diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,120120 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9998688524590165,
+  "eval_steps": 500,
+  "global_step": 17154,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00017486338797814208,
+      "grad_norm": 3.386112299979388,
+      "learning_rate": 3.883495145631068e-07,
+      "loss": 33.945,
+      "step": 1
+    },
+    {
+      "epoch": 0.00034972677595628416,
+      "grad_norm": 3.301855418072964,
+      "learning_rate": 7.766990291262136e-07,
+      "loss": 33.9354,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005245901639344263,
+      "grad_norm": 3.2872074385661554,
+      "learning_rate": 1.1650485436893206e-06,
+      "loss": 33.93,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006994535519125683,
+      "grad_norm": 3.3335351872546863,
+      "learning_rate": 1.5533980582524272e-06,
+      "loss": 33.9324,
+      "step": 4
+    },
+    {
+      "epoch": 0.0008743169398907104,
+      "grad_norm": 3.170898530779455,
+      "learning_rate": 1.941747572815534e-06,
+      "loss": 33.9394,
+      "step": 5
+    },
+    {
+      "epoch": 0.0010491803278688525,
+      "grad_norm": 3.2136639519293584,
+      "learning_rate": 2.330097087378641e-06,
+      "loss": 33.9419,
+      "step": 6
+    },
+    {
+      "epoch": 0.0012240437158469945,
+      "grad_norm": 3.3873744575021867,
+      "learning_rate": 2.718446601941748e-06,
+      "loss": 33.9203,
+      "step": 7
+    },
+    {
+      "epoch": 0.0013989071038251367,
+      "grad_norm": 3.421790349850829,
+      "learning_rate": 3.1067961165048544e-06,
+      "loss": 33.9194,
+      "step": 8
+    },
+    {
+      "epoch": 0.0015737704918032786,
+      "grad_norm": 3.2307283797832382,
+      "learning_rate": 3.4951456310679615e-06,
+      "loss": 33.9028,
+      "step": 9
+    },
+    {
+      "epoch": 0.0017486338797814208,
+      "grad_norm": 3.265863573723781,
+      "learning_rate": 3.883495145631068e-06,
+      "loss": 33.9084,
+      "step": 10
+    },
+    {
+      "epoch": 0.001923497267759563,
+      "grad_norm": 3.226620851226898,
+      "learning_rate": 4.271844660194175e-06,
+      "loss": 33.8298,
+      "step": 11
+    },
+    {
+      "epoch": 0.002098360655737705,
+      "grad_norm": 3.302267581545605,
+      "learning_rate": 4.660194174757282e-06,
+      "loss": 33.8095,
+      "step": 12
+    },
+    {
+      "epoch": 0.002273224043715847,
+      "grad_norm": 3.2136875117936348,
+      "learning_rate": 5.048543689320388e-06,
+      "loss": 33.7766,
+      "step": 13
+    },
+    {
+      "epoch": 0.002448087431693989,
+      "grad_norm": 3.308945282206079,
+      "learning_rate": 5.436893203883496e-06,
+      "loss": 33.755,
+      "step": 14
+    },
+    {
+      "epoch": 0.002622950819672131,
+      "grad_norm": 3.5294321006817015,
+      "learning_rate": 5.825242718446602e-06,
+      "loss": 33.525,
+      "step": 15
+    },
+    {
+      "epoch": 0.0027978142076502733,
+      "grad_norm": 3.4134360987755645,
+      "learning_rate": 6.213592233009709e-06,
+      "loss": 33.4709,
+      "step": 16
+    },
+    {
+      "epoch": 0.0029726775956284155,
+      "grad_norm": 3.3562989311688263,
+      "learning_rate": 6.601941747572816e-06,
+      "loss": 33.4265,
+      "step": 17
+    },
+    {
+      "epoch": 0.003147540983606557,
+      "grad_norm": 3.4528206260450807,
+      "learning_rate": 6.990291262135923e-06,
+      "loss": 33.3407,
+      "step": 18
+    },
+    {
+      "epoch": 0.0033224043715846994,
+      "grad_norm": 3.4529712126434875,
+      "learning_rate": 7.378640776699029e-06,
+      "loss": 33.2631,
+      "step": 19
+    },
+    {
+      "epoch": 0.0034972677595628415,
+      "grad_norm": 3.558122235303745,
+      "learning_rate": 7.766990291262136e-06,
+      "loss": 32.9482,
+      "step": 20
+    },
+    {
+      "epoch": 0.0036721311475409837,
+      "grad_norm": 3.418724666889899,
+      "learning_rate": 8.155339805825243e-06,
+      "loss": 32.7329,
+      "step": 21
+    },
+    {
+      "epoch": 0.003846994535519126,
+      "grad_norm": 3.338071507808524,
+      "learning_rate": 8.54368932038835e-06,
+      "loss": 32.6177,
+      "step": 22
+    },
+    {
+      "epoch": 0.004021857923497268,
+      "grad_norm": 3.1649306547042224,
+      "learning_rate": 8.932038834951456e-06,
+      "loss": 32.4122,
+      "step": 23
+    },
+    {
+      "epoch": 0.00419672131147541,
+      "grad_norm": 2.9274026063118965,
+      "learning_rate": 9.320388349514565e-06,
+      "loss": 32.2422,
+      "step": 24
+    },
+    {
+      "epoch": 0.004371584699453552,
+      "grad_norm": 2.8276164433407076,
+      "learning_rate": 9.70873786407767e-06,
+      "loss": 32.1173,
+      "step": 25
+    },
+    {
+      "epoch": 0.004546448087431694,
+      "grad_norm": 2.5951113290279118,
+      "learning_rate": 1.0097087378640776e-05,
+      "loss": 31.9519,
+      "step": 26
+    },
+    {
+      "epoch": 0.004721311475409836,
+      "grad_norm": 2.3580517871601945,
+      "learning_rate": 1.0485436893203885e-05,
+      "loss": 31.7728,
+      "step": 27
+    },
+    {
+      "epoch": 0.004896174863387978,
+      "grad_norm": 2.2116933864249795,
+      "learning_rate": 1.0873786407766991e-05,
+      "loss": 31.6051,
+      "step": 28
+    },
+    {
+      "epoch": 0.0050710382513661206,
+      "grad_norm": 2.1219056739078717,
+      "learning_rate": 1.1262135922330098e-05,
+      "loss": 31.4971,
+      "step": 29
+    },
+    {
+      "epoch": 0.005245901639344262,
+      "grad_norm": 2.0564218235534897,
+      "learning_rate": 1.1650485436893204e-05,
+      "loss": 31.4669,
+      "step": 30
+    },
+    {
+      "epoch": 0.005420765027322404,
+      "grad_norm": 1.9801950615504207,
+      "learning_rate": 1.2038834951456311e-05,
+      "loss": 31.2348,
+      "step": 31
+    },
+    {
+      "epoch": 0.005595628415300547,
+      "grad_norm": 1.8866664441781562,
+      "learning_rate": 1.2427184466019418e-05,
+      "loss": 31.1033,
+      "step": 32
+    },
+    {
+      "epoch": 0.005770491803278688,
+      "grad_norm": 1.845932748193857,
+      "learning_rate": 1.2815533980582526e-05,
+      "loss": 31.0001,
+      "step": 33
+    },
+    {
+      "epoch": 0.005945355191256831,
+      "grad_norm": 1.8017677119440227,
+      "learning_rate": 1.3203883495145633e-05,
+      "loss": 30.829,
+      "step": 34
+    },
+    {
+      "epoch": 0.006120218579234973,
+      "grad_norm": 1.7123244654110201,
+      "learning_rate": 1.3592233009708738e-05,
+      "loss": 30.7524,
+      "step": 35
+    },
+    {
+      "epoch": 0.006295081967213114,
+      "grad_norm": 1.7234106612563074,
+      "learning_rate": 1.3980582524271846e-05,
+      "loss": 30.6717,
+      "step": 36
+    },
+    {
+      "epoch": 0.006469945355191257,
+      "grad_norm": 1.6315148822019014,
+      "learning_rate": 1.4368932038834951e-05,
+      "loss": 30.5737,
+      "step": 37
+    },
+    {
+      "epoch": 0.006644808743169399,
+      "grad_norm": 1.5824498918993093,
+      "learning_rate": 1.4757281553398058e-05,
+      "loss": 30.4548,
+      "step": 38
+    },
+    {
+      "epoch": 0.006819672131147541,
+      "grad_norm": 1.5418686932322991,
+      "learning_rate": 1.5145631067961166e-05,
+      "loss": 30.3664,
+      "step": 39
+    },
+    {
+      "epoch": 0.006994535519125683,
+      "grad_norm": 1.4774275812925728,
+      "learning_rate": 1.5533980582524273e-05,
+      "loss": 30.2576,
+      "step": 40
+    },
+    {
+      "epoch": 0.007169398907103825,
+      "grad_norm": 1.5219725740013579,
+      "learning_rate": 1.5922330097087378e-05,
+      "loss": 30.1334,
+      "step": 41
+    },
+    {
+      "epoch": 0.007344262295081967,
+      "grad_norm": 1.3981678750138917,
+      "learning_rate": 1.6310679611650486e-05,
+      "loss": 30.0078,
+      "step": 42
+    },
+    {
+      "epoch": 0.007519125683060109,
+      "grad_norm": 1.347723249503582,
+      "learning_rate": 1.669902912621359e-05,
+      "loss": 29.8979,
+      "step": 43
+    },
+    {
+      "epoch": 0.007693989071038252,
+      "grad_norm": 1.3289720781490084,
+      "learning_rate": 1.70873786407767e-05,
+      "loss": 29.848,
+      "step": 44
+    },
+    {
+      "epoch": 0.007868852459016393,
+      "grad_norm": 1.2358792919316215,
+      "learning_rate": 1.7475728155339808e-05,
+      "loss": 29.7605,
+      "step": 45
+    },
+    {
+      "epoch": 0.008043715846994535,
+      "grad_norm": 1.208413150776593,
+      "learning_rate": 1.7864077669902913e-05,
+      "loss": 29.6598,
+      "step": 46
+    },
+    {
+      "epoch": 0.008218579234972677,
+      "grad_norm": 1.1624661109917669,
+      "learning_rate": 1.825242718446602e-05,
+      "loss": 29.5291,
+      "step": 47
+    },
+    {
+      "epoch": 0.00839344262295082,
+      "grad_norm": 1.1279156539134851,
+      "learning_rate": 1.864077669902913e-05,
+      "loss": 29.5274,
+      "step": 48
+    },
+    {
+      "epoch": 0.008568306010928962,
+      "grad_norm": 1.082561609802951,
+      "learning_rate": 1.9029126213592234e-05,
+      "loss": 29.4525,
+      "step": 49
+    },
+    {
+      "epoch": 0.008743169398907104,
+      "grad_norm": 1.0751477621340617,
+      "learning_rate": 1.941747572815534e-05,
+      "loss": 29.2889,
+      "step": 50
+    },
+    {
+      "epoch": 0.008918032786885246,
+      "grad_norm": 0.9981018793086815,
+      "learning_rate": 1.9805825242718447e-05,
+      "loss": 29.3316,
+      "step": 51
+    },
+    {
+      "epoch": 0.009092896174863387,
+      "grad_norm": 1.0339258702296024,
+      "learning_rate": 2.0194174757281552e-05,
+      "loss": 29.2028,
+      "step": 52
+    },
+    {
+      "epoch": 0.00926775956284153,
+      "grad_norm": 0.9616631698018512,
+      "learning_rate": 2.058252427184466e-05,
+      "loss": 29.1528,
+      "step": 53
+    },
+    {
+      "epoch": 0.009442622950819672,
+      "grad_norm": 0.9246687457061893,
+      "learning_rate": 2.097087378640777e-05,
+      "loss": 29.0442,
+      "step": 54
+    },
+    {
+      "epoch": 0.009617486338797814,
+      "grad_norm": 0.9027111521092218,
+      "learning_rate": 2.1359223300970874e-05,
+      "loss": 28.9741,
+      "step": 55
+    },
+    {
+      "epoch": 0.009792349726775956,
+      "grad_norm": 0.8859860202866106,
+      "learning_rate": 2.1747572815533982e-05,
+      "loss": 28.9628,
+      "step": 56
+    },
+    {
+      "epoch": 0.009967213114754098,
+      "grad_norm": 0.868572779385588,
+      "learning_rate": 2.213592233009709e-05,
+      "loss": 28.9051,
+      "step": 57
+    },
+    {
+      "epoch": 0.010142076502732241,
+      "grad_norm": 0.8368053100054919,
+      "learning_rate": 2.2524271844660196e-05,
+      "loss": 28.79,
+      "step": 58
+    },
+    {
+      "epoch": 0.010316939890710383,
+      "grad_norm": 0.8942429180781748,
+      "learning_rate": 2.29126213592233e-05,
+      "loss": 28.7725,
+      "step": 59
+    },
+    {
+      "epoch": 0.010491803278688525,
+      "grad_norm": 0.7957964846068506,
+      "learning_rate": 2.330097087378641e-05,
+      "loss": 28.7043,
+      "step": 60
+    },
+    {
+      "epoch": 0.010666666666666666,
+      "grad_norm": 0.829961882674078,
+      "learning_rate": 2.3689320388349514e-05,
+      "loss": 28.6705,
+      "step": 61
+    },
+    {
+      "epoch": 0.010841530054644808,
+      "grad_norm": 0.784574524335117,
+      "learning_rate": 2.4077669902912622e-05,
+      "loss": 28.6104,
+      "step": 62
+    },
+    {
+      "epoch": 0.011016393442622952,
+      "grad_norm": 0.7975808925271971,
+      "learning_rate": 2.446601941747573e-05,
+      "loss": 28.5369,
+      "step": 63
+    },
+    {
+      "epoch": 0.011191256830601093,
+      "grad_norm": 0.7448303643282296,
+      "learning_rate": 2.4854368932038836e-05,
+      "loss": 28.539,
+      "step": 64
+    },
+    {
+      "epoch": 0.011366120218579235,
+      "grad_norm": 0.7559226537590357,
+      "learning_rate": 2.5242718446601944e-05,
+      "loss": 28.5066,
+      "step": 65
+    },
+    {
+      "epoch": 0.011540983606557377,
+      "grad_norm": 0.767307177936777,
+      "learning_rate": 2.5631067961165052e-05,
+      "loss": 28.4593,
+      "step": 66
+    },
+    {
+      "epoch": 0.011715846994535518,
+      "grad_norm": 0.7295314294646044,
+      "learning_rate": 2.6019417475728157e-05,
+      "loss": 28.4519,
+      "step": 67
+    },
+    {
+      "epoch": 0.011890710382513662,
+      "grad_norm": 0.7653345671341244,
+      "learning_rate": 2.6407766990291266e-05,
+      "loss": 28.4035,
+      "step": 68
+    },
+    {
+      "epoch": 0.012065573770491804,
+      "grad_norm": 0.7159549359203016,
+      "learning_rate": 2.6796116504854367e-05,
+      "loss": 28.4183,
+      "step": 69
+    },
+    {
+      "epoch": 0.012240437158469945,
+      "grad_norm": 0.7346523236625657,
+      "learning_rate": 2.7184466019417475e-05,
+      "loss": 28.3121,
+      "step": 70
+    },
+    {
+      "epoch": 0.012415300546448087,
+      "grad_norm": 0.729876696276633,
+      "learning_rate": 2.7572815533980584e-05,
+      "loss": 28.3076,
+      "step": 71
+    },
+    {
+      "epoch": 0.012590163934426229,
+      "grad_norm": 0.6705845622058847,
+      "learning_rate": 2.7961165048543692e-05,
+      "loss": 28.3325,
+      "step": 72
+    },
+    {
+      "epoch": 0.012765027322404372,
+      "grad_norm": 0.6913278483836015,
+      "learning_rate": 2.83495145631068e-05,
+      "loss": 28.3139,
+      "step": 73
+    },
+    {
+      "epoch": 0.012939890710382514,
+      "grad_norm": 0.649379855028883,
+      "learning_rate": 2.8737864077669902e-05,
+      "loss": 28.2426,
+      "step": 74
+    },
+    {
+      "epoch": 0.013114754098360656,
+      "grad_norm": 0.6602848727865764,
+      "learning_rate": 2.912621359223301e-05,
+      "loss": 28.2219,
+      "step": 75
+    },
+    {
+      "epoch": 0.013289617486338797,
+      "grad_norm": 0.6073869401871402,
+      "learning_rate": 2.9514563106796115e-05,
+      "loss": 28.2706,
+      "step": 76
+    },
+    {
+      "epoch": 0.01346448087431694,
+      "grad_norm": 0.7162308478662635,
+      "learning_rate": 2.9902912621359224e-05,
+      "loss": 28.1173,
+      "step": 77
+    },
+    {
+      "epoch": 0.013639344262295083,
+      "grad_norm": 0.6583715235239334,
+      "learning_rate": 3.0291262135922332e-05,
+      "loss": 28.228,
+      "step": 78
+    },
+    {
+      "epoch": 0.013814207650273224,
+      "grad_norm": 0.6796022991532419,
+      "learning_rate": 3.067961165048544e-05,
+      "loss": 28.2468,
+      "step": 79
+    },
+    {
+      "epoch": 0.013989071038251366,
+      "grad_norm": 0.6290326365778236,
+      "learning_rate": 3.1067961165048545e-05,
+      "loss": 28.1854,
+      "step": 80
+    },
+    {
+      "epoch": 0.014163934426229508,
+      "grad_norm": 0.6455866614964082,
+      "learning_rate": 3.145631067961165e-05,
+      "loss": 28.1671,
+      "step": 81
+    },
+    {
+      "epoch": 0.01433879781420765,
+      "grad_norm": 0.7020625741591339,
+      "learning_rate": 3.1844660194174755e-05,
+      "loss": 28.1357,
+      "step": 82
+    },
+    {
+      "epoch": 0.014513661202185793,
+      "grad_norm": 0.6712197513610287,
+      "learning_rate": 3.2233009708737864e-05,
+      "loss": 28.1395,
+      "step": 83
+    },
+    {
+      "epoch": 0.014688524590163935,
+      "grad_norm": 0.7362056715455096,
+      "learning_rate": 3.262135922330097e-05,
+      "loss": 28.1757,
+      "step": 84
+    },
+    {
+      "epoch": 0.014863387978142076,
+      "grad_norm": 0.6696527232474568,
+      "learning_rate": 3.300970873786408e-05,
+      "loss": 28.032,
+      "step": 85
+    },
+    {
+      "epoch": 0.015038251366120218,
+      "grad_norm": 0.7688007856011254,
+      "learning_rate": 3.339805825242718e-05,
+      "loss": 28.0971,
+      "step": 86
+    },
+    {
+      "epoch": 0.01521311475409836,
+      "grad_norm": 0.65825668484645,
+      "learning_rate": 3.378640776699029e-05,
+      "loss": 28.0074,
+      "step": 87
+    },
+    {
+      "epoch": 0.015387978142076503,
+      "grad_norm": 0.8519607982356727,
+      "learning_rate": 3.41747572815534e-05,
+      "loss": 28.0693,
+      "step": 88
+    },
+    {
+      "epoch": 0.015562841530054645,
+      "grad_norm": 0.613383951658111,
+      "learning_rate": 3.456310679611651e-05,
+      "loss": 28.0525,
+      "step": 89
+    },
+    {
+      "epoch": 0.015737704918032787,
+      "grad_norm": 0.8364511908673492,
+      "learning_rate": 3.4951456310679615e-05,
+      "loss": 28.0503,
+      "step": 90
+    },
+    {
+      "epoch": 0.01591256830601093,
+      "grad_norm": 0.6174556804700497,
+      "learning_rate": 3.5339805825242724e-05,
+      "loss": 28.0559,
+      "step": 91
+    },
+    {
+      "epoch": 0.01608743169398907,
+      "grad_norm": 0.7540010552869615,
+      "learning_rate": 3.5728155339805825e-05,
+      "loss": 28.0048,
+      "step": 92
+    },
+    {
+      "epoch": 0.016262295081967214,
+      "grad_norm": 0.6921543543349493,
+      "learning_rate": 3.6116504854368933e-05,
+      "loss": 27.866,
+      "step": 93
+    },
+    {
+      "epoch": 0.016437158469945354,
+      "grad_norm": 0.6728603723803438,
+      "learning_rate": 3.650485436893204e-05,
+      "loss": 28.0057,
+      "step": 94
+    },
+    {
+      "epoch": 0.016612021857923497,
+      "grad_norm": 0.60948027178895,
+      "learning_rate": 3.689320388349515e-05,
+      "loss": 27.9646,
+      "step": 95
+    },
+    {
+      "epoch": 0.01678688524590164,
+      "grad_norm": 0.6989897672492298,
+      "learning_rate": 3.728155339805826e-05,
+      "loss": 27.9309,
+      "step": 96
+    },
+    {
+      "epoch": 0.01696174863387978,
+      "grad_norm": 0.6248044191164496,
+      "learning_rate": 3.766990291262136e-05,
+      "loss": 27.9108,
+      "step": 97
+    },
+    {
+      "epoch": 0.017136612021857924,
+      "grad_norm": 0.700318673675118,
+      "learning_rate": 3.805825242718447e-05,
+      "loss": 27.9397,
+      "step": 98
+    },
+    {
+      "epoch": 0.017311475409836064,
+      "grad_norm": 0.6824749917671127,
+      "learning_rate": 3.844660194174757e-05,
+      "loss": 27.8747,
+      "step": 99
+    },
+    {
+      "epoch": 0.017486338797814208,
+      "grad_norm": 0.7247551744277292,
+      "learning_rate": 3.883495145631068e-05,
+      "loss": 27.9164,
+      "step": 100
+    },
+    {
+      "epoch": 0.01766120218579235,
+      "grad_norm": 0.7356692420801295,
+      "learning_rate": 3.9223300970873787e-05,
+      "loss": 27.9587,
+      "step": 101
+    },
+    {
+      "epoch": 0.01783606557377049,
+      "grad_norm": 0.7292053051497158,
+      "learning_rate": 3.9611650485436895e-05,
+      "loss": 27.8391,
+      "step": 102
+    },
+    {
+      "epoch": 0.018010928961748635,
+      "grad_norm": 0.7445518321043685,
+      "learning_rate": 4e-05,
+      "loss": 27.8912,
+      "step": 103
+    },
+    {
+      "epoch": 0.018185792349726775,
+      "grad_norm": 0.6695985037568062,
+      "learning_rate": 4.0388349514563105e-05,
+      "loss": 27.8801,
+      "step": 104
+    },
+    {
+      "epoch": 0.018360655737704918,
+      "grad_norm": 0.788760787929675,
+      "learning_rate": 4.077669902912621e-05,
+      "loss": 27.8388,
+      "step": 105
+    },
+    {
+      "epoch": 0.01853551912568306,
+      "grad_norm": 0.7037278469607879,
+      "learning_rate": 4.116504854368932e-05,
+      "loss": 27.7546,
+      "step": 106
+    },
+    {
+      "epoch": 0.0187103825136612,
+      "grad_norm": 0.6601747399802127,
+      "learning_rate": 4.155339805825243e-05,
+      "loss": 27.7468,
+      "step": 107
+    },
+    {
+      "epoch": 0.018885245901639345,
+      "grad_norm": 0.626585401560701,
+      "learning_rate": 4.194174757281554e-05,
+      "loss": 27.7965,
+      "step": 108
+    },
+    {
+      "epoch": 0.019060109289617485,
+      "grad_norm": 0.7334311358388453,
+      "learning_rate": 4.2330097087378647e-05,
+      "loss": 27.7638,
+      "step": 109
+    },
+    {
+      "epoch": 0.01923497267759563,
+      "grad_norm": 0.6252407226745226,
+      "learning_rate": 4.271844660194175e-05,
+      "loss": 27.7048,
+      "step": 110
+    },
+    {
+      "epoch": 0.019409836065573772,
+      "grad_norm": 0.8099199913263514,
+      "learning_rate": 4.3106796116504856e-05,
+      "loss": 27.6691,
+      "step": 111
+    },
+    {
+      "epoch": 0.019584699453551912,
+      "grad_norm": 0.7965657440521245,
+      "learning_rate": 4.3495145631067965e-05,
+      "loss": 27.7457,
+      "step": 112
+    },
+    {
+      "epoch": 0.019759562841530055,
+      "grad_norm": 0.6596667660202103,
+      "learning_rate": 4.388349514563107e-05,
+      "loss": 27.6561,
+      "step": 113
+    },
+    {
+      "epoch": 0.019934426229508195,
+      "grad_norm": 0.6460754899538416,
+      "learning_rate": 4.427184466019418e-05,
+      "loss": 27.7172,
+      "step": 114
+    },
+    {
+      "epoch": 0.02010928961748634,
+      "grad_norm": 0.7850815140180447,
+      "learning_rate": 4.466019417475728e-05,
+      "loss": 27.796,
+      "step": 115
+    },
+    {
+      "epoch": 0.020284153005464482,
+      "grad_norm": 0.848009353890878,
+      "learning_rate": 4.504854368932039e-05,
+      "loss": 27.6449,
+      "step": 116
+    },
+    {
+      "epoch": 0.020459016393442622,
+      "grad_norm": 0.7247381055781121,
+      "learning_rate": 4.543689320388349e-05,
+      "loss": 27.6212,
+      "step": 117
+    },
+    {
+      "epoch": 0.020633879781420766,
+      "grad_norm": 1.0758631377482637,
+      "learning_rate": 4.58252427184466e-05,
+      "loss": 27.639,
+      "step": 118
+    },
+    {
+      "epoch": 0.020808743169398906,
+      "grad_norm": 0.97724592852275,
+      "learning_rate": 4.621359223300971e-05,
+      "loss": 27.6032,
+      "step": 119
+    },
+    {
+      "epoch": 0.02098360655737705,
+      "grad_norm": 0.7219194176836166,
+      "learning_rate": 4.660194174757282e-05,
+      "loss": 27.601,
+      "step": 120
+    },
+    {
+      "epoch": 0.021158469945355193,
+      "grad_norm": 1.2884836206823118,
+      "learning_rate": 4.6990291262135926e-05,
+      "loss": 27.7102,
+      "step": 121
+    },
+    {
+      "epoch": 0.021333333333333333,
+      "grad_norm": 1.340463557968345,
+      "learning_rate": 4.737864077669903e-05,
+      "loss": 27.6488,
+      "step": 122
+    },
+    {
+      "epoch": 0.021508196721311476,
+      "grad_norm": 0.7871000797954846,
+      "learning_rate": 4.7766990291262136e-05,
+      "loss": 27.6339,
+      "step": 123
+    },
+    {
+      "epoch": 0.021683060109289616,
+      "grad_norm": 0.7703341139307888,
+      "learning_rate": 4.8155339805825245e-05,
+      "loss": 27.5887,
+      "step": 124
+    },
+    {
+      "epoch": 0.02185792349726776,
+      "grad_norm": 0.8128180229402846,
+      "learning_rate": 4.854368932038835e-05,
+      "loss": 27.4695,
+      "step": 125
+    },
+    {
+      "epoch": 0.022032786885245903,
+      "grad_norm": 0.6666641138328544,
+      "learning_rate": 4.893203883495146e-05,
+      "loss": 27.51,
+      "step": 126
+    },
+    {
+      "epoch": 0.022207650273224043,
+      "grad_norm": 0.7264977654470559,
+      "learning_rate": 4.932038834951457e-05,
+      "loss": 27.5647,
+      "step": 127
+    },
+    {
+      "epoch": 0.022382513661202186,
+      "grad_norm": 0.6880544060740871,
+      "learning_rate": 4.970873786407767e-05,
+      "loss": 27.4677,
+      "step": 128
+    },
+    {
+      "epoch": 0.022557377049180326,
+      "grad_norm": 0.733993838092126,
+      "learning_rate": 5.0097087378640786e-05,
+      "loss": 27.4663,
+      "step": 129
+    },
+    {
+      "epoch": 0.02273224043715847,
+      "grad_norm": 1.0454600569712833,
+      "learning_rate": 5.048543689320389e-05,
+      "loss": 27.4795,
+      "step": 130
+    },
+    {
+      "epoch": 0.022907103825136613,
+      "grad_norm": 1.2373375677116867,
+      "learning_rate": 5.087378640776699e-05,
+      "loss": 27.5211,
+      "step": 131
+    },
+    {
+      "epoch": 0.023081967213114753,
+      "grad_norm": 0.8959994448163705,
+      "learning_rate": 5.1262135922330105e-05,
+      "loss": 27.4599,
+      "step": 132
+    },
+    {
+      "epoch": 0.023256830601092897,
+      "grad_norm": 1.050463943907856,
+      "learning_rate": 5.1650485436893206e-05,
+      "loss": 27.4536,
+      "step": 133
+    },
+    {
+      "epoch": 0.023431693989071037,
+      "grad_norm": 2.3510938102532264,
+      "learning_rate": 5.2038834951456314e-05,
+      "loss": 27.3921,
+      "step": 134
+    },
+    {
+      "epoch": 0.02360655737704918,
+      "grad_norm": 3.2167249741768886,
+      "learning_rate": 5.2427184466019416e-05,
+      "loss": 27.5069,
+      "step": 135
+    },
+    {
+      "epoch": 0.023781420765027324,
+      "grad_norm": 2.31182195865069,
+      "learning_rate": 5.281553398058253e-05,
+      "loss": 27.3039,
+      "step": 136
+    },
+    {
+      "epoch": 0.023956284153005464,
+      "grad_norm": 0.8982950806529117,
+      "learning_rate": 5.320388349514563e-05,
+      "loss": 27.3785,
+      "step": 137
+    },
+    {
+      "epoch": 0.024131147540983607,
+      "grad_norm": 2.685085903204742,
+      "learning_rate": 5.3592233009708734e-05,
+      "loss": 27.4435,
+      "step": 138
+    },
+    {
+      "epoch": 0.024306010928961747,
+      "grad_norm": 2.613951333711937,
+      "learning_rate": 5.398058252427185e-05,
+      "loss": 27.3727,
+      "step": 139
+    },
+    {
+      "epoch": 0.02448087431693989,
+      "grad_norm": 0.8910261234086992,
+      "learning_rate": 5.436893203883495e-05,
+      "loss": 27.4373,
+      "step": 140
+    },
+    {
+      "epoch": 0.024655737704918034,
+      "grad_norm": 2.012499354156556,
+      "learning_rate": 5.4757281553398066e-05,
+      "loss": 27.3795,
+      "step": 141
+    },
+    {
+      "epoch": 0.024830601092896174,
+      "grad_norm": 1.077486282763958,
+      "learning_rate": 5.514563106796117e-05,
+      "loss": 27.2662,
+      "step": 142
+    },
+    {
+      "epoch": 0.025005464480874318,
+      "grad_norm": 1.6446095378737147,
+      "learning_rate": 5.553398058252427e-05,
+      "loss": 27.2508,
+      "step": 143
+    },
+    {
+      "epoch": 0.025180327868852458,
+      "grad_norm": 1.4713187392836555,
+      "learning_rate": 5.5922330097087384e-05,
+      "loss": 27.1625,
+      "step": 144
+    },
+    {
+      "epoch": 0.0253551912568306,
+      "grad_norm": 0.9353233498505846,
+      "learning_rate": 5.6310679611650486e-05,
+      "loss": 27.2339,
+      "step": 145
+    },
+    {
+      "epoch": 0.025530054644808745,
+      "grad_norm": 1.3379180304113865,
+      "learning_rate": 5.66990291262136e-05,
+      "loss": 27.2142,
+      "step": 146
+    },
+    {
+      "epoch": 0.025704918032786885,
+      "grad_norm": 0.8249582424182282,
+      "learning_rate": 5.70873786407767e-05,
+      "loss": 27.1541,
+      "step": 147
+    },
+    {
+      "epoch": 0.025879781420765028,
+      "grad_norm": 1.1413146679112618,
+      "learning_rate": 5.7475728155339804e-05,
+      "loss": 27.2608,
+      "step": 148
+    },
+    {
+      "epoch": 0.026054644808743168,
+      "grad_norm": 0.7450374490730223,
+      "learning_rate": 5.786407766990292e-05,
+      "loss": 27.1943,
+      "step": 149
+    },
+    {
+      "epoch": 0.02622950819672131,
+      "grad_norm": 1.7422405665297696,
+      "learning_rate": 5.825242718446602e-05,
+      "loss": 27.1963,
+      "step": 150
+    },
+    {
+      "epoch": 0.026404371584699455,
+      "grad_norm": 1.2109541629433265,
+      "learning_rate": 5.864077669902913e-05,
+      "loss": 27.1779,
+      "step": 151
+    },
+    {
+      "epoch": 0.026579234972677595,
+      "grad_norm": 1.4045130752262627,
+      "learning_rate": 5.902912621359223e-05,
+      "loss": 27.1706,
+      "step": 152
+    },
+    {
+      "epoch": 0.02675409836065574,
+      "grad_norm": 1.5145977812564573,
+      "learning_rate": 5.9417475728155346e-05,
+      "loss": 27.1327,
+      "step": 153
+    },
+    {
+      "epoch": 0.02692896174863388,
+      "grad_norm": 1.1345053191014491,
+      "learning_rate": 5.980582524271845e-05,
+      "loss": 27.0112,
+      "step": 154
+    },
+    {
+      "epoch": 0.027103825136612022,
+      "grad_norm": 1.1576264163957626,
+      "learning_rate": 6.019417475728155e-05,
+      "loss": 27.0981,
+      "step": 155
+    },
+    {
+      "epoch": 0.027278688524590165,
+      "grad_norm": 1.5140159291913213,
+      "learning_rate": 6.0582524271844664e-05,
+      "loss": 27.1185,
+      "step": 156
+    },
+    {
+      "epoch": 0.027453551912568305,
+      "grad_norm": 1.1352691472838692,
+      "learning_rate": 6.0970873786407766e-05,
+      "loss": 27.097,
+      "step": 157
+    },
+    {
+      "epoch": 0.02762841530054645,
+      "grad_norm": 1.6549512762560716,
+      "learning_rate": 6.135922330097087e-05,
+      "loss": 27.0199,
+      "step": 158
+    },
+    {
+      "epoch": 0.02780327868852459,
+      "grad_norm": 1.8736652301262886,
+      "learning_rate": 6.174757281553398e-05,
+      "loss": 27.0112,
+      "step": 159
+    },
+    {
+      "epoch": 0.027978142076502732,
+      "grad_norm": 1.0656686568152762,
+      "learning_rate": 6.213592233009709e-05,
+      "loss": 27.019,
+      "step": 160
+    },
+    {
+      "epoch": 0.028153005464480876,
+      "grad_norm": 1.6364603916333416,
+      "learning_rate": 6.25242718446602e-05,
+      "loss": 27.0707,
+      "step": 161
+    },
+    {
+      "epoch": 0.028327868852459016,
+      "grad_norm": 1.0598105531006505,
+      "learning_rate": 6.29126213592233e-05,
+      "loss": 26.7652,
+      "step": 162
+    },
+    {
+      "epoch": 0.02850273224043716,
+      "grad_norm": 1.8013082307822805,
+      "learning_rate": 6.330097087378642e-05,
+      "loss": 26.7981,
+      "step": 163
+    },
+    {
+      "epoch": 0.0286775956284153,
+      "grad_norm": 1.4743372931868821,
+      "learning_rate": 6.368932038834951e-05,
+      "loss": 27.0121,
+      "step": 164
+    },
+    {
+      "epoch": 0.028852459016393443,
+      "grad_norm": 1.00474627222483,
+      "learning_rate": 6.407766990291263e-05,
+      "loss": 26.9741,
+      "step": 165
+    },
+    {
+      "epoch": 0.029027322404371586,
+      "grad_norm": 1.8226602588624101,
+      "learning_rate": 6.446601941747573e-05,
+      "loss": 26.925,
+      "step": 166
+    },
+    {
+      "epoch": 0.029202185792349726,
+      "grad_norm": 1.5666757583039996,
+      "learning_rate": 6.485436893203884e-05,
+      "loss": 26.775,
+      "step": 167
+    },
+    {
+      "epoch": 0.02937704918032787,
+      "grad_norm": 2.210776706685403,
+      "learning_rate": 6.524271844660194e-05,
+      "loss": 26.8147,
+      "step": 168
+    },
+    {
+      "epoch": 0.02955191256830601,
+      "grad_norm": 1.4478872377463647,
+      "learning_rate": 6.563106796116505e-05,
+      "loss": 26.7301,
+      "step": 169
+    },
+    {
+      "epoch": 0.029726775956284153,
+      "grad_norm": 1.7571194210688013,
+      "learning_rate": 6.601941747572816e-05,
+      "loss": 26.7594,
+      "step": 170
+    },
+    {
+      "epoch": 0.029901639344262296,
+      "grad_norm": 1.4994547217824454,
+      "learning_rate": 6.640776699029127e-05,
+      "loss": 26.8093,
+      "step": 171
+    },
+    {
+      "epoch": 0.030076502732240436,
+      "grad_norm": 1.73347989885927,
+      "learning_rate": 6.679611650485436e-05,
+      "loss": 26.756,
+      "step": 172
+    },
+    {
+      "epoch": 0.03025136612021858,
+      "grad_norm": 2.3568512776299104,
+      "learning_rate": 6.718446601941749e-05,
+      "loss": 26.8174,
+      "step": 173
+    },
+    {
+      "epoch": 0.03042622950819672,
+      "grad_norm": 1.367904823999101,
+      "learning_rate": 6.757281553398058e-05,
+      "loss": 26.5728,
+      "step": 174
+    },
+    {
+      "epoch": 0.030601092896174863,
+      "grad_norm": 1.887370202974666,
+      "learning_rate": 6.79611650485437e-05,
+      "loss": 26.6853,
+      "step": 175
+    },
+    {
+      "epoch": 0.030775956284153007,
+      "grad_norm": 1.18320666794206,
+      "learning_rate": 6.83495145631068e-05,
+      "loss": 26.6554,
+      "step": 176
+    },
+    {
+      "epoch": 0.030950819672131147,
+      "grad_norm": 1.770182479868567,
+      "learning_rate": 6.87378640776699e-05,
+      "loss": 26.5512,
+      "step": 177
+    },
+    {
+      "epoch": 0.03112568306010929,
+      "grad_norm": 2.7238126677114067,
+      "learning_rate": 6.912621359223301e-05,
+      "loss": 26.754,
+      "step": 178
+    },
+    {
+      "epoch": 0.031300546448087434,
+      "grad_norm": 3.4712430498993343,
+      "learning_rate": 6.951456310679612e-05,
+      "loss": 26.6662,
+      "step": 179
+    },
+    {
+      "epoch": 0.031475409836065574,
+      "grad_norm": 2.5475651731477913,
+      "learning_rate": 6.990291262135923e-05,
+      "loss": 26.5639,
+      "step": 180
+    },
+    {
+      "epoch": 0.031650273224043714,
+      "grad_norm": 1.1262431634300252,
+      "learning_rate": 7.029126213592233e-05,
+      "loss": 26.5284,
+      "step": 181
+    },
+    {
+      "epoch": 0.03182513661202186,
+      "grad_norm": 1.6808429494186945,
+      "learning_rate": 7.067961165048545e-05,
+      "loss": 26.6376,
+      "step": 182
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 1.2342389178456556,
+      "learning_rate": 7.106796116504854e-05,
+      "loss": 26.4409,
+      "step": 183
+    },
+    {
+      "epoch": 0.03217486338797814,
+      "grad_norm": 1.5314941838308087,
+      "learning_rate": 7.145631067961165e-05,
+      "loss": 26.5482,
+      "step": 184
+    },
+    {
+      "epoch": 0.03234972677595629,
+      "grad_norm": 1.743817006124641,
+      "learning_rate": 7.184466019417476e-05,
+      "loss": 26.4414,
+      "step": 185
+    },
+    {
+      "epoch": 0.03252459016393443,
+      "grad_norm": 1.3487327718258968,
+      "learning_rate": 7.223300970873787e-05,
+      "loss": 26.3532,
+      "step": 186
+    },
+    {
+      "epoch": 0.03269945355191257,
+      "grad_norm": 2.047847746817484,
+      "learning_rate": 7.262135922330098e-05,
+      "loss": 26.3154,
+      "step": 187
+    },
+    {
+      "epoch": 0.03287431693989071,
+      "grad_norm": 2.214710987964511,
+      "learning_rate": 7.300970873786408e-05,
+      "loss": 26.4252,
+      "step": 188
+    },
+    {
+      "epoch": 0.033049180327868855,
+      "grad_norm": 1.4348332642275505,
+      "learning_rate": 7.339805825242719e-05,
+      "loss": 26.2203,
+      "step": 189
+    },
+    {
+      "epoch": 0.033224043715846995,
+      "grad_norm": 2.6841581228614655,
+      "learning_rate": 7.37864077669903e-05,
+      "loss": 26.1603,
+      "step": 190
+    },
+    {
+      "epoch": 0.033398907103825135,
+      "grad_norm": 2.224650698910777,
+      "learning_rate": 7.41747572815534e-05,
+      "loss": 26.3499,
+      "step": 191
+    },
+    {
+      "epoch": 0.03357377049180328,
+      "grad_norm": 1.8117890632348195,
+      "learning_rate": 7.456310679611652e-05,
+      "loss": 26.2694,
+      "step": 192
+    },
+    {
+      "epoch": 0.03374863387978142,
+      "grad_norm": 1.5896707434736081,
+      "learning_rate": 7.495145631067961e-05,
+      "loss": 26.1269,
+      "step": 193
+    },
+    {
+      "epoch": 0.03392349726775956,
+      "grad_norm": 2.142528103626558,
+      "learning_rate": 7.533980582524272e-05,
+      "loss": 25.9901,
+      "step": 194
+    },
+    {
+      "epoch": 0.03409836065573771,
+      "grad_norm": 1.1518632514157712,
+      "learning_rate": 7.572815533980583e-05,
+      "loss": 26.1014,
+      "step": 195
+    },
+    {
+      "epoch": 0.03427322404371585,
+      "grad_norm": 1.916088839574567,
+      "learning_rate": 7.611650485436894e-05,
+      "loss": 25.9957,
+      "step": 196
+    },
+    {
+      "epoch": 0.03444808743169399,
+      "grad_norm": 1.410255808255757,
+      "learning_rate": 7.650485436893205e-05,
+      "loss": 26.128,
+      "step": 197
+    },
+    {
+      "epoch": 0.03462295081967213,
+      "grad_norm": 1.4333131389324751,
+      "learning_rate": 7.689320388349514e-05,
+      "loss": 26.1555,
+      "step": 198
+    },
+    {
+      "epoch": 0.034797814207650275,
+      "grad_norm": 1.4117787582905545,
+      "learning_rate": 7.728155339805826e-05,
+      "loss": 26.0967,
+      "step": 199
+    },
+    {
+      "epoch": 0.034972677595628415,
+      "grad_norm": 1.2414997316218406,
+      "learning_rate": 7.766990291262136e-05,
+      "loss": 25.9518,
+      "step": 200
+    },
+    {
+      "epoch": 0.035147540983606555,
+      "grad_norm": 1.6426669005684051,
+      "learning_rate": 7.805825242718448e-05,
+      "loss": 26.0134,
+      "step": 201
+    },
+    {
+      "epoch": 0.0353224043715847,
+      "grad_norm": 1.083842960494342,
+      "learning_rate": 7.844660194174757e-05,
+      "loss": 25.9456,
+      "step": 202
+    },
+    {
+      "epoch": 0.03549726775956284,
+      "grad_norm": 1.5256402222487864,
+      "learning_rate": 7.883495145631068e-05,
+      "loss": 25.8323,
+      "step": 203
+    },
+    {
+      "epoch": 0.03567213114754098,
+      "grad_norm": 1.5954831473009823,
+      "learning_rate": 7.922330097087379e-05,
+      "loss": 25.7245,
+      "step": 204
+    },
+    {
+      "epoch": 0.03584699453551913,
+      "grad_norm": 2.8978505804300023,
+      "learning_rate": 7.96116504854369e-05,
+      "loss": 25.7811,
+      "step": 205
+    },
+    {
+      "epoch": 0.03602185792349727,
+      "grad_norm": 5.457061757586217,
+      "learning_rate": 8e-05,
+      "loss": 25.8036,
+      "step": 206
+    },
+    {
+      "epoch": 0.03619672131147541,
+      "grad_norm": 4.915919000065065,
+      "learning_rate": 8.038834951456311e-05,
+      "loss": 26.0879,
+      "step": 207
+    },
+    {
+      "epoch": 0.03637158469945355,
+      "grad_norm": 1.7473368166946874,
+      "learning_rate": 8.077669902912621e-05,
+      "loss": 25.7102,
+      "step": 208
+    },
+    {
+      "epoch": 0.036546448087431696,
+      "grad_norm": 3.0606921539790872,
+      "learning_rate": 8.116504854368933e-05,
+      "loss": 26.02,
+      "step": 209
+    },
+    {
+      "epoch": 0.036721311475409836,
+      "grad_norm": 1.6028700063838963,
+      "learning_rate": 8.155339805825243e-05,
+      "loss": 25.8541,
+      "step": 210
+    },
+    {
+      "epoch": 0.036896174863387976,
+      "grad_norm": 1.787150028281631,
+      "learning_rate": 8.194174757281555e-05,
+      "loss": 25.7878,
+      "step": 211
+    },
+    {
+      "epoch": 0.03707103825136612,
+      "grad_norm": 1.8630917026164302,
+      "learning_rate": 8.233009708737864e-05,
+      "loss": 25.5079,
+      "step": 212
+    },
+    {
+      "epoch": 0.03724590163934426,
+      "grad_norm": 1.7345924472784835,
+      "learning_rate": 8.271844660194175e-05,
+      "loss": 25.8212,
+      "step": 213
+    },
+    {
+      "epoch": 0.0374207650273224,
+      "grad_norm": 1.4272958596984497,
+      "learning_rate": 8.310679611650486e-05,
+      "loss": 25.7483,
+      "step": 214
+    },
+    {
+      "epoch": 0.03759562841530055,
+      "grad_norm": 1.5438071688683501,
+      "learning_rate": 8.349514563106797e-05,
+      "loss": 25.4727,
+      "step": 215
+    },
+    {
+      "epoch": 0.03777049180327869,
+      "grad_norm": 1.4220295417049966,
+      "learning_rate": 8.388349514563108e-05,
+      "loss": 25.6307,
+      "step": 216
+    },
+    {
+      "epoch": 0.03794535519125683,
+      "grad_norm": 1.149225801579692,
+      "learning_rate": 8.427184466019417e-05,
+      "loss": 25.7143,
+      "step": 217
+    },
+    {
+      "epoch": 0.03812021857923497,
+      "grad_norm": 1.1938788106660905,
+      "learning_rate": 8.466019417475729e-05,
+      "loss": 25.6103,
+      "step": 218
+    },
+    {
+      "epoch": 0.03829508196721312,
+      "grad_norm": 1.328839568652738,
+      "learning_rate": 8.504854368932039e-05,
+      "loss": 25.2909,
+      "step": 219
+    },
+    {
+      "epoch": 0.03846994535519126,
+      "grad_norm": 1.1852681067069615,
+      "learning_rate": 8.54368932038835e-05,
+      "loss": 25.5815,
+      "step": 220
+    },
+    {
+      "epoch": 0.0386448087431694,
+      "grad_norm": 1.1536310788990107,
+      "learning_rate": 8.58252427184466e-05,
+      "loss": 25.3194,
+      "step": 221
+    },
+    {
+      "epoch": 0.038819672131147544,
+      "grad_norm": 1.2907167887274764,
+      "learning_rate": 8.621359223300971e-05,
+      "loss": 25.2006,
+      "step": 222
+    },
+    {
+      "epoch": 0.038994535519125684,
+      "grad_norm": 1.3111961598844903,
+      "learning_rate": 8.660194174757282e-05,
+      "loss": 25.3642,
+      "step": 223
+    },
+    {
+      "epoch": 0.039169398907103824,
+      "grad_norm": 1.0000850646336463,
+      "learning_rate": 8.699029126213593e-05,
+      "loss": 25.299,
+      "step": 224
+    },
+    {
+      "epoch": 0.03934426229508197,
+      "grad_norm": 1.543339081392779,
+      "learning_rate": 8.737864077669902e-05,
+      "loss": 25.17,
+      "step": 225
+    },
+    {
+      "epoch": 0.03951912568306011,
+      "grad_norm": 1.319169149286536,
+      "learning_rate": 8.776699029126215e-05,
+      "loss": 25.4809,
+      "step": 226
+    },
+    {
+      "epoch": 0.03969398907103825,
+      "grad_norm": 1.1202327632648483,
+      "learning_rate": 8.815533980582524e-05,
+      "loss": 25.0522,
+      "step": 227
+    },
+    {
+      "epoch": 0.03986885245901639,
+      "grad_norm": 1.3359650110951526,
+      "learning_rate": 8.854368932038836e-05,
+      "loss": 25.3521,
+      "step": 228
+    },
+    {
+      "epoch": 0.04004371584699454,
+      "grad_norm": 1.0859436410535628,
+      "learning_rate": 8.893203883495146e-05,
+      "loss": 24.9834,
+      "step": 229
+    },
+    {
+      "epoch": 0.04021857923497268,
+      "grad_norm": 1.2728100963641589,
+      "learning_rate": 8.932038834951457e-05,
+      "loss": 25.363,
+      "step": 230
+    },
+    {
+      "epoch": 0.04039344262295082,
+      "grad_norm": 1.5821484688265286,
+      "learning_rate": 8.970873786407767e-05,
+      "loss": 25.2992,
+      "step": 231
+    },
+    {
+      "epoch": 0.040568306010928964,
+      "grad_norm": 1.0044335769763397,
+      "learning_rate": 9.009708737864078e-05,
+      "loss": 25.1629,
+      "step": 232
+    },
+    {
+      "epoch": 0.040743169398907104,
+      "grad_norm": 1.7077724064648756,
+      "learning_rate": 9.048543689320389e-05,
+      "loss": 25.2518,
+      "step": 233
+    },
+    {
+      "epoch": 0.040918032786885244,
+      "grad_norm": 2.0537553622809877,
+      "learning_rate": 9.087378640776699e-05,
+      "loss": 25.2085,
+      "step": 234
+    },
+    {
+      "epoch": 0.04109289617486339,
+      "grad_norm": 1.517308432026352,
+      "learning_rate": 9.126213592233011e-05,
+      "loss": 25.2308,
+      "step": 235
+    },
+    {
+      "epoch": 0.04126775956284153,
+      "grad_norm": 1.4839258049130037,
+      "learning_rate": 9.16504854368932e-05,
+      "loss": 25.133,
+      "step": 236
+    },
+    {
+      "epoch": 0.04144262295081967,
+      "grad_norm": 1.8346806092665455,
+      "learning_rate": 9.203883495145631e-05,
+      "loss": 25.0673,
+      "step": 237
+    },
+    {
+      "epoch": 0.04161748633879781,
+      "grad_norm": 1.350100593159644,
+      "learning_rate": 9.242718446601942e-05,
+      "loss": 24.9688,
+      "step": 238
+    },
+    {
+      "epoch": 0.04179234972677596,
+      "grad_norm": 1.407191992227566,
+      "learning_rate": 9.281553398058253e-05,
+      "loss": 25.0104,
+      "step": 239
+    },
+    {
+      "epoch": 0.0419672131147541,
+      "grad_norm": 1.3786348537205404,
+      "learning_rate": 9.320388349514564e-05,
+      "loss": 25.1293,
+      "step": 240
+    },
+    {
+      "epoch": 0.04214207650273224,
+      "grad_norm": 1.404813154352224,
+      "learning_rate": 9.359223300970874e-05,
+      "loss": 25.0661,
+      "step": 241
+    },
+    {
+      "epoch": 0.042316939890710385,
+      "grad_norm": 1.2738654592961063,
+      "learning_rate": 9.398058252427185e-05,
+      "loss": 25.0409,
+      "step": 242
+    },
+    {
+      "epoch": 0.042491803278688525,
+      "grad_norm": 1.472868094396671,
+      "learning_rate": 9.436893203883496e-05,
+      "loss": 24.8169,
+      "step": 243
+    },
+    {
+      "epoch": 0.042666666666666665,
+      "grad_norm": 1.3318284402325256,
+      "learning_rate": 9.475728155339806e-05,
+      "loss": 25.0467,
+      "step": 244
+    },
+    {
+      "epoch": 0.04284153005464481,
+      "grad_norm": 1.1245143492653764,
+      "learning_rate": 9.514563106796118e-05,
+      "loss": 24.905,
+      "step": 245
+    },
+    {
+      "epoch": 0.04301639344262295,
+      "grad_norm": 1.264426704365736,
+      "learning_rate": 9.553398058252427e-05,
+      "loss": 24.9962,
+      "step": 246
+    },
+    {
+      "epoch": 0.04319125683060109,
+      "grad_norm": 1.4159985648722833,
+      "learning_rate": 9.592233009708738e-05,
+      "loss": 24.9875,
+      "step": 247
+    },
+    {
+      "epoch": 0.04336612021857923,
+      "grad_norm": 1.422588737011595,
+      "learning_rate": 9.631067961165049e-05,
+      "loss": 24.8068,
+      "step": 248
+    },
+    {
+      "epoch": 0.04354098360655738,
+      "grad_norm": 1.2570641610477307,
+      "learning_rate": 9.66990291262136e-05,
+      "loss": 24.8871,
+      "step": 249
+    },
+    {
+      "epoch": 0.04371584699453552,
+      "grad_norm": 1.4335304498476538,
+      "learning_rate": 9.70873786407767e-05,
+      "loss": 24.9094,
+      "step": 250
+    },
+    {
+      "epoch": 0.04389071038251366,
+      "grad_norm": 1.5819403991590997,
+      "learning_rate": 9.74757281553398e-05,
+      "loss": 24.9108,
+      "step": 251
+    },
+    {
+      "epoch": 0.044065573770491806,
+      "grad_norm": 1.490726930021544,
+      "learning_rate": 9.786407766990292e-05,
+      "loss": 24.6819,
+      "step": 252
+    },
+    {
+      "epoch": 0.044240437158469946,
+      "grad_norm": 1.3086712959575149,
+      "learning_rate": 9.825242718446602e-05,
+      "loss": 24.791,
+      "step": 253
+    },
+    {
+      "epoch": 0.044415300546448086,
+      "grad_norm": 1.5330342722354908,
+      "learning_rate": 9.864077669902914e-05,
+      "loss": 24.8677,
+      "step": 254
+    },
+    {
+      "epoch": 0.04459016393442623,
+      "grad_norm": 1.2860179199875001,
+      "learning_rate": 9.902912621359223e-05,
+      "loss": 24.5896,
+      "step": 255
+    },
+    {
+      "epoch": 0.04476502732240437,
+      "grad_norm": 1.6103186896406079,
+      "learning_rate": 9.941747572815534e-05,
+      "loss": 24.6407,
+      "step": 256
+    },
+    {
+      "epoch": 0.04493989071038251,
+      "grad_norm": 1.6945179998444215,
+      "learning_rate": 9.980582524271845e-05,
+      "loss": 24.8803,
+      "step": 257
+    },
+    {
+      "epoch": 0.04511475409836065,
+      "grad_norm": 1.586404031109761,
+      "learning_rate": 0.00010019417475728157,
+      "loss": 24.6385,
+      "step": 258
+    },
+    {
+      "epoch": 0.0452896174863388,
+      "grad_norm": 1.3124249440571,
+      "learning_rate": 0.00010058252427184467,
+      "loss": 24.4425,
+      "step": 259
+    },
+    {
+      "epoch": 0.04546448087431694,
+      "grad_norm": 1.227310735201249,
+      "learning_rate": 0.00010097087378640778,
+      "loss": 24.7231,
+      "step": 260
+    },
+    {
+      "epoch": 0.04563934426229508,
+      "grad_norm": 1.553504150261481,
+      "learning_rate": 0.00010135922330097087,
+      "loss": 24.4844,
+      "step": 261
+    },
+    {
+      "epoch": 0.04581420765027323,
+      "grad_norm": 1.3791254141686424,
+      "learning_rate": 0.00010174757281553398,
+      "loss": 24.8046,
+      "step": 262
+    },
+    {
+      "epoch": 0.04598907103825137,
+      "grad_norm": 1.7831245624514926,
+      "learning_rate": 0.0001021359223300971,
+      "loss": 24.3947,
+      "step": 263
+    },
+    {
+      "epoch": 0.04616393442622951,
+      "grad_norm": 2.0182101639619616,
+      "learning_rate": 0.00010252427184466021,
+      "loss": 24.4338,
+      "step": 264
+    },
+    {
+      "epoch": 0.046338797814207654,
+      "grad_norm": 1.9889745311243676,
+      "learning_rate": 0.0001029126213592233,
+      "loss": 24.5828,
+      "step": 265
+    },
+    {
+      "epoch": 0.046513661202185794,
+      "grad_norm": 1.5251114969932746,
+      "learning_rate": 0.00010330097087378641,
+      "loss": 24.1939,
+      "step": 266
+    },
+    {
+      "epoch": 0.046688524590163934,
+      "grad_norm": 2.205152074781535,
+      "learning_rate": 0.00010368932038834951,
+      "loss": 24.6243,
+      "step": 267
+    },
+    {
+      "epoch": 0.046863387978142074,
+      "grad_norm": 2.2676024997080177,
+      "learning_rate": 0.00010407766990291263,
+      "loss": 24.4939,
+      "step": 268
+    },
+    {
+      "epoch": 0.04703825136612022,
+      "grad_norm": 1.5671997774755555,
+      "learning_rate": 0.00010446601941747574,
+      "loss": 24.4961,
+      "step": 269
+    },
+    {
+      "epoch": 0.04721311475409836,
+      "grad_norm": 1.7298761544579566,
+      "learning_rate": 0.00010485436893203883,
+      "loss": 24.482,
+      "step": 270
+    },
+    {
+      "epoch": 0.0473879781420765,
+      "grad_norm": 2.2444706271457577,
+      "learning_rate": 0.00010524271844660194,
+      "loss": 24.586,
+      "step": 271
+    },
+    {
+      "epoch": 0.04756284153005465,
+      "grad_norm": 2.4874584558205446,
+      "learning_rate": 0.00010563106796116506,
+      "loss": 24.5723,
+      "step": 272
+    },
+    {
+      "epoch": 0.04773770491803279,
+      "grad_norm": 2.4311872052067733,
+      "learning_rate": 0.00010601941747572817,
+      "loss": 24.5651,
+      "step": 273
+    },
+    {
+      "epoch": 0.04791256830601093,
+      "grad_norm": 2.3231044976104283,
+      "learning_rate": 0.00010640776699029127,
+      "loss": 24.4892,
+      "step": 274
+    },
+    {
+      "epoch": 0.048087431693989074,
+      "grad_norm": 1.9697495191418115,
+      "learning_rate": 0.00010679611650485437,
+      "loss": 24.4501,
+      "step": 275
+    },
+    {
+      "epoch": 0.048262295081967214,
+      "grad_norm": 1.5108678787740186,
+      "learning_rate": 0.00010718446601941747,
+      "loss": 24.4863,
+      "step": 276
+    },
+    {
+      "epoch": 0.048437158469945354,
+      "grad_norm": 1.7175924073729367,
+      "learning_rate": 0.0001075728155339806,
+      "loss": 24.434,
+      "step": 277
+    },
+    {
+      "epoch": 0.048612021857923494,
+      "grad_norm": 1.4839769316594682,
+      "learning_rate": 0.0001079611650485437,
+      "loss": 24.5242,
+      "step": 278
+    },
+    {
+      "epoch": 0.04878688524590164,
+      "grad_norm": 1.7459338618533526,
+      "learning_rate": 0.00010834951456310681,
+      "loss": 24.3051,
+      "step": 279
+    },
+    {
+      "epoch": 0.04896174863387978,
+      "grad_norm": 1.9494681979117896,
+      "learning_rate": 0.0001087378640776699,
+      "loss": 24.1319,
+      "step": 280
+    },
+    {
+      "epoch": 0.04913661202185792,
+      "grad_norm": 1.7296677613817741,
+      "learning_rate": 0.00010912621359223301,
+      "loss": 23.981,
+      "step": 281
+    },
+    {
+      "epoch": 0.04931147540983607,
+      "grad_norm": 2.2781745136042773,
+      "learning_rate": 0.00010951456310679613,
+      "loss": 24.2243,
+      "step": 282
+    },
+    {
+      "epoch": 0.04948633879781421,
+      "grad_norm": 3.0584169767259906,
+      "learning_rate": 0.00010990291262135923,
+      "loss": 24.2669,
+      "step": 283
+    },
+    {
+      "epoch": 0.04966120218579235,
+      "grad_norm": 3.2198004903540145,
+      "learning_rate": 0.00011029126213592234,
+      "loss": 24.4346,
+      "step": 284
+    },
+    {
+      "epoch": 0.049836065573770495,
+      "grad_norm": 2.659733559049539,
+      "learning_rate": 0.00011067961165048544,
+      "loss": 24.1997,
+      "step": 285
+    },
+    {
+      "epoch": 0.050010928961748635,
+      "grad_norm": 2.402788174789463,
+      "learning_rate": 0.00011106796116504854,
+      "loss": 24.0731,
+      "step": 286
+    },
+    {
+      "epoch": 0.050185792349726775,
+      "grad_norm": 2.4900064177923125,
+      "learning_rate": 0.00011145631067961166,
+      "loss": 24.2849,
+      "step": 287
+    },
+    {
+      "epoch": 0.050360655737704915,
+      "grad_norm": 1.9230144045707052,
+      "learning_rate": 0.00011184466019417477,
+      "loss": 24.2312,
+      "step": 288
+    },
+    {
+      "epoch": 0.05053551912568306,
+      "grad_norm": 1.6639732296051115,
+      "learning_rate": 0.00011223300970873786,
+      "loss": 24.1744,
+      "step": 289
+    },
+    {
+      "epoch": 0.0507103825136612,
+      "grad_norm": 1.7380523953191724,
+      "learning_rate": 0.00011262135922330097,
+      "loss": 24.2466,
+      "step": 290
+    },
+    {
+      "epoch": 0.05088524590163934,
+      "grad_norm": 1.7986385061073795,
+      "learning_rate": 0.00011300970873786407,
+      "loss": 24.2872,
+      "step": 291
+    },
+    {
+      "epoch": 0.05106010928961749,
+      "grad_norm": 1.9953697613917631,
+      "learning_rate": 0.0001133980582524272,
+      "loss": 24.2107,
+      "step": 292
+    },
+    {
+      "epoch": 0.05123497267759563,
+      "grad_norm": 1.5721242786111598,
+      "learning_rate": 0.0001137864077669903,
+      "loss": 24.1789,
+      "step": 293
+    },
+    {
+      "epoch": 0.05140983606557377,
+      "grad_norm": 1.5091374050441484,
+      "learning_rate": 0.0001141747572815534,
+      "loss": 24.0058,
+      "step": 294
+    },
+    {
+      "epoch": 0.051584699453551916,
+      "grad_norm": 1.4330478896483803,
+      "learning_rate": 0.0001145631067961165,
+      "loss": 23.6032,
+      "step": 295
+    },
+    {
+      "epoch": 0.051759562841530056,
+      "grad_norm": 1.3936984061539133,
+      "learning_rate": 0.00011495145631067961,
+      "loss": 24.2235,
+      "step": 296
+    },
+    {
+      "epoch": 0.051934426229508196,
+      "grad_norm": 1.3488376188913822,
+      "learning_rate": 0.00011533980582524273,
+      "loss": 24.1511,
+      "step": 297
+    },
+    {
+      "epoch": 0.052109289617486336,
+      "grad_norm": 1.3138387412381427,
+      "learning_rate": 0.00011572815533980584,
+      "loss": 23.9672,
+      "step": 298
+    },
+    {
+      "epoch": 0.05228415300546448,
+      "grad_norm": 1.3666809567336977,
+      "learning_rate": 0.00011611650485436893,
+      "loss": 23.861,
+      "step": 299
+    },
+    {
+      "epoch": 0.05245901639344262,
+      "grad_norm": 1.347392906365072,
+      "learning_rate": 0.00011650485436893204,
+      "loss": 24.0448,
+      "step": 300
+    },
+    {
+      "epoch": 0.05263387978142076,
+      "grad_norm": 1.2489019297135782,
+      "learning_rate": 0.00011689320388349516,
+      "loss": 23.5865,
+      "step": 301
+    },
+    {
+      "epoch": 0.05280874316939891,
+      "grad_norm": 1.4775922463957203,
+      "learning_rate": 0.00011728155339805826,
+      "loss": 24.201,
+      "step": 302
+    },
+    {
+      "epoch": 0.05298360655737705,
+      "grad_norm": 1.5407184917281043,
+      "learning_rate": 0.00011766990291262137,
+      "loss": 23.8972,
+      "step": 303
+    },
+    {
+      "epoch": 0.05315846994535519,
+      "grad_norm": 1.6301528421047358,
+      "learning_rate": 0.00011805825242718446,
+      "loss": 23.833,
+      "step": 304
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 1.5031765842400073,
+      "learning_rate": 0.00011844660194174757,
+      "loss": 23.7747,
+      "step": 305
+    },
+    {
+      "epoch": 0.05350819672131148,
+      "grad_norm": 1.518068854826819,
+      "learning_rate": 0.00011883495145631069,
+      "loss": 23.8982,
+      "step": 306
+    },
+    {
+      "epoch": 0.05368306010928962,
+      "grad_norm": 2.0170491997658324,
+      "learning_rate": 0.0001192233009708738,
+      "loss": 23.9209,
+      "step": 307
+    },
+    {
+      "epoch": 0.05385792349726776,
+      "grad_norm": 2.6787774682685823,
+      "learning_rate": 0.0001196116504854369,
+      "loss": 23.7983,
+      "step": 308
+    },
+    {
+      "epoch": 0.054032786885245904,
+      "grad_norm": 2.833868735772404,
+      "learning_rate": 0.00012,
+      "loss": 24.0811,
+      "step": 309
+    },
+    {
+      "epoch": 0.054207650273224044,
+      "grad_norm": 2.311054331190073,
+      "learning_rate": 0.0001203883495145631,
+      "loss": 24.0544,
+      "step": 310
+    },
+    {
+      "epoch": 0.054382513661202184,
+      "grad_norm": 2.429241282228084,
+      "learning_rate": 0.00012077669902912623,
+      "loss": 24.0111,
+      "step": 311
+    },
+    {
+      "epoch": 0.05455737704918033,
+      "grad_norm": 1.7098760253361627,
+      "learning_rate": 0.00012116504854368933,
+      "loss": 23.9614,
+      "step": 312
+    },
+    {
+      "epoch": 0.05473224043715847,
+      "grad_norm": 1.5796824174154733,
+      "learning_rate": 0.00012155339805825244,
+      "loss": 23.9022,
+      "step": 313
+    },
+    {
+      "epoch": 0.05490710382513661,
+      "grad_norm": 1.6364651153737892,
+      "learning_rate": 0.00012194174757281553,
+      "loss": 23.6749,
+      "step": 314
+    },
+    {
+      "epoch": 0.05508196721311476,
+      "grad_norm": 1.35377002267063,
+      "learning_rate": 0.00012233009708737864,
+      "loss": 23.6252,
+      "step": 315
+    },
+    {
+      "epoch": 0.0552568306010929,
+      "grad_norm": 1.4989262816439655,
+      "learning_rate": 0.00012271844660194175,
+      "loss": 23.852,
+      "step": 316
+    },
+    {
+      "epoch": 0.05543169398907104,
+      "grad_norm": 1.7533657476648798,
+      "learning_rate": 0.00012310679611650486,
+      "loss": 23.798,
+      "step": 317
+    },
+    {
+      "epoch": 0.05560655737704918,
+      "grad_norm": 1.7171136934662872,
+      "learning_rate": 0.00012349514563106796,
+      "loss": 23.8312,
+      "step": 318
+    },
+    {
+      "epoch": 0.055781420765027324,
+      "grad_norm": 1.9692613157809589,
+      "learning_rate": 0.00012388349514563107,
+      "loss": 23.7644,
+      "step": 319
+    },
+    {
+      "epoch": 0.055956284153005464,
+      "grad_norm": 1.5233186976504252,
+      "learning_rate": 0.00012427184466019418,
+      "loss": 23.7924,
+      "step": 320
+    },
+    {
+      "epoch": 0.056131147540983604,
+      "grad_norm": 1.6592455275844602,
+      "learning_rate": 0.0001246601941747573,
+      "loss": 23.6883,
+      "step": 321
+    },
+    {
+      "epoch": 0.05630601092896175,
+      "grad_norm": 1.6209270619718281,
+      "learning_rate": 0.0001250485436893204,
+      "loss": 23.5422,
+      "step": 322
+    },
+    {
+      "epoch": 0.05648087431693989,
+      "grad_norm": 1.6375699629030656,
+      "learning_rate": 0.0001254368932038835,
+      "loss": 23.7402,
+      "step": 323
+    },
+    {
+      "epoch": 0.05665573770491803,
+      "grad_norm": 1.4794951746826033,
+      "learning_rate": 0.0001258252427184466,
+      "loss": 23.7669,
+      "step": 324
+    },
+    {
+      "epoch": 0.05683060109289618,
+      "grad_norm": 1.587389948582562,
+      "learning_rate": 0.00012621359223300972,
+      "loss": 23.5248,
+      "step": 325
+    },
+    {
+      "epoch": 0.05700546448087432,
+      "grad_norm": 1.7387395632346945,
+      "learning_rate": 0.00012660194174757283,
+      "loss": 23.6999,
+      "step": 326
+    },
+    {
+      "epoch": 0.05718032786885246,
+      "grad_norm": 1.6994445579615245,
+      "learning_rate": 0.00012699029126213594,
+      "loss": 23.5905,
+      "step": 327
+    },
+    {
+      "epoch": 0.0573551912568306,
+      "grad_norm": 1.5869200793273068,
+      "learning_rate": 0.00012737864077669902,
+      "loss": 23.451,
+      "step": 328
+    },
+    {
+      "epoch": 0.057530054644808745,
+      "grad_norm": 1.7706049850118701,
+      "learning_rate": 0.00012776699029126213,
+      "loss": 23.6379,
+      "step": 329
+    },
+    {
+      "epoch": 0.057704918032786885,
+      "grad_norm": 1.7994100968252111,
+      "learning_rate": 0.00012815533980582526,
+      "loss": 23.6177,
+      "step": 330
+    },
+    {
+      "epoch": 0.057879781420765025,
+      "grad_norm": 1.9575956236047787,
+      "learning_rate": 0.00012854368932038837,
+      "loss": 23.4379,
+      "step": 331
+    },
+    {
+      "epoch": 0.05805464480874317,
+      "grad_norm": 1.7895011019930787,
+      "learning_rate": 0.00012893203883495145,
+      "loss": 23.5539,
+      "step": 332
+    },
+    {
+      "epoch": 0.05822950819672131,
+      "grad_norm": 2.040533383765471,
+      "learning_rate": 0.00012932038834951456,
+      "loss": 23.6989,
+      "step": 333
+    },
+    {
+      "epoch": 0.05840437158469945,
+      "grad_norm": 1.8877476509844433,
+      "learning_rate": 0.00012970873786407767,
+      "loss": 23.4679,
+      "step": 334
+    },
+    {
+      "epoch": 0.0585792349726776,
+      "grad_norm": 1.4552629307809595,
+      "learning_rate": 0.00013009708737864078,
+      "loss": 23.5266,
+      "step": 335
+    },
+    {
+      "epoch": 0.05875409836065574,
+      "grad_norm": 1.7601144572738068,
+      "learning_rate": 0.0001304854368932039,
+      "loss": 23.1991,
+      "step": 336
+    },
+    {
+      "epoch": 0.05892896174863388,
+      "grad_norm": 1.7764459818777258,
+      "learning_rate": 0.000130873786407767,
+      "loss": 23.6786,
+      "step": 337
+    },
+    {
+      "epoch": 0.05910382513661202,
+      "grad_norm": 1.524193736161843,
+      "learning_rate": 0.0001312621359223301,
+      "loss": 23.581,
+      "step": 338
+    },
+    {
+      "epoch": 0.059278688524590166,
+      "grad_norm": 1.7686627278434328,
+      "learning_rate": 0.0001316504854368932,
+      "loss": 23.3358,
+      "step": 339
+    },
+    {
+      "epoch": 0.059453551912568306,
+      "grad_norm": 1.6824339921910365,
+      "learning_rate": 0.00013203883495145632,
+      "loss": 23.2653,
+      "step": 340
+    },
+    {
+      "epoch": 0.059628415300546446,
+      "grad_norm": 1.5482883904799893,
+      "learning_rate": 0.00013242718446601943,
+      "loss": 23.4462,
+      "step": 341
+    },
+    {
+      "epoch": 0.05980327868852459,
+      "grad_norm": 1.7902233490441095,
+      "learning_rate": 0.00013281553398058254,
+      "loss": 23.3945,
+      "step": 342
+    },
+    {
+      "epoch": 0.05997814207650273,
+      "grad_norm": 1.8108427241499443,
+      "learning_rate": 0.00013320388349514562,
+      "loss": 22.89,
+      "step": 343
+    },
+    {
+      "epoch": 0.06015300546448087,
+      "grad_norm": 2.0093294662481003,
+      "learning_rate": 0.00013359223300970873,
+      "loss": 23.4052,
+      "step": 344
+    },
+    {
+      "epoch": 0.06032786885245902,
+      "grad_norm": 2.100419867282342,
+      "learning_rate": 0.00013398058252427186,
+      "loss": 23.3609,
+      "step": 345
+    },
+    {
+      "epoch": 0.06050273224043716,
+      "grad_norm": 1.980604570638805,
+      "learning_rate": 0.00013436893203883497,
+      "loss": 23.0003,
+      "step": 346
+    },
+    {
+      "epoch": 0.0606775956284153,
+      "grad_norm": 2.1796836948144542,
+      "learning_rate": 0.00013475728155339805,
+      "loss": 23.3696,
+      "step": 347
+    },
+    {
+      "epoch": 0.06085245901639344,
+      "grad_norm": 2.337866351885552,
+      "learning_rate": 0.00013514563106796116,
+      "loss": 23.5409,
+      "step": 348
+    },
+    {
+      "epoch": 0.06102732240437159,
+      "grad_norm": 1.9902769906508848,
+      "learning_rate": 0.0001355339805825243,
+      "loss": 23.4349,
+      "step": 349
+    },
+    {
+      "epoch": 0.06120218579234973,
+      "grad_norm": 1.610352636880809,
+      "learning_rate": 0.0001359223300970874,
+      "loss": 23.4395,
+      "step": 350
+    },
+    {
+      "epoch": 0.06137704918032787,
+      "grad_norm": 1.893644138333461,
+      "learning_rate": 0.00013631067961165049,
+      "loss": 23.4562,
+      "step": 351
+    },
+    {
+      "epoch": 0.061551912568306014,
+      "grad_norm": 1.6828598021906145,
+      "learning_rate": 0.0001366990291262136,
+      "loss": 23.1564,
+      "step": 352
+    },
+    {
+      "epoch": 0.061726775956284154,
+      "grad_norm": 1.623481345764388,
+      "learning_rate": 0.0001370873786407767,
+      "loss": 23.3102,
+      "step": 353
+    },
+    {
+      "epoch": 0.061901639344262294,
+      "grad_norm": 1.8518827289877329,
+      "learning_rate": 0.0001374757281553398,
+      "loss": 23.3254,
+      "step": 354
+    },
+    {
+      "epoch": 0.06207650273224044,
+      "grad_norm": 1.6475556708956478,
+      "learning_rate": 0.00013786407766990292,
+      "loss": 23.1935,
+      "step": 355
+    },
+    {
+      "epoch": 0.06225136612021858,
+      "grad_norm": 1.9057773424800768,
+      "learning_rate": 0.00013825242718446603,
+      "loss": 23.1687,
+      "step": 356
+    },
+    {
+      "epoch": 0.06242622950819672,
+      "grad_norm": 2.210369043444016,
+      "learning_rate": 0.00013864077669902914,
+      "loss": 23.3319,
+      "step": 357
+    },
+    {
+      "epoch": 0.06260109289617487,
+      "grad_norm": 3.285275033032434,
+      "learning_rate": 0.00013902912621359224,
+      "loss": 23.3736,
+      "step": 358
+    },
+    {
+      "epoch": 0.062775956284153,
+      "grad_norm": 3.602057149266493,
+      "learning_rate": 0.00013941747572815535,
+      "loss": 23.2721,
+      "step": 359
+    },
+    {
+      "epoch": 0.06295081967213115,
+      "grad_norm": 3.093188755808431,
+      "learning_rate": 0.00013980582524271846,
+      "loss": 23.2672,
+      "step": 360
+    },
+    {
+      "epoch": 0.0631256830601093,
+      "grad_norm": 2.2432077717656256,
+      "learning_rate": 0.00014019417475728157,
+      "loss": 23.2938,
+      "step": 361
+    },
+    {
+      "epoch": 0.06330054644808743,
+      "grad_norm": 2.133305046631337,
+      "learning_rate": 0.00014058252427184465,
+      "loss": 23.2296,
+      "step": 362
+    },
+    {
+      "epoch": 0.06347540983606557,
+      "grad_norm": 2.116360355706155,
+      "learning_rate": 0.00014097087378640776,
+      "loss": 23.0299,
+      "step": 363
+    },
+    {
+      "epoch": 0.06365027322404372,
+      "grad_norm": 1.9435201296204598,
+      "learning_rate": 0.0001413592233009709,
+      "loss": 23.0754,
+      "step": 364
+    },
+    {
+      "epoch": 0.06382513661202185,
+      "grad_norm": 1.782462756947844,
+      "learning_rate": 0.000141747572815534,
+      "loss": 23.141,
+      "step": 365
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 2.0067782956008444,
+      "learning_rate": 0.00014213592233009708,
+      "loss": 23.0293,
+      "step": 366
+    },
+    {
+      "epoch": 0.06417486338797815,
+      "grad_norm": 1.6008938353874256,
+      "learning_rate": 0.0001425242718446602,
+      "loss": 23.2856,
+      "step": 367
+    },
+    {
+      "epoch": 0.06434972677595628,
+      "grad_norm": 1.6560109066526083,
+      "learning_rate": 0.0001429126213592233,
+      "loss": 23.1026,
+      "step": 368
+    },
+    {
+      "epoch": 0.06452459016393443,
+      "grad_norm": 1.6409024127309622,
+      "learning_rate": 0.0001433009708737864,
+      "loss": 22.7873,
+      "step": 369
+    },
+    {
+      "epoch": 0.06469945355191258,
+      "grad_norm": 1.5942725241881093,
+      "learning_rate": 0.00014368932038834952,
+      "loss": 23.0121,
+      "step": 370
+    },
+    {
+      "epoch": 0.06487431693989071,
+      "grad_norm": 1.5482022212031907,
+      "learning_rate": 0.00014407766990291263,
+      "loss": 22.8767,
+      "step": 371
+    },
+    {
+      "epoch": 0.06504918032786886,
+      "grad_norm": 1.4827531087440338,
+      "learning_rate": 0.00014446601941747573,
+      "loss": 22.9782,
+      "step": 372
+    },
+    {
+      "epoch": 0.06522404371584699,
+      "grad_norm": 1.6297544589264816,
+      "learning_rate": 0.00014485436893203884,
+      "loss": 22.7253,
+      "step": 373
+    },
+    {
+      "epoch": 0.06539890710382514,
+      "grad_norm": 1.6533657433495585,
+      "learning_rate": 0.00014524271844660195,
+      "loss": 22.8724,
+      "step": 374
+    },
+    {
+      "epoch": 0.06557377049180328,
+      "grad_norm": 1.4542239987036234,
+      "learning_rate": 0.00014563106796116506,
+      "loss": 22.9576,
+      "step": 375
+    },
+    {
+      "epoch": 0.06574863387978142,
+      "grad_norm": 1.5473261991368021,
+      "learning_rate": 0.00014601941747572817,
+      "loss": 23.269,
+      "step": 376
+    },
+    {
+      "epoch": 0.06592349726775956,
+      "grad_norm": 1.4865328087851244,
+      "learning_rate": 0.00014640776699029128,
+      "loss": 22.8224,
+      "step": 377
+    },
+    {
+      "epoch": 0.06609836065573771,
+      "grad_norm": 1.49477953962068,
+      "learning_rate": 0.00014679611650485438,
+      "loss": 22.5471,
+      "step": 378
+    },
+    {
+      "epoch": 0.06627322404371584,
+      "grad_norm": 1.4797072109815994,
+      "learning_rate": 0.0001471844660194175,
+      "loss": 22.8381,
+      "step": 379
+    },
+    {
+      "epoch": 0.06644808743169399,
+      "grad_norm": 1.457859675823177,
+      "learning_rate": 0.0001475728155339806,
+      "loss": 22.854,
+      "step": 380
+    },
+    {
+      "epoch": 0.06662295081967214,
+      "grad_norm": 1.5704500154639327,
+      "learning_rate": 0.00014796116504854368,
+      "loss": 23.1334,
+      "step": 381
+    },
+    {
+      "epoch": 0.06679781420765027,
+      "grad_norm": 1.419258182760531,
+      "learning_rate": 0.0001483495145631068,
+      "loss": 22.9084,
+      "step": 382
+    },
+    {
+      "epoch": 0.06697267759562842,
+      "grad_norm": 1.485552613267447,
+      "learning_rate": 0.00014873786407766993,
+      "loss": 22.7106,
+      "step": 383
+    },
+    {
+      "epoch": 0.06714754098360656,
+      "grad_norm": 1.4712083947281478,
+      "learning_rate": 0.00014912621359223303,
+      "loss": 22.9091,
+      "step": 384
+    },
+    {
+      "epoch": 0.0673224043715847,
+      "grad_norm": 1.50830354049407,
+      "learning_rate": 0.00014951456310679611,
+      "loss": 22.749,
+      "step": 385
+    },
+    {
+      "epoch": 0.06749726775956284,
+      "grad_norm": 1.4524510687224534,
+      "learning_rate": 0.00014990291262135922,
+      "loss": 22.7188,
+      "step": 386
+    },
+    {
+      "epoch": 0.06767213114754099,
+      "grad_norm": 1.4324820973997159,
+      "learning_rate": 0.00015029126213592233,
+      "loss": 22.7684,
+      "step": 387
+    },
+    {
+      "epoch": 0.06784699453551912,
+      "grad_norm": 1.5260945812074533,
+      "learning_rate": 0.00015067961165048544,
+      "loss": 22.3835,
+      "step": 388
+    },
+    {
+      "epoch": 0.06802185792349727,
+      "grad_norm": 1.7913553050133053,
+      "learning_rate": 0.00015106796116504855,
+      "loss": 22.9144,
+      "step": 389
+    },
+    {
+      "epoch": 0.06819672131147542,
+      "grad_norm": 1.4680422835901554,
+      "learning_rate": 0.00015145631067961166,
+      "loss": 22.7067,
+      "step": 390
+    },
+    {
+      "epoch": 0.06837158469945355,
+      "grad_norm": 1.45419149481162,
+      "learning_rate": 0.00015184466019417477,
+      "loss": 22.9278,
+      "step": 391
+    },
+    {
+      "epoch": 0.0685464480874317,
+      "grad_norm": 1.458564398761365,
+      "learning_rate": 0.00015223300970873787,
+      "loss": 22.7202,
+      "step": 392
+    },
+    {
+      "epoch": 0.06872131147540983,
+      "grad_norm": 1.477358594797341,
+      "learning_rate": 0.00015262135922330098,
+      "loss": 22.7036,
+      "step": 393
+    },
+    {
+      "epoch": 0.06889617486338798,
+      "grad_norm": 1.6322994189298472,
+      "learning_rate": 0.0001530097087378641,
+      "loss": 22.5482,
+      "step": 394
+    },
+    {
+      "epoch": 0.06907103825136612,
+      "grad_norm": 1.660987940553468,
+      "learning_rate": 0.0001533980582524272,
+      "loss": 22.6975,
+      "step": 395
+    },
+    {
+      "epoch": 0.06924590163934426,
+      "grad_norm": 1.4780542946271675,
+      "learning_rate": 0.00015378640776699028,
+      "loss": 22.7798,
+      "step": 396
+    },
+    {
+      "epoch": 0.0694207650273224,
+      "grad_norm": 1.673106302310641,
+      "learning_rate": 0.0001541747572815534,
+      "loss": 22.5467,
+      "step": 397
+    },
+    {
+      "epoch": 0.06959562841530055,
+      "grad_norm": 1.5343473747911236,
+      "learning_rate": 0.00015456310679611652,
+      "loss": 22.7985,
+      "step": 398
+    },
+    {
+      "epoch": 0.06977049180327868,
+      "grad_norm": 1.5398078681183458,
+      "learning_rate": 0.00015495145631067963,
+      "loss": 22.9957,
+      "step": 399
+    },
+    {
+      "epoch": 0.06994535519125683,
+      "grad_norm": 1.6833795422492999,
+      "learning_rate": 0.0001553398058252427,
+      "loss": 22.3573,
+      "step": 400
+    },
+    {
+      "epoch": 0.07012021857923498,
+      "grad_norm": 1.9082385107465012,
+      "learning_rate": 0.00015572815533980582,
+      "loss": 22.46,
+      "step": 401
+    },
+    {
+      "epoch": 0.07029508196721311,
+      "grad_norm": 1.9269261531532407,
+      "learning_rate": 0.00015611650485436896,
+      "loss": 22.667,
+      "step": 402
+    },
+    {
+      "epoch": 0.07046994535519126,
+      "grad_norm": 1.7074410886530156,
+      "learning_rate": 0.00015650485436893207,
+      "loss": 22.6482,
+      "step": 403
+    },
+    {
+      "epoch": 0.0706448087431694,
+      "grad_norm": 1.6945289388928604,
+      "learning_rate": 0.00015689320388349515,
+      "loss": 23.0029,
+      "step": 404
+    },
+    {
+      "epoch": 0.07081967213114754,
+      "grad_norm": 1.6668204568743472,
+      "learning_rate": 0.00015728155339805825,
+      "loss": 22.663,
+      "step": 405
+    },
+    {
+      "epoch": 0.07099453551912568,
+      "grad_norm": 1.5818226491846241,
+      "learning_rate": 0.00015766990291262136,
+      "loss": 22.4802,
+      "step": 406
+    },
+    {
+      "epoch": 0.07116939890710383,
+      "grad_norm": 1.7357840460300422,
+      "learning_rate": 0.00015805825242718447,
+      "loss": 22.4758,
+      "step": 407
+    },
+    {
+      "epoch": 0.07134426229508196,
+      "grad_norm": 1.6085397085700184,
+      "learning_rate": 0.00015844660194174758,
+      "loss": 22.4665,
+      "step": 408
+    },
+    {
+      "epoch": 0.07151912568306011,
+      "grad_norm": 1.7512770928715726,
+      "learning_rate": 0.0001588349514563107,
+      "loss": 22.6685,
+      "step": 409
+    },
+    {
+      "epoch": 0.07169398907103826,
+      "grad_norm": 1.6833677664858693,
+      "learning_rate": 0.0001592233009708738,
+      "loss": 22.4977,
+      "step": 410
+    },
+    {
+      "epoch": 0.07186885245901639,
+      "grad_norm": 1.4333826754829326,
+      "learning_rate": 0.0001596116504854369,
+      "loss": 22.4874,
+      "step": 411
+    },
+    {
+      "epoch": 0.07204371584699454,
+      "grad_norm": 1.837444572305455,
+      "learning_rate": 0.00016,
+      "loss": 22.8173,
+      "step": 412
+    },
+    {
+      "epoch": 0.07221857923497267,
+      "grad_norm": 2.0447148458532127,
+      "learning_rate": 0.00016038834951456312,
+      "loss": 22.7866,
+      "step": 413
+    },
+    {
+      "epoch": 0.07239344262295082,
+      "grad_norm": 2.0642972176969225,
+      "learning_rate": 0.00016077669902912623,
+      "loss": 22.6282,
+      "step": 414
+    },
+    {
+      "epoch": 0.07256830601092897,
+      "grad_norm": 1.7814513338170588,
+      "learning_rate": 0.0001611650485436893,
+      "loss": 22.4617,
+      "step": 415
+    },
+    {
+      "epoch": 0.0727431693989071,
+      "grad_norm": 1.6291149486136525,
+      "learning_rate": 0.00016155339805825242,
+      "loss": 22.6194,
+      "step": 416
+    },
+    {
+      "epoch": 0.07291803278688525,
+      "grad_norm": 1.6579115753986717,
+      "learning_rate": 0.00016194174757281555,
+      "loss": 22.4056,
+      "step": 417
+    },
+    {
+      "epoch": 0.07309289617486339,
+      "grad_norm": 1.9318331149970445,
+      "learning_rate": 0.00016233009708737866,
+      "loss": 22.5959,
+      "step": 418
+    },
+    {
+      "epoch": 0.07326775956284153,
+      "grad_norm": 1.57736618504549,
+      "learning_rate": 0.00016271844660194174,
+      "loss": 22.454,
+      "step": 419
+    },
+    {
+      "epoch": 0.07344262295081967,
+      "grad_norm": 1.7924491250195533,
+      "learning_rate": 0.00016310679611650485,
+      "loss": 22.5675,
+      "step": 420
+    },
+    {
+      "epoch": 0.07361748633879782,
+      "grad_norm": 1.6121332875080483,
+      "learning_rate": 0.00016349514563106796,
+      "loss": 22.3391,
+      "step": 421
+    },
+    {
+      "epoch": 0.07379234972677595,
+      "grad_norm": 1.7192420620006958,
+      "learning_rate": 0.0001638834951456311,
+      "loss": 22.6723,
+      "step": 422
+    },
+    {
+      "epoch": 0.0739672131147541,
+      "grad_norm": 1.8321775681577632,
+      "learning_rate": 0.00016427184466019418,
+      "loss": 22.4186,
+      "step": 423
+    },
+    {
+      "epoch": 0.07414207650273225,
+      "grad_norm": 1.7430332757298823,
+      "learning_rate": 0.00016466019417475729,
+      "loss": 22.589,
+      "step": 424
+    },
+    {
+      "epoch": 0.07431693989071038,
+      "grad_norm": 1.9405827472313628,
+      "learning_rate": 0.0001650485436893204,
+      "loss": 22.196,
+      "step": 425
+    },
+    {
+      "epoch": 0.07449180327868853,
+      "grad_norm": 2.381948600238871,
+      "learning_rate": 0.0001654368932038835,
+      "loss": 22.6273,
+      "step": 426
+    },
+    {
+      "epoch": 0.07466666666666667,
+      "grad_norm": 2.687087089323218,
+      "learning_rate": 0.0001658252427184466,
+      "loss": 22.7823,
+      "step": 427
+    },
+    {
+      "epoch": 0.0748415300546448,
+      "grad_norm": 1.9493856133651704,
+      "learning_rate": 0.00016621359223300972,
+      "loss": 22.3855,
+      "step": 428
+    },
+    {
+      "epoch": 0.07501639344262295,
+      "grad_norm": 1.8188554417287346,
+      "learning_rate": 0.00016660194174757283,
+      "loss": 22.7219,
+      "step": 429
+    },
+    {
+      "epoch": 0.0751912568306011,
+      "grad_norm": 1.7909601826205321,
+      "learning_rate": 0.00016699029126213594,
+      "loss": 22.2807,
+      "step": 430
+    },
+    {
+      "epoch": 0.07536612021857923,
+      "grad_norm": 1.657825528397177,
+      "learning_rate": 0.00016737864077669904,
+      "loss": 22.1926,
+      "step": 431
+    },
+    {
+      "epoch": 0.07554098360655738,
+      "grad_norm": 1.9550512209933935,
+      "learning_rate": 0.00016776699029126215,
+      "loss": 22.503,
+      "step": 432
+    },
+    {
+      "epoch": 0.07571584699453551,
+      "grad_norm": 1.7443837964512405,
+      "learning_rate": 0.00016815533980582526,
+      "loss": 22.5625,
+      "step": 433
+    },
+    {
+      "epoch": 0.07589071038251366,
+      "grad_norm": 1.6698551853590518,
+      "learning_rate": 0.00016854368932038834,
+      "loss": 22.453,
+      "step": 434
+    },
+    {
+      "epoch": 0.0760655737704918,
+      "grad_norm": 1.5434364600487818,
+      "learning_rate": 0.00016893203883495145,
+      "loss": 22.4266,
+      "step": 435
+    },
+    {
+      "epoch": 0.07624043715846994,
+      "grad_norm": 1.6571444591334152,
+      "learning_rate": 0.00016932038834951459,
+      "loss": 22.1424,
+      "step": 436
+    },
+    {
+      "epoch": 0.07641530054644809,
+      "grad_norm": 1.4070596511719842,
+      "learning_rate": 0.0001697087378640777,
+      "loss": 22.4079,
+      "step": 437
+    },
+    {
+      "epoch": 0.07659016393442623,
+      "grad_norm": 1.5061427026842724,
+      "learning_rate": 0.00017009708737864078,
+      "loss": 22.3731,
+      "step": 438
+    },
+    {
+      "epoch": 0.07676502732240437,
+      "grad_norm": 1.4219567438962946,
+      "learning_rate": 0.00017048543689320388,
+      "loss": 22.454,
+      "step": 439
+    },
+    {
+      "epoch": 0.07693989071038251,
+      "grad_norm": 1.64778166198791,
+      "learning_rate": 0.000170873786407767,
+      "loss": 22.2351,
+      "step": 440
+    },
+    {
+      "epoch": 0.07711475409836066,
+      "grad_norm": 1.4989250612230454,
+      "learning_rate": 0.0001712621359223301,
+      "loss": 22.3433,
+      "step": 441
+    },
+    {
+      "epoch": 0.0772896174863388,
+      "grad_norm": 1.6040734738967313,
+      "learning_rate": 0.0001716504854368932,
+      "loss": 22.2999,
+      "step": 442
+    },
+    {
+      "epoch": 0.07746448087431694,
+      "grad_norm": 1.6517096864406065,
+      "learning_rate": 0.00017203883495145632,
+      "loss": 22.3107,
+      "step": 443
+    },
+    {
+      "epoch": 0.07763934426229509,
+      "grad_norm": 1.6850823957042895,
+      "learning_rate": 0.00017242718446601943,
+      "loss": 22.5389,
+      "step": 444
+    },
+    {
+      "epoch": 0.07781420765027322,
+      "grad_norm": 1.5574828435148171,
+      "learning_rate": 0.00017281553398058253,
+      "loss": 22.363,
+      "step": 445
+    },
+    {
+      "epoch": 0.07798907103825137,
+      "grad_norm": 1.4230336002100596,
+      "learning_rate": 0.00017320388349514564,
+      "loss": 22.3149,
+      "step": 446
+    },
+    {
+      "epoch": 0.07816393442622951,
+      "grad_norm": 1.541697029452484,
+      "learning_rate": 0.00017359223300970875,
+      "loss": 22.5084,
+      "step": 447
+    },
+    {
+      "epoch": 0.07833879781420765,
+      "grad_norm": 1.7090834665408232,
+      "learning_rate": 0.00017398058252427186,
+      "loss": 22.4051,
+      "step": 448
+    },
+    {
+      "epoch": 0.0785136612021858,
+      "grad_norm": 1.5665252377267607,
+      "learning_rate": 0.00017436893203883494,
+      "loss": 22.234,
+      "step": 449
+    },
+    {
+      "epoch": 0.07868852459016394,
+      "grad_norm": 1.5315119437307703,
+      "learning_rate": 0.00017475728155339805,
+      "loss": 22.5203,
+      "step": 450
+    },
+    {
+      "epoch": 0.07886338797814207,
+      "grad_norm": 1.4053989737606034,
+      "learning_rate": 0.00017514563106796118,
+      "loss": 22.4551,
+      "step": 451
+    },
+    {
+      "epoch": 0.07903825136612022,
+      "grad_norm": 1.5720905926261681,
+      "learning_rate": 0.0001755339805825243,
+      "loss": 22.2055,
+      "step": 452
+    },
+    {
+      "epoch": 0.07921311475409835,
+      "grad_norm": 1.968095193766102,
+      "learning_rate": 0.00017592233009708737,
+      "loss": 22.3609,
+      "step": 453
+    },
+    {
+      "epoch": 0.0793879781420765,
+      "grad_norm": 1.7016683103256518,
+      "learning_rate": 0.00017631067961165048,
+      "loss": 22.4068,
+      "step": 454
+    },
+    {
+      "epoch": 0.07956284153005465,
+      "grad_norm": 1.5642285413428207,
+      "learning_rate": 0.00017669902912621362,
+      "loss": 22.1687,
+      "step": 455
+    },
+    {
+      "epoch": 0.07973770491803278,
+      "grad_norm": 1.523312140202522,
+      "learning_rate": 0.00017708737864077673,
+      "loss": 21.8166,
+      "step": 456
+    },
+    {
+      "epoch": 0.07991256830601093,
+      "grad_norm": 1.744981992664748,
+      "learning_rate": 0.0001774757281553398,
+      "loss": 22.2373,
+      "step": 457
+    },
+    {
+      "epoch": 0.08008743169398908,
+      "grad_norm": 1.5416112498721504,
+      "learning_rate": 0.00017786407766990292,
+      "loss": 22.1719,
+      "step": 458
+    },
+    {
+      "epoch": 0.08026229508196721,
+      "grad_norm": 1.6698164963395143,
+      "learning_rate": 0.00017825242718446602,
+      "loss": 22.1361,
+      "step": 459
+    },
+    {
+      "epoch": 0.08043715846994536,
+      "grad_norm": 1.5909545644710787,
+      "learning_rate": 0.00017864077669902913,
+      "loss": 22.278,
+      "step": 460
+    },
+    {
+      "epoch": 0.0806120218579235,
+      "grad_norm": 1.5186935922925675,
+      "learning_rate": 0.00017902912621359224,
+      "loss": 21.853,
+      "step": 461
+    },
+    {
+      "epoch": 0.08078688524590164,
+      "grad_norm": 1.5620419815431477,
+      "learning_rate": 0.00017941747572815535,
+      "loss": 22.0707,
+      "step": 462
+    },
+    {
+      "epoch": 0.08096174863387978,
+      "grad_norm": 1.4254644433190773,
+      "learning_rate": 0.00017980582524271846,
+      "loss": 22.2617,
+      "step": 463
+    },
+    {
+      "epoch": 0.08113661202185793,
+      "grad_norm": 1.6480149335593344,
+      "learning_rate": 0.00018019417475728157,
+      "loss": 22.0279,
+      "step": 464
+    },
+    {
+      "epoch": 0.08131147540983606,
+      "grad_norm": 1.505941090597203,
+      "learning_rate": 0.00018058252427184467,
+      "loss": 22.2981,
+      "step": 465
+    },
+    {
+      "epoch": 0.08148633879781421,
+      "grad_norm": 1.5535237343745822,
+      "learning_rate": 0.00018097087378640778,
+      "loss": 22.2173,
+      "step": 466
+    },
+    {
+      "epoch": 0.08166120218579236,
+      "grad_norm": 1.3591949382470476,
+      "learning_rate": 0.0001813592233009709,
+      "loss": 21.8488,
+      "step": 467
+    },
+    {
+      "epoch": 0.08183606557377049,
+      "grad_norm": 1.6639616080345727,
+      "learning_rate": 0.00018174757281553397,
+      "loss": 22.1584,
+      "step": 468
+    },
+    {
+      "epoch": 0.08201092896174864,
+      "grad_norm": 1.4484300201563793,
+      "learning_rate": 0.00018213592233009708,
+      "loss": 22.2266,
+      "step": 469
+    },
+    {
+      "epoch": 0.08218579234972678,
+      "grad_norm": 1.4216522198177663,
+      "learning_rate": 0.00018252427184466022,
+      "loss": 22.1438,
+      "step": 470
+    },
+    {
+      "epoch": 0.08236065573770492,
+      "grad_norm": 1.584438887330677,
+      "learning_rate": 0.00018291262135922332,
+      "loss": 22.0837,
+      "step": 471
+    },
+    {
+      "epoch": 0.08253551912568306,
+      "grad_norm": 1.761685271726996,
+      "learning_rate": 0.0001833009708737864,
+      "loss": 22.1318,
+      "step": 472
+    },
+    {
+      "epoch": 0.0827103825136612,
+      "grad_norm": 1.6943793411317238,
+      "learning_rate": 0.0001836893203883495,
+      "loss": 22.0787,
+      "step": 473
+    },
+    {
+      "epoch": 0.08288524590163934,
+      "grad_norm": 1.6216442619200884,
+      "learning_rate": 0.00018407766990291262,
+      "loss": 22.189,
+      "step": 474
+    },
+    {
+      "epoch": 0.08306010928961749,
+      "grad_norm": 2.087538504586377,
+      "learning_rate": 0.00018446601941747576,
+      "loss": 22.2204,
+      "step": 475
+    },
+    {
+      "epoch": 0.08323497267759562,
+      "grad_norm": 2.1086371928849905,
+      "learning_rate": 0.00018485436893203884,
+      "loss": 22.0829,
+      "step": 476
+    },
+    {
+      "epoch": 0.08340983606557377,
+      "grad_norm": 1.7959287129402397,
+      "learning_rate": 0.00018524271844660195,
+      "loss": 22.127,
+      "step": 477
+    },
+    {
+      "epoch": 0.08358469945355192,
+      "grad_norm": 1.654758372529112,
+      "learning_rate": 0.00018563106796116506,
+      "loss": 21.6711,
+      "step": 478
+    },
+    {
+      "epoch": 0.08375956284153005,
+      "grad_norm": 1.5678369486476094,
+      "learning_rate": 0.00018601941747572816,
+      "loss": 22.2965,
+      "step": 479
+    },
+    {
+      "epoch": 0.0839344262295082,
+      "grad_norm": 1.5858670297898745,
+      "learning_rate": 0.00018640776699029127,
+      "loss": 21.9604,
+      "step": 480
+    },
+    {
+      "epoch": 0.08410928961748634,
+      "grad_norm": 1.6007831025403958,
+      "learning_rate": 0.00018679611650485438,
+      "loss": 21.7396,
+      "step": 481
+    },
+    {
+      "epoch": 0.08428415300546448,
+      "grad_norm": 1.5916125732638868,
+      "learning_rate": 0.0001871844660194175,
+      "loss": 21.5677,
+      "step": 482
+    },
+    {
+      "epoch": 0.08445901639344262,
+      "grad_norm": 1.5827779065793333,
+      "learning_rate": 0.0001875728155339806,
+      "loss": 21.8268,
+      "step": 483
+    },
+    {
+      "epoch": 0.08463387978142077,
+      "grad_norm": 1.7717806575123736,
+      "learning_rate": 0.0001879611650485437,
+      "loss": 21.9115,
+      "step": 484
+    },
+    {
+      "epoch": 0.0848087431693989,
+      "grad_norm": 1.5391617010173666,
+      "learning_rate": 0.00018834951456310681,
+      "loss": 22.0358,
+      "step": 485
+    },
+    {
+      "epoch": 0.08498360655737705,
+      "grad_norm": 1.4442779424018597,
+      "learning_rate": 0.00018873786407766992,
+      "loss": 21.8761,
+      "step": 486
+    },
+    {
+      "epoch": 0.0851584699453552,
+      "grad_norm": 1.4503540287234666,
+      "learning_rate": 0.000189126213592233,
+      "loss": 21.9838,
+      "step": 487
+    },
+    {
+      "epoch": 0.08533333333333333,
+      "grad_norm": 1.5575966657374933,
+      "learning_rate": 0.0001895145631067961,
+      "loss": 22.1111,
+      "step": 488
+    },
+    {
+      "epoch": 0.08550819672131148,
+      "grad_norm": 1.3911356087010316,
+      "learning_rate": 0.00018990291262135925,
+      "loss": 21.8753,
+      "step": 489
+    },
+    {
+      "epoch": 0.08568306010928962,
+      "grad_norm": 1.5481988906954534,
+      "learning_rate": 0.00019029126213592236,
+      "loss": 21.7535,
+      "step": 490
+    },
+    {
+      "epoch": 0.08585792349726776,
+      "grad_norm": 1.4972477147146168,
+      "learning_rate": 0.00019067961165048544,
+      "loss": 21.9273,
+      "step": 491
+    },
+    {
+      "epoch": 0.0860327868852459,
+      "grad_norm": 1.5729456928599983,
+      "learning_rate": 0.00019106796116504854,
+      "loss": 21.6329,
+      "step": 492
+    },
+    {
+      "epoch": 0.08620765027322404,
+      "grad_norm": 1.4491150003953728,
+      "learning_rate": 0.00019145631067961165,
+      "loss": 21.9927,
+      "step": 493
+    },
+    {
+      "epoch": 0.08638251366120218,
+      "grad_norm": 1.5015452116638142,
+      "learning_rate": 0.00019184466019417476,
+      "loss": 22.0337,
+      "step": 494
+    },
+    {
+      "epoch": 0.08655737704918033,
+      "grad_norm": 1.5487470361956077,
+      "learning_rate": 0.00019223300970873787,
+      "loss": 21.5224,
+      "step": 495
+    },
+    {
+      "epoch": 0.08673224043715846,
+      "grad_norm": 1.592130962115304,
+      "learning_rate": 0.00019262135922330098,
+      "loss": 21.6164,
+      "step": 496
+    },
+    {
+      "epoch": 0.08690710382513661,
+      "grad_norm": 1.7656867103405427,
+      "learning_rate": 0.0001930097087378641,
+      "loss": 21.7937,
+      "step": 497
+    },
+    {
+      "epoch": 0.08708196721311476,
+      "grad_norm": 1.695855419010997,
+      "learning_rate": 0.0001933980582524272,
+      "loss": 21.768,
+      "step": 498
+    },
+    {
+      "epoch": 0.08725683060109289,
+      "grad_norm": 1.4586270854095689,
+      "learning_rate": 0.0001937864077669903,
+      "loss": 21.6736,
+      "step": 499
+    },
+    {
+      "epoch": 0.08743169398907104,
+      "grad_norm": 1.6315182326287516,
+      "learning_rate": 0.0001941747572815534,
+      "loss": 21.8954,
+      "step": 500
+    },
+    {
+      "epoch": 0.08760655737704919,
+      "grad_norm": 1.4768432268002138,
+      "learning_rate": 0.00019456310679611652,
+      "loss": 21.9182,
+      "step": 501
+    },
+    {
+      "epoch": 0.08778142076502732,
+      "grad_norm": 1.6325767673499638,
+      "learning_rate": 0.0001949514563106796,
+      "loss": 21.8541,
+      "step": 502
+    },
+    {
+      "epoch": 0.08795628415300547,
+      "grad_norm": 1.4992263208949212,
+      "learning_rate": 0.00019533980582524274,
+      "loss": 21.9824,
+      "step": 503
+    },
+    {
+      "epoch": 0.08813114754098361,
+      "grad_norm": 1.4803757660053558,
+      "learning_rate": 0.00019572815533980584,
+      "loss": 21.6765,
+      "step": 504
+    },
+    {
+      "epoch": 0.08830601092896175,
+      "grad_norm": 1.5981839869117533,
+      "learning_rate": 0.00019611650485436895,
+      "loss": 21.8101,
+      "step": 505
+    },
+    {
+      "epoch": 0.08848087431693989,
+      "grad_norm": 1.4930096052976118,
+      "learning_rate": 0.00019650485436893203,
+      "loss": 21.9434,
+      "step": 506
+    },
+    {
+      "epoch": 0.08865573770491804,
+      "grad_norm": 1.7389945142710554,
+      "learning_rate": 0.00019689320388349514,
+      "loss": 21.9629,
+      "step": 507
+    },
+    {
+      "epoch": 0.08883060109289617,
+      "grad_norm": 1.752527255155556,
+      "learning_rate": 0.00019728155339805828,
+      "loss": 21.7151,
+      "step": 508
+    },
+    {
+      "epoch": 0.08900546448087432,
+      "grad_norm": 2.0026530332784866,
+      "learning_rate": 0.0001976699029126214,
+      "loss": 21.2661,
+      "step": 509
+    },
+    {
+      "epoch": 0.08918032786885247,
+      "grad_norm": 1.935492570212818,
+      "learning_rate": 0.00019805825242718447,
+      "loss": 21.8792,
+      "step": 510
+    },
+    {
+      "epoch": 0.0893551912568306,
+      "grad_norm": 1.8315888908236948,
+      "learning_rate": 0.00019844660194174758,
+      "loss": 21.9572,
+      "step": 511
+    },
+    {
+      "epoch": 0.08953005464480875,
+      "grad_norm": 1.682793528428224,
+      "learning_rate": 0.00019883495145631068,
+      "loss": 21.5654,
+      "step": 512
+    },
+    {
+      "epoch": 0.08970491803278688,
+      "grad_norm": 1.4728401675844127,
+      "learning_rate": 0.0001992233009708738,
+      "loss": 21.7578,
+      "step": 513
+    },
+    {
+      "epoch": 0.08987978142076503,
+      "grad_norm": 1.4507275498768977,
+      "learning_rate": 0.0001996116504854369,
+      "loss": 21.752,
+      "step": 514
+    },
+    {
+      "epoch": 0.09005464480874317,
+      "grad_norm": 1.4624066878334696,
+      "learning_rate": 0.0002,
+      "loss": 21.5151,
+      "step": 515
+    },
+    {
+      "epoch": 0.0902295081967213,
+      "grad_norm": 1.5663501605383396,
+      "learning_rate": 0.00019999999821755844,
+      "loss": 21.7713,
+      "step": 516
+    },
+    {
+      "epoch": 0.09040437158469945,
+      "grad_norm": 1.4073414861426643,
+      "learning_rate": 0.00019999999287023375,
+      "loss": 21.5218,
+      "step": 517
+    },
+    {
+      "epoch": 0.0905792349726776,
+      "grad_norm": 1.3915402135178898,
+      "learning_rate": 0.00019999998395802616,
+      "loss": 21.4469,
+      "step": 518
+    },
+    {
+      "epoch": 0.09075409836065573,
+      "grad_norm": 1.5626973473166133,
+      "learning_rate": 0.00019999997148093597,
+      "loss": 21.3755,
+      "step": 519
+    },
+    {
+      "epoch": 0.09092896174863388,
+      "grad_norm": 1.5394659841834721,
+      "learning_rate": 0.00019999995543896367,
+      "loss": 21.404,
+      "step": 520
+    },
+    {
+      "epoch": 0.09110382513661203,
+      "grad_norm": 1.5760579878267653,
+      "learning_rate": 0.00019999993583210977,
+      "loss": 21.7745,
+      "step": 521
+    },
+    {
+      "epoch": 0.09127868852459016,
+      "grad_norm": 1.5083970614839504,
+      "learning_rate": 0.00019999991266037498,
+      "loss": 21.6818,
+      "step": 522
+    },
+    {
+      "epoch": 0.0914535519125683,
+      "grad_norm": 1.5127434140395897,
+      "learning_rate": 0.00019999988592376015,
+      "loss": 21.4476,
+      "step": 523
+    },
+    {
+      "epoch": 0.09162841530054645,
+      "grad_norm": 1.472710017782009,
+      "learning_rate": 0.00019999985562226623,
+      "loss": 21.6091,
+      "step": 524
+    },
+    {
+      "epoch": 0.09180327868852459,
+      "grad_norm": 1.4469824858596865,
+      "learning_rate": 0.0001999998217558943,
+      "loss": 21.9552,
+      "step": 525
+    },
+    {
+      "epoch": 0.09197814207650273,
+      "grad_norm": 1.4218669565320168,
+      "learning_rate": 0.00019999978432464553,
+      "loss": 21.853,
+      "step": 526
+    },
+    {
+      "epoch": 0.09215300546448088,
+      "grad_norm": 1.6120458842287857,
+      "learning_rate": 0.00019999974332852137,
+      "loss": 21.3693,
+      "step": 527
+    },
+    {
+      "epoch": 0.09232786885245901,
+      "grad_norm": 1.762108050648944,
+      "learning_rate": 0.0001999996987675231,
+      "loss": 21.5389,
+      "step": 528
+    },
+    {
+      "epoch": 0.09250273224043716,
+      "grad_norm": 1.5326531228388447,
+      "learning_rate": 0.00019999965064165247,
+      "loss": 21.6794,
+      "step": 529
+    },
+    {
+      "epoch": 0.09267759562841531,
+      "grad_norm": 1.5431208225800142,
+      "learning_rate": 0.0001999995989509111,
+      "loss": 21.8269,
+      "step": 530
+    },
+    {
+      "epoch": 0.09285245901639344,
+      "grad_norm": 1.5931870521655485,
+      "learning_rate": 0.00019999954369530088,
+      "loss": 21.6959,
+      "step": 531
+    },
+    {
+      "epoch": 0.09302732240437159,
+      "grad_norm": 1.5796237884125146,
+      "learning_rate": 0.00019999948487482377,
+      "loss": 21.277,
+      "step": 532
+    },
+    {
+      "epoch": 0.09320218579234972,
+      "grad_norm": 1.4189910891944213,
+      "learning_rate": 0.00019999942248948182,
+      "loss": 21.7232,
+      "step": 533
+    },
+    {
+      "epoch": 0.09337704918032787,
+      "grad_norm": 1.4335124751257342,
+      "learning_rate": 0.00019999935653927732,
+      "loss": 21.7186,
+      "step": 534
+    },
+    {
+      "epoch": 0.09355191256830601,
+      "grad_norm": 1.563574454532597,
+      "learning_rate": 0.00019999928702421263,
+      "loss": 21.5368,
+      "step": 535
+    },
+    {
+      "epoch": 0.09372677595628415,
+      "grad_norm": 1.456651230640845,
+      "learning_rate": 0.00019999921394429015,
+      "loss": 21.8046,
+      "step": 536
+    },
+    {
+      "epoch": 0.0939016393442623,
+      "grad_norm": 1.8631784637492115,
+      "learning_rate": 0.00019999913729951252,
+      "loss": 21.7923,
+      "step": 537
+    },
+    {
+      "epoch": 0.09407650273224044,
+      "grad_norm": 1.5930876825672058,
+      "learning_rate": 0.0001999990570898825,
+      "loss": 21.9396,
+      "step": 538
+    },
+    {
+      "epoch": 0.09425136612021857,
+      "grad_norm": 1.578186428351345,
+      "learning_rate": 0.00019999897331540294,
+      "loss": 21.5401,
+      "step": 539
+    },
+    {
+      "epoch": 0.09442622950819672,
+      "grad_norm": 1.4385164011327294,
+      "learning_rate": 0.00019999888597607682,
+      "loss": 21.7341,
+      "step": 540
+    },
+    {
+      "epoch": 0.09460109289617487,
+      "grad_norm": 1.4593660638402952,
+      "learning_rate": 0.00019999879507190724,
+      "loss": 21.3149,
+      "step": 541
+    },
+    {
+      "epoch": 0.094775956284153,
+      "grad_norm": 1.513756151100862,
+      "learning_rate": 0.00019999870060289748,
+      "loss": 21.3167,
+      "step": 542
+    },
+    {
+      "epoch": 0.09495081967213115,
+      "grad_norm": 1.6105495901633282,
+      "learning_rate": 0.00019999860256905084,
+      "loss": 21.1241,
+      "step": 543
+    },
+    {
+      "epoch": 0.0951256830601093,
+      "grad_norm": 1.3104841586103604,
+      "learning_rate": 0.00019999850097037088,
+      "loss": 21.3351,
+      "step": 544
+    },
+    {
+      "epoch": 0.09530054644808743,
+      "grad_norm": 1.544758508432627,
+      "learning_rate": 0.00019999839580686117,
+      "loss": 21.8086,
+      "step": 545
+    },
+    {
+      "epoch": 0.09547540983606558,
+      "grad_norm": 1.5611842109141982,
+      "learning_rate": 0.0001999982870785255,
+      "loss": 21.6175,
+      "step": 546
+    },
+    {
+      "epoch": 0.09565027322404372,
+      "grad_norm": 1.4456383222711473,
+      "learning_rate": 0.00019999817478536774,
+      "loss": 21.516,
+      "step": 547
+    },
+    {
+      "epoch": 0.09582513661202186,
+      "grad_norm": 1.559493918504419,
+      "learning_rate": 0.00019999805892739186,
+      "loss": 21.4345,
+      "step": 548
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 1.418048297129011,
+      "learning_rate": 0.00019999793950460206,
+      "loss": 21.7738,
+      "step": 549
+    },
+    {
+      "epoch": 0.09617486338797815,
+      "grad_norm": 1.5356870885136218,
+      "learning_rate": 0.00019999781651700252,
+      "loss": 21.6373,
+      "step": 550
+    },
+    {
+      "epoch": 0.09634972677595628,
+      "grad_norm": 1.4256807952593002,
+      "learning_rate": 0.00019999768996459765,
+      "loss": 21.6271,
+      "step": 551
+    },
+    {
+      "epoch": 0.09652459016393443,
+      "grad_norm": 1.4478158876705693,
+      "learning_rate": 0.00019999755984739196,
+      "loss": 21.5831,
+      "step": 552
+    },
+    {
+      "epoch": 0.09669945355191256,
+      "grad_norm": 1.355120742014885,
+      "learning_rate": 0.00019999742616539013,
+      "loss": 21.4753,
+      "step": 553
+    },
+    {
+      "epoch": 0.09687431693989071,
+      "grad_norm": 1.564410649407237,
+      "learning_rate": 0.00019999728891859685,
+      "loss": 21.43,
+      "step": 554
+    },
+    {
+      "epoch": 0.09704918032786886,
+      "grad_norm": 1.4682053601249963,
+      "learning_rate": 0.0001999971481070171,
+      "loss": 21.3804,
+      "step": 555
+    },
+    {
+      "epoch": 0.09722404371584699,
+      "grad_norm": 1.7146744302004981,
+      "learning_rate": 0.00019999700373065583,
+      "loss": 21.3727,
+      "step": 556
+    },
+    {
+      "epoch": 0.09739890710382514,
+      "grad_norm": 1.5378756309345263,
+      "learning_rate": 0.00019999685578951822,
+      "loss": 21.6489,
+      "step": 557
+    },
+    {
+      "epoch": 0.09757377049180328,
+      "grad_norm": 1.4460148064686693,
+      "learning_rate": 0.00019999670428360951,
+      "loss": 20.9609,
+      "step": 558
+    },
+    {
+      "epoch": 0.09774863387978142,
+      "grad_norm": 1.3998330236085832,
+      "learning_rate": 0.00019999654921293517,
+      "loss": 21.074,
+      "step": 559
+    },
+    {
+      "epoch": 0.09792349726775956,
+      "grad_norm": 1.486069912595028,
+      "learning_rate": 0.00019999639057750063,
+      "loss": 21.5626,
+      "step": 560
+    },
+    {
+      "epoch": 0.09809836065573771,
+      "grad_norm": 1.5468621568389536,
+      "learning_rate": 0.0001999962283773116,
+      "loss": 21.4322,
+      "step": 561
+    },
+    {
+      "epoch": 0.09827322404371584,
+      "grad_norm": 1.478652138456163,
+      "learning_rate": 0.00019999606261237395,
+      "loss": 21.4271,
+      "step": 562
+    },
+    {
+      "epoch": 0.09844808743169399,
+      "grad_norm": 1.4351947995028043,
+      "learning_rate": 0.0001999958932826934,
+      "loss": 21.4821,
+      "step": 563
+    },
+    {
+      "epoch": 0.09862295081967214,
+      "grad_norm": 1.6509398465793443,
+      "learning_rate": 0.00019999572038827614,
+      "loss": 21.5234,
+      "step": 564
+    },
+    {
+      "epoch": 0.09879781420765027,
+      "grad_norm": 1.7285335959051016,
+      "learning_rate": 0.00019999554392912826,
+      "loss": 21.4537,
+      "step": 565
+    },
+    {
+      "epoch": 0.09897267759562842,
+      "grad_norm": 1.6786810676171795,
+      "learning_rate": 0.0001999953639052561,
+      "loss": 21.4216,
+      "step": 566
+    },
+    {
+      "epoch": 0.09914754098360656,
+      "grad_norm": 1.4288356366044446,
+      "learning_rate": 0.00019999518031666602,
+      "loss": 21.2988,
+      "step": 567
+    },
+    {
+      "epoch": 0.0993224043715847,
+      "grad_norm": 1.406405250081455,
+      "learning_rate": 0.0001999949931633646,
+      "loss": 21.5415,
+      "step": 568
+    },
+    {
+      "epoch": 0.09949726775956284,
+      "grad_norm": 1.5612103476875205,
+      "learning_rate": 0.00019999480244535848,
+      "loss": 21.263,
+      "step": 569
+    },
+    {
+      "epoch": 0.09967213114754099,
+      "grad_norm": 1.5575393056454105,
+      "learning_rate": 0.0001999946081626545,
+      "loss": 21.6311,
+      "step": 570
+    },
+    {
+      "epoch": 0.09984699453551912,
+      "grad_norm": 1.3301399152477933,
+      "learning_rate": 0.0001999944103152596,
+      "loss": 21.4014,
+      "step": 571
+    },
+    {
+      "epoch": 0.10002185792349727,
+      "grad_norm": 1.542702442327069,
+      "learning_rate": 0.0001999942089031808,
+      "loss": 21.5682,
+      "step": 572
+    },
+    {
+      "epoch": 0.1001967213114754,
+      "grad_norm": 1.5591590149522698,
+      "learning_rate": 0.00019999400392642523,
+      "loss": 21.6319,
+      "step": 573
+    },
+    {
+      "epoch": 0.10037158469945355,
+      "grad_norm": 1.4015353622093518,
+      "learning_rate": 0.0001999937953850003,
+      "loss": 21.0777,
+      "step": 574
+    },
+    {
+      "epoch": 0.1005464480874317,
+      "grad_norm": 1.485120089767315,
+      "learning_rate": 0.0001999935832789134,
+      "loss": 21.1657,
+      "step": 575
+    },
+    {
+      "epoch": 0.10072131147540983,
+      "grad_norm": 1.631510098996298,
+      "learning_rate": 0.00019999336760817205,
+      "loss": 21.4196,
+      "step": 576
+    },
+    {
+      "epoch": 0.10089617486338798,
+      "grad_norm": 1.5607082447123368,
+      "learning_rate": 0.000199993148372784,
+      "loss": 21.4619,
+      "step": 577
+    },
+    {
+      "epoch": 0.10107103825136612,
+      "grad_norm": 1.5579605116566313,
+      "learning_rate": 0.00019999292557275702,
+      "loss": 21.5023,
+      "step": 578
+    },
+    {
+      "epoch": 0.10124590163934426,
+      "grad_norm": 1.6002542512540017,
+      "learning_rate": 0.0001999926992080991,
+      "loss": 21.0933,
+      "step": 579
+    },
+    {
+      "epoch": 0.1014207650273224,
+      "grad_norm": 1.4607468035804105,
+      "learning_rate": 0.00019999246927881824,
+      "loss": 21.3615,
+      "step": 580
+    },
+    {
+      "epoch": 0.10159562841530055,
+      "grad_norm": 1.3845131479127328,
+      "learning_rate": 0.00019999223578492275,
+      "loss": 21.4523,
+      "step": 581
+    },
+    {
+      "epoch": 0.10177049180327868,
+      "grad_norm": 1.480233332224391,
+      "learning_rate": 0.0001999919987264208,
+      "loss": 21.0257,
+      "step": 582
+    },
+    {
+      "epoch": 0.10194535519125683,
+      "grad_norm": 1.424665675503719,
+      "learning_rate": 0.00019999175810332098,
+      "loss": 21.0237,
+      "step": 583
+    },
+    {
+      "epoch": 0.10212021857923498,
+      "grad_norm": 1.48958610120618,
+      "learning_rate": 0.0001999915139156318,
+      "loss": 20.9914,
+      "step": 584
+    },
+    {
+      "epoch": 0.10229508196721311,
+      "grad_norm": 1.2961643851132956,
+      "learning_rate": 0.00019999126616336197,
+      "loss": 21.1286,
+      "step": 585
+    },
+    {
+      "epoch": 0.10246994535519126,
+      "grad_norm": 1.4526903599382266,
+      "learning_rate": 0.0001999910148465203,
+      "loss": 21.3468,
+      "step": 586
+    },
+    {
+      "epoch": 0.1026448087431694,
+      "grad_norm": 1.378300376152645,
+      "learning_rate": 0.00019999075996511584,
+      "loss": 21.2379,
+      "step": 587
+    },
+    {
+      "epoch": 0.10281967213114754,
+      "grad_norm": 1.4516308013103802,
+      "learning_rate": 0.00019999050151915757,
+      "loss": 21.0438,
+      "step": 588
+    },
+    {
+      "epoch": 0.10299453551912569,
+      "grad_norm": 1.4386989112002002,
+      "learning_rate": 0.00019999023950865476,
+      "loss": 21.1477,
+      "step": 589
+    },
+    {
+      "epoch": 0.10316939890710383,
+      "grad_norm": 1.4986606894851209,
+      "learning_rate": 0.00019998997393361672,
+      "loss": 21.3292,
+      "step": 590
+    },
+    {
+      "epoch": 0.10334426229508197,
+      "grad_norm": 1.2926603961947927,
+      "learning_rate": 0.00019998970479405298,
+      "loss": 21.2093,
+      "step": 591
+    },
+    {
+      "epoch": 0.10351912568306011,
+      "grad_norm": 1.3926820365732147,
+      "learning_rate": 0.00019998943208997304,
+      "loss": 21.3882,
+      "step": 592
+    },
+    {
+      "epoch": 0.10369398907103825,
+      "grad_norm": 1.3577396716561607,
+      "learning_rate": 0.0001999891558213867,
+      "loss": 21.392,
+      "step": 593
+    },
+    {
+      "epoch": 0.10386885245901639,
+      "grad_norm": 1.4448958959769551,
+      "learning_rate": 0.00019998887598830377,
+      "loss": 21.4463,
+      "step": 594
+    },
+    {
+      "epoch": 0.10404371584699454,
+      "grad_norm": 1.2524720447652753,
+      "learning_rate": 0.00019998859259073422,
+      "loss": 21.4549,
+      "step": 595
+    },
+    {
+      "epoch": 0.10421857923497267,
+      "grad_norm": 1.2504759349835721,
+      "learning_rate": 0.00019998830562868817,
+      "loss": 21.133,
+      "step": 596
+    },
+    {
+      "epoch": 0.10439344262295082,
+      "grad_norm": 1.3830526932745792,
+      "learning_rate": 0.0001999880151021759,
+      "loss": 21.2999,
+      "step": 597
+    },
+    {
+      "epoch": 0.10456830601092897,
+      "grad_norm": 1.3257853281186347,
+      "learning_rate": 0.00019998772101120763,
+      "loss": 21.2445,
+      "step": 598
+    },
+    {
+      "epoch": 0.1047431693989071,
+      "grad_norm": 1.3339047887075912,
+      "learning_rate": 0.000199987423355794,
+      "loss": 21.4049,
+      "step": 599
+    },
+    {
+      "epoch": 0.10491803278688525,
+      "grad_norm": 1.2862690219942938,
+      "learning_rate": 0.00019998712213594553,
+      "loss": 21.1874,
+      "step": 600
+    },
+    {
+      "epoch": 0.10509289617486339,
+      "grad_norm": 1.4239643711140693,
+      "learning_rate": 0.00019998681735167298,
+      "loss": 21.03,
+      "step": 601
+    },
+    {
+      "epoch": 0.10526775956284153,
+      "grad_norm": 1.3294669277495423,
+      "learning_rate": 0.0001999865090029872,
+      "loss": 20.6044,
+      "step": 602
+    },
+    {
+      "epoch": 0.10544262295081967,
+      "grad_norm": 1.43430472118243,
+      "learning_rate": 0.0001999861970898992,
+      "loss": 20.9213,
+      "step": 603
+    },
+    {
+      "epoch": 0.10561748633879782,
+      "grad_norm": 1.5915828260326397,
+      "learning_rate": 0.0001999858816124201,
+      "loss": 20.919,
+      "step": 604
+    },
+    {
+      "epoch": 0.10579234972677595,
+      "grad_norm": 1.6251027672666483,
+      "learning_rate": 0.00019998556257056114,
+      "loss": 21.0002,
+      "step": 605
+    },
+    {
+      "epoch": 0.1059672131147541,
+      "grad_norm": 1.5075862901908788,
+      "learning_rate": 0.00019998523996433369,
+      "loss": 21.2611,
+      "step": 606
+    },
+    {
+      "epoch": 0.10614207650273225,
+      "grad_norm": 1.4373532558919255,
+      "learning_rate": 0.00019998491379374927,
+      "loss": 21.0713,
+      "step": 607
+    },
+    {
+      "epoch": 0.10631693989071038,
+      "grad_norm": 1.5031206901355736,
+      "learning_rate": 0.00019998458405881948,
+      "loss": 20.9206,
+      "step": 608
+    },
+    {
+      "epoch": 0.10649180327868853,
+      "grad_norm": 1.545214325440606,
+      "learning_rate": 0.00019998425075955608,
+      "loss": 20.9908,
+      "step": 609
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.4923442988584725,
+      "learning_rate": 0.00019998391389597098,
+      "loss": 21.2207,
+      "step": 610
+    },
+    {
+      "epoch": 0.1068415300546448,
+      "grad_norm": 1.499314260474079,
+      "learning_rate": 0.00019998357346807618,
+      "loss": 21.4015,
+      "step": 611
+    },
+    {
+      "epoch": 0.10701639344262295,
+      "grad_norm": 1.5727498146487346,
+      "learning_rate": 0.00019998322947588376,
+      "loss": 21.5875,
+      "step": 612
+    },
+    {
+      "epoch": 0.10719125683060109,
+      "grad_norm": 1.4839985754357456,
+      "learning_rate": 0.00019998288191940607,
+      "loss": 21.0614,
+      "step": 613
+    },
+    {
+      "epoch": 0.10736612021857923,
+      "grad_norm": 1.3056970476164746,
+      "learning_rate": 0.0001999825307986554,
+      "loss": 20.815,
+      "step": 614
+    },
+    {
+      "epoch": 0.10754098360655738,
+      "grad_norm": 1.405955248823047,
+      "learning_rate": 0.0001999821761136444,
+      "loss": 20.7235,
+      "step": 615
+    },
+    {
+      "epoch": 0.10771584699453551,
+      "grad_norm": 1.2902075260350376,
+      "learning_rate": 0.00019998181786438557,
+      "loss": 21.3318,
+      "step": 616
+    },
+    {
+      "epoch": 0.10789071038251366,
+      "grad_norm": 1.3733684829974098,
+      "learning_rate": 0.00019998145605089177,
+      "loss": 21.0189,
+      "step": 617
+    },
+    {
+      "epoch": 0.10806557377049181,
+      "grad_norm": 1.2633030658402202,
+      "learning_rate": 0.0001999810906731759,
+      "loss": 20.6548,
+      "step": 618
+    },
+    {
+      "epoch": 0.10824043715846994,
+      "grad_norm": 1.42724005140538,
+      "learning_rate": 0.00019998072173125096,
+      "loss": 21.2501,
+      "step": 619
+    },
+    {
+      "epoch": 0.10841530054644809,
+      "grad_norm": 1.4342416500759791,
+      "learning_rate": 0.00019998034922513006,
+      "loss": 20.9497,
+      "step": 620
+    },
+    {
+      "epoch": 0.10859016393442623,
+      "grad_norm": 1.4335392636583382,
+      "learning_rate": 0.00019997997315482658,
+      "loss": 20.8335,
+      "step": 621
+    },
+    {
+      "epoch": 0.10876502732240437,
+      "grad_norm": 1.4739683642812622,
+      "learning_rate": 0.00019997959352035385,
+      "loss": 21.1807,
+      "step": 622
+    },
+    {
+      "epoch": 0.10893989071038251,
+      "grad_norm": 1.265243423560929,
+      "learning_rate": 0.00019997921032172543,
+      "loss": 21.3424,
+      "step": 623
+    },
+    {
+      "epoch": 0.10911475409836066,
+      "grad_norm": 1.399462684459209,
+      "learning_rate": 0.00019997882355895496,
+      "loss": 21.1475,
+      "step": 624
+    },
+    {
+      "epoch": 0.1092896174863388,
+      "grad_norm": 1.4328765689070113,
+      "learning_rate": 0.00019997843323205628,
+      "loss": 21.1664,
+      "step": 625
+    },
+    {
+      "epoch": 0.10946448087431694,
+      "grad_norm": 1.3915290552927835,
+      "learning_rate": 0.00019997803934104322,
+      "loss": 21.1486,
+      "step": 626
+    },
+    {
+      "epoch": 0.10963934426229509,
+      "grad_norm": 1.2534448867727173,
+      "learning_rate": 0.00019997764188592988,
+      "loss": 21.2436,
+      "step": 627
+    },
+    {
+      "epoch": 0.10981420765027322,
+      "grad_norm": 1.3680981821039544,
+      "learning_rate": 0.00019997724086673045,
+      "loss": 20.8907,
+      "step": 628
+    },
+    {
+      "epoch": 0.10998907103825137,
+      "grad_norm": 1.429991040691319,
+      "learning_rate": 0.00019997683628345916,
+      "loss": 20.6385,
+      "step": 629
+    },
+    {
+      "epoch": 0.11016393442622952,
+      "grad_norm": 1.473514270829343,
+      "learning_rate": 0.00019997642813613048,
+      "loss": 21.2429,
+      "step": 630
+    },
+    {
+      "epoch": 0.11033879781420765,
+      "grad_norm": 1.4418886103229043,
+      "learning_rate": 0.00019997601642475897,
+      "loss": 20.7405,
+      "step": 631
+    },
+    {
+      "epoch": 0.1105136612021858,
+      "grad_norm": 1.399503000827461,
+      "learning_rate": 0.00019997560114935924,
+      "loss": 20.7463,
+      "step": 632
+    },
+    {
+      "epoch": 0.11068852459016393,
+      "grad_norm": 1.4136010287192413,
+      "learning_rate": 0.00019997518230994613,
+      "loss": 20.9935,
+      "step": 633
+    },
+    {
+      "epoch": 0.11086338797814208,
+      "grad_norm": 1.5225226295500658,
+      "learning_rate": 0.00019997475990653462,
+      "loss": 20.3201,
+      "step": 634
+    },
+    {
+      "epoch": 0.11103825136612022,
+      "grad_norm": 1.6541404535645081,
+      "learning_rate": 0.00019997433393913975,
+      "loss": 20.8392,
+      "step": 635
+    },
+    {
+      "epoch": 0.11121311475409836,
+      "grad_norm": 1.3704741884043672,
+      "learning_rate": 0.0001999739044077766,
+      "loss": 21.0464,
+      "step": 636
+    },
+    {
+      "epoch": 0.1113879781420765,
+      "grad_norm": 1.3521531898924628,
+      "learning_rate": 0.00019997347131246062,
+      "loss": 21.1612,
+      "step": 637
+    },
+    {
+      "epoch": 0.11156284153005465,
+      "grad_norm": 1.4230459494502,
+      "learning_rate": 0.00019997303465320722,
+      "loss": 21.0674,
+      "step": 638
+    },
+    {
+      "epoch": 0.11173770491803278,
+      "grad_norm": 1.6084631716283078,
+      "learning_rate": 0.00019997259443003187,
+      "loss": 21.0788,
+      "step": 639
+    },
+    {
+      "epoch": 0.11191256830601093,
+      "grad_norm": 1.3658334408284842,
+      "learning_rate": 0.00019997215064295037,
+      "loss": 20.9051,
+      "step": 640
+    },
+    {
+      "epoch": 0.11208743169398908,
+      "grad_norm": 1.395754293206865,
+      "learning_rate": 0.00019997170329197851,
+      "loss": 21.034,
+      "step": 641
+    },
+    {
+      "epoch": 0.11226229508196721,
+      "grad_norm": 1.5200058583558713,
+      "learning_rate": 0.00019997125237713223,
+      "loss": 20.7846,
+      "step": 642
+    },
+    {
+      "epoch": 0.11243715846994536,
+      "grad_norm": 1.5109413034662726,
+      "learning_rate": 0.0001999707978984276,
+      "loss": 20.8589,
+      "step": 643
+    },
+    {
+      "epoch": 0.1126120218579235,
+      "grad_norm": 1.4334270079196423,
+      "learning_rate": 0.00019997033985588085,
+      "loss": 20.8393,
+      "step": 644
+    },
+    {
+      "epoch": 0.11278688524590164,
+      "grad_norm": 1.4698518047770226,
+      "learning_rate": 0.00019996987824950826,
+      "loss": 20.9827,
+      "step": 645
+    },
+    {
+      "epoch": 0.11296174863387978,
+      "grad_norm": 1.4859893617060778,
+      "learning_rate": 0.00019996941307932634,
+      "loss": 21.0379,
+      "step": 646
+    },
+    {
+      "epoch": 0.11313661202185793,
+      "grad_norm": 1.3432883459488068,
+      "learning_rate": 0.0001999689443453516,
+      "loss": 20.7753,
+      "step": 647
+    },
+    {
+      "epoch": 0.11331147540983606,
+      "grad_norm": 1.3688122314247202,
+      "learning_rate": 0.00019996847204760082,
+      "loss": 21.2198,
+      "step": 648
+    },
+    {
+      "epoch": 0.11348633879781421,
+      "grad_norm": 1.3187048115921922,
+      "learning_rate": 0.00019996799618609082,
+      "loss": 20.8223,
+      "step": 649
+    },
+    {
+      "epoch": 0.11366120218579236,
+      "grad_norm": 1.323459621336572,
+      "learning_rate": 0.00019996751676083853,
+      "loss": 21.1542,
+      "step": 650
+    },
+    {
+      "epoch": 0.11383606557377049,
+      "grad_norm": 1.3347004836233718,
+      "learning_rate": 0.0001999670337718611,
+      "loss": 20.7618,
+      "step": 651
+    },
+    {
+      "epoch": 0.11401092896174864,
+      "grad_norm": 1.3546133084959977,
+      "learning_rate": 0.00019996654721917572,
+      "loss": 21.2173,
+      "step": 652
+    },
+    {
+      "epoch": 0.11418579234972677,
+      "grad_norm": 1.2902484748560015,
+      "learning_rate": 0.0001999660571027997,
+      "loss": 20.6198,
+      "step": 653
+    },
+    {
+      "epoch": 0.11436065573770492,
+      "grad_norm": 1.268308427677259,
+      "learning_rate": 0.00019996556342275054,
+      "loss": 20.786,
+      "step": 654
+    },
+    {
+      "epoch": 0.11453551912568306,
+      "grad_norm": 1.3452098956061154,
+      "learning_rate": 0.00019996506617904586,
+      "loss": 20.6291,
+      "step": 655
+    },
+    {
+      "epoch": 0.1147103825136612,
+      "grad_norm": 1.2069614445264691,
+      "learning_rate": 0.00019996456537170335,
+      "loss": 21.0425,
+      "step": 656
+    },
+    {
+      "epoch": 0.11488524590163934,
+      "grad_norm": 1.326486793716623,
+      "learning_rate": 0.0001999640610007409,
+      "loss": 20.8136,
+      "step": 657
+    },
+    {
+      "epoch": 0.11506010928961749,
+      "grad_norm": 1.3694737341581176,
+      "learning_rate": 0.00019996355306617646,
+      "loss": 20.7922,
+      "step": 658
+    },
+    {
+      "epoch": 0.11523497267759562,
+      "grad_norm": 1.2847565644505237,
+      "learning_rate": 0.00019996304156802818,
+      "loss": 20.9856,
+      "step": 659
+    },
+    {
+      "epoch": 0.11540983606557377,
+      "grad_norm": 1.3085012295265666,
+      "learning_rate": 0.00019996252650631422,
+      "loss": 20.7964,
+      "step": 660
+    },
+    {
+      "epoch": 0.11558469945355192,
+      "grad_norm": 1.3793150426066287,
+      "learning_rate": 0.000199962007881053,
+      "loss": 20.8835,
+      "step": 661
+    },
+    {
+      "epoch": 0.11575956284153005,
+      "grad_norm": 1.443912111137165,
+      "learning_rate": 0.00019996148569226298,
+      "loss": 20.8484,
+      "step": 662
+    },
+    {
+      "epoch": 0.1159344262295082,
+      "grad_norm": 1.375108707427781,
+      "learning_rate": 0.0001999609599399628,
+      "loss": 20.3645,
+      "step": 663
+    },
+    {
+      "epoch": 0.11610928961748634,
+      "grad_norm": 1.2823948633971998,
+      "learning_rate": 0.00019996043062417117,
+      "loss": 20.9268,
+      "step": 664
+    },
+    {
+      "epoch": 0.11628415300546448,
+      "grad_norm": 1.3186212221347229,
+      "learning_rate": 0.00019995989774490702,
+      "loss": 20.8695,
+      "step": 665
+    },
+    {
+      "epoch": 0.11645901639344262,
+      "grad_norm": 1.444109984251212,
+      "learning_rate": 0.00019995936130218927,
+      "loss": 20.7315,
+      "step": 666
+    },
+    {
+      "epoch": 0.11663387978142077,
+      "grad_norm": 1.391582119352991,
+      "learning_rate": 0.0001999588212960371,
+      "loss": 20.7947,
+      "step": 667
+    },
+    {
+      "epoch": 0.1168087431693989,
+      "grad_norm": 1.384102557931775,
+      "learning_rate": 0.00019995827772646972,
+      "loss": 20.725,
+      "step": 668
+    },
+    {
+      "epoch": 0.11698360655737705,
+      "grad_norm": 1.4253835243449573,
+      "learning_rate": 0.00019995773059350655,
+      "loss": 20.6505,
+      "step": 669
+    },
+    {
+      "epoch": 0.1171584699453552,
+      "grad_norm": 1.4813931894153756,
+      "learning_rate": 0.00019995717989716708,
+      "loss": 20.7136,
+      "step": 670
+    },
+    {
+      "epoch": 0.11733333333333333,
+      "grad_norm": 1.292735968992483,
+      "learning_rate": 0.00019995662563747092,
+      "loss": 20.8383,
+      "step": 671
+    },
+    {
+      "epoch": 0.11750819672131148,
+      "grad_norm": 1.461263871482395,
+      "learning_rate": 0.00019995606781443784,
+      "loss": 20.8659,
+      "step": 672
+    },
+    {
+      "epoch": 0.11768306010928961,
+      "grad_norm": 1.3269208700653763,
+      "learning_rate": 0.0001999555064280877,
+      "loss": 20.9973,
+      "step": 673
+    },
+    {
+      "epoch": 0.11785792349726776,
+      "grad_norm": 1.3360510358455027,
+      "learning_rate": 0.00019995494147844058,
+      "loss": 20.9169,
+      "step": 674
+    },
+    {
+      "epoch": 0.1180327868852459,
+      "grad_norm": 1.3830616963666775,
+      "learning_rate": 0.0001999543729655166,
+      "loss": 20.7052,
+      "step": 675
+    },
+    {
+      "epoch": 0.11820765027322404,
+      "grad_norm": 1.51345700613002,
+      "learning_rate": 0.00019995380088933594,
+      "loss": 20.8428,
+      "step": 676
+    },
+    {
+      "epoch": 0.11838251366120218,
+      "grad_norm": 1.4887587609898862,
+      "learning_rate": 0.00019995322524991911,
+      "loss": 20.8145,
+      "step": 677
+    },
+    {
+      "epoch": 0.11855737704918033,
+      "grad_norm": 1.4640960323845742,
+      "learning_rate": 0.00019995264604728656,
+      "loss": 20.7119,
+      "step": 678
+    },
+    {
+      "epoch": 0.11873224043715846,
+      "grad_norm": 1.2445771528414142,
+      "learning_rate": 0.000199952063281459,
+      "loss": 20.7485,
+      "step": 679
+    },
+    {
+      "epoch": 0.11890710382513661,
+      "grad_norm": 1.3301963883020411,
+      "learning_rate": 0.00019995147695245713,
+      "loss": 20.632,
+      "step": 680
+    },
+    {
+      "epoch": 0.11908196721311476,
+      "grad_norm": 1.3562546108925047,
+      "learning_rate": 0.0001999508870603019,
+      "loss": 20.6064,
+      "step": 681
+    },
+    {
+      "epoch": 0.11925683060109289,
+      "grad_norm": 1.3390691280950568,
+      "learning_rate": 0.0001999502936050143,
+      "loss": 20.9188,
+      "step": 682
+    },
+    {
+      "epoch": 0.11943169398907104,
+      "grad_norm": 1.3328102593254942,
+      "learning_rate": 0.00019994969658661556,
+      "loss": 20.7272,
+      "step": 683
+    },
+    {
+      "epoch": 0.11960655737704919,
+      "grad_norm": 1.2997418910822645,
+      "learning_rate": 0.00019994909600512688,
+      "loss": 20.5439,
+      "step": 684
+    },
+    {
+      "epoch": 0.11978142076502732,
+      "grad_norm": 1.3162429262026676,
+      "learning_rate": 0.00019994849186056972,
+      "loss": 20.9732,
+      "step": 685
+    },
+    {
+      "epoch": 0.11995628415300547,
+      "grad_norm": 1.331119643501685,
+      "learning_rate": 0.0001999478841529656,
+      "loss": 20.9466,
+      "step": 686
+    },
+    {
+      "epoch": 0.12013114754098361,
+      "grad_norm": 1.3630607174128468,
+      "learning_rate": 0.00019994727288233615,
+      "loss": 20.6307,
+      "step": 687
+    },
+    {
+      "epoch": 0.12030601092896175,
+      "grad_norm": 1.343867661115932,
+      "learning_rate": 0.00019994665804870328,
+      "loss": 20.8521,
+      "step": 688
+    },
+    {
+      "epoch": 0.12048087431693989,
+      "grad_norm": 1.3015871290935346,
+      "learning_rate": 0.00019994603965208877,
+      "loss": 20.7935,
+      "step": 689
+    },
+    {
+      "epoch": 0.12065573770491804,
+      "grad_norm": 1.4120117832021912,
+      "learning_rate": 0.0001999454176925147,
+      "loss": 20.5387,
+      "step": 690
+    },
+    {
+      "epoch": 0.12083060109289617,
+      "grad_norm": 1.4130037878814872,
+      "learning_rate": 0.00019994479217000334,
+      "loss": 20.4618,
+      "step": 691
+    },
+    {
+      "epoch": 0.12100546448087432,
+      "grad_norm": 1.3687914148587064,
+      "learning_rate": 0.00019994416308457682,
+      "loss": 20.5442,
+      "step": 692
+    },
+    {
+      "epoch": 0.12118032786885245,
+      "grad_norm": 1.3811907095290517,
+      "learning_rate": 0.00019994353043625772,
+      "loss": 20.3624,
+      "step": 693
+    },
+    {
+      "epoch": 0.1213551912568306,
+      "grad_norm": 1.3183782419586418,
+      "learning_rate": 0.00019994289422506852,
+      "loss": 20.7026,
+      "step": 694
+    },
+    {
+      "epoch": 0.12153005464480875,
+      "grad_norm": 1.3649485108558763,
+      "learning_rate": 0.0001999422544510319,
+      "loss": 20.6981,
+      "step": 695
+    },
+    {
+      "epoch": 0.12170491803278688,
+      "grad_norm": 1.5113535330807033,
+      "learning_rate": 0.00019994161111417066,
+      "loss": 20.2814,
+      "step": 696
+    },
+    {
+      "epoch": 0.12187978142076503,
+      "grad_norm": 1.3834861872184578,
+      "learning_rate": 0.0001999409642145078,
+      "loss": 20.7604,
+      "step": 697
+    },
+    {
+      "epoch": 0.12205464480874317,
+      "grad_norm": 1.4539867393796597,
+      "learning_rate": 0.0001999403137520663,
+      "loss": 20.7304,
+      "step": 698
+    },
+    {
+      "epoch": 0.1222295081967213,
+      "grad_norm": 1.3412801368067135,
+      "learning_rate": 0.0001999396597268694,
+      "loss": 20.6878,
+      "step": 699
+    },
+    {
+      "epoch": 0.12240437158469945,
+      "grad_norm": 1.4359423267397655,
+      "learning_rate": 0.0001999390021389404,
+      "loss": 20.533,
+      "step": 700
+    },
+    {
+      "epoch": 0.1225792349726776,
+      "grad_norm": 1.49385708281787,
+      "learning_rate": 0.0001999383409883027,
+      "loss": 20.947,
+      "step": 701
+    },
+    {
+      "epoch": 0.12275409836065573,
+      "grad_norm": 1.3153449788975269,
+      "learning_rate": 0.00019993767627497997,
+      "loss": 20.9254,
+      "step": 702
+    },
+    {
+      "epoch": 0.12292896174863388,
+      "grad_norm": 1.4122049083730508,
+      "learning_rate": 0.00019993700799899583,
+      "loss": 20.8905,
+      "step": 703
+    },
+    {
+      "epoch": 0.12310382513661203,
+      "grad_norm": 1.3364514392786968,
+      "learning_rate": 0.0001999363361603741,
+      "loss": 20.3494,
+      "step": 704
+    },
+    {
+      "epoch": 0.12327868852459016,
+      "grad_norm": 1.4395285895007168,
+      "learning_rate": 0.00019993566075913872,
+      "loss": 20.9163,
+      "step": 705
+    },
+    {
+      "epoch": 0.12345355191256831,
+      "grad_norm": 1.4381314706526465,
+      "learning_rate": 0.00019993498179531386,
+      "loss": 20.557,
+      "step": 706
+    },
+    {
+      "epoch": 0.12362841530054645,
+      "grad_norm": 1.3078948236417194,
+      "learning_rate": 0.0001999342992689236,
+      "loss": 20.7446,
+      "step": 707
+    },
+    {
+      "epoch": 0.12380327868852459,
+      "grad_norm": 1.3216327354960593,
+      "learning_rate": 0.00019993361317999236,
+      "loss": 20.5877,
+      "step": 708
+    },
+    {
+      "epoch": 0.12397814207650273,
+      "grad_norm": 1.2898538186196304,
+      "learning_rate": 0.00019993292352854457,
+      "loss": 20.549,
+      "step": 709
+    },
+    {
+      "epoch": 0.12415300546448088,
+      "grad_norm": 1.3422718778715095,
+      "learning_rate": 0.00019993223031460484,
+      "loss": 20.8468,
+      "step": 710
+    },
+    {
+      "epoch": 0.12432786885245901,
+      "grad_norm": 1.2936552766196858,
+      "learning_rate": 0.00019993153353819784,
+      "loss": 20.6873,
+      "step": 711
+    },
+    {
+      "epoch": 0.12450273224043716,
+      "grad_norm": 1.3064786964509192,
+      "learning_rate": 0.0001999308331993484,
+      "loss": 20.774,
+      "step": 712
+    },
+    {
+      "epoch": 0.1246775956284153,
+      "grad_norm": 1.3836211112868535,
+      "learning_rate": 0.00019993012929808154,
+      "loss": 21.0009,
+      "step": 713
+    },
+    {
+      "epoch": 0.12485245901639344,
+      "grad_norm": 1.3191330367458096,
+      "learning_rate": 0.00019992942183442232,
+      "loss": 20.6148,
+      "step": 714
+    },
+    {
+      "epoch": 0.1250273224043716,
+      "grad_norm": 1.4727587514116907,
+      "learning_rate": 0.000199928710808396,
+      "loss": 20.5058,
+      "step": 715
+    },
+    {
+      "epoch": 0.12520218579234973,
+      "grad_norm": 1.3770834944055408,
+      "learning_rate": 0.00019992799622002783,
+      "loss": 20.2673,
+      "step": 716
+    },
+    {
+      "epoch": 0.12537704918032788,
+      "grad_norm": 1.3205544344250497,
+      "learning_rate": 0.0001999272780693434,
+      "loss": 20.3886,
+      "step": 717
+    },
+    {
+      "epoch": 0.125551912568306,
+      "grad_norm": 1.3349505129228918,
+      "learning_rate": 0.00019992655635636824,
+      "loss": 20.5046,
+      "step": 718
+    },
+    {
+      "epoch": 0.12572677595628415,
+      "grad_norm": 1.2452320965860522,
+      "learning_rate": 0.0001999258310811281,
+      "loss": 20.6643,
+      "step": 719
+    },
+    {
+      "epoch": 0.1259016393442623,
+      "grad_norm": 1.3618274164911943,
+      "learning_rate": 0.0001999251022436488,
+      "loss": 20.8388,
+      "step": 720
+    },
+    {
+      "epoch": 0.12607650273224044,
+      "grad_norm": 1.3155629157343025,
+      "learning_rate": 0.00019992436984395638,
+      "loss": 20.6447,
+      "step": 721
+    },
+    {
+      "epoch": 0.1262513661202186,
+      "grad_norm": 1.3639599991215077,
+      "learning_rate": 0.00019992363388207694,
+      "loss": 20.5149,
+      "step": 722
+    },
+    {
+      "epoch": 0.1264262295081967,
+      "grad_norm": 1.3407504819253124,
+      "learning_rate": 0.0001999228943580367,
+      "loss": 20.7141,
+      "step": 723
+    },
+    {
+      "epoch": 0.12660109289617485,
+      "grad_norm": 1.3408486312964114,
+      "learning_rate": 0.000199922151271862,
+      "loss": 20.2581,
+      "step": 724
+    },
+    {
+      "epoch": 0.126775956284153,
+      "grad_norm": 1.487800112247374,
+      "learning_rate": 0.00019992140462357936,
+      "loss": 20.6269,
+      "step": 725
+    },
+    {
+      "epoch": 0.12695081967213115,
+      "grad_norm": 1.6632384961247015,
+      "learning_rate": 0.0001999206544132154,
+      "loss": 20.7322,
+      "step": 726
+    },
+    {
+      "epoch": 0.1271256830601093,
+      "grad_norm": 1.3930035741089706,
+      "learning_rate": 0.00019991990064079687,
+      "loss": 20.3643,
+      "step": 727
+    },
+    {
+      "epoch": 0.12730054644808744,
+      "grad_norm": 1.4098288477847323,
+      "learning_rate": 0.0001999191433063506,
+      "loss": 20.7929,
+      "step": 728
+    },
+    {
+      "epoch": 0.12747540983606556,
+      "grad_norm": 1.4486219639109712,
+      "learning_rate": 0.00019991838240990362,
+      "loss": 20.5736,
+      "step": 729
+    },
+    {
+      "epoch": 0.1276502732240437,
+      "grad_norm": 1.3404003987272428,
+      "learning_rate": 0.00019991761795148304,
+      "loss": 20.4523,
+      "step": 730
+    },
+    {
+      "epoch": 0.12782513661202186,
+      "grad_norm": 1.4928674484888211,
+      "learning_rate": 0.00019991684993111614,
+      "loss": 20.2982,
+      "step": 731
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 1.386664282789252,
+      "learning_rate": 0.0001999160783488303,
+      "loss": 20.7081,
+      "step": 732
+    },
+    {
+      "epoch": 0.12817486338797815,
+      "grad_norm": 1.3155731375623738,
+      "learning_rate": 0.00019991530320465298,
+      "loss": 20.4899,
+      "step": 733
+    },
+    {
+      "epoch": 0.1283497267759563,
+      "grad_norm": 1.4508008445640799,
+      "learning_rate": 0.00019991452449861182,
+      "loss": 20.7206,
+      "step": 734
+    },
+    {
+      "epoch": 0.12852459016393442,
+      "grad_norm": 1.43440718120707,
+      "learning_rate": 0.00019991374223073463,
+      "loss": 20.5938,
+      "step": 735
+    },
+    {
+      "epoch": 0.12869945355191256,
+      "grad_norm": 1.4040737726682895,
+      "learning_rate": 0.00019991295640104924,
+      "loss": 20.6732,
+      "step": 736
+    },
+    {
+      "epoch": 0.1288743169398907,
+      "grad_norm": 1.2981595242492414,
+      "learning_rate": 0.00019991216700958369,
+      "loss": 20.4253,
+      "step": 737
+    },
+    {
+      "epoch": 0.12904918032786886,
+      "grad_norm": 1.53723524150373,
+      "learning_rate": 0.00019991137405636614,
+      "loss": 20.618,
+      "step": 738
+    },
+    {
+      "epoch": 0.129224043715847,
+      "grad_norm": 1.4393705720050005,
+      "learning_rate": 0.00019991057754142484,
+      "loss": 20.7256,
+      "step": 739
+    },
+    {
+      "epoch": 0.12939890710382515,
+      "grad_norm": 1.5177332592799948,
+      "learning_rate": 0.00019990977746478817,
+      "loss": 20.4345,
+      "step": 740
+    },
+    {
+      "epoch": 0.12957377049180327,
+      "grad_norm": 1.339063322312335,
+      "learning_rate": 0.00019990897382648467,
+      "loss": 20.4322,
+      "step": 741
+    },
+    {
+      "epoch": 0.12974863387978142,
+      "grad_norm": 1.2753124993160665,
+      "learning_rate": 0.000199908166626543,
+      "loss": 20.2751,
+      "step": 742
+    },
+    {
+      "epoch": 0.12992349726775956,
+      "grad_norm": 1.376390169988682,
+      "learning_rate": 0.0001999073558649919,
+      "loss": 20.5172,
+      "step": 743
+    },
+    {
+      "epoch": 0.1300983606557377,
+      "grad_norm": 1.301660637744646,
+      "learning_rate": 0.0001999065415418603,
+      "loss": 20.5115,
+      "step": 744
+    },
+    {
+      "epoch": 0.13027322404371586,
+      "grad_norm": 1.2855953834809757,
+      "learning_rate": 0.0001999057236571772,
+      "loss": 20.385,
+      "step": 745
+    },
+    {
+      "epoch": 0.13044808743169398,
+      "grad_norm": 1.3202912336288968,
+      "learning_rate": 0.00019990490221097183,
+      "loss": 20.0174,
+      "step": 746
+    },
+    {
+      "epoch": 0.13062295081967212,
+      "grad_norm": 1.3809839766554488,
+      "learning_rate": 0.00019990407720327339,
+      "loss": 20.6165,
+      "step": 747
+    },
+    {
+      "epoch": 0.13079781420765027,
+      "grad_norm": 1.293321804664792,
+      "learning_rate": 0.00019990324863411132,
+      "loss": 20.4854,
+      "step": 748
+    },
+    {
+      "epoch": 0.13097267759562842,
+      "grad_norm": 1.4215677753284295,
+      "learning_rate": 0.00019990241650351517,
+      "loss": 20.5352,
+      "step": 749
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 1.3600915386428387,
+      "learning_rate": 0.0001999015808115146,
+      "loss": 20.5937,
+      "step": 750
+    },
+    {
+      "epoch": 0.1313224043715847,
+      "grad_norm": 1.46994263167156,
+      "learning_rate": 0.00019990074155813936,
+      "loss": 20.6392,
+      "step": 751
+    },
+    {
+      "epoch": 0.13149726775956283,
+      "grad_norm": 1.2912988159423264,
+      "learning_rate": 0.00019989989874341944,
+      "loss": 20.7074,
+      "step": 752
+    },
+    {
+      "epoch": 0.13167213114754098,
+      "grad_norm": 1.4610704787703697,
+      "learning_rate": 0.00019989905236738483,
+      "loss": 20.8952,
+      "step": 753
+    },
+    {
+      "epoch": 0.13184699453551912,
+      "grad_norm": 1.4626117423687155,
+      "learning_rate": 0.00019989820243006574,
+      "loss": 20.123,
+      "step": 754
+    },
+    {
+      "epoch": 0.13202185792349727,
+      "grad_norm": 1.2497978947177966,
+      "learning_rate": 0.00019989734893149247,
+      "loss": 20.1232,
+      "step": 755
+    },
+    {
+      "epoch": 0.13219672131147542,
+      "grad_norm": 1.475726203357079,
+      "learning_rate": 0.00019989649187169542,
+      "loss": 20.1972,
+      "step": 756
+    },
+    {
+      "epoch": 0.13237158469945356,
+      "grad_norm": 1.3644776069487643,
+      "learning_rate": 0.00019989563125070512,
+      "loss": 20.3674,
+      "step": 757
+    },
+    {
+      "epoch": 0.13254644808743168,
+      "grad_norm": 1.3781378403337121,
+      "learning_rate": 0.0001998947670685523,
+      "loss": 20.4122,
+      "step": 758
+    },
+    {
+      "epoch": 0.13272131147540983,
+      "grad_norm": 1.2343661237627714,
+      "learning_rate": 0.00019989389932526774,
+      "loss": 20.6108,
+      "step": 759
+    },
+    {
+      "epoch": 0.13289617486338798,
+      "grad_norm": 1.3417324081687187,
+      "learning_rate": 0.00019989302802088239,
+      "loss": 20.2597,
+      "step": 760
+    },
+    {
+      "epoch": 0.13307103825136612,
+      "grad_norm": 1.289932633347394,
+      "learning_rate": 0.0001998921531554273,
+      "loss": 20.5087,
+      "step": 761
+    },
+    {
+      "epoch": 0.13324590163934427,
+      "grad_norm": 1.278470999542162,
+      "learning_rate": 0.00019989127472893364,
+      "loss": 20.4325,
+      "step": 762
+    },
+    {
+      "epoch": 0.1334207650273224,
+      "grad_norm": 1.1892211346549322,
+      "learning_rate": 0.00019989039274143279,
+      "loss": 20.4155,
+      "step": 763
+    },
+    {
+      "epoch": 0.13359562841530054,
+      "grad_norm": 1.3196603936290423,
+      "learning_rate": 0.00019988950719295614,
+      "loss": 20.6831,
+      "step": 764
+    },
+    {
+      "epoch": 0.13377049180327868,
+      "grad_norm": 1.2759085635573113,
+      "learning_rate": 0.00019988861808353526,
+      "loss": 20.5908,
+      "step": 765
+    },
+    {
+      "epoch": 0.13394535519125683,
+      "grad_norm": 1.1524509225244133,
+      "learning_rate": 0.00019988772541320183,
+      "loss": 20.5321,
+      "step": 766
+    },
+    {
+      "epoch": 0.13412021857923498,
+      "grad_norm": 1.2589090287685165,
+      "learning_rate": 0.00019988682918198772,
+      "loss": 20.2243,
+      "step": 767
+    },
+    {
+      "epoch": 0.13429508196721313,
+      "grad_norm": 1.264423114833383,
+      "learning_rate": 0.00019988592938992484,
+      "loss": 20.6382,
+      "step": 768
+    },
+    {
+      "epoch": 0.13446994535519124,
+      "grad_norm": 1.1916938794877043,
+      "learning_rate": 0.0001998850260370453,
+      "loss": 20.3778,
+      "step": 769
+    },
+    {
+      "epoch": 0.1346448087431694,
+      "grad_norm": 1.286751047187731,
+      "learning_rate": 0.00019988411912338127,
+      "loss": 20.2242,
+      "step": 770
+    },
+    {
+      "epoch": 0.13481967213114754,
+      "grad_norm": 1.2880350700110053,
+      "learning_rate": 0.00019988320864896511,
+      "loss": 20.4999,
+      "step": 771
+    },
+    {
+      "epoch": 0.13499453551912569,
+      "grad_norm": 1.2505164776041109,
+      "learning_rate": 0.00019988229461382926,
+      "loss": 20.46,
+      "step": 772
+    },
+    {
+      "epoch": 0.13516939890710383,
+      "grad_norm": 1.3679541086138753,
+      "learning_rate": 0.00019988137701800628,
+      "loss": 20.518,
+      "step": 773
+    },
+    {
+      "epoch": 0.13534426229508198,
+      "grad_norm": 1.3893850643564896,
+      "learning_rate": 0.00019988045586152894,
+      "loss": 20.171,
+      "step": 774
+    },
+    {
+      "epoch": 0.1355191256830601,
+      "grad_norm": 1.1922624516496092,
+      "learning_rate": 0.00019987953114443004,
+      "loss": 19.8321,
+      "step": 775
+    },
+    {
+      "epoch": 0.13569398907103825,
+      "grad_norm": 1.2457620799719489,
+      "learning_rate": 0.00019987860286674254,
+      "loss": 20.1735,
+      "step": 776
+    },
+    {
+      "epoch": 0.1358688524590164,
+      "grad_norm": 1.316119704414435,
+      "learning_rate": 0.00019987767102849955,
+      "loss": 19.9576,
+      "step": 777
+    },
+    {
+      "epoch": 0.13604371584699454,
+      "grad_norm": 1.2076941839331776,
+      "learning_rate": 0.00019987673562973427,
+      "loss": 20.1364,
+      "step": 778
+    },
+    {
+      "epoch": 0.1362185792349727,
+      "grad_norm": 1.3308618719803478,
+      "learning_rate": 0.0001998757966704801,
+      "loss": 20.2638,
+      "step": 779
+    },
+    {
+      "epoch": 0.13639344262295083,
+      "grad_norm": 1.2594682336315988,
+      "learning_rate": 0.00019987485415077044,
+      "loss": 20.3756,
+      "step": 780
+    },
+    {
+      "epoch": 0.13656830601092895,
+      "grad_norm": 1.236309086848712,
+      "learning_rate": 0.0001998739080706389,
+      "loss": 20.0025,
+      "step": 781
+    },
+    {
+      "epoch": 0.1367431693989071,
+      "grad_norm": 1.2777583870638514,
+      "learning_rate": 0.00019987295843011923,
+      "loss": 20.1856,
+      "step": 782
+    },
+    {
+      "epoch": 0.13691803278688525,
+      "grad_norm": 1.2895783021009863,
+      "learning_rate": 0.00019987200522924531,
+      "loss": 20.425,
+      "step": 783
+    },
+    {
+      "epoch": 0.1370928961748634,
+      "grad_norm": 1.3197848005357413,
+      "learning_rate": 0.0001998710484680511,
+      "loss": 20.383,
+      "step": 784
+    },
+    {
+      "epoch": 0.13726775956284154,
+      "grad_norm": 1.4408170821192636,
+      "learning_rate": 0.00019987008814657066,
+      "loss": 20.0618,
+      "step": 785
+    },
+    {
+      "epoch": 0.13744262295081966,
+      "grad_norm": 1.272530487556067,
+      "learning_rate": 0.0001998691242648383,
+      "loss": 20.5229,
+      "step": 786
+    },
+    {
+      "epoch": 0.1376174863387978,
+      "grad_norm": 1.3450079189131052,
+      "learning_rate": 0.00019986815682288833,
+      "loss": 20.3475,
+      "step": 787
+    },
+    {
+      "epoch": 0.13779234972677595,
+      "grad_norm": 1.61355114723381,
+      "learning_rate": 0.00019986718582075524,
+      "loss": 20.642,
+      "step": 788
+    },
+    {
+      "epoch": 0.1379672131147541,
+      "grad_norm": 1.7486805450952867,
+      "learning_rate": 0.00019986621125847367,
+      "loss": 20.5306,
+      "step": 789
+    },
+    {
+      "epoch": 0.13814207650273225,
+      "grad_norm": 1.3746379555158637,
+      "learning_rate": 0.00019986523313607837,
+      "loss": 20.4321,
+      "step": 790
+    },
+    {
+      "epoch": 0.1383169398907104,
+      "grad_norm": 1.3589934860926685,
+      "learning_rate": 0.00019986425145360418,
+      "loss": 20.295,
+      "step": 791
+    },
+    {
+      "epoch": 0.1384918032786885,
+      "grad_norm": 1.5864057496369561,
+      "learning_rate": 0.00019986326621108607,
+      "loss": 20.5492,
+      "step": 792
+    },
+    {
+      "epoch": 0.13866666666666666,
+      "grad_norm": 1.369256977439755,
+      "learning_rate": 0.00019986227740855924,
+      "loss": 20.3518,
+      "step": 793
+    },
+    {
+      "epoch": 0.1388415300546448,
+      "grad_norm": 1.4147980701609055,
+      "learning_rate": 0.00019986128504605885,
+      "loss": 20.2384,
+      "step": 794
+    },
+    {
+      "epoch": 0.13901639344262295,
+      "grad_norm": 1.5284609573170043,
+      "learning_rate": 0.00019986028912362036,
+      "loss": 20.4406,
+      "step": 795
+    },
+    {
+      "epoch": 0.1391912568306011,
+      "grad_norm": 1.3145395052145303,
+      "learning_rate": 0.00019985928964127923,
+      "loss": 20.2957,
+      "step": 796
+    },
+    {
+      "epoch": 0.13936612021857925,
+      "grad_norm": 1.381281717206995,
+      "learning_rate": 0.00019985828659907108,
+      "loss": 19.8987,
+      "step": 797
+    },
+    {
+      "epoch": 0.13954098360655737,
+      "grad_norm": 1.5172080474247256,
+      "learning_rate": 0.0001998572799970317,
+      "loss": 20.2627,
+      "step": 798
+    },
+    {
+      "epoch": 0.13971584699453551,
+      "grad_norm": 1.3682923060937577,
+      "learning_rate": 0.00019985626983519696,
+      "loss": 20.6542,
+      "step": 799
+    },
+    {
+      "epoch": 0.13989071038251366,
+      "grad_norm": 1.3822147920446548,
+      "learning_rate": 0.00019985525611360285,
+      "loss": 20.3927,
+      "step": 800
+    },
+    {
+      "epoch": 0.1400655737704918,
+      "grad_norm": 1.4792369122270443,
+      "learning_rate": 0.00019985423883228556,
+      "loss": 20.1639,
+      "step": 801
+    },
+    {
+      "epoch": 0.14024043715846995,
+      "grad_norm": 1.3011253185776044,
+      "learning_rate": 0.00019985321799128133,
+      "loss": 20.1262,
+      "step": 802
+    },
+    {
+      "epoch": 0.14041530054644807,
+      "grad_norm": 1.348865541274003,
+      "learning_rate": 0.0001998521935906265,
+      "loss": 20.26,
+      "step": 803
+    },
+    {
+      "epoch": 0.14059016393442622,
+      "grad_norm": 1.2840794257946866,
+      "learning_rate": 0.00019985116563035765,
+      "loss": 20.4856,
+      "step": 804
+    },
+    {
+      "epoch": 0.14076502732240437,
+      "grad_norm": 1.3551702209673844,
+      "learning_rate": 0.0001998501341105114,
+      "loss": 20.2278,
+      "step": 805
+    },
+    {
+      "epoch": 0.14093989071038251,
+      "grad_norm": 1.2695256942954227,
+      "learning_rate": 0.00019984909903112457,
+      "loss": 20.1948,
+      "step": 806
+    },
+    {
+      "epoch": 0.14111475409836066,
+      "grad_norm": 1.3448563278886352,
+      "learning_rate": 0.000199848060392234,
+      "loss": 20.3823,
+      "step": 807
+    },
+    {
+      "epoch": 0.1412896174863388,
+      "grad_norm": 1.2257845791651993,
+      "learning_rate": 0.00019984701819387673,
+      "loss": 20.2385,
+      "step": 808
+    },
+    {
+      "epoch": 0.14146448087431693,
+      "grad_norm": 1.3094096934157324,
+      "learning_rate": 0.00019984597243608994,
+      "loss": 20.543,
+      "step": 809
+    },
+    {
+      "epoch": 0.14163934426229507,
+      "grad_norm": 1.3083484696227599,
+      "learning_rate": 0.00019984492311891087,
+      "loss": 20.2053,
+      "step": 810
+    },
+    {
+      "epoch": 0.14181420765027322,
+      "grad_norm": 1.2119730181510122,
+      "learning_rate": 0.00019984387024237692,
+      "loss": 20.0746,
+      "step": 811
+    },
+    {
+      "epoch": 0.14198907103825137,
+      "grad_norm": 1.32481654375634,
+      "learning_rate": 0.0001998428138065257,
+      "loss": 20.2476,
+      "step": 812
+    },
+    {
+      "epoch": 0.14216393442622952,
+      "grad_norm": 1.2520574242064844,
+      "learning_rate": 0.00019984175381139481,
+      "loss": 20.4532,
+      "step": 813
+    },
+    {
+      "epoch": 0.14233879781420766,
+      "grad_norm": 1.3152856683975311,
+      "learning_rate": 0.00019984069025702204,
+      "loss": 19.8072,
+      "step": 814
+    },
+    {
+      "epoch": 0.14251366120218578,
+      "grad_norm": 1.2060586630634185,
+      "learning_rate": 0.00019983962314344531,
+      "loss": 19.8887,
+      "step": 815
+    },
+    {
+      "epoch": 0.14268852459016393,
+      "grad_norm": 1.2252391882504432,
+      "learning_rate": 0.00019983855247070266,
+      "loss": 20.0243,
+      "step": 816
+    },
+    {
+      "epoch": 0.14286338797814208,
+      "grad_norm": 1.2877005052773716,
+      "learning_rate": 0.0001998374782388323,
+      "loss": 20.0781,
+      "step": 817
+    },
+    {
+      "epoch": 0.14303825136612022,
+      "grad_norm": 1.2788213304535323,
+      "learning_rate": 0.00019983640044787243,
+      "loss": 20.1346,
+      "step": 818
+    },
+    {
+      "epoch": 0.14321311475409837,
+      "grad_norm": 1.2578860870357333,
+      "learning_rate": 0.00019983531909786156,
+      "loss": 19.817,
+      "step": 819
+    },
+    {
+      "epoch": 0.14338797814207652,
+      "grad_norm": 1.24889166592984,
+      "learning_rate": 0.00019983423418883821,
+      "loss": 20.6021,
+      "step": 820
+    },
+    {
+      "epoch": 0.14356284153005464,
+      "grad_norm": 1.2237368817469831,
+      "learning_rate": 0.00019983314572084106,
+      "loss": 19.8449,
+      "step": 821
+    },
+    {
+      "epoch": 0.14373770491803278,
+      "grad_norm": 1.3395249596156966,
+      "learning_rate": 0.00019983205369390893,
+      "loss": 20.2493,
+      "step": 822
+    },
+    {
+      "epoch": 0.14391256830601093,
+      "grad_norm": 1.283033182621752,
+      "learning_rate": 0.0001998309581080807,
+      "loss": 20.0856,
+      "step": 823
+    },
+    {
+      "epoch": 0.14408743169398908,
+      "grad_norm": 1.2191302617524613,
+      "learning_rate": 0.00019982985896339542,
+      "loss": 20.6192,
+      "step": 824
+    },
+    {
+      "epoch": 0.14426229508196722,
+      "grad_norm": 1.319803472543724,
+      "learning_rate": 0.00019982875625989237,
+      "loss": 19.7639,
+      "step": 825
+    },
+    {
+      "epoch": 0.14443715846994534,
+      "grad_norm": 1.3998644662515127,
+      "learning_rate": 0.00019982764999761077,
+      "loss": 20.2052,
+      "step": 826
+    },
+    {
+      "epoch": 0.1446120218579235,
+      "grad_norm": 1.3000497429265612,
+      "learning_rate": 0.00019982654017659007,
+      "loss": 19.7027,
+      "step": 827
+    },
+    {
+      "epoch": 0.14478688524590164,
+      "grad_norm": 1.2525306086193309,
+      "learning_rate": 0.00019982542679686986,
+      "loss": 19.9938,
+      "step": 828
+    },
+    {
+      "epoch": 0.14496174863387978,
+      "grad_norm": 1.3562633076946946,
+      "learning_rate": 0.00019982430985848984,
+      "loss": 19.9615,
+      "step": 829
+    },
+    {
+      "epoch": 0.14513661202185793,
+      "grad_norm": 1.2467775462313069,
+      "learning_rate": 0.00019982318936148977,
+      "loss": 20.488,
+      "step": 830
+    },
+    {
+      "epoch": 0.14531147540983608,
+      "grad_norm": 1.3806341375819082,
+      "learning_rate": 0.00019982206530590966,
+      "loss": 20.0818,
+      "step": 831
+    },
+    {
+      "epoch": 0.1454863387978142,
+      "grad_norm": 1.2018680423604182,
+      "learning_rate": 0.00019982093769178953,
+      "loss": 20.2856,
+      "step": 832
+    },
+    {
+      "epoch": 0.14566120218579234,
+      "grad_norm": 1.460549753616387,
+      "learning_rate": 0.0001998198065191696,
+      "loss": 20.4271,
+      "step": 833
+    },
+    {
+      "epoch": 0.1458360655737705,
+      "grad_norm": 1.2798245028252804,
+      "learning_rate": 0.0001998186717880902,
+      "loss": 20.067,
+      "step": 834
+    },
+    {
+      "epoch": 0.14601092896174864,
+      "grad_norm": 1.5167081864512044,
+      "learning_rate": 0.00019981753349859177,
+      "loss": 20.3034,
+      "step": 835
+    },
+    {
+      "epoch": 0.14618579234972678,
+      "grad_norm": 1.428926968826349,
+      "learning_rate": 0.0001998163916507149,
+      "loss": 20.2478,
+      "step": 836
+    },
+    {
+      "epoch": 0.14636065573770493,
+      "grad_norm": 1.2725960889035437,
+      "learning_rate": 0.0001998152462445003,
+      "loss": 20.0183,
+      "step": 837
+    },
+    {
+      "epoch": 0.14653551912568305,
+      "grad_norm": 1.3722204502386346,
+      "learning_rate": 0.00019981409727998876,
+      "loss": 19.9968,
+      "step": 838
+    },
+    {
+      "epoch": 0.1467103825136612,
+      "grad_norm": 1.477616691660844,
+      "learning_rate": 0.00019981294475722133,
+      "loss": 20.0477,
+      "step": 839
+    },
+    {
+      "epoch": 0.14688524590163934,
+      "grad_norm": 1.3736988747985228,
+      "learning_rate": 0.000199811788676239,
+      "loss": 20.2295,
+      "step": 840
+    },
+    {
+      "epoch": 0.1470601092896175,
+      "grad_norm": 1.2805633563563807,
+      "learning_rate": 0.000199810629037083,
+      "loss": 20.2939,
+      "step": 841
+    },
+    {
+      "epoch": 0.14723497267759564,
+      "grad_norm": 1.3097219692053876,
+      "learning_rate": 0.00019980946583979473,
+      "loss": 20.1307,
+      "step": 842
+    },
+    {
+      "epoch": 0.14740983606557376,
+      "grad_norm": 1.463789373433653,
+      "learning_rate": 0.0001998082990844156,
+      "loss": 20.3019,
+      "step": 843
+    },
+    {
+      "epoch": 0.1475846994535519,
+      "grad_norm": 1.3296266727960115,
+      "learning_rate": 0.00019980712877098723,
+      "loss": 20.1528,
+      "step": 844
+    },
+    {
+      "epoch": 0.14775956284153005,
+      "grad_norm": 1.3783632583019003,
+      "learning_rate": 0.0001998059548995513,
+      "loss": 20.192,
+      "step": 845
+    },
+    {
+      "epoch": 0.1479344262295082,
+      "grad_norm": 1.2070144730947703,
+      "learning_rate": 0.00019980477747014974,
+      "loss": 20.194,
+      "step": 846
+    },
+    {
+      "epoch": 0.14810928961748634,
+      "grad_norm": 1.508133701638163,
+      "learning_rate": 0.00019980359648282445,
+      "loss": 19.9979,
+      "step": 847
+    },
+    {
+      "epoch": 0.1482841530054645,
+      "grad_norm": 1.2801450518273803,
+      "learning_rate": 0.00019980241193761754,
+      "loss": 20.0348,
+      "step": 848
+    },
+    {
+      "epoch": 0.1484590163934426,
+      "grad_norm": 1.357308912628944,
+      "learning_rate": 0.00019980122383457126,
+      "loss": 20.1357,
+      "step": 849
+    },
+    {
+      "epoch": 0.14863387978142076,
+      "grad_norm": 1.3101012339041351,
+      "learning_rate": 0.000199800032173728,
+      "loss": 19.9805,
+      "step": 850
+    },
+    {
+      "epoch": 0.1488087431693989,
+      "grad_norm": 1.337462404258618,
+      "learning_rate": 0.00019979883695513012,
+      "loss": 20.1067,
+      "step": 851
+    },
+    {
+      "epoch": 0.14898360655737705,
+      "grad_norm": 1.325299070465483,
+      "learning_rate": 0.00019979763817882035,
+      "loss": 20.135,
+      "step": 852
+    },
+    {
+      "epoch": 0.1491584699453552,
+      "grad_norm": 1.3157959062959075,
+      "learning_rate": 0.0001997964358448414,
+      "loss": 19.8901,
+      "step": 853
+    },
+    {
+      "epoch": 0.14933333333333335,
+      "grad_norm": 1.3201867664912597,
+      "learning_rate": 0.00019979522995323604,
+      "loss": 20.5077,
+      "step": 854
+    },
+    {
+      "epoch": 0.14950819672131146,
+      "grad_norm": 1.279076210306007,
+      "learning_rate": 0.0001997940205040474,
+      "loss": 20.3993,
+      "step": 855
+    },
+    {
+      "epoch": 0.1496830601092896,
+      "grad_norm": 1.3930440771306882,
+      "learning_rate": 0.00019979280749731848,
+      "loss": 20.1275,
+      "step": 856
+    },
+    {
+      "epoch": 0.14985792349726776,
+      "grad_norm": 1.223425418622535,
+      "learning_rate": 0.00019979159093309258,
+      "loss": 20.1217,
+      "step": 857
+    },
+    {
+      "epoch": 0.1500327868852459,
+      "grad_norm": 1.3059989833207233,
+      "learning_rate": 0.00019979037081141308,
+      "loss": 20.2336,
+      "step": 858
+    },
+    {
+      "epoch": 0.15020765027322405,
+      "grad_norm": 1.3720828388085469,
+      "learning_rate": 0.00019978914713232344,
+      "loss": 20.0936,
+      "step": 859
+    },
+    {
+      "epoch": 0.1503825136612022,
+      "grad_norm": 1.3217169810258282,
+      "learning_rate": 0.0001997879198958673,
+      "loss": 20.2432,
+      "step": 860
+    },
+    {
+      "epoch": 0.15055737704918032,
+      "grad_norm": 1.2974071333101287,
+      "learning_rate": 0.00019978668910208838,
+      "loss": 20.305,
+      "step": 861
+    },
+    {
+      "epoch": 0.15073224043715847,
+      "grad_norm": 1.3645295597349223,
+      "learning_rate": 0.0001997854547510306,
+      "loss": 20.001,
+      "step": 862
+    },
+    {
+      "epoch": 0.1509071038251366,
+      "grad_norm": 1.3330755626063477,
+      "learning_rate": 0.00019978421684273794,
+      "loss": 19.9367,
+      "step": 863
+    },
+    {
+      "epoch": 0.15108196721311476,
+      "grad_norm": 1.2394584783397085,
+      "learning_rate": 0.00019978297537725452,
+      "loss": 20.2882,
+      "step": 864
+    },
+    {
+      "epoch": 0.1512568306010929,
+      "grad_norm": 1.2366590412082268,
+      "learning_rate": 0.00019978173035462463,
+      "loss": 19.9031,
+      "step": 865
+    },
+    {
+      "epoch": 0.15143169398907103,
+      "grad_norm": 1.2494870662950297,
+      "learning_rate": 0.00019978048177489265,
+      "loss": 20.3026,
+      "step": 866
+    },
+    {
+      "epoch": 0.15160655737704917,
+      "grad_norm": 1.3503021658522985,
+      "learning_rate": 0.00019977922963810307,
+      "loss": 19.7149,
+      "step": 867
+    },
+    {
+      "epoch": 0.15178142076502732,
+      "grad_norm": 1.277524157810039,
+      "learning_rate": 0.0001997779739443005,
+      "loss": 20.2412,
+      "step": 868
+    },
+    {
+      "epoch": 0.15195628415300547,
+      "grad_norm": 1.4197935251160216,
+      "learning_rate": 0.0001997767146935298,
+      "loss": 20.3355,
+      "step": 869
+    },
+    {
+      "epoch": 0.1521311475409836,
+      "grad_norm": 1.5365990375905976,
+      "learning_rate": 0.00019977545188583573,
+      "loss": 19.9331,
+      "step": 870
+    },
+    {
+      "epoch": 0.15230601092896176,
+      "grad_norm": 1.1865669252646374,
+      "learning_rate": 0.00019977418552126342,
+      "loss": 19.9512,
+      "step": 871
+    },
+    {
+      "epoch": 0.15248087431693988,
+      "grad_norm": 1.3482496130108197,
+      "learning_rate": 0.00019977291559985795,
+      "loss": 19.5843,
+      "step": 872
+    },
+    {
+      "epoch": 0.15265573770491803,
+      "grad_norm": 1.3850788671103205,
+      "learning_rate": 0.0001997716421216646,
+      "loss": 20.2252,
+      "step": 873
+    },
+    {
+      "epoch": 0.15283060109289617,
+      "grad_norm": 1.3120842829633357,
+      "learning_rate": 0.00019977036508672883,
+      "loss": 19.742,
+      "step": 874
+    },
+    {
+      "epoch": 0.15300546448087432,
+      "grad_norm": 1.3591434485444065,
+      "learning_rate": 0.00019976908449509606,
+      "loss": 19.7383,
+      "step": 875
+    },
+    {
+      "epoch": 0.15318032786885247,
+      "grad_norm": 1.4634503027553556,
+      "learning_rate": 0.00019976780034681202,
+      "loss": 19.5963,
+      "step": 876
+    },
+    {
+      "epoch": 0.15335519125683061,
+      "grad_norm": 1.3050445521224892,
+      "learning_rate": 0.00019976651264192247,
+      "loss": 20.1209,
+      "step": 877
+    },
+    {
+      "epoch": 0.15353005464480873,
+      "grad_norm": 1.2224864727294524,
+      "learning_rate": 0.00019976522138047328,
+      "loss": 20.492,
+      "step": 878
+    },
+    {
+      "epoch": 0.15370491803278688,
+      "grad_norm": 1.379863527121902,
+      "learning_rate": 0.00019976392656251052,
+      "loss": 20.0967,
+      "step": 879
+    },
+    {
+      "epoch": 0.15387978142076503,
+      "grad_norm": 1.2614172566505264,
+      "learning_rate": 0.00019976262818808036,
+      "loss": 19.9446,
+      "step": 880
+    },
+    {
+      "epoch": 0.15405464480874317,
+      "grad_norm": 1.2802221606186543,
+      "learning_rate": 0.00019976132625722906,
+      "loss": 19.8779,
+      "step": 881
+    },
+    {
+      "epoch": 0.15422950819672132,
+      "grad_norm": 1.4709328648757334,
+      "learning_rate": 0.00019976002077000298,
+      "loss": 20.0297,
+      "step": 882
+    },
+    {
+      "epoch": 0.15440437158469944,
+      "grad_norm": 1.5226619840040578,
+      "learning_rate": 0.0001997587117264488,
+      "loss": 20.0354,
+      "step": 883
+    },
+    {
+      "epoch": 0.1545792349726776,
+      "grad_norm": 1.178099673967306,
+      "learning_rate": 0.00019975739912661306,
+      "loss": 20.2382,
+      "step": 884
+    },
+    {
+      "epoch": 0.15475409836065573,
+      "grad_norm": 1.2956414927535658,
+      "learning_rate": 0.00019975608297054262,
+      "loss": 19.7736,
+      "step": 885
+    },
+    {
+      "epoch": 0.15492896174863388,
+      "grad_norm": 1.4293574079481088,
+      "learning_rate": 0.00019975476325828436,
+      "loss": 19.8823,
+      "step": 886
+    },
+    {
+      "epoch": 0.15510382513661203,
+      "grad_norm": 1.371783919476702,
+      "learning_rate": 0.00019975343998988532,
+      "loss": 19.8214,
+      "step": 887
+    },
+    {
+      "epoch": 0.15527868852459017,
+      "grad_norm": 1.3261618001580415,
+      "learning_rate": 0.00019975211316539273,
+      "loss": 19.9993,
+      "step": 888
+    },
+    {
+      "epoch": 0.1554535519125683,
+      "grad_norm": 1.4085054078546397,
+      "learning_rate": 0.00019975078278485384,
+      "loss": 20.0074,
+      "step": 889
+    },
+    {
+      "epoch": 0.15562841530054644,
+      "grad_norm": 1.3143989991824803,
+      "learning_rate": 0.0001997494488483161,
+      "loss": 19.66,
+      "step": 890
+    },
+    {
+      "epoch": 0.1558032786885246,
+      "grad_norm": 1.485910885371283,
+      "learning_rate": 0.000199748111355827,
+      "loss": 20.084,
+      "step": 891
+    },
+    {
+      "epoch": 0.15597814207650273,
+      "grad_norm": 1.369685838876942,
+      "learning_rate": 0.00019974677030743432,
+      "loss": 19.5451,
+      "step": 892
+    },
+    {
+      "epoch": 0.15615300546448088,
+      "grad_norm": 1.4360089861976701,
+      "learning_rate": 0.00019974542570318582,
+      "loss": 19.8346,
+      "step": 893
+    },
+    {
+      "epoch": 0.15632786885245903,
+      "grad_norm": 1.2952951967398003,
+      "learning_rate": 0.00019974407754312943,
+      "loss": 19.8943,
+      "step": 894
+    },
+    {
+      "epoch": 0.15650273224043715,
+      "grad_norm": 1.3008822934203657,
+      "learning_rate": 0.00019974272582731322,
+      "loss": 19.9117,
+      "step": 895
+    },
+    {
+      "epoch": 0.1566775956284153,
+      "grad_norm": 1.2916909920831638,
+      "learning_rate": 0.00019974137055578535,
+      "loss": 19.8287,
+      "step": 896
+    },
+    {
+      "epoch": 0.15685245901639344,
+      "grad_norm": 1.2986742745348634,
+      "learning_rate": 0.00019974001172859418,
+      "loss": 19.9456,
+      "step": 897
+    },
+    {
+      "epoch": 0.1570273224043716,
+      "grad_norm": 1.3260987923541057,
+      "learning_rate": 0.0001997386493457881,
+      "loss": 19.9288,
+      "step": 898
+    },
+    {
+      "epoch": 0.15720218579234974,
+      "grad_norm": 1.3274645971760635,
+      "learning_rate": 0.00019973728340741574,
+      "loss": 20.0516,
+      "step": 899
+    },
+    {
+      "epoch": 0.15737704918032788,
+      "grad_norm": 1.301260279828156,
+      "learning_rate": 0.0001997359139135257,
+      "loss": 19.5245,
+      "step": 900
+    },
+    {
+      "epoch": 0.157551912568306,
+      "grad_norm": 1.350651673955518,
+      "learning_rate": 0.0001997345408641669,
+      "loss": 19.8333,
+      "step": 901
+    },
+    {
+      "epoch": 0.15772677595628415,
+      "grad_norm": 1.3776501472130478,
+      "learning_rate": 0.00019973316425938824,
+      "loss": 19.9707,
+      "step": 902
+    },
+    {
+      "epoch": 0.1579016393442623,
+      "grad_norm": 1.4893982688906977,
+      "learning_rate": 0.0001997317840992388,
+      "loss": 20.2676,
+      "step": 903
+    },
+    {
+      "epoch": 0.15807650273224044,
+      "grad_norm": 1.6185032000042556,
+      "learning_rate": 0.00019973040038376777,
+      "loss": 20.1304,
+      "step": 904
+    },
+    {
+      "epoch": 0.1582513661202186,
+      "grad_norm": 1.4687428779681588,
+      "learning_rate": 0.0001997290131130245,
+      "loss": 20.2139,
+      "step": 905
+    },
+    {
+      "epoch": 0.1584262295081967,
+      "grad_norm": 1.3707073768357452,
+      "learning_rate": 0.0001997276222870584,
+      "loss": 20.0199,
+      "step": 906
+    },
+    {
+      "epoch": 0.15860109289617486,
+      "grad_norm": 1.3741866726858913,
+      "learning_rate": 0.00019972622790591916,
+      "loss": 19.7281,
+      "step": 907
+    },
+    {
+      "epoch": 0.158775956284153,
+      "grad_norm": 1.247709392470359,
+      "learning_rate": 0.00019972482996965637,
+      "loss": 19.6781,
+      "step": 908
+    },
+    {
+      "epoch": 0.15895081967213115,
+      "grad_norm": 1.3823284358647585,
+      "learning_rate": 0.00019972342847831987,
+      "loss": 19.9488,
+      "step": 909
+    },
+    {
+      "epoch": 0.1591256830601093,
+      "grad_norm": 1.2824541047851332,
+      "learning_rate": 0.0001997220234319597,
+      "loss": 19.8458,
+      "step": 910
+    },
+    {
+      "epoch": 0.15930054644808744,
+      "grad_norm": 1.269478876729098,
+      "learning_rate": 0.0001997206148306259,
+      "loss": 20.0976,
+      "step": 911
+    },
+    {
+      "epoch": 0.15947540983606556,
+      "grad_norm": 1.4349726399759402,
+      "learning_rate": 0.0001997192026743687,
+      "loss": 19.7584,
+      "step": 912
+    },
+    {
+      "epoch": 0.1596502732240437,
+      "grad_norm": 1.2751769856775843,
+      "learning_rate": 0.00019971778696323842,
+      "loss": 19.7502,
+      "step": 913
+    },
+    {
+      "epoch": 0.15982513661202186,
+      "grad_norm": 1.3598648592698497,
+      "learning_rate": 0.00019971636769728558,
+      "loss": 19.8254,
+      "step": 914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2328721967474663,
+      "learning_rate": 0.00019971494487656068,
+      "loss": 19.5423,
+      "step": 915
+    },
+    {
+      "epoch": 0.16017486338797815,
+      "grad_norm": 1.3147772027596407,
+      "learning_rate": 0.00019971351850111454,
+      "loss": 20.0449,
+      "step": 916
+    },
+    {
+      "epoch": 0.1603497267759563,
+      "grad_norm": 1.2744401713880884,
+      "learning_rate": 0.00019971208857099795,
+      "loss": 19.9607,
+      "step": 917
+    },
+    {
+      "epoch": 0.16052459016393442,
+      "grad_norm": 1.2852566885570487,
+      "learning_rate": 0.0001997106550862619,
+      "loss": 19.6896,
+      "step": 918
+    },
+    {
+      "epoch": 0.16069945355191256,
+      "grad_norm": 1.3667789856873371,
+      "learning_rate": 0.0001997092180469575,
+      "loss": 19.8467,
+      "step": 919
+    },
+    {
+      "epoch": 0.1608743169398907,
+      "grad_norm": 1.3174049914627273,
+      "learning_rate": 0.00019970777745313597,
+      "loss": 19.8737,
+      "step": 920
+    },
+    {
+      "epoch": 0.16104918032786886,
+      "grad_norm": 1.4002747914768499,
+      "learning_rate": 0.0001997063333048487,
+      "loss": 19.8113,
+      "step": 921
+    },
+    {
+      "epoch": 0.161224043715847,
+      "grad_norm": 1.4312661283211916,
+      "learning_rate": 0.0001997048856021471,
+      "loss": 20.0675,
+      "step": 922
+    },
+    {
+      "epoch": 0.16139890710382512,
+      "grad_norm": 1.2959986105042811,
+      "learning_rate": 0.00019970343434508284,
+      "loss": 19.598,
+      "step": 923
+    },
+    {
+      "epoch": 0.16157377049180327,
+      "grad_norm": 1.3804957206632562,
+      "learning_rate": 0.00019970197953370764,
+      "loss": 19.6202,
+      "step": 924
+    },
+    {
+      "epoch": 0.16174863387978142,
+      "grad_norm": 1.4193077530732656,
+      "learning_rate": 0.00019970052116807335,
+      "loss": 20.0281,
+      "step": 925
+    },
+    {
+      "epoch": 0.16192349726775956,
+      "grad_norm": 1.3720240432842543,
+      "learning_rate": 0.00019969905924823199,
+      "loss": 19.8349,
+      "step": 926
+    },
+    {
+      "epoch": 0.1620983606557377,
+      "grad_norm": 1.4165814614076186,
+      "learning_rate": 0.00019969759377423563,
+      "loss": 19.4714,
+      "step": 927
+    },
+    {
+      "epoch": 0.16227322404371586,
+      "grad_norm": 1.5545713766516525,
+      "learning_rate": 0.00019969612474613654,
+      "loss": 19.6454,
+      "step": 928
+    },
+    {
+      "epoch": 0.16244808743169398,
+      "grad_norm": 1.305851680941692,
+      "learning_rate": 0.00019969465216398708,
+      "loss": 19.9802,
+      "step": 929
+    },
+    {
+      "epoch": 0.16262295081967212,
+      "grad_norm": 1.2844341006382183,
+      "learning_rate": 0.00019969317602783977,
+      "loss": 19.9642,
+      "step": 930
+    },
+    {
+      "epoch": 0.16279781420765027,
+      "grad_norm": 1.2770610952766768,
+      "learning_rate": 0.00019969169633774723,
+      "loss": 19.5491,
+      "step": 931
+    },
+    {
+      "epoch": 0.16297267759562842,
+      "grad_norm": 1.3512045633053396,
+      "learning_rate": 0.00019969021309376215,
+      "loss": 19.5813,
+      "step": 932
+    },
+    {
+      "epoch": 0.16314754098360656,
+      "grad_norm": 1.4132611825431685,
+      "learning_rate": 0.00019968872629593746,
+      "loss": 20.0876,
+      "step": 933
+    },
+    {
+      "epoch": 0.1633224043715847,
+      "grad_norm": 1.2344731503213153,
+      "learning_rate": 0.00019968723594432616,
+      "loss": 20.0943,
+      "step": 934
+    },
+    {
+      "epoch": 0.16349726775956283,
+      "grad_norm": 1.5758758069363767,
+      "learning_rate": 0.00019968574203898135,
+      "loss": 19.7211,
+      "step": 935
+    },
+    {
+      "epoch": 0.16367213114754098,
+      "grad_norm": 1.488351922875125,
+      "learning_rate": 0.00019968424457995633,
+      "loss": 19.6306,
+      "step": 936
+    },
+    {
+      "epoch": 0.16384699453551912,
+      "grad_norm": 1.237805156279325,
+      "learning_rate": 0.00019968274356730445,
+      "loss": 19.8861,
+      "step": 937
+    },
+    {
+      "epoch": 0.16402185792349727,
+      "grad_norm": 1.430280823603207,
+      "learning_rate": 0.00019968123900107924,
+      "loss": 19.9332,
+      "step": 938
+    },
+    {
+      "epoch": 0.16419672131147542,
+      "grad_norm": 1.732992457315382,
+      "learning_rate": 0.00019967973088133432,
+      "loss": 19.8782,
+      "step": 939
+    },
+    {
+      "epoch": 0.16437158469945357,
+      "grad_norm": 1.4228089763449026,
+      "learning_rate": 0.00019967821920812345,
+      "loss": 19.8464,
+      "step": 940
+    },
+    {
+      "epoch": 0.16454644808743168,
+      "grad_norm": 1.4194765063238075,
+      "learning_rate": 0.00019967670398150054,
+      "loss": 19.836,
+      "step": 941
+    },
+    {
+      "epoch": 0.16472131147540983,
+      "grad_norm": 1.4848101846251736,
+      "learning_rate": 0.00019967518520151958,
+      "loss": 20.2538,
+      "step": 942
+    },
+    {
+      "epoch": 0.16489617486338798,
+      "grad_norm": 1.3378963737134957,
+      "learning_rate": 0.00019967366286823473,
+      "loss": 19.8787,
+      "step": 943
+    },
+    {
+      "epoch": 0.16507103825136613,
+      "grad_norm": 1.3758407351905813,
+      "learning_rate": 0.00019967213698170023,
+      "loss": 19.7609,
+      "step": 944
+    },
+    {
+      "epoch": 0.16524590163934427,
+      "grad_norm": 1.2832111893086628,
+      "learning_rate": 0.00019967060754197053,
+      "loss": 19.5556,
+      "step": 945
+    },
+    {
+      "epoch": 0.1654207650273224,
+      "grad_norm": 1.4523920161823773,
+      "learning_rate": 0.00019966907454910013,
+      "loss": 19.6547,
+      "step": 946
+    },
+    {
+      "epoch": 0.16559562841530054,
+      "grad_norm": 1.4550879511232486,
+      "learning_rate": 0.00019966753800314366,
+      "loss": 20.0375,
+      "step": 947
+    },
+    {
+      "epoch": 0.16577049180327869,
+      "grad_norm": 1.278634016481033,
+      "learning_rate": 0.00019966599790415592,
+      "loss": 19.6786,
+      "step": 948
+    },
+    {
+      "epoch": 0.16594535519125683,
+      "grad_norm": 1.4004026800564668,
+      "learning_rate": 0.00019966445425219177,
+      "loss": 19.8162,
+      "step": 949
+    },
+    {
+      "epoch": 0.16612021857923498,
+      "grad_norm": 1.4149498395574276,
+      "learning_rate": 0.0001996629070473063,
+      "loss": 19.6902,
+      "step": 950
+    },
+    {
+      "epoch": 0.16629508196721313,
+      "grad_norm": 1.245804803868014,
+      "learning_rate": 0.00019966135628955462,
+      "loss": 19.9225,
+      "step": 951
+    },
+    {
+      "epoch": 0.16646994535519125,
+      "grad_norm": 1.4749545779638245,
+      "learning_rate": 0.00019965980197899207,
+      "loss": 19.8565,
+      "step": 952
+    },
+    {
+      "epoch": 0.1666448087431694,
+      "grad_norm": 1.362515156019282,
+      "learning_rate": 0.00019965824411567398,
+      "loss": 19.5739,
+      "step": 953
+    },
+    {
+      "epoch": 0.16681967213114754,
+      "grad_norm": 1.2668615640874228,
+      "learning_rate": 0.00019965668269965594,
+      "loss": 19.5864,
+      "step": 954
+    },
+    {
+      "epoch": 0.1669945355191257,
+      "grad_norm": 1.3018840538541674,
+      "learning_rate": 0.0001996551177309936,
+      "loss": 19.5993,
+      "step": 955
+    },
+    {
+      "epoch": 0.16716939890710383,
+      "grad_norm": 1.3541504064073189,
+      "learning_rate": 0.00019965354920974273,
+      "loss": 19.7446,
+      "step": 956
+    },
+    {
+      "epoch": 0.16734426229508198,
+      "grad_norm": 1.319584081888164,
+      "learning_rate": 0.0001996519771359593,
+      "loss": 19.6207,
+      "step": 957
+    },
+    {
+      "epoch": 0.1675191256830601,
+      "grad_norm": 1.231539342278188,
+      "learning_rate": 0.00019965040150969927,
+      "loss": 19.4233,
+      "step": 958
+    },
+    {
+      "epoch": 0.16769398907103825,
+      "grad_norm": 1.2318519097363205,
+      "learning_rate": 0.00019964882233101887,
+      "loss": 19.566,
+      "step": 959
+    },
+    {
+      "epoch": 0.1678688524590164,
+      "grad_norm": 1.4158244763879513,
+      "learning_rate": 0.0001996472395999744,
+      "loss": 19.6809,
+      "step": 960
+    },
+    {
+      "epoch": 0.16804371584699454,
+      "grad_norm": 1.4089477271343345,
+      "learning_rate": 0.00019964565331662224,
+      "loss": 19.4136,
+      "step": 961
+    },
+    {
+      "epoch": 0.1682185792349727,
+      "grad_norm": 1.3115647505347263,
+      "learning_rate": 0.00019964406348101896,
+      "loss": 19.4166,
+      "step": 962
+    },
+    {
+      "epoch": 0.1683934426229508,
+      "grad_norm": 1.3224805388798946,
+      "learning_rate": 0.00019964247009322125,
+      "loss": 19.5034,
+      "step": 963
+    },
+    {
+      "epoch": 0.16856830601092895,
+      "grad_norm": 1.4483845490396527,
+      "learning_rate": 0.00019964087315328593,
+      "loss": 19.749,
+      "step": 964
+    },
+    {
+      "epoch": 0.1687431693989071,
+      "grad_norm": 1.3643525190982806,
+      "learning_rate": 0.00019963927266126988,
+      "loss": 19.6729,
+      "step": 965
+    },
+    {
+      "epoch": 0.16891803278688525,
+      "grad_norm": 1.3446312015907202,
+      "learning_rate": 0.00019963766861723014,
+      "loss": 20.128,
+      "step": 966
+    },
+    {
+      "epoch": 0.1690928961748634,
+      "grad_norm": 1.3058336802341861,
+      "learning_rate": 0.00019963606102122396,
+      "loss": 19.4073,
+      "step": 967
+    },
+    {
+      "epoch": 0.16926775956284154,
+      "grad_norm": 1.2142067463683872,
+      "learning_rate": 0.00019963444987330863,
+      "loss": 19.9203,
+      "step": 968
+    },
+    {
+      "epoch": 0.16944262295081966,
+      "grad_norm": 1.3727844765255597,
+      "learning_rate": 0.00019963283517354153,
+      "loss": 19.6948,
+      "step": 969
+    },
+    {
+      "epoch": 0.1696174863387978,
+      "grad_norm": 1.2735674057663253,
+      "learning_rate": 0.0001996312169219803,
+      "loss": 19.8846,
+      "step": 970
+    },
+    {
+      "epoch": 0.16979234972677595,
+      "grad_norm": 1.3495963072928474,
+      "learning_rate": 0.00019962959511868257,
+      "loss": 19.5717,
+      "step": 971
+    },
+    {
+      "epoch": 0.1699672131147541,
+      "grad_norm": 1.3390388982002612,
+      "learning_rate": 0.0001996279697637062,
+      "loss": 19.7193,
+      "step": 972
+    },
+    {
+      "epoch": 0.17014207650273225,
+      "grad_norm": 1.2068074592640192,
+      "learning_rate": 0.00019962634085710912,
+      "loss": 19.7624,
+      "step": 973
+    },
+    {
+      "epoch": 0.1703169398907104,
+      "grad_norm": 1.3292131928875004,
+      "learning_rate": 0.00019962470839894936,
+      "loss": 19.3494,
+      "step": 974
+    },
+    {
+      "epoch": 0.17049180327868851,
+      "grad_norm": 1.314370849887888,
+      "learning_rate": 0.00019962307238928517,
+      "loss": 19.3734,
+      "step": 975
+    },
+    {
+      "epoch": 0.17066666666666666,
+      "grad_norm": 1.1939802913013235,
+      "learning_rate": 0.00019962143282817484,
+      "loss": 19.4145,
+      "step": 976
+    },
+    {
+      "epoch": 0.1708415300546448,
+      "grad_norm": 1.4102711998384776,
+      "learning_rate": 0.0001996197897156768,
+      "loss": 19.3822,
+      "step": 977
+    },
+    {
+      "epoch": 0.17101639344262295,
+      "grad_norm": 1.2661873459827928,
+      "learning_rate": 0.00019961814305184968,
+      "loss": 19.6976,
+      "step": 978
+    },
+    {
+      "epoch": 0.1711912568306011,
+      "grad_norm": 1.267588121446262,
+      "learning_rate": 0.00019961649283675216,
+      "loss": 19.4422,
+      "step": 979
+    },
+    {
+      "epoch": 0.17136612021857925,
+      "grad_norm": 1.3226767679258262,
+      "learning_rate": 0.000199614839070443,
+      "loss": 19.5923,
+      "step": 980
+    },
+    {
+      "epoch": 0.17154098360655737,
+      "grad_norm": 1.3121089871591924,
+      "learning_rate": 0.00019961318175298125,
+      "loss": 19.8583,
+      "step": 981
+    },
+    {
+      "epoch": 0.17171584699453551,
+      "grad_norm": 1.184184293321143,
+      "learning_rate": 0.000199611520884426,
+      "loss": 19.3013,
+      "step": 982
+    },
+    {
+      "epoch": 0.17189071038251366,
+      "grad_norm": 1.2949859456679917,
+      "learning_rate": 0.00019960985646483634,
+      "loss": 19.7952,
+      "step": 983
+    },
+    {
+      "epoch": 0.1720655737704918,
+      "grad_norm": 1.3866526136006447,
+      "learning_rate": 0.00019960818849427172,
+      "loss": 19.7235,
+      "step": 984
+    },
+    {
+      "epoch": 0.17224043715846996,
+      "grad_norm": 1.313061898891991,
+      "learning_rate": 0.00019960651697279155,
+      "loss": 19.4775,
+      "step": 985
+    },
+    {
+      "epoch": 0.17241530054644807,
+      "grad_norm": 1.364330389531021,
+      "learning_rate": 0.00019960484190045542,
+      "loss": 19.4724,
+      "step": 986
+    },
+    {
+      "epoch": 0.17259016393442622,
+      "grad_norm": 1.6549263694421181,
+      "learning_rate": 0.00019960316327732308,
+      "loss": 19.5357,
+      "step": 987
+    },
+    {
+      "epoch": 0.17276502732240437,
+      "grad_norm": 1.97877599264236,
+      "learning_rate": 0.00019960148110345432,
+      "loss": 19.5259,
+      "step": 988
+    },
+    {
+      "epoch": 0.17293989071038252,
+      "grad_norm": 1.6186960704984417,
+      "learning_rate": 0.00019959979537890915,
+      "loss": 19.3942,
+      "step": 989
+    },
+    {
+      "epoch": 0.17311475409836066,
+      "grad_norm": 1.5635595869295358,
+      "learning_rate": 0.0001995981061037476,
+      "loss": 19.2376,
+      "step": 990
+    },
+    {
+      "epoch": 0.1732896174863388,
+      "grad_norm": 1.629277457889035,
+      "learning_rate": 0.00019959641327802995,
+      "loss": 19.7325,
+      "step": 991
+    },
+    {
+      "epoch": 0.17346448087431693,
+      "grad_norm": 1.6469395969816802,
+      "learning_rate": 0.00019959471690181657,
+      "loss": 19.7159,
+      "step": 992
+    },
+    {
+      "epoch": 0.17363934426229508,
+      "grad_norm": 1.5934094437031194,
+      "learning_rate": 0.00019959301697516788,
+      "loss": 19.3461,
+      "step": 993
+    },
+    {
+      "epoch": 0.17381420765027322,
+      "grad_norm": 1.4488013332671699,
+      "learning_rate": 0.00019959131349814448,
+      "loss": 19.4833,
+      "step": 994
+    },
+    {
+      "epoch": 0.17398907103825137,
+      "grad_norm": 1.3966607614561988,
+      "learning_rate": 0.00019958960647080713,
+      "loss": 19.736,
+      "step": 995
+    },
+    {
+      "epoch": 0.17416393442622952,
+      "grad_norm": 1.5123881535436954,
+      "learning_rate": 0.00019958789589321664,
+      "loss": 19.3524,
+      "step": 996
+    },
+    {
+      "epoch": 0.17433879781420766,
+      "grad_norm": 1.5120749993482852,
+      "learning_rate": 0.00019958618176543405,
+      "loss": 19.5458,
+      "step": 997
+    },
+    {
+      "epoch": 0.17451366120218578,
+      "grad_norm": 1.4911934783290044,
+      "learning_rate": 0.00019958446408752044,
+      "loss": 19.5937,
+      "step": 998
+    },
+    {
+      "epoch": 0.17468852459016393,
+      "grad_norm": 1.4095352969522055,
+      "learning_rate": 0.00019958274285953703,
+      "loss": 19.8468,
+      "step": 999
+    },
+    {
+      "epoch": 0.17486338797814208,
+      "grad_norm": 1.3614337014121825,
+      "learning_rate": 0.00019958101808154516,
+      "loss": 19.6515,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17503825136612022,
+      "grad_norm": 1.3441348132717117,
+      "learning_rate": 0.00019957928975360636,
+      "loss": 19.7161,
+      "step": 1001
+    },
+    {
+      "epoch": 0.17521311475409837,
+      "grad_norm": 1.4172224262806417,
+      "learning_rate": 0.00019957755787578224,
+      "loss": 19.2393,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1753879781420765,
+      "grad_norm": 1.4346317432060496,
+      "learning_rate": 0.0001995758224481345,
+      "loss": 19.6299,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17556284153005464,
+      "grad_norm": 1.5322576988714849,
+      "learning_rate": 0.00019957408347072507,
+      "loss": 19.2634,
+      "step": 1004
+    },
+    {
+      "epoch": 0.17573770491803278,
+      "grad_norm": 1.3490321050420313,
+      "learning_rate": 0.00019957234094361586,
+      "loss": 19.5492,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17591256830601093,
+      "grad_norm": 1.348372404811798,
+      "learning_rate": 0.00019957059486686905,
+      "loss": 19.6576,
+      "step": 1006
+    },
+    {
+      "epoch": 0.17608743169398908,
+      "grad_norm": 1.3265014022274648,
+      "learning_rate": 0.00019956884524054686,
+      "loss": 19.7314,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17626229508196722,
+      "grad_norm": 1.3261948974073543,
+      "learning_rate": 0.0001995670920647117,
+      "loss": 19.5822,
+      "step": 1008
+    },
+    {
+      "epoch": 0.17643715846994534,
+      "grad_norm": 1.2481071280523042,
+      "learning_rate": 0.000199565335339426,
+      "loss": 19.5337,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1766120218579235,
+      "grad_norm": 1.2789302681671852,
+      "learning_rate": 0.00019956357506475245,
+      "loss": 19.8494,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17678688524590164,
+      "grad_norm": 1.2513480780869344,
+      "learning_rate": 0.00019956181124075378,
+      "loss": 19.8701,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17696174863387978,
+      "grad_norm": 1.3964175587360528,
+      "learning_rate": 0.00019956004386749286,
+      "loss": 19.3415,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17713661202185793,
+      "grad_norm": 1.292099494997329,
+      "learning_rate": 0.0001995582729450327,
+      "loss": 19.5398,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17731147540983608,
+      "grad_norm": 1.2831239982717917,
+      "learning_rate": 0.00019955649847343642,
+      "loss": 19.6113,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1774863387978142,
+      "grad_norm": 1.3261688888699408,
+      "learning_rate": 0.00019955472045276728,
+      "loss": 19.5661,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17766120218579234,
+      "grad_norm": 1.3212626514361785,
+      "learning_rate": 0.0001995529388830887,
+      "loss": 19.5011,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1778360655737705,
+      "grad_norm": 1.4102180996445532,
+      "learning_rate": 0.00019955115376446415,
+      "loss": 19.4511,
+      "step": 1017
+    },
+    {
+      "epoch": 0.17801092896174864,
+      "grad_norm": 1.3586035507381788,
+      "learning_rate": 0.0001995493650969573,
+      "loss": 19.4661,
+      "step": 1018
+    },
+    {
+      "epoch": 0.17818579234972678,
+      "grad_norm": 1.5516317436439893,
+      "learning_rate": 0.00019954757288063185,
+      "loss": 19.1962,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17836065573770493,
+      "grad_norm": 1.4001992981020221,
+      "learning_rate": 0.0001995457771155518,
+      "loss": 19.3016,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17853551912568305,
+      "grad_norm": 1.5004753918888654,
+      "learning_rate": 0.00019954397780178106,
+      "loss": 19.6729,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1787103825136612,
+      "grad_norm": 1.7662619669872315,
+      "learning_rate": 0.00019954217493938382,
+      "loss": 19.6293,
+      "step": 1022
+    },
+    {
+      "epoch": 0.17888524590163934,
+      "grad_norm": 1.4001076292098547,
+      "learning_rate": 0.00019954036852842433,
+      "loss": 19.4689,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1790601092896175,
+      "grad_norm": 1.5219545226108333,
+      "learning_rate": 0.00019953855856896704,
+      "loss": 19.5425,
+      "step": 1024
+    },
+    {
+      "epoch": 0.17923497267759564,
+      "grad_norm": 1.5684493302018054,
+      "learning_rate": 0.00019953674506107641,
+      "loss": 19.4769,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17940983606557376,
+      "grad_norm": 1.5808877535240833,
+      "learning_rate": 0.00019953492800481713,
+      "loss": 19.686,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1795846994535519,
+      "grad_norm": 1.544148829517665,
+      "learning_rate": 0.00019953310740025395,
+      "loss": 19.5717,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17975956284153005,
+      "grad_norm": 1.5749927291595933,
+      "learning_rate": 0.00019953128324745178,
+      "loss": 19.8058,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1799344262295082,
+      "grad_norm": 1.5849530834837326,
+      "learning_rate": 0.00019952945554647564,
+      "loss": 19.6037,
+      "step": 1029
+    },
+    {
+      "epoch": 0.18010928961748635,
+      "grad_norm": 1.8447278227624593,
+      "learning_rate": 0.00019952762429739074,
+      "loss": 19.437,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1802841530054645,
+      "grad_norm": 1.742168810404768,
+      "learning_rate": 0.0001995257895002623,
+      "loss": 19.4708,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1804590163934426,
+      "grad_norm": 2.202327370572489,
+      "learning_rate": 0.00019952395115515573,
+      "loss": 19.7286,
+      "step": 1032
+    },
+    {
+      "epoch": 0.18063387978142076,
+      "grad_norm": 1.7200206273416838,
+      "learning_rate": 0.0001995221092621366,
+      "loss": 19.7494,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1808087431693989,
+      "grad_norm": 1.6674072937928308,
+      "learning_rate": 0.00019952026382127054,
+      "loss": 18.9442,
+      "step": 1034
+    },
+    {
+      "epoch": 0.18098360655737705,
+      "grad_norm": 1.9320535992821364,
+      "learning_rate": 0.00019951841483262337,
+      "loss": 19.4541,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1811584699453552,
+      "grad_norm": 1.535118375870958,
+      "learning_rate": 0.00019951656229626096,
+      "loss": 19.7926,
+      "step": 1036
+    },
+    {
+      "epoch": 0.18133333333333335,
+      "grad_norm": 1.4429284110948841,
+      "learning_rate": 0.0001995147062122494,
+      "loss": 19.3934,
+      "step": 1037
+    },
+    {
+      "epoch": 0.18150819672131147,
+      "grad_norm": 1.4650229583364678,
+      "learning_rate": 0.00019951284658065483,
+      "loss": 19.4995,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1816830601092896,
+      "grad_norm": 1.495399673197328,
+      "learning_rate": 0.00019951098340154354,
+      "loss": 19.341,
+      "step": 1039
+    },
+    {
+      "epoch": 0.18185792349726776,
+      "grad_norm": 1.4796774897308305,
+      "learning_rate": 0.000199509116674982,
+      "loss": 19.5645,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1820327868852459,
+      "grad_norm": 1.535478908196544,
+      "learning_rate": 0.00019950724640103664,
+      "loss": 19.3404,
+      "step": 1041
+    },
+    {
+      "epoch": 0.18220765027322405,
+      "grad_norm": 1.547290891631957,
+      "learning_rate": 0.00019950537257977424,
+      "loss": 19.0147,
+      "step": 1042
+    },
+    {
+      "epoch": 0.18238251366120217,
+      "grad_norm": 1.5077760909872937,
+      "learning_rate": 0.00019950349521126158,
+      "loss": 19.3107,
+      "step": 1043
+    },
+    {
+      "epoch": 0.18255737704918032,
+      "grad_norm": 1.7247717830663887,
+      "learning_rate": 0.00019950161429556555,
+      "loss": 19.2961,
+      "step": 1044
+    },
+    {
+      "epoch": 0.18273224043715847,
+      "grad_norm": 1.520874898870438,
+      "learning_rate": 0.00019949972983275327,
+      "loss": 19.4493,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1829071038251366,
+      "grad_norm": 1.6353490430056257,
+      "learning_rate": 0.00019949784182289184,
+      "loss": 19.3697,
+      "step": 1046
+    },
+    {
+      "epoch": 0.18308196721311476,
+      "grad_norm": 1.8183591470916314,
+      "learning_rate": 0.0001994959502660486,
+      "loss": 19.5211,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1832568306010929,
+      "grad_norm": 1.6230075725162287,
+      "learning_rate": 0.00019949405516229098,
+      "loss": 18.9628,
+      "step": 1048
+    },
+    {
+      "epoch": 0.18343169398907103,
+      "grad_norm": 1.5479313364578149,
+      "learning_rate": 0.00019949215651168657,
+      "loss": 19.211,
+      "step": 1049
+    },
+    {
+      "epoch": 0.18360655737704917,
+      "grad_norm": 1.4210492671589066,
+      "learning_rate": 0.000199490254314303,
+      "loss": 19.519,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18378142076502732,
+      "grad_norm": 1.5489677169880829,
+      "learning_rate": 0.0001994883485702081,
+      "loss": 19.2902,
+      "step": 1051
+    },
+    {
+      "epoch": 0.18395628415300547,
+      "grad_norm": 1.5433379298303516,
+      "learning_rate": 0.00019948643927946978,
+      "loss": 19.2067,
+      "step": 1052
+    },
+    {
+      "epoch": 0.18413114754098361,
+      "grad_norm": 1.4509743142704377,
+      "learning_rate": 0.0001994845264421562,
+      "loss": 19.3254,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18430601092896176,
+      "grad_norm": 1.5099655317073868,
+      "learning_rate": 0.00019948261005833546,
+      "loss": 19.3812,
+      "step": 1054
+    },
+    {
+      "epoch": 0.18448087431693988,
+      "grad_norm": 1.4746533922551992,
+      "learning_rate": 0.0001994806901280759,
+      "loss": 19.3512,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18465573770491803,
+      "grad_norm": 1.48618897541253,
+      "learning_rate": 0.00019947876665144597,
+      "loss": 19.3056,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18483060109289617,
+      "grad_norm": 1.9876374571132254,
+      "learning_rate": 0.00019947683962851426,
+      "loss": 19.1813,
+      "step": 1057
+    },
+    {
+      "epoch": 0.18500546448087432,
+      "grad_norm": 1.437915368076657,
+      "learning_rate": 0.0001994749090593494,
+      "loss": 19.4327,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18518032786885247,
+      "grad_norm": 1.7518160692614155,
+      "learning_rate": 0.0001994729749440203,
+      "loss": 19.5388,
+      "step": 1059
+    },
+    {
+      "epoch": 0.18535519125683061,
+      "grad_norm": 1.459306071340485,
+      "learning_rate": 0.00019947103728259588,
+      "loss": 19.4625,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18553005464480873,
+      "grad_norm": 1.3993319327201208,
+      "learning_rate": 0.00019946909607514515,
+      "loss": 19.181,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18570491803278688,
+      "grad_norm": 1.366759197132066,
+      "learning_rate": 0.0001994671513217374,
+      "loss": 19.2434,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18587978142076503,
+      "grad_norm": 1.4121635771525025,
+      "learning_rate": 0.00019946520302244186,
+      "loss": 18.9851,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18605464480874317,
+      "grad_norm": 1.5165597626555145,
+      "learning_rate": 0.00019946325117732808,
+      "loss": 19.0491,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18622950819672132,
+      "grad_norm": 1.2940489445877688,
+      "learning_rate": 0.0001994612957864656,
+      "loss": 19.4774,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18640437158469944,
+      "grad_norm": 1.4184403517418196,
+      "learning_rate": 0.00019945933684992418,
+      "loss": 19.0542,
+      "step": 1066
+    },
+    {
+      "epoch": 0.1865792349726776,
+      "grad_norm": 1.345150517862019,
+      "learning_rate": 0.00019945737436777357,
+      "loss": 19.3622,
+      "step": 1067
+    },
+    {
+      "epoch": 0.18675409836065573,
+      "grad_norm": 1.358430787626506,
+      "learning_rate": 0.00019945540834008377,
+      "loss": 19.38,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18692896174863388,
+      "grad_norm": 1.302600945772105,
+      "learning_rate": 0.00019945343876692485,
+      "loss": 19.4249,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18710382513661203,
+      "grad_norm": 1.4261746931504553,
+      "learning_rate": 0.00019945146564836705,
+      "loss": 19.0869,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18727868852459018,
+      "grad_norm": 1.3379764597496466,
+      "learning_rate": 0.0001994494889844807,
+      "loss": 19.103,
+      "step": 1071
+    },
+    {
+      "epoch": 0.1874535519125683,
+      "grad_norm": 1.3461602554963372,
+      "learning_rate": 0.00019944750877533624,
+      "loss": 19.155,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18762841530054644,
+      "grad_norm": 1.689399420445121,
+      "learning_rate": 0.00019944552502100432,
+      "loss": 18.8798,
+      "step": 1073
+    },
+    {
+      "epoch": 0.1878032786885246,
+      "grad_norm": 1.7942626995233892,
+      "learning_rate": 0.0001994435377215556,
+      "loss": 19.3715,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18797814207650274,
+      "grad_norm": 1.241770261447969,
+      "learning_rate": 0.00019944154687706096,
+      "loss": 19.3258,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18815300546448088,
+      "grad_norm": 1.6762216526012204,
+      "learning_rate": 0.00019943955248759133,
+      "loss": 18.6739,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18832786885245903,
+      "grad_norm": 1.8211961603591211,
+      "learning_rate": 0.0001994375545532179,
+      "loss": 19.483,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18850273224043715,
+      "grad_norm": 1.3554656769280469,
+      "learning_rate": 0.00019943555307401175,
+      "loss": 18.6847,
+      "step": 1078
+    },
+    {
+      "epoch": 0.1886775956284153,
+      "grad_norm": 1.447178263987568,
+      "learning_rate": 0.00019943354805004434,
+      "loss": 19.2477,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18885245901639344,
+      "grad_norm": 1.345296749366657,
+      "learning_rate": 0.00019943153948138713,
+      "loss": 19.3467,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1890273224043716,
+      "grad_norm": 1.802460253690046,
+      "learning_rate": 0.0001994295273681117,
+      "loss": 19.3436,
+      "step": 1081
+    },
+    {
+      "epoch": 0.18920218579234974,
+      "grad_norm": 1.7915346080681953,
+      "learning_rate": 0.00019942751171028982,
+      "loss": 18.9898,
+      "step": 1082
+    },
+    {
+      "epoch": 0.18937704918032786,
+      "grad_norm": 1.4519674368230169,
+      "learning_rate": 0.0001994254925079933,
+      "loss": 19.1765,
+      "step": 1083
+    },
+    {
+      "epoch": 0.189551912568306,
+      "grad_norm": 1.6874653400318131,
+      "learning_rate": 0.00019942346976129413,
+      "loss": 19.4259,
+      "step": 1084
+    },
+    {
+      "epoch": 0.18972677595628415,
+      "grad_norm": 1.740682075516354,
+      "learning_rate": 0.00019942144347026442,
+      "loss": 19.1982,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1899016393442623,
+      "grad_norm": 1.3443772092719877,
+      "learning_rate": 0.00019941941363497642,
+      "loss": 19.4253,
+      "step": 1086
+    },
+    {
+      "epoch": 0.19007650273224044,
+      "grad_norm": 1.438576694840547,
+      "learning_rate": 0.00019941738025550248,
+      "loss": 18.8751,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1902513661202186,
+      "grad_norm": 1.3931682049311325,
+      "learning_rate": 0.0001994153433319151,
+      "loss": 19.1694,
+      "step": 1088
+    },
+    {
+      "epoch": 0.1904262295081967,
+      "grad_norm": 1.4000223852031382,
+      "learning_rate": 0.00019941330286428688,
+      "loss": 19.3496,
+      "step": 1089
+    },
+    {
+      "epoch": 0.19060109289617486,
+      "grad_norm": 1.3509382042905818,
+      "learning_rate": 0.00019941125885269054,
+      "loss": 19.1443,
+      "step": 1090
+    },
+    {
+      "epoch": 0.190775956284153,
+      "grad_norm": 1.421853850169956,
+      "learning_rate": 0.000199409211297199,
+      "loss": 19.4318,
+      "step": 1091
+    },
+    {
+      "epoch": 0.19095081967213115,
+      "grad_norm": 1.5737361749018475,
+      "learning_rate": 0.0001994071601978852,
+      "loss": 18.9774,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1911256830601093,
+      "grad_norm": 1.6206968994936957,
+      "learning_rate": 0.00019940510555482234,
+      "loss": 19.3353,
+      "step": 1093
+    },
+    {
+      "epoch": 0.19130054644808744,
+      "grad_norm": 1.3156512344887465,
+      "learning_rate": 0.00019940304736808354,
+      "loss": 19.0201,
+      "step": 1094
+    },
+    {
+      "epoch": 0.19147540983606556,
+      "grad_norm": 1.3125763544238283,
+      "learning_rate": 0.0001994009856377423,
+      "loss": 19.5437,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1916502732240437,
+      "grad_norm": 1.2879974850082636,
+      "learning_rate": 0.00019939892036387202,
+      "loss": 19.1067,
+      "step": 1096
+    },
+    {
+      "epoch": 0.19182513661202186,
+      "grad_norm": 1.339094063930677,
+      "learning_rate": 0.0001993968515465464,
+      "loss": 18.5712,
+      "step": 1097
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.43148079539111,
+      "learning_rate": 0.00019939477918583912,
+      "loss": 18.8232,
+      "step": 1098
+    },
+    {
+      "epoch": 0.19217486338797815,
+      "grad_norm": 1.4653375866031668,
+      "learning_rate": 0.00019939270328182408,
+      "loss": 19.409,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1923497267759563,
+      "grad_norm": 1.3539269624252213,
+      "learning_rate": 0.00019939062383457533,
+      "loss": 19.1148,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19252459016393442,
+      "grad_norm": 1.3218165924021008,
+      "learning_rate": 0.00019938854084416695,
+      "loss": 19.287,
+      "step": 1101
+    },
+    {
+      "epoch": 0.19269945355191256,
+      "grad_norm": 1.401538371605601,
+      "learning_rate": 0.00019938645431067325,
+      "loss": 19.4156,
+      "step": 1102
+    },
+    {
+      "epoch": 0.1928743169398907,
+      "grad_norm": 1.703265551100519,
+      "learning_rate": 0.00019938436423416853,
+      "loss": 18.7401,
+      "step": 1103
+    },
+    {
+      "epoch": 0.19304918032786886,
+      "grad_norm": 1.7279397401143672,
+      "learning_rate": 0.00019938227061472733,
+      "loss": 18.8475,
+      "step": 1104
+    },
+    {
+      "epoch": 0.193224043715847,
+      "grad_norm": 1.4715151701068794,
+      "learning_rate": 0.00019938017345242436,
+      "loss": 19.277,
+      "step": 1105
+    },
+    {
+      "epoch": 0.19339890710382512,
+      "grad_norm": 1.414626155556516,
+      "learning_rate": 0.00019937807274733425,
+      "loss": 19.1569,
+      "step": 1106
+    },
+    {
+      "epoch": 0.19357377049180327,
+      "grad_norm": 1.461615910678144,
+      "learning_rate": 0.00019937596849953202,
+      "loss": 19.2301,
+      "step": 1107
+    },
+    {
+      "epoch": 0.19374863387978142,
+      "grad_norm": 1.3964442943324462,
+      "learning_rate": 0.0001993738607090926,
+      "loss": 19.5276,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19392349726775956,
+      "grad_norm": 1.818059256419435,
+      "learning_rate": 0.00019937174937609113,
+      "loss": 19.1364,
+      "step": 1109
+    },
+    {
+      "epoch": 0.1940983606557377,
+      "grad_norm": 1.960615972738547,
+      "learning_rate": 0.0001993696345006029,
+      "loss": 19.4545,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19427322404371586,
+      "grad_norm": 2.067002576921416,
+      "learning_rate": 0.0001993675160827033,
+      "loss": 18.8762,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19444808743169398,
+      "grad_norm": 1.7350782876931656,
+      "learning_rate": 0.00019936539412246788,
+      "loss": 19.2967,
+      "step": 1112
+    },
+    {
+      "epoch": 0.19462295081967212,
+      "grad_norm": 1.5493291911234606,
+      "learning_rate": 0.00019936326861997225,
+      "loss": 18.9989,
+      "step": 1113
+    },
+    {
+      "epoch": 0.19479781420765027,
+      "grad_norm": 1.8826905183878657,
+      "learning_rate": 0.00019936113957529216,
+      "loss": 19.2541,
+      "step": 1114
+    },
+    {
+      "epoch": 0.19497267759562842,
+      "grad_norm": 2.2469843831093295,
+      "learning_rate": 0.00019935900698850358,
+      "loss": 19.4937,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19514754098360657,
+      "grad_norm": 2.322481710288418,
+      "learning_rate": 0.00019935687085968244,
+      "loss": 19.1972,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1953224043715847,
+      "grad_norm": 1.6277651446306327,
+      "learning_rate": 0.00019935473118890497,
+      "loss": 18.6955,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19549726775956283,
+      "grad_norm": 2.0792155735213473,
+      "learning_rate": 0.00019935258797624741,
+      "loss": 19.2692,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19567213114754098,
+      "grad_norm": 2.1117489045649593,
+      "learning_rate": 0.00019935044122178618,
+      "loss": 19.0071,
+      "step": 1119
+    },
+    {
+      "epoch": 0.19584699453551913,
+      "grad_norm": 1.5426892788789335,
+      "learning_rate": 0.00019934829092559776,
+      "loss": 19.2058,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19602185792349727,
+      "grad_norm": 1.9973990573710716,
+      "learning_rate": 0.00019934613708775885,
+      "loss": 19.3478,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19619672131147542,
+      "grad_norm": 1.5400344630133678,
+      "learning_rate": 0.0001993439797083463,
+      "loss": 19.2388,
+      "step": 1122
+    },
+    {
+      "epoch": 0.19637158469945354,
+      "grad_norm": 2.1818493571106625,
+      "learning_rate": 0.00019934181878743687,
+      "loss": 19.0536,
+      "step": 1123
+    },
+    {
+      "epoch": 0.19654644808743169,
+      "grad_norm": 1.779951917410962,
+      "learning_rate": 0.00019933965432510767,
+      "loss": 18.8479,
+      "step": 1124
+    },
+    {
+      "epoch": 0.19672131147540983,
+      "grad_norm": 1.5897758214495117,
+      "learning_rate": 0.0001993374863214359,
+      "loss": 19.1842,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19689617486338798,
+      "grad_norm": 1.6075503995637959,
+      "learning_rate": 0.00019933531477649879,
+      "loss": 18.9811,
+      "step": 1126
+    },
+    {
+      "epoch": 0.19707103825136613,
+      "grad_norm": 1.5930678312057172,
+      "learning_rate": 0.00019933313969037374,
+      "loss": 18.8618,
+      "step": 1127
+    },
+    {
+      "epoch": 0.19724590163934427,
+      "grad_norm": 1.3550070733748825,
+      "learning_rate": 0.00019933096106313836,
+      "loss": 18.9897,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1974207650273224,
+      "grad_norm": 1.5533166084560903,
+      "learning_rate": 0.00019932877889487026,
+      "loss": 18.9144,
+      "step": 1129
+    },
+    {
+      "epoch": 0.19759562841530054,
+      "grad_norm": 1.5500758360769153,
+      "learning_rate": 0.00019932659318564724,
+      "loss": 19.0631,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1977704918032787,
+      "grad_norm": 1.4199372201454832,
+      "learning_rate": 0.00019932440393554722,
+      "loss": 18.9184,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19794535519125683,
+      "grad_norm": 1.5350176167063316,
+      "learning_rate": 0.00019932221114464827,
+      "loss": 19.0294,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19812021857923498,
+      "grad_norm": 1.5230572518358763,
+      "learning_rate": 0.0001993200148130285,
+      "loss": 18.756,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19829508196721313,
+      "grad_norm": 1.5240897874089026,
+      "learning_rate": 0.00019931781494076628,
+      "loss": 18.5926,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19846994535519125,
+      "grad_norm": 1.4747912561775705,
+      "learning_rate": 0.00019931561152793998,
+      "loss": 18.041,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1986448087431694,
+      "grad_norm": 1.5774832056309964,
+      "learning_rate": 0.00019931340457462817,
+      "loss": 18.9621,
+      "step": 1136
+    },
+    {
+      "epoch": 0.19881967213114754,
+      "grad_norm": 1.648127816397584,
+      "learning_rate": 0.00019931119408090952,
+      "loss": 19.0788,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1989945355191257,
+      "grad_norm": 1.571366790233563,
+      "learning_rate": 0.00019930898004686282,
+      "loss": 18.9714,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19916939890710383,
+      "grad_norm": 1.7788632016396098,
+      "learning_rate": 0.00019930676247256702,
+      "loss": 18.839,
+      "step": 1139
+    },
+    {
+      "epoch": 0.19934426229508198,
+      "grad_norm": 1.9511604735410486,
+      "learning_rate": 0.00019930454135810122,
+      "loss": 18.9233,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1995191256830601,
+      "grad_norm": 1.5834921486537894,
+      "learning_rate": 0.00019930231670354446,
+      "loss": 19.1752,
+      "step": 1141
+    },
+    {
+      "epoch": 0.19969398907103825,
+      "grad_norm": 1.5442604249078462,
+      "learning_rate": 0.00019930008850897618,
+      "loss": 18.7033,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1998688524590164,
+      "grad_norm": 1.4544037725638024,
+      "learning_rate": 0.00019929785677447577,
+      "loss": 18.9764,
+      "step": 1143
+    },
+    {
+      "epoch": 0.20004371584699454,
+      "grad_norm": 1.6890845383405941,
+      "learning_rate": 0.00019929562150012276,
+      "loss": 18.7224,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2002185792349727,
+      "grad_norm": 1.5732969260014604,
+      "learning_rate": 0.0001992933826859969,
+      "loss": 18.9916,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2003934426229508,
+      "grad_norm": 1.9281587925632055,
+      "learning_rate": 0.0001992911403321779,
+      "loss": 18.5169,
+      "step": 1146
+    },
+    {
+      "epoch": 0.20056830601092895,
+      "grad_norm": 2.1545267070645937,
+      "learning_rate": 0.0001992888944387458,
+      "loss": 18.628,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2007431693989071,
+      "grad_norm": 2.2528494853848935,
+      "learning_rate": 0.00019928664500578057,
+      "loss": 18.8532,
+      "step": 1148
+    },
+    {
+      "epoch": 0.20091803278688525,
+      "grad_norm": 1.491257260988266,
+      "learning_rate": 0.0001992843920333625,
+      "loss": 18.6853,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2010928961748634,
+      "grad_norm": 2.0716019236003076,
+      "learning_rate": 0.00019928213552157184,
+      "loss": 18.9287,
+      "step": 1150
+    },
+    {
+      "epoch": 0.20126775956284154,
+      "grad_norm": 2.1657383221616264,
+      "learning_rate": 0.00019927987547048903,
+      "loss": 18.5415,
+      "step": 1151
+    },
+    {
+      "epoch": 0.20144262295081966,
+      "grad_norm": 1.6139825445631686,
+      "learning_rate": 0.00019927761188019466,
+      "loss": 18.9942,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2016174863387978,
+      "grad_norm": 1.5855999325729664,
+      "learning_rate": 0.00019927534475076944,
+      "loss": 18.8998,
+      "step": 1153
+    },
+    {
+      "epoch": 0.20179234972677595,
+      "grad_norm": 2.217331694362355,
+      "learning_rate": 0.00019927307408229417,
+      "loss": 19.1417,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2019672131147541,
+      "grad_norm": 2.665658907239602,
+      "learning_rate": 0.00019927079987484977,
+      "loss": 18.9151,
+      "step": 1155
+    },
+    {
+      "epoch": 0.20214207650273225,
+      "grad_norm": 2.269017721243067,
+      "learning_rate": 0.00019926852212851737,
+      "loss": 19.0009,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2023169398907104,
+      "grad_norm": 1.9063917001271249,
+      "learning_rate": 0.00019926624084337812,
+      "loss": 18.5359,
+      "step": 1157
+    },
+    {
+      "epoch": 0.20249180327868851,
+      "grad_norm": 3.113359238391419,
+      "learning_rate": 0.00019926395601951337,
+      "loss": 18.9488,
+      "step": 1158
+    },
+    {
+      "epoch": 0.20266666666666666,
+      "grad_norm": 1.6518204584456844,
+      "learning_rate": 0.00019926166765700458,
+      "loss": 18.2722,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2028415300546448,
+      "grad_norm": 2.6998834703151986,
+      "learning_rate": 0.0001992593757559333,
+      "loss": 18.7151,
+      "step": 1160
+    },
+    {
+      "epoch": 0.20301639344262296,
+      "grad_norm": 1.8425482552325754,
+      "learning_rate": 0.00019925708031638122,
+      "loss": 18.9176,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2031912568306011,
+      "grad_norm": 1.9918570432213742,
+      "learning_rate": 0.00019925478133843024,
+      "loss": 18.7401,
+      "step": 1162
+    },
+    {
+      "epoch": 0.20336612021857922,
+      "grad_norm": 1.6525951018555898,
+      "learning_rate": 0.00019925247882216225,
+      "loss": 18.9335,
+      "step": 1163
+    },
+    {
+      "epoch": 0.20354098360655737,
+      "grad_norm": 1.840069438862545,
+      "learning_rate": 0.00019925017276765936,
+      "loss": 18.8434,
+      "step": 1164
+    },
+    {
+      "epoch": 0.20371584699453552,
+      "grad_norm": 1.573881645451688,
+      "learning_rate": 0.00019924786317500375,
+      "loss": 18.6015,
+      "step": 1165
+    },
+    {
+      "epoch": 0.20389071038251366,
+      "grad_norm": 1.5145775631961997,
+      "learning_rate": 0.0001992455500442778,
+      "loss": 18.5148,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2040655737704918,
+      "grad_norm": 1.7808038572572418,
+      "learning_rate": 0.00019924323337556395,
+      "loss": 18.9369,
+      "step": 1167
+    },
+    {
+      "epoch": 0.20424043715846996,
+      "grad_norm": 1.513468973275696,
+      "learning_rate": 0.00019924091316894478,
+      "loss": 18.8236,
+      "step": 1168
+    },
+    {
+      "epoch": 0.20441530054644808,
+      "grad_norm": 1.5689351333685422,
+      "learning_rate": 0.00019923858942450299,
+      "loss": 18.7761,
+      "step": 1169
+    },
+    {
+      "epoch": 0.20459016393442622,
+      "grad_norm": 1.7306169722182825,
+      "learning_rate": 0.00019923626214232146,
+      "loss": 18.831,
+      "step": 1170
+    },
+    {
+      "epoch": 0.20476502732240437,
+      "grad_norm": 1.469078882143747,
+      "learning_rate": 0.0001992339313224831,
+      "loss": 18.4921,
+      "step": 1171
+    },
+    {
+      "epoch": 0.20493989071038252,
+      "grad_norm": 1.5163752866757128,
+      "learning_rate": 0.00019923159696507106,
+      "loss": 18.6962,
+      "step": 1172
+    },
+    {
+      "epoch": 0.20511475409836066,
+      "grad_norm": 1.6005477793450589,
+      "learning_rate": 0.00019922925907016852,
+      "loss": 18.8574,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2052896174863388,
+      "grad_norm": 1.3824623364984747,
+      "learning_rate": 0.0001992269176378588,
+      "loss": 18.5966,
+      "step": 1174
+    },
+    {
+      "epoch": 0.20546448087431693,
+      "grad_norm": 1.538047530173973,
+      "learning_rate": 0.00019922457266822546,
+      "loss": 18.5884,
+      "step": 1175
+    },
+    {
+      "epoch": 0.20563934426229508,
+      "grad_norm": 1.464443451280581,
+      "learning_rate": 0.00019922222416135198,
+      "loss": 18.5474,
+      "step": 1176
+    },
+    {
+      "epoch": 0.20581420765027322,
+      "grad_norm": 1.3681765153093244,
+      "learning_rate": 0.00019921987211732217,
+      "loss": 18.5981,
+      "step": 1177
+    },
+    {
+      "epoch": 0.20598907103825137,
+      "grad_norm": 1.488481882622922,
+      "learning_rate": 0.00019921751653621982,
+      "loss": 18.8429,
+      "step": 1178
+    },
+    {
+      "epoch": 0.20616393442622952,
+      "grad_norm": 1.4113474881023211,
+      "learning_rate": 0.00019921515741812894,
+      "loss": 18.0455,
+      "step": 1179
+    },
+    {
+      "epoch": 0.20633879781420766,
+      "grad_norm": 1.592254559105652,
+      "learning_rate": 0.00019921279476313363,
+      "loss": 18.4208,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20651366120218578,
+      "grad_norm": 1.9590757411489361,
+      "learning_rate": 0.00019921042857131812,
+      "loss": 18.7016,
+      "step": 1181
+    },
+    {
+      "epoch": 0.20668852459016393,
+      "grad_norm": 1.904092171111077,
+      "learning_rate": 0.0001992080588427667,
+      "loss": 18.7116,
+      "step": 1182
+    },
+    {
+      "epoch": 0.20686338797814208,
+      "grad_norm": 1.6796932675803176,
+      "learning_rate": 0.00019920568557756392,
+      "loss": 18.5636,
+      "step": 1183
+    },
+    {
+      "epoch": 0.20703825136612022,
+      "grad_norm": 1.5786616124246278,
+      "learning_rate": 0.00019920330877579436,
+      "loss": 18.5112,
+      "step": 1184
+    },
+    {
+      "epoch": 0.20721311475409837,
+      "grad_norm": 1.8269944973608587,
+      "learning_rate": 0.00019920092843754274,
+      "loss": 18.2574,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2073879781420765,
+      "grad_norm": 2.039381330242516,
+      "learning_rate": 0.00019919854456289392,
+      "loss": 18.2247,
+      "step": 1186
+    },
+    {
+      "epoch": 0.20756284153005464,
+      "grad_norm": 2.0537352845047874,
+      "learning_rate": 0.00019919615715193291,
+      "loss": 18.531,
+      "step": 1187
+    },
+    {
+      "epoch": 0.20773770491803278,
+      "grad_norm": 1.6270827497368314,
+      "learning_rate": 0.0001991937662047448,
+      "loss": 18.593,
+      "step": 1188
+    },
+    {
+      "epoch": 0.20791256830601093,
+      "grad_norm": 1.6098846340891355,
+      "learning_rate": 0.0001991913717214148,
+      "loss": 18.3948,
+      "step": 1189
+    },
+    {
+      "epoch": 0.20808743169398908,
+      "grad_norm": 1.625205696601441,
+      "learning_rate": 0.00019918897370202829,
+      "loss": 18.4675,
+      "step": 1190
+    },
+    {
+      "epoch": 0.20826229508196722,
+      "grad_norm": 2.2108550639276445,
+      "learning_rate": 0.00019918657214667077,
+      "loss": 18.3112,
+      "step": 1191
+    },
+    {
+      "epoch": 0.20843715846994534,
+      "grad_norm": 3.0003819294524563,
+      "learning_rate": 0.00019918416705542782,
+      "loss": 18.3037,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2086120218579235,
+      "grad_norm": 2.4698783785197005,
+      "learning_rate": 0.00019918175842838522,
+      "loss": 18.4743,
+      "step": 1193
+    },
+    {
+      "epoch": 0.20878688524590164,
+      "grad_norm": 1.487235940421633,
+      "learning_rate": 0.00019917934626562878,
+      "loss": 17.9449,
+      "step": 1194
+    },
+    {
+      "epoch": 0.20896174863387978,
+      "grad_norm": 1.6926586808959825,
+      "learning_rate": 0.00019917693056724456,
+      "loss": 18.5681,
+      "step": 1195
+    },
+    {
+      "epoch": 0.20913661202185793,
+      "grad_norm": 1.7072470457128879,
+      "learning_rate": 0.00019917451133331864,
+      "loss": 18.0071,
+      "step": 1196
+    },
+    {
+      "epoch": 0.20931147540983608,
+      "grad_norm": 1.541243216063536,
+      "learning_rate": 0.00019917208856393725,
+      "loss": 18.1567,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2094863387978142,
+      "grad_norm": 1.5689291099581495,
+      "learning_rate": 0.00019916966225918677,
+      "loss": 18.2568,
+      "step": 1198
+    },
+    {
+      "epoch": 0.20966120218579234,
+      "grad_norm": 1.5897752037646073,
+      "learning_rate": 0.0001991672324191537,
+      "loss": 18.5607,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2098360655737705,
+      "grad_norm": 1.5930209497938275,
+      "learning_rate": 0.00019916479904392466,
+      "loss": 18.6538,
+      "step": 1200
+    },
+    {
+      "epoch": 0.21001092896174864,
+      "grad_norm": 1.7442768372431126,
+      "learning_rate": 0.0001991623621335864,
+      "loss": 18.194,
+      "step": 1201
+    },
+    {
+      "epoch": 0.21018579234972679,
+      "grad_norm": 1.5207730489229672,
+      "learning_rate": 0.0001991599216882258,
+      "loss": 18.6341,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2103606557377049,
+      "grad_norm": 1.9988805570283688,
+      "learning_rate": 0.00019915747770792982,
+      "loss": 18.1322,
+      "step": 1203
+    },
+    {
+      "epoch": 0.21053551912568305,
+      "grad_norm": 2.425260922465687,
+      "learning_rate": 0.00019915503019278562,
+      "loss": 18.4211,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2107103825136612,
+      "grad_norm": 2.4464248501420944,
+      "learning_rate": 0.00019915257914288046,
+      "loss": 18.5717,
+      "step": 1205
+    },
+    {
+      "epoch": 0.21088524590163935,
+      "grad_norm": 1.7884373601059156,
+      "learning_rate": 0.0001991501245583017,
+      "loss": 18.2678,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2110601092896175,
+      "grad_norm": 1.4806317207558461,
+      "learning_rate": 0.00019914766643913684,
+      "loss": 18.3399,
+      "step": 1207
+    },
+    {
+      "epoch": 0.21123497267759564,
+      "grad_norm": 1.710130759567135,
+      "learning_rate": 0.0001991452047854735,
+      "loss": 17.817,
+      "step": 1208
+    },
+    {
+      "epoch": 0.21140983606557376,
+      "grad_norm": 1.5483423450852498,
+      "learning_rate": 0.00019914273959739946,
+      "loss": 18.152,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2115846994535519,
+      "grad_norm": 1.6573481456335872,
+      "learning_rate": 0.00019914027087500257,
+      "loss": 18.4426,
+      "step": 1210
+    },
+    {
+      "epoch": 0.21175956284153005,
+      "grad_norm": 1.7524810804791924,
+      "learning_rate": 0.0001991377986183709,
+      "loss": 18.3412,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2119344262295082,
+      "grad_norm": 1.9595637730897768,
+      "learning_rate": 0.0001991353228275925,
+      "loss": 18.2901,
+      "step": 1212
+    },
+    {
+      "epoch": 0.21210928961748635,
+      "grad_norm": 1.9874217414033413,
+      "learning_rate": 0.00019913284350275568,
+      "loss": 18.3349,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2122841530054645,
+      "grad_norm": 1.7188878125595333,
+      "learning_rate": 0.0001991303606439488,
+      "loss": 18.2034,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2124590163934426,
+      "grad_norm": 2.973583145961686,
+      "learning_rate": 0.00019912787425126042,
+      "loss": 18.2879,
+      "step": 1215
+    },
+    {
+      "epoch": 0.21263387978142076,
+      "grad_norm": 3.8426544920451984,
+      "learning_rate": 0.00019912538432477912,
+      "loss": 17.9707,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2128087431693989,
+      "grad_norm": 2.3684423979092903,
+      "learning_rate": 0.0001991228908645937,
+      "loss": 18.0538,
+      "step": 1217
+    },
+    {
+      "epoch": 0.21298360655737705,
+      "grad_norm": 2.057033147645001,
+      "learning_rate": 0.000199120393870793,
+      "loss": 18.0997,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2131584699453552,
+      "grad_norm": 2.2427209096690737,
+      "learning_rate": 0.00019911789334346608,
+      "loss": 18.4081,
+      "step": 1219
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 1.7183706095820295,
+      "learning_rate": 0.00019911538928270205,
+      "loss": 18.1987,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21350819672131147,
+      "grad_norm": 1.8689242139152866,
+      "learning_rate": 0.00019911288168859023,
+      "loss": 18.4589,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2136830601092896,
+      "grad_norm": 1.744210846701776,
+      "learning_rate": 0.00019911037056121998,
+      "loss": 18.0588,
+      "step": 1222
+    },
+    {
+      "epoch": 0.21385792349726776,
+      "grad_norm": 1.544536015230708,
+      "learning_rate": 0.0001991078559006808,
+      "loss": 18.3034,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2140327868852459,
+      "grad_norm": 1.8682832025672105,
+      "learning_rate": 0.00019910533770706236,
+      "loss": 17.8858,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21420765027322405,
+      "grad_norm": 1.5884926876069607,
+      "learning_rate": 0.00019910281598045445,
+      "loss": 18.005,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21438251366120217,
+      "grad_norm": 1.7700040914399007,
+      "learning_rate": 0.0001991002907209469,
+      "loss": 18.3214,
+      "step": 1226
+    },
+    {
+      "epoch": 0.21455737704918032,
+      "grad_norm": 1.6023868047110896,
+      "learning_rate": 0.00019909776192862978,
+      "loss": 18.2684,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21473224043715847,
+      "grad_norm": 1.6770195583665282,
+      "learning_rate": 0.00019909522960359324,
+      "loss": 18.3238,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21490710382513661,
+      "grad_norm": 1.7162458117513508,
+      "learning_rate": 0.0001990926937459275,
+      "loss": 18.0089,
+      "step": 1229
+    },
+    {
+      "epoch": 0.21508196721311476,
+      "grad_norm": 1.640833863568782,
+      "learning_rate": 0.0001990901543557231,
+      "loss": 17.7868,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2152568306010929,
+      "grad_norm": 1.7031949439011524,
+      "learning_rate": 0.0001990876114330704,
+      "loss": 18.488,
+      "step": 1231
+    },
+    {
+      "epoch": 0.21543169398907103,
+      "grad_norm": 2.0983634201253496,
+      "learning_rate": 0.00019908506497806015,
+      "loss": 18.2721,
+      "step": 1232
+    },
+    {
+      "epoch": 0.21560655737704917,
+      "grad_norm": 2.1299073367334005,
+      "learning_rate": 0.00019908251499078307,
+      "loss": 17.882,
+      "step": 1233
+    },
+    {
+      "epoch": 0.21578142076502732,
+      "grad_norm": 1.630573615364069,
+      "learning_rate": 0.00019907996147133013,
+      "loss": 17.8605,
+      "step": 1234
+    },
+    {
+      "epoch": 0.21595628415300547,
+      "grad_norm": 2.4397802097188577,
+      "learning_rate": 0.0001990774044197923,
+      "loss": 18.1743,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21613114754098361,
+      "grad_norm": 1.8965875453049164,
+      "learning_rate": 0.00019907484383626078,
+      "loss": 17.9163,
+      "step": 1236
+    },
+    {
+      "epoch": 0.21630601092896176,
+      "grad_norm": 1.8437711783533537,
+      "learning_rate": 0.00019907227972082683,
+      "loss": 17.876,
+      "step": 1237
+    },
+    {
+      "epoch": 0.21648087431693988,
+      "grad_norm": 2.033910097519758,
+      "learning_rate": 0.00019906971207358186,
+      "loss": 17.9927,
+      "step": 1238
+    },
+    {
+      "epoch": 0.21665573770491803,
+      "grad_norm": 1.6762065406378288,
+      "learning_rate": 0.0001990671408946174,
+      "loss": 18.0149,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21683060109289617,
+      "grad_norm": 1.653459148459724,
+      "learning_rate": 0.00019906456618402514,
+      "loss": 18.0915,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21700546448087432,
+      "grad_norm": 1.833561830547301,
+      "learning_rate": 0.0001990619879418968,
+      "loss": 17.5707,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21718032786885247,
+      "grad_norm": 1.8058492782580673,
+      "learning_rate": 0.00019905940616832436,
+      "loss": 18.3882,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2173551912568306,
+      "grad_norm": 1.6381839774895255,
+      "learning_rate": 0.00019905682086339982,
+      "loss": 17.7048,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21753005464480873,
+      "grad_norm": 1.6083886210000828,
+      "learning_rate": 0.00019905423202721536,
+      "loss": 17.8312,
+      "step": 1244
+    },
+    {
+      "epoch": 0.21770491803278688,
+      "grad_norm": 1.8211038639066914,
+      "learning_rate": 0.00019905163965986327,
+      "loss": 17.545,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21787978142076503,
+      "grad_norm": 1.7896913425339498,
+      "learning_rate": 0.00019904904376143593,
+      "loss": 17.8022,
+      "step": 1246
+    },
+    {
+      "epoch": 0.21805464480874318,
+      "grad_norm": 1.9705023305052887,
+      "learning_rate": 0.0001990464443320259,
+      "loss": 17.8544,
+      "step": 1247
+    },
+    {
+      "epoch": 0.21822950819672132,
+      "grad_norm": 1.7505119619608371,
+      "learning_rate": 0.00019904384137172588,
+      "loss": 17.9201,
+      "step": 1248
+    },
+    {
+      "epoch": 0.21840437158469944,
+      "grad_norm": 1.6189935787966083,
+      "learning_rate": 0.00019904123488062863,
+      "loss": 17.7838,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2185792349726776,
+      "grad_norm": 1.8709984960330022,
+      "learning_rate": 0.00019903862485882705,
+      "loss": 17.9602,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21875409836065574,
+      "grad_norm": 2.1047889714361183,
+      "learning_rate": 0.00019903601130641426,
+      "loss": 17.3404,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21892896174863388,
+      "grad_norm": 1.8392056895293873,
+      "learning_rate": 0.0001990333942234833,
+      "loss": 17.8909,
+      "step": 1252
+    },
+    {
+      "epoch": 0.21910382513661203,
+      "grad_norm": 1.6932890691320583,
+      "learning_rate": 0.00019903077361012761,
+      "loss": 17.7794,
+      "step": 1253
+    },
+    {
+      "epoch": 0.21927868852459018,
+      "grad_norm": 1.8385094806853013,
+      "learning_rate": 0.00019902814946644052,
+      "loss": 17.8148,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2194535519125683,
+      "grad_norm": 1.644093082861578,
+      "learning_rate": 0.0001990255217925156,
+      "loss": 17.5246,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21962841530054644,
+      "grad_norm": 1.631344563234371,
+      "learning_rate": 0.00019902289058844653,
+      "loss": 17.7096,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2198032786885246,
+      "grad_norm": 1.718813676733304,
+      "learning_rate": 0.00019902025585432712,
+      "loss": 18.0137,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21997814207650274,
+      "grad_norm": 1.7547281118009983,
+      "learning_rate": 0.00019901761759025125,
+      "loss": 17.9355,
+      "step": 1258
+    },
+    {
+      "epoch": 0.22015300546448088,
+      "grad_norm": 2.289250376663363,
+      "learning_rate": 0.00019901497579631304,
+      "loss": 17.8652,
+      "step": 1259
+    },
+    {
+      "epoch": 0.22032786885245903,
+      "grad_norm": 2.4017325019137914,
+      "learning_rate": 0.00019901233047260662,
+      "loss": 17.7282,
+      "step": 1260
+    },
+    {
+      "epoch": 0.22050273224043715,
+      "grad_norm": 1.8493740707136512,
+      "learning_rate": 0.0001990096816192263,
+      "loss": 17.6025,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2206775956284153,
+      "grad_norm": 1.7599584651373186,
+      "learning_rate": 0.0001990070292362665,
+      "loss": 17.822,
+      "step": 1262
+    },
+    {
+      "epoch": 0.22085245901639344,
+      "grad_norm": 1.6327265999815253,
+      "learning_rate": 0.00019900437332382177,
+      "loss": 17.7642,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2210273224043716,
+      "grad_norm": 1.920428871811497,
+      "learning_rate": 0.00019900171388198686,
+      "loss": 17.8858,
+      "step": 1264
+    },
+    {
+      "epoch": 0.22120218579234974,
+      "grad_norm": 1.6020612093021442,
+      "learning_rate": 0.0001989990509108565,
+      "loss": 17.6456,
+      "step": 1265
+    },
+    {
+      "epoch": 0.22137704918032786,
+      "grad_norm": 1.6959567944495955,
+      "learning_rate": 0.0001989963844105256,
+      "loss": 17.5012,
+      "step": 1266
+    },
+    {
+      "epoch": 0.221551912568306,
+      "grad_norm": 1.7001455712328042,
+      "learning_rate": 0.0001989937143810893,
+      "loss": 17.5914,
+      "step": 1267
+    },
+    {
+      "epoch": 0.22172677595628415,
+      "grad_norm": 1.7944419893726093,
+      "learning_rate": 0.00019899104082264274,
+      "loss": 17.3426,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2219016393442623,
+      "grad_norm": 2.1651832329396674,
+      "learning_rate": 0.00019898836373528122,
+      "loss": 17.6099,
+      "step": 1269
+    },
+    {
+      "epoch": 0.22207650273224044,
+      "grad_norm": 1.9481155119382196,
+      "learning_rate": 0.0001989856831191002,
+      "loss": 17.6136,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2222513661202186,
+      "grad_norm": 1.6769539271759848,
+      "learning_rate": 0.00019898299897419522,
+      "loss": 17.3916,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2224262295081967,
+      "grad_norm": 1.5951437962250405,
+      "learning_rate": 0.000198980311300662,
+      "loss": 16.9676,
+      "step": 1272
+    },
+    {
+      "epoch": 0.22260109289617486,
+      "grad_norm": 1.5608574347643474,
+      "learning_rate": 0.0001989776200985963,
+      "loss": 17.7566,
+      "step": 1273
+    },
+    {
+      "epoch": 0.222775956284153,
+      "grad_norm": 1.8483897422797706,
+      "learning_rate": 0.0001989749253680941,
+      "loss": 17.6427,
+      "step": 1274
+    },
+    {
+      "epoch": 0.22295081967213115,
+      "grad_norm": 1.9347374568674047,
+      "learning_rate": 0.00019897222710925144,
+      "loss": 17.6276,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2231256830601093,
+      "grad_norm": 1.688495068280619,
+      "learning_rate": 0.00019896952532216452,
+      "loss": 17.0668,
+      "step": 1276
+    },
+    {
+      "epoch": 0.22330054644808744,
+      "grad_norm": 1.6391874871200922,
+      "learning_rate": 0.00019896682000692967,
+      "loss": 17.7267,
+      "step": 1277
+    },
+    {
+      "epoch": 0.22347540983606556,
+      "grad_norm": 1.7300087354179783,
+      "learning_rate": 0.00019896411116364333,
+      "loss": 17.3771,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2236502732240437,
+      "grad_norm": 1.6645629653545007,
+      "learning_rate": 0.00019896139879240204,
+      "loss": 17.7611,
+      "step": 1279
+    },
+    {
+      "epoch": 0.22382513661202186,
+      "grad_norm": 1.7633976036273051,
+      "learning_rate": 0.00019895868289330248,
+      "loss": 17.7086,
+      "step": 1280
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 1.8688426766712505,
+      "learning_rate": 0.0001989559634664415,
+      "loss": 17.4243,
+      "step": 1281
+    },
+    {
+      "epoch": 0.22417486338797815,
+      "grad_norm": 1.3899293223991285,
+      "learning_rate": 0.00019895324051191608,
+      "loss": 17.234,
+      "step": 1282
+    },
+    {
+      "epoch": 0.22434972677595627,
+      "grad_norm": 1.7512023279627533,
+      "learning_rate": 0.00019895051402982322,
+      "loss": 17.5204,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22452459016393442,
+      "grad_norm": 1.464026751253038,
+      "learning_rate": 0.00019894778402026016,
+      "loss": 17.1976,
+      "step": 1284
+    },
+    {
+      "epoch": 0.22469945355191256,
+      "grad_norm": 2.4639000614810107,
+      "learning_rate": 0.0001989450504833242,
+      "loss": 17.584,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2248743169398907,
+      "grad_norm": 2.738565847341316,
+      "learning_rate": 0.00019894231341911281,
+      "loss": 17.41,
+      "step": 1286
+    },
+    {
+      "epoch": 0.22504918032786886,
+      "grad_norm": 1.5431557913687002,
+      "learning_rate": 0.00019893957282772352,
+      "loss": 17.3984,
+      "step": 1287
+    },
+    {
+      "epoch": 0.225224043715847,
+      "grad_norm": 2.420160649408186,
+      "learning_rate": 0.00019893682870925409,
+      "loss": 17.4385,
+      "step": 1288
+    },
+    {
+      "epoch": 0.22539890710382512,
+      "grad_norm": 2.029306707366613,
+      "learning_rate": 0.0001989340810638023,
+      "loss": 17.3905,
+      "step": 1289
+    },
+    {
+      "epoch": 0.22557377049180327,
+      "grad_norm": 1.9772235046184272,
+      "learning_rate": 0.00019893132989146607,
+      "loss": 17.4069,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22574863387978142,
+      "grad_norm": 2.824378908171905,
+      "learning_rate": 0.0001989285751923436,
+      "loss": 17.6207,
+      "step": 1291
+    },
+    {
+      "epoch": 0.22592349726775957,
+      "grad_norm": 2.2631555417561784,
+      "learning_rate": 0.00019892581696653294,
+      "loss": 17.262,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2260983606557377,
+      "grad_norm": 1.9338772369719466,
+      "learning_rate": 0.0001989230552141325,
+      "loss": 17.6609,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22627322404371586,
+      "grad_norm": 1.8161654762442379,
+      "learning_rate": 0.00019892028993524075,
+      "loss": 17.4384,
+      "step": 1294
+    },
+    {
+      "epoch": 0.22644808743169398,
+      "grad_norm": 1.8276081427456043,
+      "learning_rate": 0.0001989175211299562,
+      "loss": 17.4305,
+      "step": 1295
+    },
+    {
+      "epoch": 0.22662295081967213,
+      "grad_norm": 1.7608593267036337,
+      "learning_rate": 0.00019891474879837762,
+      "loss": 17.4019,
+      "step": 1296
+    },
+    {
+      "epoch": 0.22679781420765027,
+      "grad_norm": 1.6793592327187377,
+      "learning_rate": 0.0001989119729406038,
+      "loss": 17.5015,
+      "step": 1297
+    },
+    {
+      "epoch": 0.22697267759562842,
+      "grad_norm": 1.786326267445353,
+      "learning_rate": 0.00019890919355673375,
+      "loss": 17.2351,
+      "step": 1298
+    },
+    {
+      "epoch": 0.22714754098360657,
+      "grad_norm": 1.9058710122057303,
+      "learning_rate": 0.00019890641064686647,
+      "loss": 17.7825,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2273224043715847,
+      "grad_norm": 1.624972438337584,
+      "learning_rate": 0.00019890362421110121,
+      "loss": 17.126,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22749726775956283,
+      "grad_norm": 1.8652025821837277,
+      "learning_rate": 0.00019890083424953733,
+      "loss": 17.4274,
+      "step": 1301
+    },
+    {
+      "epoch": 0.22767213114754098,
+      "grad_norm": 1.7778856917545003,
+      "learning_rate": 0.00019889804076227426,
+      "loss": 17.6084,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22784699453551913,
+      "grad_norm": 1.932142811042149,
+      "learning_rate": 0.00019889524374941156,
+      "loss": 17.2667,
+      "step": 1303
+    },
+    {
+      "epoch": 0.22802185792349727,
+      "grad_norm": 1.7743895202189173,
+      "learning_rate": 0.00019889244321104902,
+      "loss": 17.2274,
+      "step": 1304
+    },
+    {
+      "epoch": 0.22819672131147542,
+      "grad_norm": 1.778099931115376,
+      "learning_rate": 0.00019888963914728637,
+      "loss": 17.536,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22837158469945354,
+      "grad_norm": 1.9313188823912735,
+      "learning_rate": 0.00019888683155822365,
+      "loss": 17.4471,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2285464480874317,
+      "grad_norm": 1.6727240164493227,
+      "learning_rate": 0.0001988840204439609,
+      "loss": 17.2329,
+      "step": 1307
+    },
+    {
+      "epoch": 0.22872131147540983,
+      "grad_norm": 1.6353041354223645,
+      "learning_rate": 0.00019888120580459836,
+      "loss": 17.4505,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22889617486338798,
+      "grad_norm": 1.6207331391102167,
+      "learning_rate": 0.0001988783876402364,
+      "loss": 17.4067,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22907103825136613,
+      "grad_norm": 1.6044267719651608,
+      "learning_rate": 0.0001988755659509754,
+      "loss": 17.5522,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22924590163934427,
+      "grad_norm": 1.655731035325366,
+      "learning_rate": 0.000198872740736916,
+      "loss": 17.1713,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2294207650273224,
+      "grad_norm": 1.6139398220657495,
+      "learning_rate": 0.00019886991199815897,
+      "loss": 17.2409,
+      "step": 1312
+    },
+    {
+      "epoch": 0.22959562841530054,
+      "grad_norm": 1.655392452678039,
+      "learning_rate": 0.00019886707973480503,
+      "loss": 17.0416,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2297704918032787,
+      "grad_norm": 1.8950277005577771,
+      "learning_rate": 0.00019886424394695523,
+      "loss": 17.2721,
+      "step": 1314
+    },
+    {
+      "epoch": 0.22994535519125683,
+      "grad_norm": 1.794910738637396,
+      "learning_rate": 0.00019886140463471066,
+      "loss": 17.4707,
+      "step": 1315
+    },
+    {
+      "epoch": 0.23012021857923498,
+      "grad_norm": 1.5432502071183682,
+      "learning_rate": 0.00019885856179817252,
+      "loss": 17.2731,
+      "step": 1316
+    },
+    {
+      "epoch": 0.23029508196721313,
+      "grad_norm": 1.7407515446614226,
+      "learning_rate": 0.00019885571543744215,
+      "loss": 17.1709,
+      "step": 1317
+    },
+    {
+      "epoch": 0.23046994535519125,
+      "grad_norm": 1.6213296165681714,
+      "learning_rate": 0.00019885286555262102,
+      "loss": 17.3267,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2306448087431694,
+      "grad_norm": 1.6735581398255512,
+      "learning_rate": 0.00019885001214381073,
+      "loss": 17.2875,
+      "step": 1319
+    },
+    {
+      "epoch": 0.23081967213114754,
+      "grad_norm": 1.470303207359532,
+      "learning_rate": 0.000198847155211113,
+      "loss": 17.1371,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2309945355191257,
+      "grad_norm": 2.083031913699354,
+      "learning_rate": 0.00019884429475462966,
+      "loss": 17.2774,
+      "step": 1321
+    },
+    {
+      "epoch": 0.23116939890710383,
+      "grad_norm": 1.9373589407889298,
+      "learning_rate": 0.00019884143077446273,
+      "loss": 17.5009,
+      "step": 1322
+    },
+    {
+      "epoch": 0.23134426229508195,
+      "grad_norm": 1.8742325077228446,
+      "learning_rate": 0.0001988385632707143,
+      "loss": 17.0952,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2315191256830601,
+      "grad_norm": 1.7240117522504999,
+      "learning_rate": 0.0001988356922434865,
+      "loss": 17.0477,
+      "step": 1324
+    },
+    {
+      "epoch": 0.23169398907103825,
+      "grad_norm": 1.99384480049891,
+      "learning_rate": 0.0001988328176928818,
+      "loss": 17.3025,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2318688524590164,
+      "grad_norm": 1.7436572314134062,
+      "learning_rate": 0.00019882993961900262,
+      "loss": 16.7,
+      "step": 1326
+    },
+    {
+      "epoch": 0.23204371584699454,
+      "grad_norm": 2.540812718464482,
+      "learning_rate": 0.00019882705802195153,
+      "loss": 17.3556,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2322185792349727,
+      "grad_norm": 2.3260758117375904,
+      "learning_rate": 0.00019882417290183134,
+      "loss": 17.2322,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2323934426229508,
+      "grad_norm": 1.6997926652519386,
+      "learning_rate": 0.0001988212842587448,
+      "loss": 17.1036,
+      "step": 1329
+    },
+    {
+      "epoch": 0.23256830601092895,
+      "grad_norm": 1.7359654959939863,
+      "learning_rate": 0.00019881839209279494,
+      "loss": 17.2727,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2327431693989071,
+      "grad_norm": 1.5791617588912434,
+      "learning_rate": 0.00019881549640408488,
+      "loss": 17.1704,
+      "step": 1331
+    },
+    {
+      "epoch": 0.23291803278688525,
+      "grad_norm": 1.6788214157850958,
+      "learning_rate": 0.00019881259719271784,
+      "loss": 16.949,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2330928961748634,
+      "grad_norm": 1.542082914282521,
+      "learning_rate": 0.00019880969445879712,
+      "loss": 16.7084,
+      "step": 1333
+    },
+    {
+      "epoch": 0.23326775956284154,
+      "grad_norm": 1.712270674503248,
+      "learning_rate": 0.00019880678820242625,
+      "loss": 17.0553,
+      "step": 1334
+    },
+    {
+      "epoch": 0.23344262295081966,
+      "grad_norm": 1.5887328135574468,
+      "learning_rate": 0.00019880387842370885,
+      "loss": 16.8962,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2336174863387978,
+      "grad_norm": 1.6494694943195114,
+      "learning_rate": 0.00019880096512274863,
+      "loss": 17.1302,
+      "step": 1336
+    },
+    {
+      "epoch": 0.23379234972677596,
+      "grad_norm": 1.7404823801938127,
+      "learning_rate": 0.00019879804829964939,
+      "loss": 17.2397,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2339672131147541,
+      "grad_norm": 1.7599195516022248,
+      "learning_rate": 0.0001987951279545152,
+      "loss": 16.7862,
+      "step": 1338
+    },
+    {
+      "epoch": 0.23414207650273225,
+      "grad_norm": 1.6053404513984595,
+      "learning_rate": 0.0001987922040874501,
+      "loss": 16.2704,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2343169398907104,
+      "grad_norm": 1.7997354208063487,
+      "learning_rate": 0.0001987892766985584,
+      "loss": 17.0718,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23449180327868852,
+      "grad_norm": 1.539769055757127,
+      "learning_rate": 0.00019878634578794437,
+      "loss": 16.8221,
+      "step": 1341
+    },
+    {
+      "epoch": 0.23466666666666666,
+      "grad_norm": 1.6122219647164933,
+      "learning_rate": 0.00019878341135571255,
+      "loss": 17.1259,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2348415300546448,
+      "grad_norm": 1.6030612124192751,
+      "learning_rate": 0.0001987804734019675,
+      "loss": 17.1963,
+      "step": 1343
+    },
+    {
+      "epoch": 0.23501639344262296,
+      "grad_norm": 1.8015203042734653,
+      "learning_rate": 0.000198777531926814,
+      "loss": 16.9472,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2351912568306011,
+      "grad_norm": 2.2728836058020043,
+      "learning_rate": 0.00019877458693035691,
+      "loss": 16.8443,
+      "step": 1345
+    },
+    {
+      "epoch": 0.23536612021857922,
+      "grad_norm": 1.5833056315222012,
+      "learning_rate": 0.0001987716384127012,
+      "loss": 17.0806,
+      "step": 1346
+    },
+    {
+      "epoch": 0.23554098360655737,
+      "grad_norm": 1.8214299897702169,
+      "learning_rate": 0.00019876868637395197,
+      "loss": 16.9154,
+      "step": 1347
+    },
+    {
+      "epoch": 0.23571584699453552,
+      "grad_norm": 1.7298406242037496,
+      "learning_rate": 0.00019876573081421448,
+      "loss": 16.9033,
+      "step": 1348
+    },
+    {
+      "epoch": 0.23589071038251366,
+      "grad_norm": 1.7011866564150073,
+      "learning_rate": 0.00019876277173359408,
+      "loss": 17.4219,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2360655737704918,
+      "grad_norm": 1.6264583805941504,
+      "learning_rate": 0.00019875980913219626,
+      "loss": 17.0413,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23624043715846996,
+      "grad_norm": 1.6162391741129394,
+      "learning_rate": 0.00019875684301012662,
+      "loss": 16.9863,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23641530054644808,
+      "grad_norm": 1.7227521886564017,
+      "learning_rate": 0.00019875387336749093,
+      "loss": 17.3484,
+      "step": 1352
+    },
+    {
+      "epoch": 0.23659016393442622,
+      "grad_norm": 1.6255461972622907,
+      "learning_rate": 0.00019875090020439503,
+      "loss": 16.842,
+      "step": 1353
+    },
+    {
+      "epoch": 0.23676502732240437,
+      "grad_norm": 1.6768322156040876,
+      "learning_rate": 0.00019874792352094493,
+      "loss": 16.9243,
+      "step": 1354
+    },
+    {
+      "epoch": 0.23693989071038252,
+      "grad_norm": 1.6449519431949888,
+      "learning_rate": 0.0001987449433172467,
+      "loss": 16.8335,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23711475409836066,
+      "grad_norm": 1.5345014520681837,
+      "learning_rate": 0.00019874195959340663,
+      "loss": 16.9805,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2372896174863388,
+      "grad_norm": 1.836348128958707,
+      "learning_rate": 0.00019873897234953106,
+      "loss": 16.9078,
+      "step": 1357
+    },
+    {
+      "epoch": 0.23746448087431693,
+      "grad_norm": 1.75814796646439,
+      "learning_rate": 0.0001987359815857265,
+      "loss": 16.8962,
+      "step": 1358
+    },
+    {
+      "epoch": 0.23763934426229508,
+      "grad_norm": 1.8708936459692298,
+      "learning_rate": 0.00019873298730209955,
+      "loss": 16.8663,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23781420765027322,
+      "grad_norm": 1.8208286373185967,
+      "learning_rate": 0.00019872998949875697,
+      "loss": 16.8012,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23798907103825137,
+      "grad_norm": 1.7471085480368367,
+      "learning_rate": 0.0001987269881758056,
+      "loss": 16.7664,
+      "step": 1361
+    },
+    {
+      "epoch": 0.23816393442622952,
+      "grad_norm": 1.7681544148500326,
+      "learning_rate": 0.00019872398333335248,
+      "loss": 16.7062,
+      "step": 1362
+    },
+    {
+      "epoch": 0.23833879781420764,
+      "grad_norm": 1.5890627685111178,
+      "learning_rate": 0.00019872097497150465,
+      "loss": 16.4899,
+      "step": 1363
+    },
+    {
+      "epoch": 0.23851366120218578,
+      "grad_norm": 1.6325418847282922,
+      "learning_rate": 0.00019871796309036945,
+      "loss": 16.9979,
+      "step": 1364
+    },
+    {
+      "epoch": 0.23868852459016393,
+      "grad_norm": 1.600714662081447,
+      "learning_rate": 0.00019871494769005418,
+      "loss": 16.5346,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23886338797814208,
+      "grad_norm": 1.5964912083506693,
+      "learning_rate": 0.00019871192877066638,
+      "loss": 17.0542,
+      "step": 1366
+    },
+    {
+      "epoch": 0.23903825136612022,
+      "grad_norm": 1.6213508072702978,
+      "learning_rate": 0.00019870890633231364,
+      "loss": 16.8779,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23921311475409837,
+      "grad_norm": 1.4553922471655296,
+      "learning_rate": 0.00019870588037510374,
+      "loss": 16.8086,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2393879781420765,
+      "grad_norm": 1.6490599898927343,
+      "learning_rate": 0.0001987028508991445,
+      "loss": 17.1314,
+      "step": 1369
+    },
+    {
+      "epoch": 0.23956284153005464,
+      "grad_norm": 1.517437130801965,
+      "learning_rate": 0.00019869981790454397,
+      "loss": 16.1766,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23973770491803278,
+      "grad_norm": 1.615341317799307,
+      "learning_rate": 0.00019869678139141023,
+      "loss": 16.7646,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23991256830601093,
+      "grad_norm": 1.590400428709324,
+      "learning_rate": 0.00019869374135985158,
+      "loss": 16.5618,
+      "step": 1372
+    },
+    {
+      "epoch": 0.24008743169398908,
+      "grad_norm": 1.5278257964062274,
+      "learning_rate": 0.00019869069780997634,
+      "loss": 16.6469,
+      "step": 1373
+    },
+    {
+      "epoch": 0.24026229508196723,
+      "grad_norm": 1.5753799318363597,
+      "learning_rate": 0.000198687650741893,
+      "loss": 16.9666,
+      "step": 1374
+    },
+    {
+      "epoch": 0.24043715846994534,
+      "grad_norm": 1.5620270659972193,
+      "learning_rate": 0.00019868460015571025,
+      "loss": 16.7647,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2406120218579235,
+      "grad_norm": 1.7871696533070531,
+      "learning_rate": 0.0001986815460515368,
+      "loss": 16.6256,
+      "step": 1376
+    },
+    {
+      "epoch": 0.24078688524590164,
+      "grad_norm": 1.8327805669827983,
+      "learning_rate": 0.00019867848842948155,
+      "loss": 16.609,
+      "step": 1377
+    },
+    {
+      "epoch": 0.24096174863387979,
+      "grad_norm": 1.624768002630991,
+      "learning_rate": 0.00019867542728965347,
+      "loss": 16.9837,
+      "step": 1378
+    },
+    {
+      "epoch": 0.24113661202185793,
+      "grad_norm": 2.0696060863319583,
+      "learning_rate": 0.00019867236263216168,
+      "loss": 16.7008,
+      "step": 1379
+    },
+    {
+      "epoch": 0.24131147540983608,
+      "grad_norm": 1.7573924298731205,
+      "learning_rate": 0.00019866929445711546,
+      "loss": 16.7066,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2414863387978142,
+      "grad_norm": 1.5301952890816586,
+      "learning_rate": 0.0001986662227646242,
+      "loss": 16.61,
+      "step": 1381
+    },
+    {
+      "epoch": 0.24166120218579235,
+      "grad_norm": 1.7327672155659475,
+      "learning_rate": 0.00019866314755479734,
+      "loss": 16.7784,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2418360655737705,
+      "grad_norm": 1.783550785350083,
+      "learning_rate": 0.00019866006882774455,
+      "loss": 16.8326,
+      "step": 1383
+    },
+    {
+      "epoch": 0.24201092896174864,
+      "grad_norm": 1.5632910147137122,
+      "learning_rate": 0.0001986569865835756,
+      "loss": 16.9741,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2421857923497268,
+      "grad_norm": 1.600375947696716,
+      "learning_rate": 0.0001986539008224003,
+      "loss": 16.651,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2423606557377049,
+      "grad_norm": 1.549290940471902,
+      "learning_rate": 0.00019865081154432876,
+      "loss": 16.8505,
+      "step": 1386
+    },
+    {
+      "epoch": 0.24253551912568305,
+      "grad_norm": 1.4116630431592063,
+      "learning_rate": 0.000198647718749471,
+      "loss": 16.4504,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2427103825136612,
+      "grad_norm": 1.6544425102477824,
+      "learning_rate": 0.00019864462243793736,
+      "loss": 16.6169,
+      "step": 1388
+    },
+    {
+      "epoch": 0.24288524590163935,
+      "grad_norm": 1.5332421610794331,
+      "learning_rate": 0.00019864152260983816,
+      "loss": 16.9787,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2430601092896175,
+      "grad_norm": 2.216998515428759,
+      "learning_rate": 0.00019863841926528395,
+      "loss": 16.667,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24323497267759564,
+      "grad_norm": 1.9717754512202508,
+      "learning_rate": 0.00019863531240438533,
+      "loss": 16.6358,
+      "step": 1391
+    },
+    {
+      "epoch": 0.24340983606557376,
+      "grad_norm": 1.92841224259518,
+      "learning_rate": 0.00019863220202725305,
+      "loss": 16.7788,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2435846994535519,
+      "grad_norm": 2.31165368277974,
+      "learning_rate": 0.00019862908813399804,
+      "loss": 16.7083,
+      "step": 1393
+    },
+    {
+      "epoch": 0.24375956284153005,
+      "grad_norm": 1.5286253878723064,
+      "learning_rate": 0.00019862597072473125,
+      "loss": 16.8764,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2439344262295082,
+      "grad_norm": 1.833785820528381,
+      "learning_rate": 0.00019862284979956383,
+      "loss": 16.8119,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24410928961748635,
+      "grad_norm": 1.5964578981745985,
+      "learning_rate": 0.00019861972535860706,
+      "loss": 16.4565,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2442841530054645,
+      "grad_norm": 1.795966630447667,
+      "learning_rate": 0.0001986165974019723,
+      "loss": 16.3924,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2444590163934426,
+      "grad_norm": 1.7113047616494266,
+      "learning_rate": 0.00019861346592977105,
+      "loss": 16.6651,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24463387978142076,
+      "grad_norm": 1.6497509415054772,
+      "learning_rate": 0.000198610330942115,
+      "loss": 16.6694,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2448087431693989,
+      "grad_norm": 1.5918429020430376,
+      "learning_rate": 0.00019860719243911584,
+      "loss": 16.678,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24498360655737705,
+      "grad_norm": 1.7977014026376188,
+      "learning_rate": 0.00019860405042088547,
+      "loss": 16.4494,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2451584699453552,
+      "grad_norm": 1.6280274446328187,
+      "learning_rate": 0.00019860090488753592,
+      "loss": 16.5664,
+      "step": 1402
+    },
+    {
+      "epoch": 0.24533333333333332,
+      "grad_norm": 1.7655648253831906,
+      "learning_rate": 0.0001985977558391793,
+      "loss": 16.5676,
+      "step": 1403
+    },
+    {
+      "epoch": 0.24550819672131147,
+      "grad_norm": 1.612891245501369,
+      "learning_rate": 0.00019859460327592787,
+      "loss": 16.6567,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2456830601092896,
+      "grad_norm": 1.6808946417601947,
+      "learning_rate": 0.00019859144719789406,
+      "loss": 16.1945,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24585792349726776,
+      "grad_norm": 1.8511539180347298,
+      "learning_rate": 0.00019858828760519032,
+      "loss": 16.9862,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2460327868852459,
+      "grad_norm": 1.5134646551494664,
+      "learning_rate": 0.00019858512449792936,
+      "loss": 16.5606,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24620765027322405,
+      "grad_norm": 1.6377232795150334,
+      "learning_rate": 0.00019858195787622387,
+      "loss": 16.452,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24638251366120217,
+      "grad_norm": 1.6699295383685244,
+      "learning_rate": 0.00019857878774018676,
+      "loss": 16.6267,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24655737704918032,
+      "grad_norm": 1.7262901806514335,
+      "learning_rate": 0.00019857561408993105,
+      "loss": 16.4822,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24673224043715847,
+      "grad_norm": 1.637375644945704,
+      "learning_rate": 0.00019857243692556986,
+      "loss": 16.4902,
+      "step": 1411
+    },
+    {
+      "epoch": 0.24690710382513661,
+      "grad_norm": 1.616342266209881,
+      "learning_rate": 0.00019856925624721648,
+      "loss": 16.6332,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24708196721311476,
+      "grad_norm": 1.720155852041632,
+      "learning_rate": 0.00019856607205498428,
+      "loss": 16.4121,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2472568306010929,
+      "grad_norm": 1.5386185326001645,
+      "learning_rate": 0.00019856288434898675,
+      "loss": 16.6239,
+      "step": 1414
+    },
+    {
+      "epoch": 0.24743169398907103,
+      "grad_norm": 1.8282187030133177,
+      "learning_rate": 0.00019855969312933754,
+      "loss": 16.6885,
+      "step": 1415
+    },
+    {
+      "epoch": 0.24760655737704917,
+      "grad_norm": 1.5316979572856355,
+      "learning_rate": 0.00019855649839615048,
+      "loss": 16.5405,
+      "step": 1416
+    },
+    {
+      "epoch": 0.24778142076502732,
+      "grad_norm": 1.6181185597941847,
+      "learning_rate": 0.00019855330014953936,
+      "loss": 16.7715,
+      "step": 1417
+    },
+    {
+      "epoch": 0.24795628415300547,
+      "grad_norm": 1.8236004622542563,
+      "learning_rate": 0.00019855009838961826,
+      "loss": 16.6527,
+      "step": 1418
+    },
+    {
+      "epoch": 0.24813114754098362,
+      "grad_norm": 1.6435210971822736,
+      "learning_rate": 0.00019854689311650127,
+      "loss": 16.729,
+      "step": 1419
+    },
+    {
+      "epoch": 0.24830601092896176,
+      "grad_norm": 1.8459963881717978,
+      "learning_rate": 0.0001985436843303027,
+      "loss": 16.5433,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24848087431693988,
+      "grad_norm": 2.3942226167046594,
+      "learning_rate": 0.00019854047203113692,
+      "loss": 16.6646,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24865573770491803,
+      "grad_norm": 1.476711020418244,
+      "learning_rate": 0.00019853725621911845,
+      "loss": 16.5425,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24883060109289618,
+      "grad_norm": 2.1611887384911745,
+      "learning_rate": 0.0001985340368943619,
+      "loss": 16.6415,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24900546448087432,
+      "grad_norm": 1.6131685343629776,
+      "learning_rate": 0.0001985308140569821,
+      "loss": 16.2748,
+      "step": 1424
+    },
+    {
+      "epoch": 0.24918032786885247,
+      "grad_norm": 2.2670231772711267,
+      "learning_rate": 0.00019852758770709387,
+      "loss": 16.6695,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2493551912568306,
+      "grad_norm": 2.1582796905077566,
+      "learning_rate": 0.00019852435784481224,
+      "loss": 16.7599,
+      "step": 1426
+    },
+    {
+      "epoch": 0.24953005464480874,
+      "grad_norm": 1.6945140343571883,
+      "learning_rate": 0.00019852112447025242,
+      "loss": 16.8919,
+      "step": 1427
+    },
+    {
+      "epoch": 0.24970491803278688,
+      "grad_norm": 1.9870623696895802,
+      "learning_rate": 0.0001985178875835296,
+      "loss": 16.3145,
+      "step": 1428
+    },
+    {
+      "epoch": 0.24987978142076503,
+      "grad_norm": 1.5542700617275476,
+      "learning_rate": 0.00019851464718475916,
+      "loss": 16.3599,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2500546448087432,
+      "grad_norm": 1.6628416247163955,
+      "learning_rate": 0.00019851140327405666,
+      "loss": 16.2299,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2502295081967213,
+      "grad_norm": 2.0719824128305393,
+      "learning_rate": 0.00019850815585153775,
+      "loss": 16.8486,
+      "step": 1431
+    },
+    {
+      "epoch": 0.25040437158469947,
+      "grad_norm": 2.506644821583195,
+      "learning_rate": 0.00019850490491731817,
+      "loss": 16.7045,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2505792349726776,
+      "grad_norm": 1.4425105481746046,
+      "learning_rate": 0.0001985016504715138,
+      "loss": 16.5919,
+      "step": 1433
+    },
+    {
+      "epoch": 0.25075409836065576,
+      "grad_norm": 1.8276251115008666,
+      "learning_rate": 0.00019849839251424072,
+      "loss": 16.3187,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2509289617486339,
+      "grad_norm": 1.5930724573478343,
+      "learning_rate": 0.00019849513104561498,
+      "loss": 16.4821,
+      "step": 1435
+    },
+    {
+      "epoch": 0.251103825136612,
+      "grad_norm": 1.695799686826368,
+      "learning_rate": 0.00019849186606575296,
+      "loss": 16.4497,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2512786885245902,
+      "grad_norm": 1.5545671017854008,
+      "learning_rate": 0.00019848859757477094,
+      "loss": 16.4246,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2514535519125683,
+      "grad_norm": 1.476390357461579,
+      "learning_rate": 0.0001984853255727855,
+      "loss": 16.1889,
+      "step": 1438
+    },
+    {
+      "epoch": 0.25162841530054647,
+      "grad_norm": 1.6973138414587152,
+      "learning_rate": 0.0001984820500599133,
+      "loss": 16.2979,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2518032786885246,
+      "grad_norm": 1.4246667080840756,
+      "learning_rate": 0.00019847877103627105,
+      "loss": 16.3344,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2519781420765027,
+      "grad_norm": 1.6290243216633027,
+      "learning_rate": 0.00019847548850197567,
+      "loss": 16.01,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2521530054644809,
+      "grad_norm": 1.7163210656886583,
+      "learning_rate": 0.0001984722024571442,
+      "loss": 16.4122,
+      "step": 1442
+    },
+    {
+      "epoch": 0.252327868852459,
+      "grad_norm": 1.5240135267062265,
+      "learning_rate": 0.00019846891290189374,
+      "loss": 16.5386,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2525027322404372,
+      "grad_norm": 1.5402646385021275,
+      "learning_rate": 0.0001984656198363416,
+      "loss": 16.3906,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2526775956284153,
+      "grad_norm": 1.444155914939519,
+      "learning_rate": 0.00019846232326060513,
+      "loss": 16.1134,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2528524590163934,
+      "grad_norm": 1.8907793530929604,
+      "learning_rate": 0.00019845902317480193,
+      "loss": 16.723,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2530273224043716,
+      "grad_norm": 1.7690175780700113,
+      "learning_rate": 0.00019845571957904955,
+      "loss": 16.5949,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2532021857923497,
+      "grad_norm": 1.5357599936486643,
+      "learning_rate": 0.00019845241247346581,
+      "loss": 16.1803,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2533770491803279,
+      "grad_norm": 1.7480391654017444,
+      "learning_rate": 0.00019844910185816858,
+      "loss": 16.2731,
+      "step": 1449
+    },
+    {
+      "epoch": 0.253551912568306,
+      "grad_norm": 1.4596687756793283,
+      "learning_rate": 0.00019844578773327594,
+      "loss": 16.2703,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2537267759562842,
+      "grad_norm": 2.158334900157998,
+      "learning_rate": 0.00019844247009890596,
+      "loss": 16.3489,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2539016393442623,
+      "grad_norm": 1.9879483052177425,
+      "learning_rate": 0.00019843914895517693,
+      "loss": 16.415,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2540765027322404,
+      "grad_norm": 1.9229632066253965,
+      "learning_rate": 0.00019843582430220725,
+      "loss": 16.4318,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2542513661202186,
+      "grad_norm": 1.6337668362247826,
+      "learning_rate": 0.00019843249614011547,
+      "loss": 16.2282,
+      "step": 1454
+    },
+    {
+      "epoch": 0.2544262295081967,
+      "grad_norm": 1.6725826288551242,
+      "learning_rate": 0.0001984291644690202,
+      "loss": 16.1965,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2546010928961749,
+      "grad_norm": 1.5394197987151612,
+      "learning_rate": 0.0001984258292890402,
+      "loss": 16.2485,
+      "step": 1456
+    },
+    {
+      "epoch": 0.254775956284153,
+      "grad_norm": 1.6980900823486738,
+      "learning_rate": 0.0001984224906002944,
+      "loss": 16.3188,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2549508196721311,
+      "grad_norm": 1.4508717211120286,
+      "learning_rate": 0.00019841914840290178,
+      "loss": 16.2114,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2551256830601093,
+      "grad_norm": 1.6861719325920175,
+      "learning_rate": 0.00019841580269698155,
+      "loss": 16.2039,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2553005464480874,
+      "grad_norm": 1.7624409913449495,
+      "learning_rate": 0.00019841245348265293,
+      "loss": 16.4882,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2554754098360656,
+      "grad_norm": 1.433355676270299,
+      "learning_rate": 0.00019840910076003533,
+      "loss": 15.9816,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2556502732240437,
+      "grad_norm": 1.5975513448624128,
+      "learning_rate": 0.00019840574452924825,
+      "loss": 16.2462,
+      "step": 1462
+    },
+    {
+      "epoch": 0.25582513661202183,
+      "grad_norm": 1.5806259747808018,
+      "learning_rate": 0.00019840238479041134,
+      "loss": 16.3284,
+      "step": 1463
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.5668257756389263,
+      "learning_rate": 0.0001983990215436444,
+      "loss": 16.5311,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2561748633879781,
+      "grad_norm": 1.6609128152253052,
+      "learning_rate": 0.0001983956547890673,
+      "loss": 16.5421,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2563497267759563,
+      "grad_norm": 1.5766949932393004,
+      "learning_rate": 0.00019839228452680008,
+      "loss": 16.6109,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2565245901639344,
+      "grad_norm": 1.4981193396783188,
+      "learning_rate": 0.00019838891075696287,
+      "loss": 15.9645,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2566994535519126,
+      "grad_norm": 1.4954889449224593,
+      "learning_rate": 0.00019838553347967594,
+      "loss": 16.2731,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2568743169398907,
+      "grad_norm": 1.5654988647754264,
+      "learning_rate": 0.00019838215269505973,
+      "loss": 16.6369,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25704918032786883,
+      "grad_norm": 1.500142157811162,
+      "learning_rate": 0.00019837876840323468,
+      "loss": 16.1109,
+      "step": 1470
+    },
+    {
+      "epoch": 0.257224043715847,
+      "grad_norm": 1.6022816100408914,
+      "learning_rate": 0.0001983753806043215,
+      "loss": 16.1576,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2573989071038251,
+      "grad_norm": 1.4693951171167876,
+      "learning_rate": 0.00019837198929844094,
+      "loss": 15.9589,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2575737704918033,
+      "grad_norm": 1.4405114216718897,
+      "learning_rate": 0.0001983685944857139,
+      "loss": 15.9539,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2577486338797814,
+      "grad_norm": 1.4315420947304809,
+      "learning_rate": 0.0001983651961662614,
+      "loss": 16.3088,
+      "step": 1474
+    },
+    {
+      "epoch": 0.25792349726775954,
+      "grad_norm": 1.4489572456634403,
+      "learning_rate": 0.00019836179434020457,
+      "loss": 16.0701,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2580983606557377,
+      "grad_norm": 1.4920250758815634,
+      "learning_rate": 0.0001983583890076647,
+      "loss": 16.1924,
+      "step": 1476
+    },
+    {
+      "epoch": 0.25827322404371583,
+      "grad_norm": 1.503090529486238,
+      "learning_rate": 0.00019835498016876316,
+      "loss": 16.158,
+      "step": 1477
+    },
+    {
+      "epoch": 0.258448087431694,
+      "grad_norm": 1.636179583409921,
+      "learning_rate": 0.00019835156782362155,
+      "loss": 16.0953,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2586229508196721,
+      "grad_norm": 2.1275771248329347,
+      "learning_rate": 0.00019834815197236142,
+      "loss": 16.1118,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2587978142076503,
+      "grad_norm": 1.552827106406696,
+      "learning_rate": 0.00019834473261510458,
+      "loss": 16.1535,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2589726775956284,
+      "grad_norm": 1.593683388930233,
+      "learning_rate": 0.00019834130975197292,
+      "loss": 15.8991,
+      "step": 1481
+    },
+    {
+      "epoch": 0.25914754098360654,
+      "grad_norm": 1.617225189858768,
+      "learning_rate": 0.00019833788338308848,
+      "loss": 16.387,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2593224043715847,
+      "grad_norm": 1.5160002033657218,
+      "learning_rate": 0.00019833445350857338,
+      "loss": 16.2489,
+      "step": 1483
+    },
+    {
+      "epoch": 0.25949726775956283,
+      "grad_norm": 1.553496335850074,
+      "learning_rate": 0.00019833102012854992,
+      "loss": 16.3341,
+      "step": 1484
+    },
+    {
+      "epoch": 0.259672131147541,
+      "grad_norm": 1.56584200232332,
+      "learning_rate": 0.00019832758324314046,
+      "loss": 16.2813,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2598469945355191,
+      "grad_norm": 1.622333196760155,
+      "learning_rate": 0.00019832414285246753,
+      "loss": 16.2909,
+      "step": 1486
+    },
+    {
+      "epoch": 0.26002185792349725,
+      "grad_norm": 1.4970200651436145,
+      "learning_rate": 0.00019832069895665385,
+      "loss": 16.0422,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2601967213114754,
+      "grad_norm": 1.7430893988739387,
+      "learning_rate": 0.0001983172515558221,
+      "loss": 16.3343,
+      "step": 1488
+    },
+    {
+      "epoch": 0.26037158469945354,
+      "grad_norm": 1.690386473950431,
+      "learning_rate": 0.00019831380065009517,
+      "loss": 16.0303,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2605464480874317,
+      "grad_norm": 1.8555907264978893,
+      "learning_rate": 0.00019831034623959615,
+      "loss": 16.2651,
+      "step": 1490
+    },
+    {
+      "epoch": 0.26072131147540983,
+      "grad_norm": 1.6412164862006566,
+      "learning_rate": 0.0001983068883244481,
+      "loss": 16.1136,
+      "step": 1491
+    },
+    {
+      "epoch": 0.26089617486338795,
+      "grad_norm": 1.6652140708171532,
+      "learning_rate": 0.0001983034269047744,
+      "loss": 16.1538,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2610710382513661,
+      "grad_norm": 1.7227770307077954,
+      "learning_rate": 0.00019829996198069837,
+      "loss": 16.1584,
+      "step": 1493
+    },
+    {
+      "epoch": 0.26124590163934425,
+      "grad_norm": 1.6364508392442756,
+      "learning_rate": 0.0001982964935523435,
+      "loss": 16.3478,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2614207650273224,
+      "grad_norm": 1.5797403715849947,
+      "learning_rate": 0.00019829302161983355,
+      "loss": 16.0179,
+      "step": 1495
+    },
+    {
+      "epoch": 0.26159562841530054,
+      "grad_norm": 1.6339532283993918,
+      "learning_rate": 0.00019828954618329218,
+      "loss": 16.2061,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2617704918032787,
+      "grad_norm": 1.6667452015765643,
+      "learning_rate": 0.00019828606724284337,
+      "loss": 16.0299,
+      "step": 1497
+    },
+    {
+      "epoch": 0.26194535519125683,
+      "grad_norm": 1.528012275104366,
+      "learning_rate": 0.00019828258479861105,
+      "loss": 16.0844,
+      "step": 1498
+    },
+    {
+      "epoch": 0.26212021857923495,
+      "grad_norm": 1.618353892039659,
+      "learning_rate": 0.00019827909885071943,
+      "loss": 15.9671,
+      "step": 1499
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 1.6147310204759577,
+      "learning_rate": 0.00019827560939929276,
+      "loss": 16.1892,
+      "step": 1500
+    },
+    {
+      "epoch": 0.26246994535519125,
+      "grad_norm": 1.6099760197934725,
+      "learning_rate": 0.0001982721164444554,
+      "loss": 16.5786,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2626448087431694,
+      "grad_norm": 1.6271125549125505,
+      "learning_rate": 0.000198268619986332,
+      "loss": 16.1088,
+      "step": 1502
+    },
+    {
+      "epoch": 0.26281967213114754,
+      "grad_norm": 1.4981991276062967,
+      "learning_rate": 0.00019826512002504704,
+      "loss": 16.3142,
+      "step": 1503
+    },
+    {
+      "epoch": 0.26299453551912566,
+      "grad_norm": 1.54644494786832,
+      "learning_rate": 0.00019826161656072539,
+      "loss": 16.4257,
+      "step": 1504
+    },
+    {
+      "epoch": 0.26316939890710384,
+      "grad_norm": 1.430872622797648,
+      "learning_rate": 0.0001982581095934919,
+      "loss": 15.9351,
+      "step": 1505
+    },
+    {
+      "epoch": 0.26334426229508195,
+      "grad_norm": 1.4093899214174432,
+      "learning_rate": 0.00019825459912347162,
+      "loss": 15.7533,
+      "step": 1506
+    },
+    {
+      "epoch": 0.26351912568306013,
+      "grad_norm": 1.5787367557846999,
+      "learning_rate": 0.00019825108515078967,
+      "loss": 16.1457,
+      "step": 1507
+    },
+    {
+      "epoch": 0.26369398907103825,
+      "grad_norm": 1.6884589168334578,
+      "learning_rate": 0.00019824756767557135,
+      "loss": 15.8085,
+      "step": 1508
+    },
+    {
+      "epoch": 0.26386885245901637,
+      "grad_norm": 1.694607939399454,
+      "learning_rate": 0.000198244046697942,
+      "loss": 16.2214,
+      "step": 1509
+    },
+    {
+      "epoch": 0.26404371584699454,
+      "grad_norm": 1.4155779794480248,
+      "learning_rate": 0.00019824052221802718,
+      "loss": 16.1356,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26421857923497266,
+      "grad_norm": 1.635799773512617,
+      "learning_rate": 0.0001982369942359525,
+      "loss": 16.0507,
+      "step": 1511
+    },
+    {
+      "epoch": 0.26439344262295084,
+      "grad_norm": 1.4752661839941412,
+      "learning_rate": 0.0001982334627518438,
+      "loss": 16.118,
+      "step": 1512
+    },
+    {
+      "epoch": 0.26456830601092896,
+      "grad_norm": 1.4760023682002674,
+      "learning_rate": 0.0001982299277658269,
+      "loss": 16.2431,
+      "step": 1513
+    },
+    {
+      "epoch": 0.26474316939890713,
+      "grad_norm": 1.5388926779551777,
+      "learning_rate": 0.00019822638927802789,
+      "loss": 15.5859,
+      "step": 1514
+    },
+    {
+      "epoch": 0.26491803278688525,
+      "grad_norm": 1.4619207858711964,
+      "learning_rate": 0.0001982228472885728,
+      "loss": 16.2883,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26509289617486337,
+      "grad_norm": 1.596728649810811,
+      "learning_rate": 0.000198219301797588,
+      "loss": 16.0188,
+      "step": 1516
+    },
+    {
+      "epoch": 0.26526775956284154,
+      "grad_norm": 1.4362777210110558,
+      "learning_rate": 0.00019821575280519986,
+      "loss": 16.0934,
+      "step": 1517
+    },
+    {
+      "epoch": 0.26544262295081966,
+      "grad_norm": 1.7543127207714446,
+      "learning_rate": 0.00019821220031153485,
+      "loss": 16.1101,
+      "step": 1518
+    },
+    {
+      "epoch": 0.26561748633879784,
+      "grad_norm": 1.6601255375601607,
+      "learning_rate": 0.00019820864431671967,
+      "loss": 15.7749,
+      "step": 1519
+    },
+    {
+      "epoch": 0.26579234972677596,
+      "grad_norm": 1.6631713351789204,
+      "learning_rate": 0.00019820508482088105,
+      "loss": 15.9447,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2659672131147541,
+      "grad_norm": 1.729228281432634,
+      "learning_rate": 0.00019820152182414594,
+      "loss": 15.9674,
+      "step": 1521
+    },
+    {
+      "epoch": 0.26614207650273225,
+      "grad_norm": 1.5685616163098326,
+      "learning_rate": 0.00019819795532664128,
+      "loss": 15.8254,
+      "step": 1522
+    },
+    {
+      "epoch": 0.26631693989071037,
+      "grad_norm": 1.9414023486477148,
+      "learning_rate": 0.00019819438532849422,
+      "loss": 15.8548,
+      "step": 1523
+    },
+    {
+      "epoch": 0.26649180327868854,
+      "grad_norm": 1.5082510715832724,
+      "learning_rate": 0.0001981908118298321,
+      "loss": 16.2434,
+      "step": 1524
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.9349279196339007,
+      "learning_rate": 0.00019818723483078224,
+      "loss": 15.9414,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2668415300546448,
+      "grad_norm": 1.4649938017192718,
+      "learning_rate": 0.00019818365433147217,
+      "loss": 15.9142,
+      "step": 1526
+    },
+    {
+      "epoch": 0.26701639344262296,
+      "grad_norm": 1.6981791003561213,
+      "learning_rate": 0.0001981800703320295,
+      "loss": 15.9982,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2671912568306011,
+      "grad_norm": 1.552002458987299,
+      "learning_rate": 0.0001981764828325821,
+      "loss": 16.0986,
+      "step": 1528
+    },
+    {
+      "epoch": 0.26736612021857925,
+      "grad_norm": 1.6628272530602144,
+      "learning_rate": 0.00019817289183325778,
+      "loss": 15.523,
+      "step": 1529
+    },
+    {
+      "epoch": 0.26754098360655737,
+      "grad_norm": 1.6785689760100526,
+      "learning_rate": 0.00019816929733418456,
+      "loss": 15.8643,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26771584699453554,
+      "grad_norm": 2.0018860811347707,
+      "learning_rate": 0.0001981656993354906,
+      "loss": 15.9298,
+      "step": 1531
+    },
+    {
+      "epoch": 0.26789071038251366,
+      "grad_norm": 1.512878495307547,
+      "learning_rate": 0.00019816209783730414,
+      "loss": 15.5975,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2680655737704918,
+      "grad_norm": 1.7090004580898435,
+      "learning_rate": 0.00019815849283975355,
+      "loss": 15.9445,
+      "step": 1533
+    },
+    {
+      "epoch": 0.26824043715846996,
+      "grad_norm": 1.775936148890099,
+      "learning_rate": 0.00019815488434296743,
+      "loss": 16.1576,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2684153005464481,
+      "grad_norm": 1.4895620098749631,
+      "learning_rate": 0.00019815127234707434,
+      "loss": 15.8862,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26859016393442625,
+      "grad_norm": 1.4622439493010273,
+      "learning_rate": 0.00019814765685220308,
+      "loss": 15.9433,
+      "step": 1536
+    },
+    {
+      "epoch": 0.26876502732240437,
+      "grad_norm": 1.5243117122101288,
+      "learning_rate": 0.00019814403785848252,
+      "loss": 15.7747,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2689398907103825,
+      "grad_norm": 1.6756100164272107,
+      "learning_rate": 0.0001981404153660417,
+      "loss": 16.2599,
+      "step": 1538
+    },
+    {
+      "epoch": 0.26911475409836066,
+      "grad_norm": 1.4769064569706876,
+      "learning_rate": 0.0001981367893750097,
+      "loss": 15.9377,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2692896174863388,
+      "grad_norm": 1.5220907670674488,
+      "learning_rate": 0.00019813315988551585,
+      "loss": 15.7406,
+      "step": 1540
+    },
+    {
+      "epoch": 0.26946448087431696,
+      "grad_norm": 1.5555931346572007,
+      "learning_rate": 0.0001981295268976895,
+      "loss": 15.944,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2696393442622951,
+      "grad_norm": 1.467685559458748,
+      "learning_rate": 0.00019812589041166013,
+      "loss": 15.5414,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2698142076502732,
+      "grad_norm": 1.422380538144313,
+      "learning_rate": 0.00019812225042755745,
+      "loss": 15.87,
+      "step": 1543
+    },
+    {
+      "epoch": 0.26998907103825137,
+      "grad_norm": 1.38842405248452,
+      "learning_rate": 0.00019811860694551116,
+      "loss": 15.6931,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2701639344262295,
+      "grad_norm": 1.50696833009234,
+      "learning_rate": 0.0001981149599656512,
+      "loss": 15.6626,
+      "step": 1545
+    },
+    {
+      "epoch": 0.27033879781420767,
+      "grad_norm": 1.5434433743277605,
+      "learning_rate": 0.00019811130948810756,
+      "loss": 15.8504,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2705136612021858,
+      "grad_norm": 1.8100248840585416,
+      "learning_rate": 0.00019810765551301035,
+      "loss": 15.8706,
+      "step": 1547
+    },
+    {
+      "epoch": 0.27068852459016396,
+      "grad_norm": 1.4700888221166897,
+      "learning_rate": 0.00019810399804048984,
+      "loss": 15.7311,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2708633879781421,
+      "grad_norm": 1.51018933078465,
+      "learning_rate": 0.00019810033707067643,
+      "loss": 15.9401,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2710382513661202,
+      "grad_norm": 1.4911915504430941,
+      "learning_rate": 0.0001980966726037006,
+      "loss": 15.8297,
+      "step": 1550
+    },
+    {
+      "epoch": 0.27121311475409837,
+      "grad_norm": 1.857880818800563,
+      "learning_rate": 0.00019809300463969308,
+      "loss": 15.9538,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2713879781420765,
+      "grad_norm": 1.89554523650845,
+      "learning_rate": 0.00019808933317878448,
+      "loss": 15.9796,
+      "step": 1552
+    },
+    {
+      "epoch": 0.27156284153005467,
+      "grad_norm": 1.680191776485675,
+      "learning_rate": 0.0001980856582211058,
+      "loss": 15.7169,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2717377049180328,
+      "grad_norm": 1.4953845981361804,
+      "learning_rate": 0.00019808197976678798,
+      "loss": 15.7143,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2719125683060109,
+      "grad_norm": 2.10215084720139,
+      "learning_rate": 0.00019807829781596215,
+      "loss": 15.919,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2720874316939891,
+      "grad_norm": 1.9635354521646873,
+      "learning_rate": 0.00019807461236875966,
+      "loss": 16.1166,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2722622950819672,
+      "grad_norm": 1.5451079037298268,
+      "learning_rate": 0.00019807092342531178,
+      "loss": 16.0751,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2724371584699454,
+      "grad_norm": 1.9752348142320781,
+      "learning_rate": 0.0001980672309857501,
+      "loss": 15.7877,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2726120218579235,
+      "grad_norm": 1.7062070119915576,
+      "learning_rate": 0.00019806353505020616,
+      "loss": 15.6193,
+      "step": 1559
+    },
+    {
+      "epoch": 0.27278688524590167,
+      "grad_norm": 1.5583820657866814,
+      "learning_rate": 0.00019805983561881183,
+      "loss": 16.2562,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2729617486338798,
+      "grad_norm": 1.5502792909294119,
+      "learning_rate": 0.00019805613269169889,
+      "loss": 15.7356,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2731366120218579,
+      "grad_norm": 1.441446583910463,
+      "learning_rate": 0.00019805242626899937,
+      "loss": 15.9144,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2733114754098361,
+      "grad_norm": 1.5304499789086337,
+      "learning_rate": 0.0001980487163508455,
+      "loss": 15.5667,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2734863387978142,
+      "grad_norm": 1.72946822071907,
+      "learning_rate": 0.00019804500293736936,
+      "loss": 15.9971,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2736612021857924,
+      "grad_norm": 1.5308656852630482,
+      "learning_rate": 0.00019804128602870345,
+      "loss": 15.7705,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2738360655737705,
+      "grad_norm": 2.3286032115126885,
+      "learning_rate": 0.00019803756562498026,
+      "loss": 15.8736,
+      "step": 1566
+    },
+    {
+      "epoch": 0.2740109289617486,
+      "grad_norm": 1.5854939023687284,
+      "learning_rate": 0.00019803384172633237,
+      "loss": 15.9019,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2741857923497268,
+      "grad_norm": 1.6602458083608502,
+      "learning_rate": 0.0001980301143328926,
+      "loss": 15.7014,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2743606557377049,
+      "grad_norm": 1.6116426813071882,
+      "learning_rate": 0.00019802638344479373,
+      "loss": 15.5775,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2745355191256831,
+      "grad_norm": 1.591346810352641,
+      "learning_rate": 0.00019802264906216888,
+      "loss": 15.5568,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2747103825136612,
+      "grad_norm": 1.76101511566957,
+      "learning_rate": 0.0001980189111851511,
+      "loss": 15.9693,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2748852459016393,
+      "grad_norm": 1.5079547916258242,
+      "learning_rate": 0.00019801516981387364,
+      "loss": 15.9707,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2750601092896175,
+      "grad_norm": 1.3949501073416468,
+      "learning_rate": 0.0001980114249484699,
+      "loss": 15.6708,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2752349726775956,
+      "grad_norm": 1.591313256778152,
+      "learning_rate": 0.00019800767658907339,
+      "loss": 15.6966,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2754098360655738,
+      "grad_norm": 1.5218622525760999,
+      "learning_rate": 0.0001980039247358177,
+      "loss": 15.8968,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2755846994535519,
+      "grad_norm": 1.8346777945258883,
+      "learning_rate": 0.00019800016938883666,
+      "loss": 15.816,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2757595628415301,
+      "grad_norm": 1.4953196535885003,
+      "learning_rate": 0.00019799641054826404,
+      "loss": 15.5237,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2759344262295082,
+      "grad_norm": 1.6803995127474578,
+      "learning_rate": 0.00019799264821423386,
+      "loss": 15.52,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2761092896174863,
+      "grad_norm": 1.56614047026686,
+      "learning_rate": 0.0001979888823868803,
+      "loss": 15.8857,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2762841530054645,
+      "grad_norm": 1.696573126126465,
+      "learning_rate": 0.00019798511306633754,
+      "loss": 15.7216,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2764590163934426,
+      "grad_norm": 1.3928421004663223,
+      "learning_rate": 0.00019798134025274,
+      "loss": 15.4713,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2766338797814208,
+      "grad_norm": 1.7252025356180993,
+      "learning_rate": 0.00019797756394622216,
+      "loss": 15.6598,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2768087431693989,
+      "grad_norm": 1.3225873752971458,
+      "learning_rate": 0.00019797378414691862,
+      "loss": 15.6446,
+      "step": 1583
+    },
+    {
+      "epoch": 0.276983606557377,
+      "grad_norm": 1.6514846118216737,
+      "learning_rate": 0.00019797000085496413,
+      "loss": 15.8446,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2771584699453552,
+      "grad_norm": 1.4126951611544007,
+      "learning_rate": 0.0001979662140704936,
+      "loss": 15.5701,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2773333333333333,
+      "grad_norm": 1.612912981376172,
+      "learning_rate": 0.00019796242379364204,
+      "loss": 15.4731,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2775081967213115,
+      "grad_norm": 1.4389017472659529,
+      "learning_rate": 0.0001979586300245445,
+      "loss": 15.7569,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2776830601092896,
+      "grad_norm": 1.6861931199531617,
+      "learning_rate": 0.00019795483276333623,
+      "loss": 15.6057,
+      "step": 1588
+    },
+    {
+      "epoch": 0.27785792349726773,
+      "grad_norm": 1.439086401043075,
+      "learning_rate": 0.00019795103201015261,
+      "loss": 15.7015,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2780327868852459,
+      "grad_norm": 1.5981091875161397,
+      "learning_rate": 0.00019794722776512918,
+      "loss": 15.9013,
+      "step": 1590
+    },
+    {
+      "epoch": 0.278207650273224,
+      "grad_norm": 1.5955863601146387,
+      "learning_rate": 0.0001979434200284015,
+      "loss": 15.8063,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2783825136612022,
+      "grad_norm": 1.4712216708412615,
+      "learning_rate": 0.0001979396088001053,
+      "loss": 15.786,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2785573770491803,
+      "grad_norm": 1.547913954072815,
+      "learning_rate": 0.00019793579408037654,
+      "loss": 15.6259,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2787322404371585,
+      "grad_norm": 1.9117129214790918,
+      "learning_rate": 0.0001979319758693511,
+      "loss": 15.7584,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2789071038251366,
+      "grad_norm": 1.4160752641481151,
+      "learning_rate": 0.00019792815416716516,
+      "loss": 15.5637,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27908196721311473,
+      "grad_norm": 1.7808050288523605,
+      "learning_rate": 0.00019792432897395497,
+      "loss": 15.4255,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2792568306010929,
+      "grad_norm": 1.4419074568516335,
+      "learning_rate": 0.0001979205002898568,
+      "loss": 15.9277,
+      "step": 1597
+    },
+    {
+      "epoch": 0.27943169398907103,
+      "grad_norm": 1.8210047851144426,
+      "learning_rate": 0.00019791666811500727,
+      "loss": 15.9129,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2796065573770492,
+      "grad_norm": 1.4584276019049585,
+      "learning_rate": 0.00019791283244954285,
+      "loss": 15.6536,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2797814207650273,
+      "grad_norm": 1.6461089201084023,
+      "learning_rate": 0.00019790899329360042,
+      "loss": 15.8276,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27995628415300544,
+      "grad_norm": 1.4481429705704494,
+      "learning_rate": 0.00019790515064731676,
+      "loss": 15.6019,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2801311475409836,
+      "grad_norm": 1.8650435354927006,
+      "learning_rate": 0.00019790130451082885,
+      "loss": 15.5417,
+      "step": 1602
+    },
+    {
+      "epoch": 0.28030601092896174,
+      "grad_norm": 1.4539368343652062,
+      "learning_rate": 0.0001978974548842738,
+      "loss": 15.686,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2804808743169399,
+      "grad_norm": 1.7646246059932267,
+      "learning_rate": 0.0001978936017677889,
+      "loss": 15.707,
+      "step": 1604
+    },
+    {
+      "epoch": 0.28065573770491803,
+      "grad_norm": 1.5189187876821144,
+      "learning_rate": 0.00019788974516151146,
+      "loss": 15.8178,
+      "step": 1605
+    },
+    {
+      "epoch": 0.28083060109289615,
+      "grad_norm": 1.4943813732481501,
+      "learning_rate": 0.00019788588506557898,
+      "loss": 15.715,
+      "step": 1606
+    },
+    {
+      "epoch": 0.2810054644808743,
+      "grad_norm": 1.4411810039040656,
+      "learning_rate": 0.0001978820214801291,
+      "loss": 15.5296,
+      "step": 1607
+    },
+    {
+      "epoch": 0.28118032786885244,
+      "grad_norm": 1.863976306436966,
+      "learning_rate": 0.00019787815440529944,
+      "loss": 15.7753,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2813551912568306,
+      "grad_norm": 1.4638617947796417,
+      "learning_rate": 0.00019787428384122798,
+      "loss": 15.6945,
+      "step": 1609
+    },
+    {
+      "epoch": 0.28153005464480874,
+      "grad_norm": 1.4484816798824487,
+      "learning_rate": 0.00019787040978805267,
+      "loss": 15.3665,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2817049180327869,
+      "grad_norm": 1.5192838491598326,
+      "learning_rate": 0.0001978665322459116,
+      "loss": 15.8621,
+      "step": 1611
+    },
+    {
+      "epoch": 0.28187978142076503,
+      "grad_norm": 1.5724996407266838,
+      "learning_rate": 0.00019786265121494296,
+      "loss": 15.4976,
+      "step": 1612
+    },
+    {
+      "epoch": 0.28205464480874315,
+      "grad_norm": 1.6785420914314688,
+      "learning_rate": 0.00019785876669528517,
+      "loss": 15.3978,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2822295081967213,
+      "grad_norm": 1.4406956323840943,
+      "learning_rate": 0.00019785487868707668,
+      "loss": 15.6437,
+      "step": 1614
+    },
+    {
+      "epoch": 0.28240437158469944,
+      "grad_norm": 1.4702328586093394,
+      "learning_rate": 0.0001978509871904561,
+      "loss": 15.5968,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2825792349726776,
+      "grad_norm": 1.530285566171929,
+      "learning_rate": 0.00019784709220556213,
+      "loss": 15.6954,
+      "step": 1616
+    },
+    {
+      "epoch": 0.28275409836065574,
+      "grad_norm": 1.486724401595466,
+      "learning_rate": 0.00019784319373253366,
+      "loss": 15.3834,
+      "step": 1617
+    },
+    {
+      "epoch": 0.28292896174863386,
+      "grad_norm": 1.4516986681061115,
+      "learning_rate": 0.00019783929177150965,
+      "loss": 15.7076,
+      "step": 1618
+    },
+    {
+      "epoch": 0.28310382513661203,
+      "grad_norm": 1.4481663352660354,
+      "learning_rate": 0.0001978353863226292,
+      "loss": 15.5101,
+      "step": 1619
+    },
+    {
+      "epoch": 0.28327868852459015,
+      "grad_norm": 1.6146097705370912,
+      "learning_rate": 0.00019783147738603157,
+      "loss": 15.6021,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2834535519125683,
+      "grad_norm": 1.430332762058535,
+      "learning_rate": 0.00019782756496185603,
+      "loss": 15.7268,
+      "step": 1621
+    },
+    {
+      "epoch": 0.28362841530054644,
+      "grad_norm": 1.4089346086275016,
+      "learning_rate": 0.00019782364905024212,
+      "loss": 15.5139,
+      "step": 1622
+    },
+    {
+      "epoch": 0.28380327868852456,
+      "grad_norm": 1.4699071515057074,
+      "learning_rate": 0.0001978197296513294,
+      "loss": 15.6454,
+      "step": 1623
+    },
+    {
+      "epoch": 0.28397814207650274,
+      "grad_norm": 1.5620848172904194,
+      "learning_rate": 0.00019781580676525763,
+      "loss": 15.4689,
+      "step": 1624
+    },
+    {
+      "epoch": 0.28415300546448086,
+      "grad_norm": 1.3650284162211386,
+      "learning_rate": 0.0001978118803921666,
+      "loss": 15.8486,
+      "step": 1625
+    },
+    {
+      "epoch": 0.28432786885245903,
+      "grad_norm": 1.5198560666967997,
+      "learning_rate": 0.00019780795053219637,
+      "loss": 15.8791,
+      "step": 1626
+    },
+    {
+      "epoch": 0.28450273224043715,
+      "grad_norm": 1.4184924949413895,
+      "learning_rate": 0.00019780401718548694,
+      "loss": 15.7928,
+      "step": 1627
+    },
+    {
+      "epoch": 0.2846775956284153,
+      "grad_norm": 1.349463309728299,
+      "learning_rate": 0.0001978000803521786,
+      "loss": 15.6609,
+      "step": 1628
+    },
+    {
+      "epoch": 0.28485245901639344,
+      "grad_norm": 1.3756351066839754,
+      "learning_rate": 0.0001977961400324116,
+      "loss": 15.6725,
+      "step": 1629
+    },
+    {
+      "epoch": 0.28502732240437156,
+      "grad_norm": 1.376108600722472,
+      "learning_rate": 0.00019779219622632652,
+      "loss": 15.4726,
+      "step": 1630
+    },
+    {
+      "epoch": 0.28520218579234974,
+      "grad_norm": 1.4433070706510114,
+      "learning_rate": 0.00019778824893406388,
+      "loss": 15.1858,
+      "step": 1631
+    },
+    {
+      "epoch": 0.28537704918032786,
+      "grad_norm": 1.4844055428620277,
+      "learning_rate": 0.00019778429815576446,
+      "loss": 15.6629,
+      "step": 1632
+    },
+    {
+      "epoch": 0.28555191256830603,
+      "grad_norm": 1.3565161470728628,
+      "learning_rate": 0.00019778034389156902,
+      "loss": 15.5282,
+      "step": 1633
+    },
+    {
+      "epoch": 0.28572677595628415,
+      "grad_norm": 1.4673674971846031,
+      "learning_rate": 0.00019777638614161856,
+      "loss": 15.2829,
+      "step": 1634
+    },
+    {
+      "epoch": 0.28590163934426227,
+      "grad_norm": 1.4210440043595802,
+      "learning_rate": 0.0001977724249060542,
+      "loss": 15.8685,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28607650273224045,
+      "grad_norm": 1.3189872719213316,
+      "learning_rate": 0.00019776846018501712,
+      "loss": 15.4722,
+      "step": 1636
+    },
+    {
+      "epoch": 0.28625136612021856,
+      "grad_norm": 1.2825278764440793,
+      "learning_rate": 0.00019776449197864864,
+      "loss": 15.5114,
+      "step": 1637
+    },
+    {
+      "epoch": 0.28642622950819674,
+      "grad_norm": 1.468160236572325,
+      "learning_rate": 0.0001977605202870903,
+      "loss": 15.4728,
+      "step": 1638
+    },
+    {
+      "epoch": 0.28660109289617486,
+      "grad_norm": 1.3554600086758413,
+      "learning_rate": 0.00019775654511048355,
+      "loss": 15.4228,
+      "step": 1639
+    },
+    {
+      "epoch": 0.28677595628415303,
+      "grad_norm": 1.5497776220880999,
+      "learning_rate": 0.00019775256644897024,
+      "loss": 15.2629,
+      "step": 1640
+    },
+    {
+      "epoch": 0.28695081967213115,
+      "grad_norm": 1.4633177581591899,
+      "learning_rate": 0.00019774858430269217,
+      "loss": 15.2389,
+      "step": 1641
+    },
+    {
+      "epoch": 0.28712568306010927,
+      "grad_norm": 1.4698400678995456,
+      "learning_rate": 0.00019774459867179123,
+      "loss": 15.5013,
+      "step": 1642
+    },
+    {
+      "epoch": 0.28730054644808745,
+      "grad_norm": 1.4100653760262816,
+      "learning_rate": 0.00019774060955640957,
+      "loss": 15.3523,
+      "step": 1643
+    },
+    {
+      "epoch": 0.28747540983606557,
+      "grad_norm": 1.6847024979332206,
+      "learning_rate": 0.00019773661695668936,
+      "loss": 15.5238,
+      "step": 1644
+    },
+    {
+      "epoch": 0.28765027322404374,
+      "grad_norm": 1.3868600075674473,
+      "learning_rate": 0.00019773262087277298,
+      "loss": 15.2852,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28782513661202186,
+      "grad_norm": 1.7429392884261736,
+      "learning_rate": 0.00019772862130480284,
+      "loss": 15.8816,
+      "step": 1646
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.979669065186403,
+      "learning_rate": 0.00019772461825292152,
+      "loss": 15.5795,
+      "step": 1647
+    },
+    {
+      "epoch": 0.28817486338797815,
+      "grad_norm": 1.43365824317307,
+      "learning_rate": 0.00019772061171727175,
+      "loss": 15.1019,
+      "step": 1648
+    },
+    {
+      "epoch": 0.28834972677595627,
+      "grad_norm": 1.4539597631084242,
+      "learning_rate": 0.00019771660169799632,
+      "loss": 15.0157,
+      "step": 1649
+    },
+    {
+      "epoch": 0.28852459016393445,
+      "grad_norm": 1.5300163701945042,
+      "learning_rate": 0.00019771258819523826,
+      "loss": 15.6285,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28869945355191257,
+      "grad_norm": 1.4335188561326184,
+      "learning_rate": 0.00019770857120914056,
+      "loss": 15.6093,
+      "step": 1651
+    },
+    {
+      "epoch": 0.2888743169398907,
+      "grad_norm": 1.478067575391783,
+      "learning_rate": 0.00019770455073984646,
+      "loss": 15.6694,
+      "step": 1652
+    },
+    {
+      "epoch": 0.28904918032786886,
+      "grad_norm": 1.6447358898620401,
+      "learning_rate": 0.00019770052678749932,
+      "loss": 15.6763,
+      "step": 1653
+    },
+    {
+      "epoch": 0.289224043715847,
+      "grad_norm": 1.558562746562047,
+      "learning_rate": 0.0001976964993522425,
+      "loss": 15.321,
+      "step": 1654
+    },
+    {
+      "epoch": 0.28939890710382515,
+      "grad_norm": 1.5735219607240223,
+      "learning_rate": 0.00019769246843421963,
+      "loss": 15.4433,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2895737704918033,
+      "grad_norm": 1.7903655815642994,
+      "learning_rate": 0.00019768843403357442,
+      "loss": 15.4872,
+      "step": 1656
+    },
+    {
+      "epoch": 0.28974863387978145,
+      "grad_norm": 1.350311153490308,
+      "learning_rate": 0.00019768439615045066,
+      "loss": 15.2019,
+      "step": 1657
+    },
+    {
+      "epoch": 0.28992349726775957,
+      "grad_norm": 1.6639609662867119,
+      "learning_rate": 0.0001976803547849923,
+      "loss": 15.5233,
+      "step": 1658
+    },
+    {
+      "epoch": 0.2900983606557377,
+      "grad_norm": 1.7166702226106474,
+      "learning_rate": 0.00019767630993734346,
+      "loss": 15.7641,
+      "step": 1659
+    },
+    {
+      "epoch": 0.29027322404371586,
+      "grad_norm": 1.3810089912520673,
+      "learning_rate": 0.00019767226160764828,
+      "loss": 15.3051,
+      "step": 1660
+    },
+    {
+      "epoch": 0.290448087431694,
+      "grad_norm": 1.6324168950599192,
+      "learning_rate": 0.00019766820979605107,
+      "loss": 15.1183,
+      "step": 1661
+    },
+    {
+      "epoch": 0.29062295081967215,
+      "grad_norm": 1.494673142217879,
+      "learning_rate": 0.00019766415450269632,
+      "loss": 15.4047,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2907978142076503,
+      "grad_norm": 1.6434209657326846,
+      "learning_rate": 0.00019766009572772853,
+      "loss": 15.2376,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2909726775956284,
+      "grad_norm": 1.8139726308593944,
+      "learning_rate": 0.00019765603347129246,
+      "loss": 15.6365,
+      "step": 1664
+    },
+    {
+      "epoch": 0.29114754098360657,
+      "grad_norm": 1.5696810361240336,
+      "learning_rate": 0.0001976519677335329,
+      "loss": 15.7446,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2913224043715847,
+      "grad_norm": 1.5117033596240153,
+      "learning_rate": 0.00019764789851459477,
+      "loss": 15.4937,
+      "step": 1666
+    },
+    {
+      "epoch": 0.29149726775956286,
+      "grad_norm": 1.5912970237568276,
+      "learning_rate": 0.00019764382581462315,
+      "loss": 15.6513,
+      "step": 1667
+    },
+    {
+      "epoch": 0.291672131147541,
+      "grad_norm": 1.6851466878985737,
+      "learning_rate": 0.00019763974963376324,
+      "loss": 15.6385,
+      "step": 1668
+    },
+    {
+      "epoch": 0.2918469945355191,
+      "grad_norm": 1.5965836957314985,
+      "learning_rate": 0.0001976356699721603,
+      "loss": 15.266,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2920218579234973,
+      "grad_norm": 1.4870897575339517,
+      "learning_rate": 0.00019763158682995985,
+      "loss": 15.5346,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2921967213114754,
+      "grad_norm": 1.4911808003102367,
+      "learning_rate": 0.00019762750020730735,
+      "loss": 15.235,
+      "step": 1671
+    },
+    {
+      "epoch": 0.29237158469945357,
+      "grad_norm": 1.552177888741532,
+      "learning_rate": 0.00019762341010434855,
+      "loss": 15.4487,
+      "step": 1672
+    },
+    {
+      "epoch": 0.2925464480874317,
+      "grad_norm": 1.7133799064611066,
+      "learning_rate": 0.00019761931652122926,
+      "loss": 15.3947,
+      "step": 1673
+    },
+    {
+      "epoch": 0.29272131147540986,
+      "grad_norm": 1.4524469103799251,
+      "learning_rate": 0.00019761521945809538,
+      "loss": 15.3702,
+      "step": 1674
+    },
+    {
+      "epoch": 0.292896174863388,
+      "grad_norm": 1.5806254624039533,
+      "learning_rate": 0.00019761111891509296,
+      "loss": 15.3285,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2930710382513661,
+      "grad_norm": 1.413378250453078,
+      "learning_rate": 0.00019760701489236824,
+      "loss": 15.1188,
+      "step": 1676
+    },
+    {
+      "epoch": 0.2932459016393443,
+      "grad_norm": 1.3736824788185449,
+      "learning_rate": 0.00019760290739006744,
+      "loss": 15.5671,
+      "step": 1677
+    },
+    {
+      "epoch": 0.2934207650273224,
+      "grad_norm": 1.3473509630237506,
+      "learning_rate": 0.00019759879640833704,
+      "loss": 15.5058,
+      "step": 1678
+    },
+    {
+      "epoch": 0.29359562841530057,
+      "grad_norm": 1.360833911073522,
+      "learning_rate": 0.0001975946819473236,
+      "loss": 15.2995,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2937704918032787,
+      "grad_norm": 1.4028093322239432,
+      "learning_rate": 0.00019759056400717374,
+      "loss": 15.1375,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2939453551912568,
+      "grad_norm": 1.4011493554064443,
+      "learning_rate": 0.0001975864425880343,
+      "loss": 15.7165,
+      "step": 1681
+    },
+    {
+      "epoch": 0.294120218579235,
+      "grad_norm": 1.449600372166689,
+      "learning_rate": 0.00019758231769005224,
+      "loss": 15.4375,
+      "step": 1682
+    },
+    {
+      "epoch": 0.2942950819672131,
+      "grad_norm": 1.4134038533921116,
+      "learning_rate": 0.00019757818931337453,
+      "loss": 15.5551,
+      "step": 1683
+    },
+    {
+      "epoch": 0.2944699453551913,
+      "grad_norm": 1.6899032108117447,
+      "learning_rate": 0.00019757405745814838,
+      "loss": 15.572,
+      "step": 1684
+    },
+    {
+      "epoch": 0.2946448087431694,
+      "grad_norm": 1.370790009169781,
+      "learning_rate": 0.0001975699221245211,
+      "loss": 15.4158,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2948196721311475,
+      "grad_norm": 1.3598006230312607,
+      "learning_rate": 0.0001975657833126401,
+      "loss": 15.1601,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2949945355191257,
+      "grad_norm": 1.4681215614539929,
+      "learning_rate": 0.0001975616410226529,
+      "loss": 15.1113,
+      "step": 1687
+    },
+    {
+      "epoch": 0.2951693989071038,
+      "grad_norm": 1.5781079916937728,
+      "learning_rate": 0.0001975574952547072,
+      "loss": 15.5315,
+      "step": 1688
+    },
+    {
+      "epoch": 0.295344262295082,
+      "grad_norm": 1.3606908728419302,
+      "learning_rate": 0.00019755334600895074,
+      "loss": 15.3536,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2955191256830601,
+      "grad_norm": 1.4329870909113551,
+      "learning_rate": 0.00019754919328553153,
+      "loss": 15.534,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2956939890710383,
+      "grad_norm": 1.6801876127317592,
+      "learning_rate": 0.0001975450370845975,
+      "loss": 15.5954,
+      "step": 1691
+    },
+    {
+      "epoch": 0.2958688524590164,
+      "grad_norm": 1.8284954202816095,
+      "learning_rate": 0.0001975408774062969,
+      "loss": 15.0757,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2960437158469945,
+      "grad_norm": 1.3781481807685676,
+      "learning_rate": 0.000197536714250778,
+      "loss": 15.3119,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2962185792349727,
+      "grad_norm": 1.4589524759823662,
+      "learning_rate": 0.00019753254761818915,
+      "loss": 15.3493,
+      "step": 1694
+    },
+    {
+      "epoch": 0.2963934426229508,
+      "grad_norm": 1.3595150900967399,
+      "learning_rate": 0.00019752837750867898,
+      "loss": 15.2418,
+      "step": 1695
+    },
+    {
+      "epoch": 0.296568306010929,
+      "grad_norm": 1.4044094935928002,
+      "learning_rate": 0.00019752420392239606,
+      "loss": 15.4083,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2967431693989071,
+      "grad_norm": 1.5226111573339838,
+      "learning_rate": 0.00019752002685948923,
+      "loss": 15.4787,
+      "step": 1697
+    },
+    {
+      "epoch": 0.2969180327868852,
+      "grad_norm": 1.864382360071117,
+      "learning_rate": 0.0001975158463201074,
+      "loss": 15.5128,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2970928961748634,
+      "grad_norm": 1.4962992102579296,
+      "learning_rate": 0.00019751166230439954,
+      "loss": 15.3296,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2972677595628415,
+      "grad_norm": 1.6748351706173334,
+      "learning_rate": 0.00019750747481251489,
+      "loss": 15.322,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2974426229508197,
+      "grad_norm": 1.5106268689277014,
+      "learning_rate": 0.00019750328384460268,
+      "loss": 15.4791,
+      "step": 1701
+    },
+    {
+      "epoch": 0.2976174863387978,
+      "grad_norm": 1.6778775642622952,
+      "learning_rate": 0.00019749908940081232,
+      "loss": 15.7756,
+      "step": 1702
+    },
+    {
+      "epoch": 0.29779234972677593,
+      "grad_norm": 1.4710128638958109,
+      "learning_rate": 0.00019749489148129334,
+      "loss": 15.635,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2979672131147541,
+      "grad_norm": 1.5500851089895231,
+      "learning_rate": 0.0001974906900861954,
+      "loss": 15.3435,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2981420765027322,
+      "grad_norm": 1.7400279575117699,
+      "learning_rate": 0.00019748648521566824,
+      "loss": 15.2659,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2983169398907104,
+      "grad_norm": 1.4274467697898285,
+      "learning_rate": 0.00019748227686986177,
+      "loss": 15.4243,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2984918032786885,
+      "grad_norm": 1.4477478332334315,
+      "learning_rate": 0.00019747806504892605,
+      "loss": 15.3539,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2986666666666667,
+      "grad_norm": 1.6476438370718176,
+      "learning_rate": 0.0001974738497530112,
+      "loss": 15.3207,
+      "step": 1708
+    },
+    {
+      "epoch": 0.2988415300546448,
+      "grad_norm": 1.478018817617984,
+      "learning_rate": 0.0001974696309822675,
+      "loss": 15.325,
+      "step": 1709
+    },
+    {
+      "epoch": 0.29901639344262293,
+      "grad_norm": 1.6815783078081807,
+      "learning_rate": 0.0001974654087368453,
+      "loss": 15.3049,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2991912568306011,
+      "grad_norm": 2.0417652771569847,
+      "learning_rate": 0.00019746118301689522,
+      "loss": 15.371,
+      "step": 1711
+    },
+    {
+      "epoch": 0.2993661202185792,
+      "grad_norm": 1.6925317465130594,
+      "learning_rate": 0.0001974569538225678,
+      "loss": 15.7141,
+      "step": 1712
+    },
+    {
+      "epoch": 0.2995409836065574,
+      "grad_norm": 1.4855249776770254,
+      "learning_rate": 0.00019745272115401384,
+      "loss": 15.2898,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2997158469945355,
+      "grad_norm": 2.0444558779408193,
+      "learning_rate": 0.00019744848501138422,
+      "loss": 15.3885,
+      "step": 1714
+    },
+    {
+      "epoch": 0.29989071038251364,
+      "grad_norm": 1.262162721118693,
+      "learning_rate": 0.00019744424539483,
+      "loss": 15.0144,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3000655737704918,
+      "grad_norm": 1.7581939176186367,
+      "learning_rate": 0.00019744000230450224,
+      "loss": 15.3186,
+      "step": 1716
+    },
+    {
+      "epoch": 0.30024043715846993,
+      "grad_norm": 1.4183961166064225,
+      "learning_rate": 0.00019743575574055226,
+      "loss": 15.3079,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3004153005464481,
+      "grad_norm": 1.6454832926791176,
+      "learning_rate": 0.00019743150570313144,
+      "loss": 15.188,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3005901639344262,
+      "grad_norm": 1.4579751714635618,
+      "learning_rate": 0.00019742725219239128,
+      "loss": 15.2236,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3007650273224044,
+      "grad_norm": 1.699095760765359,
+      "learning_rate": 0.0001974229952084834,
+      "loss": 15.4178,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3009398907103825,
+      "grad_norm": 1.4834088889745052,
+      "learning_rate": 0.00019741873475155958,
+      "loss": 15.3977,
+      "step": 1721
+    },
+    {
+      "epoch": 0.30111475409836064,
+      "grad_norm": 1.5309411434541935,
+      "learning_rate": 0.00019741447082177168,
+      "loss": 15.4729,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3012896174863388,
+      "grad_norm": 1.6992557587126644,
+      "learning_rate": 0.0001974102034192717,
+      "loss": 15.2265,
+      "step": 1723
+    },
+    {
+      "epoch": 0.30146448087431693,
+      "grad_norm": 1.5491543386245066,
+      "learning_rate": 0.0001974059325442118,
+      "loss": 15.4009,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3016393442622951,
+      "grad_norm": 1.67015340516173,
+      "learning_rate": 0.00019740165819674422,
+      "loss": 15.201,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3018142076502732,
+      "grad_norm": 1.3045066144139061,
+      "learning_rate": 0.0001973973803770213,
+      "loss": 15.3174,
+      "step": 1726
+    },
+    {
+      "epoch": 0.30198907103825134,
+      "grad_norm": 1.743282108938772,
+      "learning_rate": 0.0001973930990851956,
+      "loss": 15.3069,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3021639344262295,
+      "grad_norm": 1.4626398740050295,
+      "learning_rate": 0.00019738881432141968,
+      "loss": 15.3348,
+      "step": 1728
+    },
+    {
+      "epoch": 0.30233879781420764,
+      "grad_norm": 1.493227683392788,
+      "learning_rate": 0.0001973845260858463,
+      "loss": 15.4276,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3025136612021858,
+      "grad_norm": 1.554195994958747,
+      "learning_rate": 0.00019738023437862836,
+      "loss": 15.2409,
+      "step": 1730
+    },
+    {
+      "epoch": 0.30268852459016393,
+      "grad_norm": 1.474166905468034,
+      "learning_rate": 0.00019737593919991886,
+      "loss": 15.1395,
+      "step": 1731
+    },
+    {
+      "epoch": 0.30286338797814205,
+      "grad_norm": 1.6888417459169387,
+      "learning_rate": 0.0001973716405498709,
+      "loss": 15.4349,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3030382513661202,
+      "grad_norm": 1.6790190771397218,
+      "learning_rate": 0.00019736733842863768,
+      "loss": 15.3642,
+      "step": 1733
+    },
+    {
+      "epoch": 0.30321311475409835,
+      "grad_norm": 1.294152752370562,
+      "learning_rate": 0.00019736303283637266,
+      "loss": 15.4957,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3033879781420765,
+      "grad_norm": 1.5468799928841532,
+      "learning_rate": 0.00019735872377322925,
+      "loss": 15.1703,
+      "step": 1735
+    },
+    {
+      "epoch": 0.30356284153005464,
+      "grad_norm": 1.4167340375740132,
+      "learning_rate": 0.0001973544112393611,
+      "loss": 15.3945,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3037377049180328,
+      "grad_norm": 1.3396538981515709,
+      "learning_rate": 0.00019735009523492192,
+      "loss": 15.4082,
+      "step": 1737
+    },
+    {
+      "epoch": 0.30391256830601093,
+      "grad_norm": 1.5969159548076637,
+      "learning_rate": 0.0001973457757600656,
+      "loss": 15.2806,
+      "step": 1738
+    },
+    {
+      "epoch": 0.30408743169398905,
+      "grad_norm": 1.519520897205997,
+      "learning_rate": 0.00019734145281494607,
+      "loss": 15.2859,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3042622950819672,
+      "grad_norm": 1.5681093667969128,
+      "learning_rate": 0.00019733712639971753,
+      "loss": 15.0389,
+      "step": 1740
+    },
+    {
+      "epoch": 0.30443715846994535,
+      "grad_norm": 1.5450779081669375,
+      "learning_rate": 0.00019733279651453414,
+      "loss": 15.4737,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3046120218579235,
+      "grad_norm": 1.5029110053305381,
+      "learning_rate": 0.00019732846315955027,
+      "loss": 15.4495,
+      "step": 1742
+    },
+    {
+      "epoch": 0.30478688524590164,
+      "grad_norm": 1.6784252438814309,
+      "learning_rate": 0.0001973241263349204,
+      "loss": 15.3041,
+      "step": 1743
+    },
+    {
+      "epoch": 0.30496174863387976,
+      "grad_norm": 1.3815883653474246,
+      "learning_rate": 0.00019731978604079918,
+      "loss": 15.2797,
+      "step": 1744
+    },
+    {
+      "epoch": 0.30513661202185793,
+      "grad_norm": 1.5476787162365246,
+      "learning_rate": 0.00019731544227734123,
+      "loss": 15.044,
+      "step": 1745
+    },
+    {
+      "epoch": 0.30531147540983605,
+      "grad_norm": 1.6919180750870648,
+      "learning_rate": 0.0001973110950447015,
+      "loss": 15.2557,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3054863387978142,
+      "grad_norm": 1.4136970098211805,
+      "learning_rate": 0.00019730674434303492,
+      "loss": 15.4255,
+      "step": 1747
+    },
+    {
+      "epoch": 0.30566120218579235,
+      "grad_norm": 1.4454056983634909,
+      "learning_rate": 0.0001973023901724966,
+      "loss": 15.0188,
+      "step": 1748
+    },
+    {
+      "epoch": 0.30583606557377047,
+      "grad_norm": 1.4599047302645138,
+      "learning_rate": 0.00019729803253324177,
+      "loss": 15.2744,
+      "step": 1749
+    },
+    {
+      "epoch": 0.30601092896174864,
+      "grad_norm": 1.3963895284539556,
+      "learning_rate": 0.00019729367142542575,
+      "loss": 15.25,
+      "step": 1750
+    },
+    {
+      "epoch": 0.30618579234972676,
+      "grad_norm": 1.4838625200473141,
+      "learning_rate": 0.00019728930684920402,
+      "loss": 15.0876,
+      "step": 1751
+    },
+    {
+      "epoch": 0.30636065573770493,
+      "grad_norm": 1.4332185090270269,
+      "learning_rate": 0.00019728493880473216,
+      "loss": 15.4242,
+      "step": 1752
+    },
+    {
+      "epoch": 0.30653551912568305,
+      "grad_norm": 1.4346230594068328,
+      "learning_rate": 0.00019728056729216592,
+      "loss": 15.0258,
+      "step": 1753
+    },
+    {
+      "epoch": 0.30671038251366123,
+      "grad_norm": 1.7041055687113942,
+      "learning_rate": 0.0001972761923116611,
+      "loss": 15.4431,
+      "step": 1754
+    },
+    {
+      "epoch": 0.30688524590163935,
+      "grad_norm": 1.4231528074091726,
+      "learning_rate": 0.00019727181386337364,
+      "loss": 15.4257,
+      "step": 1755
+    },
+    {
+      "epoch": 0.30706010928961747,
+      "grad_norm": 1.5223524951287408,
+      "learning_rate": 0.0001972674319474597,
+      "loss": 15.3751,
+      "step": 1756
+    },
+    {
+      "epoch": 0.30723497267759564,
+      "grad_norm": 1.4589322388570236,
+      "learning_rate": 0.00019726304656407547,
+      "loss": 15.2148,
+      "step": 1757
+    },
+    {
+      "epoch": 0.30740983606557376,
+      "grad_norm": 1.4308642699091971,
+      "learning_rate": 0.00019725865771337727,
+      "loss": 15.2039,
+      "step": 1758
+    },
+    {
+      "epoch": 0.30758469945355194,
+      "grad_norm": 1.6993846403934507,
+      "learning_rate": 0.00019725426539552154,
+      "loss": 15.0384,
+      "step": 1759
+    },
+    {
+      "epoch": 0.30775956284153005,
+      "grad_norm": 1.282379258095917,
+      "learning_rate": 0.00019724986961066487,
+      "loss": 15.3616,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3079344262295082,
+      "grad_norm": 1.8876981935768224,
+      "learning_rate": 0.000197245470358964,
+      "loss": 15.1834,
+      "step": 1761
+    },
+    {
+      "epoch": 0.30810928961748635,
+      "grad_norm": 1.5346427663385724,
+      "learning_rate": 0.00019724106764057572,
+      "loss": 15.259,
+      "step": 1762
+    },
+    {
+      "epoch": 0.30828415300546447,
+      "grad_norm": 1.416902896518776,
+      "learning_rate": 0.00019723666145565698,
+      "loss": 15.1135,
+      "step": 1763
+    },
+    {
+      "epoch": 0.30845901639344264,
+      "grad_norm": 1.4863991679720134,
+      "learning_rate": 0.0001972322518043649,
+      "loss": 15.3039,
+      "step": 1764
+    },
+    {
+      "epoch": 0.30863387978142076,
+      "grad_norm": 1.3424097535606676,
+      "learning_rate": 0.0001972278386868566,
+      "loss": 15.1451,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3088087431693989,
+      "grad_norm": 1.5728620784670708,
+      "learning_rate": 0.00019722342210328947,
+      "loss": 15.3843,
+      "step": 1766
+    },
+    {
+      "epoch": 0.30898360655737706,
+      "grad_norm": 1.3729876545068607,
+      "learning_rate": 0.00019721900205382095,
+      "loss": 14.8282,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3091584699453552,
+      "grad_norm": 1.481182606773243,
+      "learning_rate": 0.0001972145785386086,
+      "loss": 15.2695,
+      "step": 1768
+    },
+    {
+      "epoch": 0.30933333333333335,
+      "grad_norm": 1.6025820633056858,
+      "learning_rate": 0.00019721015155781008,
+      "loss": 15.1786,
+      "step": 1769
+    },
+    {
+      "epoch": 0.30950819672131147,
+      "grad_norm": 1.4699878624879936,
+      "learning_rate": 0.00019720572111158325,
+      "loss": 15.2567,
+      "step": 1770
+    },
+    {
+      "epoch": 0.30968306010928964,
+      "grad_norm": 1.3513205989470862,
+      "learning_rate": 0.00019720128720008604,
+      "loss": 15.1828,
+      "step": 1771
+    },
+    {
+      "epoch": 0.30985792349726776,
+      "grad_norm": 1.3476038248357014,
+      "learning_rate": 0.0001971968498234765,
+      "loss": 14.9964,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3100327868852459,
+      "grad_norm": 1.4261630649732147,
+      "learning_rate": 0.00019719240898191282,
+      "loss": 15.1503,
+      "step": 1773
+    },
+    {
+      "epoch": 0.31020765027322406,
+      "grad_norm": 1.4658363702025459,
+      "learning_rate": 0.00019718796467555331,
+      "loss": 15.196,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3103825136612022,
+      "grad_norm": 1.5155090941248586,
+      "learning_rate": 0.00019718351690455643,
+      "loss": 15.1993,
+      "step": 1775
+    },
+    {
+      "epoch": 0.31055737704918035,
+      "grad_norm": 1.462686625037041,
+      "learning_rate": 0.0001971790656690807,
+      "loss": 15.0612,
+      "step": 1776
+    },
+    {
+      "epoch": 0.31073224043715847,
+      "grad_norm": 1.3102257252090215,
+      "learning_rate": 0.00019717461096928487,
+      "loss": 15.3235,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3109071038251366,
+      "grad_norm": 1.429673015304483,
+      "learning_rate": 0.00019717015280532765,
+      "loss": 15.1924,
+      "step": 1778
+    },
+    {
+      "epoch": 0.31108196721311476,
+      "grad_norm": 1.438724348883116,
+      "learning_rate": 0.000197165691177368,
+      "loss": 14.9827,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3112568306010929,
+      "grad_norm": 1.3990433196985816,
+      "learning_rate": 0.00019716122608556504,
+      "loss": 15.0849,
+      "step": 1780
+    },
+    {
+      "epoch": 0.31143169398907106,
+      "grad_norm": 1.3431295348429457,
+      "learning_rate": 0.00019715675753007784,
+      "loss": 14.9668,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3116065573770492,
+      "grad_norm": 1.4548243776003993,
+      "learning_rate": 0.00019715228551106577,
+      "loss": 15.2233,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3117814207650273,
+      "grad_norm": 1.4551581521165344,
+      "learning_rate": 0.00019714781002868822,
+      "loss": 14.9507,
+      "step": 1783
+    },
+    {
+      "epoch": 0.31195628415300547,
+      "grad_norm": 1.4013761757223513,
+      "learning_rate": 0.0001971433310831048,
+      "loss": 15.2273,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3121311475409836,
+      "grad_norm": 1.5167380820336946,
+      "learning_rate": 0.00019713884867447506,
+      "loss": 15.2282,
+      "step": 1785
+    },
+    {
+      "epoch": 0.31230601092896176,
+      "grad_norm": 1.3681806522496789,
+      "learning_rate": 0.00019713436280295889,
+      "loss": 15.0537,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3124808743169399,
+      "grad_norm": 1.6278253904507054,
+      "learning_rate": 0.00019712987346871618,
+      "loss": 15.3361,
+      "step": 1787
+    },
+    {
+      "epoch": 0.31265573770491806,
+      "grad_norm": 1.5045813343744885,
+      "learning_rate": 0.00019712538067190693,
+      "loss": 14.8531,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3128306010928962,
+      "grad_norm": 1.426697465295607,
+      "learning_rate": 0.00019712088441269137,
+      "loss": 15.0651,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3130054644808743,
+      "grad_norm": 1.5795087740702447,
+      "learning_rate": 0.00019711638469122975,
+      "loss": 15.2061,
+      "step": 1790
+    },
+    {
+      "epoch": 0.31318032786885247,
+      "grad_norm": 1.4121724950223842,
+      "learning_rate": 0.0001971118815076825,
+      "loss": 15.1125,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3133551912568306,
+      "grad_norm": 1.6766573266288682,
+      "learning_rate": 0.0001971073748622101,
+      "loss": 15.0442,
+      "step": 1792
+    },
+    {
+      "epoch": 0.31353005464480876,
+      "grad_norm": 1.6062716552541574,
+      "learning_rate": 0.00019710286475497325,
+      "loss": 15.0856,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3137049180327869,
+      "grad_norm": 1.3956677956156491,
+      "learning_rate": 0.00019709835118613276,
+      "loss": 14.9683,
+      "step": 1794
+    },
+    {
+      "epoch": 0.313879781420765,
+      "grad_norm": 1.3945617513058932,
+      "learning_rate": 0.00019709383415584947,
+      "loss": 14.9478,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3140546448087432,
+      "grad_norm": 1.421814071132363,
+      "learning_rate": 0.00019708931366428444,
+      "loss": 15.0759,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3142295081967213,
+      "grad_norm": 1.3590502419489354,
+      "learning_rate": 0.00019708478971159883,
+      "loss": 15.2461,
+      "step": 1797
+    },
+    {
+      "epoch": 0.31440437158469947,
+      "grad_norm": 1.4056711686106476,
+      "learning_rate": 0.00019708026229795387,
+      "loss": 15.2499,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3145792349726776,
+      "grad_norm": 1.3947680237736413,
+      "learning_rate": 0.00019707573142351098,
+      "loss": 14.7967,
+      "step": 1799
+    },
+    {
+      "epoch": 0.31475409836065577,
+      "grad_norm": 1.6417952860710414,
+      "learning_rate": 0.00019707119708843173,
+      "loss": 15.0631,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3149289617486339,
+      "grad_norm": 1.3623776183587402,
+      "learning_rate": 0.0001970666592928777,
+      "loss": 15.2634,
+      "step": 1801
+    },
+    {
+      "epoch": 0.315103825136612,
+      "grad_norm": 1.605514095337212,
+      "learning_rate": 0.00019706211803701066,
+      "loss": 15.2723,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3152786885245902,
+      "grad_norm": 1.3898591883802944,
+      "learning_rate": 0.0001970575733209925,
+      "loss": 15.0088,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3154535519125683,
+      "grad_norm": 1.6896796141017978,
+      "learning_rate": 0.00019705302514498528,
+      "loss": 15.1235,
+      "step": 1804
+    },
+    {
+      "epoch": 0.31562841530054647,
+      "grad_norm": 1.3495652929590123,
+      "learning_rate": 0.00019704847350915112,
+      "loss": 15.2517,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3158032786885246,
+      "grad_norm": 1.7138557832713,
+      "learning_rate": 0.00019704391841365225,
+      "loss": 15.0469,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3159781420765027,
+      "grad_norm": 1.3132674690398922,
+      "learning_rate": 0.00019703935985865108,
+      "loss": 15.0156,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3161530054644809,
+      "grad_norm": 1.7095433961240707,
+      "learning_rate": 0.00019703479784431008,
+      "loss": 14.6604,
+      "step": 1808
+    },
+    {
+      "epoch": 0.316327868852459,
+      "grad_norm": 1.760864424495567,
+      "learning_rate": 0.00019703023237079194,
+      "loss": 15.1142,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3165027322404372,
+      "grad_norm": 1.5152710700116483,
+      "learning_rate": 0.00019702566343825936,
+      "loss": 15.0974,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3166775956284153,
+      "grad_norm": 1.7003749659765657,
+      "learning_rate": 0.00019702109104687528,
+      "loss": 14.7246,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3168524590163934,
+      "grad_norm": 2.043200161698302,
+      "learning_rate": 0.00019701651519680264,
+      "loss": 14.8114,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3170273224043716,
+      "grad_norm": 1.9939737837933524,
+      "learning_rate": 0.00019701193588820458,
+      "loss": 15.0315,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3172021857923497,
+      "grad_norm": 1.4076135244239756,
+      "learning_rate": 0.00019700735312124437,
+      "loss": 15.2993,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3173770491803279,
+      "grad_norm": 1.763899335402363,
+      "learning_rate": 0.00019700276689608531,
+      "loss": 15.1265,
+      "step": 1815
+    },
+    {
+      "epoch": 0.317551912568306,
+      "grad_norm": 1.5488991307708508,
+      "learning_rate": 0.000196998177212891,
+      "loss": 15.1636,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3177267759562842,
+      "grad_norm": 1.5093652180632993,
+      "learning_rate": 0.00019699358407182497,
+      "loss": 15.0986,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3179016393442623,
+      "grad_norm": 1.5874617826779704,
+      "learning_rate": 0.00019698898747305102,
+      "loss": 15.063,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3180765027322404,
+      "grad_norm": 1.5518204115242134,
+      "learning_rate": 0.00019698438741673298,
+      "loss": 15.2745,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3182513661202186,
+      "grad_norm": 1.569595388997705,
+      "learning_rate": 0.0001969797839030348,
+      "loss": 15.238,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3184262295081967,
+      "grad_norm": 1.4517044025117007,
+      "learning_rate": 0.0001969751769321207,
+      "loss": 15.3307,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3186010928961749,
+      "grad_norm": 1.5323057434173706,
+      "learning_rate": 0.00019697056650415482,
+      "loss": 15.1135,
+      "step": 1822
+    },
+    {
+      "epoch": 0.318775956284153,
+      "grad_norm": 1.6351512021742434,
+      "learning_rate": 0.00019696595261930154,
+      "loss": 15.0819,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3189508196721311,
+      "grad_norm": 1.4990934185259097,
+      "learning_rate": 0.00019696133527772534,
+      "loss": 15.2444,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3191256830601093,
+      "grad_norm": 1.7602573729924982,
+      "learning_rate": 0.00019695671447959084,
+      "loss": 15.1549,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3193005464480874,
+      "grad_norm": 1.5490574591304658,
+      "learning_rate": 0.0001969520902250627,
+      "loss": 14.9569,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3194754098360656,
+      "grad_norm": 2.088380013762828,
+      "learning_rate": 0.00019694746251430588,
+      "loss": 15.1254,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3196502732240437,
+      "grad_norm": 1.6141787193121468,
+      "learning_rate": 0.00019694283134748526,
+      "loss": 15.2856,
+      "step": 1828
+    },
+    {
+      "epoch": 0.31982513661202183,
+      "grad_norm": 2.0846014479700594,
+      "learning_rate": 0.00019693819672476597,
+      "loss": 14.9664,
+      "step": 1829
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5265083728076192,
+      "learning_rate": 0.00019693355864631323,
+      "loss": 15.1279,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3201748633879781,
+      "grad_norm": 1.860805827257303,
+      "learning_rate": 0.00019692891711229238,
+      "loss": 15.1187,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3203497267759563,
+      "grad_norm": 1.499737208216022,
+      "learning_rate": 0.00019692427212286888,
+      "loss": 14.9492,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3205245901639344,
+      "grad_norm": 1.8628990932322809,
+      "learning_rate": 0.00019691962367820832,
+      "loss": 15.2767,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3206994535519126,
+      "grad_norm": 1.5421033106724444,
+      "learning_rate": 0.00019691497177847645,
+      "loss": 15.0546,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3208743169398907,
+      "grad_norm": 1.3847236344059985,
+      "learning_rate": 0.00019691031642383902,
+      "loss": 15.142,
+      "step": 1835
+    },
+    {
+      "epoch": 0.32104918032786883,
+      "grad_norm": 1.289654153756778,
+      "learning_rate": 0.00019690565761446206,
+      "loss": 14.9178,
+      "step": 1836
+    },
+    {
+      "epoch": 0.321224043715847,
+      "grad_norm": 1.368081062564805,
+      "learning_rate": 0.00019690099535051164,
+      "loss": 14.8546,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3213989071038251,
+      "grad_norm": 1.3851887641544425,
+      "learning_rate": 0.0001968963296321539,
+      "loss": 14.6446,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3215737704918033,
+      "grad_norm": 1.3848725087386333,
+      "learning_rate": 0.0001968916604595553,
+      "loss": 15.1185,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3217486338797814,
+      "grad_norm": 1.4155888870672355,
+      "learning_rate": 0.00019688698783288216,
+      "loss": 14.8656,
+      "step": 1840
+    },
+    {
+      "epoch": 0.32192349726775954,
+      "grad_norm": 1.3873580550414621,
+      "learning_rate": 0.0001968823117523011,
+      "loss": 14.8256,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3220983606557377,
+      "grad_norm": 1.4812031041562292,
+      "learning_rate": 0.00019687763221797885,
+      "loss": 14.858,
+      "step": 1842
+    },
+    {
+      "epoch": 0.32227322404371583,
+      "grad_norm": 1.3470593631491588,
+      "learning_rate": 0.00019687294923008216,
+      "loss": 14.8193,
+      "step": 1843
+    },
+    {
+      "epoch": 0.322448087431694,
+      "grad_norm": 1.384613170243936,
+      "learning_rate": 0.00019686826278877804,
+      "loss": 14.9737,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3226229508196721,
+      "grad_norm": 1.7660412314614702,
+      "learning_rate": 0.00019686357289423355,
+      "loss": 14.8026,
+      "step": 1845
+    },
+    {
+      "epoch": 0.32279781420765025,
+      "grad_norm": 1.5459328152506693,
+      "learning_rate": 0.00019685887954661583,
+      "loss": 15.3989,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3229726775956284,
+      "grad_norm": 1.5653124094884683,
+      "learning_rate": 0.0001968541827460922,
+      "loss": 14.6973,
+      "step": 1847
+    },
+    {
+      "epoch": 0.32314754098360654,
+      "grad_norm": 1.5155494667925862,
+      "learning_rate": 0.00019684948249283014,
+      "loss": 15.106,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3233224043715847,
+      "grad_norm": 1.5134124848085053,
+      "learning_rate": 0.00019684477878699724,
+      "loss": 15.0265,
+      "step": 1849
+    },
+    {
+      "epoch": 0.32349726775956283,
+      "grad_norm": 1.892906379216677,
+      "learning_rate": 0.00019684007162876105,
+      "loss": 14.8982,
+      "step": 1850
+    },
+    {
+      "epoch": 0.323672131147541,
+      "grad_norm": 1.4429499550871525,
+      "learning_rate": 0.0001968353610182895,
+      "loss": 15.1595,
+      "step": 1851
+    },
+    {
+      "epoch": 0.32384699453551913,
+      "grad_norm": 1.765579054852258,
+      "learning_rate": 0.00019683064695575044,
+      "loss": 14.7171,
+      "step": 1852
+    },
+    {
+      "epoch": 0.32402185792349725,
+      "grad_norm": 1.3731953119496088,
+      "learning_rate": 0.00019682592944131195,
+      "loss": 14.9585,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3241967213114754,
+      "grad_norm": 1.7818136616463094,
+      "learning_rate": 0.00019682120847514222,
+      "loss": 15.1517,
+      "step": 1854
+    },
+    {
+      "epoch": 0.32437158469945354,
+      "grad_norm": 1.5692274748690416,
+      "learning_rate": 0.00019681648405740952,
+      "loss": 14.8793,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3245464480874317,
+      "grad_norm": 1.5860814922824387,
+      "learning_rate": 0.0001968117561882823,
+      "loss": 15.1491,
+      "step": 1856
+    },
+    {
+      "epoch": 0.32472131147540984,
+      "grad_norm": 1.5751966437609264,
+      "learning_rate": 0.00019680702486792905,
+      "loss": 15.3467,
+      "step": 1857
+    },
+    {
+      "epoch": 0.32489617486338795,
+      "grad_norm": 1.5965785468629625,
+      "learning_rate": 0.00019680229009651852,
+      "loss": 14.9617,
+      "step": 1858
+    },
+    {
+      "epoch": 0.32507103825136613,
+      "grad_norm": 1.5100051438117381,
+      "learning_rate": 0.0001967975518742194,
+      "loss": 15.0654,
+      "step": 1859
+    },
+    {
+      "epoch": 0.32524590163934425,
+      "grad_norm": 1.395248296032666,
+      "learning_rate": 0.0001967928102012007,
+      "loss": 14.9817,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3254207650273224,
+      "grad_norm": 1.3588929450775902,
+      "learning_rate": 0.00019678806507763137,
+      "loss": 14.925,
+      "step": 1861
+    },
+    {
+      "epoch": 0.32559562841530054,
+      "grad_norm": 1.4451349058421252,
+      "learning_rate": 0.0001967833165036806,
+      "loss": 14.7176,
+      "step": 1862
+    },
+    {
+      "epoch": 0.32577049180327866,
+      "grad_norm": 1.3394100339325643,
+      "learning_rate": 0.0001967785644795177,
+      "loss": 15.0341,
+      "step": 1863
+    },
+    {
+      "epoch": 0.32594535519125684,
+      "grad_norm": 1.4789916831473389,
+      "learning_rate": 0.00019677380900531204,
+      "loss": 15.1636,
+      "step": 1864
+    },
+    {
+      "epoch": 0.32612021857923496,
+      "grad_norm": 1.592787323505593,
+      "learning_rate": 0.00019676905008123315,
+      "loss": 15.2226,
+      "step": 1865
+    },
+    {
+      "epoch": 0.32629508196721313,
+      "grad_norm": 1.4455963834930592,
+      "learning_rate": 0.0001967642877074507,
+      "loss": 15.085,
+      "step": 1866
+    },
+    {
+      "epoch": 0.32646994535519125,
+      "grad_norm": 1.653868527208664,
+      "learning_rate": 0.00019675952188413445,
+      "loss": 14.8762,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3266448087431694,
+      "grad_norm": 1.3781058889685696,
+      "learning_rate": 0.0001967547526114543,
+      "loss": 14.8575,
+      "step": 1868
+    },
+    {
+      "epoch": 0.32681967213114754,
+      "grad_norm": 1.4485618569857965,
+      "learning_rate": 0.00019674997988958023,
+      "loss": 14.9257,
+      "step": 1869
+    },
+    {
+      "epoch": 0.32699453551912566,
+      "grad_norm": 1.4998318479978259,
+      "learning_rate": 0.00019674520371868245,
+      "loss": 15.0242,
+      "step": 1870
+    },
+    {
+      "epoch": 0.32716939890710384,
+      "grad_norm": 1.5537230093662213,
+      "learning_rate": 0.00019674042409893117,
+      "loss": 14.654,
+      "step": 1871
+    },
+    {
+      "epoch": 0.32734426229508196,
+      "grad_norm": 1.3309241267530183,
+      "learning_rate": 0.00019673564103049685,
+      "loss": 14.7704,
+      "step": 1872
+    },
+    {
+      "epoch": 0.32751912568306013,
+      "grad_norm": 1.4221530698343336,
+      "learning_rate": 0.00019673085451354992,
+      "loss": 14.8092,
+      "step": 1873
+    },
+    {
+      "epoch": 0.32769398907103825,
+      "grad_norm": 1.3828525073933555,
+      "learning_rate": 0.000196726064548261,
+      "loss": 15.1883,
+      "step": 1874
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 1.4146438961509784,
+      "learning_rate": 0.00019672127113480093,
+      "loss": 14.5338,
+      "step": 1875
+    },
+    {
+      "epoch": 0.32804371584699454,
+      "grad_norm": 1.3744722952702224,
+      "learning_rate": 0.00019671647427334056,
+      "loss": 14.8745,
+      "step": 1876
+    },
+    {
+      "epoch": 0.32821857923497266,
+      "grad_norm": 1.4608711839348996,
+      "learning_rate": 0.00019671167396405088,
+      "loss": 14.8967,
+      "step": 1877
+    },
+    {
+      "epoch": 0.32839344262295084,
+      "grad_norm": 1.5297469215958726,
+      "learning_rate": 0.00019670687020710301,
+      "loss": 15.0081,
+      "step": 1878
+    },
+    {
+      "epoch": 0.32856830601092896,
+      "grad_norm": 1.4794297152506968,
+      "learning_rate": 0.00019670206300266823,
+      "loss": 14.8551,
+      "step": 1879
+    },
+    {
+      "epoch": 0.32874316939890713,
+      "grad_norm": 1.7343482590836434,
+      "learning_rate": 0.00019669725235091785,
+      "loss": 15.1383,
+      "step": 1880
+    },
+    {
+      "epoch": 0.32891803278688525,
+      "grad_norm": 1.3573104231358488,
+      "learning_rate": 0.00019669243825202342,
+      "loss": 15.0126,
+      "step": 1881
+    },
+    {
+      "epoch": 0.32909289617486337,
+      "grad_norm": 1.885489170179754,
+      "learning_rate": 0.00019668762070615653,
+      "loss": 14.9892,
+      "step": 1882
+    },
+    {
+      "epoch": 0.32926775956284154,
+      "grad_norm": 1.419500068351377,
+      "learning_rate": 0.00019668279971348898,
+      "loss": 14.9124,
+      "step": 1883
+    },
+    {
+      "epoch": 0.32944262295081966,
+      "grad_norm": 1.587590732899521,
+      "learning_rate": 0.00019667797527419251,
+      "loss": 14.8914,
+      "step": 1884
+    },
+    {
+      "epoch": 0.32961748633879784,
+      "grad_norm": 1.517807884535486,
+      "learning_rate": 0.0001966731473884392,
+      "loss": 15.1905,
+      "step": 1885
+    },
+    {
+      "epoch": 0.32979234972677596,
+      "grad_norm": 1.531802994423991,
+      "learning_rate": 0.00019666831605640118,
+      "loss": 14.926,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3299672131147541,
+      "grad_norm": 1.3723547806607257,
+      "learning_rate": 0.0001966634812782506,
+      "loss": 14.9322,
+      "step": 1887
+    },
+    {
+      "epoch": 0.33014207650273225,
+      "grad_norm": 1.5367884031756769,
+      "learning_rate": 0.00019665864305415983,
+      "loss": 14.9903,
+      "step": 1888
+    },
+    {
+      "epoch": 0.33031693989071037,
+      "grad_norm": 1.4759247228997032,
+      "learning_rate": 0.0001966538013843014,
+      "loss": 14.9803,
+      "step": 1889
+    },
+    {
+      "epoch": 0.33049180327868855,
+      "grad_norm": 1.416858974506283,
+      "learning_rate": 0.00019664895626884788,
+      "loss": 14.7585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.33066666666666666,
+      "grad_norm": 1.2852444788862958,
+      "learning_rate": 0.00019664410770797196,
+      "loss": 14.8987,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3308415300546448,
+      "grad_norm": 1.3301098372348825,
+      "learning_rate": 0.00019663925570184655,
+      "loss": 14.9116,
+      "step": 1892
+    },
+    {
+      "epoch": 0.33101639344262296,
+      "grad_norm": 1.3804570541979038,
+      "learning_rate": 0.00019663440025064456,
+      "loss": 14.7682,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3311912568306011,
+      "grad_norm": 1.3310326529901273,
+      "learning_rate": 0.00019662954135453914,
+      "loss": 14.9101,
+      "step": 1894
+    },
+    {
+      "epoch": 0.33136612021857925,
+      "grad_norm": 1.3502838475005303,
+      "learning_rate": 0.00019662467901370345,
+      "loss": 14.8581,
+      "step": 1895
+    },
+    {
+      "epoch": 0.33154098360655737,
+      "grad_norm": 1.356482931872594,
+      "learning_rate": 0.00019661981322831085,
+      "loss": 15.071,
+      "step": 1896
+    },
+    {
+      "epoch": 0.33171584699453555,
+      "grad_norm": 1.300740080508581,
+      "learning_rate": 0.0001966149439985348,
+      "loss": 14.5984,
+      "step": 1897
+    },
+    {
+      "epoch": 0.33189071038251367,
+      "grad_norm": 1.3278083558172844,
+      "learning_rate": 0.00019661007132454886,
+      "loss": 14.89,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3320655737704918,
+      "grad_norm": 1.3175424670408047,
+      "learning_rate": 0.00019660519520652678,
+      "loss": 14.6583,
+      "step": 1899
+    },
+    {
+      "epoch": 0.33224043715846996,
+      "grad_norm": 1.431419538079511,
+      "learning_rate": 0.00019660031564464235,
+      "loss": 14.8059,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3324153005464481,
+      "grad_norm": 1.3970546186149477,
+      "learning_rate": 0.00019659543263906952,
+      "loss": 14.6413,
+      "step": 1901
+    },
+    {
+      "epoch": 0.33259016393442625,
+      "grad_norm": 1.36222799188417,
+      "learning_rate": 0.00019659054618998242,
+      "loss": 14.5629,
+      "step": 1902
+    },
+    {
+      "epoch": 0.33276502732240437,
+      "grad_norm": 1.502422451245484,
+      "learning_rate": 0.00019658565629755514,
+      "loss": 14.8166,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3329398907103825,
+      "grad_norm": 1.5549342318370265,
+      "learning_rate": 0.00019658076296196207,
+      "loss": 14.8976,
+      "step": 1904
+    },
+    {
+      "epoch": 0.33311475409836067,
+      "grad_norm": 1.3804904945873857,
+      "learning_rate": 0.00019657586618337768,
+      "loss": 15.0864,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3332896174863388,
+      "grad_norm": 1.418380261560941,
+      "learning_rate": 0.0001965709659619765,
+      "loss": 14.8646,
+      "step": 1906
+    },
+    {
+      "epoch": 0.33346448087431696,
+      "grad_norm": 1.3707569689959762,
+      "learning_rate": 0.00019656606229793318,
+      "loss": 14.9952,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3336393442622951,
+      "grad_norm": 1.3625963464190811,
+      "learning_rate": 0.0001965611551914226,
+      "loss": 14.6739,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3338142076502732,
+      "grad_norm": 1.4458012272301102,
+      "learning_rate": 0.00019655624464261963,
+      "loss": 14.9077,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3339890710382514,
+      "grad_norm": 1.6510216146854304,
+      "learning_rate": 0.00019655133065169935,
+      "loss": 14.8148,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3341639344262295,
+      "grad_norm": 1.3853433034705833,
+      "learning_rate": 0.00019654641321883697,
+      "loss": 14.8428,
+      "step": 1911
+    },
+    {
+      "epoch": 0.33433879781420767,
+      "grad_norm": 1.391903084547026,
+      "learning_rate": 0.00019654149234420772,
+      "loss": 14.867,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3345136612021858,
+      "grad_norm": 1.3588105255789633,
+      "learning_rate": 0.0001965365680279871,
+      "loss": 14.8987,
+      "step": 1913
+    },
+    {
+      "epoch": 0.33468852459016396,
+      "grad_norm": 1.4254728108990615,
+      "learning_rate": 0.0001965316402703506,
+      "loss": 14.683,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3348633879781421,
+      "grad_norm": 1.4656012278858592,
+      "learning_rate": 0.00019652670907147394,
+      "loss": 14.6061,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3350382513661202,
+      "grad_norm": 1.3015050658424718,
+      "learning_rate": 0.00019652177443153283,
+      "loss": 14.9615,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3352131147540984,
+      "grad_norm": 1.5574561680059813,
+      "learning_rate": 0.00019651683635070325,
+      "loss": 14.9352,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3353879781420765,
+      "grad_norm": 1.3618601163603645,
+      "learning_rate": 0.00019651189482916122,
+      "loss": 14.4762,
+      "step": 1918
+    },
+    {
+      "epoch": 0.33556284153005467,
+      "grad_norm": 1.322272567851972,
+      "learning_rate": 0.00019650694986708293,
+      "loss": 14.6065,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3357377049180328,
+      "grad_norm": 1.3468521726843163,
+      "learning_rate": 0.00019650200146464463,
+      "loss": 14.958,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3359125683060109,
+      "grad_norm": 1.4001251703852877,
+      "learning_rate": 0.00019649704962202273,
+      "loss": 14.7173,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3360874316939891,
+      "grad_norm": 1.3437104958565498,
+      "learning_rate": 0.00019649209433939373,
+      "loss": 14.712,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3362622950819672,
+      "grad_norm": 1.3867263996350425,
+      "learning_rate": 0.0001964871356169343,
+      "loss": 14.7642,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3364371584699454,
+      "grad_norm": 1.2798388783143038,
+      "learning_rate": 0.00019648217345482127,
+      "loss": 14.7523,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3366120218579235,
+      "grad_norm": 1.5855545487859661,
+      "learning_rate": 0.00019647720785323145,
+      "loss": 15.0669,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3367868852459016,
+      "grad_norm": 1.3107904646144906,
+      "learning_rate": 0.0001964722388123419,
+      "loss": 14.8842,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3369617486338798,
+      "grad_norm": 1.2179079470045577,
+      "learning_rate": 0.00019646726633232973,
+      "loss": 14.3816,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3371366120218579,
+      "grad_norm": 1.4030452747658206,
+      "learning_rate": 0.00019646229041337223,
+      "loss": 15.0208,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3373114754098361,
+      "grad_norm": 1.2916728832044642,
+      "learning_rate": 0.0001964573110556468,
+      "loss": 14.3861,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3374863387978142,
+      "grad_norm": 1.4295952031827919,
+      "learning_rate": 0.0001964523282593309,
+      "loss": 14.6576,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3376612021857924,
+      "grad_norm": 1.3160829729596235,
+      "learning_rate": 0.00019644734202460223,
+      "loss": 14.8454,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3378360655737705,
+      "grad_norm": 1.3979753711826433,
+      "learning_rate": 0.00019644235235163847,
+      "loss": 15.0091,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3380109289617486,
+      "grad_norm": 1.3470200802094994,
+      "learning_rate": 0.00019643735924061755,
+      "loss": 14.7379,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3381857923497268,
+      "grad_norm": 1.2939474517320313,
+      "learning_rate": 0.00019643236269171742,
+      "loss": 14.86,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3383606557377049,
+      "grad_norm": 1.2792733973533448,
+      "learning_rate": 0.00019642736270511626,
+      "loss": 14.8291,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3385355191256831,
+      "grad_norm": 1.342798733631339,
+      "learning_rate": 0.00019642235928099227,
+      "loss": 14.6662,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3387103825136612,
+      "grad_norm": 1.401714155488173,
+      "learning_rate": 0.00019641735241952385,
+      "loss": 15.0985,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3388852459016393,
+      "grad_norm": 1.2618751053837365,
+      "learning_rate": 0.00019641234212088943,
+      "loss": 14.9744,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3390601092896175,
+      "grad_norm": 1.3705272819234233,
+      "learning_rate": 0.0001964073283852677,
+      "loss": 14.6427,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3392349726775956,
+      "grad_norm": 1.2089917779180503,
+      "learning_rate": 0.00019640231121283733,
+      "loss": 14.7995,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3394098360655738,
+      "grad_norm": 1.2720595070463983,
+      "learning_rate": 0.0001963972906037772,
+      "loss": 14.5138,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3395846994535519,
+      "grad_norm": 1.3146457978735864,
+      "learning_rate": 0.00019639226655826628,
+      "loss": 15.1176,
+      "step": 1942
+    },
+    {
+      "epoch": 0.33975956284153,
+      "grad_norm": 1.2334495296762833,
+      "learning_rate": 0.0001963872390764837,
+      "loss": 14.9351,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3399344262295082,
+      "grad_norm": 1.5101145451836924,
+      "learning_rate": 0.00019638220815860864,
+      "loss": 14.602,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3401092896174863,
+      "grad_norm": 1.3882548467058133,
+      "learning_rate": 0.0001963771738048205,
+      "loss": 14.6569,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3402841530054645,
+      "grad_norm": 1.5535624506271046,
+      "learning_rate": 0.00019637213601529866,
+      "loss": 14.7715,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3404590163934426,
+      "grad_norm": 2.022705436455542,
+      "learning_rate": 0.00019636709479022284,
+      "loss": 14.717,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3406338797814208,
+      "grad_norm": 1.344748851827121,
+      "learning_rate": 0.00019636205012977262,
+      "loss": 14.6959,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3408087431693989,
+      "grad_norm": 1.802202548573162,
+      "learning_rate": 0.00019635700203412794,
+      "loss": 14.5395,
+      "step": 1949
+    },
+    {
+      "epoch": 0.34098360655737703,
+      "grad_norm": 1.4068265910288218,
+      "learning_rate": 0.00019635195050346873,
+      "loss": 14.6908,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3411584699453552,
+      "grad_norm": 1.8636516004978554,
+      "learning_rate": 0.00019634689553797504,
+      "loss": 14.8128,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3413333333333333,
+      "grad_norm": 1.5179215906072452,
+      "learning_rate": 0.00019634183713782707,
+      "loss": 14.8191,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3415081967213115,
+      "grad_norm": 1.903352529877981,
+      "learning_rate": 0.00019633677530320523,
+      "loss": 14.6031,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3416830601092896,
+      "grad_norm": 1.309344048422258,
+      "learning_rate": 0.00019633171003428986,
+      "loss": 14.7125,
+      "step": 1954
+    },
+    {
+      "epoch": 0.34185792349726773,
+      "grad_norm": 1.611585730667701,
+      "learning_rate": 0.00019632664133126157,
+      "loss": 14.9401,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3420327868852459,
+      "grad_norm": 1.3134934819556503,
+      "learning_rate": 0.0001963215691943011,
+      "loss": 14.6043,
+      "step": 1956
+    },
+    {
+      "epoch": 0.34220765027322403,
+      "grad_norm": 1.3377470759839438,
+      "learning_rate": 0.0001963164936235892,
+      "loss": 14.826,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3423825136612022,
+      "grad_norm": 1.4533691819157817,
+      "learning_rate": 0.00019631141461930684,
+      "loss": 14.8708,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3425573770491803,
+      "grad_norm": 1.4093526662030496,
+      "learning_rate": 0.0001963063321816351,
+      "loss": 14.7369,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3427322404371585,
+      "grad_norm": 1.6131207309762958,
+      "learning_rate": 0.0001963012463107551,
+      "loss": 14.9152,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3429071038251366,
+      "grad_norm": 1.309625593687913,
+      "learning_rate": 0.00019629615700684822,
+      "loss": 14.7792,
+      "step": 1961
+    },
+    {
+      "epoch": 0.34308196721311474,
+      "grad_norm": 1.5367774160268919,
+      "learning_rate": 0.00019629106427009586,
+      "loss": 14.7606,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3432568306010929,
+      "grad_norm": 1.306536374338164,
+      "learning_rate": 0.00019628596810067955,
+      "loss": 14.6118,
+      "step": 1963
+    },
+    {
+      "epoch": 0.34343169398907103,
+      "grad_norm": 1.548178505969289,
+      "learning_rate": 0.00019628086849878096,
+      "loss": 14.7743,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3436065573770492,
+      "grad_norm": 1.4819858117742184,
+      "learning_rate": 0.0001962757654645819,
+      "loss": 14.768,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3437814207650273,
+      "grad_norm": 1.4576113412910434,
+      "learning_rate": 0.00019627065899826429,
+      "loss": 15.0185,
+      "step": 1966
+    },
+    {
+      "epoch": 0.34395628415300544,
+      "grad_norm": 1.5322767855355433,
+      "learning_rate": 0.0001962655491000102,
+      "loss": 14.499,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3441311475409836,
+      "grad_norm": 1.4197471895305993,
+      "learning_rate": 0.0001962604357700017,
+      "loss": 14.8505,
+      "step": 1968
+    },
+    {
+      "epoch": 0.34430601092896174,
+      "grad_norm": 1.3464411871464448,
+      "learning_rate": 0.00019625531900842118,
+      "loss": 14.7205,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3444808743169399,
+      "grad_norm": 1.6159697382419567,
+      "learning_rate": 0.00019625019881545097,
+      "loss": 14.9496,
+      "step": 1970
+    },
+    {
+      "epoch": 0.34465573770491803,
+      "grad_norm": 1.2656752622124936,
+      "learning_rate": 0.00019624507519127367,
+      "loss": 14.5053,
+      "step": 1971
+    },
+    {
+      "epoch": 0.34483060109289615,
+      "grad_norm": 1.6455652541853243,
+      "learning_rate": 0.00019623994813607184,
+      "loss": 15.0763,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3450054644808743,
+      "grad_norm": 1.3989231986854582,
+      "learning_rate": 0.00019623481765002832,
+      "loss": 15.0151,
+      "step": 1973
+    },
+    {
+      "epoch": 0.34518032786885244,
+      "grad_norm": 1.3274224200562186,
+      "learning_rate": 0.000196229683733326,
+      "loss": 14.4122,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3453551912568306,
+      "grad_norm": 1.4994881479052369,
+      "learning_rate": 0.00019622454638614788,
+      "loss": 14.8659,
+      "step": 1975
+    },
+    {
+      "epoch": 0.34553005464480874,
+      "grad_norm": 1.3040997609588154,
+      "learning_rate": 0.0001962194056086771,
+      "loss": 15.0618,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3457049180327869,
+      "grad_norm": 1.5424813592414328,
+      "learning_rate": 0.00019621426140109695,
+      "loss": 14.6781,
+      "step": 1977
+    },
+    {
+      "epoch": 0.34587978142076503,
+      "grad_norm": 1.381660898111536,
+      "learning_rate": 0.00019620911376359078,
+      "loss": 14.7836,
+      "step": 1978
+    },
+    {
+      "epoch": 0.34605464480874315,
+      "grad_norm": 1.4702262909818928,
+      "learning_rate": 0.00019620396269634212,
+      "loss": 14.6823,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3462295081967213,
+      "grad_norm": 1.4340170255825535,
+      "learning_rate": 0.0001961988081995346,
+      "loss": 14.6133,
+      "step": 1980
+    },
+    {
+      "epoch": 0.34640437158469944,
+      "grad_norm": 1.1883181316009734,
+      "learning_rate": 0.00019619365027335195,
+      "loss": 14.365,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3465792349726776,
+      "grad_norm": 1.459528070957838,
+      "learning_rate": 0.00019618848891797804,
+      "loss": 14.6125,
+      "step": 1982
+    },
+    {
+      "epoch": 0.34675409836065574,
+      "grad_norm": 1.3391965241125472,
+      "learning_rate": 0.00019618332413359692,
+      "loss": 14.8242,
+      "step": 1983
+    },
+    {
+      "epoch": 0.34692896174863386,
+      "grad_norm": 1.697635131293985,
+      "learning_rate": 0.00019617815592039265,
+      "loss": 14.8003,
+      "step": 1984
+    },
+    {
+      "epoch": 0.34710382513661203,
+      "grad_norm": 1.4661329997271706,
+      "learning_rate": 0.00019617298427854953,
+      "loss": 14.8689,
+      "step": 1985
+    },
+    {
+      "epoch": 0.34727868852459015,
+      "grad_norm": 1.4002063885928773,
+      "learning_rate": 0.00019616780920825186,
+      "loss": 14.6477,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3474535519125683,
+      "grad_norm": 1.226728358198994,
+      "learning_rate": 0.00019616263070968413,
+      "loss": 14.7463,
+      "step": 1987
+    },
+    {
+      "epoch": 0.34762841530054645,
+      "grad_norm": 1.276561821861322,
+      "learning_rate": 0.000196157448783031,
+      "loss": 14.9414,
+      "step": 1988
+    },
+    {
+      "epoch": 0.34780327868852456,
+      "grad_norm": 1.3378405706714207,
+      "learning_rate": 0.00019615226342847714,
+      "loss": 14.8313,
+      "step": 1989
+    },
+    {
+      "epoch": 0.34797814207650274,
+      "grad_norm": 1.2746345815003266,
+      "learning_rate": 0.0001961470746462075,
+      "loss": 14.8866,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34815300546448086,
+      "grad_norm": 1.3210689040385137,
+      "learning_rate": 0.0001961418824364069,
+      "loss": 14.8579,
+      "step": 1991
+    },
+    {
+      "epoch": 0.34832786885245903,
+      "grad_norm": 1.4271802809102772,
+      "learning_rate": 0.00019613668679926055,
+      "loss": 14.5673,
+      "step": 1992
+    },
+    {
+      "epoch": 0.34850273224043715,
+      "grad_norm": 1.3090076504557777,
+      "learning_rate": 0.00019613148773495366,
+      "loss": 14.3658,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3486775956284153,
+      "grad_norm": 1.4561107568095404,
+      "learning_rate": 0.0001961262852436715,
+      "loss": 14.8367,
+      "step": 1994
+    },
+    {
+      "epoch": 0.34885245901639345,
+      "grad_norm": 1.4643995698859658,
+      "learning_rate": 0.00019612107932559964,
+      "loss": 14.7841,
+      "step": 1995
+    },
+    {
+      "epoch": 0.34902732240437156,
+      "grad_norm": 1.3920133172498834,
+      "learning_rate": 0.00019611586998092358,
+      "loss": 15.003,
+      "step": 1996
+    },
+    {
+      "epoch": 0.34920218579234974,
+      "grad_norm": 1.3833010232690344,
+      "learning_rate": 0.00019611065720982906,
+      "loss": 14.6234,
+      "step": 1997
+    },
+    {
+      "epoch": 0.34937704918032786,
+      "grad_norm": 1.5067461558090645,
+      "learning_rate": 0.00019610544101250193,
+      "loss": 14.8711,
+      "step": 1998
+    },
+    {
+      "epoch": 0.34955191256830603,
+      "grad_norm": 1.4152681389304547,
+      "learning_rate": 0.00019610022138912808,
+      "loss": 14.5929,
+      "step": 1999
+    },
+    {
+      "epoch": 0.34972677595628415,
+      "grad_norm": 1.5951747286620666,
+      "learning_rate": 0.0001960949983398936,
+      "loss": 15.1925,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34990163934426227,
+      "grad_norm": 1.2633861618045776,
+      "learning_rate": 0.00019608977186498473,
+      "loss": 14.8481,
+      "step": 2001
+    },
+    {
+      "epoch": 0.35007650273224045,
+      "grad_norm": 1.6621810586838663,
+      "learning_rate": 0.00019608454196458775,
+      "loss": 14.57,
+      "step": 2002
+    },
+    {
+      "epoch": 0.35025136612021857,
+      "grad_norm": 1.2890156161058215,
+      "learning_rate": 0.00019607930863888915,
+      "loss": 14.7158,
+      "step": 2003
+    },
+    {
+      "epoch": 0.35042622950819674,
+      "grad_norm": 1.4629865668750535,
+      "learning_rate": 0.00019607407188807543,
+      "loss": 14.6202,
+      "step": 2004
+    },
+    {
+      "epoch": 0.35060109289617486,
+      "grad_norm": 1.2705220878069092,
+      "learning_rate": 0.0001960688317123333,
+      "loss": 14.9151,
+      "step": 2005
+    },
+    {
+      "epoch": 0.350775956284153,
+      "grad_norm": 1.613679901999037,
+      "learning_rate": 0.00019606358811184954,
+      "loss": 14.8049,
+      "step": 2006
+    },
+    {
+      "epoch": 0.35095081967213115,
+      "grad_norm": 1.3757431366403075,
+      "learning_rate": 0.00019605834108681112,
+      "loss": 14.4603,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3511256830601093,
+      "grad_norm": 1.2765676757039264,
+      "learning_rate": 0.0001960530906374051,
+      "loss": 14.6091,
+      "step": 2008
+    },
+    {
+      "epoch": 0.35130054644808745,
+      "grad_norm": 1.3159939959896456,
+      "learning_rate": 0.00019604783676381858,
+      "loss": 15.0868,
+      "step": 2009
+    },
+    {
+      "epoch": 0.35147540983606557,
+      "grad_norm": 1.39946403653438,
+      "learning_rate": 0.0001960425794662389,
+      "loss": 14.8133,
+      "step": 2010
+    },
+    {
+      "epoch": 0.35165027322404374,
+      "grad_norm": 1.48352634068383,
+      "learning_rate": 0.00019603731874485352,
+      "loss": 14.9085,
+      "step": 2011
+    },
+    {
+      "epoch": 0.35182513661202186,
+      "grad_norm": 1.3398478088886328,
+      "learning_rate": 0.0001960320545998499,
+      "loss": 14.0306,
+      "step": 2012
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 1.3492777851972912,
+      "learning_rate": 0.00019602678703141573,
+      "loss": 14.6131,
+      "step": 2013
+    },
+    {
+      "epoch": 0.35217486338797815,
+      "grad_norm": 1.3022864620469587,
+      "learning_rate": 0.00019602151603973883,
+      "loss": 14.6517,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3523497267759563,
+      "grad_norm": 1.374619018523308,
+      "learning_rate": 0.00019601624162500706,
+      "loss": 14.946,
+      "step": 2015
+    },
+    {
+      "epoch": 0.35252459016393445,
+      "grad_norm": 1.2927710229044915,
+      "learning_rate": 0.00019601096378740843,
+      "loss": 14.5301,
+      "step": 2016
+    },
+    {
+      "epoch": 0.35269945355191257,
+      "grad_norm": 1.4879540814447172,
+      "learning_rate": 0.00019600568252713113,
+      "loss": 14.5034,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3528743169398907,
+      "grad_norm": 1.2362949863484294,
+      "learning_rate": 0.00019600039784436343,
+      "loss": 14.6791,
+      "step": 2018
+    },
+    {
+      "epoch": 0.35304918032786886,
+      "grad_norm": 1.3256445387036013,
+      "learning_rate": 0.0001959951097392937,
+      "loss": 14.8652,
+      "step": 2019
+    },
+    {
+      "epoch": 0.353224043715847,
+      "grad_norm": 1.4849077229692782,
+      "learning_rate": 0.0001959898182121105,
+      "loss": 14.6566,
+      "step": 2020
+    },
+    {
+      "epoch": 0.35339890710382516,
+      "grad_norm": 1.4788117696294794,
+      "learning_rate": 0.00019598452326300238,
+      "loss": 14.3945,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3535737704918033,
+      "grad_norm": 1.3490630560546217,
+      "learning_rate": 0.0001959792248921582,
+      "loss": 14.7488,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3537486338797814,
+      "grad_norm": 1.553759521721829,
+      "learning_rate": 0.00019597392309976677,
+      "loss": 14.7327,
+      "step": 2023
+    },
+    {
+      "epoch": 0.35392349726775957,
+      "grad_norm": 1.356139246244613,
+      "learning_rate": 0.00019596861788601712,
+      "loss": 14.7949,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3540983606557377,
+      "grad_norm": 1.4514631394586774,
+      "learning_rate": 0.00019596330925109837,
+      "loss": 14.7258,
+      "step": 2025
+    },
+    {
+      "epoch": 0.35427322404371586,
+      "grad_norm": 1.4207541029354396,
+      "learning_rate": 0.00019595799719519973,
+      "loss": 14.6993,
+      "step": 2026
+    },
+    {
+      "epoch": 0.354448087431694,
+      "grad_norm": 1.311209931862104,
+      "learning_rate": 0.00019595268171851066,
+      "loss": 14.558,
+      "step": 2027
+    },
+    {
+      "epoch": 0.35462295081967216,
+      "grad_norm": 1.5262484719690754,
+      "learning_rate": 0.00019594736282122058,
+      "loss": 14.5811,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3547978142076503,
+      "grad_norm": 1.4947262625486137,
+      "learning_rate": 0.0001959420405035191,
+      "loss": 14.6971,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3549726775956284,
+      "grad_norm": 1.3963634116953312,
+      "learning_rate": 0.00019593671476559598,
+      "loss": 14.5597,
+      "step": 2030
+    },
+    {
+      "epoch": 0.35514754098360657,
+      "grad_norm": 1.709802747946802,
+      "learning_rate": 0.00019593138560764108,
+      "loss": 14.8406,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3553224043715847,
+      "grad_norm": 1.4834170771637107,
+      "learning_rate": 0.00019592605302984434,
+      "loss": 14.7537,
+      "step": 2032
+    },
+    {
+      "epoch": 0.35549726775956286,
+      "grad_norm": 1.4739159170007505,
+      "learning_rate": 0.00019592071703239592,
+      "loss": 14.9898,
+      "step": 2033
+    },
+    {
+      "epoch": 0.355672131147541,
+      "grad_norm": 1.8684006168713483,
+      "learning_rate": 0.00019591537761548598,
+      "loss": 14.5713,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3558469945355191,
+      "grad_norm": 1.4596683548879263,
+      "learning_rate": 0.0001959100347793049,
+      "loss": 14.8717,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3560218579234973,
+      "grad_norm": 1.6138082325415888,
+      "learning_rate": 0.00019590468852404315,
+      "loss": 14.5786,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3561967213114754,
+      "grad_norm": 1.7029017083867848,
+      "learning_rate": 0.00019589933884989132,
+      "loss": 14.7299,
+      "step": 2037
+    },
+    {
+      "epoch": 0.35637158469945357,
+      "grad_norm": 1.588145531693942,
+      "learning_rate": 0.0001958939857570401,
+      "loss": 14.659,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3565464480874317,
+      "grad_norm": 1.805834431005455,
+      "learning_rate": 0.0001958886292456803,
+      "loss": 14.8962,
+      "step": 2039
+    },
+    {
+      "epoch": 0.35672131147540986,
+      "grad_norm": 1.5798665409882984,
+      "learning_rate": 0.00019588326931600292,
+      "loss": 14.8215,
+      "step": 2040
+    },
+    {
+      "epoch": 0.356896174863388,
+      "grad_norm": 1.9604919475663556,
+      "learning_rate": 0.00019587790596819904,
+      "loss": 14.7015,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3570710382513661,
+      "grad_norm": 1.470671950593263,
+      "learning_rate": 0.0001958725392024598,
+      "loss": 14.6874,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3572459016393443,
+      "grad_norm": 1.6012223893661266,
+      "learning_rate": 0.00019586716901897654,
+      "loss": 14.3949,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3574207650273224,
+      "grad_norm": 1.419682749044298,
+      "learning_rate": 0.00019586179541794074,
+      "loss": 14.6852,
+      "step": 2044
+    },
+    {
+      "epoch": 0.35759562841530057,
+      "grad_norm": 1.4378505948405664,
+      "learning_rate": 0.00019585641839954394,
+      "loss": 14.6452,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3577704918032787,
+      "grad_norm": 1.3036682230648253,
+      "learning_rate": 0.0001958510379639778,
+      "loss": 14.3856,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3579453551912568,
+      "grad_norm": 1.4412070479132604,
+      "learning_rate": 0.00019584565411143417,
+      "loss": 14.7684,
+      "step": 2047
+    },
+    {
+      "epoch": 0.358120218579235,
+      "grad_norm": 1.2848626526221891,
+      "learning_rate": 0.0001958402668421049,
+      "loss": 14.7422,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3582950819672131,
+      "grad_norm": 1.4906325514444818,
+      "learning_rate": 0.00019583487615618217,
+      "loss": 14.6743,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3584699453551913,
+      "grad_norm": 1.4686187834041537,
+      "learning_rate": 0.00019582948205385802,
+      "loss": 14.8988,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3586448087431694,
+      "grad_norm": 1.2825786976244655,
+      "learning_rate": 0.00019582408453532483,
+      "loss": 14.7727,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3588196721311475,
+      "grad_norm": 1.4016184133444034,
+      "learning_rate": 0.00019581868360077497,
+      "loss": 14.9677,
+      "step": 2052
+    },
+    {
+      "epoch": 0.3589945355191257,
+      "grad_norm": 1.2260793679006576,
+      "learning_rate": 0.000195813279250401,
+      "loss": 14.7384,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3591693989071038,
+      "grad_norm": 1.3486967314067153,
+      "learning_rate": 0.0001958078714843956,
+      "loss": 14.7872,
+      "step": 2054
+    },
+    {
+      "epoch": 0.359344262295082,
+      "grad_norm": 1.3115696205004863,
+      "learning_rate": 0.00019580246030295148,
+      "loss": 14.5705,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3595191256830601,
+      "grad_norm": 1.1971730075762124,
+      "learning_rate": 0.00019579704570626158,
+      "loss": 14.4269,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3596939890710383,
+      "grad_norm": 1.4604921452786508,
+      "learning_rate": 0.00019579162769451896,
+      "loss": 14.8719,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3598688524590164,
+      "grad_norm": 1.269201856527597,
+      "learning_rate": 0.0001957862062679167,
+      "loss": 14.6966,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3600437158469945,
+      "grad_norm": 1.3901170843763553,
+      "learning_rate": 0.00019578078142664813,
+      "loss": 14.4822,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3602185792349727,
+      "grad_norm": 1.5341864412218913,
+      "learning_rate": 0.00019577535317090658,
+      "loss": 14.651,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3603934426229508,
+      "grad_norm": 1.3321618540656468,
+      "learning_rate": 0.0001957699215008856,
+      "loss": 14.8104,
+      "step": 2061
+    },
+    {
+      "epoch": 0.360568306010929,
+      "grad_norm": 1.6923726591369699,
+      "learning_rate": 0.00019576448641677884,
+      "loss": 14.5032,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3607431693989071,
+      "grad_norm": 1.3043954133806661,
+      "learning_rate": 0.00019575904791878,
+      "loss": 14.8998,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3609180327868852,
+      "grad_norm": 1.3777107256018903,
+      "learning_rate": 0.00019575360600708298,
+      "loss": 14.4125,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3610928961748634,
+      "grad_norm": 1.3341895870728815,
+      "learning_rate": 0.0001957481606818818,
+      "loss": 14.5539,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3612677595628415,
+      "grad_norm": 1.432600157922383,
+      "learning_rate": 0.00019574271194337055,
+      "loss": 14.7698,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3614426229508197,
+      "grad_norm": 1.3823743046389887,
+      "learning_rate": 0.00019573725979174348,
+      "loss": 14.5155,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3616174863387978,
+      "grad_norm": 1.2944010050066685,
+      "learning_rate": 0.00019573180422719497,
+      "loss": 14.7532,
+      "step": 2068
+    },
+    {
+      "epoch": 0.36179234972677593,
+      "grad_norm": 1.3697995033730404,
+      "learning_rate": 0.0001957263452499195,
+      "loss": 14.7713,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3619672131147541,
+      "grad_norm": 1.3377695364967928,
+      "learning_rate": 0.00019572088286011163,
+      "loss": 14.5001,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3621420765027322,
+      "grad_norm": 1.395569887701733,
+      "learning_rate": 0.00019571541705796616,
+      "loss": 14.7675,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3623169398907104,
+      "grad_norm": 1.4091299694560664,
+      "learning_rate": 0.00019570994784367786,
+      "loss": 14.5642,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3624918032786885,
+      "grad_norm": 1.3964179094459386,
+      "learning_rate": 0.00019570447521744178,
+      "loss": 14.5952,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3626666666666667,
+      "grad_norm": 1.461064172770821,
+      "learning_rate": 0.00019569899917945294,
+      "loss": 14.3649,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3628415300546448,
+      "grad_norm": 1.293818468695891,
+      "learning_rate": 0.00019569351972990664,
+      "loss": 14.3534,
+      "step": 2075
+    },
+    {
+      "epoch": 0.36301639344262293,
+      "grad_norm": 1.53112563567672,
+      "learning_rate": 0.00019568803686899816,
+      "loss": 14.5196,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3631912568306011,
+      "grad_norm": 1.4946838593464862,
+      "learning_rate": 0.00019568255059692293,
+      "loss": 14.4349,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3633661202185792,
+      "grad_norm": 1.189244264631987,
+      "learning_rate": 0.0001956770609138766,
+      "loss": 14.5753,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3635409836065574,
+      "grad_norm": 1.3829953196157143,
+      "learning_rate": 0.00019567156782005483,
+      "loss": 14.4937,
+      "step": 2079
+    },
+    {
+      "epoch": 0.3637158469945355,
+      "grad_norm": 1.4515220466300212,
+      "learning_rate": 0.0001956660713156534,
+      "loss": 14.656,
+      "step": 2080
+    },
+    {
+      "epoch": 0.36389071038251364,
+      "grad_norm": 1.282655636551277,
+      "learning_rate": 0.00019566057140086836,
+      "loss": 14.2587,
+      "step": 2081
+    },
+    {
+      "epoch": 0.3640655737704918,
+      "grad_norm": 1.6459818876662669,
+      "learning_rate": 0.00019565506807589568,
+      "loss": 14.4693,
+      "step": 2082
+    },
+    {
+      "epoch": 0.36424043715846993,
+      "grad_norm": 1.3038207597519382,
+      "learning_rate": 0.00019564956134093163,
+      "loss": 14.4404,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3644153005464481,
+      "grad_norm": 1.511055920824356,
+      "learning_rate": 0.00019564405119617244,
+      "loss": 14.5629,
+      "step": 2084
+    },
+    {
+      "epoch": 0.3645901639344262,
+      "grad_norm": 1.6011093274891863,
+      "learning_rate": 0.00019563853764181457,
+      "loss": 14.6647,
+      "step": 2085
+    },
+    {
+      "epoch": 0.36476502732240434,
+      "grad_norm": 1.2565226504090952,
+      "learning_rate": 0.00019563302067805455,
+      "loss": 14.2293,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3649398907103825,
+      "grad_norm": 1.5532290114760001,
+      "learning_rate": 0.00019562750030508913,
+      "loss": 14.329,
+      "step": 2087
+    },
+    {
+      "epoch": 0.36511475409836064,
+      "grad_norm": 1.3166940574097687,
+      "learning_rate": 0.00019562197652311505,
+      "loss": 14.5681,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3652896174863388,
+      "grad_norm": 1.8293925316788369,
+      "learning_rate": 0.00019561644933232918,
+      "loss": 14.6405,
+      "step": 2089
+    },
+    {
+      "epoch": 0.36546448087431693,
+      "grad_norm": 1.452758984307232,
+      "learning_rate": 0.00019561091873292866,
+      "loss": 14.6183,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3656393442622951,
+      "grad_norm": 1.4277412881537463,
+      "learning_rate": 0.00019560538472511056,
+      "loss": 14.4972,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3658142076502732,
+      "grad_norm": 1.3908122008767771,
+      "learning_rate": 0.0001955998473090722,
+      "loss": 14.6224,
+      "step": 2092
+    },
+    {
+      "epoch": 0.36598907103825135,
+      "grad_norm": 1.465304426652499,
+      "learning_rate": 0.00019559430648501103,
+      "loss": 14.4647,
+      "step": 2093
+    },
+    {
+      "epoch": 0.3661639344262295,
+      "grad_norm": 1.635564918591344,
+      "learning_rate": 0.00019558876225312446,
+      "loss": 14.5676,
+      "step": 2094
+    },
+    {
+      "epoch": 0.36633879781420764,
+      "grad_norm": 1.6535676920370241,
+      "learning_rate": 0.00019558321461361024,
+      "loss": 14.3523,
+      "step": 2095
+    },
+    {
+      "epoch": 0.3665136612021858,
+      "grad_norm": 1.4087727457209378,
+      "learning_rate": 0.00019557766356666608,
+      "loss": 14.5793,
+      "step": 2096
+    },
+    {
+      "epoch": 0.36668852459016393,
+      "grad_norm": 1.4635500281224136,
+      "learning_rate": 0.00019557210911248988,
+      "loss": 14.6393,
+      "step": 2097
+    },
+    {
+      "epoch": 0.36686338797814205,
+      "grad_norm": 1.4678780988034394,
+      "learning_rate": 0.00019556655125127964,
+      "loss": 14.6125,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3670382513661202,
+      "grad_norm": 1.2957316971217783,
+      "learning_rate": 0.00019556098998323352,
+      "loss": 14.5012,
+      "step": 2099
+    },
+    {
+      "epoch": 0.36721311475409835,
+      "grad_norm": 1.4655772916553491,
+      "learning_rate": 0.00019555542530854978,
+      "loss": 14.4796,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3673879781420765,
+      "grad_norm": 1.279029059007525,
+      "learning_rate": 0.00019554985722742676,
+      "loss": 14.4365,
+      "step": 2101
+    },
+    {
+      "epoch": 0.36756284153005464,
+      "grad_norm": 1.3271252100600432,
+      "learning_rate": 0.00019554428574006296,
+      "loss": 14.7177,
+      "step": 2102
+    },
+    {
+      "epoch": 0.36773770491803276,
+      "grad_norm": 1.2963204460792819,
+      "learning_rate": 0.000195538710846657,
+      "loss": 14.5744,
+      "step": 2103
+    },
+    {
+      "epoch": 0.36791256830601093,
+      "grad_norm": 1.3123620269022467,
+      "learning_rate": 0.00019553313254740763,
+      "loss": 14.3918,
+      "step": 2104
+    },
+    {
+      "epoch": 0.36808743169398905,
+      "grad_norm": 1.2986262819710828,
+      "learning_rate": 0.00019552755084251372,
+      "loss": 14.5495,
+      "step": 2105
+    },
+    {
+      "epoch": 0.36826229508196723,
+      "grad_norm": 1.5325342706111704,
+      "learning_rate": 0.0001955219657321742,
+      "loss": 14.5012,
+      "step": 2106
+    },
+    {
+      "epoch": 0.36843715846994535,
+      "grad_norm": 1.2105007567547832,
+      "learning_rate": 0.00019551637721658822,
+      "loss": 14.2837,
+      "step": 2107
+    },
+    {
+      "epoch": 0.3686120218579235,
+      "grad_norm": 1.6213342155779322,
+      "learning_rate": 0.000195510785295955,
+      "loss": 14.6472,
+      "step": 2108
+    },
+    {
+      "epoch": 0.36878688524590164,
+      "grad_norm": 1.2720062541344659,
+      "learning_rate": 0.0001955051899704739,
+      "loss": 14.3193,
+      "step": 2109
+    },
+    {
+      "epoch": 0.36896174863387976,
+      "grad_norm": 1.6856547790484573,
+      "learning_rate": 0.00019549959124034433,
+      "loss": 14.4781,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36913661202185793,
+      "grad_norm": 1.2219368185361192,
+      "learning_rate": 0.0001954939891057659,
+      "loss": 14.3645,
+      "step": 2111
+    },
+    {
+      "epoch": 0.36931147540983605,
+      "grad_norm": 1.6206406272872877,
+      "learning_rate": 0.00019548838356693837,
+      "loss": 14.5518,
+      "step": 2112
+    },
+    {
+      "epoch": 0.36948633879781423,
+      "grad_norm": 1.4190977704364682,
+      "learning_rate": 0.0001954827746240615,
+      "loss": 14.5981,
+      "step": 2113
+    },
+    {
+      "epoch": 0.36966120218579235,
+      "grad_norm": 1.312964474018905,
+      "learning_rate": 0.00019547716227733529,
+      "loss": 14.6501,
+      "step": 2114
+    },
+    {
+      "epoch": 0.36983606557377047,
+      "grad_norm": 1.380378169503319,
+      "learning_rate": 0.00019547154652695977,
+      "loss": 14.3767,
+      "step": 2115
+    },
+    {
+      "epoch": 0.37001092896174864,
+      "grad_norm": 1.404008768390226,
+      "learning_rate": 0.0001954659273731352,
+      "loss": 14.6838,
+      "step": 2116
+    },
+    {
+      "epoch": 0.37018579234972676,
+      "grad_norm": 1.3744195327376445,
+      "learning_rate": 0.00019546030481606182,
+      "loss": 14.4298,
+      "step": 2117
+    },
+    {
+      "epoch": 0.37036065573770494,
+      "grad_norm": 1.3474255569216465,
+      "learning_rate": 0.00019545467885594015,
+      "loss": 14.6912,
+      "step": 2118
+    },
+    {
+      "epoch": 0.37053551912568305,
+      "grad_norm": 1.4337769464851768,
+      "learning_rate": 0.0001954490494929707,
+      "loss": 14.4285,
+      "step": 2119
+    },
+    {
+      "epoch": 0.37071038251366123,
+      "grad_norm": 1.3065122770259308,
+      "learning_rate": 0.00019544341672735413,
+      "loss": 14.5876,
+      "step": 2120
+    },
+    {
+      "epoch": 0.37088524590163935,
+      "grad_norm": 1.7071246170489875,
+      "learning_rate": 0.0001954377805592913,
+      "loss": 14.6202,
+      "step": 2121
+    },
+    {
+      "epoch": 0.37106010928961747,
+      "grad_norm": 1.3324442932481844,
+      "learning_rate": 0.00019543214098898307,
+      "loss": 14.2039,
+      "step": 2122
+    },
+    {
+      "epoch": 0.37123497267759564,
+      "grad_norm": 1.393116946066203,
+      "learning_rate": 0.00019542649801663052,
+      "loss": 14.5995,
+      "step": 2123
+    },
+    {
+      "epoch": 0.37140983606557376,
+      "grad_norm": 1.2913484843471483,
+      "learning_rate": 0.00019542085164243482,
+      "loss": 14.5505,
+      "step": 2124
+    },
+    {
+      "epoch": 0.37158469945355194,
+      "grad_norm": 1.3239605918846198,
+      "learning_rate": 0.00019541520186659725,
+      "loss": 14.6363,
+      "step": 2125
+    },
+    {
+      "epoch": 0.37175956284153006,
+      "grad_norm": 1.4245911266442501,
+      "learning_rate": 0.0001954095486893192,
+      "loss": 14.619,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3719344262295082,
+      "grad_norm": 1.3386157701032495,
+      "learning_rate": 0.00019540389211080222,
+      "loss": 14.6174,
+      "step": 2127
+    },
+    {
+      "epoch": 0.37210928961748635,
+      "grad_norm": 1.2966278510733862,
+      "learning_rate": 0.00019539823213124798,
+      "loss": 14.59,
+      "step": 2128
+    },
+    {
+      "epoch": 0.37228415300546447,
+      "grad_norm": 1.286098860983407,
+      "learning_rate": 0.00019539256875085817,
+      "loss": 14.5913,
+      "step": 2129
+    },
+    {
+      "epoch": 0.37245901639344264,
+      "grad_norm": 1.2074610448187273,
+      "learning_rate": 0.00019538690196983476,
+      "loss": 14.4688,
+      "step": 2130
+    },
+    {
+      "epoch": 0.37263387978142076,
+      "grad_norm": 1.3332853795398116,
+      "learning_rate": 0.00019538123178837978,
+      "loss": 14.45,
+      "step": 2131
+    },
+    {
+      "epoch": 0.3728087431693989,
+      "grad_norm": 1.3162665550548427,
+      "learning_rate": 0.00019537555820669527,
+      "loss": 14.7174,
+      "step": 2132
+    },
+    {
+      "epoch": 0.37298360655737706,
+      "grad_norm": 1.358962152172808,
+      "learning_rate": 0.00019536988122498356,
+      "loss": 14.499,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3731584699453552,
+      "grad_norm": 1.485646652327959,
+      "learning_rate": 0.00019536420084344705,
+      "loss": 14.5691,
+      "step": 2134
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 1.2890100440417238,
+      "learning_rate": 0.00019535851706228816,
+      "loss": 14.6745,
+      "step": 2135
+    },
+    {
+      "epoch": 0.37350819672131147,
+      "grad_norm": 1.320115897603953,
+      "learning_rate": 0.00019535282988170956,
+      "loss": 14.499,
+      "step": 2136
+    },
+    {
+      "epoch": 0.37368306010928964,
+      "grad_norm": 1.4081931581855835,
+      "learning_rate": 0.00019534713930191398,
+      "loss": 14.3921,
+      "step": 2137
+    },
+    {
+      "epoch": 0.37385792349726776,
+      "grad_norm": 1.437760014960006,
+      "learning_rate": 0.0001953414453231043,
+      "loss": 14.1352,
+      "step": 2138
+    },
+    {
+      "epoch": 0.3740327868852459,
+      "grad_norm": 1.3353545432278233,
+      "learning_rate": 0.00019533574794548347,
+      "loss": 14.5345,
+      "step": 2139
+    },
+    {
+      "epoch": 0.37420765027322406,
+      "grad_norm": 1.2614238805721287,
+      "learning_rate": 0.00019533004716925466,
+      "loss": 14.1567,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3743825136612022,
+      "grad_norm": 1.5010156086463093,
+      "learning_rate": 0.000195324342994621,
+      "loss": 14.5815,
+      "step": 2141
+    },
+    {
+      "epoch": 0.37455737704918035,
+      "grad_norm": 1.4567210372221295,
+      "learning_rate": 0.0001953186354217859,
+      "loss": 14.567,
+      "step": 2142
+    },
+    {
+      "epoch": 0.37473224043715847,
+      "grad_norm": 1.6873517156395064,
+      "learning_rate": 0.00019531292445095283,
+      "loss": 14.3584,
+      "step": 2143
+    },
+    {
+      "epoch": 0.3749071038251366,
+      "grad_norm": 1.5714406506477572,
+      "learning_rate": 0.00019530721008232532,
+      "loss": 14.7987,
+      "step": 2144
+    },
+    {
+      "epoch": 0.37508196721311476,
+      "grad_norm": 1.365236285692513,
+      "learning_rate": 0.00019530149231610717,
+      "loss": 14.5818,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3752568306010929,
+      "grad_norm": 1.441228235867867,
+      "learning_rate": 0.00019529577115250216,
+      "loss": 14.6956,
+      "step": 2146
+    },
+    {
+      "epoch": 0.37543169398907106,
+      "grad_norm": 1.407503432516305,
+      "learning_rate": 0.00019529004659171425,
+      "loss": 14.6692,
+      "step": 2147
+    },
+    {
+      "epoch": 0.3756065573770492,
+      "grad_norm": 1.2752777431644522,
+      "learning_rate": 0.00019528431863394748,
+      "loss": 14.3925,
+      "step": 2148
+    },
+    {
+      "epoch": 0.3757814207650273,
+      "grad_norm": 1.2920950553770378,
+      "learning_rate": 0.0001952785872794061,
+      "loss": 14.8852,
+      "step": 2149
+    },
+    {
+      "epoch": 0.37595628415300547,
+      "grad_norm": 1.5312200423660065,
+      "learning_rate": 0.00019527285252829443,
+      "loss": 14.5085,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3761311475409836,
+      "grad_norm": 1.4000176781606608,
+      "learning_rate": 0.00019526711438081686,
+      "loss": 14.6166,
+      "step": 2151
+    },
+    {
+      "epoch": 0.37630601092896176,
+      "grad_norm": 1.2516073862015533,
+      "learning_rate": 0.00019526137283717795,
+      "loss": 14.5483,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3764808743169399,
+      "grad_norm": 1.2899002884579716,
+      "learning_rate": 0.00019525562789758243,
+      "loss": 14.09,
+      "step": 2153
+    },
+    {
+      "epoch": 0.37665573770491806,
+      "grad_norm": 1.332142492742357,
+      "learning_rate": 0.00019524987956223507,
+      "loss": 14.6249,
+      "step": 2154
+    },
+    {
+      "epoch": 0.3768306010928962,
+      "grad_norm": 1.2639328886451406,
+      "learning_rate": 0.00019524412783134077,
+      "loss": 14.3075,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3770054644808743,
+      "grad_norm": 1.3008277816490215,
+      "learning_rate": 0.0001952383727051046,
+      "loss": 14.3213,
+      "step": 2156
+    },
+    {
+      "epoch": 0.37718032786885247,
+      "grad_norm": 1.2830098683703632,
+      "learning_rate": 0.00019523261418373172,
+      "loss": 14.4805,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3773551912568306,
+      "grad_norm": 1.2813943846179958,
+      "learning_rate": 0.00019522685226742743,
+      "loss": 14.7577,
+      "step": 2158
+    },
+    {
+      "epoch": 0.37753005464480877,
+      "grad_norm": 1.3762981300388364,
+      "learning_rate": 0.00019522108695639709,
+      "loss": 14.2876,
+      "step": 2159
+    },
+    {
+      "epoch": 0.3777049180327869,
+      "grad_norm": 1.3428755036753888,
+      "learning_rate": 0.00019521531825084627,
+      "loss": 14.571,
+      "step": 2160
+    },
+    {
+      "epoch": 0.377879781420765,
+      "grad_norm": 1.293803847020964,
+      "learning_rate": 0.0001952095461509806,
+      "loss": 14.6988,
+      "step": 2161
+    },
+    {
+      "epoch": 0.3780546448087432,
+      "grad_norm": 1.5724439578196798,
+      "learning_rate": 0.00019520377065700584,
+      "loss": 14.4867,
+      "step": 2162
+    },
+    {
+      "epoch": 0.3782295081967213,
+      "grad_norm": 1.391497296621329,
+      "learning_rate": 0.0001951979917691279,
+      "loss": 14.7746,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3784043715846995,
+      "grad_norm": 1.391927565072659,
+      "learning_rate": 0.00019519220948755278,
+      "loss": 14.4276,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3785792349726776,
+      "grad_norm": 1.5458221861838217,
+      "learning_rate": 0.0001951864238124866,
+      "loss": 14.5188,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3787540983606557,
+      "grad_norm": 1.2917917677743096,
+      "learning_rate": 0.00019518063474413563,
+      "loss": 14.5752,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3789289617486339,
+      "grad_norm": 1.3389858758700832,
+      "learning_rate": 0.00019517484228270623,
+      "loss": 14.4183,
+      "step": 2167
+    },
+    {
+      "epoch": 0.379103825136612,
+      "grad_norm": 1.3254235163310677,
+      "learning_rate": 0.00019516904642840493,
+      "loss": 14.4811,
+      "step": 2168
+    },
+    {
+      "epoch": 0.3792786885245902,
+      "grad_norm": 1.4233042398329134,
+      "learning_rate": 0.00019516324718143828,
+      "loss": 14.6186,
+      "step": 2169
+    },
+    {
+      "epoch": 0.3794535519125683,
+      "grad_norm": 1.328995514674753,
+      "learning_rate": 0.0001951574445420131,
+      "loss": 14.4451,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3796284153005465,
+      "grad_norm": 1.4255516404632071,
+      "learning_rate": 0.00019515163851033615,
+      "loss": 14.4896,
+      "step": 2171
+    },
+    {
+      "epoch": 0.3798032786885246,
+      "grad_norm": 1.3052830103085729,
+      "learning_rate": 0.00019514582908661452,
+      "loss": 14.4347,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3799781420765027,
+      "grad_norm": 1.4656170057280637,
+      "learning_rate": 0.0001951400162710552,
+      "loss": 14.5635,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3801530054644809,
+      "grad_norm": 1.309935358040408,
+      "learning_rate": 0.00019513420006386548,
+      "loss": 14.2247,
+      "step": 2174
+    },
+    {
+      "epoch": 0.380327868852459,
+      "grad_norm": 1.3952121974498495,
+      "learning_rate": 0.00019512838046525265,
+      "loss": 14.4731,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3805027322404372,
+      "grad_norm": 1.4381365881112114,
+      "learning_rate": 0.00019512255747542424,
+      "loss": 14.4967,
+      "step": 2176
+    },
+    {
+      "epoch": 0.3806775956284153,
+      "grad_norm": 1.314550116342945,
+      "learning_rate": 0.0001951167310945878,
+      "loss": 14.3826,
+      "step": 2177
+    },
+    {
+      "epoch": 0.3808524590163934,
+      "grad_norm": 1.3998129123048473,
+      "learning_rate": 0.000195110901322951,
+      "loss": 14.5204,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3810273224043716,
+      "grad_norm": 1.3174480444716092,
+      "learning_rate": 0.0001951050681607217,
+      "loss": 14.5172,
+      "step": 2179
+    },
+    {
+      "epoch": 0.3812021857923497,
+      "grad_norm": 1.2741648284625087,
+      "learning_rate": 0.00019509923160810787,
+      "loss": 14.343,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3813770491803279,
+      "grad_norm": 1.2975010190603553,
+      "learning_rate": 0.00019509339166531747,
+      "loss": 14.3676,
+      "step": 2181
+    },
+    {
+      "epoch": 0.381551912568306,
+      "grad_norm": 1.2475177825199941,
+      "learning_rate": 0.0001950875483325588,
+      "loss": 14.6398,
+      "step": 2182
+    },
+    {
+      "epoch": 0.3817267759562841,
+      "grad_norm": 1.2616791559012888,
+      "learning_rate": 0.00019508170161004016,
+      "loss": 14.6076,
+      "step": 2183
+    },
+    {
+      "epoch": 0.3819016393442623,
+      "grad_norm": 1.391436091362934,
+      "learning_rate": 0.00019507585149796994,
+      "loss": 14.1515,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3820765027322404,
+      "grad_norm": 1.3337674134003623,
+      "learning_rate": 0.00019506999799655668,
+      "loss": 14.6608,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3822513661202186,
+      "grad_norm": 1.3505203920127198,
+      "learning_rate": 0.00019506414110600907,
+      "loss": 14.7541,
+      "step": 2186
+    },
+    {
+      "epoch": 0.3824262295081967,
+      "grad_norm": 1.4221413766060111,
+      "learning_rate": 0.0001950582808265359,
+      "loss": 14.6973,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3826010928961749,
+      "grad_norm": 1.391181338203451,
+      "learning_rate": 0.0001950524171583461,
+      "loss": 14.4463,
+      "step": 2188
+    },
+    {
+      "epoch": 0.382775956284153,
+      "grad_norm": 1.3715852119619636,
+      "learning_rate": 0.00019504655010164866,
+      "loss": 14.5012,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3829508196721311,
+      "grad_norm": 1.3168396001681442,
+      "learning_rate": 0.00019504067965665277,
+      "loss": 14.5678,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3831256830601093,
+      "grad_norm": 1.4310297910141672,
+      "learning_rate": 0.00019503480582356767,
+      "loss": 14.4248,
+      "step": 2191
+    },
+    {
+      "epoch": 0.3833005464480874,
+      "grad_norm": 1.2827611216361843,
+      "learning_rate": 0.00019502892860260283,
+      "loss": 14.1721,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3834754098360656,
+      "grad_norm": 1.2515792314220222,
+      "learning_rate": 0.0001950230479939677,
+      "loss": 14.6532,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3836502732240437,
+      "grad_norm": 1.435180917612355,
+      "learning_rate": 0.00019501716399787192,
+      "loss": 14.6534,
+      "step": 2194
+    },
+    {
+      "epoch": 0.38382513661202183,
+      "grad_norm": 1.2977645591575382,
+      "learning_rate": 0.00019501127661452524,
+      "loss": 14.7265,
+      "step": 2195
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.3020010515168834,
+      "learning_rate": 0.00019500538584413756,
+      "loss": 14.6098,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3841748633879781,
+      "grad_norm": 1.4196532143368013,
+      "learning_rate": 0.0001949994916869189,
+      "loss": 14.4324,
+      "step": 2197
+    },
+    {
+      "epoch": 0.3843497267759563,
+      "grad_norm": 1.2876245000698028,
+      "learning_rate": 0.00019499359414307935,
+      "loss": 14.3118,
+      "step": 2198
+    },
+    {
+      "epoch": 0.3845245901639344,
+      "grad_norm": 1.4988673955581826,
+      "learning_rate": 0.00019498769321282913,
+      "loss": 14.2304,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3846994535519126,
+      "grad_norm": 1.2427086746459695,
+      "learning_rate": 0.0001949817888963787,
+      "loss": 14.3424,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3848743169398907,
+      "grad_norm": 1.4596236612472768,
+      "learning_rate": 0.00019497588119393838,
+      "loss": 14.3134,
+      "step": 2201
+    },
+    {
+      "epoch": 0.38504918032786883,
+      "grad_norm": 1.4794741882086404,
+      "learning_rate": 0.00019496997010571892,
+      "loss": 14.4923,
+      "step": 2202
+    },
+    {
+      "epoch": 0.385224043715847,
+      "grad_norm": 1.2907594714236577,
+      "learning_rate": 0.000194964055631931,
+      "loss": 14.3187,
+      "step": 2203
+    },
+    {
+      "epoch": 0.38539890710382513,
+      "grad_norm": 1.451599296172969,
+      "learning_rate": 0.0001949581377727854,
+      "loss": 14.7114,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3855737704918033,
+      "grad_norm": 1.8923815404208637,
+      "learning_rate": 0.00019495221652849316,
+      "loss": 14.4529,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3857486338797814,
+      "grad_norm": 1.319960257170267,
+      "learning_rate": 0.00019494629189926532,
+      "loss": 14.6252,
+      "step": 2206
+    },
+    {
+      "epoch": 0.38592349726775954,
+      "grad_norm": 1.589198093581371,
+      "learning_rate": 0.00019494036388531313,
+      "loss": 14.3772,
+      "step": 2207
+    },
+    {
+      "epoch": 0.3860983606557377,
+      "grad_norm": 1.1997165281303332,
+      "learning_rate": 0.0001949344324868479,
+      "loss": 14.5313,
+      "step": 2208
+    },
+    {
+      "epoch": 0.38627322404371583,
+      "grad_norm": 1.620339104084589,
+      "learning_rate": 0.00019492849770408105,
+      "loss": 14.6546,
+      "step": 2209
+    },
+    {
+      "epoch": 0.386448087431694,
+      "grad_norm": 1.220558064643469,
+      "learning_rate": 0.00019492255953722419,
+      "loss": 14.4266,
+      "step": 2210
+    },
+    {
+      "epoch": 0.38662295081967213,
+      "grad_norm": 1.5680382558402641,
+      "learning_rate": 0.00019491661798648897,
+      "loss": 14.4388,
+      "step": 2211
+    },
+    {
+      "epoch": 0.38679781420765025,
+      "grad_norm": 1.2203375350411219,
+      "learning_rate": 0.00019491067305208722,
+      "loss": 14.5582,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3869726775956284,
+      "grad_norm": 1.55225375858333,
+      "learning_rate": 0.00019490472473423088,
+      "loss": 14.2872,
+      "step": 2213
+    },
+    {
+      "epoch": 0.38714754098360654,
+      "grad_norm": 1.1265239656581891,
+      "learning_rate": 0.000194898773033132,
+      "loss": 14.3721,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3873224043715847,
+      "grad_norm": 1.3655425590775916,
+      "learning_rate": 0.0001948928179490027,
+      "loss": 14.3522,
+      "step": 2215
+    },
+    {
+      "epoch": 0.38749726775956284,
+      "grad_norm": 1.542877146833185,
+      "learning_rate": 0.0001948868594820553,
+      "loss": 14.5304,
+      "step": 2216
+    },
+    {
+      "epoch": 0.387672131147541,
+      "grad_norm": 1.5145896311864229,
+      "learning_rate": 0.00019488089763250224,
+      "loss": 14.6245,
+      "step": 2217
+    },
+    {
+      "epoch": 0.38784699453551913,
+      "grad_norm": 1.4008488693261774,
+      "learning_rate": 0.00019487493240055606,
+      "loss": 14.6606,
+      "step": 2218
+    },
+    {
+      "epoch": 0.38802185792349725,
+      "grad_norm": 1.2158737825965045,
+      "learning_rate": 0.00019486896378642939,
+      "loss": 14.3176,
+      "step": 2219
+    },
+    {
+      "epoch": 0.3881967213114754,
+      "grad_norm": 1.4054782798228376,
+      "learning_rate": 0.000194862991790335,
+      "loss": 14.5477,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38837158469945354,
+      "grad_norm": 1.393783687576429,
+      "learning_rate": 0.00019485701641248576,
+      "loss": 14.0242,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3885464480874317,
+      "grad_norm": 1.2759858546021083,
+      "learning_rate": 0.0001948510376530947,
+      "loss": 14.0878,
+      "step": 2222
+    },
+    {
+      "epoch": 0.38872131147540984,
+      "grad_norm": 1.5344298140302186,
+      "learning_rate": 0.000194845055512375,
+      "loss": 14.2331,
+      "step": 2223
+    },
+    {
+      "epoch": 0.38889617486338796,
+      "grad_norm": 1.285602587995282,
+      "learning_rate": 0.0001948390699905399,
+      "loss": 14.5185,
+      "step": 2224
+    },
+    {
+      "epoch": 0.38907103825136613,
+      "grad_norm": 1.38759371965707,
+      "learning_rate": 0.00019483308108780273,
+      "loss": 14.5425,
+      "step": 2225
+    },
+    {
+      "epoch": 0.38924590163934425,
+      "grad_norm": 1.392487756553308,
+      "learning_rate": 0.00019482708880437706,
+      "loss": 14.4852,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3894207650273224,
+      "grad_norm": 1.2283229587642075,
+      "learning_rate": 0.00019482109314047642,
+      "loss": 13.9746,
+      "step": 2227
+    },
+    {
+      "epoch": 0.38959562841530054,
+      "grad_norm": 1.2430899105309483,
+      "learning_rate": 0.00019481509409631463,
+      "loss": 14.0473,
+      "step": 2228
+    },
+    {
+      "epoch": 0.38977049180327866,
+      "grad_norm": 1.3767054742370097,
+      "learning_rate": 0.0001948090916721055,
+      "loss": 14.2891,
+      "step": 2229
+    },
+    {
+      "epoch": 0.38994535519125684,
+      "grad_norm": 1.6105179392872082,
+      "learning_rate": 0.00019480308586806305,
+      "loss": 14.5394,
+      "step": 2230
+    },
+    {
+      "epoch": 0.39012021857923496,
+      "grad_norm": 1.5497642555694036,
+      "learning_rate": 0.0001947970766844013,
+      "loss": 14.3276,
+      "step": 2231
+    },
+    {
+      "epoch": 0.39029508196721313,
+      "grad_norm": 1.4413599995098616,
+      "learning_rate": 0.00019479106412133456,
+      "loss": 14.1592,
+      "step": 2232
+    },
+    {
+      "epoch": 0.39046994535519125,
+      "grad_norm": 1.4818023418559778,
+      "learning_rate": 0.00019478504817907716,
+      "loss": 14.4188,
+      "step": 2233
+    },
+    {
+      "epoch": 0.3906448087431694,
+      "grad_norm": 1.3653157968725609,
+      "learning_rate": 0.00019477902885784348,
+      "loss": 14.5216,
+      "step": 2234
+    },
+    {
+      "epoch": 0.39081967213114754,
+      "grad_norm": 1.4874312671460237,
+      "learning_rate": 0.00019477300615784818,
+      "loss": 14.4962,
+      "step": 2235
+    },
+    {
+      "epoch": 0.39099453551912566,
+      "grad_norm": 1.32811264240576,
+      "learning_rate": 0.00019476698007930599,
+      "loss": 14.5426,
+      "step": 2236
+    },
+    {
+      "epoch": 0.39116939890710384,
+      "grad_norm": 1.4982764047667834,
+      "learning_rate": 0.00019476095062243161,
+      "loss": 14.3579,
+      "step": 2237
+    },
+    {
+      "epoch": 0.39134426229508196,
+      "grad_norm": 1.43461429089932,
+      "learning_rate": 0.0001947549177874401,
+      "loss": 14.421,
+      "step": 2238
+    },
+    {
+      "epoch": 0.39151912568306013,
+      "grad_norm": 1.569647173143244,
+      "learning_rate": 0.00019474888157454648,
+      "loss": 14.4645,
+      "step": 2239
+    },
+    {
+      "epoch": 0.39169398907103825,
+      "grad_norm": 1.2424044910941385,
+      "learning_rate": 0.00019474284198396592,
+      "loss": 14.1408,
+      "step": 2240
+    },
+    {
+      "epoch": 0.39186885245901637,
+      "grad_norm": 1.3898948964706888,
+      "learning_rate": 0.00019473679901591372,
+      "loss": 14.1953,
+      "step": 2241
+    },
+    {
+      "epoch": 0.39204371584699454,
+      "grad_norm": 1.401804702399091,
+      "learning_rate": 0.00019473075267060537,
+      "loss": 14.2442,
+      "step": 2242
+    },
+    {
+      "epoch": 0.39221857923497266,
+      "grad_norm": 1.2559645615464778,
+      "learning_rate": 0.00019472470294825637,
+      "loss": 14.544,
+      "step": 2243
+    },
+    {
+      "epoch": 0.39239344262295084,
+      "grad_norm": 1.2148576398140296,
+      "learning_rate": 0.00019471864984908232,
+      "loss": 14.0538,
+      "step": 2244
+    },
+    {
+      "epoch": 0.39256830601092896,
+      "grad_norm": 1.2646088386538168,
+      "learning_rate": 0.00019471259337329913,
+      "loss": 14.6328,
+      "step": 2245
+    },
+    {
+      "epoch": 0.3927431693989071,
+      "grad_norm": 1.3498135672409128,
+      "learning_rate": 0.0001947065335211226,
+      "loss": 14.2913,
+      "step": 2246
+    },
+    {
+      "epoch": 0.39291803278688525,
+      "grad_norm": 1.2884296936741326,
+      "learning_rate": 0.00019470047029276883,
+      "loss": 14.2864,
+      "step": 2247
+    },
+    {
+      "epoch": 0.39309289617486337,
+      "grad_norm": 1.2406304382958655,
+      "learning_rate": 0.00019469440368845392,
+      "loss": 14.3557,
+      "step": 2248
+    },
+    {
+      "epoch": 0.39326775956284155,
+      "grad_norm": 1.4114763060997866,
+      "learning_rate": 0.0001946883337083942,
+      "loss": 13.9915,
+      "step": 2249
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 1.2882382352155064,
+      "learning_rate": 0.00019468226035280597,
+      "loss": 14.1258,
+      "step": 2250
+    },
+    {
+      "epoch": 0.39361748633879784,
+      "grad_norm": 1.2503206702538745,
+      "learning_rate": 0.0001946761836219058,
+      "loss": 14.0355,
+      "step": 2251
+    },
+    {
+      "epoch": 0.39379234972677596,
+      "grad_norm": 1.332106061781732,
+      "learning_rate": 0.0001946701035159103,
+      "loss": 14.2354,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3939672131147541,
+      "grad_norm": 1.196026220200491,
+      "learning_rate": 0.00019466402003503623,
+      "loss": 14.2826,
+      "step": 2253
+    },
+    {
+      "epoch": 0.39414207650273225,
+      "grad_norm": 1.310548882451437,
+      "learning_rate": 0.00019465793317950043,
+      "loss": 14.3062,
+      "step": 2254
+    },
+    {
+      "epoch": 0.39431693989071037,
+      "grad_norm": 1.366297802578152,
+      "learning_rate": 0.00019465184294951992,
+      "loss": 14.5147,
+      "step": 2255
+    },
+    {
+      "epoch": 0.39449180327868855,
+      "grad_norm": 1.2481222551609268,
+      "learning_rate": 0.0001946457493453118,
+      "loss": 14.5343,
+      "step": 2256
+    },
+    {
+      "epoch": 0.39466666666666667,
+      "grad_norm": 1.5911853257288278,
+      "learning_rate": 0.0001946396523670933,
+      "loss": 14.4327,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3948415300546448,
+      "grad_norm": 1.5215496497789271,
+      "learning_rate": 0.00019463355201508175,
+      "loss": 14.3018,
+      "step": 2258
+    },
+    {
+      "epoch": 0.39501639344262296,
+      "grad_norm": 1.5236946260993527,
+      "learning_rate": 0.00019462744828949467,
+      "loss": 14.2096,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3951912568306011,
+      "grad_norm": 2.224343226041316,
+      "learning_rate": 0.00019462134119054957,
+      "loss": 14.2572,
+      "step": 2260
+    },
+    {
+      "epoch": 0.39536612021857925,
+      "grad_norm": 1.48888263244502,
+      "learning_rate": 0.00019461523071846422,
+      "loss": 14.2992,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3955409836065574,
+      "grad_norm": 1.538880581961395,
+      "learning_rate": 0.0001946091168734565,
+      "loss": 14.3284,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3957158469945355,
+      "grad_norm": 1.6668405125357222,
+      "learning_rate": 0.00019460299965574427,
+      "loss": 14.2906,
+      "step": 2263
+    },
+    {
+      "epoch": 0.39589071038251367,
+      "grad_norm": 1.3709953971499556,
+      "learning_rate": 0.0001945968790655456,
+      "loss": 14.3966,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3960655737704918,
+      "grad_norm": 1.9785067184906011,
+      "learning_rate": 0.00019459075510307874,
+      "loss": 14.3407,
+      "step": 2265
+    },
+    {
+      "epoch": 0.39624043715846996,
+      "grad_norm": 1.298384340384138,
+      "learning_rate": 0.000194584627768562,
+      "loss": 14.5178,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3964153005464481,
+      "grad_norm": 1.8572018356937254,
+      "learning_rate": 0.00019457849706221376,
+      "loss": 14.2753,
+      "step": 2267
+    },
+    {
+      "epoch": 0.39659016393442625,
+      "grad_norm": 1.3229622706708994,
+      "learning_rate": 0.0001945723629842526,
+      "loss": 14.5123,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3967650273224044,
+      "grad_norm": 1.7159669406983802,
+      "learning_rate": 0.00019456622553489723,
+      "loss": 14.4557,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3969398907103825,
+      "grad_norm": 1.3207769669176777,
+      "learning_rate": 0.0001945600847143664,
+      "loss": 14.3971,
+      "step": 2270
+    },
+    {
+      "epoch": 0.39711475409836067,
+      "grad_norm": 1.8631529252368189,
+      "learning_rate": 0.00019455394052287898,
+      "loss": 14.3529,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3972896174863388,
+      "grad_norm": 1.7309168248611855,
+      "learning_rate": 0.0001945477929606541,
+      "loss": 14.4197,
+      "step": 2272
+    },
+    {
+      "epoch": 0.39746448087431696,
+      "grad_norm": 1.5184260994861951,
+      "learning_rate": 0.00019454164202791088,
+      "loss": 14.1431,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3976393442622951,
+      "grad_norm": 1.5983448240114593,
+      "learning_rate": 0.00019453548772486854,
+      "loss": 14.2034,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3978142076502732,
+      "grad_norm": 1.3821755297210108,
+      "learning_rate": 0.00019452933005174655,
+      "loss": 14.3441,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3979890710382514,
+      "grad_norm": 1.6079457291241157,
+      "learning_rate": 0.00019452316900876436,
+      "loss": 14.4292,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3981639344262295,
+      "grad_norm": 1.2098407401426978,
+      "learning_rate": 0.00019451700459614167,
+      "loss": 14.4151,
+      "step": 2277
+    },
+    {
+      "epoch": 0.39833879781420767,
+      "grad_norm": 1.3057682681087908,
+      "learning_rate": 0.00019451083681409816,
+      "loss": 14.3507,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3985136612021858,
+      "grad_norm": 1.3993348319327088,
+      "learning_rate": 0.00019450466566285377,
+      "loss": 14.4976,
+      "step": 2279
+    },
+    {
+      "epoch": 0.39868852459016396,
+      "grad_norm": 1.288443951126012,
+      "learning_rate": 0.00019449849114262845,
+      "loss": 14.4923,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3988633879781421,
+      "grad_norm": 1.37734950007728,
+      "learning_rate": 0.00019449231325364234,
+      "loss": 14.4399,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3990382513661202,
+      "grad_norm": 1.2241366728767085,
+      "learning_rate": 0.00019448613199611565,
+      "loss": 14.2185,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3992131147540984,
+      "grad_norm": 1.2089262674341934,
+      "learning_rate": 0.00019447994737026875,
+      "loss": 14.1326,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3993879781420765,
+      "grad_norm": 1.3258461317596824,
+      "learning_rate": 0.00019447375937632212,
+      "loss": 14.4605,
+      "step": 2284
+    },
+    {
+      "epoch": 0.39956284153005467,
+      "grad_norm": 1.277966754519787,
+      "learning_rate": 0.00019446756801449637,
+      "loss": 14.4823,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3997377049180328,
+      "grad_norm": 1.3190588095498152,
+      "learning_rate": 0.0001944613732850122,
+      "loss": 14.4077,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3999125683060109,
+      "grad_norm": 1.2989165162081675,
+      "learning_rate": 0.0001944551751880904,
+      "loss": 14.3028,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4000874316939891,
+      "grad_norm": 1.2362392466143701,
+      "learning_rate": 0.00019444897372395198,
+      "loss": 14.4447,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4002622950819672,
+      "grad_norm": 1.3753096428162939,
+      "learning_rate": 0.000194442768892818,
+      "loss": 14.272,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4004371584699454,
+      "grad_norm": 1.4118532136442912,
+      "learning_rate": 0.00019443656069490964,
+      "loss": 14.2034,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4006120218579235,
+      "grad_norm": 1.4386269229353865,
+      "learning_rate": 0.00019443034913044823,
+      "loss": 14.2438,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4007868852459016,
+      "grad_norm": 1.2718690262826031,
+      "learning_rate": 0.00019442413419965525,
+      "loss": 14.1961,
+      "step": 2292
+    },
+    {
+      "epoch": 0.4009617486338798,
+      "grad_norm": 1.2944697194828119,
+      "learning_rate": 0.00019441791590275217,
+      "loss": 14.1527,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4011366120218579,
+      "grad_norm": 1.2194989712568618,
+      "learning_rate": 0.00019441169423996068,
+      "loss": 14.382,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4013114754098361,
+      "grad_norm": 1.463019671270362,
+      "learning_rate": 0.0001944054692115026,
+      "loss": 14.5256,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4014863387978142,
+      "grad_norm": 1.391305296741265,
+      "learning_rate": 0.0001943992408175999,
+      "loss": 14.4964,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4016612021857924,
+      "grad_norm": 1.485846862194901,
+      "learning_rate": 0.00019439300905847452,
+      "loss": 14.062,
+      "step": 2297
+    },
+    {
+      "epoch": 0.4018360655737705,
+      "grad_norm": 1.7781343115155528,
+      "learning_rate": 0.00019438677393434866,
+      "loss": 14.2639,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4020109289617486,
+      "grad_norm": 1.2546971525131352,
+      "learning_rate": 0.0001943805354454446,
+      "loss": 14.1678,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4021857923497268,
+      "grad_norm": 1.5609884747936316,
+      "learning_rate": 0.0001943742935919847,
+      "loss": 14.2365,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4023606557377049,
+      "grad_norm": 1.291269945072596,
+      "learning_rate": 0.00019436804837419153,
+      "loss": 14.3751,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4025355191256831,
+      "grad_norm": 1.5201543464653204,
+      "learning_rate": 0.00019436179979228765,
+      "loss": 14.3775,
+      "step": 2302
+    },
+    {
+      "epoch": 0.4027103825136612,
+      "grad_norm": 1.456905464759203,
+      "learning_rate": 0.00019435554784649588,
+      "loss": 14.3829,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4028852459016393,
+      "grad_norm": 1.3610638094765095,
+      "learning_rate": 0.00019434929253703907,
+      "loss": 14.1787,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4030601092896175,
+      "grad_norm": 1.3875729977752838,
+      "learning_rate": 0.0001943430338641402,
+      "loss": 13.9729,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4032349726775956,
+      "grad_norm": 1.3159814572861817,
+      "learning_rate": 0.00019433677182802242,
+      "loss": 13.9498,
+      "step": 2306
+    },
+    {
+      "epoch": 0.4034098360655738,
+      "grad_norm": 1.4072211878139025,
+      "learning_rate": 0.00019433050642890894,
+      "loss": 14.3657,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4035846994535519,
+      "grad_norm": 1.2683404872654593,
+      "learning_rate": 0.0001943242376670231,
+      "loss": 14.3744,
+      "step": 2308
+    },
+    {
+      "epoch": 0.40375956284153003,
+      "grad_norm": 1.4912922006667666,
+      "learning_rate": 0.00019431796554258846,
+      "loss": 14.179,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4039344262295082,
+      "grad_norm": 1.4137314214204624,
+      "learning_rate": 0.0001943116900558285,
+      "loss": 14.5483,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4041092896174863,
+      "grad_norm": 1.3305418668149835,
+      "learning_rate": 0.00019430541120696697,
+      "loss": 14.4634,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4042841530054645,
+      "grad_norm": 1.2303717791202715,
+      "learning_rate": 0.00019429912899622775,
+      "loss": 14.3479,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4044590163934426,
+      "grad_norm": 1.264538132007814,
+      "learning_rate": 0.00019429284342383472,
+      "loss": 14.1033,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4046338797814208,
+      "grad_norm": 1.452391454889947,
+      "learning_rate": 0.00019428655449001201,
+      "loss": 14.0936,
+      "step": 2314
+    },
+    {
+      "epoch": 0.4048087431693989,
+      "grad_norm": 1.3862786094002513,
+      "learning_rate": 0.00019428026219498379,
+      "loss": 14.0309,
+      "step": 2315
+    },
+    {
+      "epoch": 0.40498360655737703,
+      "grad_norm": 1.3228558390731706,
+      "learning_rate": 0.0001942739665389744,
+      "loss": 14.0458,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4051584699453552,
+      "grad_norm": 1.7338261211823827,
+      "learning_rate": 0.0001942676675222082,
+      "loss": 14.4281,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4053333333333333,
+      "grad_norm": 1.1798211121240945,
+      "learning_rate": 0.00019426136514490985,
+      "loss": 14.1195,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4055081967213115,
+      "grad_norm": 1.691892665662126,
+      "learning_rate": 0.0001942550594073039,
+      "loss": 14.0584,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4056830601092896,
+      "grad_norm": 1.2165329499055861,
+      "learning_rate": 0.00019424875030961527,
+      "loss": 14.0478,
+      "step": 2320
+    },
+    {
+      "epoch": 0.40585792349726774,
+      "grad_norm": 1.4989849683694314,
+      "learning_rate": 0.00019424243785206877,
+      "loss": 14.4494,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4060327868852459,
+      "grad_norm": 1.478738121680484,
+      "learning_rate": 0.00019423612203488947,
+      "loss": 14.3022,
+      "step": 2322
+    },
+    {
+      "epoch": 0.40620765027322403,
+      "grad_norm": 1.5805264976915128,
+      "learning_rate": 0.00019422980285830254,
+      "loss": 14.5226,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4063825136612022,
+      "grad_norm": 1.5647310800700773,
+      "learning_rate": 0.00019422348032253325,
+      "loss": 14.4713,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4065573770491803,
+      "grad_norm": 1.514221602869899,
+      "learning_rate": 0.00019421715442780694,
+      "loss": 14.4567,
+      "step": 2325
+    },
+    {
+      "epoch": 0.40673224043715844,
+      "grad_norm": 1.310662132434029,
+      "learning_rate": 0.00019421082517434918,
+      "loss": 14.2331,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4069071038251366,
+      "grad_norm": 1.3001794175956565,
+      "learning_rate": 0.00019420449256238559,
+      "loss": 14.4457,
+      "step": 2327
+    },
+    {
+      "epoch": 0.40708196721311474,
+      "grad_norm": 1.5557962718130425,
+      "learning_rate": 0.00019419815659214187,
+      "loss": 14.2325,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4072568306010929,
+      "grad_norm": 1.2101733419474199,
+      "learning_rate": 0.00019419181726384395,
+      "loss": 14.3108,
+      "step": 2329
+    },
+    {
+      "epoch": 0.40743169398907103,
+      "grad_norm": 1.368599628749081,
+      "learning_rate": 0.00019418547457771777,
+      "loss": 14.3,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4076065573770492,
+      "grad_norm": 1.261350590096406,
+      "learning_rate": 0.00019417912853398947,
+      "loss": 14.371,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4077814207650273,
+      "grad_norm": 1.2767441895629654,
+      "learning_rate": 0.0001941727791328853,
+      "loss": 14.4444,
+      "step": 2332
+    },
+    {
+      "epoch": 0.40795628415300544,
+      "grad_norm": 1.3721048044916937,
+      "learning_rate": 0.00019416642637463158,
+      "loss": 14.7513,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4081311475409836,
+      "grad_norm": 1.333152475946102,
+      "learning_rate": 0.00019416007025945477,
+      "loss": 14.2676,
+      "step": 2334
+    },
+    {
+      "epoch": 0.40830601092896174,
+      "grad_norm": 1.3068740147740077,
+      "learning_rate": 0.00019415371078758143,
+      "loss": 13.9051,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4084808743169399,
+      "grad_norm": 1.20382410457215,
+      "learning_rate": 0.0001941473479592384,
+      "loss": 14.289,
+      "step": 2336
+    },
+    {
+      "epoch": 0.40865573770491803,
+      "grad_norm": 1.365903557280566,
+      "learning_rate": 0.00019414098177465233,
+      "loss": 14.3482,
+      "step": 2337
+    },
+    {
+      "epoch": 0.40883060109289615,
+      "grad_norm": 1.3382004317130736,
+      "learning_rate": 0.0001941346122340503,
+      "loss": 14.2991,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4090054644808743,
+      "grad_norm": 1.241057626182281,
+      "learning_rate": 0.0001941282393376593,
+      "loss": 14.5589,
+      "step": 2339
+    },
+    {
+      "epoch": 0.40918032786885244,
+      "grad_norm": 1.4566889296527659,
+      "learning_rate": 0.00019412186308570654,
+      "loss": 14.2869,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4093551912568306,
+      "grad_norm": 1.4443361052680914,
+      "learning_rate": 0.00019411548347841938,
+      "loss": 14.3001,
+      "step": 2341
+    },
+    {
+      "epoch": 0.40953005464480874,
+      "grad_norm": 1.3783942917017484,
+      "learning_rate": 0.00019410910051602514,
+      "loss": 14.4105,
+      "step": 2342
+    },
+    {
+      "epoch": 0.40970491803278686,
+      "grad_norm": 1.2496190926066386,
+      "learning_rate": 0.00019410271419875144,
+      "loss": 14.1878,
+      "step": 2343
+    },
+    {
+      "epoch": 0.40987978142076503,
+      "grad_norm": 1.4729595974720375,
+      "learning_rate": 0.00019409632452682595,
+      "loss": 14.5124,
+      "step": 2344
+    },
+    {
+      "epoch": 0.41005464480874315,
+      "grad_norm": 1.39630371222355,
+      "learning_rate": 0.0001940899315004764,
+      "loss": 14.3644,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4102295081967213,
+      "grad_norm": 1.2200208544984015,
+      "learning_rate": 0.00019408353511993074,
+      "loss": 14.2061,
+      "step": 2346
+    },
+    {
+      "epoch": 0.41040437158469945,
+      "grad_norm": 1.2377800678990594,
+      "learning_rate": 0.00019407713538541695,
+      "loss": 14.1392,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4105792349726776,
+      "grad_norm": 1.2100441109241513,
+      "learning_rate": 0.00019407073229716325,
+      "loss": 14.1132,
+      "step": 2348
+    },
+    {
+      "epoch": 0.41075409836065574,
+      "grad_norm": 1.3429674310180395,
+      "learning_rate": 0.00019406432585539782,
+      "loss": 14.4442,
+      "step": 2349
+    },
+    {
+      "epoch": 0.41092896174863386,
+      "grad_norm": 1.2729838665571835,
+      "learning_rate": 0.00019405791606034905,
+      "loss": 14.1705,
+      "step": 2350
+    },
+    {
+      "epoch": 0.41110382513661203,
+      "grad_norm": 1.4200203154947677,
+      "learning_rate": 0.00019405150291224553,
+      "loss": 14.0317,
+      "step": 2351
+    },
+    {
+      "epoch": 0.41127868852459015,
+      "grad_norm": 1.3479008547799805,
+      "learning_rate": 0.00019404508641131577,
+      "loss": 14.4487,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4114535519125683,
+      "grad_norm": 1.4982709991414698,
+      "learning_rate": 0.0001940386665577886,
+      "loss": 14.4684,
+      "step": 2353
+    },
+    {
+      "epoch": 0.41162841530054645,
+      "grad_norm": 1.3322195848798741,
+      "learning_rate": 0.0001940322433518928,
+      "loss": 14.2321,
+      "step": 2354
+    },
+    {
+      "epoch": 0.41180327868852457,
+      "grad_norm": 1.489407035319849,
+      "learning_rate": 0.00019402581679385742,
+      "loss": 14.3858,
+      "step": 2355
+    },
+    {
+      "epoch": 0.41197814207650274,
+      "grad_norm": 2.0263970893456213,
+      "learning_rate": 0.00019401938688391152,
+      "loss": 14.3125,
+      "step": 2356
+    },
+    {
+      "epoch": 0.41215300546448086,
+      "grad_norm": 1.2400976516578472,
+      "learning_rate": 0.00019401295362228433,
+      "loss": 14.1785,
+      "step": 2357
+    },
+    {
+      "epoch": 0.41232786885245903,
+      "grad_norm": 2.3836475359356513,
+      "learning_rate": 0.00019400651700920518,
+      "loss": 14.395,
+      "step": 2358
+    },
+    {
+      "epoch": 0.41250273224043715,
+      "grad_norm": 1.2994644086890064,
+      "learning_rate": 0.00019400007704490352,
+      "loss": 14.4806,
+      "step": 2359
+    },
+    {
+      "epoch": 0.41267759562841533,
+      "grad_norm": 2.026392128469461,
+      "learning_rate": 0.00019399363372960893,
+      "loss": 14.4646,
+      "step": 2360
+    },
+    {
+      "epoch": 0.41285245901639345,
+      "grad_norm": 1.7444505553145528,
+      "learning_rate": 0.00019398718706355116,
+      "loss": 14.2121,
+      "step": 2361
+    },
+    {
+      "epoch": 0.41302732240437157,
+      "grad_norm": 1.4611498141698975,
+      "learning_rate": 0.00019398073704695996,
+      "loss": 14.1025,
+      "step": 2362
+    },
+    {
+      "epoch": 0.41320218579234974,
+      "grad_norm": 1.5915290496987806,
+      "learning_rate": 0.0001939742836800653,
+      "loss": 14.129,
+      "step": 2363
+    },
+    {
+      "epoch": 0.41337704918032786,
+      "grad_norm": 1.265789757068292,
+      "learning_rate": 0.0001939678269630972,
+      "loss": 13.7971,
+      "step": 2364
+    },
+    {
+      "epoch": 0.41355191256830603,
+      "grad_norm": 1.617178606471799,
+      "learning_rate": 0.0001939613668962859,
+      "loss": 14.1926,
+      "step": 2365
+    },
+    {
+      "epoch": 0.41372677595628415,
+      "grad_norm": 1.271018936626067,
+      "learning_rate": 0.0001939549034798616,
+      "loss": 14.342,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4139016393442623,
+      "grad_norm": 1.343657863468849,
+      "learning_rate": 0.0001939484367140548,
+      "loss": 14.2044,
+      "step": 2367
+    },
+    {
+      "epoch": 0.41407650273224045,
+      "grad_norm": 1.2900982998352508,
+      "learning_rate": 0.00019394196659909598,
+      "loss": 13.9665,
+      "step": 2368
+    },
+    {
+      "epoch": 0.41425136612021857,
+      "grad_norm": 1.3522684750372107,
+      "learning_rate": 0.00019393549313521583,
+      "loss": 14.2624,
+      "step": 2369
+    },
+    {
+      "epoch": 0.41442622950819674,
+      "grad_norm": 1.2753371939661025,
+      "learning_rate": 0.00019392901632264508,
+      "loss": 14.3021,
+      "step": 2370
+    },
+    {
+      "epoch": 0.41460109289617486,
+      "grad_norm": 1.4196927623595181,
+      "learning_rate": 0.00019392253616161466,
+      "loss": 14.246,
+      "step": 2371
+    },
+    {
+      "epoch": 0.414775956284153,
+      "grad_norm": 1.3223623542598104,
+      "learning_rate": 0.00019391605265235557,
+      "loss": 13.8355,
+      "step": 2372
+    },
+    {
+      "epoch": 0.41495081967213115,
+      "grad_norm": 1.4024687611106585,
+      "learning_rate": 0.00019390956579509893,
+      "loss": 14.4974,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4151256830601093,
+      "grad_norm": 1.4431073513953583,
+      "learning_rate": 0.00019390307559007596,
+      "loss": 14.2706,
+      "step": 2374
+    },
+    {
+      "epoch": 0.41530054644808745,
+      "grad_norm": 1.2926694782345516,
+      "learning_rate": 0.00019389658203751808,
+      "loss": 13.9471,
+      "step": 2375
+    },
+    {
+      "epoch": 0.41547540983606557,
+      "grad_norm": 1.2962905413021235,
+      "learning_rate": 0.00019389008513765677,
+      "loss": 14.2121,
+      "step": 2376
+    },
+    {
+      "epoch": 0.41565027322404374,
+      "grad_norm": 1.4287792939621797,
+      "learning_rate": 0.0001938835848907236,
+      "loss": 14.0407,
+      "step": 2377
+    },
+    {
+      "epoch": 0.41582513661202186,
+      "grad_norm": 1.4968346032146598,
+      "learning_rate": 0.00019387708129695033,
+      "loss": 14.4453,
+      "step": 2378
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 1.24626962188957,
+      "learning_rate": 0.0001938705743565688,
+      "loss": 14.3107,
+      "step": 2379
+    },
+    {
+      "epoch": 0.41617486338797816,
+      "grad_norm": 1.9398171785365503,
+      "learning_rate": 0.00019386406406981098,
+      "loss": 14.4012,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4163497267759563,
+      "grad_norm": 1.9892773639002697,
+      "learning_rate": 0.0001938575504369089,
+      "loss": 14.2969,
+      "step": 2381
+    },
+    {
+      "epoch": 0.41652459016393445,
+      "grad_norm": 1.1586067891040257,
+      "learning_rate": 0.00019385103345809484,
+      "loss": 14.3739,
+      "step": 2382
+    },
+    {
+      "epoch": 0.41669945355191257,
+      "grad_norm": 1.5124110221946374,
+      "learning_rate": 0.00019384451313360108,
+      "loss": 14.3004,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4168743169398907,
+      "grad_norm": 1.254968582308649,
+      "learning_rate": 0.00019383798946366008,
+      "loss": 14.1591,
+      "step": 2384
+    },
+    {
+      "epoch": 0.41704918032786886,
+      "grad_norm": 1.300196053794439,
+      "learning_rate": 0.0001938314624485044,
+      "loss": 14.1826,
+      "step": 2385
+    },
+    {
+      "epoch": 0.417224043715847,
+      "grad_norm": 1.2758566234039694,
+      "learning_rate": 0.00019382493208836671,
+      "loss": 14.0797,
+      "step": 2386
+    },
+    {
+      "epoch": 0.41739890710382516,
+      "grad_norm": 1.386775034032542,
+      "learning_rate": 0.0001938183983834798,
+      "loss": 14.2569,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4175737704918033,
+      "grad_norm": 1.3391396055302798,
+      "learning_rate": 0.00019381186133407663,
+      "loss": 14.2349,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4177486338797814,
+      "grad_norm": 1.4542118668347324,
+      "learning_rate": 0.0001938053209403902,
+      "loss": 14.3073,
+      "step": 2389
+    },
+    {
+      "epoch": 0.41792349726775957,
+      "grad_norm": 1.2045691882822098,
+      "learning_rate": 0.00019379877720265368,
+      "loss": 14.3214,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4180983606557377,
+      "grad_norm": 1.5555434892279323,
+      "learning_rate": 0.00019379223012110032,
+      "loss": 14.0971,
+      "step": 2391
+    },
+    {
+      "epoch": 0.41827322404371586,
+      "grad_norm": 1.309543520565377,
+      "learning_rate": 0.00019378567969596355,
+      "loss": 13.9392,
+      "step": 2392
+    },
+    {
+      "epoch": 0.418448087431694,
+      "grad_norm": 1.3817520406256774,
+      "learning_rate": 0.00019377912592747688,
+      "loss": 14.2074,
+      "step": 2393
+    },
+    {
+      "epoch": 0.41862295081967216,
+      "grad_norm": 1.3145428124527276,
+      "learning_rate": 0.00019377256881587395,
+      "loss": 14.2081,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4187978142076503,
+      "grad_norm": 1.2546628637109183,
+      "learning_rate": 0.0001937660083613885,
+      "loss": 14.1698,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4189726775956284,
+      "grad_norm": 1.4696611217661188,
+      "learning_rate": 0.00019375944456425441,
+      "loss": 14.0125,
+      "step": 2396
+    },
+    {
+      "epoch": 0.41914754098360657,
+      "grad_norm": 1.287598182984324,
+      "learning_rate": 0.00019375287742470564,
+      "loss": 13.9755,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4193224043715847,
+      "grad_norm": 1.4129440310064345,
+      "learning_rate": 0.00019374630694297636,
+      "loss": 14.1524,
+      "step": 2398
+    },
+    {
+      "epoch": 0.41949726775956286,
+      "grad_norm": 1.1840151639405265,
+      "learning_rate": 0.00019373973311930076,
+      "loss": 14.2268,
+      "step": 2399
+    },
+    {
+      "epoch": 0.419672131147541,
+      "grad_norm": 1.310684784493499,
+      "learning_rate": 0.0001937331559539132,
+      "loss": 14.201,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4198469945355191,
+      "grad_norm": 1.2907789995565884,
+      "learning_rate": 0.00019372657544704813,
+      "loss": 14.1758,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4200218579234973,
+      "grad_norm": 1.2756430119707747,
+      "learning_rate": 0.00019371999159894013,
+      "loss": 14.3245,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4201967213114754,
+      "grad_norm": 1.2756582340885156,
+      "learning_rate": 0.000193713404409824,
+      "loss": 14.2089,
+      "step": 2403
+    },
+    {
+      "epoch": 0.42037158469945357,
+      "grad_norm": 1.33456032855704,
+      "learning_rate": 0.00019370681387993445,
+      "loss": 14.3623,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4205464480874317,
+      "grad_norm": 1.2982480679867137,
+      "learning_rate": 0.00019370022000950648,
+      "loss": 14.2107,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4207213114754098,
+      "grad_norm": 1.3878402470690299,
+      "learning_rate": 0.00019369362279877513,
+      "loss": 14.3315,
+      "step": 2406
+    },
+    {
+      "epoch": 0.420896174863388,
+      "grad_norm": 1.5351013033264354,
+      "learning_rate": 0.0001936870222479756,
+      "loss": 14.3496,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4210710382513661,
+      "grad_norm": 1.387910482127796,
+      "learning_rate": 0.0001936804183573432,
+      "loss": 14.328,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4212459016393443,
+      "grad_norm": 1.5644932666143296,
+      "learning_rate": 0.00019367381112711334,
+      "loss": 14.3994,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4214207650273224,
+      "grad_norm": 1.518063233550762,
+      "learning_rate": 0.00019366720055752156,
+      "loss": 14.3336,
+      "step": 2410
+    },
+    {
+      "epoch": 0.42159562841530057,
+      "grad_norm": 1.362155055146043,
+      "learning_rate": 0.00019366058664880352,
+      "loss": 14.1753,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4217704918032787,
+      "grad_norm": 1.4456810553156647,
+      "learning_rate": 0.000193653969401195,
+      "loss": 13.9485,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4219453551912568,
+      "grad_norm": 1.3994478782383022,
+      "learning_rate": 0.00019364734881493188,
+      "loss": 14.3986,
+      "step": 2413
+    },
+    {
+      "epoch": 0.422120218579235,
+      "grad_norm": 1.351135515499546,
+      "learning_rate": 0.00019364072489025017,
+      "loss": 14.3073,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4222950819672131,
+      "grad_norm": 1.314771861586448,
+      "learning_rate": 0.00019363409762738606,
+      "loss": 14.1764,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4224699453551913,
+      "grad_norm": 1.2995199500111634,
+      "learning_rate": 0.00019362746702657576,
+      "loss": 14.0207,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4226448087431694,
+      "grad_norm": 1.3299452926228617,
+      "learning_rate": 0.00019362083308805566,
+      "loss": 14.2146,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4228196721311475,
+      "grad_norm": 1.1901060888205708,
+      "learning_rate": 0.00019361419581206222,
+      "loss": 14.0829,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4229945355191257,
+      "grad_norm": 1.4131854193301923,
+      "learning_rate": 0.0001936075551988321,
+      "loss": 14.3244,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4231693989071038,
+      "grad_norm": 1.3230904581694383,
+      "learning_rate": 0.00019360091124860198,
+      "loss": 14.3515,
+      "step": 2420
+    },
+    {
+      "epoch": 0.423344262295082,
+      "grad_norm": 1.3244030868600896,
+      "learning_rate": 0.00019359426396160876,
+      "loss": 14.1674,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4235191256830601,
+      "grad_norm": 1.1779488233223439,
+      "learning_rate": 0.00019358761333808936,
+      "loss": 14.258,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4236939890710382,
+      "grad_norm": 1.3908601444223523,
+      "learning_rate": 0.00019358095937828092,
+      "loss": 14.1888,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4238688524590164,
+      "grad_norm": 1.1676333292270147,
+      "learning_rate": 0.0001935743020824206,
+      "loss": 14.2069,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4240437158469945,
+      "grad_norm": 1.2763018203867063,
+      "learning_rate": 0.00019356764145074572,
+      "loss": 14.192,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4242185792349727,
+      "grad_norm": 1.3686578566058412,
+      "learning_rate": 0.0001935609774834938,
+      "loss": 14.0725,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4243934426229508,
+      "grad_norm": 1.3721678207032761,
+      "learning_rate": 0.00019355431018090228,
+      "loss": 14.247,
+      "step": 2427
+    },
+    {
+      "epoch": 0.424568306010929,
+      "grad_norm": 1.28662285037685,
+      "learning_rate": 0.00019354763954320894,
+      "loss": 14.2103,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4247431693989071,
+      "grad_norm": 1.2637751669504733,
+      "learning_rate": 0.00019354096557065156,
+      "loss": 13.9343,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4249180327868852,
+      "grad_norm": 1.227571801975648,
+      "learning_rate": 0.00019353428826346803,
+      "loss": 14.3972,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4250928961748634,
+      "grad_norm": 1.169882683193998,
+      "learning_rate": 0.0001935276076218964,
+      "loss": 14.1064,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4252677595628415,
+      "grad_norm": 1.211875874531382,
+      "learning_rate": 0.00019352092364617487,
+      "loss": 13.7963,
+      "step": 2432
+    },
+    {
+      "epoch": 0.4254426229508197,
+      "grad_norm": 1.262021541914038,
+      "learning_rate": 0.00019351423633654164,
+      "loss": 14.2446,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4256174863387978,
+      "grad_norm": 1.2826991579368374,
+      "learning_rate": 0.00019350754569323516,
+      "loss": 14.3749,
+      "step": 2434
+    },
+    {
+      "epoch": 0.42579234972677593,
+      "grad_norm": 1.2126393185179347,
+      "learning_rate": 0.00019350085171649395,
+      "loss": 14.0305,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4259672131147541,
+      "grad_norm": 1.1558140799407817,
+      "learning_rate": 0.00019349415440655658,
+      "loss": 13.973,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4261420765027322,
+      "grad_norm": 1.331285190506253,
+      "learning_rate": 0.00019348745376366187,
+      "loss": 13.7912,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4263169398907104,
+      "grad_norm": 1.2479421657728167,
+      "learning_rate": 0.00019348074978804865,
+      "loss": 14.128,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4264918032786885,
+      "grad_norm": 1.2309824852425153,
+      "learning_rate": 0.00019347404247995592,
+      "loss": 14.3792,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 1.2541550775427777,
+      "learning_rate": 0.0001934673318396228,
+      "loss": 14.2594,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4268415300546448,
+      "grad_norm": 1.277517746650007,
+      "learning_rate": 0.0001934606178672885,
+      "loss": 14.1558,
+      "step": 2441
+    },
+    {
+      "epoch": 0.42701639344262293,
+      "grad_norm": 1.2526995212370582,
+      "learning_rate": 0.00019345390056319236,
+      "loss": 14.2794,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4271912568306011,
+      "grad_norm": 1.2355452352044025,
+      "learning_rate": 0.00019344717992757386,
+      "loss": 13.9259,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4273661202185792,
+      "grad_norm": 1.277921992983003,
+      "learning_rate": 0.00019344045596067258,
+      "loss": 14.3576,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4275409836065574,
+      "grad_norm": 1.2458556134799403,
+      "learning_rate": 0.0001934337286627282,
+      "loss": 14.4477,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4277158469945355,
+      "grad_norm": 1.4596693010087316,
+      "learning_rate": 0.0001934269980339806,
+      "loss": 14.2315,
+      "step": 2446
+    },
+    {
+      "epoch": 0.42789071038251364,
+      "grad_norm": 1.203609575502491,
+      "learning_rate": 0.00019342026407466963,
+      "loss": 14.1145,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4280655737704918,
+      "grad_norm": 1.2955506981360736,
+      "learning_rate": 0.00019341352678503543,
+      "loss": 13.9923,
+      "step": 2448
+    },
+    {
+      "epoch": 0.42824043715846993,
+      "grad_norm": 1.393424677777909,
+      "learning_rate": 0.00019340678616531815,
+      "loss": 14.306,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4284153005464481,
+      "grad_norm": 1.332256219903588,
+      "learning_rate": 0.00019340004221575803,
+      "loss": 14.0993,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4285901639344262,
+      "grad_norm": 1.352426607236982,
+      "learning_rate": 0.00019339329493659558,
+      "loss": 14.3099,
+      "step": 2451
+    },
+    {
+      "epoch": 0.42876502732240435,
+      "grad_norm": 1.351667079160558,
+      "learning_rate": 0.00019338654432807126,
+      "loss": 14.2316,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4289398907103825,
+      "grad_norm": 1.1963293147134983,
+      "learning_rate": 0.00019337979039042572,
+      "loss": 14.0589,
+      "step": 2453
+    },
+    {
+      "epoch": 0.42911475409836064,
+      "grad_norm": 1.3536806279731457,
+      "learning_rate": 0.00019337303312389978,
+      "loss": 14.0496,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4292896174863388,
+      "grad_norm": 1.5938034261762926,
+      "learning_rate": 0.00019336627252873433,
+      "loss": 14.1861,
+      "step": 2455
+    },
+    {
+      "epoch": 0.42946448087431693,
+      "grad_norm": 1.251440116615712,
+      "learning_rate": 0.0001933595086051703,
+      "loss": 14.4585,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4296393442622951,
+      "grad_norm": 1.7090282988811927,
+      "learning_rate": 0.00019335274135344891,
+      "loss": 14.2176,
+      "step": 2457
+    },
+    {
+      "epoch": 0.42981420765027323,
+      "grad_norm": 1.1361504257162174,
+      "learning_rate": 0.00019334597077381137,
+      "loss": 14.0133,
+      "step": 2458
+    },
+    {
+      "epoch": 0.42998907103825135,
+      "grad_norm": 1.4092927500799095,
+      "learning_rate": 0.000193339196866499,
+      "loss": 14.3701,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4301639344262295,
+      "grad_norm": 1.1890071517467573,
+      "learning_rate": 0.00019333241963175333,
+      "loss": 13.9367,
+      "step": 2460
+    },
+    {
+      "epoch": 0.43033879781420764,
+      "grad_norm": 1.2580933059408514,
+      "learning_rate": 0.00019332563906981594,
+      "loss": 14.171,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4305136612021858,
+      "grad_norm": 1.2734201754637515,
+      "learning_rate": 0.00019331885518092856,
+      "loss": 14.4276,
+      "step": 2462
+    },
+    {
+      "epoch": 0.43068852459016393,
+      "grad_norm": 1.1914571078688596,
+      "learning_rate": 0.00019331206796533302,
+      "loss": 13.9551,
+      "step": 2463
+    },
+    {
+      "epoch": 0.43086338797814205,
+      "grad_norm": 1.659314918298205,
+      "learning_rate": 0.0001933052774232713,
+      "loss": 14.3911,
+      "step": 2464
+    },
+    {
+      "epoch": 0.43103825136612023,
+      "grad_norm": 1.3604051104638515,
+      "learning_rate": 0.00019329848355498542,
+      "loss": 13.9744,
+      "step": 2465
+    },
+    {
+      "epoch": 0.43121311475409835,
+      "grad_norm": 1.4064677928905436,
+      "learning_rate": 0.00019329168636071767,
+      "loss": 14.1554,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4313879781420765,
+      "grad_norm": 1.221431059526863,
+      "learning_rate": 0.00019328488584071022,
+      "loss": 13.8779,
+      "step": 2467
+    },
+    {
+      "epoch": 0.43156284153005464,
+      "grad_norm": 1.3564270169163823,
+      "learning_rate": 0.00019327808199520562,
+      "loss": 14.1576,
+      "step": 2468
+    },
+    {
+      "epoch": 0.43173770491803276,
+      "grad_norm": 1.3930508967828443,
+      "learning_rate": 0.0001932712748244464,
+      "loss": 14.0781,
+      "step": 2469
+    },
+    {
+      "epoch": 0.43191256830601094,
+      "grad_norm": 1.2233227601627403,
+      "learning_rate": 0.0001932644643286752,
+      "loss": 14.2066,
+      "step": 2470
+    },
+    {
+      "epoch": 0.43208743169398905,
+      "grad_norm": 1.3679119651930605,
+      "learning_rate": 0.0001932576505081348,
+      "loss": 14.3512,
+      "step": 2471
+    },
+    {
+      "epoch": 0.43226229508196723,
+      "grad_norm": 1.3145105752571724,
+      "learning_rate": 0.00019325083336306812,
+      "loss": 14.3779,
+      "step": 2472
+    },
+    {
+      "epoch": 0.43243715846994535,
+      "grad_norm": 1.2395083331745653,
+      "learning_rate": 0.0001932440128937182,
+      "loss": 13.7706,
+      "step": 2473
+    },
+    {
+      "epoch": 0.4326120218579235,
+      "grad_norm": 1.5628591049737275,
+      "learning_rate": 0.00019323718910032816,
+      "loss": 14.2119,
+      "step": 2474
+    },
+    {
+      "epoch": 0.43278688524590164,
+      "grad_norm": 1.4629716676644093,
+      "learning_rate": 0.00019323036198314126,
+      "loss": 14.0942,
+      "step": 2475
+    },
+    {
+      "epoch": 0.43296174863387976,
+      "grad_norm": 1.306096294180338,
+      "learning_rate": 0.00019322353154240088,
+      "loss": 14.1815,
+      "step": 2476
+    },
+    {
+      "epoch": 0.43313661202185794,
+      "grad_norm": 1.7084633375684057,
+      "learning_rate": 0.0001932166977783505,
+      "loss": 14.1255,
+      "step": 2477
+    },
+    {
+      "epoch": 0.43331147540983606,
+      "grad_norm": 1.2033405002302575,
+      "learning_rate": 0.0001932098606912338,
+      "loss": 14.2257,
+      "step": 2478
+    },
+    {
+      "epoch": 0.43348633879781423,
+      "grad_norm": 1.7257347185137974,
+      "learning_rate": 0.00019320302028129444,
+      "loss": 14.1348,
+      "step": 2479
+    },
+    {
+      "epoch": 0.43366120218579235,
+      "grad_norm": 1.5337144586259355,
+      "learning_rate": 0.0001931961765487763,
+      "loss": 14.2742,
+      "step": 2480
+    },
+    {
+      "epoch": 0.43383606557377047,
+      "grad_norm": 1.3473482673268058,
+      "learning_rate": 0.00019318932949392335,
+      "loss": 14.0655,
+      "step": 2481
+    },
+    {
+      "epoch": 0.43401092896174864,
+      "grad_norm": 1.5638449948337458,
+      "learning_rate": 0.0001931824791169797,
+      "loss": 14.0133,
+      "step": 2482
+    },
+    {
+      "epoch": 0.43418579234972676,
+      "grad_norm": 1.1122477118160141,
+      "learning_rate": 0.00019317562541818956,
+      "loss": 14.1741,
+      "step": 2483
+    },
+    {
+      "epoch": 0.43436065573770494,
+      "grad_norm": 1.5682998854446055,
+      "learning_rate": 0.00019316876839779723,
+      "loss": 14.1771,
+      "step": 2484
+    },
+    {
+      "epoch": 0.43453551912568306,
+      "grad_norm": 1.135251771318683,
+      "learning_rate": 0.00019316190805604712,
+      "loss": 13.9852,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4347103825136612,
+      "grad_norm": 1.3833338662220684,
+      "learning_rate": 0.00019315504439318387,
+      "loss": 13.9328,
+      "step": 2486
+    },
+    {
+      "epoch": 0.43488524590163935,
+      "grad_norm": 1.2045335954845187,
+      "learning_rate": 0.00019314817740945212,
+      "loss": 13.7993,
+      "step": 2487
+    },
+    {
+      "epoch": 0.43506010928961747,
+      "grad_norm": 1.3955105894820239,
+      "learning_rate": 0.0001931413071050967,
+      "loss": 13.8352,
+      "step": 2488
+    },
+    {
+      "epoch": 0.43523497267759564,
+      "grad_norm": 1.2721849973617156,
+      "learning_rate": 0.0001931344334803625,
+      "loss": 13.5996,
+      "step": 2489
+    },
+    {
+      "epoch": 0.43540983606557376,
+      "grad_norm": 1.6576857399323592,
+      "learning_rate": 0.00019312755653549455,
+      "loss": 14.2353,
+      "step": 2490
+    },
+    {
+      "epoch": 0.43558469945355194,
+      "grad_norm": 1.4095886503197859,
+      "learning_rate": 0.000193120676270738,
+      "loss": 13.9959,
+      "step": 2491
+    },
+    {
+      "epoch": 0.43575956284153006,
+      "grad_norm": 1.318136786726855,
+      "learning_rate": 0.00019311379268633817,
+      "loss": 14.2314,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4359344262295082,
+      "grad_norm": 1.5244379081581403,
+      "learning_rate": 0.0001931069057825404,
+      "loss": 14.0499,
+      "step": 2493
+    },
+    {
+      "epoch": 0.43610928961748635,
+      "grad_norm": 1.5005257171879849,
+      "learning_rate": 0.00019310001555959024,
+      "loss": 14.3261,
+      "step": 2494
+    },
+    {
+      "epoch": 0.43628415300546447,
+      "grad_norm": 1.432373739301482,
+      "learning_rate": 0.00019309312201773327,
+      "loss": 14.1276,
+      "step": 2495
+    },
+    {
+      "epoch": 0.43645901639344264,
+      "grad_norm": 1.415999135995586,
+      "learning_rate": 0.0001930862251572153,
+      "loss": 14.2594,
+      "step": 2496
+    },
+    {
+      "epoch": 0.43663387978142076,
+      "grad_norm": 1.3527926482535146,
+      "learning_rate": 0.00019307932497828216,
+      "loss": 13.8874,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4368087431693989,
+      "grad_norm": 1.4209416598657396,
+      "learning_rate": 0.00019307242148117984,
+      "loss": 14.1223,
+      "step": 2498
+    },
+    {
+      "epoch": 0.43698360655737706,
+      "grad_norm": 1.5223539813108815,
+      "learning_rate": 0.00019306551466615442,
+      "loss": 14.0894,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4371584699453552,
+      "grad_norm": 1.353024373761381,
+      "learning_rate": 0.00019305860453345215,
+      "loss": 14.3112,
+      "step": 2500
+    },
+    {
+      "epoch": 0.43733333333333335,
+      "grad_norm": 1.4730693202684368,
+      "learning_rate": 0.00019305169108331933,
+      "loss": 14.1158,
+      "step": 2501
+    },
+    {
+      "epoch": 0.43750819672131147,
+      "grad_norm": 1.4067558381939798,
+      "learning_rate": 0.00019304477431600247,
+      "loss": 13.9894,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4376830601092896,
+      "grad_norm": 1.1870081211098806,
+      "learning_rate": 0.0001930378542317481,
+      "loss": 13.9917,
+      "step": 2503
+    },
+    {
+      "epoch": 0.43785792349726776,
+      "grad_norm": 1.2020075683239564,
+      "learning_rate": 0.00019303093083080294,
+      "loss": 14.1565,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4380327868852459,
+      "grad_norm": 1.3920223558675995,
+      "learning_rate": 0.0001930240041134138,
+      "loss": 14.0683,
+      "step": 2505
+    },
+    {
+      "epoch": 0.43820765027322406,
+      "grad_norm": 1.2655062902331353,
+      "learning_rate": 0.00019301707407982758,
+      "loss": 14.0951,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4383825136612022,
+      "grad_norm": 1.2209362441279037,
+      "learning_rate": 0.00019301014073029138,
+      "loss": 13.8443,
+      "step": 2507
+    },
+    {
+      "epoch": 0.43855737704918035,
+      "grad_norm": 1.295336417793617,
+      "learning_rate": 0.00019300320406505229,
+      "loss": 14.2294,
+      "step": 2508
+    },
+    {
+      "epoch": 0.43873224043715847,
+      "grad_norm": 1.33877587670741,
+      "learning_rate": 0.00019299626408435767,
+      "loss": 14.2494,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4389071038251366,
+      "grad_norm": 1.158776587786046,
+      "learning_rate": 0.00019298932078845487,
+      "loss": 13.7154,
+      "step": 2510
+    },
+    {
+      "epoch": 0.43908196721311477,
+      "grad_norm": 1.420901877018778,
+      "learning_rate": 0.00019298237417759143,
+      "loss": 14.048,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4392568306010929,
+      "grad_norm": 1.2997428832283504,
+      "learning_rate": 0.00019297542425201502,
+      "loss": 14.1237,
+      "step": 2512
+    },
+    {
+      "epoch": 0.43943169398907106,
+      "grad_norm": 1.1389697781964987,
+      "learning_rate": 0.00019296847101197333,
+      "loss": 13.9533,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4396065573770492,
+      "grad_norm": 1.277106640561947,
+      "learning_rate": 0.0001929615144577143,
+      "loss": 13.8521,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4397814207650273,
+      "grad_norm": 1.1584969538684822,
+      "learning_rate": 0.0001929545545894859,
+      "loss": 14.0941,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4399562841530055,
+      "grad_norm": 1.2497966107817347,
+      "learning_rate": 0.00019294759140753618,
+      "loss": 14.0604,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4401311475409836,
+      "grad_norm": 1.4055349945575168,
+      "learning_rate": 0.00019294062491211346,
+      "loss": 14.1163,
+      "step": 2517
+    },
+    {
+      "epoch": 0.44030601092896177,
+      "grad_norm": 1.167205688702947,
+      "learning_rate": 0.00019293365510346603,
+      "loss": 13.9511,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4404808743169399,
+      "grad_norm": 1.2363212159792358,
+      "learning_rate": 0.0001929266819818424,
+      "loss": 14.1856,
+      "step": 2519
+    },
+    {
+      "epoch": 0.44065573770491806,
+      "grad_norm": 1.1435346996245321,
+      "learning_rate": 0.00019291970554749112,
+      "loss": 14.0089,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4408306010928962,
+      "grad_norm": 1.293514492068434,
+      "learning_rate": 0.00019291272580066094,
+      "loss": 14.07,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4410054644808743,
+      "grad_norm": 1.2148960962111284,
+      "learning_rate": 0.0001929057427416006,
+      "loss": 14.1427,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4411803278688525,
+      "grad_norm": 1.300597117456122,
+      "learning_rate": 0.00019289875637055906,
+      "loss": 14.1132,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4413551912568306,
+      "grad_norm": 1.3223468380922765,
+      "learning_rate": 0.00019289176668778544,
+      "loss": 14.2546,
+      "step": 2524
+    },
+    {
+      "epoch": 0.44153005464480877,
+      "grad_norm": 1.1963247827608814,
+      "learning_rate": 0.00019288477369352886,
+      "loss": 14.022,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4417049180327869,
+      "grad_norm": 1.3406863489191634,
+      "learning_rate": 0.00019287777738803862,
+      "loss": 14.2264,
+      "step": 2526
+    },
+    {
+      "epoch": 0.441879781420765,
+      "grad_norm": 1.1514808168944384,
+      "learning_rate": 0.00019287077777156413,
+      "loss": 13.8425,
+      "step": 2527
+    },
+    {
+      "epoch": 0.4420546448087432,
+      "grad_norm": 1.3222788898300526,
+      "learning_rate": 0.00019286377484435492,
+      "loss": 14.1827,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4422295081967213,
+      "grad_norm": 1.197087055014921,
+      "learning_rate": 0.00019285676860666064,
+      "loss": 13.9644,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4424043715846995,
+      "grad_norm": 1.309534389473615,
+      "learning_rate": 0.00019284975905873103,
+      "loss": 14.3249,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4425792349726776,
+      "grad_norm": 1.4376786132618704,
+      "learning_rate": 0.00019284274620081602,
+      "loss": 14.1253,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4427540983606557,
+      "grad_norm": 1.2071354447051912,
+      "learning_rate": 0.00019283573003316557,
+      "loss": 13.9765,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4429289617486339,
+      "grad_norm": 1.3495000020800252,
+      "learning_rate": 0.0001928287105560298,
+      "loss": 14.1254,
+      "step": 2533
+    },
+    {
+      "epoch": 0.443103825136612,
+      "grad_norm": 1.3037500310726544,
+      "learning_rate": 0.00019282168776965896,
+      "loss": 14.1722,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4432786885245902,
+      "grad_norm": 1.2966022635105359,
+      "learning_rate": 0.00019281466167430343,
+      "loss": 14.1775,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4434535519125683,
+      "grad_norm": 1.276326218904945,
+      "learning_rate": 0.00019280763227021364,
+      "loss": 13.7482,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4436284153005465,
+      "grad_norm": 1.2625236556611712,
+      "learning_rate": 0.0001928005995576402,
+      "loss": 14.0711,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4438032786885246,
+      "grad_norm": 1.3937068523347034,
+      "learning_rate": 0.00019279356353683378,
+      "loss": 14.2607,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4439781420765027,
+      "grad_norm": 1.3391470579468492,
+      "learning_rate": 0.0001927865242080453,
+      "loss": 13.9463,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4441530054644809,
+      "grad_norm": 1.2574596689064192,
+      "learning_rate": 0.00019277948157152557,
+      "loss": 13.9948,
+      "step": 2540
+    },
+    {
+      "epoch": 0.444327868852459,
+      "grad_norm": 1.570244680996015,
+      "learning_rate": 0.00019277243562752574,
+      "loss": 13.8516,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4445027322404372,
+      "grad_norm": 1.2538938467715954,
+      "learning_rate": 0.000192765386376297,
+      "loss": 14.1271,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4446775956284153,
+      "grad_norm": 1.3981395989062133,
+      "learning_rate": 0.00019275833381809062,
+      "loss": 14.0002,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4448524590163934,
+      "grad_norm": 1.243429804335662,
+      "learning_rate": 0.000192751277953158,
+      "loss": 14.4403,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4450273224043716,
+      "grad_norm": 1.3352506592151978,
+      "learning_rate": 0.00019274421878175067,
+      "loss": 14.1207,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4452021857923497,
+      "grad_norm": 1.3369422894139582,
+      "learning_rate": 0.00019273715630412034,
+      "loss": 14.1108,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4453770491803279,
+      "grad_norm": 1.3187849851972346,
+      "learning_rate": 0.0001927300905205187,
+      "loss": 14.0717,
+      "step": 2547
+    },
+    {
+      "epoch": 0.445551912568306,
+      "grad_norm": 1.3379809847864539,
+      "learning_rate": 0.00019272302143119772,
+      "loss": 14.3749,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4457267759562841,
+      "grad_norm": 1.1889914515889877,
+      "learning_rate": 0.00019271594903640931,
+      "loss": 13.7976,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4459016393442623,
+      "grad_norm": 1.2524395178646768,
+      "learning_rate": 0.0001927088733364057,
+      "loss": 13.7846,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4460765027322404,
+      "grad_norm": 1.2597044345902513,
+      "learning_rate": 0.00019270179433143904,
+      "loss": 14.2405,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4462513661202186,
+      "grad_norm": 1.2878940327014066,
+      "learning_rate": 0.00019269471202176174,
+      "loss": 13.9575,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4464262295081967,
+      "grad_norm": 1.323597582393801,
+      "learning_rate": 0.00019268762640762624,
+      "loss": 14.1209,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4466010928961749,
+      "grad_norm": 1.1602985888737596,
+      "learning_rate": 0.00019268053748928515,
+      "loss": 14.0795,
+      "step": 2554
+    },
+    {
+      "epoch": 0.446775956284153,
+      "grad_norm": 1.1726442753827209,
+      "learning_rate": 0.0001926734452669912,
+      "loss": 13.9039,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4469508196721311,
+      "grad_norm": 1.2699529485188732,
+      "learning_rate": 0.0001926663497409972,
+      "loss": 14.036,
+      "step": 2556
+    },
+    {
+      "epoch": 0.4471256830601093,
+      "grad_norm": 1.3207625851556415,
+      "learning_rate": 0.0001926592509115561,
+      "loss": 14.3263,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4473005464480874,
+      "grad_norm": 1.4767018136891317,
+      "learning_rate": 0.00019265214877892098,
+      "loss": 14.1306,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4474754098360656,
+      "grad_norm": 1.557851631651253,
+      "learning_rate": 0.000192645043343345,
+      "loss": 14.2456,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4476502732240437,
+      "grad_norm": 1.3008508325574022,
+      "learning_rate": 0.00019263793460508146,
+      "loss": 14.3237,
+      "step": 2560
+    },
+    {
+      "epoch": 0.44782513661202183,
+      "grad_norm": 1.4513059557375978,
+      "learning_rate": 0.00019263082256438377,
+      "loss": 13.9974,
+      "step": 2561
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.2761385869821016,
+      "learning_rate": 0.0001926237072215055,
+      "loss": 14.2237,
+      "step": 2562
+    },
+    {
+      "epoch": 0.44817486338797813,
+      "grad_norm": 1.4220376073642027,
+      "learning_rate": 0.00019261658857670032,
+      "loss": 14.1019,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4483497267759563,
+      "grad_norm": 1.6021933093017457,
+      "learning_rate": 0.00019260946663022193,
+      "loss": 14.078,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4485245901639344,
+      "grad_norm": 1.1530313682544089,
+      "learning_rate": 0.00019260234138232423,
+      "loss": 14.1927,
+      "step": 2565
+    },
+    {
+      "epoch": 0.44869945355191254,
+      "grad_norm": 1.6965587286924255,
+      "learning_rate": 0.00019259521283326132,
+      "loss": 14.1986,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4488743169398907,
+      "grad_norm": 1.2907830089689476,
+      "learning_rate": 0.00019258808098328724,
+      "loss": 14.2033,
+      "step": 2567
+    },
+    {
+      "epoch": 0.44904918032786884,
+      "grad_norm": 1.7413612440343211,
+      "learning_rate": 0.00019258094583265623,
+      "loss": 14.0868,
+      "step": 2568
+    },
+    {
+      "epoch": 0.449224043715847,
+      "grad_norm": 1.361561846083516,
+      "learning_rate": 0.00019257380738162266,
+      "loss": 14.1609,
+      "step": 2569
+    },
+    {
+      "epoch": 0.44939890710382513,
+      "grad_norm": 1.6948326159045823,
+      "learning_rate": 0.00019256666563044104,
+      "loss": 14.2691,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4495737704918033,
+      "grad_norm": 1.4387344988666062,
+      "learning_rate": 0.00019255952057936596,
+      "loss": 14.068,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4497486338797814,
+      "grad_norm": 1.4619606689002984,
+      "learning_rate": 0.00019255237222865206,
+      "loss": 14.0456,
+      "step": 2572
+    },
+    {
+      "epoch": 0.44992349726775954,
+      "grad_norm": 1.5494392017546808,
+      "learning_rate": 0.00019254522057855428,
+      "loss": 14.3057,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4500983606557377,
+      "grad_norm": 1.3453162842164461,
+      "learning_rate": 0.00019253806562932749,
+      "loss": 14.1334,
+      "step": 2574
+    },
+    {
+      "epoch": 0.45027322404371584,
+      "grad_norm": 1.5958846081731701,
+      "learning_rate": 0.0001925309073812268,
+      "loss": 13.6958,
+      "step": 2575
+    },
+    {
+      "epoch": 0.450448087431694,
+      "grad_norm": 1.3595520854561733,
+      "learning_rate": 0.00019252374583450733,
+      "loss": 14.2782,
+      "step": 2576
+    },
+    {
+      "epoch": 0.45062295081967213,
+      "grad_norm": 1.4211899062316293,
+      "learning_rate": 0.00019251658098942446,
+      "loss": 14.1955,
+      "step": 2577
+    },
+    {
+      "epoch": 0.45079781420765025,
+      "grad_norm": 1.2388439997495284,
+      "learning_rate": 0.00019250941284623357,
+      "loss": 13.9422,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4509726775956284,
+      "grad_norm": 1.347636276629752,
+      "learning_rate": 0.00019250224140519016,
+      "loss": 14.2894,
+      "step": 2579
+    },
+    {
+      "epoch": 0.45114754098360654,
+      "grad_norm": 1.1483251453267282,
+      "learning_rate": 0.00019249506666654999,
+      "loss": 14.1667,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4513224043715847,
+      "grad_norm": 1.2971223852513096,
+      "learning_rate": 0.00019248788863056868,
+      "loss": 13.939,
+      "step": 2581
+    },
+    {
+      "epoch": 0.45149726775956284,
+      "grad_norm": 1.3151563791615264,
+      "learning_rate": 0.00019248070729750225,
+      "loss": 14.0353,
+      "step": 2582
+    },
+    {
+      "epoch": 0.45167213114754096,
+      "grad_norm": 1.268436481912017,
+      "learning_rate": 0.00019247352266760665,
+      "loss": 14.0486,
+      "step": 2583
+    },
+    {
+      "epoch": 0.45184699453551913,
+      "grad_norm": 1.226119658694897,
+      "learning_rate": 0.000192466334741138,
+      "loss": 14.1641,
+      "step": 2584
+    },
+    {
+      "epoch": 0.45202185792349725,
+      "grad_norm": 1.1542559026216468,
+      "learning_rate": 0.00019245914351835258,
+      "loss": 13.8696,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4521967213114754,
+      "grad_norm": 1.1720452646404815,
+      "learning_rate": 0.0001924519489995067,
+      "loss": 14.0179,
+      "step": 2586
+    },
+    {
+      "epoch": 0.45237158469945354,
+      "grad_norm": 1.24882842596847,
+      "learning_rate": 0.00019244475118485686,
+      "loss": 14.0831,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4525464480874317,
+      "grad_norm": 1.4381568113942924,
+      "learning_rate": 0.00019243755007465965,
+      "loss": 14.1636,
+      "step": 2588
+    },
+    {
+      "epoch": 0.45272131147540984,
+      "grad_norm": 1.2953791750569945,
+      "learning_rate": 0.00019243034566917178,
+      "loss": 14.133,
+      "step": 2589
+    },
+    {
+      "epoch": 0.45289617486338796,
+      "grad_norm": 1.4543042674554227,
+      "learning_rate": 0.0001924231379686501,
+      "loss": 14.2368,
+      "step": 2590
+    },
+    {
+      "epoch": 0.45307103825136613,
+      "grad_norm": 1.2147846358299028,
+      "learning_rate": 0.0001924159269733515,
+      "loss": 13.941,
+      "step": 2591
+    },
+    {
+      "epoch": 0.45324590163934425,
+      "grad_norm": 1.5636284962761928,
+      "learning_rate": 0.0001924087126835331,
+      "loss": 14.1185,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4534207650273224,
+      "grad_norm": 1.2714404318610737,
+      "learning_rate": 0.00019240149509945205,
+      "loss": 14.1778,
+      "step": 2593
+    },
+    {
+      "epoch": 0.45359562841530054,
+      "grad_norm": 1.8513778098120897,
+      "learning_rate": 0.00019239427422136567,
+      "loss": 14.2023,
+      "step": 2594
+    },
+    {
+      "epoch": 0.45377049180327866,
+      "grad_norm": 1.243513845395729,
+      "learning_rate": 0.00019238705004953136,
+      "loss": 13.8157,
+      "step": 2595
+    },
+    {
+      "epoch": 0.45394535519125684,
+      "grad_norm": 2.1452410489360667,
+      "learning_rate": 0.00019237982258420665,
+      "loss": 13.692,
+      "step": 2596
+    },
+    {
+      "epoch": 0.45412021857923496,
+      "grad_norm": 1.2699165237362773,
+      "learning_rate": 0.0001923725918256492,
+      "loss": 14.1213,
+      "step": 2597
+    },
+    {
+      "epoch": 0.45429508196721313,
+      "grad_norm": 1.5805127360153228,
+      "learning_rate": 0.00019236535777411682,
+      "loss": 14.0852,
+      "step": 2598
+    },
+    {
+      "epoch": 0.45446994535519125,
+      "grad_norm": 1.516022274929188,
+      "learning_rate": 0.0001923581204298673,
+      "loss": 13.9528,
+      "step": 2599
+    },
+    {
+      "epoch": 0.4546448087431694,
+      "grad_norm": 1.2098776056679739,
+      "learning_rate": 0.00019235087979315872,
+      "loss": 14.0148,
+      "step": 2600
+    },
+    {
+      "epoch": 0.45481967213114755,
+      "grad_norm": 1.4442507780753517,
+      "learning_rate": 0.00019234363586424918,
+      "loss": 14.0686,
+      "step": 2601
+    },
+    {
+      "epoch": 0.45499453551912566,
+      "grad_norm": 1.3589488807167547,
+      "learning_rate": 0.00019233638864339688,
+      "loss": 13.9665,
+      "step": 2602
+    },
+    {
+      "epoch": 0.45516939890710384,
+      "grad_norm": 1.377739248111322,
+      "learning_rate": 0.00019232913813086025,
+      "loss": 14.0487,
+      "step": 2603
+    },
+    {
+      "epoch": 0.45534426229508196,
+      "grad_norm": 1.3571460718215753,
+      "learning_rate": 0.0001923218843268977,
+      "loss": 14.1586,
+      "step": 2604
+    },
+    {
+      "epoch": 0.45551912568306013,
+      "grad_norm": 1.3881830551766883,
+      "learning_rate": 0.00019231462723176784,
+      "loss": 13.9582,
+      "step": 2605
+    },
+    {
+      "epoch": 0.45569398907103825,
+      "grad_norm": 1.2191016983215417,
+      "learning_rate": 0.0001923073668457294,
+      "loss": 13.7767,
+      "step": 2606
+    },
+    {
+      "epoch": 0.45586885245901637,
+      "grad_norm": 1.361556157622327,
+      "learning_rate": 0.00019230010316904114,
+      "loss": 13.9911,
+      "step": 2607
+    },
+    {
+      "epoch": 0.45604371584699455,
+      "grad_norm": 1.531918184369747,
+      "learning_rate": 0.00019229283620196208,
+      "loss": 14.0963,
+      "step": 2608
+    },
+    {
+      "epoch": 0.45621857923497267,
+      "grad_norm": 1.2224725496252957,
+      "learning_rate": 0.00019228556594475123,
+      "loss": 13.6751,
+      "step": 2609
+    },
+    {
+      "epoch": 0.45639344262295084,
+      "grad_norm": 1.3627788523393793,
+      "learning_rate": 0.00019227829239766778,
+      "loss": 13.9522,
+      "step": 2610
+    },
+    {
+      "epoch": 0.45656830601092896,
+      "grad_norm": 1.3282225710715738,
+      "learning_rate": 0.000192271015560971,
+      "loss": 14.2712,
+      "step": 2611
+    },
+    {
+      "epoch": 0.4567431693989071,
+      "grad_norm": 1.194841387921992,
+      "learning_rate": 0.00019226373543492035,
+      "loss": 14.1409,
+      "step": 2612
+    },
+    {
+      "epoch": 0.45691803278688525,
+      "grad_norm": 1.253810285037887,
+      "learning_rate": 0.00019225645201977534,
+      "loss": 14.0073,
+      "step": 2613
+    },
+    {
+      "epoch": 0.4570928961748634,
+      "grad_norm": 1.3169002838997383,
+      "learning_rate": 0.00019224916531579558,
+      "loss": 14.056,
+      "step": 2614
+    },
+    {
+      "epoch": 0.45726775956284155,
+      "grad_norm": 1.211087111908163,
+      "learning_rate": 0.00019224187532324084,
+      "loss": 13.9067,
+      "step": 2615
+    },
+    {
+      "epoch": 0.45744262295081967,
+      "grad_norm": 1.3512386846665538,
+      "learning_rate": 0.00019223458204237104,
+      "loss": 13.9468,
+      "step": 2616
+    },
+    {
+      "epoch": 0.45761748633879784,
+      "grad_norm": 1.4447813172162298,
+      "learning_rate": 0.00019222728547344615,
+      "loss": 13.9502,
+      "step": 2617
+    },
+    {
+      "epoch": 0.45779234972677596,
+      "grad_norm": 1.2656211497946623,
+      "learning_rate": 0.00019221998561672632,
+      "loss": 13.8367,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4579672131147541,
+      "grad_norm": 1.2627738682996006,
+      "learning_rate": 0.0001922126824724717,
+      "loss": 13.983,
+      "step": 2619
+    },
+    {
+      "epoch": 0.45814207650273225,
+      "grad_norm": 1.395034668401733,
+      "learning_rate": 0.00019220537604094272,
+      "loss": 14.3124,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4583169398907104,
+      "grad_norm": 1.2675074124157233,
+      "learning_rate": 0.0001921980663223998,
+      "loss": 14.0551,
+      "step": 2621
+    },
+    {
+      "epoch": 0.45849180327868855,
+      "grad_norm": 1.313598101389853,
+      "learning_rate": 0.00019219075331710356,
+      "loss": 14.1915,
+      "step": 2622
+    },
+    {
+      "epoch": 0.45866666666666667,
+      "grad_norm": 1.273270919675609,
+      "learning_rate": 0.00019218343702531466,
+      "loss": 14.2345,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4588415300546448,
+      "grad_norm": 1.3119484730990056,
+      "learning_rate": 0.00019217611744729393,
+      "loss": 14.0518,
+      "step": 2624
+    },
+    {
+      "epoch": 0.45901639344262296,
+      "grad_norm": 1.187885980793901,
+      "learning_rate": 0.00019216879458330236,
+      "loss": 13.8297,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4591912568306011,
+      "grad_norm": 1.3280837800364131,
+      "learning_rate": 0.00019216146843360087,
+      "loss": 14.0802,
+      "step": 2626
+    },
+    {
+      "epoch": 0.45936612021857925,
+      "grad_norm": 1.1674421490428062,
+      "learning_rate": 0.0001921541389984508,
+      "loss": 13.8897,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4595409836065574,
+      "grad_norm": 1.3580952421764985,
+      "learning_rate": 0.00019214680627811326,
+      "loss": 14.083,
+      "step": 2628
+    },
+    {
+      "epoch": 0.4597158469945355,
+      "grad_norm": 1.1401070579436259,
+      "learning_rate": 0.00019213947027284977,
+      "loss": 13.693,
+      "step": 2629
+    },
+    {
+      "epoch": 0.45989071038251367,
+      "grad_norm": 1.2768196385829969,
+      "learning_rate": 0.00019213213098292182,
+      "loss": 14.082,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4600655737704918,
+      "grad_norm": 1.2908383436905595,
+      "learning_rate": 0.00019212478840859105,
+      "loss": 14.0632,
+      "step": 2631
+    },
+    {
+      "epoch": 0.46024043715846996,
+      "grad_norm": 1.3697292785525246,
+      "learning_rate": 0.0001921174425501192,
+      "loss": 14.1453,
+      "step": 2632
+    },
+    {
+      "epoch": 0.4604153005464481,
+      "grad_norm": 1.5520830591361507,
+      "learning_rate": 0.00019211009340776815,
+      "loss": 13.9241,
+      "step": 2633
+    },
+    {
+      "epoch": 0.46059016393442626,
+      "grad_norm": 1.1766200701668905,
+      "learning_rate": 0.0001921027409817999,
+      "loss": 13.7422,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4607650273224044,
+      "grad_norm": 1.3106777252891708,
+      "learning_rate": 0.00019209538527247656,
+      "loss": 14.2029,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4609398907103825,
+      "grad_norm": 1.2626404812469778,
+      "learning_rate": 0.0001920880262800603,
+      "loss": 13.9867,
+      "step": 2636
+    },
+    {
+      "epoch": 0.46111475409836067,
+      "grad_norm": 1.2480582981099482,
+      "learning_rate": 0.0001920806640048135,
+      "loss": 14.0424,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4612896174863388,
+      "grad_norm": 1.2851233779029947,
+      "learning_rate": 0.00019207329844699863,
+      "loss": 13.8875,
+      "step": 2638
+    },
+    {
+      "epoch": 0.46146448087431696,
+      "grad_norm": 1.259520832235592,
+      "learning_rate": 0.00019206592960687826,
+      "loss": 14.1773,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4616393442622951,
+      "grad_norm": 1.229511396789006,
+      "learning_rate": 0.00019205855748471504,
+      "loss": 13.7825,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4618142076502732,
+      "grad_norm": 1.1205120583520851,
+      "learning_rate": 0.00019205118208077181,
+      "loss": 14.0233,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4619890710382514,
+      "grad_norm": 1.33595015046854,
+      "learning_rate": 0.0001920438033953115,
+      "loss": 14.0047,
+      "step": 2642
+    },
+    {
+      "epoch": 0.4621639344262295,
+      "grad_norm": 1.1511966812097552,
+      "learning_rate": 0.00019203642142859713,
+      "loss": 13.9771,
+      "step": 2643
+    },
+    {
+      "epoch": 0.46233879781420767,
+      "grad_norm": 1.4331855165588603,
+      "learning_rate": 0.00019202903618089186,
+      "loss": 13.9654,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4625136612021858,
+      "grad_norm": 1.1745755044987414,
+      "learning_rate": 0.000192021647652459,
+      "loss": 13.8691,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4626885245901639,
+      "grad_norm": 1.1874750140664765,
+      "learning_rate": 0.0001920142558435619,
+      "loss": 14.1575,
+      "step": 2646
+    },
+    {
+      "epoch": 0.4628633879781421,
+      "grad_norm": 1.3009105067080318,
+      "learning_rate": 0.0001920068607544641,
+      "loss": 14.0513,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4630382513661202,
+      "grad_norm": 1.2600999469083451,
+      "learning_rate": 0.0001919994623854292,
+      "loss": 14.0171,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4632131147540984,
+      "grad_norm": 1.261267138509606,
+      "learning_rate": 0.00019199206073672092,
+      "loss": 13.9696,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4633879781420765,
+      "grad_norm": 1.4072044712684115,
+      "learning_rate": 0.00019198465580860322,
+      "loss": 14.2258,
+      "step": 2650
+    },
+    {
+      "epoch": 0.46356284153005467,
+      "grad_norm": 1.3723193338758783,
+      "learning_rate": 0.00019197724760134,
+      "loss": 14.139,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4637377049180328,
+      "grad_norm": 1.3676485918175911,
+      "learning_rate": 0.00019196983611519534,
+      "loss": 13.8849,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4639125683060109,
+      "grad_norm": 1.241013314566385,
+      "learning_rate": 0.0001919624213504335,
+      "loss": 14.1285,
+      "step": 2653
+    },
+    {
+      "epoch": 0.4640874316939891,
+      "grad_norm": 1.3915269033945277,
+      "learning_rate": 0.00019195500330731878,
+      "loss": 13.8253,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4642622950819672,
+      "grad_norm": 1.4559794955432648,
+      "learning_rate": 0.00019194758198611561,
+      "loss": 14.3291,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4644371584699454,
+      "grad_norm": 1.3113455992389693,
+      "learning_rate": 0.0001919401573870886,
+      "loss": 13.9547,
+      "step": 2656
+    },
+    {
+      "epoch": 0.4646120218579235,
+      "grad_norm": 1.2430521159582375,
+      "learning_rate": 0.00019193272951050237,
+      "loss": 13.7188,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4647868852459016,
+      "grad_norm": 1.7023790794992888,
+      "learning_rate": 0.00019192529835662176,
+      "loss": 14.2681,
+      "step": 2658
+    },
+    {
+      "epoch": 0.4649617486338798,
+      "grad_norm": 1.4354554438873337,
+      "learning_rate": 0.00019191786392571168,
+      "loss": 14.1077,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4651366120218579,
+      "grad_norm": 1.3724077022975107,
+      "learning_rate": 0.00019191042621803715,
+      "loss": 13.9138,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4653114754098361,
+      "grad_norm": 1.4683637195129973,
+      "learning_rate": 0.00019190298523386328,
+      "loss": 13.9201,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4654863387978142,
+      "grad_norm": 1.244421674767809,
+      "learning_rate": 0.00019189554097345538,
+      "loss": 13.98,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4656612021857923,
+      "grad_norm": 1.458468754547796,
+      "learning_rate": 0.0001918880934370788,
+      "loss": 14.0113,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4658360655737705,
+      "grad_norm": 1.3305525632910704,
+      "learning_rate": 0.0001918806426249991,
+      "loss": 14.1026,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4660109289617486,
+      "grad_norm": 1.415444020065837,
+      "learning_rate": 0.00019187318853748178,
+      "loss": 13.6567,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4661857923497268,
+      "grad_norm": 1.382129034584835,
+      "learning_rate": 0.00019186573117479266,
+      "loss": 14.3409,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4663606557377049,
+      "grad_norm": 1.2425344505426767,
+      "learning_rate": 0.00019185827053719753,
+      "loss": 13.8481,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4665355191256831,
+      "grad_norm": 1.1575489299712298,
+      "learning_rate": 0.0001918508066249624,
+      "loss": 13.8528,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4667103825136612,
+      "grad_norm": 1.2719238326893942,
+      "learning_rate": 0.00019184333943835335,
+      "loss": 14.0125,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4668852459016393,
+      "grad_norm": 1.169765428094273,
+      "learning_rate": 0.00019183586897763655,
+      "loss": 13.802,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4670601092896175,
+      "grad_norm": 1.2160208074708592,
+      "learning_rate": 0.0001918283952430783,
+      "loss": 13.9944,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4672349726775956,
+      "grad_norm": 1.2336316749107217,
+      "learning_rate": 0.00019182091823494505,
+      "loss": 14.2652,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4674098360655738,
+      "grad_norm": 1.2002868485891862,
+      "learning_rate": 0.00019181343795350334,
+      "loss": 13.9336,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4675846994535519,
+      "grad_norm": 1.238877529563468,
+      "learning_rate": 0.00019180595439901988,
+      "loss": 14.2388,
+      "step": 2674
+    },
+    {
+      "epoch": 0.46775956284153003,
+      "grad_norm": 1.2895292033033507,
+      "learning_rate": 0.00019179846757176136,
+      "loss": 13.8458,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4679344262295082,
+      "grad_norm": 1.1682581218973975,
+      "learning_rate": 0.00019179097747199476,
+      "loss": 13.9653,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4681092896174863,
+      "grad_norm": 1.2738397099032068,
+      "learning_rate": 0.00019178348409998704,
+      "loss": 14.089,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4682841530054645,
+      "grad_norm": 1.1728637257095087,
+      "learning_rate": 0.00019177598745600535,
+      "loss": 13.905,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4684590163934426,
+      "grad_norm": 1.235284997022892,
+      "learning_rate": 0.00019176848754031697,
+      "loss": 13.9767,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4686338797814208,
+      "grad_norm": 1.1495192996786274,
+      "learning_rate": 0.00019176098435318918,
+      "loss": 13.8297,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4688087431693989,
+      "grad_norm": 1.1966984700195498,
+      "learning_rate": 0.00019175347789488953,
+      "loss": 13.6804,
+      "step": 2681
+    },
+    {
+      "epoch": 0.46898360655737703,
+      "grad_norm": 1.3369120718560146,
+      "learning_rate": 0.0001917459681656856,
+      "loss": 14.1404,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4691584699453552,
+      "grad_norm": 1.3626757981239366,
+      "learning_rate": 0.0001917384551658451,
+      "loss": 13.8776,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4693333333333333,
+      "grad_norm": 1.5945521473502848,
+      "learning_rate": 0.00019173093889563583,
+      "loss": 14.1004,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4695081967213115,
+      "grad_norm": 1.0533281851960279,
+      "learning_rate": 0.0001917234193553258,
+      "loss": 13.9314,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4696830601092896,
+      "grad_norm": 1.5898747013563934,
+      "learning_rate": 0.00019171589654518303,
+      "loss": 13.9274,
+      "step": 2686
+    },
+    {
+      "epoch": 0.46985792349726774,
+      "grad_norm": 1.3084802088055065,
+      "learning_rate": 0.00019170837046547572,
+      "loss": 14.1889,
+      "step": 2687
+    },
+    {
+      "epoch": 0.4700327868852459,
+      "grad_norm": 1.3730504700912884,
+      "learning_rate": 0.0001917008411164721,
+      "loss": 14.0039,
+      "step": 2688
+    },
+    {
+      "epoch": 0.47020765027322403,
+      "grad_norm": 1.3435784995366167,
+      "learning_rate": 0.00019169330849844068,
+      "loss": 13.8893,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4703825136612022,
+      "grad_norm": 1.3975697450297198,
+      "learning_rate": 0.00019168577261164993,
+      "loss": 14.0401,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4705573770491803,
+      "grad_norm": 1.4487287600140792,
+      "learning_rate": 0.00019167823345636853,
+      "loss": 14.0583,
+      "step": 2691
+    },
+    {
+      "epoch": 0.47073224043715844,
+      "grad_norm": 1.3115985627472073,
+      "learning_rate": 0.00019167069103286522,
+      "loss": 14.0201,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4709071038251366,
+      "grad_norm": 1.4771254685632997,
+      "learning_rate": 0.00019166314534140884,
+      "loss": 13.9102,
+      "step": 2693
+    },
+    {
+      "epoch": 0.47108196721311474,
+      "grad_norm": 1.2325844824644336,
+      "learning_rate": 0.00019165559638226846,
+      "loss": 14.2692,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4712568306010929,
+      "grad_norm": 1.261842907657071,
+      "learning_rate": 0.00019164804415571315,
+      "loss": 13.958,
+      "step": 2695
+    },
+    {
+      "epoch": 0.47143169398907103,
+      "grad_norm": 1.5515171074334488,
+      "learning_rate": 0.00019164048866201214,
+      "loss": 14.0444,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4716065573770492,
+      "grad_norm": 1.1955582018954507,
+      "learning_rate": 0.00019163292990143478,
+      "loss": 13.833,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4717814207650273,
+      "grad_norm": 1.2727671865980306,
+      "learning_rate": 0.00019162536787425052,
+      "loss": 13.9914,
+      "step": 2698
+    },
+    {
+      "epoch": 0.47195628415300545,
+      "grad_norm": 1.1653155998843978,
+      "learning_rate": 0.00019161780258072897,
+      "loss": 13.8875,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4721311475409836,
+      "grad_norm": 1.3179998178925538,
+      "learning_rate": 0.00019161023402113978,
+      "loss": 13.8759,
+      "step": 2700
+    },
+    {
+      "epoch": 0.47230601092896174,
+      "grad_norm": 1.1842880819261987,
+      "learning_rate": 0.0001916026621957528,
+      "loss": 13.1733,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4724808743169399,
+      "grad_norm": 1.4086788633105245,
+      "learning_rate": 0.00019159508710483793,
+      "loss": 14.0626,
+      "step": 2702
+    },
+    {
+      "epoch": 0.47265573770491803,
+      "grad_norm": 1.2158417961550776,
+      "learning_rate": 0.00019158750874866525,
+      "loss": 13.724,
+      "step": 2703
+    },
+    {
+      "epoch": 0.47283060109289615,
+      "grad_norm": 1.2901175603405857,
+      "learning_rate": 0.00019157992712750486,
+      "loss": 14.0993,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4730054644808743,
+      "grad_norm": 1.3672102734156437,
+      "learning_rate": 0.0001915723422416271,
+      "loss": 13.6986,
+      "step": 2705
+    },
+    {
+      "epoch": 0.47318032786885245,
+      "grad_norm": 1.2531218312888777,
+      "learning_rate": 0.00019156475409130228,
+      "loss": 14.0632,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4733551912568306,
+      "grad_norm": 1.3530691450071894,
+      "learning_rate": 0.00019155716267680098,
+      "loss": 13.7972,
+      "step": 2707
+    },
+    {
+      "epoch": 0.47353005464480874,
+      "grad_norm": 1.5385733383808797,
+      "learning_rate": 0.00019154956799839382,
+      "loss": 13.886,
+      "step": 2708
+    },
+    {
+      "epoch": 0.47370491803278686,
+      "grad_norm": 1.3783865552145131,
+      "learning_rate": 0.00019154197005635148,
+      "loss": 14.3079,
+      "step": 2709
+    },
+    {
+      "epoch": 0.47387978142076503,
+      "grad_norm": 1.4023213882106609,
+      "learning_rate": 0.00019153436885094492,
+      "loss": 14.0184,
+      "step": 2710
+    },
+    {
+      "epoch": 0.47405464480874315,
+      "grad_norm": 1.1659472134411617,
+      "learning_rate": 0.00019152676438244503,
+      "loss": 14.0527,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4742295081967213,
+      "grad_norm": 1.1869508721537623,
+      "learning_rate": 0.00019151915665112292,
+      "loss": 14.0068,
+      "step": 2712
+    },
+    {
+      "epoch": 0.47440437158469945,
+      "grad_norm": 1.231118322115901,
+      "learning_rate": 0.00019151154565724983,
+      "loss": 13.6552,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4745792349726776,
+      "grad_norm": 1.170656071882839,
+      "learning_rate": 0.00019150393140109705,
+      "loss": 14.0466,
+      "step": 2714
+    },
+    {
+      "epoch": 0.47475409836065574,
+      "grad_norm": 1.1460022604341458,
+      "learning_rate": 0.00019149631388293598,
+      "loss": 13.9927,
+      "step": 2715
+    },
+    {
+      "epoch": 0.47492896174863386,
+      "grad_norm": 1.152777036421214,
+      "learning_rate": 0.00019148869310303828,
+      "loss": 13.4854,
+      "step": 2716
+    },
+    {
+      "epoch": 0.47510382513661203,
+      "grad_norm": 1.1613888761730258,
+      "learning_rate": 0.00019148106906167553,
+      "loss": 14.077,
+      "step": 2717
+    },
+    {
+      "epoch": 0.47527868852459015,
+      "grad_norm": 1.1447378770363217,
+      "learning_rate": 0.00019147344175911957,
+      "loss": 14.0694,
+      "step": 2718
+    },
+    {
+      "epoch": 0.47545355191256833,
+      "grad_norm": 1.2515336992045467,
+      "learning_rate": 0.0001914658111956423,
+      "loss": 13.7745,
+      "step": 2719
+    },
+    {
+      "epoch": 0.47562841530054645,
+      "grad_norm": 1.2746132822103091,
+      "learning_rate": 0.0001914581773715157,
+      "loss": 14.111,
+      "step": 2720
+    },
+    {
+      "epoch": 0.47580327868852457,
+      "grad_norm": 1.1999216275836322,
+      "learning_rate": 0.00019145054028701193,
+      "loss": 14.0436,
+      "step": 2721
+    },
+    {
+      "epoch": 0.47597814207650274,
+      "grad_norm": 1.2104418106763921,
+      "learning_rate": 0.00019144289994240326,
+      "loss": 14.0545,
+      "step": 2722
+    },
+    {
+      "epoch": 0.47615300546448086,
+      "grad_norm": 1.249971676623531,
+      "learning_rate": 0.00019143525633796203,
+      "loss": 13.9548,
+      "step": 2723
+    },
+    {
+      "epoch": 0.47632786885245904,
+      "grad_norm": 1.3965686765635776,
+      "learning_rate": 0.00019142760947396078,
+      "loss": 13.865,
+      "step": 2724
+    },
+    {
+      "epoch": 0.47650273224043715,
+      "grad_norm": 1.2356126090659816,
+      "learning_rate": 0.00019141995935067205,
+      "loss": 13.9406,
+      "step": 2725
+    },
+    {
+      "epoch": 0.4766775956284153,
+      "grad_norm": 1.2853538323727716,
+      "learning_rate": 0.00019141230596836858,
+      "loss": 14.0004,
+      "step": 2726
+    },
+    {
+      "epoch": 0.47685245901639345,
+      "grad_norm": 1.1862538910606168,
+      "learning_rate": 0.0001914046493273232,
+      "loss": 13.9758,
+      "step": 2727
+    },
+    {
+      "epoch": 0.47702732240437157,
+      "grad_norm": 1.3332162396399185,
+      "learning_rate": 0.00019139698942780887,
+      "loss": 14.0012,
+      "step": 2728
+    },
+    {
+      "epoch": 0.47720218579234974,
+      "grad_norm": 1.2454289987314502,
+      "learning_rate": 0.0001913893262700987,
+      "loss": 13.9813,
+      "step": 2729
+    },
+    {
+      "epoch": 0.47737704918032786,
+      "grad_norm": 1.293170230670198,
+      "learning_rate": 0.00019138165985446576,
+      "loss": 13.8719,
+      "step": 2730
+    },
+    {
+      "epoch": 0.47755191256830604,
+      "grad_norm": 1.295728533160236,
+      "learning_rate": 0.00019137399018118344,
+      "loss": 13.9065,
+      "step": 2731
+    },
+    {
+      "epoch": 0.47772677595628416,
+      "grad_norm": 1.2333882152354971,
+      "learning_rate": 0.00019136631725052514,
+      "loss": 14.0619,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4779016393442623,
+      "grad_norm": 1.278374520549938,
+      "learning_rate": 0.00019135864106276435,
+      "loss": 14.1069,
+      "step": 2733
+    },
+    {
+      "epoch": 0.47807650273224045,
+      "grad_norm": 1.160877430955078,
+      "learning_rate": 0.00019135096161817478,
+      "loss": 13.9965,
+      "step": 2734
+    },
+    {
+      "epoch": 0.47825136612021857,
+      "grad_norm": 1.3859084602559415,
+      "learning_rate": 0.00019134327891703014,
+      "loss": 14.1874,
+      "step": 2735
+    },
+    {
+      "epoch": 0.47842622950819674,
+      "grad_norm": 1.1968212148560113,
+      "learning_rate": 0.00019133559295960435,
+      "loss": 14.1567,
+      "step": 2736
+    },
+    {
+      "epoch": 0.47860109289617486,
+      "grad_norm": 1.2693451140851357,
+      "learning_rate": 0.00019132790374617136,
+      "loss": 13.601,
+      "step": 2737
+    },
+    {
+      "epoch": 0.478775956284153,
+      "grad_norm": 1.2253278995489099,
+      "learning_rate": 0.00019132021127700534,
+      "loss": 13.9727,
+      "step": 2738
+    },
+    {
+      "epoch": 0.47895081967213116,
+      "grad_norm": 1.4203060938572567,
+      "learning_rate": 0.00019131251555238045,
+      "loss": 13.9352,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4791256830601093,
+      "grad_norm": 1.3563855039512465,
+      "learning_rate": 0.00019130481657257109,
+      "loss": 13.7955,
+      "step": 2740
+    },
+    {
+      "epoch": 0.47930054644808745,
+      "grad_norm": 1.226956853011043,
+      "learning_rate": 0.00019129711433785168,
+      "loss": 13.9666,
+      "step": 2741
+    },
+    {
+      "epoch": 0.47947540983606557,
+      "grad_norm": 1.4336446851456082,
+      "learning_rate": 0.0001912894088484968,
+      "loss": 14.0177,
+      "step": 2742
+    },
+    {
+      "epoch": 0.4796502732240437,
+      "grad_norm": 1.416505328052403,
+      "learning_rate": 0.00019128170010478118,
+      "loss": 13.6848,
+      "step": 2743
+    },
+    {
+      "epoch": 0.47982513661202186,
+      "grad_norm": 1.1613952782571584,
+      "learning_rate": 0.00019127398810697962,
+      "loss": 13.8704,
+      "step": 2744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2577117461999212,
+      "learning_rate": 0.000191266272855367,
+      "loss": 13.8653,
+      "step": 2745
+    },
+    {
+      "epoch": 0.48017486338797816,
+      "grad_norm": 1.3746488518971267,
+      "learning_rate": 0.00019125855435021838,
+      "loss": 13.8485,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4803497267759563,
+      "grad_norm": 1.1503611356517327,
+      "learning_rate": 0.00019125083259180893,
+      "loss": 13.999,
+      "step": 2747
+    },
+    {
+      "epoch": 0.48052459016393445,
+      "grad_norm": 1.3928665169088694,
+      "learning_rate": 0.00019124310758041394,
+      "loss": 13.8425,
+      "step": 2748
+    },
+    {
+      "epoch": 0.48069945355191257,
+      "grad_norm": 1.3381552501544673,
+      "learning_rate": 0.00019123537931630872,
+      "loss": 13.9602,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4808743169398907,
+      "grad_norm": 1.2153765722027456,
+      "learning_rate": 0.00019122764779976885,
+      "loss": 14.0177,
+      "step": 2750
+    },
+    {
+      "epoch": 0.48104918032786886,
+      "grad_norm": 1.2634962830667469,
+      "learning_rate": 0.00019121991303106995,
+      "loss": 13.9964,
+      "step": 2751
+    },
+    {
+      "epoch": 0.481224043715847,
+      "grad_norm": 1.140950004660111,
+      "learning_rate": 0.0001912121750104877,
+      "loss": 14.0431,
+      "step": 2752
+    },
+    {
+      "epoch": 0.48139890710382516,
+      "grad_norm": 1.2721661290499897,
+      "learning_rate": 0.00019120443373829798,
+      "loss": 14.0971,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4815737704918033,
+      "grad_norm": 1.225995594234216,
+      "learning_rate": 0.00019119668921477673,
+      "loss": 13.8345,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4817486338797814,
+      "grad_norm": 1.2980053342456503,
+      "learning_rate": 0.00019118894144020012,
+      "loss": 13.6778,
+      "step": 2755
+    },
+    {
+      "epoch": 0.48192349726775957,
+      "grad_norm": 1.2010703864113887,
+      "learning_rate": 0.00019118119041484425,
+      "loss": 14.0549,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4820983606557377,
+      "grad_norm": 1.2144409241919103,
+      "learning_rate": 0.00019117343613898552,
+      "loss": 14.0174,
+      "step": 2757
+    },
+    {
+      "epoch": 0.48227322404371586,
+      "grad_norm": 1.3016908851229452,
+      "learning_rate": 0.00019116567861290028,
+      "loss": 13.9388,
+      "step": 2758
+    },
+    {
+      "epoch": 0.482448087431694,
+      "grad_norm": 1.2382182559188868,
+      "learning_rate": 0.0001911579178368651,
+      "loss": 13.6848,
+      "step": 2759
+    },
+    {
+      "epoch": 0.48262295081967216,
+      "grad_norm": 1.1385554263852473,
+      "learning_rate": 0.0001911501538111567,
+      "loss": 14.0055,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4827978142076503,
+      "grad_norm": 1.2858471315684987,
+      "learning_rate": 0.00019114238653605178,
+      "loss": 14.0425,
+      "step": 2761
+    },
+    {
+      "epoch": 0.4829726775956284,
+      "grad_norm": 1.213608979725418,
+      "learning_rate": 0.00019113461601182724,
+      "loss": 13.9777,
+      "step": 2762
+    },
+    {
+      "epoch": 0.48314754098360657,
+      "grad_norm": 1.4561082669846082,
+      "learning_rate": 0.0001911268422387602,
+      "loss": 13.7519,
+      "step": 2763
+    },
+    {
+      "epoch": 0.4833224043715847,
+      "grad_norm": 1.2812896759114545,
+      "learning_rate": 0.00019111906521712763,
+      "loss": 13.9535,
+      "step": 2764
+    },
+    {
+      "epoch": 0.48349726775956287,
+      "grad_norm": 1.1529262687651818,
+      "learning_rate": 0.00019111128494720688,
+      "loss": 13.7959,
+      "step": 2765
+    },
+    {
+      "epoch": 0.483672131147541,
+      "grad_norm": 1.3894508662127194,
+      "learning_rate": 0.0001911035014292753,
+      "loss": 13.7552,
+      "step": 2766
+    },
+    {
+      "epoch": 0.4838469945355191,
+      "grad_norm": 1.2338140639001536,
+      "learning_rate": 0.00019109571466361029,
+      "loss": 13.7876,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4840218579234973,
+      "grad_norm": 1.3832746298855545,
+      "learning_rate": 0.0001910879246504895,
+      "loss": 13.9657,
+      "step": 2768
+    },
+    {
+      "epoch": 0.4841967213114754,
+      "grad_norm": 1.3728956716364877,
+      "learning_rate": 0.00019108013139019063,
+      "loss": 14.0387,
+      "step": 2769
+    },
+    {
+      "epoch": 0.4843715846994536,
+      "grad_norm": 1.3089553199848079,
+      "learning_rate": 0.00019107233488299148,
+      "loss": 13.6706,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4845464480874317,
+      "grad_norm": 1.0754036280339492,
+      "learning_rate": 0.00019106453512917,
+      "loss": 13.8719,
+      "step": 2771
+    },
+    {
+      "epoch": 0.4847213114754098,
+      "grad_norm": 1.2330137682176296,
+      "learning_rate": 0.00019105673212900425,
+      "loss": 13.7861,
+      "step": 2772
+    },
+    {
+      "epoch": 0.484896174863388,
+      "grad_norm": 1.302107099893207,
+      "learning_rate": 0.0001910489258827724,
+      "loss": 13.9104,
+      "step": 2773
+    },
+    {
+      "epoch": 0.4850710382513661,
+      "grad_norm": 1.1769290455033623,
+      "learning_rate": 0.00019104111639075267,
+      "loss": 13.918,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4852459016393443,
+      "grad_norm": 1.2812733963652814,
+      "learning_rate": 0.00019103330365322356,
+      "loss": 13.8699,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4854207650273224,
+      "grad_norm": 1.0595740855657496,
+      "learning_rate": 0.00019102548767046352,
+      "loss": 13.869,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4855956284153006,
+      "grad_norm": 1.199008326797078,
+      "learning_rate": 0.0001910176684427512,
+      "loss": 14.2194,
+      "step": 2777
+    },
+    {
+      "epoch": 0.4857704918032787,
+      "grad_norm": 1.271816231731783,
+      "learning_rate": 0.00019100984597036536,
+      "loss": 13.8591,
+      "step": 2778
+    },
+    {
+      "epoch": 0.4859453551912568,
+      "grad_norm": 1.2722337213771007,
+      "learning_rate": 0.0001910020202535848,
+      "loss": 13.7346,
+      "step": 2779
+    },
+    {
+      "epoch": 0.486120218579235,
+      "grad_norm": 1.4552237378802115,
+      "learning_rate": 0.00019099419129268855,
+      "loss": 13.9854,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4862950819672131,
+      "grad_norm": 1.4599644235410012,
+      "learning_rate": 0.00019098635908795574,
+      "loss": 13.8875,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4864699453551913,
+      "grad_norm": 1.3386406609895234,
+      "learning_rate": 0.0001909785236396655,
+      "loss": 13.5961,
+      "step": 2782
+    },
+    {
+      "epoch": 0.4866448087431694,
+      "grad_norm": 1.3734704317993047,
+      "learning_rate": 0.0001909706849480972,
+      "loss": 13.9973,
+      "step": 2783
+    },
+    {
+      "epoch": 0.4868196721311475,
+      "grad_norm": 1.6130958799939505,
+      "learning_rate": 0.00019096284301353027,
+      "loss": 13.9688,
+      "step": 2784
+    },
+    {
+      "epoch": 0.4869945355191257,
+      "grad_norm": 1.518369408940572,
+      "learning_rate": 0.00019095499783624423,
+      "loss": 14.0246,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4871693989071038,
+      "grad_norm": 1.34935423176506,
+      "learning_rate": 0.00019094714941651883,
+      "loss": 13.8824,
+      "step": 2786
+    },
+    {
+      "epoch": 0.487344262295082,
+      "grad_norm": 1.4785462994838297,
+      "learning_rate": 0.00019093929775463382,
+      "loss": 14.0251,
+      "step": 2787
+    },
+    {
+      "epoch": 0.4875191256830601,
+      "grad_norm": 1.2108248308490297,
+      "learning_rate": 0.00019093144285086905,
+      "loss": 14.1462,
+      "step": 2788
+    },
+    {
+      "epoch": 0.4876939890710382,
+      "grad_norm": 1.264095217828504,
+      "learning_rate": 0.00019092358470550465,
+      "loss": 13.9098,
+      "step": 2789
+    },
+    {
+      "epoch": 0.4878688524590164,
+      "grad_norm": 1.5863451547755396,
+      "learning_rate": 0.00019091572331882064,
+      "loss": 14.2768,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4880437158469945,
+      "grad_norm": 1.0445790006734113,
+      "learning_rate": 0.00019090785869109732,
+      "loss": 13.8189,
+      "step": 2791
+    },
+    {
+      "epoch": 0.4882185792349727,
+      "grad_norm": 1.2731524600060333,
+      "learning_rate": 0.00019089999082261505,
+      "loss": 14.178,
+      "step": 2792
+    },
+    {
+      "epoch": 0.4883934426229508,
+      "grad_norm": 1.2061636305927945,
+      "learning_rate": 0.00019089211971365434,
+      "loss": 14.0535,
+      "step": 2793
+    },
+    {
+      "epoch": 0.488568306010929,
+      "grad_norm": 1.2323082493588855,
+      "learning_rate": 0.00019088424536449573,
+      "loss": 14.0032,
+      "step": 2794
+    },
+    {
+      "epoch": 0.4887431693989071,
+      "grad_norm": 1.1565889602383628,
+      "learning_rate": 0.00019087636777541996,
+      "loss": 13.7323,
+      "step": 2795
+    },
+    {
+      "epoch": 0.4889180327868852,
+      "grad_norm": 1.1262182228303437,
+      "learning_rate": 0.00019086848694670788,
+      "loss": 14.1451,
+      "step": 2796
+    },
+    {
+      "epoch": 0.4890928961748634,
+      "grad_norm": 1.242166545016731,
+      "learning_rate": 0.00019086060287864037,
+      "loss": 14.045,
+      "step": 2797
+    },
+    {
+      "epoch": 0.4892677595628415,
+      "grad_norm": 1.1780453847176704,
+      "learning_rate": 0.00019085271557149858,
+      "loss": 13.7468,
+      "step": 2798
+    },
+    {
+      "epoch": 0.4894426229508197,
+      "grad_norm": 1.2875596850189237,
+      "learning_rate": 0.0001908448250255636,
+      "loss": 13.5689,
+      "step": 2799
+    },
+    {
+      "epoch": 0.4896174863387978,
+      "grad_norm": 1.260455848818752,
+      "learning_rate": 0.00019083693124111674,
+      "loss": 14.1355,
+      "step": 2800
+    },
+    {
+      "epoch": 0.48979234972677593,
+      "grad_norm": 1.138867921473112,
+      "learning_rate": 0.00019082903421843944,
+      "loss": 14.1499,
+      "step": 2801
+    },
+    {
+      "epoch": 0.4899672131147541,
+      "grad_norm": 1.4679883232207116,
+      "learning_rate": 0.00019082113395781317,
+      "loss": 14.0324,
+      "step": 2802
+    },
+    {
+      "epoch": 0.4901420765027322,
+      "grad_norm": 1.3829562546590837,
+      "learning_rate": 0.0001908132304595196,
+      "loss": 13.7793,
+      "step": 2803
+    },
+    {
+      "epoch": 0.4903169398907104,
+      "grad_norm": 1.1653758408452894,
+      "learning_rate": 0.00019080532372384046,
+      "loss": 13.8745,
+      "step": 2804
+    },
+    {
+      "epoch": 0.4904918032786885,
+      "grad_norm": 1.286985612036507,
+      "learning_rate": 0.00019079741375105762,
+      "loss": 13.7791,
+      "step": 2805
+    },
+    {
+      "epoch": 0.49066666666666664,
+      "grad_norm": 1.1400312615969967,
+      "learning_rate": 0.00019078950054145306,
+      "loss": 13.596,
+      "step": 2806
+    },
+    {
+      "epoch": 0.4908415300546448,
+      "grad_norm": 1.2839289061941153,
+      "learning_rate": 0.00019078158409530888,
+      "loss": 13.7573,
+      "step": 2807
+    },
+    {
+      "epoch": 0.49101639344262293,
+      "grad_norm": 1.1929820143867258,
+      "learning_rate": 0.00019077366441290733,
+      "loss": 14.1584,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4911912568306011,
+      "grad_norm": 1.1382330481553684,
+      "learning_rate": 0.00019076574149453065,
+      "loss": 13.8619,
+      "step": 2809
+    },
+    {
+      "epoch": 0.4913661202185792,
+      "grad_norm": 1.2168472810901585,
+      "learning_rate": 0.00019075781534046138,
+      "loss": 14.0113,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4915409836065574,
+      "grad_norm": 1.3293652186417195,
+      "learning_rate": 0.00019074988595098202,
+      "loss": 13.8939,
+      "step": 2811
+    },
+    {
+      "epoch": 0.4917158469945355,
+      "grad_norm": 1.1385390379859033,
+      "learning_rate": 0.00019074195332637522,
+      "loss": 13.8806,
+      "step": 2812
+    },
+    {
+      "epoch": 0.49189071038251364,
+      "grad_norm": 1.665120700847162,
+      "learning_rate": 0.00019073401746692384,
+      "loss": 13.8042,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4920655737704918,
+      "grad_norm": 1.2230785637217534,
+      "learning_rate": 0.00019072607837291073,
+      "loss": 13.7785,
+      "step": 2814
+    },
+    {
+      "epoch": 0.49224043715846993,
+      "grad_norm": 1.4271677264796188,
+      "learning_rate": 0.00019071813604461897,
+      "loss": 13.941,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4924153005464481,
+      "grad_norm": 1.3358827665703934,
+      "learning_rate": 0.00019071019048233162,
+      "loss": 13.9439,
+      "step": 2816
+    },
+    {
+      "epoch": 0.49259016393442623,
+      "grad_norm": 1.1830243313491589,
+      "learning_rate": 0.00019070224168633198,
+      "loss": 13.905,
+      "step": 2817
+    },
+    {
+      "epoch": 0.49276502732240435,
+      "grad_norm": 1.403102340937304,
+      "learning_rate": 0.00019069428965690338,
+      "loss": 13.9176,
+      "step": 2818
+    },
+    {
+      "epoch": 0.4929398907103825,
+      "grad_norm": 1.7536454986706784,
+      "learning_rate": 0.0001906863343943293,
+      "loss": 13.6727,
+      "step": 2819
+    },
+    {
+      "epoch": 0.49311475409836064,
+      "grad_norm": 1.2061011745044328,
+      "learning_rate": 0.0001906783758988934,
+      "loss": 13.9793,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4932896174863388,
+      "grad_norm": 1.3159794832367089,
+      "learning_rate": 0.00019067041417087933,
+      "loss": 13.5188,
+      "step": 2821
+    },
+    {
+      "epoch": 0.49346448087431694,
+      "grad_norm": 1.196448100527925,
+      "learning_rate": 0.0001906624492105709,
+      "loss": 14.0034,
+      "step": 2822
+    },
+    {
+      "epoch": 0.49363934426229505,
+      "grad_norm": 1.4343886349926234,
+      "learning_rate": 0.00019065448101825212,
+      "loss": 13.9949,
+      "step": 2823
+    },
+    {
+      "epoch": 0.49381420765027323,
+      "grad_norm": 1.2293758772119776,
+      "learning_rate": 0.000190646509594207,
+      "loss": 14.0928,
+      "step": 2824
+    },
+    {
+      "epoch": 0.49398907103825135,
+      "grad_norm": 1.5108793383864856,
+      "learning_rate": 0.00019063853493871973,
+      "loss": 13.5891,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4941639344262295,
+      "grad_norm": 1.4162191189145734,
+      "learning_rate": 0.00019063055705207456,
+      "loss": 13.9415,
+      "step": 2826
+    },
+    {
+      "epoch": 0.49433879781420764,
+      "grad_norm": 1.28211654984096,
+      "learning_rate": 0.00019062257593455595,
+      "loss": 13.7172,
+      "step": 2827
+    },
+    {
+      "epoch": 0.4945136612021858,
+      "grad_norm": 1.3359353794741904,
+      "learning_rate": 0.0001906145915864484,
+      "loss": 13.7796,
+      "step": 2828
+    },
+    {
+      "epoch": 0.49468852459016394,
+      "grad_norm": 1.2160951815337444,
+      "learning_rate": 0.0001906066040080365,
+      "loss": 13.8888,
+      "step": 2829
+    },
+    {
+      "epoch": 0.49486338797814206,
+      "grad_norm": 1.4806448008159547,
+      "learning_rate": 0.00019059861319960502,
+      "loss": 13.8087,
+      "step": 2830
+    },
+    {
+      "epoch": 0.49503825136612023,
+      "grad_norm": 1.1365445449440972,
+      "learning_rate": 0.00019059061916143887,
+      "loss": 13.77,
+      "step": 2831
+    },
+    {
+      "epoch": 0.49521311475409835,
+      "grad_norm": 1.3780277604567965,
+      "learning_rate": 0.00019058262189382296,
+      "loss": 14.0985,
+      "step": 2832
+    },
+    {
+      "epoch": 0.4953879781420765,
+      "grad_norm": 1.2070477717844164,
+      "learning_rate": 0.00019057462139704243,
+      "loss": 13.6331,
+      "step": 2833
+    },
+    {
+      "epoch": 0.49556284153005464,
+      "grad_norm": 1.2429235493915116,
+      "learning_rate": 0.00019056661767138247,
+      "loss": 13.9853,
+      "step": 2834
+    },
+    {
+      "epoch": 0.49573770491803276,
+      "grad_norm": 1.5291472529310095,
+      "learning_rate": 0.00019055861071712839,
+      "loss": 13.6425,
+      "step": 2835
+    },
+    {
+      "epoch": 0.49591256830601094,
+      "grad_norm": 1.2028007363181514,
+      "learning_rate": 0.00019055060053456568,
+      "loss": 13.6941,
+      "step": 2836
+    },
+    {
+      "epoch": 0.49608743169398906,
+      "grad_norm": 1.2455280766336256,
+      "learning_rate": 0.00019054258712397984,
+      "loss": 14.0715,
+      "step": 2837
+    },
+    {
+      "epoch": 0.49626229508196723,
+      "grad_norm": 1.2857136888681702,
+      "learning_rate": 0.00019053457048565656,
+      "loss": 13.8673,
+      "step": 2838
+    },
+    {
+      "epoch": 0.49643715846994535,
+      "grad_norm": 1.2813539677009285,
+      "learning_rate": 0.0001905265506198816,
+      "loss": 13.7924,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4966120218579235,
+      "grad_norm": 1.1435021369641678,
+      "learning_rate": 0.0001905185275269409,
+      "loss": 13.7113,
+      "step": 2840
+    },
+    {
+      "epoch": 0.49678688524590164,
+      "grad_norm": 1.2099117556422025,
+      "learning_rate": 0.00019051050120712045,
+      "loss": 14.0972,
+      "step": 2841
+    },
+    {
+      "epoch": 0.49696174863387976,
+      "grad_norm": 1.1964304482697798,
+      "learning_rate": 0.00019050247166070637,
+      "loss": 13.9908,
+      "step": 2842
+    },
+    {
+      "epoch": 0.49713661202185794,
+      "grad_norm": 1.209687924113987,
+      "learning_rate": 0.00019049443888798494,
+      "loss": 14.1253,
+      "step": 2843
+    },
+    {
+      "epoch": 0.49731147540983606,
+      "grad_norm": 1.108045026248808,
+      "learning_rate": 0.0001904864028892425,
+      "loss": 13.6802,
+      "step": 2844
+    },
+    {
+      "epoch": 0.49748633879781423,
+      "grad_norm": 1.2072004527691764,
+      "learning_rate": 0.00019047836366476548,
+      "loss": 13.8197,
+      "step": 2845
+    },
+    {
+      "epoch": 0.49766120218579235,
+      "grad_norm": 1.112239085181912,
+      "learning_rate": 0.00019047032121484052,
+      "loss": 13.8631,
+      "step": 2846
+    },
+    {
+      "epoch": 0.49783606557377047,
+      "grad_norm": 1.1703720382742087,
+      "learning_rate": 0.0001904622755397543,
+      "loss": 13.7094,
+      "step": 2847
+    },
+    {
+      "epoch": 0.49801092896174864,
+      "grad_norm": 1.2406363002098666,
+      "learning_rate": 0.0001904542266397937,
+      "loss": 13.9789,
+      "step": 2848
+    },
+    {
+      "epoch": 0.49818579234972676,
+      "grad_norm": 1.3037041734751695,
+      "learning_rate": 0.00019044617451524558,
+      "loss": 14.1769,
+      "step": 2849
+    },
+    {
+      "epoch": 0.49836065573770494,
+      "grad_norm": 1.3938700864828997,
+      "learning_rate": 0.000190438119166397,
+      "loss": 14.073,
+      "step": 2850
+    },
+    {
+      "epoch": 0.49853551912568306,
+      "grad_norm": 1.1958996801664235,
+      "learning_rate": 0.00019043006059353513,
+      "loss": 14.0115,
+      "step": 2851
+    },
+    {
+      "epoch": 0.4987103825136612,
+      "grad_norm": 1.3015499139123632,
+      "learning_rate": 0.00019042199879694727,
+      "loss": 13.9574,
+      "step": 2852
+    },
+    {
+      "epoch": 0.49888524590163935,
+      "grad_norm": 1.1800467748985315,
+      "learning_rate": 0.0001904139337769208,
+      "loss": 13.9483,
+      "step": 2853
+    },
+    {
+      "epoch": 0.49906010928961747,
+      "grad_norm": 1.3409827765152198,
+      "learning_rate": 0.00019040586553374325,
+      "loss": 13.8788,
+      "step": 2854
+    },
+    {
+      "epoch": 0.49923497267759565,
+      "grad_norm": 1.177757896848586,
+      "learning_rate": 0.0001903977940677022,
+      "loss": 13.737,
+      "step": 2855
+    },
+    {
+      "epoch": 0.49940983606557376,
+      "grad_norm": 1.217641383915674,
+      "learning_rate": 0.00019038971937908542,
+      "loss": 13.6898,
+      "step": 2856
+    },
+    {
+      "epoch": 0.49958469945355194,
+      "grad_norm": 1.3433231004159687,
+      "learning_rate": 0.00019038164146818073,
+      "loss": 13.7248,
+      "step": 2857
+    },
+    {
+      "epoch": 0.49975956284153006,
+      "grad_norm": 1.144089517458212,
+      "learning_rate": 0.00019037356033527615,
+      "loss": 13.8596,
+      "step": 2858
+    },
+    {
+      "epoch": 0.4999344262295082,
+      "grad_norm": 1.2175678637873133,
+      "learning_rate": 0.00019036547598065974,
+      "loss": 14.0139,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5001092896174864,
+      "grad_norm": 1.2368348940379503,
+      "learning_rate": 0.00019035738840461966,
+      "loss": 13.8082,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5002841530054645,
+      "grad_norm": 1.1461221698311972,
+      "learning_rate": 0.00019034929760744425,
+      "loss": 13.8249,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5004590163934426,
+      "grad_norm": 1.1926006549769868,
+      "learning_rate": 0.00019034120358942195,
+      "loss": 14.1505,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5006338797814207,
+      "grad_norm": 1.1726045239032084,
+      "learning_rate": 0.00019033310635084132,
+      "loss": 13.9102,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5008087431693989,
+      "grad_norm": 1.2189539895471697,
+      "learning_rate": 0.00019032500589199097,
+      "loss": 13.9328,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5009836065573771,
+      "grad_norm": 1.352679394457732,
+      "learning_rate": 0.0001903169022131597,
+      "loss": 13.9434,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5011584699453552,
+      "grad_norm": 1.2110852983224147,
+      "learning_rate": 0.0001903087953146364,
+      "loss": 14.0935,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5013333333333333,
+      "grad_norm": 1.299017580743337,
+      "learning_rate": 0.00019030068519671004,
+      "loss": 13.8898,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5015081967213115,
+      "grad_norm": 1.2690203627845316,
+      "learning_rate": 0.00019029257185966975,
+      "loss": 14.0417,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5016830601092896,
+      "grad_norm": 1.174178280412931,
+      "learning_rate": 0.00019028445530380479,
+      "loss": 13.9641,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5018579234972678,
+      "grad_norm": 1.2011187498759839,
+      "learning_rate": 0.0001902763355294045,
+      "loss": 13.9561,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5020327868852459,
+      "grad_norm": 1.1244023289540204,
+      "learning_rate": 0.0001902682125367583,
+      "loss": 13.8363,
+      "step": 2871
+    },
+    {
+      "epoch": 0.502207650273224,
+      "grad_norm": 1.1859725035862998,
+      "learning_rate": 0.0001902600863261558,
+      "loss": 13.8732,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5023825136612022,
+      "grad_norm": 1.3799226157491784,
+      "learning_rate": 0.00019025195689788667,
+      "loss": 13.8756,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5025573770491804,
+      "grad_norm": 1.1524393302390896,
+      "learning_rate": 0.00019024382425224072,
+      "loss": 13.5822,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5027322404371585,
+      "grad_norm": 1.2777898303306028,
+      "learning_rate": 0.0001902356883895079,
+      "loss": 13.7607,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5029071038251366,
+      "grad_norm": 1.360456648355831,
+      "learning_rate": 0.00019022754930997818,
+      "loss": 13.8528,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5030819672131147,
+      "grad_norm": 1.2071312143686148,
+      "learning_rate": 0.0001902194070139418,
+      "loss": 13.9616,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5032568306010929,
+      "grad_norm": 1.1683610604787498,
+      "learning_rate": 0.00019021126150168893,
+      "loss": 13.9703,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5034316939890711,
+      "grad_norm": 1.1588879377213224,
+      "learning_rate": 0.00019020311277351,
+      "loss": 14.0105,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5036065573770492,
+      "grad_norm": 1.1561920755808799,
+      "learning_rate": 0.00019019496082969549,
+      "loss": 13.6089,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5037814207650273,
+      "grad_norm": 1.2167949622522187,
+      "learning_rate": 0.000190186805670536,
+      "loss": 13.6412,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5039562841530054,
+      "grad_norm": 1.1678032114453896,
+      "learning_rate": 0.00019017864729632228,
+      "loss": 13.9914,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5041311475409836,
+      "grad_norm": 1.231625890489481,
+      "learning_rate": 0.00019017048570734512,
+      "loss": 13.8864,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5043060109289618,
+      "grad_norm": 1.171982349271538,
+      "learning_rate": 0.00019016232090389548,
+      "loss": 13.8209,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5044808743169399,
+      "grad_norm": 1.3268909097934662,
+      "learning_rate": 0.0001901541528862645,
+      "loss": 13.9774,
+      "step": 2885
+    },
+    {
+      "epoch": 0.504655737704918,
+      "grad_norm": 1.160834754578347,
+      "learning_rate": 0.0001901459816547433,
+      "loss": 13.8288,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5048306010928961,
+      "grad_norm": 1.2919948420939367,
+      "learning_rate": 0.00019013780720962314,
+      "loss": 13.7821,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5050054644808744,
+      "grad_norm": 1.2841218459571861,
+      "learning_rate": 0.0001901296295511955,
+      "loss": 13.7225,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5051803278688525,
+      "grad_norm": 1.2500797423188543,
+      "learning_rate": 0.00019012144867975184,
+      "loss": 13.6021,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5053551912568306,
+      "grad_norm": 1.306690595721468,
+      "learning_rate": 0.00019011326459558383,
+      "loss": 14.019,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5055300546448087,
+      "grad_norm": 1.3849977268414941,
+      "learning_rate": 0.00019010507729898322,
+      "loss": 13.7073,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5057049180327868,
+      "grad_norm": 1.0986092200244337,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 13.9445,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5058797814207651,
+      "grad_norm": 1.3281974336123372,
+      "learning_rate": 0.00019008869306965184,
+      "loss": 13.9035,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5060546448087432,
+      "grad_norm": 1.2222054977994015,
+      "learning_rate": 0.00019008049613750513,
+      "loss": 13.7802,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5062295081967213,
+      "grad_norm": 1.3828804625191164,
+      "learning_rate": 0.000190072295994094,
+      "loss": 13.9509,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5064043715846994,
+      "grad_norm": 1.3135010891955583,
+      "learning_rate": 0.0001900640926397107,
+      "loss": 13.6918,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5065792349726777,
+      "grad_norm": 1.2572059536883085,
+      "learning_rate": 0.00019005588607464775,
+      "loss": 13.9823,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5067540983606558,
+      "grad_norm": 1.3863731587017298,
+      "learning_rate": 0.00019004767629919768,
+      "loss": 13.743,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5069289617486339,
+      "grad_norm": 1.2316463758927474,
+      "learning_rate": 0.00019003946331365318,
+      "loss": 13.7398,
+      "step": 2899
+    },
+    {
+      "epoch": 0.507103825136612,
+      "grad_norm": 1.2721799707451609,
+      "learning_rate": 0.000190031247118307,
+      "loss": 13.4961,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5072786885245901,
+      "grad_norm": 1.2488563868688978,
+      "learning_rate": 0.00019002302771345205,
+      "loss": 13.6025,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5074535519125684,
+      "grad_norm": 1.3987593268851117,
+      "learning_rate": 0.00019001480509938133,
+      "loss": 13.8392,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5076284153005465,
+      "grad_norm": 1.292024235648191,
+      "learning_rate": 0.00019000657927638798,
+      "loss": 13.8774,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5078032786885246,
+      "grad_norm": 1.255593884475766,
+      "learning_rate": 0.00018999835024476524,
+      "loss": 13.8484,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5079781420765027,
+      "grad_norm": 1.3469540204849297,
+      "learning_rate": 0.0001899901180048065,
+      "loss": 14.0248,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5081530054644808,
+      "grad_norm": 1.246069432246737,
+      "learning_rate": 0.00018998188255680515,
+      "loss": 13.892,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5083278688524591,
+      "grad_norm": 1.2239250526060876,
+      "learning_rate": 0.00018997364390105484,
+      "loss": 13.4386,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5085027322404372,
+      "grad_norm": 1.1345462599305747,
+      "learning_rate": 0.00018996540203784922,
+      "loss": 13.819,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5086775956284153,
+      "grad_norm": 1.2804627414061998,
+      "learning_rate": 0.00018995715696748218,
+      "loss": 13.6044,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5088524590163934,
+      "grad_norm": 1.247083125333989,
+      "learning_rate": 0.00018994890869024756,
+      "loss": 13.9543,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5090273224043715,
+      "grad_norm": 1.254064002678629,
+      "learning_rate": 0.00018994065720643946,
+      "loss": 13.9206,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5092021857923498,
+      "grad_norm": 1.1843763846722941,
+      "learning_rate": 0.00018993240251635203,
+      "loss": 13.8914,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5093770491803279,
+      "grad_norm": 1.1496364772410994,
+      "learning_rate": 0.00018992414462027946,
+      "loss": 13.817,
+      "step": 2913
+    },
+    {
+      "epoch": 0.509551912568306,
+      "grad_norm": 1.3140927001443021,
+      "learning_rate": 0.00018991588351851627,
+      "loss": 13.6606,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5097267759562841,
+      "grad_norm": 1.173325229512116,
+      "learning_rate": 0.00018990761921135687,
+      "loss": 13.8049,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5099016393442622,
+      "grad_norm": 1.31207807670621,
+      "learning_rate": 0.00018989935169909586,
+      "loss": 13.908,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5100765027322405,
+      "grad_norm": 1.1362180596986993,
+      "learning_rate": 0.00018989108098202802,
+      "loss": 13.8933,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5102513661202186,
+      "grad_norm": 1.0979271859964237,
+      "learning_rate": 0.0001898828070604482,
+      "loss": 13.871,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5104262295081967,
+      "grad_norm": 1.232034348696334,
+      "learning_rate": 0.0001898745299346513,
+      "loss": 13.7647,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5106010928961748,
+      "grad_norm": 1.1333619618111277,
+      "learning_rate": 0.0001898662496049324,
+      "loss": 14.0367,
+      "step": 2920
+    },
+    {
+      "epoch": 0.510775956284153,
+      "grad_norm": 1.1218015312373768,
+      "learning_rate": 0.0001898579660715867,
+      "loss": 13.7393,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5109508196721312,
+      "grad_norm": 1.1832808180566834,
+      "learning_rate": 0.00018984967933490952,
+      "loss": 14.0408,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5111256830601093,
+      "grad_norm": 1.2304788217690628,
+      "learning_rate": 0.0001898413893951962,
+      "loss": 13.609,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5113005464480874,
+      "grad_norm": 1.1589708539118033,
+      "learning_rate": 0.00018983309625274236,
+      "loss": 13.7612,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5114754098360655,
+      "grad_norm": 1.2444728573372792,
+      "learning_rate": 0.0001898247999078436,
+      "loss": 13.9508,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5116502732240437,
+      "grad_norm": 1.4289014651748693,
+      "learning_rate": 0.00018981650036079567,
+      "loss": 13.651,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5118251366120219,
+      "grad_norm": 1.1875098238718673,
+      "learning_rate": 0.00018980819761189443,
+      "loss": 13.7696,
+      "step": 2927
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 1.3055812569856053,
+      "learning_rate": 0.00018979989166143587,
+      "loss": 13.9737,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5121748633879781,
+      "grad_norm": 1.321976034760829,
+      "learning_rate": 0.00018979158250971608,
+      "loss": 13.8634,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5123497267759562,
+      "grad_norm": 1.198900516659969,
+      "learning_rate": 0.00018978327015703133,
+      "loss": 13.8762,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5125245901639345,
+      "grad_norm": 1.2106877548695654,
+      "learning_rate": 0.00018977495460367787,
+      "loss": 13.766,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5126994535519126,
+      "grad_norm": 1.2203547314546754,
+      "learning_rate": 0.00018976663584995218,
+      "loss": 13.9192,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5128743169398907,
+      "grad_norm": 1.3170312067637853,
+      "learning_rate": 0.00018975831389615077,
+      "loss": 13.7319,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5130491803278688,
+      "grad_norm": 1.1270482220898297,
+      "learning_rate": 0.00018974998874257036,
+      "loss": 13.8068,
+      "step": 2934
+    },
+    {
+      "epoch": 0.513224043715847,
+      "grad_norm": 1.3037263755413357,
+      "learning_rate": 0.00018974166038950773,
+      "loss": 13.8786,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5133989071038252,
+      "grad_norm": 1.0896580410468732,
+      "learning_rate": 0.00018973332883725976,
+      "loss": 13.8027,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5135737704918033,
+      "grad_norm": 1.418043252931441,
+      "learning_rate": 0.00018972499408612342,
+      "loss": 13.9245,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5137486338797814,
+      "grad_norm": 1.3213810742652363,
+      "learning_rate": 0.00018971665613639593,
+      "loss": 13.5499,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5139234972677595,
+      "grad_norm": 1.4678574659150916,
+      "learning_rate": 0.00018970831498837443,
+      "loss": 13.7502,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5140983606557377,
+      "grad_norm": 1.4521726692724521,
+      "learning_rate": 0.00018969997064235633,
+      "loss": 13.8744,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5142732240437159,
+      "grad_norm": 1.0954247952839378,
+      "learning_rate": 0.00018969162309863906,
+      "loss": 13.9372,
+      "step": 2941
+    },
+    {
+      "epoch": 0.514448087431694,
+      "grad_norm": 1.4443664013328197,
+      "learning_rate": 0.00018968327235752022,
+      "loss": 13.8786,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5146229508196721,
+      "grad_norm": 1.2493612695740308,
+      "learning_rate": 0.00018967491841929754,
+      "loss": 14.0713,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5147978142076503,
+      "grad_norm": 1.4101956892230298,
+      "learning_rate": 0.00018966656128426877,
+      "loss": 13.955,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5149726775956284,
+      "grad_norm": 1.4396339248547667,
+      "learning_rate": 0.00018965820095273185,
+      "loss": 13.6374,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5151475409836066,
+      "grad_norm": 1.1095385500783588,
+      "learning_rate": 0.00018964983742498484,
+      "loss": 13.8309,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5153224043715847,
+      "grad_norm": 1.4727235762565791,
+      "learning_rate": 0.00018964147070132584,
+      "loss": 14.0092,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5154972677595628,
+      "grad_norm": 1.1660317881187727,
+      "learning_rate": 0.00018963310078205317,
+      "loss": 13.8606,
+      "step": 2948
+    },
+    {
+      "epoch": 0.515672131147541,
+      "grad_norm": 1.4190572249268663,
+      "learning_rate": 0.0001896247276674652,
+      "loss": 14.025,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5158469945355191,
+      "grad_norm": 1.2191310206187107,
+      "learning_rate": 0.00018961635135786035,
+      "loss": 13.8402,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5160218579234973,
+      "grad_norm": 1.150164289460907,
+      "learning_rate": 0.00018960797185353732,
+      "loss": 13.8369,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5161967213114754,
+      "grad_norm": 1.2889902344673023,
+      "learning_rate": 0.00018959958915479476,
+      "loss": 14.0094,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5163715846994535,
+      "grad_norm": 1.3235903297894023,
+      "learning_rate": 0.00018959120326193156,
+      "loss": 13.7814,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5165464480874317,
+      "grad_norm": 1.1869330020267272,
+      "learning_rate": 0.0001895828141752466,
+      "loss": 13.6945,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5167213114754098,
+      "grad_norm": 1.2343467119366665,
+      "learning_rate": 0.00018957442189503906,
+      "loss": 13.7742,
+      "step": 2955
+    },
+    {
+      "epoch": 0.516896174863388,
+      "grad_norm": 1.202445744160247,
+      "learning_rate": 0.00018956602642160796,
+      "loss": 13.8397,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5170710382513661,
+      "grad_norm": 1.2020508067533828,
+      "learning_rate": 0.0001895576277552527,
+      "loss": 13.8279,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5172459016393443,
+      "grad_norm": 1.238653069449452,
+      "learning_rate": 0.00018954922589627262,
+      "loss": 13.8375,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5174207650273224,
+      "grad_norm": 1.2022025892264472,
+      "learning_rate": 0.0001895408208449673,
+      "loss": 13.7492,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5175956284153006,
+      "grad_norm": 1.3193513142513154,
+      "learning_rate": 0.0001895324126016363,
+      "loss": 13.9002,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5177704918032787,
+      "grad_norm": 1.1121191763484037,
+      "learning_rate": 0.0001895240011665794,
+      "loss": 13.7941,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5179453551912568,
+      "grad_norm": 1.261533630530726,
+      "learning_rate": 0.00018951558654009646,
+      "loss": 13.8914,
+      "step": 2962
+    },
+    {
+      "epoch": 0.518120218579235,
+      "grad_norm": 1.2904816143951479,
+      "learning_rate": 0.00018950716872248748,
+      "loss": 13.8479,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5182950819672131,
+      "grad_norm": 1.3498996736056883,
+      "learning_rate": 0.0001894987477140525,
+      "loss": 13.6599,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5184699453551913,
+      "grad_norm": 1.5172361690854903,
+      "learning_rate": 0.0001894903235150917,
+      "loss": 13.9095,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5186448087431694,
+      "grad_norm": 1.3225164954074484,
+      "learning_rate": 0.00018948189612590547,
+      "loss": 14.0741,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5188196721311475,
+      "grad_norm": 1.3964289867852602,
+      "learning_rate": 0.00018947346554679416,
+      "loss": 13.9047,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5189945355191257,
+      "grad_norm": 1.421889013026338,
+      "learning_rate": 0.00018946503177805834,
+      "loss": 14.0741,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5191693989071038,
+      "grad_norm": 1.1250830884827094,
+      "learning_rate": 0.00018945659481999866,
+      "loss": 13.7141,
+      "step": 2969
+    },
+    {
+      "epoch": 0.519344262295082,
+      "grad_norm": 1.2731664866857395,
+      "learning_rate": 0.00018944815467291593,
+      "loss": 13.6915,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5195191256830601,
+      "grad_norm": 1.2060892260908518,
+      "learning_rate": 0.00018943971133711095,
+      "loss": 13.8007,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5196939890710383,
+      "grad_norm": 1.2776421789294636,
+      "learning_rate": 0.0001894312648128848,
+      "loss": 13.5314,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5198688524590164,
+      "grad_norm": 1.1936252401514795,
+      "learning_rate": 0.00018942281510053854,
+      "loss": 13.6678,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5200437158469945,
+      "grad_norm": 1.1891583985767826,
+      "learning_rate": 0.0001894143622003734,
+      "loss": 13.9522,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5202185792349727,
+      "grad_norm": 1.1394469586476341,
+      "learning_rate": 0.0001894059061126907,
+      "loss": 13.9298,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5203934426229508,
+      "grad_norm": 1.1708869054922726,
+      "learning_rate": 0.00018939744683779192,
+      "loss": 13.7963,
+      "step": 2976
+    },
+    {
+      "epoch": 0.520568306010929,
+      "grad_norm": 1.1869979404543702,
+      "learning_rate": 0.0001893889843759786,
+      "loss": 13.8493,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5207431693989071,
+      "grad_norm": 1.170558415517968,
+      "learning_rate": 0.00018938051872755245,
+      "loss": 13.597,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5209180327868852,
+      "grad_norm": 1.1373272084651225,
+      "learning_rate": 0.0001893720498928152,
+      "loss": 13.6153,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5210928961748634,
+      "grad_norm": 1.2400691649874687,
+      "learning_rate": 0.0001893635778720688,
+      "loss": 13.8188,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5212677595628415,
+      "grad_norm": 1.1750382055889652,
+      "learning_rate": 0.00018935510266561529,
+      "loss": 14.0181,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5214426229508197,
+      "grad_norm": 1.2517642791531403,
+      "learning_rate": 0.00018934662427375675,
+      "loss": 13.4506,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5216174863387978,
+      "grad_norm": 1.2803031068525657,
+      "learning_rate": 0.00018933814269679546,
+      "loss": 13.6498,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5217923497267759,
+      "grad_norm": 1.1926556841652434,
+      "learning_rate": 0.00018932965793503375,
+      "loss": 13.3,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5219672131147541,
+      "grad_norm": 1.1982858006724904,
+      "learning_rate": 0.0001893211699887741,
+      "loss": 13.8608,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5221420765027323,
+      "grad_norm": 1.3271780056258597,
+      "learning_rate": 0.00018931267885831908,
+      "loss": 13.736,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5223169398907104,
+      "grad_norm": 1.1487847615735813,
+      "learning_rate": 0.00018930418454397144,
+      "loss": 13.6561,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5224918032786885,
+      "grad_norm": 1.2267886294971446,
+      "learning_rate": 0.00018929568704603397,
+      "loss": 13.8088,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5226666666666666,
+      "grad_norm": 1.1545317811195526,
+      "learning_rate": 0.00018928718636480958,
+      "loss": 13.6549,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5228415300546448,
+      "grad_norm": 1.3063244721811762,
+      "learning_rate": 0.0001892786825006013,
+      "loss": 13.8689,
+      "step": 2990
+    },
+    {
+      "epoch": 0.523016393442623,
+      "grad_norm": 1.3822422717963427,
+      "learning_rate": 0.00018927017545371233,
+      "loss": 14.0317,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5231912568306011,
+      "grad_norm": 1.3582676168534595,
+      "learning_rate": 0.00018926166522444587,
+      "loss": 13.9243,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5233661202185792,
+      "grad_norm": 1.5680643772378189,
+      "learning_rate": 0.00018925315181310537,
+      "loss": 13.8838,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5235409836065574,
+      "grad_norm": 1.2327303488815695,
+      "learning_rate": 0.00018924463521999426,
+      "loss": 13.8333,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5237158469945355,
+      "grad_norm": 1.190682259845902,
+      "learning_rate": 0.00018923611544541618,
+      "loss": 13.4787,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5238907103825137,
+      "grad_norm": 1.1982507986602027,
+      "learning_rate": 0.00018922759248967484,
+      "loss": 13.6529,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5240655737704918,
+      "grad_norm": 1.202590178221436,
+      "learning_rate": 0.0001892190663530741,
+      "loss": 13.921,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5242404371584699,
+      "grad_norm": 1.1946948974005123,
+      "learning_rate": 0.0001892105370359179,
+      "loss": 13.8333,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5244153005464481,
+      "grad_norm": 1.1809629784497837,
+      "learning_rate": 0.00018920200453851024,
+      "loss": 13.8892,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 1.3572564863523213,
+      "learning_rate": 0.00018919346886115537,
+      "loss": 13.7687,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5247650273224044,
+      "grad_norm": 1.151696911011658,
+      "learning_rate": 0.00018918493000415754,
+      "loss": 14.0057,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5249398907103825,
+      "grad_norm": 1.2292101006648233,
+      "learning_rate": 0.00018917638796782116,
+      "loss": 13.7566,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5251147540983606,
+      "grad_norm": 1.1114998570201353,
+      "learning_rate": 0.0001891678427524507,
+      "loss": 13.7065,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5252896174863388,
+      "grad_norm": 1.3237804472551213,
+      "learning_rate": 0.0001891592943583509,
+      "loss": 13.8113,
+      "step": 3004
+    },
+    {
+      "epoch": 0.525464480874317,
+      "grad_norm": 1.1762677170322886,
+      "learning_rate": 0.00018915074278582638,
+      "loss": 13.7204,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5256393442622951,
+      "grad_norm": 1.3116211173069012,
+      "learning_rate": 0.00018914218803518207,
+      "loss": 13.6716,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5258142076502732,
+      "grad_norm": 1.1694143259646128,
+      "learning_rate": 0.0001891336301067229,
+      "loss": 13.8788,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5259890710382513,
+      "grad_norm": 1.3331294284114494,
+      "learning_rate": 0.00018912506900075393,
+      "loss": 13.6962,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5261639344262296,
+      "grad_norm": 1.307672607245653,
+      "learning_rate": 0.0001891165047175804,
+      "loss": 13.7725,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5263387978142077,
+      "grad_norm": 1.2331050136467847,
+      "learning_rate": 0.00018910793725750763,
+      "loss": 13.8234,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5265136612021858,
+      "grad_norm": 1.4356042993755485,
+      "learning_rate": 0.00018909936662084094,
+      "loss": 13.6022,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5266885245901639,
+      "grad_norm": 1.259662784633729,
+      "learning_rate": 0.00018909079280788599,
+      "loss": 13.8524,
+      "step": 3012
+    },
+    {
+      "epoch": 0.526863387978142,
+      "grad_norm": 1.7770484822694672,
+      "learning_rate": 0.00018908221581894833,
+      "loss": 13.8814,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5270382513661203,
+      "grad_norm": 1.2797668722563265,
+      "learning_rate": 0.00018907363565433377,
+      "loss": 13.9568,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5272131147540984,
+      "grad_norm": 1.4914348798547032,
+      "learning_rate": 0.0001890650523143482,
+      "loss": 13.791,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5273879781420765,
+      "grad_norm": 1.4080676898156628,
+      "learning_rate": 0.00018905646579929752,
+      "loss": 13.8568,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5275628415300546,
+      "grad_norm": 1.1678879535032416,
+      "learning_rate": 0.00018904787610948793,
+      "loss": 13.8778,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5277377049180327,
+      "grad_norm": 1.2036740610931085,
+      "learning_rate": 0.0001890392832452256,
+      "loss": 13.8252,
+      "step": 3018
+    },
+    {
+      "epoch": 0.527912568306011,
+      "grad_norm": 1.2469591355288023,
+      "learning_rate": 0.0001890306872068168,
+      "loss": 13.4424,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5280874316939891,
+      "grad_norm": 1.4732295103466813,
+      "learning_rate": 0.00018902208799456807,
+      "loss": 13.7159,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5282622950819672,
+      "grad_norm": 1.2629627850785576,
+      "learning_rate": 0.0001890134856087859,
+      "loss": 13.7653,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5284371584699453,
+      "grad_norm": 1.63463902251391,
+      "learning_rate": 0.000189004880049777,
+      "loss": 13.6301,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5286120218579234,
+      "grad_norm": 1.1811010343657327,
+      "learning_rate": 0.00018899627131784808,
+      "loss": 13.6463,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5287868852459017,
+      "grad_norm": 1.3055530843172165,
+      "learning_rate": 0.00018898765941330608,
+      "loss": 13.478,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5289617486338798,
+      "grad_norm": 1.1829335964637098,
+      "learning_rate": 0.000188979044336458,
+      "loss": 13.5353,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5291366120218579,
+      "grad_norm": 1.2983629308530915,
+      "learning_rate": 0.00018897042608761094,
+      "loss": 13.5867,
+      "step": 3026
+    },
+    {
+      "epoch": 0.529311475409836,
+      "grad_norm": 1.2429006551864084,
+      "learning_rate": 0.00018896180466707212,
+      "loss": 13.8914,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5294863387978143,
+      "grad_norm": 1.3815940581934025,
+      "learning_rate": 0.00018895318007514894,
+      "loss": 13.9443,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5296612021857924,
+      "grad_norm": 1.2472541038784366,
+      "learning_rate": 0.00018894455231214881,
+      "loss": 13.6543,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5298360655737705,
+      "grad_norm": 1.3302395699202638,
+      "learning_rate": 0.0001889359213783793,
+      "loss": 13.3144,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5300109289617486,
+      "grad_norm": 1.5572661297882524,
+      "learning_rate": 0.0001889272872741481,
+      "loss": 13.7339,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5301857923497267,
+      "grad_norm": 1.282341127296372,
+      "learning_rate": 0.00018891864999976305,
+      "loss": 13.8444,
+      "step": 3032
+    },
+    {
+      "epoch": 0.530360655737705,
+      "grad_norm": 1.4494002039069231,
+      "learning_rate": 0.00018891000955553196,
+      "loss": 13.7634,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5305355191256831,
+      "grad_norm": 1.5400714890794582,
+      "learning_rate": 0.00018890136594176295,
+      "loss": 13.8013,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5307103825136612,
+      "grad_norm": 1.270598619556945,
+      "learning_rate": 0.0001888927191587641,
+      "loss": 13.5671,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5308852459016393,
+      "grad_norm": 1.2272845078375412,
+      "learning_rate": 0.00018888406920684368,
+      "loss": 13.8208,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5310601092896174,
+      "grad_norm": 1.4123035065687197,
+      "learning_rate": 0.00018887541608631004,
+      "loss": 13.6828,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5312349726775957,
+      "grad_norm": 1.235369871811384,
+      "learning_rate": 0.00018886675979747165,
+      "loss": 13.8955,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5314098360655738,
+      "grad_norm": 1.7724927921902702,
+      "learning_rate": 0.0001888581003406371,
+      "loss": 13.6205,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5315846994535519,
+      "grad_norm": 1.0277181872807781,
+      "learning_rate": 0.00018884943771611508,
+      "loss": 13.6827,
+      "step": 3040
+    },
+    {
+      "epoch": 0.53175956284153,
+      "grad_norm": 1.3221905565758667,
+      "learning_rate": 0.00018884077192421445,
+      "loss": 13.9633,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5319344262295082,
+      "grad_norm": 1.2745393286599744,
+      "learning_rate": 0.0001888321029652441,
+      "loss": 13.8935,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5321092896174864,
+      "grad_norm": 1.1563299129999771,
+      "learning_rate": 0.00018882343083951304,
+      "loss": 13.8434,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5322841530054645,
+      "grad_norm": 1.2599909897972257,
+      "learning_rate": 0.00018881475554733047,
+      "loss": 13.6737,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5324590163934426,
+      "grad_norm": 1.303196263784496,
+      "learning_rate": 0.0001888060770890056,
+      "loss": 13.6216,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5326338797814207,
+      "grad_norm": 1.0920717568524128,
+      "learning_rate": 0.0001887973954648479,
+      "loss": 13.4219,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5328087431693989,
+      "grad_norm": 1.2499661299713312,
+      "learning_rate": 0.00018878871067516678,
+      "loss": 13.6223,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5329836065573771,
+      "grad_norm": 1.1548831646473225,
+      "learning_rate": 0.0001887800227202718,
+      "loss": 13.4562,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5331584699453552,
+      "grad_norm": 1.1628765322103447,
+      "learning_rate": 0.0001887713316004728,
+      "loss": 13.7309,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.4024291049674764,
+      "learning_rate": 0.00018876263731607956,
+      "loss": 13.7643,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5335081967213114,
+      "grad_norm": 1.144531530256149,
+      "learning_rate": 0.00018875393986740195,
+      "loss": 13.7748,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5336830601092896,
+      "grad_norm": 1.2548117372388536,
+      "learning_rate": 0.00018874523925475014,
+      "loss": 13.7695,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5338579234972678,
+      "grad_norm": 1.2533587778396391,
+      "learning_rate": 0.0001887365354784342,
+      "loss": 13.7755,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5340327868852459,
+      "grad_norm": 1.1716125832271749,
+      "learning_rate": 0.00018872782853876445,
+      "loss": 13.8702,
+      "step": 3054
+    },
+    {
+      "epoch": 0.534207650273224,
+      "grad_norm": 1.1518011218816093,
+      "learning_rate": 0.0001887191184360513,
+      "loss": 13.2754,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5343825136612022,
+      "grad_norm": 1.2631601439879814,
+      "learning_rate": 0.00018871040517060524,
+      "loss": 13.9643,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5345573770491804,
+      "grad_norm": 1.2151804811245237,
+      "learning_rate": 0.00018870168874273686,
+      "loss": 13.948,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5347322404371585,
+      "grad_norm": 1.2560553878902159,
+      "learning_rate": 0.0001886929691527569,
+      "loss": 13.6429,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5349071038251366,
+      "grad_norm": 1.2113654457478873,
+      "learning_rate": 0.00018868424640097623,
+      "loss": 13.9012,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5350819672131147,
+      "grad_norm": 1.256182466025893,
+      "learning_rate": 0.0001886755204877058,
+      "loss": 13.3819,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5352568306010929,
+      "grad_norm": 1.197402403896745,
+      "learning_rate": 0.00018866679141325666,
+      "loss": 13.7503,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5354316939890711,
+      "grad_norm": 1.1242452038805537,
+      "learning_rate": 0.00018865805917794005,
+      "loss": 13.9085,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5356065573770492,
+      "grad_norm": 1.3159424173461236,
+      "learning_rate": 0.00018864932378206717,
+      "loss": 13.8932,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5357814207650273,
+      "grad_norm": 1.2179696718806694,
+      "learning_rate": 0.00018864058522594947,
+      "loss": 13.8974,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5359562841530054,
+      "grad_norm": 1.3024649512037172,
+      "learning_rate": 0.0001886318435098985,
+      "loss": 13.8422,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5361311475409836,
+      "grad_norm": 1.2293341208856792,
+      "learning_rate": 0.00018862309863422585,
+      "loss": 13.5494,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5363060109289618,
+      "grad_norm": 1.1287900268832836,
+      "learning_rate": 0.00018861435059924325,
+      "loss": 13.801,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5364808743169399,
+      "grad_norm": 1.2065109815230846,
+      "learning_rate": 0.00018860559940526262,
+      "loss": 13.5509,
+      "step": 3068
+    },
+    {
+      "epoch": 0.536655737704918,
+      "grad_norm": 1.2306597913545654,
+      "learning_rate": 0.0001885968450525959,
+      "loss": 13.6578,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5368306010928962,
+      "grad_norm": 1.2531020949639835,
+      "learning_rate": 0.00018858808754155515,
+      "loss": 13.5235,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5370054644808743,
+      "grad_norm": 1.2075952458849035,
+      "learning_rate": 0.00018857932687245258,
+      "loss": 13.5918,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5371803278688525,
+      "grad_norm": 1.2623232518601641,
+      "learning_rate": 0.0001885705630456005,
+      "loss": 13.9361,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5373551912568306,
+      "grad_norm": 1.1656260580968563,
+      "learning_rate": 0.00018856179606131134,
+      "loss": 13.8668,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5375300546448087,
+      "grad_norm": 1.3278802230154976,
+      "learning_rate": 0.00018855302591989763,
+      "loss": 13.655,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5377049180327869,
+      "grad_norm": 1.1202757702029604,
+      "learning_rate": 0.000188544252621672,
+      "loss": 13.6525,
+      "step": 3075
+    },
+    {
+      "epoch": 0.537879781420765,
+      "grad_norm": 1.2707610246773375,
+      "learning_rate": 0.00018853547616694722,
+      "loss": 13.8799,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5380546448087432,
+      "grad_norm": 1.2766003818758584,
+      "learning_rate": 0.00018852669655603616,
+      "loss": 13.966,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5382295081967213,
+      "grad_norm": 1.0661305994685653,
+      "learning_rate": 0.00018851791378925178,
+      "loss": 13.5115,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5384043715846994,
+      "grad_norm": 1.244349986526602,
+      "learning_rate": 0.00018850912786690726,
+      "loss": 13.5494,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5385792349726776,
+      "grad_norm": 1.1107721314019212,
+      "learning_rate": 0.00018850033878931567,
+      "loss": 13.8839,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5387540983606557,
+      "grad_norm": 1.201043525963248,
+      "learning_rate": 0.0001884915465567904,
+      "loss": 13.8022,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5389289617486339,
+      "grad_norm": 1.111636203951517,
+      "learning_rate": 0.00018848275116964492,
+      "loss": 13.6092,
+      "step": 3082
+    },
+    {
+      "epoch": 0.539103825136612,
+      "grad_norm": 1.448460296653779,
+      "learning_rate": 0.00018847395262819276,
+      "loss": 13.7628,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5392786885245902,
+      "grad_norm": 1.1263249931685253,
+      "learning_rate": 0.00018846515093274752,
+      "loss": 13.7235,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5394535519125683,
+      "grad_norm": 1.4016428950910071,
+      "learning_rate": 0.00018845634608362304,
+      "loss": 13.9198,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5396284153005464,
+      "grad_norm": 1.1625789625516596,
+      "learning_rate": 0.00018844753808113315,
+      "loss": 13.9951,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5398032786885246,
+      "grad_norm": 1.38526739043911,
+      "learning_rate": 0.00018843872692559188,
+      "loss": 13.7845,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5399781420765027,
+      "grad_norm": 1.1836111515216527,
+      "learning_rate": 0.00018842991261731334,
+      "loss": 13.8851,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5401530054644809,
+      "grad_norm": 1.2723069673090703,
+      "learning_rate": 0.0001884210951566117,
+      "loss": 13.7715,
+      "step": 3089
+    },
+    {
+      "epoch": 0.540327868852459,
+      "grad_norm": 1.1348786539576186,
+      "learning_rate": 0.00018841227454380135,
+      "loss": 13.6251,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5405027322404372,
+      "grad_norm": 1.4565860614115382,
+      "learning_rate": 0.0001884034507791967,
+      "loss": 13.8167,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5406775956284153,
+      "grad_norm": 1.2446522817344452,
+      "learning_rate": 0.00018839462386311233,
+      "loss": 13.3987,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5408524590163934,
+      "grad_norm": 1.2373480035639102,
+      "learning_rate": 0.0001883857937958629,
+      "loss": 13.4911,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5410273224043716,
+      "grad_norm": 1.447767068838908,
+      "learning_rate": 0.00018837696057776317,
+      "loss": 13.5377,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5412021857923497,
+      "grad_norm": 1.0547118278201317,
+      "learning_rate": 0.00018836812420912807,
+      "loss": 13.6148,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5413770491803279,
+      "grad_norm": 1.5341737799634614,
+      "learning_rate": 0.00018835928469027255,
+      "loss": 13.7783,
+      "step": 3096
+    },
+    {
+      "epoch": 0.541551912568306,
+      "grad_norm": 1.2497755398750594,
+      "learning_rate": 0.0001883504420215118,
+      "loss": 13.9261,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5417267759562842,
+      "grad_norm": 1.2125096790320515,
+      "learning_rate": 0.00018834159620316102,
+      "loss": 13.4238,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5419016393442623,
+      "grad_norm": 1.3730778950112126,
+      "learning_rate": 0.00018833274723553553,
+      "loss": 13.8791,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5420765027322404,
+      "grad_norm": 1.1052728988249139,
+      "learning_rate": 0.00018832389511895086,
+      "loss": 13.9701,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5422513661202186,
+      "grad_norm": 1.2984948127679035,
+      "learning_rate": 0.00018831503985372247,
+      "loss": 13.8549,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5424262295081967,
+      "grad_norm": 1.1735858616054722,
+      "learning_rate": 0.00018830618144016612,
+      "loss": 13.8977,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5426010928961749,
+      "grad_norm": 1.1861979379930683,
+      "learning_rate": 0.00018829731987859754,
+      "loss": 13.7865,
+      "step": 3103
+    },
+    {
+      "epoch": 0.542775956284153,
+      "grad_norm": 1.1055357348607826,
+      "learning_rate": 0.00018828845516933272,
+      "loss": 13.6072,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5429508196721311,
+      "grad_norm": 1.2334753061796393,
+      "learning_rate": 0.0001882795873126876,
+      "loss": 13.7651,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5431256830601093,
+      "grad_norm": 1.1258053037417728,
+      "learning_rate": 0.00018827071630897834,
+      "loss": 13.7295,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5433005464480875,
+      "grad_norm": 1.2556954574720034,
+      "learning_rate": 0.0001882618421585212,
+      "loss": 13.7288,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5434754098360656,
+      "grad_norm": 1.1997804822429867,
+      "learning_rate": 0.00018825296486163248,
+      "loss": 13.6874,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5436502732240437,
+      "grad_norm": 1.2424053014868113,
+      "learning_rate": 0.0001882440844186287,
+      "loss": 14.0409,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5438251366120218,
+      "grad_norm": 1.1070788150741167,
+      "learning_rate": 0.0001882352008298264,
+      "loss": 13.5177,
+      "step": 3110
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 1.1545216590732315,
+      "learning_rate": 0.00018822631409554224,
+      "loss": 13.735,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5441748633879782,
+      "grad_norm": 1.152904462068519,
+      "learning_rate": 0.00018821742421609312,
+      "loss": 13.7436,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5443497267759563,
+      "grad_norm": 1.2046622633576725,
+      "learning_rate": 0.00018820853119179588,
+      "loss": 14.0161,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5445245901639344,
+      "grad_norm": 1.2361488655249484,
+      "learning_rate": 0.00018819963502296756,
+      "loss": 13.6671,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5446994535519125,
+      "grad_norm": 1.3489421184314463,
+      "learning_rate": 0.00018819073570992533,
+      "loss": 13.7651,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5448743169398907,
+      "grad_norm": 1.2607615642528434,
+      "learning_rate": 0.00018818183325298638,
+      "loss": 13.8111,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5450491803278689,
+      "grad_norm": 1.3686446502656056,
+      "learning_rate": 0.0001881729276524681,
+      "loss": 13.8314,
+      "step": 3117
+    },
+    {
+      "epoch": 0.545224043715847,
+      "grad_norm": 1.1131459894619677,
+      "learning_rate": 0.000188164018908688,
+      "loss": 13.8387,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5453989071038251,
+      "grad_norm": 1.3470952554483446,
+      "learning_rate": 0.0001881551070219636,
+      "loss": 13.7282,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5455737704918033,
+      "grad_norm": 1.1338681136559838,
+      "learning_rate": 0.00018814619199261263,
+      "loss": 13.6187,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5457486338797815,
+      "grad_norm": 1.2041942059491892,
+      "learning_rate": 0.00018813727382095295,
+      "loss": 13.7114,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5459234972677596,
+      "grad_norm": 1.489071617950837,
+      "learning_rate": 0.0001881283525073024,
+      "loss": 13.6194,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5460983606557377,
+      "grad_norm": 1.3829154212460286,
+      "learning_rate": 0.00018811942805197905,
+      "loss": 13.7589,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5462732240437158,
+      "grad_norm": 1.5566600592800701,
+      "learning_rate": 0.00018811050045530105,
+      "loss": 13.5632,
+      "step": 3124
+    },
+    {
+      "epoch": 0.546448087431694,
+      "grad_norm": 1.1053893591534034,
+      "learning_rate": 0.00018810156971758664,
+      "loss": 13.7574,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5466229508196722,
+      "grad_norm": 1.603117318848877,
+      "learning_rate": 0.00018809263583915425,
+      "loss": 13.6876,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5467978142076503,
+      "grad_norm": 1.2441219057701782,
+      "learning_rate": 0.0001880836988203223,
+      "loss": 13.7162,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5469726775956284,
+      "grad_norm": 1.1672319066614154,
+      "learning_rate": 0.00018807475866140937,
+      "loss": 13.6385,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5471475409836065,
+      "grad_norm": 1.2221065111067413,
+      "learning_rate": 0.0001880658153627342,
+      "loss": 13.8002,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5473224043715847,
+      "grad_norm": 1.1585010664758035,
+      "learning_rate": 0.00018805686892461564,
+      "loss": 13.4907,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5474972677595629,
+      "grad_norm": 1.3190365648318876,
+      "learning_rate": 0.0001880479193473726,
+      "loss": 13.5447,
+      "step": 3131
+    },
+    {
+      "epoch": 0.547672131147541,
+      "grad_norm": 1.158558915803366,
+      "learning_rate": 0.00018803896663132407,
+      "loss": 13.9854,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5478469945355191,
+      "grad_norm": 1.4363696025086508,
+      "learning_rate": 0.00018803001077678927,
+      "loss": 13.8571,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5480218579234972,
+      "grad_norm": 1.167395998234039,
+      "learning_rate": 0.0001880210517840874,
+      "loss": 13.7641,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5481967213114755,
+      "grad_norm": 1.2074890149160442,
+      "learning_rate": 0.0001880120896535379,
+      "loss": 13.864,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5483715846994536,
+      "grad_norm": 1.1024970830822427,
+      "learning_rate": 0.00018800312438546022,
+      "loss": 13.7042,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5485464480874317,
+      "grad_norm": 1.21560015841336,
+      "learning_rate": 0.00018799415598017398,
+      "loss": 13.4547,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5487213114754098,
+      "grad_norm": 1.2243317055602712,
+      "learning_rate": 0.00018798518443799891,
+      "loss": 13.6136,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5488961748633879,
+      "grad_norm": 1.1351796679915556,
+      "learning_rate": 0.0001879762097592548,
+      "loss": 13.9359,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5490710382513662,
+      "grad_norm": 1.0729444861793345,
+      "learning_rate": 0.0001879672319442616,
+      "loss": 13.7984,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5492459016393443,
+      "grad_norm": 1.1623911458295084,
+      "learning_rate": 0.00018795825099333937,
+      "loss": 13.7022,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5494207650273224,
+      "grad_norm": 1.1583257016836928,
+      "learning_rate": 0.00018794926690680826,
+      "loss": 13.6576,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5495956284153005,
+      "grad_norm": 1.2327884902237392,
+      "learning_rate": 0.00018794027968498851,
+      "loss": 13.7111,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5497704918032786,
+      "grad_norm": 1.3118790608619282,
+      "learning_rate": 0.0001879312893282006,
+      "loss": 13.904,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5499453551912569,
+      "grad_norm": 1.1343140753580758,
+      "learning_rate": 0.0001879222958367649,
+      "loss": 13.7539,
+      "step": 3145
+    },
+    {
+      "epoch": 0.550120218579235,
+      "grad_norm": 1.324744317609858,
+      "learning_rate": 0.00018791329921100211,
+      "loss": 13.6028,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5502950819672131,
+      "grad_norm": 1.1079269000676684,
+      "learning_rate": 0.0001879042994512329,
+      "loss": 13.9182,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5504699453551912,
+      "grad_norm": 1.2024903278270411,
+      "learning_rate": 0.00018789529655777813,
+      "loss": 13.8567,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5506448087431693,
+      "grad_norm": 1.1480264445182846,
+      "learning_rate": 0.00018788629053095873,
+      "loss": 13.6432,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5508196721311476,
+      "grad_norm": 1.1467087910226226,
+      "learning_rate": 0.00018787728137109578,
+      "loss": 13.8129,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5509945355191257,
+      "grad_norm": 1.1521704679093687,
+      "learning_rate": 0.0001878682690785104,
+      "loss": 13.4941,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5511693989071038,
+      "grad_norm": 1.397662886512745,
+      "learning_rate": 0.00018785925365352388,
+      "loss": 13.6287,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5513442622950819,
+      "grad_norm": 1.1305823920169407,
+      "learning_rate": 0.00018785023509645764,
+      "loss": 13.5304,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5515191256830602,
+      "grad_norm": 1.3539835414265802,
+      "learning_rate": 0.00018784121340763315,
+      "loss": 13.4411,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5516939890710383,
+      "grad_norm": 1.4542957960392946,
+      "learning_rate": 0.00018783218858737204,
+      "loss": 13.7529,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5518688524590164,
+      "grad_norm": 1.1169374518101292,
+      "learning_rate": 0.00018782316063599604,
+      "loss": 13.7694,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5520437158469945,
+      "grad_norm": 1.5985835839134996,
+      "learning_rate": 0.00018781412955382694,
+      "loss": 13.6139,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5522185792349726,
+      "grad_norm": 1.1677691039154516,
+      "learning_rate": 0.00018780509534118675,
+      "loss": 13.9747,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5523934426229509,
+      "grad_norm": 1.4515651860050967,
+      "learning_rate": 0.0001877960579983975,
+      "loss": 13.8282,
+      "step": 3159
+    },
+    {
+      "epoch": 0.552568306010929,
+      "grad_norm": 1.1129240557512103,
+      "learning_rate": 0.00018778701752578133,
+      "loss": 13.7192,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5527431693989071,
+      "grad_norm": 1.5364578500066117,
+      "learning_rate": 0.00018777797392366058,
+      "loss": 13.4123,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5529180327868852,
+      "grad_norm": 1.1442465848769103,
+      "learning_rate": 0.0001877689271923576,
+      "loss": 13.6299,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5530928961748633,
+      "grad_norm": 1.4882930060565067,
+      "learning_rate": 0.00018775987733219495,
+      "loss": 13.8041,
+      "step": 3163
+    },
+    {
+      "epoch": 0.5532677595628416,
+      "grad_norm": 1.0888417869034321,
+      "learning_rate": 0.00018775082434349517,
+      "loss": 13.7719,
+      "step": 3164
+    },
+    {
+      "epoch": 0.5534426229508197,
+      "grad_norm": 1.5954842301326597,
+      "learning_rate": 0.00018774176822658104,
+      "loss": 13.8549,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5536174863387978,
+      "grad_norm": 1.1945339258638583,
+      "learning_rate": 0.00018773270898177539,
+      "loss": 13.6122,
+      "step": 3166
+    },
+    {
+      "epoch": 0.5537923497267759,
+      "grad_norm": 1.2435381392606097,
+      "learning_rate": 0.0001877236466094012,
+      "loss": 13.8668,
+      "step": 3167
+    },
+    {
+      "epoch": 0.553967213114754,
+      "grad_norm": 1.1469594719130833,
+      "learning_rate": 0.00018771458110978143,
+      "loss": 13.6371,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5541420765027323,
+      "grad_norm": 1.1847421761966765,
+      "learning_rate": 0.0001877055124832394,
+      "loss": 13.8408,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5543169398907104,
+      "grad_norm": 1.065175389268738,
+      "learning_rate": 0.00018769644073009828,
+      "loss": 13.4176,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5544918032786885,
+      "grad_norm": 1.2397704703906187,
+      "learning_rate": 0.00018768736585068153,
+      "loss": 13.7613,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5546666666666666,
+      "grad_norm": 1.1700131800445739,
+      "learning_rate": 0.00018767828784531262,
+      "loss": 13.5697,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5548415300546448,
+      "grad_norm": 1.2942088864961037,
+      "learning_rate": 0.00018766920671431522,
+      "loss": 13.5966,
+      "step": 3173
+    },
+    {
+      "epoch": 0.555016393442623,
+      "grad_norm": 1.3591512955557021,
+      "learning_rate": 0.000187660122458013,
+      "loss": 13.8873,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5551912568306011,
+      "grad_norm": 1.3707056613280215,
+      "learning_rate": 0.00018765103507672985,
+      "loss": 13.7077,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5553661202185792,
+      "grad_norm": 1.2877031721898924,
+      "learning_rate": 0.0001876419445707897,
+      "loss": 13.5568,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5555409836065573,
+      "grad_norm": 1.106384226197071,
+      "learning_rate": 0.0001876328509405166,
+      "loss": 13.383,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5557158469945355,
+      "grad_norm": 1.2333080190672552,
+      "learning_rate": 0.0001876237541862348,
+      "loss": 13.7639,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5558907103825137,
+      "grad_norm": 1.3264229665918148,
+      "learning_rate": 0.00018761465430826855,
+      "loss": 13.8365,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5560655737704918,
+      "grad_norm": 1.3413171424508368,
+      "learning_rate": 0.00018760555130694219,
+      "loss": 13.6122,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5562404371584699,
+      "grad_norm": 1.2894238811059864,
+      "learning_rate": 0.0001875964451825803,
+      "loss": 13.4991,
+      "step": 3181
+    },
+    {
+      "epoch": 0.556415300546448,
+      "grad_norm": 1.1736804758130284,
+      "learning_rate": 0.0001875873359355075,
+      "loss": 13.7006,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5565901639344262,
+      "grad_norm": 1.3358225329808804,
+      "learning_rate": 0.0001875782235660485,
+      "loss": 13.8717,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5567650273224044,
+      "grad_norm": 1.322134362124205,
+      "learning_rate": 0.00018756910807452815,
+      "loss": 13.7952,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5569398907103825,
+      "grad_norm": 1.6083035869638675,
+      "learning_rate": 0.0001875599894612714,
+      "loss": 13.8682,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5571147540983606,
+      "grad_norm": 1.2937124190412215,
+      "learning_rate": 0.00018755086772660336,
+      "loss": 13.7321,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5572896174863388,
+      "grad_norm": 1.202863194492976,
+      "learning_rate": 0.00018754174287084914,
+      "loss": 13.7073,
+      "step": 3187
+    },
+    {
+      "epoch": 0.557464480874317,
+      "grad_norm": 1.5479809253014278,
+      "learning_rate": 0.00018753261489433412,
+      "loss": 13.546,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5576393442622951,
+      "grad_norm": 1.1330163275297682,
+      "learning_rate": 0.0001875234837973836,
+      "loss": 13.534,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5578142076502732,
+      "grad_norm": 1.2978797858054607,
+      "learning_rate": 0.00018751434958032314,
+      "loss": 13.6357,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5579890710382513,
+      "grad_norm": 1.2421748246025976,
+      "learning_rate": 0.00018750521224347838,
+      "loss": 13.6749,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5581639344262295,
+      "grad_norm": 1.250142886780825,
+      "learning_rate": 0.00018749607178717508,
+      "loss": 13.7759,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5583387978142077,
+      "grad_norm": 1.4508524145697779,
+      "learning_rate": 0.00018748692821173902,
+      "loss": 13.9224,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5585136612021858,
+      "grad_norm": 1.0173505395845732,
+      "learning_rate": 0.00018747778151749617,
+      "loss": 13.5385,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5586885245901639,
+      "grad_norm": 1.2764847644648494,
+      "learning_rate": 0.00018746863170477265,
+      "loss": 13.7681,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5588633879781421,
+      "grad_norm": 1.2418307036371061,
+      "learning_rate": 0.0001874594787738946,
+      "loss": 13.6567,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5590382513661202,
+      "grad_norm": 1.5795633208775033,
+      "learning_rate": 0.00018745032272518826,
+      "loss": 13.5551,
+      "step": 3197
+    },
+    {
+      "epoch": 0.5592131147540984,
+      "grad_norm": 1.466398406933894,
+      "learning_rate": 0.00018744116355898016,
+      "loss": 13.339,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5593879781420765,
+      "grad_norm": 1.2255574291306202,
+      "learning_rate": 0.0001874320012755967,
+      "loss": 13.7888,
+      "step": 3199
+    },
+    {
+      "epoch": 0.5595628415300546,
+      "grad_norm": 1.3583276491271046,
+      "learning_rate": 0.00018742283587536458,
+      "loss": 13.6803,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5597377049180328,
+      "grad_norm": 1.0897897451537029,
+      "learning_rate": 0.00018741366735861047,
+      "loss": 13.3327,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5599125683060109,
+      "grad_norm": 1.3337817427379772,
+      "learning_rate": 0.00018740449572566127,
+      "loss": 13.6846,
+      "step": 3202
+    },
+    {
+      "epoch": 0.5600874316939891,
+      "grad_norm": 1.1714836988156505,
+      "learning_rate": 0.00018739532097684393,
+      "loss": 13.425,
+      "step": 3203
+    },
+    {
+      "epoch": 0.5602622950819672,
+      "grad_norm": 1.5385693020749938,
+      "learning_rate": 0.00018738614311248548,
+      "loss": 13.5832,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5604371584699454,
+      "grad_norm": 1.280708457474151,
+      "learning_rate": 0.00018737696213291317,
+      "loss": 13.7592,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5606120218579235,
+      "grad_norm": 1.5573530361690837,
+      "learning_rate": 0.00018736777803845422,
+      "loss": 13.9149,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5607868852459016,
+      "grad_norm": 1.287188969719002,
+      "learning_rate": 0.00018735859082943607,
+      "loss": 13.7136,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5609617486338798,
+      "grad_norm": 1.4792261304846532,
+      "learning_rate": 0.00018734940050618622,
+      "loss": 13.7687,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5611366120218579,
+      "grad_norm": 1.4171199594635007,
+      "learning_rate": 0.00018734020706903228,
+      "loss": 13.646,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5613114754098361,
+      "grad_norm": 1.3128667019931928,
+      "learning_rate": 0.00018733101051830202,
+      "loss": 13.7535,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5614863387978142,
+      "grad_norm": 1.3412818461940144,
+      "learning_rate": 0.00018732181085432328,
+      "loss": 13.7607,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5616612021857923,
+      "grad_norm": 1.1388198927301085,
+      "learning_rate": 0.000187312608077424,
+      "loss": 13.606,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5618360655737705,
+      "grad_norm": 1.2495924095265925,
+      "learning_rate": 0.00018730340218793226,
+      "loss": 13.5825,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5620109289617486,
+      "grad_norm": 1.1040650476445155,
+      "learning_rate": 0.00018729419318617622,
+      "loss": 13.4669,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5621857923497268,
+      "grad_norm": 1.1847797594411824,
+      "learning_rate": 0.00018728498107248418,
+      "loss": 13.5621,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5623606557377049,
+      "grad_norm": 1.131551217214803,
+      "learning_rate": 0.00018727576584718456,
+      "loss": 13.6659,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5625355191256831,
+      "grad_norm": 1.2010769570043125,
+      "learning_rate": 0.00018726654751060587,
+      "loss": 13.7379,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5627103825136612,
+      "grad_norm": 1.1195776510146982,
+      "learning_rate": 0.0001872573260630767,
+      "loss": 13.6329,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5628852459016394,
+      "grad_norm": 1.2845776240610187,
+      "learning_rate": 0.00018724810150492582,
+      "loss": 13.6742,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5630601092896175,
+      "grad_norm": 1.2826174091838747,
+      "learning_rate": 0.00018723887383648204,
+      "loss": 13.7762,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5632349726775956,
+      "grad_norm": 1.3653688028275326,
+      "learning_rate": 0.00018722964305807437,
+      "loss": 13.6718,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5634098360655738,
+      "grad_norm": 1.2270535849345123,
+      "learning_rate": 0.0001872204091700318,
+      "loss": 13.5432,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5635846994535519,
+      "grad_norm": 1.1613165968763355,
+      "learning_rate": 0.00018721117217268362,
+      "loss": 13.7546,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5637595628415301,
+      "grad_norm": 1.2898113164606886,
+      "learning_rate": 0.00018720193206635898,
+      "loss": 13.7115,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5639344262295082,
+      "grad_norm": 1.32240231364783,
+      "learning_rate": 0.00018719268885138734,
+      "loss": 13.8669,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5641092896174863,
+      "grad_norm": 1.3696834072948973,
+      "learning_rate": 0.00018718344252809826,
+      "loss": 13.2815,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5642841530054645,
+      "grad_norm": 1.1391313254229998,
+      "learning_rate": 0.0001871741930968213,
+      "loss": 13.8008,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5644590163934426,
+      "grad_norm": 1.1240917480189665,
+      "learning_rate": 0.0001871649405578862,
+      "loss": 13.4689,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5646338797814208,
+      "grad_norm": 1.2355211915968234,
+      "learning_rate": 0.0001871556849116228,
+      "loss": 13.7522,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5648087431693989,
+      "grad_norm": 1.1970318016422548,
+      "learning_rate": 0.00018714642615836108,
+      "loss": 13.6552,
+      "step": 3230
+    },
+    {
+      "epoch": 0.564983606557377,
+      "grad_norm": 1.1502493827384448,
+      "learning_rate": 0.00018713716429843112,
+      "loss": 13.3371,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5651584699453552,
+      "grad_norm": 1.357752376333775,
+      "learning_rate": 0.000187127899332163,
+      "loss": 13.6472,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5653333333333334,
+      "grad_norm": 1.244815626954618,
+      "learning_rate": 0.00018711863125988712,
+      "loss": 13.4906,
+      "step": 3233
+    },
+    {
+      "epoch": 0.5655081967213115,
+      "grad_norm": 1.6276777117388235,
+      "learning_rate": 0.0001871093600819338,
+      "loss": 13.5481,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5656830601092896,
+      "grad_norm": 1.4097382679767136,
+      "learning_rate": 0.00018710008579863357,
+      "loss": 13.9013,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5658579234972677,
+      "grad_norm": 1.3846091842391257,
+      "learning_rate": 0.00018709080841031704,
+      "loss": 13.4651,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5660327868852459,
+      "grad_norm": 1.3529839204806942,
+      "learning_rate": 0.00018708152791731495,
+      "loss": 13.8476,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5662076502732241,
+      "grad_norm": 1.1360611136580752,
+      "learning_rate": 0.00018707224431995813,
+      "loss": 13.6254,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5663825136612022,
+      "grad_norm": 1.3898605677782114,
+      "learning_rate": 0.00018706295761857753,
+      "loss": 13.5298,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5665573770491803,
+      "grad_norm": 1.0596971226514622,
+      "learning_rate": 0.00018705366781350422,
+      "loss": 13.6352,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5667322404371584,
+      "grad_norm": 1.399036477854296,
+      "learning_rate": 0.00018704437490506937,
+      "loss": 13.7621,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5669071038251366,
+      "grad_norm": 1.1440582615554877,
+      "learning_rate": 0.00018703507889360424,
+      "loss": 13.765,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5670819672131148,
+      "grad_norm": 1.4261708369072,
+      "learning_rate": 0.00018702577977944024,
+      "loss": 13.63,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5672568306010929,
+      "grad_norm": 1.2361954951262102,
+      "learning_rate": 0.0001870164775629089,
+      "loss": 13.5868,
+      "step": 3244
+    },
+    {
+      "epoch": 0.567431693989071,
+      "grad_norm": 1.246652847117018,
+      "learning_rate": 0.00018700717224434175,
+      "loss": 13.416,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5676065573770491,
+      "grad_norm": 1.280878536872333,
+      "learning_rate": 0.0001869978638240706,
+      "loss": 13.8791,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5677814207650274,
+      "grad_norm": 1.2408218359958947,
+      "learning_rate": 0.00018698855230242726,
+      "loss": 13.6524,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5679562841530055,
+      "grad_norm": 1.2388941333273307,
+      "learning_rate": 0.00018697923767974363,
+      "loss": 13.512,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5681311475409836,
+      "grad_norm": 1.2438323343424003,
+      "learning_rate": 0.00018696991995635184,
+      "loss": 13.5711,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5683060109289617,
+      "grad_norm": 1.3706945722863966,
+      "learning_rate": 0.000186960599132584,
+      "loss": 13.75,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5684808743169399,
+      "grad_norm": 1.1019114208683911,
+      "learning_rate": 0.0001869512752087724,
+      "loss": 13.793,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5686557377049181,
+      "grad_norm": 1.3650994060049184,
+      "learning_rate": 0.0001869419481852494,
+      "loss": 13.5287,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5688306010928962,
+      "grad_norm": 1.1416912106698454,
+      "learning_rate": 0.0001869326180623476,
+      "loss": 13.5599,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5690054644808743,
+      "grad_norm": 1.4203579245981808,
+      "learning_rate": 0.0001869232848403995,
+      "loss": 13.6121,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5691803278688524,
+      "grad_norm": 1.284543697333458,
+      "learning_rate": 0.00018691394851973786,
+      "loss": 13.7399,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5693551912568307,
+      "grad_norm": 1.3164475608850037,
+      "learning_rate": 0.00018690460910069547,
+      "loss": 13.9075,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5695300546448088,
+      "grad_norm": 1.4576024243741856,
+      "learning_rate": 0.00018689526658360533,
+      "loss": 13.8181,
+      "step": 3257
+    },
+    {
+      "epoch": 0.5697049180327869,
+      "grad_norm": 1.3525725581890748,
+      "learning_rate": 0.00018688592096880043,
+      "loss": 13.7881,
+      "step": 3258
+    },
+    {
+      "epoch": 0.569879781420765,
+      "grad_norm": 1.2152749762061041,
+      "learning_rate": 0.00018687657225661402,
+      "loss": 13.7247,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5700546448087431,
+      "grad_norm": 1.164471148677243,
+      "learning_rate": 0.00018686722044737928,
+      "loss": 13.4141,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5702295081967214,
+      "grad_norm": 1.3451119603454018,
+      "learning_rate": 0.00018685786554142964,
+      "loss": 13.681,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5704043715846995,
+      "grad_norm": 1.0859183823452971,
+      "learning_rate": 0.00018684850753909854,
+      "loss": 13.7469,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5705792349726776,
+      "grad_norm": 1.1239412607650912,
+      "learning_rate": 0.00018683914644071966,
+      "loss": 13.5751,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5707540983606557,
+      "grad_norm": 1.21994856626705,
+      "learning_rate": 0.00018682978224662664,
+      "loss": 13.6052,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5709289617486338,
+      "grad_norm": 1.2494087180660551,
+      "learning_rate": 0.00018682041495715333,
+      "loss": 13.6636,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5711038251366121,
+      "grad_norm": 1.0324738151687674,
+      "learning_rate": 0.0001868110445726337,
+      "loss": 13.7293,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5712786885245902,
+      "grad_norm": 1.1150294057836758,
+      "learning_rate": 0.00018680167109340175,
+      "loss": 13.9179,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5714535519125683,
+      "grad_norm": 1.1873445922606845,
+      "learning_rate": 0.00018679229451979164,
+      "loss": 13.4828,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5716284153005464,
+      "grad_norm": 1.17175407425156,
+      "learning_rate": 0.00018678291485213762,
+      "loss": 13.6041,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5718032786885245,
+      "grad_norm": 1.1367746687187246,
+      "learning_rate": 0.0001867735320907741,
+      "loss": 13.6557,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5719781420765028,
+      "grad_norm": 1.1280514436502602,
+      "learning_rate": 0.00018676414623603553,
+      "loss": 13.566,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5721530054644809,
+      "grad_norm": 1.0974538571732908,
+      "learning_rate": 0.0001867547572882565,
+      "loss": 13.6393,
+      "step": 3272
+    },
+    {
+      "epoch": 0.572327868852459,
+      "grad_norm": 1.1574780544126335,
+      "learning_rate": 0.00018674536524777177,
+      "loss": 13.4409,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5725027322404371,
+      "grad_norm": 1.0365154428658514,
+      "learning_rate": 0.00018673597011491613,
+      "loss": 13.5191,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5726775956284152,
+      "grad_norm": 0.9955278064842747,
+      "learning_rate": 0.00018672657189002447,
+      "loss": 13.5891,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5728524590163935,
+      "grad_norm": 1.2261449727758442,
+      "learning_rate": 0.00018671717057343185,
+      "loss": 13.4792,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5730273224043716,
+      "grad_norm": 1.041498959307052,
+      "learning_rate": 0.00018670776616547342,
+      "loss": 13.382,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5732021857923497,
+      "grad_norm": 1.189646992682826,
+      "learning_rate": 0.00018669835866648443,
+      "loss": 13.6984,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5733770491803278,
+      "grad_norm": 1.2125784295275244,
+      "learning_rate": 0.0001866889480768003,
+      "loss": 13.5838,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5735519125683061,
+      "grad_norm": 1.152358891088499,
+      "learning_rate": 0.00018667953439675637,
+      "loss": 13.4342,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5737267759562842,
+      "grad_norm": 1.1895872357151644,
+      "learning_rate": 0.0001866701176266884,
+      "loss": 13.7093,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5739016393442623,
+      "grad_norm": 1.1057544524356397,
+      "learning_rate": 0.00018666069776693194,
+      "loss": 13.9575,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5740765027322404,
+      "grad_norm": 1.1559840372030425,
+      "learning_rate": 0.00018665127481782285,
+      "loss": 13.6151,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5742513661202185,
+      "grad_norm": 1.3275369927233527,
+      "learning_rate": 0.00018664184877969708,
+      "loss": 13.7602,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5744262295081968,
+      "grad_norm": 1.0646529051445657,
+      "learning_rate": 0.00018663241965289061,
+      "loss": 13.2978,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5746010928961749,
+      "grad_norm": 1.5495995383316026,
+      "learning_rate": 0.00018662298743773962,
+      "loss": 13.5858,
+      "step": 3286
+    },
+    {
+      "epoch": 0.574775956284153,
+      "grad_norm": 1.1537687209699699,
+      "learning_rate": 0.00018661355213458035,
+      "loss": 13.4691,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5749508196721311,
+      "grad_norm": 1.3057044035981014,
+      "learning_rate": 0.00018660411374374914,
+      "loss": 13.4726,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5751256830601092,
+      "grad_norm": 1.3078998848601386,
+      "learning_rate": 0.00018659467226558242,
+      "loss": 13.8432,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5753005464480875,
+      "grad_norm": 1.2072805108145155,
+      "learning_rate": 0.00018658522770041683,
+      "loss": 13.6569,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5754754098360656,
+      "grad_norm": 1.2064258173884983,
+      "learning_rate": 0.00018657578004858903,
+      "loss": 13.6633,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5756502732240437,
+      "grad_norm": 1.2611907966523688,
+      "learning_rate": 0.00018656632931043584,
+      "loss": 13.5153,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5758251366120218,
+      "grad_norm": 1.2271420639969794,
+      "learning_rate": 0.00018655687548629416,
+      "loss": 13.4729,
+      "step": 3293
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.1796170484663637,
+      "learning_rate": 0.000186547418576501,
+      "loss": 13.7384,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5761748633879782,
+      "grad_norm": 1.1740798512970216,
+      "learning_rate": 0.00018653795858139348,
+      "loss": 13.5803,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5763497267759563,
+      "grad_norm": 1.2095542324992181,
+      "learning_rate": 0.0001865284955013088,
+      "loss": 13.5394,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5765245901639344,
+      "grad_norm": 1.285466754319837,
+      "learning_rate": 0.00018651902933658444,
+      "loss": 13.8087,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5766994535519125,
+      "grad_norm": 1.0576106295524625,
+      "learning_rate": 0.0001865095600875577,
+      "loss": 13.5627,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5768743169398907,
+      "grad_norm": 1.1280747207123096,
+      "learning_rate": 0.00018650008775456623,
+      "loss": 13.5863,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5770491803278689,
+      "grad_norm": 1.1020006437451926,
+      "learning_rate": 0.00018649061233794775,
+      "loss": 13.563,
+      "step": 3300
+    },
+    {
+      "epoch": 0.577224043715847,
+      "grad_norm": 1.1571204072561136,
+      "learning_rate": 0.00018648113383803994,
+      "loss": 13.9519,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5773989071038251,
+      "grad_norm": 1.135040099814093,
+      "learning_rate": 0.00018647165225518073,
+      "loss": 13.5525,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5775737704918033,
+      "grad_norm": 1.1795696462484475,
+      "learning_rate": 0.0001864621675897082,
+      "loss": 13.6996,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5777486338797814,
+      "grad_norm": 1.154617847472943,
+      "learning_rate": 0.0001864526798419604,
+      "loss": 13.7127,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5779234972677596,
+      "grad_norm": 1.2267365357331996,
+      "learning_rate": 0.00018644318901227556,
+      "loss": 13.816,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5780983606557377,
+      "grad_norm": 1.0891689746315016,
+      "learning_rate": 0.00018643369510099205,
+      "loss": 13.6958,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5782732240437158,
+      "grad_norm": 1.3643928748461573,
+      "learning_rate": 0.0001864241981084483,
+      "loss": 13.8576,
+      "step": 3307
+    },
+    {
+      "epoch": 0.578448087431694,
+      "grad_norm": 1.1667756701285175,
+      "learning_rate": 0.00018641469803498282,
+      "loss": 13.6957,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5786229508196721,
+      "grad_norm": 1.1240216446538809,
+      "learning_rate": 0.00018640519488093436,
+      "loss": 13.6346,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5787978142076503,
+      "grad_norm": 1.1191813801057022,
+      "learning_rate": 0.00018639568864664164,
+      "loss": 13.8805,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5789726775956284,
+      "grad_norm": 1.1118958209117218,
+      "learning_rate": 0.00018638617933244356,
+      "loss": 13.4591,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5791475409836065,
+      "grad_norm": 1.298821045869589,
+      "learning_rate": 0.00018637666693867915,
+      "loss": 13.526,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5793224043715847,
+      "grad_norm": 1.1759931081657973,
+      "learning_rate": 0.00018636715146568743,
+      "loss": 13.64,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5794972677595629,
+      "grad_norm": 1.216731507201774,
+      "learning_rate": 0.0001863576329138077,
+      "loss": 13.7887,
+      "step": 3314
+    },
+    {
+      "epoch": 0.579672131147541,
+      "grad_norm": 1.292782938914617,
+      "learning_rate": 0.00018634811128337927,
+      "loss": 13.7702,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5798469945355191,
+      "grad_norm": 1.1204745715682762,
+      "learning_rate": 0.00018633858657474153,
+      "loss": 13.7659,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5800218579234973,
+      "grad_norm": 1.219995564283885,
+      "learning_rate": 0.00018632905878823407,
+      "loss": 13.646,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5801967213114754,
+      "grad_norm": 1.2599338603100707,
+      "learning_rate": 0.00018631952792419655,
+      "loss": 13.8358,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5803715846994536,
+      "grad_norm": 1.1963108614216957,
+      "learning_rate": 0.00018630999398296867,
+      "loss": 13.5891,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5805464480874317,
+      "grad_norm": 1.1952667864123192,
+      "learning_rate": 0.00018630045696489038,
+      "loss": 13.7816,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5807213114754098,
+      "grad_norm": 1.2304284582105716,
+      "learning_rate": 0.0001862909168703016,
+      "loss": 13.7488,
+      "step": 3321
+    },
+    {
+      "epoch": 0.580896174863388,
+      "grad_norm": 1.239031184658597,
+      "learning_rate": 0.00018628137369954248,
+      "loss": 13.6153,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5810710382513661,
+      "grad_norm": 1.1799365146308676,
+      "learning_rate": 0.00018627182745295317,
+      "loss": 13.6517,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5812459016393443,
+      "grad_norm": 1.2215600110931715,
+      "learning_rate": 0.00018626227813087404,
+      "loss": 13.498,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5814207650273224,
+      "grad_norm": 1.1619789057316185,
+      "learning_rate": 0.00018625272573364544,
+      "loss": 13.8602,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5815956284153005,
+      "grad_norm": 1.359172216063871,
+      "learning_rate": 0.000186243170261608,
+      "loss": 13.5056,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5817704918032787,
+      "grad_norm": 1.1719245564380827,
+      "learning_rate": 0.00018623361171510227,
+      "loss": 13.3534,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5819453551912568,
+      "grad_norm": 1.3421486301385386,
+      "learning_rate": 0.00018622405009446907,
+      "loss": 13.555,
+      "step": 3328
+    },
+    {
+      "epoch": 0.582120218579235,
+      "grad_norm": 1.15126308103428,
+      "learning_rate": 0.00018621448540004918,
+      "loss": 13.7725,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5822950819672131,
+      "grad_norm": 1.4060327571664029,
+      "learning_rate": 0.00018620491763218363,
+      "loss": 13.6849,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5824699453551913,
+      "grad_norm": 1.177925147060337,
+      "learning_rate": 0.00018619534679121348,
+      "loss": 13.6929,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5826448087431694,
+      "grad_norm": 1.2361410733400036,
+      "learning_rate": 0.00018618577287747993,
+      "loss": 13.7937,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5828196721311475,
+      "grad_norm": 1.211441667100556,
+      "learning_rate": 0.0001861761958913243,
+      "loss": 13.3754,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5829945355191257,
+      "grad_norm": 1.3552646582147747,
+      "learning_rate": 0.00018616661583308795,
+      "loss": 13.5881,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5831693989071038,
+      "grad_norm": 1.089299895070179,
+      "learning_rate": 0.0001861570327031124,
+      "loss": 13.7258,
+      "step": 3335
+    },
+    {
+      "epoch": 0.583344262295082,
+      "grad_norm": 1.5730679665981862,
+      "learning_rate": 0.00018614744650173934,
+      "loss": 13.4515,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5835191256830601,
+      "grad_norm": 1.4576032337512446,
+      "learning_rate": 0.00018613785722931044,
+      "loss": 13.7792,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5836939890710382,
+      "grad_norm": 1.220859407864607,
+      "learning_rate": 0.00018612826488616756,
+      "loss": 13.5018,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5838688524590164,
+      "grad_norm": 1.4574633252686797,
+      "learning_rate": 0.0001861186694726527,
+      "loss": 13.4768,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5840437158469945,
+      "grad_norm": 1.3160535644208569,
+      "learning_rate": 0.00018610907098910785,
+      "loss": 13.7266,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5842185792349727,
+      "grad_norm": 1.2131503953282217,
+      "learning_rate": 0.00018609946943587527,
+      "loss": 13.6014,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5843934426229508,
+      "grad_norm": 1.1305851342038677,
+      "learning_rate": 0.0001860898648132972,
+      "loss": 13.534,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5845683060109289,
+      "grad_norm": 1.3306912808670175,
+      "learning_rate": 0.000186080257121716,
+      "loss": 13.2243,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5847431693989071,
+      "grad_norm": 1.130496023762683,
+      "learning_rate": 0.00018607064636147423,
+      "loss": 13.7005,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5849180327868853,
+      "grad_norm": 1.2858974344764933,
+      "learning_rate": 0.00018606103253291444,
+      "loss": 13.7989,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5850928961748634,
+      "grad_norm": 1.1535898112296297,
+      "learning_rate": 0.00018605141563637946,
+      "loss": 13.3871,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5852677595628415,
+      "grad_norm": 1.1209676619406967,
+      "learning_rate": 0.000186041795672212,
+      "loss": 13.5467,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5854426229508197,
+      "grad_norm": 1.1688557432956987,
+      "learning_rate": 0.00018603217264075508,
+      "loss": 13.6956,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5856174863387978,
+      "grad_norm": 1.2508861028879055,
+      "learning_rate": 0.00018602254654235173,
+      "loss": 13.5973,
+      "step": 3349
+    },
+    {
+      "epoch": 0.585792349726776,
+      "grad_norm": 1.0033409166577905,
+      "learning_rate": 0.00018601291737734506,
+      "loss": 13.4788,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5859672131147541,
+      "grad_norm": 1.3928634918635816,
+      "learning_rate": 0.00018600328514607842,
+      "loss": 13.8212,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5861420765027322,
+      "grad_norm": 1.063309740091956,
+      "learning_rate": 0.00018599364984889515,
+      "loss": 13.6619,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5863169398907104,
+      "grad_norm": 1.2383333241058794,
+      "learning_rate": 0.00018598401148613871,
+      "loss": 13.5828,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5864918032786886,
+      "grad_norm": 1.1816919939568726,
+      "learning_rate": 0.00018597437005815276,
+      "loss": 13.6763,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 1.424090610022146,
+      "learning_rate": 0.0001859647255652809,
+      "loss": 13.6416,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5868415300546448,
+      "grad_norm": 1.2328691754184156,
+      "learning_rate": 0.00018595507800786706,
+      "loss": 13.4893,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5870163934426229,
+      "grad_norm": 1.335994258493678,
+      "learning_rate": 0.00018594542738625513,
+      "loss": 13.5382,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5871912568306011,
+      "grad_norm": 1.4550826419076994,
+      "learning_rate": 0.0001859357737007891,
+      "loss": 13.6962,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5873661202185793,
+      "grad_norm": 1.132743409587505,
+      "learning_rate": 0.00018592611695181316,
+      "loss": 13.7879,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5875409836065574,
+      "grad_norm": 1.4441982163324687,
+      "learning_rate": 0.00018591645713967155,
+      "loss": 13.6113,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5877158469945355,
+      "grad_norm": 1.1667700190557362,
+      "learning_rate": 0.00018590679426470861,
+      "loss": 13.2595,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5878907103825136,
+      "grad_norm": 1.2414683951984455,
+      "learning_rate": 0.00018589712832726884,
+      "loss": 13.799,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5880655737704918,
+      "grad_norm": 1.0886670932446814,
+      "learning_rate": 0.00018588745932769679,
+      "loss": 13.5558,
+      "step": 3363
+    },
+    {
+      "epoch": 0.58824043715847,
+      "grad_norm": 1.1894559658679518,
+      "learning_rate": 0.00018587778726633717,
+      "loss": 13.4538,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5884153005464481,
+      "grad_norm": 1.1525908038836081,
+      "learning_rate": 0.00018586811214353478,
+      "loss": 13.806,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5885901639344262,
+      "grad_norm": 1.108603442272093,
+      "learning_rate": 0.00018585843395963448,
+      "loss": 13.3808,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5887650273224043,
+      "grad_norm": 1.1019160016186655,
+      "learning_rate": 0.00018584875271498137,
+      "loss": 13.5488,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5889398907103826,
+      "grad_norm": 1.1848271624099973,
+      "learning_rate": 0.0001858390684099205,
+      "loss": 13.5558,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5891147540983607,
+      "grad_norm": 1.1517565978387565,
+      "learning_rate": 0.00018582938104479718,
+      "loss": 13.6627,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5892896174863388,
+      "grad_norm": 1.2530122243443416,
+      "learning_rate": 0.00018581969061995667,
+      "loss": 13.5791,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5894644808743169,
+      "grad_norm": 1.156320332460005,
+      "learning_rate": 0.00018580999713574444,
+      "loss": 13.5233,
+      "step": 3371
+    },
+    {
+      "epoch": 0.589639344262295,
+      "grad_norm": 1.08320902552286,
+      "learning_rate": 0.00018580030059250612,
+      "loss": 13.7317,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5898142076502733,
+      "grad_norm": 1.0711239205347114,
+      "learning_rate": 0.0001857906009905873,
+      "loss": 13.4653,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5899890710382514,
+      "grad_norm": 1.2500856886978928,
+      "learning_rate": 0.00018578089833033379,
+      "loss": 13.67,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5901639344262295,
+      "grad_norm": 1.3074952244380984,
+      "learning_rate": 0.00018577119261209147,
+      "loss": 13.4725,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5903387978142076,
+      "grad_norm": 1.1007369136775904,
+      "learning_rate": 0.00018576148383620634,
+      "loss": 13.5877,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5905136612021858,
+      "grad_norm": 1.2469518901905956,
+      "learning_rate": 0.00018575177200302453,
+      "loss": 13.4266,
+      "step": 3377
+    },
+    {
+      "epoch": 0.590688524590164,
+      "grad_norm": 1.1110400523705453,
+      "learning_rate": 0.00018574205711289224,
+      "loss": 13.5751,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5908633879781421,
+      "grad_norm": 1.2472245298762819,
+      "learning_rate": 0.0001857323391661558,
+      "loss": 13.5726,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5910382513661202,
+      "grad_norm": 1.0913735962788642,
+      "learning_rate": 0.00018572261816316164,
+      "loss": 13.6065,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5912131147540983,
+      "grad_norm": 1.342912585833475,
+      "learning_rate": 0.00018571289410425628,
+      "loss": 13.6356,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5913879781420766,
+      "grad_norm": 1.1368784543353725,
+      "learning_rate": 0.0001857031669897864,
+      "loss": 13.6337,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5915628415300547,
+      "grad_norm": 1.2641490448492372,
+      "learning_rate": 0.00018569343682009874,
+      "loss": 13.7385,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5917377049180328,
+      "grad_norm": 1.140376368692139,
+      "learning_rate": 0.0001856837035955402,
+      "loss": 13.5313,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5919125683060109,
+      "grad_norm": 1.129262526488115,
+      "learning_rate": 0.0001856739673164577,
+      "loss": 13.4123,
+      "step": 3385
+    },
+    {
+      "epoch": 0.592087431693989,
+      "grad_norm": 1.3425549576362743,
+      "learning_rate": 0.0001856642279831984,
+      "loss": 13.5691,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5922622950819673,
+      "grad_norm": 1.2005107198695177,
+      "learning_rate": 0.00018565448559610946,
+      "loss": 13.5374,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5924371584699454,
+      "grad_norm": 1.2309232502436873,
+      "learning_rate": 0.00018564474015553818,
+      "loss": 13.5327,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5926120218579235,
+      "grad_norm": 1.1718309987192126,
+      "learning_rate": 0.00018563499166183195,
+      "loss": 13.5728,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5927868852459016,
+      "grad_norm": 1.1406686952070972,
+      "learning_rate": 0.00018562524011533834,
+      "loss": 13.4312,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5929617486338797,
+      "grad_norm": 1.1174427796607203,
+      "learning_rate": 0.00018561548551640494,
+      "loss": 13.363,
+      "step": 3391
+    },
+    {
+      "epoch": 0.593136612021858,
+      "grad_norm": 1.50259694200028,
+      "learning_rate": 0.00018560572786537955,
+      "loss": 13.1872,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5933114754098361,
+      "grad_norm": 1.0964477922882139,
+      "learning_rate": 0.00018559596716260997,
+      "loss": 13.5846,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5934863387978142,
+      "grad_norm": 1.5212894158960828,
+      "learning_rate": 0.00018558620340844418,
+      "loss": 13.9308,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5936612021857923,
+      "grad_norm": 1.1437000161449726,
+      "learning_rate": 0.0001855764366032302,
+      "loss": 13.8367,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5938360655737704,
+      "grad_norm": 1.320956169921284,
+      "learning_rate": 0.00018556666674731629,
+      "loss": 13.6937,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5940109289617487,
+      "grad_norm": 1.2648804494831412,
+      "learning_rate": 0.00018555689384105065,
+      "loss": 13.7595,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5941857923497268,
+      "grad_norm": 1.1203674775197547,
+      "learning_rate": 0.0001855471178847817,
+      "loss": 13.4594,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5943606557377049,
+      "grad_norm": 1.0239232223128065,
+      "learning_rate": 0.00018553733887885797,
+      "loss": 13.5928,
+      "step": 3399
+    },
+    {
+      "epoch": 0.594535519125683,
+      "grad_norm": 1.1561903726788638,
+      "learning_rate": 0.00018552755682362806,
+      "loss": 13.8309,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5947103825136612,
+      "grad_norm": 1.0886415528042235,
+      "learning_rate": 0.00018551777171944064,
+      "loss": 13.1358,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5948852459016394,
+      "grad_norm": 1.0940547328197818,
+      "learning_rate": 0.00018550798356664462,
+      "loss": 13.5408,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5950601092896175,
+      "grad_norm": 1.216734875193561,
+      "learning_rate": 0.00018549819236558885,
+      "loss": 13.3185,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5952349726775956,
+      "grad_norm": 1.1549707779312703,
+      "learning_rate": 0.00018548839811662244,
+      "loss": 13.5091,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5954098360655737,
+      "grad_norm": 1.3442032989820911,
+      "learning_rate": 0.0001854786008200945,
+      "loss": 13.5503,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5955846994535519,
+      "grad_norm": 1.1334696344212072,
+      "learning_rate": 0.00018546880047635433,
+      "loss": 13.5603,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5957595628415301,
+      "grad_norm": 1.2110162073888882,
+      "learning_rate": 0.00018545899708575126,
+      "loss": 13.5035,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5959344262295082,
+      "grad_norm": 1.1719358786432688,
+      "learning_rate": 0.0001854491906486348,
+      "loss": 13.598,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5961092896174863,
+      "grad_norm": 1.4512213986491935,
+      "learning_rate": 0.00018543938116535454,
+      "loss": 13.824,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5962841530054644,
+      "grad_norm": 1.2505907987890692,
+      "learning_rate": 0.00018542956863626018,
+      "loss": 13.1806,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5964590163934427,
+      "grad_norm": 1.3066759102803973,
+      "learning_rate": 0.0001854197530617015,
+      "loss": 13.591,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5966338797814208,
+      "grad_norm": 1.3902332877417056,
+      "learning_rate": 0.0001854099344420284,
+      "loss": 13.4018,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5968087431693989,
+      "grad_norm": 1.0946471702078666,
+      "learning_rate": 0.00018540011277759098,
+      "loss": 13.839,
+      "step": 3413
+    },
+    {
+      "epoch": 0.596983606557377,
+      "grad_norm": 1.1397839891963202,
+      "learning_rate": 0.00018539028806873926,
+      "loss": 13.4359,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5971584699453552,
+      "grad_norm": 1.0698805463644814,
+      "learning_rate": 0.00018538046031582356,
+      "loss": 13.2002,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5973333333333334,
+      "grad_norm": 1.2439819023487013,
+      "learning_rate": 0.00018537062951919423,
+      "loss": 13.6697,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5975081967213115,
+      "grad_norm": 1.146777420056596,
+      "learning_rate": 0.00018536079567920167,
+      "loss": 13.7101,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5976830601092896,
+      "grad_norm": 1.2049154723225795,
+      "learning_rate": 0.0001853509587961965,
+      "loss": 13.3813,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5978579234972677,
+      "grad_norm": 0.9982970664145563,
+      "learning_rate": 0.00018534111887052936,
+      "loss": 13.3232,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5980327868852459,
+      "grad_norm": 1.3018717621772986,
+      "learning_rate": 0.00018533127590255103,
+      "loss": 13.8034,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5982076502732241,
+      "grad_norm": 1.4938603037155935,
+      "learning_rate": 0.00018532142989261246,
+      "loss": 13.6255,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5983825136612022,
+      "grad_norm": 1.354724016867926,
+      "learning_rate": 0.00018531158084106455,
+      "loss": 13.4032,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5985573770491803,
+      "grad_norm": 1.1812136476087542,
+      "learning_rate": 0.00018530172874825848,
+      "loss": 13.4395,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5987322404371584,
+      "grad_norm": 1.10858975140192,
+      "learning_rate": 0.00018529187361454548,
+      "loss": 13.5871,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5989071038251366,
+      "grad_norm": 1.138779797941296,
+      "learning_rate": 0.0001852820154402768,
+      "loss": 13.7053,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5990819672131148,
+      "grad_norm": 1.1319467066239404,
+      "learning_rate": 0.0001852721542258039,
+      "loss": 13.6661,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5992568306010929,
+      "grad_norm": 1.133171786503854,
+      "learning_rate": 0.00018526228997147834,
+      "loss": 13.6169,
+      "step": 3427
+    },
+    {
+      "epoch": 0.599431693989071,
+      "grad_norm": 1.2063921718058477,
+      "learning_rate": 0.00018525242267765176,
+      "loss": 13.6743,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5996065573770492,
+      "grad_norm": 1.1080652958037664,
+      "learning_rate": 0.00018524255234467592,
+      "loss": 13.7415,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5997814207650273,
+      "grad_norm": 1.174235896134168,
+      "learning_rate": 0.00018523267897290269,
+      "loss": 13.5205,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5999562841530055,
+      "grad_norm": 1.1499781112796497,
+      "learning_rate": 0.00018522280256268406,
+      "loss": 13.8102,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6001311475409836,
+      "grad_norm": 1.1373318382090754,
+      "learning_rate": 0.00018521292311437203,
+      "loss": 13.7687,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6003060109289617,
+      "grad_norm": 1.2892809420468152,
+      "learning_rate": 0.0001852030406283189,
+      "loss": 13.5701,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6004808743169399,
+      "grad_norm": 1.0921521874289226,
+      "learning_rate": 0.0001851931551048769,
+      "loss": 13.626,
+      "step": 3434
+    },
+    {
+      "epoch": 0.600655737704918,
+      "grad_norm": 1.0853692068729033,
+      "learning_rate": 0.00018518326654439848,
+      "loss": 13.3752,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6008306010928962,
+      "grad_norm": 1.1754905567210931,
+      "learning_rate": 0.0001851733749472361,
+      "loss": 13.1123,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6010054644808743,
+      "grad_norm": 1.1513766909995768,
+      "learning_rate": 0.00018516348031374246,
+      "loss": 13.6438,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6011803278688524,
+      "grad_norm": 1.1246490841303531,
+      "learning_rate": 0.00018515358264427023,
+      "loss": 13.4715,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6013551912568306,
+      "grad_norm": 1.298175032802918,
+      "learning_rate": 0.00018514368193917226,
+      "loss": 13.7083,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6015300546448088,
+      "grad_norm": 1.181237297339856,
+      "learning_rate": 0.00018513377819880155,
+      "loss": 13.6853,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6017049180327869,
+      "grad_norm": 1.3465962000112033,
+      "learning_rate": 0.00018512387142351105,
+      "loss": 13.7468,
+      "step": 3441
+    },
+    {
+      "epoch": 0.601879781420765,
+      "grad_norm": 1.2154417288543433,
+      "learning_rate": 0.00018511396161365404,
+      "loss": 13.7335,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6020546448087432,
+      "grad_norm": 1.178670245222367,
+      "learning_rate": 0.0001851040487695837,
+      "loss": 13.7464,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6022295081967213,
+      "grad_norm": 1.277185374171119,
+      "learning_rate": 0.0001850941328916535,
+      "loss": 13.8094,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6024043715846995,
+      "grad_norm": 1.139684988522207,
+      "learning_rate": 0.00018508421398021687,
+      "loss": 13.5917,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6025792349726776,
+      "grad_norm": 1.2784575182383469,
+      "learning_rate": 0.00018507429203562742,
+      "loss": 13.6575,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6027540983606557,
+      "grad_norm": 1.2088122937562602,
+      "learning_rate": 0.00018506436705823887,
+      "loss": 13.4885,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6029289617486339,
+      "grad_norm": 1.2957666227269176,
+      "learning_rate": 0.00018505443904840498,
+      "loss": 13.6647,
+      "step": 3448
+    },
+    {
+      "epoch": 0.603103825136612,
+      "grad_norm": 1.5509261753691117,
+      "learning_rate": 0.00018504450800647972,
+      "loss": 13.7769,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6032786885245902,
+      "grad_norm": 1.0817455314770468,
+      "learning_rate": 0.00018503457393281717,
+      "loss": 13.5432,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6034535519125683,
+      "grad_norm": 1.4720281838718088,
+      "learning_rate": 0.00018502463682777137,
+      "loss": 13.5528,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6036284153005465,
+      "grad_norm": 1.0331966836366322,
+      "learning_rate": 0.0001850146966916966,
+      "loss": 13.5716,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6038032786885246,
+      "grad_norm": 1.5867315388835632,
+      "learning_rate": 0.00018500475352494725,
+      "loss": 13.4761,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6039781420765027,
+      "grad_norm": 1.1323182920978978,
+      "learning_rate": 0.00018499480732787774,
+      "loss": 13.6207,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6041530054644809,
+      "grad_norm": 1.474631584605191,
+      "learning_rate": 0.00018498485810084265,
+      "loss": 13.7353,
+      "step": 3455
+    },
+    {
+      "epoch": 0.604327868852459,
+      "grad_norm": 1.1535406575051739,
+      "learning_rate": 0.00018497490584419666,
+      "loss": 13.0831,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6045027322404372,
+      "grad_norm": 1.337130832426148,
+      "learning_rate": 0.00018496495055829457,
+      "loss": 13.4937,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6046775956284153,
+      "grad_norm": 1.2972507805115867,
+      "learning_rate": 0.00018495499224349123,
+      "loss": 13.6798,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6048524590163934,
+      "grad_norm": 1.3094256343793946,
+      "learning_rate": 0.0001849450309001417,
+      "loss": 13.7542,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6050273224043716,
+      "grad_norm": 1.3394220326367594,
+      "learning_rate": 0.0001849350665286011,
+      "loss": 13.8391,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6052021857923497,
+      "grad_norm": 1.3038630320659461,
+      "learning_rate": 0.00018492509912922459,
+      "loss": 13.6264,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6053770491803279,
+      "grad_norm": 1.1337456530352685,
+      "learning_rate": 0.0001849151287023675,
+      "loss": 13.6872,
+      "step": 3462
+    },
+    {
+      "epoch": 0.605551912568306,
+      "grad_norm": 1.2287930390467,
+      "learning_rate": 0.0001849051552483853,
+      "loss": 13.518,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6057267759562841,
+      "grad_norm": 1.273254939971413,
+      "learning_rate": 0.0001848951787676335,
+      "loss": 13.3623,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6059016393442623,
+      "grad_norm": 1.2003512811621562,
+      "learning_rate": 0.0001848851992604678,
+      "loss": 13.6268,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6060765027322405,
+      "grad_norm": 1.2725868732597525,
+      "learning_rate": 0.00018487521672724388,
+      "loss": 13.6627,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6062513661202186,
+      "grad_norm": 1.1632442555677893,
+      "learning_rate": 0.0001848652311683177,
+      "loss": 13.4457,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6064262295081967,
+      "grad_norm": 1.398948636500373,
+      "learning_rate": 0.00018485524258404515,
+      "loss": 13.625,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6066010928961748,
+      "grad_norm": 1.1975812647326158,
+      "learning_rate": 0.00018484525097478235,
+      "loss": 13.7301,
+      "step": 3469
+    },
+    {
+      "epoch": 0.606775956284153,
+      "grad_norm": 1.1875855291903812,
+      "learning_rate": 0.0001848352563408855,
+      "loss": 13.6067,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6069508196721312,
+      "grad_norm": 1.1949373463446575,
+      "learning_rate": 0.00018482525868271087,
+      "loss": 13.473,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6071256830601093,
+      "grad_norm": 1.2347881792403779,
+      "learning_rate": 0.0001848152580006149,
+      "loss": 13.7808,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6073005464480874,
+      "grad_norm": 1.053786031225198,
+      "learning_rate": 0.000184805254294954,
+      "loss": 13.2997,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6074754098360656,
+      "grad_norm": 1.0799690706472778,
+      "learning_rate": 0.00018479524756608492,
+      "loss": 13.1389,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6076502732240437,
+      "grad_norm": 1.1200593406306762,
+      "learning_rate": 0.00018478523781436437,
+      "loss": 13.5508,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6078251366120219,
+      "grad_norm": 1.1777303385352544,
+      "learning_rate": 0.0001847752250401491,
+      "loss": 13.5405,
+      "step": 3476
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 1.2015021042115492,
+      "learning_rate": 0.00018476520924379616,
+      "loss": 13.7584,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6081748633879781,
+      "grad_norm": 1.097967078561471,
+      "learning_rate": 0.0001847551904256625,
+      "loss": 13.5065,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6083497267759563,
+      "grad_norm": 1.125064620337091,
+      "learning_rate": 0.00018474516858610533,
+      "loss": 13.6467,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6085245901639345,
+      "grad_norm": 1.1455520282687257,
+      "learning_rate": 0.00018473514372548193,
+      "loss": 13.5108,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6086994535519126,
+      "grad_norm": 1.1764525775107082,
+      "learning_rate": 0.00018472511584414965,
+      "loss": 13.3052,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6088743169398907,
+      "grad_norm": 1.0854788919336873,
+      "learning_rate": 0.000184715084942466,
+      "loss": 13.4342,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6090491803278688,
+      "grad_norm": 1.1305081696806374,
+      "learning_rate": 0.00018470505102078852,
+      "loss": 13.3753,
+      "step": 3483
+    },
+    {
+      "epoch": 0.609224043715847,
+      "grad_norm": 1.3455028427154783,
+      "learning_rate": 0.00018469501407947494,
+      "loss": 13.7017,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6093989071038252,
+      "grad_norm": 1.08789524585948,
+      "learning_rate": 0.00018468497411888311,
+      "loss": 13.6711,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6095737704918033,
+      "grad_norm": 1.4048777762917908,
+      "learning_rate": 0.00018467493113937084,
+      "loss": 13.7685,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6097486338797814,
+      "grad_norm": 1.3794766486926324,
+      "learning_rate": 0.00018466488514129624,
+      "loss": 13.6256,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6099234972677595,
+      "grad_norm": 1.0813576044875859,
+      "learning_rate": 0.0001846548361250174,
+      "loss": 13.3775,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6100983606557377,
+      "grad_norm": 1.494701733409518,
+      "learning_rate": 0.00018464478409089256,
+      "loss": 13.8518,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6102732240437159,
+      "grad_norm": 1.1352088519776924,
+      "learning_rate": 0.00018463472903928004,
+      "loss": 13.4881,
+      "step": 3490
+    },
+    {
+      "epoch": 0.610448087431694,
+      "grad_norm": 1.1631881474881676,
+      "learning_rate": 0.00018462467097053834,
+      "loss": 13.6079,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6106229508196721,
+      "grad_norm": 1.1799922098643427,
+      "learning_rate": 0.000184614609885026,
+      "loss": 13.3299,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6107978142076502,
+      "grad_norm": 1.1165024218612287,
+      "learning_rate": 0.00018460454578310165,
+      "loss": 13.4646,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6109726775956285,
+      "grad_norm": 1.2378237906120204,
+      "learning_rate": 0.0001845944786651241,
+      "loss": 13.5753,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6111475409836066,
+      "grad_norm": 1.090186127139018,
+      "learning_rate": 0.00018458440853145227,
+      "loss": 13.1588,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6113224043715847,
+      "grad_norm": 1.2372127329953693,
+      "learning_rate": 0.00018457433538244505,
+      "loss": 13.4032,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6114972677595628,
+      "grad_norm": 0.9794409849675095,
+      "learning_rate": 0.0001845642592184616,
+      "loss": 13.2448,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6116721311475409,
+      "grad_norm": 1.396525454191913,
+      "learning_rate": 0.00018455418003986113,
+      "loss": 13.7563,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6118469945355192,
+      "grad_norm": 1.1314609804809281,
+      "learning_rate": 0.0001845440978470029,
+      "loss": 13.4524,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6120218579234973,
+      "grad_norm": 1.1889860987759673,
+      "learning_rate": 0.00018453401264024639,
+      "loss": 13.6805,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6121967213114754,
+      "grad_norm": 1.1238075649058945,
+      "learning_rate": 0.0001845239244199511,
+      "loss": 13.8149,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6123715846994535,
+      "grad_norm": 1.072512233354193,
+      "learning_rate": 0.00018451383318647663,
+      "loss": 13.3883,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6125464480874316,
+      "grad_norm": 1.2147934281086754,
+      "learning_rate": 0.00018450373894018278,
+      "loss": 13.6361,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6127213114754099,
+      "grad_norm": 1.1279422711199576,
+      "learning_rate": 0.00018449364168142936,
+      "loss": 13.3933,
+      "step": 3504
+    },
+    {
+      "epoch": 0.612896174863388,
+      "grad_norm": 1.200569782086159,
+      "learning_rate": 0.00018448354141057633,
+      "loss": 13.777,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6130710382513661,
+      "grad_norm": 1.116008616557573,
+      "learning_rate": 0.00018447343812798376,
+      "loss": 13.5436,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6132459016393442,
+      "grad_norm": 1.1912809563308637,
+      "learning_rate": 0.0001844633318340118,
+      "loss": 13.7053,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6134207650273225,
+      "grad_norm": 1.123231475013404,
+      "learning_rate": 0.00018445322252902074,
+      "loss": 13.6174,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6135956284153006,
+      "grad_norm": 1.2092238485120057,
+      "learning_rate": 0.00018444311021337102,
+      "loss": 13.3131,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6137704918032787,
+      "grad_norm": 1.2197599945342996,
+      "learning_rate": 0.00018443299488742304,
+      "loss": 13.8384,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6139453551912568,
+      "grad_norm": 1.235549284889929,
+      "learning_rate": 0.00018442287655153744,
+      "loss": 13.4817,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6141202185792349,
+      "grad_norm": 1.2486578239699964,
+      "learning_rate": 0.00018441275520607496,
+      "loss": 13.3025,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6142950819672132,
+      "grad_norm": 1.1578152822379064,
+      "learning_rate": 0.00018440263085139636,
+      "loss": 13.5228,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6144699453551913,
+      "grad_norm": 1.0448790165174107,
+      "learning_rate": 0.0001843925034878626,
+      "loss": 13.5041,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6146448087431694,
+      "grad_norm": 1.144822757301191,
+      "learning_rate": 0.0001843823731158347,
+      "loss": 13.5839,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6148196721311475,
+      "grad_norm": 1.1747807987521217,
+      "learning_rate": 0.00018437223973567375,
+      "loss": 13.319,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6149945355191256,
+      "grad_norm": 1.026661722759175,
+      "learning_rate": 0.00018436210334774104,
+      "loss": 13.4167,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6151693989071039,
+      "grad_norm": 1.1271363658046274,
+      "learning_rate": 0.00018435196395239794,
+      "loss": 13.4667,
+      "step": 3518
+    },
+    {
+      "epoch": 0.615344262295082,
+      "grad_norm": 1.0912318914447703,
+      "learning_rate": 0.00018434182155000584,
+      "loss": 13.6715,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6155191256830601,
+      "grad_norm": 1.079588268331472,
+      "learning_rate": 0.00018433167614092638,
+      "loss": 13.2843,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6156939890710382,
+      "grad_norm": 1.1264039643856416,
+      "learning_rate": 0.00018432152772552117,
+      "loss": 13.6269,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6158688524590163,
+      "grad_norm": 1.1765015474590221,
+      "learning_rate": 0.00018431137630415205,
+      "loss": 13.6219,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6160437158469946,
+      "grad_norm": 1.1045418944164038,
+      "learning_rate": 0.00018430122187718085,
+      "loss": 13.4068,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6162185792349727,
+      "grad_norm": 1.269784826194506,
+      "learning_rate": 0.00018429106444496957,
+      "loss": 13.7416,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6163934426229508,
+      "grad_norm": 1.0827417594202744,
+      "learning_rate": 0.00018428090400788034,
+      "loss": 13.3537,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6165683060109289,
+      "grad_norm": 1.1941697795207937,
+      "learning_rate": 0.00018427074056627535,
+      "loss": 13.4005,
+      "step": 3526
+    },
+    {
+      "epoch": 0.616743169398907,
+      "grad_norm": 1.3884747332353056,
+      "learning_rate": 0.00018426057412051697,
+      "loss": 13.2332,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6169180327868853,
+      "grad_norm": 1.2247134278519725,
+      "learning_rate": 0.00018425040467096753,
+      "loss": 13.3512,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6170928961748634,
+      "grad_norm": 1.3349091986905972,
+      "learning_rate": 0.00018424023221798956,
+      "loss": 13.5841,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6172677595628415,
+      "grad_norm": 1.4595765952204782,
+      "learning_rate": 0.0001842300567619458,
+      "loss": 13.6966,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6174426229508196,
+      "grad_norm": 1.156259775443137,
+      "learning_rate": 0.0001842198783031989,
+      "loss": 13.5445,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6176174863387978,
+      "grad_norm": 1.236632849610943,
+      "learning_rate": 0.00018420969684211172,
+      "loss": 13.8818,
+      "step": 3532
+    },
+    {
+      "epoch": 0.617792349726776,
+      "grad_norm": 1.1965264401831748,
+      "learning_rate": 0.00018419951237904725,
+      "loss": 13.6694,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6179672131147541,
+      "grad_norm": 1.1093239190514494,
+      "learning_rate": 0.00018418932491436854,
+      "loss": 13.6039,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6181420765027322,
+      "grad_norm": 1.0319959592936405,
+      "learning_rate": 0.0001841791344484388,
+      "loss": 13.5145,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6183169398907103,
+      "grad_norm": 1.2556766054324962,
+      "learning_rate": 0.00018416894098162122,
+      "loss": 13.3562,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6184918032786886,
+      "grad_norm": 1.0963042494610156,
+      "learning_rate": 0.00018415874451427927,
+      "loss": 13.5246,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6186666666666667,
+      "grad_norm": 1.301131544309476,
+      "learning_rate": 0.00018414854504677637,
+      "loss": 13.604,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6188415300546448,
+      "grad_norm": 1.0991068810660287,
+      "learning_rate": 0.00018413834257947618,
+      "loss": 13.561,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6190163934426229,
+      "grad_norm": 1.5627096185557838,
+      "learning_rate": 0.00018412813711274237,
+      "loss": 13.7967,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6191912568306011,
+      "grad_norm": 1.386565051755526,
+      "learning_rate": 0.00018411792864693878,
+      "loss": 13.3279,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6193661202185793,
+      "grad_norm": 1.2755180143704328,
+      "learning_rate": 0.0001841077171824293,
+      "loss": 13.4161,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6195409836065574,
+      "grad_norm": 1.785511157401594,
+      "learning_rate": 0.00018409750271957797,
+      "loss": 13.8902,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6197158469945355,
+      "grad_norm": 1.0919753533194605,
+      "learning_rate": 0.00018408728525874896,
+      "loss": 13.633,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6198907103825136,
+      "grad_norm": 1.6321607543421683,
+      "learning_rate": 0.00018407706480030644,
+      "loss": 13.6226,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6200655737704918,
+      "grad_norm": 1.1905137367153251,
+      "learning_rate": 0.0001840668413446148,
+      "loss": 13.5146,
+      "step": 3546
+    },
+    {
+      "epoch": 0.62024043715847,
+      "grad_norm": 1.2697051157925883,
+      "learning_rate": 0.0001840566148920385,
+      "loss": 13.6146,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6204153005464481,
+      "grad_norm": 1.2362692070785855,
+      "learning_rate": 0.00018404638544294207,
+      "loss": 13.643,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6205901639344262,
+      "grad_norm": 1.194529738140476,
+      "learning_rate": 0.00018403615299769023,
+      "loss": 13.419,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6207650273224044,
+      "grad_norm": 1.1892690085361848,
+      "learning_rate": 0.00018402591755664767,
+      "loss": 13.3825,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6209398907103825,
+      "grad_norm": 1.1274562329181563,
+      "learning_rate": 0.00018401567912017933,
+      "loss": 13.5771,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6211147540983607,
+      "grad_norm": 1.2406854393441804,
+      "learning_rate": 0.00018400543768865024,
+      "loss": 13.6427,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6212896174863388,
+      "grad_norm": 1.1176582383588316,
+      "learning_rate": 0.0001839951932624254,
+      "loss": 13.2464,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6214644808743169,
+      "grad_norm": 1.3549647870963086,
+      "learning_rate": 0.0001839849458418701,
+      "loss": 13.4051,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6216393442622951,
+      "grad_norm": 1.1304882019905114,
+      "learning_rate": 0.00018397469542734957,
+      "loss": 13.6422,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6218142076502732,
+      "grad_norm": 1.305909758611905,
+      "learning_rate": 0.0001839644420192293,
+      "loss": 13.507,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6219890710382514,
+      "grad_norm": 1.1241688417370046,
+      "learning_rate": 0.00018395418561787473,
+      "loss": 13.6394,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6221639344262295,
+      "grad_norm": 1.0859349622601775,
+      "learning_rate": 0.00018394392622365157,
+      "loss": 13.6044,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6223387978142076,
+      "grad_norm": 1.2203809853424643,
+      "learning_rate": 0.00018393366383692548,
+      "loss": 13.4772,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6225136612021858,
+      "grad_norm": 1.2261754454709395,
+      "learning_rate": 0.00018392339845806234,
+      "loss": 13.6233,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6226885245901639,
+      "grad_norm": 1.3754959352149652,
+      "learning_rate": 0.00018391313008742812,
+      "loss": 13.2379,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6228633879781421,
+      "grad_norm": 1.1676295692174345,
+      "learning_rate": 0.00018390285872538887,
+      "loss": 13.4247,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6230382513661202,
+      "grad_norm": 1.3341838762651266,
+      "learning_rate": 0.00018389258437231072,
+      "loss": 13.5484,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6232131147540984,
+      "grad_norm": 1.157250315645199,
+      "learning_rate": 0.00018388230702855995,
+      "loss": 13.6022,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6233879781420765,
+      "grad_norm": 1.3317939727969537,
+      "learning_rate": 0.00018387202669450294,
+      "loss": 13.6827,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6235628415300546,
+      "grad_norm": 1.3694889561368528,
+      "learning_rate": 0.0001838617433705062,
+      "loss": 13.5898,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6237377049180328,
+      "grad_norm": 1.242719130325765,
+      "learning_rate": 0.00018385145705693626,
+      "loss": 13.4984,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6239125683060109,
+      "grad_norm": 1.2835195288374872,
+      "learning_rate": 0.00018384116775415984,
+      "loss": 13.466,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6240874316939891,
+      "grad_norm": 1.2287923479289495,
+      "learning_rate": 0.00018383087546254377,
+      "loss": 13.2505,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6242622950819672,
+      "grad_norm": 1.517854551896419,
+      "learning_rate": 0.0001838205801824549,
+      "loss": 13.4806,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6244371584699454,
+      "grad_norm": 1.0883568354842463,
+      "learning_rate": 0.00018381028191426032,
+      "loss": 13.827,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6246120218579235,
+      "grad_norm": 1.243538646157312,
+      "learning_rate": 0.00018379998065832712,
+      "loss": 13.4716,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6247868852459016,
+      "grad_norm": 1.2231812326920612,
+      "learning_rate": 0.00018378967641502248,
+      "loss": 13.6229,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6249617486338798,
+      "grad_norm": 1.1892334554127575,
+      "learning_rate": 0.0001837793691847138,
+      "loss": 13.5496,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6251366120218579,
+      "grad_norm": 1.3184913098591091,
+      "learning_rate": 0.0001837690589677685,
+      "loss": 13.7276,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6253114754098361,
+      "grad_norm": 1.0980606226053158,
+      "learning_rate": 0.00018375874576455407,
+      "loss": 13.3523,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6254863387978142,
+      "grad_norm": 1.149408664572725,
+      "learning_rate": 0.00018374842957543828,
+      "loss": 13.6445,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6256612021857924,
+      "grad_norm": 1.066031156674473,
+      "learning_rate": 0.0001837381104007888,
+      "loss": 13.6131,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6258360655737705,
+      "grad_norm": 1.0850950915701207,
+      "learning_rate": 0.0001837277882409735,
+      "loss": 13.4788,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6260109289617486,
+      "grad_norm": 1.0610743055090108,
+      "learning_rate": 0.0001837174630963604,
+      "loss": 13.4419,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6261857923497268,
+      "grad_norm": 1.1531851196245324,
+      "learning_rate": 0.00018370713496731756,
+      "loss": 13.5774,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6263606557377049,
+      "grad_norm": 1.1105370370301848,
+      "learning_rate": 0.00018369680385421316,
+      "loss": 13.4742,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6265355191256831,
+      "grad_norm": 1.1215975252580959,
+      "learning_rate": 0.00018368646975741544,
+      "loss": 13.5054,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6267103825136612,
+      "grad_norm": 1.1264460459221477,
+      "learning_rate": 0.0001836761326772929,
+      "loss": 13.4178,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6268852459016393,
+      "grad_norm": 1.1605187460207491,
+      "learning_rate": 0.00018366579261421397,
+      "loss": 13.5259,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6270601092896175,
+      "grad_norm": 1.2215221331671227,
+      "learning_rate": 0.0001836554495685473,
+      "loss": 13.4552,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6272349726775956,
+      "grad_norm": 1.1397853549477104,
+      "learning_rate": 0.00018364510354066163,
+      "loss": 13.6044,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6274098360655738,
+      "grad_norm": 1.275940652851,
+      "learning_rate": 0.00018363475453092568,
+      "loss": 13.1909,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6275846994535519,
+      "grad_norm": 1.065552542833296,
+      "learning_rate": 0.00018362440253970848,
+      "loss": 13.2973,
+      "step": 3589
+    },
+    {
+      "epoch": 0.62775956284153,
+      "grad_norm": 1.1770410680857162,
+      "learning_rate": 0.00018361404756737907,
+      "loss": 13.5586,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6279344262295082,
+      "grad_norm": 1.030066688104203,
+      "learning_rate": 0.0001836036896143065,
+      "loss": 13.3568,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6281092896174864,
+      "grad_norm": 1.0624598122693039,
+      "learning_rate": 0.00018359332868086008,
+      "loss": 13.502,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6282841530054645,
+      "grad_norm": 1.1817554999388071,
+      "learning_rate": 0.0001835829647674092,
+      "loss": 13.46,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6284590163934426,
+      "grad_norm": 1.0994756695759,
+      "learning_rate": 0.00018357259787432325,
+      "loss": 13.5795,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6286338797814207,
+      "grad_norm": 1.3105792658124125,
+      "learning_rate": 0.00018356222800197184,
+      "loss": 13.3193,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6288087431693989,
+      "grad_norm": 1.1618201865232334,
+      "learning_rate": 0.00018355185515072465,
+      "loss": 13.1055,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6289836065573771,
+      "grad_norm": 1.1581127471867285,
+      "learning_rate": 0.0001835414793209514,
+      "loss": 13.6892,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6291584699453552,
+      "grad_norm": 1.1320519214440483,
+      "learning_rate": 0.00018353110051302207,
+      "loss": 13.4349,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6293333333333333,
+      "grad_norm": 1.168987840537311,
+      "learning_rate": 0.00018352071872730657,
+      "loss": 13.6144,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6295081967213115,
+      "grad_norm": 1.2784825298144247,
+      "learning_rate": 0.00018351033396417504,
+      "loss": 13.5016,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6296830601092896,
+      "grad_norm": 1.1895873153867476,
+      "learning_rate": 0.0001834999462239977,
+      "loss": 13.1531,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6298579234972678,
+      "grad_norm": 1.1793475197749033,
+      "learning_rate": 0.00018348955550714476,
+      "loss": 13.5347,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6300327868852459,
+      "grad_norm": 1.370536863442864,
+      "learning_rate": 0.00018347916181398678,
+      "loss": 13.5594,
+      "step": 3603
+    },
+    {
+      "epoch": 0.630207650273224,
+      "grad_norm": 1.0800253443496268,
+      "learning_rate": 0.00018346876514489422,
+      "loss": 13.3832,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6303825136612022,
+      "grad_norm": 1.142830972095037,
+      "learning_rate": 0.00018345836550023766,
+      "loss": 13.4572,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6305573770491804,
+      "grad_norm": 1.0894295185643978,
+      "learning_rate": 0.00018344796288038794,
+      "loss": 13.2357,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6307322404371585,
+      "grad_norm": 1.2547768056798378,
+      "learning_rate": 0.00018343755728571578,
+      "loss": 13.6834,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6309071038251366,
+      "grad_norm": 1.0503929394945186,
+      "learning_rate": 0.00018342714871659224,
+      "loss": 13.4001,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6310819672131147,
+      "grad_norm": 1.16687865169157,
+      "learning_rate": 0.00018341673717338827,
+      "loss": 13.6119,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6312568306010929,
+      "grad_norm": 1.0844010374864426,
+      "learning_rate": 0.00018340632265647513,
+      "loss": 13.5134,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6314316939890711,
+      "grad_norm": 1.0802085631002922,
+      "learning_rate": 0.000183395905166224,
+      "loss": 13.6171,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6316065573770492,
+      "grad_norm": 1.1559787045640633,
+      "learning_rate": 0.00018338548470300632,
+      "loss": 13.5261,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6317814207650273,
+      "grad_norm": 1.073754681499962,
+      "learning_rate": 0.00018337506126719352,
+      "loss": 13.564,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6319562841530054,
+      "grad_norm": 1.3161063570432199,
+      "learning_rate": 0.0001833646348591572,
+      "loss": 13.5113,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6321311475409837,
+      "grad_norm": 1.0416876108787616,
+      "learning_rate": 0.00018335420547926904,
+      "loss": 13.2107,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6323060109289618,
+      "grad_norm": 1.1115823958635362,
+      "learning_rate": 0.00018334377312790087,
+      "loss": 13.6883,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6324808743169399,
+      "grad_norm": 1.059352935117229,
+      "learning_rate": 0.00018333333780542454,
+      "loss": 13.5876,
+      "step": 3617
+    },
+    {
+      "epoch": 0.632655737704918,
+      "grad_norm": 1.1085302956943375,
+      "learning_rate": 0.00018332289951221212,
+      "loss": 13.0881,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6328306010928961,
+      "grad_norm": 1.0530277375320944,
+      "learning_rate": 0.00018331245824863563,
+      "loss": 13.6673,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6330054644808744,
+      "grad_norm": 1.165090467769962,
+      "learning_rate": 0.00018330201401506736,
+      "loss": 13.5816,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6331803278688525,
+      "grad_norm": 1.104951212380159,
+      "learning_rate": 0.00018329156681187966,
+      "loss": 13.5186,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6333551912568306,
+      "grad_norm": 1.3270716906972349,
+      "learning_rate": 0.00018328111663944488,
+      "loss": 13.7617,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6335300546448087,
+      "grad_norm": 1.340086297773826,
+      "learning_rate": 0.0001832706634981356,
+      "loss": 13.4905,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6337049180327868,
+      "grad_norm": 0.9963624072637088,
+      "learning_rate": 0.00018326020738832445,
+      "loss": 13.3623,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6338797814207651,
+      "grad_norm": 1.383047018396898,
+      "learning_rate": 0.0001832497483103842,
+      "loss": 13.5452,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6340546448087432,
+      "grad_norm": 1.1950217477401177,
+      "learning_rate": 0.00018323928626468768,
+      "loss": 13.3853,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6342295081967213,
+      "grad_norm": 1.2335569267031434,
+      "learning_rate": 0.00018322882125160788,
+      "loss": 13.3244,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6344043715846994,
+      "grad_norm": 1.2198230894716469,
+      "learning_rate": 0.00018321835327151783,
+      "loss": 13.5137,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6345792349726775,
+      "grad_norm": 1.1141391061444106,
+      "learning_rate": 0.00018320788232479069,
+      "loss": 13.5836,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6347540983606558,
+      "grad_norm": 1.2536763903870378,
+      "learning_rate": 0.0001831974084117998,
+      "loss": 13.2004,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6349289617486339,
+      "grad_norm": 1.1271934840570288,
+      "learning_rate": 0.00018318693153291851,
+      "loss": 13.504,
+      "step": 3631
+    },
+    {
+      "epoch": 0.635103825136612,
+      "grad_norm": 1.0755190920739264,
+      "learning_rate": 0.00018317645168852028,
+      "loss": 13.4916,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6352786885245901,
+      "grad_norm": 1.054030253623532,
+      "learning_rate": 0.00018316596887897873,
+      "loss": 13.4545,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6354535519125684,
+      "grad_norm": 1.1861295037599298,
+      "learning_rate": 0.00018315548310466758,
+      "loss": 13.3004,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6356284153005465,
+      "grad_norm": 1.1526633366705976,
+      "learning_rate": 0.00018314499436596058,
+      "loss": 13.6421,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6358032786885246,
+      "grad_norm": 1.2582711153840436,
+      "learning_rate": 0.0001831345026632317,
+      "loss": 13.3638,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6359781420765027,
+      "grad_norm": 1.2256182157600344,
+      "learning_rate": 0.00018312400799685495,
+      "loss": 13.0682,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6361530054644808,
+      "grad_norm": 1.2232056290715785,
+      "learning_rate": 0.0001831135103672044,
+      "loss": 13.7461,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6363278688524591,
+      "grad_norm": 1.0538688449838323,
+      "learning_rate": 0.0001831030097746543,
+      "loss": 13.3799,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6365027322404372,
+      "grad_norm": 1.4102680272124948,
+      "learning_rate": 0.00018309250621957905,
+      "loss": 13.6606,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6366775956284153,
+      "grad_norm": 1.1383724319381687,
+      "learning_rate": 0.000183081999702353,
+      "loss": 13.3884,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6368524590163934,
+      "grad_norm": 1.05354454150061,
+      "learning_rate": 0.00018307149022335075,
+      "loss": 13.4184,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6370273224043715,
+      "grad_norm": 1.16354307583164,
+      "learning_rate": 0.00018306097778294692,
+      "loss": 13.4367,
+      "step": 3643
+    },
+    {
+      "epoch": 0.6372021857923498,
+      "grad_norm": 1.1456281052013215,
+      "learning_rate": 0.00018305046238151627,
+      "loss": 13.5894,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6373770491803279,
+      "grad_norm": 1.3070631233050416,
+      "learning_rate": 0.0001830399440194337,
+      "loss": 13.1757,
+      "step": 3645
+    },
+    {
+      "epoch": 0.637551912568306,
+      "grad_norm": 1.0699948397132144,
+      "learning_rate": 0.00018302942269707412,
+      "loss": 13.3028,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6377267759562841,
+      "grad_norm": 1.1119676085731325,
+      "learning_rate": 0.00018301889841481264,
+      "loss": 13.6058,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6379016393442623,
+      "grad_norm": 1.2210585218635392,
+      "learning_rate": 0.00018300837117302444,
+      "loss": 13.4363,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6380765027322405,
+      "grad_norm": 1.1943717833429346,
+      "learning_rate": 0.00018299784097208476,
+      "loss": 13.3919,
+      "step": 3649
+    },
+    {
+      "epoch": 0.6382513661202186,
+      "grad_norm": 1.3373865957020399,
+      "learning_rate": 0.00018298730781236903,
+      "loss": 13.819,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6384262295081967,
+      "grad_norm": 1.3427599683377214,
+      "learning_rate": 0.00018297677169425275,
+      "loss": 13.5283,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6386010928961748,
+      "grad_norm": 1.0574842038013743,
+      "learning_rate": 0.00018296623261811152,
+      "loss": 13.089,
+      "step": 3652
+    },
+    {
+      "epoch": 0.638775956284153,
+      "grad_norm": 1.2324486335688178,
+      "learning_rate": 0.000182955690584321,
+      "loss": 13.3164,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6389508196721312,
+      "grad_norm": 1.103407440974947,
+      "learning_rate": 0.00018294514559325707,
+      "loss": 13.42,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6391256830601093,
+      "grad_norm": 1.2242238370655103,
+      "learning_rate": 0.00018293459764529558,
+      "loss": 13.4293,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6393005464480874,
+      "grad_norm": 1.1672559746041917,
+      "learning_rate": 0.0001829240467408126,
+      "loss": 13.538,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6394754098360655,
+      "grad_norm": 1.0377209477917848,
+      "learning_rate": 0.00018291349288018425,
+      "loss": 13.2948,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6396502732240437,
+      "grad_norm": 1.2623040658400493,
+      "learning_rate": 0.00018290293606378674,
+      "loss": 13.4516,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6398251366120219,
+      "grad_norm": 1.1411459017737329,
+      "learning_rate": 0.00018289237629199643,
+      "loss": 13.5397,
+      "step": 3659
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0990507488911239,
+      "learning_rate": 0.00018288181356518972,
+      "loss": 13.3086,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6401748633879781,
+      "grad_norm": 1.0951148697993385,
+      "learning_rate": 0.00018287124788374324,
+      "loss": 13.4296,
+      "step": 3661
+    },
+    {
+      "epoch": 0.6403497267759563,
+      "grad_norm": 1.1296027955600811,
+      "learning_rate": 0.00018286067924803357,
+      "loss": 13.3586,
+      "step": 3662
+    },
+    {
+      "epoch": 0.6405245901639344,
+      "grad_norm": 1.1699415932550388,
+      "learning_rate": 0.00018285010765843748,
+      "loss": 13.5081,
+      "step": 3663
+    },
+    {
+      "epoch": 0.6406994535519126,
+      "grad_norm": 1.1215498960776535,
+      "learning_rate": 0.0001828395331153319,
+      "loss": 13.3637,
+      "step": 3664
+    },
+    {
+      "epoch": 0.6408743169398907,
+      "grad_norm": 1.273019235731098,
+      "learning_rate": 0.0001828289556190937,
+      "loss": 13.5434,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6410491803278688,
+      "grad_norm": 1.0664439256610798,
+      "learning_rate": 0.00018281837517010002,
+      "loss": 13.5333,
+      "step": 3666
+    },
+    {
+      "epoch": 0.641224043715847,
+      "grad_norm": 1.132402801353701,
+      "learning_rate": 0.00018280779176872806,
+      "loss": 13.125,
+      "step": 3667
+    },
+    {
+      "epoch": 0.6413989071038252,
+      "grad_norm": 1.0784195180601692,
+      "learning_rate": 0.00018279720541535506,
+      "loss": 13.3971,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6415737704918033,
+      "grad_norm": 1.0879394581185502,
+      "learning_rate": 0.00018278661611035842,
+      "loss": 13.5336,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6417486338797814,
+      "grad_norm": 1.1425776159738898,
+      "learning_rate": 0.00018277602385411562,
+      "loss": 13.6846,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6419234972677595,
+      "grad_norm": 1.0404617029404288,
+      "learning_rate": 0.00018276542864700433,
+      "loss": 13.3511,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6420983606557377,
+      "grad_norm": 1.1817395106306305,
+      "learning_rate": 0.0001827548304894022,
+      "loss": 13.4541,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6422732240437159,
+      "grad_norm": 1.1412240415667787,
+      "learning_rate": 0.00018274422938168702,
+      "loss": 13.6468,
+      "step": 3673
+    },
+    {
+      "epoch": 0.642448087431694,
+      "grad_norm": 1.2280688088608323,
+      "learning_rate": 0.00018273362532423675,
+      "loss": 13.6444,
+      "step": 3674
+    },
+    {
+      "epoch": 0.6426229508196721,
+      "grad_norm": 1.1746508111682068,
+      "learning_rate": 0.00018272301831742942,
+      "loss": 13.4148,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6427978142076503,
+      "grad_norm": 1.123882976783784,
+      "learning_rate": 0.0001827124083616431,
+      "loss": 13.5754,
+      "step": 3676
+    },
+    {
+      "epoch": 0.6429726775956284,
+      "grad_norm": 1.2554772574704003,
+      "learning_rate": 0.00018270179545725612,
+      "loss": 13.4365,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6431475409836066,
+      "grad_norm": 1.2319015584546669,
+      "learning_rate": 0.0001826911796046467,
+      "loss": 13.6123,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6433224043715847,
+      "grad_norm": 1.169861846633121,
+      "learning_rate": 0.00018268056080419335,
+      "loss": 13.6165,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6434972677595628,
+      "grad_norm": 1.2614375044828119,
+      "learning_rate": 0.00018266993905627463,
+      "loss": 13.4491,
+      "step": 3680
+    },
+    {
+      "epoch": 0.643672131147541,
+      "grad_norm": 1.1193574447286199,
+      "learning_rate": 0.0001826593143612692,
+      "loss": 13.5037,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6438469945355191,
+      "grad_norm": 1.1784956689408606,
+      "learning_rate": 0.00018264868671955572,
+      "loss": 13.6801,
+      "step": 3682
+    },
+    {
+      "epoch": 0.6440218579234973,
+      "grad_norm": 1.2218777951537458,
+      "learning_rate": 0.00018263805613151317,
+      "loss": 13.2125,
+      "step": 3683
+    },
+    {
+      "epoch": 0.6441967213114754,
+      "grad_norm": 1.306739917170205,
+      "learning_rate": 0.00018262742259752047,
+      "loss": 13.5613,
+      "step": 3684
+    },
+    {
+      "epoch": 0.6443715846994535,
+      "grad_norm": 1.4086064360054535,
+      "learning_rate": 0.0001826167861179567,
+      "loss": 13.4062,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6445464480874317,
+      "grad_norm": 1.1238385132709063,
+      "learning_rate": 0.00018260614669320102,
+      "loss": 13.4265,
+      "step": 3686
+    },
+    {
+      "epoch": 0.6447213114754098,
+      "grad_norm": 1.018478993057905,
+      "learning_rate": 0.0001825955043236327,
+      "loss": 13.4826,
+      "step": 3687
+    },
+    {
+      "epoch": 0.644896174863388,
+      "grad_norm": 1.0611537396174475,
+      "learning_rate": 0.00018258485900963118,
+      "loss": 13.3971,
+      "step": 3688
+    },
+    {
+      "epoch": 0.6450710382513661,
+      "grad_norm": 1.1147315387880634,
+      "learning_rate": 0.00018257421075157593,
+      "loss": 13.5392,
+      "step": 3689
+    },
+    {
+      "epoch": 0.6452459016393443,
+      "grad_norm": 1.1481746251514715,
+      "learning_rate": 0.00018256355954984654,
+      "loss": 13.4556,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6454207650273224,
+      "grad_norm": 1.1178558712737368,
+      "learning_rate": 0.00018255290540482273,
+      "loss": 13.3824,
+      "step": 3691
+    },
+    {
+      "epoch": 0.6455956284153005,
+      "grad_norm": 1.121387201907994,
+      "learning_rate": 0.00018254224831688425,
+      "loss": 13.54,
+      "step": 3692
+    },
+    {
+      "epoch": 0.6457704918032787,
+      "grad_norm": 1.1868548951683981,
+      "learning_rate": 0.0001825315882864111,
+      "loss": 13.4597,
+      "step": 3693
+    },
+    {
+      "epoch": 0.6459453551912568,
+      "grad_norm": 1.028729757736023,
+      "learning_rate": 0.00018252092531378323,
+      "loss": 13.3,
+      "step": 3694
+    },
+    {
+      "epoch": 0.646120218579235,
+      "grad_norm": 1.0459853117424054,
+      "learning_rate": 0.0001825102593993808,
+      "loss": 13.4381,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6462950819672131,
+      "grad_norm": 1.0192035431602735,
+      "learning_rate": 0.00018249959054358403,
+      "loss": 13.5795,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6464699453551913,
+      "grad_norm": 1.0280679695022703,
+      "learning_rate": 0.00018248891874677323,
+      "loss": 13.4295,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6466448087431694,
+      "grad_norm": 1.0742253586670178,
+      "learning_rate": 0.00018247824400932888,
+      "loss": 13.6384,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6468196721311475,
+      "grad_norm": 1.0619512711897166,
+      "learning_rate": 0.00018246756633163148,
+      "loss": 13.4152,
+      "step": 3699
+    },
+    {
+      "epoch": 0.6469945355191257,
+      "grad_norm": 1.110308152055046,
+      "learning_rate": 0.0001824568857140617,
+      "loss": 13.5581,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6471693989071038,
+      "grad_norm": 1.200380779317526,
+      "learning_rate": 0.00018244620215700027,
+      "loss": 13.4373,
+      "step": 3701
+    },
+    {
+      "epoch": 0.647344262295082,
+      "grad_norm": 1.0680187425542933,
+      "learning_rate": 0.00018243551566082808,
+      "loss": 13.1469,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6475191256830601,
+      "grad_norm": 1.24687063687571,
+      "learning_rate": 0.0001824248262259261,
+      "loss": 13.3408,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6476939890710383,
+      "grad_norm": 1.1175382140720478,
+      "learning_rate": 0.00018241413385267533,
+      "loss": 13.347,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6478688524590164,
+      "grad_norm": 1.1437375744579255,
+      "learning_rate": 0.00018240343854145695,
+      "loss": 13.502,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6480437158469945,
+      "grad_norm": 1.0583876629026994,
+      "learning_rate": 0.00018239274029265232,
+      "loss": 13.6813,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6482185792349727,
+      "grad_norm": 1.054133300950442,
+      "learning_rate": 0.00018238203910664273,
+      "loss": 13.6389,
+      "step": 3707
+    },
+    {
+      "epoch": 0.6483934426229508,
+      "grad_norm": 1.1526402544671663,
+      "learning_rate": 0.00018237133498380973,
+      "loss": 13.519,
+      "step": 3708
+    },
+    {
+      "epoch": 0.648568306010929,
+      "grad_norm": 1.136471153823572,
+      "learning_rate": 0.00018236062792453486,
+      "loss": 13.5146,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6487431693989071,
+      "grad_norm": 1.0667526106992309,
+      "learning_rate": 0.00018234991792919983,
+      "loss": 13.4133,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6489180327868852,
+      "grad_norm": 1.0816496472920132,
+      "learning_rate": 0.00018233920499818643,
+      "loss": 13.5892,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6490928961748634,
+      "grad_norm": 1.0346628383907097,
+      "learning_rate": 0.00018232848913187657,
+      "loss": 13.3638,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6492677595628416,
+      "grad_norm": 1.0397610993952668,
+      "learning_rate": 0.0001823177703306523,
+      "loss": 13.6354,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6494426229508197,
+      "grad_norm": 1.2091224237762483,
+      "learning_rate": 0.00018230704859489567,
+      "loss": 13.3701,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6496174863387978,
+      "grad_norm": 1.3420594654631648,
+      "learning_rate": 0.00018229632392498894,
+      "loss": 13.504,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6497923497267759,
+      "grad_norm": 1.2614178707173744,
+      "learning_rate": 0.0001822855963213144,
+      "loss": 13.6286,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6499672131147541,
+      "grad_norm": 1.228163867622733,
+      "learning_rate": 0.00018227486578425447,
+      "loss": 13.4725,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6501420765027323,
+      "grad_norm": 1.2431439480284636,
+      "learning_rate": 0.00018226413231419174,
+      "loss": 13.5867,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6503169398907104,
+      "grad_norm": 1.1992516375722777,
+      "learning_rate": 0.0001822533959115088,
+      "loss": 13.4375,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6504918032786885,
+      "grad_norm": 1.1026056938253808,
+      "learning_rate": 0.00018224265657658842,
+      "loss": 13.7823,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6506666666666666,
+      "grad_norm": 1.4438658204937642,
+      "learning_rate": 0.00018223191430981336,
+      "loss": 13.4335,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6508415300546448,
+      "grad_norm": 1.1764163932444567,
+      "learning_rate": 0.00018222116911156667,
+      "loss": 13.597,
+      "step": 3722
+    },
+    {
+      "epoch": 0.651016393442623,
+      "grad_norm": 1.1968392156515237,
+      "learning_rate": 0.00018221042098223137,
+      "loss": 13.6037,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6511912568306011,
+      "grad_norm": 1.077618783334924,
+      "learning_rate": 0.0001821996699221906,
+      "loss": 13.6081,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6513661202185792,
+      "grad_norm": 1.0616604687917335,
+      "learning_rate": 0.00018218891593182768,
+      "loss": 13.764,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6515409836065573,
+      "grad_norm": 1.3186199302817554,
+      "learning_rate": 0.00018217815901152591,
+      "loss": 13.3377,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6517158469945356,
+      "grad_norm": 1.2256329754833133,
+      "learning_rate": 0.00018216739916166876,
+      "loss": 13.6997,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6518907103825137,
+      "grad_norm": 1.2759602097971556,
+      "learning_rate": 0.00018215663638263986,
+      "loss": 13.3277,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6520655737704918,
+      "grad_norm": 1.1966722467718982,
+      "learning_rate": 0.00018214587067482286,
+      "loss": 13.258,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6522404371584699,
+      "grad_norm": 1.3277156991300592,
+      "learning_rate": 0.00018213510203860154,
+      "loss": 13.4958,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6524153005464481,
+      "grad_norm": 1.3010002294607208,
+      "learning_rate": 0.0001821243304743598,
+      "loss": 13.3855,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6525901639344263,
+      "grad_norm": 1.247923976545006,
+      "learning_rate": 0.00018211355598248162,
+      "loss": 13.3254,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6527650273224044,
+      "grad_norm": 1.138286495769463,
+      "learning_rate": 0.00018210277856335112,
+      "loss": 13.3418,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6529398907103825,
+      "grad_norm": 1.4108600414422123,
+      "learning_rate": 0.00018209199821735247,
+      "loss": 13.5889,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6531147540983606,
+      "grad_norm": 1.1534070105965393,
+      "learning_rate": 0.00018208121494487002,
+      "loss": 13.6099,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6532896174863388,
+      "grad_norm": 1.1838295933776124,
+      "learning_rate": 0.00018207042874628817,
+      "loss": 13.5977,
+      "step": 3736
+    },
+    {
+      "epoch": 0.653464480874317,
+      "grad_norm": 1.3329068931868884,
+      "learning_rate": 0.0001820596396219914,
+      "loss": 13.4728,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6536393442622951,
+      "grad_norm": 1.1047840605037362,
+      "learning_rate": 0.00018204884757236437,
+      "loss": 13.1848,
+      "step": 3738
+    },
+    {
+      "epoch": 0.6538142076502732,
+      "grad_norm": 1.1802715417646985,
+      "learning_rate": 0.0001820380525977918,
+      "loss": 13.5279,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6539890710382513,
+      "grad_norm": 1.185507029226543,
+      "learning_rate": 0.00018202725469865846,
+      "loss": 13.5441,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6541639344262296,
+      "grad_norm": 1.1517523258673419,
+      "learning_rate": 0.00018201645387534935,
+      "loss": 13.3516,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6543387978142077,
+      "grad_norm": 1.1615558700527302,
+      "learning_rate": 0.0001820056501282495,
+      "loss": 13.4598,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6545136612021858,
+      "grad_norm": 1.057818890633854,
+      "learning_rate": 0.00018199484345774404,
+      "loss": 13.3196,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6546885245901639,
+      "grad_norm": 1.0544340866230137,
+      "learning_rate": 0.0001819840338642182,
+      "loss": 13.3138,
+      "step": 3744
+    },
+    {
+      "epoch": 0.654863387978142,
+      "grad_norm": 1.0320489798983339,
+      "learning_rate": 0.00018197322134805736,
+      "loss": 13.1261,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6550382513661203,
+      "grad_norm": 1.1869013938328608,
+      "learning_rate": 0.0001819624059096469,
+      "loss": 13.3141,
+      "step": 3746
+    },
+    {
+      "epoch": 0.6552131147540984,
+      "grad_norm": 1.1847153481084907,
+      "learning_rate": 0.00018195158754937248,
+      "loss": 13.2808,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6553879781420765,
+      "grad_norm": 1.1891073281377498,
+      "learning_rate": 0.0001819407662676197,
+      "loss": 13.6148,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6555628415300546,
+      "grad_norm": 1.5338806781654208,
+      "learning_rate": 0.00018192994206477433,
+      "loss": 13.5678,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 1.2355303869693297,
+      "learning_rate": 0.00018191911494122226,
+      "loss": 13.4449,
+      "step": 3750
+    },
+    {
+      "epoch": 0.655912568306011,
+      "grad_norm": 1.1664437583466725,
+      "learning_rate": 0.00018190828489734947,
+      "loss": 13.464,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6560874316939891,
+      "grad_norm": 1.281434650342634,
+      "learning_rate": 0.000181897451933542,
+      "loss": 13.372,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6562622950819672,
+      "grad_norm": 1.0391082326082286,
+      "learning_rate": 0.00018188661605018607,
+      "loss": 13.4007,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6564371584699453,
+      "grad_norm": 1.082682248916343,
+      "learning_rate": 0.00018187577724766794,
+      "loss": 13.4048,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6566120218579234,
+      "grad_norm": 1.1567814873539308,
+      "learning_rate": 0.000181864935526374,
+      "loss": 13.4933,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6567868852459017,
+      "grad_norm": 1.2506238112673018,
+      "learning_rate": 0.0001818540908866908,
+      "loss": 13.4328,
+      "step": 3756
+    },
+    {
+      "epoch": 0.6569617486338798,
+      "grad_norm": 1.1442103436340068,
+      "learning_rate": 0.00018184324332900487,
+      "loss": 13.5872,
+      "step": 3757
+    },
+    {
+      "epoch": 0.6571366120218579,
+      "grad_norm": 1.1487468061652488,
+      "learning_rate": 0.00018183239285370295,
+      "loss": 13.3409,
+      "step": 3758
+    },
+    {
+      "epoch": 0.657311475409836,
+      "grad_norm": 1.0878534428271762,
+      "learning_rate": 0.00018182153946117178,
+      "loss": 13.3938,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6574863387978143,
+      "grad_norm": 1.0622999308885939,
+      "learning_rate": 0.00018181068315179842,
+      "loss": 13.5628,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6576612021857924,
+      "grad_norm": 1.0425531287600538,
+      "learning_rate": 0.00018179982392596973,
+      "loss": 13.493,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6578360655737705,
+      "grad_norm": 1.203244393517811,
+      "learning_rate": 0.0001817889617840729,
+      "loss": 13.4876,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6580109289617486,
+      "grad_norm": 1.051004804546352,
+      "learning_rate": 0.00018177809672649515,
+      "loss": 13.4619,
+      "step": 3763
+    },
+    {
+      "epoch": 0.6581857923497267,
+      "grad_norm": 1.1130405047507759,
+      "learning_rate": 0.0001817672287536238,
+      "loss": 13.4336,
+      "step": 3764
+    },
+    {
+      "epoch": 0.658360655737705,
+      "grad_norm": 1.064121718697386,
+      "learning_rate": 0.0001817563578658463,
+      "loss": 13.5331,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6585355191256831,
+      "grad_norm": 1.101059981699715,
+      "learning_rate": 0.0001817454840635501,
+      "loss": 13.6079,
+      "step": 3766
+    },
+    {
+      "epoch": 0.6587103825136612,
+      "grad_norm": 1.00802228820857,
+      "learning_rate": 0.00018173460734712294,
+      "loss": 13.2871,
+      "step": 3767
+    },
+    {
+      "epoch": 0.6588852459016393,
+      "grad_norm": 1.1643067562863008,
+      "learning_rate": 0.00018172372771695254,
+      "loss": 13.4338,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6590601092896174,
+      "grad_norm": 1.1192192380420396,
+      "learning_rate": 0.0001817128451734267,
+      "loss": 13.4335,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6592349726775957,
+      "grad_norm": 1.1809971194846451,
+      "learning_rate": 0.00018170195971693343,
+      "loss": 13.5848,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6594098360655738,
+      "grad_norm": 1.0796206308429424,
+      "learning_rate": 0.00018169107134786074,
+      "loss": 13.7216,
+      "step": 3771
+    },
+    {
+      "epoch": 0.6595846994535519,
+      "grad_norm": 1.0197802531032276,
+      "learning_rate": 0.00018168018006659681,
+      "loss": 13.1469,
+      "step": 3772
+    },
+    {
+      "epoch": 0.65975956284153,
+      "grad_norm": 1.1248569178349022,
+      "learning_rate": 0.0001816692858735299,
+      "loss": 13.443,
+      "step": 3773
+    },
+    {
+      "epoch": 0.6599344262295082,
+      "grad_norm": 1.2773702282466302,
+      "learning_rate": 0.00018165838876904835,
+      "loss": 13.3293,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6601092896174864,
+      "grad_norm": 1.2401246683346843,
+      "learning_rate": 0.00018164748875354065,
+      "loss": 13.629,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6602841530054645,
+      "grad_norm": 1.1759416402842542,
+      "learning_rate": 0.00018163658582739538,
+      "loss": 13.4036,
+      "step": 3776
+    },
+    {
+      "epoch": 0.6604590163934426,
+      "grad_norm": 1.3356555498295266,
+      "learning_rate": 0.0001816256799910012,
+      "loss": 13.6551,
+      "step": 3777
+    },
+    {
+      "epoch": 0.6606338797814207,
+      "grad_norm": 1.2534773131636325,
+      "learning_rate": 0.0001816147712447469,
+      "loss": 13.4498,
+      "step": 3778
+    },
+    {
+      "epoch": 0.6608087431693989,
+      "grad_norm": 1.0493686941174944,
+      "learning_rate": 0.00018160385958902134,
+      "loss": 13.2949,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6609836065573771,
+      "grad_norm": 1.2986364686216127,
+      "learning_rate": 0.00018159294502421357,
+      "loss": 13.449,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6611584699453552,
+      "grad_norm": 1.1370940387170916,
+      "learning_rate": 0.00018158202755071262,
+      "loss": 13.0474,
+      "step": 3781
+    },
+    {
+      "epoch": 0.6613333333333333,
+      "grad_norm": 1.1699706079568817,
+      "learning_rate": 0.00018157110716890773,
+      "loss": 13.3924,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6615081967213114,
+      "grad_norm": 1.1767569270964835,
+      "learning_rate": 0.00018156018387918815,
+      "loss": 13.1539,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6616830601092896,
+      "grad_norm": 1.3260827619960578,
+      "learning_rate": 0.0001815492576819433,
+      "loss": 13.4347,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6618579234972678,
+      "grad_norm": 1.010370833134153,
+      "learning_rate": 0.0001815383285775627,
+      "loss": 13.6413,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6620327868852459,
+      "grad_norm": 1.2724480042053943,
+      "learning_rate": 0.00018152739656643595,
+      "loss": 13.4413,
+      "step": 3786
+    },
+    {
+      "epoch": 0.662207650273224,
+      "grad_norm": 1.1906459924855848,
+      "learning_rate": 0.0001815164616489528,
+      "loss": 13.588,
+      "step": 3787
+    },
+    {
+      "epoch": 0.6623825136612022,
+      "grad_norm": 1.2957109968584464,
+      "learning_rate": 0.000181505523825503,
+      "loss": 13.3804,
+      "step": 3788
+    },
+    {
+      "epoch": 0.6625573770491803,
+      "grad_norm": 1.055030708814444,
+      "learning_rate": 0.00018149458309647655,
+      "loss": 13.3774,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6627322404371585,
+      "grad_norm": 1.3189602006340833,
+      "learning_rate": 0.00018148363946226336,
+      "loss": 13.2257,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6629071038251366,
+      "grad_norm": 1.0381053685283288,
+      "learning_rate": 0.00018147269292325367,
+      "loss": 13.508,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6630819672131147,
+      "grad_norm": 1.4891761389826392,
+      "learning_rate": 0.00018146174347983768,
+      "loss": 13.4561,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6632568306010929,
+      "grad_norm": 1.6551233085477497,
+      "learning_rate": 0.00018145079113240567,
+      "loss": 13.4023,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6634316939890711,
+      "grad_norm": 1.0044385269896619,
+      "learning_rate": 0.00018143983588134814,
+      "loss": 13.3892,
+      "step": 3794
+    },
+    {
+      "epoch": 0.6636065573770492,
+      "grad_norm": 1.401443433389816,
+      "learning_rate": 0.00018142887772705564,
+      "loss": 13.658,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6637814207650273,
+      "grad_norm": 1.08037261942208,
+      "learning_rate": 0.00018141791666991875,
+      "loss": 13.4318,
+      "step": 3796
+    },
+    {
+      "epoch": 0.6639562841530054,
+      "grad_norm": 1.3085838488640709,
+      "learning_rate": 0.00018140695271032825,
+      "loss": 13.5901,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6641311475409836,
+      "grad_norm": 1.1952833244718908,
+      "learning_rate": 0.00018139598584867504,
+      "loss": 13.3695,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6643060109289618,
+      "grad_norm": 1.320976855817679,
+      "learning_rate": 0.00018138501608535,
+      "loss": 13.3925,
+      "step": 3799
+    },
+    {
+      "epoch": 0.6644808743169399,
+      "grad_norm": 1.338778639968441,
+      "learning_rate": 0.00018137404342074426,
+      "loss": 13.2354,
+      "step": 3800
+    },
+    {
+      "epoch": 0.664655737704918,
+      "grad_norm": 1.2376915695428672,
+      "learning_rate": 0.00018136306785524893,
+      "loss": 13.3005,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6648306010928962,
+      "grad_norm": 1.217957045499402,
+      "learning_rate": 0.00018135208938925527,
+      "loss": 13.3238,
+      "step": 3802
+    },
+    {
+      "epoch": 0.6650054644808743,
+      "grad_norm": 1.2281049294898971,
+      "learning_rate": 0.00018134110802315472,
+      "loss": 13.5627,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6651803278688525,
+      "grad_norm": 1.33406095050728,
+      "learning_rate": 0.00018133012375733866,
+      "loss": 13.521,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6653551912568306,
+      "grad_norm": 1.2369546170073358,
+      "learning_rate": 0.00018131913659219872,
+      "loss": 13.5317,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6655300546448087,
+      "grad_norm": 1.195687498857937,
+      "learning_rate": 0.0001813081465281266,
+      "loss": 13.3333,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6657049180327869,
+      "grad_norm": 1.1977762329490032,
+      "learning_rate": 0.00018129715356551403,
+      "loss": 13.3303,
+      "step": 3807
+    },
+    {
+      "epoch": 0.665879781420765,
+      "grad_norm": 1.126610337089159,
+      "learning_rate": 0.00018128615770475294,
+      "loss": 13.5821,
+      "step": 3808
+    },
+    {
+      "epoch": 0.6660546448087432,
+      "grad_norm": 1.0888054391085429,
+      "learning_rate": 0.00018127515894623527,
+      "loss": 13.3369,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6662295081967213,
+      "grad_norm": 1.2257581268162514,
+      "learning_rate": 0.0001812641572903532,
+      "loss": 13.8483,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6664043715846995,
+      "grad_norm": 1.2773124181926458,
+      "learning_rate": 0.00018125315273749882,
+      "loss": 13.5742,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6665792349726776,
+      "grad_norm": 1.2354842257641534,
+      "learning_rate": 0.00018124214528806448,
+      "loss": 13.2712,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6667540983606557,
+      "grad_norm": 1.3395341843444324,
+      "learning_rate": 0.00018123113494244262,
+      "loss": 13.4118,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6669289617486339,
+      "grad_norm": 1.5148046094482195,
+      "learning_rate": 0.00018122012170102567,
+      "loss": 13.5459,
+      "step": 3814
+    },
+    {
+      "epoch": 0.667103825136612,
+      "grad_norm": 1.1980572624945183,
+      "learning_rate": 0.00018120910556420627,
+      "loss": 13.6799,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6672786885245902,
+      "grad_norm": 1.2772754338726642,
+      "learning_rate": 0.00018119808653237717,
+      "loss": 13.1999,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6674535519125683,
+      "grad_norm": 1.0693970733098404,
+      "learning_rate": 0.00018118706460593114,
+      "loss": 13.4722,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6676284153005464,
+      "grad_norm": 1.3072619358809028,
+      "learning_rate": 0.00018117603978526115,
+      "loss": 13.2107,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6678032786885246,
+      "grad_norm": 1.1626129707748176,
+      "learning_rate": 0.0001811650120707601,
+      "loss": 13.4421,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6679781420765027,
+      "grad_norm": 1.1771226781984048,
+      "learning_rate": 0.00018115398146282128,
+      "loss": 13.1805,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6681530054644809,
+      "grad_norm": 1.3287228622151057,
+      "learning_rate": 0.00018114294796183782,
+      "loss": 13.3254,
+      "step": 3821
+    },
+    {
+      "epoch": 0.668327868852459,
+      "grad_norm": 1.120982000595319,
+      "learning_rate": 0.00018113191156820304,
+      "loss": 13.0879,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6685027322404371,
+      "grad_norm": 1.355574144256038,
+      "learning_rate": 0.00018112087228231043,
+      "loss": 13.1414,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6686775956284153,
+      "grad_norm": 1.2287212459670547,
+      "learning_rate": 0.00018110983010455353,
+      "loss": 13.551,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6688524590163935,
+      "grad_norm": 1.0990709270680643,
+      "learning_rate": 0.00018109878503532592,
+      "loss": 13.177,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6690273224043716,
+      "grad_norm": 1.2068989945774982,
+      "learning_rate": 0.00018108773707502137,
+      "loss": 13.4769,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6692021857923497,
+      "grad_norm": 1.0931546963553338,
+      "learning_rate": 0.00018107668622403375,
+      "loss": 13.4482,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6693770491803279,
+      "grad_norm": 1.1128025658152356,
+      "learning_rate": 0.00018106563248275698,
+      "loss": 13.364,
+      "step": 3828
+    },
+    {
+      "epoch": 0.669551912568306,
+      "grad_norm": 0.9975227955051569,
+      "learning_rate": 0.00018105457585158513,
+      "loss": 13.6404,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6697267759562842,
+      "grad_norm": 1.0779037317766336,
+      "learning_rate": 0.00018104351633091236,
+      "loss": 13.3381,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6699016393442623,
+      "grad_norm": 1.064452189690793,
+      "learning_rate": 0.00018103245392113294,
+      "loss": 13.3197,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6700765027322404,
+      "grad_norm": 1.0218727401550984,
+      "learning_rate": 0.0001810213886226412,
+      "loss": 13.4283,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6702513661202186,
+      "grad_norm": 1.1845556133381274,
+      "learning_rate": 0.0001810103204358316,
+      "loss": 13.5132,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6704262295081967,
+      "grad_norm": 1.204343373192663,
+      "learning_rate": 0.00018099924936109877,
+      "loss": 13.5258,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6706010928961749,
+      "grad_norm": 1.0676342748900296,
+      "learning_rate": 0.00018098817539883728,
+      "loss": 13.4079,
+      "step": 3835
+    },
+    {
+      "epoch": 0.670775956284153,
+      "grad_norm": 1.0222211530205398,
+      "learning_rate": 0.000180977098549442,
+      "loss": 13.3466,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6709508196721311,
+      "grad_norm": 1.0069185771075608,
+      "learning_rate": 0.00018096601881330775,
+      "loss": 13.4396,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6711256830601093,
+      "grad_norm": 1.1252118636174233,
+      "learning_rate": 0.00018095493619082955,
+      "loss": 13.2987,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6713005464480875,
+      "grad_norm": 1.102846537631715,
+      "learning_rate": 0.00018094385068240243,
+      "loss": 13.5624,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6714754098360656,
+      "grad_norm": 1.0880013944698927,
+      "learning_rate": 0.00018093276228842164,
+      "loss": 13.3342,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6716502732240437,
+      "grad_norm": 1.2155642913056222,
+      "learning_rate": 0.00018092167100928238,
+      "loss": 13.4164,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6718251366120218,
+      "grad_norm": 1.084664208310918,
+      "learning_rate": 0.00018091057684538017,
+      "loss": 13.4527,
+      "step": 3842
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 1.0912993780535072,
+      "learning_rate": 0.00018089947979711037,
+      "loss": 13.4663,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6721748633879782,
+      "grad_norm": 1.0457100288707268,
+      "learning_rate": 0.00018088837986486865,
+      "loss": 13.2387,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6723497267759563,
+      "grad_norm": 1.082980490133716,
+      "learning_rate": 0.00018087727704905074,
+      "loss": 13.6641,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6725245901639344,
+      "grad_norm": 1.1388130779408356,
+      "learning_rate": 0.00018086617135005235,
+      "loss": 13.5832,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6726994535519125,
+      "grad_norm": 1.2022044088719626,
+      "learning_rate": 0.00018085506276826945,
+      "loss": 13.5247,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6728743169398907,
+      "grad_norm": 1.1100006129176359,
+      "learning_rate": 0.00018084395130409803,
+      "loss": 13.0189,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6730491803278689,
+      "grad_norm": 1.3653192367493996,
+      "learning_rate": 0.00018083283695793418,
+      "loss": 13.058,
+      "step": 3849
+    },
+    {
+      "epoch": 0.673224043715847,
+      "grad_norm": 1.2270434541594915,
+      "learning_rate": 0.00018082171973017418,
+      "loss": 13.4213,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6733989071038251,
+      "grad_norm": 1.0918972595532943,
+      "learning_rate": 0.00018081059962121425,
+      "loss": 13.404,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6735737704918032,
+      "grad_norm": 1.3369359190676018,
+      "learning_rate": 0.0001807994766314509,
+      "loss": 13.2498,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6737486338797815,
+      "grad_norm": 1.3013963730122449,
+      "learning_rate": 0.0001807883507612806,
+      "loss": 13.5067,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6739234972677596,
+      "grad_norm": 1.209963909043034,
+      "learning_rate": 0.00018077722201109995,
+      "loss": 13.5344,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6740983606557377,
+      "grad_norm": 1.2455309662890324,
+      "learning_rate": 0.00018076609038130578,
+      "loss": 13.666,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6742732240437158,
+      "grad_norm": 1.2616497420903006,
+      "learning_rate": 0.00018075495587229482,
+      "loss": 13.3517,
+      "step": 3856
+    },
+    {
+      "epoch": 0.674448087431694,
+      "grad_norm": 1.1361485087349283,
+      "learning_rate": 0.00018074381848446404,
+      "loss": 13.4781,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6746229508196722,
+      "grad_norm": 1.1025779657099908,
+      "learning_rate": 0.00018073267821821048,
+      "loss": 13.2756,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6747978142076503,
+      "grad_norm": 1.370841271995651,
+      "learning_rate": 0.00018072153507393124,
+      "loss": 13.5866,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6749726775956284,
+      "grad_norm": 1.325505509143198,
+      "learning_rate": 0.0001807103890520236,
+      "loss": 13.4162,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6751475409836065,
+      "grad_norm": 1.051733917724121,
+      "learning_rate": 0.0001806992401528849,
+      "loss": 13.1435,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6753224043715848,
+      "grad_norm": 1.3086916411624554,
+      "learning_rate": 0.00018068808837691256,
+      "loss": 13.6124,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6754972677595629,
+      "grad_norm": 1.041553110958841,
+      "learning_rate": 0.00018067693372450413,
+      "loss": 13.1936,
+      "step": 3863
+    },
+    {
+      "epoch": 0.675672131147541,
+      "grad_norm": 1.2685206886203493,
+      "learning_rate": 0.00018066577619605732,
+      "loss": 13.251,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6758469945355191,
+      "grad_norm": 1.0978548558954826,
+      "learning_rate": 0.00018065461579196978,
+      "loss": 13.3597,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6760218579234972,
+      "grad_norm": 1.2601469926293265,
+      "learning_rate": 0.00018064345251263947,
+      "loss": 13.289,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6761967213114755,
+      "grad_norm": 1.2081627795173022,
+      "learning_rate": 0.00018063228635846427,
+      "loss": 13.5965,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6763715846994536,
+      "grad_norm": 1.0850343316779967,
+      "learning_rate": 0.0001806211173298423,
+      "loss": 13.6295,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6765464480874317,
+      "grad_norm": 1.1671705647742512,
+      "learning_rate": 0.00018060994542717167,
+      "loss": 13.4058,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6767213114754098,
+      "grad_norm": 1.0423261523451195,
+      "learning_rate": 0.0001805987706508507,
+      "loss": 13.2693,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6768961748633879,
+      "grad_norm": 1.2815451003101856,
+      "learning_rate": 0.00018058759300127768,
+      "loss": 13.6882,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6770710382513662,
+      "grad_norm": 1.1837753968435243,
+      "learning_rate": 0.00018057641247885115,
+      "loss": 13.514,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6772459016393443,
+      "grad_norm": 1.3390893739888792,
+      "learning_rate": 0.00018056522908396964,
+      "loss": 13.5609,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6774207650273224,
+      "grad_norm": 1.1994427280844568,
+      "learning_rate": 0.00018055404281703185,
+      "loss": 13.2628,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6775956284153005,
+      "grad_norm": 1.1173276260890046,
+      "learning_rate": 0.00018054285367843657,
+      "loss": 13.2949,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6777704918032786,
+      "grad_norm": 1.2569508517846402,
+      "learning_rate": 0.00018053166166858263,
+      "loss": 13.1001,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6779453551912569,
+      "grad_norm": 1.092274956322442,
+      "learning_rate": 0.00018052046678786909,
+      "loss": 13.5342,
+      "step": 3877
+    },
+    {
+      "epoch": 0.678120218579235,
+      "grad_norm": 1.047754371859823,
+      "learning_rate": 0.00018050926903669497,
+      "loss": 13.5213,
+      "step": 3878
+    },
+    {
+      "epoch": 0.6782950819672131,
+      "grad_norm": 1.115942927876703,
+      "learning_rate": 0.00018049806841545943,
+      "loss": 13.3881,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6784699453551912,
+      "grad_norm": 1.193625729316959,
+      "learning_rate": 0.00018048686492456187,
+      "loss": 13.5048,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6786448087431693,
+      "grad_norm": 1.1077549046436428,
+      "learning_rate": 0.00018047565856440154,
+      "loss": 13.289,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6788196721311476,
+      "grad_norm": 1.2691380679258033,
+      "learning_rate": 0.00018046444933537807,
+      "loss": 13.6367,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6789945355191257,
+      "grad_norm": 1.1894186357820593,
+      "learning_rate": 0.00018045323723789097,
+      "loss": 13.245,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6791693989071038,
+      "grad_norm": 1.4661207183263636,
+      "learning_rate": 0.00018044202227233996,
+      "loss": 13.4297,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6793442622950819,
+      "grad_norm": 1.1888640342260755,
+      "learning_rate": 0.00018043080443912483,
+      "loss": 13.4276,
+      "step": 3885
+    },
+    {
+      "epoch": 0.67951912568306,
+      "grad_norm": 1.3472429717052838,
+      "learning_rate": 0.00018041958373864552,
+      "loss": 13.6146,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6796939890710383,
+      "grad_norm": 1.2727945288085705,
+      "learning_rate": 0.000180408360171302,
+      "loss": 13.4829,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6798688524590164,
+      "grad_norm": 1.1238049679360107,
+      "learning_rate": 0.0001803971337374944,
+      "loss": 13.5353,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6800437158469945,
+      "grad_norm": 1.241967845259122,
+      "learning_rate": 0.00018038590443762291,
+      "loss": 13.4498,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6802185792349726,
+      "grad_norm": 1.3013421062733888,
+      "learning_rate": 0.00018037467227208782,
+      "loss": 13.3626,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6803934426229509,
+      "grad_norm": 1.4061957447343,
+      "learning_rate": 0.0001803634372412896,
+      "loss": 13.5474,
+      "step": 3891
+    },
+    {
+      "epoch": 0.680568306010929,
+      "grad_norm": 1.1578821092664744,
+      "learning_rate": 0.00018035219934562873,
+      "loss": 13.3081,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6807431693989071,
+      "grad_norm": 1.4364332466737597,
+      "learning_rate": 0.0001803409585855058,
+      "loss": 13.5539,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6809180327868852,
+      "grad_norm": 1.0629691820970084,
+      "learning_rate": 0.00018032971496132163,
+      "loss": 13.3797,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6810928961748633,
+      "grad_norm": 1.3602603341771533,
+      "learning_rate": 0.00018031846847347692,
+      "loss": 13.2482,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6812677595628416,
+      "grad_norm": 1.110655375748488,
+      "learning_rate": 0.0001803072191223727,
+      "loss": 13.5043,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6814426229508197,
+      "grad_norm": 1.3858146365348956,
+      "learning_rate": 0.00018029596690840994,
+      "loss": 13.536,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6816174863387978,
+      "grad_norm": 1.1195740541649162,
+      "learning_rate": 0.00018028471183198974,
+      "loss": 13.5416,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6817923497267759,
+      "grad_norm": 1.1874954781653697,
+      "learning_rate": 0.0001802734538935134,
+      "loss": 13.2698,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6819672131147541,
+      "grad_norm": 1.0265007334133331,
+      "learning_rate": 0.0001802621930933822,
+      "loss": 13.2089,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6821420765027323,
+      "grad_norm": 1.1878010139410944,
+      "learning_rate": 0.00018025092943199758,
+      "loss": 13.4009,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6823169398907104,
+      "grad_norm": 1.4324303902205868,
+      "learning_rate": 0.00018023966290976114,
+      "loss": 13.597,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6824918032786885,
+      "grad_norm": 1.1309232140957488,
+      "learning_rate": 0.00018022839352707442,
+      "loss": 13.1226,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6826666666666666,
+      "grad_norm": 1.514264892611927,
+      "learning_rate": 0.00018021712128433922,
+      "loss": 13.5352,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6828415300546448,
+      "grad_norm": 1.0469341288986134,
+      "learning_rate": 0.00018020584618195736,
+      "loss": 13.5383,
+      "step": 3905
+    },
+    {
+      "epoch": 0.683016393442623,
+      "grad_norm": 1.5584125902687858,
+      "learning_rate": 0.00018019456822033082,
+      "loss": 13.4541,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6831912568306011,
+      "grad_norm": 1.06896565465885,
+      "learning_rate": 0.00018018328739986158,
+      "loss": 13.406,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6833661202185792,
+      "grad_norm": 1.6381901190982158,
+      "learning_rate": 0.00018017200372095186,
+      "loss": 13.4273,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6835409836065574,
+      "grad_norm": 1.2937433622934862,
+      "learning_rate": 0.00018016071718400387,
+      "loss": 13.4334,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6837158469945355,
+      "grad_norm": 1.2273651306790268,
+      "learning_rate": 0.00018014942778941996,
+      "loss": 13.3669,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6838907103825137,
+      "grad_norm": 1.2513772428545242,
+      "learning_rate": 0.00018013813553760261,
+      "loss": 13.2912,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6840655737704918,
+      "grad_norm": 1.1587820092859185,
+      "learning_rate": 0.00018012684042895436,
+      "loss": 13.3518,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6842404371584699,
+      "grad_norm": 1.2943629845823392,
+      "learning_rate": 0.00018011554246387786,
+      "loss": 13.5357,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6844153005464481,
+      "grad_norm": 1.0819847297859309,
+      "learning_rate": 0.00018010424164277588,
+      "loss": 13.5118,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6845901639344262,
+      "grad_norm": 1.271529468620181,
+      "learning_rate": 0.00018009293796605126,
+      "loss": 13.5365,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6847650273224044,
+      "grad_norm": 1.1627637967344349,
+      "learning_rate": 0.000180081631434107,
+      "loss": 13.4493,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6849398907103825,
+      "grad_norm": 1.292398587029334,
+      "learning_rate": 0.00018007032204734612,
+      "loss": 13.2837,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6851147540983606,
+      "grad_norm": 1.1614721552213714,
+      "learning_rate": 0.00018005900980617184,
+      "loss": 13.2874,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6852896174863388,
+      "grad_norm": 1.1241093426697135,
+      "learning_rate": 0.00018004769471098738,
+      "loss": 13.3722,
+      "step": 3919
+    },
+    {
+      "epoch": 0.685464480874317,
+      "grad_norm": 1.1293809949028246,
+      "learning_rate": 0.00018003637676219616,
+      "loss": 13.2029,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6856393442622951,
+      "grad_norm": 0.984074673216755,
+      "learning_rate": 0.0001800250559602016,
+      "loss": 13.3006,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6858142076502732,
+      "grad_norm": 1.1958602182197688,
+      "learning_rate": 0.00018001373230540729,
+      "loss": 13.2654,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6859890710382514,
+      "grad_norm": 1.2360180312640843,
+      "learning_rate": 0.0001800024057982169,
+      "loss": 13.344,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6861639344262295,
+      "grad_norm": 1.0957709872816717,
+      "learning_rate": 0.00017999107643903423,
+      "loss": 13.1055,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6863387978142077,
+      "grad_norm": 1.3782218901315366,
+      "learning_rate": 0.00017997974422826317,
+      "loss": 13.1256,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6865136612021858,
+      "grad_norm": 1.2734130651921989,
+      "learning_rate": 0.00017996840916630763,
+      "loss": 13.4518,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6866885245901639,
+      "grad_norm": 1.2387651406576121,
+      "learning_rate": 0.00017995707125357174,
+      "loss": 13.3218,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6868633879781421,
+      "grad_norm": 1.039708327196395,
+      "learning_rate": 0.00017994573049045972,
+      "loss": 13.1454,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6870382513661202,
+      "grad_norm": 1.1282927110326888,
+      "learning_rate": 0.00017993438687737579,
+      "loss": 13.4545,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6872131147540984,
+      "grad_norm": 1.044138526498366,
+      "learning_rate": 0.00017992304041472435,
+      "loss": 13.4369,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6873879781420765,
+      "grad_norm": 1.1554446084171837,
+      "learning_rate": 0.00017991169110290992,
+      "loss": 13.4966,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6875628415300546,
+      "grad_norm": 1.174151066152567,
+      "learning_rate": 0.00017990033894233704,
+      "loss": 13.4358,
+      "step": 3932
+    },
+    {
+      "epoch": 0.6877377049180328,
+      "grad_norm": 1.0465266837426657,
+      "learning_rate": 0.00017988898393341045,
+      "loss": 13.5314,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6879125683060109,
+      "grad_norm": 1.068567060317609,
+      "learning_rate": 0.00017987762607653493,
+      "loss": 13.4022,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6880874316939891,
+      "grad_norm": 1.126974752712792,
+      "learning_rate": 0.00017986626537211536,
+      "loss": 13.6007,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6882622950819672,
+      "grad_norm": 1.0286176261713011,
+      "learning_rate": 0.00017985490182055672,
+      "loss": 13.4992,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6884371584699454,
+      "grad_norm": 1.174810163319476,
+      "learning_rate": 0.00017984353542226415,
+      "loss": 13.5726,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6886120218579235,
+      "grad_norm": 1.0688529914573788,
+      "learning_rate": 0.00017983216617764285,
+      "loss": 13.481,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6887868852459016,
+      "grad_norm": 1.0237061784025459,
+      "learning_rate": 0.00017982079408709808,
+      "loss": 13.4897,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6889617486338798,
+      "grad_norm": 1.1432744315789636,
+      "learning_rate": 0.00017980941915103526,
+      "loss": 13.3116,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6891366120218579,
+      "grad_norm": 1.0867127128582394,
+      "learning_rate": 0.0001797980413698599,
+      "loss": 13.376,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6893114754098361,
+      "grad_norm": 1.2100880770373368,
+      "learning_rate": 0.00017978666074397757,
+      "loss": 13.1917,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6894863387978142,
+      "grad_norm": 1.0562634117560812,
+      "learning_rate": 0.00017977527727379406,
+      "loss": 13.3835,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6896612021857923,
+      "grad_norm": 1.385800277488955,
+      "learning_rate": 0.0001797638909597151,
+      "loss": 13.215,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6898360655737705,
+      "grad_norm": 1.3926513937025804,
+      "learning_rate": 0.0001797525018021466,
+      "loss": 13.4665,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6900109289617486,
+      "grad_norm": 1.2794978687676886,
+      "learning_rate": 0.0001797411098014946,
+      "loss": 13.6634,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6901857923497268,
+      "grad_norm": 1.3662077330189604,
+      "learning_rate": 0.0001797297149581652,
+      "loss": 13.3612,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6903606557377049,
+      "grad_norm": 1.1819596062296613,
+      "learning_rate": 0.00017971831727256466,
+      "loss": 13.4935,
+      "step": 3948
+    },
+    {
+      "epoch": 0.690535519125683,
+      "grad_norm": 1.409489649711722,
+      "learning_rate": 0.0001797069167450992,
+      "loss": 13.245,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6907103825136612,
+      "grad_norm": 1.1302886782252874,
+      "learning_rate": 0.0001796955133761753,
+      "loss": 13.2985,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6908852459016394,
+      "grad_norm": 1.3456313222292011,
+      "learning_rate": 0.00017968410716619947,
+      "loss": 13.3604,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6910601092896175,
+      "grad_norm": 1.1257966425666175,
+      "learning_rate": 0.0001796726981155783,
+      "loss": 13.1432,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6912349726775956,
+      "grad_norm": 1.1845134444356873,
+      "learning_rate": 0.00017966128622471855,
+      "loss": 13.4193,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6914098360655738,
+      "grad_norm": 1.1412565841233737,
+      "learning_rate": 0.00017964987149402702,
+      "loss": 13.4525,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6915846994535519,
+      "grad_norm": 1.1473328330690522,
+      "learning_rate": 0.00017963845392391064,
+      "loss": 13.4512,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6917595628415301,
+      "grad_norm": 1.1309670996075774,
+      "learning_rate": 0.00017962703351477642,
+      "loss": 13.4592,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6919344262295082,
+      "grad_norm": 1.1158995524491335,
+      "learning_rate": 0.00017961561026703146,
+      "loss": 13.3278,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6921092896174863,
+      "grad_norm": 1.0658333920092482,
+      "learning_rate": 0.00017960418418108305,
+      "loss": 13.3177,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6922841530054645,
+      "grad_norm": 1.13089959671713,
+      "learning_rate": 0.00017959275525733845,
+      "loss": 13.4333,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6924590163934427,
+      "grad_norm": 1.2476183259432887,
+      "learning_rate": 0.00017958132349620517,
+      "loss": 13.3367,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6926338797814208,
+      "grad_norm": 1.0997783881264516,
+      "learning_rate": 0.00017956988889809063,
+      "loss": 13.6874,
+      "step": 3961
+    },
+    {
+      "epoch": 0.6928087431693989,
+      "grad_norm": 1.1972757952907462,
+      "learning_rate": 0.00017955845146340254,
+      "loss": 13.3413,
+      "step": 3962
+    },
+    {
+      "epoch": 0.692983606557377,
+      "grad_norm": 1.156185696141994,
+      "learning_rate": 0.00017954701119254863,
+      "loss": 13.7006,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6931584699453552,
+      "grad_norm": 1.1972528639734745,
+      "learning_rate": 0.00017953556808593667,
+      "loss": 13.5323,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 1.2252611554633501,
+      "learning_rate": 0.00017952412214397467,
+      "loss": 13.5772,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6935081967213115,
+      "grad_norm": 1.108528319057744,
+      "learning_rate": 0.00017951267336707062,
+      "loss": 12.8934,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6936830601092896,
+      "grad_norm": 1.296013154562118,
+      "learning_rate": 0.00017950122175563267,
+      "loss": 13.6614,
+      "step": 3967
+    },
+    {
+      "epoch": 0.6938579234972677,
+      "grad_norm": 1.1326948760669266,
+      "learning_rate": 0.00017948976731006904,
+      "loss": 13.2392,
+      "step": 3968
+    },
+    {
+      "epoch": 0.694032786885246,
+      "grad_norm": 1.2763048644323,
+      "learning_rate": 0.00017947831003078808,
+      "loss": 13.5237,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6942076502732241,
+      "grad_norm": 1.0739245194678677,
+      "learning_rate": 0.00017946684991819822,
+      "loss": 13.2261,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6943825136612022,
+      "grad_norm": 1.3481444785240313,
+      "learning_rate": 0.00017945538697270802,
+      "loss": 13.5524,
+      "step": 3971
+    },
+    {
+      "epoch": 0.6945573770491803,
+      "grad_norm": 1.0980041923109707,
+      "learning_rate": 0.00017944392119472613,
+      "loss": 13.5924,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6947322404371584,
+      "grad_norm": 1.118242430731248,
+      "learning_rate": 0.00017943245258466125,
+      "loss": 13.303,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6949071038251367,
+      "grad_norm": 1.1831946693025341,
+      "learning_rate": 0.00017942098114292224,
+      "loss": 13.5655,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6950819672131148,
+      "grad_norm": 1.2548359967946645,
+      "learning_rate": 0.00017940950686991803,
+      "loss": 13.3103,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6952568306010929,
+      "grad_norm": 1.0481372201340613,
+      "learning_rate": 0.0001793980297660577,
+      "loss": 13.4017,
+      "step": 3976
+    },
+    {
+      "epoch": 0.695431693989071,
+      "grad_norm": 1.1045456784670769,
+      "learning_rate": 0.00017938654983175035,
+      "loss": 13.406,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6956065573770491,
+      "grad_norm": 1.3240089933637378,
+      "learning_rate": 0.0001793750670674053,
+      "loss": 13.4742,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6957814207650274,
+      "grad_norm": 1.0766601101406037,
+      "learning_rate": 0.0001793635814734318,
+      "loss": 13.3871,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6959562841530055,
+      "grad_norm": 1.226864912995238,
+      "learning_rate": 0.0001793520930502394,
+      "loss": 13.444,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6961311475409836,
+      "grad_norm": 1.0333597746521521,
+      "learning_rate": 0.00017934060179823755,
+      "loss": 13.3952,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6963060109289617,
+      "grad_norm": 1.2498828372283037,
+      "learning_rate": 0.00017932910771783597,
+      "loss": 13.518,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6964808743169398,
+      "grad_norm": 1.1573776269513478,
+      "learning_rate": 0.00017931761080944436,
+      "loss": 13.6366,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6966557377049181,
+      "grad_norm": 1.0803472233503593,
+      "learning_rate": 0.0001793061110734726,
+      "loss": 13.4077,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6968306010928962,
+      "grad_norm": 1.1488393240794943,
+      "learning_rate": 0.00017929460851033067,
+      "loss": 13.5574,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6970054644808743,
+      "grad_norm": 1.284135828686321,
+      "learning_rate": 0.00017928310312042856,
+      "loss": 13.378,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6971803278688524,
+      "grad_norm": 1.0878710788429735,
+      "learning_rate": 0.00017927159490417648,
+      "loss": 13.6206,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6973551912568307,
+      "grad_norm": 1.3443211694094381,
+      "learning_rate": 0.00017926008386198467,
+      "loss": 13.4135,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6975300546448088,
+      "grad_norm": 1.116179132627872,
+      "learning_rate": 0.00017924856999426345,
+      "loss": 13.1214,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6977049180327869,
+      "grad_norm": 1.2689523973726184,
+      "learning_rate": 0.00017923705330142328,
+      "loss": 13.4437,
+      "step": 3990
+    },
+    {
+      "epoch": 0.697879781420765,
+      "grad_norm": 1.2325999141243518,
+      "learning_rate": 0.0001792255337838748,
+      "loss": 13.1683,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6980546448087431,
+      "grad_norm": 1.691046173173216,
+      "learning_rate": 0.00017921401144202855,
+      "loss": 13.5388,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6982295081967214,
+      "grad_norm": 1.2672191223481781,
+      "learning_rate": 0.00017920248627629538,
+      "loss": 13.3659,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6984043715846995,
+      "grad_norm": 1.7200925774078981,
+      "learning_rate": 0.0001791909582870861,
+      "loss": 13.498,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6985792349726776,
+      "grad_norm": 1.147168794874349,
+      "learning_rate": 0.0001791794274748117,
+      "loss": 13.2973,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6987540983606557,
+      "grad_norm": 1.579691663989506,
+      "learning_rate": 0.00017916789383988317,
+      "loss": 13.3282,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6989289617486338,
+      "grad_norm": 1.1063696032734527,
+      "learning_rate": 0.00017915635738271178,
+      "loss": 13.0535,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6991038251366121,
+      "grad_norm": 1.0740641832641598,
+      "learning_rate": 0.00017914481810370872,
+      "loss": 13.3404,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6992786885245902,
+      "grad_norm": 1.187461475149742,
+      "learning_rate": 0.00017913327600328532,
+      "loss": 13.3083,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6994535519125683,
+      "grad_norm": 1.0489118347493438,
+      "learning_rate": 0.00017912173108185315,
+      "loss": 13.4758,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6996284153005464,
+      "grad_norm": 1.1410384308440231,
+      "learning_rate": 0.00017911018333982367,
+      "loss": 13.3755,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6998032786885245,
+      "grad_norm": 1.0917731840047957,
+      "learning_rate": 0.00017909863277760862,
+      "loss": 13.5652,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6999781420765028,
+      "grad_norm": 1.0879459690434867,
+      "learning_rate": 0.0001790870793956197,
+      "loss": 13.6662,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7001530054644809,
+      "grad_norm": 1.0873357165694086,
+      "learning_rate": 0.00017907552319426878,
+      "loss": 13.5234,
+      "step": 4004
+    },
+    {
+      "epoch": 0.700327868852459,
+      "grad_norm": 1.044244946292674,
+      "learning_rate": 0.00017906396417396788,
+      "loss": 13.578,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7005027322404371,
+      "grad_norm": 1.0153768218586114,
+      "learning_rate": 0.00017905240233512905,
+      "loss": 13.6568,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7006775956284153,
+      "grad_norm": 1.0283479922133336,
+      "learning_rate": 0.0001790408376781644,
+      "loss": 13.2282,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7008524590163935,
+      "grad_norm": 1.0912927578305027,
+      "learning_rate": 0.00017902927020348628,
+      "loss": 13.5446,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7010273224043716,
+      "grad_norm": 1.3301818913271068,
+      "learning_rate": 0.00017901769991150696,
+      "loss": 13.4335,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7012021857923497,
+      "grad_norm": 1.1618926643465526,
+      "learning_rate": 0.000179006126802639,
+      "loss": 13.448,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7013770491803278,
+      "grad_norm": 1.2730042201683287,
+      "learning_rate": 0.0001789945508772949,
+      "loss": 13.318,
+      "step": 4011
+    },
+    {
+      "epoch": 0.701551912568306,
+      "grad_norm": 1.1887058650515676,
+      "learning_rate": 0.0001789829721358874,
+      "loss": 13.4871,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7017267759562842,
+      "grad_norm": 1.15674873603848,
+      "learning_rate": 0.0001789713905788292,
+      "loss": 13.5488,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7019016393442623,
+      "grad_norm": 1.2352532422551104,
+      "learning_rate": 0.0001789598062065332,
+      "loss": 13.5241,
+      "step": 4014
+    },
+    {
+      "epoch": 0.7020765027322404,
+      "grad_norm": 1.0344371929037872,
+      "learning_rate": 0.00017894821901941232,
+      "loss": 13.3964,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7022513661202185,
+      "grad_norm": 1.014328820956659,
+      "learning_rate": 0.00017893662901787973,
+      "loss": 13.3394,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7024262295081968,
+      "grad_norm": 1.187594110741848,
+      "learning_rate": 0.00017892503620234852,
+      "loss": 13.0745,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7026010928961749,
+      "grad_norm": 1.0648237539924639,
+      "learning_rate": 0.00017891344057323198,
+      "loss": 13.3388,
+      "step": 4018
+    },
+    {
+      "epoch": 0.702775956284153,
+      "grad_norm": 1.131727009125832,
+      "learning_rate": 0.00017890184213094348,
+      "loss": 13.2692,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7029508196721311,
+      "grad_norm": 1.1177775434601154,
+      "learning_rate": 0.00017889024087589653,
+      "loss": 13.4509,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7031256830601093,
+      "grad_norm": 1.051170586922452,
+      "learning_rate": 0.00017887863680850463,
+      "loss": 13.2689,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7033005464480875,
+      "grad_norm": 1.0293267813473057,
+      "learning_rate": 0.00017886702992918148,
+      "loss": 13.2114,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7034754098360656,
+      "grad_norm": 1.1512590923759398,
+      "learning_rate": 0.00017885542023834088,
+      "loss": 13.3532,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7036502732240437,
+      "grad_norm": 1.288361642796465,
+      "learning_rate": 0.00017884380773639668,
+      "loss": 13.4026,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7038251366120218,
+      "grad_norm": 1.0791481549775508,
+      "learning_rate": 0.00017883219242376285,
+      "loss": 13.567,
+      "step": 4025
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.074966995875059,
+      "learning_rate": 0.00017882057430085348,
+      "loss": 13.4776,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7041748633879782,
+      "grad_norm": 0.9895300249371727,
+      "learning_rate": 0.0001788089533680827,
+      "loss": 13.1461,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7043497267759563,
+      "grad_norm": 1.0750812737025062,
+      "learning_rate": 0.00017879732962586484,
+      "loss": 13.5289,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7045245901639344,
+      "grad_norm": 1.009335240955156,
+      "learning_rate": 0.00017878570307461425,
+      "loss": 13.4641,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7046994535519125,
+      "grad_norm": 1.0652945496627086,
+      "learning_rate": 0.00017877407371474536,
+      "loss": 13.5119,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7048743169398907,
+      "grad_norm": 1.0424926301648998,
+      "learning_rate": 0.0001787624415466728,
+      "loss": 13.2672,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7050491803278689,
+      "grad_norm": 1.094350375768725,
+      "learning_rate": 0.00017875080657081123,
+      "loss": 13.1526,
+      "step": 4032
+    },
+    {
+      "epoch": 0.705224043715847,
+      "grad_norm": 1.1598678278571657,
+      "learning_rate": 0.00017873916878757539,
+      "loss": 13.2519,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7053989071038251,
+      "grad_norm": 1.1190258625039073,
+      "learning_rate": 0.0001787275281973802,
+      "loss": 13.329,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7055737704918033,
+      "grad_norm": 1.0292349640941165,
+      "learning_rate": 0.00017871588480064058,
+      "loss": 13.4495,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7057486338797814,
+      "grad_norm": 0.971405773264925,
+      "learning_rate": 0.00017870423859777168,
+      "loss": 12.9148,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7059234972677596,
+      "grad_norm": 1.1306973165695946,
+      "learning_rate": 0.00017869258958918863,
+      "loss": 13.065,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7060983606557377,
+      "grad_norm": 1.0528436592528554,
+      "learning_rate": 0.00017868093777530667,
+      "loss": 13.4466,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7062732240437158,
+      "grad_norm": 1.1346719830505816,
+      "learning_rate": 0.00017866928315654124,
+      "loss": 13.464,
+      "step": 4039
+    },
+    {
+      "epoch": 0.706448087431694,
+      "grad_norm": 1.1164489048788415,
+      "learning_rate": 0.00017865762573330776,
+      "loss": 13.4602,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7066229508196721,
+      "grad_norm": 1.1666373514046584,
+      "learning_rate": 0.00017864596550602185,
+      "loss": 13.1349,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7067978142076503,
+      "grad_norm": 1.303026899107083,
+      "learning_rate": 0.0001786343024750991,
+      "loss": 13.4619,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7069726775956284,
+      "grad_norm": 1.0788492447269453,
+      "learning_rate": 0.0001786226366409554,
+      "loss": 13.23,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7071475409836065,
+      "grad_norm": 1.087290506214695,
+      "learning_rate": 0.00017861096800400654,
+      "loss": 13.3915,
+      "step": 4044
+    },
+    {
+      "epoch": 0.7073224043715847,
+      "grad_norm": 1.0696547824057192,
+      "learning_rate": 0.0001785992965646685,
+      "loss": 13.3673,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7074972677595628,
+      "grad_norm": 1.0951724111669245,
+      "learning_rate": 0.0001785876223233574,
+      "loss": 13.4158,
+      "step": 4046
+    },
+    {
+      "epoch": 0.707672131147541,
+      "grad_norm": 1.0466744566413821,
+      "learning_rate": 0.00017857594528048937,
+      "loss": 13.2555,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7078469945355191,
+      "grad_norm": 1.1598029269989767,
+      "learning_rate": 0.00017856426543648072,
+      "loss": 13.4235,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7080218579234973,
+      "grad_norm": 1.0894544856222905,
+      "learning_rate": 0.00017855258279174778,
+      "loss": 13.3735,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7081967213114754,
+      "grad_norm": 1.1994429568324905,
+      "learning_rate": 0.00017854089734670704,
+      "loss": 13.1777,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7083715846994536,
+      "grad_norm": 1.2445122320597486,
+      "learning_rate": 0.0001785292091017751,
+      "loss": 13.3826,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7085464480874317,
+      "grad_norm": 1.2483060337576728,
+      "learning_rate": 0.00017851751805736857,
+      "loss": 13.3937,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7087213114754098,
+      "grad_norm": 1.3492304578093786,
+      "learning_rate": 0.0001785058242139043,
+      "loss": 13.5479,
+      "step": 4053
+    },
+    {
+      "epoch": 0.708896174863388,
+      "grad_norm": 1.1551001922039759,
+      "learning_rate": 0.00017849412757179906,
+      "loss": 13.1395,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7090710382513661,
+      "grad_norm": 1.0780709710047316,
+      "learning_rate": 0.0001784824281314699,
+      "loss": 13.4827,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7092459016393443,
+      "grad_norm": 1.1388638300585574,
+      "learning_rate": 0.00017847072589333393,
+      "loss": 13.2079,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7094207650273224,
+      "grad_norm": 1.203764714615421,
+      "learning_rate": 0.00017845902085780822,
+      "loss": 13.0177,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7095956284153005,
+      "grad_norm": 1.1112710799110752,
+      "learning_rate": 0.0001784473130253101,
+      "loss": 13.1949,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7097704918032787,
+      "grad_norm": 1.12131268700945,
+      "learning_rate": 0.0001784356023962569,
+      "loss": 13.0085,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7099453551912568,
+      "grad_norm": 1.088169265895156,
+      "learning_rate": 0.00017842388897106615,
+      "loss": 13.2828,
+      "step": 4060
+    },
+    {
+      "epoch": 0.710120218579235,
+      "grad_norm": 1.0163430986935118,
+      "learning_rate": 0.00017841217275015538,
+      "loss": 13.1533,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7102950819672131,
+      "grad_norm": 1.095259268371509,
+      "learning_rate": 0.00017840045373394226,
+      "loss": 13.4163,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7104699453551913,
+      "grad_norm": 1.085764716891633,
+      "learning_rate": 0.00017838873192284454,
+      "loss": 13.3847,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7106448087431694,
+      "grad_norm": 0.9702085728437106,
+      "learning_rate": 0.00017837700731728018,
+      "loss": 13.3048,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7108196721311475,
+      "grad_norm": 1.0183982849561801,
+      "learning_rate": 0.00017836527991766703,
+      "loss": 13.4257,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7109945355191257,
+      "grad_norm": 1.0654956813807566,
+      "learning_rate": 0.00017835354972442323,
+      "loss": 13.197,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7111693989071038,
+      "grad_norm": 1.0696925022473691,
+      "learning_rate": 0.0001783418167379669,
+      "loss": 13.4059,
+      "step": 4067
+    },
+    {
+      "epoch": 0.711344262295082,
+      "grad_norm": 1.0176751482155386,
+      "learning_rate": 0.00017833008095871638,
+      "loss": 13.2784,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7115191256830601,
+      "grad_norm": 1.323150341120676,
+      "learning_rate": 0.00017831834238708996,
+      "loss": 13.2986,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7116939890710382,
+      "grad_norm": 1.0887123478132703,
+      "learning_rate": 0.00017830660102350615,
+      "loss": 13.4318,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7118688524590164,
+      "grad_norm": 1.170971640756222,
+      "learning_rate": 0.0001782948568683835,
+      "loss": 13.6124,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7120437158469946,
+      "grad_norm": 1.0396687125819633,
+      "learning_rate": 0.0001782831099221407,
+      "loss": 13.5718,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7122185792349727,
+      "grad_norm": 1.1102584757138223,
+      "learning_rate": 0.00017827136018519648,
+      "loss": 13.4874,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7123934426229508,
+      "grad_norm": 1.0901978958025627,
+      "learning_rate": 0.0001782596076579697,
+      "loss": 13.3374,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7125683060109289,
+      "grad_norm": 1.035466640703559,
+      "learning_rate": 0.00017824785234087938,
+      "loss": 13.2317,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7127431693989071,
+      "grad_norm": 1.0813651856256308,
+      "learning_rate": 0.00017823609423434452,
+      "loss": 13.5872,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7129180327868853,
+      "grad_norm": 1.113184421728092,
+      "learning_rate": 0.00017822433333878436,
+      "loss": 13.4142,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7130928961748634,
+      "grad_norm": 1.0472059868718335,
+      "learning_rate": 0.00017821256965461807,
+      "loss": 13.424,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7132677595628415,
+      "grad_norm": 1.2505446922339216,
+      "learning_rate": 0.00017820080318226502,
+      "loss": 13.5483,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7134426229508197,
+      "grad_norm": 1.200178506379058,
+      "learning_rate": 0.00017818903392214475,
+      "loss": 13.5161,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7136174863387978,
+      "grad_norm": 1.3414392739823633,
+      "learning_rate": 0.00017817726187467677,
+      "loss": 13.4669,
+      "step": 4081
+    },
+    {
+      "epoch": 0.713792349726776,
+      "grad_norm": 1.1971005943108568,
+      "learning_rate": 0.00017816548704028073,
+      "loss": 13.4563,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7139672131147541,
+      "grad_norm": 1.098323919430518,
+      "learning_rate": 0.00017815370941937642,
+      "loss": 13.2037,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7141420765027322,
+      "grad_norm": 1.1162875260945304,
+      "learning_rate": 0.00017814192901238366,
+      "loss": 13.5066,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7143169398907104,
+      "grad_norm": 1.0850173310497075,
+      "learning_rate": 0.00017813014581972246,
+      "loss": 13.5891,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7144918032786886,
+      "grad_norm": 1.0458743399050878,
+      "learning_rate": 0.00017811835984181282,
+      "loss": 13.3522,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7146666666666667,
+      "grad_norm": 0.9956310318252529,
+      "learning_rate": 0.00017810657107907493,
+      "loss": 13.2493,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7148415300546448,
+      "grad_norm": 1.1493380191967206,
+      "learning_rate": 0.00017809477953192906,
+      "loss": 13.3696,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7150163934426229,
+      "grad_norm": 1.1114725769844236,
+      "learning_rate": 0.00017808298520079552,
+      "loss": 13.6426,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7151912568306011,
+      "grad_norm": 0.9788799394035881,
+      "learning_rate": 0.0001780711880860948,
+      "loss": 13.4424,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7153661202185793,
+      "grad_norm": 1.1753715032639214,
+      "learning_rate": 0.00017805938818824744,
+      "loss": 13.4581,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7155409836065574,
+      "grad_norm": 1.0534674566136242,
+      "learning_rate": 0.00017804758550767405,
+      "loss": 13.3884,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7157158469945355,
+      "grad_norm": 1.0351309011251588,
+      "learning_rate": 0.00017803578004479552,
+      "loss": 13.4542,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7158907103825136,
+      "grad_norm": 1.1224353425471967,
+      "learning_rate": 0.0001780239718000325,
+      "loss": 13.3354,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7160655737704918,
+      "grad_norm": 1.0346663935720668,
+      "learning_rate": 0.00017801216077380612,
+      "loss": 13.2027,
+      "step": 4095
+    },
+    {
+      "epoch": 0.71624043715847,
+      "grad_norm": 1.075820073298326,
+      "learning_rate": 0.00017800034696653733,
+      "loss": 13.3864,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7164153005464481,
+      "grad_norm": 1.057638029458754,
+      "learning_rate": 0.00017798853037864732,
+      "loss": 13.4064,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7165901639344262,
+      "grad_norm": 1.1409746606266309,
+      "learning_rate": 0.0001779767110105573,
+      "loss": 13.2548,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7167650273224043,
+      "grad_norm": 1.04751905462506,
+      "learning_rate": 0.00017796488886268866,
+      "loss": 12.9681,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7169398907103826,
+      "grad_norm": 1.0753667728301555,
+      "learning_rate": 0.0001779530639354628,
+      "loss": 13.4802,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7171147540983607,
+      "grad_norm": 1.1346319887175274,
+      "learning_rate": 0.0001779412362293013,
+      "loss": 13.5714,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7172896174863388,
+      "grad_norm": 1.1019447162522424,
+      "learning_rate": 0.0001779294057446258,
+      "loss": 13.6048,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7174644808743169,
+      "grad_norm": 0.9932264158430129,
+      "learning_rate": 0.00017791757248185802,
+      "loss": 13.3241,
+      "step": 4103
+    },
+    {
+      "epoch": 0.717639344262295,
+      "grad_norm": 0.987358392859603,
+      "learning_rate": 0.00017790573644141982,
+      "loss": 13.1613,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7178142076502733,
+      "grad_norm": 1.108986954346081,
+      "learning_rate": 0.00017789389762373317,
+      "loss": 13.3765,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7179890710382514,
+      "grad_norm": 1.0977597664946652,
+      "learning_rate": 0.00017788205602922006,
+      "loss": 13.578,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7181639344262295,
+      "grad_norm": 1.2153525757251977,
+      "learning_rate": 0.00017787021165830265,
+      "loss": 13.2799,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7183387978142076,
+      "grad_norm": 0.9966594452977267,
+      "learning_rate": 0.00017785836451140313,
+      "loss": 13.382,
+      "step": 4108
+    },
+    {
+      "epoch": 0.7185136612021857,
+      "grad_norm": 1.2321568249045907,
+      "learning_rate": 0.00017784651458894394,
+      "loss": 13.41,
+      "step": 4109
+    },
+    {
+      "epoch": 0.718688524590164,
+      "grad_norm": 1.119021339580658,
+      "learning_rate": 0.00017783466189134745,
+      "loss": 13.4343,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7188633879781421,
+      "grad_norm": 1.1062023380300514,
+      "learning_rate": 0.00017782280641903618,
+      "loss": 13.2802,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7190382513661202,
+      "grad_norm": 1.2988177624220283,
+      "learning_rate": 0.00017781094817243283,
+      "loss": 13.5988,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7192131147540983,
+      "grad_norm": 1.038175089403046,
+      "learning_rate": 0.00017779908715196004,
+      "loss": 13.3732,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7193879781420766,
+      "grad_norm": 1.050532583623423,
+      "learning_rate": 0.00017778722335804072,
+      "loss": 13.4464,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7195628415300547,
+      "grad_norm": 1.1258799204468777,
+      "learning_rate": 0.00017777535679109775,
+      "loss": 13.3979,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7197377049180328,
+      "grad_norm": 1.1868845145634648,
+      "learning_rate": 0.00017776348745155422,
+      "loss": 13.4169,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7199125683060109,
+      "grad_norm": 1.0843895421803618,
+      "learning_rate": 0.00017775161533983318,
+      "loss": 13.3772,
+      "step": 4117
+    },
+    {
+      "epoch": 0.720087431693989,
+      "grad_norm": 1.2358964792868647,
+      "learning_rate": 0.00017773974045635793,
+      "loss": 13.2544,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7202622950819673,
+      "grad_norm": 1.1682536834859765,
+      "learning_rate": 0.00017772786280155175,
+      "loss": 13.1931,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7204371584699454,
+      "grad_norm": 1.1668569318124191,
+      "learning_rate": 0.0001777159823758381,
+      "loss": 13.4081,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7206120218579235,
+      "grad_norm": 1.1187539015914065,
+      "learning_rate": 0.0001777040991796404,
+      "loss": 13.281,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7207868852459016,
+      "grad_norm": 1.169213120981278,
+      "learning_rate": 0.00017769221321338243,
+      "loss": 13.0755,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7209617486338797,
+      "grad_norm": 1.2461948938745235,
+      "learning_rate": 0.00017768032447748779,
+      "loss": 13.4864,
+      "step": 4123
+    },
+    {
+      "epoch": 0.721136612021858,
+      "grad_norm": 1.392910495963338,
+      "learning_rate": 0.00017766843297238035,
+      "loss": 13.4846,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7213114754098361,
+      "grad_norm": 1.2193978183044833,
+      "learning_rate": 0.00017765653869848404,
+      "loss": 13.3491,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7214863387978142,
+      "grad_norm": 1.0449283101134503,
+      "learning_rate": 0.00017764464165622283,
+      "loss": 13.3313,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7216612021857923,
+      "grad_norm": 1.034096540956219,
+      "learning_rate": 0.0001776327418460209,
+      "loss": 13.2353,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7218360655737704,
+      "grad_norm": 1.0965242661601595,
+      "learning_rate": 0.0001776208392683024,
+      "loss": 13.5301,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7220109289617487,
+      "grad_norm": 1.1373146931758258,
+      "learning_rate": 0.00017760893392349168,
+      "loss": 13.2487,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7221857923497268,
+      "grad_norm": 1.0782259312863192,
+      "learning_rate": 0.00017759702581201313,
+      "loss": 13.038,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7223606557377049,
+      "grad_norm": 1.0060120566626511,
+      "learning_rate": 0.00017758511493429127,
+      "loss": 13.4416,
+      "step": 4131
+    },
+    {
+      "epoch": 0.722535519125683,
+      "grad_norm": 1.0596901691061729,
+      "learning_rate": 0.0001775732012907507,
+      "loss": 13.3342,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7227103825136612,
+      "grad_norm": 1.0603732194589899,
+      "learning_rate": 0.00017756128488181618,
+      "loss": 13.1557,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7228852459016394,
+      "grad_norm": 1.0888947550127304,
+      "learning_rate": 0.00017754936570791246,
+      "loss": 13.4032,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7230601092896175,
+      "grad_norm": 1.1280419425094472,
+      "learning_rate": 0.00017753744376946444,
+      "loss": 13.418,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7232349726775956,
+      "grad_norm": 1.0430393945414929,
+      "learning_rate": 0.00017752551906689718,
+      "loss": 13.2644,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7234098360655737,
+      "grad_norm": 1.074607202483291,
+      "learning_rate": 0.00017751359160063571,
+      "loss": 13.408,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7235846994535519,
+      "grad_norm": 1.1219062073498127,
+      "learning_rate": 0.00017750166137110525,
+      "loss": 13.449,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7237595628415301,
+      "grad_norm": 1.0312892503441693,
+      "learning_rate": 0.00017748972837873116,
+      "loss": 13.5564,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7239344262295082,
+      "grad_norm": 1.188814381556139,
+      "learning_rate": 0.00017747779262393878,
+      "loss": 13.3769,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7241092896174863,
+      "grad_norm": 0.9910925091996505,
+      "learning_rate": 0.00017746585410715357,
+      "loss": 13.1835,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7242841530054644,
+      "grad_norm": 1.1302753874633356,
+      "learning_rate": 0.0001774539128288012,
+      "loss": 13.5371,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7244590163934426,
+      "grad_norm": 1.1134805507975447,
+      "learning_rate": 0.00017744196878930737,
+      "loss": 13.2772,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7246338797814208,
+      "grad_norm": 1.2068073585587769,
+      "learning_rate": 0.00017743002198909779,
+      "loss": 13.3297,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7248087431693989,
+      "grad_norm": 1.259507432103488,
+      "learning_rate": 0.0001774180724285984,
+      "loss": 13.4062,
+      "step": 4145
+    },
+    {
+      "epoch": 0.724983606557377,
+      "grad_norm": 1.1815001630831523,
+      "learning_rate": 0.0001774061201082352,
+      "loss": 13.2462,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7251584699453552,
+      "grad_norm": 1.3540992135455494,
+      "learning_rate": 0.0001773941650284342,
+      "loss": 13.4373,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7253333333333334,
+      "grad_norm": 1.0412180798968171,
+      "learning_rate": 0.0001773822071896217,
+      "loss": 13.5683,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7255081967213115,
+      "grad_norm": 1.2128188969842522,
+      "learning_rate": 0.00017737024659222393,
+      "loss": 13.1152,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7256830601092896,
+      "grad_norm": 1.0442688758232064,
+      "learning_rate": 0.00017735828323666724,
+      "loss": 13.607,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7258579234972677,
+      "grad_norm": 1.1740088420403816,
+      "learning_rate": 0.00017734631712337814,
+      "loss": 13.2143,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7260327868852459,
+      "grad_norm": 1.1827627544669543,
+      "learning_rate": 0.0001773343482527832,
+      "loss": 13.216,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7262076502732241,
+      "grad_norm": 1.2188533882592656,
+      "learning_rate": 0.0001773223766253091,
+      "loss": 13.1723,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7263825136612022,
+      "grad_norm": 1.284505894525246,
+      "learning_rate": 0.00017731040224138264,
+      "loss": 13.3779,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7265573770491803,
+      "grad_norm": 1.0326843806309225,
+      "learning_rate": 0.00017729842510143064,
+      "loss": 13.3099,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7267322404371584,
+      "grad_norm": 1.2020621985260413,
+      "learning_rate": 0.0001772864452058801,
+      "loss": 13.2508,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7269071038251366,
+      "grad_norm": 1.1358001376436946,
+      "learning_rate": 0.00017727446255515813,
+      "loss": 13.5975,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7270819672131148,
+      "grad_norm": 1.2334948415048916,
+      "learning_rate": 0.0001772624771496918,
+      "loss": 13.4323,
+      "step": 4158
+    },
+    {
+      "epoch": 0.7272568306010929,
+      "grad_norm": 0.937571363254278,
+      "learning_rate": 0.00017725048898990843,
+      "loss": 13.1099,
+      "step": 4159
+    },
+    {
+      "epoch": 0.727431693989071,
+      "grad_norm": 1.5092913146340854,
+      "learning_rate": 0.00017723849807623542,
+      "loss": 13.2684,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7276065573770492,
+      "grad_norm": 1.2401965563117634,
+      "learning_rate": 0.0001772265044091002,
+      "loss": 13.4094,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7277814207650273,
+      "grad_norm": 1.2126134125460863,
+      "learning_rate": 0.0001772145079889303,
+      "loss": 13.2431,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7279562841530055,
+      "grad_norm": 1.2723956985627605,
+      "learning_rate": 0.0001772025088161534,
+      "loss": 13.3272,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7281311475409836,
+      "grad_norm": 1.3549000752365317,
+      "learning_rate": 0.00017719050689119729,
+      "loss": 13.5765,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7283060109289617,
+      "grad_norm": 1.4613780502784213,
+      "learning_rate": 0.0001771785022144898,
+      "loss": 13.3978,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7284808743169399,
+      "grad_norm": 1.0996794648315666,
+      "learning_rate": 0.00017716649478645885,
+      "loss": 13.5287,
+      "step": 4166
+    },
+    {
+      "epoch": 0.728655737704918,
+      "grad_norm": 1.584500614024604,
+      "learning_rate": 0.00017715448460753255,
+      "loss": 13.298,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7288306010928962,
+      "grad_norm": 0.9607949340306912,
+      "learning_rate": 0.000177142471678139,
+      "loss": 13.3669,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7290054644808743,
+      "grad_norm": 1.5463539486620645,
+      "learning_rate": 0.00017713045599870647,
+      "loss": 13.5286,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7291803278688525,
+      "grad_norm": 1.0850930710688569,
+      "learning_rate": 0.0001771184375696633,
+      "loss": 13.0614,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7293551912568306,
+      "grad_norm": 1.2696096568566515,
+      "learning_rate": 0.0001771064163914379,
+      "loss": 13.3843,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7295300546448087,
+      "grad_norm": 1.0520512712175936,
+      "learning_rate": 0.00017709439246445892,
+      "loss": 13.4216,
+      "step": 4172
+    },
+    {
+      "epoch": 0.7297049180327869,
+      "grad_norm": 1.658154604632169,
+      "learning_rate": 0.0001770823657891549,
+      "loss": 13.3523,
+      "step": 4173
+    },
+    {
+      "epoch": 0.729879781420765,
+      "grad_norm": 1.183855563671467,
+      "learning_rate": 0.00017707033636595455,
+      "loss": 13.5712,
+      "step": 4174
+    },
+    {
+      "epoch": 0.7300546448087432,
+      "grad_norm": 1.5385766061694053,
+      "learning_rate": 0.0001770583041952868,
+      "loss": 13.2083,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7302295081967213,
+      "grad_norm": 1.071231851244594,
+      "learning_rate": 0.00017704626927758052,
+      "loss": 13.2584,
+      "step": 4176
+    },
+    {
+      "epoch": 0.7304043715846995,
+      "grad_norm": 1.307302465079046,
+      "learning_rate": 0.00017703423161326477,
+      "loss": 13.3758,
+      "step": 4177
+    },
+    {
+      "epoch": 0.7305792349726776,
+      "grad_norm": 1.118107972995625,
+      "learning_rate": 0.00017702219120276867,
+      "loss": 13.3532,
+      "step": 4178
+    },
+    {
+      "epoch": 0.7307540983606557,
+      "grad_norm": 1.1131097630771614,
+      "learning_rate": 0.00017701014804652143,
+      "loss": 13.3719,
+      "step": 4179
+    },
+    {
+      "epoch": 0.7309289617486339,
+      "grad_norm": 1.091276548373919,
+      "learning_rate": 0.0001769981021449524,
+      "loss": 13.2253,
+      "step": 4180
+    },
+    {
+      "epoch": 0.731103825136612,
+      "grad_norm": 1.3611859999519424,
+      "learning_rate": 0.000176986053498491,
+      "loss": 13.5506,
+      "step": 4181
+    },
+    {
+      "epoch": 0.7312786885245902,
+      "grad_norm": 1.2433442906902836,
+      "learning_rate": 0.00017697400210756672,
+      "loss": 13.2385,
+      "step": 4182
+    },
+    {
+      "epoch": 0.7314535519125683,
+      "grad_norm": 1.0872098970831954,
+      "learning_rate": 0.00017696194797260924,
+      "loss": 13.3169,
+      "step": 4183
+    },
+    {
+      "epoch": 0.7316284153005465,
+      "grad_norm": 1.1004422591989327,
+      "learning_rate": 0.0001769498910940482,
+      "loss": 13.348,
+      "step": 4184
+    },
+    {
+      "epoch": 0.7318032786885246,
+      "grad_norm": 1.1357323091230165,
+      "learning_rate": 0.00017693783147231345,
+      "loss": 13.1525,
+      "step": 4185
+    },
+    {
+      "epoch": 0.7319781420765027,
+      "grad_norm": 1.2455546725276008,
+      "learning_rate": 0.0001769257691078349,
+      "loss": 13.3182,
+      "step": 4186
+    },
+    {
+      "epoch": 0.7321530054644809,
+      "grad_norm": 1.1490018047075639,
+      "learning_rate": 0.00017691370400104255,
+      "loss": 13.4364,
+      "step": 4187
+    },
+    {
+      "epoch": 0.732327868852459,
+      "grad_norm": 1.163949362375902,
+      "learning_rate": 0.00017690163615236655,
+      "loss": 13.3481,
+      "step": 4188
+    },
+    {
+      "epoch": 0.7325027322404372,
+      "grad_norm": 1.0764479683049926,
+      "learning_rate": 0.00017688956556223702,
+      "loss": 13.1744,
+      "step": 4189
+    },
+    {
+      "epoch": 0.7326775956284153,
+      "grad_norm": 1.050625873941137,
+      "learning_rate": 0.00017687749223108435,
+      "loss": 13.3782,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7328524590163934,
+      "grad_norm": 1.0342863211438171,
+      "learning_rate": 0.0001768654161593389,
+      "loss": 13.2605,
+      "step": 4191
+    },
+    {
+      "epoch": 0.7330273224043716,
+      "grad_norm": 1.020470863353238,
+      "learning_rate": 0.00017685333734743117,
+      "loss": 13.6007,
+      "step": 4192
+    },
+    {
+      "epoch": 0.7332021857923497,
+      "grad_norm": 0.9876235005028986,
+      "learning_rate": 0.00017684125579579178,
+      "loss": 13.1965,
+      "step": 4193
+    },
+    {
+      "epoch": 0.7333770491803279,
+      "grad_norm": 1.0146774797996458,
+      "learning_rate": 0.00017682917150485137,
+      "loss": 13.1008,
+      "step": 4194
+    },
+    {
+      "epoch": 0.733551912568306,
+      "grad_norm": 1.0967820978898668,
+      "learning_rate": 0.00017681708447504074,
+      "loss": 13.3811,
+      "step": 4195
+    },
+    {
+      "epoch": 0.7337267759562841,
+      "grad_norm": 1.0767174030242201,
+      "learning_rate": 0.00017680499470679086,
+      "loss": 13.2775,
+      "step": 4196
+    },
+    {
+      "epoch": 0.7339016393442623,
+      "grad_norm": 1.002692105604657,
+      "learning_rate": 0.00017679290220053262,
+      "loss": 13.1599,
+      "step": 4197
+    },
+    {
+      "epoch": 0.7340765027322405,
+      "grad_norm": 1.1821058983675454,
+      "learning_rate": 0.00017678080695669715,
+      "loss": 13.2611,
+      "step": 4198
+    },
+    {
+      "epoch": 0.7342513661202186,
+      "grad_norm": 1.1150225771243392,
+      "learning_rate": 0.0001767687089757156,
+      "loss": 13.3466,
+      "step": 4199
+    },
+    {
+      "epoch": 0.7344262295081967,
+      "grad_norm": 1.1260362211678672,
+      "learning_rate": 0.00017675660825801928,
+      "loss": 13.3399,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7346010928961748,
+      "grad_norm": 1.0444411132194291,
+      "learning_rate": 0.00017674450480403958,
+      "loss": 13.4377,
+      "step": 4201
+    },
+    {
+      "epoch": 0.734775956284153,
+      "grad_norm": 1.1084346563404157,
+      "learning_rate": 0.00017673239861420794,
+      "loss": 13.4237,
+      "step": 4202
+    },
+    {
+      "epoch": 0.7349508196721312,
+      "grad_norm": 1.267135299559725,
+      "learning_rate": 0.0001767202896889559,
+      "loss": 13.5139,
+      "step": 4203
+    },
+    {
+      "epoch": 0.7351256830601093,
+      "grad_norm": 1.052344490113259,
+      "learning_rate": 0.0001767081780287152,
+      "loss": 13.3432,
+      "step": 4204
+    },
+    {
+      "epoch": 0.7353005464480874,
+      "grad_norm": 1.0455176963611819,
+      "learning_rate": 0.00017669606363391758,
+      "loss": 13.4504,
+      "step": 4205
+    },
+    {
+      "epoch": 0.7354754098360655,
+      "grad_norm": 1.1168451527164451,
+      "learning_rate": 0.00017668394650499492,
+      "loss": 13.2974,
+      "step": 4206
+    },
+    {
+      "epoch": 0.7356502732240437,
+      "grad_norm": 1.0867899938800374,
+      "learning_rate": 0.00017667182664237913,
+      "loss": 13.4599,
+      "step": 4207
+    },
+    {
+      "epoch": 0.7358251366120219,
+      "grad_norm": 1.0658033241493483,
+      "learning_rate": 0.00017665970404650232,
+      "loss": 13.5748,
+      "step": 4208
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 1.2084223259923172,
+      "learning_rate": 0.00017664757871779663,
+      "loss": 13.3618,
+      "step": 4209
+    },
+    {
+      "epoch": 0.7361748633879781,
+      "grad_norm": 1.2380363767514424,
+      "learning_rate": 0.0001766354506566943,
+      "loss": 13.5645,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7363497267759563,
+      "grad_norm": 1.0830299900231013,
+      "learning_rate": 0.0001766233198636277,
+      "loss": 13.4257,
+      "step": 4211
+    },
+    {
+      "epoch": 0.7365245901639345,
+      "grad_norm": 1.1814345551062968,
+      "learning_rate": 0.00017661118633902927,
+      "loss": 13.2279,
+      "step": 4212
+    },
+    {
+      "epoch": 0.7366994535519126,
+      "grad_norm": 1.1367509403043994,
+      "learning_rate": 0.00017659905008333155,
+      "loss": 13.4402,
+      "step": 4213
+    },
+    {
+      "epoch": 0.7368743169398907,
+      "grad_norm": 1.1798679926277105,
+      "learning_rate": 0.00017658691109696722,
+      "loss": 13.3797,
+      "step": 4214
+    },
+    {
+      "epoch": 0.7370491803278688,
+      "grad_norm": 1.0878547997535737,
+      "learning_rate": 0.00017657476938036894,
+      "loss": 13.5493,
+      "step": 4215
+    },
+    {
+      "epoch": 0.737224043715847,
+      "grad_norm": 1.3481344701729998,
+      "learning_rate": 0.00017656262493396963,
+      "loss": 13.4237,
+      "step": 4216
+    },
+    {
+      "epoch": 0.7373989071038252,
+      "grad_norm": 1.1637003529186587,
+      "learning_rate": 0.0001765504777582022,
+      "loss": 13.4558,
+      "step": 4217
+    },
+    {
+      "epoch": 0.7375737704918033,
+      "grad_norm": 1.1405931816339143,
+      "learning_rate": 0.00017653832785349968,
+      "loss": 13.319,
+      "step": 4218
+    },
+    {
+      "epoch": 0.7377486338797814,
+      "grad_norm": 1.033013687909615,
+      "learning_rate": 0.00017652617522029518,
+      "loss": 13.2598,
+      "step": 4219
+    },
+    {
+      "epoch": 0.7379234972677595,
+      "grad_norm": 1.0830991169774995,
+      "learning_rate": 0.00017651401985902196,
+      "loss": 13.4972,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7380983606557378,
+      "grad_norm": 1.1262096841559428,
+      "learning_rate": 0.0001765018617701133,
+      "loss": 13.2226,
+      "step": 4221
+    },
+    {
+      "epoch": 0.7382732240437159,
+      "grad_norm": 1.0638761377743347,
+      "learning_rate": 0.0001764897009540027,
+      "loss": 13.2913,
+      "step": 4222
+    },
+    {
+      "epoch": 0.738448087431694,
+      "grad_norm": 1.1981395120275664,
+      "learning_rate": 0.00017647753741112358,
+      "loss": 13.2517,
+      "step": 4223
+    },
+    {
+      "epoch": 0.7386229508196721,
+      "grad_norm": 1.026315914815636,
+      "learning_rate": 0.00017646537114190963,
+      "loss": 13.1571,
+      "step": 4224
+    },
+    {
+      "epoch": 0.7387978142076502,
+      "grad_norm": 1.3419936288803103,
+      "learning_rate": 0.00017645320214679456,
+      "loss": 13.4378,
+      "step": 4225
+    },
+    {
+      "epoch": 0.7389726775956285,
+      "grad_norm": 0.9924207489812511,
+      "learning_rate": 0.00017644103042621211,
+      "loss": 13.3622,
+      "step": 4226
+    },
+    {
+      "epoch": 0.7391475409836066,
+      "grad_norm": 1.2772300656665627,
+      "learning_rate": 0.00017642885598059626,
+      "loss": 13.3832,
+      "step": 4227
+    },
+    {
+      "epoch": 0.7393224043715847,
+      "grad_norm": 1.1285554741961956,
+      "learning_rate": 0.00017641667881038102,
+      "loss": 13.2506,
+      "step": 4228
+    },
+    {
+      "epoch": 0.7394972677595628,
+      "grad_norm": 1.4320605870227885,
+      "learning_rate": 0.00017640449891600043,
+      "loss": 13.2332,
+      "step": 4229
+    },
+    {
+      "epoch": 0.7396721311475409,
+      "grad_norm": 1.3655396766412737,
+      "learning_rate": 0.00017639231629788875,
+      "loss": 13.5676,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7398469945355192,
+      "grad_norm": 1.2549390489538033,
+      "learning_rate": 0.00017638013095648024,
+      "loss": 13.3998,
+      "step": 4231
+    },
+    {
+      "epoch": 0.7400218579234973,
+      "grad_norm": 1.4375157366393596,
+      "learning_rate": 0.00017636794289220927,
+      "loss": 13.4697,
+      "step": 4232
+    },
+    {
+      "epoch": 0.7401967213114754,
+      "grad_norm": 1.035245894313682,
+      "learning_rate": 0.00017635575210551039,
+      "loss": 13.2351,
+      "step": 4233
+    },
+    {
+      "epoch": 0.7403715846994535,
+      "grad_norm": 1.5942971805082653,
+      "learning_rate": 0.00017634355859681814,
+      "loss": 13.4474,
+      "step": 4234
+    },
+    {
+      "epoch": 0.7405464480874316,
+      "grad_norm": 0.9792200309661176,
+      "learning_rate": 0.00017633136236656723,
+      "loss": 13.3313,
+      "step": 4235
+    },
+    {
+      "epoch": 0.7407213114754099,
+      "grad_norm": 1.4749647662566598,
+      "learning_rate": 0.00017631916341519246,
+      "loss": 13.2225,
+      "step": 4236
+    },
+    {
+      "epoch": 0.740896174863388,
+      "grad_norm": 1.0331403898115437,
+      "learning_rate": 0.00017630696174312867,
+      "loss": 13.1599,
+      "step": 4237
+    },
+    {
+      "epoch": 0.7410710382513661,
+      "grad_norm": 1.5154924018794482,
+      "learning_rate": 0.00017629475735081085,
+      "loss": 13.3112,
+      "step": 4238
+    },
+    {
+      "epoch": 0.7412459016393442,
+      "grad_norm": 1.0746830464877537,
+      "learning_rate": 0.00017628255023867406,
+      "loss": 13.3133,
+      "step": 4239
+    },
+    {
+      "epoch": 0.7414207650273225,
+      "grad_norm": 1.285233493529,
+      "learning_rate": 0.0001762703404071535,
+      "loss": 13.2773,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7415956284153006,
+      "grad_norm": 1.0525704170432553,
+      "learning_rate": 0.00017625812785668438,
+      "loss": 13.4388,
+      "step": 4241
+    },
+    {
+      "epoch": 0.7417704918032787,
+      "grad_norm": 1.0949658674240137,
+      "learning_rate": 0.00017624591258770213,
+      "loss": 13.0369,
+      "step": 4242
+    },
+    {
+      "epoch": 0.7419453551912568,
+      "grad_norm": 1.164435882345618,
+      "learning_rate": 0.00017623369460064217,
+      "loss": 13.0686,
+      "step": 4243
+    },
+    {
+      "epoch": 0.7421202185792349,
+      "grad_norm": 1.1501165471860086,
+      "learning_rate": 0.0001762214738959401,
+      "loss": 13.298,
+      "step": 4244
+    },
+    {
+      "epoch": 0.7422950819672132,
+      "grad_norm": 1.115954292200518,
+      "learning_rate": 0.0001762092504740315,
+      "loss": 13.3105,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7424699453551913,
+      "grad_norm": 1.1665951074749692,
+      "learning_rate": 0.00017619702433535215,
+      "loss": 13.1405,
+      "step": 4246
+    },
+    {
+      "epoch": 0.7426448087431694,
+      "grad_norm": 1.1678122103347228,
+      "learning_rate": 0.00017618479548033795,
+      "loss": 13.4631,
+      "step": 4247
+    },
+    {
+      "epoch": 0.7428196721311475,
+      "grad_norm": 1.1721739799637785,
+      "learning_rate": 0.0001761725639094248,
+      "loss": 13.4025,
+      "step": 4248
+    },
+    {
+      "epoch": 0.7429945355191256,
+      "grad_norm": 1.026707811339875,
+      "learning_rate": 0.00017616032962304874,
+      "loss": 13.4096,
+      "step": 4249
+    },
+    {
+      "epoch": 0.7431693989071039,
+      "grad_norm": 1.4697151716383123,
+      "learning_rate": 0.0001761480926216459,
+      "loss": 13.44,
+      "step": 4250
+    },
+    {
+      "epoch": 0.743344262295082,
+      "grad_norm": 1.2112606256449165,
+      "learning_rate": 0.00017613585290565252,
+      "loss": 13.4641,
+      "step": 4251
+    },
+    {
+      "epoch": 0.7435191256830601,
+      "grad_norm": 1.1941848394226824,
+      "learning_rate": 0.00017612361047550495,
+      "loss": 13.1605,
+      "step": 4252
+    },
+    {
+      "epoch": 0.7436939890710382,
+      "grad_norm": 1.0550518682940642,
+      "learning_rate": 0.00017611136533163962,
+      "loss": 12.9354,
+      "step": 4253
+    },
+    {
+      "epoch": 0.7438688524590163,
+      "grad_norm": 1.0382317372903374,
+      "learning_rate": 0.00017609911747449303,
+      "loss": 13.21,
+      "step": 4254
+    },
+    {
+      "epoch": 0.7440437158469946,
+      "grad_norm": 1.1623510941874624,
+      "learning_rate": 0.0001760868669045018,
+      "loss": 13.4831,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7442185792349727,
+      "grad_norm": 1.0359483439078498,
+      "learning_rate": 0.0001760746136221027,
+      "loss": 13.3809,
+      "step": 4256
+    },
+    {
+      "epoch": 0.7443934426229508,
+      "grad_norm": 0.9597437327584789,
+      "learning_rate": 0.00017606235762773246,
+      "loss": 13.3949,
+      "step": 4257
+    },
+    {
+      "epoch": 0.7445683060109289,
+      "grad_norm": 1.0610104052597784,
+      "learning_rate": 0.00017605009892182806,
+      "loss": 13.2879,
+      "step": 4258
+    },
+    {
+      "epoch": 0.7447431693989071,
+      "grad_norm": 1.0106455059465298,
+      "learning_rate": 0.0001760378375048265,
+      "loss": 13.2169,
+      "step": 4259
+    },
+    {
+      "epoch": 0.7449180327868853,
+      "grad_norm": 1.0316146910965476,
+      "learning_rate": 0.00017602557337716488,
+      "loss": 13.2288,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7450928961748634,
+      "grad_norm": 1.0255024439231326,
+      "learning_rate": 0.00017601330653928038,
+      "loss": 13.2594,
+      "step": 4261
+    },
+    {
+      "epoch": 0.7452677595628415,
+      "grad_norm": 1.165679715590607,
+      "learning_rate": 0.00017600103699161033,
+      "loss": 13.2474,
+      "step": 4262
+    },
+    {
+      "epoch": 0.7454426229508196,
+      "grad_norm": 1.0305755059548496,
+      "learning_rate": 0.00017598876473459208,
+      "loss": 13.1775,
+      "step": 4263
+    },
+    {
+      "epoch": 0.7456174863387978,
+      "grad_norm": 1.0016604632709425,
+      "learning_rate": 0.00017597648976866318,
+      "loss": 13.1246,
+      "step": 4264
+    },
+    {
+      "epoch": 0.745792349726776,
+      "grad_norm": 1.1007566935107729,
+      "learning_rate": 0.00017596421209426118,
+      "loss": 13.3767,
+      "step": 4265
+    },
+    {
+      "epoch": 0.7459672131147541,
+      "grad_norm": 1.0462675470267473,
+      "learning_rate": 0.00017595193171182375,
+      "loss": 13.0523,
+      "step": 4266
+    },
+    {
+      "epoch": 0.7461420765027322,
+      "grad_norm": 1.2433752984312119,
+      "learning_rate": 0.00017593964862178872,
+      "loss": 12.8529,
+      "step": 4267
+    },
+    {
+      "epoch": 0.7463169398907104,
+      "grad_norm": 1.2026969051209537,
+      "learning_rate": 0.00017592736282459395,
+      "loss": 13.1619,
+      "step": 4268
+    },
+    {
+      "epoch": 0.7464918032786885,
+      "grad_norm": 1.3782000143372093,
+      "learning_rate": 0.0001759150743206774,
+      "loss": 13.0333,
+      "step": 4269
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 1.18568990118911,
+      "learning_rate": 0.00017590278311047715,
+      "loss": 13.262,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7468415300546448,
+      "grad_norm": 1.300593302865863,
+      "learning_rate": 0.00017589048919443135,
+      "loss": 13.3238,
+      "step": 4271
+    },
+    {
+      "epoch": 0.7470163934426229,
+      "grad_norm": 1.2761769343291085,
+      "learning_rate": 0.00017587819257297826,
+      "loss": 13.3021,
+      "step": 4272
+    },
+    {
+      "epoch": 0.7471912568306011,
+      "grad_norm": 1.1417698494168746,
+      "learning_rate": 0.0001758658932465563,
+      "loss": 13.4388,
+      "step": 4273
+    },
+    {
+      "epoch": 0.7473661202185793,
+      "grad_norm": 1.1744168161073747,
+      "learning_rate": 0.00017585359121560386,
+      "loss": 13.3415,
+      "step": 4274
+    },
+    {
+      "epoch": 0.7475409836065574,
+      "grad_norm": 1.0375522590526776,
+      "learning_rate": 0.00017584128648055953,
+      "loss": 13.3234,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7477158469945355,
+      "grad_norm": 1.3533402361501041,
+      "learning_rate": 0.00017582897904186196,
+      "loss": 13.3219,
+      "step": 4276
+    },
+    {
+      "epoch": 0.7478907103825136,
+      "grad_norm": 1.160165509459615,
+      "learning_rate": 0.00017581666889994986,
+      "loss": 13.1046,
+      "step": 4277
+    },
+    {
+      "epoch": 0.7480655737704918,
+      "grad_norm": 1.225789815456043,
+      "learning_rate": 0.0001758043560552621,
+      "loss": 13.5402,
+      "step": 4278
+    },
+    {
+      "epoch": 0.74824043715847,
+      "grad_norm": 1.0747065245662202,
+      "learning_rate": 0.0001757920405082376,
+      "loss": 13.0669,
+      "step": 4279
+    },
+    {
+      "epoch": 0.7484153005464481,
+      "grad_norm": 1.0979730153391796,
+      "learning_rate": 0.00017577972225931544,
+      "loss": 13.0997,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7485901639344262,
+      "grad_norm": 1.1741892499359141,
+      "learning_rate": 0.00017576740130893472,
+      "loss": 13.1116,
+      "step": 4281
+    },
+    {
+      "epoch": 0.7487650273224044,
+      "grad_norm": 1.1117193932635838,
+      "learning_rate": 0.00017575507765753466,
+      "loss": 13.398,
+      "step": 4282
+    },
+    {
+      "epoch": 0.7489398907103825,
+      "grad_norm": 1.0748881423373449,
+      "learning_rate": 0.0001757427513055546,
+      "loss": 13.2051,
+      "step": 4283
+    },
+    {
+      "epoch": 0.7491147540983607,
+      "grad_norm": 1.3342069764959004,
+      "learning_rate": 0.00017573042225343393,
+      "loss": 13.232,
+      "step": 4284
+    },
+    {
+      "epoch": 0.7492896174863388,
+      "grad_norm": 1.1464580541305722,
+      "learning_rate": 0.0001757180905016122,
+      "loss": 13.46,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7494644808743169,
+      "grad_norm": 1.3010828795843605,
+      "learning_rate": 0.00017570575605052903,
+      "loss": 13.1363,
+      "step": 4286
+    },
+    {
+      "epoch": 0.7496393442622951,
+      "grad_norm": 1.2058036372023124,
+      "learning_rate": 0.00017569341890062408,
+      "loss": 13.3706,
+      "step": 4287
+    },
+    {
+      "epoch": 0.7498142076502732,
+      "grad_norm": 1.2526501170035145,
+      "learning_rate": 0.0001756810790523372,
+      "loss": 13.4098,
+      "step": 4288
+    },
+    {
+      "epoch": 0.7499890710382514,
+      "grad_norm": 1.3610675589014176,
+      "learning_rate": 0.00017566873650610827,
+      "loss": 13.2183,
+      "step": 4289
+    },
+    {
+      "epoch": 0.7501639344262295,
+      "grad_norm": 1.0449085934801086,
+      "learning_rate": 0.0001756563912623773,
+      "loss": 13.2398,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7503387978142076,
+      "grad_norm": 1.1925595661210433,
+      "learning_rate": 0.00017564404332158436,
+      "loss": 12.8962,
+      "step": 4291
+    },
+    {
+      "epoch": 0.7505136612021858,
+      "grad_norm": 1.0455169845091516,
+      "learning_rate": 0.0001756316926841697,
+      "loss": 13.2487,
+      "step": 4292
+    },
+    {
+      "epoch": 0.7506885245901639,
+      "grad_norm": 1.4246099063121855,
+      "learning_rate": 0.00017561933935057348,
+      "loss": 13.2933,
+      "step": 4293
+    },
+    {
+      "epoch": 0.7508633879781421,
+      "grad_norm": 1.0599387008281134,
+      "learning_rate": 0.00017560698332123623,
+      "loss": 13.2553,
+      "step": 4294
+    },
+    {
+      "epoch": 0.7510382513661202,
+      "grad_norm": 1.377935121057621,
+      "learning_rate": 0.00017559462459659835,
+      "loss": 13.2571,
+      "step": 4295
+    },
+    {
+      "epoch": 0.7512131147540984,
+      "grad_norm": 1.2891140262620917,
+      "learning_rate": 0.0001755822631771004,
+      "loss": 13.2454,
+      "step": 4296
+    },
+    {
+      "epoch": 0.7513879781420765,
+      "grad_norm": 1.1194104954494102,
+      "learning_rate": 0.0001755698990631831,
+      "loss": 13.4947,
+      "step": 4297
+    },
+    {
+      "epoch": 0.7515628415300546,
+      "grad_norm": 1.1285130664921834,
+      "learning_rate": 0.0001755575322552872,
+      "loss": 13.2508,
+      "step": 4298
+    },
+    {
+      "epoch": 0.7517377049180328,
+      "grad_norm": 1.0999694127904596,
+      "learning_rate": 0.00017554516275385354,
+      "loss": 13.4566,
+      "step": 4299
+    },
+    {
+      "epoch": 0.7519125683060109,
+      "grad_norm": 1.0087356400933265,
+      "learning_rate": 0.00017553279055932308,
+      "loss": 13.3014,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7520874316939891,
+      "grad_norm": 1.0780229921946787,
+      "learning_rate": 0.0001755204156721369,
+      "loss": 13.4444,
+      "step": 4301
+    },
+    {
+      "epoch": 0.7522622950819672,
+      "grad_norm": 1.013385833221533,
+      "learning_rate": 0.00017550803809273616,
+      "loss": 13.2109,
+      "step": 4302
+    },
+    {
+      "epoch": 0.7524371584699453,
+      "grad_norm": 1.0273872985166947,
+      "learning_rate": 0.00017549565782156202,
+      "loss": 13.2896,
+      "step": 4303
+    },
+    {
+      "epoch": 0.7526120218579235,
+      "grad_norm": 1.1337397420722057,
+      "learning_rate": 0.00017548327485905597,
+      "loss": 13.3296,
+      "step": 4304
+    },
+    {
+      "epoch": 0.7527868852459016,
+      "grad_norm": 1.1037237326178644,
+      "learning_rate": 0.0001754708892056593,
+      "loss": 13.3568,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7529617486338798,
+      "grad_norm": 1.1054894178159915,
+      "learning_rate": 0.00017545850086181367,
+      "loss": 13.346,
+      "step": 4306
+    },
+    {
+      "epoch": 0.7531366120218579,
+      "grad_norm": 0.9769213040026621,
+      "learning_rate": 0.00017544610982796058,
+      "loss": 13.2065,
+      "step": 4307
+    },
+    {
+      "epoch": 0.7533114754098361,
+      "grad_norm": 1.2763445279817573,
+      "learning_rate": 0.00017543371610454183,
+      "loss": 12.9914,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7534863387978142,
+      "grad_norm": 1.0351342705029216,
+      "learning_rate": 0.00017542131969199929,
+      "loss": 13.2507,
+      "step": 4309
+    },
+    {
+      "epoch": 0.7536612021857924,
+      "grad_norm": 1.0575308036345383,
+      "learning_rate": 0.00017540892059077478,
+      "loss": 13.2902,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7538360655737705,
+      "grad_norm": 1.1170966619350964,
+      "learning_rate": 0.0001753965188013104,
+      "loss": 13.4359,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7540109289617486,
+      "grad_norm": 1.1215938386000888,
+      "learning_rate": 0.00017538411432404817,
+      "loss": 12.99,
+      "step": 4312
+    },
+    {
+      "epoch": 0.7541857923497268,
+      "grad_norm": 1.0359116763283156,
+      "learning_rate": 0.00017537170715943034,
+      "loss": 13.1375,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7543606557377049,
+      "grad_norm": 1.145303078056227,
+      "learning_rate": 0.00017535929730789923,
+      "loss": 13.2435,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7545355191256831,
+      "grad_norm": 1.1777857848090187,
+      "learning_rate": 0.0001753468847698972,
+      "loss": 13.4104,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7547103825136612,
+      "grad_norm": 1.026428198773554,
+      "learning_rate": 0.0001753344695458668,
+      "loss": 13.2722,
+      "step": 4316
+    },
+    {
+      "epoch": 0.7548852459016393,
+      "grad_norm": 1.0476597905082996,
+      "learning_rate": 0.00017532205163625056,
+      "loss": 12.9956,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7550601092896175,
+      "grad_norm": 1.1627719686728173,
+      "learning_rate": 0.00017530963104149118,
+      "loss": 13.3773,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7552349726775957,
+      "grad_norm": 1.335858776077617,
+      "learning_rate": 0.00017529720776203144,
+      "loss": 13.4121,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7554098360655738,
+      "grad_norm": 1.0363811223575308,
+      "learning_rate": 0.00017528478179831424,
+      "loss": 13.1536,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7555846994535519,
+      "grad_norm": 1.2032322788062637,
+      "learning_rate": 0.00017527235315078251,
+      "loss": 13.528,
+      "step": 4321
+    },
+    {
+      "epoch": 0.75575956284153,
+      "grad_norm": 1.103881178472415,
+      "learning_rate": 0.00017525992181987934,
+      "loss": 13.4959,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7559344262295082,
+      "grad_norm": 1.0241770040294738,
+      "learning_rate": 0.0001752474878060479,
+      "loss": 13.2335,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7561092896174864,
+      "grad_norm": 1.0721903816114995,
+      "learning_rate": 0.00017523505110973142,
+      "loss": 13.2259,
+      "step": 4324
+    },
+    {
+      "epoch": 0.7562841530054645,
+      "grad_norm": 1.0370595152686675,
+      "learning_rate": 0.0001752226117313733,
+      "loss": 13.3099,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7564590163934426,
+      "grad_norm": 1.1868024510054487,
+      "learning_rate": 0.00017521016967141692,
+      "loss": 13.2088,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7566338797814207,
+      "grad_norm": 1.03746979120928,
+      "learning_rate": 0.00017519772493030587,
+      "loss": 13.2352,
+      "step": 4327
+    },
+    {
+      "epoch": 0.756808743169399,
+      "grad_norm": 1.1260414581147593,
+      "learning_rate": 0.00017518527750848381,
+      "loss": 13.052,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7569836065573771,
+      "grad_norm": 1.0869228255236716,
+      "learning_rate": 0.00017517282740639445,
+      "loss": 13.3809,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7571584699453552,
+      "grad_norm": 1.124857310056332,
+      "learning_rate": 0.0001751603746244816,
+      "loss": 13.0901,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7573333333333333,
+      "grad_norm": 1.0074190135826668,
+      "learning_rate": 0.0001751479191631892,
+      "loss": 13.1722,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7575081967213114,
+      "grad_norm": 1.028022760173983,
+      "learning_rate": 0.00017513546102296133,
+      "loss": 13.2271,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7576830601092897,
+      "grad_norm": 1.0887317082300576,
+      "learning_rate": 0.00017512300020424202,
+      "loss": 12.8424,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7578579234972678,
+      "grad_norm": 1.014470395958744,
+      "learning_rate": 0.00017511053670747555,
+      "loss": 13.0649,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7580327868852459,
+      "grad_norm": 1.1704040587778797,
+      "learning_rate": 0.00017509807053310618,
+      "loss": 13.2837,
+      "step": 4335
+    },
+    {
+      "epoch": 0.758207650273224,
+      "grad_norm": 1.166916933541748,
+      "learning_rate": 0.00017508560168157838,
+      "loss": 13.0692,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7583825136612022,
+      "grad_norm": 1.1235042320195856,
+      "learning_rate": 0.00017507313015333658,
+      "loss": 13.2592,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7585573770491804,
+      "grad_norm": 1.02559198929084,
+      "learning_rate": 0.0001750606559488254,
+      "loss": 13.1948,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7587322404371585,
+      "grad_norm": 1.287099996884149,
+      "learning_rate": 0.00017504817906848952,
+      "loss": 13.2013,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7589071038251366,
+      "grad_norm": 1.160753507683279,
+      "learning_rate": 0.00017503569951277378,
+      "loss": 13.3862,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7590819672131147,
+      "grad_norm": 1.1257797801585052,
+      "learning_rate": 0.000175023217282123,
+      "loss": 13.1441,
+      "step": 4341
+    },
+    {
+      "epoch": 0.759256830601093,
+      "grad_norm": 1.457214688677896,
+      "learning_rate": 0.00017501073237698216,
+      "loss": 13.3292,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7594316939890711,
+      "grad_norm": 0.97249630427483,
+      "learning_rate": 0.0001749982447977964,
+      "loss": 13.1808,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7596065573770492,
+      "grad_norm": 1.3063224700712084,
+      "learning_rate": 0.0001749857545450108,
+      "loss": 13.0502,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7597814207650273,
+      "grad_norm": 1.0473866267224001,
+      "learning_rate": 0.00017497326161907063,
+      "loss": 13.4083,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7599562841530054,
+      "grad_norm": 1.1435901932305381,
+      "learning_rate": 0.00017496076602042135,
+      "loss": 13.0844,
+      "step": 4346
+    },
+    {
+      "epoch": 0.7601311475409837,
+      "grad_norm": 1.1457006783701325,
+      "learning_rate": 0.00017494826774950832,
+      "loss": 13.2088,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7603060109289618,
+      "grad_norm": 1.0965567399289504,
+      "learning_rate": 0.0001749357668067771,
+      "loss": 13.1259,
+      "step": 4348
+    },
+    {
+      "epoch": 0.7604808743169399,
+      "grad_norm": 1.1018254471452031,
+      "learning_rate": 0.00017492326319267334,
+      "loss": 13.1661,
+      "step": 4349
+    },
+    {
+      "epoch": 0.760655737704918,
+      "grad_norm": 1.018902763525456,
+      "learning_rate": 0.00017491075690764284,
+      "loss": 13.2576,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7608306010928961,
+      "grad_norm": 1.2466774970378054,
+      "learning_rate": 0.0001748982479521313,
+      "loss": 13.3628,
+      "step": 4351
+    },
+    {
+      "epoch": 0.7610054644808744,
+      "grad_norm": 0.9834184049320555,
+      "learning_rate": 0.00017488573632658478,
+      "loss": 13.1891,
+      "step": 4352
+    },
+    {
+      "epoch": 0.7611803278688525,
+      "grad_norm": 1.2957082745257502,
+      "learning_rate": 0.00017487322203144925,
+      "loss": 13.2437,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7613551912568306,
+      "grad_norm": 1.1915868079077487,
+      "learning_rate": 0.00017486070506717082,
+      "loss": 13.0824,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7615300546448087,
+      "grad_norm": 1.0543677940063496,
+      "learning_rate": 0.00017484818543419574,
+      "loss": 13.17,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7617049180327868,
+      "grad_norm": 1.060966154861429,
+      "learning_rate": 0.00017483566313297028,
+      "loss": 13.0278,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7618797814207651,
+      "grad_norm": 1.0190014840661843,
+      "learning_rate": 0.00017482313816394088,
+      "loss": 13.2964,
+      "step": 4357
+    },
+    {
+      "epoch": 0.7620546448087432,
+      "grad_norm": 1.0414153018918255,
+      "learning_rate": 0.00017481061052755404,
+      "loss": 12.9208,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7622295081967213,
+      "grad_norm": 0.9933567156876972,
+      "learning_rate": 0.0001747980802242563,
+      "loss": 13.1372,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7624043715846994,
+      "grad_norm": 1.0893907176903603,
+      "learning_rate": 0.0001747855472544944,
+      "loss": 13.3071,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7625792349726775,
+      "grad_norm": 1.0781726651184464,
+      "learning_rate": 0.00017477301161871513,
+      "loss": 13.1952,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7627540983606558,
+      "grad_norm": 1.0014584080191862,
+      "learning_rate": 0.00017476047331736538,
+      "loss": 13.2536,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7629289617486339,
+      "grad_norm": 1.0338078233318073,
+      "learning_rate": 0.00017474793235089207,
+      "loss": 13.2374,
+      "step": 4363
+    },
+    {
+      "epoch": 0.763103825136612,
+      "grad_norm": 1.0363810345576476,
+      "learning_rate": 0.00017473538871974232,
+      "loss": 13.2854,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7632786885245901,
+      "grad_norm": 1.0709477282679298,
+      "learning_rate": 0.0001747228424243633,
+      "loss": 13.2606,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7634535519125683,
+      "grad_norm": 0.9904279174221331,
+      "learning_rate": 0.0001747102934652022,
+      "loss": 13.2153,
+      "step": 4366
+    },
+    {
+      "epoch": 0.7636284153005465,
+      "grad_norm": 1.101446953227734,
+      "learning_rate": 0.00017469774184270648,
+      "loss": 13.1817,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7638032786885246,
+      "grad_norm": 1.0536217434767765,
+      "learning_rate": 0.00017468518755732354,
+      "loss": 13.1949,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7639781420765027,
+      "grad_norm": 1.01246187733694,
+      "learning_rate": 0.00017467263060950088,
+      "loss": 13.2704,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7641530054644808,
+      "grad_norm": 1.0860812242775562,
+      "learning_rate": 0.00017466007099968623,
+      "loss": 13.2178,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7643278688524591,
+      "grad_norm": 1.003462353142333,
+      "learning_rate": 0.00017464750872832726,
+      "loss": 13.1894,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7645027322404372,
+      "grad_norm": 1.1953015445825577,
+      "learning_rate": 0.00017463494379587179,
+      "loss": 13.2947,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7646775956284153,
+      "grad_norm": 1.1190204323332613,
+      "learning_rate": 0.0001746223762027678,
+      "loss": 13.2403,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7648524590163934,
+      "grad_norm": 1.0115433536508769,
+      "learning_rate": 0.0001746098059494633,
+      "loss": 13.2914,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7650273224043715,
+      "grad_norm": 1.1660167257044576,
+      "learning_rate": 0.00017459723303640638,
+      "loss": 12.9424,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7652021857923498,
+      "grad_norm": 1.059695845886241,
+      "learning_rate": 0.00017458465746404524,
+      "loss": 13.2894,
+      "step": 4376
+    },
+    {
+      "epoch": 0.7653770491803279,
+      "grad_norm": 1.147824122343963,
+      "learning_rate": 0.00017457207923282823,
+      "loss": 13.527,
+      "step": 4377
+    },
+    {
+      "epoch": 0.765551912568306,
+      "grad_norm": 1.1578150614417397,
+      "learning_rate": 0.00017455949834320372,
+      "loss": 13.3908,
+      "step": 4378
+    },
+    {
+      "epoch": 0.7657267759562841,
+      "grad_norm": 1.0934807607424095,
+      "learning_rate": 0.0001745469147956202,
+      "loss": 13.1459,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7659016393442623,
+      "grad_norm": 1.0332073853511754,
+      "learning_rate": 0.00017453432859052627,
+      "loss": 13.2532,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7660765027322405,
+      "grad_norm": 1.052187910976826,
+      "learning_rate": 0.0001745217397283706,
+      "loss": 13.2555,
+      "step": 4381
+    },
+    {
+      "epoch": 0.7662513661202186,
+      "grad_norm": 1.0375933273291524,
+      "learning_rate": 0.000174509148209602,
+      "loss": 13.2551,
+      "step": 4382
+    },
+    {
+      "epoch": 0.7664262295081967,
+      "grad_norm": 1.0909615637360341,
+      "learning_rate": 0.0001744965540346693,
+      "loss": 13.2107,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7666010928961748,
+      "grad_norm": 1.1771331325930408,
+      "learning_rate": 0.00017448395720402152,
+      "loss": 13.3229,
+      "step": 4384
+    },
+    {
+      "epoch": 0.766775956284153,
+      "grad_norm": 1.1219783434873656,
+      "learning_rate": 0.00017447135771810762,
+      "loss": 13.4573,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7669508196721312,
+      "grad_norm": 0.9562598892513839,
+      "learning_rate": 0.0001744587555773769,
+      "loss": 12.989,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7671256830601093,
+      "grad_norm": 1.088974947028239,
+      "learning_rate": 0.0001744461507822785,
+      "loss": 13.4503,
+      "step": 4387
+    },
+    {
+      "epoch": 0.7673005464480874,
+      "grad_norm": 1.064659306202908,
+      "learning_rate": 0.00017443354333326183,
+      "loss": 13.0199,
+      "step": 4388
+    },
+    {
+      "epoch": 0.7674754098360655,
+      "grad_norm": 1.0805864245906895,
+      "learning_rate": 0.0001744209332307763,
+      "loss": 13.1378,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7676502732240437,
+      "grad_norm": 1.1905863651416573,
+      "learning_rate": 0.00017440832047527146,
+      "loss": 13.3131,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7678251366120219,
+      "grad_norm": 1.2625036724396932,
+      "learning_rate": 0.0001743957050671969,
+      "loss": 13.187,
+      "step": 4391
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 1.2077291333641984,
+      "learning_rate": 0.00017438308700700242,
+      "loss": 13.6207,
+      "step": 4392
+    },
+    {
+      "epoch": 0.7681748633879781,
+      "grad_norm": 1.3487371831146033,
+      "learning_rate": 0.00017437046629513776,
+      "loss": 13.4582,
+      "step": 4393
+    },
+    {
+      "epoch": 0.7683497267759563,
+      "grad_norm": 1.0557273190646381,
+      "learning_rate": 0.0001743578429320529,
+      "loss": 13.3714,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7685245901639344,
+      "grad_norm": 1.174761343203722,
+      "learning_rate": 0.00017434521691819777,
+      "loss": 13.4749,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7686994535519126,
+      "grad_norm": 1.146761396734963,
+      "learning_rate": 0.00017433258825402256,
+      "loss": 13.3389,
+      "step": 4396
+    },
+    {
+      "epoch": 0.7688743169398907,
+      "grad_norm": 1.226237892350042,
+      "learning_rate": 0.0001743199569399774,
+      "loss": 13.3662,
+      "step": 4397
+    },
+    {
+      "epoch": 0.7690491803278688,
+      "grad_norm": 1.0592394639325957,
+      "learning_rate": 0.00017430732297651262,
+      "loss": 13.1346,
+      "step": 4398
+    },
+    {
+      "epoch": 0.769224043715847,
+      "grad_norm": 1.29092016344245,
+      "learning_rate": 0.0001742946863640786,
+      "loss": 13.2863,
+      "step": 4399
+    },
+    {
+      "epoch": 0.7693989071038252,
+      "grad_norm": 1.1854129599161183,
+      "learning_rate": 0.0001742820471031258,
+      "loss": 13.3698,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7695737704918033,
+      "grad_norm": 1.0254070482834627,
+      "learning_rate": 0.00017426940519410484,
+      "loss": 13.199,
+      "step": 4401
+    },
+    {
+      "epoch": 0.7697486338797814,
+      "grad_norm": 1.1181604353044752,
+      "learning_rate": 0.00017425676063746634,
+      "loss": 13.3346,
+      "step": 4402
+    },
+    {
+      "epoch": 0.7699234972677595,
+      "grad_norm": 1.129668420196073,
+      "learning_rate": 0.00017424411343366106,
+      "loss": 12.8938,
+      "step": 4403
+    },
+    {
+      "epoch": 0.7700983606557377,
+      "grad_norm": 1.1929544972542374,
+      "learning_rate": 0.00017423146358313993,
+      "loss": 13.3744,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7702732240437159,
+      "grad_norm": 1.3227682664946063,
+      "learning_rate": 0.00017421881108635386,
+      "loss": 13.24,
+      "step": 4405
+    },
+    {
+      "epoch": 0.770448087431694,
+      "grad_norm": 1.1543094529780074,
+      "learning_rate": 0.00017420615594375383,
+      "loss": 13.0288,
+      "step": 4406
+    },
+    {
+      "epoch": 0.7706229508196721,
+      "grad_norm": 1.2480672456448385,
+      "learning_rate": 0.00017419349815579106,
+      "loss": 13.2714,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7707978142076503,
+      "grad_norm": 1.0394708749260075,
+      "learning_rate": 0.00017418083772291677,
+      "loss": 13.2667,
+      "step": 4408
+    },
+    {
+      "epoch": 0.7709726775956284,
+      "grad_norm": 1.0973642002472672,
+      "learning_rate": 0.00017416817464558233,
+      "loss": 13.2742,
+      "step": 4409
+    },
+    {
+      "epoch": 0.7711475409836066,
+      "grad_norm": 1.0676491526521117,
+      "learning_rate": 0.0001741555089242391,
+      "loss": 13.5218,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7713224043715847,
+      "grad_norm": 1.0550517277505218,
+      "learning_rate": 0.00017414284055933856,
+      "loss": 13.1554,
+      "step": 4411
+    },
+    {
+      "epoch": 0.7714972677595628,
+      "grad_norm": 1.0007066958523054,
+      "learning_rate": 0.00017413016955133243,
+      "loss": 13.0005,
+      "step": 4412
+    },
+    {
+      "epoch": 0.771672131147541,
+      "grad_norm": 1.112039937984147,
+      "learning_rate": 0.00017411749590067236,
+      "loss": 13.3601,
+      "step": 4413
+    },
+    {
+      "epoch": 0.7718469945355191,
+      "grad_norm": 0.9984040674941809,
+      "learning_rate": 0.00017410481960781015,
+      "loss": 13.1135,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7720218579234973,
+      "grad_norm": 1.1158447313390034,
+      "learning_rate": 0.0001740921406731977,
+      "loss": 13.2725,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7721967213114754,
+      "grad_norm": 1.194799976251758,
+      "learning_rate": 0.00017407945909728704,
+      "loss": 13.4915,
+      "step": 4416
+    },
+    {
+      "epoch": 0.7723715846994536,
+      "grad_norm": 1.0997120300189935,
+      "learning_rate": 0.0001740667748805302,
+      "loss": 13.3317,
+      "step": 4417
+    },
+    {
+      "epoch": 0.7725464480874317,
+      "grad_norm": 1.0707076602774193,
+      "learning_rate": 0.00017405408802337937,
+      "loss": 13.099,
+      "step": 4418
+    },
+    {
+      "epoch": 0.7727213114754098,
+      "grad_norm": 1.192543332787693,
+      "learning_rate": 0.0001740413985262868,
+      "loss": 13.361,
+      "step": 4419
+    },
+    {
+      "epoch": 0.772896174863388,
+      "grad_norm": 1.2697808193483944,
+      "learning_rate": 0.0001740287063897049,
+      "loss": 13.4621,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7730710382513661,
+      "grad_norm": 1.0003616353865457,
+      "learning_rate": 0.00017401601161408609,
+      "loss": 13.208,
+      "step": 4421
+    },
+    {
+      "epoch": 0.7732459016393443,
+      "grad_norm": 1.023736058379049,
+      "learning_rate": 0.00017400331419988298,
+      "loss": 13.1393,
+      "step": 4422
+    },
+    {
+      "epoch": 0.7734207650273224,
+      "grad_norm": 1.150502224676596,
+      "learning_rate": 0.00017399061414754817,
+      "loss": 13.4852,
+      "step": 4423
+    },
+    {
+      "epoch": 0.7735956284153005,
+      "grad_norm": 1.312518634137807,
+      "learning_rate": 0.00017397791145753438,
+      "loss": 13.0859,
+      "step": 4424
+    },
+    {
+      "epoch": 0.7737704918032787,
+      "grad_norm": 1.016181572660905,
+      "learning_rate": 0.00017396520613029455,
+      "loss": 13.2541,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7739453551912568,
+      "grad_norm": 1.1924441387135627,
+      "learning_rate": 0.0001739524981662815,
+      "loss": 13.2149,
+      "step": 4426
+    },
+    {
+      "epoch": 0.774120218579235,
+      "grad_norm": 1.1686921652183957,
+      "learning_rate": 0.0001739397875659483,
+      "loss": 13.357,
+      "step": 4427
+    },
+    {
+      "epoch": 0.7742950819672131,
+      "grad_norm": 1.0918210581078276,
+      "learning_rate": 0.00017392707432974805,
+      "loss": 13.2361,
+      "step": 4428
+    },
+    {
+      "epoch": 0.7744699453551912,
+      "grad_norm": 1.1976863407517733,
+      "learning_rate": 0.00017391435845813397,
+      "loss": 13.4217,
+      "step": 4429
+    },
+    {
+      "epoch": 0.7746448087431694,
+      "grad_norm": 0.9908341815754529,
+      "learning_rate": 0.0001739016399515594,
+      "loss": 13.3883,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7748196721311476,
+      "grad_norm": 1.2487232606113154,
+      "learning_rate": 0.00017388891881047768,
+      "loss": 13.2423,
+      "step": 4431
+    },
+    {
+      "epoch": 0.7749945355191257,
+      "grad_norm": 1.113949402855011,
+      "learning_rate": 0.00017387619503534237,
+      "loss": 13.2755,
+      "step": 4432
+    },
+    {
+      "epoch": 0.7751693989071038,
+      "grad_norm": 1.1366704014429023,
+      "learning_rate": 0.000173863468626607,
+      "loss": 13.187,
+      "step": 4433
+    },
+    {
+      "epoch": 0.775344262295082,
+      "grad_norm": 0.9933726050967537,
+      "learning_rate": 0.00017385073958472527,
+      "loss": 12.8675,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7755191256830601,
+      "grad_norm": 1.2660905925379289,
+      "learning_rate": 0.00017383800791015097,
+      "loss": 13.3401,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7756939890710383,
+      "grad_norm": 1.203805151353913,
+      "learning_rate": 0.00017382527360333794,
+      "loss": 13.5362,
+      "step": 4436
+    },
+    {
+      "epoch": 0.7758688524590164,
+      "grad_norm": 1.2195237932521876,
+      "learning_rate": 0.00017381253666474015,
+      "loss": 13.4693,
+      "step": 4437
+    },
+    {
+      "epoch": 0.7760437158469945,
+      "grad_norm": 1.19848702853387,
+      "learning_rate": 0.0001737997970948117,
+      "loss": 13.2456,
+      "step": 4438
+    },
+    {
+      "epoch": 0.7762185792349727,
+      "grad_norm": 1.0395620464230815,
+      "learning_rate": 0.0001737870548940067,
+      "loss": 13.3137,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7763934426229508,
+      "grad_norm": 1.0794509424387768,
+      "learning_rate": 0.00017377431006277943,
+      "loss": 13.2277,
+      "step": 4440
+    },
+    {
+      "epoch": 0.776568306010929,
+      "grad_norm": 1.2141079143793858,
+      "learning_rate": 0.00017376156260158417,
+      "loss": 13.4349,
+      "step": 4441
+    },
+    {
+      "epoch": 0.7767431693989071,
+      "grad_norm": 1.0400250594451081,
+      "learning_rate": 0.00017374881251087538,
+      "loss": 13.0245,
+      "step": 4442
+    },
+    {
+      "epoch": 0.7769180327868852,
+      "grad_norm": 1.075482015696506,
+      "learning_rate": 0.0001737360597911076,
+      "loss": 13.3923,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7770928961748634,
+      "grad_norm": 1.1823750686855996,
+      "learning_rate": 0.00017372330444273546,
+      "loss": 13.1826,
+      "step": 4444
+    },
+    {
+      "epoch": 0.7772677595628416,
+      "grad_norm": 0.9997327594896035,
+      "learning_rate": 0.0001737105464662136,
+      "loss": 13.2707,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7774426229508197,
+      "grad_norm": 1.1972370604897296,
+      "learning_rate": 0.00017369778586199695,
+      "loss": 12.9756,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7776174863387978,
+      "grad_norm": 1.116601357449739,
+      "learning_rate": 0.0001736850226305403,
+      "loss": 13.3125,
+      "step": 4447
+    },
+    {
+      "epoch": 0.7777923497267759,
+      "grad_norm": 0.9754594471922488,
+      "learning_rate": 0.00017367225677229867,
+      "loss": 13.1562,
+      "step": 4448
+    },
+    {
+      "epoch": 0.7779672131147541,
+      "grad_norm": 1.0604492331998907,
+      "learning_rate": 0.0001736594882877272,
+      "loss": 13.4176,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7781420765027323,
+      "grad_norm": 0.9922110871083953,
+      "learning_rate": 0.000173646717177281,
+      "loss": 13.0268,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7783169398907104,
+      "grad_norm": 1.0704723528293687,
+      "learning_rate": 0.00017363394344141543,
+      "loss": 12.92,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7784918032786885,
+      "grad_norm": 0.9822893313579842,
+      "learning_rate": 0.00017362116708058575,
+      "loss": 13.0878,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7786666666666666,
+      "grad_norm": 1.014121848640626,
+      "learning_rate": 0.00017360838809524754,
+      "loss": 13.3544,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7788415300546448,
+      "grad_norm": 1.0422587997476371,
+      "learning_rate": 0.00017359560648585624,
+      "loss": 13.2441,
+      "step": 4454
+    },
+    {
+      "epoch": 0.779016393442623,
+      "grad_norm": 1.0477679108131848,
+      "learning_rate": 0.00017358282225286758,
+      "loss": 13.2991,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7791912568306011,
+      "grad_norm": 1.0549193584325995,
+      "learning_rate": 0.00017357003539673728,
+      "loss": 13.1968,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7793661202185792,
+      "grad_norm": 1.168792092846992,
+      "learning_rate": 0.00017355724591792116,
+      "loss": 13.3162,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7795409836065573,
+      "grad_norm": 1.1144766223588214,
+      "learning_rate": 0.0001735444538168752,
+      "loss": 13.0455,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7797158469945356,
+      "grad_norm": 1.0875052519099806,
+      "learning_rate": 0.00017353165909405536,
+      "loss": 13.358,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7798907103825137,
+      "grad_norm": 1.067537872006528,
+      "learning_rate": 0.00017351886174991782,
+      "loss": 13.1955,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7800655737704918,
+      "grad_norm": 1.2134982082047925,
+      "learning_rate": 0.0001735060617849187,
+      "loss": 13.3432,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7802404371584699,
+      "grad_norm": 1.1619956228568127,
+      "learning_rate": 0.00017349325919951438,
+      "loss": 13.0888,
+      "step": 4462
+    },
+    {
+      "epoch": 0.780415300546448,
+      "grad_norm": 1.1673974303172379,
+      "learning_rate": 0.00017348045399416127,
+      "loss": 13.1271,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7805901639344263,
+      "grad_norm": 1.1841005128195274,
+      "learning_rate": 0.0001734676461693158,
+      "loss": 13.2573,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7807650273224044,
+      "grad_norm": 1.1098704624492313,
+      "learning_rate": 0.00017345483572543457,
+      "loss": 13.1798,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7809398907103825,
+      "grad_norm": 1.4578194354473213,
+      "learning_rate": 0.00017344202266297428,
+      "loss": 13.4957,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7811147540983606,
+      "grad_norm": 1.182382475154591,
+      "learning_rate": 0.0001734292069823917,
+      "loss": 13.4701,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7812896174863388,
+      "grad_norm": 1.3477014512659418,
+      "learning_rate": 0.0001734163886841437,
+      "loss": 13.2318,
+      "step": 4468
+    },
+    {
+      "epoch": 0.781464480874317,
+      "grad_norm": 1.5223072052184234,
+      "learning_rate": 0.00017340356776868721,
+      "loss": 13.2151,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7816393442622951,
+      "grad_norm": 1.1450648411971749,
+      "learning_rate": 0.00017339074423647927,
+      "loss": 13.3118,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7818142076502732,
+      "grad_norm": 1.6832775841040941,
+      "learning_rate": 0.00017337791808797707,
+      "loss": 13.43,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7819890710382513,
+      "grad_norm": 1.2141893330936602,
+      "learning_rate": 0.0001733650893236378,
+      "loss": 13.385,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7821639344262296,
+      "grad_norm": 1.3913621999735726,
+      "learning_rate": 0.00017335225794391882,
+      "loss": 13.2742,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7823387978142077,
+      "grad_norm": 1.4768032911637357,
+      "learning_rate": 0.00017333942394927757,
+      "loss": 13.0376,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7825136612021858,
+      "grad_norm": 1.1023411596136576,
+      "learning_rate": 0.0001733265873401715,
+      "loss": 12.9129,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7826885245901639,
+      "grad_norm": 1.466900333706247,
+      "learning_rate": 0.0001733137481170583,
+      "loss": 12.9852,
+      "step": 4476
+    },
+    {
+      "epoch": 0.782863387978142,
+      "grad_norm": 1.0121904730891351,
+      "learning_rate": 0.00017330090628039563,
+      "loss": 13.1913,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7830382513661203,
+      "grad_norm": 1.4768774047770987,
+      "learning_rate": 0.0001732880618306413,
+      "loss": 13.4071,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7832131147540984,
+      "grad_norm": 0.9386629161903606,
+      "learning_rate": 0.00017327521476825317,
+      "loss": 13.0795,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7833879781420765,
+      "grad_norm": 1.2418311972842977,
+      "learning_rate": 0.00017326236509368929,
+      "loss": 13.1932,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7835628415300546,
+      "grad_norm": 1.0533826584983237,
+      "learning_rate": 0.00017324951280740764,
+      "loss": 13.3318,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7837377049180327,
+      "grad_norm": 1.1238285160532293,
+      "learning_rate": 0.00017323665790986648,
+      "loss": 12.9095,
+      "step": 4482
+    },
+    {
+      "epoch": 0.783912568306011,
+      "grad_norm": 1.1393472891284349,
+      "learning_rate": 0.000173223800401524,
+      "loss": 13.1699,
+      "step": 4483
+    },
+    {
+      "epoch": 0.7840874316939891,
+      "grad_norm": 1.3491811038075343,
+      "learning_rate": 0.00017321094028283863,
+      "loss": 13.4994,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7842622950819672,
+      "grad_norm": 1.088571194466479,
+      "learning_rate": 0.00017319807755426873,
+      "loss": 13.3291,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7844371584699453,
+      "grad_norm": 1.237551570992867,
+      "learning_rate": 0.00017318521221627292,
+      "loss": 13.2941,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7846120218579234,
+      "grad_norm": 1.0439701995519355,
+      "learning_rate": 0.00017317234426930978,
+      "loss": 13.0406,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7847868852459017,
+      "grad_norm": 1.1295173049205391,
+      "learning_rate": 0.00017315947371383808,
+      "loss": 13.4777,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7849617486338798,
+      "grad_norm": 1.0695362389056249,
+      "learning_rate": 0.00017314660055031661,
+      "loss": 13.2887,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7851366120218579,
+      "grad_norm": 1.014908344094443,
+      "learning_rate": 0.0001731337247792043,
+      "loss": 13.3509,
+      "step": 4490
+    },
+    {
+      "epoch": 0.785311475409836,
+      "grad_norm": 1.1132565749048438,
+      "learning_rate": 0.00017312084640096017,
+      "loss": 13.389,
+      "step": 4491
+    },
+    {
+      "epoch": 0.7854863387978142,
+      "grad_norm": 1.1377081939368021,
+      "learning_rate": 0.00017310796541604326,
+      "loss": 13.4303,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7856612021857924,
+      "grad_norm": 1.032043271563996,
+      "learning_rate": 0.00017309508182491282,
+      "loss": 13.2811,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7858360655737705,
+      "grad_norm": 1.0838295375900635,
+      "learning_rate": 0.0001730821956280281,
+      "loss": 13.3901,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7860109289617486,
+      "grad_norm": 1.0834613348566788,
+      "learning_rate": 0.0001730693068258485,
+      "loss": 13.193,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7861857923497267,
+      "grad_norm": 1.0476990283657281,
+      "learning_rate": 0.0001730564154188335,
+      "loss": 13.4186,
+      "step": 4496
+    },
+    {
+      "epoch": 0.786360655737705,
+      "grad_norm": 1.2112879212469894,
+      "learning_rate": 0.00017304352140744262,
+      "loss": 13.057,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7865355191256831,
+      "grad_norm": 1.1863422229085996,
+      "learning_rate": 0.00017303062479213555,
+      "loss": 13.3803,
+      "step": 4498
+    },
+    {
+      "epoch": 0.7867103825136612,
+      "grad_norm": 0.9863952792115721,
+      "learning_rate": 0.00017301772557337206,
+      "loss": 13.1689,
+      "step": 4499
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 1.0962115521433535,
+      "learning_rate": 0.00017300482375161197,
+      "loss": 13.2743,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7870601092896174,
+      "grad_norm": 1.0030393504892594,
+      "learning_rate": 0.00017299191932731515,
+      "loss": 13.4123,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7872349726775957,
+      "grad_norm": 1.0260902670702765,
+      "learning_rate": 0.00017297901230094173,
+      "loss": 13.0905,
+      "step": 4502
+    },
+    {
+      "epoch": 0.7874098360655738,
+      "grad_norm": 0.9906780560242556,
+      "learning_rate": 0.0001729661026729518,
+      "loss": 13.1112,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7875846994535519,
+      "grad_norm": 0.9632768863310043,
+      "learning_rate": 0.00017295319044380552,
+      "loss": 13.3068,
+      "step": 4504
+    },
+    {
+      "epoch": 0.78775956284153,
+      "grad_norm": 0.9678049080054917,
+      "learning_rate": 0.0001729402756139633,
+      "loss": 13.1675,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7879344262295082,
+      "grad_norm": 0.9918332191575959,
+      "learning_rate": 0.0001729273581838854,
+      "loss": 13.1655,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7881092896174864,
+      "grad_norm": 1.0647298909634553,
+      "learning_rate": 0.00017291443815403246,
+      "loss": 13.2467,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7882841530054645,
+      "grad_norm": 0.9698522367534986,
+      "learning_rate": 0.00017290151552486495,
+      "loss": 13.0309,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7884590163934426,
+      "grad_norm": 1.093405349287798,
+      "learning_rate": 0.0001728885902968436,
+      "loss": 13.359,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7886338797814207,
+      "grad_norm": 1.0572706382659576,
+      "learning_rate": 0.00017287566247042917,
+      "loss": 13.225,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7888087431693989,
+      "grad_norm": 0.9769141426348387,
+      "learning_rate": 0.0001728627320460825,
+      "loss": 13.3329,
+      "step": 4511
+    },
+    {
+      "epoch": 0.7889836065573771,
+      "grad_norm": 1.0691514007164173,
+      "learning_rate": 0.00017284979902426457,
+      "loss": 13.4597,
+      "step": 4512
+    },
+    {
+      "epoch": 0.7891584699453552,
+      "grad_norm": 1.0094499722315307,
+      "learning_rate": 0.00017283686340543644,
+      "loss": 13.1114,
+      "step": 4513
+    },
+    {
+      "epoch": 0.7893333333333333,
+      "grad_norm": 1.0657089573491922,
+      "learning_rate": 0.0001728239251900592,
+      "loss": 13.2451,
+      "step": 4514
+    },
+    {
+      "epoch": 0.7895081967213115,
+      "grad_norm": 1.0691884614767238,
+      "learning_rate": 0.00017281098437859412,
+      "loss": 13.2762,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7896830601092896,
+      "grad_norm": 1.1684622713016644,
+      "learning_rate": 0.00017279804097150255,
+      "loss": 13.5182,
+      "step": 4516
+    },
+    {
+      "epoch": 0.7898579234972678,
+      "grad_norm": 0.9391779135992439,
+      "learning_rate": 0.00017278509496924583,
+      "loss": 13.2652,
+      "step": 4517
+    },
+    {
+      "epoch": 0.7900327868852459,
+      "grad_norm": 1.0543467261235109,
+      "learning_rate": 0.00017277214637228554,
+      "loss": 13.248,
+      "step": 4518
+    },
+    {
+      "epoch": 0.790207650273224,
+      "grad_norm": 1.0261045283463779,
+      "learning_rate": 0.00017275919518108324,
+      "loss": 13.4642,
+      "step": 4519
+    },
+    {
+      "epoch": 0.7903825136612022,
+      "grad_norm": 0.9326447366490629,
+      "learning_rate": 0.00017274624139610066,
+      "loss": 13.1088,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7905573770491803,
+      "grad_norm": 1.0290952497597243,
+      "learning_rate": 0.00017273328501779956,
+      "loss": 13.1885,
+      "step": 4521
+    },
+    {
+      "epoch": 0.7907322404371585,
+      "grad_norm": 0.9815256982448063,
+      "learning_rate": 0.00017272032604664185,
+      "loss": 13.0269,
+      "step": 4522
+    },
+    {
+      "epoch": 0.7909071038251366,
+      "grad_norm": 1.2150414314255429,
+      "learning_rate": 0.00017270736448308943,
+      "loss": 13.4512,
+      "step": 4523
+    },
+    {
+      "epoch": 0.7910819672131147,
+      "grad_norm": 1.1065458756079976,
+      "learning_rate": 0.0001726944003276045,
+      "loss": 13.3435,
+      "step": 4524
+    },
+    {
+      "epoch": 0.7912568306010929,
+      "grad_norm": 1.2612796479637145,
+      "learning_rate": 0.00017268143358064905,
+      "loss": 13.505,
+      "step": 4525
+    },
+    {
+      "epoch": 0.791431693989071,
+      "grad_norm": 1.0925430601137145,
+      "learning_rate": 0.00017266846424268545,
+      "loss": 13.136,
+      "step": 4526
+    },
+    {
+      "epoch": 0.7916065573770492,
+      "grad_norm": 1.117918550402917,
+      "learning_rate": 0.000172655492314176,
+      "loss": 13.0728,
+      "step": 4527
+    },
+    {
+      "epoch": 0.7917814207650273,
+      "grad_norm": 1.0166812748642862,
+      "learning_rate": 0.00017264251779558314,
+      "loss": 13.0538,
+      "step": 4528
+    },
+    {
+      "epoch": 0.7919562841530055,
+      "grad_norm": 1.2121443853634768,
+      "learning_rate": 0.00017262954068736938,
+      "loss": 12.8987,
+      "step": 4529
+    },
+    {
+      "epoch": 0.7921311475409836,
+      "grad_norm": 1.1105601881135536,
+      "learning_rate": 0.00017261656098999738,
+      "loss": 12.9943,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7923060109289618,
+      "grad_norm": 1.250337922055719,
+      "learning_rate": 0.0001726035787039298,
+      "loss": 13.2707,
+      "step": 4531
+    },
+    {
+      "epoch": 0.7924808743169399,
+      "grad_norm": 1.0350143217137526,
+      "learning_rate": 0.00017259059382962946,
+      "loss": 13.0753,
+      "step": 4532
+    },
+    {
+      "epoch": 0.792655737704918,
+      "grad_norm": 1.0463838847333524,
+      "learning_rate": 0.00017257760636755926,
+      "loss": 13.1839,
+      "step": 4533
+    },
+    {
+      "epoch": 0.7928306010928962,
+      "grad_norm": 0.9887922641808388,
+      "learning_rate": 0.00017256461631818222,
+      "loss": 13.1701,
+      "step": 4534
+    },
+    {
+      "epoch": 0.7930054644808743,
+      "grad_norm": 1.0483298571740616,
+      "learning_rate": 0.00017255162368196136,
+      "loss": 13.4219,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7931803278688525,
+      "grad_norm": 1.0081474817774019,
+      "learning_rate": 0.00017253862845935993,
+      "loss": 12.8833,
+      "step": 4536
+    },
+    {
+      "epoch": 0.7933551912568306,
+      "grad_norm": 1.1904308266389068,
+      "learning_rate": 0.0001725256306508411,
+      "loss": 13.2928,
+      "step": 4537
+    },
+    {
+      "epoch": 0.7935300546448087,
+      "grad_norm": 1.0026629542867214,
+      "learning_rate": 0.0001725126302568683,
+      "loss": 12.9651,
+      "step": 4538
+    },
+    {
+      "epoch": 0.7937049180327869,
+      "grad_norm": 1.1179050419728214,
+      "learning_rate": 0.00017249962727790497,
+      "loss": 13.3721,
+      "step": 4539
+    },
+    {
+      "epoch": 0.793879781420765,
+      "grad_norm": 0.9969169588045961,
+      "learning_rate": 0.00017248662171441458,
+      "loss": 13.3605,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7940546448087432,
+      "grad_norm": 1.0929108538376537,
+      "learning_rate": 0.00017247361356686084,
+      "loss": 13.3257,
+      "step": 4541
+    },
+    {
+      "epoch": 0.7942295081967213,
+      "grad_norm": 1.1390280049706198,
+      "learning_rate": 0.00017246060283570747,
+      "loss": 13.0797,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7944043715846995,
+      "grad_norm": 1.0490939941103141,
+      "learning_rate": 0.00017244758952141825,
+      "loss": 13.2013,
+      "step": 4543
+    },
+    {
+      "epoch": 0.7945792349726776,
+      "grad_norm": 1.189289720920158,
+      "learning_rate": 0.00017243457362445713,
+      "loss": 13.3709,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7947540983606557,
+      "grad_norm": 1.019298878156795,
+      "learning_rate": 0.00017242155514528808,
+      "loss": 13.0767,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7949289617486339,
+      "grad_norm": 1.3621242127805075,
+      "learning_rate": 0.0001724085340843752,
+      "loss": 13.113,
+      "step": 4546
+    },
+    {
+      "epoch": 0.795103825136612,
+      "grad_norm": 1.1349527343185728,
+      "learning_rate": 0.0001723955104421827,
+      "loss": 13.3538,
+      "step": 4547
+    },
+    {
+      "epoch": 0.7952786885245902,
+      "grad_norm": 1.027485275580688,
+      "learning_rate": 0.00017238248421917482,
+      "loss": 13.1146,
+      "step": 4548
+    },
+    {
+      "epoch": 0.7954535519125683,
+      "grad_norm": 1.1887370671611106,
+      "learning_rate": 0.00017236945541581594,
+      "loss": 13.3219,
+      "step": 4549
+    },
+    {
+      "epoch": 0.7956284153005464,
+      "grad_norm": 0.8876112199426491,
+      "learning_rate": 0.00017235642403257056,
+      "loss": 12.9355,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7958032786885246,
+      "grad_norm": 1.2002822854817508,
+      "learning_rate": 0.00017234339006990318,
+      "loss": 13.2561,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7959781420765027,
+      "grad_norm": 1.1152707084850997,
+      "learning_rate": 0.00017233035352827845,
+      "loss": 13.2671,
+      "step": 4552
+    },
+    {
+      "epoch": 0.7961530054644809,
+      "grad_norm": 1.066941680125543,
+      "learning_rate": 0.00017231731440816116,
+      "loss": 13.0994,
+      "step": 4553
+    },
+    {
+      "epoch": 0.796327868852459,
+      "grad_norm": 0.9743251315132265,
+      "learning_rate": 0.0001723042727100161,
+      "loss": 13.2916,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7965027322404371,
+      "grad_norm": 1.0891560613680753,
+      "learning_rate": 0.00017229122843430818,
+      "loss": 13.0356,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7966775956284153,
+      "grad_norm": 1.0426692635916404,
+      "learning_rate": 0.00017227818158150243,
+      "loss": 12.9019,
+      "step": 4556
+    },
+    {
+      "epoch": 0.7968524590163935,
+      "grad_norm": 1.1588858139147808,
+      "learning_rate": 0.00017226513215206398,
+      "loss": 13.3465,
+      "step": 4557
+    },
+    {
+      "epoch": 0.7970273224043716,
+      "grad_norm": 1.168451246586438,
+      "learning_rate": 0.00017225208014645798,
+      "loss": 13.2644,
+      "step": 4558
+    },
+    {
+      "epoch": 0.7972021857923497,
+      "grad_norm": 1.098781782616588,
+      "learning_rate": 0.00017223902556514974,
+      "loss": 13.1361,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7973770491803279,
+      "grad_norm": 1.3905602004813766,
+      "learning_rate": 0.00017222596840860463,
+      "loss": 13.2515,
+      "step": 4560
+    },
+    {
+      "epoch": 0.797551912568306,
+      "grad_norm": 1.0329652436815835,
+      "learning_rate": 0.00017221290867728816,
+      "loss": 13.0143,
+      "step": 4561
+    },
+    {
+      "epoch": 0.7977267759562842,
+      "grad_norm": 1.1338036596924934,
+      "learning_rate": 0.00017219984637166586,
+      "loss": 13.3124,
+      "step": 4562
+    },
+    {
+      "epoch": 0.7979016393442623,
+      "grad_norm": 0.9974958908354787,
+      "learning_rate": 0.0001721867814922034,
+      "loss": 13.1334,
+      "step": 4563
+    },
+    {
+      "epoch": 0.7980765027322404,
+      "grad_norm": 1.1833188930091922,
+      "learning_rate": 0.0001721737140393665,
+      "loss": 13.0481,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7982513661202186,
+      "grad_norm": 0.93829950688818,
+      "learning_rate": 0.00017216064401362103,
+      "loss": 13.0132,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7984262295081967,
+      "grad_norm": 1.32819429437167,
+      "learning_rate": 0.0001721475714154329,
+      "loss": 13.3076,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7986010928961749,
+      "grad_norm": 1.1767383760385774,
+      "learning_rate": 0.00017213449624526816,
+      "loss": 13.3794,
+      "step": 4567
+    },
+    {
+      "epoch": 0.798775956284153,
+      "grad_norm": 1.1668443940070155,
+      "learning_rate": 0.0001721214185035929,
+      "loss": 13.1902,
+      "step": 4568
+    },
+    {
+      "epoch": 0.7989508196721311,
+      "grad_norm": 1.0432494176644378,
+      "learning_rate": 0.00017210833819087334,
+      "loss": 13.2999,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7991256830601093,
+      "grad_norm": 1.462942483051223,
+      "learning_rate": 0.00017209525530757576,
+      "loss": 12.9226,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7993005464480875,
+      "grad_norm": 1.0923293987507214,
+      "learning_rate": 0.00017208216985416662,
+      "loss": 12.8879,
+      "step": 4571
+    },
+    {
+      "epoch": 0.7994754098360656,
+      "grad_norm": 1.1824646425983683,
+      "learning_rate": 0.0001720690818311123,
+      "loss": 13.3146,
+      "step": 4572
+    },
+    {
+      "epoch": 0.7996502732240437,
+      "grad_norm": 1.1357828987848213,
+      "learning_rate": 0.00017205599123887942,
+      "loss": 13.1438,
+      "step": 4573
+    },
+    {
+      "epoch": 0.7998251366120218,
+      "grad_norm": 1.0723829462009373,
+      "learning_rate": 0.0001720428980779346,
+      "loss": 13.0074,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4108630856669024,
+      "learning_rate": 0.00017202980234874468,
+      "loss": 13.3265,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8001748633879782,
+      "grad_norm": 1.0792831189469045,
+      "learning_rate": 0.00017201670405177647,
+      "loss": 13.1195,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8003497267759563,
+      "grad_norm": 1.2452358278397966,
+      "learning_rate": 0.00017200360318749688,
+      "loss": 12.791,
+      "step": 4577
+    },
+    {
+      "epoch": 0.8005245901639344,
+      "grad_norm": 1.0769461810495071,
+      "learning_rate": 0.000171990499756373,
+      "loss": 13.072,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8006994535519125,
+      "grad_norm": 1.2893586907249603,
+      "learning_rate": 0.00017197739375887189,
+      "loss": 13.4864,
+      "step": 4579
+    },
+    {
+      "epoch": 0.8008743169398908,
+      "grad_norm": 0.9705562015850532,
+      "learning_rate": 0.00017196428519546076,
+      "loss": 13.1809,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8010491803278689,
+      "grad_norm": 1.3230145090688967,
+      "learning_rate": 0.00017195117406660696,
+      "loss": 13.3468,
+      "step": 4581
+    },
+    {
+      "epoch": 0.801224043715847,
+      "grad_norm": 1.0071622818851065,
+      "learning_rate": 0.0001719380603727779,
+      "loss": 13.3143,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8013989071038251,
+      "grad_norm": 1.2417722540361373,
+      "learning_rate": 0.00017192494411444103,
+      "loss": 13.0616,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8015737704918032,
+      "grad_norm": 1.2366280669407594,
+      "learning_rate": 0.0001719118252920639,
+      "loss": 13.4264,
+      "step": 4584
+    },
+    {
+      "epoch": 0.8017486338797815,
+      "grad_norm": 1.2021731018901374,
+      "learning_rate": 0.00017189870390611428,
+      "loss": 13.0545,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8019234972677596,
+      "grad_norm": 1.141048913186511,
+      "learning_rate": 0.00017188557995705982,
+      "loss": 13.4543,
+      "step": 4586
+    },
+    {
+      "epoch": 0.8020983606557377,
+      "grad_norm": 1.354158619958263,
+      "learning_rate": 0.00017187245344536844,
+      "loss": 13.1987,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8022732240437158,
+      "grad_norm": 1.2221071537107366,
+      "learning_rate": 0.0001718593243715081,
+      "loss": 13.4435,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8024480874316939,
+      "grad_norm": 1.1794884595843984,
+      "learning_rate": 0.00017184619273594674,
+      "loss": 13.2197,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8026229508196722,
+      "grad_norm": 1.1976990628274706,
+      "learning_rate": 0.0001718330585391526,
+      "loss": 13.1109,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8027978142076503,
+      "grad_norm": 1.224491122903339,
+      "learning_rate": 0.00017181992178159385,
+      "loss": 13.2313,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8029726775956284,
+      "grad_norm": 0.9826271570764799,
+      "learning_rate": 0.00017180678246373878,
+      "loss": 13.2701,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8031475409836065,
+      "grad_norm": 1.1292961871485738,
+      "learning_rate": 0.00017179364058605583,
+      "loss": 13.3362,
+      "step": 4593
+    },
+    {
+      "epoch": 0.8033224043715848,
+      "grad_norm": 1.1262570953303135,
+      "learning_rate": 0.00017178049614901346,
+      "loss": 13.3229,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8034972677595629,
+      "grad_norm": 1.0229594117932217,
+      "learning_rate": 0.00017176734915308028,
+      "loss": 13.1324,
+      "step": 4595
+    },
+    {
+      "epoch": 0.803672131147541,
+      "grad_norm": 1.0654438377715905,
+      "learning_rate": 0.00017175419959872495,
+      "loss": 13.3989,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8038469945355191,
+      "grad_norm": 1.187584538349838,
+      "learning_rate": 0.00017174104748641624,
+      "loss": 13.3176,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8040218579234972,
+      "grad_norm": 1.1046886287221491,
+      "learning_rate": 0.000171727892816623,
+      "loss": 13.2229,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8041967213114755,
+      "grad_norm": 0.9790802716314042,
+      "learning_rate": 0.0001717147355898142,
+      "loss": 13.3817,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8043715846994536,
+      "grad_norm": 1.1959852352105165,
+      "learning_rate": 0.00017170157580645884,
+      "loss": 13.1027,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8045464480874317,
+      "grad_norm": 1.0480587326404165,
+      "learning_rate": 0.0001716884134670261,
+      "loss": 13.2454,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8047213114754098,
+      "grad_norm": 1.0707583434207593,
+      "learning_rate": 0.0001716752485719852,
+      "loss": 13.1694,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8048961748633879,
+      "grad_norm": 1.0433123365903174,
+      "learning_rate": 0.0001716620811218054,
+      "loss": 13.0852,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8050710382513662,
+      "grad_norm": 1.178265333780702,
+      "learning_rate": 0.00017164891111695613,
+      "loss": 13.3244,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8052459016393443,
+      "grad_norm": 1.0839095457993844,
+      "learning_rate": 0.00017163573855790692,
+      "loss": 13.3049,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8054207650273224,
+      "grad_norm": 1.032766113547239,
+      "learning_rate": 0.0001716225634451273,
+      "loss": 12.8118,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8055956284153005,
+      "grad_norm": 1.3388840966044846,
+      "learning_rate": 0.00017160938577908698,
+      "loss": 13.4595,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8057704918032786,
+      "grad_norm": 0.9452352126031467,
+      "learning_rate": 0.00017159620556025576,
+      "loss": 12.6382,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8059453551912569,
+      "grad_norm": 1.3406356686481946,
+      "learning_rate": 0.00017158302278910342,
+      "loss": 13.2291,
+      "step": 4609
+    },
+    {
+      "epoch": 0.806120218579235,
+      "grad_norm": 1.091367685410862,
+      "learning_rate": 0.00017156983746609994,
+      "loss": 13.2604,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8062950819672131,
+      "grad_norm": 1.0318870634732717,
+      "learning_rate": 0.00017155664959171542,
+      "loss": 13.2714,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8064699453551912,
+      "grad_norm": 1.1148355526211864,
+      "learning_rate": 0.00017154345916641991,
+      "loss": 13.0275,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8066448087431694,
+      "grad_norm": 1.0765679107304855,
+      "learning_rate": 0.0001715302661906837,
+      "loss": 13.103,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8068196721311476,
+      "grad_norm": 0.9702293740781631,
+      "learning_rate": 0.00017151707066497706,
+      "loss": 13.1033,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8069945355191257,
+      "grad_norm": 0.9823717033272834,
+      "learning_rate": 0.00017150387258977042,
+      "loss": 13.1262,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8071693989071038,
+      "grad_norm": 1.0462438358306725,
+      "learning_rate": 0.00017149067196553427,
+      "loss": 13.429,
+      "step": 4616
+    },
+    {
+      "epoch": 0.8073442622950819,
+      "grad_norm": 1.0497953118872312,
+      "learning_rate": 0.00017147746879273918,
+      "loss": 13.1208,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8075191256830601,
+      "grad_norm": 1.0168164078854183,
+      "learning_rate": 0.00017146426307185583,
+      "loss": 13.2515,
+      "step": 4618
+    },
+    {
+      "epoch": 0.8076939890710383,
+      "grad_norm": 1.0891769362165826,
+      "learning_rate": 0.000171451054803355,
+      "loss": 13.1898,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8078688524590164,
+      "grad_norm": 1.0733282385456493,
+      "learning_rate": 0.00017143784398770758,
+      "loss": 13.0045,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8080437158469945,
+      "grad_norm": 1.016473171609589,
+      "learning_rate": 0.00017142463062538445,
+      "loss": 13.4693,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8082185792349726,
+      "grad_norm": 1.1084344755023416,
+      "learning_rate": 0.0001714114147168567,
+      "loss": 13.2718,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8083934426229509,
+      "grad_norm": 1.3887172778364578,
+      "learning_rate": 0.00017139819626259546,
+      "loss": 13.1649,
+      "step": 4623
+    },
+    {
+      "epoch": 0.808568306010929,
+      "grad_norm": 1.1536341201662386,
+      "learning_rate": 0.00017138497526307194,
+      "loss": 13.2869,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8087431693989071,
+      "grad_norm": 1.4782862788523683,
+      "learning_rate": 0.00017137175171875747,
+      "loss": 13.234,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8089180327868852,
+      "grad_norm": 1.1565689369075327,
+      "learning_rate": 0.0001713585256301234,
+      "loss": 13.2469,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8090928961748634,
+      "grad_norm": 1.505039899473979,
+      "learning_rate": 0.0001713452969976413,
+      "loss": 13.2221,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8092677595628416,
+      "grad_norm": 1.1215556407520693,
+      "learning_rate": 0.00017133206582178268,
+      "loss": 13.2808,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8094426229508197,
+      "grad_norm": 1.5070539936193605,
+      "learning_rate": 0.00017131883210301928,
+      "loss": 13.0301,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8096174863387978,
+      "grad_norm": 1.0872084839078284,
+      "learning_rate": 0.00017130559584182286,
+      "loss": 13.1055,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8097923497267759,
+      "grad_norm": 1.3876469448316042,
+      "learning_rate": 0.00017129235703866526,
+      "loss": 13.2273,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8099672131147541,
+      "grad_norm": 1.076150804348979,
+      "learning_rate": 0.00017127911569401842,
+      "loss": 13.1638,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8101420765027323,
+      "grad_norm": 1.2309850974957741,
+      "learning_rate": 0.00017126587180835436,
+      "loss": 13.0262,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8103169398907104,
+      "grad_norm": 1.018506775361349,
+      "learning_rate": 0.0001712526253821453,
+      "loss": 13.0632,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8104918032786885,
+      "grad_norm": 1.5866093614135903,
+      "learning_rate": 0.00017123937641586333,
+      "loss": 13.3014,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8106666666666666,
+      "grad_norm": 1.097963629252059,
+      "learning_rate": 0.00017122612490998084,
+      "loss": 13.0343,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8108415300546448,
+      "grad_norm": 1.3749151110059166,
+      "learning_rate": 0.0001712128708649702,
+      "loss": 13.2306,
+      "step": 4637
+    },
+    {
+      "epoch": 0.811016393442623,
+      "grad_norm": 1.1943110518949531,
+      "learning_rate": 0.000171199614281304,
+      "loss": 13.0393,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8111912568306011,
+      "grad_norm": 1.2082414495486342,
+      "learning_rate": 0.00017118635515945462,
+      "loss": 13.1852,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8113661202185792,
+      "grad_norm": 1.0023724015196207,
+      "learning_rate": 0.00017117309349989492,
+      "loss": 13.0837,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8115409836065574,
+      "grad_norm": 1.2231971880402088,
+      "learning_rate": 0.0001711598293030976,
+      "loss": 13.3379,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8117158469945355,
+      "grad_norm": 1.1072298244321126,
+      "learning_rate": 0.00017114656256953548,
+      "loss": 13.3651,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8118907103825137,
+      "grad_norm": 1.2353500284002419,
+      "learning_rate": 0.00017113329329968154,
+      "loss": 13.0696,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8120655737704918,
+      "grad_norm": 1.1340508352367593,
+      "learning_rate": 0.00017112002149400878,
+      "loss": 13.1907,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8122404371584699,
+      "grad_norm": 1.2418233470771165,
+      "learning_rate": 0.0001711067471529904,
+      "loss": 13.5233,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8124153005464481,
+      "grad_norm": 1.0469933512560265,
+      "learning_rate": 0.00017109347027709953,
+      "loss": 13.3609,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8125901639344262,
+      "grad_norm": 1.0938270624665392,
+      "learning_rate": 0.00017108019086680952,
+      "loss": 13.254,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8127650273224044,
+      "grad_norm": 1.053931312784866,
+      "learning_rate": 0.00017106690892259376,
+      "loss": 13.3132,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8129398907103825,
+      "grad_norm": 1.2715912190632968,
+      "learning_rate": 0.00017105362444492573,
+      "loss": 13.2256,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8131147540983606,
+      "grad_norm": 1.0164654804137658,
+      "learning_rate": 0.00017104033743427897,
+      "loss": 13.237,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8132896174863388,
+      "grad_norm": 1.3012203637129898,
+      "learning_rate": 0.0001710270478911272,
+      "loss": 13.1723,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8134644808743169,
+      "grad_norm": 1.1218929522755734,
+      "learning_rate": 0.00017101375581594417,
+      "loss": 13.2856,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8136393442622951,
+      "grad_norm": 0.9980869850767224,
+      "learning_rate": 0.0001710004612092037,
+      "loss": 13.0299,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8138142076502732,
+      "grad_norm": 1.0631823980646573,
+      "learning_rate": 0.00017098716407137977,
+      "loss": 13.0411,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8139890710382514,
+      "grad_norm": 1.2844267873038717,
+      "learning_rate": 0.00017097386440294635,
+      "loss": 13.1627,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8141639344262295,
+      "grad_norm": 1.0559202060197888,
+      "learning_rate": 0.0001709605622043776,
+      "loss": 13.1632,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8143387978142077,
+      "grad_norm": 1.2524403243665276,
+      "learning_rate": 0.0001709472574761477,
+      "loss": 13.317,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8145136612021858,
+      "grad_norm": 1.0083595847471656,
+      "learning_rate": 0.000170933950218731,
+      "loss": 13.2786,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8146885245901639,
+      "grad_norm": 1.1911317603357232,
+      "learning_rate": 0.00017092064043260183,
+      "loss": 13.3226,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8148633879781421,
+      "grad_norm": 1.1339623621325483,
+      "learning_rate": 0.00017090732811823469,
+      "loss": 13.2789,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8150382513661202,
+      "grad_norm": 0.9711249442803134,
+      "learning_rate": 0.00017089401327610414,
+      "loss": 13.236,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8152131147540984,
+      "grad_norm": 1.1222497938788845,
+      "learning_rate": 0.00017088069590668486,
+      "loss": 13.0051,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8153879781420765,
+      "grad_norm": 1.0146307586626688,
+      "learning_rate": 0.0001708673760104516,
+      "loss": 13.1873,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8155628415300546,
+      "grad_norm": 1.0553547944336037,
+      "learning_rate": 0.00017085405358787918,
+      "loss": 13.2275,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8157377049180328,
+      "grad_norm": 1.135304193761643,
+      "learning_rate": 0.00017084072863944255,
+      "loss": 13.1321,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8159125683060109,
+      "grad_norm": 1.160644059766452,
+      "learning_rate": 0.00017082740116561667,
+      "loss": 13.2409,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8160874316939891,
+      "grad_norm": 1.0275514974261344,
+      "learning_rate": 0.00017081407116687668,
+      "loss": 13.2025,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8162622950819672,
+      "grad_norm": 1.0890741600865401,
+      "learning_rate": 0.00017080073864369783,
+      "loss": 13.2389,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8164371584699454,
+      "grad_norm": 1.0361442596096584,
+      "learning_rate": 0.00017078740359655536,
+      "loss": 13.3719,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8166120218579235,
+      "grad_norm": 1.036632870003715,
+      "learning_rate": 0.00017077406602592467,
+      "loss": 13.1234,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8167868852459016,
+      "grad_norm": 1.0469856687064367,
+      "learning_rate": 0.00017076072593228122,
+      "loss": 12.8605,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8169617486338798,
+      "grad_norm": 1.0516154497521835,
+      "learning_rate": 0.00017074738331610055,
+      "loss": 13.2497,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8171366120218579,
+      "grad_norm": 1.0404621287489573,
+      "learning_rate": 0.0001707340381778583,
+      "loss": 13.2857,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8173114754098361,
+      "grad_norm": 1.012844640709977,
+      "learning_rate": 0.00017072069051803025,
+      "loss": 13.3164,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8174863387978142,
+      "grad_norm": 1.133455712602176,
+      "learning_rate": 0.00017070734033709223,
+      "loss": 13.3412,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8176612021857923,
+      "grad_norm": 1.00612501130817,
+      "learning_rate": 0.0001706939876355201,
+      "loss": 13.4745,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8178360655737705,
+      "grad_norm": 1.1522601184347765,
+      "learning_rate": 0.00017068063241378994,
+      "loss": 13.1956,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8180109289617487,
+      "grad_norm": 1.244514440194398,
+      "learning_rate": 0.00017066727467237783,
+      "loss": 13.0218,
+      "step": 4678
+    },
+    {
+      "epoch": 0.8181857923497268,
+      "grad_norm": 1.0762217712140727,
+      "learning_rate": 0.0001706539144117599,
+      "loss": 13.2251,
+      "step": 4679
+    },
+    {
+      "epoch": 0.8183606557377049,
+      "grad_norm": 1.2172785253053886,
+      "learning_rate": 0.0001706405516324125,
+      "loss": 13.232,
+      "step": 4680
+    },
+    {
+      "epoch": 0.818535519125683,
+      "grad_norm": 1.2254360539679026,
+      "learning_rate": 0.000170627186334812,
+      "loss": 13.499,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8187103825136612,
+      "grad_norm": 1.0685387283725805,
+      "learning_rate": 0.00017061381851943475,
+      "loss": 13.0088,
+      "step": 4682
+    },
+    {
+      "epoch": 0.8188852459016394,
+      "grad_norm": 1.28667541759663,
+      "learning_rate": 0.00017060044818675745,
+      "loss": 13.0659,
+      "step": 4683
+    },
+    {
+      "epoch": 0.8190601092896175,
+      "grad_norm": 1.0685376753817937,
+      "learning_rate": 0.0001705870753372566,
+      "loss": 13.3616,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8192349726775956,
+      "grad_norm": 1.0532447009132293,
+      "learning_rate": 0.00017057369997140902,
+      "loss": 13.2143,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8194098360655737,
+      "grad_norm": 1.1178116026265774,
+      "learning_rate": 0.00017056032208969147,
+      "loss": 13.0286,
+      "step": 4686
+    },
+    {
+      "epoch": 0.819584699453552,
+      "grad_norm": 1.246025638918086,
+      "learning_rate": 0.0001705469416925809,
+      "loss": 13.3337,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8197595628415301,
+      "grad_norm": 1.0477786709971102,
+      "learning_rate": 0.00017053355878055427,
+      "loss": 13.2782,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8199344262295082,
+      "grad_norm": 1.2045180980148116,
+      "learning_rate": 0.0001705201733540887,
+      "loss": 13.3559,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8201092896174863,
+      "grad_norm": 1.1485369265484282,
+      "learning_rate": 0.0001705067854136613,
+      "loss": 13.4987,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8202841530054645,
+      "grad_norm": 1.1033502336525467,
+      "learning_rate": 0.00017049339495974942,
+      "loss": 13.1581,
+      "step": 4691
+    },
+    {
+      "epoch": 0.8204590163934427,
+      "grad_norm": 1.0006673279904088,
+      "learning_rate": 0.00017048000199283036,
+      "loss": 13.2993,
+      "step": 4692
+    },
+    {
+      "epoch": 0.8206338797814208,
+      "grad_norm": 1.1769403509661873,
+      "learning_rate": 0.00017046660651338157,
+      "loss": 13.38,
+      "step": 4693
+    },
+    {
+      "epoch": 0.8208087431693989,
+      "grad_norm": 1.0580196406335056,
+      "learning_rate": 0.0001704532085218806,
+      "loss": 13.4451,
+      "step": 4694
+    },
+    {
+      "epoch": 0.820983606557377,
+      "grad_norm": 1.3656232397407577,
+      "learning_rate": 0.00017043980801880504,
+      "loss": 13.2036,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8211584699453552,
+      "grad_norm": 1.0709824484924444,
+      "learning_rate": 0.00017042640500463264,
+      "loss": 13.1952,
+      "step": 4696
+    },
+    {
+      "epoch": 0.8213333333333334,
+      "grad_norm": 1.2677451595124665,
+      "learning_rate": 0.0001704129994798412,
+      "loss": 13.2092,
+      "step": 4697
+    },
+    {
+      "epoch": 0.8215081967213115,
+      "grad_norm": 1.1830903138560451,
+      "learning_rate": 0.00017039959144490857,
+      "loss": 13.0523,
+      "step": 4698
+    },
+    {
+      "epoch": 0.8216830601092896,
+      "grad_norm": 1.0951100399770493,
+      "learning_rate": 0.0001703861809003128,
+      "loss": 13.1045,
+      "step": 4699
+    },
+    {
+      "epoch": 0.8218579234972677,
+      "grad_norm": 1.1891279572044318,
+      "learning_rate": 0.00017037276784653187,
+      "loss": 13.2815,
+      "step": 4700
+    },
+    {
+      "epoch": 0.822032786885246,
+      "grad_norm": 1.1801484769913342,
+      "learning_rate": 0.000170359352284044,
+      "loss": 13.3245,
+      "step": 4701
+    },
+    {
+      "epoch": 0.8222076502732241,
+      "grad_norm": 1.223531723372345,
+      "learning_rate": 0.00017034593421332743,
+      "loss": 13.1514,
+      "step": 4702
+    },
+    {
+      "epoch": 0.8223825136612022,
+      "grad_norm": 1.0313881946569836,
+      "learning_rate": 0.00017033251363486054,
+      "loss": 13.2422,
+      "step": 4703
+    },
+    {
+      "epoch": 0.8225573770491803,
+      "grad_norm": 1.1073748395964127,
+      "learning_rate": 0.00017031909054912166,
+      "loss": 12.8551,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8227322404371584,
+      "grad_norm": 0.986767715844882,
+      "learning_rate": 0.00017030566495658937,
+      "loss": 13.1059,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8229071038251367,
+      "grad_norm": 1.002329552432758,
+      "learning_rate": 0.00017029223685774224,
+      "loss": 13.2875,
+      "step": 4706
+    },
+    {
+      "epoch": 0.8230819672131148,
+      "grad_norm": 1.0312746578338234,
+      "learning_rate": 0.00017027880625305905,
+      "loss": 13.0706,
+      "step": 4707
+    },
+    {
+      "epoch": 0.8232568306010929,
+      "grad_norm": 1.00503871380246,
+      "learning_rate": 0.0001702653731430185,
+      "loss": 13.1731,
+      "step": 4708
+    },
+    {
+      "epoch": 0.823431693989071,
+      "grad_norm": 1.0087103635884849,
+      "learning_rate": 0.0001702519375280995,
+      "loss": 13.2515,
+      "step": 4709
+    },
+    {
+      "epoch": 0.8236065573770491,
+      "grad_norm": 1.3118924546197457,
+      "learning_rate": 0.000170238499408781,
+      "loss": 13.1826,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8237814207650274,
+      "grad_norm": 0.9668312361062673,
+      "learning_rate": 0.00017022505878554208,
+      "loss": 12.7876,
+      "step": 4711
+    },
+    {
+      "epoch": 0.8239562841530055,
+      "grad_norm": 1.208306257840741,
+      "learning_rate": 0.00017021161565886186,
+      "loss": 13.1977,
+      "step": 4712
+    },
+    {
+      "epoch": 0.8241311475409836,
+      "grad_norm": 0.9842095918299907,
+      "learning_rate": 0.00017019817002921956,
+      "loss": 13.1951,
+      "step": 4713
+    },
+    {
+      "epoch": 0.8243060109289617,
+      "grad_norm": 1.2807448925987572,
+      "learning_rate": 0.00017018472189709452,
+      "loss": 13.2964,
+      "step": 4714
+    },
+    {
+      "epoch": 0.8244808743169398,
+      "grad_norm": 1.1446818731612376,
+      "learning_rate": 0.00017017127126296616,
+      "loss": 12.9917,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8246557377049181,
+      "grad_norm": 1.0576540821095655,
+      "learning_rate": 0.00017015781812731393,
+      "loss": 12.9596,
+      "step": 4716
+    },
+    {
+      "epoch": 0.8248306010928962,
+      "grad_norm": 1.1297225529107155,
+      "learning_rate": 0.00017014436249061752,
+      "loss": 13.1572,
+      "step": 4717
+    },
+    {
+      "epoch": 0.8250054644808743,
+      "grad_norm": 1.031485158187455,
+      "learning_rate": 0.00017013090435335647,
+      "loss": 13.048,
+      "step": 4718
+    },
+    {
+      "epoch": 0.8251803278688524,
+      "grad_norm": 1.174864487170843,
+      "learning_rate": 0.00017011744371601067,
+      "loss": 13.3435,
+      "step": 4719
+    },
+    {
+      "epoch": 0.8253551912568307,
+      "grad_norm": 1.0170158332539994,
+      "learning_rate": 0.0001701039805790599,
+      "loss": 12.9225,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8255300546448088,
+      "grad_norm": 1.3425548366468953,
+      "learning_rate": 0.00017009051494298413,
+      "loss": 13.5911,
+      "step": 4721
+    },
+    {
+      "epoch": 0.8257049180327869,
+      "grad_norm": 0.9960717177367561,
+      "learning_rate": 0.0001700770468082634,
+      "loss": 13.15,
+      "step": 4722
+    },
+    {
+      "epoch": 0.825879781420765,
+      "grad_norm": 1.3394169436350134,
+      "learning_rate": 0.00017006357617537784,
+      "loss": 13.3589,
+      "step": 4723
+    },
+    {
+      "epoch": 0.8260546448087431,
+      "grad_norm": 1.0307557533199605,
+      "learning_rate": 0.00017005010304480762,
+      "loss": 13.1063,
+      "step": 4724
+    },
+    {
+      "epoch": 0.8262295081967214,
+      "grad_norm": 1.5544983451761172,
+      "learning_rate": 0.00017003662741703308,
+      "loss": 13.1992,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8264043715846995,
+      "grad_norm": 1.0638884338855692,
+      "learning_rate": 0.00017002314929253466,
+      "loss": 13.0703,
+      "step": 4726
+    },
+    {
+      "epoch": 0.8265792349726776,
+      "grad_norm": 1.3644368339288475,
+      "learning_rate": 0.00017000966867179268,
+      "loss": 13.2917,
+      "step": 4727
+    },
+    {
+      "epoch": 0.8267540983606557,
+      "grad_norm": 0.9869058700935681,
+      "learning_rate": 0.00016999618555528788,
+      "loss": 13.1074,
+      "step": 4728
+    },
+    {
+      "epoch": 0.8269289617486338,
+      "grad_norm": 1.4046542498624752,
+      "learning_rate": 0.0001699826999435008,
+      "loss": 13.2727,
+      "step": 4729
+    },
+    {
+      "epoch": 0.8271038251366121,
+      "grad_norm": 1.0425857543772457,
+      "learning_rate": 0.00016996921183691228,
+      "loss": 13.2523,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8272786885245902,
+      "grad_norm": 1.1731690901833107,
+      "learning_rate": 0.00016995572123600309,
+      "loss": 13.0009,
+      "step": 4731
+    },
+    {
+      "epoch": 0.8274535519125683,
+      "grad_norm": 1.0510755959661622,
+      "learning_rate": 0.00016994222814125415,
+      "loss": 13.2819,
+      "step": 4732
+    },
+    {
+      "epoch": 0.8276284153005464,
+      "grad_norm": 1.1809140812549093,
+      "learning_rate": 0.00016992873255314655,
+      "loss": 13.0209,
+      "step": 4733
+    },
+    {
+      "epoch": 0.8278032786885245,
+      "grad_norm": 0.978616727178935,
+      "learning_rate": 0.00016991523447216128,
+      "loss": 13.0674,
+      "step": 4734
+    },
+    {
+      "epoch": 0.8279781420765028,
+      "grad_norm": 1.4118555511814466,
+      "learning_rate": 0.0001699017338987796,
+      "loss": 13.4172,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8281530054644809,
+      "grad_norm": 0.9495289780171372,
+      "learning_rate": 0.00016988823083348282,
+      "loss": 13.1706,
+      "step": 4736
+    },
+    {
+      "epoch": 0.828327868852459,
+      "grad_norm": 1.0885775908547497,
+      "learning_rate": 0.00016987472527675224,
+      "loss": 13.2528,
+      "step": 4737
+    },
+    {
+      "epoch": 0.8285027322404371,
+      "grad_norm": 1.070116105512822,
+      "learning_rate": 0.0001698612172290693,
+      "loss": 13.3557,
+      "step": 4738
+    },
+    {
+      "epoch": 0.8286775956284153,
+      "grad_norm": 1.0042235709832468,
+      "learning_rate": 0.00016984770669091565,
+      "loss": 13.054,
+      "step": 4739
+    },
+    {
+      "epoch": 0.8288524590163935,
+      "grad_norm": 0.965254720181685,
+      "learning_rate": 0.00016983419366277286,
+      "loss": 13.1748,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8290273224043716,
+      "grad_norm": 1.194485184270313,
+      "learning_rate": 0.00016982067814512263,
+      "loss": 12.9329,
+      "step": 4741
+    },
+    {
+      "epoch": 0.8292021857923497,
+      "grad_norm": 0.9808998289430918,
+      "learning_rate": 0.0001698071601384468,
+      "loss": 13.2592,
+      "step": 4742
+    },
+    {
+      "epoch": 0.8293770491803278,
+      "grad_norm": 1.072178382534396,
+      "learning_rate": 0.00016979363964322733,
+      "loss": 12.866,
+      "step": 4743
+    },
+    {
+      "epoch": 0.829551912568306,
+      "grad_norm": 0.9808921482293745,
+      "learning_rate": 0.00016978011665994607,
+      "loss": 13.1125,
+      "step": 4744
+    },
+    {
+      "epoch": 0.8297267759562842,
+      "grad_norm": 1.1016067573650872,
+      "learning_rate": 0.00016976659118908525,
+      "loss": 13.1887,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8299016393442623,
+      "grad_norm": 1.081011160025717,
+      "learning_rate": 0.00016975306323112693,
+      "loss": 13.0957,
+      "step": 4746
+    },
+    {
+      "epoch": 0.8300765027322404,
+      "grad_norm": 1.0868943345818762,
+      "learning_rate": 0.00016973953278655342,
+      "loss": 13.0448,
+      "step": 4747
+    },
+    {
+      "epoch": 0.8302513661202185,
+      "grad_norm": 0.9868329282459617,
+      "learning_rate": 0.000169725999855847,
+      "loss": 13.1163,
+      "step": 4748
+    },
+    {
+      "epoch": 0.8304262295081967,
+      "grad_norm": 1.0779194021908531,
+      "learning_rate": 0.0001697124644394902,
+      "loss": 12.9552,
+      "step": 4749
+    },
+    {
+      "epoch": 0.8306010928961749,
+      "grad_norm": 1.121503155445143,
+      "learning_rate": 0.0001696989265379655,
+      "loss": 13.2221,
+      "step": 4750
+    },
+    {
+      "epoch": 0.830775956284153,
+      "grad_norm": 1.0081442223661243,
+      "learning_rate": 0.00016968538615175547,
+      "loss": 13.2103,
+      "step": 4751
+    },
+    {
+      "epoch": 0.8309508196721311,
+      "grad_norm": 1.0035237266422832,
+      "learning_rate": 0.00016967184328134285,
+      "loss": 13.0049,
+      "step": 4752
+    },
+    {
+      "epoch": 0.8311256830601093,
+      "grad_norm": 1.1320413777705967,
+      "learning_rate": 0.00016965829792721043,
+      "loss": 13.1259,
+      "step": 4753
+    },
+    {
+      "epoch": 0.8313005464480875,
+      "grad_norm": 1.0604989625002292,
+      "learning_rate": 0.0001696447500898411,
+      "loss": 13.2605,
+      "step": 4754
+    },
+    {
+      "epoch": 0.8314754098360656,
+      "grad_norm": 1.043317727547491,
+      "learning_rate": 0.00016963119976971775,
+      "loss": 13.3967,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8316502732240437,
+      "grad_norm": 1.0623017067033043,
+      "learning_rate": 0.0001696176469673235,
+      "loss": 13.22,
+      "step": 4756
+    },
+    {
+      "epoch": 0.8318251366120218,
+      "grad_norm": 1.0812196748226661,
+      "learning_rate": 0.00016960409168314148,
+      "loss": 13.3487,
+      "step": 4757
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 1.1090201763362149,
+      "learning_rate": 0.00016959053391765494,
+      "loss": 13.2022,
+      "step": 4758
+    },
+    {
+      "epoch": 0.8321748633879782,
+      "grad_norm": 0.9835813876618725,
+      "learning_rate": 0.00016957697367134714,
+      "loss": 12.9108,
+      "step": 4759
+    },
+    {
+      "epoch": 0.8323497267759563,
+      "grad_norm": 1.046057346962884,
+      "learning_rate": 0.00016956341094470152,
+      "loss": 13.1988,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8325245901639344,
+      "grad_norm": 1.0031410766100706,
+      "learning_rate": 0.0001695498457382016,
+      "loss": 13.0984,
+      "step": 4761
+    },
+    {
+      "epoch": 0.8326994535519125,
+      "grad_norm": 1.0021870048455344,
+      "learning_rate": 0.0001695362780523309,
+      "loss": 13.2447,
+      "step": 4762
+    },
+    {
+      "epoch": 0.8328743169398907,
+      "grad_norm": 1.0488092637684887,
+      "learning_rate": 0.00016952270788757318,
+      "loss": 13.279,
+      "step": 4763
+    },
+    {
+      "epoch": 0.8330491803278689,
+      "grad_norm": 1.0269073347247257,
+      "learning_rate": 0.0001695091352444121,
+      "loss": 13.2977,
+      "step": 4764
+    },
+    {
+      "epoch": 0.833224043715847,
+      "grad_norm": 1.079695591947614,
+      "learning_rate": 0.00016949556012333164,
+      "loss": 13.3626,
+      "step": 4765
+    },
+    {
+      "epoch": 0.8333989071038251,
+      "grad_norm": 1.0827760249855367,
+      "learning_rate": 0.0001694819825248156,
+      "loss": 13.2417,
+      "step": 4766
+    },
+    {
+      "epoch": 0.8335737704918033,
+      "grad_norm": 1.2034627981096104,
+      "learning_rate": 0.00016946840244934808,
+      "loss": 13.2817,
+      "step": 4767
+    },
+    {
+      "epoch": 0.8337486338797814,
+      "grad_norm": 1.0856757533651453,
+      "learning_rate": 0.00016945481989741315,
+      "loss": 13.3181,
+      "step": 4768
+    },
+    {
+      "epoch": 0.8339234972677596,
+      "grad_norm": 1.1564782291677558,
+      "learning_rate": 0.00016944123486949507,
+      "loss": 13.168,
+      "step": 4769
+    },
+    {
+      "epoch": 0.8340983606557377,
+      "grad_norm": 1.1834007441813699,
+      "learning_rate": 0.00016942764736607808,
+      "loss": 13.0343,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8342732240437158,
+      "grad_norm": 0.9683996460667018,
+      "learning_rate": 0.00016941405738764658,
+      "loss": 13.0795,
+      "step": 4771
+    },
+    {
+      "epoch": 0.834448087431694,
+      "grad_norm": 1.068148778295192,
+      "learning_rate": 0.00016940046493468507,
+      "loss": 13.1549,
+      "step": 4772
+    },
+    {
+      "epoch": 0.8346229508196721,
+      "grad_norm": 1.058025953159505,
+      "learning_rate": 0.00016938687000767806,
+      "loss": 13.2243,
+      "step": 4773
+    },
+    {
+      "epoch": 0.8347978142076503,
+      "grad_norm": 1.149168013369045,
+      "learning_rate": 0.00016937327260711018,
+      "loss": 13.1226,
+      "step": 4774
+    },
+    {
+      "epoch": 0.8349726775956284,
+      "grad_norm": 0.9398682285662047,
+      "learning_rate": 0.0001693596727334662,
+      "loss": 13.1529,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8351475409836066,
+      "grad_norm": 0.9864763189047223,
+      "learning_rate": 0.00016934607038723094,
+      "loss": 13.205,
+      "step": 4776
+    },
+    {
+      "epoch": 0.8353224043715847,
+      "grad_norm": 1.0181891586164487,
+      "learning_rate": 0.00016933246556888925,
+      "loss": 13.2329,
+      "step": 4777
+    },
+    {
+      "epoch": 0.8354972677595628,
+      "grad_norm": 0.9961691893215022,
+      "learning_rate": 0.0001693188582789262,
+      "loss": 12.9232,
+      "step": 4778
+    },
+    {
+      "epoch": 0.835672131147541,
+      "grad_norm": 0.9691633178988182,
+      "learning_rate": 0.00016930524851782686,
+      "loss": 13.2404,
+      "step": 4779
+    },
+    {
+      "epoch": 0.8358469945355191,
+      "grad_norm": 1.0998945497287533,
+      "learning_rate": 0.00016929163628607636,
+      "loss": 12.9192,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8360218579234973,
+      "grad_norm": 0.9594888520618047,
+      "learning_rate": 0.00016927802158415997,
+      "loss": 13.1025,
+      "step": 4781
+    },
+    {
+      "epoch": 0.8361967213114754,
+      "grad_norm": 1.1325492505297197,
+      "learning_rate": 0.0001692644044125631,
+      "loss": 13.2435,
+      "step": 4782
+    },
+    {
+      "epoch": 0.8363715846994536,
+      "grad_norm": 1.0359668596154115,
+      "learning_rate": 0.00016925078477177108,
+      "loss": 12.942,
+      "step": 4783
+    },
+    {
+      "epoch": 0.8365464480874317,
+      "grad_norm": 1.216270607773214,
+      "learning_rate": 0.0001692371626622695,
+      "loss": 13.1835,
+      "step": 4784
+    },
+    {
+      "epoch": 0.8367213114754098,
+      "grad_norm": 1.0213873129850326,
+      "learning_rate": 0.00016922353808454402,
+      "loss": 13.0125,
+      "step": 4785
+    },
+    {
+      "epoch": 0.836896174863388,
+      "grad_norm": 1.1026252725871448,
+      "learning_rate": 0.00016920991103908025,
+      "loss": 13.1132,
+      "step": 4786
+    },
+    {
+      "epoch": 0.8370710382513661,
+      "grad_norm": 1.0527114951256653,
+      "learning_rate": 0.00016919628152636403,
+      "loss": 13.114,
+      "step": 4787
+    },
+    {
+      "epoch": 0.8372459016393443,
+      "grad_norm": 1.0263799036506065,
+      "learning_rate": 0.00016918264954688121,
+      "loss": 13.1857,
+      "step": 4788
+    },
+    {
+      "epoch": 0.8374207650273224,
+      "grad_norm": 1.0783308720849332,
+      "learning_rate": 0.00016916901510111775,
+      "loss": 13.2461,
+      "step": 4789
+    },
+    {
+      "epoch": 0.8375956284153006,
+      "grad_norm": 1.1596147683247529,
+      "learning_rate": 0.00016915537818955976,
+      "loss": 13.0421,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8377704918032787,
+      "grad_norm": 1.0614938756655832,
+      "learning_rate": 0.00016914173881269327,
+      "loss": 13.1272,
+      "step": 4791
+    },
+    {
+      "epoch": 0.8379453551912568,
+      "grad_norm": 1.343186273562059,
+      "learning_rate": 0.00016912809697100466,
+      "loss": 13.0854,
+      "step": 4792
+    },
+    {
+      "epoch": 0.838120218579235,
+      "grad_norm": 1.048396875907177,
+      "learning_rate": 0.00016911445266498006,
+      "loss": 13.186,
+      "step": 4793
+    },
+    {
+      "epoch": 0.8382950819672131,
+      "grad_norm": 1.1154605304095762,
+      "learning_rate": 0.00016910080589510606,
+      "loss": 13.1695,
+      "step": 4794
+    },
+    {
+      "epoch": 0.8384699453551913,
+      "grad_norm": 1.1144620717154656,
+      "learning_rate": 0.00016908715666186903,
+      "loss": 13.2206,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8386448087431694,
+      "grad_norm": 1.3503094243396856,
+      "learning_rate": 0.0001690735049657556,
+      "loss": 13.2591,
+      "step": 4796
+    },
+    {
+      "epoch": 0.8388196721311475,
+      "grad_norm": 1.06640471495948,
+      "learning_rate": 0.00016905985080725242,
+      "loss": 13.1534,
+      "step": 4797
+    },
+    {
+      "epoch": 0.8389945355191257,
+      "grad_norm": 1.0339923675712372,
+      "learning_rate": 0.00016904619418684624,
+      "loss": 13.2009,
+      "step": 4798
+    },
+    {
+      "epoch": 0.8391693989071038,
+      "grad_norm": 1.0477717521092498,
+      "learning_rate": 0.00016903253510502388,
+      "loss": 12.9952,
+      "step": 4799
+    },
+    {
+      "epoch": 0.839344262295082,
+      "grad_norm": 1.1284478467737393,
+      "learning_rate": 0.00016901887356227232,
+      "loss": 13.0351,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8395191256830601,
+      "grad_norm": 1.1873842762324716,
+      "learning_rate": 0.00016900520955907857,
+      "loss": 13.0763,
+      "step": 4801
+    },
+    {
+      "epoch": 0.8396939890710382,
+      "grad_norm": 1.0849600473079566,
+      "learning_rate": 0.00016899154309592973,
+      "loss": 13.1916,
+      "step": 4802
+    },
+    {
+      "epoch": 0.8398688524590164,
+      "grad_norm": 1.1021086318333497,
+      "learning_rate": 0.00016897787417331298,
+      "loss": 13.0852,
+      "step": 4803
+    },
+    {
+      "epoch": 0.8400437158469946,
+      "grad_norm": 1.0646875463494196,
+      "learning_rate": 0.0001689642027917156,
+      "loss": 13.1031,
+      "step": 4804
+    },
+    {
+      "epoch": 0.8402185792349727,
+      "grad_norm": 1.1262014643527851,
+      "learning_rate": 0.00016895052895162497,
+      "loss": 13.214,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8403934426229508,
+      "grad_norm": 1.115656135950651,
+      "learning_rate": 0.00016893685265352852,
+      "loss": 13.393,
+      "step": 4806
+    },
+    {
+      "epoch": 0.8405683060109289,
+      "grad_norm": 1.1840493185272392,
+      "learning_rate": 0.00016892317389791383,
+      "loss": 13.3853,
+      "step": 4807
+    },
+    {
+      "epoch": 0.8407431693989071,
+      "grad_norm": 1.2343806158644677,
+      "learning_rate": 0.00016890949268526855,
+      "loss": 13.146,
+      "step": 4808
+    },
+    {
+      "epoch": 0.8409180327868853,
+      "grad_norm": 0.9916461790537285,
+      "learning_rate": 0.00016889580901608033,
+      "loss": 13.2398,
+      "step": 4809
+    },
+    {
+      "epoch": 0.8410928961748634,
+      "grad_norm": 1.1530470393569383,
+      "learning_rate": 0.00016888212289083702,
+      "loss": 13.2642,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8412677595628415,
+      "grad_norm": 1.0150080554194392,
+      "learning_rate": 0.00016886843431002655,
+      "loss": 13.0395,
+      "step": 4811
+    },
+    {
+      "epoch": 0.8414426229508196,
+      "grad_norm": 1.0180751316090078,
+      "learning_rate": 0.0001688547432741368,
+      "loss": 13.2275,
+      "step": 4812
+    },
+    {
+      "epoch": 0.8416174863387978,
+      "grad_norm": 1.0559082670369528,
+      "learning_rate": 0.00016884104978365594,
+      "loss": 13.1317,
+      "step": 4813
+    },
+    {
+      "epoch": 0.841792349726776,
+      "grad_norm": 1.0852686825943518,
+      "learning_rate": 0.00016882735383907206,
+      "loss": 13.1213,
+      "step": 4814
+    },
+    {
+      "epoch": 0.8419672131147541,
+      "grad_norm": 1.0026438129938298,
+      "learning_rate": 0.00016881365544087348,
+      "loss": 13.226,
+      "step": 4815
+    },
+    {
+      "epoch": 0.8421420765027322,
+      "grad_norm": 0.9906286059839241,
+      "learning_rate": 0.00016879995458954846,
+      "loss": 13.2956,
+      "step": 4816
+    },
+    {
+      "epoch": 0.8423169398907104,
+      "grad_norm": 0.9937095317128998,
+      "learning_rate": 0.00016878625128558542,
+      "loss": 13.2553,
+      "step": 4817
+    },
+    {
+      "epoch": 0.8424918032786886,
+      "grad_norm": 1.0388274172199015,
+      "learning_rate": 0.00016877254552947292,
+      "loss": 13.3603,
+      "step": 4818
+    },
+    {
+      "epoch": 0.8426666666666667,
+      "grad_norm": 1.1329716934364644,
+      "learning_rate": 0.00016875883732169947,
+      "loss": 13.1776,
+      "step": 4819
+    },
+    {
+      "epoch": 0.8428415300546448,
+      "grad_norm": 1.070136908284888,
+      "learning_rate": 0.00016874512666275386,
+      "loss": 13.2641,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8430163934426229,
+      "grad_norm": 0.9734101552912694,
+      "learning_rate": 0.00016873141355312478,
+      "loss": 13.1521,
+      "step": 4821
+    },
+    {
+      "epoch": 0.8431912568306011,
+      "grad_norm": 1.1873185725521578,
+      "learning_rate": 0.00016871769799330113,
+      "loss": 13.2195,
+      "step": 4822
+    },
+    {
+      "epoch": 0.8433661202185793,
+      "grad_norm": 0.9420096301753423,
+      "learning_rate": 0.0001687039799837718,
+      "loss": 13.2389,
+      "step": 4823
+    },
+    {
+      "epoch": 0.8435409836065574,
+      "grad_norm": 1.0420441316770201,
+      "learning_rate": 0.00016869025952502589,
+      "loss": 13.1045,
+      "step": 4824
+    },
+    {
+      "epoch": 0.8437158469945355,
+      "grad_norm": 1.050075293674413,
+      "learning_rate": 0.00016867653661755246,
+      "loss": 13.0525,
+      "step": 4825
+    },
+    {
+      "epoch": 0.8438907103825136,
+      "grad_norm": 1.0518283157338428,
+      "learning_rate": 0.00016866281126184078,
+      "loss": 13.2082,
+      "step": 4826
+    },
+    {
+      "epoch": 0.8440655737704919,
+      "grad_norm": 1.043194753312052,
+      "learning_rate": 0.0001686490834583801,
+      "loss": 13.3298,
+      "step": 4827
+    },
+    {
+      "epoch": 0.84424043715847,
+      "grad_norm": 1.1958441897841543,
+      "learning_rate": 0.00016863535320765975,
+      "loss": 13.1917,
+      "step": 4828
+    },
+    {
+      "epoch": 0.8444153005464481,
+      "grad_norm": 1.0988463275501361,
+      "learning_rate": 0.0001686216205101693,
+      "loss": 13.1912,
+      "step": 4829
+    },
+    {
+      "epoch": 0.8445901639344262,
+      "grad_norm": 0.9569598484546458,
+      "learning_rate": 0.0001686078853663982,
+      "loss": 13.0371,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8447650273224043,
+      "grad_norm": 1.0935865471553605,
+      "learning_rate": 0.0001685941477768362,
+      "loss": 13.1936,
+      "step": 4831
+    },
+    {
+      "epoch": 0.8449398907103826,
+      "grad_norm": 1.0501270961855256,
+      "learning_rate": 0.00016858040774197293,
+      "loss": 12.6492,
+      "step": 4832
+    },
+    {
+      "epoch": 0.8451147540983607,
+      "grad_norm": 1.0790474659388116,
+      "learning_rate": 0.0001685666652622983,
+      "loss": 13.3146,
+      "step": 4833
+    },
+    {
+      "epoch": 0.8452896174863388,
+      "grad_norm": 1.0873464437657154,
+      "learning_rate": 0.00016855292033830213,
+      "loss": 13.1466,
+      "step": 4834
+    },
+    {
+      "epoch": 0.8454644808743169,
+      "grad_norm": 1.1874482051006168,
+      "learning_rate": 0.00016853917297047444,
+      "loss": 13.275,
+      "step": 4835
+    },
+    {
+      "epoch": 0.845639344262295,
+      "grad_norm": 1.0805984889990679,
+      "learning_rate": 0.00016852542315930533,
+      "loss": 13.2134,
+      "step": 4836
+    },
+    {
+      "epoch": 0.8458142076502733,
+      "grad_norm": 1.2218569875742402,
+      "learning_rate": 0.00016851167090528498,
+      "loss": 13.1139,
+      "step": 4837
+    },
+    {
+      "epoch": 0.8459890710382514,
+      "grad_norm": 1.0769562579906846,
+      "learning_rate": 0.00016849791620890354,
+      "loss": 13.1435,
+      "step": 4838
+    },
+    {
+      "epoch": 0.8461639344262295,
+      "grad_norm": 1.1696379947828737,
+      "learning_rate": 0.00016848415907065148,
+      "loss": 13.2921,
+      "step": 4839
+    },
+    {
+      "epoch": 0.8463387978142076,
+      "grad_norm": 1.0772271982111534,
+      "learning_rate": 0.00016847039949101911,
+      "loss": 13.0191,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8465136612021857,
+      "grad_norm": 1.214570075582511,
+      "learning_rate": 0.00016845663747049706,
+      "loss": 13.1014,
+      "step": 4841
+    },
+    {
+      "epoch": 0.846688524590164,
+      "grad_norm": 1.0593686044376815,
+      "learning_rate": 0.00016844287300957583,
+      "loss": 13.045,
+      "step": 4842
+    },
+    {
+      "epoch": 0.8468633879781421,
+      "grad_norm": 1.0127736960659057,
+      "learning_rate": 0.00016842910610874613,
+      "loss": 12.9817,
+      "step": 4843
+    },
+    {
+      "epoch": 0.8470382513661202,
+      "grad_norm": 1.5888178165798126,
+      "learning_rate": 0.00016841533676849877,
+      "loss": 13.1671,
+      "step": 4844
+    },
+    {
+      "epoch": 0.8472131147540983,
+      "grad_norm": 1.0448357024305255,
+      "learning_rate": 0.00016840156498932457,
+      "loss": 13.2997,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8473879781420764,
+      "grad_norm": 1.2782181143145495,
+      "learning_rate": 0.00016838779077171453,
+      "loss": 13.0885,
+      "step": 4846
+    },
+    {
+      "epoch": 0.8475628415300547,
+      "grad_norm": 1.0961251513083736,
+      "learning_rate": 0.00016837401411615963,
+      "loss": 13.3082,
+      "step": 4847
+    },
+    {
+      "epoch": 0.8477377049180328,
+      "grad_norm": 1.3172232332411096,
+      "learning_rate": 0.00016836023502315103,
+      "loss": 13.1726,
+      "step": 4848
+    },
+    {
+      "epoch": 0.8479125683060109,
+      "grad_norm": 1.1945240886827282,
+      "learning_rate": 0.00016834645349317987,
+      "loss": 13.3634,
+      "step": 4849
+    },
+    {
+      "epoch": 0.848087431693989,
+      "grad_norm": 1.1421884192358143,
+      "learning_rate": 0.00016833266952673756,
+      "loss": 13.3385,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8482622950819673,
+      "grad_norm": 1.082719486838395,
+      "learning_rate": 0.00016831888312431539,
+      "loss": 13.2179,
+      "step": 4851
+    },
+    {
+      "epoch": 0.8484371584699454,
+      "grad_norm": 1.1492382501074587,
+      "learning_rate": 0.00016830509428640488,
+      "loss": 13.1614,
+      "step": 4852
+    },
+    {
+      "epoch": 0.8486120218579235,
+      "grad_norm": 1.0118265729053904,
+      "learning_rate": 0.00016829130301349753,
+      "loss": 12.8845,
+      "step": 4853
+    },
+    {
+      "epoch": 0.8487868852459016,
+      "grad_norm": 1.024474410775125,
+      "learning_rate": 0.00016827750930608503,
+      "loss": 13.3286,
+      "step": 4854
+    },
+    {
+      "epoch": 0.8489617486338797,
+      "grad_norm": 1.1696649405934012,
+      "learning_rate": 0.00016826371316465912,
+      "loss": 13.2236,
+      "step": 4855
+    },
+    {
+      "epoch": 0.849136612021858,
+      "grad_norm": 1.1659988315163161,
+      "learning_rate": 0.0001682499145897116,
+      "loss": 13.3773,
+      "step": 4856
+    },
+    {
+      "epoch": 0.8493114754098361,
+      "grad_norm": 1.039455788501163,
+      "learning_rate": 0.0001682361135817343,
+      "loss": 13.2456,
+      "step": 4857
+    },
+    {
+      "epoch": 0.8494863387978142,
+      "grad_norm": 1.1320031967434312,
+      "learning_rate": 0.0001682223101412193,
+      "loss": 13.0937,
+      "step": 4858
+    },
+    {
+      "epoch": 0.8496612021857923,
+      "grad_norm": 1.03400502796243,
+      "learning_rate": 0.00016820850426865872,
+      "loss": 13.1921,
+      "step": 4859
+    },
+    {
+      "epoch": 0.8498360655737704,
+      "grad_norm": 1.233616281193629,
+      "learning_rate": 0.00016819469596454458,
+      "loss": 13.0777,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8500109289617487,
+      "grad_norm": 1.1158064658536937,
+      "learning_rate": 0.00016818088522936922,
+      "loss": 13.125,
+      "step": 4861
+    },
+    {
+      "epoch": 0.8501857923497268,
+      "grad_norm": 1.3546489794139422,
+      "learning_rate": 0.000168167072063625,
+      "loss": 13.268,
+      "step": 4862
+    },
+    {
+      "epoch": 0.8503606557377049,
+      "grad_norm": 1.1316033768875373,
+      "learning_rate": 0.00016815325646780428,
+      "loss": 13.2527,
+      "step": 4863
+    },
+    {
+      "epoch": 0.850535519125683,
+      "grad_norm": 1.0944750346263052,
+      "learning_rate": 0.00016813943844239958,
+      "loss": 13.2508,
+      "step": 4864
+    },
+    {
+      "epoch": 0.8507103825136612,
+      "grad_norm": 1.0772131658318589,
+      "learning_rate": 0.00016812561798790353,
+      "loss": 13.1172,
+      "step": 4865
+    },
+    {
+      "epoch": 0.8508852459016394,
+      "grad_norm": 1.0896382352951153,
+      "learning_rate": 0.0001681117951048088,
+      "loss": 13.1145,
+      "step": 4866
+    },
+    {
+      "epoch": 0.8510601092896175,
+      "grad_norm": 1.0801807109485815,
+      "learning_rate": 0.00016809796979360816,
+      "loss": 13.0329,
+      "step": 4867
+    },
+    {
+      "epoch": 0.8512349726775956,
+      "grad_norm": 1.0488510382378902,
+      "learning_rate": 0.00016808414205479443,
+      "loss": 13.3442,
+      "step": 4868
+    },
+    {
+      "epoch": 0.8514098360655737,
+      "grad_norm": 1.115818115082578,
+      "learning_rate": 0.00016807031188886059,
+      "loss": 13.2367,
+      "step": 4869
+    },
+    {
+      "epoch": 0.8515846994535519,
+      "grad_norm": 1.1112923383710984,
+      "learning_rate": 0.00016805647929629968,
+      "loss": 12.9415,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8517595628415301,
+      "grad_norm": 1.0590046676950067,
+      "learning_rate": 0.00016804264427760482,
+      "loss": 13.3431,
+      "step": 4871
+    },
+    {
+      "epoch": 0.8519344262295082,
+      "grad_norm": 0.9954135618488615,
+      "learning_rate": 0.00016802880683326917,
+      "loss": 13.2394,
+      "step": 4872
+    },
+    {
+      "epoch": 0.8521092896174863,
+      "grad_norm": 1.0971033325820008,
+      "learning_rate": 0.000168014966963786,
+      "loss": 13.2546,
+      "step": 4873
+    },
+    {
+      "epoch": 0.8522841530054645,
+      "grad_norm": 1.1646866408849095,
+      "learning_rate": 0.00016800112466964877,
+      "loss": 13.1397,
+      "step": 4874
+    },
+    {
+      "epoch": 0.8524590163934426,
+      "grad_norm": 1.1074224810058881,
+      "learning_rate": 0.0001679872799513509,
+      "loss": 13.2596,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8526338797814208,
+      "grad_norm": 1.016204791629475,
+      "learning_rate": 0.0001679734328093859,
+      "loss": 13.1608,
+      "step": 4876
+    },
+    {
+      "epoch": 0.8528087431693989,
+      "grad_norm": 1.2401802369693007,
+      "learning_rate": 0.00016795958324424746,
+      "loss": 13.4432,
+      "step": 4877
+    },
+    {
+      "epoch": 0.852983606557377,
+      "grad_norm": 1.1329779396169966,
+      "learning_rate": 0.0001679457312564293,
+      "loss": 12.9907,
+      "step": 4878
+    },
+    {
+      "epoch": 0.8531584699453552,
+      "grad_norm": 1.1413103230182977,
+      "learning_rate": 0.00016793187684642518,
+      "loss": 13.2615,
+      "step": 4879
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 1.372601642549337,
+      "learning_rate": 0.000167918020014729,
+      "loss": 13.2511,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8535081967213115,
+      "grad_norm": 1.242123909608916,
+      "learning_rate": 0.0001679041607618348,
+      "loss": 13.2245,
+      "step": 4881
+    },
+    {
+      "epoch": 0.8536830601092896,
+      "grad_norm": 1.013370409304611,
+      "learning_rate": 0.0001678902990882366,
+      "loss": 13.0279,
+      "step": 4882
+    },
+    {
+      "epoch": 0.8538579234972677,
+      "grad_norm": 1.3117380372625265,
+      "learning_rate": 0.00016787643499442854,
+      "loss": 13.2936,
+      "step": 4883
+    },
+    {
+      "epoch": 0.8540327868852459,
+      "grad_norm": 0.958446619901357,
+      "learning_rate": 0.0001678625684809049,
+      "loss": 13.1382,
+      "step": 4884
+    },
+    {
+      "epoch": 0.8542076502732241,
+      "grad_norm": 1.20231503653157,
+      "learning_rate": 0.00016784869954815994,
+      "loss": 13.3027,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8543825136612022,
+      "grad_norm": 1.1999847047316352,
+      "learning_rate": 0.00016783482819668816,
+      "loss": 13.2001,
+      "step": 4886
+    },
+    {
+      "epoch": 0.8545573770491803,
+      "grad_norm": 1.0511863298329822,
+      "learning_rate": 0.00016782095442698398,
+      "loss": 13.3084,
+      "step": 4887
+    },
+    {
+      "epoch": 0.8547322404371585,
+      "grad_norm": 1.0209681299363673,
+      "learning_rate": 0.00016780707823954206,
+      "loss": 13.2713,
+      "step": 4888
+    },
+    {
+      "epoch": 0.8549071038251366,
+      "grad_norm": 1.0214814853035792,
+      "learning_rate": 0.00016779319963485697,
+      "loss": 12.8946,
+      "step": 4889
+    },
+    {
+      "epoch": 0.8550819672131148,
+      "grad_norm": 1.086613975969627,
+      "learning_rate": 0.00016777931861342355,
+      "loss": 13.1541,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8552568306010929,
+      "grad_norm": 1.0611659678380003,
+      "learning_rate": 0.0001677654351757366,
+      "loss": 13.1536,
+      "step": 4891
+    },
+    {
+      "epoch": 0.855431693989071,
+      "grad_norm": 0.9731764744951953,
+      "learning_rate": 0.00016775154932229108,
+      "loss": 13.1364,
+      "step": 4892
+    },
+    {
+      "epoch": 0.8556065573770492,
+      "grad_norm": 1.098312935923493,
+      "learning_rate": 0.00016773766105358196,
+      "loss": 13.1982,
+      "step": 4893
+    },
+    {
+      "epoch": 0.8557814207650273,
+      "grad_norm": 1.050433583787284,
+      "learning_rate": 0.00016772377037010437,
+      "loss": 13.4044,
+      "step": 4894
+    },
+    {
+      "epoch": 0.8559562841530055,
+      "grad_norm": 1.0348876101760243,
+      "learning_rate": 0.00016770987727235352,
+      "loss": 13.1216,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8561311475409836,
+      "grad_norm": 1.0229356868344677,
+      "learning_rate": 0.00016769598176082464,
+      "loss": 13.2532,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8563060109289617,
+      "grad_norm": 1.1540465998828568,
+      "learning_rate": 0.0001676820838360131,
+      "loss": 13.0903,
+      "step": 4897
+    },
+    {
+      "epoch": 0.8564808743169399,
+      "grad_norm": 1.0006460051391728,
+      "learning_rate": 0.00016766818349841435,
+      "loss": 12.9993,
+      "step": 4898
+    },
+    {
+      "epoch": 0.856655737704918,
+      "grad_norm": 1.033246510659785,
+      "learning_rate": 0.00016765428074852395,
+      "loss": 13.3881,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8568306010928962,
+      "grad_norm": 1.0996460926621991,
+      "learning_rate": 0.00016764037558683742,
+      "loss": 12.9671,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8570054644808743,
+      "grad_norm": 1.014883832954209,
+      "learning_rate": 0.0001676264680138506,
+      "loss": 13.227,
+      "step": 4901
+    },
+    {
+      "epoch": 0.8571803278688525,
+      "grad_norm": 1.0592691980005593,
+      "learning_rate": 0.00016761255803005918,
+      "loss": 12.9378,
+      "step": 4902
+    },
+    {
+      "epoch": 0.8573551912568306,
+      "grad_norm": 0.9706707978547193,
+      "learning_rate": 0.00016759864563595906,
+      "loss": 12.8094,
+      "step": 4903
+    },
+    {
+      "epoch": 0.8575300546448087,
+      "grad_norm": 1.059923686148713,
+      "learning_rate": 0.0001675847308320462,
+      "loss": 13.1794,
+      "step": 4904
+    },
+    {
+      "epoch": 0.8577049180327869,
+      "grad_norm": 1.0738915720593951,
+      "learning_rate": 0.00016757081361881666,
+      "loss": 13.2483,
+      "step": 4905
+    },
+    {
+      "epoch": 0.857879781420765,
+      "grad_norm": 0.9901434374743401,
+      "learning_rate": 0.00016755689399676658,
+      "loss": 13.2259,
+      "step": 4906
+    },
+    {
+      "epoch": 0.8580546448087432,
+      "grad_norm": 1.095759758491536,
+      "learning_rate": 0.00016754297196639213,
+      "loss": 13.3985,
+      "step": 4907
+    },
+    {
+      "epoch": 0.8582295081967213,
+      "grad_norm": 1.198721234373483,
+      "learning_rate": 0.00016752904752818966,
+      "loss": 13.1902,
+      "step": 4908
+    },
+    {
+      "epoch": 0.8584043715846994,
+      "grad_norm": 1.0750549619150895,
+      "learning_rate": 0.00016751512068265552,
+      "loss": 12.9592,
+      "step": 4909
+    },
+    {
+      "epoch": 0.8585792349726776,
+      "grad_norm": 1.2854730517456827,
+      "learning_rate": 0.00016750119143028625,
+      "loss": 13.1854,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8587540983606557,
+      "grad_norm": 1.0460421889130573,
+      "learning_rate": 0.00016748725977157834,
+      "loss": 13.1135,
+      "step": 4911
+    },
+    {
+      "epoch": 0.8589289617486339,
+      "grad_norm": 1.0753229956323231,
+      "learning_rate": 0.00016747332570702847,
+      "loss": 13.1749,
+      "step": 4912
+    },
+    {
+      "epoch": 0.859103825136612,
+      "grad_norm": 1.0597186309625286,
+      "learning_rate": 0.00016745938923713337,
+      "loss": 13.2863,
+      "step": 4913
+    },
+    {
+      "epoch": 0.8592786885245902,
+      "grad_norm": 1.0241833300593426,
+      "learning_rate": 0.0001674454503623899,
+      "loss": 12.9243,
+      "step": 4914
+    },
+    {
+      "epoch": 0.8594535519125683,
+      "grad_norm": 1.0551664214112453,
+      "learning_rate": 0.0001674315090832949,
+      "loss": 12.9244,
+      "step": 4915
+    },
+    {
+      "epoch": 0.8596284153005465,
+      "grad_norm": 1.0545212277701494,
+      "learning_rate": 0.00016741756540034538,
+      "loss": 13.0046,
+      "step": 4916
+    },
+    {
+      "epoch": 0.8598032786885246,
+      "grad_norm": 1.0665719007423362,
+      "learning_rate": 0.00016740361931403842,
+      "loss": 13.0952,
+      "step": 4917
+    },
+    {
+      "epoch": 0.8599781420765027,
+      "grad_norm": 1.030646891687767,
+      "learning_rate": 0.00016738967082487123,
+      "loss": 12.871,
+      "step": 4918
+    },
+    {
+      "epoch": 0.8601530054644809,
+      "grad_norm": 0.9977101618744593,
+      "learning_rate": 0.00016737571993334096,
+      "loss": 13.1027,
+      "step": 4919
+    },
+    {
+      "epoch": 0.860327868852459,
+      "grad_norm": 1.1425160754943013,
+      "learning_rate": 0.000167361766639945,
+      "loss": 13.0445,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8605027322404372,
+      "grad_norm": 1.027736691805298,
+      "learning_rate": 0.0001673478109451808,
+      "loss": 12.9519,
+      "step": 4921
+    },
+    {
+      "epoch": 0.8606775956284153,
+      "grad_norm": 1.0038834063394064,
+      "learning_rate": 0.0001673338528495458,
+      "loss": 13.0172,
+      "step": 4922
+    },
+    {
+      "epoch": 0.8608524590163934,
+      "grad_norm": 1.2447534727678171,
+      "learning_rate": 0.0001673198923535376,
+      "loss": 13.2659,
+      "step": 4923
+    },
+    {
+      "epoch": 0.8610273224043716,
+      "grad_norm": 0.9867396087240805,
+      "learning_rate": 0.0001673059294576539,
+      "loss": 13.0591,
+      "step": 4924
+    },
+    {
+      "epoch": 0.8612021857923498,
+      "grad_norm": 0.9840744619359707,
+      "learning_rate": 0.0001672919641623925,
+      "loss": 12.9561,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8613770491803279,
+      "grad_norm": 1.220336551241528,
+      "learning_rate": 0.00016727799646825117,
+      "loss": 13.0826,
+      "step": 4926
+    },
+    {
+      "epoch": 0.861551912568306,
+      "grad_norm": 1.026144621959909,
+      "learning_rate": 0.00016726402637572783,
+      "loss": 13.1731,
+      "step": 4927
+    },
+    {
+      "epoch": 0.8617267759562841,
+      "grad_norm": 1.0033550587274311,
+      "learning_rate": 0.00016725005388532058,
+      "loss": 13.082,
+      "step": 4928
+    },
+    {
+      "epoch": 0.8619016393442623,
+      "grad_norm": 1.0279851799635433,
+      "learning_rate": 0.00016723607899752748,
+      "loss": 13.0645,
+      "step": 4929
+    },
+    {
+      "epoch": 0.8620765027322405,
+      "grad_norm": 1.004465232950486,
+      "learning_rate": 0.00016722210171284671,
+      "loss": 13.076,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8622513661202186,
+      "grad_norm": 1.0179204273963018,
+      "learning_rate": 0.00016720812203177657,
+      "loss": 12.9251,
+      "step": 4931
+    },
+    {
+      "epoch": 0.8624262295081967,
+      "grad_norm": 1.1581352592313168,
+      "learning_rate": 0.00016719413995481538,
+      "loss": 13.272,
+      "step": 4932
+    },
+    {
+      "epoch": 0.8626010928961748,
+      "grad_norm": 1.054040843485506,
+      "learning_rate": 0.00016718015548246163,
+      "loss": 13.1605,
+      "step": 4933
+    },
+    {
+      "epoch": 0.862775956284153,
+      "grad_norm": 0.9297707783481513,
+      "learning_rate": 0.0001671661686152138,
+      "loss": 13.0961,
+      "step": 4934
+    },
+    {
+      "epoch": 0.8629508196721312,
+      "grad_norm": 1.0305717772572935,
+      "learning_rate": 0.0001671521793535706,
+      "loss": 13.0282,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8631256830601093,
+      "grad_norm": 0.9969372817602833,
+      "learning_rate": 0.0001671381876980306,
+      "loss": 13.2128,
+      "step": 4936
+    },
+    {
+      "epoch": 0.8633005464480874,
+      "grad_norm": 1.0500309901141263,
+      "learning_rate": 0.0001671241936490927,
+      "loss": 13.3334,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8634754098360655,
+      "grad_norm": 1.0835596306277664,
+      "learning_rate": 0.0001671101972072557,
+      "loss": 13.0425,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8636502732240438,
+      "grad_norm": 1.1837905471358097,
+      "learning_rate": 0.00016709619837301855,
+      "loss": 13.121,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8638251366120219,
+      "grad_norm": 1.0225757015714316,
+      "learning_rate": 0.00016708219714688035,
+      "loss": 13.2626,
+      "step": 4940
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 1.0255574880862257,
+      "learning_rate": 0.0001670681935293402,
+      "loss": 13.2903,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8641748633879781,
+      "grad_norm": 1.0044601426963156,
+      "learning_rate": 0.0001670541875208973,
+      "loss": 13.0992,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8643497267759563,
+      "grad_norm": 1.029373246283069,
+      "learning_rate": 0.00016704017912205096,
+      "loss": 13.1348,
+      "step": 4943
+    },
+    {
+      "epoch": 0.8645245901639345,
+      "grad_norm": 1.0160345309472745,
+      "learning_rate": 0.00016702616833330058,
+      "loss": 13.1247,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8646994535519126,
+      "grad_norm": 1.0027164447048138,
+      "learning_rate": 0.0001670121551551456,
+      "loss": 12.9268,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8648743169398907,
+      "grad_norm": 0.9842168924059305,
+      "learning_rate": 0.0001669981395880856,
+      "loss": 13.2424,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8650491803278688,
+      "grad_norm": 0.9272596576083145,
+      "learning_rate": 0.00016698412163262017,
+      "loss": 13.0258,
+      "step": 4947
+    },
+    {
+      "epoch": 0.865224043715847,
+      "grad_norm": 1.0617058783509998,
+      "learning_rate": 0.0001669701012892491,
+      "loss": 13.1565,
+      "step": 4948
+    },
+    {
+      "epoch": 0.8653989071038252,
+      "grad_norm": 1.0664449800184719,
+      "learning_rate": 0.00016695607855847212,
+      "loss": 13.2642,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8655737704918033,
+      "grad_norm": 0.9935428794049602,
+      "learning_rate": 0.0001669420534407892,
+      "loss": 13.1262,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8657486338797814,
+      "grad_norm": 1.1786563771356664,
+      "learning_rate": 0.00016692802593670027,
+      "loss": 13.1477,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8659234972677595,
+      "grad_norm": 0.9877426842322178,
+      "learning_rate": 0.00016691399604670543,
+      "loss": 12.9803,
+      "step": 4952
+    },
+    {
+      "epoch": 0.8660983606557378,
+      "grad_norm": 0.9270345324337631,
+      "learning_rate": 0.00016689996377130484,
+      "loss": 12.6505,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8662732240437159,
+      "grad_norm": 1.1698696853973434,
+      "learning_rate": 0.00016688592911099867,
+      "loss": 13.1964,
+      "step": 4954
+    },
+    {
+      "epoch": 0.866448087431694,
+      "grad_norm": 1.2482999740380707,
+      "learning_rate": 0.0001668718920662873,
+      "loss": 13.1888,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8666229508196721,
+      "grad_norm": 0.9739793394584232,
+      "learning_rate": 0.00016685785263767107,
+      "loss": 13.0475,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8667978142076502,
+      "grad_norm": 1.0308163586657828,
+      "learning_rate": 0.00016684381082565053,
+      "loss": 13.1115,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8669726775956285,
+      "grad_norm": 1.11907586958608,
+      "learning_rate": 0.00016682976663072622,
+      "loss": 13.2989,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8671475409836066,
+      "grad_norm": 1.0093943672657852,
+      "learning_rate": 0.00016681572005339883,
+      "loss": 13.0687,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8673224043715847,
+      "grad_norm": 1.0773418393968262,
+      "learning_rate": 0.00016680167109416906,
+      "loss": 13.273,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8674972677595628,
+      "grad_norm": 1.0574148821421272,
+      "learning_rate": 0.0001667876197535378,
+      "loss": 13.078,
+      "step": 4961
+    },
+    {
+      "epoch": 0.8676721311475409,
+      "grad_norm": 1.2801806537439624,
+      "learning_rate": 0.0001667735660320059,
+      "loss": 13.2134,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8678469945355192,
+      "grad_norm": 1.0091367632092252,
+      "learning_rate": 0.0001667595099300744,
+      "loss": 13.1548,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8680218579234973,
+      "grad_norm": 1.1926219085807128,
+      "learning_rate": 0.00016674545144824436,
+      "loss": 13.1789,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8681967213114754,
+      "grad_norm": 1.2116434137583605,
+      "learning_rate": 0.00016673139058701698,
+      "loss": 12.9888,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8683715846994535,
+      "grad_norm": 1.1831845775356373,
+      "learning_rate": 0.00016671732734689347,
+      "loss": 13.1238,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8685464480874316,
+      "grad_norm": 1.5291810684662241,
+      "learning_rate": 0.0001667032617283752,
+      "loss": 13.1794,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8687213114754099,
+      "grad_norm": 0.988706228262596,
+      "learning_rate": 0.0001666891937319636,
+      "loss": 12.9956,
+      "step": 4968
+    },
+    {
+      "epoch": 0.868896174863388,
+      "grad_norm": 1.1620202044269017,
+      "learning_rate": 0.00016667512335816014,
+      "loss": 12.9503,
+      "step": 4969
+    },
+    {
+      "epoch": 0.8690710382513661,
+      "grad_norm": 1.1678261655742292,
+      "learning_rate": 0.00016666105060746646,
+      "loss": 13.0745,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8692459016393442,
+      "grad_norm": 1.1477706910554264,
+      "learning_rate": 0.0001666469754803842,
+      "loss": 13.16,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8694207650273224,
+      "grad_norm": 0.9964360077117844,
+      "learning_rate": 0.00016663289797741514,
+      "loss": 13.1427,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8695956284153006,
+      "grad_norm": 1.1220059690673654,
+      "learning_rate": 0.00016661881809906107,
+      "loss": 12.9187,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8697704918032787,
+      "grad_norm": 1.3214151403348124,
+      "learning_rate": 0.00016660473584582404,
+      "loss": 13.2221,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8699453551912568,
+      "grad_norm": 1.194737390368143,
+      "learning_rate": 0.00016659065121820595,
+      "loss": 13.1952,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8701202185792349,
+      "grad_norm": 1.0362173735541393,
+      "learning_rate": 0.00016657656421670893,
+      "loss": 12.8399,
+      "step": 4976
+    },
+    {
+      "epoch": 0.8702950819672132,
+      "grad_norm": 1.2415175026091017,
+      "learning_rate": 0.0001665624748418352,
+      "loss": 13.0483,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8704699453551913,
+      "grad_norm": 1.106618974970487,
+      "learning_rate": 0.00016654838309408706,
+      "loss": 13.1992,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8706448087431694,
+      "grad_norm": 1.2451012423458434,
+      "learning_rate": 0.00016653428897396677,
+      "loss": 13.3911,
+      "step": 4979
+    },
+    {
+      "epoch": 0.8708196721311475,
+      "grad_norm": 1.0516923866268926,
+      "learning_rate": 0.0001665201924819768,
+      "loss": 13.1734,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8709945355191256,
+      "grad_norm": 1.077992835899992,
+      "learning_rate": 0.00016650609361861974,
+      "loss": 13.3124,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8711693989071039,
+      "grad_norm": 1.045135841045463,
+      "learning_rate": 0.00016649199238439813,
+      "loss": 13.0415,
+      "step": 4982
+    },
+    {
+      "epoch": 0.871344262295082,
+      "grad_norm": 1.0091929517991698,
+      "learning_rate": 0.00016647788877981466,
+      "loss": 13.0945,
+      "step": 4983
+    },
+    {
+      "epoch": 0.8715191256830601,
+      "grad_norm": 0.9164757767944148,
+      "learning_rate": 0.00016646378280537211,
+      "loss": 12.9847,
+      "step": 4984
+    },
+    {
+      "epoch": 0.8716939890710382,
+      "grad_norm": 1.1622626035188592,
+      "learning_rate": 0.0001664496744615734,
+      "loss": 13.5258,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8718688524590164,
+      "grad_norm": 0.9566584350717555,
+      "learning_rate": 0.0001664355637489214,
+      "loss": 12.9651,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8720437158469946,
+      "grad_norm": 1.2029822336656022,
+      "learning_rate": 0.0001664214506679192,
+      "loss": 13.4228,
+      "step": 4987
+    },
+    {
+      "epoch": 0.8722185792349727,
+      "grad_norm": 0.9846417007453064,
+      "learning_rate": 0.00016640733521906988,
+      "loss": 13.0483,
+      "step": 4988
+    },
+    {
+      "epoch": 0.8723934426229508,
+      "grad_norm": 1.248834998105806,
+      "learning_rate": 0.00016639321740287666,
+      "loss": 13.1132,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8725683060109289,
+      "grad_norm": 0.97418156479623,
+      "learning_rate": 0.00016637909721984277,
+      "loss": 13.1437,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8727431693989071,
+      "grad_norm": 1.190759487341783,
+      "learning_rate": 0.00016636497467047167,
+      "loss": 13.1385,
+      "step": 4991
+    },
+    {
+      "epoch": 0.8729180327868853,
+      "grad_norm": 1.1370357352695186,
+      "learning_rate": 0.00016635084975526673,
+      "loss": 13.0739,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8730928961748634,
+      "grad_norm": 1.168827250047583,
+      "learning_rate": 0.00016633672247473154,
+      "loss": 12.9478,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8732677595628415,
+      "grad_norm": 1.2628701534810747,
+      "learning_rate": 0.00016632259282936968,
+      "loss": 13.2009,
+      "step": 4994
+    },
+    {
+      "epoch": 0.8734426229508196,
+      "grad_norm": 1.1737478027295685,
+      "learning_rate": 0.0001663084608196849,
+      "loss": 13.4011,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8736174863387978,
+      "grad_norm": 1.1703110183501246,
+      "learning_rate": 0.00016629432644618092,
+      "loss": 13.1534,
+      "step": 4996
+    },
+    {
+      "epoch": 0.873792349726776,
+      "grad_norm": 1.0659337111456821,
+      "learning_rate": 0.0001662801897093617,
+      "loss": 13.2309,
+      "step": 4997
+    },
+    {
+      "epoch": 0.8739672131147541,
+      "grad_norm": 1.358137325953602,
+      "learning_rate": 0.00016626605060973115,
+      "loss": 13.0259,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8741420765027322,
+      "grad_norm": 1.0169184142131926,
+      "learning_rate": 0.00016625190914779328,
+      "loss": 13.0864,
+      "step": 4999
+    },
+    {
+      "epoch": 0.8743169398907104,
+      "grad_norm": 1.3261923611387298,
+      "learning_rate": 0.00016623776532405233,
+      "loss": 13.043,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8744918032786885,
+      "grad_norm": 1.057510019859187,
+      "learning_rate": 0.0001662236191390124,
+      "loss": 13.1061,
+      "step": 5001
+    },
+    {
+      "epoch": 0.8746666666666667,
+      "grad_norm": 1.1495900526839573,
+      "learning_rate": 0.00016620947059317782,
+      "loss": 13.1116,
+      "step": 5002
+    },
+    {
+      "epoch": 0.8748415300546448,
+      "grad_norm": 0.9628683328525784,
+      "learning_rate": 0.00016619531968705293,
+      "loss": 13.1505,
+      "step": 5003
+    },
+    {
+      "epoch": 0.8750163934426229,
+      "grad_norm": 1.0776713650725103,
+      "learning_rate": 0.0001661811664211423,
+      "loss": 13.0933,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8751912568306011,
+      "grad_norm": 0.9352922164151384,
+      "learning_rate": 0.00016616701079595037,
+      "loss": 13.1008,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8753661202185792,
+      "grad_norm": 1.0848732590238002,
+      "learning_rate": 0.00016615285281198182,
+      "loss": 12.9954,
+      "step": 5006
+    },
+    {
+      "epoch": 0.8755409836065574,
+      "grad_norm": 1.0037004319098022,
+      "learning_rate": 0.00016613869246974138,
+      "loss": 13.1139,
+      "step": 5007
+    },
+    {
+      "epoch": 0.8757158469945355,
+      "grad_norm": 1.0473649949846868,
+      "learning_rate": 0.0001661245297697338,
+      "loss": 13.0738,
+      "step": 5008
+    },
+    {
+      "epoch": 0.8758907103825136,
+      "grad_norm": 1.250177176062139,
+      "learning_rate": 0.00016611036471246403,
+      "loss": 13.1476,
+      "step": 5009
+    },
+    {
+      "epoch": 0.8760655737704918,
+      "grad_norm": 1.1527807924290916,
+      "learning_rate": 0.00016609619729843696,
+      "loss": 12.9705,
+      "step": 5010
+    },
+    {
+      "epoch": 0.87624043715847,
+      "grad_norm": 1.0021565275515623,
+      "learning_rate": 0.0001660820275281577,
+      "loss": 13.0303,
+      "step": 5011
+    },
+    {
+      "epoch": 0.8764153005464481,
+      "grad_norm": 1.073534119294341,
+      "learning_rate": 0.00016606785540213136,
+      "loss": 13.1925,
+      "step": 5012
+    },
+    {
+      "epoch": 0.8765901639344262,
+      "grad_norm": 1.0934821367498766,
+      "learning_rate": 0.00016605368092086317,
+      "loss": 13.1258,
+      "step": 5013
+    },
+    {
+      "epoch": 0.8767650273224044,
+      "grad_norm": 0.951763299287275,
+      "learning_rate": 0.00016603950408485845,
+      "loss": 13.0312,
+      "step": 5014
+    },
+    {
+      "epoch": 0.8769398907103825,
+      "grad_norm": 0.9950711567122683,
+      "learning_rate": 0.00016602532489462255,
+      "loss": 13.1938,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8771147540983607,
+      "grad_norm": 1.0451095659983833,
+      "learning_rate": 0.00016601114335066094,
+      "loss": 13.1341,
+      "step": 5016
+    },
+    {
+      "epoch": 0.8772896174863388,
+      "grad_norm": 1.1496127186364797,
+      "learning_rate": 0.00016599695945347925,
+      "loss": 13.1726,
+      "step": 5017
+    },
+    {
+      "epoch": 0.8774644808743169,
+      "grad_norm": 1.1821298673827625,
+      "learning_rate": 0.000165982773203583,
+      "loss": 13.0694,
+      "step": 5018
+    },
+    {
+      "epoch": 0.8776393442622951,
+      "grad_norm": 1.1805214157008104,
+      "learning_rate": 0.00016596858460147802,
+      "loss": 13.1101,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8778142076502732,
+      "grad_norm": 1.1210637299882193,
+      "learning_rate": 0.00016595439364767005,
+      "loss": 13.2409,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8779890710382514,
+      "grad_norm": 1.2714374365664756,
+      "learning_rate": 0.000165940200342665,
+      "loss": 13.2136,
+      "step": 5021
+    },
+    {
+      "epoch": 0.8781639344262295,
+      "grad_norm": 1.0595631262140364,
+      "learning_rate": 0.00016592600468696888,
+      "loss": 13.2275,
+      "step": 5022
+    },
+    {
+      "epoch": 0.8783387978142077,
+      "grad_norm": 1.0764842891342963,
+      "learning_rate": 0.0001659118066810877,
+      "loss": 13.2553,
+      "step": 5023
+    },
+    {
+      "epoch": 0.8785136612021858,
+      "grad_norm": 1.0201751836220951,
+      "learning_rate": 0.00016589760632552766,
+      "loss": 12.8847,
+      "step": 5024
+    },
+    {
+      "epoch": 0.8786885245901639,
+      "grad_norm": 1.071859358498654,
+      "learning_rate": 0.0001658834036207949,
+      "loss": 13.1701,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8788633879781421,
+      "grad_norm": 1.0726233286425788,
+      "learning_rate": 0.0001658691985673958,
+      "loss": 13.1025,
+      "step": 5026
+    },
+    {
+      "epoch": 0.8790382513661202,
+      "grad_norm": 1.067433393660233,
+      "learning_rate": 0.00016585499116583674,
+      "loss": 13.0596,
+      "step": 5027
+    },
+    {
+      "epoch": 0.8792131147540984,
+      "grad_norm": 1.049987162450565,
+      "learning_rate": 0.00016584078141662417,
+      "loss": 12.9978,
+      "step": 5028
+    },
+    {
+      "epoch": 0.8793879781420765,
+      "grad_norm": 1.016880481788774,
+      "learning_rate": 0.00016582656932026467,
+      "loss": 13.0308,
+      "step": 5029
+    },
+    {
+      "epoch": 0.8795628415300546,
+      "grad_norm": 1.1094962304625562,
+      "learning_rate": 0.0001658123548772649,
+      "loss": 13.3978,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8797377049180328,
+      "grad_norm": 0.9818586059430351,
+      "learning_rate": 0.00016579813808813155,
+      "loss": 13.236,
+      "step": 5031
+    },
+    {
+      "epoch": 0.879912568306011,
+      "grad_norm": 0.9651423938037951,
+      "learning_rate": 0.00016578391895337146,
+      "loss": 12.8547,
+      "step": 5032
+    },
+    {
+      "epoch": 0.8800874316939891,
+      "grad_norm": 1.2000562371971777,
+      "learning_rate": 0.00016576969747349152,
+      "loss": 13.1054,
+      "step": 5033
+    },
+    {
+      "epoch": 0.8802622950819672,
+      "grad_norm": 0.9721019781374212,
+      "learning_rate": 0.0001657554736489987,
+      "loss": 13.2771,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8804371584699453,
+      "grad_norm": 1.021714063967213,
+      "learning_rate": 0.0001657412474804001,
+      "loss": 13.123,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8806120218579235,
+      "grad_norm": 1.0211681995486912,
+      "learning_rate": 0.00016572701896820283,
+      "loss": 12.9038,
+      "step": 5036
+    },
+    {
+      "epoch": 0.8807868852459017,
+      "grad_norm": 1.0266930704022665,
+      "learning_rate": 0.0001657127881129141,
+      "loss": 13.1566,
+      "step": 5037
+    },
+    {
+      "epoch": 0.8809617486338798,
+      "grad_norm": 1.036314448239098,
+      "learning_rate": 0.00016569855491504128,
+      "loss": 13.2464,
+      "step": 5038
+    },
+    {
+      "epoch": 0.8811366120218579,
+      "grad_norm": 1.0546844310218562,
+      "learning_rate": 0.00016568431937509173,
+      "loss": 12.9352,
+      "step": 5039
+    },
+    {
+      "epoch": 0.8813114754098361,
+      "grad_norm": 0.9799356644062107,
+      "learning_rate": 0.00016567008149357293,
+      "loss": 13.0859,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8814863387978142,
+      "grad_norm": 1.000656865645498,
+      "learning_rate": 0.00016565584127099244,
+      "loss": 13.0222,
+      "step": 5041
+    },
+    {
+      "epoch": 0.8816612021857924,
+      "grad_norm": 1.0905609045437992,
+      "learning_rate": 0.00016564159870785796,
+      "loss": 13.2145,
+      "step": 5042
+    },
+    {
+      "epoch": 0.8818360655737705,
+      "grad_norm": 1.0724694212166215,
+      "learning_rate": 0.00016562735380467716,
+      "loss": 13.1011,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8820109289617486,
+      "grad_norm": 0.9742415017751124,
+      "learning_rate": 0.00016561310656195785,
+      "loss": 12.948,
+      "step": 5044
+    },
+    {
+      "epoch": 0.8821857923497268,
+      "grad_norm": 1.002950375847487,
+      "learning_rate": 0.000165598856980208,
+      "loss": 13.0422,
+      "step": 5045
+    },
+    {
+      "epoch": 0.882360655737705,
+      "grad_norm": 1.1012974183477067,
+      "learning_rate": 0.00016558460505993548,
+      "loss": 13.0944,
+      "step": 5046
+    },
+    {
+      "epoch": 0.8825355191256831,
+      "grad_norm": 1.1532392289574651,
+      "learning_rate": 0.0001655703508016485,
+      "loss": 12.9183,
+      "step": 5047
+    },
+    {
+      "epoch": 0.8827103825136612,
+      "grad_norm": 1.029805639883048,
+      "learning_rate": 0.00016555609420585506,
+      "loss": 13.1968,
+      "step": 5048
+    },
+    {
+      "epoch": 0.8828852459016393,
+      "grad_norm": 1.3925111012882598,
+      "learning_rate": 0.00016554183527306345,
+      "loss": 12.8049,
+      "step": 5049
+    },
+    {
+      "epoch": 0.8830601092896175,
+      "grad_norm": 1.106225218823942,
+      "learning_rate": 0.00016552757400378204,
+      "loss": 13.2679,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8832349726775957,
+      "grad_norm": 1.1970404918906452,
+      "learning_rate": 0.00016551331039851914,
+      "loss": 13.2714,
+      "step": 5051
+    },
+    {
+      "epoch": 0.8834098360655738,
+      "grad_norm": 1.1378852740242023,
+      "learning_rate": 0.0001654990444577833,
+      "loss": 13.3906,
+      "step": 5052
+    },
+    {
+      "epoch": 0.8835846994535519,
+      "grad_norm": 1.174252990997479,
+      "learning_rate": 0.00016548477618208302,
+      "loss": 13.018,
+      "step": 5053
+    },
+    {
+      "epoch": 0.88375956284153,
+      "grad_norm": 1.3259189497685406,
+      "learning_rate": 0.00016547050557192696,
+      "loss": 13.2295,
+      "step": 5054
+    },
+    {
+      "epoch": 0.8839344262295082,
+      "grad_norm": 1.026928632490159,
+      "learning_rate": 0.00016545623262782393,
+      "loss": 13.0395,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8841092896174864,
+      "grad_norm": 1.2489543711919737,
+      "learning_rate": 0.00016544195735028266,
+      "loss": 13.4029,
+      "step": 5056
+    },
+    {
+      "epoch": 0.8842841530054645,
+      "grad_norm": 1.0294493080998206,
+      "learning_rate": 0.00016542767973981204,
+      "loss": 13.1265,
+      "step": 5057
+    },
+    {
+      "epoch": 0.8844590163934426,
+      "grad_norm": 1.403426035190648,
+      "learning_rate": 0.00016541339979692112,
+      "loss": 13.0865,
+      "step": 5058
+    },
+    {
+      "epoch": 0.8846338797814207,
+      "grad_norm": 1.1053120701707366,
+      "learning_rate": 0.00016539911752211893,
+      "loss": 13.267,
+      "step": 5059
+    },
+    {
+      "epoch": 0.884808743169399,
+      "grad_norm": 1.0907618101462577,
+      "learning_rate": 0.00016538483291591457,
+      "loss": 12.906,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8849836065573771,
+      "grad_norm": 1.1580453431677213,
+      "learning_rate": 0.00016537054597881733,
+      "loss": 12.9053,
+      "step": 5061
+    },
+    {
+      "epoch": 0.8851584699453552,
+      "grad_norm": 0.9949538425872086,
+      "learning_rate": 0.0001653562567113365,
+      "loss": 13.329,
+      "step": 5062
+    },
+    {
+      "epoch": 0.8853333333333333,
+      "grad_norm": 1.1742132992435959,
+      "learning_rate": 0.0001653419651139815,
+      "loss": 13.2617,
+      "step": 5063
+    },
+    {
+      "epoch": 0.8855081967213114,
+      "grad_norm": 1.097920367974077,
+      "learning_rate": 0.00016532767118726177,
+      "loss": 13.1991,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8856830601092897,
+      "grad_norm": 1.0729814220530995,
+      "learning_rate": 0.0001653133749316869,
+      "loss": 13.2163,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8858579234972678,
+      "grad_norm": 0.9377078114150883,
+      "learning_rate": 0.0001652990763477665,
+      "loss": 13.1464,
+      "step": 5066
+    },
+    {
+      "epoch": 0.8860327868852459,
+      "grad_norm": 1.0106783386792777,
+      "learning_rate": 0.00016528477543601035,
+      "loss": 13.1624,
+      "step": 5067
+    },
+    {
+      "epoch": 0.886207650273224,
+      "grad_norm": 1.1808346466782569,
+      "learning_rate": 0.00016527047219692823,
+      "loss": 13.1583,
+      "step": 5068
+    },
+    {
+      "epoch": 0.8863825136612021,
+      "grad_norm": 0.9344566794807014,
+      "learning_rate": 0.00016525616663103004,
+      "loss": 13.0367,
+      "step": 5069
+    },
+    {
+      "epoch": 0.8865573770491804,
+      "grad_norm": 0.978250691157319,
+      "learning_rate": 0.00016524185873882573,
+      "loss": 13.227,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8867322404371585,
+      "grad_norm": 1.051332953377195,
+      "learning_rate": 0.0001652275485208254,
+      "loss": 12.9032,
+      "step": 5071
+    },
+    {
+      "epoch": 0.8869071038251366,
+      "grad_norm": 1.0783920299418954,
+      "learning_rate": 0.00016521323597753918,
+      "loss": 13.2321,
+      "step": 5072
+    },
+    {
+      "epoch": 0.8870819672131147,
+      "grad_norm": 1.1767307727528704,
+      "learning_rate": 0.00016519892110947727,
+      "loss": 13.3177,
+      "step": 5073
+    },
+    {
+      "epoch": 0.887256830601093,
+      "grad_norm": 1.107565662545992,
+      "learning_rate": 0.00016518460391715,
+      "loss": 12.8444,
+      "step": 5074
+    },
+    {
+      "epoch": 0.8874316939890711,
+      "grad_norm": 1.0085423244409866,
+      "learning_rate": 0.00016517028440106777,
+      "loss": 12.7571,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8876065573770492,
+      "grad_norm": 1.3093647592096374,
+      "learning_rate": 0.00016515596256174103,
+      "loss": 13.1949,
+      "step": 5076
+    },
+    {
+      "epoch": 0.8877814207650273,
+      "grad_norm": 0.9973490050502184,
+      "learning_rate": 0.0001651416383996804,
+      "loss": 13.216,
+      "step": 5077
+    },
+    {
+      "epoch": 0.8879562841530054,
+      "grad_norm": 1.2777389938708077,
+      "learning_rate": 0.00016512731191539644,
+      "loss": 13.2191,
+      "step": 5078
+    },
+    {
+      "epoch": 0.8881311475409837,
+      "grad_norm": 1.0086825241912099,
+      "learning_rate": 0.00016511298310939988,
+      "loss": 13.0623,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8883060109289618,
+      "grad_norm": 1.2901482685844818,
+      "learning_rate": 0.00016509865198220157,
+      "loss": 13.1494,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8884808743169399,
+      "grad_norm": 1.0360154072962413,
+      "learning_rate": 0.00016508431853431236,
+      "loss": 13.092,
+      "step": 5081
+    },
+    {
+      "epoch": 0.888655737704918,
+      "grad_norm": 1.6037488476488795,
+      "learning_rate": 0.00016506998276624328,
+      "loss": 13.1482,
+      "step": 5082
+    },
+    {
+      "epoch": 0.8888306010928961,
+      "grad_norm": 1.0689740163888923,
+      "learning_rate": 0.0001650556446785053,
+      "loss": 13.2119,
+      "step": 5083
+    },
+    {
+      "epoch": 0.8890054644808744,
+      "grad_norm": 1.5025125149972007,
+      "learning_rate": 0.0001650413042716096,
+      "loss": 12.9603,
+      "step": 5084
+    },
+    {
+      "epoch": 0.8891803278688525,
+      "grad_norm": 1.120877894013796,
+      "learning_rate": 0.00016502696154606743,
+      "loss": 13.2423,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8893551912568306,
+      "grad_norm": 1.4100015576892306,
+      "learning_rate": 0.00016501261650239002,
+      "loss": 13.3364,
+      "step": 5086
+    },
+    {
+      "epoch": 0.8895300546448087,
+      "grad_norm": 1.1884800726183626,
+      "learning_rate": 0.0001649982691410888,
+      "loss": 13.0923,
+      "step": 5087
+    },
+    {
+      "epoch": 0.8897049180327868,
+      "grad_norm": 1.307131208519879,
+      "learning_rate": 0.00016498391946267522,
+      "loss": 13.0487,
+      "step": 5088
+    },
+    {
+      "epoch": 0.8898797814207651,
+      "grad_norm": 1.228629661259222,
+      "learning_rate": 0.00016496956746766086,
+      "loss": 13.2383,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8900546448087432,
+      "grad_norm": 1.393481466724229,
+      "learning_rate": 0.00016495521315655731,
+      "loss": 13.3623,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8902295081967213,
+      "grad_norm": 1.220940426071632,
+      "learning_rate": 0.00016494085652987633,
+      "loss": 13.162,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8904043715846994,
+      "grad_norm": 1.4450666031846735,
+      "learning_rate": 0.00016492649758812966,
+      "loss": 13.2243,
+      "step": 5092
+    },
+    {
+      "epoch": 0.8905792349726775,
+      "grad_norm": 1.0838903160143873,
+      "learning_rate": 0.0001649121363318292,
+      "loss": 13.1727,
+      "step": 5093
+    },
+    {
+      "epoch": 0.8907540983606558,
+      "grad_norm": 1.5135229677025923,
+      "learning_rate": 0.00016489777276148694,
+      "loss": 13.2355,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8909289617486339,
+      "grad_norm": 1.181262384894784,
+      "learning_rate": 0.0001648834068776149,
+      "loss": 13.2197,
+      "step": 5095
+    },
+    {
+      "epoch": 0.891103825136612,
+      "grad_norm": 1.2271705352371585,
+      "learning_rate": 0.00016486903868072523,
+      "loss": 13.1166,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8912786885245901,
+      "grad_norm": 1.096622832115133,
+      "learning_rate": 0.00016485466817133012,
+      "loss": 13.2,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8914535519125683,
+      "grad_norm": 1.2915302004002156,
+      "learning_rate": 0.00016484029534994185,
+      "loss": 13.0971,
+      "step": 5098
+    },
+    {
+      "epoch": 0.8916284153005465,
+      "grad_norm": 1.2527829938708277,
+      "learning_rate": 0.0001648259202170728,
+      "loss": 13.2639,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8918032786885246,
+      "grad_norm": 1.124811737009959,
+      "learning_rate": 0.00016481154277323546,
+      "loss": 13.1473,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8919781420765027,
+      "grad_norm": 1.001904163435099,
+      "learning_rate": 0.0001647971630189423,
+      "loss": 13.2765,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8921530054644808,
+      "grad_norm": 1.215869786701635,
+      "learning_rate": 0.00016478278095470605,
+      "loss": 12.9558,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8923278688524591,
+      "grad_norm": 1.0872732706883066,
+      "learning_rate": 0.00016476839658103928,
+      "loss": 13.0683,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8925027322404372,
+      "grad_norm": 1.1459701761737155,
+      "learning_rate": 0.00016475400989845491,
+      "loss": 13.1403,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8926775956284153,
+      "grad_norm": 0.9609402195592083,
+      "learning_rate": 0.0001647396209074657,
+      "loss": 13.0223,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8928524590163934,
+      "grad_norm": 0.9592355262282445,
+      "learning_rate": 0.0001647252296085847,
+      "loss": 12.9404,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8930273224043715,
+      "grad_norm": 1.0248109117158126,
+      "learning_rate": 0.00016471083600232483,
+      "loss": 13.1917,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8932021857923498,
+      "grad_norm": 1.0641069683180382,
+      "learning_rate": 0.00016469644008919928,
+      "loss": 13.0388,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8933770491803279,
+      "grad_norm": 1.0037842408780082,
+      "learning_rate": 0.00016468204186972124,
+      "loss": 12.8864,
+      "step": 5109
+    },
+    {
+      "epoch": 0.893551912568306,
+      "grad_norm": 1.055150626902609,
+      "learning_rate": 0.00016466764134440398,
+      "loss": 13.0586,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8937267759562841,
+      "grad_norm": 1.0784898292046317,
+      "learning_rate": 0.00016465323851376084,
+      "loss": 13.0693,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8939016393442623,
+      "grad_norm": 0.9584501287895819,
+      "learning_rate": 0.0001646388333783053,
+      "loss": 13.0657,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8940765027322405,
+      "grad_norm": 1.0406912763942242,
+      "learning_rate": 0.00016462442593855088,
+      "loss": 13.3532,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8942513661202186,
+      "grad_norm": 1.0204636863078345,
+      "learning_rate": 0.00016461001619501116,
+      "loss": 13.1713,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8944262295081967,
+      "grad_norm": 0.994889569348482,
+      "learning_rate": 0.00016459560414819988,
+      "loss": 13.0849,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8946010928961748,
+      "grad_norm": 1.162179525672018,
+      "learning_rate": 0.00016458118979863077,
+      "loss": 13.4374,
+      "step": 5116
+    },
+    {
+      "epoch": 0.894775956284153,
+      "grad_norm": 0.953277831656197,
+      "learning_rate": 0.0001645667731468177,
+      "loss": 13.2037,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8949508196721312,
+      "grad_norm": 1.0861562858985265,
+      "learning_rate": 0.0001645523541932746,
+      "loss": 13.2499,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8951256830601093,
+      "grad_norm": 1.0478528212248095,
+      "learning_rate": 0.0001645379329385155,
+      "loss": 12.9128,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8953005464480874,
+      "grad_norm": 0.9193918651553563,
+      "learning_rate": 0.0001645235093830545,
+      "loss": 12.9444,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8954754098360656,
+      "grad_norm": 0.9693833373317083,
+      "learning_rate": 0.00016450908352740577,
+      "loss": 13.249,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8956502732240437,
+      "grad_norm": 1.1077464524070006,
+      "learning_rate": 0.0001644946553720836,
+      "loss": 13.3742,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8958251366120219,
+      "grad_norm": 0.9553255545416606,
+      "learning_rate": 0.00016448022491760228,
+      "loss": 13.0816,
+      "step": 5123
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.941335284772038,
+      "learning_rate": 0.00016446579216447633,
+      "loss": 13.0304,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8961748633879781,
+      "grad_norm": 0.9857391405372814,
+      "learning_rate": 0.00016445135711322015,
+      "loss": 12.8891,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8963497267759563,
+      "grad_norm": 0.998180377920576,
+      "learning_rate": 0.00016443691976434846,
+      "loss": 13.3789,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8965245901639344,
+      "grad_norm": 1.099522478676132,
+      "learning_rate": 0.00016442248011837583,
+      "loss": 13.184,
+      "step": 5127
+    },
+    {
+      "epoch": 0.8966994535519126,
+      "grad_norm": 0.9577057873839472,
+      "learning_rate": 0.00016440803817581706,
+      "loss": 13.076,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8968743169398907,
+      "grad_norm": 1.2313674012071836,
+      "learning_rate": 0.00016439359393718696,
+      "loss": 13.2258,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8970491803278688,
+      "grad_norm": 1.0236641173014656,
+      "learning_rate": 0.00016437914740300048,
+      "loss": 13.1061,
+      "step": 5130
+    },
+    {
+      "epoch": 0.897224043715847,
+      "grad_norm": 1.7091149938279573,
+      "learning_rate": 0.00016436469857377263,
+      "loss": 13.2096,
+      "step": 5131
+    },
+    {
+      "epoch": 0.8973989071038251,
+      "grad_norm": 1.2965883373339222,
+      "learning_rate": 0.00016435024745001847,
+      "loss": 12.8878,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8975737704918033,
+      "grad_norm": 1.3819810161102948,
+      "learning_rate": 0.0001643357940322532,
+      "loss": 13.1574,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8977486338797814,
+      "grad_norm": 1.4956417147544736,
+      "learning_rate": 0.000164321338320992,
+      "loss": 13.1633,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8979234972677596,
+      "grad_norm": 1.1644793129695572,
+      "learning_rate": 0.00016430688031675024,
+      "loss": 13.2097,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8980983606557377,
+      "grad_norm": 1.2610743519560774,
+      "learning_rate": 0.00016429242002004337,
+      "loss": 13.3295,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8982732240437159,
+      "grad_norm": 1.0200619950260472,
+      "learning_rate": 0.0001642779574313868,
+      "loss": 13.0348,
+      "step": 5137
+    },
+    {
+      "epoch": 0.898448087431694,
+      "grad_norm": 1.226365143120819,
+      "learning_rate": 0.00016426349255129619,
+      "loss": 13.1294,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8986229508196721,
+      "grad_norm": 0.9800582904045035,
+      "learning_rate": 0.00016424902538028714,
+      "loss": 13.1061,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8987978142076503,
+      "grad_norm": 1.253539498638682,
+      "learning_rate": 0.0001642345559188754,
+      "loss": 13.0276,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8989726775956284,
+      "grad_norm": 1.1157975334176036,
+      "learning_rate": 0.00016422008416757677,
+      "loss": 13.2012,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8991475409836066,
+      "grad_norm": 1.0266082466832995,
+      "learning_rate": 0.0001642056101269072,
+      "loss": 13.2871,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8993224043715847,
+      "grad_norm": 1.0299618564030604,
+      "learning_rate": 0.00016419113379738262,
+      "loss": 13.0189,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8994972677595628,
+      "grad_norm": 1.0480571710938205,
+      "learning_rate": 0.00016417665517951915,
+      "loss": 13.1829,
+      "step": 5144
+    },
+    {
+      "epoch": 0.899672131147541,
+      "grad_norm": 0.9301726770987088,
+      "learning_rate": 0.0001641621742738329,
+      "loss": 13.0625,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8998469945355191,
+      "grad_norm": 1.061264831282701,
+      "learning_rate": 0.00016414769108084008,
+      "loss": 13.2299,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9000218579234973,
+      "grad_norm": 1.0283664096687777,
+      "learning_rate": 0.00016413320560105702,
+      "loss": 13.0639,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9001967213114754,
+      "grad_norm": 1.1936977833957418,
+      "learning_rate": 0.00016411871783500014,
+      "loss": 12.9593,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9003715846994536,
+      "grad_norm": 0.9451314042438957,
+      "learning_rate": 0.00016410422778318585,
+      "loss": 13.0816,
+      "step": 5149
+    },
+    {
+      "epoch": 0.9005464480874317,
+      "grad_norm": 1.003967697361481,
+      "learning_rate": 0.00016408973544613077,
+      "loss": 13.1015,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9007213114754098,
+      "grad_norm": 1.2081968599862971,
+      "learning_rate": 0.0001640752408243515,
+      "loss": 13.0075,
+      "step": 5151
+    },
+    {
+      "epoch": 0.900896174863388,
+      "grad_norm": 1.2182153692646083,
+      "learning_rate": 0.00016406074391836473,
+      "loss": 13.0395,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9010710382513661,
+      "grad_norm": 1.213275878246996,
+      "learning_rate": 0.00016404624472868732,
+      "loss": 13.1965,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9012459016393443,
+      "grad_norm": 1.3967299701403255,
+      "learning_rate": 0.0001640317432558361,
+      "loss": 13.1683,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9014207650273224,
+      "grad_norm": 1.026196235859324,
+      "learning_rate": 0.00016401723950032807,
+      "loss": 13.0648,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9015956284153005,
+      "grad_norm": 1.188335613035461,
+      "learning_rate": 0.0001640027334626802,
+      "loss": 13.0563,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9017704918032787,
+      "grad_norm": 1.085661629756283,
+      "learning_rate": 0.00016398822514340967,
+      "loss": 12.9718,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9019453551912568,
+      "grad_norm": 1.189807720549792,
+      "learning_rate": 0.0001639737145430337,
+      "loss": 13.13,
+      "step": 5158
+    },
+    {
+      "epoch": 0.902120218579235,
+      "grad_norm": 1.1451576997112716,
+      "learning_rate": 0.00016395920166206954,
+      "loss": 13.2381,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9022950819672131,
+      "grad_norm": 1.0688274112459837,
+      "learning_rate": 0.00016394468650103458,
+      "loss": 12.9817,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9024699453551912,
+      "grad_norm": 1.0737921004934665,
+      "learning_rate": 0.00016393016906044624,
+      "loss": 12.995,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9026448087431694,
+      "grad_norm": 0.9445686469170373,
+      "learning_rate": 0.0001639156493408221,
+      "loss": 13.0683,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9028196721311476,
+      "grad_norm": 1.1207688135091856,
+      "learning_rate": 0.00016390112734267968,
+      "loss": 13.1641,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9029945355191257,
+      "grad_norm": 0.9746790067467143,
+      "learning_rate": 0.00016388660306653678,
+      "loss": 13.0761,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9031693989071038,
+      "grad_norm": 1.2776729198433767,
+      "learning_rate": 0.0001638720765129111,
+      "loss": 13.0734,
+      "step": 5165
+    },
+    {
+      "epoch": 0.9033442622950819,
+      "grad_norm": 1.0682351753913377,
+      "learning_rate": 0.00016385754768232053,
+      "loss": 12.7409,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9035191256830601,
+      "grad_norm": 1.0393334588189205,
+      "learning_rate": 0.00016384301657528297,
+      "loss": 13.0527,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9036939890710383,
+      "grad_norm": 1.1057126422956254,
+      "learning_rate": 0.0001638284831923165,
+      "loss": 13.0869,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9038688524590164,
+      "grad_norm": 1.0848948241178307,
+      "learning_rate": 0.00016381394753393918,
+      "loss": 12.9858,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9040437158469945,
+      "grad_norm": 1.0550480604598704,
+      "learning_rate": 0.00016379940960066917,
+      "loss": 13.0839,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9042185792349727,
+      "grad_norm": 1.0696791924762759,
+      "learning_rate": 0.00016378486939302476,
+      "loss": 13.2367,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9043934426229508,
+      "grad_norm": 1.0597943117095383,
+      "learning_rate": 0.00016377032691152429,
+      "loss": 13.0461,
+      "step": 5172
+    },
+    {
+      "epoch": 0.904568306010929,
+      "grad_norm": 1.0049162735468513,
+      "learning_rate": 0.00016375578215668615,
+      "loss": 13.1892,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9047431693989071,
+      "grad_norm": 1.0831769790232322,
+      "learning_rate": 0.00016374123512902887,
+      "loss": 12.7707,
+      "step": 5174
+    },
+    {
+      "epoch": 0.9049180327868852,
+      "grad_norm": 1.005623413037796,
+      "learning_rate": 0.00016372668582907105,
+      "loss": 13.0737,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9050928961748634,
+      "grad_norm": 0.9393636463522114,
+      "learning_rate": 0.0001637121342573313,
+      "loss": 13.2624,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9052677595628416,
+      "grad_norm": 0.9725686650852635,
+      "learning_rate": 0.00016369758041432844,
+      "loss": 13.1471,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9054426229508197,
+      "grad_norm": 1.094595622997911,
+      "learning_rate": 0.0001636830243005812,
+      "loss": 12.9775,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9056174863387978,
+      "grad_norm": 1.078672170094051,
+      "learning_rate": 0.00016366846591660863,
+      "loss": 12.9999,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9057923497267759,
+      "grad_norm": 0.9803607716046627,
+      "learning_rate": 0.0001636539052629296,
+      "loss": 13.1072,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9059672131147541,
+      "grad_norm": 1.0895909001580415,
+      "learning_rate": 0.00016363934234006324,
+      "loss": 13.1108,
+      "step": 5181
+    },
+    {
+      "epoch": 0.9061420765027323,
+      "grad_norm": 1.0999396894297129,
+      "learning_rate": 0.00016362477714852865,
+      "loss": 13.3532,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9063169398907104,
+      "grad_norm": 1.067942558632785,
+      "learning_rate": 0.00016361020968884513,
+      "loss": 12.8807,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9064918032786885,
+      "grad_norm": 1.2523773104082052,
+      "learning_rate": 0.00016359563996153192,
+      "loss": 13.2027,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "grad_norm": 1.08182993609442,
+      "learning_rate": 0.00016358106796710847,
+      "loss": 13.1587,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9068415300546449,
+      "grad_norm": 1.2967914071851943,
+      "learning_rate": 0.0001635664937060942,
+      "loss": 13.1198,
+      "step": 5186
+    },
+    {
+      "epoch": 0.907016393442623,
+      "grad_norm": 1.0938257170473755,
+      "learning_rate": 0.00016355191717900872,
+      "loss": 13.1153,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9071912568306011,
+      "grad_norm": 1.0822687599571001,
+      "learning_rate": 0.00016353733838637163,
+      "loss": 13.0902,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9073661202185792,
+      "grad_norm": 1.2500526713359503,
+      "learning_rate": 0.0001635227573287027,
+      "loss": 13.0362,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9075409836065573,
+      "grad_norm": 1.0798054202167864,
+      "learning_rate": 0.00016350817400652168,
+      "loss": 12.9051,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9077158469945356,
+      "grad_norm": 1.1094576314292002,
+      "learning_rate": 0.00016349358842034843,
+      "loss": 12.7644,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9078907103825137,
+      "grad_norm": 1.051265617166796,
+      "learning_rate": 0.00016347900057070295,
+      "loss": 13.2505,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9080655737704918,
+      "grad_norm": 1.1220994702096547,
+      "learning_rate": 0.00016346441045810527,
+      "loss": 13.1551,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9082404371584699,
+      "grad_norm": 0.9847592825479641,
+      "learning_rate": 0.0001634498180830755,
+      "loss": 13.167,
+      "step": 5194
+    },
+    {
+      "epoch": 0.908415300546448,
+      "grad_norm": 1.1911549031222726,
+      "learning_rate": 0.00016343522344613388,
+      "loss": 13.2989,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9085901639344263,
+      "grad_norm": 0.9613639345240738,
+      "learning_rate": 0.00016342062654780062,
+      "loss": 13.163,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9087650273224044,
+      "grad_norm": 1.093411987072267,
+      "learning_rate": 0.00016340602738859614,
+      "loss": 12.9232,
+      "step": 5197
+    },
+    {
+      "epoch": 0.9089398907103825,
+      "grad_norm": 1.0929023205016408,
+      "learning_rate": 0.00016339142596904086,
+      "loss": 13.0746,
+      "step": 5198
+    },
+    {
+      "epoch": 0.9091147540983606,
+      "grad_norm": 1.1598838531386815,
+      "learning_rate": 0.00016337682228965533,
+      "loss": 13.399,
+      "step": 5199
+    },
+    {
+      "epoch": 0.9092896174863389,
+      "grad_norm": 1.1863670384398028,
+      "learning_rate": 0.0001633622163509601,
+      "loss": 13.0216,
+      "step": 5200
+    },
+    {
+      "epoch": 0.909464480874317,
+      "grad_norm": 1.1434348431035068,
+      "learning_rate": 0.00016334760815347595,
+      "loss": 13.011,
+      "step": 5201
+    },
+    {
+      "epoch": 0.9096393442622951,
+      "grad_norm": 1.0461138370782173,
+      "learning_rate": 0.0001633329976977235,
+      "loss": 13.1832,
+      "step": 5202
+    },
+    {
+      "epoch": 0.9098142076502732,
+      "grad_norm": 1.036621945805809,
+      "learning_rate": 0.00016331838498422375,
+      "loss": 13.1365,
+      "step": 5203
+    },
+    {
+      "epoch": 0.9099890710382513,
+      "grad_norm": 0.9726775974413656,
+      "learning_rate": 0.00016330377001349747,
+      "loss": 13.2714,
+      "step": 5204
+    },
+    {
+      "epoch": 0.9101639344262296,
+      "grad_norm": 1.1420383213313947,
+      "learning_rate": 0.00016328915278606584,
+      "loss": 12.979,
+      "step": 5205
+    },
+    {
+      "epoch": 0.9103387978142077,
+      "grad_norm": 0.9541275004362733,
+      "learning_rate": 0.00016327453330244982,
+      "loss": 13.1176,
+      "step": 5206
+    },
+    {
+      "epoch": 0.9105136612021858,
+      "grad_norm": 1.0355634669780587,
+      "learning_rate": 0.00016325991156317059,
+      "loss": 13.084,
+      "step": 5207
+    },
+    {
+      "epoch": 0.9106885245901639,
+      "grad_norm": 1.0020795061014172,
+      "learning_rate": 0.00016324528756874948,
+      "loss": 13.3201,
+      "step": 5208
+    },
+    {
+      "epoch": 0.910863387978142,
+      "grad_norm": 1.1077769139727354,
+      "learning_rate": 0.00016323066131970776,
+      "loss": 13.1706,
+      "step": 5209
+    },
+    {
+      "epoch": 0.9110382513661203,
+      "grad_norm": 0.9507334371260586,
+      "learning_rate": 0.00016321603281656678,
+      "loss": 13.0721,
+      "step": 5210
+    },
+    {
+      "epoch": 0.9112131147540984,
+      "grad_norm": 1.0475082401438784,
+      "learning_rate": 0.00016320140205984814,
+      "loss": 12.9265,
+      "step": 5211
+    },
+    {
+      "epoch": 0.9113879781420765,
+      "grad_norm": 1.042539849444481,
+      "learning_rate": 0.00016318676905007333,
+      "loss": 13.1194,
+      "step": 5212
+    },
+    {
+      "epoch": 0.9115628415300546,
+      "grad_norm": 1.0134933101747523,
+      "learning_rate": 0.00016317213378776405,
+      "loss": 13.0359,
+      "step": 5213
+    },
+    {
+      "epoch": 0.9117377049180327,
+      "grad_norm": 0.9325546997865468,
+      "learning_rate": 0.000163157496273442,
+      "loss": 13.1404,
+      "step": 5214
+    },
+    {
+      "epoch": 0.911912568306011,
+      "grad_norm": 0.9869040597093476,
+      "learning_rate": 0.000163142856507629,
+      "loss": 13.0762,
+      "step": 5215
+    },
+    {
+      "epoch": 0.9120874316939891,
+      "grad_norm": 1.013820026210645,
+      "learning_rate": 0.00016312821449084695,
+      "loss": 13.0755,
+      "step": 5216
+    },
+    {
+      "epoch": 0.9122622950819672,
+      "grad_norm": 0.96292071736087,
+      "learning_rate": 0.0001631135702236178,
+      "loss": 12.9245,
+      "step": 5217
+    },
+    {
+      "epoch": 0.9124371584699453,
+      "grad_norm": 1.1275384996745026,
+      "learning_rate": 0.00016309892370646362,
+      "loss": 13.0555,
+      "step": 5218
+    },
+    {
+      "epoch": 0.9126120218579234,
+      "grad_norm": 1.0735861363467107,
+      "learning_rate": 0.00016308427493990653,
+      "loss": 13.2778,
+      "step": 5219
+    },
+    {
+      "epoch": 0.9127868852459017,
+      "grad_norm": 0.993658295292285,
+      "learning_rate": 0.00016306962392446874,
+      "loss": 12.9395,
+      "step": 5220
+    },
+    {
+      "epoch": 0.9129617486338798,
+      "grad_norm": 1.082070946833284,
+      "learning_rate": 0.00016305497066067254,
+      "loss": 13.0637,
+      "step": 5221
+    },
+    {
+      "epoch": 0.9131366120218579,
+      "grad_norm": 1.175678078280184,
+      "learning_rate": 0.0001630403151490403,
+      "loss": 13.0275,
+      "step": 5222
+    },
+    {
+      "epoch": 0.913311475409836,
+      "grad_norm": 1.0672769576612011,
+      "learning_rate": 0.00016302565739009453,
+      "loss": 13.1717,
+      "step": 5223
+    },
+    {
+      "epoch": 0.9134863387978142,
+      "grad_norm": 1.299693958520185,
+      "learning_rate": 0.00016301099738435767,
+      "loss": 12.9121,
+      "step": 5224
+    },
+    {
+      "epoch": 0.9136612021857924,
+      "grad_norm": 1.1660306575682022,
+      "learning_rate": 0.00016299633513235237,
+      "loss": 13.0671,
+      "step": 5225
+    },
+    {
+      "epoch": 0.9138360655737705,
+      "grad_norm": 0.9924512878094355,
+      "learning_rate": 0.00016298167063460135,
+      "loss": 13.1255,
+      "step": 5226
+    },
+    {
+      "epoch": 0.9140109289617486,
+      "grad_norm": 1.0203421158575814,
+      "learning_rate": 0.00016296700389162734,
+      "loss": 13.0974,
+      "step": 5227
+    },
+    {
+      "epoch": 0.9141857923497267,
+      "grad_norm": 1.0345091824791073,
+      "learning_rate": 0.00016295233490395324,
+      "loss": 12.9662,
+      "step": 5228
+    },
+    {
+      "epoch": 0.9143606557377049,
+      "grad_norm": 1.2597142448078866,
+      "learning_rate": 0.0001629376636721019,
+      "loss": 12.9967,
+      "step": 5229
+    },
+    {
+      "epoch": 0.9145355191256831,
+      "grad_norm": 1.0518976939776268,
+      "learning_rate": 0.00016292299019659643,
+      "loss": 13.0607,
+      "step": 5230
+    },
+    {
+      "epoch": 0.9147103825136612,
+      "grad_norm": 1.2305300352614879,
+      "learning_rate": 0.00016290831447795983,
+      "loss": 13.2327,
+      "step": 5231
+    },
+    {
+      "epoch": 0.9148852459016393,
+      "grad_norm": 1.0806767599961948,
+      "learning_rate": 0.0001628936365167154,
+      "loss": 13.1214,
+      "step": 5232
+    },
+    {
+      "epoch": 0.9150601092896175,
+      "grad_norm": 1.0065139111289976,
+      "learning_rate": 0.00016287895631338624,
+      "loss": 13.1232,
+      "step": 5233
+    },
+    {
+      "epoch": 0.9152349726775957,
+      "grad_norm": 1.1843266644570525,
+      "learning_rate": 0.00016286427386849576,
+      "loss": 12.9782,
+      "step": 5234
+    },
+    {
+      "epoch": 0.9154098360655738,
+      "grad_norm": 1.0499649384005998,
+      "learning_rate": 0.00016284958918256738,
+      "loss": 12.9317,
+      "step": 5235
+    },
+    {
+      "epoch": 0.9155846994535519,
+      "grad_norm": 0.9919351371885171,
+      "learning_rate": 0.0001628349022561246,
+      "loss": 13.0773,
+      "step": 5236
+    },
+    {
+      "epoch": 0.91575956284153,
+      "grad_norm": 1.1303796663763577,
+      "learning_rate": 0.00016282021308969093,
+      "loss": 13.2121,
+      "step": 5237
+    },
+    {
+      "epoch": 0.9159344262295082,
+      "grad_norm": 1.0861380211134337,
+      "learning_rate": 0.00016280552168379008,
+      "loss": 13.097,
+      "step": 5238
+    },
+    {
+      "epoch": 0.9161092896174864,
+      "grad_norm": 0.9247479871892634,
+      "learning_rate": 0.00016279082803894577,
+      "loss": 13.2906,
+      "step": 5239
+    },
+    {
+      "epoch": 0.9162841530054645,
+      "grad_norm": 0.9861326622387075,
+      "learning_rate": 0.00016277613215568177,
+      "loss": 12.998,
+      "step": 5240
+    },
+    {
+      "epoch": 0.9164590163934426,
+      "grad_norm": 1.0612893099090601,
+      "learning_rate": 0.00016276143403452204,
+      "loss": 13.1041,
+      "step": 5241
+    },
+    {
+      "epoch": 0.9166338797814207,
+      "grad_norm": 1.1082611586000426,
+      "learning_rate": 0.0001627467336759905,
+      "loss": 13.0489,
+      "step": 5242
+    },
+    {
+      "epoch": 0.9168087431693989,
+      "grad_norm": 1.0519652290678172,
+      "learning_rate": 0.00016273203108061122,
+      "loss": 13.1454,
+      "step": 5243
+    },
+    {
+      "epoch": 0.9169836065573771,
+      "grad_norm": 1.0101400812987142,
+      "learning_rate": 0.00016271732624890834,
+      "loss": 12.9821,
+      "step": 5244
+    },
+    {
+      "epoch": 0.9171584699453552,
+      "grad_norm": 0.9891057795727843,
+      "learning_rate": 0.00016270261918140605,
+      "loss": 13.108,
+      "step": 5245
+    },
+    {
+      "epoch": 0.9173333333333333,
+      "grad_norm": 1.1005093886215969,
+      "learning_rate": 0.00016268790987862864,
+      "loss": 12.9244,
+      "step": 5246
+    },
+    {
+      "epoch": 0.9175081967213115,
+      "grad_norm": 0.9821464186209711,
+      "learning_rate": 0.00016267319834110047,
+      "loss": 12.8329,
+      "step": 5247
+    },
+    {
+      "epoch": 0.9176830601092896,
+      "grad_norm": 1.016991928179774,
+      "learning_rate": 0.00016265848456934606,
+      "loss": 13.1312,
+      "step": 5248
+    },
+    {
+      "epoch": 0.9178579234972678,
+      "grad_norm": 1.1638726108142752,
+      "learning_rate": 0.0001626437685638898,
+      "loss": 12.9194,
+      "step": 5249
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 0.9470440722859793,
+      "learning_rate": 0.00016262905032525646,
+      "loss": 13.0933,
+      "step": 5250
+    },
+    {
+      "epoch": 0.918207650273224,
+      "grad_norm": 1.2112566033421905,
+      "learning_rate": 0.00016261432985397058,
+      "loss": 13.0376,
+      "step": 5251
+    },
+    {
+      "epoch": 0.9183825136612022,
+      "grad_norm": 0.9899800590663986,
+      "learning_rate": 0.00016259960715055708,
+      "loss": 13.0211,
+      "step": 5252
+    },
+    {
+      "epoch": 0.9185573770491803,
+      "grad_norm": 0.9532294975143042,
+      "learning_rate": 0.00016258488221554065,
+      "loss": 12.8388,
+      "step": 5253
+    },
+    {
+      "epoch": 0.9187322404371585,
+      "grad_norm": 1.0502769551036406,
+      "learning_rate": 0.00016257015504944633,
+      "loss": 13.0509,
+      "step": 5254
+    },
+    {
+      "epoch": 0.9189071038251366,
+      "grad_norm": 1.0149268972992163,
+      "learning_rate": 0.00016255542565279908,
+      "loss": 12.9989,
+      "step": 5255
+    },
+    {
+      "epoch": 0.9190819672131147,
+      "grad_norm": 1.0955363227365538,
+      "learning_rate": 0.00016254069402612398,
+      "loss": 13.0424,
+      "step": 5256
+    },
+    {
+      "epoch": 0.9192568306010929,
+      "grad_norm": 1.042661370966323,
+      "learning_rate": 0.00016252596016994622,
+      "loss": 12.9968,
+      "step": 5257
+    },
+    {
+      "epoch": 0.919431693989071,
+      "grad_norm": 1.058765619149393,
+      "learning_rate": 0.00016251122408479104,
+      "loss": 12.8729,
+      "step": 5258
+    },
+    {
+      "epoch": 0.9196065573770492,
+      "grad_norm": 1.1540291852800764,
+      "learning_rate": 0.00016249648577118376,
+      "loss": 13.1919,
+      "step": 5259
+    },
+    {
+      "epoch": 0.9197814207650273,
+      "grad_norm": 1.053262600501537,
+      "learning_rate": 0.00016248174522964976,
+      "loss": 13.1106,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9199562841530055,
+      "grad_norm": 1.0141175451920508,
+      "learning_rate": 0.00016246700246071456,
+      "loss": 12.5659,
+      "step": 5261
+    },
+    {
+      "epoch": 0.9201311475409836,
+      "grad_norm": 1.1262544635040619,
+      "learning_rate": 0.00016245225746490372,
+      "loss": 13.1052,
+      "step": 5262
+    },
+    {
+      "epoch": 0.9203060109289618,
+      "grad_norm": 1.069890026374457,
+      "learning_rate": 0.00016243751024274286,
+      "loss": 13.1282,
+      "step": 5263
+    },
+    {
+      "epoch": 0.9204808743169399,
+      "grad_norm": 1.1778545285305002,
+      "learning_rate": 0.0001624227607947577,
+      "loss": 13.0529,
+      "step": 5264
+    },
+    {
+      "epoch": 0.920655737704918,
+      "grad_norm": 0.9911662543445142,
+      "learning_rate": 0.00016240800912147404,
+      "loss": 13.0938,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9208306010928962,
+      "grad_norm": 1.2164403011786473,
+      "learning_rate": 0.0001623932552234178,
+      "loss": 13.2376,
+      "step": 5266
+    },
+    {
+      "epoch": 0.9210054644808743,
+      "grad_norm": 1.026716043182348,
+      "learning_rate": 0.0001623784991011149,
+      "loss": 13.0209,
+      "step": 5267
+    },
+    {
+      "epoch": 0.9211803278688525,
+      "grad_norm": 1.2792831557859712,
+      "learning_rate": 0.00016236374075509138,
+      "loss": 12.8004,
+      "step": 5268
+    },
+    {
+      "epoch": 0.9213551912568306,
+      "grad_norm": 1.0798335943271136,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 13.3592,
+      "step": 5269
+    },
+    {
+      "epoch": 0.9215300546448087,
+      "grad_norm": 1.3041986353824409,
+      "learning_rate": 0.00016233421739398707,
+      "loss": 13.1667,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9217049180327869,
+      "grad_norm": 0.9658308116775829,
+      "learning_rate": 0.00016231945237995873,
+      "loss": 12.858,
+      "step": 5271
+    },
+    {
+      "epoch": 0.921879781420765,
+      "grad_norm": 0.9731658584621199,
+      "learning_rate": 0.00016230468514431473,
+      "loss": 12.9957,
+      "step": 5272
+    },
+    {
+      "epoch": 0.9220546448087432,
+      "grad_norm": 1.0560654108957857,
+      "learning_rate": 0.00016228991568758154,
+      "loss": 12.8931,
+      "step": 5273
+    },
+    {
+      "epoch": 0.9222295081967213,
+      "grad_norm": 1.0743070734937514,
+      "learning_rate": 0.00016227514401028558,
+      "loss": 13.1617,
+      "step": 5274
+    },
+    {
+      "epoch": 0.9224043715846995,
+      "grad_norm": 0.9832164658038864,
+      "learning_rate": 0.00016226037011295353,
+      "loss": 12.9456,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9225792349726776,
+      "grad_norm": 0.9793769146595501,
+      "learning_rate": 0.000162245593996112,
+      "loss": 13.1825,
+      "step": 5276
+    },
+    {
+      "epoch": 0.9227540983606557,
+      "grad_norm": 1.0601296338253827,
+      "learning_rate": 0.00016223081566028782,
+      "loss": 13.062,
+      "step": 5277
+    },
+    {
+      "epoch": 0.9229289617486339,
+      "grad_norm": 1.0089526731482483,
+      "learning_rate": 0.00016221603510600772,
+      "loss": 12.8337,
+      "step": 5278
+    },
+    {
+      "epoch": 0.923103825136612,
+      "grad_norm": 0.9967669795670093,
+      "learning_rate": 0.0001622012523337987,
+      "loss": 13.1075,
+      "step": 5279
+    },
+    {
+      "epoch": 0.9232786885245902,
+      "grad_norm": 1.0908047825134808,
+      "learning_rate": 0.00016218646734418766,
+      "loss": 13.1537,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9234535519125683,
+      "grad_norm": 1.2605328050564788,
+      "learning_rate": 0.00016217168013770174,
+      "loss": 13.1385,
+      "step": 5281
+    },
+    {
+      "epoch": 0.9236284153005464,
+      "grad_norm": 1.102121290545848,
+      "learning_rate": 0.0001621568907148681,
+      "loss": 13.1221,
+      "step": 5282
+    },
+    {
+      "epoch": 0.9238032786885246,
+      "grad_norm": 1.109872784774642,
+      "learning_rate": 0.00016214209907621386,
+      "loss": 13.1222,
+      "step": 5283
+    },
+    {
+      "epoch": 0.9239781420765028,
+      "grad_norm": 0.9736784457652242,
+      "learning_rate": 0.00016212730522226644,
+      "loss": 13.1591,
+      "step": 5284
+    },
+    {
+      "epoch": 0.9241530054644809,
+      "grad_norm": 1.0408470690885039,
+      "learning_rate": 0.00016211250915355318,
+      "loss": 13.0464,
+      "step": 5285
+    },
+    {
+      "epoch": 0.924327868852459,
+      "grad_norm": 0.9999838210039518,
+      "learning_rate": 0.00016209771087060152,
+      "loss": 13.1499,
+      "step": 5286
+    },
+    {
+      "epoch": 0.9245027322404371,
+      "grad_norm": 1.0057960558574082,
+      "learning_rate": 0.00016208291037393902,
+      "loss": 12.902,
+      "step": 5287
+    },
+    {
+      "epoch": 0.9246775956284153,
+      "grad_norm": 0.9749693066582582,
+      "learning_rate": 0.0001620681076640933,
+      "loss": 13.0342,
+      "step": 5288
+    },
+    {
+      "epoch": 0.9248524590163935,
+      "grad_norm": 0.9901887771523645,
+      "learning_rate": 0.00016205330274159208,
+      "loss": 12.8108,
+      "step": 5289
+    },
+    {
+      "epoch": 0.9250273224043716,
+      "grad_norm": 0.9627690716170393,
+      "learning_rate": 0.0001620384956069631,
+      "loss": 13.0701,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9252021857923497,
+      "grad_norm": 0.9841559931957863,
+      "learning_rate": 0.00016202368626073424,
+      "loss": 13.0946,
+      "step": 5291
+    },
+    {
+      "epoch": 0.9253770491803278,
+      "grad_norm": 1.1341505929881748,
+      "learning_rate": 0.00016200887470343343,
+      "loss": 12.7917,
+      "step": 5292
+    },
+    {
+      "epoch": 0.925551912568306,
+      "grad_norm": 1.0253769805325061,
+      "learning_rate": 0.00016199406093558868,
+      "loss": 12.896,
+      "step": 5293
+    },
+    {
+      "epoch": 0.9257267759562842,
+      "grad_norm": 0.931268621178718,
+      "learning_rate": 0.00016197924495772812,
+      "loss": 12.8766,
+      "step": 5294
+    },
+    {
+      "epoch": 0.9259016393442623,
+      "grad_norm": 1.057614255949337,
+      "learning_rate": 0.0001619644267703799,
+      "loss": 13.0255,
+      "step": 5295
+    },
+    {
+      "epoch": 0.9260765027322404,
+      "grad_norm": 1.0481528672809686,
+      "learning_rate": 0.00016194960637407223,
+      "loss": 13.227,
+      "step": 5296
+    },
+    {
+      "epoch": 0.9262513661202186,
+      "grad_norm": 0.9699942865515989,
+      "learning_rate": 0.00016193478376933348,
+      "loss": 13.2938,
+      "step": 5297
+    },
+    {
+      "epoch": 0.9264262295081968,
+      "grad_norm": 1.05239148474365,
+      "learning_rate": 0.00016191995895669203,
+      "loss": 13.0095,
+      "step": 5298
+    },
+    {
+      "epoch": 0.9266010928961749,
+      "grad_norm": 1.1285199916038746,
+      "learning_rate": 0.00016190513193667643,
+      "loss": 13.1271,
+      "step": 5299
+    },
+    {
+      "epoch": 0.926775956284153,
+      "grad_norm": 1.0193313522979972,
+      "learning_rate": 0.00016189030270981518,
+      "loss": 13.0492,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9269508196721311,
+      "grad_norm": 1.1794919900228968,
+      "learning_rate": 0.00016187547127663696,
+      "loss": 12.946,
+      "step": 5301
+    },
+    {
+      "epoch": 0.9271256830601093,
+      "grad_norm": 0.9750169221698061,
+      "learning_rate": 0.00016186063763767045,
+      "loss": 12.9647,
+      "step": 5302
+    },
+    {
+      "epoch": 0.9273005464480875,
+      "grad_norm": 1.0680774852499548,
+      "learning_rate": 0.00016184580179344452,
+      "loss": 12.9003,
+      "step": 5303
+    },
+    {
+      "epoch": 0.9274754098360656,
+      "grad_norm": 1.0204602587535911,
+      "learning_rate": 0.000161830963744488,
+      "loss": 13.0663,
+      "step": 5304
+    },
+    {
+      "epoch": 0.9276502732240437,
+      "grad_norm": 1.1691264791257072,
+      "learning_rate": 0.00016181612349132988,
+      "loss": 13.3329,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9278251366120218,
+      "grad_norm": 0.9907078787834903,
+      "learning_rate": 0.00016180128103449918,
+      "loss": 12.9926,
+      "step": 5306
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 1.087725636605843,
+      "learning_rate": 0.000161786436374525,
+      "loss": 13.1736,
+      "step": 5307
+    },
+    {
+      "epoch": 0.9281748633879782,
+      "grad_norm": 1.0149763527117772,
+      "learning_rate": 0.00016177158951193656,
+      "loss": 13.117,
+      "step": 5308
+    },
+    {
+      "epoch": 0.9283497267759563,
+      "grad_norm": 1.0279908397779276,
+      "learning_rate": 0.00016175674044726314,
+      "loss": 12.9696,
+      "step": 5309
+    },
+    {
+      "epoch": 0.9285245901639344,
+      "grad_norm": 1.1198034757028696,
+      "learning_rate": 0.00016174188918103406,
+      "loss": 12.8894,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9286994535519125,
+      "grad_norm": 0.9579423150232147,
+      "learning_rate": 0.0001617270357137788,
+      "loss": 13.0091,
+      "step": 5311
+    },
+    {
+      "epoch": 0.9288743169398908,
+      "grad_norm": 1.028038737939872,
+      "learning_rate": 0.0001617121800460268,
+      "loss": 12.9365,
+      "step": 5312
+    },
+    {
+      "epoch": 0.9290491803278689,
+      "grad_norm": 0.9395006275264274,
+      "learning_rate": 0.00016169732217830772,
+      "loss": 12.9443,
+      "step": 5313
+    },
+    {
+      "epoch": 0.929224043715847,
+      "grad_norm": 0.9422682630165028,
+      "learning_rate": 0.00016168246211115114,
+      "loss": 12.9516,
+      "step": 5314
+    },
+    {
+      "epoch": 0.9293989071038251,
+      "grad_norm": 1.080245368719705,
+      "learning_rate": 0.00016166759984508687,
+      "loss": 12.9493,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9295737704918032,
+      "grad_norm": 1.1145048996113534,
+      "learning_rate": 0.00016165273538064473,
+      "loss": 13.1521,
+      "step": 5316
+    },
+    {
+      "epoch": 0.9297486338797815,
+      "grad_norm": 1.0728748202299256,
+      "learning_rate": 0.00016163786871835461,
+      "loss": 13.2212,
+      "step": 5317
+    },
+    {
+      "epoch": 0.9299234972677596,
+      "grad_norm": 1.074048999355483,
+      "learning_rate": 0.0001616229998587465,
+      "loss": 12.8369,
+      "step": 5318
+    },
+    {
+      "epoch": 0.9300983606557377,
+      "grad_norm": 1.0820586647606927,
+      "learning_rate": 0.00016160812880235042,
+      "loss": 13.0636,
+      "step": 5319
+    },
+    {
+      "epoch": 0.9302732240437158,
+      "grad_norm": 0.9583217115103381,
+      "learning_rate": 0.00016159325554969653,
+      "loss": 12.8566,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9304480874316939,
+      "grad_norm": 0.9411620771467556,
+      "learning_rate": 0.00016157838010131506,
+      "loss": 13.1397,
+      "step": 5321
+    },
+    {
+      "epoch": 0.9306229508196722,
+      "grad_norm": 1.2545826697182731,
+      "learning_rate": 0.00016156350245773626,
+      "loss": 13.1293,
+      "step": 5322
+    },
+    {
+      "epoch": 0.9307978142076503,
+      "grad_norm": 1.0506600796350092,
+      "learning_rate": 0.00016154862261949054,
+      "loss": 12.968,
+      "step": 5323
+    },
+    {
+      "epoch": 0.9309726775956284,
+      "grad_norm": 1.1749629146021405,
+      "learning_rate": 0.00016153374058710832,
+      "loss": 13.4354,
+      "step": 5324
+    },
+    {
+      "epoch": 0.9311475409836065,
+      "grad_norm": 1.0479209421855455,
+      "learning_rate": 0.00016151885636112018,
+      "loss": 12.9382,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9313224043715846,
+      "grad_norm": 1.3490150664810263,
+      "learning_rate": 0.00016150396994205667,
+      "loss": 13.1875,
+      "step": 5326
+    },
+    {
+      "epoch": 0.9314972677595629,
+      "grad_norm": 1.2631761999283164,
+      "learning_rate": 0.00016148908133044852,
+      "loss": 13.139,
+      "step": 5327
+    },
+    {
+      "epoch": 0.931672131147541,
+      "grad_norm": 1.049782946767302,
+      "learning_rate": 0.0001614741905268264,
+      "loss": 13.1017,
+      "step": 5328
+    },
+    {
+      "epoch": 0.9318469945355191,
+      "grad_norm": 1.0335769947329747,
+      "learning_rate": 0.00016145929753172124,
+      "loss": 13.0596,
+      "step": 5329
+    },
+    {
+      "epoch": 0.9320218579234972,
+      "grad_norm": 0.9837291979832756,
+      "learning_rate": 0.00016144440234566393,
+      "loss": 13.0119,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9321967213114755,
+      "grad_norm": 1.038707178015588,
+      "learning_rate": 0.0001614295049691855,
+      "loss": 13.1649,
+      "step": 5331
+    },
+    {
+      "epoch": 0.9323715846994536,
+      "grad_norm": 1.0033069956531957,
+      "learning_rate": 0.00016141460540281693,
+      "loss": 13.0944,
+      "step": 5332
+    },
+    {
+      "epoch": 0.9325464480874317,
+      "grad_norm": 0.9976000672151868,
+      "learning_rate": 0.00016139970364708947,
+      "loss": 13.03,
+      "step": 5333
+    },
+    {
+      "epoch": 0.9327213114754098,
+      "grad_norm": 0.9842137657819452,
+      "learning_rate": 0.0001613847997025343,
+      "loss": 12.949,
+      "step": 5334
+    },
+    {
+      "epoch": 0.9328961748633879,
+      "grad_norm": 1.2216484857028502,
+      "learning_rate": 0.00016136989356968276,
+      "loss": 13.0797,
+      "step": 5335
+    },
+    {
+      "epoch": 0.9330710382513662,
+      "grad_norm": 0.9318235806609725,
+      "learning_rate": 0.0001613549852490662,
+      "loss": 13.1714,
+      "step": 5336
+    },
+    {
+      "epoch": 0.9332459016393443,
+      "grad_norm": 0.9820623406443466,
+      "learning_rate": 0.00016134007474121612,
+      "loss": 12.8534,
+      "step": 5337
+    },
+    {
+      "epoch": 0.9334207650273224,
+      "grad_norm": 1.2199105135848025,
+      "learning_rate": 0.000161325162046664,
+      "loss": 13.175,
+      "step": 5338
+    },
+    {
+      "epoch": 0.9335956284153005,
+      "grad_norm": 1.1005134162269197,
+      "learning_rate": 0.00016131024716594156,
+      "loss": 13.1105,
+      "step": 5339
+    },
+    {
+      "epoch": 0.9337704918032786,
+      "grad_norm": 0.9360721321815563,
+      "learning_rate": 0.0001612953300995804,
+      "loss": 13.0857,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9339453551912569,
+      "grad_norm": 1.1878195870396264,
+      "learning_rate": 0.00016128041084811237,
+      "loss": 12.7611,
+      "step": 5341
+    },
+    {
+      "epoch": 0.934120218579235,
+      "grad_norm": 1.040146181803212,
+      "learning_rate": 0.00016126548941206928,
+      "loss": 13.0374,
+      "step": 5342
+    },
+    {
+      "epoch": 0.9342950819672131,
+      "grad_norm": 1.3340668359794192,
+      "learning_rate": 0.00016125056579198305,
+      "loss": 12.9549,
+      "step": 5343
+    },
+    {
+      "epoch": 0.9344699453551912,
+      "grad_norm": 1.164549325996109,
+      "learning_rate": 0.00016123563998838574,
+      "loss": 13.0401,
+      "step": 5344
+    },
+    {
+      "epoch": 0.9346448087431694,
+      "grad_norm": 1.2760252453436123,
+      "learning_rate": 0.0001612207120018094,
+      "loss": 12.9537,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9348196721311476,
+      "grad_norm": 1.192818114894634,
+      "learning_rate": 0.00016120578183278621,
+      "loss": 12.9308,
+      "step": 5346
+    },
+    {
+      "epoch": 0.9349945355191257,
+      "grad_norm": 1.1100815672978748,
+      "learning_rate": 0.0001611908494818484,
+      "loss": 12.8472,
+      "step": 5347
+    },
+    {
+      "epoch": 0.9351693989071038,
+      "grad_norm": 1.4394502356583556,
+      "learning_rate": 0.0001611759149495283,
+      "loss": 13.1188,
+      "step": 5348
+    },
+    {
+      "epoch": 0.9353442622950819,
+      "grad_norm": 1.0172976406245964,
+      "learning_rate": 0.00016116097823635833,
+      "loss": 13.2269,
+      "step": 5349
+    },
+    {
+      "epoch": 0.9355191256830601,
+      "grad_norm": 1.4942190622451212,
+      "learning_rate": 0.00016114603934287093,
+      "loss": 13.1589,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9356939890710383,
+      "grad_norm": 0.991172693155545,
+      "learning_rate": 0.00016113109826959866,
+      "loss": 13.0279,
+      "step": 5351
+    },
+    {
+      "epoch": 0.9358688524590164,
+      "grad_norm": 1.4583316412027423,
+      "learning_rate": 0.00016111615501707418,
+      "loss": 13.182,
+      "step": 5352
+    },
+    {
+      "epoch": 0.9360437158469945,
+      "grad_norm": 1.0636223067404489,
+      "learning_rate": 0.00016110120958583014,
+      "loss": 13.138,
+      "step": 5353
+    },
+    {
+      "epoch": 0.9362185792349726,
+      "grad_norm": 1.1856092766758446,
+      "learning_rate": 0.00016108626197639943,
+      "loss": 13.0094,
+      "step": 5354
+    },
+    {
+      "epoch": 0.9363934426229508,
+      "grad_norm": 0.9821582878762112,
+      "learning_rate": 0.0001610713121893148,
+      "loss": 13.2487,
+      "step": 5355
+    },
+    {
+      "epoch": 0.936568306010929,
+      "grad_norm": 1.167887035850994,
+      "learning_rate": 0.0001610563602251093,
+      "loss": 13.183,
+      "step": 5356
+    },
+    {
+      "epoch": 0.9367431693989071,
+      "grad_norm": 0.9406775618451702,
+      "learning_rate": 0.00016104140608431586,
+      "loss": 13.0267,
+      "step": 5357
+    },
+    {
+      "epoch": 0.9369180327868852,
+      "grad_norm": 0.9749180536005657,
+      "learning_rate": 0.00016102644976746759,
+      "loss": 12.8873,
+      "step": 5358
+    },
+    {
+      "epoch": 0.9370928961748634,
+      "grad_norm": 1.032881288647813,
+      "learning_rate": 0.00016101149127509772,
+      "loss": 13.166,
+      "step": 5359
+    },
+    {
+      "epoch": 0.9372677595628416,
+      "grad_norm": 0.9883982489360726,
+      "learning_rate": 0.00016099653060773944,
+      "loss": 12.598,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9374426229508197,
+      "grad_norm": 0.9204395053327618,
+      "learning_rate": 0.00016098156776592613,
+      "loss": 12.9627,
+      "step": 5361
+    },
+    {
+      "epoch": 0.9376174863387978,
+      "grad_norm": 1.0634467145070756,
+      "learning_rate": 0.00016096660275019117,
+      "loss": 13.0073,
+      "step": 5362
+    },
+    {
+      "epoch": 0.9377923497267759,
+      "grad_norm": 0.9496883920018242,
+      "learning_rate": 0.00016095163556106807,
+      "loss": 13.0495,
+      "step": 5363
+    },
+    {
+      "epoch": 0.9379672131147541,
+      "grad_norm": 1.0019235497815238,
+      "learning_rate": 0.0001609366661990904,
+      "loss": 12.8558,
+      "step": 5364
+    },
+    {
+      "epoch": 0.9381420765027323,
+      "grad_norm": 1.0678293965424739,
+      "learning_rate": 0.0001609216946647917,
+      "loss": 12.8908,
+      "step": 5365
+    },
+    {
+      "epoch": 0.9383169398907104,
+      "grad_norm": 1.0358384310454427,
+      "learning_rate": 0.00016090672095870582,
+      "loss": 13.1739,
+      "step": 5366
+    },
+    {
+      "epoch": 0.9384918032786885,
+      "grad_norm": 1.0092816472890407,
+      "learning_rate": 0.0001608917450813665,
+      "loss": 13.0051,
+      "step": 5367
+    },
+    {
+      "epoch": 0.9386666666666666,
+      "grad_norm": 0.96577167068727,
+      "learning_rate": 0.00016087676703330762,
+      "loss": 12.8715,
+      "step": 5368
+    },
+    {
+      "epoch": 0.9388415300546448,
+      "grad_norm": 1.0394733350551408,
+      "learning_rate": 0.00016086178681506307,
+      "loss": 13.0869,
+      "step": 5369
+    },
+    {
+      "epoch": 0.939016393442623,
+      "grad_norm": 0.9766242973631275,
+      "learning_rate": 0.00016084680442716697,
+      "loss": 12.7958,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9391912568306011,
+      "grad_norm": 0.9712315831406569,
+      "learning_rate": 0.00016083181987015338,
+      "loss": 12.7737,
+      "step": 5371
+    },
+    {
+      "epoch": 0.9393661202185792,
+      "grad_norm": 1.2570455327139105,
+      "learning_rate": 0.00016081683314455648,
+      "loss": 13.0277,
+      "step": 5372
+    },
+    {
+      "epoch": 0.9395409836065574,
+      "grad_norm": 1.0368262609153864,
+      "learning_rate": 0.00016080184425091055,
+      "loss": 13.1939,
+      "step": 5373
+    },
+    {
+      "epoch": 0.9397158469945355,
+      "grad_norm": 1.0912071974412187,
+      "learning_rate": 0.00016078685318974992,
+      "loss": 12.7917,
+      "step": 5374
+    },
+    {
+      "epoch": 0.9398907103825137,
+      "grad_norm": 1.1693207751453396,
+      "learning_rate": 0.00016077185996160896,
+      "loss": 13.0118,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9400655737704918,
+      "grad_norm": 1.0368262171298743,
+      "learning_rate": 0.00016075686456702224,
+      "loss": 13.2949,
+      "step": 5376
+    },
+    {
+      "epoch": 0.9402404371584699,
+      "grad_norm": 1.1513671896886577,
+      "learning_rate": 0.00016074186700652422,
+      "loss": 12.9899,
+      "step": 5377
+    },
+    {
+      "epoch": 0.9404153005464481,
+      "grad_norm": 1.015884882245202,
+      "learning_rate": 0.00016072686728064967,
+      "loss": 13.0222,
+      "step": 5378
+    },
+    {
+      "epoch": 0.9405901639344262,
+      "grad_norm": 1.2715004132889136,
+      "learning_rate": 0.0001607118653899332,
+      "loss": 13.0008,
+      "step": 5379
+    },
+    {
+      "epoch": 0.9407650273224044,
+      "grad_norm": 1.2471649976641133,
+      "learning_rate": 0.00016069686133490972,
+      "loss": 13.1164,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9409398907103825,
+      "grad_norm": 1.0734668187002079,
+      "learning_rate": 0.000160681855116114,
+      "loss": 13.0177,
+      "step": 5381
+    },
+    {
+      "epoch": 0.9411147540983607,
+      "grad_norm": 1.1869515357495728,
+      "learning_rate": 0.0001606668467340811,
+      "loss": 12.9489,
+      "step": 5382
+    },
+    {
+      "epoch": 0.9412896174863388,
+      "grad_norm": 1.0305544500487678,
+      "learning_rate": 0.00016065183618934597,
+      "loss": 12.9939,
+      "step": 5383
+    },
+    {
+      "epoch": 0.9414644808743169,
+      "grad_norm": 1.005370793529981,
+      "learning_rate": 0.00016063682348244372,
+      "loss": 13.1122,
+      "step": 5384
+    },
+    {
+      "epoch": 0.9416393442622951,
+      "grad_norm": 1.2329534962693545,
+      "learning_rate": 0.00016062180861390956,
+      "loss": 13.0544,
+      "step": 5385
+    },
+    {
+      "epoch": 0.9418142076502732,
+      "grad_norm": 1.0905595600947142,
+      "learning_rate": 0.00016060679158427877,
+      "loss": 13.2256,
+      "step": 5386
+    },
+    {
+      "epoch": 0.9419890710382514,
+      "grad_norm": 1.2291107465541624,
+      "learning_rate": 0.0001605917723940867,
+      "loss": 13.1014,
+      "step": 5387
+    },
+    {
+      "epoch": 0.9421639344262295,
+      "grad_norm": 0.9536836166006816,
+      "learning_rate": 0.0001605767510438687,
+      "loss": 12.9572,
+      "step": 5388
+    },
+    {
+      "epoch": 0.9423387978142076,
+      "grad_norm": 1.3188933173793371,
+      "learning_rate": 0.0001605617275341603,
+      "loss": 13.0626,
+      "step": 5389
+    },
+    {
+      "epoch": 0.9425136612021858,
+      "grad_norm": 1.1078122577950196,
+      "learning_rate": 0.0001605467018654971,
+      "loss": 13.1484,
+      "step": 5390
+    },
+    {
+      "epoch": 0.942688524590164,
+      "grad_norm": 1.20607843762867,
+      "learning_rate": 0.00016053167403841468,
+      "loss": 12.7701,
+      "step": 5391
+    },
+    {
+      "epoch": 0.9428633879781421,
+      "grad_norm": 1.0485964600673667,
+      "learning_rate": 0.00016051664405344884,
+      "loss": 12.8348,
+      "step": 5392
+    },
+    {
+      "epoch": 0.9430382513661202,
+      "grad_norm": 1.138152187256898,
+      "learning_rate": 0.00016050161191113533,
+      "loss": 12.9639,
+      "step": 5393
+    },
+    {
+      "epoch": 0.9432131147540984,
+      "grad_norm": 1.0266620223880358,
+      "learning_rate": 0.00016048657761201006,
+      "loss": 13.1695,
+      "step": 5394
+    },
+    {
+      "epoch": 0.9433879781420765,
+      "grad_norm": 1.0910807251147023,
+      "learning_rate": 0.00016047154115660895,
+      "loss": 13.2427,
+      "step": 5395
+    },
+    {
+      "epoch": 0.9435628415300547,
+      "grad_norm": 1.043147032606643,
+      "learning_rate": 0.00016045650254546806,
+      "loss": 13.0065,
+      "step": 5396
+    },
+    {
+      "epoch": 0.9437377049180328,
+      "grad_norm": 1.0365469430710854,
+      "learning_rate": 0.0001604414617791235,
+      "loss": 13.1747,
+      "step": 5397
+    },
+    {
+      "epoch": 0.9439125683060109,
+      "grad_norm": 0.9284716507428904,
+      "learning_rate": 0.00016042641885811142,
+      "loss": 12.89,
+      "step": 5398
+    },
+    {
+      "epoch": 0.9440874316939891,
+      "grad_norm": 1.0877642840777222,
+      "learning_rate": 0.0001604113737829681,
+      "loss": 13.2985,
+      "step": 5399
+    },
+    {
+      "epoch": 0.9442622950819672,
+      "grad_norm": 0.9938847953758947,
+      "learning_rate": 0.0001603963265542299,
+      "loss": 13.2444,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9444371584699454,
+      "grad_norm": 1.0891802452788868,
+      "learning_rate": 0.00016038127717243325,
+      "loss": 12.9997,
+      "step": 5401
+    },
+    {
+      "epoch": 0.9446120218579235,
+      "grad_norm": 0.9908482493191274,
+      "learning_rate": 0.00016036622563811459,
+      "loss": 13.2089,
+      "step": 5402
+    },
+    {
+      "epoch": 0.9447868852459016,
+      "grad_norm": 0.9654982843239905,
+      "learning_rate": 0.00016035117195181052,
+      "loss": 12.9318,
+      "step": 5403
+    },
+    {
+      "epoch": 0.9449617486338798,
+      "grad_norm": 1.041151779440629,
+      "learning_rate": 0.0001603361161140577,
+      "loss": 12.9161,
+      "step": 5404
+    },
+    {
+      "epoch": 0.945136612021858,
+      "grad_norm": 1.0051769420933299,
+      "learning_rate": 0.00016032105812539284,
+      "loss": 13.2286,
+      "step": 5405
+    },
+    {
+      "epoch": 0.9453114754098361,
+      "grad_norm": 0.9594279421316193,
+      "learning_rate": 0.00016030599798635273,
+      "loss": 13.1867,
+      "step": 5406
+    },
+    {
+      "epoch": 0.9454863387978142,
+      "grad_norm": 0.9589917008355675,
+      "learning_rate": 0.00016029093569747423,
+      "loss": 12.9395,
+      "step": 5407
+    },
+    {
+      "epoch": 0.9456612021857923,
+      "grad_norm": 0.9815750208402833,
+      "learning_rate": 0.00016027587125929432,
+      "loss": 13.1275,
+      "step": 5408
+    },
+    {
+      "epoch": 0.9458360655737705,
+      "grad_norm": 1.1315387180553365,
+      "learning_rate": 0.00016026080467235003,
+      "loss": 13.3305,
+      "step": 5409
+    },
+    {
+      "epoch": 0.9460109289617487,
+      "grad_norm": 0.9969623005013643,
+      "learning_rate": 0.00016024573593717846,
+      "loss": 13.0303,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9461857923497268,
+      "grad_norm": 1.1222478839831957,
+      "learning_rate": 0.0001602306650543168,
+      "loss": 12.9464,
+      "step": 5411
+    },
+    {
+      "epoch": 0.9463606557377049,
+      "grad_norm": 0.9970085411322385,
+      "learning_rate": 0.00016021559202430229,
+      "loss": 13.2072,
+      "step": 5412
+    },
+    {
+      "epoch": 0.946535519125683,
+      "grad_norm": 1.34861307257846,
+      "learning_rate": 0.0001602005168476723,
+      "loss": 12.7178,
+      "step": 5413
+    },
+    {
+      "epoch": 0.9467103825136612,
+      "grad_norm": 0.9691594748063544,
+      "learning_rate": 0.0001601854395249642,
+      "loss": 13.0758,
+      "step": 5414
+    },
+    {
+      "epoch": 0.9468852459016394,
+      "grad_norm": 1.5219161033836515,
+      "learning_rate": 0.0001601703600567155,
+      "loss": 13.2516,
+      "step": 5415
+    },
+    {
+      "epoch": 0.9470601092896175,
+      "grad_norm": 1.1995218444443447,
+      "learning_rate": 0.00016015527844346378,
+      "loss": 13.1808,
+      "step": 5416
+    },
+    {
+      "epoch": 0.9472349726775956,
+      "grad_norm": 1.136734860855643,
+      "learning_rate": 0.00016014019468574663,
+      "loss": 13.0955,
+      "step": 5417
+    },
+    {
+      "epoch": 0.9474098360655737,
+      "grad_norm": 1.1849999034866545,
+      "learning_rate": 0.00016012510878410185,
+      "loss": 12.9587,
+      "step": 5418
+    },
+    {
+      "epoch": 0.947584699453552,
+      "grad_norm": 1.1056235819491107,
+      "learning_rate": 0.0001601100207390672,
+      "loss": 13.2416,
+      "step": 5419
+    },
+    {
+      "epoch": 0.9477595628415301,
+      "grad_norm": 1.1569717326250384,
+      "learning_rate": 0.0001600949305511805,
+      "loss": 13.2833,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9479344262295082,
+      "grad_norm": 1.0357501645411582,
+      "learning_rate": 0.00016007983822097975,
+      "loss": 12.6182,
+      "step": 5421
+    },
+    {
+      "epoch": 0.9481092896174863,
+      "grad_norm": 1.0340165472601828,
+      "learning_rate": 0.00016006474374900297,
+      "loss": 13.2372,
+      "step": 5422
+    },
+    {
+      "epoch": 0.9482841530054645,
+      "grad_norm": 1.120219775782231,
+      "learning_rate": 0.00016004964713578826,
+      "loss": 13.096,
+      "step": 5423
+    },
+    {
+      "epoch": 0.9484590163934427,
+      "grad_norm": 0.9242114160538553,
+      "learning_rate": 0.0001600345483818738,
+      "loss": 12.9627,
+      "step": 5424
+    },
+    {
+      "epoch": 0.9486338797814208,
+      "grad_norm": 1.0878325734367116,
+      "learning_rate": 0.0001600194474877978,
+      "loss": 12.9433,
+      "step": 5425
+    },
+    {
+      "epoch": 0.9488087431693989,
+      "grad_norm": 1.0292527607174962,
+      "learning_rate": 0.00016000434445409865,
+      "loss": 13.1605,
+      "step": 5426
+    },
+    {
+      "epoch": 0.948983606557377,
+      "grad_norm": 1.0831940765086687,
+      "learning_rate": 0.0001599892392813147,
+      "loss": 12.8204,
+      "step": 5427
+    },
+    {
+      "epoch": 0.9491584699453552,
+      "grad_norm": 1.05335178730812,
+      "learning_rate": 0.00015997413196998448,
+      "loss": 12.8089,
+      "step": 5428
+    },
+    {
+      "epoch": 0.9493333333333334,
+      "grad_norm": 0.966737871282216,
+      "learning_rate": 0.00015995902252064653,
+      "loss": 13.0408,
+      "step": 5429
+    },
+    {
+      "epoch": 0.9495081967213115,
+      "grad_norm": 1.108882213299146,
+      "learning_rate": 0.00015994391093383946,
+      "loss": 12.9813,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9496830601092896,
+      "grad_norm": 1.1236357711177067,
+      "learning_rate": 0.00015992879721010204,
+      "loss": 13.1687,
+      "step": 5431
+    },
+    {
+      "epoch": 0.9498579234972677,
+      "grad_norm": 1.0749788189665328,
+      "learning_rate": 0.000159913681349973,
+      "loss": 13.2253,
+      "step": 5432
+    },
+    {
+      "epoch": 0.950032786885246,
+      "grad_norm": 1.0195964139625664,
+      "learning_rate": 0.0001598985633539912,
+      "loss": 13.2476,
+      "step": 5433
+    },
+    {
+      "epoch": 0.9502076502732241,
+      "grad_norm": 1.1042465226318903,
+      "learning_rate": 0.00015988344322269565,
+      "loss": 12.8114,
+      "step": 5434
+    },
+    {
+      "epoch": 0.9503825136612022,
+      "grad_norm": 1.0403763181973484,
+      "learning_rate": 0.00015986832095662529,
+      "loss": 12.8752,
+      "step": 5435
+    },
+    {
+      "epoch": 0.9505573770491803,
+      "grad_norm": 1.008027271288761,
+      "learning_rate": 0.0001598531965563192,
+      "loss": 13.0617,
+      "step": 5436
+    },
+    {
+      "epoch": 0.9507322404371584,
+      "grad_norm": 0.8999257521463893,
+      "learning_rate": 0.00015983807002231662,
+      "loss": 12.835,
+      "step": 5437
+    },
+    {
+      "epoch": 0.9509071038251367,
+      "grad_norm": 1.0957674957979129,
+      "learning_rate": 0.00015982294135515675,
+      "loss": 12.7559,
+      "step": 5438
+    },
+    {
+      "epoch": 0.9510819672131148,
+      "grad_norm": 0.8995731312409265,
+      "learning_rate": 0.0001598078105553789,
+      "loss": 12.9983,
+      "step": 5439
+    },
+    {
+      "epoch": 0.9512568306010929,
+      "grad_norm": 1.0510436074615728,
+      "learning_rate": 0.00015979267762352252,
+      "loss": 12.9979,
+      "step": 5440
+    },
+    {
+      "epoch": 0.951431693989071,
+      "grad_norm": 1.1371525754240397,
+      "learning_rate": 0.00015977754256012703,
+      "loss": 13.074,
+      "step": 5441
+    },
+    {
+      "epoch": 0.9516065573770491,
+      "grad_norm": 0.9686209084578506,
+      "learning_rate": 0.00015976240536573198,
+      "loss": 13.0636,
+      "step": 5442
+    },
+    {
+      "epoch": 0.9517814207650274,
+      "grad_norm": 0.9660978984184397,
+      "learning_rate": 0.00015974726604087698,
+      "loss": 13.2903,
+      "step": 5443
+    },
+    {
+      "epoch": 0.9519562841530055,
+      "grad_norm": 0.9339141871780698,
+      "learning_rate": 0.0001597321245861018,
+      "loss": 12.8779,
+      "step": 5444
+    },
+    {
+      "epoch": 0.9521311475409836,
+      "grad_norm": 0.901548930425824,
+      "learning_rate": 0.00015971698100194614,
+      "loss": 12.9257,
+      "step": 5445
+    },
+    {
+      "epoch": 0.9523060109289617,
+      "grad_norm": 1.063730224168387,
+      "learning_rate": 0.00015970183528894987,
+      "loss": 12.8469,
+      "step": 5446
+    },
+    {
+      "epoch": 0.9524808743169398,
+      "grad_norm": 0.9672847620199743,
+      "learning_rate": 0.00015968668744765295,
+      "loss": 12.9243,
+      "step": 5447
+    },
+    {
+      "epoch": 0.9526557377049181,
+      "grad_norm": 0.9826748107293516,
+      "learning_rate": 0.00015967153747859536,
+      "loss": 13.17,
+      "step": 5448
+    },
+    {
+      "epoch": 0.9528306010928962,
+      "grad_norm": 0.9679332993998159,
+      "learning_rate": 0.00015965638538231714,
+      "loss": 12.9208,
+      "step": 5449
+    },
+    {
+      "epoch": 0.9530054644808743,
+      "grad_norm": 0.9443463791806203,
+      "learning_rate": 0.00015964123115935855,
+      "loss": 12.9875,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9531803278688524,
+      "grad_norm": 0.94252463578606,
+      "learning_rate": 0.00015962607481025972,
+      "loss": 13.197,
+      "step": 5451
+    },
+    {
+      "epoch": 0.9533551912568305,
+      "grad_norm": 0.9474495353461088,
+      "learning_rate": 0.00015961091633556095,
+      "loss": 12.9948,
+      "step": 5452
+    },
+    {
+      "epoch": 0.9535300546448088,
+      "grad_norm": 1.0252034028722141,
+      "learning_rate": 0.00015959575573580272,
+      "loss": 13.0311,
+      "step": 5453
+    },
+    {
+      "epoch": 0.9537049180327869,
+      "grad_norm": 0.9781290320747732,
+      "learning_rate": 0.00015958059301152537,
+      "loss": 13.213,
+      "step": 5454
+    },
+    {
+      "epoch": 0.953879781420765,
+      "grad_norm": 0.9543627358189808,
+      "learning_rate": 0.00015956542816326954,
+      "loss": 13.1886,
+      "step": 5455
+    },
+    {
+      "epoch": 0.9540546448087431,
+      "grad_norm": 0.857883359867671,
+      "learning_rate": 0.00015955026119157575,
+      "loss": 12.8132,
+      "step": 5456
+    },
+    {
+      "epoch": 0.9542295081967214,
+      "grad_norm": 0.9917290287566483,
+      "learning_rate": 0.00015953509209698478,
+      "loss": 13.0426,
+      "step": 5457
+    },
+    {
+      "epoch": 0.9544043715846995,
+      "grad_norm": 0.9086030453226982,
+      "learning_rate": 0.0001595199208800373,
+      "loss": 12.6724,
+      "step": 5458
+    },
+    {
+      "epoch": 0.9545792349726776,
+      "grad_norm": 0.9539837375196578,
+      "learning_rate": 0.00015950474754127417,
+      "loss": 13.0357,
+      "step": 5459
+    },
+    {
+      "epoch": 0.9547540983606557,
+      "grad_norm": 1.0069240646415978,
+      "learning_rate": 0.00015948957208123635,
+      "loss": 12.9342,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9549289617486338,
+      "grad_norm": 0.9760065254887342,
+      "learning_rate": 0.00015947439450046476,
+      "loss": 13.0935,
+      "step": 5461
+    },
+    {
+      "epoch": 0.9551038251366121,
+      "grad_norm": 1.0258542664978767,
+      "learning_rate": 0.0001594592147995005,
+      "loss": 13.101,
+      "step": 5462
+    },
+    {
+      "epoch": 0.9552786885245902,
+      "grad_norm": 0.9753606688795596,
+      "learning_rate": 0.0001594440329788847,
+      "loss": 12.5692,
+      "step": 5463
+    },
+    {
+      "epoch": 0.9554535519125683,
+      "grad_norm": 1.162377428740842,
+      "learning_rate": 0.0001594288490391586,
+      "loss": 12.8847,
+      "step": 5464
+    },
+    {
+      "epoch": 0.9556284153005464,
+      "grad_norm": 1.0144501983983822,
+      "learning_rate": 0.00015941366298086345,
+      "loss": 13.1213,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9558032786885245,
+      "grad_norm": 0.9631274265244782,
+      "learning_rate": 0.00015939847480454065,
+      "loss": 13.1254,
+      "step": 5466
+    },
+    {
+      "epoch": 0.9559781420765028,
+      "grad_norm": 1.0066793922184631,
+      "learning_rate": 0.0001593832845107316,
+      "loss": 13.212,
+      "step": 5467
+    },
+    {
+      "epoch": 0.9561530054644809,
+      "grad_norm": 0.993961936426745,
+      "learning_rate": 0.00015936809209997783,
+      "loss": 13.1242,
+      "step": 5468
+    },
+    {
+      "epoch": 0.956327868852459,
+      "grad_norm": 0.9758426191961634,
+      "learning_rate": 0.00015935289757282095,
+      "loss": 12.6049,
+      "step": 5469
+    },
+    {
+      "epoch": 0.9565027322404371,
+      "grad_norm": 0.9788493163317662,
+      "learning_rate": 0.00015933770092980265,
+      "loss": 13.1836,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9566775956284153,
+      "grad_norm": 1.1063827992722206,
+      "learning_rate": 0.0001593225021714646,
+      "loss": 13.0454,
+      "step": 5471
+    },
+    {
+      "epoch": 0.9568524590163935,
+      "grad_norm": 1.0586315874936396,
+      "learning_rate": 0.0001593073012983487,
+      "loss": 13.0676,
+      "step": 5472
+    },
+    {
+      "epoch": 0.9570273224043716,
+      "grad_norm": 1.0518748229294133,
+      "learning_rate": 0.00015929209831099677,
+      "loss": 13.0051,
+      "step": 5473
+    },
+    {
+      "epoch": 0.9572021857923497,
+      "grad_norm": 1.0818094907861906,
+      "learning_rate": 0.00015927689320995082,
+      "loss": 12.9434,
+      "step": 5474
+    },
+    {
+      "epoch": 0.9573770491803278,
+      "grad_norm": 1.0005114249057663,
+      "learning_rate": 0.00015926168599575286,
+      "loss": 13.0241,
+      "step": 5475
+    },
+    {
+      "epoch": 0.957551912568306,
+      "grad_norm": 1.0466744897534959,
+      "learning_rate": 0.00015924647666894508,
+      "loss": 12.8837,
+      "step": 5476
+    },
+    {
+      "epoch": 0.9577267759562842,
+      "grad_norm": 0.9928719235358001,
+      "learning_rate": 0.0001592312652300696,
+      "loss": 12.8165,
+      "step": 5477
+    },
+    {
+      "epoch": 0.9579016393442623,
+      "grad_norm": 1.2327071637469256,
+      "learning_rate": 0.0001592160516796687,
+      "loss": 13.1322,
+      "step": 5478
+    },
+    {
+      "epoch": 0.9580765027322404,
+      "grad_norm": 1.102420665820387,
+      "learning_rate": 0.00015920083601828479,
+      "loss": 13.0243,
+      "step": 5479
+    },
+    {
+      "epoch": 0.9582513661202186,
+      "grad_norm": 1.0470356507985126,
+      "learning_rate": 0.0001591856182464602,
+      "loss": 13.1166,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9584262295081967,
+      "grad_norm": 1.0848633841974544,
+      "learning_rate": 0.0001591703983647375,
+      "loss": 13.1667,
+      "step": 5481
+    },
+    {
+      "epoch": 0.9586010928961749,
+      "grad_norm": 1.0014822157452945,
+      "learning_rate": 0.0001591551763736592,
+      "loss": 12.8229,
+      "step": 5482
+    },
+    {
+      "epoch": 0.958775956284153,
+      "grad_norm": 1.108953406428386,
+      "learning_rate": 0.00015913995227376804,
+      "loss": 13.1302,
+      "step": 5483
+    },
+    {
+      "epoch": 0.9589508196721311,
+      "grad_norm": 1.047278704462408,
+      "learning_rate": 0.00015912472606560663,
+      "loss": 12.8382,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9591256830601093,
+      "grad_norm": 0.8940200628922451,
+      "learning_rate": 0.00015910949774971785,
+      "loss": 12.6449,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9593005464480874,
+      "grad_norm": 1.034390024703985,
+      "learning_rate": 0.0001590942673266445,
+      "loss": 12.9359,
+      "step": 5486
+    },
+    {
+      "epoch": 0.9594754098360656,
+      "grad_norm": 1.0904372604106782,
+      "learning_rate": 0.0001590790347969296,
+      "loss": 12.8489,
+      "step": 5487
+    },
+    {
+      "epoch": 0.9596502732240437,
+      "grad_norm": 1.0148719763054272,
+      "learning_rate": 0.00015906380016111608,
+      "loss": 12.8688,
+      "step": 5488
+    },
+    {
+      "epoch": 0.9598251366120218,
+      "grad_norm": 0.9549091446597242,
+      "learning_rate": 0.00015904856341974715,
+      "loss": 13.092,
+      "step": 5489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1659541968863483,
+      "learning_rate": 0.0001590333245733659,
+      "loss": 13.1029,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9601748633879782,
+      "grad_norm": 1.0475602910317154,
+      "learning_rate": 0.0001590180836225156,
+      "loss": 13.0846,
+      "step": 5491
+    },
+    {
+      "epoch": 0.9603497267759563,
+      "grad_norm": 1.3972551085845217,
+      "learning_rate": 0.00015900284056773957,
+      "loss": 13.2713,
+      "step": 5492
+    },
+    {
+      "epoch": 0.9605245901639344,
+      "grad_norm": 1.1775731628336537,
+      "learning_rate": 0.00015898759540958123,
+      "loss": 12.9805,
+      "step": 5493
+    },
+    {
+      "epoch": 0.9606994535519126,
+      "grad_norm": 1.1324917303819635,
+      "learning_rate": 0.00015897234814858403,
+      "loss": 12.8183,
+      "step": 5494
+    },
+    {
+      "epoch": 0.9608743169398907,
+      "grad_norm": 1.0054079838494083,
+      "learning_rate": 0.00015895709878529151,
+      "loss": 13.09,
+      "step": 5495
+    },
+    {
+      "epoch": 0.9610491803278689,
+      "grad_norm": 1.109755041141514,
+      "learning_rate": 0.0001589418473202473,
+      "loss": 13.0659,
+      "step": 5496
+    },
+    {
+      "epoch": 0.961224043715847,
+      "grad_norm": 1.0249527278420143,
+      "learning_rate": 0.0001589265937539951,
+      "loss": 13.0197,
+      "step": 5497
+    },
+    {
+      "epoch": 0.9613989071038251,
+      "grad_norm": 1.098052106719893,
+      "learning_rate": 0.0001589113380870787,
+      "loss": 13.1447,
+      "step": 5498
+    },
+    {
+      "epoch": 0.9615737704918033,
+      "grad_norm": 1.056539524631655,
+      "learning_rate": 0.0001588960803200419,
+      "loss": 13.1013,
+      "step": 5499
+    },
+    {
+      "epoch": 0.9617486338797814,
+      "grad_norm": 1.118573254803624,
+      "learning_rate": 0.00015888082045342866,
+      "loss": 12.9812,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9619234972677596,
+      "grad_norm": 1.1400016541817315,
+      "learning_rate": 0.00015886555848778296,
+      "loss": 13.1703,
+      "step": 5501
+    },
+    {
+      "epoch": 0.9620983606557377,
+      "grad_norm": 1.2177652126253558,
+      "learning_rate": 0.0001588502944236489,
+      "loss": 13.1374,
+      "step": 5502
+    },
+    {
+      "epoch": 0.9622732240437158,
+      "grad_norm": 0.9941470416394068,
+      "learning_rate": 0.00015883502826157057,
+      "loss": 12.96,
+      "step": 5503
+    },
+    {
+      "epoch": 0.962448087431694,
+      "grad_norm": 1.1855678018098552,
+      "learning_rate": 0.00015881976000209225,
+      "loss": 12.9795,
+      "step": 5504
+    },
+    {
+      "epoch": 0.9626229508196721,
+      "grad_norm": 1.065887123478693,
+      "learning_rate": 0.0001588044896457582,
+      "loss": 12.9925,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9627978142076503,
+      "grad_norm": 0.9393656318646899,
+      "learning_rate": 0.0001587892171931128,
+      "loss": 12.9248,
+      "step": 5506
+    },
+    {
+      "epoch": 0.9629726775956284,
+      "grad_norm": 1.1090600887563564,
+      "learning_rate": 0.0001587739426447005,
+      "loss": 13.1733,
+      "step": 5507
+    },
+    {
+      "epoch": 0.9631475409836066,
+      "grad_norm": 1.0569303832267012,
+      "learning_rate": 0.00015875866600106582,
+      "loss": 13.3574,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9633224043715847,
+      "grad_norm": 1.0892839791203042,
+      "learning_rate": 0.00015874338726275334,
+      "loss": 13.2502,
+      "step": 5509
+    },
+    {
+      "epoch": 0.9634972677595628,
+      "grad_norm": 1.1551623881595519,
+      "learning_rate": 0.00015872810643030773,
+      "loss": 12.9915,
+      "step": 5510
+    },
+    {
+      "epoch": 0.963672131147541,
+      "grad_norm": 1.048091609007791,
+      "learning_rate": 0.00015871282350427373,
+      "loss": 13.1979,
+      "step": 5511
+    },
+    {
+      "epoch": 0.9638469945355191,
+      "grad_norm": 1.0854868994626923,
+      "learning_rate": 0.00015869753848519623,
+      "loss": 12.8609,
+      "step": 5512
+    },
+    {
+      "epoch": 0.9640218579234973,
+      "grad_norm": 1.042765771544001,
+      "learning_rate": 0.00015868225137362,
+      "loss": 13.0085,
+      "step": 5513
+    },
+    {
+      "epoch": 0.9641967213114754,
+      "grad_norm": 1.0572250491406987,
+      "learning_rate": 0.00015866696217009012,
+      "loss": 12.9401,
+      "step": 5514
+    },
+    {
+      "epoch": 0.9643715846994535,
+      "grad_norm": 1.1295517024315993,
+      "learning_rate": 0.00015865167087515154,
+      "loss": 12.7965,
+      "step": 5515
+    },
+    {
+      "epoch": 0.9645464480874317,
+      "grad_norm": 1.0117690540302893,
+      "learning_rate": 0.00015863637748934943,
+      "loss": 13.2776,
+      "step": 5516
+    },
+    {
+      "epoch": 0.9647213114754098,
+      "grad_norm": 1.0805367727933899,
+      "learning_rate": 0.000158621082013229,
+      "loss": 12.9184,
+      "step": 5517
+    },
+    {
+      "epoch": 0.964896174863388,
+      "grad_norm": 0.950766334454316,
+      "learning_rate": 0.00015860578444733546,
+      "loss": 13.1449,
+      "step": 5518
+    },
+    {
+      "epoch": 0.9650710382513661,
+      "grad_norm": 1.0632276777353042,
+      "learning_rate": 0.00015859048479221422,
+      "loss": 12.6403,
+      "step": 5519
+    },
+    {
+      "epoch": 0.9652459016393443,
+      "grad_norm": 0.8893042397952529,
+      "learning_rate": 0.0001585751830484106,
+      "loss": 12.9407,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9654207650273224,
+      "grad_norm": 0.9070578888761115,
+      "learning_rate": 0.00015855987921647015,
+      "loss": 13.1946,
+      "step": 5521
+    },
+    {
+      "epoch": 0.9655956284153006,
+      "grad_norm": 0.9404308791655529,
+      "learning_rate": 0.00015854457329693842,
+      "loss": 12.9527,
+      "step": 5522
+    },
+    {
+      "epoch": 0.9657704918032787,
+      "grad_norm": 1.0646452948743945,
+      "learning_rate": 0.0001585292652903611,
+      "loss": 13.1344,
+      "step": 5523
+    },
+    {
+      "epoch": 0.9659453551912568,
+      "grad_norm": 0.9733494924789688,
+      "learning_rate": 0.0001585139551972838,
+      "loss": 13.0784,
+      "step": 5524
+    },
+    {
+      "epoch": 0.966120218579235,
+      "grad_norm": 0.9144297998263317,
+      "learning_rate": 0.0001584986430182524,
+      "loss": 12.9481,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9662950819672131,
+      "grad_norm": 1.0118352505783275,
+      "learning_rate": 0.00015848332875381273,
+      "loss": 13.2882,
+      "step": 5526
+    },
+    {
+      "epoch": 0.9664699453551913,
+      "grad_norm": 0.9288533865892901,
+      "learning_rate": 0.00015846801240451073,
+      "loss": 12.8913,
+      "step": 5527
+    },
+    {
+      "epoch": 0.9666448087431694,
+      "grad_norm": 0.9393982616856045,
+      "learning_rate": 0.00015845269397089234,
+      "loss": 12.875,
+      "step": 5528
+    },
+    {
+      "epoch": 0.9668196721311475,
+      "grad_norm": 0.9279770407385811,
+      "learning_rate": 0.00015843737345350376,
+      "loss": 13.0135,
+      "step": 5529
+    },
+    {
+      "epoch": 0.9669945355191257,
+      "grad_norm": 0.9867481318193133,
+      "learning_rate": 0.00015842205085289107,
+      "loss": 13.0748,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9671693989071038,
+      "grad_norm": 1.0017662209493767,
+      "learning_rate": 0.00015840672616960054,
+      "loss": 13.0458,
+      "step": 5531
+    },
+    {
+      "epoch": 0.967344262295082,
+      "grad_norm": 1.0076848375332876,
+      "learning_rate": 0.00015839139940417845,
+      "loss": 13.1258,
+      "step": 5532
+    },
+    {
+      "epoch": 0.9675191256830601,
+      "grad_norm": 0.9420601063245073,
+      "learning_rate": 0.00015837607055717122,
+      "loss": 13.0614,
+      "step": 5533
+    },
+    {
+      "epoch": 0.9676939890710382,
+      "grad_norm": 1.0865521686290167,
+      "learning_rate": 0.00015836073962912525,
+      "loss": 13.0657,
+      "step": 5534
+    },
+    {
+      "epoch": 0.9678688524590164,
+      "grad_norm": 0.9188820362741096,
+      "learning_rate": 0.00015834540662058708,
+      "loss": 12.876,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9680437158469946,
+      "grad_norm": 1.1773712723091374,
+      "learning_rate": 0.00015833007153210338,
+      "loss": 13.1847,
+      "step": 5536
+    },
+    {
+      "epoch": 0.9682185792349727,
+      "grad_norm": 0.9881625920790658,
+      "learning_rate": 0.00015831473436422075,
+      "loss": 12.9974,
+      "step": 5537
+    },
+    {
+      "epoch": 0.9683934426229508,
+      "grad_norm": 1.1128468086763903,
+      "learning_rate": 0.000158299395117486,
+      "loss": 12.9484,
+      "step": 5538
+    },
+    {
+      "epoch": 0.9685683060109289,
+      "grad_norm": 1.009292036004313,
+      "learning_rate": 0.0001582840537924459,
+      "loss": 13.02,
+      "step": 5539
+    },
+    {
+      "epoch": 0.9687431693989071,
+      "grad_norm": 1.0803630682321588,
+      "learning_rate": 0.0001582687103896474,
+      "loss": 13.0568,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9689180327868853,
+      "grad_norm": 1.0854226542158816,
+      "learning_rate": 0.00015825336490963744,
+      "loss": 13.019,
+      "step": 5541
+    },
+    {
+      "epoch": 0.9690928961748634,
+      "grad_norm": 0.9472541664840086,
+      "learning_rate": 0.0001582380173529631,
+      "loss": 12.9921,
+      "step": 5542
+    },
+    {
+      "epoch": 0.9692677595628415,
+      "grad_norm": 1.0599368033991456,
+      "learning_rate": 0.0001582226677201715,
+      "loss": 13.1585,
+      "step": 5543
+    },
+    {
+      "epoch": 0.9694426229508196,
+      "grad_norm": 1.0943502034021018,
+      "learning_rate": 0.0001582073160118098,
+      "loss": 13.0423,
+      "step": 5544
+    },
+    {
+      "epoch": 0.9696174863387979,
+      "grad_norm": 1.1096430396325465,
+      "learning_rate": 0.00015819196222842527,
+      "loss": 12.8908,
+      "step": 5545
+    },
+    {
+      "epoch": 0.969792349726776,
+      "grad_norm": 1.0182684801628847,
+      "learning_rate": 0.00015817660637056528,
+      "loss": 12.7191,
+      "step": 5546
+    },
+    {
+      "epoch": 0.9699672131147541,
+      "grad_norm": 1.0079204278236664,
+      "learning_rate": 0.0001581612484387773,
+      "loss": 12.9026,
+      "step": 5547
+    },
+    {
+      "epoch": 0.9701420765027322,
+      "grad_norm": 1.0190167181483287,
+      "learning_rate": 0.00015814588843360872,
+      "loss": 13.1468,
+      "step": 5548
+    },
+    {
+      "epoch": 0.9703169398907103,
+      "grad_norm": 1.1040070141691218,
+      "learning_rate": 0.00015813052635560716,
+      "loss": 13.1116,
+      "step": 5549
+    },
+    {
+      "epoch": 0.9704918032786886,
+      "grad_norm": 0.9567803439326276,
+      "learning_rate": 0.00015811516220532026,
+      "loss": 13.1484,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9706666666666667,
+      "grad_norm": 1.0612401181140863,
+      "learning_rate": 0.00015809979598329572,
+      "loss": 13.3471,
+      "step": 5551
+    },
+    {
+      "epoch": 0.9708415300546448,
+      "grad_norm": 0.9859626862418754,
+      "learning_rate": 0.00015808442769008133,
+      "loss": 13.0755,
+      "step": 5552
+    },
+    {
+      "epoch": 0.9710163934426229,
+      "grad_norm": 1.0079961016419334,
+      "learning_rate": 0.00015806905732622498,
+      "loss": 13.1267,
+      "step": 5553
+    },
+    {
+      "epoch": 0.9711912568306011,
+      "grad_norm": 0.9380536660107852,
+      "learning_rate": 0.00015805368489227455,
+      "loss": 12.9468,
+      "step": 5554
+    },
+    {
+      "epoch": 0.9713661202185793,
+      "grad_norm": 0.9473588759602117,
+      "learning_rate": 0.0001580383103887781,
+      "loss": 12.8689,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9715409836065574,
+      "grad_norm": 1.0644647440554031,
+      "learning_rate": 0.0001580229338162837,
+      "loss": 13.1356,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9717158469945355,
+      "grad_norm": 0.9395210951049576,
+      "learning_rate": 0.0001580075551753395,
+      "loss": 12.8246,
+      "step": 5557
+    },
+    {
+      "epoch": 0.9718907103825136,
+      "grad_norm": 0.9653120504231506,
+      "learning_rate": 0.00015799217446649374,
+      "loss": 13.1462,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9720655737704919,
+      "grad_norm": 0.9947101491661783,
+      "learning_rate": 0.00015797679169029472,
+      "loss": 13.4209,
+      "step": 5559
+    },
+    {
+      "epoch": 0.97224043715847,
+      "grad_norm": 0.9978401337440878,
+      "learning_rate": 0.0001579614068472908,
+      "loss": 13.0308,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9724153005464481,
+      "grad_norm": 1.0711585868283746,
+      "learning_rate": 0.00015794601993803046,
+      "loss": 13.1566,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9725901639344262,
+      "grad_norm": 0.9612981577378347,
+      "learning_rate": 0.0001579306309630622,
+      "loss": 12.6805,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9727650273224043,
+      "grad_norm": 1.0045452051532744,
+      "learning_rate": 0.0001579152399229346,
+      "loss": 13.2585,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9729398907103826,
+      "grad_norm": 1.0947855970904175,
+      "learning_rate": 0.00015789984681819643,
+      "loss": 13.0648,
+      "step": 5564
+    },
+    {
+      "epoch": 0.9731147540983607,
+      "grad_norm": 0.9521380399273314,
+      "learning_rate": 0.0001578844516493963,
+      "loss": 13.0538,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9732896174863388,
+      "grad_norm": 1.1876390894535453,
+      "learning_rate": 0.00015786905441708316,
+      "loss": 13.1549,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9734644808743169,
+      "grad_norm": 0.9458961000421134,
+      "learning_rate": 0.0001578536551218058,
+      "loss": 12.8601,
+      "step": 5567
+    },
+    {
+      "epoch": 0.973639344262295,
+      "grad_norm": 1.0837647371980237,
+      "learning_rate": 0.00015783825376411325,
+      "loss": 13.0356,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9738142076502733,
+      "grad_norm": 0.9573702062153486,
+      "learning_rate": 0.00015782285034455452,
+      "loss": 13.0021,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9739890710382514,
+      "grad_norm": 0.970375653662717,
+      "learning_rate": 0.0001578074448636788,
+      "loss": 13.0038,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9741639344262295,
+      "grad_norm": 0.9737029387845367,
+      "learning_rate": 0.0001577920373220351,
+      "loss": 12.8295,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9743387978142076,
+      "grad_norm": 0.9552569803934081,
+      "learning_rate": 0.00015777662772017287,
+      "loss": 13.0729,
+      "step": 5572
+    },
+    {
+      "epoch": 0.9745136612021857,
+      "grad_norm": 1.0650974111896336,
+      "learning_rate": 0.00015776121605864132,
+      "loss": 12.7384,
+      "step": 5573
+    },
+    {
+      "epoch": 0.974688524590164,
+      "grad_norm": 1.138556865977511,
+      "learning_rate": 0.00015774580233798995,
+      "loss": 13.2009,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9748633879781421,
+      "grad_norm": 0.8935628473997531,
+      "learning_rate": 0.00015773038655876817,
+      "loss": 12.8336,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9750382513661202,
+      "grad_norm": 1.1948237455457973,
+      "learning_rate": 0.00015771496872152554,
+      "loss": 13.161,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9752131147540983,
+      "grad_norm": 1.0193845572066627,
+      "learning_rate": 0.00015769954882681174,
+      "loss": 12.8362,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9753879781420765,
+      "grad_norm": 1.0799578746356011,
+      "learning_rate": 0.0001576841268751764,
+      "loss": 13.0613,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9755628415300547,
+      "grad_norm": 1.143325808246046,
+      "learning_rate": 0.00015766870286716938,
+      "loss": 13.3478,
+      "step": 5579
+    },
+    {
+      "epoch": 0.9757377049180328,
+      "grad_norm": 1.1037634008591461,
+      "learning_rate": 0.00015765327680334048,
+      "loss": 12.782,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9759125683060109,
+      "grad_norm": 1.0158454066969482,
+      "learning_rate": 0.00015763784868423958,
+      "loss": 13.0239,
+      "step": 5581
+    },
+    {
+      "epoch": 0.976087431693989,
+      "grad_norm": 1.1192272008577044,
+      "learning_rate": 0.00015762241851041675,
+      "loss": 12.9192,
+      "step": 5582
+    },
+    {
+      "epoch": 0.9762622950819673,
+      "grad_norm": 1.3707350387450123,
+      "learning_rate": 0.00015760698628242202,
+      "loss": 13.1187,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9764371584699454,
+      "grad_norm": 1.1098019949077154,
+      "learning_rate": 0.00015759155200080553,
+      "loss": 13.1187,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9766120218579235,
+      "grad_norm": 0.987406998211166,
+      "learning_rate": 0.0001575761156661175,
+      "loss": 12.9388,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9767868852459016,
+      "grad_norm": 1.1772972039049567,
+      "learning_rate": 0.00015756067727890822,
+      "loss": 13.2284,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9769617486338797,
+      "grad_norm": 0.9954656792570882,
+      "learning_rate": 0.00015754523683972807,
+      "loss": 12.9759,
+      "step": 5587
+    },
+    {
+      "epoch": 0.977136612021858,
+      "grad_norm": 1.1719817874315968,
+      "learning_rate": 0.0001575297943491274,
+      "loss": 13.2178,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9773114754098361,
+      "grad_norm": 1.167599847626033,
+      "learning_rate": 0.00015751434980765684,
+      "loss": 12.8828,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9774863387978142,
+      "grad_norm": 1.1607631625957147,
+      "learning_rate": 0.0001574989032158669,
+      "loss": 13.0252,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9776612021857923,
+      "grad_norm": 1.0607136267683406,
+      "learning_rate": 0.0001574834545743082,
+      "loss": 13.1184,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9778360655737705,
+      "grad_norm": 1.03033931498142,
+      "learning_rate": 0.00015746800388353153,
+      "loss": 12.836,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9780109289617487,
+      "grad_norm": 1.1079645669591098,
+      "learning_rate": 0.00015745255114408768,
+      "loss": 12.8562,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9781857923497268,
+      "grad_norm": 1.0178487848749074,
+      "learning_rate": 0.0001574370963565275,
+      "loss": 12.9708,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9783606557377049,
+      "grad_norm": 1.0321042666284161,
+      "learning_rate": 0.0001574216395214019,
+      "loss": 13.02,
+      "step": 5595
+    },
+    {
+      "epoch": 0.978535519125683,
+      "grad_norm": 0.9768706645801073,
+      "learning_rate": 0.000157406180639262,
+      "loss": 12.8637,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9787103825136612,
+      "grad_norm": 0.9360981925141619,
+      "learning_rate": 0.0001573907197106588,
+      "loss": 13.1354,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9788852459016394,
+      "grad_norm": 1.0552685610076833,
+      "learning_rate": 0.00015737525673614354,
+      "loss": 12.9403,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9790601092896175,
+      "grad_norm": 1.0949074340767395,
+      "learning_rate": 0.00015735979171626737,
+      "loss": 12.7516,
+      "step": 5599
+    },
+    {
+      "epoch": 0.9792349726775956,
+      "grad_norm": 1.0210813683087443,
+      "learning_rate": 0.00015734432465158165,
+      "loss": 13.08,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9794098360655737,
+      "grad_norm": 1.0325342685279488,
+      "learning_rate": 0.00015732885554263775,
+      "loss": 13.1323,
+      "step": 5601
+    },
+    {
+      "epoch": 0.9795846994535519,
+      "grad_norm": 1.0097509594371041,
+      "learning_rate": 0.00015731338438998711,
+      "loss": 13.266,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9797595628415301,
+      "grad_norm": 1.084162267427115,
+      "learning_rate": 0.00015729791119418134,
+      "loss": 13.1375,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9799344262295082,
+      "grad_norm": 0.9604007116883926,
+      "learning_rate": 0.00015728243595577193,
+      "loss": 13.1254,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9801092896174863,
+      "grad_norm": 1.075578936663742,
+      "learning_rate": 0.00015726695867531064,
+      "loss": 12.9491,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9802841530054645,
+      "grad_norm": 0.8748336153893767,
+      "learning_rate": 0.00015725147935334919,
+      "loss": 12.6999,
+      "step": 5606
+    },
+    {
+      "epoch": 0.9804590163934426,
+      "grad_norm": 1.0582417795408372,
+      "learning_rate": 0.00015723599799043937,
+      "loss": 13.0445,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9806338797814208,
+      "grad_norm": 0.9773579031741022,
+      "learning_rate": 0.0001572205145871331,
+      "loss": 13.0641,
+      "step": 5608
+    },
+    {
+      "epoch": 0.9808087431693989,
+      "grad_norm": 0.9708619474128133,
+      "learning_rate": 0.0001572050291439823,
+      "loss": 13.0307,
+      "step": 5609
+    },
+    {
+      "epoch": 0.980983606557377,
+      "grad_norm": 1.0762660914650546,
+      "learning_rate": 0.0001571895416615391,
+      "loss": 12.9824,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9811584699453552,
+      "grad_norm": 0.9715207039899709,
+      "learning_rate": 0.00015717405214035556,
+      "loss": 13.1029,
+      "step": 5611
+    },
+    {
+      "epoch": 0.9813333333333333,
+      "grad_norm": 0.9202029413283324,
+      "learning_rate": 0.00015715856058098383,
+      "loss": 13.0119,
+      "step": 5612
+    },
+    {
+      "epoch": 0.9815081967213115,
+      "grad_norm": 1.0072707708962803,
+      "learning_rate": 0.00015714306698397622,
+      "loss": 13.0721,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9816830601092896,
+      "grad_norm": 0.9106713439559978,
+      "learning_rate": 0.00015712757134988503,
+      "loss": 13.0699,
+      "step": 5614
+    },
+    {
+      "epoch": 0.9818579234972677,
+      "grad_norm": 1.1027598751548424,
+      "learning_rate": 0.00015711207367926267,
+      "loss": 13.0172,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9820327868852459,
+      "grad_norm": 0.9437004383317746,
+      "learning_rate": 0.0001570965739726616,
+      "loss": 12.9337,
+      "step": 5616
+    },
+    {
+      "epoch": 0.9822076502732241,
+      "grad_norm": 1.0344259049198226,
+      "learning_rate": 0.0001570810722306344,
+      "loss": 13.0011,
+      "step": 5617
+    },
+    {
+      "epoch": 0.9823825136612022,
+      "grad_norm": 1.1334515557540783,
+      "learning_rate": 0.00015706556845373364,
+      "loss": 13.027,
+      "step": 5618
+    },
+    {
+      "epoch": 0.9825573770491803,
+      "grad_norm": 0.9413571433267647,
+      "learning_rate": 0.00015705006264251205,
+      "loss": 13.1695,
+      "step": 5619
+    },
+    {
+      "epoch": 0.9827322404371585,
+      "grad_norm": 1.273490623400564,
+      "learning_rate": 0.0001570345547975224,
+      "loss": 13.1048,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9829071038251366,
+      "grad_norm": 1.029025013914732,
+      "learning_rate": 0.0001570190449193175,
+      "loss": 13.2209,
+      "step": 5621
+    },
+    {
+      "epoch": 0.9830819672131148,
+      "grad_norm": 1.4147616739125288,
+      "learning_rate": 0.00015700353300845024,
+      "loss": 13.0466,
+      "step": 5622
+    },
+    {
+      "epoch": 0.9832568306010929,
+      "grad_norm": 1.0009787301111261,
+      "learning_rate": 0.00015698801906547364,
+      "loss": 13.0271,
+      "step": 5623
+    },
+    {
+      "epoch": 0.983431693989071,
+      "grad_norm": 1.1957352420959284,
+      "learning_rate": 0.00015697250309094075,
+      "loss": 13.027,
+      "step": 5624
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 1.1179369074080379,
+      "learning_rate": 0.00015695698508540472,
+      "loss": 12.9148,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9837814207650273,
+      "grad_norm": 1.059421063458121,
+      "learning_rate": 0.0001569414650494187,
+      "loss": 12.8864,
+      "step": 5626
+    },
+    {
+      "epoch": 0.9839562841530055,
+      "grad_norm": 1.0580751482487216,
+      "learning_rate": 0.00015692594298353596,
+      "loss": 13.0191,
+      "step": 5627
+    },
+    {
+      "epoch": 0.9841311475409836,
+      "grad_norm": 1.1635576755451482,
+      "learning_rate": 0.00015691041888830988,
+      "loss": 13.1416,
+      "step": 5628
+    },
+    {
+      "epoch": 0.9843060109289617,
+      "grad_norm": 1.03894150397942,
+      "learning_rate": 0.00015689489276429387,
+      "loss": 13.0074,
+      "step": 5629
+    },
+    {
+      "epoch": 0.9844808743169399,
+      "grad_norm": 1.0631435696071665,
+      "learning_rate": 0.0001568793646120414,
+      "loss": 13.0891,
+      "step": 5630
+    },
+    {
+      "epoch": 0.984655737704918,
+      "grad_norm": 1.3147749123115255,
+      "learning_rate": 0.00015686383443210602,
+      "loss": 13.1688,
+      "step": 5631
+    },
+    {
+      "epoch": 0.9848306010928962,
+      "grad_norm": 0.9366906948625507,
+      "learning_rate": 0.0001568483022250414,
+      "loss": 13.0847,
+      "step": 5632
+    },
+    {
+      "epoch": 0.9850054644808743,
+      "grad_norm": 1.0289827478738942,
+      "learning_rate": 0.0001568327679914012,
+      "loss": 13.1792,
+      "step": 5633
+    },
+    {
+      "epoch": 0.9851803278688525,
+      "grad_norm": 1.0496342374819552,
+      "learning_rate": 0.00015681723173173925,
+      "loss": 12.9525,
+      "step": 5634
+    },
+    {
+      "epoch": 0.9853551912568306,
+      "grad_norm": 0.9346918813546089,
+      "learning_rate": 0.00015680169344660936,
+      "loss": 12.9413,
+      "step": 5635
+    },
+    {
+      "epoch": 0.9855300546448087,
+      "grad_norm": 0.9574561980076749,
+      "learning_rate": 0.00015678615313656548,
+      "loss": 13.0142,
+      "step": 5636
+    },
+    {
+      "epoch": 0.9857049180327869,
+      "grad_norm": 0.9777354798102345,
+      "learning_rate": 0.00015677061080216157,
+      "loss": 12.7698,
+      "step": 5637
+    },
+    {
+      "epoch": 0.985879781420765,
+      "grad_norm": 0.9192243620656491,
+      "learning_rate": 0.0001567550664439517,
+      "loss": 12.6905,
+      "step": 5638
+    },
+    {
+      "epoch": 0.9860546448087432,
+      "grad_norm": 0.9836954780580127,
+      "learning_rate": 0.00015673952006249007,
+      "loss": 12.8287,
+      "step": 5639
+    },
+    {
+      "epoch": 0.9862295081967213,
+      "grad_norm": 0.984801277367561,
+      "learning_rate": 0.0001567239716583308,
+      "loss": 12.8336,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9864043715846994,
+      "grad_norm": 0.8970584700023332,
+      "learning_rate": 0.00015670842123202824,
+      "loss": 13.0727,
+      "step": 5641
+    },
+    {
+      "epoch": 0.9865792349726776,
+      "grad_norm": 0.9465342580673601,
+      "learning_rate": 0.00015669286878413672,
+      "loss": 13.1505,
+      "step": 5642
+    },
+    {
+      "epoch": 0.9867540983606558,
+      "grad_norm": 1.0296126142391069,
+      "learning_rate": 0.00015667731431521066,
+      "loss": 13.0141,
+      "step": 5643
+    },
+    {
+      "epoch": 0.9869289617486339,
+      "grad_norm": 1.0178494669318594,
+      "learning_rate": 0.00015666175782580455,
+      "loss": 13.0044,
+      "step": 5644
+    },
+    {
+      "epoch": 0.987103825136612,
+      "grad_norm": 1.1340388432907624,
+      "learning_rate": 0.000156646199316473,
+      "loss": 13.2465,
+      "step": 5645
+    },
+    {
+      "epoch": 0.9872786885245901,
+      "grad_norm": 0.9300499443959911,
+      "learning_rate": 0.00015663063878777061,
+      "loss": 13.202,
+      "step": 5646
+    },
+    {
+      "epoch": 0.9874535519125683,
+      "grad_norm": 0.9178855367180061,
+      "learning_rate": 0.00015661507624025213,
+      "loss": 13.1335,
+      "step": 5647
+    },
+    {
+      "epoch": 0.9876284153005465,
+      "grad_norm": 0.988482866272391,
+      "learning_rate": 0.00015659951167447232,
+      "loss": 13.1364,
+      "step": 5648
+    },
+    {
+      "epoch": 0.9878032786885246,
+      "grad_norm": 1.1043787437892274,
+      "learning_rate": 0.00015658394509098604,
+      "loss": 12.6555,
+      "step": 5649
+    },
+    {
+      "epoch": 0.9879781420765027,
+      "grad_norm": 1.05307606745739,
+      "learning_rate": 0.00015656837649034822,
+      "loss": 13.2069,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9881530054644809,
+      "grad_norm": 1.1147440754535414,
+      "learning_rate": 0.00015655280587311392,
+      "loss": 12.9282,
+      "step": 5651
+    },
+    {
+      "epoch": 0.988327868852459,
+      "grad_norm": 1.0170622089376822,
+      "learning_rate": 0.00015653723323983814,
+      "loss": 13.0175,
+      "step": 5652
+    },
+    {
+      "epoch": 0.9885027322404372,
+      "grad_norm": 1.1271606431586543,
+      "learning_rate": 0.00015652165859107603,
+      "loss": 13.1175,
+      "step": 5653
+    },
+    {
+      "epoch": 0.9886775956284153,
+      "grad_norm": 0.9376481662568956,
+      "learning_rate": 0.00015650608192738284,
+      "loss": 13.1462,
+      "step": 5654
+    },
+    {
+      "epoch": 0.9888524590163934,
+      "grad_norm": 1.0095818395898033,
+      "learning_rate": 0.00015649050324931384,
+      "loss": 13.098,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9890273224043716,
+      "grad_norm": 0.9407617098358656,
+      "learning_rate": 0.00015647492255742442,
+      "loss": 12.8934,
+      "step": 5656
+    },
+    {
+      "epoch": 0.9892021857923498,
+      "grad_norm": 1.0325254240471298,
+      "learning_rate": 0.00015645933985226998,
+      "loss": 13.1896,
+      "step": 5657
+    },
+    {
+      "epoch": 0.9893770491803279,
+      "grad_norm": 1.0110571265742443,
+      "learning_rate": 0.00015644375513440605,
+      "loss": 12.8152,
+      "step": 5658
+    },
+    {
+      "epoch": 0.989551912568306,
+      "grad_norm": 0.9642579388744977,
+      "learning_rate": 0.0001564281684043882,
+      "loss": 12.8893,
+      "step": 5659
+    },
+    {
+      "epoch": 0.9897267759562841,
+      "grad_norm": 1.031622252629102,
+      "learning_rate": 0.0001564125796627721,
+      "loss": 12.8075,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9899016393442623,
+      "grad_norm": 0.8881017918406984,
+      "learning_rate": 0.00015639698891011338,
+      "loss": 13.0526,
+      "step": 5661
+    },
+    {
+      "epoch": 0.9900765027322405,
+      "grad_norm": 0.9110047674943799,
+      "learning_rate": 0.00015638139614696796,
+      "loss": 13.0442,
+      "step": 5662
+    },
+    {
+      "epoch": 0.9902513661202186,
+      "grad_norm": 0.90507542320123,
+      "learning_rate": 0.00015636580137389161,
+      "loss": 12.8243,
+      "step": 5663
+    },
+    {
+      "epoch": 0.9904262295081967,
+      "grad_norm": 1.0448862423900356,
+      "learning_rate": 0.0001563502045914403,
+      "loss": 13.1453,
+      "step": 5664
+    },
+    {
+      "epoch": 0.9906010928961748,
+      "grad_norm": 1.0346755962725191,
+      "learning_rate": 0.00015633460580017006,
+      "loss": 12.8383,
+      "step": 5665
+    },
+    {
+      "epoch": 0.990775956284153,
+      "grad_norm": 0.9542491351913789,
+      "learning_rate": 0.00015631900500063693,
+      "loss": 13.2634,
+      "step": 5666
+    },
+    {
+      "epoch": 0.9909508196721312,
+      "grad_norm": 0.9314520754847789,
+      "learning_rate": 0.00015630340219339705,
+      "loss": 12.8974,
+      "step": 5667
+    },
+    {
+      "epoch": 0.9911256830601093,
+      "grad_norm": 0.9495637096421382,
+      "learning_rate": 0.00015628779737900667,
+      "loss": 12.8488,
+      "step": 5668
+    },
+    {
+      "epoch": 0.9913005464480874,
+      "grad_norm": 0.9703171348035438,
+      "learning_rate": 0.0001562721905580221,
+      "loss": 13.1371,
+      "step": 5669
+    },
+    {
+      "epoch": 0.9914754098360655,
+      "grad_norm": 1.0447759077395051,
+      "learning_rate": 0.00015625658173099968,
+      "loss": 13.0532,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9916502732240438,
+      "grad_norm": 0.9533286907345461,
+      "learning_rate": 0.00015624097089849586,
+      "loss": 12.8795,
+      "step": 5671
+    },
+    {
+      "epoch": 0.9918251366120219,
+      "grad_norm": 1.0819510027586912,
+      "learning_rate": 0.00015622535806106714,
+      "loss": 13.044,
+      "step": 5672
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.95937312000353,
+      "learning_rate": 0.00015620974321927004,
+      "loss": 12.8812,
+      "step": 5673
+    },
+    {
+      "epoch": 0.9921748633879781,
+      "grad_norm": 0.9909471553490584,
+      "learning_rate": 0.0001561941263736613,
+      "loss": 12.8953,
+      "step": 5674
+    },
+    {
+      "epoch": 0.9923497267759562,
+      "grad_norm": 0.9901831893410414,
+      "learning_rate": 0.00015617850752479764,
+      "loss": 12.9205,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9925245901639345,
+      "grad_norm": 0.9658509067804621,
+      "learning_rate": 0.0001561628866732358,
+      "loss": 13.0271,
+      "step": 5676
+    },
+    {
+      "epoch": 0.9926994535519126,
+      "grad_norm": 0.9970565034301627,
+      "learning_rate": 0.00015614726381953266,
+      "loss": 13.0376,
+      "step": 5677
+    },
+    {
+      "epoch": 0.9928743169398907,
+      "grad_norm": 1.0683208925514063,
+      "learning_rate": 0.00015613163896424518,
+      "loss": 13.0205,
+      "step": 5678
+    },
+    {
+      "epoch": 0.9930491803278688,
+      "grad_norm": 0.9037924564749319,
+      "learning_rate": 0.00015611601210793036,
+      "loss": 12.9982,
+      "step": 5679
+    },
+    {
+      "epoch": 0.993224043715847,
+      "grad_norm": 1.0748761786192709,
+      "learning_rate": 0.00015610038325114526,
+      "loss": 13.0122,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9933989071038252,
+      "grad_norm": 0.9801616255005129,
+      "learning_rate": 0.00015608475239444703,
+      "loss": 12.892,
+      "step": 5681
+    },
+    {
+      "epoch": 0.9935737704918033,
+      "grad_norm": 1.1002439187942723,
+      "learning_rate": 0.00015606911953839294,
+      "loss": 12.9027,
+      "step": 5682
+    },
+    {
+      "epoch": 0.9937486338797814,
+      "grad_norm": 1.0754898235904273,
+      "learning_rate": 0.00015605348468354023,
+      "loss": 12.9342,
+      "step": 5683
+    },
+    {
+      "epoch": 0.9939234972677595,
+      "grad_norm": 1.016754310680494,
+      "learning_rate": 0.0001560378478304463,
+      "loss": 13.2102,
+      "step": 5684
+    },
+    {
+      "epoch": 0.9940983606557378,
+      "grad_norm": 1.1714599133029822,
+      "learning_rate": 0.00015602220897966855,
+      "loss": 12.835,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9942732240437159,
+      "grad_norm": 1.0847927010692018,
+      "learning_rate": 0.00015600656813176452,
+      "loss": 12.7414,
+      "step": 5686
+    },
+    {
+      "epoch": 0.994448087431694,
+      "grad_norm": 1.1467649512358842,
+      "learning_rate": 0.00015599092528729175,
+      "loss": 13.2224,
+      "step": 5687
+    },
+    {
+      "epoch": 0.9946229508196721,
+      "grad_norm": 1.0101400589298057,
+      "learning_rate": 0.00015597528044680793,
+      "loss": 13.1562,
+      "step": 5688
+    },
+    {
+      "epoch": 0.9947978142076502,
+      "grad_norm": 1.202817455416812,
+      "learning_rate": 0.00015595963361087075,
+      "loss": 13.1429,
+      "step": 5689
+    },
+    {
+      "epoch": 0.9949726775956285,
+      "grad_norm": 0.985432109234334,
+      "learning_rate": 0.000155943984780038,
+      "loss": 12.4911,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9951475409836066,
+      "grad_norm": 1.0059565550449576,
+      "learning_rate": 0.00015592833395486756,
+      "loss": 13.0159,
+      "step": 5691
+    },
+    {
+      "epoch": 0.9953224043715847,
+      "grad_norm": 0.9984531611344668,
+      "learning_rate": 0.0001559126811359174,
+      "loss": 13.1112,
+      "step": 5692
+    },
+    {
+      "epoch": 0.9954972677595628,
+      "grad_norm": 0.9785082705892666,
+      "learning_rate": 0.00015589702632374546,
+      "loss": 12.9155,
+      "step": 5693
+    },
+    {
+      "epoch": 0.9956721311475409,
+      "grad_norm": 0.9892066187039625,
+      "learning_rate": 0.00015588136951890985,
+      "loss": 12.8331,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9958469945355192,
+      "grad_norm": 1.0325108873141304,
+      "learning_rate": 0.00015586571072196868,
+      "loss": 12.6427,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9960218579234973,
+      "grad_norm": 1.0141880184011536,
+      "learning_rate": 0.0001558500499334802,
+      "loss": 13.0085,
+      "step": 5696
+    },
+    {
+      "epoch": 0.9961967213114754,
+      "grad_norm": 1.0011610419057102,
+      "learning_rate": 0.00015583438715400272,
+      "loss": 13.0159,
+      "step": 5697
+    },
+    {
+      "epoch": 0.9963715846994535,
+      "grad_norm": 1.2481359145466777,
+      "learning_rate": 0.00015581872238409455,
+      "loss": 13.1011,
+      "step": 5698
+    },
+    {
+      "epoch": 0.9965464480874316,
+      "grad_norm": 0.9108631709784564,
+      "learning_rate": 0.00015580305562431414,
+      "loss": 12.896,
+      "step": 5699
+    },
+    {
+      "epoch": 0.9967213114754099,
+      "grad_norm": 1.2254065418325575,
+      "learning_rate": 0.00015578738687522002,
+      "loss": 13.0631,
+      "step": 5700
+    },
+    {
+      "epoch": 0.996896174863388,
+      "grad_norm": 0.9720909818646462,
+      "learning_rate": 0.00015577171613737073,
+      "loss": 12.9959,
+      "step": 5701
+    },
+    {
+      "epoch": 0.9970710382513661,
+      "grad_norm": 1.2226367656372983,
+      "learning_rate": 0.0001557560434113249,
+      "loss": 12.8775,
+      "step": 5702
+    },
+    {
+      "epoch": 0.9972459016393442,
+      "grad_norm": 0.9739671516772868,
+      "learning_rate": 0.00015574036869764132,
+      "loss": 13.0037,
+      "step": 5703
+    },
+    {
+      "epoch": 0.9974207650273224,
+      "grad_norm": 1.22440362226296,
+      "learning_rate": 0.00015572469199687865,
+      "loss": 12.9166,
+      "step": 5704
+    },
+    {
+      "epoch": 0.9975956284153006,
+      "grad_norm": 1.0389413215352592,
+      "learning_rate": 0.00015570901330959585,
+      "loss": 12.9755,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9977704918032787,
+      "grad_norm": 1.1998174548066176,
+      "learning_rate": 0.00015569333263635183,
+      "loss": 12.9984,
+      "step": 5706
+    },
+    {
+      "epoch": 0.9979453551912568,
+      "grad_norm": 0.9334244170918876,
+      "learning_rate": 0.00015567764997770556,
+      "loss": 12.9831,
+      "step": 5707
+    },
+    {
+      "epoch": 0.9981202185792349,
+      "grad_norm": 1.0413960900115964,
+      "learning_rate": 0.00015566196533421613,
+      "loss": 13.0097,
+      "step": 5708
+    },
+    {
+      "epoch": 0.9982950819672131,
+      "grad_norm": 0.906822846245073,
+      "learning_rate": 0.00015564627870644264,
+      "loss": 12.6445,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9984699453551913,
+      "grad_norm": 1.0569605727675035,
+      "learning_rate": 0.00015563059009494437,
+      "loss": 13.1262,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9986448087431694,
+      "grad_norm": 0.9927885819675168,
+      "learning_rate": 0.00015561489950028054,
+      "loss": 12.7954,
+      "step": 5711
+    },
+    {
+      "epoch": 0.9988196721311475,
+      "grad_norm": 0.9798891394441664,
+      "learning_rate": 0.00015559920692301052,
+      "loss": 13.0252,
+      "step": 5712
+    },
+    {
+      "epoch": 0.9989945355191256,
+      "grad_norm": 1.0805317479253553,
+      "learning_rate": 0.00015558351236369373,
+      "loss": 12.7689,
+      "step": 5713
+    },
+    {
+      "epoch": 0.9991693989071039,
+      "grad_norm": 0.9097309218357689,
+      "learning_rate": 0.00015556781582288967,
+      "loss": 12.9464,
+      "step": 5714
+    },
+    {
+      "epoch": 0.999344262295082,
+      "grad_norm": 1.4937798439335195,
+      "learning_rate": 0.0001555521173011579,
+      "loss": 13.1732,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9995191256830601,
+      "grad_norm": 1.45148611502171,
+      "learning_rate": 0.00015553641679905806,
+      "loss": 13.0073,
+      "step": 5716
+    },
+    {
+      "epoch": 0.9996939890710382,
+      "grad_norm": 1.1239097159838713,
+      "learning_rate": 0.00015552071431714984,
+      "loss": 12.9456,
+      "step": 5717
+    },
+    {
+      "epoch": 0.9998688524590164,
+      "grad_norm": 1.0580405196674623,
+      "learning_rate": 0.00015550500985599304,
+      "loss": 12.9189,
+      "step": 5718
+    },
+    {
+      "epoch": 1.0000437158469946,
+      "grad_norm": 1.3117166148540125,
+      "learning_rate": 0.00015548930341614746,
+      "loss": 12.8271,
+      "step": 5719
+    },
+    {
+      "epoch": 1.0002185792349727,
+      "grad_norm": 1.0294265743924178,
+      "learning_rate": 0.00015547359499817307,
+      "loss": 12.5875,
+      "step": 5720
+    },
+    {
+      "epoch": 1.0003934426229508,
+      "grad_norm": 1.1273185764999751,
+      "learning_rate": 0.00015545788460262982,
+      "loss": 12.8105,
+      "step": 5721
+    },
+    {
+      "epoch": 1.000568306010929,
+      "grad_norm": 1.1924919554616442,
+      "learning_rate": 0.0001554421722300778,
+      "loss": 12.896,
+      "step": 5722
+    },
+    {
+      "epoch": 1.000743169398907,
+      "grad_norm": 1.041056761742461,
+      "learning_rate": 0.0001554264578810771,
+      "loss": 12.7954,
+      "step": 5723
+    },
+    {
+      "epoch": 1.0009180327868852,
+      "grad_norm": 1.0538799116087565,
+      "learning_rate": 0.00015541074155618793,
+      "loss": 12.868,
+      "step": 5724
+    },
+    {
+      "epoch": 1.0010928961748633,
+      "grad_norm": 1.0814006033419443,
+      "learning_rate": 0.00015539502325597057,
+      "loss": 12.7864,
+      "step": 5725
+    },
+    {
+      "epoch": 1.0012677595628414,
+      "grad_norm": 1.0781823218982667,
+      "learning_rate": 0.00015537930298098536,
+      "loss": 12.6844,
+      "step": 5726
+    },
+    {
+      "epoch": 1.0014426229508198,
+      "grad_norm": 0.8897328734602151,
+      "learning_rate": 0.0001553635807317927,
+      "loss": 12.7686,
+      "step": 5727
+    },
+    {
+      "epoch": 1.0016174863387979,
+      "grad_norm": 1.0724698880546701,
+      "learning_rate": 0.00015534785650895308,
+      "loss": 12.9882,
+      "step": 5728
+    },
+    {
+      "epoch": 1.001792349726776,
+      "grad_norm": 0.8825373981485454,
+      "learning_rate": 0.00015533213031302702,
+      "loss": 12.8303,
+      "step": 5729
+    },
+    {
+      "epoch": 1.0019672131147541,
+      "grad_norm": 0.9568802321492628,
+      "learning_rate": 0.00015531640214457517,
+      "loss": 12.7803,
+      "step": 5730
+    },
+    {
+      "epoch": 1.0021420765027322,
+      "grad_norm": 0.9684240533919183,
+      "learning_rate": 0.00015530067200415822,
+      "loss": 12.9904,
+      "step": 5731
+    },
+    {
+      "epoch": 1.0023169398907104,
+      "grad_norm": 0.9678819179737588,
+      "learning_rate": 0.00015528493989233692,
+      "loss": 12.9086,
+      "step": 5732
+    },
+    {
+      "epoch": 1.0024918032786885,
+      "grad_norm": 0.893972524996136,
+      "learning_rate": 0.00015526920580967213,
+      "loss": 12.592,
+      "step": 5733
+    },
+    {
+      "epoch": 1.0026666666666666,
+      "grad_norm": 0.9239554431162853,
+      "learning_rate": 0.00015525346975672467,
+      "loss": 12.6891,
+      "step": 5734
+    },
+    {
+      "epoch": 1.0028415300546447,
+      "grad_norm": 1.053507090145417,
+      "learning_rate": 0.00015523773173405564,
+      "loss": 12.5799,
+      "step": 5735
+    },
+    {
+      "epoch": 1.003016393442623,
+      "grad_norm": 1.189773025143268,
+      "learning_rate": 0.00015522199174222595,
+      "loss": 12.8722,
+      "step": 5736
+    },
+    {
+      "epoch": 1.0031912568306012,
+      "grad_norm": 0.9898629893206867,
+      "learning_rate": 0.00015520624978179683,
+      "loss": 12.752,
+      "step": 5737
+    },
+    {
+      "epoch": 1.0033661202185793,
+      "grad_norm": 1.3169853924635708,
+      "learning_rate": 0.00015519050585332937,
+      "loss": 12.6253,
+      "step": 5738
+    },
+    {
+      "epoch": 1.0035409836065574,
+      "grad_norm": 0.9422864988608778,
+      "learning_rate": 0.0001551747599573849,
+      "loss": 12.9373,
+      "step": 5739
+    },
+    {
+      "epoch": 1.0037158469945355,
+      "grad_norm": 1.3417710953815636,
+      "learning_rate": 0.00015515901209452466,
+      "loss": 12.9034,
+      "step": 5740
+    },
+    {
+      "epoch": 1.0038907103825137,
+      "grad_norm": 0.9725475644956473,
+      "learning_rate": 0.00015514326226531008,
+      "loss": 12.7519,
+      "step": 5741
+    },
+    {
+      "epoch": 1.0040655737704918,
+      "grad_norm": 1.226599694730734,
+      "learning_rate": 0.00015512751047030266,
+      "loss": 12.6712,
+      "step": 5742
+    },
+    {
+      "epoch": 1.00424043715847,
+      "grad_norm": 0.9006208434865149,
+      "learning_rate": 0.0001551117567100639,
+      "loss": 12.6737,
+      "step": 5743
+    },
+    {
+      "epoch": 1.004415300546448,
+      "grad_norm": 1.2295676787310617,
+      "learning_rate": 0.00015509600098515536,
+      "loss": 12.7293,
+      "step": 5744
+    },
+    {
+      "epoch": 1.0045901639344261,
+      "grad_norm": 1.042112617698739,
+      "learning_rate": 0.0001550802432961388,
+      "loss": 12.732,
+      "step": 5745
+    },
+    {
+      "epoch": 1.0047650273224045,
+      "grad_norm": 1.0769711551161993,
+      "learning_rate": 0.00015506448364357593,
+      "loss": 12.6015,
+      "step": 5746
+    },
+    {
+      "epoch": 1.0049398907103826,
+      "grad_norm": 1.0095002492752392,
+      "learning_rate": 0.0001550487220280285,
+      "loss": 12.8492,
+      "step": 5747
+    },
+    {
+      "epoch": 1.0051147540983607,
+      "grad_norm": 1.1305303135372011,
+      "learning_rate": 0.00015503295845005848,
+      "loss": 12.7312,
+      "step": 5748
+    },
+    {
+      "epoch": 1.0052896174863388,
+      "grad_norm": 0.8876186223046253,
+      "learning_rate": 0.0001550171929102278,
+      "loss": 12.7421,
+      "step": 5749
+    },
+    {
+      "epoch": 1.005464480874317,
+      "grad_norm": 1.3798104883223505,
+      "learning_rate": 0.00015500142540909848,
+      "loss": 12.7822,
+      "step": 5750
+    },
+    {
+      "epoch": 1.005639344262295,
+      "grad_norm": 0.9629936720722557,
+      "learning_rate": 0.0001549856559472326,
+      "loss": 12.8829,
+      "step": 5751
+    },
+    {
+      "epoch": 1.0058142076502732,
+      "grad_norm": 1.323244173182941,
+      "learning_rate": 0.00015496988452519235,
+      "loss": 12.8504,
+      "step": 5752
+    },
+    {
+      "epoch": 1.0059890710382513,
+      "grad_norm": 0.923940011209967,
+      "learning_rate": 0.00015495411114353988,
+      "loss": 12.6983,
+      "step": 5753
+    },
+    {
+      "epoch": 1.0061639344262294,
+      "grad_norm": 1.354071523851431,
+      "learning_rate": 0.0001549383358028376,
+      "loss": 12.8353,
+      "step": 5754
+    },
+    {
+      "epoch": 1.0063387978142075,
+      "grad_norm": 0.9693519433142417,
+      "learning_rate": 0.0001549225585036478,
+      "loss": 12.6676,
+      "step": 5755
+    },
+    {
+      "epoch": 1.0065136612021859,
+      "grad_norm": 1.1613891480710639,
+      "learning_rate": 0.000154906779246533,
+      "loss": 12.8078,
+      "step": 5756
+    },
+    {
+      "epoch": 1.006688524590164,
+      "grad_norm": 0.9471573763944504,
+      "learning_rate": 0.00015489099803205566,
+      "loss": 12.6295,
+      "step": 5757
+    },
+    {
+      "epoch": 1.0068633879781421,
+      "grad_norm": 1.1326288623549619,
+      "learning_rate": 0.00015487521486077838,
+      "loss": 12.8775,
+      "step": 5758
+    },
+    {
+      "epoch": 1.0070382513661202,
+      "grad_norm": 1.0455758326538487,
+      "learning_rate": 0.00015485942973326376,
+      "loss": 12.6713,
+      "step": 5759
+    },
+    {
+      "epoch": 1.0072131147540984,
+      "grad_norm": 1.1923312474311845,
+      "learning_rate": 0.00015484364265007462,
+      "loss": 12.76,
+      "step": 5760
+    },
+    {
+      "epoch": 1.0073879781420765,
+      "grad_norm": 1.079780251283543,
+      "learning_rate": 0.0001548278536117737,
+      "loss": 12.724,
+      "step": 5761
+    },
+    {
+      "epoch": 1.0075628415300546,
+      "grad_norm": 1.293186563779246,
+      "learning_rate": 0.00015481206261892382,
+      "loss": 12.7428,
+      "step": 5762
+    },
+    {
+      "epoch": 1.0077377049180327,
+      "grad_norm": 0.9807806921196431,
+      "learning_rate": 0.00015479626967208798,
+      "loss": 12.8381,
+      "step": 5763
+    },
+    {
+      "epoch": 1.0079125683060108,
+      "grad_norm": 1.1543572315908173,
+      "learning_rate": 0.00015478047477182912,
+      "loss": 12.9089,
+      "step": 5764
+    },
+    {
+      "epoch": 1.0080874316939892,
+      "grad_norm": 1.0315542561765898,
+      "learning_rate": 0.00015476467791871037,
+      "loss": 12.8783,
+      "step": 5765
+    },
+    {
+      "epoch": 1.0082622950819673,
+      "grad_norm": 1.2364914529486868,
+      "learning_rate": 0.00015474887911329482,
+      "loss": 12.9313,
+      "step": 5766
+    },
+    {
+      "epoch": 1.0084371584699454,
+      "grad_norm": 0.9776605652983369,
+      "learning_rate": 0.00015473307835614568,
+      "loss": 12.6219,
+      "step": 5767
+    },
+    {
+      "epoch": 1.0086120218579235,
+      "grad_norm": 1.1931232683061381,
+      "learning_rate": 0.00015471727564782625,
+      "loss": 12.8025,
+      "step": 5768
+    },
+    {
+      "epoch": 1.0087868852459017,
+      "grad_norm": 1.037179058237254,
+      "learning_rate": 0.0001547014709888999,
+      "loss": 12.8139,
+      "step": 5769
+    },
+    {
+      "epoch": 1.0089617486338798,
+      "grad_norm": 1.0441931591833935,
+      "learning_rate": 0.00015468566437993,
+      "loss": 12.9056,
+      "step": 5770
+    },
+    {
+      "epoch": 1.009136612021858,
+      "grad_norm": 1.0636806188889592,
+      "learning_rate": 0.00015466985582148008,
+      "loss": 12.9277,
+      "step": 5771
+    },
+    {
+      "epoch": 1.009311475409836,
+      "grad_norm": 1.0151215940215235,
+      "learning_rate": 0.00015465404531411365,
+      "loss": 13.0257,
+      "step": 5772
+    },
+    {
+      "epoch": 1.0094863387978141,
+      "grad_norm": 1.1358476658470882,
+      "learning_rate": 0.00015463823285839437,
+      "loss": 13.01,
+      "step": 5773
+    },
+    {
+      "epoch": 1.0096612021857923,
+      "grad_norm": 0.8945744913477524,
+      "learning_rate": 0.00015462241845488593,
+      "loss": 12.6328,
+      "step": 5774
+    },
+    {
+      "epoch": 1.0098360655737706,
+      "grad_norm": 1.261879043168702,
+      "learning_rate": 0.0001546066021041521,
+      "loss": 12.8961,
+      "step": 5775
+    },
+    {
+      "epoch": 1.0100109289617487,
+      "grad_norm": 0.9350881352965094,
+      "learning_rate": 0.00015459078380675672,
+      "loss": 12.9134,
+      "step": 5776
+    },
+    {
+      "epoch": 1.0101857923497268,
+      "grad_norm": 1.2009295830834532,
+      "learning_rate": 0.00015457496356326362,
+      "loss": 12.7942,
+      "step": 5777
+    },
+    {
+      "epoch": 1.010360655737705,
+      "grad_norm": 0.9902082266049297,
+      "learning_rate": 0.00015455914137423688,
+      "loss": 12.9879,
+      "step": 5778
+    },
+    {
+      "epoch": 1.010535519125683,
+      "grad_norm": 1.0827473960347689,
+      "learning_rate": 0.0001545433172402405,
+      "loss": 12.833,
+      "step": 5779
+    },
+    {
+      "epoch": 1.0107103825136612,
+      "grad_norm": 0.8712673538338591,
+      "learning_rate": 0.00015452749116183856,
+      "loss": 12.8435,
+      "step": 5780
+    },
+    {
+      "epoch": 1.0108852459016393,
+      "grad_norm": 1.3577173447369346,
+      "learning_rate": 0.00015451166313959526,
+      "loss": 12.6365,
+      "step": 5781
+    },
+    {
+      "epoch": 1.0110601092896174,
+      "grad_norm": 0.8874050173044505,
+      "learning_rate": 0.0001544958331740749,
+      "loss": 12.8458,
+      "step": 5782
+    },
+    {
+      "epoch": 1.0112349726775955,
+      "grad_norm": 1.2356665174821075,
+      "learning_rate": 0.00015448000126584173,
+      "loss": 13.1253,
+      "step": 5783
+    },
+    {
+      "epoch": 1.0114098360655737,
+      "grad_norm": 0.9548286049137854,
+      "learning_rate": 0.0001544641674154602,
+      "loss": 13.0372,
+      "step": 5784
+    },
+    {
+      "epoch": 1.011584699453552,
+      "grad_norm": 1.2045069311632426,
+      "learning_rate": 0.00015444833162349468,
+      "loss": 12.7264,
+      "step": 5785
+    },
+    {
+      "epoch": 1.0117595628415301,
+      "grad_norm": 0.9593639060663132,
+      "learning_rate": 0.0001544324938905098,
+      "loss": 12.9075,
+      "step": 5786
+    },
+    {
+      "epoch": 1.0119344262295082,
+      "grad_norm": 1.1580808697710694,
+      "learning_rate": 0.00015441665421707008,
+      "loss": 12.8029,
+      "step": 5787
+    },
+    {
+      "epoch": 1.0121092896174864,
+      "grad_norm": 0.9667916893396415,
+      "learning_rate": 0.00015440081260374025,
+      "loss": 12.8095,
+      "step": 5788
+    },
+    {
+      "epoch": 1.0122841530054645,
+      "grad_norm": 0.9882555450603694,
+      "learning_rate": 0.00015438496905108496,
+      "loss": 13.013,
+      "step": 5789
+    },
+    {
+      "epoch": 1.0124590163934426,
+      "grad_norm": 0.843560582565455,
+      "learning_rate": 0.00015436912355966913,
+      "loss": 12.6506,
+      "step": 5790
+    },
+    {
+      "epoch": 1.0126338797814207,
+      "grad_norm": 1.015472381517174,
+      "learning_rate": 0.00015435327613005752,
+      "loss": 12.8892,
+      "step": 5791
+    },
+    {
+      "epoch": 1.0128087431693988,
+      "grad_norm": 0.9070465889351418,
+      "learning_rate": 0.00015433742676281516,
+      "loss": 12.8378,
+      "step": 5792
+    },
+    {
+      "epoch": 1.012983606557377,
+      "grad_norm": 1.1519153526932906,
+      "learning_rate": 0.000154321575458507,
+      "loss": 12.9536,
+      "step": 5793
+    },
+    {
+      "epoch": 1.0131584699453553,
+      "grad_norm": 0.889223717424324,
+      "learning_rate": 0.00015430572221769812,
+      "loss": 12.8088,
+      "step": 5794
+    },
+    {
+      "epoch": 1.0133333333333334,
+      "grad_norm": 1.0449486400979666,
+      "learning_rate": 0.00015428986704095374,
+      "loss": 12.998,
+      "step": 5795
+    },
+    {
+      "epoch": 1.0135081967213115,
+      "grad_norm": 0.9917060253512682,
+      "learning_rate": 0.00015427400992883898,
+      "loss": 12.8997,
+      "step": 5796
+    },
+    {
+      "epoch": 1.0136830601092897,
+      "grad_norm": 1.0659917216743071,
+      "learning_rate": 0.00015425815088191924,
+      "loss": 12.779,
+      "step": 5797
+    },
+    {
+      "epoch": 1.0138579234972678,
+      "grad_norm": 0.8268195752651412,
+      "learning_rate": 0.00015424228990075975,
+      "loss": 12.7984,
+      "step": 5798
+    },
+    {
+      "epoch": 1.014032786885246,
+      "grad_norm": 1.0772862664520597,
+      "learning_rate": 0.00015422642698592604,
+      "loss": 13.0184,
+      "step": 5799
+    },
+    {
+      "epoch": 1.014207650273224,
+      "grad_norm": 0.9585987999505438,
+      "learning_rate": 0.00015421056213798354,
+      "loss": 12.9469,
+      "step": 5800
+    },
+    {
+      "epoch": 1.0143825136612021,
+      "grad_norm": 1.1547910996917603,
+      "learning_rate": 0.0001541946953574979,
+      "loss": 12.8862,
+      "step": 5801
+    },
+    {
+      "epoch": 1.0145573770491803,
+      "grad_norm": 0.8787958558137422,
+      "learning_rate": 0.00015417882664503464,
+      "loss": 12.6504,
+      "step": 5802
+    },
+    {
+      "epoch": 1.0147322404371584,
+      "grad_norm": 1.1499556604565417,
+      "learning_rate": 0.00015416295600115948,
+      "loss": 12.5992,
+      "step": 5803
+    },
+    {
+      "epoch": 1.0149071038251367,
+      "grad_norm": 1.001953628238658,
+      "learning_rate": 0.00015414708342643824,
+      "loss": 12.8876,
+      "step": 5804
+    },
+    {
+      "epoch": 1.0150819672131148,
+      "grad_norm": 1.096256494443402,
+      "learning_rate": 0.00015413120892143677,
+      "loss": 12.8052,
+      "step": 5805
+    },
+    {
+      "epoch": 1.015256830601093,
+      "grad_norm": 1.0277274190708123,
+      "learning_rate": 0.00015411533248672093,
+      "loss": 12.6515,
+      "step": 5806
+    },
+    {
+      "epoch": 1.015431693989071,
+      "grad_norm": 1.118399989957836,
+      "learning_rate": 0.00015409945412285674,
+      "loss": 12.9429,
+      "step": 5807
+    },
+    {
+      "epoch": 1.0156065573770492,
+      "grad_norm": 0.9127514836930853,
+      "learning_rate": 0.0001540835738304102,
+      "loss": 12.9573,
+      "step": 5808
+    },
+    {
+      "epoch": 1.0157814207650273,
+      "grad_norm": 1.1346250526021744,
+      "learning_rate": 0.0001540676916099474,
+      "loss": 12.6327,
+      "step": 5809
+    },
+    {
+      "epoch": 1.0159562841530054,
+      "grad_norm": 1.034056073069556,
+      "learning_rate": 0.00015405180746203465,
+      "loss": 12.7788,
+      "step": 5810
+    },
+    {
+      "epoch": 1.0161311475409835,
+      "grad_norm": 1.107397885013493,
+      "learning_rate": 0.00015403592138723803,
+      "loss": 12.7105,
+      "step": 5811
+    },
+    {
+      "epoch": 1.0163060109289617,
+      "grad_norm": 0.9839199124476846,
+      "learning_rate": 0.00015402003338612402,
+      "loss": 12.9244,
+      "step": 5812
+    },
+    {
+      "epoch": 1.0164808743169398,
+      "grad_norm": 1.191963433496373,
+      "learning_rate": 0.0001540041434592589,
+      "loss": 12.8046,
+      "step": 5813
+    },
+    {
+      "epoch": 1.0166557377049181,
+      "grad_norm": 0.8723784513619921,
+      "learning_rate": 0.00015398825160720916,
+      "loss": 12.3596,
+      "step": 5814
+    },
+    {
+      "epoch": 1.0168306010928962,
+      "grad_norm": 1.0272823820338426,
+      "learning_rate": 0.00015397235783054134,
+      "loss": 12.5855,
+      "step": 5815
+    },
+    {
+      "epoch": 1.0170054644808744,
+      "grad_norm": 1.123644866061793,
+      "learning_rate": 0.00015395646212982203,
+      "loss": 12.5761,
+      "step": 5816
+    },
+    {
+      "epoch": 1.0171803278688525,
+      "grad_norm": 1.1507521471209468,
+      "learning_rate": 0.00015394056450561788,
+      "loss": 12.8554,
+      "step": 5817
+    },
+    {
+      "epoch": 1.0173551912568306,
+      "grad_norm": 0.9683171676783642,
+      "learning_rate": 0.0001539246649584956,
+      "loss": 12.9958,
+      "step": 5818
+    },
+    {
+      "epoch": 1.0175300546448087,
+      "grad_norm": 1.033590165356284,
+      "learning_rate": 0.00015390876348902202,
+      "loss": 12.5621,
+      "step": 5819
+    },
+    {
+      "epoch": 1.0177049180327868,
+      "grad_norm": 0.9597267603633057,
+      "learning_rate": 0.00015389286009776404,
+      "loss": 12.8795,
+      "step": 5820
+    },
+    {
+      "epoch": 1.017879781420765,
+      "grad_norm": 0.9690833102527847,
+      "learning_rate": 0.00015387695478528854,
+      "loss": 12.9685,
+      "step": 5821
+    },
+    {
+      "epoch": 1.018054644808743,
+      "grad_norm": 0.8866948851610025,
+      "learning_rate": 0.00015386104755216258,
+      "loss": 13.0058,
+      "step": 5822
+    },
+    {
+      "epoch": 1.0182295081967214,
+      "grad_norm": 1.0852212925944018,
+      "learning_rate": 0.00015384513839895322,
+      "loss": 12.8298,
+      "step": 5823
+    },
+    {
+      "epoch": 1.0184043715846995,
+      "grad_norm": 0.8979117111090053,
+      "learning_rate": 0.00015382922732622752,
+      "loss": 12.8248,
+      "step": 5824
+    },
+    {
+      "epoch": 1.0185792349726777,
+      "grad_norm": 0.9713902597767406,
+      "learning_rate": 0.0001538133143345528,
+      "loss": 12.8296,
+      "step": 5825
+    },
+    {
+      "epoch": 1.0187540983606558,
+      "grad_norm": 0.9337364091869716,
+      "learning_rate": 0.0001537973994244963,
+      "loss": 12.8734,
+      "step": 5826
+    },
+    {
+      "epoch": 1.018928961748634,
+      "grad_norm": 0.9968806061743415,
+      "learning_rate": 0.00015378148259662536,
+      "loss": 12.6963,
+      "step": 5827
+    },
+    {
+      "epoch": 1.019103825136612,
+      "grad_norm": 0.9224058434769836,
+      "learning_rate": 0.00015376556385150738,
+      "loss": 12.8106,
+      "step": 5828
+    },
+    {
+      "epoch": 1.0192786885245901,
+      "grad_norm": 1.023721056247785,
+      "learning_rate": 0.00015374964318970992,
+      "loss": 13.0338,
+      "step": 5829
+    },
+    {
+      "epoch": 1.0194535519125683,
+      "grad_norm": 0.9620494120590283,
+      "learning_rate": 0.0001537337206118004,
+      "loss": 12.7316,
+      "step": 5830
+    },
+    {
+      "epoch": 1.0196284153005464,
+      "grad_norm": 1.029089328094101,
+      "learning_rate": 0.0001537177961183466,
+      "loss": 12.9842,
+      "step": 5831
+    },
+    {
+      "epoch": 1.0198032786885245,
+      "grad_norm": 0.9563610994203585,
+      "learning_rate": 0.0001537018697099161,
+      "loss": 12.8188,
+      "step": 5832
+    },
+    {
+      "epoch": 1.0199781420765028,
+      "grad_norm": 0.9883213413647389,
+      "learning_rate": 0.0001536859413870767,
+      "loss": 12.833,
+      "step": 5833
+    },
+    {
+      "epoch": 1.020153005464481,
+      "grad_norm": 0.9561332159964103,
+      "learning_rate": 0.00015367001115039618,
+      "loss": 13.0105,
+      "step": 5834
+    },
+    {
+      "epoch": 1.020327868852459,
+      "grad_norm": 1.1222422359848472,
+      "learning_rate": 0.0001536540790004425,
+      "loss": 12.7275,
+      "step": 5835
+    },
+    {
+      "epoch": 1.0205027322404372,
+      "grad_norm": 0.9366864649511542,
+      "learning_rate": 0.00015363814493778359,
+      "loss": 12.8117,
+      "step": 5836
+    },
+    {
+      "epoch": 1.0206775956284153,
+      "grad_norm": 1.0472685245128925,
+      "learning_rate": 0.00015362220896298747,
+      "loss": 12.8567,
+      "step": 5837
+    },
+    {
+      "epoch": 1.0208524590163934,
+      "grad_norm": 0.8327556591522323,
+      "learning_rate": 0.00015360627107662223,
+      "loss": 12.9414,
+      "step": 5838
+    },
+    {
+      "epoch": 1.0210273224043716,
+      "grad_norm": 0.8866977362473365,
+      "learning_rate": 0.00015359033127925611,
+      "loss": 12.9607,
+      "step": 5839
+    },
+    {
+      "epoch": 1.0212021857923497,
+      "grad_norm": 0.9659563725955937,
+      "learning_rate": 0.00015357438957145728,
+      "loss": 12.753,
+      "step": 5840
+    },
+    {
+      "epoch": 1.0213770491803278,
+      "grad_norm": 0.956912052674121,
+      "learning_rate": 0.00015355844595379403,
+      "loss": 12.4886,
+      "step": 5841
+    },
+    {
+      "epoch": 1.021551912568306,
+      "grad_norm": 0.9898717838625498,
+      "learning_rate": 0.0001535425004268348,
+      "loss": 12.4759,
+      "step": 5842
+    },
+    {
+      "epoch": 1.0217267759562842,
+      "grad_norm": 1.0589691708604283,
+      "learning_rate": 0.00015352655299114793,
+      "loss": 12.7954,
+      "step": 5843
+    },
+    {
+      "epoch": 1.0219016393442624,
+      "grad_norm": 0.9832970673285913,
+      "learning_rate": 0.00015351060364730202,
+      "loss": 12.5871,
+      "step": 5844
+    },
+    {
+      "epoch": 1.0220765027322405,
+      "grad_norm": 0.9686197216254843,
+      "learning_rate": 0.00015349465239586563,
+      "loss": 12.946,
+      "step": 5845
+    },
+    {
+      "epoch": 1.0222513661202186,
+      "grad_norm": 1.0359417377551494,
+      "learning_rate": 0.0001534786992374073,
+      "loss": 12.7239,
+      "step": 5846
+    },
+    {
+      "epoch": 1.0224262295081967,
+      "grad_norm": 0.9510849223939123,
+      "learning_rate": 0.0001534627441724959,
+      "loss": 12.7308,
+      "step": 5847
+    },
+    {
+      "epoch": 1.0226010928961748,
+      "grad_norm": 1.045875426862254,
+      "learning_rate": 0.00015344678720170013,
+      "loss": 12.8284,
+      "step": 5848
+    },
+    {
+      "epoch": 1.022775956284153,
+      "grad_norm": 0.9310225258882808,
+      "learning_rate": 0.00015343082832558884,
+      "loss": 12.7542,
+      "step": 5849
+    },
+    {
+      "epoch": 1.022950819672131,
+      "grad_norm": 0.9965980480196068,
+      "learning_rate": 0.00015341486754473092,
+      "loss": 12.644,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0231256830601092,
+      "grad_norm": 0.9977385171556593,
+      "learning_rate": 0.0001533989048596954,
+      "loss": 12.8762,
+      "step": 5851
+    },
+    {
+      "epoch": 1.0233005464480873,
+      "grad_norm": 0.9531537344791426,
+      "learning_rate": 0.00015338294027105134,
+      "loss": 12.7019,
+      "step": 5852
+    },
+    {
+      "epoch": 1.0234754098360657,
+      "grad_norm": 1.0808592283859422,
+      "learning_rate": 0.0001533669737793678,
+      "loss": 12.7984,
+      "step": 5853
+    },
+    {
+      "epoch": 1.0236502732240438,
+      "grad_norm": 0.9810323968370025,
+      "learning_rate": 0.000153351005385214,
+      "loss": 12.5838,
+      "step": 5854
+    },
+    {
+      "epoch": 1.023825136612022,
+      "grad_norm": 1.1036467623087924,
+      "learning_rate": 0.0001533350350891592,
+      "loss": 12.8537,
+      "step": 5855
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 1.0244676061077191,
+      "learning_rate": 0.0001533190628917727,
+      "loss": 12.7604,
+      "step": 5856
+    },
+    {
+      "epoch": 1.0241748633879781,
+      "grad_norm": 1.0744775765251473,
+      "learning_rate": 0.0001533030887936239,
+      "loss": 12.7855,
+      "step": 5857
+    },
+    {
+      "epoch": 1.0243497267759563,
+      "grad_norm": 0.9940953121798775,
+      "learning_rate": 0.00015328711279528228,
+      "loss": 12.8177,
+      "step": 5858
+    },
+    {
+      "epoch": 1.0245245901639344,
+      "grad_norm": 1.0396305234057444,
+      "learning_rate": 0.00015327113489731735,
+      "loss": 13.0094,
+      "step": 5859
+    },
+    {
+      "epoch": 1.0246994535519125,
+      "grad_norm": 0.9296994340039273,
+      "learning_rate": 0.00015325515510029872,
+      "loss": 12.7848,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0248743169398906,
+      "grad_norm": 1.1063929632127674,
+      "learning_rate": 0.00015323917340479598,
+      "loss": 12.7713,
+      "step": 5861
+    },
+    {
+      "epoch": 1.025049180327869,
+      "grad_norm": 1.0156511503819372,
+      "learning_rate": 0.00015322318981137896,
+      "loss": 12.7798,
+      "step": 5862
+    },
+    {
+      "epoch": 1.025224043715847,
+      "grad_norm": 1.0908018657577843,
+      "learning_rate": 0.00015320720432061738,
+      "loss": 12.4797,
+      "step": 5863
+    },
+    {
+      "epoch": 1.0253989071038252,
+      "grad_norm": 0.9451088723053486,
+      "learning_rate": 0.00015319121693308115,
+      "loss": 12.4758,
+      "step": 5864
+    },
+    {
+      "epoch": 1.0255737704918033,
+      "grad_norm": 1.2218103611217703,
+      "learning_rate": 0.00015317522764934013,
+      "loss": 12.7424,
+      "step": 5865
+    },
+    {
+      "epoch": 1.0257486338797814,
+      "grad_norm": 0.9020146796945147,
+      "learning_rate": 0.00015315923646996442,
+      "loss": 12.6129,
+      "step": 5866
+    },
+    {
+      "epoch": 1.0259234972677596,
+      "grad_norm": 0.9855857027207057,
+      "learning_rate": 0.00015314324339552404,
+      "loss": 12.7591,
+      "step": 5867
+    },
+    {
+      "epoch": 1.0260983606557377,
+      "grad_norm": 0.9375211295026901,
+      "learning_rate": 0.0001531272484265891,
+      "loss": 12.9173,
+      "step": 5868
+    },
+    {
+      "epoch": 1.0262732240437158,
+      "grad_norm": 0.9597848555538238,
+      "learning_rate": 0.00015311125156372985,
+      "loss": 12.9081,
+      "step": 5869
+    },
+    {
+      "epoch": 1.026448087431694,
+      "grad_norm": 0.9175783545807291,
+      "learning_rate": 0.00015309525280751655,
+      "loss": 12.6924,
+      "step": 5870
+    },
+    {
+      "epoch": 1.026622950819672,
+      "grad_norm": 0.9459301661054323,
+      "learning_rate": 0.0001530792521585195,
+      "loss": 12.8483,
+      "step": 5871
+    },
+    {
+      "epoch": 1.0267978142076504,
+      "grad_norm": 0.9308278732841,
+      "learning_rate": 0.00015306324961730913,
+      "loss": 12.5084,
+      "step": 5872
+    },
+    {
+      "epoch": 1.0269726775956285,
+      "grad_norm": 0.8757240981251544,
+      "learning_rate": 0.00015304724518445592,
+      "loss": 12.9669,
+      "step": 5873
+    },
+    {
+      "epoch": 1.0271475409836066,
+      "grad_norm": 1.0593682335545147,
+      "learning_rate": 0.00015303123886053037,
+      "loss": 12.7075,
+      "step": 5874
+    },
+    {
+      "epoch": 1.0273224043715847,
+      "grad_norm": 0.9335619124986143,
+      "learning_rate": 0.00015301523064610313,
+      "loss": 12.7148,
+      "step": 5875
+    },
+    {
+      "epoch": 1.0274972677595628,
+      "grad_norm": 1.015105736070275,
+      "learning_rate": 0.00015299922054174485,
+      "loss": 12.8383,
+      "step": 5876
+    },
+    {
+      "epoch": 1.027672131147541,
+      "grad_norm": 1.0545599165768411,
+      "learning_rate": 0.00015298320854802627,
+      "loss": 12.7095,
+      "step": 5877
+    },
+    {
+      "epoch": 1.027846994535519,
+      "grad_norm": 0.8882884707734419,
+      "learning_rate": 0.00015296719466551827,
+      "loss": 12.6591,
+      "step": 5878
+    },
+    {
+      "epoch": 1.0280218579234972,
+      "grad_norm": 0.9557249185142375,
+      "learning_rate": 0.00015295117889479163,
+      "loss": 12.6578,
+      "step": 5879
+    },
+    {
+      "epoch": 1.0281967213114753,
+      "grad_norm": 0.9333161844952702,
+      "learning_rate": 0.00015293516123641732,
+      "loss": 12.5779,
+      "step": 5880
+    },
+    {
+      "epoch": 1.0283715846994534,
+      "grad_norm": 0.8315281269519189,
+      "learning_rate": 0.00015291914169096636,
+      "loss": 12.6794,
+      "step": 5881
+    },
+    {
+      "epoch": 1.0285464480874318,
+      "grad_norm": 1.0285160790914818,
+      "learning_rate": 0.00015290312025900983,
+      "loss": 12.9342,
+      "step": 5882
+    },
+    {
+      "epoch": 1.02872131147541,
+      "grad_norm": 1.0322814539569374,
+      "learning_rate": 0.00015288709694111886,
+      "loss": 12.3953,
+      "step": 5883
+    },
+    {
+      "epoch": 1.028896174863388,
+      "grad_norm": 0.9178338836262308,
+      "learning_rate": 0.0001528710717378647,
+      "loss": 12.7817,
+      "step": 5884
+    },
+    {
+      "epoch": 1.0290710382513661,
+      "grad_norm": 0.990628015622356,
+      "learning_rate": 0.00015285504464981862,
+      "loss": 12.803,
+      "step": 5885
+    },
+    {
+      "epoch": 1.0292459016393443,
+      "grad_norm": 0.8934470789121718,
+      "learning_rate": 0.0001528390156775519,
+      "loss": 12.5094,
+      "step": 5886
+    },
+    {
+      "epoch": 1.0294207650273224,
+      "grad_norm": 0.9528971304568961,
+      "learning_rate": 0.00015282298482163607,
+      "loss": 12.9875,
+      "step": 5887
+    },
+    {
+      "epoch": 1.0295956284153005,
+      "grad_norm": 0.9975123574801872,
+      "learning_rate": 0.0001528069520826425,
+      "loss": 12.8641,
+      "step": 5888
+    },
+    {
+      "epoch": 1.0297704918032786,
+      "grad_norm": 0.9378061901127366,
+      "learning_rate": 0.00015279091746114282,
+      "loss": 12.7364,
+      "step": 5889
+    },
+    {
+      "epoch": 1.0299453551912567,
+      "grad_norm": 0.8591495732482324,
+      "learning_rate": 0.00015277488095770858,
+      "loss": 12.4817,
+      "step": 5890
+    },
+    {
+      "epoch": 1.030120218579235,
+      "grad_norm": 0.884527765042689,
+      "learning_rate": 0.0001527588425729115,
+      "loss": 12.8358,
+      "step": 5891
+    },
+    {
+      "epoch": 1.0302950819672132,
+      "grad_norm": 0.9848111750269086,
+      "learning_rate": 0.00015274280230732332,
+      "loss": 12.8365,
+      "step": 5892
+    },
+    {
+      "epoch": 1.0304699453551913,
+      "grad_norm": 0.9020982428176488,
+      "learning_rate": 0.0001527267601615159,
+      "loss": 12.7715,
+      "step": 5893
+    },
+    {
+      "epoch": 1.0306448087431694,
+      "grad_norm": 0.8989637038232652,
+      "learning_rate": 0.00015271071613606106,
+      "loss": 12.8374,
+      "step": 5894
+    },
+    {
+      "epoch": 1.0308196721311476,
+      "grad_norm": 0.9054225839325064,
+      "learning_rate": 0.0001526946702315308,
+      "loss": 12.6814,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0309945355191257,
+      "grad_norm": 1.0827762698044259,
+      "learning_rate": 0.00015267862244849707,
+      "loss": 12.6159,
+      "step": 5896
+    },
+    {
+      "epoch": 1.0311693989071038,
+      "grad_norm": 0.9218778023116809,
+      "learning_rate": 0.00015266257278753203,
+      "loss": 12.8691,
+      "step": 5897
+    },
+    {
+      "epoch": 1.031344262295082,
+      "grad_norm": 1.0040305462998123,
+      "learning_rate": 0.00015264652124920777,
+      "loss": 12.7873,
+      "step": 5898
+    },
+    {
+      "epoch": 1.03151912568306,
+      "grad_norm": 1.0787462920778053,
+      "learning_rate": 0.00015263046783409656,
+      "loss": 12.767,
+      "step": 5899
+    },
+    {
+      "epoch": 1.0316939890710382,
+      "grad_norm": 0.8854950593788634,
+      "learning_rate": 0.00015261441254277064,
+      "loss": 12.4615,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0318688524590165,
+      "grad_norm": 1.0156494298722287,
+      "learning_rate": 0.0001525983553758024,
+      "loss": 12.9772,
+      "step": 5901
+    },
+    {
+      "epoch": 1.0320437158469946,
+      "grad_norm": 1.0004323689608254,
+      "learning_rate": 0.0001525822963337643,
+      "loss": 12.8268,
+      "step": 5902
+    },
+    {
+      "epoch": 1.0322185792349727,
+      "grad_norm": 1.0611837324006714,
+      "learning_rate": 0.00015256623541722868,
+      "loss": 12.6138,
+      "step": 5903
+    },
+    {
+      "epoch": 1.0323934426229509,
+      "grad_norm": 1.0119579579821436,
+      "learning_rate": 0.00015255017262676823,
+      "loss": 12.7141,
+      "step": 5904
+    },
+    {
+      "epoch": 1.032568306010929,
+      "grad_norm": 0.9655185349478802,
+      "learning_rate": 0.00015253410796295554,
+      "loss": 12.6824,
+      "step": 5905
+    },
+    {
+      "epoch": 1.032743169398907,
+      "grad_norm": 1.0139420516748472,
+      "learning_rate": 0.00015251804142636327,
+      "loss": 12.8793,
+      "step": 5906
+    },
+    {
+      "epoch": 1.0329180327868852,
+      "grad_norm": 0.9890149190666634,
+      "learning_rate": 0.00015250197301756418,
+      "loss": 12.783,
+      "step": 5907
+    },
+    {
+      "epoch": 1.0330928961748633,
+      "grad_norm": 1.4114446312649225,
+      "learning_rate": 0.0001524859027371311,
+      "loss": 12.7544,
+      "step": 5908
+    },
+    {
+      "epoch": 1.0332677595628414,
+      "grad_norm": 0.9670231157121283,
+      "learning_rate": 0.00015246983058563686,
+      "loss": 12.9071,
+      "step": 5909
+    },
+    {
+      "epoch": 1.0334426229508196,
+      "grad_norm": 1.0737792720407409,
+      "learning_rate": 0.0001524537565636545,
+      "loss": 12.7742,
+      "step": 5910
+    },
+    {
+      "epoch": 1.033617486338798,
+      "grad_norm": 0.9417097081031416,
+      "learning_rate": 0.000152437680671757,
+      "loss": 12.9295,
+      "step": 5911
+    },
+    {
+      "epoch": 1.033792349726776,
+      "grad_norm": 1.089480960099164,
+      "learning_rate": 0.0001524216029105175,
+      "loss": 12.9062,
+      "step": 5912
+    },
+    {
+      "epoch": 1.0339672131147541,
+      "grad_norm": 0.8736040869195002,
+      "learning_rate": 0.000152405523280509,
+      "loss": 12.6475,
+      "step": 5913
+    },
+    {
+      "epoch": 1.0341420765027323,
+      "grad_norm": 1.0284034357293081,
+      "learning_rate": 0.00015238944178230492,
+      "loss": 12.8362,
+      "step": 5914
+    },
+    {
+      "epoch": 1.0343169398907104,
+      "grad_norm": 1.065019183058274,
+      "learning_rate": 0.0001523733584164784,
+      "loss": 12.8787,
+      "step": 5915
+    },
+    {
+      "epoch": 1.0344918032786885,
+      "grad_norm": 1.2086376114128479,
+      "learning_rate": 0.0001523572731836029,
+      "loss": 12.9414,
+      "step": 5916
+    },
+    {
+      "epoch": 1.0346666666666666,
+      "grad_norm": 1.011035252039666,
+      "learning_rate": 0.00015234118608425175,
+      "loss": 13.0592,
+      "step": 5917
+    },
+    {
+      "epoch": 1.0348415300546447,
+      "grad_norm": 0.8646367717453202,
+      "learning_rate": 0.00015232509711899845,
+      "loss": 12.3822,
+      "step": 5918
+    },
+    {
+      "epoch": 1.0350163934426229,
+      "grad_norm": 0.9975204843078367,
+      "learning_rate": 0.0001523090062884166,
+      "loss": 12.8622,
+      "step": 5919
+    },
+    {
+      "epoch": 1.035191256830601,
+      "grad_norm": 0.9372515329916008,
+      "learning_rate": 0.00015229291359307977,
+      "loss": 12.7617,
+      "step": 5920
+    },
+    {
+      "epoch": 1.0353661202185793,
+      "grad_norm": 1.252709016861526,
+      "learning_rate": 0.0001522768190335617,
+      "loss": 12.8798,
+      "step": 5921
+    },
+    {
+      "epoch": 1.0355409836065574,
+      "grad_norm": 1.1153987620020709,
+      "learning_rate": 0.00015226072261043604,
+      "loss": 12.742,
+      "step": 5922
+    },
+    {
+      "epoch": 1.0357158469945356,
+      "grad_norm": 1.2820265993389164,
+      "learning_rate": 0.00015224462432427677,
+      "loss": 12.8235,
+      "step": 5923
+    },
+    {
+      "epoch": 1.0358907103825137,
+      "grad_norm": 1.40420997597151,
+      "learning_rate": 0.00015222852417565762,
+      "loss": 12.9368,
+      "step": 5924
+    },
+    {
+      "epoch": 1.0360655737704918,
+      "grad_norm": 1.1479775286875498,
+      "learning_rate": 0.00015221242216515265,
+      "loss": 12.7502,
+      "step": 5925
+    },
+    {
+      "epoch": 1.03624043715847,
+      "grad_norm": 1.466967498385392,
+      "learning_rate": 0.0001521963182933358,
+      "loss": 12.8221,
+      "step": 5926
+    },
+    {
+      "epoch": 1.036415300546448,
+      "grad_norm": 0.9463316066995873,
+      "learning_rate": 0.00015218021256078117,
+      "loss": 12.7365,
+      "step": 5927
+    },
+    {
+      "epoch": 1.0365901639344262,
+      "grad_norm": 1.236805487503475,
+      "learning_rate": 0.00015216410496806296,
+      "loss": 12.9108,
+      "step": 5928
+    },
+    {
+      "epoch": 1.0367650273224043,
+      "grad_norm": 1.014644511618251,
+      "learning_rate": 0.00015214799551575538,
+      "loss": 12.2944,
+      "step": 5929
+    },
+    {
+      "epoch": 1.0369398907103826,
+      "grad_norm": 0.9562365916489023,
+      "learning_rate": 0.00015213188420443264,
+      "loss": 12.5864,
+      "step": 5930
+    },
+    {
+      "epoch": 1.0371147540983607,
+      "grad_norm": 0.9768460970980851,
+      "learning_rate": 0.0001521157710346692,
+      "loss": 12.8024,
+      "step": 5931
+    },
+    {
+      "epoch": 1.0372896174863389,
+      "grad_norm": 1.0531029031860144,
+      "learning_rate": 0.00015209965600703938,
+      "loss": 12.6982,
+      "step": 5932
+    },
+    {
+      "epoch": 1.037464480874317,
+      "grad_norm": 1.0378878351992358,
+      "learning_rate": 0.00015208353912211768,
+      "loss": 12.7901,
+      "step": 5933
+    },
+    {
+      "epoch": 1.037639344262295,
+      "grad_norm": 1.0426041097750254,
+      "learning_rate": 0.00015206742038047867,
+      "loss": 12.7765,
+      "step": 5934
+    },
+    {
+      "epoch": 1.0378142076502732,
+      "grad_norm": 0.9960120154397091,
+      "learning_rate": 0.00015205129978269696,
+      "loss": 12.8432,
+      "step": 5935
+    },
+    {
+      "epoch": 1.0379890710382513,
+      "grad_norm": 1.0536110271299453,
+      "learning_rate": 0.00015203517732934721,
+      "loss": 12.9442,
+      "step": 5936
+    },
+    {
+      "epoch": 1.0381639344262295,
+      "grad_norm": 1.0476001660794219,
+      "learning_rate": 0.0001520190530210042,
+      "loss": 12.9322,
+      "step": 5937
+    },
+    {
+      "epoch": 1.0383387978142076,
+      "grad_norm": 0.8875095454735327,
+      "learning_rate": 0.00015200292685824273,
+      "loss": 12.7782,
+      "step": 5938
+    },
+    {
+      "epoch": 1.0385136612021857,
+      "grad_norm": 1.00457106840121,
+      "learning_rate": 0.0001519867988416377,
+      "loss": 12.9423,
+      "step": 5939
+    },
+    {
+      "epoch": 1.038688524590164,
+      "grad_norm": 0.8518362636097836,
+      "learning_rate": 0.000151970668971764,
+      "loss": 12.5095,
+      "step": 5940
+    },
+    {
+      "epoch": 1.0388633879781421,
+      "grad_norm": 0.9593081240339605,
+      "learning_rate": 0.00015195453724919667,
+      "loss": 12.5927,
+      "step": 5941
+    },
+    {
+      "epoch": 1.0390382513661203,
+      "grad_norm": 0.9794414628100548,
+      "learning_rate": 0.0001519384036745108,
+      "loss": 12.917,
+      "step": 5942
+    },
+    {
+      "epoch": 1.0392131147540984,
+      "grad_norm": 0.8819758297591879,
+      "learning_rate": 0.00015192226824828152,
+      "loss": 12.7695,
+      "step": 5943
+    },
+    {
+      "epoch": 1.0393879781420765,
+      "grad_norm": 0.8810712468710736,
+      "learning_rate": 0.00015190613097108407,
+      "loss": 12.6803,
+      "step": 5944
+    },
+    {
+      "epoch": 1.0395628415300546,
+      "grad_norm": 0.9020350130587346,
+      "learning_rate": 0.00015188999184349364,
+      "loss": 12.8355,
+      "step": 5945
+    },
+    {
+      "epoch": 1.0397377049180327,
+      "grad_norm": 0.9512507037819944,
+      "learning_rate": 0.00015187385086608567,
+      "loss": 12.8648,
+      "step": 5946
+    },
+    {
+      "epoch": 1.0399125683060109,
+      "grad_norm": 1.0076652322761075,
+      "learning_rate": 0.0001518577080394355,
+      "loss": 12.9334,
+      "step": 5947
+    },
+    {
+      "epoch": 1.040087431693989,
+      "grad_norm": 0.9303118425477559,
+      "learning_rate": 0.00015184156336411865,
+      "loss": 12.7033,
+      "step": 5948
+    },
+    {
+      "epoch": 1.040262295081967,
+      "grad_norm": 1.036783769131067,
+      "learning_rate": 0.00015182541684071062,
+      "loss": 12.883,
+      "step": 5949
+    },
+    {
+      "epoch": 1.0404371584699454,
+      "grad_norm": 1.074122887265705,
+      "learning_rate": 0.00015180926846978702,
+      "loss": 12.6767,
+      "step": 5950
+    },
+    {
+      "epoch": 1.0406120218579236,
+      "grad_norm": 0.8638541952238109,
+      "learning_rate": 0.00015179311825192358,
+      "loss": 12.5695,
+      "step": 5951
+    },
+    {
+      "epoch": 1.0407868852459017,
+      "grad_norm": 0.9915491453788682,
+      "learning_rate": 0.00015177696618769594,
+      "loss": 12.7746,
+      "step": 5952
+    },
+    {
+      "epoch": 1.0409617486338798,
+      "grad_norm": 0.9059049600350142,
+      "learning_rate": 0.00015176081227768,
+      "loss": 12.9624,
+      "step": 5953
+    },
+    {
+      "epoch": 1.041136612021858,
+      "grad_norm": 0.9915128113077263,
+      "learning_rate": 0.00015174465652245155,
+      "loss": 12.9529,
+      "step": 5954
+    },
+    {
+      "epoch": 1.041311475409836,
+      "grad_norm": 0.9327982385207654,
+      "learning_rate": 0.00015172849892258652,
+      "loss": 12.7279,
+      "step": 5955
+    },
+    {
+      "epoch": 1.0414863387978142,
+      "grad_norm": 1.0557686611824362,
+      "learning_rate": 0.00015171233947866095,
+      "loss": 12.6398,
+      "step": 5956
+    },
+    {
+      "epoch": 1.0416612021857923,
+      "grad_norm": 0.9561372991122408,
+      "learning_rate": 0.00015169617819125092,
+      "loss": 13.0025,
+      "step": 5957
+    },
+    {
+      "epoch": 1.0418360655737704,
+      "grad_norm": 1.2183415917601048,
+      "learning_rate": 0.00015168001506093255,
+      "loss": 12.9229,
+      "step": 5958
+    },
+    {
+      "epoch": 1.0420109289617487,
+      "grad_norm": 0.9725506229102289,
+      "learning_rate": 0.000151663850088282,
+      "loss": 12.6269,
+      "step": 5959
+    },
+    {
+      "epoch": 1.0421857923497269,
+      "grad_norm": 1.1330558987007604,
+      "learning_rate": 0.00015164768327387558,
+      "loss": 12.7189,
+      "step": 5960
+    },
+    {
+      "epoch": 1.042360655737705,
+      "grad_norm": 0.922804734245367,
+      "learning_rate": 0.00015163151461828958,
+      "loss": 12.7669,
+      "step": 5961
+    },
+    {
+      "epoch": 1.042535519125683,
+      "grad_norm": 1.3591651409324197,
+      "learning_rate": 0.0001516153441221004,
+      "loss": 12.9676,
+      "step": 5962
+    },
+    {
+      "epoch": 1.0427103825136612,
+      "grad_norm": 0.9276170733403708,
+      "learning_rate": 0.00015159917178588455,
+      "loss": 12.5727,
+      "step": 5963
+    },
+    {
+      "epoch": 1.0428852459016393,
+      "grad_norm": 1.588223185664152,
+      "learning_rate": 0.00015158299761021846,
+      "loss": 12.8968,
+      "step": 5964
+    },
+    {
+      "epoch": 1.0430601092896175,
+      "grad_norm": 0.9392342345152118,
+      "learning_rate": 0.00015156682159567878,
+      "loss": 12.6627,
+      "step": 5965
+    },
+    {
+      "epoch": 1.0432349726775956,
+      "grad_norm": 1.8450457523758381,
+      "learning_rate": 0.00015155064374284218,
+      "loss": 12.8852,
+      "step": 5966
+    },
+    {
+      "epoch": 1.0434098360655737,
+      "grad_norm": 1.1587117205892632,
+      "learning_rate": 0.00015153446405228538,
+      "loss": 12.8924,
+      "step": 5967
+    },
+    {
+      "epoch": 1.0435846994535518,
+      "grad_norm": 1.4568463052260145,
+      "learning_rate": 0.00015151828252458512,
+      "loss": 12.9772,
+      "step": 5968
+    },
+    {
+      "epoch": 1.0437595628415302,
+      "grad_norm": 0.9406688895326896,
+      "learning_rate": 0.00015150209916031825,
+      "loss": 12.5519,
+      "step": 5969
+    },
+    {
+      "epoch": 1.0439344262295083,
+      "grad_norm": 1.442680111678496,
+      "learning_rate": 0.00015148591396006177,
+      "loss": 12.9102,
+      "step": 5970
+    },
+    {
+      "epoch": 1.0441092896174864,
+      "grad_norm": 0.95451998390197,
+      "learning_rate": 0.0001514697269243926,
+      "loss": 12.9579,
+      "step": 5971
+    },
+    {
+      "epoch": 1.0442841530054645,
+      "grad_norm": 1.2612850265761253,
+      "learning_rate": 0.0001514535380538878,
+      "loss": 12.8011,
+      "step": 5972
+    },
+    {
+      "epoch": 1.0444590163934426,
+      "grad_norm": 0.8785179014463079,
+      "learning_rate": 0.0001514373473491245,
+      "loss": 12.6816,
+      "step": 5973
+    },
+    {
+      "epoch": 1.0446338797814207,
+      "grad_norm": 1.1059277275489223,
+      "learning_rate": 0.00015142115481067982,
+      "loss": 12.6933,
+      "step": 5974
+    },
+    {
+      "epoch": 1.0448087431693989,
+      "grad_norm": 0.8810766153851258,
+      "learning_rate": 0.0001514049604391311,
+      "loss": 12.6989,
+      "step": 5975
+    },
+    {
+      "epoch": 1.044983606557377,
+      "grad_norm": 1.0424871345652813,
+      "learning_rate": 0.00015138876423505555,
+      "loss": 13.0764,
+      "step": 5976
+    },
+    {
+      "epoch": 1.045158469945355,
+      "grad_norm": 0.9242178094374971,
+      "learning_rate": 0.00015137256619903063,
+      "loss": 12.9611,
+      "step": 5977
+    },
+    {
+      "epoch": 1.0453333333333332,
+      "grad_norm": 0.9391042603063254,
+      "learning_rate": 0.00015135636633163373,
+      "loss": 12.8577,
+      "step": 5978
+    },
+    {
+      "epoch": 1.0455081967213116,
+      "grad_norm": 0.9851819175423958,
+      "learning_rate": 0.00015134016463344243,
+      "loss": 12.6126,
+      "step": 5979
+    },
+    {
+      "epoch": 1.0456830601092897,
+      "grad_norm": 0.901949398303109,
+      "learning_rate": 0.00015132396110503418,
+      "loss": 13.0703,
+      "step": 5980
+    },
+    {
+      "epoch": 1.0458579234972678,
+      "grad_norm": 0.8489517683877947,
+      "learning_rate": 0.00015130775574698669,
+      "loss": 12.6625,
+      "step": 5981
+    },
+    {
+      "epoch": 1.046032786885246,
+      "grad_norm": 0.895736492948733,
+      "learning_rate": 0.00015129154855987766,
+      "loss": 12.7455,
+      "step": 5982
+    },
+    {
+      "epoch": 1.046207650273224,
+      "grad_norm": 0.8792843954888936,
+      "learning_rate": 0.00015127533954428486,
+      "loss": 12.6225,
+      "step": 5983
+    },
+    {
+      "epoch": 1.0463825136612022,
+      "grad_norm": 0.9220273552161367,
+      "learning_rate": 0.0001512591287007861,
+      "loss": 12.6702,
+      "step": 5984
+    },
+    {
+      "epoch": 1.0465573770491803,
+      "grad_norm": 0.9343124155995092,
+      "learning_rate": 0.00015124291602995932,
+      "loss": 12.8932,
+      "step": 5985
+    },
+    {
+      "epoch": 1.0467322404371584,
+      "grad_norm": 0.8776672857427174,
+      "learning_rate": 0.00015122670153238244,
+      "loss": 12.7881,
+      "step": 5986
+    },
+    {
+      "epoch": 1.0469071038251365,
+      "grad_norm": 0.9996109829694944,
+      "learning_rate": 0.00015121048520863354,
+      "loss": 12.8868,
+      "step": 5987
+    },
+    {
+      "epoch": 1.0470819672131149,
+      "grad_norm": 0.9591990702422303,
+      "learning_rate": 0.00015119426705929062,
+      "loss": 12.6493,
+      "step": 5988
+    },
+    {
+      "epoch": 1.047256830601093,
+      "grad_norm": 0.9476639748895526,
+      "learning_rate": 0.0001511780470849319,
+      "loss": 12.812,
+      "step": 5989
+    },
+    {
+      "epoch": 1.047431693989071,
+      "grad_norm": 1.0212977171760105,
+      "learning_rate": 0.00015116182528613557,
+      "loss": 12.6406,
+      "step": 5990
+    },
+    {
+      "epoch": 1.0476065573770492,
+      "grad_norm": 0.9874327115357278,
+      "learning_rate": 0.00015114560166348,
+      "loss": 12.5053,
+      "step": 5991
+    },
+    {
+      "epoch": 1.0477814207650273,
+      "grad_norm": 0.9569244221674033,
+      "learning_rate": 0.00015112937621754346,
+      "loss": 12.8516,
+      "step": 5992
+    },
+    {
+      "epoch": 1.0479562841530055,
+      "grad_norm": 0.9558164238674131,
+      "learning_rate": 0.00015111314894890443,
+      "loss": 12.8335,
+      "step": 5993
+    },
+    {
+      "epoch": 1.0481311475409836,
+      "grad_norm": 0.9440894281006865,
+      "learning_rate": 0.00015109691985814135,
+      "loss": 12.385,
+      "step": 5994
+    },
+    {
+      "epoch": 1.0483060109289617,
+      "grad_norm": 0.9294924778062401,
+      "learning_rate": 0.00015108068894583273,
+      "loss": 12.7152,
+      "step": 5995
+    },
+    {
+      "epoch": 1.0484808743169398,
+      "grad_norm": 1.0659495685648468,
+      "learning_rate": 0.0001510644562125573,
+      "loss": 12.8351,
+      "step": 5996
+    },
+    {
+      "epoch": 1.048655737704918,
+      "grad_norm": 0.9447836045205636,
+      "learning_rate": 0.00015104822165889359,
+      "loss": 12.7943,
+      "step": 5997
+    },
+    {
+      "epoch": 1.0488306010928963,
+      "grad_norm": 1.242085376502686,
+      "learning_rate": 0.00015103198528542048,
+      "loss": 13.0398,
+      "step": 5998
+    },
+    {
+      "epoch": 1.0490054644808744,
+      "grad_norm": 1.0385791992849787,
+      "learning_rate": 0.00015101574709271671,
+      "loss": 12.7351,
+      "step": 5999
+    },
+    {
+      "epoch": 1.0491803278688525,
+      "grad_norm": 1.0308473396753508,
+      "learning_rate": 0.00015099950708136114,
+      "loss": 12.8564,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0494863387978142,
+      "grad_norm": 0.9424417923034738,
+      "learning_rate": 0.00015098326525193274,
+      "loss": 12.621,
+      "step": 6001
+    },
+    {
+      "epoch": 1.0496612021857923,
+      "grad_norm": 1.0886920630148185,
+      "learning_rate": 0.00015096702160501048,
+      "loss": 12.814,
+      "step": 6002
+    },
+    {
+      "epoch": 1.0498360655737704,
+      "grad_norm": 1.0001079121394847,
+      "learning_rate": 0.00015095077614117348,
+      "loss": 12.8616,
+      "step": 6003
+    },
+    {
+      "epoch": 1.0500109289617487,
+      "grad_norm": 1.058298243006356,
+      "learning_rate": 0.00015093452886100087,
+      "loss": 12.811,
+      "step": 6004
+    },
+    {
+      "epoch": 1.0501857923497269,
+      "grad_norm": 0.9774497030837925,
+      "learning_rate": 0.00015091827976507175,
+      "loss": 12.7978,
+      "step": 6005
+    },
+    {
+      "epoch": 1.050360655737705,
+      "grad_norm": 1.0999132148723623,
+      "learning_rate": 0.00015090202885396548,
+      "loss": 12.9275,
+      "step": 6006
+    },
+    {
+      "epoch": 1.050535519125683,
+      "grad_norm": 0.9005357632815353,
+      "learning_rate": 0.00015088577612826134,
+      "loss": 12.6272,
+      "step": 6007
+    },
+    {
+      "epoch": 1.0507103825136612,
+      "grad_norm": 1.027114488831304,
+      "learning_rate": 0.00015086952158853878,
+      "loss": 13.0013,
+      "step": 6008
+    },
+    {
+      "epoch": 1.0508852459016393,
+      "grad_norm": 1.03317731371466,
+      "learning_rate": 0.00015085326523537715,
+      "loss": 13.0301,
+      "step": 6009
+    },
+    {
+      "epoch": 1.0510601092896175,
+      "grad_norm": 1.0191649649609007,
+      "learning_rate": 0.00015083700706935605,
+      "loss": 12.9099,
+      "step": 6010
+    },
+    {
+      "epoch": 1.0512349726775956,
+      "grad_norm": 1.0386968248837256,
+      "learning_rate": 0.00015082074709105506,
+      "loss": 12.7915,
+      "step": 6011
+    },
+    {
+      "epoch": 1.0514098360655737,
+      "grad_norm": 1.2161063073242089,
+      "learning_rate": 0.00015080448530105378,
+      "loss": 12.6949,
+      "step": 6012
+    },
+    {
+      "epoch": 1.0515846994535518,
+      "grad_norm": 0.9784423939488432,
+      "learning_rate": 0.000150788221699932,
+      "loss": 12.6665,
+      "step": 6013
+    },
+    {
+      "epoch": 1.0517595628415302,
+      "grad_norm": 0.9210074586599032,
+      "learning_rate": 0.00015077195628826942,
+      "loss": 12.9884,
+      "step": 6014
+    },
+    {
+      "epoch": 1.0519344262295083,
+      "grad_norm": 0.8691904017167088,
+      "learning_rate": 0.00015075568906664595,
+      "loss": 12.9475,
+      "step": 6015
+    },
+    {
+      "epoch": 1.0521092896174864,
+      "grad_norm": 1.0994222046791524,
+      "learning_rate": 0.00015073942003564146,
+      "loss": 12.9958,
+      "step": 6016
+    },
+    {
+      "epoch": 1.0522841530054645,
+      "grad_norm": 0.9378392631227045,
+      "learning_rate": 0.00015072314919583593,
+      "loss": 12.9026,
+      "step": 6017
+    },
+    {
+      "epoch": 1.0524590163934426,
+      "grad_norm": 1.0570162032637072,
+      "learning_rate": 0.00015070687654780937,
+      "loss": 12.8405,
+      "step": 6018
+    },
+    {
+      "epoch": 1.0526338797814208,
+      "grad_norm": 1.0493203723712519,
+      "learning_rate": 0.0001506906020921419,
+      "loss": 12.6525,
+      "step": 6019
+    },
+    {
+      "epoch": 1.0528087431693989,
+      "grad_norm": 1.1571039009560602,
+      "learning_rate": 0.00015067432582941372,
+      "loss": 13.237,
+      "step": 6020
+    },
+    {
+      "epoch": 1.052983606557377,
+      "grad_norm": 1.0383958916559262,
+      "learning_rate": 0.00015065804776020504,
+      "loss": 13.0578,
+      "step": 6021
+    },
+    {
+      "epoch": 1.0531584699453551,
+      "grad_norm": 1.0869518082957417,
+      "learning_rate": 0.0001506417678850961,
+      "loss": 12.749,
+      "step": 6022
+    },
+    {
+      "epoch": 1.0533333333333332,
+      "grad_norm": 0.9387768060862951,
+      "learning_rate": 0.00015062548620466734,
+      "loss": 12.8617,
+      "step": 6023
+    },
+    {
+      "epoch": 1.0535081967213116,
+      "grad_norm": 0.980994460404087,
+      "learning_rate": 0.00015060920271949915,
+      "loss": 12.9849,
+      "step": 6024
+    },
+    {
+      "epoch": 1.0536830601092897,
+      "grad_norm": 0.9103957319506393,
+      "learning_rate": 0.00015059291743017202,
+      "loss": 12.9221,
+      "step": 6025
+    },
+    {
+      "epoch": 1.0538579234972678,
+      "grad_norm": 1.0567719914802576,
+      "learning_rate": 0.00015057663033726646,
+      "loss": 12.6549,
+      "step": 6026
+    },
+    {
+      "epoch": 1.054032786885246,
+      "grad_norm": 1.0694662301858795,
+      "learning_rate": 0.00015056034144136313,
+      "loss": 13.0015,
+      "step": 6027
+    },
+    {
+      "epoch": 1.054207650273224,
+      "grad_norm": 1.0279478283911807,
+      "learning_rate": 0.00015054405074304273,
+      "loss": 13.0466,
+      "step": 6028
+    },
+    {
+      "epoch": 1.0543825136612022,
+      "grad_norm": 1.1004635691657012,
+      "learning_rate": 0.00015052775824288595,
+      "loss": 13.0458,
+      "step": 6029
+    },
+    {
+      "epoch": 1.0545573770491803,
+      "grad_norm": 1.0385608279938439,
+      "learning_rate": 0.00015051146394147363,
+      "loss": 13.0475,
+      "step": 6030
+    },
+    {
+      "epoch": 1.0547322404371584,
+      "grad_norm": 1.1650283363714036,
+      "learning_rate": 0.00015049516783938662,
+      "loss": 12.9624,
+      "step": 6031
+    },
+    {
+      "epoch": 1.0549071038251365,
+      "grad_norm": 1.1527217764164492,
+      "learning_rate": 0.0001504788699372059,
+      "loss": 12.8245,
+      "step": 6032
+    },
+    {
+      "epoch": 1.0550819672131149,
+      "grad_norm": 1.0804217750264378,
+      "learning_rate": 0.00015046257023551243,
+      "loss": 12.9579,
+      "step": 6033
+    },
+    {
+      "epoch": 1.055256830601093,
+      "grad_norm": 1.123504877556599,
+      "learning_rate": 0.0001504462687348873,
+      "loss": 13.1039,
+      "step": 6034
+    },
+    {
+      "epoch": 1.055431693989071,
+      "grad_norm": 1.1614632267483198,
+      "learning_rate": 0.0001504299654359116,
+      "loss": 12.9676,
+      "step": 6035
+    },
+    {
+      "epoch": 1.0556065573770492,
+      "grad_norm": 0.9627493864915239,
+      "learning_rate": 0.0001504136603391666,
+      "loss": 12.9581,
+      "step": 6036
+    },
+    {
+      "epoch": 1.0557814207650273,
+      "grad_norm": 1.093620882844446,
+      "learning_rate": 0.00015039735344523348,
+      "loss": 13.139,
+      "step": 6037
+    },
+    {
+      "epoch": 1.0559562841530055,
+      "grad_norm": 1.039609199680131,
+      "learning_rate": 0.0001503810447546936,
+      "loss": 12.9682,
+      "step": 6038
+    },
+    {
+      "epoch": 1.0561311475409836,
+      "grad_norm": 1.051708303077083,
+      "learning_rate": 0.00015036473426812834,
+      "loss": 12.9045,
+      "step": 6039
+    },
+    {
+      "epoch": 1.0563060109289617,
+      "grad_norm": 1.0402957045032155,
+      "learning_rate": 0.00015034842198611913,
+      "loss": 12.7995,
+      "step": 6040
+    },
+    {
+      "epoch": 1.0564808743169398,
+      "grad_norm": 1.0758433300415693,
+      "learning_rate": 0.00015033210790924755,
+      "loss": 13.1054,
+      "step": 6041
+    },
+    {
+      "epoch": 1.056655737704918,
+      "grad_norm": 1.0179128113956342,
+      "learning_rate": 0.0001503157920380951,
+      "loss": 13.0925,
+      "step": 6042
+    },
+    {
+      "epoch": 1.0568306010928963,
+      "grad_norm": 1.232128301523086,
+      "learning_rate": 0.00015029947437324346,
+      "loss": 12.9303,
+      "step": 6043
+    },
+    {
+      "epoch": 1.0570054644808744,
+      "grad_norm": 0.9877442483970782,
+      "learning_rate": 0.0001502831549152743,
+      "loss": 12.9145,
+      "step": 6044
+    },
+    {
+      "epoch": 1.0571803278688525,
+      "grad_norm": 1.085482691036006,
+      "learning_rate": 0.0001502668336647694,
+      "loss": 12.6581,
+      "step": 6045
+    },
+    {
+      "epoch": 1.0573551912568306,
+      "grad_norm": 0.981089179671101,
+      "learning_rate": 0.00015025051062231065,
+      "loss": 12.7373,
+      "step": 6046
+    },
+    {
+      "epoch": 1.0575300546448088,
+      "grad_norm": 1.2819601753779573,
+      "learning_rate": 0.0001502341857884799,
+      "loss": 13.0458,
+      "step": 6047
+    },
+    {
+      "epoch": 1.0577049180327869,
+      "grad_norm": 1.006811290461747,
+      "learning_rate": 0.00015021785916385913,
+      "loss": 13.1035,
+      "step": 6048
+    },
+    {
+      "epoch": 1.057879781420765,
+      "grad_norm": 1.1030242220375812,
+      "learning_rate": 0.00015020153074903037,
+      "loss": 12.8326,
+      "step": 6049
+    },
+    {
+      "epoch": 1.0580546448087431,
+      "grad_norm": 0.9890687969429875,
+      "learning_rate": 0.00015018520054457566,
+      "loss": 12.7345,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0582295081967212,
+      "grad_norm": 1.1390496965960055,
+      "learning_rate": 0.0001501688685510772,
+      "loss": 12.8592,
+      "step": 6051
+    },
+    {
+      "epoch": 1.0584043715846994,
+      "grad_norm": 1.2518109826603638,
+      "learning_rate": 0.00015015253476911717,
+      "loss": 12.9349,
+      "step": 6052
+    },
+    {
+      "epoch": 1.0585792349726777,
+      "grad_norm": 0.9663904397205664,
+      "learning_rate": 0.0001501361991992779,
+      "loss": 12.8725,
+      "step": 6053
+    },
+    {
+      "epoch": 1.0587540983606558,
+      "grad_norm": 1.4518711275678262,
+      "learning_rate": 0.0001501198618421417,
+      "loss": 12.6922,
+      "step": 6054
+    },
+    {
+      "epoch": 1.058928961748634,
+      "grad_norm": 1.122455257776936,
+      "learning_rate": 0.00015010352269829096,
+      "loss": 13.0452,
+      "step": 6055
+    },
+    {
+      "epoch": 1.059103825136612,
+      "grad_norm": 1.5931936181592403,
+      "learning_rate": 0.00015008718176830817,
+      "loss": 13.0184,
+      "step": 6056
+    },
+    {
+      "epoch": 1.0592786885245902,
+      "grad_norm": 1.0625902251020982,
+      "learning_rate": 0.0001500708390527759,
+      "loss": 12.7789,
+      "step": 6057
+    },
+    {
+      "epoch": 1.0594535519125683,
+      "grad_norm": 1.393421720137965,
+      "learning_rate": 0.00015005449455227673,
+      "loss": 12.7429,
+      "step": 6058
+    },
+    {
+      "epoch": 1.0596284153005464,
+      "grad_norm": 1.1201178097580662,
+      "learning_rate": 0.0001500381482673933,
+      "loss": 12.6696,
+      "step": 6059
+    },
+    {
+      "epoch": 1.0598032786885245,
+      "grad_norm": 1.4954717583328596,
+      "learning_rate": 0.00015002180019870837,
+      "loss": 13.0347,
+      "step": 6060
+    },
+    {
+      "epoch": 1.0599781420765026,
+      "grad_norm": 0.9816460814370129,
+      "learning_rate": 0.00015000545034680467,
+      "loss": 12.3899,
+      "step": 6061
+    },
+    {
+      "epoch": 1.0601530054644808,
+      "grad_norm": 1.1965654880583994,
+      "learning_rate": 0.0001499890987122651,
+      "loss": 12.975,
+      "step": 6062
+    },
+    {
+      "epoch": 1.060327868852459,
+      "grad_norm": 1.04047995675337,
+      "learning_rate": 0.00014997274529567257,
+      "loss": 12.91,
+      "step": 6063
+    },
+    {
+      "epoch": 1.0605027322404372,
+      "grad_norm": 1.313860881218347,
+      "learning_rate": 0.00014995639009761006,
+      "loss": 12.6262,
+      "step": 6064
+    },
+    {
+      "epoch": 1.0606775956284153,
+      "grad_norm": 0.9714490649515118,
+      "learning_rate": 0.00014994003311866058,
+      "loss": 12.7236,
+      "step": 6065
+    },
+    {
+      "epoch": 1.0608524590163935,
+      "grad_norm": 1.105040227751914,
+      "learning_rate": 0.0001499236743594073,
+      "loss": 13.1272,
+      "step": 6066
+    },
+    {
+      "epoch": 1.0610273224043716,
+      "grad_norm": 0.9783731789646218,
+      "learning_rate": 0.00014990731382043336,
+      "loss": 12.9208,
+      "step": 6067
+    },
+    {
+      "epoch": 1.0612021857923497,
+      "grad_norm": 1.0733969050900545,
+      "learning_rate": 0.000149890951502322,
+      "loss": 12.9764,
+      "step": 6068
+    },
+    {
+      "epoch": 1.0613770491803278,
+      "grad_norm": 1.084431883296517,
+      "learning_rate": 0.0001498745874056565,
+      "loss": 12.9688,
+      "step": 6069
+    },
+    {
+      "epoch": 1.061551912568306,
+      "grad_norm": 1.0333482958610118,
+      "learning_rate": 0.0001498582215310202,
+      "loss": 12.933,
+      "step": 6070
+    },
+    {
+      "epoch": 1.061726775956284,
+      "grad_norm": 1.0649230830808145,
+      "learning_rate": 0.0001498418538789966,
+      "loss": 12.9565,
+      "step": 6071
+    },
+    {
+      "epoch": 1.0619016393442622,
+      "grad_norm": 1.1783848912533579,
+      "learning_rate": 0.0001498254844501691,
+      "loss": 12.7799,
+      "step": 6072
+    },
+    {
+      "epoch": 1.0620765027322405,
+      "grad_norm": 1.0783038572499637,
+      "learning_rate": 0.00014980911324512136,
+      "loss": 12.752,
+      "step": 6073
+    },
+    {
+      "epoch": 1.0622513661202186,
+      "grad_norm": 1.0597846907486763,
+      "learning_rate": 0.00014979274026443687,
+      "loss": 12.8678,
+      "step": 6074
+    },
+    {
+      "epoch": 1.0624262295081968,
+      "grad_norm": 1.0569329288024063,
+      "learning_rate": 0.00014977636550869938,
+      "loss": 12.9342,
+      "step": 6075
+    },
+    {
+      "epoch": 1.0626010928961749,
+      "grad_norm": 1.0351598911947753,
+      "learning_rate": 0.00014975998897849262,
+      "loss": 13.0293,
+      "step": 6076
+    },
+    {
+      "epoch": 1.062775956284153,
+      "grad_norm": 0.9307151820874094,
+      "learning_rate": 0.0001497436106744004,
+      "loss": 12.8825,
+      "step": 6077
+    },
+    {
+      "epoch": 1.0629508196721311,
+      "grad_norm": 1.1039063883116056,
+      "learning_rate": 0.00014972723059700655,
+      "loss": 12.7658,
+      "step": 6078
+    },
+    {
+      "epoch": 1.0631256830601092,
+      "grad_norm": 0.8944812565803056,
+      "learning_rate": 0.00014971084874689506,
+      "loss": 12.8533,
+      "step": 6079
+    },
+    {
+      "epoch": 1.0633005464480874,
+      "grad_norm": 1.0780331858121923,
+      "learning_rate": 0.0001496944651246499,
+      "loss": 12.9654,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0634754098360655,
+      "grad_norm": 0.9770373162373954,
+      "learning_rate": 0.00014967807973085507,
+      "loss": 12.7056,
+      "step": 6081
+    },
+    {
+      "epoch": 1.0636502732240438,
+      "grad_norm": 1.1018748052389913,
+      "learning_rate": 0.0001496616925660948,
+      "loss": 12.9621,
+      "step": 6082
+    },
+    {
+      "epoch": 1.063825136612022,
+      "grad_norm": 0.9420060686970838,
+      "learning_rate": 0.00014964530363095318,
+      "loss": 12.8944,
+      "step": 6083
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 0.9681799913825933,
+      "learning_rate": 0.00014962891292601447,
+      "loss": 12.6947,
+      "step": 6084
+    },
+    {
+      "epoch": 1.0641748633879782,
+      "grad_norm": 0.9481816055231097,
+      "learning_rate": 0.00014961252045186306,
+      "loss": 12.9401,
+      "step": 6085
+    },
+    {
+      "epoch": 1.0643497267759563,
+      "grad_norm": 1.0999723916577568,
+      "learning_rate": 0.00014959612620908322,
+      "loss": 12.9226,
+      "step": 6086
+    },
+    {
+      "epoch": 1.0645245901639344,
+      "grad_norm": 0.9643619577793858,
+      "learning_rate": 0.00014957973019825944,
+      "loss": 12.6197,
+      "step": 6087
+    },
+    {
+      "epoch": 1.0646994535519125,
+      "grad_norm": 0.942258704423735,
+      "learning_rate": 0.00014956333241997622,
+      "loss": 12.8789,
+      "step": 6088
+    },
+    {
+      "epoch": 1.0648743169398907,
+      "grad_norm": 0.9659810173924988,
+      "learning_rate": 0.00014954693287481808,
+      "loss": 12.8672,
+      "step": 6089
+    },
+    {
+      "epoch": 1.0650491803278688,
+      "grad_norm": 0.9422015985660918,
+      "learning_rate": 0.00014953053156336967,
+      "loss": 13.034,
+      "step": 6090
+    },
+    {
+      "epoch": 1.065224043715847,
+      "grad_norm": 1.1209982357911124,
+      "learning_rate": 0.00014951412848621568,
+      "loss": 12.8626,
+      "step": 6091
+    },
+    {
+      "epoch": 1.0653989071038252,
+      "grad_norm": 1.0429352098176539,
+      "learning_rate": 0.0001494977236439409,
+      "loss": 12.8057,
+      "step": 6092
+    },
+    {
+      "epoch": 1.0655737704918034,
+      "grad_norm": 0.8902644321432125,
+      "learning_rate": 0.00014948131703713006,
+      "loss": 12.9812,
+      "step": 6093
+    },
+    {
+      "epoch": 1.0657486338797815,
+      "grad_norm": 1.0336161783815072,
+      "learning_rate": 0.00014946490866636812,
+      "loss": 13.1031,
+      "step": 6094
+    },
+    {
+      "epoch": 1.0659234972677596,
+      "grad_norm": 1.0899490947852968,
+      "learning_rate": 0.00014944849853223993,
+      "loss": 12.9656,
+      "step": 6095
+    },
+    {
+      "epoch": 1.0660983606557377,
+      "grad_norm": 0.9033684832828107,
+      "learning_rate": 0.00014943208663533057,
+      "loss": 12.605,
+      "step": 6096
+    },
+    {
+      "epoch": 1.0662732240437158,
+      "grad_norm": 0.9869117732125442,
+      "learning_rate": 0.00014941567297622509,
+      "loss": 12.7446,
+      "step": 6097
+    },
+    {
+      "epoch": 1.066448087431694,
+      "grad_norm": 1.0135421402957598,
+      "learning_rate": 0.0001493992575555086,
+      "loss": 12.8632,
+      "step": 6098
+    },
+    {
+      "epoch": 1.066622950819672,
+      "grad_norm": 0.9828502256078951,
+      "learning_rate": 0.00014938284037376628,
+      "loss": 13.1784,
+      "step": 6099
+    },
+    {
+      "epoch": 1.0667978142076502,
+      "grad_norm": 0.9735236948737991,
+      "learning_rate": 0.0001493664214315834,
+      "loss": 13.0141,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0669726775956283,
+      "grad_norm": 0.9226852787328303,
+      "learning_rate": 0.00014935000072954525,
+      "loss": 12.5356,
+      "step": 6101
+    },
+    {
+      "epoch": 1.0671475409836066,
+      "grad_norm": 0.9445958142044685,
+      "learning_rate": 0.00014933357826823728,
+      "loss": 12.7995,
+      "step": 6102
+    },
+    {
+      "epoch": 1.0673224043715848,
+      "grad_norm": 0.8915066417741323,
+      "learning_rate": 0.00014931715404824486,
+      "loss": 12.8435,
+      "step": 6103
+    },
+    {
+      "epoch": 1.0674972677595629,
+      "grad_norm": 0.9793171431330728,
+      "learning_rate": 0.0001493007280701535,
+      "loss": 13.0439,
+      "step": 6104
+    },
+    {
+      "epoch": 1.067672131147541,
+      "grad_norm": 0.9530996045062179,
+      "learning_rate": 0.00014928430033454886,
+      "loss": 13.0039,
+      "step": 6105
+    },
+    {
+      "epoch": 1.0678469945355191,
+      "grad_norm": 0.9014927584868717,
+      "learning_rate": 0.0001492678708420164,
+      "loss": 12.635,
+      "step": 6106
+    },
+    {
+      "epoch": 1.0680218579234972,
+      "grad_norm": 0.8906875669343959,
+      "learning_rate": 0.00014925143959314198,
+      "loss": 12.8257,
+      "step": 6107
+    },
+    {
+      "epoch": 1.0681967213114754,
+      "grad_norm": 1.020536130462321,
+      "learning_rate": 0.00014923500658851126,
+      "loss": 12.9567,
+      "step": 6108
+    },
+    {
+      "epoch": 1.0683715846994535,
+      "grad_norm": 1.080333704697687,
+      "learning_rate": 0.00014921857182871007,
+      "loss": 12.9407,
+      "step": 6109
+    },
+    {
+      "epoch": 1.0685464480874316,
+      "grad_norm": 1.154088657881979,
+      "learning_rate": 0.0001492021353143243,
+      "loss": 12.8017,
+      "step": 6110
+    },
+    {
+      "epoch": 1.06872131147541,
+      "grad_norm": 0.9551138968567795,
+      "learning_rate": 0.0001491856970459399,
+      "loss": 13.0097,
+      "step": 6111
+    },
+    {
+      "epoch": 1.068896174863388,
+      "grad_norm": 1.0818227573890589,
+      "learning_rate": 0.00014916925702414285,
+      "loss": 12.796,
+      "step": 6112
+    },
+    {
+      "epoch": 1.0690710382513662,
+      "grad_norm": 1.068104435501901,
+      "learning_rate": 0.00014915281524951928,
+      "loss": 13.0147,
+      "step": 6113
+    },
+    {
+      "epoch": 1.0692459016393443,
+      "grad_norm": 1.0504169625590478,
+      "learning_rate": 0.00014913637172265522,
+      "loss": 12.9267,
+      "step": 6114
+    },
+    {
+      "epoch": 1.0694207650273224,
+      "grad_norm": 1.2312830611784227,
+      "learning_rate": 0.00014911992644413698,
+      "loss": 12.8154,
+      "step": 6115
+    },
+    {
+      "epoch": 1.0695956284153005,
+      "grad_norm": 1.0833395429194792,
+      "learning_rate": 0.0001491034794145507,
+      "loss": 12.989,
+      "step": 6116
+    },
+    {
+      "epoch": 1.0697704918032787,
+      "grad_norm": 1.2256149002191432,
+      "learning_rate": 0.00014908703063448277,
+      "loss": 13.0776,
+      "step": 6117
+    },
+    {
+      "epoch": 1.0699453551912568,
+      "grad_norm": 1.0319630441270744,
+      "learning_rate": 0.00014907058010451955,
+      "loss": 12.592,
+      "step": 6118
+    },
+    {
+      "epoch": 1.070120218579235,
+      "grad_norm": 0.9800230287816085,
+      "learning_rate": 0.00014905412782524748,
+      "loss": 12.6097,
+      "step": 6119
+    },
+    {
+      "epoch": 1.070295081967213,
+      "grad_norm": 0.985468331654974,
+      "learning_rate": 0.00014903767379725308,
+      "loss": 12.6834,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0704699453551914,
+      "grad_norm": 1.0685541411282056,
+      "learning_rate": 0.00014902121802112288,
+      "loss": 12.9151,
+      "step": 6121
+    },
+    {
+      "epoch": 1.0706448087431695,
+      "grad_norm": 1.195272927518154,
+      "learning_rate": 0.00014900476049744353,
+      "loss": 13.1284,
+      "step": 6122
+    },
+    {
+      "epoch": 1.0708196721311476,
+      "grad_norm": 1.115304936714522,
+      "learning_rate": 0.00014898830122680174,
+      "loss": 12.7367,
+      "step": 6123
+    },
+    {
+      "epoch": 1.0709945355191257,
+      "grad_norm": 1.0323435867089568,
+      "learning_rate": 0.00014897184020978426,
+      "loss": 12.7283,
+      "step": 6124
+    },
+    {
+      "epoch": 1.0711693989071038,
+      "grad_norm": 1.336917700427458,
+      "learning_rate": 0.00014895537744697787,
+      "loss": 12.7989,
+      "step": 6125
+    },
+    {
+      "epoch": 1.071344262295082,
+      "grad_norm": 0.9405115109371951,
+      "learning_rate": 0.00014893891293896947,
+      "loss": 12.5353,
+      "step": 6126
+    },
+    {
+      "epoch": 1.07151912568306,
+      "grad_norm": 1.377432234662197,
+      "learning_rate": 0.00014892244668634601,
+      "loss": 13.0454,
+      "step": 6127
+    },
+    {
+      "epoch": 1.0716939890710382,
+      "grad_norm": 1.151041987123703,
+      "learning_rate": 0.00014890597868969448,
+      "loss": 12.8257,
+      "step": 6128
+    },
+    {
+      "epoch": 1.0718688524590163,
+      "grad_norm": 1.3027205004365896,
+      "learning_rate": 0.00014888950894960198,
+      "loss": 12.9346,
+      "step": 6129
+    },
+    {
+      "epoch": 1.0720437158469944,
+      "grad_norm": 1.154327848641509,
+      "learning_rate": 0.00014887303746665556,
+      "loss": 13.1544,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0722185792349728,
+      "grad_norm": 1.2848749719270907,
+      "learning_rate": 0.0001488565642414425,
+      "loss": 12.9054,
+      "step": 6131
+    },
+    {
+      "epoch": 1.0723934426229509,
+      "grad_norm": 1.181082579608363,
+      "learning_rate": 0.00014884008927455,
+      "loss": 13.0341,
+      "step": 6132
+    },
+    {
+      "epoch": 1.072568306010929,
+      "grad_norm": 1.378898095722529,
+      "learning_rate": 0.00014882361256656537,
+      "loss": 12.7483,
+      "step": 6133
+    },
+    {
+      "epoch": 1.0727431693989071,
+      "grad_norm": 1.1555066394075084,
+      "learning_rate": 0.00014880713411807598,
+      "loss": 13.1155,
+      "step": 6134
+    },
+    {
+      "epoch": 1.0729180327868852,
+      "grad_norm": 1.0804404415349447,
+      "learning_rate": 0.0001487906539296693,
+      "loss": 12.7613,
+      "step": 6135
+    },
+    {
+      "epoch": 1.0730928961748634,
+      "grad_norm": 1.007465214824562,
+      "learning_rate": 0.00014877417200193283,
+      "loss": 13.0155,
+      "step": 6136
+    },
+    {
+      "epoch": 1.0732677595628415,
+      "grad_norm": 1.176409983510257,
+      "learning_rate": 0.0001487576883354541,
+      "loss": 12.7945,
+      "step": 6137
+    },
+    {
+      "epoch": 1.0734426229508196,
+      "grad_norm": 0.9792180296999246,
+      "learning_rate": 0.00014874120293082075,
+      "loss": 12.924,
+      "step": 6138
+    },
+    {
+      "epoch": 1.0736174863387977,
+      "grad_norm": 0.9358378636724238,
+      "learning_rate": 0.00014872471578862047,
+      "loss": 12.6821,
+      "step": 6139
+    },
+    {
+      "epoch": 1.073792349726776,
+      "grad_norm": 0.9509694722719203,
+      "learning_rate": 0.00014870822690944097,
+      "loss": 13.0074,
+      "step": 6140
+    },
+    {
+      "epoch": 1.0739672131147542,
+      "grad_norm": 1.2395818343252931,
+      "learning_rate": 0.00014869173629387012,
+      "loss": 12.9003,
+      "step": 6141
+    },
+    {
+      "epoch": 1.0741420765027323,
+      "grad_norm": 0.9911494555957945,
+      "learning_rate": 0.00014867524394249576,
+      "loss": 12.9083,
+      "step": 6142
+    },
+    {
+      "epoch": 1.0743169398907104,
+      "grad_norm": 0.9658300715511031,
+      "learning_rate": 0.00014865874985590583,
+      "loss": 12.7296,
+      "step": 6143
+    },
+    {
+      "epoch": 1.0744918032786885,
+      "grad_norm": 1.082505962308745,
+      "learning_rate": 0.00014864225403468827,
+      "loss": 13.0142,
+      "step": 6144
+    },
+    {
+      "epoch": 1.0746666666666667,
+      "grad_norm": 0.9530343085360637,
+      "learning_rate": 0.00014862575647943123,
+      "loss": 13.003,
+      "step": 6145
+    },
+    {
+      "epoch": 1.0748415300546448,
+      "grad_norm": 0.971633649745438,
+      "learning_rate": 0.00014860925719072275,
+      "loss": 12.6248,
+      "step": 6146
+    },
+    {
+      "epoch": 1.075016393442623,
+      "grad_norm": 0.9608423069120946,
+      "learning_rate": 0.0001485927561691511,
+      "loss": 13.0935,
+      "step": 6147
+    },
+    {
+      "epoch": 1.075191256830601,
+      "grad_norm": 0.9202066204177779,
+      "learning_rate": 0.00014857625341530443,
+      "loss": 12.7492,
+      "step": 6148
+    },
+    {
+      "epoch": 1.0753661202185791,
+      "grad_norm": 0.9722117826946632,
+      "learning_rate": 0.0001485597489297711,
+      "loss": 12.8472,
+      "step": 6149
+    },
+    {
+      "epoch": 1.0755409836065575,
+      "grad_norm": 0.983897359463761,
+      "learning_rate": 0.00014854324271313947,
+      "loss": 12.9082,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0757158469945356,
+      "grad_norm": 0.9204206572513022,
+      "learning_rate": 0.0001485267347659979,
+      "loss": 12.9336,
+      "step": 6151
+    },
+    {
+      "epoch": 1.0758907103825137,
+      "grad_norm": 1.0340170667675084,
+      "learning_rate": 0.00014851022508893498,
+      "loss": 12.9373,
+      "step": 6152
+    },
+    {
+      "epoch": 1.0760655737704918,
+      "grad_norm": 0.9892007017720458,
+      "learning_rate": 0.0001484937136825392,
+      "loss": 12.9774,
+      "step": 6153
+    },
+    {
+      "epoch": 1.07624043715847,
+      "grad_norm": 1.046076428614335,
+      "learning_rate": 0.00014847720054739918,
+      "loss": 12.777,
+      "step": 6154
+    },
+    {
+      "epoch": 1.076415300546448,
+      "grad_norm": 1.0337595941971989,
+      "learning_rate": 0.0001484606856841036,
+      "loss": 12.8311,
+      "step": 6155
+    },
+    {
+      "epoch": 1.0765901639344262,
+      "grad_norm": 0.9924168930722331,
+      "learning_rate": 0.0001484441690932412,
+      "loss": 13.0158,
+      "step": 6156
+    },
+    {
+      "epoch": 1.0767650273224043,
+      "grad_norm": 0.9774198550359688,
+      "learning_rate": 0.00014842765077540077,
+      "loss": 12.9466,
+      "step": 6157
+    },
+    {
+      "epoch": 1.0769398907103824,
+      "grad_norm": 0.9469104763206976,
+      "learning_rate": 0.00014841113073117115,
+      "loss": 12.6837,
+      "step": 6158
+    },
+    {
+      "epoch": 1.0771147540983605,
+      "grad_norm": 0.9349992911178122,
+      "learning_rate": 0.00014839460896114132,
+      "loss": 13.0405,
+      "step": 6159
+    },
+    {
+      "epoch": 1.077289617486339,
+      "grad_norm": 0.977984879507208,
+      "learning_rate": 0.0001483780854659002,
+      "loss": 12.8382,
+      "step": 6160
+    },
+    {
+      "epoch": 1.077464480874317,
+      "grad_norm": 0.9713118569191158,
+      "learning_rate": 0.00014836156024603686,
+      "loss": 12.9058,
+      "step": 6161
+    },
+    {
+      "epoch": 1.0776393442622951,
+      "grad_norm": 1.0144804613143459,
+      "learning_rate": 0.00014834503330214043,
+      "loss": 13.2054,
+      "step": 6162
+    },
+    {
+      "epoch": 1.0778142076502732,
+      "grad_norm": 0.9318499177506817,
+      "learning_rate": 0.00014832850463480002,
+      "loss": 13.0567,
+      "step": 6163
+    },
+    {
+      "epoch": 1.0779890710382514,
+      "grad_norm": 1.0449224844270253,
+      "learning_rate": 0.00014831197424460487,
+      "loss": 13.075,
+      "step": 6164
+    },
+    {
+      "epoch": 1.0781639344262295,
+      "grad_norm": 0.930849345105244,
+      "learning_rate": 0.0001482954421321443,
+      "loss": 13.0181,
+      "step": 6165
+    },
+    {
+      "epoch": 1.0783387978142076,
+      "grad_norm": 1.128775391932213,
+      "learning_rate": 0.00014827890829800767,
+      "loss": 12.9855,
+      "step": 6166
+    },
+    {
+      "epoch": 1.0785136612021857,
+      "grad_norm": 1.1816916452067165,
+      "learning_rate": 0.00014826237274278436,
+      "loss": 12.836,
+      "step": 6167
+    },
+    {
+      "epoch": 1.0786885245901638,
+      "grad_norm": 1.0420013758448976,
+      "learning_rate": 0.00014824583546706383,
+      "loss": 13.0986,
+      "step": 6168
+    },
+    {
+      "epoch": 1.0788633879781422,
+      "grad_norm": 0.9853214527957307,
+      "learning_rate": 0.00014822929647143566,
+      "loss": 12.9766,
+      "step": 6169
+    },
+    {
+      "epoch": 1.0790382513661203,
+      "grad_norm": 1.056887370882715,
+      "learning_rate": 0.0001482127557564894,
+      "loss": 12.8045,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0792131147540984,
+      "grad_norm": 1.0371648132319042,
+      "learning_rate": 0.0001481962133228147,
+      "loss": 13.0606,
+      "step": 6171
+    },
+    {
+      "epoch": 1.0793879781420765,
+      "grad_norm": 1.0825065749640754,
+      "learning_rate": 0.00014817966917100134,
+      "loss": 13.0564,
+      "step": 6172
+    },
+    {
+      "epoch": 1.0795628415300547,
+      "grad_norm": 1.008735910173053,
+      "learning_rate": 0.0001481631233016391,
+      "loss": 12.884,
+      "step": 6173
+    },
+    {
+      "epoch": 1.0797377049180328,
+      "grad_norm": 0.8603888378238541,
+      "learning_rate": 0.00014814657571531772,
+      "loss": 12.6507,
+      "step": 6174
+    },
+    {
+      "epoch": 1.079912568306011,
+      "grad_norm": 0.9781977016352021,
+      "learning_rate": 0.00014813002641262722,
+      "loss": 13.0,
+      "step": 6175
+    },
+    {
+      "epoch": 1.080087431693989,
+      "grad_norm": 0.9348864449706504,
+      "learning_rate": 0.00014811347539415748,
+      "loss": 12.8917,
+      "step": 6176
+    },
+    {
+      "epoch": 1.0802622950819671,
+      "grad_norm": 0.9420801014640124,
+      "learning_rate": 0.00014809692266049858,
+      "loss": 12.8496,
+      "step": 6177
+    },
+    {
+      "epoch": 1.0804371584699453,
+      "grad_norm": 0.9719495829542942,
+      "learning_rate": 0.00014808036821224057,
+      "loss": 12.9468,
+      "step": 6178
+    },
+    {
+      "epoch": 1.0806120218579236,
+      "grad_norm": 0.9272893654377178,
+      "learning_rate": 0.0001480638120499736,
+      "loss": 12.68,
+      "step": 6179
+    },
+    {
+      "epoch": 1.0807868852459017,
+      "grad_norm": 0.916741857559261,
+      "learning_rate": 0.00014804725417428787,
+      "loss": 12.8841,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0809617486338798,
+      "grad_norm": 0.9802121072854125,
+      "learning_rate": 0.0001480306945857737,
+      "loss": 13.103,
+      "step": 6181
+    },
+    {
+      "epoch": 1.081136612021858,
+      "grad_norm": 0.937282679662265,
+      "learning_rate": 0.0001480141332850214,
+      "loss": 12.7479,
+      "step": 6182
+    },
+    {
+      "epoch": 1.081311475409836,
+      "grad_norm": 1.022990481793885,
+      "learning_rate": 0.0001479975702726213,
+      "loss": 12.8758,
+      "step": 6183
+    },
+    {
+      "epoch": 1.0814863387978142,
+      "grad_norm": 1.0452329169754662,
+      "learning_rate": 0.00014798100554916393,
+      "loss": 13.0177,
+      "step": 6184
+    },
+    {
+      "epoch": 1.0816612021857923,
+      "grad_norm": 0.9334871490257489,
+      "learning_rate": 0.00014796443911523977,
+      "loss": 12.7505,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0818360655737704,
+      "grad_norm": 1.0264151671870507,
+      "learning_rate": 0.00014794787097143942,
+      "loss": 12.9823,
+      "step": 6186
+    },
+    {
+      "epoch": 1.0820109289617486,
+      "grad_norm": 1.0295772018704905,
+      "learning_rate": 0.00014793130111835344,
+      "loss": 12.999,
+      "step": 6187
+    },
+    {
+      "epoch": 1.0821857923497267,
+      "grad_norm": 1.018071888165355,
+      "learning_rate": 0.00014791472955657262,
+      "loss": 12.9607,
+      "step": 6188
+    },
+    {
+      "epoch": 1.082360655737705,
+      "grad_norm": 1.0346884779684378,
+      "learning_rate": 0.00014789815628668765,
+      "loss": 12.9482,
+      "step": 6189
+    },
+    {
+      "epoch": 1.0825355191256831,
+      "grad_norm": 1.127203636187896,
+      "learning_rate": 0.00014788158130928935,
+      "loss": 12.9467,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0827103825136613,
+      "grad_norm": 1.0490786674840173,
+      "learning_rate": 0.00014786500462496863,
+      "loss": 12.8176,
+      "step": 6191
+    },
+    {
+      "epoch": 1.0828852459016394,
+      "grad_norm": 1.1569983557765393,
+      "learning_rate": 0.00014784842623431646,
+      "loss": 13.0771,
+      "step": 6192
+    },
+    {
+      "epoch": 1.0830601092896175,
+      "grad_norm": 0.9062083786418332,
+      "learning_rate": 0.00014783184613792377,
+      "loss": 13.0031,
+      "step": 6193
+    },
+    {
+      "epoch": 1.0832349726775956,
+      "grad_norm": 1.0789263115301897,
+      "learning_rate": 0.00014781526433638167,
+      "loss": 12.8455,
+      "step": 6194
+    },
+    {
+      "epoch": 1.0834098360655737,
+      "grad_norm": 1.0288402380649737,
+      "learning_rate": 0.00014779868083028124,
+      "loss": 12.952,
+      "step": 6195
+    },
+    {
+      "epoch": 1.0835846994535518,
+      "grad_norm": 0.9654803593670325,
+      "learning_rate": 0.0001477820956202137,
+      "loss": 12.6033,
+      "step": 6196
+    },
+    {
+      "epoch": 1.08375956284153,
+      "grad_norm": 0.9899457866781652,
+      "learning_rate": 0.0001477655087067703,
+      "loss": 13.1329,
+      "step": 6197
+    },
+    {
+      "epoch": 1.0839344262295083,
+      "grad_norm": 0.9389860969444443,
+      "learning_rate": 0.00014774892009054226,
+      "loss": 12.9747,
+      "step": 6198
+    },
+    {
+      "epoch": 1.0841092896174864,
+      "grad_norm": 0.8895410967817159,
+      "learning_rate": 0.00014773232977212104,
+      "loss": 12.7516,
+      "step": 6199
+    },
+    {
+      "epoch": 1.0842841530054645,
+      "grad_norm": 1.0154595238737019,
+      "learning_rate": 0.00014771573775209802,
+      "loss": 12.7854,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0844590163934427,
+      "grad_norm": 1.0750761871196757,
+      "learning_rate": 0.00014769914403106471,
+      "loss": 12.6843,
+      "step": 6201
+    },
+    {
+      "epoch": 1.0846338797814208,
+      "grad_norm": 0.9727102305161532,
+      "learning_rate": 0.00014768254860961265,
+      "loss": 12.8725,
+      "step": 6202
+    },
+    {
+      "epoch": 1.084808743169399,
+      "grad_norm": 0.9422160060262535,
+      "learning_rate": 0.00014766595148833344,
+      "loss": 12.9991,
+      "step": 6203
+    },
+    {
+      "epoch": 1.084983606557377,
+      "grad_norm": 0.9416595221014893,
+      "learning_rate": 0.00014764935266781877,
+      "loss": 12.7515,
+      "step": 6204
+    },
+    {
+      "epoch": 1.0851584699453551,
+      "grad_norm": 0.9956531466662158,
+      "learning_rate": 0.00014763275214866033,
+      "loss": 12.9723,
+      "step": 6205
+    },
+    {
+      "epoch": 1.0853333333333333,
+      "grad_norm": 1.0754140331481776,
+      "learning_rate": 0.00014761614993144992,
+      "loss": 12.9261,
+      "step": 6206
+    },
+    {
+      "epoch": 1.0855081967213114,
+      "grad_norm": 0.9601479585758789,
+      "learning_rate": 0.00014759954601677945,
+      "loss": 12.8479,
+      "step": 6207
+    },
+    {
+      "epoch": 1.0856830601092897,
+      "grad_norm": 0.9567527660730798,
+      "learning_rate": 0.0001475829404052407,
+      "loss": 12.692,
+      "step": 6208
+    },
+    {
+      "epoch": 1.0858579234972678,
+      "grad_norm": 1.1672227765708165,
+      "learning_rate": 0.00014756633309742577,
+      "loss": 13.067,
+      "step": 6209
+    },
+    {
+      "epoch": 1.086032786885246,
+      "grad_norm": 1.0325923177859107,
+      "learning_rate": 0.0001475497240939266,
+      "loss": 12.6326,
+      "step": 6210
+    },
+    {
+      "epoch": 1.086207650273224,
+      "grad_norm": 1.1393212610782077,
+      "learning_rate": 0.00014753311339533534,
+      "loss": 13.0352,
+      "step": 6211
+    },
+    {
+      "epoch": 1.0863825136612022,
+      "grad_norm": 1.166916436742792,
+      "learning_rate": 0.00014751650100224416,
+      "loss": 12.9924,
+      "step": 6212
+    },
+    {
+      "epoch": 1.0865573770491803,
+      "grad_norm": 0.9396944666057209,
+      "learning_rate": 0.0001474998869152452,
+      "loss": 12.4169,
+      "step": 6213
+    },
+    {
+      "epoch": 1.0867322404371584,
+      "grad_norm": 0.9698878843080225,
+      "learning_rate": 0.00014748327113493082,
+      "loss": 12.6662,
+      "step": 6214
+    },
+    {
+      "epoch": 1.0869071038251366,
+      "grad_norm": 1.0223868159277263,
+      "learning_rate": 0.00014746665366189324,
+      "loss": 12.9087,
+      "step": 6215
+    },
+    {
+      "epoch": 1.0870819672131147,
+      "grad_norm": 1.0197670044601244,
+      "learning_rate": 0.00014745003449672493,
+      "loss": 12.8885,
+      "step": 6216
+    },
+    {
+      "epoch": 1.0872568306010928,
+      "grad_norm": 1.0366486395252343,
+      "learning_rate": 0.00014743341364001832,
+      "loss": 12.8135,
+      "step": 6217
+    },
+    {
+      "epoch": 1.0874316939890711,
+      "grad_norm": 1.045474857516445,
+      "learning_rate": 0.00014741679109236597,
+      "loss": 12.776,
+      "step": 6218
+    },
+    {
+      "epoch": 1.0876065573770493,
+      "grad_norm": 0.9000550002296378,
+      "learning_rate": 0.00014740016685436038,
+      "loss": 12.8341,
+      "step": 6219
+    },
+    {
+      "epoch": 1.0877814207650274,
+      "grad_norm": 0.9658809734256046,
+      "learning_rate": 0.00014738354092659424,
+      "loss": 12.8941,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0879562841530055,
+      "grad_norm": 0.9359407584509806,
+      "learning_rate": 0.00014736691330966022,
+      "loss": 13.0177,
+      "step": 6221
+    },
+    {
+      "epoch": 1.0881311475409836,
+      "grad_norm": 1.0128515183164557,
+      "learning_rate": 0.00014735028400415112,
+      "loss": 12.8298,
+      "step": 6222
+    },
+    {
+      "epoch": 1.0883060109289617,
+      "grad_norm": 0.9272134480620067,
+      "learning_rate": 0.00014733365301065969,
+      "loss": 12.8021,
+      "step": 6223
+    },
+    {
+      "epoch": 1.0884808743169399,
+      "grad_norm": 0.9490670974190152,
+      "learning_rate": 0.00014731702032977882,
+      "loss": 12.9306,
+      "step": 6224
+    },
+    {
+      "epoch": 1.088655737704918,
+      "grad_norm": 0.8779258053214238,
+      "learning_rate": 0.00014730038596210147,
+      "loss": 12.9322,
+      "step": 6225
+    },
+    {
+      "epoch": 1.088830601092896,
+      "grad_norm": 0.9724857305657879,
+      "learning_rate": 0.00014728374990822062,
+      "loss": 12.666,
+      "step": 6226
+    },
+    {
+      "epoch": 1.0890054644808744,
+      "grad_norm": 0.8885814584772828,
+      "learning_rate": 0.00014726711216872935,
+      "loss": 12.3734,
+      "step": 6227
+    },
+    {
+      "epoch": 1.0891803278688525,
+      "grad_norm": 0.9553632457172897,
+      "learning_rate": 0.00014725047274422075,
+      "loss": 12.8196,
+      "step": 6228
+    },
+    {
+      "epoch": 1.0893551912568307,
+      "grad_norm": 0.9605891488428634,
+      "learning_rate": 0.000147233831635288,
+      "loss": 12.9386,
+      "step": 6229
+    },
+    {
+      "epoch": 1.0895300546448088,
+      "grad_norm": 0.8993247912435723,
+      "learning_rate": 0.00014721718884252432,
+      "loss": 12.7064,
+      "step": 6230
+    },
+    {
+      "epoch": 1.089704918032787,
+      "grad_norm": 1.0544854333444287,
+      "learning_rate": 0.00014720054436652305,
+      "loss": 12.9889,
+      "step": 6231
+    },
+    {
+      "epoch": 1.089879781420765,
+      "grad_norm": 0.8675187566094088,
+      "learning_rate": 0.00014718389820787752,
+      "loss": 12.8884,
+      "step": 6232
+    },
+    {
+      "epoch": 1.0900546448087431,
+      "grad_norm": 0.9701883686683893,
+      "learning_rate": 0.00014716725036718117,
+      "loss": 12.6008,
+      "step": 6233
+    },
+    {
+      "epoch": 1.0902295081967213,
+      "grad_norm": 0.9692991983717634,
+      "learning_rate": 0.0001471506008450274,
+      "loss": 12.866,
+      "step": 6234
+    },
+    {
+      "epoch": 1.0904043715846994,
+      "grad_norm": 1.0295623223932253,
+      "learning_rate": 0.0001471339496420098,
+      "loss": 12.7207,
+      "step": 6235
+    },
+    {
+      "epoch": 1.0905792349726775,
+      "grad_norm": 0.9580080687246624,
+      "learning_rate": 0.000147117296758722,
+      "loss": 12.8462,
+      "step": 6236
+    },
+    {
+      "epoch": 1.0907540983606556,
+      "grad_norm": 1.1050251040389028,
+      "learning_rate": 0.00014710064219575764,
+      "loss": 12.8191,
+      "step": 6237
+    },
+    {
+      "epoch": 1.090928961748634,
+      "grad_norm": 0.9325564711217481,
+      "learning_rate": 0.00014708398595371036,
+      "loss": 12.5275,
+      "step": 6238
+    },
+    {
+      "epoch": 1.091103825136612,
+      "grad_norm": 1.1483637504373307,
+      "learning_rate": 0.00014706732803317402,
+      "loss": 13.0926,
+      "step": 6239
+    },
+    {
+      "epoch": 1.0912786885245902,
+      "grad_norm": 0.9614029119914083,
+      "learning_rate": 0.00014705066843474245,
+      "loss": 12.8544,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0914535519125683,
+      "grad_norm": 1.237099326925253,
+      "learning_rate": 0.0001470340071590095,
+      "loss": 12.8598,
+      "step": 6241
+    },
+    {
+      "epoch": 1.0916284153005464,
+      "grad_norm": 1.0258546707551828,
+      "learning_rate": 0.00014701734420656914,
+      "loss": 12.9359,
+      "step": 6242
+    },
+    {
+      "epoch": 1.0918032786885246,
+      "grad_norm": 0.9952502643185644,
+      "learning_rate": 0.0001470006795780154,
+      "loss": 12.9389,
+      "step": 6243
+    },
+    {
+      "epoch": 1.0919781420765027,
+      "grad_norm": 0.897286866631415,
+      "learning_rate": 0.00014698401327394234,
+      "loss": 12.9488,
+      "step": 6244
+    },
+    {
+      "epoch": 1.0921530054644808,
+      "grad_norm": 1.2172831178394388,
+      "learning_rate": 0.0001469673452949441,
+      "loss": 12.7134,
+      "step": 6245
+    },
+    {
+      "epoch": 1.092327868852459,
+      "grad_norm": 0.9446791388850003,
+      "learning_rate": 0.00014695067564161492,
+      "loss": 12.8018,
+      "step": 6246
+    },
+    {
+      "epoch": 1.0925027322404373,
+      "grad_norm": 0.986378188432403,
+      "learning_rate": 0.00014693400431454895,
+      "loss": 12.8149,
+      "step": 6247
+    },
+    {
+      "epoch": 1.0926775956284154,
+      "grad_norm": 1.0038110307099106,
+      "learning_rate": 0.0001469173313143406,
+      "loss": 13.0523,
+      "step": 6248
+    },
+    {
+      "epoch": 1.0928524590163935,
+      "grad_norm": 1.1343447454099547,
+      "learning_rate": 0.0001469006566415842,
+      "loss": 12.8662,
+      "step": 6249
+    },
+    {
+      "epoch": 1.0930273224043716,
+      "grad_norm": 0.9775939586215213,
+      "learning_rate": 0.0001468839802968742,
+      "loss": 12.708,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0932021857923497,
+      "grad_norm": 1.164043130696164,
+      "learning_rate": 0.00014686730228080506,
+      "loss": 12.8954,
+      "step": 6251
+    },
+    {
+      "epoch": 1.0933770491803279,
+      "grad_norm": 1.0910058610079951,
+      "learning_rate": 0.00014685062259397135,
+      "loss": 13.0169,
+      "step": 6252
+    },
+    {
+      "epoch": 1.093551912568306,
+      "grad_norm": 1.1271257113464446,
+      "learning_rate": 0.0001468339412369677,
+      "loss": 12.8888,
+      "step": 6253
+    },
+    {
+      "epoch": 1.093726775956284,
+      "grad_norm": 0.9328723588093873,
+      "learning_rate": 0.00014681725821038874,
+      "loss": 12.9985,
+      "step": 6254
+    },
+    {
+      "epoch": 1.0939016393442622,
+      "grad_norm": 1.1181288002052772,
+      "learning_rate": 0.00014680057351482923,
+      "loss": 13.0369,
+      "step": 6255
+    },
+    {
+      "epoch": 1.0940765027322406,
+      "grad_norm": 1.1090594998143832,
+      "learning_rate": 0.00014678388715088397,
+      "loss": 12.979,
+      "step": 6256
+    },
+    {
+      "epoch": 1.0942513661202187,
+      "grad_norm": 1.0578731365895846,
+      "learning_rate": 0.00014676719911914778,
+      "loss": 12.74,
+      "step": 6257
+    },
+    {
+      "epoch": 1.0944262295081968,
+      "grad_norm": 1.0689641558793628,
+      "learning_rate": 0.0001467505094202156,
+      "loss": 12.89,
+      "step": 6258
+    },
+    {
+      "epoch": 1.094601092896175,
+      "grad_norm": 1.1684075268534562,
+      "learning_rate": 0.0001467338180546824,
+      "loss": 12.9374,
+      "step": 6259
+    },
+    {
+      "epoch": 1.094775956284153,
+      "grad_norm": 0.9522908957064957,
+      "learning_rate": 0.0001467171250231431,
+      "loss": 12.7482,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0949508196721311,
+      "grad_norm": 1.0922383157252478,
+      "learning_rate": 0.00014670043032619296,
+      "loss": 12.5736,
+      "step": 6261
+    },
+    {
+      "epoch": 1.0951256830601093,
+      "grad_norm": 1.0231141512517743,
+      "learning_rate": 0.00014668373396442704,
+      "loss": 12.6768,
+      "step": 6262
+    },
+    {
+      "epoch": 1.0953005464480874,
+      "grad_norm": 0.9997982507344119,
+      "learning_rate": 0.00014666703593844052,
+      "loss": 12.9415,
+      "step": 6263
+    },
+    {
+      "epoch": 1.0954754098360655,
+      "grad_norm": 0.9309429312587131,
+      "learning_rate": 0.00014665033624882868,
+      "loss": 13.0442,
+      "step": 6264
+    },
+    {
+      "epoch": 1.0956502732240436,
+      "grad_norm": 1.1470736436466606,
+      "learning_rate": 0.00014663363489618688,
+      "loss": 12.8411,
+      "step": 6265
+    },
+    {
+      "epoch": 1.0958251366120217,
+      "grad_norm": 0.951345998703373,
+      "learning_rate": 0.0001466169318811105,
+      "loss": 12.9262,
+      "step": 6266
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 1.0570042541084064,
+      "learning_rate": 0.00014660022720419495,
+      "loss": 13.0597,
+      "step": 6267
+    },
+    {
+      "epoch": 1.0961748633879782,
+      "grad_norm": 0.9565544927381568,
+      "learning_rate": 0.00014658352086603573,
+      "loss": 12.9491,
+      "step": 6268
+    },
+    {
+      "epoch": 1.0963497267759563,
+      "grad_norm": 1.039349078758091,
+      "learning_rate": 0.00014656681286722844,
+      "loss": 12.9323,
+      "step": 6269
+    },
+    {
+      "epoch": 1.0965245901639344,
+      "grad_norm": 1.1041804285951822,
+      "learning_rate": 0.00014655010320836865,
+      "loss": 12.939,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0966994535519126,
+      "grad_norm": 1.41697118914981,
+      "learning_rate": 0.00014653339189005208,
+      "loss": 12.9599,
+      "step": 6271
+    },
+    {
+      "epoch": 1.0968743169398907,
+      "grad_norm": 0.9212249713314206,
+      "learning_rate": 0.00014651667891287445,
+      "loss": 12.8481,
+      "step": 6272
+    },
+    {
+      "epoch": 1.0970491803278688,
+      "grad_norm": 1.066060897671866,
+      "learning_rate": 0.0001464999642774316,
+      "loss": 12.7165,
+      "step": 6273
+    },
+    {
+      "epoch": 1.097224043715847,
+      "grad_norm": 0.9631813649485593,
+      "learning_rate": 0.00014648324798431932,
+      "loss": 12.6435,
+      "step": 6274
+    },
+    {
+      "epoch": 1.097398907103825,
+      "grad_norm": 0.9530750259700367,
+      "learning_rate": 0.0001464665300341336,
+      "loss": 12.9717,
+      "step": 6275
+    },
+    {
+      "epoch": 1.0975737704918034,
+      "grad_norm": 1.1199982224662333,
+      "learning_rate": 0.00014644981042747035,
+      "loss": 12.631,
+      "step": 6276
+    },
+    {
+      "epoch": 1.0977486338797815,
+      "grad_norm": 1.0320097013363771,
+      "learning_rate": 0.00014643308916492562,
+      "loss": 12.7356,
+      "step": 6277
+    },
+    {
+      "epoch": 1.0979234972677596,
+      "grad_norm": 1.0762866843910106,
+      "learning_rate": 0.00014641636624709552,
+      "loss": 12.9607,
+      "step": 6278
+    },
+    {
+      "epoch": 1.0980983606557377,
+      "grad_norm": 1.0290287814914907,
+      "learning_rate": 0.00014639964167457618,
+      "loss": 12.9939,
+      "step": 6279
+    },
+    {
+      "epoch": 1.0982732240437159,
+      "grad_norm": 1.0204743747622995,
+      "learning_rate": 0.00014638291544796385,
+      "loss": 12.9651,
+      "step": 6280
+    },
+    {
+      "epoch": 1.098448087431694,
+      "grad_norm": 1.070125513950545,
+      "learning_rate": 0.00014636618756785476,
+      "loss": 13.0469,
+      "step": 6281
+    },
+    {
+      "epoch": 1.098622950819672,
+      "grad_norm": 0.9044936885111616,
+      "learning_rate": 0.00014634945803484529,
+      "loss": 13.0077,
+      "step": 6282
+    },
+    {
+      "epoch": 1.0987978142076502,
+      "grad_norm": 0.958300983012431,
+      "learning_rate": 0.00014633272684953173,
+      "loss": 13.0353,
+      "step": 6283
+    },
+    {
+      "epoch": 1.0989726775956283,
+      "grad_norm": 1.0290138079190367,
+      "learning_rate": 0.00014631599401251065,
+      "loss": 12.9675,
+      "step": 6284
+    },
+    {
+      "epoch": 1.0991475409836067,
+      "grad_norm": 0.958603922066716,
+      "learning_rate": 0.00014629925952437846,
+      "loss": 12.8193,
+      "step": 6285
+    },
+    {
+      "epoch": 1.0993224043715848,
+      "grad_norm": 0.9395647591625631,
+      "learning_rate": 0.0001462825233857318,
+      "loss": 13.126,
+      "step": 6286
+    },
+    {
+      "epoch": 1.099497267759563,
+      "grad_norm": 1.0609061454882525,
+      "learning_rate": 0.0001462657855971672,
+      "loss": 12.7109,
+      "step": 6287
+    },
+    {
+      "epoch": 1.099672131147541,
+      "grad_norm": 1.0829507612129392,
+      "learning_rate": 0.0001462490461592815,
+      "loss": 13.1596,
+      "step": 6288
+    },
+    {
+      "epoch": 1.0998469945355192,
+      "grad_norm": 0.9983384077540056,
+      "learning_rate": 0.00014623230507267122,
+      "loss": 12.8995,
+      "step": 6289
+    },
+    {
+      "epoch": 1.1000218579234973,
+      "grad_norm": 1.0599814009250546,
+      "learning_rate": 0.00014621556233793335,
+      "loss": 13.1156,
+      "step": 6290
+    },
+    {
+      "epoch": 1.1001967213114754,
+      "grad_norm": 1.0422770517601487,
+      "learning_rate": 0.00014619881795566468,
+      "loss": 13.1361,
+      "step": 6291
+    },
+    {
+      "epoch": 1.1003715846994535,
+      "grad_norm": 0.9523472933133751,
+      "learning_rate": 0.0001461820719264621,
+      "loss": 12.8996,
+      "step": 6292
+    },
+    {
+      "epoch": 1.1005464480874316,
+      "grad_norm": 0.9395769189448356,
+      "learning_rate": 0.00014616532425092263,
+      "loss": 12.8813,
+      "step": 6293
+    },
+    {
+      "epoch": 1.1007213114754097,
+      "grad_norm": 1.0164000044440684,
+      "learning_rate": 0.00014614857492964328,
+      "loss": 12.9323,
+      "step": 6294
+    },
+    {
+      "epoch": 1.1008961748633879,
+      "grad_norm": 0.9277586016604579,
+      "learning_rate": 0.0001461318239632212,
+      "loss": 12.8218,
+      "step": 6295
+    },
+    {
+      "epoch": 1.1010710382513662,
+      "grad_norm": 1.200826364416946,
+      "learning_rate": 0.00014611507135225344,
+      "loss": 12.8692,
+      "step": 6296
+    },
+    {
+      "epoch": 1.1012459016393443,
+      "grad_norm": 1.0299552141853505,
+      "learning_rate": 0.0001460983170973373,
+      "loss": 12.7139,
+      "step": 6297
+    },
+    {
+      "epoch": 1.1014207650273224,
+      "grad_norm": 0.9145062632730554,
+      "learning_rate": 0.00014608156119906996,
+      "loss": 12.7803,
+      "step": 6298
+    },
+    {
+      "epoch": 1.1015956284153006,
+      "grad_norm": 1.0442990512108332,
+      "learning_rate": 0.00014606480365804883,
+      "loss": 13.0558,
+      "step": 6299
+    },
+    {
+      "epoch": 1.1017704918032787,
+      "grad_norm": 0.9712195799905337,
+      "learning_rate": 0.0001460480444748713,
+      "loss": 12.7606,
+      "step": 6300
+    },
+    {
+      "epoch": 1.1019453551912568,
+      "grad_norm": 0.9627520130822141,
+      "learning_rate": 0.00014603128365013476,
+      "loss": 12.663,
+      "step": 6301
+    },
+    {
+      "epoch": 1.102120218579235,
+      "grad_norm": 0.9084916804667493,
+      "learning_rate": 0.00014601452118443673,
+      "loss": 12.7617,
+      "step": 6302
+    },
+    {
+      "epoch": 1.102295081967213,
+      "grad_norm": 0.977966459847853,
+      "learning_rate": 0.00014599775707837482,
+      "loss": 12.7474,
+      "step": 6303
+    },
+    {
+      "epoch": 1.1024699453551912,
+      "grad_norm": 0.9432222899089909,
+      "learning_rate": 0.00014598099133254658,
+      "loss": 13.0001,
+      "step": 6304
+    },
+    {
+      "epoch": 1.1026448087431695,
+      "grad_norm": 0.9257843033014742,
+      "learning_rate": 0.00014596422394754974,
+      "loss": 12.8003,
+      "step": 6305
+    },
+    {
+      "epoch": 1.1028196721311476,
+      "grad_norm": 0.9753411182550991,
+      "learning_rate": 0.00014594745492398203,
+      "loss": 12.6868,
+      "step": 6306
+    },
+    {
+      "epoch": 1.1029945355191257,
+      "grad_norm": 1.0142905944253662,
+      "learning_rate": 0.0001459306842624412,
+      "loss": 12.7587,
+      "step": 6307
+    },
+    {
+      "epoch": 1.1031693989071039,
+      "grad_norm": 0.9694804459757126,
+      "learning_rate": 0.00014591391196352516,
+      "loss": 13.0686,
+      "step": 6308
+    },
+    {
+      "epoch": 1.103344262295082,
+      "grad_norm": 0.9996573984234162,
+      "learning_rate": 0.0001458971380278318,
+      "loss": 12.9108,
+      "step": 6309
+    },
+    {
+      "epoch": 1.10351912568306,
+      "grad_norm": 0.9604803322821475,
+      "learning_rate": 0.00014588036245595908,
+      "loss": 13.0201,
+      "step": 6310
+    },
+    {
+      "epoch": 1.1036939890710382,
+      "grad_norm": 0.9608926024155768,
+      "learning_rate": 0.00014586358524850508,
+      "loss": 13.062,
+      "step": 6311
+    },
+    {
+      "epoch": 1.1038688524590163,
+      "grad_norm": 0.9132810687921301,
+      "learning_rate": 0.00014584680640606783,
+      "loss": 13.0516,
+      "step": 6312
+    },
+    {
+      "epoch": 1.1040437158469945,
+      "grad_norm": 0.935752088919276,
+      "learning_rate": 0.00014583002592924552,
+      "loss": 13.0301,
+      "step": 6313
+    },
+    {
+      "epoch": 1.1042185792349726,
+      "grad_norm": 0.9620894588283724,
+      "learning_rate": 0.0001458132438186363,
+      "loss": 12.8588,
+      "step": 6314
+    },
+    {
+      "epoch": 1.104393442622951,
+      "grad_norm": 0.9563481423452398,
+      "learning_rate": 0.00014579646007483848,
+      "loss": 12.8686,
+      "step": 6315
+    },
+    {
+      "epoch": 1.104568306010929,
+      "grad_norm": 0.9098575549137109,
+      "learning_rate": 0.00014577967469845034,
+      "loss": 12.7669,
+      "step": 6316
+    },
+    {
+      "epoch": 1.1047431693989072,
+      "grad_norm": 0.9425564971887411,
+      "learning_rate": 0.00014576288769007029,
+      "loss": 13.0285,
+      "step": 6317
+    },
+    {
+      "epoch": 1.1049180327868853,
+      "grad_norm": 1.15198627695795,
+      "learning_rate": 0.00014574609905029678,
+      "loss": 12.9789,
+      "step": 6318
+    },
+    {
+      "epoch": 1.1050928961748634,
+      "grad_norm": 0.9485724877603348,
+      "learning_rate": 0.0001457293087797283,
+      "loss": 12.7038,
+      "step": 6319
+    },
+    {
+      "epoch": 1.1052677595628415,
+      "grad_norm": 1.0433277647452865,
+      "learning_rate": 0.0001457125168789634,
+      "loss": 12.656,
+      "step": 6320
+    },
+    {
+      "epoch": 1.1054426229508196,
+      "grad_norm": 0.9675982571103142,
+      "learning_rate": 0.00014569572334860066,
+      "loss": 12.9069,
+      "step": 6321
+    },
+    {
+      "epoch": 1.1056174863387978,
+      "grad_norm": 1.0890260062819028,
+      "learning_rate": 0.00014567892818923876,
+      "loss": 12.6241,
+      "step": 6322
+    },
+    {
+      "epoch": 1.1057923497267759,
+      "grad_norm": 0.928556696217548,
+      "learning_rate": 0.00014566213140147647,
+      "loss": 12.8027,
+      "step": 6323
+    },
+    {
+      "epoch": 1.105967213114754,
+      "grad_norm": 1.04646964454437,
+      "learning_rate": 0.00014564533298591252,
+      "loss": 13.0335,
+      "step": 6324
+    },
+    {
+      "epoch": 1.1061420765027323,
+      "grad_norm": 1.0716959064473872,
+      "learning_rate": 0.00014562853294314577,
+      "loss": 12.9655,
+      "step": 6325
+    },
+    {
+      "epoch": 1.1063169398907104,
+      "grad_norm": 0.9617213168633011,
+      "learning_rate": 0.00014561173127377515,
+      "loss": 12.6803,
+      "step": 6326
+    },
+    {
+      "epoch": 1.1064918032786886,
+      "grad_norm": 1.0091388257423752,
+      "learning_rate": 0.0001455949279783996,
+      "loss": 12.678,
+      "step": 6327
+    },
+    {
+      "epoch": 1.1066666666666667,
+      "grad_norm": 0.9407302598154026,
+      "learning_rate": 0.00014557812305761813,
+      "loss": 12.8729,
+      "step": 6328
+    },
+    {
+      "epoch": 1.1068415300546448,
+      "grad_norm": 1.0004672387265763,
+      "learning_rate": 0.00014556131651202985,
+      "loss": 13.1892,
+      "step": 6329
+    },
+    {
+      "epoch": 1.107016393442623,
+      "grad_norm": 0.9511663473550727,
+      "learning_rate": 0.00014554450834223384,
+      "loss": 13.1127,
+      "step": 6330
+    },
+    {
+      "epoch": 1.107191256830601,
+      "grad_norm": 0.9178780915124245,
+      "learning_rate": 0.00014552769854882938,
+      "loss": 12.6862,
+      "step": 6331
+    },
+    {
+      "epoch": 1.1073661202185792,
+      "grad_norm": 0.8796991756826232,
+      "learning_rate": 0.0001455108871324156,
+      "loss": 12.5795,
+      "step": 6332
+    },
+    {
+      "epoch": 1.1075409836065573,
+      "grad_norm": 1.0710767666405834,
+      "learning_rate": 0.0001454940740935919,
+      "loss": 12.6096,
+      "step": 6333
+    },
+    {
+      "epoch": 1.1077158469945356,
+      "grad_norm": 0.9725456155878551,
+      "learning_rate": 0.00014547725943295757,
+      "loss": 13.0882,
+      "step": 6334
+    },
+    {
+      "epoch": 1.1078907103825137,
+      "grad_norm": 0.9497935307248269,
+      "learning_rate": 0.0001454604431511121,
+      "loss": 12.8412,
+      "step": 6335
+    },
+    {
+      "epoch": 1.1080655737704919,
+      "grad_norm": 0.9592913516496624,
+      "learning_rate": 0.00014544362524865494,
+      "loss": 12.5864,
+      "step": 6336
+    },
+    {
+      "epoch": 1.10824043715847,
+      "grad_norm": 0.9454769315659806,
+      "learning_rate": 0.00014542680572618565,
+      "loss": 12.9111,
+      "step": 6337
+    },
+    {
+      "epoch": 1.108415300546448,
+      "grad_norm": 0.8891294634950007,
+      "learning_rate": 0.0001454099845843038,
+      "loss": 12.7475,
+      "step": 6338
+    },
+    {
+      "epoch": 1.1085901639344262,
+      "grad_norm": 1.0320235917162013,
+      "learning_rate": 0.00014539316182360907,
+      "loss": 12.6395,
+      "step": 6339
+    },
+    {
+      "epoch": 1.1087650273224043,
+      "grad_norm": 0.9423801795327925,
+      "learning_rate": 0.00014537633744470115,
+      "loss": 12.7841,
+      "step": 6340
+    },
+    {
+      "epoch": 1.1089398907103825,
+      "grad_norm": 0.9446386433023142,
+      "learning_rate": 0.00014535951144817983,
+      "loss": 13.0652,
+      "step": 6341
+    },
+    {
+      "epoch": 1.1091147540983606,
+      "grad_norm": 0.9832210975545934,
+      "learning_rate": 0.0001453426838346449,
+      "loss": 12.8839,
+      "step": 6342
+    },
+    {
+      "epoch": 1.1092896174863387,
+      "grad_norm": 0.9359471562762891,
+      "learning_rate": 0.00014532585460469628,
+      "loss": 13.1025,
+      "step": 6343
+    },
+    {
+      "epoch": 1.109464480874317,
+      "grad_norm": 0.9050788566038713,
+      "learning_rate": 0.00014530902375893388,
+      "loss": 13.0382,
+      "step": 6344
+    },
+    {
+      "epoch": 1.1096393442622952,
+      "grad_norm": 0.9499302251658649,
+      "learning_rate": 0.00014529219129795774,
+      "loss": 12.9109,
+      "step": 6345
+    },
+    {
+      "epoch": 1.1098142076502733,
+      "grad_norm": 1.0204348703245336,
+      "learning_rate": 0.00014527535722236792,
+      "loss": 12.8995,
+      "step": 6346
+    },
+    {
+      "epoch": 1.1099890710382514,
+      "grad_norm": 1.0113824827397144,
+      "learning_rate": 0.00014525852153276448,
+      "loss": 12.4335,
+      "step": 6347
+    },
+    {
+      "epoch": 1.1101639344262295,
+      "grad_norm": 1.0340496120895781,
+      "learning_rate": 0.00014524168422974765,
+      "loss": 13.0347,
+      "step": 6348
+    },
+    {
+      "epoch": 1.1103387978142076,
+      "grad_norm": 1.1248638810697578,
+      "learning_rate": 0.0001452248453139176,
+      "loss": 12.7543,
+      "step": 6349
+    },
+    {
+      "epoch": 1.1105136612021858,
+      "grad_norm": 0.8899636266982461,
+      "learning_rate": 0.0001452080047858747,
+      "loss": 12.7606,
+      "step": 6350
+    },
+    {
+      "epoch": 1.1106885245901639,
+      "grad_norm": 0.981202049316266,
+      "learning_rate": 0.00014519116264621922,
+      "loss": 13.0214,
+      "step": 6351
+    },
+    {
+      "epoch": 1.110863387978142,
+      "grad_norm": 0.9366726835783897,
+      "learning_rate": 0.00014517431889555162,
+      "loss": 12.3788,
+      "step": 6352
+    },
+    {
+      "epoch": 1.11103825136612,
+      "grad_norm": 1.0154186160809773,
+      "learning_rate": 0.0001451574735344723,
+      "loss": 12.7503,
+      "step": 6353
+    },
+    {
+      "epoch": 1.1112131147540985,
+      "grad_norm": 0.9302609891062468,
+      "learning_rate": 0.00014514062656358182,
+      "loss": 12.9493,
+      "step": 6354
+    },
+    {
+      "epoch": 1.1113879781420766,
+      "grad_norm": 1.030923826793076,
+      "learning_rate": 0.00014512377798348074,
+      "loss": 12.9461,
+      "step": 6355
+    },
+    {
+      "epoch": 1.1115628415300547,
+      "grad_norm": 1.1090145433464,
+      "learning_rate": 0.00014510692779476968,
+      "loss": 13.0448,
+      "step": 6356
+    },
+    {
+      "epoch": 1.1117377049180328,
+      "grad_norm": 0.9437255949481375,
+      "learning_rate": 0.00014509007599804941,
+      "loss": 12.8133,
+      "step": 6357
+    },
+    {
+      "epoch": 1.111912568306011,
+      "grad_norm": 1.0670703734986478,
+      "learning_rate": 0.00014507322259392055,
+      "loss": 12.7118,
+      "step": 6358
+    },
+    {
+      "epoch": 1.112087431693989,
+      "grad_norm": 1.0062248392589532,
+      "learning_rate": 0.000145056367582984,
+      "loss": 12.9077,
+      "step": 6359
+    },
+    {
+      "epoch": 1.1122622950819672,
+      "grad_norm": 0.9336293038361126,
+      "learning_rate": 0.00014503951096584057,
+      "loss": 12.9178,
+      "step": 6360
+    },
+    {
+      "epoch": 1.1124371584699453,
+      "grad_norm": 0.9790730404136643,
+      "learning_rate": 0.00014502265274309118,
+      "loss": 12.8311,
+      "step": 6361
+    },
+    {
+      "epoch": 1.1126120218579234,
+      "grad_norm": 1.0436651936885484,
+      "learning_rate": 0.00014500579291533682,
+      "loss": 12.8559,
+      "step": 6362
+    },
+    {
+      "epoch": 1.1127868852459017,
+      "grad_norm": 0.9974449976817464,
+      "learning_rate": 0.00014498893148317857,
+      "loss": 12.8152,
+      "step": 6363
+    },
+    {
+      "epoch": 1.1129617486338799,
+      "grad_norm": 1.114683882441767,
+      "learning_rate": 0.00014497206844721748,
+      "loss": 12.946,
+      "step": 6364
+    },
+    {
+      "epoch": 1.113136612021858,
+      "grad_norm": 1.1209641439967066,
+      "learning_rate": 0.00014495520380805466,
+      "loss": 12.7475,
+      "step": 6365
+    },
+    {
+      "epoch": 1.113311475409836,
+      "grad_norm": 1.131384132509228,
+      "learning_rate": 0.00014493833756629135,
+      "loss": 13.0722,
+      "step": 6366
+    },
+    {
+      "epoch": 1.1134863387978142,
+      "grad_norm": 0.9369518273606976,
+      "learning_rate": 0.00014492146972252883,
+      "loss": 12.5973,
+      "step": 6367
+    },
+    {
+      "epoch": 1.1136612021857923,
+      "grad_norm": 0.9492506939384171,
+      "learning_rate": 0.0001449046002773684,
+      "loss": 13.0732,
+      "step": 6368
+    },
+    {
+      "epoch": 1.1138360655737705,
+      "grad_norm": 0.9700744951451677,
+      "learning_rate": 0.00014488772923141143,
+      "loss": 12.8815,
+      "step": 6369
+    },
+    {
+      "epoch": 1.1140109289617486,
+      "grad_norm": 1.0459142466049152,
+      "learning_rate": 0.00014487085658525933,
+      "loss": 13.0273,
+      "step": 6370
+    },
+    {
+      "epoch": 1.1141857923497267,
+      "grad_norm": 0.9312395071086997,
+      "learning_rate": 0.00014485398233951365,
+      "loss": 12.6671,
+      "step": 6371
+    },
+    {
+      "epoch": 1.1143606557377048,
+      "grad_norm": 0.8681455354798004,
+      "learning_rate": 0.0001448371064947759,
+      "loss": 12.5922,
+      "step": 6372
+    },
+    {
+      "epoch": 1.1145355191256832,
+      "grad_norm": 1.0270880668953863,
+      "learning_rate": 0.0001448202290516477,
+      "loss": 12.7527,
+      "step": 6373
+    },
+    {
+      "epoch": 1.1147103825136613,
+      "grad_norm": 1.0742797570460578,
+      "learning_rate": 0.00014480335001073067,
+      "loss": 12.968,
+      "step": 6374
+    },
+    {
+      "epoch": 1.1148852459016394,
+      "grad_norm": 1.0201181038455271,
+      "learning_rate": 0.0001447864693726266,
+      "loss": 12.7847,
+      "step": 6375
+    },
+    {
+      "epoch": 1.1150601092896175,
+      "grad_norm": 1.245749283385066,
+      "learning_rate": 0.0001447695871379372,
+      "loss": 12.6743,
+      "step": 6376
+    },
+    {
+      "epoch": 1.1152349726775956,
+      "grad_norm": 0.9422023286609262,
+      "learning_rate": 0.00014475270330726434,
+      "loss": 12.9328,
+      "step": 6377
+    },
+    {
+      "epoch": 1.1154098360655738,
+      "grad_norm": 1.1838331056395668,
+      "learning_rate": 0.0001447358178812099,
+      "loss": 12.799,
+      "step": 6378
+    },
+    {
+      "epoch": 1.1155846994535519,
+      "grad_norm": 1.003869968995696,
+      "learning_rate": 0.00014471893086037577,
+      "loss": 12.9162,
+      "step": 6379
+    },
+    {
+      "epoch": 1.11575956284153,
+      "grad_norm": 1.0858836647557837,
+      "learning_rate": 0.00014470204224536402,
+      "loss": 12.9116,
+      "step": 6380
+    },
+    {
+      "epoch": 1.1159344262295081,
+      "grad_norm": 0.9586278873311045,
+      "learning_rate": 0.00014468515203677672,
+      "loss": 12.4839,
+      "step": 6381
+    },
+    {
+      "epoch": 1.1161092896174862,
+      "grad_norm": 1.0085426135829159,
+      "learning_rate": 0.00014466826023521598,
+      "loss": 12.9173,
+      "step": 6382
+    },
+    {
+      "epoch": 1.1162841530054646,
+      "grad_norm": 0.949018213652461,
+      "learning_rate": 0.00014465136684128392,
+      "loss": 12.8466,
+      "step": 6383
+    },
+    {
+      "epoch": 1.1164590163934427,
+      "grad_norm": 1.0161370484820145,
+      "learning_rate": 0.0001446344718555828,
+      "loss": 12.8546,
+      "step": 6384
+    },
+    {
+      "epoch": 1.1166338797814208,
+      "grad_norm": 1.0053997129332755,
+      "learning_rate": 0.00014461757527871492,
+      "loss": 12.9308,
+      "step": 6385
+    },
+    {
+      "epoch": 1.116808743169399,
+      "grad_norm": 0.9670227037301652,
+      "learning_rate": 0.0001446006771112826,
+      "loss": 12.7162,
+      "step": 6386
+    },
+    {
+      "epoch": 1.116983606557377,
+      "grad_norm": 1.01974986291054,
+      "learning_rate": 0.00014458377735388826,
+      "loss": 12.8361,
+      "step": 6387
+    },
+    {
+      "epoch": 1.1171584699453552,
+      "grad_norm": 0.9072435499363817,
+      "learning_rate": 0.00014456687600713434,
+      "loss": 12.656,
+      "step": 6388
+    },
+    {
+      "epoch": 1.1173333333333333,
+      "grad_norm": 1.0420588755748312,
+      "learning_rate": 0.00014454997307162333,
+      "loss": 12.8271,
+      "step": 6389
+    },
+    {
+      "epoch": 1.1175081967213114,
+      "grad_norm": 1.1019861280559637,
+      "learning_rate": 0.00014453306854795788,
+      "loss": 12.9671,
+      "step": 6390
+    },
+    {
+      "epoch": 1.1176830601092895,
+      "grad_norm": 1.3447116523139957,
+      "learning_rate": 0.00014451616243674054,
+      "loss": 13.0958,
+      "step": 6391
+    },
+    {
+      "epoch": 1.1178579234972679,
+      "grad_norm": 1.0578326067958506,
+      "learning_rate": 0.00014449925473857404,
+      "loss": 12.8351,
+      "step": 6392
+    },
+    {
+      "epoch": 1.118032786885246,
+      "grad_norm": 1.0592982220191571,
+      "learning_rate": 0.0001444823454540611,
+      "loss": 12.8234,
+      "step": 6393
+    },
+    {
+      "epoch": 1.118207650273224,
+      "grad_norm": 1.1017676111705226,
+      "learning_rate": 0.00014446543458380452,
+      "loss": 12.9138,
+      "step": 6394
+    },
+    {
+      "epoch": 1.1183825136612022,
+      "grad_norm": 1.0963187812962014,
+      "learning_rate": 0.00014444852212840714,
+      "loss": 12.9175,
+      "step": 6395
+    },
+    {
+      "epoch": 1.1185573770491803,
+      "grad_norm": 0.9108592015466924,
+      "learning_rate": 0.00014443160808847187,
+      "loss": 12.8708,
+      "step": 6396
+    },
+    {
+      "epoch": 1.1187322404371585,
+      "grad_norm": 1.2003041354969557,
+      "learning_rate": 0.0001444146924646017,
+      "loss": 12.8917,
+      "step": 6397
+    },
+    {
+      "epoch": 1.1189071038251366,
+      "grad_norm": 0.9665717501208945,
+      "learning_rate": 0.00014439777525739964,
+      "loss": 12.7063,
+      "step": 6398
+    },
+    {
+      "epoch": 1.1190819672131147,
+      "grad_norm": 1.0616984620705514,
+      "learning_rate": 0.00014438085646746872,
+      "loss": 12.8401,
+      "step": 6399
+    },
+    {
+      "epoch": 1.1192568306010928,
+      "grad_norm": 0.9315509295706816,
+      "learning_rate": 0.00014436393609541218,
+      "loss": 12.9934,
+      "step": 6400
+    },
+    {
+      "epoch": 1.119431693989071,
+      "grad_norm": 0.8971615598039472,
+      "learning_rate": 0.00014434701414183312,
+      "loss": 12.6968,
+      "step": 6401
+    },
+    {
+      "epoch": 1.1196065573770493,
+      "grad_norm": 1.0344832787664062,
+      "learning_rate": 0.00014433009060733484,
+      "loss": 12.8498,
+      "step": 6402
+    },
+    {
+      "epoch": 1.1197814207650274,
+      "grad_norm": 0.9193521123094371,
+      "learning_rate": 0.00014431316549252062,
+      "loss": 13.074,
+      "step": 6403
+    },
+    {
+      "epoch": 1.1199562841530055,
+      "grad_norm": 0.9906045258009556,
+      "learning_rate": 0.0001442962387979938,
+      "loss": 13.1379,
+      "step": 6404
+    },
+    {
+      "epoch": 1.1201311475409836,
+      "grad_norm": 0.9308858086806688,
+      "learning_rate": 0.00014427931052435783,
+      "loss": 12.7646,
+      "step": 6405
+    },
+    {
+      "epoch": 1.1203060109289618,
+      "grad_norm": 1.0568630778280672,
+      "learning_rate": 0.00014426238067221617,
+      "loss": 12.9502,
+      "step": 6406
+    },
+    {
+      "epoch": 1.1204808743169399,
+      "grad_norm": 0.873647580299579,
+      "learning_rate": 0.00014424544924217236,
+      "loss": 12.888,
+      "step": 6407
+    },
+    {
+      "epoch": 1.120655737704918,
+      "grad_norm": 0.9632745909856218,
+      "learning_rate": 0.00014422851623483,
+      "loss": 12.7534,
+      "step": 6408
+    },
+    {
+      "epoch": 1.1208306010928961,
+      "grad_norm": 1.0129396645193791,
+      "learning_rate": 0.00014421158165079268,
+      "loss": 12.6197,
+      "step": 6409
+    },
+    {
+      "epoch": 1.1210054644808742,
+      "grad_norm": 0.9262746699281231,
+      "learning_rate": 0.00014419464549066417,
+      "loss": 12.6364,
+      "step": 6410
+    },
+    {
+      "epoch": 1.1211803278688524,
+      "grad_norm": 1.0071968269610994,
+      "learning_rate": 0.00014417770775504814,
+      "loss": 12.5222,
+      "step": 6411
+    },
+    {
+      "epoch": 1.1213551912568307,
+      "grad_norm": 1.2990062985206479,
+      "learning_rate": 0.0001441607684445485,
+      "loss": 12.9567,
+      "step": 6412
+    },
+    {
+      "epoch": 1.1215300546448088,
+      "grad_norm": 1.1312525058230316,
+      "learning_rate": 0.000144143827559769,
+      "loss": 12.9621,
+      "step": 6413
+    },
+    {
+      "epoch": 1.121704918032787,
+      "grad_norm": 1.2204976483052687,
+      "learning_rate": 0.00014412688510131367,
+      "loss": 12.6713,
+      "step": 6414
+    },
+    {
+      "epoch": 1.121879781420765,
+      "grad_norm": 1.2698758368789562,
+      "learning_rate": 0.00014410994106978643,
+      "loss": 12.9339,
+      "step": 6415
+    },
+    {
+      "epoch": 1.1220546448087432,
+      "grad_norm": 1.1167058173336821,
+      "learning_rate": 0.0001440929954657913,
+      "loss": 12.8938,
+      "step": 6416
+    },
+    {
+      "epoch": 1.1222295081967213,
+      "grad_norm": 1.1504900803578217,
+      "learning_rate": 0.00014407604828993242,
+      "loss": 12.8303,
+      "step": 6417
+    },
+    {
+      "epoch": 1.1224043715846994,
+      "grad_norm": 1.4538909223370882,
+      "learning_rate": 0.00014405909954281392,
+      "loss": 12.9189,
+      "step": 6418
+    },
+    {
+      "epoch": 1.1225792349726775,
+      "grad_norm": 1.1355565655102084,
+      "learning_rate": 0.00014404214922504,
+      "loss": 13.0846,
+      "step": 6419
+    },
+    {
+      "epoch": 1.1227540983606557,
+      "grad_norm": 1.2370906102430759,
+      "learning_rate": 0.00014402519733721492,
+      "loss": 13.1167,
+      "step": 6420
+    },
+    {
+      "epoch": 1.122928961748634,
+      "grad_norm": 1.1521257990450127,
+      "learning_rate": 0.00014400824387994298,
+      "loss": 13.0763,
+      "step": 6421
+    },
+    {
+      "epoch": 1.123103825136612,
+      "grad_norm": 1.180046360747912,
+      "learning_rate": 0.00014399128885382855,
+      "loss": 12.5617,
+      "step": 6422
+    },
+    {
+      "epoch": 1.1232786885245902,
+      "grad_norm": 1.0591725059779817,
+      "learning_rate": 0.00014397433225947612,
+      "loss": 13.0939,
+      "step": 6423
+    },
+    {
+      "epoch": 1.1234535519125683,
+      "grad_norm": 1.2864453677821048,
+      "learning_rate": 0.00014395737409749008,
+      "loss": 12.867,
+      "step": 6424
+    },
+    {
+      "epoch": 1.1236284153005465,
+      "grad_norm": 1.0335774391227541,
+      "learning_rate": 0.000143940414368475,
+      "loss": 12.9329,
+      "step": 6425
+    },
+    {
+      "epoch": 1.1238032786885246,
+      "grad_norm": 1.0676868689151269,
+      "learning_rate": 0.0001439234530730355,
+      "loss": 12.6845,
+      "step": 6426
+    },
+    {
+      "epoch": 1.1239781420765027,
+      "grad_norm": 1.041373465387186,
+      "learning_rate": 0.00014390649021177624,
+      "loss": 12.8845,
+      "step": 6427
+    },
+    {
+      "epoch": 1.1241530054644808,
+      "grad_norm": 1.0335581969493821,
+      "learning_rate": 0.0001438895257853019,
+      "loss": 12.9124,
+      "step": 6428
+    },
+    {
+      "epoch": 1.124327868852459,
+      "grad_norm": 1.016760514270019,
+      "learning_rate": 0.00014387255979421724,
+      "loss": 12.9475,
+      "step": 6429
+    },
+    {
+      "epoch": 1.124502732240437,
+      "grad_norm": 0.9706985036359654,
+      "learning_rate": 0.0001438555922391271,
+      "loss": 12.8509,
+      "step": 6430
+    },
+    {
+      "epoch": 1.1246775956284152,
+      "grad_norm": 1.074576430164745,
+      "learning_rate": 0.0001438386231206363,
+      "loss": 13.1062,
+      "step": 6431
+    },
+    {
+      "epoch": 1.1248524590163935,
+      "grad_norm": 0.9814547347892936,
+      "learning_rate": 0.00014382165243934984,
+      "loss": 12.834,
+      "step": 6432
+    },
+    {
+      "epoch": 1.1250273224043716,
+      "grad_norm": 1.0382505068133778,
+      "learning_rate": 0.00014380468019587264,
+      "loss": 12.8115,
+      "step": 6433
+    },
+    {
+      "epoch": 1.1252021857923498,
+      "grad_norm": 0.9684459837390821,
+      "learning_rate": 0.0001437877063908098,
+      "loss": 12.6882,
+      "step": 6434
+    },
+    {
+      "epoch": 1.1253770491803279,
+      "grad_norm": 0.8768797156921189,
+      "learning_rate": 0.00014377073102476634,
+      "loss": 12.6754,
+      "step": 6435
+    },
+    {
+      "epoch": 1.125551912568306,
+      "grad_norm": 0.9517680984618392,
+      "learning_rate": 0.0001437537540983475,
+      "loss": 12.7623,
+      "step": 6436
+    },
+    {
+      "epoch": 1.1257267759562841,
+      "grad_norm": 0.9493805454148532,
+      "learning_rate": 0.00014373677561215842,
+      "loss": 12.9781,
+      "step": 6437
+    },
+    {
+      "epoch": 1.1259016393442622,
+      "grad_norm": 0.9574495314205402,
+      "learning_rate": 0.00014371979556680443,
+      "loss": 12.8768,
+      "step": 6438
+    },
+    {
+      "epoch": 1.1260765027322404,
+      "grad_norm": 1.032941449411265,
+      "learning_rate": 0.00014370281396289077,
+      "loss": 12.8082,
+      "step": 6439
+    },
+    {
+      "epoch": 1.1262513661202185,
+      "grad_norm": 0.9924182176931756,
+      "learning_rate": 0.00014368583080102293,
+      "loss": 12.6782,
+      "step": 6440
+    },
+    {
+      "epoch": 1.1264262295081968,
+      "grad_norm": 1.1434980295913995,
+      "learning_rate": 0.00014366884608180619,
+      "loss": 12.9386,
+      "step": 6441
+    },
+    {
+      "epoch": 1.126601092896175,
+      "grad_norm": 0.8831817120929184,
+      "learning_rate": 0.00014365185980584613,
+      "loss": 12.4554,
+      "step": 6442
+    },
+    {
+      "epoch": 1.126775956284153,
+      "grad_norm": 1.0488294318990419,
+      "learning_rate": 0.00014363487197374824,
+      "loss": 12.8442,
+      "step": 6443
+    },
+    {
+      "epoch": 1.1269508196721312,
+      "grad_norm": 0.9942428447101139,
+      "learning_rate": 0.00014361788258611817,
+      "loss": 12.7351,
+      "step": 6444
+    },
+    {
+      "epoch": 1.1271256830601093,
+      "grad_norm": 0.9964110389797936,
+      "learning_rate": 0.00014360089164356152,
+      "loss": 12.7442,
+      "step": 6445
+    },
+    {
+      "epoch": 1.1273005464480874,
+      "grad_norm": 0.9954202265405698,
+      "learning_rate": 0.00014358389914668405,
+      "loss": 13.0806,
+      "step": 6446
+    },
+    {
+      "epoch": 1.1274754098360655,
+      "grad_norm": 1.0517072984829197,
+      "learning_rate": 0.00014356690509609155,
+      "loss": 12.8548,
+      "step": 6447
+    },
+    {
+      "epoch": 1.1276502732240437,
+      "grad_norm": 1.101375665883472,
+      "learning_rate": 0.00014354990949238972,
+      "loss": 12.5517,
+      "step": 6448
+    },
+    {
+      "epoch": 1.1278251366120218,
+      "grad_norm": 1.201254313651208,
+      "learning_rate": 0.0001435329123361845,
+      "loss": 12.5152,
+      "step": 6449
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 1.073943513165616,
+      "learning_rate": 0.00014351591362808184,
+      "loss": 13.028,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1281748633879782,
+      "grad_norm": 1.1444585907878908,
+      "learning_rate": 0.00014349891336868772,
+      "loss": 12.9701,
+      "step": 6451
+    },
+    {
+      "epoch": 1.1283497267759564,
+      "grad_norm": 1.0117360145341276,
+      "learning_rate": 0.00014348191155860813,
+      "loss": 12.8778,
+      "step": 6452
+    },
+    {
+      "epoch": 1.1285245901639345,
+      "grad_norm": 1.0850977497059378,
+      "learning_rate": 0.00014346490819844917,
+      "loss": 12.8148,
+      "step": 6453
+    },
+    {
+      "epoch": 1.1286994535519126,
+      "grad_norm": 0.8315285661546942,
+      "learning_rate": 0.00014344790328881706,
+      "loss": 12.7837,
+      "step": 6454
+    },
+    {
+      "epoch": 1.1288743169398907,
+      "grad_norm": 0.914360589363778,
+      "learning_rate": 0.00014343089683031796,
+      "loss": 12.8954,
+      "step": 6455
+    },
+    {
+      "epoch": 1.1290491803278688,
+      "grad_norm": 1.0350191348069273,
+      "learning_rate": 0.0001434138888235581,
+      "loss": 12.917,
+      "step": 6456
+    },
+    {
+      "epoch": 1.129224043715847,
+      "grad_norm": 0.9910393349542456,
+      "learning_rate": 0.00014339687926914386,
+      "loss": 12.8862,
+      "step": 6457
+    },
+    {
+      "epoch": 1.129398907103825,
+      "grad_norm": 0.9043226995202885,
+      "learning_rate": 0.00014337986816768155,
+      "loss": 12.6928,
+      "step": 6458
+    },
+    {
+      "epoch": 1.1295737704918032,
+      "grad_norm": 0.9546321134562754,
+      "learning_rate": 0.00014336285551977767,
+      "loss": 12.8463,
+      "step": 6459
+    },
+    {
+      "epoch": 1.1297486338797813,
+      "grad_norm": 1.0719111610074294,
+      "learning_rate": 0.0001433458413260386,
+      "loss": 12.6827,
+      "step": 6460
+    },
+    {
+      "epoch": 1.1299234972677596,
+      "grad_norm": 1.1207838838846786,
+      "learning_rate": 0.00014332882558707098,
+      "loss": 12.9513,
+      "step": 6461
+    },
+    {
+      "epoch": 1.1300983606557378,
+      "grad_norm": 0.9031128279630334,
+      "learning_rate": 0.00014331180830348133,
+      "loss": 12.9169,
+      "step": 6462
+    },
+    {
+      "epoch": 1.1302732240437159,
+      "grad_norm": 1.220623539482139,
+      "learning_rate": 0.0001432947894758763,
+      "loss": 12.9165,
+      "step": 6463
+    },
+    {
+      "epoch": 1.130448087431694,
+      "grad_norm": 0.9315008612807775,
+      "learning_rate": 0.00014327776910486265,
+      "loss": 12.4881,
+      "step": 6464
+    },
+    {
+      "epoch": 1.1306229508196721,
+      "grad_norm": 1.088064058424339,
+      "learning_rate": 0.00014326074719104705,
+      "loss": 12.7064,
+      "step": 6465
+    },
+    {
+      "epoch": 1.1307978142076502,
+      "grad_norm": 0.9598666051348961,
+      "learning_rate": 0.00014324372373503638,
+      "loss": 12.8713,
+      "step": 6466
+    },
+    {
+      "epoch": 1.1309726775956284,
+      "grad_norm": 1.0707143112477504,
+      "learning_rate": 0.00014322669873743747,
+      "loss": 12.7194,
+      "step": 6467
+    },
+    {
+      "epoch": 1.1311475409836065,
+      "grad_norm": 0.9684616800977404,
+      "learning_rate": 0.0001432096721988573,
+      "loss": 12.8694,
+      "step": 6468
+    },
+    {
+      "epoch": 1.1313224043715846,
+      "grad_norm": 1.1991315652809633,
+      "learning_rate": 0.00014319264411990275,
+      "loss": 12.9362,
+      "step": 6469
+    },
+    {
+      "epoch": 1.131497267759563,
+      "grad_norm": 0.960766131299217,
+      "learning_rate": 0.00014317561450118092,
+      "loss": 12.9742,
+      "step": 6470
+    },
+    {
+      "epoch": 1.131672131147541,
+      "grad_norm": 1.1936158438001252,
+      "learning_rate": 0.00014315858334329885,
+      "loss": 13.1684,
+      "step": 6471
+    },
+    {
+      "epoch": 1.1318469945355192,
+      "grad_norm": 1.0305623286947165,
+      "learning_rate": 0.00014314155064686378,
+      "loss": 12.7466,
+      "step": 6472
+    },
+    {
+      "epoch": 1.1320218579234973,
+      "grad_norm": 1.1281574888523183,
+      "learning_rate": 0.00014312451641248277,
+      "loss": 12.6614,
+      "step": 6473
+    },
+    {
+      "epoch": 1.1321967213114754,
+      "grad_norm": 0.9765478662344634,
+      "learning_rate": 0.00014310748064076318,
+      "loss": 12.6269,
+      "step": 6474
+    },
+    {
+      "epoch": 1.1323715846994535,
+      "grad_norm": 1.2108152423802356,
+      "learning_rate": 0.00014309044333231223,
+      "loss": 12.5652,
+      "step": 6475
+    },
+    {
+      "epoch": 1.1325464480874317,
+      "grad_norm": 0.9203189608187743,
+      "learning_rate": 0.00014307340448773735,
+      "loss": 12.8437,
+      "step": 6476
+    },
+    {
+      "epoch": 1.1327213114754098,
+      "grad_norm": 1.4240799588524367,
+      "learning_rate": 0.00014305636410764594,
+      "loss": 12.9081,
+      "step": 6477
+    },
+    {
+      "epoch": 1.132896174863388,
+      "grad_norm": 0.9999659911894696,
+      "learning_rate": 0.00014303932219264542,
+      "loss": 12.6907,
+      "step": 6478
+    },
+    {
+      "epoch": 1.1330710382513662,
+      "grad_norm": 1.1065872433548516,
+      "learning_rate": 0.00014302227874334338,
+      "loss": 12.8326,
+      "step": 6479
+    },
+    {
+      "epoch": 1.1332459016393444,
+      "grad_norm": 1.0002032714454996,
+      "learning_rate": 0.00014300523376034732,
+      "loss": 12.7641,
+      "step": 6480
+    },
+    {
+      "epoch": 1.1334207650273225,
+      "grad_norm": 1.141390323017141,
+      "learning_rate": 0.00014298818724426497,
+      "loss": 12.7693,
+      "step": 6481
+    },
+    {
+      "epoch": 1.1335956284153006,
+      "grad_norm": 0.9135942539532641,
+      "learning_rate": 0.00014297113919570396,
+      "loss": 12.9565,
+      "step": 6482
+    },
+    {
+      "epoch": 1.1337704918032787,
+      "grad_norm": 1.2015062325555166,
+      "learning_rate": 0.00014295408961527203,
+      "loss": 12.895,
+      "step": 6483
+    },
+    {
+      "epoch": 1.1339453551912568,
+      "grad_norm": 0.898389332849536,
+      "learning_rate": 0.000142937038503577,
+      "loss": 12.8923,
+      "step": 6484
+    },
+    {
+      "epoch": 1.134120218579235,
+      "grad_norm": 1.0316222605568794,
+      "learning_rate": 0.00014291998586122673,
+      "loss": 12.721,
+      "step": 6485
+    },
+    {
+      "epoch": 1.134295081967213,
+      "grad_norm": 1.0684514052855079,
+      "learning_rate": 0.0001429029316888291,
+      "loss": 12.9148,
+      "step": 6486
+    },
+    {
+      "epoch": 1.1344699453551912,
+      "grad_norm": 0.9499253805484872,
+      "learning_rate": 0.00014288587598699208,
+      "loss": 12.9663,
+      "step": 6487
+    },
+    {
+      "epoch": 1.1346448087431693,
+      "grad_norm": 0.9948072013570985,
+      "learning_rate": 0.0001428688187563237,
+      "loss": 12.6588,
+      "step": 6488
+    },
+    {
+      "epoch": 1.1348196721311474,
+      "grad_norm": 0.9618649625321469,
+      "learning_rate": 0.00014285175999743198,
+      "loss": 12.8599,
+      "step": 6489
+    },
+    {
+      "epoch": 1.1349945355191258,
+      "grad_norm": 1.166034966115638,
+      "learning_rate": 0.0001428346997109251,
+      "loss": 12.8333,
+      "step": 6490
+    },
+    {
+      "epoch": 1.1351693989071039,
+      "grad_norm": 0.9264260346381482,
+      "learning_rate": 0.00014281763789741122,
+      "loss": 12.8937,
+      "step": 6491
+    },
+    {
+      "epoch": 1.135344262295082,
+      "grad_norm": 1.0727430796594608,
+      "learning_rate": 0.0001428005745574986,
+      "loss": 12.7757,
+      "step": 6492
+    },
+    {
+      "epoch": 1.1355191256830601,
+      "grad_norm": 0.9019240058369361,
+      "learning_rate": 0.00014278350969179547,
+      "loss": 12.4689,
+      "step": 6493
+    },
+    {
+      "epoch": 1.1356939890710382,
+      "grad_norm": 0.9602386866563427,
+      "learning_rate": 0.00014276644330091026,
+      "loss": 12.6508,
+      "step": 6494
+    },
+    {
+      "epoch": 1.1358688524590164,
+      "grad_norm": 0.916404651483836,
+      "learning_rate": 0.00014274937538545125,
+      "loss": 12.6597,
+      "step": 6495
+    },
+    {
+      "epoch": 1.1360437158469945,
+      "grad_norm": 0.9676879853261008,
+      "learning_rate": 0.000142732305946027,
+      "loss": 12.6929,
+      "step": 6496
+    },
+    {
+      "epoch": 1.1362185792349726,
+      "grad_norm": 1.0462073827096285,
+      "learning_rate": 0.00014271523498324596,
+      "loss": 13.0406,
+      "step": 6497
+    },
+    {
+      "epoch": 1.1363934426229507,
+      "grad_norm": 0.9694616889619639,
+      "learning_rate": 0.0001426981624977167,
+      "loss": 12.9042,
+      "step": 6498
+    },
+    {
+      "epoch": 1.136568306010929,
+      "grad_norm": 0.9120405605003369,
+      "learning_rate": 0.0001426810884900478,
+      "loss": 12.6996,
+      "step": 6499
+    },
+    {
+      "epoch": 1.1367431693989072,
+      "grad_norm": 1.0299978503205096,
+      "learning_rate": 0.000142664012960848,
+      "loss": 12.7555,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1369180327868853,
+      "grad_norm": 0.9726580762888906,
+      "learning_rate": 0.00014264693591072597,
+      "loss": 12.9049,
+      "step": 6501
+    },
+    {
+      "epoch": 1.1370928961748634,
+      "grad_norm": 0.9511296699960244,
+      "learning_rate": 0.00014262985734029052,
+      "loss": 12.9961,
+      "step": 6502
+    },
+    {
+      "epoch": 1.1372677595628415,
+      "grad_norm": 1.0002802309486658,
+      "learning_rate": 0.00014261277725015044,
+      "loss": 12.6852,
+      "step": 6503
+    },
+    {
+      "epoch": 1.1374426229508197,
+      "grad_norm": 0.9924940064815208,
+      "learning_rate": 0.00014259569564091466,
+      "loss": 12.8917,
+      "step": 6504
+    },
+    {
+      "epoch": 1.1376174863387978,
+      "grad_norm": 0.9207623041832929,
+      "learning_rate": 0.00014257861251319208,
+      "loss": 12.8495,
+      "step": 6505
+    },
+    {
+      "epoch": 1.137792349726776,
+      "grad_norm": 1.0051821164578862,
+      "learning_rate": 0.0001425615278675917,
+      "loss": 13.0293,
+      "step": 6506
+    },
+    {
+      "epoch": 1.137967213114754,
+      "grad_norm": 1.0270370641460322,
+      "learning_rate": 0.0001425444417047226,
+      "loss": 12.867,
+      "step": 6507
+    },
+    {
+      "epoch": 1.1381420765027324,
+      "grad_norm": 1.0373053113354893,
+      "learning_rate": 0.00014252735402519388,
+      "loss": 12.9583,
+      "step": 6508
+    },
+    {
+      "epoch": 1.1383169398907105,
+      "grad_norm": 0.8827652616122416,
+      "learning_rate": 0.0001425102648296147,
+      "loss": 12.7238,
+      "step": 6509
+    },
+    {
+      "epoch": 1.1384918032786886,
+      "grad_norm": 0.9298615409646995,
+      "learning_rate": 0.00014249317411859418,
+      "loss": 12.8962,
+      "step": 6510
+    },
+    {
+      "epoch": 1.1386666666666667,
+      "grad_norm": 0.8871983329699917,
+      "learning_rate": 0.0001424760818927417,
+      "loss": 12.8821,
+      "step": 6511
+    },
+    {
+      "epoch": 1.1388415300546448,
+      "grad_norm": 0.9981876545528537,
+      "learning_rate": 0.00014245898815266647,
+      "loss": 12.8398,
+      "step": 6512
+    },
+    {
+      "epoch": 1.139016393442623,
+      "grad_norm": 0.9844505909746346,
+      "learning_rate": 0.00014244189289897797,
+      "loss": 12.9484,
+      "step": 6513
+    },
+    {
+      "epoch": 1.139191256830601,
+      "grad_norm": 1.167437933489472,
+      "learning_rate": 0.00014242479613228558,
+      "loss": 13.0862,
+      "step": 6514
+    },
+    {
+      "epoch": 1.1393661202185792,
+      "grad_norm": 1.0833706732384536,
+      "learning_rate": 0.00014240769785319875,
+      "loss": 12.618,
+      "step": 6515
+    },
+    {
+      "epoch": 1.1395409836065573,
+      "grad_norm": 0.924644635459391,
+      "learning_rate": 0.000142390598062327,
+      "loss": 12.9121,
+      "step": 6516
+    },
+    {
+      "epoch": 1.1397158469945354,
+      "grad_norm": 0.9152488329821913,
+      "learning_rate": 0.00014237349676028002,
+      "loss": 13.0093,
+      "step": 6517
+    },
+    {
+      "epoch": 1.1398907103825136,
+      "grad_norm": 0.8931539074118952,
+      "learning_rate": 0.00014235639394766738,
+      "loss": 12.8834,
+      "step": 6518
+    },
+    {
+      "epoch": 1.140065573770492,
+      "grad_norm": 0.8713127063311402,
+      "learning_rate": 0.00014233928962509879,
+      "loss": 12.7861,
+      "step": 6519
+    },
+    {
+      "epoch": 1.14024043715847,
+      "grad_norm": 0.9570291169259808,
+      "learning_rate": 0.00014232218379318397,
+      "loss": 12.9803,
+      "step": 6520
+    },
+    {
+      "epoch": 1.1404153005464481,
+      "grad_norm": 0.9678029273821815,
+      "learning_rate": 0.00014230507645253277,
+      "loss": 12.9218,
+      "step": 6521
+    },
+    {
+      "epoch": 1.1405901639344262,
+      "grad_norm": 0.9096646561266358,
+      "learning_rate": 0.000142287967603755,
+      "loss": 12.9238,
+      "step": 6522
+    },
+    {
+      "epoch": 1.1407650273224044,
+      "grad_norm": 0.9509550117296877,
+      "learning_rate": 0.0001422708572474606,
+      "loss": 12.8814,
+      "step": 6523
+    },
+    {
+      "epoch": 1.1409398907103825,
+      "grad_norm": 0.9225505160360192,
+      "learning_rate": 0.00014225374538425953,
+      "loss": 12.6903,
+      "step": 6524
+    },
+    {
+      "epoch": 1.1411147540983606,
+      "grad_norm": 0.9923474524745476,
+      "learning_rate": 0.00014223663201476177,
+      "loss": 12.9389,
+      "step": 6525
+    },
+    {
+      "epoch": 1.1412896174863387,
+      "grad_norm": 1.0870546085688526,
+      "learning_rate": 0.00014221951713957749,
+      "loss": 12.7584,
+      "step": 6526
+    },
+    {
+      "epoch": 1.1414644808743168,
+      "grad_norm": 1.0593039533596944,
+      "learning_rate": 0.00014220240075931673,
+      "loss": 12.9774,
+      "step": 6527
+    },
+    {
+      "epoch": 1.1416393442622952,
+      "grad_norm": 0.9869994260603556,
+      "learning_rate": 0.0001421852828745897,
+      "loss": 12.8887,
+      "step": 6528
+    },
+    {
+      "epoch": 1.1418142076502733,
+      "grad_norm": 0.9451912187155117,
+      "learning_rate": 0.00014216816348600658,
+      "loss": 12.7161,
+      "step": 6529
+    },
+    {
+      "epoch": 1.1419890710382514,
+      "grad_norm": 0.9639996679986677,
+      "learning_rate": 0.00014215104259417775,
+      "loss": 12.8141,
+      "step": 6530
+    },
+    {
+      "epoch": 1.1421639344262295,
+      "grad_norm": 0.990878968835246,
+      "learning_rate": 0.00014213392019971345,
+      "loss": 12.9577,
+      "step": 6531
+    },
+    {
+      "epoch": 1.1423387978142077,
+      "grad_norm": 0.947558197397529,
+      "learning_rate": 0.00014211679630322417,
+      "loss": 12.4867,
+      "step": 6532
+    },
+    {
+      "epoch": 1.1425136612021858,
+      "grad_norm": 1.0837838605077936,
+      "learning_rate": 0.00014209967090532028,
+      "loss": 12.6717,
+      "step": 6533
+    },
+    {
+      "epoch": 1.142688524590164,
+      "grad_norm": 1.0656758830524606,
+      "learning_rate": 0.00014208254400661233,
+      "loss": 12.8119,
+      "step": 6534
+    },
+    {
+      "epoch": 1.142863387978142,
+      "grad_norm": 0.9893030235492781,
+      "learning_rate": 0.00014206541560771082,
+      "loss": 12.8167,
+      "step": 6535
+    },
+    {
+      "epoch": 1.1430382513661201,
+      "grad_norm": 0.9359430930456779,
+      "learning_rate": 0.00014204828570922644,
+      "loss": 12.842,
+      "step": 6536
+    },
+    {
+      "epoch": 1.1432131147540985,
+      "grad_norm": 0.8843059842711439,
+      "learning_rate": 0.00014203115431176975,
+      "loss": 12.4809,
+      "step": 6537
+    },
+    {
+      "epoch": 1.1433879781420766,
+      "grad_norm": 0.9664645142343307,
+      "learning_rate": 0.00014201402141595157,
+      "loss": 12.9659,
+      "step": 6538
+    },
+    {
+      "epoch": 1.1435628415300547,
+      "grad_norm": 0.8556287346029755,
+      "learning_rate": 0.0001419968870223826,
+      "loss": 12.4833,
+      "step": 6539
+    },
+    {
+      "epoch": 1.1437377049180328,
+      "grad_norm": 0.9376292982573017,
+      "learning_rate": 0.00014197975113167366,
+      "loss": 12.8292,
+      "step": 6540
+    },
+    {
+      "epoch": 1.143912568306011,
+      "grad_norm": 0.9726961485030232,
+      "learning_rate": 0.00014196261374443568,
+      "loss": 12.8491,
+      "step": 6541
+    },
+    {
+      "epoch": 1.144087431693989,
+      "grad_norm": 0.8963345881213066,
+      "learning_rate": 0.00014194547486127949,
+      "loss": 12.9398,
+      "step": 6542
+    },
+    {
+      "epoch": 1.1442622950819672,
+      "grad_norm": 0.9846870314900331,
+      "learning_rate": 0.00014192833448281616,
+      "loss": 12.7019,
+      "step": 6543
+    },
+    {
+      "epoch": 1.1444371584699453,
+      "grad_norm": 1.0374374897186391,
+      "learning_rate": 0.00014191119260965668,
+      "loss": 12.9157,
+      "step": 6544
+    },
+    {
+      "epoch": 1.1446120218579234,
+      "grad_norm": 0.8892242006578459,
+      "learning_rate": 0.00014189404924241217,
+      "loss": 12.3858,
+      "step": 6545
+    },
+    {
+      "epoch": 1.1447868852459016,
+      "grad_norm": 0.9145584017948469,
+      "learning_rate": 0.00014187690438169373,
+      "loss": 12.6983,
+      "step": 6546
+    },
+    {
+      "epoch": 1.1449617486338797,
+      "grad_norm": 0.8891046432066504,
+      "learning_rate": 0.0001418597580281126,
+      "loss": 12.7012,
+      "step": 6547
+    },
+    {
+      "epoch": 1.145136612021858,
+      "grad_norm": 0.9252519523852065,
+      "learning_rate": 0.00014184261018228,
+      "loss": 12.9811,
+      "step": 6548
+    },
+    {
+      "epoch": 1.1453114754098361,
+      "grad_norm": 0.894736818422037,
+      "learning_rate": 0.0001418254608448072,
+      "loss": 12.7437,
+      "step": 6549
+    },
+    {
+      "epoch": 1.1454863387978143,
+      "grad_norm": 0.8901828102418696,
+      "learning_rate": 0.00014180831001630564,
+      "loss": 12.8339,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1456612021857924,
+      "grad_norm": 0.8972269926387209,
+      "learning_rate": 0.00014179115769738661,
+      "loss": 12.9913,
+      "step": 6551
+    },
+    {
+      "epoch": 1.1458360655737705,
+      "grad_norm": 1.0119109960406014,
+      "learning_rate": 0.0001417740038886617,
+      "loss": 12.8073,
+      "step": 6552
+    },
+    {
+      "epoch": 1.1460109289617486,
+      "grad_norm": 0.9562213531326846,
+      "learning_rate": 0.00014175684859074234,
+      "loss": 12.9356,
+      "step": 6553
+    },
+    {
+      "epoch": 1.1461857923497267,
+      "grad_norm": 0.9494674229344139,
+      "learning_rate": 0.00014173969180424008,
+      "loss": 12.9822,
+      "step": 6554
+    },
+    {
+      "epoch": 1.1463606557377048,
+      "grad_norm": 0.9626769722578743,
+      "learning_rate": 0.00014172253352976664,
+      "loss": 12.6553,
+      "step": 6555
+    },
+    {
+      "epoch": 1.146535519125683,
+      "grad_norm": 0.9277990724590246,
+      "learning_rate": 0.0001417053737679336,
+      "loss": 12.6627,
+      "step": 6556
+    },
+    {
+      "epoch": 1.1467103825136613,
+      "grad_norm": 0.9218938060586452,
+      "learning_rate": 0.0001416882125193527,
+      "loss": 12.6898,
+      "step": 6557
+    },
+    {
+      "epoch": 1.1468852459016394,
+      "grad_norm": 1.0759749111176764,
+      "learning_rate": 0.00014167104978463576,
+      "loss": 12.7576,
+      "step": 6558
+    },
+    {
+      "epoch": 1.1470601092896175,
+      "grad_norm": 0.9384486860173089,
+      "learning_rate": 0.00014165388556439454,
+      "loss": 12.8943,
+      "step": 6559
+    },
+    {
+      "epoch": 1.1472349726775957,
+      "grad_norm": 0.857205569167023,
+      "learning_rate": 0.00014163671985924098,
+      "loss": 12.7978,
+      "step": 6560
+    },
+    {
+      "epoch": 1.1474098360655738,
+      "grad_norm": 0.9230533832890756,
+      "learning_rate": 0.000141619552669787,
+      "loss": 12.9419,
+      "step": 6561
+    },
+    {
+      "epoch": 1.147584699453552,
+      "grad_norm": 0.9771595741996724,
+      "learning_rate": 0.00014160238399664463,
+      "loss": 12.8992,
+      "step": 6562
+    },
+    {
+      "epoch": 1.14775956284153,
+      "grad_norm": 0.9440978302697506,
+      "learning_rate": 0.00014158521384042584,
+      "loss": 12.9285,
+      "step": 6563
+    },
+    {
+      "epoch": 1.1479344262295081,
+      "grad_norm": 0.8757365956972956,
+      "learning_rate": 0.0001415680422017428,
+      "loss": 12.7493,
+      "step": 6564
+    },
+    {
+      "epoch": 1.1481092896174863,
+      "grad_norm": 0.9746164361440127,
+      "learning_rate": 0.00014155086908120758,
+      "loss": 12.9188,
+      "step": 6565
+    },
+    {
+      "epoch": 1.1482841530054646,
+      "grad_norm": 0.8706689857226375,
+      "learning_rate": 0.00014153369447943246,
+      "loss": 12.7016,
+      "step": 6566
+    },
+    {
+      "epoch": 1.1484590163934425,
+      "grad_norm": 1.0049201460925261,
+      "learning_rate": 0.00014151651839702965,
+      "loss": 12.7805,
+      "step": 6567
+    },
+    {
+      "epoch": 1.1486338797814208,
+      "grad_norm": 0.966917027799729,
+      "learning_rate": 0.00014149934083461144,
+      "loss": 12.839,
+      "step": 6568
+    },
+    {
+      "epoch": 1.148808743169399,
+      "grad_norm": 0.9952337281638914,
+      "learning_rate": 0.00014148216179279022,
+      "loss": 12.8763,
+      "step": 6569
+    },
+    {
+      "epoch": 1.148983606557377,
+      "grad_norm": 0.8487357061131254,
+      "learning_rate": 0.00014146498127217842,
+      "loss": 12.8111,
+      "step": 6570
+    },
+    {
+      "epoch": 1.1491584699453552,
+      "grad_norm": 0.8354336770114451,
+      "learning_rate": 0.0001414477992733885,
+      "loss": 12.6001,
+      "step": 6571
+    },
+    {
+      "epoch": 1.1493333333333333,
+      "grad_norm": 0.9632406687723454,
+      "learning_rate": 0.00014143061579703292,
+      "loss": 12.9171,
+      "step": 6572
+    },
+    {
+      "epoch": 1.1495081967213114,
+      "grad_norm": 0.9709074763599246,
+      "learning_rate": 0.00014141343084372434,
+      "loss": 13.0112,
+      "step": 6573
+    },
+    {
+      "epoch": 1.1496830601092896,
+      "grad_norm": 0.8454967753134516,
+      "learning_rate": 0.00014139624441407528,
+      "loss": 12.8865,
+      "step": 6574
+    },
+    {
+      "epoch": 1.1498579234972677,
+      "grad_norm": 0.8869247291226243,
+      "learning_rate": 0.00014137905650869853,
+      "loss": 12.9139,
+      "step": 6575
+    },
+    {
+      "epoch": 1.1500327868852458,
+      "grad_norm": 1.0332750541960865,
+      "learning_rate": 0.00014136186712820673,
+      "loss": 12.9513,
+      "step": 6576
+    },
+    {
+      "epoch": 1.1502076502732241,
+      "grad_norm": 1.022088825907778,
+      "learning_rate": 0.0001413446762732127,
+      "loss": 12.824,
+      "step": 6577
+    },
+    {
+      "epoch": 1.1503825136612023,
+      "grad_norm": 0.9719706301732817,
+      "learning_rate": 0.00014132748394432926,
+      "loss": 13.0015,
+      "step": 6578
+    },
+    {
+      "epoch": 1.1505573770491804,
+      "grad_norm": 1.0606693908837481,
+      "learning_rate": 0.0001413102901421693,
+      "loss": 12.9814,
+      "step": 6579
+    },
+    {
+      "epoch": 1.1507322404371585,
+      "grad_norm": 0.9435148225254092,
+      "learning_rate": 0.00014129309486734577,
+      "loss": 12.6371,
+      "step": 6580
+    },
+    {
+      "epoch": 1.1509071038251366,
+      "grad_norm": 1.0067123672760994,
+      "learning_rate": 0.00014127589812047166,
+      "loss": 12.7711,
+      "step": 6581
+    },
+    {
+      "epoch": 1.1510819672131147,
+      "grad_norm": 1.0306532669966713,
+      "learning_rate": 0.00014125869990216,
+      "loss": 13.0994,
+      "step": 6582
+    },
+    {
+      "epoch": 1.1512568306010929,
+      "grad_norm": 0.9739820551413995,
+      "learning_rate": 0.00014124150021302393,
+      "loss": 12.7936,
+      "step": 6583
+    },
+    {
+      "epoch": 1.151431693989071,
+      "grad_norm": 1.0382739078347503,
+      "learning_rate": 0.0001412242990536765,
+      "loss": 12.9266,
+      "step": 6584
+    },
+    {
+      "epoch": 1.151606557377049,
+      "grad_norm": 1.0300633935139905,
+      "learning_rate": 0.000141207096424731,
+      "loss": 12.6662,
+      "step": 6585
+    },
+    {
+      "epoch": 1.1517814207650274,
+      "grad_norm": 1.0792304914849236,
+      "learning_rate": 0.0001411898923268007,
+      "loss": 12.9954,
+      "step": 6586
+    },
+    {
+      "epoch": 1.1519562841530055,
+      "grad_norm": 0.9447794106714564,
+      "learning_rate": 0.00014117268676049882,
+      "loss": 12.9268,
+      "step": 6587
+    },
+    {
+      "epoch": 1.1521311475409837,
+      "grad_norm": 0.9477753737148822,
+      "learning_rate": 0.00014115547972643874,
+      "loss": 12.6516,
+      "step": 6588
+    },
+    {
+      "epoch": 1.1523060109289618,
+      "grad_norm": 0.9113397629229537,
+      "learning_rate": 0.0001411382712252339,
+      "loss": 12.8098,
+      "step": 6589
+    },
+    {
+      "epoch": 1.15248087431694,
+      "grad_norm": 0.9931153881907534,
+      "learning_rate": 0.0001411210612574978,
+      "loss": 12.597,
+      "step": 6590
+    },
+    {
+      "epoch": 1.152655737704918,
+      "grad_norm": 0.9240203337818456,
+      "learning_rate": 0.00014110384982384387,
+      "loss": 12.9831,
+      "step": 6591
+    },
+    {
+      "epoch": 1.1528306010928961,
+      "grad_norm": 0.8810012983795746,
+      "learning_rate": 0.00014108663692488578,
+      "loss": 12.8074,
+      "step": 6592
+    },
+    {
+      "epoch": 1.1530054644808743,
+      "grad_norm": 0.9879823442414903,
+      "learning_rate": 0.00014106942256123702,
+      "loss": 12.759,
+      "step": 6593
+    },
+    {
+      "epoch": 1.1531803278688524,
+      "grad_norm": 0.9226007472125703,
+      "learning_rate": 0.00014105220673351135,
+      "loss": 12.7984,
+      "step": 6594
+    },
+    {
+      "epoch": 1.1533551912568307,
+      "grad_norm": 1.0226291520707427,
+      "learning_rate": 0.00014103498944232249,
+      "loss": 12.996,
+      "step": 6595
+    },
+    {
+      "epoch": 1.1535300546448086,
+      "grad_norm": 0.9100182412270365,
+      "learning_rate": 0.0001410177706882842,
+      "loss": 13.0264,
+      "step": 6596
+    },
+    {
+      "epoch": 1.153704918032787,
+      "grad_norm": 0.9172522985035899,
+      "learning_rate": 0.00014100055047201027,
+      "loss": 12.8712,
+      "step": 6597
+    },
+    {
+      "epoch": 1.153879781420765,
+      "grad_norm": 0.9463220240540026,
+      "learning_rate": 0.00014098332879411464,
+      "loss": 12.6966,
+      "step": 6598
+    },
+    {
+      "epoch": 1.1540546448087432,
+      "grad_norm": 0.9813674186363909,
+      "learning_rate": 0.00014096610565521122,
+      "loss": 12.7214,
+      "step": 6599
+    },
+    {
+      "epoch": 1.1542295081967213,
+      "grad_norm": 0.9762147451263238,
+      "learning_rate": 0.00014094888105591405,
+      "loss": 12.7122,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1544043715846994,
+      "grad_norm": 0.9501528812727771,
+      "learning_rate": 0.00014093165499683708,
+      "loss": 12.762,
+      "step": 6601
+    },
+    {
+      "epoch": 1.1545792349726776,
+      "grad_norm": 0.9096364505454195,
+      "learning_rate": 0.0001409144274785944,
+      "loss": 12.9704,
+      "step": 6602
+    },
+    {
+      "epoch": 1.1547540983606557,
+      "grad_norm": 0.9400453345611935,
+      "learning_rate": 0.00014089719850180022,
+      "loss": 12.6981,
+      "step": 6603
+    },
+    {
+      "epoch": 1.1549289617486338,
+      "grad_norm": 0.9821667429682015,
+      "learning_rate": 0.00014087996806706868,
+      "loss": 12.8193,
+      "step": 6604
+    },
+    {
+      "epoch": 1.155103825136612,
+      "grad_norm": 1.060545947442995,
+      "learning_rate": 0.00014086273617501402,
+      "loss": 12.7697,
+      "step": 6605
+    },
+    {
+      "epoch": 1.1552786885245903,
+      "grad_norm": 0.9247546173632069,
+      "learning_rate": 0.00014084550282625056,
+      "loss": 12.8991,
+      "step": 6606
+    },
+    {
+      "epoch": 1.1554535519125684,
+      "grad_norm": 1.141190353446948,
+      "learning_rate": 0.00014082826802139272,
+      "loss": 12.9346,
+      "step": 6607
+    },
+    {
+      "epoch": 1.1556284153005465,
+      "grad_norm": 1.0387605619679983,
+      "learning_rate": 0.00014081103176105477,
+      "loss": 12.6112,
+      "step": 6608
+    },
+    {
+      "epoch": 1.1558032786885246,
+      "grad_norm": 1.1480963092580951,
+      "learning_rate": 0.00014079379404585124,
+      "loss": 13.0622,
+      "step": 6609
+    },
+    {
+      "epoch": 1.1559781420765027,
+      "grad_norm": 0.9279846954186188,
+      "learning_rate": 0.00014077655487639658,
+      "loss": 12.6353,
+      "step": 6610
+    },
+    {
+      "epoch": 1.1561530054644809,
+      "grad_norm": 1.1135527456781857,
+      "learning_rate": 0.0001407593142533054,
+      "loss": 12.616,
+      "step": 6611
+    },
+    {
+      "epoch": 1.156327868852459,
+      "grad_norm": 0.869994000291676,
+      "learning_rate": 0.0001407420721771923,
+      "loss": 12.8149,
+      "step": 6612
+    },
+    {
+      "epoch": 1.156502732240437,
+      "grad_norm": 0.9769627890089269,
+      "learning_rate": 0.00014072482864867193,
+      "loss": 12.7982,
+      "step": 6613
+    },
+    {
+      "epoch": 1.1566775956284152,
+      "grad_norm": 0.9329823249713404,
+      "learning_rate": 0.00014070758366835899,
+      "loss": 12.7153,
+      "step": 6614
+    },
+    {
+      "epoch": 1.1568524590163936,
+      "grad_norm": 0.9470595934050233,
+      "learning_rate": 0.00014069033723686827,
+      "loss": 12.8549,
+      "step": 6615
+    },
+    {
+      "epoch": 1.1570273224043717,
+      "grad_norm": 0.9109990222735924,
+      "learning_rate": 0.00014067308935481452,
+      "loss": 12.8253,
+      "step": 6616
+    },
+    {
+      "epoch": 1.1572021857923498,
+      "grad_norm": 0.9515039694634606,
+      "learning_rate": 0.0001406558400228127,
+      "loss": 12.792,
+      "step": 6617
+    },
+    {
+      "epoch": 1.157377049180328,
+      "grad_norm": 0.8933904951567768,
+      "learning_rate": 0.0001406385892414777,
+      "loss": 12.66,
+      "step": 6618
+    },
+    {
+      "epoch": 1.157551912568306,
+      "grad_norm": 0.8977340689748541,
+      "learning_rate": 0.00014062133701142447,
+      "loss": 12.7362,
+      "step": 6619
+    },
+    {
+      "epoch": 1.1577267759562841,
+      "grad_norm": 0.8672536307610854,
+      "learning_rate": 0.00014060408333326803,
+      "loss": 12.7498,
+      "step": 6620
+    },
+    {
+      "epoch": 1.1579016393442623,
+      "grad_norm": 1.1089646151632497,
+      "learning_rate": 0.00014058682820762344,
+      "loss": 13.035,
+      "step": 6621
+    },
+    {
+      "epoch": 1.1580765027322404,
+      "grad_norm": 0.9717951324100851,
+      "learning_rate": 0.00014056957163510584,
+      "loss": 13.0343,
+      "step": 6622
+    },
+    {
+      "epoch": 1.1582513661202185,
+      "grad_norm": 1.0867307001968491,
+      "learning_rate": 0.00014055231361633042,
+      "loss": 12.9753,
+      "step": 6623
+    },
+    {
+      "epoch": 1.1584262295081966,
+      "grad_norm": 0.8994012237529518,
+      "learning_rate": 0.0001405350541519124,
+      "loss": 12.8432,
+      "step": 6624
+    },
+    {
+      "epoch": 1.1586010928961747,
+      "grad_norm": 1.0706202520015078,
+      "learning_rate": 0.00014051779324246704,
+      "loss": 12.6886,
+      "step": 6625
+    },
+    {
+      "epoch": 1.158775956284153,
+      "grad_norm": 0.9143683308731937,
+      "learning_rate": 0.00014050053088860974,
+      "loss": 12.704,
+      "step": 6626
+    },
+    {
+      "epoch": 1.1589508196721312,
+      "grad_norm": 1.1677191764384576,
+      "learning_rate": 0.0001404832670909558,
+      "loss": 12.9734,
+      "step": 6627
+    },
+    {
+      "epoch": 1.1591256830601093,
+      "grad_norm": 1.033573892450421,
+      "learning_rate": 0.0001404660018501207,
+      "loss": 12.7946,
+      "step": 6628
+    },
+    {
+      "epoch": 1.1593005464480874,
+      "grad_norm": 1.0596028526376982,
+      "learning_rate": 0.0001404487351667199,
+      "loss": 12.9314,
+      "step": 6629
+    },
+    {
+      "epoch": 1.1594754098360656,
+      "grad_norm": 0.9748260741017685,
+      "learning_rate": 0.00014043146704136898,
+      "loss": 12.7896,
+      "step": 6630
+    },
+    {
+      "epoch": 1.1596502732240437,
+      "grad_norm": 1.044952313902316,
+      "learning_rate": 0.00014041419747468347,
+      "loss": 12.6461,
+      "step": 6631
+    },
+    {
+      "epoch": 1.1598251366120218,
+      "grad_norm": 1.0138054830341148,
+      "learning_rate": 0.00014039692646727908,
+      "loss": 12.9226,
+      "step": 6632
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.98886176118161,
+      "learning_rate": 0.0001403796540197714,
+      "loss": 12.6651,
+      "step": 6633
+    },
+    {
+      "epoch": 1.160174863387978,
+      "grad_norm": 1.042020696411358,
+      "learning_rate": 0.00014036238013277624,
+      "loss": 12.8557,
+      "step": 6634
+    },
+    {
+      "epoch": 1.1603497267759564,
+      "grad_norm": 1.0391314192653303,
+      "learning_rate": 0.00014034510480690943,
+      "loss": 12.8563,
+      "step": 6635
+    },
+    {
+      "epoch": 1.1605245901639345,
+      "grad_norm": 1.0359136086580323,
+      "learning_rate": 0.00014032782804278675,
+      "loss": 12.7625,
+      "step": 6636
+    },
+    {
+      "epoch": 1.1606994535519126,
+      "grad_norm": 0.9526361263308267,
+      "learning_rate": 0.00014031054984102413,
+      "loss": 12.7177,
+      "step": 6637
+    },
+    {
+      "epoch": 1.1608743169398907,
+      "grad_norm": 1.1367985945548438,
+      "learning_rate": 0.00014029327020223748,
+      "loss": 12.8593,
+      "step": 6638
+    },
+    {
+      "epoch": 1.1610491803278689,
+      "grad_norm": 1.043537150654818,
+      "learning_rate": 0.00014027598912704284,
+      "loss": 12.8003,
+      "step": 6639
+    },
+    {
+      "epoch": 1.161224043715847,
+      "grad_norm": 1.0521837334199893,
+      "learning_rate": 0.00014025870661605623,
+      "loss": 12.967,
+      "step": 6640
+    },
+    {
+      "epoch": 1.161398907103825,
+      "grad_norm": 0.9657699181124805,
+      "learning_rate": 0.0001402414226698938,
+      "loss": 12.5411,
+      "step": 6641
+    },
+    {
+      "epoch": 1.1615737704918032,
+      "grad_norm": 1.0135364335321777,
+      "learning_rate": 0.00014022413728917164,
+      "loss": 12.7189,
+      "step": 6642
+    },
+    {
+      "epoch": 1.1617486338797813,
+      "grad_norm": 0.9680963012039417,
+      "learning_rate": 0.00014020685047450598,
+      "loss": 12.9443,
+      "step": 6643
+    },
+    {
+      "epoch": 1.1619234972677597,
+      "grad_norm": 0.980970816659346,
+      "learning_rate": 0.0001401895622265131,
+      "loss": 13.0092,
+      "step": 6644
+    },
+    {
+      "epoch": 1.1620983606557378,
+      "grad_norm": 0.9713343806842923,
+      "learning_rate": 0.00014017227254580926,
+      "loss": 12.5604,
+      "step": 6645
+    },
+    {
+      "epoch": 1.162273224043716,
+      "grad_norm": 1.0351421626318307,
+      "learning_rate": 0.00014015498143301087,
+      "loss": 12.8067,
+      "step": 6646
+    },
+    {
+      "epoch": 1.162448087431694,
+      "grad_norm": 0.9806398703467815,
+      "learning_rate": 0.00014013768888873428,
+      "loss": 12.8386,
+      "step": 6647
+    },
+    {
+      "epoch": 1.1626229508196722,
+      "grad_norm": 1.0928287158208978,
+      "learning_rate": 0.000140120394913596,
+      "loss": 13.0043,
+      "step": 6648
+    },
+    {
+      "epoch": 1.1627978142076503,
+      "grad_norm": 0.9133434448493266,
+      "learning_rate": 0.00014010309950821253,
+      "loss": 12.6629,
+      "step": 6649
+    },
+    {
+      "epoch": 1.1629726775956284,
+      "grad_norm": 0.9587442032649144,
+      "learning_rate": 0.00014008580267320037,
+      "loss": 12.6631,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1631475409836065,
+      "grad_norm": 1.158268928443757,
+      "learning_rate": 0.00014006850440917619,
+      "loss": 13.0065,
+      "step": 6651
+    },
+    {
+      "epoch": 1.1633224043715846,
+      "grad_norm": 1.0674603086803687,
+      "learning_rate": 0.00014005120471675665,
+      "loss": 12.9968,
+      "step": 6652
+    },
+    {
+      "epoch": 1.1634972677595627,
+      "grad_norm": 0.9248823199518936,
+      "learning_rate": 0.00014003390359655846,
+      "loss": 12.6934,
+      "step": 6653
+    },
+    {
+      "epoch": 1.1636721311475409,
+      "grad_norm": 0.920476494466567,
+      "learning_rate": 0.0001400166010491984,
+      "loss": 12.7691,
+      "step": 6654
+    },
+    {
+      "epoch": 1.1638469945355192,
+      "grad_norm": 1.0235780190141632,
+      "learning_rate": 0.00013999929707529324,
+      "loss": 12.9446,
+      "step": 6655
+    },
+    {
+      "epoch": 1.1640218579234973,
+      "grad_norm": 0.9466706869021523,
+      "learning_rate": 0.0001399819916754599,
+      "loss": 12.9737,
+      "step": 6656
+    },
+    {
+      "epoch": 1.1641967213114754,
+      "grad_norm": 0.9843804801692284,
+      "learning_rate": 0.00013996468485031523,
+      "loss": 12.7,
+      "step": 6657
+    },
+    {
+      "epoch": 1.1643715846994536,
+      "grad_norm": 0.8838601147784251,
+      "learning_rate": 0.00013994737660047627,
+      "loss": 12.8396,
+      "step": 6658
+    },
+    {
+      "epoch": 1.1645464480874317,
+      "grad_norm": 0.8802110093018689,
+      "learning_rate": 0.00013993006692656002,
+      "loss": 12.8409,
+      "step": 6659
+    },
+    {
+      "epoch": 1.1647213114754098,
+      "grad_norm": 0.9767056489322163,
+      "learning_rate": 0.0001399127558291835,
+      "loss": 12.8958,
+      "step": 6660
+    },
+    {
+      "epoch": 1.164896174863388,
+      "grad_norm": 1.0339719002468983,
+      "learning_rate": 0.00013989544330896388,
+      "loss": 12.8161,
+      "step": 6661
+    },
+    {
+      "epoch": 1.165071038251366,
+      "grad_norm": 1.1189331257008102,
+      "learning_rate": 0.00013987812936651833,
+      "loss": 12.8348,
+      "step": 6662
+    },
+    {
+      "epoch": 1.1652459016393442,
+      "grad_norm": 1.1892508063311784,
+      "learning_rate": 0.00013986081400246405,
+      "loss": 12.9854,
+      "step": 6663
+    },
+    {
+      "epoch": 1.1654207650273225,
+      "grad_norm": 1.0010086412137,
+      "learning_rate": 0.00013984349721741834,
+      "loss": 12.7628,
+      "step": 6664
+    },
+    {
+      "epoch": 1.1655956284153006,
+      "grad_norm": 0.9915590090558417,
+      "learning_rate": 0.00013982617901199852,
+      "loss": 12.975,
+      "step": 6665
+    },
+    {
+      "epoch": 1.1657704918032787,
+      "grad_norm": 1.0256176200239535,
+      "learning_rate": 0.0001398088593868219,
+      "loss": 12.775,
+      "step": 6666
+    },
+    {
+      "epoch": 1.1659453551912569,
+      "grad_norm": 1.0943682681784974,
+      "learning_rate": 0.000139791538342506,
+      "loss": 12.8523,
+      "step": 6667
+    },
+    {
+      "epoch": 1.166120218579235,
+      "grad_norm": 1.0107630989226442,
+      "learning_rate": 0.0001397742158796682,
+      "loss": 12.8658,
+      "step": 6668
+    },
+    {
+      "epoch": 1.166295081967213,
+      "grad_norm": 1.086517615458311,
+      "learning_rate": 0.0001397568919989261,
+      "loss": 13.0323,
+      "step": 6669
+    },
+    {
+      "epoch": 1.1664699453551912,
+      "grad_norm": 0.933382372833015,
+      "learning_rate": 0.00013973956670089725,
+      "loss": 13.049,
+      "step": 6670
+    },
+    {
+      "epoch": 1.1666448087431693,
+      "grad_norm": 0.995729263346041,
+      "learning_rate": 0.00013972223998619932,
+      "loss": 12.7585,
+      "step": 6671
+    },
+    {
+      "epoch": 1.1668196721311475,
+      "grad_norm": 1.0006393068150192,
+      "learning_rate": 0.00013970491185544986,
+      "loss": 12.7609,
+      "step": 6672
+    },
+    {
+      "epoch": 1.1669945355191258,
+      "grad_norm": 0.9739102155021805,
+      "learning_rate": 0.00013968758230926674,
+      "loss": 12.8457,
+      "step": 6673
+    },
+    {
+      "epoch": 1.167169398907104,
+      "grad_norm": 0.9068720379073797,
+      "learning_rate": 0.0001396702513482677,
+      "loss": 12.9335,
+      "step": 6674
+    },
+    {
+      "epoch": 1.167344262295082,
+      "grad_norm": 1.0258225575188826,
+      "learning_rate": 0.0001396529189730705,
+      "loss": 12.6661,
+      "step": 6675
+    },
+    {
+      "epoch": 1.1675191256830602,
+      "grad_norm": 0.9754879098103444,
+      "learning_rate": 0.0001396355851842931,
+      "loss": 12.7556,
+      "step": 6676
+    },
+    {
+      "epoch": 1.1676939890710383,
+      "grad_norm": 0.9426033725784796,
+      "learning_rate": 0.0001396182499825534,
+      "loss": 12.8007,
+      "step": 6677
+    },
+    {
+      "epoch": 1.1678688524590164,
+      "grad_norm": 1.2406283543729868,
+      "learning_rate": 0.00013960091336846935,
+      "loss": 12.8176,
+      "step": 6678
+    },
+    {
+      "epoch": 1.1680437158469945,
+      "grad_norm": 0.991812489188756,
+      "learning_rate": 0.000139583575342659,
+      "loss": 12.6974,
+      "step": 6679
+    },
+    {
+      "epoch": 1.1682185792349726,
+      "grad_norm": 0.9276505226635766,
+      "learning_rate": 0.00013956623590574048,
+      "loss": 12.5221,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1683934426229508,
+      "grad_norm": 1.229060875978901,
+      "learning_rate": 0.00013954889505833184,
+      "loss": 12.6798,
+      "step": 6681
+    },
+    {
+      "epoch": 1.1685683060109289,
+      "grad_norm": 1.0759563438434354,
+      "learning_rate": 0.00013953155280105133,
+      "loss": 12.9437,
+      "step": 6682
+    },
+    {
+      "epoch": 1.168743169398907,
+      "grad_norm": 0.9946275916189473,
+      "learning_rate": 0.00013951420913451714,
+      "loss": 12.8516,
+      "step": 6683
+    },
+    {
+      "epoch": 1.1689180327868853,
+      "grad_norm": 1.0651756608759657,
+      "learning_rate": 0.00013949686405934756,
+      "loss": 13.0381,
+      "step": 6684
+    },
+    {
+      "epoch": 1.1690928961748634,
+      "grad_norm": 1.0935273636055007,
+      "learning_rate": 0.0001394795175761609,
+      "loss": 12.4966,
+      "step": 6685
+    },
+    {
+      "epoch": 1.1692677595628416,
+      "grad_norm": 1.0554862805198368,
+      "learning_rate": 0.0001394621696855756,
+      "loss": 12.8798,
+      "step": 6686
+    },
+    {
+      "epoch": 1.1694426229508197,
+      "grad_norm": 1.0693657675516417,
+      "learning_rate": 0.00013944482038821002,
+      "loss": 12.8673,
+      "step": 6687
+    },
+    {
+      "epoch": 1.1696174863387978,
+      "grad_norm": 1.169809880282687,
+      "learning_rate": 0.0001394274696846827,
+      "loss": 12.924,
+      "step": 6688
+    },
+    {
+      "epoch": 1.169792349726776,
+      "grad_norm": 1.0767993308223291,
+      "learning_rate": 0.00013941011757561216,
+      "loss": 12.7103,
+      "step": 6689
+    },
+    {
+      "epoch": 1.169967213114754,
+      "grad_norm": 0.9922636889020657,
+      "learning_rate": 0.00013939276406161696,
+      "loss": 12.7794,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1701420765027322,
+      "grad_norm": 1.0360132398288748,
+      "learning_rate": 0.00013937540914331576,
+      "loss": 12.791,
+      "step": 6691
+    },
+    {
+      "epoch": 1.1703169398907103,
+      "grad_norm": 0.9857173595147908,
+      "learning_rate": 0.00013935805282132724,
+      "loss": 12.7123,
+      "step": 6692
+    },
+    {
+      "epoch": 1.1704918032786886,
+      "grad_norm": 0.9660454906543523,
+      "learning_rate": 0.00013934069509627007,
+      "loss": 12.6918,
+      "step": 6693
+    },
+    {
+      "epoch": 1.1706666666666667,
+      "grad_norm": 0.9242966696517444,
+      "learning_rate": 0.00013932333596876318,
+      "loss": 12.7201,
+      "step": 6694
+    },
+    {
+      "epoch": 1.1708415300546449,
+      "grad_norm": 0.971847112404912,
+      "learning_rate": 0.00013930597543942522,
+      "loss": 12.6096,
+      "step": 6695
+    },
+    {
+      "epoch": 1.171016393442623,
+      "grad_norm": 1.0322549881723733,
+      "learning_rate": 0.00013928861350887516,
+      "loss": 12.8146,
+      "step": 6696
+    },
+    {
+      "epoch": 1.171191256830601,
+      "grad_norm": 1.0169043068896255,
+      "learning_rate": 0.00013927125017773197,
+      "loss": 12.6407,
+      "step": 6697
+    },
+    {
+      "epoch": 1.1713661202185792,
+      "grad_norm": 0.9557373534265272,
+      "learning_rate": 0.0001392538854466146,
+      "loss": 12.7546,
+      "step": 6698
+    },
+    {
+      "epoch": 1.1715409836065573,
+      "grad_norm": 1.0851093026044367,
+      "learning_rate": 0.00013923651931614206,
+      "loss": 12.9191,
+      "step": 6699
+    },
+    {
+      "epoch": 1.1717158469945355,
+      "grad_norm": 1.0416248369234264,
+      "learning_rate": 0.00013921915178693344,
+      "loss": 12.5326,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1718907103825136,
+      "grad_norm": 1.1690939199907606,
+      "learning_rate": 0.0001392017828596079,
+      "loss": 12.9659,
+      "step": 6701
+    },
+    {
+      "epoch": 1.172065573770492,
+      "grad_norm": 1.0341537104236254,
+      "learning_rate": 0.00013918441253478459,
+      "loss": 12.8518,
+      "step": 6702
+    },
+    {
+      "epoch": 1.17224043715847,
+      "grad_norm": 1.0851954238905426,
+      "learning_rate": 0.00013916704081308278,
+      "loss": 12.6264,
+      "step": 6703
+    },
+    {
+      "epoch": 1.1724153005464482,
+      "grad_norm": 1.0062604979501832,
+      "learning_rate": 0.0001391496676951217,
+      "loss": 12.8608,
+      "step": 6704
+    },
+    {
+      "epoch": 1.1725901639344263,
+      "grad_norm": 1.2076675320952066,
+      "learning_rate": 0.0001391322931815207,
+      "loss": 12.6503,
+      "step": 6705
+    },
+    {
+      "epoch": 1.1727650273224044,
+      "grad_norm": 0.9879704456868904,
+      "learning_rate": 0.00013911491727289918,
+      "loss": 12.6198,
+      "step": 6706
+    },
+    {
+      "epoch": 1.1729398907103825,
+      "grad_norm": 0.9740860808692555,
+      "learning_rate": 0.00013909753996987656,
+      "loss": 12.6111,
+      "step": 6707
+    },
+    {
+      "epoch": 1.1731147540983606,
+      "grad_norm": 0.928250699006588,
+      "learning_rate": 0.00013908016127307232,
+      "loss": 12.5194,
+      "step": 6708
+    },
+    {
+      "epoch": 1.1732896174863388,
+      "grad_norm": 1.0157103116049193,
+      "learning_rate": 0.00013906278118310604,
+      "loss": 12.9143,
+      "step": 6709
+    },
+    {
+      "epoch": 1.1734644808743169,
+      "grad_norm": 0.9454333319275028,
+      "learning_rate": 0.0001390453997005972,
+      "loss": 12.941,
+      "step": 6710
+    },
+    {
+      "epoch": 1.173639344262295,
+      "grad_norm": 0.9677277815884573,
+      "learning_rate": 0.00013902801682616549,
+      "loss": 12.5767,
+      "step": 6711
+    },
+    {
+      "epoch": 1.1738142076502731,
+      "grad_norm": 0.9477466414720637,
+      "learning_rate": 0.00013901063256043056,
+      "loss": 12.8066,
+      "step": 6712
+    },
+    {
+      "epoch": 1.1739890710382515,
+      "grad_norm": 0.9243192129888016,
+      "learning_rate": 0.0001389932469040122,
+      "loss": 12.9999,
+      "step": 6713
+    },
+    {
+      "epoch": 1.1741639344262296,
+      "grad_norm": 1.1236580728543566,
+      "learning_rate": 0.0001389758598575301,
+      "loss": 12.4574,
+      "step": 6714
+    },
+    {
+      "epoch": 1.1743387978142077,
+      "grad_norm": 1.1264517192730779,
+      "learning_rate": 0.00013895847142160415,
+      "loss": 12.6287,
+      "step": 6715
+    },
+    {
+      "epoch": 1.1745136612021858,
+      "grad_norm": 1.0389601612096464,
+      "learning_rate": 0.00013894108159685423,
+      "loss": 12.7819,
+      "step": 6716
+    },
+    {
+      "epoch": 1.174688524590164,
+      "grad_norm": 1.055331730207508,
+      "learning_rate": 0.00013892369038390022,
+      "loss": 12.9206,
+      "step": 6717
+    },
+    {
+      "epoch": 1.174863387978142,
+      "grad_norm": 0.946865387433633,
+      "learning_rate": 0.00013890629778336215,
+      "loss": 12.9063,
+      "step": 6718
+    },
+    {
+      "epoch": 1.1750382513661202,
+      "grad_norm": 0.8916534401148509,
+      "learning_rate": 0.00013888890379586003,
+      "loss": 12.8898,
+      "step": 6719
+    },
+    {
+      "epoch": 1.1752131147540983,
+      "grad_norm": 0.9583343347796306,
+      "learning_rate": 0.0001388715084220139,
+      "loss": 12.6606,
+      "step": 6720
+    },
+    {
+      "epoch": 1.1753879781420764,
+      "grad_norm": 0.9203964447051272,
+      "learning_rate": 0.0001388541116624439,
+      "loss": 12.9199,
+      "step": 6721
+    },
+    {
+      "epoch": 1.1755628415300547,
+      "grad_norm": 0.9368892392416879,
+      "learning_rate": 0.00013883671351777027,
+      "loss": 12.6742,
+      "step": 6722
+    },
+    {
+      "epoch": 1.1757377049180329,
+      "grad_norm": 0.9662246649055902,
+      "learning_rate": 0.00013881931398861315,
+      "loss": 12.8001,
+      "step": 6723
+    },
+    {
+      "epoch": 1.175912568306011,
+      "grad_norm": 1.0158497350310232,
+      "learning_rate": 0.0001388019130755928,
+      "loss": 12.9007,
+      "step": 6724
+    },
+    {
+      "epoch": 1.176087431693989,
+      "grad_norm": 0.9680838940102812,
+      "learning_rate": 0.00013878451077932962,
+      "loss": 12.9357,
+      "step": 6725
+    },
+    {
+      "epoch": 1.1762622950819672,
+      "grad_norm": 0.9659835435982178,
+      "learning_rate": 0.00013876710710044397,
+      "loss": 12.9283,
+      "step": 6726
+    },
+    {
+      "epoch": 1.1764371584699453,
+      "grad_norm": 1.0088883919470208,
+      "learning_rate": 0.0001387497020395562,
+      "loss": 12.813,
+      "step": 6727
+    },
+    {
+      "epoch": 1.1766120218579235,
+      "grad_norm": 0.9721428411274903,
+      "learning_rate": 0.00013873229559728685,
+      "loss": 13.0366,
+      "step": 6728
+    },
+    {
+      "epoch": 1.1767868852459016,
+      "grad_norm": 0.9535369972952754,
+      "learning_rate": 0.00013871488777425644,
+      "loss": 13.0352,
+      "step": 6729
+    },
+    {
+      "epoch": 1.1769617486338797,
+      "grad_norm": 1.0030894245218915,
+      "learning_rate": 0.00013869747857108547,
+      "loss": 12.6655,
+      "step": 6730
+    },
+    {
+      "epoch": 1.177136612021858,
+      "grad_norm": 0.9118727979043016,
+      "learning_rate": 0.00013868006798839467,
+      "loss": 12.9318,
+      "step": 6731
+    },
+    {
+      "epoch": 1.1773114754098362,
+      "grad_norm": 1.0078147427975528,
+      "learning_rate": 0.00013866265602680456,
+      "loss": 12.9054,
+      "step": 6732
+    },
+    {
+      "epoch": 1.1774863387978143,
+      "grad_norm": 0.9171066468287343,
+      "learning_rate": 0.00013864524268693596,
+      "loss": 12.9215,
+      "step": 6733
+    },
+    {
+      "epoch": 1.1776612021857924,
+      "grad_norm": 0.9862606707995889,
+      "learning_rate": 0.0001386278279694096,
+      "loss": 12.8052,
+      "step": 6734
+    },
+    {
+      "epoch": 1.1778360655737705,
+      "grad_norm": 0.9519073536634386,
+      "learning_rate": 0.00013861041187484632,
+      "loss": 12.785,
+      "step": 6735
+    },
+    {
+      "epoch": 1.1780109289617486,
+      "grad_norm": 0.9437244714258748,
+      "learning_rate": 0.00013859299440386697,
+      "loss": 12.7438,
+      "step": 6736
+    },
+    {
+      "epoch": 1.1781857923497268,
+      "grad_norm": 1.016057989034622,
+      "learning_rate": 0.00013857557555709249,
+      "loss": 12.4872,
+      "step": 6737
+    },
+    {
+      "epoch": 1.1783606557377049,
+      "grad_norm": 1.0527471036564366,
+      "learning_rate": 0.00013855815533514375,
+      "loss": 12.822,
+      "step": 6738
+    },
+    {
+      "epoch": 1.178535519125683,
+      "grad_norm": 0.9821804904662034,
+      "learning_rate": 0.00013854073373864188,
+      "loss": 12.8875,
+      "step": 6739
+    },
+    {
+      "epoch": 1.1787103825136611,
+      "grad_norm": 1.0973443465902053,
+      "learning_rate": 0.00013852331076820784,
+      "loss": 12.8511,
+      "step": 6740
+    },
+    {
+      "epoch": 1.1788852459016392,
+      "grad_norm": 1.0734976811707018,
+      "learning_rate": 0.00013850588642446283,
+      "loss": 12.7047,
+      "step": 6741
+    },
+    {
+      "epoch": 1.1790601092896176,
+      "grad_norm": 1.0272399716570506,
+      "learning_rate": 0.0001384884607080279,
+      "loss": 12.8807,
+      "step": 6742
+    },
+    {
+      "epoch": 1.1792349726775957,
+      "grad_norm": 1.082708736906711,
+      "learning_rate": 0.00013847103361952433,
+      "loss": 12.7524,
+      "step": 6743
+    },
+    {
+      "epoch": 1.1794098360655738,
+      "grad_norm": 1.046687069296545,
+      "learning_rate": 0.00013845360515957338,
+      "loss": 12.919,
+      "step": 6744
+    },
+    {
+      "epoch": 1.179584699453552,
+      "grad_norm": 0.8633419756686521,
+      "learning_rate": 0.00013843617532879632,
+      "loss": 12.9289,
+      "step": 6745
+    },
+    {
+      "epoch": 1.17975956284153,
+      "grad_norm": 1.0910720869555703,
+      "learning_rate": 0.00013841874412781456,
+      "loss": 13.0445,
+      "step": 6746
+    },
+    {
+      "epoch": 1.1799344262295082,
+      "grad_norm": 0.9244030167826086,
+      "learning_rate": 0.0001384013115572494,
+      "loss": 12.8772,
+      "step": 6747
+    },
+    {
+      "epoch": 1.1801092896174863,
+      "grad_norm": 1.0651433713913363,
+      "learning_rate": 0.00013838387761772237,
+      "loss": 12.8579,
+      "step": 6748
+    },
+    {
+      "epoch": 1.1802841530054644,
+      "grad_norm": 0.9330798287381464,
+      "learning_rate": 0.00013836644230985496,
+      "loss": 12.8237,
+      "step": 6749
+    },
+    {
+      "epoch": 1.1804590163934425,
+      "grad_norm": 0.9767665846112978,
+      "learning_rate": 0.00013834900563426867,
+      "loss": 12.9622,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1806338797814209,
+      "grad_norm": 0.9569926087535915,
+      "learning_rate": 0.00013833156759158517,
+      "loss": 12.9212,
+      "step": 6751
+    },
+    {
+      "epoch": 1.180808743169399,
+      "grad_norm": 0.8726408414893182,
+      "learning_rate": 0.00013831412818242605,
+      "loss": 12.4184,
+      "step": 6752
+    },
+    {
+      "epoch": 1.180983606557377,
+      "grad_norm": 0.945188628265453,
+      "learning_rate": 0.00013829668740741307,
+      "loss": 12.7783,
+      "step": 6753
+    },
+    {
+      "epoch": 1.1811584699453552,
+      "grad_norm": 1.0012809410627421,
+      "learning_rate": 0.0001382792452671679,
+      "loss": 12.9012,
+      "step": 6754
+    },
+    {
+      "epoch": 1.1813333333333333,
+      "grad_norm": 1.0915617710270242,
+      "learning_rate": 0.00013826180176231234,
+      "loss": 12.7865,
+      "step": 6755
+    },
+    {
+      "epoch": 1.1815081967213115,
+      "grad_norm": 1.0404045243439866,
+      "learning_rate": 0.0001382443568934683,
+      "loss": 12.8463,
+      "step": 6756
+    },
+    {
+      "epoch": 1.1816830601092896,
+      "grad_norm": 1.1017482105348722,
+      "learning_rate": 0.00013822691066125759,
+      "loss": 12.8764,
+      "step": 6757
+    },
+    {
+      "epoch": 1.1818579234972677,
+      "grad_norm": 1.0216197484453229,
+      "learning_rate": 0.00013820946306630219,
+      "loss": 12.8561,
+      "step": 6758
+    },
+    {
+      "epoch": 1.1820327868852458,
+      "grad_norm": 1.129214598553067,
+      "learning_rate": 0.00013819201410922405,
+      "loss": 12.7863,
+      "step": 6759
+    },
+    {
+      "epoch": 1.1822076502732242,
+      "grad_norm": 1.0543424165492266,
+      "learning_rate": 0.00013817456379064521,
+      "loss": 12.7096,
+      "step": 6760
+    },
+    {
+      "epoch": 1.182382513661202,
+      "grad_norm": 1.0806362031026913,
+      "learning_rate": 0.00013815711211118785,
+      "loss": 12.665,
+      "step": 6761
+    },
+    {
+      "epoch": 1.1825573770491804,
+      "grad_norm": 1.286310310928308,
+      "learning_rate": 0.00013813965907147395,
+      "loss": 12.5795,
+      "step": 6762
+    },
+    {
+      "epoch": 1.1827322404371585,
+      "grad_norm": 1.0983156638092233,
+      "learning_rate": 0.0001381222046721258,
+      "loss": 12.7564,
+      "step": 6763
+    },
+    {
+      "epoch": 1.1829071038251366,
+      "grad_norm": 1.5303414780902687,
+      "learning_rate": 0.00013810474891376562,
+      "loss": 12.7116,
+      "step": 6764
+    },
+    {
+      "epoch": 1.1830819672131148,
+      "grad_norm": 1.0758430709585043,
+      "learning_rate": 0.00013808729179701562,
+      "loss": 12.8303,
+      "step": 6765
+    },
+    {
+      "epoch": 1.1832568306010929,
+      "grad_norm": 1.4702570630621714,
+      "learning_rate": 0.0001380698333224982,
+      "loss": 12.6288,
+      "step": 6766
+    },
+    {
+      "epoch": 1.183431693989071,
+      "grad_norm": 0.9351167819841862,
+      "learning_rate": 0.0001380523734908357,
+      "loss": 12.6503,
+      "step": 6767
+    },
+    {
+      "epoch": 1.1836065573770491,
+      "grad_norm": 1.567908758053857,
+      "learning_rate": 0.0001380349123026505,
+      "loss": 12.9233,
+      "step": 6768
+    },
+    {
+      "epoch": 1.1837814207650272,
+      "grad_norm": 0.8964290540767871,
+      "learning_rate": 0.00013801744975856515,
+      "loss": 12.7479,
+      "step": 6769
+    },
+    {
+      "epoch": 1.1839562841530054,
+      "grad_norm": 1.3954820238320083,
+      "learning_rate": 0.0001379999858592021,
+      "loss": 12.6877,
+      "step": 6770
+    },
+    {
+      "epoch": 1.1841311475409837,
+      "grad_norm": 1.1134271882949345,
+      "learning_rate": 0.000137982520605184,
+      "loss": 12.8919,
+      "step": 6771
+    },
+    {
+      "epoch": 1.1843060109289618,
+      "grad_norm": 1.240653724351266,
+      "learning_rate": 0.0001379650539971334,
+      "loss": 12.7602,
+      "step": 6772
+    },
+    {
+      "epoch": 1.18448087431694,
+      "grad_norm": 1.1385062025908537,
+      "learning_rate": 0.00013794758603567298,
+      "loss": 12.7906,
+      "step": 6773
+    },
+    {
+      "epoch": 1.184655737704918,
+      "grad_norm": 0.9665807241928888,
+      "learning_rate": 0.0001379301167214255,
+      "loss": 12.6937,
+      "step": 6774
+    },
+    {
+      "epoch": 1.1848306010928962,
+      "grad_norm": 1.076409575851837,
+      "learning_rate": 0.0001379126460550136,
+      "loss": 12.681,
+      "step": 6775
+    },
+    {
+      "epoch": 1.1850054644808743,
+      "grad_norm": 0.9308832933171476,
+      "learning_rate": 0.0001378951740370602,
+      "loss": 12.7359,
+      "step": 6776
+    },
+    {
+      "epoch": 1.1851803278688524,
+      "grad_norm": 1.1737691010368638,
+      "learning_rate": 0.00013787770066818814,
+      "loss": 13.0158,
+      "step": 6777
+    },
+    {
+      "epoch": 1.1853551912568305,
+      "grad_norm": 1.0139051997811819,
+      "learning_rate": 0.00013786022594902025,
+      "loss": 12.8348,
+      "step": 6778
+    },
+    {
+      "epoch": 1.1855300546448087,
+      "grad_norm": 1.314237461325346,
+      "learning_rate": 0.00013784274988017957,
+      "loss": 12.5726,
+      "step": 6779
+    },
+    {
+      "epoch": 1.185704918032787,
+      "grad_norm": 1.0503950609347963,
+      "learning_rate": 0.0001378252724622891,
+      "loss": 12.7913,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1858797814207651,
+      "grad_norm": 1.061671099566588,
+      "learning_rate": 0.0001378077936959718,
+      "loss": 12.5342,
+      "step": 6781
+    },
+    {
+      "epoch": 1.1860546448087432,
+      "grad_norm": 1.077709974589101,
+      "learning_rate": 0.00013779031358185088,
+      "loss": 12.5322,
+      "step": 6782
+    },
+    {
+      "epoch": 1.1862295081967213,
+      "grad_norm": 1.0171743391992545,
+      "learning_rate": 0.00013777283212054942,
+      "loss": 12.7918,
+      "step": 6783
+    },
+    {
+      "epoch": 1.1864043715846995,
+      "grad_norm": 0.9779562804555918,
+      "learning_rate": 0.00013775534931269063,
+      "loss": 12.6847,
+      "step": 6784
+    },
+    {
+      "epoch": 1.1865792349726776,
+      "grad_norm": 1.0383331248105572,
+      "learning_rate": 0.00013773786515889773,
+      "loss": 12.8552,
+      "step": 6785
+    },
+    {
+      "epoch": 1.1867540983606557,
+      "grad_norm": 0.8744972951848378,
+      "learning_rate": 0.00013772037965979408,
+      "loss": 12.9112,
+      "step": 6786
+    },
+    {
+      "epoch": 1.1869289617486338,
+      "grad_norm": 1.0450299579779962,
+      "learning_rate": 0.0001377028928160029,
+      "loss": 12.8467,
+      "step": 6787
+    },
+    {
+      "epoch": 1.187103825136612,
+      "grad_norm": 0.9228622836563185,
+      "learning_rate": 0.00013768540462814769,
+      "loss": 12.768,
+      "step": 6788
+    },
+    {
+      "epoch": 1.1872786885245903,
+      "grad_norm": 0.9351555218619839,
+      "learning_rate": 0.00013766791509685182,
+      "loss": 12.8779,
+      "step": 6789
+    },
+    {
+      "epoch": 1.1874535519125682,
+      "grad_norm": 0.9535213690194717,
+      "learning_rate": 0.00013765042422273876,
+      "loss": 12.8775,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1876284153005465,
+      "grad_norm": 0.9260027488615343,
+      "learning_rate": 0.00013763293200643213,
+      "loss": 12.6567,
+      "step": 6791
+    },
+    {
+      "epoch": 1.1878032786885246,
+      "grad_norm": 0.9730671201944312,
+      "learning_rate": 0.00013761543844855538,
+      "loss": 12.8625,
+      "step": 6792
+    },
+    {
+      "epoch": 1.1879781420765028,
+      "grad_norm": 0.888736314675643,
+      "learning_rate": 0.00013759794354973223,
+      "loss": 12.805,
+      "step": 6793
+    },
+    {
+      "epoch": 1.1881530054644809,
+      "grad_norm": 0.8878711627478881,
+      "learning_rate": 0.00013758044731058632,
+      "loss": 12.3256,
+      "step": 6794
+    },
+    {
+      "epoch": 1.188327868852459,
+      "grad_norm": 0.9644432563311535,
+      "learning_rate": 0.00013756294973174137,
+      "loss": 12.9345,
+      "step": 6795
+    },
+    {
+      "epoch": 1.1885027322404371,
+      "grad_norm": 0.8914119452444954,
+      "learning_rate": 0.00013754545081382113,
+      "loss": 12.3931,
+      "step": 6796
+    },
+    {
+      "epoch": 1.1886775956284152,
+      "grad_norm": 0.9589632738815765,
+      "learning_rate": 0.00013752795055744946,
+      "loss": 12.8326,
+      "step": 6797
+    },
+    {
+      "epoch": 1.1888524590163934,
+      "grad_norm": 0.9216109510743586,
+      "learning_rate": 0.00013751044896325018,
+      "loss": 12.9002,
+      "step": 6798
+    },
+    {
+      "epoch": 1.1890273224043715,
+      "grad_norm": 0.8575973176100514,
+      "learning_rate": 0.00013749294603184722,
+      "loss": 12.981,
+      "step": 6799
+    },
+    {
+      "epoch": 1.1892021857923498,
+      "grad_norm": 0.9397830916801594,
+      "learning_rate": 0.00013747544176386458,
+      "loss": 12.6837,
+      "step": 6800
+    },
+    {
+      "epoch": 1.189377049180328,
+      "grad_norm": 0.9257670896085133,
+      "learning_rate": 0.0001374579361599262,
+      "loss": 12.6759,
+      "step": 6801
+    },
+    {
+      "epoch": 1.189551912568306,
+      "grad_norm": 0.9243689579627582,
+      "learning_rate": 0.00013744042922065615,
+      "loss": 12.9026,
+      "step": 6802
+    },
+    {
+      "epoch": 1.1897267759562842,
+      "grad_norm": 0.9754911433042539,
+      "learning_rate": 0.00013742292094667854,
+      "loss": 12.7482,
+      "step": 6803
+    },
+    {
+      "epoch": 1.1899016393442623,
+      "grad_norm": 0.9781951444128781,
+      "learning_rate": 0.00013740541133861755,
+      "loss": 12.9754,
+      "step": 6804
+    },
+    {
+      "epoch": 1.1900765027322404,
+      "grad_norm": 0.9257969857091367,
+      "learning_rate": 0.0001373879003970973,
+      "loss": 12.4995,
+      "step": 6805
+    },
+    {
+      "epoch": 1.1902513661202185,
+      "grad_norm": 0.9555391478198427,
+      "learning_rate": 0.00013737038812274214,
+      "loss": 12.7852,
+      "step": 6806
+    },
+    {
+      "epoch": 1.1904262295081967,
+      "grad_norm": 0.9382708006486778,
+      "learning_rate": 0.00013735287451617628,
+      "loss": 12.8353,
+      "step": 6807
+    },
+    {
+      "epoch": 1.1906010928961748,
+      "grad_norm": 0.9661477726462054,
+      "learning_rate": 0.0001373353595780241,
+      "loss": 12.8105,
+      "step": 6808
+    },
+    {
+      "epoch": 1.1907759562841531,
+      "grad_norm": 0.9770588428428507,
+      "learning_rate": 0.00013731784330890995,
+      "loss": 12.9647,
+      "step": 6809
+    },
+    {
+      "epoch": 1.1909508196721312,
+      "grad_norm": 0.8994532460214584,
+      "learning_rate": 0.00013730032570945832,
+      "loss": 12.6564,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1911256830601094,
+      "grad_norm": 0.9562859903931189,
+      "learning_rate": 0.00013728280678029363,
+      "loss": 12.8376,
+      "step": 6811
+    },
+    {
+      "epoch": 1.1913005464480875,
+      "grad_norm": 0.8931773293356661,
+      "learning_rate": 0.00013726528652204044,
+      "loss": 12.6558,
+      "step": 6812
+    },
+    {
+      "epoch": 1.1914754098360656,
+      "grad_norm": 0.8030063832156455,
+      "learning_rate": 0.00013724776493532334,
+      "loss": 12.9745,
+      "step": 6813
+    },
+    {
+      "epoch": 1.1916502732240437,
+      "grad_norm": 0.9687614144518287,
+      "learning_rate": 0.0001372302420207669,
+      "loss": 12.7316,
+      "step": 6814
+    },
+    {
+      "epoch": 1.1918251366120218,
+      "grad_norm": 0.9598209613811769,
+      "learning_rate": 0.00013721271777899588,
+      "loss": 12.4413,
+      "step": 6815
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 1.0592243594222832,
+      "learning_rate": 0.00013719519221063492,
+      "loss": 12.607,
+      "step": 6816
+    },
+    {
+      "epoch": 1.192174863387978,
+      "grad_norm": 0.9304252120434033,
+      "learning_rate": 0.00013717766531630882,
+      "loss": 12.9474,
+      "step": 6817
+    },
+    {
+      "epoch": 1.1923497267759564,
+      "grad_norm": 1.0277461026199857,
+      "learning_rate": 0.0001371601370966424,
+      "loss": 12.9168,
+      "step": 6818
+    },
+    {
+      "epoch": 1.1925245901639343,
+      "grad_norm": 1.04245754902813,
+      "learning_rate": 0.00013714260755226054,
+      "loss": 12.917,
+      "step": 6819
+    },
+    {
+      "epoch": 1.1926994535519126,
+      "grad_norm": 0.932050597588475,
+      "learning_rate": 0.00013712507668378808,
+      "loss": 13.0481,
+      "step": 6820
+    },
+    {
+      "epoch": 1.1928743169398908,
+      "grad_norm": 0.8927467758389661,
+      "learning_rate": 0.00013710754449185002,
+      "loss": 12.5411,
+      "step": 6821
+    },
+    {
+      "epoch": 1.1930491803278689,
+      "grad_norm": 0.9579781134820726,
+      "learning_rate": 0.00013709001097707136,
+      "loss": 12.7251,
+      "step": 6822
+    },
+    {
+      "epoch": 1.193224043715847,
+      "grad_norm": 0.9377068285835998,
+      "learning_rate": 0.0001370724761400771,
+      "loss": 12.967,
+      "step": 6823
+    },
+    {
+      "epoch": 1.1933989071038251,
+      "grad_norm": 0.9912742715192026,
+      "learning_rate": 0.00013705493998149242,
+      "loss": 12.7652,
+      "step": 6824
+    },
+    {
+      "epoch": 1.1935737704918032,
+      "grad_norm": 0.910439183927035,
+      "learning_rate": 0.0001370374025019424,
+      "loss": 12.8065,
+      "step": 6825
+    },
+    {
+      "epoch": 1.1937486338797814,
+      "grad_norm": 0.8730973965213532,
+      "learning_rate": 0.00013701986370205229,
+      "loss": 12.9624,
+      "step": 6826
+    },
+    {
+      "epoch": 1.1939234972677595,
+      "grad_norm": 0.9096789884762311,
+      "learning_rate": 0.00013700232358244727,
+      "loss": 12.7938,
+      "step": 6827
+    },
+    {
+      "epoch": 1.1940983606557376,
+      "grad_norm": 0.9329116158168036,
+      "learning_rate": 0.00013698478214375267,
+      "loss": 13.0141,
+      "step": 6828
+    },
+    {
+      "epoch": 1.194273224043716,
+      "grad_norm": 0.9749807799251876,
+      "learning_rate": 0.00013696723938659376,
+      "loss": 12.5979,
+      "step": 6829
+    },
+    {
+      "epoch": 1.194448087431694,
+      "grad_norm": 0.8987803637357475,
+      "learning_rate": 0.00013694969531159598,
+      "loss": 12.8982,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1946229508196722,
+      "grad_norm": 0.9773821252220684,
+      "learning_rate": 0.00013693214991938475,
+      "loss": 12.7489,
+      "step": 6831
+    },
+    {
+      "epoch": 1.1947978142076503,
+      "grad_norm": 1.0047394186351617,
+      "learning_rate": 0.00013691460321058548,
+      "loss": 12.9222,
+      "step": 6832
+    },
+    {
+      "epoch": 1.1949726775956284,
+      "grad_norm": 1.0033238271580145,
+      "learning_rate": 0.00013689705518582377,
+      "loss": 13.0009,
+      "step": 6833
+    },
+    {
+      "epoch": 1.1951475409836065,
+      "grad_norm": 1.0208680441655502,
+      "learning_rate": 0.0001368795058457251,
+      "loss": 12.8975,
+      "step": 6834
+    },
+    {
+      "epoch": 1.1953224043715847,
+      "grad_norm": 1.0233044923891845,
+      "learning_rate": 0.0001368619551909152,
+      "loss": 12.5044,
+      "step": 6835
+    },
+    {
+      "epoch": 1.1954972677595628,
+      "grad_norm": 0.9250460357362105,
+      "learning_rate": 0.00013684440322201966,
+      "loss": 12.9995,
+      "step": 6836
+    },
+    {
+      "epoch": 1.195672131147541,
+      "grad_norm": 1.2658413050021535,
+      "learning_rate": 0.0001368268499396642,
+      "loss": 12.8588,
+      "step": 6837
+    },
+    {
+      "epoch": 1.1958469945355192,
+      "grad_norm": 0.9164034757849367,
+      "learning_rate": 0.00013680929534447454,
+      "loss": 12.8288,
+      "step": 6838
+    },
+    {
+      "epoch": 1.1960218579234974,
+      "grad_norm": 1.1282892754198088,
+      "learning_rate": 0.0001367917394370765,
+      "loss": 12.9536,
+      "step": 6839
+    },
+    {
+      "epoch": 1.1961967213114755,
+      "grad_norm": 1.0284528422697004,
+      "learning_rate": 0.00013677418221809596,
+      "loss": 12.8853,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1963715846994536,
+      "grad_norm": 1.0840351225582636,
+      "learning_rate": 0.00013675662368815877,
+      "loss": 12.7582,
+      "step": 6841
+    },
+    {
+      "epoch": 1.1965464480874317,
+      "grad_norm": 0.9664749153759619,
+      "learning_rate": 0.0001367390638478909,
+      "loss": 12.74,
+      "step": 6842
+    },
+    {
+      "epoch": 1.1967213114754098,
+      "grad_norm": 1.1999597451184205,
+      "learning_rate": 0.00013672150269791833,
+      "loss": 12.8992,
+      "step": 6843
+    },
+    {
+      "epoch": 1.196896174863388,
+      "grad_norm": 0.9533990412971114,
+      "learning_rate": 0.0001367039402388671,
+      "loss": 12.9318,
+      "step": 6844
+    },
+    {
+      "epoch": 1.197071038251366,
+      "grad_norm": 1.0934684012798246,
+      "learning_rate": 0.00013668637647136326,
+      "loss": 12.5986,
+      "step": 6845
+    },
+    {
+      "epoch": 1.1972459016393442,
+      "grad_norm": 1.061294103641881,
+      "learning_rate": 0.000136668811396033,
+      "loss": 12.8586,
+      "step": 6846
+    },
+    {
+      "epoch": 1.1974207650273223,
+      "grad_norm": 1.097500174313484,
+      "learning_rate": 0.0001366512450135024,
+      "loss": 12.681,
+      "step": 6847
+    },
+    {
+      "epoch": 1.1975956284153004,
+      "grad_norm": 1.0585983426447796,
+      "learning_rate": 0.0001366336773243978,
+      "loss": 12.9003,
+      "step": 6848
+    },
+    {
+      "epoch": 1.1977704918032788,
+      "grad_norm": 0.9320502911262779,
+      "learning_rate": 0.00013661610832934538,
+      "loss": 12.8294,
+      "step": 6849
+    },
+    {
+      "epoch": 1.1979453551912569,
+      "grad_norm": 0.9887633976436222,
+      "learning_rate": 0.00013659853802897145,
+      "loss": 12.7922,
+      "step": 6850
+    },
+    {
+      "epoch": 1.198120218579235,
+      "grad_norm": 0.9278376167086726,
+      "learning_rate": 0.00013658096642390244,
+      "loss": 12.5991,
+      "step": 6851
+    },
+    {
+      "epoch": 1.1982950819672131,
+      "grad_norm": 0.9563513624006176,
+      "learning_rate": 0.0001365633935147647,
+      "loss": 12.4571,
+      "step": 6852
+    },
+    {
+      "epoch": 1.1984699453551912,
+      "grad_norm": 0.9390990088887622,
+      "learning_rate": 0.00013654581930218473,
+      "loss": 12.2828,
+      "step": 6853
+    },
+    {
+      "epoch": 1.1986448087431694,
+      "grad_norm": 1.047895993172075,
+      "learning_rate": 0.00013652824378678898,
+      "loss": 12.8489,
+      "step": 6854
+    },
+    {
+      "epoch": 1.1988196721311475,
+      "grad_norm": 1.052869941330101,
+      "learning_rate": 0.000136510666969204,
+      "loss": 12.9242,
+      "step": 6855
+    },
+    {
+      "epoch": 1.1989945355191256,
+      "grad_norm": 1.343071135513479,
+      "learning_rate": 0.00013649308885005642,
+      "loss": 12.8648,
+      "step": 6856
+    },
+    {
+      "epoch": 1.1991693989071037,
+      "grad_norm": 1.0730101248949955,
+      "learning_rate": 0.00013647550942997287,
+      "loss": 12.9266,
+      "step": 6857
+    },
+    {
+      "epoch": 1.199344262295082,
+      "grad_norm": 1.0223521634491002,
+      "learning_rate": 0.00013645792870957998,
+      "loss": 12.819,
+      "step": 6858
+    },
+    {
+      "epoch": 1.1995191256830602,
+      "grad_norm": 1.0320297680001016,
+      "learning_rate": 0.00013644034668950457,
+      "loss": 13.0833,
+      "step": 6859
+    },
+    {
+      "epoch": 1.1996939890710383,
+      "grad_norm": 1.0708626373684658,
+      "learning_rate": 0.00013642276337037337,
+      "loss": 12.7023,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1998688524590164,
+      "grad_norm": 0.9851386675303585,
+      "learning_rate": 0.0001364051787528132,
+      "loss": 12.7762,
+      "step": 6861
+    },
+    {
+      "epoch": 1.2000437158469945,
+      "grad_norm": 1.1635957254232023,
+      "learning_rate": 0.00013638759283745094,
+      "loss": 12.7062,
+      "step": 6862
+    },
+    {
+      "epoch": 1.2002185792349727,
+      "grad_norm": 0.8814403824896168,
+      "learning_rate": 0.00013637000562491354,
+      "loss": 12.8909,
+      "step": 6863
+    },
+    {
+      "epoch": 1.2003934426229508,
+      "grad_norm": 1.1992615041462882,
+      "learning_rate": 0.0001363524171158279,
+      "loss": 12.6326,
+      "step": 6864
+    },
+    {
+      "epoch": 1.200568306010929,
+      "grad_norm": 0.9727849233339904,
+      "learning_rate": 0.00013633482731082109,
+      "loss": 12.6501,
+      "step": 6865
+    },
+    {
+      "epoch": 1.200743169398907,
+      "grad_norm": 1.1800777366986699,
+      "learning_rate": 0.00013631723621052014,
+      "loss": 12.7602,
+      "step": 6866
+    },
+    {
+      "epoch": 1.2009180327868854,
+      "grad_norm": 0.8798093215427212,
+      "learning_rate": 0.0001362996438155521,
+      "loss": 12.814,
+      "step": 6867
+    },
+    {
+      "epoch": 1.2010928961748635,
+      "grad_norm": 1.2641840506716469,
+      "learning_rate": 0.0001362820501265442,
+      "loss": 12.8513,
+      "step": 6868
+    },
+    {
+      "epoch": 1.2012677595628416,
+      "grad_norm": 0.909895705144345,
+      "learning_rate": 0.00013626445514412362,
+      "loss": 12.6738,
+      "step": 6869
+    },
+    {
+      "epoch": 1.2014426229508197,
+      "grad_norm": 1.2945690786234478,
+      "learning_rate": 0.0001362468588689176,
+      "loss": 12.9061,
+      "step": 6870
+    },
+    {
+      "epoch": 1.2016174863387978,
+      "grad_norm": 0.9917603857012391,
+      "learning_rate": 0.00013622926130155337,
+      "loss": 12.8995,
+      "step": 6871
+    },
+    {
+      "epoch": 1.201792349726776,
+      "grad_norm": 1.3126230381858268,
+      "learning_rate": 0.00013621166244265833,
+      "loss": 12.8241,
+      "step": 6872
+    },
+    {
+      "epoch": 1.201967213114754,
+      "grad_norm": 0.9124629837529699,
+      "learning_rate": 0.0001361940622928598,
+      "loss": 12.6508,
+      "step": 6873
+    },
+    {
+      "epoch": 1.2021420765027322,
+      "grad_norm": 1.3411430064512695,
+      "learning_rate": 0.00013617646085278525,
+      "loss": 12.7668,
+      "step": 6874
+    },
+    {
+      "epoch": 1.2023169398907103,
+      "grad_norm": 0.9784211072050053,
+      "learning_rate": 0.00013615885812306218,
+      "loss": 12.7303,
+      "step": 6875
+    },
+    {
+      "epoch": 1.2024918032786884,
+      "grad_norm": 1.1146500807654314,
+      "learning_rate": 0.00013614125410431801,
+      "loss": 12.7386,
+      "step": 6876
+    },
+    {
+      "epoch": 1.2026666666666666,
+      "grad_norm": 1.0889205580911105,
+      "learning_rate": 0.00013612364879718038,
+      "loss": 12.5834,
+      "step": 6877
+    },
+    {
+      "epoch": 1.202841530054645,
+      "grad_norm": 0.9548211955473582,
+      "learning_rate": 0.00013610604220227687,
+      "loss": 12.6014,
+      "step": 6878
+    },
+    {
+      "epoch": 1.203016393442623,
+      "grad_norm": 1.0946422704943066,
+      "learning_rate": 0.00013608843432023516,
+      "loss": 12.8477,
+      "step": 6879
+    },
+    {
+      "epoch": 1.2031912568306011,
+      "grad_norm": 0.8843845845409238,
+      "learning_rate": 0.0001360708251516829,
+      "loss": 12.6582,
+      "step": 6880
+    },
+    {
+      "epoch": 1.2033661202185792,
+      "grad_norm": 1.0692902022346427,
+      "learning_rate": 0.0001360532146972479,
+      "loss": 12.8776,
+      "step": 6881
+    },
+    {
+      "epoch": 1.2035409836065574,
+      "grad_norm": 0.9595173059197544,
+      "learning_rate": 0.00013603560295755791,
+      "loss": 12.7064,
+      "step": 6882
+    },
+    {
+      "epoch": 1.2037158469945355,
+      "grad_norm": 1.0056788469878555,
+      "learning_rate": 0.00013601798993324076,
+      "loss": 12.5573,
+      "step": 6883
+    },
+    {
+      "epoch": 1.2038907103825136,
+      "grad_norm": 0.8766997608061297,
+      "learning_rate": 0.0001360003756249244,
+      "loss": 12.6751,
+      "step": 6884
+    },
+    {
+      "epoch": 1.2040655737704917,
+      "grad_norm": 1.0311999476623805,
+      "learning_rate": 0.00013598276003323666,
+      "loss": 12.9266,
+      "step": 6885
+    },
+    {
+      "epoch": 1.2042404371584698,
+      "grad_norm": 0.9746158221464148,
+      "learning_rate": 0.0001359651431588056,
+      "loss": 12.9394,
+      "step": 6886
+    },
+    {
+      "epoch": 1.2044153005464482,
+      "grad_norm": 1.07903119418466,
+      "learning_rate": 0.00013594752500225915,
+      "loss": 12.9183,
+      "step": 6887
+    },
+    {
+      "epoch": 1.2045901639344263,
+      "grad_norm": 1.0103584142191124,
+      "learning_rate": 0.0001359299055642255,
+      "loss": 12.6891,
+      "step": 6888
+    },
+    {
+      "epoch": 1.2047650273224044,
+      "grad_norm": 0.8683827149282526,
+      "learning_rate": 0.00013591228484533265,
+      "loss": 12.6809,
+      "step": 6889
+    },
+    {
+      "epoch": 1.2049398907103825,
+      "grad_norm": 1.2149155467567763,
+      "learning_rate": 0.0001358946628462088,
+      "loss": 12.6443,
+      "step": 6890
+    },
+    {
+      "epoch": 1.2051147540983607,
+      "grad_norm": 1.065459817101849,
+      "learning_rate": 0.0001358770395674822,
+      "loss": 12.9262,
+      "step": 6891
+    },
+    {
+      "epoch": 1.2052896174863388,
+      "grad_norm": 0.9798444862448251,
+      "learning_rate": 0.00013585941500978105,
+      "loss": 12.7185,
+      "step": 6892
+    },
+    {
+      "epoch": 1.205464480874317,
+      "grad_norm": 0.8695292714681996,
+      "learning_rate": 0.00013584178917373365,
+      "loss": 12.6597,
+      "step": 6893
+    },
+    {
+      "epoch": 1.205639344262295,
+      "grad_norm": 0.9920760922960684,
+      "learning_rate": 0.00013582416205996833,
+      "loss": 12.7818,
+      "step": 6894
+    },
+    {
+      "epoch": 1.2058142076502731,
+      "grad_norm": 1.0680480076540708,
+      "learning_rate": 0.0001358065336691135,
+      "loss": 12.7698,
+      "step": 6895
+    },
+    {
+      "epoch": 1.2059890710382515,
+      "grad_norm": 0.9148525417818952,
+      "learning_rate": 0.00013578890400179754,
+      "loss": 12.8846,
+      "step": 6896
+    },
+    {
+      "epoch": 1.2061639344262296,
+      "grad_norm": 1.1529885544129168,
+      "learning_rate": 0.00013577127305864901,
+      "loss": 12.3446,
+      "step": 6897
+    },
+    {
+      "epoch": 1.2063387978142077,
+      "grad_norm": 0.9516252712835119,
+      "learning_rate": 0.0001357536408402964,
+      "loss": 12.6513,
+      "step": 6898
+    },
+    {
+      "epoch": 1.2065136612021858,
+      "grad_norm": 1.01812815238512,
+      "learning_rate": 0.00013573600734736823,
+      "loss": 12.7916,
+      "step": 6899
+    },
+    {
+      "epoch": 1.206688524590164,
+      "grad_norm": 0.9968263267572524,
+      "learning_rate": 0.00013571837258049318,
+      "loss": 12.9425,
+      "step": 6900
+    },
+    {
+      "epoch": 1.206863387978142,
+      "grad_norm": 0.9724281260372909,
+      "learning_rate": 0.00013570073654029988,
+      "loss": 12.8154,
+      "step": 6901
+    },
+    {
+      "epoch": 1.2070382513661202,
+      "grad_norm": 0.8849932389723398,
+      "learning_rate": 0.00013568309922741702,
+      "loss": 12.7089,
+      "step": 6902
+    },
+    {
+      "epoch": 1.2072131147540983,
+      "grad_norm": 0.8787879793344909,
+      "learning_rate": 0.00013566546064247334,
+      "loss": 12.5033,
+      "step": 6903
+    },
+    {
+      "epoch": 1.2073879781420764,
+      "grad_norm": 0.9309790975872455,
+      "learning_rate": 0.0001356478207860977,
+      "loss": 12.6263,
+      "step": 6904
+    },
+    {
+      "epoch": 1.2075628415300546,
+      "grad_norm": 0.9310518624190444,
+      "learning_rate": 0.00013563017965891887,
+      "loss": 12.8408,
+      "step": 6905
+    },
+    {
+      "epoch": 1.2077377049180327,
+      "grad_norm": 0.9353338439776411,
+      "learning_rate": 0.0001356125372615658,
+      "loss": 12.8715,
+      "step": 6906
+    },
+    {
+      "epoch": 1.207912568306011,
+      "grad_norm": 0.8197675505156127,
+      "learning_rate": 0.00013559489359466734,
+      "loss": 12.5463,
+      "step": 6907
+    },
+    {
+      "epoch": 1.2080874316939891,
+      "grad_norm": 0.8836971886610967,
+      "learning_rate": 0.00013557724865885255,
+      "loss": 12.7456,
+      "step": 6908
+    },
+    {
+      "epoch": 1.2082622950819673,
+      "grad_norm": 1.0711555792860596,
+      "learning_rate": 0.0001355596024547504,
+      "loss": 12.5922,
+      "step": 6909
+    },
+    {
+      "epoch": 1.2084371584699454,
+      "grad_norm": 0.9006055943991134,
+      "learning_rate": 0.00013554195498298997,
+      "loss": 12.6274,
+      "step": 6910
+    },
+    {
+      "epoch": 1.2086120218579235,
+      "grad_norm": 1.058273860037635,
+      "learning_rate": 0.00013552430624420037,
+      "loss": 12.6559,
+      "step": 6911
+    },
+    {
+      "epoch": 1.2087868852459016,
+      "grad_norm": 0.8594211043172542,
+      "learning_rate": 0.00013550665623901075,
+      "loss": 12.4421,
+      "step": 6912
+    },
+    {
+      "epoch": 1.2089617486338797,
+      "grad_norm": 0.927387273677329,
+      "learning_rate": 0.00013548900496805033,
+      "loss": 12.9779,
+      "step": 6913
+    },
+    {
+      "epoch": 1.2091366120218578,
+      "grad_norm": 0.9708645082406709,
+      "learning_rate": 0.00013547135243194833,
+      "loss": 12.467,
+      "step": 6914
+    },
+    {
+      "epoch": 1.209311475409836,
+      "grad_norm": 0.8856762292838793,
+      "learning_rate": 0.00013545369863133408,
+      "loss": 12.6366,
+      "step": 6915
+    },
+    {
+      "epoch": 1.2094863387978143,
+      "grad_norm": 0.8382941848330923,
+      "learning_rate": 0.0001354360435668369,
+      "loss": 12.5084,
+      "step": 6916
+    },
+    {
+      "epoch": 1.2096612021857924,
+      "grad_norm": 0.8899380607892434,
+      "learning_rate": 0.00013541838723908614,
+      "loss": 12.9824,
+      "step": 6917
+    },
+    {
+      "epoch": 1.2098360655737705,
+      "grad_norm": 0.9574867046885243,
+      "learning_rate": 0.00013540072964871128,
+      "loss": 12.9922,
+      "step": 6918
+    },
+    {
+      "epoch": 1.2100109289617487,
+      "grad_norm": 0.8347338616329886,
+      "learning_rate": 0.00013538307079634179,
+      "loss": 12.5822,
+      "step": 6919
+    },
+    {
+      "epoch": 1.2101857923497268,
+      "grad_norm": 1.060900418508991,
+      "learning_rate": 0.00013536541068260712,
+      "loss": 12.8105,
+      "step": 6920
+    },
+    {
+      "epoch": 1.210360655737705,
+      "grad_norm": 0.8418188393178861,
+      "learning_rate": 0.0001353477493081369,
+      "loss": 12.6111,
+      "step": 6921
+    },
+    {
+      "epoch": 1.210535519125683,
+      "grad_norm": 0.8200700183295137,
+      "learning_rate": 0.00013533008667356074,
+      "loss": 12.8045,
+      "step": 6922
+    },
+    {
+      "epoch": 1.2107103825136611,
+      "grad_norm": 0.9366053893658081,
+      "learning_rate": 0.00013531242277950822,
+      "loss": 12.8414,
+      "step": 6923
+    },
+    {
+      "epoch": 1.2108852459016393,
+      "grad_norm": 1.0677245912152866,
+      "learning_rate": 0.00013529475762660912,
+      "loss": 12.7364,
+      "step": 6924
+    },
+    {
+      "epoch": 1.2110601092896176,
+      "grad_norm": 0.8365082021134274,
+      "learning_rate": 0.00013527709121549314,
+      "loss": 12.7511,
+      "step": 6925
+    },
+    {
+      "epoch": 1.2112349726775957,
+      "grad_norm": 0.9657790921942198,
+      "learning_rate": 0.00013525942354679008,
+      "loss": 12.5682,
+      "step": 6926
+    },
+    {
+      "epoch": 1.2114098360655738,
+      "grad_norm": 0.96195607051413,
+      "learning_rate": 0.00013524175462112977,
+      "loss": 12.7594,
+      "step": 6927
+    },
+    {
+      "epoch": 1.211584699453552,
+      "grad_norm": 0.9347253392269284,
+      "learning_rate": 0.0001352240844391421,
+      "loss": 12.8036,
+      "step": 6928
+    },
+    {
+      "epoch": 1.21175956284153,
+      "grad_norm": 0.8558379425222449,
+      "learning_rate": 0.00013520641300145692,
+      "loss": 12.8164,
+      "step": 6929
+    },
+    {
+      "epoch": 1.2119344262295082,
+      "grad_norm": 0.9980084547518578,
+      "learning_rate": 0.0001351887403087043,
+      "loss": 12.7683,
+      "step": 6930
+    },
+    {
+      "epoch": 1.2121092896174863,
+      "grad_norm": 0.8735023501742085,
+      "learning_rate": 0.00013517106636151417,
+      "loss": 12.8649,
+      "step": 6931
+    },
+    {
+      "epoch": 1.2122841530054644,
+      "grad_norm": 1.047045870210984,
+      "learning_rate": 0.00013515339116051663,
+      "loss": 12.8012,
+      "step": 6932
+    },
+    {
+      "epoch": 1.2124590163934426,
+      "grad_norm": 0.9343372488362751,
+      "learning_rate": 0.00013513571470634176,
+      "loss": 12.8345,
+      "step": 6933
+    },
+    {
+      "epoch": 1.2126338797814207,
+      "grad_norm": 1.142362385846318,
+      "learning_rate": 0.00013511803699961975,
+      "loss": 12.5958,
+      "step": 6934
+    },
+    {
+      "epoch": 1.2128087431693988,
+      "grad_norm": 1.0626675400872199,
+      "learning_rate": 0.00013510035804098068,
+      "loss": 12.7639,
+      "step": 6935
+    },
+    {
+      "epoch": 1.2129836065573771,
+      "grad_norm": 1.2277746566298486,
+      "learning_rate": 0.00013508267783105492,
+      "loss": 12.7439,
+      "step": 6936
+    },
+    {
+      "epoch": 1.2131584699453553,
+      "grad_norm": 1.0301743094728437,
+      "learning_rate": 0.00013506499637047268,
+      "loss": 12.9124,
+      "step": 6937
+    },
+    {
+      "epoch": 1.2133333333333334,
+      "grad_norm": 0.9670407619739956,
+      "learning_rate": 0.00013504731365986426,
+      "loss": 12.7737,
+      "step": 6938
+    },
+    {
+      "epoch": 1.2135081967213115,
+      "grad_norm": 1.1298288885399081,
+      "learning_rate": 0.00013502962969986007,
+      "loss": 13.0591,
+      "step": 6939
+    },
+    {
+      "epoch": 1.2136830601092896,
+      "grad_norm": 0.8328505493435971,
+      "learning_rate": 0.0001350119444910905,
+      "loss": 12.6208,
+      "step": 6940
+    },
+    {
+      "epoch": 1.2138579234972677,
+      "grad_norm": 0.9817532856085991,
+      "learning_rate": 0.00013499425803418602,
+      "loss": 12.8543,
+      "step": 6941
+    },
+    {
+      "epoch": 1.2140327868852459,
+      "grad_norm": 0.8665708750291875,
+      "learning_rate": 0.00013497657032977714,
+      "loss": 12.497,
+      "step": 6942
+    },
+    {
+      "epoch": 1.214207650273224,
+      "grad_norm": 1.0596155278596882,
+      "learning_rate": 0.00013495888137849438,
+      "loss": 12.6374,
+      "step": 6943
+    },
+    {
+      "epoch": 1.214382513661202,
+      "grad_norm": 0.8474082445775147,
+      "learning_rate": 0.00013494119118096838,
+      "loss": 12.9511,
+      "step": 6944
+    },
+    {
+      "epoch": 1.2145573770491804,
+      "grad_norm": 0.9643720738516623,
+      "learning_rate": 0.00013492349973782969,
+      "loss": 12.8284,
+      "step": 6945
+    },
+    {
+      "epoch": 1.2147322404371586,
+      "grad_norm": 0.9220324880194883,
+      "learning_rate": 0.00013490580704970904,
+      "loss": 12.91,
+      "step": 6946
+    },
+    {
+      "epoch": 1.2149071038251367,
+      "grad_norm": 0.942586967022598,
+      "learning_rate": 0.0001348881131172372,
+      "loss": 12.7202,
+      "step": 6947
+    },
+    {
+      "epoch": 1.2150819672131148,
+      "grad_norm": 0.9801884630498918,
+      "learning_rate": 0.00013487041794104485,
+      "loss": 12.5907,
+      "step": 6948
+    },
+    {
+      "epoch": 1.215256830601093,
+      "grad_norm": 0.9444007782911374,
+      "learning_rate": 0.00013485272152176286,
+      "loss": 13.0393,
+      "step": 6949
+    },
+    {
+      "epoch": 1.215431693989071,
+      "grad_norm": 0.9299051533162268,
+      "learning_rate": 0.00013483502386002204,
+      "loss": 12.9128,
+      "step": 6950
+    },
+    {
+      "epoch": 1.2156065573770491,
+      "grad_norm": 0.9401011666134214,
+      "learning_rate": 0.00013481732495645334,
+      "loss": 12.7539,
+      "step": 6951
+    },
+    {
+      "epoch": 1.2157814207650273,
+      "grad_norm": 1.0878486247478167,
+      "learning_rate": 0.00013479962481168767,
+      "loss": 12.73,
+      "step": 6952
+    },
+    {
+      "epoch": 1.2159562841530054,
+      "grad_norm": 1.0414231627001462,
+      "learning_rate": 0.00013478192342635605,
+      "loss": 12.9907,
+      "step": 6953
+    },
+    {
+      "epoch": 1.2161311475409837,
+      "grad_norm": 0.886233279658895,
+      "learning_rate": 0.00013476422080108952,
+      "loss": 12.7727,
+      "step": 6954
+    },
+    {
+      "epoch": 1.2163060109289618,
+      "grad_norm": 1.1136379258212963,
+      "learning_rate": 0.0001347465169365191,
+      "loss": 12.7545,
+      "step": 6955
+    },
+    {
+      "epoch": 1.21648087431694,
+      "grad_norm": 0.9403679827650228,
+      "learning_rate": 0.00013472881183327595,
+      "loss": 12.7509,
+      "step": 6956
+    },
+    {
+      "epoch": 1.216655737704918,
+      "grad_norm": 1.0901683500229138,
+      "learning_rate": 0.00013471110549199126,
+      "loss": 12.7513,
+      "step": 6957
+    },
+    {
+      "epoch": 1.2168306010928962,
+      "grad_norm": 0.9274245778722934,
+      "learning_rate": 0.0001346933979132962,
+      "loss": 12.8303,
+      "step": 6958
+    },
+    {
+      "epoch": 1.2170054644808743,
+      "grad_norm": 1.077118264954372,
+      "learning_rate": 0.00013467568909782202,
+      "loss": 12.5546,
+      "step": 6959
+    },
+    {
+      "epoch": 1.2171803278688524,
+      "grad_norm": 1.019076550619029,
+      "learning_rate": 0.00013465797904620006,
+      "loss": 13.0121,
+      "step": 6960
+    },
+    {
+      "epoch": 1.2173551912568306,
+      "grad_norm": 0.9768982132868937,
+      "learning_rate": 0.00013464026775906164,
+      "loss": 12.5658,
+      "step": 6961
+    },
+    {
+      "epoch": 1.2175300546448087,
+      "grad_norm": 1.0120405696315398,
+      "learning_rate": 0.00013462255523703816,
+      "loss": 12.7575,
+      "step": 6962
+    },
+    {
+      "epoch": 1.2177049180327868,
+      "grad_norm": 0.9154796926963928,
+      "learning_rate": 0.000134604841480761,
+      "loss": 12.5467,
+      "step": 6963
+    },
+    {
+      "epoch": 1.217879781420765,
+      "grad_norm": 0.9688787840703305,
+      "learning_rate": 0.0001345871264908617,
+      "loss": 12.7577,
+      "step": 6964
+    },
+    {
+      "epoch": 1.2180546448087433,
+      "grad_norm": 0.9670622007515419,
+      "learning_rate": 0.00013456941026797174,
+      "loss": 12.7473,
+      "step": 6965
+    },
+    {
+      "epoch": 1.2182295081967214,
+      "grad_norm": 1.1455096742739677,
+      "learning_rate": 0.0001345516928127227,
+      "loss": 12.8364,
+      "step": 6966
+    },
+    {
+      "epoch": 1.2184043715846995,
+      "grad_norm": 1.032611526309827,
+      "learning_rate": 0.00013453397412574618,
+      "loss": 12.7646,
+      "step": 6967
+    },
+    {
+      "epoch": 1.2185792349726776,
+      "grad_norm": 1.0831751092560942,
+      "learning_rate": 0.00013451625420767384,
+      "loss": 12.7898,
+      "step": 6968
+    },
+    {
+      "epoch": 1.2187540983606557,
+      "grad_norm": 1.021089891756914,
+      "learning_rate": 0.00013449853305913734,
+      "loss": 12.3654,
+      "step": 6969
+    },
+    {
+      "epoch": 1.2189289617486339,
+      "grad_norm": 0.9953213803179233,
+      "learning_rate": 0.00013448081068076847,
+      "loss": 12.8283,
+      "step": 6970
+    },
+    {
+      "epoch": 1.219103825136612,
+      "grad_norm": 0.9516324471844034,
+      "learning_rate": 0.00013446308707319897,
+      "loss": 12.7384,
+      "step": 6971
+    },
+    {
+      "epoch": 1.21927868852459,
+      "grad_norm": 0.8321181296108124,
+      "learning_rate": 0.0001344453622370607,
+      "loss": 12.785,
+      "step": 6972
+    },
+    {
+      "epoch": 1.2194535519125682,
+      "grad_norm": 0.9257217810801244,
+      "learning_rate": 0.00013442763617298547,
+      "loss": 12.5309,
+      "step": 6973
+    },
+    {
+      "epoch": 1.2196284153005466,
+      "grad_norm": 0.817258793819358,
+      "learning_rate": 0.0001344099088816053,
+      "loss": 12.6765,
+      "step": 6974
+    },
+    {
+      "epoch": 1.2198032786885247,
+      "grad_norm": 1.1161581165428032,
+      "learning_rate": 0.00013439218036355205,
+      "loss": 13.0061,
+      "step": 6975
+    },
+    {
+      "epoch": 1.2199781420765028,
+      "grad_norm": 0.9168017319785344,
+      "learning_rate": 0.00013437445061945775,
+      "loss": 12.9022,
+      "step": 6976
+    },
+    {
+      "epoch": 1.220153005464481,
+      "grad_norm": 0.890882154978607,
+      "learning_rate": 0.00013435671964995443,
+      "loss": 12.7629,
+      "step": 6977
+    },
+    {
+      "epoch": 1.220327868852459,
+      "grad_norm": 1.0151189642501908,
+      "learning_rate": 0.00013433898745567418,
+      "loss": 12.6869,
+      "step": 6978
+    },
+    {
+      "epoch": 1.2205027322404371,
+      "grad_norm": 0.9272366205480601,
+      "learning_rate": 0.0001343212540372492,
+      "loss": 12.593,
+      "step": 6979
+    },
+    {
+      "epoch": 1.2206775956284153,
+      "grad_norm": 0.8963944255984798,
+      "learning_rate": 0.0001343035193953116,
+      "loss": 12.7931,
+      "step": 6980
+    },
+    {
+      "epoch": 1.2208524590163934,
+      "grad_norm": 0.9014681780481516,
+      "learning_rate": 0.00013428578353049363,
+      "loss": 12.7434,
+      "step": 6981
+    },
+    {
+      "epoch": 1.2210273224043715,
+      "grad_norm": 0.9502553019688095,
+      "learning_rate": 0.0001342680464434275,
+      "loss": 12.9521,
+      "step": 6982
+    },
+    {
+      "epoch": 1.2212021857923498,
+      "grad_norm": 0.9255324331097006,
+      "learning_rate": 0.00013425030813474557,
+      "loss": 12.6695,
+      "step": 6983
+    },
+    {
+      "epoch": 1.2213770491803277,
+      "grad_norm": 0.9781511058424792,
+      "learning_rate": 0.00013423256860508015,
+      "loss": 12.5981,
+      "step": 6984
+    },
+    {
+      "epoch": 1.221551912568306,
+      "grad_norm": 0.9473939309222359,
+      "learning_rate": 0.00013421482785506368,
+      "loss": 12.7189,
+      "step": 6985
+    },
+    {
+      "epoch": 1.2217267759562842,
+      "grad_norm": 1.0856750166272642,
+      "learning_rate": 0.00013419708588532854,
+      "loss": 12.5475,
+      "step": 6986
+    },
+    {
+      "epoch": 1.2219016393442623,
+      "grad_norm": 0.8848904127257012,
+      "learning_rate": 0.0001341793426965073,
+      "loss": 12.7393,
+      "step": 6987
+    },
+    {
+      "epoch": 1.2220765027322404,
+      "grad_norm": 0.9751575478353907,
+      "learning_rate": 0.00013416159828923238,
+      "loss": 12.6877,
+      "step": 6988
+    },
+    {
+      "epoch": 1.2222513661202186,
+      "grad_norm": 0.9793901786652744,
+      "learning_rate": 0.0001341438526641364,
+      "loss": 12.5893,
+      "step": 6989
+    },
+    {
+      "epoch": 1.2224262295081967,
+      "grad_norm": 0.9445715630771777,
+      "learning_rate": 0.000134126105821852,
+      "loss": 12.4192,
+      "step": 6990
+    },
+    {
+      "epoch": 1.2226010928961748,
+      "grad_norm": 0.9427693733567513,
+      "learning_rate": 0.00013410835776301179,
+      "loss": 12.8769,
+      "step": 6991
+    },
+    {
+      "epoch": 1.222775956284153,
+      "grad_norm": 0.9091804633072086,
+      "learning_rate": 0.00013409060848824846,
+      "loss": 12.7179,
+      "step": 6992
+    },
+    {
+      "epoch": 1.222950819672131,
+      "grad_norm": 0.9849038381979641,
+      "learning_rate": 0.00013407285799819476,
+      "loss": 12.9044,
+      "step": 6993
+    },
+    {
+      "epoch": 1.2231256830601094,
+      "grad_norm": 0.9395915238067825,
+      "learning_rate": 0.00013405510629348349,
+      "loss": 12.4208,
+      "step": 6994
+    },
+    {
+      "epoch": 1.2233005464480875,
+      "grad_norm": 0.9285786204519447,
+      "learning_rate": 0.00013403735337474745,
+      "loss": 12.9325,
+      "step": 6995
+    },
+    {
+      "epoch": 1.2234754098360656,
+      "grad_norm": 1.0551235457165258,
+      "learning_rate": 0.0001340195992426196,
+      "loss": 12.577,
+      "step": 6996
+    },
+    {
+      "epoch": 1.2236502732240437,
+      "grad_norm": 1.0556281194060209,
+      "learning_rate": 0.0001340018438977327,
+      "loss": 12.9339,
+      "step": 6997
+    },
+    {
+      "epoch": 1.2238251366120219,
+      "grad_norm": 1.0063640390941702,
+      "learning_rate": 0.00013398408734071983,
+      "loss": 12.8372,
+      "step": 6998
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 1.0755807883494337,
+      "learning_rate": 0.00013396632957221396,
+      "loss": 12.6706,
+      "step": 6999
+    },
+    {
+      "epoch": 1.224174863387978,
+      "grad_norm": 1.0672334919451305,
+      "learning_rate": 0.00013394857059284814,
+      "loss": 12.5695,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2243497267759562,
+      "grad_norm": 1.0370769413581027,
+      "learning_rate": 0.0001339308104032554,
+      "loss": 12.7982,
+      "step": 7001
+    },
+    {
+      "epoch": 1.2245245901639343,
+      "grad_norm": 1.0956219373633853,
+      "learning_rate": 0.00013391304900406894,
+      "loss": 12.4979,
+      "step": 7002
+    },
+    {
+      "epoch": 1.2246994535519127,
+      "grad_norm": 1.0503448339463055,
+      "learning_rate": 0.00013389528639592188,
+      "loss": 12.8438,
+      "step": 7003
+    },
+    {
+      "epoch": 1.2248743169398908,
+      "grad_norm": 1.0247922384054355,
+      "learning_rate": 0.00013387752257944748,
+      "loss": 12.6489,
+      "step": 7004
+    },
+    {
+      "epoch": 1.225049180327869,
+      "grad_norm": 0.8616044854028275,
+      "learning_rate": 0.00013385975755527899,
+      "loss": 12.7229,
+      "step": 7005
+    },
+    {
+      "epoch": 1.225224043715847,
+      "grad_norm": 1.2407267339269274,
+      "learning_rate": 0.0001338419913240497,
+      "loss": 12.6447,
+      "step": 7006
+    },
+    {
+      "epoch": 1.2253989071038252,
+      "grad_norm": 0.8788275910158206,
+      "learning_rate": 0.00013382422388639298,
+      "loss": 12.7187,
+      "step": 7007
+    },
+    {
+      "epoch": 1.2255737704918033,
+      "grad_norm": 1.1346748295159024,
+      "learning_rate": 0.00013380645524294215,
+      "loss": 12.7869,
+      "step": 7008
+    },
+    {
+      "epoch": 1.2257486338797814,
+      "grad_norm": 1.035261253055552,
+      "learning_rate": 0.00013378868539433073,
+      "loss": 12.7846,
+      "step": 7009
+    },
+    {
+      "epoch": 1.2259234972677595,
+      "grad_norm": 0.9079841609527992,
+      "learning_rate": 0.00013377091434119213,
+      "loss": 12.5962,
+      "step": 7010
+    },
+    {
+      "epoch": 1.2260983606557376,
+      "grad_norm": 0.9858579119278681,
+      "learning_rate": 0.0001337531420841599,
+      "loss": 12.9331,
+      "step": 7011
+    },
+    {
+      "epoch": 1.226273224043716,
+      "grad_norm": 1.0157192947416396,
+      "learning_rate": 0.0001337353686238676,
+      "loss": 12.8123,
+      "step": 7012
+    },
+    {
+      "epoch": 1.2264480874316939,
+      "grad_norm": 0.9688403780820822,
+      "learning_rate": 0.0001337175939609488,
+      "loss": 12.6181,
+      "step": 7013
+    },
+    {
+      "epoch": 1.2266229508196722,
+      "grad_norm": 0.9226435449030201,
+      "learning_rate": 0.00013369981809603718,
+      "loss": 12.698,
+      "step": 7014
+    },
+    {
+      "epoch": 1.2267978142076503,
+      "grad_norm": 0.8923499745232336,
+      "learning_rate": 0.00013368204102976643,
+      "loss": 12.8294,
+      "step": 7015
+    },
+    {
+      "epoch": 1.2269726775956284,
+      "grad_norm": 0.9060837829251971,
+      "learning_rate": 0.00013366426276277025,
+      "loss": 12.6637,
+      "step": 7016
+    },
+    {
+      "epoch": 1.2271475409836066,
+      "grad_norm": 0.9518933619633229,
+      "learning_rate": 0.0001336464832956825,
+      "loss": 13.0699,
+      "step": 7017
+    },
+    {
+      "epoch": 1.2273224043715847,
+      "grad_norm": 0.8989376696055529,
+      "learning_rate": 0.00013362870262913684,
+      "loss": 12.509,
+      "step": 7018
+    },
+    {
+      "epoch": 1.2274972677595628,
+      "grad_norm": 0.9353088513570145,
+      "learning_rate": 0.00013361092076376728,
+      "loss": 12.7973,
+      "step": 7019
+    },
+    {
+      "epoch": 1.227672131147541,
+      "grad_norm": 1.021394253588313,
+      "learning_rate": 0.00013359313770020762,
+      "loss": 12.9817,
+      "step": 7020
+    },
+    {
+      "epoch": 1.227846994535519,
+      "grad_norm": 0.9788806420093794,
+      "learning_rate": 0.0001335753534390919,
+      "loss": 12.6931,
+      "step": 7021
+    },
+    {
+      "epoch": 1.2280218579234972,
+      "grad_norm": 1.0248246578744062,
+      "learning_rate": 0.00013355756798105403,
+      "loss": 12.5934,
+      "step": 7022
+    },
+    {
+      "epoch": 1.2281967213114755,
+      "grad_norm": 0.9723461988429306,
+      "learning_rate": 0.00013353978132672807,
+      "loss": 12.8618,
+      "step": 7023
+    },
+    {
+      "epoch": 1.2283715846994536,
+      "grad_norm": 1.2406656446992461,
+      "learning_rate": 0.00013352199347674813,
+      "loss": 12.8269,
+      "step": 7024
+    },
+    {
+      "epoch": 1.2285464480874317,
+      "grad_norm": 1.135523051717477,
+      "learning_rate": 0.00013350420443174822,
+      "loss": 12.7661,
+      "step": 7025
+    },
+    {
+      "epoch": 1.2287213114754099,
+      "grad_norm": 1.1724275715136874,
+      "learning_rate": 0.00013348641419236263,
+      "loss": 12.8601,
+      "step": 7026
+    },
+    {
+      "epoch": 1.228896174863388,
+      "grad_norm": 1.177814603132971,
+      "learning_rate": 0.00013346862275922546,
+      "loss": 12.8931,
+      "step": 7027
+    },
+    {
+      "epoch": 1.229071038251366,
+      "grad_norm": 1.1245932069983022,
+      "learning_rate": 0.000133450830132971,
+      "loss": 12.9892,
+      "step": 7028
+    },
+    {
+      "epoch": 1.2292459016393442,
+      "grad_norm": 1.1065730230680655,
+      "learning_rate": 0.00013343303631423354,
+      "loss": 12.7383,
+      "step": 7029
+    },
+    {
+      "epoch": 1.2294207650273223,
+      "grad_norm": 1.0456675352818623,
+      "learning_rate": 0.00013341524130364743,
+      "loss": 12.8148,
+      "step": 7030
+    },
+    {
+      "epoch": 1.2295956284153005,
+      "grad_norm": 1.064425295405141,
+      "learning_rate": 0.00013339744510184695,
+      "loss": 12.7402,
+      "step": 7031
+    },
+    {
+      "epoch": 1.2297704918032788,
+      "grad_norm": 1.0733685616734108,
+      "learning_rate": 0.00013337964770946659,
+      "loss": 12.7048,
+      "step": 7032
+    },
+    {
+      "epoch": 1.229945355191257,
+      "grad_norm": 0.8887405230524729,
+      "learning_rate": 0.00013336184912714078,
+      "loss": 12.9778,
+      "step": 7033
+    },
+    {
+      "epoch": 1.230120218579235,
+      "grad_norm": 0.9654263428873432,
+      "learning_rate": 0.00013334404935550403,
+      "loss": 12.8009,
+      "step": 7034
+    },
+    {
+      "epoch": 1.2302950819672132,
+      "grad_norm": 0.9191639284917633,
+      "learning_rate": 0.0001333262483951909,
+      "loss": 12.8059,
+      "step": 7035
+    },
+    {
+      "epoch": 1.2304699453551913,
+      "grad_norm": 1.241532629597836,
+      "learning_rate": 0.00013330844624683596,
+      "loss": 12.9115,
+      "step": 7036
+    },
+    {
+      "epoch": 1.2306448087431694,
+      "grad_norm": 0.9300419392102857,
+      "learning_rate": 0.00013329064291107382,
+      "loss": 12.8243,
+      "step": 7037
+    },
+    {
+      "epoch": 1.2308196721311475,
+      "grad_norm": 1.1036375611774047,
+      "learning_rate": 0.00013327283838853916,
+      "loss": 12.782,
+      "step": 7038
+    },
+    {
+      "epoch": 1.2309945355191256,
+      "grad_norm": 0.9714679236515612,
+      "learning_rate": 0.00013325503267986664,
+      "loss": 12.7971,
+      "step": 7039
+    },
+    {
+      "epoch": 1.2311693989071038,
+      "grad_norm": 1.0281826909100609,
+      "learning_rate": 0.00013323722578569109,
+      "loss": 12.9638,
+      "step": 7040
+    },
+    {
+      "epoch": 1.2313442622950819,
+      "grad_norm": 0.9642560101784174,
+      "learning_rate": 0.0001332194177066473,
+      "loss": 12.6563,
+      "step": 7041
+    },
+    {
+      "epoch": 1.23151912568306,
+      "grad_norm": 1.0660494233665843,
+      "learning_rate": 0.00013320160844337004,
+      "loss": 12.5441,
+      "step": 7042
+    },
+    {
+      "epoch": 1.2316939890710383,
+      "grad_norm": 1.5964507516569373,
+      "learning_rate": 0.00013318379799649426,
+      "loss": 12.9723,
+      "step": 7043
+    },
+    {
+      "epoch": 1.2318688524590165,
+      "grad_norm": 0.9986062610799927,
+      "learning_rate": 0.00013316598636665483,
+      "loss": 12.4718,
+      "step": 7044
+    },
+    {
+      "epoch": 1.2320437158469946,
+      "grad_norm": 1.4096075862124138,
+      "learning_rate": 0.00013314817355448676,
+      "loss": 12.913,
+      "step": 7045
+    },
+    {
+      "epoch": 1.2322185792349727,
+      "grad_norm": 0.9727431755116899,
+      "learning_rate": 0.000133130359560625,
+      "loss": 12.8112,
+      "step": 7046
+    },
+    {
+      "epoch": 1.2323934426229508,
+      "grad_norm": 1.414063546517956,
+      "learning_rate": 0.00013311254438570468,
+      "loss": 12.7166,
+      "step": 7047
+    },
+    {
+      "epoch": 1.232568306010929,
+      "grad_norm": 1.00324417242208,
+      "learning_rate": 0.0001330947280303608,
+      "loss": 12.9161,
+      "step": 7048
+    },
+    {
+      "epoch": 1.232743169398907,
+      "grad_norm": 1.0674306153441435,
+      "learning_rate": 0.0001330769104952285,
+      "loss": 12.8772,
+      "step": 7049
+    },
+    {
+      "epoch": 1.2329180327868852,
+      "grad_norm": 0.9947589946619596,
+      "learning_rate": 0.00013305909178094308,
+      "loss": 12.8069,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2330928961748633,
+      "grad_norm": 0.9645649405342874,
+      "learning_rate": 0.0001330412718881396,
+      "loss": 12.4973,
+      "step": 7051
+    },
+    {
+      "epoch": 1.2332677595628416,
+      "grad_norm": 1.1802414951284754,
+      "learning_rate": 0.0001330234508174534,
+      "loss": 12.7304,
+      "step": 7052
+    },
+    {
+      "epoch": 1.2334426229508197,
+      "grad_norm": 1.0010022142877297,
+      "learning_rate": 0.00013300562856951976,
+      "loss": 12.6818,
+      "step": 7053
+    },
+    {
+      "epoch": 1.2336174863387979,
+      "grad_norm": 0.9418034431778709,
+      "learning_rate": 0.00013298780514497404,
+      "loss": 12.7791,
+      "step": 7054
+    },
+    {
+      "epoch": 1.233792349726776,
+      "grad_norm": 0.8987090641225224,
+      "learning_rate": 0.00013296998054445158,
+      "loss": 12.9244,
+      "step": 7055
+    },
+    {
+      "epoch": 1.233967213114754,
+      "grad_norm": 1.0106511534907083,
+      "learning_rate": 0.00013295215476858788,
+      "loss": 12.5663,
+      "step": 7056
+    },
+    {
+      "epoch": 1.2341420765027322,
+      "grad_norm": 0.869438641710853,
+      "learning_rate": 0.0001329343278180183,
+      "loss": 12.0637,
+      "step": 7057
+    },
+    {
+      "epoch": 1.2343169398907103,
+      "grad_norm": 1.0289236412977036,
+      "learning_rate": 0.00013291649969337843,
+      "loss": 12.7346,
+      "step": 7058
+    },
+    {
+      "epoch": 1.2344918032786885,
+      "grad_norm": 0.8927415871849623,
+      "learning_rate": 0.0001328986703953038,
+      "loss": 12.8804,
+      "step": 7059
+    },
+    {
+      "epoch": 1.2346666666666666,
+      "grad_norm": 0.9681431433044527,
+      "learning_rate": 0.00013288083992443004,
+      "loss": 12.8281,
+      "step": 7060
+    },
+    {
+      "epoch": 1.234841530054645,
+      "grad_norm": 1.0076852993768028,
+      "learning_rate": 0.00013286300828139273,
+      "loss": 12.9755,
+      "step": 7061
+    },
+    {
+      "epoch": 1.235016393442623,
+      "grad_norm": 1.112165233704498,
+      "learning_rate": 0.0001328451754668276,
+      "loss": 12.801,
+      "step": 7062
+    },
+    {
+      "epoch": 1.2351912568306012,
+      "grad_norm": 0.867625842793695,
+      "learning_rate": 0.00013282734148137033,
+      "loss": 12.6945,
+      "step": 7063
+    },
+    {
+      "epoch": 1.2353661202185793,
+      "grad_norm": 1.0429430219322071,
+      "learning_rate": 0.00013280950632565668,
+      "loss": 12.8739,
+      "step": 7064
+    },
+    {
+      "epoch": 1.2355409836065574,
+      "grad_norm": 0.9459087380684591,
+      "learning_rate": 0.00013279167000032244,
+      "loss": 12.7352,
+      "step": 7065
+    },
+    {
+      "epoch": 1.2357158469945355,
+      "grad_norm": 0.9607765032334675,
+      "learning_rate": 0.00013277383250600352,
+      "loss": 12.762,
+      "step": 7066
+    },
+    {
+      "epoch": 1.2358907103825136,
+      "grad_norm": 1.0587955310165764,
+      "learning_rate": 0.00013275599384333575,
+      "loss": 13.0285,
+      "step": 7067
+    },
+    {
+      "epoch": 1.2360655737704918,
+      "grad_norm": 0.9907759933697672,
+      "learning_rate": 0.00013273815401295503,
+      "loss": 12.8414,
+      "step": 7068
+    },
+    {
+      "epoch": 1.2362404371584699,
+      "grad_norm": 1.022760274618045,
+      "learning_rate": 0.00013272031301549744,
+      "loss": 12.7375,
+      "step": 7069
+    },
+    {
+      "epoch": 1.236415300546448,
+      "grad_norm": 0.9470718224578069,
+      "learning_rate": 0.00013270247085159887,
+      "loss": 13.0508,
+      "step": 7070
+    },
+    {
+      "epoch": 1.2365901639344261,
+      "grad_norm": 0.9865816774078604,
+      "learning_rate": 0.00013268462752189547,
+      "loss": 12.6713,
+      "step": 7071
+    },
+    {
+      "epoch": 1.2367650273224045,
+      "grad_norm": 0.9863030951036142,
+      "learning_rate": 0.00013266678302702322,
+      "loss": 12.7834,
+      "step": 7072
+    },
+    {
+      "epoch": 1.2369398907103826,
+      "grad_norm": 1.122255940812311,
+      "learning_rate": 0.0001326489373676184,
+      "loss": 12.5922,
+      "step": 7073
+    },
+    {
+      "epoch": 1.2371147540983607,
+      "grad_norm": 0.9661847769610816,
+      "learning_rate": 0.00013263109054431705,
+      "loss": 12.8466,
+      "step": 7074
+    },
+    {
+      "epoch": 1.2372896174863388,
+      "grad_norm": 1.1580975327880423,
+      "learning_rate": 0.00013261324255775547,
+      "loss": 12.6691,
+      "step": 7075
+    },
+    {
+      "epoch": 1.237464480874317,
+      "grad_norm": 1.0156858840856897,
+      "learning_rate": 0.00013259539340856988,
+      "loss": 12.7219,
+      "step": 7076
+    },
+    {
+      "epoch": 1.237639344262295,
+      "grad_norm": 1.004429815510842,
+      "learning_rate": 0.0001325775430973966,
+      "loss": 12.7349,
+      "step": 7077
+    },
+    {
+      "epoch": 1.2378142076502732,
+      "grad_norm": 0.8507677188344421,
+      "learning_rate": 0.000132559691624872,
+      "loss": 12.7644,
+      "step": 7078
+    },
+    {
+      "epoch": 1.2379890710382513,
+      "grad_norm": 1.0050680720176137,
+      "learning_rate": 0.00013254183899163242,
+      "loss": 12.6612,
+      "step": 7079
+    },
+    {
+      "epoch": 1.2381639344262294,
+      "grad_norm": 1.0163422558508362,
+      "learning_rate": 0.0001325239851983143,
+      "loss": 12.5567,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2383387978142077,
+      "grad_norm": 1.0491003479596057,
+      "learning_rate": 0.0001325061302455541,
+      "loss": 12.6057,
+      "step": 7081
+    },
+    {
+      "epoch": 1.2385136612021859,
+      "grad_norm": 0.9315362310916387,
+      "learning_rate": 0.00013248827413398832,
+      "loss": 12.8837,
+      "step": 7082
+    },
+    {
+      "epoch": 1.238688524590164,
+      "grad_norm": 0.9825397178734124,
+      "learning_rate": 0.00013247041686425358,
+      "loss": 12.5173,
+      "step": 7083
+    },
+    {
+      "epoch": 1.238863387978142,
+      "grad_norm": 0.8840471409644404,
+      "learning_rate": 0.00013245255843698638,
+      "loss": 12.8882,
+      "step": 7084
+    },
+    {
+      "epoch": 1.2390382513661202,
+      "grad_norm": 0.9135241405323147,
+      "learning_rate": 0.0001324346988528234,
+      "loss": 12.7666,
+      "step": 7085
+    },
+    {
+      "epoch": 1.2392131147540983,
+      "grad_norm": 0.9721744861970376,
+      "learning_rate": 0.00013241683811240126,
+      "loss": 12.8687,
+      "step": 7086
+    },
+    {
+      "epoch": 1.2393879781420765,
+      "grad_norm": 0.9355866516911652,
+      "learning_rate": 0.00013239897621635676,
+      "loss": 12.929,
+      "step": 7087
+    },
+    {
+      "epoch": 1.2395628415300546,
+      "grad_norm": 1.0125998629020199,
+      "learning_rate": 0.00013238111316532663,
+      "loss": 12.351,
+      "step": 7088
+    },
+    {
+      "epoch": 1.2397377049180327,
+      "grad_norm": 0.860117962146092,
+      "learning_rate": 0.0001323632489599476,
+      "loss": 12.6941,
+      "step": 7089
+    },
+    {
+      "epoch": 1.239912568306011,
+      "grad_norm": 0.9429477590631502,
+      "learning_rate": 0.00013234538360085659,
+      "loss": 12.605,
+      "step": 7090
+    },
+    {
+      "epoch": 1.2400874316939892,
+      "grad_norm": 0.9088965798753909,
+      "learning_rate": 0.00013232751708869042,
+      "loss": 12.7022,
+      "step": 7091
+    },
+    {
+      "epoch": 1.2402622950819673,
+      "grad_norm": 1.0268562930845793,
+      "learning_rate": 0.0001323096494240861,
+      "loss": 12.9152,
+      "step": 7092
+    },
+    {
+      "epoch": 1.2404371584699454,
+      "grad_norm": 0.9759452390077702,
+      "learning_rate": 0.00013229178060768045,
+      "loss": 12.6167,
+      "step": 7093
+    },
+    {
+      "epoch": 1.2406120218579235,
+      "grad_norm": 0.934542150682003,
+      "learning_rate": 0.00013227391064011057,
+      "loss": 12.6602,
+      "step": 7094
+    },
+    {
+      "epoch": 1.2407868852459016,
+      "grad_norm": 0.9698180880164841,
+      "learning_rate": 0.00013225603952201346,
+      "loss": 12.7276,
+      "step": 7095
+    },
+    {
+      "epoch": 1.2409617486338798,
+      "grad_norm": 0.9141897627676152,
+      "learning_rate": 0.0001322381672540263,
+      "loss": 12.9896,
+      "step": 7096
+    },
+    {
+      "epoch": 1.2411366120218579,
+      "grad_norm": 1.0610221597361615,
+      "learning_rate": 0.00013222029383678607,
+      "loss": 12.6911,
+      "step": 7097
+    },
+    {
+      "epoch": 1.241311475409836,
+      "grad_norm": 0.9359582569021324,
+      "learning_rate": 0.00013220241927093006,
+      "loss": 12.6518,
+      "step": 7098
+    },
+    {
+      "epoch": 1.2414863387978141,
+      "grad_norm": 0.9051015227373518,
+      "learning_rate": 0.00013218454355709542,
+      "loss": 12.7124,
+      "step": 7099
+    },
+    {
+      "epoch": 1.2416612021857922,
+      "grad_norm": 0.9518774048259826,
+      "learning_rate": 0.0001321666666959194,
+      "loss": 12.607,
+      "step": 7100
+    },
+    {
+      "epoch": 1.2418360655737706,
+      "grad_norm": 0.986934200415848,
+      "learning_rate": 0.0001321487886880393,
+      "loss": 12.7353,
+      "step": 7101
+    },
+    {
+      "epoch": 1.2420109289617487,
+      "grad_norm": 0.9613093148421648,
+      "learning_rate": 0.00013213090953409243,
+      "loss": 12.9504,
+      "step": 7102
+    },
+    {
+      "epoch": 1.2421857923497268,
+      "grad_norm": 0.9414896504653035,
+      "learning_rate": 0.00013211302923471618,
+      "loss": 12.796,
+      "step": 7103
+    },
+    {
+      "epoch": 1.242360655737705,
+      "grad_norm": 1.0736175593483368,
+      "learning_rate": 0.00013209514779054796,
+      "loss": 12.9263,
+      "step": 7104
+    },
+    {
+      "epoch": 1.242535519125683,
+      "grad_norm": 0.8679941843050337,
+      "learning_rate": 0.00013207726520222523,
+      "loss": 12.658,
+      "step": 7105
+    },
+    {
+      "epoch": 1.2427103825136612,
+      "grad_norm": 0.9644433223873685,
+      "learning_rate": 0.00013205938147038546,
+      "loss": 12.6925,
+      "step": 7106
+    },
+    {
+      "epoch": 1.2428852459016393,
+      "grad_norm": 1.0265021107924732,
+      "learning_rate": 0.00013204149659566623,
+      "loss": 12.9885,
+      "step": 7107
+    },
+    {
+      "epoch": 1.2430601092896174,
+      "grad_norm": 0.9810754025240068,
+      "learning_rate": 0.00013202361057870502,
+      "loss": 12.7863,
+      "step": 7108
+    },
+    {
+      "epoch": 1.2432349726775955,
+      "grad_norm": 1.0193207374714452,
+      "learning_rate": 0.00013200572342013958,
+      "loss": 12.63,
+      "step": 7109
+    },
+    {
+      "epoch": 1.2434098360655739,
+      "grad_norm": 1.0289088215288742,
+      "learning_rate": 0.00013198783512060743,
+      "loss": 12.8191,
+      "step": 7110
+    },
+    {
+      "epoch": 1.243584699453552,
+      "grad_norm": 1.258389873589706,
+      "learning_rate": 0.0001319699456807464,
+      "loss": 12.7615,
+      "step": 7111
+    },
+    {
+      "epoch": 1.24375956284153,
+      "grad_norm": 0.9735890150684879,
+      "learning_rate": 0.0001319520551011941,
+      "loss": 12.96,
+      "step": 7112
+    },
+    {
+      "epoch": 1.2439344262295082,
+      "grad_norm": 1.1839432304461923,
+      "learning_rate": 0.00013193416338258838,
+      "loss": 12.8864,
+      "step": 7113
+    },
+    {
+      "epoch": 1.2441092896174863,
+      "grad_norm": 0.8979446958022619,
+      "learning_rate": 0.00013191627052556705,
+      "loss": 12.6206,
+      "step": 7114
+    },
+    {
+      "epoch": 1.2442841530054645,
+      "grad_norm": 1.0922084179169742,
+      "learning_rate": 0.00013189837653076797,
+      "loss": 12.586,
+      "step": 7115
+    },
+    {
+      "epoch": 1.2444590163934426,
+      "grad_norm": 0.8850667030094259,
+      "learning_rate": 0.00013188048139882908,
+      "loss": 12.7921,
+      "step": 7116
+    },
+    {
+      "epoch": 1.2446338797814207,
+      "grad_norm": 0.9525535457119231,
+      "learning_rate": 0.0001318625851303882,
+      "loss": 12.738,
+      "step": 7117
+    },
+    {
+      "epoch": 1.2448087431693988,
+      "grad_norm": 0.9692024769766523,
+      "learning_rate": 0.00013184468772608346,
+      "loss": 12.7703,
+      "step": 7118
+    },
+    {
+      "epoch": 1.2449836065573772,
+      "grad_norm": 1.0904995721590065,
+      "learning_rate": 0.00013182678918655273,
+      "loss": 12.5923,
+      "step": 7119
+    },
+    {
+      "epoch": 1.2451584699453553,
+      "grad_norm": 1.0175853671490638,
+      "learning_rate": 0.00013180888951243422,
+      "loss": 12.6922,
+      "step": 7120
+    },
+    {
+      "epoch": 1.2453333333333334,
+      "grad_norm": 0.8963626026917078,
+      "learning_rate": 0.00013179098870436592,
+      "loss": 12.8182,
+      "step": 7121
+    },
+    {
+      "epoch": 1.2455081967213115,
+      "grad_norm": 0.9528368239815724,
+      "learning_rate": 0.00013177308676298602,
+      "loss": 12.9323,
+      "step": 7122
+    },
+    {
+      "epoch": 1.2456830601092896,
+      "grad_norm": 0.9042154328680907,
+      "learning_rate": 0.00013175518368893272,
+      "loss": 12.4548,
+      "step": 7123
+    },
+    {
+      "epoch": 1.2458579234972678,
+      "grad_norm": 0.9011426225627774,
+      "learning_rate": 0.00013173727948284422,
+      "loss": 13.0074,
+      "step": 7124
+    },
+    {
+      "epoch": 1.2460327868852459,
+      "grad_norm": 0.9488552589569451,
+      "learning_rate": 0.00013171937414535877,
+      "loss": 12.7539,
+      "step": 7125
+    },
+    {
+      "epoch": 1.246207650273224,
+      "grad_norm": 1.128323161760789,
+      "learning_rate": 0.00013170146767711472,
+      "loss": 12.7238,
+      "step": 7126
+    },
+    {
+      "epoch": 1.2463825136612021,
+      "grad_norm": 1.0005867469338214,
+      "learning_rate": 0.00013168356007875037,
+      "loss": 12.8537,
+      "step": 7127
+    },
+    {
+      "epoch": 1.2465573770491802,
+      "grad_norm": 1.0009907417760797,
+      "learning_rate": 0.00013166565135090411,
+      "loss": 12.774,
+      "step": 7128
+    },
+    {
+      "epoch": 1.2467322404371584,
+      "grad_norm": 0.9452016008393145,
+      "learning_rate": 0.0001316477414942144,
+      "loss": 12.6692,
+      "step": 7129
+    },
+    {
+      "epoch": 1.2469071038251367,
+      "grad_norm": 0.930337260146719,
+      "learning_rate": 0.00013162983050931965,
+      "loss": 12.7642,
+      "step": 7130
+    },
+    {
+      "epoch": 1.2470819672131148,
+      "grad_norm": 1.1821151016947846,
+      "learning_rate": 0.00013161191839685844,
+      "loss": 12.6536,
+      "step": 7131
+    },
+    {
+      "epoch": 1.247256830601093,
+      "grad_norm": 1.00971819950322,
+      "learning_rate": 0.00013159400515746925,
+      "loss": 12.7766,
+      "step": 7132
+    },
+    {
+      "epoch": 1.247431693989071,
+      "grad_norm": 1.1958323859695004,
+      "learning_rate": 0.0001315760907917907,
+      "loss": 12.6978,
+      "step": 7133
+    },
+    {
+      "epoch": 1.2476065573770492,
+      "grad_norm": 1.0088690116740182,
+      "learning_rate": 0.00013155817530046137,
+      "loss": 12.821,
+      "step": 7134
+    },
+    {
+      "epoch": 1.2477814207650273,
+      "grad_norm": 1.1479802510996968,
+      "learning_rate": 0.00013154025868412,
+      "loss": 12.9511,
+      "step": 7135
+    },
+    {
+      "epoch": 1.2479562841530054,
+      "grad_norm": 1.3027456589092854,
+      "learning_rate": 0.00013152234094340525,
+      "loss": 12.9365,
+      "step": 7136
+    },
+    {
+      "epoch": 1.2481311475409835,
+      "grad_norm": 0.8934269895431799,
+      "learning_rate": 0.0001315044220789559,
+      "loss": 12.9629,
+      "step": 7137
+    },
+    {
+      "epoch": 1.2483060109289617,
+      "grad_norm": 0.8867307525977941,
+      "learning_rate": 0.00013148650209141068,
+      "loss": 12.662,
+      "step": 7138
+    },
+    {
+      "epoch": 1.24848087431694,
+      "grad_norm": 1.038739334693772,
+      "learning_rate": 0.00013146858098140844,
+      "loss": 12.8618,
+      "step": 7139
+    },
+    {
+      "epoch": 1.2486557377049181,
+      "grad_norm": 0.8973184828122213,
+      "learning_rate": 0.00013145065874958805,
+      "loss": 12.7578,
+      "step": 7140
+    },
+    {
+      "epoch": 1.2488306010928962,
+      "grad_norm": 1.0217679880564805,
+      "learning_rate": 0.00013143273539658847,
+      "loss": 12.9241,
+      "step": 7141
+    },
+    {
+      "epoch": 1.2490054644808744,
+      "grad_norm": 0.9280699745486137,
+      "learning_rate": 0.00013141481092304855,
+      "loss": 12.495,
+      "step": 7142
+    },
+    {
+      "epoch": 1.2491803278688525,
+      "grad_norm": 1.0869911514835424,
+      "learning_rate": 0.00013139688532960736,
+      "loss": 12.7851,
+      "step": 7143
+    },
+    {
+      "epoch": 1.2493551912568306,
+      "grad_norm": 1.0062589826345467,
+      "learning_rate": 0.00013137895861690385,
+      "loss": 12.9751,
+      "step": 7144
+    },
+    {
+      "epoch": 1.2495300546448087,
+      "grad_norm": 1.2033093858458186,
+      "learning_rate": 0.0001313610307855772,
+      "loss": 13.0139,
+      "step": 7145
+    },
+    {
+      "epoch": 1.2497049180327868,
+      "grad_norm": 0.8994168646882352,
+      "learning_rate": 0.00013134310183626638,
+      "loss": 12.5421,
+      "step": 7146
+    },
+    {
+      "epoch": 1.249879781420765,
+      "grad_norm": 1.254924216760557,
+      "learning_rate": 0.00013132517176961064,
+      "loss": 12.7836,
+      "step": 7147
+    },
+    {
+      "epoch": 1.2500546448087433,
+      "grad_norm": 0.9710450846401139,
+      "learning_rate": 0.00013130724058624908,
+      "loss": 12.6353,
+      "step": 7148
+    },
+    {
+      "epoch": 1.2502295081967212,
+      "grad_norm": 1.071411410194843,
+      "learning_rate": 0.000131289308286821,
+      "loss": 12.9195,
+      "step": 7149
+    },
+    {
+      "epoch": 1.2504043715846995,
+      "grad_norm": 1.042572679437824,
+      "learning_rate": 0.00013127137487196563,
+      "loss": 12.8359,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2505792349726776,
+      "grad_norm": 0.9342257302098903,
+      "learning_rate": 0.00013125344034232227,
+      "loss": 12.7905,
+      "step": 7151
+    },
+    {
+      "epoch": 1.2507540983606558,
+      "grad_norm": 1.0721279149436784,
+      "learning_rate": 0.0001312355046985303,
+      "loss": 12.6278,
+      "step": 7152
+    },
+    {
+      "epoch": 1.2509289617486339,
+      "grad_norm": 0.8417827140729762,
+      "learning_rate": 0.00013121756794122908,
+      "loss": 12.6245,
+      "step": 7153
+    },
+    {
+      "epoch": 1.251103825136612,
+      "grad_norm": 1.3122211879710837,
+      "learning_rate": 0.00013119963007105805,
+      "loss": 12.6158,
+      "step": 7154
+    },
+    {
+      "epoch": 1.2512786885245901,
+      "grad_norm": 0.9345449028772576,
+      "learning_rate": 0.00013118169108865664,
+      "loss": 12.8226,
+      "step": 7155
+    },
+    {
+      "epoch": 1.2514535519125682,
+      "grad_norm": 1.2994148455426373,
+      "learning_rate": 0.00013116375099466438,
+      "loss": 12.5804,
+      "step": 7156
+    },
+    {
+      "epoch": 1.2516284153005466,
+      "grad_norm": 1.0164020052136387,
+      "learning_rate": 0.00013114580978972079,
+      "loss": 12.6454,
+      "step": 7157
+    },
+    {
+      "epoch": 1.2518032786885245,
+      "grad_norm": 1.0498906400252896,
+      "learning_rate": 0.00013112786747446548,
+      "loss": 12.7571,
+      "step": 7158
+    },
+    {
+      "epoch": 1.2519781420765028,
+      "grad_norm": 1.485055619284985,
+      "learning_rate": 0.00013110992404953805,
+      "loss": 12.4526,
+      "step": 7159
+    },
+    {
+      "epoch": 1.252153005464481,
+      "grad_norm": 1.0020477567291395,
+      "learning_rate": 0.00013109197951557822,
+      "loss": 12.8245,
+      "step": 7160
+    },
+    {
+      "epoch": 1.252327868852459,
+      "grad_norm": 1.4427651833290516,
+      "learning_rate": 0.0001310740338732256,
+      "loss": 12.8787,
+      "step": 7161
+    },
+    {
+      "epoch": 1.2525027322404372,
+      "grad_norm": 0.926847379226587,
+      "learning_rate": 0.00013105608712312002,
+      "loss": 12.8294,
+      "step": 7162
+    },
+    {
+      "epoch": 1.2526775956284153,
+      "grad_norm": 1.569761311206467,
+      "learning_rate": 0.0001310381392659012,
+      "loss": 12.6412,
+      "step": 7163
+    },
+    {
+      "epoch": 1.2528524590163934,
+      "grad_norm": 0.9065884044631501,
+      "learning_rate": 0.00013102019030220897,
+      "loss": 12.9299,
+      "step": 7164
+    },
+    {
+      "epoch": 1.2530273224043715,
+      "grad_norm": 1.2482084507605593,
+      "learning_rate": 0.00013100224023268324,
+      "loss": 12.8867,
+      "step": 7165
+    },
+    {
+      "epoch": 1.2532021857923497,
+      "grad_norm": 1.0735277206295022,
+      "learning_rate": 0.00013098428905796382,
+      "loss": 12.5555,
+      "step": 7166
+    },
+    {
+      "epoch": 1.2533770491803278,
+      "grad_norm": 0.8966636012685265,
+      "learning_rate": 0.00013096633677869073,
+      "loss": 12.7255,
+      "step": 7167
+    },
+    {
+      "epoch": 1.2535519125683061,
+      "grad_norm": 0.9567121210274179,
+      "learning_rate": 0.00013094838339550386,
+      "loss": 12.6556,
+      "step": 7168
+    },
+    {
+      "epoch": 1.2537267759562842,
+      "grad_norm": 1.0584211971286455,
+      "learning_rate": 0.00013093042890904334,
+      "loss": 12.8255,
+      "step": 7169
+    },
+    {
+      "epoch": 1.2539016393442624,
+      "grad_norm": 0.9688320852175827,
+      "learning_rate": 0.00013091247331994914,
+      "loss": 12.8195,
+      "step": 7170
+    },
+    {
+      "epoch": 1.2540765027322405,
+      "grad_norm": 1.051043199592745,
+      "learning_rate": 0.0001308945166288614,
+      "loss": 12.727,
+      "step": 7171
+    },
+    {
+      "epoch": 1.2542513661202186,
+      "grad_norm": 0.9237202108479016,
+      "learning_rate": 0.0001308765588364202,
+      "loss": 12.7572,
+      "step": 7172
+    },
+    {
+      "epoch": 1.2544262295081967,
+      "grad_norm": 1.0817155719401357,
+      "learning_rate": 0.0001308585999432658,
+      "loss": 12.5414,
+      "step": 7173
+    },
+    {
+      "epoch": 1.2546010928961748,
+      "grad_norm": 1.0131801330537675,
+      "learning_rate": 0.00013084063995003834,
+      "loss": 12.6966,
+      "step": 7174
+    },
+    {
+      "epoch": 1.254775956284153,
+      "grad_norm": 0.9618558828636264,
+      "learning_rate": 0.00013082267885737814,
+      "loss": 12.6328,
+      "step": 7175
+    },
+    {
+      "epoch": 1.254950819672131,
+      "grad_norm": 1.0211323105767214,
+      "learning_rate": 0.0001308047166659254,
+      "loss": 12.6533,
+      "step": 7176
+    },
+    {
+      "epoch": 1.2551256830601094,
+      "grad_norm": 0.9634351631875981,
+      "learning_rate": 0.0001307867533763205,
+      "loss": 12.6475,
+      "step": 7177
+    },
+    {
+      "epoch": 1.2553005464480873,
+      "grad_norm": 0.9727765406974627,
+      "learning_rate": 0.00013076878898920385,
+      "loss": 12.8742,
+      "step": 7178
+    },
+    {
+      "epoch": 1.2554754098360656,
+      "grad_norm": 0.8906160327306498,
+      "learning_rate": 0.0001307508235052158,
+      "loss": 12.5466,
+      "step": 7179
+    },
+    {
+      "epoch": 1.2556502732240438,
+      "grad_norm": 0.9626546825966282,
+      "learning_rate": 0.00013073285692499683,
+      "loss": 12.8164,
+      "step": 7180
+    },
+    {
+      "epoch": 1.2558251366120219,
+      "grad_norm": 0.892506221528571,
+      "learning_rate": 0.0001307148892491874,
+      "loss": 12.7445,
+      "step": 7181
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 0.9266687926257567,
+      "learning_rate": 0.0001306969204784281,
+      "loss": 12.8317,
+      "step": 7182
+    },
+    {
+      "epoch": 1.2561748633879781,
+      "grad_norm": 0.9781385482203775,
+      "learning_rate": 0.0001306789506133594,
+      "loss": 12.9283,
+      "step": 7183
+    },
+    {
+      "epoch": 1.2563497267759562,
+      "grad_norm": 0.9302294050163131,
+      "learning_rate": 0.00013066097965462196,
+      "loss": 13.04,
+      "step": 7184
+    },
+    {
+      "epoch": 1.2565245901639344,
+      "grad_norm": 0.8549929029831052,
+      "learning_rate": 0.00013064300760285644,
+      "loss": 12.5547,
+      "step": 7185
+    },
+    {
+      "epoch": 1.2566994535519127,
+      "grad_norm": 0.9088239535392652,
+      "learning_rate": 0.00013062503445870349,
+      "loss": 12.8146,
+      "step": 7186
+    },
+    {
+      "epoch": 1.2568743169398906,
+      "grad_norm": 0.9732381745462171,
+      "learning_rate": 0.00013060706022280385,
+      "loss": 13.0616,
+      "step": 7187
+    },
+    {
+      "epoch": 1.257049180327869,
+      "grad_norm": 1.1127722817561059,
+      "learning_rate": 0.00013058908489579828,
+      "loss": 12.6644,
+      "step": 7188
+    },
+    {
+      "epoch": 1.257224043715847,
+      "grad_norm": 1.1071548539004268,
+      "learning_rate": 0.00013057110847832757,
+      "loss": 12.7721,
+      "step": 7189
+    },
+    {
+      "epoch": 1.2573989071038252,
+      "grad_norm": 0.9667166472950688,
+      "learning_rate": 0.00013055313097103254,
+      "loss": 12.612,
+      "step": 7190
+    },
+    {
+      "epoch": 1.2575737704918033,
+      "grad_norm": 0.9139744817355395,
+      "learning_rate": 0.00013053515237455412,
+      "loss": 12.5972,
+      "step": 7191
+    },
+    {
+      "epoch": 1.2577486338797814,
+      "grad_norm": 0.9591746149942626,
+      "learning_rate": 0.00013051717268953316,
+      "loss": 12.7776,
+      "step": 7192
+    },
+    {
+      "epoch": 1.2579234972677595,
+      "grad_norm": 0.9515625907336246,
+      "learning_rate": 0.00013049919191661065,
+      "loss": 12.5516,
+      "step": 7193
+    },
+    {
+      "epoch": 1.2580983606557377,
+      "grad_norm": 0.8912092468102829,
+      "learning_rate": 0.00013048121005642763,
+      "loss": 12.7524,
+      "step": 7194
+    },
+    {
+      "epoch": 1.2582732240437158,
+      "grad_norm": 0.884091593311582,
+      "learning_rate": 0.00013046322710962504,
+      "loss": 12.7698,
+      "step": 7195
+    },
+    {
+      "epoch": 1.258448087431694,
+      "grad_norm": 1.0381240156496145,
+      "learning_rate": 0.000130445243076844,
+      "loss": 12.6449,
+      "step": 7196
+    },
+    {
+      "epoch": 1.2586229508196722,
+      "grad_norm": 0.8855594622766011,
+      "learning_rate": 0.00013042725795872566,
+      "loss": 12.6612,
+      "step": 7197
+    },
+    {
+      "epoch": 1.2587978142076504,
+      "grad_norm": 0.8538406331014228,
+      "learning_rate": 0.00013040927175591107,
+      "loss": 12.6915,
+      "step": 7198
+    },
+    {
+      "epoch": 1.2589726775956285,
+      "grad_norm": 0.8445656252424755,
+      "learning_rate": 0.0001303912844690415,
+      "loss": 12.4014,
+      "step": 7199
+    },
+    {
+      "epoch": 1.2591475409836066,
+      "grad_norm": 1.0873407320468667,
+      "learning_rate": 0.00013037329609875812,
+      "loss": 12.9491,
+      "step": 7200
+    },
+    {
+      "epoch": 1.2593224043715847,
+      "grad_norm": 0.931066430940458,
+      "learning_rate": 0.00013035530664570226,
+      "loss": 12.7769,
+      "step": 7201
+    },
+    {
+      "epoch": 1.2594972677595628,
+      "grad_norm": 0.8913549178270321,
+      "learning_rate": 0.00013033731611051516,
+      "loss": 12.9083,
+      "step": 7202
+    },
+    {
+      "epoch": 1.259672131147541,
+      "grad_norm": 0.9528663338315482,
+      "learning_rate": 0.00013031932449383818,
+      "loss": 12.8912,
+      "step": 7203
+    },
+    {
+      "epoch": 1.259846994535519,
+      "grad_norm": 0.9425344735655625,
+      "learning_rate": 0.00013030133179631269,
+      "loss": 12.7776,
+      "step": 7204
+    },
+    {
+      "epoch": 1.2600218579234972,
+      "grad_norm": 0.8918757116963639,
+      "learning_rate": 0.00013028333801858017,
+      "loss": 12.7718,
+      "step": 7205
+    },
+    {
+      "epoch": 1.2601967213114755,
+      "grad_norm": 0.9282233022726376,
+      "learning_rate": 0.000130265343161282,
+      "loss": 12.8504,
+      "step": 7206
+    },
+    {
+      "epoch": 1.2603715846994534,
+      "grad_norm": 0.8734122131075301,
+      "learning_rate": 0.00013024734722505977,
+      "loss": 12.6005,
+      "step": 7207
+    },
+    {
+      "epoch": 1.2605464480874318,
+      "grad_norm": 0.840634438128401,
+      "learning_rate": 0.00013022935021055488,
+      "loss": 12.807,
+      "step": 7208
+    },
+    {
+      "epoch": 1.26072131147541,
+      "grad_norm": 1.031509718265783,
+      "learning_rate": 0.000130211352118409,
+      "loss": 12.7005,
+      "step": 7209
+    },
+    {
+      "epoch": 1.260896174863388,
+      "grad_norm": 0.8892252811870535,
+      "learning_rate": 0.00013019335294926375,
+      "loss": 12.7366,
+      "step": 7210
+    },
+    {
+      "epoch": 1.2610710382513661,
+      "grad_norm": 0.8784778807693947,
+      "learning_rate": 0.00013017535270376065,
+      "loss": 12.766,
+      "step": 7211
+    },
+    {
+      "epoch": 1.2612459016393442,
+      "grad_norm": 0.9020691235679404,
+      "learning_rate": 0.00013015735138254155,
+      "loss": 12.923,
+      "step": 7212
+    },
+    {
+      "epoch": 1.2614207650273224,
+      "grad_norm": 0.9632413011805839,
+      "learning_rate": 0.00013013934898624808,
+      "loss": 12.6237,
+      "step": 7213
+    },
+    {
+      "epoch": 1.2615956284153005,
+      "grad_norm": 0.9706117765457414,
+      "learning_rate": 0.0001301213455155221,
+      "loss": 12.7637,
+      "step": 7214
+    },
+    {
+      "epoch": 1.2617704918032788,
+      "grad_norm": 0.9960872035777134,
+      "learning_rate": 0.00013010334097100524,
+      "loss": 12.6965,
+      "step": 7215
+    },
+    {
+      "epoch": 1.2619453551912567,
+      "grad_norm": 0.8708452746006209,
+      "learning_rate": 0.00013008533535333952,
+      "loss": 12.7652,
+      "step": 7216
+    },
+    {
+      "epoch": 1.262120218579235,
+      "grad_norm": 0.8352125727387708,
+      "learning_rate": 0.00013006732866316674,
+      "loss": 12.653,
+      "step": 7217
+    },
+    {
+      "epoch": 1.2622950819672132,
+      "grad_norm": 0.842698982306876,
+      "learning_rate": 0.00013004932090112883,
+      "loss": 12.7758,
+      "step": 7218
+    },
+    {
+      "epoch": 1.2624699453551913,
+      "grad_norm": 0.9488555687883907,
+      "learning_rate": 0.00013003131206786768,
+      "loss": 13.1459,
+      "step": 7219
+    },
+    {
+      "epoch": 1.2626448087431694,
+      "grad_norm": 0.9894234301388708,
+      "learning_rate": 0.0001300133021640254,
+      "loss": 12.7926,
+      "step": 7220
+    },
+    {
+      "epoch": 1.2628196721311475,
+      "grad_norm": 1.0114660111523381,
+      "learning_rate": 0.00012999529119024395,
+      "loss": 12.8922,
+      "step": 7221
+    },
+    {
+      "epoch": 1.2629945355191257,
+      "grad_norm": 0.9663648670549007,
+      "learning_rate": 0.0001299772791471654,
+      "loss": 12.8841,
+      "step": 7222
+    },
+    {
+      "epoch": 1.2631693989071038,
+      "grad_norm": 0.936949696061778,
+      "learning_rate": 0.00012995926603543188,
+      "loss": 12.6916,
+      "step": 7223
+    },
+    {
+      "epoch": 1.263344262295082,
+      "grad_norm": 0.8274145276476531,
+      "learning_rate": 0.00012994125185568554,
+      "loss": 12.5443,
+      "step": 7224
+    },
+    {
+      "epoch": 1.26351912568306,
+      "grad_norm": 0.8627218735498686,
+      "learning_rate": 0.00012992323660856856,
+      "loss": 12.7944,
+      "step": 7225
+    },
+    {
+      "epoch": 1.2636939890710384,
+      "grad_norm": 0.9173471918176785,
+      "learning_rate": 0.00012990522029472312,
+      "loss": 12.5605,
+      "step": 7226
+    },
+    {
+      "epoch": 1.2638688524590163,
+      "grad_norm": 0.9466172301206995,
+      "learning_rate": 0.00012988720291479155,
+      "loss": 12.8004,
+      "step": 7227
+    },
+    {
+      "epoch": 1.2640437158469946,
+      "grad_norm": 0.8883329677831308,
+      "learning_rate": 0.00012986918446941613,
+      "loss": 12.8537,
+      "step": 7228
+    },
+    {
+      "epoch": 1.2642185792349727,
+      "grad_norm": 0.8989808603812869,
+      "learning_rate": 0.0001298511649592391,
+      "loss": 12.6196,
+      "step": 7229
+    },
+    {
+      "epoch": 1.2643934426229508,
+      "grad_norm": 1.0654910204852972,
+      "learning_rate": 0.000129833144384903,
+      "loss": 12.7962,
+      "step": 7230
+    },
+    {
+      "epoch": 1.264568306010929,
+      "grad_norm": 1.0777472184096157,
+      "learning_rate": 0.00012981512274705016,
+      "loss": 12.8617,
+      "step": 7231
+    },
+    {
+      "epoch": 1.264743169398907,
+      "grad_norm": 0.8891535053075018,
+      "learning_rate": 0.00012979710004632297,
+      "loss": 12.3212,
+      "step": 7232
+    },
+    {
+      "epoch": 1.2649180327868852,
+      "grad_norm": 1.013139000421492,
+      "learning_rate": 0.00012977907628336404,
+      "loss": 12.9844,
+      "step": 7233
+    },
+    {
+      "epoch": 1.2650928961748633,
+      "grad_norm": 0.8580330866239956,
+      "learning_rate": 0.0001297610514588158,
+      "loss": 12.7636,
+      "step": 7234
+    },
+    {
+      "epoch": 1.2652677595628417,
+      "grad_norm": 0.8992480111975468,
+      "learning_rate": 0.00012974302557332087,
+      "loss": 12.8321,
+      "step": 7235
+    },
+    {
+      "epoch": 1.2654426229508196,
+      "grad_norm": 1.068401482190471,
+      "learning_rate": 0.0001297249986275218,
+      "loss": 12.7508,
+      "step": 7236
+    },
+    {
+      "epoch": 1.265617486338798,
+      "grad_norm": 0.8870549868560514,
+      "learning_rate": 0.00012970697062206126,
+      "loss": 12.4828,
+      "step": 7237
+    },
+    {
+      "epoch": 1.265792349726776,
+      "grad_norm": 1.0229377778551425,
+      "learning_rate": 0.00012968894155758194,
+      "loss": 12.7166,
+      "step": 7238
+    },
+    {
+      "epoch": 1.2659672131147541,
+      "grad_norm": 0.9611666894946574,
+      "learning_rate": 0.00012967091143472652,
+      "loss": 12.7474,
+      "step": 7239
+    },
+    {
+      "epoch": 1.2661420765027322,
+      "grad_norm": 1.0202369008103225,
+      "learning_rate": 0.0001296528802541378,
+      "loss": 12.6365,
+      "step": 7240
+    },
+    {
+      "epoch": 1.2663169398907104,
+      "grad_norm": 0.9764170943518881,
+      "learning_rate": 0.00012963484801645854,
+      "loss": 12.646,
+      "step": 7241
+    },
+    {
+      "epoch": 1.2664918032786885,
+      "grad_norm": 0.9905275391760269,
+      "learning_rate": 0.00012961681472233155,
+      "loss": 12.943,
+      "step": 7242
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 0.9624328175283896,
+      "learning_rate": 0.00012959878037239973,
+      "loss": 12.7468,
+      "step": 7243
+    },
+    {
+      "epoch": 1.2668415300546447,
+      "grad_norm": 1.0173628440177487,
+      "learning_rate": 0.000129580744967306,
+      "loss": 12.6663,
+      "step": 7244
+    },
+    {
+      "epoch": 1.2670163934426228,
+      "grad_norm": 0.8836845704321192,
+      "learning_rate": 0.00012956270850769325,
+      "loss": 12.8659,
+      "step": 7245
+    },
+    {
+      "epoch": 1.2671912568306012,
+      "grad_norm": 0.9579890210789309,
+      "learning_rate": 0.0001295446709942045,
+      "loss": 12.8666,
+      "step": 7246
+    },
+    {
+      "epoch": 1.2673661202185793,
+      "grad_norm": 0.8994229984613087,
+      "learning_rate": 0.00012952663242748273,
+      "loss": 12.4022,
+      "step": 7247
+    },
+    {
+      "epoch": 1.2675409836065574,
+      "grad_norm": 0.985529337286308,
+      "learning_rate": 0.000129508592808171,
+      "loss": 12.7587,
+      "step": 7248
+    },
+    {
+      "epoch": 1.2677158469945355,
+      "grad_norm": 1.0790271665300903,
+      "learning_rate": 0.00012949055213691243,
+      "loss": 12.7318,
+      "step": 7249
+    },
+    {
+      "epoch": 1.2678907103825137,
+      "grad_norm": 0.952524514616297,
+      "learning_rate": 0.00012947251041435013,
+      "loss": 12.4792,
+      "step": 7250
+    },
+    {
+      "epoch": 1.2680655737704918,
+      "grad_norm": 0.8771114942180513,
+      "learning_rate": 0.00012945446764112728,
+      "loss": 12.7966,
+      "step": 7251
+    },
+    {
+      "epoch": 1.26824043715847,
+      "grad_norm": 0.9648787689836018,
+      "learning_rate": 0.0001294364238178871,
+      "loss": 12.9597,
+      "step": 7252
+    },
+    {
+      "epoch": 1.268415300546448,
+      "grad_norm": 0.8935787826487378,
+      "learning_rate": 0.00012941837894527272,
+      "loss": 12.5409,
+      "step": 7253
+    },
+    {
+      "epoch": 1.2685901639344261,
+      "grad_norm": 1.0180478230729935,
+      "learning_rate": 0.00012940033302392755,
+      "loss": 12.7733,
+      "step": 7254
+    },
+    {
+      "epoch": 1.2687650273224045,
+      "grad_norm": 0.9129662885096282,
+      "learning_rate": 0.00012938228605449489,
+      "loss": 12.6711,
+      "step": 7255
+    },
+    {
+      "epoch": 1.2689398907103824,
+      "grad_norm": 1.1095754194626464,
+      "learning_rate": 0.00012936423803761802,
+      "loss": 12.9629,
+      "step": 7256
+    },
+    {
+      "epoch": 1.2691147540983607,
+      "grad_norm": 0.9966103291201412,
+      "learning_rate": 0.00012934618897394038,
+      "loss": 12.7748,
+      "step": 7257
+    },
+    {
+      "epoch": 1.2692896174863388,
+      "grad_norm": 0.9208516011577458,
+      "learning_rate": 0.00012932813886410536,
+      "loss": 12.4935,
+      "step": 7258
+    },
+    {
+      "epoch": 1.269464480874317,
+      "grad_norm": 0.9403206693054532,
+      "learning_rate": 0.0001293100877087565,
+      "loss": 12.7516,
+      "step": 7259
+    },
+    {
+      "epoch": 1.269639344262295,
+      "grad_norm": 0.9950257825294416,
+      "learning_rate": 0.00012929203550853723,
+      "loss": 12.4495,
+      "step": 7260
+    },
+    {
+      "epoch": 1.2698142076502732,
+      "grad_norm": 0.8802214837961195,
+      "learning_rate": 0.00012927398226409114,
+      "loss": 12.7776,
+      "step": 7261
+    },
+    {
+      "epoch": 1.2699890710382513,
+      "grad_norm": 0.8386421430306265,
+      "learning_rate": 0.0001292559279760618,
+      "loss": 12.6008,
+      "step": 7262
+    },
+    {
+      "epoch": 1.2701639344262294,
+      "grad_norm": 0.8270709314912761,
+      "learning_rate": 0.0001292378726450928,
+      "loss": 12.5777,
+      "step": 7263
+    },
+    {
+      "epoch": 1.2703387978142078,
+      "grad_norm": 0.9371279396289066,
+      "learning_rate": 0.00012921981627182777,
+      "loss": 12.7086,
+      "step": 7264
+    },
+    {
+      "epoch": 1.2705136612021857,
+      "grad_norm": 0.8371853051952268,
+      "learning_rate": 0.00012920175885691045,
+      "loss": 12.8024,
+      "step": 7265
+    },
+    {
+      "epoch": 1.270688524590164,
+      "grad_norm": 0.9546818843285667,
+      "learning_rate": 0.00012918370040098455,
+      "loss": 12.6812,
+      "step": 7266
+    },
+    {
+      "epoch": 1.2708633879781421,
+      "grad_norm": 0.9251162737156448,
+      "learning_rate": 0.0001291656409046938,
+      "loss": 12.7928,
+      "step": 7267
+    },
+    {
+      "epoch": 1.2710382513661203,
+      "grad_norm": 0.9065939852065447,
+      "learning_rate": 0.00012914758036868204,
+      "loss": 12.5786,
+      "step": 7268
+    },
+    {
+      "epoch": 1.2712131147540984,
+      "grad_norm": 0.9204601119491929,
+      "learning_rate": 0.00012912951879359312,
+      "loss": 12.6463,
+      "step": 7269
+    },
+    {
+      "epoch": 1.2713879781420765,
+      "grad_norm": 0.9225817415730266,
+      "learning_rate": 0.00012911145618007088,
+      "loss": 12.7065,
+      "step": 7270
+    },
+    {
+      "epoch": 1.2715628415300546,
+      "grad_norm": 0.9187050914645966,
+      "learning_rate": 0.00012909339252875922,
+      "loss": 12.5237,
+      "step": 7271
+    },
+    {
+      "epoch": 1.2717377049180327,
+      "grad_norm": 0.9958015024961702,
+      "learning_rate": 0.00012907532784030217,
+      "loss": 12.6916,
+      "step": 7272
+    },
+    {
+      "epoch": 1.2719125683060108,
+      "grad_norm": 0.9608020963354621,
+      "learning_rate": 0.00012905726211534358,
+      "loss": 12.8137,
+      "step": 7273
+    },
+    {
+      "epoch": 1.272087431693989,
+      "grad_norm": 0.9488000416661233,
+      "learning_rate": 0.00012903919535452757,
+      "loss": 12.8191,
+      "step": 7274
+    },
+    {
+      "epoch": 1.2722622950819673,
+      "grad_norm": 0.9919171002505768,
+      "learning_rate": 0.00012902112755849818,
+      "loss": 12.8929,
+      "step": 7275
+    },
+    {
+      "epoch": 1.2724371584699454,
+      "grad_norm": 1.078810450937001,
+      "learning_rate": 0.0001290030587278995,
+      "loss": 12.6695,
+      "step": 7276
+    },
+    {
+      "epoch": 1.2726120218579235,
+      "grad_norm": 0.911235459403405,
+      "learning_rate": 0.00012898498886337563,
+      "loss": 12.4908,
+      "step": 7277
+    },
+    {
+      "epoch": 1.2727868852459017,
+      "grad_norm": 0.954058478278222,
+      "learning_rate": 0.00012896691796557082,
+      "loss": 13.0482,
+      "step": 7278
+    },
+    {
+      "epoch": 1.2729617486338798,
+      "grad_norm": 0.8773021226898946,
+      "learning_rate": 0.00012894884603512923,
+      "loss": 12.6526,
+      "step": 7279
+    },
+    {
+      "epoch": 1.273136612021858,
+      "grad_norm": 0.9814739290552918,
+      "learning_rate": 0.00012893077307269508,
+      "loss": 12.7774,
+      "step": 7280
+    },
+    {
+      "epoch": 1.273311475409836,
+      "grad_norm": 0.9805132778367996,
+      "learning_rate": 0.00012891269907891267,
+      "loss": 12.5567,
+      "step": 7281
+    },
+    {
+      "epoch": 1.2734863387978141,
+      "grad_norm": 0.9085050256693168,
+      "learning_rate": 0.00012889462405442633,
+      "loss": 12.8297,
+      "step": 7282
+    },
+    {
+      "epoch": 1.2736612021857923,
+      "grad_norm": 0.8430624512004506,
+      "learning_rate": 0.00012887654799988038,
+      "loss": 12.681,
+      "step": 7283
+    },
+    {
+      "epoch": 1.2738360655737706,
+      "grad_norm": 0.8910243940208875,
+      "learning_rate": 0.00012885847091591925,
+      "loss": 12.7676,
+      "step": 7284
+    },
+    {
+      "epoch": 1.2740109289617485,
+      "grad_norm": 0.9296559080172977,
+      "learning_rate": 0.00012884039280318735,
+      "loss": 12.8008,
+      "step": 7285
+    },
+    {
+      "epoch": 1.2741857923497268,
+      "grad_norm": 0.9319873782005653,
+      "learning_rate": 0.00012882231366232914,
+      "loss": 12.7853,
+      "step": 7286
+    },
+    {
+      "epoch": 1.274360655737705,
+      "grad_norm": 0.8914193258633966,
+      "learning_rate": 0.00012880423349398914,
+      "loss": 12.5077,
+      "step": 7287
+    },
+    {
+      "epoch": 1.274535519125683,
+      "grad_norm": 0.8296168096673674,
+      "learning_rate": 0.00012878615229881184,
+      "loss": 12.5477,
+      "step": 7288
+    },
+    {
+      "epoch": 1.2747103825136612,
+      "grad_norm": 0.9524125081681935,
+      "learning_rate": 0.00012876807007744184,
+      "loss": 12.889,
+      "step": 7289
+    },
+    {
+      "epoch": 1.2748852459016393,
+      "grad_norm": 0.9396387987272797,
+      "learning_rate": 0.0001287499868305238,
+      "loss": 12.7999,
+      "step": 7290
+    },
+    {
+      "epoch": 1.2750601092896174,
+      "grad_norm": 0.9160187681004355,
+      "learning_rate": 0.0001287319025587023,
+      "loss": 12.6928,
+      "step": 7291
+    },
+    {
+      "epoch": 1.2752349726775956,
+      "grad_norm": 0.8893442692057532,
+      "learning_rate": 0.00012871381726262202,
+      "loss": 12.6052,
+      "step": 7292
+    },
+    {
+      "epoch": 1.275409836065574,
+      "grad_norm": 0.9407326700361326,
+      "learning_rate": 0.0001286957309429277,
+      "loss": 12.7796,
+      "step": 7293
+    },
+    {
+      "epoch": 1.2755846994535518,
+      "grad_norm": 0.9920543550002341,
+      "learning_rate": 0.00012867764360026412,
+      "loss": 12.6535,
+      "step": 7294
+    },
+    {
+      "epoch": 1.2757595628415301,
+      "grad_norm": 1.016836167957809,
+      "learning_rate": 0.00012865955523527604,
+      "loss": 12.4595,
+      "step": 7295
+    },
+    {
+      "epoch": 1.2759344262295083,
+      "grad_norm": 0.9160493028367158,
+      "learning_rate": 0.0001286414658486083,
+      "loss": 12.5124,
+      "step": 7296
+    },
+    {
+      "epoch": 1.2761092896174864,
+      "grad_norm": 0.8432927866435672,
+      "learning_rate": 0.00012862337544090579,
+      "loss": 12.6965,
+      "step": 7297
+    },
+    {
+      "epoch": 1.2762841530054645,
+      "grad_norm": 0.9460061781986843,
+      "learning_rate": 0.00012860528401281338,
+      "loss": 12.6103,
+      "step": 7298
+    },
+    {
+      "epoch": 1.2764590163934426,
+      "grad_norm": 0.90961222298048,
+      "learning_rate": 0.000128587191564976,
+      "loss": 12.594,
+      "step": 7299
+    },
+    {
+      "epoch": 1.2766338797814207,
+      "grad_norm": 1.0671531876546705,
+      "learning_rate": 0.00012856909809803865,
+      "loss": 12.6433,
+      "step": 7300
+    },
+    {
+      "epoch": 1.2768087431693989,
+      "grad_norm": 0.8547544723872627,
+      "learning_rate": 0.00012855100361264637,
+      "loss": 12.707,
+      "step": 7301
+    },
+    {
+      "epoch": 1.276983606557377,
+      "grad_norm": 1.005035361541161,
+      "learning_rate": 0.00012853290810944408,
+      "loss": 12.7542,
+      "step": 7302
+    },
+    {
+      "epoch": 1.277158469945355,
+      "grad_norm": 0.966385680812926,
+      "learning_rate": 0.00012851481158907704,
+      "loss": 12.6046,
+      "step": 7303
+    },
+    {
+      "epoch": 1.2773333333333334,
+      "grad_norm": 1.002244699169205,
+      "learning_rate": 0.00012849671405219026,
+      "loss": 12.4059,
+      "step": 7304
+    },
+    {
+      "epoch": 1.2775081967213116,
+      "grad_norm": 0.9261412212396707,
+      "learning_rate": 0.0001284786154994289,
+      "loss": 12.8483,
+      "step": 7305
+    },
+    {
+      "epoch": 1.2776830601092897,
+      "grad_norm": 0.9440549129221418,
+      "learning_rate": 0.00012846051593143817,
+      "loss": 12.7102,
+      "step": 7306
+    },
+    {
+      "epoch": 1.2778579234972678,
+      "grad_norm": 0.8824141492507447,
+      "learning_rate": 0.0001284424153488633,
+      "loss": 12.6971,
+      "step": 7307
+    },
+    {
+      "epoch": 1.278032786885246,
+      "grad_norm": 0.8885773325721616,
+      "learning_rate": 0.00012842431375234957,
+      "loss": 12.8372,
+      "step": 7308
+    },
+    {
+      "epoch": 1.278207650273224,
+      "grad_norm": 0.8989591105028164,
+      "learning_rate": 0.00012840621114254224,
+      "loss": 12.666,
+      "step": 7309
+    },
+    {
+      "epoch": 1.2783825136612021,
+      "grad_norm": 0.9338833237468541,
+      "learning_rate": 0.00012838810752008672,
+      "loss": 12.8369,
+      "step": 7310
+    },
+    {
+      "epoch": 1.2785573770491803,
+      "grad_norm": 0.8644277409493065,
+      "learning_rate": 0.0001283700028856283,
+      "loss": 12.6958,
+      "step": 7311
+    },
+    {
+      "epoch": 1.2787322404371584,
+      "grad_norm": 0.9432506542786484,
+      "learning_rate": 0.00012835189723981241,
+      "loss": 12.7528,
+      "step": 7312
+    },
+    {
+      "epoch": 1.2789071038251367,
+      "grad_norm": 0.9143351107534043,
+      "learning_rate": 0.00012833379058328453,
+      "loss": 12.5743,
+      "step": 7313
+    },
+    {
+      "epoch": 1.2790819672131146,
+      "grad_norm": 0.8846508485113439,
+      "learning_rate": 0.00012831568291669015,
+      "loss": 12.4907,
+      "step": 7314
+    },
+    {
+      "epoch": 1.279256830601093,
+      "grad_norm": 0.9448382306891285,
+      "learning_rate": 0.00012829757424067471,
+      "loss": 12.8737,
+      "step": 7315
+    },
+    {
+      "epoch": 1.279431693989071,
+      "grad_norm": 0.9017017674028667,
+      "learning_rate": 0.00012827946455588383,
+      "loss": 12.8892,
+      "step": 7316
+    },
+    {
+      "epoch": 1.2796065573770492,
+      "grad_norm": 0.8797743715477924,
+      "learning_rate": 0.00012826135386296309,
+      "loss": 12.7819,
+      "step": 7317
+    },
+    {
+      "epoch": 1.2797814207650273,
+      "grad_norm": 0.8760730081287241,
+      "learning_rate": 0.00012824324216255807,
+      "loss": 12.8474,
+      "step": 7318
+    },
+    {
+      "epoch": 1.2799562841530054,
+      "grad_norm": 0.9119718054383802,
+      "learning_rate": 0.00012822512945531448,
+      "loss": 12.6396,
+      "step": 7319
+    },
+    {
+      "epoch": 1.2801311475409836,
+      "grad_norm": 1.0045283534855312,
+      "learning_rate": 0.00012820701574187804,
+      "loss": 12.6289,
+      "step": 7320
+    },
+    {
+      "epoch": 1.2803060109289617,
+      "grad_norm": 0.9615841296827089,
+      "learning_rate": 0.00012818890102289437,
+      "loss": 12.7611,
+      "step": 7321
+    },
+    {
+      "epoch": 1.28048087431694,
+      "grad_norm": 0.8876563049947133,
+      "learning_rate": 0.00012817078529900937,
+      "loss": 12.7437,
+      "step": 7322
+    },
+    {
+      "epoch": 1.280655737704918,
+      "grad_norm": 0.9777210572562747,
+      "learning_rate": 0.00012815266857086878,
+      "loss": 12.8527,
+      "step": 7323
+    },
+    {
+      "epoch": 1.2808306010928963,
+      "grad_norm": 0.9176488695361016,
+      "learning_rate": 0.00012813455083911842,
+      "loss": 12.8019,
+      "step": 7324
+    },
+    {
+      "epoch": 1.2810054644808744,
+      "grad_norm": 0.9573510777173637,
+      "learning_rate": 0.00012811643210440424,
+      "loss": 12.5819,
+      "step": 7325
+    },
+    {
+      "epoch": 1.2811803278688525,
+      "grad_norm": 0.9014128656537733,
+      "learning_rate": 0.0001280983123673721,
+      "loss": 12.8371,
+      "step": 7326
+    },
+    {
+      "epoch": 1.2813551912568306,
+      "grad_norm": 0.9613743287792429,
+      "learning_rate": 0.00012808019162866794,
+      "loss": 12.7273,
+      "step": 7327
+    },
+    {
+      "epoch": 1.2815300546448087,
+      "grad_norm": 0.8760655118798772,
+      "learning_rate": 0.00012806206988893773,
+      "loss": 12.5168,
+      "step": 7328
+    },
+    {
+      "epoch": 1.2817049180327869,
+      "grad_norm": 0.9034870808235133,
+      "learning_rate": 0.00012804394714882754,
+      "loss": 12.8694,
+      "step": 7329
+    },
+    {
+      "epoch": 1.281879781420765,
+      "grad_norm": 1.0013352845961887,
+      "learning_rate": 0.0001280258234089834,
+      "loss": 12.606,
+      "step": 7330
+    },
+    {
+      "epoch": 1.282054644808743,
+      "grad_norm": 1.0240637398051444,
+      "learning_rate": 0.0001280076986700514,
+      "loss": 12.441,
+      "step": 7331
+    },
+    {
+      "epoch": 1.2822295081967212,
+      "grad_norm": 0.9469981035781992,
+      "learning_rate": 0.00012798957293267766,
+      "loss": 12.7436,
+      "step": 7332
+    },
+    {
+      "epoch": 1.2824043715846996,
+      "grad_norm": 1.1682110641394157,
+      "learning_rate": 0.00012797144619750836,
+      "loss": 12.7274,
+      "step": 7333
+    },
+    {
+      "epoch": 1.2825792349726777,
+      "grad_norm": 0.9012862573316809,
+      "learning_rate": 0.00012795331846518967,
+      "loss": 12.7293,
+      "step": 7334
+    },
+    {
+      "epoch": 1.2827540983606558,
+      "grad_norm": 1.0754348470421549,
+      "learning_rate": 0.00012793518973636784,
+      "loss": 12.6308,
+      "step": 7335
+    },
+    {
+      "epoch": 1.282928961748634,
+      "grad_norm": 1.0247092514999998,
+      "learning_rate": 0.00012791706001168916,
+      "loss": 12.7072,
+      "step": 7336
+    },
+    {
+      "epoch": 1.283103825136612,
+      "grad_norm": 1.0907565430532669,
+      "learning_rate": 0.00012789892929179988,
+      "loss": 12.604,
+      "step": 7337
+    },
+    {
+      "epoch": 1.2832786885245901,
+      "grad_norm": 0.9366634810746373,
+      "learning_rate": 0.00012788079757734637,
+      "loss": 12.5671,
+      "step": 7338
+    },
+    {
+      "epoch": 1.2834535519125683,
+      "grad_norm": 1.0773906929080201,
+      "learning_rate": 0.000127862664868975,
+      "loss": 12.6471,
+      "step": 7339
+    },
+    {
+      "epoch": 1.2836284153005464,
+      "grad_norm": 0.8964902324491744,
+      "learning_rate": 0.00012784453116733218,
+      "loss": 12.5744,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2838032786885245,
+      "grad_norm": 0.9101045321132113,
+      "learning_rate": 0.00012782639647306436,
+      "loss": 12.8177,
+      "step": 7341
+    },
+    {
+      "epoch": 1.2839781420765028,
+      "grad_norm": 0.995567607989419,
+      "learning_rate": 0.00012780826078681804,
+      "loss": 12.55,
+      "step": 7342
+    },
+    {
+      "epoch": 1.2841530054644807,
+      "grad_norm": 0.8647746430964968,
+      "learning_rate": 0.00012779012410923967,
+      "loss": 12.9343,
+      "step": 7343
+    },
+    {
+      "epoch": 1.284327868852459,
+      "grad_norm": 0.9869196619844803,
+      "learning_rate": 0.0001277719864409759,
+      "loss": 12.9509,
+      "step": 7344
+    },
+    {
+      "epoch": 1.2845027322404372,
+      "grad_norm": 0.9517681536430423,
+      "learning_rate": 0.0001277538477826732,
+      "loss": 12.7992,
+      "step": 7345
+    },
+    {
+      "epoch": 1.2846775956284153,
+      "grad_norm": 0.9130608682640723,
+      "learning_rate": 0.0001277357081349783,
+      "loss": 12.8586,
+      "step": 7346
+    },
+    {
+      "epoch": 1.2848524590163934,
+      "grad_norm": 1.0504855673320257,
+      "learning_rate": 0.0001277175674985378,
+      "loss": 12.8962,
+      "step": 7347
+    },
+    {
+      "epoch": 1.2850273224043716,
+      "grad_norm": 0.9524118270322817,
+      "learning_rate": 0.0001276994258739984,
+      "loss": 12.594,
+      "step": 7348
+    },
+    {
+      "epoch": 1.2852021857923497,
+      "grad_norm": 0.9956114122424041,
+      "learning_rate": 0.0001276812832620068,
+      "loss": 12.361,
+      "step": 7349
+    },
+    {
+      "epoch": 1.2853770491803278,
+      "grad_norm": 0.8578853411263113,
+      "learning_rate": 0.00012766313966320984,
+      "loss": 12.7294,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2855519125683061,
+      "grad_norm": 0.9464625461151871,
+      "learning_rate": 0.00012764499507825428,
+      "loss": 12.6731,
+      "step": 7351
+    },
+    {
+      "epoch": 1.285726775956284,
+      "grad_norm": 0.9025292809736237,
+      "learning_rate": 0.00012762684950778688,
+      "loss": 12.4921,
+      "step": 7352
+    },
+    {
+      "epoch": 1.2859016393442624,
+      "grad_norm": 1.0178437535803981,
+      "learning_rate": 0.00012760870295245463,
+      "loss": 12.9138,
+      "step": 7353
+    },
+    {
+      "epoch": 1.2860765027322405,
+      "grad_norm": 1.0400189672285707,
+      "learning_rate": 0.00012759055541290438,
+      "loss": 12.6482,
+      "step": 7354
+    },
+    {
+      "epoch": 1.2862513661202186,
+      "grad_norm": 0.8593561707493116,
+      "learning_rate": 0.00012757240688978305,
+      "loss": 12.7401,
+      "step": 7355
+    },
+    {
+      "epoch": 1.2864262295081967,
+      "grad_norm": 1.1340521653073719,
+      "learning_rate": 0.0001275542573837376,
+      "loss": 12.6406,
+      "step": 7356
+    },
+    {
+      "epoch": 1.2866010928961749,
+      "grad_norm": 0.9056525553422236,
+      "learning_rate": 0.00012753610689541508,
+      "loss": 12.593,
+      "step": 7357
+    },
+    {
+      "epoch": 1.286775956284153,
+      "grad_norm": 0.9161006108068178,
+      "learning_rate": 0.00012751795542546255,
+      "loss": 12.4948,
+      "step": 7358
+    },
+    {
+      "epoch": 1.286950819672131,
+      "grad_norm": 0.8652973941076951,
+      "learning_rate": 0.00012749980297452703,
+      "loss": 12.4379,
+      "step": 7359
+    },
+    {
+      "epoch": 1.2871256830601092,
+      "grad_norm": 1.069465374506469,
+      "learning_rate": 0.00012748164954325567,
+      "loss": 12.687,
+      "step": 7360
+    },
+    {
+      "epoch": 1.2873005464480873,
+      "grad_norm": 0.9142816726732456,
+      "learning_rate": 0.00012746349513229562,
+      "loss": 12.5287,
+      "step": 7361
+    },
+    {
+      "epoch": 1.2874754098360657,
+      "grad_norm": 0.9522044801600332,
+      "learning_rate": 0.00012744533974229407,
+      "loss": 12.717,
+      "step": 7362
+    },
+    {
+      "epoch": 1.2876502732240438,
+      "grad_norm": 1.0876494570212911,
+      "learning_rate": 0.00012742718337389823,
+      "loss": 12.4769,
+      "step": 7363
+    },
+    {
+      "epoch": 1.287825136612022,
+      "grad_norm": 1.1086157379046,
+      "learning_rate": 0.00012740902602775535,
+      "loss": 13.0353,
+      "step": 7364
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 1.1321117877520996,
+      "learning_rate": 0.00012739086770451266,
+      "loss": 12.6766,
+      "step": 7365
+    },
+    {
+      "epoch": 1.2881748633879782,
+      "grad_norm": 0.9560764740788905,
+      "learning_rate": 0.00012737270840481758,
+      "loss": 12.3924,
+      "step": 7366
+    },
+    {
+      "epoch": 1.2883497267759563,
+      "grad_norm": 1.0974834814186643,
+      "learning_rate": 0.00012735454812931742,
+      "loss": 12.2536,
+      "step": 7367
+    },
+    {
+      "epoch": 1.2885245901639344,
+      "grad_norm": 0.839110986716448,
+      "learning_rate": 0.0001273363868786596,
+      "loss": 12.7597,
+      "step": 7368
+    },
+    {
+      "epoch": 1.2886994535519125,
+      "grad_norm": 1.1472906903682403,
+      "learning_rate": 0.0001273182246534915,
+      "loss": 12.8342,
+      "step": 7369
+    },
+    {
+      "epoch": 1.2888743169398906,
+      "grad_norm": 0.8597271760040094,
+      "learning_rate": 0.0001273000614544606,
+      "loss": 12.8506,
+      "step": 7370
+    },
+    {
+      "epoch": 1.289049180327869,
+      "grad_norm": 1.1697562780056554,
+      "learning_rate": 0.00012728189728221443,
+      "loss": 12.9124,
+      "step": 7371
+    },
+    {
+      "epoch": 1.2892240437158469,
+      "grad_norm": 0.9617155722137722,
+      "learning_rate": 0.0001272637321374005,
+      "loss": 12.5391,
+      "step": 7372
+    },
+    {
+      "epoch": 1.2893989071038252,
+      "grad_norm": 1.162117765023248,
+      "learning_rate": 0.00012724556602066635,
+      "loss": 12.6809,
+      "step": 7373
+    },
+    {
+      "epoch": 1.2895737704918033,
+      "grad_norm": 0.9869014960671592,
+      "learning_rate": 0.00012722739893265963,
+      "loss": 12.7114,
+      "step": 7374
+    },
+    {
+      "epoch": 1.2897486338797814,
+      "grad_norm": 0.9631115340340691,
+      "learning_rate": 0.0001272092308740279,
+      "loss": 12.5427,
+      "step": 7375
+    },
+    {
+      "epoch": 1.2899234972677596,
+      "grad_norm": 0.9825634137452935,
+      "learning_rate": 0.00012719106184541895,
+      "loss": 12.7751,
+      "step": 7376
+    },
+    {
+      "epoch": 1.2900983606557377,
+      "grad_norm": 0.9589090503787476,
+      "learning_rate": 0.00012717289184748038,
+      "loss": 12.9558,
+      "step": 7377
+    },
+    {
+      "epoch": 1.2902732240437158,
+      "grad_norm": 1.055181720008075,
+      "learning_rate": 0.00012715472088085998,
+      "loss": 12.6259,
+      "step": 7378
+    },
+    {
+      "epoch": 1.290448087431694,
+      "grad_norm": 0.9225000210616582,
+      "learning_rate": 0.0001271365489462055,
+      "loss": 12.4137,
+      "step": 7379
+    },
+    {
+      "epoch": 1.2906229508196723,
+      "grad_norm": 0.9904745326192705,
+      "learning_rate": 0.00012711837604416473,
+      "loss": 12.5181,
+      "step": 7380
+    },
+    {
+      "epoch": 1.2907978142076502,
+      "grad_norm": 0.9205804605306508,
+      "learning_rate": 0.0001271002021753856,
+      "loss": 12.4805,
+      "step": 7381
+    },
+    {
+      "epoch": 1.2909726775956285,
+      "grad_norm": 0.9527786432532233,
+      "learning_rate": 0.00012708202734051585,
+      "loss": 12.8039,
+      "step": 7382
+    },
+    {
+      "epoch": 1.2911475409836066,
+      "grad_norm": 0.9257385068963034,
+      "learning_rate": 0.00012706385154020352,
+      "loss": 12.8842,
+      "step": 7383
+    },
+    {
+      "epoch": 1.2913224043715847,
+      "grad_norm": 0.8534557390345304,
+      "learning_rate": 0.00012704567477509645,
+      "loss": 12.6894,
+      "step": 7384
+    },
+    {
+      "epoch": 1.2914972677595629,
+      "grad_norm": 0.9697757721112075,
+      "learning_rate": 0.00012702749704584273,
+      "loss": 12.7672,
+      "step": 7385
+    },
+    {
+      "epoch": 1.291672131147541,
+      "grad_norm": 0.9643360865660973,
+      "learning_rate": 0.00012700931835309028,
+      "loss": 12.8567,
+      "step": 7386
+    },
+    {
+      "epoch": 1.291846994535519,
+      "grad_norm": 0.8870086383860479,
+      "learning_rate": 0.00012699113869748722,
+      "loss": 12.4758,
+      "step": 7387
+    },
+    {
+      "epoch": 1.2920218579234972,
+      "grad_norm": 1.0128503356394167,
+      "learning_rate": 0.00012697295807968158,
+      "loss": 12.7563,
+      "step": 7388
+    },
+    {
+      "epoch": 1.2921967213114753,
+      "grad_norm": 0.8704560399099353,
+      "learning_rate": 0.00012695477650032154,
+      "loss": 12.6096,
+      "step": 7389
+    },
+    {
+      "epoch": 1.2923715846994535,
+      "grad_norm": 1.0045144083602195,
+      "learning_rate": 0.00012693659396005515,
+      "loss": 12.6571,
+      "step": 7390
+    },
+    {
+      "epoch": 1.2925464480874318,
+      "grad_norm": 0.8667956847352619,
+      "learning_rate": 0.0001269184104595307,
+      "loss": 12.5482,
+      "step": 7391
+    },
+    {
+      "epoch": 1.29272131147541,
+      "grad_norm": 0.869830152215468,
+      "learning_rate": 0.00012690022599939633,
+      "loss": 12.6041,
+      "step": 7392
+    },
+    {
+      "epoch": 1.292896174863388,
+      "grad_norm": 0.9594317351766162,
+      "learning_rate": 0.00012688204058030036,
+      "loss": 12.6042,
+      "step": 7393
+    },
+    {
+      "epoch": 1.2930710382513662,
+      "grad_norm": 1.0082407086612393,
+      "learning_rate": 0.00012686385420289105,
+      "loss": 12.4753,
+      "step": 7394
+    },
+    {
+      "epoch": 1.2932459016393443,
+      "grad_norm": 0.8897094273608757,
+      "learning_rate": 0.00012684566686781671,
+      "loss": 12.8569,
+      "step": 7395
+    },
+    {
+      "epoch": 1.2934207650273224,
+      "grad_norm": 0.8630686299099817,
+      "learning_rate": 0.00012682747857572574,
+      "loss": 12.7697,
+      "step": 7396
+    },
+    {
+      "epoch": 1.2935956284153005,
+      "grad_norm": 0.9084877799128454,
+      "learning_rate": 0.0001268092893272665,
+      "loss": 12.6617,
+      "step": 7397
+    },
+    {
+      "epoch": 1.2937704918032786,
+      "grad_norm": 0.930265034662397,
+      "learning_rate": 0.0001267910991230874,
+      "loss": 12.5089,
+      "step": 7398
+    },
+    {
+      "epoch": 1.2939453551912568,
+      "grad_norm": 0.9061163314815036,
+      "learning_rate": 0.00012677290796383689,
+      "loss": 12.9418,
+      "step": 7399
+    },
+    {
+      "epoch": 1.294120218579235,
+      "grad_norm": 0.8558280295627545,
+      "learning_rate": 0.00012675471585016354,
+      "loss": 12.7445,
+      "step": 7400
+    },
+    {
+      "epoch": 1.294295081967213,
+      "grad_norm": 0.9204371953316287,
+      "learning_rate": 0.0001267365227827158,
+      "loss": 12.7334,
+      "step": 7401
+    },
+    {
+      "epoch": 1.2944699453551913,
+      "grad_norm": 0.9775557446026183,
+      "learning_rate": 0.0001267183287621423,
+      "loss": 12.7689,
+      "step": 7402
+    },
+    {
+      "epoch": 1.2946448087431695,
+      "grad_norm": 0.8379469232599047,
+      "learning_rate": 0.00012670013378909153,
+      "loss": 12.6826,
+      "step": 7403
+    },
+    {
+      "epoch": 1.2948196721311476,
+      "grad_norm": 0.9323312177989993,
+      "learning_rate": 0.00012668193786421225,
+      "loss": 12.5156,
+      "step": 7404
+    },
+    {
+      "epoch": 1.2949945355191257,
+      "grad_norm": 0.8967592798292138,
+      "learning_rate": 0.00012666374098815303,
+      "loss": 12.4445,
+      "step": 7405
+    },
+    {
+      "epoch": 1.2951693989071038,
+      "grad_norm": 0.9156932919174103,
+      "learning_rate": 0.00012664554316156262,
+      "loss": 12.7341,
+      "step": 7406
+    },
+    {
+      "epoch": 1.295344262295082,
+      "grad_norm": 0.8831381299893228,
+      "learning_rate": 0.0001266273443850897,
+      "loss": 12.7196,
+      "step": 7407
+    },
+    {
+      "epoch": 1.29551912568306,
+      "grad_norm": 1.034228682727626,
+      "learning_rate": 0.0001266091446593831,
+      "loss": 12.8057,
+      "step": 7408
+    },
+    {
+      "epoch": 1.2956939890710384,
+      "grad_norm": 0.9148089065773326,
+      "learning_rate": 0.00012659094398509156,
+      "loss": 12.8005,
+      "step": 7409
+    },
+    {
+      "epoch": 1.2958688524590163,
+      "grad_norm": 0.8773003365851604,
+      "learning_rate": 0.00012657274236286395,
+      "loss": 12.454,
+      "step": 7410
+    },
+    {
+      "epoch": 1.2960437158469946,
+      "grad_norm": 0.8653250580296039,
+      "learning_rate": 0.0001265545397933491,
+      "loss": 12.7101,
+      "step": 7411
+    },
+    {
+      "epoch": 1.2962185792349727,
+      "grad_norm": 0.9938842156895481,
+      "learning_rate": 0.0001265363362771959,
+      "loss": 12.6116,
+      "step": 7412
+    },
+    {
+      "epoch": 1.2963934426229509,
+      "grad_norm": 0.9758820743290891,
+      "learning_rate": 0.00012651813181505336,
+      "loss": 12.5107,
+      "step": 7413
+    },
+    {
+      "epoch": 1.296568306010929,
+      "grad_norm": 0.8526251484765087,
+      "learning_rate": 0.0001264999264075704,
+      "loss": 12.7216,
+      "step": 7414
+    },
+    {
+      "epoch": 1.296743169398907,
+      "grad_norm": 0.8756685775248594,
+      "learning_rate": 0.00012648172005539604,
+      "loss": 12.7955,
+      "step": 7415
+    },
+    {
+      "epoch": 1.2969180327868852,
+      "grad_norm": 0.8502696254750236,
+      "learning_rate": 0.00012646351275917927,
+      "loss": 12.7673,
+      "step": 7416
+    },
+    {
+      "epoch": 1.2970928961748633,
+      "grad_norm": 0.9878443211110879,
+      "learning_rate": 0.00012644530451956922,
+      "loss": 12.7097,
+      "step": 7417
+    },
+    {
+      "epoch": 1.2972677595628415,
+      "grad_norm": 0.9037821932007248,
+      "learning_rate": 0.00012642709533721493,
+      "loss": 12.6481,
+      "step": 7418
+    },
+    {
+      "epoch": 1.2974426229508196,
+      "grad_norm": 0.8516958702880278,
+      "learning_rate": 0.00012640888521276558,
+      "loss": 12.7769,
+      "step": 7419
+    },
+    {
+      "epoch": 1.297617486338798,
+      "grad_norm": 1.0309473086797154,
+      "learning_rate": 0.00012639067414687031,
+      "loss": 13.0472,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2977923497267758,
+      "grad_norm": 0.9529718958869846,
+      "learning_rate": 0.00012637246214017837,
+      "loss": 12.9243,
+      "step": 7421
+    },
+    {
+      "epoch": 1.2979672131147542,
+      "grad_norm": 0.8038522215993102,
+      "learning_rate": 0.00012635424919333897,
+      "loss": 12.6582,
+      "step": 7422
+    },
+    {
+      "epoch": 1.2981420765027323,
+      "grad_norm": 0.8809106741568995,
+      "learning_rate": 0.0001263360353070014,
+      "loss": 12.5508,
+      "step": 7423
+    },
+    {
+      "epoch": 1.2983169398907104,
+      "grad_norm": 0.8953255135108998,
+      "learning_rate": 0.00012631782048181487,
+      "loss": 12.7534,
+      "step": 7424
+    },
+    {
+      "epoch": 1.2984918032786885,
+      "grad_norm": 0.9021496961883327,
+      "learning_rate": 0.00012629960471842883,
+      "loss": 12.6819,
+      "step": 7425
+    },
+    {
+      "epoch": 1.2986666666666666,
+      "grad_norm": 0.8892945320767184,
+      "learning_rate": 0.00012628138801749263,
+      "loss": 12.619,
+      "step": 7426
+    },
+    {
+      "epoch": 1.2988415300546448,
+      "grad_norm": 0.958884383853952,
+      "learning_rate": 0.00012626317037965564,
+      "loss": 12.6767,
+      "step": 7427
+    },
+    {
+      "epoch": 1.2990163934426229,
+      "grad_norm": 0.8767964515518507,
+      "learning_rate": 0.00012624495180556727,
+      "loss": 12.6518,
+      "step": 7428
+    },
+    {
+      "epoch": 1.2991912568306012,
+      "grad_norm": 0.9351357757452532,
+      "learning_rate": 0.00012622673229587705,
+      "loss": 12.7276,
+      "step": 7429
+    },
+    {
+      "epoch": 1.2993661202185791,
+      "grad_norm": 0.8952722599629866,
+      "learning_rate": 0.0001262085118512345,
+      "loss": 13.1038,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2995409836065575,
+      "grad_norm": 0.8829684732072776,
+      "learning_rate": 0.0001261902904722891,
+      "loss": 12.6586,
+      "step": 7431
+    },
+    {
+      "epoch": 1.2997158469945356,
+      "grad_norm": 0.8467005303026097,
+      "learning_rate": 0.00012617206815969045,
+      "loss": 12.7017,
+      "step": 7432
+    },
+    {
+      "epoch": 1.2998907103825137,
+      "grad_norm": 0.8062867760023634,
+      "learning_rate": 0.00012615384491408815,
+      "loss": 12.4491,
+      "step": 7433
+    },
+    {
+      "epoch": 1.3000655737704918,
+      "grad_norm": 0.8894073003197971,
+      "learning_rate": 0.00012613562073613186,
+      "loss": 12.6934,
+      "step": 7434
+    },
+    {
+      "epoch": 1.30024043715847,
+      "grad_norm": 0.9309880754693564,
+      "learning_rate": 0.00012611739562647117,
+      "loss": 12.618,
+      "step": 7435
+    },
+    {
+      "epoch": 1.300415300546448,
+      "grad_norm": 0.9405541484895552,
+      "learning_rate": 0.0001260991695857559,
+      "loss": 12.6045,
+      "step": 7436
+    },
+    {
+      "epoch": 1.3005901639344262,
+      "grad_norm": 0.9413875974122076,
+      "learning_rate": 0.00012608094261463566,
+      "loss": 12.6338,
+      "step": 7437
+    },
+    {
+      "epoch": 1.3007650273224045,
+      "grad_norm": 0.960508593560316,
+      "learning_rate": 0.00012606271471376034,
+      "loss": 12.7041,
+      "step": 7438
+    },
+    {
+      "epoch": 1.3009398907103824,
+      "grad_norm": 0.8818405984450475,
+      "learning_rate": 0.00012604448588377965,
+      "loss": 12.8025,
+      "step": 7439
+    },
+    {
+      "epoch": 1.3011147540983607,
+      "grad_norm": 0.904933579341933,
+      "learning_rate": 0.00012602625612534354,
+      "loss": 12.8511,
+      "step": 7440
+    },
+    {
+      "epoch": 1.3012896174863389,
+      "grad_norm": 0.8637655251707168,
+      "learning_rate": 0.00012600802543910174,
+      "loss": 12.6249,
+      "step": 7441
+    },
+    {
+      "epoch": 1.301464480874317,
+      "grad_norm": 1.0306076810769043,
+      "learning_rate": 0.00012598979382570427,
+      "loss": 12.7721,
+      "step": 7442
+    },
+    {
+      "epoch": 1.301639344262295,
+      "grad_norm": 0.8732763131206391,
+      "learning_rate": 0.000125971561285801,
+      "loss": 12.4986,
+      "step": 7443
+    },
+    {
+      "epoch": 1.3018142076502732,
+      "grad_norm": 0.9234921607775054,
+      "learning_rate": 0.0001259533278200419,
+      "loss": 12.7616,
+      "step": 7444
+    },
+    {
+      "epoch": 1.3019890710382513,
+      "grad_norm": 0.88442708524128,
+      "learning_rate": 0.00012593509342907698,
+      "loss": 12.5401,
+      "step": 7445
+    },
+    {
+      "epoch": 1.3021639344262295,
+      "grad_norm": 0.9208598791362171,
+      "learning_rate": 0.0001259168581135563,
+      "loss": 12.7792,
+      "step": 7446
+    },
+    {
+      "epoch": 1.3023387978142076,
+      "grad_norm": 0.9052773745335073,
+      "learning_rate": 0.00012589862187412988,
+      "loss": 12.8496,
+      "step": 7447
+    },
+    {
+      "epoch": 1.3025136612021857,
+      "grad_norm": 0.805016038958211,
+      "learning_rate": 0.00012588038471144787,
+      "loss": 12.7189,
+      "step": 7448
+    },
+    {
+      "epoch": 1.302688524590164,
+      "grad_norm": 0.8989754734684526,
+      "learning_rate": 0.0001258621466261604,
+      "loss": 12.6244,
+      "step": 7449
+    },
+    {
+      "epoch": 1.302863387978142,
+      "grad_norm": 0.908922164341543,
+      "learning_rate": 0.0001258439076189176,
+      "loss": 12.8001,
+      "step": 7450
+    },
+    {
+      "epoch": 1.3030382513661203,
+      "grad_norm": 0.9797841129536045,
+      "learning_rate": 0.0001258256676903697,
+      "loss": 12.7811,
+      "step": 7451
+    },
+    {
+      "epoch": 1.3032131147540984,
+      "grad_norm": 0.8589425805883665,
+      "learning_rate": 0.0001258074268411669,
+      "loss": 12.9175,
+      "step": 7452
+    },
+    {
+      "epoch": 1.3033879781420765,
+      "grad_norm": 1.009738133041059,
+      "learning_rate": 0.0001257891850719595,
+      "loss": 12.64,
+      "step": 7453
+    },
+    {
+      "epoch": 1.3035628415300546,
+      "grad_norm": 0.8882157611385488,
+      "learning_rate": 0.00012577094238339778,
+      "loss": 12.7451,
+      "step": 7454
+    },
+    {
+      "epoch": 1.3037377049180328,
+      "grad_norm": 0.881589437355957,
+      "learning_rate": 0.0001257526987761321,
+      "loss": 12.7747,
+      "step": 7455
+    },
+    {
+      "epoch": 1.3039125683060109,
+      "grad_norm": 0.8835135559274517,
+      "learning_rate": 0.00012573445425081276,
+      "loss": 12.759,
+      "step": 7456
+    },
+    {
+      "epoch": 1.304087431693989,
+      "grad_norm": 0.9626800099546876,
+      "learning_rate": 0.0001257162088080902,
+      "loss": 12.7156,
+      "step": 7457
+    },
+    {
+      "epoch": 1.3042622950819673,
+      "grad_norm": 1.0224280659778078,
+      "learning_rate": 0.00012569796244861486,
+      "loss": 12.4497,
+      "step": 7458
+    },
+    {
+      "epoch": 1.3044371584699452,
+      "grad_norm": 0.9463990256354243,
+      "learning_rate": 0.00012567971517303715,
+      "loss": 12.809,
+      "step": 7459
+    },
+    {
+      "epoch": 1.3046120218579236,
+      "grad_norm": 0.9995118429238221,
+      "learning_rate": 0.00012566146698200764,
+      "loss": 12.7716,
+      "step": 7460
+    },
+    {
+      "epoch": 1.3047868852459017,
+      "grad_norm": 0.9237494327932464,
+      "learning_rate": 0.00012564321787617678,
+      "loss": 12.7795,
+      "step": 7461
+    },
+    {
+      "epoch": 1.3049617486338798,
+      "grad_norm": 0.8243422449555694,
+      "learning_rate": 0.0001256249678561952,
+      "loss": 12.7446,
+      "step": 7462
+    },
+    {
+      "epoch": 1.305136612021858,
+      "grad_norm": 1.037138278325413,
+      "learning_rate": 0.00012560671692271345,
+      "loss": 12.5458,
+      "step": 7463
+    },
+    {
+      "epoch": 1.305311475409836,
+      "grad_norm": 0.9059713469350278,
+      "learning_rate": 0.0001255884650763821,
+      "loss": 12.7063,
+      "step": 7464
+    },
+    {
+      "epoch": 1.3054863387978142,
+      "grad_norm": 0.8831590515351154,
+      "learning_rate": 0.00012557021231785194,
+      "loss": 12.832,
+      "step": 7465
+    },
+    {
+      "epoch": 1.3056612021857923,
+      "grad_norm": 1.0065745213738189,
+      "learning_rate": 0.00012555195864777355,
+      "loss": 12.394,
+      "step": 7466
+    },
+    {
+      "epoch": 1.3058360655737704,
+      "grad_norm": 0.8587351656742327,
+      "learning_rate": 0.0001255337040667977,
+      "loss": 12.7712,
+      "step": 7467
+    },
+    {
+      "epoch": 1.3060109289617485,
+      "grad_norm": 0.8542489802987356,
+      "learning_rate": 0.00012551544857557513,
+      "loss": 12.7114,
+      "step": 7468
+    },
+    {
+      "epoch": 1.3061857923497269,
+      "grad_norm": 0.8619074350743203,
+      "learning_rate": 0.00012549719217475663,
+      "loss": 12.5014,
+      "step": 7469
+    },
+    {
+      "epoch": 1.306360655737705,
+      "grad_norm": 1.0140794957969577,
+      "learning_rate": 0.00012547893486499303,
+      "loss": 12.8576,
+      "step": 7470
+    },
+    {
+      "epoch": 1.306535519125683,
+      "grad_norm": 0.9118573107044557,
+      "learning_rate": 0.00012546067664693516,
+      "loss": 12.5841,
+      "step": 7471
+    },
+    {
+      "epoch": 1.3067103825136612,
+      "grad_norm": 0.9880648174314989,
+      "learning_rate": 0.00012544241752123392,
+      "loss": 12.9307,
+      "step": 7472
+    },
+    {
+      "epoch": 1.3068852459016393,
+      "grad_norm": 0.908311454556239,
+      "learning_rate": 0.0001254241574885402,
+      "loss": 12.8124,
+      "step": 7473
+    },
+    {
+      "epoch": 1.3070601092896175,
+      "grad_norm": 0.9651362651098679,
+      "learning_rate": 0.00012540589654950497,
+      "loss": 12.8334,
+      "step": 7474
+    },
+    {
+      "epoch": 1.3072349726775956,
+      "grad_norm": 0.9031615120050126,
+      "learning_rate": 0.00012538763470477926,
+      "loss": 12.7148,
+      "step": 7475
+    },
+    {
+      "epoch": 1.3074098360655737,
+      "grad_norm": 0.8971900230787496,
+      "learning_rate": 0.00012536937195501397,
+      "loss": 12.7549,
+      "step": 7476
+    },
+    {
+      "epoch": 1.3075846994535518,
+      "grad_norm": 0.8732160952473621,
+      "learning_rate": 0.00012535110830086028,
+      "loss": 12.6435,
+      "step": 7477
+    },
+    {
+      "epoch": 1.3077595628415302,
+      "grad_norm": 0.9115620831702768,
+      "learning_rate": 0.00012533284374296915,
+      "loss": 12.7577,
+      "step": 7478
+    },
+    {
+      "epoch": 1.307934426229508,
+      "grad_norm": 1.0843806577499138,
+      "learning_rate": 0.00012531457828199175,
+      "loss": 12.6102,
+      "step": 7479
+    },
+    {
+      "epoch": 1.3081092896174864,
+      "grad_norm": 0.8409515015199379,
+      "learning_rate": 0.0001252963119185792,
+      "loss": 12.689,
+      "step": 7480
+    },
+    {
+      "epoch": 1.3082841530054645,
+      "grad_norm": 0.8363694042973903,
+      "learning_rate": 0.00012527804465338273,
+      "loss": 12.564,
+      "step": 7481
+    },
+    {
+      "epoch": 1.3084590163934426,
+      "grad_norm": 0.8746734958935993,
+      "learning_rate": 0.00012525977648705344,
+      "loss": 12.7471,
+      "step": 7482
+    },
+    {
+      "epoch": 1.3086338797814208,
+      "grad_norm": 0.90089136734978,
+      "learning_rate": 0.00012524150742024265,
+      "loss": 12.6409,
+      "step": 7483
+    },
+    {
+      "epoch": 1.3088087431693989,
+      "grad_norm": 0.995295890887571,
+      "learning_rate": 0.00012522323745360164,
+      "loss": 12.8532,
+      "step": 7484
+    },
+    {
+      "epoch": 1.308983606557377,
+      "grad_norm": 0.8506886558592721,
+      "learning_rate": 0.00012520496658778167,
+      "loss": 12.4413,
+      "step": 7485
+    },
+    {
+      "epoch": 1.3091584699453551,
+      "grad_norm": 0.928363725636955,
+      "learning_rate": 0.0001251866948234341,
+      "loss": 12.7153,
+      "step": 7486
+    },
+    {
+      "epoch": 1.3093333333333335,
+      "grad_norm": 0.9510472398320924,
+      "learning_rate": 0.00012516842216121028,
+      "loss": 12.7035,
+      "step": 7487
+    },
+    {
+      "epoch": 1.3095081967213114,
+      "grad_norm": 1.043144436021238,
+      "learning_rate": 0.0001251501486017616,
+      "loss": 12.6369,
+      "step": 7488
+    },
+    {
+      "epoch": 1.3096830601092897,
+      "grad_norm": 0.9915797195480698,
+      "learning_rate": 0.00012513187414573952,
+      "loss": 12.7135,
+      "step": 7489
+    },
+    {
+      "epoch": 1.3098579234972678,
+      "grad_norm": 1.0829721982569824,
+      "learning_rate": 0.00012511359879379552,
+      "loss": 12.6474,
+      "step": 7490
+    },
+    {
+      "epoch": 1.310032786885246,
+      "grad_norm": 0.9336136751750157,
+      "learning_rate": 0.000125095322546581,
+      "loss": 12.6079,
+      "step": 7491
+    },
+    {
+      "epoch": 1.310207650273224,
+      "grad_norm": 0.9677927641759908,
+      "learning_rate": 0.00012507704540474758,
+      "loss": 12.7239,
+      "step": 7492
+    },
+    {
+      "epoch": 1.3103825136612022,
+      "grad_norm": 1.1705446713270835,
+      "learning_rate": 0.00012505876736894678,
+      "loss": 12.7419,
+      "step": 7493
+    },
+    {
+      "epoch": 1.3105573770491803,
+      "grad_norm": 0.9027769352419879,
+      "learning_rate": 0.00012504048843983026,
+      "loss": 12.6003,
+      "step": 7494
+    },
+    {
+      "epoch": 1.3107322404371584,
+      "grad_norm": 0.8731007857347591,
+      "learning_rate": 0.0001250222086180495,
+      "loss": 12.7866,
+      "step": 7495
+    },
+    {
+      "epoch": 1.3109071038251365,
+      "grad_norm": 1.0192764037914548,
+      "learning_rate": 0.0001250039279042563,
+      "loss": 12.7469,
+      "step": 7496
+    },
+    {
+      "epoch": 1.3110819672131147,
+      "grad_norm": 0.9333772948763206,
+      "learning_rate": 0.00012498564629910227,
+      "loss": 12.3814,
+      "step": 7497
+    },
+    {
+      "epoch": 1.311256830601093,
+      "grad_norm": 1.1259562690487923,
+      "learning_rate": 0.00012496736380323913,
+      "loss": 12.701,
+      "step": 7498
+    },
+    {
+      "epoch": 1.3114316939890711,
+      "grad_norm": 0.8531860177340805,
+      "learning_rate": 0.00012494908041731865,
+      "loss": 12.5669,
+      "step": 7499
+    },
+    {
+      "epoch": 1.3116065573770492,
+      "grad_norm": 1.15028128995983,
+      "learning_rate": 0.00012493079614199262,
+      "loss": 12.6784,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3117814207650274,
+      "grad_norm": 1.048835602689186,
+      "learning_rate": 0.0001249125109779128,
+      "loss": 12.4966,
+      "step": 7501
+    },
+    {
+      "epoch": 1.3119562841530055,
+      "grad_norm": 1.0594989226212148,
+      "learning_rate": 0.00012489422492573108,
+      "loss": 12.657,
+      "step": 7502
+    },
+    {
+      "epoch": 1.3121311475409836,
+      "grad_norm": 0.8918479721707675,
+      "learning_rate": 0.00012487593798609934,
+      "loss": 12.6821,
+      "step": 7503
+    },
+    {
+      "epoch": 1.3123060109289617,
+      "grad_norm": 1.2135871458263223,
+      "learning_rate": 0.00012485765015966948,
+      "loss": 12.6514,
+      "step": 7504
+    },
+    {
+      "epoch": 1.3124808743169398,
+      "grad_norm": 1.056784348492542,
+      "learning_rate": 0.00012483936144709344,
+      "loss": 12.7196,
+      "step": 7505
+    },
+    {
+      "epoch": 1.312655737704918,
+      "grad_norm": 1.037667463182284,
+      "learning_rate": 0.00012482107184902315,
+      "loss": 12.4129,
+      "step": 7506
+    },
+    {
+      "epoch": 1.3128306010928963,
+      "grad_norm": 0.9208801097779403,
+      "learning_rate": 0.0001248027813661107,
+      "loss": 12.5269,
+      "step": 7507
+    },
+    {
+      "epoch": 1.3130054644808742,
+      "grad_norm": 1.0930191546103234,
+      "learning_rate": 0.00012478448999900802,
+      "loss": 12.6772,
+      "step": 7508
+    },
+    {
+      "epoch": 1.3131803278688525,
+      "grad_norm": 0.9439795186281558,
+      "learning_rate": 0.00012476619774836727,
+      "loss": 12.5547,
+      "step": 7509
+    },
+    {
+      "epoch": 1.3133551912568306,
+      "grad_norm": 0.9244052243205925,
+      "learning_rate": 0.00012474790461484048,
+      "loss": 12.5791,
+      "step": 7510
+    },
+    {
+      "epoch": 1.3135300546448088,
+      "grad_norm": 0.9717898694431847,
+      "learning_rate": 0.0001247296105990798,
+      "loss": 12.717,
+      "step": 7511
+    },
+    {
+      "epoch": 1.3137049180327869,
+      "grad_norm": 0.8344503181110615,
+      "learning_rate": 0.0001247113157017374,
+      "loss": 12.569,
+      "step": 7512
+    },
+    {
+      "epoch": 1.313879781420765,
+      "grad_norm": 0.9429873386143692,
+      "learning_rate": 0.00012469301992346549,
+      "loss": 12.6244,
+      "step": 7513
+    },
+    {
+      "epoch": 1.3140546448087431,
+      "grad_norm": 0.9020750917513125,
+      "learning_rate": 0.00012467472326491624,
+      "loss": 12.6494,
+      "step": 7514
+    },
+    {
+      "epoch": 1.3142295081967212,
+      "grad_norm": 0.8745913856017722,
+      "learning_rate": 0.00012465642572674195,
+      "loss": 12.8206,
+      "step": 7515
+    },
+    {
+      "epoch": 1.3144043715846996,
+      "grad_norm": 0.890607386719025,
+      "learning_rate": 0.0001246381273095949,
+      "loss": 12.6892,
+      "step": 7516
+    },
+    {
+      "epoch": 1.3145792349726775,
+      "grad_norm": 0.8963637786788858,
+      "learning_rate": 0.00012461982801412742,
+      "loss": 12.3316,
+      "step": 7517
+    },
+    {
+      "epoch": 1.3147540983606558,
+      "grad_norm": 1.0173774459028684,
+      "learning_rate": 0.0001246015278409918,
+      "loss": 12.6095,
+      "step": 7518
+    },
+    {
+      "epoch": 1.314928961748634,
+      "grad_norm": 0.8859820959062121,
+      "learning_rate": 0.00012458322679084044,
+      "loss": 12.8227,
+      "step": 7519
+    },
+    {
+      "epoch": 1.315103825136612,
+      "grad_norm": 1.1080139981938617,
+      "learning_rate": 0.0001245649248643258,
+      "loss": 12.8501,
+      "step": 7520
+    },
+    {
+      "epoch": 1.3152786885245902,
+      "grad_norm": 0.8179394596122821,
+      "learning_rate": 0.00012454662206210027,
+      "loss": 12.5468,
+      "step": 7521
+    },
+    {
+      "epoch": 1.3154535519125683,
+      "grad_norm": 1.0589685487772804,
+      "learning_rate": 0.00012452831838481637,
+      "loss": 12.7617,
+      "step": 7522
+    },
+    {
+      "epoch": 1.3156284153005464,
+      "grad_norm": 0.8750310029880283,
+      "learning_rate": 0.00012451001383312655,
+      "loss": 12.7936,
+      "step": 7523
+    },
+    {
+      "epoch": 1.3158032786885245,
+      "grad_norm": 1.008617480683122,
+      "learning_rate": 0.0001244917084076834,
+      "loss": 12.6381,
+      "step": 7524
+    },
+    {
+      "epoch": 1.3159781420765027,
+      "grad_norm": 0.8839430586071638,
+      "learning_rate": 0.00012447340210913944,
+      "loss": 12.5345,
+      "step": 7525
+    },
+    {
+      "epoch": 1.3161530054644808,
+      "grad_norm": 0.9073678294001182,
+      "learning_rate": 0.0001244550949381473,
+      "loss": 12.3062,
+      "step": 7526
+    },
+    {
+      "epoch": 1.3163278688524591,
+      "grad_norm": 0.8836793876432184,
+      "learning_rate": 0.0001244367868953596,
+      "loss": 12.7333,
+      "step": 7527
+    },
+    {
+      "epoch": 1.3165027322404372,
+      "grad_norm": 0.8960429520681795,
+      "learning_rate": 0.000124418477981429,
+      "loss": 12.7105,
+      "step": 7528
+    },
+    {
+      "epoch": 1.3166775956284154,
+      "grad_norm": 1.059044979509697,
+      "learning_rate": 0.00012440016819700815,
+      "loss": 12.423,
+      "step": 7529
+    },
+    {
+      "epoch": 1.3168524590163935,
+      "grad_norm": 0.8909811459713691,
+      "learning_rate": 0.00012438185754274988,
+      "loss": 12.4628,
+      "step": 7530
+    },
+    {
+      "epoch": 1.3170273224043716,
+      "grad_norm": 1.0138443247593014,
+      "learning_rate": 0.0001243635460193068,
+      "loss": 12.5655,
+      "step": 7531
+    },
+    {
+      "epoch": 1.3172021857923497,
+      "grad_norm": 0.9988989029306041,
+      "learning_rate": 0.00012434523362733183,
+      "loss": 12.809,
+      "step": 7532
+    },
+    {
+      "epoch": 1.3173770491803278,
+      "grad_norm": 1.0036427066254592,
+      "learning_rate": 0.0001243269203674777,
+      "loss": 12.7775,
+      "step": 7533
+    },
+    {
+      "epoch": 1.317551912568306,
+      "grad_norm": 0.879445580707031,
+      "learning_rate": 0.0001243086062403973,
+      "loss": 12.7247,
+      "step": 7534
+    },
+    {
+      "epoch": 1.317726775956284,
+      "grad_norm": 1.0815838065253318,
+      "learning_rate": 0.00012429029124674348,
+      "loss": 12.697,
+      "step": 7535
+    },
+    {
+      "epoch": 1.3179016393442624,
+      "grad_norm": 0.9224182925283757,
+      "learning_rate": 0.00012427197538716913,
+      "loss": 12.544,
+      "step": 7536
+    },
+    {
+      "epoch": 1.3180765027322403,
+      "grad_norm": 1.1648542909517594,
+      "learning_rate": 0.00012425365866232727,
+      "loss": 12.85,
+      "step": 7537
+    },
+    {
+      "epoch": 1.3182513661202186,
+      "grad_norm": 0.9268021851907212,
+      "learning_rate": 0.00012423534107287076,
+      "loss": 12.6956,
+      "step": 7538
+    },
+    {
+      "epoch": 1.3184262295081968,
+      "grad_norm": 1.4031268879288066,
+      "learning_rate": 0.0001242170226194527,
+      "loss": 12.9141,
+      "step": 7539
+    },
+    {
+      "epoch": 1.3186010928961749,
+      "grad_norm": 0.9406972308137136,
+      "learning_rate": 0.00012419870330272604,
+      "loss": 12.6601,
+      "step": 7540
+    },
+    {
+      "epoch": 1.318775956284153,
+      "grad_norm": 1.1646315093203998,
+      "learning_rate": 0.00012418038312334392,
+      "loss": 12.7429,
+      "step": 7541
+    },
+    {
+      "epoch": 1.3189508196721311,
+      "grad_norm": 0.9798871172176068,
+      "learning_rate": 0.00012416206208195937,
+      "loss": 12.7917,
+      "step": 7542
+    },
+    {
+      "epoch": 1.3191256830601092,
+      "grad_norm": 0.9317848054636018,
+      "learning_rate": 0.00012414374017922554,
+      "loss": 12.8319,
+      "step": 7543
+    },
+    {
+      "epoch": 1.3193005464480874,
+      "grad_norm": 1.057451962124646,
+      "learning_rate": 0.00012412541741579558,
+      "loss": 12.6673,
+      "step": 7544
+    },
+    {
+      "epoch": 1.3194754098360657,
+      "grad_norm": 0.8622094507862313,
+      "learning_rate": 0.00012410709379232271,
+      "loss": 12.7773,
+      "step": 7545
+    },
+    {
+      "epoch": 1.3196502732240436,
+      "grad_norm": 1.1046063390125547,
+      "learning_rate": 0.00012408876930946008,
+      "loss": 12.8696,
+      "step": 7546
+    },
+    {
+      "epoch": 1.319825136612022,
+      "grad_norm": 0.8137710409199472,
+      "learning_rate": 0.00012407044396786095,
+      "loss": 12.5671,
+      "step": 7547
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9043288791655439,
+      "learning_rate": 0.00012405211776817863,
+      "loss": 12.7606,
+      "step": 7548
+    },
+    {
+      "epoch": 1.3201748633879782,
+      "grad_norm": 1.0131689424397818,
+      "learning_rate": 0.00012403379071106643,
+      "loss": 12.7233,
+      "step": 7549
+    },
+    {
+      "epoch": 1.3203497267759563,
+      "grad_norm": 0.9816258751839482,
+      "learning_rate": 0.00012401546279717764,
+      "loss": 12.5583,
+      "step": 7550
+    },
+    {
+      "epoch": 1.3205245901639344,
+      "grad_norm": 1.0392538231317279,
+      "learning_rate": 0.0001239971340271657,
+      "loss": 12.7949,
+      "step": 7551
+    },
+    {
+      "epoch": 1.3206994535519125,
+      "grad_norm": 1.044914526960209,
+      "learning_rate": 0.00012397880440168397,
+      "loss": 12.6141,
+      "step": 7552
+    },
+    {
+      "epoch": 1.3208743169398907,
+      "grad_norm": 1.162255502565,
+      "learning_rate": 0.00012396047392138586,
+      "loss": 12.7679,
+      "step": 7553
+    },
+    {
+      "epoch": 1.3210491803278688,
+      "grad_norm": 0.913099643907041,
+      "learning_rate": 0.00012394214258692484,
+      "loss": 12.5685,
+      "step": 7554
+    },
+    {
+      "epoch": 1.321224043715847,
+      "grad_norm": 1.040917629691687,
+      "learning_rate": 0.00012392381039895442,
+      "loss": 12.4794,
+      "step": 7555
+    },
+    {
+      "epoch": 1.3213989071038252,
+      "grad_norm": 0.8812092541555466,
+      "learning_rate": 0.00012390547735812812,
+      "loss": 12.417,
+      "step": 7556
+    },
+    {
+      "epoch": 1.3215737704918034,
+      "grad_norm": 0.9805003488373776,
+      "learning_rate": 0.00012388714346509943,
+      "loss": 12.6505,
+      "step": 7557
+    },
+    {
+      "epoch": 1.3217486338797815,
+      "grad_norm": 0.9258014663046308,
+      "learning_rate": 0.00012386880872052207,
+      "loss": 12.486,
+      "step": 7558
+    },
+    {
+      "epoch": 1.3219234972677596,
+      "grad_norm": 0.9645294424802325,
+      "learning_rate": 0.0001238504731250495,
+      "loss": 12.4567,
+      "step": 7559
+    },
+    {
+      "epoch": 1.3220983606557377,
+      "grad_norm": 1.0379669277138197,
+      "learning_rate": 0.00012383213667933548,
+      "loss": 12.4502,
+      "step": 7560
+    },
+    {
+      "epoch": 1.3222732240437158,
+      "grad_norm": 0.8531817848360264,
+      "learning_rate": 0.0001238137993840336,
+      "loss": 12.4207,
+      "step": 7561
+    },
+    {
+      "epoch": 1.322448087431694,
+      "grad_norm": 1.3263191181236753,
+      "learning_rate": 0.0001237954612397976,
+      "loss": 12.6549,
+      "step": 7562
+    },
+    {
+      "epoch": 1.322622950819672,
+      "grad_norm": 0.8379391969634253,
+      "learning_rate": 0.00012377712224728122,
+      "loss": 12.5568,
+      "step": 7563
+    },
+    {
+      "epoch": 1.3227978142076502,
+      "grad_norm": 1.2019687258583152,
+      "learning_rate": 0.00012375878240713817,
+      "loss": 12.9784,
+      "step": 7564
+    },
+    {
+      "epoch": 1.3229726775956285,
+      "grad_norm": 0.8816091163593924,
+      "learning_rate": 0.00012374044172002235,
+      "loss": 12.381,
+      "step": 7565
+    },
+    {
+      "epoch": 1.3231475409836064,
+      "grad_norm": 1.0503607280607132,
+      "learning_rate": 0.0001237221001865875,
+      "loss": 12.7037,
+      "step": 7566
+    },
+    {
+      "epoch": 1.3233224043715848,
+      "grad_norm": 0.928211636389298,
+      "learning_rate": 0.0001237037578074875,
+      "loss": 12.7277,
+      "step": 7567
+    },
+    {
+      "epoch": 1.323497267759563,
+      "grad_norm": 0.9258556854791553,
+      "learning_rate": 0.00012368541458337618,
+      "loss": 12.6151,
+      "step": 7568
+    },
+    {
+      "epoch": 1.323672131147541,
+      "grad_norm": 0.9960760058711274,
+      "learning_rate": 0.00012366707051490757,
+      "loss": 12.7773,
+      "step": 7569
+    },
+    {
+      "epoch": 1.3238469945355191,
+      "grad_norm": 0.9215601282974597,
+      "learning_rate": 0.0001236487256027355,
+      "loss": 12.358,
+      "step": 7570
+    },
+    {
+      "epoch": 1.3240218579234972,
+      "grad_norm": 0.9201328464349152,
+      "learning_rate": 0.00012363037984751402,
+      "loss": 12.702,
+      "step": 7571
+    },
+    {
+      "epoch": 1.3241967213114754,
+      "grad_norm": 0.8277629523828129,
+      "learning_rate": 0.0001236120332498971,
+      "loss": 12.6702,
+      "step": 7572
+    },
+    {
+      "epoch": 1.3243715846994535,
+      "grad_norm": 1.0368500750739225,
+      "learning_rate": 0.00012359368581053875,
+      "loss": 12.5051,
+      "step": 7573
+    },
+    {
+      "epoch": 1.3245464480874318,
+      "grad_norm": 0.9592903627125127,
+      "learning_rate": 0.0001235753375300931,
+      "loss": 12.8172,
+      "step": 7574
+    },
+    {
+      "epoch": 1.3247213114754097,
+      "grad_norm": 1.0573419436192344,
+      "learning_rate": 0.0001235569884092142,
+      "loss": 12.8253,
+      "step": 7575
+    },
+    {
+      "epoch": 1.324896174863388,
+      "grad_norm": 0.8910968088939851,
+      "learning_rate": 0.00012353863844855623,
+      "loss": 12.5328,
+      "step": 7576
+    },
+    {
+      "epoch": 1.3250710382513662,
+      "grad_norm": 0.9835227374032318,
+      "learning_rate": 0.00012352028764877325,
+      "loss": 12.7056,
+      "step": 7577
+    },
+    {
+      "epoch": 1.3252459016393443,
+      "grad_norm": 0.9329564424431268,
+      "learning_rate": 0.0001235019360105195,
+      "loss": 12.6893,
+      "step": 7578
+    },
+    {
+      "epoch": 1.3254207650273224,
+      "grad_norm": 0.895828697605026,
+      "learning_rate": 0.00012348358353444922,
+      "loss": 12.6605,
+      "step": 7579
+    },
+    {
+      "epoch": 1.3255956284153005,
+      "grad_norm": 0.9866549343496374,
+      "learning_rate": 0.00012346523022121661,
+      "loss": 12.4402,
+      "step": 7580
+    },
+    {
+      "epoch": 1.3257704918032787,
+      "grad_norm": 0.9291039584699675,
+      "learning_rate": 0.00012344687607147595,
+      "loss": 12.7594,
+      "step": 7581
+    },
+    {
+      "epoch": 1.3259453551912568,
+      "grad_norm": 1.0549658131980133,
+      "learning_rate": 0.00012342852108588155,
+      "loss": 12.7427,
+      "step": 7582
+    },
+    {
+      "epoch": 1.326120218579235,
+      "grad_norm": 1.0286201303312659,
+      "learning_rate": 0.00012341016526508776,
+      "loss": 12.8253,
+      "step": 7583
+    },
+    {
+      "epoch": 1.326295081967213,
+      "grad_norm": 1.052605488181669,
+      "learning_rate": 0.00012339180860974891,
+      "loss": 12.7165,
+      "step": 7584
+    },
+    {
+      "epoch": 1.3264699453551914,
+      "grad_norm": 1.0677190618607857,
+      "learning_rate": 0.00012337345112051947,
+      "loss": 12.5794,
+      "step": 7585
+    },
+    {
+      "epoch": 1.3266448087431695,
+      "grad_norm": 0.8851642749774572,
+      "learning_rate": 0.00012335509279805376,
+      "loss": 12.605,
+      "step": 7586
+    },
+    {
+      "epoch": 1.3268196721311476,
+      "grad_norm": 0.9746943702073416,
+      "learning_rate": 0.00012333673364300628,
+      "loss": 12.6272,
+      "step": 7587
+    },
+    {
+      "epoch": 1.3269945355191257,
+      "grad_norm": 0.8737927865214103,
+      "learning_rate": 0.00012331837365603152,
+      "loss": 12.7749,
+      "step": 7588
+    },
+    {
+      "epoch": 1.3271693989071038,
+      "grad_norm": 1.3376215097695257,
+      "learning_rate": 0.00012330001283778397,
+      "loss": 12.3738,
+      "step": 7589
+    },
+    {
+      "epoch": 1.327344262295082,
+      "grad_norm": 0.8349814097419166,
+      "learning_rate": 0.0001232816511889182,
+      "loss": 12.4378,
+      "step": 7590
+    },
+    {
+      "epoch": 1.32751912568306,
+      "grad_norm": 1.0255091890666859,
+      "learning_rate": 0.00012326328871008877,
+      "loss": 12.6493,
+      "step": 7591
+    },
+    {
+      "epoch": 1.3276939890710382,
+      "grad_norm": 0.9614040488646675,
+      "learning_rate": 0.00012324492540195025,
+      "loss": 12.7698,
+      "step": 7592
+    },
+    {
+      "epoch": 1.3278688524590163,
+      "grad_norm": 0.9140379721668453,
+      "learning_rate": 0.0001232265612651573,
+      "loss": 12.3876,
+      "step": 7593
+    },
+    {
+      "epoch": 1.3280437158469947,
+      "grad_norm": 0.9055934229349352,
+      "learning_rate": 0.0001232081963003646,
+      "loss": 12.5847,
+      "step": 7594
+    },
+    {
+      "epoch": 1.3282185792349726,
+      "grad_norm": 0.8970966494708524,
+      "learning_rate": 0.00012318983050822684,
+      "loss": 12.6568,
+      "step": 7595
+    },
+    {
+      "epoch": 1.328393442622951,
+      "grad_norm": 1.0276853224243823,
+      "learning_rate": 0.0001231714638893987,
+      "loss": 12.6937,
+      "step": 7596
+    },
+    {
+      "epoch": 1.328568306010929,
+      "grad_norm": 0.8886179826089685,
+      "learning_rate": 0.00012315309644453493,
+      "loss": 12.5494,
+      "step": 7597
+    },
+    {
+      "epoch": 1.3287431693989071,
+      "grad_norm": 0.9868805986857655,
+      "learning_rate": 0.00012313472817429032,
+      "loss": 12.7929,
+      "step": 7598
+    },
+    {
+      "epoch": 1.3289180327868853,
+      "grad_norm": 0.9049412370421427,
+      "learning_rate": 0.0001231163590793197,
+      "loss": 12.7899,
+      "step": 7599
+    },
+    {
+      "epoch": 1.3290928961748634,
+      "grad_norm": 1.0393173060304044,
+      "learning_rate": 0.00012309798916027787,
+      "loss": 12.6644,
+      "step": 7600
+    },
+    {
+      "epoch": 1.3292677595628415,
+      "grad_norm": 0.9040814960553488,
+      "learning_rate": 0.00012307961841781973,
+      "loss": 12.6156,
+      "step": 7601
+    },
+    {
+      "epoch": 1.3294426229508196,
+      "grad_norm": 0.9041042083881761,
+      "learning_rate": 0.00012306124685260015,
+      "loss": 12.5528,
+      "step": 7602
+    },
+    {
+      "epoch": 1.329617486338798,
+      "grad_norm": 0.8853742106982211,
+      "learning_rate": 0.00012304287446527408,
+      "loss": 12.7864,
+      "step": 7603
+    },
+    {
+      "epoch": 1.3297923497267758,
+      "grad_norm": 0.9069882379508433,
+      "learning_rate": 0.0001230245012564964,
+      "loss": 12.7249,
+      "step": 7604
+    },
+    {
+      "epoch": 1.3299672131147542,
+      "grad_norm": 0.8784371846464542,
+      "learning_rate": 0.00012300612722692222,
+      "loss": 12.6181,
+      "step": 7605
+    },
+    {
+      "epoch": 1.3301420765027323,
+      "grad_norm": 1.0459841635834188,
+      "learning_rate": 0.00012298775237720646,
+      "loss": 12.7404,
+      "step": 7606
+    },
+    {
+      "epoch": 1.3303169398907104,
+      "grad_norm": 0.9727325398839504,
+      "learning_rate": 0.00012296937670800418,
+      "loss": 12.8129,
+      "step": 7607
+    },
+    {
+      "epoch": 1.3304918032786885,
+      "grad_norm": 0.9107104280640168,
+      "learning_rate": 0.00012295100021997047,
+      "loss": 12.4528,
+      "step": 7608
+    },
+    {
+      "epoch": 1.3306666666666667,
+      "grad_norm": 0.7971257553224294,
+      "learning_rate": 0.00012293262291376036,
+      "loss": 12.7203,
+      "step": 7609
+    },
+    {
+      "epoch": 1.3308415300546448,
+      "grad_norm": 1.041541933231847,
+      "learning_rate": 0.0001229142447900291,
+      "loss": 12.6237,
+      "step": 7610
+    },
+    {
+      "epoch": 1.331016393442623,
+      "grad_norm": 0.9749629416540315,
+      "learning_rate": 0.00012289586584943177,
+      "loss": 12.5244,
+      "step": 7611
+    },
+    {
+      "epoch": 1.331191256830601,
+      "grad_norm": 0.9664386755507637,
+      "learning_rate": 0.00012287748609262356,
+      "loss": 12.6182,
+      "step": 7612
+    },
+    {
+      "epoch": 1.3313661202185791,
+      "grad_norm": 1.0837038941288515,
+      "learning_rate": 0.00012285910552025975,
+      "loss": 12.6653,
+      "step": 7613
+    },
+    {
+      "epoch": 1.3315409836065575,
+      "grad_norm": 0.916936804901147,
+      "learning_rate": 0.0001228407241329955,
+      "loss": 12.7484,
+      "step": 7614
+    },
+    {
+      "epoch": 1.3317158469945356,
+      "grad_norm": 1.1066226707750926,
+      "learning_rate": 0.0001228223419314861,
+      "loss": 12.4514,
+      "step": 7615
+    },
+    {
+      "epoch": 1.3318907103825137,
+      "grad_norm": 0.856526615857491,
+      "learning_rate": 0.00012280395891638694,
+      "loss": 12.6703,
+      "step": 7616
+    },
+    {
+      "epoch": 1.3320655737704918,
+      "grad_norm": 0.9350926861796075,
+      "learning_rate": 0.00012278557508835322,
+      "loss": 12.4199,
+      "step": 7617
+    },
+    {
+      "epoch": 1.33224043715847,
+      "grad_norm": 0.9285413909151308,
+      "learning_rate": 0.0001227671904480404,
+      "loss": 12.6351,
+      "step": 7618
+    },
+    {
+      "epoch": 1.332415300546448,
+      "grad_norm": 0.9484254437336883,
+      "learning_rate": 0.00012274880499610382,
+      "loss": 12.5004,
+      "step": 7619
+    },
+    {
+      "epoch": 1.3325901639344262,
+      "grad_norm": 0.9904899306748566,
+      "learning_rate": 0.00012273041873319895,
+      "loss": 12.4448,
+      "step": 7620
+    },
+    {
+      "epoch": 1.3327650273224043,
+      "grad_norm": 0.9773904491953352,
+      "learning_rate": 0.0001227120316599812,
+      "loss": 12.5402,
+      "step": 7621
+    },
+    {
+      "epoch": 1.3329398907103824,
+      "grad_norm": 1.0234471203313564,
+      "learning_rate": 0.00012269364377710606,
+      "loss": 12.5823,
+      "step": 7622
+    },
+    {
+      "epoch": 1.3331147540983608,
+      "grad_norm": 0.9012043393653772,
+      "learning_rate": 0.00012267525508522902,
+      "loss": 12.7253,
+      "step": 7623
+    },
+    {
+      "epoch": 1.3332896174863387,
+      "grad_norm": 0.9998796945177708,
+      "learning_rate": 0.00012265686558500565,
+      "loss": 12.6605,
+      "step": 7624
+    },
+    {
+      "epoch": 1.333464480874317,
+      "grad_norm": 0.8794886199293482,
+      "learning_rate": 0.00012263847527709147,
+      "loss": 12.8312,
+      "step": 7625
+    },
+    {
+      "epoch": 1.3336393442622951,
+      "grad_norm": 0.9091873011008116,
+      "learning_rate": 0.00012262008416214211,
+      "loss": 12.483,
+      "step": 7626
+    },
+    {
+      "epoch": 1.3338142076502733,
+      "grad_norm": 0.8740294084260081,
+      "learning_rate": 0.0001226016922408132,
+      "loss": 12.7175,
+      "step": 7627
+    },
+    {
+      "epoch": 1.3339890710382514,
+      "grad_norm": 0.9118397033010246,
+      "learning_rate": 0.00012258329951376034,
+      "loss": 12.5265,
+      "step": 7628
+    },
+    {
+      "epoch": 1.3341639344262295,
+      "grad_norm": 0.8339779042598247,
+      "learning_rate": 0.00012256490598163924,
+      "loss": 12.6186,
+      "step": 7629
+    },
+    {
+      "epoch": 1.3343387978142076,
+      "grad_norm": 0.9025455273200661,
+      "learning_rate": 0.0001225465116451056,
+      "loss": 12.6417,
+      "step": 7630
+    },
+    {
+      "epoch": 1.3345136612021857,
+      "grad_norm": 0.91856898136822,
+      "learning_rate": 0.00012252811650481518,
+      "loss": 12.6129,
+      "step": 7631
+    },
+    {
+      "epoch": 1.334688524590164,
+      "grad_norm": 0.8552693304452379,
+      "learning_rate": 0.0001225097205614237,
+      "loss": 12.4477,
+      "step": 7632
+    },
+    {
+      "epoch": 1.334863387978142,
+      "grad_norm": 0.8576679649709076,
+      "learning_rate": 0.00012249132381558705,
+      "loss": 12.4593,
+      "step": 7633
+    },
+    {
+      "epoch": 1.3350382513661203,
+      "grad_norm": 0.8386565115363278,
+      "learning_rate": 0.00012247292626796092,
+      "loss": 12.6277,
+      "step": 7634
+    },
+    {
+      "epoch": 1.3352131147540984,
+      "grad_norm": 0.8945355706843519,
+      "learning_rate": 0.00012245452791920125,
+      "loss": 12.6742,
+      "step": 7635
+    },
+    {
+      "epoch": 1.3353879781420765,
+      "grad_norm": 0.8118725330228259,
+      "learning_rate": 0.00012243612876996387,
+      "loss": 12.3814,
+      "step": 7636
+    },
+    {
+      "epoch": 1.3355628415300547,
+      "grad_norm": 0.8865259279239474,
+      "learning_rate": 0.0001224177288209047,
+      "loss": 12.375,
+      "step": 7637
+    },
+    {
+      "epoch": 1.3357377049180328,
+      "grad_norm": 0.889031065245286,
+      "learning_rate": 0.00012239932807267973,
+      "loss": 12.7573,
+      "step": 7638
+    },
+    {
+      "epoch": 1.335912568306011,
+      "grad_norm": 0.9353265562165927,
+      "learning_rate": 0.0001223809265259449,
+      "loss": 12.5049,
+      "step": 7639
+    },
+    {
+      "epoch": 1.336087431693989,
+      "grad_norm": 0.9636573078690902,
+      "learning_rate": 0.00012236252418135614,
+      "loss": 12.5288,
+      "step": 7640
+    },
+    {
+      "epoch": 1.3362622950819671,
+      "grad_norm": 0.9704934046470483,
+      "learning_rate": 0.00012234412103956957,
+      "loss": 12.625,
+      "step": 7641
+    },
+    {
+      "epoch": 1.3364371584699453,
+      "grad_norm": 0.9041846248808965,
+      "learning_rate": 0.00012232571710124113,
+      "loss": 12.6094,
+      "step": 7642
+    },
+    {
+      "epoch": 1.3366120218579236,
+      "grad_norm": 1.0019683602035392,
+      "learning_rate": 0.000122307312367027,
+      "loss": 12.8057,
+      "step": 7643
+    },
+    {
+      "epoch": 1.3367868852459015,
+      "grad_norm": 0.8314010090609112,
+      "learning_rate": 0.00012228890683758326,
+      "loss": 12.6977,
+      "step": 7644
+    },
+    {
+      "epoch": 1.3369617486338798,
+      "grad_norm": 0.9543433542413339,
+      "learning_rate": 0.00012227050051356604,
+      "loss": 12.2319,
+      "step": 7645
+    },
+    {
+      "epoch": 1.337136612021858,
+      "grad_norm": 0.8495550331447579,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 12.6864,
+      "step": 7646
+    },
+    {
+      "epoch": 1.337311475409836,
+      "grad_norm": 1.245529899750992,
+      "learning_rate": 0.00012223368548443575,
+      "loss": 12.3512,
+      "step": 7647
+    },
+    {
+      "epoch": 1.3374863387978142,
+      "grad_norm": 0.9135974967080202,
+      "learning_rate": 0.00012221527678063515,
+      "loss": 12.4463,
+      "step": 7648
+    },
+    {
+      "epoch": 1.3376612021857923,
+      "grad_norm": 1.046034533091607,
+      "learning_rate": 0.00012219686728488587,
+      "loss": 12.7019,
+      "step": 7649
+    },
+    {
+      "epoch": 1.3378360655737704,
+      "grad_norm": 0.9437126256486696,
+      "learning_rate": 0.00012217845699784424,
+      "loss": 12.764,
+      "step": 7650
+    },
+    {
+      "epoch": 1.3380109289617486,
+      "grad_norm": 0.9843739862726263,
+      "learning_rate": 0.00012216004592016647,
+      "loss": 12.6972,
+      "step": 7651
+    },
+    {
+      "epoch": 1.338185792349727,
+      "grad_norm": 1.0606720699707182,
+      "learning_rate": 0.00012214163405250902,
+      "loss": 12.5347,
+      "step": 7652
+    },
+    {
+      "epoch": 1.3383606557377048,
+      "grad_norm": 0.890131447463472,
+      "learning_rate": 0.00012212322139552815,
+      "loss": 12.7508,
+      "step": 7653
+    },
+    {
+      "epoch": 1.3385355191256831,
+      "grad_norm": 1.3823504957186352,
+      "learning_rate": 0.0001221048079498803,
+      "loss": 12.5231,
+      "step": 7654
+    },
+    {
+      "epoch": 1.3387103825136613,
+      "grad_norm": 0.8637849634844073,
+      "learning_rate": 0.00012208639371622185,
+      "loss": 12.8312,
+      "step": 7655
+    },
+    {
+      "epoch": 1.3388852459016394,
+      "grad_norm": 1.0956446772397719,
+      "learning_rate": 0.00012206797869520929,
+      "loss": 12.7699,
+      "step": 7656
+    },
+    {
+      "epoch": 1.3390601092896175,
+      "grad_norm": 0.9587316077404023,
+      "learning_rate": 0.00012204956288749905,
+      "loss": 12.4712,
+      "step": 7657
+    },
+    {
+      "epoch": 1.3392349726775956,
+      "grad_norm": 1.1628184821936538,
+      "learning_rate": 0.0001220311462937477,
+      "loss": 12.6894,
+      "step": 7658
+    },
+    {
+      "epoch": 1.3394098360655737,
+      "grad_norm": 0.7962949590051458,
+      "learning_rate": 0.00012201272891461169,
+      "loss": 12.4529,
+      "step": 7659
+    },
+    {
+      "epoch": 1.3395846994535519,
+      "grad_norm": 1.0573227310106867,
+      "learning_rate": 0.00012199431075074763,
+      "loss": 12.8741,
+      "step": 7660
+    },
+    {
+      "epoch": 1.33975956284153,
+      "grad_norm": 0.8747703208570907,
+      "learning_rate": 0.00012197589180281208,
+      "loss": 12.7981,
+      "step": 7661
+    },
+    {
+      "epoch": 1.339934426229508,
+      "grad_norm": 0.8371233804660783,
+      "learning_rate": 0.00012195747207146166,
+      "loss": 12.4415,
+      "step": 7662
+    },
+    {
+      "epoch": 1.3401092896174864,
+      "grad_norm": 0.8914901292952766,
+      "learning_rate": 0.00012193905155735302,
+      "loss": 12.5022,
+      "step": 7663
+    },
+    {
+      "epoch": 1.3402841530054646,
+      "grad_norm": 0.8459279614650566,
+      "learning_rate": 0.00012192063026114282,
+      "loss": 12.6288,
+      "step": 7664
+    },
+    {
+      "epoch": 1.3404590163934427,
+      "grad_norm": 0.871456503648344,
+      "learning_rate": 0.00012190220818348777,
+      "loss": 12.5644,
+      "step": 7665
+    },
+    {
+      "epoch": 1.3406338797814208,
+      "grad_norm": 0.8233701386280605,
+      "learning_rate": 0.00012188378532504456,
+      "loss": 12.6001,
+      "step": 7666
+    },
+    {
+      "epoch": 1.340808743169399,
+      "grad_norm": 0.8866411404440134,
+      "learning_rate": 0.00012186536168647003,
+      "loss": 12.4371,
+      "step": 7667
+    },
+    {
+      "epoch": 1.340983606557377,
+      "grad_norm": 0.9983983402793032,
+      "learning_rate": 0.00012184693726842085,
+      "loss": 12.6384,
+      "step": 7668
+    },
+    {
+      "epoch": 1.3411584699453551,
+      "grad_norm": 0.9300139404631219,
+      "learning_rate": 0.00012182851207155393,
+      "loss": 12.6595,
+      "step": 7669
+    },
+    {
+      "epoch": 1.3413333333333333,
+      "grad_norm": 0.9206163599478269,
+      "learning_rate": 0.00012181008609652603,
+      "loss": 12.6839,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3415081967213114,
+      "grad_norm": 0.988545506337875,
+      "learning_rate": 0.00012179165934399403,
+      "loss": 12.5646,
+      "step": 7671
+    },
+    {
+      "epoch": 1.3416830601092897,
+      "grad_norm": 0.8289822309775691,
+      "learning_rate": 0.00012177323181461484,
+      "loss": 12.5158,
+      "step": 7672
+    },
+    {
+      "epoch": 1.3418579234972676,
+      "grad_norm": 0.8940493869108551,
+      "learning_rate": 0.00012175480350904538,
+      "loss": 12.7608,
+      "step": 7673
+    },
+    {
+      "epoch": 1.342032786885246,
+      "grad_norm": 0.8136258835166864,
+      "learning_rate": 0.00012173637442794259,
+      "loss": 12.4901,
+      "step": 7674
+    },
+    {
+      "epoch": 1.342207650273224,
+      "grad_norm": 0.8957043454320405,
+      "learning_rate": 0.00012171794457196347,
+      "loss": 12.7205,
+      "step": 7675
+    },
+    {
+      "epoch": 1.3423825136612022,
+      "grad_norm": 0.9131924617513347,
+      "learning_rate": 0.000121699513941765,
+      "loss": 12.7489,
+      "step": 7676
+    },
+    {
+      "epoch": 1.3425573770491803,
+      "grad_norm": 0.997285505606792,
+      "learning_rate": 0.00012168108253800418,
+      "loss": 12.6065,
+      "step": 7677
+    },
+    {
+      "epoch": 1.3427322404371584,
+      "grad_norm": 0.820548679229875,
+      "learning_rate": 0.00012166265036133812,
+      "loss": 12.7266,
+      "step": 7678
+    },
+    {
+      "epoch": 1.3429071038251366,
+      "grad_norm": 0.9319627315916242,
+      "learning_rate": 0.00012164421741242388,
+      "loss": 12.6227,
+      "step": 7679
+    },
+    {
+      "epoch": 1.3430819672131147,
+      "grad_norm": 0.9106570609328712,
+      "learning_rate": 0.0001216257836919186,
+      "loss": 12.5819,
+      "step": 7680
+    },
+    {
+      "epoch": 1.343256830601093,
+      "grad_norm": 0.9369481888930027,
+      "learning_rate": 0.00012160734920047935,
+      "loss": 12.485,
+      "step": 7681
+    },
+    {
+      "epoch": 1.343431693989071,
+      "grad_norm": 0.9625066074444513,
+      "learning_rate": 0.00012158891393876338,
+      "loss": 12.5841,
+      "step": 7682
+    },
+    {
+      "epoch": 1.3436065573770493,
+      "grad_norm": 0.9924010092147063,
+      "learning_rate": 0.00012157047790742781,
+      "loss": 12.6202,
+      "step": 7683
+    },
+    {
+      "epoch": 1.3437814207650274,
+      "grad_norm": 0.9951411648504125,
+      "learning_rate": 0.00012155204110712997,
+      "loss": 12.8119,
+      "step": 7684
+    },
+    {
+      "epoch": 1.3439562841530055,
+      "grad_norm": 0.9317857490133182,
+      "learning_rate": 0.000121533603538527,
+      "loss": 12.4208,
+      "step": 7685
+    },
+    {
+      "epoch": 1.3441311475409836,
+      "grad_norm": 0.9711538258932574,
+      "learning_rate": 0.00012151516520227626,
+      "loss": 12.7294,
+      "step": 7686
+    },
+    {
+      "epoch": 1.3443060109289617,
+      "grad_norm": 1.035327721953838,
+      "learning_rate": 0.00012149672609903497,
+      "loss": 12.5542,
+      "step": 7687
+    },
+    {
+      "epoch": 1.3444808743169399,
+      "grad_norm": 0.9773368494007096,
+      "learning_rate": 0.00012147828622946056,
+      "loss": 12.8032,
+      "step": 7688
+    },
+    {
+      "epoch": 1.344655737704918,
+      "grad_norm": 0.9120878630404244,
+      "learning_rate": 0.00012145984559421031,
+      "loss": 12.4674,
+      "step": 7689
+    },
+    {
+      "epoch": 1.344830601092896,
+      "grad_norm": 1.176985871099031,
+      "learning_rate": 0.00012144140419394165,
+      "loss": 12.9191,
+      "step": 7690
+    },
+    {
+      "epoch": 1.3450054644808742,
+      "grad_norm": 0.9988126576798323,
+      "learning_rate": 0.00012142296202931198,
+      "loss": 12.8025,
+      "step": 7691
+    },
+    {
+      "epoch": 1.3451803278688526,
+      "grad_norm": 1.128797775097035,
+      "learning_rate": 0.00012140451910097872,
+      "loss": 12.3842,
+      "step": 7692
+    },
+    {
+      "epoch": 1.3453551912568307,
+      "grad_norm": 0.8712111582446104,
+      "learning_rate": 0.00012138607540959942,
+      "loss": 12.695,
+      "step": 7693
+    },
+    {
+      "epoch": 1.3455300546448088,
+      "grad_norm": 1.2628325497764001,
+      "learning_rate": 0.00012136763095583145,
+      "loss": 12.8658,
+      "step": 7694
+    },
+    {
+      "epoch": 1.345704918032787,
+      "grad_norm": 0.8573305170255978,
+      "learning_rate": 0.00012134918574033245,
+      "loss": 12.5886,
+      "step": 7695
+    },
+    {
+      "epoch": 1.345879781420765,
+      "grad_norm": 1.330519031119997,
+      "learning_rate": 0.0001213307397637599,
+      "loss": 12.76,
+      "step": 7696
+    },
+    {
+      "epoch": 1.3460546448087432,
+      "grad_norm": 0.9486183011970546,
+      "learning_rate": 0.00012131229302677143,
+      "loss": 12.5906,
+      "step": 7697
+    },
+    {
+      "epoch": 1.3462295081967213,
+      "grad_norm": 1.084374567906662,
+      "learning_rate": 0.00012129384553002457,
+      "loss": 12.5189,
+      "step": 7698
+    },
+    {
+      "epoch": 1.3464043715846994,
+      "grad_norm": 0.9436074204189772,
+      "learning_rate": 0.00012127539727417702,
+      "loss": 12.3913,
+      "step": 7699
+    },
+    {
+      "epoch": 1.3465792349726775,
+      "grad_norm": 1.0348070964870246,
+      "learning_rate": 0.0001212569482598864,
+      "loss": 12.5885,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3467540983606558,
+      "grad_norm": 0.9022933225594578,
+      "learning_rate": 0.00012123849848781042,
+      "loss": 12.7248,
+      "step": 7701
+    },
+    {
+      "epoch": 1.3469289617486337,
+      "grad_norm": 0.9363330778012222,
+      "learning_rate": 0.00012122004795860678,
+      "loss": 12.6746,
+      "step": 7702
+    },
+    {
+      "epoch": 1.347103825136612,
+      "grad_norm": 1.1564631082753023,
+      "learning_rate": 0.00012120159667293325,
+      "loss": 12.7915,
+      "step": 7703
+    },
+    {
+      "epoch": 1.3472786885245902,
+      "grad_norm": 1.0057513080105784,
+      "learning_rate": 0.00012118314463144753,
+      "loss": 12.5998,
+      "step": 7704
+    },
+    {
+      "epoch": 1.3474535519125683,
+      "grad_norm": 1.0272502519174338,
+      "learning_rate": 0.0001211646918348075,
+      "loss": 12.6578,
+      "step": 7705
+    },
+    {
+      "epoch": 1.3476284153005464,
+      "grad_norm": 0.8484429540640807,
+      "learning_rate": 0.00012114623828367089,
+      "loss": 12.8235,
+      "step": 7706
+    },
+    {
+      "epoch": 1.3478032786885246,
+      "grad_norm": 0.9125419829094858,
+      "learning_rate": 0.00012112778397869562,
+      "loss": 12.6478,
+      "step": 7707
+    },
+    {
+      "epoch": 1.3479781420765027,
+      "grad_norm": 0.8824349279945002,
+      "learning_rate": 0.00012110932892053952,
+      "loss": 12.7882,
+      "step": 7708
+    },
+    {
+      "epoch": 1.3481530054644808,
+      "grad_norm": 0.9197427208191605,
+      "learning_rate": 0.00012109087310986049,
+      "loss": 12.6734,
+      "step": 7709
+    },
+    {
+      "epoch": 1.3483278688524591,
+      "grad_norm": 0.9021794190183039,
+      "learning_rate": 0.00012107241654731652,
+      "loss": 12.561,
+      "step": 7710
+    },
+    {
+      "epoch": 1.348502732240437,
+      "grad_norm": 0.8619805243583376,
+      "learning_rate": 0.00012105395923356548,
+      "loss": 12.2361,
+      "step": 7711
+    },
+    {
+      "epoch": 1.3486775956284154,
+      "grad_norm": 1.0021816771727,
+      "learning_rate": 0.00012103550116926543,
+      "loss": 12.6505,
+      "step": 7712
+    },
+    {
+      "epoch": 1.3488524590163935,
+      "grad_norm": 0.8732308550084493,
+      "learning_rate": 0.00012101704235507432,
+      "loss": 12.6976,
+      "step": 7713
+    },
+    {
+      "epoch": 1.3490273224043716,
+      "grad_norm": 0.8728360934357123,
+      "learning_rate": 0.00012099858279165022,
+      "loss": 12.8562,
+      "step": 7714
+    },
+    {
+      "epoch": 1.3492021857923497,
+      "grad_norm": 0.8830775143261113,
+      "learning_rate": 0.00012098012247965113,
+      "loss": 12.5783,
+      "step": 7715
+    },
+    {
+      "epoch": 1.3493770491803279,
+      "grad_norm": 0.9434112870629697,
+      "learning_rate": 0.00012096166141973526,
+      "loss": 12.7398,
+      "step": 7716
+    },
+    {
+      "epoch": 1.349551912568306,
+      "grad_norm": 0.9574946007267329,
+      "learning_rate": 0.0001209431996125606,
+      "loss": 12.5551,
+      "step": 7717
+    },
+    {
+      "epoch": 1.349726775956284,
+      "grad_norm": 1.1000546504415858,
+      "learning_rate": 0.00012092473705878533,
+      "loss": 12.9254,
+      "step": 7718
+    },
+    {
+      "epoch": 1.3499016393442622,
+      "grad_norm": 0.8642040840582633,
+      "learning_rate": 0.00012090627375906767,
+      "loss": 12.7717,
+      "step": 7719
+    },
+    {
+      "epoch": 1.3500765027322403,
+      "grad_norm": 0.8100703366337746,
+      "learning_rate": 0.00012088780971406577,
+      "loss": 12.5134,
+      "step": 7720
+    },
+    {
+      "epoch": 1.3502513661202187,
+      "grad_norm": 0.9218470759427396,
+      "learning_rate": 0.00012086934492443785,
+      "loss": 12.6532,
+      "step": 7721
+    },
+    {
+      "epoch": 1.3504262295081968,
+      "grad_norm": 0.815631343923478,
+      "learning_rate": 0.00012085087939084219,
+      "loss": 12.5894,
+      "step": 7722
+    },
+    {
+      "epoch": 1.350601092896175,
+      "grad_norm": 0.9767744529207633,
+      "learning_rate": 0.00012083241311393703,
+      "loss": 12.794,
+      "step": 7723
+    },
+    {
+      "epoch": 1.350775956284153,
+      "grad_norm": 0.9105012288421024,
+      "learning_rate": 0.00012081394609438066,
+      "loss": 12.6682,
+      "step": 7724
+    },
+    {
+      "epoch": 1.3509508196721312,
+      "grad_norm": 0.8998528258410792,
+      "learning_rate": 0.00012079547833283144,
+      "loss": 12.3793,
+      "step": 7725
+    },
+    {
+      "epoch": 1.3511256830601093,
+      "grad_norm": 0.9056160792663196,
+      "learning_rate": 0.00012077700982994773,
+      "loss": 12.5486,
+      "step": 7726
+    },
+    {
+      "epoch": 1.3513005464480874,
+      "grad_norm": 1.0502096840494337,
+      "learning_rate": 0.00012075854058638787,
+      "loss": 12.9456,
+      "step": 7727
+    },
+    {
+      "epoch": 1.3514754098360655,
+      "grad_norm": 0.8919581420630961,
+      "learning_rate": 0.0001207400706028103,
+      "loss": 12.7092,
+      "step": 7728
+    },
+    {
+      "epoch": 1.3516502732240436,
+      "grad_norm": 0.9719569833930369,
+      "learning_rate": 0.00012072159987987346,
+      "loss": 12.6972,
+      "step": 7729
+    },
+    {
+      "epoch": 1.351825136612022,
+      "grad_norm": 0.8685707575886957,
+      "learning_rate": 0.00012070312841823576,
+      "loss": 11.998,
+      "step": 7730
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 0.8956580480587271,
+      "learning_rate": 0.00012068465621855578,
+      "loss": 12.4884,
+      "step": 7731
+    },
+    {
+      "epoch": 1.3521748633879782,
+      "grad_norm": 0.8252507638271495,
+      "learning_rate": 0.00012066618328149193,
+      "loss": 12.6409,
+      "step": 7732
+    },
+    {
+      "epoch": 1.3523497267759563,
+      "grad_norm": 0.9245423683014288,
+      "learning_rate": 0.0001206477096077028,
+      "loss": 12.7214,
+      "step": 7733
+    },
+    {
+      "epoch": 1.3525245901639344,
+      "grad_norm": 0.8670064940003833,
+      "learning_rate": 0.00012062923519784694,
+      "loss": 12.4679,
+      "step": 7734
+    },
+    {
+      "epoch": 1.3526994535519126,
+      "grad_norm": 0.8188924996236454,
+      "learning_rate": 0.00012061076005258296,
+      "loss": 12.4029,
+      "step": 7735
+    },
+    {
+      "epoch": 1.3528743169398907,
+      "grad_norm": 0.8810225533712441,
+      "learning_rate": 0.00012059228417256947,
+      "loss": 12.6359,
+      "step": 7736
+    },
+    {
+      "epoch": 1.3530491803278688,
+      "grad_norm": 0.8433012735344511,
+      "learning_rate": 0.00012057380755846509,
+      "loss": 12.7553,
+      "step": 7737
+    },
+    {
+      "epoch": 1.353224043715847,
+      "grad_norm": 0.9562134066961638,
+      "learning_rate": 0.00012055533021092853,
+      "loss": 12.6086,
+      "step": 7738
+    },
+    {
+      "epoch": 1.3533989071038253,
+      "grad_norm": 0.8090709567325499,
+      "learning_rate": 0.00012053685213061843,
+      "loss": 12.3495,
+      "step": 7739
+    },
+    {
+      "epoch": 1.3535737704918032,
+      "grad_norm": 0.9412448805893343,
+      "learning_rate": 0.0001205183733181936,
+      "loss": 12.6715,
+      "step": 7740
+    },
+    {
+      "epoch": 1.3537486338797815,
+      "grad_norm": 1.008951322807329,
+      "learning_rate": 0.00012049989377431268,
+      "loss": 12.7238,
+      "step": 7741
+    },
+    {
+      "epoch": 1.3539234972677596,
+      "grad_norm": 0.9260050439353585,
+      "learning_rate": 0.00012048141349963453,
+      "loss": 12.7024,
+      "step": 7742
+    },
+    {
+      "epoch": 1.3540983606557377,
+      "grad_norm": 0.8978664230746478,
+      "learning_rate": 0.00012046293249481791,
+      "loss": 12.6822,
+      "step": 7743
+    },
+    {
+      "epoch": 1.3542732240437159,
+      "grad_norm": 0.8665097264325973,
+      "learning_rate": 0.00012044445076052167,
+      "loss": 12.635,
+      "step": 7744
+    },
+    {
+      "epoch": 1.354448087431694,
+      "grad_norm": 0.9035435641250393,
+      "learning_rate": 0.00012042596829740462,
+      "loss": 12.5661,
+      "step": 7745
+    },
+    {
+      "epoch": 1.354622950819672,
+      "grad_norm": 0.8461278006002175,
+      "learning_rate": 0.00012040748510612567,
+      "loss": 12.5505,
+      "step": 7746
+    },
+    {
+      "epoch": 1.3547978142076502,
+      "grad_norm": 0.85969996767854,
+      "learning_rate": 0.00012038900118734371,
+      "loss": 12.721,
+      "step": 7747
+    },
+    {
+      "epoch": 1.3549726775956283,
+      "grad_norm": 0.8539833739352206,
+      "learning_rate": 0.00012037051654171771,
+      "loss": 12.5612,
+      "step": 7748
+    },
+    {
+      "epoch": 1.3551475409836065,
+      "grad_norm": 0.8388202965747236,
+      "learning_rate": 0.00012035203116990658,
+      "loss": 12.6811,
+      "step": 7749
+    },
+    {
+      "epoch": 1.3553224043715848,
+      "grad_norm": 0.8682367013603537,
+      "learning_rate": 0.00012033354507256933,
+      "loss": 12.6299,
+      "step": 7750
+    },
+    {
+      "epoch": 1.355497267759563,
+      "grad_norm": 0.912624572263596,
+      "learning_rate": 0.00012031505825036493,
+      "loss": 12.9217,
+      "step": 7751
+    },
+    {
+      "epoch": 1.355672131147541,
+      "grad_norm": 0.8262194735271031,
+      "learning_rate": 0.00012029657070395248,
+      "loss": 12.5238,
+      "step": 7752
+    },
+    {
+      "epoch": 1.3558469945355192,
+      "grad_norm": 0.8645357239770095,
+      "learning_rate": 0.00012027808243399096,
+      "loss": 12.7491,
+      "step": 7753
+    },
+    {
+      "epoch": 1.3560218579234973,
+      "grad_norm": 0.8537406629895447,
+      "learning_rate": 0.0001202595934411395,
+      "loss": 12.4792,
+      "step": 7754
+    },
+    {
+      "epoch": 1.3561967213114754,
+      "grad_norm": 0.8671858857150034,
+      "learning_rate": 0.00012024110372605722,
+      "loss": 12.6555,
+      "step": 7755
+    },
+    {
+      "epoch": 1.3563715846994535,
+      "grad_norm": 0.8260993577054578,
+      "learning_rate": 0.00012022261328940323,
+      "loss": 12.6514,
+      "step": 7756
+    },
+    {
+      "epoch": 1.3565464480874316,
+      "grad_norm": 0.8848307618421489,
+      "learning_rate": 0.00012020412213183671,
+      "loss": 12.7544,
+      "step": 7757
+    },
+    {
+      "epoch": 1.3567213114754098,
+      "grad_norm": 1.0045059372469998,
+      "learning_rate": 0.00012018563025401684,
+      "loss": 12.7384,
+      "step": 7758
+    },
+    {
+      "epoch": 1.356896174863388,
+      "grad_norm": 0.8786522303694287,
+      "learning_rate": 0.00012016713765660284,
+      "loss": 12.6759,
+      "step": 7759
+    },
+    {
+      "epoch": 1.357071038251366,
+      "grad_norm": 0.9698631334432638,
+      "learning_rate": 0.00012014864434025392,
+      "loss": 12.6962,
+      "step": 7760
+    },
+    {
+      "epoch": 1.3572459016393443,
+      "grad_norm": 0.9212766868737665,
+      "learning_rate": 0.00012013015030562942,
+      "loss": 12.3748,
+      "step": 7761
+    },
+    {
+      "epoch": 1.3574207650273225,
+      "grad_norm": 0.8356372923576197,
+      "learning_rate": 0.00012011165555338855,
+      "loss": 12.6745,
+      "step": 7762
+    },
+    {
+      "epoch": 1.3575956284153006,
+      "grad_norm": 0.9174607037801241,
+      "learning_rate": 0.00012009316008419063,
+      "loss": 12.5089,
+      "step": 7763
+    },
+    {
+      "epoch": 1.3577704918032787,
+      "grad_norm": 0.8597744995908848,
+      "learning_rate": 0.00012007466389869505,
+      "loss": 12.4191,
+      "step": 7764
+    },
+    {
+      "epoch": 1.3579453551912568,
+      "grad_norm": 1.0395153774211867,
+      "learning_rate": 0.0001200561669975612,
+      "loss": 12.6771,
+      "step": 7765
+    },
+    {
+      "epoch": 1.358120218579235,
+      "grad_norm": 1.0396669467752633,
+      "learning_rate": 0.00012003766938144836,
+      "loss": 12.7332,
+      "step": 7766
+    },
+    {
+      "epoch": 1.358295081967213,
+      "grad_norm": 0.8712325312146028,
+      "learning_rate": 0.00012001917105101607,
+      "loss": 12.6888,
+      "step": 7767
+    },
+    {
+      "epoch": 1.3584699453551914,
+      "grad_norm": 0.9299811611775006,
+      "learning_rate": 0.00012000067200692369,
+      "loss": 12.7511,
+      "step": 7768
+    },
+    {
+      "epoch": 1.3586448087431693,
+      "grad_norm": 0.9724499539427498,
+      "learning_rate": 0.00011998217224983075,
+      "loss": 12.7461,
+      "step": 7769
+    },
+    {
+      "epoch": 1.3588196721311476,
+      "grad_norm": 0.9205640475544216,
+      "learning_rate": 0.00011996367178039671,
+      "loss": 12.818,
+      "step": 7770
+    },
+    {
+      "epoch": 1.3589945355191257,
+      "grad_norm": 1.0574992365055247,
+      "learning_rate": 0.00011994517059928107,
+      "loss": 12.728,
+      "step": 7771
+    },
+    {
+      "epoch": 1.3591693989071039,
+      "grad_norm": 0.9265451005351516,
+      "learning_rate": 0.00011992666870714341,
+      "loss": 12.8123,
+      "step": 7772
+    },
+    {
+      "epoch": 1.359344262295082,
+      "grad_norm": 1.0861145671299648,
+      "learning_rate": 0.00011990816610464331,
+      "loss": 12.5592,
+      "step": 7773
+    },
+    {
+      "epoch": 1.35951912568306,
+      "grad_norm": 0.8499054697044062,
+      "learning_rate": 0.00011988966279244036,
+      "loss": 12.5378,
+      "step": 7774
+    },
+    {
+      "epoch": 1.3596939890710382,
+      "grad_norm": 1.0590806005812914,
+      "learning_rate": 0.00011987115877119413,
+      "loss": 12.8062,
+      "step": 7775
+    },
+    {
+      "epoch": 1.3598688524590163,
+      "grad_norm": 1.0764032917778037,
+      "learning_rate": 0.00011985265404156433,
+      "loss": 12.6635,
+      "step": 7776
+    },
+    {
+      "epoch": 1.3600437158469945,
+      "grad_norm": 0.9119501193632505,
+      "learning_rate": 0.00011983414860421061,
+      "loss": 12.5646,
+      "step": 7777
+    },
+    {
+      "epoch": 1.3602185792349726,
+      "grad_norm": 1.1534778469895366,
+      "learning_rate": 0.00011981564245979266,
+      "loss": 12.6468,
+      "step": 7778
+    },
+    {
+      "epoch": 1.360393442622951,
+      "grad_norm": 0.9253041795539357,
+      "learning_rate": 0.00011979713560897021,
+      "loss": 12.7498,
+      "step": 7779
+    },
+    {
+      "epoch": 1.360568306010929,
+      "grad_norm": 0.9327518620757932,
+      "learning_rate": 0.00011977862805240303,
+      "loss": 12.4705,
+      "step": 7780
+    },
+    {
+      "epoch": 1.3607431693989072,
+      "grad_norm": 0.8269504881294485,
+      "learning_rate": 0.00011976011979075084,
+      "loss": 12.8029,
+      "step": 7781
+    },
+    {
+      "epoch": 1.3609180327868853,
+      "grad_norm": 0.8448073821365879,
+      "learning_rate": 0.00011974161082467345,
+      "loss": 12.4386,
+      "step": 7782
+    },
+    {
+      "epoch": 1.3610928961748634,
+      "grad_norm": 0.9790268173690487,
+      "learning_rate": 0.00011972310115483072,
+      "loss": 12.5794,
+      "step": 7783
+    },
+    {
+      "epoch": 1.3612677595628415,
+      "grad_norm": 0.9844042754135798,
+      "learning_rate": 0.00011970459078188247,
+      "loss": 12.7075,
+      "step": 7784
+    },
+    {
+      "epoch": 1.3614426229508196,
+      "grad_norm": 0.8648345088013241,
+      "learning_rate": 0.00011968607970648862,
+      "loss": 12.5686,
+      "step": 7785
+    },
+    {
+      "epoch": 1.3616174863387978,
+      "grad_norm": 0.9596239977637347,
+      "learning_rate": 0.00011966756792930897,
+      "loss": 12.7442,
+      "step": 7786
+    },
+    {
+      "epoch": 1.3617923497267759,
+      "grad_norm": 0.963692403413722,
+      "learning_rate": 0.00011964905545100355,
+      "loss": 12.78,
+      "step": 7787
+    },
+    {
+      "epoch": 1.3619672131147542,
+      "grad_norm": 0.9265431055348666,
+      "learning_rate": 0.00011963054227223225,
+      "loss": 12.5935,
+      "step": 7788
+    },
+    {
+      "epoch": 1.3621420765027321,
+      "grad_norm": 0.9874231361293097,
+      "learning_rate": 0.00011961202839365507,
+      "loss": 12.7565,
+      "step": 7789
+    },
+    {
+      "epoch": 1.3623169398907105,
+      "grad_norm": 0.8840529996894472,
+      "learning_rate": 0.00011959351381593193,
+      "loss": 12.6351,
+      "step": 7790
+    },
+    {
+      "epoch": 1.3624918032786886,
+      "grad_norm": 0.9538711648673945,
+      "learning_rate": 0.00011957499853972296,
+      "loss": 12.6057,
+      "step": 7791
+    },
+    {
+      "epoch": 1.3626666666666667,
+      "grad_norm": 0.8950484655971558,
+      "learning_rate": 0.00011955648256568814,
+      "loss": 12.4563,
+      "step": 7792
+    },
+    {
+      "epoch": 1.3628415300546448,
+      "grad_norm": 0.9777278614698798,
+      "learning_rate": 0.0001195379658944876,
+      "loss": 12.4935,
+      "step": 7793
+    },
+    {
+      "epoch": 1.363016393442623,
+      "grad_norm": 0.9369883284009081,
+      "learning_rate": 0.00011951944852678138,
+      "loss": 12.5751,
+      "step": 7794
+    },
+    {
+      "epoch": 1.363191256830601,
+      "grad_norm": 0.8503147211842604,
+      "learning_rate": 0.00011950093046322964,
+      "loss": 12.4568,
+      "step": 7795
+    },
+    {
+      "epoch": 1.3633661202185792,
+      "grad_norm": 0.8669770231505273,
+      "learning_rate": 0.0001194824117044925,
+      "loss": 12.6456,
+      "step": 7796
+    },
+    {
+      "epoch": 1.3635409836065575,
+      "grad_norm": 0.8873307297309517,
+      "learning_rate": 0.00011946389225123015,
+      "loss": 12.5274,
+      "step": 7797
+    },
+    {
+      "epoch": 1.3637158469945354,
+      "grad_norm": 0.9534164888857005,
+      "learning_rate": 0.00011944537210410279,
+      "loss": 12.6363,
+      "step": 7798
+    },
+    {
+      "epoch": 1.3638907103825137,
+      "grad_norm": 0.8603947143988157,
+      "learning_rate": 0.00011942685126377062,
+      "loss": 12.2886,
+      "step": 7799
+    },
+    {
+      "epoch": 1.3640655737704919,
+      "grad_norm": 0.9412851775984222,
+      "learning_rate": 0.00011940832973089391,
+      "loss": 12.501,
+      "step": 7800
+    },
+    {
+      "epoch": 1.36424043715847,
+      "grad_norm": 1.012515211485684,
+      "learning_rate": 0.00011938980750613292,
+      "loss": 12.5301,
+      "step": 7801
+    },
+    {
+      "epoch": 1.364415300546448,
+      "grad_norm": 0.8820182490302628,
+      "learning_rate": 0.00011937128459014795,
+      "loss": 12.6149,
+      "step": 7802
+    },
+    {
+      "epoch": 1.3645901639344262,
+      "grad_norm": 1.0186108550720492,
+      "learning_rate": 0.00011935276098359929,
+      "loss": 12.6671,
+      "step": 7803
+    },
+    {
+      "epoch": 1.3647650273224043,
+      "grad_norm": 0.8635372780117059,
+      "learning_rate": 0.00011933423668714736,
+      "loss": 12.3091,
+      "step": 7804
+    },
+    {
+      "epoch": 1.3649398907103825,
+      "grad_norm": 1.0677333985157607,
+      "learning_rate": 0.00011931571170145244,
+      "loss": 12.3335,
+      "step": 7805
+    },
+    {
+      "epoch": 1.3651147540983606,
+      "grad_norm": 0.9169677872665041,
+      "learning_rate": 0.000119297186027175,
+      "loss": 12.6359,
+      "step": 7806
+    },
+    {
+      "epoch": 1.3652896174863387,
+      "grad_norm": 0.8561395997783093,
+      "learning_rate": 0.0001192786596649754,
+      "loss": 12.5565,
+      "step": 7807
+    },
+    {
+      "epoch": 1.365464480874317,
+      "grad_norm": 0.9182099629375526,
+      "learning_rate": 0.0001192601326155141,
+      "loss": 12.6331,
+      "step": 7808
+    },
+    {
+      "epoch": 1.3656393442622952,
+      "grad_norm": 1.5236426776034022,
+      "learning_rate": 0.00011924160487945157,
+      "loss": 12.533,
+      "step": 7809
+    },
+    {
+      "epoch": 1.3658142076502733,
+      "grad_norm": 0.8424455856282069,
+      "learning_rate": 0.00011922307645744834,
+      "loss": 12.4999,
+      "step": 7810
+    },
+    {
+      "epoch": 1.3659890710382514,
+      "grad_norm": 0.814139129701583,
+      "learning_rate": 0.00011920454735016489,
+      "loss": 12.4926,
+      "step": 7811
+    },
+    {
+      "epoch": 1.3661639344262295,
+      "grad_norm": 0.836455357794008,
+      "learning_rate": 0.00011918601755826175,
+      "loss": 12.5097,
+      "step": 7812
+    },
+    {
+      "epoch": 1.3663387978142076,
+      "grad_norm": 0.9340242933094057,
+      "learning_rate": 0.00011916748708239951,
+      "loss": 12.2564,
+      "step": 7813
+    },
+    {
+      "epoch": 1.3665136612021858,
+      "grad_norm": 0.8102736797283079,
+      "learning_rate": 0.00011914895592323876,
+      "loss": 12.6333,
+      "step": 7814
+    },
+    {
+      "epoch": 1.3666885245901639,
+      "grad_norm": 0.8727106513172793,
+      "learning_rate": 0.00011913042408144011,
+      "loss": 12.6906,
+      "step": 7815
+    },
+    {
+      "epoch": 1.366863387978142,
+      "grad_norm": 0.8955214625867272,
+      "learning_rate": 0.00011911189155766415,
+      "loss": 12.6469,
+      "step": 7816
+    },
+    {
+      "epoch": 1.3670382513661203,
+      "grad_norm": 0.7953103122326275,
+      "learning_rate": 0.00011909335835257161,
+      "loss": 12.509,
+      "step": 7817
+    },
+    {
+      "epoch": 1.3672131147540982,
+      "grad_norm": 0.8673130334980008,
+      "learning_rate": 0.00011907482446682315,
+      "loss": 12.5109,
+      "step": 7818
+    },
+    {
+      "epoch": 1.3673879781420766,
+      "grad_norm": 0.8268057830268651,
+      "learning_rate": 0.00011905628990107948,
+      "loss": 12.5107,
+      "step": 7819
+    },
+    {
+      "epoch": 1.3675628415300547,
+      "grad_norm": 0.9166349891660935,
+      "learning_rate": 0.00011903775465600135,
+      "loss": 12.7628,
+      "step": 7820
+    },
+    {
+      "epoch": 1.3677377049180328,
+      "grad_norm": 0.8389877429474274,
+      "learning_rate": 0.00011901921873224952,
+      "loss": 12.6503,
+      "step": 7821
+    },
+    {
+      "epoch": 1.367912568306011,
+      "grad_norm": 0.8208703600414191,
+      "learning_rate": 0.00011900068213048474,
+      "loss": 12.5585,
+      "step": 7822
+    },
+    {
+      "epoch": 1.368087431693989,
+      "grad_norm": 0.869109961002426,
+      "learning_rate": 0.00011898214485136787,
+      "loss": 12.628,
+      "step": 7823
+    },
+    {
+      "epoch": 1.3682622950819672,
+      "grad_norm": 0.8375259849910043,
+      "learning_rate": 0.00011896360689555969,
+      "loss": 12.4653,
+      "step": 7824
+    },
+    {
+      "epoch": 1.3684371584699453,
+      "grad_norm": 0.8581175611950336,
+      "learning_rate": 0.0001189450682637211,
+      "loss": 12.3777,
+      "step": 7825
+    },
+    {
+      "epoch": 1.3686120218579236,
+      "grad_norm": 0.8360790418176772,
+      "learning_rate": 0.00011892652895651296,
+      "loss": 12.6755,
+      "step": 7826
+    },
+    {
+      "epoch": 1.3687868852459015,
+      "grad_norm": 1.0757475759932231,
+      "learning_rate": 0.00011890798897459614,
+      "loss": 12.5064,
+      "step": 7827
+    },
+    {
+      "epoch": 1.3689617486338799,
+      "grad_norm": 0.8958272251030032,
+      "learning_rate": 0.00011888944831863163,
+      "loss": 12.5777,
+      "step": 7828
+    },
+    {
+      "epoch": 1.369136612021858,
+      "grad_norm": 1.0046313981024826,
+      "learning_rate": 0.00011887090698928036,
+      "loss": 12.5332,
+      "step": 7829
+    },
+    {
+      "epoch": 1.369311475409836,
+      "grad_norm": 0.9082566082554551,
+      "learning_rate": 0.0001188523649872033,
+      "loss": 12.578,
+      "step": 7830
+    },
+    {
+      "epoch": 1.3694863387978142,
+      "grad_norm": 1.004753443623804,
+      "learning_rate": 0.00011883382231306145,
+      "loss": 12.6287,
+      "step": 7831
+    },
+    {
+      "epoch": 1.3696612021857923,
+      "grad_norm": 0.9600461097670322,
+      "learning_rate": 0.00011881527896751587,
+      "loss": 12.6846,
+      "step": 7832
+    },
+    {
+      "epoch": 1.3698360655737705,
+      "grad_norm": 0.9534633139462342,
+      "learning_rate": 0.00011879673495122754,
+      "loss": 12.4696,
+      "step": 7833
+    },
+    {
+      "epoch": 1.3700109289617486,
+      "grad_norm": 1.163388115408535,
+      "learning_rate": 0.00011877819026485759,
+      "loss": 12.7515,
+      "step": 7834
+    },
+    {
+      "epoch": 1.3701857923497267,
+      "grad_norm": 0.872259090793141,
+      "learning_rate": 0.00011875964490906709,
+      "loss": 12.5811,
+      "step": 7835
+    },
+    {
+      "epoch": 1.3703606557377048,
+      "grad_norm": 0.9896534338992459,
+      "learning_rate": 0.00011874109888451717,
+      "loss": 12.7433,
+      "step": 7836
+    },
+    {
+      "epoch": 1.3705355191256832,
+      "grad_norm": 0.8640335969724767,
+      "learning_rate": 0.00011872255219186894,
+      "loss": 12.5491,
+      "step": 7837
+    },
+    {
+      "epoch": 1.3707103825136613,
+      "grad_norm": 0.8281071690797066,
+      "learning_rate": 0.00011870400483178365,
+      "loss": 12.6302,
+      "step": 7838
+    },
+    {
+      "epoch": 1.3708852459016394,
+      "grad_norm": 1.0181072451543118,
+      "learning_rate": 0.0001186854568049224,
+      "loss": 12.6972,
+      "step": 7839
+    },
+    {
+      "epoch": 1.3710601092896175,
+      "grad_norm": 0.9788166715777312,
+      "learning_rate": 0.0001186669081119465,
+      "loss": 12.4063,
+      "step": 7840
+    },
+    {
+      "epoch": 1.3712349726775956,
+      "grad_norm": 0.8947629133445,
+      "learning_rate": 0.00011864835875351708,
+      "loss": 12.6357,
+      "step": 7841
+    },
+    {
+      "epoch": 1.3714098360655738,
+      "grad_norm": 0.9111518013018266,
+      "learning_rate": 0.0001186298087302955,
+      "loss": 12.6965,
+      "step": 7842
+    },
+    {
+      "epoch": 1.3715846994535519,
+      "grad_norm": 0.8986562185590965,
+      "learning_rate": 0.00011861125804294298,
+      "loss": 12.7792,
+      "step": 7843
+    },
+    {
+      "epoch": 1.37175956284153,
+      "grad_norm": 0.8636197617306978,
+      "learning_rate": 0.00011859270669212088,
+      "loss": 12.6809,
+      "step": 7844
+    },
+    {
+      "epoch": 1.3719344262295081,
+      "grad_norm": 1.0086042681185245,
+      "learning_rate": 0.00011857415467849047,
+      "loss": 12.7306,
+      "step": 7845
+    },
+    {
+      "epoch": 1.3721092896174865,
+      "grad_norm": 0.9327098287745239,
+      "learning_rate": 0.00011855560200271316,
+      "loss": 12.6622,
+      "step": 7846
+    },
+    {
+      "epoch": 1.3722841530054644,
+      "grad_norm": 0.8753970259468623,
+      "learning_rate": 0.00011853704866545034,
+      "loss": 12.571,
+      "step": 7847
+    },
+    {
+      "epoch": 1.3724590163934427,
+      "grad_norm": 1.062422880622898,
+      "learning_rate": 0.00011851849466736338,
+      "loss": 12.4793,
+      "step": 7848
+    },
+    {
+      "epoch": 1.3726338797814208,
+      "grad_norm": 0.9465067186583179,
+      "learning_rate": 0.00011849994000911373,
+      "loss": 12.5359,
+      "step": 7849
+    },
+    {
+      "epoch": 1.372808743169399,
+      "grad_norm": 1.0725696988724776,
+      "learning_rate": 0.00011848138469136282,
+      "loss": 12.7783,
+      "step": 7850
+    },
+    {
+      "epoch": 1.372983606557377,
+      "grad_norm": 0.843156219173075,
+      "learning_rate": 0.00011846282871477217,
+      "loss": 12.5699,
+      "step": 7851
+    },
+    {
+      "epoch": 1.3731584699453552,
+      "grad_norm": 1.2059129591900153,
+      "learning_rate": 0.00011844427208000319,
+      "loss": 12.6995,
+      "step": 7852
+    },
+    {
+      "epoch": 1.3733333333333333,
+      "grad_norm": 0.9001703620168554,
+      "learning_rate": 0.00011842571478771752,
+      "loss": 12.7657,
+      "step": 7853
+    },
+    {
+      "epoch": 1.3735081967213114,
+      "grad_norm": 1.1581600383347315,
+      "learning_rate": 0.0001184071568385766,
+      "loss": 12.6385,
+      "step": 7854
+    },
+    {
+      "epoch": 1.3736830601092898,
+      "grad_norm": 1.055748750209106,
+      "learning_rate": 0.00011838859823324207,
+      "loss": 12.3759,
+      "step": 7855
+    },
+    {
+      "epoch": 1.3738579234972677,
+      "grad_norm": 1.144292393998499,
+      "learning_rate": 0.00011837003897237549,
+      "loss": 12.3197,
+      "step": 7856
+    },
+    {
+      "epoch": 1.374032786885246,
+      "grad_norm": 1.0408386346716512,
+      "learning_rate": 0.00011835147905663848,
+      "loss": 12.6327,
+      "step": 7857
+    },
+    {
+      "epoch": 1.3742076502732241,
+      "grad_norm": 1.0982955417384408,
+      "learning_rate": 0.00011833291848669267,
+      "loss": 12.3114,
+      "step": 7858
+    },
+    {
+      "epoch": 1.3743825136612022,
+      "grad_norm": 1.0926837225310502,
+      "learning_rate": 0.0001183143572631998,
+      "loss": 12.6209,
+      "step": 7859
+    },
+    {
+      "epoch": 1.3745573770491804,
+      "grad_norm": 1.005119632773031,
+      "learning_rate": 0.00011829579538682143,
+      "loss": 12.5351,
+      "step": 7860
+    },
+    {
+      "epoch": 1.3747322404371585,
+      "grad_norm": 1.1158287761797065,
+      "learning_rate": 0.00011827723285821934,
+      "loss": 12.5783,
+      "step": 7861
+    },
+    {
+      "epoch": 1.3749071038251366,
+      "grad_norm": 0.9338173015877503,
+      "learning_rate": 0.00011825866967805526,
+      "loss": 12.8744,
+      "step": 7862
+    },
+    {
+      "epoch": 1.3750819672131147,
+      "grad_norm": 1.104584723047831,
+      "learning_rate": 0.00011824010584699094,
+      "loss": 12.622,
+      "step": 7863
+    },
+    {
+      "epoch": 1.3752568306010928,
+      "grad_norm": 1.0418864733461954,
+      "learning_rate": 0.00011822154136568818,
+      "loss": 12.7454,
+      "step": 7864
+    },
+    {
+      "epoch": 1.375431693989071,
+      "grad_norm": 1.1653417333211344,
+      "learning_rate": 0.00011820297623480872,
+      "loss": 12.7228,
+      "step": 7865
+    },
+    {
+      "epoch": 1.3756065573770493,
+      "grad_norm": 1.0020311108191862,
+      "learning_rate": 0.00011818441045501445,
+      "loss": 12.5822,
+      "step": 7866
+    },
+    {
+      "epoch": 1.3757814207650272,
+      "grad_norm": 1.2741442369606824,
+      "learning_rate": 0.00011816584402696718,
+      "loss": 12.9718,
+      "step": 7867
+    },
+    {
+      "epoch": 1.3759562841530055,
+      "grad_norm": 0.9320003947245341,
+      "learning_rate": 0.0001181472769513288,
+      "loss": 12.5998,
+      "step": 7868
+    },
+    {
+      "epoch": 1.3761311475409836,
+      "grad_norm": 0.9127210715803713,
+      "learning_rate": 0.0001181287092287612,
+      "loss": 12.7065,
+      "step": 7869
+    },
+    {
+      "epoch": 1.3763060109289618,
+      "grad_norm": 1.1355961103807166,
+      "learning_rate": 0.00011811014085992631,
+      "loss": 12.6824,
+      "step": 7870
+    },
+    {
+      "epoch": 1.3764808743169399,
+      "grad_norm": 0.8727345804946404,
+      "learning_rate": 0.00011809157184548602,
+      "loss": 12.2006,
+      "step": 7871
+    },
+    {
+      "epoch": 1.376655737704918,
+      "grad_norm": 1.3268784193153638,
+      "learning_rate": 0.00011807300218610235,
+      "loss": 12.7717,
+      "step": 7872
+    },
+    {
+      "epoch": 1.3768306010928961,
+      "grad_norm": 0.8433449150304362,
+      "learning_rate": 0.00011805443188243728,
+      "loss": 12.458,
+      "step": 7873
+    },
+    {
+      "epoch": 1.3770054644808742,
+      "grad_norm": 1.1036814195247544,
+      "learning_rate": 0.00011803586093515282,
+      "loss": 12.3928,
+      "step": 7874
+    },
+    {
+      "epoch": 1.3771803278688526,
+      "grad_norm": 0.8962175984158945,
+      "learning_rate": 0.00011801728934491096,
+      "loss": 12.5673,
+      "step": 7875
+    },
+    {
+      "epoch": 1.3773551912568305,
+      "grad_norm": 0.9711220514729638,
+      "learning_rate": 0.0001179987171123738,
+      "loss": 12.8345,
+      "step": 7876
+    },
+    {
+      "epoch": 1.3775300546448088,
+      "grad_norm": 1.0531639900461418,
+      "learning_rate": 0.0001179801442382034,
+      "loss": 12.3819,
+      "step": 7877
+    },
+    {
+      "epoch": 1.377704918032787,
+      "grad_norm": 0.972145021793926,
+      "learning_rate": 0.00011796157072306189,
+      "loss": 12.6654,
+      "step": 7878
+    },
+    {
+      "epoch": 1.377879781420765,
+      "grad_norm": 1.092132929535,
+      "learning_rate": 0.00011794299656761134,
+      "loss": 12.7852,
+      "step": 7879
+    },
+    {
+      "epoch": 1.3780546448087432,
+      "grad_norm": 0.9100101867903442,
+      "learning_rate": 0.00011792442177251393,
+      "loss": 12.6399,
+      "step": 7880
+    },
+    {
+      "epoch": 1.3782295081967213,
+      "grad_norm": 1.3317901284946552,
+      "learning_rate": 0.00011790584633843182,
+      "loss": 12.8299,
+      "step": 7881
+    },
+    {
+      "epoch": 1.3784043715846994,
+      "grad_norm": 0.8606689359914058,
+      "learning_rate": 0.00011788727026602721,
+      "loss": 12.5371,
+      "step": 7882
+    },
+    {
+      "epoch": 1.3785792349726775,
+      "grad_norm": 1.148504157028118,
+      "learning_rate": 0.00011786869355596235,
+      "loss": 12.6218,
+      "step": 7883
+    },
+    {
+      "epoch": 1.3787540983606557,
+      "grad_norm": 0.9190703934725587,
+      "learning_rate": 0.00011785011620889941,
+      "loss": 12.7129,
+      "step": 7884
+    },
+    {
+      "epoch": 1.3789289617486338,
+      "grad_norm": 0.9894718395716963,
+      "learning_rate": 0.00011783153822550068,
+      "loss": 12.6709,
+      "step": 7885
+    },
+    {
+      "epoch": 1.3791038251366121,
+      "grad_norm": 0.9817490997820352,
+      "learning_rate": 0.00011781295960642845,
+      "loss": 12.6277,
+      "step": 7886
+    },
+    {
+      "epoch": 1.3792786885245902,
+      "grad_norm": 0.8782559113558503,
+      "learning_rate": 0.00011779438035234504,
+      "loss": 12.6958,
+      "step": 7887
+    },
+    {
+      "epoch": 1.3794535519125684,
+      "grad_norm": 0.994603057810569,
+      "learning_rate": 0.00011777580046391274,
+      "loss": 12.6179,
+      "step": 7888
+    },
+    {
+      "epoch": 1.3796284153005465,
+      "grad_norm": 0.8600297398714492,
+      "learning_rate": 0.00011775721994179393,
+      "loss": 12.5806,
+      "step": 7889
+    },
+    {
+      "epoch": 1.3798032786885246,
+      "grad_norm": 1.0547806562826456,
+      "learning_rate": 0.00011773863878665099,
+      "loss": 12.5368,
+      "step": 7890
+    },
+    {
+      "epoch": 1.3799781420765027,
+      "grad_norm": 0.9115800579100909,
+      "learning_rate": 0.00011772005699914628,
+      "loss": 12.7304,
+      "step": 7891
+    },
+    {
+      "epoch": 1.3801530054644808,
+      "grad_norm": 1.0215560245206745,
+      "learning_rate": 0.00011770147457994223,
+      "loss": 12.4876,
+      "step": 7892
+    },
+    {
+      "epoch": 1.380327868852459,
+      "grad_norm": 0.9247823833097195,
+      "learning_rate": 0.00011768289152970133,
+      "loss": 12.5487,
+      "step": 7893
+    },
+    {
+      "epoch": 1.380502732240437,
+      "grad_norm": 0.9158988688692056,
+      "learning_rate": 0.000117664307849086,
+      "loss": 12.5663,
+      "step": 7894
+    },
+    {
+      "epoch": 1.3806775956284154,
+      "grad_norm": 0.9913481919364826,
+      "learning_rate": 0.00011764572353875871,
+      "loss": 12.5689,
+      "step": 7895
+    },
+    {
+      "epoch": 1.3808524590163933,
+      "grad_norm": 0.9775730384117036,
+      "learning_rate": 0.00011762713859938202,
+      "loss": 12.6235,
+      "step": 7896
+    },
+    {
+      "epoch": 1.3810273224043716,
+      "grad_norm": 0.947214691673919,
+      "learning_rate": 0.00011760855303161841,
+      "loss": 12.6046,
+      "step": 7897
+    },
+    {
+      "epoch": 1.3812021857923498,
+      "grad_norm": 0.8853777583747012,
+      "learning_rate": 0.00011758996683613046,
+      "loss": 12.5324,
+      "step": 7898
+    },
+    {
+      "epoch": 1.3813770491803279,
+      "grad_norm": 0.9422511771020968,
+      "learning_rate": 0.00011757138001358075,
+      "loss": 12.5569,
+      "step": 7899
+    },
+    {
+      "epoch": 1.381551912568306,
+      "grad_norm": 0.893379091358036,
+      "learning_rate": 0.00011755279256463187,
+      "loss": 12.7007,
+      "step": 7900
+    },
+    {
+      "epoch": 1.3817267759562841,
+      "grad_norm": 0.8275049726256694,
+      "learning_rate": 0.00011753420448994645,
+      "loss": 12.6593,
+      "step": 7901
+    },
+    {
+      "epoch": 1.3819016393442622,
+      "grad_norm": 0.8444329406386587,
+      "learning_rate": 0.00011751561579018715,
+      "loss": 12.2365,
+      "step": 7902
+    },
+    {
+      "epoch": 1.3820765027322404,
+      "grad_norm": 0.8270051973597067,
+      "learning_rate": 0.00011749702646601657,
+      "loss": 12.7264,
+      "step": 7903
+    },
+    {
+      "epoch": 1.3822513661202187,
+      "grad_norm": 0.9095231921756851,
+      "learning_rate": 0.00011747843651809747,
+      "loss": 12.9029,
+      "step": 7904
+    },
+    {
+      "epoch": 1.3824262295081966,
+      "grad_norm": 0.9089073619801391,
+      "learning_rate": 0.0001174598459470925,
+      "loss": 12.8196,
+      "step": 7905
+    },
+    {
+      "epoch": 1.382601092896175,
+      "grad_norm": 0.9131029892023901,
+      "learning_rate": 0.00011744125475366446,
+      "loss": 12.6088,
+      "step": 7906
+    },
+    {
+      "epoch": 1.382775956284153,
+      "grad_norm": 0.9529255311427445,
+      "learning_rate": 0.00011742266293847603,
+      "loss": 12.6615,
+      "step": 7907
+    },
+    {
+      "epoch": 1.3829508196721312,
+      "grad_norm": 1.0308733509951202,
+      "learning_rate": 0.00011740407050219003,
+      "loss": 12.7084,
+      "step": 7908
+    },
+    {
+      "epoch": 1.3831256830601093,
+      "grad_norm": 0.9907155924493855,
+      "learning_rate": 0.00011738547744546926,
+      "loss": 12.6182,
+      "step": 7909
+    },
+    {
+      "epoch": 1.3833005464480874,
+      "grad_norm": 1.1329131402258212,
+      "learning_rate": 0.00011736688376897653,
+      "loss": 12.489,
+      "step": 7910
+    },
+    {
+      "epoch": 1.3834754098360655,
+      "grad_norm": 0.9403609561146012,
+      "learning_rate": 0.00011734828947337469,
+      "loss": 12.7053,
+      "step": 7911
+    },
+    {
+      "epoch": 1.3836502732240437,
+      "grad_norm": 1.1764581596923591,
+      "learning_rate": 0.00011732969455932658,
+      "loss": 12.721,
+      "step": 7912
+    },
+    {
+      "epoch": 1.3838251366120218,
+      "grad_norm": 0.8908020752447745,
+      "learning_rate": 0.00011731109902749514,
+      "loss": 12.8447,
+      "step": 7913
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 1.1343462451652946,
+      "learning_rate": 0.00011729250287854321,
+      "loss": 12.7451,
+      "step": 7914
+    },
+    {
+      "epoch": 1.3841748633879782,
+      "grad_norm": 0.9792181697494778,
+      "learning_rate": 0.00011727390611313378,
+      "loss": 12.6123,
+      "step": 7915
+    },
+    {
+      "epoch": 1.3843497267759564,
+      "grad_norm": 0.8597535635477715,
+      "learning_rate": 0.00011725530873192975,
+      "loss": 12.4225,
+      "step": 7916
+    },
+    {
+      "epoch": 1.3845245901639345,
+      "grad_norm": 0.9874289397649618,
+      "learning_rate": 0.00011723671073559414,
+      "loss": 12.4141,
+      "step": 7917
+    },
+    {
+      "epoch": 1.3846994535519126,
+      "grad_norm": 0.8421589883358082,
+      "learning_rate": 0.00011721811212478995,
+      "loss": 12.4798,
+      "step": 7918
+    },
+    {
+      "epoch": 1.3848743169398907,
+      "grad_norm": 0.9548579485864868,
+      "learning_rate": 0.00011719951290018016,
+      "loss": 12.5525,
+      "step": 7919
+    },
+    {
+      "epoch": 1.3850491803278688,
+      "grad_norm": 0.9198298617152987,
+      "learning_rate": 0.00011718091306242784,
+      "loss": 12.6496,
+      "step": 7920
+    },
+    {
+      "epoch": 1.385224043715847,
+      "grad_norm": 0.8514643665634862,
+      "learning_rate": 0.00011716231261219606,
+      "loss": 12.5642,
+      "step": 7921
+    },
+    {
+      "epoch": 1.385398907103825,
+      "grad_norm": 1.123645358787463,
+      "learning_rate": 0.00011714371155014785,
+      "loss": 12.8132,
+      "step": 7922
+    },
+    {
+      "epoch": 1.3855737704918032,
+      "grad_norm": 0.9502701764687163,
+      "learning_rate": 0.00011712510987694638,
+      "loss": 12.5999,
+      "step": 7923
+    },
+    {
+      "epoch": 1.3857486338797815,
+      "grad_norm": 0.9834596055010704,
+      "learning_rate": 0.00011710650759325474,
+      "loss": 12.7273,
+      "step": 7924
+    },
+    {
+      "epoch": 1.3859234972677594,
+      "grad_norm": 0.9372660442391497,
+      "learning_rate": 0.00011708790469973609,
+      "loss": 12.5925,
+      "step": 7925
+    },
+    {
+      "epoch": 1.3860983606557378,
+      "grad_norm": 1.0452017102530384,
+      "learning_rate": 0.00011706930119705361,
+      "loss": 12.6416,
+      "step": 7926
+    },
+    {
+      "epoch": 1.386273224043716,
+      "grad_norm": 1.1980081574686838,
+      "learning_rate": 0.00011705069708587046,
+      "loss": 12.7017,
+      "step": 7927
+    },
+    {
+      "epoch": 1.386448087431694,
+      "grad_norm": 0.8641248803606547,
+      "learning_rate": 0.00011703209236684993,
+      "loss": 12.6358,
+      "step": 7928
+    },
+    {
+      "epoch": 1.3866229508196721,
+      "grad_norm": 1.1654227415405718,
+      "learning_rate": 0.00011701348704065516,
+      "loss": 12.6022,
+      "step": 7929
+    },
+    {
+      "epoch": 1.3867978142076502,
+      "grad_norm": 0.9654692122290163,
+      "learning_rate": 0.0001169948811079495,
+      "loss": 12.644,
+      "step": 7930
+    },
+    {
+      "epoch": 1.3869726775956284,
+      "grad_norm": 0.8802927529762463,
+      "learning_rate": 0.00011697627456939615,
+      "loss": 12.4708,
+      "step": 7931
+    },
+    {
+      "epoch": 1.3871475409836065,
+      "grad_norm": 0.8799131035438351,
+      "learning_rate": 0.00011695766742565845,
+      "loss": 12.6453,
+      "step": 7932
+    },
+    {
+      "epoch": 1.3873224043715848,
+      "grad_norm": 1.1423561218132539,
+      "learning_rate": 0.00011693905967739973,
+      "loss": 12.4796,
+      "step": 7933
+    },
+    {
+      "epoch": 1.3874972677595627,
+      "grad_norm": 0.8827624758173159,
+      "learning_rate": 0.00011692045132528334,
+      "loss": 12.7193,
+      "step": 7934
+    },
+    {
+      "epoch": 1.387672131147541,
+      "grad_norm": 0.8615053958119717,
+      "learning_rate": 0.00011690184236997259,
+      "loss": 12.7476,
+      "step": 7935
+    },
+    {
+      "epoch": 1.3878469945355192,
+      "grad_norm": 0.9604299891312083,
+      "learning_rate": 0.00011688323281213091,
+      "loss": 12.8099,
+      "step": 7936
+    },
+    {
+      "epoch": 1.3880218579234973,
+      "grad_norm": 0.9464052923124322,
+      "learning_rate": 0.00011686462265242171,
+      "loss": 12.5566,
+      "step": 7937
+    },
+    {
+      "epoch": 1.3881967213114754,
+      "grad_norm": 0.8669280836354059,
+      "learning_rate": 0.00011684601189150846,
+      "loss": 12.7413,
+      "step": 7938
+    },
+    {
+      "epoch": 1.3883715846994535,
+      "grad_norm": 0.8664925177975221,
+      "learning_rate": 0.00011682740053005451,
+      "loss": 12.2908,
+      "step": 7939
+    },
+    {
+      "epoch": 1.3885464480874317,
+      "grad_norm": 0.9421812529328959,
+      "learning_rate": 0.00011680878856872343,
+      "loss": 12.3747,
+      "step": 7940
+    },
+    {
+      "epoch": 1.3887213114754098,
+      "grad_norm": 0.9392036518611391,
+      "learning_rate": 0.00011679017600817868,
+      "loss": 12.4302,
+      "step": 7941
+    },
+    {
+      "epoch": 1.388896174863388,
+      "grad_norm": 0.9723632881289356,
+      "learning_rate": 0.00011677156284908374,
+      "loss": 12.6947,
+      "step": 7942
+    },
+    {
+      "epoch": 1.389071038251366,
+      "grad_norm": 0.8897490924623733,
+      "learning_rate": 0.00011675294909210221,
+      "loss": 12.6963,
+      "step": 7943
+    },
+    {
+      "epoch": 1.3892459016393444,
+      "grad_norm": 1.050841188742159,
+      "learning_rate": 0.00011673433473789761,
+      "loss": 12.6457,
+      "step": 7944
+    },
+    {
+      "epoch": 1.3894207650273225,
+      "grad_norm": 0.8707267301441691,
+      "learning_rate": 0.00011671571978713351,
+      "loss": 12.2238,
+      "step": 7945
+    },
+    {
+      "epoch": 1.3895956284153006,
+      "grad_norm": 1.032806682886381,
+      "learning_rate": 0.00011669710424047354,
+      "loss": 12.3315,
+      "step": 7946
+    },
+    {
+      "epoch": 1.3897704918032787,
+      "grad_norm": 0.8833834070846938,
+      "learning_rate": 0.00011667848809858132,
+      "loss": 12.5249,
+      "step": 7947
+    },
+    {
+      "epoch": 1.3899453551912568,
+      "grad_norm": 0.9106086689421099,
+      "learning_rate": 0.00011665987136212049,
+      "loss": 12.6183,
+      "step": 7948
+    },
+    {
+      "epoch": 1.390120218579235,
+      "grad_norm": 0.9795115762515854,
+      "learning_rate": 0.0001166412540317547,
+      "loss": 12.4765,
+      "step": 7949
+    },
+    {
+      "epoch": 1.390295081967213,
+      "grad_norm": 0.9077778342022489,
+      "learning_rate": 0.00011662263610814764,
+      "loss": 12.4281,
+      "step": 7950
+    },
+    {
+      "epoch": 1.3904699453551912,
+      "grad_norm": 0.9423315888084247,
+      "learning_rate": 0.00011660401759196304,
+      "loss": 12.6356,
+      "step": 7951
+    },
+    {
+      "epoch": 1.3906448087431693,
+      "grad_norm": 0.9035094142861478,
+      "learning_rate": 0.00011658539848386461,
+      "loss": 12.7288,
+      "step": 7952
+    },
+    {
+      "epoch": 1.3908196721311477,
+      "grad_norm": 1.0777137927674068,
+      "learning_rate": 0.00011656677878451609,
+      "loss": 12.7031,
+      "step": 7953
+    },
+    {
+      "epoch": 1.3909945355191256,
+      "grad_norm": 0.8776020102263408,
+      "learning_rate": 0.00011654815849458125,
+      "loss": 12.7291,
+      "step": 7954
+    },
+    {
+      "epoch": 1.391169398907104,
+      "grad_norm": 0.8503199831813772,
+      "learning_rate": 0.00011652953761472393,
+      "loss": 12.4776,
+      "step": 7955
+    },
+    {
+      "epoch": 1.391344262295082,
+      "grad_norm": 1.064154936823554,
+      "learning_rate": 0.00011651091614560791,
+      "loss": 12.5655,
+      "step": 7956
+    },
+    {
+      "epoch": 1.3915191256830601,
+      "grad_norm": 0.9223199807508609,
+      "learning_rate": 0.00011649229408789699,
+      "loss": 12.6679,
+      "step": 7957
+    },
+    {
+      "epoch": 1.3916939890710383,
+      "grad_norm": 0.9284232172799366,
+      "learning_rate": 0.00011647367144225509,
+      "loss": 12.3927,
+      "step": 7958
+    },
+    {
+      "epoch": 1.3918688524590164,
+      "grad_norm": 0.8546042638400625,
+      "learning_rate": 0.000116455048209346,
+      "loss": 12.4286,
+      "step": 7959
+    },
+    {
+      "epoch": 1.3920437158469945,
+      "grad_norm": 0.9538522769695175,
+      "learning_rate": 0.00011643642438983374,
+      "loss": 12.4074,
+      "step": 7960
+    },
+    {
+      "epoch": 1.3922185792349726,
+      "grad_norm": 0.9917552937520397,
+      "learning_rate": 0.00011641779998438209,
+      "loss": 12.7839,
+      "step": 7961
+    },
+    {
+      "epoch": 1.392393442622951,
+      "grad_norm": 0.9468468770626227,
+      "learning_rate": 0.00011639917499365507,
+      "loss": 12.3297,
+      "step": 7962
+    },
+    {
+      "epoch": 1.3925683060109288,
+      "grad_norm": 0.8458935464203001,
+      "learning_rate": 0.0001163805494183166,
+      "loss": 12.8068,
+      "step": 7963
+    },
+    {
+      "epoch": 1.3927431693989072,
+      "grad_norm": 1.0121195178854396,
+      "learning_rate": 0.00011636192325903073,
+      "loss": 12.4265,
+      "step": 7964
+    },
+    {
+      "epoch": 1.3929180327868853,
+      "grad_norm": 0.8913651478574601,
+      "learning_rate": 0.00011634329651646136,
+      "loss": 12.5387,
+      "step": 7965
+    },
+    {
+      "epoch": 1.3930928961748634,
+      "grad_norm": 0.9347919849440812,
+      "learning_rate": 0.00011632466919127263,
+      "loss": 12.5983,
+      "step": 7966
+    },
+    {
+      "epoch": 1.3932677595628415,
+      "grad_norm": 0.8531780799124907,
+      "learning_rate": 0.00011630604128412845,
+      "loss": 12.2353,
+      "step": 7967
+    },
+    {
+      "epoch": 1.3934426229508197,
+      "grad_norm": 0.8921113142407534,
+      "learning_rate": 0.00011628741279569299,
+      "loss": 12.397,
+      "step": 7968
+    },
+    {
+      "epoch": 1.3936174863387978,
+      "grad_norm": 0.926219055058836,
+      "learning_rate": 0.00011626878372663028,
+      "loss": 12.2978,
+      "step": 7969
+    },
+    {
+      "epoch": 1.393792349726776,
+      "grad_norm": 0.8797016032694451,
+      "learning_rate": 0.00011625015407760447,
+      "loss": 12.3892,
+      "step": 7970
+    },
+    {
+      "epoch": 1.393967213114754,
+      "grad_norm": 0.8611911004604361,
+      "learning_rate": 0.0001162315238492796,
+      "loss": 12.5649,
+      "step": 7971
+    },
+    {
+      "epoch": 1.3941420765027321,
+      "grad_norm": 0.8901923008601879,
+      "learning_rate": 0.00011621289304231988,
+      "loss": 12.5613,
+      "step": 7972
+    },
+    {
+      "epoch": 1.3943169398907105,
+      "grad_norm": 1.0125876045249942,
+      "learning_rate": 0.00011619426165738949,
+      "loss": 12.6438,
+      "step": 7973
+    },
+    {
+      "epoch": 1.3944918032786886,
+      "grad_norm": 0.9031508726129753,
+      "learning_rate": 0.00011617562969515259,
+      "loss": 12.76,
+      "step": 7974
+    },
+    {
+      "epoch": 1.3946666666666667,
+      "grad_norm": 1.0055400827526453,
+      "learning_rate": 0.0001161569971562734,
+      "loss": 12.6096,
+      "step": 7975
+    },
+    {
+      "epoch": 1.3948415300546448,
+      "grad_norm": 0.8803613852173099,
+      "learning_rate": 0.00011613836404141612,
+      "loss": 12.5201,
+      "step": 7976
+    },
+    {
+      "epoch": 1.395016393442623,
+      "grad_norm": 1.0305456766746206,
+      "learning_rate": 0.00011611973035124502,
+      "loss": 12.4493,
+      "step": 7977
+    },
+    {
+      "epoch": 1.395191256830601,
+      "grad_norm": 0.8968774105421765,
+      "learning_rate": 0.00011610109608642434,
+      "loss": 12.4391,
+      "step": 7978
+    },
+    {
+      "epoch": 1.3953661202185792,
+      "grad_norm": 1.0112860116186733,
+      "learning_rate": 0.00011608246124761846,
+      "loss": 12.467,
+      "step": 7979
+    },
+    {
+      "epoch": 1.3955409836065573,
+      "grad_norm": 0.9073185846513934,
+      "learning_rate": 0.00011606382583549157,
+      "loss": 12.4794,
+      "step": 7980
+    },
+    {
+      "epoch": 1.3957158469945354,
+      "grad_norm": 0.8848009583369179,
+      "learning_rate": 0.00011604518985070808,
+      "loss": 12.5492,
+      "step": 7981
+    },
+    {
+      "epoch": 1.3958907103825138,
+      "grad_norm": 0.979123195116063,
+      "learning_rate": 0.00011602655329393228,
+      "loss": 12.5547,
+      "step": 7982
+    },
+    {
+      "epoch": 1.3960655737704917,
+      "grad_norm": 0.9197889280110713,
+      "learning_rate": 0.00011600791616582864,
+      "loss": 12.518,
+      "step": 7983
+    },
+    {
+      "epoch": 1.39624043715847,
+      "grad_norm": 0.8926352842722569,
+      "learning_rate": 0.00011598927846706147,
+      "loss": 12.6655,
+      "step": 7984
+    },
+    {
+      "epoch": 1.3964153005464481,
+      "grad_norm": 0.7943970932645256,
+      "learning_rate": 0.0001159706401982952,
+      "loss": 12.4877,
+      "step": 7985
+    },
+    {
+      "epoch": 1.3965901639344263,
+      "grad_norm": 0.912504611348873,
+      "learning_rate": 0.00011595200136019427,
+      "loss": 12.6975,
+      "step": 7986
+    },
+    {
+      "epoch": 1.3967650273224044,
+      "grad_norm": 0.8346170332585905,
+      "learning_rate": 0.00011593336195342314,
+      "loss": 12.6479,
+      "step": 7987
+    },
+    {
+      "epoch": 1.3969398907103825,
+      "grad_norm": 0.8399099375430827,
+      "learning_rate": 0.00011591472197864628,
+      "loss": 12.6438,
+      "step": 7988
+    },
+    {
+      "epoch": 1.3971147540983606,
+      "grad_norm": 0.8519913266368846,
+      "learning_rate": 0.00011589608143652815,
+      "loss": 12.5495,
+      "step": 7989
+    },
+    {
+      "epoch": 1.3972896174863387,
+      "grad_norm": 0.9465141261536565,
+      "learning_rate": 0.00011587744032773326,
+      "loss": 12.6467,
+      "step": 7990
+    },
+    {
+      "epoch": 1.397464480874317,
+      "grad_norm": 0.9862958952235513,
+      "learning_rate": 0.00011585879865292621,
+      "loss": 12.3817,
+      "step": 7991
+    },
+    {
+      "epoch": 1.397639344262295,
+      "grad_norm": 0.8780140850428705,
+      "learning_rate": 0.00011584015641277153,
+      "loss": 12.4058,
+      "step": 7992
+    },
+    {
+      "epoch": 1.3978142076502733,
+      "grad_norm": 0.9403512542414041,
+      "learning_rate": 0.00011582151360793376,
+      "loss": 12.594,
+      "step": 7993
+    },
+    {
+      "epoch": 1.3979890710382514,
+      "grad_norm": 0.8669989358492076,
+      "learning_rate": 0.00011580287023907755,
+      "loss": 12.6506,
+      "step": 7994
+    },
+    {
+      "epoch": 1.3981639344262295,
+      "grad_norm": 0.8577463130354455,
+      "learning_rate": 0.00011578422630686743,
+      "loss": 12.6793,
+      "step": 7995
+    },
+    {
+      "epoch": 1.3983387978142077,
+      "grad_norm": 0.8593438236418616,
+      "learning_rate": 0.00011576558181196812,
+      "loss": 12.6281,
+      "step": 7996
+    },
+    {
+      "epoch": 1.3985136612021858,
+      "grad_norm": 1.0466295585992318,
+      "learning_rate": 0.00011574693675504422,
+      "loss": 12.6437,
+      "step": 7997
+    },
+    {
+      "epoch": 1.398688524590164,
+      "grad_norm": 1.0583871057342003,
+      "learning_rate": 0.0001157282911367604,
+      "loss": 12.7407,
+      "step": 7998
+    },
+    {
+      "epoch": 1.398863387978142,
+      "grad_norm": 0.91251418368265,
+      "learning_rate": 0.00011570964495778141,
+      "loss": 12.6795,
+      "step": 7999
+    },
+    {
+      "epoch": 1.3990382513661201,
+      "grad_norm": 0.9585482688276755,
+      "learning_rate": 0.00011569099821877194,
+      "loss": 12.517,
+      "step": 8000
+    },
+    {
+      "epoch": 1.3992131147540983,
+      "grad_norm": 0.9665517518920412,
+      "learning_rate": 0.0001156723509203967,
+      "loss": 12.2801,
+      "step": 8001
+    },
+    {
+      "epoch": 1.3993879781420766,
+      "grad_norm": 0.9033650770668575,
+      "learning_rate": 0.00011565370306332046,
+      "loss": 12.6792,
+      "step": 8002
+    },
+    {
+      "epoch": 1.3995628415300547,
+      "grad_norm": 1.216865025799233,
+      "learning_rate": 0.00011563505464820802,
+      "loss": 12.7375,
+      "step": 8003
+    },
+    {
+      "epoch": 1.3997377049180328,
+      "grad_norm": 0.8479897698115457,
+      "learning_rate": 0.00011561640567572412,
+      "loss": 12.6458,
+      "step": 8004
+    },
+    {
+      "epoch": 1.399912568306011,
+      "grad_norm": 1.0718468462564548,
+      "learning_rate": 0.00011559775614653364,
+      "loss": 12.4865,
+      "step": 8005
+    },
+    {
+      "epoch": 1.400087431693989,
+      "grad_norm": 0.856247618848395,
+      "learning_rate": 0.00011557910606130133,
+      "loss": 12.7268,
+      "step": 8006
+    },
+    {
+      "epoch": 1.4002622950819672,
+      "grad_norm": 0.9770208599704708,
+      "learning_rate": 0.0001155604554206921,
+      "loss": 12.4304,
+      "step": 8007
+    },
+    {
+      "epoch": 1.4004371584699453,
+      "grad_norm": 0.9528334720336898,
+      "learning_rate": 0.00011554180422537083,
+      "loss": 12.4172,
+      "step": 8008
+    },
+    {
+      "epoch": 1.4006120218579234,
+      "grad_norm": 0.8211156037317089,
+      "learning_rate": 0.0001155231524760024,
+      "loss": 12.5029,
+      "step": 8009
+    },
+    {
+      "epoch": 1.4007868852459016,
+      "grad_norm": 0.9325524105996914,
+      "learning_rate": 0.00011550450017325169,
+      "loss": 12.4458,
+      "step": 8010
+    },
+    {
+      "epoch": 1.40096174863388,
+      "grad_norm": 0.8660123596393129,
+      "learning_rate": 0.00011548584731778371,
+      "loss": 12.4431,
+      "step": 8011
+    },
+    {
+      "epoch": 1.4011366120218578,
+      "grad_norm": 0.8578025275478971,
+      "learning_rate": 0.00011546719391026335,
+      "loss": 12.6814,
+      "step": 8012
+    },
+    {
+      "epoch": 1.4013114754098361,
+      "grad_norm": 0.9272406704472043,
+      "learning_rate": 0.00011544853995135558,
+      "loss": 12.7633,
+      "step": 8013
+    },
+    {
+      "epoch": 1.4014863387978143,
+      "grad_norm": 0.9199913491520667,
+      "learning_rate": 0.00011542988544172544,
+      "loss": 12.7577,
+      "step": 8014
+    },
+    {
+      "epoch": 1.4016612021857924,
+      "grad_norm": 0.8764163785035407,
+      "learning_rate": 0.00011541123038203789,
+      "loss": 12.2416,
+      "step": 8015
+    },
+    {
+      "epoch": 1.4018360655737705,
+      "grad_norm": 0.8482641135191635,
+      "learning_rate": 0.000115392574772958,
+      "loss": 12.6046,
+      "step": 8016
+    },
+    {
+      "epoch": 1.4020109289617486,
+      "grad_norm": 0.8747309634099323,
+      "learning_rate": 0.00011537391861515077,
+      "loss": 12.4284,
+      "step": 8017
+    },
+    {
+      "epoch": 1.4021857923497267,
+      "grad_norm": 0.8722437730997091,
+      "learning_rate": 0.00011535526190928134,
+      "loss": 12.5248,
+      "step": 8018
+    },
+    {
+      "epoch": 1.4023606557377049,
+      "grad_norm": 0.9133766187395744,
+      "learning_rate": 0.00011533660465601475,
+      "loss": 12.5714,
+      "step": 8019
+    },
+    {
+      "epoch": 1.4025355191256832,
+      "grad_norm": 0.8631774623989653,
+      "learning_rate": 0.00011531794685601614,
+      "loss": 12.6192,
+      "step": 8020
+    },
+    {
+      "epoch": 1.402710382513661,
+      "grad_norm": 0.9356658145921931,
+      "learning_rate": 0.00011529928850995057,
+      "loss": 12.5894,
+      "step": 8021
+    },
+    {
+      "epoch": 1.4028852459016394,
+      "grad_norm": 0.8686346458850325,
+      "learning_rate": 0.0001152806296184833,
+      "loss": 12.4725,
+      "step": 8022
+    },
+    {
+      "epoch": 1.4030601092896176,
+      "grad_norm": 0.9777206669878337,
+      "learning_rate": 0.0001152619701822794,
+      "loss": 12.3073,
+      "step": 8023
+    },
+    {
+      "epoch": 1.4032349726775957,
+      "grad_norm": 0.8834434964812903,
+      "learning_rate": 0.00011524331020200412,
+      "loss": 12.2598,
+      "step": 8024
+    },
+    {
+      "epoch": 1.4034098360655738,
+      "grad_norm": 0.9490806153005406,
+      "learning_rate": 0.00011522464967832262,
+      "loss": 12.6085,
+      "step": 8025
+    },
+    {
+      "epoch": 1.403584699453552,
+      "grad_norm": 0.8972070049129305,
+      "learning_rate": 0.00011520598861190015,
+      "loss": 12.6946,
+      "step": 8026
+    },
+    {
+      "epoch": 1.40375956284153,
+      "grad_norm": 0.995678617654657,
+      "learning_rate": 0.00011518732700340196,
+      "loss": 12.4725,
+      "step": 8027
+    },
+    {
+      "epoch": 1.4039344262295081,
+      "grad_norm": 0.8526531440181404,
+      "learning_rate": 0.00011516866485349331,
+      "loss": 12.7587,
+      "step": 8028
+    },
+    {
+      "epoch": 1.4041092896174863,
+      "grad_norm": 0.8458201136649123,
+      "learning_rate": 0.00011515000216283947,
+      "loss": 12.7064,
+      "step": 8029
+    },
+    {
+      "epoch": 1.4042841530054644,
+      "grad_norm": 0.8642909455588227,
+      "learning_rate": 0.00011513133893210578,
+      "loss": 12.6469,
+      "step": 8030
+    },
+    {
+      "epoch": 1.4044590163934427,
+      "grad_norm": 0.8758453786330523,
+      "learning_rate": 0.00011511267516195749,
+      "loss": 12.2833,
+      "step": 8031
+    },
+    {
+      "epoch": 1.4046338797814208,
+      "grad_norm": 0.8523067464793508,
+      "learning_rate": 0.00011509401085306002,
+      "loss": 12.4334,
+      "step": 8032
+    },
+    {
+      "epoch": 1.404808743169399,
+      "grad_norm": 0.852813966849903,
+      "learning_rate": 0.00011507534600607869,
+      "loss": 12.4188,
+      "step": 8033
+    },
+    {
+      "epoch": 1.404983606557377,
+      "grad_norm": 0.8325072565066433,
+      "learning_rate": 0.0001150566806216789,
+      "loss": 12.3678,
+      "step": 8034
+    },
+    {
+      "epoch": 1.4051584699453552,
+      "grad_norm": 0.8940028765002682,
+      "learning_rate": 0.00011503801470052603,
+      "loss": 12.6453,
+      "step": 8035
+    },
+    {
+      "epoch": 1.4053333333333333,
+      "grad_norm": 0.8296742257197316,
+      "learning_rate": 0.00011501934824328549,
+      "loss": 12.3804,
+      "step": 8036
+    },
+    {
+      "epoch": 1.4055081967213114,
+      "grad_norm": 0.8377614859113718,
+      "learning_rate": 0.00011500068125062276,
+      "loss": 12.3476,
+      "step": 8037
+    },
+    {
+      "epoch": 1.4056830601092896,
+      "grad_norm": 0.8961936924633441,
+      "learning_rate": 0.00011498201372320323,
+      "loss": 12.4096,
+      "step": 8038
+    },
+    {
+      "epoch": 1.4058579234972677,
+      "grad_norm": 0.9643673073913197,
+      "learning_rate": 0.00011496334566169247,
+      "loss": 12.7254,
+      "step": 8039
+    },
+    {
+      "epoch": 1.406032786885246,
+      "grad_norm": 0.9915101562299861,
+      "learning_rate": 0.00011494467706675587,
+      "loss": 12.6477,
+      "step": 8040
+    },
+    {
+      "epoch": 1.406207650273224,
+      "grad_norm": 0.8247933278577888,
+      "learning_rate": 0.00011492600793905903,
+      "loss": 12.8147,
+      "step": 8041
+    },
+    {
+      "epoch": 1.4063825136612023,
+      "grad_norm": 0.8336812140707922,
+      "learning_rate": 0.00011490733827926743,
+      "loss": 12.685,
+      "step": 8042
+    },
+    {
+      "epoch": 1.4065573770491804,
+      "grad_norm": 1.1527840369446487,
+      "learning_rate": 0.00011488866808804664,
+      "loss": 12.7705,
+      "step": 8043
+    },
+    {
+      "epoch": 1.4067322404371585,
+      "grad_norm": 0.8126145695233117,
+      "learning_rate": 0.0001148699973660622,
+      "loss": 12.5866,
+      "step": 8044
+    },
+    {
+      "epoch": 1.4069071038251366,
+      "grad_norm": 0.9410575781358811,
+      "learning_rate": 0.00011485132611397978,
+      "loss": 12.6587,
+      "step": 8045
+    },
+    {
+      "epoch": 1.4070819672131147,
+      "grad_norm": 0.9848551679955331,
+      "learning_rate": 0.00011483265433246488,
+      "loss": 12.491,
+      "step": 8046
+    },
+    {
+      "epoch": 1.4072568306010929,
+      "grad_norm": 0.9006759113550658,
+      "learning_rate": 0.00011481398202218323,
+      "loss": 12.5296,
+      "step": 8047
+    },
+    {
+      "epoch": 1.407431693989071,
+      "grad_norm": 0.9418464108024169,
+      "learning_rate": 0.00011479530918380042,
+      "loss": 12.6155,
+      "step": 8048
+    },
+    {
+      "epoch": 1.4076065573770493,
+      "grad_norm": 0.9308131157104319,
+      "learning_rate": 0.00011477663581798213,
+      "loss": 12.6283,
+      "step": 8049
+    },
+    {
+      "epoch": 1.4077814207650272,
+      "grad_norm": 1.0065801516252006,
+      "learning_rate": 0.00011475796192539403,
+      "loss": 12.696,
+      "step": 8050
+    },
+    {
+      "epoch": 1.4079562841530056,
+      "grad_norm": 0.9562460532151865,
+      "learning_rate": 0.00011473928750670182,
+      "loss": 12.9542,
+      "step": 8051
+    },
+    {
+      "epoch": 1.4081311475409837,
+      "grad_norm": 0.9501937208254427,
+      "learning_rate": 0.00011472061256257122,
+      "loss": 12.6394,
+      "step": 8052
+    },
+    {
+      "epoch": 1.4083060109289618,
+      "grad_norm": 0.8967756032808688,
+      "learning_rate": 0.00011470193709366799,
+      "loss": 12.129,
+      "step": 8053
+    },
+    {
+      "epoch": 1.40848087431694,
+      "grad_norm": 0.928604627628387,
+      "learning_rate": 0.0001146832611006579,
+      "loss": 12.5535,
+      "step": 8054
+    },
+    {
+      "epoch": 1.408655737704918,
+      "grad_norm": 0.9027606325808583,
+      "learning_rate": 0.0001146645845842067,
+      "loss": 12.6121,
+      "step": 8055
+    },
+    {
+      "epoch": 1.4088306010928962,
+      "grad_norm": 0.9141159130154033,
+      "learning_rate": 0.00011464590754498019,
+      "loss": 12.5691,
+      "step": 8056
+    },
+    {
+      "epoch": 1.4090054644808743,
+      "grad_norm": 0.9016570353533077,
+      "learning_rate": 0.00011462722998364417,
+      "loss": 12.7891,
+      "step": 8057
+    },
+    {
+      "epoch": 1.4091803278688524,
+      "grad_norm": 0.9747415137211107,
+      "learning_rate": 0.00011460855190086453,
+      "loss": 12.5256,
+      "step": 8058
+    },
+    {
+      "epoch": 1.4093551912568305,
+      "grad_norm": 1.12396074627741,
+      "learning_rate": 0.00011458987329730707,
+      "loss": 12.6154,
+      "step": 8059
+    },
+    {
+      "epoch": 1.4095300546448088,
+      "grad_norm": 0.8902079571065487,
+      "learning_rate": 0.00011457119417363767,
+      "loss": 12.6641,
+      "step": 8060
+    },
+    {
+      "epoch": 1.4097049180327867,
+      "grad_norm": 0.9456689248549719,
+      "learning_rate": 0.0001145525145305222,
+      "loss": 12.5362,
+      "step": 8061
+    },
+    {
+      "epoch": 1.409879781420765,
+      "grad_norm": 0.8746256928648977,
+      "learning_rate": 0.0001145338343686266,
+      "loss": 12.7886,
+      "step": 8062
+    },
+    {
+      "epoch": 1.4100546448087432,
+      "grad_norm": 0.9324851195912152,
+      "learning_rate": 0.00011451515368861679,
+      "loss": 12.6587,
+      "step": 8063
+    },
+    {
+      "epoch": 1.4102295081967213,
+      "grad_norm": 0.9195890721004482,
+      "learning_rate": 0.00011449647249115872,
+      "loss": 12.5763,
+      "step": 8064
+    },
+    {
+      "epoch": 1.4104043715846994,
+      "grad_norm": 0.8103223552353603,
+      "learning_rate": 0.00011447779077691833,
+      "loss": 12.4925,
+      "step": 8065
+    },
+    {
+      "epoch": 1.4105792349726776,
+      "grad_norm": 0.828524914098734,
+      "learning_rate": 0.00011445910854656161,
+      "loss": 12.4764,
+      "step": 8066
+    },
+    {
+      "epoch": 1.4107540983606557,
+      "grad_norm": 0.8829637359439767,
+      "learning_rate": 0.00011444042580075461,
+      "loss": 12.7371,
+      "step": 8067
+    },
+    {
+      "epoch": 1.4109289617486338,
+      "grad_norm": 0.8646473611645243,
+      "learning_rate": 0.00011442174254016325,
+      "loss": 12.4606,
+      "step": 8068
+    },
+    {
+      "epoch": 1.4111038251366121,
+      "grad_norm": 0.8658246802698257,
+      "learning_rate": 0.00011440305876545363,
+      "loss": 12.3427,
+      "step": 8069
+    },
+    {
+      "epoch": 1.41127868852459,
+      "grad_norm": 0.8755309807934462,
+      "learning_rate": 0.0001143843744772918,
+      "loss": 12.7166,
+      "step": 8070
+    },
+    {
+      "epoch": 1.4114535519125684,
+      "grad_norm": 0.9374762784819018,
+      "learning_rate": 0.0001143656896763438,
+      "loss": 12.7516,
+      "step": 8071
+    },
+    {
+      "epoch": 1.4116284153005465,
+      "grad_norm": 0.9458425745546445,
+      "learning_rate": 0.00011434700436327577,
+      "loss": 12.5859,
+      "step": 8072
+    },
+    {
+      "epoch": 1.4118032786885246,
+      "grad_norm": 0.9786455472519712,
+      "learning_rate": 0.00011432831853875381,
+      "loss": 12.6702,
+      "step": 8073
+    },
+    {
+      "epoch": 1.4119781420765027,
+      "grad_norm": 0.8352619019557761,
+      "learning_rate": 0.00011430963220344402,
+      "loss": 12.6435,
+      "step": 8074
+    },
+    {
+      "epoch": 1.4121530054644809,
+      "grad_norm": 0.952555721311562,
+      "learning_rate": 0.00011429094535801257,
+      "loss": 12.5181,
+      "step": 8075
+    },
+    {
+      "epoch": 1.412327868852459,
+      "grad_norm": 0.9126960482938375,
+      "learning_rate": 0.0001142722580031256,
+      "loss": 12.5624,
+      "step": 8076
+    },
+    {
+      "epoch": 1.412502732240437,
+      "grad_norm": 0.8799934509494677,
+      "learning_rate": 0.00011425357013944933,
+      "loss": 12.8317,
+      "step": 8077
+    },
+    {
+      "epoch": 1.4126775956284154,
+      "grad_norm": 0.8880488154128516,
+      "learning_rate": 0.00011423488176764992,
+      "loss": 12.743,
+      "step": 8078
+    },
+    {
+      "epoch": 1.4128524590163933,
+      "grad_norm": 0.8332363345407132,
+      "learning_rate": 0.00011421619288839363,
+      "loss": 12.5051,
+      "step": 8079
+    },
+    {
+      "epoch": 1.4130273224043717,
+      "grad_norm": 0.8342887622831718,
+      "learning_rate": 0.00011419750350234663,
+      "loss": 12.5122,
+      "step": 8080
+    },
+    {
+      "epoch": 1.4132021857923498,
+      "grad_norm": 0.9269612864315407,
+      "learning_rate": 0.00011417881361017523,
+      "loss": 12.46,
+      "step": 8081
+    },
+    {
+      "epoch": 1.413377049180328,
+      "grad_norm": 0.8104570360999335,
+      "learning_rate": 0.00011416012321254572,
+      "loss": 12.2303,
+      "step": 8082
+    },
+    {
+      "epoch": 1.413551912568306,
+      "grad_norm": 0.9051638559688675,
+      "learning_rate": 0.00011414143231012435,
+      "loss": 12.4973,
+      "step": 8083
+    },
+    {
+      "epoch": 1.4137267759562842,
+      "grad_norm": 0.880279748096964,
+      "learning_rate": 0.00011412274090357744,
+      "loss": 12.661,
+      "step": 8084
+    },
+    {
+      "epoch": 1.4139016393442623,
+      "grad_norm": 0.8802102030472613,
+      "learning_rate": 0.0001141040489935713,
+      "loss": 12.5251,
+      "step": 8085
+    },
+    {
+      "epoch": 1.4140765027322404,
+      "grad_norm": 0.7790676774498854,
+      "learning_rate": 0.00011408535658077233,
+      "loss": 12.3273,
+      "step": 8086
+    },
+    {
+      "epoch": 1.4142513661202185,
+      "grad_norm": 0.8961780556547811,
+      "learning_rate": 0.00011406666366584682,
+      "loss": 12.4995,
+      "step": 8087
+    },
+    {
+      "epoch": 1.4144262295081966,
+      "grad_norm": 0.8486492103471115,
+      "learning_rate": 0.0001140479702494612,
+      "loss": 12.6161,
+      "step": 8088
+    },
+    {
+      "epoch": 1.414601092896175,
+      "grad_norm": 0.8780154614023069,
+      "learning_rate": 0.00011402927633228185,
+      "loss": 12.576,
+      "step": 8089
+    },
+    {
+      "epoch": 1.4147759562841529,
+      "grad_norm": 0.88605365563496,
+      "learning_rate": 0.00011401058191497518,
+      "loss": 12.2518,
+      "step": 8090
+    },
+    {
+      "epoch": 1.4149508196721312,
+      "grad_norm": 0.9646645828088128,
+      "learning_rate": 0.00011399188699820764,
+      "loss": 12.7532,
+      "step": 8091
+    },
+    {
+      "epoch": 1.4151256830601093,
+      "grad_norm": 0.948733824493873,
+      "learning_rate": 0.00011397319158264568,
+      "loss": 12.5499,
+      "step": 8092
+    },
+    {
+      "epoch": 1.4153005464480874,
+      "grad_norm": 0.9261464798679311,
+      "learning_rate": 0.00011395449566895578,
+      "loss": 12.3039,
+      "step": 8093
+    },
+    {
+      "epoch": 1.4154754098360656,
+      "grad_norm": 0.869767531572524,
+      "learning_rate": 0.0001139357992578044,
+      "loss": 12.5532,
+      "step": 8094
+    },
+    {
+      "epoch": 1.4156502732240437,
+      "grad_norm": 0.9951594132063838,
+      "learning_rate": 0.00011391710234985804,
+      "loss": 12.3969,
+      "step": 8095
+    },
+    {
+      "epoch": 1.4158251366120218,
+      "grad_norm": 0.8873689299589889,
+      "learning_rate": 0.00011389840494578327,
+      "loss": 12.686,
+      "step": 8096
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 1.0416450133689998,
+      "learning_rate": 0.0001138797070462466,
+      "loss": 12.6327,
+      "step": 8097
+    },
+    {
+      "epoch": 1.4161748633879783,
+      "grad_norm": 1.0369998279581205,
+      "learning_rate": 0.00011386100865191455,
+      "loss": 12.7018,
+      "step": 8098
+    },
+    {
+      "epoch": 1.4163497267759562,
+      "grad_norm": 0.8854018172506729,
+      "learning_rate": 0.0001138423097634538,
+      "loss": 12.5463,
+      "step": 8099
+    },
+    {
+      "epoch": 1.4165245901639345,
+      "grad_norm": 0.8695860028290375,
+      "learning_rate": 0.00011382361038153084,
+      "loss": 12.6931,
+      "step": 8100
+    },
+    {
+      "epoch": 1.4166994535519126,
+      "grad_norm": 0.9391391732424031,
+      "learning_rate": 0.00011380491050681234,
+      "loss": 12.6103,
+      "step": 8101
+    },
+    {
+      "epoch": 1.4168743169398907,
+      "grad_norm": 0.8943869640561095,
+      "learning_rate": 0.00011378621013996489,
+      "loss": 12.5224,
+      "step": 8102
+    },
+    {
+      "epoch": 1.4170491803278689,
+      "grad_norm": 1.0731185824975769,
+      "learning_rate": 0.00011376750928165519,
+      "loss": 12.6222,
+      "step": 8103
+    },
+    {
+      "epoch": 1.417224043715847,
+      "grad_norm": 0.8585067979650871,
+      "learning_rate": 0.00011374880793254982,
+      "loss": 12.3392,
+      "step": 8104
+    },
+    {
+      "epoch": 1.417398907103825,
+      "grad_norm": 0.9889931554998888,
+      "learning_rate": 0.00011373010609331555,
+      "loss": 12.6585,
+      "step": 8105
+    },
+    {
+      "epoch": 1.4175737704918032,
+      "grad_norm": 0.8575702859788513,
+      "learning_rate": 0.00011371140376461901,
+      "loss": 12.6221,
+      "step": 8106
+    },
+    {
+      "epoch": 1.4177486338797813,
+      "grad_norm": 1.1186855303436487,
+      "learning_rate": 0.00011369270094712696,
+      "loss": 12.5837,
+      "step": 8107
+    },
+    {
+      "epoch": 1.4179234972677595,
+      "grad_norm": 0.8696899256161291,
+      "learning_rate": 0.00011367399764150613,
+      "loss": 12.6469,
+      "step": 8108
+    },
+    {
+      "epoch": 1.4180983606557378,
+      "grad_norm": 0.906135176655088,
+      "learning_rate": 0.00011365529384842326,
+      "loss": 12.4604,
+      "step": 8109
+    },
+    {
+      "epoch": 1.418273224043716,
+      "grad_norm": 0.9265097941885898,
+      "learning_rate": 0.0001136365895685451,
+      "loss": 12.292,
+      "step": 8110
+    },
+    {
+      "epoch": 1.418448087431694,
+      "grad_norm": 0.9094979494982218,
+      "learning_rate": 0.00011361788480253848,
+      "loss": 12.4828,
+      "step": 8111
+    },
+    {
+      "epoch": 1.4186229508196722,
+      "grad_norm": 0.9598045174155673,
+      "learning_rate": 0.00011359917955107017,
+      "loss": 12.5349,
+      "step": 8112
+    },
+    {
+      "epoch": 1.4187978142076503,
+      "grad_norm": 0.8391712764832526,
+      "learning_rate": 0.00011358047381480699,
+      "loss": 12.5061,
+      "step": 8113
+    },
+    {
+      "epoch": 1.4189726775956284,
+      "grad_norm": 0.9077163144172787,
+      "learning_rate": 0.0001135617675944158,
+      "loss": 12.3795,
+      "step": 8114
+    },
+    {
+      "epoch": 1.4191475409836065,
+      "grad_norm": 0.892088336561997,
+      "learning_rate": 0.00011354306089056342,
+      "loss": 12.3122,
+      "step": 8115
+    },
+    {
+      "epoch": 1.4193224043715846,
+      "grad_norm": 0.8501359716928516,
+      "learning_rate": 0.00011352435370391675,
+      "loss": 12.5237,
+      "step": 8116
+    },
+    {
+      "epoch": 1.4194972677595628,
+      "grad_norm": 0.8707332542429412,
+      "learning_rate": 0.00011350564603514264,
+      "loss": 12.6053,
+      "step": 8117
+    },
+    {
+      "epoch": 1.419672131147541,
+      "grad_norm": 0.8092575721569489,
+      "learning_rate": 0.00011348693788490807,
+      "loss": 12.5359,
+      "step": 8118
+    },
+    {
+      "epoch": 1.419846994535519,
+      "grad_norm": 0.887432216929935,
+      "learning_rate": 0.00011346822925387992,
+      "loss": 12.5681,
+      "step": 8119
+    },
+    {
+      "epoch": 1.4200218579234973,
+      "grad_norm": 1.0054672971838376,
+      "learning_rate": 0.00011344952014272515,
+      "loss": 12.6712,
+      "step": 8120
+    },
+    {
+      "epoch": 1.4201967213114755,
+      "grad_norm": 0.807582386456399,
+      "learning_rate": 0.00011343081055211065,
+      "loss": 12.4431,
+      "step": 8121
+    },
+    {
+      "epoch": 1.4203715846994536,
+      "grad_norm": 0.8987103998402867,
+      "learning_rate": 0.00011341210048270347,
+      "loss": 12.7128,
+      "step": 8122
+    },
+    {
+      "epoch": 1.4205464480874317,
+      "grad_norm": 0.8426311608851698,
+      "learning_rate": 0.0001133933899351706,
+      "loss": 12.5331,
+      "step": 8123
+    },
+    {
+      "epoch": 1.4207213114754098,
+      "grad_norm": 0.9525781542291504,
+      "learning_rate": 0.00011337467891017901,
+      "loss": 12.5033,
+      "step": 8124
+    },
+    {
+      "epoch": 1.420896174863388,
+      "grad_norm": 0.8661739777526897,
+      "learning_rate": 0.00011335596740839573,
+      "loss": 12.633,
+      "step": 8125
+    },
+    {
+      "epoch": 1.421071038251366,
+      "grad_norm": 0.9418071119879229,
+      "learning_rate": 0.00011333725543048781,
+      "loss": 12.6478,
+      "step": 8126
+    },
+    {
+      "epoch": 1.4212459016393444,
+      "grad_norm": 0.878991301878814,
+      "learning_rate": 0.00011331854297712237,
+      "loss": 12.6775,
+      "step": 8127
+    },
+    {
+      "epoch": 1.4214207650273223,
+      "grad_norm": 0.8923255968805869,
+      "learning_rate": 0.00011329983004896638,
+      "loss": 12.6989,
+      "step": 8128
+    },
+    {
+      "epoch": 1.4215956284153006,
+      "grad_norm": 0.8899260494189457,
+      "learning_rate": 0.00011328111664668701,
+      "loss": 12.5418,
+      "step": 8129
+    },
+    {
+      "epoch": 1.4217704918032787,
+      "grad_norm": 0.9713595537548666,
+      "learning_rate": 0.00011326240277095134,
+      "loss": 12.3521,
+      "step": 8130
+    },
+    {
+      "epoch": 1.4219453551912569,
+      "grad_norm": 0.843714060885214,
+      "learning_rate": 0.00011324368842242651,
+      "loss": 12.6993,
+      "step": 8131
+    },
+    {
+      "epoch": 1.422120218579235,
+      "grad_norm": 0.9683851818248936,
+      "learning_rate": 0.00011322497360177968,
+      "loss": 12.6151,
+      "step": 8132
+    },
+    {
+      "epoch": 1.422295081967213,
+      "grad_norm": 0.8787080474863871,
+      "learning_rate": 0.000113206258309678,
+      "loss": 12.4742,
+      "step": 8133
+    },
+    {
+      "epoch": 1.4224699453551912,
+      "grad_norm": 0.8489982701288066,
+      "learning_rate": 0.00011318754254678858,
+      "loss": 12.4644,
+      "step": 8134
+    },
+    {
+      "epoch": 1.4226448087431693,
+      "grad_norm": 0.9784225036733697,
+      "learning_rate": 0.00011316882631377872,
+      "loss": 12.5307,
+      "step": 8135
+    },
+    {
+      "epoch": 1.4228196721311475,
+      "grad_norm": 0.8412977267442197,
+      "learning_rate": 0.00011315010961131557,
+      "loss": 12.555,
+      "step": 8136
+    },
+    {
+      "epoch": 1.4229945355191256,
+      "grad_norm": 1.0871964672002001,
+      "learning_rate": 0.0001131313924400664,
+      "loss": 12.6879,
+      "step": 8137
+    },
+    {
+      "epoch": 1.423169398907104,
+      "grad_norm": 0.8889520791226282,
+      "learning_rate": 0.0001131126748006984,
+      "loss": 12.643,
+      "step": 8138
+    },
+    {
+      "epoch": 1.423344262295082,
+      "grad_norm": 1.0082819245200307,
+      "learning_rate": 0.0001130939566938789,
+      "loss": 12.4495,
+      "step": 8139
+    },
+    {
+      "epoch": 1.4235191256830602,
+      "grad_norm": 0.8113140307407997,
+      "learning_rate": 0.00011307523812027514,
+      "loss": 12.5798,
+      "step": 8140
+    },
+    {
+      "epoch": 1.4236939890710383,
+      "grad_norm": 1.0043760850562167,
+      "learning_rate": 0.00011305651908055441,
+      "loss": 12.5781,
+      "step": 8141
+    },
+    {
+      "epoch": 1.4238688524590164,
+      "grad_norm": 0.9315844371016179,
+      "learning_rate": 0.00011303779957538401,
+      "loss": 12.6495,
+      "step": 8142
+    },
+    {
+      "epoch": 1.4240437158469945,
+      "grad_norm": 0.8839042536204116,
+      "learning_rate": 0.00011301907960543129,
+      "loss": 12.635,
+      "step": 8143
+    },
+    {
+      "epoch": 1.4242185792349726,
+      "grad_norm": 0.9551248265761464,
+      "learning_rate": 0.00011300035917136363,
+      "loss": 12.3969,
+      "step": 8144
+    },
+    {
+      "epoch": 1.4243934426229508,
+      "grad_norm": 0.8755619251640476,
+      "learning_rate": 0.0001129816382738483,
+      "loss": 12.5121,
+      "step": 8145
+    },
+    {
+      "epoch": 1.4245683060109289,
+      "grad_norm": 0.9970114196982752,
+      "learning_rate": 0.00011296291691355278,
+      "loss": 12.5627,
+      "step": 8146
+    },
+    {
+      "epoch": 1.4247431693989072,
+      "grad_norm": 0.8885138512797107,
+      "learning_rate": 0.0001129441950911444,
+      "loss": 12.3329,
+      "step": 8147
+    },
+    {
+      "epoch": 1.4249180327868851,
+      "grad_norm": 0.9647084702768811,
+      "learning_rate": 0.0001129254728072906,
+      "loss": 12.7507,
+      "step": 8148
+    },
+    {
+      "epoch": 1.4250928961748635,
+      "grad_norm": 0.7917041491729312,
+      "learning_rate": 0.00011290675006265879,
+      "loss": 12.5387,
+      "step": 8149
+    },
+    {
+      "epoch": 1.4252677595628416,
+      "grad_norm": 0.9277469269809915,
+      "learning_rate": 0.00011288802685791642,
+      "loss": 12.3164,
+      "step": 8150
+    },
+    {
+      "epoch": 1.4254426229508197,
+      "grad_norm": 0.9122711837505145,
+      "learning_rate": 0.00011286930319373095,
+      "loss": 12.5856,
+      "step": 8151
+    },
+    {
+      "epoch": 1.4256174863387978,
+      "grad_norm": 0.838662746495406,
+      "learning_rate": 0.00011285057907076983,
+      "loss": 12.7596,
+      "step": 8152
+    },
+    {
+      "epoch": 1.425792349726776,
+      "grad_norm": 0.8485581821598445,
+      "learning_rate": 0.0001128318544897006,
+      "loss": 12.4258,
+      "step": 8153
+    },
+    {
+      "epoch": 1.425967213114754,
+      "grad_norm": 0.8412854899868941,
+      "learning_rate": 0.00011281312945119075,
+      "loss": 12.4059,
+      "step": 8154
+    },
+    {
+      "epoch": 1.4261420765027322,
+      "grad_norm": 0.8716142596750819,
+      "learning_rate": 0.00011279440395590782,
+      "loss": 12.1698,
+      "step": 8155
+    },
+    {
+      "epoch": 1.4263169398907105,
+      "grad_norm": 0.8863233738357288,
+      "learning_rate": 0.00011277567800451934,
+      "loss": 12.4989,
+      "step": 8156
+    },
+    {
+      "epoch": 1.4264918032786884,
+      "grad_norm": 0.8150124653936976,
+      "learning_rate": 0.00011275695159769285,
+      "loss": 12.7125,
+      "step": 8157
+    },
+    {
+      "epoch": 1.4266666666666667,
+      "grad_norm": 0.8893006270928008,
+      "learning_rate": 0.00011273822473609595,
+      "loss": 12.6586,
+      "step": 8158
+    },
+    {
+      "epoch": 1.4268415300546449,
+      "grad_norm": 0.9261199011314879,
+      "learning_rate": 0.00011271949742039623,
+      "loss": 12.5982,
+      "step": 8159
+    },
+    {
+      "epoch": 1.427016393442623,
+      "grad_norm": 0.8545666038394567,
+      "learning_rate": 0.00011270076965126126,
+      "loss": 12.6697,
+      "step": 8160
+    },
+    {
+      "epoch": 1.427191256830601,
+      "grad_norm": 0.9467677842412786,
+      "learning_rate": 0.0001126820414293587,
+      "loss": 12.3578,
+      "step": 8161
+    },
+    {
+      "epoch": 1.4273661202185792,
+      "grad_norm": 0.8932694300845122,
+      "learning_rate": 0.00011266331275535617,
+      "loss": 12.6269,
+      "step": 8162
+    },
+    {
+      "epoch": 1.4275409836065573,
+      "grad_norm": 0.8898826828530598,
+      "learning_rate": 0.00011264458362992137,
+      "loss": 12.8032,
+      "step": 8163
+    },
+    {
+      "epoch": 1.4277158469945355,
+      "grad_norm": 0.849456279583039,
+      "learning_rate": 0.00011262585405372191,
+      "loss": 12.5669,
+      "step": 8164
+    },
+    {
+      "epoch": 1.4278907103825136,
+      "grad_norm": 0.7930684645090522,
+      "learning_rate": 0.00011260712402742552,
+      "loss": 12.5642,
+      "step": 8165
+    },
+    {
+      "epoch": 1.4280655737704917,
+      "grad_norm": 0.960620333228305,
+      "learning_rate": 0.00011258839355169986,
+      "loss": 12.2933,
+      "step": 8166
+    },
+    {
+      "epoch": 1.42824043715847,
+      "grad_norm": 0.9425512755110634,
+      "learning_rate": 0.00011256966262721272,
+      "loss": 12.6515,
+      "step": 8167
+    },
+    {
+      "epoch": 1.4284153005464482,
+      "grad_norm": 0.8673983999385199,
+      "learning_rate": 0.00011255093125463177,
+      "loss": 12.4937,
+      "step": 8168
+    },
+    {
+      "epoch": 1.4285901639344263,
+      "grad_norm": 0.9056283956727258,
+      "learning_rate": 0.00011253219943462478,
+      "loss": 12.6599,
+      "step": 8169
+    },
+    {
+      "epoch": 1.4287650273224044,
+      "grad_norm": 0.8079345561894602,
+      "learning_rate": 0.00011251346716785953,
+      "loss": 12.6645,
+      "step": 8170
+    },
+    {
+      "epoch": 1.4289398907103825,
+      "grad_norm": 0.9074806210132881,
+      "learning_rate": 0.00011249473445500378,
+      "loss": 12.4554,
+      "step": 8171
+    },
+    {
+      "epoch": 1.4291147540983606,
+      "grad_norm": 0.8607091575427953,
+      "learning_rate": 0.00011247600129672535,
+      "loss": 12.4899,
+      "step": 8172
+    },
+    {
+      "epoch": 1.4292896174863388,
+      "grad_norm": 0.8090554594957733,
+      "learning_rate": 0.0001124572676936921,
+      "loss": 12.5643,
+      "step": 8173
+    },
+    {
+      "epoch": 1.4294644808743169,
+      "grad_norm": 1.008544446125213,
+      "learning_rate": 0.00011243853364657178,
+      "loss": 12.7625,
+      "step": 8174
+    },
+    {
+      "epoch": 1.429639344262295,
+      "grad_norm": 0.9046235709027177,
+      "learning_rate": 0.00011241979915603225,
+      "loss": 12.5869,
+      "step": 8175
+    },
+    {
+      "epoch": 1.4298142076502733,
+      "grad_norm": 0.8176429452438262,
+      "learning_rate": 0.00011240106422274142,
+      "loss": 12.4626,
+      "step": 8176
+    },
+    {
+      "epoch": 1.4299890710382512,
+      "grad_norm": 0.8706958742511044,
+      "learning_rate": 0.0001123823288473671,
+      "loss": 12.6339,
+      "step": 8177
+    },
+    {
+      "epoch": 1.4301639344262296,
+      "grad_norm": 0.8313188824103054,
+      "learning_rate": 0.00011236359303057726,
+      "loss": 12.3302,
+      "step": 8178
+    },
+    {
+      "epoch": 1.4303387978142077,
+      "grad_norm": 0.9163365259054718,
+      "learning_rate": 0.00011234485677303974,
+      "loss": 12.473,
+      "step": 8179
+    },
+    {
+      "epoch": 1.4305136612021858,
+      "grad_norm": 0.9121535286959248,
+      "learning_rate": 0.00011232612007542251,
+      "loss": 12.7717,
+      "step": 8180
+    },
+    {
+      "epoch": 1.430688524590164,
+      "grad_norm": 0.865530707622351,
+      "learning_rate": 0.00011230738293839351,
+      "loss": 12.3854,
+      "step": 8181
+    },
+    {
+      "epoch": 1.430863387978142,
+      "grad_norm": 0.885546754795118,
+      "learning_rate": 0.00011228864536262069,
+      "loss": 12.7556,
+      "step": 8182
+    },
+    {
+      "epoch": 1.4310382513661202,
+      "grad_norm": 0.8354639827496596,
+      "learning_rate": 0.00011226990734877201,
+      "loss": 12.3392,
+      "step": 8183
+    },
+    {
+      "epoch": 1.4312131147540983,
+      "grad_norm": 0.840550636890545,
+      "learning_rate": 0.0001122511688975155,
+      "loss": 12.5408,
+      "step": 8184
+    },
+    {
+      "epoch": 1.4313879781420766,
+      "grad_norm": 0.8623195144122584,
+      "learning_rate": 0.00011223243000951909,
+      "loss": 12.2644,
+      "step": 8185
+    },
+    {
+      "epoch": 1.4315628415300545,
+      "grad_norm": 0.8619104391015985,
+      "learning_rate": 0.00011221369068545087,
+      "loss": 12.5957,
+      "step": 8186
+    },
+    {
+      "epoch": 1.4317377049180329,
+      "grad_norm": 0.9089856323990927,
+      "learning_rate": 0.00011219495092597882,
+      "loss": 12.3645,
+      "step": 8187
+    },
+    {
+      "epoch": 1.431912568306011,
+      "grad_norm": 0.8464108194777812,
+      "learning_rate": 0.00011217621073177105,
+      "loss": 12.6891,
+      "step": 8188
+    },
+    {
+      "epoch": 1.432087431693989,
+      "grad_norm": 0.8720665464461965,
+      "learning_rate": 0.00011215747010349558,
+      "loss": 12.7463,
+      "step": 8189
+    },
+    {
+      "epoch": 1.4322622950819672,
+      "grad_norm": 0.9855385637120933,
+      "learning_rate": 0.0001121387290418205,
+      "loss": 12.7881,
+      "step": 8190
+    },
+    {
+      "epoch": 1.4324371584699453,
+      "grad_norm": 0.7992730078735458,
+      "learning_rate": 0.00011211998754741394,
+      "loss": 12.1977,
+      "step": 8191
+    },
+    {
+      "epoch": 1.4326120218579235,
+      "grad_norm": 1.0143543761663922,
+      "learning_rate": 0.00011210124562094396,
+      "loss": 12.5677,
+      "step": 8192
+    },
+    {
+      "epoch": 1.4327868852459016,
+      "grad_norm": 0.8784635105706641,
+      "learning_rate": 0.00011208250326307871,
+      "loss": 12.5534,
+      "step": 8193
+    },
+    {
+      "epoch": 1.4329617486338797,
+      "grad_norm": 1.0547664821346636,
+      "learning_rate": 0.00011206376047448632,
+      "loss": 12.5903,
+      "step": 8194
+    },
+    {
+      "epoch": 1.4331366120218578,
+      "grad_norm": 0.8373810824129753,
+      "learning_rate": 0.000112045017255835,
+      "loss": 12.6022,
+      "step": 8195
+    },
+    {
+      "epoch": 1.4333114754098362,
+      "grad_norm": 1.0823008054840473,
+      "learning_rate": 0.00011202627360779288,
+      "loss": 12.6379,
+      "step": 8196
+    },
+    {
+      "epoch": 1.4334863387978143,
+      "grad_norm": 0.8953073571946643,
+      "learning_rate": 0.00011200752953102813,
+      "loss": 12.5131,
+      "step": 8197
+    },
+    {
+      "epoch": 1.4336612021857924,
+      "grad_norm": 1.035864540891144,
+      "learning_rate": 0.000111988785026209,
+      "loss": 12.6488,
+      "step": 8198
+    },
+    {
+      "epoch": 1.4338360655737705,
+      "grad_norm": 0.9273510809698687,
+      "learning_rate": 0.00011197004009400369,
+      "loss": 12.4827,
+      "step": 8199
+    },
+    {
+      "epoch": 1.4340109289617486,
+      "grad_norm": 0.8625537402527969,
+      "learning_rate": 0.00011195129473508043,
+      "loss": 12.458,
+      "step": 8200
+    },
+    {
+      "epoch": 1.4341857923497268,
+      "grad_norm": 0.994643528136931,
+      "learning_rate": 0.00011193254895010752,
+      "loss": 12.6414,
+      "step": 8201
+    },
+    {
+      "epoch": 1.4343606557377049,
+      "grad_norm": 0.9148498198696154,
+      "learning_rate": 0.00011191380273975311,
+      "loss": 12.4741,
+      "step": 8202
+    },
+    {
+      "epoch": 1.434535519125683,
+      "grad_norm": 0.9588159778696707,
+      "learning_rate": 0.00011189505610468562,
+      "loss": 12.469,
+      "step": 8203
+    },
+    {
+      "epoch": 1.4347103825136611,
+      "grad_norm": 0.852521097174195,
+      "learning_rate": 0.00011187630904557323,
+      "loss": 12.3513,
+      "step": 8204
+    },
+    {
+      "epoch": 1.4348852459016395,
+      "grad_norm": 0.8887173432461027,
+      "learning_rate": 0.00011185756156308433,
+      "loss": 12.1379,
+      "step": 8205
+    },
+    {
+      "epoch": 1.4350601092896174,
+      "grad_norm": 0.8224277612608998,
+      "learning_rate": 0.0001118388136578872,
+      "loss": 12.2075,
+      "step": 8206
+    },
+    {
+      "epoch": 1.4352349726775957,
+      "grad_norm": 0.865531875712152,
+      "learning_rate": 0.00011182006533065017,
+      "loss": 12.0202,
+      "step": 8207
+    },
+    {
+      "epoch": 1.4354098360655738,
+      "grad_norm": 1.0292193144743382,
+      "learning_rate": 0.00011180131658204166,
+      "loss": 12.6322,
+      "step": 8208
+    },
+    {
+      "epoch": 1.435584699453552,
+      "grad_norm": 0.9628094273519001,
+      "learning_rate": 0.00011178256741272998,
+      "loss": 12.4483,
+      "step": 8209
+    },
+    {
+      "epoch": 1.43575956284153,
+      "grad_norm": 0.9118800619087868,
+      "learning_rate": 0.00011176381782338357,
+      "loss": 12.6018,
+      "step": 8210
+    },
+    {
+      "epoch": 1.4359344262295082,
+      "grad_norm": 0.9058832517017643,
+      "learning_rate": 0.00011174506781467076,
+      "loss": 12.481,
+      "step": 8211
+    },
+    {
+      "epoch": 1.4361092896174863,
+      "grad_norm": 1.098464629667907,
+      "learning_rate": 0.00011172631738726002,
+      "loss": 12.7587,
+      "step": 8212
+    },
+    {
+      "epoch": 1.4362841530054644,
+      "grad_norm": 0.8418175935565474,
+      "learning_rate": 0.00011170756654181977,
+      "loss": 12.5446,
+      "step": 8213
+    },
+    {
+      "epoch": 1.4364590163934428,
+      "grad_norm": 1.020115459012143,
+      "learning_rate": 0.00011168881527901844,
+      "loss": 12.7111,
+      "step": 8214
+    },
+    {
+      "epoch": 1.4366338797814207,
+      "grad_norm": 0.9263566845351348,
+      "learning_rate": 0.00011167006359952452,
+      "loss": 12.3483,
+      "step": 8215
+    },
+    {
+      "epoch": 1.436808743169399,
+      "grad_norm": 0.9917080140934179,
+      "learning_rate": 0.00011165131150400645,
+      "loss": 12.5197,
+      "step": 8216
+    },
+    {
+      "epoch": 1.4369836065573771,
+      "grad_norm": 1.086867199808763,
+      "learning_rate": 0.00011163255899313272,
+      "loss": 12.4978,
+      "step": 8217
+    },
+    {
+      "epoch": 1.4371584699453552,
+      "grad_norm": 0.9523060185667697,
+      "learning_rate": 0.0001116138060675719,
+      "loss": 12.7009,
+      "step": 8218
+    },
+    {
+      "epoch": 1.4373333333333334,
+      "grad_norm": 1.1088335457152076,
+      "learning_rate": 0.00011159505272799244,
+      "loss": 12.5119,
+      "step": 8219
+    },
+    {
+      "epoch": 1.4375081967213115,
+      "grad_norm": 0.8762656149552628,
+      "learning_rate": 0.0001115762989750629,
+      "loss": 12.3886,
+      "step": 8220
+    },
+    {
+      "epoch": 1.4376830601092896,
+      "grad_norm": 1.2234613927688245,
+      "learning_rate": 0.00011155754480945184,
+      "loss": 12.4305,
+      "step": 8221
+    },
+    {
+      "epoch": 1.4378579234972677,
+      "grad_norm": 0.8980467025187463,
+      "learning_rate": 0.00011153879023182778,
+      "loss": 12.5637,
+      "step": 8222
+    },
+    {
+      "epoch": 1.4380327868852458,
+      "grad_norm": 1.162253333624586,
+      "learning_rate": 0.00011152003524285936,
+      "loss": 12.5476,
+      "step": 8223
+    },
+    {
+      "epoch": 1.438207650273224,
+      "grad_norm": 1.0390690034387178,
+      "learning_rate": 0.00011150127984321512,
+      "loss": 12.5199,
+      "step": 8224
+    },
+    {
+      "epoch": 1.4383825136612023,
+      "grad_norm": 0.9413391021756108,
+      "learning_rate": 0.0001114825240335637,
+      "loss": 12.3318,
+      "step": 8225
+    },
+    {
+      "epoch": 1.4385573770491804,
+      "grad_norm": 1.0870461410548735,
+      "learning_rate": 0.00011146376781457371,
+      "loss": 12.6965,
+      "step": 8226
+    },
+    {
+      "epoch": 1.4387322404371585,
+      "grad_norm": 0.8648061271779185,
+      "learning_rate": 0.00011144501118691382,
+      "loss": 12.7017,
+      "step": 8227
+    },
+    {
+      "epoch": 1.4389071038251366,
+      "grad_norm": 1.0314497856138465,
+      "learning_rate": 0.0001114262541512526,
+      "loss": 12.3038,
+      "step": 8228
+    },
+    {
+      "epoch": 1.4390819672131148,
+      "grad_norm": 1.0172373501562404,
+      "learning_rate": 0.00011140749670825883,
+      "loss": 12.498,
+      "step": 8229
+    },
+    {
+      "epoch": 1.4392568306010929,
+      "grad_norm": 0.9535577373178917,
+      "learning_rate": 0.00011138873885860108,
+      "loss": 12.5782,
+      "step": 8230
+    },
+    {
+      "epoch": 1.439431693989071,
+      "grad_norm": 0.9639143976571551,
+      "learning_rate": 0.00011136998060294814,
+      "loss": 12.3886,
+      "step": 8231
+    },
+    {
+      "epoch": 1.4396065573770491,
+      "grad_norm": 0.8468447687811376,
+      "learning_rate": 0.00011135122194196863,
+      "loss": 12.2588,
+      "step": 8232
+    },
+    {
+      "epoch": 1.4397814207650272,
+      "grad_norm": 0.963574047260045,
+      "learning_rate": 0.00011133246287633136,
+      "loss": 12.5079,
+      "step": 8233
+    },
+    {
+      "epoch": 1.4399562841530056,
+      "grad_norm": 0.8722905010933919,
+      "learning_rate": 0.00011131370340670502,
+      "loss": 12.534,
+      "step": 8234
+    },
+    {
+      "epoch": 1.4401311475409835,
+      "grad_norm": 1.0143719596891065,
+      "learning_rate": 0.00011129494353375834,
+      "loss": 12.5732,
+      "step": 8235
+    },
+    {
+      "epoch": 1.4403060109289618,
+      "grad_norm": 0.8855957106966378,
+      "learning_rate": 0.00011127618325816016,
+      "loss": 12.3555,
+      "step": 8236
+    },
+    {
+      "epoch": 1.44048087431694,
+      "grad_norm": 0.8524775169312052,
+      "learning_rate": 0.00011125742258057918,
+      "loss": 12.6597,
+      "step": 8237
+    },
+    {
+      "epoch": 1.440655737704918,
+      "grad_norm": 1.0397570052909293,
+      "learning_rate": 0.00011123866150168429,
+      "loss": 12.5388,
+      "step": 8238
+    },
+    {
+      "epoch": 1.4408306010928962,
+      "grad_norm": 0.8334926950162171,
+      "learning_rate": 0.00011121990002214422,
+      "loss": 12.536,
+      "step": 8239
+    },
+    {
+      "epoch": 1.4410054644808743,
+      "grad_norm": 1.1174548511968991,
+      "learning_rate": 0.00011120113814262782,
+      "loss": 12.6025,
+      "step": 8240
+    },
+    {
+      "epoch": 1.4411803278688524,
+      "grad_norm": 0.9253796986676236,
+      "learning_rate": 0.00011118237586380393,
+      "loss": 12.4794,
+      "step": 8241
+    },
+    {
+      "epoch": 1.4413551912568305,
+      "grad_norm": 0.9572104165795813,
+      "learning_rate": 0.0001111636131863414,
+      "loss": 12.7264,
+      "step": 8242
+    },
+    {
+      "epoch": 1.4415300546448089,
+      "grad_norm": 0.8513732140973793,
+      "learning_rate": 0.00011114485011090912,
+      "loss": 12.5161,
+      "step": 8243
+    },
+    {
+      "epoch": 1.4417049180327868,
+      "grad_norm": 0.8641349051782954,
+      "learning_rate": 0.00011112608663817597,
+      "loss": 12.6944,
+      "step": 8244
+    },
+    {
+      "epoch": 1.4418797814207651,
+      "grad_norm": 0.8737551038462615,
+      "learning_rate": 0.00011110732276881079,
+      "loss": 12.3615,
+      "step": 8245
+    },
+    {
+      "epoch": 1.4420546448087432,
+      "grad_norm": 0.8603169201343597,
+      "learning_rate": 0.00011108855850348256,
+      "loss": 12.6157,
+      "step": 8246
+    },
+    {
+      "epoch": 1.4422295081967214,
+      "grad_norm": 0.8911518247362692,
+      "learning_rate": 0.00011106979384286016,
+      "loss": 12.4412,
+      "step": 8247
+    },
+    {
+      "epoch": 1.4424043715846995,
+      "grad_norm": 0.8812362134467291,
+      "learning_rate": 0.00011105102878761257,
+      "loss": 12.781,
+      "step": 8248
+    },
+    {
+      "epoch": 1.4425792349726776,
+      "grad_norm": 0.9229004319211899,
+      "learning_rate": 0.00011103226333840869,
+      "loss": 12.4725,
+      "step": 8249
+    },
+    {
+      "epoch": 1.4427540983606557,
+      "grad_norm": 0.8631472908801788,
+      "learning_rate": 0.00011101349749591754,
+      "loss": 12.4528,
+      "step": 8250
+    },
+    {
+      "epoch": 1.4429289617486338,
+      "grad_norm": 0.881912666312339,
+      "learning_rate": 0.00011099473126080804,
+      "loss": 12.5944,
+      "step": 8251
+    },
+    {
+      "epoch": 1.443103825136612,
+      "grad_norm": 0.819686831880575,
+      "learning_rate": 0.00011097596463374922,
+      "loss": 12.6318,
+      "step": 8252
+    },
+    {
+      "epoch": 1.44327868852459,
+      "grad_norm": 0.8948145735881394,
+      "learning_rate": 0.00011095719761541011,
+      "loss": 12.6346,
+      "step": 8253
+    },
+    {
+      "epoch": 1.4434535519125684,
+      "grad_norm": 0.8683615319967234,
+      "learning_rate": 0.00011093843020645966,
+      "loss": 12.2992,
+      "step": 8254
+    },
+    {
+      "epoch": 1.4436284153005465,
+      "grad_norm": 0.864552973296335,
+      "learning_rate": 0.00011091966240756699,
+      "loss": 12.5541,
+      "step": 8255
+    },
+    {
+      "epoch": 1.4438032786885246,
+      "grad_norm": 0.9488267887328384,
+      "learning_rate": 0.00011090089421940111,
+      "loss": 12.6688,
+      "step": 8256
+    },
+    {
+      "epoch": 1.4439781420765028,
+      "grad_norm": 0.819067512832053,
+      "learning_rate": 0.00011088212564263109,
+      "loss": 12.4009,
+      "step": 8257
+    },
+    {
+      "epoch": 1.4441530054644809,
+      "grad_norm": 0.9251125038055008,
+      "learning_rate": 0.00011086335667792598,
+      "loss": 12.4906,
+      "step": 8258
+    },
+    {
+      "epoch": 1.444327868852459,
+      "grad_norm": 0.9948136215774953,
+      "learning_rate": 0.00011084458732595493,
+      "loss": 12.3306,
+      "step": 8259
+    },
+    {
+      "epoch": 1.4445027322404371,
+      "grad_norm": 0.8331789811338214,
+      "learning_rate": 0.00011082581758738699,
+      "loss": 12.6098,
+      "step": 8260
+    },
+    {
+      "epoch": 1.4446775956284152,
+      "grad_norm": 0.840957711582271,
+      "learning_rate": 0.0001108070474628913,
+      "loss": 12.4867,
+      "step": 8261
+    },
+    {
+      "epoch": 1.4448524590163934,
+      "grad_norm": 0.9021658067137824,
+      "learning_rate": 0.000110788276953137,
+      "loss": 12.8783,
+      "step": 8262
+    },
+    {
+      "epoch": 1.4450273224043717,
+      "grad_norm": 0.8410526962843238,
+      "learning_rate": 0.00011076950605879324,
+      "loss": 12.5104,
+      "step": 8263
+    },
+    {
+      "epoch": 1.4452021857923496,
+      "grad_norm": 0.8768022095889711,
+      "learning_rate": 0.00011075073478052918,
+      "loss": 12.4708,
+      "step": 8264
+    },
+    {
+      "epoch": 1.445377049180328,
+      "grad_norm": 0.8939099428947376,
+      "learning_rate": 0.00011073196311901397,
+      "loss": 12.5075,
+      "step": 8265
+    },
+    {
+      "epoch": 1.445551912568306,
+      "grad_norm": 0.9595791401625436,
+      "learning_rate": 0.0001107131910749168,
+      "loss": 12.7381,
+      "step": 8266
+    },
+    {
+      "epoch": 1.4457267759562842,
+      "grad_norm": 0.8478735512023919,
+      "learning_rate": 0.00011069441864890693,
+      "loss": 12.2274,
+      "step": 8267
+    },
+    {
+      "epoch": 1.4459016393442623,
+      "grad_norm": 0.848566399955693,
+      "learning_rate": 0.0001106756458416535,
+      "loss": 12.2352,
+      "step": 8268
+    },
+    {
+      "epoch": 1.4460765027322404,
+      "grad_norm": 0.8858879268337193,
+      "learning_rate": 0.00011065687265382577,
+      "loss": 12.6805,
+      "step": 8269
+    },
+    {
+      "epoch": 1.4462513661202185,
+      "grad_norm": 0.9299784676550779,
+      "learning_rate": 0.00011063809908609298,
+      "loss": 12.4298,
+      "step": 8270
+    },
+    {
+      "epoch": 1.4464262295081967,
+      "grad_norm": 0.882527242103597,
+      "learning_rate": 0.00011061932513912436,
+      "loss": 12.527,
+      "step": 8271
+    },
+    {
+      "epoch": 1.446601092896175,
+      "grad_norm": 0.8633010835544755,
+      "learning_rate": 0.00011060055081358927,
+      "loss": 12.5658,
+      "step": 8272
+    },
+    {
+      "epoch": 1.446775956284153,
+      "grad_norm": 0.9871081209097997,
+      "learning_rate": 0.00011058177611015687,
+      "loss": 12.4277,
+      "step": 8273
+    },
+    {
+      "epoch": 1.4469508196721312,
+      "grad_norm": 0.8385530694950053,
+      "learning_rate": 0.00011056300102949655,
+      "loss": 12.4792,
+      "step": 8274
+    },
+    {
+      "epoch": 1.4471256830601094,
+      "grad_norm": 1.207726593023715,
+      "learning_rate": 0.00011054422557227758,
+      "loss": 12.7587,
+      "step": 8275
+    },
+    {
+      "epoch": 1.4473005464480875,
+      "grad_norm": 0.9301614174671998,
+      "learning_rate": 0.00011052544973916928,
+      "loss": 12.6298,
+      "step": 8276
+    },
+    {
+      "epoch": 1.4474754098360656,
+      "grad_norm": 0.9591657465385437,
+      "learning_rate": 0.00011050667353084099,
+      "loss": 12.6385,
+      "step": 8277
+    },
+    {
+      "epoch": 1.4476502732240437,
+      "grad_norm": 0.8997375612444214,
+      "learning_rate": 0.00011048789694796209,
+      "loss": 12.7008,
+      "step": 8278
+    },
+    {
+      "epoch": 1.4478251366120218,
+      "grad_norm": 0.8993016280826812,
+      "learning_rate": 0.00011046911999120189,
+      "loss": 12.5011,
+      "step": 8279
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 0.9512894169031425,
+      "learning_rate": 0.00011045034266122978,
+      "loss": 12.682,
+      "step": 8280
+    },
+    {
+      "epoch": 1.448174863387978,
+      "grad_norm": 0.8164913515025698,
+      "learning_rate": 0.00011043156495871519,
+      "loss": 12.6227,
+      "step": 8281
+    },
+    {
+      "epoch": 1.4483497267759562,
+      "grad_norm": 0.8648224228131356,
+      "learning_rate": 0.0001104127868843275,
+      "loss": 12.5501,
+      "step": 8282
+    },
+    {
+      "epoch": 1.4485245901639345,
+      "grad_norm": 0.9032181688341264,
+      "learning_rate": 0.00011039400843873613,
+      "loss": 12.552,
+      "step": 8283
+    },
+    {
+      "epoch": 1.4486994535519124,
+      "grad_norm": 1.077126157498355,
+      "learning_rate": 0.00011037522962261049,
+      "loss": 12.6611,
+      "step": 8284
+    },
+    {
+      "epoch": 1.4488743169398908,
+      "grad_norm": 0.9499391515425297,
+      "learning_rate": 0.00011035645043662004,
+      "loss": 12.5804,
+      "step": 8285
+    },
+    {
+      "epoch": 1.449049180327869,
+      "grad_norm": 1.0910985340960078,
+      "learning_rate": 0.00011033767088143423,
+      "loss": 12.5586,
+      "step": 8286
+    },
+    {
+      "epoch": 1.449224043715847,
+      "grad_norm": 0.9517592099086474,
+      "learning_rate": 0.0001103188909577225,
+      "loss": 12.6571,
+      "step": 8287
+    },
+    {
+      "epoch": 1.4493989071038251,
+      "grad_norm": 0.9407599997414418,
+      "learning_rate": 0.00011030011066615439,
+      "loss": 12.7254,
+      "step": 8288
+    },
+    {
+      "epoch": 1.4495737704918032,
+      "grad_norm": 1.1809029818754755,
+      "learning_rate": 0.0001102813300073994,
+      "loss": 12.5729,
+      "step": 8289
+    },
+    {
+      "epoch": 1.4497486338797814,
+      "grad_norm": 0.905815551367805,
+      "learning_rate": 0.00011026254898212695,
+      "loss": 12.5142,
+      "step": 8290
+    },
+    {
+      "epoch": 1.4499234972677595,
+      "grad_norm": 1.1397408662529789,
+      "learning_rate": 0.00011024376759100667,
+      "loss": 12.7431,
+      "step": 8291
+    },
+    {
+      "epoch": 1.4500983606557378,
+      "grad_norm": 0.8870999538505454,
+      "learning_rate": 0.00011022498583470802,
+      "loss": 12.6536,
+      "step": 8292
+    },
+    {
+      "epoch": 1.4502732240437157,
+      "grad_norm": 1.3249077260956008,
+      "learning_rate": 0.00011020620371390057,
+      "loss": 12.1322,
+      "step": 8293
+    },
+    {
+      "epoch": 1.450448087431694,
+      "grad_norm": 0.8636486094369135,
+      "learning_rate": 0.00011018742122925389,
+      "loss": 12.7748,
+      "step": 8294
+    },
+    {
+      "epoch": 1.4506229508196722,
+      "grad_norm": 1.1487385568570287,
+      "learning_rate": 0.00011016863838143756,
+      "loss": 12.6511,
+      "step": 8295
+    },
+    {
+      "epoch": 1.4507978142076503,
+      "grad_norm": 0.9414045628427169,
+      "learning_rate": 0.00011014985517112112,
+      "loss": 12.4471,
+      "step": 8296
+    },
+    {
+      "epoch": 1.4509726775956284,
+      "grad_norm": 0.9838340604196542,
+      "learning_rate": 0.00011013107159897422,
+      "loss": 12.7347,
+      "step": 8297
+    },
+    {
+      "epoch": 1.4511475409836065,
+      "grad_norm": 1.0583932785739791,
+      "learning_rate": 0.00011011228766566645,
+      "loss": 12.6628,
+      "step": 8298
+    },
+    {
+      "epoch": 1.4513224043715847,
+      "grad_norm": 0.8372812598627761,
+      "learning_rate": 0.00011009350337186744,
+      "loss": 12.4661,
+      "step": 8299
+    },
+    {
+      "epoch": 1.4514972677595628,
+      "grad_norm": 1.24788680139167,
+      "learning_rate": 0.00011007471871824684,
+      "loss": 12.4837,
+      "step": 8300
+    },
+    {
+      "epoch": 1.451672131147541,
+      "grad_norm": 1.0538888259431491,
+      "learning_rate": 0.00011005593370547427,
+      "loss": 12.4979,
+      "step": 8301
+    },
+    {
+      "epoch": 1.451846994535519,
+      "grad_norm": 0.9740096376199042,
+      "learning_rate": 0.00011003714833421944,
+      "loss": 12.6322,
+      "step": 8302
+    },
+    {
+      "epoch": 1.4520218579234974,
+      "grad_norm": 0.940332678858345,
+      "learning_rate": 0.00011001836260515198,
+      "loss": 12.3814,
+      "step": 8303
+    },
+    {
+      "epoch": 1.4521967213114755,
+      "grad_norm": 0.8207806511136034,
+      "learning_rate": 0.0001099995765189416,
+      "loss": 12.4309,
+      "step": 8304
+    },
+    {
+      "epoch": 1.4523715846994536,
+      "grad_norm": 1.0809907884783445,
+      "learning_rate": 0.00010998079007625798,
+      "loss": 12.5418,
+      "step": 8305
+    },
+    {
+      "epoch": 1.4525464480874317,
+      "grad_norm": 0.9392384167425176,
+      "learning_rate": 0.00010996200327777089,
+      "loss": 12.6477,
+      "step": 8306
+    },
+    {
+      "epoch": 1.4527213114754098,
+      "grad_norm": 0.9636199423495737,
+      "learning_rate": 0.00010994321612414999,
+      "loss": 12.5828,
+      "step": 8307
+    },
+    {
+      "epoch": 1.452896174863388,
+      "grad_norm": 0.9370880846330513,
+      "learning_rate": 0.00010992442861606507,
+      "loss": 12.6829,
+      "step": 8308
+    },
+    {
+      "epoch": 1.453071038251366,
+      "grad_norm": 0.9599538139134756,
+      "learning_rate": 0.00010990564075418586,
+      "loss": 12.4907,
+      "step": 8309
+    },
+    {
+      "epoch": 1.4532459016393442,
+      "grad_norm": 0.9713423812723952,
+      "learning_rate": 0.00010988685253918214,
+      "loss": 12.5782,
+      "step": 8310
+    },
+    {
+      "epoch": 1.4534207650273223,
+      "grad_norm": 0.8936544389131059,
+      "learning_rate": 0.00010986806397172367,
+      "loss": 12.649,
+      "step": 8311
+    },
+    {
+      "epoch": 1.4535956284153007,
+      "grad_norm": 0.9934255537813076,
+      "learning_rate": 0.00010984927505248026,
+      "loss": 12.6352,
+      "step": 8312
+    },
+    {
+      "epoch": 1.4537704918032786,
+      "grad_norm": 0.8950697879233503,
+      "learning_rate": 0.00010983048578212169,
+      "loss": 12.2237,
+      "step": 8313
+    },
+    {
+      "epoch": 1.453945355191257,
+      "grad_norm": 0.9124869221377818,
+      "learning_rate": 0.00010981169616131781,
+      "loss": 12.246,
+      "step": 8314
+    },
+    {
+      "epoch": 1.454120218579235,
+      "grad_norm": 0.7608855036135511,
+      "learning_rate": 0.00010979290619073838,
+      "loss": 12.6155,
+      "step": 8315
+    },
+    {
+      "epoch": 1.4542950819672131,
+      "grad_norm": 0.8822364265342264,
+      "learning_rate": 0.00010977411587105334,
+      "loss": 12.621,
+      "step": 8316
+    },
+    {
+      "epoch": 1.4544699453551913,
+      "grad_norm": 0.9304041245721779,
+      "learning_rate": 0.00010975532520293246,
+      "loss": 12.3976,
+      "step": 8317
+    },
+    {
+      "epoch": 1.4546448087431694,
+      "grad_norm": 0.8388511111270569,
+      "learning_rate": 0.00010973653418704561,
+      "loss": 12.5814,
+      "step": 8318
+    },
+    {
+      "epoch": 1.4548196721311475,
+      "grad_norm": 0.926682652740001,
+      "learning_rate": 0.00010971774282406275,
+      "loss": 12.5571,
+      "step": 8319
+    },
+    {
+      "epoch": 1.4549945355191256,
+      "grad_norm": 0.9423391611055525,
+      "learning_rate": 0.00010969895111465368,
+      "loss": 12.4701,
+      "step": 8320
+    },
+    {
+      "epoch": 1.455169398907104,
+      "grad_norm": 0.8827089892967771,
+      "learning_rate": 0.00010968015905948834,
+      "loss": 12.4893,
+      "step": 8321
+    },
+    {
+      "epoch": 1.4553442622950818,
+      "grad_norm": 1.1238089724076248,
+      "learning_rate": 0.00010966136665923663,
+      "loss": 12.6516,
+      "step": 8322
+    },
+    {
+      "epoch": 1.4555191256830602,
+      "grad_norm": 0.8113634716756719,
+      "learning_rate": 0.00010964257391456849,
+      "loss": 12.4725,
+      "step": 8323
+    },
+    {
+      "epoch": 1.4556939890710383,
+      "grad_norm": 0.9456919367927055,
+      "learning_rate": 0.00010962378082615386,
+      "loss": 12.2335,
+      "step": 8324
+    },
+    {
+      "epoch": 1.4558688524590164,
+      "grad_norm": 0.9074325671028495,
+      "learning_rate": 0.00010960498739466267,
+      "loss": 12.4915,
+      "step": 8325
+    },
+    {
+      "epoch": 1.4560437158469945,
+      "grad_norm": 0.839297662437774,
+      "learning_rate": 0.00010958619362076492,
+      "loss": 12.5931,
+      "step": 8326
+    },
+    {
+      "epoch": 1.4562185792349727,
+      "grad_norm": 0.9263992138869221,
+      "learning_rate": 0.00010956739950513057,
+      "loss": 12.2145,
+      "step": 8327
+    },
+    {
+      "epoch": 1.4563934426229508,
+      "grad_norm": 0.8061151975063716,
+      "learning_rate": 0.00010954860504842959,
+      "loss": 12.391,
+      "step": 8328
+    },
+    {
+      "epoch": 1.456568306010929,
+      "grad_norm": 0.8485929792187269,
+      "learning_rate": 0.00010952981025133203,
+      "loss": 12.7061,
+      "step": 8329
+    },
+    {
+      "epoch": 1.456743169398907,
+      "grad_norm": 0.8700281173163149,
+      "learning_rate": 0.00010951101511450786,
+      "loss": 12.6877,
+      "step": 8330
+    },
+    {
+      "epoch": 1.4569180327868851,
+      "grad_norm": 0.8263079246603727,
+      "learning_rate": 0.00010949221963862708,
+      "loss": 12.5506,
+      "step": 8331
+    },
+    {
+      "epoch": 1.4570928961748635,
+      "grad_norm": 0.8881703621171391,
+      "learning_rate": 0.00010947342382435979,
+      "loss": 12.537,
+      "step": 8332
+    },
+    {
+      "epoch": 1.4572677595628416,
+      "grad_norm": 0.8350186984458714,
+      "learning_rate": 0.00010945462767237597,
+      "loss": 12.417,
+      "step": 8333
+    },
+    {
+      "epoch": 1.4574426229508197,
+      "grad_norm": 0.969874070210571,
+      "learning_rate": 0.00010943583118334577,
+      "loss": 12.4114,
+      "step": 8334
+    },
+    {
+      "epoch": 1.4576174863387978,
+      "grad_norm": 0.8870740817991929,
+      "learning_rate": 0.0001094170343579392,
+      "loss": 12.4238,
+      "step": 8335
+    },
+    {
+      "epoch": 1.457792349726776,
+      "grad_norm": 0.7989623105675551,
+      "learning_rate": 0.00010939823719682636,
+      "loss": 12.3658,
+      "step": 8336
+    },
+    {
+      "epoch": 1.457967213114754,
+      "grad_norm": 0.8640957567273615,
+      "learning_rate": 0.00010937943970067733,
+      "loss": 12.5414,
+      "step": 8337
+    },
+    {
+      "epoch": 1.4581420765027322,
+      "grad_norm": 0.9310863489737742,
+      "learning_rate": 0.00010936064187016224,
+      "loss": 12.7395,
+      "step": 8338
+    },
+    {
+      "epoch": 1.4583169398907103,
+      "grad_norm": 0.9198421988632046,
+      "learning_rate": 0.0001093418437059512,
+      "loss": 12.5124,
+      "step": 8339
+    },
+    {
+      "epoch": 1.4584918032786884,
+      "grad_norm": 0.8572937528113912,
+      "learning_rate": 0.00010932304520871435,
+      "loss": 12.6384,
+      "step": 8340
+    },
+    {
+      "epoch": 1.4586666666666668,
+      "grad_norm": 0.8217174873943454,
+      "learning_rate": 0.00010930424637912183,
+      "loss": 12.7232,
+      "step": 8341
+    },
+    {
+      "epoch": 1.4588415300546447,
+      "grad_norm": 0.8412113872271085,
+      "learning_rate": 0.00010928544721784377,
+      "loss": 12.6029,
+      "step": 8342
+    },
+    {
+      "epoch": 1.459016393442623,
+      "grad_norm": 0.8608998877578838,
+      "learning_rate": 0.0001092666477255504,
+      "loss": 12.3506,
+      "step": 8343
+    },
+    {
+      "epoch": 1.4591912568306011,
+      "grad_norm": 0.9611375595229498,
+      "learning_rate": 0.00010924784790291187,
+      "loss": 12.541,
+      "step": 8344
+    },
+    {
+      "epoch": 1.4593661202185793,
+      "grad_norm": 0.8883472253391482,
+      "learning_rate": 0.00010922904775059836,
+      "loss": 12.4942,
+      "step": 8345
+    },
+    {
+      "epoch": 1.4595409836065574,
+      "grad_norm": 0.9677212001974193,
+      "learning_rate": 0.00010921024726928009,
+      "loss": 12.5542,
+      "step": 8346
+    },
+    {
+      "epoch": 1.4597158469945355,
+      "grad_norm": 0.7549707358262479,
+      "learning_rate": 0.00010919144645962725,
+      "loss": 12.2712,
+      "step": 8347
+    },
+    {
+      "epoch": 1.4598907103825136,
+      "grad_norm": 0.9586344422888592,
+      "learning_rate": 0.00010917264532231008,
+      "loss": 12.4336,
+      "step": 8348
+    },
+    {
+      "epoch": 1.4600655737704917,
+      "grad_norm": 0.8499994044551441,
+      "learning_rate": 0.00010915384385799882,
+      "loss": 12.6123,
+      "step": 8349
+    },
+    {
+      "epoch": 1.46024043715847,
+      "grad_norm": 0.8259006097030974,
+      "learning_rate": 0.00010913504206736373,
+      "loss": 12.6529,
+      "step": 8350
+    },
+    {
+      "epoch": 1.460415300546448,
+      "grad_norm": 0.8598112460795396,
+      "learning_rate": 0.00010911623995107504,
+      "loss": 12.3972,
+      "step": 8351
+    },
+    {
+      "epoch": 1.4605901639344263,
+      "grad_norm": 0.8354845128317215,
+      "learning_rate": 0.00010909743750980307,
+      "loss": 12.3156,
+      "step": 8352
+    },
+    {
+      "epoch": 1.4607650273224044,
+      "grad_norm": 0.9187121107126787,
+      "learning_rate": 0.00010907863474421811,
+      "loss": 12.7363,
+      "step": 8353
+    },
+    {
+      "epoch": 1.4609398907103825,
+      "grad_norm": 0.807361724969711,
+      "learning_rate": 0.00010905983165499039,
+      "loss": 12.5207,
+      "step": 8354
+    },
+    {
+      "epoch": 1.4611147540983607,
+      "grad_norm": 0.9460465585125636,
+      "learning_rate": 0.00010904102824279028,
+      "loss": 12.4558,
+      "step": 8355
+    },
+    {
+      "epoch": 1.4612896174863388,
+      "grad_norm": 0.942667279126698,
+      "learning_rate": 0.00010902222450828805,
+      "loss": 12.3674,
+      "step": 8356
+    },
+    {
+      "epoch": 1.461464480874317,
+      "grad_norm": 0.8727837494113009,
+      "learning_rate": 0.00010900342045215408,
+      "loss": 12.6874,
+      "step": 8357
+    },
+    {
+      "epoch": 1.461639344262295,
+      "grad_norm": 0.9198510356093967,
+      "learning_rate": 0.00010898461607505867,
+      "loss": 12.2881,
+      "step": 8358
+    },
+    {
+      "epoch": 1.4618142076502731,
+      "grad_norm": 0.898817003902528,
+      "learning_rate": 0.00010896581137767221,
+      "loss": 12.6092,
+      "step": 8359
+    },
+    {
+      "epoch": 1.4619890710382513,
+      "grad_norm": 0.9649184774900371,
+      "learning_rate": 0.00010894700636066503,
+      "loss": 12.4885,
+      "step": 8360
+    },
+    {
+      "epoch": 1.4621639344262296,
+      "grad_norm": 0.9449004961637909,
+      "learning_rate": 0.00010892820102470755,
+      "loss": 12.4816,
+      "step": 8361
+    },
+    {
+      "epoch": 1.4623387978142077,
+      "grad_norm": 0.8401469844454763,
+      "learning_rate": 0.00010890939537047014,
+      "loss": 12.5023,
+      "step": 8362
+    },
+    {
+      "epoch": 1.4625136612021858,
+      "grad_norm": 0.875713561379008,
+      "learning_rate": 0.00010889058939862315,
+      "loss": 12.4536,
+      "step": 8363
+    },
+    {
+      "epoch": 1.462688524590164,
+      "grad_norm": 0.9361511554677743,
+      "learning_rate": 0.0001088717831098371,
+      "loss": 12.7189,
+      "step": 8364
+    },
+    {
+      "epoch": 1.462863387978142,
+      "grad_norm": 0.8652180914513555,
+      "learning_rate": 0.0001088529765047823,
+      "loss": 12.6368,
+      "step": 8365
+    },
+    {
+      "epoch": 1.4630382513661202,
+      "grad_norm": 0.9642940122432373,
+      "learning_rate": 0.00010883416958412926,
+      "loss": 12.4741,
+      "step": 8366
+    },
+    {
+      "epoch": 1.4632131147540983,
+      "grad_norm": 0.9336441313181538,
+      "learning_rate": 0.00010881536234854836,
+      "loss": 12.5181,
+      "step": 8367
+    },
+    {
+      "epoch": 1.4633879781420764,
+      "grad_norm": 0.9491830618716803,
+      "learning_rate": 0.00010879655479871014,
+      "loss": 12.6859,
+      "step": 8368
+    },
+    {
+      "epoch": 1.4635628415300546,
+      "grad_norm": 1.042446420171934,
+      "learning_rate": 0.00010877774693528498,
+      "loss": 12.5463,
+      "step": 8369
+    },
+    {
+      "epoch": 1.463737704918033,
+      "grad_norm": 0.8692644077261414,
+      "learning_rate": 0.0001087589387589434,
+      "loss": 12.4102,
+      "step": 8370
+    },
+    {
+      "epoch": 1.4639125683060108,
+      "grad_norm": 0.8922819211285749,
+      "learning_rate": 0.00010874013027035588,
+      "loss": 12.6442,
+      "step": 8371
+    },
+    {
+      "epoch": 1.4640874316939891,
+      "grad_norm": 0.8691476040253241,
+      "learning_rate": 0.00010872132147019296,
+      "loss": 12.3828,
+      "step": 8372
+    },
+    {
+      "epoch": 1.4642622950819673,
+      "grad_norm": 0.9310082754714143,
+      "learning_rate": 0.0001087025123591251,
+      "loss": 12.8292,
+      "step": 8373
+    },
+    {
+      "epoch": 1.4644371584699454,
+      "grad_norm": 0.8044269922976581,
+      "learning_rate": 0.00010868370293782286,
+      "loss": 12.3818,
+      "step": 8374
+    },
+    {
+      "epoch": 1.4646120218579235,
+      "grad_norm": 0.8936400347492616,
+      "learning_rate": 0.00010866489320695672,
+      "loss": 12.2486,
+      "step": 8375
+    },
+    {
+      "epoch": 1.4647868852459016,
+      "grad_norm": 0.8734606676001041,
+      "learning_rate": 0.00010864608316719731,
+      "loss": 12.6883,
+      "step": 8376
+    },
+    {
+      "epoch": 1.4649617486338797,
+      "grad_norm": 0.8268689338631273,
+      "learning_rate": 0.0001086272728192151,
+      "loss": 12.6972,
+      "step": 8377
+    },
+    {
+      "epoch": 1.4651366120218579,
+      "grad_norm": 0.9739916370746787,
+      "learning_rate": 0.00010860846216368069,
+      "loss": 12.4462,
+      "step": 8378
+    },
+    {
+      "epoch": 1.4653114754098362,
+      "grad_norm": 0.9679981950178671,
+      "learning_rate": 0.00010858965120126469,
+      "loss": 12.5099,
+      "step": 8379
+    },
+    {
+      "epoch": 1.465486338797814,
+      "grad_norm": 0.8584248397539386,
+      "learning_rate": 0.00010857083993263764,
+      "loss": 12.5277,
+      "step": 8380
+    },
+    {
+      "epoch": 1.4656612021857924,
+      "grad_norm": 0.9533495103187719,
+      "learning_rate": 0.00010855202835847018,
+      "loss": 12.4618,
+      "step": 8381
+    },
+    {
+      "epoch": 1.4658360655737706,
+      "grad_norm": 0.9687107385352647,
+      "learning_rate": 0.00010853321647943286,
+      "loss": 12.6619,
+      "step": 8382
+    },
+    {
+      "epoch": 1.4660109289617487,
+      "grad_norm": 0.9330629425278357,
+      "learning_rate": 0.00010851440429619639,
+      "loss": 12.125,
+      "step": 8383
+    },
+    {
+      "epoch": 1.4661857923497268,
+      "grad_norm": 0.8739631576029804,
+      "learning_rate": 0.00010849559180943135,
+      "loss": 12.8244,
+      "step": 8384
+    },
+    {
+      "epoch": 1.466360655737705,
+      "grad_norm": 0.8974160447751439,
+      "learning_rate": 0.00010847677901980837,
+      "loss": 12.3206,
+      "step": 8385
+    },
+    {
+      "epoch": 1.466535519125683,
+      "grad_norm": 0.9762805407831742,
+      "learning_rate": 0.0001084579659279981,
+      "loss": 12.4412,
+      "step": 8386
+    },
+    {
+      "epoch": 1.4667103825136611,
+      "grad_norm": 0.990246563011631,
+      "learning_rate": 0.00010843915253467125,
+      "loss": 12.4709,
+      "step": 8387
+    },
+    {
+      "epoch": 1.4668852459016393,
+      "grad_norm": 0.8681134721600716,
+      "learning_rate": 0.00010842033884049847,
+      "loss": 12.267,
+      "step": 8388
+    },
+    {
+      "epoch": 1.4670601092896174,
+      "grad_norm": 0.8576618103547837,
+      "learning_rate": 0.00010840152484615048,
+      "loss": 12.5157,
+      "step": 8389
+    },
+    {
+      "epoch": 1.4672349726775957,
+      "grad_norm": 0.8138608516816535,
+      "learning_rate": 0.00010838271055229789,
+      "loss": 12.7549,
+      "step": 8390
+    },
+    {
+      "epoch": 1.4674098360655738,
+      "grad_norm": 0.8325286552990326,
+      "learning_rate": 0.0001083638959596115,
+      "loss": 12.4352,
+      "step": 8391
+    },
+    {
+      "epoch": 1.467584699453552,
+      "grad_norm": 0.8661996495839224,
+      "learning_rate": 0.000108345081068762,
+      "loss": 12.7634,
+      "step": 8392
+    },
+    {
+      "epoch": 1.46775956284153,
+      "grad_norm": 0.9549223462554179,
+      "learning_rate": 0.00010832626588042008,
+      "loss": 12.368,
+      "step": 8393
+    },
+    {
+      "epoch": 1.4679344262295082,
+      "grad_norm": 0.8286907459352781,
+      "learning_rate": 0.00010830745039525654,
+      "loss": 12.5081,
+      "step": 8394
+    },
+    {
+      "epoch": 1.4681092896174863,
+      "grad_norm": 0.9262136854017711,
+      "learning_rate": 0.00010828863461394208,
+      "loss": 12.5925,
+      "step": 8395
+    },
+    {
+      "epoch": 1.4682841530054644,
+      "grad_norm": 1.0404770593148323,
+      "learning_rate": 0.00010826981853714746,
+      "loss": 12.4679,
+      "step": 8396
+    },
+    {
+      "epoch": 1.4684590163934426,
+      "grad_norm": 0.9177770417759035,
+      "learning_rate": 0.00010825100216554349,
+      "loss": 12.5117,
+      "step": 8397
+    },
+    {
+      "epoch": 1.4686338797814207,
+      "grad_norm": 0.9765998634366251,
+      "learning_rate": 0.00010823218549980095,
+      "loss": 12.4666,
+      "step": 8398
+    },
+    {
+      "epoch": 1.468808743169399,
+      "grad_norm": 0.8756697574533419,
+      "learning_rate": 0.00010821336854059058,
+      "loss": 12.2857,
+      "step": 8399
+    },
+    {
+      "epoch": 1.468983606557377,
+      "grad_norm": 1.2866670800084514,
+      "learning_rate": 0.00010819455128858324,
+      "loss": 12.5846,
+      "step": 8400
+    },
+    {
+      "epoch": 1.4691584699453553,
+      "grad_norm": 0.8256360624121984,
+      "learning_rate": 0.00010817573374444972,
+      "loss": 12.4484,
+      "step": 8401
+    },
+    {
+      "epoch": 1.4693333333333334,
+      "grad_norm": 1.127383288559482,
+      "learning_rate": 0.00010815691590886085,
+      "loss": 12.6591,
+      "step": 8402
+    },
+    {
+      "epoch": 1.4695081967213115,
+      "grad_norm": 0.9083046451571061,
+      "learning_rate": 0.00010813809778248744,
+      "loss": 12.5042,
+      "step": 8403
+    },
+    {
+      "epoch": 1.4696830601092896,
+      "grad_norm": 1.029870758914921,
+      "learning_rate": 0.0001081192793660004,
+      "loss": 12.4649,
+      "step": 8404
+    },
+    {
+      "epoch": 1.4698579234972677,
+      "grad_norm": 0.8519179061658847,
+      "learning_rate": 0.00010810046066007049,
+      "loss": 12.7104,
+      "step": 8405
+    },
+    {
+      "epoch": 1.4700327868852459,
+      "grad_norm": 1.0157676885886522,
+      "learning_rate": 0.00010808164166536862,
+      "loss": 12.5763,
+      "step": 8406
+    },
+    {
+      "epoch": 1.470207650273224,
+      "grad_norm": 0.8869962766211789,
+      "learning_rate": 0.00010806282238256569,
+      "loss": 12.4207,
+      "step": 8407
+    },
+    {
+      "epoch": 1.4703825136612023,
+      "grad_norm": 0.9321738239616519,
+      "learning_rate": 0.00010804400281233253,
+      "loss": 12.642,
+      "step": 8408
+    },
+    {
+      "epoch": 1.4705573770491802,
+      "grad_norm": 0.89838281410065,
+      "learning_rate": 0.00010802518295534012,
+      "loss": 12.6035,
+      "step": 8409
+    },
+    {
+      "epoch": 1.4707322404371586,
+      "grad_norm": 0.938831239857002,
+      "learning_rate": 0.00010800636281225926,
+      "loss": 12.5851,
+      "step": 8410
+    },
+    {
+      "epoch": 1.4709071038251367,
+      "grad_norm": 0.836783212060901,
+      "learning_rate": 0.00010798754238376096,
+      "loss": 12.5281,
+      "step": 8411
+    },
+    {
+      "epoch": 1.4710819672131148,
+      "grad_norm": 0.9106864040633825,
+      "learning_rate": 0.0001079687216705161,
+      "loss": 12.7971,
+      "step": 8412
+    },
+    {
+      "epoch": 1.471256830601093,
+      "grad_norm": 0.9459149616555652,
+      "learning_rate": 0.00010794990067319562,
+      "loss": 12.4718,
+      "step": 8413
+    },
+    {
+      "epoch": 1.471431693989071,
+      "grad_norm": 0.8849166092351991,
+      "learning_rate": 0.00010793107939247044,
+      "loss": 12.5884,
+      "step": 8414
+    },
+    {
+      "epoch": 1.4716065573770492,
+      "grad_norm": 0.9069393564152191,
+      "learning_rate": 0.00010791225782901158,
+      "loss": 12.3827,
+      "step": 8415
+    },
+    {
+      "epoch": 1.4717814207650273,
+      "grad_norm": 0.876943170989315,
+      "learning_rate": 0.00010789343598348996,
+      "loss": 12.5418,
+      "step": 8416
+    },
+    {
+      "epoch": 1.4719562841530054,
+      "grad_norm": 0.8615398857660733,
+      "learning_rate": 0.00010787461385657658,
+      "loss": 12.4776,
+      "step": 8417
+    },
+    {
+      "epoch": 1.4721311475409835,
+      "grad_norm": 0.9139613504495553,
+      "learning_rate": 0.0001078557914489424,
+      "loss": 12.4289,
+      "step": 8418
+    },
+    {
+      "epoch": 1.4723060109289619,
+      "grad_norm": 0.7968223180447925,
+      "learning_rate": 0.00010783696876125846,
+      "loss": 11.7791,
+      "step": 8419
+    },
+    {
+      "epoch": 1.47248087431694,
+      "grad_norm": 0.958856817819427,
+      "learning_rate": 0.00010781814579419572,
+      "loss": 12.5106,
+      "step": 8420
+    },
+    {
+      "epoch": 1.472655737704918,
+      "grad_norm": 0.8276992910090525,
+      "learning_rate": 0.00010779932254842521,
+      "loss": 12.3193,
+      "step": 8421
+    },
+    {
+      "epoch": 1.4728306010928962,
+      "grad_norm": 1.090079886963483,
+      "learning_rate": 0.00010778049902461798,
+      "loss": 12.6399,
+      "step": 8422
+    },
+    {
+      "epoch": 1.4730054644808743,
+      "grad_norm": 0.8737018349741387,
+      "learning_rate": 0.00010776167522344504,
+      "loss": 12.205,
+      "step": 8423
+    },
+    {
+      "epoch": 1.4731803278688524,
+      "grad_norm": 0.9668398955289376,
+      "learning_rate": 0.00010774285114557745,
+      "loss": 12.5858,
+      "step": 8424
+    },
+    {
+      "epoch": 1.4733551912568306,
+      "grad_norm": 0.9968248270112592,
+      "learning_rate": 0.00010772402679168625,
+      "loss": 12.3964,
+      "step": 8425
+    },
+    {
+      "epoch": 1.4735300546448087,
+      "grad_norm": 0.9327626912100624,
+      "learning_rate": 0.00010770520216244255,
+      "loss": 12.4233,
+      "step": 8426
+    },
+    {
+      "epoch": 1.4737049180327868,
+      "grad_norm": 0.9909546550429474,
+      "learning_rate": 0.00010768637725851735,
+      "loss": 12.7674,
+      "step": 8427
+    },
+    {
+      "epoch": 1.4738797814207651,
+      "grad_norm": 0.8829759761431353,
+      "learning_rate": 0.00010766755208058184,
+      "loss": 12.5372,
+      "step": 8428
+    },
+    {
+      "epoch": 1.474054644808743,
+      "grad_norm": 0.9317050434888644,
+      "learning_rate": 0.00010764872662930703,
+      "loss": 12.5855,
+      "step": 8429
+    },
+    {
+      "epoch": 1.4742295081967214,
+      "grad_norm": 0.924976805577369,
+      "learning_rate": 0.00010762990090536404,
+      "loss": 12.5285,
+      "step": 8430
+    },
+    {
+      "epoch": 1.4744043715846995,
+      "grad_norm": 0.8168476626973477,
+      "learning_rate": 0.00010761107490942401,
+      "loss": 12.2142,
+      "step": 8431
+    },
+    {
+      "epoch": 1.4745792349726776,
+      "grad_norm": 0.8810797342852928,
+      "learning_rate": 0.00010759224864215805,
+      "loss": 12.6416,
+      "step": 8432
+    },
+    {
+      "epoch": 1.4747540983606557,
+      "grad_norm": 0.8514251857505607,
+      "learning_rate": 0.00010757342210423726,
+      "loss": 12.5571,
+      "step": 8433
+    },
+    {
+      "epoch": 1.4749289617486339,
+      "grad_norm": 0.8671351779586285,
+      "learning_rate": 0.00010755459529633287,
+      "loss": 12.1281,
+      "step": 8434
+    },
+    {
+      "epoch": 1.475103825136612,
+      "grad_norm": 0.8279000437962102,
+      "learning_rate": 0.00010753576821911597,
+      "loss": 12.5757,
+      "step": 8435
+    },
+    {
+      "epoch": 1.47527868852459,
+      "grad_norm": 0.8718147671721369,
+      "learning_rate": 0.00010751694087325774,
+      "loss": 12.5916,
+      "step": 8436
+    },
+    {
+      "epoch": 1.4754535519125684,
+      "grad_norm": 0.8486728424471771,
+      "learning_rate": 0.00010749811325942936,
+      "loss": 12.3582,
+      "step": 8437
+    },
+    {
+      "epoch": 1.4756284153005463,
+      "grad_norm": 0.9030307292302165,
+      "learning_rate": 0.00010747928537830198,
+      "loss": 12.6237,
+      "step": 8438
+    },
+    {
+      "epoch": 1.4758032786885247,
+      "grad_norm": 0.9290911345924985,
+      "learning_rate": 0.00010746045723054685,
+      "loss": 12.609,
+      "step": 8439
+    },
+    {
+      "epoch": 1.4759781420765028,
+      "grad_norm": 0.9935177093503251,
+      "learning_rate": 0.0001074416288168351,
+      "loss": 12.6394,
+      "step": 8440
+    },
+    {
+      "epoch": 1.476153005464481,
+      "grad_norm": 0.8936653705729091,
+      "learning_rate": 0.00010742280013783799,
+      "loss": 12.565,
+      "step": 8441
+    },
+    {
+      "epoch": 1.476327868852459,
+      "grad_norm": 0.9911204861230385,
+      "learning_rate": 0.00010740397119422671,
+      "loss": 12.4295,
+      "step": 8442
+    },
+    {
+      "epoch": 1.4765027322404372,
+      "grad_norm": 0.8362798458825972,
+      "learning_rate": 0.00010738514198667253,
+      "loss": 12.5511,
+      "step": 8443
+    },
+    {
+      "epoch": 1.4766775956284153,
+      "grad_norm": 1.024332006748974,
+      "learning_rate": 0.00010736631251584666,
+      "loss": 12.5535,
+      "step": 8444
+    },
+    {
+      "epoch": 1.4768524590163934,
+      "grad_norm": 0.8860920505575397,
+      "learning_rate": 0.00010734748278242038,
+      "loss": 12.484,
+      "step": 8445
+    },
+    {
+      "epoch": 1.4770273224043715,
+      "grad_norm": 1.051100469967993,
+      "learning_rate": 0.00010732865278706488,
+      "loss": 12.585,
+      "step": 8446
+    },
+    {
+      "epoch": 1.4772021857923496,
+      "grad_norm": 1.0872891764986927,
+      "learning_rate": 0.0001073098225304515,
+      "loss": 12.5603,
+      "step": 8447
+    },
+    {
+      "epoch": 1.477377049180328,
+      "grad_norm": 0.8462531912352655,
+      "learning_rate": 0.00010729099201325148,
+      "loss": 12.4742,
+      "step": 8448
+    },
+    {
+      "epoch": 1.477551912568306,
+      "grad_norm": 1.027351536806053,
+      "learning_rate": 0.00010727216123613612,
+      "loss": 12.5619,
+      "step": 8449
+    },
+    {
+      "epoch": 1.4777267759562842,
+      "grad_norm": 0.8586827808652006,
+      "learning_rate": 0.00010725333019977669,
+      "loss": 12.6301,
+      "step": 8450
+    },
+    {
+      "epoch": 1.4779016393442623,
+      "grad_norm": 1.02854218597266,
+      "learning_rate": 0.00010723449890484452,
+      "loss": 12.6666,
+      "step": 8451
+    },
+    {
+      "epoch": 1.4780765027322404,
+      "grad_norm": 0.8697922852781875,
+      "learning_rate": 0.0001072156673520109,
+      "loss": 12.5527,
+      "step": 8452
+    },
+    {
+      "epoch": 1.4782513661202186,
+      "grad_norm": 0.9119039217986623,
+      "learning_rate": 0.00010719683554194722,
+      "loss": 12.7676,
+      "step": 8453
+    },
+    {
+      "epoch": 1.4784262295081967,
+      "grad_norm": 0.909564893880191,
+      "learning_rate": 0.00010717800347532473,
+      "loss": 12.7152,
+      "step": 8454
+    },
+    {
+      "epoch": 1.4786010928961748,
+      "grad_norm": 0.8004217912627506,
+      "learning_rate": 0.0001071591711528148,
+      "loss": 12.2763,
+      "step": 8455
+    },
+    {
+      "epoch": 1.478775956284153,
+      "grad_norm": 0.9934871970979868,
+      "learning_rate": 0.00010714033857508879,
+      "loss": 12.5464,
+      "step": 8456
+    },
+    {
+      "epoch": 1.4789508196721313,
+      "grad_norm": 0.9288039913561553,
+      "learning_rate": 0.00010712150574281805,
+      "loss": 12.5187,
+      "step": 8457
+    },
+    {
+      "epoch": 1.4791256830601092,
+      "grad_norm": 0.88426980301783,
+      "learning_rate": 0.00010710267265667395,
+      "loss": 12.3585,
+      "step": 8458
+    },
+    {
+      "epoch": 1.4793005464480875,
+      "grad_norm": 0.8768683932396324,
+      "learning_rate": 0.00010708383931732786,
+      "loss": 12.5526,
+      "step": 8459
+    },
+    {
+      "epoch": 1.4794754098360656,
+      "grad_norm": 0.8738952767659309,
+      "learning_rate": 0.00010706500572545118,
+      "loss": 12.568,
+      "step": 8460
+    },
+    {
+      "epoch": 1.4796502732240437,
+      "grad_norm": 0.8379072828399788,
+      "learning_rate": 0.00010704617188171529,
+      "loss": 12.2701,
+      "step": 8461
+    },
+    {
+      "epoch": 1.4798251366120219,
+      "grad_norm": 0.8440924130402738,
+      "learning_rate": 0.00010702733778679164,
+      "loss": 12.4865,
+      "step": 8462
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9442852300440827,
+      "learning_rate": 0.00010700850344135158,
+      "loss": 12.4589,
+      "step": 8463
+    },
+    {
+      "epoch": 1.480174863387978,
+      "grad_norm": 0.8568612001023266,
+      "learning_rate": 0.00010698966884606659,
+      "loss": 12.3983,
+      "step": 8464
+    },
+    {
+      "epoch": 1.4803497267759562,
+      "grad_norm": 0.8934555387286246,
+      "learning_rate": 0.00010697083400160801,
+      "loss": 12.6002,
+      "step": 8465
+    },
+    {
+      "epoch": 1.4805245901639346,
+      "grad_norm": 0.8737359288966402,
+      "learning_rate": 0.0001069519989086474,
+      "loss": 12.4667,
+      "step": 8466
+    },
+    {
+      "epoch": 1.4806994535519125,
+      "grad_norm": 0.8200668118424213,
+      "learning_rate": 0.00010693316356785615,
+      "loss": 12.5233,
+      "step": 8467
+    },
+    {
+      "epoch": 1.4808743169398908,
+      "grad_norm": 1.0180276577457714,
+      "learning_rate": 0.0001069143279799057,
+      "loss": 12.6525,
+      "step": 8468
+    },
+    {
+      "epoch": 1.481049180327869,
+      "grad_norm": 0.952120279347043,
+      "learning_rate": 0.00010689549214546753,
+      "loss": 12.5679,
+      "step": 8469
+    },
+    {
+      "epoch": 1.481224043715847,
+      "grad_norm": 0.9722644484899204,
+      "learning_rate": 0.00010687665606521311,
+      "loss": 12.6755,
+      "step": 8470
+    },
+    {
+      "epoch": 1.4813989071038252,
+      "grad_norm": 0.842626801125624,
+      "learning_rate": 0.00010685781973981398,
+      "loss": 12.6258,
+      "step": 8471
+    },
+    {
+      "epoch": 1.4815737704918033,
+      "grad_norm": 0.9396322202559235,
+      "learning_rate": 0.00010683898316994156,
+      "loss": 12.4811,
+      "step": 8472
+    },
+    {
+      "epoch": 1.4817486338797814,
+      "grad_norm": 0.8741241391875951,
+      "learning_rate": 0.00010682014635626741,
+      "loss": 12.2736,
+      "step": 8473
+    },
+    {
+      "epoch": 1.4819234972677595,
+      "grad_norm": 0.8439563103247604,
+      "learning_rate": 0.00010680130929946296,
+      "loss": 12.6784,
+      "step": 8474
+    },
+    {
+      "epoch": 1.4820983606557376,
+      "grad_norm": 0.9149211222788804,
+      "learning_rate": 0.00010678247200019981,
+      "loss": 12.5777,
+      "step": 8475
+    },
+    {
+      "epoch": 1.4822732240437158,
+      "grad_norm": 0.8990721503049465,
+      "learning_rate": 0.00010676363445914945,
+      "loss": 12.4141,
+      "step": 8476
+    },
+    {
+      "epoch": 1.482448087431694,
+      "grad_norm": 0.8545966122299057,
+      "learning_rate": 0.00010674479667698342,
+      "loss": 12.3099,
+      "step": 8477
+    },
+    {
+      "epoch": 1.4826229508196722,
+      "grad_norm": 0.8963376791044619,
+      "learning_rate": 0.00010672595865437327,
+      "loss": 12.6503,
+      "step": 8478
+    },
+    {
+      "epoch": 1.4827978142076503,
+      "grad_norm": 0.9654401975235286,
+      "learning_rate": 0.00010670712039199057,
+      "loss": 12.6388,
+      "step": 8479
+    },
+    {
+      "epoch": 1.4829726775956285,
+      "grad_norm": 0.8346336788436199,
+      "learning_rate": 0.00010668828189050686,
+      "loss": 12.5677,
+      "step": 8480
+    },
+    {
+      "epoch": 1.4831475409836066,
+      "grad_norm": 0.8494644923232404,
+      "learning_rate": 0.00010666944315059371,
+      "loss": 12.3624,
+      "step": 8481
+    },
+    {
+      "epoch": 1.4833224043715847,
+      "grad_norm": 0.8164900726960362,
+      "learning_rate": 0.0001066506041729227,
+      "loss": 12.5797,
+      "step": 8482
+    },
+    {
+      "epoch": 1.4834972677595628,
+      "grad_norm": 0.8797731308887554,
+      "learning_rate": 0.00010663176495816543,
+      "loss": 12.4292,
+      "step": 8483
+    },
+    {
+      "epoch": 1.483672131147541,
+      "grad_norm": 0.9142594647871727,
+      "learning_rate": 0.00010661292550699346,
+      "loss": 12.3663,
+      "step": 8484
+    },
+    {
+      "epoch": 1.483846994535519,
+      "grad_norm": 0.8513139739763689,
+      "learning_rate": 0.00010659408582007848,
+      "loss": 12.4148,
+      "step": 8485
+    },
+    {
+      "epoch": 1.4840218579234974,
+      "grad_norm": 0.9141400246655257,
+      "learning_rate": 0.000106575245898092,
+      "loss": 12.496,
+      "step": 8486
+    },
+    {
+      "epoch": 1.4841967213114753,
+      "grad_norm": 0.9326564123702514,
+      "learning_rate": 0.00010655640574170568,
+      "loss": 12.6182,
+      "step": 8487
+    },
+    {
+      "epoch": 1.4843715846994536,
+      "grad_norm": 0.8462721073238955,
+      "learning_rate": 0.00010653756535159118,
+      "loss": 12.3578,
+      "step": 8488
+    },
+    {
+      "epoch": 1.4845464480874317,
+      "grad_norm": 0.8303629917311007,
+      "learning_rate": 0.00010651872472842011,
+      "loss": 12.5014,
+      "step": 8489
+    },
+    {
+      "epoch": 1.4847213114754099,
+      "grad_norm": 0.9061090485822794,
+      "learning_rate": 0.00010649988387286411,
+      "loss": 12.387,
+      "step": 8490
+    },
+    {
+      "epoch": 1.484896174863388,
+      "grad_norm": 0.8766125374055391,
+      "learning_rate": 0.00010648104278559485,
+      "loss": 12.5321,
+      "step": 8491
+    },
+    {
+      "epoch": 1.485071038251366,
+      "grad_norm": 0.8464518782204247,
+      "learning_rate": 0.00010646220146728399,
+      "loss": 12.5802,
+      "step": 8492
+    },
+    {
+      "epoch": 1.4852459016393442,
+      "grad_norm": 0.83929464583899,
+      "learning_rate": 0.00010644335991860319,
+      "loss": 12.4319,
+      "step": 8493
+    },
+    {
+      "epoch": 1.4854207650273223,
+      "grad_norm": 0.809335152128572,
+      "learning_rate": 0.00010642451814022416,
+      "loss": 12.518,
+      "step": 8494
+    },
+    {
+      "epoch": 1.4855956284153007,
+      "grad_norm": 0.9057589567283063,
+      "learning_rate": 0.00010640567613281854,
+      "loss": 12.7841,
+      "step": 8495
+    },
+    {
+      "epoch": 1.4857704918032786,
+      "grad_norm": 0.9195024796470672,
+      "learning_rate": 0.00010638683389705804,
+      "loss": 12.4455,
+      "step": 8496
+    },
+    {
+      "epoch": 1.485945355191257,
+      "grad_norm": 0.9763659507668692,
+      "learning_rate": 0.00010636799143361438,
+      "loss": 12.3569,
+      "step": 8497
+    },
+    {
+      "epoch": 1.486120218579235,
+      "grad_norm": 0.9302109010099469,
+      "learning_rate": 0.00010634914874315929,
+      "loss": 12.5468,
+      "step": 8498
+    },
+    {
+      "epoch": 1.4862950819672132,
+      "grad_norm": 1.0308055960859823,
+      "learning_rate": 0.00010633030582636443,
+      "loss": 12.4831,
+      "step": 8499
+    },
+    {
+      "epoch": 1.4864699453551913,
+      "grad_norm": 1.0988092385261878,
+      "learning_rate": 0.0001063114626839016,
+      "loss": 12.252,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4866448087431694,
+      "grad_norm": 1.0222601955858854,
+      "learning_rate": 0.0001062926193164425,
+      "loss": 12.5872,
+      "step": 8501
+    },
+    {
+      "epoch": 1.4868196721311475,
+      "grad_norm": 1.0940014291994988,
+      "learning_rate": 0.00010627377572465884,
+      "loss": 12.5335,
+      "step": 8502
+    },
+    {
+      "epoch": 1.4869945355191256,
+      "grad_norm": 0.9678982523285841,
+      "learning_rate": 0.00010625493190922244,
+      "loss": 12.5678,
+      "step": 8503
+    },
+    {
+      "epoch": 1.4871693989071038,
+      "grad_norm": 1.0827263861828915,
+      "learning_rate": 0.000106236087870805,
+      "loss": 12.4706,
+      "step": 8504
+    },
+    {
+      "epoch": 1.4873442622950819,
+      "grad_norm": 0.8582958643335811,
+      "learning_rate": 0.0001062172436100783,
+      "loss": 12.6327,
+      "step": 8505
+    },
+    {
+      "epoch": 1.4875191256830602,
+      "grad_norm": 0.9032065409285445,
+      "learning_rate": 0.00010619839912771414,
+      "loss": 12.7883,
+      "step": 8506
+    },
+    {
+      "epoch": 1.4876939890710381,
+      "grad_norm": 0.8874862722943533,
+      "learning_rate": 0.00010617955442438432,
+      "loss": 12.5287,
+      "step": 8507
+    },
+    {
+      "epoch": 1.4878688524590165,
+      "grad_norm": 1.1807181699239793,
+      "learning_rate": 0.00010616070950076058,
+      "loss": 12.8144,
+      "step": 8508
+    },
+    {
+      "epoch": 1.4880437158469946,
+      "grad_norm": 0.801418298633424,
+      "learning_rate": 0.00010614186435751476,
+      "loss": 12.4778,
+      "step": 8509
+    },
+    {
+      "epoch": 1.4882185792349727,
+      "grad_norm": 1.0616603942250389,
+      "learning_rate": 0.00010612301899531865,
+      "loss": 12.7147,
+      "step": 8510
+    },
+    {
+      "epoch": 1.4883934426229508,
+      "grad_norm": 0.9397516074444198,
+      "learning_rate": 0.00010610417341484406,
+      "loss": 12.6676,
+      "step": 8511
+    },
+    {
+      "epoch": 1.488568306010929,
+      "grad_norm": 0.9079461604677818,
+      "learning_rate": 0.00010608532761676282,
+      "loss": 12.5958,
+      "step": 8512
+    },
+    {
+      "epoch": 1.488743169398907,
+      "grad_norm": 1.0013602733624352,
+      "learning_rate": 0.00010606648160174677,
+      "loss": 12.437,
+      "step": 8513
+    },
+    {
+      "epoch": 1.4889180327868852,
+      "grad_norm": 0.8763489308209385,
+      "learning_rate": 0.00010604763537046773,
+      "loss": 12.774,
+      "step": 8514
+    },
+    {
+      "epoch": 1.4890928961748635,
+      "grad_norm": 1.0491776144463287,
+      "learning_rate": 0.00010602878892359753,
+      "loss": 12.6252,
+      "step": 8515
+    },
+    {
+      "epoch": 1.4892677595628414,
+      "grad_norm": 0.8485751449738611,
+      "learning_rate": 0.0001060099422618081,
+      "loss": 12.4271,
+      "step": 8516
+    },
+    {
+      "epoch": 1.4894426229508198,
+      "grad_norm": 0.8292188382249565,
+      "learning_rate": 0.00010599109538577122,
+      "loss": 12.2698,
+      "step": 8517
+    },
+    {
+      "epoch": 1.4896174863387979,
+      "grad_norm": 0.8734346299661377,
+      "learning_rate": 0.00010597224829615879,
+      "loss": 12.6663,
+      "step": 8518
+    },
+    {
+      "epoch": 1.489792349726776,
+      "grad_norm": 0.7838163795533105,
+      "learning_rate": 0.00010595340099364268,
+      "loss": 12.7631,
+      "step": 8519
+    },
+    {
+      "epoch": 1.489967213114754,
+      "grad_norm": 0.9735907952293417,
+      "learning_rate": 0.0001059345534788948,
+      "loss": 12.6575,
+      "step": 8520
+    },
+    {
+      "epoch": 1.4901420765027322,
+      "grad_norm": 0.8438013297221556,
+      "learning_rate": 0.00010591570575258702,
+      "loss": 12.4665,
+      "step": 8521
+    },
+    {
+      "epoch": 1.4903169398907103,
+      "grad_norm": 0.9413011806639835,
+      "learning_rate": 0.00010589685781539122,
+      "loss": 12.4926,
+      "step": 8522
+    },
+    {
+      "epoch": 1.4904918032786885,
+      "grad_norm": 0.8669659232094992,
+      "learning_rate": 0.00010587800966797934,
+      "loss": 12.473,
+      "step": 8523
+    },
+    {
+      "epoch": 1.4906666666666666,
+      "grad_norm": 0.8749630216237756,
+      "learning_rate": 0.00010585916131102329,
+      "loss": 12.2792,
+      "step": 8524
+    },
+    {
+      "epoch": 1.4908415300546447,
+      "grad_norm": 0.8861553791893306,
+      "learning_rate": 0.00010584031274519497,
+      "loss": 12.4509,
+      "step": 8525
+    },
+    {
+      "epoch": 1.491016393442623,
+      "grad_norm": 0.8703024556185357,
+      "learning_rate": 0.00010582146397116634,
+      "loss": 12.7835,
+      "step": 8526
+    },
+    {
+      "epoch": 1.4911912568306012,
+      "grad_norm": 0.8325538153326252,
+      "learning_rate": 0.00010580261498960933,
+      "loss": 12.4708,
+      "step": 8527
+    },
+    {
+      "epoch": 1.4913661202185793,
+      "grad_norm": 0.9795680740184397,
+      "learning_rate": 0.00010578376580119587,
+      "loss": 12.6434,
+      "step": 8528
+    },
+    {
+      "epoch": 1.4915409836065574,
+      "grad_norm": 0.8892036972176605,
+      "learning_rate": 0.00010576491640659789,
+      "loss": 12.5754,
+      "step": 8529
+    },
+    {
+      "epoch": 1.4917158469945355,
+      "grad_norm": 0.9272664995298787,
+      "learning_rate": 0.0001057460668064874,
+      "loss": 12.4846,
+      "step": 8530
+    },
+    {
+      "epoch": 1.4918907103825136,
+      "grad_norm": 0.8430925554450315,
+      "learning_rate": 0.00010572721700153634,
+      "loss": 12.4525,
+      "step": 8531
+    },
+    {
+      "epoch": 1.4920655737704918,
+      "grad_norm": 0.8879217722358399,
+      "learning_rate": 0.00010570836699241667,
+      "loss": 12.3934,
+      "step": 8532
+    },
+    {
+      "epoch": 1.4922404371584699,
+      "grad_norm": 0.979653051097447,
+      "learning_rate": 0.00010568951677980041,
+      "loss": 12.5662,
+      "step": 8533
+    },
+    {
+      "epoch": 1.492415300546448,
+      "grad_norm": 0.9578608381693213,
+      "learning_rate": 0.0001056706663643595,
+      "loss": 12.5469,
+      "step": 8534
+    },
+    {
+      "epoch": 1.4925901639344263,
+      "grad_norm": 1.1201123498175347,
+      "learning_rate": 0.00010565181574676598,
+      "loss": 12.6094,
+      "step": 8535
+    },
+    {
+      "epoch": 1.4927650273224042,
+      "grad_norm": 1.0190610874681014,
+      "learning_rate": 0.0001056329649276918,
+      "loss": 12.533,
+      "step": 8536
+    },
+    {
+      "epoch": 1.4929398907103826,
+      "grad_norm": 0.9198614816909826,
+      "learning_rate": 0.00010561411390780904,
+      "loss": 12.2663,
+      "step": 8537
+    },
+    {
+      "epoch": 1.4931147540983607,
+      "grad_norm": 0.8714124589762254,
+      "learning_rate": 0.00010559526268778967,
+      "loss": 12.5868,
+      "step": 8538
+    },
+    {
+      "epoch": 1.4932896174863388,
+      "grad_norm": 0.9406126864094495,
+      "learning_rate": 0.00010557641126830573,
+      "loss": 12.1862,
+      "step": 8539
+    },
+    {
+      "epoch": 1.493464480874317,
+      "grad_norm": 0.8563689690513447,
+      "learning_rate": 0.00010555755965002922,
+      "loss": 12.656,
+      "step": 8540
+    },
+    {
+      "epoch": 1.493639344262295,
+      "grad_norm": 0.9830121402804619,
+      "learning_rate": 0.0001055387078336322,
+      "loss": 12.5622,
+      "step": 8541
+    },
+    {
+      "epoch": 1.4938142076502732,
+      "grad_norm": 0.8574666698167875,
+      "learning_rate": 0.00010551985581978672,
+      "loss": 12.6827,
+      "step": 8542
+    },
+    {
+      "epoch": 1.4939890710382513,
+      "grad_norm": 0.8941245933484989,
+      "learning_rate": 0.00010550100360916484,
+      "loss": 12.1278,
+      "step": 8543
+    },
+    {
+      "epoch": 1.4941639344262296,
+      "grad_norm": 0.8281530318728838,
+      "learning_rate": 0.00010548215120243861,
+      "loss": 12.5057,
+      "step": 8544
+    },
+    {
+      "epoch": 1.4943387978142075,
+      "grad_norm": 0.8619672349425133,
+      "learning_rate": 0.0001054632986002801,
+      "loss": 12.4237,
+      "step": 8545
+    },
+    {
+      "epoch": 1.4945136612021859,
+      "grad_norm": 0.8191158669488459,
+      "learning_rate": 0.00010544444580336135,
+      "loss": 12.4431,
+      "step": 8546
+    },
+    {
+      "epoch": 1.494688524590164,
+      "grad_norm": 0.9013448130674392,
+      "learning_rate": 0.00010542559281235452,
+      "loss": 12.5492,
+      "step": 8547
+    },
+    {
+      "epoch": 1.494863387978142,
+      "grad_norm": 0.9037637078488759,
+      "learning_rate": 0.00010540673962793161,
+      "loss": 12.4715,
+      "step": 8548
+    },
+    {
+      "epoch": 1.4950382513661202,
+      "grad_norm": 0.8771135542432359,
+      "learning_rate": 0.00010538788625076475,
+      "loss": 12.443,
+      "step": 8549
+    },
+    {
+      "epoch": 1.4952131147540983,
+      "grad_norm": 0.8063438857103185,
+      "learning_rate": 0.00010536903268152602,
+      "loss": 12.7181,
+      "step": 8550
+    },
+    {
+      "epoch": 1.4953879781420765,
+      "grad_norm": 0.8843832443480406,
+      "learning_rate": 0.00010535017892088757,
+      "loss": 12.2871,
+      "step": 8551
+    },
+    {
+      "epoch": 1.4955628415300546,
+      "grad_norm": 0.8760688556532847,
+      "learning_rate": 0.00010533132496952152,
+      "loss": 12.6647,
+      "step": 8552
+    },
+    {
+      "epoch": 1.4957377049180327,
+      "grad_norm": 0.8449461478134513,
+      "learning_rate": 0.00010531247082809992,
+      "loss": 12.2294,
+      "step": 8553
+    },
+    {
+      "epoch": 1.4959125683060108,
+      "grad_norm": 0.8433527094787162,
+      "learning_rate": 0.00010529361649729497,
+      "loss": 12.3077,
+      "step": 8554
+    },
+    {
+      "epoch": 1.4960874316939892,
+      "grad_norm": 1.0010874494714348,
+      "learning_rate": 0.00010527476197777875,
+      "loss": 12.6883,
+      "step": 8555
+    },
+    {
+      "epoch": 1.4962622950819673,
+      "grad_norm": 0.9365179685498033,
+      "learning_rate": 0.00010525590727022345,
+      "loss": 12.4663,
+      "step": 8556
+    },
+    {
+      "epoch": 1.4964371584699454,
+      "grad_norm": 0.9794423640982062,
+      "learning_rate": 0.0001052370523753012,
+      "loss": 12.4915,
+      "step": 8557
+    },
+    {
+      "epoch": 1.4966120218579235,
+      "grad_norm": 0.9071409375423495,
+      "learning_rate": 0.00010521819729368414,
+      "loss": 12.3416,
+      "step": 8558
+    },
+    {
+      "epoch": 1.4967868852459016,
+      "grad_norm": 0.9133093591536828,
+      "learning_rate": 0.00010519934202604444,
+      "loss": 12.7477,
+      "step": 8559
+    },
+    {
+      "epoch": 1.4969617486338798,
+      "grad_norm": 1.0257835966263094,
+      "learning_rate": 0.00010518048657305426,
+      "loss": 12.5972,
+      "step": 8560
+    },
+    {
+      "epoch": 1.4971366120218579,
+      "grad_norm": 0.9136943828073743,
+      "learning_rate": 0.0001051616309353858,
+      "loss": 12.7159,
+      "step": 8561
+    },
+    {
+      "epoch": 1.497311475409836,
+      "grad_norm": 0.9366086016795039,
+      "learning_rate": 0.00010514277511371124,
+      "loss": 12.3729,
+      "step": 8562
+    },
+    {
+      "epoch": 1.4974863387978141,
+      "grad_norm": 1.041246690203112,
+      "learning_rate": 0.00010512391910870276,
+      "loss": 12.4003,
+      "step": 8563
+    },
+    {
+      "epoch": 1.4976612021857925,
+      "grad_norm": 0.8273169240648313,
+      "learning_rate": 0.00010510506292103252,
+      "loss": 12.5366,
+      "step": 8564
+    },
+    {
+      "epoch": 1.4978360655737704,
+      "grad_norm": 1.0695979274251854,
+      "learning_rate": 0.0001050862065513728,
+      "loss": 12.3874,
+      "step": 8565
+    },
+    {
+      "epoch": 1.4980109289617487,
+      "grad_norm": 0.9581045848200255,
+      "learning_rate": 0.00010506735000039572,
+      "loss": 12.587,
+      "step": 8566
+    },
+    {
+      "epoch": 1.4981857923497268,
+      "grad_norm": 1.2356969748126798,
+      "learning_rate": 0.00010504849326877355,
+      "loss": 12.7864,
+      "step": 8567
+    },
+    {
+      "epoch": 1.498360655737705,
+      "grad_norm": 0.8818187806713084,
+      "learning_rate": 0.00010502963635717846,
+      "loss": 12.6576,
+      "step": 8568
+    },
+    {
+      "epoch": 1.498535519125683,
+      "grad_norm": 1.1050374131051082,
+      "learning_rate": 0.00010501077926628273,
+      "loss": 12.6849,
+      "step": 8569
+    },
+    {
+      "epoch": 1.4987103825136612,
+      "grad_norm": 0.979542168298192,
+      "learning_rate": 0.00010499192199675857,
+      "loss": 12.5976,
+      "step": 8570
+    },
+    {
+      "epoch": 1.4988852459016393,
+      "grad_norm": 0.9961664535171065,
+      "learning_rate": 0.00010497306454927825,
+      "loss": 12.5833,
+      "step": 8571
+    },
+    {
+      "epoch": 1.4990601092896174,
+      "grad_norm": 1.0857838171998946,
+      "learning_rate": 0.00010495420692451397,
+      "loss": 12.5696,
+      "step": 8572
+    },
+    {
+      "epoch": 1.4992349726775958,
+      "grad_norm": 0.8658776558068718,
+      "learning_rate": 0.000104935349123138,
+      "loss": 12.3784,
+      "step": 8573
+    },
+    {
+      "epoch": 1.4994098360655737,
+      "grad_norm": 1.0423904711318905,
+      "learning_rate": 0.0001049164911458226,
+      "loss": 12.3768,
+      "step": 8574
+    },
+    {
+      "epoch": 1.499584699453552,
+      "grad_norm": 0.8719561097535383,
+      "learning_rate": 0.00010489763299324004,
+      "loss": 12.4324,
+      "step": 8575
+    },
+    {
+      "epoch": 1.4997595628415301,
+      "grad_norm": 1.0305503472449464,
+      "learning_rate": 0.0001048787746660626,
+      "loss": 12.5596,
+      "step": 8576
+    },
+    {
+      "epoch": 1.4999344262295082,
+      "grad_norm": 0.9303733249303655,
+      "learning_rate": 0.00010485991616496252,
+      "loss": 12.5848,
+      "step": 8577
+    },
+    {
+      "epoch": 1.5001092896174864,
+      "grad_norm": 1.115903216728464,
+      "learning_rate": 0.0001048410574906121,
+      "loss": 12.4351,
+      "step": 8578
+    },
+    {
+      "epoch": 1.5002841530054645,
+      "grad_norm": 0.9052909922047911,
+      "learning_rate": 0.00010482219864368365,
+      "loss": 12.5073,
+      "step": 8579
+    },
+    {
+      "epoch": 1.5004590163934426,
+      "grad_norm": 1.1411658065417103,
+      "learning_rate": 0.00010480333962484948,
+      "loss": 12.7337,
+      "step": 8580
+    },
+    {
+      "epoch": 1.5006338797814207,
+      "grad_norm": 0.9574161826486545,
+      "learning_rate": 0.00010478448043478182,
+      "loss": 12.6036,
+      "step": 8581
+    },
+    {
+      "epoch": 1.500808743169399,
+      "grad_norm": 1.0283381049693847,
+      "learning_rate": 0.00010476562107415305,
+      "loss": 12.5621,
+      "step": 8582
+    },
+    {
+      "epoch": 1.500983606557377,
+      "grad_norm": 1.1589314666512702,
+      "learning_rate": 0.00010474676154363543,
+      "loss": 12.5771,
+      "step": 8583
+    },
+    {
+      "epoch": 1.5011584699453553,
+      "grad_norm": 0.966710895792901,
+      "learning_rate": 0.00010472790184390136,
+      "loss": 12.6887,
+      "step": 8584
+    },
+    {
+      "epoch": 1.5013333333333332,
+      "grad_norm": 1.0535964337353871,
+      "learning_rate": 0.00010470904197562305,
+      "loss": 12.5065,
+      "step": 8585
+    },
+    {
+      "epoch": 1.5015081967213115,
+      "grad_norm": 0.9536509096188037,
+      "learning_rate": 0.0001046901819394729,
+      "loss": 12.6288,
+      "step": 8586
+    },
+    {
+      "epoch": 1.5016830601092896,
+      "grad_norm": 1.2318677954147836,
+      "learning_rate": 0.00010467132173612329,
+      "loss": 12.6417,
+      "step": 8587
+    },
+    {
+      "epoch": 1.5018579234972678,
+      "grad_norm": 0.851192229228795,
+      "learning_rate": 0.00010465246136624649,
+      "loss": 12.6014,
+      "step": 8588
+    },
+    {
+      "epoch": 1.5020327868852459,
+      "grad_norm": 1.0283230073505745,
+      "learning_rate": 0.00010463360083051488,
+      "loss": 12.5336,
+      "step": 8589
+    },
+    {
+      "epoch": 1.502207650273224,
+      "grad_norm": 0.8220637414668986,
+      "learning_rate": 0.00010461474012960082,
+      "loss": 12.4671,
+      "step": 8590
+    },
+    {
+      "epoch": 1.5023825136612023,
+      "grad_norm": 0.9356422498189204,
+      "learning_rate": 0.00010459587926417667,
+      "loss": 12.5395,
+      "step": 8591
+    },
+    {
+      "epoch": 1.5025573770491802,
+      "grad_norm": 0.9167277670361936,
+      "learning_rate": 0.0001045770182349148,
+      "loss": 12.2158,
+      "step": 8592
+    },
+    {
+      "epoch": 1.5027322404371586,
+      "grad_norm": 0.8847969461791523,
+      "learning_rate": 0.00010455815704248755,
+      "loss": 12.4517,
+      "step": 8593
+    },
+    {
+      "epoch": 1.5029071038251365,
+      "grad_norm": 0.8642151967277056,
+      "learning_rate": 0.00010453929568756735,
+      "loss": 12.5887,
+      "step": 8594
+    },
+    {
+      "epoch": 1.5030819672131148,
+      "grad_norm": 0.981067040369997,
+      "learning_rate": 0.00010452043417082654,
+      "loss": 12.6438,
+      "step": 8595
+    },
+    {
+      "epoch": 1.503256830601093,
+      "grad_norm": 0.7740227462316293,
+      "learning_rate": 0.00010450157249293754,
+      "loss": 12.6774,
+      "step": 8596
+    },
+    {
+      "epoch": 1.503431693989071,
+      "grad_norm": 0.9069887150834689,
+      "learning_rate": 0.00010448271065457275,
+      "loss": 12.7,
+      "step": 8597
+    },
+    {
+      "epoch": 1.5036065573770492,
+      "grad_norm": 0.8906401217890463,
+      "learning_rate": 0.00010446384865640456,
+      "loss": 12.2871,
+      "step": 8598
+    },
+    {
+      "epoch": 1.5037814207650273,
+      "grad_norm": 0.8689609861147126,
+      "learning_rate": 0.00010444498649910539,
+      "loss": 12.3318,
+      "step": 8599
+    },
+    {
+      "epoch": 1.5039562841530054,
+      "grad_norm": 0.9141613916101471,
+      "learning_rate": 0.00010442612418334763,
+      "loss": 12.6846,
+      "step": 8600
+    },
+    {
+      "epoch": 1.5041311475409835,
+      "grad_norm": 0.8445754663902966,
+      "learning_rate": 0.00010440726170980372,
+      "loss": 12.4911,
+      "step": 8601
+    },
+    {
+      "epoch": 1.5043060109289619,
+      "grad_norm": 0.8013522560595361,
+      "learning_rate": 0.00010438839907914608,
+      "loss": 12.5217,
+      "step": 8602
+    },
+    {
+      "epoch": 1.5044808743169398,
+      "grad_norm": 0.9182071543674741,
+      "learning_rate": 0.00010436953629204715,
+      "loss": 12.6598,
+      "step": 8603
+    },
+    {
+      "epoch": 1.5046557377049181,
+      "grad_norm": 0.8371055221250275,
+      "learning_rate": 0.00010435067334917934,
+      "loss": 12.5089,
+      "step": 8604
+    },
+    {
+      "epoch": 1.504830601092896,
+      "grad_norm": 0.9585417269763045,
+      "learning_rate": 0.0001043318102512151,
+      "loss": 12.3824,
+      "step": 8605
+    },
+    {
+      "epoch": 1.5050054644808744,
+      "grad_norm": 0.8092737156881172,
+      "learning_rate": 0.0001043129469988269,
+      "loss": 12.3966,
+      "step": 8606
+    },
+    {
+      "epoch": 1.5051803278688525,
+      "grad_norm": 1.0288096660516857,
+      "learning_rate": 0.00010429408359268718,
+      "loss": 12.2816,
+      "step": 8607
+    },
+    {
+      "epoch": 1.5053551912568306,
+      "grad_norm": 0.8956996108604248,
+      "learning_rate": 0.00010427522003346839,
+      "loss": 12.6771,
+      "step": 8608
+    },
+    {
+      "epoch": 1.5055300546448087,
+      "grad_norm": 0.9252121265936708,
+      "learning_rate": 0.00010425635632184303,
+      "loss": 12.4171,
+      "step": 8609
+    },
+    {
+      "epoch": 1.5057049180327868,
+      "grad_norm": 0.867113083999361,
+      "learning_rate": 0.00010423749245848351,
+      "loss": 12.6296,
+      "step": 8610
+    },
+    {
+      "epoch": 1.5058797814207652,
+      "grad_norm": 0.8725824646600967,
+      "learning_rate": 0.00010421862844406234,
+      "loss": 12.5548,
+      "step": 8611
+    },
+    {
+      "epoch": 1.506054644808743,
+      "grad_norm": 1.0125079784906785,
+      "learning_rate": 0.000104199764279252,
+      "loss": 12.4263,
+      "step": 8612
+    },
+    {
+      "epoch": 1.5062295081967214,
+      "grad_norm": 0.855833370906108,
+      "learning_rate": 0.00010418089996472497,
+      "loss": 12.498,
+      "step": 8613
+    },
+    {
+      "epoch": 1.5064043715846993,
+      "grad_norm": 0.9573233558026336,
+      "learning_rate": 0.00010416203550115373,
+      "loss": 12.3454,
+      "step": 8614
+    },
+    {
+      "epoch": 1.5065792349726777,
+      "grad_norm": 0.9331812572114606,
+      "learning_rate": 0.00010414317088921078,
+      "loss": 12.5894,
+      "step": 8615
+    },
+    {
+      "epoch": 1.5067540983606558,
+      "grad_norm": 0.9993847686101508,
+      "learning_rate": 0.00010412430612956864,
+      "loss": 12.3331,
+      "step": 8616
+    },
+    {
+      "epoch": 1.5069289617486339,
+      "grad_norm": 1.2109868815570608,
+      "learning_rate": 0.00010410544122289983,
+      "loss": 12.4082,
+      "step": 8617
+    },
+    {
+      "epoch": 1.507103825136612,
+      "grad_norm": 0.8077439717586005,
+      "learning_rate": 0.0001040865761698768,
+      "loss": 12.1907,
+      "step": 8618
+    },
+    {
+      "epoch": 1.5072786885245901,
+      "grad_norm": 1.1961363067313746,
+      "learning_rate": 0.00010406771097117212,
+      "loss": 12.2353,
+      "step": 8619
+    },
+    {
+      "epoch": 1.5074535519125685,
+      "grad_norm": 0.877167199020539,
+      "learning_rate": 0.00010404884562745832,
+      "loss": 12.4911,
+      "step": 8620
+    },
+    {
+      "epoch": 1.5076284153005464,
+      "grad_norm": 0.992397691118107,
+      "learning_rate": 0.00010402998013940786,
+      "loss": 12.4921,
+      "step": 8621
+    },
+    {
+      "epoch": 1.5078032786885247,
+      "grad_norm": 1.0956608826809215,
+      "learning_rate": 0.00010401111450769334,
+      "loss": 12.4764,
+      "step": 8622
+    },
+    {
+      "epoch": 1.5079781420765026,
+      "grad_norm": 1.0647162393362157,
+      "learning_rate": 0.00010399224873298725,
+      "loss": 12.6425,
+      "step": 8623
+    },
+    {
+      "epoch": 1.508153005464481,
+      "grad_norm": 0.9919790118773141,
+      "learning_rate": 0.00010397338281596219,
+      "loss": 12.6299,
+      "step": 8624
+    },
+    {
+      "epoch": 1.508327868852459,
+      "grad_norm": 0.8017602448379579,
+      "learning_rate": 0.00010395451675729069,
+      "loss": 12.1785,
+      "step": 8625
+    },
+    {
+      "epoch": 1.5085027322404372,
+      "grad_norm": 1.0557181967581906,
+      "learning_rate": 0.00010393565055764524,
+      "loss": 12.5207,
+      "step": 8626
+    },
+    {
+      "epoch": 1.5086775956284153,
+      "grad_norm": 0.8169081785825584,
+      "learning_rate": 0.00010391678421769848,
+      "loss": 12.2814,
+      "step": 8627
+    },
+    {
+      "epoch": 1.5088524590163934,
+      "grad_norm": 1.034426444446928,
+      "learning_rate": 0.00010389791773812292,
+      "loss": 12.6176,
+      "step": 8628
+    },
+    {
+      "epoch": 1.5090273224043715,
+      "grad_norm": 0.8737784953201214,
+      "learning_rate": 0.00010387905111959117,
+      "loss": 12.5602,
+      "step": 8629
+    },
+    {
+      "epoch": 1.5092021857923497,
+      "grad_norm": 0.9629875169794179,
+      "learning_rate": 0.00010386018436277576,
+      "loss": 12.5666,
+      "step": 8630
+    },
+    {
+      "epoch": 1.509377049180328,
+      "grad_norm": 0.8489478432517394,
+      "learning_rate": 0.00010384131746834927,
+      "loss": 12.4782,
+      "step": 8631
+    },
+    {
+      "epoch": 1.509551912568306,
+      "grad_norm": 1.0859574350005199,
+      "learning_rate": 0.00010382245043698431,
+      "loss": 12.3396,
+      "step": 8632
+    },
+    {
+      "epoch": 1.5097267759562842,
+      "grad_norm": 0.8217936732064257,
+      "learning_rate": 0.00010380358326935349,
+      "loss": 12.4577,
+      "step": 8633
+    },
+    {
+      "epoch": 1.5099016393442621,
+      "grad_norm": 0.9927098843605152,
+      "learning_rate": 0.00010378471596612933,
+      "loss": 12.5088,
+      "step": 8634
+    },
+    {
+      "epoch": 1.5100765027322405,
+      "grad_norm": 0.8429665406657225,
+      "learning_rate": 0.0001037658485279845,
+      "loss": 12.5813,
+      "step": 8635
+    },
+    {
+      "epoch": 1.5102513661202186,
+      "grad_norm": 0.8265701215092331,
+      "learning_rate": 0.00010374698095559154,
+      "loss": 12.5493,
+      "step": 8636
+    },
+    {
+      "epoch": 1.5104262295081967,
+      "grad_norm": 0.8652278632089414,
+      "learning_rate": 0.00010372811324962313,
+      "loss": 12.3876,
+      "step": 8637
+    },
+    {
+      "epoch": 1.5106010928961748,
+      "grad_norm": 0.8878831187299165,
+      "learning_rate": 0.0001037092454107518,
+      "loss": 12.7006,
+      "step": 8638
+    },
+    {
+      "epoch": 1.510775956284153,
+      "grad_norm": 0.8495438224966414,
+      "learning_rate": 0.00010369037743965022,
+      "loss": 12.4601,
+      "step": 8639
+    },
+    {
+      "epoch": 1.5109508196721313,
+      "grad_norm": 0.8711327352171817,
+      "learning_rate": 0.00010367150933699099,
+      "loss": 12.6737,
+      "step": 8640
+    },
+    {
+      "epoch": 1.5111256830601092,
+      "grad_norm": 0.8877560046718799,
+      "learning_rate": 0.00010365264110344674,
+      "loss": 12.2774,
+      "step": 8641
+    },
+    {
+      "epoch": 1.5113005464480875,
+      "grad_norm": 0.8603948047824703,
+      "learning_rate": 0.00010363377273969011,
+      "loss": 12.4644,
+      "step": 8642
+    },
+    {
+      "epoch": 1.5114754098360654,
+      "grad_norm": 0.9369275918353296,
+      "learning_rate": 0.00010361490424639373,
+      "loss": 12.6257,
+      "step": 8643
+    },
+    {
+      "epoch": 1.5116502732240438,
+      "grad_norm": 0.9557847324961447,
+      "learning_rate": 0.00010359603562423024,
+      "loss": 12.2009,
+      "step": 8644
+    },
+    {
+      "epoch": 1.511825136612022,
+      "grad_norm": 0.9225441860501614,
+      "learning_rate": 0.00010357716687387227,
+      "loss": 12.417,
+      "step": 8645
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.878734161390327,
+      "learning_rate": 0.00010355829799599248,
+      "loss": 12.625,
+      "step": 8646
+    },
+    {
+      "epoch": 1.5121748633879781,
+      "grad_norm": 0.8710184544361252,
+      "learning_rate": 0.00010353942899126352,
+      "loss": 12.5593,
+      "step": 8647
+    },
+    {
+      "epoch": 1.5123497267759562,
+      "grad_norm": 0.9612989230425423,
+      "learning_rate": 0.00010352055986035806,
+      "loss": 12.6047,
+      "step": 8648
+    },
+    {
+      "epoch": 1.5125245901639346,
+      "grad_norm": 1.0696587275816931,
+      "learning_rate": 0.00010350169060394877,
+      "loss": 12.4681,
+      "step": 8649
+    },
+    {
+      "epoch": 1.5126994535519125,
+      "grad_norm": 0.8864383529399072,
+      "learning_rate": 0.00010348282122270827,
+      "loss": 12.5043,
+      "step": 8650
+    },
+    {
+      "epoch": 1.5128743169398908,
+      "grad_norm": 0.8128381785323298,
+      "learning_rate": 0.00010346395171730927,
+      "loss": 12.3854,
+      "step": 8651
+    },
+    {
+      "epoch": 1.5130491803278687,
+      "grad_norm": 0.8846358144719854,
+      "learning_rate": 0.00010344508208842445,
+      "loss": 12.521,
+      "step": 8652
+    },
+    {
+      "epoch": 1.513224043715847,
+      "grad_norm": 0.916910486064855,
+      "learning_rate": 0.00010342621233672647,
+      "loss": 12.6,
+      "step": 8653
+    },
+    {
+      "epoch": 1.5133989071038252,
+      "grad_norm": 0.8814254606506835,
+      "learning_rate": 0.00010340734246288804,
+      "loss": 12.475,
+      "step": 8654
+    },
+    {
+      "epoch": 1.5135737704918033,
+      "grad_norm": 0.8758832398419388,
+      "learning_rate": 0.00010338847246758182,
+      "loss": 12.6491,
+      "step": 8655
+    },
+    {
+      "epoch": 1.5137486338797814,
+      "grad_norm": 0.9328008287928363,
+      "learning_rate": 0.00010336960235148052,
+      "loss": 12.2148,
+      "step": 8656
+    },
+    {
+      "epoch": 1.5139234972677595,
+      "grad_norm": 0.9442208992983571,
+      "learning_rate": 0.00010335073211525684,
+      "loss": 12.4254,
+      "step": 8657
+    },
+    {
+      "epoch": 1.5140983606557377,
+      "grad_norm": 0.8885882040887102,
+      "learning_rate": 0.00010333186175958344,
+      "loss": 12.5811,
+      "step": 8658
+    },
+    {
+      "epoch": 1.5142732240437158,
+      "grad_norm": 0.8287174403721649,
+      "learning_rate": 0.0001033129912851331,
+      "loss": 12.6503,
+      "step": 8659
+    },
+    {
+      "epoch": 1.5144480874316941,
+      "grad_norm": 0.9043805546191482,
+      "learning_rate": 0.00010329412069257847,
+      "loss": 12.5299,
+      "step": 8660
+    },
+    {
+      "epoch": 1.514622950819672,
+      "grad_norm": 0.8458337455681579,
+      "learning_rate": 0.00010327524998259229,
+      "loss": 12.7322,
+      "step": 8661
+    },
+    {
+      "epoch": 1.5147978142076504,
+      "grad_norm": 0.8627504469547747,
+      "learning_rate": 0.00010325637915584728,
+      "loss": 12.5889,
+      "step": 8662
+    },
+    {
+      "epoch": 1.5149726775956283,
+      "grad_norm": 0.9144592093645277,
+      "learning_rate": 0.00010323750821301616,
+      "loss": 12.3555,
+      "step": 8663
+    },
+    {
+      "epoch": 1.5151475409836066,
+      "grad_norm": 0.8885233826196194,
+      "learning_rate": 0.00010321863715477165,
+      "loss": 12.5299,
+      "step": 8664
+    },
+    {
+      "epoch": 1.5153224043715847,
+      "grad_norm": 0.864253478530642,
+      "learning_rate": 0.00010319976598178649,
+      "loss": 12.6357,
+      "step": 8665
+    },
+    {
+      "epoch": 1.5154972677595628,
+      "grad_norm": 0.8580010683952881,
+      "learning_rate": 0.0001031808946947334,
+      "loss": 12.5093,
+      "step": 8666
+    },
+    {
+      "epoch": 1.515672131147541,
+      "grad_norm": 0.8672072731636734,
+      "learning_rate": 0.00010316202329428512,
+      "loss": 12.6601,
+      "step": 8667
+    },
+    {
+      "epoch": 1.515846994535519,
+      "grad_norm": 0.9294094535820496,
+      "learning_rate": 0.00010314315178111442,
+      "loss": 12.5227,
+      "step": 8668
+    },
+    {
+      "epoch": 1.5160218579234974,
+      "grad_norm": 0.851652181198173,
+      "learning_rate": 0.00010312428015589403,
+      "loss": 12.5364,
+      "step": 8669
+    },
+    {
+      "epoch": 1.5161967213114753,
+      "grad_norm": 0.9599323673714785,
+      "learning_rate": 0.0001031054084192967,
+      "loss": 12.606,
+      "step": 8670
+    },
+    {
+      "epoch": 1.5163715846994537,
+      "grad_norm": 0.962931548142491,
+      "learning_rate": 0.0001030865365719952,
+      "loss": 12.4016,
+      "step": 8671
+    },
+    {
+      "epoch": 1.5165464480874316,
+      "grad_norm": 0.9054030567412402,
+      "learning_rate": 0.0001030676646146623,
+      "loss": 12.3054,
+      "step": 8672
+    },
+    {
+      "epoch": 1.51672131147541,
+      "grad_norm": 1.0541276666484098,
+      "learning_rate": 0.00010304879254797071,
+      "loss": 12.502,
+      "step": 8673
+    },
+    {
+      "epoch": 1.516896174863388,
+      "grad_norm": 0.8660904686363017,
+      "learning_rate": 0.00010302992037259325,
+      "loss": 12.4511,
+      "step": 8674
+    },
+    {
+      "epoch": 1.5170710382513661,
+      "grad_norm": 1.1171702174880105,
+      "learning_rate": 0.00010301104808920263,
+      "loss": 12.5162,
+      "step": 8675
+    },
+    {
+      "epoch": 1.5172459016393443,
+      "grad_norm": 0.8285807870141748,
+      "learning_rate": 0.00010299217569847166,
+      "loss": 12.4511,
+      "step": 8676
+    },
+    {
+      "epoch": 1.5174207650273224,
+      "grad_norm": 1.0072162951330352,
+      "learning_rate": 0.00010297330320107314,
+      "loss": 12.431,
+      "step": 8677
+    },
+    {
+      "epoch": 1.5175956284153007,
+      "grad_norm": 0.8502378054791302,
+      "learning_rate": 0.00010295443059767986,
+      "loss": 12.5103,
+      "step": 8678
+    },
+    {
+      "epoch": 1.5177704918032786,
+      "grad_norm": 0.8713749535663187,
+      "learning_rate": 0.00010293555788896453,
+      "loss": 12.5518,
+      "step": 8679
+    },
+    {
+      "epoch": 1.517945355191257,
+      "grad_norm": 0.8092457321324069,
+      "learning_rate": 0.00010291668507560006,
+      "loss": 12.575,
+      "step": 8680
+    },
+    {
+      "epoch": 1.5181202185792348,
+      "grad_norm": 0.8664708636889497,
+      "learning_rate": 0.00010289781215825912,
+      "loss": 12.5515,
+      "step": 8681
+    },
+    {
+      "epoch": 1.5182950819672132,
+      "grad_norm": 0.8589952472388163,
+      "learning_rate": 0.0001028789391376146,
+      "loss": 12.257,
+      "step": 8682
+    },
+    {
+      "epoch": 1.5184699453551913,
+      "grad_norm": 0.8869461672934923,
+      "learning_rate": 0.00010286006601433922,
+      "loss": 12.567,
+      "step": 8683
+    },
+    {
+      "epoch": 1.5186448087431694,
+      "grad_norm": 0.9143880291078902,
+      "learning_rate": 0.00010284119278910586,
+      "loss": 12.746,
+      "step": 8684
+    },
+    {
+      "epoch": 1.5188196721311475,
+      "grad_norm": 0.7850952156604951,
+      "learning_rate": 0.0001028223194625873,
+      "loss": 12.5986,
+      "step": 8685
+    },
+    {
+      "epoch": 1.5189945355191257,
+      "grad_norm": 0.8439458950727965,
+      "learning_rate": 0.0001028034460354563,
+      "loss": 12.7532,
+      "step": 8686
+    },
+    {
+      "epoch": 1.5191693989071038,
+      "grad_norm": 0.8229892175252025,
+      "learning_rate": 0.00010278457250838578,
+      "loss": 12.4858,
+      "step": 8687
+    },
+    {
+      "epoch": 1.519344262295082,
+      "grad_norm": 0.8303752866159646,
+      "learning_rate": 0.00010276569888204847,
+      "loss": 12.4127,
+      "step": 8688
+    },
+    {
+      "epoch": 1.5195191256830602,
+      "grad_norm": 0.9289007078872603,
+      "learning_rate": 0.00010274682515711727,
+      "loss": 12.4485,
+      "step": 8689
+    },
+    {
+      "epoch": 1.5196939890710381,
+      "grad_norm": 0.7804486840392998,
+      "learning_rate": 0.00010272795133426491,
+      "loss": 12.2249,
+      "step": 8690
+    },
+    {
+      "epoch": 1.5198688524590165,
+      "grad_norm": 0.8710700812512094,
+      "learning_rate": 0.00010270907741416431,
+      "loss": 12.3086,
+      "step": 8691
+    },
+    {
+      "epoch": 1.5200437158469944,
+      "grad_norm": 0.9053458789652518,
+      "learning_rate": 0.00010269020339748825,
+      "loss": 12.6222,
+      "step": 8692
+    },
+    {
+      "epoch": 1.5202185792349727,
+      "grad_norm": 0.8738901223459843,
+      "learning_rate": 0.0001026713292849096,
+      "loss": 12.5501,
+      "step": 8693
+    },
+    {
+      "epoch": 1.5203934426229508,
+      "grad_norm": 0.9509771072359767,
+      "learning_rate": 0.00010265245507710114,
+      "loss": 12.5008,
+      "step": 8694
+    },
+    {
+      "epoch": 1.520568306010929,
+      "grad_norm": 0.9066516332022959,
+      "learning_rate": 0.00010263358077473578,
+      "loss": 12.5698,
+      "step": 8695
+    },
+    {
+      "epoch": 1.520743169398907,
+      "grad_norm": 0.9589324390288597,
+      "learning_rate": 0.00010261470637848633,
+      "loss": 12.3545,
+      "step": 8696
+    },
+    {
+      "epoch": 1.5209180327868852,
+      "grad_norm": 0.8406401131856995,
+      "learning_rate": 0.00010259583188902569,
+      "loss": 12.3896,
+      "step": 8697
+    },
+    {
+      "epoch": 1.5210928961748635,
+      "grad_norm": 0.827758627900525,
+      "learning_rate": 0.00010257695730702663,
+      "loss": 12.6114,
+      "step": 8698
+    },
+    {
+      "epoch": 1.5212677595628414,
+      "grad_norm": 0.7869805033202723,
+      "learning_rate": 0.00010255808263316209,
+      "loss": 12.6453,
+      "step": 8699
+    },
+    {
+      "epoch": 1.5214426229508198,
+      "grad_norm": 0.8151924856122014,
+      "learning_rate": 0.00010253920786810487,
+      "loss": 12.1755,
+      "step": 8700
+    },
+    {
+      "epoch": 1.5216174863387977,
+      "grad_norm": 0.8492925320180307,
+      "learning_rate": 0.00010252033301252786,
+      "loss": 12.259,
+      "step": 8701
+    },
+    {
+      "epoch": 1.521792349726776,
+      "grad_norm": 0.791231676017581,
+      "learning_rate": 0.00010250145806710391,
+      "loss": 12.0354,
+      "step": 8702
+    },
+    {
+      "epoch": 1.5219672131147541,
+      "grad_norm": 0.857359642073224,
+      "learning_rate": 0.00010248258303250595,
+      "loss": 12.5948,
+      "step": 8703
+    },
+    {
+      "epoch": 1.5221420765027323,
+      "grad_norm": 0.958978023594799,
+      "learning_rate": 0.00010246370790940676,
+      "loss": 12.3718,
+      "step": 8704
+    },
+    {
+      "epoch": 1.5223169398907104,
+      "grad_norm": 0.873305752261052,
+      "learning_rate": 0.00010244483269847926,
+      "loss": 12.3965,
+      "step": 8705
+    },
+    {
+      "epoch": 1.5224918032786885,
+      "grad_norm": 0.8621068726313083,
+      "learning_rate": 0.00010242595740039636,
+      "loss": 12.4664,
+      "step": 8706
+    },
+    {
+      "epoch": 1.5226666666666666,
+      "grad_norm": 0.9032543908726034,
+      "learning_rate": 0.00010240708201583089,
+      "loss": 12.3694,
+      "step": 8707
+    },
+    {
+      "epoch": 1.5228415300546447,
+      "grad_norm": 0.8938829048609823,
+      "learning_rate": 0.00010238820654545578,
+      "loss": 12.5588,
+      "step": 8708
+    },
+    {
+      "epoch": 1.523016393442623,
+      "grad_norm": 0.9793068232330431,
+      "learning_rate": 0.00010236933098994388,
+      "loss": 12.6965,
+      "step": 8709
+    },
+    {
+      "epoch": 1.523191256830601,
+      "grad_norm": 0.9668073175295789,
+      "learning_rate": 0.00010235045534996812,
+      "loss": 12.5533,
+      "step": 8710
+    },
+    {
+      "epoch": 1.5233661202185793,
+      "grad_norm": 0.7771544692642193,
+      "learning_rate": 0.00010233157962620134,
+      "loss": 12.5441,
+      "step": 8711
+    },
+    {
+      "epoch": 1.5235409836065574,
+      "grad_norm": 0.9264560632886151,
+      "learning_rate": 0.0001023127038193165,
+      "loss": 12.586,
+      "step": 8712
+    },
+    {
+      "epoch": 1.5237158469945355,
+      "grad_norm": 1.094389796621491,
+      "learning_rate": 0.00010229382792998646,
+      "loss": 12.248,
+      "step": 8713
+    },
+    {
+      "epoch": 1.5238907103825137,
+      "grad_norm": 0.8757847751050941,
+      "learning_rate": 0.00010227495195888411,
+      "loss": 12.2501,
+      "step": 8714
+    },
+    {
+      "epoch": 1.5240655737704918,
+      "grad_norm": 0.9415752018278628,
+      "learning_rate": 0.0001022560759066824,
+      "loss": 12.5739,
+      "step": 8715
+    },
+    {
+      "epoch": 1.52424043715847,
+      "grad_norm": 0.9088876463668217,
+      "learning_rate": 0.00010223719977405424,
+      "loss": 12.5398,
+      "step": 8716
+    },
+    {
+      "epoch": 1.524415300546448,
+      "grad_norm": 1.111754470296551,
+      "learning_rate": 0.0001022183235616725,
+      "loss": 12.6236,
+      "step": 8717
+    },
+    {
+      "epoch": 1.5245901639344264,
+      "grad_norm": 0.9204610050181499,
+      "learning_rate": 0.00010219944727021011,
+      "loss": 12.4739,
+      "step": 8718
+    },
+    {
+      "epoch": 1.5247650273224043,
+      "grad_norm": 0.9085548268769885,
+      "learning_rate": 0.00010218057090034004,
+      "loss": 12.6623,
+      "step": 8719
+    },
+    {
+      "epoch": 1.5249398907103826,
+      "grad_norm": 1.1013458281324078,
+      "learning_rate": 0.00010216169445273512,
+      "loss": 12.4837,
+      "step": 8720
+    },
+    {
+      "epoch": 1.5251147540983605,
+      "grad_norm": 0.8517396822686177,
+      "learning_rate": 0.0001021428179280683,
+      "loss": 12.4604,
+      "step": 8721
+    },
+    {
+      "epoch": 1.5252896174863388,
+      "grad_norm": 1.0349393885057685,
+      "learning_rate": 0.00010212394132701255,
+      "loss": 12.5353,
+      "step": 8722
+    },
+    {
+      "epoch": 1.525464480874317,
+      "grad_norm": 0.7577994910921078,
+      "learning_rate": 0.0001021050646502408,
+      "loss": 12.4606,
+      "step": 8723
+    },
+    {
+      "epoch": 1.525639344262295,
+      "grad_norm": 0.8717855032423774,
+      "learning_rate": 0.00010208618789842593,
+      "loss": 12.3735,
+      "step": 8724
+    },
+    {
+      "epoch": 1.5258142076502732,
+      "grad_norm": 0.8701124960567699,
+      "learning_rate": 0.00010206731107224091,
+      "loss": 12.6166,
+      "step": 8725
+    },
+    {
+      "epoch": 1.5259890710382513,
+      "grad_norm": 0.8584280730306354,
+      "learning_rate": 0.00010204843417235862,
+      "loss": 12.3721,
+      "step": 8726
+    },
+    {
+      "epoch": 1.5261639344262297,
+      "grad_norm": 0.82298860679162,
+      "learning_rate": 0.00010202955719945211,
+      "loss": 12.5042,
+      "step": 8727
+    },
+    {
+      "epoch": 1.5263387978142076,
+      "grad_norm": 0.8760896047856774,
+      "learning_rate": 0.00010201068015419422,
+      "loss": 12.4873,
+      "step": 8728
+    },
+    {
+      "epoch": 1.526513661202186,
+      "grad_norm": 0.8993766409219954,
+      "learning_rate": 0.00010199180303725795,
+      "loss": 12.3262,
+      "step": 8729
+    },
+    {
+      "epoch": 1.5266885245901638,
+      "grad_norm": 0.9411953377303098,
+      "learning_rate": 0.00010197292584931622,
+      "loss": 12.5335,
+      "step": 8730
+    },
+    {
+      "epoch": 1.5268633879781421,
+      "grad_norm": 0.934859505185347,
+      "learning_rate": 0.00010195404859104196,
+      "loss": 12.5066,
+      "step": 8731
+    },
+    {
+      "epoch": 1.5270382513661203,
+      "grad_norm": 1.0239981826116673,
+      "learning_rate": 0.00010193517126310819,
+      "loss": 12.694,
+      "step": 8732
+    },
+    {
+      "epoch": 1.5272131147540984,
+      "grad_norm": 0.8376916906273218,
+      "learning_rate": 0.00010191629386618782,
+      "loss": 12.4754,
+      "step": 8733
+    },
+    {
+      "epoch": 1.5273879781420765,
+      "grad_norm": 1.0666174544320797,
+      "learning_rate": 0.0001018974164009538,
+      "loss": 12.5312,
+      "step": 8734
+    },
+    {
+      "epoch": 1.5275628415300546,
+      "grad_norm": 0.9356248515194583,
+      "learning_rate": 0.00010187853886807912,
+      "loss": 12.5437,
+      "step": 8735
+    },
+    {
+      "epoch": 1.5277377049180327,
+      "grad_norm": 1.0985742559159213,
+      "learning_rate": 0.00010185966126823672,
+      "loss": 12.5601,
+      "step": 8736
+    },
+    {
+      "epoch": 1.5279125683060109,
+      "grad_norm": 0.7704810152378437,
+      "learning_rate": 0.00010184078360209955,
+      "loss": 12.2448,
+      "step": 8737
+    },
+    {
+      "epoch": 1.5280874316939892,
+      "grad_norm": 0.9794240926081157,
+      "learning_rate": 0.00010182190587034061,
+      "loss": 12.4099,
+      "step": 8738
+    },
+    {
+      "epoch": 1.528262295081967,
+      "grad_norm": 0.8673544639856227,
+      "learning_rate": 0.00010180302807363284,
+      "loss": 12.4384,
+      "step": 8739
+    },
+    {
+      "epoch": 1.5284371584699454,
+      "grad_norm": 0.7951178236471188,
+      "learning_rate": 0.00010178415021264924,
+      "loss": 12.3561,
+      "step": 8740
+    },
+    {
+      "epoch": 1.5286120218579233,
+      "grad_norm": 1.0779108701962834,
+      "learning_rate": 0.00010176527228806277,
+      "loss": 12.3083,
+      "step": 8741
+    },
+    {
+      "epoch": 1.5287868852459017,
+      "grad_norm": 0.7949567212544616,
+      "learning_rate": 0.0001017463943005464,
+      "loss": 12.1871,
+      "step": 8742
+    },
+    {
+      "epoch": 1.5289617486338798,
+      "grad_norm": 0.9048945551700367,
+      "learning_rate": 0.0001017275162507731,
+      "loss": 12.217,
+      "step": 8743
+    },
+    {
+      "epoch": 1.529136612021858,
+      "grad_norm": 0.9770795831810701,
+      "learning_rate": 0.00010170863813941591,
+      "loss": 12.3164,
+      "step": 8744
+    },
+    {
+      "epoch": 1.529311475409836,
+      "grad_norm": 0.8773980947635844,
+      "learning_rate": 0.00010168975996714772,
+      "loss": 12.5615,
+      "step": 8745
+    },
+    {
+      "epoch": 1.5294863387978141,
+      "grad_norm": 1.205361118569565,
+      "learning_rate": 0.00010167088173464159,
+      "loss": 12.6899,
+      "step": 8746
+    },
+    {
+      "epoch": 1.5296612021857925,
+      "grad_norm": 0.7954538655309519,
+      "learning_rate": 0.00010165200344257047,
+      "loss": 12.4721,
+      "step": 8747
+    },
+    {
+      "epoch": 1.5298360655737704,
+      "grad_norm": 1.1040433025112915,
+      "learning_rate": 0.00010163312509160737,
+      "loss": 11.9795,
+      "step": 8748
+    },
+    {
+      "epoch": 1.5300109289617487,
+      "grad_norm": 0.9004246944033603,
+      "learning_rate": 0.00010161424668242523,
+      "loss": 12.4346,
+      "step": 8749
+    },
+    {
+      "epoch": 1.5301857923497266,
+      "grad_norm": 0.9756792997010076,
+      "learning_rate": 0.00010159536821569709,
+      "loss": 12.6367,
+      "step": 8750
+    },
+    {
+      "epoch": 1.530360655737705,
+      "grad_norm": 0.9060741515861062,
+      "learning_rate": 0.00010157648969209596,
+      "loss": 12.4788,
+      "step": 8751
+    },
+    {
+      "epoch": 1.530535519125683,
+      "grad_norm": 0.838128038230085,
+      "learning_rate": 0.0001015576111122948,
+      "loss": 12.5124,
+      "step": 8752
+    },
+    {
+      "epoch": 1.5307103825136612,
+      "grad_norm": 1.0268059594828656,
+      "learning_rate": 0.00010153873247696663,
+      "loss": 12.3003,
+      "step": 8753
+    },
+    {
+      "epoch": 1.5308852459016393,
+      "grad_norm": 0.9362550022002653,
+      "learning_rate": 0.00010151985378678446,
+      "loss": 12.5202,
+      "step": 8754
+    },
+    {
+      "epoch": 1.5310601092896174,
+      "grad_norm": 0.8877627076485769,
+      "learning_rate": 0.00010150097504242126,
+      "loss": 12.3675,
+      "step": 8755
+    },
+    {
+      "epoch": 1.5312349726775958,
+      "grad_norm": 0.8691856575896528,
+      "learning_rate": 0.00010148209624455005,
+      "loss": 12.4828,
+      "step": 8756
+    },
+    {
+      "epoch": 1.5314098360655737,
+      "grad_norm": 0.956777611898186,
+      "learning_rate": 0.00010146321739384385,
+      "loss": 12.2288,
+      "step": 8757
+    },
+    {
+      "epoch": 1.531584699453552,
+      "grad_norm": 0.792382076916739,
+      "learning_rate": 0.00010144433849097566,
+      "loss": 12.4351,
+      "step": 8758
+    },
+    {
+      "epoch": 1.53175956284153,
+      "grad_norm": 1.012738328952214,
+      "learning_rate": 0.00010142545953661846,
+      "loss": 12.596,
+      "step": 8759
+    },
+    {
+      "epoch": 1.5319344262295083,
+      "grad_norm": 0.9380342676394612,
+      "learning_rate": 0.00010140658053144531,
+      "loss": 12.5147,
+      "step": 8760
+    },
+    {
+      "epoch": 1.5321092896174864,
+      "grad_norm": 1.0139866492832168,
+      "learning_rate": 0.00010138770147612924,
+      "loss": 12.563,
+      "step": 8761
+    },
+    {
+      "epoch": 1.5322841530054645,
+      "grad_norm": 0.865820059506088,
+      "learning_rate": 0.00010136882237134319,
+      "loss": 12.4406,
+      "step": 8762
+    },
+    {
+      "epoch": 1.5324590163934426,
+      "grad_norm": 0.8587259150217248,
+      "learning_rate": 0.00010134994321776025,
+      "loss": 12.382,
+      "step": 8763
+    },
+    {
+      "epoch": 1.5326338797814207,
+      "grad_norm": 0.8739927923367914,
+      "learning_rate": 0.00010133106401605341,
+      "loss": 12.1941,
+      "step": 8764
+    },
+    {
+      "epoch": 1.5328087431693989,
+      "grad_norm": 0.8299171447867978,
+      "learning_rate": 0.00010131218476689571,
+      "loss": 12.4067,
+      "step": 8765
+    },
+    {
+      "epoch": 1.532983606557377,
+      "grad_norm": 0.9145339877366716,
+      "learning_rate": 0.00010129330547096013,
+      "loss": 12.2467,
+      "step": 8766
+    },
+    {
+      "epoch": 1.5331584699453553,
+      "grad_norm": 0.838624417686024,
+      "learning_rate": 0.00010127442612891971,
+      "loss": 12.4518,
+      "step": 8767
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 0.9511478885621287,
+      "learning_rate": 0.0001012555467414475,
+      "loss": 12.4555,
+      "step": 8768
+    },
+    {
+      "epoch": 1.5335081967213116,
+      "grad_norm": 0.9130709989362289,
+      "learning_rate": 0.0001012366673092165,
+      "loss": 12.5248,
+      "step": 8769
+    },
+    {
+      "epoch": 1.5336830601092895,
+      "grad_norm": 1.16561831975681,
+      "learning_rate": 0.00010121778783289979,
+      "loss": 12.5655,
+      "step": 8770
+    },
+    {
+      "epoch": 1.5338579234972678,
+      "grad_norm": 0.8977666358996825,
+      "learning_rate": 0.00010119890831317034,
+      "loss": 12.4406,
+      "step": 8771
+    },
+    {
+      "epoch": 1.534032786885246,
+      "grad_norm": 0.906660043610029,
+      "learning_rate": 0.00010118002875070123,
+      "loss": 12.5858,
+      "step": 8772
+    },
+    {
+      "epoch": 1.534207650273224,
+      "grad_norm": 0.9240009995988522,
+      "learning_rate": 0.00010116114914616546,
+      "loss": 11.965,
+      "step": 8773
+    },
+    {
+      "epoch": 1.5343825136612022,
+      "grad_norm": 0.9352421133749256,
+      "learning_rate": 0.00010114226950023608,
+      "loss": 12.6056,
+      "step": 8774
+    },
+    {
+      "epoch": 1.5345573770491803,
+      "grad_norm": 1.033127764020646,
+      "learning_rate": 0.00010112338981358613,
+      "loss": 12.6504,
+      "step": 8775
+    },
+    {
+      "epoch": 1.5347322404371586,
+      "grad_norm": 0.8299106019994102,
+      "learning_rate": 0.00010110451008688863,
+      "loss": 12.3964,
+      "step": 8776
+    },
+    {
+      "epoch": 1.5349071038251365,
+      "grad_norm": 1.0749805851098289,
+      "learning_rate": 0.00010108563032081666,
+      "loss": 12.6103,
+      "step": 8777
+    },
+    {
+      "epoch": 1.5350819672131149,
+      "grad_norm": 0.8189775779194022,
+      "learning_rate": 0.00010106675051604325,
+      "loss": 12.1776,
+      "step": 8778
+    },
+    {
+      "epoch": 1.5352568306010927,
+      "grad_norm": 1.0605979747289829,
+      "learning_rate": 0.00010104787067324137,
+      "loss": 12.523,
+      "step": 8779
+    },
+    {
+      "epoch": 1.535431693989071,
+      "grad_norm": 0.8623901002502062,
+      "learning_rate": 0.00010102899079308418,
+      "loss": 12.6644,
+      "step": 8780
+    },
+    {
+      "epoch": 1.5356065573770492,
+      "grad_norm": 0.893142589423873,
+      "learning_rate": 0.00010101011087624468,
+      "loss": 12.6044,
+      "step": 8781
+    },
+    {
+      "epoch": 1.5357814207650273,
+      "grad_norm": 0.974441319589175,
+      "learning_rate": 0.00010099123092339587,
+      "loss": 12.6272,
+      "step": 8782
+    },
+    {
+      "epoch": 1.5359562841530054,
+      "grad_norm": 0.8220641488725463,
+      "learning_rate": 0.00010097235093521084,
+      "loss": 12.6098,
+      "step": 8783
+    },
+    {
+      "epoch": 1.5361311475409836,
+      "grad_norm": 0.977922673600657,
+      "learning_rate": 0.00010095347091236265,
+      "loss": 12.3426,
+      "step": 8784
+    },
+    {
+      "epoch": 1.536306010928962,
+      "grad_norm": 0.9269014688920683,
+      "learning_rate": 0.0001009345908555243,
+      "loss": 12.5232,
+      "step": 8785
+    },
+    {
+      "epoch": 1.5364808743169398,
+      "grad_norm": 0.9555767788575813,
+      "learning_rate": 0.0001009157107653689,
+      "loss": 12.299,
+      "step": 8786
+    },
+    {
+      "epoch": 1.5366557377049181,
+      "grad_norm": 1.0027193573107882,
+      "learning_rate": 0.00010089683064256952,
+      "loss": 12.3999,
+      "step": 8787
+    },
+    {
+      "epoch": 1.536830601092896,
+      "grad_norm": 0.7972746650702294,
+      "learning_rate": 0.00010087795048779911,
+      "loss": 12.2387,
+      "step": 8788
+    },
+    {
+      "epoch": 1.5370054644808744,
+      "grad_norm": 1.0442899370818888,
+      "learning_rate": 0.00010085907030173084,
+      "loss": 12.3519,
+      "step": 8789
+    },
+    {
+      "epoch": 1.5371803278688525,
+      "grad_norm": 0.9426023229672832,
+      "learning_rate": 0.00010084019008503769,
+      "loss": 12.6934,
+      "step": 8790
+    },
+    {
+      "epoch": 1.5373551912568306,
+      "grad_norm": 0.8513189877096606,
+      "learning_rate": 0.00010082130983839278,
+      "loss": 12.5961,
+      "step": 8791
+    },
+    {
+      "epoch": 1.5375300546448087,
+      "grad_norm": 1.0407663402992882,
+      "learning_rate": 0.0001008024295624691,
+      "loss": 12.437,
+      "step": 8792
+    },
+    {
+      "epoch": 1.5377049180327869,
+      "grad_norm": 0.832404818199253,
+      "learning_rate": 0.00010078354925793976,
+      "loss": 12.4444,
+      "step": 8793
+    },
+    {
+      "epoch": 1.537879781420765,
+      "grad_norm": 0.9029675302328394,
+      "learning_rate": 0.00010076466892547781,
+      "loss": 12.6059,
+      "step": 8794
+    },
+    {
+      "epoch": 1.538054644808743,
+      "grad_norm": 0.9238303975368473,
+      "learning_rate": 0.00010074578856575627,
+      "loss": 12.6738,
+      "step": 8795
+    },
+    {
+      "epoch": 1.5382295081967214,
+      "grad_norm": 0.7903431719065386,
+      "learning_rate": 0.00010072690817944828,
+      "loss": 12.3104,
+      "step": 8796
+    },
+    {
+      "epoch": 1.5384043715846993,
+      "grad_norm": 0.8985967039076203,
+      "learning_rate": 0.00010070802776722683,
+      "loss": 12.2306,
+      "step": 8797
+    },
+    {
+      "epoch": 1.5385792349726777,
+      "grad_norm": 0.8934880154888546,
+      "learning_rate": 0.00010068914732976504,
+      "loss": 12.5915,
+      "step": 8798
+    },
+    {
+      "epoch": 1.5387540983606556,
+      "grad_norm": 0.8219472704773545,
+      "learning_rate": 0.00010067026686773593,
+      "loss": 12.5812,
+      "step": 8799
+    },
+    {
+      "epoch": 1.538928961748634,
+      "grad_norm": 1.0404350692404587,
+      "learning_rate": 0.00010065138638181262,
+      "loss": 12.4085,
+      "step": 8800
+    },
+    {
+      "epoch": 1.539103825136612,
+      "grad_norm": 0.8877705540021121,
+      "learning_rate": 0.00010063250587266813,
+      "loss": 12.4564,
+      "step": 8801
+    },
+    {
+      "epoch": 1.5392786885245902,
+      "grad_norm": 1.0175042554780118,
+      "learning_rate": 0.00010061362534097556,
+      "loss": 12.4896,
+      "step": 8802
+    },
+    {
+      "epoch": 1.5394535519125683,
+      "grad_norm": 0.9851734373541233,
+      "learning_rate": 0.00010059474478740794,
+      "loss": 12.6063,
+      "step": 8803
+    },
+    {
+      "epoch": 1.5396284153005464,
+      "grad_norm": 1.0030100322902102,
+      "learning_rate": 0.00010057586421263836,
+      "loss": 12.732,
+      "step": 8804
+    },
+    {
+      "epoch": 1.5398032786885247,
+      "grad_norm": 0.9873104630668095,
+      "learning_rate": 0.0001005569836173399,
+      "loss": 12.4985,
+      "step": 8805
+    },
+    {
+      "epoch": 1.5399781420765026,
+      "grad_norm": 0.8472868953247298,
+      "learning_rate": 0.00010053810300218564,
+      "loss": 12.6111,
+      "step": 8806
+    },
+    {
+      "epoch": 1.540153005464481,
+      "grad_norm": 1.0650641893714754,
+      "learning_rate": 0.00010051922236784863,
+      "loss": 12.4468,
+      "step": 8807
+    },
+    {
+      "epoch": 1.5403278688524589,
+      "grad_norm": 0.8999864486801297,
+      "learning_rate": 0.00010050034171500196,
+      "loss": 12.332,
+      "step": 8808
+    },
+    {
+      "epoch": 1.5405027322404372,
+      "grad_norm": 0.8770151172026236,
+      "learning_rate": 0.00010048146104431866,
+      "loss": 12.5405,
+      "step": 8809
+    },
+    {
+      "epoch": 1.5406775956284153,
+      "grad_norm": 0.9856594671217037,
+      "learning_rate": 0.00010046258035647187,
+      "loss": 12.1747,
+      "step": 8810
+    },
+    {
+      "epoch": 1.5408524590163934,
+      "grad_norm": 0.9492807482429887,
+      "learning_rate": 0.00010044369965213461,
+      "loss": 12.2329,
+      "step": 8811
+    },
+    {
+      "epoch": 1.5410273224043716,
+      "grad_norm": 1.0212692124387595,
+      "learning_rate": 0.00010042481893197997,
+      "loss": 12.2993,
+      "step": 8812
+    },
+    {
+      "epoch": 1.5412021857923497,
+      "grad_norm": 0.9993383338662069,
+      "learning_rate": 0.00010040593819668107,
+      "loss": 12.3194,
+      "step": 8813
+    },
+    {
+      "epoch": 1.541377049180328,
+      "grad_norm": 0.8913334485256652,
+      "learning_rate": 0.0001003870574469109,
+      "loss": 12.5227,
+      "step": 8814
+    },
+    {
+      "epoch": 1.541551912568306,
+      "grad_norm": 0.9056243269154375,
+      "learning_rate": 0.00010036817668334262,
+      "loss": 12.5697,
+      "step": 8815
+    },
+    {
+      "epoch": 1.5417267759562843,
+      "grad_norm": 0.942047688664532,
+      "learning_rate": 0.00010034929590664925,
+      "loss": 12.1829,
+      "step": 8816
+    },
+    {
+      "epoch": 1.5419016393442622,
+      "grad_norm": 0.9298039575245354,
+      "learning_rate": 0.00010033041511750391,
+      "loss": 12.4624,
+      "step": 8817
+    },
+    {
+      "epoch": 1.5420765027322405,
+      "grad_norm": 0.9174868545788346,
+      "learning_rate": 0.00010031153431657962,
+      "loss": 12.7226,
+      "step": 8818
+    },
+    {
+      "epoch": 1.5422513661202186,
+      "grad_norm": 0.882654377095613,
+      "learning_rate": 0.00010029265350454956,
+      "loss": 12.5461,
+      "step": 8819
+    },
+    {
+      "epoch": 1.5424262295081967,
+      "grad_norm": 0.8493737575030523,
+      "learning_rate": 0.00010027377268208668,
+      "loss": 12.6397,
+      "step": 8820
+    },
+    {
+      "epoch": 1.5426010928961749,
+      "grad_norm": 0.8653315200812223,
+      "learning_rate": 0.00010025489184986413,
+      "loss": 12.5584,
+      "step": 8821
+    },
+    {
+      "epoch": 1.542775956284153,
+      "grad_norm": 0.8389727446618159,
+      "learning_rate": 0.000100236011008555,
+      "loss": 12.2783,
+      "step": 8822
+    },
+    {
+      "epoch": 1.542950819672131,
+      "grad_norm": 0.8740235745176038,
+      "learning_rate": 0.00010021713015883237,
+      "loss": 12.5347,
+      "step": 8823
+    },
+    {
+      "epoch": 1.5431256830601092,
+      "grad_norm": 0.8567735682021159,
+      "learning_rate": 0.00010019824930136927,
+      "loss": 12.4828,
+      "step": 8824
+    },
+    {
+      "epoch": 1.5433005464480876,
+      "grad_norm": 0.8673178041638963,
+      "learning_rate": 0.00010017936843683884,
+      "loss": 12.4762,
+      "step": 8825
+    },
+    {
+      "epoch": 1.5434754098360655,
+      "grad_norm": 0.8614855505387793,
+      "learning_rate": 0.00010016048756591412,
+      "loss": 12.4501,
+      "step": 8826
+    },
+    {
+      "epoch": 1.5436502732240438,
+      "grad_norm": 1.0526056386503218,
+      "learning_rate": 0.00010014160668926824,
+      "loss": 12.7379,
+      "step": 8827
+    },
+    {
+      "epoch": 1.5438251366120217,
+      "grad_norm": 0.8392781739557277,
+      "learning_rate": 0.00010012272580757423,
+      "loss": 12.3297,
+      "step": 8828
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 0.896508970604959,
+      "learning_rate": 0.00010010384492150516,
+      "loss": 12.5345,
+      "step": 8829
+    },
+    {
+      "epoch": 1.5441748633879782,
+      "grad_norm": 0.8672413820781159,
+      "learning_rate": 0.00010008496403173415,
+      "loss": 12.5226,
+      "step": 8830
+    },
+    {
+      "epoch": 1.5443497267759563,
+      "grad_norm": 0.8589334261864096,
+      "learning_rate": 0.00010006608313893427,
+      "loss": 12.7431,
+      "step": 8831
+    },
+    {
+      "epoch": 1.5445245901639344,
+      "grad_norm": 0.9381424052145109,
+      "learning_rate": 0.00010004720224377861,
+      "loss": 12.4516,
+      "step": 8832
+    },
+    {
+      "epoch": 1.5446994535519125,
+      "grad_norm": 0.8658836471603493,
+      "learning_rate": 0.00010002832134694024,
+      "loss": 12.5433,
+      "step": 8833
+    },
+    {
+      "epoch": 1.5448743169398909,
+      "grad_norm": 1.0355854910153963,
+      "learning_rate": 0.00010000944044909229,
+      "loss": 12.5556,
+      "step": 8834
+    },
+    {
+      "epoch": 1.5450491803278688,
+      "grad_norm": 0.8763382731215065,
+      "learning_rate": 9.999055955090777e-05,
+      "loss": 12.5109,
+      "step": 8835
+    },
+    {
+      "epoch": 1.545224043715847,
+      "grad_norm": 0.8920833386212459,
+      "learning_rate": 9.997167865305977e-05,
+      "loss": 12.6132,
+      "step": 8836
+    },
+    {
+      "epoch": 1.545398907103825,
+      "grad_norm": 0.9076204150279168,
+      "learning_rate": 9.99527977562214e-05,
+      "loss": 12.5039,
+      "step": 8837
+    },
+    {
+      "epoch": 1.5455737704918033,
+      "grad_norm": 0.9696970002357128,
+      "learning_rate": 9.993391686106574e-05,
+      "loss": 12.4132,
+      "step": 8838
+    },
+    {
+      "epoch": 1.5457486338797815,
+      "grad_norm": 0.9285078680746449,
+      "learning_rate": 9.991503596826586e-05,
+      "loss": 12.4851,
+      "step": 8839
+    },
+    {
+      "epoch": 1.5459234972677596,
+      "grad_norm": 0.8949619534210004,
+      "learning_rate": 9.989615507849485e-05,
+      "loss": 12.3289,
+      "step": 8840
+    },
+    {
+      "epoch": 1.5460983606557377,
+      "grad_norm": 0.875721925431564,
+      "learning_rate": 9.98772741924258e-05,
+      "loss": 12.4953,
+      "step": 8841
+    },
+    {
+      "epoch": 1.5462732240437158,
+      "grad_norm": 0.8285945621823395,
+      "learning_rate": 9.985839331073177e-05,
+      "loss": 12.3683,
+      "step": 8842
+    },
+    {
+      "epoch": 1.5464480874316942,
+      "grad_norm": 0.8702954361637661,
+      "learning_rate": 9.983951243408589e-05,
+      "loss": 12.4646,
+      "step": 8843
+    },
+    {
+      "epoch": 1.546622950819672,
+      "grad_norm": 0.8696834400845689,
+      "learning_rate": 9.982063156316117e-05,
+      "loss": 12.4753,
+      "step": 8844
+    },
+    {
+      "epoch": 1.5467978142076504,
+      "grad_norm": 0.8430763294461332,
+      "learning_rate": 9.980175069863073e-05,
+      "loss": 12.5093,
+      "step": 8845
+    },
+    {
+      "epoch": 1.5469726775956283,
+      "grad_norm": 0.8293266429462607,
+      "learning_rate": 9.978286984116768e-05,
+      "loss": 12.422,
+      "step": 8846
+    },
+    {
+      "epoch": 1.5471475409836066,
+      "grad_norm": 0.8341846705784909,
+      "learning_rate": 9.976398899144503e-05,
+      "loss": 12.5706,
+      "step": 8847
+    },
+    {
+      "epoch": 1.5473224043715847,
+      "grad_norm": 0.8449114927385544,
+      "learning_rate": 9.974510815013588e-05,
+      "loss": 12.3122,
+      "step": 8848
+    },
+    {
+      "epoch": 1.5474972677595629,
+      "grad_norm": 0.871814609192964,
+      "learning_rate": 9.972622731791334e-05,
+      "loss": 12.2792,
+      "step": 8849
+    },
+    {
+      "epoch": 1.547672131147541,
+      "grad_norm": 0.9013312763225872,
+      "learning_rate": 9.97073464954505e-05,
+      "loss": 12.7615,
+      "step": 8850
+    },
+    {
+      "epoch": 1.547846994535519,
+      "grad_norm": 0.8471576073724528,
+      "learning_rate": 9.968846568342039e-05,
+      "loss": 12.6157,
+      "step": 8851
+    },
+    {
+      "epoch": 1.5480218579234972,
+      "grad_norm": 0.8822383765039529,
+      "learning_rate": 9.966958488249611e-05,
+      "loss": 12.5626,
+      "step": 8852
+    },
+    {
+      "epoch": 1.5481967213114753,
+      "grad_norm": 0.8397756221061787,
+      "learning_rate": 9.965070409335074e-05,
+      "loss": 12.6414,
+      "step": 8853
+    },
+    {
+      "epoch": 1.5483715846994537,
+      "grad_norm": 0.8627660777161763,
+      "learning_rate": 9.963182331665743e-05,
+      "loss": 12.5116,
+      "step": 8854
+    },
+    {
+      "epoch": 1.5485464480874316,
+      "grad_norm": 0.8880817186633981,
+      "learning_rate": 9.961294255308912e-05,
+      "loss": 12.2733,
+      "step": 8855
+    },
+    {
+      "epoch": 1.54872131147541,
+      "grad_norm": 0.8820772774152056,
+      "learning_rate": 9.959406180331895e-05,
+      "loss": 12.3754,
+      "step": 8856
+    },
+    {
+      "epoch": 1.5488961748633878,
+      "grad_norm": 0.8714086466111773,
+      "learning_rate": 9.957518106802003e-05,
+      "loss": 12.717,
+      "step": 8857
+    },
+    {
+      "epoch": 1.5490710382513662,
+      "grad_norm": 0.8807426475425526,
+      "learning_rate": 9.95563003478654e-05,
+      "loss": 12.6133,
+      "step": 8858
+    },
+    {
+      "epoch": 1.5492459016393443,
+      "grad_norm": 0.8963547893685956,
+      "learning_rate": 9.953741964352816e-05,
+      "loss": 12.422,
+      "step": 8859
+    },
+    {
+      "epoch": 1.5494207650273224,
+      "grad_norm": 0.9011514395505278,
+      "learning_rate": 9.951853895568134e-05,
+      "loss": 12.483,
+      "step": 8860
+    },
+    {
+      "epoch": 1.5495956284153005,
+      "grad_norm": 0.8951855347206065,
+      "learning_rate": 9.949965828499809e-05,
+      "loss": 12.4506,
+      "step": 8861
+    },
+    {
+      "epoch": 1.5497704918032786,
+      "grad_norm": 0.9761813889459976,
+      "learning_rate": 9.94807776321514e-05,
+      "loss": 12.6563,
+      "step": 8862
+    },
+    {
+      "epoch": 1.549945355191257,
+      "grad_norm": 0.8881471088393866,
+      "learning_rate": 9.946189699781438e-05,
+      "loss": 12.4952,
+      "step": 8863
+    },
+    {
+      "epoch": 1.5501202185792349,
+      "grad_norm": 0.8952224582227682,
+      "learning_rate": 9.94430163826601e-05,
+      "loss": 12.3704,
+      "step": 8864
+    },
+    {
+      "epoch": 1.5502950819672132,
+      "grad_norm": 0.9157034109272056,
+      "learning_rate": 9.942413578736166e-05,
+      "loss": 12.6923,
+      "step": 8865
+    },
+    {
+      "epoch": 1.5504699453551911,
+      "grad_norm": 1.0051322015789124,
+      "learning_rate": 9.940525521259208e-05,
+      "loss": 12.6365,
+      "step": 8866
+    },
+    {
+      "epoch": 1.5506448087431695,
+      "grad_norm": 0.9108063698119713,
+      "learning_rate": 9.938637465902447e-05,
+      "loss": 12.3484,
+      "step": 8867
+    },
+    {
+      "epoch": 1.5508196721311476,
+      "grad_norm": 0.9765680071072966,
+      "learning_rate": 9.936749412733188e-05,
+      "loss": 12.58,
+      "step": 8868
+    },
+    {
+      "epoch": 1.5509945355191257,
+      "grad_norm": 0.884464242375809,
+      "learning_rate": 9.934861361818741e-05,
+      "loss": 12.2945,
+      "step": 8869
+    },
+    {
+      "epoch": 1.5511693989071038,
+      "grad_norm": 0.8626053718383789,
+      "learning_rate": 9.932973313226408e-05,
+      "loss": 12.4247,
+      "step": 8870
+    },
+    {
+      "epoch": 1.551344262295082,
+      "grad_norm": 0.8446808195128959,
+      "learning_rate": 9.931085267023498e-05,
+      "loss": 12.2111,
+      "step": 8871
+    },
+    {
+      "epoch": 1.5515191256830603,
+      "grad_norm": 0.8446149979418724,
+      "learning_rate": 9.929197223277317e-05,
+      "loss": 12.2541,
+      "step": 8872
+    },
+    {
+      "epoch": 1.5516939890710382,
+      "grad_norm": 0.9378289157920111,
+      "learning_rate": 9.927309182055178e-05,
+      "loss": 12.5572,
+      "step": 8873
+    },
+    {
+      "epoch": 1.5518688524590165,
+      "grad_norm": 0.8252251368496185,
+      "learning_rate": 9.925421143424377e-05,
+      "loss": 12.5237,
+      "step": 8874
+    },
+    {
+      "epoch": 1.5520437158469944,
+      "grad_norm": 0.9699485673942745,
+      "learning_rate": 9.923533107452223e-05,
+      "loss": 12.359,
+      "step": 8875
+    },
+    {
+      "epoch": 1.5522185792349728,
+      "grad_norm": 0.9748295760534885,
+      "learning_rate": 9.921645074206029e-05,
+      "loss": 12.7394,
+      "step": 8876
+    },
+    {
+      "epoch": 1.5523934426229509,
+      "grad_norm": 0.8416213842722176,
+      "learning_rate": 9.919757043753093e-05,
+      "loss": 12.5605,
+      "step": 8877
+    },
+    {
+      "epoch": 1.552568306010929,
+      "grad_norm": 0.9644789920935984,
+      "learning_rate": 9.917869016160724e-05,
+      "loss": 12.5427,
+      "step": 8878
+    },
+    {
+      "epoch": 1.552743169398907,
+      "grad_norm": 0.8228571531758789,
+      "learning_rate": 9.91598099149623e-05,
+      "loss": 12.2201,
+      "step": 8879
+    },
+    {
+      "epoch": 1.5529180327868852,
+      "grad_norm": 0.9201200540519665,
+      "learning_rate": 9.91409296982692e-05,
+      "loss": 12.3139,
+      "step": 8880
+    },
+    {
+      "epoch": 1.5530928961748633,
+      "grad_norm": 0.8937549173792652,
+      "learning_rate": 9.91220495122009e-05,
+      "loss": 12.5178,
+      "step": 8881
+    },
+    {
+      "epoch": 1.5532677595628415,
+      "grad_norm": 0.8858341051897222,
+      "learning_rate": 9.910316935743051e-05,
+      "loss": 12.487,
+      "step": 8882
+    },
+    {
+      "epoch": 1.5534426229508198,
+      "grad_norm": 0.8288354884949143,
+      "learning_rate": 9.90842892346311e-05,
+      "loss": 12.6163,
+      "step": 8883
+    },
+    {
+      "epoch": 1.5536174863387977,
+      "grad_norm": 0.8277959415401892,
+      "learning_rate": 9.90654091444757e-05,
+      "loss": 12.4223,
+      "step": 8884
+    },
+    {
+      "epoch": 1.553792349726776,
+      "grad_norm": 0.8715847974575123,
+      "learning_rate": 9.904652908763738e-05,
+      "loss": 12.6564,
+      "step": 8885
+    },
+    {
+      "epoch": 1.553967213114754,
+      "grad_norm": 0.9068916294525298,
+      "learning_rate": 9.902764906478918e-05,
+      "loss": 12.383,
+      "step": 8886
+    },
+    {
+      "epoch": 1.5541420765027323,
+      "grad_norm": 0.9739721769718316,
+      "learning_rate": 9.900876907660413e-05,
+      "loss": 12.6202,
+      "step": 8887
+    },
+    {
+      "epoch": 1.5543169398907104,
+      "grad_norm": 0.8738850351781622,
+      "learning_rate": 9.898988912375537e-05,
+      "loss": 12.2688,
+      "step": 8888
+    },
+    {
+      "epoch": 1.5544918032786885,
+      "grad_norm": 0.846003800281422,
+      "learning_rate": 9.897100920691584e-05,
+      "loss": 12.5105,
+      "step": 8889
+    },
+    {
+      "epoch": 1.5546666666666666,
+      "grad_norm": 0.8956490990704042,
+      "learning_rate": 9.895212932675862e-05,
+      "loss": 12.3528,
+      "step": 8890
+    },
+    {
+      "epoch": 1.5548415300546448,
+      "grad_norm": 0.8466414975948449,
+      "learning_rate": 9.893324948395682e-05,
+      "loss": 12.3069,
+      "step": 8891
+    },
+    {
+      "epoch": 1.555016393442623,
+      "grad_norm": 0.9003551151055573,
+      "learning_rate": 9.891436967918338e-05,
+      "loss": 12.6031,
+      "step": 8892
+    },
+    {
+      "epoch": 1.555191256830601,
+      "grad_norm": 0.8595256858905451,
+      "learning_rate": 9.889548991311139e-05,
+      "loss": 12.4882,
+      "step": 8893
+    },
+    {
+      "epoch": 1.5553661202185793,
+      "grad_norm": 0.8441683264617886,
+      "learning_rate": 9.887661018641388e-05,
+      "loss": 12.2916,
+      "step": 8894
+    },
+    {
+      "epoch": 1.5555409836065572,
+      "grad_norm": 0.893727862238372,
+      "learning_rate": 9.885773049976394e-05,
+      "loss": 12.2458,
+      "step": 8895
+    },
+    {
+      "epoch": 1.5557158469945356,
+      "grad_norm": 0.9085150235601644,
+      "learning_rate": 9.883885085383456e-05,
+      "loss": 12.5631,
+      "step": 8896
+    },
+    {
+      "epoch": 1.5558907103825137,
+      "grad_norm": 0.8663848425199862,
+      "learning_rate": 9.881997124929878e-05,
+      "loss": 12.5889,
+      "step": 8897
+    },
+    {
+      "epoch": 1.5560655737704918,
+      "grad_norm": 0.896783087422355,
+      "learning_rate": 9.880109168682966e-05,
+      "loss": 12.4324,
+      "step": 8898
+    },
+    {
+      "epoch": 1.55624043715847,
+      "grad_norm": 0.9600732003057358,
+      "learning_rate": 9.878221216710026e-05,
+      "loss": 12.311,
+      "step": 8899
+    },
+    {
+      "epoch": 1.556415300546448,
+      "grad_norm": 0.8783923937123727,
+      "learning_rate": 9.876333269078352e-05,
+      "loss": 12.4349,
+      "step": 8900
+    },
+    {
+      "epoch": 1.5565901639344262,
+      "grad_norm": 0.8612333573237096,
+      "learning_rate": 9.874445325855252e-05,
+      "loss": 12.6713,
+      "step": 8901
+    },
+    {
+      "epoch": 1.5567650273224043,
+      "grad_norm": 0.9000060868189774,
+      "learning_rate": 9.87255738710803e-05,
+      "loss": 12.5733,
+      "step": 8902
+    },
+    {
+      "epoch": 1.5569398907103826,
+      "grad_norm": 0.8672259749218155,
+      "learning_rate": 9.870669452903991e-05,
+      "loss": 12.6409,
+      "step": 8903
+    },
+    {
+      "epoch": 1.5571147540983605,
+      "grad_norm": 0.9064013778522149,
+      "learning_rate": 9.868781523310432e-05,
+      "loss": 12.4883,
+      "step": 8904
+    },
+    {
+      "epoch": 1.5572896174863389,
+      "grad_norm": 0.8887463961531094,
+      "learning_rate": 9.866893598394658e-05,
+      "loss": 12.5415,
+      "step": 8905
+    },
+    {
+      "epoch": 1.557464480874317,
+      "grad_norm": 0.8560185305417563,
+      "learning_rate": 9.865005678223977e-05,
+      "loss": 12.3581,
+      "step": 8906
+    },
+    {
+      "epoch": 1.5576393442622951,
+      "grad_norm": 0.8679169869092218,
+      "learning_rate": 9.863117762865682e-05,
+      "loss": 12.3468,
+      "step": 8907
+    },
+    {
+      "epoch": 1.5578142076502732,
+      "grad_norm": 0.8256402775895497,
+      "learning_rate": 9.861229852387078e-05,
+      "loss": 12.4269,
+      "step": 8908
+    },
+    {
+      "epoch": 1.5579890710382513,
+      "grad_norm": 0.9128362678673252,
+      "learning_rate": 9.859341946855468e-05,
+      "loss": 12.3656,
+      "step": 8909
+    },
+    {
+      "epoch": 1.5581639344262295,
+      "grad_norm": 0.8325224266300113,
+      "learning_rate": 9.857454046338155e-05,
+      "loss": 12.5737,
+      "step": 8910
+    },
+    {
+      "epoch": 1.5583387978142076,
+      "grad_norm": 0.9371234750579664,
+      "learning_rate": 9.855566150902436e-05,
+      "loss": 12.6807,
+      "step": 8911
+    },
+    {
+      "epoch": 1.558513661202186,
+      "grad_norm": 0.8109294408483373,
+      "learning_rate": 9.853678260615617e-05,
+      "loss": 12.3804,
+      "step": 8912
+    },
+    {
+      "epoch": 1.5586885245901638,
+      "grad_norm": 0.910724032255098,
+      "learning_rate": 9.851790375544994e-05,
+      "loss": 12.548,
+      "step": 8913
+    },
+    {
+      "epoch": 1.5588633879781422,
+      "grad_norm": 0.8267533527349472,
+      "learning_rate": 9.849902495757878e-05,
+      "loss": 12.4802,
+      "step": 8914
+    },
+    {
+      "epoch": 1.55903825136612,
+      "grad_norm": 0.8425187029126691,
+      "learning_rate": 9.848014621321557e-05,
+      "loss": 12.2244,
+      "step": 8915
+    },
+    {
+      "epoch": 1.5592131147540984,
+      "grad_norm": 0.8617406217157798,
+      "learning_rate": 9.846126752303338e-05,
+      "loss": 12.1511,
+      "step": 8916
+    },
+    {
+      "epoch": 1.5593879781420765,
+      "grad_norm": 0.8398458250420293,
+      "learning_rate": 9.844238888770519e-05,
+      "loss": 12.5081,
+      "step": 8917
+    },
+    {
+      "epoch": 1.5595628415300546,
+      "grad_norm": 0.8913641038070237,
+      "learning_rate": 9.842351030790408e-05,
+      "loss": 12.4807,
+      "step": 8918
+    },
+    {
+      "epoch": 1.5597377049180328,
+      "grad_norm": 0.9138595059901576,
+      "learning_rate": 9.840463178430294e-05,
+      "loss": 12.1133,
+      "step": 8919
+    },
+    {
+      "epoch": 1.5599125683060109,
+      "grad_norm": 0.8664531824840374,
+      "learning_rate": 9.83857533175748e-05,
+      "loss": 12.4313,
+      "step": 8920
+    },
+    {
+      "epoch": 1.5600874316939892,
+      "grad_norm": 0.9561787335393952,
+      "learning_rate": 9.83668749083927e-05,
+      "loss": 12.2084,
+      "step": 8921
+    },
+    {
+      "epoch": 1.5602622950819671,
+      "grad_norm": 0.8434773080611726,
+      "learning_rate": 9.834799655742957e-05,
+      "loss": 12.3823,
+      "step": 8922
+    },
+    {
+      "epoch": 1.5604371584699455,
+      "grad_norm": 0.9984646624390666,
+      "learning_rate": 9.832911826535843e-05,
+      "loss": 12.5418,
+      "step": 8923
+    },
+    {
+      "epoch": 1.5606120218579234,
+      "grad_norm": 0.903541141223209,
+      "learning_rate": 9.831024003285229e-05,
+      "loss": 12.6253,
+      "step": 8924
+    },
+    {
+      "epoch": 1.5607868852459017,
+      "grad_norm": 0.9804002647592783,
+      "learning_rate": 9.829136186058414e-05,
+      "loss": 12.3851,
+      "step": 8925
+    },
+    {
+      "epoch": 1.5609617486338798,
+      "grad_norm": 0.8337655461194936,
+      "learning_rate": 9.827248374922691e-05,
+      "loss": 12.5047,
+      "step": 8926
+    },
+    {
+      "epoch": 1.561136612021858,
+      "grad_norm": 0.8282712487257637,
+      "learning_rate": 9.825360569945362e-05,
+      "loss": 12.408,
+      "step": 8927
+    },
+    {
+      "epoch": 1.561311475409836,
+      "grad_norm": 0.836761013727824,
+      "learning_rate": 9.823472771193724e-05,
+      "loss": 12.5143,
+      "step": 8928
+    },
+    {
+      "epoch": 1.5614863387978142,
+      "grad_norm": 0.8588558505771349,
+      "learning_rate": 9.821584978735078e-05,
+      "loss": 12.5503,
+      "step": 8929
+    },
+    {
+      "epoch": 1.5616612021857923,
+      "grad_norm": 0.8172956070808483,
+      "learning_rate": 9.819697192636717e-05,
+      "loss": 12.3414,
+      "step": 8930
+    },
+    {
+      "epoch": 1.5618360655737704,
+      "grad_norm": 0.8190169890680373,
+      "learning_rate": 9.81780941296594e-05,
+      "loss": 12.3452,
+      "step": 8931
+    },
+    {
+      "epoch": 1.5620109289617488,
+      "grad_norm": 0.8040659802818045,
+      "learning_rate": 9.815921639790045e-05,
+      "loss": 12.2491,
+      "step": 8932
+    },
+    {
+      "epoch": 1.5621857923497267,
+      "grad_norm": 0.8347233331392608,
+      "learning_rate": 9.814033873176332e-05,
+      "loss": 12.3364,
+      "step": 8933
+    },
+    {
+      "epoch": 1.562360655737705,
+      "grad_norm": 0.9209303811782419,
+      "learning_rate": 9.812146113192091e-05,
+      "loss": 12.4633,
+      "step": 8934
+    },
+    {
+      "epoch": 1.5625355191256831,
+      "grad_norm": 0.9021341859109129,
+      "learning_rate": 9.810258359904621e-05,
+      "loss": 12.5371,
+      "step": 8935
+    },
+    {
+      "epoch": 1.5627103825136612,
+      "grad_norm": 0.8193068229587129,
+      "learning_rate": 9.808370613381222e-05,
+      "loss": 12.4511,
+      "step": 8936
+    },
+    {
+      "epoch": 1.5628852459016394,
+      "grad_norm": 0.8926392570998246,
+      "learning_rate": 9.806482873689184e-05,
+      "loss": 12.4483,
+      "step": 8937
+    },
+    {
+      "epoch": 1.5630601092896175,
+      "grad_norm": 0.9061600983317484,
+      "learning_rate": 9.804595140895805e-05,
+      "loss": 12.5061,
+      "step": 8938
+    },
+    {
+      "epoch": 1.5632349726775956,
+      "grad_norm": 0.8401440206041181,
+      "learning_rate": 9.80270741506838e-05,
+      "loss": 12.462,
+      "step": 8939
+    },
+    {
+      "epoch": 1.5634098360655737,
+      "grad_norm": 0.9172725797863391,
+      "learning_rate": 9.80081969627421e-05,
+      "loss": 12.3717,
+      "step": 8940
+    },
+    {
+      "epoch": 1.563584699453552,
+      "grad_norm": 0.8487609769668235,
+      "learning_rate": 9.79893198458058e-05,
+      "loss": 12.5269,
+      "step": 8941
+    },
+    {
+      "epoch": 1.56375956284153,
+      "grad_norm": 0.9726400546557555,
+      "learning_rate": 9.797044280054791e-05,
+      "loss": 12.446,
+      "step": 8942
+    },
+    {
+      "epoch": 1.5639344262295083,
+      "grad_norm": 0.965135992705086,
+      "learning_rate": 9.795156582764137e-05,
+      "loss": 12.6517,
+      "step": 8943
+    },
+    {
+      "epoch": 1.5641092896174862,
+      "grad_norm": 0.9484612796883747,
+      "learning_rate": 9.793268892775914e-05,
+      "loss": 12.1601,
+      "step": 8944
+    },
+    {
+      "epoch": 1.5642841530054645,
+      "grad_norm": 0.9827950029094329,
+      "learning_rate": 9.791381210157411e-05,
+      "loss": 12.5616,
+      "step": 8945
+    },
+    {
+      "epoch": 1.5644590163934426,
+      "grad_norm": 0.897805294049497,
+      "learning_rate": 9.789493534975923e-05,
+      "loss": 12.307,
+      "step": 8946
+    },
+    {
+      "epoch": 1.5646338797814208,
+      "grad_norm": 0.8992230755256304,
+      "learning_rate": 9.787605867298744e-05,
+      "loss": 12.5649,
+      "step": 8947
+    },
+    {
+      "epoch": 1.5648087431693989,
+      "grad_norm": 0.9161209405590759,
+      "learning_rate": 9.78571820719317e-05,
+      "loss": 12.4877,
+      "step": 8948
+    },
+    {
+      "epoch": 1.564983606557377,
+      "grad_norm": 0.8495420890885654,
+      "learning_rate": 9.783830554726492e-05,
+      "loss": 12.2022,
+      "step": 8949
+    },
+    {
+      "epoch": 1.5651584699453553,
+      "grad_norm": 1.190028514195336,
+      "learning_rate": 9.781942909966e-05,
+      "loss": 12.4373,
+      "step": 8950
+    },
+    {
+      "epoch": 1.5653333333333332,
+      "grad_norm": 0.8456976948840981,
+      "learning_rate": 9.780055272978988e-05,
+      "loss": 12.2385,
+      "step": 8951
+    },
+    {
+      "epoch": 1.5655081967213116,
+      "grad_norm": 1.2936031405725024,
+      "learning_rate": 9.778167643832754e-05,
+      "loss": 12.3247,
+      "step": 8952
+    },
+    {
+      "epoch": 1.5656830601092895,
+      "grad_norm": 0.8271377231472544,
+      "learning_rate": 9.77628002259458e-05,
+      "loss": 12.7102,
+      "step": 8953
+    },
+    {
+      "epoch": 1.5658579234972678,
+      "grad_norm": 1.193554184509257,
+      "learning_rate": 9.77439240933176e-05,
+      "loss": 12.2711,
+      "step": 8954
+    },
+    {
+      "epoch": 1.566032786885246,
+      "grad_norm": 0.9721406504474502,
+      "learning_rate": 9.77250480411159e-05,
+      "loss": 12.6082,
+      "step": 8955
+    },
+    {
+      "epoch": 1.566207650273224,
+      "grad_norm": 0.9469194702558226,
+      "learning_rate": 9.770617207001357e-05,
+      "loss": 12.4561,
+      "step": 8956
+    },
+    {
+      "epoch": 1.5663825136612022,
+      "grad_norm": 1.0665145036121586,
+      "learning_rate": 9.768729618068352e-05,
+      "loss": 12.2742,
+      "step": 8957
+    },
+    {
+      "epoch": 1.5665573770491803,
+      "grad_norm": 0.8178881813342123,
+      "learning_rate": 9.766842037379866e-05,
+      "loss": 12.5248,
+      "step": 8958
+    },
+    {
+      "epoch": 1.5667322404371584,
+      "grad_norm": 1.0855946940117762,
+      "learning_rate": 9.764954465003193e-05,
+      "loss": 12.5763,
+      "step": 8959
+    },
+    {
+      "epoch": 1.5669071038251365,
+      "grad_norm": 0.8969078151731266,
+      "learning_rate": 9.763066901005615e-05,
+      "loss": 12.4709,
+      "step": 8960
+    },
+    {
+      "epoch": 1.5670819672131149,
+      "grad_norm": 0.8926057256407498,
+      "learning_rate": 9.761179345454424e-05,
+      "loss": 12.4311,
+      "step": 8961
+    },
+    {
+      "epoch": 1.5672568306010928,
+      "grad_norm": 0.9753855581067705,
+      "learning_rate": 9.75929179841691e-05,
+      "loss": 12.4063,
+      "step": 8962
+    },
+    {
+      "epoch": 1.5674316939890711,
+      "grad_norm": 0.8330964097383005,
+      "learning_rate": 9.757404259960368e-05,
+      "loss": 12.2602,
+      "step": 8963
+    },
+    {
+      "epoch": 1.567606557377049,
+      "grad_norm": 0.971808183872899,
+      "learning_rate": 9.755516730152075e-05,
+      "loss": 12.6999,
+      "step": 8964
+    },
+    {
+      "epoch": 1.5677814207650274,
+      "grad_norm": 0.9069212402627571,
+      "learning_rate": 9.753629209059326e-05,
+      "loss": 12.4504,
+      "step": 8965
+    },
+    {
+      "epoch": 1.5679562841530055,
+      "grad_norm": 0.9151926732114558,
+      "learning_rate": 9.751741696749408e-05,
+      "loss": 12.3813,
+      "step": 8966
+    },
+    {
+      "epoch": 1.5681311475409836,
+      "grad_norm": 1.0122386254986975,
+      "learning_rate": 9.749854193289611e-05,
+      "loss": 12.3693,
+      "step": 8967
+    },
+    {
+      "epoch": 1.5683060109289617,
+      "grad_norm": 0.8455543859756642,
+      "learning_rate": 9.747966698747215e-05,
+      "loss": 12.5628,
+      "step": 8968
+    },
+    {
+      "epoch": 1.5684808743169398,
+      "grad_norm": 0.9854007196462461,
+      "learning_rate": 9.746079213189514e-05,
+      "loss": 12.6373,
+      "step": 8969
+    },
+    {
+      "epoch": 1.5686557377049182,
+      "grad_norm": 0.8704866727986921,
+      "learning_rate": 9.744191736683796e-05,
+      "loss": 12.3799,
+      "step": 8970
+    },
+    {
+      "epoch": 1.568830601092896,
+      "grad_norm": 0.8340467629373529,
+      "learning_rate": 9.74230426929734e-05,
+      "loss": 12.3521,
+      "step": 8971
+    },
+    {
+      "epoch": 1.5690054644808744,
+      "grad_norm": 0.8721702978985637,
+      "learning_rate": 9.740416811097435e-05,
+      "loss": 12.4205,
+      "step": 8972
+    },
+    {
+      "epoch": 1.5691803278688523,
+      "grad_norm": 0.8547022699096425,
+      "learning_rate": 9.738529362151368e-05,
+      "loss": 12.5086,
+      "step": 8973
+    },
+    {
+      "epoch": 1.5693551912568307,
+      "grad_norm": 0.8831657745393852,
+      "learning_rate": 9.736641922526425e-05,
+      "loss": 12.7038,
+      "step": 8974
+    },
+    {
+      "epoch": 1.5695300546448088,
+      "grad_norm": 0.8810962122565086,
+      "learning_rate": 9.734754492289888e-05,
+      "loss": 12.6048,
+      "step": 8975
+    },
+    {
+      "epoch": 1.5697049180327869,
+      "grad_norm": 0.8308111832108258,
+      "learning_rate": 9.732867071509043e-05,
+      "loss": 12.6145,
+      "step": 8976
+    },
+    {
+      "epoch": 1.569879781420765,
+      "grad_norm": 0.818468472699485,
+      "learning_rate": 9.730979660251175e-05,
+      "loss": 12.572,
+      "step": 8977
+    },
+    {
+      "epoch": 1.5700546448087431,
+      "grad_norm": 0.863501748772249,
+      "learning_rate": 9.729092258583573e-05,
+      "loss": 12.2737,
+      "step": 8978
+    },
+    {
+      "epoch": 1.5702295081967215,
+      "grad_norm": 0.8461830742443198,
+      "learning_rate": 9.727204866573511e-05,
+      "loss": 12.4746,
+      "step": 8979
+    },
+    {
+      "epoch": 1.5704043715846994,
+      "grad_norm": 0.9051144195625321,
+      "learning_rate": 9.725317484288275e-05,
+      "loss": 12.5978,
+      "step": 8980
+    },
+    {
+      "epoch": 1.5705792349726777,
+      "grad_norm": 0.8981569441519959,
+      "learning_rate": 9.723430111795152e-05,
+      "loss": 12.4337,
+      "step": 8981
+    },
+    {
+      "epoch": 1.5707540983606556,
+      "grad_norm": 0.8980186635668037,
+      "learning_rate": 9.721542749161425e-05,
+      "loss": 12.3445,
+      "step": 8982
+    },
+    {
+      "epoch": 1.570928961748634,
+      "grad_norm": 0.8967017006701873,
+      "learning_rate": 9.719655396454371e-05,
+      "loss": 12.4915,
+      "step": 8983
+    },
+    {
+      "epoch": 1.571103825136612,
+      "grad_norm": 0.8744114858356888,
+      "learning_rate": 9.717768053741273e-05,
+      "loss": 12.505,
+      "step": 8984
+    },
+    {
+      "epoch": 1.5712786885245902,
+      "grad_norm": 1.010335856177983,
+      "learning_rate": 9.715880721089418e-05,
+      "loss": 12.7549,
+      "step": 8985
+    },
+    {
+      "epoch": 1.5714535519125683,
+      "grad_norm": 0.8820675960909026,
+      "learning_rate": 9.713993398566079e-05,
+      "loss": 12.2048,
+      "step": 8986
+    },
+    {
+      "epoch": 1.5716284153005464,
+      "grad_norm": 0.9105383250805773,
+      "learning_rate": 9.712106086238542e-05,
+      "loss": 12.4068,
+      "step": 8987
+    },
+    {
+      "epoch": 1.5718032786885245,
+      "grad_norm": 0.8612222748803015,
+      "learning_rate": 9.710218784174088e-05,
+      "loss": 12.4731,
+      "step": 8988
+    },
+    {
+      "epoch": 1.5719781420765027,
+      "grad_norm": 0.9093864982894899,
+      "learning_rate": 9.70833149244e-05,
+      "loss": 12.3915,
+      "step": 8989
+    },
+    {
+      "epoch": 1.572153005464481,
+      "grad_norm": 0.8712131459034899,
+      "learning_rate": 9.706444211103548e-05,
+      "loss": 12.4413,
+      "step": 8990
+    },
+    {
+      "epoch": 1.572327868852459,
+      "grad_norm": 0.8841315772802362,
+      "learning_rate": 9.704556940232017e-05,
+      "loss": 12.1211,
+      "step": 8991
+    },
+    {
+      "epoch": 1.5725027322404372,
+      "grad_norm": 0.9130623539372285,
+      "learning_rate": 9.702669679892686e-05,
+      "loss": 12.3619,
+      "step": 8992
+    },
+    {
+      "epoch": 1.5726775956284151,
+      "grad_norm": 0.873043061764684,
+      "learning_rate": 9.700782430152835e-05,
+      "loss": 12.4776,
+      "step": 8993
+    },
+    {
+      "epoch": 1.5728524590163935,
+      "grad_norm": 0.8633278740675929,
+      "learning_rate": 9.698895191079739e-05,
+      "loss": 12.3028,
+      "step": 8994
+    },
+    {
+      "epoch": 1.5730273224043716,
+      "grad_norm": 0.9002673489745153,
+      "learning_rate": 9.697007962740679e-05,
+      "loss": 12.2501,
+      "step": 8995
+    },
+    {
+      "epoch": 1.5732021857923497,
+      "grad_norm": 0.9000118017263571,
+      "learning_rate": 9.69512074520293e-05,
+      "loss": 12.5546,
+      "step": 8996
+    },
+    {
+      "epoch": 1.5733770491803278,
+      "grad_norm": 0.8731859135994953,
+      "learning_rate": 9.693233538533775e-05,
+      "loss": 12.4119,
+      "step": 8997
+    },
+    {
+      "epoch": 1.573551912568306,
+      "grad_norm": 0.8849449534848706,
+      "learning_rate": 9.69134634280048e-05,
+      "loss": 12.2144,
+      "step": 8998
+    },
+    {
+      "epoch": 1.5737267759562843,
+      "grad_norm": 0.8978099026965912,
+      "learning_rate": 9.689459158070329e-05,
+      "loss": 12.5352,
+      "step": 8999
+    },
+    {
+      "epoch": 1.5739016393442622,
+      "grad_norm": 0.9017363215065652,
+      "learning_rate": 9.687571984410598e-05,
+      "loss": 12.7338,
+      "step": 9000
+    },
+    {
+      "epoch": 1.5740765027322405,
+      "grad_norm": 0.8986520013228205,
+      "learning_rate": 9.685684821888561e-05,
+      "loss": 12.4719,
+      "step": 9001
+    },
+    {
+      "epoch": 1.5742513661202184,
+      "grad_norm": 1.0529449312291592,
+      "learning_rate": 9.683797670571489e-05,
+      "loss": 12.5639,
+      "step": 9002
+    },
+    {
+      "epoch": 1.5744262295081968,
+      "grad_norm": 1.1022031163055543,
+      "learning_rate": 9.681910530526662e-05,
+      "loss": 12.1113,
+      "step": 9003
+    },
+    {
+      "epoch": 1.574601092896175,
+      "grad_norm": 0.949951391604054,
+      "learning_rate": 9.680023401821355e-05,
+      "loss": 12.4021,
+      "step": 9004
+    },
+    {
+      "epoch": 1.574775956284153,
+      "grad_norm": 1.1389937771586909,
+      "learning_rate": 9.678136284522839e-05,
+      "loss": 12.3611,
+      "step": 9005
+    },
+    {
+      "epoch": 1.5749508196721311,
+      "grad_norm": 0.7673003048906618,
+      "learning_rate": 9.676249178698386e-05,
+      "loss": 12.3475,
+      "step": 9006
+    },
+    {
+      "epoch": 1.5751256830601092,
+      "grad_norm": 1.042241624547605,
+      "learning_rate": 9.674362084415273e-05,
+      "loss": 12.69,
+      "step": 9007
+    },
+    {
+      "epoch": 1.5753005464480876,
+      "grad_norm": 0.895345735018858,
+      "learning_rate": 9.672475001740775e-05,
+      "loss": 12.3679,
+      "step": 9008
+    },
+    {
+      "epoch": 1.5754754098360655,
+      "grad_norm": 0.900509013980459,
+      "learning_rate": 9.670587930742157e-05,
+      "loss": 12.4743,
+      "step": 9009
+    },
+    {
+      "epoch": 1.5756502732240438,
+      "grad_norm": 0.9242769421483537,
+      "learning_rate": 9.668700871486694e-05,
+      "loss": 12.3237,
+      "step": 9010
+    },
+    {
+      "epoch": 1.5758251366120217,
+      "grad_norm": 0.8926681264501338,
+      "learning_rate": 9.666813824041657e-05,
+      "loss": 12.2871,
+      "step": 9011
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 0.9753279694005403,
+      "learning_rate": 9.66492678847432e-05,
+      "loss": 12.5973,
+      "step": 9012
+    },
+    {
+      "epoch": 1.5761748633879782,
+      "grad_norm": 0.8724160313059977,
+      "learning_rate": 9.66303976485195e-05,
+      "loss": 12.3882,
+      "step": 9013
+    },
+    {
+      "epoch": 1.5763497267759563,
+      "grad_norm": 0.8259979913970612,
+      "learning_rate": 9.66115275324182e-05,
+      "loss": 12.3395,
+      "step": 9014
+    },
+    {
+      "epoch": 1.5765245901639344,
+      "grad_norm": 0.9794969067229902,
+      "learning_rate": 9.659265753711201e-05,
+      "loss": 12.5391,
+      "step": 9015
+    },
+    {
+      "epoch": 1.5766994535519125,
+      "grad_norm": 0.7763206603830205,
+      "learning_rate": 9.657378766327355e-05,
+      "loss": 12.4569,
+      "step": 9016
+    },
+    {
+      "epoch": 1.5768743169398907,
+      "grad_norm": 0.7966280464233246,
+      "learning_rate": 9.655491791157558e-05,
+      "loss": 12.4247,
+      "step": 9017
+    },
+    {
+      "epoch": 1.5770491803278688,
+      "grad_norm": 0.8989204513167535,
+      "learning_rate": 9.653604828269074e-05,
+      "loss": 12.3904,
+      "step": 9018
+    },
+    {
+      "epoch": 1.5772240437158471,
+      "grad_norm": 0.8004653109933985,
+      "learning_rate": 9.651717877729176e-05,
+      "loss": 12.7062,
+      "step": 9019
+    },
+    {
+      "epoch": 1.577398907103825,
+      "grad_norm": 0.861152374393377,
+      "learning_rate": 9.649830939605127e-05,
+      "loss": 12.3317,
+      "step": 9020
+    },
+    {
+      "epoch": 1.5775737704918034,
+      "grad_norm": 0.8230374334543521,
+      "learning_rate": 9.647944013964195e-05,
+      "loss": 12.5508,
+      "step": 9021
+    },
+    {
+      "epoch": 1.5777486338797813,
+      "grad_norm": 0.8243608404774936,
+      "learning_rate": 9.646057100873648e-05,
+      "loss": 12.5215,
+      "step": 9022
+    },
+    {
+      "epoch": 1.5779234972677596,
+      "grad_norm": 0.9440635216481033,
+      "learning_rate": 9.644170200400756e-05,
+      "loss": 12.5985,
+      "step": 9023
+    },
+    {
+      "epoch": 1.5780983606557377,
+      "grad_norm": 0.7834341088059104,
+      "learning_rate": 9.642283312612777e-05,
+      "loss": 12.5394,
+      "step": 9024
+    },
+    {
+      "epoch": 1.5782732240437158,
+      "grad_norm": 0.8631792727506581,
+      "learning_rate": 9.640396437576978e-05,
+      "loss": 12.6271,
+      "step": 9025
+    },
+    {
+      "epoch": 1.578448087431694,
+      "grad_norm": 0.9081231856195582,
+      "learning_rate": 9.638509575360628e-05,
+      "loss": 12.5076,
+      "step": 9026
+    },
+    {
+      "epoch": 1.578622950819672,
+      "grad_norm": 0.8383834508307483,
+      "learning_rate": 9.636622726030993e-05,
+      "loss": 12.4185,
+      "step": 9027
+    },
+    {
+      "epoch": 1.5787978142076504,
+      "grad_norm": 0.7819664418065406,
+      "learning_rate": 9.634735889655328e-05,
+      "loss": 12.6922,
+      "step": 9028
+    },
+    {
+      "epoch": 1.5789726775956283,
+      "grad_norm": 0.9711383181871011,
+      "learning_rate": 9.632849066300903e-05,
+      "loss": 12.3223,
+      "step": 9029
+    },
+    {
+      "epoch": 1.5791475409836067,
+      "grad_norm": 0.9124559645016405,
+      "learning_rate": 9.630962256034982e-05,
+      "loss": 12.3096,
+      "step": 9030
+    },
+    {
+      "epoch": 1.5793224043715846,
+      "grad_norm": 0.8733066884582611,
+      "learning_rate": 9.629075458924822e-05,
+      "loss": 12.4539,
+      "step": 9031
+    },
+    {
+      "epoch": 1.579497267759563,
+      "grad_norm": 0.9666460601274756,
+      "learning_rate": 9.627188675037689e-05,
+      "loss": 12.6204,
+      "step": 9032
+    },
+    {
+      "epoch": 1.579672131147541,
+      "grad_norm": 0.8523995722557371,
+      "learning_rate": 9.625301904440845e-05,
+      "loss": 12.577,
+      "step": 9033
+    },
+    {
+      "epoch": 1.5798469945355191,
+      "grad_norm": 1.0394037447213784,
+      "learning_rate": 9.623415147201554e-05,
+      "loss": 12.5821,
+      "step": 9034
+    },
+    {
+      "epoch": 1.5800218579234973,
+      "grad_norm": 0.8469717540450956,
+      "learning_rate": 9.62152840338707e-05,
+      "loss": 12.472,
+      "step": 9035
+    },
+    {
+      "epoch": 1.5801967213114754,
+      "grad_norm": 1.0108046903509402,
+      "learning_rate": 9.619641673064654e-05,
+      "loss": 12.6067,
+      "step": 9036
+    },
+    {
+      "epoch": 1.5803715846994537,
+      "grad_norm": 0.8283605948345499,
+      "learning_rate": 9.617754956301568e-05,
+      "loss": 12.4504,
+      "step": 9037
+    },
+    {
+      "epoch": 1.5805464480874316,
+      "grad_norm": 0.892128303923831,
+      "learning_rate": 9.615868253165076e-05,
+      "loss": 12.6223,
+      "step": 9038
+    },
+    {
+      "epoch": 1.58072131147541,
+      "grad_norm": 0.8921197584350866,
+      "learning_rate": 9.613981563722427e-05,
+      "loss": 12.5885,
+      "step": 9039
+    },
+    {
+      "epoch": 1.5808961748633878,
+      "grad_norm": 0.8267109778985544,
+      "learning_rate": 9.612094888040885e-05,
+      "loss": 12.4436,
+      "step": 9040
+    },
+    {
+      "epoch": 1.5810710382513662,
+      "grad_norm": 0.8326287790711234,
+      "learning_rate": 9.610208226187707e-05,
+      "loss": 12.4838,
+      "step": 9041
+    },
+    {
+      "epoch": 1.5812459016393443,
+      "grad_norm": 0.8352650275743513,
+      "learning_rate": 9.608321578230157e-05,
+      "loss": 12.369,
+      "step": 9042
+    },
+    {
+      "epoch": 1.5814207650273224,
+      "grad_norm": 0.8853330107961841,
+      "learning_rate": 9.606434944235478e-05,
+      "loss": 12.6864,
+      "step": 9043
+    },
+    {
+      "epoch": 1.5815956284153005,
+      "grad_norm": 0.8680459362405647,
+      "learning_rate": 9.604548324270935e-05,
+      "loss": 12.2961,
+      "step": 9044
+    },
+    {
+      "epoch": 1.5817704918032787,
+      "grad_norm": 0.8365915670246777,
+      "learning_rate": 9.602661718403785e-05,
+      "loss": 12.2079,
+      "step": 9045
+    },
+    {
+      "epoch": 1.5819453551912568,
+      "grad_norm": 0.9178047566540193,
+      "learning_rate": 9.600775126701277e-05,
+      "loss": 12.3682,
+      "step": 9046
+    },
+    {
+      "epoch": 1.582120218579235,
+      "grad_norm": 0.9873943529292327,
+      "learning_rate": 9.59888854923067e-05,
+      "loss": 12.6405,
+      "step": 9047
+    },
+    {
+      "epoch": 1.5822950819672132,
+      "grad_norm": 0.8247774867990797,
+      "learning_rate": 9.597001986059215e-05,
+      "loss": 12.5221,
+      "step": 9048
+    },
+    {
+      "epoch": 1.5824699453551911,
+      "grad_norm": 0.8463711380212453,
+      "learning_rate": 9.595115437254173e-05,
+      "loss": 12.5271,
+      "step": 9049
+    },
+    {
+      "epoch": 1.5826448087431695,
+      "grad_norm": 0.933542033564877,
+      "learning_rate": 9.593228902882789e-05,
+      "loss": 12.625,
+      "step": 9050
+    },
+    {
+      "epoch": 1.5828196721311474,
+      "grad_norm": 0.9004820994496542,
+      "learning_rate": 9.591342383012321e-05,
+      "loss": 12.161,
+      "step": 9051
+    },
+    {
+      "epoch": 1.5829945355191257,
+      "grad_norm": 0.7963530759436676,
+      "learning_rate": 9.589455877710018e-05,
+      "loss": 12.4422,
+      "step": 9052
+    },
+    {
+      "epoch": 1.5831693989071038,
+      "grad_norm": 0.967234920270152,
+      "learning_rate": 9.587569387043138e-05,
+      "loss": 12.5551,
+      "step": 9053
+    },
+    {
+      "epoch": 1.583344262295082,
+      "grad_norm": 0.8325077482351456,
+      "learning_rate": 9.585682911078923e-05,
+      "loss": 12.322,
+      "step": 9054
+    },
+    {
+      "epoch": 1.58351912568306,
+      "grad_norm": 0.8339228756462773,
+      "learning_rate": 9.58379644988463e-05,
+      "loss": 12.5839,
+      "step": 9055
+    },
+    {
+      "epoch": 1.5836939890710382,
+      "grad_norm": 0.9184216031725403,
+      "learning_rate": 9.581910003527505e-05,
+      "loss": 12.2295,
+      "step": 9056
+    },
+    {
+      "epoch": 1.5838688524590165,
+      "grad_norm": 0.8832283059564053,
+      "learning_rate": 9.580023572074803e-05,
+      "loss": 12.2917,
+      "step": 9057
+    },
+    {
+      "epoch": 1.5840437158469944,
+      "grad_norm": 0.9244456591107132,
+      "learning_rate": 9.578137155593768e-05,
+      "loss": 12.4364,
+      "step": 9058
+    },
+    {
+      "epoch": 1.5842185792349728,
+      "grad_norm": 0.7974513329226408,
+      "learning_rate": 9.57625075415165e-05,
+      "loss": 12.4683,
+      "step": 9059
+    },
+    {
+      "epoch": 1.5843934426229507,
+      "grad_norm": 0.8313927393843281,
+      "learning_rate": 9.574364367815698e-05,
+      "loss": 12.3713,
+      "step": 9060
+    },
+    {
+      "epoch": 1.584568306010929,
+      "grad_norm": 0.8410503894672294,
+      "learning_rate": 9.572477996653164e-05,
+      "loss": 12.0423,
+      "step": 9061
+    },
+    {
+      "epoch": 1.5847431693989071,
+      "grad_norm": 0.9088459075965052,
+      "learning_rate": 9.570591640731283e-05,
+      "loss": 12.5024,
+      "step": 9062
+    },
+    {
+      "epoch": 1.5849180327868853,
+      "grad_norm": 0.9075083274391992,
+      "learning_rate": 9.56870530011731e-05,
+      "loss": 12.5576,
+      "step": 9063
+    },
+    {
+      "epoch": 1.5850928961748634,
+      "grad_norm": 0.7936951487187851,
+      "learning_rate": 9.566818974878491e-05,
+      "loss": 12.2531,
+      "step": 9064
+    },
+    {
+      "epoch": 1.5852677595628415,
+      "grad_norm": 0.8727306554903502,
+      "learning_rate": 9.564932665082068e-05,
+      "loss": 12.4479,
+      "step": 9065
+    },
+    {
+      "epoch": 1.5854426229508198,
+      "grad_norm": 0.8900353168383224,
+      "learning_rate": 9.563046370795287e-05,
+      "loss": 12.5118,
+      "step": 9066
+    },
+    {
+      "epoch": 1.5856174863387977,
+      "grad_norm": 0.8085027722050077,
+      "learning_rate": 9.561160092085392e-05,
+      "loss": 12.4972,
+      "step": 9067
+    },
+    {
+      "epoch": 1.585792349726776,
+      "grad_norm": 0.8008809813929234,
+      "learning_rate": 9.559273829019631e-05,
+      "loss": 12.3947,
+      "step": 9068
+    },
+    {
+      "epoch": 1.585967213114754,
+      "grad_norm": 0.870196795733139,
+      "learning_rate": 9.55738758166524e-05,
+      "loss": 12.6319,
+      "step": 9069
+    },
+    {
+      "epoch": 1.5861420765027323,
+      "grad_norm": 0.831395957203009,
+      "learning_rate": 9.555501350089463e-05,
+      "loss": 12.5075,
+      "step": 9070
+    },
+    {
+      "epoch": 1.5863169398907104,
+      "grad_norm": 0.8463108784720633,
+      "learning_rate": 9.553615134359544e-05,
+      "loss": 12.4845,
+      "step": 9071
+    },
+    {
+      "epoch": 1.5864918032786886,
+      "grad_norm": 0.8742643777627114,
+      "learning_rate": 9.551728934542728e-05,
+      "loss": 12.4602,
+      "step": 9072
+    },
+    {
+      "epoch": 1.5866666666666667,
+      "grad_norm": 0.870615523497562,
+      "learning_rate": 9.549842750706248e-05,
+      "loss": 12.4975,
+      "step": 9073
+    },
+    {
+      "epoch": 1.5868415300546448,
+      "grad_norm": 0.8286114735263692,
+      "learning_rate": 9.547956582917347e-05,
+      "loss": 12.314,
+      "step": 9074
+    },
+    {
+      "epoch": 1.587016393442623,
+      "grad_norm": 0.8405123766372462,
+      "learning_rate": 9.546070431243269e-05,
+      "loss": 12.3999,
+      "step": 9075
+    },
+    {
+      "epoch": 1.587191256830601,
+      "grad_norm": 0.8315690560311912,
+      "learning_rate": 9.544184295751248e-05,
+      "loss": 12.5531,
+      "step": 9076
+    },
+    {
+      "epoch": 1.5873661202185794,
+      "grad_norm": 0.8598289965093098,
+      "learning_rate": 9.542298176508524e-05,
+      "loss": 12.5457,
+      "step": 9077
+    },
+    {
+      "epoch": 1.5875409836065573,
+      "grad_norm": 0.8533391703207498,
+      "learning_rate": 9.540412073582334e-05,
+      "loss": 12.4609,
+      "step": 9078
+    },
+    {
+      "epoch": 1.5877158469945356,
+      "grad_norm": 0.8794252389214742,
+      "learning_rate": 9.538525987039923e-05,
+      "loss": 11.9555,
+      "step": 9079
+    },
+    {
+      "epoch": 1.5878907103825135,
+      "grad_norm": 0.8913029583366625,
+      "learning_rate": 9.536639916948514e-05,
+      "loss": 12.6049,
+      "step": 9080
+    },
+    {
+      "epoch": 1.5880655737704918,
+      "grad_norm": 0.8933941858166918,
+      "learning_rate": 9.534753863375353e-05,
+      "loss": 12.4691,
+      "step": 9081
+    },
+    {
+      "epoch": 1.58824043715847,
+      "grad_norm": 0.8274151412838159,
+      "learning_rate": 9.532867826387672e-05,
+      "loss": 12.2968,
+      "step": 9082
+    },
+    {
+      "epoch": 1.588415300546448,
+      "grad_norm": 0.8892355622929394,
+      "learning_rate": 9.530981806052711e-05,
+      "loss": 12.5446,
+      "step": 9083
+    },
+    {
+      "epoch": 1.5885901639344262,
+      "grad_norm": 0.8920104786506649,
+      "learning_rate": 9.529095802437697e-05,
+      "loss": 12.2938,
+      "step": 9084
+    },
+    {
+      "epoch": 1.5887650273224043,
+      "grad_norm": 0.8184119029314825,
+      "learning_rate": 9.527209815609868e-05,
+      "loss": 12.4442,
+      "step": 9085
+    },
+    {
+      "epoch": 1.5889398907103827,
+      "grad_norm": 0.8581141064777039,
+      "learning_rate": 9.525323845636456e-05,
+      "loss": 12.4101,
+      "step": 9086
+    },
+    {
+      "epoch": 1.5891147540983606,
+      "grad_norm": 0.973832257554075,
+      "learning_rate": 9.523437892584699e-05,
+      "loss": 12.5021,
+      "step": 9087
+    },
+    {
+      "epoch": 1.589289617486339,
+      "grad_norm": 0.9088902605733656,
+      "learning_rate": 9.52155195652182e-05,
+      "loss": 12.4065,
+      "step": 9088
+    },
+    {
+      "epoch": 1.5894644808743168,
+      "grad_norm": 0.8872026803660271,
+      "learning_rate": 9.519666037515055e-05,
+      "loss": 12.3396,
+      "step": 9089
+    },
+    {
+      "epoch": 1.5896393442622951,
+      "grad_norm": 0.8157882427795301,
+      "learning_rate": 9.517780135631635e-05,
+      "loss": 12.6172,
+      "step": 9090
+    },
+    {
+      "epoch": 1.5898142076502733,
+      "grad_norm": 1.038791878200524,
+      "learning_rate": 9.515894250938793e-05,
+      "loss": 12.3453,
+      "step": 9091
+    },
+    {
+      "epoch": 1.5899890710382514,
+      "grad_norm": 0.8363525308639861,
+      "learning_rate": 9.514008383503752e-05,
+      "loss": 12.5453,
+      "step": 9092
+    },
+    {
+      "epoch": 1.5901639344262295,
+      "grad_norm": 0.868885836509341,
+      "learning_rate": 9.512122533393744e-05,
+      "loss": 12.2911,
+      "step": 9093
+    },
+    {
+      "epoch": 1.5903387978142076,
+      "grad_norm": 0.8627763229030093,
+      "learning_rate": 9.510236700675998e-05,
+      "loss": 12.4425,
+      "step": 9094
+    },
+    {
+      "epoch": 1.590513661202186,
+      "grad_norm": 0.9247487646014687,
+      "learning_rate": 9.508350885417742e-05,
+      "loss": 12.2613,
+      "step": 9095
+    },
+    {
+      "epoch": 1.5906885245901639,
+      "grad_norm": 0.8656546840106795,
+      "learning_rate": 9.506465087686202e-05,
+      "loss": 12.3981,
+      "step": 9096
+    },
+    {
+      "epoch": 1.5908633879781422,
+      "grad_norm": 0.784881234317142,
+      "learning_rate": 9.504579307548603e-05,
+      "loss": 12.4106,
+      "step": 9097
+    },
+    {
+      "epoch": 1.59103825136612,
+      "grad_norm": 0.8049295615110604,
+      "learning_rate": 9.502693545072179e-05,
+      "loss": 12.5169,
+      "step": 9098
+    },
+    {
+      "epoch": 1.5912131147540984,
+      "grad_norm": 0.9062309001343105,
+      "learning_rate": 9.500807800324143e-05,
+      "loss": 12.4649,
+      "step": 9099
+    },
+    {
+      "epoch": 1.5913879781420766,
+      "grad_norm": 0.8552631334361153,
+      "learning_rate": 9.498922073371728e-05,
+      "loss": 12.5036,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5915628415300547,
+      "grad_norm": 1.0063060221622697,
+      "learning_rate": 9.497036364282155e-05,
+      "loss": 12.4932,
+      "step": 9101
+    },
+    {
+      "epoch": 1.5917377049180328,
+      "grad_norm": 0.8677169906913941,
+      "learning_rate": 9.49515067312265e-05,
+      "loss": 12.3972,
+      "step": 9102
+    },
+    {
+      "epoch": 1.591912568306011,
+      "grad_norm": 1.1708261060778746,
+      "learning_rate": 9.49326499996043e-05,
+      "loss": 12.3336,
+      "step": 9103
+    },
+    {
+      "epoch": 1.592087431693989,
+      "grad_norm": 0.8980400827159744,
+      "learning_rate": 9.491379344862723e-05,
+      "loss": 12.4835,
+      "step": 9104
+    },
+    {
+      "epoch": 1.5922622950819671,
+      "grad_norm": 0.9467510993665783,
+      "learning_rate": 9.489493707896746e-05,
+      "loss": 12.3918,
+      "step": 9105
+    },
+    {
+      "epoch": 1.5924371584699455,
+      "grad_norm": 0.9589427841984315,
+      "learning_rate": 9.487608089129729e-05,
+      "loss": 12.4072,
+      "step": 9106
+    },
+    {
+      "epoch": 1.5926120218579234,
+      "grad_norm": 0.9202204517939178,
+      "learning_rate": 9.485722488628878e-05,
+      "loss": 12.4493,
+      "step": 9107
+    },
+    {
+      "epoch": 1.5927868852459017,
+      "grad_norm": 1.0680938916435512,
+      "learning_rate": 9.48383690646142e-05,
+      "loss": 12.3427,
+      "step": 9108
+    },
+    {
+      "epoch": 1.5929617486338796,
+      "grad_norm": 0.8740385786912839,
+      "learning_rate": 9.481951342694575e-05,
+      "loss": 12.304,
+      "step": 9109
+    },
+    {
+      "epoch": 1.593136612021858,
+      "grad_norm": 1.1143558421610373,
+      "learning_rate": 9.480065797395558e-05,
+      "loss": 11.9776,
+      "step": 9110
+    },
+    {
+      "epoch": 1.593311475409836,
+      "grad_norm": 0.8517889187662474,
+      "learning_rate": 9.478180270631589e-05,
+      "loss": 12.4196,
+      "step": 9111
+    },
+    {
+      "epoch": 1.5934863387978142,
+      "grad_norm": 0.9687257801374577,
+      "learning_rate": 9.47629476246988e-05,
+      "loss": 12.7338,
+      "step": 9112
+    },
+    {
+      "epoch": 1.5936612021857923,
+      "grad_norm": 0.978512498688004,
+      "learning_rate": 9.474409272977659e-05,
+      "loss": 12.68,
+      "step": 9113
+    },
+    {
+      "epoch": 1.5938360655737704,
+      "grad_norm": 0.8758463366996168,
+      "learning_rate": 9.472523802222128e-05,
+      "loss": 12.5247,
+      "step": 9114
+    },
+    {
+      "epoch": 1.5940109289617488,
+      "grad_norm": 0.9362798191335562,
+      "learning_rate": 9.470638350270505e-05,
+      "loss": 12.621,
+      "step": 9115
+    },
+    {
+      "epoch": 1.5941857923497267,
+      "grad_norm": 0.8350235518824751,
+      "learning_rate": 9.468752917190009e-05,
+      "loss": 12.378,
+      "step": 9116
+    },
+    {
+      "epoch": 1.594360655737705,
+      "grad_norm": 0.7759432934070447,
+      "learning_rate": 9.466867503047853e-05,
+      "loss": 12.4521,
+      "step": 9117
+    },
+    {
+      "epoch": 1.594535519125683,
+      "grad_norm": 0.8397690855827101,
+      "learning_rate": 9.464982107911245e-05,
+      "loss": 12.7081,
+      "step": 9118
+    },
+    {
+      "epoch": 1.5947103825136613,
+      "grad_norm": 0.7967771490088285,
+      "learning_rate": 9.4630967318474e-05,
+      "loss": 12.0911,
+      "step": 9119
+    },
+    {
+      "epoch": 1.5948852459016394,
+      "grad_norm": 0.7818902888496055,
+      "learning_rate": 9.461211374923527e-05,
+      "loss": 12.3914,
+      "step": 9120
+    },
+    {
+      "epoch": 1.5950601092896175,
+      "grad_norm": 0.8227705450770788,
+      "learning_rate": 9.459326037206842e-05,
+      "loss": 12.1416,
+      "step": 9121
+    },
+    {
+      "epoch": 1.5952349726775956,
+      "grad_norm": 0.950806389992197,
+      "learning_rate": 9.457440718764552e-05,
+      "loss": 12.391,
+      "step": 9122
+    },
+    {
+      "epoch": 1.5954098360655737,
+      "grad_norm": 0.8950979362730641,
+      "learning_rate": 9.455555419663865e-05,
+      "loss": 12.4636,
+      "step": 9123
+    },
+    {
+      "epoch": 1.5955846994535519,
+      "grad_norm": 0.8454099781690252,
+      "learning_rate": 9.453670139971995e-05,
+      "loss": 12.4009,
+      "step": 9124
+    },
+    {
+      "epoch": 1.59575956284153,
+      "grad_norm": 0.7998160841264482,
+      "learning_rate": 9.451784879756141e-05,
+      "loss": 12.377,
+      "step": 9125
+    },
+    {
+      "epoch": 1.5959344262295083,
+      "grad_norm": 0.8911315469439166,
+      "learning_rate": 9.449899639083517e-05,
+      "loss": 12.4463,
+      "step": 9126
+    },
+    {
+      "epoch": 1.5961092896174862,
+      "grad_norm": 0.8293074890483108,
+      "learning_rate": 9.448014418021328e-05,
+      "loss": 12.629,
+      "step": 9127
+    },
+    {
+      "epoch": 1.5962841530054646,
+      "grad_norm": 0.7827429138740363,
+      "learning_rate": 9.446129216636782e-05,
+      "loss": 12.0536,
+      "step": 9128
+    },
+    {
+      "epoch": 1.5964590163934427,
+      "grad_norm": 0.8443578277327008,
+      "learning_rate": 9.44424403499708e-05,
+      "loss": 12.4106,
+      "step": 9129
+    },
+    {
+      "epoch": 1.5966338797814208,
+      "grad_norm": 0.8484249317136138,
+      "learning_rate": 9.44235887316943e-05,
+      "loss": 12.2182,
+      "step": 9130
+    },
+    {
+      "epoch": 1.596808743169399,
+      "grad_norm": 0.8302306822565308,
+      "learning_rate": 9.440473731221034e-05,
+      "loss": 12.6653,
+      "step": 9131
+    },
+    {
+      "epoch": 1.596983606557377,
+      "grad_norm": 0.7879489677488947,
+      "learning_rate": 9.4385886092191e-05,
+      "loss": 12.3698,
+      "step": 9132
+    },
+    {
+      "epoch": 1.5971584699453552,
+      "grad_norm": 0.8551207513111081,
+      "learning_rate": 9.43670350723082e-05,
+      "loss": 12.0978,
+      "step": 9133
+    },
+    {
+      "epoch": 1.5973333333333333,
+      "grad_norm": 0.8333316952959201,
+      "learning_rate": 9.434818425323403e-05,
+      "loss": 12.4851,
+      "step": 9134
+    },
+    {
+      "epoch": 1.5975081967213116,
+      "grad_norm": 0.8869094259457061,
+      "learning_rate": 9.43293336356405e-05,
+      "loss": 12.5453,
+      "step": 9135
+    },
+    {
+      "epoch": 1.5976830601092895,
+      "grad_norm": 0.9627226700311413,
+      "learning_rate": 9.431048322019964e-05,
+      "loss": 12.3008,
+      "step": 9136
+    },
+    {
+      "epoch": 1.5978579234972679,
+      "grad_norm": 0.828364813878328,
+      "learning_rate": 9.429163300758336e-05,
+      "loss": 12.293,
+      "step": 9137
+    },
+    {
+      "epoch": 1.5980327868852457,
+      "grad_norm": 1.1941648129684135,
+      "learning_rate": 9.42727829984637e-05,
+      "loss": 12.6042,
+      "step": 9138
+    },
+    {
+      "epoch": 1.598207650273224,
+      "grad_norm": 0.9047815572073054,
+      "learning_rate": 9.425393319351264e-05,
+      "loss": 12.491,
+      "step": 9139
+    },
+    {
+      "epoch": 1.5983825136612022,
+      "grad_norm": 0.8027740316318034,
+      "learning_rate": 9.423508359340213e-05,
+      "loss": 12.2377,
+      "step": 9140
+    },
+    {
+      "epoch": 1.5985573770491803,
+      "grad_norm": 0.9713996294886109,
+      "learning_rate": 9.421623419880417e-05,
+      "loss": 12.3114,
+      "step": 9141
+    },
+    {
+      "epoch": 1.5987322404371584,
+      "grad_norm": 0.8810147581748409,
+      "learning_rate": 9.419738501039068e-05,
+      "loss": 12.5076,
+      "step": 9142
+    },
+    {
+      "epoch": 1.5989071038251366,
+      "grad_norm": 1.0448210877555044,
+      "learning_rate": 9.41785360288337e-05,
+      "loss": 12.5632,
+      "step": 9143
+    },
+    {
+      "epoch": 1.599081967213115,
+      "grad_norm": 0.891085462731894,
+      "learning_rate": 9.415968725480506e-05,
+      "loss": 12.5641,
+      "step": 9144
+    },
+    {
+      "epoch": 1.5992568306010928,
+      "grad_norm": 0.9772724801465503,
+      "learning_rate": 9.414083868897672e-05,
+      "loss": 12.5012,
+      "step": 9145
+    },
+    {
+      "epoch": 1.5994316939890711,
+      "grad_norm": 0.945352200134951,
+      "learning_rate": 9.412199033202067e-05,
+      "loss": 12.4978,
+      "step": 9146
+    },
+    {
+      "epoch": 1.599606557377049,
+      "grad_norm": 1.1887425777117213,
+      "learning_rate": 9.41031421846088e-05,
+      "loss": 12.6282,
+      "step": 9147
+    },
+    {
+      "epoch": 1.5997814207650274,
+      "grad_norm": 1.1052531792161464,
+      "learning_rate": 9.4084294247413e-05,
+      "loss": 12.3821,
+      "step": 9148
+    },
+    {
+      "epoch": 1.5999562841530055,
+      "grad_norm": 1.000856787033423,
+      "learning_rate": 9.406544652110521e-05,
+      "loss": 12.5807,
+      "step": 9149
+    },
+    {
+      "epoch": 1.6001311475409836,
+      "grad_norm": 1.1513135142247821,
+      "learning_rate": 9.40465990063573e-05,
+      "loss": 12.6099,
+      "step": 9150
+    },
+    {
+      "epoch": 1.6003060109289617,
+      "grad_norm": 1.0031614651159482,
+      "learning_rate": 9.402775170384125e-05,
+      "loss": 12.2912,
+      "step": 9151
+    },
+    {
+      "epoch": 1.6004808743169399,
+      "grad_norm": 0.9860450641782019,
+      "learning_rate": 9.40089046142288e-05,
+      "loss": 12.5327,
+      "step": 9152
+    },
+    {
+      "epoch": 1.600655737704918,
+      "grad_norm": 0.9706838169479816,
+      "learning_rate": 9.399005773819193e-05,
+      "loss": 12.208,
+      "step": 9153
+    },
+    {
+      "epoch": 1.600830601092896,
+      "grad_norm": 0.8911704737703784,
+      "learning_rate": 9.397121107640248e-05,
+      "loss": 11.9634,
+      "step": 9154
+    },
+    {
+      "epoch": 1.6010054644808744,
+      "grad_norm": 1.0621536406273033,
+      "learning_rate": 9.39523646295323e-05,
+      "loss": 12.5084,
+      "step": 9155
+    },
+    {
+      "epoch": 1.6011803278688523,
+      "grad_norm": 0.9475976807941711,
+      "learning_rate": 9.393351839825325e-05,
+      "loss": 12.3474,
+      "step": 9156
+    },
+    {
+      "epoch": 1.6013551912568307,
+      "grad_norm": 0.8677473812426292,
+      "learning_rate": 9.391467238323718e-05,
+      "loss": 12.5107,
+      "step": 9157
+    },
+    {
+      "epoch": 1.6015300546448088,
+      "grad_norm": 0.8810014879144001,
+      "learning_rate": 9.389582658515598e-05,
+      "loss": 12.4087,
+      "step": 9158
+    },
+    {
+      "epoch": 1.601704918032787,
+      "grad_norm": 0.8934431574199322,
+      "learning_rate": 9.387698100468137e-05,
+      "loss": 12.5847,
+      "step": 9159
+    },
+    {
+      "epoch": 1.601879781420765,
+      "grad_norm": 0.9426174604528907,
+      "learning_rate": 9.385813564248526e-05,
+      "loss": 12.5891,
+      "step": 9160
+    },
+    {
+      "epoch": 1.6020546448087432,
+      "grad_norm": 0.8773714493147355,
+      "learning_rate": 9.383929049923942e-05,
+      "loss": 12.5986,
+      "step": 9161
+    },
+    {
+      "epoch": 1.6022295081967213,
+      "grad_norm": 0.9489013501472907,
+      "learning_rate": 9.382044557561572e-05,
+      "loss": 12.6748,
+      "step": 9162
+    },
+    {
+      "epoch": 1.6024043715846994,
+      "grad_norm": 0.9212329520364723,
+      "learning_rate": 9.380160087228588e-05,
+      "loss": 12.4548,
+      "step": 9163
+    },
+    {
+      "epoch": 1.6025792349726777,
+      "grad_norm": 0.8256087013859104,
+      "learning_rate": 9.378275638992172e-05,
+      "loss": 12.555,
+      "step": 9164
+    },
+    {
+      "epoch": 1.6027540983606556,
+      "grad_norm": 0.9210042143004221,
+      "learning_rate": 9.376391212919503e-05,
+      "loss": 12.3014,
+      "step": 9165
+    },
+    {
+      "epoch": 1.602928961748634,
+      "grad_norm": 0.9152215077808141,
+      "learning_rate": 9.374506809077761e-05,
+      "loss": 12.5469,
+      "step": 9166
+    },
+    {
+      "epoch": 1.6031038251366119,
+      "grad_norm": 0.8910031975986785,
+      "learning_rate": 9.372622427534118e-05,
+      "loss": 12.5688,
+      "step": 9167
+    },
+    {
+      "epoch": 1.6032786885245902,
+      "grad_norm": 0.9174577653029024,
+      "learning_rate": 9.370738068355753e-05,
+      "loss": 12.4325,
+      "step": 9168
+    },
+    {
+      "epoch": 1.6034535519125683,
+      "grad_norm": 0.8512999492404971,
+      "learning_rate": 9.36885373160984e-05,
+      "loss": 12.3809,
+      "step": 9169
+    },
+    {
+      "epoch": 1.6036284153005465,
+      "grad_norm": 0.8031276710547048,
+      "learning_rate": 9.36696941736356e-05,
+      "loss": 12.4612,
+      "step": 9170
+    },
+    {
+      "epoch": 1.6038032786885246,
+      "grad_norm": 0.8698881883505654,
+      "learning_rate": 9.365085125684075e-05,
+      "loss": 12.3364,
+      "step": 9171
+    },
+    {
+      "epoch": 1.6039781420765027,
+      "grad_norm": 0.8594588739832696,
+      "learning_rate": 9.363200856638563e-05,
+      "loss": 12.4644,
+      "step": 9172
+    },
+    {
+      "epoch": 1.604153005464481,
+      "grad_norm": 0.8617345877545352,
+      "learning_rate": 9.361316610294199e-05,
+      "loss": 12.5356,
+      "step": 9173
+    },
+    {
+      "epoch": 1.604327868852459,
+      "grad_norm": 0.8538853630262284,
+      "learning_rate": 9.35943238671815e-05,
+      "loss": 11.9978,
+      "step": 9174
+    },
+    {
+      "epoch": 1.6045027322404373,
+      "grad_norm": 0.9138287512673946,
+      "learning_rate": 9.357548185977587e-05,
+      "loss": 12.3738,
+      "step": 9175
+    },
+    {
+      "epoch": 1.6046775956284152,
+      "grad_norm": 0.8549791421339404,
+      "learning_rate": 9.35566400813968e-05,
+      "loss": 12.562,
+      "step": 9176
+    },
+    {
+      "epoch": 1.6048524590163935,
+      "grad_norm": 0.9553125376547841,
+      "learning_rate": 9.353779853271605e-05,
+      "loss": 12.6147,
+      "step": 9177
+    },
+    {
+      "epoch": 1.6050273224043716,
+      "grad_norm": 0.93963318128475,
+      "learning_rate": 9.351895721440517e-05,
+      "loss": 12.6503,
+      "step": 9178
+    },
+    {
+      "epoch": 1.6052021857923497,
+      "grad_norm": 0.805294240686669,
+      "learning_rate": 9.350011612713591e-05,
+      "loss": 12.4713,
+      "step": 9179
+    },
+    {
+      "epoch": 1.6053770491803279,
+      "grad_norm": 0.9016180152390687,
+      "learning_rate": 9.34812752715799e-05,
+      "loss": 12.5634,
+      "step": 9180
+    },
+    {
+      "epoch": 1.605551912568306,
+      "grad_norm": 0.8530752231134118,
+      "learning_rate": 9.346243464840885e-05,
+      "loss": 12.4223,
+      "step": 9181
+    },
+    {
+      "epoch": 1.605726775956284,
+      "grad_norm": 0.9756898553188798,
+      "learning_rate": 9.344359425829434e-05,
+      "loss": 12.1848,
+      "step": 9182
+    },
+    {
+      "epoch": 1.6059016393442622,
+      "grad_norm": 0.8840354301336354,
+      "learning_rate": 9.342475410190803e-05,
+      "loss": 12.369,
+      "step": 9183
+    },
+    {
+      "epoch": 1.6060765027322406,
+      "grad_norm": 1.019663358049355,
+      "learning_rate": 9.340591417992155e-05,
+      "loss": 12.5015,
+      "step": 9184
+    },
+    {
+      "epoch": 1.6062513661202185,
+      "grad_norm": 0.8379411377104817,
+      "learning_rate": 9.338707449300655e-05,
+      "loss": 12.3347,
+      "step": 9185
+    },
+    {
+      "epoch": 1.6064262295081968,
+      "grad_norm": 1.1274819349882819,
+      "learning_rate": 9.336823504183458e-05,
+      "loss": 12.458,
+      "step": 9186
+    },
+    {
+      "epoch": 1.6066010928961747,
+      "grad_norm": 0.9821725799443815,
+      "learning_rate": 9.33493958270773e-05,
+      "loss": 12.5881,
+      "step": 9187
+    },
+    {
+      "epoch": 1.606775956284153,
+      "grad_norm": 0.8907944737894422,
+      "learning_rate": 9.333055684940634e-05,
+      "loss": 12.4629,
+      "step": 9188
+    },
+    {
+      "epoch": 1.6069508196721312,
+      "grad_norm": 1.0302305984248619,
+      "learning_rate": 9.331171810949318e-05,
+      "loss": 12.2219,
+      "step": 9189
+    },
+    {
+      "epoch": 1.6071256830601093,
+      "grad_norm": 0.8463831402884162,
+      "learning_rate": 9.329287960800944e-05,
+      "loss": 12.6355,
+      "step": 9190
+    },
+    {
+      "epoch": 1.6073005464480874,
+      "grad_norm": 0.8815554686903941,
+      "learning_rate": 9.327404134562673e-05,
+      "loss": 12.19,
+      "step": 9191
+    },
+    {
+      "epoch": 1.6074754098360655,
+      "grad_norm": 0.823243169026288,
+      "learning_rate": 9.32552033230166e-05,
+      "loss": 12.0459,
+      "step": 9192
+    },
+    {
+      "epoch": 1.6076502732240439,
+      "grad_norm": 0.8733996240457796,
+      "learning_rate": 9.323636554085057e-05,
+      "loss": 12.353,
+      "step": 9193
+    },
+    {
+      "epoch": 1.6078251366120218,
+      "grad_norm": 0.8758918773788839,
+      "learning_rate": 9.32175279998002e-05,
+      "loss": 12.3893,
+      "step": 9194
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.9151784695132588,
+      "learning_rate": 9.319869070053705e-05,
+      "loss": 12.5729,
+      "step": 9195
+    },
+    {
+      "epoch": 1.608174863387978,
+      "grad_norm": 0.8851144216260559,
+      "learning_rate": 9.317985364373266e-05,
+      "loss": 12.4362,
+      "step": 9196
+    },
+    {
+      "epoch": 1.6083497267759563,
+      "grad_norm": 0.8405834885527884,
+      "learning_rate": 9.316101683005846e-05,
+      "loss": 12.5267,
+      "step": 9197
+    },
+    {
+      "epoch": 1.6085245901639345,
+      "grad_norm": 0.9346443437756583,
+      "learning_rate": 9.314218026018603e-05,
+      "loss": 12.3936,
+      "step": 9198
+    },
+    {
+      "epoch": 1.6086994535519126,
+      "grad_norm": 0.8792295269499714,
+      "learning_rate": 9.312334393478688e-05,
+      "loss": 12.2187,
+      "step": 9199
+    },
+    {
+      "epoch": 1.6088743169398907,
+      "grad_norm": 0.8809021914641177,
+      "learning_rate": 9.31045078545325e-05,
+      "loss": 12.3528,
+      "step": 9200
+    },
+    {
+      "epoch": 1.6090491803278688,
+      "grad_norm": 0.83331779456664,
+      "learning_rate": 9.308567202009432e-05,
+      "loss": 12.2628,
+      "step": 9201
+    },
+    {
+      "epoch": 1.6092240437158472,
+      "grad_norm": 0.8533453320553586,
+      "learning_rate": 9.306683643214386e-05,
+      "loss": 12.5551,
+      "step": 9202
+    },
+    {
+      "epoch": 1.609398907103825,
+      "grad_norm": 0.9352583135875334,
+      "learning_rate": 9.304800109135263e-05,
+      "loss": 12.5922,
+      "step": 9203
+    },
+    {
+      "epoch": 1.6095737704918034,
+      "grad_norm": 0.8879327585158995,
+      "learning_rate": 9.3029165998392e-05,
+      "loss": 12.5313,
+      "step": 9204
+    },
+    {
+      "epoch": 1.6097486338797813,
+      "grad_norm": 0.8917995030285945,
+      "learning_rate": 9.301033115393345e-05,
+      "loss": 12.4673,
+      "step": 9205
+    },
+    {
+      "epoch": 1.6099234972677596,
+      "grad_norm": 0.8161012796040856,
+      "learning_rate": 9.299149655864843e-05,
+      "loss": 12.2217,
+      "step": 9206
+    },
+    {
+      "epoch": 1.6100983606557377,
+      "grad_norm": 0.8781404086379982,
+      "learning_rate": 9.29726622132084e-05,
+      "loss": 12.663,
+      "step": 9207
+    },
+    {
+      "epoch": 1.6102732240437159,
+      "grad_norm": 0.9625553450546545,
+      "learning_rate": 9.295382811828472e-05,
+      "loss": 12.3889,
+      "step": 9208
+    },
+    {
+      "epoch": 1.610448087431694,
+      "grad_norm": 0.8769078393487654,
+      "learning_rate": 9.293499427454884e-05,
+      "loss": 12.5203,
+      "step": 9209
+    },
+    {
+      "epoch": 1.610622950819672,
+      "grad_norm": 0.815079194974489,
+      "learning_rate": 9.291616068267215e-05,
+      "loss": 12.2187,
+      "step": 9210
+    },
+    {
+      "epoch": 1.6107978142076502,
+      "grad_norm": 0.9042406640808806,
+      "learning_rate": 9.289732734332609e-05,
+      "loss": 12.3772,
+      "step": 9211
+    },
+    {
+      "epoch": 1.6109726775956283,
+      "grad_norm": 0.8901117329381276,
+      "learning_rate": 9.287849425718198e-05,
+      "loss": 12.4652,
+      "step": 9212
+    },
+    {
+      "epoch": 1.6111475409836067,
+      "grad_norm": 0.9297556656212488,
+      "learning_rate": 9.285966142491122e-05,
+      "loss": 12.0605,
+      "step": 9213
+    },
+    {
+      "epoch": 1.6113224043715846,
+      "grad_norm": 0.8504354099962553,
+      "learning_rate": 9.284082884718521e-05,
+      "loss": 12.2946,
+      "step": 9214
+    },
+    {
+      "epoch": 1.611497267759563,
+      "grad_norm": 0.8506932788579944,
+      "learning_rate": 9.282199652467532e-05,
+      "loss": 12.2067,
+      "step": 9215
+    },
+    {
+      "epoch": 1.6116721311475408,
+      "grad_norm": 0.9526720041892951,
+      "learning_rate": 9.280316445805282e-05,
+      "loss": 12.6135,
+      "step": 9216
+    },
+    {
+      "epoch": 1.6118469945355192,
+      "grad_norm": 0.8907670282228053,
+      "learning_rate": 9.27843326479891e-05,
+      "loss": 12.3808,
+      "step": 9217
+    },
+    {
+      "epoch": 1.6120218579234973,
+      "grad_norm": 0.9291729410306363,
+      "learning_rate": 9.27655010951555e-05,
+      "loss": 12.5781,
+      "step": 9218
+    },
+    {
+      "epoch": 1.6121967213114754,
+      "grad_norm": 0.8623482443961867,
+      "learning_rate": 9.274666980022332e-05,
+      "loss": 12.6569,
+      "step": 9219
+    },
+    {
+      "epoch": 1.6123715846994535,
+      "grad_norm": 0.9929961909937107,
+      "learning_rate": 9.27278387638639e-05,
+      "loss": 12.2892,
+      "step": 9220
+    },
+    {
+      "epoch": 1.6125464480874316,
+      "grad_norm": 0.7888702160069027,
+      "learning_rate": 9.270900798674853e-05,
+      "loss": 12.5402,
+      "step": 9221
+    },
+    {
+      "epoch": 1.61272131147541,
+      "grad_norm": 0.9320579876001506,
+      "learning_rate": 9.269017746954853e-05,
+      "loss": 12.2163,
+      "step": 9222
+    },
+    {
+      "epoch": 1.6128961748633879,
+      "grad_norm": 0.9661294400113405,
+      "learning_rate": 9.267134721293514e-05,
+      "loss": 12.6429,
+      "step": 9223
+    },
+    {
+      "epoch": 1.6130710382513662,
+      "grad_norm": 0.9001271859475077,
+      "learning_rate": 9.265251721757964e-05,
+      "loss": 12.4442,
+      "step": 9224
+    },
+    {
+      "epoch": 1.6132459016393441,
+      "grad_norm": 0.9136439013773435,
+      "learning_rate": 9.263368748415334e-05,
+      "loss": 12.5176,
+      "step": 9225
+    },
+    {
+      "epoch": 1.6134207650273225,
+      "grad_norm": 0.8958938884445352,
+      "learning_rate": 9.261485801332749e-05,
+      "loss": 12.5305,
+      "step": 9226
+    },
+    {
+      "epoch": 1.6135956284153006,
+      "grad_norm": 0.8803635828099189,
+      "learning_rate": 9.259602880577331e-05,
+      "loss": 12.2536,
+      "step": 9227
+    },
+    {
+      "epoch": 1.6137704918032787,
+      "grad_norm": 0.9814415516307501,
+      "learning_rate": 9.257719986216204e-05,
+      "loss": 12.7285,
+      "step": 9228
+    },
+    {
+      "epoch": 1.6139453551912568,
+      "grad_norm": 0.8612659980231753,
+      "learning_rate": 9.255837118316491e-05,
+      "loss": 12.4352,
+      "step": 9229
+    },
+    {
+      "epoch": 1.614120218579235,
+      "grad_norm": 0.9517924639727676,
+      "learning_rate": 9.253954276945319e-05,
+      "loss": 12.1532,
+      "step": 9230
+    },
+    {
+      "epoch": 1.6142950819672133,
+      "grad_norm": 0.8656487429690173,
+      "learning_rate": 9.252071462169803e-05,
+      "loss": 12.4289,
+      "step": 9231
+    },
+    {
+      "epoch": 1.6144699453551912,
+      "grad_norm": 0.8056651847403206,
+      "learning_rate": 9.250188674057065e-05,
+      "loss": 12.4904,
+      "step": 9232
+    },
+    {
+      "epoch": 1.6146448087431695,
+      "grad_norm": 0.9107453097692721,
+      "learning_rate": 9.248305912674228e-05,
+      "loss": 12.4557,
+      "step": 9233
+    },
+    {
+      "epoch": 1.6148196721311474,
+      "grad_norm": 0.8191687374600801,
+      "learning_rate": 9.246423178088405e-05,
+      "loss": 12.24,
+      "step": 9234
+    },
+    {
+      "epoch": 1.6149945355191258,
+      "grad_norm": 0.8489867860838588,
+      "learning_rate": 9.244540470366714e-05,
+      "loss": 12.3728,
+      "step": 9235
+    },
+    {
+      "epoch": 1.6151693989071039,
+      "grad_norm": 0.8041996749546387,
+      "learning_rate": 9.242657789576272e-05,
+      "loss": 12.3576,
+      "step": 9236
+    },
+    {
+      "epoch": 1.615344262295082,
+      "grad_norm": 0.8699013117600289,
+      "learning_rate": 9.240775135784199e-05,
+      "loss": 12.5921,
+      "step": 9237
+    },
+    {
+      "epoch": 1.61551912568306,
+      "grad_norm": 0.8093413324202431,
+      "learning_rate": 9.238892509057601e-05,
+      "loss": 12.2571,
+      "step": 9238
+    },
+    {
+      "epoch": 1.6156939890710382,
+      "grad_norm": 0.8479053571445598,
+      "learning_rate": 9.237009909463598e-05,
+      "loss": 12.5604,
+      "step": 9239
+    },
+    {
+      "epoch": 1.6158688524590163,
+      "grad_norm": 0.8757474093506225,
+      "learning_rate": 9.235127337069298e-05,
+      "loss": 12.5071,
+      "step": 9240
+    },
+    {
+      "epoch": 1.6160437158469945,
+      "grad_norm": 0.8128772150928795,
+      "learning_rate": 9.23324479194182e-05,
+      "loss": 12.2634,
+      "step": 9241
+    },
+    {
+      "epoch": 1.6162185792349728,
+      "grad_norm": 0.7972758719458545,
+      "learning_rate": 9.231362274148266e-05,
+      "loss": 12.6128,
+      "step": 9242
+    },
+    {
+      "epoch": 1.6163934426229507,
+      "grad_norm": 0.8578859767857246,
+      "learning_rate": 9.229479783755748e-05,
+      "loss": 12.2563,
+      "step": 9243
+    },
+    {
+      "epoch": 1.616568306010929,
+      "grad_norm": 0.8813210026620237,
+      "learning_rate": 9.227597320831375e-05,
+      "loss": 12.32,
+      "step": 9244
+    },
+    {
+      "epoch": 1.616743169398907,
+      "grad_norm": 0.9036173921427253,
+      "learning_rate": 9.225714885442257e-05,
+      "loss": 12.0721,
+      "step": 9245
+    },
+    {
+      "epoch": 1.6169180327868853,
+      "grad_norm": 0.8901810329627683,
+      "learning_rate": 9.223832477655497e-05,
+      "loss": 12.1709,
+      "step": 9246
+    },
+    {
+      "epoch": 1.6170928961748634,
+      "grad_norm": 0.8507356957236244,
+      "learning_rate": 9.221950097538203e-05,
+      "loss": 12.4966,
+      "step": 9247
+    },
+    {
+      "epoch": 1.6172677595628415,
+      "grad_norm": 0.9371343390432378,
+      "learning_rate": 9.220067745157482e-05,
+      "loss": 12.5592,
+      "step": 9248
+    },
+    {
+      "epoch": 1.6174426229508196,
+      "grad_norm": 0.9006083721532335,
+      "learning_rate": 9.218185420580432e-05,
+      "loss": 12.4015,
+      "step": 9249
+    },
+    {
+      "epoch": 1.6176174863387978,
+      "grad_norm": 1.0726264584381047,
+      "learning_rate": 9.216303123874157e-05,
+      "loss": 12.7023,
+      "step": 9250
+    },
+    {
+      "epoch": 1.617792349726776,
+      "grad_norm": 0.8986597284755948,
+      "learning_rate": 9.214420855105761e-05,
+      "loss": 12.5428,
+      "step": 9251
+    },
+    {
+      "epoch": 1.617967213114754,
+      "grad_norm": 0.9143637018594418,
+      "learning_rate": 9.212538614342346e-05,
+      "loss": 12.4761,
+      "step": 9252
+    },
+    {
+      "epoch": 1.6181420765027323,
+      "grad_norm": 0.8754918234965825,
+      "learning_rate": 9.210656401651008e-05,
+      "loss": 12.4962,
+      "step": 9253
+    },
+    {
+      "epoch": 1.6183169398907102,
+      "grad_norm": 0.9347122127224794,
+      "learning_rate": 9.208774217098844e-05,
+      "loss": 12.1458,
+      "step": 9254
+    },
+    {
+      "epoch": 1.6184918032786886,
+      "grad_norm": 0.9286915685827619,
+      "learning_rate": 9.206892060752957e-05,
+      "loss": 12.4038,
+      "step": 9255
+    },
+    {
+      "epoch": 1.6186666666666667,
+      "grad_norm": 0.9167107461889161,
+      "learning_rate": 9.205009932680442e-05,
+      "loss": 12.5204,
+      "step": 9256
+    },
+    {
+      "epoch": 1.6188415300546448,
+      "grad_norm": 1.1442516058543728,
+      "learning_rate": 9.203127832948393e-05,
+      "loss": 12.4912,
+      "step": 9257
+    },
+    {
+      "epoch": 1.619016393442623,
+      "grad_norm": 1.0006994824160118,
+      "learning_rate": 9.201245761623905e-05,
+      "loss": 12.5359,
+      "step": 9258
+    },
+    {
+      "epoch": 1.619191256830601,
+      "grad_norm": 1.0066080467056222,
+      "learning_rate": 9.199363718774072e-05,
+      "loss": 12.06,
+      "step": 9259
+    },
+    {
+      "epoch": 1.6193661202185794,
+      "grad_norm": 0.9777766222425417,
+      "learning_rate": 9.197481704465993e-05,
+      "loss": 12.3325,
+      "step": 9260
+    },
+    {
+      "epoch": 1.6195409836065573,
+      "grad_norm": 0.9835503986500688,
+      "learning_rate": 9.195599718766748e-05,
+      "loss": 12.7063,
+      "step": 9261
+    },
+    {
+      "epoch": 1.6197158469945356,
+      "grad_norm": 0.8995794296055497,
+      "learning_rate": 9.193717761743433e-05,
+      "loss": 12.4535,
+      "step": 9262
+    },
+    {
+      "epoch": 1.6198907103825135,
+      "grad_norm": 0.9735370975552687,
+      "learning_rate": 9.19183583346314e-05,
+      "loss": 12.4655,
+      "step": 9263
+    },
+    {
+      "epoch": 1.6200655737704919,
+      "grad_norm": 0.907076392982167,
+      "learning_rate": 9.189953933992954e-05,
+      "loss": 12.4236,
+      "step": 9264
+    },
+    {
+      "epoch": 1.62024043715847,
+      "grad_norm": 0.8431806119287445,
+      "learning_rate": 9.188072063399963e-05,
+      "loss": 12.5236,
+      "step": 9265
+    },
+    {
+      "epoch": 1.6204153005464481,
+      "grad_norm": 0.8751977733346993,
+      "learning_rate": 9.186190221751254e-05,
+      "loss": 12.4924,
+      "step": 9266
+    },
+    {
+      "epoch": 1.6205901639344262,
+      "grad_norm": 0.917558206700827,
+      "learning_rate": 9.184308409113918e-05,
+      "loss": 12.3722,
+      "step": 9267
+    },
+    {
+      "epoch": 1.6207650273224044,
+      "grad_norm": 0.7980283101726278,
+      "learning_rate": 9.182426625555029e-05,
+      "loss": 12.3016,
+      "step": 9268
+    },
+    {
+      "epoch": 1.6209398907103825,
+      "grad_norm": 0.851998467534572,
+      "learning_rate": 9.180544871141677e-05,
+      "loss": 12.4869,
+      "step": 9269
+    },
+    {
+      "epoch": 1.6211147540983606,
+      "grad_norm": 0.9257329882335664,
+      "learning_rate": 9.178663145940942e-05,
+      "loss": 12.4625,
+      "step": 9270
+    },
+    {
+      "epoch": 1.621289617486339,
+      "grad_norm": 0.9195497541807373,
+      "learning_rate": 9.17678145001991e-05,
+      "loss": 12.1433,
+      "step": 9271
+    },
+    {
+      "epoch": 1.6214644808743168,
+      "grad_norm": 0.9181198045849859,
+      "learning_rate": 9.174899783445654e-05,
+      "loss": 12.291,
+      "step": 9272
+    },
+    {
+      "epoch": 1.6216393442622952,
+      "grad_norm": 0.8411052418580699,
+      "learning_rate": 9.173018146285256e-05,
+      "loss": 12.5278,
+      "step": 9273
+    },
+    {
+      "epoch": 1.621814207650273,
+      "grad_norm": 0.8760194596008416,
+      "learning_rate": 9.171136538605796e-05,
+      "loss": 12.4293,
+      "step": 9274
+    },
+    {
+      "epoch": 1.6219890710382514,
+      "grad_norm": 0.9083510317717781,
+      "learning_rate": 9.16925496047435e-05,
+      "loss": 12.4847,
+      "step": 9275
+    },
+    {
+      "epoch": 1.6221639344262295,
+      "grad_norm": 1.067125737551947,
+      "learning_rate": 9.167373411957993e-05,
+      "loss": 12.5335,
+      "step": 9276
+    },
+    {
+      "epoch": 1.6223387978142076,
+      "grad_norm": 0.8614873981292401,
+      "learning_rate": 9.165491893123802e-05,
+      "loss": 12.4044,
+      "step": 9277
+    },
+    {
+      "epoch": 1.6225136612021858,
+      "grad_norm": 1.1480331192012303,
+      "learning_rate": 9.163610404038849e-05,
+      "loss": 12.5425,
+      "step": 9278
+    },
+    {
+      "epoch": 1.6226885245901639,
+      "grad_norm": 0.8727742682448972,
+      "learning_rate": 9.161728944770214e-05,
+      "loss": 12.1556,
+      "step": 9279
+    },
+    {
+      "epoch": 1.6228633879781422,
+      "grad_norm": 1.0067996805456954,
+      "learning_rate": 9.159847515384956e-05,
+      "loss": 12.2501,
+      "step": 9280
+    },
+    {
+      "epoch": 1.6230382513661201,
+      "grad_norm": 1.0632686532511422,
+      "learning_rate": 9.157966115950154e-05,
+      "loss": 12.4617,
+      "step": 9281
+    },
+    {
+      "epoch": 1.6232131147540985,
+      "grad_norm": 0.8535847301517938,
+      "learning_rate": 9.156084746532877e-05,
+      "loss": 12.5103,
+      "step": 9282
+    },
+    {
+      "epoch": 1.6233879781420764,
+      "grad_norm": 0.8827735530563324,
+      "learning_rate": 9.154203407200192e-05,
+      "loss": 12.5549,
+      "step": 9283
+    },
+    {
+      "epoch": 1.6235628415300547,
+      "grad_norm": 0.9000624710997517,
+      "learning_rate": 9.152322098019166e-05,
+      "loss": 12.43,
+      "step": 9284
+    },
+    {
+      "epoch": 1.6237377049180328,
+      "grad_norm": 0.8416121911713254,
+      "learning_rate": 9.150440819056866e-05,
+      "loss": 12.3876,
+      "step": 9285
+    },
+    {
+      "epoch": 1.623912568306011,
+      "grad_norm": 0.8820744658851195,
+      "learning_rate": 9.148559570380365e-05,
+      "loss": 12.3645,
+      "step": 9286
+    },
+    {
+      "epoch": 1.624087431693989,
+      "grad_norm": 0.9119608610700415,
+      "learning_rate": 9.146678352056715e-05,
+      "loss": 12.1696,
+      "step": 9287
+    },
+    {
+      "epoch": 1.6242622950819672,
+      "grad_norm": 0.947251374202448,
+      "learning_rate": 9.144797164152986e-05,
+      "loss": 12.4045,
+      "step": 9288
+    },
+    {
+      "epoch": 1.6244371584699455,
+      "grad_norm": 0.8773814468822657,
+      "learning_rate": 9.142916006736237e-05,
+      "loss": 12.7251,
+      "step": 9289
+    },
+    {
+      "epoch": 1.6246120218579234,
+      "grad_norm": 0.7983314570208858,
+      "learning_rate": 9.141034879873533e-05,
+      "loss": 12.3392,
+      "step": 9290
+    },
+    {
+      "epoch": 1.6247868852459018,
+      "grad_norm": 0.8991308858611409,
+      "learning_rate": 9.139153783631932e-05,
+      "loss": 12.5514,
+      "step": 9291
+    },
+    {
+      "epoch": 1.6249617486338797,
+      "grad_norm": 0.8991262021689161,
+      "learning_rate": 9.137272718078493e-05,
+      "loss": 12.4748,
+      "step": 9292
+    },
+    {
+      "epoch": 1.625136612021858,
+      "grad_norm": 0.8693422680770767,
+      "learning_rate": 9.135391683280272e-05,
+      "loss": 12.4904,
+      "step": 9293
+    },
+    {
+      "epoch": 1.6253114754098361,
+      "grad_norm": 0.8558128801165167,
+      "learning_rate": 9.133510679304329e-05,
+      "loss": 12.2518,
+      "step": 9294
+    },
+    {
+      "epoch": 1.6254863387978142,
+      "grad_norm": 0.7985685660971039,
+      "learning_rate": 9.131629706217718e-05,
+      "loss": 12.5775,
+      "step": 9295
+    },
+    {
+      "epoch": 1.6256612021857924,
+      "grad_norm": 0.9418320967635477,
+      "learning_rate": 9.12974876408749e-05,
+      "loss": 12.5272,
+      "step": 9296
+    },
+    {
+      "epoch": 1.6258360655737705,
+      "grad_norm": 0.8414045049865506,
+      "learning_rate": 9.127867852980708e-05,
+      "loss": 12.3978,
+      "step": 9297
+    },
+    {
+      "epoch": 1.6260109289617486,
+      "grad_norm": 0.8381442558909467,
+      "learning_rate": 9.125986972964413e-05,
+      "loss": 12.2811,
+      "step": 9298
+    },
+    {
+      "epoch": 1.6261857923497267,
+      "grad_norm": 0.9016316953030443,
+      "learning_rate": 9.124106124105662e-05,
+      "loss": 12.5262,
+      "step": 9299
+    },
+    {
+      "epoch": 1.626360655737705,
+      "grad_norm": 0.8567453260880977,
+      "learning_rate": 9.122225306471505e-05,
+      "loss": 12.3728,
+      "step": 9300
+    },
+    {
+      "epoch": 1.626535519125683,
+      "grad_norm": 0.8593390170910412,
+      "learning_rate": 9.120344520128991e-05,
+      "loss": 12.4617,
+      "step": 9301
+    },
+    {
+      "epoch": 1.6267103825136613,
+      "grad_norm": 0.8051878355506716,
+      "learning_rate": 9.118463765145166e-05,
+      "loss": 12.2269,
+      "step": 9302
+    },
+    {
+      "epoch": 1.6268852459016392,
+      "grad_norm": 0.841161106491591,
+      "learning_rate": 9.116583041587076e-05,
+      "loss": 12.4111,
+      "step": 9303
+    },
+    {
+      "epoch": 1.6270601092896175,
+      "grad_norm": 0.8181935532873947,
+      "learning_rate": 9.11470234952177e-05,
+      "loss": 12.4,
+      "step": 9304
+    },
+    {
+      "epoch": 1.6272349726775956,
+      "grad_norm": 0.8575754583763077,
+      "learning_rate": 9.112821689016296e-05,
+      "loss": 12.5232,
+      "step": 9305
+    },
+    {
+      "epoch": 1.6274098360655738,
+      "grad_norm": 0.7874656821807219,
+      "learning_rate": 9.110941060137686e-05,
+      "loss": 12.1618,
+      "step": 9306
+    },
+    {
+      "epoch": 1.6275846994535519,
+      "grad_norm": 0.820373573689476,
+      "learning_rate": 9.109060462952988e-05,
+      "loss": 12.1662,
+      "step": 9307
+    },
+    {
+      "epoch": 1.62775956284153,
+      "grad_norm": 0.9246893921338271,
+      "learning_rate": 9.107179897529246e-05,
+      "loss": 12.4209,
+      "step": 9308
+    },
+    {
+      "epoch": 1.6279344262295083,
+      "grad_norm": 0.8040426169650693,
+      "learning_rate": 9.105299363933498e-05,
+      "loss": 12.357,
+      "step": 9309
+    },
+    {
+      "epoch": 1.6281092896174862,
+      "grad_norm": 0.8605333865526013,
+      "learning_rate": 9.103418862232781e-05,
+      "loss": 12.4433,
+      "step": 9310
+    },
+    {
+      "epoch": 1.6282841530054646,
+      "grad_norm": 0.7928500624788642,
+      "learning_rate": 9.101538392494134e-05,
+      "loss": 12.3202,
+      "step": 9311
+    },
+    {
+      "epoch": 1.6284590163934425,
+      "grad_norm": 0.9013991481662553,
+      "learning_rate": 9.099657954784596e-05,
+      "loss": 12.4815,
+      "step": 9312
+    },
+    {
+      "epoch": 1.6286338797814208,
+      "grad_norm": 0.8321721619629551,
+      "learning_rate": 9.097777549171198e-05,
+      "loss": 12.2112,
+      "step": 9313
+    },
+    {
+      "epoch": 1.628808743169399,
+      "grad_norm": 0.8030026797763067,
+      "learning_rate": 9.095897175720975e-05,
+      "loss": 11.9916,
+      "step": 9314
+    },
+    {
+      "epoch": 1.628983606557377,
+      "grad_norm": 0.8860835739735432,
+      "learning_rate": 9.094016834500961e-05,
+      "loss": 12.559,
+      "step": 9315
+    },
+    {
+      "epoch": 1.6291584699453552,
+      "grad_norm": 0.9017675135225521,
+      "learning_rate": 9.092136525578194e-05,
+      "loss": 12.4015,
+      "step": 9316
+    },
+    {
+      "epoch": 1.6293333333333333,
+      "grad_norm": 0.8783731669976412,
+      "learning_rate": 9.090256249019694e-05,
+      "loss": 12.5179,
+      "step": 9317
+    },
+    {
+      "epoch": 1.6295081967213116,
+      "grad_norm": 0.8375161283911363,
+      "learning_rate": 9.088376004892496e-05,
+      "loss": 12.3926,
+      "step": 9318
+    },
+    {
+      "epoch": 1.6296830601092895,
+      "grad_norm": 0.8540116786566325,
+      "learning_rate": 9.086495793263628e-05,
+      "loss": 12.0698,
+      "step": 9319
+    },
+    {
+      "epoch": 1.6298579234972679,
+      "grad_norm": 0.9278450462777708,
+      "learning_rate": 9.08461561420012e-05,
+      "loss": 12.4377,
+      "step": 9320
+    },
+    {
+      "epoch": 1.6300327868852458,
+      "grad_norm": 0.8326219514612957,
+      "learning_rate": 9.082735467768994e-05,
+      "loss": 12.4762,
+      "step": 9321
+    },
+    {
+      "epoch": 1.6302076502732241,
+      "grad_norm": 0.872470611038733,
+      "learning_rate": 9.080855354037276e-05,
+      "loss": 12.3379,
+      "step": 9322
+    },
+    {
+      "epoch": 1.6303825136612022,
+      "grad_norm": 0.8896116962254348,
+      "learning_rate": 9.078975273071992e-05,
+      "loss": 12.3721,
+      "step": 9323
+    },
+    {
+      "epoch": 1.6305573770491804,
+      "grad_norm": 0.922358598326129,
+      "learning_rate": 9.077095224940167e-05,
+      "loss": 12.0624,
+      "step": 9324
+    },
+    {
+      "epoch": 1.6307322404371585,
+      "grad_norm": 0.9198560845933128,
+      "learning_rate": 9.075215209708815e-05,
+      "loss": 12.5745,
+      "step": 9325
+    },
+    {
+      "epoch": 1.6309071038251366,
+      "grad_norm": 0.8314067058939317,
+      "learning_rate": 9.07333522744496e-05,
+      "loss": 12.3322,
+      "step": 9326
+    },
+    {
+      "epoch": 1.6310819672131147,
+      "grad_norm": 0.8614635691499651,
+      "learning_rate": 9.071455278215624e-05,
+      "loss": 12.5032,
+      "step": 9327
+    },
+    {
+      "epoch": 1.6312568306010928,
+      "grad_norm": 0.8608186904537055,
+      "learning_rate": 9.06957536208782e-05,
+      "loss": 12.4742,
+      "step": 9328
+    },
+    {
+      "epoch": 1.6314316939890712,
+      "grad_norm": 0.8736512455377207,
+      "learning_rate": 9.067695479128568e-05,
+      "loss": 12.5316,
+      "step": 9329
+    },
+    {
+      "epoch": 1.631606557377049,
+      "grad_norm": 0.8594576111313248,
+      "learning_rate": 9.065815629404881e-05,
+      "loss": 12.4503,
+      "step": 9330
+    },
+    {
+      "epoch": 1.6317814207650274,
+      "grad_norm": 0.8568148197493665,
+      "learning_rate": 9.063935812983781e-05,
+      "loss": 12.5382,
+      "step": 9331
+    },
+    {
+      "epoch": 1.6319562841530053,
+      "grad_norm": 0.8499541451088,
+      "learning_rate": 9.062056029932271e-05,
+      "loss": 12.2829,
+      "step": 9332
+    },
+    {
+      "epoch": 1.6321311475409837,
+      "grad_norm": 0.889159562751674,
+      "learning_rate": 9.060176280317368e-05,
+      "loss": 12.1869,
+      "step": 9333
+    },
+    {
+      "epoch": 1.6323060109289618,
+      "grad_norm": 0.8456076567844809,
+      "learning_rate": 9.058296564206081e-05,
+      "loss": 12.6312,
+      "step": 9334
+    },
+    {
+      "epoch": 1.63248087431694,
+      "grad_norm": 0.828113652321329,
+      "learning_rate": 9.056416881665427e-05,
+      "loss": 12.4942,
+      "step": 9335
+    },
+    {
+      "epoch": 1.632655737704918,
+      "grad_norm": 0.8833199087648398,
+      "learning_rate": 9.054537232762405e-05,
+      "loss": 12.0296,
+      "step": 9336
+    },
+    {
+      "epoch": 1.6328306010928961,
+      "grad_norm": 0.8884258671439023,
+      "learning_rate": 9.052657617564025e-05,
+      "loss": 12.5718,
+      "step": 9337
+    },
+    {
+      "epoch": 1.6330054644808745,
+      "grad_norm": 0.9196662771229902,
+      "learning_rate": 9.050778036137293e-05,
+      "loss": 12.434,
+      "step": 9338
+    },
+    {
+      "epoch": 1.6331803278688524,
+      "grad_norm": 0.9574630569288348,
+      "learning_rate": 9.048898488549219e-05,
+      "loss": 12.4494,
+      "step": 9339
+    },
+    {
+      "epoch": 1.6333551912568307,
+      "grad_norm": 0.9423106375237535,
+      "learning_rate": 9.0470189748668e-05,
+      "loss": 12.6897,
+      "step": 9340
+    },
+    {
+      "epoch": 1.6335300546448086,
+      "grad_norm": 0.9514055450535158,
+      "learning_rate": 9.04513949515704e-05,
+      "loss": 12.4104,
+      "step": 9341
+    },
+    {
+      "epoch": 1.633704918032787,
+      "grad_norm": 0.8522675118804346,
+      "learning_rate": 9.043260049486946e-05,
+      "loss": 12.2878,
+      "step": 9342
+    },
+    {
+      "epoch": 1.633879781420765,
+      "grad_norm": 0.8912142244406059,
+      "learning_rate": 9.041380637923512e-05,
+      "loss": 12.4942,
+      "step": 9343
+    },
+    {
+      "epoch": 1.6340546448087432,
+      "grad_norm": 0.8697147462985901,
+      "learning_rate": 9.039501260533735e-05,
+      "loss": 12.3453,
+      "step": 9344
+    },
+    {
+      "epoch": 1.6342295081967213,
+      "grad_norm": 0.8796211386267054,
+      "learning_rate": 9.037621917384616e-05,
+      "loss": 12.1113,
+      "step": 9345
+    },
+    {
+      "epoch": 1.6344043715846994,
+      "grad_norm": 0.8112428092157553,
+      "learning_rate": 9.035742608543153e-05,
+      "loss": 12.3963,
+      "step": 9346
+    },
+    {
+      "epoch": 1.6345792349726775,
+      "grad_norm": 0.8380877170405395,
+      "learning_rate": 9.033863334076341e-05,
+      "loss": 12.5511,
+      "step": 9347
+    },
+    {
+      "epoch": 1.6347540983606557,
+      "grad_norm": 0.8817844850186087,
+      "learning_rate": 9.031984094051168e-05,
+      "loss": 12.1416,
+      "step": 9348
+    },
+    {
+      "epoch": 1.634928961748634,
+      "grad_norm": 0.8777133509607908,
+      "learning_rate": 9.030104888534633e-05,
+      "loss": 12.3754,
+      "step": 9349
+    },
+    {
+      "epoch": 1.635103825136612,
+      "grad_norm": 0.8087735072590904,
+      "learning_rate": 9.02822571759373e-05,
+      "loss": 12.4583,
+      "step": 9350
+    },
+    {
+      "epoch": 1.6352786885245902,
+      "grad_norm": 0.9155965581664437,
+      "learning_rate": 9.02634658129544e-05,
+      "loss": 12.3589,
+      "step": 9351
+    },
+    {
+      "epoch": 1.6354535519125684,
+      "grad_norm": 0.9060654364088596,
+      "learning_rate": 9.024467479706756e-05,
+      "loss": 12.1691,
+      "step": 9352
+    },
+    {
+      "epoch": 1.6356284153005465,
+      "grad_norm": 0.8883116924273173,
+      "learning_rate": 9.022588412894667e-05,
+      "loss": 12.5802,
+      "step": 9353
+    },
+    {
+      "epoch": 1.6358032786885246,
+      "grad_norm": 1.0056847148188894,
+      "learning_rate": 9.020709380926163e-05,
+      "loss": 12.3042,
+      "step": 9354
+    },
+    {
+      "epoch": 1.6359781420765027,
+      "grad_norm": 1.02705257606901,
+      "learning_rate": 9.018830383868221e-05,
+      "loss": 11.9655,
+      "step": 9355
+    },
+    {
+      "epoch": 1.6361530054644808,
+      "grad_norm": 0.9035901488616217,
+      "learning_rate": 9.01695142178783e-05,
+      "loss": 12.6761,
+      "step": 9356
+    },
+    {
+      "epoch": 1.636327868852459,
+      "grad_norm": 1.1167878058551135,
+      "learning_rate": 9.015072494751978e-05,
+      "loss": 12.3402,
+      "step": 9357
+    },
+    {
+      "epoch": 1.6365027322404373,
+      "grad_norm": 0.9405479471751221,
+      "learning_rate": 9.013193602827634e-05,
+      "loss": 12.6078,
+      "step": 9358
+    },
+    {
+      "epoch": 1.6366775956284152,
+      "grad_norm": 0.8654043353552054,
+      "learning_rate": 9.011314746081787e-05,
+      "loss": 12.3457,
+      "step": 9359
+    },
+    {
+      "epoch": 1.6368524590163935,
+      "grad_norm": 0.9903909920078074,
+      "learning_rate": 9.009435924581414e-05,
+      "loss": 12.397,
+      "step": 9360
+    },
+    {
+      "epoch": 1.6370273224043714,
+      "grad_norm": 0.8753583799369794,
+      "learning_rate": 9.007557138393497e-05,
+      "loss": 12.3232,
+      "step": 9361
+    },
+    {
+      "epoch": 1.6372021857923498,
+      "grad_norm": 1.0578065553257043,
+      "learning_rate": 9.005678387585003e-05,
+      "loss": 12.5107,
+      "step": 9362
+    },
+    {
+      "epoch": 1.637377049180328,
+      "grad_norm": 0.8105861751534481,
+      "learning_rate": 9.003799672222914e-05,
+      "loss": 12.1338,
+      "step": 9363
+    },
+    {
+      "epoch": 1.637551912568306,
+      "grad_norm": 0.9271873441176671,
+      "learning_rate": 9.001920992374204e-05,
+      "loss": 12.2527,
+      "step": 9364
+    },
+    {
+      "epoch": 1.6377267759562841,
+      "grad_norm": 0.9016373078337595,
+      "learning_rate": 9.000042348105844e-05,
+      "loss": 12.5657,
+      "step": 9365
+    },
+    {
+      "epoch": 1.6379016393442623,
+      "grad_norm": 0.8542389490057284,
+      "learning_rate": 8.998163739484805e-05,
+      "loss": 12.3742,
+      "step": 9366
+    },
+    {
+      "epoch": 1.6380765027322406,
+      "grad_norm": 0.8787631188562384,
+      "learning_rate": 8.996285166578057e-05,
+      "loss": 12.2757,
+      "step": 9367
+    },
+    {
+      "epoch": 1.6382513661202185,
+      "grad_norm": 0.8540554103209741,
+      "learning_rate": 8.994406629452571e-05,
+      "loss": 12.7014,
+      "step": 9368
+    },
+    {
+      "epoch": 1.6384262295081968,
+      "grad_norm": 0.8377726998662034,
+      "learning_rate": 8.99252812817532e-05,
+      "loss": 12.4714,
+      "step": 9369
+    },
+    {
+      "epoch": 1.6386010928961747,
+      "grad_norm": 0.8815084875784641,
+      "learning_rate": 8.990649662813257e-05,
+      "loss": 12.0364,
+      "step": 9370
+    },
+    {
+      "epoch": 1.638775956284153,
+      "grad_norm": 0.8523017064531755,
+      "learning_rate": 8.988771233433356e-05,
+      "loss": 12.327,
+      "step": 9371
+    },
+    {
+      "epoch": 1.6389508196721312,
+      "grad_norm": 0.9056855925916698,
+      "learning_rate": 8.98689284010258e-05,
+      "loss": 12.3716,
+      "step": 9372
+    },
+    {
+      "epoch": 1.6391256830601093,
+      "grad_norm": 0.841408180136612,
+      "learning_rate": 8.985014482887889e-05,
+      "loss": 12.2794,
+      "step": 9373
+    },
+    {
+      "epoch": 1.6393005464480874,
+      "grad_norm": 0.8602814548610891,
+      "learning_rate": 8.983136161856246e-05,
+      "loss": 12.5011,
+      "step": 9374
+    },
+    {
+      "epoch": 1.6394754098360655,
+      "grad_norm": 0.8329085287604946,
+      "learning_rate": 8.98125787707461e-05,
+      "loss": 12.2982,
+      "step": 9375
+    },
+    {
+      "epoch": 1.6396502732240437,
+      "grad_norm": 0.9034339511015033,
+      "learning_rate": 8.979379628609945e-05,
+      "loss": 12.4103,
+      "step": 9376
+    },
+    {
+      "epoch": 1.6398251366120218,
+      "grad_norm": 0.8507166331427156,
+      "learning_rate": 8.9775014165292e-05,
+      "loss": 12.5058,
+      "step": 9377
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.8060027579147369,
+      "learning_rate": 8.975623240899335e-05,
+      "loss": 12.2238,
+      "step": 9378
+    },
+    {
+      "epoch": 1.640174863387978,
+      "grad_norm": 0.8241265311824252,
+      "learning_rate": 8.973745101787304e-05,
+      "loss": 12.4076,
+      "step": 9379
+    },
+    {
+      "epoch": 1.6403497267759564,
+      "grad_norm": 0.8982509455914411,
+      "learning_rate": 8.971866999260066e-05,
+      "loss": 12.2888,
+      "step": 9380
+    },
+    {
+      "epoch": 1.6405245901639343,
+      "grad_norm": 0.8911376772658594,
+      "learning_rate": 8.969988933384563e-05,
+      "loss": 12.3927,
+      "step": 9381
+    },
+    {
+      "epoch": 1.6406994535519126,
+      "grad_norm": 0.8487348023535088,
+      "learning_rate": 8.968110904227751e-05,
+      "loss": 12.3254,
+      "step": 9382
+    },
+    {
+      "epoch": 1.6408743169398907,
+      "grad_norm": 0.8478984654356327,
+      "learning_rate": 8.96623291185658e-05,
+      "loss": 12.4497,
+      "step": 9383
+    },
+    {
+      "epoch": 1.6410491803278688,
+      "grad_norm": 0.9174201711612727,
+      "learning_rate": 8.964354956338e-05,
+      "loss": 12.5178,
+      "step": 9384
+    },
+    {
+      "epoch": 1.641224043715847,
+      "grad_norm": 0.7887127381570163,
+      "learning_rate": 8.962477037738955e-05,
+      "loss": 12.1173,
+      "step": 9385
+    },
+    {
+      "epoch": 1.641398907103825,
+      "grad_norm": 0.8021653512941338,
+      "learning_rate": 8.960599156126389e-05,
+      "loss": 12.3559,
+      "step": 9386
+    },
+    {
+      "epoch": 1.6415737704918034,
+      "grad_norm": 0.8304369996627394,
+      "learning_rate": 8.95872131156725e-05,
+      "loss": 12.432,
+      "step": 9387
+    },
+    {
+      "epoch": 1.6417486338797813,
+      "grad_norm": 0.840633665805651,
+      "learning_rate": 8.956843504128484e-05,
+      "loss": 12.5649,
+      "step": 9388
+    },
+    {
+      "epoch": 1.6419234972677597,
+      "grad_norm": 0.7886112442964481,
+      "learning_rate": 8.954965733877023e-05,
+      "loss": 12.2935,
+      "step": 9389
+    },
+    {
+      "epoch": 1.6420983606557376,
+      "grad_norm": 0.8134702767985551,
+      "learning_rate": 8.953088000879815e-05,
+      "loss": 12.3253,
+      "step": 9390
+    },
+    {
+      "epoch": 1.642273224043716,
+      "grad_norm": 0.9254792631382897,
+      "learning_rate": 8.951210305203797e-05,
+      "loss": 12.5492,
+      "step": 9391
+    },
+    {
+      "epoch": 1.642448087431694,
+      "grad_norm": 0.8982664272576596,
+      "learning_rate": 8.949332646915903e-05,
+      "loss": 12.5238,
+      "step": 9392
+    },
+    {
+      "epoch": 1.6426229508196721,
+      "grad_norm": 0.8175368005449031,
+      "learning_rate": 8.947455026083073e-05,
+      "loss": 12.3197,
+      "step": 9393
+    },
+    {
+      "epoch": 1.6427978142076503,
+      "grad_norm": 0.8954463704214838,
+      "learning_rate": 8.945577442772243e-05,
+      "loss": 12.4464,
+      "step": 9394
+    },
+    {
+      "epoch": 1.6429726775956284,
+      "grad_norm": 0.8128634275427028,
+      "learning_rate": 8.943699897050349e-05,
+      "loss": 12.4204,
+      "step": 9395
+    },
+    {
+      "epoch": 1.6431475409836067,
+      "grad_norm": 0.8595991425049263,
+      "learning_rate": 8.941822388984315e-05,
+      "loss": 12.4206,
+      "step": 9396
+    },
+    {
+      "epoch": 1.6433224043715846,
+      "grad_norm": 0.8553751113218426,
+      "learning_rate": 8.939944918641077e-05,
+      "loss": 12.5768,
+      "step": 9397
+    },
+    {
+      "epoch": 1.643497267759563,
+      "grad_norm": 0.9330729367666841,
+      "learning_rate": 8.938067486087564e-05,
+      "loss": 12.4341,
+      "step": 9398
+    },
+    {
+      "epoch": 1.6436721311475408,
+      "grad_norm": 0.875672728769871,
+      "learning_rate": 8.936190091390705e-05,
+      "loss": 12.4358,
+      "step": 9399
+    },
+    {
+      "epoch": 1.6438469945355192,
+      "grad_norm": 0.7742307282076821,
+      "learning_rate": 8.934312734617424e-05,
+      "loss": 12.6027,
+      "step": 9400
+    },
+    {
+      "epoch": 1.6440218579234973,
+      "grad_norm": 0.8672472238526467,
+      "learning_rate": 8.932435415834652e-05,
+      "loss": 12.1809,
+      "step": 9401
+    },
+    {
+      "epoch": 1.6441967213114754,
+      "grad_norm": 0.8918870301551834,
+      "learning_rate": 8.930558135109308e-05,
+      "loss": 12.3889,
+      "step": 9402
+    },
+    {
+      "epoch": 1.6443715846994535,
+      "grad_norm": 0.8630819154368348,
+      "learning_rate": 8.928680892508321e-05,
+      "loss": 12.3023,
+      "step": 9403
+    },
+    {
+      "epoch": 1.6445464480874317,
+      "grad_norm": 0.7973643818436305,
+      "learning_rate": 8.926803688098604e-05,
+      "loss": 12.3931,
+      "step": 9404
+    },
+    {
+      "epoch": 1.6447213114754098,
+      "grad_norm": 0.8184326188050267,
+      "learning_rate": 8.924926521947083e-05,
+      "loss": 12.4479,
+      "step": 9405
+    },
+    {
+      "epoch": 1.644896174863388,
+      "grad_norm": 0.794109975694462,
+      "learning_rate": 8.92304939412068e-05,
+      "loss": 12.3564,
+      "step": 9406
+    },
+    {
+      "epoch": 1.6450710382513662,
+      "grad_norm": 0.9466937647724862,
+      "learning_rate": 8.921172304686302e-05,
+      "loss": 12.4372,
+      "step": 9407
+    },
+    {
+      "epoch": 1.6452459016393441,
+      "grad_norm": 0.8160859367390103,
+      "learning_rate": 8.919295253710871e-05,
+      "loss": 12.4242,
+      "step": 9408
+    },
+    {
+      "epoch": 1.6454207650273225,
+      "grad_norm": 0.9037346457279796,
+      "learning_rate": 8.917418241261303e-05,
+      "loss": 12.3138,
+      "step": 9409
+    },
+    {
+      "epoch": 1.6455956284153004,
+      "grad_norm": 0.8318799214369152,
+      "learning_rate": 8.91554126740451e-05,
+      "loss": 12.5076,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6457704918032787,
+      "grad_norm": 0.8437133299663867,
+      "learning_rate": 8.913664332207403e-05,
+      "loss": 12.3462,
+      "step": 9411
+    },
+    {
+      "epoch": 1.6459453551912568,
+      "grad_norm": 0.8483700820951627,
+      "learning_rate": 8.911787435736893e-05,
+      "loss": 12.2226,
+      "step": 9412
+    },
+    {
+      "epoch": 1.646120218579235,
+      "grad_norm": 0.9729652813325591,
+      "learning_rate": 8.90991057805989e-05,
+      "loss": 12.3816,
+      "step": 9413
+    },
+    {
+      "epoch": 1.646295081967213,
+      "grad_norm": 0.8614746855460623,
+      "learning_rate": 8.908033759243305e-05,
+      "loss": 12.4613,
+      "step": 9414
+    },
+    {
+      "epoch": 1.6464699453551912,
+      "grad_norm": 0.9106725189751098,
+      "learning_rate": 8.906156979354036e-05,
+      "loss": 12.4413,
+      "step": 9415
+    },
+    {
+      "epoch": 1.6466448087431695,
+      "grad_norm": 0.8893266079063001,
+      "learning_rate": 8.904280238458993e-05,
+      "loss": 12.5912,
+      "step": 9416
+    },
+    {
+      "epoch": 1.6468196721311474,
+      "grad_norm": 0.8591562597253415,
+      "learning_rate": 8.902403536625079e-05,
+      "loss": 12.3778,
+      "step": 9417
+    },
+    {
+      "epoch": 1.6469945355191258,
+      "grad_norm": 0.7928378557535352,
+      "learning_rate": 8.9005268739192e-05,
+      "loss": 12.5516,
+      "step": 9418
+    },
+    {
+      "epoch": 1.6471693989071037,
+      "grad_norm": 0.9067873518902194,
+      "learning_rate": 8.89865025040825e-05,
+      "loss": 12.404,
+      "step": 9419
+    },
+    {
+      "epoch": 1.647344262295082,
+      "grad_norm": 0.8414209540650591,
+      "learning_rate": 8.896773666159131e-05,
+      "loss": 12.0241,
+      "step": 9420
+    },
+    {
+      "epoch": 1.6475191256830601,
+      "grad_norm": 0.8760228629266117,
+      "learning_rate": 8.894897121238747e-05,
+      "loss": 12.3097,
+      "step": 9421
+    },
+    {
+      "epoch": 1.6476939890710383,
+      "grad_norm": 0.9340961358475026,
+      "learning_rate": 8.893020615713985e-05,
+      "loss": 12.2671,
+      "step": 9422
+    },
+    {
+      "epoch": 1.6478688524590164,
+      "grad_norm": 0.8254992771266029,
+      "learning_rate": 8.891144149651745e-05,
+      "loss": 12.3679,
+      "step": 9423
+    },
+    {
+      "epoch": 1.6480437158469945,
+      "grad_norm": 1.005846982301549,
+      "learning_rate": 8.889267723118921e-05,
+      "loss": 12.6217,
+      "step": 9424
+    },
+    {
+      "epoch": 1.6482185792349728,
+      "grad_norm": 0.8558081452513971,
+      "learning_rate": 8.887391336182408e-05,
+      "loss": 12.616,
+      "step": 9425
+    },
+    {
+      "epoch": 1.6483934426229507,
+      "grad_norm": 0.9269872751320422,
+      "learning_rate": 8.88551498890909e-05,
+      "loss": 12.4675,
+      "step": 9426
+    },
+    {
+      "epoch": 1.648568306010929,
+      "grad_norm": 0.9260980950861534,
+      "learning_rate": 8.88363868136586e-05,
+      "loss": 12.4154,
+      "step": 9427
+    },
+    {
+      "epoch": 1.648743169398907,
+      "grad_norm": 0.8595578236046979,
+      "learning_rate": 8.881762413619608e-05,
+      "loss": 12.3782,
+      "step": 9428
+    },
+    {
+      "epoch": 1.6489180327868853,
+      "grad_norm": 0.8205036984050875,
+      "learning_rate": 8.879886185737221e-05,
+      "loss": 12.5461,
+      "step": 9429
+    },
+    {
+      "epoch": 1.6490928961748634,
+      "grad_norm": 0.8317799736078726,
+      "learning_rate": 8.878009997785581e-05,
+      "loss": 12.3656,
+      "step": 9430
+    },
+    {
+      "epoch": 1.6492677595628416,
+      "grad_norm": 0.8764319380001973,
+      "learning_rate": 8.876133849831572e-05,
+      "loss": 12.5232,
+      "step": 9431
+    },
+    {
+      "epoch": 1.6494426229508197,
+      "grad_norm": 0.9309297860981803,
+      "learning_rate": 8.87425774194208e-05,
+      "loss": 12.3684,
+      "step": 9432
+    },
+    {
+      "epoch": 1.6496174863387978,
+      "grad_norm": 0.9218250279049235,
+      "learning_rate": 8.872381674183989e-05,
+      "loss": 12.469,
+      "step": 9433
+    },
+    {
+      "epoch": 1.649792349726776,
+      "grad_norm": 0.9747748921504878,
+      "learning_rate": 8.870505646624167e-05,
+      "loss": 12.5851,
+      "step": 9434
+    },
+    {
+      "epoch": 1.649967213114754,
+      "grad_norm": 0.9056473911242849,
+      "learning_rate": 8.868629659329502e-05,
+      "loss": 12.3858,
+      "step": 9435
+    },
+    {
+      "epoch": 1.6501420765027324,
+      "grad_norm": 0.9403373533974563,
+      "learning_rate": 8.866753712366867e-05,
+      "loss": 12.4903,
+      "step": 9436
+    },
+    {
+      "epoch": 1.6503169398907103,
+      "grad_norm": 0.9144114142012154,
+      "learning_rate": 8.864877805803138e-05,
+      "loss": 12.3959,
+      "step": 9437
+    },
+    {
+      "epoch": 1.6504918032786886,
+      "grad_norm": 1.0708544095817738,
+      "learning_rate": 8.863001939705189e-05,
+      "loss": 12.6251,
+      "step": 9438
+    },
+    {
+      "epoch": 1.6506666666666665,
+      "grad_norm": 0.8910344870319512,
+      "learning_rate": 8.861126114139892e-05,
+      "loss": 12.3146,
+      "step": 9439
+    },
+    {
+      "epoch": 1.6508415300546448,
+      "grad_norm": 0.8223277792664022,
+      "learning_rate": 8.859250329174123e-05,
+      "loss": 12.5721,
+      "step": 9440
+    },
+    {
+      "epoch": 1.651016393442623,
+      "grad_norm": 1.035704297204595,
+      "learning_rate": 8.857374584874741e-05,
+      "loss": 12.5157,
+      "step": 9441
+    },
+    {
+      "epoch": 1.651191256830601,
+      "grad_norm": 0.8603774476464121,
+      "learning_rate": 8.855498881308621e-05,
+      "loss": 12.5515,
+      "step": 9442
+    },
+    {
+      "epoch": 1.6513661202185792,
+      "grad_norm": 0.8336322049632011,
+      "learning_rate": 8.853623218542629e-05,
+      "loss": 12.6927,
+      "step": 9443
+    },
+    {
+      "epoch": 1.6515409836065573,
+      "grad_norm": 0.953065592030103,
+      "learning_rate": 8.851747596643631e-05,
+      "loss": 12.2601,
+      "step": 9444
+    },
+    {
+      "epoch": 1.6517158469945357,
+      "grad_norm": 0.9266583904304027,
+      "learning_rate": 8.849872015678489e-05,
+      "loss": 12.5959,
+      "step": 9445
+    },
+    {
+      "epoch": 1.6518907103825136,
+      "grad_norm": 0.9175886203531197,
+      "learning_rate": 8.847996475714065e-05,
+      "loss": 12.3061,
+      "step": 9446
+    },
+    {
+      "epoch": 1.652065573770492,
+      "grad_norm": 0.9280142826949787,
+      "learning_rate": 8.84612097681722e-05,
+      "loss": 12.1824,
+      "step": 9447
+    },
+    {
+      "epoch": 1.6522404371584698,
+      "grad_norm": 0.8872990048524403,
+      "learning_rate": 8.84424551905482e-05,
+      "loss": 12.38,
+      "step": 9448
+    },
+    {
+      "epoch": 1.6524153005464481,
+      "grad_norm": 1.029758547774045,
+      "learning_rate": 8.84237010249371e-05,
+      "loss": 12.3283,
+      "step": 9449
+    },
+    {
+      "epoch": 1.6525901639344263,
+      "grad_norm": 0.8465106788162057,
+      "learning_rate": 8.840494727200757e-05,
+      "loss": 12.2872,
+      "step": 9450
+    },
+    {
+      "epoch": 1.6527650273224044,
+      "grad_norm": 1.06079967945314,
+      "learning_rate": 8.838619393242814e-05,
+      "loss": 12.2652,
+      "step": 9451
+    },
+    {
+      "epoch": 1.6529398907103825,
+      "grad_norm": 0.8996106568399752,
+      "learning_rate": 8.836744100686728e-05,
+      "loss": 12.5569,
+      "step": 9452
+    },
+    {
+      "epoch": 1.6531147540983606,
+      "grad_norm": 1.0378212905170257,
+      "learning_rate": 8.834868849599358e-05,
+      "loss": 12.5242,
+      "step": 9453
+    },
+    {
+      "epoch": 1.653289617486339,
+      "grad_norm": 0.9691955161879316,
+      "learning_rate": 8.832993640047551e-05,
+      "loss": 12.4273,
+      "step": 9454
+    },
+    {
+      "epoch": 1.6534644808743169,
+      "grad_norm": 0.83481818627919,
+      "learning_rate": 8.831118472098158e-05,
+      "loss": 12.4053,
+      "step": 9455
+    },
+    {
+      "epoch": 1.6536393442622952,
+      "grad_norm": 1.2021761134723414,
+      "learning_rate": 8.829243345818025e-05,
+      "loss": 12.1933,
+      "step": 9456
+    },
+    {
+      "epoch": 1.653814207650273,
+      "grad_norm": 0.8091561232972599,
+      "learning_rate": 8.827368261273999e-05,
+      "loss": 12.4493,
+      "step": 9457
+    },
+    {
+      "epoch": 1.6539890710382514,
+      "grad_norm": 1.0206031081206737,
+      "learning_rate": 8.825493218532924e-05,
+      "loss": 12.4606,
+      "step": 9458
+    },
+    {
+      "epoch": 1.6541639344262296,
+      "grad_norm": 0.9813153240720285,
+      "learning_rate": 8.823618217661648e-05,
+      "loss": 12.2285,
+      "step": 9459
+    },
+    {
+      "epoch": 1.6543387978142077,
+      "grad_norm": 0.8914820590984277,
+      "learning_rate": 8.821743258727004e-05,
+      "loss": 12.3885,
+      "step": 9460
+    },
+    {
+      "epoch": 1.6545136612021858,
+      "grad_norm": 1.0201243061649647,
+      "learning_rate": 8.819868341795836e-05,
+      "loss": 12.3489,
+      "step": 9461
+    },
+    {
+      "epoch": 1.654688524590164,
+      "grad_norm": 0.8526050992464609,
+      "learning_rate": 8.817993466934984e-05,
+      "loss": 12.2516,
+      "step": 9462
+    },
+    {
+      "epoch": 1.654863387978142,
+      "grad_norm": 0.9589221388682012,
+      "learning_rate": 8.816118634211284e-05,
+      "loss": 12.0934,
+      "step": 9463
+    },
+    {
+      "epoch": 1.6550382513661202,
+      "grad_norm": 0.8465526692759612,
+      "learning_rate": 8.81424384369157e-05,
+      "loss": 12.3066,
+      "step": 9464
+    },
+    {
+      "epoch": 1.6552131147540985,
+      "grad_norm": 0.8887916400776031,
+      "learning_rate": 8.812369095442677e-05,
+      "loss": 12.2337,
+      "step": 9465
+    },
+    {
+      "epoch": 1.6553879781420764,
+      "grad_norm": 0.9083632836211323,
+      "learning_rate": 8.810494389531443e-05,
+      "loss": 12.4899,
+      "step": 9466
+    },
+    {
+      "epoch": 1.6555628415300547,
+      "grad_norm": 0.8424056409518047,
+      "learning_rate": 8.80861972602469e-05,
+      "loss": 12.4839,
+      "step": 9467
+    },
+    {
+      "epoch": 1.6557377049180326,
+      "grad_norm": 0.9069156873926751,
+      "learning_rate": 8.806745104989251e-05,
+      "loss": 12.3761,
+      "step": 9468
+    },
+    {
+      "epoch": 1.655912568306011,
+      "grad_norm": 0.8740495314017528,
+      "learning_rate": 8.804870526491957e-05,
+      "loss": 12.4434,
+      "step": 9469
+    },
+    {
+      "epoch": 1.656087431693989,
+      "grad_norm": 0.9145175164619912,
+      "learning_rate": 8.802995990599634e-05,
+      "loss": 12.2823,
+      "step": 9470
+    },
+    {
+      "epoch": 1.6562622950819672,
+      "grad_norm": 0.8590374027518649,
+      "learning_rate": 8.801121497379103e-05,
+      "loss": 12.3734,
+      "step": 9471
+    },
+    {
+      "epoch": 1.6564371584699453,
+      "grad_norm": 0.8387695811204341,
+      "learning_rate": 8.799247046897189e-05,
+      "loss": 12.3565,
+      "step": 9472
+    },
+    {
+      "epoch": 1.6566120218579234,
+      "grad_norm": 0.9292240405876124,
+      "learning_rate": 8.797372639220715e-05,
+      "loss": 12.3547,
+      "step": 9473
+    },
+    {
+      "epoch": 1.6567868852459018,
+      "grad_norm": 0.8410040370802538,
+      "learning_rate": 8.795498274416504e-05,
+      "loss": 12.3191,
+      "step": 9474
+    },
+    {
+      "epoch": 1.6569617486338797,
+      "grad_norm": 0.8603657533355824,
+      "learning_rate": 8.793623952551369e-05,
+      "loss": 12.5682,
+      "step": 9475
+    },
+    {
+      "epoch": 1.657136612021858,
+      "grad_norm": 0.9320715559951456,
+      "learning_rate": 8.791749673692131e-05,
+      "loss": 12.2997,
+      "step": 9476
+    },
+    {
+      "epoch": 1.657311475409836,
+      "grad_norm": 0.8809501595342889,
+      "learning_rate": 8.789875437905605e-05,
+      "loss": 12.328,
+      "step": 9477
+    },
+    {
+      "epoch": 1.6574863387978143,
+      "grad_norm": 0.8374250400343395,
+      "learning_rate": 8.788001245258611e-05,
+      "loss": 12.4853,
+      "step": 9478
+    },
+    {
+      "epoch": 1.6576612021857924,
+      "grad_norm": 0.858639361567042,
+      "learning_rate": 8.786127095817952e-05,
+      "loss": 12.4785,
+      "step": 9479
+    },
+    {
+      "epoch": 1.6578360655737705,
+      "grad_norm": 0.9664068913064389,
+      "learning_rate": 8.784252989650443e-05,
+      "loss": 12.4189,
+      "step": 9480
+    },
+    {
+      "epoch": 1.6580109289617486,
+      "grad_norm": 0.7886775555589691,
+      "learning_rate": 8.782378926822898e-05,
+      "loss": 12.463,
+      "step": 9481
+    },
+    {
+      "epoch": 1.6581857923497267,
+      "grad_norm": 0.8586586000423226,
+      "learning_rate": 8.780504907402119e-05,
+      "loss": 12.3814,
+      "step": 9482
+    },
+    {
+      "epoch": 1.658360655737705,
+      "grad_norm": 0.8091491453013238,
+      "learning_rate": 8.778630931454914e-05,
+      "loss": 12.4949,
+      "step": 9483
+    },
+    {
+      "epoch": 1.658535519125683,
+      "grad_norm": 0.8176837673796474,
+      "learning_rate": 8.776756999048091e-05,
+      "loss": 12.5727,
+      "step": 9484
+    },
+    {
+      "epoch": 1.6587103825136613,
+      "grad_norm": 0.7929312041906066,
+      "learning_rate": 8.774883110248455e-05,
+      "loss": 12.3044,
+      "step": 9485
+    },
+    {
+      "epoch": 1.6588852459016392,
+      "grad_norm": 0.8805916760095298,
+      "learning_rate": 8.773009265122801e-05,
+      "loss": 12.4404,
+      "step": 9486
+    },
+    {
+      "epoch": 1.6590601092896176,
+      "grad_norm": 0.8855381582890577,
+      "learning_rate": 8.771135463737933e-05,
+      "loss": 12.3908,
+      "step": 9487
+    },
+    {
+      "epoch": 1.6592349726775957,
+      "grad_norm": 0.8142561577748099,
+      "learning_rate": 8.76926170616065e-05,
+      "loss": 12.5112,
+      "step": 9488
+    },
+    {
+      "epoch": 1.6594098360655738,
+      "grad_norm": 0.8633048045193886,
+      "learning_rate": 8.767387992457751e-05,
+      "loss": 12.6667,
+      "step": 9489
+    },
+    {
+      "epoch": 1.659584699453552,
+      "grad_norm": 0.8013948060884099,
+      "learning_rate": 8.765514322696028e-05,
+      "loss": 12.1754,
+      "step": 9490
+    },
+    {
+      "epoch": 1.65975956284153,
+      "grad_norm": 0.8542848802866024,
+      "learning_rate": 8.763640696942277e-05,
+      "loss": 12.4241,
+      "step": 9491
+    },
+    {
+      "epoch": 1.6599344262295082,
+      "grad_norm": 0.9611286572026894,
+      "learning_rate": 8.761767115263291e-05,
+      "loss": 12.2602,
+      "step": 9492
+    },
+    {
+      "epoch": 1.6601092896174863,
+      "grad_norm": 0.8586574631928995,
+      "learning_rate": 8.759893577725864e-05,
+      "loss": 12.6089,
+      "step": 9493
+    },
+    {
+      "epoch": 1.6602841530054646,
+      "grad_norm": 0.8970002866998756,
+      "learning_rate": 8.758020084396778e-05,
+      "loss": 12.2958,
+      "step": 9494
+    },
+    {
+      "epoch": 1.6604590163934425,
+      "grad_norm": 0.8846020374794965,
+      "learning_rate": 8.756146635342826e-05,
+      "loss": 12.5943,
+      "step": 9495
+    },
+    {
+      "epoch": 1.6606338797814209,
+      "grad_norm": 0.8515639348639372,
+      "learning_rate": 8.754273230630792e-05,
+      "loss": 12.4128,
+      "step": 9496
+    },
+    {
+      "epoch": 1.6608087431693987,
+      "grad_norm": 0.882672835733929,
+      "learning_rate": 8.752399870327466e-05,
+      "loss": 12.3118,
+      "step": 9497
+    },
+    {
+      "epoch": 1.660983606557377,
+      "grad_norm": 0.8618120287888688,
+      "learning_rate": 8.750526554499623e-05,
+      "loss": 12.3683,
+      "step": 9498
+    },
+    {
+      "epoch": 1.6611584699453552,
+      "grad_norm": 0.9160051121861648,
+      "learning_rate": 8.748653283214049e-05,
+      "loss": 12.0311,
+      "step": 9499
+    },
+    {
+      "epoch": 1.6613333333333333,
+      "grad_norm": 0.8360244397448144,
+      "learning_rate": 8.746780056537524e-05,
+      "loss": 12.2856,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6615081967213114,
+      "grad_norm": 0.9662314114704509,
+      "learning_rate": 8.744906874536825e-05,
+      "loss": 12.0997,
+      "step": 9501
+    },
+    {
+      "epoch": 1.6616830601092896,
+      "grad_norm": 0.8561099757106927,
+      "learning_rate": 8.74303373727873e-05,
+      "loss": 12.3436,
+      "step": 9502
+    },
+    {
+      "epoch": 1.661857923497268,
+      "grad_norm": 0.9759305958117179,
+      "learning_rate": 8.741160644830013e-05,
+      "loss": 12.5914,
+      "step": 9503
+    },
+    {
+      "epoch": 1.6620327868852458,
+      "grad_norm": 0.9671701318646592,
+      "learning_rate": 8.739287597257453e-05,
+      "loss": 12.4194,
+      "step": 9504
+    },
+    {
+      "epoch": 1.6622076502732241,
+      "grad_norm": 0.9568677996211298,
+      "learning_rate": 8.737414594627811e-05,
+      "loss": 12.5054,
+      "step": 9505
+    },
+    {
+      "epoch": 1.662382513661202,
+      "grad_norm": 0.9197831380546796,
+      "learning_rate": 8.735541637007865e-05,
+      "loss": 12.2854,
+      "step": 9506
+    },
+    {
+      "epoch": 1.6625573770491804,
+      "grad_norm": 0.8777972045152752,
+      "learning_rate": 8.733668724464383e-05,
+      "loss": 12.3477,
+      "step": 9507
+    },
+    {
+      "epoch": 1.6627322404371585,
+      "grad_norm": 0.8469349016549714,
+      "learning_rate": 8.731795857064134e-05,
+      "loss": 12.1375,
+      "step": 9508
+    },
+    {
+      "epoch": 1.6629071038251366,
+      "grad_norm": 0.8309966927857881,
+      "learning_rate": 8.729923034873877e-05,
+      "loss": 12.5223,
+      "step": 9509
+    },
+    {
+      "epoch": 1.6630819672131147,
+      "grad_norm": 0.9354696390844488,
+      "learning_rate": 8.728050257960379e-05,
+      "loss": 12.4202,
+      "step": 9510
+    },
+    {
+      "epoch": 1.6632568306010929,
+      "grad_norm": 0.8181894428825042,
+      "learning_rate": 8.726177526390406e-05,
+      "loss": 12.371,
+      "step": 9511
+    },
+    {
+      "epoch": 1.6634316939890712,
+      "grad_norm": 0.9247643897297729,
+      "learning_rate": 8.724304840230718e-05,
+      "loss": 12.3749,
+      "step": 9512
+    },
+    {
+      "epoch": 1.663606557377049,
+      "grad_norm": 0.8167010147459076,
+      "learning_rate": 8.722432199548068e-05,
+      "loss": 12.6153,
+      "step": 9513
+    },
+    {
+      "epoch": 1.6637814207650274,
+      "grad_norm": 0.8456607150327055,
+      "learning_rate": 8.720559604409219e-05,
+      "loss": 12.395,
+      "step": 9514
+    },
+    {
+      "epoch": 1.6639562841530053,
+      "grad_norm": 0.8952876511439045,
+      "learning_rate": 8.718687054880927e-05,
+      "loss": 12.5726,
+      "step": 9515
+    },
+    {
+      "epoch": 1.6641311475409837,
+      "grad_norm": 0.8708058830598523,
+      "learning_rate": 8.716814551029942e-05,
+      "loss": 12.2674,
+      "step": 9516
+    },
+    {
+      "epoch": 1.6643060109289618,
+      "grad_norm": 0.828386566837521,
+      "learning_rate": 8.714942092923018e-05,
+      "loss": 12.3477,
+      "step": 9517
+    },
+    {
+      "epoch": 1.66448087431694,
+      "grad_norm": 0.9104053214972829,
+      "learning_rate": 8.713069680626909e-05,
+      "loss": 12.149,
+      "step": 9518
+    },
+    {
+      "epoch": 1.664655737704918,
+      "grad_norm": 0.8939072300069097,
+      "learning_rate": 8.711197314208363e-05,
+      "loss": 12.193,
+      "step": 9519
+    },
+    {
+      "epoch": 1.6648306010928962,
+      "grad_norm": 0.953269030605929,
+      "learning_rate": 8.709324993734124e-05,
+      "loss": 12.3019,
+      "step": 9520
+    },
+    {
+      "epoch": 1.6650054644808743,
+      "grad_norm": 0.8842656536095821,
+      "learning_rate": 8.707452719270942e-05,
+      "loss": 12.444,
+      "step": 9521
+    },
+    {
+      "epoch": 1.6651803278688524,
+      "grad_norm": 0.9363520582919254,
+      "learning_rate": 8.70558049088556e-05,
+      "loss": 12.5055,
+      "step": 9522
+    },
+    {
+      "epoch": 1.6653551912568307,
+      "grad_norm": 0.8041628816412986,
+      "learning_rate": 8.703708308644725e-05,
+      "loss": 12.4753,
+      "step": 9523
+    },
+    {
+      "epoch": 1.6655300546448086,
+      "grad_norm": 0.8695152335383938,
+      "learning_rate": 8.70183617261517e-05,
+      "loss": 12.2934,
+      "step": 9524
+    },
+    {
+      "epoch": 1.665704918032787,
+      "grad_norm": 0.8623544781222179,
+      "learning_rate": 8.69996408286364e-05,
+      "loss": 12.2068,
+      "step": 9525
+    },
+    {
+      "epoch": 1.6658797814207649,
+      "grad_norm": 0.8823211430551698,
+      "learning_rate": 8.698092039456872e-05,
+      "loss": 12.5096,
+      "step": 9526
+    },
+    {
+      "epoch": 1.6660546448087432,
+      "grad_norm": 0.8721790359779125,
+      "learning_rate": 8.696220042461602e-05,
+      "loss": 12.2753,
+      "step": 9527
+    },
+    {
+      "epoch": 1.6662295081967213,
+      "grad_norm": 0.9509547122064781,
+      "learning_rate": 8.694348091944562e-05,
+      "loss": 12.7363,
+      "step": 9528
+    },
+    {
+      "epoch": 1.6664043715846995,
+      "grad_norm": 0.8797537005923901,
+      "learning_rate": 8.692476187972487e-05,
+      "loss": 12.5848,
+      "step": 9529
+    },
+    {
+      "epoch": 1.6665792349726776,
+      "grad_norm": 0.855329945206678,
+      "learning_rate": 8.690604330612114e-05,
+      "loss": 12.2179,
+      "step": 9530
+    },
+    {
+      "epoch": 1.6667540983606557,
+      "grad_norm": 0.8804151694004223,
+      "learning_rate": 8.68873251993016e-05,
+      "loss": 12.3827,
+      "step": 9531
+    },
+    {
+      "epoch": 1.666928961748634,
+      "grad_norm": 0.8514305666480497,
+      "learning_rate": 8.686860755993362e-05,
+      "loss": 12.5224,
+      "step": 9532
+    },
+    {
+      "epoch": 1.667103825136612,
+      "grad_norm": 0.8972632055846413,
+      "learning_rate": 8.684989038868444e-05,
+      "loss": 12.6163,
+      "step": 9533
+    },
+    {
+      "epoch": 1.6672786885245903,
+      "grad_norm": 0.7914041838007753,
+      "learning_rate": 8.68311736862213e-05,
+      "loss": 12.0862,
+      "step": 9534
+    },
+    {
+      "epoch": 1.6674535519125682,
+      "grad_norm": 0.9377433739536636,
+      "learning_rate": 8.681245745321143e-05,
+      "loss": 12.465,
+      "step": 9535
+    },
+    {
+      "epoch": 1.6676284153005465,
+      "grad_norm": 0.9196358076370018,
+      "learning_rate": 8.679374169032204e-05,
+      "loss": 12.1878,
+      "step": 9536
+    },
+    {
+      "epoch": 1.6678032786885246,
+      "grad_norm": 0.8330923261828096,
+      "learning_rate": 8.677502639822033e-05,
+      "loss": 12.4289,
+      "step": 9537
+    },
+    {
+      "epoch": 1.6679781420765027,
+      "grad_norm": 0.8903864401926884,
+      "learning_rate": 8.675631157757351e-05,
+      "loss": 12.096,
+      "step": 9538
+    },
+    {
+      "epoch": 1.6681530054644809,
+      "grad_norm": 0.7892724475947079,
+      "learning_rate": 8.673759722904868e-05,
+      "loss": 12.2806,
+      "step": 9539
+    },
+    {
+      "epoch": 1.668327868852459,
+      "grad_norm": 0.7942380734455243,
+      "learning_rate": 8.6718883353313e-05,
+      "loss": 12.0517,
+      "step": 9540
+    },
+    {
+      "epoch": 1.668502732240437,
+      "grad_norm": 0.8367197582390018,
+      "learning_rate": 8.670016995103363e-05,
+      "loss": 12.1277,
+      "step": 9541
+    },
+    {
+      "epoch": 1.6686775956284152,
+      "grad_norm": 0.8451829527352458,
+      "learning_rate": 8.668145702287769e-05,
+      "loss": 12.5189,
+      "step": 9542
+    },
+    {
+      "epoch": 1.6688524590163936,
+      "grad_norm": 0.848488418274663,
+      "learning_rate": 8.66627445695122e-05,
+      "loss": 12.1559,
+      "step": 9543
+    },
+    {
+      "epoch": 1.6690273224043715,
+      "grad_norm": 0.8671063835144477,
+      "learning_rate": 8.66440325916043e-05,
+      "loss": 12.4875,
+      "step": 9544
+    },
+    {
+      "epoch": 1.6692021857923498,
+      "grad_norm": 0.8296202312323183,
+      "learning_rate": 8.662532108982104e-05,
+      "loss": 12.4083,
+      "step": 9545
+    },
+    {
+      "epoch": 1.669377049180328,
+      "grad_norm": 0.8010123141182754,
+      "learning_rate": 8.660661006482943e-05,
+      "loss": 12.3589,
+      "step": 9546
+    },
+    {
+      "epoch": 1.669551912568306,
+      "grad_norm": 0.8135721659106544,
+      "learning_rate": 8.658789951729653e-05,
+      "loss": 12.6416,
+      "step": 9547
+    },
+    {
+      "epoch": 1.6697267759562842,
+      "grad_norm": 0.8295827802172475,
+      "learning_rate": 8.656918944788935e-05,
+      "loss": 12.3269,
+      "step": 9548
+    },
+    {
+      "epoch": 1.6699016393442623,
+      "grad_norm": 0.8096770146944436,
+      "learning_rate": 8.655047985727492e-05,
+      "loss": 12.3477,
+      "step": 9549
+    },
+    {
+      "epoch": 1.6700765027322404,
+      "grad_norm": 0.798838047969943,
+      "learning_rate": 8.653177074612011e-05,
+      "loss": 12.4344,
+      "step": 9550
+    },
+    {
+      "epoch": 1.6702513661202185,
+      "grad_norm": 0.8431899420793462,
+      "learning_rate": 8.651306211509194e-05,
+      "loss": 12.4777,
+      "step": 9551
+    },
+    {
+      "epoch": 1.6704262295081969,
+      "grad_norm": 0.8341497988968088,
+      "learning_rate": 8.649435396485735e-05,
+      "loss": 12.5218,
+      "step": 9552
+    },
+    {
+      "epoch": 1.6706010928961748,
+      "grad_norm": 0.833216162840044,
+      "learning_rate": 8.647564629608329e-05,
+      "loss": 12.3759,
+      "step": 9553
+    },
+    {
+      "epoch": 1.670775956284153,
+      "grad_norm": 0.7871436568242034,
+      "learning_rate": 8.64569391094366e-05,
+      "loss": 12.3814,
+      "step": 9554
+    },
+    {
+      "epoch": 1.670950819672131,
+      "grad_norm": 0.8126883830063427,
+      "learning_rate": 8.643823240558421e-05,
+      "loss": 12.4191,
+      "step": 9555
+    },
+    {
+      "epoch": 1.6711256830601093,
+      "grad_norm": 0.9257497695569287,
+      "learning_rate": 8.6419526185193e-05,
+      "loss": 12.2797,
+      "step": 9556
+    },
+    {
+      "epoch": 1.6713005464480875,
+      "grad_norm": 0.8639333962769389,
+      "learning_rate": 8.640082044892986e-05,
+      "loss": 12.5195,
+      "step": 9557
+    },
+    {
+      "epoch": 1.6714754098360656,
+      "grad_norm": 0.8852797041501307,
+      "learning_rate": 8.638211519746154e-05,
+      "loss": 12.3198,
+      "step": 9558
+    },
+    {
+      "epoch": 1.6716502732240437,
+      "grad_norm": 0.8184511845328485,
+      "learning_rate": 8.636341043145491e-05,
+      "loss": 12.4053,
+      "step": 9559
+    },
+    {
+      "epoch": 1.6718251366120218,
+      "grad_norm": 0.878208069040532,
+      "learning_rate": 8.634470615157679e-05,
+      "loss": 12.4168,
+      "step": 9560
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 0.8369201381123921,
+      "learning_rate": 8.632600235849389e-05,
+      "loss": 12.5063,
+      "step": 9561
+    },
+    {
+      "epoch": 1.672174863387978,
+      "grad_norm": 0.8383235112429234,
+      "learning_rate": 8.630729905287305e-05,
+      "loss": 12.2931,
+      "step": 9562
+    },
+    {
+      "epoch": 1.6723497267759564,
+      "grad_norm": 0.927606887109522,
+      "learning_rate": 8.6288596235381e-05,
+      "loss": 12.6162,
+      "step": 9563
+    },
+    {
+      "epoch": 1.6725245901639343,
+      "grad_norm": 0.9448098600557481,
+      "learning_rate": 8.626989390668448e-05,
+      "loss": 12.4762,
+      "step": 9564
+    },
+    {
+      "epoch": 1.6726994535519126,
+      "grad_norm": 0.9798253207565408,
+      "learning_rate": 8.62511920674502e-05,
+      "loss": 12.463,
+      "step": 9565
+    },
+    {
+      "epoch": 1.6728743169398907,
+      "grad_norm": 0.822983929755255,
+      "learning_rate": 8.623249071834483e-05,
+      "loss": 11.9987,
+      "step": 9566
+    },
+    {
+      "epoch": 1.6730491803278689,
+      "grad_norm": 0.9452536303086394,
+      "learning_rate": 8.62137898600351e-05,
+      "loss": 12.0574,
+      "step": 9567
+    },
+    {
+      "epoch": 1.673224043715847,
+      "grad_norm": 0.8805562853159419,
+      "learning_rate": 8.619508949318769e-05,
+      "loss": 12.4205,
+      "step": 9568
+    },
+    {
+      "epoch": 1.673398907103825,
+      "grad_norm": 0.7787216687938341,
+      "learning_rate": 8.617638961846918e-05,
+      "loss": 12.4076,
+      "step": 9569
+    },
+    {
+      "epoch": 1.6735737704918032,
+      "grad_norm": 1.1228181837004705,
+      "learning_rate": 8.615769023654623e-05,
+      "loss": 12.2418,
+      "step": 9570
+    },
+    {
+      "epoch": 1.6737486338797813,
+      "grad_norm": 0.8517917477275708,
+      "learning_rate": 8.613899134808543e-05,
+      "loss": 12.4595,
+      "step": 9571
+    },
+    {
+      "epoch": 1.6739234972677597,
+      "grad_norm": 0.961676410699922,
+      "learning_rate": 8.612029295375343e-05,
+      "loss": 12.4825,
+      "step": 9572
+    },
+    {
+      "epoch": 1.6740983606557376,
+      "grad_norm": 0.9873627964636547,
+      "learning_rate": 8.610159505421674e-05,
+      "loss": 12.6218,
+      "step": 9573
+    },
+    {
+      "epoch": 1.674273224043716,
+      "grad_norm": 0.8696692394811056,
+      "learning_rate": 8.608289765014196e-05,
+      "loss": 12.2953,
+      "step": 9574
+    },
+    {
+      "epoch": 1.674448087431694,
+      "grad_norm": 1.1435616784950644,
+      "learning_rate": 8.606420074219565e-05,
+      "loss": 12.5098,
+      "step": 9575
+    },
+    {
+      "epoch": 1.6746229508196722,
+      "grad_norm": 0.8921943982730288,
+      "learning_rate": 8.604550433104426e-05,
+      "loss": 12.2709,
+      "step": 9576
+    },
+    {
+      "epoch": 1.6747978142076503,
+      "grad_norm": 0.8850364611448914,
+      "learning_rate": 8.602680841735433e-05,
+      "loss": 12.4623,
+      "step": 9577
+    },
+    {
+      "epoch": 1.6749726775956284,
+      "grad_norm": 1.0729487509514157,
+      "learning_rate": 8.600811300179237e-05,
+      "loss": 12.424,
+      "step": 9578
+    },
+    {
+      "epoch": 1.6751475409836065,
+      "grad_norm": 0.8127270489743716,
+      "learning_rate": 8.598941808502484e-05,
+      "loss": 12.1025,
+      "step": 9579
+    },
+    {
+      "epoch": 1.6753224043715846,
+      "grad_norm": 1.0401647957685363,
+      "learning_rate": 8.597072366771818e-05,
+      "loss": 12.4472,
+      "step": 9580
+    },
+    {
+      "epoch": 1.675497267759563,
+      "grad_norm": 0.8512282922785127,
+      "learning_rate": 8.595202975053882e-05,
+      "loss": 12.192,
+      "step": 9581
+    },
+    {
+      "epoch": 1.6756721311475409,
+      "grad_norm": 0.9393565087798113,
+      "learning_rate": 8.59333363341532e-05,
+      "loss": 12.2342,
+      "step": 9582
+    },
+    {
+      "epoch": 1.6758469945355192,
+      "grad_norm": 0.9283939315300735,
+      "learning_rate": 8.591464341922772e-05,
+      "loss": 12.2964,
+      "step": 9583
+    },
+    {
+      "epoch": 1.6760218579234971,
+      "grad_norm": 0.8445107501023972,
+      "learning_rate": 8.589595100642871e-05,
+      "loss": 12.2645,
+      "step": 9584
+    },
+    {
+      "epoch": 1.6761967213114755,
+      "grad_norm": 0.9538369163788109,
+      "learning_rate": 8.587725909642258e-05,
+      "loss": 12.5875,
+      "step": 9585
+    },
+    {
+      "epoch": 1.6763715846994536,
+      "grad_norm": 0.8774658242403546,
+      "learning_rate": 8.585856768987566e-05,
+      "loss": 12.6113,
+      "step": 9586
+    },
+    {
+      "epoch": 1.6765464480874317,
+      "grad_norm": 0.9963274320715264,
+      "learning_rate": 8.583987678745432e-05,
+      "loss": 12.3462,
+      "step": 9587
+    },
+    {
+      "epoch": 1.6767213114754098,
+      "grad_norm": 0.8553809543451761,
+      "learning_rate": 8.582118638982478e-05,
+      "loss": 12.2962,
+      "step": 9588
+    },
+    {
+      "epoch": 1.676896174863388,
+      "grad_norm": 0.9866648108006186,
+      "learning_rate": 8.580249649765338e-05,
+      "loss": 12.6153,
+      "step": 9589
+    },
+    {
+      "epoch": 1.6770710382513663,
+      "grad_norm": 0.8966364994927224,
+      "learning_rate": 8.578380711160643e-05,
+      "loss": 12.4671,
+      "step": 9590
+    },
+    {
+      "epoch": 1.6772459016393442,
+      "grad_norm": 0.8247005462377744,
+      "learning_rate": 8.57651182323501e-05,
+      "loss": 12.5493,
+      "step": 9591
+    },
+    {
+      "epoch": 1.6774207650273225,
+      "grad_norm": 0.955757041825362,
+      "learning_rate": 8.57464298605507e-05,
+      "loss": 12.2103,
+      "step": 9592
+    },
+    {
+      "epoch": 1.6775956284153004,
+      "grad_norm": 0.9415665789554071,
+      "learning_rate": 8.57277419968744e-05,
+      "loss": 12.2748,
+      "step": 9593
+    },
+    {
+      "epoch": 1.6777704918032788,
+      "grad_norm": 0.8775544220924428,
+      "learning_rate": 8.570905464198747e-05,
+      "loss": 12.0317,
+      "step": 9594
+    },
+    {
+      "epoch": 1.6779453551912569,
+      "grad_norm": 0.8489148411277712,
+      "learning_rate": 8.5690367796556e-05,
+      "loss": 12.5046,
+      "step": 9595
+    },
+    {
+      "epoch": 1.678120218579235,
+      "grad_norm": 1.0249559121590346,
+      "learning_rate": 8.567168146124621e-05,
+      "loss": 12.4962,
+      "step": 9596
+    },
+    {
+      "epoch": 1.678295081967213,
+      "grad_norm": 0.8537354430969748,
+      "learning_rate": 8.565299563672423e-05,
+      "loss": 12.4098,
+      "step": 9597
+    },
+    {
+      "epoch": 1.6784699453551912,
+      "grad_norm": 0.9513899472220745,
+      "learning_rate": 8.56343103236562e-05,
+      "loss": 12.4721,
+      "step": 9598
+    },
+    {
+      "epoch": 1.6786448087431693,
+      "grad_norm": 0.8903629490130955,
+      "learning_rate": 8.561562552270823e-05,
+      "loss": 12.3283,
+      "step": 9599
+    },
+    {
+      "epoch": 1.6788196721311475,
+      "grad_norm": 0.9031842978856202,
+      "learning_rate": 8.559694123454638e-05,
+      "loss": 12.5229,
+      "step": 9600
+    },
+    {
+      "epoch": 1.6789945355191258,
+      "grad_norm": 1.0681985560445244,
+      "learning_rate": 8.557825745983676e-05,
+      "loss": 12.2209,
+      "step": 9601
+    },
+    {
+      "epoch": 1.6791693989071037,
+      "grad_norm": 0.8445469714598773,
+      "learning_rate": 8.555957419924545e-05,
+      "loss": 12.4332,
+      "step": 9602
+    },
+    {
+      "epoch": 1.679344262295082,
+      "grad_norm": 0.835089101905763,
+      "learning_rate": 8.55408914534384e-05,
+      "loss": 12.4733,
+      "step": 9603
+    },
+    {
+      "epoch": 1.67951912568306,
+      "grad_norm": 0.9858217472467418,
+      "learning_rate": 8.552220922308169e-05,
+      "loss": 12.6056,
+      "step": 9604
+    },
+    {
+      "epoch": 1.6796939890710383,
+      "grad_norm": 0.8418095307759471,
+      "learning_rate": 8.550352750884133e-05,
+      "loss": 12.4325,
+      "step": 9605
+    },
+    {
+      "epoch": 1.6798688524590164,
+      "grad_norm": 0.9795908561306873,
+      "learning_rate": 8.548484631138325e-05,
+      "loss": 12.5705,
+      "step": 9606
+    },
+    {
+      "epoch": 1.6800437158469945,
+      "grad_norm": 1.2051617648724804,
+      "learning_rate": 8.546616563137343e-05,
+      "loss": 12.4329,
+      "step": 9607
+    },
+    {
+      "epoch": 1.6802185792349726,
+      "grad_norm": 0.9692397249060085,
+      "learning_rate": 8.544748546947783e-05,
+      "loss": 12.3605,
+      "step": 9608
+    },
+    {
+      "epoch": 1.6803934426229508,
+      "grad_norm": 1.1289977680775873,
+      "learning_rate": 8.54288058263624e-05,
+      "loss": 12.4863,
+      "step": 9609
+    },
+    {
+      "epoch": 1.680568306010929,
+      "grad_norm": 0.8522933974395096,
+      "learning_rate": 8.541012670269297e-05,
+      "loss": 12.3384,
+      "step": 9610
+    },
+    {
+      "epoch": 1.680743169398907,
+      "grad_norm": 1.1647686666901234,
+      "learning_rate": 8.539144809913548e-05,
+      "loss": 12.4799,
+      "step": 9611
+    },
+    {
+      "epoch": 1.6809180327868853,
+      "grad_norm": 0.974574518860736,
+      "learning_rate": 8.537277001635582e-05,
+      "loss": 12.4222,
+      "step": 9612
+    },
+    {
+      "epoch": 1.6810928961748632,
+      "grad_norm": 0.8401830434248883,
+      "learning_rate": 8.535409245501985e-05,
+      "loss": 12.2364,
+      "step": 9613
+    },
+    {
+      "epoch": 1.6812677595628416,
+      "grad_norm": 1.3510510026387637,
+      "learning_rate": 8.533541541579333e-05,
+      "loss": 12.5058,
+      "step": 9614
+    },
+    {
+      "epoch": 1.6814426229508197,
+      "grad_norm": 0.8682727286390431,
+      "learning_rate": 8.531673889934212e-05,
+      "loss": 12.5385,
+      "step": 9615
+    },
+    {
+      "epoch": 1.6816174863387978,
+      "grad_norm": 1.1030537752383525,
+      "learning_rate": 8.5298062906332e-05,
+      "loss": 12.5206,
+      "step": 9616
+    },
+    {
+      "epoch": 1.681792349726776,
+      "grad_norm": 1.0663572894657027,
+      "learning_rate": 8.527938743742879e-05,
+      "loss": 12.2404,
+      "step": 9617
+    },
+    {
+      "epoch": 1.681967213114754,
+      "grad_norm": 0.799070017993735,
+      "learning_rate": 8.52607124932982e-05,
+      "loss": 12.2321,
+      "step": 9618
+    },
+    {
+      "epoch": 1.6821420765027324,
+      "grad_norm": 1.1773773236679521,
+      "learning_rate": 8.524203807460599e-05,
+      "loss": 12.3912,
+      "step": 9619
+    },
+    {
+      "epoch": 1.6823169398907103,
+      "grad_norm": 0.9146312020202072,
+      "learning_rate": 8.522336418201787e-05,
+      "loss": 12.6135,
+      "step": 9620
+    },
+    {
+      "epoch": 1.6824918032786886,
+      "grad_norm": 0.9920228681414673,
+      "learning_rate": 8.52046908161996e-05,
+      "loss": 12.1281,
+      "step": 9621
+    },
+    {
+      "epoch": 1.6826666666666665,
+      "grad_norm": 0.8409526390270133,
+      "learning_rate": 8.518601797781678e-05,
+      "loss": 12.4974,
+      "step": 9622
+    },
+    {
+      "epoch": 1.6828415300546449,
+      "grad_norm": 0.8142110304649607,
+      "learning_rate": 8.51673456675351e-05,
+      "loss": 12.5749,
+      "step": 9623
+    },
+    {
+      "epoch": 1.683016393442623,
+      "grad_norm": 0.9149554891669859,
+      "learning_rate": 8.514867388602027e-05,
+      "loss": 12.3848,
+      "step": 9624
+    },
+    {
+      "epoch": 1.6831912568306011,
+      "grad_norm": 0.8337246764123352,
+      "learning_rate": 8.513000263393782e-05,
+      "loss": 12.3985,
+      "step": 9625
+    },
+    {
+      "epoch": 1.6833661202185792,
+      "grad_norm": 0.8971279502743026,
+      "learning_rate": 8.51113319119534e-05,
+      "loss": 12.3451,
+      "step": 9626
+    },
+    {
+      "epoch": 1.6835409836065574,
+      "grad_norm": 0.9361206378091889,
+      "learning_rate": 8.50926617207326e-05,
+      "loss": 12.4359,
+      "step": 9627
+    },
+    {
+      "epoch": 1.6837158469945355,
+      "grad_norm": 0.8556999701073332,
+      "learning_rate": 8.507399206094101e-05,
+      "loss": 12.3254,
+      "step": 9628
+    },
+    {
+      "epoch": 1.6838907103825136,
+      "grad_norm": 0.9304075718137647,
+      "learning_rate": 8.505532293324414e-05,
+      "loss": 12.2088,
+      "step": 9629
+    },
+    {
+      "epoch": 1.684065573770492,
+      "grad_norm": 0.8512398144899571,
+      "learning_rate": 8.503665433830757e-05,
+      "loss": 12.3743,
+      "step": 9630
+    },
+    {
+      "epoch": 1.6842404371584698,
+      "grad_norm": 0.917197134491199,
+      "learning_rate": 8.501798627679675e-05,
+      "loss": 12.5273,
+      "step": 9631
+    },
+    {
+      "epoch": 1.6844153005464482,
+      "grad_norm": 0.8605927366389857,
+      "learning_rate": 8.499931874937729e-05,
+      "loss": 12.4499,
+      "step": 9632
+    },
+    {
+      "epoch": 1.684590163934426,
+      "grad_norm": 0.9285462125080516,
+      "learning_rate": 8.498065175671455e-05,
+      "loss": 12.5333,
+      "step": 9633
+    },
+    {
+      "epoch": 1.6847650273224044,
+      "grad_norm": 0.8211163929925291,
+      "learning_rate": 8.496198529947401e-05,
+      "loss": 12.4375,
+      "step": 9634
+    },
+    {
+      "epoch": 1.6849398907103825,
+      "grad_norm": 0.9980139762883339,
+      "learning_rate": 8.494331937832112e-05,
+      "loss": 12.2563,
+      "step": 9635
+    },
+    {
+      "epoch": 1.6851147540983606,
+      "grad_norm": 0.8616115729820928,
+      "learning_rate": 8.492465399392133e-05,
+      "loss": 12.3074,
+      "step": 9636
+    },
+    {
+      "epoch": 1.6852896174863388,
+      "grad_norm": 0.9629199553757837,
+      "learning_rate": 8.490598914694e-05,
+      "loss": 12.3393,
+      "step": 9637
+    },
+    {
+      "epoch": 1.6854644808743169,
+      "grad_norm": 0.9616430685779083,
+      "learning_rate": 8.488732483804252e-05,
+      "loss": 12.1984,
+      "step": 9638
+    },
+    {
+      "epoch": 1.6856393442622952,
+      "grad_norm": 0.8186997451539337,
+      "learning_rate": 8.486866106789429e-05,
+      "loss": 12.349,
+      "step": 9639
+    },
+    {
+      "epoch": 1.6858142076502731,
+      "grad_norm": 0.9210898043262764,
+      "learning_rate": 8.484999783716057e-05,
+      "loss": 12.2988,
+      "step": 9640
+    },
+    {
+      "epoch": 1.6859890710382515,
+      "grad_norm": 0.9275177552340566,
+      "learning_rate": 8.483133514650671e-05,
+      "loss": 12.3054,
+      "step": 9641
+    },
+    {
+      "epoch": 1.6861639344262294,
+      "grad_norm": 0.8348063767663174,
+      "learning_rate": 8.481267299659805e-05,
+      "loss": 12.1272,
+      "step": 9642
+    },
+    {
+      "epoch": 1.6863387978142077,
+      "grad_norm": 0.9907455966431447,
+      "learning_rate": 8.479401138809987e-05,
+      "loss": 12.0302,
+      "step": 9643
+    },
+    {
+      "epoch": 1.6865136612021858,
+      "grad_norm": 1.0691546950743864,
+      "learning_rate": 8.47753503216774e-05,
+      "loss": 12.4429,
+      "step": 9644
+    },
+    {
+      "epoch": 1.686688524590164,
+      "grad_norm": 0.8516467008357832,
+      "learning_rate": 8.475668979799589e-05,
+      "loss": 12.317,
+      "step": 9645
+    },
+    {
+      "epoch": 1.686863387978142,
+      "grad_norm": 0.827448703704944,
+      "learning_rate": 8.47380298177206e-05,
+      "loss": 12.1797,
+      "step": 9646
+    },
+    {
+      "epoch": 1.6870382513661202,
+      "grad_norm": 0.8661566571177207,
+      "learning_rate": 8.471937038151674e-05,
+      "loss": 12.4284,
+      "step": 9647
+    },
+    {
+      "epoch": 1.6872131147540985,
+      "grad_norm": 0.8213897817083132,
+      "learning_rate": 8.470071149004944e-05,
+      "loss": 12.435,
+      "step": 9648
+    },
+    {
+      "epoch": 1.6873879781420764,
+      "grad_norm": 0.8580258067311062,
+      "learning_rate": 8.46820531439839e-05,
+      "loss": 12.4783,
+      "step": 9649
+    },
+    {
+      "epoch": 1.6875628415300548,
+      "grad_norm": 0.8485199712341666,
+      "learning_rate": 8.466339534398526e-05,
+      "loss": 12.4291,
+      "step": 9650
+    },
+    {
+      "epoch": 1.6877377049180327,
+      "grad_norm": 0.8588755306789766,
+      "learning_rate": 8.46447380907187e-05,
+      "loss": 12.4856,
+      "step": 9651
+    },
+    {
+      "epoch": 1.687912568306011,
+      "grad_norm": 0.945534968277794,
+      "learning_rate": 8.462608138484924e-05,
+      "loss": 12.3789,
+      "step": 9652
+    },
+    {
+      "epoch": 1.6880874316939891,
+      "grad_norm": 0.8779949644830021,
+      "learning_rate": 8.460742522704204e-05,
+      "loss": 12.5388,
+      "step": 9653
+    },
+    {
+      "epoch": 1.6882622950819672,
+      "grad_norm": 0.839380892081775,
+      "learning_rate": 8.458876961796214e-05,
+      "loss": 12.5203,
+      "step": 9654
+    },
+    {
+      "epoch": 1.6884371584699454,
+      "grad_norm": 0.8448927843065802,
+      "learning_rate": 8.45701145582746e-05,
+      "loss": 12.5048,
+      "step": 9655
+    },
+    {
+      "epoch": 1.6886120218579235,
+      "grad_norm": 0.7845470212464833,
+      "learning_rate": 8.455146004864443e-05,
+      "loss": 12.5055,
+      "step": 9656
+    },
+    {
+      "epoch": 1.6887868852459016,
+      "grad_norm": 0.7918469109157767,
+      "learning_rate": 8.453280608973666e-05,
+      "loss": 12.5001,
+      "step": 9657
+    },
+    {
+      "epoch": 1.6889617486338797,
+      "grad_norm": 0.8355868198510873,
+      "learning_rate": 8.451415268221633e-05,
+      "loss": 12.3044,
+      "step": 9658
+    },
+    {
+      "epoch": 1.689136612021858,
+      "grad_norm": 0.8535830340858777,
+      "learning_rate": 8.449549982674832e-05,
+      "loss": 12.4048,
+      "step": 9659
+    },
+    {
+      "epoch": 1.689311475409836,
+      "grad_norm": 0.9276216253897264,
+      "learning_rate": 8.447684752399763e-05,
+      "loss": 12.2217,
+      "step": 9660
+    },
+    {
+      "epoch": 1.6894863387978143,
+      "grad_norm": 0.785318568759971,
+      "learning_rate": 8.445819577462918e-05,
+      "loss": 12.4178,
+      "step": 9661
+    },
+    {
+      "epoch": 1.6896612021857922,
+      "grad_norm": 0.9369049291127156,
+      "learning_rate": 8.443954457930791e-05,
+      "loss": 12.1632,
+      "step": 9662
+    },
+    {
+      "epoch": 1.6898360655737705,
+      "grad_norm": 0.8372106312680911,
+      "learning_rate": 8.44208939386987e-05,
+      "loss": 12.4733,
+      "step": 9663
+    },
+    {
+      "epoch": 1.6900109289617486,
+      "grad_norm": 0.9661941250920283,
+      "learning_rate": 8.440224385346639e-05,
+      "loss": 12.5662,
+      "step": 9664
+    },
+    {
+      "epoch": 1.6901857923497268,
+      "grad_norm": 0.7995190984726264,
+      "learning_rate": 8.438359432427588e-05,
+      "loss": 12.4184,
+      "step": 9665
+    },
+    {
+      "epoch": 1.6903606557377049,
+      "grad_norm": 0.9243627978639564,
+      "learning_rate": 8.436494535179203e-05,
+      "loss": 12.4543,
+      "step": 9666
+    },
+    {
+      "epoch": 1.690535519125683,
+      "grad_norm": 0.8340795689487667,
+      "learning_rate": 8.434629693667954e-05,
+      "loss": 12.2575,
+      "step": 9667
+    },
+    {
+      "epoch": 1.6907103825136613,
+      "grad_norm": 0.9026972731431608,
+      "learning_rate": 8.43276490796033e-05,
+      "loss": 12.2582,
+      "step": 9668
+    },
+    {
+      "epoch": 1.6908852459016392,
+      "grad_norm": 1.0228500548275867,
+      "learning_rate": 8.43090017812281e-05,
+      "loss": 12.3556,
+      "step": 9669
+    },
+    {
+      "epoch": 1.6910601092896176,
+      "grad_norm": 0.8580223047256874,
+      "learning_rate": 8.42903550422186e-05,
+      "loss": 12.0977,
+      "step": 9670
+    },
+    {
+      "epoch": 1.6912349726775955,
+      "grad_norm": 0.9069998706130605,
+      "learning_rate": 8.427170886323961e-05,
+      "loss": 12.434,
+      "step": 9671
+    },
+    {
+      "epoch": 1.6914098360655738,
+      "grad_norm": 1.0839225293583898,
+      "learning_rate": 8.425306324495582e-05,
+      "loss": 12.3767,
+      "step": 9672
+    },
+    {
+      "epoch": 1.691584699453552,
+      "grad_norm": 0.852997214321001,
+      "learning_rate": 8.423441818803193e-05,
+      "loss": 12.4734,
+      "step": 9673
+    },
+    {
+      "epoch": 1.69175956284153,
+      "grad_norm": 1.1005314367766597,
+      "learning_rate": 8.42157736931326e-05,
+      "loss": 12.4675,
+      "step": 9674
+    },
+    {
+      "epoch": 1.6919344262295082,
+      "grad_norm": 0.8376374747430659,
+      "learning_rate": 8.419712976092249e-05,
+      "loss": 12.3228,
+      "step": 9675
+    },
+    {
+      "epoch": 1.6921092896174863,
+      "grad_norm": 0.8913363654785464,
+      "learning_rate": 8.417848639206624e-05,
+      "loss": 12.3119,
+      "step": 9676
+    },
+    {
+      "epoch": 1.6922841530054646,
+      "grad_norm": 0.9870932357374819,
+      "learning_rate": 8.41598435872285e-05,
+      "loss": 12.3406,
+      "step": 9677
+    },
+    {
+      "epoch": 1.6924590163934425,
+      "grad_norm": 0.8550621328336686,
+      "learning_rate": 8.41412013470738e-05,
+      "loss": 12.2255,
+      "step": 9678
+    },
+    {
+      "epoch": 1.6926338797814209,
+      "grad_norm": 0.8572370911636539,
+      "learning_rate": 8.412255967226675e-05,
+      "loss": 12.6729,
+      "step": 9679
+    },
+    {
+      "epoch": 1.6928087431693988,
+      "grad_norm": 0.890674958207373,
+      "learning_rate": 8.410391856347189e-05,
+      "loss": 12.3989,
+      "step": 9680
+    },
+    {
+      "epoch": 1.6929836065573771,
+      "grad_norm": 0.8950222174109327,
+      "learning_rate": 8.408527802135377e-05,
+      "loss": 12.5987,
+      "step": 9681
+    },
+    {
+      "epoch": 1.6931584699453552,
+      "grad_norm": 0.8853168969679093,
+      "learning_rate": 8.406663804657688e-05,
+      "loss": 12.5454,
+      "step": 9682
+    },
+    {
+      "epoch": 1.6933333333333334,
+      "grad_norm": 0.8100881046534579,
+      "learning_rate": 8.404799863980574e-05,
+      "loss": 12.5704,
+      "step": 9683
+    },
+    {
+      "epoch": 1.6935081967213115,
+      "grad_norm": 0.7644463187648582,
+      "learning_rate": 8.402935980170483e-05,
+      "loss": 11.9385,
+      "step": 9684
+    },
+    {
+      "epoch": 1.6936830601092896,
+      "grad_norm": 0.8646597456500006,
+      "learning_rate": 8.401072153293856e-05,
+      "loss": 12.6653,
+      "step": 9685
+    },
+    {
+      "epoch": 1.6938579234972677,
+      "grad_norm": 0.8692876153297349,
+      "learning_rate": 8.399208383417138e-05,
+      "loss": 12.1576,
+      "step": 9686
+    },
+    {
+      "epoch": 1.6940327868852458,
+      "grad_norm": 0.8979391840910621,
+      "learning_rate": 8.39734467060677e-05,
+      "loss": 12.5148,
+      "step": 9687
+    },
+    {
+      "epoch": 1.6942076502732242,
+      "grad_norm": 0.8771023243596415,
+      "learning_rate": 8.395481014929195e-05,
+      "loss": 12.2426,
+      "step": 9688
+    },
+    {
+      "epoch": 1.694382513661202,
+      "grad_norm": 0.9106802526037356,
+      "learning_rate": 8.393617416450845e-05,
+      "loss": 12.4457,
+      "step": 9689
+    },
+    {
+      "epoch": 1.6945573770491804,
+      "grad_norm": 1.0429772893699187,
+      "learning_rate": 8.391753875238158e-05,
+      "loss": 12.5829,
+      "step": 9690
+    },
+    {
+      "epoch": 1.6947322404371583,
+      "grad_norm": 0.7716643475077776,
+      "learning_rate": 8.389890391357564e-05,
+      "loss": 12.3476,
+      "step": 9691
+    },
+    {
+      "epoch": 1.6949071038251367,
+      "grad_norm": 1.1944163417244884,
+      "learning_rate": 8.388026964875503e-05,
+      "loss": 12.5024,
+      "step": 9692
+    },
+    {
+      "epoch": 1.6950819672131148,
+      "grad_norm": 0.8155541167727286,
+      "learning_rate": 8.386163595858392e-05,
+      "loss": 12.3197,
+      "step": 9693
+    },
+    {
+      "epoch": 1.695256830601093,
+      "grad_norm": 0.9368870022354092,
+      "learning_rate": 8.384300284372664e-05,
+      "loss": 12.4827,
+      "step": 9694
+    },
+    {
+      "epoch": 1.695431693989071,
+      "grad_norm": 0.9670226554868012,
+      "learning_rate": 8.382437030484742e-05,
+      "loss": 12.3915,
+      "step": 9695
+    },
+    {
+      "epoch": 1.6956065573770491,
+      "grad_norm": 0.864849935910566,
+      "learning_rate": 8.380573834261054e-05,
+      "loss": 12.3909,
+      "step": 9696
+    },
+    {
+      "epoch": 1.6957814207650275,
+      "grad_norm": 1.0534030721685128,
+      "learning_rate": 8.378710695768013e-05,
+      "loss": 12.369,
+      "step": 9697
+    },
+    {
+      "epoch": 1.6959562841530054,
+      "grad_norm": 0.9844468997560816,
+      "learning_rate": 8.376847615072042e-05,
+      "loss": 12.3643,
+      "step": 9698
+    },
+    {
+      "epoch": 1.6961311475409837,
+      "grad_norm": 0.8998962663196872,
+      "learning_rate": 8.374984592239558e-05,
+      "loss": 12.4232,
+      "step": 9699
+    },
+    {
+      "epoch": 1.6963060109289616,
+      "grad_norm": 1.118151174010567,
+      "learning_rate": 8.373121627336974e-05,
+      "loss": 12.5168,
+      "step": 9700
+    },
+    {
+      "epoch": 1.69648087431694,
+      "grad_norm": 0.8810347238214952,
+      "learning_rate": 8.371258720430702e-05,
+      "loss": 12.5981,
+      "step": 9701
+    },
+    {
+      "epoch": 1.696655737704918,
+      "grad_norm": 0.9259864786260742,
+      "learning_rate": 8.369395871587153e-05,
+      "loss": 12.3999,
+      "step": 9702
+    },
+    {
+      "epoch": 1.6968306010928962,
+      "grad_norm": 1.0350692631678338,
+      "learning_rate": 8.367533080872743e-05,
+      "loss": 12.5565,
+      "step": 9703
+    },
+    {
+      "epoch": 1.6970054644808743,
+      "grad_norm": 0.8553529492057262,
+      "learning_rate": 8.365670348353865e-05,
+      "loss": 12.3499,
+      "step": 9704
+    },
+    {
+      "epoch": 1.6971803278688524,
+      "grad_norm": 1.3062167493475174,
+      "learning_rate": 8.36380767409693e-05,
+      "loss": 12.5913,
+      "step": 9705
+    },
+    {
+      "epoch": 1.6973551912568308,
+      "grad_norm": 0.8434956774078605,
+      "learning_rate": 8.361945058168339e-05,
+      "loss": 12.3769,
+      "step": 9706
+    },
+    {
+      "epoch": 1.6975300546448087,
+      "grad_norm": 1.044677407141339,
+      "learning_rate": 8.360082500634496e-05,
+      "loss": 12.0818,
+      "step": 9707
+    },
+    {
+      "epoch": 1.697704918032787,
+      "grad_norm": 1.005379355912065,
+      "learning_rate": 8.358220001561793e-05,
+      "loss": 12.4562,
+      "step": 9708
+    },
+    {
+      "epoch": 1.697879781420765,
+      "grad_norm": 0.8332143903847055,
+      "learning_rate": 8.35635756101663e-05,
+      "loss": 12.1427,
+      "step": 9709
+    },
+    {
+      "epoch": 1.6980546448087432,
+      "grad_norm": 1.1081103222320987,
+      "learning_rate": 8.354495179065398e-05,
+      "loss": 12.5002,
+      "step": 9710
+    },
+    {
+      "epoch": 1.6982295081967214,
+      "grad_norm": 0.8539991226409985,
+      "learning_rate": 8.352632855774496e-05,
+      "loss": 12.4231,
+      "step": 9711
+    },
+    {
+      "epoch": 1.6984043715846995,
+      "grad_norm": 0.9673052511542999,
+      "learning_rate": 8.350770591210302e-05,
+      "loss": 12.4337,
+      "step": 9712
+    },
+    {
+      "epoch": 1.6985792349726776,
+      "grad_norm": 0.8911133684195757,
+      "learning_rate": 8.348908385439211e-05,
+      "loss": 12.3109,
+      "step": 9713
+    },
+    {
+      "epoch": 1.6987540983606557,
+      "grad_norm": 0.9879031446655476,
+      "learning_rate": 8.34704623852761e-05,
+      "loss": 12.317,
+      "step": 9714
+    },
+    {
+      "epoch": 1.6989289617486338,
+      "grad_norm": 0.9188835203377719,
+      "learning_rate": 8.345184150541876e-05,
+      "loss": 12.0505,
+      "step": 9715
+    },
+    {
+      "epoch": 1.699103825136612,
+      "grad_norm": 0.8288755752152905,
+      "learning_rate": 8.343322121548393e-05,
+      "loss": 12.2953,
+      "step": 9716
+    },
+    {
+      "epoch": 1.6992786885245903,
+      "grad_norm": 0.8400366282108314,
+      "learning_rate": 8.341460151613542e-05,
+      "loss": 12.3752,
+      "step": 9717
+    },
+    {
+      "epoch": 1.6994535519125682,
+      "grad_norm": 0.8286566137107336,
+      "learning_rate": 8.339598240803698e-05,
+      "loss": 12.4892,
+      "step": 9718
+    },
+    {
+      "epoch": 1.6996284153005465,
+      "grad_norm": 0.8127078679804278,
+      "learning_rate": 8.337736389185237e-05,
+      "loss": 12.3923,
+      "step": 9719
+    },
+    {
+      "epoch": 1.6998032786885244,
+      "grad_norm": 0.9465423838113034,
+      "learning_rate": 8.335874596824532e-05,
+      "loss": 12.5495,
+      "step": 9720
+    },
+    {
+      "epoch": 1.6999781420765028,
+      "grad_norm": 0.858665016578235,
+      "learning_rate": 8.334012863787952e-05,
+      "loss": 12.6384,
+      "step": 9721
+    },
+    {
+      "epoch": 1.700153005464481,
+      "grad_norm": 0.8439369115707926,
+      "learning_rate": 8.332151190141873e-05,
+      "loss": 12.523,
+      "step": 9722
+    },
+    {
+      "epoch": 1.700327868852459,
+      "grad_norm": 0.9331037891251578,
+      "learning_rate": 8.330289575952649e-05,
+      "loss": 12.5383,
+      "step": 9723
+    },
+    {
+      "epoch": 1.7005027322404371,
+      "grad_norm": 0.9046213745291241,
+      "learning_rate": 8.328428021286651e-05,
+      "loss": 12.6687,
+      "step": 9724
+    },
+    {
+      "epoch": 1.7006775956284153,
+      "grad_norm": 0.9532852473299519,
+      "learning_rate": 8.326566526210241e-05,
+      "loss": 12.2758,
+      "step": 9725
+    },
+    {
+      "epoch": 1.7008524590163936,
+      "grad_norm": 0.9800014409671435,
+      "learning_rate": 8.324705090789782e-05,
+      "loss": 12.5712,
+      "step": 9726
+    },
+    {
+      "epoch": 1.7010273224043715,
+      "grad_norm": 0.9558911878604928,
+      "learning_rate": 8.322843715091627e-05,
+      "loss": 12.449,
+      "step": 9727
+    },
+    {
+      "epoch": 1.7012021857923498,
+      "grad_norm": 1.0293949988503577,
+      "learning_rate": 8.320982399182134e-05,
+      "loss": 12.4792,
+      "step": 9728
+    },
+    {
+      "epoch": 1.7013770491803277,
+      "grad_norm": 0.8884035485495154,
+      "learning_rate": 8.319121143127658e-05,
+      "loss": 12.3039,
+      "step": 9729
+    },
+    {
+      "epoch": 1.701551912568306,
+      "grad_norm": 1.04675050859568,
+      "learning_rate": 8.317259946994551e-05,
+      "loss": 12.5472,
+      "step": 9730
+    },
+    {
+      "epoch": 1.7017267759562842,
+      "grad_norm": 0.9611162880751182,
+      "learning_rate": 8.315398810849158e-05,
+      "loss": 12.5324,
+      "step": 9731
+    },
+    {
+      "epoch": 1.7019016393442623,
+      "grad_norm": 1.0367760562669115,
+      "learning_rate": 8.313537734757828e-05,
+      "loss": 12.5555,
+      "step": 9732
+    },
+    {
+      "epoch": 1.7020765027322404,
+      "grad_norm": 1.1926753689609189,
+      "learning_rate": 8.311676718786911e-05,
+      "loss": 12.4145,
+      "step": 9733
+    },
+    {
+      "epoch": 1.7022513661202185,
+      "grad_norm": 0.8919266562816889,
+      "learning_rate": 8.309815763002744e-05,
+      "loss": 12.3939,
+      "step": 9734
+    },
+    {
+      "epoch": 1.7024262295081969,
+      "grad_norm": 1.037991106153887,
+      "learning_rate": 8.30795486747167e-05,
+      "loss": 12.1006,
+      "step": 9735
+    },
+    {
+      "epoch": 1.7026010928961748,
+      "grad_norm": 0.8639259968958813,
+      "learning_rate": 8.306094032260028e-05,
+      "loss": 12.2631,
+      "step": 9736
+    },
+    {
+      "epoch": 1.7027759562841531,
+      "grad_norm": 0.9292492953648239,
+      "learning_rate": 8.304233257434158e-05,
+      "loss": 12.2673,
+      "step": 9737
+    },
+    {
+      "epoch": 1.702950819672131,
+      "grad_norm": 0.9550916742357102,
+      "learning_rate": 8.302372543060389e-05,
+      "loss": 12.4325,
+      "step": 9738
+    },
+    {
+      "epoch": 1.7031256830601094,
+      "grad_norm": 0.841741625832111,
+      "learning_rate": 8.300511889205052e-05,
+      "loss": 12.3067,
+      "step": 9739
+    },
+    {
+      "epoch": 1.7033005464480875,
+      "grad_norm": 0.7994922129324903,
+      "learning_rate": 8.298651295934484e-05,
+      "loss": 12.2692,
+      "step": 9740
+    },
+    {
+      "epoch": 1.7034754098360656,
+      "grad_norm": 0.9871973993252253,
+      "learning_rate": 8.296790763315012e-05,
+      "loss": 12.3367,
+      "step": 9741
+    },
+    {
+      "epoch": 1.7036502732240437,
+      "grad_norm": 0.9159702213878206,
+      "learning_rate": 8.294930291412955e-05,
+      "loss": 12.4072,
+      "step": 9742
+    },
+    {
+      "epoch": 1.7038251366120218,
+      "grad_norm": 0.8722477047869988,
+      "learning_rate": 8.293069880294641e-05,
+      "loss": 12.6339,
+      "step": 9743
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.859171951689099,
+      "learning_rate": 8.291209530026392e-05,
+      "loss": 12.4197,
+      "step": 9744
+    },
+    {
+      "epoch": 1.704174863387978,
+      "grad_norm": 0.8060040114796714,
+      "learning_rate": 8.289349240674528e-05,
+      "loss": 12.2226,
+      "step": 9745
+    },
+    {
+      "epoch": 1.7043497267759564,
+      "grad_norm": 0.9052241521439028,
+      "learning_rate": 8.287489012305364e-05,
+      "loss": 12.5481,
+      "step": 9746
+    },
+    {
+      "epoch": 1.7045245901639343,
+      "grad_norm": 0.9445585311095407,
+      "learning_rate": 8.285628844985215e-05,
+      "loss": 12.5041,
+      "step": 9747
+    },
+    {
+      "epoch": 1.7046994535519127,
+      "grad_norm": 0.9848983073605247,
+      "learning_rate": 8.283768738780399e-05,
+      "loss": 12.4976,
+      "step": 9748
+    },
+    {
+      "epoch": 1.7048743169398906,
+      "grad_norm": 0.8936639559945521,
+      "learning_rate": 8.28190869375722e-05,
+      "loss": 12.3233,
+      "step": 9749
+    },
+    {
+      "epoch": 1.705049180327869,
+      "grad_norm": 0.8642266252605626,
+      "learning_rate": 8.280048709981985e-05,
+      "loss": 12.236,
+      "step": 9750
+    },
+    {
+      "epoch": 1.705224043715847,
+      "grad_norm": 0.9613945799246977,
+      "learning_rate": 8.278188787521006e-05,
+      "loss": 12.3174,
+      "step": 9751
+    },
+    {
+      "epoch": 1.7053989071038251,
+      "grad_norm": 0.9774537670993793,
+      "learning_rate": 8.276328926440587e-05,
+      "loss": 12.3296,
+      "step": 9752
+    },
+    {
+      "epoch": 1.7055737704918033,
+      "grad_norm": 0.7952742286892003,
+      "learning_rate": 8.274469126807026e-05,
+      "loss": 12.5072,
+      "step": 9753
+    },
+    {
+      "epoch": 1.7057486338797814,
+      "grad_norm": 0.8298950761111222,
+      "learning_rate": 8.272609388686624e-05,
+      "loss": 12.0161,
+      "step": 9754
+    },
+    {
+      "epoch": 1.7059234972677597,
+      "grad_norm": 0.9353524194867014,
+      "learning_rate": 8.270749712145679e-05,
+      "loss": 12.0611,
+      "step": 9755
+    },
+    {
+      "epoch": 1.7060983606557376,
+      "grad_norm": 0.9336043447244885,
+      "learning_rate": 8.268890097250491e-05,
+      "loss": 12.5007,
+      "step": 9756
+    },
+    {
+      "epoch": 1.706273224043716,
+      "grad_norm": 0.938369585154999,
+      "learning_rate": 8.267030544067345e-05,
+      "loss": 12.4921,
+      "step": 9757
+    },
+    {
+      "epoch": 1.7064480874316938,
+      "grad_norm": 1.0697316950244635,
+      "learning_rate": 8.265171052662533e-05,
+      "loss": 12.4945,
+      "step": 9758
+    },
+    {
+      "epoch": 1.7066229508196722,
+      "grad_norm": 0.848396709926546,
+      "learning_rate": 8.263311623102348e-05,
+      "loss": 12.1413,
+      "step": 9759
+    },
+    {
+      "epoch": 1.7067978142076503,
+      "grad_norm": 0.9594548050771615,
+      "learning_rate": 8.261452255453078e-05,
+      "loss": 12.4285,
+      "step": 9760
+    },
+    {
+      "epoch": 1.7069726775956284,
+      "grad_norm": 0.8530281745711932,
+      "learning_rate": 8.259592949781e-05,
+      "loss": 12.1631,
+      "step": 9761
+    },
+    {
+      "epoch": 1.7071475409836065,
+      "grad_norm": 1.0173429274799015,
+      "learning_rate": 8.2577337061524e-05,
+      "loss": 12.4177,
+      "step": 9762
+    },
+    {
+      "epoch": 1.7073224043715847,
+      "grad_norm": 0.9340275542782659,
+      "learning_rate": 8.25587452463356e-05,
+      "loss": 12.4298,
+      "step": 9763
+    },
+    {
+      "epoch": 1.7074972677595628,
+      "grad_norm": 0.8500487070014348,
+      "learning_rate": 8.254015405290751e-05,
+      "loss": 12.4585,
+      "step": 9764
+    },
+    {
+      "epoch": 1.707672131147541,
+      "grad_norm": 0.9225878314860328,
+      "learning_rate": 8.252156348190255e-05,
+      "loss": 12.2477,
+      "step": 9765
+    },
+    {
+      "epoch": 1.7078469945355192,
+      "grad_norm": 0.9917571006211306,
+      "learning_rate": 8.250297353398343e-05,
+      "loss": 12.4083,
+      "step": 9766
+    },
+    {
+      "epoch": 1.7080218579234971,
+      "grad_norm": 0.8720680724394299,
+      "learning_rate": 8.248438420981291e-05,
+      "loss": 12.3945,
+      "step": 9767
+    },
+    {
+      "epoch": 1.7081967213114755,
+      "grad_norm": 1.0378963450594805,
+      "learning_rate": 8.246579551005357e-05,
+      "loss": 12.21,
+      "step": 9768
+    },
+    {
+      "epoch": 1.7083715846994536,
+      "grad_norm": 0.8558970600678352,
+      "learning_rate": 8.244720743536814e-05,
+      "loss": 12.365,
+      "step": 9769
+    },
+    {
+      "epoch": 1.7085464480874317,
+      "grad_norm": 0.9850922004952328,
+      "learning_rate": 8.242861998641925e-05,
+      "loss": 12.4593,
+      "step": 9770
+    },
+    {
+      "epoch": 1.7087213114754098,
+      "grad_norm": 0.9653005627188489,
+      "learning_rate": 8.241003316386956e-05,
+      "loss": 12.4459,
+      "step": 9771
+    },
+    {
+      "epoch": 1.708896174863388,
+      "grad_norm": 0.7954428010470325,
+      "learning_rate": 8.239144696838161e-05,
+      "loss": 12.1666,
+      "step": 9772
+    },
+    {
+      "epoch": 1.709071038251366,
+      "grad_norm": 1.0545346006026335,
+      "learning_rate": 8.2372861400618e-05,
+      "loss": 12.5689,
+      "step": 9773
+    },
+    {
+      "epoch": 1.7092459016393442,
+      "grad_norm": 0.8709263736329381,
+      "learning_rate": 8.235427646124129e-05,
+      "loss": 12.2136,
+      "step": 9774
+    },
+    {
+      "epoch": 1.7094207650273225,
+      "grad_norm": 1.08872891162241,
+      "learning_rate": 8.233569215091405e-05,
+      "loss": 12.044,
+      "step": 9775
+    },
+    {
+      "epoch": 1.7095956284153004,
+      "grad_norm": 0.836560530986481,
+      "learning_rate": 8.231710847029869e-05,
+      "loss": 12.2044,
+      "step": 9776
+    },
+    {
+      "epoch": 1.7097704918032788,
+      "grad_norm": 0.9526806974630336,
+      "learning_rate": 8.229852542005777e-05,
+      "loss": 12.0505,
+      "step": 9777
+    },
+    {
+      "epoch": 1.7099453551912567,
+      "grad_norm": 0.9860075158037472,
+      "learning_rate": 8.227994300085376e-05,
+      "loss": 12.2786,
+      "step": 9778
+    },
+    {
+      "epoch": 1.710120218579235,
+      "grad_norm": 0.7960076456531503,
+      "learning_rate": 8.226136121334904e-05,
+      "loss": 12.2374,
+      "step": 9779
+    },
+    {
+      "epoch": 1.7102950819672131,
+      "grad_norm": 1.083373930048826,
+      "learning_rate": 8.224278005820609e-05,
+      "loss": 12.4575,
+      "step": 9780
+    },
+    {
+      "epoch": 1.7104699453551913,
+      "grad_norm": 0.9014254522968917,
+      "learning_rate": 8.222419953608727e-05,
+      "loss": 12.4429,
+      "step": 9781
+    },
+    {
+      "epoch": 1.7106448087431694,
+      "grad_norm": 0.8710717225198525,
+      "learning_rate": 8.220561964765501e-05,
+      "loss": 12.3555,
+      "step": 9782
+    },
+    {
+      "epoch": 1.7108196721311475,
+      "grad_norm": 0.9937798677170512,
+      "learning_rate": 8.218704039357156e-05,
+      "loss": 12.4193,
+      "step": 9783
+    },
+    {
+      "epoch": 1.7109945355191258,
+      "grad_norm": 0.8510487908021956,
+      "learning_rate": 8.216846177449932e-05,
+      "loss": 12.2407,
+      "step": 9784
+    },
+    {
+      "epoch": 1.7111693989071037,
+      "grad_norm": 1.011127696490448,
+      "learning_rate": 8.21498837911006e-05,
+      "loss": 12.3386,
+      "step": 9785
+    },
+    {
+      "epoch": 1.711344262295082,
+      "grad_norm": 0.9742681631377862,
+      "learning_rate": 8.21313064440377e-05,
+      "loss": 12.3166,
+      "step": 9786
+    },
+    {
+      "epoch": 1.71151912568306,
+      "grad_norm": 0.9385268743200814,
+      "learning_rate": 8.211272973397281e-05,
+      "loss": 12.3038,
+      "step": 9787
+    },
+    {
+      "epoch": 1.7116939890710383,
+      "grad_norm": 0.9900460928612158,
+      "learning_rate": 8.20941536615682e-05,
+      "loss": 12.4451,
+      "step": 9788
+    },
+    {
+      "epoch": 1.7118688524590164,
+      "grad_norm": 0.874093794651664,
+      "learning_rate": 8.207557822748608e-05,
+      "loss": 12.5902,
+      "step": 9789
+    },
+    {
+      "epoch": 1.7120437158469946,
+      "grad_norm": 0.9096831821801822,
+      "learning_rate": 8.20570034323887e-05,
+      "loss": 12.6054,
+      "step": 9790
+    },
+    {
+      "epoch": 1.7122185792349727,
+      "grad_norm": 0.9964176725484617,
+      "learning_rate": 8.203842927693815e-05,
+      "loss": 12.4949,
+      "step": 9791
+    },
+    {
+      "epoch": 1.7123934426229508,
+      "grad_norm": 0.8399987073684347,
+      "learning_rate": 8.20198557617966e-05,
+      "loss": 12.307,
+      "step": 9792
+    },
+    {
+      "epoch": 1.712568306010929,
+      "grad_norm": 0.9937056254100337,
+      "learning_rate": 8.200128288762624e-05,
+      "loss": 12.2427,
+      "step": 9793
+    },
+    {
+      "epoch": 1.712743169398907,
+      "grad_norm": 0.9327291406366751,
+      "learning_rate": 8.198271065508907e-05,
+      "loss": 12.5728,
+      "step": 9794
+    },
+    {
+      "epoch": 1.7129180327868854,
+      "grad_norm": 0.8996777550055025,
+      "learning_rate": 8.19641390648472e-05,
+      "loss": 12.4469,
+      "step": 9795
+    },
+    {
+      "epoch": 1.7130928961748633,
+      "grad_norm": 0.9023604605961693,
+      "learning_rate": 8.194556811756272e-05,
+      "loss": 12.4882,
+      "step": 9796
+    },
+    {
+      "epoch": 1.7132677595628416,
+      "grad_norm": 0.9822364158295833,
+      "learning_rate": 8.192699781389766e-05,
+      "loss": 12.5214,
+      "step": 9797
+    },
+    {
+      "epoch": 1.7134426229508197,
+      "grad_norm": 0.9685558413790663,
+      "learning_rate": 8.190842815451399e-05,
+      "loss": 12.5316,
+      "step": 9798
+    },
+    {
+      "epoch": 1.7136174863387978,
+      "grad_norm": 0.8884164847585488,
+      "learning_rate": 8.188985914007371e-05,
+      "loss": 12.4484,
+      "step": 9799
+    },
+    {
+      "epoch": 1.713792349726776,
+      "grad_norm": 0.7960617201796198,
+      "learning_rate": 8.187129077123881e-05,
+      "loss": 12.5238,
+      "step": 9800
+    },
+    {
+      "epoch": 1.713967213114754,
+      "grad_norm": 0.8731261161062576,
+      "learning_rate": 8.185272304867124e-05,
+      "loss": 12.2411,
+      "step": 9801
+    },
+    {
+      "epoch": 1.7141420765027322,
+      "grad_norm": 0.8781075630560602,
+      "learning_rate": 8.183415597303285e-05,
+      "loss": 12.5632,
+      "step": 9802
+    },
+    {
+      "epoch": 1.7143169398907103,
+      "grad_norm": 0.8658380362565137,
+      "learning_rate": 8.181558954498557e-05,
+      "loss": 12.5888,
+      "step": 9803
+    },
+    {
+      "epoch": 1.7144918032786887,
+      "grad_norm": 0.8191954257906889,
+      "learning_rate": 8.179702376519129e-05,
+      "loss": 12.3949,
+      "step": 9804
+    },
+    {
+      "epoch": 1.7146666666666666,
+      "grad_norm": 0.8599713024574686,
+      "learning_rate": 8.177845863431189e-05,
+      "loss": 12.3093,
+      "step": 9805
+    },
+    {
+      "epoch": 1.714841530054645,
+      "grad_norm": 0.9169096418088455,
+      "learning_rate": 8.175989415300908e-05,
+      "loss": 12.2478,
+      "step": 9806
+    },
+    {
+      "epoch": 1.7150163934426228,
+      "grad_norm": 0.9820932724851201,
+      "learning_rate": 8.174133032194475e-05,
+      "loss": 12.6216,
+      "step": 9807
+    },
+    {
+      "epoch": 1.7151912568306011,
+      "grad_norm": 0.8841498900998308,
+      "learning_rate": 8.172276714178068e-05,
+      "loss": 12.5057,
+      "step": 9808
+    },
+    {
+      "epoch": 1.7153661202185793,
+      "grad_norm": 0.8337865025341601,
+      "learning_rate": 8.17042046131786e-05,
+      "loss": 12.4992,
+      "step": 9809
+    },
+    {
+      "epoch": 1.7155409836065574,
+      "grad_norm": 0.8314225410346628,
+      "learning_rate": 8.168564273680024e-05,
+      "loss": 12.392,
+      "step": 9810
+    },
+    {
+      "epoch": 1.7157158469945355,
+      "grad_norm": 0.8846699523680919,
+      "learning_rate": 8.166708151330731e-05,
+      "loss": 12.5034,
+      "step": 9811
+    },
+    {
+      "epoch": 1.7158907103825136,
+      "grad_norm": 0.8410482066741652,
+      "learning_rate": 8.164852094336156e-05,
+      "loss": 12.4191,
+      "step": 9812
+    },
+    {
+      "epoch": 1.716065573770492,
+      "grad_norm": 0.9739821465530538,
+      "learning_rate": 8.162996102762455e-05,
+      "loss": 12.2317,
+      "step": 9813
+    },
+    {
+      "epoch": 1.7162404371584699,
+      "grad_norm": 0.8361367049529421,
+      "learning_rate": 8.161140176675795e-05,
+      "loss": 12.4012,
+      "step": 9814
+    },
+    {
+      "epoch": 1.7164153005464482,
+      "grad_norm": 0.8636009205013571,
+      "learning_rate": 8.159284316142341e-05,
+      "loss": 12.4715,
+      "step": 9815
+    },
+    {
+      "epoch": 1.716590163934426,
+      "grad_norm": 0.9219168680038073,
+      "learning_rate": 8.157428521228252e-05,
+      "loss": 12.1483,
+      "step": 9816
+    },
+    {
+      "epoch": 1.7167650273224044,
+      "grad_norm": 0.8120292534863384,
+      "learning_rate": 8.155572791999682e-05,
+      "loss": 12.0038,
+      "step": 9817
+    },
+    {
+      "epoch": 1.7169398907103826,
+      "grad_norm": 0.8512319408004001,
+      "learning_rate": 8.153717128522787e-05,
+      "loss": 12.4882,
+      "step": 9818
+    },
+    {
+      "epoch": 1.7171147540983607,
+      "grad_norm": 0.8955366127997854,
+      "learning_rate": 8.151861530863719e-05,
+      "loss": 12.5042,
+      "step": 9819
+    },
+    {
+      "epoch": 1.7172896174863388,
+      "grad_norm": 0.8260534794373441,
+      "learning_rate": 8.15000599908863e-05,
+      "loss": 12.5907,
+      "step": 9820
+    },
+    {
+      "epoch": 1.717464480874317,
+      "grad_norm": 0.8342486594827818,
+      "learning_rate": 8.148150533263664e-05,
+      "loss": 12.3704,
+      "step": 9821
+    },
+    {
+      "epoch": 1.717639344262295,
+      "grad_norm": 0.8829072779802116,
+      "learning_rate": 8.146295133454967e-05,
+      "loss": 12.213,
+      "step": 9822
+    },
+    {
+      "epoch": 1.7178142076502732,
+      "grad_norm": 0.8973420604312676,
+      "learning_rate": 8.144439799728685e-05,
+      "loss": 12.4031,
+      "step": 9823
+    },
+    {
+      "epoch": 1.7179890710382515,
+      "grad_norm": 0.9050180033546156,
+      "learning_rate": 8.142584532150954e-05,
+      "loss": 12.5729,
+      "step": 9824
+    },
+    {
+      "epoch": 1.7181639344262294,
+      "grad_norm": 0.8695254049056623,
+      "learning_rate": 8.140729330787915e-05,
+      "loss": 12.24,
+      "step": 9825
+    },
+    {
+      "epoch": 1.7183387978142077,
+      "grad_norm": 0.9718708896447065,
+      "learning_rate": 8.138874195705702e-05,
+      "loss": 12.4713,
+      "step": 9826
+    },
+    {
+      "epoch": 1.7185136612021856,
+      "grad_norm": 0.8256793110331689,
+      "learning_rate": 8.137019126970454e-05,
+      "loss": 12.4445,
+      "step": 9827
+    },
+    {
+      "epoch": 1.718688524590164,
+      "grad_norm": 0.9525717748455899,
+      "learning_rate": 8.135164124648293e-05,
+      "loss": 12.4463,
+      "step": 9828
+    },
+    {
+      "epoch": 1.718863387978142,
+      "grad_norm": 0.9993877299666701,
+      "learning_rate": 8.133309188805352e-05,
+      "loss": 12.3333,
+      "step": 9829
+    },
+    {
+      "epoch": 1.7190382513661202,
+      "grad_norm": 0.9121969380540185,
+      "learning_rate": 8.131454319507758e-05,
+      "loss": 12.5828,
+      "step": 9830
+    },
+    {
+      "epoch": 1.7192131147540983,
+      "grad_norm": 1.081499372023399,
+      "learning_rate": 8.129599516821639e-05,
+      "loss": 12.4156,
+      "step": 9831
+    },
+    {
+      "epoch": 1.7193879781420764,
+      "grad_norm": 0.8011405498380547,
+      "learning_rate": 8.127744780813107e-05,
+      "loss": 12.4883,
+      "step": 9832
+    },
+    {
+      "epoch": 1.7195628415300548,
+      "grad_norm": 1.1086666280430568,
+      "learning_rate": 8.125890111548286e-05,
+      "loss": 12.4437,
+      "step": 9833
+    },
+    {
+      "epoch": 1.7197377049180327,
+      "grad_norm": 0.8218628231211554,
+      "learning_rate": 8.124035509093292e-05,
+      "loss": 12.4497,
+      "step": 9834
+    },
+    {
+      "epoch": 1.719912568306011,
+      "grad_norm": 0.8837678684372642,
+      "learning_rate": 8.122180973514244e-05,
+      "loss": 12.3753,
+      "step": 9835
+    },
+    {
+      "epoch": 1.720087431693989,
+      "grad_norm": 0.8976111274835056,
+      "learning_rate": 8.120326504877248e-05,
+      "loss": 12.2845,
+      "step": 9836
+    },
+    {
+      "epoch": 1.7202622950819673,
+      "grad_norm": 0.7946618009538426,
+      "learning_rate": 8.118472103248415e-05,
+      "loss": 12.2975,
+      "step": 9837
+    },
+    {
+      "epoch": 1.7204371584699454,
+      "grad_norm": 0.9221363189569418,
+      "learning_rate": 8.116617768693856e-05,
+      "loss": 12.397,
+      "step": 9838
+    },
+    {
+      "epoch": 1.7206120218579235,
+      "grad_norm": 0.9019681199623205,
+      "learning_rate": 8.114763501279673e-05,
+      "loss": 12.2691,
+      "step": 9839
+    },
+    {
+      "epoch": 1.7207868852459016,
+      "grad_norm": 0.9379489994704956,
+      "learning_rate": 8.112909301071967e-05,
+      "loss": 12.1004,
+      "step": 9840
+    },
+    {
+      "epoch": 1.7209617486338797,
+      "grad_norm": 0.8432634306751471,
+      "learning_rate": 8.111055168136838e-05,
+      "loss": 12.5039,
+      "step": 9841
+    },
+    {
+      "epoch": 1.721136612021858,
+      "grad_norm": 0.8713257405805582,
+      "learning_rate": 8.109201102540389e-05,
+      "loss": 12.3799,
+      "step": 9842
+    },
+    {
+      "epoch": 1.721311475409836,
+      "grad_norm": 0.9036374455824125,
+      "learning_rate": 8.107347104348708e-05,
+      "loss": 12.3598,
+      "step": 9843
+    },
+    {
+      "epoch": 1.7214863387978143,
+      "grad_norm": 0.8368452887087084,
+      "learning_rate": 8.105493173627892e-05,
+      "loss": 12.4052,
+      "step": 9844
+    },
+    {
+      "epoch": 1.7216612021857922,
+      "grad_norm": 0.7385425586581146,
+      "learning_rate": 8.103639310444032e-05,
+      "loss": 12.3564,
+      "step": 9845
+    },
+    {
+      "epoch": 1.7218360655737706,
+      "grad_norm": 0.830102759668923,
+      "learning_rate": 8.101785514863218e-05,
+      "loss": 12.6235,
+      "step": 9846
+    },
+    {
+      "epoch": 1.7220109289617487,
+      "grad_norm": 0.9770540482313803,
+      "learning_rate": 8.099931786951528e-05,
+      "loss": 12.2675,
+      "step": 9847
+    },
+    {
+      "epoch": 1.7221857923497268,
+      "grad_norm": 0.8481157640456408,
+      "learning_rate": 8.09807812677505e-05,
+      "loss": 12.1058,
+      "step": 9848
+    },
+    {
+      "epoch": 1.722360655737705,
+      "grad_norm": 0.806002291941742,
+      "learning_rate": 8.096224534399865e-05,
+      "loss": 12.496,
+      "step": 9849
+    },
+    {
+      "epoch": 1.722535519125683,
+      "grad_norm": 0.8616167293145633,
+      "learning_rate": 8.094371009892054e-05,
+      "loss": 12.3505,
+      "step": 9850
+    },
+    {
+      "epoch": 1.7227103825136612,
+      "grad_norm": 0.8462330005367146,
+      "learning_rate": 8.092517553317688e-05,
+      "loss": 12.1726,
+      "step": 9851
+    },
+    {
+      "epoch": 1.7228852459016393,
+      "grad_norm": 0.8774009401003662,
+      "learning_rate": 8.09066416474284e-05,
+      "loss": 12.38,
+      "step": 9852
+    },
+    {
+      "epoch": 1.7230601092896176,
+      "grad_norm": 0.8396182337774254,
+      "learning_rate": 8.088810844233587e-05,
+      "loss": 12.4711,
+      "step": 9853
+    },
+    {
+      "epoch": 1.7232349726775955,
+      "grad_norm": 0.8061055058991128,
+      "learning_rate": 8.086957591855994e-05,
+      "loss": 12.3072,
+      "step": 9854
+    },
+    {
+      "epoch": 1.7234098360655739,
+      "grad_norm": 0.8607728618651385,
+      "learning_rate": 8.085104407676126e-05,
+      "loss": 12.4626,
+      "step": 9855
+    },
+    {
+      "epoch": 1.7235846994535517,
+      "grad_norm": 0.8740977778880837,
+      "learning_rate": 8.083251291760048e-05,
+      "loss": 12.5089,
+      "step": 9856
+    },
+    {
+      "epoch": 1.72375956284153,
+      "grad_norm": 0.8183729033147088,
+      "learning_rate": 8.081398244173827e-05,
+      "loss": 12.6228,
+      "step": 9857
+    },
+    {
+      "epoch": 1.7239344262295082,
+      "grad_norm": 0.8209207499999868,
+      "learning_rate": 8.079545264983514e-05,
+      "loss": 12.4272,
+      "step": 9858
+    },
+    {
+      "epoch": 1.7241092896174863,
+      "grad_norm": 0.8300815232355327,
+      "learning_rate": 8.077692354255167e-05,
+      "loss": 12.2358,
+      "step": 9859
+    },
+    {
+      "epoch": 1.7242841530054644,
+      "grad_norm": 0.8187940986760461,
+      "learning_rate": 8.075839512054841e-05,
+      "loss": 12.5706,
+      "step": 9860
+    },
+    {
+      "epoch": 1.7244590163934426,
+      "grad_norm": 0.8417236132832854,
+      "learning_rate": 8.073986738448592e-05,
+      "loss": 12.3521,
+      "step": 9861
+    },
+    {
+      "epoch": 1.724633879781421,
+      "grad_norm": 0.850525452659404,
+      "learning_rate": 8.072134033502464e-05,
+      "loss": 12.3671,
+      "step": 9862
+    },
+    {
+      "epoch": 1.7248087431693988,
+      "grad_norm": 0.8986905188589351,
+      "learning_rate": 8.070281397282502e-05,
+      "loss": 12.3904,
+      "step": 9863
+    },
+    {
+      "epoch": 1.7249836065573771,
+      "grad_norm": 0.9984258530854577,
+      "learning_rate": 8.068428829854755e-05,
+      "loss": 12.347,
+      "step": 9864
+    },
+    {
+      "epoch": 1.725158469945355,
+      "grad_norm": 0.8134140202463028,
+      "learning_rate": 8.066576331285269e-05,
+      "loss": 12.4448,
+      "step": 9865
+    },
+    {
+      "epoch": 1.7253333333333334,
+      "grad_norm": 0.9416077036038696,
+      "learning_rate": 8.064723901640072e-05,
+      "loss": 12.5402,
+      "step": 9866
+    },
+    {
+      "epoch": 1.7255081967213115,
+      "grad_norm": 0.9305488675591077,
+      "learning_rate": 8.062871540985207e-05,
+      "loss": 12.1564,
+      "step": 9867
+    },
+    {
+      "epoch": 1.7256830601092896,
+      "grad_norm": 0.8154106590798668,
+      "learning_rate": 8.06101924938671e-05,
+      "loss": 12.6406,
+      "step": 9868
+    },
+    {
+      "epoch": 1.7258579234972677,
+      "grad_norm": 1.0183452922426235,
+      "learning_rate": 8.059167026910612e-05,
+      "loss": 12.2745,
+      "step": 9869
+    },
+    {
+      "epoch": 1.7260327868852459,
+      "grad_norm": 0.9575126638153983,
+      "learning_rate": 8.057314873622939e-05,
+      "loss": 12.2169,
+      "step": 9870
+    },
+    {
+      "epoch": 1.7262076502732242,
+      "grad_norm": 0.8122121381089652,
+      "learning_rate": 8.055462789589721e-05,
+      "loss": 12.2315,
+      "step": 9871
+    },
+    {
+      "epoch": 1.726382513661202,
+      "grad_norm": 1.0503140426801925,
+      "learning_rate": 8.053610774876988e-05,
+      "loss": 12.3672,
+      "step": 9872
+    },
+    {
+      "epoch": 1.7265573770491804,
+      "grad_norm": 0.934143384054797,
+      "learning_rate": 8.051758829550753e-05,
+      "loss": 12.4164,
+      "step": 9873
+    },
+    {
+      "epoch": 1.7267322404371583,
+      "grad_norm": 0.9376944875550816,
+      "learning_rate": 8.049906953677037e-05,
+      "loss": 12.346,
+      "step": 9874
+    },
+    {
+      "epoch": 1.7269071038251367,
+      "grad_norm": 1.0537618385056644,
+      "learning_rate": 8.048055147321863e-05,
+      "loss": 12.6032,
+      "step": 9875
+    },
+    {
+      "epoch": 1.7270819672131148,
+      "grad_norm": 0.852281786790957,
+      "learning_rate": 8.046203410551244e-05,
+      "loss": 12.4758,
+      "step": 9876
+    },
+    {
+      "epoch": 1.727256830601093,
+      "grad_norm": 1.080952497911354,
+      "learning_rate": 8.044351743431187e-05,
+      "loss": 12.2229,
+      "step": 9877
+    },
+    {
+      "epoch": 1.727431693989071,
+      "grad_norm": 0.8011948391622095,
+      "learning_rate": 8.042500146027707e-05,
+      "loss": 12.3121,
+      "step": 9878
+    },
+    {
+      "epoch": 1.7276065573770492,
+      "grad_norm": 0.8966080269777857,
+      "learning_rate": 8.040648618406808e-05,
+      "loss": 12.4306,
+      "step": 9879
+    },
+    {
+      "epoch": 1.7277814207650273,
+      "grad_norm": 0.9115543846625433,
+      "learning_rate": 8.0387971606345e-05,
+      "loss": 12.3162,
+      "step": 9880
+    },
+    {
+      "epoch": 1.7279562841530054,
+      "grad_norm": 0.8376664881259674,
+      "learning_rate": 8.036945772776777e-05,
+      "loss": 12.3034,
+      "step": 9881
+    },
+    {
+      "epoch": 1.7281311475409837,
+      "grad_norm": 0.8847402365774073,
+      "learning_rate": 8.035094454899646e-05,
+      "loss": 12.5366,
+      "step": 9882
+    },
+    {
+      "epoch": 1.7283060109289616,
+      "grad_norm": 0.8823721703383288,
+      "learning_rate": 8.033243207069101e-05,
+      "loss": 12.3907,
+      "step": 9883
+    },
+    {
+      "epoch": 1.72848087431694,
+      "grad_norm": 0.8990877561543723,
+      "learning_rate": 8.031392029351143e-05,
+      "loss": 12.5539,
+      "step": 9884
+    },
+    {
+      "epoch": 1.7286557377049179,
+      "grad_norm": 0.9139489398857075,
+      "learning_rate": 8.029540921811754e-05,
+      "loss": 12.3293,
+      "step": 9885
+    },
+    {
+      "epoch": 1.7288306010928962,
+      "grad_norm": 0.8696102798674582,
+      "learning_rate": 8.027689884516929e-05,
+      "loss": 12.4578,
+      "step": 9886
+    },
+    {
+      "epoch": 1.7290054644808743,
+      "grad_norm": 0.8506600761278358,
+      "learning_rate": 8.025838917532657e-05,
+      "loss": 12.5192,
+      "step": 9887
+    },
+    {
+      "epoch": 1.7291803278688525,
+      "grad_norm": 0.9452221848812178,
+      "learning_rate": 8.02398802092492e-05,
+      "loss": 12.0965,
+      "step": 9888
+    },
+    {
+      "epoch": 1.7293551912568306,
+      "grad_norm": 0.8401834884412872,
+      "learning_rate": 8.0221371947597e-05,
+      "loss": 12.4375,
+      "step": 9889
+    },
+    {
+      "epoch": 1.7295300546448087,
+      "grad_norm": 0.9909146810485099,
+      "learning_rate": 8.020286439102978e-05,
+      "loss": 12.4704,
+      "step": 9890
+    },
+    {
+      "epoch": 1.729704918032787,
+      "grad_norm": 0.8564746214514465,
+      "learning_rate": 8.018435754020736e-05,
+      "loss": 12.3875,
+      "step": 9891
+    },
+    {
+      "epoch": 1.729879781420765,
+      "grad_norm": 0.8558041699027128,
+      "learning_rate": 8.01658513957894e-05,
+      "loss": 12.5948,
+      "step": 9892
+    },
+    {
+      "epoch": 1.7300546448087433,
+      "grad_norm": 0.9710200190988577,
+      "learning_rate": 8.014734595843568e-05,
+      "loss": 12.2247,
+      "step": 9893
+    },
+    {
+      "epoch": 1.7302295081967212,
+      "grad_norm": 0.8741579580488155,
+      "learning_rate": 8.012884122880586e-05,
+      "loss": 12.3113,
+      "step": 9894
+    },
+    {
+      "epoch": 1.7304043715846995,
+      "grad_norm": 0.8798533836059111,
+      "learning_rate": 8.011033720755969e-05,
+      "loss": 12.3168,
+      "step": 9895
+    },
+    {
+      "epoch": 1.7305792349726776,
+      "grad_norm": 0.8820306733076453,
+      "learning_rate": 8.00918338953567e-05,
+      "loss": 12.3697,
+      "step": 9896
+    },
+    {
+      "epoch": 1.7307540983606557,
+      "grad_norm": 0.9067551686856967,
+      "learning_rate": 8.00733312928566e-05,
+      "loss": 12.4171,
+      "step": 9897
+    },
+    {
+      "epoch": 1.7309289617486339,
+      "grad_norm": 0.8123105143890113,
+      "learning_rate": 8.005482940071894e-05,
+      "loss": 12.2462,
+      "step": 9898
+    },
+    {
+      "epoch": 1.731103825136612,
+      "grad_norm": 0.8770920744798041,
+      "learning_rate": 8.003632821960332e-05,
+      "loss": 12.5522,
+      "step": 9899
+    },
+    {
+      "epoch": 1.7312786885245903,
+      "grad_norm": 0.912305671472167,
+      "learning_rate": 8.001782775016926e-05,
+      "loss": 12.32,
+      "step": 9900
+    },
+    {
+      "epoch": 1.7314535519125682,
+      "grad_norm": 0.8727895229540586,
+      "learning_rate": 7.999932799307631e-05,
+      "loss": 12.3683,
+      "step": 9901
+    },
+    {
+      "epoch": 1.7316284153005466,
+      "grad_norm": 0.9114316783517343,
+      "learning_rate": 7.998082894898398e-05,
+      "loss": 12.4466,
+      "step": 9902
+    },
+    {
+      "epoch": 1.7318032786885245,
+      "grad_norm": 0.8891353332560631,
+      "learning_rate": 7.996233061855165e-05,
+      "loss": 12.1787,
+      "step": 9903
+    },
+    {
+      "epoch": 1.7319781420765028,
+      "grad_norm": 0.8119964729581549,
+      "learning_rate": 7.994383300243884e-05,
+      "loss": 12.3373,
+      "step": 9904
+    },
+    {
+      "epoch": 1.732153005464481,
+      "grad_norm": 0.9138961730964142,
+      "learning_rate": 7.992533610130495e-05,
+      "loss": 12.4668,
+      "step": 9905
+    },
+    {
+      "epoch": 1.732327868852459,
+      "grad_norm": 0.8987553269460932,
+      "learning_rate": 7.990683991580938e-05,
+      "loss": 12.3604,
+      "step": 9906
+    },
+    {
+      "epoch": 1.7325027322404372,
+      "grad_norm": 0.826559368344819,
+      "learning_rate": 7.988834444661148e-05,
+      "loss": 12.2812,
+      "step": 9907
+    },
+    {
+      "epoch": 1.7326775956284153,
+      "grad_norm": 0.8746397107786374,
+      "learning_rate": 7.986984969437061e-05,
+      "loss": 12.4545,
+      "step": 9908
+    },
+    {
+      "epoch": 1.7328524590163934,
+      "grad_norm": 0.8301336341872637,
+      "learning_rate": 7.985135565974606e-05,
+      "loss": 12.3228,
+      "step": 9909
+    },
+    {
+      "epoch": 1.7330273224043715,
+      "grad_norm": 0.9001027510190436,
+      "learning_rate": 7.98328623433972e-05,
+      "loss": 12.6015,
+      "step": 9910
+    },
+    {
+      "epoch": 1.7332021857923499,
+      "grad_norm": 0.8853280575754922,
+      "learning_rate": 7.981436974598319e-05,
+      "loss": 12.2933,
+      "step": 9911
+    },
+    {
+      "epoch": 1.7333770491803278,
+      "grad_norm": 0.8019088541609968,
+      "learning_rate": 7.97958778681633e-05,
+      "loss": 12.17,
+      "step": 9912
+    },
+    {
+      "epoch": 1.733551912568306,
+      "grad_norm": 0.9563778936185866,
+      "learning_rate": 7.977738671059678e-05,
+      "loss": 12.3686,
+      "step": 9913
+    },
+    {
+      "epoch": 1.733726775956284,
+      "grad_norm": 0.8304005748885537,
+      "learning_rate": 7.975889627394282e-05,
+      "loss": 12.3021,
+      "step": 9914
+    },
+    {
+      "epoch": 1.7339016393442623,
+      "grad_norm": 0.8547618468464427,
+      "learning_rate": 7.974040655886053e-05,
+      "loss": 12.2551,
+      "step": 9915
+    },
+    {
+      "epoch": 1.7340765027322405,
+      "grad_norm": 0.9469896897521157,
+      "learning_rate": 7.972191756600906e-05,
+      "loss": 12.26,
+      "step": 9916
+    },
+    {
+      "epoch": 1.7342513661202186,
+      "grad_norm": 0.8965100121125897,
+      "learning_rate": 7.970342929604757e-05,
+      "loss": 12.298,
+      "step": 9917
+    },
+    {
+      "epoch": 1.7344262295081967,
+      "grad_norm": 0.8550816173744045,
+      "learning_rate": 7.968494174963508e-05,
+      "loss": 12.3165,
+      "step": 9918
+    },
+    {
+      "epoch": 1.7346010928961748,
+      "grad_norm": 0.8743106451941778,
+      "learning_rate": 7.96664549274307e-05,
+      "loss": 12.4719,
+      "step": 9919
+    },
+    {
+      "epoch": 1.7347759562841532,
+      "grad_norm": 1.0014303906573605,
+      "learning_rate": 7.964796883009343e-05,
+      "loss": 12.4471,
+      "step": 9920
+    },
+    {
+      "epoch": 1.734950819672131,
+      "grad_norm": 0.9109559368098225,
+      "learning_rate": 7.962948345828232e-05,
+      "loss": 12.4823,
+      "step": 9921
+    },
+    {
+      "epoch": 1.7351256830601094,
+      "grad_norm": 0.8334476136363769,
+      "learning_rate": 7.96109988126563e-05,
+      "loss": 12.3344,
+      "step": 9922
+    },
+    {
+      "epoch": 1.7353005464480873,
+      "grad_norm": 0.9250888928516039,
+      "learning_rate": 7.959251489387436e-05,
+      "loss": 12.4815,
+      "step": 9923
+    },
+    {
+      "epoch": 1.7354754098360656,
+      "grad_norm": 0.8263132183088572,
+      "learning_rate": 7.95740317025954e-05,
+      "loss": 12.2944,
+      "step": 9924
+    },
+    {
+      "epoch": 1.7356502732240437,
+      "grad_norm": 0.8687609600833712,
+      "learning_rate": 7.955554923947838e-05,
+      "loss": 12.539,
+      "step": 9925
+    },
+    {
+      "epoch": 1.7358251366120219,
+      "grad_norm": 0.908373725116514,
+      "learning_rate": 7.953706750518212e-05,
+      "loss": 12.585,
+      "step": 9926
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 0.8780356509661779,
+      "learning_rate": 7.951858650036549e-05,
+      "loss": 12.3419,
+      "step": 9927
+    },
+    {
+      "epoch": 1.736174863387978,
+      "grad_norm": 0.823199800087204,
+      "learning_rate": 7.950010622568733e-05,
+      "loss": 12.602,
+      "step": 9928
+    },
+    {
+      "epoch": 1.7363497267759564,
+      "grad_norm": 0.9555173315269105,
+      "learning_rate": 7.948162668180646e-05,
+      "loss": 12.4623,
+      "step": 9929
+    },
+    {
+      "epoch": 1.7365245901639343,
+      "grad_norm": 0.8116630486758316,
+      "learning_rate": 7.946314786938158e-05,
+      "loss": 12.3028,
+      "step": 9930
+    },
+    {
+      "epoch": 1.7366994535519127,
+      "grad_norm": 0.9752439560654673,
+      "learning_rate": 7.94446697890715e-05,
+      "loss": 12.3991,
+      "step": 9931
+    },
+    {
+      "epoch": 1.7368743169398906,
+      "grad_norm": 0.8709619098573846,
+      "learning_rate": 7.942619244153494e-05,
+      "loss": 12.4732,
+      "step": 9932
+    },
+    {
+      "epoch": 1.737049180327869,
+      "grad_norm": 0.9235615711958589,
+      "learning_rate": 7.940771582743057e-05,
+      "loss": 12.5866,
+      "step": 9933
+    },
+    {
+      "epoch": 1.737224043715847,
+      "grad_norm": 0.8454726341111284,
+      "learning_rate": 7.938923994741705e-05,
+      "loss": 12.4991,
+      "step": 9934
+    },
+    {
+      "epoch": 1.7373989071038252,
+      "grad_norm": 0.7969335912692073,
+      "learning_rate": 7.937076480215307e-05,
+      "loss": 12.5383,
+      "step": 9935
+    },
+    {
+      "epoch": 1.7375737704918033,
+      "grad_norm": 1.0503538379266484,
+      "learning_rate": 7.935229039229724e-05,
+      "loss": 12.3334,
+      "step": 9936
+    },
+    {
+      "epoch": 1.7377486338797814,
+      "grad_norm": 0.8309186395024852,
+      "learning_rate": 7.93338167185081e-05,
+      "loss": 12.3397,
+      "step": 9937
+    },
+    {
+      "epoch": 1.7379234972677595,
+      "grad_norm": 0.9322778514421374,
+      "learning_rate": 7.931534378144425e-05,
+      "loss": 12.5247,
+      "step": 9938
+    },
+    {
+      "epoch": 1.7380983606557376,
+      "grad_norm": 0.9253259811629766,
+      "learning_rate": 7.929687158176423e-05,
+      "loss": 12.281,
+      "step": 9939
+    },
+    {
+      "epoch": 1.738273224043716,
+      "grad_norm": 0.8188511704457806,
+      "learning_rate": 7.927840012012659e-05,
+      "loss": 12.2753,
+      "step": 9940
+    },
+    {
+      "epoch": 1.7384480874316939,
+      "grad_norm": 0.9358891859552572,
+      "learning_rate": 7.925992939718974e-05,
+      "loss": 12.2375,
+      "step": 9941
+    },
+    {
+      "epoch": 1.7386229508196722,
+      "grad_norm": 0.9120599353243241,
+      "learning_rate": 7.924145941361216e-05,
+      "loss": 12.1878,
+      "step": 9942
+    },
+    {
+      "epoch": 1.7387978142076501,
+      "grad_norm": 0.9101669560734833,
+      "learning_rate": 7.92229901700523e-05,
+      "loss": 12.3587,
+      "step": 9943
+    },
+    {
+      "epoch": 1.7389726775956285,
+      "grad_norm": 0.990003897900018,
+      "learning_rate": 7.920452166716858e-05,
+      "loss": 12.4174,
+      "step": 9944
+    },
+    {
+      "epoch": 1.7391475409836066,
+      "grad_norm": 0.81034716919191,
+      "learning_rate": 7.918605390561937e-05,
+      "loss": 12.4507,
+      "step": 9945
+    },
+    {
+      "epoch": 1.7393224043715847,
+      "grad_norm": 0.9060163174753216,
+      "learning_rate": 7.9167586886063e-05,
+      "loss": 12.3062,
+      "step": 9946
+    },
+    {
+      "epoch": 1.7394972677595628,
+      "grad_norm": 0.9367897564218024,
+      "learning_rate": 7.914912060915782e-05,
+      "loss": 12.2478,
+      "step": 9947
+    },
+    {
+      "epoch": 1.739672131147541,
+      "grad_norm": 0.94305821101593,
+      "learning_rate": 7.913065507556218e-05,
+      "loss": 12.5712,
+      "step": 9948
+    },
+    {
+      "epoch": 1.7398469945355193,
+      "grad_norm": 0.8575620130514269,
+      "learning_rate": 7.911219028593425e-05,
+      "loss": 12.3815,
+      "step": 9949
+    },
+    {
+      "epoch": 1.7400218579234972,
+      "grad_norm": 0.8373990577620696,
+      "learning_rate": 7.909372624093234e-05,
+      "loss": 12.5367,
+      "step": 9950
+    },
+    {
+      "epoch": 1.7401967213114755,
+      "grad_norm": 0.8447276554314496,
+      "learning_rate": 7.907526294121468e-05,
+      "loss": 12.3377,
+      "step": 9951
+    },
+    {
+      "epoch": 1.7403715846994534,
+      "grad_norm": 0.8592060464838016,
+      "learning_rate": 7.905680038743943e-05,
+      "loss": 12.4984,
+      "step": 9952
+    },
+    {
+      "epoch": 1.7405464480874318,
+      "grad_norm": 0.8332040020228689,
+      "learning_rate": 7.903833858026477e-05,
+      "loss": 12.425,
+      "step": 9953
+    },
+    {
+      "epoch": 1.7407213114754099,
+      "grad_norm": 0.8533113209173488,
+      "learning_rate": 7.901987752034885e-05,
+      "loss": 12.2454,
+      "step": 9954
+    },
+    {
+      "epoch": 1.740896174863388,
+      "grad_norm": 0.959980046541722,
+      "learning_rate": 7.900141720834984e-05,
+      "loss": 12.24,
+      "step": 9955
+    },
+    {
+      "epoch": 1.741071038251366,
+      "grad_norm": 0.8137404196552177,
+      "learning_rate": 7.898295764492572e-05,
+      "loss": 12.3617,
+      "step": 9956
+    },
+    {
+      "epoch": 1.7412459016393442,
+      "grad_norm": 0.9259224968579802,
+      "learning_rate": 7.896449883073458e-05,
+      "loss": 12.342,
+      "step": 9957
+    },
+    {
+      "epoch": 1.7414207650273226,
+      "grad_norm": 0.8984119952656182,
+      "learning_rate": 7.894604076643451e-05,
+      "loss": 12.3761,
+      "step": 9958
+    },
+    {
+      "epoch": 1.7415956284153005,
+      "grad_norm": 0.8541842985368908,
+      "learning_rate": 7.892758345268353e-05,
+      "loss": 12.4625,
+      "step": 9959
+    },
+    {
+      "epoch": 1.7417704918032788,
+      "grad_norm": 1.0274239390890036,
+      "learning_rate": 7.890912689013952e-05,
+      "loss": 12.0801,
+      "step": 9960
+    },
+    {
+      "epoch": 1.7419453551912567,
+      "grad_norm": 0.8954776595476818,
+      "learning_rate": 7.889067107946051e-05,
+      "loss": 12.108,
+      "step": 9961
+    },
+    {
+      "epoch": 1.742120218579235,
+      "grad_norm": 0.8661717384405677,
+      "learning_rate": 7.887221602130441e-05,
+      "loss": 12.3595,
+      "step": 9962
+    },
+    {
+      "epoch": 1.7422950819672132,
+      "grad_norm": 0.9481858560928419,
+      "learning_rate": 7.885376171632913e-05,
+      "loss": 12.3587,
+      "step": 9963
+    },
+    {
+      "epoch": 1.7424699453551913,
+      "grad_norm": 0.8844819208987719,
+      "learning_rate": 7.883530816519253e-05,
+      "loss": 12.2293,
+      "step": 9964
+    },
+    {
+      "epoch": 1.7426448087431694,
+      "grad_norm": 0.8518039761618845,
+      "learning_rate": 7.881685536855246e-05,
+      "loss": 12.507,
+      "step": 9965
+    },
+    {
+      "epoch": 1.7428196721311475,
+      "grad_norm": 1.0054691126549598,
+      "learning_rate": 7.87984033270668e-05,
+      "loss": 12.4337,
+      "step": 9966
+    },
+    {
+      "epoch": 1.7429945355191256,
+      "grad_norm": 0.810888081513747,
+      "learning_rate": 7.877995204139323e-05,
+      "loss": 12.5026,
+      "step": 9967
+    },
+    {
+      "epoch": 1.7431693989071038,
+      "grad_norm": 1.1807637154712298,
+      "learning_rate": 7.876150151218959e-05,
+      "loss": 12.4756,
+      "step": 9968
+    },
+    {
+      "epoch": 1.743344262295082,
+      "grad_norm": 0.9017182788330572,
+      "learning_rate": 7.87430517401136e-05,
+      "loss": 12.4758,
+      "step": 9969
+    },
+    {
+      "epoch": 1.74351912568306,
+      "grad_norm": 0.8410354176175338,
+      "learning_rate": 7.872460272582301e-05,
+      "loss": 12.1672,
+      "step": 9970
+    },
+    {
+      "epoch": 1.7436939890710383,
+      "grad_norm": 0.8688796126317997,
+      "learning_rate": 7.870615446997544e-05,
+      "loss": 12.0362,
+      "step": 9971
+    },
+    {
+      "epoch": 1.7438688524590162,
+      "grad_norm": 0.8532190168483099,
+      "learning_rate": 7.86877069732286e-05,
+      "loss": 12.2069,
+      "step": 9972
+    },
+    {
+      "epoch": 1.7440437158469946,
+      "grad_norm": 0.8976956626275584,
+      "learning_rate": 7.866926023624011e-05,
+      "loss": 12.4991,
+      "step": 9973
+    },
+    {
+      "epoch": 1.7442185792349727,
+      "grad_norm": 1.0298203921627656,
+      "learning_rate": 7.865081425966759e-05,
+      "loss": 12.481,
+      "step": 9974
+    },
+    {
+      "epoch": 1.7443934426229508,
+      "grad_norm": 0.8127952204825583,
+      "learning_rate": 7.863236904416856e-05,
+      "loss": 12.4591,
+      "step": 9975
+    },
+    {
+      "epoch": 1.744568306010929,
+      "grad_norm": 1.060002351668258,
+      "learning_rate": 7.861392459040062e-05,
+      "loss": 12.3791,
+      "step": 9976
+    },
+    {
+      "epoch": 1.744743169398907,
+      "grad_norm": 0.9781502264970059,
+      "learning_rate": 7.859548089902126e-05,
+      "loss": 12.2695,
+      "step": 9977
+    },
+    {
+      "epoch": 1.7449180327868854,
+      "grad_norm": 0.9378035282498857,
+      "learning_rate": 7.857703797068805e-05,
+      "loss": 12.2875,
+      "step": 9978
+    },
+    {
+      "epoch": 1.7450928961748633,
+      "grad_norm": 0.926121338901127,
+      "learning_rate": 7.855859580605837e-05,
+      "loss": 12.3304,
+      "step": 9979
+    },
+    {
+      "epoch": 1.7452677595628416,
+      "grad_norm": 0.8531065104144323,
+      "learning_rate": 7.854015440578968e-05,
+      "loss": 12.3522,
+      "step": 9980
+    },
+    {
+      "epoch": 1.7454426229508195,
+      "grad_norm": 0.8932862665591524,
+      "learning_rate": 7.852171377053949e-05,
+      "loss": 12.2603,
+      "step": 9981
+    },
+    {
+      "epoch": 1.7456174863387979,
+      "grad_norm": 0.9523452050276265,
+      "learning_rate": 7.850327390096504e-05,
+      "loss": 12.2361,
+      "step": 9982
+    },
+    {
+      "epoch": 1.745792349726776,
+      "grad_norm": 0.7965272652265383,
+      "learning_rate": 7.848483479772378e-05,
+      "loss": 12.465,
+      "step": 9983
+    },
+    {
+      "epoch": 1.7459672131147541,
+      "grad_norm": 0.9412569940349029,
+      "learning_rate": 7.8466396461473e-05,
+      "loss": 12.1879,
+      "step": 9984
+    },
+    {
+      "epoch": 1.7461420765027322,
+      "grad_norm": 0.8399470705097324,
+      "learning_rate": 7.844795889287007e-05,
+      "loss": 11.925,
+      "step": 9985
+    },
+    {
+      "epoch": 1.7463169398907104,
+      "grad_norm": 0.9145698239357601,
+      "learning_rate": 7.84295220925722e-05,
+      "loss": 12.1971,
+      "step": 9986
+    },
+    {
+      "epoch": 1.7464918032786885,
+      "grad_norm": 0.9653318211990837,
+      "learning_rate": 7.841108606123665e-05,
+      "loss": 12.1304,
+      "step": 9987
+    },
+    {
+      "epoch": 1.7466666666666666,
+      "grad_norm": 0.7909926952956304,
+      "learning_rate": 7.839265079952067e-05,
+      "loss": 12.3177,
+      "step": 9988
+    },
+    {
+      "epoch": 1.746841530054645,
+      "grad_norm": 0.8658200999810512,
+      "learning_rate": 7.837421630808146e-05,
+      "loss": 12.4285,
+      "step": 9989
+    },
+    {
+      "epoch": 1.7470163934426228,
+      "grad_norm": 0.853466435268171,
+      "learning_rate": 7.835578258757614e-05,
+      "loss": 12.3877,
+      "step": 9990
+    },
+    {
+      "epoch": 1.7471912568306012,
+      "grad_norm": 1.055026000148489,
+      "learning_rate": 7.83373496386619e-05,
+      "loss": 12.5071,
+      "step": 9991
+    },
+    {
+      "epoch": 1.7473661202185793,
+      "grad_norm": 0.9233528811440364,
+      "learning_rate": 7.831891746199582e-05,
+      "loss": 12.412,
+      "step": 9992
+    },
+    {
+      "epoch": 1.7475409836065574,
+      "grad_norm": 0.8141451559590042,
+      "learning_rate": 7.830048605823506e-05,
+      "loss": 12.3773,
+      "step": 9993
+    },
+    {
+      "epoch": 1.7477158469945355,
+      "grad_norm": 0.9460340665680821,
+      "learning_rate": 7.828205542803656e-05,
+      "loss": 12.3753,
+      "step": 9994
+    },
+    {
+      "epoch": 1.7478907103825136,
+      "grad_norm": 0.9305826445241819,
+      "learning_rate": 7.82636255720574e-05,
+      "loss": 12.1713,
+      "step": 9995
+    },
+    {
+      "epoch": 1.7480655737704918,
+      "grad_norm": 0.8787450246766436,
+      "learning_rate": 7.824519649095464e-05,
+      "loss": 12.5947,
+      "step": 9996
+    },
+    {
+      "epoch": 1.7482404371584699,
+      "grad_norm": 1.007439383257801,
+      "learning_rate": 7.822676818538517e-05,
+      "loss": 12.0603,
+      "step": 9997
+    },
+    {
+      "epoch": 1.7484153005464482,
+      "grad_norm": 0.8131623964070317,
+      "learning_rate": 7.820834065600599e-05,
+      "loss": 12.1959,
+      "step": 9998
+    },
+    {
+      "epoch": 1.7485901639344261,
+      "grad_norm": 1.0423970160423672,
+      "learning_rate": 7.818991390347398e-05,
+      "loss": 11.9978,
+      "step": 9999
+    },
+    {
+      "epoch": 1.7487650273224045,
+      "grad_norm": 0.8823624654815809,
+      "learning_rate": 7.817148792844612e-05,
+      "loss": 12.4823,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7489398907103824,
+      "grad_norm": 0.8800264044734495,
+      "learning_rate": 7.815306273157916e-05,
+      "loss": 12.1814,
+      "step": 10001
+    },
+    {
+      "epoch": 1.7491147540983607,
+      "grad_norm": 0.8610582572046467,
+      "learning_rate": 7.813463831353e-05,
+      "loss": 12.3135,
+      "step": 10002
+    },
+    {
+      "epoch": 1.7492896174863388,
+      "grad_norm": 0.8476815445897343,
+      "learning_rate": 7.811621467495542e-05,
+      "loss": 12.4792,
+      "step": 10003
+    },
+    {
+      "epoch": 1.749464480874317,
+      "grad_norm": 0.8050581351262762,
+      "learning_rate": 7.809779181651228e-05,
+      "loss": 12.1703,
+      "step": 10004
+    },
+    {
+      "epoch": 1.749639344262295,
+      "grad_norm": 0.7927298455211145,
+      "learning_rate": 7.807936973885722e-05,
+      "loss": 12.4474,
+      "step": 10005
+    },
+    {
+      "epoch": 1.7498142076502732,
+      "grad_norm": 0.9204431016214613,
+      "learning_rate": 7.8060948442647e-05,
+      "loss": 12.4976,
+      "step": 10006
+    },
+    {
+      "epoch": 1.7499890710382515,
+      "grad_norm": 0.8076122586475984,
+      "learning_rate": 7.804252792853835e-05,
+      "loss": 12.2986,
+      "step": 10007
+    },
+    {
+      "epoch": 1.7501639344262294,
+      "grad_norm": 0.8520070433721578,
+      "learning_rate": 7.802410819718795e-05,
+      "loss": 12.2932,
+      "step": 10008
+    },
+    {
+      "epoch": 1.7503387978142078,
+      "grad_norm": 0.752768417492086,
+      "learning_rate": 7.80056892492524e-05,
+      "loss": 11.956,
+      "step": 10009
+    },
+    {
+      "epoch": 1.7505136612021857,
+      "grad_norm": 0.7665717704483362,
+      "learning_rate": 7.798727108538832e-05,
+      "loss": 12.352,
+      "step": 10010
+    },
+    {
+      "epoch": 1.750688524590164,
+      "grad_norm": 0.8484156721874053,
+      "learning_rate": 7.796885370625235e-05,
+      "loss": 12.2933,
+      "step": 10011
+    },
+    {
+      "epoch": 1.7508633879781421,
+      "grad_norm": 0.8536111440759431,
+      "learning_rate": 7.795043711250096e-05,
+      "loss": 12.3302,
+      "step": 10012
+    },
+    {
+      "epoch": 1.7510382513661202,
+      "grad_norm": 0.8340810026642778,
+      "learning_rate": 7.793202130479073e-05,
+      "loss": 12.3003,
+      "step": 10013
+    },
+    {
+      "epoch": 1.7512131147540984,
+      "grad_norm": 0.843980531430175,
+      "learning_rate": 7.791360628377816e-05,
+      "loss": 12.2867,
+      "step": 10014
+    },
+    {
+      "epoch": 1.7513879781420765,
+      "grad_norm": 0.8881119021865044,
+      "learning_rate": 7.789519205011973e-05,
+      "loss": 12.5358,
+      "step": 10015
+    },
+    {
+      "epoch": 1.7515628415300546,
+      "grad_norm": 0.8276428684543649,
+      "learning_rate": 7.787677860447187e-05,
+      "loss": 12.2594,
+      "step": 10016
+    },
+    {
+      "epoch": 1.7517377049180327,
+      "grad_norm": 0.8677352709997171,
+      "learning_rate": 7.785836594749099e-05,
+      "loss": 12.5492,
+      "step": 10017
+    },
+    {
+      "epoch": 1.751912568306011,
+      "grad_norm": 0.7868619801864761,
+      "learning_rate": 7.783995407983351e-05,
+      "loss": 12.3722,
+      "step": 10018
+    },
+    {
+      "epoch": 1.752087431693989,
+      "grad_norm": 0.8075899668660831,
+      "learning_rate": 7.782154300215581e-05,
+      "loss": 12.5,
+      "step": 10019
+    },
+    {
+      "epoch": 1.7522622950819673,
+      "grad_norm": 0.813567460736643,
+      "learning_rate": 7.780313271511414e-05,
+      "loss": 12.3065,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7524371584699452,
+      "grad_norm": 0.8917353401588598,
+      "learning_rate": 7.778472321936487e-05,
+      "loss": 12.4042,
+      "step": 10021
+    },
+    {
+      "epoch": 1.7526120218579235,
+      "grad_norm": 0.8344904765161062,
+      "learning_rate": 7.776631451556425e-05,
+      "loss": 12.4048,
+      "step": 10022
+    },
+    {
+      "epoch": 1.7527868852459016,
+      "grad_norm": 0.8294535500314978,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 12.4272,
+      "step": 10023
+    },
+    {
+      "epoch": 1.7529617486338798,
+      "grad_norm": 0.838571066012496,
+      "learning_rate": 7.7729499486434e-05,
+      "loss": 12.4223,
+      "step": 10024
+    },
+    {
+      "epoch": 1.7531366120218579,
+      "grad_norm": 0.8738789141433841,
+      "learning_rate": 7.771109316241673e-05,
+      "loss": 12.3677,
+      "step": 10025
+    },
+    {
+      "epoch": 1.753311475409836,
+      "grad_norm": 0.8447274313026758,
+      "learning_rate": 7.769268763297301e-05,
+      "loss": 12.0749,
+      "step": 10026
+    },
+    {
+      "epoch": 1.7534863387978143,
+      "grad_norm": 0.8688117814373041,
+      "learning_rate": 7.767428289875888e-05,
+      "loss": 12.3335,
+      "step": 10027
+    },
+    {
+      "epoch": 1.7536612021857922,
+      "grad_norm": 0.8911426646868257,
+      "learning_rate": 7.765587896043045e-05,
+      "loss": 12.3432,
+      "step": 10028
+    },
+    {
+      "epoch": 1.7538360655737706,
+      "grad_norm": 0.8752620443635036,
+      "learning_rate": 7.763747581864386e-05,
+      "loss": 12.4575,
+      "step": 10029
+    },
+    {
+      "epoch": 1.7540109289617485,
+      "grad_norm": 0.9084345732305614,
+      "learning_rate": 7.761907347405514e-05,
+      "loss": 12.1018,
+      "step": 10030
+    },
+    {
+      "epoch": 1.7541857923497268,
+      "grad_norm": 0.8177820776273991,
+      "learning_rate": 7.760067192732029e-05,
+      "loss": 12.2085,
+      "step": 10031
+    },
+    {
+      "epoch": 1.754360655737705,
+      "grad_norm": 0.921527473518992,
+      "learning_rate": 7.75822711790953e-05,
+      "loss": 12.3096,
+      "step": 10032
+    },
+    {
+      "epoch": 1.754535519125683,
+      "grad_norm": 0.9614503930746846,
+      "learning_rate": 7.756387123003616e-05,
+      "loss": 12.3755,
+      "step": 10033
+    },
+    {
+      "epoch": 1.7547103825136612,
+      "grad_norm": 1.0476682187510593,
+      "learning_rate": 7.75454720807988e-05,
+      "loss": 12.335,
+      "step": 10034
+    },
+    {
+      "epoch": 1.7548852459016393,
+      "grad_norm": 1.0321380384523062,
+      "learning_rate": 7.752707373203912e-05,
+      "loss": 12.1038,
+      "step": 10035
+    },
+    {
+      "epoch": 1.7550601092896176,
+      "grad_norm": 0.9079831938345465,
+      "learning_rate": 7.7508676184413e-05,
+      "loss": 12.4597,
+      "step": 10036
+    },
+    {
+      "epoch": 1.7552349726775955,
+      "grad_norm": 1.0220515531971595,
+      "learning_rate": 7.749027943857629e-05,
+      "loss": 12.4346,
+      "step": 10037
+    },
+    {
+      "epoch": 1.7554098360655739,
+      "grad_norm": 0.9415902658763914,
+      "learning_rate": 7.747188349518487e-05,
+      "loss": 12.2377,
+      "step": 10038
+    },
+    {
+      "epoch": 1.7555846994535518,
+      "grad_norm": 0.990918806214673,
+      "learning_rate": 7.745348835489442e-05,
+      "loss": 12.5513,
+      "step": 10039
+    },
+    {
+      "epoch": 1.7557595628415301,
+      "grad_norm": 1.0837896281876453,
+      "learning_rate": 7.743509401836078e-05,
+      "loss": 12.578,
+      "step": 10040
+    },
+    {
+      "epoch": 1.7559344262295082,
+      "grad_norm": 0.8356894799165906,
+      "learning_rate": 7.74167004862397e-05,
+      "loss": 12.3654,
+      "step": 10041
+    },
+    {
+      "epoch": 1.7561092896174864,
+      "grad_norm": 1.0681467556426523,
+      "learning_rate": 7.739830775918683e-05,
+      "loss": 12.3523,
+      "step": 10042
+    },
+    {
+      "epoch": 1.7562841530054645,
+      "grad_norm": 0.9621518113723072,
+      "learning_rate": 7.73799158378579e-05,
+      "loss": 12.3708,
+      "step": 10043
+    },
+    {
+      "epoch": 1.7564590163934426,
+      "grad_norm": 0.8648770054286784,
+      "learning_rate": 7.736152472290854e-05,
+      "loss": 12.3059,
+      "step": 10044
+    },
+    {
+      "epoch": 1.7566338797814207,
+      "grad_norm": 0.9280693002313251,
+      "learning_rate": 7.734313441499439e-05,
+      "loss": 12.3177,
+      "step": 10045
+    },
+    {
+      "epoch": 1.7568087431693988,
+      "grad_norm": 1.0740435399460044,
+      "learning_rate": 7.7324744914771e-05,
+      "loss": 12.0463,
+      "step": 10046
+    },
+    {
+      "epoch": 1.7569836065573772,
+      "grad_norm": 0.9134294373697178,
+      "learning_rate": 7.730635622289396e-05,
+      "loss": 12.4546,
+      "step": 10047
+    },
+    {
+      "epoch": 1.757158469945355,
+      "grad_norm": 0.8785630468464015,
+      "learning_rate": 7.728796834001881e-05,
+      "loss": 12.1482,
+      "step": 10048
+    },
+    {
+      "epoch": 1.7573333333333334,
+      "grad_norm": 0.8129393437079206,
+      "learning_rate": 7.726958126680108e-05,
+      "loss": 12.2748,
+      "step": 10049
+    },
+    {
+      "epoch": 1.7575081967213113,
+      "grad_norm": 0.897104079046457,
+      "learning_rate": 7.725119500389619e-05,
+      "loss": 12.3142,
+      "step": 10050
+    },
+    {
+      "epoch": 1.7576830601092897,
+      "grad_norm": 0.8455780209353997,
+      "learning_rate": 7.723280955195963e-05,
+      "loss": 11.8317,
+      "step": 10051
+    },
+    {
+      "epoch": 1.7578579234972678,
+      "grad_norm": 0.9256653487845473,
+      "learning_rate": 7.721442491164679e-05,
+      "loss": 12.1582,
+      "step": 10052
+    },
+    {
+      "epoch": 1.758032786885246,
+      "grad_norm": 0.9309121343868234,
+      "learning_rate": 7.719604108361311e-05,
+      "loss": 12.3305,
+      "step": 10053
+    },
+    {
+      "epoch": 1.758207650273224,
+      "grad_norm": 0.945364017997954,
+      "learning_rate": 7.717765806851389e-05,
+      "loss": 12.1265,
+      "step": 10054
+    },
+    {
+      "epoch": 1.7583825136612021,
+      "grad_norm": 0.799796884371419,
+      "learning_rate": 7.71592758670045e-05,
+      "loss": 12.2417,
+      "step": 10055
+    },
+    {
+      "epoch": 1.7585573770491805,
+      "grad_norm": 0.8718919656414126,
+      "learning_rate": 7.714089447974026e-05,
+      "loss": 12.2932,
+      "step": 10056
+    },
+    {
+      "epoch": 1.7587322404371584,
+      "grad_norm": 0.8489954292131163,
+      "learning_rate": 7.712251390737644e-05,
+      "loss": 12.2653,
+      "step": 10057
+    },
+    {
+      "epoch": 1.7589071038251367,
+      "grad_norm": 0.8562915736267475,
+      "learning_rate": 7.710413415056824e-05,
+      "loss": 12.4061,
+      "step": 10058
+    },
+    {
+      "epoch": 1.7590819672131146,
+      "grad_norm": 0.8094739139212522,
+      "learning_rate": 7.70857552099709e-05,
+      "loss": 12.2751,
+      "step": 10059
+    },
+    {
+      "epoch": 1.759256830601093,
+      "grad_norm": 0.8490399441988866,
+      "learning_rate": 7.706737708623963e-05,
+      "loss": 12.3901,
+      "step": 10060
+    },
+    {
+      "epoch": 1.759431693989071,
+      "grad_norm": 0.8306482690047966,
+      "learning_rate": 7.704899978002957e-05,
+      "loss": 12.2889,
+      "step": 10061
+    },
+    {
+      "epoch": 1.7596065573770492,
+      "grad_norm": 0.8583623826183472,
+      "learning_rate": 7.703062329199584e-05,
+      "loss": 12.1431,
+      "step": 10062
+    },
+    {
+      "epoch": 1.7597814207650273,
+      "grad_norm": 0.8209387053904033,
+      "learning_rate": 7.701224762279355e-05,
+      "loss": 12.4191,
+      "step": 10063
+    },
+    {
+      "epoch": 1.7599562841530054,
+      "grad_norm": 0.8637847031447371,
+      "learning_rate": 7.699387277307782e-05,
+      "loss": 12.2061,
+      "step": 10064
+    },
+    {
+      "epoch": 1.7601311475409838,
+      "grad_norm": 0.798011239459263,
+      "learning_rate": 7.697549874350361e-05,
+      "loss": 12.3052,
+      "step": 10065
+    },
+    {
+      "epoch": 1.7603060109289617,
+      "grad_norm": 0.7765346174266636,
+      "learning_rate": 7.695712553472595e-05,
+      "loss": 12.2247,
+      "step": 10066
+    },
+    {
+      "epoch": 1.76048087431694,
+      "grad_norm": 0.809975158597081,
+      "learning_rate": 7.693875314739986e-05,
+      "loss": 12.2145,
+      "step": 10067
+    },
+    {
+      "epoch": 1.760655737704918,
+      "grad_norm": 0.85667642794853,
+      "learning_rate": 7.69203815821803e-05,
+      "loss": 12.3625,
+      "step": 10068
+    },
+    {
+      "epoch": 1.7608306010928962,
+      "grad_norm": 0.8282760925989751,
+      "learning_rate": 7.690201083972214e-05,
+      "loss": 12.4264,
+      "step": 10069
+    },
+    {
+      "epoch": 1.7610054644808744,
+      "grad_norm": 0.817588488806022,
+      "learning_rate": 7.688364092068031e-05,
+      "loss": 12.2751,
+      "step": 10070
+    },
+    {
+      "epoch": 1.7611803278688525,
+      "grad_norm": 0.9028314576446479,
+      "learning_rate": 7.686527182570968e-05,
+      "loss": 12.248,
+      "step": 10071
+    },
+    {
+      "epoch": 1.7613551912568306,
+      "grad_norm": 0.8353881433016068,
+      "learning_rate": 7.684690355546511e-05,
+      "loss": 12.1144,
+      "step": 10072
+    },
+    {
+      "epoch": 1.7615300546448087,
+      "grad_norm": 0.7937154077858155,
+      "learning_rate": 7.682853611060134e-05,
+      "loss": 12.2797,
+      "step": 10073
+    },
+    {
+      "epoch": 1.7617049180327868,
+      "grad_norm": 0.8452124521495025,
+      "learning_rate": 7.681016949177317e-05,
+      "loss": 12.1479,
+      "step": 10074
+    },
+    {
+      "epoch": 1.761879781420765,
+      "grad_norm": 0.8362795032525148,
+      "learning_rate": 7.679180369963543e-05,
+      "loss": 12.4311,
+      "step": 10075
+    },
+    {
+      "epoch": 1.7620546448087433,
+      "grad_norm": 0.8909946425206584,
+      "learning_rate": 7.67734387348427e-05,
+      "loss": 11.9433,
+      "step": 10076
+    },
+    {
+      "epoch": 1.7622295081967212,
+      "grad_norm": 0.7670535020894903,
+      "learning_rate": 7.675507459804978e-05,
+      "loss": 12.2527,
+      "step": 10077
+    },
+    {
+      "epoch": 1.7624043715846995,
+      "grad_norm": 0.8836759053493778,
+      "learning_rate": 7.673671128991125e-05,
+      "loss": 12.4003,
+      "step": 10078
+    },
+    {
+      "epoch": 1.7625792349726774,
+      "grad_norm": 0.821890173562785,
+      "learning_rate": 7.671834881108185e-05,
+      "loss": 12.2492,
+      "step": 10079
+    },
+    {
+      "epoch": 1.7627540983606558,
+      "grad_norm": 0.900083328460929,
+      "learning_rate": 7.669998716221604e-05,
+      "loss": 12.3214,
+      "step": 10080
+    },
+    {
+      "epoch": 1.762928961748634,
+      "grad_norm": 0.8385260378970179,
+      "learning_rate": 7.66816263439685e-05,
+      "loss": 12.327,
+      "step": 10081
+    },
+    {
+      "epoch": 1.763103825136612,
+      "grad_norm": 0.8459857247938632,
+      "learning_rate": 7.666326635699372e-05,
+      "loss": 12.3833,
+      "step": 10082
+    },
+    {
+      "epoch": 1.7632786885245901,
+      "grad_norm": 0.9238380815463983,
+      "learning_rate": 7.664490720194629e-05,
+      "loss": 12.3476,
+      "step": 10083
+    },
+    {
+      "epoch": 1.7634535519125683,
+      "grad_norm": 0.8059380712027975,
+      "learning_rate": 7.662654887948057e-05,
+      "loss": 12.3635,
+      "step": 10084
+    },
+    {
+      "epoch": 1.7636284153005466,
+      "grad_norm": 0.8516113749990566,
+      "learning_rate": 7.660819139025108e-05,
+      "loss": 12.2735,
+      "step": 10085
+    },
+    {
+      "epoch": 1.7638032786885245,
+      "grad_norm": 0.8404452037373568,
+      "learning_rate": 7.658983473491224e-05,
+      "loss": 12.27,
+      "step": 10086
+    },
+    {
+      "epoch": 1.7639781420765028,
+      "grad_norm": 0.8310779191850326,
+      "learning_rate": 7.657147891411845e-05,
+      "loss": 12.3558,
+      "step": 10087
+    },
+    {
+      "epoch": 1.7641530054644807,
+      "grad_norm": 0.8535009879169967,
+      "learning_rate": 7.655312392852406e-05,
+      "loss": 12.2863,
+      "step": 10088
+    },
+    {
+      "epoch": 1.764327868852459,
+      "grad_norm": 0.8424989106871978,
+      "learning_rate": 7.65347697787834e-05,
+      "loss": 12.2843,
+      "step": 10089
+    },
+    {
+      "epoch": 1.7645027322404372,
+      "grad_norm": 0.8474617989436589,
+      "learning_rate": 7.651641646555082e-05,
+      "loss": 12.3905,
+      "step": 10090
+    },
+    {
+      "epoch": 1.7646775956284153,
+      "grad_norm": 0.8201353308213163,
+      "learning_rate": 7.64980639894805e-05,
+      "loss": 12.3562,
+      "step": 10091
+    },
+    {
+      "epoch": 1.7648524590163934,
+      "grad_norm": 0.8165750684439111,
+      "learning_rate": 7.647971235122676e-05,
+      "loss": 12.436,
+      "step": 10092
+    },
+    {
+      "epoch": 1.7650273224043715,
+      "grad_norm": 0.8377354645897709,
+      "learning_rate": 7.646136155144378e-05,
+      "loss": 12.0008,
+      "step": 10093
+    },
+    {
+      "epoch": 1.7652021857923499,
+      "grad_norm": 0.8057819795188729,
+      "learning_rate": 7.644301159078581e-05,
+      "loss": 12.3295,
+      "step": 10094
+    },
+    {
+      "epoch": 1.7653770491803278,
+      "grad_norm": 0.8483426077217217,
+      "learning_rate": 7.642466246990692e-05,
+      "loss": 12.6013,
+      "step": 10095
+    },
+    {
+      "epoch": 1.7655519125683061,
+      "grad_norm": 0.8115328332659062,
+      "learning_rate": 7.640631418946126e-05,
+      "loss": 12.3389,
+      "step": 10096
+    },
+    {
+      "epoch": 1.765726775956284,
+      "grad_norm": 0.856883186760859,
+      "learning_rate": 7.638796675010293e-05,
+      "loss": 12.2436,
+      "step": 10097
+    },
+    {
+      "epoch": 1.7659016393442624,
+      "grad_norm": 0.8072608033874045,
+      "learning_rate": 7.6369620152486e-05,
+      "loss": 12.3534,
+      "step": 10098
+    },
+    {
+      "epoch": 1.7660765027322405,
+      "grad_norm": 0.8178324088969892,
+      "learning_rate": 7.635127439726451e-05,
+      "loss": 12.3164,
+      "step": 10099
+    },
+    {
+      "epoch": 1.7662513661202186,
+      "grad_norm": 0.7549243256742258,
+      "learning_rate": 7.633292948509245e-05,
+      "loss": 12.3907,
+      "step": 10100
+    },
+    {
+      "epoch": 1.7664262295081967,
+      "grad_norm": 0.8144827061372932,
+      "learning_rate": 7.63145854166238e-05,
+      "loss": 12.2853,
+      "step": 10101
+    },
+    {
+      "epoch": 1.7666010928961748,
+      "grad_norm": 1.2788446181261952,
+      "learning_rate": 7.629624219251256e-05,
+      "loss": 12.3803,
+      "step": 10102
+    },
+    {
+      "epoch": 1.766775956284153,
+      "grad_norm": 0.8463592064803668,
+      "learning_rate": 7.627789981341253e-05,
+      "loss": 12.5061,
+      "step": 10103
+    },
+    {
+      "epoch": 1.766950819672131,
+      "grad_norm": 0.8281519241271459,
+      "learning_rate": 7.625955827997766e-05,
+      "loss": 12.1364,
+      "step": 10104
+    },
+    {
+      "epoch": 1.7671256830601094,
+      "grad_norm": 0.8682573346584741,
+      "learning_rate": 7.624121759286182e-05,
+      "loss": 12.4412,
+      "step": 10105
+    },
+    {
+      "epoch": 1.7673005464480873,
+      "grad_norm": 0.8047010930194447,
+      "learning_rate": 7.62228777527188e-05,
+      "loss": 12.1304,
+      "step": 10106
+    },
+    {
+      "epoch": 1.7674754098360657,
+      "grad_norm": 0.8683374145632052,
+      "learning_rate": 7.62045387602024e-05,
+      "loss": 12.2354,
+      "step": 10107
+    },
+    {
+      "epoch": 1.7676502732240436,
+      "grad_norm": 0.8589513202868204,
+      "learning_rate": 7.618620061596639e-05,
+      "loss": 12.3486,
+      "step": 10108
+    },
+    {
+      "epoch": 1.767825136612022,
+      "grad_norm": 0.883882978500959,
+      "learning_rate": 7.616786332066457e-05,
+      "loss": 12.2498,
+      "step": 10109
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 0.8472515930433594,
+      "learning_rate": 7.614952687495051e-05,
+      "loss": 12.6601,
+      "step": 10110
+    },
+    {
+      "epoch": 1.7681748633879781,
+      "grad_norm": 0.8424565554178013,
+      "learning_rate": 7.613119127947795e-05,
+      "loss": 12.5338,
+      "step": 10111
+    },
+    {
+      "epoch": 1.7683497267759563,
+      "grad_norm": 0.8706734687884085,
+      "learning_rate": 7.611285653490055e-05,
+      "loss": 12.3674,
+      "step": 10112
+    },
+    {
+      "epoch": 1.7685245901639344,
+      "grad_norm": 0.8740153806705518,
+      "learning_rate": 7.609452264187191e-05,
+      "loss": 12.552,
+      "step": 10113
+    },
+    {
+      "epoch": 1.7686994535519127,
+      "grad_norm": 0.8445622639292512,
+      "learning_rate": 7.60761896010456e-05,
+      "loss": 12.3761,
+      "step": 10114
+    },
+    {
+      "epoch": 1.7688743169398906,
+      "grad_norm": 0.8971217654193938,
+      "learning_rate": 7.605785741307516e-05,
+      "loss": 12.3883,
+      "step": 10115
+    },
+    {
+      "epoch": 1.769049180327869,
+      "grad_norm": 0.8187233216599277,
+      "learning_rate": 7.603952607861415e-05,
+      "loss": 12.1733,
+      "step": 10116
+    },
+    {
+      "epoch": 1.7692240437158469,
+      "grad_norm": 0.9208757019801354,
+      "learning_rate": 7.602119559831608e-05,
+      "loss": 12.3564,
+      "step": 10117
+    },
+    {
+      "epoch": 1.7693989071038252,
+      "grad_norm": 0.9051425545026021,
+      "learning_rate": 7.600286597283433e-05,
+      "loss": 12.4223,
+      "step": 10118
+    },
+    {
+      "epoch": 1.7695737704918033,
+      "grad_norm": 0.8666024024893177,
+      "learning_rate": 7.598453720282234e-05,
+      "loss": 12.3052,
+      "step": 10119
+    },
+    {
+      "epoch": 1.7697486338797814,
+      "grad_norm": 0.8765718188867544,
+      "learning_rate": 7.596620928893362e-05,
+      "loss": 12.3675,
+      "step": 10120
+    },
+    {
+      "epoch": 1.7699234972677595,
+      "grad_norm": 0.8798761576411179,
+      "learning_rate": 7.59478822318214e-05,
+      "loss": 11.9667,
+      "step": 10121
+    },
+    {
+      "epoch": 1.7700983606557377,
+      "grad_norm": 0.8423861696381536,
+      "learning_rate": 7.592955603213907e-05,
+      "loss": 12.4032,
+      "step": 10122
+    },
+    {
+      "epoch": 1.770273224043716,
+      "grad_norm": 0.7945085748831644,
+      "learning_rate": 7.591123069053995e-05,
+      "loss": 12.317,
+      "step": 10123
+    },
+    {
+      "epoch": 1.770448087431694,
+      "grad_norm": 0.8193499276771006,
+      "learning_rate": 7.589290620767735e-05,
+      "loss": 12.096,
+      "step": 10124
+    },
+    {
+      "epoch": 1.7706229508196722,
+      "grad_norm": 0.8115320984004047,
+      "learning_rate": 7.587458258420443e-05,
+      "loss": 12.3641,
+      "step": 10125
+    },
+    {
+      "epoch": 1.7707978142076501,
+      "grad_norm": 0.8239949510496585,
+      "learning_rate": 7.585625982077446e-05,
+      "loss": 12.386,
+      "step": 10126
+    },
+    {
+      "epoch": 1.7709726775956285,
+      "grad_norm": 0.8912941571717625,
+      "learning_rate": 7.583793791804064e-05,
+      "loss": 12.3581,
+      "step": 10127
+    },
+    {
+      "epoch": 1.7711475409836066,
+      "grad_norm": 0.9236291787996269,
+      "learning_rate": 7.581961687665612e-05,
+      "loss": 12.5927,
+      "step": 10128
+    },
+    {
+      "epoch": 1.7713224043715847,
+      "grad_norm": 0.8385289425955899,
+      "learning_rate": 7.580129669727397e-05,
+      "loss": 12.2236,
+      "step": 10129
+    },
+    {
+      "epoch": 1.7714972677595628,
+      "grad_norm": 0.8291638777179041,
+      "learning_rate": 7.578297738054733e-05,
+      "loss": 12.0964,
+      "step": 10130
+    },
+    {
+      "epoch": 1.771672131147541,
+      "grad_norm": 0.9195577303254183,
+      "learning_rate": 7.576465892712924e-05,
+      "loss": 12.3824,
+      "step": 10131
+    },
+    {
+      "epoch": 1.771846994535519,
+      "grad_norm": 0.8439736250432072,
+      "learning_rate": 7.574634133767277e-05,
+      "loss": 12.1472,
+      "step": 10132
+    },
+    {
+      "epoch": 1.7720218579234972,
+      "grad_norm": 0.8515741766176567,
+      "learning_rate": 7.572802461283086e-05,
+      "loss": 12.2773,
+      "step": 10133
+    },
+    {
+      "epoch": 1.7721967213114755,
+      "grad_norm": 0.8749393216112381,
+      "learning_rate": 7.570970875325654e-05,
+      "loss": 12.5103,
+      "step": 10134
+    },
+    {
+      "epoch": 1.7723715846994534,
+      "grad_norm": 0.8146234009123889,
+      "learning_rate": 7.569139375960274e-05,
+      "loss": 12.4329,
+      "step": 10135
+    },
+    {
+      "epoch": 1.7725464480874318,
+      "grad_norm": 0.8068502899691988,
+      "learning_rate": 7.567307963252233e-05,
+      "loss": 12.2316,
+      "step": 10136
+    },
+    {
+      "epoch": 1.7727213114754097,
+      "grad_norm": 0.8651809427748928,
+      "learning_rate": 7.565476637266818e-05,
+      "loss": 12.4024,
+      "step": 10137
+    },
+    {
+      "epoch": 1.772896174863388,
+      "grad_norm": 0.8786224100179436,
+      "learning_rate": 7.563645398069319e-05,
+      "loss": 12.4692,
+      "step": 10138
+    },
+    {
+      "epoch": 1.7730710382513661,
+      "grad_norm": 0.8357073127729152,
+      "learning_rate": 7.561814245725019e-05,
+      "loss": 12.2719,
+      "step": 10139
+    },
+    {
+      "epoch": 1.7732459016393443,
+      "grad_norm": 0.8537068267569561,
+      "learning_rate": 7.559983180299186e-05,
+      "loss": 12.2559,
+      "step": 10140
+    },
+    {
+      "epoch": 1.7734207650273224,
+      "grad_norm": 0.811012540608638,
+      "learning_rate": 7.558152201857104e-05,
+      "loss": 12.5413,
+      "step": 10141
+    },
+    {
+      "epoch": 1.7735956284153005,
+      "grad_norm": 0.8355491418538514,
+      "learning_rate": 7.556321310464043e-05,
+      "loss": 12.1382,
+      "step": 10142
+    },
+    {
+      "epoch": 1.7737704918032788,
+      "grad_norm": 0.7766015816649869,
+      "learning_rate": 7.554490506185273e-05,
+      "loss": 12.3969,
+      "step": 10143
+    },
+    {
+      "epoch": 1.7739453551912567,
+      "grad_norm": 0.8240362147608006,
+      "learning_rate": 7.552659789086057e-05,
+      "loss": 12.2939,
+      "step": 10144
+    },
+    {
+      "epoch": 1.774120218579235,
+      "grad_norm": 0.8387039299668904,
+      "learning_rate": 7.550829159231661e-05,
+      "loss": 12.4449,
+      "step": 10145
+    },
+    {
+      "epoch": 1.774295081967213,
+      "grad_norm": 0.7873925107234222,
+      "learning_rate": 7.548998616687344e-05,
+      "loss": 12.2991,
+      "step": 10146
+    },
+    {
+      "epoch": 1.7744699453551913,
+      "grad_norm": 0.8432729066853588,
+      "learning_rate": 7.547168161518367e-05,
+      "loss": 12.5086,
+      "step": 10147
+    },
+    {
+      "epoch": 1.7746448087431694,
+      "grad_norm": 0.8020709660970681,
+      "learning_rate": 7.545337793789974e-05,
+      "loss": 12.4938,
+      "step": 10148
+    },
+    {
+      "epoch": 1.7748196721311476,
+      "grad_norm": 0.7606583681741466,
+      "learning_rate": 7.543507513567421e-05,
+      "loss": 12.3234,
+      "step": 10149
+    },
+    {
+      "epoch": 1.7749945355191257,
+      "grad_norm": 0.8764861878662756,
+      "learning_rate": 7.541677320915958e-05,
+      "loss": 12.3099,
+      "step": 10150
+    },
+    {
+      "epoch": 1.7751693989071038,
+      "grad_norm": 0.8116566682256513,
+      "learning_rate": 7.539847215900823e-05,
+      "loss": 12.1988,
+      "step": 10151
+    },
+    {
+      "epoch": 1.7753442622950821,
+      "grad_norm": 0.7818897598764276,
+      "learning_rate": 7.53801719858726e-05,
+      "loss": 12.0198,
+      "step": 10152
+    },
+    {
+      "epoch": 1.77551912568306,
+      "grad_norm": 0.8557435722388745,
+      "learning_rate": 7.536187269040509e-05,
+      "loss": 12.374,
+      "step": 10153
+    },
+    {
+      "epoch": 1.7756939890710384,
+      "grad_norm": 0.9598455857154677,
+      "learning_rate": 7.534357427325807e-05,
+      "loss": 12.5395,
+      "step": 10154
+    },
+    {
+      "epoch": 1.7758688524590163,
+      "grad_norm": 0.8769840919911334,
+      "learning_rate": 7.532527673508377e-05,
+      "loss": 12.4307,
+      "step": 10155
+    },
+    {
+      "epoch": 1.7760437158469946,
+      "grad_norm": 0.8998131387153184,
+      "learning_rate": 7.530698007653452e-05,
+      "loss": 12.2648,
+      "step": 10156
+    },
+    {
+      "epoch": 1.7762185792349727,
+      "grad_norm": 0.8183706470968718,
+      "learning_rate": 7.52886842982626e-05,
+      "loss": 12.3927,
+      "step": 10157
+    },
+    {
+      "epoch": 1.7763934426229508,
+      "grad_norm": 0.9681464307729655,
+      "learning_rate": 7.527038940092022e-05,
+      "loss": 12.2396,
+      "step": 10158
+    },
+    {
+      "epoch": 1.776568306010929,
+      "grad_norm": 0.865821487168898,
+      "learning_rate": 7.525209538515955e-05,
+      "loss": 12.4731,
+      "step": 10159
+    },
+    {
+      "epoch": 1.776743169398907,
+      "grad_norm": 0.8825004714396416,
+      "learning_rate": 7.523380225163276e-05,
+      "loss": 12.1577,
+      "step": 10160
+    },
+    {
+      "epoch": 1.7769180327868852,
+      "grad_norm": 0.8834931998765317,
+      "learning_rate": 7.521551000099199e-05,
+      "loss": 12.4921,
+      "step": 10161
+    },
+    {
+      "epoch": 1.7770928961748633,
+      "grad_norm": 0.902124011440694,
+      "learning_rate": 7.519721863388936e-05,
+      "loss": 12.2711,
+      "step": 10162
+    },
+    {
+      "epoch": 1.7772677595628417,
+      "grad_norm": 0.8614701492455467,
+      "learning_rate": 7.517892815097687e-05,
+      "loss": 12.3862,
+      "step": 10163
+    },
+    {
+      "epoch": 1.7774426229508196,
+      "grad_norm": 0.922952366334399,
+      "learning_rate": 7.516063855290659e-05,
+      "loss": 12.0826,
+      "step": 10164
+    },
+    {
+      "epoch": 1.777617486338798,
+      "grad_norm": 0.8864362990633226,
+      "learning_rate": 7.514234984033053e-05,
+      "loss": 12.3852,
+      "step": 10165
+    },
+    {
+      "epoch": 1.7777923497267758,
+      "grad_norm": 0.8940857893823594,
+      "learning_rate": 7.512406201390068e-05,
+      "loss": 12.2565,
+      "step": 10166
+    },
+    {
+      "epoch": 1.7779672131147541,
+      "grad_norm": 0.8160972712889456,
+      "learning_rate": 7.510577507426893e-05,
+      "loss": 12.5125,
+      "step": 10167
+    },
+    {
+      "epoch": 1.7781420765027323,
+      "grad_norm": 0.8273723854713493,
+      "learning_rate": 7.508748902208722e-05,
+      "loss": 12.0824,
+      "step": 10168
+    },
+    {
+      "epoch": 1.7783169398907104,
+      "grad_norm": 0.9190526000080453,
+      "learning_rate": 7.506920385800742e-05,
+      "loss": 12.0052,
+      "step": 10169
+    },
+    {
+      "epoch": 1.7784918032786885,
+      "grad_norm": 0.8001747312565544,
+      "learning_rate": 7.505091958268137e-05,
+      "loss": 12.1923,
+      "step": 10170
+    },
+    {
+      "epoch": 1.7786666666666666,
+      "grad_norm": 0.9197999985591154,
+      "learning_rate": 7.503263619676088e-05,
+      "loss": 12.4966,
+      "step": 10171
+    },
+    {
+      "epoch": 1.778841530054645,
+      "grad_norm": 0.8333017651974164,
+      "learning_rate": 7.501435370089774e-05,
+      "loss": 12.3095,
+      "step": 10172
+    },
+    {
+      "epoch": 1.7790163934426229,
+      "grad_norm": 0.9001205339148077,
+      "learning_rate": 7.499607209574373e-05,
+      "loss": 12.362,
+      "step": 10173
+    },
+    {
+      "epoch": 1.7791912568306012,
+      "grad_norm": 0.872838878180893,
+      "learning_rate": 7.49777913819505e-05,
+      "loss": 12.244,
+      "step": 10174
+    },
+    {
+      "epoch": 1.779366120218579,
+      "grad_norm": 0.8216477334458177,
+      "learning_rate": 7.495951156016978e-05,
+      "loss": 12.4007,
+      "step": 10175
+    },
+    {
+      "epoch": 1.7795409836065574,
+      "grad_norm": 0.9432143679672689,
+      "learning_rate": 7.49412326310532e-05,
+      "loss": 12.032,
+      "step": 10176
+    },
+    {
+      "epoch": 1.7797158469945356,
+      "grad_norm": 0.8151284523478253,
+      "learning_rate": 7.492295459525243e-05,
+      "loss": 12.4184,
+      "step": 10177
+    },
+    {
+      "epoch": 1.7798907103825137,
+      "grad_norm": 0.9973890322065332,
+      "learning_rate": 7.4904677453419e-05,
+      "loss": 12.2996,
+      "step": 10178
+    },
+    {
+      "epoch": 1.7800655737704918,
+      "grad_norm": 0.8432502717618321,
+      "learning_rate": 7.488640120620452e-05,
+      "loss": 12.4276,
+      "step": 10179
+    },
+    {
+      "epoch": 1.78024043715847,
+      "grad_norm": 0.8462153214305808,
+      "learning_rate": 7.486812585426047e-05,
+      "loss": 12.1705,
+      "step": 10180
+    },
+    {
+      "epoch": 1.780415300546448,
+      "grad_norm": 0.8322632823776632,
+      "learning_rate": 7.484985139823844e-05,
+      "loss": 12.2514,
+      "step": 10181
+    },
+    {
+      "epoch": 1.7805901639344262,
+      "grad_norm": 0.9233582816193504,
+      "learning_rate": 7.483157783878976e-05,
+      "loss": 12.3897,
+      "step": 10182
+    },
+    {
+      "epoch": 1.7807650273224045,
+      "grad_norm": 0.8729164466989453,
+      "learning_rate": 7.481330517656592e-05,
+      "loss": 12.2727,
+      "step": 10183
+    },
+    {
+      "epoch": 1.7809398907103824,
+      "grad_norm": 0.9752012391767,
+      "learning_rate": 7.479503341221838e-05,
+      "loss": 12.5271,
+      "step": 10184
+    },
+    {
+      "epoch": 1.7811147540983607,
+      "grad_norm": 0.8280678612814525,
+      "learning_rate": 7.47767625463984e-05,
+      "loss": 12.571,
+      "step": 10185
+    },
+    {
+      "epoch": 1.7812896174863388,
+      "grad_norm": 0.9361137336301735,
+      "learning_rate": 7.475849257975736e-05,
+      "loss": 12.2599,
+      "step": 10186
+    },
+    {
+      "epoch": 1.781464480874317,
+      "grad_norm": 1.0164533474215354,
+      "learning_rate": 7.474022351294656e-05,
+      "loss": 12.2865,
+      "step": 10187
+    },
+    {
+      "epoch": 1.781639344262295,
+      "grad_norm": 0.86794160358336,
+      "learning_rate": 7.472195534661732e-05,
+      "loss": 12.3944,
+      "step": 10188
+    },
+    {
+      "epoch": 1.7818142076502732,
+      "grad_norm": 0.8561567565699876,
+      "learning_rate": 7.470368808142081e-05,
+      "loss": 12.4277,
+      "step": 10189
+    },
+    {
+      "epoch": 1.7819890710382513,
+      "grad_norm": 0.9528747136698381,
+      "learning_rate": 7.468542171800826e-05,
+      "loss": 12.441,
+      "step": 10190
+    },
+    {
+      "epoch": 1.7821639344262294,
+      "grad_norm": 0.7774449790006132,
+      "learning_rate": 7.466715625703086e-05,
+      "loss": 12.3724,
+      "step": 10191
+    },
+    {
+      "epoch": 1.7823387978142078,
+      "grad_norm": 1.0323833744959987,
+      "learning_rate": 7.464889169913977e-05,
+      "loss": 12.1198,
+      "step": 10192
+    },
+    {
+      "epoch": 1.7825136612021857,
+      "grad_norm": 0.8668498425979604,
+      "learning_rate": 7.463062804498603e-05,
+      "loss": 11.9888,
+      "step": 10193
+    },
+    {
+      "epoch": 1.782688524590164,
+      "grad_norm": 0.8493626690936099,
+      "learning_rate": 7.461236529522077e-05,
+      "loss": 12.088,
+      "step": 10194
+    },
+    {
+      "epoch": 1.782863387978142,
+      "grad_norm": 0.9750962249206766,
+      "learning_rate": 7.459410345049502e-05,
+      "loss": 12.2894,
+      "step": 10195
+    },
+    {
+      "epoch": 1.7830382513661203,
+      "grad_norm": 1.0349224219674198,
+      "learning_rate": 7.457584251145982e-05,
+      "loss": 12.4559,
+      "step": 10196
+    },
+    {
+      "epoch": 1.7832131147540984,
+      "grad_norm": 0.8399918682883613,
+      "learning_rate": 7.45575824787661e-05,
+      "loss": 12.2163,
+      "step": 10197
+    },
+    {
+      "epoch": 1.7833879781420765,
+      "grad_norm": 1.3126972326532405,
+      "learning_rate": 7.453932335306486e-05,
+      "loss": 12.2673,
+      "step": 10198
+    },
+    {
+      "epoch": 1.7835628415300546,
+      "grad_norm": 0.8624218678491117,
+      "learning_rate": 7.452106513500702e-05,
+      "loss": 12.4354,
+      "step": 10199
+    },
+    {
+      "epoch": 1.7837377049180327,
+      "grad_norm": 1.061173191239623,
+      "learning_rate": 7.45028078252434e-05,
+      "loss": 12.0488,
+      "step": 10200
+    },
+    {
+      "epoch": 1.783912568306011,
+      "grad_norm": 1.0709508488451664,
+      "learning_rate": 7.448455142442488e-05,
+      "loss": 12.2864,
+      "step": 10201
+    },
+    {
+      "epoch": 1.784087431693989,
+      "grad_norm": 0.8943268956606715,
+      "learning_rate": 7.446629593320232e-05,
+      "loss": 12.5714,
+      "step": 10202
+    },
+    {
+      "epoch": 1.7842622950819673,
+      "grad_norm": 1.078971880802225,
+      "learning_rate": 7.444804135222647e-05,
+      "loss": 12.366,
+      "step": 10203
+    },
+    {
+      "epoch": 1.7844371584699452,
+      "grad_norm": 0.9371448036469651,
+      "learning_rate": 7.44297876821481e-05,
+      "loss": 12.3127,
+      "step": 10204
+    },
+    {
+      "epoch": 1.7846120218579236,
+      "grad_norm": 0.8687986638369404,
+      "learning_rate": 7.44115349236179e-05,
+      "loss": 12.0904,
+      "step": 10205
+    },
+    {
+      "epoch": 1.7847868852459017,
+      "grad_norm": 1.040393173818483,
+      "learning_rate": 7.439328307728659e-05,
+      "loss": 12.4955,
+      "step": 10206
+    },
+    {
+      "epoch": 1.7849617486338798,
+      "grad_norm": 0.8704641543911663,
+      "learning_rate": 7.437503214380484e-05,
+      "loss": 12.3389,
+      "step": 10207
+    },
+    {
+      "epoch": 1.785136612021858,
+      "grad_norm": 1.008284425081215,
+      "learning_rate": 7.435678212382322e-05,
+      "loss": 12.4632,
+      "step": 10208
+    },
+    {
+      "epoch": 1.785311475409836,
+      "grad_norm": 0.8962720314479901,
+      "learning_rate": 7.433853301799237e-05,
+      "loss": 12.4359,
+      "step": 10209
+    },
+    {
+      "epoch": 1.7854863387978142,
+      "grad_norm": 0.9031193754867678,
+      "learning_rate": 7.432028482696283e-05,
+      "loss": 12.433,
+      "step": 10210
+    },
+    {
+      "epoch": 1.7856612021857923,
+      "grad_norm": 0.9224005576292607,
+      "learning_rate": 7.430203755138519e-05,
+      "loss": 12.3879,
+      "step": 10211
+    },
+    {
+      "epoch": 1.7858360655737706,
+      "grad_norm": 0.908806597285641,
+      "learning_rate": 7.428379119190982e-05,
+      "loss": 12.5155,
+      "step": 10212
+    },
+    {
+      "epoch": 1.7860109289617485,
+      "grad_norm": 0.9168945541406177,
+      "learning_rate": 7.426554574918726e-05,
+      "loss": 12.274,
+      "step": 10213
+    },
+    {
+      "epoch": 1.7861857923497269,
+      "grad_norm": 0.9469692512195691,
+      "learning_rate": 7.424730122386793e-05,
+      "loss": 12.4857,
+      "step": 10214
+    },
+    {
+      "epoch": 1.786360655737705,
+      "grad_norm": 0.9066305336169709,
+      "learning_rate": 7.422905761660224e-05,
+      "loss": 12.083,
+      "step": 10215
+    },
+    {
+      "epoch": 1.786535519125683,
+      "grad_norm": 0.9611812940803102,
+      "learning_rate": 7.42108149280405e-05,
+      "loss": 12.4591,
+      "step": 10216
+    },
+    {
+      "epoch": 1.7867103825136612,
+      "grad_norm": 0.9142252212621171,
+      "learning_rate": 7.41925731588331e-05,
+      "loss": 12.2538,
+      "step": 10217
+    },
+    {
+      "epoch": 1.7868852459016393,
+      "grad_norm": 0.8627601084827409,
+      "learning_rate": 7.417433230963035e-05,
+      "loss": 12.3385,
+      "step": 10218
+    },
+    {
+      "epoch": 1.7870601092896174,
+      "grad_norm": 0.8925934380373322,
+      "learning_rate": 7.415609238108244e-05,
+      "loss": 12.4855,
+      "step": 10219
+    },
+    {
+      "epoch": 1.7872349726775956,
+      "grad_norm": 0.9633689366433486,
+      "learning_rate": 7.413785337383962e-05,
+      "loss": 12.1705,
+      "step": 10220
+    },
+    {
+      "epoch": 1.787409836065574,
+      "grad_norm": 0.8576689755769843,
+      "learning_rate": 7.411961528855214e-05,
+      "loss": 12.245,
+      "step": 10221
+    },
+    {
+      "epoch": 1.7875846994535518,
+      "grad_norm": 0.9508468460099264,
+      "learning_rate": 7.410137812587012e-05,
+      "loss": 12.3576,
+      "step": 10222
+    },
+    {
+      "epoch": 1.7877595628415301,
+      "grad_norm": 0.9166685137017778,
+      "learning_rate": 7.408314188644372e-05,
+      "loss": 12.2973,
+      "step": 10223
+    },
+    {
+      "epoch": 1.787934426229508,
+      "grad_norm": 0.8499312386006763,
+      "learning_rate": 7.406490657092303e-05,
+      "loss": 12.2081,
+      "step": 10224
+    },
+    {
+      "epoch": 1.7881092896174864,
+      "grad_norm": 1.07179250644009,
+      "learning_rate": 7.404667217995811e-05,
+      "loss": 12.3125,
+      "step": 10225
+    },
+    {
+      "epoch": 1.7882841530054645,
+      "grad_norm": 0.8463956865033282,
+      "learning_rate": 7.402843871419905e-05,
+      "loss": 12.1689,
+      "step": 10226
+    },
+    {
+      "epoch": 1.7884590163934426,
+      "grad_norm": 0.9424327346534143,
+      "learning_rate": 7.401020617429577e-05,
+      "loss": 12.3664,
+      "step": 10227
+    },
+    {
+      "epoch": 1.7886338797814207,
+      "grad_norm": 0.87148777640361,
+      "learning_rate": 7.399197456089825e-05,
+      "loss": 12.3185,
+      "step": 10228
+    },
+    {
+      "epoch": 1.7888087431693989,
+      "grad_norm": 0.8932287924410085,
+      "learning_rate": 7.397374387465653e-05,
+      "loss": 12.4359,
+      "step": 10229
+    },
+    {
+      "epoch": 1.7889836065573772,
+      "grad_norm": 0.845955132095978,
+      "learning_rate": 7.395551411622035e-05,
+      "loss": 12.5069,
+      "step": 10230
+    },
+    {
+      "epoch": 1.789158469945355,
+      "grad_norm": 0.7818147424799535,
+      "learning_rate": 7.39372852862397e-05,
+      "loss": 12.2347,
+      "step": 10231
+    },
+    {
+      "epoch": 1.7893333333333334,
+      "grad_norm": 0.9186031731287486,
+      "learning_rate": 7.391905738536435e-05,
+      "loss": 12.2688,
+      "step": 10232
+    },
+    {
+      "epoch": 1.7895081967213113,
+      "grad_norm": 0.8319732061241333,
+      "learning_rate": 7.390083041424416e-05,
+      "loss": 12.3675,
+      "step": 10233
+    },
+    {
+      "epoch": 1.7896830601092897,
+      "grad_norm": 0.8758514294313138,
+      "learning_rate": 7.388260437352885e-05,
+      "loss": 12.5461,
+      "step": 10234
+    },
+    {
+      "epoch": 1.7898579234972678,
+      "grad_norm": 0.8002690478446246,
+      "learning_rate": 7.386437926386818e-05,
+      "loss": 12.3461,
+      "step": 10235
+    },
+    {
+      "epoch": 1.790032786885246,
+      "grad_norm": 0.8390452552139146,
+      "learning_rate": 7.384615508591186e-05,
+      "loss": 12.3733,
+      "step": 10236
+    },
+    {
+      "epoch": 1.790207650273224,
+      "grad_norm": 0.9288931992406997,
+      "learning_rate": 7.382793184030958e-05,
+      "loss": 12.5929,
+      "step": 10237
+    },
+    {
+      "epoch": 1.7903825136612022,
+      "grad_norm": 0.8044780868928011,
+      "learning_rate": 7.380970952771093e-05,
+      "loss": 12.2512,
+      "step": 10238
+    },
+    {
+      "epoch": 1.7905573770491803,
+      "grad_norm": 0.8809185759831305,
+      "learning_rate": 7.379148814876552e-05,
+      "loss": 12.2611,
+      "step": 10239
+    },
+    {
+      "epoch": 1.7907322404371584,
+      "grad_norm": 0.7923338447505514,
+      "learning_rate": 7.377326770412293e-05,
+      "loss": 12.1617,
+      "step": 10240
+    },
+    {
+      "epoch": 1.7909071038251367,
+      "grad_norm": 0.9051748845404274,
+      "learning_rate": 7.375504819443275e-05,
+      "loss": 12.5188,
+      "step": 10241
+    },
+    {
+      "epoch": 1.7910819672131146,
+      "grad_norm": 0.850801967031577,
+      "learning_rate": 7.37368296203444e-05,
+      "loss": 12.4327,
+      "step": 10242
+    },
+    {
+      "epoch": 1.791256830601093,
+      "grad_norm": 0.920230671282139,
+      "learning_rate": 7.37186119825074e-05,
+      "loss": 12.5842,
+      "step": 10243
+    },
+    {
+      "epoch": 1.7914316939890709,
+      "grad_norm": 0.8478023700466523,
+      "learning_rate": 7.37003952815712e-05,
+      "loss": 12.2682,
+      "step": 10244
+    },
+    {
+      "epoch": 1.7916065573770492,
+      "grad_norm": 0.8241102180911624,
+      "learning_rate": 7.368217951818515e-05,
+      "loss": 12.2242,
+      "step": 10245
+    },
+    {
+      "epoch": 1.7917814207650273,
+      "grad_norm": 0.845700239909394,
+      "learning_rate": 7.366396469299865e-05,
+      "loss": 12.1594,
+      "step": 10246
+    },
+    {
+      "epoch": 1.7919562841530055,
+      "grad_norm": 0.9991237331055037,
+      "learning_rate": 7.364575080666104e-05,
+      "loss": 11.9861,
+      "step": 10247
+    },
+    {
+      "epoch": 1.7921311475409836,
+      "grad_norm": 0.9062609378581191,
+      "learning_rate": 7.362753785982165e-05,
+      "loss": 12.0006,
+      "step": 10248
+    },
+    {
+      "epoch": 1.7923060109289617,
+      "grad_norm": 0.8988543651585625,
+      "learning_rate": 7.360932585312971e-05,
+      "loss": 12.3627,
+      "step": 10249
+    },
+    {
+      "epoch": 1.79248087431694,
+      "grad_norm": 0.9735481175907836,
+      "learning_rate": 7.359111478723444e-05,
+      "loss": 12.2043,
+      "step": 10250
+    },
+    {
+      "epoch": 1.792655737704918,
+      "grad_norm": 0.8577816363917664,
+      "learning_rate": 7.357290466278509e-05,
+      "loss": 12.2734,
+      "step": 10251
+    },
+    {
+      "epoch": 1.7928306010928963,
+      "grad_norm": 0.9057281375529576,
+      "learning_rate": 7.355469548043082e-05,
+      "loss": 12.3163,
+      "step": 10252
+    },
+    {
+      "epoch": 1.7930054644808742,
+      "grad_norm": 0.8300885981089025,
+      "learning_rate": 7.353648724082075e-05,
+      "loss": 12.5282,
+      "step": 10253
+    },
+    {
+      "epoch": 1.7931803278688525,
+      "grad_norm": 0.8380617018797041,
+      "learning_rate": 7.351827994460398e-05,
+      "loss": 11.9711,
+      "step": 10254
+    },
+    {
+      "epoch": 1.7933551912568306,
+      "grad_norm": 1.0169945444449526,
+      "learning_rate": 7.350007359242959e-05,
+      "loss": 12.3505,
+      "step": 10255
+    },
+    {
+      "epoch": 1.7935300546448087,
+      "grad_norm": 0.9034716005550919,
+      "learning_rate": 7.348186818494667e-05,
+      "loss": 12.047,
+      "step": 10256
+    },
+    {
+      "epoch": 1.7937049180327869,
+      "grad_norm": 1.075867205239166,
+      "learning_rate": 7.34636637228041e-05,
+      "loss": 12.4427,
+      "step": 10257
+    },
+    {
+      "epoch": 1.793879781420765,
+      "grad_norm": 0.8510398138944657,
+      "learning_rate": 7.344546020665096e-05,
+      "loss": 12.4413,
+      "step": 10258
+    },
+    {
+      "epoch": 1.7940546448087433,
+      "grad_norm": 0.8907158326028128,
+      "learning_rate": 7.342725763713611e-05,
+      "loss": 12.4056,
+      "step": 10259
+    },
+    {
+      "epoch": 1.7942295081967212,
+      "grad_norm": 0.8518501628316215,
+      "learning_rate": 7.340905601490848e-05,
+      "loss": 12.2314,
+      "step": 10260
+    },
+    {
+      "epoch": 1.7944043715846996,
+      "grad_norm": 0.8518401752480195,
+      "learning_rate": 7.339085534061692e-05,
+      "loss": 12.3434,
+      "step": 10261
+    },
+    {
+      "epoch": 1.7945792349726775,
+      "grad_norm": 0.8927701428424435,
+      "learning_rate": 7.337265561491029e-05,
+      "loss": 12.4939,
+      "step": 10262
+    },
+    {
+      "epoch": 1.7947540983606558,
+      "grad_norm": 0.8165147789374615,
+      "learning_rate": 7.335445683843743e-05,
+      "loss": 12.1919,
+      "step": 10263
+    },
+    {
+      "epoch": 1.794928961748634,
+      "grad_norm": 0.8655497341596886,
+      "learning_rate": 7.333625901184698e-05,
+      "loss": 12.2057,
+      "step": 10264
+    },
+    {
+      "epoch": 1.795103825136612,
+      "grad_norm": 0.8613894480550462,
+      "learning_rate": 7.331806213578776e-05,
+      "loss": 12.4998,
+      "step": 10265
+    },
+    {
+      "epoch": 1.7952786885245902,
+      "grad_norm": 0.8627450404008175,
+      "learning_rate": 7.329986621090845e-05,
+      "loss": 12.1647,
+      "step": 10266
+    },
+    {
+      "epoch": 1.7954535519125683,
+      "grad_norm": 0.9061656109095474,
+      "learning_rate": 7.328167123785774e-05,
+      "loss": 12.4308,
+      "step": 10267
+    },
+    {
+      "epoch": 1.7956284153005464,
+      "grad_norm": 0.8188798940834372,
+      "learning_rate": 7.326347721728421e-05,
+      "loss": 12.0532,
+      "step": 10268
+    },
+    {
+      "epoch": 1.7958032786885245,
+      "grad_norm": 0.9113268222600343,
+      "learning_rate": 7.324528414983647e-05,
+      "loss": 12.3668,
+      "step": 10269
+    },
+    {
+      "epoch": 1.7959781420765029,
+      "grad_norm": 0.8547797636818184,
+      "learning_rate": 7.32270920361631e-05,
+      "loss": 12.3983,
+      "step": 10270
+    },
+    {
+      "epoch": 1.7961530054644808,
+      "grad_norm": 0.8390399848725346,
+      "learning_rate": 7.320890087691265e-05,
+      "loss": 12.1418,
+      "step": 10271
+    },
+    {
+      "epoch": 1.796327868852459,
+      "grad_norm": 0.9790804981893908,
+      "learning_rate": 7.319071067273355e-05,
+      "loss": 12.4468,
+      "step": 10272
+    },
+    {
+      "epoch": 1.796502732240437,
+      "grad_norm": 0.7942550955313098,
+      "learning_rate": 7.317252142427429e-05,
+      "loss": 12.1761,
+      "step": 10273
+    },
+    {
+      "epoch": 1.7966775956284153,
+      "grad_norm": 0.8106095063505774,
+      "learning_rate": 7.315433313218328e-05,
+      "loss": 12.0012,
+      "step": 10274
+    },
+    {
+      "epoch": 1.7968524590163935,
+      "grad_norm": 0.9193562102788935,
+      "learning_rate": 7.313614579710899e-05,
+      "loss": 12.4064,
+      "step": 10275
+    },
+    {
+      "epoch": 1.7970273224043716,
+      "grad_norm": 0.8828249751874218,
+      "learning_rate": 7.311795941969966e-05,
+      "loss": 12.3206,
+      "step": 10276
+    },
+    {
+      "epoch": 1.7972021857923497,
+      "grad_norm": 0.7900123155803881,
+      "learning_rate": 7.309977400060368e-05,
+      "loss": 12.2658,
+      "step": 10277
+    },
+    {
+      "epoch": 1.7973770491803278,
+      "grad_norm": 0.9203220311679309,
+      "learning_rate": 7.308158954046935e-05,
+      "loss": 12.3875,
+      "step": 10278
+    },
+    {
+      "epoch": 1.7975519125683062,
+      "grad_norm": 0.8735838866750704,
+      "learning_rate": 7.306340603994487e-05,
+      "loss": 12.1128,
+      "step": 10279
+    },
+    {
+      "epoch": 1.797726775956284,
+      "grad_norm": 0.8370482724037431,
+      "learning_rate": 7.304522349967851e-05,
+      "loss": 12.4351,
+      "step": 10280
+    },
+    {
+      "epoch": 1.7979016393442624,
+      "grad_norm": 0.8282655408848165,
+      "learning_rate": 7.302704192031842e-05,
+      "loss": 12.2725,
+      "step": 10281
+    },
+    {
+      "epoch": 1.7980765027322403,
+      "grad_norm": 0.8362454572042515,
+      "learning_rate": 7.300886130251281e-05,
+      "loss": 12.1694,
+      "step": 10282
+    },
+    {
+      "epoch": 1.7982513661202186,
+      "grad_norm": 0.8732886499325971,
+      "learning_rate": 7.299068164690972e-05,
+      "loss": 12.1648,
+      "step": 10283
+    },
+    {
+      "epoch": 1.7984262295081967,
+      "grad_norm": 0.878336324315316,
+      "learning_rate": 7.297250295415729e-05,
+      "loss": 12.4092,
+      "step": 10284
+    },
+    {
+      "epoch": 1.7986010928961749,
+      "grad_norm": 0.9642000562432985,
+      "learning_rate": 7.295432522490353e-05,
+      "loss": 12.4229,
+      "step": 10285
+    },
+    {
+      "epoch": 1.798775956284153,
+      "grad_norm": 0.7969246604746993,
+      "learning_rate": 7.293614845979652e-05,
+      "loss": 12.274,
+      "step": 10286
+    },
+    {
+      "epoch": 1.798950819672131,
+      "grad_norm": 0.8194800638962813,
+      "learning_rate": 7.291797265948416e-05,
+      "loss": 12.3431,
+      "step": 10287
+    },
+    {
+      "epoch": 1.7991256830601094,
+      "grad_norm": 0.9167762946575484,
+      "learning_rate": 7.289979782461443e-05,
+      "loss": 12.0137,
+      "step": 10288
+    },
+    {
+      "epoch": 1.7993005464480873,
+      "grad_norm": 0.7977875592615656,
+      "learning_rate": 7.288162395583525e-05,
+      "loss": 12.051,
+      "step": 10289
+    },
+    {
+      "epoch": 1.7994754098360657,
+      "grad_norm": 0.7991984086136696,
+      "learning_rate": 7.286345105379453e-05,
+      "loss": 12.4233,
+      "step": 10290
+    },
+    {
+      "epoch": 1.7996502732240436,
+      "grad_norm": 0.8427938808207396,
+      "learning_rate": 7.284527911914004e-05,
+      "loss": 12.2328,
+      "step": 10291
+    },
+    {
+      "epoch": 1.799825136612022,
+      "grad_norm": 0.8784887727132893,
+      "learning_rate": 7.282710815251961e-05,
+      "loss": 12.1422,
+      "step": 10292
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8175247416624171,
+      "learning_rate": 7.280893815458109e-05,
+      "loss": 12.425,
+      "step": 10293
+    },
+    {
+      "epoch": 1.8001748633879782,
+      "grad_norm": 0.8763598427103595,
+      "learning_rate": 7.27907691259721e-05,
+      "loss": 12.2017,
+      "step": 10294
+    },
+    {
+      "epoch": 1.8003497267759563,
+      "grad_norm": 0.8058466609988468,
+      "learning_rate": 7.277260106734039e-05,
+      "loss": 11.9263,
+      "step": 10295
+    },
+    {
+      "epoch": 1.8005245901639344,
+      "grad_norm": 0.8167609033059259,
+      "learning_rate": 7.275443397933366e-05,
+      "loss": 12.1884,
+      "step": 10296
+    },
+    {
+      "epoch": 1.8006994535519125,
+      "grad_norm": 0.930815875523796,
+      "learning_rate": 7.273626786259952e-05,
+      "loss": 12.5641,
+      "step": 10297
+    },
+    {
+      "epoch": 1.8008743169398906,
+      "grad_norm": 0.8301231069611339,
+      "learning_rate": 7.271810271778559e-05,
+      "loss": 12.3136,
+      "step": 10298
+    },
+    {
+      "epoch": 1.801049180327869,
+      "grad_norm": 0.9803629680483931,
+      "learning_rate": 7.26999385455394e-05,
+      "loss": 12.4588,
+      "step": 10299
+    },
+    {
+      "epoch": 1.8012240437158469,
+      "grad_norm": 0.7934410772447484,
+      "learning_rate": 7.26817753465085e-05,
+      "loss": 12.4501,
+      "step": 10300
+    },
+    {
+      "epoch": 1.8013989071038252,
+      "grad_norm": 0.9826506311463088,
+      "learning_rate": 7.266361312134045e-05,
+      "loss": 12.163,
+      "step": 10301
+    },
+    {
+      "epoch": 1.8015737704918031,
+      "grad_norm": 0.8270796180949644,
+      "learning_rate": 7.264545187068262e-05,
+      "loss": 12.5143,
+      "step": 10302
+    },
+    {
+      "epoch": 1.8017486338797815,
+      "grad_norm": 0.761706009525359,
+      "learning_rate": 7.262729159518244e-05,
+      "loss": 12.1829,
+      "step": 10303
+    },
+    {
+      "epoch": 1.8019234972677596,
+      "grad_norm": 0.8289756033715447,
+      "learning_rate": 7.260913229548735e-05,
+      "loss": 12.5232,
+      "step": 10304
+    },
+    {
+      "epoch": 1.8020983606557377,
+      "grad_norm": 0.8512502136638614,
+      "learning_rate": 7.25909739722447e-05,
+      "loss": 12.3429,
+      "step": 10305
+    },
+    {
+      "epoch": 1.8022732240437158,
+      "grad_norm": 0.8712639723968792,
+      "learning_rate": 7.25728166261018e-05,
+      "loss": 12.5541,
+      "step": 10306
+    },
+    {
+      "epoch": 1.802448087431694,
+      "grad_norm": 0.859719232711297,
+      "learning_rate": 7.255466025770593e-05,
+      "loss": 12.3246,
+      "step": 10307
+    },
+    {
+      "epoch": 1.8026229508196723,
+      "grad_norm": 0.9393345984749552,
+      "learning_rate": 7.25365048677044e-05,
+      "loss": 12.2136,
+      "step": 10308
+    },
+    {
+      "epoch": 1.8027978142076502,
+      "grad_norm": 0.8299570440567844,
+      "learning_rate": 7.251835045674434e-05,
+      "loss": 12.3953,
+      "step": 10309
+    },
+    {
+      "epoch": 1.8029726775956285,
+      "grad_norm": 0.8215415123959144,
+      "learning_rate": 7.250019702547298e-05,
+      "loss": 12.4069,
+      "step": 10310
+    },
+    {
+      "epoch": 1.8031475409836064,
+      "grad_norm": 0.896648014815074,
+      "learning_rate": 7.248204457453746e-05,
+      "loss": 12.4766,
+      "step": 10311
+    },
+    {
+      "epoch": 1.8033224043715848,
+      "grad_norm": 0.8949173999462613,
+      "learning_rate": 7.246389310458493e-05,
+      "loss": 12.4207,
+      "step": 10312
+    },
+    {
+      "epoch": 1.8034972677595629,
+      "grad_norm": 0.7605743372999363,
+      "learning_rate": 7.244574261626243e-05,
+      "loss": 12.2829,
+      "step": 10313
+    },
+    {
+      "epoch": 1.803672131147541,
+      "grad_norm": 0.863648990610636,
+      "learning_rate": 7.242759311021699e-05,
+      "loss": 12.4662,
+      "step": 10314
+    },
+    {
+      "epoch": 1.803846994535519,
+      "grad_norm": 0.9533251505142829,
+      "learning_rate": 7.240944458709564e-05,
+      "loss": 12.4237,
+      "step": 10315
+    },
+    {
+      "epoch": 1.8040218579234972,
+      "grad_norm": 0.8801261155304421,
+      "learning_rate": 7.239129704754541e-05,
+      "loss": 12.3383,
+      "step": 10316
+    },
+    {
+      "epoch": 1.8041967213114756,
+      "grad_norm": 0.9083623189852397,
+      "learning_rate": 7.237315049221313e-05,
+      "loss": 12.5068,
+      "step": 10317
+    },
+    {
+      "epoch": 1.8043715846994535,
+      "grad_norm": 0.8007302378734926,
+      "learning_rate": 7.235500492174576e-05,
+      "loss": 12.2126,
+      "step": 10318
+    },
+    {
+      "epoch": 1.8045464480874318,
+      "grad_norm": 0.8905288279759441,
+      "learning_rate": 7.233686033679017e-05,
+      "loss": 12.3601,
+      "step": 10319
+    },
+    {
+      "epoch": 1.8047213114754097,
+      "grad_norm": 0.8646825774984003,
+      "learning_rate": 7.231871673799322e-05,
+      "loss": 12.3335,
+      "step": 10320
+    },
+    {
+      "epoch": 1.804896174863388,
+      "grad_norm": 0.7952237167689241,
+      "learning_rate": 7.230057412600165e-05,
+      "loss": 12.2572,
+      "step": 10321
+    },
+    {
+      "epoch": 1.8050710382513662,
+      "grad_norm": 1.0144488811323187,
+      "learning_rate": 7.228243250146223e-05,
+      "loss": 12.4561,
+      "step": 10322
+    },
+    {
+      "epoch": 1.8052459016393443,
+      "grad_norm": 0.863116437936078,
+      "learning_rate": 7.226429186502173e-05,
+      "loss": 12.4047,
+      "step": 10323
+    },
+    {
+      "epoch": 1.8054207650273224,
+      "grad_norm": 0.8810617613314512,
+      "learning_rate": 7.22461522173268e-05,
+      "loss": 11.971,
+      "step": 10324
+    },
+    {
+      "epoch": 1.8055956284153005,
+      "grad_norm": 0.8569593428111462,
+      "learning_rate": 7.222801355902413e-05,
+      "loss": 12.5428,
+      "step": 10325
+    },
+    {
+      "epoch": 1.8057704918032786,
+      "grad_norm": 0.8230959945208861,
+      "learning_rate": 7.220987589076031e-05,
+      "loss": 11.8568,
+      "step": 10326
+    },
+    {
+      "epoch": 1.8059453551912568,
+      "grad_norm": 0.8487186495772262,
+      "learning_rate": 7.219173921318201e-05,
+      "loss": 12.3007,
+      "step": 10327
+    },
+    {
+      "epoch": 1.806120218579235,
+      "grad_norm": 0.8668646892128924,
+      "learning_rate": 7.217360352693565e-05,
+      "loss": 12.2216,
+      "step": 10328
+    },
+    {
+      "epoch": 1.806295081967213,
+      "grad_norm": 0.7992104714865973,
+      "learning_rate": 7.215546883266784e-05,
+      "loss": 12.4226,
+      "step": 10329
+    },
+    {
+      "epoch": 1.8064699453551913,
+      "grad_norm": 0.9601732650911196,
+      "learning_rate": 7.213733513102501e-05,
+      "loss": 12.1005,
+      "step": 10330
+    },
+    {
+      "epoch": 1.8066448087431692,
+      "grad_norm": 0.8034826197370097,
+      "learning_rate": 7.211920242265366e-05,
+      "loss": 12.2453,
+      "step": 10331
+    },
+    {
+      "epoch": 1.8068196721311476,
+      "grad_norm": 0.8219743457314211,
+      "learning_rate": 7.210107070820014e-05,
+      "loss": 12.2333,
+      "step": 10332
+    },
+    {
+      "epoch": 1.8069945355191257,
+      "grad_norm": 0.9011070814264512,
+      "learning_rate": 7.208293998831085e-05,
+      "loss": 12.2694,
+      "step": 10333
+    },
+    {
+      "epoch": 1.8071693989071038,
+      "grad_norm": 0.826217005994298,
+      "learning_rate": 7.206481026363214e-05,
+      "loss": 12.5788,
+      "step": 10334
+    },
+    {
+      "epoch": 1.807344262295082,
+      "grad_norm": 0.8340134442381696,
+      "learning_rate": 7.204668153481035e-05,
+      "loss": 12.2586,
+      "step": 10335
+    },
+    {
+      "epoch": 1.80751912568306,
+      "grad_norm": 0.862053538821597,
+      "learning_rate": 7.202855380249167e-05,
+      "loss": 12.3928,
+      "step": 10336
+    },
+    {
+      "epoch": 1.8076939890710384,
+      "grad_norm": 0.8575260314451341,
+      "learning_rate": 7.201042706732235e-05,
+      "loss": 12.2594,
+      "step": 10337
+    },
+    {
+      "epoch": 1.8078688524590163,
+      "grad_norm": 0.9007278470334384,
+      "learning_rate": 7.199230132994865e-05,
+      "loss": 12.0935,
+      "step": 10338
+    },
+    {
+      "epoch": 1.8080437158469946,
+      "grad_norm": 0.8301463323546879,
+      "learning_rate": 7.197417659101664e-05,
+      "loss": 12.5881,
+      "step": 10339
+    },
+    {
+      "epoch": 1.8082185792349725,
+      "grad_norm": 1.1063094158191777,
+      "learning_rate": 7.195605285117248e-05,
+      "loss": 12.4239,
+      "step": 10340
+    },
+    {
+      "epoch": 1.8083934426229509,
+      "grad_norm": 0.869617682220201,
+      "learning_rate": 7.193793011106229e-05,
+      "loss": 12.2381,
+      "step": 10341
+    },
+    {
+      "epoch": 1.808568306010929,
+      "grad_norm": 1.1115063538881929,
+      "learning_rate": 7.191980837133213e-05,
+      "loss": 12.395,
+      "step": 10342
+    },
+    {
+      "epoch": 1.8087431693989071,
+      "grad_norm": 0.8909255908147582,
+      "learning_rate": 7.190168763262795e-05,
+      "loss": 12.3524,
+      "step": 10343
+    },
+    {
+      "epoch": 1.8089180327868852,
+      "grad_norm": 0.985906423893338,
+      "learning_rate": 7.188356789559579e-05,
+      "loss": 12.3592,
+      "step": 10344
+    },
+    {
+      "epoch": 1.8090928961748634,
+      "grad_norm": 0.9146424150781998,
+      "learning_rate": 7.186544916088156e-05,
+      "loss": 12.2743,
+      "step": 10345
+    },
+    {
+      "epoch": 1.8092677595628417,
+      "grad_norm": 0.851701808370673,
+      "learning_rate": 7.184733142913127e-05,
+      "loss": 12.4044,
+      "step": 10346
+    },
+    {
+      "epoch": 1.8094426229508196,
+      "grad_norm": 0.8890441768399008,
+      "learning_rate": 7.182921470099066e-05,
+      "loss": 12.0989,
+      "step": 10347
+    },
+    {
+      "epoch": 1.809617486338798,
+      "grad_norm": 0.8355706775027694,
+      "learning_rate": 7.181109897710564e-05,
+      "loss": 12.1809,
+      "step": 10348
+    },
+    {
+      "epoch": 1.8097923497267758,
+      "grad_norm": 0.8304037522989456,
+      "learning_rate": 7.179298425812199e-05,
+      "loss": 12.284,
+      "step": 10349
+    },
+    {
+      "epoch": 1.8099672131147542,
+      "grad_norm": 0.7559889823063336,
+      "learning_rate": 7.177487054468554e-05,
+      "loss": 12.3567,
+      "step": 10350
+    },
+    {
+      "epoch": 1.8101420765027323,
+      "grad_norm": 0.7595064887550593,
+      "learning_rate": 7.175675783744195e-05,
+      "loss": 12.2008,
+      "step": 10351
+    },
+    {
+      "epoch": 1.8103169398907104,
+      "grad_norm": 0.8870730361880169,
+      "learning_rate": 7.173864613703694e-05,
+      "loss": 12.199,
+      "step": 10352
+    },
+    {
+      "epoch": 1.8104918032786885,
+      "grad_norm": 0.8303385114146704,
+      "learning_rate": 7.172053544411622e-05,
+      "loss": 12.431,
+      "step": 10353
+    },
+    {
+      "epoch": 1.8106666666666666,
+      "grad_norm": 0.8701787485764741,
+      "learning_rate": 7.170242575932534e-05,
+      "loss": 12.1725,
+      "step": 10354
+    },
+    {
+      "epoch": 1.8108415300546448,
+      "grad_norm": 0.8280358539188255,
+      "learning_rate": 7.16843170833099e-05,
+      "loss": 12.2642,
+      "step": 10355
+    },
+    {
+      "epoch": 1.8110163934426229,
+      "grad_norm": 0.8631687043115647,
+      "learning_rate": 7.166620941671547e-05,
+      "loss": 12.1309,
+      "step": 10356
+    },
+    {
+      "epoch": 1.8111912568306012,
+      "grad_norm": 0.8610516777978625,
+      "learning_rate": 7.164810276018761e-05,
+      "loss": 12.2981,
+      "step": 10357
+    },
+    {
+      "epoch": 1.8113661202185791,
+      "grad_norm": 0.8124773248408139,
+      "learning_rate": 7.162999711437173e-05,
+      "loss": 12.2579,
+      "step": 10358
+    },
+    {
+      "epoch": 1.8115409836065575,
+      "grad_norm": 0.837409646718785,
+      "learning_rate": 7.161189247991329e-05,
+      "loss": 12.4149,
+      "step": 10359
+    },
+    {
+      "epoch": 1.8117158469945354,
+      "grad_norm": 0.9249687076436647,
+      "learning_rate": 7.159378885745773e-05,
+      "loss": 12.4301,
+      "step": 10360
+    },
+    {
+      "epoch": 1.8118907103825137,
+      "grad_norm": 0.8483212076999986,
+      "learning_rate": 7.157568624765045e-05,
+      "loss": 12.189,
+      "step": 10361
+    },
+    {
+      "epoch": 1.8120655737704918,
+      "grad_norm": 0.7997920356426348,
+      "learning_rate": 7.155758465113671e-05,
+      "loss": 12.3037,
+      "step": 10362
+    },
+    {
+      "epoch": 1.81224043715847,
+      "grad_norm": 0.8997137236171063,
+      "learning_rate": 7.153948406856184e-05,
+      "loss": 12.5762,
+      "step": 10363
+    },
+    {
+      "epoch": 1.812415300546448,
+      "grad_norm": 0.8468783475615723,
+      "learning_rate": 7.152138450057111e-05,
+      "loss": 12.4624,
+      "step": 10364
+    },
+    {
+      "epoch": 1.8125901639344262,
+      "grad_norm": 0.8140957318981709,
+      "learning_rate": 7.150328594780979e-05,
+      "loss": 12.3494,
+      "step": 10365
+    },
+    {
+      "epoch": 1.8127650273224045,
+      "grad_norm": 0.8906842445250763,
+      "learning_rate": 7.1485188410923e-05,
+      "loss": 12.4084,
+      "step": 10366
+    },
+    {
+      "epoch": 1.8129398907103824,
+      "grad_norm": 0.8035500788996435,
+      "learning_rate": 7.146709189055591e-05,
+      "loss": 12.3128,
+      "step": 10367
+    },
+    {
+      "epoch": 1.8131147540983608,
+      "grad_norm": 0.8643415453852973,
+      "learning_rate": 7.14489963873537e-05,
+      "loss": 12.3851,
+      "step": 10368
+    },
+    {
+      "epoch": 1.8132896174863387,
+      "grad_norm": 0.8555174808542164,
+      "learning_rate": 7.143090190196137e-05,
+      "loss": 12.1832,
+      "step": 10369
+    },
+    {
+      "epoch": 1.813464480874317,
+      "grad_norm": 0.7777560377847419,
+      "learning_rate": 7.1412808435024e-05,
+      "loss": 12.4112,
+      "step": 10370
+    },
+    {
+      "epoch": 1.8136393442622951,
+      "grad_norm": 0.8250825488997574,
+      "learning_rate": 7.139471598718663e-05,
+      "loss": 12.2069,
+      "step": 10371
+    },
+    {
+      "epoch": 1.8138142076502732,
+      "grad_norm": 0.9850517617325762,
+      "learning_rate": 7.137662455909424e-05,
+      "loss": 12.1464,
+      "step": 10372
+    },
+    {
+      "epoch": 1.8139890710382514,
+      "grad_norm": 0.8576683372890707,
+      "learning_rate": 7.135853415139171e-05,
+      "loss": 12.2624,
+      "step": 10373
+    },
+    {
+      "epoch": 1.8141639344262295,
+      "grad_norm": 0.9814073541028026,
+      "learning_rate": 7.134044476472397e-05,
+      "loss": 12.2695,
+      "step": 10374
+    },
+    {
+      "epoch": 1.8143387978142078,
+      "grad_norm": 0.9832650197531111,
+      "learning_rate": 7.13223563997359e-05,
+      "loss": 12.3996,
+      "step": 10375
+    },
+    {
+      "epoch": 1.8145136612021857,
+      "grad_norm": 0.9594903261438422,
+      "learning_rate": 7.130426905707232e-05,
+      "loss": 12.3972,
+      "step": 10376
+    },
+    {
+      "epoch": 1.814688524590164,
+      "grad_norm": 0.943872789465858,
+      "learning_rate": 7.128618273737801e-05,
+      "loss": 12.4266,
+      "step": 10377
+    },
+    {
+      "epoch": 1.814863387978142,
+      "grad_norm": 0.9214043844360776,
+      "learning_rate": 7.126809744129774e-05,
+      "loss": 12.4171,
+      "step": 10378
+    },
+    {
+      "epoch": 1.8150382513661203,
+      "grad_norm": 0.9135857002195819,
+      "learning_rate": 7.125001316947622e-05,
+      "loss": 12.317,
+      "step": 10379
+    },
+    {
+      "epoch": 1.8152131147540984,
+      "grad_norm": 0.9288694475650756,
+      "learning_rate": 7.123192992255818e-05,
+      "loss": 12.1585,
+      "step": 10380
+    },
+    {
+      "epoch": 1.8153879781420765,
+      "grad_norm": 0.8906645446203482,
+      "learning_rate": 7.12138477011882e-05,
+      "loss": 12.3258,
+      "step": 10381
+    },
+    {
+      "epoch": 1.8155628415300546,
+      "grad_norm": 0.8351521176549369,
+      "learning_rate": 7.11957665060109e-05,
+      "loss": 12.4144,
+      "step": 10382
+    },
+    {
+      "epoch": 1.8157377049180328,
+      "grad_norm": 1.0571246124491074,
+      "learning_rate": 7.117768633767087e-05,
+      "loss": 12.1388,
+      "step": 10383
+    },
+    {
+      "epoch": 1.8159125683060109,
+      "grad_norm": 0.8807021419428783,
+      "learning_rate": 7.11596071968127e-05,
+      "loss": 12.3337,
+      "step": 10384
+    },
+    {
+      "epoch": 1.816087431693989,
+      "grad_norm": 0.9263355836544078,
+      "learning_rate": 7.114152908408079e-05,
+      "loss": 12.3361,
+      "step": 10385
+    },
+    {
+      "epoch": 1.8162622950819673,
+      "grad_norm": 1.0210017857246936,
+      "learning_rate": 7.112345200011964e-05,
+      "loss": 12.3477,
+      "step": 10386
+    },
+    {
+      "epoch": 1.8164371584699452,
+      "grad_norm": 0.8979348869817056,
+      "learning_rate": 7.110537594557372e-05,
+      "loss": 12.5214,
+      "step": 10387
+    },
+    {
+      "epoch": 1.8166120218579236,
+      "grad_norm": 1.0987160790537842,
+      "learning_rate": 7.108730092108736e-05,
+      "loss": 12.2733,
+      "step": 10388
+    },
+    {
+      "epoch": 1.8167868852459015,
+      "grad_norm": 0.9321828914171868,
+      "learning_rate": 7.106922692730494e-05,
+      "loss": 11.9746,
+      "step": 10389
+    },
+    {
+      "epoch": 1.8169617486338798,
+      "grad_norm": 0.927344023191489,
+      "learning_rate": 7.105115396487078e-05,
+      "loss": 12.2874,
+      "step": 10390
+    },
+    {
+      "epoch": 1.817136612021858,
+      "grad_norm": 0.9878874655233649,
+      "learning_rate": 7.103308203442923e-05,
+      "loss": 12.379,
+      "step": 10391
+    },
+    {
+      "epoch": 1.817311475409836,
+      "grad_norm": 0.8493416845579638,
+      "learning_rate": 7.101501113662438e-05,
+      "loss": 12.4272,
+      "step": 10392
+    },
+    {
+      "epoch": 1.8174863387978142,
+      "grad_norm": 1.0649299421753777,
+      "learning_rate": 7.099694127210053e-05,
+      "loss": 12.5006,
+      "step": 10393
+    },
+    {
+      "epoch": 1.8176612021857923,
+      "grad_norm": 0.8947058715649522,
+      "learning_rate": 7.097887244150182e-05,
+      "loss": 12.5721,
+      "step": 10394
+    },
+    {
+      "epoch": 1.8178360655737706,
+      "grad_norm": 0.8496096789345634,
+      "learning_rate": 7.096080464547246e-05,
+      "loss": 12.3174,
+      "step": 10395
+    },
+    {
+      "epoch": 1.8180109289617485,
+      "grad_norm": 0.9089628102821888,
+      "learning_rate": 7.094273788465644e-05,
+      "loss": 12.1804,
+      "step": 10396
+    },
+    {
+      "epoch": 1.8181857923497269,
+      "grad_norm": 0.9539711524054936,
+      "learning_rate": 7.092467215969787e-05,
+      "loss": 12.4023,
+      "step": 10397
+    },
+    {
+      "epoch": 1.8183606557377048,
+      "grad_norm": 0.8371109186186878,
+      "learning_rate": 7.090660747124077e-05,
+      "loss": 12.386,
+      "step": 10398
+    },
+    {
+      "epoch": 1.8185355191256831,
+      "grad_norm": 0.9673743866340031,
+      "learning_rate": 7.088854381992915e-05,
+      "loss": 12.551,
+      "step": 10399
+    },
+    {
+      "epoch": 1.8187103825136612,
+      "grad_norm": 0.8854912838534904,
+      "learning_rate": 7.087048120640689e-05,
+      "loss": 12.1104,
+      "step": 10400
+    },
+    {
+      "epoch": 1.8188852459016394,
+      "grad_norm": 0.8254734029372828,
+      "learning_rate": 7.085241963131794e-05,
+      "loss": 12.1535,
+      "step": 10401
+    },
+    {
+      "epoch": 1.8190601092896175,
+      "grad_norm": 1.0398932999431256,
+      "learning_rate": 7.083435909530622e-05,
+      "loss": 12.4232,
+      "step": 10402
+    },
+    {
+      "epoch": 1.8192349726775956,
+      "grad_norm": 0.7932677526684231,
+      "learning_rate": 7.081629959901549e-05,
+      "loss": 12.3766,
+      "step": 10403
+    },
+    {
+      "epoch": 1.8194098360655737,
+      "grad_norm": 1.0342166252619942,
+      "learning_rate": 7.079824114308955e-05,
+      "loss": 12.1636,
+      "step": 10404
+    },
+    {
+      "epoch": 1.8195846994535518,
+      "grad_norm": 0.8939304405967597,
+      "learning_rate": 7.078018372817224e-05,
+      "loss": 12.443,
+      "step": 10405
+    },
+    {
+      "epoch": 1.8197595628415302,
+      "grad_norm": 0.8792916938761457,
+      "learning_rate": 7.076212735490726e-05,
+      "loss": 12.4231,
+      "step": 10406
+    },
+    {
+      "epoch": 1.819934426229508,
+      "grad_norm": 0.8927469425495834,
+      "learning_rate": 7.074407202393823e-05,
+      "loss": 12.4188,
+      "step": 10407
+    },
+    {
+      "epoch": 1.8201092896174864,
+      "grad_norm": 0.9712665684763472,
+      "learning_rate": 7.072601773590885e-05,
+      "loss": 12.5669,
+      "step": 10408
+    },
+    {
+      "epoch": 1.8202841530054645,
+      "grad_norm": 0.8797962683527357,
+      "learning_rate": 7.070796449146276e-05,
+      "loss": 12.2454,
+      "step": 10409
+    },
+    {
+      "epoch": 1.8204590163934427,
+      "grad_norm": 0.8807897320241405,
+      "learning_rate": 7.068991229124353e-05,
+      "loss": 12.4196,
+      "step": 10410
+    },
+    {
+      "epoch": 1.8206338797814208,
+      "grad_norm": 0.9624888148563097,
+      "learning_rate": 7.067186113589466e-05,
+      "loss": 12.4881,
+      "step": 10411
+    },
+    {
+      "epoch": 1.820808743169399,
+      "grad_norm": 0.9077002639256008,
+      "learning_rate": 7.065381102605965e-05,
+      "loss": 12.5485,
+      "step": 10412
+    },
+    {
+      "epoch": 1.820983606557377,
+      "grad_norm": 0.7866313607326276,
+      "learning_rate": 7.0635761962382e-05,
+      "loss": 12.3135,
+      "step": 10413
+    },
+    {
+      "epoch": 1.8211584699453551,
+      "grad_norm": 0.9573356693539152,
+      "learning_rate": 7.061771394550516e-05,
+      "loss": 12.3553,
+      "step": 10414
+    },
+    {
+      "epoch": 1.8213333333333335,
+      "grad_norm": 0.7647341921993502,
+      "learning_rate": 7.059966697607246e-05,
+      "loss": 12.3812,
+      "step": 10415
+    },
+    {
+      "epoch": 1.8215081967213114,
+      "grad_norm": 1.0450214386034322,
+      "learning_rate": 7.058162105472728e-05,
+      "loss": 12.1469,
+      "step": 10416
+    },
+    {
+      "epoch": 1.8216830601092897,
+      "grad_norm": 0.9139206413920422,
+      "learning_rate": 7.056357618211298e-05,
+      "loss": 12.2891,
+      "step": 10417
+    },
+    {
+      "epoch": 1.8218579234972676,
+      "grad_norm": 0.9119807083039579,
+      "learning_rate": 7.054553235887276e-05,
+      "loss": 12.3766,
+      "step": 10418
+    },
+    {
+      "epoch": 1.822032786885246,
+      "grad_norm": 0.9465637958538345,
+      "learning_rate": 7.052748958564988e-05,
+      "loss": 12.4194,
+      "step": 10419
+    },
+    {
+      "epoch": 1.822207650273224,
+      "grad_norm": 0.8283277892828732,
+      "learning_rate": 7.050944786308757e-05,
+      "loss": 12.2575,
+      "step": 10420
+    },
+    {
+      "epoch": 1.8223825136612022,
+      "grad_norm": 0.9553343625046395,
+      "learning_rate": 7.049140719182902e-05,
+      "loss": 12.3499,
+      "step": 10421
+    },
+    {
+      "epoch": 1.8225573770491803,
+      "grad_norm": 0.8803914890520061,
+      "learning_rate": 7.047336757251729e-05,
+      "loss": 12.0045,
+      "step": 10422
+    },
+    {
+      "epoch": 1.8227322404371584,
+      "grad_norm": 0.851215715887077,
+      "learning_rate": 7.045532900579552e-05,
+      "loss": 12.2315,
+      "step": 10423
+    },
+    {
+      "epoch": 1.8229071038251368,
+      "grad_norm": 0.8581347412442004,
+      "learning_rate": 7.043729149230675e-05,
+      "loss": 12.4555,
+      "step": 10424
+    },
+    {
+      "epoch": 1.8230819672131147,
+      "grad_norm": 0.9777319894820808,
+      "learning_rate": 7.041925503269404e-05,
+      "loss": 12.2475,
+      "step": 10425
+    },
+    {
+      "epoch": 1.823256830601093,
+      "grad_norm": 0.7879899295698269,
+      "learning_rate": 7.040121962760026e-05,
+      "loss": 12.3321,
+      "step": 10426
+    },
+    {
+      "epoch": 1.823431693989071,
+      "grad_norm": 0.9915115290304952,
+      "learning_rate": 7.038318527766845e-05,
+      "loss": 12.3861,
+      "step": 10427
+    },
+    {
+      "epoch": 1.8236065573770492,
+      "grad_norm": 0.8550358755999851,
+      "learning_rate": 7.036515198354148e-05,
+      "loss": 12.2175,
+      "step": 10428
+    },
+    {
+      "epoch": 1.8237814207650274,
+      "grad_norm": 0.8275761017889598,
+      "learning_rate": 7.034711974586223e-05,
+      "loss": 11.9957,
+      "step": 10429
+    },
+    {
+      "epoch": 1.8239562841530055,
+      "grad_norm": 0.7692474062595089,
+      "learning_rate": 7.03290885652735e-05,
+      "loss": 12.381,
+      "step": 10430
+    },
+    {
+      "epoch": 1.8241311475409836,
+      "grad_norm": 0.7422833399256152,
+      "learning_rate": 7.03110584424181e-05,
+      "loss": 12.3518,
+      "step": 10431
+    },
+    {
+      "epoch": 1.8243060109289617,
+      "grad_norm": 0.8241275497833097,
+      "learning_rate": 7.029302937793877e-05,
+      "loss": 12.3792,
+      "step": 10432
+    },
+    {
+      "epoch": 1.8244808743169398,
+      "grad_norm": 0.8200414296904475,
+      "learning_rate": 7.027500137247825e-05,
+      "loss": 12.1919,
+      "step": 10433
+    },
+    {
+      "epoch": 1.824655737704918,
+      "grad_norm": 0.8694529917226467,
+      "learning_rate": 7.025697442667918e-05,
+      "loss": 12.1207,
+      "step": 10434
+    },
+    {
+      "epoch": 1.8248306010928963,
+      "grad_norm": 0.8186837095417424,
+      "learning_rate": 7.023894854118421e-05,
+      "loss": 12.3171,
+      "step": 10435
+    },
+    {
+      "epoch": 1.8250054644808742,
+      "grad_norm": 0.8566605019305793,
+      "learning_rate": 7.0220923716636e-05,
+      "loss": 12.1878,
+      "step": 10436
+    },
+    {
+      "epoch": 1.8251803278688525,
+      "grad_norm": 0.8461265137285099,
+      "learning_rate": 7.020289995367704e-05,
+      "loss": 12.4617,
+      "step": 10437
+    },
+    {
+      "epoch": 1.8253551912568307,
+      "grad_norm": 0.8258490384699018,
+      "learning_rate": 7.018487725294988e-05,
+      "loss": 12.0537,
+      "step": 10438
+    },
+    {
+      "epoch": 1.8255300546448088,
+      "grad_norm": 0.8967512643469905,
+      "learning_rate": 7.016685561509701e-05,
+      "loss": 12.6829,
+      "step": 10439
+    },
+    {
+      "epoch": 1.825704918032787,
+      "grad_norm": 0.8066536408213533,
+      "learning_rate": 7.014883504076088e-05,
+      "loss": 12.2796,
+      "step": 10440
+    },
+    {
+      "epoch": 1.825879781420765,
+      "grad_norm": 0.8869346412068297,
+      "learning_rate": 7.013081553058392e-05,
+      "loss": 12.4882,
+      "step": 10441
+    },
+    {
+      "epoch": 1.8260546448087431,
+      "grad_norm": 0.8278998729116505,
+      "learning_rate": 7.011279708520846e-05,
+      "loss": 12.2161,
+      "step": 10442
+    },
+    {
+      "epoch": 1.8262295081967213,
+      "grad_norm": 0.8599287201982083,
+      "learning_rate": 7.009477970527688e-05,
+      "loss": 12.2681,
+      "step": 10443
+    },
+    {
+      "epoch": 1.8264043715846996,
+      "grad_norm": 0.9746028985150255,
+      "learning_rate": 7.007676339143149e-05,
+      "loss": 12.1658,
+      "step": 10444
+    },
+    {
+      "epoch": 1.8265792349726775,
+      "grad_norm": 0.8999063348209112,
+      "learning_rate": 7.005874814431448e-05,
+      "loss": 12.4449,
+      "step": 10445
+    },
+    {
+      "epoch": 1.8267540983606558,
+      "grad_norm": 1.0159606592769284,
+      "learning_rate": 7.004073396456813e-05,
+      "loss": 12.2668,
+      "step": 10446
+    },
+    {
+      "epoch": 1.8269289617486337,
+      "grad_norm": 0.8903232331426724,
+      "learning_rate": 7.002272085283463e-05,
+      "loss": 12.3927,
+      "step": 10447
+    },
+    {
+      "epoch": 1.827103825136612,
+      "grad_norm": 0.8066660014105542,
+      "learning_rate": 7.000470880975606e-05,
+      "loss": 12.3625,
+      "step": 10448
+    },
+    {
+      "epoch": 1.8272786885245902,
+      "grad_norm": 0.986503715091503,
+      "learning_rate": 6.998669783597461e-05,
+      "loss": 12.0291,
+      "step": 10449
+    },
+    {
+      "epoch": 1.8274535519125683,
+      "grad_norm": 0.8810656695036704,
+      "learning_rate": 6.996868793213231e-05,
+      "loss": 12.3495,
+      "step": 10450
+    },
+    {
+      "epoch": 1.8276284153005464,
+      "grad_norm": 0.8556422088520595,
+      "learning_rate": 6.995067909887123e-05,
+      "loss": 12.1282,
+      "step": 10451
+    },
+    {
+      "epoch": 1.8278032786885245,
+      "grad_norm": 0.8832076092178609,
+      "learning_rate": 6.993267133683329e-05,
+      "loss": 12.2864,
+      "step": 10452
+    },
+    {
+      "epoch": 1.8279781420765029,
+      "grad_norm": 0.9777176994955169,
+      "learning_rate": 6.991466464666048e-05,
+      "loss": 12.4753,
+      "step": 10453
+    },
+    {
+      "epoch": 1.8281530054644808,
+      "grad_norm": 0.8310561539974617,
+      "learning_rate": 6.989665902899474e-05,
+      "loss": 12.3038,
+      "step": 10454
+    },
+    {
+      "epoch": 1.8283278688524591,
+      "grad_norm": 0.8305113458572507,
+      "learning_rate": 6.987865448447796e-05,
+      "loss": 12.4326,
+      "step": 10455
+    },
+    {
+      "epoch": 1.828502732240437,
+      "grad_norm": 0.9227778562782453,
+      "learning_rate": 6.986065101375194e-05,
+      "loss": 12.4705,
+      "step": 10456
+    },
+    {
+      "epoch": 1.8286775956284154,
+      "grad_norm": 0.8506592007223809,
+      "learning_rate": 6.984264861745846e-05,
+      "loss": 12.2105,
+      "step": 10457
+    },
+    {
+      "epoch": 1.8288524590163935,
+      "grad_norm": 0.8030030517191978,
+      "learning_rate": 6.982464729623934e-05,
+      "loss": 12.3529,
+      "step": 10458
+    },
+    {
+      "epoch": 1.8290273224043716,
+      "grad_norm": 0.9025827843196401,
+      "learning_rate": 6.980664705073632e-05,
+      "loss": 12.0597,
+      "step": 10459
+    },
+    {
+      "epoch": 1.8292021857923497,
+      "grad_norm": 0.9213223196249417,
+      "learning_rate": 6.978864788159102e-05,
+      "loss": 12.4508,
+      "step": 10460
+    },
+    {
+      "epoch": 1.8293770491803278,
+      "grad_norm": 0.9081620850049806,
+      "learning_rate": 6.977064978944511e-05,
+      "loss": 11.9828,
+      "step": 10461
+    },
+    {
+      "epoch": 1.829551912568306,
+      "grad_norm": 0.8108022667089976,
+      "learning_rate": 6.97526527749403e-05,
+      "loss": 12.2059,
+      "step": 10462
+    },
+    {
+      "epoch": 1.829726775956284,
+      "grad_norm": 0.918602903944313,
+      "learning_rate": 6.9734656838718e-05,
+      "loss": 12.3006,
+      "step": 10463
+    },
+    {
+      "epoch": 1.8299016393442624,
+      "grad_norm": 0.8272319652138208,
+      "learning_rate": 6.971666198141984e-05,
+      "loss": 12.1473,
+      "step": 10464
+    },
+    {
+      "epoch": 1.8300765027322403,
+      "grad_norm": 0.8697883591878587,
+      "learning_rate": 6.969866820368728e-05,
+      "loss": 12.1757,
+      "step": 10465
+    },
+    {
+      "epoch": 1.8302513661202187,
+      "grad_norm": 0.8273142072505941,
+      "learning_rate": 6.968067550616184e-05,
+      "loss": 12.2545,
+      "step": 10466
+    },
+    {
+      "epoch": 1.8304262295081966,
+      "grad_norm": 0.828149197845791,
+      "learning_rate": 6.966268388948488e-05,
+      "loss": 12.0841,
+      "step": 10467
+    },
+    {
+      "epoch": 1.830601092896175,
+      "grad_norm": 0.8620958683728057,
+      "learning_rate": 6.964469335429775e-05,
+      "loss": 12.3437,
+      "step": 10468
+    },
+    {
+      "epoch": 1.830775956284153,
+      "grad_norm": 0.8509620979925345,
+      "learning_rate": 6.962670390124186e-05,
+      "loss": 12.2925,
+      "step": 10469
+    },
+    {
+      "epoch": 1.8309508196721311,
+      "grad_norm": 0.8116544503522306,
+      "learning_rate": 6.960871553095854e-05,
+      "loss": 12.1546,
+      "step": 10470
+    },
+    {
+      "epoch": 1.8311256830601093,
+      "grad_norm": 0.8593204132257913,
+      "learning_rate": 6.959072824408895e-05,
+      "loss": 12.2439,
+      "step": 10471
+    },
+    {
+      "epoch": 1.8313005464480874,
+      "grad_norm": 0.8246298690024486,
+      "learning_rate": 6.957274204127438e-05,
+      "loss": 12.3567,
+      "step": 10472
+    },
+    {
+      "epoch": 1.8314754098360657,
+      "grad_norm": 0.8845713756291407,
+      "learning_rate": 6.9554756923156e-05,
+      "loss": 12.5383,
+      "step": 10473
+    },
+    {
+      "epoch": 1.8316502732240436,
+      "grad_norm": 0.7511157933533888,
+      "learning_rate": 6.953677289037498e-05,
+      "loss": 12.3571,
+      "step": 10474
+    },
+    {
+      "epoch": 1.831825136612022,
+      "grad_norm": 0.8221026460730847,
+      "learning_rate": 6.951878994357242e-05,
+      "loss": 12.4525,
+      "step": 10475
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 0.8644741676219417,
+      "learning_rate": 6.950080808338934e-05,
+      "loss": 12.279,
+      "step": 10476
+    },
+    {
+      "epoch": 1.8321748633879782,
+      "grad_norm": 0.7845585181575598,
+      "learning_rate": 6.948282731046687e-05,
+      "loss": 12.0816,
+      "step": 10477
+    },
+    {
+      "epoch": 1.8323497267759563,
+      "grad_norm": 0.8014288004802609,
+      "learning_rate": 6.946484762544592e-05,
+      "loss": 12.3599,
+      "step": 10478
+    },
+    {
+      "epoch": 1.8325245901639344,
+      "grad_norm": 0.8009390476679396,
+      "learning_rate": 6.944686902896747e-05,
+      "loss": 12.2777,
+      "step": 10479
+    },
+    {
+      "epoch": 1.8326994535519125,
+      "grad_norm": 0.8106519059540933,
+      "learning_rate": 6.942889152167244e-05,
+      "loss": 12.4019,
+      "step": 10480
+    },
+    {
+      "epoch": 1.8328743169398907,
+      "grad_norm": 0.8492960328789662,
+      "learning_rate": 6.941091510420177e-05,
+      "loss": 12.4055,
+      "step": 10481
+    },
+    {
+      "epoch": 1.833049180327869,
+      "grad_norm": 0.8874536241178516,
+      "learning_rate": 6.939293977719617e-05,
+      "loss": 12.4239,
+      "step": 10482
+    },
+    {
+      "epoch": 1.833224043715847,
+      "grad_norm": 0.7863302674168603,
+      "learning_rate": 6.937496554129652e-05,
+      "loss": 12.503,
+      "step": 10483
+    },
+    {
+      "epoch": 1.8333989071038252,
+      "grad_norm": 0.8709991089009794,
+      "learning_rate": 6.935699239714357e-05,
+      "loss": 12.3626,
+      "step": 10484
+    },
+    {
+      "epoch": 1.8335737704918031,
+      "grad_norm": 0.8476585190290585,
+      "learning_rate": 6.933902034537805e-05,
+      "loss": 12.415,
+      "step": 10485
+    },
+    {
+      "epoch": 1.8337486338797815,
+      "grad_norm": 0.8658331082350927,
+      "learning_rate": 6.932104938664062e-05,
+      "loss": 12.4374,
+      "step": 10486
+    },
+    {
+      "epoch": 1.8339234972677596,
+      "grad_norm": 0.8122199501204901,
+      "learning_rate": 6.930307952157193e-05,
+      "loss": 12.3014,
+      "step": 10487
+    },
+    {
+      "epoch": 1.8340983606557377,
+      "grad_norm": 0.834577682057897,
+      "learning_rate": 6.928511075081259e-05,
+      "loss": 12.2163,
+      "step": 10488
+    },
+    {
+      "epoch": 1.8342732240437158,
+      "grad_norm": 0.826986874805727,
+      "learning_rate": 6.92671430750032e-05,
+      "loss": 12.2148,
+      "step": 10489
+    },
+    {
+      "epoch": 1.834448087431694,
+      "grad_norm": 0.8328179506190684,
+      "learning_rate": 6.924917649478422e-05,
+      "loss": 12.2751,
+      "step": 10490
+    },
+    {
+      "epoch": 1.834622950819672,
+      "grad_norm": 0.8857092302124824,
+      "learning_rate": 6.923121101079617e-05,
+      "loss": 12.3525,
+      "step": 10491
+    },
+    {
+      "epoch": 1.8347978142076502,
+      "grad_norm": 0.8255270795099259,
+      "learning_rate": 6.921324662367949e-05,
+      "loss": 12.2818,
+      "step": 10492
+    },
+    {
+      "epoch": 1.8349726775956285,
+      "grad_norm": 0.7966541142139357,
+      "learning_rate": 6.919528333407462e-05,
+      "loss": 12.3027,
+      "step": 10493
+    },
+    {
+      "epoch": 1.8351475409836064,
+      "grad_norm": 0.9060870811466809,
+      "learning_rate": 6.91773211426219e-05,
+      "loss": 12.3719,
+      "step": 10494
+    },
+    {
+      "epoch": 1.8353224043715848,
+      "grad_norm": 0.7710983277870033,
+      "learning_rate": 6.915936004996167e-05,
+      "loss": 12.4343,
+      "step": 10495
+    },
+    {
+      "epoch": 1.8354972677595627,
+      "grad_norm": 0.8853414815988397,
+      "learning_rate": 6.914140005673423e-05,
+      "loss": 12.1023,
+      "step": 10496
+    },
+    {
+      "epoch": 1.835672131147541,
+      "grad_norm": 0.8428654328737115,
+      "learning_rate": 6.912344116357981e-05,
+      "loss": 12.3924,
+      "step": 10497
+    },
+    {
+      "epoch": 1.8358469945355191,
+      "grad_norm": 0.8744377656180209,
+      "learning_rate": 6.910548337113863e-05,
+      "loss": 12.0367,
+      "step": 10498
+    },
+    {
+      "epoch": 1.8360218579234973,
+      "grad_norm": 0.7395657102051567,
+      "learning_rate": 6.908752668005086e-05,
+      "loss": 12.2888,
+      "step": 10499
+    },
+    {
+      "epoch": 1.8361967213114754,
+      "grad_norm": 0.9668270917917522,
+      "learning_rate": 6.906957109095671e-05,
+      "loss": 12.3493,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8363715846994535,
+      "grad_norm": 0.8251324334835735,
+      "learning_rate": 6.905161660449615e-05,
+      "loss": 12.1143,
+      "step": 10501
+    },
+    {
+      "epoch": 1.8365464480874318,
+      "grad_norm": 0.8898784419108068,
+      "learning_rate": 6.903366322130931e-05,
+      "loss": 12.2467,
+      "step": 10502
+    },
+    {
+      "epoch": 1.8367213114754097,
+      "grad_norm": 0.8934123804327111,
+      "learning_rate": 6.901571094203619e-05,
+      "loss": 12.091,
+      "step": 10503
+    },
+    {
+      "epoch": 1.836896174863388,
+      "grad_norm": 0.8286821610468502,
+      "learning_rate": 6.89977597673168e-05,
+      "loss": 12.2628,
+      "step": 10504
+    },
+    {
+      "epoch": 1.837071038251366,
+      "grad_norm": 0.9669211989693376,
+      "learning_rate": 6.897980969779103e-05,
+      "loss": 12.2888,
+      "step": 10505
+    },
+    {
+      "epoch": 1.8372459016393443,
+      "grad_norm": 0.8327521404580168,
+      "learning_rate": 6.89618607340988e-05,
+      "loss": 12.312,
+      "step": 10506
+    },
+    {
+      "epoch": 1.8374207650273224,
+      "grad_norm": 0.9297322856815488,
+      "learning_rate": 6.894391287687998e-05,
+      "loss": 12.3835,
+      "step": 10507
+    },
+    {
+      "epoch": 1.8375956284153006,
+      "grad_norm": 0.8162505883449878,
+      "learning_rate": 6.892596612677442e-05,
+      "loss": 12.1851,
+      "step": 10508
+    },
+    {
+      "epoch": 1.8377704918032787,
+      "grad_norm": 0.9088375133033475,
+      "learning_rate": 6.89080204844218e-05,
+      "loss": 12.1562,
+      "step": 10509
+    },
+    {
+      "epoch": 1.8379453551912568,
+      "grad_norm": 0.9839899238423985,
+      "learning_rate": 6.889007595046194e-05,
+      "loss": 12.1843,
+      "step": 10510
+    },
+    {
+      "epoch": 1.8381202185792351,
+      "grad_norm": 0.8761771210445494,
+      "learning_rate": 6.887213252553455e-05,
+      "loss": 12.3085,
+      "step": 10511
+    },
+    {
+      "epoch": 1.838295081967213,
+      "grad_norm": 0.9205004251936182,
+      "learning_rate": 6.885419021027924e-05,
+      "loss": 12.3227,
+      "step": 10512
+    },
+    {
+      "epoch": 1.8384699453551914,
+      "grad_norm": 0.8403637618920609,
+      "learning_rate": 6.883624900533565e-05,
+      "loss": 12.3591,
+      "step": 10513
+    },
+    {
+      "epoch": 1.8386448087431693,
+      "grad_norm": 0.8651555683124033,
+      "learning_rate": 6.881830891134338e-05,
+      "loss": 12.3743,
+      "step": 10514
+    },
+    {
+      "epoch": 1.8388196721311476,
+      "grad_norm": 0.8726767356683347,
+      "learning_rate": 6.8800369928942e-05,
+      "loss": 12.3321,
+      "step": 10515
+    },
+    {
+      "epoch": 1.8389945355191257,
+      "grad_norm": 0.8294357310828436,
+      "learning_rate": 6.878243205877094e-05,
+      "loss": 12.3482,
+      "step": 10516
+    },
+    {
+      "epoch": 1.8391693989071038,
+      "grad_norm": 0.8547378033311738,
+      "learning_rate": 6.876449530146971e-05,
+      "loss": 12.0703,
+      "step": 10517
+    },
+    {
+      "epoch": 1.839344262295082,
+      "grad_norm": 0.9159266276088324,
+      "learning_rate": 6.874655965767773e-05,
+      "loss": 12.1709,
+      "step": 10518
+    },
+    {
+      "epoch": 1.83951912568306,
+      "grad_norm": 0.9068139387307611,
+      "learning_rate": 6.872862512803442e-05,
+      "loss": 12.1985,
+      "step": 10519
+    },
+    {
+      "epoch": 1.8396939890710382,
+      "grad_norm": 0.8207571076228222,
+      "learning_rate": 6.871069171317903e-05,
+      "loss": 12.3514,
+      "step": 10520
+    },
+    {
+      "epoch": 1.8398688524590163,
+      "grad_norm": 0.9170275379361238,
+      "learning_rate": 6.869275941375094e-05,
+      "loss": 12.2282,
+      "step": 10521
+    },
+    {
+      "epoch": 1.8400437158469947,
+      "grad_norm": 0.9749689302214418,
+      "learning_rate": 6.86748282303894e-05,
+      "loss": 12.2851,
+      "step": 10522
+    },
+    {
+      "epoch": 1.8402185792349726,
+      "grad_norm": 0.9000345893304592,
+      "learning_rate": 6.865689816373364e-05,
+      "loss": 12.3539,
+      "step": 10523
+    },
+    {
+      "epoch": 1.840393442622951,
+      "grad_norm": 0.983746829053226,
+      "learning_rate": 6.863896921442283e-05,
+      "loss": 12.5169,
+      "step": 10524
+    },
+    {
+      "epoch": 1.8405683060109288,
+      "grad_norm": 0.8817507601153671,
+      "learning_rate": 6.862104138309613e-05,
+      "loss": 12.4311,
+      "step": 10525
+    },
+    {
+      "epoch": 1.8407431693989071,
+      "grad_norm": 0.8777610191570602,
+      "learning_rate": 6.860311467039268e-05,
+      "loss": 12.2974,
+      "step": 10526
+    },
+    {
+      "epoch": 1.8409180327868853,
+      "grad_norm": 0.8489540220347904,
+      "learning_rate": 6.858518907695146e-05,
+      "loss": 12.3644,
+      "step": 10527
+    },
+    {
+      "epoch": 1.8410928961748634,
+      "grad_norm": 0.7883490257665654,
+      "learning_rate": 6.856726460341157e-05,
+      "loss": 12.4127,
+      "step": 10528
+    },
+    {
+      "epoch": 1.8412677595628415,
+      "grad_norm": 0.8023427632242428,
+      "learning_rate": 6.854934125041194e-05,
+      "loss": 12.1986,
+      "step": 10529
+    },
+    {
+      "epoch": 1.8414426229508196,
+      "grad_norm": 0.80465648480669,
+      "learning_rate": 6.853141901859159e-05,
+      "loss": 12.3659,
+      "step": 10530
+    },
+    {
+      "epoch": 1.841617486338798,
+      "grad_norm": 0.8407646782530662,
+      "learning_rate": 6.851349790858934e-05,
+      "loss": 12.3249,
+      "step": 10531
+    },
+    {
+      "epoch": 1.8417923497267759,
+      "grad_norm": 0.8517748052532418,
+      "learning_rate": 6.849557792104412e-05,
+      "loss": 12.2635,
+      "step": 10532
+    },
+    {
+      "epoch": 1.8419672131147542,
+      "grad_norm": 0.8614240509416563,
+      "learning_rate": 6.847765905659474e-05,
+      "loss": 12.3653,
+      "step": 10533
+    },
+    {
+      "epoch": 1.842142076502732,
+      "grad_norm": 0.8662064775672966,
+      "learning_rate": 6.845974131588002e-05,
+      "loss": 12.4368,
+      "step": 10534
+    },
+    {
+      "epoch": 1.8423169398907104,
+      "grad_norm": 0.8778985842798118,
+      "learning_rate": 6.844182469953863e-05,
+      "loss": 12.3805,
+      "step": 10535
+    },
+    {
+      "epoch": 1.8424918032786886,
+      "grad_norm": 0.8541936460794287,
+      "learning_rate": 6.842390920820932e-05,
+      "loss": 12.4842,
+      "step": 10536
+    },
+    {
+      "epoch": 1.8426666666666667,
+      "grad_norm": 0.8652075581049274,
+      "learning_rate": 6.840599484253076e-05,
+      "loss": 12.3267,
+      "step": 10537
+    },
+    {
+      "epoch": 1.8428415300546448,
+      "grad_norm": 0.8652130438047211,
+      "learning_rate": 6.838808160314161e-05,
+      "loss": 12.4533,
+      "step": 10538
+    },
+    {
+      "epoch": 1.843016393442623,
+      "grad_norm": 0.807361412318147,
+      "learning_rate": 6.837016949068036e-05,
+      "loss": 12.3029,
+      "step": 10539
+    },
+    {
+      "epoch": 1.8431912568306013,
+      "grad_norm": 0.9063240619894546,
+      "learning_rate": 6.835225850578562e-05,
+      "loss": 12.3852,
+      "step": 10540
+    },
+    {
+      "epoch": 1.8433661202185792,
+      "grad_norm": 0.7849673185219948,
+      "learning_rate": 6.833434864909591e-05,
+      "loss": 12.3849,
+      "step": 10541
+    },
+    {
+      "epoch": 1.8435409836065575,
+      "grad_norm": 0.8421951014285121,
+      "learning_rate": 6.831643992124966e-05,
+      "loss": 12.22,
+      "step": 10542
+    },
+    {
+      "epoch": 1.8437158469945354,
+      "grad_norm": 0.8831022107487797,
+      "learning_rate": 6.82985323228853e-05,
+      "loss": 12.1427,
+      "step": 10543
+    },
+    {
+      "epoch": 1.8438907103825137,
+      "grad_norm": 0.8013646391374627,
+      "learning_rate": 6.828062585464123e-05,
+      "loss": 12.3456,
+      "step": 10544
+    },
+    {
+      "epoch": 1.8440655737704919,
+      "grad_norm": 0.8310680049690529,
+      "learning_rate": 6.826272051715583e-05,
+      "loss": 12.4665,
+      "step": 10545
+    },
+    {
+      "epoch": 1.84424043715847,
+      "grad_norm": 0.8903196934762273,
+      "learning_rate": 6.824481631106732e-05,
+      "loss": 12.2453,
+      "step": 10546
+    },
+    {
+      "epoch": 1.844415300546448,
+      "grad_norm": 0.8552933199227096,
+      "learning_rate": 6.8226913237014e-05,
+      "loss": 12.3441,
+      "step": 10547
+    },
+    {
+      "epoch": 1.8445901639344262,
+      "grad_norm": 0.7971743777011189,
+      "learning_rate": 6.820901129563409e-05,
+      "loss": 12.1919,
+      "step": 10548
+    },
+    {
+      "epoch": 1.8447650273224043,
+      "grad_norm": 0.8686848652380336,
+      "learning_rate": 6.819111048756582e-05,
+      "loss": 12.3031,
+      "step": 10549
+    },
+    {
+      "epoch": 1.8449398907103824,
+      "grad_norm": 0.7914515461754862,
+      "learning_rate": 6.817321081344728e-05,
+      "loss": 11.8302,
+      "step": 10550
+    },
+    {
+      "epoch": 1.8451147540983608,
+      "grad_norm": 0.8338322149810359,
+      "learning_rate": 6.815531227391658e-05,
+      "loss": 12.4602,
+      "step": 10551
+    },
+    {
+      "epoch": 1.8452896174863387,
+      "grad_norm": 0.8043127913370328,
+      "learning_rate": 6.81374148696118e-05,
+      "loss": 12.2884,
+      "step": 10552
+    },
+    {
+      "epoch": 1.845464480874317,
+      "grad_norm": 0.8172572572588722,
+      "learning_rate": 6.811951860117099e-05,
+      "loss": 12.4281,
+      "step": 10553
+    },
+    {
+      "epoch": 1.845639344262295,
+      "grad_norm": 0.914135057498397,
+      "learning_rate": 6.810162346923203e-05,
+      "loss": 12.2948,
+      "step": 10554
+    },
+    {
+      "epoch": 1.8458142076502733,
+      "grad_norm": 0.8407204489852937,
+      "learning_rate": 6.808372947443296e-05,
+      "loss": 12.1945,
+      "step": 10555
+    },
+    {
+      "epoch": 1.8459890710382514,
+      "grad_norm": 0.8502111146349697,
+      "learning_rate": 6.806583661741164e-05,
+      "loss": 12.2507,
+      "step": 10556
+    },
+    {
+      "epoch": 1.8461639344262295,
+      "grad_norm": 0.8134207935863995,
+      "learning_rate": 6.804794489880592e-05,
+      "loss": 12.457,
+      "step": 10557
+    },
+    {
+      "epoch": 1.8463387978142076,
+      "grad_norm": 0.8683125676648293,
+      "learning_rate": 6.803005431925363e-05,
+      "loss": 12.1924,
+      "step": 10558
+    },
+    {
+      "epoch": 1.8465136612021857,
+      "grad_norm": 0.9092542166569593,
+      "learning_rate": 6.801216487939255e-05,
+      "loss": 12.2155,
+      "step": 10559
+    },
+    {
+      "epoch": 1.846688524590164,
+      "grad_norm": 0.7930848558421137,
+      "learning_rate": 6.799427657986047e-05,
+      "loss": 12.1761,
+      "step": 10560
+    },
+    {
+      "epoch": 1.846863387978142,
+      "grad_norm": 0.8659831333750595,
+      "learning_rate": 6.797638942129499e-05,
+      "loss": 12.1268,
+      "step": 10561
+    },
+    {
+      "epoch": 1.8470382513661203,
+      "grad_norm": 0.8715582287587573,
+      "learning_rate": 6.79585034043338e-05,
+      "loss": 12.2971,
+      "step": 10562
+    },
+    {
+      "epoch": 1.8472131147540982,
+      "grad_norm": 0.8348640601778209,
+      "learning_rate": 6.794061852961455e-05,
+      "loss": 12.4451,
+      "step": 10563
+    },
+    {
+      "epoch": 1.8473879781420766,
+      "grad_norm": 0.8458810892769101,
+      "learning_rate": 6.792273479777481e-05,
+      "loss": 12.2581,
+      "step": 10564
+    },
+    {
+      "epoch": 1.8475628415300547,
+      "grad_norm": 1.0571011961835413,
+      "learning_rate": 6.790485220945206e-05,
+      "loss": 12.4252,
+      "step": 10565
+    },
+    {
+      "epoch": 1.8477377049180328,
+      "grad_norm": 0.8050770220388342,
+      "learning_rate": 6.788697076528383e-05,
+      "loss": 12.361,
+      "step": 10566
+    },
+    {
+      "epoch": 1.847912568306011,
+      "grad_norm": 0.9756280077747103,
+      "learning_rate": 6.786909046590758e-05,
+      "loss": 12.5049,
+      "step": 10567
+    },
+    {
+      "epoch": 1.848087431693989,
+      "grad_norm": 1.012687458208825,
+      "learning_rate": 6.785121131196073e-05,
+      "loss": 12.4215,
+      "step": 10568
+    },
+    {
+      "epoch": 1.8482622950819674,
+      "grad_norm": 0.8488766645617807,
+      "learning_rate": 6.783333330408061e-05,
+      "loss": 12.3097,
+      "step": 10569
+    },
+    {
+      "epoch": 1.8484371584699453,
+      "grad_norm": 1.3521584047644934,
+      "learning_rate": 6.781545644290459e-05,
+      "loss": 12.2983,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8486120218579236,
+      "grad_norm": 0.8752156478756365,
+      "learning_rate": 6.779758072906998e-05,
+      "loss": 12.0069,
+      "step": 10571
+    },
+    {
+      "epoch": 1.8487868852459015,
+      "grad_norm": 1.07188191155273,
+      "learning_rate": 6.777970616321394e-05,
+      "loss": 12.4705,
+      "step": 10572
+    },
+    {
+      "epoch": 1.8489617486338799,
+      "grad_norm": 1.306680963671546,
+      "learning_rate": 6.776183274597374e-05,
+      "loss": 12.3213,
+      "step": 10573
+    },
+    {
+      "epoch": 1.849136612021858,
+      "grad_norm": 0.8588071506867766,
+      "learning_rate": 6.774396047798652e-05,
+      "loss": 12.4889,
+      "step": 10574
+    },
+    {
+      "epoch": 1.849311475409836,
+      "grad_norm": 1.1091104645651328,
+      "learning_rate": 6.772608935988946e-05,
+      "loss": 12.3766,
+      "step": 10575
+    },
+    {
+      "epoch": 1.8494863387978142,
+      "grad_norm": 0.9237575931845798,
+      "learning_rate": 6.770821939231959e-05,
+      "loss": 12.1766,
+      "step": 10576
+    },
+    {
+      "epoch": 1.8496612021857923,
+      "grad_norm": 0.8783937615332432,
+      "learning_rate": 6.769035057591396e-05,
+      "loss": 12.2951,
+      "step": 10577
+    },
+    {
+      "epoch": 1.8498360655737704,
+      "grad_norm": 1.0872881962722227,
+      "learning_rate": 6.767248291130957e-05,
+      "loss": 12.1206,
+      "step": 10578
+    },
+    {
+      "epoch": 1.8500109289617486,
+      "grad_norm": 0.8483813575815637,
+      "learning_rate": 6.765461639914345e-05,
+      "loss": 12.1747,
+      "step": 10579
+    },
+    {
+      "epoch": 1.850185792349727,
+      "grad_norm": 0.8485856415116522,
+      "learning_rate": 6.763675104005243e-05,
+      "loss": 12.3744,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8503606557377048,
+      "grad_norm": 0.9199106596901012,
+      "learning_rate": 6.761888683467341e-05,
+      "loss": 12.3659,
+      "step": 10581
+    },
+    {
+      "epoch": 1.8505355191256831,
+      "grad_norm": 0.8325365462827281,
+      "learning_rate": 6.760102378364323e-05,
+      "loss": 12.3924,
+      "step": 10582
+    },
+    {
+      "epoch": 1.850710382513661,
+      "grad_norm": 0.8355717494792648,
+      "learning_rate": 6.758316188759876e-05,
+      "loss": 12.2328,
+      "step": 10583
+    },
+    {
+      "epoch": 1.8508852459016394,
+      "grad_norm": 0.8422758842993864,
+      "learning_rate": 6.756530114717665e-05,
+      "loss": 12.2432,
+      "step": 10584
+    },
+    {
+      "epoch": 1.8510601092896175,
+      "grad_norm": 0.849461221449101,
+      "learning_rate": 6.754744156301366e-05,
+      "loss": 12.1724,
+      "step": 10585
+    },
+    {
+      "epoch": 1.8512349726775956,
+      "grad_norm": 0.8115712590090164,
+      "learning_rate": 6.752958313574646e-05,
+      "loss": 12.464,
+      "step": 10586
+    },
+    {
+      "epoch": 1.8514098360655737,
+      "grad_norm": 0.8399964712498778,
+      "learning_rate": 6.751172586601169e-05,
+      "loss": 12.3532,
+      "step": 10587
+    },
+    {
+      "epoch": 1.8515846994535519,
+      "grad_norm": 0.8825690596170592,
+      "learning_rate": 6.749386975444593e-05,
+      "loss": 12.0698,
+      "step": 10588
+    },
+    {
+      "epoch": 1.8517595628415302,
+      "grad_norm": 0.8494701964860599,
+      "learning_rate": 6.747601480168573e-05,
+      "loss": 12.4342,
+      "step": 10589
+    },
+    {
+      "epoch": 1.851934426229508,
+      "grad_norm": 0.8747738053734121,
+      "learning_rate": 6.745816100836764e-05,
+      "loss": 12.4111,
+      "step": 10590
+    },
+    {
+      "epoch": 1.8521092896174864,
+      "grad_norm": 0.9931060229197741,
+      "learning_rate": 6.744030837512805e-05,
+      "loss": 12.3396,
+      "step": 10591
+    },
+    {
+      "epoch": 1.8522841530054643,
+      "grad_norm": 0.8569104165915573,
+      "learning_rate": 6.742245690260342e-05,
+      "loss": 12.3101,
+      "step": 10592
+    },
+    {
+      "epoch": 1.8524590163934427,
+      "grad_norm": 0.7594265290955434,
+      "learning_rate": 6.740460659143013e-05,
+      "loss": 12.4149,
+      "step": 10593
+    },
+    {
+      "epoch": 1.8526338797814208,
+      "grad_norm": 0.8948763205764393,
+      "learning_rate": 6.738675744224458e-05,
+      "loss": 12.2749,
+      "step": 10594
+    },
+    {
+      "epoch": 1.852808743169399,
+      "grad_norm": 0.8974498332465565,
+      "learning_rate": 6.736890945568297e-05,
+      "loss": 12.5933,
+      "step": 10595
+    },
+    {
+      "epoch": 1.852983606557377,
+      "grad_norm": 0.9010171465753564,
+      "learning_rate": 6.735106263238163e-05,
+      "loss": 12.122,
+      "step": 10596
+    },
+    {
+      "epoch": 1.8531584699453552,
+      "grad_norm": 0.7914498501275443,
+      "learning_rate": 6.733321697297676e-05,
+      "loss": 12.3819,
+      "step": 10597
+    },
+    {
+      "epoch": 1.8533333333333335,
+      "grad_norm": 0.8642431838954521,
+      "learning_rate": 6.731537247810458e-05,
+      "loss": 12.3778,
+      "step": 10598
+    },
+    {
+      "epoch": 1.8535081967213114,
+      "grad_norm": 0.9793071382593393,
+      "learning_rate": 6.729752914840115e-05,
+      "loss": 12.3649,
+      "step": 10599
+    },
+    {
+      "epoch": 1.8536830601092897,
+      "grad_norm": 0.8531735770562499,
+      "learning_rate": 6.727968698450257e-05,
+      "loss": 12.2039,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8538579234972676,
+      "grad_norm": 0.8153159315184321,
+      "learning_rate": 6.726184598704495e-05,
+      "loss": 12.4353,
+      "step": 10601
+    },
+    {
+      "epoch": 1.854032786885246,
+      "grad_norm": 0.8547485069010969,
+      "learning_rate": 6.724400615666428e-05,
+      "loss": 12.2676,
+      "step": 10602
+    },
+    {
+      "epoch": 1.854207650273224,
+      "grad_norm": 0.8817251888024591,
+      "learning_rate": 6.722616749399649e-05,
+      "loss": 12.4208,
+      "step": 10603
+    },
+    {
+      "epoch": 1.8543825136612022,
+      "grad_norm": 0.7929368060103341,
+      "learning_rate": 6.720832999967754e-05,
+      "loss": 12.3469,
+      "step": 10604
+    },
+    {
+      "epoch": 1.8545573770491803,
+      "grad_norm": 0.9307606691747488,
+      "learning_rate": 6.719049367434336e-05,
+      "loss": 12.4712,
+      "step": 10605
+    },
+    {
+      "epoch": 1.8547322404371585,
+      "grad_norm": 0.9985181946854012,
+      "learning_rate": 6.717265851862971e-05,
+      "loss": 12.4448,
+      "step": 10606
+    },
+    {
+      "epoch": 1.8549071038251366,
+      "grad_norm": 0.832120258968788,
+      "learning_rate": 6.715482453317243e-05,
+      "loss": 12.0459,
+      "step": 10607
+    },
+    {
+      "epoch": 1.8550819672131147,
+      "grad_norm": 0.8766006612167874,
+      "learning_rate": 6.713699171860726e-05,
+      "loss": 12.2551,
+      "step": 10608
+    },
+    {
+      "epoch": 1.855256830601093,
+      "grad_norm": 0.8597852838707307,
+      "learning_rate": 6.711916007556999e-05,
+      "loss": 12.2963,
+      "step": 10609
+    },
+    {
+      "epoch": 1.855431693989071,
+      "grad_norm": 0.8077671924334235,
+      "learning_rate": 6.710132960469622e-05,
+      "loss": 12.2878,
+      "step": 10610
+    },
+    {
+      "epoch": 1.8556065573770493,
+      "grad_norm": 0.8379840964878651,
+      "learning_rate": 6.708350030662158e-05,
+      "loss": 12.3145,
+      "step": 10611
+    },
+    {
+      "epoch": 1.8557814207650272,
+      "grad_norm": 0.872053543217442,
+      "learning_rate": 6.706567218198172e-05,
+      "loss": 12.5222,
+      "step": 10612
+    },
+    {
+      "epoch": 1.8559562841530055,
+      "grad_norm": 0.8779861817624293,
+      "learning_rate": 6.704784523141219e-05,
+      "loss": 12.289,
+      "step": 10613
+    },
+    {
+      "epoch": 1.8561311475409836,
+      "grad_norm": 0.875213881455269,
+      "learning_rate": 6.703001945554845e-05,
+      "loss": 12.4224,
+      "step": 10614
+    },
+    {
+      "epoch": 1.8563060109289617,
+      "grad_norm": 0.9348117622354503,
+      "learning_rate": 6.701219485502598e-05,
+      "loss": 12.1649,
+      "step": 10615
+    },
+    {
+      "epoch": 1.8564808743169399,
+      "grad_norm": 0.9235369470273656,
+      "learning_rate": 6.699437143048023e-05,
+      "loss": 12.1545,
+      "step": 10616
+    },
+    {
+      "epoch": 1.856655737704918,
+      "grad_norm": 0.912461233087628,
+      "learning_rate": 6.697654918254664e-05,
+      "loss": 12.5488,
+      "step": 10617
+    },
+    {
+      "epoch": 1.8568306010928963,
+      "grad_norm": 0.8090178999255551,
+      "learning_rate": 6.695872811186041e-05,
+      "loss": 12.1528,
+      "step": 10618
+    },
+    {
+      "epoch": 1.8570054644808742,
+      "grad_norm": 0.8026913310673309,
+      "learning_rate": 6.694090821905695e-05,
+      "loss": 12.3575,
+      "step": 10619
+    },
+    {
+      "epoch": 1.8571803278688526,
+      "grad_norm": 0.7998469820322193,
+      "learning_rate": 6.692308950477148e-05,
+      "loss": 12.1341,
+      "step": 10620
+    },
+    {
+      "epoch": 1.8573551912568305,
+      "grad_norm": 0.8054379987237197,
+      "learning_rate": 6.690527196963923e-05,
+      "loss": 11.958,
+      "step": 10621
+    },
+    {
+      "epoch": 1.8575300546448088,
+      "grad_norm": 0.8501088532050544,
+      "learning_rate": 6.688745561429534e-05,
+      "loss": 12.2555,
+      "step": 10622
+    },
+    {
+      "epoch": 1.857704918032787,
+      "grad_norm": 0.8792315013320382,
+      "learning_rate": 6.686964043937498e-05,
+      "loss": 12.3458,
+      "step": 10623
+    },
+    {
+      "epoch": 1.857879781420765,
+      "grad_norm": 0.8063836695001297,
+      "learning_rate": 6.685182644551329e-05,
+      "loss": 12.4233,
+      "step": 10624
+    },
+    {
+      "epoch": 1.8580546448087432,
+      "grad_norm": 0.9297123113529661,
+      "learning_rate": 6.683401363334519e-05,
+      "loss": 12.5362,
+      "step": 10625
+    },
+    {
+      "epoch": 1.8582295081967213,
+      "grad_norm": 0.890564712367902,
+      "learning_rate": 6.681620200350575e-05,
+      "loss": 12.3489,
+      "step": 10626
+    },
+    {
+      "epoch": 1.8584043715846994,
+      "grad_norm": 0.8328363669905767,
+      "learning_rate": 6.679839155662997e-05,
+      "loss": 12.0023,
+      "step": 10627
+    },
+    {
+      "epoch": 1.8585792349726775,
+      "grad_norm": 0.9627972802188004,
+      "learning_rate": 6.678058229335276e-05,
+      "loss": 12.3238,
+      "step": 10628
+    },
+    {
+      "epoch": 1.8587540983606559,
+      "grad_norm": 0.876037574405016,
+      "learning_rate": 6.676277421430894e-05,
+      "loss": 12.2791,
+      "step": 10629
+    },
+    {
+      "epoch": 1.8589289617486338,
+      "grad_norm": 0.8973058494017542,
+      "learning_rate": 6.674496732013337e-05,
+      "loss": 12.3767,
+      "step": 10630
+    },
+    {
+      "epoch": 1.859103825136612,
+      "grad_norm": 0.9646461835845513,
+      "learning_rate": 6.672716161146088e-05,
+      "loss": 12.4192,
+      "step": 10631
+    },
+    {
+      "epoch": 1.8592786885245902,
+      "grad_norm": 0.8178665591735637,
+      "learning_rate": 6.670935708892623e-05,
+      "loss": 12.1308,
+      "step": 10632
+    },
+    {
+      "epoch": 1.8594535519125683,
+      "grad_norm": 0.8280258962007468,
+      "learning_rate": 6.669155375316407e-05,
+      "loss": 12.115,
+      "step": 10633
+    },
+    {
+      "epoch": 1.8596284153005465,
+      "grad_norm": 0.8451095377826365,
+      "learning_rate": 6.66737516048091e-05,
+      "loss": 12.1692,
+      "step": 10634
+    },
+    {
+      "epoch": 1.8598032786885246,
+      "grad_norm": 0.9871963271138425,
+      "learning_rate": 6.665595064449599e-05,
+      "loss": 12.2791,
+      "step": 10635
+    },
+    {
+      "epoch": 1.8599781420765027,
+      "grad_norm": 0.8305379643491608,
+      "learning_rate": 6.663815087285924e-05,
+      "loss": 11.9773,
+      "step": 10636
+    },
+    {
+      "epoch": 1.8601530054644808,
+      "grad_norm": 0.8634557877793637,
+      "learning_rate": 6.662035229053344e-05,
+      "loss": 12.2688,
+      "step": 10637
+    },
+    {
+      "epoch": 1.8603278688524592,
+      "grad_norm": 0.962052545598278,
+      "learning_rate": 6.660255489815309e-05,
+      "loss": 12.1287,
+      "step": 10638
+    },
+    {
+      "epoch": 1.860502732240437,
+      "grad_norm": 0.8550874297353687,
+      "learning_rate": 6.658475869635264e-05,
+      "loss": 12.1238,
+      "step": 10639
+    },
+    {
+      "epoch": 1.8606775956284154,
+      "grad_norm": 0.9297096776795786,
+      "learning_rate": 6.656696368576647e-05,
+      "loss": 12.2017,
+      "step": 10640
+    },
+    {
+      "epoch": 1.8608524590163933,
+      "grad_norm": 0.7990644952043101,
+      "learning_rate": 6.654916986702899e-05,
+      "loss": 12.3731,
+      "step": 10641
+    },
+    {
+      "epoch": 1.8610273224043716,
+      "grad_norm": 0.9325382816035438,
+      "learning_rate": 6.653137724077454e-05,
+      "loss": 12.1876,
+      "step": 10642
+    },
+    {
+      "epoch": 1.8612021857923498,
+      "grad_norm": 0.8724917848246369,
+      "learning_rate": 6.651358580763742e-05,
+      "loss": 12.1452,
+      "step": 10643
+    },
+    {
+      "epoch": 1.8613770491803279,
+      "grad_norm": 1.009730440587401,
+      "learning_rate": 6.649579556825179e-05,
+      "loss": 12.1335,
+      "step": 10644
+    },
+    {
+      "epoch": 1.861551912568306,
+      "grad_norm": 0.868013814649603,
+      "learning_rate": 6.647800652325192e-05,
+      "loss": 12.3185,
+      "step": 10645
+    },
+    {
+      "epoch": 1.861726775956284,
+      "grad_norm": 0.9037677154022569,
+      "learning_rate": 6.646021867327192e-05,
+      "loss": 12.2098,
+      "step": 10646
+    },
+    {
+      "epoch": 1.8619016393442624,
+      "grad_norm": 0.7774280342168283,
+      "learning_rate": 6.644243201894599e-05,
+      "loss": 12.2372,
+      "step": 10647
+    },
+    {
+      "epoch": 1.8620765027322403,
+      "grad_norm": 0.8848731487724426,
+      "learning_rate": 6.642464656090812e-05,
+      "loss": 12.266,
+      "step": 10648
+    },
+    {
+      "epoch": 1.8622513661202187,
+      "grad_norm": 0.8307803466059175,
+      "learning_rate": 6.640686229979236e-05,
+      "loss": 12.0765,
+      "step": 10649
+    },
+    {
+      "epoch": 1.8624262295081966,
+      "grad_norm": 0.8780936612636244,
+      "learning_rate": 6.638907923623276e-05,
+      "loss": 12.3541,
+      "step": 10650
+    },
+    {
+      "epoch": 1.862601092896175,
+      "grad_norm": 0.812206725876964,
+      "learning_rate": 6.637129737086318e-05,
+      "loss": 12.3111,
+      "step": 10651
+    },
+    {
+      "epoch": 1.862775956284153,
+      "grad_norm": 0.8060766891243994,
+      "learning_rate": 6.635351670431755e-05,
+      "loss": 12.284,
+      "step": 10652
+    },
+    {
+      "epoch": 1.8629508196721312,
+      "grad_norm": 0.8540626724054814,
+      "learning_rate": 6.633573723722974e-05,
+      "loss": 12.2201,
+      "step": 10653
+    },
+    {
+      "epoch": 1.8631256830601093,
+      "grad_norm": 0.9120896860027538,
+      "learning_rate": 6.631795897023361e-05,
+      "loss": 12.3657,
+      "step": 10654
+    },
+    {
+      "epoch": 1.8633005464480874,
+      "grad_norm": 0.8928354574384839,
+      "learning_rate": 6.630018190396284e-05,
+      "loss": 12.4518,
+      "step": 10655
+    },
+    {
+      "epoch": 1.8634754098360655,
+      "grad_norm": 0.9123242315499519,
+      "learning_rate": 6.628240603905122e-05,
+      "loss": 12.1583,
+      "step": 10656
+    },
+    {
+      "epoch": 1.8636502732240436,
+      "grad_norm": 0.7948486602154636,
+      "learning_rate": 6.626463137613242e-05,
+      "loss": 12.2836,
+      "step": 10657
+    },
+    {
+      "epoch": 1.863825136612022,
+      "grad_norm": 0.8509749497734365,
+      "learning_rate": 6.624685791584014e-05,
+      "loss": 12.3734,
+      "step": 10658
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 0.885528073327053,
+      "learning_rate": 6.622908565880788e-05,
+      "loss": 12.4851,
+      "step": 10659
+    },
+    {
+      "epoch": 1.8641748633879782,
+      "grad_norm": 0.8808370467109746,
+      "learning_rate": 6.621131460566928e-05,
+      "loss": 12.1464,
+      "step": 10660
+    },
+    {
+      "epoch": 1.8643497267759563,
+      "grad_norm": 0.8410311478341113,
+      "learning_rate": 6.619354475705783e-05,
+      "loss": 12.2696,
+      "step": 10661
+    },
+    {
+      "epoch": 1.8645245901639345,
+      "grad_norm": 0.8685644188920065,
+      "learning_rate": 6.617577611360707e-05,
+      "loss": 12.2933,
+      "step": 10662
+    },
+    {
+      "epoch": 1.8646994535519126,
+      "grad_norm": 0.827861565815456,
+      "learning_rate": 6.615800867595032e-05,
+      "loss": 12.0763,
+      "step": 10663
+    },
+    {
+      "epoch": 1.8648743169398907,
+      "grad_norm": 0.842813173101257,
+      "learning_rate": 6.614024244472101e-05,
+      "loss": 12.4002,
+      "step": 10664
+    },
+    {
+      "epoch": 1.8650491803278688,
+      "grad_norm": 0.8169162759273408,
+      "learning_rate": 6.612247742055253e-05,
+      "loss": 12.2151,
+      "step": 10665
+    },
+    {
+      "epoch": 1.865224043715847,
+      "grad_norm": 0.8323859023808404,
+      "learning_rate": 6.610471360407813e-05,
+      "loss": 12.301,
+      "step": 10666
+    },
+    {
+      "epoch": 1.8653989071038253,
+      "grad_norm": 0.8483395701788813,
+      "learning_rate": 6.608695099593107e-05,
+      "loss": 12.3702,
+      "step": 10667
+    },
+    {
+      "epoch": 1.8655737704918032,
+      "grad_norm": 0.8130072134351058,
+      "learning_rate": 6.60691895967446e-05,
+      "loss": 12.3068,
+      "step": 10668
+    },
+    {
+      "epoch": 1.8657486338797815,
+      "grad_norm": 0.8655347249707372,
+      "learning_rate": 6.605142940715192e-05,
+      "loss": 12.2305,
+      "step": 10669
+    },
+    {
+      "epoch": 1.8659234972677594,
+      "grad_norm": 0.7898642160995322,
+      "learning_rate": 6.603367042778606e-05,
+      "loss": 12.1365,
+      "step": 10670
+    },
+    {
+      "epoch": 1.8660983606557378,
+      "grad_norm": 0.7915674211667713,
+      "learning_rate": 6.601591265928018e-05,
+      "loss": 11.8762,
+      "step": 10671
+    },
+    {
+      "epoch": 1.8662732240437159,
+      "grad_norm": 0.8174452794074722,
+      "learning_rate": 6.599815610226729e-05,
+      "loss": 12.3379,
+      "step": 10672
+    },
+    {
+      "epoch": 1.866448087431694,
+      "grad_norm": 0.8582684221996301,
+      "learning_rate": 6.598040075738048e-05,
+      "loss": 12.3151,
+      "step": 10673
+    },
+    {
+      "epoch": 1.866622950819672,
+      "grad_norm": 0.8341150486851976,
+      "learning_rate": 6.596264662525257e-05,
+      "loss": 12.1969,
+      "step": 10674
+    },
+    {
+      "epoch": 1.8667978142076502,
+      "grad_norm": 0.8084620925589945,
+      "learning_rate": 6.594489370651654e-05,
+      "loss": 12.2984,
+      "step": 10675
+    },
+    {
+      "epoch": 1.8669726775956286,
+      "grad_norm": 0.8796381712153659,
+      "learning_rate": 6.592714200180526e-05,
+      "loss": 12.426,
+      "step": 10676
+    },
+    {
+      "epoch": 1.8671475409836065,
+      "grad_norm": 0.7659144379234968,
+      "learning_rate": 6.590939151175159e-05,
+      "loss": 12.2812,
+      "step": 10677
+    },
+    {
+      "epoch": 1.8673224043715848,
+      "grad_norm": 0.8901139140886518,
+      "learning_rate": 6.589164223698825e-05,
+      "loss": 12.4081,
+      "step": 10678
+    },
+    {
+      "epoch": 1.8674972677595627,
+      "grad_norm": 0.8218752908629163,
+      "learning_rate": 6.587389417814803e-05,
+      "loss": 12.2338,
+      "step": 10679
+    },
+    {
+      "epoch": 1.867672131147541,
+      "grad_norm": 0.7873179958344586,
+      "learning_rate": 6.585614733586363e-05,
+      "loss": 12.4192,
+      "step": 10680
+    },
+    {
+      "epoch": 1.8678469945355192,
+      "grad_norm": 0.7962777919371647,
+      "learning_rate": 6.583840171076765e-05,
+      "loss": 12.3676,
+      "step": 10681
+    },
+    {
+      "epoch": 1.8680218579234973,
+      "grad_norm": 0.8203267954557086,
+      "learning_rate": 6.582065730349274e-05,
+      "loss": 12.2802,
+      "step": 10682
+    },
+    {
+      "epoch": 1.8681967213114754,
+      "grad_norm": 0.845607118884225,
+      "learning_rate": 6.580291411467145e-05,
+      "loss": 12.2041,
+      "step": 10683
+    },
+    {
+      "epoch": 1.8683715846994535,
+      "grad_norm": 1.0006474387970228,
+      "learning_rate": 6.578517214493636e-05,
+      "loss": 12.1889,
+      "step": 10684
+    },
+    {
+      "epoch": 1.8685464480874316,
+      "grad_norm": 0.8541526890923835,
+      "learning_rate": 6.576743139491986e-05,
+      "loss": 12.3548,
+      "step": 10685
+    },
+    {
+      "epoch": 1.8687213114754098,
+      "grad_norm": 0.855935116528316,
+      "learning_rate": 6.574969186525446e-05,
+      "loss": 12.0423,
+      "step": 10686
+    },
+    {
+      "epoch": 1.868896174863388,
+      "grad_norm": 0.9899142067775027,
+      "learning_rate": 6.573195355657251e-05,
+      "loss": 12.0007,
+      "step": 10687
+    },
+    {
+      "epoch": 1.869071038251366,
+      "grad_norm": 0.8490265276307096,
+      "learning_rate": 6.571421646950642e-05,
+      "loss": 12.1903,
+      "step": 10688
+    },
+    {
+      "epoch": 1.8692459016393443,
+      "grad_norm": 0.9907930113337557,
+      "learning_rate": 6.569648060468843e-05,
+      "loss": 12.3079,
+      "step": 10689
+    },
+    {
+      "epoch": 1.8694207650273222,
+      "grad_norm": 0.938359046530837,
+      "learning_rate": 6.567874596275082e-05,
+      "loss": 12.2771,
+      "step": 10690
+    },
+    {
+      "epoch": 1.8695956284153006,
+      "grad_norm": 0.8054138393559677,
+      "learning_rate": 6.56610125443258e-05,
+      "loss": 12.0948,
+      "step": 10691
+    },
+    {
+      "epoch": 1.8697704918032787,
+      "grad_norm": 0.8954235114182872,
+      "learning_rate": 6.564328035004558e-05,
+      "loss": 12.3762,
+      "step": 10692
+    },
+    {
+      "epoch": 1.8699453551912568,
+      "grad_norm": 0.8440231142263578,
+      "learning_rate": 6.562554938054229e-05,
+      "loss": 12.3469,
+      "step": 10693
+    },
+    {
+      "epoch": 1.870120218579235,
+      "grad_norm": 0.7981594053278486,
+      "learning_rate": 6.560781963644797e-05,
+      "loss": 12.0608,
+      "step": 10694
+    },
+    {
+      "epoch": 1.870295081967213,
+      "grad_norm": 0.8790147924438338,
+      "learning_rate": 6.559009111839476e-05,
+      "loss": 12.2292,
+      "step": 10695
+    },
+    {
+      "epoch": 1.8704699453551914,
+      "grad_norm": 1.0042569881723327,
+      "learning_rate": 6.557236382701454e-05,
+      "loss": 12.3442,
+      "step": 10696
+    },
+    {
+      "epoch": 1.8706448087431693,
+      "grad_norm": 0.9035937761243358,
+      "learning_rate": 6.555463776293933e-05,
+      "loss": 12.4901,
+      "step": 10697
+    },
+    {
+      "epoch": 1.8708196721311476,
+      "grad_norm": 0.872933214897039,
+      "learning_rate": 6.553691292680104e-05,
+      "loss": 12.2608,
+      "step": 10698
+    },
+    {
+      "epoch": 1.8709945355191255,
+      "grad_norm": 1.1852426916614782,
+      "learning_rate": 6.551918931923158e-05,
+      "loss": 12.4893,
+      "step": 10699
+    },
+    {
+      "epoch": 1.8711693989071039,
+      "grad_norm": 0.8392321130282046,
+      "learning_rate": 6.550146694086268e-05,
+      "loss": 12.1646,
+      "step": 10700
+    },
+    {
+      "epoch": 1.871344262295082,
+      "grad_norm": 0.9620823974730051,
+      "learning_rate": 6.54837457923262e-05,
+      "loss": 12.3222,
+      "step": 10701
+    },
+    {
+      "epoch": 1.8715191256830601,
+      "grad_norm": 0.8804490412919502,
+      "learning_rate": 6.546602587425383e-05,
+      "loss": 12.1761,
+      "step": 10702
+    },
+    {
+      "epoch": 1.8716939890710382,
+      "grad_norm": 0.8574826078806529,
+      "learning_rate": 6.544830718727733e-05,
+      "loss": 12.664,
+      "step": 10703
+    },
+    {
+      "epoch": 1.8718688524590164,
+      "grad_norm": 0.9212453070609018,
+      "learning_rate": 6.543058973202828e-05,
+      "loss": 12.1371,
+      "step": 10704
+    },
+    {
+      "epoch": 1.8720437158469947,
+      "grad_norm": 0.9524750749184752,
+      "learning_rate": 6.541287350913832e-05,
+      "loss": 12.5376,
+      "step": 10705
+    },
+    {
+      "epoch": 1.8722185792349726,
+      "grad_norm": 0.7829864438088935,
+      "learning_rate": 6.5395158519239e-05,
+      "loss": 12.1735,
+      "step": 10706
+    },
+    {
+      "epoch": 1.872393442622951,
+      "grad_norm": 0.9822613618677235,
+      "learning_rate": 6.537744476296189e-05,
+      "loss": 12.2731,
+      "step": 10707
+    },
+    {
+      "epoch": 1.8725683060109288,
+      "grad_norm": 0.9108804567467401,
+      "learning_rate": 6.535973224093838e-05,
+      "loss": 12.3564,
+      "step": 10708
+    },
+    {
+      "epoch": 1.8727431693989072,
+      "grad_norm": 0.8945636840192817,
+      "learning_rate": 6.534202095379996e-05,
+      "loss": 12.3125,
+      "step": 10709
+    },
+    {
+      "epoch": 1.8729180327868853,
+      "grad_norm": 0.9952794153797375,
+      "learning_rate": 6.532431090217799e-05,
+      "loss": 12.2416,
+      "step": 10710
+    },
+    {
+      "epoch": 1.8730928961748634,
+      "grad_norm": 0.8748731360819888,
+      "learning_rate": 6.530660208670383e-05,
+      "loss": 12.1099,
+      "step": 10711
+    },
+    {
+      "epoch": 1.8732677595628415,
+      "grad_norm": 0.9127869480814272,
+      "learning_rate": 6.528889450800876e-05,
+      "loss": 12.4124,
+      "step": 10712
+    },
+    {
+      "epoch": 1.8734426229508196,
+      "grad_norm": 0.9130532569009562,
+      "learning_rate": 6.527118816672406e-05,
+      "loss": 12.5307,
+      "step": 10713
+    },
+    {
+      "epoch": 1.8736174863387978,
+      "grad_norm": 0.7994479476276319,
+      "learning_rate": 6.525348306348093e-05,
+      "loss": 12.2791,
+      "step": 10714
+    },
+    {
+      "epoch": 1.8737923497267759,
+      "grad_norm": 0.9676296761291234,
+      "learning_rate": 6.523577919891053e-05,
+      "loss": 12.4419,
+      "step": 10715
+    },
+    {
+      "epoch": 1.8739672131147542,
+      "grad_norm": 0.8954899089016255,
+      "learning_rate": 6.521807657364396e-05,
+      "loss": 12.217,
+      "step": 10716
+    },
+    {
+      "epoch": 1.8741420765027321,
+      "grad_norm": 0.913126420215513,
+      "learning_rate": 6.520037518831233e-05,
+      "loss": 12.2546,
+      "step": 10717
+    },
+    {
+      "epoch": 1.8743169398907105,
+      "grad_norm": 0.9150746463730837,
+      "learning_rate": 6.518267504354671e-05,
+      "loss": 12.1453,
+      "step": 10718
+    },
+    {
+      "epoch": 1.8744918032786884,
+      "grad_norm": 0.8555668340660046,
+      "learning_rate": 6.5164976139978e-05,
+      "loss": 12.2796,
+      "step": 10719
+    },
+    {
+      "epoch": 1.8746666666666667,
+      "grad_norm": 0.84622931836164,
+      "learning_rate": 6.514727847823718e-05,
+      "loss": 12.3024,
+      "step": 10720
+    },
+    {
+      "epoch": 1.8748415300546448,
+      "grad_norm": 0.8713699154881622,
+      "learning_rate": 6.512958205895517e-05,
+      "loss": 12.3672,
+      "step": 10721
+    },
+    {
+      "epoch": 1.875016393442623,
+      "grad_norm": 0.8572154031613081,
+      "learning_rate": 6.511188688276283e-05,
+      "loss": 12.2297,
+      "step": 10722
+    },
+    {
+      "epoch": 1.875191256830601,
+      "grad_norm": 0.850992688366582,
+      "learning_rate": 6.509419295029097e-05,
+      "loss": 12.2745,
+      "step": 10723
+    },
+    {
+      "epoch": 1.8753661202185792,
+      "grad_norm": 0.8864958853480386,
+      "learning_rate": 6.507650026217031e-05,
+      "loss": 12.1887,
+      "step": 10724
+    },
+    {
+      "epoch": 1.8755409836065575,
+      "grad_norm": 0.8408937865333684,
+      "learning_rate": 6.505880881903163e-05,
+      "loss": 12.2147,
+      "step": 10725
+    },
+    {
+      "epoch": 1.8757158469945354,
+      "grad_norm": 0.8314894240865415,
+      "learning_rate": 6.504111862150563e-05,
+      "loss": 12.2583,
+      "step": 10726
+    },
+    {
+      "epoch": 1.8758907103825138,
+      "grad_norm": 0.9067176671342666,
+      "learning_rate": 6.502342967022288e-05,
+      "loss": 12.3556,
+      "step": 10727
+    },
+    {
+      "epoch": 1.8760655737704917,
+      "grad_norm": 0.8047681842981359,
+      "learning_rate": 6.500574196581398e-05,
+      "loss": 12.1728,
+      "step": 10728
+    },
+    {
+      "epoch": 1.87624043715847,
+      "grad_norm": 0.8417533270254295,
+      "learning_rate": 6.498805550890952e-05,
+      "loss": 12.2123,
+      "step": 10729
+    },
+    {
+      "epoch": 1.8764153005464481,
+      "grad_norm": 0.8596854731310849,
+      "learning_rate": 6.497037030013995e-05,
+      "loss": 12.3654,
+      "step": 10730
+    },
+    {
+      "epoch": 1.8765901639344262,
+      "grad_norm": 0.8870546003999198,
+      "learning_rate": 6.495268634013575e-05,
+      "loss": 12.2451,
+      "step": 10731
+    },
+    {
+      "epoch": 1.8767650273224044,
+      "grad_norm": 0.8282609747290559,
+      "learning_rate": 6.493500362952733e-05,
+      "loss": 12.2015,
+      "step": 10732
+    },
+    {
+      "epoch": 1.8769398907103825,
+      "grad_norm": 0.8249233018583778,
+      "learning_rate": 6.49173221689451e-05,
+      "loss": 12.3794,
+      "step": 10733
+    },
+    {
+      "epoch": 1.8771147540983608,
+      "grad_norm": 0.8629819480883509,
+      "learning_rate": 6.489964195901933e-05,
+      "loss": 12.2819,
+      "step": 10734
+    },
+    {
+      "epoch": 1.8772896174863387,
+      "grad_norm": 1.0153561221526348,
+      "learning_rate": 6.488196300038028e-05,
+      "loss": 12.2986,
+      "step": 10735
+    },
+    {
+      "epoch": 1.877464480874317,
+      "grad_norm": 0.8580566977942091,
+      "learning_rate": 6.486428529365824e-05,
+      "loss": 12.2774,
+      "step": 10736
+    },
+    {
+      "epoch": 1.877639344262295,
+      "grad_norm": 0.8343663516589719,
+      "learning_rate": 6.48466088394834e-05,
+      "loss": 12.2264,
+      "step": 10737
+    },
+    {
+      "epoch": 1.8778142076502733,
+      "grad_norm": 0.8595771959097412,
+      "learning_rate": 6.482893363848583e-05,
+      "loss": 12.3956,
+      "step": 10738
+    },
+    {
+      "epoch": 1.8779890710382514,
+      "grad_norm": 0.8818596034744856,
+      "learning_rate": 6.481125969129572e-05,
+      "loss": 12.378,
+      "step": 10739
+    },
+    {
+      "epoch": 1.8781639344262295,
+      "grad_norm": 0.860350912515887,
+      "learning_rate": 6.479358699854306e-05,
+      "loss": 12.3888,
+      "step": 10740
+    },
+    {
+      "epoch": 1.8783387978142077,
+      "grad_norm": 0.8313535464530174,
+      "learning_rate": 6.477591556085795e-05,
+      "loss": 12.4655,
+      "step": 10741
+    },
+    {
+      "epoch": 1.8785136612021858,
+      "grad_norm": 0.7939775427245966,
+      "learning_rate": 6.475824537887025e-05,
+      "loss": 12.0622,
+      "step": 10742
+    },
+    {
+      "epoch": 1.8786885245901639,
+      "grad_norm": 0.8845139737817305,
+      "learning_rate": 6.474057645320992e-05,
+      "loss": 12.3288,
+      "step": 10743
+    },
+    {
+      "epoch": 1.878863387978142,
+      "grad_norm": 0.8203973212869795,
+      "learning_rate": 6.472290878450688e-05,
+      "loss": 12.2777,
+      "step": 10744
+    },
+    {
+      "epoch": 1.8790382513661203,
+      "grad_norm": 0.7589754690041041,
+      "learning_rate": 6.47052423733909e-05,
+      "loss": 12.2819,
+      "step": 10745
+    },
+    {
+      "epoch": 1.8792131147540982,
+      "grad_norm": 0.9005495332904406,
+      "learning_rate": 6.46875772204918e-05,
+      "loss": 12.1828,
+      "step": 10746
+    },
+    {
+      "epoch": 1.8793879781420766,
+      "grad_norm": 0.8311859370130545,
+      "learning_rate": 6.466991332643929e-05,
+      "loss": 12.1844,
+      "step": 10747
+    },
+    {
+      "epoch": 1.8795628415300545,
+      "grad_norm": 0.8611246599576098,
+      "learning_rate": 6.465225069186312e-05,
+      "loss": 12.4874,
+      "step": 10748
+    },
+    {
+      "epoch": 1.8797377049180328,
+      "grad_norm": 0.8336554476400025,
+      "learning_rate": 6.46345893173929e-05,
+      "loss": 12.3778,
+      "step": 10749
+    },
+    {
+      "epoch": 1.879912568306011,
+      "grad_norm": 0.847340143094919,
+      "learning_rate": 6.461692920365824e-05,
+      "loss": 12.0351,
+      "step": 10750
+    },
+    {
+      "epoch": 1.880087431693989,
+      "grad_norm": 0.8022464939865721,
+      "learning_rate": 6.459927035128872e-05,
+      "loss": 12.2813,
+      "step": 10751
+    },
+    {
+      "epoch": 1.8802622950819672,
+      "grad_norm": 0.8179878584133047,
+      "learning_rate": 6.458161276091388e-05,
+      "loss": 12.4265,
+      "step": 10752
+    },
+    {
+      "epoch": 1.8804371584699453,
+      "grad_norm": 0.893612935569724,
+      "learning_rate": 6.456395643316314e-05,
+      "loss": 12.2035,
+      "step": 10753
+    },
+    {
+      "epoch": 1.8806120218579236,
+      "grad_norm": 0.8011606751012206,
+      "learning_rate": 6.454630136866595e-05,
+      "loss": 12.0978,
+      "step": 10754
+    },
+    {
+      "epoch": 1.8807868852459015,
+      "grad_norm": 0.9074685757703779,
+      "learning_rate": 6.452864756805167e-05,
+      "loss": 12.3482,
+      "step": 10755
+    },
+    {
+      "epoch": 1.8809617486338799,
+      "grad_norm": 0.8683754378334868,
+      "learning_rate": 6.45109950319497e-05,
+      "loss": 12.4086,
+      "step": 10756
+    },
+    {
+      "epoch": 1.8811366120218578,
+      "grad_norm": 0.7803434915586183,
+      "learning_rate": 6.449334376098927e-05,
+      "loss": 12.1018,
+      "step": 10757
+    },
+    {
+      "epoch": 1.8813114754098361,
+      "grad_norm": 0.8451482946643494,
+      "learning_rate": 6.447569375579964e-05,
+      "loss": 12.3125,
+      "step": 10758
+    },
+    {
+      "epoch": 1.8814863387978142,
+      "grad_norm": 0.8377504529107129,
+      "learning_rate": 6.445804501701007e-05,
+      "loss": 12.2256,
+      "step": 10759
+    },
+    {
+      "epoch": 1.8816612021857924,
+      "grad_norm": 0.832169482560039,
+      "learning_rate": 6.444039754524965e-05,
+      "loss": 12.3523,
+      "step": 10760
+    },
+    {
+      "epoch": 1.8818360655737705,
+      "grad_norm": 0.8741894358214393,
+      "learning_rate": 6.442275134114747e-05,
+      "loss": 12.2596,
+      "step": 10761
+    },
+    {
+      "epoch": 1.8820109289617486,
+      "grad_norm": 0.8660474007674159,
+      "learning_rate": 6.440510640533266e-05,
+      "loss": 12.1541,
+      "step": 10762
+    },
+    {
+      "epoch": 1.882185792349727,
+      "grad_norm": 0.8089877456050284,
+      "learning_rate": 6.438746273843426e-05,
+      "loss": 12.2323,
+      "step": 10763
+    },
+    {
+      "epoch": 1.8823606557377048,
+      "grad_norm": 1.1128784057800825,
+      "learning_rate": 6.436982034108115e-05,
+      "loss": 12.3007,
+      "step": 10764
+    },
+    {
+      "epoch": 1.8825355191256832,
+      "grad_norm": 0.8692221828693789,
+      "learning_rate": 6.435217921390233e-05,
+      "loss": 12.0779,
+      "step": 10765
+    },
+    {
+      "epoch": 1.882710382513661,
+      "grad_norm": 1.0080944433862784,
+      "learning_rate": 6.433453935752667e-05,
+      "loss": 12.3694,
+      "step": 10766
+    },
+    {
+      "epoch": 1.8828852459016394,
+      "grad_norm": 0.8967909795961685,
+      "learning_rate": 6.431690077258303e-05,
+      "loss": 11.9751,
+      "step": 10767
+    },
+    {
+      "epoch": 1.8830601092896175,
+      "grad_norm": 0.8528167375802318,
+      "learning_rate": 6.429926345970017e-05,
+      "loss": 12.3314,
+      "step": 10768
+    },
+    {
+      "epoch": 1.8832349726775957,
+      "grad_norm": 0.8779058416069523,
+      "learning_rate": 6.428162741950684e-05,
+      "loss": 12.4055,
+      "step": 10769
+    },
+    {
+      "epoch": 1.8834098360655738,
+      "grad_norm": 1.0779316352054398,
+      "learning_rate": 6.426399265263177e-05,
+      "loss": 12.5463,
+      "step": 10770
+    },
+    {
+      "epoch": 1.883584699453552,
+      "grad_norm": 0.8796563478472336,
+      "learning_rate": 6.424635915970365e-05,
+      "loss": 12.1802,
+      "step": 10771
+    },
+    {
+      "epoch": 1.88375956284153,
+      "grad_norm": 1.1187416574129032,
+      "learning_rate": 6.422872694135101e-05,
+      "loss": 12.3606,
+      "step": 10772
+    },
+    {
+      "epoch": 1.8839344262295081,
+      "grad_norm": 0.9651626960009807,
+      "learning_rate": 6.421109599820246e-05,
+      "loss": 12.172,
+      "step": 10773
+    },
+    {
+      "epoch": 1.8841092896174865,
+      "grad_norm": 0.8641798885227696,
+      "learning_rate": 6.419346633088654e-05,
+      "loss": 12.5764,
+      "step": 10774
+    },
+    {
+      "epoch": 1.8842841530054644,
+      "grad_norm": 1.0033339695316486,
+      "learning_rate": 6.417583794003171e-05,
+      "loss": 12.3115,
+      "step": 10775
+    },
+    {
+      "epoch": 1.8844590163934427,
+      "grad_norm": 0.8859757047466569,
+      "learning_rate": 6.415821082626636e-05,
+      "loss": 12.2673,
+      "step": 10776
+    },
+    {
+      "epoch": 1.8846338797814206,
+      "grad_norm": 0.8510466473944368,
+      "learning_rate": 6.414058499021895e-05,
+      "loss": 12.431,
+      "step": 10777
+    },
+    {
+      "epoch": 1.884808743169399,
+      "grad_norm": 0.9957360993408949,
+      "learning_rate": 6.412296043251782e-05,
+      "loss": 12.0339,
+      "step": 10778
+    },
+    {
+      "epoch": 1.884983606557377,
+      "grad_norm": 1.102489996180215,
+      "learning_rate": 6.410533715379118e-05,
+      "loss": 12.0845,
+      "step": 10779
+    },
+    {
+      "epoch": 1.8851584699453552,
+      "grad_norm": 0.9977632768309702,
+      "learning_rate": 6.408771515466736e-05,
+      "loss": 12.5025,
+      "step": 10780
+    },
+    {
+      "epoch": 1.8853333333333333,
+      "grad_norm": 1.0925298188805157,
+      "learning_rate": 6.407009443577452e-05,
+      "loss": 12.4352,
+      "step": 10781
+    },
+    {
+      "epoch": 1.8855081967213114,
+      "grad_norm": 0.9558647768224844,
+      "learning_rate": 6.405247499774085e-05,
+      "loss": 12.3292,
+      "step": 10782
+    },
+    {
+      "epoch": 1.8856830601092898,
+      "grad_norm": 0.8818445216948603,
+      "learning_rate": 6.403485684119444e-05,
+      "loss": 12.4181,
+      "step": 10783
+    },
+    {
+      "epoch": 1.8858579234972677,
+      "grad_norm": 0.8750274549906047,
+      "learning_rate": 6.401723996676335e-05,
+      "loss": 12.3788,
+      "step": 10784
+    },
+    {
+      "epoch": 1.886032786885246,
+      "grad_norm": 0.8749750992126025,
+      "learning_rate": 6.399962437507564e-05,
+      "loss": 12.3288,
+      "step": 10785
+    },
+    {
+      "epoch": 1.886207650273224,
+      "grad_norm": 0.8252662309592312,
+      "learning_rate": 6.398201006675926e-05,
+      "loss": 12.323,
+      "step": 10786
+    },
+    {
+      "epoch": 1.8863825136612022,
+      "grad_norm": 0.8647822330487758,
+      "learning_rate": 6.396439704244211e-05,
+      "loss": 12.2173,
+      "step": 10787
+    },
+    {
+      "epoch": 1.8865573770491804,
+      "grad_norm": 0.8772604316759826,
+      "learning_rate": 6.394678530275211e-05,
+      "loss": 12.3755,
+      "step": 10788
+    },
+    {
+      "epoch": 1.8867322404371585,
+      "grad_norm": 0.8530246658238239,
+      "learning_rate": 6.392917484831713e-05,
+      "loss": 12.037,
+      "step": 10789
+    },
+    {
+      "epoch": 1.8869071038251366,
+      "grad_norm": 0.9591323645654981,
+      "learning_rate": 6.391156567976489e-05,
+      "loss": 12.33,
+      "step": 10790
+    },
+    {
+      "epoch": 1.8870819672131147,
+      "grad_norm": 0.9609927096851214,
+      "learning_rate": 6.389395779772315e-05,
+      "loss": 12.4546,
+      "step": 10791
+    },
+    {
+      "epoch": 1.887256830601093,
+      "grad_norm": 0.8233453089778682,
+      "learning_rate": 6.387635120281963e-05,
+      "loss": 12.0599,
+      "step": 10792
+    },
+    {
+      "epoch": 1.887431693989071,
+      "grad_norm": 0.9997583587579623,
+      "learning_rate": 6.385874589568201e-05,
+      "loss": 11.8665,
+      "step": 10793
+    },
+    {
+      "epoch": 1.8876065573770493,
+      "grad_norm": 0.9649152485379596,
+      "learning_rate": 6.384114187693787e-05,
+      "loss": 12.3294,
+      "step": 10794
+    },
+    {
+      "epoch": 1.8877814207650272,
+      "grad_norm": 0.986027394128267,
+      "learning_rate": 6.382353914721475e-05,
+      "loss": 12.3667,
+      "step": 10795
+    },
+    {
+      "epoch": 1.8879562841530055,
+      "grad_norm": 0.9247348867367596,
+      "learning_rate": 6.380593770714021e-05,
+      "loss": 12.4207,
+      "step": 10796
+    },
+    {
+      "epoch": 1.8881311475409837,
+      "grad_norm": 0.7911963784051624,
+      "learning_rate": 6.378833755734173e-05,
+      "loss": 12.2706,
+      "step": 10797
+    },
+    {
+      "epoch": 1.8883060109289618,
+      "grad_norm": 1.2494871296173713,
+      "learning_rate": 6.377073869844666e-05,
+      "loss": 12.3003,
+      "step": 10798
+    },
+    {
+      "epoch": 1.88848087431694,
+      "grad_norm": 0.888208355347475,
+      "learning_rate": 6.375314113108244e-05,
+      "loss": 12.294,
+      "step": 10799
+    },
+    {
+      "epoch": 1.888655737704918,
+      "grad_norm": 0.8485795683478718,
+      "learning_rate": 6.373554485587638e-05,
+      "loss": 12.3128,
+      "step": 10800
+    },
+    {
+      "epoch": 1.8888306010928961,
+      "grad_norm": 0.9025713954882507,
+      "learning_rate": 6.371794987345581e-05,
+      "loss": 12.3386,
+      "step": 10801
+    },
+    {
+      "epoch": 1.8890054644808743,
+      "grad_norm": 0.7737308913522108,
+      "learning_rate": 6.370035618444789e-05,
+      "loss": 12.1043,
+      "step": 10802
+    },
+    {
+      "epoch": 1.8891803278688526,
+      "grad_norm": 1.0400354048092375,
+      "learning_rate": 6.36827637894799e-05,
+      "loss": 12.4534,
+      "step": 10803
+    },
+    {
+      "epoch": 1.8893551912568305,
+      "grad_norm": 0.9226031660563947,
+      "learning_rate": 6.366517268917895e-05,
+      "loss": 12.4372,
+      "step": 10804
+    },
+    {
+      "epoch": 1.8895300546448088,
+      "grad_norm": 0.844450878703697,
+      "learning_rate": 6.364758288417212e-05,
+      "loss": 12.2702,
+      "step": 10805
+    },
+    {
+      "epoch": 1.8897049180327867,
+      "grad_norm": 0.8419982964290488,
+      "learning_rate": 6.362999437508649e-05,
+      "loss": 12.2309,
+      "step": 10806
+    },
+    {
+      "epoch": 1.889879781420765,
+      "grad_norm": 1.001385875659587,
+      "learning_rate": 6.361240716254907e-05,
+      "loss": 12.4373,
+      "step": 10807
+    },
+    {
+      "epoch": 1.8900546448087432,
+      "grad_norm": 0.9446154094434277,
+      "learning_rate": 6.359482124718685e-05,
+      "loss": 12.4445,
+      "step": 10808
+    },
+    {
+      "epoch": 1.8902295081967213,
+      "grad_norm": 0.9667497013075737,
+      "learning_rate": 6.357723662962666e-05,
+      "loss": 12.3486,
+      "step": 10809
+    },
+    {
+      "epoch": 1.8904043715846994,
+      "grad_norm": 0.8941213350372106,
+      "learning_rate": 6.355965331049546e-05,
+      "loss": 12.2895,
+      "step": 10810
+    },
+    {
+      "epoch": 1.8905792349726775,
+      "grad_norm": 1.0200379640125419,
+      "learning_rate": 6.354207129042003e-05,
+      "loss": 12.3555,
+      "step": 10811
+    },
+    {
+      "epoch": 1.8907540983606559,
+      "grad_norm": 0.842247582720843,
+      "learning_rate": 6.352449057002718e-05,
+      "loss": 12.4101,
+      "step": 10812
+    },
+    {
+      "epoch": 1.8909289617486338,
+      "grad_norm": 0.9272470008465794,
+      "learning_rate": 6.35069111499436e-05,
+      "loss": 12.3797,
+      "step": 10813
+    },
+    {
+      "epoch": 1.8911038251366121,
+      "grad_norm": 1.0590182521277454,
+      "learning_rate": 6.348933303079602e-05,
+      "loss": 12.2698,
+      "step": 10814
+    },
+    {
+      "epoch": 1.89127868852459,
+      "grad_norm": 0.8856150234660799,
+      "learning_rate": 6.347175621321104e-05,
+      "loss": 12.3147,
+      "step": 10815
+    },
+    {
+      "epoch": 1.8914535519125684,
+      "grad_norm": 0.7966478322694542,
+      "learning_rate": 6.345418069781532e-05,
+      "loss": 12.2737,
+      "step": 10816
+    },
+    {
+      "epoch": 1.8916284153005465,
+      "grad_norm": 1.086898186921641,
+      "learning_rate": 6.343660648523532e-05,
+      "loss": 12.4003,
+      "step": 10817
+    },
+    {
+      "epoch": 1.8918032786885246,
+      "grad_norm": 0.9043623055326161,
+      "learning_rate": 6.341903357609757e-05,
+      "loss": 12.2768,
+      "step": 10818
+    },
+    {
+      "epoch": 1.8919781420765027,
+      "grad_norm": 0.8472385554711792,
+      "learning_rate": 6.340146197102854e-05,
+      "loss": 12.4694,
+      "step": 10819
+    },
+    {
+      "epoch": 1.8921530054644808,
+      "grad_norm": 1.0221313810323147,
+      "learning_rate": 6.338389167065466e-05,
+      "loss": 12.1762,
+      "step": 10820
+    },
+    {
+      "epoch": 1.8923278688524592,
+      "grad_norm": 0.9262948074473999,
+      "learning_rate": 6.336632267560222e-05,
+      "loss": 12.2249,
+      "step": 10821
+    },
+    {
+      "epoch": 1.892502732240437,
+      "grad_norm": 0.9143637580206688,
+      "learning_rate": 6.33487549864976e-05,
+      "loss": 12.3046,
+      "step": 10822
+    },
+    {
+      "epoch": 1.8926775956284154,
+      "grad_norm": 0.8993890809826561,
+      "learning_rate": 6.333118860396705e-05,
+      "loss": 12.1954,
+      "step": 10823
+    },
+    {
+      "epoch": 1.8928524590163933,
+      "grad_norm": 0.8755449465289793,
+      "learning_rate": 6.331362352863677e-05,
+      "loss": 12.1548,
+      "step": 10824
+    },
+    {
+      "epoch": 1.8930273224043717,
+      "grad_norm": 0.9282991174177369,
+      "learning_rate": 6.329605976113293e-05,
+      "loss": 12.321,
+      "step": 10825
+    },
+    {
+      "epoch": 1.8932021857923498,
+      "grad_norm": 0.8432564680472435,
+      "learning_rate": 6.327849730208169e-05,
+      "loss": 12.294,
+      "step": 10826
+    },
+    {
+      "epoch": 1.893377049180328,
+      "grad_norm": 0.8892392039342358,
+      "learning_rate": 6.326093615210915e-05,
+      "loss": 12.0531,
+      "step": 10827
+    },
+    {
+      "epoch": 1.893551912568306,
+      "grad_norm": 0.8311054147907054,
+      "learning_rate": 6.324337631184127e-05,
+      "loss": 12.2239,
+      "step": 10828
+    },
+    {
+      "epoch": 1.8937267759562841,
+      "grad_norm": 0.9690434478327566,
+      "learning_rate": 6.322581778190409e-05,
+      "loss": 12.2331,
+      "step": 10829
+    },
+    {
+      "epoch": 1.8939016393442623,
+      "grad_norm": 0.8407652329539045,
+      "learning_rate": 6.320826056292353e-05,
+      "loss": 12.2472,
+      "step": 10830
+    },
+    {
+      "epoch": 1.8940765027322404,
+      "grad_norm": 0.8752678407576392,
+      "learning_rate": 6.31907046555255e-05,
+      "loss": 12.4986,
+      "step": 10831
+    },
+    {
+      "epoch": 1.8942513661202187,
+      "grad_norm": 0.8609359583210369,
+      "learning_rate": 6.317315006033583e-05,
+      "loss": 12.3189,
+      "step": 10832
+    },
+    {
+      "epoch": 1.8944262295081966,
+      "grad_norm": 0.8043361281269135,
+      "learning_rate": 6.315559677798035e-05,
+      "loss": 12.2049,
+      "step": 10833
+    },
+    {
+      "epoch": 1.894601092896175,
+      "grad_norm": 0.9561296655596749,
+      "learning_rate": 6.313804480908478e-05,
+      "loss": 12.558,
+      "step": 10834
+    },
+    {
+      "epoch": 1.8947759562841529,
+      "grad_norm": 0.8391536948421467,
+      "learning_rate": 6.31204941542749e-05,
+      "loss": 12.4123,
+      "step": 10835
+    },
+    {
+      "epoch": 1.8949508196721312,
+      "grad_norm": 0.8259367453642504,
+      "learning_rate": 6.310294481417627e-05,
+      "loss": 12.4106,
+      "step": 10836
+    },
+    {
+      "epoch": 1.8951256830601093,
+      "grad_norm": 0.9226272871221105,
+      "learning_rate": 6.308539678941454e-05,
+      "loss": 12.0432,
+      "step": 10837
+    },
+    {
+      "epoch": 1.8953005464480874,
+      "grad_norm": 0.8371732073413168,
+      "learning_rate": 6.306785008061531e-05,
+      "loss": 12.1175,
+      "step": 10838
+    },
+    {
+      "epoch": 1.8954754098360656,
+      "grad_norm": 0.8940558654399988,
+      "learning_rate": 6.305030468840405e-05,
+      "loss": 12.3789,
+      "step": 10839
+    },
+    {
+      "epoch": 1.8956502732240437,
+      "grad_norm": 0.8591914536489308,
+      "learning_rate": 6.303276061340625e-05,
+      "loss": 12.4643,
+      "step": 10840
+    },
+    {
+      "epoch": 1.895825136612022,
+      "grad_norm": 0.8375538794031384,
+      "learning_rate": 6.301521785624736e-05,
+      "loss": 12.242,
+      "step": 10841
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.8452184149841255,
+      "learning_rate": 6.299767641755277e-05,
+      "loss": 12.2139,
+      "step": 10842
+    },
+    {
+      "epoch": 1.8961748633879782,
+      "grad_norm": 0.8382523900103953,
+      "learning_rate": 6.298013629794775e-05,
+      "loss": 12.0797,
+      "step": 10843
+    },
+    {
+      "epoch": 1.8963497267759561,
+      "grad_norm": 0.9635494355555214,
+      "learning_rate": 6.29625974980576e-05,
+      "loss": 12.5081,
+      "step": 10844
+    },
+    {
+      "epoch": 1.8965245901639345,
+      "grad_norm": 0.877594053992503,
+      "learning_rate": 6.29450600185076e-05,
+      "loss": 12.2966,
+      "step": 10845
+    },
+    {
+      "epoch": 1.8966994535519126,
+      "grad_norm": 0.8446228543644849,
+      "learning_rate": 6.29275238599229e-05,
+      "loss": 12.2712,
+      "step": 10846
+    },
+    {
+      "epoch": 1.8968743169398907,
+      "grad_norm": 0.82257697602346,
+      "learning_rate": 6.290998902292869e-05,
+      "loss": 12.3798,
+      "step": 10847
+    },
+    {
+      "epoch": 1.8970491803278688,
+      "grad_norm": 0.9900538390207403,
+      "learning_rate": 6.289245550814999e-05,
+      "loss": 12.3222,
+      "step": 10848
+    },
+    {
+      "epoch": 1.897224043715847,
+      "grad_norm": 0.8590629259052771,
+      "learning_rate": 6.287492331621192e-05,
+      "loss": 12.3993,
+      "step": 10849
+    },
+    {
+      "epoch": 1.897398907103825,
+      "grad_norm": 0.8155934832909203,
+      "learning_rate": 6.285739244773949e-05,
+      "loss": 12.0675,
+      "step": 10850
+    },
+    {
+      "epoch": 1.8975737704918032,
+      "grad_norm": 0.8482675128692526,
+      "learning_rate": 6.28398629033576e-05,
+      "loss": 12.273,
+      "step": 10851
+    },
+    {
+      "epoch": 1.8977486338797815,
+      "grad_norm": 0.9270372509655945,
+      "learning_rate": 6.282233468369116e-05,
+      "loss": 12.3173,
+      "step": 10852
+    },
+    {
+      "epoch": 1.8979234972677594,
+      "grad_norm": 0.9221404059219835,
+      "learning_rate": 6.28048077893651e-05,
+      "loss": 12.316,
+      "step": 10853
+    },
+    {
+      "epoch": 1.8980983606557378,
+      "grad_norm": 0.8807183948240684,
+      "learning_rate": 6.278728222100416e-05,
+      "loss": 12.5038,
+      "step": 10854
+    },
+    {
+      "epoch": 1.898273224043716,
+      "grad_norm": 0.8277623118475081,
+      "learning_rate": 6.27697579792331e-05,
+      "loss": 12.2416,
+      "step": 10855
+    },
+    {
+      "epoch": 1.898448087431694,
+      "grad_norm": 0.8382884497269205,
+      "learning_rate": 6.275223506467669e-05,
+      "loss": 12.3148,
+      "step": 10856
+    },
+    {
+      "epoch": 1.8986229508196721,
+      "grad_norm": 0.8286788253311251,
+      "learning_rate": 6.273471347795959e-05,
+      "loss": 12.302,
+      "step": 10857
+    },
+    {
+      "epoch": 1.8987978142076503,
+      "grad_norm": 0.8584240343945087,
+      "learning_rate": 6.271719321970639e-05,
+      "loss": 12.1591,
+      "step": 10858
+    },
+    {
+      "epoch": 1.8989726775956284,
+      "grad_norm": 0.8377023566024651,
+      "learning_rate": 6.269967429054171e-05,
+      "loss": 12.3803,
+      "step": 10859
+    },
+    {
+      "epoch": 1.8991475409836065,
+      "grad_norm": 0.8386557039063995,
+      "learning_rate": 6.268215669109004e-05,
+      "loss": 12.4465,
+      "step": 10860
+    },
+    {
+      "epoch": 1.8993224043715848,
+      "grad_norm": 0.803344034441615,
+      "learning_rate": 6.266464042197594e-05,
+      "loss": 12.1927,
+      "step": 10861
+    },
+    {
+      "epoch": 1.8994972677595627,
+      "grad_norm": 0.8850568499368178,
+      "learning_rate": 6.264712548382374e-05,
+      "loss": 12.3666,
+      "step": 10862
+    },
+    {
+      "epoch": 1.899672131147541,
+      "grad_norm": 0.7850263825339603,
+      "learning_rate": 6.262961187725787e-05,
+      "loss": 12.2801,
+      "step": 10863
+    },
+    {
+      "epoch": 1.899846994535519,
+      "grad_norm": 0.9144039986605087,
+      "learning_rate": 6.261209960290268e-05,
+      "loss": 12.4227,
+      "step": 10864
+    },
+    {
+      "epoch": 1.9000218579234973,
+      "grad_norm": 0.8217303488326204,
+      "learning_rate": 6.259458866138247e-05,
+      "loss": 12.2558,
+      "step": 10865
+    },
+    {
+      "epoch": 1.9001967213114754,
+      "grad_norm": 0.7937898212508487,
+      "learning_rate": 6.257707905332146e-05,
+      "loss": 12.0943,
+      "step": 10866
+    },
+    {
+      "epoch": 1.9003715846994536,
+      "grad_norm": 0.7859746173607471,
+      "learning_rate": 6.255957077934386e-05,
+      "loss": 12.3236,
+      "step": 10867
+    },
+    {
+      "epoch": 1.9005464480874317,
+      "grad_norm": 0.7655110302857147,
+      "learning_rate": 6.254206384007384e-05,
+      "loss": 12.3438,
+      "step": 10868
+    },
+    {
+      "epoch": 1.9007213114754098,
+      "grad_norm": 0.880155898959594,
+      "learning_rate": 6.252455823613546e-05,
+      "loss": 12.2247,
+      "step": 10869
+    },
+    {
+      "epoch": 1.9008961748633881,
+      "grad_norm": 0.8108016052465822,
+      "learning_rate": 6.250705396815277e-05,
+      "loss": 12.2038,
+      "step": 10870
+    },
+    {
+      "epoch": 1.901071038251366,
+      "grad_norm": 0.9754752750419199,
+      "learning_rate": 6.248955103674983e-05,
+      "loss": 12.3878,
+      "step": 10871
+    },
+    {
+      "epoch": 1.9012459016393444,
+      "grad_norm": 0.817259378014999,
+      "learning_rate": 6.247204944255058e-05,
+      "loss": 12.3572,
+      "step": 10872
+    },
+    {
+      "epoch": 1.9014207650273223,
+      "grad_norm": 0.8329024920357093,
+      "learning_rate": 6.24545491861789e-05,
+      "loss": 12.2072,
+      "step": 10873
+    },
+    {
+      "epoch": 1.9015956284153006,
+      "grad_norm": 1.0244737187442778,
+      "learning_rate": 6.243705026825867e-05,
+      "loss": 12.1654,
+      "step": 10874
+    },
+    {
+      "epoch": 1.9017704918032787,
+      "grad_norm": 0.8293916023986773,
+      "learning_rate": 6.241955268941371e-05,
+      "loss": 12.1492,
+      "step": 10875
+    },
+    {
+      "epoch": 1.9019453551912568,
+      "grad_norm": 0.9580609895435346,
+      "learning_rate": 6.24020564502678e-05,
+      "loss": 12.2968,
+      "step": 10876
+    },
+    {
+      "epoch": 1.902120218579235,
+      "grad_norm": 0.8916745927549105,
+      "learning_rate": 6.238456155144463e-05,
+      "loss": 12.4364,
+      "step": 10877
+    },
+    {
+      "epoch": 1.902295081967213,
+      "grad_norm": 0.9248140493007405,
+      "learning_rate": 6.236706799356791e-05,
+      "loss": 12.21,
+      "step": 10878
+    },
+    {
+      "epoch": 1.9024699453551912,
+      "grad_norm": 0.862409248773275,
+      "learning_rate": 6.234957577726123e-05,
+      "loss": 12.1929,
+      "step": 10879
+    },
+    {
+      "epoch": 1.9026448087431693,
+      "grad_norm": 0.8753153945725162,
+      "learning_rate": 6.233208490314822e-05,
+      "loss": 12.2519,
+      "step": 10880
+    },
+    {
+      "epoch": 1.9028196721311477,
+      "grad_norm": 0.8624522777444089,
+      "learning_rate": 6.231459537185235e-05,
+      "loss": 12.374,
+      "step": 10881
+    },
+    {
+      "epoch": 1.9029945355191256,
+      "grad_norm": 0.9285225428824988,
+      "learning_rate": 6.22971071839971e-05,
+      "loss": 12.242,
+      "step": 10882
+    },
+    {
+      "epoch": 1.903169398907104,
+      "grad_norm": 0.8659749894695746,
+      "learning_rate": 6.227962034020598e-05,
+      "loss": 12.199,
+      "step": 10883
+    },
+    {
+      "epoch": 1.9033442622950818,
+      "grad_norm": 1.0050832621917414,
+      "learning_rate": 6.226213484110228e-05,
+      "loss": 11.9043,
+      "step": 10884
+    },
+    {
+      "epoch": 1.9035191256830601,
+      "grad_norm": 0.8908439073040441,
+      "learning_rate": 6.224465068730938e-05,
+      "loss": 12.2218,
+      "step": 10885
+    },
+    {
+      "epoch": 1.9036939890710383,
+      "grad_norm": 0.8793672458078501,
+      "learning_rate": 6.222716787945059e-05,
+      "loss": 12.142,
+      "step": 10886
+    },
+    {
+      "epoch": 1.9038688524590164,
+      "grad_norm": 0.8576081416558379,
+      "learning_rate": 6.220968641814915e-05,
+      "loss": 12.0774,
+      "step": 10887
+    },
+    {
+      "epoch": 1.9040437158469945,
+      "grad_norm": 0.833860414624755,
+      "learning_rate": 6.219220630402821e-05,
+      "loss": 12.29,
+      "step": 10888
+    },
+    {
+      "epoch": 1.9042185792349726,
+      "grad_norm": 0.8447710405983789,
+      "learning_rate": 6.217472753771095e-05,
+      "loss": 12.4181,
+      "step": 10889
+    },
+    {
+      "epoch": 1.904393442622951,
+      "grad_norm": 0.8583776080494234,
+      "learning_rate": 6.215725011982043e-05,
+      "loss": 12.2218,
+      "step": 10890
+    },
+    {
+      "epoch": 1.9045683060109289,
+      "grad_norm": 0.8846686488137185,
+      "learning_rate": 6.213977405097978e-05,
+      "loss": 12.3479,
+      "step": 10891
+    },
+    {
+      "epoch": 1.9047431693989072,
+      "grad_norm": 0.8700152528227358,
+      "learning_rate": 6.21222993318119e-05,
+      "loss": 11.8978,
+      "step": 10892
+    },
+    {
+      "epoch": 1.904918032786885,
+      "grad_norm": 0.7987041045081542,
+      "learning_rate": 6.210482596293982e-05,
+      "loss": 12.3155,
+      "step": 10893
+    },
+    {
+      "epoch": 1.9050928961748634,
+      "grad_norm": 0.8289159111939113,
+      "learning_rate": 6.20873539449864e-05,
+      "loss": 12.4768,
+      "step": 10894
+    },
+    {
+      "epoch": 1.9052677595628416,
+      "grad_norm": 0.8532093807509568,
+      "learning_rate": 6.206988327857457e-05,
+      "loss": 12.316,
+      "step": 10895
+    },
+    {
+      "epoch": 1.9054426229508197,
+      "grad_norm": 0.8381600967921212,
+      "learning_rate": 6.205241396432703e-05,
+      "loss": 12.2005,
+      "step": 10896
+    },
+    {
+      "epoch": 1.9056174863387978,
+      "grad_norm": 0.8383858213093603,
+      "learning_rate": 6.203494600286661e-05,
+      "loss": 12.1866,
+      "step": 10897
+    },
+    {
+      "epoch": 1.905792349726776,
+      "grad_norm": 0.8572940469314656,
+      "learning_rate": 6.201747939481604e-05,
+      "loss": 12.3016,
+      "step": 10898
+    },
+    {
+      "epoch": 1.9059672131147543,
+      "grad_norm": 0.8372507940018635,
+      "learning_rate": 6.200001414079791e-05,
+      "loss": 12.3147,
+      "step": 10899
+    },
+    {
+      "epoch": 1.9061420765027322,
+      "grad_norm": 0.9614147913378795,
+      "learning_rate": 6.198255024143488e-05,
+      "loss": 12.4806,
+      "step": 10900
+    },
+    {
+      "epoch": 1.9063169398907105,
+      "grad_norm": 0.8471709675290133,
+      "learning_rate": 6.196508769734953e-05,
+      "loss": 12.0425,
+      "step": 10901
+    },
+    {
+      "epoch": 1.9064918032786884,
+      "grad_norm": 0.8881006543529384,
+      "learning_rate": 6.194762650916435e-05,
+      "loss": 12.3445,
+      "step": 10902
+    },
+    {
+      "epoch": 1.9066666666666667,
+      "grad_norm": 0.8981040035362979,
+      "learning_rate": 6.193016667750184e-05,
+      "loss": 12.289,
+      "step": 10903
+    },
+    {
+      "epoch": 1.9068415300546449,
+      "grad_norm": 1.0098211950477538,
+      "learning_rate": 6.191270820298439e-05,
+      "loss": 12.2599,
+      "step": 10904
+    },
+    {
+      "epoch": 1.907016393442623,
+      "grad_norm": 0.8888062638785366,
+      "learning_rate": 6.18952510862344e-05,
+      "loss": 12.2911,
+      "step": 10905
+    },
+    {
+      "epoch": 1.907191256830601,
+      "grad_norm": 0.8918762227298584,
+      "learning_rate": 6.187779532787423e-05,
+      "loss": 12.2518,
+      "step": 10906
+    },
+    {
+      "epoch": 1.9073661202185792,
+      "grad_norm": 0.8811811369595453,
+      "learning_rate": 6.186034092852606e-05,
+      "loss": 12.1926,
+      "step": 10907
+    },
+    {
+      "epoch": 1.9075409836065573,
+      "grad_norm": 0.8430924010207436,
+      "learning_rate": 6.184288788881218e-05,
+      "loss": 12.0714,
+      "step": 10908
+    },
+    {
+      "epoch": 1.9077158469945354,
+      "grad_norm": 0.816580934645475,
+      "learning_rate": 6.182543620935477e-05,
+      "loss": 11.9793,
+      "step": 10909
+    },
+    {
+      "epoch": 1.9078907103825138,
+      "grad_norm": 0.8582232145567085,
+      "learning_rate": 6.180798589077597e-05,
+      "loss": 12.4353,
+      "step": 10910
+    },
+    {
+      "epoch": 1.9080655737704917,
+      "grad_norm": 0.851569856235332,
+      "learning_rate": 6.179053693369785e-05,
+      "loss": 12.3153,
+      "step": 10911
+    },
+    {
+      "epoch": 1.90824043715847,
+      "grad_norm": 0.8145012771930719,
+      "learning_rate": 6.177308933874242e-05,
+      "loss": 12.3717,
+      "step": 10912
+    },
+    {
+      "epoch": 1.908415300546448,
+      "grad_norm": 0.8293106265017478,
+      "learning_rate": 6.175564310653174e-05,
+      "loss": 12.4713,
+      "step": 10913
+    },
+    {
+      "epoch": 1.9085901639344263,
+      "grad_norm": 0.8483742532113551,
+      "learning_rate": 6.173819823768767e-05,
+      "loss": 12.3913,
+      "step": 10914
+    },
+    {
+      "epoch": 1.9087650273224044,
+      "grad_norm": 0.8486844447785163,
+      "learning_rate": 6.172075473283212e-05,
+      "loss": 12.1171,
+      "step": 10915
+    },
+    {
+      "epoch": 1.9089398907103825,
+      "grad_norm": 0.9192282181697299,
+      "learning_rate": 6.170331259258695e-05,
+      "loss": 12.2719,
+      "step": 10916
+    },
+    {
+      "epoch": 1.9091147540983606,
+      "grad_norm": 0.8925869969629322,
+      "learning_rate": 6.168587181757397e-05,
+      "loss": 12.5211,
+      "step": 10917
+    },
+    {
+      "epoch": 1.9092896174863387,
+      "grad_norm": 0.8201643467659439,
+      "learning_rate": 6.166843240841485e-05,
+      "loss": 12.2247,
+      "step": 10918
+    },
+    {
+      "epoch": 1.909464480874317,
+      "grad_norm": 0.8984974576914804,
+      "learning_rate": 6.165099436573134e-05,
+      "loss": 12.1748,
+      "step": 10919
+    },
+    {
+      "epoch": 1.909639344262295,
+      "grad_norm": 0.887893077206699,
+      "learning_rate": 6.163355769014507e-05,
+      "loss": 12.3524,
+      "step": 10920
+    },
+    {
+      "epoch": 1.9098142076502733,
+      "grad_norm": 0.8584185788739924,
+      "learning_rate": 6.161612238227768e-05,
+      "loss": 12.2891,
+      "step": 10921
+    },
+    {
+      "epoch": 1.9099890710382512,
+      "grad_norm": 0.9909185064224524,
+      "learning_rate": 6.159868844275063e-05,
+      "loss": 12.43,
+      "step": 10922
+    },
+    {
+      "epoch": 1.9101639344262296,
+      "grad_norm": 0.9289680757689105,
+      "learning_rate": 6.158125587218548e-05,
+      "loss": 12.1932,
+      "step": 10923
+    },
+    {
+      "epoch": 1.9103387978142077,
+      "grad_norm": 0.8214618497249393,
+      "learning_rate": 6.156382467120367e-05,
+      "loss": 12.3254,
+      "step": 10924
+    },
+    {
+      "epoch": 1.9105136612021858,
+      "grad_norm": 1.005234596044205,
+      "learning_rate": 6.154639484042665e-05,
+      "loss": 12.3089,
+      "step": 10925
+    },
+    {
+      "epoch": 1.910688524590164,
+      "grad_norm": 0.8839704742609666,
+      "learning_rate": 6.152896638047568e-05,
+      "loss": 12.481,
+      "step": 10926
+    },
+    {
+      "epoch": 1.910863387978142,
+      "grad_norm": 0.8625225618921266,
+      "learning_rate": 6.151153929197211e-05,
+      "loss": 12.3519,
+      "step": 10927
+    },
+    {
+      "epoch": 1.9110382513661204,
+      "grad_norm": 0.8782476843920616,
+      "learning_rate": 6.149411357553723e-05,
+      "loss": 12.2779,
+      "step": 10928
+    },
+    {
+      "epoch": 1.9112131147540983,
+      "grad_norm": 0.8536287679359156,
+      "learning_rate": 6.147668923179217e-05,
+      "loss": 12.0915,
+      "step": 10929
+    },
+    {
+      "epoch": 1.9113879781420766,
+      "grad_norm": 0.8620122133426223,
+      "learning_rate": 6.145926626135814e-05,
+      "loss": 12.2196,
+      "step": 10930
+    },
+    {
+      "epoch": 1.9115628415300545,
+      "grad_norm": 0.8541321997044491,
+      "learning_rate": 6.144184466485623e-05,
+      "loss": 12.2336,
+      "step": 10931
+    },
+    {
+      "epoch": 1.9117377049180329,
+      "grad_norm": 0.837624380973538,
+      "learning_rate": 6.142442444290756e-05,
+      "loss": 12.3557,
+      "step": 10932
+    },
+    {
+      "epoch": 1.911912568306011,
+      "grad_norm": 0.8410357764529042,
+      "learning_rate": 6.140700559613304e-05,
+      "loss": 12.2249,
+      "step": 10933
+    },
+    {
+      "epoch": 1.912087431693989,
+      "grad_norm": 0.8443643562019841,
+      "learning_rate": 6.138958812515367e-05,
+      "loss": 12.2519,
+      "step": 10934
+    },
+    {
+      "epoch": 1.9122622950819672,
+      "grad_norm": 0.8384421091034681,
+      "learning_rate": 6.13721720305904e-05,
+      "loss": 12.1514,
+      "step": 10935
+    },
+    {
+      "epoch": 1.9124371584699453,
+      "grad_norm": 0.9100518183522472,
+      "learning_rate": 6.135475731306406e-05,
+      "loss": 12.1846,
+      "step": 10936
+    },
+    {
+      "epoch": 1.9126120218579234,
+      "grad_norm": 0.8820749686874612,
+      "learning_rate": 6.133734397319547e-05,
+      "loss": 12.4458,
+      "step": 10937
+    },
+    {
+      "epoch": 1.9127868852459016,
+      "grad_norm": 0.8568057243475411,
+      "learning_rate": 6.131993201160537e-05,
+      "loss": 12.0987,
+      "step": 10938
+    },
+    {
+      "epoch": 1.91296174863388,
+      "grad_norm": 0.892601546417402,
+      "learning_rate": 6.130252142891452e-05,
+      "loss": 12.1791,
+      "step": 10939
+    },
+    {
+      "epoch": 1.9131366120218578,
+      "grad_norm": 0.8600995019765232,
+      "learning_rate": 6.12851122257436e-05,
+      "loss": 12.1334,
+      "step": 10940
+    },
+    {
+      "epoch": 1.9133114754098361,
+      "grad_norm": 0.8161767981392547,
+      "learning_rate": 6.126770440271316e-05,
+      "loss": 12.366,
+      "step": 10941
+    },
+    {
+      "epoch": 1.913486338797814,
+      "grad_norm": 0.850044213603243,
+      "learning_rate": 6.125029796044379e-05,
+      "loss": 12.0542,
+      "step": 10942
+    },
+    {
+      "epoch": 1.9136612021857924,
+      "grad_norm": 0.9609328460146946,
+      "learning_rate": 6.123289289955604e-05,
+      "loss": 12.1575,
+      "step": 10943
+    },
+    {
+      "epoch": 1.9138360655737705,
+      "grad_norm": 0.8211292467401703,
+      "learning_rate": 6.121548922067039e-05,
+      "loss": 12.2293,
+      "step": 10944
+    },
+    {
+      "epoch": 1.9140109289617486,
+      "grad_norm": 0.9109442546053526,
+      "learning_rate": 6.11980869244072e-05,
+      "loss": 12.3016,
+      "step": 10945
+    },
+    {
+      "epoch": 1.9141857923497267,
+      "grad_norm": 0.8372510657220408,
+      "learning_rate": 6.118068601138689e-05,
+      "loss": 12.1658,
+      "step": 10946
+    },
+    {
+      "epoch": 1.9143606557377049,
+      "grad_norm": 0.8766869348799818,
+      "learning_rate": 6.116328648222978e-05,
+      "loss": 12.1269,
+      "step": 10947
+    },
+    {
+      "epoch": 1.9145355191256832,
+      "grad_norm": 1.0109847660958247,
+      "learning_rate": 6.11458883375561e-05,
+      "loss": 12.2141,
+      "step": 10948
+    },
+    {
+      "epoch": 1.914710382513661,
+      "grad_norm": 0.8929093041282883,
+      "learning_rate": 6.112849157798612e-05,
+      "loss": 12.2832,
+      "step": 10949
+    },
+    {
+      "epoch": 1.9148852459016394,
+      "grad_norm": 0.7724589786404207,
+      "learning_rate": 6.111109620414e-05,
+      "loss": 12.3293,
+      "step": 10950
+    },
+    {
+      "epoch": 1.9150601092896173,
+      "grad_norm": 0.874986361293151,
+      "learning_rate": 6.109370221663788e-05,
+      "loss": 12.36,
+      "step": 10951
+    },
+    {
+      "epoch": 1.9152349726775957,
+      "grad_norm": 0.9258035730903107,
+      "learning_rate": 6.10763096160998e-05,
+      "loss": 12.1461,
+      "step": 10952
+    },
+    {
+      "epoch": 1.9154098360655738,
+      "grad_norm": 0.8521525529558457,
+      "learning_rate": 6.10589184031458e-05,
+      "loss": 12.1131,
+      "step": 10953
+    },
+    {
+      "epoch": 1.915584699453552,
+      "grad_norm": 0.8469425997247821,
+      "learning_rate": 6.104152857839586e-05,
+      "loss": 12.307,
+      "step": 10954
+    },
+    {
+      "epoch": 1.91575956284153,
+      "grad_norm": 0.8759185360600372,
+      "learning_rate": 6.102414014246993e-05,
+      "loss": 12.3596,
+      "step": 10955
+    },
+    {
+      "epoch": 1.9159344262295082,
+      "grad_norm": 0.9641256564120689,
+      "learning_rate": 6.1006753095987845e-05,
+      "loss": 12.1913,
+      "step": 10956
+    },
+    {
+      "epoch": 1.9161092896174865,
+      "grad_norm": 0.9297590520542015,
+      "learning_rate": 6.098936743956945e-05,
+      "loss": 12.497,
+      "step": 10957
+    },
+    {
+      "epoch": 1.9162841530054644,
+      "grad_norm": 0.8483543759504661,
+      "learning_rate": 6.0971983173834524e-05,
+      "loss": 12.2116,
+      "step": 10958
+    },
+    {
+      "epoch": 1.9164590163934427,
+      "grad_norm": 0.8965208802275759,
+      "learning_rate": 6.0954600299402844e-05,
+      "loss": 12.268,
+      "step": 10959
+    },
+    {
+      "epoch": 1.9166338797814206,
+      "grad_norm": 0.8476730033668306,
+      "learning_rate": 6.0937218816894e-05,
+      "loss": 12.2257,
+      "step": 10960
+    },
+    {
+      "epoch": 1.916808743169399,
+      "grad_norm": 0.9107661051923197,
+      "learning_rate": 6.091983872692767e-05,
+      "loss": 12.296,
+      "step": 10961
+    },
+    {
+      "epoch": 1.916983606557377,
+      "grad_norm": 0.7686433406362608,
+      "learning_rate": 6.0902460030123465e-05,
+      "loss": 12.1343,
+      "step": 10962
+    },
+    {
+      "epoch": 1.9171584699453552,
+      "grad_norm": 0.8312095855610594,
+      "learning_rate": 6.088508272710085e-05,
+      "loss": 12.296,
+      "step": 10963
+    },
+    {
+      "epoch": 1.9173333333333333,
+      "grad_norm": 0.94259648486346,
+      "learning_rate": 6.086770681847932e-05,
+      "loss": 12.0885,
+      "step": 10964
+    },
+    {
+      "epoch": 1.9175081967213115,
+      "grad_norm": 0.7941607913535185,
+      "learning_rate": 6.085033230487833e-05,
+      "loss": 11.9966,
+      "step": 10965
+    },
+    {
+      "epoch": 1.9176830601092896,
+      "grad_norm": 0.8357854284589631,
+      "learning_rate": 6.0832959186917274e-05,
+      "loss": 12.3067,
+      "step": 10966
+    },
+    {
+      "epoch": 1.9178579234972677,
+      "grad_norm": 0.8746057143798704,
+      "learning_rate": 6.081558746521543e-05,
+      "loss": 12.026,
+      "step": 10967
+    },
+    {
+      "epoch": 1.918032786885246,
+      "grad_norm": 0.7715153010445156,
+      "learning_rate": 6.079821714039211e-05,
+      "loss": 12.3211,
+      "step": 10968
+    },
+    {
+      "epoch": 1.918207650273224,
+      "grad_norm": 0.8700910045705993,
+      "learning_rate": 6.0780848213066554e-05,
+      "loss": 12.2202,
+      "step": 10969
+    },
+    {
+      "epoch": 1.9183825136612023,
+      "grad_norm": 0.87249127891326,
+      "learning_rate": 6.076348068385797e-05,
+      "loss": 12.2263,
+      "step": 10970
+    },
+    {
+      "epoch": 1.9185573770491802,
+      "grad_norm": 0.8864084062274343,
+      "learning_rate": 6.074611455338544e-05,
+      "loss": 12.0735,
+      "step": 10971
+    },
+    {
+      "epoch": 1.9187322404371585,
+      "grad_norm": 0.9033783888506872,
+      "learning_rate": 6.0728749822268036e-05,
+      "loss": 12.1826,
+      "step": 10972
+    },
+    {
+      "epoch": 1.9189071038251366,
+      "grad_norm": 0.9701482938456613,
+      "learning_rate": 6.0711386491124825e-05,
+      "loss": 12.2309,
+      "step": 10973
+    },
+    {
+      "epoch": 1.9190819672131147,
+      "grad_norm": 0.7780101452342744,
+      "learning_rate": 6.0694024560574805e-05,
+      "loss": 12.2162,
+      "step": 10974
+    },
+    {
+      "epoch": 1.9192568306010929,
+      "grad_norm": 0.8911809734169515,
+      "learning_rate": 6.0676664031236864e-05,
+      "loss": 12.179,
+      "step": 10975
+    },
+    {
+      "epoch": 1.919431693989071,
+      "grad_norm": 0.890648033540949,
+      "learning_rate": 6.0659304903729905e-05,
+      "loss": 12.0903,
+      "step": 10976
+    },
+    {
+      "epoch": 1.9196065573770493,
+      "grad_norm": 0.8538655051842282,
+      "learning_rate": 6.06419471786728e-05,
+      "loss": 12.3915,
+      "step": 10977
+    },
+    {
+      "epoch": 1.9197814207650272,
+      "grad_norm": 0.8825789479525016,
+      "learning_rate": 6.062459085668426e-05,
+      "loss": 12.2598,
+      "step": 10978
+    },
+    {
+      "epoch": 1.9199562841530056,
+      "grad_norm": 0.7890001672125353,
+      "learning_rate": 6.0607235938383046e-05,
+      "loss": 11.8272,
+      "step": 10979
+    },
+    {
+      "epoch": 1.9201311475409835,
+      "grad_norm": 0.8291895724522736,
+      "learning_rate": 6.0589882424387845e-05,
+      "loss": 12.3617,
+      "step": 10980
+    },
+    {
+      "epoch": 1.9203060109289618,
+      "grad_norm": 0.9328620089918747,
+      "learning_rate": 6.057253031531731e-05,
+      "loss": 12.243,
+      "step": 10981
+    },
+    {
+      "epoch": 1.92048087431694,
+      "grad_norm": 0.8676044252706024,
+      "learning_rate": 6.055517961178998e-05,
+      "loss": 12.2301,
+      "step": 10982
+    },
+    {
+      "epoch": 1.920655737704918,
+      "grad_norm": 0.8815551145962758,
+      "learning_rate": 6.0537830314424414e-05,
+      "loss": 12.2716,
+      "step": 10983
+    },
+    {
+      "epoch": 1.9208306010928962,
+      "grad_norm": 0.8126844454450843,
+      "learning_rate": 6.05204824238391e-05,
+      "loss": 12.3894,
+      "step": 10984
+    },
+    {
+      "epoch": 1.9210054644808743,
+      "grad_norm": 0.8739995322728398,
+      "learning_rate": 6.0503135940652486e-05,
+      "loss": 12.1455,
+      "step": 10985
+    },
+    {
+      "epoch": 1.9211803278688526,
+      "grad_norm": 0.9304539478321003,
+      "learning_rate": 6.048579086548289e-05,
+      "loss": 12.0091,
+      "step": 10986
+    },
+    {
+      "epoch": 1.9213551912568305,
+      "grad_norm": 0.8255937894830452,
+      "learning_rate": 6.046844719894869e-05,
+      "loss": 12.5608,
+      "step": 10987
+    },
+    {
+      "epoch": 1.9215300546448089,
+      "grad_norm": 0.945820928491554,
+      "learning_rate": 6.0451104941668156e-05,
+      "loss": 12.3827,
+      "step": 10988
+    },
+    {
+      "epoch": 1.9217049180327868,
+      "grad_norm": 0.9548807918609383,
+      "learning_rate": 6.043376409425956e-05,
+      "loss": 12.1049,
+      "step": 10989
+    },
+    {
+      "epoch": 1.921879781420765,
+      "grad_norm": 1.012652078906236,
+      "learning_rate": 6.041642465734101e-05,
+      "loss": 12.2198,
+      "step": 10990
+    },
+    {
+      "epoch": 1.9220546448087432,
+      "grad_norm": 0.9498844935729711,
+      "learning_rate": 6.039908663153068e-05,
+      "loss": 12.0985,
+      "step": 10991
+    },
+    {
+      "epoch": 1.9222295081967213,
+      "grad_norm": 0.837963401347531,
+      "learning_rate": 6.038175001744665e-05,
+      "loss": 12.3435,
+      "step": 10992
+    },
+    {
+      "epoch": 1.9224043715846995,
+      "grad_norm": 0.9335931677600102,
+      "learning_rate": 6.0364414815706924e-05,
+      "loss": 12.1565,
+      "step": 10993
+    },
+    {
+      "epoch": 1.9225792349726776,
+      "grad_norm": 0.8404197446759898,
+      "learning_rate": 6.034708102692951e-05,
+      "loss": 12.4355,
+      "step": 10994
+    },
+    {
+      "epoch": 1.9227540983606557,
+      "grad_norm": 0.9461031365253384,
+      "learning_rate": 6.032974865173232e-05,
+      "loss": 12.2399,
+      "step": 10995
+    },
+    {
+      "epoch": 1.9229289617486338,
+      "grad_norm": 0.8593090102991852,
+      "learning_rate": 6.031241769073328e-05,
+      "loss": 12.0649,
+      "step": 10996
+    },
+    {
+      "epoch": 1.9231038251366122,
+      "grad_norm": 0.8833824424303988,
+      "learning_rate": 6.0295088144550135e-05,
+      "loss": 12.2669,
+      "step": 10997
+    },
+    {
+      "epoch": 1.92327868852459,
+      "grad_norm": 0.8660877335883567,
+      "learning_rate": 6.027776001380072e-05,
+      "loss": 12.3602,
+      "step": 10998
+    },
+    {
+      "epoch": 1.9234535519125684,
+      "grad_norm": 0.9224705590670178,
+      "learning_rate": 6.026043329910275e-05,
+      "loss": 12.3268,
+      "step": 10999
+    },
+    {
+      "epoch": 1.9236284153005463,
+      "grad_norm": 0.8490994713555499,
+      "learning_rate": 6.024310800107391e-05,
+      "loss": 12.3255,
+      "step": 11000
+    },
+    {
+      "epoch": 1.9238032786885246,
+      "grad_norm": 0.8652224666220857,
+      "learning_rate": 6.022578412033181e-05,
+      "loss": 12.271,
+      "step": 11001
+    },
+    {
+      "epoch": 1.9239781420765028,
+      "grad_norm": 0.8892610081852007,
+      "learning_rate": 6.020846165749403e-05,
+      "loss": 12.3416,
+      "step": 11002
+    },
+    {
+      "epoch": 1.9241530054644809,
+      "grad_norm": 0.8245325996002593,
+      "learning_rate": 6.01911406131781e-05,
+      "loss": 12.2394,
+      "step": 11003
+    },
+    {
+      "epoch": 1.924327868852459,
+      "grad_norm": 0.9555637960732583,
+      "learning_rate": 6.017382098800153e-05,
+      "loss": 12.3573,
+      "step": 11004
+    },
+    {
+      "epoch": 1.924502732240437,
+      "grad_norm": 0.8551371473941503,
+      "learning_rate": 6.015650278258168e-05,
+      "loss": 12.1031,
+      "step": 11005
+    },
+    {
+      "epoch": 1.9246775956284154,
+      "grad_norm": 0.8192872977336978,
+      "learning_rate": 6.013918599753595e-05,
+      "loss": 12.256,
+      "step": 11006
+    },
+    {
+      "epoch": 1.9248524590163933,
+      "grad_norm": 0.8762594058258517,
+      "learning_rate": 6.0121870633481714e-05,
+      "loss": 11.9938,
+      "step": 11007
+    },
+    {
+      "epoch": 1.9250273224043717,
+      "grad_norm": 0.7945563570257301,
+      "learning_rate": 6.010455669103614e-05,
+      "loss": 12.3169,
+      "step": 11008
+    },
+    {
+      "epoch": 1.9252021857923496,
+      "grad_norm": 0.8634508285477696,
+      "learning_rate": 6.008724417081652e-05,
+      "loss": 12.2532,
+      "step": 11009
+    },
+    {
+      "epoch": 1.925377049180328,
+      "grad_norm": 0.9603075446968895,
+      "learning_rate": 6.006993307344001e-05,
+      "loss": 11.9219,
+      "step": 11010
+    },
+    {
+      "epoch": 1.925551912568306,
+      "grad_norm": 0.789532978629523,
+      "learning_rate": 6.0052623399523755e-05,
+      "loss": 12.127,
+      "step": 11011
+    },
+    {
+      "epoch": 1.9257267759562842,
+      "grad_norm": 0.7891559957265806,
+      "learning_rate": 6.003531514968478e-05,
+      "loss": 12.0623,
+      "step": 11012
+    },
+    {
+      "epoch": 1.9259016393442623,
+      "grad_norm": 0.8910975462027739,
+      "learning_rate": 6.001800832454012e-05,
+      "loss": 12.2062,
+      "step": 11013
+    },
+    {
+      "epoch": 1.9260765027322404,
+      "grad_norm": 0.8171563397197296,
+      "learning_rate": 6.0000702924706765e-05,
+      "loss": 12.3992,
+      "step": 11014
+    },
+    {
+      "epoch": 1.9262513661202187,
+      "grad_norm": 0.939512254137984,
+      "learning_rate": 5.998339895080164e-05,
+      "loss": 12.4613,
+      "step": 11015
+    },
+    {
+      "epoch": 1.9264262295081966,
+      "grad_norm": 0.8347457394336618,
+      "learning_rate": 5.996609640344155e-05,
+      "loss": 12.2164,
+      "step": 11016
+    },
+    {
+      "epoch": 1.926601092896175,
+      "grad_norm": 0.8608907697701501,
+      "learning_rate": 5.9948795283243356e-05,
+      "loss": 12.303,
+      "step": 11017
+    },
+    {
+      "epoch": 1.9267759562841529,
+      "grad_norm": 0.8640491683199534,
+      "learning_rate": 5.993149559082382e-05,
+      "loss": 12.2356,
+      "step": 11018
+    },
+    {
+      "epoch": 1.9269508196721312,
+      "grad_norm": 0.802224492286506,
+      "learning_rate": 5.991419732679966e-05,
+      "loss": 12.1662,
+      "step": 11019
+    },
+    {
+      "epoch": 1.9271256830601093,
+      "grad_norm": 0.7916104947750318,
+      "learning_rate": 5.989690049178751e-05,
+      "loss": 12.1941,
+      "step": 11020
+    },
+    {
+      "epoch": 1.9273005464480875,
+      "grad_norm": 0.8541347001203861,
+      "learning_rate": 5.9879605086404003e-05,
+      "loss": 12.0423,
+      "step": 11021
+    },
+    {
+      "epoch": 1.9274754098360656,
+      "grad_norm": 0.9141419762715097,
+      "learning_rate": 5.986231111126575e-05,
+      "loss": 12.2205,
+      "step": 11022
+    },
+    {
+      "epoch": 1.9276502732240437,
+      "grad_norm": 0.9497248690650949,
+      "learning_rate": 5.984501856698917e-05,
+      "loss": 12.5016,
+      "step": 11023
+    },
+    {
+      "epoch": 1.9278251366120218,
+      "grad_norm": 0.9876691464959881,
+      "learning_rate": 5.982772745419074e-05,
+      "loss": 12.2327,
+      "step": 11024
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 0.8940725523757261,
+      "learning_rate": 5.9810437773486906e-05,
+      "loss": 12.3106,
+      "step": 11025
+    },
+    {
+      "epoch": 1.9281748633879783,
+      "grad_norm": 1.0840672826666091,
+      "learning_rate": 5.979314952549402e-05,
+      "loss": 12.3255,
+      "step": 11026
+    },
+    {
+      "epoch": 1.9283497267759562,
+      "grad_norm": 1.0958213520270355,
+      "learning_rate": 5.977586271082837e-05,
+      "loss": 12.0949,
+      "step": 11027
+    },
+    {
+      "epoch": 1.9285245901639345,
+      "grad_norm": 0.9069712711799168,
+      "learning_rate": 5.975857733010621e-05,
+      "loss": 12.0493,
+      "step": 11028
+    },
+    {
+      "epoch": 1.9286994535519124,
+      "grad_norm": 1.0278671965564983,
+      "learning_rate": 5.9741293383943756e-05,
+      "loss": 12.2147,
+      "step": 11029
+    },
+    {
+      "epoch": 1.9288743169398908,
+      "grad_norm": 1.0636509191652934,
+      "learning_rate": 5.9724010872957184e-05,
+      "loss": 12.125,
+      "step": 11030
+    },
+    {
+      "epoch": 1.9290491803278689,
+      "grad_norm": 0.8095562182251459,
+      "learning_rate": 5.970672979776254e-05,
+      "loss": 12.1555,
+      "step": 11031
+    },
+    {
+      "epoch": 1.929224043715847,
+      "grad_norm": 0.9535076406602162,
+      "learning_rate": 5.968945015897589e-05,
+      "loss": 12.1783,
+      "step": 11032
+    },
+    {
+      "epoch": 1.9293989071038251,
+      "grad_norm": 1.1019782436845906,
+      "learning_rate": 5.967217195721325e-05,
+      "loss": 12.1976,
+      "step": 11033
+    },
+    {
+      "epoch": 1.9295737704918032,
+      "grad_norm": 0.8959111975759978,
+      "learning_rate": 5.9654895193090596e-05,
+      "loss": 12.2972,
+      "step": 11034
+    },
+    {
+      "epoch": 1.9297486338797816,
+      "grad_norm": 0.9825086762696388,
+      "learning_rate": 5.963761986722376e-05,
+      "loss": 12.4105,
+      "step": 11035
+    },
+    {
+      "epoch": 1.9299234972677595,
+      "grad_norm": 0.948791109266962,
+      "learning_rate": 5.962034598022862e-05,
+      "loss": 12.0775,
+      "step": 11036
+    },
+    {
+      "epoch": 1.9300983606557378,
+      "grad_norm": 0.8640728550865847,
+      "learning_rate": 5.9603073532720974e-05,
+      "loss": 12.3013,
+      "step": 11037
+    },
+    {
+      "epoch": 1.9302732240437157,
+      "grad_norm": 0.9580873853604052,
+      "learning_rate": 5.9585802525316546e-05,
+      "loss": 12.1064,
+      "step": 11038
+    },
+    {
+      "epoch": 1.930448087431694,
+      "grad_norm": 0.8628841058181292,
+      "learning_rate": 5.956853295863104e-05,
+      "loss": 12.3765,
+      "step": 11039
+    },
+    {
+      "epoch": 1.9306229508196722,
+      "grad_norm": 0.8328776698133872,
+      "learning_rate": 5.955126483328008e-05,
+      "loss": 12.3169,
+      "step": 11040
+    },
+    {
+      "epoch": 1.9307978142076503,
+      "grad_norm": 0.9196536946427678,
+      "learning_rate": 5.953399814987933e-05,
+      "loss": 12.1793,
+      "step": 11041
+    },
+    {
+      "epoch": 1.9309726775956284,
+      "grad_norm": 1.0386614491009036,
+      "learning_rate": 5.951673290904423e-05,
+      "loss": 12.5673,
+      "step": 11042
+    },
+    {
+      "epoch": 1.9311475409836065,
+      "grad_norm": 0.8231590932550457,
+      "learning_rate": 5.9499469111390284e-05,
+      "loss": 12.1521,
+      "step": 11043
+    },
+    {
+      "epoch": 1.9313224043715846,
+      "grad_norm": 1.1208871005414247,
+      "learning_rate": 5.948220675753295e-05,
+      "loss": 12.3653,
+      "step": 11044
+    },
+    {
+      "epoch": 1.9314972677595628,
+      "grad_norm": 0.9089576193975838,
+      "learning_rate": 5.9464945848087615e-05,
+      "loss": 12.3509,
+      "step": 11045
+    },
+    {
+      "epoch": 1.931672131147541,
+      "grad_norm": 0.8206146528987119,
+      "learning_rate": 5.9447686383669596e-05,
+      "loss": 12.2933,
+      "step": 11046
+    },
+    {
+      "epoch": 1.931846994535519,
+      "grad_norm": 1.0077612047638134,
+      "learning_rate": 5.943042836489417e-05,
+      "loss": 12.1991,
+      "step": 11047
+    },
+    {
+      "epoch": 1.9320218579234973,
+      "grad_norm": 0.9560096574788073,
+      "learning_rate": 5.941317179237658e-05,
+      "loss": 12.3023,
+      "step": 11048
+    },
+    {
+      "epoch": 1.9321967213114755,
+      "grad_norm": 0.903056162450902,
+      "learning_rate": 5.939591666673203e-05,
+      "loss": 12.3917,
+      "step": 11049
+    },
+    {
+      "epoch": 1.9323715846994536,
+      "grad_norm": 0.990863930656871,
+      "learning_rate": 5.937866298857557e-05,
+      "loss": 12.3135,
+      "step": 11050
+    },
+    {
+      "epoch": 1.9325464480874317,
+      "grad_norm": 0.9232719498743704,
+      "learning_rate": 5.9361410758522305e-05,
+      "loss": 12.1946,
+      "step": 11051
+    },
+    {
+      "epoch": 1.9327213114754098,
+      "grad_norm": 0.8124387745988718,
+      "learning_rate": 5.934415997718727e-05,
+      "loss": 12.1639,
+      "step": 11052
+    },
+    {
+      "epoch": 1.932896174863388,
+      "grad_norm": 1.0913883007917873,
+      "learning_rate": 5.932691064518549e-05,
+      "loss": 12.2192,
+      "step": 11053
+    },
+    {
+      "epoch": 1.933071038251366,
+      "grad_norm": 0.9187586573282348,
+      "learning_rate": 5.930966276313177e-05,
+      "loss": 12.3954,
+      "step": 11054
+    },
+    {
+      "epoch": 1.9332459016393444,
+      "grad_norm": 0.8170846265297577,
+      "learning_rate": 5.929241633164102e-05,
+      "loss": 12.057,
+      "step": 11055
+    },
+    {
+      "epoch": 1.9334207650273223,
+      "grad_norm": 1.052918991219279,
+      "learning_rate": 5.9275171351328096e-05,
+      "loss": 12.3434,
+      "step": 11056
+    },
+    {
+      "epoch": 1.9335956284153006,
+      "grad_norm": 0.9271423122723286,
+      "learning_rate": 5.925792782280771e-05,
+      "loss": 12.3187,
+      "step": 11057
+    },
+    {
+      "epoch": 1.9337704918032785,
+      "grad_norm": 0.8013342306131593,
+      "learning_rate": 5.924068574669459e-05,
+      "loss": 12.3207,
+      "step": 11058
+    },
+    {
+      "epoch": 1.9339453551912569,
+      "grad_norm": 0.9168656252006515,
+      "learning_rate": 5.922344512360341e-05,
+      "loss": 11.9718,
+      "step": 11059
+    },
+    {
+      "epoch": 1.934120218579235,
+      "grad_norm": 0.9786288572221411,
+      "learning_rate": 5.920620595414881e-05,
+      "loss": 12.2706,
+      "step": 11060
+    },
+    {
+      "epoch": 1.9342950819672131,
+      "grad_norm": 1.0375450317911805,
+      "learning_rate": 5.918896823894526e-05,
+      "loss": 12.1156,
+      "step": 11061
+    },
+    {
+      "epoch": 1.9344699453551912,
+      "grad_norm": 0.7719969001324037,
+      "learning_rate": 5.9171731978607304e-05,
+      "loss": 12.2672,
+      "step": 11062
+    },
+    {
+      "epoch": 1.9346448087431694,
+      "grad_norm": 1.2064858680281048,
+      "learning_rate": 5.915449717374941e-05,
+      "loss": 12.1152,
+      "step": 11063
+    },
+    {
+      "epoch": 1.9348196721311477,
+      "grad_norm": 0.9820166591910283,
+      "learning_rate": 5.913726382498599e-05,
+      "loss": 12.1613,
+      "step": 11064
+    },
+    {
+      "epoch": 1.9349945355191256,
+      "grad_norm": 0.8926746696179138,
+      "learning_rate": 5.912003193293134e-05,
+      "loss": 12.079,
+      "step": 11065
+    },
+    {
+      "epoch": 1.935169398907104,
+      "grad_norm": 1.0997036636068749,
+      "learning_rate": 5.91028014981998e-05,
+      "loss": 12.3135,
+      "step": 11066
+    },
+    {
+      "epoch": 1.9353442622950818,
+      "grad_norm": 0.9363687300806661,
+      "learning_rate": 5.908557252140561e-05,
+      "loss": 12.3826,
+      "step": 11067
+    },
+    {
+      "epoch": 1.9355191256830602,
+      "grad_norm": 0.9134663581672803,
+      "learning_rate": 5.9068345003162985e-05,
+      "loss": 12.3387,
+      "step": 11068
+    },
+    {
+      "epoch": 1.9356939890710383,
+      "grad_norm": 1.0205638111380826,
+      "learning_rate": 5.905111894408599e-05,
+      "loss": 12.2484,
+      "step": 11069
+    },
+    {
+      "epoch": 1.9358688524590164,
+      "grad_norm": 0.8117283576662402,
+      "learning_rate": 5.903389434478877e-05,
+      "loss": 12.3941,
+      "step": 11070
+    },
+    {
+      "epoch": 1.9360437158469945,
+      "grad_norm": 0.9047889880861556,
+      "learning_rate": 5.9016671205885365e-05,
+      "loss": 12.3419,
+      "step": 11071
+    },
+    {
+      "epoch": 1.9362185792349726,
+      "grad_norm": 0.9128657414152593,
+      "learning_rate": 5.8999449527989745e-05,
+      "loss": 12.2393,
+      "step": 11072
+    },
+    {
+      "epoch": 1.9363934426229508,
+      "grad_norm": 0.89662673363326,
+      "learning_rate": 5.898222931171583e-05,
+      "loss": 12.4291,
+      "step": 11073
+    },
+    {
+      "epoch": 1.9365683060109289,
+      "grad_norm": 0.896765976301233,
+      "learning_rate": 5.896501055767752e-05,
+      "loss": 12.4098,
+      "step": 11074
+    },
+    {
+      "epoch": 1.9367431693989072,
+      "grad_norm": 0.8441360465558105,
+      "learning_rate": 5.8947793266488674e-05,
+      "loss": 12.273,
+      "step": 11075
+    },
+    {
+      "epoch": 1.9369180327868851,
+      "grad_norm": 0.8555383823943771,
+      "learning_rate": 5.8930577438763004e-05,
+      "loss": 12.1131,
+      "step": 11076
+    },
+    {
+      "epoch": 1.9370928961748635,
+      "grad_norm": 0.8419860332338182,
+      "learning_rate": 5.891336307511426e-05,
+      "loss": 12.3006,
+      "step": 11077
+    },
+    {
+      "epoch": 1.9372677595628416,
+      "grad_norm": 0.8064063286162753,
+      "learning_rate": 5.889615017615612e-05,
+      "loss": 11.8253,
+      "step": 11078
+    },
+    {
+      "epoch": 1.9374426229508197,
+      "grad_norm": 0.8566911797858311,
+      "learning_rate": 5.887893874250223e-05,
+      "loss": 12.2157,
+      "step": 11079
+    },
+    {
+      "epoch": 1.9376174863387978,
+      "grad_norm": 0.9049797475431768,
+      "learning_rate": 5.88617287747661e-05,
+      "loss": 12.0653,
+      "step": 11080
+    },
+    {
+      "epoch": 1.937792349726776,
+      "grad_norm": 0.818760470681919,
+      "learning_rate": 5.8844520273561274e-05,
+      "loss": 12.3015,
+      "step": 11081
+    },
+    {
+      "epoch": 1.937967213114754,
+      "grad_norm": 0.8675745650364827,
+      "learning_rate": 5.8827313239501216e-05,
+      "loss": 12.0188,
+      "step": 11082
+    },
+    {
+      "epoch": 1.9381420765027322,
+      "grad_norm": 0.9540841723924383,
+      "learning_rate": 5.881010767319936e-05,
+      "loss": 12.016,
+      "step": 11083
+    },
+    {
+      "epoch": 1.9383169398907105,
+      "grad_norm": 0.828788637735626,
+      "learning_rate": 5.879290357526901e-05,
+      "loss": 12.4105,
+      "step": 11084
+    },
+    {
+      "epoch": 1.9384918032786884,
+      "grad_norm": 0.8822597409162242,
+      "learning_rate": 5.87757009463235e-05,
+      "loss": 12.1624,
+      "step": 11085
+    },
+    {
+      "epoch": 1.9386666666666668,
+      "grad_norm": 0.8068591949816425,
+      "learning_rate": 5.875849978697614e-05,
+      "loss": 12.0988,
+      "step": 11086
+    },
+    {
+      "epoch": 1.9388415300546447,
+      "grad_norm": 0.8655839765640958,
+      "learning_rate": 5.874130009784002e-05,
+      "loss": 12.29,
+      "step": 11087
+    },
+    {
+      "epoch": 1.939016393442623,
+      "grad_norm": 0.8979314777124198,
+      "learning_rate": 5.872410187952835e-05,
+      "loss": 12.0005,
+      "step": 11088
+    },
+    {
+      "epoch": 1.9391912568306011,
+      "grad_norm": 0.8186660796640594,
+      "learning_rate": 5.8706905132654235e-05,
+      "loss": 11.9694,
+      "step": 11089
+    },
+    {
+      "epoch": 1.9393661202185792,
+      "grad_norm": 0.9060825093144664,
+      "learning_rate": 5.868970985783071e-05,
+      "loss": 12.1497,
+      "step": 11090
+    },
+    {
+      "epoch": 1.9395409836065574,
+      "grad_norm": 0.9683489726675992,
+      "learning_rate": 5.8672516055670754e-05,
+      "loss": 12.4312,
+      "step": 11091
+    },
+    {
+      "epoch": 1.9397158469945355,
+      "grad_norm": 0.8675230240341153,
+      "learning_rate": 5.8655323726787313e-05,
+      "loss": 11.9711,
+      "step": 11092
+    },
+    {
+      "epoch": 1.9398907103825138,
+      "grad_norm": 0.9121740385513901,
+      "learning_rate": 5.863813287179327e-05,
+      "loss": 12.1631,
+      "step": 11093
+    },
+    {
+      "epoch": 1.9400655737704917,
+      "grad_norm": 0.9280750668054052,
+      "learning_rate": 5.8620943491301516e-05,
+      "loss": 12.4073,
+      "step": 11094
+    },
+    {
+      "epoch": 1.94024043715847,
+      "grad_norm": 0.7879596828362316,
+      "learning_rate": 5.860375558592473e-05,
+      "loss": 12.1742,
+      "step": 11095
+    },
+    {
+      "epoch": 1.940415300546448,
+      "grad_norm": 0.8070988444157706,
+      "learning_rate": 5.8586569156275696e-05,
+      "loss": 12.2327,
+      "step": 11096
+    },
+    {
+      "epoch": 1.9405901639344263,
+      "grad_norm": 0.9225752984755143,
+      "learning_rate": 5.856938420296707e-05,
+      "loss": 12.2271,
+      "step": 11097
+    },
+    {
+      "epoch": 1.9407650273224044,
+      "grad_norm": 0.9162584008928971,
+      "learning_rate": 5.855220072661154e-05,
+      "loss": 12.2232,
+      "step": 11098
+    },
+    {
+      "epoch": 1.9409398907103825,
+      "grad_norm": 0.9719910921344835,
+      "learning_rate": 5.853501872782159e-05,
+      "loss": 12.2534,
+      "step": 11099
+    },
+    {
+      "epoch": 1.9411147540983607,
+      "grad_norm": 0.8080083834525469,
+      "learning_rate": 5.851783820720977e-05,
+      "loss": 12.118,
+      "step": 11100
+    },
+    {
+      "epoch": 1.9412896174863388,
+      "grad_norm": 0.8488092086956848,
+      "learning_rate": 5.8500659165388585e-05,
+      "loss": 12.1916,
+      "step": 11101
+    },
+    {
+      "epoch": 1.941464480874317,
+      "grad_norm": 1.03184009439455,
+      "learning_rate": 5.848348160297039e-05,
+      "loss": 12.3526,
+      "step": 11102
+    },
+    {
+      "epoch": 1.941639344262295,
+      "grad_norm": 0.8230400346575164,
+      "learning_rate": 5.8466305520567564e-05,
+      "loss": 12.2375,
+      "step": 11103
+    },
+    {
+      "epoch": 1.9418142076502733,
+      "grad_norm": 0.9916500153458733,
+      "learning_rate": 5.8449130918792424e-05,
+      "loss": 12.3771,
+      "step": 11104
+    },
+    {
+      "epoch": 1.9419890710382512,
+      "grad_norm": 1.120259968448504,
+      "learning_rate": 5.8431957798257254e-05,
+      "loss": 12.249,
+      "step": 11105
+    },
+    {
+      "epoch": 1.9421639344262296,
+      "grad_norm": 0.7746994107861325,
+      "learning_rate": 5.841478615957419e-05,
+      "loss": 12.1862,
+      "step": 11106
+    },
+    {
+      "epoch": 1.9423387978142075,
+      "grad_norm": 0.9930422435376264,
+      "learning_rate": 5.8397616003355404e-05,
+      "loss": 12.2689,
+      "step": 11107
+    },
+    {
+      "epoch": 1.9425136612021858,
+      "grad_norm": 0.9647409836926931,
+      "learning_rate": 5.838044733021297e-05,
+      "loss": 12.3217,
+      "step": 11108
+    },
+    {
+      "epoch": 1.942688524590164,
+      "grad_norm": 0.8283201924833815,
+      "learning_rate": 5.8363280140759043e-05,
+      "loss": 11.9874,
+      "step": 11109
+    },
+    {
+      "epoch": 1.942863387978142,
+      "grad_norm": 0.8906206821355735,
+      "learning_rate": 5.834611443560548e-05,
+      "loss": 11.9406,
+      "step": 11110
+    },
+    {
+      "epoch": 1.9430382513661202,
+      "grad_norm": 0.7984826525507134,
+      "learning_rate": 5.8328950215364266e-05,
+      "loss": 12.1782,
+      "step": 11111
+    },
+    {
+      "epoch": 1.9432131147540983,
+      "grad_norm": 0.9077530660816675,
+      "learning_rate": 5.83117874806473e-05,
+      "loss": 12.3477,
+      "step": 11112
+    },
+    {
+      "epoch": 1.9433879781420766,
+      "grad_norm": 0.9857202609405887,
+      "learning_rate": 5.829462623206644e-05,
+      "loss": 12.3711,
+      "step": 11113
+    },
+    {
+      "epoch": 1.9435628415300545,
+      "grad_norm": 0.7999080768286219,
+      "learning_rate": 5.8277466470233376e-05,
+      "loss": 12.2309,
+      "step": 11114
+    },
+    {
+      "epoch": 1.9437377049180329,
+      "grad_norm": 0.838157978257301,
+      "learning_rate": 5.82603081957599e-05,
+      "loss": 12.3751,
+      "step": 11115
+    },
+    {
+      "epoch": 1.9439125683060108,
+      "grad_norm": 0.7904185164462837,
+      "learning_rate": 5.8243151409257715e-05,
+      "loss": 12.1012,
+      "step": 11116
+    },
+    {
+      "epoch": 1.9440874316939891,
+      "grad_norm": 0.897274934501879,
+      "learning_rate": 5.8225996111338344e-05,
+      "loss": 12.474,
+      "step": 11117
+    },
+    {
+      "epoch": 1.9442622950819672,
+      "grad_norm": 0.832761919104939,
+      "learning_rate": 5.8208842302613395e-05,
+      "loss": 12.4753,
+      "step": 11118
+    },
+    {
+      "epoch": 1.9444371584699454,
+      "grad_norm": 0.8667337014196432,
+      "learning_rate": 5.81916899836944e-05,
+      "loss": 12.2141,
+      "step": 11119
+    },
+    {
+      "epoch": 1.9446120218579235,
+      "grad_norm": 0.8114406123857004,
+      "learning_rate": 5.8174539155192845e-05,
+      "loss": 12.4576,
+      "step": 11120
+    },
+    {
+      "epoch": 1.9447868852459016,
+      "grad_norm": 0.8277565550854526,
+      "learning_rate": 5.815738981772007e-05,
+      "loss": 12.1736,
+      "step": 11121
+    },
+    {
+      "epoch": 1.94496174863388,
+      "grad_norm": 0.8638855575742672,
+      "learning_rate": 5.8140241971887455e-05,
+      "loss": 12.0402,
+      "step": 11122
+    },
+    {
+      "epoch": 1.9451366120218578,
+      "grad_norm": 0.9552508249010787,
+      "learning_rate": 5.81230956183063e-05,
+      "loss": 12.4608,
+      "step": 11123
+    },
+    {
+      "epoch": 1.9453114754098362,
+      "grad_norm": 0.8433434081930071,
+      "learning_rate": 5.8105950757587866e-05,
+      "loss": 12.4097,
+      "step": 11124
+    },
+    {
+      "epoch": 1.945486338797814,
+      "grad_norm": 0.832310529977346,
+      "learning_rate": 5.8088807390343334e-05,
+      "loss": 12.1564,
+      "step": 11125
+    },
+    {
+      "epoch": 1.9456612021857924,
+      "grad_norm": 0.8156188895163821,
+      "learning_rate": 5.807166551718386e-05,
+      "loss": 12.3455,
+      "step": 11126
+    },
+    {
+      "epoch": 1.9458360655737705,
+      "grad_norm": 0.9407325396038491,
+      "learning_rate": 5.805452513872052e-05,
+      "loss": 12.4859,
+      "step": 11127
+    },
+    {
+      "epoch": 1.9460109289617487,
+      "grad_norm": 0.9272640770951925,
+      "learning_rate": 5.8037386255564385e-05,
+      "loss": 12.2322,
+      "step": 11128
+    },
+    {
+      "epoch": 1.9461857923497268,
+      "grad_norm": 0.8424486125662835,
+      "learning_rate": 5.8020248868326355e-05,
+      "loss": 12.1777,
+      "step": 11129
+    },
+    {
+      "epoch": 1.946360655737705,
+      "grad_norm": 0.9543838172520963,
+      "learning_rate": 5.8003112977617414e-05,
+      "loss": 12.4262,
+      "step": 11130
+    },
+    {
+      "epoch": 1.946535519125683,
+      "grad_norm": 0.8284793318935559,
+      "learning_rate": 5.798597858404846e-05,
+      "loss": 11.9367,
+      "step": 11131
+    },
+    {
+      "epoch": 1.9467103825136611,
+      "grad_norm": 0.9828265406504,
+      "learning_rate": 5.796884568823026e-05,
+      "loss": 12.2657,
+      "step": 11132
+    },
+    {
+      "epoch": 1.9468852459016395,
+      "grad_norm": 0.8676873059947362,
+      "learning_rate": 5.7951714290773584e-05,
+      "loss": 12.4223,
+      "step": 11133
+    },
+    {
+      "epoch": 1.9470601092896174,
+      "grad_norm": 0.8579316495356852,
+      "learning_rate": 5.793458439228918e-05,
+      "loss": 12.3598,
+      "step": 11134
+    },
+    {
+      "epoch": 1.9472349726775957,
+      "grad_norm": 0.9504463274607737,
+      "learning_rate": 5.791745599338772e-05,
+      "loss": 12.2068,
+      "step": 11135
+    },
+    {
+      "epoch": 1.9474098360655736,
+      "grad_norm": 0.9221883068943194,
+      "learning_rate": 5.790032909467975e-05,
+      "loss": 12.1595,
+      "step": 11136
+    },
+    {
+      "epoch": 1.947584699453552,
+      "grad_norm": 0.8218262543881443,
+      "learning_rate": 5.7883203696775865e-05,
+      "loss": 12.4571,
+      "step": 11137
+    },
+    {
+      "epoch": 1.94775956284153,
+      "grad_norm": 0.9127630906493763,
+      "learning_rate": 5.7866079800286555e-05,
+      "loss": 12.4713,
+      "step": 11138
+    },
+    {
+      "epoch": 1.9479344262295082,
+      "grad_norm": 0.8485891386620642,
+      "learning_rate": 5.7848957405822325e-05,
+      "loss": 11.7684,
+      "step": 11139
+    },
+    {
+      "epoch": 1.9481092896174863,
+      "grad_norm": 0.9015612094201875,
+      "learning_rate": 5.7831836513993466e-05,
+      "loss": 12.4338,
+      "step": 11140
+    },
+    {
+      "epoch": 1.9482841530054644,
+      "grad_norm": 0.8379867671028731,
+      "learning_rate": 5.781471712541036e-05,
+      "loss": 12.3092,
+      "step": 11141
+    },
+    {
+      "epoch": 1.9484590163934428,
+      "grad_norm": 0.9366189149544168,
+      "learning_rate": 5.779759924068331e-05,
+      "loss": 12.207,
+      "step": 11142
+    },
+    {
+      "epoch": 1.9486338797814207,
+      "grad_norm": 0.8071899544703058,
+      "learning_rate": 5.778048286042252e-05,
+      "loss": 12.1297,
+      "step": 11143
+    },
+    {
+      "epoch": 1.948808743169399,
+      "grad_norm": 0.9680488603879278,
+      "learning_rate": 5.7763367985238204e-05,
+      "loss": 12.3306,
+      "step": 11144
+    },
+    {
+      "epoch": 1.948983606557377,
+      "grad_norm": 0.9356155168158086,
+      "learning_rate": 5.774625461574048e-05,
+      "loss": 12.0401,
+      "step": 11145
+    },
+    {
+      "epoch": 1.9491584699453552,
+      "grad_norm": 0.7924082839980219,
+      "learning_rate": 5.772914275253943e-05,
+      "loss": 12.046,
+      "step": 11146
+    },
+    {
+      "epoch": 1.9493333333333334,
+      "grad_norm": 0.9004703672466366,
+      "learning_rate": 5.771203239624503e-05,
+      "loss": 12.2706,
+      "step": 11147
+    },
+    {
+      "epoch": 1.9495081967213115,
+      "grad_norm": 0.9181869378738174,
+      "learning_rate": 5.769492354746726e-05,
+      "loss": 12.1606,
+      "step": 11148
+    },
+    {
+      "epoch": 1.9496830601092896,
+      "grad_norm": 0.8846857248511301,
+      "learning_rate": 5.7677816206816026e-05,
+      "loss": 12.3545,
+      "step": 11149
+    },
+    {
+      "epoch": 1.9498579234972677,
+      "grad_norm": 0.8706235545476242,
+      "learning_rate": 5.7660710374901264e-05,
+      "loss": 12.4051,
+      "step": 11150
+    },
+    {
+      "epoch": 1.950032786885246,
+      "grad_norm": 0.9688161976244143,
+      "learning_rate": 5.7643606052332644e-05,
+      "loss": 12.4731,
+      "step": 11151
+    },
+    {
+      "epoch": 1.950207650273224,
+      "grad_norm": 0.8341385593318809,
+      "learning_rate": 5.762650323971999e-05,
+      "loss": 12.0449,
+      "step": 11152
+    },
+    {
+      "epoch": 1.9503825136612023,
+      "grad_norm": 0.891684892301519,
+      "learning_rate": 5.760940193767299e-05,
+      "loss": 12.1365,
+      "step": 11153
+    },
+    {
+      "epoch": 1.9505573770491802,
+      "grad_norm": 0.8513608190135402,
+      "learning_rate": 5.759230214680131e-05,
+      "loss": 12.3008,
+      "step": 11154
+    },
+    {
+      "epoch": 1.9507322404371585,
+      "grad_norm": 0.793095950206773,
+      "learning_rate": 5.7575203867714475e-05,
+      "loss": 12.0987,
+      "step": 11155
+    },
+    {
+      "epoch": 1.9509071038251367,
+      "grad_norm": 0.858097617720932,
+      "learning_rate": 5.755810710102206e-05,
+      "loss": 11.9752,
+      "step": 11156
+    },
+    {
+      "epoch": 1.9510819672131148,
+      "grad_norm": 0.7838399815803725,
+      "learning_rate": 5.7541011847333536e-05,
+      "loss": 12.2875,
+      "step": 11157
+    },
+    {
+      "epoch": 1.951256830601093,
+      "grad_norm": 0.814430041466282,
+      "learning_rate": 5.7523918107258335e-05,
+      "loss": 12.2206,
+      "step": 11158
+    },
+    {
+      "epoch": 1.951431693989071,
+      "grad_norm": 0.8784563350254526,
+      "learning_rate": 5.750682588140583e-05,
+      "loss": 12.2651,
+      "step": 11159
+    },
+    {
+      "epoch": 1.9516065573770491,
+      "grad_norm": 0.8350147452857877,
+      "learning_rate": 5.7489735170385314e-05,
+      "loss": 12.2166,
+      "step": 11160
+    },
+    {
+      "epoch": 1.9517814207650273,
+      "grad_norm": 0.8575214428832652,
+      "learning_rate": 5.747264597480609e-05,
+      "loss": 12.5102,
+      "step": 11161
+    },
+    {
+      "epoch": 1.9519562841530056,
+      "grad_norm": 0.7997033707273009,
+      "learning_rate": 5.745555829527738e-05,
+      "loss": 12.1158,
+      "step": 11162
+    },
+    {
+      "epoch": 1.9521311475409835,
+      "grad_norm": 0.8631455922303708,
+      "learning_rate": 5.743847213240827e-05,
+      "loss": 12.1769,
+      "step": 11163
+    },
+    {
+      "epoch": 1.9523060109289618,
+      "grad_norm": 0.8514941770805735,
+      "learning_rate": 5.742138748680791e-05,
+      "loss": 12.079,
+      "step": 11164
+    },
+    {
+      "epoch": 1.9524808743169397,
+      "grad_norm": 0.8482716682703184,
+      "learning_rate": 5.740430435908537e-05,
+      "loss": 12.1445,
+      "step": 11165
+    },
+    {
+      "epoch": 1.952655737704918,
+      "grad_norm": 0.865352207359588,
+      "learning_rate": 5.7387222749849576e-05,
+      "loss": 12.4015,
+      "step": 11166
+    },
+    {
+      "epoch": 1.9528306010928962,
+      "grad_norm": 0.7982624013753612,
+      "learning_rate": 5.73701426597095e-05,
+      "loss": 12.099,
+      "step": 11167
+    },
+    {
+      "epoch": 1.9530054644808743,
+      "grad_norm": 0.95385103697066,
+      "learning_rate": 5.735306408927403e-05,
+      "loss": 12.2347,
+      "step": 11168
+    },
+    {
+      "epoch": 1.9531803278688524,
+      "grad_norm": 0.8847328490705907,
+      "learning_rate": 5.733598703915204e-05,
+      "loss": 12.4405,
+      "step": 11169
+    },
+    {
+      "epoch": 1.9533551912568305,
+      "grad_norm": 0.8245309947868766,
+      "learning_rate": 5.731891150995222e-05,
+      "loss": 12.2326,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9535300546448089,
+      "grad_norm": 0.954534082200947,
+      "learning_rate": 5.730183750228335e-05,
+      "loss": 12.2224,
+      "step": 11171
+    },
+    {
+      "epoch": 1.9537049180327868,
+      "grad_norm": 0.9662213534711139,
+      "learning_rate": 5.728476501675407e-05,
+      "loss": 12.3383,
+      "step": 11172
+    },
+    {
+      "epoch": 1.9538797814207651,
+      "grad_norm": 0.8697585139268442,
+      "learning_rate": 5.726769405397305e-05,
+      "loss": 12.3938,
+      "step": 11173
+    },
+    {
+      "epoch": 1.954054644808743,
+      "grad_norm": 0.9056982678103023,
+      "learning_rate": 5.725062461454877e-05,
+      "loss": 12.0941,
+      "step": 11174
+    },
+    {
+      "epoch": 1.9542295081967214,
+      "grad_norm": 0.8907214489350211,
+      "learning_rate": 5.7233556699089786e-05,
+      "loss": 12.2799,
+      "step": 11175
+    },
+    {
+      "epoch": 1.9544043715846995,
+      "grad_norm": 0.8582222588514684,
+      "learning_rate": 5.721649030820454e-05,
+      "loss": 11.9254,
+      "step": 11176
+    },
+    {
+      "epoch": 1.9545792349726776,
+      "grad_norm": 0.9938700348394264,
+      "learning_rate": 5.719942544250142e-05,
+      "loss": 12.2594,
+      "step": 11177
+    },
+    {
+      "epoch": 1.9547540983606557,
+      "grad_norm": 0.9051890409037727,
+      "learning_rate": 5.718236210258878e-05,
+      "loss": 12.1681,
+      "step": 11178
+    },
+    {
+      "epoch": 1.9549289617486338,
+      "grad_norm": 0.8378217243476324,
+      "learning_rate": 5.7165300289074896e-05,
+      "loss": 12.2978,
+      "step": 11179
+    },
+    {
+      "epoch": 1.9551038251366122,
+      "grad_norm": 1.1308945790921212,
+      "learning_rate": 5.7148240002568054e-05,
+      "loss": 12.2855,
+      "step": 11180
+    },
+    {
+      "epoch": 1.95527868852459,
+      "grad_norm": 0.8360919454146528,
+      "learning_rate": 5.713118124367636e-05,
+      "loss": 11.7681,
+      "step": 11181
+    },
+    {
+      "epoch": 1.9554535519125684,
+      "grad_norm": 0.8996940005594299,
+      "learning_rate": 5.7114124013007954e-05,
+      "loss": 12.0697,
+      "step": 11182
+    },
+    {
+      "epoch": 1.9556284153005463,
+      "grad_norm": 0.9686205711492755,
+      "learning_rate": 5.709706831117092e-05,
+      "loss": 12.3275,
+      "step": 11183
+    },
+    {
+      "epoch": 1.9558032786885247,
+      "grad_norm": 0.8422090899053709,
+      "learning_rate": 5.7080014138773305e-05,
+      "loss": 12.3684,
+      "step": 11184
+    },
+    {
+      "epoch": 1.9559781420765028,
+      "grad_norm": 0.9198342179260941,
+      "learning_rate": 5.706296149642303e-05,
+      "loss": 12.3846,
+      "step": 11185
+    },
+    {
+      "epoch": 1.956153005464481,
+      "grad_norm": 0.9255659691710914,
+      "learning_rate": 5.7045910384727985e-05,
+      "loss": 12.335,
+      "step": 11186
+    },
+    {
+      "epoch": 1.956327868852459,
+      "grad_norm": 0.8238153677166291,
+      "learning_rate": 5.702886080429606e-05,
+      "loss": 11.7849,
+      "step": 11187
+    },
+    {
+      "epoch": 1.9565027322404371,
+      "grad_norm": 0.873348770406156,
+      "learning_rate": 5.7011812755735076e-05,
+      "loss": 12.4025,
+      "step": 11188
+    },
+    {
+      "epoch": 1.9566775956284153,
+      "grad_norm": 0.9217882156078671,
+      "learning_rate": 5.69947662396527e-05,
+      "loss": 12.2426,
+      "step": 11189
+    },
+    {
+      "epoch": 1.9568524590163934,
+      "grad_norm": 0.8257621102652903,
+      "learning_rate": 5.697772125665666e-05,
+      "loss": 12.2713,
+      "step": 11190
+    },
+    {
+      "epoch": 1.9570273224043717,
+      "grad_norm": 0.8407793523308688,
+      "learning_rate": 5.696067780735459e-05,
+      "loss": 12.1758,
+      "step": 11191
+    },
+    {
+      "epoch": 1.9572021857923496,
+      "grad_norm": 0.8057141302320048,
+      "learning_rate": 5.694363589235412e-05,
+      "loss": 12.1255,
+      "step": 11192
+    },
+    {
+      "epoch": 1.957377049180328,
+      "grad_norm": 0.8339179360316713,
+      "learning_rate": 5.692659551226267e-05,
+      "loss": 12.2183,
+      "step": 11193
+    },
+    {
+      "epoch": 1.9575519125683059,
+      "grad_norm": 0.8454542315408923,
+      "learning_rate": 5.690955666768778e-05,
+      "loss": 12.0487,
+      "step": 11194
+    },
+    {
+      "epoch": 1.9577267759562842,
+      "grad_norm": 0.8226699623758691,
+      "learning_rate": 5.689251935923684e-05,
+      "loss": 12.0617,
+      "step": 11195
+    },
+    {
+      "epoch": 1.9579016393442623,
+      "grad_norm": 0.8175140965656276,
+      "learning_rate": 5.687548358751722e-05,
+      "loss": 12.3724,
+      "step": 11196
+    },
+    {
+      "epoch": 1.9580765027322404,
+      "grad_norm": 0.7829632453499972,
+      "learning_rate": 5.685844935313622e-05,
+      "loss": 12.2047,
+      "step": 11197
+    },
+    {
+      "epoch": 1.9582513661202186,
+      "grad_norm": 0.9682527593815305,
+      "learning_rate": 5.6841416656701105e-05,
+      "loss": 12.3136,
+      "step": 11198
+    },
+    {
+      "epoch": 1.9584262295081967,
+      "grad_norm": 0.8263370834338384,
+      "learning_rate": 5.68243854988191e-05,
+      "loss": 12.3454,
+      "step": 11199
+    },
+    {
+      "epoch": 1.958601092896175,
+      "grad_norm": 0.9180965509065332,
+      "learning_rate": 5.680735588009727e-05,
+      "loss": 12.0516,
+      "step": 11200
+    },
+    {
+      "epoch": 1.958775956284153,
+      "grad_norm": 0.8711495128471796,
+      "learning_rate": 5.6790327801142726e-05,
+      "loss": 12.4088,
+      "step": 11201
+    },
+    {
+      "epoch": 1.9589508196721312,
+      "grad_norm": 0.8961448051101066,
+      "learning_rate": 5.677330126256252e-05,
+      "loss": 12.0683,
+      "step": 11202
+    },
+    {
+      "epoch": 1.9591256830601091,
+      "grad_norm": 0.9879148445123819,
+      "learning_rate": 5.675627626496365e-05,
+      "loss": 11.8914,
+      "step": 11203
+    },
+    {
+      "epoch": 1.9593005464480875,
+      "grad_norm": 0.8247422303979386,
+      "learning_rate": 5.673925280895297e-05,
+      "loss": 12.1689,
+      "step": 11204
+    },
+    {
+      "epoch": 1.9594754098360656,
+      "grad_norm": 0.8305157342008883,
+      "learning_rate": 5.6722230895137384e-05,
+      "loss": 12.078,
+      "step": 11205
+    },
+    {
+      "epoch": 1.9596502732240437,
+      "grad_norm": 0.8776186576113338,
+      "learning_rate": 5.67052105241237e-05,
+      "loss": 12.0962,
+      "step": 11206
+    },
+    {
+      "epoch": 1.9598251366120218,
+      "grad_norm": 0.8743207454621139,
+      "learning_rate": 5.668819169651871e-05,
+      "loss": 12.3515,
+      "step": 11207
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8242309217923381,
+      "learning_rate": 5.667117441292906e-05,
+      "loss": 12.2866,
+      "step": 11208
+    },
+    {
+      "epoch": 1.9601748633879783,
+      "grad_norm": 0.8677967342318483,
+      "learning_rate": 5.66541586739614e-05,
+      "loss": 12.3244,
+      "step": 11209
+    },
+    {
+      "epoch": 1.9603497267759562,
+      "grad_norm": 0.9162076008754557,
+      "learning_rate": 5.6637144480222396e-05,
+      "loss": 12.48,
+      "step": 11210
+    },
+    {
+      "epoch": 1.9605245901639345,
+      "grad_norm": 0.7943273635203332,
+      "learning_rate": 5.662013183231847e-05,
+      "loss": 12.1466,
+      "step": 11211
+    },
+    {
+      "epoch": 1.9606994535519124,
+      "grad_norm": 0.8875878803256271,
+      "learning_rate": 5.660312073085617e-05,
+      "loss": 12.0162,
+      "step": 11212
+    },
+    {
+      "epoch": 1.9608743169398908,
+      "grad_norm": 0.8535100412898243,
+      "learning_rate": 5.658611117644192e-05,
+      "loss": 12.3476,
+      "step": 11213
+    },
+    {
+      "epoch": 1.961049180327869,
+      "grad_norm": 0.8486997681151002,
+      "learning_rate": 5.656910316968207e-05,
+      "loss": 12.2247,
+      "step": 11214
+    },
+    {
+      "epoch": 1.961224043715847,
+      "grad_norm": 0.9083211675850384,
+      "learning_rate": 5.655209671118294e-05,
+      "loss": 12.2237,
+      "step": 11215
+    },
+    {
+      "epoch": 1.9613989071038251,
+      "grad_norm": 0.9134036149849396,
+      "learning_rate": 5.653509180155081e-05,
+      "loss": 12.3546,
+      "step": 11216
+    },
+    {
+      "epoch": 1.9615737704918033,
+      "grad_norm": 0.9780723719447683,
+      "learning_rate": 5.651808844139188e-05,
+      "loss": 12.3501,
+      "step": 11217
+    },
+    {
+      "epoch": 1.9617486338797814,
+      "grad_norm": 0.7985032688621129,
+      "learning_rate": 5.6501086631312326e-05,
+      "loss": 12.225,
+      "step": 11218
+    },
+    {
+      "epoch": 1.9619234972677595,
+      "grad_norm": 0.8736398959617375,
+      "learning_rate": 5.648408637191818e-05,
+      "loss": 12.315,
+      "step": 11219
+    },
+    {
+      "epoch": 1.9620983606557378,
+      "grad_norm": 0.9037560038847615,
+      "learning_rate": 5.64670876638155e-05,
+      "loss": 12.2242,
+      "step": 11220
+    },
+    {
+      "epoch": 1.9622732240437157,
+      "grad_norm": 0.8790611228609938,
+      "learning_rate": 5.645009050761029e-05,
+      "loss": 12.2088,
+      "step": 11221
+    },
+    {
+      "epoch": 1.962448087431694,
+      "grad_norm": 0.9012414893237451,
+      "learning_rate": 5.64330949039085e-05,
+      "loss": 12.2071,
+      "step": 11222
+    },
+    {
+      "epoch": 1.962622950819672,
+      "grad_norm": 0.8924743692251643,
+      "learning_rate": 5.641610085331594e-05,
+      "loss": 12.198,
+      "step": 11223
+    },
+    {
+      "epoch": 1.9627978142076503,
+      "grad_norm": 0.8497105657071982,
+      "learning_rate": 5.6399108356438466e-05,
+      "loss": 12.1303,
+      "step": 11224
+    },
+    {
+      "epoch": 1.9629726775956284,
+      "grad_norm": 0.9590389783595135,
+      "learning_rate": 5.6382117413881884e-05,
+      "loss": 12.3703,
+      "step": 11225
+    },
+    {
+      "epoch": 1.9631475409836066,
+      "grad_norm": 0.9834176528408225,
+      "learning_rate": 5.6365128026251804e-05,
+      "loss": 12.5454,
+      "step": 11226
+    },
+    {
+      "epoch": 1.9633224043715847,
+      "grad_norm": 0.8746381831036794,
+      "learning_rate": 5.634814019415393e-05,
+      "loss": 12.4042,
+      "step": 11227
+    },
+    {
+      "epoch": 1.9634972677595628,
+      "grad_norm": 0.8485076461023081,
+      "learning_rate": 5.6331153918193844e-05,
+      "loss": 12.1675,
+      "step": 11228
+    },
+    {
+      "epoch": 1.9636721311475411,
+      "grad_norm": 0.8611873678540886,
+      "learning_rate": 5.6314169198977166e-05,
+      "loss": 12.4073,
+      "step": 11229
+    },
+    {
+      "epoch": 1.963846994535519,
+      "grad_norm": 0.8360844024316305,
+      "learning_rate": 5.6297186037109254e-05,
+      "loss": 12.0092,
+      "step": 11230
+    },
+    {
+      "epoch": 1.9640218579234974,
+      "grad_norm": 0.8481052995593172,
+      "learning_rate": 5.6280204433195595e-05,
+      "loss": 12.2228,
+      "step": 11231
+    },
+    {
+      "epoch": 1.9641967213114753,
+      "grad_norm": 0.8757814931712266,
+      "learning_rate": 5.626322438784157e-05,
+      "loss": 12.1542,
+      "step": 11232
+    },
+    {
+      "epoch": 1.9643715846994536,
+      "grad_norm": 0.9137917513990141,
+      "learning_rate": 5.6246245901652505e-05,
+      "loss": 12.0111,
+      "step": 11233
+    },
+    {
+      "epoch": 1.9645464480874317,
+      "grad_norm": 0.878774884991079,
+      "learning_rate": 5.6229268975233653e-05,
+      "loss": 12.439,
+      "step": 11234
+    },
+    {
+      "epoch": 1.9647213114754098,
+      "grad_norm": 0.8168065636730765,
+      "learning_rate": 5.6212293609190214e-05,
+      "loss": 12.1924,
+      "step": 11235
+    },
+    {
+      "epoch": 1.964896174863388,
+      "grad_norm": 0.9354733135157628,
+      "learning_rate": 5.619531980412735e-05,
+      "loss": 12.3422,
+      "step": 11236
+    },
+    {
+      "epoch": 1.965071038251366,
+      "grad_norm": 0.9109170544882589,
+      "learning_rate": 5.617834756065019e-05,
+      "loss": 11.7652,
+      "step": 11237
+    },
+    {
+      "epoch": 1.9652459016393444,
+      "grad_norm": 0.9352847488936943,
+      "learning_rate": 5.6161376879363715e-05,
+      "loss": 12.2395,
+      "step": 11238
+    },
+    {
+      "epoch": 1.9654207650273223,
+      "grad_norm": 0.8707840455218763,
+      "learning_rate": 5.614440776087292e-05,
+      "loss": 12.3957,
+      "step": 11239
+    },
+    {
+      "epoch": 1.9655956284153007,
+      "grad_norm": 0.8481439791172248,
+      "learning_rate": 5.612744020578279e-05,
+      "loss": 12.1911,
+      "step": 11240
+    },
+    {
+      "epoch": 1.9657704918032786,
+      "grad_norm": 0.9076059582121181,
+      "learning_rate": 5.611047421469812e-05,
+      "loss": 12.2899,
+      "step": 11241
+    },
+    {
+      "epoch": 1.965945355191257,
+      "grad_norm": 0.8826956896622233,
+      "learning_rate": 5.609350978822377e-05,
+      "loss": 12.2933,
+      "step": 11242
+    },
+    {
+      "epoch": 1.966120218579235,
+      "grad_norm": 0.8326489469943024,
+      "learning_rate": 5.607654692696449e-05,
+      "loss": 12.2096,
+      "step": 11243
+    },
+    {
+      "epoch": 1.9662950819672131,
+      "grad_norm": 0.9016830604500502,
+      "learning_rate": 5.6059585631525024e-05,
+      "loss": 12.4979,
+      "step": 11244
+    },
+    {
+      "epoch": 1.9664699453551913,
+      "grad_norm": 0.8429959668426574,
+      "learning_rate": 5.604262590250997e-05,
+      "loss": 12.1215,
+      "step": 11245
+    },
+    {
+      "epoch": 1.9666448087431694,
+      "grad_norm": 0.9060558863288589,
+      "learning_rate": 5.602566774052393e-05,
+      "loss": 12.0984,
+      "step": 11246
+    },
+    {
+      "epoch": 1.9668196721311475,
+      "grad_norm": 0.8080461112491408,
+      "learning_rate": 5.6008711146171454e-05,
+      "loss": 12.2109,
+      "step": 11247
+    },
+    {
+      "epoch": 1.9669945355191256,
+      "grad_norm": 0.8289926132400716,
+      "learning_rate": 5.5991756120057035e-05,
+      "loss": 12.3054,
+      "step": 11248
+    },
+    {
+      "epoch": 1.967169398907104,
+      "grad_norm": 0.8273629499412067,
+      "learning_rate": 5.59748026627851e-05,
+      "loss": 12.2843,
+      "step": 11249
+    },
+    {
+      "epoch": 1.9673442622950819,
+      "grad_norm": 0.8560728521223296,
+      "learning_rate": 5.595785077496001e-05,
+      "loss": 12.3719,
+      "step": 11250
+    },
+    {
+      "epoch": 1.9675191256830602,
+      "grad_norm": 0.8923282991258019,
+      "learning_rate": 5.594090045718607e-05,
+      "loss": 12.3311,
+      "step": 11251
+    },
+    {
+      "epoch": 1.967693989071038,
+      "grad_norm": 0.88506743972148,
+      "learning_rate": 5.592395171006759e-05,
+      "loss": 12.2525,
+      "step": 11252
+    },
+    {
+      "epoch": 1.9678688524590164,
+      "grad_norm": 0.8203579792231537,
+      "learning_rate": 5.590700453420871e-05,
+      "loss": 12.1039,
+      "step": 11253
+    },
+    {
+      "epoch": 1.9680437158469946,
+      "grad_norm": 0.8656057786517357,
+      "learning_rate": 5.5890058930213596e-05,
+      "loss": 12.4048,
+      "step": 11254
+    },
+    {
+      "epoch": 1.9682185792349727,
+      "grad_norm": 0.8615736891320792,
+      "learning_rate": 5.5873114898686375e-05,
+      "loss": 12.2049,
+      "step": 11255
+    },
+    {
+      "epoch": 1.9683934426229508,
+      "grad_norm": 0.8388853365831425,
+      "learning_rate": 5.5856172440231014e-05,
+      "loss": 12.1681,
+      "step": 11256
+    },
+    {
+      "epoch": 1.968568306010929,
+      "grad_norm": 0.843489810152766,
+      "learning_rate": 5.5839231555451535e-05,
+      "loss": 12.2807,
+      "step": 11257
+    },
+    {
+      "epoch": 1.9687431693989073,
+      "grad_norm": 0.8138613411401638,
+      "learning_rate": 5.582229224495186e-05,
+      "loss": 12.229,
+      "step": 11258
+    },
+    {
+      "epoch": 1.9689180327868852,
+      "grad_norm": 0.8225847586881782,
+      "learning_rate": 5.580535450933587e-05,
+      "loss": 12.2776,
+      "step": 11259
+    },
+    {
+      "epoch": 1.9690928961748635,
+      "grad_norm": 0.815860884086137,
+      "learning_rate": 5.5788418349207336e-05,
+      "loss": 12.1811,
+      "step": 11260
+    },
+    {
+      "epoch": 1.9692677595628414,
+      "grad_norm": 0.8410228538155001,
+      "learning_rate": 5.577148376517002e-05,
+      "loss": 12.3578,
+      "step": 11261
+    },
+    {
+      "epoch": 1.9694426229508197,
+      "grad_norm": 0.8517087952227382,
+      "learning_rate": 5.575455075782763e-05,
+      "loss": 12.1986,
+      "step": 11262
+    },
+    {
+      "epoch": 1.9696174863387979,
+      "grad_norm": 0.8634391660408379,
+      "learning_rate": 5.5737619327783866e-05,
+      "loss": 12.0621,
+      "step": 11263
+    },
+    {
+      "epoch": 1.969792349726776,
+      "grad_norm": 0.932349286150795,
+      "learning_rate": 5.572068947564221e-05,
+      "loss": 11.9143,
+      "step": 11264
+    },
+    {
+      "epoch": 1.969967213114754,
+      "grad_norm": 0.8258717666958844,
+      "learning_rate": 5.570376120200623e-05,
+      "loss": 12.0633,
+      "step": 11265
+    },
+    {
+      "epoch": 1.9701420765027322,
+      "grad_norm": 0.8716591876457411,
+      "learning_rate": 5.5686834507479424e-05,
+      "loss": 12.3568,
+      "step": 11266
+    },
+    {
+      "epoch": 1.9703169398907103,
+      "grad_norm": 0.8740350324297245,
+      "learning_rate": 5.566990939266519e-05,
+      "loss": 12.2349,
+      "step": 11267
+    },
+    {
+      "epoch": 1.9704918032786884,
+      "grad_norm": 0.8120852296938585,
+      "learning_rate": 5.565298585816687e-05,
+      "loss": 12.3569,
+      "step": 11268
+    },
+    {
+      "epoch": 1.9706666666666668,
+      "grad_norm": 0.8987900716102915,
+      "learning_rate": 5.563606390458782e-05,
+      "loss": 12.5033,
+      "step": 11269
+    },
+    {
+      "epoch": 1.9708415300546447,
+      "grad_norm": 0.8568125674636936,
+      "learning_rate": 5.5619143532531235e-05,
+      "loss": 12.2304,
+      "step": 11270
+    },
+    {
+      "epoch": 1.971016393442623,
+      "grad_norm": 0.8883772117420818,
+      "learning_rate": 5.560222474260039e-05,
+      "loss": 12.3144,
+      "step": 11271
+    },
+    {
+      "epoch": 1.9711912568306011,
+      "grad_norm": 0.8836690239825735,
+      "learning_rate": 5.5585307535398315e-05,
+      "loss": 12.1785,
+      "step": 11272
+    },
+    {
+      "epoch": 1.9713661202185793,
+      "grad_norm": 0.9340062547945843,
+      "learning_rate": 5.5568391911528126e-05,
+      "loss": 12.1662,
+      "step": 11273
+    },
+    {
+      "epoch": 1.9715409836065574,
+      "grad_norm": 0.8264199576098263,
+      "learning_rate": 5.555147787159289e-05,
+      "loss": 12.36,
+      "step": 11274
+    },
+    {
+      "epoch": 1.9717158469945355,
+      "grad_norm": 0.9563283165668096,
+      "learning_rate": 5.5534565416195506e-05,
+      "loss": 12.1146,
+      "step": 11275
+    },
+    {
+      "epoch": 1.9718907103825136,
+      "grad_norm": 0.9126666628533331,
+      "learning_rate": 5.551765454593891e-05,
+      "loss": 12.3868,
+      "step": 11276
+    },
+    {
+      "epoch": 1.9720655737704917,
+      "grad_norm": 0.8759887793122597,
+      "learning_rate": 5.550074526142594e-05,
+      "loss": 12.6006,
+      "step": 11277
+    },
+    {
+      "epoch": 1.97224043715847,
+      "grad_norm": 1.1945915253117978,
+      "learning_rate": 5.5483837563259475e-05,
+      "loss": 12.181,
+      "step": 11278
+    },
+    {
+      "epoch": 1.972415300546448,
+      "grad_norm": 0.9511080148365217,
+      "learning_rate": 5.546693145204214e-05,
+      "loss": 12.3104,
+      "step": 11279
+    },
+    {
+      "epoch": 1.9725901639344263,
+      "grad_norm": 0.8815071359443518,
+      "learning_rate": 5.545002692837666e-05,
+      "loss": 11.8942,
+      "step": 11280
+    },
+    {
+      "epoch": 1.9727650273224042,
+      "grad_norm": 1.0400035648053538,
+      "learning_rate": 5.543312399286569e-05,
+      "loss": 12.5102,
+      "step": 11281
+    },
+    {
+      "epoch": 1.9729398907103826,
+      "grad_norm": 0.977304907830253,
+      "learning_rate": 5.541622264611179e-05,
+      "loss": 12.2132,
+      "step": 11282
+    },
+    {
+      "epoch": 1.9731147540983607,
+      "grad_norm": 0.8945031093372915,
+      "learning_rate": 5.539932288871744e-05,
+      "loss": 12.2954,
+      "step": 11283
+    },
+    {
+      "epoch": 1.9732896174863388,
+      "grad_norm": 0.988134218225382,
+      "learning_rate": 5.538242472128512e-05,
+      "loss": 12.3439,
+      "step": 11284
+    },
+    {
+      "epoch": 1.973464480874317,
+      "grad_norm": 0.8580908996308882,
+      "learning_rate": 5.536552814441722e-05,
+      "loss": 12.0976,
+      "step": 11285
+    },
+    {
+      "epoch": 1.973639344262295,
+      "grad_norm": 0.8849276835525739,
+      "learning_rate": 5.53486331587161e-05,
+      "loss": 12.2678,
+      "step": 11286
+    },
+    {
+      "epoch": 1.9738142076502734,
+      "grad_norm": 0.8089646025299159,
+      "learning_rate": 5.533173976478403e-05,
+      "loss": 12.2156,
+      "step": 11287
+    },
+    {
+      "epoch": 1.9739890710382513,
+      "grad_norm": 0.8351154785319297,
+      "learning_rate": 5.531484796322326e-05,
+      "loss": 12.1961,
+      "step": 11288
+    },
+    {
+      "epoch": 1.9741639344262296,
+      "grad_norm": 0.8544330956390268,
+      "learning_rate": 5.529795775463598e-05,
+      "loss": 12.088,
+      "step": 11289
+    },
+    {
+      "epoch": 1.9743387978142075,
+      "grad_norm": 0.846373367010418,
+      "learning_rate": 5.528106913962423e-05,
+      "loss": 12.2915,
+      "step": 11290
+    },
+    {
+      "epoch": 1.9745136612021859,
+      "grad_norm": 0.9276902398757745,
+      "learning_rate": 5.526418211879014e-05,
+      "loss": 11.9427,
+      "step": 11291
+    },
+    {
+      "epoch": 1.974688524590164,
+      "grad_norm": 0.9434594122430858,
+      "learning_rate": 5.5247296692735675e-05,
+      "loss": 12.4425,
+      "step": 11292
+    },
+    {
+      "epoch": 1.974863387978142,
+      "grad_norm": 0.9375033930182723,
+      "learning_rate": 5.523041286206283e-05,
+      "loss": 12.0712,
+      "step": 11293
+    },
+    {
+      "epoch": 1.9750382513661202,
+      "grad_norm": 0.8625214977679031,
+      "learning_rate": 5.521353062737342e-05,
+      "loss": 12.3412,
+      "step": 11294
+    },
+    {
+      "epoch": 1.9752131147540983,
+      "grad_norm": 0.8806110013539836,
+      "learning_rate": 5.519664998926932e-05,
+      "loss": 12.085,
+      "step": 11295
+    },
+    {
+      "epoch": 1.9753879781420765,
+      "grad_norm": 0.9901919083024505,
+      "learning_rate": 5.517977094835232e-05,
+      "loss": 12.2767,
+      "step": 11296
+    },
+    {
+      "epoch": 1.9755628415300546,
+      "grad_norm": 0.9150907649524236,
+      "learning_rate": 5.516289350522413e-05,
+      "loss": 12.4956,
+      "step": 11297
+    },
+    {
+      "epoch": 1.975737704918033,
+      "grad_norm": 0.9480141074897204,
+      "learning_rate": 5.514601766048637e-05,
+      "loss": 11.9807,
+      "step": 11298
+    },
+    {
+      "epoch": 1.9759125683060108,
+      "grad_norm": 0.8639951736005552,
+      "learning_rate": 5.512914341474067e-05,
+      "loss": 12.2577,
+      "step": 11299
+    },
+    {
+      "epoch": 1.9760874316939891,
+      "grad_norm": 1.0485521526229407,
+      "learning_rate": 5.511227076858859e-05,
+      "loss": 12.1394,
+      "step": 11300
+    },
+    {
+      "epoch": 1.9762622950819673,
+      "grad_norm": 0.7792924880838396,
+      "learning_rate": 5.509539972263165e-05,
+      "loss": 12.3657,
+      "step": 11301
+    },
+    {
+      "epoch": 1.9764371584699454,
+      "grad_norm": 0.9523504382440009,
+      "learning_rate": 5.507853027747121e-05,
+      "loss": 12.2846,
+      "step": 11302
+    },
+    {
+      "epoch": 1.9766120218579235,
+      "grad_norm": 0.8333906524003072,
+      "learning_rate": 5.506166243370867e-05,
+      "loss": 12.1713,
+      "step": 11303
+    },
+    {
+      "epoch": 1.9767868852459016,
+      "grad_norm": 0.8524504379167062,
+      "learning_rate": 5.504479619194536e-05,
+      "loss": 12.3873,
+      "step": 11304
+    },
+    {
+      "epoch": 1.9769617486338797,
+      "grad_norm": 0.9209810449507406,
+      "learning_rate": 5.502793155278254e-05,
+      "loss": 12.2588,
+      "step": 11305
+    },
+    {
+      "epoch": 1.9771366120218579,
+      "grad_norm": 0.8228035953034171,
+      "learning_rate": 5.501106851682142e-05,
+      "loss": 12.4178,
+      "step": 11306
+    },
+    {
+      "epoch": 1.9773114754098362,
+      "grad_norm": 0.8551185434098869,
+      "learning_rate": 5.499420708466314e-05,
+      "loss": 12.0924,
+      "step": 11307
+    },
+    {
+      "epoch": 1.977486338797814,
+      "grad_norm": 0.860051814316108,
+      "learning_rate": 5.497734725690884e-05,
+      "loss": 12.2584,
+      "step": 11308
+    },
+    {
+      "epoch": 1.9776612021857924,
+      "grad_norm": 0.8175369425673785,
+      "learning_rate": 5.496048903415947e-05,
+      "loss": 12.2674,
+      "step": 11309
+    },
+    {
+      "epoch": 1.9778360655737703,
+      "grad_norm": 0.8527881674266078,
+      "learning_rate": 5.494363241701603e-05,
+      "loss": 12.1309,
+      "step": 11310
+    },
+    {
+      "epoch": 1.9780109289617487,
+      "grad_norm": 0.877090185777507,
+      "learning_rate": 5.4926777406079454e-05,
+      "loss": 12.0228,
+      "step": 11311
+    },
+    {
+      "epoch": 1.9781857923497268,
+      "grad_norm": 0.8686496822620678,
+      "learning_rate": 5.490992400195065e-05,
+      "loss": 12.1914,
+      "step": 11312
+    },
+    {
+      "epoch": 1.978360655737705,
+      "grad_norm": 0.8738341303010867,
+      "learning_rate": 5.4893072205230325e-05,
+      "loss": 12.2802,
+      "step": 11313
+    },
+    {
+      "epoch": 1.978535519125683,
+      "grad_norm": 0.8378678829383389,
+      "learning_rate": 5.487622201651927e-05,
+      "loss": 12.1175,
+      "step": 11314
+    },
+    {
+      "epoch": 1.9787103825136612,
+      "grad_norm": 0.8190159229381305,
+      "learning_rate": 5.485937343641819e-05,
+      "loss": 12.3759,
+      "step": 11315
+    },
+    {
+      "epoch": 1.9788852459016395,
+      "grad_norm": 0.9669900361022432,
+      "learning_rate": 5.484252646552774e-05,
+      "loss": 12.1593,
+      "step": 11316
+    },
+    {
+      "epoch": 1.9790601092896174,
+      "grad_norm": 0.8640306169484377,
+      "learning_rate": 5.482568110444843e-05,
+      "loss": 11.9667,
+      "step": 11317
+    },
+    {
+      "epoch": 1.9792349726775957,
+      "grad_norm": 0.8165558270136228,
+      "learning_rate": 5.48088373537808e-05,
+      "loss": 12.3505,
+      "step": 11318
+    },
+    {
+      "epoch": 1.9794098360655736,
+      "grad_norm": 0.951765814117522,
+      "learning_rate": 5.4791995214125355e-05,
+      "loss": 12.3549,
+      "step": 11319
+    },
+    {
+      "epoch": 1.979584699453552,
+      "grad_norm": 0.9319934481884067,
+      "learning_rate": 5.477515468608243e-05,
+      "loss": 12.488,
+      "step": 11320
+    },
+    {
+      "epoch": 1.97975956284153,
+      "grad_norm": 0.9135944142356557,
+      "learning_rate": 5.475831577025239e-05,
+      "loss": 12.3323,
+      "step": 11321
+    },
+    {
+      "epoch": 1.9799344262295082,
+      "grad_norm": 0.8726965956826739,
+      "learning_rate": 5.474147846723554e-05,
+      "loss": 12.4035,
+      "step": 11322
+    },
+    {
+      "epoch": 1.9801092896174863,
+      "grad_norm": 0.8308488461215615,
+      "learning_rate": 5.4724642777632105e-05,
+      "loss": 12.1794,
+      "step": 11323
+    },
+    {
+      "epoch": 1.9802841530054645,
+      "grad_norm": 0.8520499201393715,
+      "learning_rate": 5.470780870204225e-05,
+      "loss": 11.9778,
+      "step": 11324
+    },
+    {
+      "epoch": 1.9804590163934426,
+      "grad_norm": 0.998267385495656,
+      "learning_rate": 5.469097624106611e-05,
+      "loss": 12.3087,
+      "step": 11325
+    },
+    {
+      "epoch": 1.9806338797814207,
+      "grad_norm": 0.8310462910948193,
+      "learning_rate": 5.467414539530371e-05,
+      "loss": 12.2631,
+      "step": 11326
+    },
+    {
+      "epoch": 1.980808743169399,
+      "grad_norm": 0.8985748905689394,
+      "learning_rate": 5.465731616535512e-05,
+      "loss": 12.2408,
+      "step": 11327
+    },
+    {
+      "epoch": 1.980983606557377,
+      "grad_norm": 0.8660057857274777,
+      "learning_rate": 5.464048855182019e-05,
+      "loss": 12.1938,
+      "step": 11328
+    },
+    {
+      "epoch": 1.9811584699453553,
+      "grad_norm": 0.8146395866101184,
+      "learning_rate": 5.462366255529885e-05,
+      "loss": 12.3391,
+      "step": 11329
+    },
+    {
+      "epoch": 1.9813333333333332,
+      "grad_norm": 0.8314263624857369,
+      "learning_rate": 5.4606838176390916e-05,
+      "loss": 12.2278,
+      "step": 11330
+    },
+    {
+      "epoch": 1.9815081967213115,
+      "grad_norm": 0.8692131480798383,
+      "learning_rate": 5.459001541569622e-05,
+      "loss": 12.3119,
+      "step": 11331
+    },
+    {
+      "epoch": 1.9816830601092896,
+      "grad_norm": 0.8942928785934642,
+      "learning_rate": 5.4573194273814356e-05,
+      "loss": 12.3211,
+      "step": 11332
+    },
+    {
+      "epoch": 1.9818579234972677,
+      "grad_norm": 0.9102499894925764,
+      "learning_rate": 5.455637475134505e-05,
+      "loss": 12.2457,
+      "step": 11333
+    },
+    {
+      "epoch": 1.9820327868852459,
+      "grad_norm": 0.8011414451062313,
+      "learning_rate": 5.453955684888794e-05,
+      "loss": 12.2388,
+      "step": 11334
+    },
+    {
+      "epoch": 1.982207650273224,
+      "grad_norm": 0.8638839147018844,
+      "learning_rate": 5.4522740567042454e-05,
+      "loss": 12.1734,
+      "step": 11335
+    },
+    {
+      "epoch": 1.9823825136612023,
+      "grad_norm": 0.8272704975816485,
+      "learning_rate": 5.450592590640815e-05,
+      "loss": 12.2805,
+      "step": 11336
+    },
+    {
+      "epoch": 1.9825573770491802,
+      "grad_norm": 0.8173583611585263,
+      "learning_rate": 5.4489112867584426e-05,
+      "loss": 12.4163,
+      "step": 11337
+    },
+    {
+      "epoch": 1.9827322404371586,
+      "grad_norm": 0.8352362437665914,
+      "learning_rate": 5.447230145117066e-05,
+      "loss": 12.3475,
+      "step": 11338
+    },
+    {
+      "epoch": 1.9829071038251365,
+      "grad_norm": 0.8591285524102215,
+      "learning_rate": 5.4455491657766135e-05,
+      "loss": 12.4095,
+      "step": 11339
+    },
+    {
+      "epoch": 1.9830819672131148,
+      "grad_norm": 0.8147169122105877,
+      "learning_rate": 5.443868348797014e-05,
+      "loss": 12.2582,
+      "step": 11340
+    },
+    {
+      "epoch": 1.983256830601093,
+      "grad_norm": 0.8326204781369309,
+      "learning_rate": 5.4421876942381836e-05,
+      "loss": 12.2589,
+      "step": 11341
+    },
+    {
+      "epoch": 1.983431693989071,
+      "grad_norm": 0.8439395376486967,
+      "learning_rate": 5.440507202160041e-05,
+      "loss": 12.319,
+      "step": 11342
+    },
+    {
+      "epoch": 1.9836065573770492,
+      "grad_norm": 0.8642564387270083,
+      "learning_rate": 5.438826872622485e-05,
+      "loss": 12.1326,
+      "step": 11343
+    },
+    {
+      "epoch": 1.9837814207650273,
+      "grad_norm": 0.8793202200446176,
+      "learning_rate": 5.437146705685423e-05,
+      "loss": 12.1135,
+      "step": 11344
+    },
+    {
+      "epoch": 1.9839562841530056,
+      "grad_norm": 0.8368003325575047,
+      "learning_rate": 5.435466701408749e-05,
+      "loss": 12.2517,
+      "step": 11345
+    },
+    {
+      "epoch": 1.9841311475409835,
+      "grad_norm": 0.9137572199843982,
+      "learning_rate": 5.433786859852357e-05,
+      "loss": 12.3287,
+      "step": 11346
+    },
+    {
+      "epoch": 1.9843060109289619,
+      "grad_norm": 0.8535376235884267,
+      "learning_rate": 5.432107181076126e-05,
+      "loss": 12.2415,
+      "step": 11347
+    },
+    {
+      "epoch": 1.9844808743169398,
+      "grad_norm": 0.856006319953132,
+      "learning_rate": 5.430427665139937e-05,
+      "loss": 12.3469,
+      "step": 11348
+    },
+    {
+      "epoch": 1.984655737704918,
+      "grad_norm": 0.8970087957174441,
+      "learning_rate": 5.4287483121036656e-05,
+      "loss": 12.3967,
+      "step": 11349
+    },
+    {
+      "epoch": 1.9848306010928962,
+      "grad_norm": 0.8644787677582932,
+      "learning_rate": 5.427069122027173e-05,
+      "loss": 12.3134,
+      "step": 11350
+    },
+    {
+      "epoch": 1.9850054644808743,
+      "grad_norm": 0.8500892607200596,
+      "learning_rate": 5.425390094970322e-05,
+      "loss": 12.3853,
+      "step": 11351
+    },
+    {
+      "epoch": 1.9851803278688525,
+      "grad_norm": 0.8745003517491262,
+      "learning_rate": 5.42371123099297e-05,
+      "loss": 12.2081,
+      "step": 11352
+    },
+    {
+      "epoch": 1.9853551912568306,
+      "grad_norm": 0.819353222144728,
+      "learning_rate": 5.42203253015497e-05,
+      "loss": 12.1944,
+      "step": 11353
+    },
+    {
+      "epoch": 1.9855300546448087,
+      "grad_norm": 0.8682814230314225,
+      "learning_rate": 5.420353992516158e-05,
+      "loss": 12.2664,
+      "step": 11354
+    },
+    {
+      "epoch": 1.9857049180327868,
+      "grad_norm": 0.8477198639822803,
+      "learning_rate": 5.4186756181363754e-05,
+      "loss": 12.006,
+      "step": 11355
+    },
+    {
+      "epoch": 1.9858797814207652,
+      "grad_norm": 0.7982008370108242,
+      "learning_rate": 5.4169974070754526e-05,
+      "loss": 11.9644,
+      "step": 11356
+    },
+    {
+      "epoch": 1.986054644808743,
+      "grad_norm": 0.8278482499994546,
+      "learning_rate": 5.415319359393218e-05,
+      "loss": 12.0973,
+      "step": 11357
+    },
+    {
+      "epoch": 1.9862295081967214,
+      "grad_norm": 0.8412027503883197,
+      "learning_rate": 5.413641475149493e-05,
+      "loss": 12.106,
+      "step": 11358
+    },
+    {
+      "epoch": 1.9864043715846993,
+      "grad_norm": 0.7863687949161491,
+      "learning_rate": 5.4119637544040904e-05,
+      "loss": 12.3384,
+      "step": 11359
+    },
+    {
+      "epoch": 1.9865792349726776,
+      "grad_norm": 0.8367441628785277,
+      "learning_rate": 5.410286197216818e-05,
+      "loss": 12.3552,
+      "step": 11360
+    },
+    {
+      "epoch": 1.9867540983606558,
+      "grad_norm": 0.9016132405646389,
+      "learning_rate": 5.4086088036474856e-05,
+      "loss": 12.1961,
+      "step": 11361
+    },
+    {
+      "epoch": 1.9869289617486339,
+      "grad_norm": 0.8829468807263899,
+      "learning_rate": 5.406931573755881e-05,
+      "loss": 12.147,
+      "step": 11362
+    },
+    {
+      "epoch": 1.987103825136612,
+      "grad_norm": 0.87815264235494,
+      "learning_rate": 5.4052545076018e-05,
+      "loss": 12.4008,
+      "step": 11363
+    },
+    {
+      "epoch": 1.98727868852459,
+      "grad_norm": 0.9051908491121355,
+      "learning_rate": 5.4035776052450296e-05,
+      "loss": 12.459,
+      "step": 11364
+    },
+    {
+      "epoch": 1.9874535519125685,
+      "grad_norm": 0.8299440660783297,
+      "learning_rate": 5.401900866745344e-05,
+      "loss": 12.3871,
+      "step": 11365
+    },
+    {
+      "epoch": 1.9876284153005463,
+      "grad_norm": 0.9278550550182606,
+      "learning_rate": 5.4002242921625214e-05,
+      "loss": 12.3522,
+      "step": 11366
+    },
+    {
+      "epoch": 1.9878032786885247,
+      "grad_norm": 0.8513922486268201,
+      "learning_rate": 5.398547881556326e-05,
+      "loss": 11.9079,
+      "step": 11367
+    },
+    {
+      "epoch": 1.9879781420765026,
+      "grad_norm": 0.843288676325621,
+      "learning_rate": 5.3968716349865286e-05,
+      "loss": 12.423,
+      "step": 11368
+    },
+    {
+      "epoch": 1.988153005464481,
+      "grad_norm": 0.8478770883905097,
+      "learning_rate": 5.395195552512874e-05,
+      "loss": 12.2179,
+      "step": 11369
+    },
+    {
+      "epoch": 1.988327868852459,
+      "grad_norm": 0.8071801260334267,
+      "learning_rate": 5.3935196341951186e-05,
+      "loss": 12.2515,
+      "step": 11370
+    },
+    {
+      "epoch": 1.9885027322404372,
+      "grad_norm": 0.8682461845644124,
+      "learning_rate": 5.391843880093006e-05,
+      "loss": 12.3226,
+      "step": 11371
+    },
+    {
+      "epoch": 1.9886775956284153,
+      "grad_norm": 0.7995179955155082,
+      "learning_rate": 5.3901682902662775e-05,
+      "loss": 12.3883,
+      "step": 11372
+    },
+    {
+      "epoch": 1.9888524590163934,
+      "grad_norm": 0.8127249758724902,
+      "learning_rate": 5.388492864774661e-05,
+      "loss": 12.3603,
+      "step": 11373
+    },
+    {
+      "epoch": 1.9890273224043717,
+      "grad_norm": 0.8091776578300516,
+      "learning_rate": 5.386817603677886e-05,
+      "loss": 12.1704,
+      "step": 11374
+    },
+    {
+      "epoch": 1.9892021857923496,
+      "grad_norm": 0.866391600511762,
+      "learning_rate": 5.385142507035673e-05,
+      "loss": 12.3362,
+      "step": 11375
+    },
+    {
+      "epoch": 1.989377049180328,
+      "grad_norm": 0.83674540794791,
+      "learning_rate": 5.383467574907738e-05,
+      "loss": 12.0292,
+      "step": 11376
+    },
+    {
+      "epoch": 1.9895519125683059,
+      "grad_norm": 0.8562605936490946,
+      "learning_rate": 5.381792807353789e-05,
+      "loss": 12.145,
+      "step": 11377
+    },
+    {
+      "epoch": 1.9897267759562842,
+      "grad_norm": 0.878085998448763,
+      "learning_rate": 5.380118204433533e-05,
+      "loss": 12.0473,
+      "step": 11378
+    },
+    {
+      "epoch": 1.9899016393442623,
+      "grad_norm": 0.7880152768453864,
+      "learning_rate": 5.3784437662066634e-05,
+      "loss": 12.3433,
+      "step": 11379
+    },
+    {
+      "epoch": 1.9900765027322405,
+      "grad_norm": 0.8812936324796734,
+      "learning_rate": 5.3767694927328784e-05,
+      "loss": 12.2383,
+      "step": 11380
+    },
+    {
+      "epoch": 1.9902513661202186,
+      "grad_norm": 0.8830389558806537,
+      "learning_rate": 5.375095384071855e-05,
+      "loss": 12.0833,
+      "step": 11381
+    },
+    {
+      "epoch": 1.9904262295081967,
+      "grad_norm": 0.8261855581979557,
+      "learning_rate": 5.373421440283277e-05,
+      "loss": 12.3606,
+      "step": 11382
+    },
+    {
+      "epoch": 1.9906010928961748,
+      "grad_norm": 0.8455916270426153,
+      "learning_rate": 5.371747661426825e-05,
+      "loss": 12.0351,
+      "step": 11383
+    },
+    {
+      "epoch": 1.990775956284153,
+      "grad_norm": 0.9992744452694671,
+      "learning_rate": 5.370074047562156e-05,
+      "loss": 12.5254,
+      "step": 11384
+    },
+    {
+      "epoch": 1.9909508196721313,
+      "grad_norm": 0.7653588356820331,
+      "learning_rate": 5.368400598748937e-05,
+      "loss": 12.1795,
+      "step": 11385
+    },
+    {
+      "epoch": 1.9911256830601092,
+      "grad_norm": 0.8266245378337351,
+      "learning_rate": 5.366727315046827e-05,
+      "loss": 12.1454,
+      "step": 11386
+    },
+    {
+      "epoch": 1.9913005464480875,
+      "grad_norm": 0.9218828859932822,
+      "learning_rate": 5.365054196515478e-05,
+      "loss": 12.371,
+      "step": 11387
+    },
+    {
+      "epoch": 1.9914754098360654,
+      "grad_norm": 0.8803623810594634,
+      "learning_rate": 5.363381243214527e-05,
+      "loss": 12.2693,
+      "step": 11388
+    },
+    {
+      "epoch": 1.9916502732240438,
+      "grad_norm": 0.881686328545193,
+      "learning_rate": 5.361708455203618e-05,
+      "loss": 12.099,
+      "step": 11389
+    },
+    {
+      "epoch": 1.9918251366120219,
+      "grad_norm": 0.833730140901784,
+      "learning_rate": 5.3600358325423825e-05,
+      "loss": 12.2309,
+      "step": 11390
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.8816903991964089,
+      "learning_rate": 5.358363375290454e-05,
+      "loss": 12.1044,
+      "step": 11391
+    },
+    {
+      "epoch": 1.9921748633879781,
+      "grad_norm": 0.855121709949791,
+      "learning_rate": 5.3566910835074425e-05,
+      "loss": 12.151,
+      "step": 11392
+    },
+    {
+      "epoch": 1.9923497267759562,
+      "grad_norm": 0.8428920755437617,
+      "learning_rate": 5.3550189572529685e-05,
+      "loss": 12.0869,
+      "step": 11393
+    },
+    {
+      "epoch": 1.9925245901639346,
+      "grad_norm": 0.8132015854408841,
+      "learning_rate": 5.353346996586643e-05,
+      "loss": 12.3026,
+      "step": 11394
+    },
+    {
+      "epoch": 1.9926994535519125,
+      "grad_norm": 0.8776273529408958,
+      "learning_rate": 5.351675201568067e-05,
+      "loss": 12.2637,
+      "step": 11395
+    },
+    {
+      "epoch": 1.9928743169398908,
+      "grad_norm": 0.8531162948835644,
+      "learning_rate": 5.3500035722568396e-05,
+      "loss": 12.2244,
+      "step": 11396
+    },
+    {
+      "epoch": 1.9930491803278687,
+      "grad_norm": 0.7952921888185005,
+      "learning_rate": 5.3483321087125525e-05,
+      "loss": 12.2495,
+      "step": 11397
+    },
+    {
+      "epoch": 1.993224043715847,
+      "grad_norm": 0.9107242568653583,
+      "learning_rate": 5.3466608109947935e-05,
+      "loss": 12.2397,
+      "step": 11398
+    },
+    {
+      "epoch": 1.9933989071038252,
+      "grad_norm": 0.9128385405802119,
+      "learning_rate": 5.344989679163137e-05,
+      "loss": 12.1042,
+      "step": 11399
+    },
+    {
+      "epoch": 1.9935737704918033,
+      "grad_norm": 0.890375477781187,
+      "learning_rate": 5.343318713277159e-05,
+      "loss": 12.0924,
+      "step": 11400
+    },
+    {
+      "epoch": 1.9937486338797814,
+      "grad_norm": 0.8487820609310212,
+      "learning_rate": 5.341647913396428e-05,
+      "loss": 12.1551,
+      "step": 11401
+    },
+    {
+      "epoch": 1.9939234972677595,
+      "grad_norm": 0.8789500223443876,
+      "learning_rate": 5.3399772795805106e-05,
+      "loss": 12.4406,
+      "step": 11402
+    },
+    {
+      "epoch": 1.9940983606557379,
+      "grad_norm": 0.9633899426963765,
+      "learning_rate": 5.338306811888954e-05,
+      "loss": 11.9506,
+      "step": 11403
+    },
+    {
+      "epoch": 1.9942732240437158,
+      "grad_norm": 0.7992966745657283,
+      "learning_rate": 5.336636510381312e-05,
+      "loss": 11.9401,
+      "step": 11404
+    },
+    {
+      "epoch": 1.994448087431694,
+      "grad_norm": 0.870665270575908,
+      "learning_rate": 5.334966375117132e-05,
+      "loss": 12.4416,
+      "step": 11405
+    },
+    {
+      "epoch": 1.994622950819672,
+      "grad_norm": 0.8507352272181526,
+      "learning_rate": 5.333296406155952e-05,
+      "loss": 12.4094,
+      "step": 11406
+    },
+    {
+      "epoch": 1.9947978142076503,
+      "grad_norm": 0.8648792200541738,
+      "learning_rate": 5.3316266035573e-05,
+      "loss": 12.3689,
+      "step": 11407
+    },
+    {
+      "epoch": 1.9949726775956285,
+      "grad_norm": 0.8265523130260706,
+      "learning_rate": 5.329956967380706e-05,
+      "loss": 11.6727,
+      "step": 11408
+    },
+    {
+      "epoch": 1.9951475409836066,
+      "grad_norm": 0.9570356344334237,
+      "learning_rate": 5.328287497685689e-05,
+      "loss": 12.2376,
+      "step": 11409
+    },
+    {
+      "epoch": 1.9953224043715847,
+      "grad_norm": 0.8241855317212115,
+      "learning_rate": 5.326618194531767e-05,
+      "loss": 12.3233,
+      "step": 11410
+    },
+    {
+      "epoch": 1.9954972677595628,
+      "grad_norm": 0.9406111780209117,
+      "learning_rate": 5.324949057978444e-05,
+      "loss": 12.2049,
+      "step": 11411
+    },
+    {
+      "epoch": 1.995672131147541,
+      "grad_norm": 0.8484339236678463,
+      "learning_rate": 5.323280088085225e-05,
+      "loss": 12.1011,
+      "step": 11412
+    },
+    {
+      "epoch": 1.995846994535519,
+      "grad_norm": 0.8425003802604275,
+      "learning_rate": 5.321611284911605e-05,
+      "loss": 11.9169,
+      "step": 11413
+    },
+    {
+      "epoch": 1.9960218579234974,
+      "grad_norm": 0.9119105611513538,
+      "learning_rate": 5.319942648517078e-05,
+      "loss": 12.1517,
+      "step": 11414
+    },
+    {
+      "epoch": 1.9961967213114753,
+      "grad_norm": 0.8629787133287546,
+      "learning_rate": 5.318274178961127e-05,
+      "loss": 12.1787,
+      "step": 11415
+    },
+    {
+      "epoch": 1.9963715846994536,
+      "grad_norm": 0.8828332296296444,
+      "learning_rate": 5.3166058763032315e-05,
+      "loss": 12.3241,
+      "step": 11416
+    },
+    {
+      "epoch": 1.9965464480874315,
+      "grad_norm": 0.8346288426712679,
+      "learning_rate": 5.314937740602868e-05,
+      "loss": 12.1436,
+      "step": 11417
+    },
+    {
+      "epoch": 1.9967213114754099,
+      "grad_norm": 0.8418686829345926,
+      "learning_rate": 5.313269771919498e-05,
+      "loss": 12.2536,
+      "step": 11418
+    },
+    {
+      "epoch": 1.996896174863388,
+      "grad_norm": 0.8497124162393768,
+      "learning_rate": 5.311601970312583e-05,
+      "loss": 12.2375,
+      "step": 11419
+    },
+    {
+      "epoch": 1.9970710382513661,
+      "grad_norm": 0.836629451946742,
+      "learning_rate": 5.30993433584158e-05,
+      "loss": 12.1056,
+      "step": 11420
+    },
+    {
+      "epoch": 1.9972459016393442,
+      "grad_norm": 0.9721544128185875,
+      "learning_rate": 5.3082668685659434e-05,
+      "loss": 12.2443,
+      "step": 11421
+    },
+    {
+      "epoch": 1.9974207650273224,
+      "grad_norm": 0.887583409758862,
+      "learning_rate": 5.3065995685451055e-05,
+      "loss": 12.1435,
+      "step": 11422
+    },
+    {
+      "epoch": 1.9975956284153007,
+      "grad_norm": 0.9162957304380673,
+      "learning_rate": 5.30493243583851e-05,
+      "loss": 12.2011,
+      "step": 11423
+    },
+    {
+      "epoch": 1.9977704918032786,
+      "grad_norm": 0.9168731986950565,
+      "learning_rate": 5.3032654705055886e-05,
+      "loss": 12.2716,
+      "step": 11424
+    },
+    {
+      "epoch": 1.997945355191257,
+      "grad_norm": 0.8750205624321047,
+      "learning_rate": 5.3015986726057686e-05,
+      "loss": 12.2686,
+      "step": 11425
+    },
+    {
+      "epoch": 1.9981202185792348,
+      "grad_norm": 0.9019533613851235,
+      "learning_rate": 5.299932042198463e-05,
+      "loss": 12.2421,
+      "step": 11426
+    },
+    {
+      "epoch": 1.9982950819672132,
+      "grad_norm": 0.928841387173018,
+      "learning_rate": 5.298265579343088e-05,
+      "loss": 11.9413,
+      "step": 11427
+    },
+    {
+      "epoch": 1.9984699453551913,
+      "grad_norm": 0.9387324022987977,
+      "learning_rate": 5.2965992840990565e-05,
+      "loss": 12.3378,
+      "step": 11428
+    },
+    {
+      "epoch": 1.9986448087431694,
+      "grad_norm": 0.8747086556598522,
+      "learning_rate": 5.294933156525761e-05,
+      "loss": 12.0219,
+      "step": 11429
+    },
+    {
+      "epoch": 1.9988196721311475,
+      "grad_norm": 0.8283200650983346,
+      "learning_rate": 5.293267196682601e-05,
+      "loss": 12.2362,
+      "step": 11430
+    },
+    {
+      "epoch": 1.9989945355191256,
+      "grad_norm": 0.8481214681477444,
+      "learning_rate": 5.291601404628965e-05,
+      "loss": 12.0379,
+      "step": 11431
+    },
+    {
+      "epoch": 1.999169398907104,
+      "grad_norm": 0.792896099604959,
+      "learning_rate": 5.28993578042424e-05,
+      "loss": 12.2522,
+      "step": 11432
+    },
+    {
+      "epoch": 1.9993442622950819,
+      "grad_norm": 0.8269789227706517,
+      "learning_rate": 5.2882703241277995e-05,
+      "loss": 12.3865,
+      "step": 11433
+    },
+    {
+      "epoch": 1.9995191256830602,
+      "grad_norm": 0.8825761015675465,
+      "learning_rate": 5.286605035799018e-05,
+      "loss": 12.2169,
+      "step": 11434
+    },
+    {
+      "epoch": 1.9996939890710381,
+      "grad_norm": 0.826713870079857,
+      "learning_rate": 5.28493991549726e-05,
+      "loss": 12.1736,
+      "step": 11435
+    },
+    {
+      "epoch": 1.9998688524590165,
+      "grad_norm": 0.8804796201662017,
+      "learning_rate": 5.283274963281888e-05,
+      "loss": 12.1193,
+      "step": 11436
+    },
+    {
+      "epoch": 2.0000437158469944,
+      "grad_norm": 0.9162320401544003,
+      "learning_rate": 5.28161017921225e-05,
+      "loss": 11.7641,
+      "step": 11437
+    },
+    {
+      "epoch": 2.0002185792349727,
+      "grad_norm": 0.872223966386479,
+      "learning_rate": 5.2799455633476946e-05,
+      "loss": 12.0886,
+      "step": 11438
+    },
+    {
+      "epoch": 2.0003934426229506,
+      "grad_norm": 0.7880921009809397,
+      "learning_rate": 5.2782811157475665e-05,
+      "loss": 12.0586,
+      "step": 11439
+    },
+    {
+      "epoch": 2.000568306010929,
+      "grad_norm": 0.8538654922905673,
+      "learning_rate": 5.2766168364712033e-05,
+      "loss": 11.9395,
+      "step": 11440
+    },
+    {
+      "epoch": 2.0007431693989073,
+      "grad_norm": 0.9393243421518569,
+      "learning_rate": 5.274952725577927e-05,
+      "loss": 12.0192,
+      "step": 11441
+    },
+    {
+      "epoch": 2.000918032786885,
+      "grad_norm": 0.8320361867955063,
+      "learning_rate": 5.273288783127066e-05,
+      "loss": 12.1779,
+      "step": 11442
+    },
+    {
+      "epoch": 2.0010928961748635,
+      "grad_norm": 0.9348801092001878,
+      "learning_rate": 5.2716250091779407e-05,
+      "loss": 12.2274,
+      "step": 11443
+    },
+    {
+      "epoch": 2.0012677595628414,
+      "grad_norm": 0.9878847696223728,
+      "learning_rate": 5.269961403789856e-05,
+      "loss": 12.4037,
+      "step": 11444
+    },
+    {
+      "epoch": 2.0014426229508198,
+      "grad_norm": 0.7680330537445444,
+      "learning_rate": 5.26829796702212e-05,
+      "loss": 12.0357,
+      "step": 11445
+    },
+    {
+      "epoch": 2.0016174863387977,
+      "grad_norm": 1.0261177421508012,
+      "learning_rate": 5.2666346989340344e-05,
+      "loss": 12.2361,
+      "step": 11446
+    },
+    {
+      "epoch": 2.001792349726776,
+      "grad_norm": 0.9977142947026689,
+      "learning_rate": 5.2649715995848904e-05,
+      "loss": 12.058,
+      "step": 11447
+    },
+    {
+      "epoch": 2.001967213114754,
+      "grad_norm": 0.8425969683777143,
+      "learning_rate": 5.263308669033977e-05,
+      "loss": 12.1171,
+      "step": 11448
+    },
+    {
+      "epoch": 2.0021420765027322,
+      "grad_norm": 0.8900261566044434,
+      "learning_rate": 5.261645907340575e-05,
+      "loss": 12.0204,
+      "step": 11449
+    },
+    {
+      "epoch": 2.0023169398907106,
+      "grad_norm": 0.8148013270655887,
+      "learning_rate": 5.25998331456396e-05,
+      "loss": 11.9691,
+      "step": 11450
+    },
+    {
+      "epoch": 2.0024918032786885,
+      "grad_norm": 0.7825519918961045,
+      "learning_rate": 5.2583208907634065e-05,
+      "loss": 12.0092,
+      "step": 11451
+    },
+    {
+      "epoch": 2.002666666666667,
+      "grad_norm": 1.0423256007059467,
+      "learning_rate": 5.2566586359981685e-05,
+      "loss": 12.0384,
+      "step": 11452
+    },
+    {
+      "epoch": 2.0028415300546447,
+      "grad_norm": 0.9471389995629305,
+      "learning_rate": 5.2549965503275076e-05,
+      "loss": 12.2087,
+      "step": 11453
+    },
+    {
+      "epoch": 2.003016393442623,
+      "grad_norm": 1.004364687340277,
+      "learning_rate": 5.253334633810676e-05,
+      "loss": 12.2701,
+      "step": 11454
+    },
+    {
+      "epoch": 2.003191256830601,
+      "grad_norm": 0.9882817502065865,
+      "learning_rate": 5.251672886506924e-05,
+      "loss": 12.0804,
+      "step": 11455
+    },
+    {
+      "epoch": 2.0033661202185793,
+      "grad_norm": 0.8680798115056488,
+      "learning_rate": 5.250011308475481e-05,
+      "loss": 12.1791,
+      "step": 11456
+    },
+    {
+      "epoch": 2.003540983606557,
+      "grad_norm": 0.9746445721269879,
+      "learning_rate": 5.2483498997755854e-05,
+      "loss": 12.214,
+      "step": 11457
+    },
+    {
+      "epoch": 2.0037158469945355,
+      "grad_norm": 0.8380125878047513,
+      "learning_rate": 5.246688660466468e-05,
+      "loss": 12.0135,
+      "step": 11458
+    },
+    {
+      "epoch": 2.003890710382514,
+      "grad_norm": 0.8292656833160694,
+      "learning_rate": 5.245027590607341e-05,
+      "loss": 12.1473,
+      "step": 11459
+    },
+    {
+      "epoch": 2.0040655737704918,
+      "grad_norm": 0.811888216370582,
+      "learning_rate": 5.243366690257427e-05,
+      "loss": 12.2067,
+      "step": 11460
+    },
+    {
+      "epoch": 2.00424043715847,
+      "grad_norm": 0.8552111657793885,
+      "learning_rate": 5.241705959475931e-05,
+      "loss": 12.1164,
+      "step": 11461
+    },
+    {
+      "epoch": 2.004415300546448,
+      "grad_norm": 0.8133000375976763,
+      "learning_rate": 5.2400453983220645e-05,
+      "loss": 12.1969,
+      "step": 11462
+    },
+    {
+      "epoch": 2.0045901639344264,
+      "grad_norm": 0.8175503153118707,
+      "learning_rate": 5.2383850068550114e-05,
+      "loss": 12.178,
+      "step": 11463
+    },
+    {
+      "epoch": 2.0047650273224042,
+      "grad_norm": 0.7846209464683862,
+      "learning_rate": 5.236724785133971e-05,
+      "loss": 12.2014,
+      "step": 11464
+    },
+    {
+      "epoch": 2.0049398907103826,
+      "grad_norm": 0.8427989702030071,
+      "learning_rate": 5.235064733218127e-05,
+      "loss": 12.1795,
+      "step": 11465
+    },
+    {
+      "epoch": 2.0051147540983605,
+      "grad_norm": 0.8435451120833807,
+      "learning_rate": 5.2334048511666564e-05,
+      "loss": 12.1118,
+      "step": 11466
+    },
+    {
+      "epoch": 2.005289617486339,
+      "grad_norm": 0.9054402415909678,
+      "learning_rate": 5.231745139038735e-05,
+      "loss": 12.1552,
+      "step": 11467
+    },
+    {
+      "epoch": 2.0054644808743167,
+      "grad_norm": 0.8119923822291346,
+      "learning_rate": 5.230085596893528e-05,
+      "loss": 12.1514,
+      "step": 11468
+    },
+    {
+      "epoch": 2.005639344262295,
+      "grad_norm": 0.9428751888151696,
+      "learning_rate": 5.228426224790197e-05,
+      "loss": 12.0806,
+      "step": 11469
+    },
+    {
+      "epoch": 2.0058142076502734,
+      "grad_norm": 0.8426159727365372,
+      "learning_rate": 5.2267670227878994e-05,
+      "loss": 12.1592,
+      "step": 11470
+    },
+    {
+      "epoch": 2.0059890710382513,
+      "grad_norm": 0.7969995249648528,
+      "learning_rate": 5.2251079909457765e-05,
+      "loss": 12.038,
+      "step": 11471
+    },
+    {
+      "epoch": 2.0061639344262296,
+      "grad_norm": 0.870340490985051,
+      "learning_rate": 5.2234491293229747e-05,
+      "loss": 12.141,
+      "step": 11472
+    },
+    {
+      "epoch": 2.0063387978142075,
+      "grad_norm": 0.8071942908183259,
+      "learning_rate": 5.2217904379786356e-05,
+      "loss": 11.9991,
+      "step": 11473
+    },
+    {
+      "epoch": 2.006513661202186,
+      "grad_norm": 0.8453335665010714,
+      "learning_rate": 5.2201319169718795e-05,
+      "loss": 12.1156,
+      "step": 11474
+    },
+    {
+      "epoch": 2.006688524590164,
+      "grad_norm": 0.8946034280655394,
+      "learning_rate": 5.218473566361837e-05,
+      "loss": 12.0019,
+      "step": 11475
+    },
+    {
+      "epoch": 2.006863387978142,
+      "grad_norm": 0.8148059859252423,
+      "learning_rate": 5.216815386207623e-05,
+      "loss": 12.0604,
+      "step": 11476
+    },
+    {
+      "epoch": 2.00703825136612,
+      "grad_norm": 0.8372630385840281,
+      "learning_rate": 5.215157376568358e-05,
+      "loss": 12.068,
+      "step": 11477
+    },
+    {
+      "epoch": 2.0072131147540984,
+      "grad_norm": 0.9375354799337461,
+      "learning_rate": 5.213499537503137e-05,
+      "loss": 12.0532,
+      "step": 11478
+    },
+    {
+      "epoch": 2.0073879781420767,
+      "grad_norm": 0.8256463597961361,
+      "learning_rate": 5.211841869071066e-05,
+      "loss": 12.2572,
+      "step": 11479
+    },
+    {
+      "epoch": 2.0075628415300546,
+      "grad_norm": 0.8449029350079251,
+      "learning_rate": 5.210184371331238e-05,
+      "loss": 12.0357,
+      "step": 11480
+    },
+    {
+      "epoch": 2.007737704918033,
+      "grad_norm": 1.0571356790525288,
+      "learning_rate": 5.208527044342744e-05,
+      "loss": 12.2408,
+      "step": 11481
+    },
+    {
+      "epoch": 2.007912568306011,
+      "grad_norm": 0.9009675475201214,
+      "learning_rate": 5.2068698881646596e-05,
+      "loss": 12.0123,
+      "step": 11482
+    },
+    {
+      "epoch": 2.008087431693989,
+      "grad_norm": 0.8855486381304796,
+      "learning_rate": 5.2052129028560625e-05,
+      "loss": 12.1592,
+      "step": 11483
+    },
+    {
+      "epoch": 2.008262295081967,
+      "grad_norm": 0.8083167995716016,
+      "learning_rate": 5.2035560884760246e-05,
+      "loss": 12.1336,
+      "step": 11484
+    },
+    {
+      "epoch": 2.0084371584699454,
+      "grad_norm": 0.8082083936645514,
+      "learning_rate": 5.2018994450836066e-05,
+      "loss": 11.82,
+      "step": 11485
+    },
+    {
+      "epoch": 2.0086120218579233,
+      "grad_norm": 0.8448789820079377,
+      "learning_rate": 5.200242972737869e-05,
+      "loss": 12.2245,
+      "step": 11486
+    },
+    {
+      "epoch": 2.0087868852459017,
+      "grad_norm": 0.8475793231531575,
+      "learning_rate": 5.198586671497861e-05,
+      "loss": 12.0537,
+      "step": 11487
+    },
+    {
+      "epoch": 2.00896174863388,
+      "grad_norm": 0.7898276052929052,
+      "learning_rate": 5.1969305414226275e-05,
+      "loss": 12.1699,
+      "step": 11488
+    },
+    {
+      "epoch": 2.009136612021858,
+      "grad_norm": 0.8447846949015401,
+      "learning_rate": 5.195274582571212e-05,
+      "loss": 12.1678,
+      "step": 11489
+    },
+    {
+      "epoch": 2.0093114754098362,
+      "grad_norm": 0.7994439033766871,
+      "learning_rate": 5.193618795002642e-05,
+      "loss": 11.8036,
+      "step": 11490
+    },
+    {
+      "epoch": 2.009486338797814,
+      "grad_norm": 0.8989616932784362,
+      "learning_rate": 5.191963178775945e-05,
+      "loss": 11.9515,
+      "step": 11491
+    },
+    {
+      "epoch": 2.0096612021857925,
+      "grad_norm": 0.8503780700106004,
+      "learning_rate": 5.190307733950146e-05,
+      "loss": 12.1291,
+      "step": 11492
+    },
+    {
+      "epoch": 2.0098360655737704,
+      "grad_norm": 0.8659320572020215,
+      "learning_rate": 5.188652460584254e-05,
+      "loss": 12.0547,
+      "step": 11493
+    },
+    {
+      "epoch": 2.0100109289617487,
+      "grad_norm": 0.8591914149721381,
+      "learning_rate": 5.186997358737281e-05,
+      "loss": 12.0141,
+      "step": 11494
+    },
+    {
+      "epoch": 2.0101857923497266,
+      "grad_norm": 0.8017939262023638,
+      "learning_rate": 5.1853424284682284e-05,
+      "loss": 12.0903,
+      "step": 11495
+    },
+    {
+      "epoch": 2.010360655737705,
+      "grad_norm": 0.8396973294717929,
+      "learning_rate": 5.183687669836097e-05,
+      "loss": 12.354,
+      "step": 11496
+    },
+    {
+      "epoch": 2.010535519125683,
+      "grad_norm": 0.8722844167663003,
+      "learning_rate": 5.182033082899868e-05,
+      "loss": 12.0664,
+      "step": 11497
+    },
+    {
+      "epoch": 2.010710382513661,
+      "grad_norm": 0.8094348491480677,
+      "learning_rate": 5.1803786677185306e-05,
+      "loss": 11.8523,
+      "step": 11498
+    },
+    {
+      "epoch": 2.0108852459016395,
+      "grad_norm": 0.8394238377535376,
+      "learning_rate": 5.178724424351063e-05,
+      "loss": 12.1626,
+      "step": 11499
+    },
+    {
+      "epoch": 2.0110601092896174,
+      "grad_norm": 0.9536124605698751,
+      "learning_rate": 5.17707035285644e-05,
+      "loss": 12.2723,
+      "step": 11500
+    },
+    {
+      "epoch": 2.0112349726775958,
+      "grad_norm": 0.7958911816454702,
+      "learning_rate": 5.17541645329362e-05,
+      "loss": 12.1088,
+      "step": 11501
+    },
+    {
+      "epoch": 2.0114098360655737,
+      "grad_norm": 0.8085148923398242,
+      "learning_rate": 5.173762725721567e-05,
+      "loss": 12.255,
+      "step": 11502
+    },
+    {
+      "epoch": 2.011584699453552,
+      "grad_norm": 0.8411398070419089,
+      "learning_rate": 5.1721091701992355e-05,
+      "loss": 11.9592,
+      "step": 11503
+    },
+    {
+      "epoch": 2.01175956284153,
+      "grad_norm": 0.8884592988766087,
+      "learning_rate": 5.170455786785568e-05,
+      "loss": 12.2751,
+      "step": 11504
+    },
+    {
+      "epoch": 2.0119344262295082,
+      "grad_norm": 0.8176132871236438,
+      "learning_rate": 5.168802575539511e-05,
+      "loss": 11.977,
+      "step": 11505
+    },
+    {
+      "epoch": 2.012109289617486,
+      "grad_norm": 0.8878020100669458,
+      "learning_rate": 5.167149536519997e-05,
+      "loss": 12.1818,
+      "step": 11506
+    },
+    {
+      "epoch": 2.0122841530054645,
+      "grad_norm": 0.8194108305284789,
+      "learning_rate": 5.165496669785961e-05,
+      "loss": 12.0982,
+      "step": 11507
+    },
+    {
+      "epoch": 2.012459016393443,
+      "grad_norm": 0.8185546380752089,
+      "learning_rate": 5.163843975396314e-05,
+      "loss": 12.1175,
+      "step": 11508
+    },
+    {
+      "epoch": 2.0126338797814207,
+      "grad_norm": 0.8576954376218585,
+      "learning_rate": 5.162191453409979e-05,
+      "loss": 12.0106,
+      "step": 11509
+    },
+    {
+      "epoch": 2.012808743169399,
+      "grad_norm": 0.8178454579466525,
+      "learning_rate": 5.160539103885869e-05,
+      "loss": 12.2328,
+      "step": 11510
+    },
+    {
+      "epoch": 2.012983606557377,
+      "grad_norm": 0.7675450002763233,
+      "learning_rate": 5.158886926882886e-05,
+      "loss": 12.1311,
+      "step": 11511
+    },
+    {
+      "epoch": 2.0131584699453553,
+      "grad_norm": 0.8810984821791683,
+      "learning_rate": 5.157234922459926e-05,
+      "loss": 11.9637,
+      "step": 11512
+    },
+    {
+      "epoch": 2.013333333333333,
+      "grad_norm": 0.8814031914797863,
+      "learning_rate": 5.155583090675882e-05,
+      "loss": 11.7912,
+      "step": 11513
+    },
+    {
+      "epoch": 2.0135081967213115,
+      "grad_norm": 0.8479303187490702,
+      "learning_rate": 5.153931431589641e-05,
+      "loss": 12.2001,
+      "step": 11514
+    },
+    {
+      "epoch": 2.0136830601092894,
+      "grad_norm": 0.7916603023343071,
+      "learning_rate": 5.1522799452600866e-05,
+      "loss": 12.1269,
+      "step": 11515
+    },
+    {
+      "epoch": 2.0138579234972678,
+      "grad_norm": 0.8562281145039391,
+      "learning_rate": 5.150628631746084e-05,
+      "loss": 11.9385,
+      "step": 11516
+    },
+    {
+      "epoch": 2.014032786885246,
+      "grad_norm": 0.8160671713058405,
+      "learning_rate": 5.148977491106506e-05,
+      "loss": 12.1681,
+      "step": 11517
+    },
+    {
+      "epoch": 2.014207650273224,
+      "grad_norm": 0.84129938612902,
+      "learning_rate": 5.147326523400211e-05,
+      "loss": 12.0679,
+      "step": 11518
+    },
+    {
+      "epoch": 2.0143825136612024,
+      "grad_norm": 0.9879580114625444,
+      "learning_rate": 5.1456757286860604e-05,
+      "loss": 12.1999,
+      "step": 11519
+    },
+    {
+      "epoch": 2.0145573770491803,
+      "grad_norm": 0.830036422427295,
+      "learning_rate": 5.144025107022894e-05,
+      "loss": 12.1439,
+      "step": 11520
+    },
+    {
+      "epoch": 2.0147322404371586,
+      "grad_norm": 0.9491012603731753,
+      "learning_rate": 5.142374658469559e-05,
+      "loss": 12.083,
+      "step": 11521
+    },
+    {
+      "epoch": 2.0149071038251365,
+      "grad_norm": 0.8548134754230599,
+      "learning_rate": 5.1407243830848915e-05,
+      "loss": 12.1091,
+      "step": 11522
+    },
+    {
+      "epoch": 2.015081967213115,
+      "grad_norm": 1.3095576909899689,
+      "learning_rate": 5.1390742809277223e-05,
+      "loss": 12.0537,
+      "step": 11523
+    },
+    {
+      "epoch": 2.0152568306010927,
+      "grad_norm": 0.8565999429381388,
+      "learning_rate": 5.137424352056877e-05,
+      "loss": 12.1009,
+      "step": 11524
+    },
+    {
+      "epoch": 2.015431693989071,
+      "grad_norm": 0.7945019468583112,
+      "learning_rate": 5.135774596531171e-05,
+      "loss": 11.8842,
+      "step": 11525
+    },
+    {
+      "epoch": 2.015606557377049,
+      "grad_norm": 0.8294736761168225,
+      "learning_rate": 5.134125014409422e-05,
+      "loss": 12.1266,
+      "step": 11526
+    },
+    {
+      "epoch": 2.0157814207650273,
+      "grad_norm": 0.9919602694521141,
+      "learning_rate": 5.1324756057504274e-05,
+      "loss": 12.239,
+      "step": 11527
+    },
+    {
+      "epoch": 2.0159562841530057,
+      "grad_norm": 0.817200380590453,
+      "learning_rate": 5.13082637061299e-05,
+      "loss": 11.8776,
+      "step": 11528
+    },
+    {
+      "epoch": 2.0161311475409835,
+      "grad_norm": 0.8418544892345262,
+      "learning_rate": 5.129177309055904e-05,
+      "loss": 12.236,
+      "step": 11529
+    },
+    {
+      "epoch": 2.016306010928962,
+      "grad_norm": 0.9897635021815742,
+      "learning_rate": 5.127528421137958e-05,
+      "loss": 12.233,
+      "step": 11530
+    },
+    {
+      "epoch": 2.01648087431694,
+      "grad_norm": 0.8244372662383517,
+      "learning_rate": 5.1258797069179285e-05,
+      "loss": 12.2372,
+      "step": 11531
+    },
+    {
+      "epoch": 2.016655737704918,
+      "grad_norm": 0.8183105142577372,
+      "learning_rate": 5.124231166454592e-05,
+      "loss": 12.0143,
+      "step": 11532
+    },
+    {
+      "epoch": 2.016830601092896,
+      "grad_norm": 0.8316084320811433,
+      "learning_rate": 5.122582799806719e-05,
+      "loss": 12.0757,
+      "step": 11533
+    },
+    {
+      "epoch": 2.0170054644808744,
+      "grad_norm": 0.8252877693484376,
+      "learning_rate": 5.120934607033072e-05,
+      "loss": 12.1457,
+      "step": 11534
+    },
+    {
+      "epoch": 2.0171803278688523,
+      "grad_norm": 0.8113986983069235,
+      "learning_rate": 5.1192865881924044e-05,
+      "loss": 12.044,
+      "step": 11535
+    },
+    {
+      "epoch": 2.0173551912568306,
+      "grad_norm": 0.812755098335897,
+      "learning_rate": 5.117638743343466e-05,
+      "loss": 12.3531,
+      "step": 11536
+    },
+    {
+      "epoch": 2.017530054644809,
+      "grad_norm": 0.9083406240059557,
+      "learning_rate": 5.115991072545002e-05,
+      "loss": 12.0258,
+      "step": 11537
+    },
+    {
+      "epoch": 2.017704918032787,
+      "grad_norm": 0.8995624444373247,
+      "learning_rate": 5.1143435758557515e-05,
+      "loss": 11.969,
+      "step": 11538
+    },
+    {
+      "epoch": 2.017879781420765,
+      "grad_norm": 0.8125128297775147,
+      "learning_rate": 5.1126962533344415e-05,
+      "loss": 12.1589,
+      "step": 11539
+    },
+    {
+      "epoch": 2.018054644808743,
+      "grad_norm": 0.9261333361017888,
+      "learning_rate": 5.111049105039803e-05,
+      "loss": 12.041,
+      "step": 11540
+    },
+    {
+      "epoch": 2.0182295081967214,
+      "grad_norm": 0.8666248851789801,
+      "learning_rate": 5.1094021310305526e-05,
+      "loss": 11.9574,
+      "step": 11541
+    },
+    {
+      "epoch": 2.0184043715846993,
+      "grad_norm": 0.899436094970062,
+      "learning_rate": 5.1077553313654e-05,
+      "loss": 12.1706,
+      "step": 11542
+    },
+    {
+      "epoch": 2.0185792349726777,
+      "grad_norm": 0.8236457618935118,
+      "learning_rate": 5.1061087061030536e-05,
+      "loss": 11.9673,
+      "step": 11543
+    },
+    {
+      "epoch": 2.0187540983606556,
+      "grad_norm": 0.7779485264066582,
+      "learning_rate": 5.1044622553022135e-05,
+      "loss": 12.0903,
+      "step": 11544
+    },
+    {
+      "epoch": 2.018928961748634,
+      "grad_norm": 0.8089881879434202,
+      "learning_rate": 5.102815979021578e-05,
+      "loss": 12.1148,
+      "step": 11545
+    },
+    {
+      "epoch": 2.0191038251366122,
+      "grad_norm": 0.9131337180579007,
+      "learning_rate": 5.101169877319828e-05,
+      "loss": 12.1481,
+      "step": 11546
+    },
+    {
+      "epoch": 2.01927868852459,
+      "grad_norm": 0.8145069799275922,
+      "learning_rate": 5.099523950255648e-05,
+      "loss": 12.2384,
+      "step": 11547
+    },
+    {
+      "epoch": 2.0194535519125685,
+      "grad_norm": 0.8750425604094472,
+      "learning_rate": 5.097878197887713e-05,
+      "loss": 12.2774,
+      "step": 11548
+    },
+    {
+      "epoch": 2.0196284153005464,
+      "grad_norm": 0.8290847697640132,
+      "learning_rate": 5.096232620274697e-05,
+      "loss": 12.0336,
+      "step": 11549
+    },
+    {
+      "epoch": 2.0198032786885247,
+      "grad_norm": 0.8712455751135948,
+      "learning_rate": 5.094587217475255e-05,
+      "loss": 12.2157,
+      "step": 11550
+    },
+    {
+      "epoch": 2.0199781420765026,
+      "grad_norm": 0.893581799777159,
+      "learning_rate": 5.092941989548048e-05,
+      "loss": 12.1884,
+      "step": 11551
+    },
+    {
+      "epoch": 2.020153005464481,
+      "grad_norm": 0.8550700501495322,
+      "learning_rate": 5.091296936551728e-05,
+      "loss": 11.9344,
+      "step": 11552
+    },
+    {
+      "epoch": 2.020327868852459,
+      "grad_norm": 0.8172989818821782,
+      "learning_rate": 5.089652058544934e-05,
+      "loss": 12.1816,
+      "step": 11553
+    },
+    {
+      "epoch": 2.020502732240437,
+      "grad_norm": 0.7826927997670895,
+      "learning_rate": 5.0880073555863075e-05,
+      "loss": 11.6386,
+      "step": 11554
+    },
+    {
+      "epoch": 2.020677595628415,
+      "grad_norm": 0.8401219633831088,
+      "learning_rate": 5.0863628277344786e-05,
+      "loss": 11.9352,
+      "step": 11555
+    },
+    {
+      "epoch": 2.0208524590163934,
+      "grad_norm": 0.8572055065144848,
+      "learning_rate": 5.0847184750480745e-05,
+      "loss": 12.1488,
+      "step": 11556
+    },
+    {
+      "epoch": 2.0210273224043718,
+      "grad_norm": 0.7981129663813998,
+      "learning_rate": 5.083074297585714e-05,
+      "loss": 11.9769,
+      "step": 11557
+    },
+    {
+      "epoch": 2.0212021857923497,
+      "grad_norm": 0.8106620989385613,
+      "learning_rate": 5.08143029540601e-05,
+      "loss": 11.8522,
+      "step": 11558
+    },
+    {
+      "epoch": 2.021377049180328,
+      "grad_norm": 0.8497225707655083,
+      "learning_rate": 5.079786468567569e-05,
+      "loss": 12.1835,
+      "step": 11559
+    },
+    {
+      "epoch": 2.021551912568306,
+      "grad_norm": 0.8667820601787474,
+      "learning_rate": 5.0781428171289955e-05,
+      "loss": 12.0141,
+      "step": 11560
+    },
+    {
+      "epoch": 2.0217267759562842,
+      "grad_norm": 0.8745682159192102,
+      "learning_rate": 5.076499341148877e-05,
+      "loss": 12.1305,
+      "step": 11561
+    },
+    {
+      "epoch": 2.021901639344262,
+      "grad_norm": 0.8316722139889942,
+      "learning_rate": 5.074856040685803e-05,
+      "loss": 12.029,
+      "step": 11562
+    },
+    {
+      "epoch": 2.0220765027322405,
+      "grad_norm": 0.8463488448504508,
+      "learning_rate": 5.073212915798358e-05,
+      "loss": 12.2279,
+      "step": 11563
+    },
+    {
+      "epoch": 2.0222513661202184,
+      "grad_norm": 0.799364704959339,
+      "learning_rate": 5.07156996654512e-05,
+      "loss": 11.9643,
+      "step": 11564
+    },
+    {
+      "epoch": 2.0224262295081967,
+      "grad_norm": 0.8269099124305924,
+      "learning_rate": 5.069927192984649e-05,
+      "loss": 12.1134,
+      "step": 11565
+    },
+    {
+      "epoch": 2.022601092896175,
+      "grad_norm": 0.8043997541737333,
+      "learning_rate": 5.068284595175515e-05,
+      "loss": 12.2828,
+      "step": 11566
+    },
+    {
+      "epoch": 2.022775956284153,
+      "grad_norm": 0.8950139097460252,
+      "learning_rate": 5.066642173176277e-05,
+      "loss": 12.2698,
+      "step": 11567
+    },
+    {
+      "epoch": 2.0229508196721313,
+      "grad_norm": 0.8325849468559869,
+      "learning_rate": 5.0649999270454776e-05,
+      "loss": 12.1449,
+      "step": 11568
+    },
+    {
+      "epoch": 2.023125683060109,
+      "grad_norm": 0.8938893801856487,
+      "learning_rate": 5.063357856841664e-05,
+      "loss": 12.2289,
+      "step": 11569
+    },
+    {
+      "epoch": 2.0233005464480875,
+      "grad_norm": 0.95630880171737,
+      "learning_rate": 5.061715962623375e-05,
+      "loss": 12.2864,
+      "step": 11570
+    },
+    {
+      "epoch": 2.0234754098360654,
+      "grad_norm": 0.8315096155042504,
+      "learning_rate": 5.060074244449147e-05,
+      "loss": 12.2378,
+      "step": 11571
+    },
+    {
+      "epoch": 2.023650273224044,
+      "grad_norm": 0.7816952032303686,
+      "learning_rate": 5.058432702377496e-05,
+      "loss": 12.1337,
+      "step": 11572
+    },
+    {
+      "epoch": 2.0238251366120217,
+      "grad_norm": 0.7820386789966037,
+      "learning_rate": 5.056791336466946e-05,
+      "loss": 11.9115,
+      "step": 11573
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 0.8298081757715681,
+      "learning_rate": 5.055150146776009e-05,
+      "loss": 11.9962,
+      "step": 11574
+    },
+    {
+      "epoch": 2.0241748633879784,
+      "grad_norm": 0.8251188971470595,
+      "learning_rate": 5.053509133363191e-05,
+      "loss": 11.9649,
+      "step": 11575
+    },
+    {
+      "epoch": 2.0243497267759563,
+      "grad_norm": 0.8933595850265007,
+      "learning_rate": 5.0518682962869946e-05,
+      "loss": 12.2077,
+      "step": 11576
+    },
+    {
+      "epoch": 2.0245245901639346,
+      "grad_norm": 0.837559761987165,
+      "learning_rate": 5.050227635605911e-05,
+      "loss": 12.0528,
+      "step": 11577
+    },
+    {
+      "epoch": 2.0246994535519125,
+      "grad_norm": 0.8595928627378718,
+      "learning_rate": 5.04858715137843e-05,
+      "loss": 12.153,
+      "step": 11578
+    },
+    {
+      "epoch": 2.024874316939891,
+      "grad_norm": 0.8773589461569743,
+      "learning_rate": 5.046946843663034e-05,
+      "loss": 12.0379,
+      "step": 11579
+    },
+    {
+      "epoch": 2.0250491803278687,
+      "grad_norm": 0.9079292244903326,
+      "learning_rate": 5.045306712518194e-05,
+      "loss": 12.032,
+      "step": 11580
+    },
+    {
+      "epoch": 2.025224043715847,
+      "grad_norm": 0.8453971131420468,
+      "learning_rate": 5.043666758002381e-05,
+      "loss": 12.2197,
+      "step": 11581
+    },
+    {
+      "epoch": 2.025398907103825,
+      "grad_norm": 0.8025715653638271,
+      "learning_rate": 5.0420269801740596e-05,
+      "loss": 12.1293,
+      "step": 11582
+    },
+    {
+      "epoch": 2.0255737704918033,
+      "grad_norm": 0.8865154690782008,
+      "learning_rate": 5.04038737909168e-05,
+      "loss": 12.0971,
+      "step": 11583
+    },
+    {
+      "epoch": 2.025748633879781,
+      "grad_norm": 0.9060326760626177,
+      "learning_rate": 5.038747954813696e-05,
+      "loss": 12.0131,
+      "step": 11584
+    },
+    {
+      "epoch": 2.0259234972677596,
+      "grad_norm": 0.8290125490404008,
+      "learning_rate": 5.037108707398551e-05,
+      "loss": 12.0413,
+      "step": 11585
+    },
+    {
+      "epoch": 2.026098360655738,
+      "grad_norm": 0.9601593278392974,
+      "learning_rate": 5.035469636904686e-05,
+      "loss": 12.1599,
+      "step": 11586
+    },
+    {
+      "epoch": 2.026273224043716,
+      "grad_norm": 0.9346139023833078,
+      "learning_rate": 5.033830743390524e-05,
+      "loss": 12.1537,
+      "step": 11587
+    },
+    {
+      "epoch": 2.026448087431694,
+      "grad_norm": 0.8513847273901649,
+      "learning_rate": 5.032192026914493e-05,
+      "loss": 12.1515,
+      "step": 11588
+    },
+    {
+      "epoch": 2.026622950819672,
+      "grad_norm": 0.9632212974484677,
+      "learning_rate": 5.030553487535013e-05,
+      "loss": 12.0416,
+      "step": 11589
+    },
+    {
+      "epoch": 2.0267978142076504,
+      "grad_norm": 0.9919357150860879,
+      "learning_rate": 5.0289151253104983e-05,
+      "loss": 12.078,
+      "step": 11590
+    },
+    {
+      "epoch": 2.0269726775956283,
+      "grad_norm": 0.8794395340013474,
+      "learning_rate": 5.027276940299348e-05,
+      "loss": 12.1875,
+      "step": 11591
+    },
+    {
+      "epoch": 2.0271475409836066,
+      "grad_norm": 0.8593658583407442,
+      "learning_rate": 5.025638932559964e-05,
+      "loss": 12.1194,
+      "step": 11592
+    },
+    {
+      "epoch": 2.0273224043715845,
+      "grad_norm": 0.8046091126588863,
+      "learning_rate": 5.0240011021507403e-05,
+      "loss": 12.0614,
+      "step": 11593
+    },
+    {
+      "epoch": 2.027497267759563,
+      "grad_norm": 0.8272668037576592,
+      "learning_rate": 5.022363449130063e-05,
+      "loss": 12.1975,
+      "step": 11594
+    },
+    {
+      "epoch": 2.027672131147541,
+      "grad_norm": 0.764247185096677,
+      "learning_rate": 5.020725973556315e-05,
+      "loss": 11.6858,
+      "step": 11595
+    },
+    {
+      "epoch": 2.027846994535519,
+      "grad_norm": 0.8909080772290399,
+      "learning_rate": 5.0190886754878665e-05,
+      "loss": 12.2319,
+      "step": 11596
+    },
+    {
+      "epoch": 2.0280218579234974,
+      "grad_norm": 0.8113983963428315,
+      "learning_rate": 5.017451554983088e-05,
+      "loss": 11.9559,
+      "step": 11597
+    },
+    {
+      "epoch": 2.0281967213114753,
+      "grad_norm": 0.8332426149408084,
+      "learning_rate": 5.015814612100342e-05,
+      "loss": 12.1313,
+      "step": 11598
+    },
+    {
+      "epoch": 2.0283715846994537,
+      "grad_norm": 0.9649106754040947,
+      "learning_rate": 5.01417784689798e-05,
+      "loss": 12.1919,
+      "step": 11599
+    },
+    {
+      "epoch": 2.0285464480874316,
+      "grad_norm": 0.8240480983745241,
+      "learning_rate": 5.012541259434352e-05,
+      "loss": 11.986,
+      "step": 11600
+    },
+    {
+      "epoch": 2.02872131147541,
+      "grad_norm": 0.8240354987542744,
+      "learning_rate": 5.0109048497678046e-05,
+      "loss": 12.0279,
+      "step": 11601
+    },
+    {
+      "epoch": 2.028896174863388,
+      "grad_norm": 0.9444841240827441,
+      "learning_rate": 5.009268617956666e-05,
+      "loss": 12.0062,
+      "step": 11602
+    },
+    {
+      "epoch": 2.029071038251366,
+      "grad_norm": 0.9340231889299242,
+      "learning_rate": 5.0076325640592704e-05,
+      "loss": 12.1728,
+      "step": 11603
+    },
+    {
+      "epoch": 2.029245901639344,
+      "grad_norm": 0.9969499838762136,
+      "learning_rate": 5.0059966881339413e-05,
+      "loss": 12.2419,
+      "step": 11604
+    },
+    {
+      "epoch": 2.0294207650273224,
+      "grad_norm": 0.8094489638922827,
+      "learning_rate": 5.004360990238999e-05,
+      "loss": 11.9345,
+      "step": 11605
+    },
+    {
+      "epoch": 2.0295956284153007,
+      "grad_norm": 0.8395534167694031,
+      "learning_rate": 5.0027254704327474e-05,
+      "loss": 12.2955,
+      "step": 11606
+    },
+    {
+      "epoch": 2.0297704918032786,
+      "grad_norm": 0.9570051420771795,
+      "learning_rate": 5.001090128773493e-05,
+      "loss": 12.08,
+      "step": 11607
+    },
+    {
+      "epoch": 2.029945355191257,
+      "grad_norm": 0.9311724672632875,
+      "learning_rate": 4.999454965319536e-05,
+      "loss": 12.1425,
+      "step": 11608
+    },
+    {
+      "epoch": 2.030120218579235,
+      "grad_norm": 0.788977598069167,
+      "learning_rate": 4.997819980129169e-05,
+      "loss": 12.0502,
+      "step": 11609
+    },
+    {
+      "epoch": 2.030295081967213,
+      "grad_norm": 0.8511836873606785,
+      "learning_rate": 4.996185173260674e-05,
+      "loss": 12.0198,
+      "step": 11610
+    },
+    {
+      "epoch": 2.030469945355191,
+      "grad_norm": 0.9448934466045638,
+      "learning_rate": 4.9945505447723285e-05,
+      "loss": 12.0597,
+      "step": 11611
+    },
+    {
+      "epoch": 2.0306448087431694,
+      "grad_norm": 0.8814964862490376,
+      "learning_rate": 4.99291609472241e-05,
+      "loss": 12.2056,
+      "step": 11612
+    },
+    {
+      "epoch": 2.0308196721311473,
+      "grad_norm": 0.8648860627316209,
+      "learning_rate": 4.991281823169182e-05,
+      "loss": 12.1542,
+      "step": 11613
+    },
+    {
+      "epoch": 2.0309945355191257,
+      "grad_norm": 0.8595032613829806,
+      "learning_rate": 4.9896477301709046e-05,
+      "loss": 11.8625,
+      "step": 11614
+    },
+    {
+      "epoch": 2.031169398907104,
+      "grad_norm": 0.9896627181343652,
+      "learning_rate": 4.988013815785831e-05,
+      "loss": 12.3306,
+      "step": 11615
+    },
+    {
+      "epoch": 2.031344262295082,
+      "grad_norm": 0.8354061903381554,
+      "learning_rate": 4.986380080072214e-05,
+      "loss": 11.7951,
+      "step": 11616
+    },
+    {
+      "epoch": 2.0315191256830603,
+      "grad_norm": 1.0521906521469815,
+      "learning_rate": 4.984746523088285e-05,
+      "loss": 12.049,
+      "step": 11617
+    },
+    {
+      "epoch": 2.031693989071038,
+      "grad_norm": 0.8739166001537023,
+      "learning_rate": 4.9831131448922816e-05,
+      "loss": 12.1453,
+      "step": 11618
+    },
+    {
+      "epoch": 2.0318688524590165,
+      "grad_norm": 0.8171846789606124,
+      "learning_rate": 4.9814799455424354e-05,
+      "loss": 12.266,
+      "step": 11619
+    },
+    {
+      "epoch": 2.0320437158469944,
+      "grad_norm": 1.0448498097765475,
+      "learning_rate": 4.979846925096968e-05,
+      "loss": 12.1875,
+      "step": 11620
+    },
+    {
+      "epoch": 2.0322185792349727,
+      "grad_norm": 0.9639484402758619,
+      "learning_rate": 4.978214083614089e-05,
+      "loss": 11.9461,
+      "step": 11621
+    },
+    {
+      "epoch": 2.0323934426229506,
+      "grad_norm": 0.8716907679152626,
+      "learning_rate": 4.97658142115201e-05,
+      "loss": 12.181,
+      "step": 11622
+    },
+    {
+      "epoch": 2.032568306010929,
+      "grad_norm": 0.9792232562063573,
+      "learning_rate": 4.974948937768934e-05,
+      "loss": 11.8204,
+      "step": 11623
+    },
+    {
+      "epoch": 2.0327431693989073,
+      "grad_norm": 0.9800303194250521,
+      "learning_rate": 4.9733166335230616e-05,
+      "loss": 12.1177,
+      "step": 11624
+    },
+    {
+      "epoch": 2.032918032786885,
+      "grad_norm": 0.87888671122578,
+      "learning_rate": 4.971684508472575e-05,
+      "loss": 12.2974,
+      "step": 11625
+    },
+    {
+      "epoch": 2.0330928961748636,
+      "grad_norm": 0.8804704644870891,
+      "learning_rate": 4.97005256267566e-05,
+      "loss": 11.6749,
+      "step": 11626
+    },
+    {
+      "epoch": 2.0332677595628414,
+      "grad_norm": 0.9221253182116044,
+      "learning_rate": 4.968420796190494e-05,
+      "loss": 11.8895,
+      "step": 11627
+    },
+    {
+      "epoch": 2.03344262295082,
+      "grad_norm": 0.9663169915000723,
+      "learning_rate": 4.9667892090752476e-05,
+      "loss": 12.1376,
+      "step": 11628
+    },
+    {
+      "epoch": 2.0336174863387977,
+      "grad_norm": 0.8886248318281458,
+      "learning_rate": 4.965157801388086e-05,
+      "loss": 12.142,
+      "step": 11629
+    },
+    {
+      "epoch": 2.033792349726776,
+      "grad_norm": 0.988159222859319,
+      "learning_rate": 4.963526573187166e-05,
+      "loss": 12.1477,
+      "step": 11630
+    },
+    {
+      "epoch": 2.033967213114754,
+      "grad_norm": 0.8404718460853456,
+      "learning_rate": 4.961895524530643e-05,
+      "loss": 11.8941,
+      "step": 11631
+    },
+    {
+      "epoch": 2.0341420765027323,
+      "grad_norm": 0.8602796084100226,
+      "learning_rate": 4.9602646554766544e-05,
+      "loss": 12.0656,
+      "step": 11632
+    },
+    {
+      "epoch": 2.0343169398907106,
+      "grad_norm": 0.9474283338318651,
+      "learning_rate": 4.958633966083341e-05,
+      "loss": 12.0209,
+      "step": 11633
+    },
+    {
+      "epoch": 2.0344918032786885,
+      "grad_norm": 0.859162930279025,
+      "learning_rate": 4.957003456408839e-05,
+      "loss": 12.2376,
+      "step": 11634
+    },
+    {
+      "epoch": 2.034666666666667,
+      "grad_norm": 0.9057239380968366,
+      "learning_rate": 4.9553731265112745e-05,
+      "loss": 12.1109,
+      "step": 11635
+    },
+    {
+      "epoch": 2.0348415300546447,
+      "grad_norm": 0.8915342853708876,
+      "learning_rate": 4.95374297644876e-05,
+      "loss": 12.2732,
+      "step": 11636
+    },
+    {
+      "epoch": 2.035016393442623,
+      "grad_norm": 0.8127368287578514,
+      "learning_rate": 4.952113006279412e-05,
+      "loss": 12.0676,
+      "step": 11637
+    },
+    {
+      "epoch": 2.035191256830601,
+      "grad_norm": 0.7965284675712353,
+      "learning_rate": 4.950483216061338e-05,
+      "loss": 12.0793,
+      "step": 11638
+    },
+    {
+      "epoch": 2.0353661202185793,
+      "grad_norm": 0.8476733691407229,
+      "learning_rate": 4.9488536058526416e-05,
+      "loss": 12.0063,
+      "step": 11639
+    },
+    {
+      "epoch": 2.035540983606557,
+      "grad_norm": 0.8620696613710723,
+      "learning_rate": 4.9472241757114094e-05,
+      "loss": 12.2393,
+      "step": 11640
+    },
+    {
+      "epoch": 2.0357158469945356,
+      "grad_norm": 0.8391404588166911,
+      "learning_rate": 4.94559492569573e-05,
+      "loss": 12.0363,
+      "step": 11641
+    },
+    {
+      "epoch": 2.0358907103825135,
+      "grad_norm": 1.0316548414812572,
+      "learning_rate": 4.943965855863687e-05,
+      "loss": 12.1301,
+      "step": 11642
+    },
+    {
+      "epoch": 2.036065573770492,
+      "grad_norm": 0.820026846217024,
+      "learning_rate": 4.9423369662733574e-05,
+      "loss": 12.2132,
+      "step": 11643
+    },
+    {
+      "epoch": 2.03624043715847,
+      "grad_norm": 0.8070347246675351,
+      "learning_rate": 4.940708256982803e-05,
+      "loss": 12.0704,
+      "step": 11644
+    },
+    {
+      "epoch": 2.036415300546448,
+      "grad_norm": 0.8171789364131505,
+      "learning_rate": 4.9390797280500876e-05,
+      "loss": 11.9972,
+      "step": 11645
+    },
+    {
+      "epoch": 2.0365901639344264,
+      "grad_norm": 0.8478891684233358,
+      "learning_rate": 4.937451379533267e-05,
+      "loss": 12.1053,
+      "step": 11646
+    },
+    {
+      "epoch": 2.0367650273224043,
+      "grad_norm": 0.9239264135253089,
+      "learning_rate": 4.9358232114903894e-05,
+      "loss": 12.0785,
+      "step": 11647
+    },
+    {
+      "epoch": 2.0369398907103826,
+      "grad_norm": 0.8909638419711186,
+      "learning_rate": 4.9341952239794965e-05,
+      "loss": 12.184,
+      "step": 11648
+    },
+    {
+      "epoch": 2.0371147540983605,
+      "grad_norm": 0.9229591214640982,
+      "learning_rate": 4.9325674170586266e-05,
+      "loss": 12.1761,
+      "step": 11649
+    },
+    {
+      "epoch": 2.037289617486339,
+      "grad_norm": 0.8604295316075751,
+      "learning_rate": 4.93093979078581e-05,
+      "loss": 12.0886,
+      "step": 11650
+    },
+    {
+      "epoch": 2.0374644808743168,
+      "grad_norm": 0.8421084752493309,
+      "learning_rate": 4.929312345219065e-05,
+      "loss": 12.2718,
+      "step": 11651
+    },
+    {
+      "epoch": 2.037639344262295,
+      "grad_norm": 0.8877516302557857,
+      "learning_rate": 4.92768508041641e-05,
+      "loss": 12.0347,
+      "step": 11652
+    },
+    {
+      "epoch": 2.0378142076502734,
+      "grad_norm": 0.8832269870763751,
+      "learning_rate": 4.926057996435855e-05,
+      "loss": 12.3455,
+      "step": 11653
+    },
+    {
+      "epoch": 2.0379890710382513,
+      "grad_norm": 0.8547306358838793,
+      "learning_rate": 4.9244310933354074e-05,
+      "loss": 12.142,
+      "step": 11654
+    },
+    {
+      "epoch": 2.0381639344262297,
+      "grad_norm": 0.949470639672955,
+      "learning_rate": 4.922804371173059e-05,
+      "loss": 12.2879,
+      "step": 11655
+    },
+    {
+      "epoch": 2.0383387978142076,
+      "grad_norm": 0.9728272977962659,
+      "learning_rate": 4.9211778300068024e-05,
+      "loss": 12.2816,
+      "step": 11656
+    },
+    {
+      "epoch": 2.038513661202186,
+      "grad_norm": 0.8112382659225091,
+      "learning_rate": 4.919551469894621e-05,
+      "loss": 11.9874,
+      "step": 11657
+    },
+    {
+      "epoch": 2.038688524590164,
+      "grad_norm": 0.8517691482091337,
+      "learning_rate": 4.9179252908944985e-05,
+      "loss": 12.0689,
+      "step": 11658
+    },
+    {
+      "epoch": 2.038863387978142,
+      "grad_norm": 0.830890161750312,
+      "learning_rate": 4.916299293064398e-05,
+      "loss": 12.1531,
+      "step": 11659
+    },
+    {
+      "epoch": 2.03903825136612,
+      "grad_norm": 0.8238745765526878,
+      "learning_rate": 4.914673476462287e-05,
+      "loss": 12.1838,
+      "step": 11660
+    },
+    {
+      "epoch": 2.0392131147540984,
+      "grad_norm": 0.8914462895621383,
+      "learning_rate": 4.91304784114613e-05,
+      "loss": 12.0192,
+      "step": 11661
+    },
+    {
+      "epoch": 2.0393879781420763,
+      "grad_norm": 0.922514780382417,
+      "learning_rate": 4.91142238717387e-05,
+      "loss": 11.8475,
+      "step": 11662
+    },
+    {
+      "epoch": 2.0395628415300546,
+      "grad_norm": 0.8421796176647846,
+      "learning_rate": 4.909797114603456e-05,
+      "loss": 12.267,
+      "step": 11663
+    },
+    {
+      "epoch": 2.039737704918033,
+      "grad_norm": 0.8203056858001239,
+      "learning_rate": 4.908172023492828e-05,
+      "loss": 11.9229,
+      "step": 11664
+    },
+    {
+      "epoch": 2.039912568306011,
+      "grad_norm": 0.8605266674824325,
+      "learning_rate": 4.906547113899918e-05,
+      "loss": 11.714,
+      "step": 11665
+    },
+    {
+      "epoch": 2.040087431693989,
+      "grad_norm": 0.8774676238181531,
+      "learning_rate": 4.904922385882652e-05,
+      "loss": 12.2115,
+      "step": 11666
+    },
+    {
+      "epoch": 2.040262295081967,
+      "grad_norm": 0.8401749016703374,
+      "learning_rate": 4.9032978394989505e-05,
+      "loss": 12.1048,
+      "step": 11667
+    },
+    {
+      "epoch": 2.0404371584699454,
+      "grad_norm": 0.9398374949055076,
+      "learning_rate": 4.901673474806726e-05,
+      "loss": 12.1427,
+      "step": 11668
+    },
+    {
+      "epoch": 2.0406120218579233,
+      "grad_norm": 0.8718468391437486,
+      "learning_rate": 4.900049291863889e-05,
+      "loss": 12.0676,
+      "step": 11669
+    },
+    {
+      "epoch": 2.0407868852459017,
+      "grad_norm": 0.8088892096735718,
+      "learning_rate": 4.898425290728332e-05,
+      "loss": 12.08,
+      "step": 11670
+    },
+    {
+      "epoch": 2.0409617486338796,
+      "grad_norm": 0.8827638285330437,
+      "learning_rate": 4.896801471457953e-05,
+      "loss": 11.8015,
+      "step": 11671
+    },
+    {
+      "epoch": 2.041136612021858,
+      "grad_norm": 0.9377145235504677,
+      "learning_rate": 4.89517783411064e-05,
+      "loss": 12.2019,
+      "step": 11672
+    },
+    {
+      "epoch": 2.0413114754098363,
+      "grad_norm": 0.8457120849625256,
+      "learning_rate": 4.893554378744276e-05,
+      "loss": 11.967,
+      "step": 11673
+    },
+    {
+      "epoch": 2.041486338797814,
+      "grad_norm": 1.023677745178166,
+      "learning_rate": 4.8919311054167284e-05,
+      "loss": 12.1346,
+      "step": 11674
+    },
+    {
+      "epoch": 2.0416612021857925,
+      "grad_norm": 0.9100137510572328,
+      "learning_rate": 4.8903080141858694e-05,
+      "loss": 12.1398,
+      "step": 11675
+    },
+    {
+      "epoch": 2.0418360655737704,
+      "grad_norm": 0.8407209044625448,
+      "learning_rate": 4.888685105109563e-05,
+      "loss": 12.0873,
+      "step": 11676
+    },
+    {
+      "epoch": 2.0420109289617487,
+      "grad_norm": 0.8691279389178417,
+      "learning_rate": 4.887062378245656e-05,
+      "loss": 12.1531,
+      "step": 11677
+    },
+    {
+      "epoch": 2.0421857923497266,
+      "grad_norm": 1.0928833824030315,
+      "learning_rate": 4.885439833652002e-05,
+      "loss": 12.5315,
+      "step": 11678
+    },
+    {
+      "epoch": 2.042360655737705,
+      "grad_norm": 0.7846710648460072,
+      "learning_rate": 4.8838174713864425e-05,
+      "loss": 11.9375,
+      "step": 11679
+    },
+    {
+      "epoch": 2.042535519125683,
+      "grad_norm": 0.8586258185812929,
+      "learning_rate": 4.882195291506816e-05,
+      "loss": 12.1964,
+      "step": 11680
+    },
+    {
+      "epoch": 2.042710382513661,
+      "grad_norm": 0.8659719559942896,
+      "learning_rate": 4.880573294070944e-05,
+      "loss": 11.8162,
+      "step": 11681
+    },
+    {
+      "epoch": 2.0428852459016396,
+      "grad_norm": 0.8791016205710175,
+      "learning_rate": 4.878951479136653e-05,
+      "loss": 11.9805,
+      "step": 11682
+    },
+    {
+      "epoch": 2.0430601092896175,
+      "grad_norm": 0.8298963443218791,
+      "learning_rate": 4.877329846761758e-05,
+      "loss": 11.908,
+      "step": 11683
+    },
+    {
+      "epoch": 2.043234972677596,
+      "grad_norm": 0.8581133071705419,
+      "learning_rate": 4.875708397004069e-05,
+      "loss": 12.0223,
+      "step": 11684
+    },
+    {
+      "epoch": 2.0434098360655737,
+      "grad_norm": 0.8644391582887795,
+      "learning_rate": 4.874087129921389e-05,
+      "loss": 12.0458,
+      "step": 11685
+    },
+    {
+      "epoch": 2.043584699453552,
+      "grad_norm": 0.8090856454424012,
+      "learning_rate": 4.872466045571513e-05,
+      "loss": 11.933,
+      "step": 11686
+    },
+    {
+      "epoch": 2.04375956284153,
+      "grad_norm": 1.0001832488811249,
+      "learning_rate": 4.8708451440122326e-05,
+      "loss": 12.3448,
+      "step": 11687
+    },
+    {
+      "epoch": 2.0439344262295083,
+      "grad_norm": 0.8535476710906427,
+      "learning_rate": 4.869224425301333e-05,
+      "loss": 12.0962,
+      "step": 11688
+    },
+    {
+      "epoch": 2.044109289617486,
+      "grad_norm": 1.328103614569209,
+      "learning_rate": 4.867603889496585e-05,
+      "loss": 12.0705,
+      "step": 11689
+    },
+    {
+      "epoch": 2.0442841530054645,
+      "grad_norm": 1.042765384023151,
+      "learning_rate": 4.865983536655761e-05,
+      "loss": 11.902,
+      "step": 11690
+    },
+    {
+      "epoch": 2.0444590163934424,
+      "grad_norm": 0.8982349855288919,
+      "learning_rate": 4.8643633668366295e-05,
+      "loss": 11.8845,
+      "step": 11691
+    },
+    {
+      "epoch": 2.0446338797814207,
+      "grad_norm": 0.8087759216576449,
+      "learning_rate": 4.862743380096939e-05,
+      "loss": 11.9742,
+      "step": 11692
+    },
+    {
+      "epoch": 2.044808743169399,
+      "grad_norm": 0.8297223274699687,
+      "learning_rate": 4.8611235764944466e-05,
+      "loss": 11.9283,
+      "step": 11693
+    },
+    {
+      "epoch": 2.044983606557377,
+      "grad_norm": 1.0114601125384568,
+      "learning_rate": 4.859503956086893e-05,
+      "loss": 12.0324,
+      "step": 11694
+    },
+    {
+      "epoch": 2.0451584699453553,
+      "grad_norm": 0.8656152543779624,
+      "learning_rate": 4.857884518932022e-05,
+      "loss": 12.0023,
+      "step": 11695
+    },
+    {
+      "epoch": 2.0453333333333332,
+      "grad_norm": 0.9076939515227826,
+      "learning_rate": 4.856265265087555e-05,
+      "loss": 11.8444,
+      "step": 11696
+    },
+    {
+      "epoch": 2.0455081967213116,
+      "grad_norm": 0.8595160982953063,
+      "learning_rate": 4.8546461946112234e-05,
+      "loss": 12.0163,
+      "step": 11697
+    },
+    {
+      "epoch": 2.0456830601092895,
+      "grad_norm": 0.8378950056682104,
+      "learning_rate": 4.853027307560741e-05,
+      "loss": 12.2894,
+      "step": 11698
+    },
+    {
+      "epoch": 2.045857923497268,
+      "grad_norm": 0.8668371893864443,
+      "learning_rate": 4.851408603993828e-05,
+      "loss": 11.9167,
+      "step": 11699
+    },
+    {
+      "epoch": 2.0460327868852457,
+      "grad_norm": 0.894084896548796,
+      "learning_rate": 4.849790083968177e-05,
+      "loss": 12.1828,
+      "step": 11700
+    },
+    {
+      "epoch": 2.046207650273224,
+      "grad_norm": 0.911510709959484,
+      "learning_rate": 4.848171747541492e-05,
+      "loss": 12.3085,
+      "step": 11701
+    },
+    {
+      "epoch": 2.0463825136612024,
+      "grad_norm": 0.8070128389958798,
+      "learning_rate": 4.846553594771466e-05,
+      "loss": 11.7241,
+      "step": 11702
+    },
+    {
+      "epoch": 2.0465573770491803,
+      "grad_norm": 0.9095722848171389,
+      "learning_rate": 4.844935625715783e-05,
+      "loss": 12.2302,
+      "step": 11703
+    },
+    {
+      "epoch": 2.0467322404371586,
+      "grad_norm": 0.8484548596638241,
+      "learning_rate": 4.843317840432121e-05,
+      "loss": 11.9661,
+      "step": 11704
+    },
+    {
+      "epoch": 2.0469071038251365,
+      "grad_norm": 0.7830503018349436,
+      "learning_rate": 4.841700238978154e-05,
+      "loss": 12.1064,
+      "step": 11705
+    },
+    {
+      "epoch": 2.047081967213115,
+      "grad_norm": 0.7835894570995653,
+      "learning_rate": 4.840082821411547e-05,
+      "loss": 12.073,
+      "step": 11706
+    },
+    {
+      "epoch": 2.0472568306010928,
+      "grad_norm": 0.9295289533587182,
+      "learning_rate": 4.838465587789961e-05,
+      "loss": 12.2776,
+      "step": 11707
+    },
+    {
+      "epoch": 2.047431693989071,
+      "grad_norm": 0.8647687452369979,
+      "learning_rate": 4.836848538171044e-05,
+      "loss": 11.8746,
+      "step": 11708
+    },
+    {
+      "epoch": 2.047606557377049,
+      "grad_norm": 0.856828487414748,
+      "learning_rate": 4.835231672612443e-05,
+      "loss": 11.9508,
+      "step": 11709
+    },
+    {
+      "epoch": 2.0477814207650273,
+      "grad_norm": 0.8561495294558744,
+      "learning_rate": 4.8336149911718035e-05,
+      "loss": 11.9427,
+      "step": 11710
+    },
+    {
+      "epoch": 2.0479562841530057,
+      "grad_norm": 0.8530321269354314,
+      "learning_rate": 4.8319984939067476e-05,
+      "loss": 12.346,
+      "step": 11711
+    },
+    {
+      "epoch": 2.0481311475409836,
+      "grad_norm": 0.8702303399705901,
+      "learning_rate": 4.8303821808749085e-05,
+      "loss": 12.1038,
+      "step": 11712
+    },
+    {
+      "epoch": 2.048306010928962,
+      "grad_norm": 0.8003780466889453,
+      "learning_rate": 4.828766052133904e-05,
+      "loss": 11.8422,
+      "step": 11713
+    },
+    {
+      "epoch": 2.04848087431694,
+      "grad_norm": 0.7897273855056495,
+      "learning_rate": 4.827150107741353e-05,
+      "loss": 11.9261,
+      "step": 11714
+    },
+    {
+      "epoch": 2.048655737704918,
+      "grad_norm": 0.8629692632260452,
+      "learning_rate": 4.825534347754851e-05,
+      "loss": 11.5944,
+      "step": 11715
+    },
+    {
+      "epoch": 2.048830601092896,
+      "grad_norm": 0.8191056035336025,
+      "learning_rate": 4.823918772232005e-05,
+      "loss": 11.9053,
+      "step": 11716
+    },
+    {
+      "epoch": 2.0490054644808744,
+      "grad_norm": 0.9468799636927392,
+      "learning_rate": 4.8223033812304074e-05,
+      "loss": 11.7389,
+      "step": 11717
+    },
+    {
+      "epoch": 2.0491803278688523,
+      "grad_norm": 0.9197360521479546,
+      "learning_rate": 4.8206881748076435e-05,
+      "loss": 12.1537,
+      "step": 11718
+    },
+    {
+      "epoch": 2.0493551912568306,
+      "grad_norm": 0.8059525371948063,
+      "learning_rate": 4.8190731530212996e-05,
+      "loss": 12.0107,
+      "step": 11719
+    },
+    {
+      "epoch": 2.0495300546448085,
+      "grad_norm": 1.0122946207079764,
+      "learning_rate": 4.817458315928941e-05,
+      "loss": 12.1418,
+      "step": 11720
+    },
+    {
+      "epoch": 2.049704918032787,
+      "grad_norm": 0.9399111335258142,
+      "learning_rate": 4.815843663588138e-05,
+      "loss": 11.9413,
+      "step": 11721
+    },
+    {
+      "epoch": 2.049879781420765,
+      "grad_norm": 0.9611955229525736,
+      "learning_rate": 4.814229196056451e-05,
+      "loss": 12.0461,
+      "step": 11722
+    },
+    {
+      "epoch": 2.050054644808743,
+      "grad_norm": 0.9639692525259873,
+      "learning_rate": 4.8126149133914344e-05,
+      "loss": 11.9509,
+      "step": 11723
+    },
+    {
+      "epoch": 2.0502295081967215,
+      "grad_norm": 0.9032676218733949,
+      "learning_rate": 4.811000815650636e-05,
+      "loss": 12.0289,
+      "step": 11724
+    },
+    {
+      "epoch": 2.0504043715846993,
+      "grad_norm": 0.8448954593860305,
+      "learning_rate": 4.809386902891598e-05,
+      "loss": 12.0573,
+      "step": 11725
+    },
+    {
+      "epoch": 2.0505792349726777,
+      "grad_norm": 0.9635983766843074,
+      "learning_rate": 4.8077731751718505e-05,
+      "loss": 12.2942,
+      "step": 11726
+    },
+    {
+      "epoch": 2.0507540983606556,
+      "grad_norm": 0.8299370596044973,
+      "learning_rate": 4.806159632548921e-05,
+      "loss": 11.7755,
+      "step": 11727
+    },
+    {
+      "epoch": 2.050928961748634,
+      "grad_norm": 0.9028512878468439,
+      "learning_rate": 4.804546275080334e-05,
+      "loss": 12.227,
+      "step": 11728
+    },
+    {
+      "epoch": 2.051103825136612,
+      "grad_norm": 0.8794017820765485,
+      "learning_rate": 4.802933102823605e-05,
+      "loss": 12.2011,
+      "step": 11729
+    },
+    {
+      "epoch": 2.05127868852459,
+      "grad_norm": 0.8551578093419351,
+      "learning_rate": 4.801320115836234e-05,
+      "loss": 12.0742,
+      "step": 11730
+    },
+    {
+      "epoch": 2.0514535519125685,
+      "grad_norm": 0.8587308311633169,
+      "learning_rate": 4.799707314175729e-05,
+      "loss": 12.2571,
+      "step": 11731
+    },
+    {
+      "epoch": 2.0516284153005464,
+      "grad_norm": 0.857349881025083,
+      "learning_rate": 4.79809469789958e-05,
+      "loss": 12.0342,
+      "step": 11732
+    },
+    {
+      "epoch": 2.0518032786885247,
+      "grad_norm": 0.8572897359804669,
+      "learning_rate": 4.796482267065283e-05,
+      "loss": 11.9887,
+      "step": 11733
+    },
+    {
+      "epoch": 2.0519781420765026,
+      "grad_norm": 0.8893772863578157,
+      "learning_rate": 4.794870021730309e-05,
+      "loss": 12.2438,
+      "step": 11734
+    },
+    {
+      "epoch": 2.052153005464481,
+      "grad_norm": 0.8576465057471664,
+      "learning_rate": 4.7932579619521366e-05,
+      "loss": 12.1192,
+      "step": 11735
+    },
+    {
+      "epoch": 2.052327868852459,
+      "grad_norm": 0.8472754085017836,
+      "learning_rate": 4.791646087788235e-05,
+      "loss": 12.1224,
+      "step": 11736
+    },
+    {
+      "epoch": 2.0525027322404372,
+      "grad_norm": 0.8271635813307388,
+      "learning_rate": 4.790034399296065e-05,
+      "loss": 12.1214,
+      "step": 11737
+    },
+    {
+      "epoch": 2.052677595628415,
+      "grad_norm": 0.8619095568776548,
+      "learning_rate": 4.788422896533081e-05,
+      "loss": 12.1647,
+      "step": 11738
+    },
+    {
+      "epoch": 2.0528524590163935,
+      "grad_norm": 0.9614002630294498,
+      "learning_rate": 4.7868115795567325e-05,
+      "loss": 11.9305,
+      "step": 11739
+    },
+    {
+      "epoch": 2.053027322404372,
+      "grad_norm": 0.9159997546447983,
+      "learning_rate": 4.7852004484244636e-05,
+      "loss": 12.1096,
+      "step": 11740
+    },
+    {
+      "epoch": 2.0532021857923497,
+      "grad_norm": 0.8561182310673668,
+      "learning_rate": 4.7835895031937025e-05,
+      "loss": 12.2478,
+      "step": 11741
+    },
+    {
+      "epoch": 2.053377049180328,
+      "grad_norm": 0.8632099545792968,
+      "learning_rate": 4.781978743921881e-05,
+      "loss": 11.7096,
+      "step": 11742
+    },
+    {
+      "epoch": 2.053551912568306,
+      "grad_norm": 0.9335030577978888,
+      "learning_rate": 4.780368170666419e-05,
+      "loss": 12.207,
+      "step": 11743
+    },
+    {
+      "epoch": 2.0537267759562843,
+      "grad_norm": 0.8323618276200071,
+      "learning_rate": 4.778757783484739e-05,
+      "loss": 12.0273,
+      "step": 11744
+    },
+    {
+      "epoch": 2.053901639344262,
+      "grad_norm": 0.807510930129049,
+      "learning_rate": 4.777147582434239e-05,
+      "loss": 12.157,
+      "step": 11745
+    },
+    {
+      "epoch": 2.0540765027322405,
+      "grad_norm": 0.918121987615914,
+      "learning_rate": 4.7755375675723256e-05,
+      "loss": 11.9454,
+      "step": 11746
+    },
+    {
+      "epoch": 2.0542513661202184,
+      "grad_norm": 0.816511848887646,
+      "learning_rate": 4.773927738956394e-05,
+      "loss": 12.3015,
+      "step": 11747
+    },
+    {
+      "epoch": 2.0544262295081968,
+      "grad_norm": 0.8068073380824923,
+      "learning_rate": 4.7723180966438365e-05,
+      "loss": 12.2606,
+      "step": 11748
+    },
+    {
+      "epoch": 2.0546010928961747,
+      "grad_norm": 0.931304095477533,
+      "learning_rate": 4.770708640692026e-05,
+      "loss": 12.1419,
+      "step": 11749
+    },
+    {
+      "epoch": 2.054775956284153,
+      "grad_norm": 0.8597103185837534,
+      "learning_rate": 4.769099371158343e-05,
+      "loss": 12.2357,
+      "step": 11750
+    },
+    {
+      "epoch": 2.0549508196721313,
+      "grad_norm": 0.869786191476617,
+      "learning_rate": 4.767490288100157e-05,
+      "loss": 11.8034,
+      "step": 11751
+    },
+    {
+      "epoch": 2.0551256830601092,
+      "grad_norm": 0.8810934813179335,
+      "learning_rate": 4.765881391574831e-05,
+      "loss": 12.323,
+      "step": 11752
+    },
+    {
+      "epoch": 2.0553005464480876,
+      "grad_norm": 0.9280282300751793,
+      "learning_rate": 4.7642726816397156e-05,
+      "loss": 12.1739,
+      "step": 11753
+    },
+    {
+      "epoch": 2.0554754098360655,
+      "grad_norm": 0.843914761010326,
+      "learning_rate": 4.76266415835216e-05,
+      "loss": 12.2017,
+      "step": 11754
+    },
+    {
+      "epoch": 2.055650273224044,
+      "grad_norm": 0.8310765266376204,
+      "learning_rate": 4.76105582176951e-05,
+      "loss": 12.1754,
+      "step": 11755
+    },
+    {
+      "epoch": 2.0558251366120217,
+      "grad_norm": 0.8584569146366896,
+      "learning_rate": 4.7594476719490976e-05,
+      "loss": 11.9163,
+      "step": 11756
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 0.8282445104908365,
+      "learning_rate": 4.757839708948253e-05,
+      "loss": 12.0224,
+      "step": 11757
+    },
+    {
+      "epoch": 2.056174863387978,
+      "grad_norm": 0.819366279576755,
+      "learning_rate": 4.7562319328242976e-05,
+      "loss": 11.9417,
+      "step": 11758
+    },
+    {
+      "epoch": 2.0563497267759563,
+      "grad_norm": 0.8168072537815648,
+      "learning_rate": 4.754624343634552e-05,
+      "loss": 12.1078,
+      "step": 11759
+    },
+    {
+      "epoch": 2.0565245901639346,
+      "grad_norm": 0.8433457015718595,
+      "learning_rate": 4.7530169414363154e-05,
+      "loss": 12.3039,
+      "step": 11760
+    },
+    {
+      "epoch": 2.0566994535519125,
+      "grad_norm": 0.832829794202146,
+      "learning_rate": 4.751409726286895e-05,
+      "loss": 12.257,
+      "step": 11761
+    },
+    {
+      "epoch": 2.056874316939891,
+      "grad_norm": 0.8640646839978708,
+      "learning_rate": 4.749802698243585e-05,
+      "loss": 12.1542,
+      "step": 11762
+    },
+    {
+      "epoch": 2.0570491803278688,
+      "grad_norm": 0.8367191421118054,
+      "learning_rate": 4.748195857363679e-05,
+      "loss": 12.1506,
+      "step": 11763
+    },
+    {
+      "epoch": 2.057224043715847,
+      "grad_norm": 0.8288678542199118,
+      "learning_rate": 4.746589203704449e-05,
+      "loss": 12.0039,
+      "step": 11764
+    },
+    {
+      "epoch": 2.057398907103825,
+      "grad_norm": 0.8753296974262795,
+      "learning_rate": 4.744982737323178e-05,
+      "loss": 12.1113,
+      "step": 11765
+    },
+    {
+      "epoch": 2.0575737704918033,
+      "grad_norm": 0.844235543543705,
+      "learning_rate": 4.7433764582771314e-05,
+      "loss": 12.3144,
+      "step": 11766
+    },
+    {
+      "epoch": 2.0577486338797812,
+      "grad_norm": 0.9431764048601471,
+      "learning_rate": 4.741770366623577e-05,
+      "loss": 12.327,
+      "step": 11767
+    },
+    {
+      "epoch": 2.0579234972677596,
+      "grad_norm": 0.8447140901342387,
+      "learning_rate": 4.7401644624197604e-05,
+      "loss": 12.0413,
+      "step": 11768
+    },
+    {
+      "epoch": 2.0580983606557375,
+      "grad_norm": 0.8206908469749725,
+      "learning_rate": 4.738558745722936e-05,
+      "loss": 12.0775,
+      "step": 11769
+    },
+    {
+      "epoch": 2.058273224043716,
+      "grad_norm": 0.8590190870876443,
+      "learning_rate": 4.7369532165903494e-05,
+      "loss": 12.175,
+      "step": 11770
+    },
+    {
+      "epoch": 2.058448087431694,
+      "grad_norm": 0.7962556403742745,
+      "learning_rate": 4.735347875079227e-05,
+      "loss": 11.9645,
+      "step": 11771
+    },
+    {
+      "epoch": 2.058622950819672,
+      "grad_norm": 0.7709698749296537,
+      "learning_rate": 4.733742721246802e-05,
+      "loss": 12.0775,
+      "step": 11772
+    },
+    {
+      "epoch": 2.0587978142076504,
+      "grad_norm": 0.9020346547472197,
+      "learning_rate": 4.732137755150296e-05,
+      "loss": 12.1135,
+      "step": 11773
+    },
+    {
+      "epoch": 2.0589726775956283,
+      "grad_norm": 0.9858727518901752,
+      "learning_rate": 4.730532976846924e-05,
+      "loss": 12.0821,
+      "step": 11774
+    },
+    {
+      "epoch": 2.0591475409836066,
+      "grad_norm": 0.7806546431807486,
+      "learning_rate": 4.728928386393895e-05,
+      "loss": 12.0348,
+      "step": 11775
+    },
+    {
+      "epoch": 2.0593224043715845,
+      "grad_norm": 0.8006049278755726,
+      "learning_rate": 4.727323983848411e-05,
+      "loss": 12.1599,
+      "step": 11776
+    },
+    {
+      "epoch": 2.059497267759563,
+      "grad_norm": 0.8436131862952752,
+      "learning_rate": 4.7257197692676646e-05,
+      "loss": 12.1803,
+      "step": 11777
+    },
+    {
+      "epoch": 2.0596721311475408,
+      "grad_norm": 0.8201126945216597,
+      "learning_rate": 4.724115742708851e-05,
+      "loss": 12.0144,
+      "step": 11778
+    },
+    {
+      "epoch": 2.059846994535519,
+      "grad_norm": 0.9147450411793533,
+      "learning_rate": 4.7225119042291435e-05,
+      "loss": 12.2948,
+      "step": 11779
+    },
+    {
+      "epoch": 2.0600218579234975,
+      "grad_norm": 0.8804103113922062,
+      "learning_rate": 4.72090825388572e-05,
+      "loss": 11.8811,
+      "step": 11780
+    },
+    {
+      "epoch": 2.0601967213114754,
+      "grad_norm": 1.0102315488465343,
+      "learning_rate": 4.7193047917357494e-05,
+      "loss": 11.9956,
+      "step": 11781
+    },
+    {
+      "epoch": 2.0603715846994537,
+      "grad_norm": 0.8560286495933473,
+      "learning_rate": 4.717701517836397e-05,
+      "loss": 12.0419,
+      "step": 11782
+    },
+    {
+      "epoch": 2.0605464480874316,
+      "grad_norm": 0.8638756037830033,
+      "learning_rate": 4.71609843224481e-05,
+      "loss": 12.1948,
+      "step": 11783
+    },
+    {
+      "epoch": 2.06072131147541,
+      "grad_norm": 0.8854229848795575,
+      "learning_rate": 4.7144955350181405e-05,
+      "loss": 11.8741,
+      "step": 11784
+    },
+    {
+      "epoch": 2.060896174863388,
+      "grad_norm": 0.8540444659555694,
+      "learning_rate": 4.712892826213533e-05,
+      "loss": 12.245,
+      "step": 11785
+    },
+    {
+      "epoch": 2.061071038251366,
+      "grad_norm": 0.8637664242495906,
+      "learning_rate": 4.7112903058881166e-05,
+      "loss": 11.9614,
+      "step": 11786
+    },
+    {
+      "epoch": 2.061245901639344,
+      "grad_norm": 0.8581025296418885,
+      "learning_rate": 4.70968797409902e-05,
+      "loss": 12.329,
+      "step": 11787
+    },
+    {
+      "epoch": 2.0614207650273224,
+      "grad_norm": 0.8629255551213562,
+      "learning_rate": 4.708085830903366e-05,
+      "loss": 12.0139,
+      "step": 11788
+    },
+    {
+      "epoch": 2.0615956284153008,
+      "grad_norm": 0.9008204636815625,
+      "learning_rate": 4.706483876358274e-05,
+      "loss": 12.0094,
+      "step": 11789
+    },
+    {
+      "epoch": 2.0617704918032786,
+      "grad_norm": 0.9105462169334351,
+      "learning_rate": 4.704882110520843e-05,
+      "loss": 12.1263,
+      "step": 11790
+    },
+    {
+      "epoch": 2.061945355191257,
+      "grad_norm": 0.9585512418751968,
+      "learning_rate": 4.703280533448178e-05,
+      "loss": 11.9565,
+      "step": 11791
+    },
+    {
+      "epoch": 2.062120218579235,
+      "grad_norm": 0.9178722603575272,
+      "learning_rate": 4.701679145197373e-05,
+      "loss": 11.958,
+      "step": 11792
+    },
+    {
+      "epoch": 2.0622950819672132,
+      "grad_norm": 0.841214667830272,
+      "learning_rate": 4.700077945825516e-05,
+      "loss": 12.1922,
+      "step": 11793
+    },
+    {
+      "epoch": 2.062469945355191,
+      "grad_norm": 0.8883608665596968,
+      "learning_rate": 4.6984769353896884e-05,
+      "loss": 12.1942,
+      "step": 11794
+    },
+    {
+      "epoch": 2.0626448087431695,
+      "grad_norm": 0.8672446818767255,
+      "learning_rate": 4.696876113946963e-05,
+      "loss": 12.0929,
+      "step": 11795
+    },
+    {
+      "epoch": 2.0628196721311474,
+      "grad_norm": 0.8538547929341738,
+      "learning_rate": 4.69527548155441e-05,
+      "loss": 12.1358,
+      "step": 11796
+    },
+    {
+      "epoch": 2.0629945355191257,
+      "grad_norm": 0.8975981561135465,
+      "learning_rate": 4.69367503826909e-05,
+      "loss": 12.083,
+      "step": 11797
+    },
+    {
+      "epoch": 2.063169398907104,
+      "grad_norm": 0.8112890330123689,
+      "learning_rate": 4.692074784148052e-05,
+      "loss": 11.964,
+      "step": 11798
+    },
+    {
+      "epoch": 2.063344262295082,
+      "grad_norm": 0.8839446997139307,
+      "learning_rate": 4.690474719248347e-05,
+      "loss": 12.3218,
+      "step": 11799
+    },
+    {
+      "epoch": 2.0635191256830603,
+      "grad_norm": 0.7853017959692234,
+      "learning_rate": 4.688874843627018e-05,
+      "loss": 12.1908,
+      "step": 11800
+    },
+    {
+      "epoch": 2.063693989071038,
+      "grad_norm": 0.9409474833005593,
+      "learning_rate": 4.687275157341091e-05,
+      "loss": 12.023,
+      "step": 11801
+    },
+    {
+      "epoch": 2.0638688524590165,
+      "grad_norm": 0.8848575795353368,
+      "learning_rate": 4.6856756604475984e-05,
+      "loss": 12.0907,
+      "step": 11802
+    },
+    {
+      "epoch": 2.0640437158469944,
+      "grad_norm": 0.8972172770497159,
+      "learning_rate": 4.684076353003559e-05,
+      "loss": 12.1395,
+      "step": 11803
+    },
+    {
+      "epoch": 2.0642185792349728,
+      "grad_norm": 0.9209939437607145,
+      "learning_rate": 4.682477235065989e-05,
+      "loss": 12.0509,
+      "step": 11804
+    },
+    {
+      "epoch": 2.0643934426229507,
+      "grad_norm": 0.9035835773772734,
+      "learning_rate": 4.6808783066918915e-05,
+      "loss": 11.8353,
+      "step": 11805
+    },
+    {
+      "epoch": 2.064568306010929,
+      "grad_norm": 0.9242467455104891,
+      "learning_rate": 4.679279567938266e-05,
+      "loss": 12.077,
+      "step": 11806
+    },
+    {
+      "epoch": 2.064743169398907,
+      "grad_norm": 0.9298351282454711,
+      "learning_rate": 4.677681018862108e-05,
+      "loss": 12.0259,
+      "step": 11807
+    },
+    {
+      "epoch": 2.0649180327868852,
+      "grad_norm": 0.8524589468907634,
+      "learning_rate": 4.676082659520407e-05,
+      "loss": 12.1276,
+      "step": 11808
+    },
+    {
+      "epoch": 2.0650928961748636,
+      "grad_norm": 0.8312559819474196,
+      "learning_rate": 4.674484489970135e-05,
+      "loss": 12.2566,
+      "step": 11809
+    },
+    {
+      "epoch": 2.0652677595628415,
+      "grad_norm": 0.8927357369150666,
+      "learning_rate": 4.6728865102682684e-05,
+      "loss": 12.1241,
+      "step": 11810
+    },
+    {
+      "epoch": 2.06544262295082,
+      "grad_norm": 0.8243814906952329,
+      "learning_rate": 4.6712887204717735e-05,
+      "loss": 12.1758,
+      "step": 11811
+    },
+    {
+      "epoch": 2.0656174863387977,
+      "grad_norm": 0.8255314187481771,
+      "learning_rate": 4.66969112063761e-05,
+      "loss": 12.1347,
+      "step": 11812
+    },
+    {
+      "epoch": 2.065792349726776,
+      "grad_norm": 0.8269624811255227,
+      "learning_rate": 4.668093710822731e-05,
+      "loss": 12.29,
+      "step": 11813
+    },
+    {
+      "epoch": 2.065967213114754,
+      "grad_norm": 1.0024102828463395,
+      "learning_rate": 4.66649649108408e-05,
+      "loss": 12.1884,
+      "step": 11814
+    },
+    {
+      "epoch": 2.0661420765027323,
+      "grad_norm": 0.8428682023017623,
+      "learning_rate": 4.664899461478599e-05,
+      "loss": 11.9105,
+      "step": 11815
+    },
+    {
+      "epoch": 2.06631693989071,
+      "grad_norm": 0.7973502412763976,
+      "learning_rate": 4.663302622063221e-05,
+      "loss": 11.8394,
+      "step": 11816
+    },
+    {
+      "epoch": 2.0664918032786885,
+      "grad_norm": 0.8557955439856201,
+      "learning_rate": 4.661705972894867e-05,
+      "loss": 11.8723,
+      "step": 11817
+    },
+    {
+      "epoch": 2.066666666666667,
+      "grad_norm": 0.7854232618185207,
+      "learning_rate": 4.660109514030459e-05,
+      "loss": 12.164,
+      "step": 11818
+    },
+    {
+      "epoch": 2.0668415300546448,
+      "grad_norm": 0.8299933719518354,
+      "learning_rate": 4.6585132455269086e-05,
+      "loss": 12.1736,
+      "step": 11819
+    },
+    {
+      "epoch": 2.067016393442623,
+      "grad_norm": 0.8487535234474364,
+      "learning_rate": 4.6569171674411185e-05,
+      "loss": 12.0452,
+      "step": 11820
+    },
+    {
+      "epoch": 2.067191256830601,
+      "grad_norm": 0.860938325979823,
+      "learning_rate": 4.6553212798299875e-05,
+      "loss": 12.058,
+      "step": 11821
+    },
+    {
+      "epoch": 2.0673661202185794,
+      "grad_norm": 0.8356676858839188,
+      "learning_rate": 4.653725582750409e-05,
+      "loss": 11.8985,
+      "step": 11822
+    },
+    {
+      "epoch": 2.0675409836065572,
+      "grad_norm": 0.8709705018065415,
+      "learning_rate": 4.65213007625927e-05,
+      "loss": 12.2463,
+      "step": 11823
+    },
+    {
+      "epoch": 2.0677158469945356,
+      "grad_norm": 0.904190220403658,
+      "learning_rate": 4.650534760413443e-05,
+      "loss": 12.115,
+      "step": 11824
+    },
+    {
+      "epoch": 2.0678907103825135,
+      "grad_norm": 1.322610933404603,
+      "learning_rate": 4.648939635269801e-05,
+      "loss": 12.1617,
+      "step": 11825
+    },
+    {
+      "epoch": 2.068065573770492,
+      "grad_norm": 0.8349652006883865,
+      "learning_rate": 4.647344700885209e-05,
+      "loss": 12.1516,
+      "step": 11826
+    },
+    {
+      "epoch": 2.0682404371584697,
+      "grad_norm": 0.944389772791108,
+      "learning_rate": 4.6457499573165245e-05,
+      "loss": 12.1366,
+      "step": 11827
+    },
+    {
+      "epoch": 2.068415300546448,
+      "grad_norm": 0.8246223284305755,
+      "learning_rate": 4.644155404620597e-05,
+      "loss": 12.0932,
+      "step": 11828
+    },
+    {
+      "epoch": 2.0685901639344264,
+      "grad_norm": 0.8474265091879616,
+      "learning_rate": 4.642561042854273e-05,
+      "loss": 12.0249,
+      "step": 11829
+    },
+    {
+      "epoch": 2.0687650273224043,
+      "grad_norm": 0.8870069502152348,
+      "learning_rate": 4.640966872074388e-05,
+      "loss": 12.0715,
+      "step": 11830
+    },
+    {
+      "epoch": 2.0689398907103826,
+      "grad_norm": 0.8406473035735443,
+      "learning_rate": 4.639372892337777e-05,
+      "loss": 12.1086,
+      "step": 11831
+    },
+    {
+      "epoch": 2.0691147540983605,
+      "grad_norm": 0.9360266699778828,
+      "learning_rate": 4.637779103701255e-05,
+      "loss": 11.9595,
+      "step": 11832
+    },
+    {
+      "epoch": 2.069289617486339,
+      "grad_norm": 0.8690916250875389,
+      "learning_rate": 4.636185506221643e-05,
+      "loss": 12.1243,
+      "step": 11833
+    },
+    {
+      "epoch": 2.069464480874317,
+      "grad_norm": 0.783124982065805,
+      "learning_rate": 4.6345920999557534e-05,
+      "loss": 11.909,
+      "step": 11834
+    },
+    {
+      "epoch": 2.069639344262295,
+      "grad_norm": 0.8554257734626515,
+      "learning_rate": 4.632998884960385e-05,
+      "loss": 12.0893,
+      "step": 11835
+    },
+    {
+      "epoch": 2.069814207650273,
+      "grad_norm": 0.8401946135478692,
+      "learning_rate": 4.6314058612923337e-05,
+      "loss": 11.9389,
+      "step": 11836
+    },
+    {
+      "epoch": 2.0699890710382514,
+      "grad_norm": 0.9158215642890853,
+      "learning_rate": 4.629813029008392e-05,
+      "loss": 12.2346,
+      "step": 11837
+    },
+    {
+      "epoch": 2.0701639344262297,
+      "grad_norm": 0.8809113794078139,
+      "learning_rate": 4.628220388165345e-05,
+      "loss": 11.7844,
+      "step": 11838
+    },
+    {
+      "epoch": 2.0703387978142076,
+      "grad_norm": 0.8394265393256718,
+      "learning_rate": 4.626627938819961e-05,
+      "loss": 12.0663,
+      "step": 11839
+    },
+    {
+      "epoch": 2.070513661202186,
+      "grad_norm": 0.8156597202081676,
+      "learning_rate": 4.625035681029013e-05,
+      "loss": 11.6936,
+      "step": 11840
+    },
+    {
+      "epoch": 2.070688524590164,
+      "grad_norm": 0.8862769656649763,
+      "learning_rate": 4.623443614849263e-05,
+      "loss": 11.9158,
+      "step": 11841
+    },
+    {
+      "epoch": 2.070863387978142,
+      "grad_norm": 0.8262460623767074,
+      "learning_rate": 4.621851740337469e-05,
+      "loss": 12.1223,
+      "step": 11842
+    },
+    {
+      "epoch": 2.07103825136612,
+      "grad_norm": 0.9386546159781398,
+      "learning_rate": 4.620260057550375e-05,
+      "loss": 12.1339,
+      "step": 11843
+    },
+    {
+      "epoch": 2.0712131147540984,
+      "grad_norm": 0.8862622613163298,
+      "learning_rate": 4.6186685665447225e-05,
+      "loss": 11.9786,
+      "step": 11844
+    },
+    {
+      "epoch": 2.0713879781420763,
+      "grad_norm": 0.8461899456570282,
+      "learning_rate": 4.61707726737725e-05,
+      "loss": 12.0284,
+      "step": 11845
+    },
+    {
+      "epoch": 2.0715628415300547,
+      "grad_norm": 0.977036550356543,
+      "learning_rate": 4.615486160104683e-05,
+      "loss": 12.0938,
+      "step": 11846
+    },
+    {
+      "epoch": 2.071737704918033,
+      "grad_norm": 0.8968998878865806,
+      "learning_rate": 4.613895244783743e-05,
+      "loss": 12.2412,
+      "step": 11847
+    },
+    {
+      "epoch": 2.071912568306011,
+      "grad_norm": 0.8091937043243969,
+      "learning_rate": 4.6123045214711436e-05,
+      "loss": 12.2818,
+      "step": 11848
+    },
+    {
+      "epoch": 2.0720874316939892,
+      "grad_norm": 0.8042460917199047,
+      "learning_rate": 4.6107139902235986e-05,
+      "loss": 12.1317,
+      "step": 11849
+    },
+    {
+      "epoch": 2.072262295081967,
+      "grad_norm": 0.8294368642011385,
+      "learning_rate": 4.6091236510977974e-05,
+      "loss": 11.9523,
+      "step": 11850
+    },
+    {
+      "epoch": 2.0724371584699455,
+      "grad_norm": 0.8491189413963667,
+      "learning_rate": 4.607533504150441e-05,
+      "loss": 11.7553,
+      "step": 11851
+    },
+    {
+      "epoch": 2.0726120218579234,
+      "grad_norm": 0.8730636419972932,
+      "learning_rate": 4.6059435494382154e-05,
+      "loss": 11.9419,
+      "step": 11852
+    },
+    {
+      "epoch": 2.0727868852459017,
+      "grad_norm": 0.8742599325070947,
+      "learning_rate": 4.604353787017802e-05,
+      "loss": 11.8978,
+      "step": 11853
+    },
+    {
+      "epoch": 2.0729617486338796,
+      "grad_norm": 0.8389602342958057,
+      "learning_rate": 4.602764216945869e-05,
+      "loss": 12.1884,
+      "step": 11854
+    },
+    {
+      "epoch": 2.073136612021858,
+      "grad_norm": 0.8469118920514478,
+      "learning_rate": 4.601174839279085e-05,
+      "loss": 12.0533,
+      "step": 11855
+    },
+    {
+      "epoch": 2.0733114754098363,
+      "grad_norm": 0.8641143791361385,
+      "learning_rate": 4.599585654074111e-05,
+      "loss": 11.8874,
+      "step": 11856
+    },
+    {
+      "epoch": 2.073486338797814,
+      "grad_norm": 0.9605160197350736,
+      "learning_rate": 4.5979966613876026e-05,
+      "loss": 12.2026,
+      "step": 11857
+    },
+    {
+      "epoch": 2.0736612021857925,
+      "grad_norm": 0.891889744934542,
+      "learning_rate": 4.596407861276197e-05,
+      "loss": 12.1623,
+      "step": 11858
+    },
+    {
+      "epoch": 2.0738360655737704,
+      "grad_norm": 0.904028367628193,
+      "learning_rate": 4.59481925379654e-05,
+      "loss": 12.2201,
+      "step": 11859
+    },
+    {
+      "epoch": 2.0740109289617488,
+      "grad_norm": 0.8519102582138899,
+      "learning_rate": 4.593230839005259e-05,
+      "loss": 12.1993,
+      "step": 11860
+    },
+    {
+      "epoch": 2.0741857923497267,
+      "grad_norm": 0.824190518110466,
+      "learning_rate": 4.591642616958986e-05,
+      "loss": 12.1297,
+      "step": 11861
+    },
+    {
+      "epoch": 2.074360655737705,
+      "grad_norm": 0.8063659419841226,
+      "learning_rate": 4.590054587714331e-05,
+      "loss": 11.9765,
+      "step": 11862
+    },
+    {
+      "epoch": 2.074535519125683,
+      "grad_norm": 0.8025460576471501,
+      "learning_rate": 4.588466751327908e-05,
+      "loss": 12.1579,
+      "step": 11863
+    },
+    {
+      "epoch": 2.0747103825136612,
+      "grad_norm": 0.8913343956336299,
+      "learning_rate": 4.586879107856323e-05,
+      "loss": 12.279,
+      "step": 11864
+    },
+    {
+      "epoch": 2.074885245901639,
+      "grad_norm": 0.8378555770281966,
+      "learning_rate": 4.5852916573561746e-05,
+      "loss": 12.1784,
+      "step": 11865
+    },
+    {
+      "epoch": 2.0750601092896175,
+      "grad_norm": 0.8889694774255978,
+      "learning_rate": 4.583704399884051e-05,
+      "loss": 11.9422,
+      "step": 11866
+    },
+    {
+      "epoch": 2.075234972677596,
+      "grad_norm": 0.822353084175967,
+      "learning_rate": 4.5821173354965386e-05,
+      "loss": 12.0548,
+      "step": 11867
+    },
+    {
+      "epoch": 2.0754098360655737,
+      "grad_norm": 0.8667626354728437,
+      "learning_rate": 4.5805304642502156e-05,
+      "loss": 11.982,
+      "step": 11868
+    },
+    {
+      "epoch": 2.075584699453552,
+      "grad_norm": 0.8169362460582144,
+      "learning_rate": 4.578943786201646e-05,
+      "loss": 12.1183,
+      "step": 11869
+    },
+    {
+      "epoch": 2.07575956284153,
+      "grad_norm": 0.8348584003777162,
+      "learning_rate": 4.5773573014073965e-05,
+      "loss": 12.1614,
+      "step": 11870
+    },
+    {
+      "epoch": 2.0759344262295083,
+      "grad_norm": 0.8830802469660812,
+      "learning_rate": 4.5757710099240235e-05,
+      "loss": 11.9399,
+      "step": 11871
+    },
+    {
+      "epoch": 2.076109289617486,
+      "grad_norm": 0.8256834573668682,
+      "learning_rate": 4.57418491180808e-05,
+      "loss": 11.8537,
+      "step": 11872
+    },
+    {
+      "epoch": 2.0762841530054645,
+      "grad_norm": 0.8526754604999347,
+      "learning_rate": 4.572599007116103e-05,
+      "loss": 12.1706,
+      "step": 11873
+    },
+    {
+      "epoch": 2.0764590163934424,
+      "grad_norm": 0.8143458347939364,
+      "learning_rate": 4.571013295904628e-05,
+      "loss": 12.1479,
+      "step": 11874
+    },
+    {
+      "epoch": 2.0766338797814208,
+      "grad_norm": 0.9152765779178229,
+      "learning_rate": 4.5694277782301874e-05,
+      "loss": 11.8636,
+      "step": 11875
+    },
+    {
+      "epoch": 2.076808743169399,
+      "grad_norm": 0.9169743987526647,
+      "learning_rate": 4.567842454149305e-05,
+      "loss": 12.1559,
+      "step": 11876
+    },
+    {
+      "epoch": 2.076983606557377,
+      "grad_norm": 0.9126428837198305,
+      "learning_rate": 4.5662573237184883e-05,
+      "loss": 11.9738,
+      "step": 11877
+    },
+    {
+      "epoch": 2.0771584699453554,
+      "grad_norm": 0.9442089782410408,
+      "learning_rate": 4.5646723869942486e-05,
+      "loss": 11.9432,
+      "step": 11878
+    },
+    {
+      "epoch": 2.0773333333333333,
+      "grad_norm": 0.8223782688057631,
+      "learning_rate": 4.563087644033093e-05,
+      "loss": 12.0581,
+      "step": 11879
+    },
+    {
+      "epoch": 2.0775081967213116,
+      "grad_norm": 0.9994715058262906,
+      "learning_rate": 4.561503094891506e-05,
+      "loss": 12.2868,
+      "step": 11880
+    },
+    {
+      "epoch": 2.0776830601092895,
+      "grad_norm": 0.8158238341107303,
+      "learning_rate": 4.5599187396259793e-05,
+      "loss": 12.1262,
+      "step": 11881
+    },
+    {
+      "epoch": 2.077857923497268,
+      "grad_norm": 0.9008441849380452,
+      "learning_rate": 4.558334578292993e-05,
+      "loss": 12.1767,
+      "step": 11882
+    },
+    {
+      "epoch": 2.0780327868852457,
+      "grad_norm": 0.8617610455898472,
+      "learning_rate": 4.556750610949022e-05,
+      "loss": 12.1132,
+      "step": 11883
+    },
+    {
+      "epoch": 2.078207650273224,
+      "grad_norm": 0.8966006311738725,
+      "learning_rate": 4.555166837650532e-05,
+      "loss": 12.15,
+      "step": 11884
+    },
+    {
+      "epoch": 2.078382513661202,
+      "grad_norm": 0.9921726673272249,
+      "learning_rate": 4.553583258453983e-05,
+      "loss": 11.9497,
+      "step": 11885
+    },
+    {
+      "epoch": 2.0785573770491803,
+      "grad_norm": 0.7987184633129951,
+      "learning_rate": 4.5519998734158264e-05,
+      "loss": 12.0264,
+      "step": 11886
+    },
+    {
+      "epoch": 2.0787322404371587,
+      "grad_norm": 0.9352763103546535,
+      "learning_rate": 4.5504166825925134e-05,
+      "loss": 11.93,
+      "step": 11887
+    },
+    {
+      "epoch": 2.0789071038251365,
+      "grad_norm": 0.8846624556930434,
+      "learning_rate": 4.5488336860404746e-05,
+      "loss": 12.008,
+      "step": 11888
+    },
+    {
+      "epoch": 2.079081967213115,
+      "grad_norm": 0.9999945177974814,
+      "learning_rate": 4.5472508838161453e-05,
+      "loss": 11.8252,
+      "step": 11889
+    },
+    {
+      "epoch": 2.079256830601093,
+      "grad_norm": 0.8748458694985315,
+      "learning_rate": 4.5456682759759516e-05,
+      "loss": 11.9968,
+      "step": 11890
+    },
+    {
+      "epoch": 2.079431693989071,
+      "grad_norm": 0.9208400948329296,
+      "learning_rate": 4.544085862576315e-05,
+      "loss": 12.269,
+      "step": 11891
+    },
+    {
+      "epoch": 2.079606557377049,
+      "grad_norm": 0.799958748288513,
+      "learning_rate": 4.542503643673639e-05,
+      "loss": 12.0509,
+      "step": 11892
+    },
+    {
+      "epoch": 2.0797814207650274,
+      "grad_norm": 0.8332141688951814,
+      "learning_rate": 4.540921619324332e-05,
+      "loss": 12.4082,
+      "step": 11893
+    },
+    {
+      "epoch": 2.0799562841530053,
+      "grad_norm": 0.900170655781252,
+      "learning_rate": 4.539339789584795e-05,
+      "loss": 11.8121,
+      "step": 11894
+    },
+    {
+      "epoch": 2.0801311475409836,
+      "grad_norm": 0.9311780694422042,
+      "learning_rate": 4.53775815451141e-05,
+      "loss": 12.153,
+      "step": 11895
+    },
+    {
+      "epoch": 2.080306010928962,
+      "grad_norm": 0.897142710731805,
+      "learning_rate": 4.5361767141605646e-05,
+      "loss": 11.985,
+      "step": 11896
+    },
+    {
+      "epoch": 2.08048087431694,
+      "grad_norm": 0.8621369776087975,
+      "learning_rate": 4.534595468588636e-05,
+      "loss": 11.972,
+      "step": 11897
+    },
+    {
+      "epoch": 2.080655737704918,
+      "grad_norm": 0.8340550829086899,
+      "learning_rate": 4.5330144178519975e-05,
+      "loss": 12.3924,
+      "step": 11898
+    },
+    {
+      "epoch": 2.080830601092896,
+      "grad_norm": 0.7805296707905417,
+      "learning_rate": 4.5314335620070025e-05,
+      "loss": 12.0237,
+      "step": 11899
+    },
+    {
+      "epoch": 2.0810054644808744,
+      "grad_norm": 0.8703870643753021,
+      "learning_rate": 4.5298529011100136e-05,
+      "loss": 11.9093,
+      "step": 11900
+    },
+    {
+      "epoch": 2.0811803278688523,
+      "grad_norm": 0.8021374127559514,
+      "learning_rate": 4.528272435217375e-05,
+      "loss": 11.9974,
+      "step": 11901
+    },
+    {
+      "epoch": 2.0813551912568307,
+      "grad_norm": 0.858367258712762,
+      "learning_rate": 4.526692164385433e-05,
+      "loss": 11.8696,
+      "step": 11902
+    },
+    {
+      "epoch": 2.0815300546448086,
+      "grad_norm": 0.9505547587441526,
+      "learning_rate": 4.52511208867052e-05,
+      "loss": 12.2265,
+      "step": 11903
+    },
+    {
+      "epoch": 2.081704918032787,
+      "grad_norm": 0.8770300920469363,
+      "learning_rate": 4.523532208128964e-05,
+      "loss": 12.1962,
+      "step": 11904
+    },
+    {
+      "epoch": 2.0818797814207652,
+      "grad_norm": 0.7892656041428492,
+      "learning_rate": 4.521952522817087e-05,
+      "loss": 11.9762,
+      "step": 11905
+    },
+    {
+      "epoch": 2.082054644808743,
+      "grad_norm": 0.7975612700967817,
+      "learning_rate": 4.520373032791205e-05,
+      "loss": 12.1646,
+      "step": 11906
+    },
+    {
+      "epoch": 2.0822295081967215,
+      "grad_norm": 0.8622461937776966,
+      "learning_rate": 4.5187937381076195e-05,
+      "loss": 12.0244,
+      "step": 11907
+    },
+    {
+      "epoch": 2.0824043715846994,
+      "grad_norm": 0.8913641827066242,
+      "learning_rate": 4.517214638822632e-05,
+      "loss": 12.0006,
+      "step": 11908
+    },
+    {
+      "epoch": 2.0825792349726777,
+      "grad_norm": 0.7701713786792563,
+      "learning_rate": 4.5156357349925414e-05,
+      "loss": 12.0788,
+      "step": 11909
+    },
+    {
+      "epoch": 2.0827540983606556,
+      "grad_norm": 0.8205128597259057,
+      "learning_rate": 4.514057026673625e-05,
+      "loss": 11.8326,
+      "step": 11910
+    },
+    {
+      "epoch": 2.082928961748634,
+      "grad_norm": 0.8389413747012539,
+      "learning_rate": 4.5124785139221646e-05,
+      "loss": 12.137,
+      "step": 11911
+    },
+    {
+      "epoch": 2.083103825136612,
+      "grad_norm": 0.9533814343184058,
+      "learning_rate": 4.510900196794435e-05,
+      "loss": 12.091,
+      "step": 11912
+    },
+    {
+      "epoch": 2.08327868852459,
+      "grad_norm": 0.9816628845756077,
+      "learning_rate": 4.509322075346704e-05,
+      "loss": 12.012,
+      "step": 11913
+    },
+    {
+      "epoch": 2.083453551912568,
+      "grad_norm": 0.9748365787153827,
+      "learning_rate": 4.5077441496352215e-05,
+      "loss": 12.1871,
+      "step": 11914
+    },
+    {
+      "epoch": 2.0836284153005464,
+      "grad_norm": 0.8018399284075753,
+      "learning_rate": 4.5061664197162433e-05,
+      "loss": 12.0212,
+      "step": 11915
+    },
+    {
+      "epoch": 2.0838032786885248,
+      "grad_norm": 0.9101180085773096,
+      "learning_rate": 4.504588885646014e-05,
+      "loss": 12.3027,
+      "step": 11916
+    },
+    {
+      "epoch": 2.0839781420765027,
+      "grad_norm": 0.9153796574020454,
+      "learning_rate": 4.503011547480769e-05,
+      "loss": 12.1011,
+      "step": 11917
+    },
+    {
+      "epoch": 2.084153005464481,
+      "grad_norm": 0.8212627214320157,
+      "learning_rate": 4.501434405276741e-05,
+      "loss": 12.1066,
+      "step": 11918
+    },
+    {
+      "epoch": 2.084327868852459,
+      "grad_norm": 0.8474325460618247,
+      "learning_rate": 4.499857459090151e-05,
+      "loss": 11.8937,
+      "step": 11919
+    },
+    {
+      "epoch": 2.0845027322404373,
+      "grad_norm": 0.816924098253567,
+      "learning_rate": 4.4982807089772175e-05,
+      "loss": 12.1574,
+      "step": 11920
+    },
+    {
+      "epoch": 2.084677595628415,
+      "grad_norm": 0.8759918966911758,
+      "learning_rate": 4.496704154994151e-05,
+      "loss": 12.1993,
+      "step": 11921
+    },
+    {
+      "epoch": 2.0848524590163935,
+      "grad_norm": 0.761068628087142,
+      "learning_rate": 4.495127797197148e-05,
+      "loss": 12.0906,
+      "step": 11922
+    },
+    {
+      "epoch": 2.0850273224043714,
+      "grad_norm": 0.8381386973206605,
+      "learning_rate": 4.493551635642408e-05,
+      "loss": 12.0347,
+      "step": 11923
+    },
+    {
+      "epoch": 2.0852021857923497,
+      "grad_norm": 0.8335088476860991,
+      "learning_rate": 4.491975670386118e-05,
+      "loss": 12.1999,
+      "step": 11924
+    },
+    {
+      "epoch": 2.085377049180328,
+      "grad_norm": 0.8293694129450234,
+      "learning_rate": 4.4903999014844645e-05,
+      "loss": 12.2012,
+      "step": 11925
+    },
+    {
+      "epoch": 2.085551912568306,
+      "grad_norm": 0.8875145680472445,
+      "learning_rate": 4.4888243289936124e-05,
+      "loss": 12.0876,
+      "step": 11926
+    },
+    {
+      "epoch": 2.0857267759562843,
+      "grad_norm": 0.8343836768160657,
+      "learning_rate": 4.487248952969735e-05,
+      "loss": 11.8696,
+      "step": 11927
+    },
+    {
+      "epoch": 2.085901639344262,
+      "grad_norm": 0.8425925218922935,
+      "learning_rate": 4.4856737734689946e-05,
+      "loss": 11.9338,
+      "step": 11928
+    },
+    {
+      "epoch": 2.0860765027322405,
+      "grad_norm": 0.8582413840135041,
+      "learning_rate": 4.484098790547537e-05,
+      "loss": 12.2768,
+      "step": 11929
+    },
+    {
+      "epoch": 2.0862513661202184,
+      "grad_norm": 0.8872163348289575,
+      "learning_rate": 4.4825240042615144e-05,
+      "loss": 12.1328,
+      "step": 11930
+    },
+    {
+      "epoch": 2.086426229508197,
+      "grad_norm": 0.8771981117181378,
+      "learning_rate": 4.480949414667064e-05,
+      "loss": 12.038,
+      "step": 11931
+    },
+    {
+      "epoch": 2.0866010928961747,
+      "grad_norm": 0.8694262364322919,
+      "learning_rate": 4.479375021820321e-05,
+      "loss": 11.7092,
+      "step": 11932
+    },
+    {
+      "epoch": 2.086775956284153,
+      "grad_norm": 0.940491485725119,
+      "learning_rate": 4.477800825777406e-05,
+      "loss": 11.8834,
+      "step": 11933
+    },
+    {
+      "epoch": 2.086950819672131,
+      "grad_norm": 0.8253499120225924,
+      "learning_rate": 4.4762268265944394e-05,
+      "loss": 11.9767,
+      "step": 11934
+    },
+    {
+      "epoch": 2.0871256830601093,
+      "grad_norm": 0.8479508215012265,
+      "learning_rate": 4.474653024327532e-05,
+      "loss": 12.1695,
+      "step": 11935
+    },
+    {
+      "epoch": 2.0873005464480876,
+      "grad_norm": 0.8410872743579668,
+      "learning_rate": 4.4730794190327896e-05,
+      "loss": 11.989,
+      "step": 11936
+    },
+    {
+      "epoch": 2.0874754098360655,
+      "grad_norm": 0.9555190780392878,
+      "learning_rate": 4.471506010766308e-05,
+      "loss": 11.7937,
+      "step": 11937
+    },
+    {
+      "epoch": 2.087650273224044,
+      "grad_norm": 0.9017232787075581,
+      "learning_rate": 4.469932799584178e-05,
+      "loss": 12.1548,
+      "step": 11938
+    },
+    {
+      "epoch": 2.0878251366120217,
+      "grad_norm": 0.9275314829903726,
+      "learning_rate": 4.468359785542482e-05,
+      "loss": 11.9309,
+      "step": 11939
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 0.8008081476323591,
+      "learning_rate": 4.4667869686972995e-05,
+      "loss": 11.9737,
+      "step": 11940
+    },
+    {
+      "epoch": 2.088174863387978,
+      "grad_norm": 0.868351753396977,
+      "learning_rate": 4.465214349104694e-05,
+      "loss": 12.1348,
+      "step": 11941
+    },
+    {
+      "epoch": 2.0883497267759563,
+      "grad_norm": 0.8089997957869987,
+      "learning_rate": 4.463641926820731e-05,
+      "loss": 12.1491,
+      "step": 11942
+    },
+    {
+      "epoch": 2.088524590163934,
+      "grad_norm": 0.9083410946593738,
+      "learning_rate": 4.4620697019014666e-05,
+      "loss": 12.2224,
+      "step": 11943
+    },
+    {
+      "epoch": 2.0886994535519126,
+      "grad_norm": 0.8033905488816815,
+      "learning_rate": 4.460497674402945e-05,
+      "loss": 12.0852,
+      "step": 11944
+    },
+    {
+      "epoch": 2.088874316939891,
+      "grad_norm": 0.8681505444522736,
+      "learning_rate": 4.458925844381209e-05,
+      "loss": 12.1571,
+      "step": 11945
+    },
+    {
+      "epoch": 2.089049180327869,
+      "grad_norm": 0.8812395740509532,
+      "learning_rate": 4.4573542118922915e-05,
+      "loss": 12.1693,
+      "step": 11946
+    },
+    {
+      "epoch": 2.089224043715847,
+      "grad_norm": 0.906209800829551,
+      "learning_rate": 4.455782776992224e-05,
+      "loss": 11.8912,
+      "step": 11947
+    },
+    {
+      "epoch": 2.089398907103825,
+      "grad_norm": 0.7929278105500595,
+      "learning_rate": 4.4542115397370196e-05,
+      "loss": 12.0255,
+      "step": 11948
+    },
+    {
+      "epoch": 2.0895737704918034,
+      "grad_norm": 1.0323307639918706,
+      "learning_rate": 4.452640500182696e-05,
+      "loss": 12.2723,
+      "step": 11949
+    },
+    {
+      "epoch": 2.0897486338797813,
+      "grad_norm": 0.8089533452875071,
+      "learning_rate": 4.451069658385254e-05,
+      "loss": 12.0748,
+      "step": 11950
+    },
+    {
+      "epoch": 2.0899234972677596,
+      "grad_norm": 0.8623182216385574,
+      "learning_rate": 4.449499014400701e-05,
+      "loss": 11.9373,
+      "step": 11951
+    },
+    {
+      "epoch": 2.0900983606557375,
+      "grad_norm": 1.0430811987026767,
+      "learning_rate": 4.447928568285019e-05,
+      "loss": 11.9706,
+      "step": 11952
+    },
+    {
+      "epoch": 2.090273224043716,
+      "grad_norm": 0.9638515208600379,
+      "learning_rate": 4.446358320094197e-05,
+      "loss": 12.2226,
+      "step": 11953
+    },
+    {
+      "epoch": 2.090448087431694,
+      "grad_norm": 0.871163794669213,
+      "learning_rate": 4.4447882698842117e-05,
+      "loss": 11.9087,
+      "step": 11954
+    },
+    {
+      "epoch": 2.090622950819672,
+      "grad_norm": 0.8838734603309314,
+      "learning_rate": 4.443218417711035e-05,
+      "loss": 12.1238,
+      "step": 11955
+    },
+    {
+      "epoch": 2.0907978142076504,
+      "grad_norm": 0.8240086292508452,
+      "learning_rate": 4.441648763630628e-05,
+      "loss": 12.0486,
+      "step": 11956
+    },
+    {
+      "epoch": 2.0909726775956283,
+      "grad_norm": 0.9045078338297504,
+      "learning_rate": 4.4400793076989485e-05,
+      "loss": 12.1665,
+      "step": 11957
+    },
+    {
+      "epoch": 2.0911475409836067,
+      "grad_norm": 0.8316960801443436,
+      "learning_rate": 4.43851004997195e-05,
+      "loss": 11.9853,
+      "step": 11958
+    },
+    {
+      "epoch": 2.0913224043715846,
+      "grad_norm": 0.904327055799402,
+      "learning_rate": 4.436940990505566e-05,
+      "loss": 11.9631,
+      "step": 11959
+    },
+    {
+      "epoch": 2.091497267759563,
+      "grad_norm": 0.8138283556815865,
+      "learning_rate": 4.435372129355736e-05,
+      "loss": 11.7985,
+      "step": 11960
+    },
+    {
+      "epoch": 2.091672131147541,
+      "grad_norm": 0.757283087287403,
+      "learning_rate": 4.433803466578389e-05,
+      "loss": 11.7704,
+      "step": 11961
+    },
+    {
+      "epoch": 2.091846994535519,
+      "grad_norm": 0.9291661749156359,
+      "learning_rate": 4.432235002229448e-05,
+      "loss": 12.2984,
+      "step": 11962
+    },
+    {
+      "epoch": 2.0920218579234975,
+      "grad_norm": 0.891995388239148,
+      "learning_rate": 4.430666736364819e-05,
+      "loss": 12.1616,
+      "step": 11963
+    },
+    {
+      "epoch": 2.0921967213114754,
+      "grad_norm": 0.8628171202237124,
+      "learning_rate": 4.4290986690404156e-05,
+      "loss": 12.1814,
+      "step": 11964
+    },
+    {
+      "epoch": 2.0923715846994537,
+      "grad_norm": 0.8755203848539038,
+      "learning_rate": 4.427530800312135e-05,
+      "loss": 12.0405,
+      "step": 11965
+    },
+    {
+      "epoch": 2.0925464480874316,
+      "grad_norm": 0.8672043431701262,
+      "learning_rate": 4.425963130235875e-05,
+      "loss": 11.9355,
+      "step": 11966
+    },
+    {
+      "epoch": 2.09272131147541,
+      "grad_norm": 0.8161629700239414,
+      "learning_rate": 4.4243956588675116e-05,
+      "loss": 11.9945,
+      "step": 11967
+    },
+    {
+      "epoch": 2.092896174863388,
+      "grad_norm": 0.7944055961183331,
+      "learning_rate": 4.42282838626293e-05,
+      "loss": 12.0348,
+      "step": 11968
+    },
+    {
+      "epoch": 2.093071038251366,
+      "grad_norm": 1.0058678904862242,
+      "learning_rate": 4.421261312478e-05,
+      "loss": 12.1851,
+      "step": 11969
+    },
+    {
+      "epoch": 2.093245901639344,
+      "grad_norm": 0.8468734951832103,
+      "learning_rate": 4.419694437568589e-05,
+      "loss": 12.1404,
+      "step": 11970
+    },
+    {
+      "epoch": 2.0934207650273224,
+      "grad_norm": 0.8403302553221277,
+      "learning_rate": 4.418127761590549e-05,
+      "loss": 12.0337,
+      "step": 11971
+    },
+    {
+      "epoch": 2.0935956284153003,
+      "grad_norm": 0.8741633783272427,
+      "learning_rate": 4.416561284599732e-05,
+      "loss": 12.0556,
+      "step": 11972
+    },
+    {
+      "epoch": 2.0937704918032787,
+      "grad_norm": 0.8324424590244093,
+      "learning_rate": 4.414995006651981e-05,
+      "loss": 12.1531,
+      "step": 11973
+    },
+    {
+      "epoch": 2.093945355191257,
+      "grad_norm": 0.8271520378641155,
+      "learning_rate": 4.413428927803134e-05,
+      "loss": 12.0057,
+      "step": 11974
+    },
+    {
+      "epoch": 2.094120218579235,
+      "grad_norm": 0.9367189005251737,
+      "learning_rate": 4.4118630481090174e-05,
+      "loss": 11.8981,
+      "step": 11975
+    },
+    {
+      "epoch": 2.0942950819672133,
+      "grad_norm": 0.8218004857993753,
+      "learning_rate": 4.4102973676254544e-05,
+      "loss": 12.0751,
+      "step": 11976
+    },
+    {
+      "epoch": 2.094469945355191,
+      "grad_norm": 0.8285176255258976,
+      "learning_rate": 4.4087318864082626e-05,
+      "loss": 11.9845,
+      "step": 11977
+    },
+    {
+      "epoch": 2.0946448087431695,
+      "grad_norm": 0.8646921642358701,
+      "learning_rate": 4.407166604513243e-05,
+      "loss": 12.1162,
+      "step": 11978
+    },
+    {
+      "epoch": 2.0948196721311474,
+      "grad_norm": 0.826640987484581,
+      "learning_rate": 4.4056015219962e-05,
+      "loss": 12.2579,
+      "step": 11979
+    },
+    {
+      "epoch": 2.0949945355191257,
+      "grad_norm": 0.906937388700659,
+      "learning_rate": 4.404036638912926e-05,
+      "loss": 12.1865,
+      "step": 11980
+    },
+    {
+      "epoch": 2.0951693989071036,
+      "grad_norm": 0.813549042336065,
+      "learning_rate": 4.402471955319212e-05,
+      "loss": 12.0311,
+      "step": 11981
+    },
+    {
+      "epoch": 2.095344262295082,
+      "grad_norm": 0.8377044828973185,
+      "learning_rate": 4.4009074712708274e-05,
+      "loss": 12.2416,
+      "step": 11982
+    },
+    {
+      "epoch": 2.0955191256830603,
+      "grad_norm": 0.8215525024232547,
+      "learning_rate": 4.399343186823551e-05,
+      "loss": 11.8151,
+      "step": 11983
+    },
+    {
+      "epoch": 2.095693989071038,
+      "grad_norm": 0.845793557090761,
+      "learning_rate": 4.397779102033146e-05,
+      "loss": 12.168,
+      "step": 11984
+    },
+    {
+      "epoch": 2.0958688524590166,
+      "grad_norm": 0.8856016382817637,
+      "learning_rate": 4.3962152169553735e-05,
+      "loss": 12.054,
+      "step": 11985
+    },
+    {
+      "epoch": 2.0960437158469944,
+      "grad_norm": 0.9178237676415065,
+      "learning_rate": 4.3946515316459794e-05,
+      "loss": 12.1969,
+      "step": 11986
+    },
+    {
+      "epoch": 2.096218579234973,
+      "grad_norm": 0.8322139030651309,
+      "learning_rate": 4.393088046160707e-05,
+      "loss": 12.1104,
+      "step": 11987
+    },
+    {
+      "epoch": 2.0963934426229507,
+      "grad_norm": 0.8481475522088542,
+      "learning_rate": 4.3915247605552986e-05,
+      "loss": 11.9652,
+      "step": 11988
+    },
+    {
+      "epoch": 2.096568306010929,
+      "grad_norm": 0.9022350626645209,
+      "learning_rate": 4.3899616748854774e-05,
+      "loss": 11.8004,
+      "step": 11989
+    },
+    {
+      "epoch": 2.096743169398907,
+      "grad_norm": 0.8644198642225273,
+      "learning_rate": 4.388398789206967e-05,
+      "loss": 11.9939,
+      "step": 11990
+    },
+    {
+      "epoch": 2.0969180327868853,
+      "grad_norm": 0.8502871433919141,
+      "learning_rate": 4.3868361035754834e-05,
+      "loss": 12.0385,
+      "step": 11991
+    },
+    {
+      "epoch": 2.097092896174863,
+      "grad_norm": 0.8617759444502106,
+      "learning_rate": 4.385273618046735e-05,
+      "loss": 12.2647,
+      "step": 11992
+    },
+    {
+      "epoch": 2.0972677595628415,
+      "grad_norm": 0.8107242270291791,
+      "learning_rate": 4.383711332676421e-05,
+      "loss": 12.0099,
+      "step": 11993
+    },
+    {
+      "epoch": 2.09744262295082,
+      "grad_norm": 0.8759216474473052,
+      "learning_rate": 4.382149247520236e-05,
+      "loss": 12.2833,
+      "step": 11994
+    },
+    {
+      "epoch": 2.0976174863387977,
+      "grad_norm": 0.8177897564232889,
+      "learning_rate": 4.3805873626338675e-05,
+      "loss": 11.8044,
+      "step": 11995
+    },
+    {
+      "epoch": 2.097792349726776,
+      "grad_norm": 0.9332160244277444,
+      "learning_rate": 4.379025678072998e-05,
+      "loss": 12.2227,
+      "step": 11996
+    },
+    {
+      "epoch": 2.097967213114754,
+      "grad_norm": 0.7905306527216682,
+      "learning_rate": 4.37746419389329e-05,
+      "loss": 12.0821,
+      "step": 11997
+    },
+    {
+      "epoch": 2.0981420765027323,
+      "grad_norm": 0.8048086859193996,
+      "learning_rate": 4.375902910150416e-05,
+      "loss": 12.2117,
+      "step": 11998
+    },
+    {
+      "epoch": 2.09831693989071,
+      "grad_norm": 0.8746753926214917,
+      "learning_rate": 4.374341826900031e-05,
+      "loss": 12.2389,
+      "step": 11999
+    },
+    {
+      "epoch": 2.0984918032786886,
+      "grad_norm": 0.8399495416954106,
+      "learning_rate": 4.372780944197792e-05,
+      "loss": 12.1611,
+      "step": 12000
+    },
+    {
+      "epoch": 2.0986666666666665,
+      "grad_norm": 0.8500855622935243,
+      "learning_rate": 4.371220262099334e-05,
+      "loss": 12.0821,
+      "step": 12001
+    },
+    {
+      "epoch": 2.098841530054645,
+      "grad_norm": 0.831237031257971,
+      "learning_rate": 4.3696597806602956e-05,
+      "loss": 12.0205,
+      "step": 12002
+    },
+    {
+      "epoch": 2.099016393442623,
+      "grad_norm": 0.8911527586809314,
+      "learning_rate": 4.368099499936312e-05,
+      "loss": 12.0975,
+      "step": 12003
+    },
+    {
+      "epoch": 2.099191256830601,
+      "grad_norm": 0.8510445809686693,
+      "learning_rate": 4.366539419982998e-05,
+      "loss": 12.1882,
+      "step": 12004
+    },
+    {
+      "epoch": 2.0993661202185794,
+      "grad_norm": 0.7913864815078623,
+      "learning_rate": 4.364979540855971e-05,
+      "loss": 12.1256,
+      "step": 12005
+    },
+    {
+      "epoch": 2.0995409836065573,
+      "grad_norm": 0.7636990539640274,
+      "learning_rate": 4.36341986261084e-05,
+      "loss": 11.7984,
+      "step": 12006
+    },
+    {
+      "epoch": 2.0997158469945356,
+      "grad_norm": 0.9217214886362757,
+      "learning_rate": 4.361860385303206e-05,
+      "loss": 11.9399,
+      "step": 12007
+    },
+    {
+      "epoch": 2.0998907103825135,
+      "grad_norm": 0.8302365737377213,
+      "learning_rate": 4.36030110898866e-05,
+      "loss": 11.8919,
+      "step": 12008
+    },
+    {
+      "epoch": 2.100065573770492,
+      "grad_norm": 0.8507719446799782,
+      "learning_rate": 4.358742033722794e-05,
+      "loss": 12.1347,
+      "step": 12009
+    },
+    {
+      "epoch": 2.1002404371584698,
+      "grad_norm": 0.8813034465897187,
+      "learning_rate": 4.357183159561181e-05,
+      "loss": 12.1408,
+      "step": 12010
+    },
+    {
+      "epoch": 2.100415300546448,
+      "grad_norm": 0.8185147579457319,
+      "learning_rate": 4.3556244865593955e-05,
+      "loss": 11.9234,
+      "step": 12011
+    },
+    {
+      "epoch": 2.1005901639344264,
+      "grad_norm": 0.8468147418894016,
+      "learning_rate": 4.354066014773002e-05,
+      "loss": 12.1607,
+      "step": 12012
+    },
+    {
+      "epoch": 2.1007650273224043,
+      "grad_norm": 0.8985669373000853,
+      "learning_rate": 4.352507744257558e-05,
+      "loss": 12.1235,
+      "step": 12013
+    },
+    {
+      "epoch": 2.1009398907103827,
+      "grad_norm": 0.8576739383708666,
+      "learning_rate": 4.350949675068615e-05,
+      "loss": 12.2119,
+      "step": 12014
+    },
+    {
+      "epoch": 2.1011147540983606,
+      "grad_norm": 0.8133234359343455,
+      "learning_rate": 4.349391807261719e-05,
+      "loss": 11.7856,
+      "step": 12015
+    },
+    {
+      "epoch": 2.101289617486339,
+      "grad_norm": 0.8315391235034162,
+      "learning_rate": 4.347834140892399e-05,
+      "loss": 11.9737,
+      "step": 12016
+    },
+    {
+      "epoch": 2.101464480874317,
+      "grad_norm": 0.913501360894979,
+      "learning_rate": 4.34627667601619e-05,
+      "loss": 12.0659,
+      "step": 12017
+    },
+    {
+      "epoch": 2.101639344262295,
+      "grad_norm": 0.9548490600908927,
+      "learning_rate": 4.3447194126886135e-05,
+      "loss": 12.256,
+      "step": 12018
+    },
+    {
+      "epoch": 2.101814207650273,
+      "grad_norm": 0.898326238614734,
+      "learning_rate": 4.34316235096518e-05,
+      "loss": 12.0682,
+      "step": 12019
+    },
+    {
+      "epoch": 2.1019890710382514,
+      "grad_norm": 0.8224190693424356,
+      "learning_rate": 4.341605490901398e-05,
+      "loss": 12.2275,
+      "step": 12020
+    },
+    {
+      "epoch": 2.1021639344262297,
+      "grad_norm": 0.8753334214634932,
+      "learning_rate": 4.340048832552771e-05,
+      "loss": 12.0011,
+      "step": 12021
+    },
+    {
+      "epoch": 2.1023387978142076,
+      "grad_norm": 0.9167503163715941,
+      "learning_rate": 4.338492375974792e-05,
+      "loss": 12.1232,
+      "step": 12022
+    },
+    {
+      "epoch": 2.102513661202186,
+      "grad_norm": 0.7965382040633766,
+      "learning_rate": 4.336936121222943e-05,
+      "loss": 12.1291,
+      "step": 12023
+    },
+    {
+      "epoch": 2.102688524590164,
+      "grad_norm": 0.819725214716866,
+      "learning_rate": 4.335380068352704e-05,
+      "loss": 12.2194,
+      "step": 12024
+    },
+    {
+      "epoch": 2.102863387978142,
+      "grad_norm": 0.8085360524966648,
+      "learning_rate": 4.333824217419546e-05,
+      "loss": 12.1171,
+      "step": 12025
+    },
+    {
+      "epoch": 2.10303825136612,
+      "grad_norm": 0.8201022341428373,
+      "learning_rate": 4.3322685684789367e-05,
+      "loss": 12.3697,
+      "step": 12026
+    },
+    {
+      "epoch": 2.1032131147540984,
+      "grad_norm": 0.8435532843599971,
+      "learning_rate": 4.330713121586329e-05,
+      "loss": 12.1963,
+      "step": 12027
+    },
+    {
+      "epoch": 2.1033879781420763,
+      "grad_norm": 0.8658967528740608,
+      "learning_rate": 4.329157876797175e-05,
+      "loss": 12.1483,
+      "step": 12028
+    },
+    {
+      "epoch": 2.1035628415300547,
+      "grad_norm": 0.8389215051604817,
+      "learning_rate": 4.3276028341669185e-05,
+      "loss": 11.9879,
+      "step": 12029
+    },
+    {
+      "epoch": 2.1037377049180326,
+      "grad_norm": 0.9696236446846754,
+      "learning_rate": 4.326047993750996e-05,
+      "loss": 12.1468,
+      "step": 12030
+    },
+    {
+      "epoch": 2.103912568306011,
+      "grad_norm": 0.9241328020912011,
+      "learning_rate": 4.32449335560483e-05,
+      "loss": 12.0973,
+      "step": 12031
+    },
+    {
+      "epoch": 2.1040874316939893,
+      "grad_norm": 0.8584629465687126,
+      "learning_rate": 4.322938919783843e-05,
+      "loss": 11.9571,
+      "step": 12032
+    },
+    {
+      "epoch": 2.104262295081967,
+      "grad_norm": 0.8584162338126011,
+      "learning_rate": 4.321384686343453e-05,
+      "loss": 12.1709,
+      "step": 12033
+    },
+    {
+      "epoch": 2.1044371584699455,
+      "grad_norm": 0.8833425111524239,
+      "learning_rate": 4.3198306553390666e-05,
+      "loss": 12.1137,
+      "step": 12034
+    },
+    {
+      "epoch": 2.1046120218579234,
+      "grad_norm": 0.8191366507638079,
+      "learning_rate": 4.3182768268260764e-05,
+      "loss": 11.9453,
+      "step": 12035
+    },
+    {
+      "epoch": 2.1047868852459017,
+      "grad_norm": 0.8361388713620261,
+      "learning_rate": 4.3167232008598804e-05,
+      "loss": 12.1827,
+      "step": 12036
+    },
+    {
+      "epoch": 2.1049617486338796,
+      "grad_norm": 0.8040536821294442,
+      "learning_rate": 4.315169777495866e-05,
+      "loss": 12.1195,
+      "step": 12037
+    },
+    {
+      "epoch": 2.105136612021858,
+      "grad_norm": 0.8423532040977776,
+      "learning_rate": 4.313616556789402e-05,
+      "loss": 12.1529,
+      "step": 12038
+    },
+    {
+      "epoch": 2.105311475409836,
+      "grad_norm": 0.8428565847974528,
+      "learning_rate": 4.312063538795864e-05,
+      "loss": 11.8966,
+      "step": 12039
+    },
+    {
+      "epoch": 2.105486338797814,
+      "grad_norm": 0.8833797093672888,
+      "learning_rate": 4.3105107235706156e-05,
+      "loss": 11.8797,
+      "step": 12040
+    },
+    {
+      "epoch": 2.1056612021857926,
+      "grad_norm": 0.8547728777786161,
+      "learning_rate": 4.308958111169017e-05,
+      "loss": 12.3025,
+      "step": 12041
+    },
+    {
+      "epoch": 2.1058360655737705,
+      "grad_norm": 0.8792601708653234,
+      "learning_rate": 4.3074057016464075e-05,
+      "loss": 11.7923,
+      "step": 12042
+    },
+    {
+      "epoch": 2.106010928961749,
+      "grad_norm": 0.8468934509432596,
+      "learning_rate": 4.3058534950581344e-05,
+      "loss": 12.1166,
+      "step": 12043
+    },
+    {
+      "epoch": 2.1061857923497267,
+      "grad_norm": 0.8132155124522168,
+      "learning_rate": 4.30430149145953e-05,
+      "loss": 12.0325,
+      "step": 12044
+    },
+    {
+      "epoch": 2.106360655737705,
+      "grad_norm": 0.8210248947282365,
+      "learning_rate": 4.302749690905925e-05,
+      "loss": 12.2575,
+      "step": 12045
+    },
+    {
+      "epoch": 2.106535519125683,
+      "grad_norm": 0.8053798336834486,
+      "learning_rate": 4.3011980934526354e-05,
+      "loss": 11.956,
+      "step": 12046
+    },
+    {
+      "epoch": 2.1067103825136613,
+      "grad_norm": 0.8336652429821607,
+      "learning_rate": 4.299646699154975e-05,
+      "loss": 11.9888,
+      "step": 12047
+    },
+    {
+      "epoch": 2.106885245901639,
+      "grad_norm": 0.8456822435174728,
+      "learning_rate": 4.2980955080682515e-05,
+      "loss": 12.0308,
+      "step": 12048
+    },
+    {
+      "epoch": 2.1070601092896175,
+      "grad_norm": 0.8211553269338716,
+      "learning_rate": 4.296544520247764e-05,
+      "loss": 12.1546,
+      "step": 12049
+    },
+    {
+      "epoch": 2.1072349726775954,
+      "grad_norm": 0.832992392822977,
+      "learning_rate": 4.294993735748796e-05,
+      "loss": 11.9708,
+      "step": 12050
+    },
+    {
+      "epoch": 2.1074098360655737,
+      "grad_norm": 0.8408228565397176,
+      "learning_rate": 4.2934431546266366e-05,
+      "loss": 12.0107,
+      "step": 12051
+    },
+    {
+      "epoch": 2.107584699453552,
+      "grad_norm": 0.7965811271198324,
+      "learning_rate": 4.2918927769365644e-05,
+      "loss": 12.108,
+      "step": 12052
+    },
+    {
+      "epoch": 2.10775956284153,
+      "grad_norm": 0.894939719546122,
+      "learning_rate": 4.290342602733842e-05,
+      "loss": 12.1981,
+      "step": 12053
+    },
+    {
+      "epoch": 2.1079344262295083,
+      "grad_norm": 0.8042082011407592,
+      "learning_rate": 4.288792632073736e-05,
+      "loss": 12.1362,
+      "step": 12054
+    },
+    {
+      "epoch": 2.1081092896174862,
+      "grad_norm": 1.1257331592367472,
+      "learning_rate": 4.287242865011498e-05,
+      "loss": 12.0785,
+      "step": 12055
+    },
+    {
+      "epoch": 2.1082841530054646,
+      "grad_norm": 0.8536543713419771,
+      "learning_rate": 4.285693301602383e-05,
+      "loss": 12.2426,
+      "step": 12056
+    },
+    {
+      "epoch": 2.1084590163934425,
+      "grad_norm": 0.8141797865965266,
+      "learning_rate": 4.2841439419016194e-05,
+      "loss": 12.0998,
+      "step": 12057
+    },
+    {
+      "epoch": 2.108633879781421,
+      "grad_norm": 0.8067752045072957,
+      "learning_rate": 4.2825947859644476e-05,
+      "loss": 12.2029,
+      "step": 12058
+    },
+    {
+      "epoch": 2.1088087431693987,
+      "grad_norm": 0.8063628852887815,
+      "learning_rate": 4.281045833846091e-05,
+      "loss": 12.0536,
+      "step": 12059
+    },
+    {
+      "epoch": 2.108983606557377,
+      "grad_norm": 0.8586095007283981,
+      "learning_rate": 4.279497085601772e-05,
+      "loss": 12.1848,
+      "step": 12060
+    },
+    {
+      "epoch": 2.1091584699453554,
+      "grad_norm": 0.8584895988232425,
+      "learning_rate": 4.277948541286695e-05,
+      "loss": 12.1064,
+      "step": 12061
+    },
+    {
+      "epoch": 2.1093333333333333,
+      "grad_norm": 0.7775033761246734,
+      "learning_rate": 4.276400200956067e-05,
+      "loss": 12.2268,
+      "step": 12062
+    },
+    {
+      "epoch": 2.1095081967213116,
+      "grad_norm": 0.8222213403080607,
+      "learning_rate": 4.2748520646650845e-05,
+      "loss": 11.7659,
+      "step": 12063
+    },
+    {
+      "epoch": 2.1096830601092895,
+      "grad_norm": 0.8682272043796141,
+      "learning_rate": 4.273304132468936e-05,
+      "loss": 12.0357,
+      "step": 12064
+    },
+    {
+      "epoch": 2.109857923497268,
+      "grad_norm": 0.8925212549596842,
+      "learning_rate": 4.2717564044228056e-05,
+      "loss": 11.8739,
+      "step": 12065
+    },
+    {
+      "epoch": 2.1100327868852458,
+      "grad_norm": 0.8748669809395033,
+      "learning_rate": 4.270208880581866e-05,
+      "loss": 11.8192,
+      "step": 12066
+    },
+    {
+      "epoch": 2.110207650273224,
+      "grad_norm": 0.8141878741029384,
+      "learning_rate": 4.268661561001288e-05,
+      "loss": 12.1086,
+      "step": 12067
+    },
+    {
+      "epoch": 2.110382513661202,
+      "grad_norm": 0.8456853825293337,
+      "learning_rate": 4.267114445736227e-05,
+      "loss": 11.7032,
+      "step": 12068
+    },
+    {
+      "epoch": 2.1105573770491803,
+      "grad_norm": 0.8702145967758425,
+      "learning_rate": 4.265567534841837e-05,
+      "loss": 11.7384,
+      "step": 12069
+    },
+    {
+      "epoch": 2.1107322404371587,
+      "grad_norm": 0.8879571916998786,
+      "learning_rate": 4.264020828373264e-05,
+      "loss": 12.1331,
+      "step": 12070
+    },
+    {
+      "epoch": 2.1109071038251366,
+      "grad_norm": 0.8970172715395069,
+      "learning_rate": 4.262474326385651e-05,
+      "loss": 12.1698,
+      "step": 12071
+    },
+    {
+      "epoch": 2.111081967213115,
+      "grad_norm": 0.8756003024390021,
+      "learning_rate": 4.2609280289341214e-05,
+      "loss": 12.0351,
+      "step": 12072
+    },
+    {
+      "epoch": 2.111256830601093,
+      "grad_norm": 0.978136220243092,
+      "learning_rate": 4.259381936073802e-05,
+      "loss": 12.2869,
+      "step": 12073
+    },
+    {
+      "epoch": 2.111431693989071,
+      "grad_norm": 0.8574241252032129,
+      "learning_rate": 4.257836047859809e-05,
+      "loss": 12.0968,
+      "step": 12074
+    },
+    {
+      "epoch": 2.111606557377049,
+      "grad_norm": 0.8737962440022039,
+      "learning_rate": 4.256290364347255e-05,
+      "loss": 12.1255,
+      "step": 12075
+    },
+    {
+      "epoch": 2.1117814207650274,
+      "grad_norm": 0.925445361540188,
+      "learning_rate": 4.2547448855912356e-05,
+      "loss": 12.0985,
+      "step": 12076
+    },
+    {
+      "epoch": 2.1119562841530053,
+      "grad_norm": 0.8546011970266806,
+      "learning_rate": 4.253199611646849e-05,
+      "loss": 12.0203,
+      "step": 12077
+    },
+    {
+      "epoch": 2.1121311475409836,
+      "grad_norm": 0.8855882363366561,
+      "learning_rate": 4.251654542569181e-05,
+      "loss": 12.2158,
+      "step": 12078
+    },
+    {
+      "epoch": 2.112306010928962,
+      "grad_norm": 0.883133872998754,
+      "learning_rate": 4.250109678413316e-05,
+      "loss": 12.1501,
+      "step": 12079
+    },
+    {
+      "epoch": 2.11248087431694,
+      "grad_norm": 0.9867492977390363,
+      "learning_rate": 4.2485650192343193e-05,
+      "loss": 12.2318,
+      "step": 12080
+    },
+    {
+      "epoch": 2.112655737704918,
+      "grad_norm": 0.8479531968147244,
+      "learning_rate": 4.2470205650872604e-05,
+      "loss": 12.0213,
+      "step": 12081
+    },
+    {
+      "epoch": 2.112830601092896,
+      "grad_norm": 0.8832867427399992,
+      "learning_rate": 4.245476316027196e-05,
+      "loss": 12.2664,
+      "step": 12082
+    },
+    {
+      "epoch": 2.1130054644808745,
+      "grad_norm": 0.8693414035046508,
+      "learning_rate": 4.243932272109178e-05,
+      "loss": 12.0773,
+      "step": 12083
+    },
+    {
+      "epoch": 2.1131803278688523,
+      "grad_norm": 0.924097663971912,
+      "learning_rate": 4.2423884333882493e-05,
+      "loss": 12.2516,
+      "step": 12084
+    },
+    {
+      "epoch": 2.1133551912568307,
+      "grad_norm": 0.8639055838524533,
+      "learning_rate": 4.2408447999194455e-05,
+      "loss": 11.8729,
+      "step": 12085
+    },
+    {
+      "epoch": 2.1135300546448086,
+      "grad_norm": 0.8583247765706166,
+      "learning_rate": 4.2393013717578e-05,
+      "loss": 12.2438,
+      "step": 12086
+    },
+    {
+      "epoch": 2.113704918032787,
+      "grad_norm": 0.8883679881921278,
+      "learning_rate": 4.237758148958326e-05,
+      "loss": 12.1312,
+      "step": 12087
+    },
+    {
+      "epoch": 2.113879781420765,
+      "grad_norm": 0.8873680916929348,
+      "learning_rate": 4.236215131576041e-05,
+      "loss": 12.3081,
+      "step": 12088
+    },
+    {
+      "epoch": 2.114054644808743,
+      "grad_norm": 0.8485501182497236,
+      "learning_rate": 4.234672319665953e-05,
+      "loss": 12.0187,
+      "step": 12089
+    },
+    {
+      "epoch": 2.1142295081967215,
+      "grad_norm": 0.8530141879494058,
+      "learning_rate": 4.233129713283065e-05,
+      "loss": 12.1194,
+      "step": 12090
+    },
+    {
+      "epoch": 2.1144043715846994,
+      "grad_norm": 0.8185226239636512,
+      "learning_rate": 4.23158731248236e-05,
+      "loss": 12.0758,
+      "step": 12091
+    },
+    {
+      "epoch": 2.1145792349726777,
+      "grad_norm": 0.8321708156802045,
+      "learning_rate": 4.2300451173188285e-05,
+      "loss": 11.845,
+      "step": 12092
+    },
+    {
+      "epoch": 2.1147540983606556,
+      "grad_norm": 0.9008297790395617,
+      "learning_rate": 4.2285031278474466e-05,
+      "loss": 12.0126,
+      "step": 12093
+    },
+    {
+      "epoch": 2.114928961748634,
+      "grad_norm": 0.8832732760363637,
+      "learning_rate": 4.226961344123188e-05,
+      "loss": 12.2324,
+      "step": 12094
+    },
+    {
+      "epoch": 2.115103825136612,
+      "grad_norm": 0.8251294074908271,
+      "learning_rate": 4.22541976620101e-05,
+      "loss": 12.0847,
+      "step": 12095
+    },
+    {
+      "epoch": 2.1152786885245902,
+      "grad_norm": 0.8718218526889548,
+      "learning_rate": 4.223878394135869e-05,
+      "loss": 12.1868,
+      "step": 12096
+    },
+    {
+      "epoch": 2.115453551912568,
+      "grad_norm": 0.7743947268109684,
+      "learning_rate": 4.222337227982719e-05,
+      "loss": 12.0948,
+      "step": 12097
+    },
+    {
+      "epoch": 2.1156284153005465,
+      "grad_norm": 0.8728093146782223,
+      "learning_rate": 4.220796267796493e-05,
+      "loss": 11.9212,
+      "step": 12098
+    },
+    {
+      "epoch": 2.115803278688525,
+      "grad_norm": 0.8160437074512754,
+      "learning_rate": 4.2192555136321276e-05,
+      "loss": 11.9576,
+      "step": 12099
+    },
+    {
+      "epoch": 2.1159781420765027,
+      "grad_norm": 0.840638994171586,
+      "learning_rate": 4.217714965544549e-05,
+      "loss": 11.894,
+      "step": 12100
+    },
+    {
+      "epoch": 2.116153005464481,
+      "grad_norm": 0.9772376481897073,
+      "learning_rate": 4.2161746235886755e-05,
+      "loss": 11.9785,
+      "step": 12101
+    },
+    {
+      "epoch": 2.116327868852459,
+      "grad_norm": 0.8388928505040012,
+      "learning_rate": 4.21463448781942e-05,
+      "loss": 12.0781,
+      "step": 12102
+    },
+    {
+      "epoch": 2.1165027322404373,
+      "grad_norm": 0.8881732793893092,
+      "learning_rate": 4.213094558291685e-05,
+      "loss": 12.1128,
+      "step": 12103
+    },
+    {
+      "epoch": 2.116677595628415,
+      "grad_norm": 0.9179317927428691,
+      "learning_rate": 4.211554835060367e-05,
+      "loss": 11.8931,
+      "step": 12104
+    },
+    {
+      "epoch": 2.1168524590163935,
+      "grad_norm": 0.8316628842162318,
+      "learning_rate": 4.210015318180361e-05,
+      "loss": 11.9237,
+      "step": 12105
+    },
+    {
+      "epoch": 2.1170273224043714,
+      "grad_norm": 0.8768518909824002,
+      "learning_rate": 4.20847600770654e-05,
+      "loss": 12.074,
+      "step": 12106
+    },
+    {
+      "epoch": 2.1172021857923498,
+      "grad_norm": 0.9702360790216447,
+      "learning_rate": 4.206936903693782e-05,
+      "loss": 11.9113,
+      "step": 12107
+    },
+    {
+      "epoch": 2.1173770491803277,
+      "grad_norm": 0.8628301082852675,
+      "learning_rate": 4.205398006196955e-05,
+      "loss": 12.0726,
+      "step": 12108
+    },
+    {
+      "epoch": 2.117551912568306,
+      "grad_norm": 0.8594474293622038,
+      "learning_rate": 4.203859315270924e-05,
+      "loss": 12.1062,
+      "step": 12109
+    },
+    {
+      "epoch": 2.1177267759562843,
+      "grad_norm": 0.8241886719159521,
+      "learning_rate": 4.202320830970531e-05,
+      "loss": 12.1534,
+      "step": 12110
+    },
+    {
+      "epoch": 2.1179016393442622,
+      "grad_norm": 0.9736924948954373,
+      "learning_rate": 4.200782553350627e-05,
+      "loss": 12.2028,
+      "step": 12111
+    },
+    {
+      "epoch": 2.1180765027322406,
+      "grad_norm": 0.8633689599186891,
+      "learning_rate": 4.199244482466054e-05,
+      "loss": 12.0266,
+      "step": 12112
+    },
+    {
+      "epoch": 2.1182513661202185,
+      "grad_norm": 0.901800662330018,
+      "learning_rate": 4.197706618371633e-05,
+      "loss": 11.9633,
+      "step": 12113
+    },
+    {
+      "epoch": 2.118426229508197,
+      "grad_norm": 0.9273406551023773,
+      "learning_rate": 4.196168961122191e-05,
+      "loss": 11.8045,
+      "step": 12114
+    },
+    {
+      "epoch": 2.1186010928961747,
+      "grad_norm": 0.9157355600673849,
+      "learning_rate": 4.194631510772546e-05,
+      "loss": 11.8312,
+      "step": 12115
+    },
+    {
+      "epoch": 2.118775956284153,
+      "grad_norm": 0.8617042262997628,
+      "learning_rate": 4.1930942673775045e-05,
+      "loss": 12.1216,
+      "step": 12116
+    },
+    {
+      "epoch": 2.118950819672131,
+      "grad_norm": 0.9100316553042627,
+      "learning_rate": 4.191557230991867e-05,
+      "loss": 12.1662,
+      "step": 12117
+    },
+    {
+      "epoch": 2.1191256830601093,
+      "grad_norm": 0.8515907618049461,
+      "learning_rate": 4.190020401670428e-05,
+      "loss": 11.929,
+      "step": 12118
+    },
+    {
+      "epoch": 2.1193005464480876,
+      "grad_norm": 0.8538258278088792,
+      "learning_rate": 4.188483779467974e-05,
+      "loss": 12.1252,
+      "step": 12119
+    },
+    {
+      "epoch": 2.1194754098360655,
+      "grad_norm": 0.9243149891159356,
+      "learning_rate": 4.186947364439287e-05,
+      "loss": 12.1882,
+      "step": 12120
+    },
+    {
+      "epoch": 2.119650273224044,
+      "grad_norm": 0.8106433762913967,
+      "learning_rate": 4.1854111566391294e-05,
+      "loss": 12.0052,
+      "step": 12121
+    },
+    {
+      "epoch": 2.1198251366120218,
+      "grad_norm": 0.8523301694011952,
+      "learning_rate": 4.183875156122272e-05,
+      "loss": 12.1279,
+      "step": 12122
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.840754161836623,
+      "learning_rate": 4.1823393629434694e-05,
+      "loss": 12.0915,
+      "step": 12123
+    },
+    {
+      "epoch": 2.120174863387978,
+      "grad_norm": 0.83293612121378,
+      "learning_rate": 4.180803777157475e-05,
+      "loss": 12.2412,
+      "step": 12124
+    },
+    {
+      "epoch": 2.1203497267759563,
+      "grad_norm": 0.8694185905286768,
+      "learning_rate": 4.179268398819024e-05,
+      "loss": 12.0005,
+      "step": 12125
+    },
+    {
+      "epoch": 2.1205245901639342,
+      "grad_norm": 0.8337785925191948,
+      "learning_rate": 4.177733227982853e-05,
+      "loss": 12.0267,
+      "step": 12126
+    },
+    {
+      "epoch": 2.1206994535519126,
+      "grad_norm": 0.7922375819664274,
+      "learning_rate": 4.176198264703694e-05,
+      "loss": 12.063,
+      "step": 12127
+    },
+    {
+      "epoch": 2.120874316939891,
+      "grad_norm": 0.876293191823484,
+      "learning_rate": 4.174663509036257e-05,
+      "loss": 11.9819,
+      "step": 12128
+    },
+    {
+      "epoch": 2.121049180327869,
+      "grad_norm": 0.8454498467010894,
+      "learning_rate": 4.173128961035262e-05,
+      "loss": 11.8199,
+      "step": 12129
+    },
+    {
+      "epoch": 2.121224043715847,
+      "grad_norm": 0.803778862804482,
+      "learning_rate": 4.1715946207554105e-05,
+      "loss": 11.9911,
+      "step": 12130
+    },
+    {
+      "epoch": 2.121398907103825,
+      "grad_norm": 0.8721367721963388,
+      "learning_rate": 4.170060488251405e-05,
+      "loss": 12.2638,
+      "step": 12131
+    },
+    {
+      "epoch": 2.1215737704918034,
+      "grad_norm": 0.8937730321671318,
+      "learning_rate": 4.168526563577928e-05,
+      "loss": 12.0173,
+      "step": 12132
+    },
+    {
+      "epoch": 2.1217486338797813,
+      "grad_norm": 0.9022952695035246,
+      "learning_rate": 4.166992846789666e-05,
+      "loss": 12.0788,
+      "step": 12133
+    },
+    {
+      "epoch": 2.1219234972677596,
+      "grad_norm": 0.8431250547695819,
+      "learning_rate": 4.165459337941292e-05,
+      "loss": 11.9473,
+      "step": 12134
+    },
+    {
+      "epoch": 2.1220983606557375,
+      "grad_norm": 0.9349634211680662,
+      "learning_rate": 4.1639260370874775e-05,
+      "loss": 12.0901,
+      "step": 12135
+    },
+    {
+      "epoch": 2.122273224043716,
+      "grad_norm": 0.8928603405907701,
+      "learning_rate": 4.1623929442828804e-05,
+      "loss": 12.1286,
+      "step": 12136
+    },
+    {
+      "epoch": 2.1224480874316938,
+      "grad_norm": 0.8718415538761642,
+      "learning_rate": 4.160860059582154e-05,
+      "loss": 12.2034,
+      "step": 12137
+    },
+    {
+      "epoch": 2.122622950819672,
+      "grad_norm": 0.8605008130359882,
+      "learning_rate": 4.159327383039945e-05,
+      "loss": 12.2652,
+      "step": 12138
+    },
+    {
+      "epoch": 2.1227978142076505,
+      "grad_norm": 0.830634996283248,
+      "learning_rate": 4.157794914710894e-05,
+      "loss": 11.9705,
+      "step": 12139
+    },
+    {
+      "epoch": 2.1229726775956284,
+      "grad_norm": 0.9774430710694515,
+      "learning_rate": 4.156262654649625e-05,
+      "loss": 12.2937,
+      "step": 12140
+    },
+    {
+      "epoch": 2.1231475409836067,
+      "grad_norm": 0.8643775545647153,
+      "learning_rate": 4.154730602910765e-05,
+      "loss": 12.1542,
+      "step": 12141
+    },
+    {
+      "epoch": 2.1233224043715846,
+      "grad_norm": 0.8767680629277746,
+      "learning_rate": 4.153198759548929e-05,
+      "loss": 12.2821,
+      "step": 12142
+    },
+    {
+      "epoch": 2.123497267759563,
+      "grad_norm": 0.8078804398278655,
+      "learning_rate": 4.151667124618729e-05,
+      "loss": 12.0109,
+      "step": 12143
+    },
+    {
+      "epoch": 2.123672131147541,
+      "grad_norm": 0.9330977769460601,
+      "learning_rate": 4.1501356981747597e-05,
+      "loss": 11.8696,
+      "step": 12144
+    },
+    {
+      "epoch": 2.123846994535519,
+      "grad_norm": 0.9407834442430864,
+      "learning_rate": 4.1486044802716186e-05,
+      "loss": 12.0629,
+      "step": 12145
+    },
+    {
+      "epoch": 2.124021857923497,
+      "grad_norm": 0.9035184019710328,
+      "learning_rate": 4.1470734709638934e-05,
+      "loss": 12.1729,
+      "step": 12146
+    },
+    {
+      "epoch": 2.1241967213114754,
+      "grad_norm": 0.8534223910382601,
+      "learning_rate": 4.1455426703061586e-05,
+      "loss": 11.5566,
+      "step": 12147
+    },
+    {
+      "epoch": 2.1243715846994538,
+      "grad_norm": 0.8871453614942676,
+      "learning_rate": 4.144012078352987e-05,
+      "loss": 12.2264,
+      "step": 12148
+    },
+    {
+      "epoch": 2.1245464480874316,
+      "grad_norm": 0.9075718365591325,
+      "learning_rate": 4.142481695158942e-05,
+      "loss": 12.0529,
+      "step": 12149
+    },
+    {
+      "epoch": 2.12472131147541,
+      "grad_norm": 0.8486328990872938,
+      "learning_rate": 4.1409515207785854e-05,
+      "loss": 12.1929,
+      "step": 12150
+    },
+    {
+      "epoch": 2.124896174863388,
+      "grad_norm": 0.9130931668883329,
+      "learning_rate": 4.1394215552664575e-05,
+      "loss": 12.1312,
+      "step": 12151
+    },
+    {
+      "epoch": 2.1250710382513662,
+      "grad_norm": 0.8795545486587911,
+      "learning_rate": 4.1378917986771026e-05,
+      "loss": 12.2248,
+      "step": 12152
+    },
+    {
+      "epoch": 2.125245901639344,
+      "grad_norm": 0.9165982543071108,
+      "learning_rate": 4.136362251065058e-05,
+      "loss": 12.4307,
+      "step": 12153
+    },
+    {
+      "epoch": 2.1254207650273225,
+      "grad_norm": 0.8943159539466045,
+      "learning_rate": 4.134832912484847e-05,
+      "loss": 12.1292,
+      "step": 12154
+    },
+    {
+      "epoch": 2.1255956284153004,
+      "grad_norm": 0.945991366219717,
+      "learning_rate": 4.1333037829909905e-05,
+      "loss": 12.1468,
+      "step": 12155
+    },
+    {
+      "epoch": 2.1257704918032787,
+      "grad_norm": 0.8514919470322315,
+      "learning_rate": 4.1317748626379994e-05,
+      "loss": 12.3056,
+      "step": 12156
+    },
+    {
+      "epoch": 2.1259453551912566,
+      "grad_norm": 0.8080699280187712,
+      "learning_rate": 4.130246151480378e-05,
+      "loss": 11.8518,
+      "step": 12157
+    },
+    {
+      "epoch": 2.126120218579235,
+      "grad_norm": 0.9215026243349487,
+      "learning_rate": 4.1287176495726264e-05,
+      "loss": 12.2728,
+      "step": 12158
+    },
+    {
+      "epoch": 2.1262950819672133,
+      "grad_norm": 0.8227729953801136,
+      "learning_rate": 4.127189356969228e-05,
+      "loss": 12.1947,
+      "step": 12159
+    },
+    {
+      "epoch": 2.126469945355191,
+      "grad_norm": 0.7929235767256814,
+      "learning_rate": 4.125661273724667e-05,
+      "loss": 12.0602,
+      "step": 12160
+    },
+    {
+      "epoch": 2.1266448087431695,
+      "grad_norm": 0.8738228093712188,
+      "learning_rate": 4.1241333998934216e-05,
+      "loss": 12.0134,
+      "step": 12161
+    },
+    {
+      "epoch": 2.1268196721311474,
+      "grad_norm": 0.8182680403473366,
+      "learning_rate": 4.122605735529952e-05,
+      "loss": 12.0157,
+      "step": 12162
+    },
+    {
+      "epoch": 2.1269945355191258,
+      "grad_norm": 0.7923667187246866,
+      "learning_rate": 4.12107828068872e-05,
+      "loss": 12.0215,
+      "step": 12163
+    },
+    {
+      "epoch": 2.1271693989071037,
+      "grad_norm": 0.8833123252817839,
+      "learning_rate": 4.1195510354241806e-05,
+      "loss": 12.2502,
+      "step": 12164
+    },
+    {
+      "epoch": 2.127344262295082,
+      "grad_norm": 0.8197367274875983,
+      "learning_rate": 4.118023999790778e-05,
+      "loss": 12.0976,
+      "step": 12165
+    },
+    {
+      "epoch": 2.12751912568306,
+      "grad_norm": 0.8314579910821963,
+      "learning_rate": 4.116497173842945e-05,
+      "loss": 12.0327,
+      "step": 12166
+    },
+    {
+      "epoch": 2.1276939890710382,
+      "grad_norm": 0.9059126231233082,
+      "learning_rate": 4.1149705576351125e-05,
+      "loss": 12.0067,
+      "step": 12167
+    },
+    {
+      "epoch": 2.1278688524590166,
+      "grad_norm": 0.8550361963091693,
+      "learning_rate": 4.113444151221705e-05,
+      "loss": 12.2528,
+      "step": 12168
+    },
+    {
+      "epoch": 2.1280437158469945,
+      "grad_norm": 0.981560687602554,
+      "learning_rate": 4.111917954657139e-05,
+      "loss": 12.2059,
+      "step": 12169
+    },
+    {
+      "epoch": 2.128218579234973,
+      "grad_norm": 0.8873028371027133,
+      "learning_rate": 4.1103919679958134e-05,
+      "loss": 12.0926,
+      "step": 12170
+    },
+    {
+      "epoch": 2.1283934426229507,
+      "grad_norm": 0.8064175871688869,
+      "learning_rate": 4.108866191292134e-05,
+      "loss": 12.1748,
+      "step": 12171
+    },
+    {
+      "epoch": 2.128568306010929,
+      "grad_norm": 0.8245937986085583,
+      "learning_rate": 4.107340624600492e-05,
+      "loss": 12.2932,
+      "step": 12172
+    },
+    {
+      "epoch": 2.128743169398907,
+      "grad_norm": 0.824258282793452,
+      "learning_rate": 4.105815267975272e-05,
+      "loss": 12.1139,
+      "step": 12173
+    },
+    {
+      "epoch": 2.1289180327868853,
+      "grad_norm": 0.871128762638985,
+      "learning_rate": 4.104290121470851e-05,
+      "loss": 12.0422,
+      "step": 12174
+    },
+    {
+      "epoch": 2.129092896174863,
+      "grad_norm": 0.9272022227765827,
+      "learning_rate": 4.102765185141598e-05,
+      "loss": 11.82,
+      "step": 12175
+    },
+    {
+      "epoch": 2.1292677595628415,
+      "grad_norm": 0.855367172891774,
+      "learning_rate": 4.10124045904188e-05,
+      "loss": 12.2047,
+      "step": 12176
+    },
+    {
+      "epoch": 2.12944262295082,
+      "grad_norm": 0.8319336929040031,
+      "learning_rate": 4.0997159432260444e-05,
+      "loss": 11.8752,
+      "step": 12177
+    },
+    {
+      "epoch": 2.1296174863387978,
+      "grad_norm": 0.8213724103966594,
+      "learning_rate": 4.0981916377484406e-05,
+      "loss": 12.2594,
+      "step": 12178
+    },
+    {
+      "epoch": 2.129792349726776,
+      "grad_norm": 0.8916714031954979,
+      "learning_rate": 4.0966675426634104e-05,
+      "loss": 12.0397,
+      "step": 12179
+    },
+    {
+      "epoch": 2.129967213114754,
+      "grad_norm": 0.8730633111484996,
+      "learning_rate": 4.095143658025289e-05,
+      "loss": 12.1415,
+      "step": 12180
+    },
+    {
+      "epoch": 2.1301420765027324,
+      "grad_norm": 0.8027586269839493,
+      "learning_rate": 4.093619983888393e-05,
+      "loss": 11.7577,
+      "step": 12181
+    },
+    {
+      "epoch": 2.1303169398907102,
+      "grad_norm": 0.8916256548200909,
+      "learning_rate": 4.092096520307044e-05,
+      "loss": 11.9501,
+      "step": 12182
+    },
+    {
+      "epoch": 2.1304918032786886,
+      "grad_norm": 0.8000644088322459,
+      "learning_rate": 4.0905732673355514e-05,
+      "loss": 12.1255,
+      "step": 12183
+    },
+    {
+      "epoch": 2.1306666666666665,
+      "grad_norm": 0.8297474095433607,
+      "learning_rate": 4.0890502250282206e-05,
+      "loss": 12.0813,
+      "step": 12184
+    },
+    {
+      "epoch": 2.130841530054645,
+      "grad_norm": 0.8294431510248556,
+      "learning_rate": 4.087527393439339e-05,
+      "loss": 12.0517,
+      "step": 12185
+    },
+    {
+      "epoch": 2.131016393442623,
+      "grad_norm": 0.8072015650372023,
+      "learning_rate": 4.086004772623199e-05,
+      "loss": 12.0242,
+      "step": 12186
+    },
+    {
+      "epoch": 2.131191256830601,
+      "grad_norm": 0.8845534457468173,
+      "learning_rate": 4.084482362634078e-05,
+      "loss": 12.3073,
+      "step": 12187
+    },
+    {
+      "epoch": 2.1313661202185794,
+      "grad_norm": 0.84502602698676,
+      "learning_rate": 4.082960163526253e-05,
+      "loss": 12.0274,
+      "step": 12188
+    },
+    {
+      "epoch": 2.1315409836065573,
+      "grad_norm": 0.831882868995073,
+      "learning_rate": 4.0814381753539824e-05,
+      "loss": 12.2523,
+      "step": 12189
+    },
+    {
+      "epoch": 2.1317158469945356,
+      "grad_norm": 0.8148099352248124,
+      "learning_rate": 4.079916398171524e-05,
+      "loss": 11.8686,
+      "step": 12190
+    },
+    {
+      "epoch": 2.1318907103825135,
+      "grad_norm": 0.9360030077410563,
+      "learning_rate": 4.07839483203313e-05,
+      "loss": 12.3329,
+      "step": 12191
+    },
+    {
+      "epoch": 2.132065573770492,
+      "grad_norm": 0.8892689871663471,
+      "learning_rate": 4.076873476993042e-05,
+      "loss": 12.1221,
+      "step": 12192
+    },
+    {
+      "epoch": 2.13224043715847,
+      "grad_norm": 0.978559676147468,
+      "learning_rate": 4.075352333105493e-05,
+      "loss": 12.1065,
+      "step": 12193
+    },
+    {
+      "epoch": 2.132415300546448,
+      "grad_norm": 0.9364082657409455,
+      "learning_rate": 4.0738314004247126e-05,
+      "loss": 12.1772,
+      "step": 12194
+    },
+    {
+      "epoch": 2.1325901639344265,
+      "grad_norm": 0.8242151154151971,
+      "learning_rate": 4.072310679004921e-05,
+      "loss": 11.8017,
+      "step": 12195
+    },
+    {
+      "epoch": 2.1327650273224044,
+      "grad_norm": 0.7776899161042652,
+      "learning_rate": 4.070790168900326e-05,
+      "loss": 11.7972,
+      "step": 12196
+    },
+    {
+      "epoch": 2.1329398907103827,
+      "grad_norm": 0.9334259124331553,
+      "learning_rate": 4.069269870165132e-05,
+      "loss": 12.0419,
+      "step": 12197
+    },
+    {
+      "epoch": 2.1331147540983606,
+      "grad_norm": 0.9000849783138496,
+      "learning_rate": 4.067749782853539e-05,
+      "loss": 12.1737,
+      "step": 12198
+    },
+    {
+      "epoch": 2.133289617486339,
+      "grad_norm": 0.8661255161587893,
+      "learning_rate": 4.0662299070197384e-05,
+      "loss": 12.1666,
+      "step": 12199
+    },
+    {
+      "epoch": 2.133464480874317,
+      "grad_norm": 0.9102003305419814,
+      "learning_rate": 4.064710242717905e-05,
+      "loss": 12.0732,
+      "step": 12200
+    },
+    {
+      "epoch": 2.133639344262295,
+      "grad_norm": 0.9700039994715612,
+      "learning_rate": 4.063190790002218e-05,
+      "loss": 12.2608,
+      "step": 12201
+    },
+    {
+      "epoch": 2.133814207650273,
+      "grad_norm": 0.8055529895692661,
+      "learning_rate": 4.061671548926842e-05,
+      "loss": 12.2432,
+      "step": 12202
+    },
+    {
+      "epoch": 2.1339890710382514,
+      "grad_norm": 0.8446876349107239,
+      "learning_rate": 4.06015251954594e-05,
+      "loss": 11.8051,
+      "step": 12203
+    },
+    {
+      "epoch": 2.1341639344262293,
+      "grad_norm": 1.0602508810879376,
+      "learning_rate": 4.058633701913658e-05,
+      "loss": 11.9095,
+      "step": 12204
+    },
+    {
+      "epoch": 2.1343387978142077,
+      "grad_norm": 0.814927722488744,
+      "learning_rate": 4.057115096084143e-05,
+      "loss": 12.23,
+      "step": 12205
+    },
+    {
+      "epoch": 2.134513661202186,
+      "grad_norm": 0.8152568184476389,
+      "learning_rate": 4.0555967021115304e-05,
+      "loss": 12.0833,
+      "step": 12206
+    },
+    {
+      "epoch": 2.134688524590164,
+      "grad_norm": 0.8593180754479574,
+      "learning_rate": 4.054078520049951e-05,
+      "loss": 12.0297,
+      "step": 12207
+    },
+    {
+      "epoch": 2.1348633879781422,
+      "grad_norm": 0.8665775388572096,
+      "learning_rate": 4.0525605499535245e-05,
+      "loss": 12.1468,
+      "step": 12208
+    },
+    {
+      "epoch": 2.13503825136612,
+      "grad_norm": 0.8073961556374808,
+      "learning_rate": 4.051042791876366e-05,
+      "loss": 12.2524,
+      "step": 12209
+    },
+    {
+      "epoch": 2.1352131147540985,
+      "grad_norm": 0.8468354431152721,
+      "learning_rate": 4.049525245872585e-05,
+      "loss": 12.0383,
+      "step": 12210
+    },
+    {
+      "epoch": 2.1353879781420764,
+      "grad_norm": 0.8539261009905378,
+      "learning_rate": 4.048007911996272e-05,
+      "loss": 12.3443,
+      "step": 12211
+    },
+    {
+      "epoch": 2.1355628415300547,
+      "grad_norm": 0.8167390725413942,
+      "learning_rate": 4.046490790301525e-05,
+      "loss": 12.2272,
+      "step": 12212
+    },
+    {
+      "epoch": 2.1357377049180326,
+      "grad_norm": 0.8166815941884268,
+      "learning_rate": 4.0449738808424234e-05,
+      "loss": 12.1794,
+      "step": 12213
+    },
+    {
+      "epoch": 2.135912568306011,
+      "grad_norm": 0.8526182728269418,
+      "learning_rate": 4.0434571836730494e-05,
+      "loss": 11.8557,
+      "step": 12214
+    },
+    {
+      "epoch": 2.136087431693989,
+      "grad_norm": 0.8331338169928626,
+      "learning_rate": 4.041940698847464e-05,
+      "loss": 12.2594,
+      "step": 12215
+    },
+    {
+      "epoch": 2.136262295081967,
+      "grad_norm": 0.8327281986511244,
+      "learning_rate": 4.040424426419731e-05,
+      "loss": 11.9378,
+      "step": 12216
+    },
+    {
+      "epoch": 2.1364371584699455,
+      "grad_norm": 0.7622790292666045,
+      "learning_rate": 4.038908366443904e-05,
+      "loss": 11.873,
+      "step": 12217
+    },
+    {
+      "epoch": 2.1366120218579234,
+      "grad_norm": 0.8609540683841915,
+      "learning_rate": 4.037392518974034e-05,
+      "loss": 12.0846,
+      "step": 12218
+    },
+    {
+      "epoch": 2.1367868852459018,
+      "grad_norm": 0.8789394989026239,
+      "learning_rate": 4.035876884064149e-05,
+      "loss": 12.1225,
+      "step": 12219
+    },
+    {
+      "epoch": 2.1369617486338797,
+      "grad_norm": 0.7893030738070597,
+      "learning_rate": 4.0343614617682855e-05,
+      "loss": 11.9696,
+      "step": 12220
+    },
+    {
+      "epoch": 2.137136612021858,
+      "grad_norm": 0.7835918987685337,
+      "learning_rate": 4.0328462521404684e-05,
+      "loss": 12.1029,
+      "step": 12221
+    },
+    {
+      "epoch": 2.137311475409836,
+      "grad_norm": 0.8498117974804291,
+      "learning_rate": 4.031331255234708e-05,
+      "loss": 12.3353,
+      "step": 12222
+    },
+    {
+      "epoch": 2.1374863387978142,
+      "grad_norm": 0.8027040837426086,
+      "learning_rate": 4.0298164711050144e-05,
+      "loss": 12.1092,
+      "step": 12223
+    },
+    {
+      "epoch": 2.137661202185792,
+      "grad_norm": 0.8646139411785132,
+      "learning_rate": 4.028301899805389e-05,
+      "loss": 11.9536,
+      "step": 12224
+    },
+    {
+      "epoch": 2.1378360655737705,
+      "grad_norm": 0.9059285351653328,
+      "learning_rate": 4.0267875413898224e-05,
+      "loss": 11.9601,
+      "step": 12225
+    },
+    {
+      "epoch": 2.138010928961749,
+      "grad_norm": 0.8072539295092409,
+      "learning_rate": 4.025273395912301e-05,
+      "loss": 12.0884,
+      "step": 12226
+    },
+    {
+      "epoch": 2.1381857923497267,
+      "grad_norm": 0.9341845356368864,
+      "learning_rate": 4.023759463426804e-05,
+      "loss": 12.1225,
+      "step": 12227
+    },
+    {
+      "epoch": 2.138360655737705,
+      "grad_norm": 0.8232963516670224,
+      "learning_rate": 4.022245743987298e-05,
+      "loss": 12.119,
+      "step": 12228
+    },
+    {
+      "epoch": 2.138535519125683,
+      "grad_norm": 0.8183453442321661,
+      "learning_rate": 4.0207322376477506e-05,
+      "loss": 11.8558,
+      "step": 12229
+    },
+    {
+      "epoch": 2.1387103825136613,
+      "grad_norm": 0.8644710434932642,
+      "learning_rate": 4.01921894446211e-05,
+      "loss": 12.203,
+      "step": 12230
+    },
+    {
+      "epoch": 2.138885245901639,
+      "grad_norm": 0.9238144643546993,
+      "learning_rate": 4.017705864484326e-05,
+      "loss": 11.9577,
+      "step": 12231
+    },
+    {
+      "epoch": 2.1390601092896175,
+      "grad_norm": 0.8967244319970029,
+      "learning_rate": 4.016192997768339e-05,
+      "loss": 12.1025,
+      "step": 12232
+    },
+    {
+      "epoch": 2.1392349726775954,
+      "grad_norm": 0.8824191143681033,
+      "learning_rate": 4.0146803443680826e-05,
+      "loss": 12.2799,
+      "step": 12233
+    },
+    {
+      "epoch": 2.1394098360655738,
+      "grad_norm": 0.9232032620274444,
+      "learning_rate": 4.013167904337476e-05,
+      "loss": 12.086,
+      "step": 12234
+    },
+    {
+      "epoch": 2.139584699453552,
+      "grad_norm": 0.8783065197028098,
+      "learning_rate": 4.011655677730438e-05,
+      "loss": 11.9639,
+      "step": 12235
+    },
+    {
+      "epoch": 2.13975956284153,
+      "grad_norm": 0.8199544308775416,
+      "learning_rate": 4.010143664600883e-05,
+      "loss": 12.1196,
+      "step": 12236
+    },
+    {
+      "epoch": 2.1399344262295084,
+      "grad_norm": 0.9549493758860148,
+      "learning_rate": 4.008631865002704e-05,
+      "loss": 12.177,
+      "step": 12237
+    },
+    {
+      "epoch": 2.1401092896174863,
+      "grad_norm": 0.8126602273195473,
+      "learning_rate": 4.007120278989799e-05,
+      "loss": 12.0911,
+      "step": 12238
+    },
+    {
+      "epoch": 2.1402841530054646,
+      "grad_norm": 0.8127821419743475,
+      "learning_rate": 4.005608906616054e-05,
+      "loss": 12.1875,
+      "step": 12239
+    },
+    {
+      "epoch": 2.1404590163934425,
+      "grad_norm": 0.8837033400885511,
+      "learning_rate": 4.004097747935351e-05,
+      "loss": 12.0635,
+      "step": 12240
+    },
+    {
+      "epoch": 2.140633879781421,
+      "grad_norm": 0.8143997467637608,
+      "learning_rate": 4.002586803001556e-05,
+      "loss": 12.0706,
+      "step": 12241
+    },
+    {
+      "epoch": 2.1408087431693987,
+      "grad_norm": 0.9086932577347775,
+      "learning_rate": 4.001076071868533e-05,
+      "loss": 12.2515,
+      "step": 12242
+    },
+    {
+      "epoch": 2.140983606557377,
+      "grad_norm": 0.8754722156099146,
+      "learning_rate": 3.9995655545901375e-05,
+      "loss": 12.1889,
+      "step": 12243
+    },
+    {
+      "epoch": 2.1411584699453554,
+      "grad_norm": 0.8098027266110391,
+      "learning_rate": 3.998055251220221e-05,
+      "loss": 12.1279,
+      "step": 12244
+    },
+    {
+      "epoch": 2.1413333333333333,
+      "grad_norm": 0.8316557744996524,
+      "learning_rate": 3.9965451618126225e-05,
+      "loss": 12.0644,
+      "step": 12245
+    },
+    {
+      "epoch": 2.1415081967213117,
+      "grad_norm": 0.8708713723015055,
+      "learning_rate": 3.9950352864211736e-05,
+      "loss": 12.0685,
+      "step": 12246
+    },
+    {
+      "epoch": 2.1416830601092895,
+      "grad_norm": 0.8424767084795344,
+      "learning_rate": 3.993525625099701e-05,
+      "loss": 12.1018,
+      "step": 12247
+    },
+    {
+      "epoch": 2.141857923497268,
+      "grad_norm": 0.8425411238184324,
+      "learning_rate": 3.992016177902026e-05,
+      "loss": 11.8428,
+      "step": 12248
+    },
+    {
+      "epoch": 2.142032786885246,
+      "grad_norm": 0.8635143023696542,
+      "learning_rate": 3.990506944881951e-05,
+      "loss": 11.9952,
+      "step": 12249
+    },
+    {
+      "epoch": 2.142207650273224,
+      "grad_norm": 0.7968626333901472,
+      "learning_rate": 3.988997926093282e-05,
+      "loss": 12.0238,
+      "step": 12250
+    },
+    {
+      "epoch": 2.142382513661202,
+      "grad_norm": 0.8453687976559802,
+      "learning_rate": 3.9874891215898136e-05,
+      "loss": 12.1959,
+      "step": 12251
+    },
+    {
+      "epoch": 2.1425573770491804,
+      "grad_norm": 0.876201848798529,
+      "learning_rate": 3.985980531425337e-05,
+      "loss": 11.9257,
+      "step": 12252
+    },
+    {
+      "epoch": 2.1427322404371583,
+      "grad_norm": 0.9509821802548178,
+      "learning_rate": 3.984472155653625e-05,
+      "loss": 11.9493,
+      "step": 12253
+    },
+    {
+      "epoch": 2.1429071038251366,
+      "grad_norm": 0.8626071238558182,
+      "learning_rate": 3.9829639943284515e-05,
+      "loss": 12.1085,
+      "step": 12254
+    },
+    {
+      "epoch": 2.143081967213115,
+      "grad_norm": 1.0128075998333887,
+      "learning_rate": 3.9814560475035844e-05,
+      "loss": 12.1304,
+      "step": 12255
+    },
+    {
+      "epoch": 2.143256830601093,
+      "grad_norm": 0.83373405616906,
+      "learning_rate": 3.9799483152327746e-05,
+      "loss": 12.0398,
+      "step": 12256
+    },
+    {
+      "epoch": 2.143431693989071,
+      "grad_norm": 0.9224611500308549,
+      "learning_rate": 3.978440797569774e-05,
+      "loss": 12.0523,
+      "step": 12257
+    },
+    {
+      "epoch": 2.143606557377049,
+      "grad_norm": 0.9987146197027341,
+      "learning_rate": 3.976933494568322e-05,
+      "loss": 12.2457,
+      "step": 12258
+    },
+    {
+      "epoch": 2.1437814207650274,
+      "grad_norm": 0.9011373130307292,
+      "learning_rate": 3.975426406282159e-05,
+      "loss": 12.2151,
+      "step": 12259
+    },
+    {
+      "epoch": 2.1439562841530053,
+      "grad_norm": 0.8410320138363152,
+      "learning_rate": 3.973919532765002e-05,
+      "loss": 11.9977,
+      "step": 12260
+    },
+    {
+      "epoch": 2.1441311475409837,
+      "grad_norm": 1.0052689365819156,
+      "learning_rate": 3.972412874070571e-05,
+      "loss": 12.2774,
+      "step": 12261
+    },
+    {
+      "epoch": 2.1443060109289616,
+      "grad_norm": 1.0504691632279592,
+      "learning_rate": 3.9709064302525803e-05,
+      "loss": 11.9707,
+      "step": 12262
+    },
+    {
+      "epoch": 2.14448087431694,
+      "grad_norm": 0.8302167636999515,
+      "learning_rate": 3.96940020136473e-05,
+      "loss": 12.0461,
+      "step": 12263
+    },
+    {
+      "epoch": 2.144655737704918,
+      "grad_norm": 0.8984787725076211,
+      "learning_rate": 3.967894187460718e-05,
+      "loss": 12.0052,
+      "step": 12264
+    },
+    {
+      "epoch": 2.144830601092896,
+      "grad_norm": 0.9237677079528185,
+      "learning_rate": 3.96638838859423e-05,
+      "loss": 11.7745,
+      "step": 12265
+    },
+    {
+      "epoch": 2.1450054644808745,
+      "grad_norm": 0.8298984438542139,
+      "learning_rate": 3.9648828048189455e-05,
+      "loss": 11.9992,
+      "step": 12266
+    },
+    {
+      "epoch": 2.1451803278688524,
+      "grad_norm": 0.8450968321823507,
+      "learning_rate": 3.9633774361885423e-05,
+      "loss": 11.991,
+      "step": 12267
+    },
+    {
+      "epoch": 2.1453551912568307,
+      "grad_norm": 0.8222432483129412,
+      "learning_rate": 3.9618722827566755e-05,
+      "loss": 12.1484,
+      "step": 12268
+    },
+    {
+      "epoch": 2.1455300546448086,
+      "grad_norm": 1.0078016873250266,
+      "learning_rate": 3.9603673445770083e-05,
+      "loss": 12.1536,
+      "step": 12269
+    },
+    {
+      "epoch": 2.145704918032787,
+      "grad_norm": 0.8253906610746427,
+      "learning_rate": 3.958862621703192e-05,
+      "loss": 11.9096,
+      "step": 12270
+    },
+    {
+      "epoch": 2.145879781420765,
+      "grad_norm": 0.8397697574591424,
+      "learning_rate": 3.9573581141888615e-05,
+      "loss": 12.1394,
+      "step": 12271
+    },
+    {
+      "epoch": 2.146054644808743,
+      "grad_norm": 0.8100219186359421,
+      "learning_rate": 3.955853822087654e-05,
+      "loss": 11.9132,
+      "step": 12272
+    },
+    {
+      "epoch": 2.146229508196721,
+      "grad_norm": 0.8768190098141944,
+      "learning_rate": 3.9543497454531964e-05,
+      "loss": 12.1166,
+      "step": 12273
+    },
+    {
+      "epoch": 2.1464043715846994,
+      "grad_norm": 0.8122670256494077,
+      "learning_rate": 3.952845884339109e-05,
+      "loss": 12.0528,
+      "step": 12274
+    },
+    {
+      "epoch": 2.1465792349726778,
+      "grad_norm": 0.8064942853917108,
+      "learning_rate": 3.951342238798999e-05,
+      "loss": 12.1654,
+      "step": 12275
+    },
+    {
+      "epoch": 2.1467540983606557,
+      "grad_norm": 0.9631728016469966,
+      "learning_rate": 3.949838808886469e-05,
+      "loss": 12.2347,
+      "step": 12276
+    },
+    {
+      "epoch": 2.146928961748634,
+      "grad_norm": 0.9628419457315419,
+      "learning_rate": 3.948335594655118e-05,
+      "loss": 12.115,
+      "step": 12277
+    },
+    {
+      "epoch": 2.147103825136612,
+      "grad_norm": 0.8960194677251645,
+      "learning_rate": 3.946832596158535e-05,
+      "loss": 12.2176,
+      "step": 12278
+    },
+    {
+      "epoch": 2.1472786885245903,
+      "grad_norm": 0.9249985315775112,
+      "learning_rate": 3.945329813450295e-05,
+      "loss": 12.12,
+      "step": 12279
+    },
+    {
+      "epoch": 2.147453551912568,
+      "grad_norm": 0.8736958791277655,
+      "learning_rate": 3.943827246583972e-05,
+      "loss": 12.1909,
+      "step": 12280
+    },
+    {
+      "epoch": 2.1476284153005465,
+      "grad_norm": 0.8233932294896839,
+      "learning_rate": 3.942324895613132e-05,
+      "loss": 11.9768,
+      "step": 12281
+    },
+    {
+      "epoch": 2.1478032786885244,
+      "grad_norm": 0.9360750567232106,
+      "learning_rate": 3.940822760591333e-05,
+      "loss": 11.8983,
+      "step": 12282
+    },
+    {
+      "epoch": 2.1479781420765027,
+      "grad_norm": 0.792599136320049,
+      "learning_rate": 3.939320841572122e-05,
+      "loss": 12.1209,
+      "step": 12283
+    },
+    {
+      "epoch": 2.148153005464481,
+      "grad_norm": 0.8273905385274052,
+      "learning_rate": 3.937819138609042e-05,
+      "loss": 12.1057,
+      "step": 12284
+    },
+    {
+      "epoch": 2.148327868852459,
+      "grad_norm": 0.8194381773643723,
+      "learning_rate": 3.936317651755631e-05,
+      "loss": 12.2028,
+      "step": 12285
+    },
+    {
+      "epoch": 2.1485027322404373,
+      "grad_norm": 0.8305875612592168,
+      "learning_rate": 3.934816381065407e-05,
+      "loss": 11.9268,
+      "step": 12286
+    },
+    {
+      "epoch": 2.148677595628415,
+      "grad_norm": 0.8120547237066168,
+      "learning_rate": 3.933315326591892e-05,
+      "loss": 12.0479,
+      "step": 12287
+    },
+    {
+      "epoch": 2.1488524590163935,
+      "grad_norm": 0.8752687062756533,
+      "learning_rate": 3.931814488388599e-05,
+      "loss": 12.3005,
+      "step": 12288
+    },
+    {
+      "epoch": 2.1490273224043714,
+      "grad_norm": 0.7686391511311305,
+      "learning_rate": 3.930313866509032e-05,
+      "loss": 11.9538,
+      "step": 12289
+    },
+    {
+      "epoch": 2.14920218579235,
+      "grad_norm": 0.8442470473942454,
+      "learning_rate": 3.928813461006681e-05,
+      "loss": 12.17,
+      "step": 12290
+    },
+    {
+      "epoch": 2.1493770491803277,
+      "grad_norm": 0.7913585618960675,
+      "learning_rate": 3.9273132719350356e-05,
+      "loss": 12.0998,
+      "step": 12291
+    },
+    {
+      "epoch": 2.149551912568306,
+      "grad_norm": 0.9046445115207745,
+      "learning_rate": 3.925813299347578e-05,
+      "loss": 12.1911,
+      "step": 12292
+    },
+    {
+      "epoch": 2.1497267759562844,
+      "grad_norm": 0.8385719076676562,
+      "learning_rate": 3.924313543297783e-05,
+      "loss": 11.9724,
+      "step": 12293
+    },
+    {
+      "epoch": 2.1499016393442623,
+      "grad_norm": 0.875360679061945,
+      "learning_rate": 3.922814003839107e-05,
+      "loss": 12.2194,
+      "step": 12294
+    },
+    {
+      "epoch": 2.1500765027322406,
+      "grad_norm": 0.8233488353898644,
+      "learning_rate": 3.9213146810250126e-05,
+      "loss": 12.1738,
+      "step": 12295
+    },
+    {
+      "epoch": 2.1502513661202185,
+      "grad_norm": 0.9032061459283371,
+      "learning_rate": 3.919815574908946e-05,
+      "loss": 12.171,
+      "step": 12296
+    },
+    {
+      "epoch": 2.150426229508197,
+      "grad_norm": 0.8437551916237183,
+      "learning_rate": 3.918316685544351e-05,
+      "loss": 11.8657,
+      "step": 12297
+    },
+    {
+      "epoch": 2.1506010928961747,
+      "grad_norm": 0.8968178051175065,
+      "learning_rate": 3.916818012984661e-05,
+      "loss": 12.1337,
+      "step": 12298
+    },
+    {
+      "epoch": 2.150775956284153,
+      "grad_norm": 0.8207787847441221,
+      "learning_rate": 3.915319557283301e-05,
+      "loss": 11.8147,
+      "step": 12299
+    },
+    {
+      "epoch": 2.150950819672131,
+      "grad_norm": 0.8662785516574382,
+      "learning_rate": 3.913821318493693e-05,
+      "loss": 12.0981,
+      "step": 12300
+    },
+    {
+      "epoch": 2.1511256830601093,
+      "grad_norm": 0.8298793405434042,
+      "learning_rate": 3.912323296669242e-05,
+      "loss": 11.7365,
+      "step": 12301
+    },
+    {
+      "epoch": 2.1513005464480877,
+      "grad_norm": 0.8247401924489834,
+      "learning_rate": 3.91082549186335e-05,
+      "loss": 12.0919,
+      "step": 12302
+    },
+    {
+      "epoch": 2.1514754098360656,
+      "grad_norm": 0.803387224272253,
+      "learning_rate": 3.909327904129416e-05,
+      "loss": 12.1615,
+      "step": 12303
+    },
+    {
+      "epoch": 2.151650273224044,
+      "grad_norm": 0.8548298930237938,
+      "learning_rate": 3.90783053352083e-05,
+      "loss": 12.1081,
+      "step": 12304
+    },
+    {
+      "epoch": 2.151825136612022,
+      "grad_norm": 0.8928777676655609,
+      "learning_rate": 3.906333380090964e-05,
+      "loss": 11.8562,
+      "step": 12305
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 0.8208560704555234,
+      "learning_rate": 3.904836443893193e-05,
+      "loss": 12.103,
+      "step": 12306
+    },
+    {
+      "epoch": 2.152174863387978,
+      "grad_norm": 0.8559086643939443,
+      "learning_rate": 3.9033397249808823e-05,
+      "loss": 12.0766,
+      "step": 12307
+    },
+    {
+      "epoch": 2.1523497267759564,
+      "grad_norm": 0.8840696140641798,
+      "learning_rate": 3.9018432234073896e-05,
+      "loss": 12.1304,
+      "step": 12308
+    },
+    {
+      "epoch": 2.1525245901639343,
+      "grad_norm": 0.9183769912974346,
+      "learning_rate": 3.9003469392260574e-05,
+      "loss": 12.1767,
+      "step": 12309
+    },
+    {
+      "epoch": 2.1526994535519126,
+      "grad_norm": 0.8860963666952577,
+      "learning_rate": 3.8988508724902305e-05,
+      "loss": 12.2287,
+      "step": 12310
+    },
+    {
+      "epoch": 2.1528743169398905,
+      "grad_norm": 0.8204400175654781,
+      "learning_rate": 3.897355023253242e-05,
+      "loss": 12.0574,
+      "step": 12311
+    },
+    {
+      "epoch": 2.153049180327869,
+      "grad_norm": 0.7682388587465341,
+      "learning_rate": 3.895859391568419e-05,
+      "loss": 12.2176,
+      "step": 12312
+    },
+    {
+      "epoch": 2.153224043715847,
+      "grad_norm": 0.8522005398705343,
+      "learning_rate": 3.894363977489075e-05,
+      "loss": 12.2417,
+      "step": 12313
+    },
+    {
+      "epoch": 2.153398907103825,
+      "grad_norm": 0.7835447133833278,
+      "learning_rate": 3.89286878106852e-05,
+      "loss": 12.0815,
+      "step": 12314
+    },
+    {
+      "epoch": 2.1535737704918034,
+      "grad_norm": 0.8485576050003056,
+      "learning_rate": 3.8913738023600585e-05,
+      "loss": 12.3007,
+      "step": 12315
+    },
+    {
+      "epoch": 2.1537486338797813,
+      "grad_norm": 0.8544443731268925,
+      "learning_rate": 3.889879041416984e-05,
+      "loss": 12.1216,
+      "step": 12316
+    },
+    {
+      "epoch": 2.1539234972677597,
+      "grad_norm": 0.7855770815456335,
+      "learning_rate": 3.888384498292582e-05,
+      "loss": 11.8788,
+      "step": 12317
+    },
+    {
+      "epoch": 2.1540983606557376,
+      "grad_norm": 0.8554299595415479,
+      "learning_rate": 3.886890173040133e-05,
+      "loss": 12.039,
+      "step": 12318
+    },
+    {
+      "epoch": 2.154273224043716,
+      "grad_norm": 0.8273319512831667,
+      "learning_rate": 3.88539606571291e-05,
+      "loss": 12.045,
+      "step": 12319
+    },
+    {
+      "epoch": 2.154448087431694,
+      "grad_norm": 0.7948030586889114,
+      "learning_rate": 3.883902176364168e-05,
+      "loss": 12.2002,
+      "step": 12320
+    },
+    {
+      "epoch": 2.154622950819672,
+      "grad_norm": 0.8536809868378867,
+      "learning_rate": 3.882408505047169e-05,
+      "loss": 12.1379,
+      "step": 12321
+    },
+    {
+      "epoch": 2.15479781420765,
+      "grad_norm": 0.8649982848182547,
+      "learning_rate": 3.880915051815159e-05,
+      "loss": 12.1358,
+      "step": 12322
+    },
+    {
+      "epoch": 2.1549726775956284,
+      "grad_norm": 0.8308378886569011,
+      "learning_rate": 3.8794218167213816e-05,
+      "loss": 11.9662,
+      "step": 12323
+    },
+    {
+      "epoch": 2.1551475409836067,
+      "grad_norm": 0.8329189806453073,
+      "learning_rate": 3.877928799819062e-05,
+      "loss": 12.1343,
+      "step": 12324
+    },
+    {
+      "epoch": 2.1553224043715846,
+      "grad_norm": 0.8723048230941424,
+      "learning_rate": 3.8764360011614273e-05,
+      "loss": 12.1168,
+      "step": 12325
+    },
+    {
+      "epoch": 2.155497267759563,
+      "grad_norm": 0.8332421160000681,
+      "learning_rate": 3.874943420801695e-05,
+      "loss": 12.1843,
+      "step": 12326
+    },
+    {
+      "epoch": 2.155672131147541,
+      "grad_norm": 0.8615888981618689,
+      "learning_rate": 3.873451058793076e-05,
+      "loss": 12.124,
+      "step": 12327
+    },
+    {
+      "epoch": 2.155846994535519,
+      "grad_norm": 0.8938924230290737,
+      "learning_rate": 3.8719589151887655e-05,
+      "loss": 11.9991,
+      "step": 12328
+    },
+    {
+      "epoch": 2.156021857923497,
+      "grad_norm": 0.8527977869079857,
+      "learning_rate": 3.870466990041961e-05,
+      "loss": 12.2811,
+      "step": 12329
+    },
+    {
+      "epoch": 2.1561967213114754,
+      "grad_norm": 0.8884558347040574,
+      "learning_rate": 3.868975283405849e-05,
+      "loss": 12.1677,
+      "step": 12330
+    },
+    {
+      "epoch": 2.1563715846994533,
+      "grad_norm": 0.8112442365671289,
+      "learning_rate": 3.867483795333602e-05,
+      "loss": 12.1769,
+      "step": 12331
+    },
+    {
+      "epoch": 2.1565464480874317,
+      "grad_norm": 0.8392383510911844,
+      "learning_rate": 3.865992525878392e-05,
+      "loss": 12.1805,
+      "step": 12332
+    },
+    {
+      "epoch": 2.15672131147541,
+      "grad_norm": 1.0069825411262874,
+      "learning_rate": 3.864501475093382e-05,
+      "loss": 12.1977,
+      "step": 12333
+    },
+    {
+      "epoch": 2.156896174863388,
+      "grad_norm": 0.9312626000810306,
+      "learning_rate": 3.863010643031726e-05,
+      "loss": 11.9888,
+      "step": 12334
+    },
+    {
+      "epoch": 2.1570710382513663,
+      "grad_norm": 0.9483846302618033,
+      "learning_rate": 3.86152002974657e-05,
+      "loss": 12.3039,
+      "step": 12335
+    },
+    {
+      "epoch": 2.157245901639344,
+      "grad_norm": 0.8365880550929337,
+      "learning_rate": 3.8600296352910524e-05,
+      "loss": 12.1313,
+      "step": 12336
+    },
+    {
+      "epoch": 2.1574207650273225,
+      "grad_norm": 0.826575588085355,
+      "learning_rate": 3.858539459718306e-05,
+      "loss": 12.0574,
+      "step": 12337
+    },
+    {
+      "epoch": 2.1575956284153004,
+      "grad_norm": 0.8875063862827994,
+      "learning_rate": 3.857049503081455e-05,
+      "loss": 12.1218,
+      "step": 12338
+    },
+    {
+      "epoch": 2.1577704918032787,
+      "grad_norm": 0.8487365837871486,
+      "learning_rate": 3.8555597654336075e-05,
+      "loss": 12.1201,
+      "step": 12339
+    },
+    {
+      "epoch": 2.1579453551912566,
+      "grad_norm": 0.8245459783306424,
+      "learning_rate": 3.8540702468278764e-05,
+      "loss": 12.2095,
+      "step": 12340
+    },
+    {
+      "epoch": 2.158120218579235,
+      "grad_norm": 0.823817910018912,
+      "learning_rate": 3.8525809473173605e-05,
+      "loss": 11.9332,
+      "step": 12341
+    },
+    {
+      "epoch": 2.1582950819672133,
+      "grad_norm": 0.8492799357651072,
+      "learning_rate": 3.8510918669551545e-05,
+      "loss": 11.9523,
+      "step": 12342
+    },
+    {
+      "epoch": 2.158469945355191,
+      "grad_norm": 0.8170480933621804,
+      "learning_rate": 3.8496030057943356e-05,
+      "loss": 12.0978,
+      "step": 12343
+    },
+    {
+      "epoch": 2.1586448087431696,
+      "grad_norm": 0.8272228154981064,
+      "learning_rate": 3.8481143638879836e-05,
+      "loss": 12.2384,
+      "step": 12344
+    },
+    {
+      "epoch": 2.1588196721311474,
+      "grad_norm": 0.8575358693371642,
+      "learning_rate": 3.8466259412891695e-05,
+      "loss": 11.8624,
+      "step": 12345
+    },
+    {
+      "epoch": 2.158994535519126,
+      "grad_norm": 0.8448992931121758,
+      "learning_rate": 3.8451377380509493e-05,
+      "loss": 12.0454,
+      "step": 12346
+    },
+    {
+      "epoch": 2.1591693989071037,
+      "grad_norm": 0.8539436603565668,
+      "learning_rate": 3.843649754226376e-05,
+      "loss": 12.1298,
+      "step": 12347
+    },
+    {
+      "epoch": 2.159344262295082,
+      "grad_norm": 0.8493744755841435,
+      "learning_rate": 3.8421619898684966e-05,
+      "loss": 12.1762,
+      "step": 12348
+    },
+    {
+      "epoch": 2.15951912568306,
+      "grad_norm": 0.8183592211040935,
+      "learning_rate": 3.840674445030351e-05,
+      "loss": 12.2078,
+      "step": 12349
+    },
+    {
+      "epoch": 2.1596939890710383,
+      "grad_norm": 0.9241501566714664,
+      "learning_rate": 3.839187119764962e-05,
+      "loss": 12.0995,
+      "step": 12350
+    },
+    {
+      "epoch": 2.1598688524590166,
+      "grad_norm": 0.8129890339161848,
+      "learning_rate": 3.837700014125354e-05,
+      "loss": 12.1437,
+      "step": 12351
+    },
+    {
+      "epoch": 2.1600437158469945,
+      "grad_norm": 0.806504663855038,
+      "learning_rate": 3.836213128164541e-05,
+      "loss": 12.0025,
+      "step": 12352
+    },
+    {
+      "epoch": 2.160218579234973,
+      "grad_norm": 0.8815159157948905,
+      "learning_rate": 3.8347264619355275e-05,
+      "loss": 12.1991,
+      "step": 12353
+    },
+    {
+      "epoch": 2.1603934426229507,
+      "grad_norm": 0.8433578632896673,
+      "learning_rate": 3.833240015491313e-05,
+      "loss": 12.1238,
+      "step": 12354
+    },
+    {
+      "epoch": 2.160568306010929,
+      "grad_norm": 0.8531962436747548,
+      "learning_rate": 3.831753788884885e-05,
+      "loss": 12.0096,
+      "step": 12355
+    },
+    {
+      "epoch": 2.160743169398907,
+      "grad_norm": 0.8772231432173621,
+      "learning_rate": 3.83026778216923e-05,
+      "loss": 11.9246,
+      "step": 12356
+    },
+    {
+      "epoch": 2.1609180327868853,
+      "grad_norm": 0.7570404140025239,
+      "learning_rate": 3.828781995397322e-05,
+      "loss": 12.0567,
+      "step": 12357
+    },
+    {
+      "epoch": 2.161092896174863,
+      "grad_norm": 0.8668038014806753,
+      "learning_rate": 3.8272964286221234e-05,
+      "loss": 12.1245,
+      "step": 12358
+    },
+    {
+      "epoch": 2.1612677595628416,
+      "grad_norm": 0.8863643025310195,
+      "learning_rate": 3.825811081896594e-05,
+      "loss": 11.7774,
+      "step": 12359
+    },
+    {
+      "epoch": 2.16144262295082,
+      "grad_norm": 0.8464853476593398,
+      "learning_rate": 3.82432595527369e-05,
+      "loss": 11.9857,
+      "step": 12360
+    },
+    {
+      "epoch": 2.161617486338798,
+      "grad_norm": 0.828172167119074,
+      "learning_rate": 3.822841048806346e-05,
+      "loss": 11.8507,
+      "step": 12361
+    },
+    {
+      "epoch": 2.161792349726776,
+      "grad_norm": 0.8432668345063999,
+      "learning_rate": 3.8213563625475015e-05,
+      "loss": 11.8605,
+      "step": 12362
+    },
+    {
+      "epoch": 2.161967213114754,
+      "grad_norm": 0.7920550928611596,
+      "learning_rate": 3.819871896550083e-05,
+      "loss": 12.0983,
+      "step": 12363
+    },
+    {
+      "epoch": 2.1621420765027324,
+      "grad_norm": 0.8366671130003107,
+      "learning_rate": 3.818387650867016e-05,
+      "loss": 12.0848,
+      "step": 12364
+    },
+    {
+      "epoch": 2.1623169398907103,
+      "grad_norm": 0.8516229370317611,
+      "learning_rate": 3.816903625551203e-05,
+      "loss": 12.1759,
+      "step": 12365
+    },
+    {
+      "epoch": 2.1624918032786886,
+      "grad_norm": 0.8805877288961056,
+      "learning_rate": 3.8154198206555506e-05,
+      "loss": 11.9086,
+      "step": 12366
+    },
+    {
+      "epoch": 2.1626666666666665,
+      "grad_norm": 0.8577414535533796,
+      "learning_rate": 3.813936236232956e-05,
+      "loss": 11.9614,
+      "step": 12367
+    },
+    {
+      "epoch": 2.162841530054645,
+      "grad_norm": 0.8276104899369088,
+      "learning_rate": 3.8124528723363095e-05,
+      "loss": 11.7867,
+      "step": 12368
+    },
+    {
+      "epoch": 2.1630163934426228,
+      "grad_norm": 0.8958534056676638,
+      "learning_rate": 3.8109697290184865e-05,
+      "loss": 12.2958,
+      "step": 12369
+    },
+    {
+      "epoch": 2.163191256830601,
+      "grad_norm": 0.9111986235279929,
+      "learning_rate": 3.809486806332362e-05,
+      "loss": 11.9419,
+      "step": 12370
+    },
+    {
+      "epoch": 2.1633661202185794,
+      "grad_norm": 0.8068836680521363,
+      "learning_rate": 3.808004104330799e-05,
+      "loss": 12.1084,
+      "step": 12371
+    },
+    {
+      "epoch": 2.1635409836065573,
+      "grad_norm": 0.7920601277849434,
+      "learning_rate": 3.8065216230666546e-05,
+      "loss": 12.0901,
+      "step": 12372
+    },
+    {
+      "epoch": 2.1637158469945357,
+      "grad_norm": 0.7964240487360539,
+      "learning_rate": 3.8050393625927795e-05,
+      "loss": 12.0426,
+      "step": 12373
+    },
+    {
+      "epoch": 2.1638907103825136,
+      "grad_norm": 0.8146042063392415,
+      "learning_rate": 3.803557322962013e-05,
+      "loss": 12.1229,
+      "step": 12374
+    },
+    {
+      "epoch": 2.164065573770492,
+      "grad_norm": 0.9616582862437453,
+      "learning_rate": 3.8020755042271874e-05,
+      "loss": 12.2822,
+      "step": 12375
+    },
+    {
+      "epoch": 2.16424043715847,
+      "grad_norm": 0.9096594306287193,
+      "learning_rate": 3.8005939064411314e-05,
+      "loss": 12.1597,
+      "step": 12376
+    },
+    {
+      "epoch": 2.164415300546448,
+      "grad_norm": 0.8135644597562823,
+      "learning_rate": 3.799112529656657e-05,
+      "loss": 12.1205,
+      "step": 12377
+    },
+    {
+      "epoch": 2.164590163934426,
+      "grad_norm": 0.8096693555826145,
+      "learning_rate": 3.797631373926576e-05,
+      "loss": 11.8651,
+      "step": 12378
+    },
+    {
+      "epoch": 2.1647650273224044,
+      "grad_norm": 0.861959675474802,
+      "learning_rate": 3.796150439303693e-05,
+      "loss": 12.1365,
+      "step": 12379
+    },
+    {
+      "epoch": 2.1649398907103823,
+      "grad_norm": 0.8629969896378262,
+      "learning_rate": 3.7946697258407945e-05,
+      "loss": 11.9415,
+      "step": 12380
+    },
+    {
+      "epoch": 2.1651147540983606,
+      "grad_norm": 0.8528866372170659,
+      "learning_rate": 3.7931892335906704e-05,
+      "loss": 11.8839,
+      "step": 12381
+    },
+    {
+      "epoch": 2.165289617486339,
+      "grad_norm": 0.8709864930635097,
+      "learning_rate": 3.7917089626060984e-05,
+      "loss": 12.1157,
+      "step": 12382
+    },
+    {
+      "epoch": 2.165464480874317,
+      "grad_norm": 0.8321239875147706,
+      "learning_rate": 3.7902289129398514e-05,
+      "loss": 11.9794,
+      "step": 12383
+    },
+    {
+      "epoch": 2.165639344262295,
+      "grad_norm": 0.8827306503864164,
+      "learning_rate": 3.788749084644686e-05,
+      "loss": 11.8957,
+      "step": 12384
+    },
+    {
+      "epoch": 2.165814207650273,
+      "grad_norm": 0.839268311744715,
+      "learning_rate": 3.7872694777733575e-05,
+      "loss": 11.95,
+      "step": 12385
+    },
+    {
+      "epoch": 2.1659890710382514,
+      "grad_norm": 0.8837744567258037,
+      "learning_rate": 3.7857900923786135e-05,
+      "loss": 11.868,
+      "step": 12386
+    },
+    {
+      "epoch": 2.1661639344262293,
+      "grad_norm": 0.7962631656930127,
+      "learning_rate": 3.784310928513196e-05,
+      "loss": 11.8444,
+      "step": 12387
+    },
+    {
+      "epoch": 2.1663387978142077,
+      "grad_norm": 0.8605407715510459,
+      "learning_rate": 3.782831986229828e-05,
+      "loss": 11.8674,
+      "step": 12388
+    },
+    {
+      "epoch": 2.1665136612021856,
+      "grad_norm": 0.8317645793959927,
+      "learning_rate": 3.781353265581235e-05,
+      "loss": 12.0101,
+      "step": 12389
+    },
+    {
+      "epoch": 2.166688524590164,
+      "grad_norm": 0.964766798864091,
+      "learning_rate": 3.7798747666201337e-05,
+      "loss": 12.2366,
+      "step": 12390
+    },
+    {
+      "epoch": 2.1668633879781423,
+      "grad_norm": 0.8143426194439874,
+      "learning_rate": 3.7783964893992296e-05,
+      "loss": 12.237,
+      "step": 12391
+    },
+    {
+      "epoch": 2.16703825136612,
+      "grad_norm": 0.8583977855830965,
+      "learning_rate": 3.7769184339712207e-05,
+      "loss": 11.8184,
+      "step": 12392
+    },
+    {
+      "epoch": 2.1672131147540985,
+      "grad_norm": 0.9095396528165531,
+      "learning_rate": 3.7754406003887985e-05,
+      "loss": 12.0394,
+      "step": 12393
+    },
+    {
+      "epoch": 2.1673879781420764,
+      "grad_norm": 0.880100569860126,
+      "learning_rate": 3.77396298870465e-05,
+      "loss": 12.029,
+      "step": 12394
+    },
+    {
+      "epoch": 2.1675628415300547,
+      "grad_norm": 0.850807690095779,
+      "learning_rate": 3.772485598971443e-05,
+      "loss": 11.9829,
+      "step": 12395
+    },
+    {
+      "epoch": 2.1677377049180326,
+      "grad_norm": 0.84649798770177,
+      "learning_rate": 3.7710084312418493e-05,
+      "loss": 11.9708,
+      "step": 12396
+    },
+    {
+      "epoch": 2.167912568306011,
+      "grad_norm": 0.8550179474660684,
+      "learning_rate": 3.769531485568526e-05,
+      "loss": 12.2177,
+      "step": 12397
+    },
+    {
+      "epoch": 2.168087431693989,
+      "grad_norm": 0.9775868348100525,
+      "learning_rate": 3.76805476200413e-05,
+      "loss": 12.1715,
+      "step": 12398
+    },
+    {
+      "epoch": 2.168262295081967,
+      "grad_norm": 0.8673337870250888,
+      "learning_rate": 3.7665782606012966e-05,
+      "loss": 12.3422,
+      "step": 12399
+    },
+    {
+      "epoch": 2.1684371584699456,
+      "grad_norm": 0.827064320591541,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 12.0058,
+      "step": 12400
+    },
+    {
+      "epoch": 2.1686120218579235,
+      "grad_norm": 0.9192679441278421,
+      "learning_rate": 3.7636259244908645e-05,
+      "loss": 12.189,
+      "step": 12401
+    },
+    {
+      "epoch": 2.168786885245902,
+      "grad_norm": 0.8356302780734979,
+      "learning_rate": 3.762150089888515e-05,
+      "loss": 12.2116,
+      "step": 12402
+    },
+    {
+      "epoch": 2.1689617486338797,
+      "grad_norm": 0.8580791923977357,
+      "learning_rate": 3.7606744776582236e-05,
+      "loss": 12.086,
+      "step": 12403
+    },
+    {
+      "epoch": 2.169136612021858,
+      "grad_norm": 0.8745240643577853,
+      "learning_rate": 3.7591990878525986e-05,
+      "loss": 11.9855,
+      "step": 12404
+    },
+    {
+      "epoch": 2.169311475409836,
+      "grad_norm": 0.9137250330401141,
+      "learning_rate": 3.7577239205242334e-05,
+      "loss": 11.9111,
+      "step": 12405
+    },
+    {
+      "epoch": 2.1694863387978143,
+      "grad_norm": 0.8432812918970647,
+      "learning_rate": 3.756248975725717e-05,
+      "loss": 11.9213,
+      "step": 12406
+    },
+    {
+      "epoch": 2.169661202185792,
+      "grad_norm": 0.7605095912223471,
+      "learning_rate": 3.7547742535096294e-05,
+      "loss": 12.0035,
+      "step": 12407
+    },
+    {
+      "epoch": 2.1698360655737705,
+      "grad_norm": 0.825748029131528,
+      "learning_rate": 3.753299753928543e-05,
+      "loss": 12.2747,
+      "step": 12408
+    },
+    {
+      "epoch": 2.170010928961749,
+      "grad_norm": 0.7831836422220148,
+      "learning_rate": 3.751825477035026e-05,
+      "loss": 12.0664,
+      "step": 12409
+    },
+    {
+      "epoch": 2.1701857923497267,
+      "grad_norm": 0.9187950977619983,
+      "learning_rate": 3.750351422881627e-05,
+      "loss": 12.1914,
+      "step": 12410
+    },
+    {
+      "epoch": 2.170360655737705,
+      "grad_norm": 0.902692079904999,
+      "learning_rate": 3.748877591520897e-05,
+      "loss": 12.1362,
+      "step": 12411
+    },
+    {
+      "epoch": 2.170535519125683,
+      "grad_norm": 0.9077371206990341,
+      "learning_rate": 3.747403983005377e-05,
+      "loss": 11.9641,
+      "step": 12412
+    },
+    {
+      "epoch": 2.1707103825136613,
+      "grad_norm": 0.8379103798973823,
+      "learning_rate": 3.7459305973876046e-05,
+      "loss": 12.0941,
+      "step": 12413
+    },
+    {
+      "epoch": 2.1708852459016392,
+      "grad_norm": 0.811404236038627,
+      "learning_rate": 3.744457434720095e-05,
+      "loss": 12.189,
+      "step": 12414
+    },
+    {
+      "epoch": 2.1710601092896176,
+      "grad_norm": 0.8190138472013903,
+      "learning_rate": 3.7429844950553695e-05,
+      "loss": 12.1256,
+      "step": 12415
+    },
+    {
+      "epoch": 2.1712349726775955,
+      "grad_norm": 0.8778342142958937,
+      "learning_rate": 3.7415117784459364e-05,
+      "loss": 11.8568,
+      "step": 12416
+    },
+    {
+      "epoch": 2.171409836065574,
+      "grad_norm": 0.8685921137758255,
+      "learning_rate": 3.7400392849442986e-05,
+      "loss": 11.9061,
+      "step": 12417
+    },
+    {
+      "epoch": 2.171584699453552,
+      "grad_norm": 0.8226551831345957,
+      "learning_rate": 3.738567014602943e-05,
+      "loss": 11.9661,
+      "step": 12418
+    },
+    {
+      "epoch": 2.17175956284153,
+      "grad_norm": 0.8356956521546436,
+      "learning_rate": 3.7370949674743585e-05,
+      "loss": 12.0293,
+      "step": 12419
+    },
+    {
+      "epoch": 2.1719344262295084,
+      "grad_norm": 0.8533504213484686,
+      "learning_rate": 3.73562314361102e-05,
+      "loss": 12.0639,
+      "step": 12420
+    },
+    {
+      "epoch": 2.1721092896174863,
+      "grad_norm": 0.9221985796575057,
+      "learning_rate": 3.7341515430654e-05,
+      "loss": 12.1607,
+      "step": 12421
+    },
+    {
+      "epoch": 2.1722841530054646,
+      "grad_norm": 0.814532783768149,
+      "learning_rate": 3.7326801658899556e-05,
+      "loss": 11.9827,
+      "step": 12422
+    },
+    {
+      "epoch": 2.1724590163934425,
+      "grad_norm": 0.8309759920199952,
+      "learning_rate": 3.731209012137139e-05,
+      "loss": 12.1643,
+      "step": 12423
+    },
+    {
+      "epoch": 2.172633879781421,
+      "grad_norm": 0.8462839793358493,
+      "learning_rate": 3.729738081859398e-05,
+      "loss": 11.8953,
+      "step": 12424
+    },
+    {
+      "epoch": 2.1728087431693988,
+      "grad_norm": 0.8216473104449308,
+      "learning_rate": 3.728267375109167e-05,
+      "loss": 11.6404,
+      "step": 12425
+    },
+    {
+      "epoch": 2.172983606557377,
+      "grad_norm": 0.8687473274619579,
+      "learning_rate": 3.7267968919388774e-05,
+      "loss": 12.1592,
+      "step": 12426
+    },
+    {
+      "epoch": 2.173158469945355,
+      "grad_norm": 0.8545015884889244,
+      "learning_rate": 3.725326632400948e-05,
+      "loss": 12.1119,
+      "step": 12427
+    },
+    {
+      "epoch": 2.1733333333333333,
+      "grad_norm": 0.8067179606818728,
+      "learning_rate": 3.723856596547798e-05,
+      "loss": 12.0852,
+      "step": 12428
+    },
+    {
+      "epoch": 2.1735081967213117,
+      "grad_norm": 0.8147209929733261,
+      "learning_rate": 3.7223867844318224e-05,
+      "loss": 11.8784,
+      "step": 12429
+    },
+    {
+      "epoch": 2.1736830601092896,
+      "grad_norm": 0.8657381178690217,
+      "learning_rate": 3.7209171961054245e-05,
+      "loss": 11.945,
+      "step": 12430
+    },
+    {
+      "epoch": 2.173857923497268,
+      "grad_norm": 0.8380080321638204,
+      "learning_rate": 3.7194478316209926e-05,
+      "loss": 12.2049,
+      "step": 12431
+    },
+    {
+      "epoch": 2.174032786885246,
+      "grad_norm": 0.8178559215107123,
+      "learning_rate": 3.717978691030909e-05,
+      "loss": 11.676,
+      "step": 12432
+    },
+    {
+      "epoch": 2.174207650273224,
+      "grad_norm": 0.8454237889841865,
+      "learning_rate": 3.716509774387543e-05,
+      "loss": 12.2444,
+      "step": 12433
+    },
+    {
+      "epoch": 2.174382513661202,
+      "grad_norm": 0.9014224408116562,
+      "learning_rate": 3.7150410817432625e-05,
+      "loss": 12.1735,
+      "step": 12434
+    },
+    {
+      "epoch": 2.1745573770491804,
+      "grad_norm": 0.8953193647764196,
+      "learning_rate": 3.713572613150423e-05,
+      "loss": 12.0417,
+      "step": 12435
+    },
+    {
+      "epoch": 2.1747322404371583,
+      "grad_norm": 0.8152160731889645,
+      "learning_rate": 3.71210436866138e-05,
+      "loss": 11.9539,
+      "step": 12436
+    },
+    {
+      "epoch": 2.1749071038251366,
+      "grad_norm": 0.9963747221533819,
+      "learning_rate": 3.7106363483284656e-05,
+      "loss": 12.2736,
+      "step": 12437
+    },
+    {
+      "epoch": 2.1750819672131145,
+      "grad_norm": 1.057419346518465,
+      "learning_rate": 3.709168552204016e-05,
+      "loss": 11.9611,
+      "step": 12438
+    },
+    {
+      "epoch": 2.175256830601093,
+      "grad_norm": 0.891115421518192,
+      "learning_rate": 3.707700980340363e-05,
+      "loss": 12.1911,
+      "step": 12439
+    },
+    {
+      "epoch": 2.175431693989071,
+      "grad_norm": 0.9115493677987123,
+      "learning_rate": 3.706233632789813e-05,
+      "loss": 12.1863,
+      "step": 12440
+    },
+    {
+      "epoch": 2.175606557377049,
+      "grad_norm": 0.9288868149389645,
+      "learning_rate": 3.704766509604681e-05,
+      "loss": 11.9408,
+      "step": 12441
+    },
+    {
+      "epoch": 2.1757814207650275,
+      "grad_norm": 0.8168934422836523,
+      "learning_rate": 3.703299610837269e-05,
+      "loss": 11.9637,
+      "step": 12442
+    },
+    {
+      "epoch": 2.1759562841530053,
+      "grad_norm": 0.8123835120353063,
+      "learning_rate": 3.7018329365398674e-05,
+      "loss": 12.0606,
+      "step": 12443
+    },
+    {
+      "epoch": 2.1761311475409837,
+      "grad_norm": 0.9512262875357688,
+      "learning_rate": 3.700366486764762e-05,
+      "loss": 12.0844,
+      "step": 12444
+    },
+    {
+      "epoch": 2.1763060109289616,
+      "grad_norm": 0.8562678755976958,
+      "learning_rate": 3.698900261564233e-05,
+      "loss": 11.9544,
+      "step": 12445
+    },
+    {
+      "epoch": 2.17648087431694,
+      "grad_norm": 0.9111805253998574,
+      "learning_rate": 3.6974342609905475e-05,
+      "loss": 12.1594,
+      "step": 12446
+    },
+    {
+      "epoch": 2.176655737704918,
+      "grad_norm": 0.7965628993176419,
+      "learning_rate": 3.69596848509597e-05,
+      "loss": 12.0568,
+      "step": 12447
+    },
+    {
+      "epoch": 2.176830601092896,
+      "grad_norm": 0.8335459120517217,
+      "learning_rate": 3.694502933932746e-05,
+      "loss": 12.2852,
+      "step": 12448
+    },
+    {
+      "epoch": 2.1770054644808745,
+      "grad_norm": 1.004276514909726,
+      "learning_rate": 3.693037607553127e-05,
+      "loss": 12.1259,
+      "step": 12449
+    },
+    {
+      "epoch": 2.1771803278688524,
+      "grad_norm": 0.822695473109056,
+      "learning_rate": 3.6915725060093475e-05,
+      "loss": 12.2522,
+      "step": 12450
+    },
+    {
+      "epoch": 2.1773551912568307,
+      "grad_norm": 0.856389524037574,
+      "learning_rate": 3.690107629353641e-05,
+      "loss": 11.8215,
+      "step": 12451
+    },
+    {
+      "epoch": 2.1775300546448086,
+      "grad_norm": 0.9380145511609989,
+      "learning_rate": 3.6886429776382226e-05,
+      "loss": 11.9256,
+      "step": 12452
+    },
+    {
+      "epoch": 2.177704918032787,
+      "grad_norm": 0.8617669262618299,
+      "learning_rate": 3.687178550915307e-05,
+      "loss": 12.0946,
+      "step": 12453
+    },
+    {
+      "epoch": 2.177879781420765,
+      "grad_norm": 0.8365541436807541,
+      "learning_rate": 3.685714349237104e-05,
+      "loss": 12.2237,
+      "step": 12454
+    },
+    {
+      "epoch": 2.1780546448087432,
+      "grad_norm": 0.8300961951669528,
+      "learning_rate": 3.684250372655803e-05,
+      "loss": 11.9817,
+      "step": 12455
+    },
+    {
+      "epoch": 2.178229508196721,
+      "grad_norm": 0.83995653697823,
+      "learning_rate": 3.6827866212235984e-05,
+      "loss": 11.8583,
+      "step": 12456
+    },
+    {
+      "epoch": 2.1784043715846995,
+      "grad_norm": 0.798911206334524,
+      "learning_rate": 3.681323094992668e-05,
+      "loss": 12.0759,
+      "step": 12457
+    },
+    {
+      "epoch": 2.178579234972678,
+      "grad_norm": 0.7649498905045049,
+      "learning_rate": 3.6798597940151916e-05,
+      "loss": 12.0378,
+      "step": 12458
+    },
+    {
+      "epoch": 2.1787540983606557,
+      "grad_norm": 0.8792405238058496,
+      "learning_rate": 3.6783967183433254e-05,
+      "loss": 12.1833,
+      "step": 12459
+    },
+    {
+      "epoch": 2.178928961748634,
+      "grad_norm": 0.8299895928987983,
+      "learning_rate": 3.676933868029231e-05,
+      "loss": 12.0436,
+      "step": 12460
+    },
+    {
+      "epoch": 2.179103825136612,
+      "grad_norm": 0.8296358906965573,
+      "learning_rate": 3.675471243125055e-05,
+      "loss": 11.7396,
+      "step": 12461
+    },
+    {
+      "epoch": 2.1792786885245903,
+      "grad_norm": 0.856204635810614,
+      "learning_rate": 3.674008843682941e-05,
+      "loss": 11.9627,
+      "step": 12462
+    },
+    {
+      "epoch": 2.179453551912568,
+      "grad_norm": 0.8996546232955709,
+      "learning_rate": 3.672546669755019e-05,
+      "loss": 12.0308,
+      "step": 12463
+    },
+    {
+      "epoch": 2.1796284153005465,
+      "grad_norm": 0.7937519876347741,
+      "learning_rate": 3.671084721393417e-05,
+      "loss": 12.1246,
+      "step": 12464
+    },
+    {
+      "epoch": 2.1798032786885244,
+      "grad_norm": 0.9090332490580615,
+      "learning_rate": 3.6696229986502494e-05,
+      "loss": 12.1469,
+      "step": 12465
+    },
+    {
+      "epoch": 2.1799781420765028,
+      "grad_norm": 0.9075092627664674,
+      "learning_rate": 3.668161501577629e-05,
+      "loss": 12.0759,
+      "step": 12466
+    },
+    {
+      "epoch": 2.180153005464481,
+      "grad_norm": 1.069186539984476,
+      "learning_rate": 3.666700230227651e-05,
+      "loss": 11.9695,
+      "step": 12467
+    },
+    {
+      "epoch": 2.180327868852459,
+      "grad_norm": 0.878352773659358,
+      "learning_rate": 3.6652391846524094e-05,
+      "loss": 12.3036,
+      "step": 12468
+    },
+    {
+      "epoch": 2.1805027322404373,
+      "grad_norm": 0.9839392529538632,
+      "learning_rate": 3.663778364903993e-05,
+      "loss": 11.9881,
+      "step": 12469
+    },
+    {
+      "epoch": 2.1806775956284152,
+      "grad_norm": 1.0190033848468596,
+      "learning_rate": 3.66231777103447e-05,
+      "loss": 12.0481,
+      "step": 12470
+    },
+    {
+      "epoch": 2.1808524590163936,
+      "grad_norm": 0.7877195466665797,
+      "learning_rate": 3.6608574030959145e-05,
+      "loss": 12.1132,
+      "step": 12471
+    },
+    {
+      "epoch": 2.1810273224043715,
+      "grad_norm": 0.8695467084364371,
+      "learning_rate": 3.659397261140386e-05,
+      "loss": 12.132,
+      "step": 12472
+    },
+    {
+      "epoch": 2.18120218579235,
+      "grad_norm": 0.943322785447394,
+      "learning_rate": 3.6579373452199405e-05,
+      "loss": 12.047,
+      "step": 12473
+    },
+    {
+      "epoch": 2.1813770491803277,
+      "grad_norm": 0.9567722189657806,
+      "learning_rate": 3.6564776553866164e-05,
+      "loss": 12.1123,
+      "step": 12474
+    },
+    {
+      "epoch": 2.181551912568306,
+      "grad_norm": 0.8634500571188226,
+      "learning_rate": 3.655018191692451e-05,
+      "loss": 11.8528,
+      "step": 12475
+    },
+    {
+      "epoch": 2.181726775956284,
+      "grad_norm": 0.8667637441793327,
+      "learning_rate": 3.653558954189474e-05,
+      "loss": 12.4516,
+      "step": 12476
+    },
+    {
+      "epoch": 2.1819016393442623,
+      "grad_norm": 0.9648500374948674,
+      "learning_rate": 3.652099942929709e-05,
+      "loss": 12.055,
+      "step": 12477
+    },
+    {
+      "epoch": 2.1820765027322406,
+      "grad_norm": 0.9341286465958981,
+      "learning_rate": 3.65064115796516e-05,
+      "loss": 12.1538,
+      "step": 12478
+    },
+    {
+      "epoch": 2.1822513661202185,
+      "grad_norm": 0.8362181992199449,
+      "learning_rate": 3.649182599347837e-05,
+      "loss": 11.935,
+      "step": 12479
+    },
+    {
+      "epoch": 2.182426229508197,
+      "grad_norm": 1.0978618373995257,
+      "learning_rate": 3.6477242671297326e-05,
+      "loss": 12.153,
+      "step": 12480
+    },
+    {
+      "epoch": 2.1826010928961748,
+      "grad_norm": 0.994974342772615,
+      "learning_rate": 3.646266161362837e-05,
+      "loss": 12.1594,
+      "step": 12481
+    },
+    {
+      "epoch": 2.182775956284153,
+      "grad_norm": 0.8255315918879981,
+      "learning_rate": 3.6448082820991284e-05,
+      "loss": 12.2063,
+      "step": 12482
+    },
+    {
+      "epoch": 2.182950819672131,
+      "grad_norm": 0.8178004390079172,
+      "learning_rate": 3.6433506293905795e-05,
+      "loss": 11.9415,
+      "step": 12483
+    },
+    {
+      "epoch": 2.1831256830601093,
+      "grad_norm": 0.8862147994043471,
+      "learning_rate": 3.641893203289155e-05,
+      "loss": 12.1393,
+      "step": 12484
+    },
+    {
+      "epoch": 2.1833005464480872,
+      "grad_norm": 0.8646491182713851,
+      "learning_rate": 3.6404360038468114e-05,
+      "loss": 12.2502,
+      "step": 12485
+    },
+    {
+      "epoch": 2.1834754098360656,
+      "grad_norm": 0.787669771738758,
+      "learning_rate": 3.6389790311154905e-05,
+      "loss": 12.1186,
+      "step": 12486
+    },
+    {
+      "epoch": 2.1836502732240435,
+      "grad_norm": 0.8857077161995235,
+      "learning_rate": 3.6375222851471356e-05,
+      "loss": 12.1156,
+      "step": 12487
+    },
+    {
+      "epoch": 2.183825136612022,
+      "grad_norm": 0.7873855093806409,
+      "learning_rate": 3.63606576599368e-05,
+      "loss": 12.006,
+      "step": 12488
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 0.8019029850395522,
+      "learning_rate": 3.634609473707042e-05,
+      "loss": 12.146,
+      "step": 12489
+    },
+    {
+      "epoch": 2.184174863387978,
+      "grad_norm": 0.8512888041552193,
+      "learning_rate": 3.6331534083391385e-05,
+      "loss": 11.8262,
+      "step": 12490
+    },
+    {
+      "epoch": 2.1843497267759564,
+      "grad_norm": 0.8746123678607995,
+      "learning_rate": 3.631697569941878e-05,
+      "loss": 11.6473,
+      "step": 12491
+    },
+    {
+      "epoch": 2.1845245901639343,
+      "grad_norm": 0.809504850994876,
+      "learning_rate": 3.630241958567161e-05,
+      "loss": 11.7252,
+      "step": 12492
+    },
+    {
+      "epoch": 2.1846994535519126,
+      "grad_norm": 0.7881012590629408,
+      "learning_rate": 3.628786574266872e-05,
+      "loss": 12.0153,
+      "step": 12493
+    },
+    {
+      "epoch": 2.1848743169398905,
+      "grad_norm": 0.831034841418906,
+      "learning_rate": 3.6273314170928994e-05,
+      "loss": 12.1105,
+      "step": 12494
+    },
+    {
+      "epoch": 2.185049180327869,
+      "grad_norm": 0.848369512730443,
+      "learning_rate": 3.625876487097115e-05,
+      "loss": 12.0528,
+      "step": 12495
+    },
+    {
+      "epoch": 2.1852240437158468,
+      "grad_norm": 0.8917049754735004,
+      "learning_rate": 3.6244217843313864e-05,
+      "loss": 11.89,
+      "step": 12496
+    },
+    {
+      "epoch": 2.185398907103825,
+      "grad_norm": 0.8273843602427987,
+      "learning_rate": 3.6229673088475725e-05,
+      "loss": 12.0001,
+      "step": 12497
+    },
+    {
+      "epoch": 2.1855737704918035,
+      "grad_norm": 0.847786925091858,
+      "learning_rate": 3.621513060697523e-05,
+      "loss": 12.2991,
+      "step": 12498
+    },
+    {
+      "epoch": 2.1857486338797814,
+      "grad_norm": 0.8003141208113811,
+      "learning_rate": 3.6200590399330816e-05,
+      "loss": 12.101,
+      "step": 12499
+    },
+    {
+      "epoch": 2.1859234972677597,
+      "grad_norm": 0.8744630321676524,
+      "learning_rate": 3.618605246606083e-05,
+      "loss": 11.9863,
+      "step": 12500
+    },
+    {
+      "epoch": 2.1860983606557376,
+      "grad_norm": 0.8077026973426334,
+      "learning_rate": 3.617151680768349e-05,
+      "loss": 11.9732,
+      "step": 12501
+    },
+    {
+      "epoch": 2.186273224043716,
+      "grad_norm": 0.8319446819890597,
+      "learning_rate": 3.6156983424717e-05,
+      "loss": 11.8618,
+      "step": 12502
+    },
+    {
+      "epoch": 2.186448087431694,
+      "grad_norm": 0.9419559378529009,
+      "learning_rate": 3.614245231767949e-05,
+      "loss": 12.0025,
+      "step": 12503
+    },
+    {
+      "epoch": 2.186622950819672,
+      "grad_norm": 0.8827484989936087,
+      "learning_rate": 3.612792348708891e-05,
+      "loss": 11.9083,
+      "step": 12504
+    },
+    {
+      "epoch": 2.18679781420765,
+      "grad_norm": 0.8467872945032586,
+      "learning_rate": 3.611339693346323e-05,
+      "loss": 12.0549,
+      "step": 12505
+    },
+    {
+      "epoch": 2.1869726775956284,
+      "grad_norm": 0.9376351994808783,
+      "learning_rate": 3.6098872657320306e-05,
+      "loss": 12.1416,
+      "step": 12506
+    },
+    {
+      "epoch": 2.1871475409836068,
+      "grad_norm": 0.8660113299864252,
+      "learning_rate": 3.6084350659177946e-05,
+      "loss": 11.9623,
+      "step": 12507
+    },
+    {
+      "epoch": 2.1873224043715846,
+      "grad_norm": 0.830800074221431,
+      "learning_rate": 3.606983093955378e-05,
+      "loss": 11.8188,
+      "step": 12508
+    },
+    {
+      "epoch": 2.187497267759563,
+      "grad_norm": 0.9019023592993132,
+      "learning_rate": 3.6055313498965434e-05,
+      "loss": 12.1677,
+      "step": 12509
+    },
+    {
+      "epoch": 2.187672131147541,
+      "grad_norm": 0.8460492038699463,
+      "learning_rate": 3.604079833793047e-05,
+      "loss": 12.1995,
+      "step": 12510
+    },
+    {
+      "epoch": 2.1878469945355192,
+      "grad_norm": 0.869744042788408,
+      "learning_rate": 3.6026285456966337e-05,
+      "loss": 12.016,
+      "step": 12511
+    },
+    {
+      "epoch": 2.188021857923497,
+      "grad_norm": 0.7895395662122443,
+      "learning_rate": 3.6011774856590354e-05,
+      "loss": 12.0771,
+      "step": 12512
+    },
+    {
+      "epoch": 2.1881967213114755,
+      "grad_norm": 0.8570959436152916,
+      "learning_rate": 3.599726653731983e-05,
+      "loss": 12.2374,
+      "step": 12513
+    },
+    {
+      "epoch": 2.1883715846994534,
+      "grad_norm": 0.8615138969734083,
+      "learning_rate": 3.5982760499671974e-05,
+      "loss": 11.9353,
+      "step": 12514
+    },
+    {
+      "epoch": 2.1885464480874317,
+      "grad_norm": 0.8478330346726265,
+      "learning_rate": 3.596825674416391e-05,
+      "loss": 12.022,
+      "step": 12515
+    },
+    {
+      "epoch": 2.18872131147541,
+      "grad_norm": 0.8138233530197847,
+      "learning_rate": 3.595375527131268e-05,
+      "loss": 12.1741,
+      "step": 12516
+    },
+    {
+      "epoch": 2.188896174863388,
+      "grad_norm": 0.8552394334557343,
+      "learning_rate": 3.5939256081635256e-05,
+      "loss": 12.0964,
+      "step": 12517
+    },
+    {
+      "epoch": 2.1890710382513663,
+      "grad_norm": 0.8160923518579172,
+      "learning_rate": 3.5924759175648524e-05,
+      "loss": 11.9598,
+      "step": 12518
+    },
+    {
+      "epoch": 2.189245901639344,
+      "grad_norm": 0.8969236759412129,
+      "learning_rate": 3.591026455386924e-05,
+      "loss": 12.0359,
+      "step": 12519
+    },
+    {
+      "epoch": 2.1894207650273225,
+      "grad_norm": 0.8268137728525363,
+      "learning_rate": 3.5895772216814135e-05,
+      "loss": 12.1303,
+      "step": 12520
+    },
+    {
+      "epoch": 2.1895956284153004,
+      "grad_norm": 0.8444585505911837,
+      "learning_rate": 3.5881282164999866e-05,
+      "loss": 12.117,
+      "step": 12521
+    },
+    {
+      "epoch": 2.1897704918032788,
+      "grad_norm": 0.7982354207469804,
+      "learning_rate": 3.5866794398942995e-05,
+      "loss": 12.1037,
+      "step": 12522
+    },
+    {
+      "epoch": 2.1899453551912567,
+      "grad_norm": 0.8347602327500715,
+      "learning_rate": 3.585230891915994e-05,
+      "loss": 12.2697,
+      "step": 12523
+    },
+    {
+      "epoch": 2.190120218579235,
+      "grad_norm": 0.8296202021677679,
+      "learning_rate": 3.583782572616713e-05,
+      "loss": 12.2436,
+      "step": 12524
+    },
+    {
+      "epoch": 2.1902950819672133,
+      "grad_norm": 0.8306273104500085,
+      "learning_rate": 3.582334482048086e-05,
+      "loss": 11.9475,
+      "step": 12525
+    },
+    {
+      "epoch": 2.1904699453551912,
+      "grad_norm": 0.8636092963167402,
+      "learning_rate": 3.5808866202617405e-05,
+      "loss": 12.0022,
+      "step": 12526
+    },
+    {
+      "epoch": 2.1906448087431696,
+      "grad_norm": 0.8765953393057019,
+      "learning_rate": 3.579438987309284e-05,
+      "loss": 11.9567,
+      "step": 12527
+    },
+    {
+      "epoch": 2.1908196721311475,
+      "grad_norm": 0.8355264009820338,
+      "learning_rate": 3.577991583242325e-05,
+      "loss": 12.1413,
+      "step": 12528
+    },
+    {
+      "epoch": 2.190994535519126,
+      "grad_norm": 0.8659829008108736,
+      "learning_rate": 3.576544408112462e-05,
+      "loss": 11.9061,
+      "step": 12529
+    },
+    {
+      "epoch": 2.1911693989071037,
+      "grad_norm": 0.8278417455031829,
+      "learning_rate": 3.5750974619712905e-05,
+      "loss": 12.012,
+      "step": 12530
+    },
+    {
+      "epoch": 2.191344262295082,
+      "grad_norm": 0.9206331406768454,
+      "learning_rate": 3.573650744870385e-05,
+      "loss": 12.4095,
+      "step": 12531
+    },
+    {
+      "epoch": 2.19151912568306,
+      "grad_norm": 0.8057175669837838,
+      "learning_rate": 3.5722042568613214e-05,
+      "loss": 12.0287,
+      "step": 12532
+    },
+    {
+      "epoch": 2.1916939890710383,
+      "grad_norm": 0.9778847864521459,
+      "learning_rate": 3.570757997995665e-05,
+      "loss": 12.0612,
+      "step": 12533
+    },
+    {
+      "epoch": 2.191868852459016,
+      "grad_norm": 0.8505687891072983,
+      "learning_rate": 3.569311968324975e-05,
+      "loss": 12.1146,
+      "step": 12534
+    },
+    {
+      "epoch": 2.1920437158469945,
+      "grad_norm": 1.0444547710241738,
+      "learning_rate": 3.5678661679008005e-05,
+      "loss": 11.9114,
+      "step": 12535
+    },
+    {
+      "epoch": 2.192218579234973,
+      "grad_norm": 0.8430242827193227,
+      "learning_rate": 3.566420596774681e-05,
+      "loss": 11.9586,
+      "step": 12536
+    },
+    {
+      "epoch": 2.1923934426229508,
+      "grad_norm": 1.0203196339103886,
+      "learning_rate": 3.564975254998154e-05,
+      "loss": 12.0671,
+      "step": 12537
+    },
+    {
+      "epoch": 2.192568306010929,
+      "grad_norm": 0.8548320502405999,
+      "learning_rate": 3.563530142622738e-05,
+      "loss": 12.2127,
+      "step": 12538
+    },
+    {
+      "epoch": 2.192743169398907,
+      "grad_norm": 0.9301084770424641,
+      "learning_rate": 3.562085259699952e-05,
+      "loss": 11.9316,
+      "step": 12539
+    },
+    {
+      "epoch": 2.1929180327868854,
+      "grad_norm": 0.9485497376272741,
+      "learning_rate": 3.560640606281304e-05,
+      "loss": 12.0694,
+      "step": 12540
+    },
+    {
+      "epoch": 2.1930928961748632,
+      "grad_norm": 0.953451606770814,
+      "learning_rate": 3.559196182418299e-05,
+      "loss": 12.2202,
+      "step": 12541
+    },
+    {
+      "epoch": 2.1932677595628416,
+      "grad_norm": 0.8542066571965617,
+      "learning_rate": 3.55775198816242e-05,
+      "loss": 12.0829,
+      "step": 12542
+    },
+    {
+      "epoch": 2.1934426229508195,
+      "grad_norm": 0.8314250886056742,
+      "learning_rate": 3.556308023565157e-05,
+      "loss": 11.7252,
+      "step": 12543
+    },
+    {
+      "epoch": 2.193617486338798,
+      "grad_norm": 0.8596646349746476,
+      "learning_rate": 3.5548642886779835e-05,
+      "loss": 11.8728,
+      "step": 12544
+    },
+    {
+      "epoch": 2.1937923497267757,
+      "grad_norm": 0.9038469544902817,
+      "learning_rate": 3.5534207835523716e-05,
+      "loss": 12.0962,
+      "step": 12545
+    },
+    {
+      "epoch": 2.193967213114754,
+      "grad_norm": 0.8694681885306448,
+      "learning_rate": 3.5519775082397735e-05,
+      "loss": 11.9286,
+      "step": 12546
+    },
+    {
+      "epoch": 2.1941420765027324,
+      "grad_norm": 0.8785818200865717,
+      "learning_rate": 3.5505344627916434e-05,
+      "loss": 11.9818,
+      "step": 12547
+    },
+    {
+      "epoch": 2.1943169398907103,
+      "grad_norm": 0.8809302755134739,
+      "learning_rate": 3.5490916472594274e-05,
+      "loss": 12.2044,
+      "step": 12548
+    },
+    {
+      "epoch": 2.1944918032786886,
+      "grad_norm": 0.8963624676058694,
+      "learning_rate": 3.547649061694553e-05,
+      "loss": 12.2206,
+      "step": 12549
+    },
+    {
+      "epoch": 2.1946666666666665,
+      "grad_norm": 0.8666209315926819,
+      "learning_rate": 3.5462067061484526e-05,
+      "loss": 12.1845,
+      "step": 12550
+    },
+    {
+      "epoch": 2.194841530054645,
+      "grad_norm": 0.8754036814359618,
+      "learning_rate": 3.544764580672541e-05,
+      "loss": 11.9451,
+      "step": 12551
+    },
+    {
+      "epoch": 2.195016393442623,
+      "grad_norm": 0.8517708765291077,
+      "learning_rate": 3.543322685318231e-05,
+      "loss": 12.0582,
+      "step": 12552
+    },
+    {
+      "epoch": 2.195191256830601,
+      "grad_norm": 0.8634650248407033,
+      "learning_rate": 3.5418810201369237e-05,
+      "loss": 12.1925,
+      "step": 12553
+    },
+    {
+      "epoch": 2.195366120218579,
+      "grad_norm": 0.8856472063706903,
+      "learning_rate": 3.540439585180012e-05,
+      "loss": 11.7964,
+      "step": 12554
+    },
+    {
+      "epoch": 2.1955409836065574,
+      "grad_norm": 0.8630760115681976,
+      "learning_rate": 3.5389983804988826e-05,
+      "loss": 12.1384,
+      "step": 12555
+    },
+    {
+      "epoch": 2.1957158469945357,
+      "grad_norm": 0.8762184329192965,
+      "learning_rate": 3.5375574061449146e-05,
+      "loss": 11.9259,
+      "step": 12556
+    },
+    {
+      "epoch": 2.1958907103825136,
+      "grad_norm": 0.8285273980726865,
+      "learning_rate": 3.536116662169471e-05,
+      "loss": 12.1734,
+      "step": 12557
+    },
+    {
+      "epoch": 2.196065573770492,
+      "grad_norm": 0.8287545474797198,
+      "learning_rate": 3.534676148623917e-05,
+      "loss": 12.2756,
+      "step": 12558
+    },
+    {
+      "epoch": 2.19624043715847,
+      "grad_norm": 0.9141400514183546,
+      "learning_rate": 3.533235865559603e-05,
+      "loss": 12.2799,
+      "step": 12559
+    },
+    {
+      "epoch": 2.196415300546448,
+      "grad_norm": 0.8318994985373482,
+      "learning_rate": 3.531795813027879e-05,
+      "loss": 12.0508,
+      "step": 12560
+    },
+    {
+      "epoch": 2.196590163934426,
+      "grad_norm": 0.8670036564389537,
+      "learning_rate": 3.5303559910800745e-05,
+      "loss": 11.9594,
+      "step": 12561
+    },
+    {
+      "epoch": 2.1967650273224044,
+      "grad_norm": 0.8484342797692843,
+      "learning_rate": 3.528916399767518e-05,
+      "loss": 12.2294,
+      "step": 12562
+    },
+    {
+      "epoch": 2.1969398907103823,
+      "grad_norm": 0.8538833545903761,
+      "learning_rate": 3.5274770391415356e-05,
+      "loss": 12.1243,
+      "step": 12563
+    },
+    {
+      "epoch": 2.1971147540983607,
+      "grad_norm": 0.8262361258833949,
+      "learning_rate": 3.526037909253431e-05,
+      "loss": 12.2507,
+      "step": 12564
+    },
+    {
+      "epoch": 2.197289617486339,
+      "grad_norm": 0.8472831035134398,
+      "learning_rate": 3.5245990101545115e-05,
+      "loss": 12.0042,
+      "step": 12565
+    },
+    {
+      "epoch": 2.197464480874317,
+      "grad_norm": 0.8399991752812425,
+      "learning_rate": 3.5231603418960715e-05,
+      "loss": 12.157,
+      "step": 12566
+    },
+    {
+      "epoch": 2.1976393442622952,
+      "grad_norm": 0.8144603346327378,
+      "learning_rate": 3.5217219045294e-05,
+      "loss": 11.928,
+      "step": 12567
+    },
+    {
+      "epoch": 2.197814207650273,
+      "grad_norm": 0.9011187861453381,
+      "learning_rate": 3.5202836981057717e-05,
+      "loss": 12.1749,
+      "step": 12568
+    },
+    {
+      "epoch": 2.1979890710382515,
+      "grad_norm": 0.8444707159497966,
+      "learning_rate": 3.5188457226764584e-05,
+      "loss": 12.1283,
+      "step": 12569
+    },
+    {
+      "epoch": 2.1981639344262294,
+      "grad_norm": 0.8146550053518657,
+      "learning_rate": 3.5174079782927226e-05,
+      "loss": 11.9886,
+      "step": 12570
+    },
+    {
+      "epoch": 2.1983387978142077,
+      "grad_norm": 0.8098185399890031,
+      "learning_rate": 3.515970465005818e-05,
+      "loss": 11.8757,
+      "step": 12571
+    },
+    {
+      "epoch": 2.1985136612021856,
+      "grad_norm": 0.8112674925629977,
+      "learning_rate": 3.5145331828669905e-05,
+      "loss": 11.9551,
+      "step": 12572
+    },
+    {
+      "epoch": 2.198688524590164,
+      "grad_norm": 0.8487424509188425,
+      "learning_rate": 3.513096131927478e-05,
+      "loss": 11.931,
+      "step": 12573
+    },
+    {
+      "epoch": 2.1988633879781423,
+      "grad_norm": 0.8670344376403731,
+      "learning_rate": 3.511659312238509e-05,
+      "loss": 12.2645,
+      "step": 12574
+    },
+    {
+      "epoch": 2.19903825136612,
+      "grad_norm": 0.8136708018105152,
+      "learning_rate": 3.5102227238513085e-05,
+      "loss": 11.8884,
+      "step": 12575
+    },
+    {
+      "epoch": 2.1992131147540985,
+      "grad_norm": 0.8502468115391458,
+      "learning_rate": 3.508786366817082e-05,
+      "loss": 12.163,
+      "step": 12576
+    },
+    {
+      "epoch": 2.1993879781420764,
+      "grad_norm": 0.8461395820777367,
+      "learning_rate": 3.507350241187036e-05,
+      "loss": 11.7977,
+      "step": 12577
+    },
+    {
+      "epoch": 2.1995628415300548,
+      "grad_norm": 0.9550628392496237,
+      "learning_rate": 3.5059143470123725e-05,
+      "loss": 11.9163,
+      "step": 12578
+    },
+    {
+      "epoch": 2.1997377049180327,
+      "grad_norm": 0.8654177042734593,
+      "learning_rate": 3.504478684344272e-05,
+      "loss": 11.8194,
+      "step": 12579
+    },
+    {
+      "epoch": 2.199912568306011,
+      "grad_norm": 0.8450469302569387,
+      "learning_rate": 3.5030432532339164e-05,
+      "loss": 12.1012,
+      "step": 12580
+    },
+    {
+      "epoch": 2.200087431693989,
+      "grad_norm": 0.9112595664645307,
+      "learning_rate": 3.501608053732478e-05,
+      "loss": 12.2233,
+      "step": 12581
+    },
+    {
+      "epoch": 2.2002622950819672,
+      "grad_norm": 0.8872362337340778,
+      "learning_rate": 3.500173085891124e-05,
+      "loss": 12.0763,
+      "step": 12582
+    },
+    {
+      "epoch": 2.2004371584699456,
+      "grad_norm": 0.8573214400743001,
+      "learning_rate": 3.498738349761001e-05,
+      "loss": 12.1033,
+      "step": 12583
+    },
+    {
+      "epoch": 2.2006120218579235,
+      "grad_norm": 0.9124884126543674,
+      "learning_rate": 3.4973038453932616e-05,
+      "loss": 11.7715,
+      "step": 12584
+    },
+    {
+      "epoch": 2.200786885245902,
+      "grad_norm": 0.8636379326125609,
+      "learning_rate": 3.495869572839041e-05,
+      "loss": 12.3335,
+      "step": 12585
+    },
+    {
+      "epoch": 2.2009617486338797,
+      "grad_norm": 0.8613179858575577,
+      "learning_rate": 3.494435532149472e-05,
+      "loss": 11.9723,
+      "step": 12586
+    },
+    {
+      "epoch": 2.201136612021858,
+      "grad_norm": 0.8715631334873828,
+      "learning_rate": 3.493001723375674e-05,
+      "loss": 12.0616,
+      "step": 12587
+    },
+    {
+      "epoch": 2.201311475409836,
+      "grad_norm": 0.886961463098513,
+      "learning_rate": 3.4915681465687625e-05,
+      "loss": 11.7877,
+      "step": 12588
+    },
+    {
+      "epoch": 2.2014863387978143,
+      "grad_norm": 0.916358894932217,
+      "learning_rate": 3.490134801779842e-05,
+      "loss": 12.1238,
+      "step": 12589
+    },
+    {
+      "epoch": 2.201661202185792,
+      "grad_norm": 0.9544145853429232,
+      "learning_rate": 3.488701689060013e-05,
+      "loss": 11.9049,
+      "step": 12590
+    },
+    {
+      "epoch": 2.2018360655737705,
+      "grad_norm": 0.8568449566160123,
+      "learning_rate": 3.487268808460359e-05,
+      "loss": 12.0469,
+      "step": 12591
+    },
+    {
+      "epoch": 2.2020109289617484,
+      "grad_norm": 0.9153267696726443,
+      "learning_rate": 3.4858361600319624e-05,
+      "loss": 11.9486,
+      "step": 12592
+    },
+    {
+      "epoch": 2.202185792349727,
+      "grad_norm": 0.8759347611601985,
+      "learning_rate": 3.484403743825895e-05,
+      "loss": 11.6942,
+      "step": 12593
+    },
+    {
+      "epoch": 2.202360655737705,
+      "grad_norm": 0.8029832002173255,
+      "learning_rate": 3.4829715598932244e-05,
+      "loss": 11.9025,
+      "step": 12594
+    },
+    {
+      "epoch": 2.202535519125683,
+      "grad_norm": 0.8723340368169346,
+      "learning_rate": 3.4815396082850015e-05,
+      "loss": 12.0187,
+      "step": 12595
+    },
+    {
+      "epoch": 2.2027103825136614,
+      "grad_norm": 0.9041126659783272,
+      "learning_rate": 3.4801078890522744e-05,
+      "loss": 12.1823,
+      "step": 12596
+    },
+    {
+      "epoch": 2.2028852459016393,
+      "grad_norm": 0.9192204900169453,
+      "learning_rate": 3.478676402246087e-05,
+      "loss": 12.1538,
+      "step": 12597
+    },
+    {
+      "epoch": 2.2030601092896176,
+      "grad_norm": 0.7942910485666179,
+      "learning_rate": 3.477245147917464e-05,
+      "loss": 12.206,
+      "step": 12598
+    },
+    {
+      "epoch": 2.2032349726775955,
+      "grad_norm": 0.9652789563160129,
+      "learning_rate": 3.4758141261174295e-05,
+      "loss": 11.9842,
+      "step": 12599
+    },
+    {
+      "epoch": 2.203409836065574,
+      "grad_norm": 0.8234346699771984,
+      "learning_rate": 3.474383336896999e-05,
+      "loss": 12.0567,
+      "step": 12600
+    },
+    {
+      "epoch": 2.2035846994535517,
+      "grad_norm": 0.8495991772012551,
+      "learning_rate": 3.4729527803071816e-05,
+      "loss": 12.171,
+      "step": 12601
+    },
+    {
+      "epoch": 2.20375956284153,
+      "grad_norm": 0.8734879417403144,
+      "learning_rate": 3.471522456398968e-05,
+      "loss": 11.9847,
+      "step": 12602
+    },
+    {
+      "epoch": 2.203934426229508,
+      "grad_norm": 0.9585400536228376,
+      "learning_rate": 3.470092365223351e-05,
+      "loss": 12.0251,
+      "step": 12603
+    },
+    {
+      "epoch": 2.2041092896174863,
+      "grad_norm": 0.7772343200138372,
+      "learning_rate": 3.4686625068313115e-05,
+      "loss": 12.1536,
+      "step": 12604
+    },
+    {
+      "epoch": 2.2042841530054647,
+      "grad_norm": 0.9160549869601518,
+      "learning_rate": 3.467232881273823e-05,
+      "loss": 12.109,
+      "step": 12605
+    },
+    {
+      "epoch": 2.2044590163934425,
+      "grad_norm": 1.0462493637547465,
+      "learning_rate": 3.46580348860185e-05,
+      "loss": 12.0689,
+      "step": 12606
+    },
+    {
+      "epoch": 2.204633879781421,
+      "grad_norm": 0.8252891044665477,
+      "learning_rate": 3.4643743288663474e-05,
+      "loss": 12.0841,
+      "step": 12607
+    },
+    {
+      "epoch": 2.204808743169399,
+      "grad_norm": 0.7851385254679606,
+      "learning_rate": 3.462945402118265e-05,
+      "loss": 12.0478,
+      "step": 12608
+    },
+    {
+      "epoch": 2.204983606557377,
+      "grad_norm": 0.839614235655714,
+      "learning_rate": 3.4615167084085445e-05,
+      "loss": 12.1411,
+      "step": 12609
+    },
+    {
+      "epoch": 2.205158469945355,
+      "grad_norm": 0.9226697442187887,
+      "learning_rate": 3.46008824778811e-05,
+      "loss": 12.1098,
+      "step": 12610
+    },
+    {
+      "epoch": 2.2053333333333334,
+      "grad_norm": 0.8620858159473727,
+      "learning_rate": 3.458660020307888e-05,
+      "loss": 12.1195,
+      "step": 12611
+    },
+    {
+      "epoch": 2.2055081967213113,
+      "grad_norm": 0.8679439078434729,
+      "learning_rate": 3.457232026018797e-05,
+      "loss": 12.0292,
+      "step": 12612
+    },
+    {
+      "epoch": 2.2056830601092896,
+      "grad_norm": 0.9067634757499937,
+      "learning_rate": 3.455804264971738e-05,
+      "loss": 12.1862,
+      "step": 12613
+    },
+    {
+      "epoch": 2.205857923497268,
+      "grad_norm": 0.835840881346912,
+      "learning_rate": 3.4543767372176095e-05,
+      "loss": 12.0657,
+      "step": 12614
+    },
+    {
+      "epoch": 2.206032786885246,
+      "grad_norm": 0.8400567309613594,
+      "learning_rate": 3.452949442807303e-05,
+      "loss": 12.2502,
+      "step": 12615
+    },
+    {
+      "epoch": 2.206207650273224,
+      "grad_norm": 0.9192057640072414,
+      "learning_rate": 3.451522381791703e-05,
+      "loss": 12.1355,
+      "step": 12616
+    },
+    {
+      "epoch": 2.206382513661202,
+      "grad_norm": 0.9642187452545291,
+      "learning_rate": 3.450095554221675e-05,
+      "loss": 11.8551,
+      "step": 12617
+    },
+    {
+      "epoch": 2.2065573770491804,
+      "grad_norm": 0.9065922321563363,
+      "learning_rate": 3.448668960148088e-05,
+      "loss": 12.2634,
+      "step": 12618
+    },
+    {
+      "epoch": 2.2067322404371583,
+      "grad_norm": 0.9090674518015217,
+      "learning_rate": 3.447242599621798e-05,
+      "loss": 11.9888,
+      "step": 12619
+    },
+    {
+      "epoch": 2.2069071038251367,
+      "grad_norm": 0.8666098139933499,
+      "learning_rate": 3.4458164726936574e-05,
+      "loss": 12.0348,
+      "step": 12620
+    },
+    {
+      "epoch": 2.2070819672131146,
+      "grad_norm": 0.9060457279135992,
+      "learning_rate": 3.444390579414498e-05,
+      "loss": 12.0515,
+      "step": 12621
+    },
+    {
+      "epoch": 2.207256830601093,
+      "grad_norm": 0.8003495351646877,
+      "learning_rate": 3.4429649198351556e-05,
+      "loss": 11.9604,
+      "step": 12622
+    },
+    {
+      "epoch": 2.2074316939890712,
+      "grad_norm": 0.9892727488466393,
+      "learning_rate": 3.441539494006451e-05,
+      "loss": 11.9938,
+      "step": 12623
+    },
+    {
+      "epoch": 2.207606557377049,
+      "grad_norm": 0.916962556770131,
+      "learning_rate": 3.4401143019792025e-05,
+      "loss": 12.0683,
+      "step": 12624
+    },
+    {
+      "epoch": 2.2077814207650275,
+      "grad_norm": 0.9085090166937814,
+      "learning_rate": 3.4386893438042135e-05,
+      "loss": 11.9944,
+      "step": 12625
+    },
+    {
+      "epoch": 2.2079562841530054,
+      "grad_norm": 0.8882605840053002,
+      "learning_rate": 3.4372646195322846e-05,
+      "loss": 11.9055,
+      "step": 12626
+    },
+    {
+      "epoch": 2.2081311475409837,
+      "grad_norm": 0.865610019559679,
+      "learning_rate": 3.435840129214206e-05,
+      "loss": 12.018,
+      "step": 12627
+    },
+    {
+      "epoch": 2.2083060109289616,
+      "grad_norm": 0.8946835426246833,
+      "learning_rate": 3.434415872900756e-05,
+      "loss": 12.1801,
+      "step": 12628
+    },
+    {
+      "epoch": 2.20848087431694,
+      "grad_norm": 0.8625780269196983,
+      "learning_rate": 3.4329918506427075e-05,
+      "loss": 12.0002,
+      "step": 12629
+    },
+    {
+      "epoch": 2.208655737704918,
+      "grad_norm": 0.8748467407331302,
+      "learning_rate": 3.4315680624908273e-05,
+      "loss": 12.1544,
+      "step": 12630
+    },
+    {
+      "epoch": 2.208830601092896,
+      "grad_norm": 0.9015640300493818,
+      "learning_rate": 3.430144508495875e-05,
+      "loss": 12.2271,
+      "step": 12631
+    },
+    {
+      "epoch": 2.2090054644808745,
+      "grad_norm": 0.8473141191323307,
+      "learning_rate": 3.428721188708591e-05,
+      "loss": 11.6793,
+      "step": 12632
+    },
+    {
+      "epoch": 2.2091803278688524,
+      "grad_norm": 0.9753165846946821,
+      "learning_rate": 3.427298103179719e-05,
+      "loss": 12.217,
+      "step": 12633
+    },
+    {
+      "epoch": 2.2093551912568308,
+      "grad_norm": 0.8393174735894894,
+      "learning_rate": 3.42587525195999e-05,
+      "loss": 11.9669,
+      "step": 12634
+    },
+    {
+      "epoch": 2.2095300546448087,
+      "grad_norm": 0.8476483007996513,
+      "learning_rate": 3.424452635100132e-05,
+      "loss": 12.0174,
+      "step": 12635
+    },
+    {
+      "epoch": 2.209704918032787,
+      "grad_norm": 0.8560735605503804,
+      "learning_rate": 3.423030252650849e-05,
+      "loss": 12.2452,
+      "step": 12636
+    },
+    {
+      "epoch": 2.209879781420765,
+      "grad_norm": 0.8848993287559681,
+      "learning_rate": 3.4216081046628555e-05,
+      "loss": 12.0795,
+      "step": 12637
+    },
+    {
+      "epoch": 2.2100546448087433,
+      "grad_norm": 0.8315685648114548,
+      "learning_rate": 3.420186191186846e-05,
+      "loss": 12.1274,
+      "step": 12638
+    },
+    {
+      "epoch": 2.210229508196721,
+      "grad_norm": 0.8858185075252867,
+      "learning_rate": 3.418764512273514e-05,
+      "loss": 12.0664,
+      "step": 12639
+    },
+    {
+      "epoch": 2.2104043715846995,
+      "grad_norm": 0.8618759841313671,
+      "learning_rate": 3.4173430679735364e-05,
+      "loss": 12.0287,
+      "step": 12640
+    },
+    {
+      "epoch": 2.210579234972678,
+      "grad_norm": 0.8200414125979227,
+      "learning_rate": 3.415921858337585e-05,
+      "loss": 11.8866,
+      "step": 12641
+    },
+    {
+      "epoch": 2.2107540983606557,
+      "grad_norm": 0.7764136288878128,
+      "learning_rate": 3.4145008834163286e-05,
+      "loss": 12.1231,
+      "step": 12642
+    },
+    {
+      "epoch": 2.210928961748634,
+      "grad_norm": 0.7969246778524792,
+      "learning_rate": 3.41308014326042e-05,
+      "loss": 11.9605,
+      "step": 12643
+    },
+    {
+      "epoch": 2.211103825136612,
+      "grad_norm": 0.8598983505777944,
+      "learning_rate": 3.411659637920509e-05,
+      "loss": 12.0205,
+      "step": 12644
+    },
+    {
+      "epoch": 2.2112786885245903,
+      "grad_norm": 0.9526269725918606,
+      "learning_rate": 3.410239367447234e-05,
+      "loss": 12.0193,
+      "step": 12645
+    },
+    {
+      "epoch": 2.211453551912568,
+      "grad_norm": 0.8115263594503315,
+      "learning_rate": 3.40881933189123e-05,
+      "loss": 11.932,
+      "step": 12646
+    },
+    {
+      "epoch": 2.2116284153005465,
+      "grad_norm": 0.9138567139620521,
+      "learning_rate": 3.4073995313031115e-05,
+      "loss": 12.187,
+      "step": 12647
+    },
+    {
+      "epoch": 2.2118032786885244,
+      "grad_norm": 1.049634786882868,
+      "learning_rate": 3.405979965733498e-05,
+      "loss": 12.0696,
+      "step": 12648
+    },
+    {
+      "epoch": 2.211978142076503,
+      "grad_norm": 0.8247677482362628,
+      "learning_rate": 3.404560635232994e-05,
+      "loss": 12.2231,
+      "step": 12649
+    },
+    {
+      "epoch": 2.2121530054644807,
+      "grad_norm": 0.8113861315802017,
+      "learning_rate": 3.403141539852202e-05,
+      "loss": 12.1485,
+      "step": 12650
+    },
+    {
+      "epoch": 2.212327868852459,
+      "grad_norm": 0.9568305085603739,
+      "learning_rate": 3.401722679641701e-05,
+      "loss": 12.1213,
+      "step": 12651
+    },
+    {
+      "epoch": 2.2125027322404374,
+      "grad_norm": 0.8561922341938222,
+      "learning_rate": 3.4003040546520785e-05,
+      "loss": 11.8301,
+      "step": 12652
+    },
+    {
+      "epoch": 2.2126775956284153,
+      "grad_norm": 0.827558451170652,
+      "learning_rate": 3.398885664933905e-05,
+      "loss": 12.1797,
+      "step": 12653
+    },
+    {
+      "epoch": 2.2128524590163936,
+      "grad_norm": 0.8539346222831721,
+      "learning_rate": 3.397467510537748e-05,
+      "loss": 12.216,
+      "step": 12654
+    },
+    {
+      "epoch": 2.2130273224043715,
+      "grad_norm": 0.9153093205998806,
+      "learning_rate": 3.396049591514159e-05,
+      "loss": 12.1838,
+      "step": 12655
+    },
+    {
+      "epoch": 2.21320218579235,
+      "grad_norm": 0.8535847900792789,
+      "learning_rate": 3.3946319079136834e-05,
+      "loss": 11.9881,
+      "step": 12656
+    },
+    {
+      "epoch": 2.2133770491803277,
+      "grad_norm": 0.8284763105143869,
+      "learning_rate": 3.393214459786868e-05,
+      "loss": 12.1259,
+      "step": 12657
+    },
+    {
+      "epoch": 2.213551912568306,
+      "grad_norm": 0.8435099499545218,
+      "learning_rate": 3.391797247184233e-05,
+      "loss": 12.0468,
+      "step": 12658
+    },
+    {
+      "epoch": 2.213726775956284,
+      "grad_norm": 1.106197284028302,
+      "learning_rate": 3.390380270156306e-05,
+      "loss": 12.2306,
+      "step": 12659
+    },
+    {
+      "epoch": 2.2139016393442623,
+      "grad_norm": 0.8679089497965824,
+      "learning_rate": 3.3889635287536004e-05,
+      "loss": 12.1043,
+      "step": 12660
+    },
+    {
+      "epoch": 2.21407650273224,
+      "grad_norm": 0.8149456462422592,
+      "learning_rate": 3.387547023026621e-05,
+      "loss": 12.0557,
+      "step": 12661
+    },
+    {
+      "epoch": 2.2142513661202186,
+      "grad_norm": 0.9943739360053336,
+      "learning_rate": 3.3861307530258623e-05,
+      "loss": 11.9627,
+      "step": 12662
+    },
+    {
+      "epoch": 2.214426229508197,
+      "grad_norm": 0.8376560299495892,
+      "learning_rate": 3.384714718801817e-05,
+      "loss": 12.0904,
+      "step": 12663
+    },
+    {
+      "epoch": 2.214601092896175,
+      "grad_norm": 0.8914386713327371,
+      "learning_rate": 3.3832989204049616e-05,
+      "loss": 12.1077,
+      "step": 12664
+    },
+    {
+      "epoch": 2.214775956284153,
+      "grad_norm": 0.839788959969399,
+      "learning_rate": 3.3818833578857734e-05,
+      "loss": 12.0805,
+      "step": 12665
+    },
+    {
+      "epoch": 2.214950819672131,
+      "grad_norm": 0.9984015792691974,
+      "learning_rate": 3.3804680312947067e-05,
+      "loss": 12.2178,
+      "step": 12666
+    },
+    {
+      "epoch": 2.2151256830601094,
+      "grad_norm": 0.8652429213884594,
+      "learning_rate": 3.37905294068222e-05,
+      "loss": 12.0091,
+      "step": 12667
+    },
+    {
+      "epoch": 2.2153005464480873,
+      "grad_norm": 0.8246844777343871,
+      "learning_rate": 3.377638086098762e-05,
+      "loss": 11.9404,
+      "step": 12668
+    },
+    {
+      "epoch": 2.2154754098360656,
+      "grad_norm": 0.8636032349576364,
+      "learning_rate": 3.3762234675947714e-05,
+      "loss": 12.0486,
+      "step": 12669
+    },
+    {
+      "epoch": 2.2156502732240435,
+      "grad_norm": 0.8988145746735959,
+      "learning_rate": 3.3748090852206715e-05,
+      "loss": 12.0467,
+      "step": 12670
+    },
+    {
+      "epoch": 2.215825136612022,
+      "grad_norm": 0.8298478010853424,
+      "learning_rate": 3.373394939026887e-05,
+      "loss": 12.0332,
+      "step": 12671
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 0.8281805648436187,
+      "learning_rate": 3.3719810290638345e-05,
+      "loss": 12.1472,
+      "step": 12672
+    },
+    {
+      "epoch": 2.216174863387978,
+      "grad_norm": 0.9226296113697374,
+      "learning_rate": 3.370567355381911e-05,
+      "loss": 11.9777,
+      "step": 12673
+    },
+    {
+      "epoch": 2.2163497267759564,
+      "grad_norm": 0.8881940922235241,
+      "learning_rate": 3.369153918031515e-05,
+      "loss": 12.199,
+      "step": 12674
+    },
+    {
+      "epoch": 2.2165245901639343,
+      "grad_norm": 0.8650863575624429,
+      "learning_rate": 3.367740717063034e-05,
+      "loss": 11.8082,
+      "step": 12675
+    },
+    {
+      "epoch": 2.2166994535519127,
+      "grad_norm": 0.8714112259205369,
+      "learning_rate": 3.366327752526851e-05,
+      "loss": 12.0035,
+      "step": 12676
+    },
+    {
+      "epoch": 2.2168743169398906,
+      "grad_norm": 0.8782455216683014,
+      "learning_rate": 3.364915024473331e-05,
+      "loss": 12.1301,
+      "step": 12677
+    },
+    {
+      "epoch": 2.217049180327869,
+      "grad_norm": 0.8644764690282688,
+      "learning_rate": 3.363502532952837e-05,
+      "loss": 12.0807,
+      "step": 12678
+    },
+    {
+      "epoch": 2.217224043715847,
+      "grad_norm": 0.8887940762614538,
+      "learning_rate": 3.362090278015724e-05,
+      "loss": 12.1311,
+      "step": 12679
+    },
+    {
+      "epoch": 2.217398907103825,
+      "grad_norm": 0.8244457203907517,
+      "learning_rate": 3.360678259712337e-05,
+      "loss": 12.189,
+      "step": 12680
+    },
+    {
+      "epoch": 2.2175737704918035,
+      "grad_norm": 0.8806758402758609,
+      "learning_rate": 3.3592664780930126e-05,
+      "loss": 12.1354,
+      "step": 12681
+    },
+    {
+      "epoch": 2.2177486338797814,
+      "grad_norm": 0.8797769205043738,
+      "learning_rate": 3.35785493320808e-05,
+      "loss": 11.7991,
+      "step": 12682
+    },
+    {
+      "epoch": 2.2179234972677597,
+      "grad_norm": 0.8289511480740758,
+      "learning_rate": 3.356443625107858e-05,
+      "loss": 11.8991,
+      "step": 12683
+    },
+    {
+      "epoch": 2.2180983606557376,
+      "grad_norm": 0.8258783105317786,
+      "learning_rate": 3.355032553842663e-05,
+      "loss": 12.1692,
+      "step": 12684
+    },
+    {
+      "epoch": 2.218273224043716,
+      "grad_norm": 0.8565204267949194,
+      "learning_rate": 3.3536217194627896e-05,
+      "loss": 12.0278,
+      "step": 12685
+    },
+    {
+      "epoch": 2.218448087431694,
+      "grad_norm": 0.8543994330807912,
+      "learning_rate": 3.352211122018536e-05,
+      "loss": 12.1649,
+      "step": 12686
+    },
+    {
+      "epoch": 2.218622950819672,
+      "grad_norm": 0.8694168622266288,
+      "learning_rate": 3.350800761560192e-05,
+      "loss": 11.9045,
+      "step": 12687
+    },
+    {
+      "epoch": 2.21879781420765,
+      "grad_norm": 0.8209161686482742,
+      "learning_rate": 3.349390638138029e-05,
+      "loss": 12.2361,
+      "step": 12688
+    },
+    {
+      "epoch": 2.2189726775956284,
+      "grad_norm": 1.024265797768057,
+      "learning_rate": 3.3479807518023196e-05,
+      "loss": 12.1928,
+      "step": 12689
+    },
+    {
+      "epoch": 2.219147540983607,
+      "grad_norm": 0.9302714788468349,
+      "learning_rate": 3.346571102603324e-05,
+      "loss": 12.1683,
+      "step": 12690
+    },
+    {
+      "epoch": 2.2193224043715847,
+      "grad_norm": 0.841320702608908,
+      "learning_rate": 3.3451616905913e-05,
+      "loss": 12.083,
+      "step": 12691
+    },
+    {
+      "epoch": 2.219497267759563,
+      "grad_norm": 0.9254893529229182,
+      "learning_rate": 3.3437525158164803e-05,
+      "loss": 12.1867,
+      "step": 12692
+    },
+    {
+      "epoch": 2.219672131147541,
+      "grad_norm": 0.8493052210852227,
+      "learning_rate": 3.342343578329108e-05,
+      "loss": 12.1095,
+      "step": 12693
+    },
+    {
+      "epoch": 2.2198469945355193,
+      "grad_norm": 0.8475267176940205,
+      "learning_rate": 3.340934878179408e-05,
+      "loss": 12.3152,
+      "step": 12694
+    },
+    {
+      "epoch": 2.220021857923497,
+      "grad_norm": 0.8636215980831696,
+      "learning_rate": 3.3395264154176e-05,
+      "loss": 12.0937,
+      "step": 12695
+    },
+    {
+      "epoch": 2.2201967213114755,
+      "grad_norm": 0.8507446670895292,
+      "learning_rate": 3.338118190093893e-05,
+      "loss": 12.0164,
+      "step": 12696
+    },
+    {
+      "epoch": 2.2203715846994534,
+      "grad_norm": 0.796320570005185,
+      "learning_rate": 3.3367102022584887e-05,
+      "loss": 11.7681,
+      "step": 12697
+    },
+    {
+      "epoch": 2.2205464480874317,
+      "grad_norm": 0.7898928853823267,
+      "learning_rate": 3.335302451961581e-05,
+      "loss": 12.2833,
+      "step": 12698
+    },
+    {
+      "epoch": 2.2207213114754096,
+      "grad_norm": 0.839078679843137,
+      "learning_rate": 3.333894939253357e-05,
+      "loss": 12.2375,
+      "step": 12699
+    },
+    {
+      "epoch": 2.220896174863388,
+      "grad_norm": 1.045408532862678,
+      "learning_rate": 3.3324876641839864e-05,
+      "loss": 12.104,
+      "step": 12700
+    },
+    {
+      "epoch": 2.2210710382513663,
+      "grad_norm": 0.8470755266494004,
+      "learning_rate": 3.3310806268036396e-05,
+      "loss": 11.7254,
+      "step": 12701
+    },
+    {
+      "epoch": 2.221245901639344,
+      "grad_norm": 0.8220549402238763,
+      "learning_rate": 3.3296738271624784e-05,
+      "loss": 11.9411,
+      "step": 12702
+    },
+    {
+      "epoch": 2.2214207650273226,
+      "grad_norm": 0.8948814917616124,
+      "learning_rate": 3.328267265310654e-05,
+      "loss": 12.0515,
+      "step": 12703
+    },
+    {
+      "epoch": 2.2215956284153004,
+      "grad_norm": 0.875187112920895,
+      "learning_rate": 3.326860941298303e-05,
+      "loss": 11.7374,
+      "step": 12704
+    },
+    {
+      "epoch": 2.221770491803279,
+      "grad_norm": 0.9041386447009567,
+      "learning_rate": 3.325454855175564e-05,
+      "loss": 12.1188,
+      "step": 12705
+    },
+    {
+      "epoch": 2.2219453551912567,
+      "grad_norm": 0.9852345829882537,
+      "learning_rate": 3.324049006992563e-05,
+      "loss": 12.0715,
+      "step": 12706
+    },
+    {
+      "epoch": 2.222120218579235,
+      "grad_norm": 0.8806507973194807,
+      "learning_rate": 3.322643396799413e-05,
+      "loss": 12.0949,
+      "step": 12707
+    },
+    {
+      "epoch": 2.222295081967213,
+      "grad_norm": 0.8614204472004723,
+      "learning_rate": 3.3212380246462226e-05,
+      "loss": 12.1123,
+      "step": 12708
+    },
+    {
+      "epoch": 2.2224699453551913,
+      "grad_norm": 0.9294854402376905,
+      "learning_rate": 3.3198328905830946e-05,
+      "loss": 12.0988,
+      "step": 12709
+    },
+    {
+      "epoch": 2.222644808743169,
+      "grad_norm": 0.8282161511149496,
+      "learning_rate": 3.318427994660122e-05,
+      "loss": 11.976,
+      "step": 12710
+    },
+    {
+      "epoch": 2.2228196721311475,
+      "grad_norm": 0.8436644562912993,
+      "learning_rate": 3.317023336927382e-05,
+      "loss": 12.1651,
+      "step": 12711
+    },
+    {
+      "epoch": 2.222994535519126,
+      "grad_norm": 0.9059204850876775,
+      "learning_rate": 3.3156189174349504e-05,
+      "loss": 11.9298,
+      "step": 12712
+    },
+    {
+      "epoch": 2.2231693989071037,
+      "grad_norm": 0.9145151067806727,
+      "learning_rate": 3.314214736232896e-05,
+      "loss": 12.027,
+      "step": 12713
+    },
+    {
+      "epoch": 2.223344262295082,
+      "grad_norm": 0.8281343042455268,
+      "learning_rate": 3.312810793371274e-05,
+      "loss": 12.1351,
+      "step": 12714
+    },
+    {
+      "epoch": 2.22351912568306,
+      "grad_norm": 0.8369176396797885,
+      "learning_rate": 3.311407088900135e-05,
+      "loss": 12.1732,
+      "step": 12715
+    },
+    {
+      "epoch": 2.2236939890710383,
+      "grad_norm": 0.8002853345477196,
+      "learning_rate": 3.310003622869517e-05,
+      "loss": 11.9655,
+      "step": 12716
+    },
+    {
+      "epoch": 2.223868852459016,
+      "grad_norm": 0.8039147674633605,
+      "learning_rate": 3.308600395329454e-05,
+      "loss": 11.8555,
+      "step": 12717
+    },
+    {
+      "epoch": 2.2240437158469946,
+      "grad_norm": 0.8359085714755426,
+      "learning_rate": 3.307197406329973e-05,
+      "loss": 11.8824,
+      "step": 12718
+    },
+    {
+      "epoch": 2.2242185792349725,
+      "grad_norm": 0.8315413063740129,
+      "learning_rate": 3.3057946559210806e-05,
+      "loss": 12.0884,
+      "step": 12719
+    },
+    {
+      "epoch": 2.224393442622951,
+      "grad_norm": 0.817216637218115,
+      "learning_rate": 3.304392144152787e-05,
+      "loss": 12.1332,
+      "step": 12720
+    },
+    {
+      "epoch": 2.224568306010929,
+      "grad_norm": 0.912892091707893,
+      "learning_rate": 3.302989871075095e-05,
+      "loss": 11.8102,
+      "step": 12721
+    },
+    {
+      "epoch": 2.224743169398907,
+      "grad_norm": 0.8777584832382274,
+      "learning_rate": 3.301587836737985e-05,
+      "loss": 12.1467,
+      "step": 12722
+    },
+    {
+      "epoch": 2.2249180327868854,
+      "grad_norm": 0.9325360727507142,
+      "learning_rate": 3.300186041191443e-05,
+      "loss": 11.7525,
+      "step": 12723
+    },
+    {
+      "epoch": 2.2250928961748633,
+      "grad_norm": 0.835885744167706,
+      "learning_rate": 3.298784484485441e-05,
+      "loss": 12.2067,
+      "step": 12724
+    },
+    {
+      "epoch": 2.2252677595628416,
+      "grad_norm": 0.928203761842843,
+      "learning_rate": 3.2973831666699454e-05,
+      "loss": 12.1222,
+      "step": 12725
+    },
+    {
+      "epoch": 2.2254426229508195,
+      "grad_norm": 0.8551650151133864,
+      "learning_rate": 3.295982087794906e-05,
+      "loss": 11.9438,
+      "step": 12726
+    },
+    {
+      "epoch": 2.225617486338798,
+      "grad_norm": 0.8671980778800428,
+      "learning_rate": 3.294581247910271e-05,
+      "loss": 12.0704,
+      "step": 12727
+    },
+    {
+      "epoch": 2.2257923497267758,
+      "grad_norm": 0.8867450983894294,
+      "learning_rate": 3.293180647065982e-05,
+      "loss": 11.9012,
+      "step": 12728
+    },
+    {
+      "epoch": 2.225967213114754,
+      "grad_norm": 0.7498820590801535,
+      "learning_rate": 3.291780285311968e-05,
+      "loss": 11.8856,
+      "step": 12729
+    },
+    {
+      "epoch": 2.2261420765027324,
+      "grad_norm": 0.8145087082351099,
+      "learning_rate": 3.2903801626981476e-05,
+      "loss": 11.9388,
+      "step": 12730
+    },
+    {
+      "epoch": 2.2263169398907103,
+      "grad_norm": 0.8402063880071137,
+      "learning_rate": 3.288980279274435e-05,
+      "loss": 12.027,
+      "step": 12731
+    },
+    {
+      "epoch": 2.2264918032786887,
+      "grad_norm": 0.8245820890588241,
+      "learning_rate": 3.2875806350907337e-05,
+      "loss": 11.8769,
+      "step": 12732
+    },
+    {
+      "epoch": 2.2266666666666666,
+      "grad_norm": 0.7808052454622144,
+      "learning_rate": 3.28618123019694e-05,
+      "loss": 11.9466,
+      "step": 12733
+    },
+    {
+      "epoch": 2.226841530054645,
+      "grad_norm": 0.8724698755541661,
+      "learning_rate": 3.284782064642942e-05,
+      "loss": 12.1257,
+      "step": 12734
+    },
+    {
+      "epoch": 2.227016393442623,
+      "grad_norm": 0.8676436969537922,
+      "learning_rate": 3.283383138478617e-05,
+      "loss": 12.0657,
+      "step": 12735
+    },
+    {
+      "epoch": 2.227191256830601,
+      "grad_norm": 0.8530832527742693,
+      "learning_rate": 3.281984451753839e-05,
+      "loss": 12.0137,
+      "step": 12736
+    },
+    {
+      "epoch": 2.227366120218579,
+      "grad_norm": 0.8098608695441138,
+      "learning_rate": 3.280586004518462e-05,
+      "loss": 12.2706,
+      "step": 12737
+    },
+    {
+      "epoch": 2.2275409836065574,
+      "grad_norm": 0.7972193234517623,
+      "learning_rate": 3.279187796822344e-05,
+      "loss": 12.2152,
+      "step": 12738
+    },
+    {
+      "epoch": 2.2277158469945357,
+      "grad_norm": 0.8441071260769053,
+      "learning_rate": 3.277789828715329e-05,
+      "loss": 11.9528,
+      "step": 12739
+    },
+    {
+      "epoch": 2.2278907103825136,
+      "grad_norm": 0.8144398853401547,
+      "learning_rate": 3.2763921002472554e-05,
+      "loss": 12.0856,
+      "step": 12740
+    },
+    {
+      "epoch": 2.228065573770492,
+      "grad_norm": 0.845797765850054,
+      "learning_rate": 3.274994611467944e-05,
+      "loss": 12.1781,
+      "step": 12741
+    },
+    {
+      "epoch": 2.22824043715847,
+      "grad_norm": 0.8928559603231505,
+      "learning_rate": 3.273597362427218e-05,
+      "loss": 12.137,
+      "step": 12742
+    },
+    {
+      "epoch": 2.228415300546448,
+      "grad_norm": 0.8459270804167159,
+      "learning_rate": 3.272200353174887e-05,
+      "loss": 11.7007,
+      "step": 12743
+    },
+    {
+      "epoch": 2.228590163934426,
+      "grad_norm": 0.9057017272075429,
+      "learning_rate": 3.2708035837607566e-05,
+      "loss": 12.2103,
+      "step": 12744
+    },
+    {
+      "epoch": 2.2287650273224044,
+      "grad_norm": 0.854125724260545,
+      "learning_rate": 3.269407054234611e-05,
+      "loss": 11.7607,
+      "step": 12745
+    },
+    {
+      "epoch": 2.2289398907103823,
+      "grad_norm": 0.8829854287254592,
+      "learning_rate": 3.268010764646242e-05,
+      "loss": 12.2322,
+      "step": 12746
+    },
+    {
+      "epoch": 2.2291147540983607,
+      "grad_norm": 0.8406804532854971,
+      "learning_rate": 3.266614715045423e-05,
+      "loss": 12.1781,
+      "step": 12747
+    },
+    {
+      "epoch": 2.229289617486339,
+      "grad_norm": 0.8547401564741897,
+      "learning_rate": 3.265218905481926e-05,
+      "loss": 12.1927,
+      "step": 12748
+    },
+    {
+      "epoch": 2.229464480874317,
+      "grad_norm": 0.7750696991348198,
+      "learning_rate": 3.263823336005503e-05,
+      "loss": 12.0851,
+      "step": 12749
+    },
+    {
+      "epoch": 2.2296393442622953,
+      "grad_norm": 0.9897502467777952,
+      "learning_rate": 3.2624280066659064e-05,
+      "loss": 12.1975,
+      "step": 12750
+    },
+    {
+      "epoch": 2.229814207650273,
+      "grad_norm": 0.8705033712124803,
+      "learning_rate": 3.2610329175128806e-05,
+      "loss": 11.7119,
+      "step": 12751
+    },
+    {
+      "epoch": 2.2299890710382515,
+      "grad_norm": 0.846339132032982,
+      "learning_rate": 3.259638068596158e-05,
+      "loss": 11.6847,
+      "step": 12752
+    },
+    {
+      "epoch": 2.2301639344262294,
+      "grad_norm": 0.9139586478673416,
+      "learning_rate": 3.258243459965462e-05,
+      "loss": 11.7687,
+      "step": 12753
+    },
+    {
+      "epoch": 2.2303387978142077,
+      "grad_norm": 0.8593322056926659,
+      "learning_rate": 3.25684909167051e-05,
+      "loss": 11.869,
+      "step": 12754
+    },
+    {
+      "epoch": 2.2305136612021856,
+      "grad_norm": 0.8424140427937049,
+      "learning_rate": 3.255454963761013e-05,
+      "loss": 12.1341,
+      "step": 12755
+    },
+    {
+      "epoch": 2.230688524590164,
+      "grad_norm": 0.8637253916357438,
+      "learning_rate": 3.254061076286663e-05,
+      "loss": 11.9875,
+      "step": 12756
+    },
+    {
+      "epoch": 2.230863387978142,
+      "grad_norm": 0.9083409537776906,
+      "learning_rate": 3.2526674292971536e-05,
+      "loss": 12.3032,
+      "step": 12757
+    },
+    {
+      "epoch": 2.23103825136612,
+      "grad_norm": 0.8368292934232069,
+      "learning_rate": 3.2512740228421666e-05,
+      "loss": 12.2892,
+      "step": 12758
+    },
+    {
+      "epoch": 2.2312131147540986,
+      "grad_norm": 0.9010359590285908,
+      "learning_rate": 3.24988085697138e-05,
+      "loss": 12.2621,
+      "step": 12759
+    },
+    {
+      "epoch": 2.2313879781420765,
+      "grad_norm": 0.849962098387883,
+      "learning_rate": 3.24848793173445e-05,
+      "loss": 11.8561,
+      "step": 12760
+    },
+    {
+      "epoch": 2.231562841530055,
+      "grad_norm": 0.8124580875434869,
+      "learning_rate": 3.2470952471810365e-05,
+      "loss": 11.7889,
+      "step": 12761
+    },
+    {
+      "epoch": 2.2317377049180327,
+      "grad_norm": 0.8507622549864295,
+      "learning_rate": 3.245702803360788e-05,
+      "loss": 11.809,
+      "step": 12762
+    },
+    {
+      "epoch": 2.231912568306011,
+      "grad_norm": 0.8090941914870413,
+      "learning_rate": 3.244310600323347e-05,
+      "loss": 12.0151,
+      "step": 12763
+    },
+    {
+      "epoch": 2.232087431693989,
+      "grad_norm": 0.8887956698491197,
+      "learning_rate": 3.2429186381183366e-05,
+      "loss": 12.055,
+      "step": 12764
+    },
+    {
+      "epoch": 2.2322622950819673,
+      "grad_norm": 0.9316949828324699,
+      "learning_rate": 3.241526916795381e-05,
+      "loss": 12.0087,
+      "step": 12765
+    },
+    {
+      "epoch": 2.232437158469945,
+      "grad_norm": 0.919945841016226,
+      "learning_rate": 3.240135436404098e-05,
+      "loss": 12.1043,
+      "step": 12766
+    },
+    {
+      "epoch": 2.2326120218579235,
+      "grad_norm": 0.8273128541268167,
+      "learning_rate": 3.2387441969940856e-05,
+      "loss": 12.0026,
+      "step": 12767
+    },
+    {
+      "epoch": 2.2327868852459014,
+      "grad_norm": 0.9539754266805427,
+      "learning_rate": 3.237353198614943e-05,
+      "loss": 11.9986,
+      "step": 12768
+    },
+    {
+      "epoch": 2.2329617486338798,
+      "grad_norm": 0.8301587270725462,
+      "learning_rate": 3.235962441316258e-05,
+      "loss": 11.9298,
+      "step": 12769
+    },
+    {
+      "epoch": 2.233136612021858,
+      "grad_norm": 0.8527588951919716,
+      "learning_rate": 3.234571925147608e-05,
+      "loss": 11.8477,
+      "step": 12770
+    },
+    {
+      "epoch": 2.233311475409836,
+      "grad_norm": 0.8716697101709479,
+      "learning_rate": 3.2331816501585655e-05,
+      "loss": 12.0712,
+      "step": 12771
+    },
+    {
+      "epoch": 2.2334863387978143,
+      "grad_norm": 0.8897256984246057,
+      "learning_rate": 3.2317916163986895e-05,
+      "loss": 12.0374,
+      "step": 12772
+    },
+    {
+      "epoch": 2.2336612021857922,
+      "grad_norm": 0.872589169475148,
+      "learning_rate": 3.230401823917536e-05,
+      "loss": 12.186,
+      "step": 12773
+    },
+    {
+      "epoch": 2.2338360655737706,
+      "grad_norm": 0.8538221157865223,
+      "learning_rate": 3.22901227276465e-05,
+      "loss": 12.156,
+      "step": 12774
+    },
+    {
+      "epoch": 2.2340109289617485,
+      "grad_norm": 0.8608640677773264,
+      "learning_rate": 3.2276229629895625e-05,
+      "loss": 12.1919,
+      "step": 12775
+    },
+    {
+      "epoch": 2.234185792349727,
+      "grad_norm": 0.8665560227212441,
+      "learning_rate": 3.226233894641805e-05,
+      "loss": 11.9841,
+      "step": 12776
+    },
+    {
+      "epoch": 2.2343606557377047,
+      "grad_norm": 0.8840432615495836,
+      "learning_rate": 3.224845067770893e-05,
+      "loss": 11.9134,
+      "step": 12777
+    },
+    {
+      "epoch": 2.234535519125683,
+      "grad_norm": 0.8289391412306008,
+      "learning_rate": 3.223456482426343e-05,
+      "loss": 12.1661,
+      "step": 12778
+    },
+    {
+      "epoch": 2.2347103825136614,
+      "grad_norm": 0.8262073932172109,
+      "learning_rate": 3.222068138657648e-05,
+      "loss": 11.9658,
+      "step": 12779
+    },
+    {
+      "epoch": 2.2348852459016393,
+      "grad_norm": 0.8430177033324028,
+      "learning_rate": 3.2206800365143054e-05,
+      "loss": 11.8961,
+      "step": 12780
+    },
+    {
+      "epoch": 2.2350601092896176,
+      "grad_norm": 0.8257810190618781,
+      "learning_rate": 3.219292176045801e-05,
+      "loss": 11.973,
+      "step": 12781
+    },
+    {
+      "epoch": 2.2352349726775955,
+      "grad_norm": 0.8921487336255416,
+      "learning_rate": 3.2179045573016055e-05,
+      "loss": 12.1408,
+      "step": 12782
+    },
+    {
+      "epoch": 2.235409836065574,
+      "grad_norm": 0.8313420670567149,
+      "learning_rate": 3.216517180331188e-05,
+      "loss": 12.1451,
+      "step": 12783
+    },
+    {
+      "epoch": 2.2355846994535518,
+      "grad_norm": 0.8463681615915065,
+      "learning_rate": 3.2151300451840075e-05,
+      "loss": 12.0225,
+      "step": 12784
+    },
+    {
+      "epoch": 2.23575956284153,
+      "grad_norm": 0.8333711034954164,
+      "learning_rate": 3.213743151909514e-05,
+      "loss": 11.968,
+      "step": 12785
+    },
+    {
+      "epoch": 2.235934426229508,
+      "grad_norm": 0.8626479702325521,
+      "learning_rate": 3.2123565005571474e-05,
+      "loss": 12.2773,
+      "step": 12786
+    },
+    {
+      "epoch": 2.2361092896174863,
+      "grad_norm": 1.1069551166889458,
+      "learning_rate": 3.210970091176342e-05,
+      "loss": 12.1209,
+      "step": 12787
+    },
+    {
+      "epoch": 2.2362841530054647,
+      "grad_norm": 0.934354626630896,
+      "learning_rate": 3.2095839238165206e-05,
+      "loss": 11.807,
+      "step": 12788
+    },
+    {
+      "epoch": 2.2364590163934426,
+      "grad_norm": 0.891602052085218,
+      "learning_rate": 3.208197998527101e-05,
+      "loss": 12.149,
+      "step": 12789
+    },
+    {
+      "epoch": 2.236633879781421,
+      "grad_norm": 0.8794661388070764,
+      "learning_rate": 3.206812315357485e-05,
+      "loss": 12.1385,
+      "step": 12790
+    },
+    {
+      "epoch": 2.236808743169399,
+      "grad_norm": 0.8140479042069774,
+      "learning_rate": 3.205426874357073e-05,
+      "loss": 11.9317,
+      "step": 12791
+    },
+    {
+      "epoch": 2.236983606557377,
+      "grad_norm": 0.8700137964560113,
+      "learning_rate": 3.2040416755752546e-05,
+      "loss": 12.051,
+      "step": 12792
+    },
+    {
+      "epoch": 2.237158469945355,
+      "grad_norm": 0.8878484979275546,
+      "learning_rate": 3.202656719061412e-05,
+      "loss": 12.1609,
+      "step": 12793
+    },
+    {
+      "epoch": 2.2373333333333334,
+      "grad_norm": 0.8874284109310924,
+      "learning_rate": 3.2012720048649135e-05,
+      "loss": 12.0,
+      "step": 12794
+    },
+    {
+      "epoch": 2.2375081967213113,
+      "grad_norm": 0.8682538125379723,
+      "learning_rate": 3.199887533035124e-05,
+      "loss": 12.0202,
+      "step": 12795
+    },
+    {
+      "epoch": 2.2376830601092896,
+      "grad_norm": 0.8128668306313835,
+      "learning_rate": 3.198503303621403e-05,
+      "loss": 12.1267,
+      "step": 12796
+    },
+    {
+      "epoch": 2.237857923497268,
+      "grad_norm": 0.8429931775468193,
+      "learning_rate": 3.1971193166730884e-05,
+      "loss": 12.0588,
+      "step": 12797
+    },
+    {
+      "epoch": 2.238032786885246,
+      "grad_norm": 0.9223649865940717,
+      "learning_rate": 3.1957355722395224e-05,
+      "loss": 12.1355,
+      "step": 12798
+    },
+    {
+      "epoch": 2.238207650273224,
+      "grad_norm": 0.8368904675147281,
+      "learning_rate": 3.194352070370033e-05,
+      "loss": 12.1387,
+      "step": 12799
+    },
+    {
+      "epoch": 2.238382513661202,
+      "grad_norm": 0.83400873689065,
+      "learning_rate": 3.192968811113943e-05,
+      "loss": 12.2314,
+      "step": 12800
+    },
+    {
+      "epoch": 2.2385573770491805,
+      "grad_norm": 0.8571813831163696,
+      "learning_rate": 3.19158579452056e-05,
+      "loss": 12.1291,
+      "step": 12801
+    },
+    {
+      "epoch": 2.2387322404371583,
+      "grad_norm": 0.8473185165275993,
+      "learning_rate": 3.190203020639187e-05,
+      "loss": 11.7871,
+      "step": 12802
+    },
+    {
+      "epoch": 2.2389071038251367,
+      "grad_norm": 0.8697740221681892,
+      "learning_rate": 3.188820489519122e-05,
+      "loss": 12.1807,
+      "step": 12803
+    },
+    {
+      "epoch": 2.2390819672131146,
+      "grad_norm": 0.8465802250499794,
+      "learning_rate": 3.187438201209648e-05,
+      "loss": 12.2525,
+      "step": 12804
+    },
+    {
+      "epoch": 2.239256830601093,
+      "grad_norm": 0.8410061202872339,
+      "learning_rate": 3.186056155760042e-05,
+      "loss": 11.5628,
+      "step": 12805
+    },
+    {
+      "epoch": 2.2394316939890713,
+      "grad_norm": 0.901737718467728,
+      "learning_rate": 3.184674353219572e-05,
+      "loss": 11.5938,
+      "step": 12806
+    },
+    {
+      "epoch": 2.239606557377049,
+      "grad_norm": 0.7941108742132292,
+      "learning_rate": 3.1832927936375e-05,
+      "loss": 12.0941,
+      "step": 12807
+    },
+    {
+      "epoch": 2.2397814207650275,
+      "grad_norm": 0.807460642153718,
+      "learning_rate": 3.181911477063078e-05,
+      "loss": 12.1224,
+      "step": 12808
+    },
+    {
+      "epoch": 2.2399562841530054,
+      "grad_norm": 0.8202970294295778,
+      "learning_rate": 3.180530403545543e-05,
+      "loss": 11.8373,
+      "step": 12809
+    },
+    {
+      "epoch": 2.2401311475409837,
+      "grad_norm": 0.8547026877716559,
+      "learning_rate": 3.1791495731341306e-05,
+      "loss": 11.9621,
+      "step": 12810
+    },
+    {
+      "epoch": 2.2403060109289616,
+      "grad_norm": 0.8002528810947858,
+      "learning_rate": 3.1777689858780666e-05,
+      "loss": 12.1093,
+      "step": 12811
+    },
+    {
+      "epoch": 2.24048087431694,
+      "grad_norm": 0.8706451144647533,
+      "learning_rate": 3.176388641826571e-05,
+      "loss": 12.1505,
+      "step": 12812
+    },
+    {
+      "epoch": 2.240655737704918,
+      "grad_norm": 0.7907915284875133,
+      "learning_rate": 3.1750085410288447e-05,
+      "loss": 12.1731,
+      "step": 12813
+    },
+    {
+      "epoch": 2.2408306010928962,
+      "grad_norm": 0.884517530619929,
+      "learning_rate": 3.17362868353409e-05,
+      "loss": 12.1598,
+      "step": 12814
+    },
+    {
+      "epoch": 2.241005464480874,
+      "grad_norm": 0.8705549267720422,
+      "learning_rate": 3.1722490693915e-05,
+      "loss": 12.2191,
+      "step": 12815
+    },
+    {
+      "epoch": 2.2411803278688525,
+      "grad_norm": 0.8838481826348631,
+      "learning_rate": 3.17086969865025e-05,
+      "loss": 12.2775,
+      "step": 12816
+    },
+    {
+      "epoch": 2.241355191256831,
+      "grad_norm": 0.8939009084792872,
+      "learning_rate": 3.169490571359516e-05,
+      "loss": 12.011,
+      "step": 12817
+    },
+    {
+      "epoch": 2.2415300546448087,
+      "grad_norm": 0.921027319116303,
+      "learning_rate": 3.1681116875684624e-05,
+      "loss": 12.0553,
+      "step": 12818
+    },
+    {
+      "epoch": 2.241704918032787,
+      "grad_norm": 0.8348610517254273,
+      "learning_rate": 3.166733047326248e-05,
+      "loss": 12.0668,
+      "step": 12819
+    },
+    {
+      "epoch": 2.241879781420765,
+      "grad_norm": 0.8660518644768737,
+      "learning_rate": 3.165354650682014e-05,
+      "loss": 11.9452,
+      "step": 12820
+    },
+    {
+      "epoch": 2.2420546448087433,
+      "grad_norm": 0.8729143934978726,
+      "learning_rate": 3.163976497684901e-05,
+      "loss": 12.1059,
+      "step": 12821
+    },
+    {
+      "epoch": 2.242229508196721,
+      "grad_norm": 0.8306139436267218,
+      "learning_rate": 3.1625985883840395e-05,
+      "loss": 11.9784,
+      "step": 12822
+    },
+    {
+      "epoch": 2.2424043715846995,
+      "grad_norm": 0.8229756059047266,
+      "learning_rate": 3.161220922828548e-05,
+      "loss": 12.1332,
+      "step": 12823
+    },
+    {
+      "epoch": 2.2425792349726774,
+      "grad_norm": 0.8918780137940043,
+      "learning_rate": 3.159843501067542e-05,
+      "loss": 12.0965,
+      "step": 12824
+    },
+    {
+      "epoch": 2.2427540983606558,
+      "grad_norm": 0.9508666455874856,
+      "learning_rate": 3.1584663231501235e-05,
+      "loss": 12.1665,
+      "step": 12825
+    },
+    {
+      "epoch": 2.2429289617486337,
+      "grad_norm": 0.8343329225503064,
+      "learning_rate": 3.157089389125386e-05,
+      "loss": 12.0628,
+      "step": 12826
+    },
+    {
+      "epoch": 2.243103825136612,
+      "grad_norm": 0.8692199082067584,
+      "learning_rate": 3.15571269904242e-05,
+      "loss": 12.1983,
+      "step": 12827
+    },
+    {
+      "epoch": 2.2432786885245903,
+      "grad_norm": 0.9208266654937008,
+      "learning_rate": 3.1543362529502965e-05,
+      "loss": 11.9141,
+      "step": 12828
+    },
+    {
+      "epoch": 2.2434535519125682,
+      "grad_norm": 0.90503757225166,
+      "learning_rate": 3.1529600508980875e-05,
+      "loss": 12.1115,
+      "step": 12829
+    },
+    {
+      "epoch": 2.2436284153005466,
+      "grad_norm": 1.0402073862601398,
+      "learning_rate": 3.151584092934856e-05,
+      "loss": 12.2199,
+      "step": 12830
+    },
+    {
+      "epoch": 2.2438032786885245,
+      "grad_norm": 0.9031409499409565,
+      "learning_rate": 3.150208379109647e-05,
+      "loss": 11.9437,
+      "step": 12831
+    },
+    {
+      "epoch": 2.243978142076503,
+      "grad_norm": 0.8818708879741305,
+      "learning_rate": 3.148832909471506e-05,
+      "loss": 11.8778,
+      "step": 12832
+    },
+    {
+      "epoch": 2.2441530054644807,
+      "grad_norm": 0.8171194667487122,
+      "learning_rate": 3.147457684069467e-05,
+      "loss": 11.8999,
+      "step": 12833
+    },
+    {
+      "epoch": 2.244327868852459,
+      "grad_norm": 0.8513811658419363,
+      "learning_rate": 3.1460827029525586e-05,
+      "loss": 12.2418,
+      "step": 12834
+    },
+    {
+      "epoch": 2.244502732240437,
+      "grad_norm": 0.8181064550946746,
+      "learning_rate": 3.14470796616979e-05,
+      "loss": 12.0801,
+      "step": 12835
+    },
+    {
+      "epoch": 2.2446775956284153,
+      "grad_norm": 0.8768793638066877,
+      "learning_rate": 3.143333473770174e-05,
+      "loss": 12.0781,
+      "step": 12836
+    },
+    {
+      "epoch": 2.2448524590163936,
+      "grad_norm": 0.8530373268818036,
+      "learning_rate": 3.141959225802708e-05,
+      "loss": 11.7867,
+      "step": 12837
+    },
+    {
+      "epoch": 2.2450273224043715,
+      "grad_norm": 0.8865618721739034,
+      "learning_rate": 3.140585222316386e-05,
+      "loss": 12.1962,
+      "step": 12838
+    },
+    {
+      "epoch": 2.24520218579235,
+      "grad_norm": 0.8726632865765696,
+      "learning_rate": 3.139211463360182e-05,
+      "loss": 12.1011,
+      "step": 12839
+    },
+    {
+      "epoch": 2.2453770491803278,
+      "grad_norm": 0.9024032617898454,
+      "learning_rate": 3.137837948983076e-05,
+      "loss": 11.8421,
+      "step": 12840
+    },
+    {
+      "epoch": 2.245551912568306,
+      "grad_norm": 0.9289675304969274,
+      "learning_rate": 3.136464679234027e-05,
+      "loss": 12.1518,
+      "step": 12841
+    },
+    {
+      "epoch": 2.245726775956284,
+      "grad_norm": 0.8115191110730365,
+      "learning_rate": 3.1350916541619947e-05,
+      "loss": 12.0368,
+      "step": 12842
+    },
+    {
+      "epoch": 2.2459016393442623,
+      "grad_norm": 0.9252519409468836,
+      "learning_rate": 3.133718873815923e-05,
+      "loss": 12.2725,
+      "step": 12843
+    },
+    {
+      "epoch": 2.2460765027322402,
+      "grad_norm": 0.8895476409029272,
+      "learning_rate": 3.132346338244752e-05,
+      "loss": 12.1994,
+      "step": 12844
+    },
+    {
+      "epoch": 2.2462513661202186,
+      "grad_norm": 0.8402230015095233,
+      "learning_rate": 3.1309740474974124e-05,
+      "loss": 12.2195,
+      "step": 12845
+    },
+    {
+      "epoch": 2.246426229508197,
+      "grad_norm": 0.8553708868438126,
+      "learning_rate": 3.12960200162282e-05,
+      "loss": 11.925,
+      "step": 12846
+    },
+    {
+      "epoch": 2.246601092896175,
+      "grad_norm": 0.8952547367566637,
+      "learning_rate": 3.128230200669888e-05,
+      "loss": 12.1025,
+      "step": 12847
+    },
+    {
+      "epoch": 2.246775956284153,
+      "grad_norm": 0.8374721815645847,
+      "learning_rate": 3.126858644687521e-05,
+      "loss": 12.2005,
+      "step": 12848
+    },
+    {
+      "epoch": 2.246950819672131,
+      "grad_norm": 0.8635107867017408,
+      "learning_rate": 3.125487333724617e-05,
+      "loss": 12.2533,
+      "step": 12849
+    },
+    {
+      "epoch": 2.2471256830601094,
+      "grad_norm": 0.8505825321540382,
+      "learning_rate": 3.124116267830054e-05,
+      "loss": 11.9669,
+      "step": 12850
+    },
+    {
+      "epoch": 2.2473005464480873,
+      "grad_norm": 0.8558864973171364,
+      "learning_rate": 3.122745447052712e-05,
+      "loss": 12.0775,
+      "step": 12851
+    },
+    {
+      "epoch": 2.2474754098360656,
+      "grad_norm": 0.8156733392821259,
+      "learning_rate": 3.1213748714414605e-05,
+      "loss": 12.0808,
+      "step": 12852
+    },
+    {
+      "epoch": 2.2476502732240435,
+      "grad_norm": 0.8304165926396554,
+      "learning_rate": 3.1200045410451596e-05,
+      "loss": 12.0867,
+      "step": 12853
+    },
+    {
+      "epoch": 2.247825136612022,
+      "grad_norm": 0.8317648464103328,
+      "learning_rate": 3.118634455912657e-05,
+      "loss": 12.1808,
+      "step": 12854
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 0.7805981215224739,
+      "learning_rate": 3.1172646160927945e-05,
+      "loss": 11.9382,
+      "step": 12855
+    },
+    {
+      "epoch": 2.248174863387978,
+      "grad_norm": 0.8940479919496156,
+      "learning_rate": 3.1158950216344074e-05,
+      "loss": 11.9509,
+      "step": 12856
+    },
+    {
+      "epoch": 2.2483497267759565,
+      "grad_norm": 0.9695190848383867,
+      "learning_rate": 3.1145256725863225e-05,
+      "loss": 12.027,
+      "step": 12857
+    },
+    {
+      "epoch": 2.2485245901639344,
+      "grad_norm": 1.0181027877037832,
+      "learning_rate": 3.11315656899735e-05,
+      "loss": 11.8628,
+      "step": 12858
+    },
+    {
+      "epoch": 2.2486994535519127,
+      "grad_norm": 0.8799926624894925,
+      "learning_rate": 3.1117877109163e-05,
+      "loss": 12.0847,
+      "step": 12859
+    },
+    {
+      "epoch": 2.2488743169398906,
+      "grad_norm": 0.8592492887038827,
+      "learning_rate": 3.110419098391969e-05,
+      "loss": 12.2748,
+      "step": 12860
+    },
+    {
+      "epoch": 2.249049180327869,
+      "grad_norm": 0.8561062396798559,
+      "learning_rate": 3.109050731473148e-05,
+      "loss": 11.9514,
+      "step": 12861
+    },
+    {
+      "epoch": 2.249224043715847,
+      "grad_norm": 0.8637109255867187,
+      "learning_rate": 3.1076826102086164e-05,
+      "loss": 12.0231,
+      "step": 12862
+    },
+    {
+      "epoch": 2.249398907103825,
+      "grad_norm": 0.9014399326536477,
+      "learning_rate": 3.106314734647147e-05,
+      "loss": 11.9821,
+      "step": 12863
+    },
+    {
+      "epoch": 2.249573770491803,
+      "grad_norm": 0.9224140040413038,
+      "learning_rate": 3.104947104837507e-05,
+      "loss": 12.192,
+      "step": 12864
+    },
+    {
+      "epoch": 2.2497486338797814,
+      "grad_norm": 0.9652279214798163,
+      "learning_rate": 3.103579720828442e-05,
+      "loss": 11.9455,
+      "step": 12865
+    },
+    {
+      "epoch": 2.2499234972677598,
+      "grad_norm": 0.7915819243156879,
+      "learning_rate": 3.1022125826687044e-05,
+      "loss": 12.1323,
+      "step": 12866
+    },
+    {
+      "epoch": 2.2500983606557377,
+      "grad_norm": 0.8152792071456124,
+      "learning_rate": 3.100845690407028e-05,
+      "loss": 12.0483,
+      "step": 12867
+    },
+    {
+      "epoch": 2.250273224043716,
+      "grad_norm": 0.8071261952070927,
+      "learning_rate": 3.099479044092145e-05,
+      "loss": 12.0022,
+      "step": 12868
+    },
+    {
+      "epoch": 2.250448087431694,
+      "grad_norm": 0.8468226211846889,
+      "learning_rate": 3.098112643772768e-05,
+      "loss": 12.0359,
+      "step": 12869
+    },
+    {
+      "epoch": 2.2506229508196722,
+      "grad_norm": 0.8418881171758219,
+      "learning_rate": 3.096746489497613e-05,
+      "loss": 12.1408,
+      "step": 12870
+    },
+    {
+      "epoch": 2.25079781420765,
+      "grad_norm": 0.8037335074220817,
+      "learning_rate": 3.095380581315379e-05,
+      "loss": 12.1941,
+      "step": 12871
+    },
+    {
+      "epoch": 2.2509726775956285,
+      "grad_norm": 0.9485282855758796,
+      "learning_rate": 3.0940149192747625e-05,
+      "loss": 12.2307,
+      "step": 12872
+    },
+    {
+      "epoch": 2.2511475409836064,
+      "grad_norm": 0.8906687748615866,
+      "learning_rate": 3.0926495034244426e-05,
+      "loss": 12.0698,
+      "step": 12873
+    },
+    {
+      "epoch": 2.2513224043715847,
+      "grad_norm": 0.8534882801175264,
+      "learning_rate": 3.091284333813098e-05,
+      "loss": 12.3663,
+      "step": 12874
+    },
+    {
+      "epoch": 2.2514972677595626,
+      "grad_norm": 0.9807760242315275,
+      "learning_rate": 3.089919410489396e-05,
+      "loss": 12.345,
+      "step": 12875
+    },
+    {
+      "epoch": 2.251672131147541,
+      "grad_norm": 0.9272390706457035,
+      "learning_rate": 3.088554733501992e-05,
+      "loss": 12.1347,
+      "step": 12876
+    },
+    {
+      "epoch": 2.2518469945355193,
+      "grad_norm": 0.8649005762893046,
+      "learning_rate": 3.087190302899536e-05,
+      "loss": 12.0775,
+      "step": 12877
+    },
+    {
+      "epoch": 2.252021857923497,
+      "grad_norm": 0.9157236965626002,
+      "learning_rate": 3.08582611873067e-05,
+      "loss": 11.8907,
+      "step": 12878
+    },
+    {
+      "epoch": 2.2521967213114755,
+      "grad_norm": 0.8770400488409529,
+      "learning_rate": 3.084462181044028e-05,
+      "loss": 11.9609,
+      "step": 12879
+    },
+    {
+      "epoch": 2.2523715846994534,
+      "grad_norm": 0.8492303377346948,
+      "learning_rate": 3.083098489888225e-05,
+      "loss": 12.0682,
+      "step": 12880
+    },
+    {
+      "epoch": 2.2525464480874318,
+      "grad_norm": 0.8774577910180653,
+      "learning_rate": 3.0817350453118795e-05,
+      "loss": 12.1263,
+      "step": 12881
+    },
+    {
+      "epoch": 2.2527213114754097,
+      "grad_norm": 0.9803711829496223,
+      "learning_rate": 3.0803718473635976e-05,
+      "loss": 11.9811,
+      "step": 12882
+    },
+    {
+      "epoch": 2.252896174863388,
+      "grad_norm": 0.8170082109350317,
+      "learning_rate": 3.079008896091977e-05,
+      "loss": 11.9985,
+      "step": 12883
+    },
+    {
+      "epoch": 2.253071038251366,
+      "grad_norm": 0.8149453023514561,
+      "learning_rate": 3.0776461915456004e-05,
+      "loss": 11.7113,
+      "step": 12884
+    },
+    {
+      "epoch": 2.2532459016393442,
+      "grad_norm": 0.8137390833867906,
+      "learning_rate": 3.0762837337730485e-05,
+      "loss": 11.9069,
+      "step": 12885
+    },
+    {
+      "epoch": 2.2534207650273226,
+      "grad_norm": 0.8147765115524876,
+      "learning_rate": 3.074921522822892e-05,
+      "loss": 12.0359,
+      "step": 12886
+    },
+    {
+      "epoch": 2.2535956284153005,
+      "grad_norm": 0.9486728645959789,
+      "learning_rate": 3.0735595587436963e-05,
+      "loss": 12.0396,
+      "step": 12887
+    },
+    {
+      "epoch": 2.253770491803279,
+      "grad_norm": 0.891464620209064,
+      "learning_rate": 3.0721978415840046e-05,
+      "loss": 12.0598,
+      "step": 12888
+    },
+    {
+      "epoch": 2.2539453551912567,
+      "grad_norm": 0.7601319080671691,
+      "learning_rate": 3.070836371392367e-05,
+      "loss": 12.1736,
+      "step": 12889
+    },
+    {
+      "epoch": 2.254120218579235,
+      "grad_norm": 0.9543057607751585,
+      "learning_rate": 3.06947514821732e-05,
+      "loss": 11.9465,
+      "step": 12890
+    },
+    {
+      "epoch": 2.254295081967213,
+      "grad_norm": 0.9848325255238769,
+      "learning_rate": 3.068114172107383e-05,
+      "loss": 12.0956,
+      "step": 12891
+    },
+    {
+      "epoch": 2.2544699453551913,
+      "grad_norm": 0.8789633182562583,
+      "learning_rate": 3.0667534431110765e-05,
+      "loss": 12.0954,
+      "step": 12892
+    },
+    {
+      "epoch": 2.254644808743169,
+      "grad_norm": 0.8741791358178752,
+      "learning_rate": 3.065392961276909e-05,
+      "loss": 12.0138,
+      "step": 12893
+    },
+    {
+      "epoch": 2.2548196721311475,
+      "grad_norm": 0.8576054584768613,
+      "learning_rate": 3.064032726653381e-05,
+      "loss": 12.1158,
+      "step": 12894
+    },
+    {
+      "epoch": 2.254994535519126,
+      "grad_norm": 0.8941864170447923,
+      "learning_rate": 3.062672739288982e-05,
+      "loss": 12.0341,
+      "step": 12895
+    },
+    {
+      "epoch": 2.2551693989071038,
+      "grad_norm": 0.9022362827227072,
+      "learning_rate": 3.061312999232195e-05,
+      "loss": 12.2007,
+      "step": 12896
+    },
+    {
+      "epoch": 2.255344262295082,
+      "grad_norm": 0.9043837254736968,
+      "learning_rate": 3.059953506531492e-05,
+      "loss": 12.1392,
+      "step": 12897
+    },
+    {
+      "epoch": 2.25551912568306,
+      "grad_norm": 0.8456058142606387,
+      "learning_rate": 3.058594261235341e-05,
+      "loss": 11.9938,
+      "step": 12898
+    },
+    {
+      "epoch": 2.2556939890710384,
+      "grad_norm": 0.9480560283871408,
+      "learning_rate": 3.057235263392192e-05,
+      "loss": 12.195,
+      "step": 12899
+    },
+    {
+      "epoch": 2.2558688524590162,
+      "grad_norm": 0.8962785952923736,
+      "learning_rate": 3.055876513050493e-05,
+      "loss": 12.0573,
+      "step": 12900
+    },
+    {
+      "epoch": 2.2560437158469946,
+      "grad_norm": 0.8248551106630213,
+      "learning_rate": 3.054518010258684e-05,
+      "loss": 11.9762,
+      "step": 12901
+    },
+    {
+      "epoch": 2.2562185792349725,
+      "grad_norm": 0.8324139185884613,
+      "learning_rate": 3.053159755065196e-05,
+      "loss": 12.1516,
+      "step": 12902
+    },
+    {
+      "epoch": 2.256393442622951,
+      "grad_norm": 0.8128950127895954,
+      "learning_rate": 3.051801747518442e-05,
+      "loss": 11.9337,
+      "step": 12903
+    },
+    {
+      "epoch": 2.256568306010929,
+      "grad_norm": 0.9124613410688993,
+      "learning_rate": 3.050443987666839e-05,
+      "loss": 12.0521,
+      "step": 12904
+    },
+    {
+      "epoch": 2.256743169398907,
+      "grad_norm": 0.8522073461965881,
+      "learning_rate": 3.0490864755587912e-05,
+      "loss": 12.0064,
+      "step": 12905
+    },
+    {
+      "epoch": 2.2569180327868854,
+      "grad_norm": 0.980273621119885,
+      "learning_rate": 3.0477292112426848e-05,
+      "loss": 11.8829,
+      "step": 12906
+    },
+    {
+      "epoch": 2.2570928961748633,
+      "grad_norm": 0.8326885271436021,
+      "learning_rate": 3.04637219476691e-05,
+      "loss": 12.0133,
+      "step": 12907
+    },
+    {
+      "epoch": 2.2572677595628416,
+      "grad_norm": 0.8631729447505501,
+      "learning_rate": 3.0450154261798425e-05,
+      "loss": 11.9495,
+      "step": 12908
+    },
+    {
+      "epoch": 2.2574426229508195,
+      "grad_norm": 0.8181846802096049,
+      "learning_rate": 3.0436589055298515e-05,
+      "loss": 12.0611,
+      "step": 12909
+    },
+    {
+      "epoch": 2.257617486338798,
+      "grad_norm": 0.8187359431260096,
+      "learning_rate": 3.0423026328652905e-05,
+      "loss": 12.2881,
+      "step": 12910
+    },
+    {
+      "epoch": 2.257792349726776,
+      "grad_norm": 0.8470814731043622,
+      "learning_rate": 3.0409466082345105e-05,
+      "loss": 11.9791,
+      "step": 12911
+    },
+    {
+      "epoch": 2.257967213114754,
+      "grad_norm": 0.8762019634626583,
+      "learning_rate": 3.0395908316858536e-05,
+      "loss": 12.1976,
+      "step": 12912
+    },
+    {
+      "epoch": 2.2581420765027325,
+      "grad_norm": 0.8875509558788486,
+      "learning_rate": 3.0382353032676512e-05,
+      "loss": 11.9833,
+      "step": 12913
+    },
+    {
+      "epoch": 2.2583169398907104,
+      "grad_norm": 0.8780795204193153,
+      "learning_rate": 3.0368800230282258e-05,
+      "loss": 12.1399,
+      "step": 12914
+    },
+    {
+      "epoch": 2.2584918032786887,
+      "grad_norm": 0.8820409278534461,
+      "learning_rate": 3.0355249910158922e-05,
+      "loss": 11.9354,
+      "step": 12915
+    },
+    {
+      "epoch": 2.2586666666666666,
+      "grad_norm": 0.8345708378903154,
+      "learning_rate": 3.034170207278957e-05,
+      "loss": 11.8819,
+      "step": 12916
+    },
+    {
+      "epoch": 2.258841530054645,
+      "grad_norm": 0.8065708174236222,
+      "learning_rate": 3.0328156718657164e-05,
+      "loss": 12.1553,
+      "step": 12917
+    },
+    {
+      "epoch": 2.259016393442623,
+      "grad_norm": 0.8511521622139839,
+      "learning_rate": 3.0314613848244545e-05,
+      "loss": 12.2075,
+      "step": 12918
+    },
+    {
+      "epoch": 2.259191256830601,
+      "grad_norm": 0.9282106623006036,
+      "learning_rate": 3.0301073462034524e-05,
+      "loss": 11.9787,
+      "step": 12919
+    },
+    {
+      "epoch": 2.259366120218579,
+      "grad_norm": 0.8974611629537453,
+      "learning_rate": 3.0287535560509796e-05,
+      "loss": 12.0654,
+      "step": 12920
+    },
+    {
+      "epoch": 2.2595409836065574,
+      "grad_norm": 0.8228890448449963,
+      "learning_rate": 3.0274000144153015e-05,
+      "loss": 12.1385,
+      "step": 12921
+    },
+    {
+      "epoch": 2.2597158469945358,
+      "grad_norm": 0.8601568298753731,
+      "learning_rate": 3.026046721344662e-05,
+      "loss": 11.9811,
+      "step": 12922
+    },
+    {
+      "epoch": 2.2598907103825137,
+      "grad_norm": 0.8361583838558205,
+      "learning_rate": 3.0246936768873092e-05,
+      "loss": 11.9602,
+      "step": 12923
+    },
+    {
+      "epoch": 2.2600655737704916,
+      "grad_norm": 0.9053652352552864,
+      "learning_rate": 3.0233408810914808e-05,
+      "loss": 12.0149,
+      "step": 12924
+    },
+    {
+      "epoch": 2.26024043715847,
+      "grad_norm": 0.859488938156791,
+      "learning_rate": 3.0219883340053945e-05,
+      "loss": 12.0122,
+      "step": 12925
+    },
+    {
+      "epoch": 2.2604153005464482,
+      "grad_norm": 0.8132783634486125,
+      "learning_rate": 3.0206360356772724e-05,
+      "loss": 12.0152,
+      "step": 12926
+    },
+    {
+      "epoch": 2.260590163934426,
+      "grad_norm": 0.9073203938994129,
+      "learning_rate": 3.0192839861553203e-05,
+      "loss": 11.9282,
+      "step": 12927
+    },
+    {
+      "epoch": 2.2607650273224045,
+      "grad_norm": 0.7752511741103251,
+      "learning_rate": 3.0179321854877406e-05,
+      "loss": 11.7625,
+      "step": 12928
+    },
+    {
+      "epoch": 2.2609398907103824,
+      "grad_norm": 0.8491226293938456,
+      "learning_rate": 3.0165806337227186e-05,
+      "loss": 12.1839,
+      "step": 12929
+    },
+    {
+      "epoch": 2.2611147540983607,
+      "grad_norm": 0.8534985682823938,
+      "learning_rate": 3.0152293309084366e-05,
+      "loss": 12.0494,
+      "step": 12930
+    },
+    {
+      "epoch": 2.2612896174863386,
+      "grad_norm": 0.8192565903447441,
+      "learning_rate": 3.0138782770930697e-05,
+      "loss": 12.313,
+      "step": 12931
+    },
+    {
+      "epoch": 2.261464480874317,
+      "grad_norm": 0.8446103026995326,
+      "learning_rate": 3.0125274723247786e-05,
+      "loss": 11.9856,
+      "step": 12932
+    },
+    {
+      "epoch": 2.261639344262295,
+      "grad_norm": 0.8256237462781437,
+      "learning_rate": 3.011176916651719e-05,
+      "loss": 12.2824,
+      "step": 12933
+    },
+    {
+      "epoch": 2.261814207650273,
+      "grad_norm": 0.8637889693256772,
+      "learning_rate": 3.0098266101220373e-05,
+      "loss": 12.2017,
+      "step": 12934
+    },
+    {
+      "epoch": 2.2619890710382515,
+      "grad_norm": 0.8244419129964321,
+      "learning_rate": 3.00847655278387e-05,
+      "loss": 12.0424,
+      "step": 12935
+    },
+    {
+      "epoch": 2.2621639344262294,
+      "grad_norm": 0.8454497584001356,
+      "learning_rate": 3.007126744685348e-05,
+      "loss": 12.0564,
+      "step": 12936
+    },
+    {
+      "epoch": 2.2623387978142078,
+      "grad_norm": 0.823661946311377,
+      "learning_rate": 3.0057771858745844e-05,
+      "loss": 11.9042,
+      "step": 12937
+    },
+    {
+      "epoch": 2.2625136612021857,
+      "grad_norm": 0.8583281153335589,
+      "learning_rate": 3.0044278763996914e-05,
+      "loss": 11.7822,
+      "step": 12938
+    },
+    {
+      "epoch": 2.262688524590164,
+      "grad_norm": 0.861175919719825,
+      "learning_rate": 3.0030788163087752e-05,
+      "loss": 12.0739,
+      "step": 12939
+    },
+    {
+      "epoch": 2.262863387978142,
+      "grad_norm": 0.8326485081708429,
+      "learning_rate": 3.00173000564992e-05,
+      "loss": 12.0051,
+      "step": 12940
+    },
+    {
+      "epoch": 2.2630382513661202,
+      "grad_norm": 0.805668503156057,
+      "learning_rate": 3.0003814444712142e-05,
+      "loss": 11.9489,
+      "step": 12941
+    },
+    {
+      "epoch": 2.263213114754098,
+      "grad_norm": 0.833825304909628,
+      "learning_rate": 2.999033132820732e-05,
+      "loss": 12.0233,
+      "step": 12942
+    },
+    {
+      "epoch": 2.2633879781420765,
+      "grad_norm": 0.8892809189223377,
+      "learning_rate": 2.9976850707465408e-05,
+      "loss": 12.0623,
+      "step": 12943
+    },
+    {
+      "epoch": 2.263562841530055,
+      "grad_norm": 0.8963120607558218,
+      "learning_rate": 2.996337258296693e-05,
+      "loss": 12.2133,
+      "step": 12944
+    },
+    {
+      "epoch": 2.2637377049180327,
+      "grad_norm": 0.8404862681875819,
+      "learning_rate": 2.9949896955192402e-05,
+      "loss": 11.9578,
+      "step": 12945
+    },
+    {
+      "epoch": 2.263912568306011,
+      "grad_norm": 0.8224395966975523,
+      "learning_rate": 2.9936423824622183e-05,
+      "loss": 12.0851,
+      "step": 12946
+    },
+    {
+      "epoch": 2.264087431693989,
+      "grad_norm": 0.8819885874629065,
+      "learning_rate": 2.9922953191736637e-05,
+      "loss": 12.3209,
+      "step": 12947
+    },
+    {
+      "epoch": 2.2642622950819673,
+      "grad_norm": 0.9488946763410495,
+      "learning_rate": 2.9909485057015906e-05,
+      "loss": 12.1757,
+      "step": 12948
+    },
+    {
+      "epoch": 2.264437158469945,
+      "grad_norm": 0.7727482092806596,
+      "learning_rate": 2.9896019420940137e-05,
+      "loss": 11.9569,
+      "step": 12949
+    },
+    {
+      "epoch": 2.2646120218579235,
+      "grad_norm": 0.7758363852675598,
+      "learning_rate": 2.988255628398936e-05,
+      "loss": 12.3143,
+      "step": 12950
+    },
+    {
+      "epoch": 2.2647868852459014,
+      "grad_norm": 0.9272313889238037,
+      "learning_rate": 2.986909564664353e-05,
+      "loss": 12.1842,
+      "step": 12951
+    },
+    {
+      "epoch": 2.26496174863388,
+      "grad_norm": 0.8851021245454141,
+      "learning_rate": 2.9855637509382504e-05,
+      "loss": 12.0198,
+      "step": 12952
+    },
+    {
+      "epoch": 2.265136612021858,
+      "grad_norm": 0.8896632371796577,
+      "learning_rate": 2.9842181872686047e-05,
+      "loss": 12.2656,
+      "step": 12953
+    },
+    {
+      "epoch": 2.265311475409836,
+      "grad_norm": 0.9298526601921041,
+      "learning_rate": 2.9828728737033872e-05,
+      "loss": 12.1706,
+      "step": 12954
+    },
+    {
+      "epoch": 2.2654863387978144,
+      "grad_norm": 0.9203188116156037,
+      "learning_rate": 2.9815278102905496e-05,
+      "loss": 12.0921,
+      "step": 12955
+    },
+    {
+      "epoch": 2.2656612021857923,
+      "grad_norm": 0.8394465302731736,
+      "learning_rate": 2.980182997078045e-05,
+      "loss": 12.1896,
+      "step": 12956
+    },
+    {
+      "epoch": 2.2658360655737706,
+      "grad_norm": 0.8390582557381216,
+      "learning_rate": 2.9788384341138153e-05,
+      "loss": 12.2456,
+      "step": 12957
+    },
+    {
+      "epoch": 2.2660109289617485,
+      "grad_norm": 0.984157146529398,
+      "learning_rate": 2.9774941214457952e-05,
+      "loss": 11.8354,
+      "step": 12958
+    },
+    {
+      "epoch": 2.266185792349727,
+      "grad_norm": 0.9243019886933824,
+      "learning_rate": 2.9761500591219005e-05,
+      "loss": 12.0792,
+      "step": 12959
+    },
+    {
+      "epoch": 2.2663606557377047,
+      "grad_norm": 0.8206526758706383,
+      "learning_rate": 2.9748062471900507e-05,
+      "loss": 12.0926,
+      "step": 12960
+    },
+    {
+      "epoch": 2.266535519125683,
+      "grad_norm": 0.876106749453926,
+      "learning_rate": 2.9734626856981494e-05,
+      "loss": 12.1335,
+      "step": 12961
+    },
+    {
+      "epoch": 2.2667103825136614,
+      "grad_norm": 0.834998835485231,
+      "learning_rate": 2.972119374694098e-05,
+      "loss": 12.1842,
+      "step": 12962
+    },
+    {
+      "epoch": 2.2668852459016393,
+      "grad_norm": 0.8271017645413988,
+      "learning_rate": 2.970776314225776e-05,
+      "loss": 11.8494,
+      "step": 12963
+    },
+    {
+      "epoch": 2.2670601092896177,
+      "grad_norm": 0.8460695365062342,
+      "learning_rate": 2.969433504341066e-05,
+      "loss": 11.8692,
+      "step": 12964
+    },
+    {
+      "epoch": 2.2672349726775956,
+      "grad_norm": 0.8068440244680367,
+      "learning_rate": 2.968090945087837e-05,
+      "loss": 12.1844,
+      "step": 12965
+    },
+    {
+      "epoch": 2.267409836065574,
+      "grad_norm": 0.7844929621024801,
+      "learning_rate": 2.966748636513953e-05,
+      "loss": 12.1837,
+      "step": 12966
+    },
+    {
+      "epoch": 2.267584699453552,
+      "grad_norm": 0.897272069227403,
+      "learning_rate": 2.9654065786672592e-05,
+      "loss": 12.0622,
+      "step": 12967
+    },
+    {
+      "epoch": 2.26775956284153,
+      "grad_norm": 0.8348520442626691,
+      "learning_rate": 2.9640647715956016e-05,
+      "loss": 12.1357,
+      "step": 12968
+    },
+    {
+      "epoch": 2.267934426229508,
+      "grad_norm": 0.860974452830736,
+      "learning_rate": 2.9627232153468144e-05,
+      "loss": 11.928,
+      "step": 12969
+    },
+    {
+      "epoch": 2.2681092896174864,
+      "grad_norm": 0.835349944882934,
+      "learning_rate": 2.9613819099687236e-05,
+      "loss": 12.0942,
+      "step": 12970
+    },
+    {
+      "epoch": 2.2682841530054647,
+      "grad_norm": 0.8764101638347285,
+      "learning_rate": 2.960040855509143e-05,
+      "loss": 11.7809,
+      "step": 12971
+    },
+    {
+      "epoch": 2.2684590163934426,
+      "grad_norm": 0.8139366719462771,
+      "learning_rate": 2.9587000520158803e-05,
+      "loss": 11.9444,
+      "step": 12972
+    },
+    {
+      "epoch": 2.268633879781421,
+      "grad_norm": 0.8804525887094613,
+      "learning_rate": 2.957359499536737e-05,
+      "loss": 12.0872,
+      "step": 12973
+    },
+    {
+      "epoch": 2.268808743169399,
+      "grad_norm": 0.8918769179595818,
+      "learning_rate": 2.9560191981194964e-05,
+      "loss": 12.3074,
+      "step": 12974
+    },
+    {
+      "epoch": 2.268983606557377,
+      "grad_norm": 0.8348041106700741,
+      "learning_rate": 2.9546791478119418e-05,
+      "loss": 12.1253,
+      "step": 12975
+    },
+    {
+      "epoch": 2.269158469945355,
+      "grad_norm": 0.8635903741802808,
+      "learning_rate": 2.953339348661843e-05,
+      "loss": 11.9297,
+      "step": 12976
+    },
+    {
+      "epoch": 2.2693333333333334,
+      "grad_norm": 0.8156032227607928,
+      "learning_rate": 2.951999800716967e-05,
+      "loss": 11.7247,
+      "step": 12977
+    },
+    {
+      "epoch": 2.2695081967213113,
+      "grad_norm": 0.9445574365729816,
+      "learning_rate": 2.9506605040250603e-05,
+      "loss": 12.1539,
+      "step": 12978
+    },
+    {
+      "epoch": 2.2696830601092897,
+      "grad_norm": 0.8531321362897243,
+      "learning_rate": 2.9493214586338702e-05,
+      "loss": 12.1111,
+      "step": 12979
+    },
+    {
+      "epoch": 2.269857923497268,
+      "grad_norm": 0.8704525184586092,
+      "learning_rate": 2.9479826645911323e-05,
+      "loss": 12.1196,
+      "step": 12980
+    },
+    {
+      "epoch": 2.270032786885246,
+      "grad_norm": 0.8491871624555383,
+      "learning_rate": 2.946644121944576e-05,
+      "loss": 12.2548,
+      "step": 12981
+    },
+    {
+      "epoch": 2.270207650273224,
+      "grad_norm": 0.9876635241231734,
+      "learning_rate": 2.9453058307419135e-05,
+      "loss": 12.0074,
+      "step": 12982
+    },
+    {
+      "epoch": 2.270382513661202,
+      "grad_norm": 0.7877158750015029,
+      "learning_rate": 2.943967791030855e-05,
+      "loss": 12.1333,
+      "step": 12983
+    },
+    {
+      "epoch": 2.2705573770491805,
+      "grad_norm": 0.8639025751289586,
+      "learning_rate": 2.9426300028591004e-05,
+      "loss": 12.2368,
+      "step": 12984
+    },
+    {
+      "epoch": 2.2707322404371584,
+      "grad_norm": 0.8376102972204789,
+      "learning_rate": 2.9412924662743403e-05,
+      "loss": 11.986,
+      "step": 12985
+    },
+    {
+      "epoch": 2.2709071038251367,
+      "grad_norm": 0.7788321598169663,
+      "learning_rate": 2.9399551813242575e-05,
+      "loss": 11.958,
+      "step": 12986
+    },
+    {
+      "epoch": 2.2710819672131146,
+      "grad_norm": 0.7945446991458258,
+      "learning_rate": 2.9386181480565235e-05,
+      "loss": 12.0923,
+      "step": 12987
+    },
+    {
+      "epoch": 2.271256830601093,
+      "grad_norm": 0.8126166611918731,
+      "learning_rate": 2.9372813665188048e-05,
+      "loss": 12.139,
+      "step": 12988
+    },
+    {
+      "epoch": 2.271431693989071,
+      "grad_norm": 0.8189768392201109,
+      "learning_rate": 2.9359448367587505e-05,
+      "loss": 12.0677,
+      "step": 12989
+    },
+    {
+      "epoch": 2.271606557377049,
+      "grad_norm": 0.8245322080471147,
+      "learning_rate": 2.934608558824009e-05,
+      "loss": 12.0279,
+      "step": 12990
+    },
+    {
+      "epoch": 2.271781420765027,
+      "grad_norm": 0.9650115787134257,
+      "learning_rate": 2.933272532762218e-05,
+      "loss": 12.1167,
+      "step": 12991
+    },
+    {
+      "epoch": 2.2719562841530054,
+      "grad_norm": 0.8394586037699784,
+      "learning_rate": 2.931936758621008e-05,
+      "loss": 11.9868,
+      "step": 12992
+    },
+    {
+      "epoch": 2.2721311475409838,
+      "grad_norm": 0.7838177142849466,
+      "learning_rate": 2.9306012364479907e-05,
+      "loss": 11.7298,
+      "step": 12993
+    },
+    {
+      "epoch": 2.2723060109289617,
+      "grad_norm": 0.8455361935348674,
+      "learning_rate": 2.9292659662907794e-05,
+      "loss": 11.7089,
+      "step": 12994
+    },
+    {
+      "epoch": 2.27248087431694,
+      "grad_norm": 0.8172452862316263,
+      "learning_rate": 2.9279309481969754e-05,
+      "loss": 12.1416,
+      "step": 12995
+    },
+    {
+      "epoch": 2.272655737704918,
+      "grad_norm": 0.7967390433265376,
+      "learning_rate": 2.926596182214173e-05,
+      "loss": 12.2695,
+      "step": 12996
+    },
+    {
+      "epoch": 2.2728306010928963,
+      "grad_norm": 0.8223823807815035,
+      "learning_rate": 2.9252616683899503e-05,
+      "loss": 12.1168,
+      "step": 12997
+    },
+    {
+      "epoch": 2.273005464480874,
+      "grad_norm": 0.8217029018366802,
+      "learning_rate": 2.9239274067718826e-05,
+      "loss": 12.1149,
+      "step": 12998
+    },
+    {
+      "epoch": 2.2731803278688525,
+      "grad_norm": 0.7902619706275137,
+      "learning_rate": 2.922593397407538e-05,
+      "loss": 12.1491,
+      "step": 12999
+    },
+    {
+      "epoch": 2.2733551912568304,
+      "grad_norm": 0.8466921575810922,
+      "learning_rate": 2.921259640344467e-05,
+      "loss": 12.1371,
+      "step": 13000
+    },
+    {
+      "epoch": 2.2735300546448087,
+      "grad_norm": 0.9298156935221452,
+      "learning_rate": 2.9199261356302187e-05,
+      "loss": 11.9313,
+      "step": 13001
+    },
+    {
+      "epoch": 2.273704918032787,
+      "grad_norm": 0.7932472839290818,
+      "learning_rate": 2.9185928833123322e-05,
+      "loss": 11.9155,
+      "step": 13002
+    },
+    {
+      "epoch": 2.273879781420765,
+      "grad_norm": 0.8660200668240852,
+      "learning_rate": 2.9172598834383358e-05,
+      "loss": 11.8727,
+      "step": 13003
+    },
+    {
+      "epoch": 2.2740546448087433,
+      "grad_norm": 0.9474191442918932,
+      "learning_rate": 2.915927136055748e-05,
+      "loss": 11.9322,
+      "step": 13004
+    },
+    {
+      "epoch": 2.274229508196721,
+      "grad_norm": 0.8454784056952059,
+      "learning_rate": 2.914594641212083e-05,
+      "loss": 12.1621,
+      "step": 13005
+    },
+    {
+      "epoch": 2.2744043715846995,
+      "grad_norm": 0.8426925629252864,
+      "learning_rate": 2.913262398954839e-05,
+      "loss": 12.0367,
+      "step": 13006
+    },
+    {
+      "epoch": 2.2745792349726774,
+      "grad_norm": 0.9382884129440389,
+      "learning_rate": 2.9119304093315148e-05,
+      "loss": 12.1691,
+      "step": 13007
+    },
+    {
+      "epoch": 2.274754098360656,
+      "grad_norm": 0.894863285147315,
+      "learning_rate": 2.910598672389585e-05,
+      "loss": 12.0957,
+      "step": 13008
+    },
+    {
+      "epoch": 2.2749289617486337,
+      "grad_norm": 0.8097945358291004,
+      "learning_rate": 2.909267188176531e-05,
+      "loss": 11.7673,
+      "step": 13009
+    },
+    {
+      "epoch": 2.275103825136612,
+      "grad_norm": 0.8280309179683489,
+      "learning_rate": 2.907935956739817e-05,
+      "loss": 12.0025,
+      "step": 13010
+    },
+    {
+      "epoch": 2.2752786885245904,
+      "grad_norm": 0.8851200678020283,
+      "learning_rate": 2.906604978126902e-05,
+      "loss": 12.1041,
+      "step": 13011
+    },
+    {
+      "epoch": 2.2754535519125683,
+      "grad_norm": 0.8719431317823629,
+      "learning_rate": 2.9052742523852306e-05,
+      "loss": 12.0264,
+      "step": 13012
+    },
+    {
+      "epoch": 2.2756284153005466,
+      "grad_norm": 0.8734654485157795,
+      "learning_rate": 2.9039437795622414e-05,
+      "loss": 12.0713,
+      "step": 13013
+    },
+    {
+      "epoch": 2.2758032786885245,
+      "grad_norm": 0.8053127641817104,
+      "learning_rate": 2.9026135597053693e-05,
+      "loss": 12.1357,
+      "step": 13014
+    },
+    {
+      "epoch": 2.275978142076503,
+      "grad_norm": 0.9255817152831223,
+      "learning_rate": 2.9012835928620276e-05,
+      "loss": 11.9229,
+      "step": 13015
+    },
+    {
+      "epoch": 2.2761530054644807,
+      "grad_norm": 0.9653861289006438,
+      "learning_rate": 2.8999538790796322e-05,
+      "loss": 12.1828,
+      "step": 13016
+    },
+    {
+      "epoch": 2.276327868852459,
+      "grad_norm": 0.8487287463041404,
+      "learning_rate": 2.8986244184055856e-05,
+      "loss": 11.6454,
+      "step": 13017
+    },
+    {
+      "epoch": 2.276502732240437,
+      "grad_norm": 0.8047268642639789,
+      "learning_rate": 2.8972952108872842e-05,
+      "loss": 12.0457,
+      "step": 13018
+    },
+    {
+      "epoch": 2.2766775956284153,
+      "grad_norm": 0.8850671617518158,
+      "learning_rate": 2.895966256572107e-05,
+      "loss": 12.0192,
+      "step": 13019
+    },
+    {
+      "epoch": 2.2768524590163937,
+      "grad_norm": 0.8972286651082022,
+      "learning_rate": 2.8946375555074334e-05,
+      "loss": 12.0389,
+      "step": 13020
+    },
+    {
+      "epoch": 2.2770273224043716,
+      "grad_norm": 0.8391740926137814,
+      "learning_rate": 2.8933091077406283e-05,
+      "loss": 11.9305,
+      "step": 13021
+    },
+    {
+      "epoch": 2.27720218579235,
+      "grad_norm": 0.9054922025136478,
+      "learning_rate": 2.89198091331905e-05,
+      "loss": 12.2474,
+      "step": 13022
+    },
+    {
+      "epoch": 2.277377049180328,
+      "grad_norm": 0.8918837696969971,
+      "learning_rate": 2.8906529722900477e-05,
+      "loss": 11.9637,
+      "step": 13023
+    },
+    {
+      "epoch": 2.277551912568306,
+      "grad_norm": 0.8798730223261083,
+      "learning_rate": 2.8893252847009612e-05,
+      "loss": 12.0185,
+      "step": 13024
+    },
+    {
+      "epoch": 2.277726775956284,
+      "grad_norm": 0.8426941443360343,
+      "learning_rate": 2.887997850599119e-05,
+      "loss": 11.9177,
+      "step": 13025
+    },
+    {
+      "epoch": 2.2779016393442624,
+      "grad_norm": 0.8580013587505719,
+      "learning_rate": 2.886670670031848e-05,
+      "loss": 11.9275,
+      "step": 13026
+    },
+    {
+      "epoch": 2.2780765027322403,
+      "grad_norm": 0.862962025448585,
+      "learning_rate": 2.8853437430464538e-05,
+      "loss": 12.155,
+      "step": 13027
+    },
+    {
+      "epoch": 2.2782513661202186,
+      "grad_norm": 0.973166968715972,
+      "learning_rate": 2.8840170696902412e-05,
+      "loss": 12.1275,
+      "step": 13028
+    },
+    {
+      "epoch": 2.278426229508197,
+      "grad_norm": 0.8043737020482826,
+      "learning_rate": 2.8826906500105077e-05,
+      "loss": 12.0723,
+      "step": 13029
+    },
+    {
+      "epoch": 2.278601092896175,
+      "grad_norm": 0.8289918558077266,
+      "learning_rate": 2.8813644840545385e-05,
+      "loss": 12.04,
+      "step": 13030
+    },
+    {
+      "epoch": 2.278775956284153,
+      "grad_norm": 0.8688473046242918,
+      "learning_rate": 2.8800385718696053e-05,
+      "loss": 12.1419,
+      "step": 13031
+    },
+    {
+      "epoch": 2.278950819672131,
+      "grad_norm": 0.8906998423178494,
+      "learning_rate": 2.8787129135029788e-05,
+      "loss": 12.1129,
+      "step": 13032
+    },
+    {
+      "epoch": 2.2791256830601094,
+      "grad_norm": 0.93389123623186,
+      "learning_rate": 2.8773875090019196e-05,
+      "loss": 11.9044,
+      "step": 13033
+    },
+    {
+      "epoch": 2.2793005464480873,
+      "grad_norm": 0.8146426701264753,
+      "learning_rate": 2.8760623584136702e-05,
+      "loss": 12.0395,
+      "step": 13034
+    },
+    {
+      "epoch": 2.2794754098360657,
+      "grad_norm": 0.7941720137879412,
+      "learning_rate": 2.874737461785475e-05,
+      "loss": 12.1096,
+      "step": 13035
+    },
+    {
+      "epoch": 2.2796502732240436,
+      "grad_norm": 0.8083802294973186,
+      "learning_rate": 2.8734128191645636e-05,
+      "loss": 12.0217,
+      "step": 13036
+    },
+    {
+      "epoch": 2.279825136612022,
+      "grad_norm": 0.9378595778999188,
+      "learning_rate": 2.8720884305981634e-05,
+      "loss": 12.0335,
+      "step": 13037
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.8806697732792731,
+      "learning_rate": 2.870764296133478e-05,
+      "loss": 11.8418,
+      "step": 13038
+    },
+    {
+      "epoch": 2.280174863387978,
+      "grad_norm": 0.9344308534818359,
+      "learning_rate": 2.8694404158177157e-05,
+      "loss": 12.2838,
+      "step": 13039
+    },
+    {
+      "epoch": 2.280349726775956,
+      "grad_norm": 0.8418506399134555,
+      "learning_rate": 2.868116789698072e-05,
+      "loss": 12.0002,
+      "step": 13040
+    },
+    {
+      "epoch": 2.2805245901639344,
+      "grad_norm": 0.8063076601667165,
+      "learning_rate": 2.8667934178217314e-05,
+      "loss": 11.9835,
+      "step": 13041
+    },
+    {
+      "epoch": 2.2806994535519127,
+      "grad_norm": 0.8308799987000758,
+      "learning_rate": 2.865470300235872e-05,
+      "loss": 12.173,
+      "step": 13042
+    },
+    {
+      "epoch": 2.2808743169398906,
+      "grad_norm": 0.844794530552277,
+      "learning_rate": 2.8641474369876597e-05,
+      "loss": 12.0361,
+      "step": 13043
+    },
+    {
+      "epoch": 2.281049180327869,
+      "grad_norm": 0.8329436718961369,
+      "learning_rate": 2.862824828124254e-05,
+      "loss": 12.2284,
+      "step": 13044
+    },
+    {
+      "epoch": 2.281224043715847,
+      "grad_norm": 0.8579406627851199,
+      "learning_rate": 2.8615024736928085e-05,
+      "loss": 11.943,
+      "step": 13045
+    },
+    {
+      "epoch": 2.281398907103825,
+      "grad_norm": 0.8186211232605328,
+      "learning_rate": 2.8601803737404552e-05,
+      "loss": 12.1102,
+      "step": 13046
+    },
+    {
+      "epoch": 2.281573770491803,
+      "grad_norm": 0.7953696954857865,
+      "learning_rate": 2.858858528314331e-05,
+      "loss": 12.0617,
+      "step": 13047
+    },
+    {
+      "epoch": 2.2817486338797814,
+      "grad_norm": 0.9237390009541817,
+      "learning_rate": 2.857536937461558e-05,
+      "loss": 11.9063,
+      "step": 13048
+    },
+    {
+      "epoch": 2.2819234972677593,
+      "grad_norm": 0.8674664091521914,
+      "learning_rate": 2.8562156012292463e-05,
+      "loss": 12.0724,
+      "step": 13049
+    },
+    {
+      "epoch": 2.2820983606557377,
+      "grad_norm": 0.9098289585476657,
+      "learning_rate": 2.8548945196645018e-05,
+      "loss": 12.2399,
+      "step": 13050
+    },
+    {
+      "epoch": 2.282273224043716,
+      "grad_norm": 0.910868372423646,
+      "learning_rate": 2.8535736928144186e-05,
+      "loss": 12.163,
+      "step": 13051
+    },
+    {
+      "epoch": 2.282448087431694,
+      "grad_norm": 0.7835041527486472,
+      "learning_rate": 2.8522531207260873e-05,
+      "loss": 12.0417,
+      "step": 13052
+    },
+    {
+      "epoch": 2.2826229508196723,
+      "grad_norm": 0.8234560535011876,
+      "learning_rate": 2.8509328034465776e-05,
+      "loss": 12.1124,
+      "step": 13053
+    },
+    {
+      "epoch": 2.28279781420765,
+      "grad_norm": 0.8236262674710952,
+      "learning_rate": 2.84961274102296e-05,
+      "loss": 12.0377,
+      "step": 13054
+    },
+    {
+      "epoch": 2.2829726775956285,
+      "grad_norm": 0.7853681274491795,
+      "learning_rate": 2.8482929335022958e-05,
+      "loss": 12.1072,
+      "step": 13055
+    },
+    {
+      "epoch": 2.2831475409836064,
+      "grad_norm": 0.8794267557102889,
+      "learning_rate": 2.8469733809316335e-05,
+      "loss": 11.675,
+      "step": 13056
+    },
+    {
+      "epoch": 2.2833224043715847,
+      "grad_norm": 0.8711036438683145,
+      "learning_rate": 2.8456540833580113e-05,
+      "loss": 12.1084,
+      "step": 13057
+    },
+    {
+      "epoch": 2.2834972677595626,
+      "grad_norm": 0.8471940106233856,
+      "learning_rate": 2.8443350408284608e-05,
+      "loss": 11.8432,
+      "step": 13058
+    },
+    {
+      "epoch": 2.283672131147541,
+      "grad_norm": 0.8474112840282765,
+      "learning_rate": 2.8430162533900073e-05,
+      "loss": 11.9871,
+      "step": 13059
+    },
+    {
+      "epoch": 2.2838469945355193,
+      "grad_norm": 0.9344477494401127,
+      "learning_rate": 2.841697721089661e-05,
+      "loss": 12.2833,
+      "step": 13060
+    },
+    {
+      "epoch": 2.284021857923497,
+      "grad_norm": 0.7753255343524558,
+      "learning_rate": 2.840379443974428e-05,
+      "loss": 11.8608,
+      "step": 13061
+    },
+    {
+      "epoch": 2.2841967213114756,
+      "grad_norm": 0.8164583755443781,
+      "learning_rate": 2.8390614220913013e-05,
+      "loss": 11.901,
+      "step": 13062
+    },
+    {
+      "epoch": 2.2843715846994535,
+      "grad_norm": 0.8062687998035187,
+      "learning_rate": 2.8377436554872726e-05,
+      "loss": 12.1556,
+      "step": 13063
+    },
+    {
+      "epoch": 2.284546448087432,
+      "grad_norm": 0.9166535381103726,
+      "learning_rate": 2.836426144209312e-05,
+      "loss": 12.0478,
+      "step": 13064
+    },
+    {
+      "epoch": 2.2847213114754097,
+      "grad_norm": 0.8840275848040142,
+      "learning_rate": 2.8351088883043886e-05,
+      "loss": 12.2016,
+      "step": 13065
+    },
+    {
+      "epoch": 2.284896174863388,
+      "grad_norm": 0.8372203207524656,
+      "learning_rate": 2.8337918878194613e-05,
+      "loss": 12.068,
+      "step": 13066
+    },
+    {
+      "epoch": 2.285071038251366,
+      "grad_norm": 0.893134705531078,
+      "learning_rate": 2.8324751428014852e-05,
+      "loss": 11.9166,
+      "step": 13067
+    },
+    {
+      "epoch": 2.2852459016393443,
+      "grad_norm": 0.8804613986566013,
+      "learning_rate": 2.831158653297392e-05,
+      "loss": 11.9131,
+      "step": 13068
+    },
+    {
+      "epoch": 2.2854207650273226,
+      "grad_norm": 0.8285413740662099,
+      "learning_rate": 2.8298424193541163e-05,
+      "loss": 12.008,
+      "step": 13069
+    },
+    {
+      "epoch": 2.2855956284153005,
+      "grad_norm": 0.8714701418015772,
+      "learning_rate": 2.8285264410185818e-05,
+      "loss": 12.0366,
+      "step": 13070
+    },
+    {
+      "epoch": 2.285770491803279,
+      "grad_norm": 0.8425781441256257,
+      "learning_rate": 2.827210718337704e-05,
+      "loss": 11.9657,
+      "step": 13071
+    },
+    {
+      "epoch": 2.2859453551912567,
+      "grad_norm": 0.9251346093458096,
+      "learning_rate": 2.825895251358379e-05,
+      "loss": 11.9011,
+      "step": 13072
+    },
+    {
+      "epoch": 2.286120218579235,
+      "grad_norm": 0.8299593054749272,
+      "learning_rate": 2.824580040127508e-05,
+      "loss": 11.9126,
+      "step": 13073
+    },
+    {
+      "epoch": 2.286295081967213,
+      "grad_norm": 0.9763833355361766,
+      "learning_rate": 2.8232650846919738e-05,
+      "loss": 12.2859,
+      "step": 13074
+    },
+    {
+      "epoch": 2.2864699453551913,
+      "grad_norm": 0.8636271984399464,
+      "learning_rate": 2.821950385098655e-05,
+      "loss": 12.1877,
+      "step": 13075
+    },
+    {
+      "epoch": 2.286644808743169,
+      "grad_norm": 0.8354645097505151,
+      "learning_rate": 2.8206359413944184e-05,
+      "loss": 12.2493,
+      "step": 13076
+    },
+    {
+      "epoch": 2.2868196721311476,
+      "grad_norm": 0.8179052277261785,
+      "learning_rate": 2.8193217536261217e-05,
+      "loss": 11.9462,
+      "step": 13077
+    },
+    {
+      "epoch": 2.286994535519126,
+      "grad_norm": 0.7957911084487097,
+      "learning_rate": 2.818007821840618e-05,
+      "loss": 12.1573,
+      "step": 13078
+    },
+    {
+      "epoch": 2.287169398907104,
+      "grad_norm": 0.8432020741446375,
+      "learning_rate": 2.8166941460847406e-05,
+      "loss": 11.999,
+      "step": 13079
+    },
+    {
+      "epoch": 2.287344262295082,
+      "grad_norm": 0.8327984946772752,
+      "learning_rate": 2.8153807264053255e-05,
+      "loss": 11.885,
+      "step": 13080
+    },
+    {
+      "epoch": 2.28751912568306,
+      "grad_norm": 0.8358037874250014,
+      "learning_rate": 2.814067562849192e-05,
+      "loss": 11.8778,
+      "step": 13081
+    },
+    {
+      "epoch": 2.2876939890710384,
+      "grad_norm": 0.927887522680803,
+      "learning_rate": 2.8127546554631577e-05,
+      "loss": 12.3117,
+      "step": 13082
+    },
+    {
+      "epoch": 2.2878688524590163,
+      "grad_norm": 0.797380427048137,
+      "learning_rate": 2.8114420042940194e-05,
+      "loss": 11.9976,
+      "step": 13083
+    },
+    {
+      "epoch": 2.2880437158469946,
+      "grad_norm": 0.8122385347462205,
+      "learning_rate": 2.8101296093885742e-05,
+      "loss": 11.8568,
+      "step": 13084
+    },
+    {
+      "epoch": 2.2882185792349725,
+      "grad_norm": 0.8651343956590838,
+      "learning_rate": 2.8088174707936075e-05,
+      "loss": 12.1769,
+      "step": 13085
+    },
+    {
+      "epoch": 2.288393442622951,
+      "grad_norm": 0.7860696194424928,
+      "learning_rate": 2.807505588555901e-05,
+      "loss": 12.125,
+      "step": 13086
+    },
+    {
+      "epoch": 2.288568306010929,
+      "grad_norm": 0.8496095039317618,
+      "learning_rate": 2.8061939627222124e-05,
+      "loss": 12.158,
+      "step": 13087
+    },
+    {
+      "epoch": 2.288743169398907,
+      "grad_norm": 0.859555743336023,
+      "learning_rate": 2.8048825933393042e-05,
+      "loss": 11.6331,
+      "step": 13088
+    },
+    {
+      "epoch": 2.2889180327868854,
+      "grad_norm": 0.805737407214944,
+      "learning_rate": 2.8035714804539237e-05,
+      "loss": 11.9029,
+      "step": 13089
+    },
+    {
+      "epoch": 2.2890928961748633,
+      "grad_norm": 0.8718592884530497,
+      "learning_rate": 2.8022606241128167e-05,
+      "loss": 12.0271,
+      "step": 13090
+    },
+    {
+      "epoch": 2.2892677595628417,
+      "grad_norm": 0.8694552123751249,
+      "learning_rate": 2.8009500243627052e-05,
+      "loss": 12.0676,
+      "step": 13091
+    },
+    {
+      "epoch": 2.2894426229508196,
+      "grad_norm": 0.8227224931229248,
+      "learning_rate": 2.7996396812503135e-05,
+      "loss": 12.0129,
+      "step": 13092
+    },
+    {
+      "epoch": 2.289617486338798,
+      "grad_norm": 0.8117081798133353,
+      "learning_rate": 2.7983295948223552e-05,
+      "loss": 11.976,
+      "step": 13093
+    },
+    {
+      "epoch": 2.289792349726776,
+      "grad_norm": 0.8733529767905962,
+      "learning_rate": 2.7970197651255326e-05,
+      "loss": 11.9883,
+      "step": 13094
+    },
+    {
+      "epoch": 2.289967213114754,
+      "grad_norm": 0.9981986481899319,
+      "learning_rate": 2.795710192206539e-05,
+      "loss": 12.0344,
+      "step": 13095
+    },
+    {
+      "epoch": 2.290142076502732,
+      "grad_norm": 0.8504719303704301,
+      "learning_rate": 2.79440087611206e-05,
+      "loss": 11.9335,
+      "step": 13096
+    },
+    {
+      "epoch": 2.2903169398907104,
+      "grad_norm": 0.8832831174623211,
+      "learning_rate": 2.7930918168887744e-05,
+      "loss": 11.9467,
+      "step": 13097
+    },
+    {
+      "epoch": 2.2904918032786883,
+      "grad_norm": 0.918986790853134,
+      "learning_rate": 2.7917830145833413e-05,
+      "loss": 11.877,
+      "step": 13098
+    },
+    {
+      "epoch": 2.2906666666666666,
+      "grad_norm": 0.7737492342170744,
+      "learning_rate": 2.7904744692424224e-05,
+      "loss": 12.1302,
+      "step": 13099
+    },
+    {
+      "epoch": 2.290841530054645,
+      "grad_norm": 0.7817712488603134,
+      "learning_rate": 2.7891661809126647e-05,
+      "loss": 12.0605,
+      "step": 13100
+    },
+    {
+      "epoch": 2.291016393442623,
+      "grad_norm": 0.9001565031759268,
+      "learning_rate": 2.7878581496407107e-05,
+      "loss": 12.0459,
+      "step": 13101
+    },
+    {
+      "epoch": 2.291191256830601,
+      "grad_norm": 0.8017079154644617,
+      "learning_rate": 2.7865503754731846e-05,
+      "loss": 12.2355,
+      "step": 13102
+    },
+    {
+      "epoch": 2.291366120218579,
+      "grad_norm": 0.8227510094812877,
+      "learning_rate": 2.7852428584567105e-05,
+      "loss": 12.042,
+      "step": 13103
+    },
+    {
+      "epoch": 2.2915409836065574,
+      "grad_norm": 0.7866209585881019,
+      "learning_rate": 2.7839355986378978e-05,
+      "loss": 12.2004,
+      "step": 13104
+    },
+    {
+      "epoch": 2.2917158469945353,
+      "grad_norm": 0.7891743906710016,
+      "learning_rate": 2.7826285960633537e-05,
+      "loss": 12.0686,
+      "step": 13105
+    },
+    {
+      "epoch": 2.2918907103825137,
+      "grad_norm": 0.8081552980049712,
+      "learning_rate": 2.7813218507796634e-05,
+      "loss": 11.9019,
+      "step": 13106
+    },
+    {
+      "epoch": 2.2920655737704916,
+      "grad_norm": 0.819446749444548,
+      "learning_rate": 2.7800153628334156e-05,
+      "loss": 12.1142,
+      "step": 13107
+    },
+    {
+      "epoch": 2.29224043715847,
+      "grad_norm": 0.8586084851039963,
+      "learning_rate": 2.7787091322711878e-05,
+      "loss": 12.1882,
+      "step": 13108
+    },
+    {
+      "epoch": 2.2924153005464483,
+      "grad_norm": 0.903289877351853,
+      "learning_rate": 2.777403159139539e-05,
+      "loss": 12.0537,
+      "step": 13109
+    },
+    {
+      "epoch": 2.292590163934426,
+      "grad_norm": 0.848429049181026,
+      "learning_rate": 2.7760974434850286e-05,
+      "loss": 11.7541,
+      "step": 13110
+    },
+    {
+      "epoch": 2.2927650273224045,
+      "grad_norm": 0.792641160924746,
+      "learning_rate": 2.7747919853542048e-05,
+      "loss": 11.9504,
+      "step": 13111
+    },
+    {
+      "epoch": 2.2929398907103824,
+      "grad_norm": 0.8781035179255177,
+      "learning_rate": 2.773486784793604e-05,
+      "loss": 12.1099,
+      "step": 13112
+    },
+    {
+      "epoch": 2.2931147540983607,
+      "grad_norm": 0.8826262782515071,
+      "learning_rate": 2.7721818418497568e-05,
+      "loss": 11.9172,
+      "step": 13113
+    },
+    {
+      "epoch": 2.2932896174863386,
+      "grad_norm": 0.8056640803661386,
+      "learning_rate": 2.7708771565691817e-05,
+      "loss": 11.9617,
+      "step": 13114
+    },
+    {
+      "epoch": 2.293464480874317,
+      "grad_norm": 0.8643506733343319,
+      "learning_rate": 2.7695727289983898e-05,
+      "loss": 12.0998,
+      "step": 13115
+    },
+    {
+      "epoch": 2.293639344262295,
+      "grad_norm": 0.8246285995508863,
+      "learning_rate": 2.768268559183885e-05,
+      "loss": 11.9989,
+      "step": 13116
+    },
+    {
+      "epoch": 2.293814207650273,
+      "grad_norm": 0.8164426140169062,
+      "learning_rate": 2.7669646471721545e-05,
+      "loss": 12.2072,
+      "step": 13117
+    },
+    {
+      "epoch": 2.2939890710382516,
+      "grad_norm": 0.9141807817526164,
+      "learning_rate": 2.7656609930096834e-05,
+      "loss": 12.0406,
+      "step": 13118
+    },
+    {
+      "epoch": 2.2941639344262295,
+      "grad_norm": 0.8797639356719835,
+      "learning_rate": 2.7643575967429447e-05,
+      "loss": 12.0246,
+      "step": 13119
+    },
+    {
+      "epoch": 2.294338797814208,
+      "grad_norm": 0.8711082908014132,
+      "learning_rate": 2.763054458418407e-05,
+      "loss": 12.1374,
+      "step": 13120
+    },
+    {
+      "epoch": 2.2945136612021857,
+      "grad_norm": 0.821648301221871,
+      "learning_rate": 2.76175157808252e-05,
+      "loss": 12.0822,
+      "step": 13121
+    },
+    {
+      "epoch": 2.294688524590164,
+      "grad_norm": 0.8291093884907937,
+      "learning_rate": 2.7604489557817315e-05,
+      "loss": 12.0797,
+      "step": 13122
+    },
+    {
+      "epoch": 2.294863387978142,
+      "grad_norm": 0.8844963970961692,
+      "learning_rate": 2.7591465915624824e-05,
+      "loss": 12.0335,
+      "step": 13123
+    },
+    {
+      "epoch": 2.2950382513661203,
+      "grad_norm": 1.000996975867511,
+      "learning_rate": 2.757844485471195e-05,
+      "loss": 11.7383,
+      "step": 13124
+    },
+    {
+      "epoch": 2.295213114754098,
+      "grad_norm": 0.8772937856053556,
+      "learning_rate": 2.7565426375542892e-05,
+      "loss": 11.8461,
+      "step": 13125
+    },
+    {
+      "epoch": 2.2953879781420765,
+      "grad_norm": 0.8167728379324596,
+      "learning_rate": 2.7552410478581758e-05,
+      "loss": 12.1798,
+      "step": 13126
+    },
+    {
+      "epoch": 2.295562841530055,
+      "grad_norm": 0.9306764350134378,
+      "learning_rate": 2.7539397164292568e-05,
+      "loss": 11.6974,
+      "step": 13127
+    },
+    {
+      "epoch": 2.2957377049180328,
+      "grad_norm": 0.9847101100507433,
+      "learning_rate": 2.7526386433139184e-05,
+      "loss": 12.2192,
+      "step": 13128
+    },
+    {
+      "epoch": 2.295912568306011,
+      "grad_norm": 0.8629982983386881,
+      "learning_rate": 2.7513378285585445e-05,
+      "loss": 12.1258,
+      "step": 13129
+    },
+    {
+      "epoch": 2.296087431693989,
+      "grad_norm": 0.9535389344307,
+      "learning_rate": 2.750037272209508e-05,
+      "loss": 12.0739,
+      "step": 13130
+    },
+    {
+      "epoch": 2.2962622950819673,
+      "grad_norm": 0.8830514312351999,
+      "learning_rate": 2.7487369743131718e-05,
+      "loss": 12.046,
+      "step": 13131
+    },
+    {
+      "epoch": 2.2964371584699452,
+      "grad_norm": 0.8079282248746688,
+      "learning_rate": 2.74743693491589e-05,
+      "loss": 11.8581,
+      "step": 13132
+    },
+    {
+      "epoch": 2.2966120218579236,
+      "grad_norm": 0.849095315481795,
+      "learning_rate": 2.746137154064009e-05,
+      "loss": 11.9327,
+      "step": 13133
+    },
+    {
+      "epoch": 2.2967868852459015,
+      "grad_norm": 0.8223314750224693,
+      "learning_rate": 2.7448376318038616e-05,
+      "loss": 12.06,
+      "step": 13134
+    },
+    {
+      "epoch": 2.29696174863388,
+      "grad_norm": 0.8839463965211576,
+      "learning_rate": 2.7435383681817807e-05,
+      "loss": 12.1482,
+      "step": 13135
+    },
+    {
+      "epoch": 2.297136612021858,
+      "grad_norm": 0.8924490817033977,
+      "learning_rate": 2.7422393632440746e-05,
+      "loss": 11.829,
+      "step": 13136
+    },
+    {
+      "epoch": 2.297311475409836,
+      "grad_norm": 0.7607133731983738,
+      "learning_rate": 2.7409406170370556e-05,
+      "loss": 12.2371,
+      "step": 13137
+    },
+    {
+      "epoch": 2.2974863387978144,
+      "grad_norm": 0.77241665070177,
+      "learning_rate": 2.7396421296070217e-05,
+      "loss": 11.7896,
+      "step": 13138
+    },
+    {
+      "epoch": 2.2976612021857923,
+      "grad_norm": 0.8320237487860381,
+      "learning_rate": 2.7383439010002666e-05,
+      "loss": 12.0193,
+      "step": 13139
+    },
+    {
+      "epoch": 2.2978360655737706,
+      "grad_norm": 0.8979645410545836,
+      "learning_rate": 2.7370459312630648e-05,
+      "loss": 11.8035,
+      "step": 13140
+    },
+    {
+      "epoch": 2.2980109289617485,
+      "grad_norm": 0.9305183912446936,
+      "learning_rate": 2.7357482204416885e-05,
+      "loss": 11.922,
+      "step": 13141
+    },
+    {
+      "epoch": 2.298185792349727,
+      "grad_norm": 0.8706536126388273,
+      "learning_rate": 2.734450768582405e-05,
+      "loss": 12.0864,
+      "step": 13142
+    },
+    {
+      "epoch": 2.2983606557377048,
+      "grad_norm": 0.9952746589253939,
+      "learning_rate": 2.7331535757314587e-05,
+      "loss": 12.1415,
+      "step": 13143
+    },
+    {
+      "epoch": 2.298535519125683,
+      "grad_norm": 0.9291726951729999,
+      "learning_rate": 2.7318566419350976e-05,
+      "loss": 12.1214,
+      "step": 13144
+    },
+    {
+      "epoch": 2.2987103825136614,
+      "grad_norm": 0.8402606355950363,
+      "learning_rate": 2.7305599672395542e-05,
+      "loss": 12.2227,
+      "step": 13145
+    },
+    {
+      "epoch": 2.2988852459016393,
+      "grad_norm": 0.8601298750634483,
+      "learning_rate": 2.7292635516910593e-05,
+      "loss": 11.8751,
+      "step": 13146
+    },
+    {
+      "epoch": 2.2990601092896172,
+      "grad_norm": 1.0490812756853822,
+      "learning_rate": 2.7279673953358196e-05,
+      "loss": 12.2003,
+      "step": 13147
+    },
+    {
+      "epoch": 2.2992349726775956,
+      "grad_norm": 0.9042509275773537,
+      "learning_rate": 2.726671498220046e-05,
+      "loss": 12.0296,
+      "step": 13148
+    },
+    {
+      "epoch": 2.299409836065574,
+      "grad_norm": 0.9407780787660709,
+      "learning_rate": 2.7253758603899358e-05,
+      "loss": 12.0439,
+      "step": 13149
+    },
+    {
+      "epoch": 2.299584699453552,
+      "grad_norm": 0.9035777149144066,
+      "learning_rate": 2.7240804818916765e-05,
+      "loss": 12.0942,
+      "step": 13150
+    },
+    {
+      "epoch": 2.29975956284153,
+      "grad_norm": 0.947664864997244,
+      "learning_rate": 2.7227853627714473e-05,
+      "loss": 12.0608,
+      "step": 13151
+    },
+    {
+      "epoch": 2.299934426229508,
+      "grad_norm": 0.9041956963565335,
+      "learning_rate": 2.7214905030754167e-05,
+      "loss": 12.0011,
+      "step": 13152
+    },
+    {
+      "epoch": 2.3001092896174864,
+      "grad_norm": 0.7864587573617998,
+      "learning_rate": 2.720195902849746e-05,
+      "loss": 12.0999,
+      "step": 13153
+    },
+    {
+      "epoch": 2.3002841530054643,
+      "grad_norm": 0.8438536043539836,
+      "learning_rate": 2.718901562140589e-05,
+      "loss": 12.0961,
+      "step": 13154
+    },
+    {
+      "epoch": 2.3004590163934426,
+      "grad_norm": 0.9369841555682275,
+      "learning_rate": 2.717607480994081e-05,
+      "loss": 11.8879,
+      "step": 13155
+    },
+    {
+      "epoch": 2.3006338797814205,
+      "grad_norm": 0.9390157353307376,
+      "learning_rate": 2.7163136594563575e-05,
+      "loss": 12.2289,
+      "step": 13156
+    },
+    {
+      "epoch": 2.300808743169399,
+      "grad_norm": 0.8679548694772461,
+      "learning_rate": 2.7150200975735463e-05,
+      "loss": 12.0621,
+      "step": 13157
+    },
+    {
+      "epoch": 2.300983606557377,
+      "grad_norm": 0.9059466762327654,
+      "learning_rate": 2.7137267953917533e-05,
+      "loss": 12.0643,
+      "step": 13158
+    },
+    {
+      "epoch": 2.301158469945355,
+      "grad_norm": 0.8622429314940718,
+      "learning_rate": 2.7124337529570865e-05,
+      "loss": 11.9543,
+      "step": 13159
+    },
+    {
+      "epoch": 2.3013333333333335,
+      "grad_norm": 0.9021991443379047,
+      "learning_rate": 2.711140970315642e-05,
+      "loss": 11.8784,
+      "step": 13160
+    },
+    {
+      "epoch": 2.3015081967213114,
+      "grad_norm": 0.8757632029079265,
+      "learning_rate": 2.709848447513509e-05,
+      "loss": 12.0616,
+      "step": 13161
+    },
+    {
+      "epoch": 2.3016830601092897,
+      "grad_norm": 0.8236695589351978,
+      "learning_rate": 2.7085561845967577e-05,
+      "loss": 12.0823,
+      "step": 13162
+    },
+    {
+      "epoch": 2.3018579234972676,
+      "grad_norm": 1.022429981015836,
+      "learning_rate": 2.7072641816114596e-05,
+      "loss": 12.0559,
+      "step": 13163
+    },
+    {
+      "epoch": 2.302032786885246,
+      "grad_norm": 0.8766922819276097,
+      "learning_rate": 2.7059724386036733e-05,
+      "loss": 12.1047,
+      "step": 13164
+    },
+    {
+      "epoch": 2.302207650273224,
+      "grad_norm": 0.8011221298161653,
+      "learning_rate": 2.7046809556194462e-05,
+      "loss": 12.133,
+      "step": 13165
+    },
+    {
+      "epoch": 2.302382513661202,
+      "grad_norm": 0.8720745962629096,
+      "learning_rate": 2.7033897327048207e-05,
+      "loss": 12.0425,
+      "step": 13166
+    },
+    {
+      "epoch": 2.3025573770491805,
+      "grad_norm": 0.9478616095356814,
+      "learning_rate": 2.7020987699058252e-05,
+      "loss": 11.8317,
+      "step": 13167
+    },
+    {
+      "epoch": 2.3027322404371584,
+      "grad_norm": 0.8197345874074919,
+      "learning_rate": 2.700808067268482e-05,
+      "loss": 12.0851,
+      "step": 13168
+    },
+    {
+      "epoch": 2.3029071038251367,
+      "grad_norm": 0.8398087498766615,
+      "learning_rate": 2.6995176248388064e-05,
+      "loss": 11.9742,
+      "step": 13169
+    },
+    {
+      "epoch": 2.3030819672131146,
+      "grad_norm": 0.8477862444500667,
+      "learning_rate": 2.698227442662795e-05,
+      "loss": 11.9173,
+      "step": 13170
+    },
+    {
+      "epoch": 2.303256830601093,
+      "grad_norm": 0.8034077281731352,
+      "learning_rate": 2.6969375207864434e-05,
+      "loss": 12.1267,
+      "step": 13171
+    },
+    {
+      "epoch": 2.303431693989071,
+      "grad_norm": 0.7826891935634578,
+      "learning_rate": 2.69564785925574e-05,
+      "loss": 12.2283,
+      "step": 13172
+    },
+    {
+      "epoch": 2.3036065573770492,
+      "grad_norm": 0.8122616250558737,
+      "learning_rate": 2.6943584581166524e-05,
+      "loss": 11.8698,
+      "step": 13173
+    },
+    {
+      "epoch": 2.303781420765027,
+      "grad_norm": 1.0043275660665498,
+      "learning_rate": 2.6930693174151513e-05,
+      "loss": 11.9215,
+      "step": 13174
+    },
+    {
+      "epoch": 2.3039562841530055,
+      "grad_norm": 0.8456401669898426,
+      "learning_rate": 2.69178043719719e-05,
+      "loss": 12.0534,
+      "step": 13175
+    },
+    {
+      "epoch": 2.304131147540984,
+      "grad_norm": 0.8707486473879569,
+      "learning_rate": 2.6904918175087212e-05,
+      "loss": 12.1277,
+      "step": 13176
+    },
+    {
+      "epoch": 2.3043060109289617,
+      "grad_norm": 0.8534897921267324,
+      "learning_rate": 2.689203458395676e-05,
+      "loss": 11.9356,
+      "step": 13177
+    },
+    {
+      "epoch": 2.30448087431694,
+      "grad_norm": 0.9318103063217791,
+      "learning_rate": 2.6879153599039864e-05,
+      "loss": 11.8877,
+      "step": 13178
+    },
+    {
+      "epoch": 2.304655737704918,
+      "grad_norm": 0.8761985192031521,
+      "learning_rate": 2.68662752207957e-05,
+      "loss": 12.1087,
+      "step": 13179
+    },
+    {
+      "epoch": 2.3048306010928963,
+      "grad_norm": 0.8523194294419338,
+      "learning_rate": 2.6853399449683415e-05,
+      "loss": 12.1771,
+      "step": 13180
+    },
+    {
+      "epoch": 2.305005464480874,
+      "grad_norm": 0.7851507616989404,
+      "learning_rate": 2.684052628616194e-05,
+      "loss": 11.778,
+      "step": 13181
+    },
+    {
+      "epoch": 2.3051803278688525,
+      "grad_norm": 0.870846187381708,
+      "learning_rate": 2.6827655730690225e-05,
+      "loss": 11.8286,
+      "step": 13182
+    },
+    {
+      "epoch": 2.3053551912568304,
+      "grad_norm": 0.8164177561411872,
+      "learning_rate": 2.68147877837271e-05,
+      "loss": 11.9026,
+      "step": 13183
+    },
+    {
+      "epoch": 2.3055300546448088,
+      "grad_norm": 0.8436457890282143,
+      "learning_rate": 2.6801922445731277e-05,
+      "loss": 12.1938,
+      "step": 13184
+    },
+    {
+      "epoch": 2.305704918032787,
+      "grad_norm": 0.8196689782565912,
+      "learning_rate": 2.6789059717161392e-05,
+      "loss": 12.0563,
+      "step": 13185
+    },
+    {
+      "epoch": 2.305879781420765,
+      "grad_norm": 0.831870596412315,
+      "learning_rate": 2.6776199598475994e-05,
+      "loss": 11.9147,
+      "step": 13186
+    },
+    {
+      "epoch": 2.3060546448087433,
+      "grad_norm": 0.844824905352645,
+      "learning_rate": 2.676334209013356e-05,
+      "loss": 12.051,
+      "step": 13187
+    },
+    {
+      "epoch": 2.3062295081967212,
+      "grad_norm": 0.8165364160966496,
+      "learning_rate": 2.6750487192592377e-05,
+      "loss": 12.1737,
+      "step": 13188
+    },
+    {
+      "epoch": 2.3064043715846996,
+      "grad_norm": 0.8997467272792194,
+      "learning_rate": 2.6737634906310748e-05,
+      "loss": 12.2515,
+      "step": 13189
+    },
+    {
+      "epoch": 2.3065792349726775,
+      "grad_norm": 0.8372496690786155,
+      "learning_rate": 2.672478523174683e-05,
+      "loss": 12.1363,
+      "step": 13190
+    },
+    {
+      "epoch": 2.306754098360656,
+      "grad_norm": 0.8000222154196955,
+      "learning_rate": 2.671193816935874e-05,
+      "loss": 12.0152,
+      "step": 13191
+    },
+    {
+      "epoch": 2.3069289617486337,
+      "grad_norm": 0.8824176689897008,
+      "learning_rate": 2.6699093719604397e-05,
+      "loss": 12.1526,
+      "step": 13192
+    },
+    {
+      "epoch": 2.307103825136612,
+      "grad_norm": 0.8441656434759962,
+      "learning_rate": 2.6686251882941715e-05,
+      "loss": 12.2028,
+      "step": 13193
+    },
+    {
+      "epoch": 2.3072786885245904,
+      "grad_norm": 0.8767478200422885,
+      "learning_rate": 2.6673412659828502e-05,
+      "loss": 12.0116,
+      "step": 13194
+    },
+    {
+      "epoch": 2.3074535519125683,
+      "grad_norm": 0.7587849401396048,
+      "learning_rate": 2.6660576050722486e-05,
+      "loss": 11.7693,
+      "step": 13195
+    },
+    {
+      "epoch": 2.3076284153005466,
+      "grad_norm": 0.885634857110189,
+      "learning_rate": 2.6647742056081205e-05,
+      "loss": 11.9615,
+      "step": 13196
+    },
+    {
+      "epoch": 2.3078032786885245,
+      "grad_norm": 0.8841490819547817,
+      "learning_rate": 2.6634910676362224e-05,
+      "loss": 12.0227,
+      "step": 13197
+    },
+    {
+      "epoch": 2.307978142076503,
+      "grad_norm": 0.9071727058360819,
+      "learning_rate": 2.6622081912022955e-05,
+      "loss": 12.1204,
+      "step": 13198
+    },
+    {
+      "epoch": 2.3081530054644808,
+      "grad_norm": 0.9006953607971995,
+      "learning_rate": 2.6609255763520767e-05,
+      "loss": 12.0305,
+      "step": 13199
+    },
+    {
+      "epoch": 2.308327868852459,
+      "grad_norm": 0.8220158834293284,
+      "learning_rate": 2.659643223131284e-05,
+      "loss": 12.3105,
+      "step": 13200
+    },
+    {
+      "epoch": 2.308502732240437,
+      "grad_norm": 0.8761544852702527,
+      "learning_rate": 2.6583611315856337e-05,
+      "loss": 12.1822,
+      "step": 13201
+    },
+    {
+      "epoch": 2.3086775956284153,
+      "grad_norm": 0.8840772289645761,
+      "learning_rate": 2.6570793017608308e-05,
+      "loss": 12.0828,
+      "step": 13202
+    },
+    {
+      "epoch": 2.3088524590163937,
+      "grad_norm": 0.843449480226199,
+      "learning_rate": 2.655797733702572e-05,
+      "loss": 12.0775,
+      "step": 13203
+    },
+    {
+      "epoch": 2.3090273224043716,
+      "grad_norm": 0.9005935770347975,
+      "learning_rate": 2.6545164274565428e-05,
+      "loss": 11.9597,
+      "step": 13204
+    },
+    {
+      "epoch": 2.3092021857923495,
+      "grad_norm": 0.8425617396995052,
+      "learning_rate": 2.653235383068421e-05,
+      "loss": 12.1109,
+      "step": 13205
+    },
+    {
+      "epoch": 2.309377049180328,
+      "grad_norm": 0.8521728482986827,
+      "learning_rate": 2.651954600583877e-05,
+      "loss": 12.1085,
+      "step": 13206
+    },
+    {
+      "epoch": 2.309551912568306,
+      "grad_norm": 0.7955894415211772,
+      "learning_rate": 2.650674080048563e-05,
+      "loss": 12.0021,
+      "step": 13207
+    },
+    {
+      "epoch": 2.309726775956284,
+      "grad_norm": 0.8429844819848661,
+      "learning_rate": 2.6493938215081304e-05,
+      "loss": 12.0752,
+      "step": 13208
+    },
+    {
+      "epoch": 2.3099016393442624,
+      "grad_norm": 0.8114444121349333,
+      "learning_rate": 2.6481138250082216e-05,
+      "loss": 11.9357,
+      "step": 13209
+    },
+    {
+      "epoch": 2.3100765027322403,
+      "grad_norm": 0.7709697431225541,
+      "learning_rate": 2.6468340905944666e-05,
+      "loss": 11.6805,
+      "step": 13210
+    },
+    {
+      "epoch": 2.3102513661202186,
+      "grad_norm": 0.8615397442857735,
+      "learning_rate": 2.6455546183124834e-05,
+      "loss": 12.2049,
+      "step": 13211
+    },
+    {
+      "epoch": 2.3104262295081965,
+      "grad_norm": 0.9002729867771196,
+      "learning_rate": 2.6442754082078846e-05,
+      "loss": 12.0461,
+      "step": 13212
+    },
+    {
+      "epoch": 2.310601092896175,
+      "grad_norm": 0.912197566748279,
+      "learning_rate": 2.6429964603262734e-05,
+      "loss": 12.1518,
+      "step": 13213
+    },
+    {
+      "epoch": 2.3107759562841528,
+      "grad_norm": 0.7984214792853227,
+      "learning_rate": 2.641717774713245e-05,
+      "loss": 12.0492,
+      "step": 13214
+    },
+    {
+      "epoch": 2.310950819672131,
+      "grad_norm": 0.9012314888629829,
+      "learning_rate": 2.640439351414379e-05,
+      "loss": 11.8369,
+      "step": 13215
+    },
+    {
+      "epoch": 2.3111256830601095,
+      "grad_norm": 0.9710941599932936,
+      "learning_rate": 2.63916119047525e-05,
+      "loss": 12.1117,
+      "step": 13216
+    },
+    {
+      "epoch": 2.3113005464480874,
+      "grad_norm": 0.8274238631482413,
+      "learning_rate": 2.6378832919414287e-05,
+      "loss": 11.7376,
+      "step": 13217
+    },
+    {
+      "epoch": 2.3114754098360657,
+      "grad_norm": 0.781890461414457,
+      "learning_rate": 2.6366056558584616e-05,
+      "loss": 12.0848,
+      "step": 13218
+    },
+    {
+      "epoch": 2.3116502732240436,
+      "grad_norm": 0.8072004039251133,
+      "learning_rate": 2.6353282822719017e-05,
+      "loss": 12.1295,
+      "step": 13219
+    },
+    {
+      "epoch": 2.311825136612022,
+      "grad_norm": 0.8258501264605221,
+      "learning_rate": 2.634051171227282e-05,
+      "loss": 12.011,
+      "step": 13220
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 0.8748330108045398,
+      "learning_rate": 2.6327743227701328e-05,
+      "loss": 11.9136,
+      "step": 13221
+    },
+    {
+      "epoch": 2.312174863387978,
+      "grad_norm": 0.822740298567402,
+      "learning_rate": 2.6314977369459714e-05,
+      "loss": 12.0371,
+      "step": 13222
+    },
+    {
+      "epoch": 2.312349726775956,
+      "grad_norm": 0.8055443653644325,
+      "learning_rate": 2.630221413800307e-05,
+      "loss": 12.1446,
+      "step": 13223
+    },
+    {
+      "epoch": 2.3125245901639344,
+      "grad_norm": 0.8293858415365772,
+      "learning_rate": 2.6289453533786378e-05,
+      "loss": 12.1342,
+      "step": 13224
+    },
+    {
+      "epoch": 2.3126994535519128,
+      "grad_norm": 0.8005540740473375,
+      "learning_rate": 2.6276695557264573e-05,
+      "loss": 12.1039,
+      "step": 13225
+    },
+    {
+      "epoch": 2.3128743169398907,
+      "grad_norm": 0.8444978800836704,
+      "learning_rate": 2.6263940208892413e-05,
+      "loss": 12.1597,
+      "step": 13226
+    },
+    {
+      "epoch": 2.313049180327869,
+      "grad_norm": 0.9042616417581115,
+      "learning_rate": 2.625118748912463e-05,
+      "loss": 12.0364,
+      "step": 13227
+    },
+    {
+      "epoch": 2.313224043715847,
+      "grad_norm": 0.8229218583698127,
+      "learning_rate": 2.623843739841584e-05,
+      "loss": 12.1541,
+      "step": 13228
+    },
+    {
+      "epoch": 2.3133989071038252,
+      "grad_norm": 0.7623585081705954,
+      "learning_rate": 2.622568993722061e-05,
+      "loss": 11.7607,
+      "step": 13229
+    },
+    {
+      "epoch": 2.313573770491803,
+      "grad_norm": 1.0113714792934738,
+      "learning_rate": 2.6212945105993313e-05,
+      "loss": 11.9646,
+      "step": 13230
+    },
+    {
+      "epoch": 2.3137486338797815,
+      "grad_norm": 0.9386608010685316,
+      "learning_rate": 2.6200202905188307e-05,
+      "loss": 12.1405,
+      "step": 13231
+    },
+    {
+      "epoch": 2.3139234972677594,
+      "grad_norm": 0.8755290238988808,
+      "learning_rate": 2.6187463335259877e-05,
+      "loss": 12.0188,
+      "step": 13232
+    },
+    {
+      "epoch": 2.3140983606557377,
+      "grad_norm": 0.8985622759100119,
+      "learning_rate": 2.6174726396662097e-05,
+      "loss": 11.7738,
+      "step": 13233
+    },
+    {
+      "epoch": 2.314273224043716,
+      "grad_norm": 0.83025771874578,
+      "learning_rate": 2.6161992089849073e-05,
+      "loss": 11.7704,
+      "step": 13234
+    },
+    {
+      "epoch": 2.314448087431694,
+      "grad_norm": 0.8425407207282207,
+      "learning_rate": 2.6149260415274756e-05,
+      "loss": 12.3467,
+      "step": 13235
+    },
+    {
+      "epoch": 2.3146229508196723,
+      "grad_norm": 0.8951112566043298,
+      "learning_rate": 2.613653137339306e-05,
+      "loss": 12.1305,
+      "step": 13236
+    },
+    {
+      "epoch": 2.31479781420765,
+      "grad_norm": 0.8321445676498314,
+      "learning_rate": 2.612380496465768e-05,
+      "loss": 11.915,
+      "step": 13237
+    },
+    {
+      "epoch": 2.3149726775956285,
+      "grad_norm": 0.8716457173577508,
+      "learning_rate": 2.6111081189522347e-05,
+      "loss": 11.999,
+      "step": 13238
+    },
+    {
+      "epoch": 2.3151475409836064,
+      "grad_norm": 0.91301349426775,
+      "learning_rate": 2.609836004844063e-05,
+      "loss": 12.0519,
+      "step": 13239
+    },
+    {
+      "epoch": 2.3153224043715848,
+      "grad_norm": 0.8289749283996211,
+      "learning_rate": 2.6085641541866037e-05,
+      "loss": 12.0688,
+      "step": 13240
+    },
+    {
+      "epoch": 2.3154972677595627,
+      "grad_norm": 0.8334558260863228,
+      "learning_rate": 2.6072925670251968e-05,
+      "loss": 12.0428,
+      "step": 13241
+    },
+    {
+      "epoch": 2.315672131147541,
+      "grad_norm": 0.8693670853843293,
+      "learning_rate": 2.606021243405172e-05,
+      "loss": 12.0785,
+      "step": 13242
+    },
+    {
+      "epoch": 2.3158469945355193,
+      "grad_norm": 0.933221642039754,
+      "learning_rate": 2.604750183371851e-05,
+      "loss": 12.0682,
+      "step": 13243
+    },
+    {
+      "epoch": 2.3160218579234972,
+      "grad_norm": 0.8135385670820428,
+      "learning_rate": 2.6034793869705486e-05,
+      "loss": 12.1254,
+      "step": 13244
+    },
+    {
+      "epoch": 2.3161967213114756,
+      "grad_norm": 0.9041425906443503,
+      "learning_rate": 2.6022088542465605e-05,
+      "loss": 12.0244,
+      "step": 13245
+    },
+    {
+      "epoch": 2.3163715846994535,
+      "grad_norm": 0.9437314967387811,
+      "learning_rate": 2.6009385852451838e-05,
+      "loss": 11.9448,
+      "step": 13246
+    },
+    {
+      "epoch": 2.316546448087432,
+      "grad_norm": 0.9147875208052166,
+      "learning_rate": 2.5996685800117016e-05,
+      "loss": 12.05,
+      "step": 13247
+    },
+    {
+      "epoch": 2.3167213114754097,
+      "grad_norm": 0.815505209383659,
+      "learning_rate": 2.5983988385913914e-05,
+      "loss": 12.0482,
+      "step": 13248
+    },
+    {
+      "epoch": 2.316896174863388,
+      "grad_norm": 0.880757751877002,
+      "learning_rate": 2.5971293610295123e-05,
+      "loss": 12.1962,
+      "step": 13249
+    },
+    {
+      "epoch": 2.317071038251366,
+      "grad_norm": 0.7756570522228822,
+      "learning_rate": 2.5958601473713207e-05,
+      "loss": 11.9997,
+      "step": 13250
+    },
+    {
+      "epoch": 2.3172459016393443,
+      "grad_norm": 0.9242578242884386,
+      "learning_rate": 2.5945911976620686e-05,
+      "loss": 12.1002,
+      "step": 13251
+    },
+    {
+      "epoch": 2.3174207650273226,
+      "grad_norm": 0.8927618474526131,
+      "learning_rate": 2.593322511946984e-05,
+      "loss": 11.9718,
+      "step": 13252
+    },
+    {
+      "epoch": 2.3175956284153005,
+      "grad_norm": 0.901982644439734,
+      "learning_rate": 2.5920540902712986e-05,
+      "loss": 11.8467,
+      "step": 13253
+    },
+    {
+      "epoch": 2.317770491803279,
+      "grad_norm": 0.8574947935049159,
+      "learning_rate": 2.5907859326802287e-05,
+      "loss": 11.7966,
+      "step": 13254
+    },
+    {
+      "epoch": 2.3179453551912568,
+      "grad_norm": 0.9359164994009554,
+      "learning_rate": 2.5895180392189876e-05,
+      "loss": 12.1236,
+      "step": 13255
+    },
+    {
+      "epoch": 2.318120218579235,
+      "grad_norm": 0.9206039375106346,
+      "learning_rate": 2.5882504099327663e-05,
+      "loss": 11.946,
+      "step": 13256
+    },
+    {
+      "epoch": 2.318295081967213,
+      "grad_norm": 0.8769267237386313,
+      "learning_rate": 2.58698304486676e-05,
+      "loss": 11.9475,
+      "step": 13257
+    },
+    {
+      "epoch": 2.3184699453551914,
+      "grad_norm": 0.8806152109543423,
+      "learning_rate": 2.585715944066145e-05,
+      "loss": 11.9996,
+      "step": 13258
+    },
+    {
+      "epoch": 2.3186448087431692,
+      "grad_norm": 0.8314776485472241,
+      "learning_rate": 2.5844491075760946e-05,
+      "loss": 11.8502,
+      "step": 13259
+    },
+    {
+      "epoch": 2.3188196721311476,
+      "grad_norm": 0.8175759800910087,
+      "learning_rate": 2.5831825354417695e-05,
+      "loss": 12.1534,
+      "step": 13260
+    },
+    {
+      "epoch": 2.318994535519126,
+      "grad_norm": 0.855592731864825,
+      "learning_rate": 2.5819162277083208e-05,
+      "loss": 12.0842,
+      "step": 13261
+    },
+    {
+      "epoch": 2.319169398907104,
+      "grad_norm": 0.8583533414843691,
+      "learning_rate": 2.580650184420892e-05,
+      "loss": 12.0626,
+      "step": 13262
+    },
+    {
+      "epoch": 2.3193442622950817,
+      "grad_norm": 0.83754510143883,
+      "learning_rate": 2.5793844056246187e-05,
+      "loss": 12.1068,
+      "step": 13263
+    },
+    {
+      "epoch": 2.31951912568306,
+      "grad_norm": 0.8250546023310049,
+      "learning_rate": 2.578118891364618e-05,
+      "loss": 11.8424,
+      "step": 13264
+    },
+    {
+      "epoch": 2.3196939890710384,
+      "grad_norm": 0.8755350575408429,
+      "learning_rate": 2.576853641686008e-05,
+      "loss": 12.012,
+      "step": 13265
+    },
+    {
+      "epoch": 2.3198688524590163,
+      "grad_norm": 0.8026436918847438,
+      "learning_rate": 2.5755886566338948e-05,
+      "loss": 12.067,
+      "step": 13266
+    },
+    {
+      "epoch": 2.3200437158469946,
+      "grad_norm": 0.8106602649320158,
+      "learning_rate": 2.5743239362533687e-05,
+      "loss": 12.2601,
+      "step": 13267
+    },
+    {
+      "epoch": 2.3202185792349725,
+      "grad_norm": 1.0388341514041965,
+      "learning_rate": 2.573059480589518e-05,
+      "loss": 12.0531,
+      "step": 13268
+    },
+    {
+      "epoch": 2.320393442622951,
+      "grad_norm": 0.8232179645518654,
+      "learning_rate": 2.5717952896874198e-05,
+      "loss": 12.0866,
+      "step": 13269
+    },
+    {
+      "epoch": 2.320568306010929,
+      "grad_norm": 0.8173574018073321,
+      "learning_rate": 2.5705313635921435e-05,
+      "loss": 11.9725,
+      "step": 13270
+    },
+    {
+      "epoch": 2.320743169398907,
+      "grad_norm": 0.8559042642402218,
+      "learning_rate": 2.569267702348741e-05,
+      "loss": 12.1561,
+      "step": 13271
+    },
+    {
+      "epoch": 2.320918032786885,
+      "grad_norm": 0.872763123087168,
+      "learning_rate": 2.568004306002263e-05,
+      "loss": 12.1657,
+      "step": 13272
+    },
+    {
+      "epoch": 2.3210928961748634,
+      "grad_norm": 0.8320361886420583,
+      "learning_rate": 2.5667411745977476e-05,
+      "loss": 12.0991,
+      "step": 13273
+    },
+    {
+      "epoch": 2.3212677595628417,
+      "grad_norm": 0.8748792611733361,
+      "learning_rate": 2.5654783081802248e-05,
+      "loss": 12.1939,
+      "step": 13274
+    },
+    {
+      "epoch": 2.3214426229508196,
+      "grad_norm": 0.855975966637078,
+      "learning_rate": 2.5642157067947135e-05,
+      "loss": 11.9459,
+      "step": 13275
+    },
+    {
+      "epoch": 2.321617486338798,
+      "grad_norm": 0.9382646887060632,
+      "learning_rate": 2.562953370486225e-05,
+      "loss": 12.1216,
+      "step": 13276
+    },
+    {
+      "epoch": 2.321792349726776,
+      "grad_norm": 0.8381964974633748,
+      "learning_rate": 2.56169129929976e-05,
+      "loss": 12.1033,
+      "step": 13277
+    },
+    {
+      "epoch": 2.321967213114754,
+      "grad_norm": 0.8187538971340198,
+      "learning_rate": 2.560429493280312e-05,
+      "loss": 12.0063,
+      "step": 13278
+    },
+    {
+      "epoch": 2.322142076502732,
+      "grad_norm": 0.8613100174748194,
+      "learning_rate": 2.5591679524728563e-05,
+      "loss": 12.1228,
+      "step": 13279
+    },
+    {
+      "epoch": 2.3223169398907104,
+      "grad_norm": 0.9041758191971293,
+      "learning_rate": 2.5579066769223715e-05,
+      "loss": 12.1138,
+      "step": 13280
+    },
+    {
+      "epoch": 2.3224918032786883,
+      "grad_norm": 0.8556486256398657,
+      "learning_rate": 2.5566456666738203e-05,
+      "loss": 12.0137,
+      "step": 13281
+    },
+    {
+      "epoch": 2.3226666666666667,
+      "grad_norm": 0.8855207697304304,
+      "learning_rate": 2.5553849217721515e-05,
+      "loss": 12.1387,
+      "step": 13282
+    },
+    {
+      "epoch": 2.322841530054645,
+      "grad_norm": 0.9305111217298289,
+      "learning_rate": 2.5541244422623124e-05,
+      "loss": 12.2095,
+      "step": 13283
+    },
+    {
+      "epoch": 2.323016393442623,
+      "grad_norm": 0.9457787717200766,
+      "learning_rate": 2.5528642281892368e-05,
+      "loss": 11.9873,
+      "step": 13284
+    },
+    {
+      "epoch": 2.3231912568306012,
+      "grad_norm": 0.8354892705791233,
+      "learning_rate": 2.5516042795978534e-05,
+      "loss": 12.0375,
+      "step": 13285
+    },
+    {
+      "epoch": 2.323366120218579,
+      "grad_norm": 0.8378917367077342,
+      "learning_rate": 2.550344596533073e-05,
+      "loss": 12.216,
+      "step": 13286
+    },
+    {
+      "epoch": 2.3235409836065575,
+      "grad_norm": 0.8919944477522265,
+      "learning_rate": 2.549085179039803e-05,
+      "loss": 12.0438,
+      "step": 13287
+    },
+    {
+      "epoch": 2.3237158469945354,
+      "grad_norm": 0.8787771702904036,
+      "learning_rate": 2.5478260271629406e-05,
+      "loss": 11.9275,
+      "step": 13288
+    },
+    {
+      "epoch": 2.3238907103825137,
+      "grad_norm": 0.797846170389105,
+      "learning_rate": 2.546567140947377e-05,
+      "loss": 11.9101,
+      "step": 13289
+    },
+    {
+      "epoch": 2.3240655737704916,
+      "grad_norm": 0.8077230166757471,
+      "learning_rate": 2.5453085204379835e-05,
+      "loss": 11.5772,
+      "step": 13290
+    },
+    {
+      "epoch": 2.32424043715847,
+      "grad_norm": 0.8095999895550328,
+      "learning_rate": 2.544050165679631e-05,
+      "loss": 11.9223,
+      "step": 13291
+    },
+    {
+      "epoch": 2.3244153005464483,
+      "grad_norm": 0.9299451089797123,
+      "learning_rate": 2.5427920767171786e-05,
+      "loss": 12.2001,
+      "step": 13292
+    },
+    {
+      "epoch": 2.324590163934426,
+      "grad_norm": 0.9296660876585324,
+      "learning_rate": 2.5415342535954756e-05,
+      "loss": 12.0137,
+      "step": 13293
+    },
+    {
+      "epoch": 2.3247650273224045,
+      "grad_norm": 0.8386492122601451,
+      "learning_rate": 2.5402766963593626e-05,
+      "loss": 11.8676,
+      "step": 13294
+    },
+    {
+      "epoch": 2.3249398907103824,
+      "grad_norm": 0.88809799024852,
+      "learning_rate": 2.539019405053671e-05,
+      "loss": 12.1714,
+      "step": 13295
+    },
+    {
+      "epoch": 2.3251147540983608,
+      "grad_norm": 0.8719944146386638,
+      "learning_rate": 2.537762379723221e-05,
+      "loss": 12.045,
+      "step": 13296
+    },
+    {
+      "epoch": 2.3252896174863387,
+      "grad_norm": 0.9688634117138604,
+      "learning_rate": 2.5365056204128223e-05,
+      "loss": 11.8202,
+      "step": 13297
+    },
+    {
+      "epoch": 2.325464480874317,
+      "grad_norm": 0.8736890763084785,
+      "learning_rate": 2.5352491271672774e-05,
+      "loss": 12.1511,
+      "step": 13298
+    },
+    {
+      "epoch": 2.325639344262295,
+      "grad_norm": 0.8971527732137702,
+      "learning_rate": 2.533992900031379e-05,
+      "loss": 11.8682,
+      "step": 13299
+    },
+    {
+      "epoch": 2.3258142076502732,
+      "grad_norm": 0.9766463294487595,
+      "learning_rate": 2.5327369390499146e-05,
+      "loss": 11.9632,
+      "step": 13300
+    },
+    {
+      "epoch": 2.3259890710382516,
+      "grad_norm": 0.8963575110487996,
+      "learning_rate": 2.5314812442676494e-05,
+      "loss": 11.9194,
+      "step": 13301
+    },
+    {
+      "epoch": 2.3261639344262295,
+      "grad_norm": 0.8739308954490594,
+      "learning_rate": 2.5302258157293536e-05,
+      "loss": 11.9825,
+      "step": 13302
+    },
+    {
+      "epoch": 2.326338797814208,
+      "grad_norm": 0.8507240690185198,
+      "learning_rate": 2.5289706534797784e-05,
+      "loss": 12.0802,
+      "step": 13303
+    },
+    {
+      "epoch": 2.3265136612021857,
+      "grad_norm": 0.8580692355742443,
+      "learning_rate": 2.5277157575636745e-05,
+      "loss": 12.0543,
+      "step": 13304
+    },
+    {
+      "epoch": 2.326688524590164,
+      "grad_norm": 0.8262690680626797,
+      "learning_rate": 2.52646112802577e-05,
+      "loss": 11.9886,
+      "step": 13305
+    },
+    {
+      "epoch": 2.326863387978142,
+      "grad_norm": 0.8887758030806902,
+      "learning_rate": 2.5252067649107936e-05,
+      "loss": 12.2303,
+      "step": 13306
+    },
+    {
+      "epoch": 2.3270382513661203,
+      "grad_norm": 0.9729889482929313,
+      "learning_rate": 2.523952668263464e-05,
+      "loss": 12.1825,
+      "step": 13307
+    },
+    {
+      "epoch": 2.327213114754098,
+      "grad_norm": 0.8310433524779841,
+      "learning_rate": 2.522698838128489e-05,
+      "loss": 12.2668,
+      "step": 13308
+    },
+    {
+      "epoch": 2.3273879781420765,
+      "grad_norm": 0.8702376982550022,
+      "learning_rate": 2.5214452745505612e-05,
+      "loss": 12.0786,
+      "step": 13309
+    },
+    {
+      "epoch": 2.327562841530055,
+      "grad_norm": 0.8383953713632997,
+      "learning_rate": 2.5201919775743722e-05,
+      "loss": 11.7567,
+      "step": 13310
+    },
+    {
+      "epoch": 2.327737704918033,
+      "grad_norm": 0.8794280138592209,
+      "learning_rate": 2.5189389472446002e-05,
+      "loss": 12.1867,
+      "step": 13311
+    },
+    {
+      "epoch": 2.327912568306011,
+      "grad_norm": 0.8223621028804409,
+      "learning_rate": 2.517686183605913e-05,
+      "loss": 12.0724,
+      "step": 13312
+    },
+    {
+      "epoch": 2.328087431693989,
+      "grad_norm": 0.8672157374079908,
+      "learning_rate": 2.5164336867029715e-05,
+      "loss": 12.245,
+      "step": 13313
+    },
+    {
+      "epoch": 2.3282622950819674,
+      "grad_norm": 0.8580409487281765,
+      "learning_rate": 2.5151814565804267e-05,
+      "loss": 12.1532,
+      "step": 13314
+    },
+    {
+      "epoch": 2.3284371584699453,
+      "grad_norm": 0.8472136843046706,
+      "learning_rate": 2.5139294932829194e-05,
+      "loss": 12.1004,
+      "step": 13315
+    },
+    {
+      "epoch": 2.3286120218579236,
+      "grad_norm": 0.7955229973051477,
+      "learning_rate": 2.5126777968550762e-05,
+      "loss": 12.0139,
+      "step": 13316
+    },
+    {
+      "epoch": 2.3287868852459015,
+      "grad_norm": 0.9596886725192404,
+      "learning_rate": 2.511426367341523e-05,
+      "loss": 12.3213,
+      "step": 13317
+    },
+    {
+      "epoch": 2.32896174863388,
+      "grad_norm": 0.8885786880309365,
+      "learning_rate": 2.5101752047868688e-05,
+      "loss": 11.9245,
+      "step": 13318
+    },
+    {
+      "epoch": 2.3291366120218577,
+      "grad_norm": 0.8185560058612333,
+      "learning_rate": 2.5089243092357208e-05,
+      "loss": 11.9899,
+      "step": 13319
+    },
+    {
+      "epoch": 2.329311475409836,
+      "grad_norm": 0.816421629322506,
+      "learning_rate": 2.507673680732666e-05,
+      "loss": 11.9847,
+      "step": 13320
+    },
+    {
+      "epoch": 2.329486338797814,
+      "grad_norm": 0.8689427744697301,
+      "learning_rate": 2.506423319322292e-05,
+      "loss": 11.8026,
+      "step": 13321
+    },
+    {
+      "epoch": 2.3296612021857923,
+      "grad_norm": 0.8729853203359426,
+      "learning_rate": 2.50517322504917e-05,
+      "loss": 12.177,
+      "step": 13322
+    },
+    {
+      "epoch": 2.3298360655737707,
+      "grad_norm": 0.8701286266862287,
+      "learning_rate": 2.503923397957868e-05,
+      "loss": 11.9017,
+      "step": 13323
+    },
+    {
+      "epoch": 2.3300109289617486,
+      "grad_norm": 0.8626603773035303,
+      "learning_rate": 2.502673838092937e-05,
+      "loss": 12.1216,
+      "step": 13324
+    },
+    {
+      "epoch": 2.330185792349727,
+      "grad_norm": 0.9302126113280452,
+      "learning_rate": 2.5014245454989238e-05,
+      "loss": 12.2712,
+      "step": 13325
+    },
+    {
+      "epoch": 2.330360655737705,
+      "grad_norm": 0.9561064821428031,
+      "learning_rate": 2.500175520220367e-05,
+      "loss": 12.0238,
+      "step": 13326
+    },
+    {
+      "epoch": 2.330535519125683,
+      "grad_norm": 0.8574473375335441,
+      "learning_rate": 2.4989267623017877e-05,
+      "loss": 11.9663,
+      "step": 13327
+    },
+    {
+      "epoch": 2.330710382513661,
+      "grad_norm": 0.8903035352839141,
+      "learning_rate": 2.4976782717877047e-05,
+      "loss": 11.8265,
+      "step": 13328
+    },
+    {
+      "epoch": 2.3308852459016394,
+      "grad_norm": 0.8417626814908084,
+      "learning_rate": 2.4964300487226268e-05,
+      "loss": 12.0026,
+      "step": 13329
+    },
+    {
+      "epoch": 2.3310601092896173,
+      "grad_norm": 0.9059993175158563,
+      "learning_rate": 2.4951820931510496e-05,
+      "loss": 12.0382,
+      "step": 13330
+    },
+    {
+      "epoch": 2.3312349726775956,
+      "grad_norm": 0.824035134083214,
+      "learning_rate": 2.4939344051174628e-05,
+      "loss": 12.2543,
+      "step": 13331
+    },
+    {
+      "epoch": 2.331409836065574,
+      "grad_norm": 0.818779153404372,
+      "learning_rate": 2.4926869846663446e-05,
+      "loss": 12.1212,
+      "step": 13332
+    },
+    {
+      "epoch": 2.331584699453552,
+      "grad_norm": 1.031179336641682,
+      "learning_rate": 2.4914398318421638e-05,
+      "loss": 11.8967,
+      "step": 13333
+    },
+    {
+      "epoch": 2.33175956284153,
+      "grad_norm": 0.8625336278979103,
+      "learning_rate": 2.4901929466893836e-05,
+      "loss": 12.1954,
+      "step": 13334
+    },
+    {
+      "epoch": 2.331934426229508,
+      "grad_norm": 0.8336592991138468,
+      "learning_rate": 2.4889463292524473e-05,
+      "loss": 11.9941,
+      "step": 13335
+    },
+    {
+      "epoch": 2.3321092896174864,
+      "grad_norm": 0.9142479116419824,
+      "learning_rate": 2.4876999795757984e-05,
+      "loss": 12.0901,
+      "step": 13336
+    },
+    {
+      "epoch": 2.3322841530054643,
+      "grad_norm": 0.8312845296157174,
+      "learning_rate": 2.4864538977038677e-05,
+      "loss": 11.8691,
+      "step": 13337
+    },
+    {
+      "epoch": 2.3324590163934427,
+      "grad_norm": 0.8060361251726218,
+      "learning_rate": 2.485208083681081e-05,
+      "loss": 12.1786,
+      "step": 13338
+    },
+    {
+      "epoch": 2.3326338797814206,
+      "grad_norm": 0.8292116032173055,
+      "learning_rate": 2.4839625375518427e-05,
+      "loss": 12.1576,
+      "step": 13339
+    },
+    {
+      "epoch": 2.332808743169399,
+      "grad_norm": 0.8423846509305245,
+      "learning_rate": 2.4827172593605574e-05,
+      "loss": 12.0886,
+      "step": 13340
+    },
+    {
+      "epoch": 2.3329836065573772,
+      "grad_norm": 0.8845252339862472,
+      "learning_rate": 2.481472249151623e-05,
+      "loss": 12.0663,
+      "step": 13341
+    },
+    {
+      "epoch": 2.333158469945355,
+      "grad_norm": 0.8575175674412301,
+      "learning_rate": 2.480227506969415e-05,
+      "loss": 11.8583,
+      "step": 13342
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.8348780498295613,
+      "learning_rate": 2.4789830328583108e-05,
+      "loss": 12.1238,
+      "step": 13343
+    },
+    {
+      "epoch": 2.3335081967213114,
+      "grad_norm": 0.9305410500066715,
+      "learning_rate": 2.4777388268626733e-05,
+      "loss": 12.179,
+      "step": 13344
+    },
+    {
+      "epoch": 2.3336830601092897,
+      "grad_norm": 0.8499100553853651,
+      "learning_rate": 2.4764948890268614e-05,
+      "loss": 11.9309,
+      "step": 13345
+    },
+    {
+      "epoch": 2.3338579234972676,
+      "grad_norm": 0.8277464718790709,
+      "learning_rate": 2.4752512193952137e-05,
+      "loss": 11.8141,
+      "step": 13346
+    },
+    {
+      "epoch": 2.334032786885246,
+      "grad_norm": 0.8383687511570382,
+      "learning_rate": 2.4740078180120686e-05,
+      "loss": 11.9822,
+      "step": 13347
+    },
+    {
+      "epoch": 2.334207650273224,
+      "grad_norm": 0.8752814053828429,
+      "learning_rate": 2.4727646849217512e-05,
+      "loss": 12.0146,
+      "step": 13348
+    },
+    {
+      "epoch": 2.334382513661202,
+      "grad_norm": 0.8325318342358831,
+      "learning_rate": 2.4715218201685774e-05,
+      "loss": 11.6663,
+      "step": 13349
+    },
+    {
+      "epoch": 2.3345573770491805,
+      "grad_norm": 1.0014164413158295,
+      "learning_rate": 2.470279223796855e-05,
+      "loss": 12.0313,
+      "step": 13350
+    },
+    {
+      "epoch": 2.3347322404371584,
+      "grad_norm": 0.8626573950083841,
+      "learning_rate": 2.4690368958508815e-05,
+      "loss": 12.065,
+      "step": 13351
+    },
+    {
+      "epoch": 2.3349071038251368,
+      "grad_norm": 0.8503372355238288,
+      "learning_rate": 2.4677948363749427e-05,
+      "loss": 12.1375,
+      "step": 13352
+    },
+    {
+      "epoch": 2.3350819672131147,
+      "grad_norm": 0.8959356244282035,
+      "learning_rate": 2.4665530454133213e-05,
+      "loss": 12.0469,
+      "step": 13353
+    },
+    {
+      "epoch": 2.335256830601093,
+      "grad_norm": 0.8175478091198162,
+      "learning_rate": 2.4653115230102786e-05,
+      "loss": 11.9699,
+      "step": 13354
+    },
+    {
+      "epoch": 2.335431693989071,
+      "grad_norm": 0.8803634622819159,
+      "learning_rate": 2.464070269210077e-05,
+      "loss": 11.8916,
+      "step": 13355
+    },
+    {
+      "epoch": 2.3356065573770493,
+      "grad_norm": 0.9357607191958612,
+      "learning_rate": 2.4628292840569643e-05,
+      "loss": 11.8502,
+      "step": 13356
+    },
+    {
+      "epoch": 2.335781420765027,
+      "grad_norm": 0.8719818952251251,
+      "learning_rate": 2.4615885675951857e-05,
+      "loss": 12.1573,
+      "step": 13357
+    },
+    {
+      "epoch": 2.3359562841530055,
+      "grad_norm": 0.8398246300236192,
+      "learning_rate": 2.4603481198689637e-05,
+      "loss": 11.9132,
+      "step": 13358
+    },
+    {
+      "epoch": 2.336131147540984,
+      "grad_norm": 0.8101156882351699,
+      "learning_rate": 2.4591079409225227e-05,
+      "loss": 11.9854,
+      "step": 13359
+    },
+    {
+      "epoch": 2.3363060109289617,
+      "grad_norm": 0.8417871636714859,
+      "learning_rate": 2.4578680308000747e-05,
+      "loss": 11.939,
+      "step": 13360
+    },
+    {
+      "epoch": 2.33648087431694,
+      "grad_norm": 0.8134221431462848,
+      "learning_rate": 2.4566283895458176e-05,
+      "loss": 12.1475,
+      "step": 13361
+    },
+    {
+      "epoch": 2.336655737704918,
+      "grad_norm": 0.7839719891956963,
+      "learning_rate": 2.455389017203944e-05,
+      "loss": 12.086,
+      "step": 13362
+    },
+    {
+      "epoch": 2.3368306010928963,
+      "grad_norm": 0.851830343489759,
+      "learning_rate": 2.4541499138186374e-05,
+      "loss": 11.9398,
+      "step": 13363
+    },
+    {
+      "epoch": 2.337005464480874,
+      "grad_norm": 0.8210864543828253,
+      "learning_rate": 2.4529110794340694e-05,
+      "loss": 12.382,
+      "step": 13364
+    },
+    {
+      "epoch": 2.3371803278688525,
+      "grad_norm": 0.8428244283443626,
+      "learning_rate": 2.451672514094404e-05,
+      "loss": 12.1128,
+      "step": 13365
+    },
+    {
+      "epoch": 2.3373551912568304,
+      "grad_norm": 0.865171800196604,
+      "learning_rate": 2.4504342178437946e-05,
+      "loss": 12.0246,
+      "step": 13366
+    },
+    {
+      "epoch": 2.337530054644809,
+      "grad_norm": 0.8613990544055984,
+      "learning_rate": 2.449196190726385e-05,
+      "loss": 11.9576,
+      "step": 13367
+    },
+    {
+      "epoch": 2.337704918032787,
+      "grad_norm": 0.8930280509478427,
+      "learning_rate": 2.4479584327863115e-05,
+      "loss": 12.1045,
+      "step": 13368
+    },
+    {
+      "epoch": 2.337879781420765,
+      "grad_norm": 0.8456623278617688,
+      "learning_rate": 2.446720944067693e-05,
+      "loss": 12.149,
+      "step": 13369
+    },
+    {
+      "epoch": 2.338054644808743,
+      "grad_norm": 0.8600418369956184,
+      "learning_rate": 2.4454837246146476e-05,
+      "loss": 12.0777,
+      "step": 13370
+    },
+    {
+      "epoch": 2.3382295081967213,
+      "grad_norm": 0.9210621446011747,
+      "learning_rate": 2.4442467744712816e-05,
+      "loss": 12.0807,
+      "step": 13371
+    },
+    {
+      "epoch": 2.3384043715846996,
+      "grad_norm": 0.7837109757569086,
+      "learning_rate": 2.4430100936816924e-05,
+      "loss": 11.626,
+      "step": 13372
+    },
+    {
+      "epoch": 2.3385792349726775,
+      "grad_norm": 0.9358667238294556,
+      "learning_rate": 2.4417736822899617e-05,
+      "loss": 12.0586,
+      "step": 13373
+    },
+    {
+      "epoch": 2.338754098360656,
+      "grad_norm": 0.7958534474037173,
+      "learning_rate": 2.4405375403401677e-05,
+      "loss": 11.9759,
+      "step": 13374
+    },
+    {
+      "epoch": 2.3389289617486337,
+      "grad_norm": 0.8385583005655304,
+      "learning_rate": 2.4393016678763813e-05,
+      "loss": 12.1147,
+      "step": 13375
+    },
+    {
+      "epoch": 2.339103825136612,
+      "grad_norm": 0.8969593245693805,
+      "learning_rate": 2.4380660649426533e-05,
+      "loss": 12.2305,
+      "step": 13376
+    },
+    {
+      "epoch": 2.33927868852459,
+      "grad_norm": 0.8637465379515069,
+      "learning_rate": 2.4368307315830364e-05,
+      "loss": 12.0622,
+      "step": 13377
+    },
+    {
+      "epoch": 2.3394535519125683,
+      "grad_norm": 0.8436699409112446,
+      "learning_rate": 2.435595667841566e-05,
+      "loss": 12.1131,
+      "step": 13378
+    },
+    {
+      "epoch": 2.339628415300546,
+      "grad_norm": 0.8092904324614943,
+      "learning_rate": 2.434360873762276e-05,
+      "loss": 12.0406,
+      "step": 13379
+    },
+    {
+      "epoch": 2.3398032786885246,
+      "grad_norm": 0.8676304866702825,
+      "learning_rate": 2.4331263493891777e-05,
+      "loss": 12.1673,
+      "step": 13380
+    },
+    {
+      "epoch": 2.339978142076503,
+      "grad_norm": 0.8780440216772432,
+      "learning_rate": 2.4318920947662836e-05,
+      "loss": 12.0752,
+      "step": 13381
+    },
+    {
+      "epoch": 2.340153005464481,
+      "grad_norm": 0.8693543890548161,
+      "learning_rate": 2.430658109937595e-05,
+      "loss": 11.9363,
+      "step": 13382
+    },
+    {
+      "epoch": 2.340327868852459,
+      "grad_norm": 0.811617175514788,
+      "learning_rate": 2.4294243949470997e-05,
+      "loss": 12.1073,
+      "step": 13383
+    },
+    {
+      "epoch": 2.340502732240437,
+      "grad_norm": 0.8372451360517987,
+      "learning_rate": 2.42819094983878e-05,
+      "loss": 12.2859,
+      "step": 13384
+    },
+    {
+      "epoch": 2.3406775956284154,
+      "grad_norm": 0.827655190531325,
+      "learning_rate": 2.4269577746566064e-05,
+      "loss": 11.9849,
+      "step": 13385
+    },
+    {
+      "epoch": 2.3408524590163933,
+      "grad_norm": 0.8856015080177497,
+      "learning_rate": 2.42572486944454e-05,
+      "loss": 12.0117,
+      "step": 13386
+    },
+    {
+      "epoch": 2.3410273224043716,
+      "grad_norm": 0.8602156180849126,
+      "learning_rate": 2.424492234246536e-05,
+      "loss": 12.0177,
+      "step": 13387
+    },
+    {
+      "epoch": 2.3412021857923495,
+      "grad_norm": 0.965217785371542,
+      "learning_rate": 2.4232598691065288e-05,
+      "loss": 12.1243,
+      "step": 13388
+    },
+    {
+      "epoch": 2.341377049180328,
+      "grad_norm": 0.8676853979107836,
+      "learning_rate": 2.422027774068455e-05,
+      "loss": 11.9811,
+      "step": 13389
+    },
+    {
+      "epoch": 2.341551912568306,
+      "grad_norm": 0.8344968693424133,
+      "learning_rate": 2.42079594917624e-05,
+      "loss": 11.8405,
+      "step": 13390
+    },
+    {
+      "epoch": 2.341726775956284,
+      "grad_norm": 1.1313467103567876,
+      "learning_rate": 2.4195643944737923e-05,
+      "loss": 11.8582,
+      "step": 13391
+    },
+    {
+      "epoch": 2.3419016393442624,
+      "grad_norm": 0.8251939385484589,
+      "learning_rate": 2.4183331100050154e-05,
+      "loss": 12.067,
+      "step": 13392
+    },
+    {
+      "epoch": 2.3420765027322403,
+      "grad_norm": 0.9102321326706452,
+      "learning_rate": 2.4171020958138057e-05,
+      "loss": 12.1696,
+      "step": 13393
+    },
+    {
+      "epoch": 2.3422513661202187,
+      "grad_norm": 0.8410624026655225,
+      "learning_rate": 2.41587135194405e-05,
+      "loss": 11.9749,
+      "step": 13394
+    },
+    {
+      "epoch": 2.3424262295081966,
+      "grad_norm": 0.8371446115317405,
+      "learning_rate": 2.4146408784396167e-05,
+      "loss": 12.0434,
+      "step": 13395
+    },
+    {
+      "epoch": 2.342601092896175,
+      "grad_norm": 0.876062074825613,
+      "learning_rate": 2.413410675344373e-05,
+      "loss": 12.0971,
+      "step": 13396
+    },
+    {
+      "epoch": 2.342775956284153,
+      "grad_norm": 0.8575097553018197,
+      "learning_rate": 2.412180742702175e-05,
+      "loss": 12.1495,
+      "step": 13397
+    },
+    {
+      "epoch": 2.342950819672131,
+      "grad_norm": 0.8019872566984028,
+      "learning_rate": 2.4109510805568703e-05,
+      "loss": 12.1579,
+      "step": 13398
+    },
+    {
+      "epoch": 2.3431256830601095,
+      "grad_norm": 0.8068418811847881,
+      "learning_rate": 2.40972168895229e-05,
+      "loss": 11.9554,
+      "step": 13399
+    },
+    {
+      "epoch": 2.3433005464480874,
+      "grad_norm": 0.8926918592796037,
+      "learning_rate": 2.4084925679322635e-05,
+      "loss": 11.9521,
+      "step": 13400
+    },
+    {
+      "epoch": 2.3434754098360657,
+      "grad_norm": 0.7775836918858025,
+      "learning_rate": 2.4072637175406075e-05,
+      "loss": 12.1416,
+      "step": 13401
+    },
+    {
+      "epoch": 2.3436502732240436,
+      "grad_norm": 0.8333329040168425,
+      "learning_rate": 2.4060351378211287e-05,
+      "loss": 11.856,
+      "step": 13402
+    },
+    {
+      "epoch": 2.343825136612022,
+      "grad_norm": 0.8944386227012071,
+      "learning_rate": 2.4048068288176238e-05,
+      "loss": 11.7943,
+      "step": 13403
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 0.8198605869322858,
+      "learning_rate": 2.4035787905738828e-05,
+      "loss": 11.9028,
+      "step": 13404
+    },
+    {
+      "epoch": 2.344174863387978,
+      "grad_norm": 0.864046662309102,
+      "learning_rate": 2.4023510231336843e-05,
+      "loss": 11.8633,
+      "step": 13405
+    },
+    {
+      "epoch": 2.344349726775956,
+      "grad_norm": 0.8180755939680983,
+      "learning_rate": 2.401123526540793e-05,
+      "loss": 12.02,
+      "step": 13406
+    },
+    {
+      "epoch": 2.3445245901639344,
+      "grad_norm": 0.7897503435980534,
+      "learning_rate": 2.3998963008389686e-05,
+      "loss": 11.9779,
+      "step": 13407
+    },
+    {
+      "epoch": 2.344699453551913,
+      "grad_norm": 0.8347884126338254,
+      "learning_rate": 2.3986693460719624e-05,
+      "loss": 11.918,
+      "step": 13408
+    },
+    {
+      "epoch": 2.3448743169398907,
+      "grad_norm": 0.8643961671861139,
+      "learning_rate": 2.3974426622835144e-05,
+      "loss": 11.9484,
+      "step": 13409
+    },
+    {
+      "epoch": 2.345049180327869,
+      "grad_norm": 0.8315316721626621,
+      "learning_rate": 2.396216249517351e-05,
+      "loss": 11.8543,
+      "step": 13410
+    },
+    {
+      "epoch": 2.345224043715847,
+      "grad_norm": 0.7878220748793924,
+      "learning_rate": 2.3949901078171934e-05,
+      "loss": 12.0562,
+      "step": 13411
+    },
+    {
+      "epoch": 2.3453989071038253,
+      "grad_norm": 0.8259796535752063,
+      "learning_rate": 2.3937642372267532e-05,
+      "loss": 12.1235,
+      "step": 13412
+    },
+    {
+      "epoch": 2.345573770491803,
+      "grad_norm": 0.8383313860697753,
+      "learning_rate": 2.392538637789734e-05,
+      "loss": 11.7622,
+      "step": 13413
+    },
+    {
+      "epoch": 2.3457486338797815,
+      "grad_norm": 0.7695944824038503,
+      "learning_rate": 2.3913133095498218e-05,
+      "loss": 11.9043,
+      "step": 13414
+    },
+    {
+      "epoch": 2.3459234972677594,
+      "grad_norm": 0.92921059092116,
+      "learning_rate": 2.3900882525506997e-05,
+      "loss": 11.9064,
+      "step": 13415
+    },
+    {
+      "epoch": 2.3460983606557377,
+      "grad_norm": 0.9362980551947135,
+      "learning_rate": 2.38886346683604e-05,
+      "loss": 12.2249,
+      "step": 13416
+    },
+    {
+      "epoch": 2.346273224043716,
+      "grad_norm": 0.8223394443513612,
+      "learning_rate": 2.387638952449508e-05,
+      "loss": 11.8029,
+      "step": 13417
+    },
+    {
+      "epoch": 2.346448087431694,
+      "grad_norm": 0.8357133661922566,
+      "learning_rate": 2.386414709434751e-05,
+      "loss": 12.0355,
+      "step": 13418
+    },
+    {
+      "epoch": 2.3466229508196723,
+      "grad_norm": 0.8085181788779195,
+      "learning_rate": 2.3851907378354134e-05,
+      "loss": 12.1223,
+      "step": 13419
+    },
+    {
+      "epoch": 2.34679781420765,
+      "grad_norm": 0.8681311116050825,
+      "learning_rate": 2.3839670376951294e-05,
+      "loss": 12.1437,
+      "step": 13420
+    },
+    {
+      "epoch": 2.3469726775956286,
+      "grad_norm": 0.9761248862565451,
+      "learning_rate": 2.382743609057522e-05,
+      "loss": 12.1621,
+      "step": 13421
+    },
+    {
+      "epoch": 2.3471475409836065,
+      "grad_norm": 0.8471502753559924,
+      "learning_rate": 2.3815204519662048e-05,
+      "loss": 12.0121,
+      "step": 13422
+    },
+    {
+      "epoch": 2.347322404371585,
+      "grad_norm": 0.844109620291802,
+      "learning_rate": 2.380297566464783e-05,
+      "loss": 12.1352,
+      "step": 13423
+    },
+    {
+      "epoch": 2.3474972677595627,
+      "grad_norm": 0.9282737504019473,
+      "learning_rate": 2.3790749525968527e-05,
+      "loss": 11.9319,
+      "step": 13424
+    },
+    {
+      "epoch": 2.347672131147541,
+      "grad_norm": 0.8628896745507159,
+      "learning_rate": 2.377852610405993e-05,
+      "loss": 11.8911,
+      "step": 13425
+    },
+    {
+      "epoch": 2.3478469945355194,
+      "grad_norm": 0.8352847484895735,
+      "learning_rate": 2.3766305399357824e-05,
+      "loss": 12.2252,
+      "step": 13426
+    },
+    {
+      "epoch": 2.3480218579234973,
+      "grad_norm": 0.9782999338632414,
+      "learning_rate": 2.3754087412297866e-05,
+      "loss": 12.1463,
+      "step": 13427
+    },
+    {
+      "epoch": 2.348196721311475,
+      "grad_norm": 0.9345009825226078,
+      "learning_rate": 2.3741872143315634e-05,
+      "loss": 12.1038,
+      "step": 13428
+    },
+    {
+      "epoch": 2.3483715846994535,
+      "grad_norm": 0.9189743044876593,
+      "learning_rate": 2.372965959284653e-05,
+      "loss": 12.0333,
+      "step": 13429
+    },
+    {
+      "epoch": 2.348546448087432,
+      "grad_norm": 0.8507578645768092,
+      "learning_rate": 2.371744976132595e-05,
+      "loss": 12.0685,
+      "step": 13430
+    },
+    {
+      "epoch": 2.3487213114754097,
+      "grad_norm": 0.9538018860070078,
+      "learning_rate": 2.3705242649189164e-05,
+      "loss": 11.9662,
+      "step": 13431
+    },
+    {
+      "epoch": 2.348896174863388,
+      "grad_norm": 0.826734536074834,
+      "learning_rate": 2.369303825687137e-05,
+      "loss": 12.0032,
+      "step": 13432
+    },
+    {
+      "epoch": 2.349071038251366,
+      "grad_norm": 1.0321971754843218,
+      "learning_rate": 2.3680836584807563e-05,
+      "loss": 11.858,
+      "step": 13433
+    },
+    {
+      "epoch": 2.3492459016393443,
+      "grad_norm": 0.8850932213448499,
+      "learning_rate": 2.3668637633432777e-05,
+      "loss": 12.1136,
+      "step": 13434
+    },
+    {
+      "epoch": 2.3494207650273222,
+      "grad_norm": 0.9401973122009789,
+      "learning_rate": 2.3656441403181894e-05,
+      "loss": 12.1667,
+      "step": 13435
+    },
+    {
+      "epoch": 2.3495956284153006,
+      "grad_norm": 0.8548150001864083,
+      "learning_rate": 2.3644247894489657e-05,
+      "loss": 11.8184,
+      "step": 13436
+    },
+    {
+      "epoch": 2.3497704918032785,
+      "grad_norm": 0.7941667121530465,
+      "learning_rate": 2.3632057107790763e-05,
+      "loss": 11.9085,
+      "step": 13437
+    },
+    {
+      "epoch": 2.349945355191257,
+      "grad_norm": 0.9039761073462464,
+      "learning_rate": 2.3619869043519817e-05,
+      "loss": 11.9973,
+      "step": 13438
+    },
+    {
+      "epoch": 2.350120218579235,
+      "grad_norm": 0.7899159525782193,
+      "learning_rate": 2.3607683702111283e-05,
+      "loss": 11.815,
+      "step": 13439
+    },
+    {
+      "epoch": 2.350295081967213,
+      "grad_norm": 0.9792136646392243,
+      "learning_rate": 2.3595501083999584e-05,
+      "loss": 12.1577,
+      "step": 13440
+    },
+    {
+      "epoch": 2.3504699453551914,
+      "grad_norm": 0.846071229975628,
+      "learning_rate": 2.358332118961899e-05,
+      "loss": 12.1962,
+      "step": 13441
+    },
+    {
+      "epoch": 2.3506448087431693,
+      "grad_norm": 0.8982866918393939,
+      "learning_rate": 2.3571144019403725e-05,
+      "loss": 12.1347,
+      "step": 13442
+    },
+    {
+      "epoch": 2.3508196721311476,
+      "grad_norm": 0.8750128597966694,
+      "learning_rate": 2.3558969573787902e-05,
+      "loss": 12.0126,
+      "step": 13443
+    },
+    {
+      "epoch": 2.3509945355191255,
+      "grad_norm": 0.8380366423824663,
+      "learning_rate": 2.354679785320547e-05,
+      "loss": 12.0665,
+      "step": 13444
+    },
+    {
+      "epoch": 2.351169398907104,
+      "grad_norm": 0.8613302250265903,
+      "learning_rate": 2.353462885809037e-05,
+      "loss": 11.9907,
+      "step": 13445
+    },
+    {
+      "epoch": 2.3513442622950818,
+      "grad_norm": 0.8798482251670454,
+      "learning_rate": 2.3522462588876414e-05,
+      "loss": 12.029,
+      "step": 13446
+    },
+    {
+      "epoch": 2.35151912568306,
+      "grad_norm": 0.8568765447983026,
+      "learning_rate": 2.3510299045997343e-05,
+      "loss": 12.0906,
+      "step": 13447
+    },
+    {
+      "epoch": 2.3516939890710384,
+      "grad_norm": 0.7766712398747664,
+      "learning_rate": 2.3498138229886702e-05,
+      "loss": 12.1171,
+      "step": 13448
+    },
+    {
+      "epoch": 2.3518688524590163,
+      "grad_norm": 0.846732821135939,
+      "learning_rate": 2.348598014097806e-05,
+      "loss": 11.8759,
+      "step": 13449
+    },
+    {
+      "epoch": 2.3520437158469947,
+      "grad_norm": 0.9376888860533041,
+      "learning_rate": 2.3473824779704857e-05,
+      "loss": 12.0749,
+      "step": 13450
+    },
+    {
+      "epoch": 2.3522185792349726,
+      "grad_norm": 0.8294720854249402,
+      "learning_rate": 2.3461672146500358e-05,
+      "loss": 12.0494,
+      "step": 13451
+    },
+    {
+      "epoch": 2.352393442622951,
+      "grad_norm": 0.8166235325262664,
+      "learning_rate": 2.3449522241797826e-05,
+      "loss": 12.1227,
+      "step": 13452
+    },
+    {
+      "epoch": 2.352568306010929,
+      "grad_norm": 0.876361868034754,
+      "learning_rate": 2.3437375066030388e-05,
+      "loss": 12.0876,
+      "step": 13453
+    },
+    {
+      "epoch": 2.352743169398907,
+      "grad_norm": 0.8508696295327914,
+      "learning_rate": 2.342523061963107e-05,
+      "loss": 12.0486,
+      "step": 13454
+    },
+    {
+      "epoch": 2.352918032786885,
+      "grad_norm": 0.824317569282533,
+      "learning_rate": 2.3413088903032808e-05,
+      "loss": 11.9001,
+      "step": 13455
+    },
+    {
+      "epoch": 2.3530928961748634,
+      "grad_norm": 0.8093004220344003,
+      "learning_rate": 2.3400949916668446e-05,
+      "loss": 11.8441,
+      "step": 13456
+    },
+    {
+      "epoch": 2.3532677595628417,
+      "grad_norm": 0.8157200512191094,
+      "learning_rate": 2.3388813660970733e-05,
+      "loss": 12.1877,
+      "step": 13457
+    },
+    {
+      "epoch": 2.3534426229508196,
+      "grad_norm": 0.779034557465957,
+      "learning_rate": 2.337668013637232e-05,
+      "loss": 12.0794,
+      "step": 13458
+    },
+    {
+      "epoch": 2.353617486338798,
+      "grad_norm": 0.8533461959804268,
+      "learning_rate": 2.336454934330572e-05,
+      "loss": 12.1098,
+      "step": 13459
+    },
+    {
+      "epoch": 2.353792349726776,
+      "grad_norm": 0.8419944789150503,
+      "learning_rate": 2.3352421282203386e-05,
+      "loss": 11.9717,
+      "step": 13460
+    },
+    {
+      "epoch": 2.353967213114754,
+      "grad_norm": 0.8396955924300114,
+      "learning_rate": 2.334029595349768e-05,
+      "loss": 12.1872,
+      "step": 13461
+    },
+    {
+      "epoch": 2.354142076502732,
+      "grad_norm": 0.8507799437541534,
+      "learning_rate": 2.3328173357620886e-05,
+      "loss": 11.9708,
+      "step": 13462
+    },
+    {
+      "epoch": 2.3543169398907104,
+      "grad_norm": 0.8146893655880147,
+      "learning_rate": 2.331605349500511e-05,
+      "loss": 12.0973,
+      "step": 13463
+    },
+    {
+      "epoch": 2.3544918032786883,
+      "grad_norm": 0.8254890536917034,
+      "learning_rate": 2.3303936366082413e-05,
+      "loss": 12.0796,
+      "step": 13464
+    },
+    {
+      "epoch": 2.3546666666666667,
+      "grad_norm": 0.8080494653278139,
+      "learning_rate": 2.329182197128479e-05,
+      "loss": 11.8273,
+      "step": 13465
+    },
+    {
+      "epoch": 2.354841530054645,
+      "grad_norm": 0.8166451863390021,
+      "learning_rate": 2.3279710311044113e-05,
+      "loss": 11.7555,
+      "step": 13466
+    },
+    {
+      "epoch": 2.355016393442623,
+      "grad_norm": 0.8240432081200515,
+      "learning_rate": 2.3267601385792103e-05,
+      "loss": 12.0268,
+      "step": 13467
+    },
+    {
+      "epoch": 2.3551912568306013,
+      "grad_norm": 0.8542780748000901,
+      "learning_rate": 2.3255495195960443e-05,
+      "loss": 12.0283,
+      "step": 13468
+    },
+    {
+      "epoch": 2.355366120218579,
+      "grad_norm": 0.8983801939541074,
+      "learning_rate": 2.3243391741980747e-05,
+      "loss": 12.2243,
+      "step": 13469
+    },
+    {
+      "epoch": 2.3555409836065575,
+      "grad_norm": 0.9059192300421178,
+      "learning_rate": 2.323129102428443e-05,
+      "loss": 12.0952,
+      "step": 13470
+    },
+    {
+      "epoch": 2.3557158469945354,
+      "grad_norm": 0.9881463310974019,
+      "learning_rate": 2.321919304330288e-05,
+      "loss": 11.8561,
+      "step": 13471
+    },
+    {
+      "epoch": 2.3558907103825137,
+      "grad_norm": 0.8950525978468995,
+      "learning_rate": 2.3207097799467404e-05,
+      "loss": 11.7238,
+      "step": 13472
+    },
+    {
+      "epoch": 2.3560655737704916,
+      "grad_norm": 0.8469313056069571,
+      "learning_rate": 2.319500529320916e-05,
+      "loss": 12.1704,
+      "step": 13473
+    },
+    {
+      "epoch": 2.35624043715847,
+      "grad_norm": 0.8186527235408058,
+      "learning_rate": 2.3182915524959236e-05,
+      "loss": 11.7994,
+      "step": 13474
+    },
+    {
+      "epoch": 2.3564153005464483,
+      "grad_norm": 0.8771547894374032,
+      "learning_rate": 2.3170828495148632e-05,
+      "loss": 12.1484,
+      "step": 13475
+    },
+    {
+      "epoch": 2.356590163934426,
+      "grad_norm": 0.8806114778572337,
+      "learning_rate": 2.315874420420823e-05,
+      "loss": 12.1264,
+      "step": 13476
+    },
+    {
+      "epoch": 2.3567650273224046,
+      "grad_norm": 0.8703371865683434,
+      "learning_rate": 2.314666265256883e-05,
+      "loss": 12.0105,
+      "step": 13477
+    },
+    {
+      "epoch": 2.3569398907103825,
+      "grad_norm": 0.8465144952599694,
+      "learning_rate": 2.3134583840661095e-05,
+      "loss": 12.1002,
+      "step": 13478
+    },
+    {
+      "epoch": 2.357114754098361,
+      "grad_norm": 0.8551482844470008,
+      "learning_rate": 2.3122507768915635e-05,
+      "loss": 12.0918,
+      "step": 13479
+    },
+    {
+      "epoch": 2.3572896174863387,
+      "grad_norm": 0.8001154699796322,
+      "learning_rate": 2.3110434437762963e-05,
+      "loss": 12.1907,
+      "step": 13480
+    },
+    {
+      "epoch": 2.357464480874317,
+      "grad_norm": 0.83925990796642,
+      "learning_rate": 2.3098363847633488e-05,
+      "loss": 12.0436,
+      "step": 13481
+    },
+    {
+      "epoch": 2.357639344262295,
+      "grad_norm": 1.0519488015398917,
+      "learning_rate": 2.3086295998957453e-05,
+      "loss": 11.9694,
+      "step": 13482
+    },
+    {
+      "epoch": 2.3578142076502733,
+      "grad_norm": 0.8846428273877859,
+      "learning_rate": 2.3074230892165115e-05,
+      "loss": 12.2181,
+      "step": 13483
+    },
+    {
+      "epoch": 2.357989071038251,
+      "grad_norm": 0.8791495373474026,
+      "learning_rate": 2.3062168527686588e-05,
+      "loss": 12.046,
+      "step": 13484
+    },
+    {
+      "epoch": 2.3581639344262295,
+      "grad_norm": 0.8422775470428837,
+      "learning_rate": 2.305010890595184e-05,
+      "loss": 11.7497,
+      "step": 13485
+    },
+    {
+      "epoch": 2.3583387978142074,
+      "grad_norm": 0.848370571741209,
+      "learning_rate": 2.30380520273908e-05,
+      "loss": 11.9656,
+      "step": 13486
+    },
+    {
+      "epoch": 2.3585136612021858,
+      "grad_norm": 0.818312318962297,
+      "learning_rate": 2.302599789243328e-05,
+      "loss": 11.9769,
+      "step": 13487
+    },
+    {
+      "epoch": 2.358688524590164,
+      "grad_norm": 0.8832224788215676,
+      "learning_rate": 2.3013946501509044e-05,
+      "loss": 12.0851,
+      "step": 13488
+    },
+    {
+      "epoch": 2.358863387978142,
+      "grad_norm": 0.8853980132557273,
+      "learning_rate": 2.3001897855047626e-05,
+      "loss": 12.1052,
+      "step": 13489
+    },
+    {
+      "epoch": 2.3590382513661203,
+      "grad_norm": 0.8991740788507105,
+      "learning_rate": 2.298985195347858e-05,
+      "loss": 12.0474,
+      "step": 13490
+    },
+    {
+      "epoch": 2.3592131147540982,
+      "grad_norm": 0.8944947854576859,
+      "learning_rate": 2.2977808797231348e-05,
+      "loss": 12.2406,
+      "step": 13491
+    },
+    {
+      "epoch": 2.3593879781420766,
+      "grad_norm": 0.839325972082787,
+      "learning_rate": 2.2965768386735242e-05,
+      "loss": 11.9905,
+      "step": 13492
+    },
+    {
+      "epoch": 2.3595628415300545,
+      "grad_norm": 0.8642164027206382,
+      "learning_rate": 2.2953730722419475e-05,
+      "loss": 11.8055,
+      "step": 13493
+    },
+    {
+      "epoch": 2.359737704918033,
+      "grad_norm": 0.8873393012044887,
+      "learning_rate": 2.2941695804713203e-05,
+      "loss": 11.9733,
+      "step": 13494
+    },
+    {
+      "epoch": 2.3599125683060107,
+      "grad_norm": 0.8113732005546687,
+      "learning_rate": 2.292966363404544e-05,
+      "loss": 12.0585,
+      "step": 13495
+    },
+    {
+      "epoch": 2.360087431693989,
+      "grad_norm": 0.8090720737836317,
+      "learning_rate": 2.291763421084514e-05,
+      "loss": 11.9083,
+      "step": 13496
+    },
+    {
+      "epoch": 2.3602622950819674,
+      "grad_norm": 0.8967563868066871,
+      "learning_rate": 2.2905607535541096e-05,
+      "loss": 11.9763,
+      "step": 13497
+    },
+    {
+      "epoch": 2.3604371584699453,
+      "grad_norm": 0.8733511983169072,
+      "learning_rate": 2.2893583608562074e-05,
+      "loss": 11.9305,
+      "step": 13498
+    },
+    {
+      "epoch": 2.3606120218579236,
+      "grad_norm": 0.8152001938742794,
+      "learning_rate": 2.2881562430336733e-05,
+      "loss": 12.1779,
+      "step": 13499
+    },
+    {
+      "epoch": 2.3607868852459015,
+      "grad_norm": 0.8665359963603033,
+      "learning_rate": 2.2869544001293554e-05,
+      "loss": 12.11,
+      "step": 13500
+    },
+    {
+      "epoch": 2.36096174863388,
+      "grad_norm": 0.880381094670043,
+      "learning_rate": 2.285752832186102e-05,
+      "loss": 12.1234,
+      "step": 13501
+    },
+    {
+      "epoch": 2.3611366120218578,
+      "grad_norm": 0.8660788745976062,
+      "learning_rate": 2.2845515392467466e-05,
+      "loss": 12.125,
+      "step": 13502
+    },
+    {
+      "epoch": 2.361311475409836,
+      "grad_norm": 0.8399396624737504,
+      "learning_rate": 2.2833505213541175e-05,
+      "loss": 12.1146,
+      "step": 13503
+    },
+    {
+      "epoch": 2.361486338797814,
+      "grad_norm": 0.8694419269707886,
+      "learning_rate": 2.2821497785510237e-05,
+      "loss": 11.9201,
+      "step": 13504
+    },
+    {
+      "epoch": 2.3616612021857923,
+      "grad_norm": 0.8089295725113208,
+      "learning_rate": 2.2809493108802727e-05,
+      "loss": 12.1271,
+      "step": 13505
+    },
+    {
+      "epoch": 2.3618360655737707,
+      "grad_norm": 0.8111217779523553,
+      "learning_rate": 2.27974911838466e-05,
+      "loss": 12.0699,
+      "step": 13506
+    },
+    {
+      "epoch": 2.3620109289617486,
+      "grad_norm": 0.846955397239367,
+      "learning_rate": 2.2785492011069744e-05,
+      "loss": 11.9139,
+      "step": 13507
+    },
+    {
+      "epoch": 2.362185792349727,
+      "grad_norm": 0.8066933504223549,
+      "learning_rate": 2.277349559089984e-05,
+      "loss": 12.0151,
+      "step": 13508
+    },
+    {
+      "epoch": 2.362360655737705,
+      "grad_norm": 0.8755676842750763,
+      "learning_rate": 2.2761501923764606e-05,
+      "loss": 12.083,
+      "step": 13509
+    },
+    {
+      "epoch": 2.362535519125683,
+      "grad_norm": 0.8744785466891387,
+      "learning_rate": 2.2749511010091574e-05,
+      "loss": 12.3521,
+      "step": 13510
+    },
+    {
+      "epoch": 2.362710382513661,
+      "grad_norm": 0.8413054274194686,
+      "learning_rate": 2.2737522850308214e-05,
+      "loss": 12.0497,
+      "step": 13511
+    },
+    {
+      "epoch": 2.3628852459016394,
+      "grad_norm": 0.9759781159641227,
+      "learning_rate": 2.2725537444841906e-05,
+      "loss": 12.247,
+      "step": 13512
+    },
+    {
+      "epoch": 2.3630601092896173,
+      "grad_norm": 0.8166139240901241,
+      "learning_rate": 2.2713554794119896e-05,
+      "loss": 12.0958,
+      "step": 13513
+    },
+    {
+      "epoch": 2.3632349726775956,
+      "grad_norm": 0.903625793945707,
+      "learning_rate": 2.2701574898569378e-05,
+      "loss": 11.9732,
+      "step": 13514
+    },
+    {
+      "epoch": 2.363409836065574,
+      "grad_norm": 0.8637078255732841,
+      "learning_rate": 2.2689597758617387e-05,
+      "loss": 11.9432,
+      "step": 13515
+    },
+    {
+      "epoch": 2.363584699453552,
+      "grad_norm": 0.9188541744966677,
+      "learning_rate": 2.2677623374690903e-05,
+      "loss": 12.0921,
+      "step": 13516
+    },
+    {
+      "epoch": 2.36375956284153,
+      "grad_norm": 1.0006505265145247,
+      "learning_rate": 2.2665651747216797e-05,
+      "loss": 12.0252,
+      "step": 13517
+    },
+    {
+      "epoch": 2.363934426229508,
+      "grad_norm": 0.8222809319691587,
+      "learning_rate": 2.2653682876621885e-05,
+      "loss": 12.137,
+      "step": 13518
+    },
+    {
+      "epoch": 2.3641092896174865,
+      "grad_norm": 0.8242114439258678,
+      "learning_rate": 2.2641716763332788e-05,
+      "loss": 12.0831,
+      "step": 13519
+    },
+    {
+      "epoch": 2.3642841530054644,
+      "grad_norm": 0.8549923568883974,
+      "learning_rate": 2.2629753407776087e-05,
+      "loss": 12.0564,
+      "step": 13520
+    },
+    {
+      "epoch": 2.3644590163934427,
+      "grad_norm": 0.8314919481144283,
+      "learning_rate": 2.261779281037829e-05,
+      "loss": 12.1236,
+      "step": 13521
+    },
+    {
+      "epoch": 2.3646338797814206,
+      "grad_norm": 0.9405834236301422,
+      "learning_rate": 2.2605834971565797e-05,
+      "loss": 12.2073,
+      "step": 13522
+    },
+    {
+      "epoch": 2.364808743169399,
+      "grad_norm": 0.9018711119872009,
+      "learning_rate": 2.2593879891764835e-05,
+      "loss": 12.0074,
+      "step": 13523
+    },
+    {
+      "epoch": 2.3649836065573773,
+      "grad_norm": 0.8829872443647495,
+      "learning_rate": 2.2581927571401618e-05,
+      "loss": 11.986,
+      "step": 13524
+    },
+    {
+      "epoch": 2.365158469945355,
+      "grad_norm": 0.8652053846626341,
+      "learning_rate": 2.2569978010902228e-05,
+      "loss": 11.9574,
+      "step": 13525
+    },
+    {
+      "epoch": 2.3653333333333335,
+      "grad_norm": 0.8934740938593152,
+      "learning_rate": 2.2558031210692677e-05,
+      "loss": 11.9956,
+      "step": 13526
+    },
+    {
+      "epoch": 2.3655081967213114,
+      "grad_norm": 0.8669633812628049,
+      "learning_rate": 2.2546087171198805e-05,
+      "loss": 12.0015,
+      "step": 13527
+    },
+    {
+      "epoch": 2.3656830601092897,
+      "grad_norm": 0.8625453124573486,
+      "learning_rate": 2.2534145892846437e-05,
+      "loss": 12.0771,
+      "step": 13528
+    },
+    {
+      "epoch": 2.3658579234972676,
+      "grad_norm": 0.8456280987273752,
+      "learning_rate": 2.2522207376061255e-05,
+      "loss": 12.0421,
+      "step": 13529
+    },
+    {
+      "epoch": 2.366032786885246,
+      "grad_norm": 0.7838683195963522,
+      "learning_rate": 2.2510271621268852e-05,
+      "loss": 11.9843,
+      "step": 13530
+    },
+    {
+      "epoch": 2.366207650273224,
+      "grad_norm": 1.0279216472109387,
+      "learning_rate": 2.249833862889473e-05,
+      "loss": 12.0258,
+      "step": 13531
+    },
+    {
+      "epoch": 2.3663825136612022,
+      "grad_norm": 1.0631659648004406,
+      "learning_rate": 2.248640839936429e-05,
+      "loss": 12.0801,
+      "step": 13532
+    },
+    {
+      "epoch": 2.3665573770491806,
+      "grad_norm": 0.8901940867951976,
+      "learning_rate": 2.2474480933102847e-05,
+      "loss": 12.0889,
+      "step": 13533
+    },
+    {
+      "epoch": 2.3667322404371585,
+      "grad_norm": 0.87674660825702,
+      "learning_rate": 2.246255623053556e-05,
+      "loss": 12.0976,
+      "step": 13534
+    },
+    {
+      "epoch": 2.366907103825137,
+      "grad_norm": 0.8595872248287717,
+      "learning_rate": 2.2450634292087546e-05,
+      "loss": 12.0304,
+      "step": 13535
+    },
+    {
+      "epoch": 2.3670819672131147,
+      "grad_norm": 0.8891532782048179,
+      "learning_rate": 2.2438715118183817e-05,
+      "loss": 12.0125,
+      "step": 13536
+    },
+    {
+      "epoch": 2.367256830601093,
+      "grad_norm": 0.8208512364402389,
+      "learning_rate": 2.2426798709249297e-05,
+      "loss": 11.9795,
+      "step": 13537
+    },
+    {
+      "epoch": 2.367431693989071,
+      "grad_norm": 0.8081952110755495,
+      "learning_rate": 2.2414885065708747e-05,
+      "loss": 11.8247,
+      "step": 13538
+    },
+    {
+      "epoch": 2.3676065573770493,
+      "grad_norm": 0.9263520360778775,
+      "learning_rate": 2.2402974187986892e-05,
+      "loss": 11.8459,
+      "step": 13539
+    },
+    {
+      "epoch": 2.367781420765027,
+      "grad_norm": 1.0120540430198703,
+      "learning_rate": 2.239106607650834e-05,
+      "loss": 11.9954,
+      "step": 13540
+    },
+    {
+      "epoch": 2.3679562841530055,
+      "grad_norm": 0.8310662407129568,
+      "learning_rate": 2.2379160731697636e-05,
+      "loss": 12.1222,
+      "step": 13541
+    },
+    {
+      "epoch": 2.3681311475409834,
+      "grad_norm": 0.975449164364088,
+      "learning_rate": 2.236725815397913e-05,
+      "loss": 12.0389,
+      "step": 13542
+    },
+    {
+      "epoch": 2.3683060109289618,
+      "grad_norm": 0.867846116067861,
+      "learning_rate": 2.2355358343777176e-05,
+      "loss": 11.7961,
+      "step": 13543
+    },
+    {
+      "epoch": 2.3684808743169397,
+      "grad_norm": 1.1257264228867467,
+      "learning_rate": 2.2343461301515977e-05,
+      "loss": 12.1504,
+      "step": 13544
+    },
+    {
+      "epoch": 2.368655737704918,
+      "grad_norm": 0.8411291137303946,
+      "learning_rate": 2.2331567027619672e-05,
+      "loss": 12.0908,
+      "step": 13545
+    },
+    {
+      "epoch": 2.3688306010928963,
+      "grad_norm": 0.8267461912066381,
+      "learning_rate": 2.231967552251224e-05,
+      "loss": 12.1872,
+      "step": 13546
+    },
+    {
+      "epoch": 2.3690054644808742,
+      "grad_norm": 1.0583638524129733,
+      "learning_rate": 2.2307786786617602e-05,
+      "loss": 12.2179,
+      "step": 13547
+    },
+    {
+      "epoch": 2.3691803278688526,
+      "grad_norm": 0.8712979969325148,
+      "learning_rate": 2.2295900820359606e-05,
+      "loss": 12.0004,
+      "step": 13548
+    },
+    {
+      "epoch": 2.3693551912568305,
+      "grad_norm": 0.8243372824020792,
+      "learning_rate": 2.2284017624161945e-05,
+      "loss": 12.0186,
+      "step": 13549
+    },
+    {
+      "epoch": 2.369530054644809,
+      "grad_norm": 0.8443769922940614,
+      "learning_rate": 2.227213719844826e-05,
+      "loss": 12.0106,
+      "step": 13550
+    },
+    {
+      "epoch": 2.3697049180327867,
+      "grad_norm": 0.8823665461352038,
+      "learning_rate": 2.2260259543642072e-05,
+      "loss": 12.0096,
+      "step": 13551
+    },
+    {
+      "epoch": 2.369879781420765,
+      "grad_norm": 0.8553735865922159,
+      "learning_rate": 2.224838466016682e-05,
+      "loss": 12.0201,
+      "step": 13552
+    },
+    {
+      "epoch": 2.370054644808743,
+      "grad_norm": 0.8440739130319683,
+      "learning_rate": 2.22365125484458e-05,
+      "loss": 12.071,
+      "step": 13553
+    },
+    {
+      "epoch": 2.3702295081967213,
+      "grad_norm": 0.7616062413663331,
+      "learning_rate": 2.222464320890224e-05,
+      "loss": 11.9459,
+      "step": 13554
+    },
+    {
+      "epoch": 2.3704043715846996,
+      "grad_norm": 1.0093751027046234,
+      "learning_rate": 2.2212776641959277e-05,
+      "loss": 11.882,
+      "step": 13555
+    },
+    {
+      "epoch": 2.3705792349726775,
+      "grad_norm": 0.8971650405646548,
+      "learning_rate": 2.220091284803998e-05,
+      "loss": 11.9927,
+      "step": 13556
+    },
+    {
+      "epoch": 2.370754098360656,
+      "grad_norm": 0.8552178509092108,
+      "learning_rate": 2.21890518275672e-05,
+      "loss": 12.221,
+      "step": 13557
+    },
+    {
+      "epoch": 2.3709289617486338,
+      "grad_norm": 0.8496435937220906,
+      "learning_rate": 2.2177193580963818e-05,
+      "loss": 11.9694,
+      "step": 13558
+    },
+    {
+      "epoch": 2.371103825136612,
+      "grad_norm": 0.9653681769189565,
+      "learning_rate": 2.216533810865259e-05,
+      "loss": 12.1074,
+      "step": 13559
+    },
+    {
+      "epoch": 2.37127868852459,
+      "grad_norm": 0.8434626710939879,
+      "learning_rate": 2.2153485411056087e-05,
+      "loss": 11.9461,
+      "step": 13560
+    },
+    {
+      "epoch": 2.3714535519125683,
+      "grad_norm": 0.8328140788269438,
+      "learning_rate": 2.214163548859687e-05,
+      "loss": 12.0712,
+      "step": 13561
+    },
+    {
+      "epoch": 2.3716284153005462,
+      "grad_norm": 0.8136062354775511,
+      "learning_rate": 2.2129788341697387e-05,
+      "loss": 12.0498,
+      "step": 13562
+    },
+    {
+      "epoch": 2.3718032786885246,
+      "grad_norm": 0.8743529929189215,
+      "learning_rate": 2.2117943970779963e-05,
+      "loss": 11.9441,
+      "step": 13563
+    },
+    {
+      "epoch": 2.371978142076503,
+      "grad_norm": 0.8086957838347429,
+      "learning_rate": 2.2106102376266845e-05,
+      "loss": 12.0929,
+      "step": 13564
+    },
+    {
+      "epoch": 2.372153005464481,
+      "grad_norm": 0.8474579593767468,
+      "learning_rate": 2.2094263558580165e-05,
+      "loss": 12.0239,
+      "step": 13565
+    },
+    {
+      "epoch": 2.372327868852459,
+      "grad_norm": 0.8866805398241027,
+      "learning_rate": 2.2082427518141968e-05,
+      "loss": 12.0994,
+      "step": 13566
+    },
+    {
+      "epoch": 2.372502732240437,
+      "grad_norm": 0.8555261250876418,
+      "learning_rate": 2.2070594255374222e-05,
+      "loss": 11.8967,
+      "step": 13567
+    },
+    {
+      "epoch": 2.3726775956284154,
+      "grad_norm": 0.8683079789403861,
+      "learning_rate": 2.205876377069871e-05,
+      "loss": 12.0937,
+      "step": 13568
+    },
+    {
+      "epoch": 2.3728524590163933,
+      "grad_norm": 0.8397706586329228,
+      "learning_rate": 2.2046936064537205e-05,
+      "loss": 11.8732,
+      "step": 13569
+    },
+    {
+      "epoch": 2.3730273224043716,
+      "grad_norm": 0.9114631285562815,
+      "learning_rate": 2.2035111137311348e-05,
+      "loss": 12.2468,
+      "step": 13570
+    },
+    {
+      "epoch": 2.3732021857923495,
+      "grad_norm": 0.8585083933095229,
+      "learning_rate": 2.2023288989442713e-05,
+      "loss": 12.1006,
+      "step": 13571
+    },
+    {
+      "epoch": 2.373377049180328,
+      "grad_norm": 0.8562909764964266,
+      "learning_rate": 2.2011469621352697e-05,
+      "loss": 11.8868,
+      "step": 13572
+    },
+    {
+      "epoch": 2.373551912568306,
+      "grad_norm": 0.8122469085617645,
+      "learning_rate": 2.1999653033462674e-05,
+      "loss": 12.1091,
+      "step": 13573
+    },
+    {
+      "epoch": 2.373726775956284,
+      "grad_norm": 0.8311941193395035,
+      "learning_rate": 2.1987839226193884e-05,
+      "loss": 12.0685,
+      "step": 13574
+    },
+    {
+      "epoch": 2.3739016393442625,
+      "grad_norm": 0.834499340938885,
+      "learning_rate": 2.1976028199967502e-05,
+      "loss": 12.0754,
+      "step": 13575
+    },
+    {
+      "epoch": 2.3740765027322404,
+      "grad_norm": 0.7982453768179,
+      "learning_rate": 2.1964219955204524e-05,
+      "loss": 11.9237,
+      "step": 13576
+    },
+    {
+      "epoch": 2.3742513661202187,
+      "grad_norm": 0.821269923034882,
+      "learning_rate": 2.1952414492325935e-05,
+      "loss": 12.0744,
+      "step": 13577
+    },
+    {
+      "epoch": 2.3744262295081966,
+      "grad_norm": 0.8084986375481263,
+      "learning_rate": 2.194061181175261e-05,
+      "loss": 11.9594,
+      "step": 13578
+    },
+    {
+      "epoch": 2.374601092896175,
+      "grad_norm": 0.8234713131006964,
+      "learning_rate": 2.1928811913905235e-05,
+      "loss": 12.1115,
+      "step": 13579
+    },
+    {
+      "epoch": 2.374775956284153,
+      "grad_norm": 0.8880753749042104,
+      "learning_rate": 2.1917014799204505e-05,
+      "loss": 12.1384,
+      "step": 13580
+    },
+    {
+      "epoch": 2.374950819672131,
+      "grad_norm": 0.877290196804364,
+      "learning_rate": 2.1905220468070965e-05,
+      "loss": 12.1015,
+      "step": 13581
+    },
+    {
+      "epoch": 2.3751256830601095,
+      "grad_norm": 0.8601081611845441,
+      "learning_rate": 2.1893428920925075e-05,
+      "loss": 12.2039,
+      "step": 13582
+    },
+    {
+      "epoch": 2.3753005464480874,
+      "grad_norm": 0.8931133646953077,
+      "learning_rate": 2.1881640158187188e-05,
+      "loss": 12.117,
+      "step": 13583
+    },
+    {
+      "epoch": 2.3754754098360658,
+      "grad_norm": 0.7745138204916165,
+      "learning_rate": 2.186985418027755e-05,
+      "loss": 11.8801,
+      "step": 13584
+    },
+    {
+      "epoch": 2.3756502732240437,
+      "grad_norm": 0.821005723684895,
+      "learning_rate": 2.185807098761633e-05,
+      "loss": 12.1776,
+      "step": 13585
+    },
+    {
+      "epoch": 2.375825136612022,
+      "grad_norm": 0.8150997312433756,
+      "learning_rate": 2.1846290580623606e-05,
+      "loss": 11.8178,
+      "step": 13586
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 0.8899960704703832,
+      "learning_rate": 2.183451295971928e-05,
+      "loss": 12.0407,
+      "step": 13587
+    },
+    {
+      "epoch": 2.3761748633879782,
+      "grad_norm": 0.8096246278545622,
+      "learning_rate": 2.182273812532325e-05,
+      "loss": 12.1994,
+      "step": 13588
+    },
+    {
+      "epoch": 2.376349726775956,
+      "grad_norm": 0.8662793701916599,
+      "learning_rate": 2.1810966077855256e-05,
+      "loss": 12.2757,
+      "step": 13589
+    },
+    {
+      "epoch": 2.3765245901639345,
+      "grad_norm": 0.8984884597563276,
+      "learning_rate": 2.179919681773499e-05,
+      "loss": 11.9281,
+      "step": 13590
+    },
+    {
+      "epoch": 2.376699453551913,
+      "grad_norm": 0.8343708786859748,
+      "learning_rate": 2.1787430345381977e-05,
+      "loss": 11.7392,
+      "step": 13591
+    },
+    {
+      "epoch": 2.3768743169398907,
+      "grad_norm": 0.927799474750672,
+      "learning_rate": 2.1775666661215676e-05,
+      "loss": 12.1446,
+      "step": 13592
+    },
+    {
+      "epoch": 2.3770491803278686,
+      "grad_norm": 0.8474768683858059,
+      "learning_rate": 2.17639057656555e-05,
+      "loss": 11.9144,
+      "step": 13593
+    },
+    {
+      "epoch": 2.377224043715847,
+      "grad_norm": 0.9699225067084624,
+      "learning_rate": 2.1752147659120646e-05,
+      "loss": 11.5919,
+      "step": 13594
+    },
+    {
+      "epoch": 2.3773989071038253,
+      "grad_norm": 0.8388400118870694,
+      "learning_rate": 2.1740392342030302e-05,
+      "loss": 12.0807,
+      "step": 13595
+    },
+    {
+      "epoch": 2.377573770491803,
+      "grad_norm": 0.8288215167927921,
+      "learning_rate": 2.172863981480354e-05,
+      "loss": 12.0315,
+      "step": 13596
+    },
+    {
+      "epoch": 2.3777486338797815,
+      "grad_norm": 0.8812990320529427,
+      "learning_rate": 2.171689007785934e-05,
+      "loss": 11.9006,
+      "step": 13597
+    },
+    {
+      "epoch": 2.3779234972677594,
+      "grad_norm": 0.8874117982150612,
+      "learning_rate": 2.1705143131616524e-05,
+      "loss": 11.9394,
+      "step": 13598
+    },
+    {
+      "epoch": 2.3780983606557378,
+      "grad_norm": 0.8723511316658887,
+      "learning_rate": 2.1693398976493872e-05,
+      "loss": 11.8149,
+      "step": 13599
+    },
+    {
+      "epoch": 2.3782732240437157,
+      "grad_norm": 0.9396411014887524,
+      "learning_rate": 2.1681657612910054e-05,
+      "loss": 12.2105,
+      "step": 13600
+    },
+    {
+      "epoch": 2.378448087431694,
+      "grad_norm": 0.8705209304488492,
+      "learning_rate": 2.1669919041283636e-05,
+      "loss": 12.0189,
+      "step": 13601
+    },
+    {
+      "epoch": 2.378622950819672,
+      "grad_norm": 0.8664921535714503,
+      "learning_rate": 2.1658183262033094e-05,
+      "loss": 11.9754,
+      "step": 13602
+    },
+    {
+      "epoch": 2.3787978142076502,
+      "grad_norm": 0.9137157676413318,
+      "learning_rate": 2.164645027557678e-05,
+      "loss": 11.8391,
+      "step": 13603
+    },
+    {
+      "epoch": 2.3789726775956286,
+      "grad_norm": 0.8472284501033758,
+      "learning_rate": 2.163472008233297e-05,
+      "loss": 12.0822,
+      "step": 13604
+    },
+    {
+      "epoch": 2.3791475409836065,
+      "grad_norm": 0.8412488675710906,
+      "learning_rate": 2.1622992682719856e-05,
+      "loss": 12.0889,
+      "step": 13605
+    },
+    {
+      "epoch": 2.379322404371585,
+      "grad_norm": 0.7981775554156962,
+      "learning_rate": 2.1611268077155445e-05,
+      "loss": 11.8638,
+      "step": 13606
+    },
+    {
+      "epoch": 2.3794972677595627,
+      "grad_norm": 0.8514786107528086,
+      "learning_rate": 2.1599546266057746e-05,
+      "loss": 12.2626,
+      "step": 13607
+    },
+    {
+      "epoch": 2.379672131147541,
+      "grad_norm": 1.061317549164936,
+      "learning_rate": 2.158782724984465e-05,
+      "loss": 12.1501,
+      "step": 13608
+    },
+    {
+      "epoch": 2.379846994535519,
+      "grad_norm": 0.8138923705827935,
+      "learning_rate": 2.1576111028933865e-05,
+      "loss": 11.8929,
+      "step": 13609
+    },
+    {
+      "epoch": 2.3800218579234973,
+      "grad_norm": 0.8372899373334693,
+      "learning_rate": 2.1564397603743104e-05,
+      "loss": 12.1295,
+      "step": 13610
+    },
+    {
+      "epoch": 2.380196721311475,
+      "grad_norm": 0.8878735223862583,
+      "learning_rate": 2.1552686974689916e-05,
+      "loss": 11.8193,
+      "step": 13611
+    },
+    {
+      "epoch": 2.3803715846994535,
+      "grad_norm": 0.903841392581458,
+      "learning_rate": 2.154097914219181e-05,
+      "loss": 12.1205,
+      "step": 13612
+    },
+    {
+      "epoch": 2.380546448087432,
+      "grad_norm": 0.8479605196646534,
+      "learning_rate": 2.1529274106666096e-05,
+      "loss": 12.286,
+      "step": 13613
+    },
+    {
+      "epoch": 2.3807213114754098,
+      "grad_norm": 0.8360254483594122,
+      "learning_rate": 2.1517571868530084e-05,
+      "loss": 11.9351,
+      "step": 13614
+    },
+    {
+      "epoch": 2.380896174863388,
+      "grad_norm": 0.9343756082653305,
+      "learning_rate": 2.1505872428200946e-05,
+      "loss": 11.928,
+      "step": 13615
+    },
+    {
+      "epoch": 2.381071038251366,
+      "grad_norm": 0.9519138885011444,
+      "learning_rate": 2.149417578609576e-05,
+      "loss": 11.8988,
+      "step": 13616
+    },
+    {
+      "epoch": 2.3812459016393444,
+      "grad_norm": 0.8368711781030005,
+      "learning_rate": 2.1482481942631462e-05,
+      "loss": 12.0904,
+      "step": 13617
+    },
+    {
+      "epoch": 2.3814207650273223,
+      "grad_norm": 0.9106519647048015,
+      "learning_rate": 2.1470790898224946e-05,
+      "loss": 12.245,
+      "step": 13618
+    },
+    {
+      "epoch": 2.3815956284153006,
+      "grad_norm": 0.8767266977163493,
+      "learning_rate": 2.1459102653292973e-05,
+      "loss": 12.08,
+      "step": 13619
+    },
+    {
+      "epoch": 2.3817704918032785,
+      "grad_norm": 0.8906624751132837,
+      "learning_rate": 2.144741720825223e-05,
+      "loss": 12.0922,
+      "step": 13620
+    },
+    {
+      "epoch": 2.381945355191257,
+      "grad_norm": 0.8602939240716297,
+      "learning_rate": 2.1435734563519295e-05,
+      "loss": 12.0009,
+      "step": 13621
+    },
+    {
+      "epoch": 2.382120218579235,
+      "grad_norm": 0.8982819169734719,
+      "learning_rate": 2.142405471951061e-05,
+      "loss": 12.2047,
+      "step": 13622
+    },
+    {
+      "epoch": 2.382295081967213,
+      "grad_norm": 0.8963938647726862,
+      "learning_rate": 2.1412377676642613e-05,
+      "loss": 11.9947,
+      "step": 13623
+    },
+    {
+      "epoch": 2.3824699453551914,
+      "grad_norm": 0.867978466211863,
+      "learning_rate": 2.1400703435331504e-05,
+      "loss": 12.0535,
+      "step": 13624
+    },
+    {
+      "epoch": 2.3826448087431693,
+      "grad_norm": 0.8451321235337419,
+      "learning_rate": 2.138903199599347e-05,
+      "loss": 11.7971,
+      "step": 13625
+    },
+    {
+      "epoch": 2.3828196721311476,
+      "grad_norm": 0.8997192391027784,
+      "learning_rate": 2.137736335904461e-05,
+      "loss": 11.7899,
+      "step": 13626
+    },
+    {
+      "epoch": 2.3829945355191255,
+      "grad_norm": 0.8578617344469531,
+      "learning_rate": 2.1365697524900906e-05,
+      "loss": 12.1769,
+      "step": 13627
+    },
+    {
+      "epoch": 2.383169398907104,
+      "grad_norm": 0.8105686480404273,
+      "learning_rate": 2.1354034493978192e-05,
+      "loss": 11.8836,
+      "step": 13628
+    },
+    {
+      "epoch": 2.383344262295082,
+      "grad_norm": 0.8764864907681031,
+      "learning_rate": 2.1342374266692257e-05,
+      "loss": 11.9648,
+      "step": 13629
+    },
+    {
+      "epoch": 2.38351912568306,
+      "grad_norm": 0.8153043397575765,
+      "learning_rate": 2.1330716843458776e-05,
+      "loss": 11.9021,
+      "step": 13630
+    },
+    {
+      "epoch": 2.3836939890710385,
+      "grad_norm": 0.8447675347444183,
+      "learning_rate": 2.1319062224693355e-05,
+      "loss": 12.0944,
+      "step": 13631
+    },
+    {
+      "epoch": 2.3838688524590164,
+      "grad_norm": 0.8607296205663447,
+      "learning_rate": 2.1307410410811402e-05,
+      "loss": 12.1099,
+      "step": 13632
+    },
+    {
+      "epoch": 2.3840437158469947,
+      "grad_norm": 0.8137850037161333,
+      "learning_rate": 2.129576140222833e-05,
+      "loss": 12.087,
+      "step": 13633
+    },
+    {
+      "epoch": 2.3842185792349726,
+      "grad_norm": 0.8145937092357591,
+      "learning_rate": 2.1284115199359412e-05,
+      "loss": 12.1559,
+      "step": 13634
+    },
+    {
+      "epoch": 2.384393442622951,
+      "grad_norm": 0.9328818518812302,
+      "learning_rate": 2.1272471802619843e-05,
+      "loss": 12.1046,
+      "step": 13635
+    },
+    {
+      "epoch": 2.384568306010929,
+      "grad_norm": 0.8545093332064592,
+      "learning_rate": 2.1260831212424648e-05,
+      "loss": 12.1746,
+      "step": 13636
+    },
+    {
+      "epoch": 2.384743169398907,
+      "grad_norm": 0.7752012647751313,
+      "learning_rate": 2.1249193429188807e-05,
+      "loss": 11.7875,
+      "step": 13637
+    },
+    {
+      "epoch": 2.384918032786885,
+      "grad_norm": 0.8469628107175152,
+      "learning_rate": 2.1237558453327222e-05,
+      "loss": 11.8158,
+      "step": 13638
+    },
+    {
+      "epoch": 2.3850928961748634,
+      "grad_norm": 0.8243071522096116,
+      "learning_rate": 2.122592628525466e-05,
+      "loss": 12.1132,
+      "step": 13639
+    },
+    {
+      "epoch": 2.3852677595628418,
+      "grad_norm": 0.8274521612587858,
+      "learning_rate": 2.121429692538578e-05,
+      "loss": 11.9157,
+      "step": 13640
+    },
+    {
+      "epoch": 2.3854426229508197,
+      "grad_norm": 0.8137643117467167,
+      "learning_rate": 2.1202670374135158e-05,
+      "loss": 12.116,
+      "step": 13641
+    },
+    {
+      "epoch": 2.385617486338798,
+      "grad_norm": 0.847957172561437,
+      "learning_rate": 2.119104663191731e-05,
+      "loss": 12.2286,
+      "step": 13642
+    },
+    {
+      "epoch": 2.385792349726776,
+      "grad_norm": 0.87156627802442,
+      "learning_rate": 2.117942569914654e-05,
+      "loss": 12.0488,
+      "step": 13643
+    },
+    {
+      "epoch": 2.3859672131147542,
+      "grad_norm": 0.8306012857755256,
+      "learning_rate": 2.1167807576237154e-05,
+      "loss": 12.1185,
+      "step": 13644
+    },
+    {
+      "epoch": 2.386142076502732,
+      "grad_norm": 0.8216569179234139,
+      "learning_rate": 2.1156192263603315e-05,
+      "loss": 11.8801,
+      "step": 13645
+    },
+    {
+      "epoch": 2.3863169398907105,
+      "grad_norm": 0.8390434215629656,
+      "learning_rate": 2.1144579761659134e-05,
+      "loss": 11.8826,
+      "step": 13646
+    },
+    {
+      "epoch": 2.3864918032786884,
+      "grad_norm": 0.882943823213301,
+      "learning_rate": 2.1132970070818524e-05,
+      "loss": 12.2613,
+      "step": 13647
+    },
+    {
+      "epoch": 2.3866666666666667,
+      "grad_norm": 0.8250197262003435,
+      "learning_rate": 2.1121363191495392e-05,
+      "loss": 12.0073,
+      "step": 13648
+    },
+    {
+      "epoch": 2.386841530054645,
+      "grad_norm": 0.8841021337081846,
+      "learning_rate": 2.110975912410349e-05,
+      "loss": 12.331,
+      "step": 13649
+    },
+    {
+      "epoch": 2.387016393442623,
+      "grad_norm": 0.8364701371292533,
+      "learning_rate": 2.1098157869056546e-05,
+      "loss": 11.9222,
+      "step": 13650
+    },
+    {
+      "epoch": 2.387191256830601,
+      "grad_norm": 0.8061896333069936,
+      "learning_rate": 2.108655942676805e-05,
+      "loss": 11.9649,
+      "step": 13651
+    },
+    {
+      "epoch": 2.387366120218579,
+      "grad_norm": 0.83825992274517,
+      "learning_rate": 2.1074963797651502e-05,
+      "loss": 12.1471,
+      "step": 13652
+    },
+    {
+      "epoch": 2.3875409836065575,
+      "grad_norm": 0.8294035567220757,
+      "learning_rate": 2.106337098212029e-05,
+      "loss": 11.9154,
+      "step": 13653
+    },
+    {
+      "epoch": 2.3877158469945354,
+      "grad_norm": 0.8101535140423348,
+      "learning_rate": 2.1051780980587677e-05,
+      "loss": 12.0859,
+      "step": 13654
+    },
+    {
+      "epoch": 2.3878907103825138,
+      "grad_norm": 0.8122233330509346,
+      "learning_rate": 2.1040193793466824e-05,
+      "loss": 11.6725,
+      "step": 13655
+    },
+    {
+      "epoch": 2.3880655737704917,
+      "grad_norm": 0.7886708399680648,
+      "learning_rate": 2.1028609421170807e-05,
+      "loss": 12.0561,
+      "step": 13656
+    },
+    {
+      "epoch": 2.38824043715847,
+      "grad_norm": 0.8226066702126613,
+      "learning_rate": 2.1017027864112627e-05,
+      "loss": 12.0389,
+      "step": 13657
+    },
+    {
+      "epoch": 2.388415300546448,
+      "grad_norm": 0.794503382977211,
+      "learning_rate": 2.10054491227051e-05,
+      "loss": 11.9804,
+      "step": 13658
+    },
+    {
+      "epoch": 2.3885901639344262,
+      "grad_norm": 0.8414266543149291,
+      "learning_rate": 2.0993873197361013e-05,
+      "loss": 12.1585,
+      "step": 13659
+    },
+    {
+      "epoch": 2.388765027322404,
+      "grad_norm": 0.8706396420634528,
+      "learning_rate": 2.0982300088493033e-05,
+      "loss": 12.0903,
+      "step": 13660
+    },
+    {
+      "epoch": 2.3889398907103825,
+      "grad_norm": 0.8107823049157017,
+      "learning_rate": 2.0970729796513766e-05,
+      "loss": 12.0517,
+      "step": 13661
+    },
+    {
+      "epoch": 2.389114754098361,
+      "grad_norm": 0.7983891936845315,
+      "learning_rate": 2.0959162321835613e-05,
+      "loss": 12.0513,
+      "step": 13662
+    },
+    {
+      "epoch": 2.3892896174863387,
+      "grad_norm": 0.8840742563479751,
+      "learning_rate": 2.0947597664870978e-05,
+      "loss": 12.0449,
+      "step": 13663
+    },
+    {
+      "epoch": 2.389464480874317,
+      "grad_norm": 0.8547455489704994,
+      "learning_rate": 2.093603582603212e-05,
+      "loss": 12.1986,
+      "step": 13664
+    },
+    {
+      "epoch": 2.389639344262295,
+      "grad_norm": 0.8483243866725483,
+      "learning_rate": 2.092447680573124e-05,
+      "loss": 12.0477,
+      "step": 13665
+    },
+    {
+      "epoch": 2.3898142076502733,
+      "grad_norm": 1.2813961370333748,
+      "learning_rate": 2.0912920604380338e-05,
+      "loss": 12.1467,
+      "step": 13666
+    },
+    {
+      "epoch": 2.389989071038251,
+      "grad_norm": 0.9022921341674176,
+      "learning_rate": 2.090136722239142e-05,
+      "loss": 11.9205,
+      "step": 13667
+    },
+    {
+      "epoch": 2.3901639344262295,
+      "grad_norm": 0.9531074981167296,
+      "learning_rate": 2.0889816660176366e-05,
+      "loss": 12.0091,
+      "step": 13668
+    },
+    {
+      "epoch": 2.3903387978142074,
+      "grad_norm": 0.8529839577025224,
+      "learning_rate": 2.087826891814689e-05,
+      "loss": 11.8907,
+      "step": 13669
+    },
+    {
+      "epoch": 2.390513661202186,
+      "grad_norm": 0.8893513442496023,
+      "learning_rate": 2.0866723996714688e-05,
+      "loss": 12.0817,
+      "step": 13670
+    },
+    {
+      "epoch": 2.390688524590164,
+      "grad_norm": 0.9623291309942105,
+      "learning_rate": 2.085518189629132e-05,
+      "loss": 12.1741,
+      "step": 13671
+    },
+    {
+      "epoch": 2.390863387978142,
+      "grad_norm": 0.8466749100170002,
+      "learning_rate": 2.084364261728824e-05,
+      "loss": 12.1069,
+      "step": 13672
+    },
+    {
+      "epoch": 2.3910382513661204,
+      "grad_norm": 0.8155038739676946,
+      "learning_rate": 2.0832106160116826e-05,
+      "loss": 12.2287,
+      "step": 13673
+    },
+    {
+      "epoch": 2.3912131147540983,
+      "grad_norm": 0.883349496075758,
+      "learning_rate": 2.082057252518832e-05,
+      "loss": 12.0243,
+      "step": 13674
+    },
+    {
+      "epoch": 2.3913879781420766,
+      "grad_norm": 0.8389600369825561,
+      "learning_rate": 2.080904171291389e-05,
+      "loss": 12.1021,
+      "step": 13675
+    },
+    {
+      "epoch": 2.3915628415300545,
+      "grad_norm": 0.8015834827353748,
+      "learning_rate": 2.0797513723704633e-05,
+      "loss": 11.9087,
+      "step": 13676
+    },
+    {
+      "epoch": 2.391737704918033,
+      "grad_norm": 0.9178833403658988,
+      "learning_rate": 2.078598855797145e-05,
+      "loss": 11.9961,
+      "step": 13677
+    },
+    {
+      "epoch": 2.3919125683060107,
+      "grad_norm": 0.8755903353010823,
+      "learning_rate": 2.077446621612521e-05,
+      "loss": 12.141,
+      "step": 13678
+    },
+    {
+      "epoch": 2.392087431693989,
+      "grad_norm": 0.8280639282853189,
+      "learning_rate": 2.076294669857669e-05,
+      "loss": 11.8324,
+      "step": 13679
+    },
+    {
+      "epoch": 2.3922622950819674,
+      "grad_norm": 0.8796901228546548,
+      "learning_rate": 2.0751430005736573e-05,
+      "loss": 11.8646,
+      "step": 13680
+    },
+    {
+      "epoch": 2.3924371584699453,
+      "grad_norm": 0.8069759298453103,
+      "learning_rate": 2.0739916138015357e-05,
+      "loss": 11.8344,
+      "step": 13681
+    },
+    {
+      "epoch": 2.3926120218579237,
+      "grad_norm": 0.8425307078518801,
+      "learning_rate": 2.0728405095823523e-05,
+      "loss": 11.9679,
+      "step": 13682
+    },
+    {
+      "epoch": 2.3927868852459016,
+      "grad_norm": 0.8986331992950226,
+      "learning_rate": 2.0716896879571434e-05,
+      "loss": 12.0496,
+      "step": 13683
+    },
+    {
+      "epoch": 2.39296174863388,
+      "grad_norm": 0.8965670970124886,
+      "learning_rate": 2.0705391489669356e-05,
+      "loss": 11.9582,
+      "step": 13684
+    },
+    {
+      "epoch": 2.393136612021858,
+      "grad_norm": 0.8252475000858733,
+      "learning_rate": 2.0693888926527405e-05,
+      "loss": 11.7784,
+      "step": 13685
+    },
+    {
+      "epoch": 2.393311475409836,
+      "grad_norm": 0.85750760875023,
+      "learning_rate": 2.068238919055565e-05,
+      "loss": 11.8965,
+      "step": 13686
+    },
+    {
+      "epoch": 2.393486338797814,
+      "grad_norm": 0.9258331999778258,
+      "learning_rate": 2.067089228216408e-05,
+      "loss": 11.9715,
+      "step": 13687
+    },
+    {
+      "epoch": 2.3936612021857924,
+      "grad_norm": 0.8200951764008158,
+      "learning_rate": 2.065939820176248e-05,
+      "loss": 11.8548,
+      "step": 13688
+    },
+    {
+      "epoch": 2.3938360655737707,
+      "grad_norm": 0.830357314629034,
+      "learning_rate": 2.064790694976064e-05,
+      "loss": 12.1763,
+      "step": 13689
+    },
+    {
+      "epoch": 2.3940109289617486,
+      "grad_norm": 0.85665433625807,
+      "learning_rate": 2.0636418526568203e-05,
+      "loss": 12.1387,
+      "step": 13690
+    },
+    {
+      "epoch": 2.394185792349727,
+      "grad_norm": 0.8188190416499621,
+      "learning_rate": 2.062493293259472e-05,
+      "loss": 11.6451,
+      "step": 13691
+    },
+    {
+      "epoch": 2.394360655737705,
+      "grad_norm": 0.8361327944995123,
+      "learning_rate": 2.0613450168249636e-05,
+      "loss": 11.8494,
+      "step": 13692
+    },
+    {
+      "epoch": 2.394535519125683,
+      "grad_norm": 0.8022212527461469,
+      "learning_rate": 2.0601970233942302e-05,
+      "loss": 12.0968,
+      "step": 13693
+    },
+    {
+      "epoch": 2.394710382513661,
+      "grad_norm": 0.7853819136542282,
+      "learning_rate": 2.0590493130081956e-05,
+      "loss": 12.1613,
+      "step": 13694
+    },
+    {
+      "epoch": 2.3948852459016394,
+      "grad_norm": 0.8795358135186023,
+      "learning_rate": 2.0579018857077792e-05,
+      "loss": 12.211,
+      "step": 13695
+    },
+    {
+      "epoch": 2.3950601092896173,
+      "grad_norm": 0.8056518513122892,
+      "learning_rate": 2.056754741533877e-05,
+      "loss": 11.8838,
+      "step": 13696
+    },
+    {
+      "epoch": 2.3952349726775957,
+      "grad_norm": 0.8198812535198279,
+      "learning_rate": 2.0556078805273882e-05,
+      "loss": 11.9709,
+      "step": 13697
+    },
+    {
+      "epoch": 2.395409836065574,
+      "grad_norm": 0.8101120787751107,
+      "learning_rate": 2.054461302729197e-05,
+      "loss": 12.1616,
+      "step": 13698
+    },
+    {
+      "epoch": 2.395584699453552,
+      "grad_norm": 0.836065222489115,
+      "learning_rate": 2.053315008180179e-05,
+      "loss": 12.1584,
+      "step": 13699
+    },
+    {
+      "epoch": 2.3957595628415302,
+      "grad_norm": 0.8168498652590676,
+      "learning_rate": 2.052168996921193e-05,
+      "loss": 12.0201,
+      "step": 13700
+    },
+    {
+      "epoch": 2.395934426229508,
+      "grad_norm": 0.8676863079264167,
+      "learning_rate": 2.0510232689930975e-05,
+      "loss": 11.8564,
+      "step": 13701
+    },
+    {
+      "epoch": 2.3961092896174865,
+      "grad_norm": 0.80233357971387,
+      "learning_rate": 2.0498778244367366e-05,
+      "loss": 12.1185,
+      "step": 13702
+    },
+    {
+      "epoch": 2.3962841530054644,
+      "grad_norm": 0.8483230123055863,
+      "learning_rate": 2.0487326632929406e-05,
+      "loss": 11.9216,
+      "step": 13703
+    },
+    {
+      "epoch": 2.3964590163934427,
+      "grad_norm": 0.7792557470262427,
+      "learning_rate": 2.047587785602535e-05,
+      "loss": 11.8552,
+      "step": 13704
+    },
+    {
+      "epoch": 2.3966338797814206,
+      "grad_norm": 0.8107085848690795,
+      "learning_rate": 2.046443191406333e-05,
+      "loss": 11.9773,
+      "step": 13705
+    },
+    {
+      "epoch": 2.396808743169399,
+      "grad_norm": 0.8490959497620556,
+      "learning_rate": 2.0452988807451422e-05,
+      "loss": 12.0743,
+      "step": 13706
+    },
+    {
+      "epoch": 2.396983606557377,
+      "grad_norm": 0.9085434674321157,
+      "learning_rate": 2.044154853659749e-05,
+      "loss": 12.0661,
+      "step": 13707
+    },
+    {
+      "epoch": 2.397158469945355,
+      "grad_norm": 0.873069561087552,
+      "learning_rate": 2.04301111019094e-05,
+      "loss": 12.0068,
+      "step": 13708
+    },
+    {
+      "epoch": 2.397333333333333,
+      "grad_norm": 0.8940394228682992,
+      "learning_rate": 2.0418676503794874e-05,
+      "loss": 11.9836,
+      "step": 13709
+    },
+    {
+      "epoch": 2.3975081967213114,
+      "grad_norm": 0.857033551033228,
+      "learning_rate": 2.0407244742661558e-05,
+      "loss": 12.2432,
+      "step": 13710
+    },
+    {
+      "epoch": 2.3976830601092898,
+      "grad_norm": 0.815022967344167,
+      "learning_rate": 2.0395815818916962e-05,
+      "loss": 11.9338,
+      "step": 13711
+    },
+    {
+      "epoch": 2.3978579234972677,
+      "grad_norm": 0.8298170371429524,
+      "learning_rate": 2.0384389732968534e-05,
+      "loss": 12.0729,
+      "step": 13712
+    },
+    {
+      "epoch": 2.398032786885246,
+      "grad_norm": 0.8856441665277777,
+      "learning_rate": 2.037296648522359e-05,
+      "loss": 12.2103,
+      "step": 13713
+    },
+    {
+      "epoch": 2.398207650273224,
+      "grad_norm": 0.8810574698960607,
+      "learning_rate": 2.0361546076089387e-05,
+      "loss": 12.1481,
+      "step": 13714
+    },
+    {
+      "epoch": 2.3983825136612023,
+      "grad_norm": 0.8359148841864394,
+      "learning_rate": 2.0350128505972986e-05,
+      "loss": 11.9543,
+      "step": 13715
+    },
+    {
+      "epoch": 2.39855737704918,
+      "grad_norm": 0.9436412001234358,
+      "learning_rate": 2.0338713775281448e-05,
+      "loss": 12.3586,
+      "step": 13716
+    },
+    {
+      "epoch": 2.3987322404371585,
+      "grad_norm": 0.8076343062084255,
+      "learning_rate": 2.0327301884421714e-05,
+      "loss": 12.1289,
+      "step": 13717
+    },
+    {
+      "epoch": 2.3989071038251364,
+      "grad_norm": 0.9233115922725631,
+      "learning_rate": 2.0315892833800552e-05,
+      "loss": 12.2078,
+      "step": 13718
+    },
+    {
+      "epoch": 2.3990819672131147,
+      "grad_norm": 0.8076310317025134,
+      "learning_rate": 2.0304486623824702e-05,
+      "loss": 11.8842,
+      "step": 13719
+    },
+    {
+      "epoch": 2.399256830601093,
+      "grad_norm": 0.8121009037221975,
+      "learning_rate": 2.0293083254900804e-05,
+      "loss": 12.0206,
+      "step": 13720
+    },
+    {
+      "epoch": 2.399431693989071,
+      "grad_norm": 0.8642478978755161,
+      "learning_rate": 2.028168272743538e-05,
+      "loss": 12.0958,
+      "step": 13721
+    },
+    {
+      "epoch": 2.3996065573770493,
+      "grad_norm": 0.839715379757018,
+      "learning_rate": 2.0270285041834802e-05,
+      "loss": 11.8484,
+      "step": 13722
+    },
+    {
+      "epoch": 2.399781420765027,
+      "grad_norm": 0.8749631397659922,
+      "learning_rate": 2.025889019850541e-05,
+      "loss": 12.0306,
+      "step": 13723
+    },
+    {
+      "epoch": 2.3999562841530055,
+      "grad_norm": 0.8179194899710117,
+      "learning_rate": 2.024749819785341e-05,
+      "loss": 12.1453,
+      "step": 13724
+    },
+    {
+      "epoch": 2.4001311475409834,
+      "grad_norm": 0.7955341940713958,
+      "learning_rate": 2.0236109040284947e-05,
+      "loss": 11.9397,
+      "step": 13725
+    },
+    {
+      "epoch": 2.400306010928962,
+      "grad_norm": 0.8494591834110702,
+      "learning_rate": 2.022472272620597e-05,
+      "loss": 11.8699,
+      "step": 13726
+    },
+    {
+      "epoch": 2.4004808743169397,
+      "grad_norm": 0.8374352120350232,
+      "learning_rate": 2.0213339256022422e-05,
+      "loss": 12.0803,
+      "step": 13727
+    },
+    {
+      "epoch": 2.400655737704918,
+      "grad_norm": 0.8320500512291247,
+      "learning_rate": 2.0201958630140118e-05,
+      "loss": 12.1427,
+      "step": 13728
+    },
+    {
+      "epoch": 2.4008306010928964,
+      "grad_norm": 0.8326371965584926,
+      "learning_rate": 2.0190580848964746e-05,
+      "loss": 11.8479,
+      "step": 13729
+    },
+    {
+      "epoch": 2.4010054644808743,
+      "grad_norm": 0.854074159237639,
+      "learning_rate": 2.017920591290192e-05,
+      "loss": 11.9289,
+      "step": 13730
+    },
+    {
+      "epoch": 2.4011803278688526,
+      "grad_norm": 0.9341760505526685,
+      "learning_rate": 2.016783382235715e-05,
+      "loss": 12.0151,
+      "step": 13731
+    },
+    {
+      "epoch": 2.4013551912568305,
+      "grad_norm": 0.8384658993649411,
+      "learning_rate": 2.0156464577735858e-05,
+      "loss": 12.0798,
+      "step": 13732
+    },
+    {
+      "epoch": 2.401530054644809,
+      "grad_norm": 0.9194100742286121,
+      "learning_rate": 2.014509817944328e-05,
+      "loss": 11.8682,
+      "step": 13733
+    },
+    {
+      "epoch": 2.4017049180327867,
+      "grad_norm": 0.8193422796322982,
+      "learning_rate": 2.013373462788466e-05,
+      "loss": 11.7828,
+      "step": 13734
+    },
+    {
+      "epoch": 2.401879781420765,
+      "grad_norm": 0.916474894048269,
+      "learning_rate": 2.0122373923465075e-05,
+      "loss": 12.0179,
+      "step": 13735
+    },
+    {
+      "epoch": 2.402054644808743,
+      "grad_norm": 0.8430446098375985,
+      "learning_rate": 2.011101606658957e-05,
+      "loss": 12.1573,
+      "step": 13736
+    },
+    {
+      "epoch": 2.4022295081967213,
+      "grad_norm": 0.8850037663759557,
+      "learning_rate": 2.0099661057662977e-05,
+      "loss": 12.0783,
+      "step": 13737
+    },
+    {
+      "epoch": 2.4024043715846997,
+      "grad_norm": 0.8486839068302077,
+      "learning_rate": 2.00883088970901e-05,
+      "loss": 11.9439,
+      "step": 13738
+    },
+    {
+      "epoch": 2.4025792349726776,
+      "grad_norm": 0.8612586281228095,
+      "learning_rate": 2.0076959585275655e-05,
+      "loss": 12.1416,
+      "step": 13739
+    },
+    {
+      "epoch": 2.402754098360656,
+      "grad_norm": 0.8309431613904699,
+      "learning_rate": 2.006561312262425e-05,
+      "loss": 11.9464,
+      "step": 13740
+    },
+    {
+      "epoch": 2.402928961748634,
+      "grad_norm": 0.8522029966188429,
+      "learning_rate": 2.0054269509540314e-05,
+      "loss": 11.8537,
+      "step": 13741
+    },
+    {
+      "epoch": 2.403103825136612,
+      "grad_norm": 0.8478098057887772,
+      "learning_rate": 2.0042928746428256e-05,
+      "loss": 12.1863,
+      "step": 13742
+    },
+    {
+      "epoch": 2.40327868852459,
+      "grad_norm": 0.7963192399857035,
+      "learning_rate": 2.003159083369238e-05,
+      "loss": 11.8212,
+      "step": 13743
+    },
+    {
+      "epoch": 2.4034535519125684,
+      "grad_norm": 0.8263304605231886,
+      "learning_rate": 2.002025577173685e-05,
+      "loss": 12.1337,
+      "step": 13744
+    },
+    {
+      "epoch": 2.4036284153005463,
+      "grad_norm": 0.8501895278366856,
+      "learning_rate": 2.0008923560965764e-05,
+      "loss": 12.1598,
+      "step": 13745
+    },
+    {
+      "epoch": 2.4038032786885246,
+      "grad_norm": 0.8772501483483334,
+      "learning_rate": 1.999759420178309e-05,
+      "loss": 11.8342,
+      "step": 13746
+    },
+    {
+      "epoch": 2.403978142076503,
+      "grad_norm": 0.8664119192593548,
+      "learning_rate": 1.998626769459273e-05,
+      "loss": 12.1642,
+      "step": 13747
+    },
+    {
+      "epoch": 2.404153005464481,
+      "grad_norm": 0.8165395110985523,
+      "learning_rate": 1.9974944039798417e-05,
+      "loss": 12.1138,
+      "step": 13748
+    },
+    {
+      "epoch": 2.404327868852459,
+      "grad_norm": 0.8199465122949869,
+      "learning_rate": 1.9963623237803852e-05,
+      "loss": 12.1587,
+      "step": 13749
+    },
+    {
+      "epoch": 2.404502732240437,
+      "grad_norm": 0.812089723794043,
+      "learning_rate": 1.9952305289012608e-05,
+      "loss": 11.8114,
+      "step": 13750
+    },
+    {
+      "epoch": 2.4046775956284154,
+      "grad_norm": 0.841361059511196,
+      "learning_rate": 1.994099019382818e-05,
+      "loss": 11.9923,
+      "step": 13751
+    },
+    {
+      "epoch": 2.4048524590163933,
+      "grad_norm": 0.8954361257449956,
+      "learning_rate": 1.992967795265388e-05,
+      "loss": 12.1462,
+      "step": 13752
+    },
+    {
+      "epoch": 2.4050273224043717,
+      "grad_norm": 0.7890348818242796,
+      "learning_rate": 1.991836856589301e-05,
+      "loss": 12.0538,
+      "step": 13753
+    },
+    {
+      "epoch": 2.4052021857923496,
+      "grad_norm": 0.8377443456078362,
+      "learning_rate": 1.9907062033948742e-05,
+      "loss": 12.1362,
+      "step": 13754
+    },
+    {
+      "epoch": 2.405377049180328,
+      "grad_norm": 0.7706433773539286,
+      "learning_rate": 1.989575835722416e-05,
+      "loss": 11.8397,
+      "step": 13755
+    },
+    {
+      "epoch": 2.4055519125683062,
+      "grad_norm": 0.8652650597582738,
+      "learning_rate": 1.988445753612217e-05,
+      "loss": 11.9791,
+      "step": 13756
+    },
+    {
+      "epoch": 2.405726775956284,
+      "grad_norm": 0.9035552094550607,
+      "learning_rate": 1.987315957104566e-05,
+      "loss": 12.0975,
+      "step": 13757
+    },
+    {
+      "epoch": 2.4059016393442625,
+      "grad_norm": 0.8679743707538495,
+      "learning_rate": 1.9861864462397395e-05,
+      "loss": 11.9201,
+      "step": 13758
+    },
+    {
+      "epoch": 2.4060765027322404,
+      "grad_norm": 0.8153838965377661,
+      "learning_rate": 1.9850572210580064e-05,
+      "loss": 12.0691,
+      "step": 13759
+    },
+    {
+      "epoch": 2.4062513661202187,
+      "grad_norm": 0.8697763556401312,
+      "learning_rate": 1.983928281599616e-05,
+      "loss": 12.0836,
+      "step": 13760
+    },
+    {
+      "epoch": 2.4064262295081966,
+      "grad_norm": 0.840957818475048,
+      "learning_rate": 1.9827996279048167e-05,
+      "loss": 12.006,
+      "step": 13761
+    },
+    {
+      "epoch": 2.406601092896175,
+      "grad_norm": 0.9209119280699782,
+      "learning_rate": 1.981671260013843e-05,
+      "loss": 11.9823,
+      "step": 13762
+    },
+    {
+      "epoch": 2.406775956284153,
+      "grad_norm": 0.8701698680695922,
+      "learning_rate": 1.9805431779669213e-05,
+      "loss": 11.84,
+      "step": 13763
+    },
+    {
+      "epoch": 2.406950819672131,
+      "grad_norm": 0.8882727075774265,
+      "learning_rate": 1.9794153818042647e-05,
+      "loss": 12.0496,
+      "step": 13764
+    },
+    {
+      "epoch": 2.407125683060109,
+      "grad_norm": 0.8379047443035091,
+      "learning_rate": 1.978287871566079e-05,
+      "loss": 12.0405,
+      "step": 13765
+    },
+    {
+      "epoch": 2.4073005464480874,
+      "grad_norm": 0.9907562671790476,
+      "learning_rate": 1.9771606472925608e-05,
+      "loss": 12.0416,
+      "step": 13766
+    },
+    {
+      "epoch": 2.4074754098360653,
+      "grad_norm": 0.8649938515143926,
+      "learning_rate": 1.97603370902389e-05,
+      "loss": 11.9509,
+      "step": 13767
+    },
+    {
+      "epoch": 2.4076502732240437,
+      "grad_norm": 0.8380382635698128,
+      "learning_rate": 1.9749070568002414e-05,
+      "loss": 11.6372,
+      "step": 13768
+    },
+    {
+      "epoch": 2.407825136612022,
+      "grad_norm": 0.8548341344348036,
+      "learning_rate": 1.9737806906617808e-05,
+      "loss": 12.027,
+      "step": 13769
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 0.7990905301039185,
+      "learning_rate": 1.972654610648663e-05,
+      "loss": 11.9926,
+      "step": 13770
+    },
+    {
+      "epoch": 2.4081748633879783,
+      "grad_norm": 0.8321308383054767,
+      "learning_rate": 1.971528816801027e-05,
+      "loss": 12.0553,
+      "step": 13771
+    },
+    {
+      "epoch": 2.408349726775956,
+      "grad_norm": 0.9706334265674035,
+      "learning_rate": 1.970403309159009e-05,
+      "loss": 11.9301,
+      "step": 13772
+    },
+    {
+      "epoch": 2.4085245901639345,
+      "grad_norm": 0.8170846465024478,
+      "learning_rate": 1.9692780877627304e-05,
+      "loss": 11.8292,
+      "step": 13773
+    },
+    {
+      "epoch": 2.4086994535519124,
+      "grad_norm": 0.8547110221029416,
+      "learning_rate": 1.968153152652308e-05,
+      "loss": 12.2007,
+      "step": 13774
+    },
+    {
+      "epoch": 2.4088743169398907,
+      "grad_norm": 0.8495707774545492,
+      "learning_rate": 1.9670285038678393e-05,
+      "loss": 12.0952,
+      "step": 13775
+    },
+    {
+      "epoch": 2.4090491803278686,
+      "grad_norm": 0.8251578001401545,
+      "learning_rate": 1.9659041414494185e-05,
+      "loss": 12.1542,
+      "step": 13776
+    },
+    {
+      "epoch": 2.409224043715847,
+      "grad_norm": 0.7909240894924671,
+      "learning_rate": 1.9647800654371307e-05,
+      "loss": 11.9064,
+      "step": 13777
+    },
+    {
+      "epoch": 2.4093989071038253,
+      "grad_norm": 0.8617817028925975,
+      "learning_rate": 1.963656275871043e-05,
+      "loss": 12.0139,
+      "step": 13778
+    },
+    {
+      "epoch": 2.409573770491803,
+      "grad_norm": 0.8904114308672725,
+      "learning_rate": 1.96253277279122e-05,
+      "loss": 11.7339,
+      "step": 13779
+    },
+    {
+      "epoch": 2.4097486338797816,
+      "grad_norm": 0.8654070558948552,
+      "learning_rate": 1.9614095562377123e-05,
+      "loss": 12.0055,
+      "step": 13780
+    },
+    {
+      "epoch": 2.4099234972677595,
+      "grad_norm": 0.8703419617816188,
+      "learning_rate": 1.9602866262505626e-05,
+      "loss": 11.9358,
+      "step": 13781
+    },
+    {
+      "epoch": 2.410098360655738,
+      "grad_norm": 0.842935469185245,
+      "learning_rate": 1.9591639828698005e-05,
+      "loss": 12.019,
+      "step": 13782
+    },
+    {
+      "epoch": 2.4102732240437157,
+      "grad_norm": 0.8360384908424641,
+      "learning_rate": 1.9580416261354483e-05,
+      "loss": 12.0289,
+      "step": 13783
+    },
+    {
+      "epoch": 2.410448087431694,
+      "grad_norm": 0.8621220367520432,
+      "learning_rate": 1.9569195560875153e-05,
+      "loss": 11.9997,
+      "step": 13784
+    },
+    {
+      "epoch": 2.410622950819672,
+      "grad_norm": 0.9005524063832634,
+      "learning_rate": 1.955797772766006e-05,
+      "loss": 11.8652,
+      "step": 13785
+    },
+    {
+      "epoch": 2.4107978142076503,
+      "grad_norm": 0.8797264940821354,
+      "learning_rate": 1.954676276210905e-05,
+      "loss": 11.9498,
+      "step": 13786
+    },
+    {
+      "epoch": 2.4109726775956286,
+      "grad_norm": 0.8545517211163607,
+      "learning_rate": 1.953555066462195e-05,
+      "loss": 11.9021,
+      "step": 13787
+    },
+    {
+      "epoch": 2.4111475409836065,
+      "grad_norm": 0.9019674408207826,
+      "learning_rate": 1.9524341435598447e-05,
+      "loss": 12.2878,
+      "step": 13788
+    },
+    {
+      "epoch": 2.411322404371585,
+      "grad_norm": 0.8025354453190974,
+      "learning_rate": 1.951313507543818e-05,
+      "loss": 11.9816,
+      "step": 13789
+    },
+    {
+      "epoch": 2.4114972677595627,
+      "grad_norm": 0.835681126999471,
+      "learning_rate": 1.950193158454058e-05,
+      "loss": 12.032,
+      "step": 13790
+    },
+    {
+      "epoch": 2.411672131147541,
+      "grad_norm": 0.7929303375344381,
+      "learning_rate": 1.949073096330506e-05,
+      "loss": 12.1572,
+      "step": 13791
+    },
+    {
+      "epoch": 2.411846994535519,
+      "grad_norm": 0.8041362124471634,
+      "learning_rate": 1.947953321213095e-05,
+      "loss": 11.9059,
+      "step": 13792
+    },
+    {
+      "epoch": 2.4120218579234973,
+      "grad_norm": 0.8852895821672665,
+      "learning_rate": 1.946833833141738e-05,
+      "loss": 11.8386,
+      "step": 13793
+    },
+    {
+      "epoch": 2.4121967213114752,
+      "grad_norm": 0.9246858652956352,
+      "learning_rate": 1.9457146321563458e-05,
+      "loss": 12.0114,
+      "step": 13794
+    },
+    {
+      "epoch": 2.4123715846994536,
+      "grad_norm": 0.8918152171598163,
+      "learning_rate": 1.9445957182968155e-05,
+      "loss": 12.1051,
+      "step": 13795
+    },
+    {
+      "epoch": 2.412546448087432,
+      "grad_norm": 0.901918287017431,
+      "learning_rate": 1.9434770916030388e-05,
+      "loss": 11.9055,
+      "step": 13796
+    },
+    {
+      "epoch": 2.41272131147541,
+      "grad_norm": 0.8763807209507082,
+      "learning_rate": 1.942358752114889e-05,
+      "loss": 12.1735,
+      "step": 13797
+    },
+    {
+      "epoch": 2.412896174863388,
+      "grad_norm": 0.861892916909663,
+      "learning_rate": 1.9412406998722356e-05,
+      "loss": 11.9292,
+      "step": 13798
+    },
+    {
+      "epoch": 2.413071038251366,
+      "grad_norm": 0.7919689416439095,
+      "learning_rate": 1.940122934914934e-05,
+      "loss": 12.1403,
+      "step": 13799
+    },
+    {
+      "epoch": 2.4132459016393444,
+      "grad_norm": 0.8438361695127756,
+      "learning_rate": 1.939005457282834e-05,
+      "loss": 12.1637,
+      "step": 13800
+    },
+    {
+      "epoch": 2.4134207650273223,
+      "grad_norm": 0.8785991665250827,
+      "learning_rate": 1.9378882670157715e-05,
+      "loss": 12.004,
+      "step": 13801
+    },
+    {
+      "epoch": 2.4135956284153006,
+      "grad_norm": 1.0193875062833269,
+      "learning_rate": 1.9367713641535724e-05,
+      "loss": 12.2028,
+      "step": 13802
+    },
+    {
+      "epoch": 2.4137704918032785,
+      "grad_norm": 0.8060606417765401,
+      "learning_rate": 1.935654748736052e-05,
+      "loss": 12.1221,
+      "step": 13803
+    },
+    {
+      "epoch": 2.413945355191257,
+      "grad_norm": 0.8701686420479509,
+      "learning_rate": 1.9345384208030213e-05,
+      "loss": 12.1883,
+      "step": 13804
+    },
+    {
+      "epoch": 2.414120218579235,
+      "grad_norm": 0.8861087961857332,
+      "learning_rate": 1.933422380394271e-05,
+      "loss": 11.8969,
+      "step": 13805
+    },
+    {
+      "epoch": 2.414295081967213,
+      "grad_norm": 0.79678916154649,
+      "learning_rate": 1.9323066275495862e-05,
+      "loss": 11.971,
+      "step": 13806
+    },
+    {
+      "epoch": 2.4144699453551914,
+      "grad_norm": 0.8625942333689921,
+      "learning_rate": 1.9311911623087442e-05,
+      "loss": 12.1035,
+      "step": 13807
+    },
+    {
+      "epoch": 2.4146448087431693,
+      "grad_norm": 0.8168355259153656,
+      "learning_rate": 1.9300759847115134e-05,
+      "loss": 11.905,
+      "step": 13808
+    },
+    {
+      "epoch": 2.4148196721311477,
+      "grad_norm": 0.8551453884913569,
+      "learning_rate": 1.928961094797642e-05,
+      "loss": 12.1433,
+      "step": 13809
+    },
+    {
+      "epoch": 2.4149945355191256,
+      "grad_norm": 0.8330724693807107,
+      "learning_rate": 1.927846492606877e-05,
+      "loss": 12.0136,
+      "step": 13810
+    },
+    {
+      "epoch": 2.415169398907104,
+      "grad_norm": 0.8542873618890862,
+      "learning_rate": 1.9267321781789572e-05,
+      "loss": 11.76,
+      "step": 13811
+    },
+    {
+      "epoch": 2.415344262295082,
+      "grad_norm": 0.8635024818553754,
+      "learning_rate": 1.9256181515535997e-05,
+      "loss": 11.9301,
+      "step": 13812
+    },
+    {
+      "epoch": 2.41551912568306,
+      "grad_norm": 0.8593564069356119,
+      "learning_rate": 1.9245044127705202e-05,
+      "loss": 11.8204,
+      "step": 13813
+    },
+    {
+      "epoch": 2.4156939890710385,
+      "grad_norm": 0.8072916621409845,
+      "learning_rate": 1.9233909618694247e-05,
+      "loss": 12.0075,
+      "step": 13814
+    },
+    {
+      "epoch": 2.4158688524590164,
+      "grad_norm": 0.8068639077754076,
+      "learning_rate": 1.9222777988900065e-05,
+      "loss": 11.958,
+      "step": 13815
+    },
+    {
+      "epoch": 2.4160437158469943,
+      "grad_norm": 0.7996612599656643,
+      "learning_rate": 1.9211649238719443e-05,
+      "loss": 11.9407,
+      "step": 13816
+    },
+    {
+      "epoch": 2.4162185792349726,
+      "grad_norm": 0.8591018995984762,
+      "learning_rate": 1.9200523368549128e-05,
+      "loss": 11.7789,
+      "step": 13817
+    },
+    {
+      "epoch": 2.416393442622951,
+      "grad_norm": 0.8750061576223885,
+      "learning_rate": 1.918940037878576e-05,
+      "loss": 11.9395,
+      "step": 13818
+    },
+    {
+      "epoch": 2.416568306010929,
+      "grad_norm": 0.799712026286618,
+      "learning_rate": 1.9178280269825853e-05,
+      "loss": 11.8531,
+      "step": 13819
+    },
+    {
+      "epoch": 2.416743169398907,
+      "grad_norm": 0.8388890103523564,
+      "learning_rate": 1.916716304206583e-05,
+      "loss": 12.1568,
+      "step": 13820
+    },
+    {
+      "epoch": 2.416918032786885,
+      "grad_norm": 0.8731387007613259,
+      "learning_rate": 1.9156048695901984e-05,
+      "loss": 11.9349,
+      "step": 13821
+    },
+    {
+      "epoch": 2.4170928961748634,
+      "grad_norm": 0.9419086530499051,
+      "learning_rate": 1.9144937231730554e-05,
+      "loss": 11.9205,
+      "step": 13822
+    },
+    {
+      "epoch": 2.4172677595628413,
+      "grad_norm": 0.8657056518415698,
+      "learning_rate": 1.9133828649947672e-05,
+      "loss": 11.9219,
+      "step": 13823
+    },
+    {
+      "epoch": 2.4174426229508197,
+      "grad_norm": 0.8209276058494246,
+      "learning_rate": 1.912272295094929e-05,
+      "loss": 11.9395,
+      "step": 13824
+    },
+    {
+      "epoch": 2.4176174863387976,
+      "grad_norm": 0.8553956661990318,
+      "learning_rate": 1.9111620135131337e-05,
+      "loss": 12.1747,
+      "step": 13825
+    },
+    {
+      "epoch": 2.417792349726776,
+      "grad_norm": 0.876981310173566,
+      "learning_rate": 1.910052020288965e-05,
+      "loss": 11.8563,
+      "step": 13826
+    },
+    {
+      "epoch": 2.4179672131147543,
+      "grad_norm": 1.1534435544280681,
+      "learning_rate": 1.9089423154619868e-05,
+      "loss": 11.9834,
+      "step": 13827
+    },
+    {
+      "epoch": 2.418142076502732,
+      "grad_norm": 0.8446061107305469,
+      "learning_rate": 1.907832899071761e-05,
+      "loss": 12.013,
+      "step": 13828
+    },
+    {
+      "epoch": 2.4183169398907105,
+      "grad_norm": 0.8478638181374288,
+      "learning_rate": 1.906723771157838e-05,
+      "loss": 11.9066,
+      "step": 13829
+    },
+    {
+      "epoch": 2.4184918032786884,
+      "grad_norm": 0.8020962230052466,
+      "learning_rate": 1.9056149317597595e-05,
+      "loss": 12.1505,
+      "step": 13830
+    },
+    {
+      "epoch": 2.4186666666666667,
+      "grad_norm": 0.9080808949839093,
+      "learning_rate": 1.9045063809170483e-05,
+      "loss": 11.9303,
+      "step": 13831
+    },
+    {
+      "epoch": 2.4188415300546446,
+      "grad_norm": 0.8303396581855162,
+      "learning_rate": 1.9033981186692274e-05,
+      "loss": 11.8552,
+      "step": 13832
+    },
+    {
+      "epoch": 2.419016393442623,
+      "grad_norm": 0.8470705587682364,
+      "learning_rate": 1.9022901450558017e-05,
+      "loss": 12.1034,
+      "step": 13833
+    },
+    {
+      "epoch": 2.419191256830601,
+      "grad_norm": 0.8390964224206198,
+      "learning_rate": 1.901182460116273e-05,
+      "loss": 12.0468,
+      "step": 13834
+    },
+    {
+      "epoch": 2.419366120218579,
+      "grad_norm": 0.8559439080347737,
+      "learning_rate": 1.9000750638901288e-05,
+      "loss": 12.0075,
+      "step": 13835
+    },
+    {
+      "epoch": 2.4195409836065576,
+      "grad_norm": 0.8872726414393268,
+      "learning_rate": 1.8989679564168427e-05,
+      "loss": 12.1009,
+      "step": 13836
+    },
+    {
+      "epoch": 2.4197158469945355,
+      "grad_norm": 0.8637500881445949,
+      "learning_rate": 1.8978611377358834e-05,
+      "loss": 11.9173,
+      "step": 13837
+    },
+    {
+      "epoch": 2.419890710382514,
+      "grad_norm": 0.8077589706989823,
+      "learning_rate": 1.8967546078867082e-05,
+      "loss": 12.0513,
+      "step": 13838
+    },
+    {
+      "epoch": 2.4200655737704917,
+      "grad_norm": 0.8360091950471201,
+      "learning_rate": 1.8956483669087643e-05,
+      "loss": 11.7335,
+      "step": 13839
+    },
+    {
+      "epoch": 2.42024043715847,
+      "grad_norm": 0.8097164985110059,
+      "learning_rate": 1.894542414841486e-05,
+      "loss": 11.9711,
+      "step": 13840
+    },
+    {
+      "epoch": 2.420415300546448,
+      "grad_norm": 0.8889638010292313,
+      "learning_rate": 1.8934367517243033e-05,
+      "loss": 11.9643,
+      "step": 13841
+    },
+    {
+      "epoch": 2.4205901639344263,
+      "grad_norm": 0.7818624673898266,
+      "learning_rate": 1.8923313775966278e-05,
+      "loss": 12.1459,
+      "step": 13842
+    },
+    {
+      "epoch": 2.420765027322404,
+      "grad_norm": 0.9174065971211713,
+      "learning_rate": 1.891226292497864e-05,
+      "loss": 12.0849,
+      "step": 13843
+    },
+    {
+      "epoch": 2.4209398907103825,
+      "grad_norm": 0.8740452873515209,
+      "learning_rate": 1.8901214964674097e-05,
+      "loss": 11.7382,
+      "step": 13844
+    },
+    {
+      "epoch": 2.421114754098361,
+      "grad_norm": 0.8871083574091639,
+      "learning_rate": 1.8890169895446517e-05,
+      "loss": 12.0253,
+      "step": 13845
+    },
+    {
+      "epoch": 2.4212896174863388,
+      "grad_norm": 0.8886674005604035,
+      "learning_rate": 1.887912771768958e-05,
+      "loss": 12.0276,
+      "step": 13846
+    },
+    {
+      "epoch": 2.421464480874317,
+      "grad_norm": 0.8741345928836088,
+      "learning_rate": 1.886808843179696e-05,
+      "loss": 12.0959,
+      "step": 13847
+    },
+    {
+      "epoch": 2.421639344262295,
+      "grad_norm": 0.8308528806526491,
+      "learning_rate": 1.8857052038162204e-05,
+      "loss": 11.9453,
+      "step": 13848
+    },
+    {
+      "epoch": 2.4218142076502733,
+      "grad_norm": 0.9065650488326357,
+      "learning_rate": 1.8846018537178756e-05,
+      "loss": 12.0327,
+      "step": 13849
+    },
+    {
+      "epoch": 2.4219890710382512,
+      "grad_norm": 0.8526588634788543,
+      "learning_rate": 1.88349879292399e-05,
+      "loss": 12.0811,
+      "step": 13850
+    },
+    {
+      "epoch": 2.4221639344262296,
+      "grad_norm": 0.9706933655654693,
+      "learning_rate": 1.88239602147389e-05,
+      "loss": 11.8784,
+      "step": 13851
+    },
+    {
+      "epoch": 2.4223387978142075,
+      "grad_norm": 0.848873998394952,
+      "learning_rate": 1.8812935394068865e-05,
+      "loss": 12.2587,
+      "step": 13852
+    },
+    {
+      "epoch": 2.422513661202186,
+      "grad_norm": 0.917818188641187,
+      "learning_rate": 1.8801913467622834e-05,
+      "loss": 11.9772,
+      "step": 13853
+    },
+    {
+      "epoch": 2.422688524590164,
+      "grad_norm": 0.8580834584277807,
+      "learning_rate": 1.8790894435793715e-05,
+      "loss": 12.1208,
+      "step": 13854
+    },
+    {
+      "epoch": 2.422863387978142,
+      "grad_norm": 0.9131275370074143,
+      "learning_rate": 1.877987829897433e-05,
+      "loss": 12.0657,
+      "step": 13855
+    },
+    {
+      "epoch": 2.4230382513661204,
+      "grad_norm": 0.8301331248567826,
+      "learning_rate": 1.8768865057557417e-05,
+      "loss": 12.2033,
+      "step": 13856
+    },
+    {
+      "epoch": 2.4232131147540983,
+      "grad_norm": 0.8892864881512925,
+      "learning_rate": 1.875785471193552e-05,
+      "loss": 12.0116,
+      "step": 13857
+    },
+    {
+      "epoch": 2.4233879781420766,
+      "grad_norm": 0.7917885711132592,
+      "learning_rate": 1.874684726250119e-05,
+      "loss": 11.7844,
+      "step": 13858
+    },
+    {
+      "epoch": 2.4235628415300545,
+      "grad_norm": 0.854809019984679,
+      "learning_rate": 1.873584270964682e-05,
+      "loss": 12.0252,
+      "step": 13859
+    },
+    {
+      "epoch": 2.423737704918033,
+      "grad_norm": 0.8609563740837705,
+      "learning_rate": 1.872484105376473e-05,
+      "loss": 12.0288,
+      "step": 13860
+    },
+    {
+      "epoch": 2.4239125683060108,
+      "grad_norm": 0.7988518684488783,
+      "learning_rate": 1.8713842295247075e-05,
+      "loss": 12.0101,
+      "step": 13861
+    },
+    {
+      "epoch": 2.424087431693989,
+      "grad_norm": 0.8582239780078201,
+      "learning_rate": 1.8702846434485976e-05,
+      "loss": 12.1799,
+      "step": 13862
+    },
+    {
+      "epoch": 2.4242622950819674,
+      "grad_norm": 0.8718709706645436,
+      "learning_rate": 1.869185347187341e-05,
+      "loss": 12.02,
+      "step": 13863
+    },
+    {
+      "epoch": 2.4244371584699453,
+      "grad_norm": 0.8410314467154411,
+      "learning_rate": 1.8680863407801297e-05,
+      "loss": 11.9311,
+      "step": 13864
+    },
+    {
+      "epoch": 2.4246120218579237,
+      "grad_norm": 0.91050099434597,
+      "learning_rate": 1.8669876242661356e-05,
+      "loss": 12.1875,
+      "step": 13865
+    },
+    {
+      "epoch": 2.4247868852459016,
+      "grad_norm": 0.776145235560072,
+      "learning_rate": 1.8658891976845318e-05,
+      "loss": 12.0477,
+      "step": 13866
+    },
+    {
+      "epoch": 2.42496174863388,
+      "grad_norm": 0.8142716866867595,
+      "learning_rate": 1.8647910610744733e-05,
+      "loss": 11.9475,
+      "step": 13867
+    },
+    {
+      "epoch": 2.425136612021858,
+      "grad_norm": 0.8901122614769386,
+      "learning_rate": 1.863693214475112e-05,
+      "loss": 12.0533,
+      "step": 13868
+    },
+    {
+      "epoch": 2.425311475409836,
+      "grad_norm": 0.8448005498468737,
+      "learning_rate": 1.8625956579255777e-05,
+      "loss": 12.0283,
+      "step": 13869
+    },
+    {
+      "epoch": 2.425486338797814,
+      "grad_norm": 0.818083861090633,
+      "learning_rate": 1.861498391465002e-05,
+      "loss": 12.0067,
+      "step": 13870
+    },
+    {
+      "epoch": 2.4256612021857924,
+      "grad_norm": 0.7996305607581397,
+      "learning_rate": 1.860401415132499e-05,
+      "loss": 12.0634,
+      "step": 13871
+    },
+    {
+      "epoch": 2.4258360655737707,
+      "grad_norm": 0.822404120353225,
+      "learning_rate": 1.859304728967175e-05,
+      "loss": 12.2239,
+      "step": 13872
+    },
+    {
+      "epoch": 2.4260109289617486,
+      "grad_norm": 0.8361868112993157,
+      "learning_rate": 1.858208333008127e-05,
+      "loss": 12.1157,
+      "step": 13873
+    },
+    {
+      "epoch": 2.4261857923497265,
+      "grad_norm": 0.9509393693958963,
+      "learning_rate": 1.8571122272944386e-05,
+      "loss": 11.9919,
+      "step": 13874
+    },
+    {
+      "epoch": 2.426360655737705,
+      "grad_norm": 0.8202483799749206,
+      "learning_rate": 1.8560164118651878e-05,
+      "loss": 12.0193,
+      "step": 13875
+    },
+    {
+      "epoch": 2.426535519125683,
+      "grad_norm": 0.848098195927585,
+      "learning_rate": 1.8549208867594347e-05,
+      "loss": 12.1177,
+      "step": 13876
+    },
+    {
+      "epoch": 2.426710382513661,
+      "grad_norm": 0.8732014360472141,
+      "learning_rate": 1.8538256520162355e-05,
+      "loss": 12.0842,
+      "step": 13877
+    },
+    {
+      "epoch": 2.4268852459016395,
+      "grad_norm": 0.829708895621316,
+      "learning_rate": 1.8527307076746335e-05,
+      "loss": 12.0138,
+      "step": 13878
+    },
+    {
+      "epoch": 2.4270601092896174,
+      "grad_norm": 0.8173595426201441,
+      "learning_rate": 1.8516360537736656e-05,
+      "loss": 12.0524,
+      "step": 13879
+    },
+    {
+      "epoch": 2.4272349726775957,
+      "grad_norm": 0.8579870547053039,
+      "learning_rate": 1.8505416903523497e-05,
+      "loss": 12.1558,
+      "step": 13880
+    },
+    {
+      "epoch": 2.4274098360655736,
+      "grad_norm": 0.8206062515856097,
+      "learning_rate": 1.8494476174496998e-05,
+      "loss": 11.8712,
+      "step": 13881
+    },
+    {
+      "epoch": 2.427584699453552,
+      "grad_norm": 0.8212550722060338,
+      "learning_rate": 1.8483538351047215e-05,
+      "loss": 11.946,
+      "step": 13882
+    },
+    {
+      "epoch": 2.42775956284153,
+      "grad_norm": 0.8794638539610143,
+      "learning_rate": 1.847260343356406e-05,
+      "loss": 12.1517,
+      "step": 13883
+    },
+    {
+      "epoch": 2.427934426229508,
+      "grad_norm": 0.8811788729702089,
+      "learning_rate": 1.846167142243731e-05,
+      "loss": 12.0879,
+      "step": 13884
+    },
+    {
+      "epoch": 2.4281092896174865,
+      "grad_norm": 0.8967910027561229,
+      "learning_rate": 1.8450742318056712e-05,
+      "loss": 11.9829,
+      "step": 13885
+    },
+    {
+      "epoch": 2.4282841530054644,
+      "grad_norm": 0.8874217476545195,
+      "learning_rate": 1.8439816120811892e-05,
+      "loss": 11.9905,
+      "step": 13886
+    },
+    {
+      "epoch": 2.4284590163934427,
+      "grad_norm": 0.8591727684584636,
+      "learning_rate": 1.8428892831092316e-05,
+      "loss": 12.1465,
+      "step": 13887
+    },
+    {
+      "epoch": 2.4286338797814206,
+      "grad_norm": 0.8919870506830135,
+      "learning_rate": 1.8417972449287403e-05,
+      "loss": 11.8445,
+      "step": 13888
+    },
+    {
+      "epoch": 2.428808743169399,
+      "grad_norm": 0.8500016755679347,
+      "learning_rate": 1.840705497578644e-05,
+      "loss": 12.1295,
+      "step": 13889
+    },
+    {
+      "epoch": 2.428983606557377,
+      "grad_norm": 0.9819465225688039,
+      "learning_rate": 1.8396140410978656e-05,
+      "loss": 12.2009,
+      "step": 13890
+    },
+    {
+      "epoch": 2.4291584699453552,
+      "grad_norm": 0.8206853610826271,
+      "learning_rate": 1.8385228755253114e-05,
+      "loss": 12.1292,
+      "step": 13891
+    },
+    {
+      "epoch": 2.429333333333333,
+      "grad_norm": 0.8024099652857772,
+      "learning_rate": 1.8374320008998803e-05,
+      "loss": 12.0163,
+      "step": 13892
+    },
+    {
+      "epoch": 2.4295081967213115,
+      "grad_norm": 0.818921688031067,
+      "learning_rate": 1.836341417260462e-05,
+      "loss": 11.9346,
+      "step": 13893
+    },
+    {
+      "epoch": 2.42968306010929,
+      "grad_norm": 0.8188945044580852,
+      "learning_rate": 1.835251124645937e-05,
+      "loss": 11.9119,
+      "step": 13894
+    },
+    {
+      "epoch": 2.4298579234972677,
+      "grad_norm": 0.8196213497057803,
+      "learning_rate": 1.8341611230951672e-05,
+      "loss": 12.0892,
+      "step": 13895
+    },
+    {
+      "epoch": 2.430032786885246,
+      "grad_norm": 0.8168979156170739,
+      "learning_rate": 1.833071412647013e-05,
+      "loss": 11.817,
+      "step": 13896
+    },
+    {
+      "epoch": 2.430207650273224,
+      "grad_norm": 0.8959234631631986,
+      "learning_rate": 1.831981993340319e-05,
+      "loss": 12.0818,
+      "step": 13897
+    },
+    {
+      "epoch": 2.4303825136612023,
+      "grad_norm": 0.8470157609293208,
+      "learning_rate": 1.830892865213928e-05,
+      "loss": 11.9987,
+      "step": 13898
+    },
+    {
+      "epoch": 2.43055737704918,
+      "grad_norm": 0.8508074804234325,
+      "learning_rate": 1.8298040283066586e-05,
+      "loss": 11.8597,
+      "step": 13899
+    },
+    {
+      "epoch": 2.4307322404371585,
+      "grad_norm": 0.8626386366020571,
+      "learning_rate": 1.8287154826573304e-05,
+      "loss": 11.9736,
+      "step": 13900
+    },
+    {
+      "epoch": 2.4309071038251364,
+      "grad_norm": 0.8685589208160558,
+      "learning_rate": 1.82762722830475e-05,
+      "loss": 12.1453,
+      "step": 13901
+    },
+    {
+      "epoch": 2.4310819672131148,
+      "grad_norm": 0.9201478396056914,
+      "learning_rate": 1.826539265287708e-05,
+      "loss": 11.9724,
+      "step": 13902
+    },
+    {
+      "epoch": 2.431256830601093,
+      "grad_norm": 0.8387181518296628,
+      "learning_rate": 1.8254515936449913e-05,
+      "loss": 11.9774,
+      "step": 13903
+    },
+    {
+      "epoch": 2.431431693989071,
+      "grad_norm": 0.8380453122757678,
+      "learning_rate": 1.824364213415375e-05,
+      "loss": 11.8655,
+      "step": 13904
+    },
+    {
+      "epoch": 2.4316065573770493,
+      "grad_norm": 0.818929871185808,
+      "learning_rate": 1.8232771246376246e-05,
+      "loss": 11.8502,
+      "step": 13905
+    },
+    {
+      "epoch": 2.4317814207650272,
+      "grad_norm": 0.9193060233365565,
+      "learning_rate": 1.8221903273504883e-05,
+      "loss": 12.0809,
+      "step": 13906
+    },
+    {
+      "epoch": 2.4319562841530056,
+      "grad_norm": 0.8512316661516591,
+      "learning_rate": 1.8211038215927122e-05,
+      "loss": 12.0677,
+      "step": 13907
+    },
+    {
+      "epoch": 2.4321311475409835,
+      "grad_norm": 0.7858441313762179,
+      "learning_rate": 1.8200176074030297e-05,
+      "loss": 11.9553,
+      "step": 13908
+    },
+    {
+      "epoch": 2.432306010928962,
+      "grad_norm": 0.88975376908314,
+      "learning_rate": 1.8189316848201608e-05,
+      "loss": 12.0404,
+      "step": 13909
+    },
+    {
+      "epoch": 2.4324808743169397,
+      "grad_norm": 0.807281318700732,
+      "learning_rate": 1.81784605388282e-05,
+      "loss": 11.8425,
+      "step": 13910
+    },
+    {
+      "epoch": 2.432655737704918,
+      "grad_norm": 0.9037608340511751,
+      "learning_rate": 1.8167607146297072e-05,
+      "loss": 12.1238,
+      "step": 13911
+    },
+    {
+      "epoch": 2.4328306010928964,
+      "grad_norm": 0.8707980538420368,
+      "learning_rate": 1.8156756670995135e-05,
+      "loss": 12.0363,
+      "step": 13912
+    },
+    {
+      "epoch": 2.4330054644808743,
+      "grad_norm": 0.8258591893858824,
+      "learning_rate": 1.8145909113309223e-05,
+      "loss": 11.9551,
+      "step": 13913
+    },
+    {
+      "epoch": 2.4331803278688526,
+      "grad_norm": 0.8010073177278266,
+      "learning_rate": 1.8135064473625994e-05,
+      "loss": 11.9526,
+      "step": 13914
+    },
+    {
+      "epoch": 2.4333551912568305,
+      "grad_norm": 0.9040497809931494,
+      "learning_rate": 1.8124222752332077e-05,
+      "loss": 11.9037,
+      "step": 13915
+    },
+    {
+      "epoch": 2.433530054644809,
+      "grad_norm": 0.8920975286999033,
+      "learning_rate": 1.8113383949813943e-05,
+      "loss": 12.0053,
+      "step": 13916
+    },
+    {
+      "epoch": 2.4337049180327868,
+      "grad_norm": 0.8709398314796433,
+      "learning_rate": 1.8102548066458025e-05,
+      "loss": 12.245,
+      "step": 13917
+    },
+    {
+      "epoch": 2.433879781420765,
+      "grad_norm": 0.8338369896226462,
+      "learning_rate": 1.8091715102650553e-05,
+      "loss": 11.8797,
+      "step": 13918
+    },
+    {
+      "epoch": 2.434054644808743,
+      "grad_norm": 0.7806957917244427,
+      "learning_rate": 1.808088505877775e-05,
+      "loss": 11.9283,
+      "step": 13919
+    },
+    {
+      "epoch": 2.4342295081967213,
+      "grad_norm": 0.8109338271663377,
+      "learning_rate": 1.8070057935225693e-05,
+      "loss": 11.633,
+      "step": 13920
+    },
+    {
+      "epoch": 2.4344043715846997,
+      "grad_norm": 0.9438830335391467,
+      "learning_rate": 1.805923373238033e-05,
+      "loss": 11.9207,
+      "step": 13921
+    },
+    {
+      "epoch": 2.4345792349726776,
+      "grad_norm": 0.8561464267244093,
+      "learning_rate": 1.804841245062755e-05,
+      "loss": 12.0875,
+      "step": 13922
+    },
+    {
+      "epoch": 2.434754098360656,
+      "grad_norm": 0.8351958003951419,
+      "learning_rate": 1.8037594090353105e-05,
+      "loss": 12.1532,
+      "step": 13923
+    },
+    {
+      "epoch": 2.434928961748634,
+      "grad_norm": 0.8833529584290901,
+      "learning_rate": 1.80267786519427e-05,
+      "loss": 11.9422,
+      "step": 13924
+    },
+    {
+      "epoch": 2.435103825136612,
+      "grad_norm": 0.8614980346526454,
+      "learning_rate": 1.8015966135781836e-05,
+      "loss": 11.9625,
+      "step": 13925
+    },
+    {
+      "epoch": 2.43527868852459,
+      "grad_norm": 0.8536829337174795,
+      "learning_rate": 1.8005156542255986e-05,
+      "loss": 11.7905,
+      "step": 13926
+    },
+    {
+      "epoch": 2.4354535519125684,
+      "grad_norm": 0.8112281710364591,
+      "learning_rate": 1.7994349871750516e-05,
+      "loss": 12.0643,
+      "step": 13927
+    },
+    {
+      "epoch": 2.4356284153005463,
+      "grad_norm": 0.886982050565766,
+      "learning_rate": 1.7983546124650653e-05,
+      "loss": 12.1878,
+      "step": 13928
+    },
+    {
+      "epoch": 2.4358032786885246,
+      "grad_norm": 0.9026833549643519,
+      "learning_rate": 1.7972745301341544e-05,
+      "loss": 12.1293,
+      "step": 13929
+    },
+    {
+      "epoch": 2.4359781420765025,
+      "grad_norm": 0.8818024277411237,
+      "learning_rate": 1.7961947402208225e-05,
+      "loss": 12.0905,
+      "step": 13930
+    },
+    {
+      "epoch": 2.436153005464481,
+      "grad_norm": 0.8096406671867207,
+      "learning_rate": 1.7951152427635633e-05,
+      "loss": 12.2285,
+      "step": 13931
+    },
+    {
+      "epoch": 2.4363278688524588,
+      "grad_norm": 0.8193955052875594,
+      "learning_rate": 1.7940360378008615e-05,
+      "loss": 11.7984,
+      "step": 13932
+    },
+    {
+      "epoch": 2.436502732240437,
+      "grad_norm": 0.8826149123961662,
+      "learning_rate": 1.7929571253711842e-05,
+      "loss": 12.0135,
+      "step": 13933
+    },
+    {
+      "epoch": 2.4366775956284155,
+      "grad_norm": 0.8137550844346492,
+      "learning_rate": 1.791878505512997e-05,
+      "loss": 11.7698,
+      "step": 13934
+    },
+    {
+      "epoch": 2.4368524590163934,
+      "grad_norm": 0.842116588017146,
+      "learning_rate": 1.790800178264754e-05,
+      "loss": 12.2534,
+      "step": 13935
+    },
+    {
+      "epoch": 2.4370273224043717,
+      "grad_norm": 0.8517075366880708,
+      "learning_rate": 1.7897221436648902e-05,
+      "loss": 12.035,
+      "step": 13936
+    },
+    {
+      "epoch": 2.4372021857923496,
+      "grad_norm": 0.8220820271653508,
+      "learning_rate": 1.78864440175184e-05,
+      "loss": 11.8556,
+      "step": 13937
+    },
+    {
+      "epoch": 2.437377049180328,
+      "grad_norm": 0.8601491251309403,
+      "learning_rate": 1.7875669525640225e-05,
+      "loss": 11.957,
+      "step": 13938
+    },
+    {
+      "epoch": 2.437551912568306,
+      "grad_norm": 0.8691128360234954,
+      "learning_rate": 1.7864897961398496e-05,
+      "loss": 12.1673,
+      "step": 13939
+    },
+    {
+      "epoch": 2.437726775956284,
+      "grad_norm": 0.8332273463973718,
+      "learning_rate": 1.785412932517717e-05,
+      "loss": 12.1044,
+      "step": 13940
+    },
+    {
+      "epoch": 2.437901639344262,
+      "grad_norm": 0.8619783367621605,
+      "learning_rate": 1.7843363617360165e-05,
+      "loss": 12.0346,
+      "step": 13941
+    },
+    {
+      "epoch": 2.4380765027322404,
+      "grad_norm": 0.89226645890678,
+      "learning_rate": 1.783260083833125e-05,
+      "loss": 11.8892,
+      "step": 13942
+    },
+    {
+      "epoch": 2.4382513661202188,
+      "grad_norm": 0.880849638150434,
+      "learning_rate": 1.7821840988474116e-05,
+      "loss": 11.9512,
+      "step": 13943
+    },
+    {
+      "epoch": 2.4384262295081967,
+      "grad_norm": 0.8642966250219023,
+      "learning_rate": 1.7811084068172335e-05,
+      "loss": 11.9879,
+      "step": 13944
+    },
+    {
+      "epoch": 2.438601092896175,
+      "grad_norm": 0.8116870579370996,
+      "learning_rate": 1.780033007780938e-05,
+      "loss": 12.2012,
+      "step": 13945
+    },
+    {
+      "epoch": 2.438775956284153,
+      "grad_norm": 0.868755803024688,
+      "learning_rate": 1.778957901776861e-05,
+      "loss": 12.0288,
+      "step": 13946
+    },
+    {
+      "epoch": 2.4389508196721312,
+      "grad_norm": 0.834454193152602,
+      "learning_rate": 1.7778830888433328e-05,
+      "loss": 12.2382,
+      "step": 13947
+    },
+    {
+      "epoch": 2.439125683060109,
+      "grad_norm": 0.7999799226670318,
+      "learning_rate": 1.7768085690186632e-05,
+      "loss": 12.1765,
+      "step": 13948
+    },
+    {
+      "epoch": 2.4393005464480875,
+      "grad_norm": 0.8404628798852937,
+      "learning_rate": 1.775734342341161e-05,
+      "loss": 12.0226,
+      "step": 13949
+    },
+    {
+      "epoch": 2.4394754098360654,
+      "grad_norm": 0.7816169169041121,
+      "learning_rate": 1.7746604088491224e-05,
+      "loss": 11.9936,
+      "step": 13950
+    },
+    {
+      "epoch": 2.4396502732240437,
+      "grad_norm": 0.8671896239935304,
+      "learning_rate": 1.7735867685808273e-05,
+      "loss": 12.0699,
+      "step": 13951
+    },
+    {
+      "epoch": 2.439825136612022,
+      "grad_norm": 0.8436921884180728,
+      "learning_rate": 1.7725134215745533e-05,
+      "loss": 11.7187,
+      "step": 13952
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.9009524580042505,
+      "learning_rate": 1.771440367868562e-05,
+      "loss": 12.0033,
+      "step": 13953
+    },
+    {
+      "epoch": 2.4401748633879783,
+      "grad_norm": 0.8202264867148044,
+      "learning_rate": 1.77036760750111e-05,
+      "loss": 12.0667,
+      "step": 13954
+    },
+    {
+      "epoch": 2.440349726775956,
+      "grad_norm": 0.8331428107147062,
+      "learning_rate": 1.7692951405104353e-05,
+      "loss": 12.0366,
+      "step": 13955
+    },
+    {
+      "epoch": 2.4405245901639345,
+      "grad_norm": 0.8567295682257888,
+      "learning_rate": 1.7682229669347726e-05,
+      "loss": 12.1612,
+      "step": 13956
+    },
+    {
+      "epoch": 2.4406994535519124,
+      "grad_norm": 0.8336166542337609,
+      "learning_rate": 1.767151086812343e-05,
+      "loss": 12.1513,
+      "step": 13957
+    },
+    {
+      "epoch": 2.4408743169398908,
+      "grad_norm": 0.7922045047532668,
+      "learning_rate": 1.76607950018136e-05,
+      "loss": 11.8169,
+      "step": 13958
+    },
+    {
+      "epoch": 2.4410491803278687,
+      "grad_norm": 0.8349132154196158,
+      "learning_rate": 1.765008207080021e-05,
+      "loss": 12.1247,
+      "step": 13959
+    },
+    {
+      "epoch": 2.441224043715847,
+      "grad_norm": 0.8436607770046674,
+      "learning_rate": 1.7639372075465177e-05,
+      "loss": 12.1372,
+      "step": 13960
+    },
+    {
+      "epoch": 2.4413989071038253,
+      "grad_norm": 0.8972154776122848,
+      "learning_rate": 1.7628665016190303e-05,
+      "loss": 12.0098,
+      "step": 13961
+    },
+    {
+      "epoch": 2.4415737704918032,
+      "grad_norm": 0.8815715130250056,
+      "learning_rate": 1.7617960893357278e-05,
+      "loss": 11.9497,
+      "step": 13962
+    },
+    {
+      "epoch": 2.4417486338797816,
+      "grad_norm": 0.9934678150308169,
+      "learning_rate": 1.760725970734769e-05,
+      "loss": 12.114,
+      "step": 13963
+    },
+    {
+      "epoch": 2.4419234972677595,
+      "grad_norm": 0.8180570564838864,
+      "learning_rate": 1.7596561458543038e-05,
+      "loss": 12.1045,
+      "step": 13964
+    },
+    {
+      "epoch": 2.442098360655738,
+      "grad_norm": 0.7690295588165138,
+      "learning_rate": 1.7585866147324702e-05,
+      "loss": 12.1005,
+      "step": 13965
+    },
+    {
+      "epoch": 2.4422732240437157,
+      "grad_norm": 0.8851146378971309,
+      "learning_rate": 1.7575173774073937e-05,
+      "loss": 11.9503,
+      "step": 13966
+    },
+    {
+      "epoch": 2.442448087431694,
+      "grad_norm": 0.8818046313106964,
+      "learning_rate": 1.7564484339171915e-05,
+      "loss": 12.02,
+      "step": 13967
+    },
+    {
+      "epoch": 2.442622950819672,
+      "grad_norm": 0.8555221000642443,
+      "learning_rate": 1.7553797842999708e-05,
+      "loss": 12.0004,
+      "step": 13968
+    },
+    {
+      "epoch": 2.4427978142076503,
+      "grad_norm": 0.8183081254328811,
+      "learning_rate": 1.7543114285938312e-05,
+      "loss": 12.3381,
+      "step": 13969
+    },
+    {
+      "epoch": 2.4429726775956286,
+      "grad_norm": 0.8289435562412607,
+      "learning_rate": 1.753243366836853e-05,
+      "loss": 12.1871,
+      "step": 13970
+    },
+    {
+      "epoch": 2.4431475409836065,
+      "grad_norm": 0.9272142412490993,
+      "learning_rate": 1.752175599067113e-05,
+      "loss": 12.1064,
+      "step": 13971
+    },
+    {
+      "epoch": 2.443322404371585,
+      "grad_norm": 0.8684881766402471,
+      "learning_rate": 1.7511081253226758e-05,
+      "loss": 11.74,
+      "step": 13972
+    },
+    {
+      "epoch": 2.4434972677595628,
+      "grad_norm": 0.8418900225601018,
+      "learning_rate": 1.750040945641599e-05,
+      "loss": 12.0482,
+      "step": 13973
+    },
+    {
+      "epoch": 2.443672131147541,
+      "grad_norm": 0.8278910309364457,
+      "learning_rate": 1.748974060061921e-05,
+      "loss": 12.1845,
+      "step": 13974
+    },
+    {
+      "epoch": 2.443846994535519,
+      "grad_norm": 0.824310328430994,
+      "learning_rate": 1.747907468621678e-05,
+      "loss": 12.1751,
+      "step": 13975
+    },
+    {
+      "epoch": 2.4440218579234974,
+      "grad_norm": 0.8883868582273574,
+      "learning_rate": 1.7468411713588917e-05,
+      "loss": 12.022,
+      "step": 13976
+    },
+    {
+      "epoch": 2.4441967213114753,
+      "grad_norm": 0.8814596568260498,
+      "learning_rate": 1.7457751683115775e-05,
+      "loss": 11.936,
+      "step": 13977
+    },
+    {
+      "epoch": 2.4443715846994536,
+      "grad_norm": 0.8246704323077948,
+      "learning_rate": 1.7447094595177317e-05,
+      "loss": 12.0361,
+      "step": 13978
+    },
+    {
+      "epoch": 2.444546448087432,
+      "grad_norm": 0.7946230202462341,
+      "learning_rate": 1.7436440450153492e-05,
+      "loss": 12.1312,
+      "step": 13979
+    },
+    {
+      "epoch": 2.44472131147541,
+      "grad_norm": 0.7822287020066696,
+      "learning_rate": 1.7425789248424086e-05,
+      "loss": 11.9322,
+      "step": 13980
+    },
+    {
+      "epoch": 2.444896174863388,
+      "grad_norm": 0.8982642531544075,
+      "learning_rate": 1.7415140990368826e-05,
+      "loss": 12.0246,
+      "step": 13981
+    },
+    {
+      "epoch": 2.445071038251366,
+      "grad_norm": 0.8663025198894438,
+      "learning_rate": 1.7404495676367293e-05,
+      "loss": 12.2439,
+      "step": 13982
+    },
+    {
+      "epoch": 2.4452459016393444,
+      "grad_norm": 0.8593212916800391,
+      "learning_rate": 1.739385330679899e-05,
+      "loss": 11.8795,
+      "step": 13983
+    },
+    {
+      "epoch": 2.4454207650273223,
+      "grad_norm": 0.8822395765531136,
+      "learning_rate": 1.7383213882043335e-05,
+      "loss": 11.986,
+      "step": 13984
+    },
+    {
+      "epoch": 2.4455956284153006,
+      "grad_norm": 0.8572445084491963,
+      "learning_rate": 1.737257740247954e-05,
+      "loss": 11.853,
+      "step": 13985
+    },
+    {
+      "epoch": 2.4457704918032785,
+      "grad_norm": 0.7937799929933647,
+      "learning_rate": 1.736194386848683e-05,
+      "loss": 12.1484,
+      "step": 13986
+    },
+    {
+      "epoch": 2.445945355191257,
+      "grad_norm": 0.8695745626149354,
+      "learning_rate": 1.735131328044426e-05,
+      "loss": 12.1852,
+      "step": 13987
+    },
+    {
+      "epoch": 2.446120218579235,
+      "grad_norm": 0.8220258093738604,
+      "learning_rate": 1.7340685638730836e-05,
+      "loss": 11.8157,
+      "step": 13988
+    },
+    {
+      "epoch": 2.446295081967213,
+      "grad_norm": 0.8167841067405838,
+      "learning_rate": 1.733006094372537e-05,
+      "loss": 11.6274,
+      "step": 13989
+    },
+    {
+      "epoch": 2.446469945355191,
+      "grad_norm": 0.9192239440111144,
+      "learning_rate": 1.731943919580664e-05,
+      "loss": 12.146,
+      "step": 13990
+    },
+    {
+      "epoch": 2.4466448087431694,
+      "grad_norm": 0.8090649668297218,
+      "learning_rate": 1.7308820395353296e-05,
+      "loss": 11.9195,
+      "step": 13991
+    },
+    {
+      "epoch": 2.4468196721311477,
+      "grad_norm": 0.7978734522637115,
+      "learning_rate": 1.7298204542743924e-05,
+      "loss": 12.0661,
+      "step": 13992
+    },
+    {
+      "epoch": 2.4469945355191256,
+      "grad_norm": 0.8354059839690776,
+      "learning_rate": 1.7287591638356903e-05,
+      "loss": 11.9975,
+      "step": 13993
+    },
+    {
+      "epoch": 2.447169398907104,
+      "grad_norm": 0.7969169738022828,
+      "learning_rate": 1.7276981682570603e-05,
+      "loss": 12.139,
+      "step": 13994
+    },
+    {
+      "epoch": 2.447344262295082,
+      "grad_norm": 0.8920065387529037,
+      "learning_rate": 1.7266374675763285e-05,
+      "loss": 12.18,
+      "step": 13995
+    },
+    {
+      "epoch": 2.44751912568306,
+      "grad_norm": 0.8534821130024761,
+      "learning_rate": 1.7255770618313018e-05,
+      "loss": 12.1247,
+      "step": 13996
+    },
+    {
+      "epoch": 2.447693989071038,
+      "grad_norm": 0.8344348898332618,
+      "learning_rate": 1.7245169510597848e-05,
+      "loss": 12.0802,
+      "step": 13997
+    },
+    {
+      "epoch": 2.4478688524590164,
+      "grad_norm": 0.8497977445779108,
+      "learning_rate": 1.723457135299569e-05,
+      "loss": 11.8153,
+      "step": 13998
+    },
+    {
+      "epoch": 2.4480437158469943,
+      "grad_norm": 0.884481190490763,
+      "learning_rate": 1.722397614588437e-05,
+      "loss": 11.7507,
+      "step": 13999
+    },
+    {
+      "epoch": 2.4482185792349727,
+      "grad_norm": 0.8311842451486293,
+      "learning_rate": 1.7213383889641588e-05,
+      "loss": 12.1751,
+      "step": 14000
+    },
+    {
+      "epoch": 2.448393442622951,
+      "grad_norm": 0.8634784770450333,
+      "learning_rate": 1.7202794584644943e-05,
+      "loss": 12.1591,
+      "step": 14001
+    },
+    {
+      "epoch": 2.448568306010929,
+      "grad_norm": 0.868136193651095,
+      "learning_rate": 1.7192208231271934e-05,
+      "loss": 11.7593,
+      "step": 14002
+    },
+    {
+      "epoch": 2.4487431693989072,
+      "grad_norm": 0.902337476009848,
+      "learning_rate": 1.718162482989998e-05,
+      "loss": 12.1915,
+      "step": 14003
+    },
+    {
+      "epoch": 2.448918032786885,
+      "grad_norm": 0.8815842615960325,
+      "learning_rate": 1.717104438090631e-05,
+      "loss": 11.9918,
+      "step": 14004
+    },
+    {
+      "epoch": 2.4490928961748635,
+      "grad_norm": 0.93422373147784,
+      "learning_rate": 1.7160466884668125e-05,
+      "loss": 11.9169,
+      "step": 14005
+    },
+    {
+      "epoch": 2.4492677595628414,
+      "grad_norm": 0.8620253896866427,
+      "learning_rate": 1.7149892341562514e-05,
+      "loss": 12.0594,
+      "step": 14006
+    },
+    {
+      "epoch": 2.4494426229508197,
+      "grad_norm": 0.8684987978608767,
+      "learning_rate": 1.713932075196647e-05,
+      "loss": 11.9523,
+      "step": 14007
+    },
+    {
+      "epoch": 2.4496174863387976,
+      "grad_norm": 0.8580966107162058,
+      "learning_rate": 1.7128752116256796e-05,
+      "loss": 12.1419,
+      "step": 14008
+    },
+    {
+      "epoch": 2.449792349726776,
+      "grad_norm": 0.8248009041072707,
+      "learning_rate": 1.711818643481029e-05,
+      "loss": 11.7654,
+      "step": 14009
+    },
+    {
+      "epoch": 2.4499672131147543,
+      "grad_norm": 0.8309466914689931,
+      "learning_rate": 1.7107623708003617e-05,
+      "loss": 12.1327,
+      "step": 14010
+    },
+    {
+      "epoch": 2.450142076502732,
+      "grad_norm": 0.7951634354159391,
+      "learning_rate": 1.7097063936213297e-05,
+      "loss": 11.8524,
+      "step": 14011
+    },
+    {
+      "epoch": 2.4503169398907105,
+      "grad_norm": 0.7908384603644906,
+      "learning_rate": 1.7086507119815774e-05,
+      "loss": 12.1252,
+      "step": 14012
+    },
+    {
+      "epoch": 2.4504918032786884,
+      "grad_norm": 0.7695094127977165,
+      "learning_rate": 1.7075953259187405e-05,
+      "loss": 12.0078,
+      "step": 14013
+    },
+    {
+      "epoch": 2.4506666666666668,
+      "grad_norm": 0.8194741423537921,
+      "learning_rate": 1.7065402354704442e-05,
+      "loss": 11.762,
+      "step": 14014
+    },
+    {
+      "epoch": 2.4508415300546447,
+      "grad_norm": 0.9049843147989916,
+      "learning_rate": 1.705485440674296e-05,
+      "loss": 11.9776,
+      "step": 14015
+    },
+    {
+      "epoch": 2.451016393442623,
+      "grad_norm": 0.8480715371195527,
+      "learning_rate": 1.7044309415679007e-05,
+      "loss": 12.2529,
+      "step": 14016
+    },
+    {
+      "epoch": 2.451191256830601,
+      "grad_norm": 0.9119874796702035,
+      "learning_rate": 1.70337673818885e-05,
+      "loss": 12.2392,
+      "step": 14017
+    },
+    {
+      "epoch": 2.4513661202185792,
+      "grad_norm": 0.8730423806136086,
+      "learning_rate": 1.7023228305747252e-05,
+      "loss": 11.9904,
+      "step": 14018
+    },
+    {
+      "epoch": 2.4515409836065576,
+      "grad_norm": 0.8225140114252999,
+      "learning_rate": 1.7012692187630963e-05,
+      "loss": 12.0728,
+      "step": 14019
+    },
+    {
+      "epoch": 2.4517158469945355,
+      "grad_norm": 0.8638158204526025,
+      "learning_rate": 1.7002159027915244e-05,
+      "loss": 11.9766,
+      "step": 14020
+    },
+    {
+      "epoch": 2.451890710382514,
+      "grad_norm": 0.9300630639828481,
+      "learning_rate": 1.699162882697558e-05,
+      "loss": 11.8553,
+      "step": 14021
+    },
+    {
+      "epoch": 2.4520655737704917,
+      "grad_norm": 0.9110490374224073,
+      "learning_rate": 1.6981101585187387e-05,
+      "loss": 11.9545,
+      "step": 14022
+    },
+    {
+      "epoch": 2.45224043715847,
+      "grad_norm": 0.8976399041163534,
+      "learning_rate": 1.6970577302925894e-05,
+      "loss": 12.1484,
+      "step": 14023
+    },
+    {
+      "epoch": 2.452415300546448,
+      "grad_norm": 0.851083572862655,
+      "learning_rate": 1.6960055980566327e-05,
+      "loss": 12.0249,
+      "step": 14024
+    },
+    {
+      "epoch": 2.4525901639344263,
+      "grad_norm": 0.861143497385287,
+      "learning_rate": 1.6949537618483724e-05,
+      "loss": 12.102,
+      "step": 14025
+    },
+    {
+      "epoch": 2.452765027322404,
+      "grad_norm": 0.840521878312951,
+      "learning_rate": 1.6939022217053103e-05,
+      "loss": 11.8746,
+      "step": 14026
+    },
+    {
+      "epoch": 2.4529398907103825,
+      "grad_norm": 0.8647360727177139,
+      "learning_rate": 1.692850977664927e-05,
+      "loss": 11.8618,
+      "step": 14027
+    },
+    {
+      "epoch": 2.453114754098361,
+      "grad_norm": 0.8647046625489658,
+      "learning_rate": 1.6918000297647008e-05,
+      "loss": 12.1625,
+      "step": 14028
+    },
+    {
+      "epoch": 2.453289617486339,
+      "grad_norm": 1.1943454326621525,
+      "learning_rate": 1.690749378042099e-05,
+      "loss": 11.9198,
+      "step": 14029
+    },
+    {
+      "epoch": 2.453464480874317,
+      "grad_norm": 0.8452721421535664,
+      "learning_rate": 1.68969902253457e-05,
+      "loss": 12.112,
+      "step": 14030
+    },
+    {
+      "epoch": 2.453639344262295,
+      "grad_norm": 0.8211247961204183,
+      "learning_rate": 1.6886489632795623e-05,
+      "loss": 12.2583,
+      "step": 14031
+    },
+    {
+      "epoch": 2.4538142076502734,
+      "grad_norm": 0.8466051912586319,
+      "learning_rate": 1.687599200314508e-05,
+      "loss": 11.8651,
+      "step": 14032
+    },
+    {
+      "epoch": 2.4539890710382513,
+      "grad_norm": 0.8428603277640969,
+      "learning_rate": 1.6865497336768297e-05,
+      "loss": 12.0298,
+      "step": 14033
+    },
+    {
+      "epoch": 2.4541639344262296,
+      "grad_norm": 0.8574435305699839,
+      "learning_rate": 1.6855005634039412e-05,
+      "loss": 11.7124,
+      "step": 14034
+    },
+    {
+      "epoch": 2.4543387978142075,
+      "grad_norm": 0.8970207952554653,
+      "learning_rate": 1.684451689533243e-05,
+      "loss": 12.1686,
+      "step": 14035
+    },
+    {
+      "epoch": 2.454513661202186,
+      "grad_norm": 0.8016609498667167,
+      "learning_rate": 1.6834031121021255e-05,
+      "loss": 12.014,
+      "step": 14036
+    },
+    {
+      "epoch": 2.454688524590164,
+      "grad_norm": 0.8209233770730979,
+      "learning_rate": 1.682354831147973e-05,
+      "loss": 11.9427,
+      "step": 14037
+    },
+    {
+      "epoch": 2.454863387978142,
+      "grad_norm": 0.7917795420768712,
+      "learning_rate": 1.6813068467081505e-05,
+      "loss": 11.9432,
+      "step": 14038
+    },
+    {
+      "epoch": 2.45503825136612,
+      "grad_norm": 0.8727522627352766,
+      "learning_rate": 1.68025915882002e-05,
+      "loss": 12.0438,
+      "step": 14039
+    },
+    {
+      "epoch": 2.4552131147540983,
+      "grad_norm": 0.9012844801857087,
+      "learning_rate": 1.6792117675209286e-05,
+      "loss": 12.2104,
+      "step": 14040
+    },
+    {
+      "epoch": 2.4553879781420767,
+      "grad_norm": 0.8544123625833889,
+      "learning_rate": 1.6781646728482193e-05,
+      "loss": 11.9692,
+      "step": 14041
+    },
+    {
+      "epoch": 2.4555628415300546,
+      "grad_norm": 0.8540666284919562,
+      "learning_rate": 1.6771178748392135e-05,
+      "loss": 12.1391,
+      "step": 14042
+    },
+    {
+      "epoch": 2.455737704918033,
+      "grad_norm": 0.8829964070030033,
+      "learning_rate": 1.6760713735312316e-05,
+      "loss": 12.1666,
+      "step": 14043
+    },
+    {
+      "epoch": 2.455912568306011,
+      "grad_norm": 0.8674901712185284,
+      "learning_rate": 1.6750251689615825e-05,
+      "loss": 12.2663,
+      "step": 14044
+    },
+    {
+      "epoch": 2.456087431693989,
+      "grad_norm": 0.8288035427157253,
+      "learning_rate": 1.6739792611675563e-05,
+      "loss": 12.1662,
+      "step": 14045
+    },
+    {
+      "epoch": 2.456262295081967,
+      "grad_norm": 0.8275128907731687,
+      "learning_rate": 1.6729336501864413e-05,
+      "loss": 12.0045,
+      "step": 14046
+    },
+    {
+      "epoch": 2.4564371584699454,
+      "grad_norm": 0.8611711229084236,
+      "learning_rate": 1.6718883360555136e-05,
+      "loss": 11.9059,
+      "step": 14047
+    },
+    {
+      "epoch": 2.4566120218579233,
+      "grad_norm": 0.8755108256032931,
+      "learning_rate": 1.6708433188120377e-05,
+      "loss": 11.93,
+      "step": 14048
+    },
+    {
+      "epoch": 2.4567868852459016,
+      "grad_norm": 0.8041767714609241,
+      "learning_rate": 1.669798598493264e-05,
+      "loss": 11.9079,
+      "step": 14049
+    },
+    {
+      "epoch": 2.45696174863388,
+      "grad_norm": 0.8694181863790654,
+      "learning_rate": 1.668754175136438e-05,
+      "loss": 12.1603,
+      "step": 14050
+    },
+    {
+      "epoch": 2.457136612021858,
+      "grad_norm": 0.8670471463565392,
+      "learning_rate": 1.667710048778791e-05,
+      "loss": 12.1713,
+      "step": 14051
+    },
+    {
+      "epoch": 2.457311475409836,
+      "grad_norm": 0.8190043295918797,
+      "learning_rate": 1.6666662194575455e-05,
+      "loss": 12.1363,
+      "step": 14052
+    },
+    {
+      "epoch": 2.457486338797814,
+      "grad_norm": 0.8235023923542741,
+      "learning_rate": 1.6656226872099125e-05,
+      "loss": 11.9286,
+      "step": 14053
+    },
+    {
+      "epoch": 2.4576612021857924,
+      "grad_norm": 0.8453922177537444,
+      "learning_rate": 1.664579452073094e-05,
+      "loss": 11.8608,
+      "step": 14054
+    },
+    {
+      "epoch": 2.4578360655737703,
+      "grad_norm": 0.8639415798700154,
+      "learning_rate": 1.6635365140842786e-05,
+      "loss": 12.0795,
+      "step": 14055
+    },
+    {
+      "epoch": 2.4580109289617487,
+      "grad_norm": 0.8321626982356215,
+      "learning_rate": 1.6624938732806483e-05,
+      "loss": 12.0969,
+      "step": 14056
+    },
+    {
+      "epoch": 2.4581857923497266,
+      "grad_norm": 0.8565867608278317,
+      "learning_rate": 1.661451529699368e-05,
+      "loss": 11.9084,
+      "step": 14057
+    },
+    {
+      "epoch": 2.458360655737705,
+      "grad_norm": 0.9208975710022785,
+      "learning_rate": 1.6604094833775987e-05,
+      "loss": 11.9888,
+      "step": 14058
+    },
+    {
+      "epoch": 2.4585355191256832,
+      "grad_norm": 0.7971568481636142,
+      "learning_rate": 1.6593677343524895e-05,
+      "loss": 12.0581,
+      "step": 14059
+    },
+    {
+      "epoch": 2.458710382513661,
+      "grad_norm": 0.8014041342054683,
+      "learning_rate": 1.6583262826611735e-05,
+      "loss": 11.7093,
+      "step": 14060
+    },
+    {
+      "epoch": 2.4588852459016395,
+      "grad_norm": 0.8366735417751555,
+      "learning_rate": 1.6572851283407786e-05,
+      "loss": 12.0042,
+      "step": 14061
+    },
+    {
+      "epoch": 2.4590601092896174,
+      "grad_norm": 0.8437206981162864,
+      "learning_rate": 1.656244271428422e-05,
+      "loss": 12.0423,
+      "step": 14062
+    },
+    {
+      "epoch": 2.4592349726775957,
+      "grad_norm": 0.8018665448703333,
+      "learning_rate": 1.6552037119612106e-05,
+      "loss": 11.8308,
+      "step": 14063
+    },
+    {
+      "epoch": 2.4594098360655736,
+      "grad_norm": 0.8119635372332689,
+      "learning_rate": 1.654163449976235e-05,
+      "loss": 12.0394,
+      "step": 14064
+    },
+    {
+      "epoch": 2.459584699453552,
+      "grad_norm": 0.8693174770241733,
+      "learning_rate": 1.6531234855105817e-05,
+      "loss": 11.7139,
+      "step": 14065
+    },
+    {
+      "epoch": 2.45975956284153,
+      "grad_norm": 0.7862988095225854,
+      "learning_rate": 1.6520838186013233e-05,
+      "loss": 11.8374,
+      "step": 14066
+    },
+    {
+      "epoch": 2.459934426229508,
+      "grad_norm": 0.8514342288077302,
+      "learning_rate": 1.651044449285525e-05,
+      "loss": 12.0301,
+      "step": 14067
+    },
+    {
+      "epoch": 2.4601092896174865,
+      "grad_norm": 0.8163909911129581,
+      "learning_rate": 1.6500053776002356e-05,
+      "loss": 11.8383,
+      "step": 14068
+    },
+    {
+      "epoch": 2.4602841530054644,
+      "grad_norm": 0.8622651319136716,
+      "learning_rate": 1.648966603582499e-05,
+      "loss": 11.7199,
+      "step": 14069
+    },
+    {
+      "epoch": 2.4604590163934428,
+      "grad_norm": 0.8134090951985606,
+      "learning_rate": 1.6479281272693446e-05,
+      "loss": 11.978,
+      "step": 14070
+    },
+    {
+      "epoch": 2.4606338797814207,
+      "grad_norm": 0.8700056449496598,
+      "learning_rate": 1.6468899486977952e-05,
+      "loss": 12.0727,
+      "step": 14071
+    },
+    {
+      "epoch": 2.460808743169399,
+      "grad_norm": 0.771033362269402,
+      "learning_rate": 1.6458520679048594e-05,
+      "loss": 12.1334,
+      "step": 14072
+    },
+    {
+      "epoch": 2.460983606557377,
+      "grad_norm": 0.8441235460512088,
+      "learning_rate": 1.6448144849275372e-05,
+      "loss": 11.8979,
+      "step": 14073
+    },
+    {
+      "epoch": 2.4611584699453553,
+      "grad_norm": 0.7887045140739162,
+      "learning_rate": 1.6437771998028186e-05,
+      "loss": 11.8473,
+      "step": 14074
+    },
+    {
+      "epoch": 2.461333333333333,
+      "grad_norm": 0.8963295886038992,
+      "learning_rate": 1.6427402125676762e-05,
+      "loss": 12.0163,
+      "step": 14075
+    },
+    {
+      "epoch": 2.4615081967213115,
+      "grad_norm": 0.8557336211366077,
+      "learning_rate": 1.6417035232590817e-05,
+      "loss": 12.0962,
+      "step": 14076
+    },
+    {
+      "epoch": 2.46168306010929,
+      "grad_norm": 0.8666803341553063,
+      "learning_rate": 1.640667131913991e-05,
+      "loss": 12.0697,
+      "step": 14077
+    },
+    {
+      "epoch": 2.4618579234972677,
+      "grad_norm": 0.8609434557907524,
+      "learning_rate": 1.6396310385693527e-05,
+      "loss": 11.619,
+      "step": 14078
+    },
+    {
+      "epoch": 2.462032786885246,
+      "grad_norm": 0.8821872289278693,
+      "learning_rate": 1.6385952432620977e-05,
+      "loss": 11.9468,
+      "step": 14079
+    },
+    {
+      "epoch": 2.462207650273224,
+      "grad_norm": 0.7996219836312449,
+      "learning_rate": 1.6375597460291525e-05,
+      "loss": 12.1786,
+      "step": 14080
+    },
+    {
+      "epoch": 2.4623825136612023,
+      "grad_norm": 0.8259933663185641,
+      "learning_rate": 1.636524546907432e-05,
+      "loss": 11.886,
+      "step": 14081
+    },
+    {
+      "epoch": 2.46255737704918,
+      "grad_norm": 0.811159446887945,
+      "learning_rate": 1.635489645933842e-05,
+      "loss": 11.9875,
+      "step": 14082
+    },
+    {
+      "epoch": 2.4627322404371585,
+      "grad_norm": 0.828990619318374,
+      "learning_rate": 1.634455043145271e-05,
+      "loss": 11.7135,
+      "step": 14083
+    },
+    {
+      "epoch": 2.4629071038251364,
+      "grad_norm": 0.8509647644928455,
+      "learning_rate": 1.6334207385786027e-05,
+      "loss": 12.0093,
+      "step": 14084
+    },
+    {
+      "epoch": 2.463081967213115,
+      "grad_norm": 0.8165169045727252,
+      "learning_rate": 1.632386732270711e-05,
+      "loss": 11.9924,
+      "step": 14085
+    },
+    {
+      "epoch": 2.463256830601093,
+      "grad_norm": 0.7788580438045126,
+      "learning_rate": 1.6313530242584586e-05,
+      "loss": 11.8638,
+      "step": 14086
+    },
+    {
+      "epoch": 2.463431693989071,
+      "grad_norm": 0.8405900492682794,
+      "learning_rate": 1.6303196145786892e-05,
+      "loss": 11.652,
+      "step": 14087
+    },
+    {
+      "epoch": 2.4636065573770494,
+      "grad_norm": 0.8231758725517625,
+      "learning_rate": 1.629286503268248e-05,
+      "loss": 11.6433,
+      "step": 14088
+    },
+    {
+      "epoch": 2.4637814207650273,
+      "grad_norm": 0.9033378626865459,
+      "learning_rate": 1.628253690363962e-05,
+      "loss": 12.1482,
+      "step": 14089
+    },
+    {
+      "epoch": 2.4639562841530056,
+      "grad_norm": 0.9042466679701783,
+      "learning_rate": 1.6272211759026502e-05,
+      "loss": 12.2503,
+      "step": 14090
+    },
+    {
+      "epoch": 2.4641311475409835,
+      "grad_norm": 0.8776983113124055,
+      "learning_rate": 1.6261889599211223e-05,
+      "loss": 12.1398,
+      "step": 14091
+    },
+    {
+      "epoch": 2.464306010928962,
+      "grad_norm": 0.8238928150880283,
+      "learning_rate": 1.6251570424561736e-05,
+      "loss": 12.0956,
+      "step": 14092
+    },
+    {
+      "epoch": 2.4644808743169397,
+      "grad_norm": 0.8831374977006817,
+      "learning_rate": 1.6241254235445934e-05,
+      "loss": 12.1842,
+      "step": 14093
+    },
+    {
+      "epoch": 2.464655737704918,
+      "grad_norm": 0.8295503862231559,
+      "learning_rate": 1.6230941032231538e-05,
+      "loss": 11.8111,
+      "step": 14094
+    },
+    {
+      "epoch": 2.4648306010928964,
+      "grad_norm": 0.804323185718119,
+      "learning_rate": 1.622063081528622e-05,
+      "loss": 12.0534,
+      "step": 14095
+    },
+    {
+      "epoch": 2.4650054644808743,
+      "grad_norm": 0.8671022360329359,
+      "learning_rate": 1.6210323584977527e-05,
+      "loss": 11.8567,
+      "step": 14096
+    },
+    {
+      "epoch": 2.465180327868852,
+      "grad_norm": 0.8129860182131238,
+      "learning_rate": 1.6200019341672924e-05,
+      "loss": 12.0722,
+      "step": 14097
+    },
+    {
+      "epoch": 2.4653551912568306,
+      "grad_norm": 0.8443215668164441,
+      "learning_rate": 1.618971808573969e-05,
+      "loss": 11.9494,
+      "step": 14098
+    },
+    {
+      "epoch": 2.465530054644809,
+      "grad_norm": 0.8582546409440859,
+      "learning_rate": 1.6179419817545093e-05,
+      "loss": 11.9035,
+      "step": 14099
+    },
+    {
+      "epoch": 2.465704918032787,
+      "grad_norm": 0.8795262060005672,
+      "learning_rate": 1.616912453745625e-05,
+      "loss": 12.0374,
+      "step": 14100
+    },
+    {
+      "epoch": 2.465879781420765,
+      "grad_norm": 0.818642359775391,
+      "learning_rate": 1.6158832245840182e-05,
+      "loss": 12.0388,
+      "step": 14101
+    },
+    {
+      "epoch": 2.466054644808743,
+      "grad_norm": 0.8499369146927583,
+      "learning_rate": 1.614854294306377e-05,
+      "loss": 12.211,
+      "step": 14102
+    },
+    {
+      "epoch": 2.4662295081967214,
+      "grad_norm": 0.8765209200701947,
+      "learning_rate": 1.613825662949383e-05,
+      "loss": 12.2706,
+      "step": 14103
+    },
+    {
+      "epoch": 2.4664043715846993,
+      "grad_norm": 1.0279984769914607,
+      "learning_rate": 1.612797330549709e-05,
+      "loss": 12.1629,
+      "step": 14104
+    },
+    {
+      "epoch": 2.4665792349726776,
+      "grad_norm": 0.9545083299203281,
+      "learning_rate": 1.611769297144008e-05,
+      "loss": 12.0246,
+      "step": 14105
+    },
+    {
+      "epoch": 2.4667540983606555,
+      "grad_norm": 0.8056243175686011,
+      "learning_rate": 1.6107415627689316e-05,
+      "loss": 11.9318,
+      "step": 14106
+    },
+    {
+      "epoch": 2.466928961748634,
+      "grad_norm": 0.8251536959147813,
+      "learning_rate": 1.609714127461116e-05,
+      "loss": 11.9814,
+      "step": 14107
+    },
+    {
+      "epoch": 2.467103825136612,
+      "grad_norm": 0.8426566238031199,
+      "learning_rate": 1.608686991257188e-05,
+      "loss": 12.1371,
+      "step": 14108
+    },
+    {
+      "epoch": 2.46727868852459,
+      "grad_norm": 0.8538729377930189,
+      "learning_rate": 1.6076601541937654e-05,
+      "loss": 12.0937,
+      "step": 14109
+    },
+    {
+      "epoch": 2.4674535519125684,
+      "grad_norm": 0.8987267709485351,
+      "learning_rate": 1.6066336163074534e-05,
+      "loss": 12.0568,
+      "step": 14110
+    },
+    {
+      "epoch": 2.4676284153005463,
+      "grad_norm": 0.9831817383599994,
+      "learning_rate": 1.6056073776348458e-05,
+      "loss": 11.8753,
+      "step": 14111
+    },
+    {
+      "epoch": 2.4678032786885247,
+      "grad_norm": 0.8987902733533717,
+      "learning_rate": 1.604581438212529e-05,
+      "loss": 12.1558,
+      "step": 14112
+    },
+    {
+      "epoch": 2.4679781420765026,
+      "grad_norm": 0.8439571861397718,
+      "learning_rate": 1.6035557980770744e-05,
+      "loss": 11.9358,
+      "step": 14113
+    },
+    {
+      "epoch": 2.468153005464481,
+      "grad_norm": 0.9673853272988142,
+      "learning_rate": 1.6025304572650436e-05,
+      "loss": 12.1301,
+      "step": 14114
+    },
+    {
+      "epoch": 2.468327868852459,
+      "grad_norm": 0.8249830088321025,
+      "learning_rate": 1.6015054158129917e-05,
+      "loss": 12.0532,
+      "step": 14115
+    },
+    {
+      "epoch": 2.468502732240437,
+      "grad_norm": 0.780286180593911,
+      "learning_rate": 1.6004806737574606e-05,
+      "loss": 11.9423,
+      "step": 14116
+    },
+    {
+      "epoch": 2.4686775956284155,
+      "grad_norm": 0.8508177163992127,
+      "learning_rate": 1.5994562311349783e-05,
+      "loss": 11.8864,
+      "step": 14117
+    },
+    {
+      "epoch": 2.4688524590163934,
+      "grad_norm": 0.8361463832303272,
+      "learning_rate": 1.5984320879820658e-05,
+      "loss": 12.1148,
+      "step": 14118
+    },
+    {
+      "epoch": 2.4690273224043717,
+      "grad_norm": 0.805005306749119,
+      "learning_rate": 1.5974082443352356e-05,
+      "loss": 12.16,
+      "step": 14119
+    },
+    {
+      "epoch": 2.4692021857923496,
+      "grad_norm": 0.8224455119146019,
+      "learning_rate": 1.5963847002309818e-05,
+      "loss": 12.018,
+      "step": 14120
+    },
+    {
+      "epoch": 2.469377049180328,
+      "grad_norm": 0.7970831142256983,
+      "learning_rate": 1.595361455705795e-05,
+      "loss": 12.1177,
+      "step": 14121
+    },
+    {
+      "epoch": 2.469551912568306,
+      "grad_norm": 0.8280330648175819,
+      "learning_rate": 1.5943385107961528e-05,
+      "loss": 12.2361,
+      "step": 14122
+    },
+    {
+      "epoch": 2.469726775956284,
+      "grad_norm": 0.8477514551943989,
+      "learning_rate": 1.5933158655385218e-05,
+      "loss": 12.2193,
+      "step": 14123
+    },
+    {
+      "epoch": 2.469901639344262,
+      "grad_norm": 0.8661781008149443,
+      "learning_rate": 1.592293519969358e-05,
+      "loss": 12.1894,
+      "step": 14124
+    },
+    {
+      "epoch": 2.4700765027322404,
+      "grad_norm": 0.8048365751657117,
+      "learning_rate": 1.5912714741251078e-05,
+      "loss": 12.0008,
+      "step": 14125
+    },
+    {
+      "epoch": 2.470251366120219,
+      "grad_norm": 0.8132631313842109,
+      "learning_rate": 1.5902497280422045e-05,
+      "loss": 11.8996,
+      "step": 14126
+    },
+    {
+      "epoch": 2.4704262295081967,
+      "grad_norm": 0.8591516039403597,
+      "learning_rate": 1.5892282817570713e-05,
+      "loss": 11.7663,
+      "step": 14127
+    },
+    {
+      "epoch": 2.470601092896175,
+      "grad_norm": 0.9052087480349496,
+      "learning_rate": 1.5882071353061234e-05,
+      "loss": 12.2074,
+      "step": 14128
+    },
+    {
+      "epoch": 2.470775956284153,
+      "grad_norm": 0.8238887571202832,
+      "learning_rate": 1.587186288725764e-05,
+      "loss": 11.8073,
+      "step": 14129
+    },
+    {
+      "epoch": 2.4709508196721313,
+      "grad_norm": 0.769302914437135,
+      "learning_rate": 1.5861657420523825e-05,
+      "loss": 12.0937,
+      "step": 14130
+    },
+    {
+      "epoch": 2.471125683060109,
+      "grad_norm": 0.8314832480847898,
+      "learning_rate": 1.5851454953223655e-05,
+      "loss": 11.9981,
+      "step": 14131
+    },
+    {
+      "epoch": 2.4713005464480875,
+      "grad_norm": 0.9072334312676719,
+      "learning_rate": 1.5841255485720765e-05,
+      "loss": 12.0553,
+      "step": 14132
+    },
+    {
+      "epoch": 2.4714754098360654,
+      "grad_norm": 0.8309337377234888,
+      "learning_rate": 1.583105901837879e-05,
+      "loss": 12.1114,
+      "step": 14133
+    },
+    {
+      "epoch": 2.4716502732240437,
+      "grad_norm": 0.8805444974888716,
+      "learning_rate": 1.5820865551561225e-05,
+      "loss": 12.0419,
+      "step": 14134
+    },
+    {
+      "epoch": 2.471825136612022,
+      "grad_norm": 0.854404593359292,
+      "learning_rate": 1.5810675085631466e-05,
+      "loss": 12.0184,
+      "step": 14135
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 0.8020610415073928,
+      "learning_rate": 1.5800487620952752e-05,
+      "loss": 11.9583,
+      "step": 14136
+    },
+    {
+      "epoch": 2.4721748633879783,
+      "grad_norm": 0.9656933251597355,
+      "learning_rate": 1.5790303157888285e-05,
+      "loss": 11.9754,
+      "step": 14137
+    },
+    {
+      "epoch": 2.472349726775956,
+      "grad_norm": 0.8459993036904807,
+      "learning_rate": 1.5780121696801143e-05,
+      "loss": 12.1916,
+      "step": 14138
+    },
+    {
+      "epoch": 2.4725245901639346,
+      "grad_norm": 0.829936286591682,
+      "learning_rate": 1.576994323805424e-05,
+      "loss": 12.0823,
+      "step": 14139
+    },
+    {
+      "epoch": 2.4726994535519125,
+      "grad_norm": 0.7924579830149562,
+      "learning_rate": 1.5759767782010447e-05,
+      "loss": 12.1323,
+      "step": 14140
+    },
+    {
+      "epoch": 2.472874316939891,
+      "grad_norm": 0.9116689724746272,
+      "learning_rate": 1.574959532903251e-05,
+      "loss": 12.2492,
+      "step": 14141
+    },
+    {
+      "epoch": 2.4730491803278687,
+      "grad_norm": 0.8812642689353741,
+      "learning_rate": 1.573942587948306e-05,
+      "loss": 11.8601,
+      "step": 14142
+    },
+    {
+      "epoch": 2.473224043715847,
+      "grad_norm": 0.8632592568696817,
+      "learning_rate": 1.572925943372463e-05,
+      "loss": 12.18,
+      "step": 14143
+    },
+    {
+      "epoch": 2.4733989071038254,
+      "grad_norm": 0.8941779340591725,
+      "learning_rate": 1.571909599211965e-05,
+      "loss": 12.0947,
+      "step": 14144
+    },
+    {
+      "epoch": 2.4735737704918033,
+      "grad_norm": 0.8216042240030033,
+      "learning_rate": 1.5708935555030414e-05,
+      "loss": 12.1892,
+      "step": 14145
+    },
+    {
+      "epoch": 2.4737486338797816,
+      "grad_norm": 0.8286874283213197,
+      "learning_rate": 1.569877812281917e-05,
+      "loss": 11.8394,
+      "step": 14146
+    },
+    {
+      "epoch": 2.4739234972677595,
+      "grad_norm": 0.8282552607018769,
+      "learning_rate": 1.568862369584797e-05,
+      "loss": 11.9936,
+      "step": 14147
+    },
+    {
+      "epoch": 2.474098360655738,
+      "grad_norm": 0.8624935619513338,
+      "learning_rate": 1.5678472274478827e-05,
+      "loss": 11.9729,
+      "step": 14148
+    },
+    {
+      "epoch": 2.4742732240437157,
+      "grad_norm": 0.8905305526818139,
+      "learning_rate": 1.566832385907362e-05,
+      "loss": 12.2743,
+      "step": 14149
+    },
+    {
+      "epoch": 2.474448087431694,
+      "grad_norm": 0.8589565604287432,
+      "learning_rate": 1.5658178449994166e-05,
+      "loss": 12.1126,
+      "step": 14150
+    },
+    {
+      "epoch": 2.474622950819672,
+      "grad_norm": 0.8966794342995452,
+      "learning_rate": 1.5648036047602088e-05,
+      "loss": 11.899,
+      "step": 14151
+    },
+    {
+      "epoch": 2.4747978142076503,
+      "grad_norm": 0.9022247030460742,
+      "learning_rate": 1.5637896652258965e-05,
+      "loss": 12.1603,
+      "step": 14152
+    },
+    {
+      "epoch": 2.4749726775956282,
+      "grad_norm": 0.8663239867220537,
+      "learning_rate": 1.5627760264326284e-05,
+      "loss": 11.9351,
+      "step": 14153
+    },
+    {
+      "epoch": 2.4751475409836066,
+      "grad_norm": 0.9151099241099055,
+      "learning_rate": 1.561762688416535e-05,
+      "loss": 12.1758,
+      "step": 14154
+    },
+    {
+      "epoch": 2.4753224043715845,
+      "grad_norm": 0.8187887941081475,
+      "learning_rate": 1.5607496512137433e-05,
+      "loss": 12.0192,
+      "step": 14155
+    },
+    {
+      "epoch": 2.475497267759563,
+      "grad_norm": 0.8595592710219212,
+      "learning_rate": 1.5597369148603646e-05,
+      "loss": 12.01,
+      "step": 14156
+    },
+    {
+      "epoch": 2.475672131147541,
+      "grad_norm": 0.8477427331026254,
+      "learning_rate": 1.558724479392507e-05,
+      "loss": 12.1236,
+      "step": 14157
+    },
+    {
+      "epoch": 2.475846994535519,
+      "grad_norm": 0.8311346708985488,
+      "learning_rate": 1.5577123448462573e-05,
+      "loss": 12.1326,
+      "step": 14158
+    },
+    {
+      "epoch": 2.4760218579234974,
+      "grad_norm": 0.8028270872752064,
+      "learning_rate": 1.5567005112576983e-05,
+      "loss": 12.1413,
+      "step": 14159
+    },
+    {
+      "epoch": 2.4761967213114753,
+      "grad_norm": 0.8650309653796693,
+      "learning_rate": 1.5556889786629e-05,
+      "loss": 12.1908,
+      "step": 14160
+    },
+    {
+      "epoch": 2.4763715846994536,
+      "grad_norm": 0.8363752187032013,
+      "learning_rate": 1.5546777470979245e-05,
+      "loss": 12.2564,
+      "step": 14161
+    },
+    {
+      "epoch": 2.4765464480874315,
+      "grad_norm": 0.8483188558123059,
+      "learning_rate": 1.55366681659882e-05,
+      "loss": 12.2492,
+      "step": 14162
+    },
+    {
+      "epoch": 2.47672131147541,
+      "grad_norm": 0.899014798647105,
+      "learning_rate": 1.5526561872016253e-05,
+      "loss": 12.0157,
+      "step": 14163
+    },
+    {
+      "epoch": 2.4768961748633878,
+      "grad_norm": 0.8644514497308261,
+      "learning_rate": 1.5516458589423677e-05,
+      "loss": 12.0117,
+      "step": 14164
+    },
+    {
+      "epoch": 2.477071038251366,
+      "grad_norm": 0.8878841176105102,
+      "learning_rate": 1.550635831857067e-05,
+      "loss": 11.953,
+      "step": 14165
+    },
+    {
+      "epoch": 2.4772459016393444,
+      "grad_norm": 0.8655183110167635,
+      "learning_rate": 1.5496261059817242e-05,
+      "loss": 11.7904,
+      "step": 14166
+    },
+    {
+      "epoch": 2.4774207650273223,
+      "grad_norm": 0.8532408213586079,
+      "learning_rate": 1.5486166813523373e-05,
+      "loss": 12.121,
+      "step": 14167
+    },
+    {
+      "epoch": 2.4775956284153007,
+      "grad_norm": 0.8103308539383337,
+      "learning_rate": 1.5476075580048932e-05,
+      "loss": 11.9806,
+      "step": 14168
+    },
+    {
+      "epoch": 2.4777704918032786,
+      "grad_norm": 0.8174337482496412,
+      "learning_rate": 1.546598735975363e-05,
+      "loss": 12.051,
+      "step": 14169
+    },
+    {
+      "epoch": 2.477945355191257,
+      "grad_norm": 0.8565089793869871,
+      "learning_rate": 1.5455902152997102e-05,
+      "loss": 12.0681,
+      "step": 14170
+    },
+    {
+      "epoch": 2.478120218579235,
+      "grad_norm": 0.9359968782650055,
+      "learning_rate": 1.5445819960138885e-05,
+      "loss": 12.147,
+      "step": 14171
+    },
+    {
+      "epoch": 2.478295081967213,
+      "grad_norm": 0.8250999100145674,
+      "learning_rate": 1.543574078153842e-05,
+      "loss": 12.0333,
+      "step": 14172
+    },
+    {
+      "epoch": 2.478469945355191,
+      "grad_norm": 0.8333568351770377,
+      "learning_rate": 1.542566461755497e-05,
+      "loss": 12.3167,
+      "step": 14173
+    },
+    {
+      "epoch": 2.4786448087431694,
+      "grad_norm": 0.83764662418744,
+      "learning_rate": 1.5415591468547765e-05,
+      "loss": 12.1655,
+      "step": 14174
+    },
+    {
+      "epoch": 2.4788196721311477,
+      "grad_norm": 0.7916320851766809,
+      "learning_rate": 1.5405521334875893e-05,
+      "loss": 11.7367,
+      "step": 14175
+    },
+    {
+      "epoch": 2.4789945355191256,
+      "grad_norm": 0.809454067884911,
+      "learning_rate": 1.5395454216898364e-05,
+      "loss": 12.045,
+      "step": 14176
+    },
+    {
+      "epoch": 2.479169398907104,
+      "grad_norm": 0.787658356628102,
+      "learning_rate": 1.538539011497403e-05,
+      "loss": 12.0557,
+      "step": 14177
+    },
+    {
+      "epoch": 2.479344262295082,
+      "grad_norm": 0.8948286019015722,
+      "learning_rate": 1.537532902946167e-05,
+      "loss": 12.2456,
+      "step": 14178
+    },
+    {
+      "epoch": 2.47951912568306,
+      "grad_norm": 0.8694514243059689,
+      "learning_rate": 1.536527096071997e-05,
+      "loss": 11.9141,
+      "step": 14179
+    },
+    {
+      "epoch": 2.479693989071038,
+      "grad_norm": 0.8439888330289235,
+      "learning_rate": 1.5355215909107468e-05,
+      "loss": 11.9965,
+      "step": 14180
+    },
+    {
+      "epoch": 2.4798688524590164,
+      "grad_norm": 0.893431558600318,
+      "learning_rate": 1.5345163874982614e-05,
+      "loss": 12.1005,
+      "step": 14181
+    },
+    {
+      "epoch": 2.4800437158469943,
+      "grad_norm": 0.8097243349432027,
+      "learning_rate": 1.533511485870377e-05,
+      "loss": 11.9288,
+      "step": 14182
+    },
+    {
+      "epoch": 2.4802185792349727,
+      "grad_norm": 0.8149927632318641,
+      "learning_rate": 1.5325068860629177e-05,
+      "loss": 11.9013,
+      "step": 14183
+    },
+    {
+      "epoch": 2.480393442622951,
+      "grad_norm": 0.8430195507636017,
+      "learning_rate": 1.531502588111693e-05,
+      "loss": 12.1511,
+      "step": 14184
+    },
+    {
+      "epoch": 2.480568306010929,
+      "grad_norm": 0.9364242155684532,
+      "learning_rate": 1.5304985920525062e-05,
+      "loss": 12.0647,
+      "step": 14185
+    },
+    {
+      "epoch": 2.4807431693989073,
+      "grad_norm": 0.8864085918166759,
+      "learning_rate": 1.5294948979211486e-05,
+      "loss": 12.1277,
+      "step": 14186
+    },
+    {
+      "epoch": 2.480918032786885,
+      "grad_norm": 0.8086262887459199,
+      "learning_rate": 1.5284915057534043e-05,
+      "loss": 12.1595,
+      "step": 14187
+    },
+    {
+      "epoch": 2.4810928961748635,
+      "grad_norm": 0.8665938614792738,
+      "learning_rate": 1.5274884155850377e-05,
+      "loss": 12.0926,
+      "step": 14188
+    },
+    {
+      "epoch": 2.4812677595628414,
+      "grad_norm": 0.8888797068121624,
+      "learning_rate": 1.5264856274518093e-05,
+      "loss": 12.1077,
+      "step": 14189
+    },
+    {
+      "epoch": 2.4814426229508197,
+      "grad_norm": 0.7993771202957876,
+      "learning_rate": 1.5254831413894687e-05,
+      "loss": 12.0981,
+      "step": 14190
+    },
+    {
+      "epoch": 2.4816174863387976,
+      "grad_norm": 0.8871250764036466,
+      "learning_rate": 1.5244809574337537e-05,
+      "loss": 11.7069,
+      "step": 14191
+    },
+    {
+      "epoch": 2.481792349726776,
+      "grad_norm": 0.8400621857815671,
+      "learning_rate": 1.5234790756203888e-05,
+      "loss": 12.1002,
+      "step": 14192
+    },
+    {
+      "epoch": 2.4819672131147543,
+      "grad_norm": 0.8649077463525767,
+      "learning_rate": 1.5224774959850918e-05,
+      "loss": 11.8762,
+      "step": 14193
+    },
+    {
+      "epoch": 2.482142076502732,
+      "grad_norm": 0.8458733425663252,
+      "learning_rate": 1.5214762185635656e-05,
+      "loss": 12.2019,
+      "step": 14194
+    },
+    {
+      "epoch": 2.4823169398907106,
+      "grad_norm": 0.8461721313208116,
+      "learning_rate": 1.5204752433915093e-05,
+      "loss": 12.1545,
+      "step": 14195
+    },
+    {
+      "epoch": 2.4824918032786885,
+      "grad_norm": 0.8207045107765252,
+      "learning_rate": 1.5194745705046009e-05,
+      "loss": 12.1182,
+      "step": 14196
+    },
+    {
+      "epoch": 2.482666666666667,
+      "grad_norm": 0.8023620169116262,
+      "learning_rate": 1.5184741999385154e-05,
+      "loss": 12.1106,
+      "step": 14197
+    },
+    {
+      "epoch": 2.4828415300546447,
+      "grad_norm": 0.8124406930866402,
+      "learning_rate": 1.5174741317289155e-05,
+      "loss": 11.947,
+      "step": 14198
+    },
+    {
+      "epoch": 2.483016393442623,
+      "grad_norm": 0.8073077418351303,
+      "learning_rate": 1.5164743659114522e-05,
+      "loss": 11.9088,
+      "step": 14199
+    },
+    {
+      "epoch": 2.483191256830601,
+      "grad_norm": 0.8323764071081361,
+      "learning_rate": 1.5154749025217651e-05,
+      "loss": 12.0034,
+      "step": 14200
+    },
+    {
+      "epoch": 2.4833661202185793,
+      "grad_norm": 0.8727893817743572,
+      "learning_rate": 1.5144757415954846e-05,
+      "loss": 11.9943,
+      "step": 14201
+    },
+    {
+      "epoch": 2.4835409836065576,
+      "grad_norm": 0.8596703958890097,
+      "learning_rate": 1.5134768831682322e-05,
+      "loss": 11.9814,
+      "step": 14202
+    },
+    {
+      "epoch": 2.4837158469945355,
+      "grad_norm": 0.852424506223901,
+      "learning_rate": 1.5124783272756115e-05,
+      "loss": 12.0106,
+      "step": 14203
+    },
+    {
+      "epoch": 2.4838907103825134,
+      "grad_norm": 0.7711879865483311,
+      "learning_rate": 1.511480073953222e-05,
+      "loss": 12.0601,
+      "step": 14204
+    },
+    {
+      "epoch": 2.4840655737704918,
+      "grad_norm": 0.8167867227743754,
+      "learning_rate": 1.5104821232366495e-05,
+      "loss": 11.9778,
+      "step": 14205
+    },
+    {
+      "epoch": 2.48424043715847,
+      "grad_norm": 0.802189987748144,
+      "learning_rate": 1.5094844751614733e-05,
+      "loss": 12.018,
+      "step": 14206
+    },
+    {
+      "epoch": 2.484415300546448,
+      "grad_norm": 0.7428518936742079,
+      "learning_rate": 1.508487129763252e-05,
+      "loss": 11.9833,
+      "step": 14207
+    },
+    {
+      "epoch": 2.4845901639344263,
+      "grad_norm": 0.8114181046702512,
+      "learning_rate": 1.5074900870775444e-05,
+      "loss": 12.168,
+      "step": 14208
+    },
+    {
+      "epoch": 2.4847650273224042,
+      "grad_norm": 0.7982042581209732,
+      "learning_rate": 1.5064933471398923e-05,
+      "loss": 12.0015,
+      "step": 14209
+    },
+    {
+      "epoch": 2.4849398907103826,
+      "grad_norm": 0.8295378108452242,
+      "learning_rate": 1.5054969099858307e-05,
+      "loss": 12.1431,
+      "step": 14210
+    },
+    {
+      "epoch": 2.4851147540983605,
+      "grad_norm": 0.8517508378322369,
+      "learning_rate": 1.5045007756508778e-05,
+      "loss": 12.2061,
+      "step": 14211
+    },
+    {
+      "epoch": 2.485289617486339,
+      "grad_norm": 0.8314877646433687,
+      "learning_rate": 1.5035049441705463e-05,
+      "loss": 12.0819,
+      "step": 14212
+    },
+    {
+      "epoch": 2.4854644808743167,
+      "grad_norm": 0.8391002811914183,
+      "learning_rate": 1.5025094155803387e-05,
+      "loss": 12.1236,
+      "step": 14213
+    },
+    {
+      "epoch": 2.485639344262295,
+      "grad_norm": 0.8100829547357279,
+      "learning_rate": 1.5015141899157392e-05,
+      "loss": 11.9344,
+      "step": 14214
+    },
+    {
+      "epoch": 2.4858142076502734,
+      "grad_norm": 0.8458635406184274,
+      "learning_rate": 1.5005192672122304e-05,
+      "loss": 12.0181,
+      "step": 14215
+    },
+    {
+      "epoch": 2.4859890710382513,
+      "grad_norm": 0.832024491189584,
+      "learning_rate": 1.4995246475052781e-05,
+      "loss": 12.1271,
+      "step": 14216
+    },
+    {
+      "epoch": 2.4861639344262296,
+      "grad_norm": 0.8292233962895056,
+      "learning_rate": 1.4985303308303411e-05,
+      "loss": 12.0091,
+      "step": 14217
+    },
+    {
+      "epoch": 2.4863387978142075,
+      "grad_norm": 0.8816309416036948,
+      "learning_rate": 1.4975363172228652e-05,
+      "loss": 11.8995,
+      "step": 14218
+    },
+    {
+      "epoch": 2.486513661202186,
+      "grad_norm": 0.8372565716526388,
+      "learning_rate": 1.4965426067182852e-05,
+      "loss": 12.2779,
+      "step": 14219
+    },
+    {
+      "epoch": 2.4866885245901638,
+      "grad_norm": 0.8358607454946596,
+      "learning_rate": 1.4955491993520265e-05,
+      "loss": 11.953,
+      "step": 14220
+    },
+    {
+      "epoch": 2.486863387978142,
+      "grad_norm": 0.8638971085630207,
+      "learning_rate": 1.494556095159504e-05,
+      "loss": 11.662,
+      "step": 14221
+    },
+    {
+      "epoch": 2.48703825136612,
+      "grad_norm": 0.8279222782745013,
+      "learning_rate": 1.4935632941761169e-05,
+      "loss": 11.987,
+      "step": 14222
+    },
+    {
+      "epoch": 2.4872131147540983,
+      "grad_norm": 0.8506862825544187,
+      "learning_rate": 1.4925707964372604e-05,
+      "loss": 11.9619,
+      "step": 14223
+    },
+    {
+      "epoch": 2.4873879781420767,
+      "grad_norm": 0.7879218600491983,
+      "learning_rate": 1.491578601978314e-05,
+      "loss": 12.1831,
+      "step": 14224
+    },
+    {
+      "epoch": 2.4875628415300546,
+      "grad_norm": 0.7866008681142058,
+      "learning_rate": 1.4905867108346528e-05,
+      "loss": 12.1507,
+      "step": 14225
+    },
+    {
+      "epoch": 2.487737704918033,
+      "grad_norm": 0.799843970700542,
+      "learning_rate": 1.4895951230416305e-05,
+      "loss": 11.8846,
+      "step": 14226
+    },
+    {
+      "epoch": 2.487912568306011,
+      "grad_norm": 0.8246891186770332,
+      "learning_rate": 1.488603838634598e-05,
+      "loss": 12.1259,
+      "step": 14227
+    },
+    {
+      "epoch": 2.488087431693989,
+      "grad_norm": 0.7617320012423292,
+      "learning_rate": 1.4876128576488968e-05,
+      "loss": 11.9755,
+      "step": 14228
+    },
+    {
+      "epoch": 2.488262295081967,
+      "grad_norm": 0.8609269239511902,
+      "learning_rate": 1.48662218011985e-05,
+      "loss": 12.1418,
+      "step": 14229
+    },
+    {
+      "epoch": 2.4884371584699454,
+      "grad_norm": 0.8885469304403418,
+      "learning_rate": 1.4856318060827756e-05,
+      "loss": 12.3196,
+      "step": 14230
+    },
+    {
+      "epoch": 2.4886120218579233,
+      "grad_norm": 0.8160605859108232,
+      "learning_rate": 1.4846417355729781e-05,
+      "loss": 11.7655,
+      "step": 14231
+    },
+    {
+      "epoch": 2.4887868852459016,
+      "grad_norm": 0.8459267661394202,
+      "learning_rate": 1.4836519686257544e-05,
+      "loss": 11.9993,
+      "step": 14232
+    },
+    {
+      "epoch": 2.48896174863388,
+      "grad_norm": 0.8411648759499434,
+      "learning_rate": 1.4826625052763888e-05,
+      "loss": 12.1038,
+      "step": 14233
+    },
+    {
+      "epoch": 2.489136612021858,
+      "grad_norm": 0.843285795798156,
+      "learning_rate": 1.4816733455601518e-05,
+      "loss": 12.0797,
+      "step": 14234
+    },
+    {
+      "epoch": 2.489311475409836,
+      "grad_norm": 0.8321139538196679,
+      "learning_rate": 1.4806844895123084e-05,
+      "loss": 11.9989,
+      "step": 14235
+    },
+    {
+      "epoch": 2.489486338797814,
+      "grad_norm": 0.8863519703957866,
+      "learning_rate": 1.4796959371681107e-05,
+      "loss": 11.9461,
+      "step": 14236
+    },
+    {
+      "epoch": 2.4896612021857925,
+      "grad_norm": 0.8088005512253386,
+      "learning_rate": 1.478707688562796e-05,
+      "loss": 11.9124,
+      "step": 14237
+    },
+    {
+      "epoch": 2.4898360655737704,
+      "grad_norm": 0.7977379524718783,
+      "learning_rate": 1.477719743731597e-05,
+      "loss": 12.105,
+      "step": 14238
+    },
+    {
+      "epoch": 2.4900109289617487,
+      "grad_norm": 0.8801867396476233,
+      "learning_rate": 1.4767321027097303e-05,
+      "loss": 12.0827,
+      "step": 14239
+    },
+    {
+      "epoch": 2.4901857923497266,
+      "grad_norm": 0.8221185649847154,
+      "learning_rate": 1.4757447655324098e-05,
+      "loss": 12.0819,
+      "step": 14240
+    },
+    {
+      "epoch": 2.490360655737705,
+      "grad_norm": 0.8171616829209221,
+      "learning_rate": 1.4747577322348249e-05,
+      "loss": 11.9892,
+      "step": 14241
+    },
+    {
+      "epoch": 2.4905355191256833,
+      "grad_norm": 0.8189916488239518,
+      "learning_rate": 1.4737710028521678e-05,
+      "loss": 12.1689,
+      "step": 14242
+    },
+    {
+      "epoch": 2.490710382513661,
+      "grad_norm": 0.8319812818512291,
+      "learning_rate": 1.4727845774196114e-05,
+      "loss": 12.1122,
+      "step": 14243
+    },
+    {
+      "epoch": 2.4908852459016395,
+      "grad_norm": 0.8119084382646167,
+      "learning_rate": 1.471798455972324e-05,
+      "loss": 12.0915,
+      "step": 14244
+    },
+    {
+      "epoch": 2.4910601092896174,
+      "grad_norm": 0.8680583595965472,
+      "learning_rate": 1.4708126385454569e-05,
+      "loss": 12.1477,
+      "step": 14245
+    },
+    {
+      "epoch": 2.4912349726775957,
+      "grad_norm": 0.9078523457532877,
+      "learning_rate": 1.4698271251741524e-05,
+      "loss": 11.9127,
+      "step": 14246
+    },
+    {
+      "epoch": 2.4914098360655736,
+      "grad_norm": 0.8382102515668052,
+      "learning_rate": 1.4688419158935473e-05,
+      "loss": 11.8166,
+      "step": 14247
+    },
+    {
+      "epoch": 2.491584699453552,
+      "grad_norm": 0.7956358073075227,
+      "learning_rate": 1.4678570107387579e-05,
+      "loss": 12.146,
+      "step": 14248
+    },
+    {
+      "epoch": 2.49175956284153,
+      "grad_norm": 0.8435225178994729,
+      "learning_rate": 1.4668724097448972e-05,
+      "loss": 12.163,
+      "step": 14249
+    },
+    {
+      "epoch": 2.4919344262295082,
+      "grad_norm": 0.8542544968640635,
+      "learning_rate": 1.4658881129470647e-05,
+      "loss": 11.8816,
+      "step": 14250
+    },
+    {
+      "epoch": 2.4921092896174866,
+      "grad_norm": 0.8699384687857089,
+      "learning_rate": 1.4649041203803503e-05,
+      "loss": 11.958,
+      "step": 14251
+    },
+    {
+      "epoch": 2.4922841530054645,
+      "grad_norm": 0.8446834227924018,
+      "learning_rate": 1.4639204320798328e-05,
+      "loss": 11.931,
+      "step": 14252
+    },
+    {
+      "epoch": 2.492459016393443,
+      "grad_norm": 0.8520900032449533,
+      "learning_rate": 1.4629370480805771e-05,
+      "loss": 11.875,
+      "step": 14253
+    },
+    {
+      "epoch": 2.4926338797814207,
+      "grad_norm": 0.9687052640284997,
+      "learning_rate": 1.4619539684176419e-05,
+      "loss": 12.1456,
+      "step": 14254
+    },
+    {
+      "epoch": 2.492808743169399,
+      "grad_norm": 0.8394816989070129,
+      "learning_rate": 1.460971193126075e-05,
+      "loss": 12.1725,
+      "step": 14255
+    },
+    {
+      "epoch": 2.492983606557377,
+      "grad_norm": 0.9114771816403713,
+      "learning_rate": 1.4599887222409048e-05,
+      "loss": 11.9009,
+      "step": 14256
+    },
+    {
+      "epoch": 2.4931584699453553,
+      "grad_norm": 0.9782054008715094,
+      "learning_rate": 1.4590065557971599e-05,
+      "loss": 11.9981,
+      "step": 14257
+    },
+    {
+      "epoch": 2.493333333333333,
+      "grad_norm": 0.8466595580177537,
+      "learning_rate": 1.4580246938298513e-05,
+      "loss": 11.9827,
+      "step": 14258
+    },
+    {
+      "epoch": 2.4935081967213115,
+      "grad_norm": 0.8755398005550925,
+      "learning_rate": 1.4570431363739844e-05,
+      "loss": 11.9431,
+      "step": 14259
+    },
+    {
+      "epoch": 2.49368306010929,
+      "grad_norm": 0.8433365316709222,
+      "learning_rate": 1.4560618834645468e-05,
+      "loss": 12.0567,
+      "step": 14260
+    },
+    {
+      "epoch": 2.4938579234972678,
+      "grad_norm": 0.793452539732546,
+      "learning_rate": 1.4550809351365202e-05,
+      "loss": 12.0703,
+      "step": 14261
+    },
+    {
+      "epoch": 2.4940327868852457,
+      "grad_norm": 0.8179544470827043,
+      "learning_rate": 1.4541002914248758e-05,
+      "loss": 12.11,
+      "step": 14262
+    },
+    {
+      "epoch": 2.494207650273224,
+      "grad_norm": 0.8789989516912081,
+      "learning_rate": 1.45311995236457e-05,
+      "loss": 11.9258,
+      "step": 14263
+    },
+    {
+      "epoch": 2.4943825136612023,
+      "grad_norm": 0.8555469767500892,
+      "learning_rate": 1.4521399179905526e-05,
+      "loss": 12.0077,
+      "step": 14264
+    },
+    {
+      "epoch": 2.4945573770491802,
+      "grad_norm": 0.8102944696802623,
+      "learning_rate": 1.451160188337759e-05,
+      "loss": 11.7907,
+      "step": 14265
+    },
+    {
+      "epoch": 2.4947322404371586,
+      "grad_norm": 0.8099141379044894,
+      "learning_rate": 1.450180763441119e-05,
+      "loss": 11.9913,
+      "step": 14266
+    },
+    {
+      "epoch": 2.4949071038251365,
+      "grad_norm": 0.8593032480299199,
+      "learning_rate": 1.4492016433355427e-05,
+      "loss": 11.9905,
+      "step": 14267
+    },
+    {
+      "epoch": 2.495081967213115,
+      "grad_norm": 0.9646792711485435,
+      "learning_rate": 1.4482228280559374e-05,
+      "loss": 12.151,
+      "step": 14268
+    },
+    {
+      "epoch": 2.4952568306010927,
+      "grad_norm": 0.8731158310393567,
+      "learning_rate": 1.4472443176371975e-05,
+      "loss": 11.9942,
+      "step": 14269
+    },
+    {
+      "epoch": 2.495431693989071,
+      "grad_norm": 0.8228957992026007,
+      "learning_rate": 1.446266112114204e-05,
+      "loss": 12.1192,
+      "step": 14270
+    },
+    {
+      "epoch": 2.495606557377049,
+      "grad_norm": 0.8283756420037787,
+      "learning_rate": 1.4452882115218291e-05,
+      "loss": 12.001,
+      "step": 14271
+    },
+    {
+      "epoch": 2.4957814207650273,
+      "grad_norm": 0.966998940266509,
+      "learning_rate": 1.4443106158949348e-05,
+      "loss": 11.8832,
+      "step": 14272
+    },
+    {
+      "epoch": 2.4959562841530056,
+      "grad_norm": 0.884079760104023,
+      "learning_rate": 1.4433333252683711e-05,
+      "loss": 12.2225,
+      "step": 14273
+    },
+    {
+      "epoch": 2.4961311475409835,
+      "grad_norm": 0.8181449384830484,
+      "learning_rate": 1.4423563396769801e-05,
+      "loss": 11.704,
+      "step": 14274
+    },
+    {
+      "epoch": 2.496306010928962,
+      "grad_norm": 0.9349176615214085,
+      "learning_rate": 1.441379659155584e-05,
+      "loss": 12.2109,
+      "step": 14275
+    },
+    {
+      "epoch": 2.4964808743169398,
+      "grad_norm": 0.7895253351061364,
+      "learning_rate": 1.4404032837390025e-05,
+      "loss": 11.9495,
+      "step": 14276
+    },
+    {
+      "epoch": 2.496655737704918,
+      "grad_norm": 0.8487267771622343,
+      "learning_rate": 1.4394272134620468e-05,
+      "loss": 12.0505,
+      "step": 14277
+    },
+    {
+      "epoch": 2.496830601092896,
+      "grad_norm": 0.8532782148703725,
+      "learning_rate": 1.4384514483595058e-05,
+      "loss": 11.9917,
+      "step": 14278
+    },
+    {
+      "epoch": 2.4970054644808743,
+      "grad_norm": 0.8230614118447994,
+      "learning_rate": 1.4374759884661682e-05,
+      "loss": 11.7435,
+      "step": 14279
+    },
+    {
+      "epoch": 2.4971803278688522,
+      "grad_norm": 0.8424644191266911,
+      "learning_rate": 1.4365008338168062e-05,
+      "loss": 11.9749,
+      "step": 14280
+    },
+    {
+      "epoch": 2.4973551912568306,
+      "grad_norm": 0.8161977657446756,
+      "learning_rate": 1.4355259844461876e-05,
+      "loss": 12.2359,
+      "step": 14281
+    },
+    {
+      "epoch": 2.497530054644809,
+      "grad_norm": 0.8199669753779821,
+      "learning_rate": 1.4345514403890581e-05,
+      "loss": 11.9507,
+      "step": 14282
+    },
+    {
+      "epoch": 2.497704918032787,
+      "grad_norm": 0.8263100606046103,
+      "learning_rate": 1.433577201680163e-05,
+      "loss": 12.0584,
+      "step": 14283
+    },
+    {
+      "epoch": 2.497879781420765,
+      "grad_norm": 0.8397198595751789,
+      "learning_rate": 1.4326032683542313e-05,
+      "loss": 11.9982,
+      "step": 14284
+    },
+    {
+      "epoch": 2.498054644808743,
+      "grad_norm": 0.883114465925353,
+      "learning_rate": 1.431629640445985e-05,
+      "loss": 12.0798,
+      "step": 14285
+    },
+    {
+      "epoch": 2.4982295081967214,
+      "grad_norm": 0.8109253099953818,
+      "learning_rate": 1.4306563179901289e-05,
+      "loss": 11.8936,
+      "step": 14286
+    },
+    {
+      "epoch": 2.4984043715846993,
+      "grad_norm": 0.8897159713020344,
+      "learning_rate": 1.4296833010213639e-05,
+      "loss": 12.0645,
+      "step": 14287
+    },
+    {
+      "epoch": 2.4985792349726776,
+      "grad_norm": 0.9263163107298291,
+      "learning_rate": 1.4287105895743746e-05,
+      "loss": 12.086,
+      "step": 14288
+    },
+    {
+      "epoch": 2.4987540983606555,
+      "grad_norm": 0.8513298819961559,
+      "learning_rate": 1.4277381836838389e-05,
+      "loss": 11.8809,
+      "step": 14289
+    },
+    {
+      "epoch": 2.498928961748634,
+      "grad_norm": 0.8405806593779701,
+      "learning_rate": 1.426766083384421e-05,
+      "loss": 11.9037,
+      "step": 14290
+    },
+    {
+      "epoch": 2.4991038251366122,
+      "grad_norm": 0.8375746760282244,
+      "learning_rate": 1.4257942887107756e-05,
+      "loss": 12.2966,
+      "step": 14291
+    },
+    {
+      "epoch": 2.49927868852459,
+      "grad_norm": 0.8019686806782712,
+      "learning_rate": 1.4248227996975483e-05,
+      "loss": 11.8564,
+      "step": 14292
+    },
+    {
+      "epoch": 2.4994535519125685,
+      "grad_norm": 0.8535284162936164,
+      "learning_rate": 1.423851616379367e-05,
+      "loss": 12.0321,
+      "step": 14293
+    },
+    {
+      "epoch": 2.4996284153005464,
+      "grad_norm": 0.8332611971307627,
+      "learning_rate": 1.422880738790855e-05,
+      "loss": 11.999,
+      "step": 14294
+    },
+    {
+      "epoch": 2.4998032786885247,
+      "grad_norm": 0.7847850877187824,
+      "learning_rate": 1.4219101669666235e-05,
+      "loss": 11.4705,
+      "step": 14295
+    },
+    {
+      "epoch": 2.4999781420765026,
+      "grad_norm": 0.8053465715234045,
+      "learning_rate": 1.4209399009412749e-05,
+      "loss": 11.9347,
+      "step": 14296
+    },
+    {
+      "epoch": 2.500153005464481,
+      "grad_norm": 0.8689926109309667,
+      "learning_rate": 1.4199699407493927e-05,
+      "loss": 12.0491,
+      "step": 14297
+    },
+    {
+      "epoch": 2.500327868852459,
+      "grad_norm": 0.8686445931674737,
+      "learning_rate": 1.419000286425557e-05,
+      "loss": 12.0105,
+      "step": 14298
+    },
+    {
+      "epoch": 2.500502732240437,
+      "grad_norm": 0.8584474310057889,
+      "learning_rate": 1.4180309380043367e-05,
+      "loss": 11.8114,
+      "step": 14299
+    },
+    {
+      "epoch": 2.5006775956284155,
+      "grad_norm": 0.7843592957704091,
+      "learning_rate": 1.4170618955202875e-05,
+      "loss": 12.0016,
+      "step": 14300
+    },
+    {
+      "epoch": 2.5008524590163934,
+      "grad_norm": 0.8718288452077836,
+      "learning_rate": 1.4160931590079518e-05,
+      "loss": 11.8085,
+      "step": 14301
+    },
+    {
+      "epoch": 2.5010273224043713,
+      "grad_norm": 0.8387040489868766,
+      "learning_rate": 1.4151247285018655e-05,
+      "loss": 11.8428,
+      "step": 14302
+    },
+    {
+      "epoch": 2.5012021857923497,
+      "grad_norm": 1.023472000802801,
+      "learning_rate": 1.4141566040365517e-05,
+      "loss": 12.126,
+      "step": 14303
+    },
+    {
+      "epoch": 2.501377049180328,
+      "grad_norm": 0.8632073138282902,
+      "learning_rate": 1.4131887856465264e-05,
+      "loss": 12.1693,
+      "step": 14304
+    },
+    {
+      "epoch": 2.501551912568306,
+      "grad_norm": 0.8250728772614667,
+      "learning_rate": 1.4122212733662866e-05,
+      "loss": 12.1153,
+      "step": 14305
+    },
+    {
+      "epoch": 2.5017267759562842,
+      "grad_norm": 0.8543001503697948,
+      "learning_rate": 1.4112540672303242e-05,
+      "loss": 11.8292,
+      "step": 14306
+    },
+    {
+      "epoch": 2.501901639344262,
+      "grad_norm": 0.7854931967632404,
+      "learning_rate": 1.4102871672731187e-05,
+      "loss": 11.9782,
+      "step": 14307
+    },
+    {
+      "epoch": 2.5020765027322405,
+      "grad_norm": 0.7868733129980179,
+      "learning_rate": 1.4093205735291403e-05,
+      "loss": 12.1272,
+      "step": 14308
+    },
+    {
+      "epoch": 2.502251366120219,
+      "grad_norm": 0.9626862677038666,
+      "learning_rate": 1.408354286032847e-05,
+      "loss": 12.2928,
+      "step": 14309
+    },
+    {
+      "epoch": 2.5024262295081967,
+      "grad_norm": 0.9065020071360941,
+      "learning_rate": 1.4073883048186841e-05,
+      "loss": 11.9438,
+      "step": 14310
+    },
+    {
+      "epoch": 2.5026010928961746,
+      "grad_norm": 0.9105473149509085,
+      "learning_rate": 1.4064226299210913e-05,
+      "loss": 11.8851,
+      "step": 14311
+    },
+    {
+      "epoch": 2.502775956284153,
+      "grad_norm": 0.8749089858416782,
+      "learning_rate": 1.4054572613744887e-05,
+      "loss": 11.9119,
+      "step": 14312
+    },
+    {
+      "epoch": 2.5029508196721313,
+      "grad_norm": 0.8277719279738605,
+      "learning_rate": 1.4044921992132931e-05,
+      "loss": 11.9167,
+      "step": 14313
+    },
+    {
+      "epoch": 2.503125683060109,
+      "grad_norm": 0.8362323852369725,
+      "learning_rate": 1.403527443471908e-05,
+      "loss": 11.9895,
+      "step": 14314
+    },
+    {
+      "epoch": 2.5033005464480875,
+      "grad_norm": 0.8256323798691045,
+      "learning_rate": 1.4025629941847285e-05,
+      "loss": 12.1628,
+      "step": 14315
+    },
+    {
+      "epoch": 2.5034754098360654,
+      "grad_norm": 0.8534615838041749,
+      "learning_rate": 1.401598851386131e-05,
+      "loss": 11.9808,
+      "step": 14316
+    },
+    {
+      "epoch": 2.5036502732240438,
+      "grad_norm": 0.801202490019147,
+      "learning_rate": 1.4006350151104874e-05,
+      "loss": 12.0378,
+      "step": 14317
+    },
+    {
+      "epoch": 2.503825136612022,
+      "grad_norm": 0.8090610041218035,
+      "learning_rate": 1.399671485392159e-05,
+      "loss": 12.1157,
+      "step": 14318
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 0.8056433938128063,
+      "learning_rate": 1.3987082622654957e-05,
+      "loss": 11.9754,
+      "step": 14319
+    },
+    {
+      "epoch": 2.504174863387978,
+      "grad_norm": 0.7958309870197471,
+      "learning_rate": 1.3977453457648304e-05,
+      "loss": 11.8214,
+      "step": 14320
+    },
+    {
+      "epoch": 2.5043497267759562,
+      "grad_norm": 0.8237417890630118,
+      "learning_rate": 1.3967827359244934e-05,
+      "loss": 12.1357,
+      "step": 14321
+    },
+    {
+      "epoch": 2.5045245901639346,
+      "grad_norm": 0.8410319563983555,
+      "learning_rate": 1.3958204327788006e-05,
+      "loss": 11.9551,
+      "step": 14322
+    },
+    {
+      "epoch": 2.5046994535519125,
+      "grad_norm": 0.8843782687495858,
+      "learning_rate": 1.3948584363620565e-05,
+      "loss": 12.1514,
+      "step": 14323
+    },
+    {
+      "epoch": 2.504874316939891,
+      "grad_norm": 0.868093074691411,
+      "learning_rate": 1.3938967467085539e-05,
+      "loss": 12.0661,
+      "step": 14324
+    },
+    {
+      "epoch": 2.5050491803278687,
+      "grad_norm": 0.7971958200976379,
+      "learning_rate": 1.3929353638525778e-05,
+      "loss": 11.7483,
+      "step": 14325
+    },
+    {
+      "epoch": 2.505224043715847,
+      "grad_norm": 0.8399783173476901,
+      "learning_rate": 1.3919742878284026e-05,
+      "loss": 11.5738,
+      "step": 14326
+    },
+    {
+      "epoch": 2.5053989071038254,
+      "grad_norm": 0.8486299391556723,
+      "learning_rate": 1.3910135186702833e-05,
+      "loss": 11.6918,
+      "step": 14327
+    },
+    {
+      "epoch": 2.5055737704918033,
+      "grad_norm": 0.8337514630367758,
+      "learning_rate": 1.3900530564124748e-05,
+      "loss": 12.0981,
+      "step": 14328
+    },
+    {
+      "epoch": 2.505748633879781,
+      "grad_norm": 0.895412724024044,
+      "learning_rate": 1.389092901089214e-05,
+      "loss": 12.1766,
+      "step": 14329
+    },
+    {
+      "epoch": 2.5059234972677595,
+      "grad_norm": 0.8290531856808979,
+      "learning_rate": 1.3881330527347337e-05,
+      "loss": 12.0501,
+      "step": 14330
+    },
+    {
+      "epoch": 2.506098360655738,
+      "grad_norm": 0.9323764937454811,
+      "learning_rate": 1.3871735113832452e-05,
+      "loss": 11.9755,
+      "step": 14331
+    },
+    {
+      "epoch": 2.5062732240437158,
+      "grad_norm": 0.7825062913178339,
+      "learning_rate": 1.3862142770689579e-05,
+      "loss": 12.0373,
+      "step": 14332
+    },
+    {
+      "epoch": 2.506448087431694,
+      "grad_norm": 0.8536733558227121,
+      "learning_rate": 1.3852553498260678e-05,
+      "loss": 11.8373,
+      "step": 14333
+    },
+    {
+      "epoch": 2.506622950819672,
+      "grad_norm": 0.8259043782251957,
+      "learning_rate": 1.384296729688762e-05,
+      "loss": 12.1469,
+      "step": 14334
+    },
+    {
+      "epoch": 2.5067978142076504,
+      "grad_norm": 0.7796821911168321,
+      "learning_rate": 1.3833384166912077e-05,
+      "loss": 12.0614,
+      "step": 14335
+    },
+    {
+      "epoch": 2.5069726775956283,
+      "grad_norm": 0.8914445088265378,
+      "learning_rate": 1.382380410867573e-05,
+      "loss": 11.9377,
+      "step": 14336
+    },
+    {
+      "epoch": 2.5071475409836066,
+      "grad_norm": 0.8089087438835985,
+      "learning_rate": 1.3814227122520097e-05,
+      "loss": 12.1792,
+      "step": 14337
+    },
+    {
+      "epoch": 2.5073224043715845,
+      "grad_norm": 0.8257029362149121,
+      "learning_rate": 1.3804653208786544e-05,
+      "loss": 12.0551,
+      "step": 14338
+    },
+    {
+      "epoch": 2.507497267759563,
+      "grad_norm": 0.8187048059937506,
+      "learning_rate": 1.3795082367816404e-05,
+      "loss": 12.0435,
+      "step": 14339
+    },
+    {
+      "epoch": 2.507672131147541,
+      "grad_norm": 0.8342799409042951,
+      "learning_rate": 1.3785514599950843e-05,
+      "loss": 11.8002,
+      "step": 14340
+    },
+    {
+      "epoch": 2.507846994535519,
+      "grad_norm": 0.8499558352991928,
+      "learning_rate": 1.377594990553097e-05,
+      "loss": 11.9495,
+      "step": 14341
+    },
+    {
+      "epoch": 2.5080218579234974,
+      "grad_norm": 0.9018401569116018,
+      "learning_rate": 1.3766388284897736e-05,
+      "loss": 12.0861,
+      "step": 14342
+    },
+    {
+      "epoch": 2.5081967213114753,
+      "grad_norm": 0.8178322223706719,
+      "learning_rate": 1.3756829738392008e-05,
+      "loss": 11.9027,
+      "step": 14343
+    },
+    {
+      "epoch": 2.5083715846994536,
+      "grad_norm": 0.871995593679521,
+      "learning_rate": 1.3747274266354537e-05,
+      "loss": 11.7502,
+      "step": 14344
+    },
+    {
+      "epoch": 2.5085464480874315,
+      "grad_norm": 0.8797471898428681,
+      "learning_rate": 1.3737721869125975e-05,
+      "loss": 12.1618,
+      "step": 14345
+    },
+    {
+      "epoch": 2.50872131147541,
+      "grad_norm": 0.857356593558665,
+      "learning_rate": 1.3728172547046825e-05,
+      "loss": 12.0355,
+      "step": 14346
+    },
+    {
+      "epoch": 2.508896174863388,
+      "grad_norm": 0.8463428832775198,
+      "learning_rate": 1.3718626300457537e-05,
+      "loss": 11.6683,
+      "step": 14347
+    },
+    {
+      "epoch": 2.509071038251366,
+      "grad_norm": 0.9160331237181124,
+      "learning_rate": 1.3709083129698397e-05,
+      "loss": 11.9897,
+      "step": 14348
+    },
+    {
+      "epoch": 2.5092459016393445,
+      "grad_norm": 0.8481564100265192,
+      "learning_rate": 1.3699543035109652e-05,
+      "loss": 11.96,
+      "step": 14349
+    },
+    {
+      "epoch": 2.5094207650273224,
+      "grad_norm": 0.8568151170010275,
+      "learning_rate": 1.3690006017031343e-05,
+      "loss": 12.0346,
+      "step": 14350
+    },
+    {
+      "epoch": 2.5095956284153007,
+      "grad_norm": 0.840765693902795,
+      "learning_rate": 1.3680472075803485e-05,
+      "loss": 12.0572,
+      "step": 14351
+    },
+    {
+      "epoch": 2.5097704918032786,
+      "grad_norm": 0.7897625506934481,
+      "learning_rate": 1.3670941211765932e-05,
+      "loss": 12.2373,
+      "step": 14352
+    },
+    {
+      "epoch": 2.509945355191257,
+      "grad_norm": 0.8333745983767858,
+      "learning_rate": 1.3661413425258485e-05,
+      "loss": 11.9107,
+      "step": 14353
+    },
+    {
+      "epoch": 2.510120218579235,
+      "grad_norm": 0.8863543828482716,
+      "learning_rate": 1.3651888716620754e-05,
+      "loss": 12.0085,
+      "step": 14354
+    },
+    {
+      "epoch": 2.510295081967213,
+      "grad_norm": 0.8449562661625603,
+      "learning_rate": 1.3642367086192309e-05,
+      "loss": 11.7806,
+      "step": 14355
+    },
+    {
+      "epoch": 2.510469945355191,
+      "grad_norm": 0.9163896122149292,
+      "learning_rate": 1.3632848534312592e-05,
+      "loss": 11.8734,
+      "step": 14356
+    },
+    {
+      "epoch": 2.5106448087431694,
+      "grad_norm": 0.8349716579566002,
+      "learning_rate": 1.3623333061320897e-05,
+      "loss": 11.9902,
+      "step": 14357
+    },
+    {
+      "epoch": 2.5108196721311478,
+      "grad_norm": 0.8719131759729781,
+      "learning_rate": 1.3613820667556454e-05,
+      "loss": 12.1553,
+      "step": 14358
+    },
+    {
+      "epoch": 2.5109945355191257,
+      "grad_norm": 0.8617535534173499,
+      "learning_rate": 1.3604311353358379e-05,
+      "loss": 11.9411,
+      "step": 14359
+    },
+    {
+      "epoch": 2.5111693989071036,
+      "grad_norm": 0.7805402809337041,
+      "learning_rate": 1.3594805119065657e-05,
+      "loss": 11.6626,
+      "step": 14360
+    },
+    {
+      "epoch": 2.511344262295082,
+      "grad_norm": 0.9479918657890278,
+      "learning_rate": 1.3585301965017182e-05,
+      "loss": 11.9186,
+      "step": 14361
+    },
+    {
+      "epoch": 2.5115191256830602,
+      "grad_norm": 0.894073278331747,
+      "learning_rate": 1.3575801891551731e-05,
+      "loss": 12.188,
+      "step": 14362
+    },
+    {
+      "epoch": 2.511693989071038,
+      "grad_norm": 0.8557226242465168,
+      "learning_rate": 1.3566304899007965e-05,
+      "loss": 12.0316,
+      "step": 14363
+    },
+    {
+      "epoch": 2.5118688524590165,
+      "grad_norm": 0.8747919886801071,
+      "learning_rate": 1.3556810987724455e-05,
+      "loss": 12.1488,
+      "step": 14364
+    },
+    {
+      "epoch": 2.5120437158469944,
+      "grad_norm": 0.9177045446003529,
+      "learning_rate": 1.3547320158039622e-05,
+      "loss": 12.1419,
+      "step": 14365
+    },
+    {
+      "epoch": 2.5122185792349727,
+      "grad_norm": 0.7859749506679483,
+      "learning_rate": 1.3537832410291818e-05,
+      "loss": 11.9281,
+      "step": 14366
+    },
+    {
+      "epoch": 2.512393442622951,
+      "grad_norm": 0.9123759862414618,
+      "learning_rate": 1.3528347744819258e-05,
+      "loss": 12.0898,
+      "step": 14367
+    },
+    {
+      "epoch": 2.512568306010929,
+      "grad_norm": 0.8319938348606775,
+      "learning_rate": 1.3518866161960098e-05,
+      "loss": 11.8288,
+      "step": 14368
+    },
+    {
+      "epoch": 2.512743169398907,
+      "grad_norm": 0.8362702784828185,
+      "learning_rate": 1.3509387662052298e-05,
+      "loss": 12.0261,
+      "step": 14369
+    },
+    {
+      "epoch": 2.512918032786885,
+      "grad_norm": 0.9073125991930466,
+      "learning_rate": 1.3499912245433777e-05,
+      "loss": 12.0802,
+      "step": 14370
+    },
+    {
+      "epoch": 2.5130928961748635,
+      "grad_norm": 0.9132113636511721,
+      "learning_rate": 1.3490439912442331e-05,
+      "loss": 11.9428,
+      "step": 14371
+    },
+    {
+      "epoch": 2.5132677595628414,
+      "grad_norm": 0.8318106992912343,
+      "learning_rate": 1.3480970663415615e-05,
+      "loss": 12.1673,
+      "step": 14372
+    },
+    {
+      "epoch": 2.5134426229508198,
+      "grad_norm": 0.8516418698136359,
+      "learning_rate": 1.34715044986912e-05,
+      "loss": 12.0021,
+      "step": 14373
+    },
+    {
+      "epoch": 2.5136174863387977,
+      "grad_norm": 0.8081541790687566,
+      "learning_rate": 1.3462041418606552e-05,
+      "loss": 11.7535,
+      "step": 14374
+    },
+    {
+      "epoch": 2.513792349726776,
+      "grad_norm": 0.8174292169055387,
+      "learning_rate": 1.3452581423499044e-05,
+      "loss": 11.9511,
+      "step": 14375
+    },
+    {
+      "epoch": 2.5139672131147544,
+      "grad_norm": 0.8155033163484899,
+      "learning_rate": 1.3443124513705863e-05,
+      "loss": 11.7271,
+      "step": 14376
+    },
+    {
+      "epoch": 2.5141420765027322,
+      "grad_norm": 0.9190750534591126,
+      "learning_rate": 1.343367068956417e-05,
+      "loss": 11.9923,
+      "step": 14377
+    },
+    {
+      "epoch": 2.51431693989071,
+      "grad_norm": 0.9314387100226824,
+      "learning_rate": 1.3424219951410965e-05,
+      "loss": 11.9646,
+      "step": 14378
+    },
+    {
+      "epoch": 2.5144918032786885,
+      "grad_norm": 0.8809963176803275,
+      "learning_rate": 1.3414772299583177e-05,
+      "loss": 11.7437,
+      "step": 14379
+    },
+    {
+      "epoch": 2.514666666666667,
+      "grad_norm": 0.8037815748212039,
+      "learning_rate": 1.3405327734417583e-05,
+      "loss": 11.7248,
+      "step": 14380
+    },
+    {
+      "epoch": 2.5148415300546447,
+      "grad_norm": 0.850645416230728,
+      "learning_rate": 1.3395886256250889e-05,
+      "loss": 12.0028,
+      "step": 14381
+    },
+    {
+      "epoch": 2.515016393442623,
+      "grad_norm": 0.8478980493875239,
+      "learning_rate": 1.338644786541966e-05,
+      "loss": 11.8569,
+      "step": 14382
+    },
+    {
+      "epoch": 2.515191256830601,
+      "grad_norm": 0.8564086166710826,
+      "learning_rate": 1.3377012562260383e-05,
+      "loss": 12.1215,
+      "step": 14383
+    },
+    {
+      "epoch": 2.5153661202185793,
+      "grad_norm": 0.876686376722307,
+      "learning_rate": 1.3367580347109388e-05,
+      "loss": 11.9714,
+      "step": 14384
+    },
+    {
+      "epoch": 2.5155409836065576,
+      "grad_norm": 0.9182008620606379,
+      "learning_rate": 1.3358151220302929e-05,
+      "loss": 12.1969,
+      "step": 14385
+    },
+    {
+      "epoch": 2.5157158469945355,
+      "grad_norm": 0.8044257285439147,
+      "learning_rate": 1.3348725182177168e-05,
+      "loss": 11.8371,
+      "step": 14386
+    },
+    {
+      "epoch": 2.5158907103825134,
+      "grad_norm": 0.8364947394190778,
+      "learning_rate": 1.3339302233068097e-05,
+      "loss": 11.7821,
+      "step": 14387
+    },
+    {
+      "epoch": 2.516065573770492,
+      "grad_norm": 0.9400230205577828,
+      "learning_rate": 1.3329882373311653e-05,
+      "loss": 12.1051,
+      "step": 14388
+    },
+    {
+      "epoch": 2.51624043715847,
+      "grad_norm": 0.8588776622259858,
+      "learning_rate": 1.3320465603243626e-05,
+      "loss": 11.9931,
+      "step": 14389
+    },
+    {
+      "epoch": 2.516415300546448,
+      "grad_norm": 0.8928632503860809,
+      "learning_rate": 1.3311051923199757e-05,
+      "loss": 12.0088,
+      "step": 14390
+    },
+    {
+      "epoch": 2.5165901639344264,
+      "grad_norm": 0.8617507213840829,
+      "learning_rate": 1.330164133351558e-05,
+      "loss": 12.0453,
+      "step": 14391
+    },
+    {
+      "epoch": 2.5167650273224043,
+      "grad_norm": 0.8375734103184235,
+      "learning_rate": 1.329223383452659e-05,
+      "loss": 12.0838,
+      "step": 14392
+    },
+    {
+      "epoch": 2.5169398907103826,
+      "grad_norm": 0.8313902232077007,
+      "learning_rate": 1.3282829426568166e-05,
+      "loss": 12.1512,
+      "step": 14393
+    },
+    {
+      "epoch": 2.5171147540983605,
+      "grad_norm": 0.8185133882029396,
+      "learning_rate": 1.327342810997556e-05,
+      "loss": 12.1215,
+      "step": 14394
+    },
+    {
+      "epoch": 2.517289617486339,
+      "grad_norm": 0.75317449124121,
+      "learning_rate": 1.3264029885083907e-05,
+      "loss": 11.847,
+      "step": 14395
+    },
+    {
+      "epoch": 2.5174644808743167,
+      "grad_norm": 0.8360760234115978,
+      "learning_rate": 1.3254634752228246e-05,
+      "loss": 12.1982,
+      "step": 14396
+    },
+    {
+      "epoch": 2.517639344262295,
+      "grad_norm": 0.7978141489720804,
+      "learning_rate": 1.32452427117435e-05,
+      "loss": 12.0789,
+      "step": 14397
+    },
+    {
+      "epoch": 2.5178142076502734,
+      "grad_norm": 0.8159249590018562,
+      "learning_rate": 1.323585376396449e-05,
+      "loss": 12.11,
+      "step": 14398
+    },
+    {
+      "epoch": 2.5179890710382513,
+      "grad_norm": 0.8995560099383605,
+      "learning_rate": 1.3226467909225915e-05,
+      "loss": 11.7882,
+      "step": 14399
+    },
+    {
+      "epoch": 2.5181639344262297,
+      "grad_norm": 0.8489706290947807,
+      "learning_rate": 1.3217085147862384e-05,
+      "loss": 11.9677,
+      "step": 14400
+    },
+    {
+      "epoch": 2.5183387978142076,
+      "grad_norm": 0.8760293960423787,
+      "learning_rate": 1.3207705480208388e-05,
+      "loss": 11.9561,
+      "step": 14401
+    },
+    {
+      "epoch": 2.518513661202186,
+      "grad_norm": 0.8611263857590663,
+      "learning_rate": 1.3198328906598268e-05,
+      "loss": 12.1044,
+      "step": 14402
+    },
+    {
+      "epoch": 2.518688524590164,
+      "grad_norm": 0.8299166590574841,
+      "learning_rate": 1.3188955427366302e-05,
+      "loss": 12.0986,
+      "step": 14403
+    },
+    {
+      "epoch": 2.518863387978142,
+      "grad_norm": 0.8933765727108648,
+      "learning_rate": 1.3179585042846654e-05,
+      "loss": 12.0183,
+      "step": 14404
+    },
+    {
+      "epoch": 2.51903825136612,
+      "grad_norm": 0.883198617835937,
+      "learning_rate": 1.3170217753373371e-05,
+      "loss": 12.2161,
+      "step": 14405
+    },
+    {
+      "epoch": 2.5192131147540984,
+      "grad_norm": 0.781582987279108,
+      "learning_rate": 1.3160853559280362e-05,
+      "loss": 12.1467,
+      "step": 14406
+    },
+    {
+      "epoch": 2.5193879781420767,
+      "grad_norm": 0.8096808224445208,
+      "learning_rate": 1.315149246090146e-05,
+      "loss": 12.0115,
+      "step": 14407
+    },
+    {
+      "epoch": 2.5195628415300546,
+      "grad_norm": 0.8432146081640696,
+      "learning_rate": 1.3142134458570377e-05,
+      "loss": 12.0377,
+      "step": 14408
+    },
+    {
+      "epoch": 2.519737704918033,
+      "grad_norm": 0.8483521528294677,
+      "learning_rate": 1.3132779552620744e-05,
+      "loss": 12.1041,
+      "step": 14409
+    },
+    {
+      "epoch": 2.519912568306011,
+      "grad_norm": 0.8484533191949541,
+      "learning_rate": 1.3123427743386007e-05,
+      "loss": 12.0591,
+      "step": 14410
+    },
+    {
+      "epoch": 2.520087431693989,
+      "grad_norm": 0.8484371292868151,
+      "learning_rate": 1.3114079031199567e-05,
+      "loss": 11.8989,
+      "step": 14411
+    },
+    {
+      "epoch": 2.520262295081967,
+      "grad_norm": 0.8376085226456307,
+      "learning_rate": 1.3104733416394687e-05,
+      "loss": 11.8579,
+      "step": 14412
+    },
+    {
+      "epoch": 2.5204371584699454,
+      "grad_norm": 0.7921101839711177,
+      "learning_rate": 1.3095390899304549e-05,
+      "loss": 11.8965,
+      "step": 14413
+    },
+    {
+      "epoch": 2.5206120218579233,
+      "grad_norm": 0.832840333737735,
+      "learning_rate": 1.3086051480262173e-05,
+      "loss": 12.0437,
+      "step": 14414
+    },
+    {
+      "epoch": 2.5207868852459017,
+      "grad_norm": 0.8853783706718761,
+      "learning_rate": 1.3076715159600517e-05,
+      "loss": 11.8643,
+      "step": 14415
+    },
+    {
+      "epoch": 2.52096174863388,
+      "grad_norm": 0.8979747935166583,
+      "learning_rate": 1.3067381937652434e-05,
+      "loss": 12.08,
+      "step": 14416
+    },
+    {
+      "epoch": 2.521136612021858,
+      "grad_norm": 0.8782880321448272,
+      "learning_rate": 1.3058051814750594e-05,
+      "loss": 11.9091,
+      "step": 14417
+    },
+    {
+      "epoch": 2.521311475409836,
+      "grad_norm": 0.9523440982442614,
+      "learning_rate": 1.3048724791227618e-05,
+      "loss": 12.0725,
+      "step": 14418
+    },
+    {
+      "epoch": 2.521486338797814,
+      "grad_norm": 0.8262993797650416,
+      "learning_rate": 1.3039400867416018e-05,
+      "loss": 12.2073,
+      "step": 14419
+    },
+    {
+      "epoch": 2.5216612021857925,
+      "grad_norm": 0.8654502093699902,
+      "learning_rate": 1.3030080043648185e-05,
+      "loss": 11.5859,
+      "step": 14420
+    },
+    {
+      "epoch": 2.5218360655737704,
+      "grad_norm": 0.8198951214908269,
+      "learning_rate": 1.3020762320256374e-05,
+      "loss": 12.0142,
+      "step": 14421
+    },
+    {
+      "epoch": 2.5220109289617487,
+      "grad_norm": 0.8071780548263199,
+      "learning_rate": 1.3011447697572765e-05,
+      "loss": 11.8195,
+      "step": 14422
+    },
+    {
+      "epoch": 2.5221857923497266,
+      "grad_norm": 0.9073567628996326,
+      "learning_rate": 1.3002136175929403e-05,
+      "loss": 12.2565,
+      "step": 14423
+    },
+    {
+      "epoch": 2.522360655737705,
+      "grad_norm": 0.8411341611438509,
+      "learning_rate": 1.2992827755658266e-05,
+      "loss": 11.8228,
+      "step": 14424
+    },
+    {
+      "epoch": 2.5225355191256833,
+      "grad_norm": 0.9099790995323949,
+      "learning_rate": 1.2983522437091134e-05,
+      "loss": 12.0235,
+      "step": 14425
+    },
+    {
+      "epoch": 2.522710382513661,
+      "grad_norm": 0.8905631961058169,
+      "learning_rate": 1.2974220220559763e-05,
+      "loss": 12.0956,
+      "step": 14426
+    },
+    {
+      "epoch": 2.522885245901639,
+      "grad_norm": 0.8656892953953399,
+      "learning_rate": 1.2964921106395767e-05,
+      "loss": 12.1485,
+      "step": 14427
+    },
+    {
+      "epoch": 2.5230601092896174,
+      "grad_norm": 0.8126991362633821,
+      "learning_rate": 1.2955625094930667e-05,
+      "loss": 11.931,
+      "step": 14428
+    },
+    {
+      "epoch": 2.5232349726775958,
+      "grad_norm": 0.8407007634323834,
+      "learning_rate": 1.2946332186495802e-05,
+      "loss": 12.0746,
+      "step": 14429
+    },
+    {
+      "epoch": 2.5234098360655737,
+      "grad_norm": 0.9716967560562049,
+      "learning_rate": 1.2937042381422482e-05,
+      "loss": 11.7008,
+      "step": 14430
+    },
+    {
+      "epoch": 2.523584699453552,
+      "grad_norm": 0.8166516676630017,
+      "learning_rate": 1.2927755680041886e-05,
+      "loss": 12.092,
+      "step": 14431
+    },
+    {
+      "epoch": 2.52375956284153,
+      "grad_norm": 0.9048046537390959,
+      "learning_rate": 1.2918472082685063e-05,
+      "loss": 12.083,
+      "step": 14432
+    },
+    {
+      "epoch": 2.5239344262295083,
+      "grad_norm": 0.793609272694295,
+      "learning_rate": 1.2909191589682968e-05,
+      "loss": 11.8747,
+      "step": 14433
+    },
+    {
+      "epoch": 2.5241092896174866,
+      "grad_norm": 0.891293945632264,
+      "learning_rate": 1.2899914201366447e-05,
+      "loss": 11.7509,
+      "step": 14434
+    },
+    {
+      "epoch": 2.5242841530054645,
+      "grad_norm": 0.8443486165471226,
+      "learning_rate": 1.2890639918066228e-05,
+      "loss": 12.1366,
+      "step": 14435
+    },
+    {
+      "epoch": 2.5244590163934424,
+      "grad_norm": 0.8191099788511197,
+      "learning_rate": 1.2881368740112897e-05,
+      "loss": 12.089,
+      "step": 14436
+    },
+    {
+      "epoch": 2.5246338797814207,
+      "grad_norm": 0.8277832083565948,
+      "learning_rate": 1.2872100667836995e-05,
+      "loss": 12.0508,
+      "step": 14437
+    },
+    {
+      "epoch": 2.524808743169399,
+      "grad_norm": 0.8478917103118231,
+      "learning_rate": 1.2862835701568898e-05,
+      "loss": 12.067,
+      "step": 14438
+    },
+    {
+      "epoch": 2.524983606557377,
+      "grad_norm": 0.8829250042640304,
+      "learning_rate": 1.2853573841638922e-05,
+      "loss": 12.1652,
+      "step": 14439
+    },
+    {
+      "epoch": 2.5251584699453553,
+      "grad_norm": 0.8612513010429051,
+      "learning_rate": 1.2844315088377201e-05,
+      "loss": 12.0845,
+      "step": 14440
+    },
+    {
+      "epoch": 2.525333333333333,
+      "grad_norm": 0.8556124901038092,
+      "learning_rate": 1.2835059442113807e-05,
+      "loss": 11.8479,
+      "step": 14441
+    },
+    {
+      "epoch": 2.5255081967213115,
+      "grad_norm": 0.8643898292109791,
+      "learning_rate": 1.2825806903178716e-05,
+      "loss": 11.8535,
+      "step": 14442
+    },
+    {
+      "epoch": 2.52568306010929,
+      "grad_norm": 0.820218445512538,
+      "learning_rate": 1.281655747190177e-05,
+      "loss": 12.0856,
+      "step": 14443
+    },
+    {
+      "epoch": 2.525857923497268,
+      "grad_norm": 0.8830738124480426,
+      "learning_rate": 1.2807311148612666e-05,
+      "loss": 12.1495,
+      "step": 14444
+    },
+    {
+      "epoch": 2.5260327868852457,
+      "grad_norm": 0.7936553553928469,
+      "learning_rate": 1.2798067933641045e-05,
+      "loss": 11.886,
+      "step": 14445
+    },
+    {
+      "epoch": 2.526207650273224,
+      "grad_norm": 0.9686909179139485,
+      "learning_rate": 1.278882782731644e-05,
+      "loss": 12.0099,
+      "step": 14446
+    },
+    {
+      "epoch": 2.5263825136612024,
+      "grad_norm": 0.8393849349109228,
+      "learning_rate": 1.2779590829968213e-05,
+      "loss": 12.1984,
+      "step": 14447
+    },
+    {
+      "epoch": 2.5265573770491803,
+      "grad_norm": 0.8690072739904925,
+      "learning_rate": 1.2770356941925654e-05,
+      "loss": 12.0359,
+      "step": 14448
+    },
+    {
+      "epoch": 2.5267322404371586,
+      "grad_norm": 0.8633159320865008,
+      "learning_rate": 1.2761126163517967e-05,
+      "loss": 12.1852,
+      "step": 14449
+    },
+    {
+      "epoch": 2.5269071038251365,
+      "grad_norm": 0.8334634319773465,
+      "learning_rate": 1.27518984950742e-05,
+      "loss": 12.2502,
+      "step": 14450
+    },
+    {
+      "epoch": 2.527081967213115,
+      "grad_norm": 0.9300191583486968,
+      "learning_rate": 1.2742673936923311e-05,
+      "loss": 12.0653,
+      "step": 14451
+    },
+    {
+      "epoch": 2.5272568306010927,
+      "grad_norm": 0.825743294204182,
+      "learning_rate": 1.2733452489394137e-05,
+      "loss": 12.0794,
+      "step": 14452
+    },
+    {
+      "epoch": 2.527431693989071,
+      "grad_norm": 0.868813587419344,
+      "learning_rate": 1.2724234152815439e-05,
+      "loss": 12.0776,
+      "step": 14453
+    },
+    {
+      "epoch": 2.527606557377049,
+      "grad_norm": 0.8310073727486854,
+      "learning_rate": 1.2715018927515831e-05,
+      "loss": 11.9211,
+      "step": 14454
+    },
+    {
+      "epoch": 2.5277814207650273,
+      "grad_norm": 0.795700258282017,
+      "learning_rate": 1.2705806813823807e-05,
+      "loss": 12.1038,
+      "step": 14455
+    },
+    {
+      "epoch": 2.5279562841530057,
+      "grad_norm": 0.8165015549621423,
+      "learning_rate": 1.2696597812067768e-05,
+      "loss": 12.0944,
+      "step": 14456
+    },
+    {
+      "epoch": 2.5281311475409836,
+      "grad_norm": 0.8186855029753786,
+      "learning_rate": 1.268739192257602e-05,
+      "loss": 11.7192,
+      "step": 14457
+    },
+    {
+      "epoch": 2.528306010928962,
+      "grad_norm": 0.8394294287553563,
+      "learning_rate": 1.2678189145676755e-05,
+      "loss": 11.9659,
+      "step": 14458
+    },
+    {
+      "epoch": 2.52848087431694,
+      "grad_norm": 0.8644911768848009,
+      "learning_rate": 1.2668989481698001e-05,
+      "loss": 12.0148,
+      "step": 14459
+    },
+    {
+      "epoch": 2.528655737704918,
+      "grad_norm": 0.8273262143267528,
+      "learning_rate": 1.265979293096774e-05,
+      "loss": 11.8751,
+      "step": 14460
+    },
+    {
+      "epoch": 2.528830601092896,
+      "grad_norm": 0.8332398751799902,
+      "learning_rate": 1.265059949381381e-05,
+      "loss": 12.1706,
+      "step": 14461
+    },
+    {
+      "epoch": 2.5290054644808744,
+      "grad_norm": 0.8531614777671932,
+      "learning_rate": 1.2641409170563967e-05,
+      "loss": 11.9569,
+      "step": 14462
+    },
+    {
+      "epoch": 2.5291803278688523,
+      "grad_norm": 0.8436440559895505,
+      "learning_rate": 1.263222196154581e-05,
+      "loss": 12.1339,
+      "step": 14463
+    },
+    {
+      "epoch": 2.5293551912568306,
+      "grad_norm": 0.8600438270239404,
+      "learning_rate": 1.2623037867086851e-05,
+      "loss": 12.0062,
+      "step": 14464
+    },
+    {
+      "epoch": 2.529530054644809,
+      "grad_norm": 0.8037752987879557,
+      "learning_rate": 1.2613856887514542e-05,
+      "loss": 11.7845,
+      "step": 14465
+    },
+    {
+      "epoch": 2.529704918032787,
+      "grad_norm": 0.7946145382165952,
+      "learning_rate": 1.2604679023156096e-05,
+      "loss": 12.1634,
+      "step": 14466
+    },
+    {
+      "epoch": 2.5298797814207648,
+      "grad_norm": 0.8264902911354237,
+      "learning_rate": 1.2595504274338743e-05,
+      "loss": 12.2105,
+      "step": 14467
+    },
+    {
+      "epoch": 2.530054644808743,
+      "grad_norm": 0.7623068917518191,
+      "learning_rate": 1.258633264138953e-05,
+      "loss": 12.1062,
+      "step": 14468
+    },
+    {
+      "epoch": 2.5302295081967214,
+      "grad_norm": 0.8042818990280584,
+      "learning_rate": 1.2577164124635432e-05,
+      "loss": 11.7544,
+      "step": 14469
+    },
+    {
+      "epoch": 2.5304043715846993,
+      "grad_norm": 0.8196893733267264,
+      "learning_rate": 1.2567998724403297e-05,
+      "loss": 11.8728,
+      "step": 14470
+    },
+    {
+      "epoch": 2.5305792349726777,
+      "grad_norm": 0.8775394885922082,
+      "learning_rate": 1.255883644101985e-05,
+      "loss": 12.048,
+      "step": 14471
+    },
+    {
+      "epoch": 2.5307540983606556,
+      "grad_norm": 0.8742902288898048,
+      "learning_rate": 1.2549677274811722e-05,
+      "loss": 12.1849,
+      "step": 14472
+    },
+    {
+      "epoch": 2.530928961748634,
+      "grad_norm": 0.738589214588346,
+      "learning_rate": 1.2540521226105439e-05,
+      "loss": 12.0504,
+      "step": 14473
+    },
+    {
+      "epoch": 2.5311038251366123,
+      "grad_norm": 0.8041078450996841,
+      "learning_rate": 1.2531368295227375e-05,
+      "loss": 11.9684,
+      "step": 14474
+    },
+    {
+      "epoch": 2.53127868852459,
+      "grad_norm": 0.810270245365187,
+      "learning_rate": 1.2522218482503834e-05,
+      "loss": 12.031,
+      "step": 14475
+    },
+    {
+      "epoch": 2.531453551912568,
+      "grad_norm": 0.8379670619269619,
+      "learning_rate": 1.2513071788260988e-05,
+      "loss": 12.0329,
+      "step": 14476
+    },
+    {
+      "epoch": 2.5316284153005464,
+      "grad_norm": 0.7805242785732378,
+      "learning_rate": 1.2503928212824945e-05,
+      "loss": 11.8642,
+      "step": 14477
+    },
+    {
+      "epoch": 2.5318032786885247,
+      "grad_norm": 0.8759866019885075,
+      "learning_rate": 1.2494787756521609e-05,
+      "loss": 12.0377,
+      "step": 14478
+    },
+    {
+      "epoch": 2.5319781420765026,
+      "grad_norm": 0.743488200020444,
+      "learning_rate": 1.2485650419676853e-05,
+      "loss": 11.8856,
+      "step": 14479
+    },
+    {
+      "epoch": 2.532153005464481,
+      "grad_norm": 0.8666460553404534,
+      "learning_rate": 1.2476516202616429e-05,
+      "loss": 12.0,
+      "step": 14480
+    },
+    {
+      "epoch": 2.532327868852459,
+      "grad_norm": 0.8731033458131752,
+      "learning_rate": 1.2467385105665918e-05,
+      "loss": 12.089,
+      "step": 14481
+    },
+    {
+      "epoch": 2.532502732240437,
+      "grad_norm": 0.8566561120915023,
+      "learning_rate": 1.2458257129150864e-05,
+      "loss": 12.0024,
+      "step": 14482
+    },
+    {
+      "epoch": 2.5326775956284155,
+      "grad_norm": 0.843453699352374,
+      "learning_rate": 1.2449132273396658e-05,
+      "loss": 11.9674,
+      "step": 14483
+    },
+    {
+      "epoch": 2.5328524590163934,
+      "grad_norm": 0.8412543406240792,
+      "learning_rate": 1.244001053872862e-05,
+      "loss": 12.0292,
+      "step": 14484
+    },
+    {
+      "epoch": 2.5330273224043713,
+      "grad_norm": 0.897391199440545,
+      "learning_rate": 1.2430891925471876e-05,
+      "loss": 11.9388,
+      "step": 14485
+    },
+    {
+      "epoch": 2.5332021857923497,
+      "grad_norm": 0.8744303826161771,
+      "learning_rate": 1.2421776433951526e-05,
+      "loss": 12.0928,
+      "step": 14486
+    },
+    {
+      "epoch": 2.533377049180328,
+      "grad_norm": 0.9025377396393691,
+      "learning_rate": 1.2412664064492518e-05,
+      "loss": 11.8941,
+      "step": 14487
+    },
+    {
+      "epoch": 2.533551912568306,
+      "grad_norm": 1.0702058429744357,
+      "learning_rate": 1.2403554817419705e-05,
+      "loss": 11.8588,
+      "step": 14488
+    },
+    {
+      "epoch": 2.5337267759562843,
+      "grad_norm": 0.8004986681455113,
+      "learning_rate": 1.2394448693057815e-05,
+      "loss": 11.5243,
+      "step": 14489
+    },
+    {
+      "epoch": 2.533901639344262,
+      "grad_norm": 0.8256661324656362,
+      "learning_rate": 1.2385345691731475e-05,
+      "loss": 11.9005,
+      "step": 14490
+    },
+    {
+      "epoch": 2.5340765027322405,
+      "grad_norm": 0.9213603920757508,
+      "learning_rate": 1.2376245813765186e-05,
+      "loss": 11.8911,
+      "step": 14491
+    },
+    {
+      "epoch": 2.534251366120219,
+      "grad_norm": 0.84791638278136,
+      "learning_rate": 1.2367149059483385e-05,
+      "loss": 12.1742,
+      "step": 14492
+    },
+    {
+      "epoch": 2.5344262295081967,
+      "grad_norm": 0.8105543203691473,
+      "learning_rate": 1.2358055429210314e-05,
+      "loss": 11.8822,
+      "step": 14493
+    },
+    {
+      "epoch": 2.5346010928961746,
+      "grad_norm": 0.847642008788115,
+      "learning_rate": 1.2348964923270168e-05,
+      "loss": 12.0728,
+      "step": 14494
+    },
+    {
+      "epoch": 2.534775956284153,
+      "grad_norm": 0.8424582664234572,
+      "learning_rate": 1.2339877541987022e-05,
+      "loss": 11.8528,
+      "step": 14495
+    },
+    {
+      "epoch": 2.5349508196721313,
+      "grad_norm": 0.8392861204167138,
+      "learning_rate": 1.2330793285684816e-05,
+      "loss": 12.0416,
+      "step": 14496
+    },
+    {
+      "epoch": 2.535125683060109,
+      "grad_norm": 0.8805431909662275,
+      "learning_rate": 1.2321712154687392e-05,
+      "loss": 12.0776,
+      "step": 14497
+    },
+    {
+      "epoch": 2.5353005464480876,
+      "grad_norm": 0.8287265479852349,
+      "learning_rate": 1.2312634149318491e-05,
+      "loss": 12.0286,
+      "step": 14498
+    },
+    {
+      "epoch": 2.5354754098360655,
+      "grad_norm": 0.8674648885725158,
+      "learning_rate": 1.2303559269901743e-05,
+      "loss": 12.0221,
+      "step": 14499
+    },
+    {
+      "epoch": 2.535650273224044,
+      "grad_norm": 0.867380635697549,
+      "learning_rate": 1.2294487516760634e-05,
+      "loss": 11.8856,
+      "step": 14500
+    },
+    {
+      "epoch": 2.5359562841530057,
+      "grad_norm": 0.7797692159009011,
+      "learning_rate": 1.228541889021857e-05,
+      "loss": 12.0909,
+      "step": 14501
+    },
+    {
+      "epoch": 2.5361311475409836,
+      "grad_norm": 0.7770621550045421,
+      "learning_rate": 1.2276353390598838e-05,
+      "loss": 11.7776,
+      "step": 14502
+    },
+    {
+      "epoch": 2.536306010928962,
+      "grad_norm": 0.8146343102757505,
+      "learning_rate": 1.2267291018224636e-05,
+      "loss": 11.9686,
+      "step": 14503
+    },
+    {
+      "epoch": 2.53648087431694,
+      "grad_norm": 0.7837351261708863,
+      "learning_rate": 1.2258231773418983e-05,
+      "loss": 11.7081,
+      "step": 14504
+    },
+    {
+      "epoch": 2.536655737704918,
+      "grad_norm": 0.785458815420535,
+      "learning_rate": 1.224917565650484e-05,
+      "loss": 11.83,
+      "step": 14505
+    },
+    {
+      "epoch": 2.536830601092896,
+      "grad_norm": 0.7428561220784052,
+      "learning_rate": 1.2240122667805077e-05,
+      "loss": 11.6636,
+      "step": 14506
+    },
+    {
+      "epoch": 2.5370054644808744,
+      "grad_norm": 0.764927681042228,
+      "learning_rate": 1.2231072807642397e-05,
+      "loss": 11.7668,
+      "step": 14507
+    },
+    {
+      "epoch": 2.5371803278688523,
+      "grad_norm": 0.8183350585251268,
+      "learning_rate": 1.222202607633942e-05,
+      "loss": 12.1238,
+      "step": 14508
+    },
+    {
+      "epoch": 2.5373551912568306,
+      "grad_norm": 0.840133540931417,
+      "learning_rate": 1.2212982474218659e-05,
+      "loss": 11.9926,
+      "step": 14509
+    },
+    {
+      "epoch": 2.537530054644809,
+      "grad_norm": 0.775244490601205,
+      "learning_rate": 1.220394200160253e-05,
+      "loss": 11.8872,
+      "step": 14510
+    },
+    {
+      "epoch": 2.537704918032787,
+      "grad_norm": 0.742732346734578,
+      "learning_rate": 1.2194904658813266e-05,
+      "loss": 11.9065,
+      "step": 14511
+    },
+    {
+      "epoch": 2.5378797814207648,
+      "grad_norm": 0.8389808712849172,
+      "learning_rate": 1.2185870446173064e-05,
+      "loss": 12.0319,
+      "step": 14512
+    },
+    {
+      "epoch": 2.538054644808743,
+      "grad_norm": 0.8203912618513468,
+      "learning_rate": 1.2176839364003978e-05,
+      "loss": 12.1031,
+      "step": 14513
+    },
+    {
+      "epoch": 2.5382295081967214,
+      "grad_norm": 0.7742101225917355,
+      "learning_rate": 1.2167811412627983e-05,
+      "loss": 11.7531,
+      "step": 14514
+    },
+    {
+      "epoch": 2.5384043715846993,
+      "grad_norm": 0.8536933157323605,
+      "learning_rate": 1.2158786592366867e-05,
+      "loss": 11.6023,
+      "step": 14515
+    },
+    {
+      "epoch": 2.5385792349726777,
+      "grad_norm": 0.8104633131631888,
+      "learning_rate": 1.214976490354237e-05,
+      "loss": 11.9692,
+      "step": 14516
+    },
+    {
+      "epoch": 2.5387540983606556,
+      "grad_norm": 0.8044530478749136,
+      "learning_rate": 1.2140746346476128e-05,
+      "loss": 12.0329,
+      "step": 14517
+    },
+    {
+      "epoch": 2.538928961748634,
+      "grad_norm": 0.7811600237117143,
+      "learning_rate": 1.2131730921489637e-05,
+      "loss": 11.8399,
+      "step": 14518
+    },
+    {
+      "epoch": 2.5391038251366123,
+      "grad_norm": 0.8388040723414867,
+      "learning_rate": 1.2122718628904251e-05,
+      "loss": 11.7738,
+      "step": 14519
+    },
+    {
+      "epoch": 2.53927868852459,
+      "grad_norm": 0.8168996494217835,
+      "learning_rate": 1.211370946904128e-05,
+      "loss": 11.898,
+      "step": 14520
+    },
+    {
+      "epoch": 2.539453551912568,
+      "grad_norm": 0.8629703925573174,
+      "learning_rate": 1.2104703442221876e-05,
+      "loss": 11.9348,
+      "step": 14521
+    },
+    {
+      "epoch": 2.5396284153005464,
+      "grad_norm": 0.8552030496302867,
+      "learning_rate": 1.2095700548767109e-05,
+      "loss": 12.0906,
+      "step": 14522
+    },
+    {
+      "epoch": 2.5398032786885247,
+      "grad_norm": 0.8072135457053555,
+      "learning_rate": 1.2086700788997906e-05,
+      "loss": 11.8729,
+      "step": 14523
+    },
+    {
+      "epoch": 2.5399781420765026,
+      "grad_norm": 0.82471300921939,
+      "learning_rate": 1.2077704163235105e-05,
+      "loss": 12.0083,
+      "step": 14524
+    },
+    {
+      "epoch": 2.540153005464481,
+      "grad_norm": 0.8200850387276617,
+      "learning_rate": 1.2068710671799443e-05,
+      "loss": 11.8478,
+      "step": 14525
+    },
+    {
+      "epoch": 2.540327868852459,
+      "grad_norm": 0.8232829301500321,
+      "learning_rate": 1.2059720315011491e-05,
+      "loss": 11.7437,
+      "step": 14526
+    },
+    {
+      "epoch": 2.540502732240437,
+      "grad_norm": 0.7970651073824148,
+      "learning_rate": 1.2050733093191757e-05,
+      "loss": 11.9609,
+      "step": 14527
+    },
+    {
+      "epoch": 2.5406775956284156,
+      "grad_norm": 0.85898675857805,
+      "learning_rate": 1.2041749006660641e-05,
+      "loss": 11.6343,
+      "step": 14528
+    },
+    {
+      "epoch": 2.5408524590163934,
+      "grad_norm": 0.7978528069065047,
+      "learning_rate": 1.2032768055738419e-05,
+      "loss": 11.6495,
+      "step": 14529
+    },
+    {
+      "epoch": 2.5410273224043713,
+      "grad_norm": 0.8166728441442744,
+      "learning_rate": 1.2023790240745215e-05,
+      "loss": 11.7375,
+      "step": 14530
+    },
+    {
+      "epoch": 2.5412021857923497,
+      "grad_norm": 0.7714675701999761,
+      "learning_rate": 1.2014815562001102e-05,
+      "loss": 11.7273,
+      "step": 14531
+    },
+    {
+      "epoch": 2.541377049180328,
+      "grad_norm": 0.8285108708288393,
+      "learning_rate": 1.2005844019826017e-05,
+      "loss": 11.9085,
+      "step": 14532
+    },
+    {
+      "epoch": 2.541551912568306,
+      "grad_norm": 0.8885203175753955,
+      "learning_rate": 1.1996875614539793e-05,
+      "loss": 11.9244,
+      "step": 14533
+    },
+    {
+      "epoch": 2.5417267759562843,
+      "grad_norm": 0.8080974447429445,
+      "learning_rate": 1.1987910346462129e-05,
+      "loss": 11.5435,
+      "step": 14534
+    },
+    {
+      "epoch": 2.541901639344262,
+      "grad_norm": 0.8599373676497108,
+      "learning_rate": 1.1978948215912623e-05,
+      "loss": 11.8212,
+      "step": 14535
+    },
+    {
+      "epoch": 2.5420765027322405,
+      "grad_norm": 0.8067474374503738,
+      "learning_rate": 1.1969989223210765e-05,
+      "loss": 12.1408,
+      "step": 14536
+    },
+    {
+      "epoch": 2.542251366120219,
+      "grad_norm": 0.8537149230161587,
+      "learning_rate": 1.1961033368675967e-05,
+      "loss": 11.9557,
+      "step": 14537
+    },
+    {
+      "epoch": 2.5424262295081967,
+      "grad_norm": 0.8070928328755395,
+      "learning_rate": 1.195208065262744e-05,
+      "loss": 12.0198,
+      "step": 14538
+    },
+    {
+      "epoch": 2.5426010928961746,
+      "grad_norm": 0.8291855064794004,
+      "learning_rate": 1.1943131075384361e-05,
+      "loss": 11.9496,
+      "step": 14539
+    },
+    {
+      "epoch": 2.542775956284153,
+      "grad_norm": 0.7396726870039485,
+      "learning_rate": 1.1934184637265788e-05,
+      "loss": 11.6694,
+      "step": 14540
+    },
+    {
+      "epoch": 2.5429508196721313,
+      "grad_norm": 0.8439060427514061,
+      "learning_rate": 1.192524133859063e-05,
+      "loss": 11.9073,
+      "step": 14541
+    },
+    {
+      "epoch": 2.543125683060109,
+      "grad_norm": 0.7912868568981541,
+      "learning_rate": 1.1916301179677725e-05,
+      "loss": 11.8838,
+      "step": 14542
+    },
+    {
+      "epoch": 2.5433005464480876,
+      "grad_norm": 0.8465664460482419,
+      "learning_rate": 1.1907364160845758e-05,
+      "loss": 11.8978,
+      "step": 14543
+    },
+    {
+      "epoch": 2.5434754098360655,
+      "grad_norm": 0.8123809994625503,
+      "learning_rate": 1.1898430282413353e-05,
+      "loss": 11.893,
+      "step": 14544
+    },
+    {
+      "epoch": 2.543650273224044,
+      "grad_norm": 0.8509467454347385,
+      "learning_rate": 1.1889499544698956e-05,
+      "loss": 12.1226,
+      "step": 14545
+    },
+    {
+      "epoch": 2.5438251366120217,
+      "grad_norm": 0.856845106307132,
+      "learning_rate": 1.1880571948020957e-05,
+      "loss": 11.7796,
+      "step": 14546
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.8124371909806221,
+      "learning_rate": 1.1871647492697614e-05,
+      "loss": 11.991,
+      "step": 14547
+    },
+    {
+      "epoch": 2.544174863387978,
+      "grad_norm": 0.8415931845871354,
+      "learning_rate": 1.1862726179047078e-05,
+      "loss": 11.9099,
+      "step": 14548
+    },
+    {
+      "epoch": 2.5443497267759563,
+      "grad_norm": 0.9162518697955981,
+      "learning_rate": 1.1853808007387367e-05,
+      "loss": 12.1552,
+      "step": 14549
+    },
+    {
+      "epoch": 2.5445245901639346,
+      "grad_norm": 0.8265165593218117,
+      "learning_rate": 1.1844892978036415e-05,
+      "loss": 11.8677,
+      "step": 14550
+    },
+    {
+      "epoch": 2.5446994535519125,
+      "grad_norm": 1.0017436454623914,
+      "learning_rate": 1.183598109131202e-05,
+      "loss": 11.9677,
+      "step": 14551
+    },
+    {
+      "epoch": 2.544874316939891,
+      "grad_norm": 0.8320601714080034,
+      "learning_rate": 1.182707234753192e-05,
+      "loss": 11.9588,
+      "step": 14552
+    },
+    {
+      "epoch": 2.5450491803278688,
+      "grad_norm": 0.8699675104270459,
+      "learning_rate": 1.1818166747013649e-05,
+      "loss": 11.9202,
+      "step": 14553
+    },
+    {
+      "epoch": 2.545224043715847,
+      "grad_norm": 0.8225406631352061,
+      "learning_rate": 1.1809264290074695e-05,
+      "loss": 12.0484,
+      "step": 14554
+    },
+    {
+      "epoch": 2.545398907103825,
+      "grad_norm": 0.8179765440874924,
+      "learning_rate": 1.1800364977032463e-05,
+      "loss": 11.9376,
+      "step": 14555
+    },
+    {
+      "epoch": 2.5455737704918033,
+      "grad_norm": 0.8006187743369648,
+      "learning_rate": 1.1791468808204142e-05,
+      "loss": 11.8346,
+      "step": 14556
+    },
+    {
+      "epoch": 2.5457486338797812,
+      "grad_norm": 0.8092803532013677,
+      "learning_rate": 1.1782575783906902e-05,
+      "loss": 11.8542,
+      "step": 14557
+    },
+    {
+      "epoch": 2.5459234972677596,
+      "grad_norm": 0.914118850720199,
+      "learning_rate": 1.1773685904457754e-05,
+      "loss": 11.7387,
+      "step": 14558
+    },
+    {
+      "epoch": 2.546098360655738,
+      "grad_norm": 0.7979557635556724,
+      "learning_rate": 1.1764799170173636e-05,
+      "loss": 11.9438,
+      "step": 14559
+    },
+    {
+      "epoch": 2.546273224043716,
+      "grad_norm": 0.7987204018908702,
+      "learning_rate": 1.1755915581371324e-05,
+      "loss": 11.8038,
+      "step": 14560
+    },
+    {
+      "epoch": 2.546448087431694,
+      "grad_norm": 0.814080534821857,
+      "learning_rate": 1.1747035138367523e-05,
+      "loss": 11.9019,
+      "step": 14561
+    },
+    {
+      "epoch": 2.546622950819672,
+      "grad_norm": 0.8264993034169837,
+      "learning_rate": 1.1738157841478815e-05,
+      "loss": 11.8668,
+      "step": 14562
+    },
+    {
+      "epoch": 2.5467978142076504,
+      "grad_norm": 0.7721054851206397,
+      "learning_rate": 1.1729283691021664e-05,
+      "loss": 11.9678,
+      "step": 14563
+    },
+    {
+      "epoch": 2.5469726775956283,
+      "grad_norm": 0.8622423556664361,
+      "learning_rate": 1.1720412687312398e-05,
+      "loss": 11.8356,
+      "step": 14564
+    },
+    {
+      "epoch": 2.5471475409836066,
+      "grad_norm": 0.8596774690000157,
+      "learning_rate": 1.1711544830667286e-05,
+      "loss": 11.9822,
+      "step": 14565
+    },
+    {
+      "epoch": 2.5473224043715845,
+      "grad_norm": 0.8306942766754448,
+      "learning_rate": 1.1702680121402443e-05,
+      "loss": 11.7155,
+      "step": 14566
+    },
+    {
+      "epoch": 2.547497267759563,
+      "grad_norm": 0.8500045601628675,
+      "learning_rate": 1.1693818559833903e-05,
+      "loss": 11.631,
+      "step": 14567
+    },
+    {
+      "epoch": 2.547672131147541,
+      "grad_norm": 0.7956007482812069,
+      "learning_rate": 1.1684960146277546e-05,
+      "loss": 12.1707,
+      "step": 14568
+    },
+    {
+      "epoch": 2.547846994535519,
+      "grad_norm": 0.8391576200793512,
+      "learning_rate": 1.1676104881049166e-05,
+      "loss": 12.0124,
+      "step": 14569
+    },
+    {
+      "epoch": 2.548021857923497,
+      "grad_norm": 0.8476879263220316,
+      "learning_rate": 1.1667252764464453e-05,
+      "loss": 12.0041,
+      "step": 14570
+    },
+    {
+      "epoch": 2.5481967213114753,
+      "grad_norm": 0.804617422001917,
+      "learning_rate": 1.1658403796839001e-05,
+      "loss": 12.0936,
+      "step": 14571
+    },
+    {
+      "epoch": 2.5483715846994537,
+      "grad_norm": 0.779066030826822,
+      "learning_rate": 1.1649557978488212e-05,
+      "loss": 11.9368,
+      "step": 14572
+    },
+    {
+      "epoch": 2.5485464480874316,
+      "grad_norm": 0.8251433000993806,
+      "learning_rate": 1.1640715309727456e-05,
+      "loss": 11.6995,
+      "step": 14573
+    },
+    {
+      "epoch": 2.54872131147541,
+      "grad_norm": 0.8042577085071579,
+      "learning_rate": 1.1631875790871982e-05,
+      "loss": 11.7579,
+      "step": 14574
+    },
+    {
+      "epoch": 2.548896174863388,
+      "grad_norm": 0.805361181544143,
+      "learning_rate": 1.162303942223687e-05,
+      "loss": 12.1495,
+      "step": 14575
+    },
+    {
+      "epoch": 2.549071038251366,
+      "grad_norm": 0.7631949504257614,
+      "learning_rate": 1.1614206204137135e-05,
+      "loss": 12.07,
+      "step": 14576
+    },
+    {
+      "epoch": 2.5492459016393445,
+      "grad_norm": 0.8400308361767597,
+      "learning_rate": 1.1605376136887692e-05,
+      "loss": 11.792,
+      "step": 14577
+    },
+    {
+      "epoch": 2.5494207650273224,
+      "grad_norm": 0.8176795211675949,
+      "learning_rate": 1.159654922080331e-05,
+      "loss": 11.9314,
+      "step": 14578
+    },
+    {
+      "epoch": 2.5495956284153003,
+      "grad_norm": 0.7753691622092185,
+      "learning_rate": 1.1587725456198661e-05,
+      "loss": 11.8523,
+      "step": 14579
+    },
+    {
+      "epoch": 2.5497704918032786,
+      "grad_norm": 0.8927920275745432,
+      "learning_rate": 1.1578904843388294e-05,
+      "loss": 11.9814,
+      "step": 14580
+    },
+    {
+      "epoch": 2.549945355191257,
+      "grad_norm": 0.8004589029273571,
+      "learning_rate": 1.1570087382686667e-05,
+      "loss": 11.8816,
+      "step": 14581
+    },
+    {
+      "epoch": 2.550120218579235,
+      "grad_norm": 0.8609463759258136,
+      "learning_rate": 1.156127307440813e-05,
+      "loss": 11.7798,
+      "step": 14582
+    },
+    {
+      "epoch": 2.550295081967213,
+      "grad_norm": 0.8302258640869294,
+      "learning_rate": 1.1552461918866852e-05,
+      "loss": 12.1309,
+      "step": 14583
+    },
+    {
+      "epoch": 2.550469945355191,
+      "grad_norm": 0.8235983752111492,
+      "learning_rate": 1.1543653916376962e-05,
+      "loss": 12.0717,
+      "step": 14584
+    },
+    {
+      "epoch": 2.5506448087431695,
+      "grad_norm": 0.876994533935794,
+      "learning_rate": 1.1534849067252474e-05,
+      "loss": 11.7042,
+      "step": 14585
+    },
+    {
+      "epoch": 2.550819672131148,
+      "grad_norm": 0.7926456462607278,
+      "learning_rate": 1.1526047371807258e-05,
+      "loss": 12.0031,
+      "step": 14586
+    },
+    {
+      "epoch": 2.5509945355191257,
+      "grad_norm": 0.8018879299565222,
+      "learning_rate": 1.1517248830355077e-05,
+      "loss": 11.7068,
+      "step": 14587
+    },
+    {
+      "epoch": 2.5511693989071036,
+      "grad_norm": 0.7991017929025095,
+      "learning_rate": 1.1508453443209588e-05,
+      "loss": 11.8276,
+      "step": 14588
+    },
+    {
+      "epoch": 2.551344262295082,
+      "grad_norm": 0.8306937056050937,
+      "learning_rate": 1.1499661210684364e-05,
+      "loss": 11.6077,
+      "step": 14589
+    },
+    {
+      "epoch": 2.5515191256830603,
+      "grad_norm": 0.8298703451456056,
+      "learning_rate": 1.149087213309279e-05,
+      "loss": 11.7283,
+      "step": 14590
+    },
+    {
+      "epoch": 2.551693989071038,
+      "grad_norm": 0.8396017451669641,
+      "learning_rate": 1.1482086210748223e-05,
+      "loss": 11.9786,
+      "step": 14591
+    },
+    {
+      "epoch": 2.5518688524590165,
+      "grad_norm": 0.7896172109466267,
+      "learning_rate": 1.1473303443963856e-05,
+      "loss": 11.9652,
+      "step": 14592
+    },
+    {
+      "epoch": 2.5520437158469944,
+      "grad_norm": 0.7789745269526177,
+      "learning_rate": 1.1464523833052809e-05,
+      "loss": 11.7866,
+      "step": 14593
+    },
+    {
+      "epoch": 2.5522185792349728,
+      "grad_norm": 0.8244460195346037,
+      "learning_rate": 1.145574737832802e-05,
+      "loss": 12.1248,
+      "step": 14594
+    },
+    {
+      "epoch": 2.552393442622951,
+      "grad_norm": 0.8429062601328134,
+      "learning_rate": 1.1446974080102391e-05,
+      "loss": 11.9806,
+      "step": 14595
+    },
+    {
+      "epoch": 2.552568306010929,
+      "grad_norm": 0.8075163739755024,
+      "learning_rate": 1.1438203938688675e-05,
+      "loss": 11.9959,
+      "step": 14596
+    },
+    {
+      "epoch": 2.552743169398907,
+      "grad_norm": 0.7991103338525558,
+      "learning_rate": 1.1429436954399508e-05,
+      "loss": 11.6821,
+      "step": 14597
+    },
+    {
+      "epoch": 2.5529180327868852,
+      "grad_norm": 0.8378372809554719,
+      "learning_rate": 1.1420673127547433e-05,
+      "loss": 11.685,
+      "step": 14598
+    },
+    {
+      "epoch": 2.5530928961748636,
+      "grad_norm": 0.8062924938266526,
+      "learning_rate": 1.1411912458444863e-05,
+      "loss": 11.9536,
+      "step": 14599
+    },
+    {
+      "epoch": 2.5532677595628415,
+      "grad_norm": 0.8869309793963901,
+      "learning_rate": 1.1403154947404116e-05,
+      "loss": 11.9039,
+      "step": 14600
+    },
+    {
+      "epoch": 2.55344262295082,
+      "grad_norm": 0.7651161373480635,
+      "learning_rate": 1.1394400594737398e-05,
+      "loss": 12.0736,
+      "step": 14601
+    },
+    {
+      "epoch": 2.5536174863387977,
+      "grad_norm": 0.7815697342416787,
+      "learning_rate": 1.138564940075676e-05,
+      "loss": 11.8441,
+      "step": 14602
+    },
+    {
+      "epoch": 2.553792349726776,
+      "grad_norm": 0.7979295773593003,
+      "learning_rate": 1.1376901365774174e-05,
+      "loss": 12.0684,
+      "step": 14603
+    },
+    {
+      "epoch": 2.553967213114754,
+      "grad_norm": 0.8310071682747306,
+      "learning_rate": 1.1368156490101544e-05,
+      "loss": 11.7859,
+      "step": 14604
+    },
+    {
+      "epoch": 2.5541420765027323,
+      "grad_norm": 0.830376536930217,
+      "learning_rate": 1.1359414774050547e-05,
+      "loss": 12.0092,
+      "step": 14605
+    },
+    {
+      "epoch": 2.55431693989071,
+      "grad_norm": 0.8072344041670004,
+      "learning_rate": 1.1350676217932854e-05,
+      "loss": 11.6749,
+      "step": 14606
+    },
+    {
+      "epoch": 2.5544918032786885,
+      "grad_norm": 0.821738094771203,
+      "learning_rate": 1.1341940822059982e-05,
+      "loss": 11.8863,
+      "step": 14607
+    },
+    {
+      "epoch": 2.554666666666667,
+      "grad_norm": 0.8072670285161213,
+      "learning_rate": 1.1333208586743348e-05,
+      "loss": 11.8224,
+      "step": 14608
+    },
+    {
+      "epoch": 2.5548415300546448,
+      "grad_norm": 0.8444583476323573,
+      "learning_rate": 1.1324479512294206e-05,
+      "loss": 11.6735,
+      "step": 14609
+    },
+    {
+      "epoch": 2.555016393442623,
+      "grad_norm": 0.8227759153608477,
+      "learning_rate": 1.1315753599023771e-05,
+      "loss": 12.0171,
+      "step": 14610
+    },
+    {
+      "epoch": 2.555191256830601,
+      "grad_norm": 0.8394482880656998,
+      "learning_rate": 1.1307030847243105e-05,
+      "loss": 11.8719,
+      "step": 14611
+    },
+    {
+      "epoch": 2.5553661202185793,
+      "grad_norm": 0.8417850382216504,
+      "learning_rate": 1.129831125726316e-05,
+      "loss": 11.689,
+      "step": 14612
+    },
+    {
+      "epoch": 2.5555409836065572,
+      "grad_norm": 0.7797728578357307,
+      "learning_rate": 1.1289594829394778e-05,
+      "loss": 11.7078,
+      "step": 14613
+    },
+    {
+      "epoch": 2.5557158469945356,
+      "grad_norm": 0.838594786544878,
+      "learning_rate": 1.1280881563948697e-05,
+      "loss": 11.9611,
+      "step": 14614
+    },
+    {
+      "epoch": 2.5558907103825135,
+      "grad_norm": 0.8180028253715357,
+      "learning_rate": 1.1272171461235536e-05,
+      "loss": 12.0065,
+      "step": 14615
+    },
+    {
+      "epoch": 2.556065573770492,
+      "grad_norm": 0.8800831630345267,
+      "learning_rate": 1.1263464521565814e-05,
+      "loss": 11.8387,
+      "step": 14616
+    },
+    {
+      "epoch": 2.55624043715847,
+      "grad_norm": 0.8475699268030322,
+      "learning_rate": 1.1254760745249882e-05,
+      "loss": 11.6624,
+      "step": 14617
+    },
+    {
+      "epoch": 2.556415300546448,
+      "grad_norm": 0.8786648838443416,
+      "learning_rate": 1.1246060132598035e-05,
+      "loss": 11.9004,
+      "step": 14618
+    },
+    {
+      "epoch": 2.556590163934426,
+      "grad_norm": 0.8379300577378556,
+      "learning_rate": 1.1237362683920483e-05,
+      "loss": 12.0748,
+      "step": 14619
+    },
+    {
+      "epoch": 2.5567650273224043,
+      "grad_norm": 0.7902829876283236,
+      "learning_rate": 1.1228668399527209e-05,
+      "loss": 12.0248,
+      "step": 14620
+    },
+    {
+      "epoch": 2.5569398907103826,
+      "grad_norm": 0.8384902140989414,
+      "learning_rate": 1.1219977279728189e-05,
+      "loss": 12.0351,
+      "step": 14621
+    },
+    {
+      "epoch": 2.5571147540983605,
+      "grad_norm": 0.9015477458728895,
+      "learning_rate": 1.1211289324833251e-05,
+      "loss": 11.9167,
+      "step": 14622
+    },
+    {
+      "epoch": 2.557289617486339,
+      "grad_norm": 0.7850329230993469,
+      "learning_rate": 1.1202604535152139e-05,
+      "loss": 11.9989,
+      "step": 14623
+    },
+    {
+      "epoch": 2.5574644808743168,
+      "grad_norm": 0.8815400473409242,
+      "learning_rate": 1.1193922910994404e-05,
+      "loss": 11.8013,
+      "step": 14624
+    },
+    {
+      "epoch": 2.557639344262295,
+      "grad_norm": 0.8022486160202894,
+      "learning_rate": 1.1185244452669552e-05,
+      "loss": 11.7385,
+      "step": 14625
+    },
+    {
+      "epoch": 2.5578142076502735,
+      "grad_norm": 0.821413861579686,
+      "learning_rate": 1.1176569160486971e-05,
+      "loss": 11.8194,
+      "step": 14626
+    },
+    {
+      "epoch": 2.5579890710382513,
+      "grad_norm": 0.8257125481470793,
+      "learning_rate": 1.1167897034755936e-05,
+      "loss": 11.7838,
+      "step": 14627
+    },
+    {
+      "epoch": 2.5581639344262292,
+      "grad_norm": 0.7901943695877518,
+      "learning_rate": 1.1159228075785566e-05,
+      "loss": 12.0259,
+      "step": 14628
+    },
+    {
+      "epoch": 2.5583387978142076,
+      "grad_norm": 0.8119804288446171,
+      "learning_rate": 1.1150562283884914e-05,
+      "loss": 12.1007,
+      "step": 14629
+    },
+    {
+      "epoch": 2.558513661202186,
+      "grad_norm": 0.8014439068385333,
+      "learning_rate": 1.114189965936291e-05,
+      "loss": 11.8184,
+      "step": 14630
+    },
+    {
+      "epoch": 2.558688524590164,
+      "grad_norm": 0.7940636738972774,
+      "learning_rate": 1.1133240202528361e-05,
+      "loss": 11.9541,
+      "step": 14631
+    },
+    {
+      "epoch": 2.558863387978142,
+      "grad_norm": 0.8390644149940556,
+      "learning_rate": 1.1124583913689968e-05,
+      "loss": 11.9164,
+      "step": 14632
+    },
+    {
+      "epoch": 2.55903825136612,
+      "grad_norm": 0.8476593006489656,
+      "learning_rate": 1.1115930793156326e-05,
+      "loss": 11.6245,
+      "step": 14633
+    },
+    {
+      "epoch": 2.5592131147540984,
+      "grad_norm": 0.7774504800714795,
+      "learning_rate": 1.1107280841235911e-05,
+      "loss": 11.5191,
+      "step": 14634
+    },
+    {
+      "epoch": 2.5593879781420767,
+      "grad_norm": 0.848800346754076,
+      "learning_rate": 1.1098634058237056e-05,
+      "loss": 11.9455,
+      "step": 14635
+    },
+    {
+      "epoch": 2.5595628415300546,
+      "grad_norm": 0.8081949251934964,
+      "learning_rate": 1.1089990444468035e-05,
+      "loss": 11.865,
+      "step": 14636
+    },
+    {
+      "epoch": 2.5597377049180325,
+      "grad_norm": 0.8337447686848433,
+      "learning_rate": 1.1081350000236968e-05,
+      "loss": 11.4225,
+      "step": 14637
+    },
+    {
+      "epoch": 2.559912568306011,
+      "grad_norm": 0.8390383262273197,
+      "learning_rate": 1.1072712725851909e-05,
+      "loss": 11.879,
+      "step": 14638
+    },
+    {
+      "epoch": 2.5600874316939892,
+      "grad_norm": 0.7631371656906987,
+      "learning_rate": 1.106407862162071e-05,
+      "loss": 11.6426,
+      "step": 14639
+    },
+    {
+      "epoch": 2.560262295081967,
+      "grad_norm": 0.7874409287389341,
+      "learning_rate": 1.1055447687851205e-05,
+      "loss": 11.8004,
+      "step": 14640
+    },
+    {
+      "epoch": 2.5604371584699455,
+      "grad_norm": 0.8006020905587778,
+      "learning_rate": 1.1046819924851071e-05,
+      "loss": 11.9151,
+      "step": 14641
+    },
+    {
+      "epoch": 2.5606120218579234,
+      "grad_norm": 0.9173789888125422,
+      "learning_rate": 1.1038195332927892e-05,
+      "loss": 11.9993,
+      "step": 14642
+    },
+    {
+      "epoch": 2.5607868852459017,
+      "grad_norm": 0.8905578333997873,
+      "learning_rate": 1.1029573912389091e-05,
+      "loss": 11.7562,
+      "step": 14643
+    },
+    {
+      "epoch": 2.56096174863388,
+      "grad_norm": 0.7990851839831199,
+      "learning_rate": 1.1020955663542031e-05,
+      "loss": 11.9176,
+      "step": 14644
+    },
+    {
+      "epoch": 2.561136612021858,
+      "grad_norm": 0.833810018273656,
+      "learning_rate": 1.1012340586693936e-05,
+      "loss": 11.8328,
+      "step": 14645
+    },
+    {
+      "epoch": 2.561311475409836,
+      "grad_norm": 0.8197725963128151,
+      "learning_rate": 1.1003728682151947e-05,
+      "loss": 11.9265,
+      "step": 14646
+    },
+    {
+      "epoch": 2.561486338797814,
+      "grad_norm": 0.8203973095472009,
+      "learning_rate": 1.0995119950223042e-05,
+      "loss": 11.9396,
+      "step": 14647
+    },
+    {
+      "epoch": 2.5616612021857925,
+      "grad_norm": 0.8427940585606543,
+      "learning_rate": 1.0986514391214109e-05,
+      "loss": 11.7194,
+      "step": 14648
+    },
+    {
+      "epoch": 2.5618360655737704,
+      "grad_norm": 0.7790611544996754,
+      "learning_rate": 1.0977912005431934e-05,
+      "loss": 11.7852,
+      "step": 14649
+    },
+    {
+      "epoch": 2.5620109289617488,
+      "grad_norm": 0.8195002852168047,
+      "learning_rate": 1.0969312793183195e-05,
+      "loss": 11.7011,
+      "step": 14650
+    },
+    {
+      "epoch": 2.5621857923497267,
+      "grad_norm": 0.7877627237555451,
+      "learning_rate": 1.0960716754774426e-05,
+      "loss": 11.7319,
+      "step": 14651
+    },
+    {
+      "epoch": 2.562360655737705,
+      "grad_norm": 0.861218651267365,
+      "learning_rate": 1.0952123890512078e-05,
+      "loss": 11.848,
+      "step": 14652
+    },
+    {
+      "epoch": 2.5625355191256833,
+      "grad_norm": 0.8210433362946695,
+      "learning_rate": 1.0943534200702488e-05,
+      "loss": 11.9795,
+      "step": 14653
+    },
+    {
+      "epoch": 2.5627103825136612,
+      "grad_norm": 0.8302142965676339,
+      "learning_rate": 1.093494768565183e-05,
+      "loss": 11.8919,
+      "step": 14654
+    },
+    {
+      "epoch": 2.562885245901639,
+      "grad_norm": 0.8235987807906904,
+      "learning_rate": 1.0926364345666229e-05,
+      "loss": 11.8382,
+      "step": 14655
+    },
+    {
+      "epoch": 2.5630601092896175,
+      "grad_norm": 0.8549096527249336,
+      "learning_rate": 1.0917784181051672e-05,
+      "loss": 11.8793,
+      "step": 14656
+    },
+    {
+      "epoch": 2.563234972677596,
+      "grad_norm": 0.817708745897742,
+      "learning_rate": 1.0909207192114047e-05,
+      "loss": 11.8651,
+      "step": 14657
+    },
+    {
+      "epoch": 2.5634098360655737,
+      "grad_norm": 0.8277378423369663,
+      "learning_rate": 1.0900633379159065e-05,
+      "loss": 11.736,
+      "step": 14658
+    },
+    {
+      "epoch": 2.563584699453552,
+      "grad_norm": 0.8613108059326163,
+      "learning_rate": 1.0892062742492415e-05,
+      "loss": 11.9539,
+      "step": 14659
+    },
+    {
+      "epoch": 2.56375956284153,
+      "grad_norm": 0.9026000231125532,
+      "learning_rate": 1.0883495282419609e-05,
+      "loss": 11.8023,
+      "step": 14660
+    },
+    {
+      "epoch": 2.5639344262295083,
+      "grad_norm": 0.9115478546611142,
+      "learning_rate": 1.0874930999246091e-05,
+      "loss": 12.1025,
+      "step": 14661
+    },
+    {
+      "epoch": 2.564109289617486,
+      "grad_norm": 0.8270579556973482,
+      "learning_rate": 1.086636989327714e-05,
+      "loss": 11.6061,
+      "step": 14662
+    },
+    {
+      "epoch": 2.5642841530054645,
+      "grad_norm": 0.8475123252227219,
+      "learning_rate": 1.0857811964817954e-05,
+      "loss": 11.9587,
+      "step": 14663
+    },
+    {
+      "epoch": 2.5644590163934424,
+      "grad_norm": 0.7967367260019812,
+      "learning_rate": 1.0849257214173636e-05,
+      "loss": 11.7261,
+      "step": 14664
+    },
+    {
+      "epoch": 2.5646338797814208,
+      "grad_norm": 0.8213991016030041,
+      "learning_rate": 1.084070564164913e-05,
+      "loss": 11.9905,
+      "step": 14665
+    },
+    {
+      "epoch": 2.564808743169399,
+      "grad_norm": 0.8338465115218143,
+      "learning_rate": 1.083215724754929e-05,
+      "loss": 11.9485,
+      "step": 14666
+    },
+    {
+      "epoch": 2.564983606557377,
+      "grad_norm": 0.8376815489791919,
+      "learning_rate": 1.0823612032178865e-05,
+      "loss": 11.6336,
+      "step": 14667
+    },
+    {
+      "epoch": 2.5651584699453553,
+      "grad_norm": 0.8130549976655856,
+      "learning_rate": 1.0815069995842475e-05,
+      "loss": 11.8022,
+      "step": 14668
+    },
+    {
+      "epoch": 2.5653333333333332,
+      "grad_norm": 0.8943778365136794,
+      "learning_rate": 1.0806531138844633e-05,
+      "loss": 11.6141,
+      "step": 14669
+    },
+    {
+      "epoch": 2.5655081967213116,
+      "grad_norm": 0.8582921202642814,
+      "learning_rate": 1.0797995461489752e-05,
+      "loss": 11.7085,
+      "step": 14670
+    },
+    {
+      "epoch": 2.5656830601092895,
+      "grad_norm": 0.8103898097780365,
+      "learning_rate": 1.078946296408212e-05,
+      "loss": 12.1628,
+      "step": 14671
+    },
+    {
+      "epoch": 2.565857923497268,
+      "grad_norm": 0.8680616271986469,
+      "learning_rate": 1.0780933646925905e-05,
+      "loss": 11.6894,
+      "step": 14672
+    },
+    {
+      "epoch": 2.5660327868852457,
+      "grad_norm": 0.802181850589803,
+      "learning_rate": 1.0772407510325155e-05,
+      "loss": 12.0469,
+      "step": 14673
+    },
+    {
+      "epoch": 2.566207650273224,
+      "grad_norm": 0.8121520953517612,
+      "learning_rate": 1.0763884554583826e-05,
+      "loss": 11.8992,
+      "step": 14674
+    },
+    {
+      "epoch": 2.5663825136612024,
+      "grad_norm": 0.8495487894372823,
+      "learning_rate": 1.075536478000575e-05,
+      "loss": 11.6745,
+      "step": 14675
+    },
+    {
+      "epoch": 2.5665573770491803,
+      "grad_norm": 0.8879826841942726,
+      "learning_rate": 1.0746848186894665e-05,
+      "loss": 11.9855,
+      "step": 14676
+    },
+    {
+      "epoch": 2.566732240437158,
+      "grad_norm": 0.9672982360319965,
+      "learning_rate": 1.0738334775554149e-05,
+      "loss": 11.9758,
+      "step": 14677
+    },
+    {
+      "epoch": 2.5669071038251365,
+      "grad_norm": 0.8264770736187197,
+      "learning_rate": 1.0729824546287693e-05,
+      "loss": 11.8789,
+      "step": 14678
+    },
+    {
+      "epoch": 2.567081967213115,
+      "grad_norm": 0.8741994546506474,
+      "learning_rate": 1.0721317499398698e-05,
+      "loss": 11.8324,
+      "step": 14679
+    },
+    {
+      "epoch": 2.5672568306010928,
+      "grad_norm": 0.8534063888125434,
+      "learning_rate": 1.0712813635190455e-05,
+      "loss": 11.822,
+      "step": 14680
+    },
+    {
+      "epoch": 2.567431693989071,
+      "grad_norm": 0.8473632046460071,
+      "learning_rate": 1.0704312953966045e-05,
+      "loss": 11.7232,
+      "step": 14681
+    },
+    {
+      "epoch": 2.567606557377049,
+      "grad_norm": 0.845687757435066,
+      "learning_rate": 1.0695815456028557e-05,
+      "loss": 12.0671,
+      "step": 14682
+    },
+    {
+      "epoch": 2.5677814207650274,
+      "grad_norm": 0.8272413074501622,
+      "learning_rate": 1.0687321141680929e-05,
+      "loss": 11.8819,
+      "step": 14683
+    },
+    {
+      "epoch": 2.5679562841530057,
+      "grad_norm": 0.8198470593654157,
+      "learning_rate": 1.0678830011225938e-05,
+      "loss": 11.8475,
+      "step": 14684
+    },
+    {
+      "epoch": 2.5681311475409836,
+      "grad_norm": 0.8031591989850162,
+      "learning_rate": 1.0670342064966287e-05,
+      "loss": 11.785,
+      "step": 14685
+    },
+    {
+      "epoch": 2.5683060109289615,
+      "grad_norm": 0.8042502870013768,
+      "learning_rate": 1.066185730320457e-05,
+      "loss": 12.0257,
+      "step": 14686
+    },
+    {
+      "epoch": 2.56848087431694,
+      "grad_norm": 0.8415440099471577,
+      "learning_rate": 1.0653375726243264e-05,
+      "loss": 12.028,
+      "step": 14687
+    },
+    {
+      "epoch": 2.568655737704918,
+      "grad_norm": 0.8511821389056726,
+      "learning_rate": 1.0644897334384718e-05,
+      "loss": 11.8057,
+      "step": 14688
+    },
+    {
+      "epoch": 2.568830601092896,
+      "grad_norm": 0.8469342124417937,
+      "learning_rate": 1.063642212793119e-05,
+      "loss": 11.7682,
+      "step": 14689
+    },
+    {
+      "epoch": 2.5690054644808744,
+      "grad_norm": 0.8024134969649985,
+      "learning_rate": 1.0627950107184791e-05,
+      "loss": 11.8208,
+      "step": 14690
+    },
+    {
+      "epoch": 2.5691803278688523,
+      "grad_norm": 0.8259461846050326,
+      "learning_rate": 1.0619481272447585e-05,
+      "loss": 11.9,
+      "step": 14691
+    },
+    {
+      "epoch": 2.5693551912568307,
+      "grad_norm": 0.845852464844825,
+      "learning_rate": 1.0611015624021413e-05,
+      "loss": 12.1379,
+      "step": 14692
+    },
+    {
+      "epoch": 2.569530054644809,
+      "grad_norm": 0.8902214344982817,
+      "learning_rate": 1.0602553162208096e-05,
+      "loss": 11.9786,
+      "step": 14693
+    },
+    {
+      "epoch": 2.569704918032787,
+      "grad_norm": 0.8205057650504964,
+      "learning_rate": 1.0594093887309308e-05,
+      "loss": 12.0741,
+      "step": 14694
+    },
+    {
+      "epoch": 2.569879781420765,
+      "grad_norm": 0.85981016262945,
+      "learning_rate": 1.0585637799626636e-05,
+      "loss": 12.0174,
+      "step": 14695
+    },
+    {
+      "epoch": 2.570054644808743,
+      "grad_norm": 0.812811874864117,
+      "learning_rate": 1.0577184899461478e-05,
+      "loss": 11.6983,
+      "step": 14696
+    },
+    {
+      "epoch": 2.5702295081967215,
+      "grad_norm": 0.8604643463551781,
+      "learning_rate": 1.056873518711521e-05,
+      "loss": 11.8804,
+      "step": 14697
+    },
+    {
+      "epoch": 2.5704043715846994,
+      "grad_norm": 0.9171040114786864,
+      "learning_rate": 1.0560288662889062e-05,
+      "loss": 12.0187,
+      "step": 14698
+    },
+    {
+      "epoch": 2.5705792349726777,
+      "grad_norm": 0.8339789802813952,
+      "learning_rate": 1.0551845327084098e-05,
+      "loss": 11.8415,
+      "step": 14699
+    },
+    {
+      "epoch": 2.5707540983606556,
+      "grad_norm": 0.8963452959135042,
+      "learning_rate": 1.0543405180001342e-05,
+      "loss": 11.7796,
+      "step": 14700
+    },
+    {
+      "epoch": 2.570928961748634,
+      "grad_norm": 0.832514541212896,
+      "learning_rate": 1.0534968221941676e-05,
+      "loss": 11.8948,
+      "step": 14701
+    },
+    {
+      "epoch": 2.5711038251366123,
+      "grad_norm": 0.8026178518028516,
+      "learning_rate": 1.052653445320586e-05,
+      "loss": 11.9415,
+      "step": 14702
+    },
+    {
+      "epoch": 2.57127868852459,
+      "grad_norm": 0.8417402666251058,
+      "learning_rate": 1.0518103874094554e-05,
+      "loss": 12.1636,
+      "step": 14703
+    },
+    {
+      "epoch": 2.571453551912568,
+      "grad_norm": 0.8208994095469848,
+      "learning_rate": 1.0509676484908293e-05,
+      "loss": 11.6307,
+      "step": 14704
+    },
+    {
+      "epoch": 2.5716284153005464,
+      "grad_norm": 0.8895644290968812,
+      "learning_rate": 1.0501252285947516e-05,
+      "loss": 11.758,
+      "step": 14705
+    },
+    {
+      "epoch": 2.5718032786885248,
+      "grad_norm": 0.8619028502247336,
+      "learning_rate": 1.0492831277512538e-05,
+      "loss": 11.9526,
+      "step": 14706
+    },
+    {
+      "epoch": 2.5719781420765027,
+      "grad_norm": 0.8222289224587206,
+      "learning_rate": 1.048441345990353e-05,
+      "loss": 11.8129,
+      "step": 14707
+    },
+    {
+      "epoch": 2.572153005464481,
+      "grad_norm": 0.8226592652930667,
+      "learning_rate": 1.0475998833420596e-05,
+      "loss": 11.8456,
+      "step": 14708
+    },
+    {
+      "epoch": 2.572327868852459,
+      "grad_norm": 0.8062524023407815,
+      "learning_rate": 1.0467587398363699e-05,
+      "loss": 11.5558,
+      "step": 14709
+    },
+    {
+      "epoch": 2.5725027322404372,
+      "grad_norm": 0.8542522078789656,
+      "learning_rate": 1.0459179155032727e-05,
+      "loss": 11.765,
+      "step": 14710
+    },
+    {
+      "epoch": 2.572677595628415,
+      "grad_norm": 0.8939189811546208,
+      "learning_rate": 1.045077410372739e-05,
+      "loss": 11.9147,
+      "step": 14711
+    },
+    {
+      "epoch": 2.5728524590163935,
+      "grad_norm": 0.813309511277867,
+      "learning_rate": 1.0442372244747323e-05,
+      "loss": 11.7706,
+      "step": 14712
+    },
+    {
+      "epoch": 2.5730273224043714,
+      "grad_norm": 0.7757906539753696,
+      "learning_rate": 1.0433973578392064e-05,
+      "loss": 11.7062,
+      "step": 14713
+    },
+    {
+      "epoch": 2.5732021857923497,
+      "grad_norm": 0.8714812013248859,
+      "learning_rate": 1.0425578104960988e-05,
+      "loss": 11.9512,
+      "step": 14714
+    },
+    {
+      "epoch": 2.573377049180328,
+      "grad_norm": 0.8535591169392536,
+      "learning_rate": 1.0417185824753395e-05,
+      "loss": 11.8488,
+      "step": 14715
+    },
+    {
+      "epoch": 2.573551912568306,
+      "grad_norm": 0.7809912972957923,
+      "learning_rate": 1.0408796738068449e-05,
+      "loss": 11.6736,
+      "step": 14716
+    },
+    {
+      "epoch": 2.5737267759562843,
+      "grad_norm": 0.8757972564519231,
+      "learning_rate": 1.0400410845205255e-05,
+      "loss": 11.9359,
+      "step": 14717
+    },
+    {
+      "epoch": 2.573901639344262,
+      "grad_norm": 0.8617940885268094,
+      "learning_rate": 1.0392028146462706e-05,
+      "loss": 12.1314,
+      "step": 14718
+    },
+    {
+      "epoch": 2.5740765027322405,
+      "grad_norm": 0.8109275993453647,
+      "learning_rate": 1.0383648642139665e-05,
+      "loss": 11.8921,
+      "step": 14719
+    },
+    {
+      "epoch": 2.5742513661202184,
+      "grad_norm": 0.8537503698050222,
+      "learning_rate": 1.0375272332534836e-05,
+      "loss": 11.945,
+      "step": 14720
+    },
+    {
+      "epoch": 2.5744262295081968,
+      "grad_norm": 0.7754330816397608,
+      "learning_rate": 1.0366899217946835e-05,
+      "loss": 11.5542,
+      "step": 14721
+    },
+    {
+      "epoch": 2.5746010928961747,
+      "grad_norm": 0.83564320525542,
+      "learning_rate": 1.0358529298674146e-05,
+      "loss": 11.8196,
+      "step": 14722
+    },
+    {
+      "epoch": 2.574775956284153,
+      "grad_norm": 0.8503593642189757,
+      "learning_rate": 1.0350162575015166e-05,
+      "loss": 11.7984,
+      "step": 14723
+    },
+    {
+      "epoch": 2.5749508196721314,
+      "grad_norm": 0.7836841360034961,
+      "learning_rate": 1.034179904726813e-05,
+      "loss": 11.8341,
+      "step": 14724
+    },
+    {
+      "epoch": 2.5751256830601092,
+      "grad_norm": 0.7816459100673164,
+      "learning_rate": 1.0333438715731237e-05,
+      "loss": 12.1506,
+      "step": 14725
+    },
+    {
+      "epoch": 2.5753005464480876,
+      "grad_norm": 0.8259923281098951,
+      "learning_rate": 1.0325081580702456e-05,
+      "loss": 11.7479,
+      "step": 14726
+    },
+    {
+      "epoch": 2.5754754098360655,
+      "grad_norm": 0.825533102403895,
+      "learning_rate": 1.0316727642479752e-05,
+      "loss": 11.8768,
+      "step": 14727
+    },
+    {
+      "epoch": 2.575650273224044,
+      "grad_norm": 0.906837827567421,
+      "learning_rate": 1.030837690136095e-05,
+      "loss": 11.7499,
+      "step": 14728
+    },
+    {
+      "epoch": 2.5758251366120217,
+      "grad_norm": 0.8304782386100006,
+      "learning_rate": 1.0300029357643693e-05,
+      "loss": 11.68,
+      "step": 14729
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 0.8346389674760287,
+      "learning_rate": 1.0291685011625574e-05,
+      "loss": 12.0162,
+      "step": 14730
+    },
+    {
+      "epoch": 2.576174863387978,
+      "grad_norm": 0.8687464013924203,
+      "learning_rate": 1.028334386360409e-05,
+      "loss": 11.7197,
+      "step": 14731
+    },
+    {
+      "epoch": 2.5763497267759563,
+      "grad_norm": 0.8089366511784553,
+      "learning_rate": 1.0275005913876579e-05,
+      "loss": 11.7594,
+      "step": 14732
+    },
+    {
+      "epoch": 2.5765245901639346,
+      "grad_norm": 0.8266261865293257,
+      "learning_rate": 1.026667116274026e-05,
+      "loss": 11.8467,
+      "step": 14733
+    },
+    {
+      "epoch": 2.5766994535519125,
+      "grad_norm": 0.7801523417371344,
+      "learning_rate": 1.0258339610492274e-05,
+      "loss": 11.9297,
+      "step": 14734
+    },
+    {
+      "epoch": 2.5768743169398904,
+      "grad_norm": 0.7788918164247653,
+      "learning_rate": 1.0250011257429637e-05,
+      "loss": 11.8748,
+      "step": 14735
+    },
+    {
+      "epoch": 2.577049180327869,
+      "grad_norm": 0.7993715220633423,
+      "learning_rate": 1.0241686103849247e-05,
+      "loss": 11.8225,
+      "step": 14736
+    },
+    {
+      "epoch": 2.577224043715847,
+      "grad_norm": 0.8152758704630523,
+      "learning_rate": 1.0233364150047852e-05,
+      "loss": 12.1491,
+      "step": 14737
+    },
+    {
+      "epoch": 2.577398907103825,
+      "grad_norm": 0.8650318965525322,
+      "learning_rate": 1.0225045396322153e-05,
+      "loss": 11.8003,
+      "step": 14738
+    },
+    {
+      "epoch": 2.5775737704918034,
+      "grad_norm": 0.7674860543905637,
+      "learning_rate": 1.0216729842968698e-05,
+      "loss": 12.0156,
+      "step": 14739
+    },
+    {
+      "epoch": 2.5777486338797813,
+      "grad_norm": 0.8491388960150726,
+      "learning_rate": 1.0208417490283917e-05,
+      "loss": 11.8959,
+      "step": 14740
+    },
+    {
+      "epoch": 2.5779234972677596,
+      "grad_norm": 0.8471227185230937,
+      "learning_rate": 1.020010833856414e-05,
+      "loss": 11.9811,
+      "step": 14741
+    },
+    {
+      "epoch": 2.578098360655738,
+      "grad_norm": 0.77101518603483,
+      "learning_rate": 1.0191802388105587e-05,
+      "loss": 12.008,
+      "step": 14742
+    },
+    {
+      "epoch": 2.578273224043716,
+      "grad_norm": 0.7803410607093112,
+      "learning_rate": 1.0183499639204363e-05,
+      "loss": 12.034,
+      "step": 14743
+    },
+    {
+      "epoch": 2.5784480874316937,
+      "grad_norm": 0.8402459023024766,
+      "learning_rate": 1.0175200092156411e-05,
+      "loss": 11.9054,
+      "step": 14744
+    },
+    {
+      "epoch": 2.578622950819672,
+      "grad_norm": 0.8224206319150873,
+      "learning_rate": 1.0166903747257639e-05,
+      "loss": 11.854,
+      "step": 14745
+    },
+    {
+      "epoch": 2.5787978142076504,
+      "grad_norm": 0.8076825889159379,
+      "learning_rate": 1.0158610604803786e-05,
+      "loss": 12.1156,
+      "step": 14746
+    },
+    {
+      "epoch": 2.5789726775956283,
+      "grad_norm": 0.8223430743085375,
+      "learning_rate": 1.0150320665090517e-05,
+      "loss": 11.7421,
+      "step": 14747
+    },
+    {
+      "epoch": 2.5791475409836067,
+      "grad_norm": 0.8447165111264919,
+      "learning_rate": 1.0142033928413319e-05,
+      "loss": 11.7199,
+      "step": 14748
+    },
+    {
+      "epoch": 2.5793224043715846,
+      "grad_norm": 0.9097250761382519,
+      "learning_rate": 1.0133750395067631e-05,
+      "loss": 11.896,
+      "step": 14749
+    },
+    {
+      "epoch": 2.579497267759563,
+      "grad_norm": 0.7711295281209676,
+      "learning_rate": 1.012547006534873e-05,
+      "loss": 12.1081,
+      "step": 14750
+    },
+    {
+      "epoch": 2.5796721311475412,
+      "grad_norm": 0.7917614123634249,
+      "learning_rate": 1.0117192939551846e-05,
+      "loss": 12.0245,
+      "step": 14751
+    },
+    {
+      "epoch": 2.579846994535519,
+      "grad_norm": 0.8389638418630575,
+      "learning_rate": 1.0108919017971985e-05,
+      "loss": 11.9752,
+      "step": 14752
+    },
+    {
+      "epoch": 2.580021857923497,
+      "grad_norm": 0.8533932924744715,
+      "learning_rate": 1.0100648300904147e-05,
+      "loss": 11.883,
+      "step": 14753
+    },
+    {
+      "epoch": 2.5801967213114754,
+      "grad_norm": 0.8899541489335638,
+      "learning_rate": 1.0092380788643152e-05,
+      "loss": 11.9491,
+      "step": 14754
+    },
+    {
+      "epoch": 2.5803715846994537,
+      "grad_norm": 0.7971383014570477,
+      "learning_rate": 1.0084116481483752e-05,
+      "loss": 11.931,
+      "step": 14755
+    },
+    {
+      "epoch": 2.5805464480874316,
+      "grad_norm": 0.8313054858314983,
+      "learning_rate": 1.0075855379720533e-05,
+      "loss": 12.0535,
+      "step": 14756
+    },
+    {
+      "epoch": 2.58072131147541,
+      "grad_norm": 0.7945896250588595,
+      "learning_rate": 1.0067597483648005e-05,
+      "loss": 12.0326,
+      "step": 14757
+    },
+    {
+      "epoch": 2.580896174863388,
+      "grad_norm": 0.8254091430464964,
+      "learning_rate": 1.0059342793560544e-05,
+      "loss": 11.8923,
+      "step": 14758
+    },
+    {
+      "epoch": 2.581071038251366,
+      "grad_norm": 0.8464661526211619,
+      "learning_rate": 1.0051091309752435e-05,
+      "loss": 11.8866,
+      "step": 14759
+    },
+    {
+      "epoch": 2.5812459016393445,
+      "grad_norm": 0.9398187107800459,
+      "learning_rate": 1.0042843032517824e-05,
+      "loss": 11.7988,
+      "step": 14760
+    },
+    {
+      "epoch": 2.5814207650273224,
+      "grad_norm": 0.9016751775505922,
+      "learning_rate": 1.003459796215076e-05,
+      "loss": 12.0751,
+      "step": 14761
+    },
+    {
+      "epoch": 2.5815956284153003,
+      "grad_norm": 0.902718793357796,
+      "learning_rate": 1.0026356098945177e-05,
+      "loss": 11.7194,
+      "step": 14762
+    },
+    {
+      "epoch": 2.5817704918032787,
+      "grad_norm": 0.9082399589285106,
+      "learning_rate": 1.0018117443194863e-05,
+      "loss": 11.5984,
+      "step": 14763
+    },
+    {
+      "epoch": 2.581945355191257,
+      "grad_norm": 0.8707460763891973,
+      "learning_rate": 1.0009881995193526e-05,
+      "loss": 11.7399,
+      "step": 14764
+    },
+    {
+      "epoch": 2.582120218579235,
+      "grad_norm": 0.8631414267918252,
+      "learning_rate": 1.0001649755234766e-05,
+      "loss": 12.0874,
+      "step": 14765
+    },
+    {
+      "epoch": 2.5822950819672132,
+      "grad_norm": 0.9153207015617882,
+      "learning_rate": 9.993420723612045e-06,
+      "loss": 11.9571,
+      "step": 14766
+    },
+    {
+      "epoch": 2.582469945355191,
+      "grad_norm": 0.8225785021546543,
+      "learning_rate": 9.9851949006187e-06,
+      "loss": 11.9646,
+      "step": 14767
+    },
+    {
+      "epoch": 2.5826448087431695,
+      "grad_norm": 0.8271455919437956,
+      "learning_rate": 9.976972286547981e-06,
+      "loss": 12.0223,
+      "step": 14768
+    },
+    {
+      "epoch": 2.5828196721311474,
+      "grad_norm": 0.7991392547102548,
+      "learning_rate": 9.968752881693023e-06,
+      "loss": 11.6094,
+      "step": 14769
+    },
+    {
+      "epoch": 2.5829945355191257,
+      "grad_norm": 0.7748895245814529,
+      "learning_rate": 9.960536686346856e-06,
+      "loss": 11.904,
+      "step": 14770
+    },
+    {
+      "epoch": 2.5831693989071036,
+      "grad_norm": 0.8233513068148274,
+      "learning_rate": 9.952323700802335e-06,
+      "loss": 11.9418,
+      "step": 14771
+    },
+    {
+      "epoch": 2.583344262295082,
+      "grad_norm": 0.7971908544033074,
+      "learning_rate": 9.94411392535226e-06,
+      "loss": 11.7673,
+      "step": 14772
+    },
+    {
+      "epoch": 2.5835191256830603,
+      "grad_norm": 0.8355097013666452,
+      "learning_rate": 9.935907360289332e-06,
+      "loss": 12.0271,
+      "step": 14773
+    },
+    {
+      "epoch": 2.583693989071038,
+      "grad_norm": 0.8987781514876154,
+      "learning_rate": 9.927704005906058e-06,
+      "loss": 11.541,
+      "step": 14774
+    },
+    {
+      "epoch": 2.5838688524590165,
+      "grad_norm": 0.8301357816883358,
+      "learning_rate": 9.919503862494894e-06,
+      "loss": 11.7128,
+      "step": 14775
+    },
+    {
+      "epoch": 2.5840437158469944,
+      "grad_norm": 0.8858008410618796,
+      "learning_rate": 9.911306930348163e-06,
+      "loss": 11.7575,
+      "step": 14776
+    },
+    {
+      "epoch": 2.584218579234973,
+      "grad_norm": 0.7498892605134886,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 11.9512,
+      "step": 14777
+    },
+    {
+      "epoch": 2.5843934426229507,
+      "grad_norm": 0.818305636691371,
+      "learning_rate": 9.894922701016762e-06,
+      "loss": 11.8071,
+      "step": 14778
+    },
+    {
+      "epoch": 2.584568306010929,
+      "grad_norm": 0.8309593268082437,
+      "learning_rate": 9.886735404416169e-06,
+      "loss": 11.4892,
+      "step": 14779
+    },
+    {
+      "epoch": 2.584743169398907,
+      "grad_norm": 0.8518409729221027,
+      "learning_rate": 9.878551320248175e-06,
+      "loss": 11.8655,
+      "step": 14780
+    },
+    {
+      "epoch": 2.5849180327868853,
+      "grad_norm": 0.8544907164481026,
+      "learning_rate": 9.870370448804545e-06,
+      "loss": 11.94,
+      "step": 14781
+    },
+    {
+      "epoch": 2.5850928961748636,
+      "grad_norm": 0.8522709751579479,
+      "learning_rate": 9.862192790376879e-06,
+      "loss": 11.6972,
+      "step": 14782
+    },
+    {
+      "epoch": 2.5852677595628415,
+      "grad_norm": 0.8612138109800376,
+      "learning_rate": 9.854018345256733e-06,
+      "loss": 11.8921,
+      "step": 14783
+    },
+    {
+      "epoch": 2.58544262295082,
+      "grad_norm": 0.8535029044375751,
+      "learning_rate": 9.845847113735496e-06,
+      "loss": 11.9669,
+      "step": 14784
+    },
+    {
+      "epoch": 2.5856174863387977,
+      "grad_norm": 0.8056993922066717,
+      "learning_rate": 9.837679096104512e-06,
+      "loss": 11.9963,
+      "step": 14785
+    },
+    {
+      "epoch": 2.585792349726776,
+      "grad_norm": 0.8325035591805139,
+      "learning_rate": 9.829514292654896e-06,
+      "loss": 11.8632,
+      "step": 14786
+    },
+    {
+      "epoch": 2.585967213114754,
+      "grad_norm": 0.8601272136202539,
+      "learning_rate": 9.821352703677745e-06,
+      "loss": 12.0273,
+      "step": 14787
+    },
+    {
+      "epoch": 2.5861420765027323,
+      "grad_norm": 0.7955615187668019,
+      "learning_rate": 9.813194329464003e-06,
+      "loss": 11.9389,
+      "step": 14788
+    },
+    {
+      "epoch": 2.58631693989071,
+      "grad_norm": 0.7669577959193455,
+      "learning_rate": 9.805039170304541e-06,
+      "loss": 11.941,
+      "step": 14789
+    },
+    {
+      "epoch": 2.5864918032786886,
+      "grad_norm": 0.8281735688302899,
+      "learning_rate": 9.796887226490026e-06,
+      "loss": 11.8751,
+      "step": 14790
+    },
+    {
+      "epoch": 2.586666666666667,
+      "grad_norm": 0.8007561066411424,
+      "learning_rate": 9.78873849831109e-06,
+      "loss": 11.9632,
+      "step": 14791
+    },
+    {
+      "epoch": 2.586841530054645,
+      "grad_norm": 0.8397720616829925,
+      "learning_rate": 9.780592986058234e-06,
+      "loss": 11.7585,
+      "step": 14792
+    },
+    {
+      "epoch": 2.5870163934426227,
+      "grad_norm": 0.8396880881892382,
+      "learning_rate": 9.772450690021828e-06,
+      "loss": 11.8248,
+      "step": 14793
+    },
+    {
+      "epoch": 2.587191256830601,
+      "grad_norm": 0.8365480229057369,
+      "learning_rate": 9.764311610492128e-06,
+      "loss": 11.9775,
+      "step": 14794
+    },
+    {
+      "epoch": 2.5873661202185794,
+      "grad_norm": 0.8375689144567582,
+      "learning_rate": 9.756175747759288e-06,
+      "loss": 11.9711,
+      "step": 14795
+    },
+    {
+      "epoch": 2.5875409836065573,
+      "grad_norm": 0.9101854016603185,
+      "learning_rate": 9.748043102113347e-06,
+      "loss": 11.8774,
+      "step": 14796
+    },
+    {
+      "epoch": 2.5877158469945356,
+      "grad_norm": 0.8199997259276778,
+      "learning_rate": 9.739913673844214e-06,
+      "loss": 11.3718,
+      "step": 14797
+    },
+    {
+      "epoch": 2.5878907103825135,
+      "grad_norm": 0.8704250822131836,
+      "learning_rate": 9.731787463241704e-06,
+      "loss": 12.0013,
+      "step": 14798
+    },
+    {
+      "epoch": 2.588065573770492,
+      "grad_norm": 0.8010496826574243,
+      "learning_rate": 9.723664470595506e-06,
+      "loss": 11.9325,
+      "step": 14799
+    },
+    {
+      "epoch": 2.58824043715847,
+      "grad_norm": 0.804218190486647,
+      "learning_rate": 9.715544696195211e-06,
+      "loss": 11.6992,
+      "step": 14800
+    },
+    {
+      "epoch": 2.588415300546448,
+      "grad_norm": 0.839620519631635,
+      "learning_rate": 9.707428140330243e-06,
+      "loss": 11.9024,
+      "step": 14801
+    },
+    {
+      "epoch": 2.588590163934426,
+      "grad_norm": 0.8376856895767071,
+      "learning_rate": 9.699314803289972e-06,
+      "loss": 11.6846,
+      "step": 14802
+    },
+    {
+      "epoch": 2.5887650273224043,
+      "grad_norm": 0.8410749546304472,
+      "learning_rate": 9.69120468536362e-06,
+      "loss": 11.9136,
+      "step": 14803
+    },
+    {
+      "epoch": 2.5889398907103827,
+      "grad_norm": 0.7957852360040717,
+      "learning_rate": 9.683097786840323e-06,
+      "loss": 11.8577,
+      "step": 14804
+    },
+    {
+      "epoch": 2.5891147540983606,
+      "grad_norm": 0.8553332262137678,
+      "learning_rate": 9.67499410800905e-06,
+      "loss": 11.8561,
+      "step": 14805
+    },
+    {
+      "epoch": 2.589289617486339,
+      "grad_norm": 0.9341874818310049,
+      "learning_rate": 9.666893649158693e-06,
+      "loss": 11.7758,
+      "step": 14806
+    },
+    {
+      "epoch": 2.589464480874317,
+      "grad_norm": 0.8323460672949311,
+      "learning_rate": 9.658796410578064e-06,
+      "loss": 11.7439,
+      "step": 14807
+    },
+    {
+      "epoch": 2.589639344262295,
+      "grad_norm": 0.7782563218492281,
+      "learning_rate": 9.650702392555765e-06,
+      "loss": 12.1032,
+      "step": 14808
+    },
+    {
+      "epoch": 2.5898142076502735,
+      "grad_norm": 0.7910582691462601,
+      "learning_rate": 9.642611595380368e-06,
+      "loss": 11.8105,
+      "step": 14809
+    },
+    {
+      "epoch": 2.5899890710382514,
+      "grad_norm": 0.8615752877201849,
+      "learning_rate": 9.634524019340308e-06,
+      "loss": 11.9989,
+      "step": 14810
+    },
+    {
+      "epoch": 2.5901639344262293,
+      "grad_norm": 0.8639369917461631,
+      "learning_rate": 9.626439664723874e-06,
+      "loss": 11.6805,
+      "step": 14811
+    },
+    {
+      "epoch": 2.5903387978142076,
+      "grad_norm": 0.8061799989844256,
+      "learning_rate": 9.618358531819271e-06,
+      "loss": 11.882,
+      "step": 14812
+    },
+    {
+      "epoch": 2.590513661202186,
+      "grad_norm": 0.7778325344040535,
+      "learning_rate": 9.610280620914592e-06,
+      "loss": 11.6746,
+      "step": 14813
+    },
+    {
+      "epoch": 2.590688524590164,
+      "grad_norm": 0.834763045204745,
+      "learning_rate": 9.602205932297802e-06,
+      "loss": 11.837,
+      "step": 14814
+    },
+    {
+      "epoch": 2.590863387978142,
+      "grad_norm": 0.794848319559798,
+      "learning_rate": 9.594134466256765e-06,
+      "loss": 11.88,
+      "step": 14815
+    },
+    {
+      "epoch": 2.59103825136612,
+      "grad_norm": 0.7903932492421196,
+      "learning_rate": 9.586066223079193e-06,
+      "loss": 11.9412,
+      "step": 14816
+    },
+    {
+      "epoch": 2.5912131147540984,
+      "grad_norm": 0.8774933081576408,
+      "learning_rate": 9.578001203052722e-06,
+      "loss": 11.815,
+      "step": 14817
+    },
+    {
+      "epoch": 2.5913879781420768,
+      "grad_norm": 0.8170565602691432,
+      "learning_rate": 9.569939406464856e-06,
+      "loss": 11.9895,
+      "step": 14818
+    },
+    {
+      "epoch": 2.5915628415300547,
+      "grad_norm": 0.8308685154012875,
+      "learning_rate": 9.56188083360302e-06,
+      "loss": 11.8923,
+      "step": 14819
+    },
+    {
+      "epoch": 2.5917377049180326,
+      "grad_norm": 0.8853167003373348,
+      "learning_rate": 9.553825484754442e-06,
+      "loss": 11.7755,
+      "step": 14820
+    },
+    {
+      "epoch": 2.591912568306011,
+      "grad_norm": 0.8799988330578846,
+      "learning_rate": 9.545773360206323e-06,
+      "loss": 11.7411,
+      "step": 14821
+    },
+    {
+      "epoch": 2.5920874316939893,
+      "grad_norm": 0.9147701209161291,
+      "learning_rate": 9.5377244602457e-06,
+      "loss": 11.9222,
+      "step": 14822
+    },
+    {
+      "epoch": 2.592262295081967,
+      "grad_norm": 0.8879527122389755,
+      "learning_rate": 9.529678785159502e-06,
+      "loss": 11.7745,
+      "step": 14823
+    },
+    {
+      "epoch": 2.5924371584699455,
+      "grad_norm": 0.8840439702943493,
+      "learning_rate": 9.52163633523454e-06,
+      "loss": 11.8082,
+      "step": 14824
+    },
+    {
+      "epoch": 2.5926120218579234,
+      "grad_norm": 0.7890461765261055,
+      "learning_rate": 9.513597110757544e-06,
+      "loss": 11.8868,
+      "step": 14825
+    },
+    {
+      "epoch": 2.5927868852459017,
+      "grad_norm": 0.8553166453914561,
+      "learning_rate": 9.505561112015094e-06,
+      "loss": 11.7594,
+      "step": 14826
+    },
+    {
+      "epoch": 2.5929617486338796,
+      "grad_norm": 0.8387099264733473,
+      "learning_rate": 9.497528339293648e-06,
+      "loss": 11.7876,
+      "step": 14827
+    },
+    {
+      "epoch": 2.593136612021858,
+      "grad_norm": 0.8248198925933339,
+      "learning_rate": 9.48949879287957e-06,
+      "loss": 11.3462,
+      "step": 14828
+    },
+    {
+      "epoch": 2.593311475409836,
+      "grad_norm": 0.8396693273803769,
+      "learning_rate": 9.481472473059116e-06,
+      "loss": 11.8394,
+      "step": 14829
+    },
+    {
+      "epoch": 2.593486338797814,
+      "grad_norm": 0.8383714997980889,
+      "learning_rate": 9.473449380118404e-06,
+      "loss": 12.1266,
+      "step": 14830
+    },
+    {
+      "epoch": 2.5936612021857925,
+      "grad_norm": 0.8578595522525921,
+      "learning_rate": 9.46542951434346e-06,
+      "loss": 12.0575,
+      "step": 14831
+    },
+    {
+      "epoch": 2.5938360655737704,
+      "grad_norm": 0.8750396093674078,
+      "learning_rate": 9.457412876020166e-06,
+      "loss": 11.9408,
+      "step": 14832
+    },
+    {
+      "epoch": 2.594010928961749,
+      "grad_norm": 0.8584615410603993,
+      "learning_rate": 9.449399465434327e-06,
+      "loss": 12.0328,
+      "step": 14833
+    },
+    {
+      "epoch": 2.5941857923497267,
+      "grad_norm": 0.7713801083719447,
+      "learning_rate": 9.441389282871615e-06,
+      "loss": 11.8525,
+      "step": 14834
+    },
+    {
+      "epoch": 2.594360655737705,
+      "grad_norm": 0.7632412858277726,
+      "learning_rate": 9.433382328617546e-06,
+      "loss": 11.9323,
+      "step": 14835
+    },
+    {
+      "epoch": 2.594535519125683,
+      "grad_norm": 0.7827943861611822,
+      "learning_rate": 9.425378602957579e-06,
+      "loss": 12.1541,
+      "step": 14836
+    },
+    {
+      "epoch": 2.5947103825136613,
+      "grad_norm": 0.8105335271055497,
+      "learning_rate": 9.417378106177066e-06,
+      "loss": 11.5788,
+      "step": 14837
+    },
+    {
+      "epoch": 2.594885245901639,
+      "grad_norm": 0.8200719817853713,
+      "learning_rate": 9.409380838561166e-06,
+      "loss": 11.8102,
+      "step": 14838
+    },
+    {
+      "epoch": 2.5950601092896175,
+      "grad_norm": 0.8329285932504068,
+      "learning_rate": 9.401386800394996e-06,
+      "loss": 11.5413,
+      "step": 14839
+    },
+    {
+      "epoch": 2.595234972677596,
+      "grad_norm": 0.8178802613107816,
+      "learning_rate": 9.393395991963527e-06,
+      "loss": 11.8355,
+      "step": 14840
+    },
+    {
+      "epoch": 2.5954098360655737,
+      "grad_norm": 0.8707458759278588,
+      "learning_rate": 9.385408413551655e-06,
+      "loss": 11.8716,
+      "step": 14841
+    },
+    {
+      "epoch": 2.5955846994535516,
+      "grad_norm": 0.8365588610724182,
+      "learning_rate": 9.377424065444075e-06,
+      "loss": 11.79,
+      "step": 14842
+    },
+    {
+      "epoch": 2.59575956284153,
+      "grad_norm": 0.7844268947172478,
+      "learning_rate": 9.369442947925456e-06,
+      "loss": 11.8461,
+      "step": 14843
+    },
+    {
+      "epoch": 2.5959344262295083,
+      "grad_norm": 0.8813084024501043,
+      "learning_rate": 9.361465061280295e-06,
+      "loss": 11.8393,
+      "step": 14844
+    },
+    {
+      "epoch": 2.596109289617486,
+      "grad_norm": 0.767540780711802,
+      "learning_rate": 9.353490405793031e-06,
+      "loss": 12.0888,
+      "step": 14845
+    },
+    {
+      "epoch": 2.5962841530054646,
+      "grad_norm": 0.7724708337271475,
+      "learning_rate": 9.345518981747903e-06,
+      "loss": 11.4746,
+      "step": 14846
+    },
+    {
+      "epoch": 2.5964590163934425,
+      "grad_norm": 0.816090242368683,
+      "learning_rate": 9.337550789429116e-06,
+      "loss": 11.8458,
+      "step": 14847
+    },
+    {
+      "epoch": 2.596633879781421,
+      "grad_norm": 0.8168459195333749,
+      "learning_rate": 9.329585829120701e-06,
+      "loss": 11.6248,
+      "step": 14848
+    },
+    {
+      "epoch": 2.596808743169399,
+      "grad_norm": 0.8235299235654943,
+      "learning_rate": 9.32162410110662e-06,
+      "loss": 12.0928,
+      "step": 14849
+    },
+    {
+      "epoch": 2.596983606557377,
+      "grad_norm": 0.8265234767245229,
+      "learning_rate": 9.313665605670697e-06,
+      "loss": 11.8555,
+      "step": 14850
+    },
+    {
+      "epoch": 2.597158469945355,
+      "grad_norm": 0.7769161207026007,
+      "learning_rate": 9.30571034309663e-06,
+      "loss": 11.5293,
+      "step": 14851
+    },
+    {
+      "epoch": 2.5973333333333333,
+      "grad_norm": 0.8239074488008755,
+      "learning_rate": 9.297758313668048e-06,
+      "loss": 11.9231,
+      "step": 14852
+    },
+    {
+      "epoch": 2.5975081967213116,
+      "grad_norm": 0.8494854129453281,
+      "learning_rate": 9.28980951766839e-06,
+      "loss": 11.9367,
+      "step": 14853
+    },
+    {
+      "epoch": 2.5976830601092895,
+      "grad_norm": 0.8014702827538401,
+      "learning_rate": 9.281863955381044e-06,
+      "loss": 11.7551,
+      "step": 14854
+    },
+    {
+      "epoch": 2.597857923497268,
+      "grad_norm": 0.809261137465035,
+      "learning_rate": 9.273921627089243e-06,
+      "loss": 11.7479,
+      "step": 14855
+    },
+    {
+      "epoch": 2.5980327868852457,
+      "grad_norm": 0.8981219654926355,
+      "learning_rate": 9.265982533076167e-06,
+      "loss": 11.9746,
+      "step": 14856
+    },
+    {
+      "epoch": 2.598207650273224,
+      "grad_norm": 0.8392131860505518,
+      "learning_rate": 9.258046673624777e-06,
+      "loss": 11.9031,
+      "step": 14857
+    },
+    {
+      "epoch": 2.5983825136612024,
+      "grad_norm": 0.7576685632628083,
+      "learning_rate": 9.250114049018e-06,
+      "loss": 11.6791,
+      "step": 14858
+    },
+    {
+      "epoch": 2.5985573770491803,
+      "grad_norm": 0.9008611894977023,
+      "learning_rate": 9.242184659538632e-06,
+      "loss": 11.7182,
+      "step": 14859
+    },
+    {
+      "epoch": 2.5987322404371582,
+      "grad_norm": 0.9337541142795127,
+      "learning_rate": 9.234258505469352e-06,
+      "loss": 11.9518,
+      "step": 14860
+    },
+    {
+      "epoch": 2.5989071038251366,
+      "grad_norm": 0.8458816360004267,
+      "learning_rate": 9.226335587092694e-06,
+      "loss": 11.9887,
+      "step": 14861
+    },
+    {
+      "epoch": 2.599081967213115,
+      "grad_norm": 0.8363738379031741,
+      "learning_rate": 9.218415904691124e-06,
+      "loss": 12.0162,
+      "step": 14862
+    },
+    {
+      "epoch": 2.599256830601093,
+      "grad_norm": 0.8547743348235335,
+      "learning_rate": 9.210499458546951e-06,
+      "loss": 11.9923,
+      "step": 14863
+    },
+    {
+      "epoch": 2.599431693989071,
+      "grad_norm": 0.936959722306092,
+      "learning_rate": 9.202586248942413e-06,
+      "loss": 11.8958,
+      "step": 14864
+    },
+    {
+      "epoch": 2.599606557377049,
+      "grad_norm": 0.8872207460466875,
+      "learning_rate": 9.194676276159574e-06,
+      "loss": 12.0398,
+      "step": 14865
+    },
+    {
+      "epoch": 2.5997814207650274,
+      "grad_norm": 0.851770541175191,
+      "learning_rate": 9.186769540480433e-06,
+      "loss": 11.797,
+      "step": 14866
+    },
+    {
+      "epoch": 2.5999562841530057,
+      "grad_norm": 0.812056049581802,
+      "learning_rate": 9.178866042186852e-06,
+      "loss": 11.9787,
+      "step": 14867
+    },
+    {
+      "epoch": 2.6001311475409836,
+      "grad_norm": 0.834209139034322,
+      "learning_rate": 9.17096578156058e-06,
+      "loss": 11.9825,
+      "step": 14868
+    },
+    {
+      "epoch": 2.6003060109289615,
+      "grad_norm": 0.9392491478522907,
+      "learning_rate": 9.163068758883253e-06,
+      "loss": 11.6095,
+      "step": 14869
+    },
+    {
+      "epoch": 2.60048087431694,
+      "grad_norm": 0.8098750470286762,
+      "learning_rate": 9.155174974436408e-06,
+      "loss": 11.9899,
+      "step": 14870
+    },
+    {
+      "epoch": 2.600655737704918,
+      "grad_norm": 0.811397342570658,
+      "learning_rate": 9.147284428501446e-06,
+      "loss": 11.6183,
+      "step": 14871
+    },
+    {
+      "epoch": 2.600830601092896,
+      "grad_norm": 0.8346665410489699,
+      "learning_rate": 9.139397121359616e-06,
+      "loss": 11.3737,
+      "step": 14872
+    },
+    {
+      "epoch": 2.6010054644808744,
+      "grad_norm": 0.8309265963728097,
+      "learning_rate": 9.13151305329213e-06,
+      "loss": 11.8762,
+      "step": 14873
+    },
+    {
+      "epoch": 2.6011803278688523,
+      "grad_norm": 0.9686643996927793,
+      "learning_rate": 9.123632224580037e-06,
+      "loss": 11.7485,
+      "step": 14874
+    },
+    {
+      "epoch": 2.6013551912568307,
+      "grad_norm": 0.8553765688367885,
+      "learning_rate": 9.115754635504281e-06,
+      "loss": 11.9096,
+      "step": 14875
+    },
+    {
+      "epoch": 2.601530054644809,
+      "grad_norm": 0.8732501880336204,
+      "learning_rate": 9.107880286345683e-06,
+      "loss": 11.718,
+      "step": 14876
+    },
+    {
+      "epoch": 2.601704918032787,
+      "grad_norm": 0.8793131635773398,
+      "learning_rate": 9.10000917738495e-06,
+      "loss": 11.9698,
+      "step": 14877
+    },
+    {
+      "epoch": 2.601879781420765,
+      "grad_norm": 0.8383233390649603,
+      "learning_rate": 9.092141308902692e-06,
+      "loss": 12.0016,
+      "step": 14878
+    },
+    {
+      "epoch": 2.602054644808743,
+      "grad_norm": 0.8115416388344148,
+      "learning_rate": 9.084276681179393e-06,
+      "loss": 12.0408,
+      "step": 14879
+    },
+    {
+      "epoch": 2.6022295081967215,
+      "grad_norm": 0.8795829966921108,
+      "learning_rate": 9.076415294495388e-06,
+      "loss": 12.0793,
+      "step": 14880
+    },
+    {
+      "epoch": 2.6024043715846994,
+      "grad_norm": 0.7820045838778276,
+      "learning_rate": 9.06855714913094e-06,
+      "loss": 11.8924,
+      "step": 14881
+    },
+    {
+      "epoch": 2.6025792349726777,
+      "grad_norm": 0.8318429191558538,
+      "learning_rate": 9.060702245366215e-06,
+      "loss": 12.018,
+      "step": 14882
+    },
+    {
+      "epoch": 2.6027540983606556,
+      "grad_norm": 0.792875345261788,
+      "learning_rate": 9.052850583481188e-06,
+      "loss": 11.7442,
+      "step": 14883
+    },
+    {
+      "epoch": 2.602928961748634,
+      "grad_norm": 0.888639675205333,
+      "learning_rate": 9.045002163755766e-06,
+      "loss": 11.965,
+      "step": 14884
+    },
+    {
+      "epoch": 2.603103825136612,
+      "grad_norm": 0.9103881899318412,
+      "learning_rate": 9.037156986469753e-06,
+      "loss": 11.9398,
+      "step": 14885
+    },
+    {
+      "epoch": 2.60327868852459,
+      "grad_norm": 0.8125774972599032,
+      "learning_rate": 9.02931505190282e-06,
+      "loss": 11.8441,
+      "step": 14886
+    },
+    {
+      "epoch": 2.603453551912568,
+      "grad_norm": 0.8533800743400496,
+      "learning_rate": 9.021476360334513e-06,
+      "loss": 11.8568,
+      "step": 14887
+    },
+    {
+      "epoch": 2.6036284153005465,
+      "grad_norm": 0.8056723328922859,
+      "learning_rate": 9.013640912044275e-06,
+      "loss": 11.8707,
+      "step": 14888
+    },
+    {
+      "epoch": 2.603803278688525,
+      "grad_norm": 0.8089806063141151,
+      "learning_rate": 9.005808707311437e-06,
+      "loss": 11.8167,
+      "step": 14889
+    },
+    {
+      "epoch": 2.6039781420765027,
+      "grad_norm": 0.8684210419179593,
+      "learning_rate": 8.997979746415208e-06,
+      "loss": 11.8675,
+      "step": 14890
+    },
+    {
+      "epoch": 2.604153005464481,
+      "grad_norm": 0.8716566588224084,
+      "learning_rate": 8.99015402963468e-06,
+      "loss": 11.9063,
+      "step": 14891
+    },
+    {
+      "epoch": 2.604327868852459,
+      "grad_norm": 0.8370371190747371,
+      "learning_rate": 8.982331557248814e-06,
+      "loss": 11.4576,
+      "step": 14892
+    },
+    {
+      "epoch": 2.6045027322404373,
+      "grad_norm": 0.8626129522460274,
+      "learning_rate": 8.974512329536489e-06,
+      "loss": 11.7952,
+      "step": 14893
+    },
+    {
+      "epoch": 2.604677595628415,
+      "grad_norm": 0.8263976330452882,
+      "learning_rate": 8.96669634677646e-06,
+      "loss": 11.9938,
+      "step": 14894
+    },
+    {
+      "epoch": 2.6048524590163935,
+      "grad_norm": 0.8403889321448865,
+      "learning_rate": 8.958883609247326e-06,
+      "loss": 12.0043,
+      "step": 14895
+    },
+    {
+      "epoch": 2.6050273224043714,
+      "grad_norm": 0.9031268590183238,
+      "learning_rate": 8.95107411722763e-06,
+      "loss": 12.0316,
+      "step": 14896
+    },
+    {
+      "epoch": 2.6052021857923497,
+      "grad_norm": 0.8040640830713969,
+      "learning_rate": 8.943267870995753e-06,
+      "loss": 11.9147,
+      "step": 14897
+    },
+    {
+      "epoch": 2.605377049180328,
+      "grad_norm": 0.7820085290865719,
+      "learning_rate": 8.935464870830012e-06,
+      "loss": 12.0642,
+      "step": 14898
+    },
+    {
+      "epoch": 2.605551912568306,
+      "grad_norm": 0.8157293917682039,
+      "learning_rate": 8.927665117008544e-06,
+      "loss": 11.8714,
+      "step": 14899
+    },
+    {
+      "epoch": 2.605726775956284,
+      "grad_norm": 0.8238117121919942,
+      "learning_rate": 8.91986860980939e-06,
+      "loss": 11.5975,
+      "step": 14900
+    },
+    {
+      "epoch": 2.605901639344262,
+      "grad_norm": 0.8833456737135869,
+      "learning_rate": 8.912075349510508e-06,
+      "loss": 11.634,
+      "step": 14901
+    },
+    {
+      "epoch": 2.6060765027322406,
+      "grad_norm": 0.852644193629201,
+      "learning_rate": 8.904285336389717e-06,
+      "loss": 11.949,
+      "step": 14902
+    },
+    {
+      "epoch": 2.6062513661202185,
+      "grad_norm": 0.8417886429608732,
+      "learning_rate": 8.896498570724721e-06,
+      "loss": 11.7956,
+      "step": 14903
+    },
+    {
+      "epoch": 2.606426229508197,
+      "grad_norm": 0.8815229753132163,
+      "learning_rate": 8.888715052793117e-06,
+      "loss": 11.8449,
+      "step": 14904
+    },
+    {
+      "epoch": 2.6066010928961747,
+      "grad_norm": 0.8511686271968386,
+      "learning_rate": 8.880934782872375e-06,
+      "loss": 11.9811,
+      "step": 14905
+    },
+    {
+      "epoch": 2.606775956284153,
+      "grad_norm": 0.9230759034415759,
+      "learning_rate": 8.873157761239825e-06,
+      "loss": 11.8555,
+      "step": 14906
+    },
+    {
+      "epoch": 2.6069508196721314,
+      "grad_norm": 0.8354102129962465,
+      "learning_rate": 8.865383988172738e-06,
+      "loss": 11.5592,
+      "step": 14907
+    },
+    {
+      "epoch": 2.6071256830601093,
+      "grad_norm": 0.8379156679502621,
+      "learning_rate": 8.857613463948233e-06,
+      "loss": 12.07,
+      "step": 14908
+    },
+    {
+      "epoch": 2.607300546448087,
+      "grad_norm": 0.8694472970428662,
+      "learning_rate": 8.849846188843336e-06,
+      "loss": 11.6176,
+      "step": 14909
+    },
+    {
+      "epoch": 2.6074754098360655,
+      "grad_norm": 0.7895317246933885,
+      "learning_rate": 8.842082163134912e-06,
+      "loss": 11.5121,
+      "step": 14910
+    },
+    {
+      "epoch": 2.607650273224044,
+      "grad_norm": 0.8464204131224263,
+      "learning_rate": 8.834321387099742e-06,
+      "loss": 11.7816,
+      "step": 14911
+    },
+    {
+      "epoch": 2.6078251366120218,
+      "grad_norm": 0.8571474887672452,
+      "learning_rate": 8.826563861014513e-06,
+      "loss": 11.7567,
+      "step": 14912
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 0.8525611057125597,
+      "learning_rate": 8.818809585155763e-06,
+      "loss": 11.9836,
+      "step": 14913
+    },
+    {
+      "epoch": 2.608174863387978,
+      "grad_norm": 0.7817166031571948,
+      "learning_rate": 8.8110585597999e-06,
+      "loss": 11.9173,
+      "step": 14914
+    },
+    {
+      "epoch": 2.6083497267759563,
+      "grad_norm": 0.7916021397127616,
+      "learning_rate": 8.803310785223262e-06,
+      "loss": 11.9686,
+      "step": 14915
+    },
+    {
+      "epoch": 2.6085245901639347,
+      "grad_norm": 0.840591844518209,
+      "learning_rate": 8.79556626170206e-06,
+      "loss": 11.8311,
+      "step": 14916
+    },
+    {
+      "epoch": 2.6086994535519126,
+      "grad_norm": 0.8673047958400307,
+      "learning_rate": 8.787824989512339e-06,
+      "loss": 11.623,
+      "step": 14917
+    },
+    {
+      "epoch": 2.6088743169398905,
+      "grad_norm": 0.8760951195315916,
+      "learning_rate": 8.78008696893009e-06,
+      "loss": 11.7444,
+      "step": 14918
+    },
+    {
+      "epoch": 2.609049180327869,
+      "grad_norm": 0.8397621221698901,
+      "learning_rate": 8.77235220023116e-06,
+      "loss": 11.6967,
+      "step": 14919
+    },
+    {
+      "epoch": 2.609224043715847,
+      "grad_norm": 0.77868916039689,
+      "learning_rate": 8.764620683691282e-06,
+      "loss": 11.9961,
+      "step": 14920
+    },
+    {
+      "epoch": 2.609398907103825,
+      "grad_norm": 0.8598211506075688,
+      "learning_rate": 8.756892419586094e-06,
+      "loss": 12.0406,
+      "step": 14921
+    },
+    {
+      "epoch": 2.6095737704918034,
+      "grad_norm": 1.0510919986863656,
+      "learning_rate": 8.749167408191073e-06,
+      "loss": 11.8958,
+      "step": 14922
+    },
+    {
+      "epoch": 2.6097486338797813,
+      "grad_norm": 0.8957038163210832,
+      "learning_rate": 8.74144564978162e-06,
+      "loss": 11.8274,
+      "step": 14923
+    },
+    {
+      "epoch": 2.6099234972677596,
+      "grad_norm": 0.7870096808950701,
+      "learning_rate": 8.73372714463302e-06,
+      "loss": 11.6377,
+      "step": 14924
+    },
+    {
+      "epoch": 2.610098360655738,
+      "grad_norm": 0.8351005070666533,
+      "learning_rate": 8.726011893020403e-06,
+      "loss": 12.0701,
+      "step": 14925
+    },
+    {
+      "epoch": 2.610273224043716,
+      "grad_norm": 0.8160193928286678,
+      "learning_rate": 8.718299895218818e-06,
+      "loss": 11.8221,
+      "step": 14926
+    },
+    {
+      "epoch": 2.6104480874316938,
+      "grad_norm": 0.8484149063608092,
+      "learning_rate": 8.710591151503189e-06,
+      "loss": 11.9389,
+      "step": 14927
+    },
+    {
+      "epoch": 2.610622950819672,
+      "grad_norm": 0.8310829616072972,
+      "learning_rate": 8.702885662148342e-06,
+      "loss": 11.6933,
+      "step": 14928
+    },
+    {
+      "epoch": 2.6107978142076504,
+      "grad_norm": 0.8171236739849919,
+      "learning_rate": 8.695183427428932e-06,
+      "loss": 11.8012,
+      "step": 14929
+    },
+    {
+      "epoch": 2.6109726775956283,
+      "grad_norm": 0.8685190884799413,
+      "learning_rate": 8.687484447619564e-06,
+      "loss": 11.896,
+      "step": 14930
+    },
+    {
+      "epoch": 2.6111475409836067,
+      "grad_norm": 0.7678272614688999,
+      "learning_rate": 8.679788722994698e-06,
+      "loss": 11.5153,
+      "step": 14931
+    },
+    {
+      "epoch": 2.6113224043715846,
+      "grad_norm": 0.8148346039211414,
+      "learning_rate": 8.672096253828654e-06,
+      "loss": 11.7314,
+      "step": 14932
+    },
+    {
+      "epoch": 2.611497267759563,
+      "grad_norm": 0.7852983070426872,
+      "learning_rate": 8.664407040395673e-06,
+      "loss": 11.6744,
+      "step": 14933
+    },
+    {
+      "epoch": 2.611672131147541,
+      "grad_norm": 0.8280646270491498,
+      "learning_rate": 8.656721082969866e-06,
+      "loss": 12.0053,
+      "step": 14934
+    },
+    {
+      "epoch": 2.611846994535519,
+      "grad_norm": 0.8439443099658098,
+      "learning_rate": 8.649038381825248e-06,
+      "loss": 11.8316,
+      "step": 14935
+    },
+    {
+      "epoch": 2.612021857923497,
+      "grad_norm": 0.9088623940276663,
+      "learning_rate": 8.641358937235656e-06,
+      "loss": 12.0064,
+      "step": 14936
+    },
+    {
+      "epoch": 2.6121967213114754,
+      "grad_norm": 0.8706972289071513,
+      "learning_rate": 8.633682749474881e-06,
+      "loss": 12.108,
+      "step": 14937
+    },
+    {
+      "epoch": 2.6123715846994537,
+      "grad_norm": 0.7942071305638314,
+      "learning_rate": 8.62600981881656e-06,
+      "loss": 11.7716,
+      "step": 14938
+    },
+    {
+      "epoch": 2.6125464480874316,
+      "grad_norm": 0.7742016949084759,
+      "learning_rate": 8.618340145534243e-06,
+      "loss": 12.0124,
+      "step": 14939
+    },
+    {
+      "epoch": 2.61272131147541,
+      "grad_norm": 0.9553016576755602,
+      "learning_rate": 8.61067372990132e-06,
+      "loss": 11.5857,
+      "step": 14940
+    },
+    {
+      "epoch": 2.612896174863388,
+      "grad_norm": 0.8783420704609399,
+      "learning_rate": 8.603010572191106e-06,
+      "loss": 12.0616,
+      "step": 14941
+    },
+    {
+      "epoch": 2.613071038251366,
+      "grad_norm": 0.8073008724193541,
+      "learning_rate": 8.595350672676782e-06,
+      "loss": 11.8577,
+      "step": 14942
+    },
+    {
+      "epoch": 2.613245901639344,
+      "grad_norm": 0.8473001298155753,
+      "learning_rate": 8.587694031631422e-06,
+      "loss": 11.8729,
+      "step": 14943
+    },
+    {
+      "epoch": 2.6134207650273225,
+      "grad_norm": 0.8494017207524522,
+      "learning_rate": 8.580040649327958e-06,
+      "loss": 11.9386,
+      "step": 14944
+    },
+    {
+      "epoch": 2.6135956284153004,
+      "grad_norm": 0.7845074959643483,
+      "learning_rate": 8.572390526039231e-06,
+      "loss": 11.7171,
+      "step": 14945
+    },
+    {
+      "epoch": 2.6137704918032787,
+      "grad_norm": 0.8404527678973054,
+      "learning_rate": 8.564743662037965e-06,
+      "loss": 12.1559,
+      "step": 14946
+    },
+    {
+      "epoch": 2.613945355191257,
+      "grad_norm": 0.769006112089856,
+      "learning_rate": 8.557100057596756e-06,
+      "loss": 11.9004,
+      "step": 14947
+    },
+    {
+      "epoch": 2.614120218579235,
+      "grad_norm": 0.8596586040673321,
+      "learning_rate": 8.549459712988084e-06,
+      "loss": 11.5231,
+      "step": 14948
+    },
+    {
+      "epoch": 2.6142950819672133,
+      "grad_norm": 0.8269595101193089,
+      "learning_rate": 8.541822628484319e-06,
+      "loss": 11.8597,
+      "step": 14949
+    },
+    {
+      "epoch": 2.614469945355191,
+      "grad_norm": 0.7727605939116106,
+      "learning_rate": 8.534188804357746e-06,
+      "loss": 12.0014,
+      "step": 14950
+    },
+    {
+      "epoch": 2.6146448087431695,
+      "grad_norm": 0.85285911951678,
+      "learning_rate": 8.526558240880445e-06,
+      "loss": 11.8701,
+      "step": 14951
+    },
+    {
+      "epoch": 2.6148196721311474,
+      "grad_norm": 0.8050818580412012,
+      "learning_rate": 8.518930938324477e-06,
+      "loss": 11.7007,
+      "step": 14952
+    },
+    {
+      "epoch": 2.6149945355191258,
+      "grad_norm": 0.8173436763913189,
+      "learning_rate": 8.511306896961735e-06,
+      "loss": 11.8633,
+      "step": 14953
+    },
+    {
+      "epoch": 2.6151693989071036,
+      "grad_norm": 0.814718427798427,
+      "learning_rate": 8.503686117064036e-06,
+      "loss": 11.8243,
+      "step": 14954
+    },
+    {
+      "epoch": 2.615344262295082,
+      "grad_norm": 0.7936865169868467,
+      "learning_rate": 8.496068598902996e-06,
+      "loss": 12.0457,
+      "step": 14955
+    },
+    {
+      "epoch": 2.6155191256830603,
+      "grad_norm": 0.827926832564677,
+      "learning_rate": 8.488454342750196e-06,
+      "loss": 11.7172,
+      "step": 14956
+    },
+    {
+      "epoch": 2.6156939890710382,
+      "grad_norm": 0.8411663160271791,
+      "learning_rate": 8.480843348877089e-06,
+      "loss": 12.0284,
+      "step": 14957
+    },
+    {
+      "epoch": 2.615868852459016,
+      "grad_norm": 0.8519773456212383,
+      "learning_rate": 8.473235617554986e-06,
+      "loss": 11.9348,
+      "step": 14958
+    },
+    {
+      "epoch": 2.6160437158469945,
+      "grad_norm": 0.7856265062938883,
+      "learning_rate": 8.465631149055097e-06,
+      "loss": 11.7094,
+      "step": 14959
+    },
+    {
+      "epoch": 2.616218579234973,
+      "grad_norm": 0.8185608615811851,
+      "learning_rate": 8.458029943648504e-06,
+      "loss": 12.088,
+      "step": 14960
+    },
+    {
+      "epoch": 2.6163934426229507,
+      "grad_norm": 0.8339237793286086,
+      "learning_rate": 8.450432001606212e-06,
+      "loss": 11.6904,
+      "step": 14961
+    },
+    {
+      "epoch": 2.616568306010929,
+      "grad_norm": 0.8320241538977604,
+      "learning_rate": 8.442837323199037e-06,
+      "loss": 11.7369,
+      "step": 14962
+    },
+    {
+      "epoch": 2.616743169398907,
+      "grad_norm": 0.8355105108295622,
+      "learning_rate": 8.43524590869773e-06,
+      "loss": 11.4674,
+      "step": 14963
+    },
+    {
+      "epoch": 2.6169180327868853,
+      "grad_norm": 0.8820531864400785,
+      "learning_rate": 8.427657758372943e-06,
+      "loss": 11.5312,
+      "step": 14964
+    },
+    {
+      "epoch": 2.6170928961748636,
+      "grad_norm": 0.8214787100942403,
+      "learning_rate": 8.420072872495167e-06,
+      "loss": 11.9843,
+      "step": 14965
+    },
+    {
+      "epoch": 2.6172677595628415,
+      "grad_norm": 0.8469328961200576,
+      "learning_rate": 8.412491251334775e-06,
+      "loss": 11.9648,
+      "step": 14966
+    },
+    {
+      "epoch": 2.6174426229508194,
+      "grad_norm": 0.8070638724302295,
+      "learning_rate": 8.404912895162075e-06,
+      "loss": 11.7647,
+      "step": 14967
+    },
+    {
+      "epoch": 2.6176174863387978,
+      "grad_norm": 0.8716445010595333,
+      "learning_rate": 8.397337804247207e-06,
+      "loss": 12.1181,
+      "step": 14968
+    },
+    {
+      "epoch": 2.617792349726776,
+      "grad_norm": 0.8389285034157056,
+      "learning_rate": 8.389765978860232e-06,
+      "loss": 11.9376,
+      "step": 14969
+    },
+    {
+      "epoch": 2.617967213114754,
+      "grad_norm": 0.8200358240539538,
+      "learning_rate": 8.382197419271054e-06,
+      "loss": 11.8518,
+      "step": 14970
+    },
+    {
+      "epoch": 2.6181420765027323,
+      "grad_norm": 0.7773236966006933,
+      "learning_rate": 8.374632125749494e-06,
+      "loss": 11.9869,
+      "step": 14971
+    },
+    {
+      "epoch": 2.6183169398907102,
+      "grad_norm": 0.9111026249632193,
+      "learning_rate": 8.367070098565232e-06,
+      "loss": 11.5563,
+      "step": 14972
+    },
+    {
+      "epoch": 2.6184918032786886,
+      "grad_norm": 0.7927067801353682,
+      "learning_rate": 8.359511337987891e-06,
+      "loss": 11.8484,
+      "step": 14973
+    },
+    {
+      "epoch": 2.618666666666667,
+      "grad_norm": 0.843019676819014,
+      "learning_rate": 8.351955844286884e-06,
+      "loss": 11.9683,
+      "step": 14974
+    },
+    {
+      "epoch": 2.618841530054645,
+      "grad_norm": 0.8701047128823646,
+      "learning_rate": 8.344403617731566e-06,
+      "loss": 11.9406,
+      "step": 14975
+    },
+    {
+      "epoch": 2.6190163934426227,
+      "grad_norm": 0.8708854177591571,
+      "learning_rate": 8.336854658591175e-06,
+      "loss": 11.916,
+      "step": 14976
+    },
+    {
+      "epoch": 2.619191256830601,
+      "grad_norm": 0.8288260488817234,
+      "learning_rate": 8.32930896713482e-06,
+      "loss": 11.4792,
+      "step": 14977
+    },
+    {
+      "epoch": 2.6193661202185794,
+      "grad_norm": 0.8232585377380505,
+      "learning_rate": 8.321766543631483e-06,
+      "loss": 11.7831,
+      "step": 14978
+    },
+    {
+      "epoch": 2.6195409836065573,
+      "grad_norm": 0.9118685905909565,
+      "learning_rate": 8.31422738835006e-06,
+      "loss": 12.0586,
+      "step": 14979
+    },
+    {
+      "epoch": 2.6197158469945356,
+      "grad_norm": 0.8516309720491421,
+      "learning_rate": 8.306691501559326e-06,
+      "loss": 11.9063,
+      "step": 14980
+    },
+    {
+      "epoch": 2.6198907103825135,
+      "grad_norm": 0.8392829154288586,
+      "learning_rate": 8.299158883527902e-06,
+      "loss": 11.8658,
+      "step": 14981
+    },
+    {
+      "epoch": 2.620065573770492,
+      "grad_norm": 0.8082112843588309,
+      "learning_rate": 8.291629534524303e-06,
+      "loss": 11.8416,
+      "step": 14982
+    },
+    {
+      "epoch": 2.62024043715847,
+      "grad_norm": 0.7909095640436087,
+      "learning_rate": 8.284103454816971e-06,
+      "loss": 11.95,
+      "step": 14983
+    },
+    {
+      "epoch": 2.620415300546448,
+      "grad_norm": 0.8604513000247042,
+      "learning_rate": 8.276580644674215e-06,
+      "loss": 11.886,
+      "step": 14984
+    },
+    {
+      "epoch": 2.620590163934426,
+      "grad_norm": 0.8355797912877886,
+      "learning_rate": 8.269061104364161e-06,
+      "loss": 11.789,
+      "step": 14985
+    },
+    {
+      "epoch": 2.6207650273224044,
+      "grad_norm": 0.8245034347924372,
+      "learning_rate": 8.261544834154922e-06,
+      "loss": 11.7606,
+      "step": 14986
+    },
+    {
+      "epoch": 2.6209398907103827,
+      "grad_norm": 0.8982289756465844,
+      "learning_rate": 8.254031834314403e-06,
+      "loss": 11.9558,
+      "step": 14987
+    },
+    {
+      "epoch": 2.6211147540983606,
+      "grad_norm": 0.8354638030399333,
+      "learning_rate": 8.246522105110488e-06,
+      "loss": 11.8295,
+      "step": 14988
+    },
+    {
+      "epoch": 2.621289617486339,
+      "grad_norm": 0.870200854688507,
+      "learning_rate": 8.239015646810833e-06,
+      "loss": 11.5652,
+      "step": 14989
+    },
+    {
+      "epoch": 2.621464480874317,
+      "grad_norm": 0.8227476851755536,
+      "learning_rate": 8.231512459683066e-06,
+      "loss": 11.6922,
+      "step": 14990
+    },
+    {
+      "epoch": 2.621639344262295,
+      "grad_norm": 0.8140829181107422,
+      "learning_rate": 8.224012543994653e-06,
+      "loss": 12.0033,
+      "step": 14991
+    },
+    {
+      "epoch": 2.621814207650273,
+      "grad_norm": 0.8611762533523449,
+      "learning_rate": 8.216515900012967e-06,
+      "loss": 11.8942,
+      "step": 14992
+    },
+    {
+      "epoch": 2.6219890710382514,
+      "grad_norm": 0.9121138754634285,
+      "learning_rate": 8.20902252800525e-06,
+      "loss": 11.9082,
+      "step": 14993
+    },
+    {
+      "epoch": 2.6221639344262293,
+      "grad_norm": 0.779445831691425,
+      "learning_rate": 8.201532428238633e-06,
+      "loss": 12.0087,
+      "step": 14994
+    },
+    {
+      "epoch": 2.6223387978142076,
+      "grad_norm": 0.8428484907671536,
+      "learning_rate": 8.194045600980148e-06,
+      "loss": 11.8391,
+      "step": 14995
+    },
+    {
+      "epoch": 2.622513661202186,
+      "grad_norm": 0.8543156090683948,
+      "learning_rate": 8.186562046496659e-06,
+      "loss": 11.99,
+      "step": 14996
+    },
+    {
+      "epoch": 2.622688524590164,
+      "grad_norm": 0.7913120258649828,
+      "learning_rate": 8.179081765054953e-06,
+      "loss": 11.6095,
+      "step": 14997
+    },
+    {
+      "epoch": 2.6228633879781422,
+      "grad_norm": 0.8822697780227857,
+      "learning_rate": 8.171604756921702e-06,
+      "loss": 11.646,
+      "step": 14998
+    },
+    {
+      "epoch": 2.62303825136612,
+      "grad_norm": 0.8197608741543965,
+      "learning_rate": 8.164131022363474e-06,
+      "loss": 11.9206,
+      "step": 14999
+    },
+    {
+      "epoch": 2.6232131147540985,
+      "grad_norm": 0.8221484416685482,
+      "learning_rate": 8.156660561646668e-06,
+      "loss": 11.9506,
+      "step": 15000
+    },
+    {
+      "epoch": 2.6233879781420764,
+      "grad_norm": 0.7860536727963783,
+      "learning_rate": 8.149193375037589e-06,
+      "loss": 11.9798,
+      "step": 15001
+    },
+    {
+      "epoch": 2.6235628415300547,
+      "grad_norm": 0.854602604561164,
+      "learning_rate": 8.141729462802461e-06,
+      "loss": 11.8374,
+      "step": 15002
+    },
+    {
+      "epoch": 2.6237377049180326,
+      "grad_norm": 0.8493090094127927,
+      "learning_rate": 8.134268825207358e-06,
+      "loss": 11.7963,
+      "step": 15003
+    },
+    {
+      "epoch": 2.623912568306011,
+      "grad_norm": 0.8251866390255355,
+      "learning_rate": 8.126811462518236e-06,
+      "loss": 11.8541,
+      "step": 15004
+    },
+    {
+      "epoch": 2.6240874316939893,
+      "grad_norm": 0.8424302558043925,
+      "learning_rate": 8.119357375000935e-06,
+      "loss": 11.5843,
+      "step": 15005
+    },
+    {
+      "epoch": 2.624262295081967,
+      "grad_norm": 0.79975324927842,
+      "learning_rate": 8.111906562921189e-06,
+      "loss": 11.851,
+      "step": 15006
+    },
+    {
+      "epoch": 2.6244371584699455,
+      "grad_norm": 0.7877493464309997,
+      "learning_rate": 8.104459026544631e-06,
+      "loss": 12.1742,
+      "step": 15007
+    },
+    {
+      "epoch": 2.6246120218579234,
+      "grad_norm": 0.7599654168152462,
+      "learning_rate": 8.097014766136734e-06,
+      "loss": 11.8249,
+      "step": 15008
+    },
+    {
+      "epoch": 2.6247868852459018,
+      "grad_norm": 0.8069859190185158,
+      "learning_rate": 8.089573781962878e-06,
+      "loss": 12.0239,
+      "step": 15009
+    },
+    {
+      "epoch": 2.6249617486338797,
+      "grad_norm": 0.8395616475164578,
+      "learning_rate": 8.082136074288327e-06,
+      "loss": 11.8718,
+      "step": 15010
+    },
+    {
+      "epoch": 2.625136612021858,
+      "grad_norm": 0.8340825528430915,
+      "learning_rate": 8.074701643378236e-06,
+      "loss": 11.8459,
+      "step": 15011
+    },
+    {
+      "epoch": 2.625311475409836,
+      "grad_norm": 0.8297122432488783,
+      "learning_rate": 8.067270489497624e-06,
+      "loss": 11.7088,
+      "step": 15012
+    },
+    {
+      "epoch": 2.6254863387978142,
+      "grad_norm": 0.8127766499571353,
+      "learning_rate": 8.059842612911406e-06,
+      "loss": 12.0585,
+      "step": 15013
+    },
+    {
+      "epoch": 2.6256612021857926,
+      "grad_norm": 0.8357931662900504,
+      "learning_rate": 8.0524180138844e-06,
+      "loss": 11.9642,
+      "step": 15014
+    },
+    {
+      "epoch": 2.6258360655737705,
+      "grad_norm": 0.7811169999480079,
+      "learning_rate": 8.04499669268124e-06,
+      "loss": 11.8357,
+      "step": 15015
+    },
+    {
+      "epoch": 2.6260109289617484,
+      "grad_norm": 0.8547497477582086,
+      "learning_rate": 8.037578649566513e-06,
+      "loss": 11.71,
+      "step": 15016
+    },
+    {
+      "epoch": 2.6261857923497267,
+      "grad_norm": 0.8830766581929118,
+      "learning_rate": 8.030163884804664e-06,
+      "loss": 11.9326,
+      "step": 15017
+    },
+    {
+      "epoch": 2.626360655737705,
+      "grad_norm": 0.8104216273601365,
+      "learning_rate": 8.022752398660027e-06,
+      "loss": 11.8526,
+      "step": 15018
+    },
+    {
+      "epoch": 2.626535519125683,
+      "grad_norm": 0.7666611815178742,
+      "learning_rate": 8.01534419139679e-06,
+      "loss": 11.9141,
+      "step": 15019
+    },
+    {
+      "epoch": 2.6267103825136613,
+      "grad_norm": 0.8772673988149772,
+      "learning_rate": 8.007939263279063e-06,
+      "loss": 11.6651,
+      "step": 15020
+    },
+    {
+      "epoch": 2.626885245901639,
+      "grad_norm": 0.8596125049755342,
+      "learning_rate": 8.000537614570825e-06,
+      "loss": 11.8109,
+      "step": 15021
+    },
+    {
+      "epoch": 2.6270601092896175,
+      "grad_norm": 0.805627365328295,
+      "learning_rate": 7.99313924553594e-06,
+      "loss": 11.8459,
+      "step": 15022
+    },
+    {
+      "epoch": 2.627234972677596,
+      "grad_norm": 0.841786673225906,
+      "learning_rate": 7.985744156438125e-06,
+      "loss": 11.9227,
+      "step": 15023
+    },
+    {
+      "epoch": 2.6274098360655738,
+      "grad_norm": 0.8434743274376555,
+      "learning_rate": 7.978352347541018e-06,
+      "loss": 11.6446,
+      "step": 15024
+    },
+    {
+      "epoch": 2.6275846994535517,
+      "grad_norm": 0.8171945806416886,
+      "learning_rate": 7.970963819108157e-06,
+      "loss": 11.6519,
+      "step": 15025
+    },
+    {
+      "epoch": 2.62775956284153,
+      "grad_norm": 0.8498189082390926,
+      "learning_rate": 7.963578571402896e-06,
+      "loss": 11.8065,
+      "step": 15026
+    },
+    {
+      "epoch": 2.6279344262295083,
+      "grad_norm": 0.7990401975729359,
+      "learning_rate": 7.956196604688526e-06,
+      "loss": 11.8577,
+      "step": 15027
+    },
+    {
+      "epoch": 2.6281092896174862,
+      "grad_norm": 0.8063134707552877,
+      "learning_rate": 7.948817919228214e-06,
+      "loss": 11.9142,
+      "step": 15028
+    },
+    {
+      "epoch": 2.6282841530054646,
+      "grad_norm": 0.8210223695364833,
+      "learning_rate": 7.94144251528498e-06,
+      "loss": 11.7923,
+      "step": 15029
+    },
+    {
+      "epoch": 2.6284590163934425,
+      "grad_norm": 0.8296345879610139,
+      "learning_rate": 7.934070393121762e-06,
+      "loss": 11.9437,
+      "step": 15030
+    },
+    {
+      "epoch": 2.628633879781421,
+      "grad_norm": 0.7787897976101607,
+      "learning_rate": 7.92670155300137e-06,
+      "loss": 11.6785,
+      "step": 15031
+    },
+    {
+      "epoch": 2.628808743169399,
+      "grad_norm": 0.7498036426719279,
+      "learning_rate": 7.919335995186494e-06,
+      "loss": 11.4815,
+      "step": 15032
+    },
+    {
+      "epoch": 2.628983606557377,
+      "grad_norm": 0.8449261409211215,
+      "learning_rate": 7.911973719939725e-06,
+      "loss": 12.0059,
+      "step": 15033
+    },
+    {
+      "epoch": 2.629158469945355,
+      "grad_norm": 0.8042676140284631,
+      "learning_rate": 7.904614727523473e-06,
+      "loss": 11.8738,
+      "step": 15034
+    },
+    {
+      "epoch": 2.6293333333333333,
+      "grad_norm": 0.9140446530346795,
+      "learning_rate": 7.89725901820011e-06,
+      "loss": 11.9285,
+      "step": 15035
+    },
+    {
+      "epoch": 2.6295081967213116,
+      "grad_norm": 0.8070344181294953,
+      "learning_rate": 7.889906592231844e-06,
+      "loss": 11.8424,
+      "step": 15036
+    },
+    {
+      "epoch": 2.6296830601092895,
+      "grad_norm": 0.7882854715433644,
+      "learning_rate": 7.882557449880823e-06,
+      "loss": 11.5712,
+      "step": 15037
+    },
+    {
+      "epoch": 2.629857923497268,
+      "grad_norm": 0.8198208085295822,
+      "learning_rate": 7.87521159140897e-06,
+      "loss": 11.8733,
+      "step": 15038
+    },
+    {
+      "epoch": 2.6300327868852458,
+      "grad_norm": 0.8465355170421689,
+      "learning_rate": 7.867869017078188e-06,
+      "loss": 11.9123,
+      "step": 15039
+    },
+    {
+      "epoch": 2.630207650273224,
+      "grad_norm": 0.8211340384206121,
+      "learning_rate": 7.860529727150256e-06,
+      "loss": 11.8376,
+      "step": 15040
+    },
+    {
+      "epoch": 2.6303825136612025,
+      "grad_norm": 0.8628838904304089,
+      "learning_rate": 7.853193721886764e-06,
+      "loss": 11.8074,
+      "step": 15041
+    },
+    {
+      "epoch": 2.6305573770491804,
+      "grad_norm": 0.7903172257230229,
+      "learning_rate": 7.84586100154925e-06,
+      "loss": 11.4763,
+      "step": 15042
+    },
+    {
+      "epoch": 2.6307322404371583,
+      "grad_norm": 0.8740111855438755,
+      "learning_rate": 7.83853156639912e-06,
+      "loss": 11.9029,
+      "step": 15043
+    },
+    {
+      "epoch": 2.6309071038251366,
+      "grad_norm": 0.7891515844723185,
+      "learning_rate": 7.831205416697685e-06,
+      "loss": 11.8095,
+      "step": 15044
+    },
+    {
+      "epoch": 2.631081967213115,
+      "grad_norm": 0.7714065816588905,
+      "learning_rate": 7.823882552706075e-06,
+      "loss": 11.9203,
+      "step": 15045
+    },
+    {
+      "epoch": 2.631256830601093,
+      "grad_norm": 0.8170149736809706,
+      "learning_rate": 7.81656297468535e-06,
+      "loss": 11.9826,
+      "step": 15046
+    },
+    {
+      "epoch": 2.631431693989071,
+      "grad_norm": 0.827752761241212,
+      "learning_rate": 7.809246682896453e-06,
+      "loss": 11.993,
+      "step": 15047
+    },
+    {
+      "epoch": 2.631606557377049,
+      "grad_norm": 0.8363308890408037,
+      "learning_rate": 7.801933677600192e-06,
+      "loss": 11.8822,
+      "step": 15048
+    },
+    {
+      "epoch": 2.6317814207650274,
+      "grad_norm": 0.8186199012876777,
+      "learning_rate": 7.794623959057279e-06,
+      "loss": 12.0029,
+      "step": 15049
+    },
+    {
+      "epoch": 2.6319562841530053,
+      "grad_norm": 0.8636851065950727,
+      "learning_rate": 7.787317527528293e-06,
+      "loss": 11.7107,
+      "step": 15050
+    },
+    {
+      "epoch": 2.6321311475409837,
+      "grad_norm": 0.8699820954584425,
+      "learning_rate": 7.780014383273693e-06,
+      "loss": 11.6031,
+      "step": 15051
+    },
+    {
+      "epoch": 2.6323060109289615,
+      "grad_norm": 0.7988786456331856,
+      "learning_rate": 7.772714526553848e-06,
+      "loss": 12.0664,
+      "step": 15052
+    },
+    {
+      "epoch": 2.63248087431694,
+      "grad_norm": 0.8796480470872985,
+      "learning_rate": 7.765417957628962e-06,
+      "loss": 11.9293,
+      "step": 15053
+    },
+    {
+      "epoch": 2.6326557377049182,
+      "grad_norm": 0.8851487518163026,
+      "learning_rate": 7.75812467675916e-06,
+      "loss": 11.4614,
+      "step": 15054
+    },
+    {
+      "epoch": 2.632830601092896,
+      "grad_norm": 0.8315997237308315,
+      "learning_rate": 7.750834684204455e-06,
+      "loss": 11.9494,
+      "step": 15055
+    },
+    {
+      "epoch": 2.6330054644808745,
+      "grad_norm": 0.9214971117688645,
+      "learning_rate": 7.743547980224697e-06,
+      "loss": 11.7928,
+      "step": 15056
+    },
+    {
+      "epoch": 2.6331803278688524,
+      "grad_norm": 0.797526435370143,
+      "learning_rate": 7.736264565079664e-06,
+      "loss": 11.9084,
+      "step": 15057
+    },
+    {
+      "epoch": 2.6333551912568307,
+      "grad_norm": 0.8717688767235385,
+      "learning_rate": 7.728984439028997e-06,
+      "loss": 12.1431,
+      "step": 15058
+    },
+    {
+      "epoch": 2.6335300546448086,
+      "grad_norm": 0.8231455944325284,
+      "learning_rate": 7.72170760233225e-06,
+      "loss": 11.8861,
+      "step": 15059
+    },
+    {
+      "epoch": 2.633704918032787,
+      "grad_norm": 0.8605167571911748,
+      "learning_rate": 7.714434055248798e-06,
+      "loss": 11.7326,
+      "step": 15060
+    },
+    {
+      "epoch": 2.633879781420765,
+      "grad_norm": 0.7893579138999312,
+      "learning_rate": 7.707163798037943e-06,
+      "loss": 11.9459,
+      "step": 15061
+    },
+    {
+      "epoch": 2.634054644808743,
+      "grad_norm": 0.8515616065206558,
+      "learning_rate": 7.699896830958864e-06,
+      "loss": 11.8129,
+      "step": 15062
+    },
+    {
+      "epoch": 2.6342295081967215,
+      "grad_norm": 0.9006006908428409,
+      "learning_rate": 7.692633154270634e-06,
+      "loss": 11.4661,
+      "step": 15063
+    },
+    {
+      "epoch": 2.6344043715846994,
+      "grad_norm": 0.8182945616169072,
+      "learning_rate": 7.685372768232179e-06,
+      "loss": 11.8098,
+      "step": 15064
+    },
+    {
+      "epoch": 2.6345792349726773,
+      "grad_norm": 0.839100427232209,
+      "learning_rate": 7.678115673102315e-06,
+      "loss": 12.0105,
+      "step": 15065
+    },
+    {
+      "epoch": 2.6347540983606557,
+      "grad_norm": 0.8244279520528541,
+      "learning_rate": 7.670861869139767e-06,
+      "loss": 11.5882,
+      "step": 15066
+    },
+    {
+      "epoch": 2.634928961748634,
+      "grad_norm": 0.8597501278872651,
+      "learning_rate": 7.663611356603118e-06,
+      "loss": 11.8345,
+      "step": 15067
+    },
+    {
+      "epoch": 2.635103825136612,
+      "grad_norm": 0.8257519895656761,
+      "learning_rate": 7.656364135750838e-06,
+      "loss": 11.9259,
+      "step": 15068
+    },
+    {
+      "epoch": 2.6352786885245902,
+      "grad_norm": 0.8732581455194346,
+      "learning_rate": 7.649120206841288e-06,
+      "loss": 11.7527,
+      "step": 15069
+    },
+    {
+      "epoch": 2.635453551912568,
+      "grad_norm": 0.8672497665015364,
+      "learning_rate": 7.641879570132715e-06,
+      "loss": 11.543,
+      "step": 15070
+    },
+    {
+      "epoch": 2.6356284153005465,
+      "grad_norm": 0.7810963628410091,
+      "learning_rate": 7.634642225883204e-06,
+      "loss": 12.0565,
+      "step": 15071
+    },
+    {
+      "epoch": 2.635803278688525,
+      "grad_norm": 0.8792534895399116,
+      "learning_rate": 7.627408174350792e-06,
+      "loss": 11.7105,
+      "step": 15072
+    },
+    {
+      "epoch": 2.6359781420765027,
+      "grad_norm": 0.8831092757199196,
+      "learning_rate": 7.62017741579335e-06,
+      "loss": 11.3631,
+      "step": 15073
+    },
+    {
+      "epoch": 2.6361530054644806,
+      "grad_norm": 0.7648969107148597,
+      "learning_rate": 7.612949950468662e-06,
+      "loss": 12.1621,
+      "step": 15074
+    },
+    {
+      "epoch": 2.636327868852459,
+      "grad_norm": 0.8122929920760062,
+      "learning_rate": 7.605725778634354e-06,
+      "loss": 11.8129,
+      "step": 15075
+    },
+    {
+      "epoch": 2.6365027322404373,
+      "grad_norm": 0.8518086278659559,
+      "learning_rate": 7.598504900547965e-06,
+      "loss": 12.066,
+      "step": 15076
+    },
+    {
+      "epoch": 2.636677595628415,
+      "grad_norm": 0.8339406983818753,
+      "learning_rate": 7.591287316466922e-06,
+      "loss": 11.822,
+      "step": 15077
+    },
+    {
+      "epoch": 2.6368524590163935,
+      "grad_norm": 0.8322386726635093,
+      "learning_rate": 7.58407302664853e-06,
+      "loss": 11.8374,
+      "step": 15078
+    },
+    {
+      "epoch": 2.6370273224043714,
+      "grad_norm": 0.7926350057189026,
+      "learning_rate": 7.576862031349941e-06,
+      "loss": 11.7419,
+      "step": 15079
+    },
+    {
+      "epoch": 2.6372021857923498,
+      "grad_norm": 0.8637514232126547,
+      "learning_rate": 7.5696543308282465e-06,
+      "loss": 11.9525,
+      "step": 15080
+    },
+    {
+      "epoch": 2.637377049180328,
+      "grad_norm": 0.8261807880325912,
+      "learning_rate": 7.5624499253403645e-06,
+      "loss": 11.5877,
+      "step": 15081
+    },
+    {
+      "epoch": 2.637551912568306,
+      "grad_norm": 0.8148850770647234,
+      "learning_rate": 7.555248815143168e-06,
+      "loss": 11.6566,
+      "step": 15082
+    },
+    {
+      "epoch": 2.637726775956284,
+      "grad_norm": 0.888365221738737,
+      "learning_rate": 7.548051000493328e-06,
+      "loss": 11.9574,
+      "step": 15083
+    },
+    {
+      "epoch": 2.6379016393442623,
+      "grad_norm": 0.833407365623943,
+      "learning_rate": 7.540856481647452e-06,
+      "loss": 11.8046,
+      "step": 15084
+    },
+    {
+      "epoch": 2.6380765027322406,
+      "grad_norm": 0.8664162382449685,
+      "learning_rate": 7.5336652588619995e-06,
+      "loss": 11.7245,
+      "step": 15085
+    },
+    {
+      "epoch": 2.6382513661202185,
+      "grad_norm": 0.8109383313560323,
+      "learning_rate": 7.5264773323933555e-06,
+      "loss": 12.1437,
+      "step": 15086
+    },
+    {
+      "epoch": 2.638426229508197,
+      "grad_norm": 0.8704877285041138,
+      "learning_rate": 7.51929270249776e-06,
+      "loss": 11.9289,
+      "step": 15087
+    },
+    {
+      "epoch": 2.6386010928961747,
+      "grad_norm": 0.8124481340622546,
+      "learning_rate": 7.512111369431318e-06,
+      "loss": 11.5173,
+      "step": 15088
+    },
+    {
+      "epoch": 2.638775956284153,
+      "grad_norm": 0.8284168222607866,
+      "learning_rate": 7.5049333334500596e-06,
+      "loss": 11.8087,
+      "step": 15089
+    },
+    {
+      "epoch": 2.6389508196721314,
+      "grad_norm": 0.8147693146094733,
+      "learning_rate": 7.497758594809845e-06,
+      "loss": 11.831,
+      "step": 15090
+    },
+    {
+      "epoch": 2.6391256830601093,
+      "grad_norm": 0.782619505814614,
+      "learning_rate": 7.49058715376646e-06,
+      "loss": 11.7095,
+      "step": 15091
+    },
+    {
+      "epoch": 2.639300546448087,
+      "grad_norm": 0.8252408027676441,
+      "learning_rate": 7.483419010575554e-06,
+      "loss": 11.9569,
+      "step": 15092
+    },
+    {
+      "epoch": 2.6394754098360655,
+      "grad_norm": 0.8481597655863877,
+      "learning_rate": 7.476254165492691e-06,
+      "loss": 11.738,
+      "step": 15093
+    },
+    {
+      "epoch": 2.639650273224044,
+      "grad_norm": 0.8134783643214111,
+      "learning_rate": 7.469092618773243e-06,
+      "loss": 11.8893,
+      "step": 15094
+    },
+    {
+      "epoch": 2.639825136612022,
+      "grad_norm": 0.850506121065754,
+      "learning_rate": 7.461934370672529e-06,
+      "loss": 11.9939,
+      "step": 15095
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.824806625443698,
+      "learning_rate": 7.454779421445746e-06,
+      "loss": 11.7133,
+      "step": 15096
+    },
+    {
+      "epoch": 2.640174863387978,
+      "grad_norm": 0.8075719685127626,
+      "learning_rate": 7.447627771347954e-06,
+      "loss": 11.8596,
+      "step": 15097
+    },
+    {
+      "epoch": 2.6403497267759564,
+      "grad_norm": 0.8194774720005993,
+      "learning_rate": 7.440479420634083e-06,
+      "loss": 11.6929,
+      "step": 15098
+    },
+    {
+      "epoch": 2.6405245901639343,
+      "grad_norm": 0.9073165348961973,
+      "learning_rate": 7.433334369558986e-06,
+      "loss": 11.7655,
+      "step": 15099
+    },
+    {
+      "epoch": 2.6406994535519126,
+      "grad_norm": 0.7945215068028392,
+      "learning_rate": 7.4261926183773475e-06,
+      "loss": 11.8022,
+      "step": 15100
+    },
+    {
+      "epoch": 2.6408743169398905,
+      "grad_norm": 0.7773387410680699,
+      "learning_rate": 7.419054167343797e-06,
+      "loss": 11.9242,
+      "step": 15101
+    },
+    {
+      "epoch": 2.641049180327869,
+      "grad_norm": 0.8433779983018361,
+      "learning_rate": 7.411919016712787e-06,
+      "loss": 11.9685,
+      "step": 15102
+    },
+    {
+      "epoch": 2.641224043715847,
+      "grad_norm": 0.768535519505352,
+      "learning_rate": 7.4047871667386916e-06,
+      "loss": 11.6452,
+      "step": 15103
+    },
+    {
+      "epoch": 2.641398907103825,
+      "grad_norm": 0.8100182613018007,
+      "learning_rate": 7.3976586176757625e-06,
+      "loss": 11.849,
+      "step": 15104
+    },
+    {
+      "epoch": 2.6415737704918034,
+      "grad_norm": 0.8524735379049018,
+      "learning_rate": 7.390533369778096e-06,
+      "loss": 11.8711,
+      "step": 15105
+    },
+    {
+      "epoch": 2.6417486338797813,
+      "grad_norm": 0.8116466566057793,
+      "learning_rate": 7.383411423299713e-06,
+      "loss": 12.0158,
+      "step": 15106
+    },
+    {
+      "epoch": 2.6419234972677597,
+      "grad_norm": 0.7862107109886474,
+      "learning_rate": 7.376292778494498e-06,
+      "loss": 11.778,
+      "step": 15107
+    },
+    {
+      "epoch": 2.6420983606557376,
+      "grad_norm": 0.8176690151518679,
+      "learning_rate": 7.369177435616248e-06,
+      "loss": 11.8157,
+      "step": 15108
+    },
+    {
+      "epoch": 2.642273224043716,
+      "grad_norm": 0.823125638440204,
+      "learning_rate": 7.3620653949185715e-06,
+      "loss": 11.9885,
+      "step": 15109
+    },
+    {
+      "epoch": 2.642448087431694,
+      "grad_norm": 0.8902172128131899,
+      "learning_rate": 7.354956656655032e-06,
+      "loss": 11.9022,
+      "step": 15110
+    },
+    {
+      "epoch": 2.642622950819672,
+      "grad_norm": 0.7967762728936711,
+      "learning_rate": 7.347851221079039e-06,
+      "loss": 11.8076,
+      "step": 15111
+    },
+    {
+      "epoch": 2.6427978142076505,
+      "grad_norm": 0.8627616825419515,
+      "learning_rate": 7.34074908844391e-06,
+      "loss": 11.8632,
+      "step": 15112
+    },
+    {
+      "epoch": 2.6429726775956284,
+      "grad_norm": 0.7944350217943962,
+      "learning_rate": 7.333650259002811e-06,
+      "loss": 11.929,
+      "step": 15113
+    },
+    {
+      "epoch": 2.6431475409836067,
+      "grad_norm": 0.8241851604894413,
+      "learning_rate": 7.326554733008806e-06,
+      "loss": 11.8379,
+      "step": 15114
+    },
+    {
+      "epoch": 2.6433224043715846,
+      "grad_norm": 0.8484351715367187,
+      "learning_rate": 7.319462510714869e-06,
+      "loss": 12.048,
+      "step": 15115
+    },
+    {
+      "epoch": 2.643497267759563,
+      "grad_norm": 0.8154160951409911,
+      "learning_rate": 7.312373592373778e-06,
+      "loss": 11.8871,
+      "step": 15116
+    },
+    {
+      "epoch": 2.643672131147541,
+      "grad_norm": 0.8041708544055172,
+      "learning_rate": 7.305287978238284e-06,
+      "loss": 11.8942,
+      "step": 15117
+    },
+    {
+      "epoch": 2.643846994535519,
+      "grad_norm": 0.7627561510751408,
+      "learning_rate": 7.298205668560976e-06,
+      "loss": 12.1042,
+      "step": 15118
+    },
+    {
+      "epoch": 2.644021857923497,
+      "grad_norm": 0.84333537561062,
+      "learning_rate": 7.291126663594317e-06,
+      "loss": 11.6355,
+      "step": 15119
+    },
+    {
+      "epoch": 2.6441967213114754,
+      "grad_norm": 0.8104128803981083,
+      "learning_rate": 7.284050963590672e-06,
+      "loss": 11.8074,
+      "step": 15120
+    },
+    {
+      "epoch": 2.6443715846994538,
+      "grad_norm": 0.8012071137771691,
+      "learning_rate": 7.276978568802295e-06,
+      "loss": 11.7583,
+      "step": 15121
+    },
+    {
+      "epoch": 2.6445464480874317,
+      "grad_norm": 0.7638794560858969,
+      "learning_rate": 7.269909479481285e-06,
+      "loss": 11.857,
+      "step": 15122
+    },
+    {
+      "epoch": 2.6447213114754096,
+      "grad_norm": 0.7769794527267032,
+      "learning_rate": 7.262843695879684e-06,
+      "loss": 11.8959,
+      "step": 15123
+    },
+    {
+      "epoch": 2.644896174863388,
+      "grad_norm": 0.8057043357589112,
+      "learning_rate": 7.2557812182493355e-06,
+      "loss": 11.8123,
+      "step": 15124
+    },
+    {
+      "epoch": 2.6450710382513662,
+      "grad_norm": 0.8573201924617998,
+      "learning_rate": 7.248722046842015e-06,
+      "loss": 11.8025,
+      "step": 15125
+    },
+    {
+      "epoch": 2.645245901639344,
+      "grad_norm": 0.8024175002401581,
+      "learning_rate": 7.241666181909401e-06,
+      "loss": 11.9327,
+      "step": 15126
+    },
+    {
+      "epoch": 2.6454207650273225,
+      "grad_norm": 0.8249604814704927,
+      "learning_rate": 7.234613623703024e-06,
+      "loss": 11.7521,
+      "step": 15127
+    },
+    {
+      "epoch": 2.6455956284153004,
+      "grad_norm": 0.8543662005336915,
+      "learning_rate": 7.2275643724742605e-06,
+      "loss": 11.9258,
+      "step": 15128
+    },
+    {
+      "epoch": 2.6457704918032787,
+      "grad_norm": 0.8575892904977154,
+      "learning_rate": 7.2205184284744436e-06,
+      "loss": 11.7753,
+      "step": 15129
+    },
+    {
+      "epoch": 2.645945355191257,
+      "grad_norm": 0.8269589174476317,
+      "learning_rate": 7.213475791954738e-06,
+      "loss": 11.6429,
+      "step": 15130
+    },
+    {
+      "epoch": 2.646120218579235,
+      "grad_norm": 0.8633503160750923,
+      "learning_rate": 7.206436463166233e-06,
+      "loss": 11.8097,
+      "step": 15131
+    },
+    {
+      "epoch": 2.646295081967213,
+      "grad_norm": 0.8050243900039792,
+      "learning_rate": 7.199400442359827e-06,
+      "loss": 11.9172,
+      "step": 15132
+    },
+    {
+      "epoch": 2.646469945355191,
+      "grad_norm": 0.8434066796648761,
+      "learning_rate": 7.192367729786375e-06,
+      "loss": 11.8899,
+      "step": 15133
+    },
+    {
+      "epoch": 2.6466448087431695,
+      "grad_norm": 0.9273883209366862,
+      "learning_rate": 7.1853383256965885e-06,
+      "loss": 12.0558,
+      "step": 15134
+    },
+    {
+      "epoch": 2.6468196721311474,
+      "grad_norm": 0.7965717843506839,
+      "learning_rate": 7.178312230341044e-06,
+      "loss": 11.832,
+      "step": 15135
+    },
+    {
+      "epoch": 2.646994535519126,
+      "grad_norm": 0.9079719630072004,
+      "learning_rate": 7.171289443970209e-06,
+      "loss": 12.0503,
+      "step": 15136
+    },
+    {
+      "epoch": 2.6471693989071037,
+      "grad_norm": 0.9078827473886443,
+      "learning_rate": 7.164269966834447e-06,
+      "loss": 11.8041,
+      "step": 15137
+    },
+    {
+      "epoch": 2.647344262295082,
+      "grad_norm": 0.8897911587388053,
+      "learning_rate": 7.157253799184005e-06,
+      "loss": 11.3944,
+      "step": 15138
+    },
+    {
+      "epoch": 2.6475191256830604,
+      "grad_norm": 0.8489656951071028,
+      "learning_rate": 7.150240941268971e-06,
+      "loss": 11.755,
+      "step": 15139
+    },
+    {
+      "epoch": 2.6476939890710383,
+      "grad_norm": 0.8257654766024564,
+      "learning_rate": 7.143231393339378e-06,
+      "loss": 11.6446,
+      "step": 15140
+    },
+    {
+      "epoch": 2.647868852459016,
+      "grad_norm": 0.7739727654630202,
+      "learning_rate": 7.136225155645093e-06,
+      "loss": 11.8294,
+      "step": 15141
+    },
+    {
+      "epoch": 2.6480437158469945,
+      "grad_norm": 0.8012444462892941,
+      "learning_rate": 7.129222228435894e-06,
+      "loss": 12.0806,
+      "step": 15142
+    },
+    {
+      "epoch": 2.648218579234973,
+      "grad_norm": 0.7914901741594199,
+      "learning_rate": 7.1222226119614046e-06,
+      "loss": 12.0884,
+      "step": 15143
+    },
+    {
+      "epoch": 2.6483934426229507,
+      "grad_norm": 0.8091119819503385,
+      "learning_rate": 7.1152263064711566e-06,
+      "loss": 11.8974,
+      "step": 15144
+    },
+    {
+      "epoch": 2.648568306010929,
+      "grad_norm": 0.8548173117124325,
+      "learning_rate": 7.1082333122145625e-06,
+      "loss": 11.7889,
+      "step": 15145
+    },
+    {
+      "epoch": 2.648743169398907,
+      "grad_norm": 0.7976070161288197,
+      "learning_rate": 7.1012436294409455e-06,
+      "loss": 11.8669,
+      "step": 15146
+    },
+    {
+      "epoch": 2.6489180327868853,
+      "grad_norm": 0.8524782236629231,
+      "learning_rate": 7.094257258399428e-06,
+      "loss": 12.0046,
+      "step": 15147
+    },
+    {
+      "epoch": 2.6490928961748637,
+      "grad_norm": 0.8189441901191146,
+      "learning_rate": 7.087274199339089e-06,
+      "loss": 11.8718,
+      "step": 15148
+    },
+    {
+      "epoch": 2.6492677595628416,
+      "grad_norm": 0.8356351684272939,
+      "learning_rate": 7.080294452508895e-06,
+      "loss": 11.9593,
+      "step": 15149
+    },
+    {
+      "epoch": 2.6494426229508194,
+      "grad_norm": 0.8585149989833287,
+      "learning_rate": 7.0733180181576155e-06,
+      "loss": 11.8084,
+      "step": 15150
+    },
+    {
+      "epoch": 2.649617486338798,
+      "grad_norm": 0.864434557141495,
+      "learning_rate": 7.066344896533972e-06,
+      "loss": 11.8498,
+      "step": 15151
+    },
+    {
+      "epoch": 2.649792349726776,
+      "grad_norm": 0.839463947222603,
+      "learning_rate": 7.059375087886555e-06,
+      "loss": 12.0162,
+      "step": 15152
+    },
+    {
+      "epoch": 2.649967213114754,
+      "grad_norm": 0.8731630950822759,
+      "learning_rate": 7.052408592463844e-06,
+      "loss": 11.7219,
+      "step": 15153
+    },
+    {
+      "epoch": 2.6501420765027324,
+      "grad_norm": 0.8544834276583652,
+      "learning_rate": 7.045445410514162e-06,
+      "loss": 11.916,
+      "step": 15154
+    },
+    {
+      "epoch": 2.6503169398907103,
+      "grad_norm": 0.8016005206679067,
+      "learning_rate": 7.038485542285733e-06,
+      "loss": 11.9018,
+      "step": 15155
+    },
+    {
+      "epoch": 2.6504918032786886,
+      "grad_norm": 0.8146878602502591,
+      "learning_rate": 7.031528988026681e-06,
+      "loss": 12.014,
+      "step": 15156
+    },
+    {
+      "epoch": 2.6506666666666665,
+      "grad_norm": 0.8038437567816485,
+      "learning_rate": 7.024575747985007e-06,
+      "loss": 11.7512,
+      "step": 15157
+    },
+    {
+      "epoch": 2.650841530054645,
+      "grad_norm": 0.8137063570686769,
+      "learning_rate": 7.017625822408569e-06,
+      "loss": 12.0344,
+      "step": 15158
+    },
+    {
+      "epoch": 2.6510163934426227,
+      "grad_norm": 0.8447960688850831,
+      "learning_rate": 7.010679211545135e-06,
+      "loss": 11.9504,
+      "step": 15159
+    },
+    {
+      "epoch": 2.651191256830601,
+      "grad_norm": 0.8232320010967477,
+      "learning_rate": 7.00373591564234e-06,
+      "loss": 11.9803,
+      "step": 15160
+    },
+    {
+      "epoch": 2.6513661202185794,
+      "grad_norm": 0.8869937247410501,
+      "learning_rate": 6.996795934947719e-06,
+      "loss": 12.1643,
+      "step": 15161
+    },
+    {
+      "epoch": 2.6515409836065573,
+      "grad_norm": 0.8350973777223798,
+      "learning_rate": 6.989859269708654e-06,
+      "loss": 11.6968,
+      "step": 15162
+    },
+    {
+      "epoch": 2.6517158469945357,
+      "grad_norm": 0.9124417342777476,
+      "learning_rate": 6.9829259201724225e-06,
+      "loss": 12.0084,
+      "step": 15163
+    },
+    {
+      "epoch": 2.6518907103825136,
+      "grad_norm": 0.8276086964113561,
+      "learning_rate": 6.9759958865862284e-06,
+      "loss": 11.7992,
+      "step": 15164
+    },
+    {
+      "epoch": 2.652065573770492,
+      "grad_norm": 0.8722023649043984,
+      "learning_rate": 6.969069169197073e-06,
+      "loss": 11.5588,
+      "step": 15165
+    },
+    {
+      "epoch": 2.65224043715847,
+      "grad_norm": 0.8782284153698875,
+      "learning_rate": 6.962145768251904e-06,
+      "loss": 11.8185,
+      "step": 15166
+    },
+    {
+      "epoch": 2.652415300546448,
+      "grad_norm": 0.819685737432756,
+      "learning_rate": 6.955225683997546e-06,
+      "loss": 11.7982,
+      "step": 15167
+    },
+    {
+      "epoch": 2.652590163934426,
+      "grad_norm": 0.8241688091752476,
+      "learning_rate": 6.948308916680691e-06,
+      "loss": 11.7409,
+      "step": 15168
+    },
+    {
+      "epoch": 2.6527650273224044,
+      "grad_norm": 0.8195761405342036,
+      "learning_rate": 6.941395466547884e-06,
+      "loss": 11.6835,
+      "step": 15169
+    },
+    {
+      "epoch": 2.6529398907103827,
+      "grad_norm": 0.8682657806522617,
+      "learning_rate": 6.9344853338456086e-06,
+      "loss": 12.024,
+      "step": 15170
+    },
+    {
+      "epoch": 2.6531147540983606,
+      "grad_norm": 0.8444127035409664,
+      "learning_rate": 6.927578518820188e-06,
+      "loss": 11.9221,
+      "step": 15171
+    },
+    {
+      "epoch": 2.653289617486339,
+      "grad_norm": 0.8294662190963231,
+      "learning_rate": 6.920675021717871e-06,
+      "loss": 11.8406,
+      "step": 15172
+    },
+    {
+      "epoch": 2.653464480874317,
+      "grad_norm": 0.8483829603738361,
+      "learning_rate": 6.913774842784726e-06,
+      "loss": 11.8432,
+      "step": 15173
+    },
+    {
+      "epoch": 2.653639344262295,
+      "grad_norm": 0.8330304824482361,
+      "learning_rate": 6.906877982266735e-06,
+      "loss": 11.6114,
+      "step": 15174
+    },
+    {
+      "epoch": 2.653814207650273,
+      "grad_norm": 0.7866671677142041,
+      "learning_rate": 6.899984440409791e-06,
+      "loss": 11.914,
+      "step": 15175
+    },
+    {
+      "epoch": 2.6539890710382514,
+      "grad_norm": 0.8385589446910081,
+      "learning_rate": 6.893094217459617e-06,
+      "loss": 11.8954,
+      "step": 15176
+    },
+    {
+      "epoch": 2.6541639344262293,
+      "grad_norm": 0.852997088948036,
+      "learning_rate": 6.886207313661852e-06,
+      "loss": 11.5827,
+      "step": 15177
+    },
+    {
+      "epoch": 2.6543387978142077,
+      "grad_norm": 0.8167294649322964,
+      "learning_rate": 6.8793237292620105e-06,
+      "loss": 11.8403,
+      "step": 15178
+    },
+    {
+      "epoch": 2.654513661202186,
+      "grad_norm": 0.8878131742806221,
+      "learning_rate": 6.8724434645054844e-06,
+      "loss": 11.8215,
+      "step": 15179
+    },
+    {
+      "epoch": 2.654688524590164,
+      "grad_norm": 0.8165846312910394,
+      "learning_rate": 6.865566519637534e-06,
+      "loss": 11.7208,
+      "step": 15180
+    },
+    {
+      "epoch": 2.654863387978142,
+      "grad_norm": 0.7654381551277791,
+      "learning_rate": 6.858692894903318e-06,
+      "loss": 11.5639,
+      "step": 15181
+    },
+    {
+      "epoch": 2.65503825136612,
+      "grad_norm": 0.8413637115595264,
+      "learning_rate": 6.851822590547874e-06,
+      "loss": 11.7499,
+      "step": 15182
+    },
+    {
+      "epoch": 2.6552131147540985,
+      "grad_norm": 0.7722110836379704,
+      "learning_rate": 6.844955606816139e-06,
+      "loss": 11.692,
+      "step": 15183
+    },
+    {
+      "epoch": 2.6553879781420764,
+      "grad_norm": 0.8381513177775318,
+      "learning_rate": 6.838091943952885e-06,
+      "loss": 11.9118,
+      "step": 15184
+    },
+    {
+      "epoch": 2.6555628415300547,
+      "grad_norm": 0.7963110530095747,
+      "learning_rate": 6.831231602202803e-06,
+      "loss": 11.9585,
+      "step": 15185
+    },
+    {
+      "epoch": 2.6557377049180326,
+      "grad_norm": 0.9047580333507158,
+      "learning_rate": 6.824374581810456e-06,
+      "loss": 11.6897,
+      "step": 15186
+    },
+    {
+      "epoch": 2.655912568306011,
+      "grad_norm": 0.885974704459003,
+      "learning_rate": 6.817520883020301e-06,
+      "loss": 11.8853,
+      "step": 15187
+    },
+    {
+      "epoch": 2.6560874316939893,
+      "grad_norm": 0.8010151386874135,
+      "learning_rate": 6.810670506076655e-06,
+      "loss": 11.75,
+      "step": 15188
+    },
+    {
+      "epoch": 2.656262295081967,
+      "grad_norm": 0.8440156450767212,
+      "learning_rate": 6.803823451223712e-06,
+      "loss": 11.8285,
+      "step": 15189
+    },
+    {
+      "epoch": 2.656437158469945,
+      "grad_norm": 0.8003903280658123,
+      "learning_rate": 6.7969797187055765e-06,
+      "loss": 11.8429,
+      "step": 15190
+    },
+    {
+      "epoch": 2.6566120218579234,
+      "grad_norm": 0.8687515462033628,
+      "learning_rate": 6.790139308766219e-06,
+      "loss": 11.7164,
+      "step": 15191
+    },
+    {
+      "epoch": 2.656786885245902,
+      "grad_norm": 0.8451195368750652,
+      "learning_rate": 6.783302221649501e-06,
+      "loss": 11.7214,
+      "step": 15192
+    },
+    {
+      "epoch": 2.6569617486338797,
+      "grad_norm": 0.79261299826229,
+      "learning_rate": 6.776468457599139e-06,
+      "loss": 12.0704,
+      "step": 15193
+    },
+    {
+      "epoch": 2.657136612021858,
+      "grad_norm": 0.906847201354928,
+      "learning_rate": 6.7696380168587705e-06,
+      "loss": 11.7687,
+      "step": 15194
+    },
+    {
+      "epoch": 2.657311475409836,
+      "grad_norm": 0.8179377589743039,
+      "learning_rate": 6.762810899671856e-06,
+      "loss": 11.7827,
+      "step": 15195
+    },
+    {
+      "epoch": 2.6574863387978143,
+      "grad_norm": 0.8030768067355875,
+      "learning_rate": 6.755987106281814e-06,
+      "loss": 11.9759,
+      "step": 15196
+    },
+    {
+      "epoch": 2.6576612021857926,
+      "grad_norm": 0.8045645641753711,
+      "learning_rate": 6.7491666369318804e-06,
+      "loss": 11.9382,
+      "step": 15197
+    },
+    {
+      "epoch": 2.6578360655737705,
+      "grad_norm": 0.8947128218629635,
+      "learning_rate": 6.742349491865219e-06,
+      "loss": 11.7979,
+      "step": 15198
+    },
+    {
+      "epoch": 2.6580109289617484,
+      "grad_norm": 0.8023974398743741,
+      "learning_rate": 6.7355356713248215e-06,
+      "loss": 11.9622,
+      "step": 15199
+    },
+    {
+      "epoch": 2.6581857923497267,
+      "grad_norm": 0.8053583343007221,
+      "learning_rate": 6.728725175553607e-06,
+      "loss": 11.8047,
+      "step": 15200
+    },
+    {
+      "epoch": 2.658360655737705,
+      "grad_norm": 0.8222687488854795,
+      "learning_rate": 6.72191800479437e-06,
+      "loss": 11.9691,
+      "step": 15201
+    },
+    {
+      "epoch": 2.658535519125683,
+      "grad_norm": 0.7925996241849872,
+      "learning_rate": 6.7151141592897815e-06,
+      "loss": 12.0773,
+      "step": 15202
+    },
+    {
+      "epoch": 2.6587103825136613,
+      "grad_norm": 0.8281322332725111,
+      "learning_rate": 6.708313639282371e-06,
+      "loss": 11.7739,
+      "step": 15203
+    },
+    {
+      "epoch": 2.658885245901639,
+      "grad_norm": 0.8294342613236103,
+      "learning_rate": 6.701516445014577e-06,
+      "loss": 11.8886,
+      "step": 15204
+    },
+    {
+      "epoch": 2.6590601092896176,
+      "grad_norm": 0.7992347478058688,
+      "learning_rate": 6.694722576728718e-06,
+      "loss": 11.8451,
+      "step": 15205
+    },
+    {
+      "epoch": 2.659234972677596,
+      "grad_norm": 0.8562534898155612,
+      "learning_rate": 6.687932034666999e-06,
+      "loss": 11.9185,
+      "step": 15206
+    },
+    {
+      "epoch": 2.659409836065574,
+      "grad_norm": 0.8704000610567124,
+      "learning_rate": 6.681144819071461e-06,
+      "loss": 12.0616,
+      "step": 15207
+    },
+    {
+      "epoch": 2.6595846994535517,
+      "grad_norm": 0.8647446604671186,
+      "learning_rate": 6.674360930184076e-06,
+      "loss": 11.63,
+      "step": 15208
+    },
+    {
+      "epoch": 2.65975956284153,
+      "grad_norm": 0.8353579612658666,
+      "learning_rate": 6.667580368246695e-06,
+      "loss": 11.9021,
+      "step": 15209
+    },
+    {
+      "epoch": 2.6599344262295084,
+      "grad_norm": 0.8915664117975434,
+      "learning_rate": 6.660803133501015e-06,
+      "loss": 11.6486,
+      "step": 15210
+    },
+    {
+      "epoch": 2.6601092896174863,
+      "grad_norm": 0.8337635172928788,
+      "learning_rate": 6.654029226188651e-06,
+      "loss": 12.0616,
+      "step": 15211
+    },
+    {
+      "epoch": 2.6602841530054646,
+      "grad_norm": 0.8730281987692969,
+      "learning_rate": 6.6472586465510915e-06,
+      "loss": 11.7349,
+      "step": 15212
+    },
+    {
+      "epoch": 2.6604590163934425,
+      "grad_norm": 0.8956079895656999,
+      "learning_rate": 6.640491394829695e-06,
+      "loss": 12.0229,
+      "step": 15213
+    },
+    {
+      "epoch": 2.660633879781421,
+      "grad_norm": 0.9197956409714025,
+      "learning_rate": 6.633727471265694e-06,
+      "loss": 11.8425,
+      "step": 15214
+    },
+    {
+      "epoch": 2.6608087431693987,
+      "grad_norm": 0.8342321547684338,
+      "learning_rate": 6.626966876100216e-06,
+      "loss": 11.777,
+      "step": 15215
+    },
+    {
+      "epoch": 2.660983606557377,
+      "grad_norm": 0.808448263345983,
+      "learning_rate": 6.620209609574268e-06,
+      "loss": 11.8426,
+      "step": 15216
+    },
+    {
+      "epoch": 2.661158469945355,
+      "grad_norm": 0.8049116950594356,
+      "learning_rate": 6.61345567192877e-06,
+      "loss": 11.4916,
+      "step": 15217
+    },
+    {
+      "epoch": 2.6613333333333333,
+      "grad_norm": 0.8207400147878711,
+      "learning_rate": 6.606705063404451e-06,
+      "loss": 11.7177,
+      "step": 15218
+    },
+    {
+      "epoch": 2.6615081967213117,
+      "grad_norm": 0.8000182626280259,
+      "learning_rate": 6.599957784241972e-06,
+      "loss": 11.6144,
+      "step": 15219
+    },
+    {
+      "epoch": 2.6616830601092896,
+      "grad_norm": 0.8561660071342286,
+      "learning_rate": 6.5932138346818774e-06,
+      "loss": 11.7772,
+      "step": 15220
+    },
+    {
+      "epoch": 2.661857923497268,
+      "grad_norm": 0.7998312829804487,
+      "learning_rate": 6.586473214964594e-06,
+      "loss": 12.0734,
+      "step": 15221
+    },
+    {
+      "epoch": 2.662032786885246,
+      "grad_norm": 0.7756997599194078,
+      "learning_rate": 6.579735925330377e-06,
+      "loss": 11.8898,
+      "step": 15222
+    },
+    {
+      "epoch": 2.662207650273224,
+      "grad_norm": 0.871640452615283,
+      "learning_rate": 6.573001966019421e-06,
+      "loss": 11.8888,
+      "step": 15223
+    },
+    {
+      "epoch": 2.662382513661202,
+      "grad_norm": 0.8253845366868511,
+      "learning_rate": 6.566271337271812e-06,
+      "loss": 11.7219,
+      "step": 15224
+    },
+    {
+      "epoch": 2.6625573770491804,
+      "grad_norm": 0.8768120452039412,
+      "learning_rate": 6.5595440393274475e-06,
+      "loss": 11.7704,
+      "step": 15225
+    },
+    {
+      "epoch": 2.6627322404371583,
+      "grad_norm": 0.843616229038178,
+      "learning_rate": 6.552820072426158e-06,
+      "loss": 11.5496,
+      "step": 15226
+    },
+    {
+      "epoch": 2.6629071038251366,
+      "grad_norm": 0.8567394432325395,
+      "learning_rate": 6.546099436807651e-06,
+      "loss": 11.9912,
+      "step": 15227
+    },
+    {
+      "epoch": 2.663081967213115,
+      "grad_norm": 0.8930115645786408,
+      "learning_rate": 6.539382132711513e-06,
+      "loss": 11.8651,
+      "step": 15228
+    },
+    {
+      "epoch": 2.663256830601093,
+      "grad_norm": 0.7982876421633521,
+      "learning_rate": 6.532668160377209e-06,
+      "loss": 11.8427,
+      "step": 15229
+    },
+    {
+      "epoch": 2.663431693989071,
+      "grad_norm": 0.8430110467138959,
+      "learning_rate": 6.5259575200440795e-06,
+      "loss": 11.8447,
+      "step": 15230
+    },
+    {
+      "epoch": 2.663606557377049,
+      "grad_norm": 0.826274105958434,
+      "learning_rate": 6.5192502119513445e-06,
+      "loss": 12.1039,
+      "step": 15231
+    },
+    {
+      "epoch": 2.6637814207650274,
+      "grad_norm": 0.7803956490392661,
+      "learning_rate": 6.512546236338146e-06,
+      "loss": 11.8662,
+      "step": 15232
+    },
+    {
+      "epoch": 2.6639562841530053,
+      "grad_norm": 0.8313216842965627,
+      "learning_rate": 6.505845593443416e-06,
+      "loss": 12.0321,
+      "step": 15233
+    },
+    {
+      "epoch": 2.6641311475409837,
+      "grad_norm": 0.844932831848499,
+      "learning_rate": 6.4991482835060625e-06,
+      "loss": 11.7405,
+      "step": 15234
+    },
+    {
+      "epoch": 2.6643060109289616,
+      "grad_norm": 0.8006531136684001,
+      "learning_rate": 6.492454306764828e-06,
+      "loss": 11.8101,
+      "step": 15235
+    },
+    {
+      "epoch": 2.66448087431694,
+      "grad_norm": 0.8773652310187584,
+      "learning_rate": 6.485763663458367e-06,
+      "loss": 11.5693,
+      "step": 15236
+    },
+    {
+      "epoch": 2.6646557377049183,
+      "grad_norm": 0.8727142007274263,
+      "learning_rate": 6.479076353825153e-06,
+      "loss": 11.5861,
+      "step": 15237
+    },
+    {
+      "epoch": 2.664830601092896,
+      "grad_norm": 0.8200003144834953,
+      "learning_rate": 6.472392378103597e-06,
+      "loss": 11.7157,
+      "step": 15238
+    },
+    {
+      "epoch": 2.665005464480874,
+      "grad_norm": 0.8935393955354042,
+      "learning_rate": 6.4657117365319765e-06,
+      "loss": 11.8596,
+      "step": 15239
+    },
+    {
+      "epoch": 2.6651803278688524,
+      "grad_norm": 0.7921092832023569,
+      "learning_rate": 6.459034429348465e-06,
+      "loss": 11.996,
+      "step": 15240
+    },
+    {
+      "epoch": 2.6653551912568307,
+      "grad_norm": 0.8329516617352127,
+      "learning_rate": 6.452360456791074e-06,
+      "loss": 11.9466,
+      "step": 15241
+    },
+    {
+      "epoch": 2.6655300546448086,
+      "grad_norm": 0.8478459090096543,
+      "learning_rate": 6.445689819097723e-06,
+      "loss": 11.7462,
+      "step": 15242
+    },
+    {
+      "epoch": 2.665704918032787,
+      "grad_norm": 0.7986708697836761,
+      "learning_rate": 6.439022516506254e-06,
+      "loss": 11.6115,
+      "step": 15243
+    },
+    {
+      "epoch": 2.665879781420765,
+      "grad_norm": 0.9222627777733775,
+      "learning_rate": 6.432358549254292e-06,
+      "loss": 11.8974,
+      "step": 15244
+    },
+    {
+      "epoch": 2.666054644808743,
+      "grad_norm": 0.8762605372255746,
+      "learning_rate": 6.425697917579432e-06,
+      "loss": 11.7566,
+      "step": 15245
+    },
+    {
+      "epoch": 2.6662295081967216,
+      "grad_norm": 0.8854815151257008,
+      "learning_rate": 6.419040621719108e-06,
+      "loss": 12.1407,
+      "step": 15246
+    },
+    {
+      "epoch": 2.6664043715846995,
+      "grad_norm": 0.8340976850597696,
+      "learning_rate": 6.412386661910641e-06,
+      "loss": 12.0161,
+      "step": 15247
+    },
+    {
+      "epoch": 2.6665792349726773,
+      "grad_norm": 0.8421755154955448,
+      "learning_rate": 6.405736038391252e-06,
+      "loss": 11.6907,
+      "step": 15248
+    },
+    {
+      "epoch": 2.6667540983606557,
+      "grad_norm": 0.8249425831874783,
+      "learning_rate": 6.399088751398019e-06,
+      "loss": 11.8638,
+      "step": 15249
+    },
+    {
+      "epoch": 2.666928961748634,
+      "grad_norm": 0.7870753372337861,
+      "learning_rate": 6.3924448011679185e-06,
+      "loss": 12.0293,
+      "step": 15250
+    },
+    {
+      "epoch": 2.667103825136612,
+      "grad_norm": 0.8919370460112124,
+      "learning_rate": 6.385804187937794e-06,
+      "loss": 11.9845,
+      "step": 15251
+    },
+    {
+      "epoch": 2.6672786885245903,
+      "grad_norm": 0.7963415533274482,
+      "learning_rate": 6.379166911944368e-06,
+      "loss": 11.6011,
+      "step": 15252
+    },
+    {
+      "epoch": 2.667453551912568,
+      "grad_norm": 0.8318330864687213,
+      "learning_rate": 6.3725329734242615e-06,
+      "loss": 11.941,
+      "step": 15253
+    },
+    {
+      "epoch": 2.6676284153005465,
+      "grad_norm": 0.821429876539019,
+      "learning_rate": 6.365902372613952e-06,
+      "loss": 11.6171,
+      "step": 15254
+    },
+    {
+      "epoch": 2.667803278688525,
+      "grad_norm": 0.8189535474258643,
+      "learning_rate": 6.359275109749841e-06,
+      "loss": 11.9072,
+      "step": 15255
+    },
+    {
+      "epoch": 2.6679781420765027,
+      "grad_norm": 0.8678204585048129,
+      "learning_rate": 6.352651185068148e-06,
+      "loss": 11.4962,
+      "step": 15256
+    },
+    {
+      "epoch": 2.6681530054644806,
+      "grad_norm": 0.7972694352902638,
+      "learning_rate": 6.34603059880503e-06,
+      "loss": 11.8003,
+      "step": 15257
+    },
+    {
+      "epoch": 2.668327868852459,
+      "grad_norm": 0.7917776812854141,
+      "learning_rate": 6.3394133511965104e-06,
+      "loss": 11.5409,
+      "step": 15258
+    },
+    {
+      "epoch": 2.6685027322404373,
+      "grad_norm": 0.7998809634715439,
+      "learning_rate": 6.332799442478465e-06,
+      "loss": 11.6283,
+      "step": 15259
+    },
+    {
+      "epoch": 2.6686775956284152,
+      "grad_norm": 0.8213678034124117,
+      "learning_rate": 6.326188872886673e-06,
+      "loss": 11.9894,
+      "step": 15260
+    },
+    {
+      "epoch": 2.6688524590163936,
+      "grad_norm": 0.8051665190962437,
+      "learning_rate": 6.319581642656803e-06,
+      "loss": 11.6093,
+      "step": 15261
+    },
+    {
+      "epoch": 2.6690273224043715,
+      "grad_norm": 0.8233985949172756,
+      "learning_rate": 6.312977752024407e-06,
+      "loss": 11.9649,
+      "step": 15262
+    },
+    {
+      "epoch": 2.66920218579235,
+      "grad_norm": 0.8131248049441794,
+      "learning_rate": 6.306377201224889e-06,
+      "loss": 11.8924,
+      "step": 15263
+    },
+    {
+      "epoch": 2.669377049180328,
+      "grad_norm": 0.7929315880364693,
+      "learning_rate": 6.299779990493537e-06,
+      "loss": 11.8554,
+      "step": 15264
+    },
+    {
+      "epoch": 2.669551912568306,
+      "grad_norm": 0.7638658905767152,
+      "learning_rate": 6.293186120065564e-06,
+      "loss": 12.1649,
+      "step": 15265
+    },
+    {
+      "epoch": 2.669726775956284,
+      "grad_norm": 0.7704617921449008,
+      "learning_rate": 6.286595590176026e-06,
+      "loss": 11.8164,
+      "step": 15266
+    },
+    {
+      "epoch": 2.6699016393442623,
+      "grad_norm": 0.8194816205933502,
+      "learning_rate": 6.280008401059845e-06,
+      "loss": 11.8015,
+      "step": 15267
+    },
+    {
+      "epoch": 2.6700765027322406,
+      "grad_norm": 0.777016379952076,
+      "learning_rate": 6.273424552951879e-06,
+      "loss": 11.9401,
+      "step": 15268
+    },
+    {
+      "epoch": 2.6702513661202185,
+      "grad_norm": 0.8342337883038325,
+      "learning_rate": 6.266844046086817e-06,
+      "loss": 11.9171,
+      "step": 15269
+    },
+    {
+      "epoch": 2.670426229508197,
+      "grad_norm": 0.8147980134532113,
+      "learning_rate": 6.2602668806992614e-06,
+      "loss": 12.0036,
+      "step": 15270
+    },
+    {
+      "epoch": 2.6706010928961748,
+      "grad_norm": 0.8196472229900389,
+      "learning_rate": 6.253693057023657e-06,
+      "loss": 11.8493,
+      "step": 15271
+    },
+    {
+      "epoch": 2.670775956284153,
+      "grad_norm": 0.8057174518954181,
+      "learning_rate": 6.247122575294362e-06,
+      "loss": 11.9041,
+      "step": 15272
+    },
+    {
+      "epoch": 2.670950819672131,
+      "grad_norm": 0.805192406295088,
+      "learning_rate": 6.2405554357456205e-06,
+      "loss": 11.897,
+      "step": 15273
+    },
+    {
+      "epoch": 2.6711256830601093,
+      "grad_norm": 0.8200683515069894,
+      "learning_rate": 6.233991638611514e-06,
+      "loss": 11.7502,
+      "step": 15274
+    },
+    {
+      "epoch": 2.6713005464480872,
+      "grad_norm": 0.8770487707072543,
+      "learning_rate": 6.227431184126064e-06,
+      "loss": 11.9729,
+      "step": 15275
+    },
+    {
+      "epoch": 2.6714754098360656,
+      "grad_norm": 0.8226109752038467,
+      "learning_rate": 6.22087407252312e-06,
+      "loss": 11.8217,
+      "step": 15276
+    },
+    {
+      "epoch": 2.671650273224044,
+      "grad_norm": 0.8122789151405257,
+      "learning_rate": 6.21432030403647e-06,
+      "loss": 11.8784,
+      "step": 15277
+    },
+    {
+      "epoch": 2.671825136612022,
+      "grad_norm": 0.7945768314238918,
+      "learning_rate": 6.207769878899705e-06,
+      "loss": 11.8257,
+      "step": 15278
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 0.8318949525798685,
+      "learning_rate": 6.201222797346362e-06,
+      "loss": 11.9924,
+      "step": 15279
+    },
+    {
+      "epoch": 2.672174863387978,
+      "grad_norm": 0.7788700342779804,
+      "learning_rate": 6.194679059609831e-06,
+      "loss": 11.8251,
+      "step": 15280
+    },
+    {
+      "epoch": 2.6723497267759564,
+      "grad_norm": 0.8510238366871319,
+      "learning_rate": 6.188138665923382e-06,
+      "loss": 12.0463,
+      "step": 15281
+    },
+    {
+      "epoch": 2.6725245901639343,
+      "grad_norm": 0.8001665963772224,
+      "learning_rate": 6.181601616520194e-06,
+      "loss": 11.88,
+      "step": 15282
+    },
+    {
+      "epoch": 2.6726994535519126,
+      "grad_norm": 0.8222827794806891,
+      "learning_rate": 6.175067911633292e-06,
+      "loss": 11.8813,
+      "step": 15283
+    },
+    {
+      "epoch": 2.6728743169398905,
+      "grad_norm": 0.8084585310536422,
+      "learning_rate": 6.168537551495601e-06,
+      "loss": 11.5027,
+      "step": 15284
+    },
+    {
+      "epoch": 2.673049180327869,
+      "grad_norm": 0.8282796312594172,
+      "learning_rate": 6.162010536339924e-06,
+      "loss": 11.484,
+      "step": 15285
+    },
+    {
+      "epoch": 2.673224043715847,
+      "grad_norm": 0.8097087777306272,
+      "learning_rate": 6.155486866398919e-06,
+      "loss": 11.8722,
+      "step": 15286
+    },
+    {
+      "epoch": 2.673398907103825,
+      "grad_norm": 0.776515294560346,
+      "learning_rate": 6.148966541905165e-06,
+      "loss": 11.8923,
+      "step": 15287
+    },
+    {
+      "epoch": 2.673573770491803,
+      "grad_norm": 0.84139630877266,
+      "learning_rate": 6.142449563091113e-06,
+      "loss": 11.6443,
+      "step": 15288
+    },
+    {
+      "epoch": 2.6737486338797813,
+      "grad_norm": 0.8440837214236074,
+      "learning_rate": 6.1359359301890625e-06,
+      "loss": 11.9056,
+      "step": 15289
+    },
+    {
+      "epoch": 2.6739234972677597,
+      "grad_norm": 0.8271535787907234,
+      "learning_rate": 6.1294256434312195e-06,
+      "loss": 11.9495,
+      "step": 15290
+    },
+    {
+      "epoch": 2.6740983606557376,
+      "grad_norm": 0.8961773620217831,
+      "learning_rate": 6.1229187030496845e-06,
+      "loss": 12.0304,
+      "step": 15291
+    },
+    {
+      "epoch": 2.674273224043716,
+      "grad_norm": 0.8268860639602308,
+      "learning_rate": 6.116415109276419e-06,
+      "loss": 11.7602,
+      "step": 15292
+    },
+    {
+      "epoch": 2.674448087431694,
+      "grad_norm": 0.8143963950068661,
+      "learning_rate": 6.109914862343258e-06,
+      "loss": 12.0012,
+      "step": 15293
+    },
+    {
+      "epoch": 2.674622950819672,
+      "grad_norm": 0.8474180167089838,
+      "learning_rate": 6.1034179624819275e-06,
+      "loss": 11.6693,
+      "step": 15294
+    },
+    {
+      "epoch": 2.6747978142076505,
+      "grad_norm": 0.8920563581947136,
+      "learning_rate": 6.0969244099240434e-06,
+      "loss": 11.8594,
+      "step": 15295
+    },
+    {
+      "epoch": 2.6749726775956284,
+      "grad_norm": 0.842724474047965,
+      "learning_rate": 6.090434204901108e-06,
+      "loss": 11.902,
+      "step": 15296
+    },
+    {
+      "epoch": 2.6751475409836063,
+      "grad_norm": 0.7993680757206643,
+      "learning_rate": 6.0839473476444585e-06,
+      "loss": 11.5763,
+      "step": 15297
+    },
+    {
+      "epoch": 2.6753224043715846,
+      "grad_norm": 0.8176339758455264,
+      "learning_rate": 6.077463838385344e-06,
+      "loss": 11.8748,
+      "step": 15298
+    },
+    {
+      "epoch": 2.675497267759563,
+      "grad_norm": 0.8473730328207222,
+      "learning_rate": 6.0709836773549245e-06,
+      "loss": 11.6753,
+      "step": 15299
+    },
+    {
+      "epoch": 2.675672131147541,
+      "grad_norm": 0.8009122667313054,
+      "learning_rate": 6.064506864784181e-06,
+      "loss": 11.7081,
+      "step": 15300
+    },
+    {
+      "epoch": 2.675846994535519,
+      "grad_norm": 0.8144527579750002,
+      "learning_rate": 6.058033400904017e-06,
+      "loss": 11.6935,
+      "step": 15301
+    },
+    {
+      "epoch": 2.676021857923497,
+      "grad_norm": 0.8515558296074611,
+      "learning_rate": 6.051563285945205e-06,
+      "loss": 11.7631,
+      "step": 15302
+    },
+    {
+      "epoch": 2.6761967213114755,
+      "grad_norm": 0.825776200628687,
+      "learning_rate": 6.045096520138404e-06,
+      "loss": 12.0678,
+      "step": 15303
+    },
+    {
+      "epoch": 2.676371584699454,
+      "grad_norm": 0.8738980944581015,
+      "learning_rate": 6.0386331037141305e-06,
+      "loss": 12.0493,
+      "step": 15304
+    },
+    {
+      "epoch": 2.6765464480874317,
+      "grad_norm": 0.7961687944453885,
+      "learning_rate": 6.0321730369027995e-06,
+      "loss": 11.7858,
+      "step": 15305
+    },
+    {
+      "epoch": 2.6767213114754096,
+      "grad_norm": 0.8040755426162154,
+      "learning_rate": 6.0257163199347154e-06,
+      "loss": 11.8167,
+      "step": 15306
+    },
+    {
+      "epoch": 2.676896174863388,
+      "grad_norm": 0.9614291450564706,
+      "learning_rate": 6.019262953040061e-06,
+      "loss": 12.0248,
+      "step": 15307
+    },
+    {
+      "epoch": 2.6770710382513663,
+      "grad_norm": 0.8337776924383841,
+      "learning_rate": 6.012812936448853e-06,
+      "loss": 11.9382,
+      "step": 15308
+    },
+    {
+      "epoch": 2.677245901639344,
+      "grad_norm": 0.8096825059661125,
+      "learning_rate": 6.006366270391062e-06,
+      "loss": 12.0631,
+      "step": 15309
+    },
+    {
+      "epoch": 2.6774207650273225,
+      "grad_norm": 0.8418262552922267,
+      "learning_rate": 5.9999229550964934e-06,
+      "loss": 11.637,
+      "step": 15310
+    },
+    {
+      "epoch": 2.6775956284153004,
+      "grad_norm": 0.8356735304772643,
+      "learning_rate": 5.993482990794852e-06,
+      "loss": 11.6931,
+      "step": 15311
+    },
+    {
+      "epoch": 2.6777704918032788,
+      "grad_norm": 0.9255284470251551,
+      "learning_rate": 5.987046377715699e-06,
+      "loss": 11.4684,
+      "step": 15312
+    },
+    {
+      "epoch": 2.677945355191257,
+      "grad_norm": 1.020765783915535,
+      "learning_rate": 5.980613116088496e-06,
+      "loss": 11.9562,
+      "step": 15313
+    },
+    {
+      "epoch": 2.678120218579235,
+      "grad_norm": 0.7941910380967743,
+      "learning_rate": 5.974183206142592e-06,
+      "loss": 11.967,
+      "step": 15314
+    },
+    {
+      "epoch": 2.678295081967213,
+      "grad_norm": 0.8229594177059424,
+      "learning_rate": 5.967756648107214e-06,
+      "loss": 11.8928,
+      "step": 15315
+    },
+    {
+      "epoch": 2.6784699453551912,
+      "grad_norm": 0.9186416549081764,
+      "learning_rate": 5.961333442211425e-06,
+      "loss": 11.9305,
+      "step": 15316
+    },
+    {
+      "epoch": 2.6786448087431696,
+      "grad_norm": 0.8727016278086329,
+      "learning_rate": 5.95491358868423e-06,
+      "loss": 11.8002,
+      "step": 15317
+    },
+    {
+      "epoch": 2.6788196721311475,
+      "grad_norm": 0.8673499235650216,
+      "learning_rate": 5.9484970877544785e-06,
+      "loss": 11.8939,
+      "step": 15318
+    },
+    {
+      "epoch": 2.678994535519126,
+      "grad_norm": 0.8513622572068483,
+      "learning_rate": 5.9420839396509335e-06,
+      "loss": 11.6799,
+      "step": 15319
+    },
+    {
+      "epoch": 2.6791693989071037,
+      "grad_norm": 0.817882366234724,
+      "learning_rate": 5.935674144602188e-06,
+      "loss": 11.931,
+      "step": 15320
+    },
+    {
+      "epoch": 2.679344262295082,
+      "grad_norm": 0.8397463854247245,
+      "learning_rate": 5.92926770283676e-06,
+      "loss": 12.0032,
+      "step": 15321
+    },
+    {
+      "epoch": 2.67951912568306,
+      "grad_norm": 0.7853766464703033,
+      "learning_rate": 5.922864614583046e-06,
+      "loss": 12.0796,
+      "step": 15322
+    },
+    {
+      "epoch": 2.6796939890710383,
+      "grad_norm": 0.8408120809635332,
+      "learning_rate": 5.9164648800692726e-06,
+      "loss": 11.9292,
+      "step": 15323
+    },
+    {
+      "epoch": 2.679868852459016,
+      "grad_norm": 0.8728007172561071,
+      "learning_rate": 5.9100684995236135e-06,
+      "loss": 12.0756,
+      "step": 15324
+    },
+    {
+      "epoch": 2.6800437158469945,
+      "grad_norm": 0.7813144170257678,
+      "learning_rate": 5.903675473174064e-06,
+      "loss": 11.9138,
+      "step": 15325
+    },
+    {
+      "epoch": 2.680218579234973,
+      "grad_norm": 0.8035050568300852,
+      "learning_rate": 5.897285801248564e-06,
+      "loss": 11.8134,
+      "step": 15326
+    },
+    {
+      "epoch": 2.6803934426229508,
+      "grad_norm": 0.8942222587463393,
+      "learning_rate": 5.890899483974865e-06,
+      "loss": 11.9086,
+      "step": 15327
+    },
+    {
+      "epoch": 2.680568306010929,
+      "grad_norm": 0.8329580726028263,
+      "learning_rate": 5.8845165215806515e-06,
+      "loss": 11.8487,
+      "step": 15328
+    },
+    {
+      "epoch": 2.680743169398907,
+      "grad_norm": 0.8797166906391657,
+      "learning_rate": 5.878136914293452e-06,
+      "loss": 11.8843,
+      "step": 15329
+    },
+    {
+      "epoch": 2.6809180327868853,
+      "grad_norm": 0.8129340372376536,
+      "learning_rate": 5.87176066234072e-06,
+      "loss": 11.9013,
+      "step": 15330
+    },
+    {
+      "epoch": 2.6810928961748632,
+      "grad_norm": 0.8671513931946832,
+      "learning_rate": 5.865387765949726e-06,
+      "loss": 11.6714,
+      "step": 15331
+    },
+    {
+      "epoch": 2.6812677595628416,
+      "grad_norm": 0.8290379830399542,
+      "learning_rate": 5.859018225347679e-06,
+      "loss": 11.9816,
+      "step": 15332
+    },
+    {
+      "epoch": 2.6814426229508195,
+      "grad_norm": 0.817414127706377,
+      "learning_rate": 5.8526520407616524e-06,
+      "loss": 12.0388,
+      "step": 15333
+    },
+    {
+      "epoch": 2.681617486338798,
+      "grad_norm": 0.8474095252395626,
+      "learning_rate": 5.846289212418565e-06,
+      "loss": 11.9272,
+      "step": 15334
+    },
+    {
+      "epoch": 2.681792349726776,
+      "grad_norm": 0.8883041769688429,
+      "learning_rate": 5.839929740545258e-06,
+      "loss": 11.6818,
+      "step": 15335
+    },
+    {
+      "epoch": 2.681967213114754,
+      "grad_norm": 0.8152546363543386,
+      "learning_rate": 5.833573625368449e-06,
+      "loss": 11.725,
+      "step": 15336
+    },
+    {
+      "epoch": 2.6821420765027324,
+      "grad_norm": 0.8614318578652594,
+      "learning_rate": 5.827220867114713e-06,
+      "loss": 11.7814,
+      "step": 15337
+    },
+    {
+      "epoch": 2.6823169398907103,
+      "grad_norm": 0.822998473325969,
+      "learning_rate": 5.8208714660105245e-06,
+      "loss": 12.0995,
+      "step": 15338
+    },
+    {
+      "epoch": 2.6824918032786886,
+      "grad_norm": 0.8673563069597154,
+      "learning_rate": 5.814525422282236e-06,
+      "loss": 11.5939,
+      "step": 15339
+    },
+    {
+      "epoch": 2.6826666666666665,
+      "grad_norm": 0.8402062991637315,
+      "learning_rate": 5.808182736156065e-06,
+      "loss": 11.96,
+      "step": 15340
+    },
+    {
+      "epoch": 2.682841530054645,
+      "grad_norm": 0.7544969269805986,
+      "learning_rate": 5.801843407858143e-06,
+      "loss": 12.0496,
+      "step": 15341
+    },
+    {
+      "epoch": 2.6830163934426228,
+      "grad_norm": 0.9048307865814933,
+      "learning_rate": 5.795507437614434e-06,
+      "loss": 11.8074,
+      "step": 15342
+    },
+    {
+      "epoch": 2.683191256830601,
+      "grad_norm": 0.7812656333917154,
+      "learning_rate": 5.789174825650823e-06,
+      "loss": 11.8968,
+      "step": 15343
+    },
+    {
+      "epoch": 2.6833661202185795,
+      "grad_norm": 0.8405663176972334,
+      "learning_rate": 5.782845572193052e-06,
+      "loss": 11.7607,
+      "step": 15344
+    },
+    {
+      "epoch": 2.6835409836065574,
+      "grad_norm": 0.8035547322068625,
+      "learning_rate": 5.776519677466763e-06,
+      "loss": 11.9236,
+      "step": 15345
+    },
+    {
+      "epoch": 2.6837158469945352,
+      "grad_norm": 0.8119665574039193,
+      "learning_rate": 5.770197141697453e-06,
+      "loss": 11.7473,
+      "step": 15346
+    },
+    {
+      "epoch": 2.6838907103825136,
+      "grad_norm": 0.8969201052985206,
+      "learning_rate": 5.7638779651105204e-06,
+      "loss": 11.6649,
+      "step": 15347
+    },
+    {
+      "epoch": 2.684065573770492,
+      "grad_norm": 0.8141373620615797,
+      "learning_rate": 5.757562147931228e-06,
+      "loss": 11.8352,
+      "step": 15348
+    },
+    {
+      "epoch": 2.68424043715847,
+      "grad_norm": 0.8242379860939469,
+      "learning_rate": 5.7512496903847526e-06,
+      "loss": 12.0157,
+      "step": 15349
+    },
+    {
+      "epoch": 2.684415300546448,
+      "grad_norm": 0.8109902863989671,
+      "learning_rate": 5.744940592696102e-06,
+      "loss": 11.9449,
+      "step": 15350
+    },
+    {
+      "epoch": 2.684590163934426,
+      "grad_norm": 0.8277092333942895,
+      "learning_rate": 5.738634855090186e-06,
+      "loss": 12.0112,
+      "step": 15351
+    },
+    {
+      "epoch": 2.6847650273224044,
+      "grad_norm": 0.8404319552008976,
+      "learning_rate": 5.732332477791813e-06,
+      "loss": 11.9049,
+      "step": 15352
+    },
+    {
+      "epoch": 2.6849398907103827,
+      "grad_norm": 0.8331168714041703,
+      "learning_rate": 5.726033461025637e-06,
+      "loss": 11.6806,
+      "step": 15353
+    },
+    {
+      "epoch": 2.6851147540983606,
+      "grad_norm": 0.7992283081592727,
+      "learning_rate": 5.7197378050162235e-06,
+      "loss": 11.8196,
+      "step": 15354
+    },
+    {
+      "epoch": 2.6852896174863385,
+      "grad_norm": 0.7650262066882353,
+      "learning_rate": 5.713445509988002e-06,
+      "loss": 11.8284,
+      "step": 15355
+    },
+    {
+      "epoch": 2.685464480874317,
+      "grad_norm": 0.8496733701461858,
+      "learning_rate": 5.707156576165284e-06,
+      "loss": 11.5953,
+      "step": 15356
+    },
+    {
+      "epoch": 2.6856393442622952,
+      "grad_norm": 0.7528932369066619,
+      "learning_rate": 5.700871003772268e-06,
+      "loss": 11.8345,
+      "step": 15357
+    },
+    {
+      "epoch": 2.685814207650273,
+      "grad_norm": 0.7948180107596798,
+      "learning_rate": 5.694588793033029e-06,
+      "loss": 11.7737,
+      "step": 15358
+    },
+    {
+      "epoch": 2.6859890710382515,
+      "grad_norm": 0.8138968862798335,
+      "learning_rate": 5.6883099441715105e-06,
+      "loss": 11.7236,
+      "step": 15359
+    },
+    {
+      "epoch": 2.6861639344262294,
+      "grad_norm": 0.837552534358638,
+      "learning_rate": 5.682034457411567e-06,
+      "loss": 11.6132,
+      "step": 15360
+    },
+    {
+      "epoch": 2.6863387978142077,
+      "grad_norm": 0.8566136884614433,
+      "learning_rate": 5.675762332976875e-06,
+      "loss": 11.4674,
+      "step": 15361
+    },
+    {
+      "epoch": 2.686513661202186,
+      "grad_norm": 0.9907661073832925,
+      "learning_rate": 5.669493571091055e-06,
+      "loss": 11.8297,
+      "step": 15362
+    },
+    {
+      "epoch": 2.686688524590164,
+      "grad_norm": 0.8425192648012403,
+      "learning_rate": 5.663228171977575e-06,
+      "loss": 11.7899,
+      "step": 15363
+    },
+    {
+      "epoch": 2.686863387978142,
+      "grad_norm": 0.8160218276125376,
+      "learning_rate": 5.6569661358597985e-06,
+      "loss": 11.6659,
+      "step": 15364
+    },
+    {
+      "epoch": 2.68703825136612,
+      "grad_norm": 0.8630873278340655,
+      "learning_rate": 5.650707462960947e-06,
+      "loss": 11.8982,
+      "step": 15365
+    },
+    {
+      "epoch": 2.6872131147540985,
+      "grad_norm": 0.8536492067956462,
+      "learning_rate": 5.6444521535041314e-06,
+      "loss": 11.9052,
+      "step": 15366
+    },
+    {
+      "epoch": 2.6873879781420764,
+      "grad_norm": 0.8764173038363605,
+      "learning_rate": 5.638200207712374e-06,
+      "loss": 11.9345,
+      "step": 15367
+    },
+    {
+      "epoch": 2.6875628415300548,
+      "grad_norm": 0.8524426034552443,
+      "learning_rate": 5.631951625808518e-06,
+      "loss": 11.8839,
+      "step": 15368
+    },
+    {
+      "epoch": 2.6877377049180327,
+      "grad_norm": 0.832518377213954,
+      "learning_rate": 5.625706408015319e-06,
+      "loss": 11.9749,
+      "step": 15369
+    },
+    {
+      "epoch": 2.687912568306011,
+      "grad_norm": 0.8931580302522469,
+      "learning_rate": 5.619464554555432e-06,
+      "loss": 11.7834,
+      "step": 15370
+    },
+    {
+      "epoch": 2.6880874316939893,
+      "grad_norm": 0.8514260682270414,
+      "learning_rate": 5.6132260656513665e-06,
+      "loss": 11.9775,
+      "step": 15371
+    },
+    {
+      "epoch": 2.6882622950819672,
+      "grad_norm": 0.7910261158738412,
+      "learning_rate": 5.6069909415255025e-06,
+      "loss": 11.9947,
+      "step": 15372
+    },
+    {
+      "epoch": 2.688437158469945,
+      "grad_norm": 0.840691681911538,
+      "learning_rate": 5.600759182400128e-06,
+      "loss": 11.9531,
+      "step": 15373
+    },
+    {
+      "epoch": 2.6886120218579235,
+      "grad_norm": 0.8221840335544224,
+      "learning_rate": 5.594530788497388e-06,
+      "loss": 11.995,
+      "step": 15374
+    },
+    {
+      "epoch": 2.688786885245902,
+      "grad_norm": 0.7860244429900329,
+      "learning_rate": 5.588305760039336e-06,
+      "loss": 12.0313,
+      "step": 15375
+    },
+    {
+      "epoch": 2.6889617486338797,
+      "grad_norm": 0.8524751857705619,
+      "learning_rate": 5.5820840972478636e-06,
+      "loss": 11.724,
+      "step": 15376
+    },
+    {
+      "epoch": 2.689136612021858,
+      "grad_norm": 0.8004184559370836,
+      "learning_rate": 5.575865800344782e-06,
+      "loss": 11.8713,
+      "step": 15377
+    },
+    {
+      "epoch": 2.689311475409836,
+      "grad_norm": 0.8549364967378225,
+      "learning_rate": 5.569650869551757e-06,
+      "loss": 11.6814,
+      "step": 15378
+    },
+    {
+      "epoch": 2.6894863387978143,
+      "grad_norm": 0.7859038958925114,
+      "learning_rate": 5.563439305090368e-06,
+      "loss": 11.9083,
+      "step": 15379
+    },
+    {
+      "epoch": 2.689661202185792,
+      "grad_norm": 0.8875835576606461,
+      "learning_rate": 5.557231107182026e-06,
+      "loss": 11.5893,
+      "step": 15380
+    },
+    {
+      "epoch": 2.6898360655737705,
+      "grad_norm": 0.8539198590331467,
+      "learning_rate": 5.551026276048033e-06,
+      "loss": 11.9456,
+      "step": 15381
+    },
+    {
+      "epoch": 2.6900109289617484,
+      "grad_norm": 0.9302892437352475,
+      "learning_rate": 5.544824811909621e-06,
+      "loss": 11.9523,
+      "step": 15382
+    },
+    {
+      "epoch": 2.6901857923497268,
+      "grad_norm": 0.7940774123893523,
+      "learning_rate": 5.538626714987838e-06,
+      "loss": 11.902,
+      "step": 15383
+    },
+    {
+      "epoch": 2.690360655737705,
+      "grad_norm": 0.8363076955308061,
+      "learning_rate": 5.532431985503639e-06,
+      "loss": 11.9014,
+      "step": 15384
+    },
+    {
+      "epoch": 2.690535519125683,
+      "grad_norm": 0.8240806562981574,
+      "learning_rate": 5.52624062367787e-06,
+      "loss": 11.726,
+      "step": 15385
+    },
+    {
+      "epoch": 2.6907103825136613,
+      "grad_norm": 0.9048740608569935,
+      "learning_rate": 5.520052629731254e-06,
+      "loss": 11.5762,
+      "step": 15386
+    },
+    {
+      "epoch": 2.6908852459016392,
+      "grad_norm": 0.8270017613231346,
+      "learning_rate": 5.51386800388437e-06,
+      "loss": 11.8252,
+      "step": 15387
+    },
+    {
+      "epoch": 2.6910601092896176,
+      "grad_norm": 0.8180092108783293,
+      "learning_rate": 5.507686746357687e-06,
+      "loss": 11.5501,
+      "step": 15388
+    },
+    {
+      "epoch": 2.6912349726775955,
+      "grad_norm": 0.8424074723426634,
+      "learning_rate": 5.501508857371562e-06,
+      "loss": 11.9291,
+      "step": 15389
+    },
+    {
+      "epoch": 2.691409836065574,
+      "grad_norm": 0.9938246989819763,
+      "learning_rate": 5.495334337146252e-06,
+      "loss": 11.7696,
+      "step": 15390
+    },
+    {
+      "epoch": 2.6915846994535517,
+      "grad_norm": 0.7917388698885299,
+      "learning_rate": 5.489163185901847e-06,
+      "loss": 11.9652,
+      "step": 15391
+    },
+    {
+      "epoch": 2.69175956284153,
+      "grad_norm": 0.8443768095094542,
+      "learning_rate": 5.482995403858349e-06,
+      "loss": 11.9206,
+      "step": 15392
+    },
+    {
+      "epoch": 2.6919344262295084,
+      "grad_norm": 0.8248851909195105,
+      "learning_rate": 5.476830991235638e-06,
+      "loss": 11.8185,
+      "step": 15393
+    },
+    {
+      "epoch": 2.6921092896174863,
+      "grad_norm": 0.7972397717861361,
+      "learning_rate": 5.470669948253471e-06,
+      "loss": 11.7448,
+      "step": 15394
+    },
+    {
+      "epoch": 2.6922841530054646,
+      "grad_norm": 0.8987596601742476,
+      "learning_rate": 5.464512275131473e-06,
+      "loss": 11.7174,
+      "step": 15395
+    },
+    {
+      "epoch": 2.6924590163934425,
+      "grad_norm": 0.7985176559303324,
+      "learning_rate": 5.458357972089146e-06,
+      "loss": 11.6586,
+      "step": 15396
+    },
+    {
+      "epoch": 2.692633879781421,
+      "grad_norm": 0.7880892015191694,
+      "learning_rate": 5.452207039345914e-06,
+      "loss": 12.1502,
+      "step": 15397
+    },
+    {
+      "epoch": 2.6928087431693988,
+      "grad_norm": 0.8215101376490352,
+      "learning_rate": 5.446059477121024e-06,
+      "loss": 11.8641,
+      "step": 15398
+    },
+    {
+      "epoch": 2.692983606557377,
+      "grad_norm": 0.8745447865712958,
+      "learning_rate": 5.4399152856336455e-06,
+      "loss": 11.9871,
+      "step": 15399
+    },
+    {
+      "epoch": 2.693158469945355,
+      "grad_norm": 0.8172665454820262,
+      "learning_rate": 5.433774465102792e-06,
+      "loss": 11.9767,
+      "step": 15400
+    },
+    {
+      "epoch": 2.6933333333333334,
+      "grad_norm": 0.8046609725757573,
+      "learning_rate": 5.427637015747411e-06,
+      "loss": 12.0393,
+      "step": 15401
+    },
+    {
+      "epoch": 2.6935081967213117,
+      "grad_norm": 0.7349960608189972,
+      "learning_rate": 5.4215029377862605e-06,
+      "loss": 11.4442,
+      "step": 15402
+    },
+    {
+      "epoch": 2.6936830601092896,
+      "grad_norm": 0.7993356276255512,
+      "learning_rate": 5.415372231438032e-06,
+      "loss": 12.1412,
+      "step": 15403
+    },
+    {
+      "epoch": 2.6938579234972675,
+      "grad_norm": 0.8689667404228189,
+      "learning_rate": 5.409244896921262e-06,
+      "loss": 11.5913,
+      "step": 15404
+    },
+    {
+      "epoch": 2.694032786885246,
+      "grad_norm": 0.8041659736912505,
+      "learning_rate": 5.4031209344544195e-06,
+      "loss": 11.9613,
+      "step": 15405
+    },
+    {
+      "epoch": 2.694207650273224,
+      "grad_norm": 0.9857642290598091,
+      "learning_rate": 5.397000344255765e-06,
+      "loss": 11.7246,
+      "step": 15406
+    },
+    {
+      "epoch": 2.694382513661202,
+      "grad_norm": 0.904426392398236,
+      "learning_rate": 5.390883126543522e-06,
+      "loss": 11.8906,
+      "step": 15407
+    },
+    {
+      "epoch": 2.6945573770491804,
+      "grad_norm": 0.8438610310688544,
+      "learning_rate": 5.384769281535762e-06,
+      "loss": 12.0473,
+      "step": 15408
+    },
+    {
+      "epoch": 2.6947322404371583,
+      "grad_norm": 0.8231760789972646,
+      "learning_rate": 5.378658809450421e-06,
+      "loss": 11.8176,
+      "step": 15409
+    },
+    {
+      "epoch": 2.6949071038251367,
+      "grad_norm": 0.8645904866471448,
+      "learning_rate": 5.372551710505347e-06,
+      "loss": 11.9239,
+      "step": 15410
+    },
+    {
+      "epoch": 2.695081967213115,
+      "grad_norm": 0.8236386807737772,
+      "learning_rate": 5.366447984918243e-06,
+      "loss": 11.8054,
+      "step": 15411
+    },
+    {
+      "epoch": 2.695256830601093,
+      "grad_norm": 0.7740550434898931,
+      "learning_rate": 5.360347632906715e-06,
+      "loss": 12.0026,
+      "step": 15412
+    },
+    {
+      "epoch": 2.695431693989071,
+      "grad_norm": 0.8166464838156825,
+      "learning_rate": 5.354250654688209e-06,
+      "loss": 11.8448,
+      "step": 15413
+    },
+    {
+      "epoch": 2.695606557377049,
+      "grad_norm": 0.8800372584479599,
+      "learning_rate": 5.3481570504800845e-06,
+      "loss": 11.81,
+      "step": 15414
+    },
+    {
+      "epoch": 2.6957814207650275,
+      "grad_norm": 0.8827682627203676,
+      "learning_rate": 5.3420668204995694e-06,
+      "loss": 11.8059,
+      "step": 15415
+    },
+    {
+      "epoch": 2.6959562841530054,
+      "grad_norm": 0.8744363950363064,
+      "learning_rate": 5.335979964963789e-06,
+      "loss": 11.7997,
+      "step": 15416
+    },
+    {
+      "epoch": 2.6961311475409837,
+      "grad_norm": 0.8565275260068598,
+      "learning_rate": 5.3298964840897024e-06,
+      "loss": 11.9306,
+      "step": 15417
+    },
+    {
+      "epoch": 2.6963060109289616,
+      "grad_norm": 0.8765727287733697,
+      "learning_rate": 5.323816378094204e-06,
+      "loss": 11.8945,
+      "step": 15418
+    },
+    {
+      "epoch": 2.69648087431694,
+      "grad_norm": 0.9318773877755219,
+      "learning_rate": 5.317739647194031e-06,
+      "loss": 12.0412,
+      "step": 15419
+    },
+    {
+      "epoch": 2.6966557377049183,
+      "grad_norm": 0.823721666892233,
+      "learning_rate": 5.311666291605833e-06,
+      "loss": 11.8732,
+      "step": 15420
+    },
+    {
+      "epoch": 2.696830601092896,
+      "grad_norm": 0.8094894617333493,
+      "learning_rate": 5.305596311546079e-06,
+      "loss": 12.0089,
+      "step": 15421
+    },
+    {
+      "epoch": 2.697005464480874,
+      "grad_norm": 0.8499035390371645,
+      "learning_rate": 5.299529707231177e-06,
+      "loss": 11.7748,
+      "step": 15422
+    },
+    {
+      "epoch": 2.6971803278688524,
+      "grad_norm": 0.8676299884620221,
+      "learning_rate": 5.293466478877407e-06,
+      "loss": 11.9663,
+      "step": 15423
+    },
+    {
+      "epoch": 2.6973551912568308,
+      "grad_norm": 0.8118798732508995,
+      "learning_rate": 5.287406626700908e-06,
+      "loss": 11.8067,
+      "step": 15424
+    },
+    {
+      "epoch": 2.6975300546448087,
+      "grad_norm": 0.7971640424515628,
+      "learning_rate": 5.281350150917696e-06,
+      "loss": 11.5878,
+      "step": 15425
+    },
+    {
+      "epoch": 2.697704918032787,
+      "grad_norm": 0.7796157235357704,
+      "learning_rate": 5.275297051743678e-06,
+      "loss": 11.9971,
+      "step": 15426
+    },
+    {
+      "epoch": 2.697879781420765,
+      "grad_norm": 0.7691940083964467,
+      "learning_rate": 5.269247329394644e-06,
+      "loss": 11.6098,
+      "step": 15427
+    },
+    {
+      "epoch": 2.6980546448087432,
+      "grad_norm": 0.8694717129535168,
+      "learning_rate": 5.2632009840862695e-06,
+      "loss": 11.9287,
+      "step": 15428
+    },
+    {
+      "epoch": 2.6982295081967216,
+      "grad_norm": 0.79550193348652,
+      "learning_rate": 5.257158016034092e-06,
+      "loss": 11.9271,
+      "step": 15429
+    },
+    {
+      "epoch": 2.6984043715846995,
+      "grad_norm": 0.8788029449489799,
+      "learning_rate": 5.2511184254535275e-06,
+      "loss": 11.8561,
+      "step": 15430
+    },
+    {
+      "epoch": 2.6985792349726774,
+      "grad_norm": 0.8630056590758124,
+      "learning_rate": 5.245082212559915e-06,
+      "loss": 11.756,
+      "step": 15431
+    },
+    {
+      "epoch": 2.6987540983606557,
+      "grad_norm": 0.871001669296069,
+      "learning_rate": 5.239049377568384e-06,
+      "loss": 11.7812,
+      "step": 15432
+    },
+    {
+      "epoch": 2.698928961748634,
+      "grad_norm": 0.8445348189798667,
+      "learning_rate": 5.233019920694038e-06,
+      "loss": 11.5331,
+      "step": 15433
+    },
+    {
+      "epoch": 2.699103825136612,
+      "grad_norm": 0.8765897464819714,
+      "learning_rate": 5.2269938421518065e-06,
+      "loss": 11.7175,
+      "step": 15434
+    },
+    {
+      "epoch": 2.6992786885245903,
+      "grad_norm": 0.7985089431608888,
+      "learning_rate": 5.220971142156528e-06,
+      "loss": 11.9019,
+      "step": 15435
+    },
+    {
+      "epoch": 2.699453551912568,
+      "grad_norm": 0.8493713401271799,
+      "learning_rate": 5.214951820922876e-06,
+      "loss": 11.9587,
+      "step": 15436
+    },
+    {
+      "epoch": 2.6996284153005465,
+      "grad_norm": 0.8060450147297293,
+      "learning_rate": 5.2089358786654445e-06,
+      "loss": 11.8803,
+      "step": 15437
+    },
+    {
+      "epoch": 2.6998032786885244,
+      "grad_norm": 1.0013951684902846,
+      "learning_rate": 5.202923315598695e-06,
+      "loss": 11.9842,
+      "step": 15438
+    },
+    {
+      "epoch": 2.6999781420765028,
+      "grad_norm": 0.8704677299327649,
+      "learning_rate": 5.196914131936992e-06,
+      "loss": 12.1112,
+      "step": 15439
+    },
+    {
+      "epoch": 2.7001530054644807,
+      "grad_norm": 0.824173293449467,
+      "learning_rate": 5.190908327894517e-06,
+      "loss": 12.0045,
+      "step": 15440
+    },
+    {
+      "epoch": 2.700327868852459,
+      "grad_norm": 0.7810756939914115,
+      "learning_rate": 5.184905903685389e-06,
+      "loss": 12.0009,
+      "step": 15441
+    },
+    {
+      "epoch": 2.7005027322404374,
+      "grad_norm": 0.8139209142164705,
+      "learning_rate": 5.178906859523602e-06,
+      "loss": 12.1327,
+      "step": 15442
+    },
+    {
+      "epoch": 2.7006775956284153,
+      "grad_norm": 0.833981801416105,
+      "learning_rate": 5.172911195622976e-06,
+      "loss": 11.7993,
+      "step": 15443
+    },
+    {
+      "epoch": 2.7008524590163936,
+      "grad_norm": 0.8410898265998095,
+      "learning_rate": 5.166918912197283e-06,
+      "loss": 12.0678,
+      "step": 15444
+    },
+    {
+      "epoch": 2.7010273224043715,
+      "grad_norm": 0.7883540629190843,
+      "learning_rate": 5.16093000946013e-06,
+      "loss": 11.9853,
+      "step": 15445
+    },
+    {
+      "epoch": 2.70120218579235,
+      "grad_norm": 0.8439478529253435,
+      "learning_rate": 5.1549444876250016e-06,
+      "loss": 11.9781,
+      "step": 15446
+    },
+    {
+      "epoch": 2.7013770491803277,
+      "grad_norm": 0.8253095003229886,
+      "learning_rate": 5.148962346905306e-06,
+      "loss": 11.7779,
+      "step": 15447
+    },
+    {
+      "epoch": 2.701551912568306,
+      "grad_norm": 0.7715750980413824,
+      "learning_rate": 5.142983587514271e-06,
+      "loss": 12.0747,
+      "step": 15448
+    },
+    {
+      "epoch": 2.701726775956284,
+      "grad_norm": 0.9354230444197351,
+      "learning_rate": 5.13700820966504e-06,
+      "loss": 12.03,
+      "step": 15449
+    },
+    {
+      "epoch": 2.7019016393442623,
+      "grad_norm": 0.8386419758363534,
+      "learning_rate": 5.13103621357065e-06,
+      "loss": 12.0439,
+      "step": 15450
+    },
+    {
+      "epoch": 2.7020765027322406,
+      "grad_norm": 0.9099720684070316,
+      "learning_rate": 5.1250675994439555e-06,
+      "loss": 11.8875,
+      "step": 15451
+    },
+    {
+      "epoch": 2.7022513661202185,
+      "grad_norm": 0.8668451772631796,
+      "learning_rate": 5.119102367497764e-06,
+      "loss": 11.8752,
+      "step": 15452
+    },
+    {
+      "epoch": 2.702426229508197,
+      "grad_norm": 0.8363600528077757,
+      "learning_rate": 5.113140517944703e-06,
+      "loss": 11.5737,
+      "step": 15453
+    },
+    {
+      "epoch": 2.702601092896175,
+      "grad_norm": 0.8698451849064767,
+      "learning_rate": 5.107182050997339e-06,
+      "loss": 11.624,
+      "step": 15454
+    },
+    {
+      "epoch": 2.702775956284153,
+      "grad_norm": 0.7915404694884947,
+      "learning_rate": 5.101226966868056e-06,
+      "loss": 11.7436,
+      "step": 15455
+    },
+    {
+      "epoch": 2.702950819672131,
+      "grad_norm": 0.8326634459427416,
+      "learning_rate": 5.0952752657691505e-06,
+      "loss": 11.8626,
+      "step": 15456
+    },
+    {
+      "epoch": 2.7031256830601094,
+      "grad_norm": 0.7980993576870864,
+      "learning_rate": 5.0893269479127866e-06,
+      "loss": 11.8292,
+      "step": 15457
+    },
+    {
+      "epoch": 2.7033005464480873,
+      "grad_norm": 0.8252381554741285,
+      "learning_rate": 5.083382013511051e-06,
+      "loss": 11.7731,
+      "step": 15458
+    },
+    {
+      "epoch": 2.7034754098360656,
+      "grad_norm": 0.917564984953018,
+      "learning_rate": 5.07744046277584e-06,
+      "loss": 11.7525,
+      "step": 15459
+    },
+    {
+      "epoch": 2.703650273224044,
+      "grad_norm": 0.8555385281305353,
+      "learning_rate": 5.071502295918962e-06,
+      "loss": 11.8634,
+      "step": 15460
+    },
+    {
+      "epoch": 2.703825136612022,
+      "grad_norm": 0.8237116826481986,
+      "learning_rate": 5.065567513152136e-06,
+      "loss": 12.0969,
+      "step": 15461
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 0.8744023721568821,
+      "learning_rate": 5.059636114686883e-06,
+      "loss": 11.8269,
+      "step": 15462
+    },
+    {
+      "epoch": 2.704174863387978,
+      "grad_norm": 0.7919472581317389,
+      "learning_rate": 5.053708100734689e-06,
+      "loss": 11.7499,
+      "step": 15463
+    },
+    {
+      "epoch": 2.7043497267759564,
+      "grad_norm": 0.835847628229974,
+      "learning_rate": 5.047783471506862e-06,
+      "loss": 11.9986,
+      "step": 15464
+    },
+    {
+      "epoch": 2.7045245901639343,
+      "grad_norm": 0.8433411255508475,
+      "learning_rate": 5.041862227214611e-06,
+      "loss": 11.9807,
+      "step": 15465
+    },
+    {
+      "epoch": 2.7046994535519127,
+      "grad_norm": 0.8060181549315139,
+      "learning_rate": 5.035944368069034e-06,
+      "loss": 11.981,
+      "step": 15466
+    },
+    {
+      "epoch": 2.7048743169398906,
+      "grad_norm": 0.8021605374251419,
+      "learning_rate": 5.030029894281086e-06,
+      "loss": 11.7695,
+      "step": 15467
+    },
+    {
+      "epoch": 2.705049180327869,
+      "grad_norm": 0.8933619573491177,
+      "learning_rate": 5.024118806061606e-06,
+      "loss": 11.7582,
+      "step": 15468
+    },
+    {
+      "epoch": 2.7052240437158472,
+      "grad_norm": 0.8135869904942155,
+      "learning_rate": 5.01821110362134e-06,
+      "loss": 11.7893,
+      "step": 15469
+    },
+    {
+      "epoch": 2.705398907103825,
+      "grad_norm": 0.8027969811931798,
+      "learning_rate": 5.012306787170861e-06,
+      "loss": 11.8221,
+      "step": 15470
+    },
+    {
+      "epoch": 2.705573770491803,
+      "grad_norm": 0.765925276245392,
+      "learning_rate": 5.00640585692066e-06,
+      "loss": 12.0486,
+      "step": 15471
+    },
+    {
+      "epoch": 2.7057486338797814,
+      "grad_norm": 0.7925362200069741,
+      "learning_rate": 5.000508313081109e-06,
+      "loss": 11.5099,
+      "step": 15472
+    },
+    {
+      "epoch": 2.7059234972677597,
+      "grad_norm": 0.9883398220236108,
+      "learning_rate": 4.994614155862442e-06,
+      "loss": 11.5173,
+      "step": 15473
+    },
+    {
+      "epoch": 2.7060983606557376,
+      "grad_norm": 0.7604576014699029,
+      "learning_rate": 4.98872338547478e-06,
+      "loss": 12.022,
+      "step": 15474
+    },
+    {
+      "epoch": 2.706273224043716,
+      "grad_norm": 0.8296498007959637,
+      "learning_rate": 4.98283600212811e-06,
+      "loss": 11.9314,
+      "step": 15475
+    },
+    {
+      "epoch": 2.706448087431694,
+      "grad_norm": 0.8721610160008808,
+      "learning_rate": 4.976952006032343e-06,
+      "loss": 11.9079,
+      "step": 15476
+    },
+    {
+      "epoch": 2.706622950819672,
+      "grad_norm": 0.809026580506253,
+      "learning_rate": 4.9710713973971975e-06,
+      "loss": 11.5975,
+      "step": 15477
+    },
+    {
+      "epoch": 2.7067978142076505,
+      "grad_norm": 0.8746476647628354,
+      "learning_rate": 4.96519417643233e-06,
+      "loss": 11.862,
+      "step": 15478
+    },
+    {
+      "epoch": 2.7069726775956284,
+      "grad_norm": 0.8298382342149091,
+      "learning_rate": 4.95932034334724e-06,
+      "loss": 11.6445,
+      "step": 15479
+    },
+    {
+      "epoch": 2.7071475409836063,
+      "grad_norm": 0.8747038723046291,
+      "learning_rate": 4.953449898351348e-06,
+      "loss": 11.9135,
+      "step": 15480
+    },
+    {
+      "epoch": 2.7073224043715847,
+      "grad_norm": 0.8263461620333833,
+      "learning_rate": 4.947582841653919e-06,
+      "loss": 11.869,
+      "step": 15481
+    },
+    {
+      "epoch": 2.707497267759563,
+      "grad_norm": 0.8438801277038245,
+      "learning_rate": 4.941719173464099e-06,
+      "loss": 11.8946,
+      "step": 15482
+    },
+    {
+      "epoch": 2.707672131147541,
+      "grad_norm": 0.8106600007808635,
+      "learning_rate": 4.93585889399093e-06,
+      "loss": 11.6792,
+      "step": 15483
+    },
+    {
+      "epoch": 2.7078469945355192,
+      "grad_norm": 0.8195920096582513,
+      "learning_rate": 4.9300020034433345e-06,
+      "loss": 11.8523,
+      "step": 15484
+    },
+    {
+      "epoch": 2.708021857923497,
+      "grad_norm": 0.7662382083656873,
+      "learning_rate": 4.924148502030079e-06,
+      "loss": 11.8868,
+      "step": 15485
+    },
+    {
+      "epoch": 2.7081967213114755,
+      "grad_norm": 0.8518980840320032,
+      "learning_rate": 4.918298389959841e-06,
+      "loss": 11.6363,
+      "step": 15486
+    },
+    {
+      "epoch": 2.708371584699454,
+      "grad_norm": 0.8404130272998386,
+      "learning_rate": 4.9124516674411755e-06,
+      "loss": 11.8076,
+      "step": 15487
+    },
+    {
+      "epoch": 2.7085464480874317,
+      "grad_norm": 0.8053374110808176,
+      "learning_rate": 4.906608334682528e-06,
+      "loss": 11.9669,
+      "step": 15488
+    },
+    {
+      "epoch": 2.7087213114754096,
+      "grad_norm": 0.8218491987024482,
+      "learning_rate": 4.900768391892175e-06,
+      "loss": 11.8537,
+      "step": 15489
+    },
+    {
+      "epoch": 2.708896174863388,
+      "grad_norm": 0.8299949604365656,
+      "learning_rate": 4.894931839278305e-06,
+      "loss": 11.6542,
+      "step": 15490
+    },
+    {
+      "epoch": 2.7090710382513663,
+      "grad_norm": 0.7752301314312261,
+      "learning_rate": 4.88909867704902e-06,
+      "loss": 12.0856,
+      "step": 15491
+    },
+    {
+      "epoch": 2.709245901639344,
+      "grad_norm": 0.8592429614839512,
+      "learning_rate": 4.88326890541223e-06,
+      "loss": 11.6819,
+      "step": 15492
+    },
+    {
+      "epoch": 2.7094207650273225,
+      "grad_norm": 0.9060132287369547,
+      "learning_rate": 4.87744252457577e-06,
+      "loss": 11.4921,
+      "step": 15493
+    },
+    {
+      "epoch": 2.7095956284153004,
+      "grad_norm": 0.8449518832160849,
+      "learning_rate": 4.8716195347473406e-06,
+      "loss": 11.6225,
+      "step": 15494
+    },
+    {
+      "epoch": 2.709770491803279,
+      "grad_norm": 0.8149263724461346,
+      "learning_rate": 4.865799936134552e-06,
+      "loss": 11.5283,
+      "step": 15495
+    },
+    {
+      "epoch": 2.7099453551912567,
+      "grad_norm": 0.8486756288910914,
+      "learning_rate": 4.859983728944817e-06,
+      "loss": 11.702,
+      "step": 15496
+    },
+    {
+      "epoch": 2.710120218579235,
+      "grad_norm": 0.8461772292540499,
+      "learning_rate": 4.854170913385514e-06,
+      "loss": 11.7612,
+      "step": 15497
+    },
+    {
+      "epoch": 2.710295081967213,
+      "grad_norm": 0.8594625644942304,
+      "learning_rate": 4.848361489663844e-06,
+      "loss": 11.9312,
+      "step": 15498
+    },
+    {
+      "epoch": 2.7104699453551913,
+      "grad_norm": 0.8520551824079509,
+      "learning_rate": 4.842555457986919e-06,
+      "loss": 11.9244,
+      "step": 15499
+    },
+    {
+      "epoch": 2.7106448087431696,
+      "grad_norm": 0.7793749923439319,
+      "learning_rate": 4.836752818561718e-06,
+      "loss": 11.8748,
+      "step": 15500
+    },
+    {
+      "epoch": 2.7108196721311475,
+      "grad_norm": 0.7719942603490426,
+      "learning_rate": 4.830953571595076e-06,
+      "loss": 11.9345,
+      "step": 15501
+    },
+    {
+      "epoch": 2.710994535519126,
+      "grad_norm": 0.7859547945220572,
+      "learning_rate": 4.82515771729376e-06,
+      "loss": 11.7081,
+      "step": 15502
+    },
+    {
+      "epoch": 2.7111693989071037,
+      "grad_norm": 0.8489403001862866,
+      "learning_rate": 4.819365255864383e-06,
+      "loss": 11.6986,
+      "step": 15503
+    },
+    {
+      "epoch": 2.711344262295082,
+      "grad_norm": 0.8528367505145581,
+      "learning_rate": 4.813576187513413e-06,
+      "loss": 11.7994,
+      "step": 15504
+    },
+    {
+      "epoch": 2.71151912568306,
+      "grad_norm": 0.8749415817088766,
+      "learning_rate": 4.807790512447241e-06,
+      "loss": 11.7357,
+      "step": 15505
+    },
+    {
+      "epoch": 2.7116939890710383,
+      "grad_norm": 0.8471230540582106,
+      "learning_rate": 4.802008230872124e-06,
+      "loss": 11.9152,
+      "step": 15506
+    },
+    {
+      "epoch": 2.711868852459016,
+      "grad_norm": 0.8718982592463321,
+      "learning_rate": 4.796229342994174e-06,
+      "loss": 12.0217,
+      "step": 15507
+    },
+    {
+      "epoch": 2.7120437158469946,
+      "grad_norm": 0.7924667566421383,
+      "learning_rate": 4.7904538490194164e-06,
+      "loss": 12.0967,
+      "step": 15508
+    },
+    {
+      "epoch": 2.712218579234973,
+      "grad_norm": 0.9031493110516594,
+      "learning_rate": 4.784681749153741e-06,
+      "loss": 11.9657,
+      "step": 15509
+    },
+    {
+      "epoch": 2.712393442622951,
+      "grad_norm": 0.8174194846437145,
+      "learning_rate": 4.778913043602928e-06,
+      "loss": 11.7463,
+      "step": 15510
+    },
+    {
+      "epoch": 2.7125683060109287,
+      "grad_norm": 0.8402870260087947,
+      "learning_rate": 4.773147732572592e-06,
+      "loss": 11.7014,
+      "step": 15511
+    },
+    {
+      "epoch": 2.712743169398907,
+      "grad_norm": 0.8931468290679045,
+      "learning_rate": 4.767385816268288e-06,
+      "loss": 12.0027,
+      "step": 15512
+    },
+    {
+      "epoch": 2.7129180327868854,
+      "grad_norm": 0.8303612876346294,
+      "learning_rate": 4.761627294895399e-06,
+      "loss": 11.9383,
+      "step": 15513
+    },
+    {
+      "epoch": 2.7130928961748633,
+      "grad_norm": 0.847224179790549,
+      "learning_rate": 4.755872168659248e-06,
+      "loss": 11.9867,
+      "step": 15514
+    },
+    {
+      "epoch": 2.7132677595628416,
+      "grad_norm": 0.9409817019055869,
+      "learning_rate": 4.75012043776496e-06,
+      "loss": 11.9497,
+      "step": 15515
+    },
+    {
+      "epoch": 2.7134426229508195,
+      "grad_norm": 0.8580517409140896,
+      "learning_rate": 4.744372102417583e-06,
+      "loss": 11.9853,
+      "step": 15516
+    },
+    {
+      "epoch": 2.713617486338798,
+      "grad_norm": 0.8562464952933087,
+      "learning_rate": 4.738627162822051e-06,
+      "loss": 11.8867,
+      "step": 15517
+    },
+    {
+      "epoch": 2.713792349726776,
+      "grad_norm": 0.8221279630691966,
+      "learning_rate": 4.732885619183159e-06,
+      "loss": 12.0048,
+      "step": 15518
+    },
+    {
+      "epoch": 2.713967213114754,
+      "grad_norm": 0.8061730371606154,
+      "learning_rate": 4.727147471705584e-06,
+      "loss": 11.7042,
+      "step": 15519
+    },
+    {
+      "epoch": 2.714142076502732,
+      "grad_norm": 0.8126213211574411,
+      "learning_rate": 4.721412720593888e-06,
+      "loss": 12.0493,
+      "step": 15520
+    },
+    {
+      "epoch": 2.7143169398907103,
+      "grad_norm": 0.8147617784266533,
+      "learning_rate": 4.715681366052527e-06,
+      "loss": 12.0772,
+      "step": 15521
+    },
+    {
+      "epoch": 2.7144918032786887,
+      "grad_norm": 0.8167797715711088,
+      "learning_rate": 4.709953408285783e-06,
+      "loss": 11.8902,
+      "step": 15522
+    },
+    {
+      "epoch": 2.7146666666666666,
+      "grad_norm": 0.8529593754300329,
+      "learning_rate": 4.7042288474978605e-06,
+      "loss": 11.8308,
+      "step": 15523
+    },
+    {
+      "epoch": 2.714841530054645,
+      "grad_norm": 0.8599947570681419,
+      "learning_rate": 4.698507683892839e-06,
+      "loss": 11.5957,
+      "step": 15524
+    },
+    {
+      "epoch": 2.715016393442623,
+      "grad_norm": 0.8388161589225266,
+      "learning_rate": 4.69278991767469e-06,
+      "loss": 12.0674,
+      "step": 15525
+    },
+    {
+      "epoch": 2.715191256830601,
+      "grad_norm": 0.7712835456610052,
+      "learning_rate": 4.687075549047204e-06,
+      "loss": 12.0119,
+      "step": 15526
+    },
+    {
+      "epoch": 2.7153661202185795,
+      "grad_norm": 0.8304409702148787,
+      "learning_rate": 4.68136457821412e-06,
+      "loss": 11.9571,
+      "step": 15527
+    },
+    {
+      "epoch": 2.7155409836065574,
+      "grad_norm": 0.813731300817505,
+      "learning_rate": 4.675657005379019e-06,
+      "loss": 11.8614,
+      "step": 15528
+    },
+    {
+      "epoch": 2.7157158469945353,
+      "grad_norm": 0.8336817325077568,
+      "learning_rate": 4.6699528307453834e-06,
+      "loss": 12.0162,
+      "step": 15529
+    },
+    {
+      "epoch": 2.7158907103825136,
+      "grad_norm": 0.8663614408457976,
+      "learning_rate": 4.664252054516527e-06,
+      "loss": 11.9014,
+      "step": 15530
+    },
+    {
+      "epoch": 2.716065573770492,
+      "grad_norm": 0.8644118960196778,
+      "learning_rate": 4.65855467689571e-06,
+      "loss": 11.7022,
+      "step": 15531
+    },
+    {
+      "epoch": 2.71624043715847,
+      "grad_norm": 0.8881236769608819,
+      "learning_rate": 4.652860698086015e-06,
+      "loss": 11.8428,
+      "step": 15532
+    },
+    {
+      "epoch": 2.716415300546448,
+      "grad_norm": 0.828729611537401,
+      "learning_rate": 4.647170118290456e-06,
+      "loss": 11.981,
+      "step": 15533
+    },
+    {
+      "epoch": 2.716590163934426,
+      "grad_norm": 0.8269050186984221,
+      "learning_rate": 4.641482937711861e-06,
+      "loss": 11.5858,
+      "step": 15534
+    },
+    {
+      "epoch": 2.7167650273224044,
+      "grad_norm": 0.8129870066794233,
+      "learning_rate": 4.635799156552978e-06,
+      "loss": 11.4899,
+      "step": 15535
+    },
+    {
+      "epoch": 2.7169398907103828,
+      "grad_norm": 0.8096951302610333,
+      "learning_rate": 4.630118775016434e-06,
+      "loss": 11.9524,
+      "step": 15536
+    },
+    {
+      "epoch": 2.7171147540983607,
+      "grad_norm": 0.8778671293101106,
+      "learning_rate": 4.624441793304735e-06,
+      "loss": 11.9141,
+      "step": 15537
+    },
+    {
+      "epoch": 2.7172896174863386,
+      "grad_norm": 0.8330034061770211,
+      "learning_rate": 4.618768211620251e-06,
+      "loss": 12.0789,
+      "step": 15538
+    },
+    {
+      "epoch": 2.717464480874317,
+      "grad_norm": 0.8274874892511639,
+      "learning_rate": 4.6130980301652415e-06,
+      "loss": 11.8597,
+      "step": 15539
+    },
+    {
+      "epoch": 2.7176393442622953,
+      "grad_norm": 0.8046441603523584,
+      "learning_rate": 4.607431249141847e-06,
+      "loss": 11.6972,
+      "step": 15540
+    },
+    {
+      "epoch": 2.717814207650273,
+      "grad_norm": 0.8204867536915023,
+      "learning_rate": 4.601767868752061e-06,
+      "loss": 11.8756,
+      "step": 15541
+    },
+    {
+      "epoch": 2.7179890710382515,
+      "grad_norm": 0.8523018587087109,
+      "learning_rate": 4.596107889197787e-06,
+      "loss": 12.0039,
+      "step": 15542
+    },
+    {
+      "epoch": 2.7181639344262294,
+      "grad_norm": 0.8658718017975509,
+      "learning_rate": 4.590451310680799e-06,
+      "loss": 11.664,
+      "step": 15543
+    },
+    {
+      "epoch": 2.7183387978142077,
+      "grad_norm": 0.7593591060820408,
+      "learning_rate": 4.584798133402768e-06,
+      "loss": 12.0221,
+      "step": 15544
+    },
+    {
+      "epoch": 2.7185136612021856,
+      "grad_norm": 0.7993170069531406,
+      "learning_rate": 4.5791483575651886e-06,
+      "loss": 11.9693,
+      "step": 15545
+    },
+    {
+      "epoch": 2.718688524590164,
+      "grad_norm": 0.8670957128765495,
+      "learning_rate": 4.573501983369477e-06,
+      "loss": 11.8635,
+      "step": 15546
+    },
+    {
+      "epoch": 2.718863387978142,
+      "grad_norm": 0.9014397428766192,
+      "learning_rate": 4.567859011016928e-06,
+      "loss": 11.8222,
+      "step": 15547
+    },
+    {
+      "epoch": 2.71903825136612,
+      "grad_norm": 0.8303467635906939,
+      "learning_rate": 4.562219440708726e-06,
+      "loss": 12.0423,
+      "step": 15548
+    },
+    {
+      "epoch": 2.7192131147540985,
+      "grad_norm": 0.7989294613414023,
+      "learning_rate": 4.556583272645875e-06,
+      "loss": 11.9106,
+      "step": 15549
+    },
+    {
+      "epoch": 2.7193879781420764,
+      "grad_norm": 0.8635895245242065,
+      "learning_rate": 4.550950507029317e-06,
+      "loss": 11.9814,
+      "step": 15550
+    },
+    {
+      "epoch": 2.719562841530055,
+      "grad_norm": 0.899196894361761,
+      "learning_rate": 4.5453211440598666e-06,
+      "loss": 11.8895,
+      "step": 15551
+    },
+    {
+      "epoch": 2.7197377049180327,
+      "grad_norm": 0.8095947052819437,
+      "learning_rate": 4.539695183938175e-06,
+      "loss": 11.9068,
+      "step": 15552
+    },
+    {
+      "epoch": 2.719912568306011,
+      "grad_norm": 0.7899626923199893,
+      "learning_rate": 4.534072626864816e-06,
+      "loss": 11.8219,
+      "step": 15553
+    },
+    {
+      "epoch": 2.720087431693989,
+      "grad_norm": 0.8562234599149052,
+      "learning_rate": 4.528453473040228e-06,
+      "loss": 11.7141,
+      "step": 15554
+    },
+    {
+      "epoch": 2.7202622950819673,
+      "grad_norm": 0.7475345893735468,
+      "learning_rate": 4.52283772266473e-06,
+      "loss": 11.8313,
+      "step": 15555
+    },
+    {
+      "epoch": 2.720437158469945,
+      "grad_norm": 0.8117433953452347,
+      "learning_rate": 4.517225375938517e-06,
+      "loss": 11.9037,
+      "step": 15556
+    },
+    {
+      "epoch": 2.7206120218579235,
+      "grad_norm": 0.8017601162706471,
+      "learning_rate": 4.5116164330616495e-06,
+      "loss": 11.684,
+      "step": 15557
+    },
+    {
+      "epoch": 2.720786885245902,
+      "grad_norm": 0.8240130822440017,
+      "learning_rate": 4.506010894234092e-06,
+      "loss": 11.5883,
+      "step": 15558
+    },
+    {
+      "epoch": 2.7209617486338797,
+      "grad_norm": 0.8825685729165531,
+      "learning_rate": 4.500408759655694e-06,
+      "loss": 11.9798,
+      "step": 15559
+    },
+    {
+      "epoch": 2.721136612021858,
+      "grad_norm": 0.8477701684605984,
+      "learning_rate": 4.494810029526131e-06,
+      "loss": 11.7888,
+      "step": 15560
+    },
+    {
+      "epoch": 2.721311475409836,
+      "grad_norm": 0.8594074121969685,
+      "learning_rate": 4.489214704044997e-06,
+      "loss": 11.8293,
+      "step": 15561
+    },
+    {
+      "epoch": 2.7214863387978143,
+      "grad_norm": 0.8086549584956774,
+      "learning_rate": 4.483622783411778e-06,
+      "loss": 11.8885,
+      "step": 15562
+    },
+    {
+      "epoch": 2.721661202185792,
+      "grad_norm": 0.778264508032574,
+      "learning_rate": 4.478034267825815e-06,
+      "loss": 11.8977,
+      "step": 15563
+    },
+    {
+      "epoch": 2.7218360655737706,
+      "grad_norm": 0.7913404546019513,
+      "learning_rate": 4.472449157486303e-06,
+      "loss": 12.1549,
+      "step": 15564
+    },
+    {
+      "epoch": 2.7220109289617485,
+      "grad_norm": 0.7700570611731851,
+      "learning_rate": 4.466867452592371e-06,
+      "loss": 11.7694,
+      "step": 15565
+    },
+    {
+      "epoch": 2.722185792349727,
+      "grad_norm": 0.8492571420904582,
+      "learning_rate": 4.461289153343007e-06,
+      "loss": 11.5757,
+      "step": 15566
+    },
+    {
+      "epoch": 2.722360655737705,
+      "grad_norm": 0.8612174542545178,
+      "learning_rate": 4.45571425993706e-06,
+      "loss": 12.0188,
+      "step": 15567
+    },
+    {
+      "epoch": 2.722535519125683,
+      "grad_norm": 0.7847593261387269,
+      "learning_rate": 4.450142772573262e-06,
+      "loss": 11.8241,
+      "step": 15568
+    },
+    {
+      "epoch": 2.722710382513661,
+      "grad_norm": 0.8625597844515662,
+      "learning_rate": 4.4445746914502314e-06,
+      "loss": 11.6577,
+      "step": 15569
+    },
+    {
+      "epoch": 2.7228852459016393,
+      "grad_norm": 0.8072570381029687,
+      "learning_rate": 4.439010016766476e-06,
+      "loss": 11.8166,
+      "step": 15570
+    },
+    {
+      "epoch": 2.7230601092896176,
+      "grad_norm": 0.8055398314867863,
+      "learning_rate": 4.433448748720359e-06,
+      "loss": 11.9687,
+      "step": 15571
+    },
+    {
+      "epoch": 2.7232349726775955,
+      "grad_norm": 0.8442868013952921,
+      "learning_rate": 4.427890887510133e-06,
+      "loss": 11.8057,
+      "step": 15572
+    },
+    {
+      "epoch": 2.723409836065574,
+      "grad_norm": 0.8271871107639015,
+      "learning_rate": 4.422336433333929e-06,
+      "loss": 11.9358,
+      "step": 15573
+    },
+    {
+      "epoch": 2.7235846994535517,
+      "grad_norm": 0.8369230280076985,
+      "learning_rate": 4.4167853863897876e-06,
+      "loss": 11.9741,
+      "step": 15574
+    },
+    {
+      "epoch": 2.72375956284153,
+      "grad_norm": 0.7990890368498895,
+      "learning_rate": 4.411237746875551e-06,
+      "loss": 12.138,
+      "step": 15575
+    },
+    {
+      "epoch": 2.7239344262295084,
+      "grad_norm": 0.8443300063437323,
+      "learning_rate": 4.405693514988996e-06,
+      "loss": 11.896,
+      "step": 15576
+    },
+    {
+      "epoch": 2.7241092896174863,
+      "grad_norm": 0.7549569112387026,
+      "learning_rate": 4.400152690927783e-06,
+      "loss": 11.7611,
+      "step": 15577
+    },
+    {
+      "epoch": 2.7242841530054642,
+      "grad_norm": 0.8265596436069791,
+      "learning_rate": 4.394615274889447e-06,
+      "loss": 12.055,
+      "step": 15578
+    },
+    {
+      "epoch": 2.7244590163934426,
+      "grad_norm": 0.8474385965744606,
+      "learning_rate": 4.38908126707136e-06,
+      "loss": 11.8143,
+      "step": 15579
+    },
+    {
+      "epoch": 2.724633879781421,
+      "grad_norm": 0.8133154763035744,
+      "learning_rate": 4.383550667670822e-06,
+      "loss": 11.8614,
+      "step": 15580
+    },
+    {
+      "epoch": 2.724808743169399,
+      "grad_norm": 0.8578394128706024,
+      "learning_rate": 4.3780234768849735e-06,
+      "loss": 11.8209,
+      "step": 15581
+    },
+    {
+      "epoch": 2.724983606557377,
+      "grad_norm": 0.8268200025765893,
+      "learning_rate": 4.372499694910881e-06,
+      "loss": 11.827,
+      "step": 15582
+    },
+    {
+      "epoch": 2.725158469945355,
+      "grad_norm": 0.8532757290142686,
+      "learning_rate": 4.366979321945441e-06,
+      "loss": 11.9023,
+      "step": 15583
+    },
+    {
+      "epoch": 2.7253333333333334,
+      "grad_norm": 0.7799777423809791,
+      "learning_rate": 4.361462358185453e-06,
+      "loss": 12.0203,
+      "step": 15584
+    },
+    {
+      "epoch": 2.7255081967213117,
+      "grad_norm": 0.790660168970853,
+      "learning_rate": 4.355948803827592e-06,
+      "loss": 11.6801,
+      "step": 15585
+    },
+    {
+      "epoch": 2.7256830601092896,
+      "grad_norm": 0.8345335785233778,
+      "learning_rate": 4.350438659068401e-06,
+      "loss": 12.1409,
+      "step": 15586
+    },
+    {
+      "epoch": 2.7258579234972675,
+      "grad_norm": 0.8628156204897328,
+      "learning_rate": 4.344931924104323e-06,
+      "loss": 11.7872,
+      "step": 15587
+    },
+    {
+      "epoch": 2.726032786885246,
+      "grad_norm": 0.8957969412507015,
+      "learning_rate": 4.3394285991316564e-06,
+      "loss": 11.6568,
+      "step": 15588
+    },
+    {
+      "epoch": 2.726207650273224,
+      "grad_norm": 0.8132456802200546,
+      "learning_rate": 4.3339286843465885e-06,
+      "loss": 11.6978,
+      "step": 15589
+    },
+    {
+      "epoch": 2.726382513661202,
+      "grad_norm": 0.8505117362241357,
+      "learning_rate": 4.328432179945196e-06,
+      "loss": 11.7895,
+      "step": 15590
+    },
+    {
+      "epoch": 2.7265573770491804,
+      "grad_norm": 0.8089730204013819,
+      "learning_rate": 4.322939086123412e-06,
+      "loss": 11.8919,
+      "step": 15591
+    },
+    {
+      "epoch": 2.7267322404371583,
+      "grad_norm": 0.7516622276104103,
+      "learning_rate": 4.317449403077067e-06,
+      "loss": 11.8775,
+      "step": 15592
+    },
+    {
+      "epoch": 2.7269071038251367,
+      "grad_norm": 0.8827637075469638,
+      "learning_rate": 4.3119631310018725e-06,
+      "loss": 12.0684,
+      "step": 15593
+    },
+    {
+      "epoch": 2.727081967213115,
+      "grad_norm": 0.8845895342271379,
+      "learning_rate": 4.306480270093372e-06,
+      "loss": 11.9488,
+      "step": 15594
+    },
+    {
+      "epoch": 2.727256830601093,
+      "grad_norm": 0.7884412756378576,
+      "learning_rate": 4.301000820547052e-06,
+      "loss": 11.7459,
+      "step": 15595
+    },
+    {
+      "epoch": 2.727431693989071,
+      "grad_norm": 0.8062735290533877,
+      "learning_rate": 4.2955247825582355e-06,
+      "loss": 11.8038,
+      "step": 15596
+    },
+    {
+      "epoch": 2.727606557377049,
+      "grad_norm": 0.8174485270649882,
+      "learning_rate": 4.290052156322155e-06,
+      "loss": 11.8726,
+      "step": 15597
+    },
+    {
+      "epoch": 2.7277814207650275,
+      "grad_norm": 0.780225244202816,
+      "learning_rate": 4.284582942033877e-06,
+      "loss": 11.8231,
+      "step": 15598
+    },
+    {
+      "epoch": 2.7279562841530054,
+      "grad_norm": 0.8596895134196779,
+      "learning_rate": 4.279117139888389e-06,
+      "loss": 11.7581,
+      "step": 15599
+    },
+    {
+      "epoch": 2.7281311475409837,
+      "grad_norm": 0.8520852924747505,
+      "learning_rate": 4.273654750080547e-06,
+      "loss": 11.9803,
+      "step": 15600
+    },
+    {
+      "epoch": 2.7283060109289616,
+      "grad_norm": 0.7972270197526126,
+      "learning_rate": 4.26819577280505e-06,
+      "loss": 11.89,
+      "step": 15601
+    },
+    {
+      "epoch": 2.72848087431694,
+      "grad_norm": 0.868753100125759,
+      "learning_rate": 4.2627402082565324e-06,
+      "loss": 11.967,
+      "step": 15602
+    },
+    {
+      "epoch": 2.728655737704918,
+      "grad_norm": 0.8363514399113409,
+      "learning_rate": 4.2572880566294604e-06,
+      "loss": 11.8067,
+      "step": 15603
+    },
+    {
+      "epoch": 2.728830601092896,
+      "grad_norm": 0.805655303800289,
+      "learning_rate": 4.251839318118233e-06,
+      "loss": 11.9919,
+      "step": 15604
+    },
+    {
+      "epoch": 2.729005464480874,
+      "grad_norm": 0.8353235540321112,
+      "learning_rate": 4.246393992917042e-06,
+      "loss": 11.9537,
+      "step": 15605
+    },
+    {
+      "epoch": 2.7291803278688525,
+      "grad_norm": 0.8711911485177831,
+      "learning_rate": 4.2409520812200285e-06,
+      "loss": 11.518,
+      "step": 15606
+    },
+    {
+      "epoch": 2.729355191256831,
+      "grad_norm": 0.836419120171018,
+      "learning_rate": 4.235513583221195e-06,
+      "loss": 11.9444,
+      "step": 15607
+    },
+    {
+      "epoch": 2.7295300546448087,
+      "grad_norm": 0.8108260667009777,
+      "learning_rate": 4.230078499114409e-06,
+      "loss": 11.9447,
+      "step": 15608
+    },
+    {
+      "epoch": 2.729704918032787,
+      "grad_norm": 0.8407045858736395,
+      "learning_rate": 4.224646829093426e-06,
+      "loss": 11.8414,
+      "step": 15609
+    },
+    {
+      "epoch": 2.729879781420765,
+      "grad_norm": 0.80409163092481,
+      "learning_rate": 4.219218573351891e-06,
+      "loss": 12.1128,
+      "step": 15610
+    },
+    {
+      "epoch": 2.7300546448087433,
+      "grad_norm": 0.8786075250148231,
+      "learning_rate": 4.213793732083304e-06,
+      "loss": 11.6735,
+      "step": 15611
+    },
+    {
+      "epoch": 2.730229508196721,
+      "grad_norm": 0.91998259475869,
+      "learning_rate": 4.208372305481067e-06,
+      "loss": 11.8141,
+      "step": 15612
+    },
+    {
+      "epoch": 2.7304043715846995,
+      "grad_norm": 0.8400916403513464,
+      "learning_rate": 4.202954293738426e-06,
+      "loss": 11.7782,
+      "step": 15613
+    },
+    {
+      "epoch": 2.7305792349726774,
+      "grad_norm": 0.8223763600917954,
+      "learning_rate": 4.197539697048536e-06,
+      "loss": 11.8557,
+      "step": 15614
+    },
+    {
+      "epoch": 2.7307540983606557,
+      "grad_norm": 0.7871047524487456,
+      "learning_rate": 4.192128515604432e-06,
+      "loss": 11.9004,
+      "step": 15615
+    },
+    {
+      "epoch": 2.730928961748634,
+      "grad_norm": 0.8436126496423331,
+      "learning_rate": 4.186720749598994e-06,
+      "loss": 11.6753,
+      "step": 15616
+    },
+    {
+      "epoch": 2.731103825136612,
+      "grad_norm": 0.8326300862867867,
+      "learning_rate": 4.181316399225022e-06,
+      "loss": 12.0024,
+      "step": 15617
+    },
+    {
+      "epoch": 2.7312786885245903,
+      "grad_norm": 0.7821942973624768,
+      "learning_rate": 4.175915464675162e-06,
+      "loss": 11.8513,
+      "step": 15618
+    },
+    {
+      "epoch": 2.7314535519125682,
+      "grad_norm": 0.898550661357881,
+      "learning_rate": 4.170517946141972e-06,
+      "loss": 11.7969,
+      "step": 15619
+    },
+    {
+      "epoch": 2.7316284153005466,
+      "grad_norm": 0.8126764338080396,
+      "learning_rate": 4.165123843817842e-06,
+      "loss": 11.9362,
+      "step": 15620
+    },
+    {
+      "epoch": 2.7318032786885245,
+      "grad_norm": 0.8891481743168584,
+      "learning_rate": 4.159733157895085e-06,
+      "loss": 11.6237,
+      "step": 15621
+    },
+    {
+      "epoch": 2.731978142076503,
+      "grad_norm": 0.8652827271142609,
+      "learning_rate": 4.1543458885658474e-06,
+      "loss": 11.8484,
+      "step": 15622
+    },
+    {
+      "epoch": 2.7321530054644807,
+      "grad_norm": 0.905697421605436,
+      "learning_rate": 4.148962036022219e-06,
+      "loss": 11.8825,
+      "step": 15623
+    },
+    {
+      "epoch": 2.732327868852459,
+      "grad_norm": 0.869028933540322,
+      "learning_rate": 4.143581600456081e-06,
+      "loss": 11.7784,
+      "step": 15624
+    },
+    {
+      "epoch": 2.7325027322404374,
+      "grad_norm": 0.8042882872079689,
+      "learning_rate": 4.138204582059269e-06,
+      "loss": 11.7851,
+      "step": 15625
+    },
+    {
+      "epoch": 2.7326775956284153,
+      "grad_norm": 0.8140708765742762,
+      "learning_rate": 4.132830981023461e-06,
+      "loss": 11.9336,
+      "step": 15626
+    },
+    {
+      "epoch": 2.732852459016393,
+      "grad_norm": 0.8710595586142156,
+      "learning_rate": 4.127460797540217e-06,
+      "loss": 11.8046,
+      "step": 15627
+    },
+    {
+      "epoch": 2.7330273224043715,
+      "grad_norm": 0.8404095936760951,
+      "learning_rate": 4.122094031800982e-06,
+      "loss": 12.0829,
+      "step": 15628
+    },
+    {
+      "epoch": 2.73320218579235,
+      "grad_norm": 0.8630518525828228,
+      "learning_rate": 4.1167306839970835e-06,
+      "loss": 11.8254,
+      "step": 15629
+    },
+    {
+      "epoch": 2.7333770491803278,
+      "grad_norm": 0.8320495736506619,
+      "learning_rate": 4.11137075431971e-06,
+      "loss": 11.6374,
+      "step": 15630
+    },
+    {
+      "epoch": 2.733551912568306,
+      "grad_norm": 0.8683661821839523,
+      "learning_rate": 4.106014242959921e-06,
+      "loss": 11.7935,
+      "step": 15631
+    },
+    {
+      "epoch": 2.733726775956284,
+      "grad_norm": 0.7950875345725554,
+      "learning_rate": 4.1006611501086865e-06,
+      "loss": 11.7437,
+      "step": 15632
+    },
+    {
+      "epoch": 2.7339016393442623,
+      "grad_norm": 0.8537606002770096,
+      "learning_rate": 4.095311475956842e-06,
+      "loss": 11.7433,
+      "step": 15633
+    },
+    {
+      "epoch": 2.7340765027322407,
+      "grad_norm": 0.8473128480675101,
+      "learning_rate": 4.089965220695091e-06,
+      "loss": 11.7195,
+      "step": 15634
+    },
+    {
+      "epoch": 2.7342513661202186,
+      "grad_norm": 0.9688574278486112,
+      "learning_rate": 4.084622384514014e-06,
+      "loss": 11.7192,
+      "step": 15635
+    },
+    {
+      "epoch": 2.7344262295081965,
+      "grad_norm": 0.8984070078636774,
+      "learning_rate": 4.079282967604092e-06,
+      "loss": 11.7998,
+      "step": 15636
+    },
+    {
+      "epoch": 2.734601092896175,
+      "grad_norm": 0.8704758439158591,
+      "learning_rate": 4.073946970155651e-06,
+      "loss": 11.9011,
+      "step": 15637
+    },
+    {
+      "epoch": 2.734775956284153,
+      "grad_norm": 0.8446343813198509,
+      "learning_rate": 4.068614392358938e-06,
+      "loss": 11.9319,
+      "step": 15638
+    },
+    {
+      "epoch": 2.734950819672131,
+      "grad_norm": 0.9168486700812436,
+      "learning_rate": 4.063285234404035e-06,
+      "loss": 11.9251,
+      "step": 15639
+    },
+    {
+      "epoch": 2.7351256830601094,
+      "grad_norm": 0.8412607118831462,
+      "learning_rate": 4.057959496480912e-06,
+      "loss": 11.8332,
+      "step": 15640
+    },
+    {
+      "epoch": 2.7353005464480873,
+      "grad_norm": 0.7714895455616335,
+      "learning_rate": 4.05263717877944e-06,
+      "loss": 11.9886,
+      "step": 15641
+    },
+    {
+      "epoch": 2.7354754098360656,
+      "grad_norm": 0.7895166094607927,
+      "learning_rate": 4.047318281489365e-06,
+      "loss": 11.7703,
+      "step": 15642
+    },
+    {
+      "epoch": 2.735650273224044,
+      "grad_norm": 0.8799169292097464,
+      "learning_rate": 4.04200280480026e-06,
+      "loss": 12.0474,
+      "step": 15643
+    },
+    {
+      "epoch": 2.735825136612022,
+      "grad_norm": 0.9133234516793397,
+      "learning_rate": 4.03669074890165e-06,
+      "loss": 12.0263,
+      "step": 15644
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.8798682065304622,
+      "learning_rate": 4.031382113982895e-06,
+      "loss": 11.7954,
+      "step": 15645
+    },
+    {
+      "epoch": 2.736174863387978,
+      "grad_norm": 0.8030134060558248,
+      "learning_rate": 4.026076900233244e-06,
+      "loss": 12.0932,
+      "step": 15646
+    },
+    {
+      "epoch": 2.7363497267759564,
+      "grad_norm": 0.83627948289696,
+      "learning_rate": 4.020775107841812e-06,
+      "loss": 11.9505,
+      "step": 15647
+    },
+    {
+      "epoch": 2.7365245901639343,
+      "grad_norm": 0.826419605005346,
+      "learning_rate": 4.0154767369976035e-06,
+      "loss": 11.821,
+      "step": 15648
+    },
+    {
+      "epoch": 2.7366994535519127,
+      "grad_norm": 0.8627450869622281,
+      "learning_rate": 4.010181787889522e-06,
+      "loss": 11.8384,
+      "step": 15649
+    },
+    {
+      "epoch": 2.7368743169398906,
+      "grad_norm": 0.8075618245817532,
+      "learning_rate": 4.004890260706295e-06,
+      "loss": 11.9921,
+      "step": 15650
+    },
+    {
+      "epoch": 2.737049180327869,
+      "grad_norm": 0.8217475551205984,
+      "learning_rate": 3.9996021556365725e-06,
+      "loss": 12.0118,
+      "step": 15651
+    },
+    {
+      "epoch": 2.7372240437158473,
+      "grad_norm": 0.8017404251860832,
+      "learning_rate": 3.994317472868869e-06,
+      "loss": 11.9925,
+      "step": 15652
+    },
+    {
+      "epoch": 2.737398907103825,
+      "grad_norm": 0.7884948441077213,
+      "learning_rate": 3.98903621259158e-06,
+      "loss": 12.0785,
+      "step": 15653
+    },
+    {
+      "epoch": 2.737573770491803,
+      "grad_norm": 0.8532920527909599,
+      "learning_rate": 3.983758374992974e-06,
+      "loss": 11.7438,
+      "step": 15654
+    },
+    {
+      "epoch": 2.7377486338797814,
+      "grad_norm": 0.8202211001602573,
+      "learning_rate": 3.978483960261192e-06,
+      "loss": 11.8549,
+      "step": 15655
+    },
+    {
+      "epoch": 2.7379234972677597,
+      "grad_norm": 0.869537251688767,
+      "learning_rate": 3.973212968584272e-06,
+      "loss": 11.9862,
+      "step": 15656
+    },
+    {
+      "epoch": 2.7380983606557376,
+      "grad_norm": 0.8210689812308655,
+      "learning_rate": 3.967945400150119e-06,
+      "loss": 11.7475,
+      "step": 15657
+    },
+    {
+      "epoch": 2.738273224043716,
+      "grad_norm": 0.8240764888803481,
+      "learning_rate": 3.962681255146505e-06,
+      "loss": 11.7972,
+      "step": 15658
+    },
+    {
+      "epoch": 2.738448087431694,
+      "grad_norm": 0.8646325031511477,
+      "learning_rate": 3.9574205337610916e-06,
+      "loss": 11.66,
+      "step": 15659
+    },
+    {
+      "epoch": 2.738622950819672,
+      "grad_norm": 0.7845376921702875,
+      "learning_rate": 3.952163236181428e-06,
+      "loss": 11.6876,
+      "step": 15660
+    },
+    {
+      "epoch": 2.73879781420765,
+      "grad_norm": 0.8889075062654325,
+      "learning_rate": 3.946909362594942e-06,
+      "loss": 11.7892,
+      "step": 15661
+    },
+    {
+      "epoch": 2.7389726775956285,
+      "grad_norm": 0.7761046985250322,
+      "learning_rate": 3.9416589131888945e-06,
+      "loss": 11.928,
+      "step": 15662
+    },
+    {
+      "epoch": 2.7391475409836064,
+      "grad_norm": 0.8838313276549233,
+      "learning_rate": 3.936411888150471e-06,
+      "loss": 11.921,
+      "step": 15663
+    },
+    {
+      "epoch": 2.7393224043715847,
+      "grad_norm": 0.839148099465914,
+      "learning_rate": 3.93116828766672e-06,
+      "loss": 11.7625,
+      "step": 15664
+    },
+    {
+      "epoch": 2.739497267759563,
+      "grad_norm": 0.7659896356723445,
+      "learning_rate": 3.925928111924581e-06,
+      "loss": 11.7627,
+      "step": 15665
+    },
+    {
+      "epoch": 2.739672131147541,
+      "grad_norm": 0.9078809925247756,
+      "learning_rate": 3.920691361110862e-06,
+      "loss": 12.0003,
+      "step": 15666
+    },
+    {
+      "epoch": 2.7398469945355193,
+      "grad_norm": 0.8144262933342807,
+      "learning_rate": 3.9154580354122335e-06,
+      "loss": 11.8848,
+      "step": 15667
+    },
+    {
+      "epoch": 2.740021857923497,
+      "grad_norm": 0.8098636649164741,
+      "learning_rate": 3.910228135015271e-06,
+      "loss": 12.0472,
+      "step": 15668
+    },
+    {
+      "epoch": 2.7401967213114755,
+      "grad_norm": 0.7886895158536884,
+      "learning_rate": 3.9050016601064e-06,
+      "loss": 11.8835,
+      "step": 15669
+    },
+    {
+      "epoch": 2.7403715846994534,
+      "grad_norm": 0.786541541158548,
+      "learning_rate": 3.89977861087194e-06,
+      "loss": 12.0169,
+      "step": 15670
+    },
+    {
+      "epoch": 2.7405464480874318,
+      "grad_norm": 0.7272343372347625,
+      "learning_rate": 3.894558987498098e-06,
+      "loss": 11.9584,
+      "step": 15671
+    },
+    {
+      "epoch": 2.7407213114754096,
+      "grad_norm": 0.8333224703825313,
+      "learning_rate": 3.889342790170947e-06,
+      "loss": 11.6898,
+      "step": 15672
+    },
+    {
+      "epoch": 2.740896174863388,
+      "grad_norm": 0.8264659099229538,
+      "learning_rate": 3.8841300190764265e-06,
+      "loss": 11.7499,
+      "step": 15673
+    },
+    {
+      "epoch": 2.7410710382513663,
+      "grad_norm": 0.8724729759769949,
+      "learning_rate": 3.878920674400366e-06,
+      "loss": 11.8642,
+      "step": 15674
+    },
+    {
+      "epoch": 2.7412459016393442,
+      "grad_norm": 0.8822812090883947,
+      "learning_rate": 3.873714756328484e-06,
+      "loss": 11.814,
+      "step": 15675
+    },
+    {
+      "epoch": 2.7414207650273226,
+      "grad_norm": 0.7660484236248114,
+      "learning_rate": 3.868512265046364e-06,
+      "loss": 11.9153,
+      "step": 15676
+    },
+    {
+      "epoch": 2.7415956284153005,
+      "grad_norm": 0.8548247677800268,
+      "learning_rate": 3.86331320073946e-06,
+      "loss": 11.9171,
+      "step": 15677
+    },
+    {
+      "epoch": 2.741770491803279,
+      "grad_norm": 0.7821201065630047,
+      "learning_rate": 3.858117563593111e-06,
+      "loss": 11.5789,
+      "step": 15678
+    },
+    {
+      "epoch": 2.7419453551912567,
+      "grad_norm": 0.8837383043380017,
+      "learning_rate": 3.852925353792547e-06,
+      "loss": 11.5339,
+      "step": 15679
+    },
+    {
+      "epoch": 2.742120218579235,
+      "grad_norm": 0.8054502959695093,
+      "learning_rate": 3.847736571522853e-06,
+      "loss": 11.8405,
+      "step": 15680
+    },
+    {
+      "epoch": 2.742295081967213,
+      "grad_norm": 0.8640219077427183,
+      "learning_rate": 3.8425512169690035e-06,
+      "loss": 11.7923,
+      "step": 15681
+    },
+    {
+      "epoch": 2.7424699453551913,
+      "grad_norm": 0.7832929810327016,
+      "learning_rate": 3.837369290315862e-06,
+      "loss": 11.7533,
+      "step": 15682
+    },
+    {
+      "epoch": 2.7426448087431696,
+      "grad_norm": 0.8328528066717115,
+      "learning_rate": 3.8321907917481695e-06,
+      "loss": 11.9544,
+      "step": 15683
+    },
+    {
+      "epoch": 2.7428196721311475,
+      "grad_norm": 0.7983490325178295,
+      "learning_rate": 3.827015721450489e-06,
+      "loss": 11.8889,
+      "step": 15684
+    },
+    {
+      "epoch": 2.7429945355191254,
+      "grad_norm": 0.8097717583251222,
+      "learning_rate": 3.8218440796073395e-06,
+      "loss": 12.0028,
+      "step": 15685
+    },
+    {
+      "epoch": 2.7431693989071038,
+      "grad_norm": 0.8338414523486561,
+      "learning_rate": 3.816675866403075e-06,
+      "loss": 11.9533,
+      "step": 15686
+    },
+    {
+      "epoch": 2.743344262295082,
+      "grad_norm": 0.90358351118349,
+      "learning_rate": 3.811511082021957e-06,
+      "loss": 11.9124,
+      "step": 15687
+    },
+    {
+      "epoch": 2.74351912568306,
+      "grad_norm": 0.7750280073077421,
+      "learning_rate": 3.806349726648062e-06,
+      "loss": 11.6414,
+      "step": 15688
+    },
+    {
+      "epoch": 2.7436939890710383,
+      "grad_norm": 0.8156779538850513,
+      "learning_rate": 3.801191800465409e-06,
+      "loss": 11.5548,
+      "step": 15689
+    },
+    {
+      "epoch": 2.7438688524590162,
+      "grad_norm": 0.795157291796279,
+      "learning_rate": 3.7960373036578846e-06,
+      "loss": 11.7328,
+      "step": 15690
+    },
+    {
+      "epoch": 2.7440437158469946,
+      "grad_norm": 0.8706093296779608,
+      "learning_rate": 3.79088623640923e-06,
+      "loss": 11.9691,
+      "step": 15691
+    },
+    {
+      "epoch": 2.744218579234973,
+      "grad_norm": 0.815276986536006,
+      "learning_rate": 3.785738598903066e-06,
+      "loss": 12.0065,
+      "step": 15692
+    },
+    {
+      "epoch": 2.744393442622951,
+      "grad_norm": 0.8361287565647082,
+      "learning_rate": 3.7805943913229003e-06,
+      "loss": 11.9532,
+      "step": 15693
+    },
+    {
+      "epoch": 2.7445683060109287,
+      "grad_norm": 0.8126333099325989,
+      "learning_rate": 3.7754536138521425e-06,
+      "loss": 11.8748,
+      "step": 15694
+    },
+    {
+      "epoch": 2.744743169398907,
+      "grad_norm": 0.8301253775868079,
+      "learning_rate": 3.770316266674023e-06,
+      "loss": 11.7677,
+      "step": 15695
+    },
+    {
+      "epoch": 2.7449180327868854,
+      "grad_norm": 0.8170907829411355,
+      "learning_rate": 3.7651823499716855e-06,
+      "loss": 11.7968,
+      "step": 15696
+    },
+    {
+      "epoch": 2.7450928961748633,
+      "grad_norm": 0.8348481771656407,
+      "learning_rate": 3.7600518639281713e-06,
+      "loss": 11.8374,
+      "step": 15697
+    },
+    {
+      "epoch": 2.7452677595628416,
+      "grad_norm": 0.7702942548879592,
+      "learning_rate": 3.7549248087263676e-06,
+      "loss": 11.9053,
+      "step": 15698
+    },
+    {
+      "epoch": 2.7454426229508195,
+      "grad_norm": 0.8398572601368448,
+      "learning_rate": 3.7498011845490287e-06,
+      "loss": 11.7579,
+      "step": 15699
+    },
+    {
+      "epoch": 2.745617486338798,
+      "grad_norm": 0.8157753278775955,
+      "learning_rate": 3.7446809915788306e-06,
+      "loss": 11.731,
+      "step": 15700
+    },
+    {
+      "epoch": 2.745792349726776,
+      "grad_norm": 0.8110964955707791,
+      "learning_rate": 3.7395642299982934e-06,
+      "loss": 11.9632,
+      "step": 15701
+    },
+    {
+      "epoch": 2.745967213114754,
+      "grad_norm": 0.7769331631920973,
+      "learning_rate": 3.7344508999898387e-06,
+      "loss": 11.7398,
+      "step": 15702
+    },
+    {
+      "epoch": 2.746142076502732,
+      "grad_norm": 0.8580054995844849,
+      "learning_rate": 3.72934100173572e-06,
+      "loss": 11.3737,
+      "step": 15703
+    },
+    {
+      "epoch": 2.7463169398907104,
+      "grad_norm": 0.8798545701451488,
+      "learning_rate": 3.724234535418114e-06,
+      "loss": 11.6431,
+      "step": 15704
+    },
+    {
+      "epoch": 2.7464918032786887,
+      "grad_norm": 0.8174295708929082,
+      "learning_rate": 3.7191315012190532e-06,
+      "loss": 11.652,
+      "step": 15705
+    },
+    {
+      "epoch": 2.7466666666666666,
+      "grad_norm": 0.8047264057416044,
+      "learning_rate": 3.7140318993204916e-06,
+      "loss": 11.8166,
+      "step": 15706
+    },
+    {
+      "epoch": 2.746841530054645,
+      "grad_norm": 0.8040251116902474,
+      "learning_rate": 3.7089357299041616e-06,
+      "loss": 11.9631,
+      "step": 15707
+    },
+    {
+      "epoch": 2.747016393442623,
+      "grad_norm": 0.8529004536597248,
+      "learning_rate": 3.703842993151785e-06,
+      "loss": 11.8775,
+      "step": 15708
+    },
+    {
+      "epoch": 2.747191256830601,
+      "grad_norm": 0.8892977159628354,
+      "learning_rate": 3.698753689244905e-06,
+      "loss": 11.9432,
+      "step": 15709
+    },
+    {
+      "epoch": 2.7473661202185795,
+      "grad_norm": 0.8597392830565885,
+      "learning_rate": 3.693667818364921e-06,
+      "loss": 11.9107,
+      "step": 15710
+    },
+    {
+      "epoch": 2.7475409836065574,
+      "grad_norm": 0.8254417041364496,
+      "learning_rate": 3.6885853806931656e-06,
+      "loss": 11.8656,
+      "step": 15711
+    },
+    {
+      "epoch": 2.7477158469945353,
+      "grad_norm": 0.8107269430481171,
+      "learning_rate": 3.6835063764108057e-06,
+      "loss": 11.866,
+      "step": 15712
+    },
+    {
+      "epoch": 2.7478907103825136,
+      "grad_norm": 0.8232508663089344,
+      "learning_rate": 3.6784308056989293e-06,
+      "loss": 11.6654,
+      "step": 15713
+    },
+    {
+      "epoch": 2.748065573770492,
+      "grad_norm": 0.8544563737847082,
+      "learning_rate": 3.673358668738436e-06,
+      "loss": 12.0555,
+      "step": 15714
+    },
+    {
+      "epoch": 2.74824043715847,
+      "grad_norm": 0.7943874056527837,
+      "learning_rate": 3.668289965710159e-06,
+      "loss": 11.5418,
+      "step": 15715
+    },
+    {
+      "epoch": 2.7484153005464482,
+      "grad_norm": 0.8027812315010039,
+      "learning_rate": 3.6632246967947983e-06,
+      "loss": 11.7101,
+      "step": 15716
+    },
+    {
+      "epoch": 2.748590163934426,
+      "grad_norm": 0.8434682507826995,
+      "learning_rate": 3.6581628621729202e-06,
+      "loss": 11.4035,
+      "step": 15717
+    },
+    {
+      "epoch": 2.7487650273224045,
+      "grad_norm": 0.8565048713474731,
+      "learning_rate": 3.6531044620249812e-06,
+      "loss": 11.9709,
+      "step": 15718
+    },
+    {
+      "epoch": 2.7489398907103824,
+      "grad_norm": 0.8265877863376859,
+      "learning_rate": 3.6480494965312807e-06,
+      "loss": 11.6546,
+      "step": 15719
+    },
+    {
+      "epoch": 2.7491147540983607,
+      "grad_norm": 0.7853194170610297,
+      "learning_rate": 3.6429979658720524e-06,
+      "loss": 11.8524,
+      "step": 15720
+    },
+    {
+      "epoch": 2.7492896174863386,
+      "grad_norm": 0.7962300653710779,
+      "learning_rate": 3.637949870227375e-06,
+      "loss": 11.9503,
+      "step": 15721
+    },
+    {
+      "epoch": 2.749464480874317,
+      "grad_norm": 0.7958736765625336,
+      "learning_rate": 3.632905209777182e-06,
+      "loss": 11.6896,
+      "step": 15722
+    },
+    {
+      "epoch": 2.7496393442622953,
+      "grad_norm": 0.82692272555272,
+      "learning_rate": 3.6278639847013296e-06,
+      "loss": 11.9272,
+      "step": 15723
+    },
+    {
+      "epoch": 2.749814207650273,
+      "grad_norm": 0.8793898567067108,
+      "learning_rate": 3.6228261951795406e-06,
+      "loss": 11.9433,
+      "step": 15724
+    },
+    {
+      "epoch": 2.7499890710382515,
+      "grad_norm": 0.7945093195761472,
+      "learning_rate": 3.6177918413913824e-06,
+      "loss": 11.8459,
+      "step": 15725
+    },
+    {
+      "epoch": 2.7501639344262294,
+      "grad_norm": 0.85509840591196,
+      "learning_rate": 3.612760923516323e-06,
+      "loss": 11.7408,
+      "step": 15726
+    },
+    {
+      "epoch": 2.7503387978142078,
+      "grad_norm": 0.7640655706720301,
+      "learning_rate": 3.6077334417337294e-06,
+      "loss": 11.4419,
+      "step": 15727
+    },
+    {
+      "epoch": 2.7505136612021857,
+      "grad_norm": 0.7611282970371427,
+      "learning_rate": 3.6027093962228366e-06,
+      "loss": 11.9006,
+      "step": 15728
+    },
+    {
+      "epoch": 2.750688524590164,
+      "grad_norm": 0.9413569901076668,
+      "learning_rate": 3.5976887871627006e-06,
+      "loss": 11.7655,
+      "step": 15729
+    },
+    {
+      "epoch": 2.750863387978142,
+      "grad_norm": 0.8663627368923627,
+      "learning_rate": 3.5926716147323236e-06,
+      "loss": 11.7886,
+      "step": 15730
+    },
+    {
+      "epoch": 2.7510382513661202,
+      "grad_norm": 0.7935205975650533,
+      "learning_rate": 3.587657879110573e-06,
+      "loss": 11.8233,
+      "step": 15731
+    },
+    {
+      "epoch": 2.7512131147540986,
+      "grad_norm": 0.8785217728675181,
+      "learning_rate": 3.5826475804761837e-06,
+      "loss": 11.7693,
+      "step": 15732
+    },
+    {
+      "epoch": 2.7513879781420765,
+      "grad_norm": 0.7921608527360304,
+      "learning_rate": 3.5776407190077354e-06,
+      "loss": 12.0536,
+      "step": 15733
+    },
+    {
+      "epoch": 2.7515628415300544,
+      "grad_norm": 0.8243234286258313,
+      "learning_rate": 3.5726372948837517e-06,
+      "loss": 11.7143,
+      "step": 15734
+    },
+    {
+      "epoch": 2.7517377049180327,
+      "grad_norm": 0.8514063363947224,
+      "learning_rate": 3.567637308282579e-06,
+      "loss": 12.0192,
+      "step": 15735
+    },
+    {
+      "epoch": 2.751912568306011,
+      "grad_norm": 0.7897813677313626,
+      "learning_rate": 3.562640759382463e-06,
+      "loss": 11.9157,
+      "step": 15736
+    },
+    {
+      "epoch": 2.752087431693989,
+      "grad_norm": 0.8023627390793941,
+      "learning_rate": 3.5576476483615396e-06,
+      "loss": 12.0266,
+      "step": 15737
+    },
+    {
+      "epoch": 2.7522622950819673,
+      "grad_norm": 0.8020379087395987,
+      "learning_rate": 3.5526579753977883e-06,
+      "loss": 11.8262,
+      "step": 15738
+    },
+    {
+      "epoch": 2.752437158469945,
+      "grad_norm": 0.8139937576189875,
+      "learning_rate": 3.547671740669101e-06,
+      "loss": 11.9361,
+      "step": 15739
+    },
+    {
+      "epoch": 2.7526120218579235,
+      "grad_norm": 0.8168731167792523,
+      "learning_rate": 3.542688944353223e-06,
+      "loss": 11.8893,
+      "step": 15740
+    },
+    {
+      "epoch": 2.752786885245902,
+      "grad_norm": 0.7997572507737364,
+      "learning_rate": 3.53770958662778e-06,
+      "loss": 11.9264,
+      "step": 15741
+    },
+    {
+      "epoch": 2.7529617486338798,
+      "grad_norm": 0.863695871638748,
+      "learning_rate": 3.532733667670285e-06,
+      "loss": 11.9086,
+      "step": 15742
+    },
+    {
+      "epoch": 2.7531366120218577,
+      "grad_norm": 0.7740174705879369,
+      "learning_rate": 3.5277611876581406e-06,
+      "loss": 11.9233,
+      "step": 15743
+    },
+    {
+      "epoch": 2.753311475409836,
+      "grad_norm": 0.8130841118002271,
+      "learning_rate": 3.522792146768572e-06,
+      "loss": 11.5725,
+      "step": 15744
+    },
+    {
+      "epoch": 2.7534863387978143,
+      "grad_norm": 0.8614215669681667,
+      "learning_rate": 3.5178265451787596e-06,
+      "loss": 11.8106,
+      "step": 15745
+    },
+    {
+      "epoch": 2.7536612021857922,
+      "grad_norm": 0.8273130935222656,
+      "learning_rate": 3.512864383065695e-06,
+      "loss": 11.7788,
+      "step": 15746
+    },
+    {
+      "epoch": 2.7538360655737706,
+      "grad_norm": 0.8273361527865064,
+      "learning_rate": 3.5079056606062923e-06,
+      "loss": 11.9174,
+      "step": 15747
+    },
+    {
+      "epoch": 2.7540109289617485,
+      "grad_norm": 0.8487969704748238,
+      "learning_rate": 3.5029503779772986e-06,
+      "loss": 11.6183,
+      "step": 15748
+    },
+    {
+      "epoch": 2.754185792349727,
+      "grad_norm": 0.8716691887142473,
+      "learning_rate": 3.4979985353553846e-06,
+      "loss": 11.6709,
+      "step": 15749
+    },
+    {
+      "epoch": 2.754360655737705,
+      "grad_norm": 0.8319713106648681,
+      "learning_rate": 3.493050132917064e-06,
+      "loss": 11.8023,
+      "step": 15750
+    },
+    {
+      "epoch": 2.754535519125683,
+      "grad_norm": 0.8564835034799165,
+      "learning_rate": 3.4881051708387735e-06,
+      "loss": 11.8479,
+      "step": 15751
+    },
+    {
+      "epoch": 2.754710382513661,
+      "grad_norm": 0.798840582251687,
+      "learning_rate": 3.48316364929675e-06,
+      "loss": 11.8182,
+      "step": 15752
+    },
+    {
+      "epoch": 2.7548852459016393,
+      "grad_norm": 0.7783540672033833,
+      "learning_rate": 3.478225568467186e-06,
+      "loss": 11.5995,
+      "step": 15753
+    },
+    {
+      "epoch": 2.7550601092896176,
+      "grad_norm": 0.8153405269920185,
+      "learning_rate": 3.4732909285260964e-06,
+      "loss": 11.9228,
+      "step": 15754
+    },
+    {
+      "epoch": 2.7552349726775955,
+      "grad_norm": 0.8312431212900911,
+      "learning_rate": 3.4683597296494177e-06,
+      "loss": 11.9228,
+      "step": 15755
+    },
+    {
+      "epoch": 2.755409836065574,
+      "grad_norm": 0.8461529976156609,
+      "learning_rate": 3.4634319720129203e-06,
+      "loss": 11.7101,
+      "step": 15756
+    },
+    {
+      "epoch": 2.755584699453552,
+      "grad_norm": 0.8770693106652995,
+      "learning_rate": 3.4585076557922756e-06,
+      "loss": 12.0439,
+      "step": 15757
+    },
+    {
+      "epoch": 2.75575956284153,
+      "grad_norm": 0.8238426501057171,
+      "learning_rate": 3.453586781163065e-06,
+      "loss": 12.0974,
+      "step": 15758
+    },
+    {
+      "epoch": 2.7559344262295085,
+      "grad_norm": 0.7942619098676408,
+      "learning_rate": 3.4486693483006593e-06,
+      "loss": 11.8427,
+      "step": 15759
+    },
+    {
+      "epoch": 2.7561092896174864,
+      "grad_norm": 0.8516083976870745,
+      "learning_rate": 3.443755357380385e-06,
+      "loss": 11.8812,
+      "step": 15760
+    },
+    {
+      "epoch": 2.7562841530054643,
+      "grad_norm": 0.8803538152925586,
+      "learning_rate": 3.438844808577424e-06,
+      "loss": 11.8233,
+      "step": 15761
+    },
+    {
+      "epoch": 2.7564590163934426,
+      "grad_norm": 0.8588385886421679,
+      "learning_rate": 3.433937702066825e-06,
+      "loss": 11.7718,
+      "step": 15762
+    },
+    {
+      "epoch": 2.756633879781421,
+      "grad_norm": 0.8271555948140127,
+      "learning_rate": 3.4290340380235265e-06,
+      "loss": 11.7985,
+      "step": 15763
+    },
+    {
+      "epoch": 2.756808743169399,
+      "grad_norm": 0.835136461335621,
+      "learning_rate": 3.4241338166223213e-06,
+      "loss": 11.5244,
+      "step": 15764
+    },
+    {
+      "epoch": 2.756983606557377,
+      "grad_norm": 0.8181912406963159,
+      "learning_rate": 3.4192370380379146e-06,
+      "loss": 11.92,
+      "step": 15765
+    },
+    {
+      "epoch": 2.757158469945355,
+      "grad_norm": 0.8320039032163287,
+      "learning_rate": 3.414343702444878e-06,
+      "loss": 11.6573,
+      "step": 15766
+    },
+    {
+      "epoch": 2.7573333333333334,
+      "grad_norm": 0.8538069014754442,
+      "learning_rate": 3.4094538100176264e-06,
+      "loss": 11.7934,
+      "step": 15767
+    },
+    {
+      "epoch": 2.7575081967213113,
+      "grad_norm": 0.7983870555623431,
+      "learning_rate": 3.4045673609304883e-06,
+      "loss": 11.814,
+      "step": 15768
+    },
+    {
+      "epoch": 2.7576830601092897,
+      "grad_norm": 0.8045062960864707,
+      "learning_rate": 3.3996843553576685e-06,
+      "loss": 11.282,
+      "step": 15769
+    },
+    {
+      "epoch": 2.7578579234972675,
+      "grad_norm": 0.803969270380976,
+      "learning_rate": 3.3948047934732386e-06,
+      "loss": 11.6858,
+      "step": 15770
+    },
+    {
+      "epoch": 2.758032786885246,
+      "grad_norm": 0.8691728489191425,
+      "learning_rate": 3.3899286754511372e-06,
+      "loss": 11.7978,
+      "step": 15771
+    },
+    {
+      "epoch": 2.7582076502732242,
+      "grad_norm": 0.8010942965983536,
+      "learning_rate": 3.3850560014652146e-06,
+      "loss": 11.6272,
+      "step": 15772
+    },
+    {
+      "epoch": 2.758382513661202,
+      "grad_norm": 0.8166863748421583,
+      "learning_rate": 3.3801867716891645e-06,
+      "loss": 11.7385,
+      "step": 15773
+    },
+    {
+      "epoch": 2.7585573770491805,
+      "grad_norm": 0.8285208790143234,
+      "learning_rate": 3.3753209862965596e-06,
+      "loss": 11.8098,
+      "step": 15774
+    },
+    {
+      "epoch": 2.7587322404371584,
+      "grad_norm": 0.7789643099778406,
+      "learning_rate": 3.3704586454608613e-06,
+      "loss": 11.7856,
+      "step": 15775
+    },
+    {
+      "epoch": 2.7589071038251367,
+      "grad_norm": 0.8543579417850826,
+      "learning_rate": 3.365599749355419e-06,
+      "loss": 11.8485,
+      "step": 15776
+    },
+    {
+      "epoch": 2.7590819672131146,
+      "grad_norm": 0.8001748072716595,
+      "learning_rate": 3.360744298153462e-06,
+      "loss": 11.8069,
+      "step": 15777
+    },
+    {
+      "epoch": 2.759256830601093,
+      "grad_norm": 0.8512490194698998,
+      "learning_rate": 3.355892292028029e-06,
+      "loss": 11.8975,
+      "step": 15778
+    },
+    {
+      "epoch": 2.759431693989071,
+      "grad_norm": 0.7973388227748305,
+      "learning_rate": 3.351043731152137e-06,
+      "loss": 11.758,
+      "step": 15779
+    },
+    {
+      "epoch": 2.759606557377049,
+      "grad_norm": 0.8363384366960619,
+      "learning_rate": 3.3461986156986035e-06,
+      "loss": 11.6413,
+      "step": 15780
+    },
+    {
+      "epoch": 2.7597814207650275,
+      "grad_norm": 0.7969930374195392,
+      "learning_rate": 3.3413569458401795e-06,
+      "loss": 11.9091,
+      "step": 15781
+    },
+    {
+      "epoch": 2.7599562841530054,
+      "grad_norm": 0.8333653713054573,
+      "learning_rate": 3.3365187217494266e-06,
+      "loss": 11.7201,
+      "step": 15782
+    },
+    {
+      "epoch": 2.7601311475409838,
+      "grad_norm": 0.7983395897934864,
+      "learning_rate": 3.331683943598851e-06,
+      "loss": 11.8237,
+      "step": 15783
+    },
+    {
+      "epoch": 2.7603060109289617,
+      "grad_norm": 0.795106383348524,
+      "learning_rate": 3.3268526115607934e-06,
+      "loss": 11.7393,
+      "step": 15784
+    },
+    {
+      "epoch": 2.76048087431694,
+      "grad_norm": 0.8383642994183044,
+      "learning_rate": 3.3220247258074934e-06,
+      "loss": 11.7348,
+      "step": 15785
+    },
+    {
+      "epoch": 2.760655737704918,
+      "grad_norm": 0.8245249189687871,
+      "learning_rate": 3.3172002865110575e-06,
+      "loss": 11.871,
+      "step": 15786
+    },
+    {
+      "epoch": 2.7608306010928962,
+      "grad_norm": 0.8382263067817732,
+      "learning_rate": 3.3123792938434593e-06,
+      "loss": 11.892,
+      "step": 15787
+    },
+    {
+      "epoch": 2.761005464480874,
+      "grad_norm": 0.8612569324411191,
+      "learning_rate": 3.3075617479765842e-06,
+      "loss": 11.7489,
+      "step": 15788
+    },
+    {
+      "epoch": 2.7611803278688525,
+      "grad_norm": 0.8353789997054182,
+      "learning_rate": 3.3027476490821496e-06,
+      "loss": 11.7302,
+      "step": 15789
+    },
+    {
+      "epoch": 2.761355191256831,
+      "grad_norm": 0.8525516948618592,
+      "learning_rate": 3.2979369973317854e-06,
+      "loss": 11.6236,
+      "step": 15790
+    },
+    {
+      "epoch": 2.7615300546448087,
+      "grad_norm": 0.7861562981089592,
+      "learning_rate": 3.2931297928969875e-06,
+      "loss": 11.7629,
+      "step": 15791
+    },
+    {
+      "epoch": 2.7617049180327866,
+      "grad_norm": 0.8327535736637184,
+      "learning_rate": 3.2883260359491297e-06,
+      "loss": 11.6454,
+      "step": 15792
+    },
+    {
+      "epoch": 2.761879781420765,
+      "grad_norm": 0.8215249507492224,
+      "learning_rate": 3.283525726659442e-06,
+      "loss": 11.9718,
+      "step": 15793
+    },
+    {
+      "epoch": 2.7620546448087433,
+      "grad_norm": 0.9486456331509118,
+      "learning_rate": 3.2787288651990545e-06,
+      "loss": 11.3573,
+      "step": 15794
+    },
+    {
+      "epoch": 2.762229508196721,
+      "grad_norm": 0.7960708874439041,
+      "learning_rate": 3.2739354517389855e-06,
+      "loss": 11.7911,
+      "step": 15795
+    },
+    {
+      "epoch": 2.7624043715846995,
+      "grad_norm": 0.8252705682465772,
+      "learning_rate": 3.2691454864501203e-06,
+      "loss": 11.9057,
+      "step": 15796
+    },
+    {
+      "epoch": 2.7625792349726774,
+      "grad_norm": 0.7772636382859855,
+      "learning_rate": 3.2643589695031784e-06,
+      "loss": 11.7673,
+      "step": 15797
+    },
+    {
+      "epoch": 2.7627540983606558,
+      "grad_norm": 0.8298219971849221,
+      "learning_rate": 3.2595759010688234e-06,
+      "loss": 11.8227,
+      "step": 15798
+    },
+    {
+      "epoch": 2.762928961748634,
+      "grad_norm": 0.8490371235475158,
+      "learning_rate": 3.2547962813175515e-06,
+      "loss": 11.806,
+      "step": 15799
+    },
+    {
+      "epoch": 2.763103825136612,
+      "grad_norm": 0.8287363179976258,
+      "learning_rate": 3.2500201104197824e-06,
+      "loss": 11.9016,
+      "step": 15800
+    },
+    {
+      "epoch": 2.76327868852459,
+      "grad_norm": 0.8696186885663012,
+      "learning_rate": 3.245247388545736e-06,
+      "loss": 11.8075,
+      "step": 15801
+    },
+    {
+      "epoch": 2.7634535519125683,
+      "grad_norm": 0.7860995208829102,
+      "learning_rate": 3.240478115865564e-06,
+      "loss": 11.8701,
+      "step": 15802
+    },
+    {
+      "epoch": 2.7636284153005466,
+      "grad_norm": 0.8158286629221627,
+      "learning_rate": 3.23571229254932e-06,
+      "loss": 11.7893,
+      "step": 15803
+    },
+    {
+      "epoch": 2.7638032786885245,
+      "grad_norm": 0.8318648265703249,
+      "learning_rate": 3.230949918766868e-06,
+      "loss": 11.7763,
+      "step": 15804
+    },
+    {
+      "epoch": 2.763978142076503,
+      "grad_norm": 0.7922373141363376,
+      "learning_rate": 3.2261909946879833e-06,
+      "loss": 11.8507,
+      "step": 15805
+    },
+    {
+      "epoch": 2.7641530054644807,
+      "grad_norm": 0.8815905919611918,
+      "learning_rate": 3.2214355204823186e-06,
+      "loss": 11.7785,
+      "step": 15806
+    },
+    {
+      "epoch": 2.764327868852459,
+      "grad_norm": 0.8250128422439158,
+      "learning_rate": 3.216683496319406e-06,
+      "loss": 11.7934,
+      "step": 15807
+    },
+    {
+      "epoch": 2.7645027322404374,
+      "grad_norm": 0.8680551943603478,
+      "learning_rate": 3.2119349223686422e-06,
+      "loss": 11.8647,
+      "step": 15808
+    },
+    {
+      "epoch": 2.7646775956284153,
+      "grad_norm": 0.8396296461296877,
+      "learning_rate": 3.207189798799326e-06,
+      "loss": 11.8596,
+      "step": 15809
+    },
+    {
+      "epoch": 2.764852459016393,
+      "grad_norm": 0.7905210664558027,
+      "learning_rate": 3.2024481257805883e-06,
+      "loss": 11.9864,
+      "step": 15810
+    },
+    {
+      "epoch": 2.7650273224043715,
+      "grad_norm": 0.8393247626425642,
+      "learning_rate": 3.197709903481505e-06,
+      "loss": 11.4746,
+      "step": 15811
+    },
+    {
+      "epoch": 2.76520218579235,
+      "grad_norm": 0.823774970749283,
+      "learning_rate": 3.1929751320709413e-06,
+      "loss": 11.8378,
+      "step": 15812
+    },
+    {
+      "epoch": 2.765377049180328,
+      "grad_norm": 0.8508474879547898,
+      "learning_rate": 3.1882438117177173e-06,
+      "loss": 12.0934,
+      "step": 15813
+    },
+    {
+      "epoch": 2.765551912568306,
+      "grad_norm": 0.8290837359498663,
+      "learning_rate": 3.183515942590476e-06,
+      "loss": 11.7951,
+      "step": 15814
+    },
+    {
+      "epoch": 2.765726775956284,
+      "grad_norm": 0.8207953605992375,
+      "learning_rate": 3.1787915248577938e-06,
+      "loss": 11.7523,
+      "step": 15815
+    },
+    {
+      "epoch": 2.7659016393442624,
+      "grad_norm": 0.7974272784367854,
+      "learning_rate": 3.174070558688058e-06,
+      "loss": 11.8255,
+      "step": 15816
+    },
+    {
+      "epoch": 2.7660765027322407,
+      "grad_norm": 0.8052514363736288,
+      "learning_rate": 3.1693530442495787e-06,
+      "loss": 11.8017,
+      "step": 15817
+    },
+    {
+      "epoch": 2.7662513661202186,
+      "grad_norm": 0.8042553838481178,
+      "learning_rate": 3.1646389817105327e-06,
+      "loss": 11.9348,
+      "step": 15818
+    },
+    {
+      "epoch": 2.7664262295081965,
+      "grad_norm": 0.7882496763516599,
+      "learning_rate": 3.1599283712389625e-06,
+      "loss": 11.8329,
+      "step": 15819
+    },
+    {
+      "epoch": 2.766601092896175,
+      "grad_norm": 0.8159705016034523,
+      "learning_rate": 3.155221213002801e-06,
+      "loss": 11.8403,
+      "step": 15820
+    },
+    {
+      "epoch": 2.766775956284153,
+      "grad_norm": 0.8161102046774616,
+      "learning_rate": 3.1505175071698477e-06,
+      "loss": 11.9601,
+      "step": 15821
+    },
+    {
+      "epoch": 2.766950819672131,
+      "grad_norm": 0.8049243183896958,
+      "learning_rate": 3.1458172539078014e-06,
+      "loss": 11.6523,
+      "step": 15822
+    },
+    {
+      "epoch": 2.7671256830601094,
+      "grad_norm": 0.8427288120962814,
+      "learning_rate": 3.141120453384194e-06,
+      "loss": 11.885,
+      "step": 15823
+    },
+    {
+      "epoch": 2.7673005464480873,
+      "grad_norm": 0.7801313596942929,
+      "learning_rate": 3.136427105766482e-06,
+      "loss": 11.6884,
+      "step": 15824
+    },
+    {
+      "epoch": 2.7674754098360657,
+      "grad_norm": 0.8550256633376712,
+      "learning_rate": 3.1317372112219633e-06,
+      "loss": 11.6966,
+      "step": 15825
+    },
+    {
+      "epoch": 2.7676502732240436,
+      "grad_norm": 0.8080784919387496,
+      "learning_rate": 3.1270507699178387e-06,
+      "loss": 11.8313,
+      "step": 15826
+    },
+    {
+      "epoch": 2.767825136612022,
+      "grad_norm": 0.7924828843252214,
+      "learning_rate": 3.122367782021174e-06,
+      "loss": 11.7283,
+      "step": 15827
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 0.8091118367868632,
+      "learning_rate": 3.117688247698902e-06,
+      "loss": 12.1757,
+      "step": 15828
+    },
+    {
+      "epoch": 2.768174863387978,
+      "grad_norm": 0.8095319574062125,
+      "learning_rate": 3.113012167117857e-06,
+      "loss": 11.983,
+      "step": 15829
+    },
+    {
+      "epoch": 2.7683497267759565,
+      "grad_norm": 0.831392071482095,
+      "learning_rate": 3.108339540444727e-06,
+      "loss": 11.8709,
+      "step": 15830
+    },
+    {
+      "epoch": 2.7685245901639344,
+      "grad_norm": 0.829354874363949,
+      "learning_rate": 3.10367036784609e-06,
+      "loss": 12.0177,
+      "step": 15831
+    },
+    {
+      "epoch": 2.7686994535519127,
+      "grad_norm": 0.8077919966725857,
+      "learning_rate": 3.0990046494883794e-06,
+      "loss": 11.8834,
+      "step": 15832
+    },
+    {
+      "epoch": 2.7688743169398906,
+      "grad_norm": 0.8582398348501777,
+      "learning_rate": 3.0943423855379516e-06,
+      "loss": 11.8093,
+      "step": 15833
+    },
+    {
+      "epoch": 2.769049180327869,
+      "grad_norm": 0.7790012049079507,
+      "learning_rate": 3.0896835761609955e-06,
+      "loss": 11.6835,
+      "step": 15834
+    },
+    {
+      "epoch": 2.769224043715847,
+      "grad_norm": 0.8699684947088071,
+      "learning_rate": 3.085028221523578e-06,
+      "loss": 11.8534,
+      "step": 15835
+    },
+    {
+      "epoch": 2.769398907103825,
+      "grad_norm": 0.8736610433658558,
+      "learning_rate": 3.0803763217916782e-06,
+      "loss": 11.8842,
+      "step": 15836
+    },
+    {
+      "epoch": 2.769573770491803,
+      "grad_norm": 0.8361353688786757,
+      "learning_rate": 3.0757278771311294e-06,
+      "loss": 11.829,
+      "step": 15837
+    },
+    {
+      "epoch": 2.7697486338797814,
+      "grad_norm": 0.8970661187043558,
+      "learning_rate": 3.0710828877076325e-06,
+      "loss": 11.7903,
+      "step": 15838
+    },
+    {
+      "epoch": 2.7699234972677598,
+      "grad_norm": 0.8246980229310844,
+      "learning_rate": 3.0664413536867887e-06,
+      "loss": 11.4737,
+      "step": 15839
+    },
+    {
+      "epoch": 2.7700983606557377,
+      "grad_norm": 0.8119328869233998,
+      "learning_rate": 3.061803275234043e-06,
+      "loss": 11.8574,
+      "step": 15840
+    },
+    {
+      "epoch": 2.770273224043716,
+      "grad_norm": 0.7771157894508163,
+      "learning_rate": 3.0571686525147634e-06,
+      "loss": 11.8486,
+      "step": 15841
+    },
+    {
+      "epoch": 2.770448087431694,
+      "grad_norm": 0.8078086080955503,
+      "learning_rate": 3.0525374856941515e-06,
+      "loss": 11.5752,
+      "step": 15842
+    },
+    {
+      "epoch": 2.7706229508196722,
+      "grad_norm": 0.8023963227395093,
+      "learning_rate": 3.0479097749373078e-06,
+      "loss": 11.8489,
+      "step": 15843
+    },
+    {
+      "epoch": 2.77079781420765,
+      "grad_norm": 0.7857324978232005,
+      "learning_rate": 3.043285520409189e-06,
+      "loss": 11.8778,
+      "step": 15844
+    },
+    {
+      "epoch": 2.7709726775956285,
+      "grad_norm": 0.8261504528256981,
+      "learning_rate": 3.0386647222746754e-06,
+      "loss": 11.8426,
+      "step": 15845
+    },
+    {
+      "epoch": 2.7711475409836064,
+      "grad_norm": 0.8361462304134022,
+      "learning_rate": 3.034047380698468e-06,
+      "loss": 12.0725,
+      "step": 15846
+    },
+    {
+      "epoch": 2.7713224043715847,
+      "grad_norm": 0.8231537077950044,
+      "learning_rate": 3.029433495845191e-06,
+      "loss": 11.6844,
+      "step": 15847
+    },
+    {
+      "epoch": 2.771497267759563,
+      "grad_norm": 0.813330219310228,
+      "learning_rate": 3.0248230678793123e-06,
+      "loss": 11.6379,
+      "step": 15848
+    },
+    {
+      "epoch": 2.771672131147541,
+      "grad_norm": 0.8610222037642199,
+      "learning_rate": 3.0202160969651782e-06,
+      "loss": 11.8464,
+      "step": 15849
+    },
+    {
+      "epoch": 2.771846994535519,
+      "grad_norm": 0.8331003249784005,
+      "learning_rate": 3.0156125832670356e-06,
+      "loss": 11.6293,
+      "step": 15850
+    },
+    {
+      "epoch": 2.772021857923497,
+      "grad_norm": 0.8945231678170419,
+      "learning_rate": 3.0110125269489865e-06,
+      "loss": 11.7628,
+      "step": 15851
+    },
+    {
+      "epoch": 2.7721967213114755,
+      "grad_norm": 0.8555898761245053,
+      "learning_rate": 3.0064159281750325e-06,
+      "loss": 11.9531,
+      "step": 15852
+    },
+    {
+      "epoch": 2.7723715846994534,
+      "grad_norm": 0.8445641310610017,
+      "learning_rate": 3.001822787109021e-06,
+      "loss": 11.9215,
+      "step": 15853
+    },
+    {
+      "epoch": 2.772546448087432,
+      "grad_norm": 0.8318728201725039,
+      "learning_rate": 2.9972331039146874e-06,
+      "loss": 11.7401,
+      "step": 15854
+    },
+    {
+      "epoch": 2.7727213114754097,
+      "grad_norm": 0.8905090796811007,
+      "learning_rate": 2.9926468787556673e-06,
+      "loss": 11.8459,
+      "step": 15855
+    },
+    {
+      "epoch": 2.772896174863388,
+      "grad_norm": 0.8519121413881368,
+      "learning_rate": 2.988064111795441e-06,
+      "loss": 11.8805,
+      "step": 15856
+    },
+    {
+      "epoch": 2.7730710382513664,
+      "grad_norm": 0.8248359626114022,
+      "learning_rate": 2.9834848031973895e-06,
+      "loss": 11.7842,
+      "step": 15857
+    },
+    {
+      "epoch": 2.7732459016393443,
+      "grad_norm": 0.8119263861966811,
+      "learning_rate": 2.978908953124748e-06,
+      "loss": 11.7647,
+      "step": 15858
+    },
+    {
+      "epoch": 2.773420765027322,
+      "grad_norm": 0.7668595347362812,
+      "learning_rate": 2.9743365617406427e-06,
+      "loss": 12.0398,
+      "step": 15859
+    },
+    {
+      "epoch": 2.7735956284153005,
+      "grad_norm": 0.8641800600998749,
+      "learning_rate": 2.9697676292080755e-06,
+      "loss": 11.6499,
+      "step": 15860
+    },
+    {
+      "epoch": 2.773770491803279,
+      "grad_norm": 0.8245125315646578,
+      "learning_rate": 2.9652021556899277e-06,
+      "loss": 11.9417,
+      "step": 15861
+    },
+    {
+      "epoch": 2.7739453551912567,
+      "grad_norm": 0.8398232529111007,
+      "learning_rate": 2.960640141348947e-06,
+      "loss": 11.7868,
+      "step": 15862
+    },
+    {
+      "epoch": 2.774120218579235,
+      "grad_norm": 0.8531438841081757,
+      "learning_rate": 2.956081586347781e-06,
+      "loss": 11.9383,
+      "step": 15863
+    },
+    {
+      "epoch": 2.774295081967213,
+      "grad_norm": 0.8205223505335102,
+      "learning_rate": 2.9515264908489106e-06,
+      "loss": 11.818,
+      "step": 15864
+    },
+    {
+      "epoch": 2.7744699453551913,
+      "grad_norm": 0.8179877175198983,
+      "learning_rate": 2.946974855014728e-06,
+      "loss": 12.0121,
+      "step": 15865
+    },
+    {
+      "epoch": 2.7746448087431697,
+      "grad_norm": 0.7771656603220166,
+      "learning_rate": 2.9424266790075037e-06,
+      "loss": 12.0102,
+      "step": 15866
+    },
+    {
+      "epoch": 2.7748196721311476,
+      "grad_norm": 0.7777236949669907,
+      "learning_rate": 2.937881962989375e-06,
+      "loss": 11.8562,
+      "step": 15867
+    },
+    {
+      "epoch": 2.7749945355191254,
+      "grad_norm": 0.8690844736368617,
+      "learning_rate": 2.933340707122334e-06,
+      "loss": 11.7773,
+      "step": 15868
+    },
+    {
+      "epoch": 2.775169398907104,
+      "grad_norm": 0.8262225581793124,
+      "learning_rate": 2.928802911568296e-06,
+      "loss": 11.6883,
+      "step": 15869
+    },
+    {
+      "epoch": 2.775344262295082,
+      "grad_norm": 0.8318492884185524,
+      "learning_rate": 2.9242685764890197e-06,
+      "loss": 11.5465,
+      "step": 15870
+    },
+    {
+      "epoch": 2.77551912568306,
+      "grad_norm": 0.8395463081823653,
+      "learning_rate": 2.9197377020461548e-06,
+      "loss": 11.87,
+      "step": 15871
+    },
+    {
+      "epoch": 2.7756939890710384,
+      "grad_norm": 0.91706196366321,
+      "learning_rate": 2.9152102884012045e-06,
+      "loss": 11.947,
+      "step": 15872
+    },
+    {
+      "epoch": 2.7758688524590163,
+      "grad_norm": 0.830080901600782,
+      "learning_rate": 2.9106863357155735e-06,
+      "loss": 11.9151,
+      "step": 15873
+    },
+    {
+      "epoch": 2.7760437158469946,
+      "grad_norm": 0.8703387187425448,
+      "learning_rate": 2.9061658441505436e-06,
+      "loss": 11.7113,
+      "step": 15874
+    },
+    {
+      "epoch": 2.776218579234973,
+      "grad_norm": 0.8789822668587134,
+      "learning_rate": 2.9016488138672637e-06,
+      "loss": 11.8693,
+      "step": 15875
+    },
+    {
+      "epoch": 2.776393442622951,
+      "grad_norm": 0.79539929945557,
+      "learning_rate": 2.8971352450267497e-06,
+      "loss": 11.6736,
+      "step": 15876
+    },
+    {
+      "epoch": 2.7765683060109287,
+      "grad_norm": 0.8882832247301763,
+      "learning_rate": 2.892625137789906e-06,
+      "loss": 11.9423,
+      "step": 15877
+    },
+    {
+      "epoch": 2.776743169398907,
+      "grad_norm": 0.8790921646915029,
+      "learning_rate": 2.8881184923175265e-06,
+      "loss": 11.6861,
+      "step": 15878
+    },
+    {
+      "epoch": 2.7769180327868854,
+      "grad_norm": 0.8103979089355227,
+      "learning_rate": 2.883615308770249e-06,
+      "loss": 12.0084,
+      "step": 15879
+    },
+    {
+      "epoch": 2.7770928961748633,
+      "grad_norm": 0.8005020497795322,
+      "learning_rate": 2.879115587308623e-06,
+      "loss": 11.7186,
+      "step": 15880
+    },
+    {
+      "epoch": 2.7772677595628417,
+      "grad_norm": 0.7993360288840892,
+      "learning_rate": 2.8746193280930535e-06,
+      "loss": 11.9095,
+      "step": 15881
+    },
+    {
+      "epoch": 2.7774426229508196,
+      "grad_norm": 0.8655672157935358,
+      "learning_rate": 2.8701265312838455e-06,
+      "loss": 11.5472,
+      "step": 15882
+    },
+    {
+      "epoch": 2.777617486338798,
+      "grad_norm": 0.8817723471465082,
+      "learning_rate": 2.8656371970411157e-06,
+      "loss": 11.8925,
+      "step": 15883
+    },
+    {
+      "epoch": 2.777792349726776,
+      "grad_norm": 0.8195915607818478,
+      "learning_rate": 2.8611513255249467e-06,
+      "loss": 11.7897,
+      "step": 15884
+    },
+    {
+      "epoch": 2.777967213114754,
+      "grad_norm": 0.8169662269754809,
+      "learning_rate": 2.856668916895222e-06,
+      "loss": 12.0097,
+      "step": 15885
+    },
+    {
+      "epoch": 2.778142076502732,
+      "grad_norm": 0.848171587566898,
+      "learning_rate": 2.8521899713117697e-06,
+      "loss": 11.616,
+      "step": 15886
+    },
+    {
+      "epoch": 2.7783169398907104,
+      "grad_norm": 0.8981159596694044,
+      "learning_rate": 2.8477144889342277e-06,
+      "loss": 11.472,
+      "step": 15887
+    },
+    {
+      "epoch": 2.7784918032786887,
+      "grad_norm": 0.8054204688043444,
+      "learning_rate": 2.8432424699221584e-06,
+      "loss": 11.7219,
+      "step": 15888
+    },
+    {
+      "epoch": 2.7786666666666666,
+      "grad_norm": 0.820066640313244,
+      "learning_rate": 2.838773914434978e-06,
+      "loss": 11.9892,
+      "step": 15889
+    },
+    {
+      "epoch": 2.778841530054645,
+      "grad_norm": 0.8274821357215576,
+      "learning_rate": 2.8343088226319926e-06,
+      "loss": 11.8215,
+      "step": 15890
+    },
+    {
+      "epoch": 2.779016393442623,
+      "grad_norm": 0.8484913887992345,
+      "learning_rate": 2.8298471946723636e-06,
+      "loss": 11.8722,
+      "step": 15891
+    },
+    {
+      "epoch": 2.779191256830601,
+      "grad_norm": 0.829414007486459,
+      "learning_rate": 2.825389030715153e-06,
+      "loss": 11.7698,
+      "step": 15892
+    },
+    {
+      "epoch": 2.779366120218579,
+      "grad_norm": 0.802337310793998,
+      "learning_rate": 2.8209343309192893e-06,
+      "loss": 11.9,
+      "step": 15893
+    },
+    {
+      "epoch": 2.7795409836065574,
+      "grad_norm": 0.8245806953910222,
+      "learning_rate": 2.8164830954435783e-06,
+      "loss": 11.4619,
+      "step": 15894
+    },
+    {
+      "epoch": 2.7797158469945353,
+      "grad_norm": 0.8570679334935982,
+      "learning_rate": 2.8120353244466936e-06,
+      "loss": 11.8995,
+      "step": 15895
+    },
+    {
+      "epoch": 2.7798907103825137,
+      "grad_norm": 0.8598450589320616,
+      "learning_rate": 2.807591018087197e-06,
+      "loss": 11.7598,
+      "step": 15896
+    },
+    {
+      "epoch": 2.780065573770492,
+      "grad_norm": 0.803748172467799,
+      "learning_rate": 2.803150176523528e-06,
+      "loss": 11.9441,
+      "step": 15897
+    },
+    {
+      "epoch": 2.78024043715847,
+      "grad_norm": 0.8087125418713298,
+      "learning_rate": 2.798712799913983e-06,
+      "loss": 11.6567,
+      "step": 15898
+    },
+    {
+      "epoch": 2.780415300546448,
+      "grad_norm": 0.8213708569384496,
+      "learning_rate": 2.794278888416757e-06,
+      "loss": 11.7879,
+      "step": 15899
+    },
+    {
+      "epoch": 2.780590163934426,
+      "grad_norm": 0.8486441409634515,
+      "learning_rate": 2.789848442189924e-06,
+      "loss": 11.8264,
+      "step": 15900
+    },
+    {
+      "epoch": 2.7807650273224045,
+      "grad_norm": 0.8825394253709611,
+      "learning_rate": 2.7854214613914244e-06,
+      "loss": 11.7617,
+      "step": 15901
+    },
+    {
+      "epoch": 2.7809398907103824,
+      "grad_norm": 0.8594559552715696,
+      "learning_rate": 2.7809979461790646e-06,
+      "loss": 12.0251,
+      "step": 15902
+    },
+    {
+      "epoch": 2.7811147540983607,
+      "grad_norm": 0.7809780552794297,
+      "learning_rate": 2.7765778967105303e-06,
+      "loss": 12.1016,
+      "step": 15903
+    },
+    {
+      "epoch": 2.7812896174863386,
+      "grad_norm": 0.8196596418274841,
+      "learning_rate": 2.7721613131433955e-06,
+      "loss": 11.7268,
+      "step": 15904
+    },
+    {
+      "epoch": 2.781464480874317,
+      "grad_norm": 0.8631312841233644,
+      "learning_rate": 2.7677481956351335e-06,
+      "loss": 11.7299,
+      "step": 15905
+    },
+    {
+      "epoch": 2.7816393442622953,
+      "grad_norm": 0.8507769654284738,
+      "learning_rate": 2.7633385443430305e-06,
+      "loss": 11.8783,
+      "step": 15906
+    },
+    {
+      "epoch": 2.781814207650273,
+      "grad_norm": 0.8420237505575894,
+      "learning_rate": 2.7589323594242933e-06,
+      "loss": 11.922,
+      "step": 15907
+    },
+    {
+      "epoch": 2.781989071038251,
+      "grad_norm": 1.0008285415108356,
+      "learning_rate": 2.7545296410360077e-06,
+      "loss": 11.9015,
+      "step": 15908
+    },
+    {
+      "epoch": 2.7821639344262294,
+      "grad_norm": 0.7880524960039292,
+      "learning_rate": 2.7501303893351372e-06,
+      "loss": 11.9109,
+      "step": 15909
+    },
+    {
+      "epoch": 2.782338797814208,
+      "grad_norm": 0.8599317743301529,
+      "learning_rate": 2.7457346044784783e-06,
+      "loss": 11.542,
+      "step": 15910
+    },
+    {
+      "epoch": 2.7825136612021857,
+      "grad_norm": 0.7320123352656867,
+      "learning_rate": 2.7413422866227503e-06,
+      "loss": 11.5555,
+      "step": 15911
+    },
+    {
+      "epoch": 2.782688524590164,
+      "grad_norm": 0.8433365383534115,
+      "learning_rate": 2.7369534359245496e-06,
+      "loss": 11.6056,
+      "step": 15912
+    },
+    {
+      "epoch": 2.782863387978142,
+      "grad_norm": 0.8852476687914703,
+      "learning_rate": 2.7325680525403075e-06,
+      "loss": 11.7937,
+      "step": 15913
+    },
+    {
+      "epoch": 2.7830382513661203,
+      "grad_norm": 0.8471405687743471,
+      "learning_rate": 2.728186136626365e-06,
+      "loss": 11.9424,
+      "step": 15914
+    },
+    {
+      "epoch": 2.7832131147540986,
+      "grad_norm": 0.7983315364072125,
+      "learning_rate": 2.723807688338942e-06,
+      "loss": 11.7553,
+      "step": 15915
+    },
+    {
+      "epoch": 2.7833879781420765,
+      "grad_norm": 0.9254307846276733,
+      "learning_rate": 2.719432707834113e-06,
+      "loss": 11.7815,
+      "step": 15916
+    },
+    {
+      "epoch": 2.7835628415300544,
+      "grad_norm": 0.7880093217433354,
+      "learning_rate": 2.715061195267854e-06,
+      "loss": 11.9612,
+      "step": 15917
+    },
+    {
+      "epoch": 2.7837377049180327,
+      "grad_norm": 0.8432261955134575,
+      "learning_rate": 2.7106931507959955e-06,
+      "loss": 11.5592,
+      "step": 15918
+    },
+    {
+      "epoch": 2.783912568306011,
+      "grad_norm": 0.7919104163140571,
+      "learning_rate": 2.7063285745742573e-06,
+      "loss": 11.7868,
+      "step": 15919
+    },
+    {
+      "epoch": 2.784087431693989,
+      "grad_norm": 0.8378156074061506,
+      "learning_rate": 2.7019674667582483e-06,
+      "loss": 12.0269,
+      "step": 15920
+    },
+    {
+      "epoch": 2.7842622950819673,
+      "grad_norm": 0.8501612505847531,
+      "learning_rate": 2.6976098275034e-06,
+      "loss": 11.8228,
+      "step": 15921
+    },
+    {
+      "epoch": 2.784437158469945,
+      "grad_norm": 0.8683145407651476,
+      "learning_rate": 2.693255656965077e-06,
+      "loss": 11.7687,
+      "step": 15922
+    },
+    {
+      "epoch": 2.7846120218579236,
+      "grad_norm": 0.8225882165859051,
+      "learning_rate": 2.688904955298488e-06,
+      "loss": 11.5747,
+      "step": 15923
+    },
+    {
+      "epoch": 2.784786885245902,
+      "grad_norm": 0.8648683502322105,
+      "learning_rate": 2.684557722658765e-06,
+      "loss": 11.9178,
+      "step": 15924
+    },
+    {
+      "epoch": 2.78496174863388,
+      "grad_norm": 0.8854930276412775,
+      "learning_rate": 2.6802139592008503e-06,
+      "loss": 11.8053,
+      "step": 15925
+    },
+    {
+      "epoch": 2.7851366120218577,
+      "grad_norm": 0.8552383606647871,
+      "learning_rate": 2.675873665079598e-06,
+      "loss": 11.9139,
+      "step": 15926
+    },
+    {
+      "epoch": 2.785311475409836,
+      "grad_norm": 0.8723691593909138,
+      "learning_rate": 2.6715368404497397e-06,
+      "loss": 11.9169,
+      "step": 15927
+    },
+    {
+      "epoch": 2.7854863387978144,
+      "grad_norm": 0.8962193585219099,
+      "learning_rate": 2.6672034854658744e-06,
+      "loss": 11.9002,
+      "step": 15928
+    },
+    {
+      "epoch": 2.7856612021857923,
+      "grad_norm": 0.8732244471417562,
+      "learning_rate": 2.6628736002824896e-06,
+      "loss": 11.8903,
+      "step": 15929
+    },
+    {
+      "epoch": 2.7858360655737706,
+      "grad_norm": 0.78631905215612,
+      "learning_rate": 2.658547185053928e-06,
+      "loss": 12.0322,
+      "step": 15930
+    },
+    {
+      "epoch": 2.7860109289617485,
+      "grad_norm": 0.8545727783985294,
+      "learning_rate": 2.654224239934433e-06,
+      "loss": 11.7287,
+      "step": 15931
+    },
+    {
+      "epoch": 2.786185792349727,
+      "grad_norm": 0.850916139318431,
+      "learning_rate": 2.649904765078104e-06,
+      "loss": 11.9364,
+      "step": 15932
+    },
+    {
+      "epoch": 2.786360655737705,
+      "grad_norm": 0.8946671425925711,
+      "learning_rate": 2.6455887606389284e-06,
+      "loss": 11.5055,
+      "step": 15933
+    },
+    {
+      "epoch": 2.786535519125683,
+      "grad_norm": 0.8445213602469748,
+      "learning_rate": 2.6412762267707724e-06,
+      "loss": 11.9335,
+      "step": 15934
+    },
+    {
+      "epoch": 2.786710382513661,
+      "grad_norm": 0.8238381769151182,
+      "learning_rate": 2.6369671636273574e-06,
+      "loss": 11.7951,
+      "step": 15935
+    },
+    {
+      "epoch": 2.7868852459016393,
+      "grad_norm": 0.9240748050606565,
+      "learning_rate": 2.632661571362316e-06,
+      "loss": 11.8332,
+      "step": 15936
+    },
+    {
+      "epoch": 2.7870601092896177,
+      "grad_norm": 0.7838865906595591,
+      "learning_rate": 2.6283594501291253e-06,
+      "loss": 12.0149,
+      "step": 15937
+    },
+    {
+      "epoch": 2.7872349726775956,
+      "grad_norm": 0.8248377324068142,
+      "learning_rate": 2.6240608000811408e-06,
+      "loss": 11.6838,
+      "step": 15938
+    },
+    {
+      "epoch": 2.787409836065574,
+      "grad_norm": 0.86462251388068,
+      "learning_rate": 2.619765621371639e-06,
+      "loss": 11.7154,
+      "step": 15939
+    },
+    {
+      "epoch": 2.787584699453552,
+      "grad_norm": 0.8202354820954432,
+      "learning_rate": 2.61547391415371e-06,
+      "loss": 11.8311,
+      "step": 15940
+    },
+    {
+      "epoch": 2.78775956284153,
+      "grad_norm": 0.7768878096838502,
+      "learning_rate": 2.6111856785803523e-06,
+      "loss": 11.8516,
+      "step": 15941
+    },
+    {
+      "epoch": 2.787934426229508,
+      "grad_norm": 0.8161009642389877,
+      "learning_rate": 2.6069009148044442e-06,
+      "loss": 11.6902,
+      "step": 15942
+    },
+    {
+      "epoch": 2.7881092896174864,
+      "grad_norm": 0.8254739761060299,
+      "learning_rate": 2.602619622978719e-06,
+      "loss": 11.7646,
+      "step": 15943
+    },
+    {
+      "epoch": 2.7882841530054643,
+      "grad_norm": 0.803702312468237,
+      "learning_rate": 2.598341803255799e-06,
+      "loss": 11.7013,
+      "step": 15944
+    },
+    {
+      "epoch": 2.7884590163934426,
+      "grad_norm": 0.8264331862107867,
+      "learning_rate": 2.594067455788207e-06,
+      "loss": 11.7955,
+      "step": 15945
+    },
+    {
+      "epoch": 2.788633879781421,
+      "grad_norm": 0.8736998463552076,
+      "learning_rate": 2.58979658072831e-06,
+      "loss": 11.8223,
+      "step": 15946
+    },
+    {
+      "epoch": 2.788808743169399,
+      "grad_norm": 0.8061538733594772,
+      "learning_rate": 2.58552917822833e-06,
+      "loss": 11.9632,
+      "step": 15947
+    },
+    {
+      "epoch": 2.788983606557377,
+      "grad_norm": 0.8303372921206827,
+      "learning_rate": 2.581265248440434e-06,
+      "loss": 11.9534,
+      "step": 15948
+    },
+    {
+      "epoch": 2.789158469945355,
+      "grad_norm": 0.7764614643402344,
+      "learning_rate": 2.5770047915166018e-06,
+      "loss": 11.7908,
+      "step": 15949
+    },
+    {
+      "epoch": 2.7893333333333334,
+      "grad_norm": 0.8888102005201515,
+      "learning_rate": 2.5727478076087218e-06,
+      "loss": 11.6795,
+      "step": 15950
+    },
+    {
+      "epoch": 2.7895081967213113,
+      "grad_norm": 0.7656921390078608,
+      "learning_rate": 2.568494296868573e-06,
+      "loss": 11.8924,
+      "step": 15951
+    },
+    {
+      "epoch": 2.7896830601092897,
+      "grad_norm": 0.805047133599811,
+      "learning_rate": 2.5642442594477455e-06,
+      "loss": 12.0187,
+      "step": 15952
+    },
+    {
+      "epoch": 2.7898579234972676,
+      "grad_norm": 0.8136377423176758,
+      "learning_rate": 2.559997695497762e-06,
+      "loss": 11.8241,
+      "step": 15953
+    },
+    {
+      "epoch": 2.790032786885246,
+      "grad_norm": 0.835057498896639,
+      "learning_rate": 2.555754605170024e-06,
+      "loss": 11.9162,
+      "step": 15954
+    },
+    {
+      "epoch": 2.7902076502732243,
+      "grad_norm": 0.8432455555169807,
+      "learning_rate": 2.5515149886157885e-06,
+      "loss": 12.0503,
+      "step": 15955
+    },
+    {
+      "epoch": 2.790382513661202,
+      "grad_norm": 0.7818353855349132,
+      "learning_rate": 2.5472788459861784e-06,
+      "loss": 11.8054,
+      "step": 15956
+    },
+    {
+      "epoch": 2.79055737704918,
+      "grad_norm": 0.808468486200658,
+      "learning_rate": 2.543046177432229e-06,
+      "loss": 11.783,
+      "step": 15957
+    },
+    {
+      "epoch": 2.7907322404371584,
+      "grad_norm": 0.8267849127896787,
+      "learning_rate": 2.538816983104808e-06,
+      "loss": 11.7214,
+      "step": 15958
+    },
+    {
+      "epoch": 2.7909071038251367,
+      "grad_norm": 0.8502375211711952,
+      "learning_rate": 2.5345912631546843e-06,
+      "loss": 12.0357,
+      "step": 15959
+    },
+    {
+      "epoch": 2.7910819672131146,
+      "grad_norm": 0.8310499790591078,
+      "learning_rate": 2.5303690177325146e-06,
+      "loss": 11.9339,
+      "step": 15960
+    },
+    {
+      "epoch": 2.791256830601093,
+      "grad_norm": 0.8878063987796794,
+      "learning_rate": 2.526150246988812e-06,
+      "loss": 12.0527,
+      "step": 15961
+    },
+    {
+      "epoch": 2.791431693989071,
+      "grad_norm": 0.8061781635908071,
+      "learning_rate": 2.5219349510739563e-06,
+      "loss": 11.809,
+      "step": 15962
+    },
+    {
+      "epoch": 2.791606557377049,
+      "grad_norm": 0.760689349225026,
+      "learning_rate": 2.5177231301382276e-06,
+      "loss": 11.7818,
+      "step": 15963
+    },
+    {
+      "epoch": 2.7917814207650276,
+      "grad_norm": 0.8109361358411292,
+      "learning_rate": 2.513514784331783e-06,
+      "loss": 11.6721,
+      "step": 15964
+    },
+    {
+      "epoch": 2.7919562841530055,
+      "grad_norm": 0.8336669179276373,
+      "learning_rate": 2.509309913804636e-06,
+      "loss": 11.4841,
+      "step": 15965
+    },
+    {
+      "epoch": 2.7921311475409833,
+      "grad_norm": 0.8300527681541179,
+      "learning_rate": 2.505108518706678e-06,
+      "loss": 11.4255,
+      "step": 15966
+    },
+    {
+      "epoch": 2.7923060109289617,
+      "grad_norm": 0.8305164931425393,
+      "learning_rate": 2.5009105991877e-06,
+      "loss": 11.8727,
+      "step": 15967
+    },
+    {
+      "epoch": 2.79248087431694,
+      "grad_norm": 0.7808659882677655,
+      "learning_rate": 2.4967161553973275e-06,
+      "loss": 11.7354,
+      "step": 15968
+    },
+    {
+      "epoch": 2.792655737704918,
+      "grad_norm": 0.8467962975900415,
+      "learning_rate": 2.4925251874851176e-06,
+      "loss": 11.7932,
+      "step": 15969
+    },
+    {
+      "epoch": 2.7928306010928963,
+      "grad_norm": 0.7981084646828388,
+      "learning_rate": 2.488337695600451e-06,
+      "loss": 11.8614,
+      "step": 15970
+    },
+    {
+      "epoch": 2.793005464480874,
+      "grad_norm": 0.7965107815017316,
+      "learning_rate": 2.4841536798926203e-06,
+      "loss": 12.0497,
+      "step": 15971
+    },
+    {
+      "epoch": 2.7931803278688525,
+      "grad_norm": 0.8472164719099183,
+      "learning_rate": 2.479973140510783e-06,
+      "loss": 11.4684,
+      "step": 15972
+    },
+    {
+      "epoch": 2.793355191256831,
+      "grad_norm": 0.8506766976368983,
+      "learning_rate": 2.4757960776039537e-06,
+      "loss": 11.8124,
+      "step": 15973
+    },
+    {
+      "epoch": 2.7935300546448087,
+      "grad_norm": 0.8024021499556909,
+      "learning_rate": 2.4716224913210464e-06,
+      "loss": 11.5224,
+      "step": 15974
+    },
+    {
+      "epoch": 2.7937049180327866,
+      "grad_norm": 0.9153367370763295,
+      "learning_rate": 2.467452381810853e-06,
+      "loss": 11.8976,
+      "step": 15975
+    },
+    {
+      "epoch": 2.793879781420765,
+      "grad_norm": 0.7956822018910218,
+      "learning_rate": 2.463285749222033e-06,
+      "loss": 11.915,
+      "step": 15976
+    },
+    {
+      "epoch": 2.7940546448087433,
+      "grad_norm": 0.8527272782019509,
+      "learning_rate": 2.4591225937031115e-06,
+      "loss": 11.9158,
+      "step": 15977
+    },
+    {
+      "epoch": 2.7942295081967212,
+      "grad_norm": 0.8168185769612555,
+      "learning_rate": 2.4549629154025033e-06,
+      "loss": 11.8033,
+      "step": 15978
+    },
+    {
+      "epoch": 2.7944043715846996,
+      "grad_norm": 0.7876676183993446,
+      "learning_rate": 2.450806714468501e-06,
+      "loss": 11.8806,
+      "step": 15979
+    },
+    {
+      "epoch": 2.7945792349726775,
+      "grad_norm": 0.8755144979737546,
+      "learning_rate": 2.4466539910492637e-06,
+      "loss": 12.0355,
+      "step": 15980
+    },
+    {
+      "epoch": 2.794754098360656,
+      "grad_norm": 0.7750396467202774,
+      "learning_rate": 2.4425047452928397e-06,
+      "loss": 11.7088,
+      "step": 15981
+    },
+    {
+      "epoch": 2.794928961748634,
+      "grad_norm": 0.793117869247311,
+      "learning_rate": 2.4383589773471217e-06,
+      "loss": 11.7189,
+      "step": 15982
+    },
+    {
+      "epoch": 2.795103825136612,
+      "grad_norm": 0.789964771929561,
+      "learning_rate": 2.4342166873599246e-06,
+      "loss": 12.0006,
+      "step": 15983
+    },
+    {
+      "epoch": 2.79527868852459,
+      "grad_norm": 0.8669727289672997,
+      "learning_rate": 2.4300778754789198e-06,
+      "loss": 11.6496,
+      "step": 15984
+    },
+    {
+      "epoch": 2.7954535519125683,
+      "grad_norm": 0.8605361078343361,
+      "learning_rate": 2.4259425418516336e-06,
+      "loss": 11.9468,
+      "step": 15985
+    },
+    {
+      "epoch": 2.7956284153005466,
+      "grad_norm": 0.7332788928577162,
+      "learning_rate": 2.4218106866254918e-06,
+      "loss": 11.5637,
+      "step": 15986
+    },
+    {
+      "epoch": 2.7958032786885245,
+      "grad_norm": 0.8163302214880888,
+      "learning_rate": 2.4176823099477886e-06,
+      "loss": 11.8645,
+      "step": 15987
+    },
+    {
+      "epoch": 2.795978142076503,
+      "grad_norm": 0.8145129919997155,
+      "learning_rate": 2.4135574119656945e-06,
+      "loss": 11.9181,
+      "step": 15988
+    },
+    {
+      "epoch": 2.7961530054644808,
+      "grad_norm": 0.8444468329431297,
+      "learning_rate": 2.40943599282627e-06,
+      "loss": 11.6446,
+      "step": 15989
+    },
+    {
+      "epoch": 2.796327868852459,
+      "grad_norm": 0.8210683389200903,
+      "learning_rate": 2.4053180526764197e-06,
+      "loss": 11.9543,
+      "step": 15990
+    },
+    {
+      "epoch": 2.796502732240437,
+      "grad_norm": 0.78085963947175,
+      "learning_rate": 2.4012035916629706e-06,
+      "loss": 11.7156,
+      "step": 15991
+    },
+    {
+      "epoch": 2.7966775956284153,
+      "grad_norm": 0.8126780237119786,
+      "learning_rate": 2.397092609932572e-06,
+      "loss": 11.538,
+      "step": 15992
+    },
+    {
+      "epoch": 2.7968524590163932,
+      "grad_norm": 0.9568448132292574,
+      "learning_rate": 2.3929851076317846e-06,
+      "loss": 11.8216,
+      "step": 15993
+    },
+    {
+      "epoch": 2.7970273224043716,
+      "grad_norm": 0.857080686245073,
+      "learning_rate": 2.388881084907035e-06,
+      "loss": 11.788,
+      "step": 15994
+    },
+    {
+      "epoch": 2.79720218579235,
+      "grad_norm": 0.8356960962143861,
+      "learning_rate": 2.384780541904641e-06,
+      "loss": 11.7993,
+      "step": 15995
+    },
+    {
+      "epoch": 2.797377049180328,
+      "grad_norm": 0.8411580618107606,
+      "learning_rate": 2.380683478770751e-06,
+      "loss": 11.9383,
+      "step": 15996
+    },
+    {
+      "epoch": 2.797551912568306,
+      "grad_norm": 0.8680996813336924,
+      "learning_rate": 2.376589895651449e-06,
+      "loss": 11.6542,
+      "step": 15997
+    },
+    {
+      "epoch": 2.797726775956284,
+      "grad_norm": 0.859452285161187,
+      "learning_rate": 2.372499792692651e-06,
+      "loss": 11.9463,
+      "step": 15998
+    },
+    {
+      "epoch": 2.7979016393442624,
+      "grad_norm": 0.813499527435541,
+      "learning_rate": 2.3684131700401847e-06,
+      "loss": 11.7988,
+      "step": 15999
+    },
+    {
+      "epoch": 2.7980765027322403,
+      "grad_norm": 0.87759863243467,
+      "learning_rate": 2.3643300278397008e-06,
+      "loss": 11.6554,
+      "step": 16000
+    },
+    {
+      "epoch": 2.7982513661202186,
+      "grad_norm": 0.8898645304127641,
+      "learning_rate": 2.3602503662367824e-06,
+      "loss": 11.6476,
+      "step": 16001
+    },
+    {
+      "epoch": 2.7984262295081965,
+      "grad_norm": 0.8595013944732536,
+      "learning_rate": 2.3561741853768583e-06,
+      "loss": 11.8787,
+      "step": 16002
+    },
+    {
+      "epoch": 2.798601092896175,
+      "grad_norm": 0.913976788405632,
+      "learning_rate": 2.352101485405245e-06,
+      "loss": 11.8118,
+      "step": 16003
+    },
+    {
+      "epoch": 2.798775956284153,
+      "grad_norm": 0.8118346169106508,
+      "learning_rate": 2.348032266467115e-06,
+      "loss": 11.7507,
+      "step": 16004
+    },
+    {
+      "epoch": 2.798950819672131,
+      "grad_norm": 0.8197015543656457,
+      "learning_rate": 2.343966528707553e-06,
+      "loss": 11.848,
+      "step": 16005
+    },
+    {
+      "epoch": 2.7991256830601094,
+      "grad_norm": 0.8415656511243589,
+      "learning_rate": 2.339904272271476e-06,
+      "loss": 11.4788,
+      "step": 16006
+    },
+    {
+      "epoch": 2.7993005464480873,
+      "grad_norm": 0.7762269167849881,
+      "learning_rate": 2.3358454973037015e-06,
+      "loss": 11.6009,
+      "step": 16007
+    },
+    {
+      "epoch": 2.7994754098360657,
+      "grad_norm": 0.8032409806324273,
+      "learning_rate": 2.331790203948936e-06,
+      "loss": 11.9421,
+      "step": 16008
+    },
+    {
+      "epoch": 2.7996502732240436,
+      "grad_norm": 0.839467167459028,
+      "learning_rate": 2.3277383923517415e-06,
+      "loss": 11.7531,
+      "step": 16009
+    },
+    {
+      "epoch": 2.799825136612022,
+      "grad_norm": 0.8651030843455862,
+      "learning_rate": 2.323690062656547e-06,
+      "loss": 11.6541,
+      "step": 16010
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8553659349932619,
+      "learning_rate": 2.3196452150076817e-06,
+      "loss": 11.9588,
+      "step": 16011
+    },
+    {
+      "epoch": 2.800174863387978,
+      "grad_norm": 0.8383856415915126,
+      "learning_rate": 2.3156038495493413e-06,
+      "loss": 11.672,
+      "step": 16012
+    },
+    {
+      "epoch": 2.8003497267759565,
+      "grad_norm": 0.7698419669523264,
+      "learning_rate": 2.3115659664255883e-06,
+      "loss": 11.4622,
+      "step": 16013
+    },
+    {
+      "epoch": 2.8005245901639344,
+      "grad_norm": 0.8195099108364723,
+      "learning_rate": 2.3075315657803744e-06,
+      "loss": 11.6811,
+      "step": 16014
+    },
+    {
+      "epoch": 2.8006994535519123,
+      "grad_norm": 0.811735470608096,
+      "learning_rate": 2.303500647757517e-06,
+      "loss": 12.0643,
+      "step": 16015
+    },
+    {
+      "epoch": 2.8008743169398906,
+      "grad_norm": 0.8097331354350444,
+      "learning_rate": 2.2994732125007134e-06,
+      "loss": 11.865,
+      "step": 16016
+    },
+    {
+      "epoch": 2.801049180327869,
+      "grad_norm": 0.8342123926768905,
+      "learning_rate": 2.295449260153537e-06,
+      "loss": 11.9836,
+      "step": 16017
+    },
+    {
+      "epoch": 2.801224043715847,
+      "grad_norm": 0.7871533057140528,
+      "learning_rate": 2.291428790859451e-06,
+      "loss": 11.9762,
+      "step": 16018
+    },
+    {
+      "epoch": 2.801398907103825,
+      "grad_norm": 0.8329570478117173,
+      "learning_rate": 2.2874118047617634e-06,
+      "loss": 11.67,
+      "step": 16019
+    },
+    {
+      "epoch": 2.801573770491803,
+      "grad_norm": 0.8596987695333514,
+      "learning_rate": 2.2833983020036698e-06,
+      "loss": 12.0297,
+      "step": 16020
+    },
+    {
+      "epoch": 2.8017486338797815,
+      "grad_norm": 0.7464967839559903,
+      "learning_rate": 2.2793882827282787e-06,
+      "loss": 11.7358,
+      "step": 16021
+    },
+    {
+      "epoch": 2.80192349726776,
+      "grad_norm": 0.805909144328293,
+      "learning_rate": 2.2753817470784977e-06,
+      "loss": 12.0586,
+      "step": 16022
+    },
+    {
+      "epoch": 2.8020983606557377,
+      "grad_norm": 0.8495220443512524,
+      "learning_rate": 2.2713786951971904e-06,
+      "loss": 11.8711,
+      "step": 16023
+    },
+    {
+      "epoch": 2.8022732240437156,
+      "grad_norm": 0.8507173679457443,
+      "learning_rate": 2.267379127227043e-06,
+      "loss": 12.0714,
+      "step": 16024
+    },
+    {
+      "epoch": 2.802448087431694,
+      "grad_norm": 0.8566432369154385,
+      "learning_rate": 2.2633830433106407e-06,
+      "loss": 11.8464,
+      "step": 16025
+    },
+    {
+      "epoch": 2.8026229508196723,
+      "grad_norm": 0.8885571825240092,
+      "learning_rate": 2.2593904435904367e-06,
+      "loss": 11.6796,
+      "step": 16026
+    },
+    {
+      "epoch": 2.80279781420765,
+      "grad_norm": 0.7985536479211449,
+      "learning_rate": 2.2554013282087726e-06,
+      "loss": 11.9524,
+      "step": 16027
+    },
+    {
+      "epoch": 2.8029726775956285,
+      "grad_norm": 0.7678408630325202,
+      "learning_rate": 2.251415697307846e-06,
+      "loss": 11.9461,
+      "step": 16028
+    },
+    {
+      "epoch": 2.8031475409836064,
+      "grad_norm": 0.8222230897941994,
+      "learning_rate": 2.247433551029754e-06,
+      "loss": 11.9893,
+      "step": 16029
+    },
+    {
+      "epoch": 2.8033224043715848,
+      "grad_norm": 0.8288723914485929,
+      "learning_rate": 2.2434548895164385e-06,
+      "loss": 11.8886,
+      "step": 16030
+    },
+    {
+      "epoch": 2.803497267759563,
+      "grad_norm": 0.7914722387343239,
+      "learning_rate": 2.239479712909731e-06,
+      "loss": 11.8348,
+      "step": 16031
+    },
+    {
+      "epoch": 2.803672131147541,
+      "grad_norm": 0.8454761577030359,
+      "learning_rate": 2.235508021351351e-06,
+      "loss": 11.9492,
+      "step": 16032
+    },
+    {
+      "epoch": 2.803846994535519,
+      "grad_norm": 0.8357242533897232,
+      "learning_rate": 2.231539814982908e-06,
+      "loss": 11.9369,
+      "step": 16033
+    },
+    {
+      "epoch": 2.8040218579234972,
+      "grad_norm": 0.8223108445095038,
+      "learning_rate": 2.2275750939458108e-06,
+      "loss": 11.8448,
+      "step": 16034
+    },
+    {
+      "epoch": 2.8041967213114756,
+      "grad_norm": 0.7948696769411293,
+      "learning_rate": 2.223613858381446e-06,
+      "loss": 12.0187,
+      "step": 16035
+    },
+    {
+      "epoch": 2.8043715846994535,
+      "grad_norm": 0.8127509620364618,
+      "learning_rate": 2.219656108431001e-06,
+      "loss": 11.7626,
+      "step": 16036
+    },
+    {
+      "epoch": 2.804546448087432,
+      "grad_norm": 0.8327429758925281,
+      "learning_rate": 2.2157018442355736e-06,
+      "loss": 11.8723,
+      "step": 16037
+    },
+    {
+      "epoch": 2.8047213114754097,
+      "grad_norm": 0.7912490818150656,
+      "learning_rate": 2.2117510659361184e-06,
+      "loss": 11.8765,
+      "step": 16038
+    },
+    {
+      "epoch": 2.804896174863388,
+      "grad_norm": 0.842975581134004,
+      "learning_rate": 2.2078037736734893e-06,
+      "loss": 11.7883,
+      "step": 16039
+    },
+    {
+      "epoch": 2.8050710382513664,
+      "grad_norm": 0.7911191904311634,
+      "learning_rate": 2.2038599675884066e-06,
+      "loss": 11.9898,
+      "step": 16040
+    },
+    {
+      "epoch": 2.8052459016393443,
+      "grad_norm": 0.8782698425207868,
+      "learning_rate": 2.1999196478214467e-06,
+      "loss": 11.9315,
+      "step": 16041
+    },
+    {
+      "epoch": 2.805420765027322,
+      "grad_norm": 0.8314873700139163,
+      "learning_rate": 2.1959828145130757e-06,
+      "loss": 11.5063,
+      "step": 16042
+    },
+    {
+      "epoch": 2.8055956284153005,
+      "grad_norm": 0.8408391436619793,
+      "learning_rate": 2.1920494678036584e-06,
+      "loss": 12.0576,
+      "step": 16043
+    },
+    {
+      "epoch": 2.805770491803279,
+      "grad_norm": 0.7903608522437822,
+      "learning_rate": 2.1881196078333944e-06,
+      "loss": 11.3778,
+      "step": 16044
+    },
+    {
+      "epoch": 2.8059453551912568,
+      "grad_norm": 0.8028406138108456,
+      "learning_rate": 2.184193234742382e-06,
+      "loss": 11.8125,
+      "step": 16045
+    },
+    {
+      "epoch": 2.806120218579235,
+      "grad_norm": 0.860201620098555,
+      "learning_rate": 2.180270348670599e-06,
+      "loss": 11.7267,
+      "step": 16046
+    },
+    {
+      "epoch": 2.806295081967213,
+      "grad_norm": 0.787303254377701,
+      "learning_rate": 2.176350949757888e-06,
+      "loss": 11.9644,
+      "step": 16047
+    },
+    {
+      "epoch": 2.8064699453551913,
+      "grad_norm": 0.8298534592472631,
+      "learning_rate": 2.172435038143983e-06,
+      "loss": 11.6069,
+      "step": 16048
+    },
+    {
+      "epoch": 2.8066448087431692,
+      "grad_norm": 0.7840890821979232,
+      "learning_rate": 2.168522613968449e-06,
+      "loss": 11.7991,
+      "step": 16049
+    },
+    {
+      "epoch": 2.8068196721311476,
+      "grad_norm": 0.8376211675534592,
+      "learning_rate": 2.164613677370797e-06,
+      "loss": 11.7267,
+      "step": 16050
+    },
+    {
+      "epoch": 2.8069945355191255,
+      "grad_norm": 0.8013311854678025,
+      "learning_rate": 2.16070822849036e-06,
+      "loss": 11.8069,
+      "step": 16051
+    },
+    {
+      "epoch": 2.807169398907104,
+      "grad_norm": 0.7799762464411619,
+      "learning_rate": 2.156806267466349e-06,
+      "loss": 12.1086,
+      "step": 16052
+    },
+    {
+      "epoch": 2.807344262295082,
+      "grad_norm": 0.8168086092580891,
+      "learning_rate": 2.1529077944378752e-06,
+      "loss": 11.7997,
+      "step": 16053
+    },
+    {
+      "epoch": 2.80751912568306,
+      "grad_norm": 0.8296429847294725,
+      "learning_rate": 2.1490128095439267e-06,
+      "loss": 11.8915,
+      "step": 16054
+    },
+    {
+      "epoch": 2.8076939890710384,
+      "grad_norm": 0.8520726179422368,
+      "learning_rate": 2.1451213129233483e-06,
+      "loss": 11.7368,
+      "step": 16055
+    },
+    {
+      "epoch": 2.8078688524590163,
+      "grad_norm": 0.8309534742698026,
+      "learning_rate": 2.1412333047148514e-06,
+      "loss": 11.599,
+      "step": 16056
+    },
+    {
+      "epoch": 2.8080437158469946,
+      "grad_norm": 0.8358518096885157,
+      "learning_rate": 2.1373487850570584e-06,
+      "loss": 12.0985,
+      "step": 16057
+    },
+    {
+      "epoch": 2.8082185792349725,
+      "grad_norm": 0.8571797407666991,
+      "learning_rate": 2.1334677540884363e-06,
+      "loss": 11.9511,
+      "step": 16058
+    },
+    {
+      "epoch": 2.808393442622951,
+      "grad_norm": 0.8308541875352484,
+      "learning_rate": 2.1295902119473408e-06,
+      "loss": 11.7189,
+      "step": 16059
+    },
+    {
+      "epoch": 2.8085683060109288,
+      "grad_norm": 0.8384905390692889,
+      "learning_rate": 2.1257161587720174e-06,
+      "loss": 11.8586,
+      "step": 16060
+    },
+    {
+      "epoch": 2.808743169398907,
+      "grad_norm": 0.8163583667520236,
+      "learning_rate": 2.1218455947005443e-06,
+      "loss": 11.9032,
+      "step": 16061
+    },
+    {
+      "epoch": 2.8089180327868855,
+      "grad_norm": 0.8534907997320043,
+      "learning_rate": 2.1179785198709332e-06,
+      "loss": 11.8264,
+      "step": 16062
+    },
+    {
+      "epoch": 2.8090928961748634,
+      "grad_norm": 0.8464568651302957,
+      "learning_rate": 2.1141149344210185e-06,
+      "loss": 11.7815,
+      "step": 16063
+    },
+    {
+      "epoch": 2.8092677595628417,
+      "grad_norm": 0.8339704701872771,
+      "learning_rate": 2.1102548384885345e-06,
+      "loss": 11.9086,
+      "step": 16064
+    },
+    {
+      "epoch": 2.8094426229508196,
+      "grad_norm": 0.8866514443123636,
+      "learning_rate": 2.106398232211104e-06,
+      "loss": 11.5648,
+      "step": 16065
+    },
+    {
+      "epoch": 2.809617486338798,
+      "grad_norm": 0.8636154412183045,
+      "learning_rate": 2.1025451157261957e-06,
+      "loss": 11.6422,
+      "step": 16066
+    },
+    {
+      "epoch": 2.809792349726776,
+      "grad_norm": 0.8435152705249161,
+      "learning_rate": 2.0986954891711764e-06,
+      "loss": 11.805,
+      "step": 16067
+    },
+    {
+      "epoch": 2.809967213114754,
+      "grad_norm": 0.7690359128558447,
+      "learning_rate": 2.0948493526832703e-06,
+      "loss": 11.9294,
+      "step": 16068
+    },
+    {
+      "epoch": 2.810142076502732,
+      "grad_norm": 0.7460969941081382,
+      "learning_rate": 2.09100670639959e-06,
+      "loss": 11.7775,
+      "step": 16069
+    },
+    {
+      "epoch": 2.8103169398907104,
+      "grad_norm": 0.8260040805702914,
+      "learning_rate": 2.0871675504571364e-06,
+      "loss": 11.7279,
+      "step": 16070
+    },
+    {
+      "epoch": 2.8104918032786887,
+      "grad_norm": 0.843697414470126,
+      "learning_rate": 2.083331884992756e-06,
+      "loss": 11.9546,
+      "step": 16071
+    },
+    {
+      "epoch": 2.8106666666666666,
+      "grad_norm": 0.7763426687557085,
+      "learning_rate": 2.0794997101431958e-06,
+      "loss": 11.7191,
+      "step": 16072
+    },
+    {
+      "epoch": 2.8108415300546445,
+      "grad_norm": 0.8337687530695002,
+      "learning_rate": 2.0756710260450562e-06,
+      "loss": 11.7713,
+      "step": 16073
+    },
+    {
+      "epoch": 2.811016393442623,
+      "grad_norm": 0.8328058116109764,
+      "learning_rate": 2.07184583283484e-06,
+      "loss": 11.6562,
+      "step": 16074
+    },
+    {
+      "epoch": 2.8111912568306012,
+      "grad_norm": 0.8201159761392508,
+      "learning_rate": 2.0680241306489048e-06,
+      "loss": 11.7827,
+      "step": 16075
+    },
+    {
+      "epoch": 2.811366120218579,
+      "grad_norm": 0.7982859086826335,
+      "learning_rate": 2.0642059196234743e-06,
+      "loss": 11.8053,
+      "step": 16076
+    },
+    {
+      "epoch": 2.8115409836065575,
+      "grad_norm": 0.8138292875346692,
+      "learning_rate": 2.0603911998946846e-06,
+      "loss": 11.9316,
+      "step": 16077
+    },
+    {
+      "epoch": 2.8117158469945354,
+      "grad_norm": 0.8250024073739017,
+      "learning_rate": 2.056579971598516e-06,
+      "loss": 11.937,
+      "step": 16078
+    },
+    {
+      "epoch": 2.8118907103825137,
+      "grad_norm": 0.7830665623701503,
+      "learning_rate": 2.0527722348708366e-06,
+      "loss": 11.7092,
+      "step": 16079
+    },
+    {
+      "epoch": 2.812065573770492,
+      "grad_norm": 0.826353648691931,
+      "learning_rate": 2.0489679898473947e-06,
+      "loss": 11.849,
+      "step": 16080
+    },
+    {
+      "epoch": 2.81224043715847,
+      "grad_norm": 0.8186124077278218,
+      "learning_rate": 2.045167236663803e-06,
+      "loss": 12.0582,
+      "step": 16081
+    },
+    {
+      "epoch": 2.812415300546448,
+      "grad_norm": 0.8294980907341801,
+      "learning_rate": 2.0413699754555425e-06,
+      "loss": 11.978,
+      "step": 16082
+    },
+    {
+      "epoch": 2.812590163934426,
+      "grad_norm": 0.824997514183264,
+      "learning_rate": 2.037576206357983e-06,
+      "loss": 11.8811,
+      "step": 16083
+    },
+    {
+      "epoch": 2.8127650273224045,
+      "grad_norm": 0.8659169025531491,
+      "learning_rate": 2.033785929506382e-06,
+      "loss": 11.8626,
+      "step": 16084
+    },
+    {
+      "epoch": 2.8129398907103824,
+      "grad_norm": 0.7915177367154949,
+      "learning_rate": 2.0299991450358657e-06,
+      "loss": 11.8261,
+      "step": 16085
+    },
+    {
+      "epoch": 2.8131147540983608,
+      "grad_norm": 0.8685755437097137,
+      "learning_rate": 2.0262158530814037e-06,
+      "loss": 11.8827,
+      "step": 16086
+    },
+    {
+      "epoch": 2.8132896174863387,
+      "grad_norm": 0.8131696185698543,
+      "learning_rate": 2.0224360537778652e-06,
+      "loss": 11.6943,
+      "step": 16087
+    },
+    {
+      "epoch": 2.813464480874317,
+      "grad_norm": 0.7930417754449063,
+      "learning_rate": 2.018659747260021e-06,
+      "loss": 11.9383,
+      "step": 16088
+    },
+    {
+      "epoch": 2.8136393442622953,
+      "grad_norm": 0.818358084489303,
+      "learning_rate": 2.0148869336624855e-06,
+      "loss": 11.7408,
+      "step": 16089
+    },
+    {
+      "epoch": 2.8138142076502732,
+      "grad_norm": 0.9241776457935922,
+      "learning_rate": 2.0111176131197285e-06,
+      "loss": 11.587,
+      "step": 16090
+    },
+    {
+      "epoch": 2.813989071038251,
+      "grad_norm": 0.8498767971803267,
+      "learning_rate": 2.007351785766143e-06,
+      "loss": 11.7864,
+      "step": 16091
+    },
+    {
+      "epoch": 2.8141639344262295,
+      "grad_norm": 0.825627258970623,
+      "learning_rate": 2.0035894517359765e-06,
+      "loss": 11.792,
+      "step": 16092
+    },
+    {
+      "epoch": 2.814338797814208,
+      "grad_norm": 0.7959819247906155,
+      "learning_rate": 1.999830611163367e-06,
+      "loss": 11.9096,
+      "step": 16093
+    },
+    {
+      "epoch": 2.8145136612021857,
+      "grad_norm": 0.9118985832091588,
+      "learning_rate": 1.9960752641822733e-06,
+      "loss": 11.8495,
+      "step": 16094
+    },
+    {
+      "epoch": 2.814688524590164,
+      "grad_norm": 0.8199697171120754,
+      "learning_rate": 1.9923234109266e-06,
+      "loss": 11.9322,
+      "step": 16095
+    },
+    {
+      "epoch": 2.814863387978142,
+      "grad_norm": 0.79411172381213,
+      "learning_rate": 1.988575051530084e-06,
+      "loss": 11.9659,
+      "step": 16096
+    },
+    {
+      "epoch": 2.8150382513661203,
+      "grad_norm": 0.7777861446731247,
+      "learning_rate": 1.984830186126352e-06,
+      "loss": 11.7937,
+      "step": 16097
+    },
+    {
+      "epoch": 2.8152131147540986,
+      "grad_norm": 0.7593382960301266,
+      "learning_rate": 1.981088814848908e-06,
+      "loss": 11.7303,
+      "step": 16098
+    },
+    {
+      "epoch": 2.8153879781420765,
+      "grad_norm": 1.108492833826254,
+      "learning_rate": 1.977350937831124e-06,
+      "loss": 11.8403,
+      "step": 16099
+    },
+    {
+      "epoch": 2.8155628415300544,
+      "grad_norm": 0.8189731485254502,
+      "learning_rate": 1.973616555206259e-06,
+      "loss": 11.9418,
+      "step": 16100
+    },
+    {
+      "epoch": 2.8157377049180328,
+      "grad_norm": 0.8467751218712043,
+      "learning_rate": 1.9698856671074294e-06,
+      "loss": 11.6711,
+      "step": 16101
+    },
+    {
+      "epoch": 2.815912568306011,
+      "grad_norm": 0.8742471207068915,
+      "learning_rate": 1.966158273667629e-06,
+      "loss": 11.8111,
+      "step": 16102
+    },
+    {
+      "epoch": 2.816087431693989,
+      "grad_norm": 0.8548467691597037,
+      "learning_rate": 1.9624343750197506e-06,
+      "loss": 11.8575,
+      "step": 16103
+    },
+    {
+      "epoch": 2.8162622950819673,
+      "grad_norm": 0.8487431346664526,
+      "learning_rate": 1.958713971296555e-06,
+      "loss": 11.8465,
+      "step": 16104
+    },
+    {
+      "epoch": 2.8164371584699452,
+      "grad_norm": 0.8328664914034621,
+      "learning_rate": 1.954997062630648e-06,
+      "loss": 12.033,
+      "step": 16105
+    },
+    {
+      "epoch": 2.8166120218579236,
+      "grad_norm": 0.8006184887852312,
+      "learning_rate": 1.951283649154545e-06,
+      "loss": 11.8287,
+      "step": 16106
+    },
+    {
+      "epoch": 2.8167868852459015,
+      "grad_norm": 0.7566211944556328,
+      "learning_rate": 1.9475737310006182e-06,
+      "loss": 11.5575,
+      "step": 16107
+    },
+    {
+      "epoch": 2.81696174863388,
+      "grad_norm": 0.7928199287044642,
+      "learning_rate": 1.9438673083011284e-06,
+      "loss": 11.7581,
+      "step": 16108
+    },
+    {
+      "epoch": 2.8171366120218577,
+      "grad_norm": 0.8975668101366199,
+      "learning_rate": 1.940164381188203e-06,
+      "loss": 11.863,
+      "step": 16109
+    },
+    {
+      "epoch": 2.817311475409836,
+      "grad_norm": 0.8444529366794525,
+      "learning_rate": 1.936464949793848e-06,
+      "loss": 11.9003,
+      "step": 16110
+    },
+    {
+      "epoch": 2.8174863387978144,
+      "grad_norm": 0.895554858138711,
+      "learning_rate": 1.9327690142499357e-06,
+      "loss": 11.9535,
+      "step": 16111
+    },
+    {
+      "epoch": 2.8176612021857923,
+      "grad_norm": 0.8576695669515757,
+      "learning_rate": 1.929076574688238e-06,
+      "loss": 12.0589,
+      "step": 16112
+    },
+    {
+      "epoch": 2.8178360655737706,
+      "grad_norm": 0.8469506616055533,
+      "learning_rate": 1.925387631240372e-06,
+      "loss": 11.7962,
+      "step": 16113
+    },
+    {
+      "epoch": 2.8180109289617485,
+      "grad_norm": 0.7710540878347927,
+      "learning_rate": 1.9217021840378434e-06,
+      "loss": 11.7433,
+      "step": 16114
+    },
+    {
+      "epoch": 2.818185792349727,
+      "grad_norm": 0.8259832877857325,
+      "learning_rate": 1.9180202332120476e-06,
+      "loss": 11.9439,
+      "step": 16115
+    },
+    {
+      "epoch": 2.818360655737705,
+      "grad_norm": 0.7882723526742709,
+      "learning_rate": 1.9143417788942243e-06,
+      "loss": 11.9385,
+      "step": 16116
+    },
+    {
+      "epoch": 2.818535519125683,
+      "grad_norm": 0.8734901629938748,
+      "learning_rate": 1.9106668212155233e-06,
+      "loss": 12.0225,
+      "step": 16117
+    },
+    {
+      "epoch": 2.818710382513661,
+      "grad_norm": 0.8236347328425107,
+      "learning_rate": 1.9069953603069512e-06,
+      "loss": 11.573,
+      "step": 16118
+    },
+    {
+      "epoch": 2.8188852459016394,
+      "grad_norm": 0.8165824922680814,
+      "learning_rate": 1.9033273962993814e-06,
+      "loss": 11.6695,
+      "step": 16119
+    },
+    {
+      "epoch": 2.8190601092896177,
+      "grad_norm": 0.8589244409928094,
+      "learning_rate": 1.8996629293235758e-06,
+      "loss": 11.8862,
+      "step": 16120
+    },
+    {
+      "epoch": 2.8192349726775956,
+      "grad_norm": 0.7858577968077873,
+      "learning_rate": 1.896001959510163e-06,
+      "loss": 11.9402,
+      "step": 16121
+    },
+    {
+      "epoch": 2.8194098360655735,
+      "grad_norm": 0.8796066314778217,
+      "learning_rate": 1.892344486989661e-06,
+      "loss": 11.654,
+      "step": 16122
+    },
+    {
+      "epoch": 2.819584699453552,
+      "grad_norm": 0.9205089679780057,
+      "learning_rate": 1.8886905118924547e-06,
+      "loss": 11.9405,
+      "step": 16123
+    },
+    {
+      "epoch": 2.81975956284153,
+      "grad_norm": 0.8276577348589754,
+      "learning_rate": 1.8850400343488062e-06,
+      "loss": 11.9298,
+      "step": 16124
+    },
+    {
+      "epoch": 2.819934426229508,
+      "grad_norm": 0.8895247995075448,
+      "learning_rate": 1.8813930544888336e-06,
+      "loss": 11.9023,
+      "step": 16125
+    },
+    {
+      "epoch": 2.8201092896174864,
+      "grad_norm": 0.9480801965458349,
+      "learning_rate": 1.8777495724425553e-06,
+      "loss": 11.9815,
+      "step": 16126
+    },
+    {
+      "epoch": 2.8202841530054643,
+      "grad_norm": 0.8498807917849077,
+      "learning_rate": 1.8741095883398786e-06,
+      "loss": 11.6688,
+      "step": 16127
+    },
+    {
+      "epoch": 2.8204590163934427,
+      "grad_norm": 0.9022313232415203,
+      "learning_rate": 1.8704731023105436e-06,
+      "loss": 11.9327,
+      "step": 16128
+    },
+    {
+      "epoch": 2.820633879781421,
+      "grad_norm": 0.9286668994137411,
+      "learning_rate": 1.8668401144841807e-06,
+      "loss": 11.9473,
+      "step": 16129
+    },
+    {
+      "epoch": 2.820808743169399,
+      "grad_norm": 0.8161885690145055,
+      "learning_rate": 1.8632106249903192e-06,
+      "loss": 12.0615,
+      "step": 16130
+    },
+    {
+      "epoch": 2.820983606557377,
+      "grad_norm": 0.8002334953761895,
+      "learning_rate": 1.8595846339583333e-06,
+      "loss": 11.8421,
+      "step": 16131
+    },
+    {
+      "epoch": 2.821158469945355,
+      "grad_norm": 0.8281139514280894,
+      "learning_rate": 1.8559621415174976e-06,
+      "loss": 11.848,
+      "step": 16132
+    },
+    {
+      "epoch": 2.8213333333333335,
+      "grad_norm": 0.7611721257200676,
+      "learning_rate": 1.8523431477969311e-06,
+      "loss": 11.9686,
+      "step": 16133
+    },
+    {
+      "epoch": 2.8215081967213114,
+      "grad_norm": 0.8688097954989691,
+      "learning_rate": 1.8487276529256636e-06,
+      "loss": 11.6321,
+      "step": 16134
+    },
+    {
+      "epoch": 2.8216830601092897,
+      "grad_norm": 0.7970226018034468,
+      "learning_rate": 1.8451156570325812e-06,
+      "loss": 11.8584,
+      "step": 16135
+    },
+    {
+      "epoch": 2.8218579234972676,
+      "grad_norm": 0.8439463016767291,
+      "learning_rate": 1.8415071602464362e-06,
+      "loss": 11.8505,
+      "step": 16136
+    },
+    {
+      "epoch": 2.822032786885246,
+      "grad_norm": 0.8639560993475481,
+      "learning_rate": 1.8379021626958814e-06,
+      "loss": 11.9378,
+      "step": 16137
+    },
+    {
+      "epoch": 2.8222076502732243,
+      "grad_norm": 0.8076615400149353,
+      "learning_rate": 1.8343006645094251e-06,
+      "loss": 11.7816,
+      "step": 16138
+    },
+    {
+      "epoch": 2.822382513661202,
+      "grad_norm": 0.9098663059952402,
+      "learning_rate": 1.8307026658154536e-06,
+      "loss": 11.7931,
+      "step": 16139
+    },
+    {
+      "epoch": 2.82255737704918,
+      "grad_norm": 0.8068381972292431,
+      "learning_rate": 1.8271081667422306e-06,
+      "loss": 11.5578,
+      "step": 16140
+    },
+    {
+      "epoch": 2.8227322404371584,
+      "grad_norm": 0.8383632889096481,
+      "learning_rate": 1.8235171674178986e-06,
+      "loss": 11.7396,
+      "step": 16141
+    },
+    {
+      "epoch": 2.8229071038251368,
+      "grad_norm": 0.9164819255357688,
+      "learning_rate": 1.819929667970488e-06,
+      "loss": 11.9937,
+      "step": 16142
+    },
+    {
+      "epoch": 2.8230819672131147,
+      "grad_norm": 0.7940368445241801,
+      "learning_rate": 1.8163456685278524e-06,
+      "loss": 11.7806,
+      "step": 16143
+    },
+    {
+      "epoch": 2.823256830601093,
+      "grad_norm": 0.7860970720191922,
+      "learning_rate": 1.8127651692177893e-06,
+      "loss": 11.9012,
+      "step": 16144
+    },
+    {
+      "epoch": 2.823431693989071,
+      "grad_norm": 0.7576718915481188,
+      "learning_rate": 1.8091881701679415e-06,
+      "loss": 11.9379,
+      "step": 16145
+    },
+    {
+      "epoch": 2.8236065573770492,
+      "grad_norm": 0.8101756722040289,
+      "learning_rate": 1.8056146715057952e-06,
+      "loss": 11.7374,
+      "step": 16146
+    },
+    {
+      "epoch": 2.8237814207650276,
+      "grad_norm": 0.8240529947531984,
+      "learning_rate": 1.8020446733587492e-06,
+      "loss": 11.5228,
+      "step": 16147
+    },
+    {
+      "epoch": 2.8239562841530055,
+      "grad_norm": 0.7582594996318445,
+      "learning_rate": 1.7984781758540904e-06,
+      "loss": 11.9454,
+      "step": 16148
+    },
+    {
+      "epoch": 2.8241311475409834,
+      "grad_norm": 0.7724356197554998,
+      "learning_rate": 1.79491517911895e-06,
+      "loss": 11.9217,
+      "step": 16149
+    },
+    {
+      "epoch": 2.8243060109289617,
+      "grad_norm": 0.8206729808773847,
+      "learning_rate": 1.791355683280338e-06,
+      "loss": 11.8956,
+      "step": 16150
+    },
+    {
+      "epoch": 2.82448087431694,
+      "grad_norm": 0.7625684142530503,
+      "learning_rate": 1.7877996884651526e-06,
+      "loss": 11.7598,
+      "step": 16151
+    },
+    {
+      "epoch": 2.824655737704918,
+      "grad_norm": 0.9324565438452143,
+      "learning_rate": 1.7842471948001593e-06,
+      "loss": 11.6535,
+      "step": 16152
+    },
+    {
+      "epoch": 2.8248306010928963,
+      "grad_norm": 0.7843006639957536,
+      "learning_rate": 1.7806982024120123e-06,
+      "loss": 11.8141,
+      "step": 16153
+    },
+    {
+      "epoch": 2.825005464480874,
+      "grad_norm": 0.8194239511005758,
+      "learning_rate": 1.7771527114271992e-06,
+      "loss": 11.682,
+      "step": 16154
+    },
+    {
+      "epoch": 2.8251803278688525,
+      "grad_norm": 0.814938306790503,
+      "learning_rate": 1.7736107219721409e-06,
+      "loss": 11.9563,
+      "step": 16155
+    },
+    {
+      "epoch": 2.825355191256831,
+      "grad_norm": 0.8485531792324985,
+      "learning_rate": 1.7700722341730925e-06,
+      "loss": 11.5514,
+      "step": 16156
+    },
+    {
+      "epoch": 2.8255300546448088,
+      "grad_norm": 0.8923481918539112,
+      "learning_rate": 1.766537248156208e-06,
+      "loss": 12.1982,
+      "step": 16157
+    },
+    {
+      "epoch": 2.8257049180327867,
+      "grad_norm": 0.8152973656034193,
+      "learning_rate": 1.7630057640474873e-06,
+      "loss": 11.8154,
+      "step": 16158
+    },
+    {
+      "epoch": 2.825879781420765,
+      "grad_norm": 0.7968631658415997,
+      "learning_rate": 1.7594777819728292e-06,
+      "loss": 12.0315,
+      "step": 16159
+    },
+    {
+      "epoch": 2.8260546448087434,
+      "grad_norm": 0.8171164095117917,
+      "learning_rate": 1.7559533020580222e-06,
+      "loss": 11.7336,
+      "step": 16160
+    },
+    {
+      "epoch": 2.8262295081967213,
+      "grad_norm": 0.8689989180020976,
+      "learning_rate": 1.7524323244286877e-06,
+      "loss": 11.7338,
+      "step": 16161
+    },
+    {
+      "epoch": 2.8264043715846996,
+      "grad_norm": 0.8223830750093349,
+      "learning_rate": 1.748914849210348e-06,
+      "loss": 11.6669,
+      "step": 16162
+    },
+    {
+      "epoch": 2.8265792349726775,
+      "grad_norm": 0.84735890625328,
+      "learning_rate": 1.7454008765283913e-06,
+      "loss": 11.9164,
+      "step": 16163
+    },
+    {
+      "epoch": 2.826754098360656,
+      "grad_norm": 0.8436250930874359,
+      "learning_rate": 1.7418904065081176e-06,
+      "loss": 11.7485,
+      "step": 16164
+    },
+    {
+      "epoch": 2.8269289617486337,
+      "grad_norm": 0.8001647406307679,
+      "learning_rate": 1.7383834392746268e-06,
+      "loss": 11.9509,
+      "step": 16165
+    },
+    {
+      "epoch": 2.827103825136612,
+      "grad_norm": 0.7888736395912359,
+      "learning_rate": 1.7348799749529742e-06,
+      "loss": 11.9144,
+      "step": 16166
+    },
+    {
+      "epoch": 2.82727868852459,
+      "grad_norm": 0.8193363683074948,
+      "learning_rate": 1.731380013668027e-06,
+      "loss": 11.4562,
+      "step": 16167
+    },
+    {
+      "epoch": 2.8274535519125683,
+      "grad_norm": 0.8523334406720959,
+      "learning_rate": 1.7278835555445738e-06,
+      "loss": 11.8034,
+      "step": 16168
+    },
+    {
+      "epoch": 2.8276284153005466,
+      "grad_norm": 0.8218533654017233,
+      "learning_rate": 1.7243906007072598e-06,
+      "loss": 11.662,
+      "step": 16169
+    },
+    {
+      "epoch": 2.8278032786885245,
+      "grad_norm": 0.8283668492342139,
+      "learning_rate": 1.720901149280585e-06,
+      "loss": 11.7956,
+      "step": 16170
+    },
+    {
+      "epoch": 2.827978142076503,
+      "grad_norm": 0.9491589089250776,
+      "learning_rate": 1.717415201388961e-06,
+      "loss": 11.9635,
+      "step": 16171
+    },
+    {
+      "epoch": 2.828153005464481,
+      "grad_norm": 0.8444012751191952,
+      "learning_rate": 1.7139327571566665e-06,
+      "loss": 11.8192,
+      "step": 16172
+    },
+    {
+      "epoch": 2.828327868852459,
+      "grad_norm": 0.7699593353799258,
+      "learning_rate": 1.7104538167078243e-06,
+      "loss": 11.9842,
+      "step": 16173
+    },
+    {
+      "epoch": 2.828502732240437,
+      "grad_norm": 0.8298744888841075,
+      "learning_rate": 1.7069783801664685e-06,
+      "loss": 11.951,
+      "step": 16174
+    },
+    {
+      "epoch": 2.8286775956284154,
+      "grad_norm": 0.8311627617318068,
+      "learning_rate": 1.7035064476564887e-06,
+      "loss": 11.7356,
+      "step": 16175
+    },
+    {
+      "epoch": 2.8288524590163933,
+      "grad_norm": 0.7680089984632756,
+      "learning_rate": 1.700038019301664e-06,
+      "loss": 11.94,
+      "step": 16176
+    },
+    {
+      "epoch": 2.8290273224043716,
+      "grad_norm": 0.8595179366933711,
+      "learning_rate": 1.6965730952256175e-06,
+      "loss": 11.5301,
+      "step": 16177
+    },
+    {
+      "epoch": 2.82920218579235,
+      "grad_norm": 0.8421170263428543,
+      "learning_rate": 1.693111675551895e-06,
+      "loss": 11.978,
+      "step": 16178
+    },
+    {
+      "epoch": 2.829377049180328,
+      "grad_norm": 0.8007729288543904,
+      "learning_rate": 1.6896537604038864e-06,
+      "loss": 11.507,
+      "step": 16179
+    },
+    {
+      "epoch": 2.8295519125683057,
+      "grad_norm": 0.8039729150864099,
+      "learning_rate": 1.686199349904849e-06,
+      "loss": 11.7303,
+      "step": 16180
+    },
+    {
+      "epoch": 2.829726775956284,
+      "grad_norm": 0.7991615938393606,
+      "learning_rate": 1.6827484441779396e-06,
+      "loss": 11.779,
+      "step": 16181
+    },
+    {
+      "epoch": 2.8299016393442624,
+      "grad_norm": 0.8355137782571048,
+      "learning_rate": 1.679301043346171e-06,
+      "loss": 11.6806,
+      "step": 16182
+    },
+    {
+      "epoch": 2.8300765027322403,
+      "grad_norm": 0.8362995669980651,
+      "learning_rate": 1.675857147532456e-06,
+      "loss": 11.7088,
+      "step": 16183
+    },
+    {
+      "epoch": 2.8302513661202187,
+      "grad_norm": 0.7995351100363547,
+      "learning_rate": 1.6724167568595518e-06,
+      "loss": 11.7611,
+      "step": 16184
+    },
+    {
+      "epoch": 2.8304262295081966,
+      "grad_norm": 0.7894066959374042,
+      "learning_rate": 1.6689798714501048e-06,
+      "loss": 11.6094,
+      "step": 16185
+    },
+    {
+      "epoch": 2.830601092896175,
+      "grad_norm": 0.8970157350636005,
+      "learning_rate": 1.6655464914266284e-06,
+      "loss": 11.8261,
+      "step": 16186
+    },
+    {
+      "epoch": 2.8307759562841532,
+      "grad_norm": 0.8495558489497941,
+      "learning_rate": 1.6621166169115354e-06,
+      "loss": 11.7715,
+      "step": 16187
+    },
+    {
+      "epoch": 2.830950819672131,
+      "grad_norm": 0.836928645620872,
+      "learning_rate": 1.658690248027095e-06,
+      "loss": 11.6597,
+      "step": 16188
+    },
+    {
+      "epoch": 2.831125683060109,
+      "grad_norm": 0.8275893085594117,
+      "learning_rate": 1.6552673848954425e-06,
+      "loss": 11.7528,
+      "step": 16189
+    },
+    {
+      "epoch": 2.8313005464480874,
+      "grad_norm": 0.7960394278704037,
+      "learning_rate": 1.6518480276386028e-06,
+      "loss": 11.8963,
+      "step": 16190
+    },
+    {
+      "epoch": 2.8314754098360657,
+      "grad_norm": 0.8432063793101965,
+      "learning_rate": 1.6484321763784783e-06,
+      "loss": 12.0253,
+      "step": 16191
+    },
+    {
+      "epoch": 2.8316502732240436,
+      "grad_norm": 0.7582983068456187,
+      "learning_rate": 1.6450198312368382e-06,
+      "loss": 11.9095,
+      "step": 16192
+    },
+    {
+      "epoch": 2.831825136612022,
+      "grad_norm": 0.818266054153907,
+      "learning_rate": 1.6416109923353185e-06,
+      "loss": 11.9683,
+      "step": 16193
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.8367268063042487,
+      "learning_rate": 1.6382056597954553e-06,
+      "loss": 11.723,
+      "step": 16194
+    },
+    {
+      "epoch": 2.832174863387978,
+      "grad_norm": 0.7868937409287899,
+      "learning_rate": 1.6348038337386295e-06,
+      "loss": 11.65,
+      "step": 16195
+    },
+    {
+      "epoch": 2.8323497267759565,
+      "grad_norm": 0.7879107827545375,
+      "learning_rate": 1.6314055142861106e-06,
+      "loss": 11.9133,
+      "step": 16196
+    },
+    {
+      "epoch": 2.8325245901639344,
+      "grad_norm": 0.7846930861419821,
+      "learning_rate": 1.6280107015590685e-06,
+      "loss": 11.8105,
+      "step": 16197
+    },
+    {
+      "epoch": 2.8326994535519123,
+      "grad_norm": 0.8799624052967153,
+      "learning_rate": 1.624619395678506e-06,
+      "loss": 11.9384,
+      "step": 16198
+    },
+    {
+      "epoch": 2.8328743169398907,
+      "grad_norm": 0.8718359303785524,
+      "learning_rate": 1.621231596765327e-06,
+      "loss": 11.8834,
+      "step": 16199
+    },
+    {
+      "epoch": 2.833049180327869,
+      "grad_norm": 0.801799726045335,
+      "learning_rate": 1.6178473049402899e-06,
+      "loss": 11.9708,
+      "step": 16200
+    },
+    {
+      "epoch": 2.833224043715847,
+      "grad_norm": 0.8134895127811209,
+      "learning_rate": 1.614466520324054e-06,
+      "loss": 11.9992,
+      "step": 16201
+    },
+    {
+      "epoch": 2.8333989071038252,
+      "grad_norm": 0.8294412826987502,
+      "learning_rate": 1.6110892430371339e-06,
+      "loss": 11.8569,
+      "step": 16202
+    },
+    {
+      "epoch": 2.833573770491803,
+      "grad_norm": 0.7962006119501784,
+      "learning_rate": 1.6077154731999332e-06,
+      "loss": 11.9415,
+      "step": 16203
+    },
+    {
+      "epoch": 2.8337486338797815,
+      "grad_norm": 0.9069484879098637,
+      "learning_rate": 1.6043452109327007e-06,
+      "loss": 11.9154,
+      "step": 16204
+    },
+    {
+      "epoch": 2.83392349726776,
+      "grad_norm": 0.8480473116015413,
+      "learning_rate": 1.6009784563556175e-06,
+      "loss": 11.8315,
+      "step": 16205
+    },
+    {
+      "epoch": 2.8340983606557377,
+      "grad_norm": 0.8146155617129561,
+      "learning_rate": 1.5976152095886654e-06,
+      "loss": 11.7715,
+      "step": 16206
+    },
+    {
+      "epoch": 2.8342732240437156,
+      "grad_norm": 0.8425013233609873,
+      "learning_rate": 1.594255470751771e-06,
+      "loss": 11.7237,
+      "step": 16207
+    },
+    {
+      "epoch": 2.834448087431694,
+      "grad_norm": 0.8170626174526268,
+      "learning_rate": 1.5908992399646938e-06,
+      "loss": 11.8121,
+      "step": 16208
+    },
+    {
+      "epoch": 2.8346229508196723,
+      "grad_norm": 0.8574613020180789,
+      "learning_rate": 1.587546517347094e-06,
+      "loss": 11.8423,
+      "step": 16209
+    },
+    {
+      "epoch": 2.83479781420765,
+      "grad_norm": 0.7973176192299652,
+      "learning_rate": 1.5841973030184643e-06,
+      "loss": 11.8053,
+      "step": 16210
+    },
+    {
+      "epoch": 2.8349726775956285,
+      "grad_norm": 0.8413347969834617,
+      "learning_rate": 1.5808515970982097e-06,
+      "loss": 11.8542,
+      "step": 16211
+    },
+    {
+      "epoch": 2.8351475409836064,
+      "grad_norm": 0.8048764705651417,
+      "learning_rate": 1.5775093997056124e-06,
+      "loss": 11.9019,
+      "step": 16212
+    },
+    {
+      "epoch": 2.835322404371585,
+      "grad_norm": 0.7895188178194967,
+      "learning_rate": 1.5741707109598103e-06,
+      "loss": 12.021,
+      "step": 16213
+    },
+    {
+      "epoch": 2.8354972677595627,
+      "grad_norm": 0.8509942542943542,
+      "learning_rate": 1.5708355309798195e-06,
+      "loss": 11.6339,
+      "step": 16214
+    },
+    {
+      "epoch": 2.835672131147541,
+      "grad_norm": 0.8052614108666595,
+      "learning_rate": 1.5675038598845448e-06,
+      "loss": 11.8947,
+      "step": 16215
+    },
+    {
+      "epoch": 2.835846994535519,
+      "grad_norm": 0.8148608578700701,
+      "learning_rate": 1.5641756977927468e-06,
+      "loss": 11.5626,
+      "step": 16216
+    },
+    {
+      "epoch": 2.8360218579234973,
+      "grad_norm": 0.7985529263271403,
+      "learning_rate": 1.5608510448230863e-06,
+      "loss": 11.8911,
+      "step": 16217
+    },
+    {
+      "epoch": 2.8361967213114756,
+      "grad_norm": 0.9548143679612636,
+      "learning_rate": 1.557529901094068e-06,
+      "loss": 11.7862,
+      "step": 16218
+    },
+    {
+      "epoch": 2.8363715846994535,
+      "grad_norm": 0.8163944640313756,
+      "learning_rate": 1.5542122667240867e-06,
+      "loss": 11.6247,
+      "step": 16219
+    },
+    {
+      "epoch": 2.836546448087432,
+      "grad_norm": 0.9585127174615614,
+      "learning_rate": 1.5508981418314138e-06,
+      "loss": 11.6882,
+      "step": 16220
+    },
+    {
+      "epoch": 2.8367213114754097,
+      "grad_norm": 0.850115749005684,
+      "learning_rate": 1.5475875265341998e-06,
+      "loss": 11.5628,
+      "step": 16221
+    },
+    {
+      "epoch": 2.836896174863388,
+      "grad_norm": 0.8252362186343245,
+      "learning_rate": 1.5442804209504613e-06,
+      "loss": 11.8166,
+      "step": 16222
+    },
+    {
+      "epoch": 2.837071038251366,
+      "grad_norm": 0.8711558048013509,
+      "learning_rate": 1.5409768251980927e-06,
+      "loss": 11.7861,
+      "step": 16223
+    },
+    {
+      "epoch": 2.8372459016393443,
+      "grad_norm": 0.853739891242758,
+      "learning_rate": 1.5376767393948666e-06,
+      "loss": 11.8031,
+      "step": 16224
+    },
+    {
+      "epoch": 2.837420765027322,
+      "grad_norm": 0.7908071580710386,
+      "learning_rate": 1.534380163658422e-06,
+      "loss": 11.9235,
+      "step": 16225
+    },
+    {
+      "epoch": 2.8375956284153006,
+      "grad_norm": 0.8421276020165898,
+      "learning_rate": 1.5310870981062764e-06,
+      "loss": 11.7025,
+      "step": 16226
+    },
+    {
+      "epoch": 2.837770491803279,
+      "grad_norm": 0.8775548055281243,
+      "learning_rate": 1.5277975428558245e-06,
+      "loss": 11.6144,
+      "step": 16227
+    },
+    {
+      "epoch": 2.837945355191257,
+      "grad_norm": 0.9281051891188313,
+      "learning_rate": 1.5245114980243503e-06,
+      "loss": 11.6531,
+      "step": 16228
+    },
+    {
+      "epoch": 2.838120218579235,
+      "grad_norm": 0.8682336608153904,
+      "learning_rate": 1.5212289637289712e-06,
+      "loss": 11.8353,
+      "step": 16229
+    },
+    {
+      "epoch": 2.838295081967213,
+      "grad_norm": 0.8690461232706229,
+      "learning_rate": 1.517949940086727e-06,
+      "loss": 11.8513,
+      "step": 16230
+    },
+    {
+      "epoch": 2.8384699453551914,
+      "grad_norm": 0.8611696703142361,
+      "learning_rate": 1.514674427214502e-06,
+      "loss": 11.8675,
+      "step": 16231
+    },
+    {
+      "epoch": 2.8386448087431693,
+      "grad_norm": 0.8362035389132811,
+      "learning_rate": 1.5114024252290693e-06,
+      "loss": 11.8916,
+      "step": 16232
+    },
+    {
+      "epoch": 2.8388196721311476,
+      "grad_norm": 0.8493821372262771,
+      "learning_rate": 1.508133934247069e-06,
+      "loss": 11.8727,
+      "step": 16233
+    },
+    {
+      "epoch": 2.8389945355191255,
+      "grad_norm": 0.81390529242504,
+      "learning_rate": 1.504868954385008e-06,
+      "loss": 11.8638,
+      "step": 16234
+    },
+    {
+      "epoch": 2.839169398907104,
+      "grad_norm": 0.832904742940924,
+      "learning_rate": 1.5016074857592931e-06,
+      "loss": 11.5932,
+      "step": 16235
+    },
+    {
+      "epoch": 2.839344262295082,
+      "grad_norm": 0.8871443354547801,
+      "learning_rate": 1.4983495284861982e-06,
+      "loss": 11.6557,
+      "step": 16236
+    },
+    {
+      "epoch": 2.83951912568306,
+      "grad_norm": 0.7763476120917934,
+      "learning_rate": 1.4950950826818522e-06,
+      "loss": 11.7349,
+      "step": 16237
+    },
+    {
+      "epoch": 2.839693989071038,
+      "grad_norm": 0.8337599812450969,
+      "learning_rate": 1.4918441484622736e-06,
+      "loss": 11.8879,
+      "step": 16238
+    },
+    {
+      "epoch": 2.8398688524590163,
+      "grad_norm": 0.8288656613351648,
+      "learning_rate": 1.4885967259433475e-06,
+      "loss": 11.742,
+      "step": 16239
+    },
+    {
+      "epoch": 2.8400437158469947,
+      "grad_norm": 0.8380162040968786,
+      "learning_rate": 1.4853528152408592e-06,
+      "loss": 11.7939,
+      "step": 16240
+    },
+    {
+      "epoch": 2.8402185792349726,
+      "grad_norm": 0.8755377212045705,
+      "learning_rate": 1.4821124164704382e-06,
+      "loss": 11.8778,
+      "step": 16241
+    },
+    {
+      "epoch": 2.840393442622951,
+      "grad_norm": 0.8445327747895812,
+      "learning_rate": 1.4788755297476141e-06,
+      "loss": 12.0071,
+      "step": 16242
+    },
+    {
+      "epoch": 2.840568306010929,
+      "grad_norm": 0.8750743798201668,
+      "learning_rate": 1.4756421551877508e-06,
+      "loss": 11.8848,
+      "step": 16243
+    },
+    {
+      "epoch": 2.840743169398907,
+      "grad_norm": 0.8665807210717675,
+      "learning_rate": 1.4724122929061446e-06,
+      "loss": 11.8296,
+      "step": 16244
+    },
+    {
+      "epoch": 2.8409180327868855,
+      "grad_norm": 0.8701509148647146,
+      "learning_rate": 1.4691859430179145e-06,
+      "loss": 11.8782,
+      "step": 16245
+    },
+    {
+      "epoch": 2.8410928961748634,
+      "grad_norm": 0.771349667984444,
+      "learning_rate": 1.4659631056380908e-06,
+      "loss": 11.9779,
+      "step": 16246
+    },
+    {
+      "epoch": 2.8412677595628413,
+      "grad_norm": 0.7934382447112657,
+      "learning_rate": 1.4627437808815593e-06,
+      "loss": 11.75,
+      "step": 16247
+    },
+    {
+      "epoch": 2.8414426229508196,
+      "grad_norm": 0.8584620893174193,
+      "learning_rate": 1.4595279688630726e-06,
+      "loss": 11.8828,
+      "step": 16248
+    },
+    {
+      "epoch": 2.841617486338798,
+      "grad_norm": 0.8253383721192752,
+      "learning_rate": 1.4563156696972946e-06,
+      "loss": 11.8761,
+      "step": 16249
+    },
+    {
+      "epoch": 2.841792349726776,
+      "grad_norm": 0.8307016603645238,
+      "learning_rate": 1.4531068834987226e-06,
+      "loss": 11.7889,
+      "step": 16250
+    },
+    {
+      "epoch": 2.841967213114754,
+      "grad_norm": 0.8677349237082049,
+      "learning_rate": 1.4499016103817541e-06,
+      "loss": 11.895,
+      "step": 16251
+    },
+    {
+      "epoch": 2.842142076502732,
+      "grad_norm": 0.8200789090296966,
+      "learning_rate": 1.446699850460642e-06,
+      "loss": 11.9368,
+      "step": 16252
+    },
+    {
+      "epoch": 2.8423169398907104,
+      "grad_norm": 0.7833990674603862,
+      "learning_rate": 1.4435016038495285e-06,
+      "loss": 11.8992,
+      "step": 16253
+    },
+    {
+      "epoch": 2.8424918032786888,
+      "grad_norm": 0.8263528867597257,
+      "learning_rate": 1.4403068706624445e-06,
+      "loss": 12.0162,
+      "step": 16254
+    },
+    {
+      "epoch": 2.8426666666666667,
+      "grad_norm": 0.8272300182857654,
+      "learning_rate": 1.4371156510132655e-06,
+      "loss": 11.8599,
+      "step": 16255
+    },
+    {
+      "epoch": 2.8428415300546446,
+      "grad_norm": 0.8144065196410167,
+      "learning_rate": 1.433927945015745e-06,
+      "loss": 11.9825,
+      "step": 16256
+    },
+    {
+      "epoch": 2.843016393442623,
+      "grad_norm": 0.8034533490217045,
+      "learning_rate": 1.4307437527835366e-06,
+      "loss": 11.8469,
+      "step": 16257
+    },
+    {
+      "epoch": 2.8431912568306013,
+      "grad_norm": 0.8137420877302832,
+      "learning_rate": 1.4275630744301494e-06,
+      "loss": 11.9078,
+      "step": 16258
+    },
+    {
+      "epoch": 2.843366120218579,
+      "grad_norm": 0.7876873808538482,
+      "learning_rate": 1.4243859100689593e-06,
+      "loss": 11.9222,
+      "step": 16259
+    },
+    {
+      "epoch": 2.8435409836065575,
+      "grad_norm": 0.8495195052332907,
+      "learning_rate": 1.4212122598132539e-06,
+      "loss": 11.6798,
+      "step": 16260
+    },
+    {
+      "epoch": 2.8437158469945354,
+      "grad_norm": 0.8972065225209321,
+      "learning_rate": 1.418042123776142e-06,
+      "loss": 11.602,
+      "step": 16261
+    },
+    {
+      "epoch": 2.8438907103825137,
+      "grad_norm": 0.8115629021631422,
+      "learning_rate": 1.4148755020706562e-06,
+      "loss": 11.8823,
+      "step": 16262
+    },
+    {
+      "epoch": 2.844065573770492,
+      "grad_norm": 0.8353954329722,
+      "learning_rate": 1.4117123948096723e-06,
+      "loss": 11.9911,
+      "step": 16263
+    },
+    {
+      "epoch": 2.84424043715847,
+      "grad_norm": 0.815212456891734,
+      "learning_rate": 1.4085528021059446e-06,
+      "loss": 11.7365,
+      "step": 16264
+    },
+    {
+      "epoch": 2.844415300546448,
+      "grad_norm": 0.8198825769592382,
+      "learning_rate": 1.4053967240721277e-06,
+      "loss": 11.8971,
+      "step": 16265
+    },
+    {
+      "epoch": 2.844590163934426,
+      "grad_norm": 0.7906264577823807,
+      "learning_rate": 1.4022441608207205e-06,
+      "loss": 11.7726,
+      "step": 16266
+    },
+    {
+      "epoch": 2.8447650273224045,
+      "grad_norm": 0.8256572412972684,
+      "learning_rate": 1.3990951124641105e-06,
+      "loss": 11.8112,
+      "step": 16267
+    },
+    {
+      "epoch": 2.8449398907103824,
+      "grad_norm": 0.7642856417569955,
+      "learning_rate": 1.3959495791145527e-06,
+      "loss": 11.4022,
+      "step": 16268
+    },
+    {
+      "epoch": 2.845114754098361,
+      "grad_norm": 0.7877849856978087,
+      "learning_rate": 1.3928075608842017e-06,
+      "loss": 11.989,
+      "step": 16269
+    },
+    {
+      "epoch": 2.8452896174863387,
+      "grad_norm": 0.741339349231008,
+      "learning_rate": 1.3896690578850348e-06,
+      "loss": 11.8738,
+      "step": 16270
+    },
+    {
+      "epoch": 2.845464480874317,
+      "grad_norm": 0.7896420893964226,
+      "learning_rate": 1.3865340702289508e-06,
+      "loss": 11.9758,
+      "step": 16271
+    },
+    {
+      "epoch": 2.845639344262295,
+      "grad_norm": 0.864829990761678,
+      "learning_rate": 1.3834025980277165e-06,
+      "loss": 11.7485,
+      "step": 16272
+    },
+    {
+      "epoch": 2.8458142076502733,
+      "grad_norm": 0.8282136992920996,
+      "learning_rate": 1.3802746413929536e-06,
+      "loss": 11.7064,
+      "step": 16273
+    },
+    {
+      "epoch": 2.845989071038251,
+      "grad_norm": 0.854788026437016,
+      "learning_rate": 1.3771502004361835e-06,
+      "loss": 11.7756,
+      "step": 16274
+    },
+    {
+      "epoch": 2.8461639344262295,
+      "grad_norm": 0.8266502695079211,
+      "learning_rate": 1.3740292752687734e-06,
+      "loss": 11.9984,
+      "step": 16275
+    },
+    {
+      "epoch": 2.846338797814208,
+      "grad_norm": 0.8424277034017154,
+      "learning_rate": 1.3709118660019782e-06,
+      "loss": 11.7482,
+      "step": 16276
+    },
+    {
+      "epoch": 2.8465136612021857,
+      "grad_norm": 0.854075673517405,
+      "learning_rate": 1.3677979727469536e-06,
+      "loss": 11.6735,
+      "step": 16277
+    },
+    {
+      "epoch": 2.846688524590164,
+      "grad_norm": 0.8286850309514718,
+      "learning_rate": 1.3646875956146887e-06,
+      "loss": 11.7289,
+      "step": 16278
+    },
+    {
+      "epoch": 2.846863387978142,
+      "grad_norm": 0.8673530933255498,
+      "learning_rate": 1.3615807347160613e-06,
+      "loss": 11.6679,
+      "step": 16279
+    },
+    {
+      "epoch": 2.8470382513661203,
+      "grad_norm": 0.778873354277685,
+      "learning_rate": 1.3584773901618387e-06,
+      "loss": 11.8362,
+      "step": 16280
+    },
+    {
+      "epoch": 2.847213114754098,
+      "grad_norm": 0.8369579193469672,
+      "learning_rate": 1.355377562062654e-06,
+      "loss": 11.9507,
+      "step": 16281
+    },
+    {
+      "epoch": 2.8473879781420766,
+      "grad_norm": 0.8078918736484676,
+      "learning_rate": 1.3522812505289972e-06,
+      "loss": 11.7948,
+      "step": 16282
+    },
+    {
+      "epoch": 2.8475628415300545,
+      "grad_norm": 0.8143465391563818,
+      "learning_rate": 1.3491884556712576e-06,
+      "loss": 11.9446,
+      "step": 16283
+    },
+    {
+      "epoch": 2.847737704918033,
+      "grad_norm": 0.7995074201713408,
+      "learning_rate": 1.3460991775997023e-06,
+      "loss": 11.9337,
+      "step": 16284
+    },
+    {
+      "epoch": 2.847912568306011,
+      "grad_norm": 0.8268435462895758,
+      "learning_rate": 1.3430134164244324e-06,
+      "loss": 12.0538,
+      "step": 16285
+    },
+    {
+      "epoch": 2.848087431693989,
+      "grad_norm": 0.8472276286748511,
+      "learning_rate": 1.3399311722554597e-06,
+      "loss": 11.9333,
+      "step": 16286
+    },
+    {
+      "epoch": 2.8482622950819674,
+      "grad_norm": 0.8775751871441294,
+      "learning_rate": 1.3368524452026743e-06,
+      "loss": 11.8469,
+      "step": 16287
+    },
+    {
+      "epoch": 2.8484371584699453,
+      "grad_norm": 0.8489596222634049,
+      "learning_rate": 1.3337772353758327e-06,
+      "loss": 11.7827,
+      "step": 16288
+    },
+    {
+      "epoch": 2.8486120218579236,
+      "grad_norm": 0.8715213555218356,
+      "learning_rate": 1.3307055428845472e-06,
+      "loss": 11.5189,
+      "step": 16289
+    },
+    {
+      "epoch": 2.8487868852459015,
+      "grad_norm": 0.8308873457342977,
+      "learning_rate": 1.3276373678383192e-06,
+      "loss": 12.0099,
+      "step": 16290
+    },
+    {
+      "epoch": 2.84896174863388,
+      "grad_norm": 0.8904438161520836,
+      "learning_rate": 1.3245727103465389e-06,
+      "loss": 11.7773,
+      "step": 16291
+    },
+    {
+      "epoch": 2.8491366120218578,
+      "grad_norm": 0.8139540630376721,
+      "learning_rate": 1.3215115705184634e-06,
+      "loss": 11.9827,
+      "step": 16292
+    },
+    {
+      "epoch": 2.849311475409836,
+      "grad_norm": 0.8338470148063035,
+      "learning_rate": 1.3184539484631942e-06,
+      "loss": 11.8811,
+      "step": 16293
+    },
+    {
+      "epoch": 2.8494863387978144,
+      "grad_norm": 0.8229496895416127,
+      "learning_rate": 1.3153998442897441e-06,
+      "loss": 11.6787,
+      "step": 16294
+    },
+    {
+      "epoch": 2.8496612021857923,
+      "grad_norm": 0.8617316524043397,
+      "learning_rate": 1.3123492581069928e-06,
+      "loss": 11.7678,
+      "step": 16295
+    },
+    {
+      "epoch": 2.8498360655737702,
+      "grad_norm": 0.9112931259907858,
+      "learning_rate": 1.3093021900236756e-06,
+      "loss": 11.5746,
+      "step": 16296
+    },
+    {
+      "epoch": 2.8500109289617486,
+      "grad_norm": 0.859727607175371,
+      "learning_rate": 1.306258640148439e-06,
+      "loss": 11.6717,
+      "step": 16297
+    },
+    {
+      "epoch": 2.850185792349727,
+      "grad_norm": 0.8401658113072482,
+      "learning_rate": 1.3032186085897624e-06,
+      "loss": 11.8884,
+      "step": 16298
+    },
+    {
+      "epoch": 2.850360655737705,
+      "grad_norm": 0.8102477195848488,
+      "learning_rate": 1.3001820954560485e-06,
+      "loss": 11.9077,
+      "step": 16299
+    },
+    {
+      "epoch": 2.850535519125683,
+      "grad_norm": 0.8625345412828977,
+      "learning_rate": 1.2971491008554993e-06,
+      "loss": 11.9394,
+      "step": 16300
+    },
+    {
+      "epoch": 2.850710382513661,
+      "grad_norm": 0.8373811381205132,
+      "learning_rate": 1.2941196248962727e-06,
+      "loss": 11.7666,
+      "step": 16301
+    },
+    {
+      "epoch": 2.8508852459016394,
+      "grad_norm": 0.8255483646190473,
+      "learning_rate": 1.2910936676863606e-06,
+      "loss": 11.7373,
+      "step": 16302
+    },
+    {
+      "epoch": 2.8510601092896177,
+      "grad_norm": 0.8265867500286805,
+      "learning_rate": 1.2880712293336316e-06,
+      "loss": 11.6898,
+      "step": 16303
+    },
+    {
+      "epoch": 2.8512349726775956,
+      "grad_norm": 0.8131753222069193,
+      "learning_rate": 1.2850523099458223e-06,
+      "loss": 11.9999,
+      "step": 16304
+    },
+    {
+      "epoch": 2.8514098360655735,
+      "grad_norm": 0.8228593236273967,
+      "learning_rate": 1.2820369096305573e-06,
+      "loss": 11.8689,
+      "step": 16305
+    },
+    {
+      "epoch": 2.851584699453552,
+      "grad_norm": 0.8426550680558417,
+      "learning_rate": 1.279025028495351e-06,
+      "loss": 11.5801,
+      "step": 16306
+    },
+    {
+      "epoch": 2.85175956284153,
+      "grad_norm": 0.8581354888574589,
+      "learning_rate": 1.2760166666475614e-06,
+      "loss": 11.9377,
+      "step": 16307
+    },
+    {
+      "epoch": 2.851934426229508,
+      "grad_norm": 0.8180750350775493,
+      "learning_rate": 1.2730118241944144e-06,
+      "loss": 11.9654,
+      "step": 16308
+    },
+    {
+      "epoch": 2.8521092896174864,
+      "grad_norm": 0.9227508725559388,
+      "learning_rate": 1.2700105012430574e-06,
+      "loss": 11.7749,
+      "step": 16309
+    },
+    {
+      "epoch": 2.8522841530054643,
+      "grad_norm": 0.808268642390108,
+      "learning_rate": 1.2670126979004603e-06,
+      "loss": 11.897,
+      "step": 16310
+    },
+    {
+      "epoch": 2.8524590163934427,
+      "grad_norm": 0.7739882912934882,
+      "learning_rate": 1.2640184142735268e-06,
+      "loss": 11.9861,
+      "step": 16311
+    },
+    {
+      "epoch": 2.852633879781421,
+      "grad_norm": 0.8569820076160176,
+      "learning_rate": 1.26102765046896e-06,
+      "loss": 11.7733,
+      "step": 16312
+    },
+    {
+      "epoch": 2.852808743169399,
+      "grad_norm": 0.832597645434707,
+      "learning_rate": 1.2580404065933859e-06,
+      "loss": 12.1411,
+      "step": 16313
+    },
+    {
+      "epoch": 2.852983606557377,
+      "grad_norm": 0.920121276970357,
+      "learning_rate": 1.2550566827533194e-06,
+      "loss": 11.5996,
+      "step": 16314
+    },
+    {
+      "epoch": 2.853158469945355,
+      "grad_norm": 0.8255998930904771,
+      "learning_rate": 1.2520764790550975e-06,
+      "loss": 11.9272,
+      "step": 16315
+    },
+    {
+      "epoch": 2.8533333333333335,
+      "grad_norm": 0.8717902806216946,
+      "learning_rate": 1.24909979560498e-06,
+      "loss": 11.8659,
+      "step": 16316
+    },
+    {
+      "epoch": 2.8535081967213114,
+      "grad_norm": 0.8266794773960883,
+      "learning_rate": 1.2461266325090815e-06,
+      "loss": 11.8732,
+      "step": 16317
+    },
+    {
+      "epoch": 2.8536830601092897,
+      "grad_norm": 0.8284847656421606,
+      "learning_rate": 1.243156989873384e-06,
+      "loss": 11.7609,
+      "step": 16318
+    },
+    {
+      "epoch": 2.8538579234972676,
+      "grad_norm": 0.8077592741960985,
+      "learning_rate": 1.2401908678037589e-06,
+      "loss": 11.9761,
+      "step": 16319
+    },
+    {
+      "epoch": 2.854032786885246,
+      "grad_norm": 0.7900572176407794,
+      "learning_rate": 1.237228266405932e-06,
+      "loss": 11.8078,
+      "step": 16320
+    },
+    {
+      "epoch": 2.8542076502732243,
+      "grad_norm": 0.807748398342076,
+      "learning_rate": 1.2342691857855304e-06,
+      "loss": 11.9294,
+      "step": 16321
+    },
+    {
+      "epoch": 2.854382513661202,
+      "grad_norm": 0.7790371344777202,
+      "learning_rate": 1.2313136260480361e-06,
+      "loss": 11.9075,
+      "step": 16322
+    },
+    {
+      "epoch": 2.85455737704918,
+      "grad_norm": 0.8590664806908127,
+      "learning_rate": 1.2283615872988098e-06,
+      "loss": 11.9758,
+      "step": 16323
+    },
+    {
+      "epoch": 2.8547322404371585,
+      "grad_norm": 0.8285862523150866,
+      "learning_rate": 1.2254130696430998e-06,
+      "loss": 11.9484,
+      "step": 16324
+    },
+    {
+      "epoch": 2.854907103825137,
+      "grad_norm": 0.8311805784909809,
+      "learning_rate": 1.2224680731860005e-06,
+      "loss": 11.5798,
+      "step": 16325
+    },
+    {
+      "epoch": 2.8550819672131147,
+      "grad_norm": 0.8641333946880083,
+      "learning_rate": 1.2195265980325054e-06,
+      "loss": 11.7276,
+      "step": 16326
+    },
+    {
+      "epoch": 2.855256830601093,
+      "grad_norm": 0.855699024249133,
+      "learning_rate": 1.2165886442874752e-06,
+      "loss": 11.8199,
+      "step": 16327
+    },
+    {
+      "epoch": 2.855431693989071,
+      "grad_norm": 0.7980923237162766,
+      "learning_rate": 1.2136542120556482e-06,
+      "loss": 11.8116,
+      "step": 16328
+    },
+    {
+      "epoch": 2.8556065573770493,
+      "grad_norm": 0.8189721503026551,
+      "learning_rate": 1.2107233014416187e-06,
+      "loss": 11.8395,
+      "step": 16329
+    },
+    {
+      "epoch": 2.855781420765027,
+      "grad_norm": 0.8576757437475019,
+      "learning_rate": 1.2077959125498916e-06,
+      "loss": 12.0521,
+      "step": 16330
+    },
+    {
+      "epoch": 2.8559562841530055,
+      "grad_norm": 0.8093891657223871,
+      "learning_rate": 1.2048720454848173e-06,
+      "loss": 11.8706,
+      "step": 16331
+    },
+    {
+      "epoch": 2.8561311475409834,
+      "grad_norm": 0.810395417515377,
+      "learning_rate": 1.201951700350612e-06,
+      "loss": 11.9529,
+      "step": 16332
+    },
+    {
+      "epoch": 2.8563060109289617,
+      "grad_norm": 0.8883347719210122,
+      "learning_rate": 1.1990348772514037e-06,
+      "loss": 11.6523,
+      "step": 16333
+    },
+    {
+      "epoch": 2.85648087431694,
+      "grad_norm": 0.8545365017849298,
+      "learning_rate": 1.1961215762911538e-06,
+      "loss": 11.6876,
+      "step": 16334
+    },
+    {
+      "epoch": 2.856655737704918,
+      "grad_norm": 0.8797147710511023,
+      "learning_rate": 1.1932117975737346e-06,
+      "loss": 12.0477,
+      "step": 16335
+    },
+    {
+      "epoch": 2.8568306010928963,
+      "grad_norm": 0.8192167576204449,
+      "learning_rate": 1.190305541202874e-06,
+      "loss": 11.738,
+      "step": 16336
+    },
+    {
+      "epoch": 2.8570054644808742,
+      "grad_norm": 0.8029526877609259,
+      "learning_rate": 1.1874028072821896e-06,
+      "loss": 11.8914,
+      "step": 16337
+    },
+    {
+      "epoch": 2.8571803278688526,
+      "grad_norm": 0.8284092914407141,
+      "learning_rate": 1.1845035959151207e-06,
+      "loss": 11.6952,
+      "step": 16338
+    },
+    {
+      "epoch": 2.8573551912568305,
+      "grad_norm": 0.7811939428389323,
+      "learning_rate": 1.1816079072050623e-06,
+      "loss": 11.5032,
+      "step": 16339
+    },
+    {
+      "epoch": 2.857530054644809,
+      "grad_norm": 0.7915410603628505,
+      "learning_rate": 1.1787157412552096e-06,
+      "loss": 11.7839,
+      "step": 16340
+    },
+    {
+      "epoch": 2.8577049180327867,
+      "grad_norm": 0.869091655488294,
+      "learning_rate": 1.1758270981687026e-06,
+      "loss": 11.866,
+      "step": 16341
+    },
+    {
+      "epoch": 2.857879781420765,
+      "grad_norm": 0.8063398074868122,
+      "learning_rate": 1.1729419780484696e-06,
+      "loss": 11.9905,
+      "step": 16342
+    },
+    {
+      "epoch": 2.8580546448087434,
+      "grad_norm": 0.8478284609601949,
+      "learning_rate": 1.1700603809974064e-06,
+      "loss": 12.0231,
+      "step": 16343
+    },
+    {
+      "epoch": 2.8582295081967213,
+      "grad_norm": 0.8467943006536827,
+      "learning_rate": 1.1671823071182086e-06,
+      "loss": 11.8689,
+      "step": 16344
+    },
+    {
+      "epoch": 2.858404371584699,
+      "grad_norm": 0.7834750698676856,
+      "learning_rate": 1.1643077565135052e-06,
+      "loss": 11.5217,
+      "step": 16345
+    },
+    {
+      "epoch": 2.8585792349726775,
+      "grad_norm": 0.8308216258978391,
+      "learning_rate": 1.1614367292857363e-06,
+      "loss": 11.8459,
+      "step": 16346
+    },
+    {
+      "epoch": 2.858754098360656,
+      "grad_norm": 0.775969610300894,
+      "learning_rate": 1.158569225537276e-06,
+      "loss": 11.8313,
+      "step": 16347
+    },
+    {
+      "epoch": 2.8589289617486338,
+      "grad_norm": 0.7930830432211922,
+      "learning_rate": 1.155705245370331e-06,
+      "loss": 11.9381,
+      "step": 16348
+    },
+    {
+      "epoch": 2.859103825136612,
+      "grad_norm": 0.9242937458301685,
+      "learning_rate": 1.1528447888870087e-06,
+      "loss": 11.9574,
+      "step": 16349
+    },
+    {
+      "epoch": 2.85927868852459,
+      "grad_norm": 0.8224449972695493,
+      "learning_rate": 1.1499878561892828e-06,
+      "loss": 11.7013,
+      "step": 16350
+    },
+    {
+      "epoch": 2.8594535519125683,
+      "grad_norm": 0.8020103925163635,
+      "learning_rate": 1.1471344473789946e-06,
+      "loss": 11.6548,
+      "step": 16351
+    },
+    {
+      "epoch": 2.8596284153005467,
+      "grad_norm": 0.7803232037871186,
+      "learning_rate": 1.1442845625578736e-06,
+      "loss": 11.7264,
+      "step": 16352
+    },
+    {
+      "epoch": 2.8598032786885246,
+      "grad_norm": 0.8774002307652491,
+      "learning_rate": 1.1414382018274938e-06,
+      "loss": 11.7606,
+      "step": 16353
+    },
+    {
+      "epoch": 2.8599781420765025,
+      "grad_norm": 0.8431312820625003,
+      "learning_rate": 1.1385953652893412e-06,
+      "loss": 11.4725,
+      "step": 16354
+    },
+    {
+      "epoch": 2.860153005464481,
+      "grad_norm": 0.8116517038764267,
+      "learning_rate": 1.1357560530447675e-06,
+      "loss": 11.7803,
+      "step": 16355
+    },
+    {
+      "epoch": 2.860327868852459,
+      "grad_norm": 0.8178227127638172,
+      "learning_rate": 1.132920265194981e-06,
+      "loss": 11.6583,
+      "step": 16356
+    },
+    {
+      "epoch": 2.860502732240437,
+      "grad_norm": 0.8204952861697725,
+      "learning_rate": 1.1300880018410565e-06,
+      "loss": 11.6988,
+      "step": 16357
+    },
+    {
+      "epoch": 2.8606775956284154,
+      "grad_norm": 0.8332203821531932,
+      "learning_rate": 1.1272592630839906e-06,
+      "loss": 11.709,
+      "step": 16358
+    },
+    {
+      "epoch": 2.8608524590163933,
+      "grad_norm": 0.8015100872769846,
+      "learning_rate": 1.124434049024603e-06,
+      "loss": 11.9409,
+      "step": 16359
+    },
+    {
+      "epoch": 2.8610273224043716,
+      "grad_norm": 0.810404087305534,
+      "learning_rate": 1.1216123597636242e-06,
+      "loss": 11.7,
+      "step": 16360
+    },
+    {
+      "epoch": 2.86120218579235,
+      "grad_norm": 0.7932947561372221,
+      "learning_rate": 1.1187941954016402e-06,
+      "loss": 11.7055,
+      "step": 16361
+    },
+    {
+      "epoch": 2.861377049180328,
+      "grad_norm": 0.9081016032248155,
+      "learning_rate": 1.115979556039104e-06,
+      "loss": 11.5758,
+      "step": 16362
+    },
+    {
+      "epoch": 2.8615519125683058,
+      "grad_norm": 0.8542276965614417,
+      "learning_rate": 1.11316844177638e-06,
+      "loss": 11.8694,
+      "step": 16363
+    },
+    {
+      "epoch": 2.861726775956284,
+      "grad_norm": 0.822890837623642,
+      "learning_rate": 1.1103608527136433e-06,
+      "loss": 11.7337,
+      "step": 16364
+    },
+    {
+      "epoch": 2.8619016393442624,
+      "grad_norm": 0.7849864157242695,
+      "learning_rate": 1.107556788951014e-06,
+      "loss": 11.8109,
+      "step": 16365
+    },
+    {
+      "epoch": 2.8620765027322403,
+      "grad_norm": 0.836465124665044,
+      "learning_rate": 1.1047562505884345e-06,
+      "loss": 11.7965,
+      "step": 16366
+    },
+    {
+      "epoch": 2.8622513661202187,
+      "grad_norm": 0.8472117128199338,
+      "learning_rate": 1.101959237725747e-06,
+      "loss": 11.6342,
+      "step": 16367
+    },
+    {
+      "epoch": 2.8624262295081966,
+      "grad_norm": 0.8040130278364233,
+      "learning_rate": 1.0991657504626717e-06,
+      "loss": 11.8856,
+      "step": 16368
+    },
+    {
+      "epoch": 2.862601092896175,
+      "grad_norm": 0.8013031063543312,
+      "learning_rate": 1.0963757888987736e-06,
+      "loss": 11.8845,
+      "step": 16369
+    },
+    {
+      "epoch": 2.8627759562841533,
+      "grad_norm": 0.7604183782919703,
+      "learning_rate": 1.0935893531335284e-06,
+      "loss": 11.8699,
+      "step": 16370
+    },
+    {
+      "epoch": 2.862950819672131,
+      "grad_norm": 0.8658814757324963,
+      "learning_rate": 1.090806443266268e-06,
+      "loss": 11.7599,
+      "step": 16371
+    },
+    {
+      "epoch": 2.863125683060109,
+      "grad_norm": 0.8305120500116788,
+      "learning_rate": 1.0880270593961906e-06,
+      "loss": 11.874,
+      "step": 16372
+    },
+    {
+      "epoch": 2.8633005464480874,
+      "grad_norm": 0.8493268932535929,
+      "learning_rate": 1.0852512016223836e-06,
+      "loss": 11.948,
+      "step": 16373
+    },
+    {
+      "epoch": 2.8634754098360657,
+      "grad_norm": 0.9053428369415987,
+      "learning_rate": 1.0824788700438016e-06,
+      "loss": 11.6843,
+      "step": 16374
+    },
+    {
+      "epoch": 2.8636502732240436,
+      "grad_norm": 0.8009659412164388,
+      "learning_rate": 1.0797100647592762e-06,
+      "loss": 11.8546,
+      "step": 16375
+    },
+    {
+      "epoch": 2.863825136612022,
+      "grad_norm": 0.8694124038525826,
+      "learning_rate": 1.0769447858675063e-06,
+      "loss": 11.9136,
+      "step": 16376
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 0.8217216042112822,
+      "learning_rate": 1.07418303346708e-06,
+      "loss": 12.0355,
+      "step": 16377
+    },
+    {
+      "epoch": 2.864174863387978,
+      "grad_norm": 0.8087894670133982,
+      "learning_rate": 1.0714248076564403e-06,
+      "loss": 11.612,
+      "step": 16378
+    },
+    {
+      "epoch": 2.8643497267759566,
+      "grad_norm": 0.8126757387381781,
+      "learning_rate": 1.0686701085339313e-06,
+      "loss": 11.7897,
+      "step": 16379
+    },
+    {
+      "epoch": 2.8645245901639345,
+      "grad_norm": 0.822403152744339,
+      "learning_rate": 1.0659189361977296e-06,
+      "loss": 11.8058,
+      "step": 16380
+    },
+    {
+      "epoch": 2.8646994535519124,
+      "grad_norm": 0.8064381587064404,
+      "learning_rate": 1.0631712907459346e-06,
+      "loss": 11.6415,
+      "step": 16381
+    },
+    {
+      "epoch": 2.8648743169398907,
+      "grad_norm": 0.7646660293668933,
+      "learning_rate": 1.0604271722764903e-06,
+      "loss": 11.9431,
+      "step": 16382
+    },
+    {
+      "epoch": 2.865049180327869,
+      "grad_norm": 0.7705319129728784,
+      "learning_rate": 1.0576865808872184e-06,
+      "loss": 11.7514,
+      "step": 16383
+    },
+    {
+      "epoch": 2.865224043715847,
+      "grad_norm": 0.8131192393777636,
+      "learning_rate": 1.0549495166758072e-06,
+      "loss": 11.8244,
+      "step": 16384
+    },
+    {
+      "epoch": 2.8653989071038253,
+      "grad_norm": 0.8250330776246947,
+      "learning_rate": 1.0522159797398457e-06,
+      "loss": 11.8976,
+      "step": 16385
+    },
+    {
+      "epoch": 2.865573770491803,
+      "grad_norm": 0.8387933399636077,
+      "learning_rate": 1.049485970176778e-06,
+      "loss": 11.8433,
+      "step": 16386
+    },
+    {
+      "epoch": 2.8657486338797815,
+      "grad_norm": 0.8704414073573251,
+      "learning_rate": 1.046759488083926e-06,
+      "loss": 11.7266,
+      "step": 16387
+    },
+    {
+      "epoch": 2.8659234972677594,
+      "grad_norm": 0.7785498649212081,
+      "learning_rate": 1.044036533558479e-06,
+      "loss": 11.6788,
+      "step": 16388
+    },
+    {
+      "epoch": 2.8660983606557378,
+      "grad_norm": 0.7756222067066479,
+      "learning_rate": 1.0413171066975148e-06,
+      "loss": 11.4432,
+      "step": 16389
+    },
+    {
+      "epoch": 2.8662732240437157,
+      "grad_norm": 0.7982759013704642,
+      "learning_rate": 1.0386012075979889e-06,
+      "loss": 11.8777,
+      "step": 16390
+    },
+    {
+      "epoch": 2.866448087431694,
+      "grad_norm": 0.8555609039967447,
+      "learning_rate": 1.035888836356691e-06,
+      "loss": 11.8285,
+      "step": 16391
+    },
+    {
+      "epoch": 2.8666229508196723,
+      "grad_norm": 0.8472660561999648,
+      "learning_rate": 1.0331799930703324e-06,
+      "loss": 11.6825,
+      "step": 16392
+    },
+    {
+      "epoch": 2.8667978142076502,
+      "grad_norm": 0.8380460479124354,
+      "learning_rate": 1.0304746778354801e-06,
+      "loss": 11.8594,
+      "step": 16393
+    },
+    {
+      "epoch": 2.8669726775956286,
+      "grad_norm": 0.9433913281634381,
+      "learning_rate": 1.0277728907485574e-06,
+      "loss": 11.9146,
+      "step": 16394
+    },
+    {
+      "epoch": 2.8671475409836065,
+      "grad_norm": 0.7867907360591933,
+      "learning_rate": 1.025074631905909e-06,
+      "loss": 11.87,
+      "step": 16395
+    },
+    {
+      "epoch": 2.867322404371585,
+      "grad_norm": 0.8241778711959667,
+      "learning_rate": 1.0223799014037027e-06,
+      "loss": 11.919,
+      "step": 16396
+    },
+    {
+      "epoch": 2.8674972677595627,
+      "grad_norm": 0.8249112877291681,
+      "learning_rate": 1.0196886993380173e-06,
+      "loss": 11.7296,
+      "step": 16397
+    },
+    {
+      "epoch": 2.867672131147541,
+      "grad_norm": 0.8083090031252976,
+      "learning_rate": 1.0170010258047757e-06,
+      "loss": 12.0091,
+      "step": 16398
+    },
+    {
+      "epoch": 2.867846994535519,
+      "grad_norm": 0.7763755807333492,
+      "learning_rate": 1.0143168808998016e-06,
+      "loss": 11.9642,
+      "step": 16399
+    },
+    {
+      "epoch": 2.8680218579234973,
+      "grad_norm": 0.8179332865283545,
+      "learning_rate": 1.011636264718785e-06,
+      "loss": 11.839,
+      "step": 16400
+    },
+    {
+      "epoch": 2.8681967213114756,
+      "grad_norm": 0.8161499754653364,
+      "learning_rate": 1.0089591773572715e-06,
+      "loss": 11.7614,
+      "step": 16401
+    },
+    {
+      "epoch": 2.8683715846994535,
+      "grad_norm": 0.8299495322398421,
+      "learning_rate": 1.0062856189107073e-06,
+      "loss": 11.7098,
+      "step": 16402
+    },
+    {
+      "epoch": 2.8685464480874314,
+      "grad_norm": 0.8624147223540971,
+      "learning_rate": 1.0036155894744049e-06,
+      "loss": 11.8816,
+      "step": 16403
+    },
+    {
+      "epoch": 2.8687213114754098,
+      "grad_norm": 0.8429163763757302,
+      "learning_rate": 1.0009490891435324e-06,
+      "loss": 11.5444,
+      "step": 16404
+    },
+    {
+      "epoch": 2.868896174863388,
+      "grad_norm": 0.7933728316609038,
+      "learning_rate": 9.982861180131587e-07,
+      "loss": 11.5258,
+      "step": 16405
+    },
+    {
+      "epoch": 2.869071038251366,
+      "grad_norm": 0.7870259390869383,
+      "learning_rate": 9.956266761782073e-07,
+      "loss": 11.7041,
+      "step": 16406
+    },
+    {
+      "epoch": 2.8692459016393443,
+      "grad_norm": 0.919137191218168,
+      "learning_rate": 9.929707637335028e-07,
+      "loss": 11.7724,
+      "step": 16407
+    },
+    {
+      "epoch": 2.8694207650273222,
+      "grad_norm": 0.8099834056211996,
+      "learning_rate": 9.903183807737138e-07,
+      "loss": 11.8155,
+      "step": 16408
+    },
+    {
+      "epoch": 2.8695956284153006,
+      "grad_norm": 0.7832298167511523,
+      "learning_rate": 9.876695273933866e-07,
+      "loss": 11.6964,
+      "step": 16409
+    },
+    {
+      "epoch": 2.869770491803279,
+      "grad_norm": 0.8868896229654226,
+      "learning_rate": 9.850242036869572e-07,
+      "loss": 11.9195,
+      "step": 16410
+    },
+    {
+      "epoch": 2.869945355191257,
+      "grad_norm": 0.7799932843707521,
+      "learning_rate": 9.82382409748739e-07,
+      "loss": 11.915,
+      "step": 16411
+    },
+    {
+      "epoch": 2.8701202185792347,
+      "grad_norm": 0.7763224601999306,
+      "learning_rate": 9.797441456728896e-07,
+      "loss": 11.6459,
+      "step": 16412
+    },
+    {
+      "epoch": 2.870295081967213,
+      "grad_norm": 0.8633588728250479,
+      "learning_rate": 9.771094115534673e-07,
+      "loss": 11.7408,
+      "step": 16413
+    },
+    {
+      "epoch": 2.8704699453551914,
+      "grad_norm": 0.8530264303298759,
+      "learning_rate": 9.744782074844084e-07,
+      "loss": 11.8546,
+      "step": 16414
+    },
+    {
+      "epoch": 2.8706448087431693,
+      "grad_norm": 0.8829945273246533,
+      "learning_rate": 9.71850533559493e-07,
+      "loss": 11.9848,
+      "step": 16415
+    },
+    {
+      "epoch": 2.8708196721311476,
+      "grad_norm": 0.8454586835773971,
+      "learning_rate": 9.692263898724019e-07,
+      "loss": 11.7607,
+      "step": 16416
+    },
+    {
+      "epoch": 2.8709945355191255,
+      "grad_norm": 0.8519754634692234,
+      "learning_rate": 9.666057765166937e-07,
+      "loss": 11.9696,
+      "step": 16417
+    },
+    {
+      "epoch": 2.871169398907104,
+      "grad_norm": 0.821925007059319,
+      "learning_rate": 9.63988693585771e-07,
+      "loss": 11.6511,
+      "step": 16418
+    },
+    {
+      "epoch": 2.871344262295082,
+      "grad_norm": 0.8245805398226563,
+      "learning_rate": 9.613751411729488e-07,
+      "loss": 11.8969,
+      "step": 16419
+    },
+    {
+      "epoch": 2.87151912568306,
+      "grad_norm": 0.765037184681705,
+      "learning_rate": 9.587651193713964e-07,
+      "loss": 11.7363,
+      "step": 16420
+    },
+    {
+      "epoch": 2.871693989071038,
+      "grad_norm": 0.8226535158080627,
+      "learning_rate": 9.561586282741398e-07,
+      "loss": 12.1721,
+      "step": 16421
+    },
+    {
+      "epoch": 2.8718688524590164,
+      "grad_norm": 0.7961230108712716,
+      "learning_rate": 9.535556679741043e-07,
+      "loss": 11.6847,
+      "step": 16422
+    },
+    {
+      "epoch": 2.8720437158469947,
+      "grad_norm": 0.9467887015488918,
+      "learning_rate": 9.509562385640825e-07,
+      "loss": 12.0103,
+      "step": 16423
+    },
+    {
+      "epoch": 2.8722185792349726,
+      "grad_norm": 0.796929280695322,
+      "learning_rate": 9.483603401367558e-07,
+      "loss": 11.7699,
+      "step": 16424
+    },
+    {
+      "epoch": 2.872393442622951,
+      "grad_norm": 0.8049202850132261,
+      "learning_rate": 9.457679727846502e-07,
+      "loss": 11.8001,
+      "step": 16425
+    },
+    {
+      "epoch": 2.872568306010929,
+      "grad_norm": 0.8393195796107369,
+      "learning_rate": 9.431791366001808e-07,
+      "loss": 11.907,
+      "step": 16426
+    },
+    {
+      "epoch": 2.872743169398907,
+      "grad_norm": 0.7824983620404689,
+      "learning_rate": 9.405938316756402e-07,
+      "loss": 11.8707,
+      "step": 16427
+    },
+    {
+      "epoch": 2.8729180327868855,
+      "grad_norm": 0.8017356352270201,
+      "learning_rate": 9.380120581031993e-07,
+      "loss": 11.7813,
+      "step": 16428
+    },
+    {
+      "epoch": 2.8730928961748634,
+      "grad_norm": 0.8413340852234095,
+      "learning_rate": 9.354338159748732e-07,
+      "loss": 11.6379,
+      "step": 16429
+    },
+    {
+      "epoch": 2.8732677595628413,
+      "grad_norm": 0.8351130410050953,
+      "learning_rate": 9.328591053825997e-07,
+      "loss": 12.0049,
+      "step": 16430
+    },
+    {
+      "epoch": 2.8734426229508196,
+      "grad_norm": 0.8145007309956814,
+      "learning_rate": 9.302879264181496e-07,
+      "loss": 12.0385,
+      "step": 16431
+    },
+    {
+      "epoch": 2.873617486338798,
+      "grad_norm": 0.8204883624613228,
+      "learning_rate": 9.277202791731721e-07,
+      "loss": 11.8118,
+      "step": 16432
+    },
+    {
+      "epoch": 2.873792349726776,
+      "grad_norm": 0.8087254160376511,
+      "learning_rate": 9.25156163739227e-07,
+      "loss": 12.0061,
+      "step": 16433
+    },
+    {
+      "epoch": 2.8739672131147542,
+      "grad_norm": 0.8477385968029024,
+      "learning_rate": 9.225955802076968e-07,
+      "loss": 11.7589,
+      "step": 16434
+    },
+    {
+      "epoch": 2.874142076502732,
+      "grad_norm": 0.9096129693148679,
+      "learning_rate": 9.200385286698865e-07,
+      "loss": 11.7556,
+      "step": 16435
+    },
+    {
+      "epoch": 2.8743169398907105,
+      "grad_norm": 0.8353283878810638,
+      "learning_rate": 9.174850092169341e-07,
+      "loss": 11.6866,
+      "step": 16436
+    },
+    {
+      "epoch": 2.8744918032786884,
+      "grad_norm": 0.822842685022766,
+      "learning_rate": 9.149350219398778e-07,
+      "loss": 11.8351,
+      "step": 16437
+    },
+    {
+      "epoch": 2.8746666666666667,
+      "grad_norm": 0.7969250470416865,
+      "learning_rate": 9.123885669296117e-07,
+      "loss": 11.8632,
+      "step": 16438
+    },
+    {
+      "epoch": 2.8748415300546446,
+      "grad_norm": 0.8023549227388022,
+      "learning_rate": 9.098456442769298e-07,
+      "loss": 11.9255,
+      "step": 16439
+    },
+    {
+      "epoch": 2.875016393442623,
+      "grad_norm": 0.8203361953430759,
+      "learning_rate": 9.073062540724709e-07,
+      "loss": 11.7638,
+      "step": 16440
+    },
+    {
+      "epoch": 2.8751912568306013,
+      "grad_norm": 0.7667164523380345,
+      "learning_rate": 9.047703964067733e-07,
+      "loss": 11.826,
+      "step": 16441
+    },
+    {
+      "epoch": 2.875366120218579,
+      "grad_norm": 0.8348894650424736,
+      "learning_rate": 9.022380713702316e-07,
+      "loss": 11.7326,
+      "step": 16442
+    },
+    {
+      "epoch": 2.8755409836065575,
+      "grad_norm": 0.8277886217550489,
+      "learning_rate": 8.997092790531181e-07,
+      "loss": 11.7373,
+      "step": 16443
+    },
+    {
+      "epoch": 2.8757158469945354,
+      "grad_norm": 0.812547729093721,
+      "learning_rate": 8.971840195455716e-07,
+      "loss": 11.7938,
+      "step": 16444
+    },
+    {
+      "epoch": 2.8758907103825138,
+      "grad_norm": 0.839786328072859,
+      "learning_rate": 8.946622929376314e-07,
+      "loss": 11.9588,
+      "step": 16445
+    },
+    {
+      "epoch": 2.8760655737704917,
+      "grad_norm": 0.8583461471981705,
+      "learning_rate": 8.921440993192032e-07,
+      "loss": 11.7267,
+      "step": 16446
+    },
+    {
+      "epoch": 2.87624043715847,
+      "grad_norm": 0.7964729883111639,
+      "learning_rate": 8.896294387800264e-07,
+      "loss": 11.777,
+      "step": 16447
+    },
+    {
+      "epoch": 2.876415300546448,
+      "grad_norm": 0.9108125622850477,
+      "learning_rate": 8.871183114097625e-07,
+      "loss": 11.8914,
+      "step": 16448
+    },
+    {
+      "epoch": 2.8765901639344262,
+      "grad_norm": 0.9908766376877637,
+      "learning_rate": 8.84610717297929e-07,
+      "loss": 11.7734,
+      "step": 16449
+    },
+    {
+      "epoch": 2.8767650273224046,
+      "grad_norm": 0.8138973756755322,
+      "learning_rate": 8.82106656533932e-07,
+      "loss": 11.7656,
+      "step": 16450
+    },
+    {
+      "epoch": 2.8769398907103825,
+      "grad_norm": 0.8696442822857977,
+      "learning_rate": 8.796061292070224e-07,
+      "loss": 11.9231,
+      "step": 16451
+    },
+    {
+      "epoch": 2.877114754098361,
+      "grad_norm": 0.8125606676689533,
+      "learning_rate": 8.771091354063287e-07,
+      "loss": 11.8444,
+      "step": 16452
+    },
+    {
+      "epoch": 2.8772896174863387,
+      "grad_norm": 0.8863007725957851,
+      "learning_rate": 8.746156752209022e-07,
+      "loss": 11.7811,
+      "step": 16453
+    },
+    {
+      "epoch": 2.877464480874317,
+      "grad_norm": 0.866949381943481,
+      "learning_rate": 8.72125748739605e-07,
+      "loss": 11.8512,
+      "step": 16454
+    },
+    {
+      "epoch": 2.877639344262295,
+      "grad_norm": 0.8157393702612493,
+      "learning_rate": 8.696393560511995e-07,
+      "loss": 11.7442,
+      "step": 16455
+    },
+    {
+      "epoch": 2.8778142076502733,
+      "grad_norm": 0.8539004246926196,
+      "learning_rate": 8.671564972443258e-07,
+      "loss": 11.9535,
+      "step": 16456
+    },
+    {
+      "epoch": 2.877989071038251,
+      "grad_norm": 0.8610465929218132,
+      "learning_rate": 8.646771724075131e-07,
+      "loss": 11.921,
+      "step": 16457
+    },
+    {
+      "epoch": 2.8781639344262295,
+      "grad_norm": 0.8473142511958486,
+      "learning_rate": 8.622013816291241e-07,
+      "loss": 11.9101,
+      "step": 16458
+    },
+    {
+      "epoch": 2.878338797814208,
+      "grad_norm": 0.7924114384521214,
+      "learning_rate": 8.597291249974215e-07,
+      "loss": 12.054,
+      "step": 16459
+    },
+    {
+      "epoch": 2.8785136612021858,
+      "grad_norm": 0.8059093644537957,
+      "learning_rate": 8.57260402600546e-07,
+      "loss": 11.6139,
+      "step": 16460
+    },
+    {
+      "epoch": 2.8786885245901637,
+      "grad_norm": 0.8668694992873535,
+      "learning_rate": 8.547952145265048e-07,
+      "loss": 11.8363,
+      "step": 16461
+    },
+    {
+      "epoch": 2.878863387978142,
+      "grad_norm": 0.8740958961044484,
+      "learning_rate": 8.523335608631722e-07,
+      "loss": 11.8247,
+      "step": 16462
+    },
+    {
+      "epoch": 2.8790382513661203,
+      "grad_norm": 0.8113463509752764,
+      "learning_rate": 8.498754416983112e-07,
+      "loss": 11.8691,
+      "step": 16463
+    },
+    {
+      "epoch": 2.8792131147540982,
+      "grad_norm": 0.8343829355270972,
+      "learning_rate": 8.474208571195408e-07,
+      "loss": 11.6983,
+      "step": 16464
+    },
+    {
+      "epoch": 2.8793879781420766,
+      "grad_norm": 0.8199226062674074,
+      "learning_rate": 8.449698072143797e-07,
+      "loss": 11.7037,
+      "step": 16465
+    },
+    {
+      "epoch": 2.8795628415300545,
+      "grad_norm": 0.954240028434136,
+      "learning_rate": 8.425222920701803e-07,
+      "loss": 11.9584,
+      "step": 16466
+    },
+    {
+      "epoch": 2.879737704918033,
+      "grad_norm": 0.8118969945853555,
+      "learning_rate": 8.400783117742172e-07,
+      "loss": 11.8938,
+      "step": 16467
+    },
+    {
+      "epoch": 2.879912568306011,
+      "grad_norm": 0.8519945130113418,
+      "learning_rate": 8.376378664135986e-07,
+      "loss": 11.5826,
+      "step": 16468
+    },
+    {
+      "epoch": 2.880087431693989,
+      "grad_norm": 0.8083105003329851,
+      "learning_rate": 8.352009560753437e-07,
+      "loss": 11.8346,
+      "step": 16469
+    },
+    {
+      "epoch": 2.880262295081967,
+      "grad_norm": 0.798937787165178,
+      "learning_rate": 8.327675808463054e-07,
+      "loss": 11.9614,
+      "step": 16470
+    },
+    {
+      "epoch": 2.8804371584699453,
+      "grad_norm": 0.836183364952261,
+      "learning_rate": 8.303377408132362e-07,
+      "loss": 11.699,
+      "step": 16471
+    },
+    {
+      "epoch": 2.8806120218579236,
+      "grad_norm": 0.8122180784397358,
+      "learning_rate": 8.279114360627671e-07,
+      "loss": 11.6721,
+      "step": 16472
+    },
+    {
+      "epoch": 2.8807868852459015,
+      "grad_norm": 0.8095076314168891,
+      "learning_rate": 8.254886666813844e-07,
+      "loss": 11.9301,
+      "step": 16473
+    },
+    {
+      "epoch": 2.88096174863388,
+      "grad_norm": 0.8266830918480491,
+      "learning_rate": 8.230694327554522e-07,
+      "loss": 11.9542,
+      "step": 16474
+    },
+    {
+      "epoch": 2.881136612021858,
+      "grad_norm": 0.8193272095336239,
+      "learning_rate": 8.206537343712128e-07,
+      "loss": 11.6789,
+      "step": 16475
+    },
+    {
+      "epoch": 2.881311475409836,
+      "grad_norm": 0.7758387079283148,
+      "learning_rate": 8.182415716147974e-07,
+      "loss": 11.8864,
+      "step": 16476
+    },
+    {
+      "epoch": 2.8814863387978145,
+      "grad_norm": 0.8279269287161828,
+      "learning_rate": 8.158329445721924e-07,
+      "loss": 11.7972,
+      "step": 16477
+    },
+    {
+      "epoch": 2.8816612021857924,
+      "grad_norm": 0.778123283252979,
+      "learning_rate": 8.134278533292516e-07,
+      "loss": 11.9131,
+      "step": 16478
+    },
+    {
+      "epoch": 2.8818360655737703,
+      "grad_norm": 0.858797938428298,
+      "learning_rate": 8.110262979717176e-07,
+      "loss": 11.8372,
+      "step": 16479
+    },
+    {
+      "epoch": 2.8820109289617486,
+      "grad_norm": 0.8452925973499412,
+      "learning_rate": 8.086282785852217e-07,
+      "loss": 11.6843,
+      "step": 16480
+    },
+    {
+      "epoch": 2.882185792349727,
+      "grad_norm": 0.852976927891676,
+      "learning_rate": 8.062337952552179e-07,
+      "loss": 11.7875,
+      "step": 16481
+    },
+    {
+      "epoch": 2.882360655737705,
+      "grad_norm": 0.8668914500447891,
+      "learning_rate": 8.038428480670934e-07,
+      "loss": 11.8623,
+      "step": 16482
+    },
+    {
+      "epoch": 2.882535519125683,
+      "grad_norm": 0.8287830005622822,
+      "learning_rate": 8.01455437106069e-07,
+      "loss": 11.5987,
+      "step": 16483
+    },
+    {
+      "epoch": 2.882710382513661,
+      "grad_norm": 0.7985695686745518,
+      "learning_rate": 7.990715624572653e-07,
+      "loss": 11.9023,
+      "step": 16484
+    },
+    {
+      "epoch": 2.8828852459016394,
+      "grad_norm": 0.7486857919111977,
+      "learning_rate": 7.966912242056479e-07,
+      "loss": 11.5494,
+      "step": 16485
+    },
+    {
+      "epoch": 2.8830601092896178,
+      "grad_norm": 0.8507188623281694,
+      "learning_rate": 7.943144224360821e-07,
+      "loss": 11.8023,
+      "step": 16486
+    },
+    {
+      "epoch": 2.8832349726775957,
+      "grad_norm": 0.8640721359361401,
+      "learning_rate": 7.919411572333002e-07,
+      "loss": 11.9283,
+      "step": 16487
+    },
+    {
+      "epoch": 2.8834098360655736,
+      "grad_norm": 0.8562785757087584,
+      "learning_rate": 7.895714286819012e-07,
+      "loss": 12.0566,
+      "step": 16488
+    },
+    {
+      "epoch": 2.883584699453552,
+      "grad_norm": 0.8642329273343842,
+      "learning_rate": 7.872052368663729e-07,
+      "loss": 11.6738,
+      "step": 16489
+    },
+    {
+      "epoch": 2.8837595628415302,
+      "grad_norm": 0.8302866598638422,
+      "learning_rate": 7.848425818710481e-07,
+      "loss": 11.8661,
+      "step": 16490
+    },
+    {
+      "epoch": 2.883934426229508,
+      "grad_norm": 0.8305222100685113,
+      "learning_rate": 7.824834637801815e-07,
+      "loss": 11.6816,
+      "step": 16491
+    },
+    {
+      "epoch": 2.8841092896174865,
+      "grad_norm": 0.8109217873729382,
+      "learning_rate": 7.80127882677839e-07,
+      "loss": 12.1283,
+      "step": 16492
+    },
+    {
+      "epoch": 2.8842841530054644,
+      "grad_norm": 0.858169411556592,
+      "learning_rate": 7.777758386480206e-07,
+      "loss": 11.8556,
+      "step": 16493
+    },
+    {
+      "epoch": 2.8844590163934427,
+      "grad_norm": 0.8349600454370142,
+      "learning_rate": 7.754273317745586e-07,
+      "loss": 11.8163,
+      "step": 16494
+    },
+    {
+      "epoch": 2.8846338797814206,
+      "grad_norm": 0.795282390302048,
+      "learning_rate": 7.730823621411864e-07,
+      "loss": 12.0025,
+      "step": 16495
+    },
+    {
+      "epoch": 2.884808743169399,
+      "grad_norm": 0.8133906355209437,
+      "learning_rate": 7.707409298314927e-07,
+      "loss": 11.5708,
+      "step": 16496
+    },
+    {
+      "epoch": 2.884983606557377,
+      "grad_norm": 0.841922494545897,
+      "learning_rate": 7.684030349289439e-07,
+      "loss": 11.6113,
+      "step": 16497
+    },
+    {
+      "epoch": 2.885158469945355,
+      "grad_norm": 0.8491338609655849,
+      "learning_rate": 7.660686775168957e-07,
+      "loss": 12.0122,
+      "step": 16498
+    },
+    {
+      "epoch": 2.8853333333333335,
+      "grad_norm": 0.8272088057475848,
+      "learning_rate": 7.637378576785592e-07,
+      "loss": 11.9669,
+      "step": 16499
+    },
+    {
+      "epoch": 2.8855081967213114,
+      "grad_norm": 0.8658151404705833,
+      "learning_rate": 7.614105754970124e-07,
+      "loss": 11.8306,
+      "step": 16500
+    },
+    {
+      "epoch": 2.8856830601092898,
+      "grad_norm": 0.864970660307742,
+      "learning_rate": 7.590868310552335e-07,
+      "loss": 11.955,
+      "step": 16501
+    },
+    {
+      "epoch": 2.8858579234972677,
+      "grad_norm": 0.8083643483398851,
+      "learning_rate": 7.567666244360671e-07,
+      "loss": 11.9738,
+      "step": 16502
+    },
+    {
+      "epoch": 2.886032786885246,
+      "grad_norm": 0.8177613137733184,
+      "learning_rate": 7.544499557222029e-07,
+      "loss": 11.8307,
+      "step": 16503
+    },
+    {
+      "epoch": 2.886207650273224,
+      "grad_norm": 0.8147743206586037,
+      "learning_rate": 7.521368249962523e-07,
+      "loss": 11.868,
+      "step": 16504
+    },
+    {
+      "epoch": 2.8863825136612022,
+      "grad_norm": 0.7856941371959844,
+      "learning_rate": 7.498272323406608e-07,
+      "loss": 11.812,
+      "step": 16505
+    },
+    {
+      "epoch": 2.88655737704918,
+      "grad_norm": 0.7937267374088196,
+      "learning_rate": 7.475211778377733e-07,
+      "loss": 11.9522,
+      "step": 16506
+    },
+    {
+      "epoch": 2.8867322404371585,
+      "grad_norm": 0.8397960433476317,
+      "learning_rate": 7.452186615697798e-07,
+      "loss": 11.5726,
+      "step": 16507
+    },
+    {
+      "epoch": 2.886907103825137,
+      "grad_norm": 0.8379221956080942,
+      "learning_rate": 7.429196836187812e-07,
+      "loss": 11.8399,
+      "step": 16508
+    },
+    {
+      "epoch": 2.8870819672131147,
+      "grad_norm": 0.8204221047214099,
+      "learning_rate": 7.406242440667232e-07,
+      "loss": 11.9961,
+      "step": 16509
+    },
+    {
+      "epoch": 2.887256830601093,
+      "grad_norm": 0.8177709307243047,
+      "learning_rate": 7.383323429954514e-07,
+      "loss": 11.601,
+      "step": 16510
+    },
+    {
+      "epoch": 2.887431693989071,
+      "grad_norm": 0.7942601719106728,
+      "learning_rate": 7.360439804866448e-07,
+      "loss": 11.3929,
+      "step": 16511
+    },
+    {
+      "epoch": 2.8876065573770493,
+      "grad_norm": 0.8681385449799717,
+      "learning_rate": 7.337591566218937e-07,
+      "loss": 11.8362,
+      "step": 16512
+    },
+    {
+      "epoch": 2.887781420765027,
+      "grad_norm": 0.7899153698278437,
+      "learning_rate": 7.314778714826442e-07,
+      "loss": 11.9141,
+      "step": 16513
+    },
+    {
+      "epoch": 2.8879562841530055,
+      "grad_norm": 0.797633801668413,
+      "learning_rate": 7.292001251502312e-07,
+      "loss": 12.001,
+      "step": 16514
+    },
+    {
+      "epoch": 2.8881311475409834,
+      "grad_norm": 0.7885570353942518,
+      "learning_rate": 7.269259177058451e-07,
+      "loss": 11.8603,
+      "step": 16515
+    },
+    {
+      "epoch": 2.8883060109289618,
+      "grad_norm": 0.9319516428785821,
+      "learning_rate": 7.246552492305659e-07,
+      "loss": 11.8117,
+      "step": 16516
+    },
+    {
+      "epoch": 2.88848087431694,
+      "grad_norm": 0.8721601837474433,
+      "learning_rate": 7.223881198053395e-07,
+      "loss": 11.8079,
+      "step": 16517
+    },
+    {
+      "epoch": 2.888655737704918,
+      "grad_norm": 0.8064099742479198,
+      "learning_rate": 7.201245295109793e-07,
+      "loss": 11.8626,
+      "step": 16518
+    },
+    {
+      "epoch": 2.888830601092896,
+      "grad_norm": 0.8511873704965383,
+      "learning_rate": 7.178644784281763e-07,
+      "loss": 11.8483,
+      "step": 16519
+    },
+    {
+      "epoch": 2.8890054644808743,
+      "grad_norm": 0.7907459587963983,
+      "learning_rate": 7.156079666375215e-07,
+      "loss": 11.6628,
+      "step": 16520
+    },
+    {
+      "epoch": 2.8891803278688526,
+      "grad_norm": 0.8193968146360129,
+      "learning_rate": 7.133549942194285e-07,
+      "loss": 12.0244,
+      "step": 16521
+    },
+    {
+      "epoch": 2.8893551912568305,
+      "grad_norm": 0.8293292146586232,
+      "learning_rate": 7.111055612542328e-07,
+      "loss": 11.9868,
+      "step": 16522
+    },
+    {
+      "epoch": 2.889530054644809,
+      "grad_norm": 0.8330769811308862,
+      "learning_rate": 7.088596678221148e-07,
+      "loss": 11.8267,
+      "step": 16523
+    },
+    {
+      "epoch": 2.8897049180327867,
+      "grad_norm": 0.8268275853489333,
+      "learning_rate": 7.06617314003144e-07,
+      "loss": 11.797,
+      "step": 16524
+    },
+    {
+      "epoch": 2.889879781420765,
+      "grad_norm": 0.8317447783635328,
+      "learning_rate": 7.043784998772451e-07,
+      "loss": 11.9868,
+      "step": 16525
+    },
+    {
+      "epoch": 2.8900546448087434,
+      "grad_norm": 0.8608593440467721,
+      "learning_rate": 7.021432255242433e-07,
+      "loss": 11.9498,
+      "step": 16526
+    },
+    {
+      "epoch": 2.8902295081967213,
+      "grad_norm": 0.7857481475352233,
+      "learning_rate": 6.999114910238191e-07,
+      "loss": 11.9281,
+      "step": 16527
+    },
+    {
+      "epoch": 2.890404371584699,
+      "grad_norm": 1.0141618513603659,
+      "learning_rate": 6.976832964555313e-07,
+      "loss": 11.7937,
+      "step": 16528
+    },
+    {
+      "epoch": 2.8905792349726775,
+      "grad_norm": 0.8270480256319858,
+      "learning_rate": 6.954586418988051e-07,
+      "loss": 11.9181,
+      "step": 16529
+    },
+    {
+      "epoch": 2.890754098360656,
+      "grad_norm": 0.8006399358020311,
+      "learning_rate": 6.932375274329661e-07,
+      "loss": 11.9929,
+      "step": 16530
+    },
+    {
+      "epoch": 2.890928961748634,
+      "grad_norm": 0.9652699055758982,
+      "learning_rate": 6.910199531371731e-07,
+      "loss": 11.9191,
+      "step": 16531
+    },
+    {
+      "epoch": 2.891103825136612,
+      "grad_norm": 0.89851049380192,
+      "learning_rate": 6.88805919090485e-07,
+      "loss": 11.7366,
+      "step": 16532
+    },
+    {
+      "epoch": 2.89127868852459,
+      "grad_norm": 0.8613160210113543,
+      "learning_rate": 6.865954253718387e-07,
+      "loss": 11.8452,
+      "step": 16533
+    },
+    {
+      "epoch": 2.8914535519125684,
+      "grad_norm": 0.8487308722129577,
+      "learning_rate": 6.84388472060038e-07,
+      "loss": 11.8597,
+      "step": 16534
+    },
+    {
+      "epoch": 2.8916284153005467,
+      "grad_norm": 0.8391932842857172,
+      "learning_rate": 6.821850592337309e-07,
+      "loss": 11.9444,
+      "step": 16535
+    },
+    {
+      "epoch": 2.8918032786885246,
+      "grad_norm": 0.8961013033684483,
+      "learning_rate": 6.799851869715102e-07,
+      "loss": 11.7726,
+      "step": 16536
+    },
+    {
+      "epoch": 2.8919781420765025,
+      "grad_norm": 0.8201447972281859,
+      "learning_rate": 6.777888553517575e-07,
+      "loss": 12.0433,
+      "step": 16537
+    },
+    {
+      "epoch": 2.892153005464481,
+      "grad_norm": 0.8445150128288915,
+      "learning_rate": 6.755960644527881e-07,
+      "loss": 11.7324,
+      "step": 16538
+    },
+    {
+      "epoch": 2.892327868852459,
+      "grad_norm": 0.8753923907105894,
+      "learning_rate": 6.734068143527727e-07,
+      "loss": 11.7806,
+      "step": 16539
+    },
+    {
+      "epoch": 2.892502732240437,
+      "grad_norm": 0.8666157995374647,
+      "learning_rate": 6.712211051297601e-07,
+      "loss": 11.7981,
+      "step": 16540
+    },
+    {
+      "epoch": 2.8926775956284154,
+      "grad_norm": 0.8128193063621056,
+      "learning_rate": 6.690389368616545e-07,
+      "loss": 11.745,
+      "step": 16541
+    },
+    {
+      "epoch": 2.8928524590163933,
+      "grad_norm": 0.8145312571361274,
+      "learning_rate": 6.668603096262604e-07,
+      "loss": 11.7179,
+      "step": 16542
+    },
+    {
+      "epoch": 2.8930273224043717,
+      "grad_norm": 0.845773860246359,
+      "learning_rate": 6.646852235012268e-07,
+      "loss": 11.8089,
+      "step": 16543
+    },
+    {
+      "epoch": 2.89320218579235,
+      "grad_norm": 0.7818955854712936,
+      "learning_rate": 6.62513678564114e-07,
+      "loss": 11.9323,
+      "step": 16544
+    },
+    {
+      "epoch": 2.893377049180328,
+      "grad_norm": 0.8525860839144848,
+      "learning_rate": 6.603456748923265e-07,
+      "loss": 11.55,
+      "step": 16545
+    },
+    {
+      "epoch": 2.893551912568306,
+      "grad_norm": 0.8652630829186698,
+      "learning_rate": 6.58181212563147e-07,
+      "loss": 11.7588,
+      "step": 16546
+    },
+    {
+      "epoch": 2.893726775956284,
+      "grad_norm": 0.8352287021781185,
+      "learning_rate": 6.560202916537362e-07,
+      "loss": 11.7703,
+      "step": 16547
+    },
+    {
+      "epoch": 2.8939016393442625,
+      "grad_norm": 0.8257684106597777,
+      "learning_rate": 6.538629122411322e-07,
+      "loss": 11.7609,
+      "step": 16548
+    },
+    {
+      "epoch": 2.8940765027322404,
+      "grad_norm": 0.8160233778613755,
+      "learning_rate": 6.517090744022403e-07,
+      "loss": 12.0373,
+      "step": 16549
+    },
+    {
+      "epoch": 2.8942513661202187,
+      "grad_norm": 0.8397727794286545,
+      "learning_rate": 6.495587782138435e-07,
+      "loss": 11.8453,
+      "step": 16550
+    },
+    {
+      "epoch": 2.8944262295081966,
+      "grad_norm": 0.8442779375505073,
+      "learning_rate": 6.474120237526027e-07,
+      "loss": 11.7383,
+      "step": 16551
+    },
+    {
+      "epoch": 2.894601092896175,
+      "grad_norm": 0.9153672521533492,
+      "learning_rate": 6.452688110950456e-07,
+      "loss": 11.9814,
+      "step": 16552
+    },
+    {
+      "epoch": 2.894775956284153,
+      "grad_norm": 0.8089337319729598,
+      "learning_rate": 6.431291403175666e-07,
+      "loss": 11.9529,
+      "step": 16553
+    },
+    {
+      "epoch": 2.894950819672131,
+      "grad_norm": 0.859013689249741,
+      "learning_rate": 6.40993011496438e-07,
+      "loss": 11.9372,
+      "step": 16554
+    },
+    {
+      "epoch": 2.895125683060109,
+      "grad_norm": 0.8809572723766197,
+      "learning_rate": 6.388604247078322e-07,
+      "loss": 11.5592,
+      "step": 16555
+    },
+    {
+      "epoch": 2.8953005464480874,
+      "grad_norm": 0.8051126821899086,
+      "learning_rate": 6.367313800277664e-07,
+      "loss": 11.6604,
+      "step": 16556
+    },
+    {
+      "epoch": 2.8954754098360658,
+      "grad_norm": 0.8294101971616902,
+      "learning_rate": 6.34605877532124e-07,
+      "loss": 11.9226,
+      "step": 16557
+    },
+    {
+      "epoch": 2.8956502732240437,
+      "grad_norm": 0.859916517028797,
+      "learning_rate": 6.324839172966779e-07,
+      "loss": 11.9692,
+      "step": 16558
+    },
+    {
+      "epoch": 2.895825136612022,
+      "grad_norm": 0.821802820034991,
+      "learning_rate": 6.303654993971008e-07,
+      "loss": 11.7032,
+      "step": 16559
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 0.827245649522575,
+      "learning_rate": 6.282506239088771e-07,
+      "loss": 11.763,
+      "step": 16560
+    },
+    {
+      "epoch": 2.8961748633879782,
+      "grad_norm": 0.8263134683060814,
+      "learning_rate": 6.261392909074238e-07,
+      "loss": 11.6304,
+      "step": 16561
+    },
+    {
+      "epoch": 2.896349726775956,
+      "grad_norm": 0.8350256261445265,
+      "learning_rate": 6.24031500467992e-07,
+      "loss": 12.0048,
+      "step": 16562
+    },
+    {
+      "epoch": 2.8965245901639345,
+      "grad_norm": 0.7942081805149999,
+      "learning_rate": 6.219272526657439e-07,
+      "loss": 11.7786,
+      "step": 16563
+    },
+    {
+      "epoch": 2.8966994535519124,
+      "grad_norm": 0.8154854063824329,
+      "learning_rate": 6.198265475756637e-07,
+      "loss": 11.7723,
+      "step": 16564
+    },
+    {
+      "epoch": 2.8968743169398907,
+      "grad_norm": 0.828255689655274,
+      "learning_rate": 6.177293852726584e-07,
+      "loss": 11.9041,
+      "step": 16565
+    },
+    {
+      "epoch": 2.897049180327869,
+      "grad_norm": 0.9215535367134075,
+      "learning_rate": 6.156357658314793e-07,
+      "loss": 11.85,
+      "step": 16566
+    },
+    {
+      "epoch": 2.897224043715847,
+      "grad_norm": 0.8068633235388643,
+      "learning_rate": 6.135456893267777e-07,
+      "loss": 11.9346,
+      "step": 16567
+    },
+    {
+      "epoch": 2.897398907103825,
+      "grad_norm": 0.8244119851595609,
+      "learning_rate": 6.114591558330384e-07,
+      "loss": 11.6208,
+      "step": 16568
+    },
+    {
+      "epoch": 2.897573770491803,
+      "grad_norm": 0.8943965142847081,
+      "learning_rate": 6.093761654246688e-07,
+      "loss": 11.7778,
+      "step": 16569
+    },
+    {
+      "epoch": 2.8977486338797815,
+      "grad_norm": 0.8675612839877085,
+      "learning_rate": 6.072967181759093e-07,
+      "loss": 11.8519,
+      "step": 16570
+    },
+    {
+      "epoch": 2.8979234972677594,
+      "grad_norm": 0.9332093361590231,
+      "learning_rate": 6.052208141608895e-07,
+      "loss": 11.8152,
+      "step": 16571
+    },
+    {
+      "epoch": 2.898098360655738,
+      "grad_norm": 0.8539593920027423,
+      "learning_rate": 6.031484534536281e-07,
+      "loss": 12.0164,
+      "step": 16572
+    },
+    {
+      "epoch": 2.8982732240437157,
+      "grad_norm": 0.7797468897535587,
+      "learning_rate": 6.010796361279769e-07,
+      "loss": 11.8306,
+      "step": 16573
+    },
+    {
+      "epoch": 2.898448087431694,
+      "grad_norm": 0.8372526553885636,
+      "learning_rate": 5.990143622577215e-07,
+      "loss": 11.8921,
+      "step": 16574
+    },
+    {
+      "epoch": 2.8986229508196724,
+      "grad_norm": 0.801259471317866,
+      "learning_rate": 5.969526319164476e-07,
+      "loss": 11.859,
+      "step": 16575
+    },
+    {
+      "epoch": 2.8987978142076503,
+      "grad_norm": 0.8081152309102073,
+      "learning_rate": 5.948944451776739e-07,
+      "loss": 11.6919,
+      "step": 16576
+    },
+    {
+      "epoch": 2.898972677595628,
+      "grad_norm": 0.8170931616903359,
+      "learning_rate": 5.928398021147863e-07,
+      "loss": 11.92,
+      "step": 16577
+    },
+    {
+      "epoch": 2.8991475409836065,
+      "grad_norm": 0.8102057089804644,
+      "learning_rate": 5.907887028010039e-07,
+      "loss": 12.0063,
+      "step": 16578
+    },
+    {
+      "epoch": 2.899322404371585,
+      "grad_norm": 0.8367175860785131,
+      "learning_rate": 5.887411473094573e-07,
+      "loss": 11.7263,
+      "step": 16579
+    },
+    {
+      "epoch": 2.8994972677595627,
+      "grad_norm": 0.8227713056854247,
+      "learning_rate": 5.866971357131434e-07,
+      "loss": 11.8949,
+      "step": 16580
+    },
+    {
+      "epoch": 2.899672131147541,
+      "grad_norm": 0.7705227001253832,
+      "learning_rate": 5.846566680849153e-07,
+      "loss": 11.8373,
+      "step": 16581
+    },
+    {
+      "epoch": 2.899846994535519,
+      "grad_norm": 0.8836961567337445,
+      "learning_rate": 5.826197444975368e-07,
+      "loss": 11.9283,
+      "step": 16582
+    },
+    {
+      "epoch": 2.9000218579234973,
+      "grad_norm": 0.8501624032365152,
+      "learning_rate": 5.805863650235943e-07,
+      "loss": 11.8194,
+      "step": 16583
+    },
+    {
+      "epoch": 2.9001967213114757,
+      "grad_norm": 0.8258828036353832,
+      "learning_rate": 5.785565297355966e-07,
+      "loss": 11.6544,
+      "step": 16584
+    },
+    {
+      "epoch": 2.9003715846994536,
+      "grad_norm": 0.8195452574110362,
+      "learning_rate": 5.765302387058857e-07,
+      "loss": 11.8871,
+      "step": 16585
+    },
+    {
+      "epoch": 2.9005464480874315,
+      "grad_norm": 0.8338425511182096,
+      "learning_rate": 5.745074920067261e-07,
+      "loss": 11.9347,
+      "step": 16586
+    },
+    {
+      "epoch": 2.90072131147541,
+      "grad_norm": 0.8511130384979397,
+      "learning_rate": 5.724882897101935e-07,
+      "loss": 11.8181,
+      "step": 16587
+    },
+    {
+      "epoch": 2.900896174863388,
+      "grad_norm": 0.8323821030231929,
+      "learning_rate": 5.704726318882858e-07,
+      "loss": 11.7433,
+      "step": 16588
+    },
+    {
+      "epoch": 2.901071038251366,
+      "grad_norm": 0.8121857737615749,
+      "learning_rate": 5.684605186128678e-07,
+      "loss": 11.955,
+      "step": 16589
+    },
+    {
+      "epoch": 2.9012459016393444,
+      "grad_norm": 0.8058540315684425,
+      "learning_rate": 5.664519499556487e-07,
+      "loss": 11.938,
+      "step": 16590
+    },
+    {
+      "epoch": 2.9014207650273223,
+      "grad_norm": 0.8085838361095868,
+      "learning_rate": 5.64446925988249e-07,
+      "loss": 11.7359,
+      "step": 16591
+    },
+    {
+      "epoch": 2.9015956284153006,
+      "grad_norm": 0.9846035487338815,
+      "learning_rate": 5.624454467821338e-07,
+      "loss": 11.673,
+      "step": 16592
+    },
+    {
+      "epoch": 2.901770491803279,
+      "grad_norm": 0.8522345845733036,
+      "learning_rate": 5.604475124086684e-07,
+      "loss": 11.656,
+      "step": 16593
+    },
+    {
+      "epoch": 2.901945355191257,
+      "grad_norm": 0.873223134874854,
+      "learning_rate": 5.58453122939051e-07,
+      "loss": 11.8277,
+      "step": 16594
+    },
+    {
+      "epoch": 2.9021202185792347,
+      "grad_norm": 0.805815041552021,
+      "learning_rate": 5.564622784444029e-07,
+      "loss": 11.9868,
+      "step": 16595
+    },
+    {
+      "epoch": 2.902295081967213,
+      "grad_norm": 0.8365431552503099,
+      "learning_rate": 5.544749789956894e-07,
+      "loss": 11.7851,
+      "step": 16596
+    },
+    {
+      "epoch": 2.9024699453551914,
+      "grad_norm": 0.8039882263218426,
+      "learning_rate": 5.524912246637537e-07,
+      "loss": 11.7748,
+      "step": 16597
+    },
+    {
+      "epoch": 2.9026448087431693,
+      "grad_norm": 0.8105620635725035,
+      "learning_rate": 5.505110155193061e-07,
+      "loss": 11.8038,
+      "step": 16598
+    },
+    {
+      "epoch": 2.9028196721311477,
+      "grad_norm": 0.8249405066449668,
+      "learning_rate": 5.485343516329566e-07,
+      "loss": 11.9732,
+      "step": 16599
+    },
+    {
+      "epoch": 2.9029945355191256,
+      "grad_norm": 0.9013227520654123,
+      "learning_rate": 5.465612330751602e-07,
+      "loss": 11.747,
+      "step": 16600
+    },
+    {
+      "epoch": 2.903169398907104,
+      "grad_norm": 0.8531792721825353,
+      "learning_rate": 5.445916599162493e-07,
+      "loss": 11.6895,
+      "step": 16601
+    },
+    {
+      "epoch": 2.903344262295082,
+      "grad_norm": 0.9065509121220929,
+      "learning_rate": 5.426256322264456e-07,
+      "loss": 11.3908,
+      "step": 16602
+    },
+    {
+      "epoch": 2.90351912568306,
+      "grad_norm": 0.8364762042755263,
+      "learning_rate": 5.406631500758374e-07,
+      "loss": 11.7388,
+      "step": 16603
+    },
+    {
+      "epoch": 2.903693989071038,
+      "grad_norm": 0.8354685944866417,
+      "learning_rate": 5.3870421353438e-07,
+      "loss": 11.6511,
+      "step": 16604
+    },
+    {
+      "epoch": 2.9038688524590164,
+      "grad_norm": 0.854637579784364,
+      "learning_rate": 5.367488226719064e-07,
+      "loss": 11.5526,
+      "step": 16605
+    },
+    {
+      "epoch": 2.9040437158469947,
+      "grad_norm": 0.8017135879825894,
+      "learning_rate": 5.347969775581385e-07,
+      "loss": 11.8483,
+      "step": 16606
+    },
+    {
+      "epoch": 2.9042185792349726,
+      "grad_norm": 0.835969760381752,
+      "learning_rate": 5.328486782626319e-07,
+      "loss": 11.9437,
+      "step": 16607
+    },
+    {
+      "epoch": 2.904393442622951,
+      "grad_norm": 0.8757673053537832,
+      "learning_rate": 5.309039248548642e-07,
+      "loss": 11.7708,
+      "step": 16608
+    },
+    {
+      "epoch": 2.904568306010929,
+      "grad_norm": 0.8453957419973035,
+      "learning_rate": 5.289627174041579e-07,
+      "loss": 11.8954,
+      "step": 16609
+    },
+    {
+      "epoch": 2.904743169398907,
+      "grad_norm": 0.8411527645581164,
+      "learning_rate": 5.270250559797018e-07,
+      "loss": 11.4409,
+      "step": 16610
+    },
+    {
+      "epoch": 2.904918032786885,
+      "grad_norm": 0.8261810513630916,
+      "learning_rate": 5.250909406505855e-07,
+      "loss": 11.9237,
+      "step": 16611
+    },
+    {
+      "epoch": 2.9050928961748634,
+      "grad_norm": 0.7846283635279668,
+      "learning_rate": 5.231603714857535e-07,
+      "loss": 12.0317,
+      "step": 16612
+    },
+    {
+      "epoch": 2.9052677595628413,
+      "grad_norm": 0.8179823565801445,
+      "learning_rate": 5.212333485540288e-07,
+      "loss": 11.8652,
+      "step": 16613
+    },
+    {
+      "epoch": 2.9054426229508197,
+      "grad_norm": 0.8279955917031854,
+      "learning_rate": 5.193098719241007e-07,
+      "loss": 11.7705,
+      "step": 16614
+    },
+    {
+      "epoch": 2.905617486338798,
+      "grad_norm": 0.8039024241865763,
+      "learning_rate": 5.173899416645478e-07,
+      "loss": 11.7405,
+      "step": 16615
+    },
+    {
+      "epoch": 2.905792349726776,
+      "grad_norm": 0.8072359941896033,
+      "learning_rate": 5.154735578438153e-07,
+      "loss": 11.8482,
+      "step": 16616
+    },
+    {
+      "epoch": 2.9059672131147543,
+      "grad_norm": 0.8103974796287863,
+      "learning_rate": 5.135607205302039e-07,
+      "loss": 11.8592,
+      "step": 16617
+    },
+    {
+      "epoch": 2.906142076502732,
+      "grad_norm": 0.9026320017832697,
+      "learning_rate": 5.116514297919261e-07,
+      "loss": 11.9738,
+      "step": 16618
+    },
+    {
+      "epoch": 2.9063169398907105,
+      "grad_norm": 0.8539359878469924,
+      "learning_rate": 5.097456856970384e-07,
+      "loss": 11.5485,
+      "step": 16619
+    },
+    {
+      "epoch": 2.9064918032786884,
+      "grad_norm": 0.834993771806954,
+      "learning_rate": 5.078434883134642e-07,
+      "loss": 11.8946,
+      "step": 16620
+    },
+    {
+      "epoch": 2.9066666666666667,
+      "grad_norm": 0.8387035360846745,
+      "learning_rate": 5.05944837709027e-07,
+      "loss": 11.8294,
+      "step": 16621
+    },
+    {
+      "epoch": 2.9068415300546446,
+      "grad_norm": 0.9171854142908227,
+      "learning_rate": 5.040497339514172e-07,
+      "loss": 11.745,
+      "step": 16622
+    },
+    {
+      "epoch": 2.907016393442623,
+      "grad_norm": 0.8664757345476007,
+      "learning_rate": 5.021581771081807e-07,
+      "loss": 11.8218,
+      "step": 16623
+    },
+    {
+      "epoch": 2.9071912568306013,
+      "grad_norm": 0.8247363010640454,
+      "learning_rate": 5.002701672467524e-07,
+      "loss": 11.7733,
+      "step": 16624
+    },
+    {
+      "epoch": 2.907366120218579,
+      "grad_norm": 0.8152883825453051,
+      "learning_rate": 4.983857044344453e-07,
+      "loss": 11.7266,
+      "step": 16625
+    },
+    {
+      "epoch": 2.907540983606557,
+      "grad_norm": 0.8420071947997381,
+      "learning_rate": 4.965047887384278e-07,
+      "loss": 11.5982,
+      "step": 16626
+    },
+    {
+      "epoch": 2.9077158469945354,
+      "grad_norm": 0.7854610115793793,
+      "learning_rate": 4.946274202257573e-07,
+      "loss": 11.5632,
+      "step": 16627
+    },
+    {
+      "epoch": 2.907890710382514,
+      "grad_norm": 0.8510121168716208,
+      "learning_rate": 4.927535989633691e-07,
+      "loss": 11.9835,
+      "step": 16628
+    },
+    {
+      "epoch": 2.9080655737704917,
+      "grad_norm": 0.8425128566153517,
+      "learning_rate": 4.908833250180434e-07,
+      "loss": 11.8482,
+      "step": 16629
+    },
+    {
+      "epoch": 2.90824043715847,
+      "grad_norm": 0.840543566090685,
+      "learning_rate": 4.89016598456471e-07,
+      "loss": 11.9289,
+      "step": 16630
+    },
+    {
+      "epoch": 2.908415300546448,
+      "grad_norm": 0.8371010430114926,
+      "learning_rate": 4.871534193451766e-07,
+      "loss": 12.0284,
+      "step": 16631
+    },
+    {
+      "epoch": 2.9085901639344263,
+      "grad_norm": 0.8234081378605547,
+      "learning_rate": 4.85293787750607e-07,
+      "loss": 11.9581,
+      "step": 16632
+    },
+    {
+      "epoch": 2.9087650273224046,
+      "grad_norm": 0.8069967786612552,
+      "learning_rate": 4.834377037390425e-07,
+      "loss": 11.6988,
+      "step": 16633
+    },
+    {
+      "epoch": 2.9089398907103825,
+      "grad_norm": 0.8531719337436052,
+      "learning_rate": 4.815851673766525e-07,
+      "loss": 11.8107,
+      "step": 16634
+    },
+    {
+      "epoch": 2.9091147540983604,
+      "grad_norm": 0.8308593013363801,
+      "learning_rate": 4.797361787294729e-07,
+      "loss": 12.0214,
+      "step": 16635
+    },
+    {
+      "epoch": 2.9092896174863387,
+      "grad_norm": 0.82405648797555,
+      "learning_rate": 4.778907378634178e-07,
+      "loss": 11.7937,
+      "step": 16636
+    },
+    {
+      "epoch": 2.909464480874317,
+      "grad_norm": 1.0235962777189893,
+      "learning_rate": 4.7604884484427905e-07,
+      "loss": 11.6999,
+      "step": 16637
+    },
+    {
+      "epoch": 2.909639344262295,
+      "grad_norm": 0.824946340178238,
+      "learning_rate": 4.742104997377261e-07,
+      "loss": 11.8976,
+      "step": 16638
+    },
+    {
+      "epoch": 2.9098142076502733,
+      "grad_norm": 0.9141168446001635,
+      "learning_rate": 4.723757026092734e-07,
+      "loss": 11.7707,
+      "step": 16639
+    },
+    {
+      "epoch": 2.909989071038251,
+      "grad_norm": 0.8317652393046661,
+      "learning_rate": 4.705444535243464e-07,
+      "loss": 11.9743,
+      "step": 16640
+    },
+    {
+      "epoch": 2.9101639344262296,
+      "grad_norm": 0.853809679299271,
+      "learning_rate": 4.687167525482372e-07,
+      "loss": 11.7061,
+      "step": 16641
+    },
+    {
+      "epoch": 2.910338797814208,
+      "grad_norm": 0.8118687561809298,
+      "learning_rate": 4.6689259974607157e-07,
+      "loss": 11.9156,
+      "step": 16642
+    },
+    {
+      "epoch": 2.910513661202186,
+      "grad_norm": 0.7947798014334521,
+      "learning_rate": 4.6507199518288633e-07,
+      "loss": 11.9057,
+      "step": 16643
+    },
+    {
+      "epoch": 2.9106885245901637,
+      "grad_norm": 0.8310156242172202,
+      "learning_rate": 4.632549389235963e-07,
+      "loss": 12.0147,
+      "step": 16644
+    },
+    {
+      "epoch": 2.910863387978142,
+      "grad_norm": 0.7969471701785061,
+      "learning_rate": 4.6144143103298286e-07,
+      "loss": 11.8959,
+      "step": 16645
+    },
+    {
+      "epoch": 2.9110382513661204,
+      "grad_norm": 0.7916785913455199,
+      "learning_rate": 4.596314715756722e-07,
+      "loss": 11.8282,
+      "step": 16646
+    },
+    {
+      "epoch": 2.9112131147540983,
+      "grad_norm": 0.8155813686835164,
+      "learning_rate": 4.578250606162016e-07,
+      "loss": 11.6224,
+      "step": 16647
+    },
+    {
+      "epoch": 2.9113879781420766,
+      "grad_norm": 0.8649343545381377,
+      "learning_rate": 4.5602219821896385e-07,
+      "loss": 11.7367,
+      "step": 16648
+    },
+    {
+      "epoch": 2.9115628415300545,
+      "grad_norm": 0.8073146087963116,
+      "learning_rate": 4.542228844482299e-07,
+      "loss": 11.8098,
+      "step": 16649
+    },
+    {
+      "epoch": 2.911737704918033,
+      "grad_norm": 0.8071497748755576,
+      "learning_rate": 4.5242711936814843e-07,
+      "loss": 11.9183,
+      "step": 16650
+    },
+    {
+      "epoch": 2.911912568306011,
+      "grad_norm": 0.9576515795635324,
+      "learning_rate": 4.506349030427237e-07,
+      "loss": 11.7119,
+      "step": 16651
+    },
+    {
+      "epoch": 2.912087431693989,
+      "grad_norm": 0.8233606566703615,
+      "learning_rate": 4.4884623553586025e-07,
+      "loss": 11.7977,
+      "step": 16652
+    },
+    {
+      "epoch": 2.912262295081967,
+      "grad_norm": 0.8400020167133451,
+      "learning_rate": 4.4706111691131815e-07,
+      "loss": 11.7099,
+      "step": 16653
+    },
+    {
+      "epoch": 2.9124371584699453,
+      "grad_norm": 0.8627881029431174,
+      "learning_rate": 4.4527954723272423e-07,
+      "loss": 11.6721,
+      "step": 16654
+    },
+    {
+      "epoch": 2.9126120218579237,
+      "grad_norm": 0.8698738878063353,
+      "learning_rate": 4.4350152656359444e-07,
+      "loss": 11.9987,
+      "step": 16655
+    },
+    {
+      "epoch": 2.9127868852459016,
+      "grad_norm": 0.827049798589559,
+      "learning_rate": 4.4172705496732246e-07,
+      "loss": 11.6347,
+      "step": 16656
+    },
+    {
+      "epoch": 2.91296174863388,
+      "grad_norm": 0.8190961902929121,
+      "learning_rate": 4.399561325071577e-07,
+      "loss": 11.6855,
+      "step": 16657
+    },
+    {
+      "epoch": 2.913136612021858,
+      "grad_norm": 0.8398341847510112,
+      "learning_rate": 4.381887592462386e-07,
+      "loss": 11.6423,
+      "step": 16658
+    },
+    {
+      "epoch": 2.913311475409836,
+      "grad_norm": 0.8238484311670083,
+      "learning_rate": 4.3642493524755914e-07,
+      "loss": 11.9433,
+      "step": 16659
+    },
+    {
+      "epoch": 2.913486338797814,
+      "grad_norm": 0.8021086178318086,
+      "learning_rate": 4.346646605740023e-07,
+      "loss": 11.6052,
+      "step": 16660
+    },
+    {
+      "epoch": 2.9136612021857924,
+      "grad_norm": 0.9780652799316204,
+      "learning_rate": 4.3290793528831805e-07,
+      "loss": 11.6572,
+      "step": 16661
+    },
+    {
+      "epoch": 2.9138360655737703,
+      "grad_norm": 0.8079997627603218,
+      "learning_rate": 4.311547594531451e-07,
+      "loss": 11.7551,
+      "step": 16662
+    },
+    {
+      "epoch": 2.9140109289617486,
+      "grad_norm": 0.8384588023137611,
+      "learning_rate": 4.2940513313096676e-07,
+      "loss": 11.8669,
+      "step": 16663
+    },
+    {
+      "epoch": 2.914185792349727,
+      "grad_norm": 0.8153237318741915,
+      "learning_rate": 4.2765905638415535e-07,
+      "loss": 11.7531,
+      "step": 16664
+    },
+    {
+      "epoch": 2.914360655737705,
+      "grad_norm": 0.915400621910665,
+      "learning_rate": 4.259165292749612e-07,
+      "loss": 11.6284,
+      "step": 16665
+    },
+    {
+      "epoch": 2.914535519125683,
+      "grad_norm": 0.8376179892641762,
+      "learning_rate": 4.2417755186551223e-07,
+      "loss": 11.7525,
+      "step": 16666
+    },
+    {
+      "epoch": 2.914710382513661,
+      "grad_norm": 0.8429862870635015,
+      "learning_rate": 4.2244212421778116e-07,
+      "loss": 11.7736,
+      "step": 16667
+    },
+    {
+      "epoch": 2.9148852459016394,
+      "grad_norm": 0.7969083308931606,
+      "learning_rate": 4.2071024639365184e-07,
+      "loss": 11.9178,
+      "step": 16668
+    },
+    {
+      "epoch": 2.9150601092896173,
+      "grad_norm": 0.8104662192773179,
+      "learning_rate": 4.189819184548527e-07,
+      "loss": 11.9561,
+      "step": 16669
+    },
+    {
+      "epoch": 2.9152349726775957,
+      "grad_norm": 0.7998571290268728,
+      "learning_rate": 4.172571404629899e-07,
+      "loss": 11.7291,
+      "step": 16670
+    },
+    {
+      "epoch": 2.9154098360655736,
+      "grad_norm": 0.8395666092702779,
+      "learning_rate": 4.155359124795699e-07,
+      "loss": 11.663,
+      "step": 16671
+    },
+    {
+      "epoch": 2.915584699453552,
+      "grad_norm": 0.8229389197693012,
+      "learning_rate": 4.138182345659547e-07,
+      "loss": 11.8957,
+      "step": 16672
+    },
+    {
+      "epoch": 2.9157595628415303,
+      "grad_norm": 0.8682370757114445,
+      "learning_rate": 4.1210410678335085e-07,
+      "loss": 11.876,
+      "step": 16673
+    },
+    {
+      "epoch": 2.915934426229508,
+      "grad_norm": 0.9382385074687432,
+      "learning_rate": 4.1039352919287623e-07,
+      "loss": 11.6641,
+      "step": 16674
+    },
+    {
+      "epoch": 2.9161092896174865,
+      "grad_norm": 0.8068472846833563,
+      "learning_rate": 4.0868650185552636e-07,
+      "loss": 12.0793,
+      "step": 16675
+    },
+    {
+      "epoch": 2.9162841530054644,
+      "grad_norm": 0.7703428024912603,
+      "learning_rate": 4.069830248321305e-07,
+      "loss": 11.8159,
+      "step": 16676
+    },
+    {
+      "epoch": 2.9164590163934427,
+      "grad_norm": 0.8756344956386618,
+      "learning_rate": 4.052830981834399e-07,
+      "loss": 11.8005,
+      "step": 16677
+    },
+    {
+      "epoch": 2.9166338797814206,
+      "grad_norm": 0.8923443136326895,
+      "learning_rate": 4.035867219700284e-07,
+      "loss": 11.7663,
+      "step": 16678
+    },
+    {
+      "epoch": 2.916808743169399,
+      "grad_norm": 0.872680646183626,
+      "learning_rate": 4.018938962524033e-07,
+      "loss": 11.7934,
+      "step": 16679
+    },
+    {
+      "epoch": 2.916983606557377,
+      "grad_norm": 0.8039057011174767,
+      "learning_rate": 4.0020462109088275e-07,
+      "loss": 11.7119,
+      "step": 16680
+    },
+    {
+      "epoch": 2.917158469945355,
+      "grad_norm": 0.8331353470510732,
+      "learning_rate": 3.985188965456965e-07,
+      "loss": 11.8525,
+      "step": 16681
+    },
+    {
+      "epoch": 2.9173333333333336,
+      "grad_norm": 0.8398345805250682,
+      "learning_rate": 3.968367226769409e-07,
+      "loss": 11.6335,
+      "step": 16682
+    },
+    {
+      "epoch": 2.9175081967213115,
+      "grad_norm": 0.8255272740980659,
+      "learning_rate": 3.9515809954457917e-07,
+      "loss": 11.5542,
+      "step": 16683
+    },
+    {
+      "epoch": 2.9176830601092894,
+      "grad_norm": 0.8966783354983613,
+      "learning_rate": 3.934830272084633e-07,
+      "loss": 11.8861,
+      "step": 16684
+    },
+    {
+      "epoch": 2.9178579234972677,
+      "grad_norm": 0.8489160413027804,
+      "learning_rate": 3.9181150572829006e-07,
+      "loss": 11.5511,
+      "step": 16685
+    },
+    {
+      "epoch": 2.918032786885246,
+      "grad_norm": 0.8034328823937174,
+      "learning_rate": 3.901435351636673e-07,
+      "loss": 11.9216,
+      "step": 16686
+    },
+    {
+      "epoch": 2.918207650273224,
+      "grad_norm": 0.9335610325508664,
+      "learning_rate": 3.8847911557403636e-07,
+      "loss": 11.7667,
+      "step": 16687
+    },
+    {
+      "epoch": 2.9183825136612023,
+      "grad_norm": 0.8507665575779584,
+      "learning_rate": 3.8681824701874983e-07,
+      "loss": 11.7487,
+      "step": 16688
+    },
+    {
+      "epoch": 2.91855737704918,
+      "grad_norm": 0.7875511759127966,
+      "learning_rate": 3.8516092955700466e-07,
+      "loss": 11.6571,
+      "step": 16689
+    },
+    {
+      "epoch": 2.9187322404371585,
+      "grad_norm": 0.848459867655781,
+      "learning_rate": 3.8350716324788705e-07,
+      "loss": 11.7012,
+      "step": 16690
+    },
+    {
+      "epoch": 2.918907103825137,
+      "grad_norm": 0.8022937549802417,
+      "learning_rate": 3.8185694815033866e-07,
+      "loss": 11.8154,
+      "step": 16691
+    },
+    {
+      "epoch": 2.9190819672131147,
+      "grad_norm": 0.7983392479530553,
+      "learning_rate": 3.802102843232125e-07,
+      "loss": 11.7755,
+      "step": 16692
+    },
+    {
+      "epoch": 2.9192568306010926,
+      "grad_norm": 0.8122371254305892,
+      "learning_rate": 3.785671718251838e-07,
+      "loss": 11.7077,
+      "step": 16693
+    },
+    {
+      "epoch": 2.919431693989071,
+      "grad_norm": 0.8408192632131036,
+      "learning_rate": 3.7692761071485005e-07,
+      "loss": 11.6161,
+      "step": 16694
+    },
+    {
+      "epoch": 2.9196065573770493,
+      "grad_norm": 0.7942074194728717,
+      "learning_rate": 3.752916010506424e-07,
+      "loss": 11.9847,
+      "step": 16695
+    },
+    {
+      "epoch": 2.9197814207650272,
+      "grad_norm": 0.8239291957165706,
+      "learning_rate": 3.73659142890892e-07,
+      "loss": 11.8203,
+      "step": 16696
+    },
+    {
+      "epoch": 2.9199562841530056,
+      "grad_norm": 0.8103417454646333,
+      "learning_rate": 3.7203023629379665e-07,
+      "loss": 11.4124,
+      "step": 16697
+    },
+    {
+      "epoch": 2.9201311475409835,
+      "grad_norm": 0.7747383487019288,
+      "learning_rate": 3.70404881317421e-07,
+      "loss": 11.9515,
+      "step": 16698
+    },
+    {
+      "epoch": 2.920306010928962,
+      "grad_norm": 0.8961809054535264,
+      "learning_rate": 3.687830780197077e-07,
+      "loss": 11.7272,
+      "step": 16699
+    },
+    {
+      "epoch": 2.92048087431694,
+      "grad_norm": 0.8802147215173276,
+      "learning_rate": 3.6716482645846595e-07,
+      "loss": 11.7775,
+      "step": 16700
+    },
+    {
+      "epoch": 2.920655737704918,
+      "grad_norm": 0.8072797357928078,
+      "learning_rate": 3.655501266913941e-07,
+      "loss": 11.8162,
+      "step": 16701
+    },
+    {
+      "epoch": 2.920830601092896,
+      "grad_norm": 0.8011947853716509,
+      "learning_rate": 3.639389787760461e-07,
+      "loss": 11.9636,
+      "step": 16702
+    },
+    {
+      "epoch": 2.9210054644808743,
+      "grad_norm": 0.8510817984770745,
+      "learning_rate": 3.623313827698649e-07,
+      "loss": 11.6501,
+      "step": 16703
+    },
+    {
+      "epoch": 2.9211803278688526,
+      "grad_norm": 0.7890836242039794,
+      "learning_rate": 3.607273387301491e-07,
+      "loss": 11.6035,
+      "step": 16704
+    },
+    {
+      "epoch": 2.9213551912568305,
+      "grad_norm": 0.7696790858658498,
+      "learning_rate": 3.5912684671409733e-07,
+      "loss": 12.1792,
+      "step": 16705
+    },
+    {
+      "epoch": 2.921530054644809,
+      "grad_norm": 0.8109032771025402,
+      "learning_rate": 3.575299067787419e-07,
+      "loss": 11.9332,
+      "step": 16706
+    },
+    {
+      "epoch": 2.9217049180327868,
+      "grad_norm": 0.8385557712097125,
+      "learning_rate": 3.5593651898103707e-07,
+      "loss": 11.6629,
+      "step": 16707
+    },
+    {
+      "epoch": 2.921879781420765,
+      "grad_norm": 0.841422557725114,
+      "learning_rate": 3.543466833777709e-07,
+      "loss": 11.8202,
+      "step": 16708
+    },
+    {
+      "epoch": 2.9220546448087434,
+      "grad_norm": 0.8248399118826236,
+      "learning_rate": 3.5276040002562016e-07,
+      "loss": 11.6969,
+      "step": 16709
+    },
+    {
+      "epoch": 2.9222295081967213,
+      "grad_norm": 0.8133197179671954,
+      "learning_rate": 3.5117766898113966e-07,
+      "loss": 11.8902,
+      "step": 16710
+    },
+    {
+      "epoch": 2.9224043715846992,
+      "grad_norm": 0.7836214750596131,
+      "learning_rate": 3.495984903007399e-07,
+      "loss": 11.7377,
+      "step": 16711
+    },
+    {
+      "epoch": 2.9225792349726776,
+      "grad_norm": 0.8250803482382199,
+      "learning_rate": 3.4802286404073125e-07,
+      "loss": 12.0373,
+      "step": 16712
+    },
+    {
+      "epoch": 2.922754098360656,
+      "grad_norm": 0.8405612914491692,
+      "learning_rate": 3.4645079025728e-07,
+      "loss": 11.8077,
+      "step": 16713
+    },
+    {
+      "epoch": 2.922928961748634,
+      "grad_norm": 0.8160397173598231,
+      "learning_rate": 3.4488226900641905e-07,
+      "loss": 11.5947,
+      "step": 16714
+    },
+    {
+      "epoch": 2.923103825136612,
+      "grad_norm": 0.8383067167485696,
+      "learning_rate": 3.433173003440704e-07,
+      "loss": 11.8118,
+      "step": 16715
+    },
+    {
+      "epoch": 2.92327868852459,
+      "grad_norm": 0.8559327197979952,
+      "learning_rate": 3.4175588432602267e-07,
+      "loss": 11.924,
+      "step": 16716
+    },
+    {
+      "epoch": 2.9234535519125684,
+      "grad_norm": 0.8623946655955073,
+      "learning_rate": 3.401980210079536e-07,
+      "loss": 11.8555,
+      "step": 16717
+    },
+    {
+      "epoch": 2.9236284153005463,
+      "grad_norm": 0.8875798194716781,
+      "learning_rate": 3.3864371044537435e-07,
+      "loss": 11.899,
+      "step": 16718
+    },
+    {
+      "epoch": 2.9238032786885246,
+      "grad_norm": 0.8446275751872346,
+      "learning_rate": 3.370929526937072e-07,
+      "loss": 11.831,
+      "step": 16719
+    },
+    {
+      "epoch": 2.9239781420765025,
+      "grad_norm": 0.8272710396894407,
+      "learning_rate": 3.3554574780823025e-07,
+      "loss": 11.8897,
+      "step": 16720
+    },
+    {
+      "epoch": 2.924153005464481,
+      "grad_norm": 0.823638618293121,
+      "learning_rate": 3.340020958440992e-07,
+      "loss": 11.7938,
+      "step": 16721
+    },
+    {
+      "epoch": 2.924327868852459,
+      "grad_norm": 0.9553792821293833,
+      "learning_rate": 3.32461996856348e-07,
+      "loss": 11.9095,
+      "step": 16722
+    },
+    {
+      "epoch": 2.924502732240437,
+      "grad_norm": 0.8212445712686604,
+      "learning_rate": 3.3092545089987713e-07,
+      "loss": 11.6666,
+      "step": 16723
+    },
+    {
+      "epoch": 2.9246775956284154,
+      "grad_norm": 0.7998083961266241,
+      "learning_rate": 3.2939245802947604e-07,
+      "loss": 11.8428,
+      "step": 16724
+    },
+    {
+      "epoch": 2.9248524590163933,
+      "grad_norm": 0.8567181944224281,
+      "learning_rate": 3.2786301829976773e-07,
+      "loss": 11.579,
+      "step": 16725
+    },
+    {
+      "epoch": 2.9250273224043717,
+      "grad_norm": 0.8276982746421884,
+      "learning_rate": 3.2633713176528634e-07,
+      "loss": 11.9044,
+      "step": 16726
+    },
+    {
+      "epoch": 2.9252021857923496,
+      "grad_norm": 0.8644554040162923,
+      "learning_rate": 3.2481479848043285e-07,
+      "loss": 11.7656,
+      "step": 16727
+    },
+    {
+      "epoch": 2.925377049180328,
+      "grad_norm": 0.8499073854098698,
+      "learning_rate": 3.232960184994749e-07,
+      "loss": 11.4585,
+      "step": 16728
+    },
+    {
+      "epoch": 2.925551912568306,
+      "grad_norm": 0.764430172927889,
+      "learning_rate": 3.2178079187655805e-07,
+      "loss": 11.7096,
+      "step": 16729
+    },
+    {
+      "epoch": 2.925726775956284,
+      "grad_norm": 0.7519984651789364,
+      "learning_rate": 3.2026911866569476e-07,
+      "loss": 11.6747,
+      "step": 16730
+    },
+    {
+      "epoch": 2.9259016393442625,
+      "grad_norm": 0.8192599931961451,
+      "learning_rate": 3.1876099892076404e-07,
+      "loss": 11.7387,
+      "step": 16731
+    },
+    {
+      "epoch": 2.9260765027322404,
+      "grad_norm": 0.8191233472728607,
+      "learning_rate": 3.172564326955452e-07,
+      "loss": 11.9721,
+      "step": 16732
+    },
+    {
+      "epoch": 2.9262513661202187,
+      "grad_norm": 0.8938666971533813,
+      "learning_rate": 3.15755420043673e-07,
+      "loss": 11.9987,
+      "step": 16733
+    },
+    {
+      "epoch": 2.9264262295081966,
+      "grad_norm": 0.8309663714170736,
+      "learning_rate": 3.1425796101864915e-07,
+      "loss": 11.7842,
+      "step": 16734
+    },
+    {
+      "epoch": 2.926601092896175,
+      "grad_norm": 0.8831433594708415,
+      "learning_rate": 3.1276405567385314e-07,
+      "loss": 11.8291,
+      "step": 16735
+    },
+    {
+      "epoch": 2.926775956284153,
+      "grad_norm": 0.8442406851147997,
+      "learning_rate": 3.1127370406255354e-07,
+      "loss": 11.7841,
+      "step": 16736
+    },
+    {
+      "epoch": 2.926950819672131,
+      "grad_norm": 0.7925759090055795,
+      "learning_rate": 3.0978690623786334e-07,
+      "loss": 11.775,
+      "step": 16737
+    },
+    {
+      "epoch": 2.927125683060109,
+      "grad_norm": 0.7568086539489256,
+      "learning_rate": 3.0830366225279574e-07,
+      "loss": 11.8207,
+      "step": 16738
+    },
+    {
+      "epoch": 2.9273005464480875,
+      "grad_norm": 0.8034706228098143,
+      "learning_rate": 3.0682397216024175e-07,
+      "loss": 11.5935,
+      "step": 16739
+    },
+    {
+      "epoch": 2.927475409836066,
+      "grad_norm": 0.8794898091885727,
+      "learning_rate": 3.053478360129147e-07,
+      "loss": 11.7498,
+      "step": 16740
+    },
+    {
+      "epoch": 2.9276502732240437,
+      "grad_norm": 0.809465803370231,
+      "learning_rate": 3.038752538634615e-07,
+      "loss": 12.0478,
+      "step": 16741
+    },
+    {
+      "epoch": 2.9278251366120216,
+      "grad_norm": 0.8082679812430387,
+      "learning_rate": 3.0240622576437337e-07,
+      "loss": 11.8121,
+      "step": 16742
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 0.9040589129762151,
+      "learning_rate": 3.0094075176803073e-07,
+      "loss": 11.8105,
+      "step": 16743
+    },
+    {
+      "epoch": 2.9281748633879783,
+      "grad_norm": 0.8204717972451746,
+      "learning_rate": 2.994788319266473e-07,
+      "loss": 11.9135,
+      "step": 16744
+    },
+    {
+      "epoch": 2.928349726775956,
+      "grad_norm": 0.8959494984747582,
+      "learning_rate": 2.980204662923702e-07,
+      "loss": 11.6158,
+      "step": 16745
+    },
+    {
+      "epoch": 2.9285245901639345,
+      "grad_norm": 0.9054960025705192,
+      "learning_rate": 2.9656565491716915e-07,
+      "loss": 11.5457,
+      "step": 16746
+    },
+    {
+      "epoch": 2.9286994535519124,
+      "grad_norm": 0.8370483636225067,
+      "learning_rate": 2.9511439785290253e-07,
+      "loss": 11.7407,
+      "step": 16747
+    },
+    {
+      "epoch": 2.9288743169398908,
+      "grad_norm": 0.8465292119064087,
+      "learning_rate": 2.936666951513178e-07,
+      "loss": 11.6547,
+      "step": 16748
+    },
+    {
+      "epoch": 2.929049180327869,
+      "grad_norm": 0.7979257646280837,
+      "learning_rate": 2.922225468640294e-07,
+      "loss": 11.7343,
+      "step": 16749
+    },
+    {
+      "epoch": 2.929224043715847,
+      "grad_norm": 0.8194670686194916,
+      "learning_rate": 2.907819530425071e-07,
+      "loss": 11.7652,
+      "step": 16750
+    },
+    {
+      "epoch": 2.929398907103825,
+      "grad_norm": 0.8682057271322378,
+      "learning_rate": 2.893449137381099e-07,
+      "loss": 11.7459,
+      "step": 16751
+    },
+    {
+      "epoch": 2.9295737704918032,
+      "grad_norm": 0.8544918859792119,
+      "learning_rate": 2.879114290020746e-07,
+      "loss": 11.8527,
+      "step": 16752
+    },
+    {
+      "epoch": 2.9297486338797816,
+      "grad_norm": 0.7892509936891964,
+      "learning_rate": 2.8648149888548246e-07,
+      "loss": 11.946,
+      "step": 16753
+    },
+    {
+      "epoch": 2.9299234972677595,
+      "grad_norm": 0.787848351527813,
+      "learning_rate": 2.850551234393373e-07,
+      "loss": 11.6981,
+      "step": 16754
+    },
+    {
+      "epoch": 2.930098360655738,
+      "grad_norm": 0.8021094529820916,
+      "learning_rate": 2.836323027144649e-07,
+      "loss": 11.8661,
+      "step": 16755
+    },
+    {
+      "epoch": 2.9302732240437157,
+      "grad_norm": 0.8029174918859913,
+      "learning_rate": 2.822130367615916e-07,
+      "loss": 11.6913,
+      "step": 16756
+    },
+    {
+      "epoch": 2.930448087431694,
+      "grad_norm": 0.7839185880897044,
+      "learning_rate": 2.8079732563131013e-07,
+      "loss": 11.9669,
+      "step": 16757
+    },
+    {
+      "epoch": 2.9306229508196724,
+      "grad_norm": 0.8223566971215485,
+      "learning_rate": 2.793851693741134e-07,
+      "loss": 11.8734,
+      "step": 16758
+    },
+    {
+      "epoch": 2.9307978142076503,
+      "grad_norm": 0.8053806516437774,
+      "learning_rate": 2.779765680403057e-07,
+      "loss": 11.7701,
+      "step": 16759
+    },
+    {
+      "epoch": 2.930972677595628,
+      "grad_norm": 0.9127981903168471,
+      "learning_rate": 2.7657152168013565e-07,
+      "loss": 12.0835,
+      "step": 16760
+    },
+    {
+      "epoch": 2.9311475409836065,
+      "grad_norm": 0.7913207671331354,
+      "learning_rate": 2.751700303436633e-07,
+      "loss": 11.7305,
+      "step": 16761
+    },
+    {
+      "epoch": 2.931322404371585,
+      "grad_norm": 0.8587981858929006,
+      "learning_rate": 2.737720940808708e-07,
+      "loss": 11.8902,
+      "step": 16762
+    },
+    {
+      "epoch": 2.9314972677595628,
+      "grad_norm": 0.8111784476632065,
+      "learning_rate": 2.72377712941585e-07,
+      "loss": 11.9175,
+      "step": 16763
+    },
+    {
+      "epoch": 2.931672131147541,
+      "grad_norm": 0.837952546718848,
+      "learning_rate": 2.7098688697551053e-07,
+      "loss": 11.8804,
+      "step": 16764
+    },
+    {
+      "epoch": 2.931846994535519,
+      "grad_norm": 0.8368518109214804,
+      "learning_rate": 2.6959961623224115e-07,
+      "loss": 11.7415,
+      "step": 16765
+    },
+    {
+      "epoch": 2.9320218579234973,
+      "grad_norm": 0.8150188458267335,
+      "learning_rate": 2.6821590076121507e-07,
+      "loss": 11.8998,
+      "step": 16766
+    },
+    {
+      "epoch": 2.9321967213114757,
+      "grad_norm": 0.8219336767016415,
+      "learning_rate": 2.6683574061177053e-07,
+      "loss": 11.9799,
+      "step": 16767
+    },
+    {
+      "epoch": 2.9323715846994536,
+      "grad_norm": 0.8056723823153136,
+      "learning_rate": 2.6545913583310157e-07,
+      "loss": 11.9199,
+      "step": 16768
+    },
+    {
+      "epoch": 2.9325464480874315,
+      "grad_norm": 0.8124671079918443,
+      "learning_rate": 2.640860864743022e-07,
+      "loss": 11.8056,
+      "step": 16769
+    },
+    {
+      "epoch": 2.93272131147541,
+      "grad_norm": 0.8101306132607292,
+      "learning_rate": 2.627165925842889e-07,
+      "loss": 11.7561,
+      "step": 16770
+    },
+    {
+      "epoch": 2.932896174863388,
+      "grad_norm": 0.8466849944749238,
+      "learning_rate": 2.6135065421191153e-07,
+      "loss": 11.7499,
+      "step": 16771
+    },
+    {
+      "epoch": 2.933071038251366,
+      "grad_norm": 0.7863734229940815,
+      "learning_rate": 2.5998827140584214e-07,
+      "loss": 11.9779,
+      "step": 16772
+    },
+    {
+      "epoch": 2.9332459016393444,
+      "grad_norm": 0.8000688481022854,
+      "learning_rate": 2.5862944421465305e-07,
+      "loss": 11.6588,
+      "step": 16773
+    },
+    {
+      "epoch": 2.9334207650273223,
+      "grad_norm": 0.881324472870993,
+      "learning_rate": 2.5727417268679445e-07,
+      "loss": 11.8916,
+      "step": 16774
+    },
+    {
+      "epoch": 2.9335956284153006,
+      "grad_norm": 0.8725408100911555,
+      "learning_rate": 2.5592245687058316e-07,
+      "loss": 11.8566,
+      "step": 16775
+    },
+    {
+      "epoch": 2.9337704918032785,
+      "grad_norm": 0.7867745075613823,
+      "learning_rate": 2.5457429681419173e-07,
+      "loss": 11.9517,
+      "step": 16776
+    },
+    {
+      "epoch": 2.933945355191257,
+      "grad_norm": 0.911109857943046,
+      "learning_rate": 2.5322969256568184e-07,
+      "loss": 11.5098,
+      "step": 16777
+    },
+    {
+      "epoch": 2.9341202185792348,
+      "grad_norm": 0.8510341068462639,
+      "learning_rate": 2.5188864417299284e-07,
+      "loss": 11.8144,
+      "step": 16778
+    },
+    {
+      "epoch": 2.934295081967213,
+      "grad_norm": 0.8396165896196338,
+      "learning_rate": 2.5055115168393095e-07,
+      "loss": 11.6313,
+      "step": 16779
+    },
+    {
+      "epoch": 2.9344699453551915,
+      "grad_norm": 0.7653883345358193,
+      "learning_rate": 2.492172151461802e-07,
+      "loss": 11.8621,
+      "step": 16780
+    },
+    {
+      "epoch": 2.9346448087431694,
+      "grad_norm": 0.8528415947379994,
+      "learning_rate": 2.478868346072916e-07,
+      "loss": 11.6598,
+      "step": 16781
+    },
+    {
+      "epoch": 2.9348196721311477,
+      "grad_norm": 0.7793212190696331,
+      "learning_rate": 2.4656001011468253e-07,
+      "loss": 11.7849,
+      "step": 16782
+    },
+    {
+      "epoch": 2.9349945355191256,
+      "grad_norm": 0.9414417337107991,
+      "learning_rate": 2.452367417156598e-07,
+      "loss": 11.6459,
+      "step": 16783
+    },
+    {
+      "epoch": 2.935169398907104,
+      "grad_norm": 0.7907975526868601,
+      "learning_rate": 2.439170294573967e-07,
+      "loss": 11.9,
+      "step": 16784
+    },
+    {
+      "epoch": 2.935344262295082,
+      "grad_norm": 0.841751422338589,
+      "learning_rate": 2.426008733869445e-07,
+      "loss": 11.9263,
+      "step": 16785
+    },
+    {
+      "epoch": 2.93551912568306,
+      "grad_norm": 0.8454812494669244,
+      "learning_rate": 2.4128827355121007e-07,
+      "loss": 11.8735,
+      "step": 16786
+    },
+    {
+      "epoch": 2.935693989071038,
+      "grad_norm": 0.7697792316725361,
+      "learning_rate": 2.399792299970005e-07,
+      "loss": 11.8249,
+      "step": 16787
+    },
+    {
+      "epoch": 2.9358688524590164,
+      "grad_norm": 0.8220561023242307,
+      "learning_rate": 2.386737427709673e-07,
+      "loss": 11.9494,
+      "step": 16788
+    },
+    {
+      "epoch": 2.9360437158469948,
+      "grad_norm": 0.8444458439618542,
+      "learning_rate": 2.3737181191966218e-07,
+      "loss": 11.8992,
+      "step": 16789
+    },
+    {
+      "epoch": 2.9362185792349726,
+      "grad_norm": 0.820432803822098,
+      "learning_rate": 2.360734374894813e-07,
+      "loss": 11.8314,
+      "step": 16790
+    },
+    {
+      "epoch": 2.9363934426229505,
+      "grad_norm": 0.9191373189381946,
+      "learning_rate": 2.3477861952673208e-07,
+      "loss": 11.9341,
+      "step": 16791
+    },
+    {
+      "epoch": 2.936568306010929,
+      "grad_norm": 0.7819934747390189,
+      "learning_rate": 2.334873580775665e-07,
+      "loss": 11.9985,
+      "step": 16792
+    },
+    {
+      "epoch": 2.9367431693989072,
+      "grad_norm": 0.8393964879207905,
+      "learning_rate": 2.321996531880033e-07,
+      "loss": 11.8762,
+      "step": 16793
+    },
+    {
+      "epoch": 2.936918032786885,
+      "grad_norm": 0.8057418161685755,
+      "learning_rate": 2.309155049039502e-07,
+      "loss": 11.6458,
+      "step": 16794
+    },
+    {
+      "epoch": 2.9370928961748635,
+      "grad_norm": 0.8415737965783977,
+      "learning_rate": 2.296349132712039e-07,
+      "loss": 11.8206,
+      "step": 16795
+    },
+    {
+      "epoch": 2.9372677595628414,
+      "grad_norm": 0.7764245967117093,
+      "learning_rate": 2.2835787833539457e-07,
+      "loss": 11.3959,
+      "step": 16796
+    },
+    {
+      "epoch": 2.9374426229508197,
+      "grad_norm": 0.909993044348292,
+      "learning_rate": 2.2708440014206356e-07,
+      "loss": 11.7837,
+      "step": 16797
+    },
+    {
+      "epoch": 2.937617486338798,
+      "grad_norm": 0.8741010859482411,
+      "learning_rate": 2.258144787365968e-07,
+      "loss": 11.5493,
+      "step": 16798
+    },
+    {
+      "epoch": 2.937792349726776,
+      "grad_norm": 0.7840172164926745,
+      "learning_rate": 2.2454811416426912e-07,
+      "loss": 11.8882,
+      "step": 16799
+    },
+    {
+      "epoch": 2.937967213114754,
+      "grad_norm": 0.8241504606307524,
+      "learning_rate": 2.232853064702334e-07,
+      "loss": 11.5534,
+      "step": 16800
+    },
+    {
+      "epoch": 2.938142076502732,
+      "grad_norm": 0.8851670362162616,
+      "learning_rate": 2.2202605569948687e-07,
+      "loss": 11.5389,
+      "step": 16801
+    },
+    {
+      "epoch": 2.9383169398907105,
+      "grad_norm": 0.783010944173616,
+      "learning_rate": 2.2077036189694923e-07,
+      "loss": 12.0188,
+      "step": 16802
+    },
+    {
+      "epoch": 2.9384918032786884,
+      "grad_norm": 0.7945513853710878,
+      "learning_rate": 2.195182251073513e-07,
+      "loss": 11.7208,
+      "step": 16803
+    },
+    {
+      "epoch": 2.9386666666666668,
+      "grad_norm": 0.8049835088510313,
+      "learning_rate": 2.1826964537535743e-07,
+      "loss": 11.6963,
+      "step": 16804
+    },
+    {
+      "epoch": 2.9388415300546447,
+      "grad_norm": 0.8008198700509723,
+      "learning_rate": 2.170246227454653e-07,
+      "loss": 11.8721,
+      "step": 16805
+    },
+    {
+      "epoch": 2.939016393442623,
+      "grad_norm": 0.8460397761931203,
+      "learning_rate": 2.1578315726207277e-07,
+      "loss": 11.5823,
+      "step": 16806
+    },
+    {
+      "epoch": 2.9391912568306013,
+      "grad_norm": 0.7619327010511785,
+      "learning_rate": 2.145452489694111e-07,
+      "loss": 11.5601,
+      "step": 16807
+    },
+    {
+      "epoch": 2.9393661202185792,
+      "grad_norm": 0.8675892118307602,
+      "learning_rate": 2.1331089791163384e-07,
+      "loss": 11.7104,
+      "step": 16808
+    },
+    {
+      "epoch": 2.939540983606557,
+      "grad_norm": 0.8639512253033231,
+      "learning_rate": 2.1208010413272806e-07,
+      "loss": 11.9925,
+      "step": 16809
+    },
+    {
+      "epoch": 2.9397158469945355,
+      "grad_norm": 0.8205784788205007,
+      "learning_rate": 2.1085286767658086e-07,
+      "loss": 11.5201,
+      "step": 16810
+    },
+    {
+      "epoch": 2.939890710382514,
+      "grad_norm": 0.8907205151494989,
+      "learning_rate": 2.0962918858694613e-07,
+      "loss": 11.7135,
+      "step": 16811
+    },
+    {
+      "epoch": 2.9400655737704917,
+      "grad_norm": 0.9043349445540547,
+      "learning_rate": 2.0840906690742234e-07,
+      "loss": 11.8929,
+      "step": 16812
+    },
+    {
+      "epoch": 2.94024043715847,
+      "grad_norm": 0.8032042604916539,
+      "learning_rate": 2.0719250268153024e-07,
+      "loss": 11.7627,
+      "step": 16813
+    },
+    {
+      "epoch": 2.940415300546448,
+      "grad_norm": 0.8118298188947959,
+      "learning_rate": 2.05979495952624e-07,
+      "loss": 11.8037,
+      "step": 16814
+    },
+    {
+      "epoch": 2.9405901639344263,
+      "grad_norm": 0.8223256787316507,
+      "learning_rate": 2.0477004676395795e-07,
+      "loss": 11.7725,
+      "step": 16815
+    },
+    {
+      "epoch": 2.9407650273224046,
+      "grad_norm": 0.8558812053795591,
+      "learning_rate": 2.0356415515864203e-07,
+      "loss": 11.7266,
+      "step": 16816
+    },
+    {
+      "epoch": 2.9409398907103825,
+      "grad_norm": 0.8300670711795889,
+      "learning_rate": 2.0236182117966406e-07,
+      "loss": 11.8135,
+      "step": 16817
+    },
+    {
+      "epoch": 2.9411147540983604,
+      "grad_norm": 0.8159150640496428,
+      "learning_rate": 2.0116304486987868e-07,
+      "loss": 11.6636,
+      "step": 16818
+    },
+    {
+      "epoch": 2.9412896174863388,
+      "grad_norm": 0.857269527521661,
+      "learning_rate": 1.9996782627202949e-07,
+      "loss": 11.737,
+      "step": 16819
+    },
+    {
+      "epoch": 2.941464480874317,
+      "grad_norm": 0.808413089865865,
+      "learning_rate": 1.9877616542872678e-07,
+      "loss": 11.9128,
+      "step": 16820
+    },
+    {
+      "epoch": 2.941639344262295,
+      "grad_norm": 0.8309030286024355,
+      "learning_rate": 1.9758806238245885e-07,
+      "loss": 11.8017,
+      "step": 16821
+    },
+    {
+      "epoch": 2.9418142076502733,
+      "grad_norm": 0.8527140510491478,
+      "learning_rate": 1.964035171755585e-07,
+      "loss": 11.8674,
+      "step": 16822
+    },
+    {
+      "epoch": 2.9419890710382512,
+      "grad_norm": 0.8344193755973661,
+      "learning_rate": 1.9522252985026968e-07,
+      "loss": 11.7398,
+      "step": 16823
+    },
+    {
+      "epoch": 2.9421639344262296,
+      "grad_norm": 0.8013201190705419,
+      "learning_rate": 1.9404510044868096e-07,
+      "loss": 11.8061,
+      "step": 16824
+    },
+    {
+      "epoch": 2.9423387978142075,
+      "grad_norm": 0.8333186522209541,
+      "learning_rate": 1.9287122901278098e-07,
+      "loss": 11.8292,
+      "step": 16825
+    },
+    {
+      "epoch": 2.942513661202186,
+      "grad_norm": 0.8310201194172274,
+      "learning_rate": 1.9170091558440296e-07,
+      "loss": 11.8406,
+      "step": 16826
+    },
+    {
+      "epoch": 2.9426885245901637,
+      "grad_norm": 0.7823655927225861,
+      "learning_rate": 1.9053416020528013e-07,
+      "loss": 11.61,
+      "step": 16827
+    },
+    {
+      "epoch": 2.942863387978142,
+      "grad_norm": 0.8482957999966495,
+      "learning_rate": 1.893709629169904e-07,
+      "loss": 11.4796,
+      "step": 16828
+    },
+    {
+      "epoch": 2.9430382513661204,
+      "grad_norm": 0.8435448936568655,
+      "learning_rate": 1.8821132376102278e-07,
+      "loss": 11.7507,
+      "step": 16829
+    },
+    {
+      "epoch": 2.9432131147540983,
+      "grad_norm": 0.868845723810858,
+      "learning_rate": 1.870552427786998e-07,
+      "loss": 11.8784,
+      "step": 16830
+    },
+    {
+      "epoch": 2.9433879781420766,
+      "grad_norm": 0.8316853867671832,
+      "learning_rate": 1.8590272001123288e-07,
+      "loss": 11.9276,
+      "step": 16831
+    },
+    {
+      "epoch": 2.9435628415300545,
+      "grad_norm": 0.804731425872179,
+      "learning_rate": 1.847537554997114e-07,
+      "loss": 11.8263,
+      "step": 16832
+    },
+    {
+      "epoch": 2.943737704918033,
+      "grad_norm": 0.828333137538221,
+      "learning_rate": 1.8360834928511373e-07,
+      "loss": 11.9217,
+      "step": 16833
+    },
+    {
+      "epoch": 2.943912568306011,
+      "grad_norm": 0.7837064920515823,
+      "learning_rate": 1.824665014082405e-07,
+      "loss": 11.6743,
+      "step": 16834
+    },
+    {
+      "epoch": 2.944087431693989,
+      "grad_norm": 0.8817265613208674,
+      "learning_rate": 1.8132821190981476e-07,
+      "loss": 12.023,
+      "step": 16835
+    },
+    {
+      "epoch": 2.944262295081967,
+      "grad_norm": 0.7718283607589802,
+      "learning_rate": 1.8019348083041509e-07,
+      "loss": 12.0787,
+      "step": 16836
+    },
+    {
+      "epoch": 2.9444371584699454,
+      "grad_norm": 0.8192158843994373,
+      "learning_rate": 1.7906230821048697e-07,
+      "loss": 11.7817,
+      "step": 16837
+    },
+    {
+      "epoch": 2.9446120218579237,
+      "grad_norm": 0.8225627246731213,
+      "learning_rate": 1.779346940903648e-07,
+      "loss": 12.044,
+      "step": 16838
+    },
+    {
+      "epoch": 2.9447868852459016,
+      "grad_norm": 0.8258252681255318,
+      "learning_rate": 1.768106385102386e-07,
+      "loss": 11.7111,
+      "step": 16839
+    },
+    {
+      "epoch": 2.94496174863388,
+      "grad_norm": 0.8366519908482815,
+      "learning_rate": 1.7569014151017638e-07,
+      "loss": 11.5867,
+      "step": 16840
+    },
+    {
+      "epoch": 2.945136612021858,
+      "grad_norm": 0.8523042443417753,
+      "learning_rate": 1.7457320313013502e-07,
+      "loss": 12.0127,
+      "step": 16841
+    },
+    {
+      "epoch": 2.945311475409836,
+      "grad_norm": 0.8164908955126096,
+      "learning_rate": 1.7345982340992717e-07,
+      "loss": 11.9883,
+      "step": 16842
+    },
+    {
+      "epoch": 2.945486338797814,
+      "grad_norm": 0.7946023629252482,
+      "learning_rate": 1.7235000238924326e-07,
+      "loss": 11.7436,
+      "step": 16843
+    },
+    {
+      "epoch": 2.9456612021857924,
+      "grad_norm": 0.8283308393974991,
+      "learning_rate": 1.7124374010764055e-07,
+      "loss": 11.9525,
+      "step": 16844
+    },
+    {
+      "epoch": 2.9458360655737703,
+      "grad_norm": 0.8911657750081872,
+      "learning_rate": 1.7014103660456525e-07,
+      "loss": 12.0097,
+      "step": 16845
+    },
+    {
+      "epoch": 2.9460109289617487,
+      "grad_norm": 0.887742448896431,
+      "learning_rate": 1.6904189191931928e-07,
+      "loss": 11.775,
+      "step": 16846
+    },
+    {
+      "epoch": 2.946185792349727,
+      "grad_norm": 0.8296508453944152,
+      "learning_rate": 1.6794630609108242e-07,
+      "loss": 11.7615,
+      "step": 16847
+    },
+    {
+      "epoch": 2.946360655737705,
+      "grad_norm": 0.8237768328118151,
+      "learning_rate": 1.6685427915893448e-07,
+      "loss": 11.9878,
+      "step": 16848
+    },
+    {
+      "epoch": 2.946535519125683,
+      "grad_norm": 0.8465304731583583,
+      "learning_rate": 1.657658111617777e-07,
+      "loss": 11.5257,
+      "step": 16849
+    },
+    {
+      "epoch": 2.946710382513661,
+      "grad_norm": 0.8040453913114642,
+      "learning_rate": 1.6468090213843656e-07,
+      "loss": 11.7824,
+      "step": 16850
+    },
+    {
+      "epoch": 2.9468852459016395,
+      "grad_norm": 0.8385187403679405,
+      "learning_rate": 1.6359955212756906e-07,
+      "loss": 11.9364,
+      "step": 16851
+    },
+    {
+      "epoch": 2.9470601092896174,
+      "grad_norm": 0.8518729780867365,
+      "learning_rate": 1.625217611677332e-07,
+      "loss": 11.932,
+      "step": 16852
+    },
+    {
+      "epoch": 2.9472349726775957,
+      "grad_norm": 0.8729618263928707,
+      "learning_rate": 1.6144752929734275e-07,
+      "loss": 11.6658,
+      "step": 16853
+    },
+    {
+      "epoch": 2.9474098360655736,
+      "grad_norm": 0.8575707626147939,
+      "learning_rate": 1.6037685655470037e-07,
+      "loss": 11.6939,
+      "step": 16854
+    },
+    {
+      "epoch": 2.947584699453552,
+      "grad_norm": 0.7998686017720421,
+      "learning_rate": 1.593097429779755e-07,
+      "loss": 12.0527,
+      "step": 16855
+    },
+    {
+      "epoch": 2.9477595628415303,
+      "grad_norm": 0.8153943282389554,
+      "learning_rate": 1.5824618860521557e-07,
+      "loss": 12.0206,
+      "step": 16856
+    },
+    {
+      "epoch": 2.947934426229508,
+      "grad_norm": 0.8285242104109136,
+      "learning_rate": 1.571861934743124e-07,
+      "loss": 11.3207,
+      "step": 16857
+    },
+    {
+      "epoch": 2.948109289617486,
+      "grad_norm": 0.8873772346777408,
+      "learning_rate": 1.5612975762308024e-07,
+      "loss": 11.9521,
+      "step": 16858
+    },
+    {
+      "epoch": 2.9482841530054644,
+      "grad_norm": 0.8589120416507513,
+      "learning_rate": 1.5507688108915564e-07,
+      "loss": 11.8744,
+      "step": 16859
+    },
+    {
+      "epoch": 2.9484590163934428,
+      "grad_norm": 0.7813916458543383,
+      "learning_rate": 1.5402756391008633e-07,
+      "loss": 11.8043,
+      "step": 16860
+    },
+    {
+      "epoch": 2.9486338797814207,
+      "grad_norm": 0.8113409827484033,
+      "learning_rate": 1.5298180612328683e-07,
+      "loss": 11.707,
+      "step": 16861
+    },
+    {
+      "epoch": 2.948808743169399,
+      "grad_norm": 0.894908802297204,
+      "learning_rate": 1.5193960776601623e-07,
+      "loss": 11.825,
+      "step": 16862
+    },
+    {
+      "epoch": 2.948983606557377,
+      "grad_norm": 0.8703686674977119,
+      "learning_rate": 1.5090096887544481e-07,
+      "loss": 11.6086,
+      "step": 16863
+    },
+    {
+      "epoch": 2.9491584699453552,
+      "grad_norm": 0.7883532058786589,
+      "learning_rate": 1.4986588948858737e-07,
+      "loss": 11.6652,
+      "step": 16864
+    },
+    {
+      "epoch": 2.9493333333333336,
+      "grad_norm": 0.8192058159714009,
+      "learning_rate": 1.4883436964235885e-07,
+      "loss": 11.8591,
+      "step": 16865
+    },
+    {
+      "epoch": 2.9495081967213115,
+      "grad_norm": 0.7956176300180147,
+      "learning_rate": 1.4780640937350764e-07,
+      "loss": 11.7282,
+      "step": 16866
+    },
+    {
+      "epoch": 2.9496830601092894,
+      "grad_norm": 0.8721461211632241,
+      "learning_rate": 1.467820087187044e-07,
+      "loss": 11.9041,
+      "step": 16867
+    },
+    {
+      "epoch": 2.9498579234972677,
+      "grad_norm": 0.8228413970538543,
+      "learning_rate": 1.4576116771445326e-07,
+      "loss": 11.969,
+      "step": 16868
+    },
+    {
+      "epoch": 2.950032786885246,
+      "grad_norm": 0.829025701740683,
+      "learning_rate": 1.447438863971473e-07,
+      "loss": 12.028,
+      "step": 16869
+    },
+    {
+      "epoch": 2.950207650273224,
+      "grad_norm": 0.8846687653035193,
+      "learning_rate": 1.4373016480305755e-07,
+      "loss": 11.626,
+      "step": 16870
+    },
+    {
+      "epoch": 2.9503825136612023,
+      "grad_norm": 0.8464090238218157,
+      "learning_rate": 1.4272000296831068e-07,
+      "loss": 11.6939,
+      "step": 16871
+    },
+    {
+      "epoch": 2.95055737704918,
+      "grad_norm": 0.8085022823595499,
+      "learning_rate": 1.4171340092893337e-07,
+      "loss": 11.9155,
+      "step": 16872
+    },
+    {
+      "epoch": 2.9507322404371585,
+      "grad_norm": 0.7473073506692088,
+      "learning_rate": 1.4071035872079698e-07,
+      "loss": 11.7255,
+      "step": 16873
+    },
+    {
+      "epoch": 2.950907103825137,
+      "grad_norm": 0.8384872174019729,
+      "learning_rate": 1.3971087637966174e-07,
+      "loss": 11.56,
+      "step": 16874
+    },
+    {
+      "epoch": 2.9510819672131148,
+      "grad_norm": 0.7580686159659087,
+      "learning_rate": 1.3871495394115474e-07,
+      "loss": 11.9322,
+      "step": 16875
+    },
+    {
+      "epoch": 2.9512568306010927,
+      "grad_norm": 0.8061525123313101,
+      "learning_rate": 1.3772259144079202e-07,
+      "loss": 11.8206,
+      "step": 16876
+    },
+    {
+      "epoch": 2.951431693989071,
+      "grad_norm": 0.8858131005923673,
+      "learning_rate": 1.3673378891394528e-07,
+      "loss": 11.8166,
+      "step": 16877
+    },
+    {
+      "epoch": 2.9516065573770494,
+      "grad_norm": 0.8018506717537887,
+      "learning_rate": 1.35748546395853e-07,
+      "loss": 11.802,
+      "step": 16878
+    },
+    {
+      "epoch": 2.9517814207650273,
+      "grad_norm": 0.8086253569875033,
+      "learning_rate": 1.3476686392165373e-07,
+      "loss": 12.0765,
+      "step": 16879
+    },
+    {
+      "epoch": 2.9519562841530056,
+      "grad_norm": 0.8150838698258202,
+      "learning_rate": 1.337887415263306e-07,
+      "loss": 11.7166,
+      "step": 16880
+    },
+    {
+      "epoch": 2.9521311475409835,
+      "grad_norm": 0.8047383288430684,
+      "learning_rate": 1.328141792447557e-07,
+      "loss": 11.7602,
+      "step": 16881
+    },
+    {
+      "epoch": 2.952306010928962,
+      "grad_norm": 0.8382493633171639,
+      "learning_rate": 1.3184317711167905e-07,
+      "loss": 11.6685,
+      "step": 16882
+    },
+    {
+      "epoch": 2.9524808743169397,
+      "grad_norm": 0.7864917836706732,
+      "learning_rate": 1.3087573516170627e-07,
+      "loss": 11.7633,
+      "step": 16883
+    },
+    {
+      "epoch": 2.952655737704918,
+      "grad_norm": 0.8199781867250031,
+      "learning_rate": 1.29911853429332e-07,
+      "loss": 11.984,
+      "step": 16884
+    },
+    {
+      "epoch": 2.952830601092896,
+      "grad_norm": 0.7904641191702731,
+      "learning_rate": 1.2895153194891764e-07,
+      "loss": 11.6709,
+      "step": 16885
+    },
+    {
+      "epoch": 2.9530054644808743,
+      "grad_norm": 0.8179185438419357,
+      "learning_rate": 1.2799477075468025e-07,
+      "loss": 11.8189,
+      "step": 16886
+    },
+    {
+      "epoch": 2.9531803278688527,
+      "grad_norm": 0.8337953543936618,
+      "learning_rate": 1.270415698807592e-07,
+      "loss": 12.0513,
+      "step": 16887
+    },
+    {
+      "epoch": 2.9533551912568305,
+      "grad_norm": 0.7651969890861688,
+      "learning_rate": 1.2609192936110515e-07,
+      "loss": 11.8089,
+      "step": 16888
+    },
+    {
+      "epoch": 2.953530054644809,
+      "grad_norm": 0.8399509644537461,
+      "learning_rate": 1.2514584922959094e-07,
+      "loss": 11.7175,
+      "step": 16889
+    },
+    {
+      "epoch": 2.953704918032787,
+      "grad_norm": 0.8794659570300504,
+      "learning_rate": 1.2420332951992298e-07,
+      "loss": 11.8687,
+      "step": 16890
+    },
+    {
+      "epoch": 2.953879781420765,
+      "grad_norm": 0.8374942413959322,
+      "learning_rate": 1.232643702657299e-07,
+      "loss": 11.9417,
+      "step": 16891
+    },
+    {
+      "epoch": 2.954054644808743,
+      "grad_norm": 0.7664802681407277,
+      "learning_rate": 1.2232897150046273e-07,
+      "loss": 11.7341,
+      "step": 16892
+    },
+    {
+      "epoch": 2.9542295081967214,
+      "grad_norm": 0.8192870359250807,
+      "learning_rate": 1.213971332574726e-07,
+      "loss": 11.85,
+      "step": 16893
+    },
+    {
+      "epoch": 2.9544043715846993,
+      "grad_norm": 0.8166969915252561,
+      "learning_rate": 1.2046885556997733e-07,
+      "loss": 11.5018,
+      "step": 16894
+    },
+    {
+      "epoch": 2.9545792349726776,
+      "grad_norm": 0.8422920015690782,
+      "learning_rate": 1.1954413847107271e-07,
+      "loss": 11.8167,
+      "step": 16895
+    },
+    {
+      "epoch": 2.954754098360656,
+      "grad_norm": 0.8409368043361298,
+      "learning_rate": 1.1862298199371014e-07,
+      "loss": 11.7047,
+      "step": 16896
+    },
+    {
+      "epoch": 2.954928961748634,
+      "grad_norm": 0.8147311053965606,
+      "learning_rate": 1.1770538617075223e-07,
+      "loss": 11.8571,
+      "step": 16897
+    },
+    {
+      "epoch": 2.955103825136612,
+      "grad_norm": 0.8216470719563965,
+      "learning_rate": 1.1679135103489502e-07,
+      "loss": 11.8455,
+      "step": 16898
+    },
+    {
+      "epoch": 2.95527868852459,
+      "grad_norm": 0.8119885862914894,
+      "learning_rate": 1.158808766187347e-07,
+      "loss": 11.359,
+      "step": 16899
+    },
+    {
+      "epoch": 2.9554535519125684,
+      "grad_norm": 0.8929297565788724,
+      "learning_rate": 1.1497396295470087e-07,
+      "loss": 11.6141,
+      "step": 16900
+    },
+    {
+      "epoch": 2.9556284153005463,
+      "grad_norm": 0.8707394532647463,
+      "learning_rate": 1.1407061007515651e-07,
+      "loss": 11.8641,
+      "step": 16901
+    },
+    {
+      "epoch": 2.9558032786885247,
+      "grad_norm": 0.7917370583825116,
+      "learning_rate": 1.1317081801228701e-07,
+      "loss": 11.972,
+      "step": 16902
+    },
+    {
+      "epoch": 2.9559781420765026,
+      "grad_norm": 0.8134336407305023,
+      "learning_rate": 1.1227458679816671e-07,
+      "loss": 11.9486,
+      "step": 16903
+    },
+    {
+      "epoch": 2.956153005464481,
+      "grad_norm": 0.7852689568853938,
+      "learning_rate": 1.1138191646475893e-07,
+      "loss": 11.909,
+      "step": 16904
+    },
+    {
+      "epoch": 2.9563278688524592,
+      "grad_norm": 0.8226517687241087,
+      "learning_rate": 1.1049280704387155e-07,
+      "loss": 11.3653,
+      "step": 16905
+    },
+    {
+      "epoch": 2.956502732240437,
+      "grad_norm": 0.7587554350553442,
+      "learning_rate": 1.0960725856721255e-07,
+      "loss": 11.9926,
+      "step": 16906
+    },
+    {
+      "epoch": 2.956677595628415,
+      "grad_norm": 0.8144300415458452,
+      "learning_rate": 1.0872527106634556e-07,
+      "loss": 11.8179,
+      "step": 16907
+    },
+    {
+      "epoch": 2.9568524590163934,
+      "grad_norm": 0.8230400297164588,
+      "learning_rate": 1.078468445727121e-07,
+      "loss": 11.8212,
+      "step": 16908
+    },
+    {
+      "epoch": 2.9570273224043717,
+      "grad_norm": 0.8196435649908351,
+      "learning_rate": 1.0697197911762047e-07,
+      "loss": 11.74,
+      "step": 16909
+    },
+    {
+      "epoch": 2.9572021857923496,
+      "grad_norm": 0.8299217152060527,
+      "learning_rate": 1.0610067473226792e-07,
+      "loss": 11.672,
+      "step": 16910
+    },
+    {
+      "epoch": 2.957377049180328,
+      "grad_norm": 0.7941566290927997,
+      "learning_rate": 1.0523293144771851e-07,
+      "loss": 11.7862,
+      "step": 16911
+    },
+    {
+      "epoch": 2.957551912568306,
+      "grad_norm": 0.8185700780650372,
+      "learning_rate": 1.0436874929489193e-07,
+      "loss": 11.6304,
+      "step": 16912
+    },
+    {
+      "epoch": 2.957726775956284,
+      "grad_norm": 0.8199329011894364,
+      "learning_rate": 1.03508128304608e-07,
+      "loss": 11.673,
+      "step": 16913
+    },
+    {
+      "epoch": 2.9579016393442625,
+      "grad_norm": 0.8312362656162837,
+      "learning_rate": 1.0265106850755324e-07,
+      "loss": 11.9669,
+      "step": 16914
+    },
+    {
+      "epoch": 2.9580765027322404,
+      "grad_norm": 0.7861699465989802,
+      "learning_rate": 1.0179756993425882e-07,
+      "loss": 11.82,
+      "step": 16915
+    },
+    {
+      "epoch": 2.9582513661202183,
+      "grad_norm": 0.9502613866928433,
+      "learning_rate": 1.0094763261516705e-07,
+      "loss": 11.8589,
+      "step": 16916
+    },
+    {
+      "epoch": 2.9584262295081967,
+      "grad_norm": 0.8694256130769183,
+      "learning_rate": 1.0010125658056479e-07,
+      "loss": 11.9163,
+      "step": 16917
+    },
+    {
+      "epoch": 2.958601092896175,
+      "grad_norm": 0.8253190711523787,
+      "learning_rate": 9.925844186063904e-08,
+      "loss": 11.6206,
+      "step": 16918
+    },
+    {
+      "epoch": 2.958775956284153,
+      "grad_norm": 0.7797860300633985,
+      "learning_rate": 9.84191884854213e-08,
+      "loss": 12.0234,
+      "step": 16919
+    },
+    {
+      "epoch": 2.9589508196721312,
+      "grad_norm": 0.8590270572779337,
+      "learning_rate": 9.758349648484321e-08,
+      "loss": 11.6498,
+      "step": 16920
+    },
+    {
+      "epoch": 2.959125683060109,
+      "grad_norm": 0.8662013145494504,
+      "learning_rate": 9.675136588869204e-08,
+      "loss": 11.4579,
+      "step": 16921
+    },
+    {
+      "epoch": 2.9593005464480875,
+      "grad_norm": 0.8121393088995996,
+      "learning_rate": 9.592279672663296e-08,
+      "loss": 11.7779,
+      "step": 16922
+    },
+    {
+      "epoch": 2.959475409836066,
+      "grad_norm": 0.7752638715819413,
+      "learning_rate": 9.509778902819788e-08,
+      "loss": 11.6765,
+      "step": 16923
+    },
+    {
+      "epoch": 2.9596502732240437,
+      "grad_norm": 0.8204480206895899,
+      "learning_rate": 9.427634282279662e-08,
+      "loss": 11.6819,
+      "step": 16924
+    },
+    {
+      "epoch": 2.9598251366120216,
+      "grad_norm": 0.8546226035486755,
+      "learning_rate": 9.345845813971688e-08,
+      "loss": 11.9049,
+      "step": 16925
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.814879335153419,
+      "learning_rate": 9.264413500811309e-08,
+      "loss": 11.8542,
+      "step": 16926
+    },
+    {
+      "epoch": 2.9601748633879783,
+      "grad_norm": 0.8413215574564328,
+      "learning_rate": 9.18333734570176e-08,
+      "loss": 11.9061,
+      "step": 16927
+    },
+    {
+      "epoch": 2.960349726775956,
+      "grad_norm": 0.8108571870549617,
+      "learning_rate": 9.102617351532949e-08,
+      "loss": 12.0338,
+      "step": 16928
+    },
+    {
+      "epoch": 2.9605245901639345,
+      "grad_norm": 0.7834555024898794,
+      "learning_rate": 9.022253521183687e-08,
+      "loss": 11.7502,
+      "step": 16929
+    },
+    {
+      "epoch": 2.9606994535519124,
+      "grad_norm": 0.8004508042180721,
+      "learning_rate": 8.94224585751724e-08,
+      "loss": 11.5447,
+      "step": 16930
+    },
+    {
+      "epoch": 2.960874316939891,
+      "grad_norm": 0.7868859035682324,
+      "learning_rate": 8.862594363386878e-08,
+      "loss": 11.9618,
+      "step": 16931
+    },
+    {
+      "epoch": 2.961049180327869,
+      "grad_norm": 0.8536174182249268,
+      "learning_rate": 8.783299041631444e-08,
+      "loss": 11.7602,
+      "step": 16932
+    },
+    {
+      "epoch": 2.961224043715847,
+      "grad_norm": 0.8502106946883525,
+      "learning_rate": 8.704359895077563e-08,
+      "loss": 11.7569,
+      "step": 16933
+    },
+    {
+      "epoch": 2.961398907103825,
+      "grad_norm": 0.8207209254106581,
+      "learning_rate": 8.625776926539653e-08,
+      "loss": 11.9051,
+      "step": 16934
+    },
+    {
+      "epoch": 2.9615737704918033,
+      "grad_norm": 0.847780083504116,
+      "learning_rate": 8.547550138819915e-08,
+      "loss": 11.9069,
+      "step": 16935
+    },
+    {
+      "epoch": 2.9617486338797816,
+      "grad_norm": 0.8088619638698804,
+      "learning_rate": 8.46967953470501e-08,
+      "loss": 11.8616,
+      "step": 16936
+    },
+    {
+      "epoch": 2.9619234972677595,
+      "grad_norm": 0.931279385483136,
+      "learning_rate": 8.392165116972717e-08,
+      "loss": 11.8752,
+      "step": 16937
+    },
+    {
+      "epoch": 2.962098360655738,
+      "grad_norm": 0.9338850562893409,
+      "learning_rate": 8.315006888386379e-08,
+      "loss": 11.6789,
+      "step": 16938
+    },
+    {
+      "epoch": 2.9622732240437157,
+      "grad_norm": 0.8792690086301204,
+      "learning_rate": 8.238204851696019e-08,
+      "loss": 11.7917,
+      "step": 16939
+    },
+    {
+      "epoch": 2.962448087431694,
+      "grad_norm": 0.768322461913467,
+      "learning_rate": 8.161759009638337e-08,
+      "loss": 11.7938,
+      "step": 16940
+    },
+    {
+      "epoch": 2.962622950819672,
+      "grad_norm": 0.9203830244753912,
+      "learning_rate": 8.085669364941151e-08,
+      "loss": 11.7773,
+      "step": 16941
+    },
+    {
+      "epoch": 2.9627978142076503,
+      "grad_norm": 0.8204609137018294,
+      "learning_rate": 8.009935920314516e-08,
+      "loss": 11.7084,
+      "step": 16942
+    },
+    {
+      "epoch": 2.962972677595628,
+      "grad_norm": 0.8401084953247371,
+      "learning_rate": 7.934558678459603e-08,
+      "loss": 11.9525,
+      "step": 16943
+    },
+    {
+      "epoch": 2.9631475409836066,
+      "grad_norm": 0.8488619941531557,
+      "learning_rate": 7.859537642063153e-08,
+      "loss": 12.0968,
+      "step": 16944
+    },
+    {
+      "epoch": 2.963322404371585,
+      "grad_norm": 0.8639341299321126,
+      "learning_rate": 7.784872813799693e-08,
+      "loss": 11.9606,
+      "step": 16945
+    },
+    {
+      "epoch": 2.963497267759563,
+      "grad_norm": 0.7730519418465762,
+      "learning_rate": 7.710564196331537e-08,
+      "loss": 11.748,
+      "step": 16946
+    },
+    {
+      "epoch": 2.963672131147541,
+      "grad_norm": 0.8491240986669121,
+      "learning_rate": 7.636611792306569e-08,
+      "loss": 11.9777,
+      "step": 16947
+    },
+    {
+      "epoch": 2.963846994535519,
+      "grad_norm": 0.7844411316637152,
+      "learning_rate": 7.563015604360457e-08,
+      "loss": 11.5671,
+      "step": 16948
+    },
+    {
+      "epoch": 2.9640218579234974,
+      "grad_norm": 0.8060101931264176,
+      "learning_rate": 7.489775635118878e-08,
+      "loss": 11.7757,
+      "step": 16949
+    },
+    {
+      "epoch": 2.9641967213114753,
+      "grad_norm": 0.8125877786042509,
+      "learning_rate": 7.416891887191968e-08,
+      "loss": 11.7158,
+      "step": 16950
+    },
+    {
+      "epoch": 2.9643715846994536,
+      "grad_norm": 0.8477670353393172,
+      "learning_rate": 7.344364363177647e-08,
+      "loss": 11.5413,
+      "step": 16951
+    },
+    {
+      "epoch": 2.9645464480874315,
+      "grad_norm": 0.8591948449095553,
+      "learning_rate": 7.272193065661625e-08,
+      "loss": 11.9486,
+      "step": 16952
+    },
+    {
+      "epoch": 2.96472131147541,
+      "grad_norm": 0.7737586834990686,
+      "learning_rate": 7.2003779972174e-08,
+      "loss": 11.8401,
+      "step": 16953
+    },
+    {
+      "epoch": 2.964896174863388,
+      "grad_norm": 0.8909730451172952,
+      "learning_rate": 7.128919160402924e-08,
+      "loss": 11.886,
+      "step": 16954
+    },
+    {
+      "epoch": 2.965071038251366,
+      "grad_norm": 0.8356211159390309,
+      "learning_rate": 7.057816557768382e-08,
+      "loss": 11.324,
+      "step": 16955
+    },
+    {
+      "epoch": 2.9652459016393444,
+      "grad_norm": 0.8117190836586392,
+      "learning_rate": 6.98707019184619e-08,
+      "loss": 11.8242,
+      "step": 16956
+    },
+    {
+      "epoch": 2.9654207650273223,
+      "grad_norm": 0.8302974838928209,
+      "learning_rate": 6.916680065159886e-08,
+      "loss": 11.9531,
+      "step": 16957
+    },
+    {
+      "epoch": 2.9655956284153007,
+      "grad_norm": 0.8910606180743026,
+      "learning_rate": 6.846646180217464e-08,
+      "loss": 11.7682,
+      "step": 16958
+    },
+    {
+      "epoch": 2.9657704918032786,
+      "grad_norm": 0.957793083673802,
+      "learning_rate": 6.776968539516926e-08,
+      "loss": 11.8162,
+      "step": 16959
+    },
+    {
+      "epoch": 2.965945355191257,
+      "grad_norm": 0.8054108093853106,
+      "learning_rate": 6.70764714554184e-08,
+      "loss": 11.8859,
+      "step": 16960
+    },
+    {
+      "epoch": 2.966120218579235,
+      "grad_norm": 0.7847897010058779,
+      "learning_rate": 6.638682000763562e-08,
+      "loss": 11.826,
+      "step": 16961
+    },
+    {
+      "epoch": 2.966295081967213,
+      "grad_norm": 0.8615404654530852,
+      "learning_rate": 6.570073107639019e-08,
+      "loss": 12.0272,
+      "step": 16962
+    },
+    {
+      "epoch": 2.9664699453551915,
+      "grad_norm": 0.7828303083854282,
+      "learning_rate": 6.501820468616248e-08,
+      "loss": 11.7188,
+      "step": 16963
+    },
+    {
+      "epoch": 2.9666448087431694,
+      "grad_norm": 0.8577783174509296,
+      "learning_rate": 6.433924086127752e-08,
+      "loss": 11.6511,
+      "step": 16964
+    },
+    {
+      "epoch": 2.9668196721311473,
+      "grad_norm": 0.7707434152094131,
+      "learning_rate": 6.366383962592703e-08,
+      "loss": 11.8149,
+      "step": 16965
+    },
+    {
+      "epoch": 2.9669945355191256,
+      "grad_norm": 0.8273527570782554,
+      "learning_rate": 6.299200100420289e-08,
+      "loss": 11.8916,
+      "step": 16966
+    },
+    {
+      "epoch": 2.967169398907104,
+      "grad_norm": 0.8246951852131876,
+      "learning_rate": 6.232372502004147e-08,
+      "loss": 11.8832,
+      "step": 16967
+    },
+    {
+      "epoch": 2.967344262295082,
+      "grad_norm": 0.8554148739756164,
+      "learning_rate": 6.165901169729038e-08,
+      "loss": 11.9862,
+      "step": 16968
+    },
+    {
+      "epoch": 2.96751912568306,
+      "grad_norm": 0.8064987235391482,
+      "learning_rate": 6.099786105961958e-08,
+      "loss": 11.9366,
+      "step": 16969
+    },
+    {
+      "epoch": 2.967693989071038,
+      "grad_norm": 0.875266890674951,
+      "learning_rate": 6.03402731306102e-08,
+      "loss": 11.8067,
+      "step": 16970
+    },
+    {
+      "epoch": 2.9678688524590164,
+      "grad_norm": 0.75578627507385,
+      "learning_rate": 5.968624793369903e-08,
+      "loss": 11.7419,
+      "step": 16971
+    },
+    {
+      "epoch": 2.968043715846995,
+      "grad_norm": 0.8022671586297876,
+      "learning_rate": 5.903578549221189e-08,
+      "loss": 11.9699,
+      "step": 16972
+    },
+    {
+      "epoch": 2.9682185792349727,
+      "grad_norm": 0.823224982250502,
+      "learning_rate": 5.838888582933022e-08,
+      "loss": 11.7815,
+      "step": 16973
+    },
+    {
+      "epoch": 2.9683934426229506,
+      "grad_norm": 0.797892854862712,
+      "learning_rate": 5.774554896811335e-08,
+      "loss": 11.7587,
+      "step": 16974
+    },
+    {
+      "epoch": 2.968568306010929,
+      "grad_norm": 0.8108641709997378,
+      "learning_rate": 5.710577493149849e-08,
+      "loss": 11.8992,
+      "step": 16975
+    },
+    {
+      "epoch": 2.9687431693989073,
+      "grad_norm": 0.7873635223801765,
+      "learning_rate": 5.646956374230072e-08,
+      "loss": 11.8307,
+      "step": 16976
+    },
+    {
+      "epoch": 2.968918032786885,
+      "grad_norm": 0.8181179604198947,
+      "learning_rate": 5.58369154231797e-08,
+      "loss": 11.8885,
+      "step": 16977
+    },
+    {
+      "epoch": 2.9690928961748635,
+      "grad_norm": 0.7776248459771156,
+      "learning_rate": 5.520782999670626e-08,
+      "loss": 11.77,
+      "step": 16978
+    },
+    {
+      "epoch": 2.9692677595628414,
+      "grad_norm": 0.81625011270914,
+      "learning_rate": 5.45823074852958e-08,
+      "loss": 11.9393,
+      "step": 16979
+    },
+    {
+      "epoch": 2.9694426229508197,
+      "grad_norm": 0.809624023112868,
+      "learning_rate": 5.3960347911252705e-08,
+      "loss": 11.7481,
+      "step": 16980
+    },
+    {
+      "epoch": 2.969617486338798,
+      "grad_norm": 0.8598671357776931,
+      "learning_rate": 5.334195129674813e-08,
+      "loss": 11.5732,
+      "step": 16981
+    },
+    {
+      "epoch": 2.969792349726776,
+      "grad_norm": 0.8491890933272156,
+      "learning_rate": 5.2727117663831097e-08,
+      "loss": 11.4778,
+      "step": 16982
+    },
+    {
+      "epoch": 2.969967213114754,
+      "grad_norm": 0.8090777477820557,
+      "learning_rate": 5.2115847034417406e-08,
+      "loss": 11.6193,
+      "step": 16983
+    },
+    {
+      "epoch": 2.970142076502732,
+      "grad_norm": 0.8797097217733348,
+      "learning_rate": 5.150813943028965e-08,
+      "loss": 11.9228,
+      "step": 16984
+    },
+    {
+      "epoch": 2.9703169398907106,
+      "grad_norm": 0.8160776009986154,
+      "learning_rate": 5.0903994873130464e-08,
+      "loss": 11.7878,
+      "step": 16985
+    },
+    {
+      "epoch": 2.9704918032786884,
+      "grad_norm": 0.7995267669307974,
+      "learning_rate": 5.0303413384455986e-08,
+      "loss": 11.9404,
+      "step": 16986
+    },
+    {
+      "epoch": 2.970666666666667,
+      "grad_norm": 0.8965963832620711,
+      "learning_rate": 4.970639498569351e-08,
+      "loss": 12.0609,
+      "step": 16987
+    },
+    {
+      "epoch": 2.9708415300546447,
+      "grad_norm": 0.7972298757265436,
+      "learning_rate": 4.911293969811492e-08,
+      "loss": 11.8083,
+      "step": 16988
+    },
+    {
+      "epoch": 2.971016393442623,
+      "grad_norm": 0.8814491185626652,
+      "learning_rate": 4.852304754288106e-08,
+      "loss": 11.8596,
+      "step": 16989
+    },
+    {
+      "epoch": 2.9711912568306014,
+      "grad_norm": 0.8688208284038095,
+      "learning_rate": 4.7936718541019555e-08,
+      "loss": 11.745,
+      "step": 16990
+    },
+    {
+      "epoch": 2.9713661202185793,
+      "grad_norm": 0.8237445073839315,
+      "learning_rate": 4.73539527134359e-08,
+      "loss": 11.7708,
+      "step": 16991
+    },
+    {
+      "epoch": 2.971540983606557,
+      "grad_norm": 0.8126556247769201,
+      "learning_rate": 4.6774750080902373e-08,
+      "loss": 11.9542,
+      "step": 16992
+    },
+    {
+      "epoch": 2.9717158469945355,
+      "grad_norm": 0.8140858869197868,
+      "learning_rate": 4.619911066405802e-08,
+      "loss": 11.7442,
+      "step": 16993
+    },
+    {
+      "epoch": 2.971890710382514,
+      "grad_norm": 0.801094021221078,
+      "learning_rate": 4.5627034483441963e-08,
+      "loss": 11.9624,
+      "step": 16994
+    },
+    {
+      "epoch": 2.9720655737704917,
+      "grad_norm": 0.8341211977115585,
+      "learning_rate": 4.5058521559426804e-08,
+      "loss": 12.1904,
+      "step": 16995
+    },
+    {
+      "epoch": 2.97224043715847,
+      "grad_norm": 0.8697426157973974,
+      "learning_rate": 4.4493571912296307e-08,
+      "loss": 11.7114,
+      "step": 16996
+    },
+    {
+      "epoch": 2.972415300546448,
+      "grad_norm": 0.8273581848101742,
+      "learning_rate": 4.393218556218992e-08,
+      "loss": 11.8683,
+      "step": 16997
+    },
+    {
+      "epoch": 2.9725901639344263,
+      "grad_norm": 0.8340364969145454,
+      "learning_rate": 4.337436252910276e-08,
+      "loss": 11.4367,
+      "step": 16998
+    },
+    {
+      "epoch": 2.972765027322404,
+      "grad_norm": 0.8111242870315082,
+      "learning_rate": 4.282010283294113e-08,
+      "loss": 12.0673,
+      "step": 16999
+    },
+    {
+      "epoch": 2.9729398907103826,
+      "grad_norm": 0.8975876042375583,
+      "learning_rate": 4.22694064934559e-08,
+      "loss": 11.7636,
+      "step": 17000
+    },
+    {
+      "epoch": 2.9731147540983605,
+      "grad_norm": 0.8817518935460475,
+      "learning_rate": 4.1722273530275804e-08,
+      "loss": 11.8319,
+      "step": 17001
+    },
+    {
+      "epoch": 2.973289617486339,
+      "grad_norm": 0.8295708537021773,
+      "learning_rate": 4.117870396290746e-08,
+      "loss": 11.9074,
+      "step": 17002
+    },
+    {
+      "epoch": 2.973464480874317,
+      "grad_norm": 0.7757830643449918,
+      "learning_rate": 4.063869781073537e-08,
+      "loss": 11.7177,
+      "step": 17003
+    },
+    {
+      "epoch": 2.973639344262295,
+      "grad_norm": 0.8592677169495526,
+      "learning_rate": 4.0102255092988596e-08,
+      "loss": 11.8383,
+      "step": 17004
+    },
+    {
+      "epoch": 2.9738142076502734,
+      "grad_norm": 0.8222188953704419,
+      "learning_rate": 3.9569375828818475e-08,
+      "loss": 11.7994,
+      "step": 17005
+    },
+    {
+      "epoch": 2.9739890710382513,
+      "grad_norm": 0.7737948675622716,
+      "learning_rate": 3.9040060037209837e-08,
+      "loss": 11.7797,
+      "step": 17006
+    },
+    {
+      "epoch": 2.9741639344262296,
+      "grad_norm": 0.8219716839504907,
+      "learning_rate": 3.8514307737025356e-08,
+      "loss": 11.6434,
+      "step": 17007
+    },
+    {
+      "epoch": 2.9743387978142075,
+      "grad_norm": 0.8021023747475531,
+      "learning_rate": 3.799211894701671e-08,
+      "loss": 11.8761,
+      "step": 17008
+    },
+    {
+      "epoch": 2.974513661202186,
+      "grad_norm": 0.8927603636355264,
+      "learning_rate": 3.747349368579123e-08,
+      "loss": 11.514,
+      "step": 17009
+    },
+    {
+      "epoch": 2.9746885245901638,
+      "grad_norm": 0.8360395795479363,
+      "learning_rate": 3.695843197184523e-08,
+      "loss": 12.0031,
+      "step": 17010
+    },
+    {
+      "epoch": 2.974863387978142,
+      "grad_norm": 0.7731407395851727,
+      "learning_rate": 3.644693382354181e-08,
+      "loss": 11.6869,
+      "step": 17011
+    },
+    {
+      "epoch": 2.9750382513661204,
+      "grad_norm": 0.8668516157035026,
+      "learning_rate": 3.5938999259099714e-08,
+      "loss": 11.8975,
+      "step": 17012
+    },
+    {
+      "epoch": 2.9752131147540983,
+      "grad_norm": 0.8061085336273904,
+      "learning_rate": 3.54346282966489e-08,
+      "loss": 11.6628,
+      "step": 17013
+    },
+    {
+      "epoch": 2.9753879781420762,
+      "grad_norm": 0.8848906453190795,
+      "learning_rate": 3.493382095415276e-08,
+      "loss": 11.793,
+      "step": 17014
+    },
+    {
+      "epoch": 2.9755628415300546,
+      "grad_norm": 0.8248868906256728,
+      "learning_rate": 3.44365772494637e-08,
+      "loss": 12.0351,
+      "step": 17015
+    },
+    {
+      "epoch": 2.975737704918033,
+      "grad_norm": 0.9078976431043173,
+      "learning_rate": 3.394289720031196e-08,
+      "loss": 11.5302,
+      "step": 17016
+    },
+    {
+      "epoch": 2.975912568306011,
+      "grad_norm": 0.8140313766493372,
+      "learning_rate": 3.3452780824305695e-08,
+      "loss": 11.8336,
+      "step": 17017
+    },
+    {
+      "epoch": 2.976087431693989,
+      "grad_norm": 0.8745734299315744,
+      "learning_rate": 3.2966228138908704e-08,
+      "loss": 11.6708,
+      "step": 17018
+    },
+    {
+      "epoch": 2.976262295081967,
+      "grad_norm": 0.7340373928178521,
+      "learning_rate": 3.248323916146268e-08,
+      "loss": 12.005,
+      "step": 17019
+    },
+    {
+      "epoch": 2.9764371584699454,
+      "grad_norm": 0.8883453031280952,
+      "learning_rate": 3.200381390919827e-08,
+      "loss": 11.7844,
+      "step": 17020
+    },
+    {
+      "epoch": 2.9766120218579237,
+      "grad_norm": 0.7865233561366882,
+      "learning_rate": 3.152795239919071e-08,
+      "loss": 11.7963,
+      "step": 17021
+    },
+    {
+      "epoch": 2.9767868852459016,
+      "grad_norm": 0.8065040118786088,
+      "learning_rate": 3.1055654648415314e-08,
+      "loss": 11.9628,
+      "step": 17022
+    },
+    {
+      "epoch": 2.9769617486338795,
+      "grad_norm": 0.8337588317867587,
+      "learning_rate": 3.058692067369195e-08,
+      "loss": 11.8667,
+      "step": 17023
+    },
+    {
+      "epoch": 2.977136612021858,
+      "grad_norm": 0.8871555725790347,
+      "learning_rate": 3.01217504917517e-08,
+      "loss": 11.9833,
+      "step": 17024
+    },
+    {
+      "epoch": 2.977311475409836,
+      "grad_norm": 0.860789147603569,
+      "learning_rate": 2.9660144119170176e-08,
+      "loss": 11.6695,
+      "step": 17025
+    },
+    {
+      "epoch": 2.977486338797814,
+      "grad_norm": 0.8000290868249745,
+      "learning_rate": 2.9202101572400886e-08,
+      "loss": 11.8361,
+      "step": 17026
+    },
+    {
+      "epoch": 2.9776612021857924,
+      "grad_norm": 0.8515577922293239,
+      "learning_rate": 2.8747622867775215e-08,
+      "loss": 11.8424,
+      "step": 17027
+    },
+    {
+      "epoch": 2.9778360655737703,
+      "grad_norm": 0.8252847276203803,
+      "learning_rate": 2.8296708021491315e-08,
+      "loss": 11.7594,
+      "step": 17028
+    },
+    {
+      "epoch": 2.9780109289617487,
+      "grad_norm": 0.8740492974079407,
+      "learning_rate": 2.7849357049625213e-08,
+      "loss": 11.5905,
+      "step": 17029
+    },
+    {
+      "epoch": 2.978185792349727,
+      "grad_norm": 0.8853615070446827,
+      "learning_rate": 2.7405569968130818e-08,
+      "loss": 11.7317,
+      "step": 17030
+    },
+    {
+      "epoch": 2.978360655737705,
+      "grad_norm": 0.8517453765443248,
+      "learning_rate": 2.69653467928177e-08,
+      "loss": 11.8794,
+      "step": 17031
+    },
+    {
+      "epoch": 2.978535519125683,
+      "grad_norm": 0.797605108717759,
+      "learning_rate": 2.652868753938442e-08,
+      "loss": 11.7139,
+      "step": 17032
+    },
+    {
+      "epoch": 2.978710382513661,
+      "grad_norm": 0.8235976502401448,
+      "learning_rate": 2.6095592223396304e-08,
+      "loss": 11.9503,
+      "step": 17033
+    },
+    {
+      "epoch": 2.9788852459016395,
+      "grad_norm": 0.7711855282440394,
+      "learning_rate": 2.5666060860285445e-08,
+      "loss": 11.7505,
+      "step": 17034
+    },
+    {
+      "epoch": 2.9790601092896174,
+      "grad_norm": 0.8383637267393949,
+      "learning_rate": 2.5240093465384028e-08,
+      "loss": 11.5547,
+      "step": 17035
+    },
+    {
+      "epoch": 2.9792349726775957,
+      "grad_norm": 0.8098935302375844,
+      "learning_rate": 2.48176900538577e-08,
+      "loss": 11.9489,
+      "step": 17036
+    },
+    {
+      "epoch": 2.9794098360655736,
+      "grad_norm": 0.7928730532693233,
+      "learning_rate": 2.4398850640772187e-08,
+      "loss": 11.9231,
+      "step": 17037
+    },
+    {
+      "epoch": 2.979584699453552,
+      "grad_norm": 0.8927834675094878,
+      "learning_rate": 2.3983575241059987e-08,
+      "loss": 12.0678,
+      "step": 17038
+    },
+    {
+      "epoch": 2.9797595628415303,
+      "grad_norm": 0.8122229200391253,
+      "learning_rate": 2.3571863869531473e-08,
+      "loss": 11.8992,
+      "step": 17039
+    },
+    {
+      "epoch": 2.979934426229508,
+      "grad_norm": 0.7901193651273525,
+      "learning_rate": 2.31637165408527e-08,
+      "loss": 12.0336,
+      "step": 17040
+    },
+    {
+      "epoch": 2.980109289617486,
+      "grad_norm": 0.8035816031853151,
+      "learning_rate": 2.275913326956758e-08,
+      "loss": 11.7707,
+      "step": 17041
+    },
+    {
+      "epoch": 2.9802841530054645,
+      "grad_norm": 0.8138626871112297,
+      "learning_rate": 2.2358114070120118e-08,
+      "loss": 11.6211,
+      "step": 17042
+    },
+    {
+      "epoch": 2.980459016393443,
+      "grad_norm": 0.8000722448096202,
+      "learning_rate": 2.196065895678778e-08,
+      "loss": 11.9246,
+      "step": 17043
+    },
+    {
+      "epoch": 2.9806338797814207,
+      "grad_norm": 0.8601483024472187,
+      "learning_rate": 2.1566767943748124e-08,
+      "loss": 11.809,
+      "step": 17044
+    },
+    {
+      "epoch": 2.980808743169399,
+      "grad_norm": 0.8956661144646264,
+      "learning_rate": 2.117644104504546e-08,
+      "loss": 11.7625,
+      "step": 17045
+    },
+    {
+      "epoch": 2.980983606557377,
+      "grad_norm": 0.864839858453375,
+      "learning_rate": 2.078967827457978e-08,
+      "loss": 11.7396,
+      "step": 17046
+    },
+    {
+      "epoch": 2.9811584699453553,
+      "grad_norm": 0.804602053090235,
+      "learning_rate": 2.0406479646151167e-08,
+      "loss": 11.9351,
+      "step": 17047
+    },
+    {
+      "epoch": 2.981333333333333,
+      "grad_norm": 0.7852657420436275,
+      "learning_rate": 2.0026845173426457e-08,
+      "loss": 11.8391,
+      "step": 17048
+    },
+    {
+      "epoch": 2.9815081967213115,
+      "grad_norm": 0.838351884309178,
+      "learning_rate": 1.9650774869928168e-08,
+      "loss": 11.8509,
+      "step": 17049
+    },
+    {
+      "epoch": 2.9816830601092894,
+      "grad_norm": 0.8072463033666571,
+      "learning_rate": 1.9278268749056694e-08,
+      "loss": 11.9008,
+      "step": 17050
+    },
+    {
+      "epoch": 2.9818579234972677,
+      "grad_norm": 0.8307673170094526,
+      "learning_rate": 1.8909326824112506e-08,
+      "loss": 11.8164,
+      "step": 17051
+    },
+    {
+      "epoch": 2.982032786885246,
+      "grad_norm": 0.7562429666246886,
+      "learning_rate": 1.8543949108229543e-08,
+      "loss": 11.8754,
+      "step": 17052
+    },
+    {
+      "epoch": 2.982207650273224,
+      "grad_norm": 0.9013846613977533,
+      "learning_rate": 1.818213561443072e-08,
+      "loss": 11.711,
+      "step": 17053
+    },
+    {
+      "epoch": 2.9823825136612023,
+      "grad_norm": 0.8412917853191562,
+      "learning_rate": 1.782388635562793e-08,
+      "loss": 11.8874,
+      "step": 17054
+    },
+    {
+      "epoch": 2.9825573770491802,
+      "grad_norm": 0.7931826857541678,
+      "learning_rate": 1.7469201344588736e-08,
+      "loss": 12.0395,
+      "step": 17055
+    },
+    {
+      "epoch": 2.9827322404371586,
+      "grad_norm": 0.8120755266432312,
+      "learning_rate": 1.7118080593958585e-08,
+      "loss": 11.9286,
+      "step": 17056
+    },
+    {
+      "epoch": 2.9829071038251365,
+      "grad_norm": 0.7968640459792795,
+      "learning_rate": 1.6770524116249686e-08,
+      "loss": 12.0105,
+      "step": 17057
+    },
+    {
+      "epoch": 2.983081967213115,
+      "grad_norm": 0.8101564918257264,
+      "learning_rate": 1.642653192385213e-08,
+      "loss": 11.878,
+      "step": 17058
+    },
+    {
+      "epoch": 2.9832568306010927,
+      "grad_norm": 0.7803223856646607,
+      "learning_rate": 1.6086104029033877e-08,
+      "loss": 11.8688,
+      "step": 17059
+    },
+    {
+      "epoch": 2.983431693989071,
+      "grad_norm": 0.7919313464694406,
+      "learning_rate": 1.5749240443918567e-08,
+      "loss": 11.956,
+      "step": 17060
+    },
+    {
+      "epoch": 2.9836065573770494,
+      "grad_norm": 0.8345055511732673,
+      "learning_rate": 1.5415941180529914e-08,
+      "loss": 11.7184,
+      "step": 17061
+    },
+    {
+      "epoch": 2.9837814207650273,
+      "grad_norm": 0.8407430118709174,
+      "learning_rate": 1.5086206250747302e-08,
+      "loss": 11.671,
+      "step": 17062
+    },
+    {
+      "epoch": 2.9839562841530056,
+      "grad_norm": 0.8170337178504729,
+      "learning_rate": 1.4760035666316895e-08,
+      "loss": 11.82,
+      "step": 17063
+    },
+    {
+      "epoch": 2.9841311475409835,
+      "grad_norm": 0.8541389748124856,
+      "learning_rate": 1.4437429438873828e-08,
+      "loss": 11.9009,
+      "step": 17064
+    },
+    {
+      "epoch": 2.984306010928962,
+      "grad_norm": 0.8501067343555323,
+      "learning_rate": 1.411838757990891e-08,
+      "loss": 11.8382,
+      "step": 17065
+    },
+    {
+      "epoch": 2.9844808743169398,
+      "grad_norm": 0.8253352721948636,
+      "learning_rate": 1.3802910100813027e-08,
+      "loss": 11.9337,
+      "step": 17066
+    },
+    {
+      "epoch": 2.984655737704918,
+      "grad_norm": 0.8802200368516925,
+      "learning_rate": 1.3490997012810535e-08,
+      "loss": 11.9496,
+      "step": 17067
+    },
+    {
+      "epoch": 2.984830601092896,
+      "grad_norm": 0.833258939468415,
+      "learning_rate": 1.3182648327048075e-08,
+      "loss": 11.8702,
+      "step": 17068
+    },
+    {
+      "epoch": 2.9850054644808743,
+      "grad_norm": 0.879932006997533,
+      "learning_rate": 1.2877864054483546e-08,
+      "loss": 11.9679,
+      "step": 17069
+    },
+    {
+      "epoch": 2.9851803278688527,
+      "grad_norm": 0.8443139462119794,
+      "learning_rate": 1.2576644206008236e-08,
+      "loss": 11.7885,
+      "step": 17070
+    },
+    {
+      "epoch": 2.9853551912568306,
+      "grad_norm": 0.8196841931082243,
+      "learning_rate": 1.2278988792358004e-08,
+      "loss": 11.8135,
+      "step": 17071
+    },
+    {
+      "epoch": 2.9855300546448085,
+      "grad_norm": 0.8095652260174275,
+      "learning_rate": 1.1984897824135478e-08,
+      "loss": 11.8519,
+      "step": 17072
+    },
+    {
+      "epoch": 2.985704918032787,
+      "grad_norm": 0.817780737290742,
+      "learning_rate": 1.1694371311832264e-08,
+      "loss": 11.6282,
+      "step": 17073
+    },
+    {
+      "epoch": 2.985879781420765,
+      "grad_norm": 0.7752644868518271,
+      "learning_rate": 1.1407409265795644e-08,
+      "loss": 11.5847,
+      "step": 17074
+    },
+    {
+      "epoch": 2.986054644808743,
+      "grad_norm": 0.8195198958310199,
+      "learning_rate": 1.1124011696250769e-08,
+      "loss": 11.7068,
+      "step": 17075
+    },
+    {
+      "epoch": 2.9862295081967214,
+      "grad_norm": 0.8152418579399487,
+      "learning_rate": 1.0844178613322875e-08,
+      "loss": 11.719,
+      "step": 17076
+    },
+    {
+      "epoch": 2.9864043715846993,
+      "grad_norm": 0.7635297367388209,
+      "learning_rate": 1.0567910026970662e-08,
+      "loss": 11.9712,
+      "step": 17077
+    },
+    {
+      "epoch": 2.9865792349726776,
+      "grad_norm": 0.8057796847649377,
+      "learning_rate": 1.029520594705291e-08,
+      "loss": 11.9486,
+      "step": 17078
+    },
+    {
+      "epoch": 2.986754098360656,
+      "grad_norm": 0.8669557847936744,
+      "learning_rate": 1.0026066383284072e-08,
+      "loss": 11.7398,
+      "step": 17079
+    },
+    {
+      "epoch": 2.986928961748634,
+      "grad_norm": 0.8537965165214995,
+      "learning_rate": 9.760491345256473e-09,
+      "loss": 11.6887,
+      "step": 17080
+    },
+    {
+      "epoch": 2.9871038251366118,
+      "grad_norm": 0.8235188557468386,
+      "learning_rate": 9.498480842440316e-09,
+      "loss": 11.951,
+      "step": 17081
+    },
+    {
+      "epoch": 2.98727868852459,
+      "grad_norm": 0.8265860187273264,
+      "learning_rate": 9.24003488418368e-09,
+      "loss": 12.0622,
+      "step": 17082
+    },
+    {
+      "epoch": 2.9874535519125685,
+      "grad_norm": 0.7783802992056859,
+      "learning_rate": 8.985153479690312e-09,
+      "loss": 12.0033,
+      "step": 17083
+    },
+    {
+      "epoch": 2.9876284153005463,
+      "grad_norm": 0.831452882158957,
+      "learning_rate": 8.733836638052938e-09,
+      "loss": 11.9335,
+      "step": 17084
+    },
+    {
+      "epoch": 2.9878032786885247,
+      "grad_norm": 0.8156444407281337,
+      "learning_rate": 8.48608436821996e-09,
+      "loss": 11.4855,
+      "step": 17085
+    },
+    {
+      "epoch": 2.9879781420765026,
+      "grad_norm": 0.8420897184187767,
+      "learning_rate": 8.241896679028748e-09,
+      "loss": 11.9657,
+      "step": 17086
+    },
+    {
+      "epoch": 2.988153005464481,
+      "grad_norm": 0.85068671844195,
+      "learning_rate": 8.001273579194558e-09,
+      "loss": 11.8561,
+      "step": 17087
+    },
+    {
+      "epoch": 2.9883278688524593,
+      "grad_norm": 0.8059543455417748,
+      "learning_rate": 7.764215077277203e-09,
+      "loss": 11.8623,
+      "step": 17088
+    },
+    {
+      "epoch": 2.988502732240437,
+      "grad_norm": 0.8692383931026236,
+      "learning_rate": 7.530721181747691e-09,
+      "loss": 11.8942,
+      "step": 17089
+    },
+    {
+      "epoch": 2.988677595628415,
+      "grad_norm": 0.7725983009879168,
+      "learning_rate": 7.300791900910486e-09,
+      "loss": 12.0037,
+      "step": 17090
+    },
+    {
+      "epoch": 2.9888524590163934,
+      "grad_norm": 0.8257710632188697,
+      "learning_rate": 7.074427242981241e-09,
+      "loss": 11.9831,
+      "step": 17091
+    },
+    {
+      "epoch": 2.9890273224043717,
+      "grad_norm": 0.7983328476452153,
+      "learning_rate": 6.851627216009071e-09,
+      "loss": 11.8,
+      "step": 17092
+    },
+    {
+      "epoch": 2.9892021857923496,
+      "grad_norm": 0.8160412126919727,
+      "learning_rate": 6.632391827954276e-09,
+      "loss": 11.9055,
+      "step": 17093
+    },
+    {
+      "epoch": 2.989377049180328,
+      "grad_norm": 0.8170102441588686,
+      "learning_rate": 6.416721086621724e-09,
+      "loss": 11.6018,
+      "step": 17094
+    },
+    {
+      "epoch": 2.989551912568306,
+      "grad_norm": 0.8182838225884195,
+      "learning_rate": 6.20461499970526e-09,
+      "loss": 11.7433,
+      "step": 17095
+    },
+    {
+      "epoch": 2.989726775956284,
+      "grad_norm": 0.8445245508742795,
+      "learning_rate": 5.996073574754402e-09,
+      "loss": 11.6251,
+      "step": 17096
+    },
+    {
+      "epoch": 2.9899016393442626,
+      "grad_norm": 0.798359142730155,
+      "learning_rate": 5.791096819218744e-09,
+      "loss": 11.9895,
+      "step": 17097
+    },
+    {
+      "epoch": 2.9900765027322405,
+      "grad_norm": 0.8054638863901955,
+      "learning_rate": 5.5896847404035555e-09,
+      "loss": 11.8302,
+      "step": 17098
+    },
+    {
+      "epoch": 2.9902513661202184,
+      "grad_norm": 0.7955839466306442,
+      "learning_rate": 5.391837345491979e-09,
+      "loss": 11.7126,
+      "step": 17099
+    },
+    {
+      "epoch": 2.9904262295081967,
+      "grad_norm": 0.8123473277953999,
+      "learning_rate": 5.197554641511726e-09,
+      "loss": 11.9595,
+      "step": 17100
+    },
+    {
+      "epoch": 2.990601092896175,
+      "grad_norm": 0.8571317590342239,
+      "learning_rate": 5.0068366354238945e-09,
+      "loss": 11.6242,
+      "step": 17101
+    },
+    {
+      "epoch": 2.990775956284153,
+      "grad_norm": 0.825077654980232,
+      "learning_rate": 4.819683334000847e-09,
+      "loss": 12.1147,
+      "step": 17102
+    },
+    {
+      "epoch": 2.9909508196721313,
+      "grad_norm": 0.7772188061156408,
+      "learning_rate": 4.6360947439261225e-09,
+      "loss": 11.7902,
+      "step": 17103
+    },
+    {
+      "epoch": 2.991125683060109,
+      "grad_norm": 0.792997441705122,
+      "learning_rate": 4.45607087175004e-09,
+      "loss": 11.7651,
+      "step": 17104
+    },
+    {
+      "epoch": 2.9913005464480875,
+      "grad_norm": 0.800513640435826,
+      "learning_rate": 4.279611723867483e-09,
+      "loss": 11.9412,
+      "step": 17105
+    },
+    {
+      "epoch": 2.9914754098360654,
+      "grad_norm": 0.8832956730709312,
+      "learning_rate": 4.10671730659562e-09,
+      "loss": 11.8389,
+      "step": 17106
+    },
+    {
+      "epoch": 2.9916502732240438,
+      "grad_norm": 0.8224450599485446,
+      "learning_rate": 3.937387626085087e-09,
+      "loss": 11.6625,
+      "step": 17107
+    },
+    {
+      "epoch": 2.9918251366120217,
+      "grad_norm": 0.8351962866523949,
+      "learning_rate": 3.771622688375498e-09,
+      "loss": 11.8108,
+      "step": 17108
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 0.8030153350421738,
+      "learning_rate": 3.6094224993732383e-09,
+      "loss": 11.7275,
+      "step": 17109
+    },
+    {
+      "epoch": 2.9921748633879783,
+      "grad_norm": 0.8141952501915447,
+      "learning_rate": 3.4507870648625704e-09,
+      "loss": 11.7583,
+      "step": 17110
+    },
+    {
+      "epoch": 2.9923497267759562,
+      "grad_norm": 0.8065205093886566,
+      "learning_rate": 3.29571639049453e-09,
+      "loss": 11.6517,
+      "step": 17111
+    },
+    {
+      "epoch": 2.9925245901639346,
+      "grad_norm": 0.7938319544895345,
+      "learning_rate": 3.1442104817980267e-09,
+      "loss": 11.9257,
+      "step": 17112
+    },
+    {
+      "epoch": 2.9926994535519125,
+      "grad_norm": 0.8859875258374207,
+      "learning_rate": 2.9962693441798474e-09,
+      "loss": 11.8318,
+      "step": 17113
+    },
+    {
+      "epoch": 2.992874316939891,
+      "grad_norm": 0.8707149517460102,
+      "learning_rate": 2.851892982913551e-09,
+      "loss": 11.7834,
+      "step": 17114
+    },
+    {
+      "epoch": 2.9930491803278687,
+      "grad_norm": 0.7875099169971858,
+      "learning_rate": 2.7110814031394703e-09,
+      "loss": 11.8711,
+      "step": 17115
+    },
+    {
+      "epoch": 2.993224043715847,
+      "grad_norm": 0.8967337774396502,
+      "learning_rate": 2.5738346098869158e-09,
+      "loss": 11.7954,
+      "step": 17116
+    },
+    {
+      "epoch": 2.993398907103825,
+      "grad_norm": 0.8354333127723104,
+      "learning_rate": 2.4401526080408686e-09,
+      "loss": 11.6927,
+      "step": 17117
+    },
+    {
+      "epoch": 2.9935737704918033,
+      "grad_norm": 0.8534990044721269,
+      "learning_rate": 2.310035402364186e-09,
+      "loss": 11.6486,
+      "step": 17118
+    },
+    {
+      "epoch": 2.9937486338797816,
+      "grad_norm": 0.8621224463544073,
+      "learning_rate": 2.183482997497599e-09,
+      "loss": 11.6996,
+      "step": 17119
+    },
+    {
+      "epoch": 2.9939234972677595,
+      "grad_norm": 0.8245040090864781,
+      "learning_rate": 2.0604953979597163e-09,
+      "loss": 12.0305,
+      "step": 17120
+    },
+    {
+      "epoch": 2.994098360655738,
+      "grad_norm": 0.9144283892145952,
+      "learning_rate": 1.9410726081359188e-09,
+      "loss": 11.4748,
+      "step": 17121
+    },
+    {
+      "epoch": 2.9942732240437158,
+      "grad_norm": 0.7787232077853423,
+      "learning_rate": 1.8252146322672581e-09,
+      "loss": 11.5691,
+      "step": 17122
+    },
+    {
+      "epoch": 2.994448087431694,
+      "grad_norm": 0.8453458854783795,
+      "learning_rate": 1.7129214745059684e-09,
+      "loss": 12.0581,
+      "step": 17123
+    },
+    {
+      "epoch": 2.994622950819672,
+      "grad_norm": 0.8055004891477201,
+      "learning_rate": 1.6041931388377506e-09,
+      "loss": 11.9916,
+      "step": 17124
+    },
+    {
+      "epoch": 2.9947978142076503,
+      "grad_norm": 0.8275352018411818,
+      "learning_rate": 1.499029629137283e-09,
+      "loss": 11.9332,
+      "step": 17125
+    },
+    {
+      "epoch": 2.9949726775956282,
+      "grad_norm": 0.7981124791025819,
+      "learning_rate": 1.3974309491682214e-09,
+      "loss": 11.2805,
+      "step": 17126
+    },
+    {
+      "epoch": 2.9951475409836066,
+      "grad_norm": 0.8272129292578727,
+      "learning_rate": 1.2993971025387908e-09,
+      "loss": 11.8069,
+      "step": 17127
+    },
+    {
+      "epoch": 2.995322404371585,
+      "grad_norm": 0.8373878327745999,
+      "learning_rate": 1.2049280927572958e-09,
+      "loss": 11.9125,
+      "step": 17128
+    },
+    {
+      "epoch": 2.995497267759563,
+      "grad_norm": 0.8422777033553197,
+      "learning_rate": 1.1140239231877125e-09,
+      "loss": 11.7739,
+      "step": 17129
+    },
+    {
+      "epoch": 2.9956721311475407,
+      "grad_norm": 0.8242999476892006,
+      "learning_rate": 1.0266845970607897e-09,
+      "loss": 11.6898,
+      "step": 17130
+    },
+    {
+      "epoch": 2.995846994535519,
+      "grad_norm": 0.857052443262927,
+      "learning_rate": 9.42910117496254e-10,
+      "loss": 11.4936,
+      "step": 17131
+    },
+    {
+      "epoch": 2.9960218579234974,
+      "grad_norm": 0.8939780672622614,
+      "learning_rate": 8.627004874806055e-10,
+      "loss": 11.6617,
+      "step": 17132
+    },
+    {
+      "epoch": 2.9961967213114753,
+      "grad_norm": 0.8616617758025709,
+      "learning_rate": 7.860557098671173e-10,
+      "loss": 11.7325,
+      "step": 17133
+    },
+    {
+      "epoch": 2.9963715846994536,
+      "grad_norm": 0.9444777844129263,
+      "learning_rate": 7.129757873980402e-10,
+      "loss": 11.9406,
+      "step": 17134
+    },
+    {
+      "epoch": 2.9965464480874315,
+      "grad_norm": 0.8094966934662343,
+      "learning_rate": 6.434607226823985e-10,
+      "loss": 11.7354,
+      "step": 17135
+    },
+    {
+      "epoch": 2.99672131147541,
+      "grad_norm": 0.8329888464645734,
+      "learning_rate": 5.775105181848872e-10,
+      "loss": 11.8654,
+      "step": 17136
+    },
+    {
+      "epoch": 2.996896174863388,
+      "grad_norm": 0.8886689783194492,
+      "learning_rate": 5.151251762591791e-10,
+      "loss": 11.848,
+      "step": 17137
+    },
+    {
+      "epoch": 2.997071038251366,
+      "grad_norm": 0.8309524682457651,
+      "learning_rate": 4.563046991368225e-10,
+      "loss": 11.6543,
+      "step": 17138
+    },
+    {
+      "epoch": 2.997245901639344,
+      "grad_norm": 0.8143905618468589,
+      "learning_rate": 4.010490889050367e-10,
+      "loss": 11.8199,
+      "step": 17139
+    },
+    {
+      "epoch": 2.9974207650273224,
+      "grad_norm": 0.798423615929347,
+      "learning_rate": 3.4935834755112085e-10,
+      "loss": 11.7616,
+      "step": 17140
+    },
+    {
+      "epoch": 2.9975956284153007,
+      "grad_norm": 0.8398474338883574,
+      "learning_rate": 3.012324768958408e-10,
+      "loss": 11.7726,
+      "step": 17141
+    },
+    {
+      "epoch": 2.9977704918032786,
+      "grad_norm": 0.8015733821071303,
+      "learning_rate": 2.5667147866004214e-10,
+      "loss": 11.883,
+      "step": 17142
+    },
+    {
+      "epoch": 2.997945355191257,
+      "grad_norm": 0.7999453537996289,
+      "learning_rate": 2.1567535445354837e-10,
+      "loss": 11.8633,
+      "step": 17143
+    },
+    {
+      "epoch": 2.998120218579235,
+      "grad_norm": 0.8389095295666273,
+      "learning_rate": 1.782441057085471e-10,
+      "loss": 11.8059,
+      "step": 17144
+    },
+    {
+      "epoch": 2.998295081967213,
+      "grad_norm": 0.7969452385588177,
+      "learning_rate": 1.4437773376840824e-10,
+      "loss": 11.5303,
+      "step": 17145
+    },
+    {
+      "epoch": 2.9984699453551915,
+      "grad_norm": 0.8930060321852812,
+      "learning_rate": 1.140762398543771e-10,
+      "loss": 11.8604,
+      "step": 17146
+    },
+    {
+      "epoch": 2.9986448087431694,
+      "grad_norm": 0.8536865186111589,
+      "learning_rate": 8.733962503226778e-11,
+      "loss": 11.5824,
+      "step": 17147
+    },
+    {
+      "epoch": 2.9988196721311473,
+      "grad_norm": 0.8558844193212226,
+      "learning_rate": 6.416789025687209e-11,
+      "loss": 11.7924,
+      "step": 17148
+    },
+    {
+      "epoch": 2.9989945355191256,
+      "grad_norm": 0.7236751164688391,
+      "learning_rate": 4.4561036349755056e-11,
+      "loss": 11.6718,
+      "step": 17149
+    },
+    {
+      "epoch": 2.999169398907104,
+      "grad_norm": 0.792682897022954,
+      "learning_rate": 2.851906403256166e-11,
+      "loss": 11.8893,
+      "step": 17150
+    },
+    {
+      "epoch": 2.999344262295082,
+      "grad_norm": 0.8286926252238437,
+      "learning_rate": 1.6041973849301172e-11,
+      "loss": 11.9826,
+      "step": 17151
+    },
+    {
+      "epoch": 2.9995191256830602,
+      "grad_norm": 0.8368587266143013,
+      "learning_rate": 7.129766266267268e-12,
+      "loss": 11.8145,
+      "step": 17152
+    },
+    {
+      "epoch": 2.999693989071038,
+      "grad_norm": 0.8083503919894249,
+      "learning_rate": 1.7824415832201624e-12,
+      "loss": 11.7724,
+      "step": 17153
+    },
+    {
+      "epoch": 2.9998688524590165,
+      "grad_norm": 0.8460169312876804,
+      "learning_rate": 0.0,
+      "loss": 11.6819,
+      "step": 17154
+    },
+    {
+      "epoch": 2.9998688524590165,
+      "step": 17154,
+      "total_flos": 1407801407143936.0,
+      "train_loss": 0.10603032340616478,
+      "train_runtime": 808.1821,
+      "train_samples_per_second": 2717.172,
+      "train_steps_per_second": 21.225
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17154,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1407801407143936.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}