diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,63102 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.9991117033089054,
+  "eval_steps": 500,
+  "global_step": 9004,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00044414834554741284,
+      "grad_norm": 14.241565838210393,
+      "learning_rate": 1.1098779134295229e-08,
+      "loss": 0.533,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008882966910948257,
+      "grad_norm": 16.278912065775813,
+      "learning_rate": 2.2197558268590458e-08,
+      "loss": 0.5359,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013324450366422385,
+      "grad_norm": 15.38532645021989,
+      "learning_rate": 3.329633740288568e-08,
+      "loss": 0.6437,
+      "step": 3
+    },
+    {
+      "epoch": 0.0017765933821896514,
+      "grad_norm": 13.00040105537638,
+      "learning_rate": 4.4395116537180915e-08,
+      "loss": 0.5256,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022207417277370642,
+      "grad_norm": 13.99913175512498,
+      "learning_rate": 5.549389567147614e-08,
+      "loss": 0.559,
+      "step": 5
+    },
+    {
+      "epoch": 0.002664890073284477,
+      "grad_norm": 14.827666239037775,
+      "learning_rate": 6.659267480577137e-08,
+      "loss": 0.5342,
+      "step": 6
+    },
+    {
+      "epoch": 0.00310903841883189,
+      "grad_norm": 14.56618533431443,
+      "learning_rate": 7.76914539400666e-08,
+      "loss": 0.5818,
+      "step": 7
+    },
+    {
+      "epoch": 0.0035531867643793028,
+      "grad_norm": 14.021029607148948,
+      "learning_rate": 8.879023307436183e-08,
+      "loss": 0.5116,
+      "step": 8
+    },
+    {
+      "epoch": 0.003997335109926716,
+      "grad_norm": 13.230591443503124,
+      "learning_rate": 9.988901220865707e-08,
+      "loss": 0.5254,
+      "step": 9
+    },
+    {
+      "epoch": 0.0044414834554741284,
+      "grad_norm": 14.105717100587825,
+      "learning_rate": 1.1098779134295228e-07,
+      "loss": 0.5469,
+      "step": 10
+    },
+    {
+      "epoch": 0.004885631801021541,
+      "grad_norm": 13.393742170063549,
+      "learning_rate": 1.220865704772475e-07,
+      "loss": 0.5273,
+      "step": 11
+    },
+    {
+      "epoch": 0.005329780146568954,
+      "grad_norm": 13.614380052118197,
+      "learning_rate": 1.3318534961154273e-07,
+      "loss": 0.5127,
+      "step": 12
+    },
+    {
+      "epoch": 0.005773928492116367,
+      "grad_norm": 12.784423466221073,
+      "learning_rate": 1.4428412874583796e-07,
+      "loss": 0.4306,
+      "step": 13
+    },
+    {
+      "epoch": 0.00621807683766378,
+      "grad_norm": 13.80866223671461,
+      "learning_rate": 1.553829078801332e-07,
+      "loss": 0.5439,
+      "step": 14
+    },
+    {
+      "epoch": 0.006662225183211193,
+      "grad_norm": 14.041859557487046,
+      "learning_rate": 1.6648168701442844e-07,
+      "loss": 0.5188,
+      "step": 15
+    },
+    {
+      "epoch": 0.0071063735287586055,
+      "grad_norm": 14.936881953932252,
+      "learning_rate": 1.7758046614872366e-07,
+      "loss": 0.5683,
+      "step": 16
+    },
+    {
+      "epoch": 0.007550521874306018,
+      "grad_norm": 15.650434961515957,
+      "learning_rate": 1.886792452830189e-07,
+      "loss": 0.5353,
+      "step": 17
+    },
+    {
+      "epoch": 0.007994670219853431,
+      "grad_norm": 14.344720416498966,
+      "learning_rate": 1.9977802441731414e-07,
+      "loss": 0.4958,
+      "step": 18
+    },
+    {
+      "epoch": 0.008438818565400843,
+      "grad_norm": 15.791401491309836,
+      "learning_rate": 2.1087680355160934e-07,
+      "loss": 0.5652,
+      "step": 19
+    },
+    {
+      "epoch": 0.008882966910948257,
+      "grad_norm": 13.010603153221817,
+      "learning_rate": 2.2197558268590456e-07,
+      "loss": 0.4343,
+      "step": 20
+    },
+    {
+      "epoch": 0.009327115256495669,
+      "grad_norm": 15.52141534233184,
+      "learning_rate": 2.330743618201998e-07,
+      "loss": 0.5389,
+      "step": 21
+    },
+    {
+      "epoch": 0.009771263602043083,
+      "grad_norm": 15.756642888500245,
+      "learning_rate": 2.44173140954495e-07,
+      "loss": 0.5326,
+      "step": 22
+    },
+    {
+      "epoch": 0.010215411947590495,
+      "grad_norm": 15.268952002568577,
+      "learning_rate": 2.5527192008879024e-07,
+      "loss": 0.4824,
+      "step": 23
+    },
+    {
+      "epoch": 0.010659560293137908,
+      "grad_norm": 15.932949745262599,
+      "learning_rate": 2.6637069922308547e-07,
+      "loss": 0.5079,
+      "step": 24
+    },
+    {
+      "epoch": 0.01110370863868532,
+      "grad_norm": 15.58611205142026,
+      "learning_rate": 2.7746947835738074e-07,
+      "loss": 0.476,
+      "step": 25
+    },
+    {
+      "epoch": 0.011547856984232734,
+      "grad_norm": 15.375627217870338,
+      "learning_rate": 2.885682574916759e-07,
+      "loss": 0.4744,
+      "step": 26
+    },
+    {
+      "epoch": 0.011992005329780146,
+      "grad_norm": 16.173866872463734,
+      "learning_rate": 2.9966703662597114e-07,
+      "loss": 0.5543,
+      "step": 27
+    },
+    {
+      "epoch": 0.01243615367532756,
+      "grad_norm": 11.889556738869189,
+      "learning_rate": 3.107658157602664e-07,
+      "loss": 0.459,
+      "step": 28
+    },
+    {
+      "epoch": 0.012880302020874972,
+      "grad_norm": 9.805467442732954,
+      "learning_rate": 3.218645948945616e-07,
+      "loss": 0.4039,
+      "step": 29
+    },
+    {
+      "epoch": 0.013324450366422385,
+      "grad_norm": 9.109396128992588,
+      "learning_rate": 3.3296337402885687e-07,
+      "loss": 0.374,
+      "step": 30
+    },
+    {
+      "epoch": 0.013768598711969797,
+      "grad_norm": 10.263226704180568,
+      "learning_rate": 3.440621531631521e-07,
+      "loss": 0.5212,
+      "step": 31
+    },
+    {
+      "epoch": 0.014212747057517211,
+      "grad_norm": 8.67404802135948,
+      "learning_rate": 3.551609322974473e-07,
+      "loss": 0.449,
+      "step": 32
+    },
+    {
+      "epoch": 0.014656895403064623,
+      "grad_norm": 9.21127577620048,
+      "learning_rate": 3.6625971143174255e-07,
+      "loss": 0.4436,
+      "step": 33
+    },
+    {
+      "epoch": 0.015101043748612037,
+      "grad_norm": 8.76521126068558,
+      "learning_rate": 3.773584905660378e-07,
+      "loss": 0.4802,
+      "step": 34
+    },
+    {
+      "epoch": 0.015545192094159449,
+      "grad_norm": 8.042704299848994,
+      "learning_rate": 3.8845726970033295e-07,
+      "loss": 0.4144,
+      "step": 35
+    },
+    {
+      "epoch": 0.015989340439706862,
+      "grad_norm": 7.470228322799545,
+      "learning_rate": 3.995560488346283e-07,
+      "loss": 0.3999,
+      "step": 36
+    },
+    {
+      "epoch": 0.016433488785254274,
+      "grad_norm": 8.608115328544926,
+      "learning_rate": 4.1065482796892345e-07,
+      "loss": 0.4636,
+      "step": 37
+    },
+    {
+      "epoch": 0.016877637130801686,
+      "grad_norm": 7.697028697048977,
+      "learning_rate": 4.217536071032187e-07,
+      "loss": 0.3997,
+      "step": 38
+    },
+    {
+      "epoch": 0.017321785476349102,
+      "grad_norm": 5.663561386344377,
+      "learning_rate": 4.328523862375139e-07,
+      "loss": 0.3691,
+      "step": 39
+    },
+    {
+      "epoch": 0.017765933821896514,
+      "grad_norm": 5.938577871272599,
+      "learning_rate": 4.4395116537180913e-07,
+      "loss": 0.3926,
+      "step": 40
+    },
+    {
+      "epoch": 0.018210082167443926,
+      "grad_norm": 6.0845722806591365,
+      "learning_rate": 4.5504994450610435e-07,
+      "loss": 0.4168,
+      "step": 41
+    },
+    {
+      "epoch": 0.018654230512991338,
+      "grad_norm": 4.988124447126365,
+      "learning_rate": 4.661487236403996e-07,
+      "loss": 0.3995,
+      "step": 42
+    },
+    {
+      "epoch": 0.019098378858538753,
+      "grad_norm": 5.157391164042898,
+      "learning_rate": 4.772475027746949e-07,
+      "loss": 0.3846,
+      "step": 43
+    },
+    {
+      "epoch": 0.019542527204086165,
+      "grad_norm": 4.524592140728303,
+      "learning_rate": 4.8834628190899e-07,
+      "loss": 0.3393,
+      "step": 44
+    },
+    {
+      "epoch": 0.019986675549633577,
+      "grad_norm": 4.748116191926127,
+      "learning_rate": 4.994450610432853e-07,
+      "loss": 0.3278,
+      "step": 45
+    },
+    {
+      "epoch": 0.02043082389518099,
+      "grad_norm": 5.1431403204013755,
+      "learning_rate": 5.105438401775805e-07,
+      "loss": 0.3638,
+      "step": 46
+    },
+    {
+      "epoch": 0.020874972240728405,
+      "grad_norm": 5.298031590299299,
+      "learning_rate": 5.216426193118758e-07,
+      "loss": 0.3682,
+      "step": 47
+    },
+    {
+      "epoch": 0.021319120586275817,
+      "grad_norm": 4.678508163207935,
+      "learning_rate": 5.327413984461709e-07,
+      "loss": 0.3772,
+      "step": 48
+    },
+    {
+      "epoch": 0.02176326893182323,
+      "grad_norm": 4.39773491079716,
+      "learning_rate": 5.438401775804662e-07,
+      "loss": 0.3682,
+      "step": 49
+    },
+    {
+      "epoch": 0.02220741727737064,
+      "grad_norm": 4.313057042541789,
+      "learning_rate": 5.549389567147615e-07,
+      "loss": 0.3237,
+      "step": 50
+    },
+    {
+      "epoch": 0.022651565622918056,
+      "grad_norm": 4.449086461913763,
+      "learning_rate": 5.660377358490567e-07,
+      "loss": 0.3623,
+      "step": 51
+    },
+    {
+      "epoch": 0.023095713968465468,
+      "grad_norm": 2.9654000926481783,
+      "learning_rate": 5.771365149833518e-07,
+      "loss": 0.2696,
+      "step": 52
+    },
+    {
+      "epoch": 0.02353986231401288,
+      "grad_norm": 2.9606096375757986,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.2518,
+      "step": 53
+    },
+    {
+      "epoch": 0.023984010659560292,
+      "grad_norm": 3.4611785952391463,
+      "learning_rate": 5.993340732519423e-07,
+      "loss": 0.2782,
+      "step": 54
+    },
+    {
+      "epoch": 0.024428159005107707,
+      "grad_norm": 3.1814895723143284,
+      "learning_rate": 6.104328523862376e-07,
+      "loss": 0.2906,
+      "step": 55
+    },
+    {
+      "epoch": 0.02487230735065512,
+      "grad_norm": 3.1808671717848496,
+      "learning_rate": 6.215316315205328e-07,
+      "loss": 0.2983,
+      "step": 56
+    },
+    {
+      "epoch": 0.02531645569620253,
+      "grad_norm": 2.118090185635995,
+      "learning_rate": 6.32630410654828e-07,
+      "loss": 0.2158,
+      "step": 57
+    },
+    {
+      "epoch": 0.025760604041749943,
+      "grad_norm": 2.697428855954616,
+      "learning_rate": 6.437291897891232e-07,
+      "loss": 0.2815,
+      "step": 58
+    },
+    {
+      "epoch": 0.02620475238729736,
+      "grad_norm": 2.8924667859746926,
+      "learning_rate": 6.548279689234186e-07,
+      "loss": 0.2487,
+      "step": 59
+    },
+    {
+      "epoch": 0.02664890073284477,
+      "grad_norm": 2.346290297240456,
+      "learning_rate": 6.659267480577137e-07,
+      "loss": 0.2496,
+      "step": 60
+    },
+    {
+      "epoch": 0.027093049078392183,
+      "grad_norm": 2.498409492992046,
+      "learning_rate": 6.77025527192009e-07,
+      "loss": 0.2728,
+      "step": 61
+    },
+    {
+      "epoch": 0.027537197423939595,
+      "grad_norm": 3.674582885160151,
+      "learning_rate": 6.881243063263042e-07,
+      "loss": 0.2383,
+      "step": 62
+    },
+    {
+      "epoch": 0.02798134576948701,
+      "grad_norm": 2.799425394175707,
+      "learning_rate": 6.992230854605994e-07,
+      "loss": 0.3199,
+      "step": 63
+    },
+    {
+      "epoch": 0.028425494115034422,
+      "grad_norm": 2.4559947409849805,
+      "learning_rate": 7.103218645948946e-07,
+      "loss": 0.2693,
+      "step": 64
+    },
+    {
+      "epoch": 0.028869642460581834,
+      "grad_norm": 2.293048008449068,
+      "learning_rate": 7.214206437291898e-07,
+      "loss": 0.2519,
+      "step": 65
+    },
+    {
+      "epoch": 0.029313790806129246,
+      "grad_norm": 2.442411284742022,
+      "learning_rate": 7.325194228634851e-07,
+      "loss": 0.2574,
+      "step": 66
+    },
+    {
+      "epoch": 0.02975793915167666,
+      "grad_norm": 1.9858437264830013,
+      "learning_rate": 7.436182019977803e-07,
+      "loss": 0.2059,
+      "step": 67
+    },
+    {
+      "epoch": 0.030202087497224073,
+      "grad_norm": 2.561701423318645,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.2526,
+      "step": 68
+    },
+    {
+      "epoch": 0.030646235842771485,
+      "grad_norm": 1.7160142067734883,
+      "learning_rate": 7.658157602663707e-07,
+      "loss": 0.2103,
+      "step": 69
+    },
+    {
+      "epoch": 0.031090384188318897,
+      "grad_norm": 3.0262045436370473,
+      "learning_rate": 7.769145394006659e-07,
+      "loss": 0.278,
+      "step": 70
+    },
+    {
+      "epoch": 0.03153453253386631,
+      "grad_norm": 2.1467589387452786,
+      "learning_rate": 7.880133185349612e-07,
+      "loss": 0.2033,
+      "step": 71
+    },
+    {
+      "epoch": 0.031978680879413725,
+      "grad_norm": 2.161876762654834,
+      "learning_rate": 7.991120976692566e-07,
+      "loss": 0.2319,
+      "step": 72
+    },
+    {
+      "epoch": 0.03242282922496114,
+      "grad_norm": 2.376008135141836,
+      "learning_rate": 8.102108768035517e-07,
+      "loss": 0.2775,
+      "step": 73
+    },
+    {
+      "epoch": 0.03286697757050855,
+      "grad_norm": 2.12961751744709,
+      "learning_rate": 8.213096559378469e-07,
+      "loss": 0.2066,
+      "step": 74
+    },
+    {
+      "epoch": 0.033311125916055964,
+      "grad_norm": 2.087503322703807,
+      "learning_rate": 8.324084350721422e-07,
+      "loss": 0.2432,
+      "step": 75
+    },
+    {
+      "epoch": 0.03375527426160337,
+      "grad_norm": 2.5677509345382354,
+      "learning_rate": 8.435072142064374e-07,
+      "loss": 0.2027,
+      "step": 76
+    },
+    {
+      "epoch": 0.03419942260715079,
+      "grad_norm": 1.7422813109091475,
+      "learning_rate": 8.546059933407326e-07,
+      "loss": 0.1981,
+      "step": 77
+    },
+    {
+      "epoch": 0.034643570952698204,
+      "grad_norm": 1.9135429042700127,
+      "learning_rate": 8.657047724750278e-07,
+      "loss": 0.1914,
+      "step": 78
+    },
+    {
+      "epoch": 0.03508771929824561,
+      "grad_norm": 2.0631119197996357,
+      "learning_rate": 8.768035516093231e-07,
+      "loss": 0.2247,
+      "step": 79
+    },
+    {
+      "epoch": 0.03553186764379303,
+      "grad_norm": 2.3180619400309452,
+      "learning_rate": 8.879023307436183e-07,
+      "loss": 0.2491,
+      "step": 80
+    },
+    {
+      "epoch": 0.03597601598934044,
+      "grad_norm": 1.9425485940179281,
+      "learning_rate": 8.990011098779134e-07,
+      "loss": 0.2187,
+      "step": 81
+    },
+    {
+      "epoch": 0.03642016433488785,
+      "grad_norm": 1.8295253723979448,
+      "learning_rate": 9.100998890122087e-07,
+      "loss": 0.2012,
+      "step": 82
+    },
+    {
+      "epoch": 0.03686431268043527,
+      "grad_norm": 1.9729843745153348,
+      "learning_rate": 9.211986681465039e-07,
+      "loss": 0.209,
+      "step": 83
+    },
+    {
+      "epoch": 0.037308461025982675,
+      "grad_norm": 2.037422711387131,
+      "learning_rate": 9.322974472807992e-07,
+      "loss": 0.2094,
+      "step": 84
+    },
+    {
+      "epoch": 0.03775260937153009,
+      "grad_norm": 1.577701916798112,
+      "learning_rate": 9.433962264150944e-07,
+      "loss": 0.1858,
+      "step": 85
+    },
+    {
+      "epoch": 0.038196757717077506,
+      "grad_norm": 2.139670658650865,
+      "learning_rate": 9.544950055493897e-07,
+      "loss": 0.2254,
+      "step": 86
+    },
+    {
+      "epoch": 0.038640906062624915,
+      "grad_norm": 2.8502450993225956,
+      "learning_rate": 9.65593784683685e-07,
+      "loss": 0.2145,
+      "step": 87
+    },
+    {
+      "epoch": 0.03908505440817233,
+      "grad_norm": 1.7436844528376316,
+      "learning_rate": 9.7669256381798e-07,
+      "loss": 0.1927,
+      "step": 88
+    },
+    {
+      "epoch": 0.039529202753719746,
+      "grad_norm": 2.2129521928377796,
+      "learning_rate": 9.877913429522753e-07,
+      "loss": 0.2449,
+      "step": 89
+    },
+    {
+      "epoch": 0.039973351099267154,
+      "grad_norm": 1.7063140977463,
+      "learning_rate": 9.988901220865706e-07,
+      "loss": 0.2002,
+      "step": 90
+    },
+    {
+      "epoch": 0.04041749944481457,
+      "grad_norm": 1.7274557393115937,
+      "learning_rate": 1.009988901220866e-06,
+      "loss": 0.2143,
+      "step": 91
+    },
+    {
+      "epoch": 0.04086164779036198,
+      "grad_norm": 1.6036310126204065,
+      "learning_rate": 1.021087680355161e-06,
+      "loss": 0.209,
+      "step": 92
+    },
+    {
+      "epoch": 0.041305796135909394,
+      "grad_norm": 1.6067518053520686,
+      "learning_rate": 1.0321864594894562e-06,
+      "loss": 0.2051,
+      "step": 93
+    },
+    {
+      "epoch": 0.04174994448145681,
+      "grad_norm": 2.193990373798874,
+      "learning_rate": 1.0432852386237515e-06,
+      "loss": 0.2124,
+      "step": 94
+    },
+    {
+      "epoch": 0.04219409282700422,
+      "grad_norm": 2.0886683470736536,
+      "learning_rate": 1.0543840177580466e-06,
+      "loss": 0.2283,
+      "step": 95
+    },
+    {
+      "epoch": 0.04263824117255163,
+      "grad_norm": 2.111784187748928,
+      "learning_rate": 1.0654827968923419e-06,
+      "loss": 0.2035,
+      "step": 96
+    },
+    {
+      "epoch": 0.04308238951809905,
+      "grad_norm": 2.082762904011713,
+      "learning_rate": 1.0765815760266371e-06,
+      "loss": 0.1979,
+      "step": 97
+    },
+    {
+      "epoch": 0.04352653786364646,
+      "grad_norm": 1.9566752129472822,
+      "learning_rate": 1.0876803551609324e-06,
+      "loss": 0.2146,
+      "step": 98
+    },
+    {
+      "epoch": 0.04397068620919387,
+      "grad_norm": 1.553353783765945,
+      "learning_rate": 1.0987791342952277e-06,
+      "loss": 0.1682,
+      "step": 99
+    },
+    {
+      "epoch": 0.04441483455474128,
+      "grad_norm": 1.9387075500236326,
+      "learning_rate": 1.109877913429523e-06,
+      "loss": 0.1792,
+      "step": 100
+    },
+    {
+      "epoch": 0.044858982900288696,
+      "grad_norm": 1.980937561976891,
+      "learning_rate": 1.120976692563818e-06,
+      "loss": 0.1896,
+      "step": 101
+    },
+    {
+      "epoch": 0.04530313124583611,
+      "grad_norm": 2.4054051662002824,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.2097,
+      "step": 102
+    },
+    {
+      "epoch": 0.04574727959138352,
+      "grad_norm": 1.8504294534254195,
+      "learning_rate": 1.1431742508324086e-06,
+      "loss": 0.1795,
+      "step": 103
+    },
+    {
+      "epoch": 0.046191427936930936,
+      "grad_norm": 1.8842730303029716,
+      "learning_rate": 1.1542730299667037e-06,
+      "loss": 0.1883,
+      "step": 104
+    },
+    {
+      "epoch": 0.046635576282478344,
+      "grad_norm": 1.5610944241824418,
+      "learning_rate": 1.165371809100999e-06,
+      "loss": 0.1655,
+      "step": 105
+    },
+    {
+      "epoch": 0.04707972462802576,
+      "grad_norm": 1.5819842411294622,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.1811,
+      "step": 106
+    },
+    {
+      "epoch": 0.047523872973573175,
+      "grad_norm": 1.9979272057896553,
+      "learning_rate": 1.1875693673695895e-06,
+      "loss": 0.1823,
+      "step": 107
+    },
+    {
+      "epoch": 0.047968021319120584,
+      "grad_norm": 1.9458215900555198,
+      "learning_rate": 1.1986681465038846e-06,
+      "loss": 0.2136,
+      "step": 108
+    },
+    {
+      "epoch": 0.048412169664668,
+      "grad_norm": 1.8191486796264364,
+      "learning_rate": 1.2097669256381799e-06,
+      "loss": 0.2082,
+      "step": 109
+    },
+    {
+      "epoch": 0.048856318010215415,
+      "grad_norm": 1.9290831624594054,
+      "learning_rate": 1.2208657047724751e-06,
+      "loss": 0.1731,
+      "step": 110
+    },
+    {
+      "epoch": 0.04930046635576282,
+      "grad_norm": 2.1157513592152233,
+      "learning_rate": 1.2319644839067704e-06,
+      "loss": 0.1883,
+      "step": 111
+    },
+    {
+      "epoch": 0.04974461470131024,
+      "grad_norm": 2.0021290506932026,
+      "learning_rate": 1.2430632630410657e-06,
+      "loss": 0.1906,
+      "step": 112
+    },
+    {
+      "epoch": 0.05018876304685765,
+      "grad_norm": 1.9159228327922788,
+      "learning_rate": 1.254162042175361e-06,
+      "loss": 0.2176,
+      "step": 113
+    },
+    {
+      "epoch": 0.05063291139240506,
+      "grad_norm": 1.7650876352312228,
+      "learning_rate": 1.265260821309656e-06,
+      "loss": 0.1718,
+      "step": 114
+    },
+    {
+      "epoch": 0.05107705973795248,
+      "grad_norm": 2.03898072034729,
+      "learning_rate": 1.2763596004439513e-06,
+      "loss": 0.1898,
+      "step": 115
+    },
+    {
+      "epoch": 0.051521208083499886,
+      "grad_norm": 1.6693420710972704,
+      "learning_rate": 1.2874583795782464e-06,
+      "loss": 0.1759,
+      "step": 116
+    },
+    {
+      "epoch": 0.0519653564290473,
+      "grad_norm": 2.335357803271554,
+      "learning_rate": 1.2985571587125417e-06,
+      "loss": 0.1837,
+      "step": 117
+    },
+    {
+      "epoch": 0.05240950477459472,
+      "grad_norm": 1.682419177219224,
+      "learning_rate": 1.3096559378468371e-06,
+      "loss": 0.1808,
+      "step": 118
+    },
+    {
+      "epoch": 0.052853653120142126,
+      "grad_norm": 2.0519984686691126,
+      "learning_rate": 1.3207547169811322e-06,
+      "loss": 0.1982,
+      "step": 119
+    },
+    {
+      "epoch": 0.05329780146568954,
+      "grad_norm": 1.6368796114014708,
+      "learning_rate": 1.3318534961154275e-06,
+      "loss": 0.1758,
+      "step": 120
+    },
+    {
+      "epoch": 0.05374194981123695,
+      "grad_norm": 1.6286801352711888,
+      "learning_rate": 1.3429522752497226e-06,
+      "loss": 0.1936,
+      "step": 121
+    },
+    {
+      "epoch": 0.054186098156784365,
+      "grad_norm": 1.8633084178414114,
+      "learning_rate": 1.354051054384018e-06,
+      "loss": 0.202,
+      "step": 122
+    },
+    {
+      "epoch": 0.05463024650233178,
+      "grad_norm": 1.9496618888186672,
+      "learning_rate": 1.3651498335183131e-06,
+      "loss": 0.2025,
+      "step": 123
+    },
+    {
+      "epoch": 0.05507439484787919,
+      "grad_norm": 2.0859989784182353,
+      "learning_rate": 1.3762486126526084e-06,
+      "loss": 0.1903,
+      "step": 124
+    },
+    {
+      "epoch": 0.055518543193426605,
+      "grad_norm": 1.993272387486733,
+      "learning_rate": 1.3873473917869035e-06,
+      "loss": 0.2123,
+      "step": 125
+    },
+    {
+      "epoch": 0.05596269153897402,
+      "grad_norm": 2.2645503131575735,
+      "learning_rate": 1.3984461709211987e-06,
+      "loss": 0.2069,
+      "step": 126
+    },
+    {
+      "epoch": 0.05640683988452143,
+      "grad_norm": 1.581545113462678,
+      "learning_rate": 1.409544950055494e-06,
+      "loss": 0.1804,
+      "step": 127
+    },
+    {
+      "epoch": 0.056850988230068844,
+      "grad_norm": 1.9654703741649047,
+      "learning_rate": 1.4206437291897893e-06,
+      "loss": 0.1625,
+      "step": 128
+    },
+    {
+      "epoch": 0.05729513657561625,
+      "grad_norm": 2.452801004200468,
+      "learning_rate": 1.4317425083240844e-06,
+      "loss": 0.1779,
+      "step": 129
+    },
+    {
+      "epoch": 0.05773928492116367,
+      "grad_norm": 1.847246271423822,
+      "learning_rate": 1.4428412874583796e-06,
+      "loss": 0.1749,
+      "step": 130
+    },
+    {
+      "epoch": 0.05818343326671108,
+      "grad_norm": 1.9357616103684274,
+      "learning_rate": 1.4539400665926751e-06,
+      "loss": 0.2213,
+      "step": 131
+    },
+    {
+      "epoch": 0.05862758161225849,
+      "grad_norm": 1.495082490211886,
+      "learning_rate": 1.4650388457269702e-06,
+      "loss": 0.1579,
+      "step": 132
+    },
+    {
+      "epoch": 0.05907172995780591,
+      "grad_norm": 1.5859431774202346,
+      "learning_rate": 1.4761376248612655e-06,
+      "loss": 0.1606,
+      "step": 133
+    },
+    {
+      "epoch": 0.05951587830335332,
+      "grad_norm": 2.8376409272544687,
+      "learning_rate": 1.4872364039955605e-06,
+      "loss": 0.2481,
+      "step": 134
+    },
+    {
+      "epoch": 0.05996002664890073,
+      "grad_norm": 1.7612190558454792,
+      "learning_rate": 1.498335183129856e-06,
+      "loss": 0.1637,
+      "step": 135
+    },
+    {
+      "epoch": 0.06040417499444815,
+      "grad_norm": 2.070041455462641,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 0.2122,
+      "step": 136
+    },
+    {
+      "epoch": 0.060848323339995555,
+      "grad_norm": 2.3780145580102627,
+      "learning_rate": 1.5205327413984464e-06,
+      "loss": 0.186,
+      "step": 137
+    },
+    {
+      "epoch": 0.06129247168554297,
+      "grad_norm": 1.8658905992142587,
+      "learning_rate": 1.5316315205327414e-06,
+      "loss": 0.1673,
+      "step": 138
+    },
+    {
+      "epoch": 0.061736620031090386,
+      "grad_norm": 2.8172930803107077,
+      "learning_rate": 1.5427302996670367e-06,
+      "loss": 0.244,
+      "step": 139
+    },
+    {
+      "epoch": 0.062180768376637795,
+      "grad_norm": 1.6725142692022914,
+      "learning_rate": 1.5538290788013318e-06,
+      "loss": 0.157,
+      "step": 140
+    },
+    {
+      "epoch": 0.0626249167221852,
+      "grad_norm": 1.392484954569343,
+      "learning_rate": 1.5649278579356273e-06,
+      "loss": 0.1558,
+      "step": 141
+    },
+    {
+      "epoch": 0.06306906506773262,
+      "grad_norm": 1.4158061250880813,
+      "learning_rate": 1.5760266370699223e-06,
+      "loss": 0.148,
+      "step": 142
+    },
+    {
+      "epoch": 0.06351321341328003,
+      "grad_norm": 1.9695160587869651,
+      "learning_rate": 1.5871254162042176e-06,
+      "loss": 0.1468,
+      "step": 143
+    },
+    {
+      "epoch": 0.06395736175882745,
+      "grad_norm": 1.6387389663448713,
+      "learning_rate": 1.5982241953385131e-06,
+      "loss": 0.1363,
+      "step": 144
+    },
+    {
+      "epoch": 0.06440151010437487,
+      "grad_norm": 1.977141381047165,
+      "learning_rate": 1.6093229744728082e-06,
+      "loss": 0.1775,
+      "step": 145
+    },
+    {
+      "epoch": 0.06484565844992228,
+      "grad_norm": 1.866914454872281,
+      "learning_rate": 1.6204217536071035e-06,
+      "loss": 0.1676,
+      "step": 146
+    },
+    {
+      "epoch": 0.06528980679546968,
+      "grad_norm": 1.8479866982339428,
+      "learning_rate": 1.6315205327413985e-06,
+      "loss": 0.1795,
+      "step": 147
+    },
+    {
+      "epoch": 0.0657339551410171,
+      "grad_norm": 2.447758458580566,
+      "learning_rate": 1.6426193118756938e-06,
+      "loss": 0.1921,
+      "step": 148
+    },
+    {
+      "epoch": 0.06617810348656451,
+      "grad_norm": 1.928627100482751,
+      "learning_rate": 1.653718091009989e-06,
+      "loss": 0.1236,
+      "step": 149
+    },
+    {
+      "epoch": 0.06662225183211193,
+      "grad_norm": 2.496031778783847,
+      "learning_rate": 1.6648168701442844e-06,
+      "loss": 0.1263,
+      "step": 150
+    },
+    {
+      "epoch": 0.06706640017765934,
+      "grad_norm": 2.543270527791031,
+      "learning_rate": 1.6759156492785794e-06,
+      "loss": 0.1798,
+      "step": 151
+    },
+    {
+      "epoch": 0.06751054852320675,
+      "grad_norm": 2.6478561509220637,
+      "learning_rate": 1.6870144284128747e-06,
+      "loss": 0.1976,
+      "step": 152
+    },
+    {
+      "epoch": 0.06795469686875416,
+      "grad_norm": 2.046113685099785,
+      "learning_rate": 1.6981132075471698e-06,
+      "loss": 0.1412,
+      "step": 153
+    },
+    {
+      "epoch": 0.06839884521430158,
+      "grad_norm": 4.398993310024598,
+      "learning_rate": 1.7092119866814653e-06,
+      "loss": 0.2004,
+      "step": 154
+    },
+    {
+      "epoch": 0.06884299355984899,
+      "grad_norm": 1.7977539055103122,
+      "learning_rate": 1.7203107658157603e-06,
+      "loss": 0.1803,
+      "step": 155
+    },
+    {
+      "epoch": 0.06928714190539641,
+      "grad_norm": 1.9000289734288398,
+      "learning_rate": 1.7314095449500556e-06,
+      "loss": 0.1725,
+      "step": 156
+    },
+    {
+      "epoch": 0.06973129025094381,
+      "grad_norm": 2.8300001805659956,
+      "learning_rate": 1.742508324084351e-06,
+      "loss": 0.1396,
+      "step": 157
+    },
+    {
+      "epoch": 0.07017543859649122,
+      "grad_norm": 1.6221647641516457,
+      "learning_rate": 1.7536071032186462e-06,
+      "loss": 0.1377,
+      "step": 158
+    },
+    {
+      "epoch": 0.07061958694203864,
+      "grad_norm": 1.8070988364351561,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.1466,
+      "step": 159
+    },
+    {
+      "epoch": 0.07106373528758606,
+      "grad_norm": 1.522016454575872,
+      "learning_rate": 1.7758046614872365e-06,
+      "loss": 0.1508,
+      "step": 160
+    },
+    {
+      "epoch": 0.07150788363313347,
+      "grad_norm": 2.5966599384816393,
+      "learning_rate": 1.7869034406215318e-06,
+      "loss": 0.1946,
+      "step": 161
+    },
+    {
+      "epoch": 0.07195203197868089,
+      "grad_norm": 2.055913601931369,
+      "learning_rate": 1.7980022197558269e-06,
+      "loss": 0.1485,
+      "step": 162
+    },
+    {
+      "epoch": 0.07239618032422829,
+      "grad_norm": 1.716589810835508,
+      "learning_rate": 1.8091009988901223e-06,
+      "loss": 0.1552,
+      "step": 163
+    },
+    {
+      "epoch": 0.0728403286697757,
+      "grad_norm": 1.324188540188503,
+      "learning_rate": 1.8201997780244174e-06,
+      "loss": 0.1217,
+      "step": 164
+    },
+    {
+      "epoch": 0.07328447701532312,
+      "grad_norm": 1.7875490213167293,
+      "learning_rate": 1.8312985571587127e-06,
+      "loss": 0.1814,
+      "step": 165
+    },
+    {
+      "epoch": 0.07372862536087053,
+      "grad_norm": 2.111076485427071,
+      "learning_rate": 1.8423973362930078e-06,
+      "loss": 0.1858,
+      "step": 166
+    },
+    {
+      "epoch": 0.07417277370641795,
+      "grad_norm": 1.8246301721111953,
+      "learning_rate": 1.8534961154273032e-06,
+      "loss": 0.1328,
+      "step": 167
+    },
+    {
+      "epoch": 0.07461692205196535,
+      "grad_norm": 7.017994935548065,
+      "learning_rate": 1.8645948945615983e-06,
+      "loss": 0.2175,
+      "step": 168
+    },
+    {
+      "epoch": 0.07506107039751277,
+      "grad_norm": 1.8833148111605775,
+      "learning_rate": 1.8756936736958936e-06,
+      "loss": 0.169,
+      "step": 169
+    },
+    {
+      "epoch": 0.07550521874306018,
+      "grad_norm": 2.1063106621517833,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.1723,
+      "step": 170
+    },
+    {
+      "epoch": 0.0759493670886076,
+      "grad_norm": 2.1080508911518603,
+      "learning_rate": 1.8978912319644842e-06,
+      "loss": 0.1313,
+      "step": 171
+    },
+    {
+      "epoch": 0.07639351543415501,
+      "grad_norm": 2.042694432552121,
+      "learning_rate": 1.9089900110987794e-06,
+      "loss": 0.1561,
+      "step": 172
+    },
+    {
+      "epoch": 0.07683766377970241,
+      "grad_norm": 2.275654963885769,
+      "learning_rate": 1.9200887902330745e-06,
+      "loss": 0.2056,
+      "step": 173
+    },
+    {
+      "epoch": 0.07728181212524983,
+      "grad_norm": 1.4521350560986936,
+      "learning_rate": 1.93118756936737e-06,
+      "loss": 0.1539,
+      "step": 174
+    },
+    {
+      "epoch": 0.07772596047079725,
+      "grad_norm": 1.4721774252433903,
+      "learning_rate": 1.942286348501665e-06,
+      "loss": 0.1392,
+      "step": 175
+    },
+    {
+      "epoch": 0.07817010881634466,
+      "grad_norm": 2.236429259701826,
+      "learning_rate": 1.95338512763596e-06,
+      "loss": 0.1851,
+      "step": 176
+    },
+    {
+      "epoch": 0.07861425716189208,
+      "grad_norm": 2.0509681089981973,
+      "learning_rate": 1.964483906770255e-06,
+      "loss": 0.1613,
+      "step": 177
+    },
+    {
+      "epoch": 0.07905840550743949,
+      "grad_norm": 1.921297270021361,
+      "learning_rate": 1.9755826859045507e-06,
+      "loss": 0.1411,
+      "step": 178
+    },
+    {
+      "epoch": 0.07950255385298689,
+      "grad_norm": 2.3334407767674534,
+      "learning_rate": 1.9866814650388457e-06,
+      "loss": 0.1673,
+      "step": 179
+    },
+    {
+      "epoch": 0.07994670219853431,
+      "grad_norm": 2.4173674731585035,
+      "learning_rate": 1.9977802441731412e-06,
+      "loss": 0.1769,
+      "step": 180
+    },
+    {
+      "epoch": 0.08039085054408172,
+      "grad_norm": 2.672814520854723,
+      "learning_rate": 2.0088790233074363e-06,
+      "loss": 0.1882,
+      "step": 181
+    },
+    {
+      "epoch": 0.08083499888962914,
+      "grad_norm": 2.082354232015489,
+      "learning_rate": 2.019977802441732e-06,
+      "loss": 0.1594,
+      "step": 182
+    },
+    {
+      "epoch": 0.08127914723517655,
+      "grad_norm": 2.0548698535266965,
+      "learning_rate": 2.031076581576027e-06,
+      "loss": 0.162,
+      "step": 183
+    },
+    {
+      "epoch": 0.08172329558072396,
+      "grad_norm": 2.352430010700108,
+      "learning_rate": 2.042175360710322e-06,
+      "loss": 0.1429,
+      "step": 184
+    },
+    {
+      "epoch": 0.08216744392627137,
+      "grad_norm": 1.8421875981209739,
+      "learning_rate": 2.0532741398446174e-06,
+      "loss": 0.1575,
+      "step": 185
+    },
+    {
+      "epoch": 0.08261159227181879,
+      "grad_norm": 1.8037755953700159,
+      "learning_rate": 2.0643729189789125e-06,
+      "loss": 0.1474,
+      "step": 186
+    },
+    {
+      "epoch": 0.0830557406173662,
+      "grad_norm": 2.2301663182454887,
+      "learning_rate": 2.075471698113208e-06,
+      "loss": 0.1381,
+      "step": 187
+    },
+    {
+      "epoch": 0.08349988896291362,
+      "grad_norm": 1.8820242106298288,
+      "learning_rate": 2.086570477247503e-06,
+      "loss": 0.136,
+      "step": 188
+    },
+    {
+      "epoch": 0.08394403730846102,
+      "grad_norm": 2.0480577593839935,
+      "learning_rate": 2.097669256381798e-06,
+      "loss": 0.146,
+      "step": 189
+    },
+    {
+      "epoch": 0.08438818565400844,
+      "grad_norm": 1.844554763649842,
+      "learning_rate": 2.108768035516093e-06,
+      "loss": 0.1663,
+      "step": 190
+    },
+    {
+      "epoch": 0.08483233399955585,
+      "grad_norm": 1.5207370248204364,
+      "learning_rate": 2.1198668146503887e-06,
+      "loss": 0.1453,
+      "step": 191
+    },
+    {
+      "epoch": 0.08527648234510327,
+      "grad_norm": 1.7288600299730912,
+      "learning_rate": 2.1309655937846837e-06,
+      "loss": 0.1492,
+      "step": 192
+    },
+    {
+      "epoch": 0.08572063069065068,
+      "grad_norm": 1.6054586875361534,
+      "learning_rate": 2.1420643729189792e-06,
+      "loss": 0.1702,
+      "step": 193
+    },
+    {
+      "epoch": 0.0861647790361981,
+      "grad_norm": 2.082434204878279,
+      "learning_rate": 2.1531631520532743e-06,
+      "loss": 0.1417,
+      "step": 194
+    },
+    {
+      "epoch": 0.0866089273817455,
+      "grad_norm": 1.893129069527839,
+      "learning_rate": 2.1642619311875694e-06,
+      "loss": 0.1746,
+      "step": 195
+    },
+    {
+      "epoch": 0.08705307572729291,
+      "grad_norm": 1.5433606197712828,
+      "learning_rate": 2.175360710321865e-06,
+      "loss": 0.1434,
+      "step": 196
+    },
+    {
+      "epoch": 0.08749722407284033,
+      "grad_norm": 1.8343180586393633,
+      "learning_rate": 2.18645948945616e-06,
+      "loss": 0.1635,
+      "step": 197
+    },
+    {
+      "epoch": 0.08794137241838774,
+      "grad_norm": 1.9644492187331304,
+      "learning_rate": 2.1975582685904554e-06,
+      "loss": 0.1613,
+      "step": 198
+    },
+    {
+      "epoch": 0.08838552076393516,
+      "grad_norm": 2.3183987673393784,
+      "learning_rate": 2.2086570477247505e-06,
+      "loss": 0.1451,
+      "step": 199
+    },
+    {
+      "epoch": 0.08882966910948256,
+      "grad_norm": 1.8970811571077912,
+      "learning_rate": 2.219755826859046e-06,
+      "loss": 0.1693,
+      "step": 200
+    },
+    {
+      "epoch": 0.08927381745502998,
+      "grad_norm": 1.3438831493644725,
+      "learning_rate": 2.230854605993341e-06,
+      "loss": 0.1442,
+      "step": 201
+    },
+    {
+      "epoch": 0.08971796580057739,
+      "grad_norm": 2.0647130607040687,
+      "learning_rate": 2.241953385127636e-06,
+      "loss": 0.1483,
+      "step": 202
+    },
+    {
+      "epoch": 0.09016211414612481,
+      "grad_norm": 1.8137712195107252,
+      "learning_rate": 2.253052164261931e-06,
+      "loss": 0.1898,
+      "step": 203
+    },
+    {
+      "epoch": 0.09060626249167222,
+      "grad_norm": 1.5677896114201355,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 0.1105,
+      "step": 204
+    },
+    {
+      "epoch": 0.09105041083721963,
+      "grad_norm": 1.5774232800494377,
+      "learning_rate": 2.2752497225305217e-06,
+      "loss": 0.1439,
+      "step": 205
+    },
+    {
+      "epoch": 0.09149455918276704,
+      "grad_norm": 2.191208018434963,
+      "learning_rate": 2.286348501664817e-06,
+      "loss": 0.1739,
+      "step": 206
+    },
+    {
+      "epoch": 0.09193870752831446,
+      "grad_norm": 1.8759698513356193,
+      "learning_rate": 2.2974472807991123e-06,
+      "loss": 0.1238,
+      "step": 207
+    },
+    {
+      "epoch": 0.09238285587386187,
+      "grad_norm": 2.193936812351504,
+      "learning_rate": 2.3085460599334073e-06,
+      "loss": 0.149,
+      "step": 208
+    },
+    {
+      "epoch": 0.09282700421940929,
+      "grad_norm": 1.7624384709819947,
+      "learning_rate": 2.319644839067703e-06,
+      "loss": 0.1395,
+      "step": 209
+    },
+    {
+      "epoch": 0.09327115256495669,
+      "grad_norm": 1.940248444895634,
+      "learning_rate": 2.330743618201998e-06,
+      "loss": 0.1321,
+      "step": 210
+    },
+    {
+      "epoch": 0.0937153009105041,
+      "grad_norm": 1.623293710115476,
+      "learning_rate": 2.3418423973362934e-06,
+      "loss": 0.1411,
+      "step": 211
+    },
+    {
+      "epoch": 0.09415944925605152,
+      "grad_norm": 2.311237234944747,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.1451,
+      "step": 212
+    },
+    {
+      "epoch": 0.09460359760159893,
+      "grad_norm": 2.0964379988774366,
+      "learning_rate": 2.364039955604884e-06,
+      "loss": 0.1628,
+      "step": 213
+    },
+    {
+      "epoch": 0.09504774594714635,
+      "grad_norm": 1.57604767113886,
+      "learning_rate": 2.375138734739179e-06,
+      "loss": 0.1282,
+      "step": 214
+    },
+    {
+      "epoch": 0.09549189429269377,
+      "grad_norm": 1.446630991649051,
+      "learning_rate": 2.386237513873474e-06,
+      "loss": 0.1392,
+      "step": 215
+    },
+    {
+      "epoch": 0.09593604263824117,
+      "grad_norm": 2.274345473449785,
+      "learning_rate": 2.397336293007769e-06,
+      "loss": 0.1962,
+      "step": 216
+    },
+    {
+      "epoch": 0.09638019098378858,
+      "grad_norm": 1.5946628028114809,
+      "learning_rate": 2.4084350721420646e-06,
+      "loss": 0.138,
+      "step": 217
+    },
+    {
+      "epoch": 0.096824339329336,
+      "grad_norm": 1.882389417953473,
+      "learning_rate": 2.4195338512763597e-06,
+      "loss": 0.1903,
+      "step": 218
+    },
+    {
+      "epoch": 0.09726848767488341,
+      "grad_norm": 1.904432577295931,
+      "learning_rate": 2.430632630410655e-06,
+      "loss": 0.1481,
+      "step": 219
+    },
+    {
+      "epoch": 0.09771263602043083,
+      "grad_norm": 2.080561584835578,
+      "learning_rate": 2.4417314095449503e-06,
+      "loss": 0.1514,
+      "step": 220
+    },
+    {
+      "epoch": 0.09815678436597823,
+      "grad_norm": 1.6677943098084358,
+      "learning_rate": 2.4528301886792453e-06,
+      "loss": 0.1347,
+      "step": 221
+    },
+    {
+      "epoch": 0.09860093271152565,
+      "grad_norm": 1.7294352905463155,
+      "learning_rate": 2.463928967813541e-06,
+      "loss": 0.1386,
+      "step": 222
+    },
+    {
+      "epoch": 0.09904508105707306,
+      "grad_norm": 2.0928510956520934,
+      "learning_rate": 2.475027746947836e-06,
+      "loss": 0.1611,
+      "step": 223
+    },
+    {
+      "epoch": 0.09948922940262048,
+      "grad_norm": 1.6018044625885859,
+      "learning_rate": 2.4861265260821314e-06,
+      "loss": 0.1338,
+      "step": 224
+    },
+    {
+      "epoch": 0.09993337774816789,
+      "grad_norm": 1.9068721122681065,
+      "learning_rate": 2.4972253052164264e-06,
+      "loss": 0.1557,
+      "step": 225
+    },
+    {
+      "epoch": 0.1003775260937153,
+      "grad_norm": 1.5854054800345856,
+      "learning_rate": 2.508324084350722e-06,
+      "loss": 0.1582,
+      "step": 226
+    },
+    {
+      "epoch": 0.10082167443926271,
+      "grad_norm": 1.6327772348060883,
+      "learning_rate": 2.519422863485017e-06,
+      "loss": 0.153,
+      "step": 227
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.7221707573000986,
+      "learning_rate": 2.530521642619312e-06,
+      "loss": 0.1521,
+      "step": 228
+    },
+    {
+      "epoch": 0.10170997113035754,
+      "grad_norm": 3.8860493727438605,
+      "learning_rate": 2.541620421753607e-06,
+      "loss": 0.185,
+      "step": 229
+    },
+    {
+      "epoch": 0.10215411947590496,
+      "grad_norm": 2.2982967806121057,
+      "learning_rate": 2.5527192008879026e-06,
+      "loss": 0.1268,
+      "step": 230
+    },
+    {
+      "epoch": 0.10259826782145237,
+      "grad_norm": 2.633523229110552,
+      "learning_rate": 2.563817980022198e-06,
+      "loss": 0.1312,
+      "step": 231
+    },
+    {
+      "epoch": 0.10304241616699977,
+      "grad_norm": 1.8405348072939953,
+      "learning_rate": 2.5749167591564928e-06,
+      "loss": 0.1352,
+      "step": 232
+    },
+    {
+      "epoch": 0.10348656451254719,
+      "grad_norm": 1.5009853192812423,
+      "learning_rate": 2.5860155382907882e-06,
+      "loss": 0.1191,
+      "step": 233
+    },
+    {
+      "epoch": 0.1039307128580946,
+      "grad_norm": 1.3280238597160159,
+      "learning_rate": 2.5971143174250833e-06,
+      "loss": 0.1207,
+      "step": 234
+    },
+    {
+      "epoch": 0.10437486120364202,
+      "grad_norm": 1.412771085327836,
+      "learning_rate": 2.608213096559379e-06,
+      "loss": 0.1343,
+      "step": 235
+    },
+    {
+      "epoch": 0.10481900954918943,
+      "grad_norm": 1.4216505684340854,
+      "learning_rate": 2.6193118756936743e-06,
+      "loss": 0.1312,
+      "step": 236
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.5128985222362534,
+      "learning_rate": 2.630410654827969e-06,
+      "loss": 0.1351,
+      "step": 237
+    },
+    {
+      "epoch": 0.10570730624028425,
+      "grad_norm": 1.9599268561293408,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.1403,
+      "step": 238
+    },
+    {
+      "epoch": 0.10615145458583167,
+      "grad_norm": 1.4181236006554954,
+      "learning_rate": 2.6526082130965595e-06,
+      "loss": 0.1377,
+      "step": 239
+    },
+    {
+      "epoch": 0.10659560293137908,
+      "grad_norm": 1.9917215037873872,
+      "learning_rate": 2.663706992230855e-06,
+      "loss": 0.1451,
+      "step": 240
+    },
+    {
+      "epoch": 0.1070397512769265,
+      "grad_norm": 1.683475509658225,
+      "learning_rate": 2.67480577136515e-06,
+      "loss": 0.1375,
+      "step": 241
+    },
+    {
+      "epoch": 0.1074838996224739,
+      "grad_norm": 1.8301967264376793,
+      "learning_rate": 2.685904550499445e-06,
+      "loss": 0.1464,
+      "step": 242
+    },
+    {
+      "epoch": 0.10792804796802132,
+      "grad_norm": 1.6846261045693358,
+      "learning_rate": 2.6970033296337406e-06,
+      "loss": 0.1332,
+      "step": 243
+    },
+    {
+      "epoch": 0.10837219631356873,
+      "grad_norm": 1.2774719377840502,
+      "learning_rate": 2.708102108768036e-06,
+      "loss": 0.1292,
+      "step": 244
+    },
+    {
+      "epoch": 0.10881634465911615,
+      "grad_norm": 1.595633542836346,
+      "learning_rate": 2.7192008879023307e-06,
+      "loss": 0.1386,
+      "step": 245
+    },
+    {
+      "epoch": 0.10926049300466356,
+      "grad_norm": 1.4528662121606482,
+      "learning_rate": 2.7302996670366262e-06,
+      "loss": 0.1392,
+      "step": 246
+    },
+    {
+      "epoch": 0.10970464135021098,
+      "grad_norm": 1.5309789173048087,
+      "learning_rate": 2.7413984461709213e-06,
+      "loss": 0.1271,
+      "step": 247
+    },
+    {
+      "epoch": 0.11014878969575838,
+      "grad_norm": 1.518105261331223,
+      "learning_rate": 2.7524972253052168e-06,
+      "loss": 0.1237,
+      "step": 248
+    },
+    {
+      "epoch": 0.1105929380413058,
+      "grad_norm": 1.4676341827820025,
+      "learning_rate": 2.7635960044395123e-06,
+      "loss": 0.1317,
+      "step": 249
+    },
+    {
+      "epoch": 0.11103708638685321,
+      "grad_norm": 2.146666992664208,
+      "learning_rate": 2.774694783573807e-06,
+      "loss": 0.1266,
+      "step": 250
+    },
+    {
+      "epoch": 0.11148123473240062,
+      "grad_norm": 1.450876621074019,
+      "learning_rate": 2.7857935627081024e-06,
+      "loss": 0.1124,
+      "step": 251
+    },
+    {
+      "epoch": 0.11192538307794804,
+      "grad_norm": 1.6447214379715893,
+      "learning_rate": 2.7968923418423975e-06,
+      "loss": 0.1176,
+      "step": 252
+    },
+    {
+      "epoch": 0.11236953142349544,
+      "grad_norm": 1.85196408048202,
+      "learning_rate": 2.807991120976693e-06,
+      "loss": 0.1196,
+      "step": 253
+    },
+    {
+      "epoch": 0.11281367976904286,
+      "grad_norm": 1.8297536053418253,
+      "learning_rate": 2.819089900110988e-06,
+      "loss": 0.1312,
+      "step": 254
+    },
+    {
+      "epoch": 0.11325782811459027,
+      "grad_norm": 1.7773730595281947,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.1343,
+      "step": 255
+    },
+    {
+      "epoch": 0.11370197646013769,
+      "grad_norm": 1.5584909202093926,
+      "learning_rate": 2.8412874583795786e-06,
+      "loss": 0.1075,
+      "step": 256
+    },
+    {
+      "epoch": 0.1141461248056851,
+      "grad_norm": 1.9920621657295152,
+      "learning_rate": 2.852386237513874e-06,
+      "loss": 0.1591,
+      "step": 257
+    },
+    {
+      "epoch": 0.1145902731512325,
+      "grad_norm": 2.0535793857620264,
+      "learning_rate": 2.8634850166481687e-06,
+      "loss": 0.1459,
+      "step": 258
+    },
+    {
+      "epoch": 0.11503442149677992,
+      "grad_norm": 1.432401039912359,
+      "learning_rate": 2.8745837957824642e-06,
+      "loss": 0.1235,
+      "step": 259
+    },
+    {
+      "epoch": 0.11547856984232734,
+      "grad_norm": 2.3190551632714427,
+      "learning_rate": 2.8856825749167593e-06,
+      "loss": 0.1608,
+      "step": 260
+    },
+    {
+      "epoch": 0.11592271818787475,
+      "grad_norm": 1.4461950992247072,
+      "learning_rate": 2.8967813540510548e-06,
+      "loss": 0.1242,
+      "step": 261
+    },
+    {
+      "epoch": 0.11636686653342217,
+      "grad_norm": 1.60132726584782,
+      "learning_rate": 2.9078801331853503e-06,
+      "loss": 0.1057,
+      "step": 262
+    },
+    {
+      "epoch": 0.11681101487896958,
+      "grad_norm": 2.5527110858786553,
+      "learning_rate": 2.918978912319645e-06,
+      "loss": 0.1735,
+      "step": 263
+    },
+    {
+      "epoch": 0.11725516322451698,
+      "grad_norm": 1.587535094612724,
+      "learning_rate": 2.9300776914539404e-06,
+      "loss": 0.1145,
+      "step": 264
+    },
+    {
+      "epoch": 0.1176993115700644,
+      "grad_norm": 1.8628254342286168,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.1475,
+      "step": 265
+    },
+    {
+      "epoch": 0.11814345991561181,
+      "grad_norm": 2.029208074213613,
+      "learning_rate": 2.952275249722531e-06,
+      "loss": 0.1448,
+      "step": 266
+    },
+    {
+      "epoch": 0.11858760826115923,
+      "grad_norm": 1.788916557618341,
+      "learning_rate": 2.9633740288568256e-06,
+      "loss": 0.1223,
+      "step": 267
+    },
+    {
+      "epoch": 0.11903175660670665,
+      "grad_norm": 2.0663932534564147,
+      "learning_rate": 2.974472807991121e-06,
+      "loss": 0.1456,
+      "step": 268
+    },
+    {
+      "epoch": 0.11947590495225405,
+      "grad_norm": 1.5929695265003,
+      "learning_rate": 2.9855715871254166e-06,
+      "loss": 0.1427,
+      "step": 269
+    },
+    {
+      "epoch": 0.11992005329780146,
+      "grad_norm": 1.8366971238631566,
+      "learning_rate": 2.996670366259712e-06,
+      "loss": 0.1029,
+      "step": 270
+    },
+    {
+      "epoch": 0.12036420164334888,
+      "grad_norm": 1.4941973532109012,
+      "learning_rate": 3.0077691453940067e-06,
+      "loss": 0.1115,
+      "step": 271
+    },
+    {
+      "epoch": 0.1208083499888963,
+      "grad_norm": 2.1604567983633403,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.11,
+      "step": 272
+    },
+    {
+      "epoch": 0.12125249833444371,
+      "grad_norm": 1.5440708951155746,
+      "learning_rate": 3.0299667036625973e-06,
+      "loss": 0.1312,
+      "step": 273
+    },
+    {
+      "epoch": 0.12169664667999111,
+      "grad_norm": 1.1320500690500013,
+      "learning_rate": 3.0410654827968928e-06,
+      "loss": 0.1172,
+      "step": 274
+    },
+    {
+      "epoch": 0.12214079502553853,
+      "grad_norm": 2.1326279999447326,
+      "learning_rate": 3.0521642619311882e-06,
+      "loss": 0.1099,
+      "step": 275
+    },
+    {
+      "epoch": 0.12258494337108594,
+      "grad_norm": 1.8624808272019928,
+      "learning_rate": 3.063263041065483e-06,
+      "loss": 0.1061,
+      "step": 276
+    },
+    {
+      "epoch": 0.12302909171663336,
+      "grad_norm": 2.836562163833866,
+      "learning_rate": 3.0743618201997784e-06,
+      "loss": 0.1173,
+      "step": 277
+    },
+    {
+      "epoch": 0.12347324006218077,
+      "grad_norm": 1.5779920497766018,
+      "learning_rate": 3.0854605993340734e-06,
+      "loss": 0.1305,
+      "step": 278
+    },
+    {
+      "epoch": 0.12391738840772819,
+      "grad_norm": 1.4061427971159166,
+      "learning_rate": 3.096559378468369e-06,
+      "loss": 0.1063,
+      "step": 279
+    },
+    {
+      "epoch": 0.12436153675327559,
+      "grad_norm": 1.4735808311660463,
+      "learning_rate": 3.1076581576026636e-06,
+      "loss": 0.1478,
+      "step": 280
+    },
+    {
+      "epoch": 0.124805685098823,
+      "grad_norm": 1.9063172097566101,
+      "learning_rate": 3.118756936736959e-06,
+      "loss": 0.152,
+      "step": 281
+    },
+    {
+      "epoch": 0.1252498334443704,
+      "grad_norm": 1.2460125667748942,
+      "learning_rate": 3.1298557158712546e-06,
+      "loss": 0.1078,
+      "step": 282
+    },
+    {
+      "epoch": 0.12569398178991784,
+      "grad_norm": 1.422370758153891,
+      "learning_rate": 3.1409544950055496e-06,
+      "loss": 0.1143,
+      "step": 283
+    },
+    {
+      "epoch": 0.12613813013546524,
+      "grad_norm": 1.3901208342210212,
+      "learning_rate": 3.1520532741398447e-06,
+      "loss": 0.1232,
+      "step": 284
+    },
+    {
+      "epoch": 0.12658227848101267,
+      "grad_norm": 1.2571280817929795,
+      "learning_rate": 3.16315205327414e-06,
+      "loss": 0.1183,
+      "step": 285
+    },
+    {
+      "epoch": 0.12702642682656007,
+      "grad_norm": 1.7211835588268667,
+      "learning_rate": 3.1742508324084352e-06,
+      "loss": 0.1376,
+      "step": 286
+    },
+    {
+      "epoch": 0.1274705751721075,
+      "grad_norm": 1.7352722841271955,
+      "learning_rate": 3.1853496115427307e-06,
+      "loss": 0.1221,
+      "step": 287
+    },
+    {
+      "epoch": 0.1279147235176549,
+      "grad_norm": 1.608485934770127,
+      "learning_rate": 3.1964483906770262e-06,
+      "loss": 0.1287,
+      "step": 288
+    },
+    {
+      "epoch": 0.1283588718632023,
+      "grad_norm": 1.4070749655660284,
+      "learning_rate": 3.207547169811321e-06,
+      "loss": 0.1107,
+      "step": 289
+    },
+    {
+      "epoch": 0.12880302020874973,
+      "grad_norm": 1.8061966734954316,
+      "learning_rate": 3.2186459489456164e-06,
+      "loss": 0.1209,
+      "step": 290
+    },
+    {
+      "epoch": 0.12924716855429713,
+      "grad_norm": 1.9290505056364757,
+      "learning_rate": 3.2297447280799114e-06,
+      "loss": 0.1088,
+      "step": 291
+    },
+    {
+      "epoch": 0.12969131689984456,
+      "grad_norm": 1.2873703020140206,
+      "learning_rate": 3.240843507214207e-06,
+      "loss": 0.1107,
+      "step": 292
+    },
+    {
+      "epoch": 0.13013546524539196,
+      "grad_norm": 2.100967763487988,
+      "learning_rate": 3.2519422863485016e-06,
+      "loss": 0.1746,
+      "step": 293
+    },
+    {
+      "epoch": 0.13057961359093936,
+      "grad_norm": 2.9637529933084785,
+      "learning_rate": 3.263041065482797e-06,
+      "loss": 0.1408,
+      "step": 294
+    },
+    {
+      "epoch": 0.1310237619364868,
+      "grad_norm": 1.8502103362104685,
+      "learning_rate": 3.2741398446170925e-06,
+      "loss": 0.108,
+      "step": 295
+    },
+    {
+      "epoch": 0.1314679102820342,
+      "grad_norm": 1.5072896588301588,
+      "learning_rate": 3.2852386237513876e-06,
+      "loss": 0.1233,
+      "step": 296
+    },
+    {
+      "epoch": 0.13191205862758162,
+      "grad_norm": 1.9060937237646072,
+      "learning_rate": 3.2963374028856827e-06,
+      "loss": 0.1137,
+      "step": 297
+    },
+    {
+      "epoch": 0.13235620697312903,
+      "grad_norm": 1.4214687758215054,
+      "learning_rate": 3.307436182019978e-06,
+      "loss": 0.1181,
+      "step": 298
+    },
+    {
+      "epoch": 0.13280035531867643,
+      "grad_norm": 1.5173189244791243,
+      "learning_rate": 3.3185349611542732e-06,
+      "loss": 0.1221,
+      "step": 299
+    },
+    {
+      "epoch": 0.13324450366422386,
+      "grad_norm": 1.4086327372245158,
+      "learning_rate": 3.3296337402885687e-06,
+      "loss": 0.1598,
+      "step": 300
+    },
+    {
+      "epoch": 0.13368865200977126,
+      "grad_norm": 1.3949120100912162,
+      "learning_rate": 3.3407325194228642e-06,
+      "loss": 0.0996,
+      "step": 301
+    },
+    {
+      "epoch": 0.1341328003553187,
+      "grad_norm": 1.6249998744801628,
+      "learning_rate": 3.351831298557159e-06,
+      "loss": 0.0959,
+      "step": 302
+    },
+    {
+      "epoch": 0.1345769487008661,
+      "grad_norm": 1.7178562509007014,
+      "learning_rate": 3.3629300776914543e-06,
+      "loss": 0.1348,
+      "step": 303
+    },
+    {
+      "epoch": 0.1350210970464135,
+      "grad_norm": 1.7790098039504103,
+      "learning_rate": 3.3740288568257494e-06,
+      "loss": 0.1011,
+      "step": 304
+    },
+    {
+      "epoch": 0.13546524539196092,
+      "grad_norm": 1.4533709920798474,
+      "learning_rate": 3.385127635960045e-06,
+      "loss": 0.1177,
+      "step": 305
+    },
+    {
+      "epoch": 0.13590939373750832,
+      "grad_norm": 1.7170638072373428,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 0.1264,
+      "step": 306
+    },
+    {
+      "epoch": 0.13635354208305575,
+      "grad_norm": 1.1622578542744249,
+      "learning_rate": 3.407325194228635e-06,
+      "loss": 0.1164,
+      "step": 307
+    },
+    {
+      "epoch": 0.13679769042860315,
+      "grad_norm": 1.7861497563291042,
+      "learning_rate": 3.4184239733629305e-06,
+      "loss": 0.1328,
+      "step": 308
+    },
+    {
+      "epoch": 0.13724183877415055,
+      "grad_norm": 1.2393311320446403,
+      "learning_rate": 3.4295227524972256e-06,
+      "loss": 0.0994,
+      "step": 309
+    },
+    {
+      "epoch": 0.13768598711969798,
+      "grad_norm": 1.779362058176627,
+      "learning_rate": 3.4406215316315207e-06,
+      "loss": 0.131,
+      "step": 310
+    },
+    {
+      "epoch": 0.13813013546524538,
+      "grad_norm": 1.384763433835653,
+      "learning_rate": 3.4517203107658157e-06,
+      "loss": 0.1011,
+      "step": 311
+    },
+    {
+      "epoch": 0.13857428381079281,
+      "grad_norm": 1.5455433688862117,
+      "learning_rate": 3.4628190899001112e-06,
+      "loss": 0.1216,
+      "step": 312
+    },
+    {
+      "epoch": 0.13901843215634022,
+      "grad_norm": 1.3658352699008705,
+      "learning_rate": 3.4739178690344067e-06,
+      "loss": 0.123,
+      "step": 313
+    },
+    {
+      "epoch": 0.13946258050188762,
+      "grad_norm": 1.3724682796873768,
+      "learning_rate": 3.485016648168702e-06,
+      "loss": 0.1209,
+      "step": 314
+    },
+    {
+      "epoch": 0.13990672884743505,
+      "grad_norm": 1.608691375904217,
+      "learning_rate": 3.496115427302997e-06,
+      "loss": 0.1141,
+      "step": 315
+    },
+    {
+      "epoch": 0.14035087719298245,
+      "grad_norm": 1.3598637431605427,
+      "learning_rate": 3.5072142064372923e-06,
+      "loss": 0.1528,
+      "step": 316
+    },
+    {
+      "epoch": 0.14079502553852988,
+      "grad_norm": 1.0876962111896626,
+      "learning_rate": 3.5183129855715874e-06,
+      "loss": 0.1049,
+      "step": 317
+    },
+    {
+      "epoch": 0.14123917388407728,
+      "grad_norm": 1.3385892109435766,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.1037,
+      "step": 318
+    },
+    {
+      "epoch": 0.14168332222962468,
+      "grad_norm": 1.7433775937165439,
+      "learning_rate": 3.5405105438401775e-06,
+      "loss": 0.1486,
+      "step": 319
+    },
+    {
+      "epoch": 0.1421274705751721,
+      "grad_norm": 1.5533508842477224,
+      "learning_rate": 3.551609322974473e-06,
+      "loss": 0.1049,
+      "step": 320
+    },
+    {
+      "epoch": 0.1425716189207195,
+      "grad_norm": 1.2029122587877374,
+      "learning_rate": 3.5627081021087685e-06,
+      "loss": 0.1098,
+      "step": 321
+    },
+    {
+      "epoch": 0.14301576726626694,
+      "grad_norm": 1.8995176312013884,
+      "learning_rate": 3.5738068812430636e-06,
+      "loss": 0.0944,
+      "step": 322
+    },
+    {
+      "epoch": 0.14345991561181434,
+      "grad_norm": 1.6602519149722867,
+      "learning_rate": 3.5849056603773586e-06,
+      "loss": 0.1243,
+      "step": 323
+    },
+    {
+      "epoch": 0.14390406395736177,
+      "grad_norm": 1.6075958194272566,
+      "learning_rate": 3.5960044395116537e-06,
+      "loss": 0.1196,
+      "step": 324
+    },
+    {
+      "epoch": 0.14434821230290917,
+      "grad_norm": 1.510226320322185,
+      "learning_rate": 3.607103218645949e-06,
+      "loss": 0.1422,
+      "step": 325
+    },
+    {
+      "epoch": 0.14479236064845658,
+      "grad_norm": 1.2802794537606514,
+      "learning_rate": 3.6182019977802447e-06,
+      "loss": 0.1244,
+      "step": 326
+    },
+    {
+      "epoch": 0.145236508994004,
+      "grad_norm": 1.1595946058732067,
+      "learning_rate": 3.6293007769145398e-06,
+      "loss": 0.0929,
+      "step": 327
+    },
+    {
+      "epoch": 0.1456806573395514,
+      "grad_norm": 1.2381936978069086,
+      "learning_rate": 3.640399556048835e-06,
+      "loss": 0.1002,
+      "step": 328
+    },
+    {
+      "epoch": 0.14612480568509884,
+      "grad_norm": 2.4354535778190742,
+      "learning_rate": 3.6514983351831303e-06,
+      "loss": 0.1377,
+      "step": 329
+    },
+    {
+      "epoch": 0.14656895403064624,
+      "grad_norm": 1.5157062223087485,
+      "learning_rate": 3.6625971143174254e-06,
+      "loss": 0.1206,
+      "step": 330
+    },
+    {
+      "epoch": 0.14701310237619364,
+      "grad_norm": 1.3681542301294034,
+      "learning_rate": 3.673695893451721e-06,
+      "loss": 0.1047,
+      "step": 331
+    },
+    {
+      "epoch": 0.14745725072174107,
+      "grad_norm": 1.9762614541590338,
+      "learning_rate": 3.6847946725860155e-06,
+      "loss": 0.1249,
+      "step": 332
+    },
+    {
+      "epoch": 0.14790139906728847,
+      "grad_norm": 1.391859368616253,
+      "learning_rate": 3.695893451720311e-06,
+      "loss": 0.1015,
+      "step": 333
+    },
+    {
+      "epoch": 0.1483455474128359,
+      "grad_norm": 1.1072542539549668,
+      "learning_rate": 3.7069922308546065e-06,
+      "loss": 0.0931,
+      "step": 334
+    },
+    {
+      "epoch": 0.1487896957583833,
+      "grad_norm": 1.4909584737380348,
+      "learning_rate": 3.7180910099889016e-06,
+      "loss": 0.1141,
+      "step": 335
+    },
+    {
+      "epoch": 0.1492338441039307,
+      "grad_norm": 1.7478929922992545,
+      "learning_rate": 3.7291897891231966e-06,
+      "loss": 0.0891,
+      "step": 336
+    },
+    {
+      "epoch": 0.14967799244947813,
+      "grad_norm": 1.5597867645297776,
+      "learning_rate": 3.7402885682574917e-06,
+      "loss": 0.128,
+      "step": 337
+    },
+    {
+      "epoch": 0.15012214079502553,
+      "grad_norm": 1.4083772110340225,
+      "learning_rate": 3.751387347391787e-06,
+      "loss": 0.1337,
+      "step": 338
+    },
+    {
+      "epoch": 0.15056628914057296,
+      "grad_norm": 2.011956681151715,
+      "learning_rate": 3.7624861265260827e-06,
+      "loss": 0.1036,
+      "step": 339
+    },
+    {
+      "epoch": 0.15101043748612036,
+      "grad_norm": 2.154404794046358,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.1104,
+      "step": 340
+    },
+    {
+      "epoch": 0.15145458583166777,
+      "grad_norm": 2.1106357904010316,
+      "learning_rate": 3.784683684794673e-06,
+      "loss": 0.1686,
+      "step": 341
+    },
+    {
+      "epoch": 0.1518987341772152,
+      "grad_norm": 1.6933676844964125,
+      "learning_rate": 3.7957824639289683e-06,
+      "loss": 0.1375,
+      "step": 342
+    },
+    {
+      "epoch": 0.1523428825227626,
+      "grad_norm": 2.205537988403305,
+      "learning_rate": 3.8068812430632634e-06,
+      "loss": 0.1513,
+      "step": 343
+    },
+    {
+      "epoch": 0.15278703086831003,
+      "grad_norm": 1.2058069327729946,
+      "learning_rate": 3.817980022197559e-06,
+      "loss": 0.091,
+      "step": 344
+    },
+    {
+      "epoch": 0.15323117921385743,
+      "grad_norm": 1.400652563163732,
+      "learning_rate": 3.829078801331854e-06,
+      "loss": 0.0826,
+      "step": 345
+    },
+    {
+      "epoch": 0.15367532755940483,
+      "grad_norm": 2.057959821320217,
+      "learning_rate": 3.840177580466149e-06,
+      "loss": 0.118,
+      "step": 346
+    },
+    {
+      "epoch": 0.15411947590495226,
+      "grad_norm": 1.6604930714816526,
+      "learning_rate": 3.851276359600444e-06,
+      "loss": 0.1301,
+      "step": 347
+    },
+    {
+      "epoch": 0.15456362425049966,
+      "grad_norm": 1.033671229980745,
+      "learning_rate": 3.86237513873474e-06,
+      "loss": 0.0924,
+      "step": 348
+    },
+    {
+      "epoch": 0.1550077725960471,
+      "grad_norm": 1.17967777436608,
+      "learning_rate": 3.873473917869034e-06,
+      "loss": 0.0978,
+      "step": 349
+    },
+    {
+      "epoch": 0.1554519209415945,
+      "grad_norm": 1.650470297879078,
+      "learning_rate": 3.88457269700333e-06,
+      "loss": 0.133,
+      "step": 350
+    },
+    {
+      "epoch": 0.1558960692871419,
+      "grad_norm": 1.47519341296619,
+      "learning_rate": 3.895671476137625e-06,
+      "loss": 0.1254,
+      "step": 351
+    },
+    {
+      "epoch": 0.15634021763268932,
+      "grad_norm": 1.4035761818876917,
+      "learning_rate": 3.90677025527192e-06,
+      "loss": 0.1183,
+      "step": 352
+    },
+    {
+      "epoch": 0.15678436597823672,
+      "grad_norm": 2.2555395385843036,
+      "learning_rate": 3.917869034406216e-06,
+      "loss": 0.1202,
+      "step": 353
+    },
+    {
+      "epoch": 0.15722851432378415,
+      "grad_norm": 1.4759637016708067,
+      "learning_rate": 3.92896781354051e-06,
+      "loss": 0.1104,
+      "step": 354
+    },
+    {
+      "epoch": 0.15767266266933155,
+      "grad_norm": 2.952687567266444,
+      "learning_rate": 3.940066592674806e-06,
+      "loss": 0.122,
+      "step": 355
+    },
+    {
+      "epoch": 0.15811681101487898,
+      "grad_norm": 1.1679111801200106,
+      "learning_rate": 3.951165371809101e-06,
+      "loss": 0.0905,
+      "step": 356
+    },
+    {
+      "epoch": 0.15856095936042638,
+      "grad_norm": 1.660729161757867,
+      "learning_rate": 3.962264150943396e-06,
+      "loss": 0.1521,
+      "step": 357
+    },
+    {
+      "epoch": 0.15900510770597379,
+      "grad_norm": 1.3321535227222834,
+      "learning_rate": 3.9733629300776915e-06,
+      "loss": 0.1315,
+      "step": 358
+    },
+    {
+      "epoch": 0.15944925605152122,
+      "grad_norm": 1.1798749965091133,
+      "learning_rate": 3.9844617092119866e-06,
+      "loss": 0.0852,
+      "step": 359
+    },
+    {
+      "epoch": 0.15989340439706862,
+      "grad_norm": 1.198919817717748,
+      "learning_rate": 3.9955604883462825e-06,
+      "loss": 0.1029,
+      "step": 360
+    },
+    {
+      "epoch": 0.16033755274261605,
+      "grad_norm": 1.1799761966591196,
+      "learning_rate": 4.0066592674805775e-06,
+      "loss": 0.1009,
+      "step": 361
+    },
+    {
+      "epoch": 0.16078170108816345,
+      "grad_norm": 1.3843920770849105,
+      "learning_rate": 4.017758046614873e-06,
+      "loss": 0.1078,
+      "step": 362
+    },
+    {
+      "epoch": 0.16122584943371085,
+      "grad_norm": 1.230254034984121,
+      "learning_rate": 4.028856825749168e-06,
+      "loss": 0.1218,
+      "step": 363
+    },
+    {
+      "epoch": 0.16166999777925828,
+      "grad_norm": 1.0307816837555734,
+      "learning_rate": 4.039955604883464e-06,
+      "loss": 0.1138,
+      "step": 364
+    },
+    {
+      "epoch": 0.16211414612480568,
+      "grad_norm": 1.1369945329768874,
+      "learning_rate": 4.051054384017759e-06,
+      "loss": 0.0917,
+      "step": 365
+    },
+    {
+      "epoch": 0.1625582944703531,
+      "grad_norm": 1.3026077981860287,
+      "learning_rate": 4.062153163152054e-06,
+      "loss": 0.1209,
+      "step": 366
+    },
+    {
+      "epoch": 0.1630024428159005,
+      "grad_norm": 1.8570215631965663,
+      "learning_rate": 4.073251942286349e-06,
+      "loss": 0.103,
+      "step": 367
+    },
+    {
+      "epoch": 0.1634465911614479,
+      "grad_norm": 1.4261357373143895,
+      "learning_rate": 4.084350721420644e-06,
+      "loss": 0.1117,
+      "step": 368
+    },
+    {
+      "epoch": 0.16389073950699534,
+      "grad_norm": 1.2842977064487127,
+      "learning_rate": 4.09544950055494e-06,
+      "loss": 0.0945,
+      "step": 369
+    },
+    {
+      "epoch": 0.16433488785254274,
+      "grad_norm": 1.7942409210847103,
+      "learning_rate": 4.106548279689235e-06,
+      "loss": 0.1658,
+      "step": 370
+    },
+    {
+      "epoch": 0.16477903619809017,
+      "grad_norm": 1.2570893224456356,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.1238,
+      "step": 371
+    },
+    {
+      "epoch": 0.16522318454363757,
+      "grad_norm": 1.0685833693361235,
+      "learning_rate": 4.128745837957825e-06,
+      "loss": 0.0964,
+      "step": 372
+    },
+    {
+      "epoch": 0.16566733288918498,
+      "grad_norm": 1.2815898248385025,
+      "learning_rate": 4.13984461709212e-06,
+      "loss": 0.105,
+      "step": 373
+    },
+    {
+      "epoch": 0.1661114812347324,
+      "grad_norm": 1.64061710866675,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.1096,
+      "step": 374
+    },
+    {
+      "epoch": 0.1665556295802798,
+      "grad_norm": 1.3088826974530428,
+      "learning_rate": 4.16204217536071e-06,
+      "loss": 0.1211,
+      "step": 375
+    },
+    {
+      "epoch": 0.16699977792582724,
+      "grad_norm": 1.3003151707348597,
+      "learning_rate": 4.173140954495006e-06,
+      "loss": 0.1111,
+      "step": 376
+    },
+    {
+      "epoch": 0.16744392627137464,
+      "grad_norm": 1.3124364780409907,
+      "learning_rate": 4.184239733629301e-06,
+      "loss": 0.1085,
+      "step": 377
+    },
+    {
+      "epoch": 0.16788807461692204,
+      "grad_norm": 1.3347095844879298,
+      "learning_rate": 4.195338512763596e-06,
+      "loss": 0.1228,
+      "step": 378
+    },
+    {
+      "epoch": 0.16833222296246947,
+      "grad_norm": 1.3151237942729317,
+      "learning_rate": 4.206437291897892e-06,
+      "loss": 0.1041,
+      "step": 379
+    },
+    {
+      "epoch": 0.16877637130801687,
+      "grad_norm": 1.6082774402292863,
+      "learning_rate": 4.217536071032186e-06,
+      "loss": 0.0915,
+      "step": 380
+    },
+    {
+      "epoch": 0.1692205196535643,
+      "grad_norm": 1.2544764918032303,
+      "learning_rate": 4.228634850166482e-06,
+      "loss": 0.0903,
+      "step": 381
+    },
+    {
+      "epoch": 0.1696646679991117,
+      "grad_norm": 1.4709993024116135,
+      "learning_rate": 4.239733629300777e-06,
+      "loss": 0.1095,
+      "step": 382
+    },
+    {
+      "epoch": 0.1701088163446591,
+      "grad_norm": 1.090059538683836,
+      "learning_rate": 4.250832408435072e-06,
+      "loss": 0.09,
+      "step": 383
+    },
+    {
+      "epoch": 0.17055296469020653,
+      "grad_norm": 1.936001842072003,
+      "learning_rate": 4.2619311875693675e-06,
+      "loss": 0.125,
+      "step": 384
+    },
+    {
+      "epoch": 0.17099711303575393,
+      "grad_norm": 1.7118256852712324,
+      "learning_rate": 4.2730299667036625e-06,
+      "loss": 0.095,
+      "step": 385
+    },
+    {
+      "epoch": 0.17144126138130136,
+      "grad_norm": 2.0311105495228268,
+      "learning_rate": 4.2841287458379584e-06,
+      "loss": 0.1029,
+      "step": 386
+    },
+    {
+      "epoch": 0.17188540972684876,
+      "grad_norm": 1.444490027117435,
+      "learning_rate": 4.2952275249722535e-06,
+      "loss": 0.1232,
+      "step": 387
+    },
+    {
+      "epoch": 0.1723295580723962,
+      "grad_norm": 1.4378003146516516,
+      "learning_rate": 4.3063263041065486e-06,
+      "loss": 0.0932,
+      "step": 388
+    },
+    {
+      "epoch": 0.1727737064179436,
+      "grad_norm": 1.1043097695262434,
+      "learning_rate": 4.317425083240844e-06,
+      "loss": 0.0994,
+      "step": 389
+    },
+    {
+      "epoch": 0.173217854763491,
+      "grad_norm": 1.1954032097891434,
+      "learning_rate": 4.328523862375139e-06,
+      "loss": 0.101,
+      "step": 390
+    },
+    {
+      "epoch": 0.17366200310903843,
+      "grad_norm": 1.6447959614931191,
+      "learning_rate": 4.339622641509435e-06,
+      "loss": 0.1343,
+      "step": 391
+    },
+    {
+      "epoch": 0.17410615145458583,
+      "grad_norm": 1.5660886998105679,
+      "learning_rate": 4.35072142064373e-06,
+      "loss": 0.1052,
+      "step": 392
+    },
+    {
+      "epoch": 0.17455029980013326,
+      "grad_norm": 1.129037326613576,
+      "learning_rate": 4.361820199778025e-06,
+      "loss": 0.0949,
+      "step": 393
+    },
+    {
+      "epoch": 0.17499444814568066,
+      "grad_norm": 0.9299986418563283,
+      "learning_rate": 4.37291897891232e-06,
+      "loss": 0.091,
+      "step": 394
+    },
+    {
+      "epoch": 0.17543859649122806,
+      "grad_norm": 1.5748400271295198,
+      "learning_rate": 4.384017758046616e-06,
+      "loss": 0.1688,
+      "step": 395
+    },
+    {
+      "epoch": 0.1758827448367755,
+      "grad_norm": 0.8778899838246078,
+      "learning_rate": 4.395116537180911e-06,
+      "loss": 0.0983,
+      "step": 396
+    },
+    {
+      "epoch": 0.1763268931823229,
+      "grad_norm": 1.1516100437124952,
+      "learning_rate": 4.406215316315206e-06,
+      "loss": 0.1201,
+      "step": 397
+    },
+    {
+      "epoch": 0.17677104152787032,
+      "grad_norm": 0.9883123409205935,
+      "learning_rate": 4.417314095449501e-06,
+      "loss": 0.0693,
+      "step": 398
+    },
+    {
+      "epoch": 0.17721518987341772,
+      "grad_norm": 1.367128447993114,
+      "learning_rate": 4.428412874583796e-06,
+      "loss": 0.1009,
+      "step": 399
+    },
+    {
+      "epoch": 0.17765933821896512,
+      "grad_norm": 1.2639479044263988,
+      "learning_rate": 4.439511653718092e-06,
+      "loss": 0.093,
+      "step": 400
+    },
+    {
+      "epoch": 0.17810348656451255,
+      "grad_norm": 1.284067214420426,
+      "learning_rate": 4.450610432852386e-06,
+      "loss": 0.1029,
+      "step": 401
+    },
+    {
+      "epoch": 0.17854763491005995,
+      "grad_norm": 0.9723989283162116,
+      "learning_rate": 4.461709211986682e-06,
+      "loss": 0.093,
+      "step": 402
+    },
+    {
+      "epoch": 0.17899178325560738,
+      "grad_norm": 1.1814808630811793,
+      "learning_rate": 4.472807991120977e-06,
+      "loss": 0.1004,
+      "step": 403
+    },
+    {
+      "epoch": 0.17943593160115479,
+      "grad_norm": 1.2100875095234747,
+      "learning_rate": 4.483906770255272e-06,
+      "loss": 0.1165,
+      "step": 404
+    },
+    {
+      "epoch": 0.1798800799467022,
+      "grad_norm": 1.2480781051362362,
+      "learning_rate": 4.495005549389568e-06,
+      "loss": 0.1086,
+      "step": 405
+    },
+    {
+      "epoch": 0.18032422829224962,
+      "grad_norm": 1.4659555944931602,
+      "learning_rate": 4.506104328523862e-06,
+      "loss": 0.0966,
+      "step": 406
+    },
+    {
+      "epoch": 0.18076837663779702,
+      "grad_norm": 1.2905749286816455,
+      "learning_rate": 4.517203107658158e-06,
+      "loss": 0.1308,
+      "step": 407
+    },
+    {
+      "epoch": 0.18121252498334445,
+      "grad_norm": 1.2083135245558752,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 0.139,
+      "step": 408
+    },
+    {
+      "epoch": 0.18165667332889185,
+      "grad_norm": 1.2677821305643562,
+      "learning_rate": 4.539400665926748e-06,
+      "loss": 0.1522,
+      "step": 409
+    },
+    {
+      "epoch": 0.18210082167443925,
+      "grad_norm": 0.8309366049632025,
+      "learning_rate": 4.5504994450610434e-06,
+      "loss": 0.0955,
+      "step": 410
+    },
+    {
+      "epoch": 0.18254497001998668,
+      "grad_norm": 1.6708204443941888,
+      "learning_rate": 4.5615982241953385e-06,
+      "loss": 0.0983,
+      "step": 411
+    },
+    {
+      "epoch": 0.18298911836553408,
+      "grad_norm": 1.8028863536801523,
+      "learning_rate": 4.572697003329634e-06,
+      "loss": 0.1141,
+      "step": 412
+    },
+    {
+      "epoch": 0.1834332667110815,
+      "grad_norm": 1.6039168610165708,
+      "learning_rate": 4.5837957824639295e-06,
+      "loss": 0.14,
+      "step": 413
+    },
+    {
+      "epoch": 0.1838774150566289,
+      "grad_norm": 1.3909894579389195,
+      "learning_rate": 4.5948945615982245e-06,
+      "loss": 0.0924,
+      "step": 414
+    },
+    {
+      "epoch": 0.1843215634021763,
+      "grad_norm": 1.444002762358773,
+      "learning_rate": 4.60599334073252e-06,
+      "loss": 0.0882,
+      "step": 415
+    },
+    {
+      "epoch": 0.18476571174772374,
+      "grad_norm": 1.3517930787179577,
+      "learning_rate": 4.617092119866815e-06,
+      "loss": 0.1198,
+      "step": 416
+    },
+    {
+      "epoch": 0.18520986009327114,
+      "grad_norm": 0.9651918865616642,
+      "learning_rate": 4.628190899001111e-06,
+      "loss": 0.0794,
+      "step": 417
+    },
+    {
+      "epoch": 0.18565400843881857,
+      "grad_norm": 1.3087574560794024,
+      "learning_rate": 4.639289678135406e-06,
+      "loss": 0.1166,
+      "step": 418
+    },
+    {
+      "epoch": 0.18609815678436598,
+      "grad_norm": 1.1029726712835008,
+      "learning_rate": 4.650388457269701e-06,
+      "loss": 0.106,
+      "step": 419
+    },
+    {
+      "epoch": 0.18654230512991338,
+      "grad_norm": 1.481760865085893,
+      "learning_rate": 4.661487236403996e-06,
+      "loss": 0.1225,
+      "step": 420
+    },
+    {
+      "epoch": 0.1869864534754608,
+      "grad_norm": 1.125656650735202,
+      "learning_rate": 4.672586015538291e-06,
+      "loss": 0.0894,
+      "step": 421
+    },
+    {
+      "epoch": 0.1874306018210082,
+      "grad_norm": 1.1643460761006563,
+      "learning_rate": 4.683684794672587e-06,
+      "loss": 0.1049,
+      "step": 422
+    },
+    {
+      "epoch": 0.18787475016655564,
+      "grad_norm": 1.2081839286402132,
+      "learning_rate": 4.694783573806882e-06,
+      "loss": 0.0984,
+      "step": 423
+    },
+    {
+      "epoch": 0.18831889851210304,
+      "grad_norm": 1.0952768068994485,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.1083,
+      "step": 424
+    },
+    {
+      "epoch": 0.18876304685765047,
+      "grad_norm": 1.993217784490899,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.1198,
+      "step": 425
+    },
+    {
+      "epoch": 0.18920719520319787,
+      "grad_norm": 1.4709865866324339,
+      "learning_rate": 4.728079911209768e-06,
+      "loss": 0.1004,
+      "step": 426
+    },
+    {
+      "epoch": 0.18965134354874527,
+      "grad_norm": 1.1800698065694786,
+      "learning_rate": 4.739178690344062e-06,
+      "loss": 0.0942,
+      "step": 427
+    },
+    {
+      "epoch": 0.1900954918942927,
+      "grad_norm": 0.9004404155709735,
+      "learning_rate": 4.750277469478358e-06,
+      "loss": 0.0879,
+      "step": 428
+    },
+    {
+      "epoch": 0.1905396402398401,
+      "grad_norm": 1.3988615855599944,
+      "learning_rate": 4.761376248612653e-06,
+      "loss": 0.1313,
+      "step": 429
+    },
+    {
+      "epoch": 0.19098378858538753,
+      "grad_norm": 1.2097973380846636,
+      "learning_rate": 4.772475027746948e-06,
+      "loss": 0.1192,
+      "step": 430
+    },
+    {
+      "epoch": 0.19142793693093493,
+      "grad_norm": 1.795772492183597,
+      "learning_rate": 4.783573806881244e-06,
+      "loss": 0.1319,
+      "step": 431
+    },
+    {
+      "epoch": 0.19187208527648233,
+      "grad_norm": 1.5438576982175642,
+      "learning_rate": 4.794672586015538e-06,
+      "loss": 0.156,
+      "step": 432
+    },
+    {
+      "epoch": 0.19231623362202976,
+      "grad_norm": 1.4462148356085274,
+      "learning_rate": 4.805771365149834e-06,
+      "loss": 0.0977,
+      "step": 433
+    },
+    {
+      "epoch": 0.19276038196757717,
+      "grad_norm": 1.1013117212825567,
+      "learning_rate": 4.816870144284129e-06,
+      "loss": 0.123,
+      "step": 434
+    },
+    {
+      "epoch": 0.1932045303131246,
+      "grad_norm": 1.473729194266236,
+      "learning_rate": 4.827968923418424e-06,
+      "loss": 0.1219,
+      "step": 435
+    },
+    {
+      "epoch": 0.193648678658672,
+      "grad_norm": 1.060585950094457,
+      "learning_rate": 4.839067702552719e-06,
+      "loss": 0.0867,
+      "step": 436
+    },
+    {
+      "epoch": 0.1940928270042194,
+      "grad_norm": 1.1939948263664504,
+      "learning_rate": 4.8501664816870145e-06,
+      "loss": 0.1221,
+      "step": 437
+    },
+    {
+      "epoch": 0.19453697534976683,
+      "grad_norm": 1.0411525410331977,
+      "learning_rate": 4.86126526082131e-06,
+      "loss": 0.0955,
+      "step": 438
+    },
+    {
+      "epoch": 0.19498112369531423,
+      "grad_norm": 1.3074349171537876,
+      "learning_rate": 4.8723640399556054e-06,
+      "loss": 0.1092,
+      "step": 439
+    },
+    {
+      "epoch": 0.19542527204086166,
+      "grad_norm": 2.556790721991839,
+      "learning_rate": 4.8834628190899005e-06,
+      "loss": 0.1144,
+      "step": 440
+    },
+    {
+      "epoch": 0.19586942038640906,
+      "grad_norm": 1.5219706664834065,
+      "learning_rate": 4.894561598224196e-06,
+      "loss": 0.1353,
+      "step": 441
+    },
+    {
+      "epoch": 0.19631356873195646,
+      "grad_norm": 1.5771009476283926,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.1139,
+      "step": 442
+    },
+    {
+      "epoch": 0.1967577170775039,
+      "grad_norm": 1.7247841251255984,
+      "learning_rate": 4.9167591564927866e-06,
+      "loss": 0.0969,
+      "step": 443
+    },
+    {
+      "epoch": 0.1972018654230513,
+      "grad_norm": 1.118860834302166,
+      "learning_rate": 4.927857935627082e-06,
+      "loss": 0.0955,
+      "step": 444
+    },
+    {
+      "epoch": 0.19764601376859872,
+      "grad_norm": 1.4715029598755733,
+      "learning_rate": 4.938956714761377e-06,
+      "loss": 0.1001,
+      "step": 445
+    },
+    {
+      "epoch": 0.19809016211414612,
+      "grad_norm": 1.5680153762667055,
+      "learning_rate": 4.950055493895672e-06,
+      "loss": 0.1222,
+      "step": 446
+    },
+    {
+      "epoch": 0.19853431045969352,
+      "grad_norm": 1.7877852329091748,
+      "learning_rate": 4.961154273029967e-06,
+      "loss": 0.1057,
+      "step": 447
+    },
+    {
+      "epoch": 0.19897845880524095,
+      "grad_norm": 1.1650327458270198,
+      "learning_rate": 4.972253052164263e-06,
+      "loss": 0.1242,
+      "step": 448
+    },
+    {
+      "epoch": 0.19942260715078836,
+      "grad_norm": 1.0753006649574548,
+      "learning_rate": 4.983351831298557e-06,
+      "loss": 0.0912,
+      "step": 449
+    },
+    {
+      "epoch": 0.19986675549633579,
+      "grad_norm": 1.546863382694724,
+      "learning_rate": 4.994450610432853e-06,
+      "loss": 0.123,
+      "step": 450
+    },
+    {
+      "epoch": 0.2003109038418832,
+      "grad_norm": 1.3219021070497294,
+      "learning_rate": 5.005549389567148e-06,
+      "loss": 0.1167,
+      "step": 451
+    },
+    {
+      "epoch": 0.2007550521874306,
+      "grad_norm": 1.3154364501774296,
+      "learning_rate": 5.016648168701444e-06,
+      "loss": 0.1023,
+      "step": 452
+    },
+    {
+      "epoch": 0.20119920053297802,
+      "grad_norm": 1.2732051452073168,
+      "learning_rate": 5.027746947835739e-06,
+      "loss": 0.0717,
+      "step": 453
+    },
+    {
+      "epoch": 0.20164334887852542,
+      "grad_norm": 1.3206735580145712,
+      "learning_rate": 5.038845726970034e-06,
+      "loss": 0.1095,
+      "step": 454
+    },
+    {
+      "epoch": 0.20208749722407285,
+      "grad_norm": 1.0332614998616947,
+      "learning_rate": 5.049944506104328e-06,
+      "loss": 0.0845,
+      "step": 455
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 1.211744462644113,
+      "learning_rate": 5.061043285238624e-06,
+      "loss": 0.1066,
+      "step": 456
+    },
+    {
+      "epoch": 0.20297579391516768,
+      "grad_norm": 1.3376009797970827,
+      "learning_rate": 5.072142064372919e-06,
+      "loss": 0.0905,
+      "step": 457
+    },
+    {
+      "epoch": 0.20341994226071508,
+      "grad_norm": 1.2905140024192596,
+      "learning_rate": 5.083240843507214e-06,
+      "loss": 0.1011,
+      "step": 458
+    },
+    {
+      "epoch": 0.20386409060626248,
+      "grad_norm": 1.1671668945164224,
+      "learning_rate": 5.09433962264151e-06,
+      "loss": 0.1027,
+      "step": 459
+    },
+    {
+      "epoch": 0.2043082389518099,
+      "grad_norm": 2.1761489486755363,
+      "learning_rate": 5.105438401775805e-06,
+      "loss": 0.1093,
+      "step": 460
+    },
+    {
+      "epoch": 0.2047523872973573,
+      "grad_norm": 1.5496188187186781,
+      "learning_rate": 5.1165371809101e-06,
+      "loss": 0.1241,
+      "step": 461
+    },
+    {
+      "epoch": 0.20519653564290474,
+      "grad_norm": 1.555761458410687,
+      "learning_rate": 5.127635960044396e-06,
+      "loss": 0.1119,
+      "step": 462
+    },
+    {
+      "epoch": 0.20564068398845214,
+      "grad_norm": 1.4018925493874805,
+      "learning_rate": 5.138734739178691e-06,
+      "loss": 0.149,
+      "step": 463
+    },
+    {
+      "epoch": 0.20608483233399955,
+      "grad_norm": 1.2908877898264874,
+      "learning_rate": 5.1498335183129855e-06,
+      "loss": 0.1255,
+      "step": 464
+    },
+    {
+      "epoch": 0.20652898067954698,
+      "grad_norm": 1.2526168814511602,
+      "learning_rate": 5.1609322974472806e-06,
+      "loss": 0.1145,
+      "step": 465
+    },
+    {
+      "epoch": 0.20697312902509438,
+      "grad_norm": 1.1641091501138674,
+      "learning_rate": 5.1720310765815765e-06,
+      "loss": 0.1136,
+      "step": 466
+    },
+    {
+      "epoch": 0.2074172773706418,
+      "grad_norm": 1.1082496575030598,
+      "learning_rate": 5.1831298557158716e-06,
+      "loss": 0.0897,
+      "step": 467
+    },
+    {
+      "epoch": 0.2078614257161892,
+      "grad_norm": 1.174903547638247,
+      "learning_rate": 5.194228634850167e-06,
+      "loss": 0.1121,
+      "step": 468
+    },
+    {
+      "epoch": 0.2083055740617366,
+      "grad_norm": 1.0600460968712155,
+      "learning_rate": 5.2053274139844625e-06,
+      "loss": 0.0819,
+      "step": 469
+    },
+    {
+      "epoch": 0.20874972240728404,
+      "grad_norm": 1.3263800742507028,
+      "learning_rate": 5.216426193118758e-06,
+      "loss": 0.084,
+      "step": 470
+    },
+    {
+      "epoch": 0.20919387075283144,
+      "grad_norm": 1.604108989764318,
+      "learning_rate": 5.227524972253053e-06,
+      "loss": 0.1337,
+      "step": 471
+    },
+    {
+      "epoch": 0.20963801909837887,
+      "grad_norm": 0.9363413921696734,
+      "learning_rate": 5.238623751387349e-06,
+      "loss": 0.0863,
+      "step": 472
+    },
+    {
+      "epoch": 0.21008216744392627,
+      "grad_norm": 2.0569016097307578,
+      "learning_rate": 5.249722530521643e-06,
+      "loss": 0.1218,
+      "step": 473
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 2.0907599336843674,
+      "learning_rate": 5.260821309655938e-06,
+      "loss": 0.1032,
+      "step": 474
+    },
+    {
+      "epoch": 0.2109704641350211,
+      "grad_norm": 1.03795187986254,
+      "learning_rate": 5.271920088790233e-06,
+      "loss": 0.1002,
+      "step": 475
+    },
+    {
+      "epoch": 0.2114146124805685,
+      "grad_norm": 1.6168543989206048,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 0.1292,
+      "step": 476
+    },
+    {
+      "epoch": 0.21185876082611593,
+      "grad_norm": 1.555835178594497,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.0956,
+      "step": 477
+    },
+    {
+      "epoch": 0.21230290917166333,
+      "grad_norm": 1.4750911257012305,
+      "learning_rate": 5.305216426193119e-06,
+      "loss": 0.1025,
+      "step": 478
+    },
+    {
+      "epoch": 0.21274705751721074,
+      "grad_norm": 1.1903302332979158,
+      "learning_rate": 5.316315205327415e-06,
+      "loss": 0.0927,
+      "step": 479
+    },
+    {
+      "epoch": 0.21319120586275817,
+      "grad_norm": 1.799907699381259,
+      "learning_rate": 5.32741398446171e-06,
+      "loss": 0.1554,
+      "step": 480
+    },
+    {
+      "epoch": 0.21363535420830557,
+      "grad_norm": 1.1084539632711692,
+      "learning_rate": 5.338512763596004e-06,
+      "loss": 0.0963,
+      "step": 481
+    },
+    {
+      "epoch": 0.214079502553853,
+      "grad_norm": 1.0531894918937277,
+      "learning_rate": 5.3496115427303e-06,
+      "loss": 0.0808,
+      "step": 482
+    },
+    {
+      "epoch": 0.2145236508994004,
+      "grad_norm": 1.4912911807270512,
+      "learning_rate": 5.360710321864595e-06,
+      "loss": 0.1053,
+      "step": 483
+    },
+    {
+      "epoch": 0.2149677992449478,
+      "grad_norm": 1.1603558318841067,
+      "learning_rate": 5.37180910099889e-06,
+      "loss": 0.0859,
+      "step": 484
+    },
+    {
+      "epoch": 0.21541194759049523,
+      "grad_norm": 1.2647862898402058,
+      "learning_rate": 5.382907880133186e-06,
+      "loss": 0.1165,
+      "step": 485
+    },
+    {
+      "epoch": 0.21585609593604263,
+      "grad_norm": 1.3294179512793625,
+      "learning_rate": 5.394006659267481e-06,
+      "loss": 0.1006,
+      "step": 486
+    },
+    {
+      "epoch": 0.21630024428159006,
+      "grad_norm": 1.288870748090468,
+      "learning_rate": 5.405105438401776e-06,
+      "loss": 0.1174,
+      "step": 487
+    },
+    {
+      "epoch": 0.21674439262713746,
+      "grad_norm": 1.6526073205290404,
+      "learning_rate": 5.416204217536072e-06,
+      "loss": 0.1278,
+      "step": 488
+    },
+    {
+      "epoch": 0.2171885409726849,
+      "grad_norm": 1.1625080059645232,
+      "learning_rate": 5.427302996670367e-06,
+      "loss": 0.0962,
+      "step": 489
+    },
+    {
+      "epoch": 0.2176326893182323,
+      "grad_norm": 1.0239562074655422,
+      "learning_rate": 5.4384017758046615e-06,
+      "loss": 0.1137,
+      "step": 490
+    },
+    {
+      "epoch": 0.2180768376637797,
+      "grad_norm": 1.1775834154919058,
+      "learning_rate": 5.4495005549389565e-06,
+      "loss": 0.1167,
+      "step": 491
+    },
+    {
+      "epoch": 0.21852098600932712,
+      "grad_norm": 1.0527464833567977,
+      "learning_rate": 5.4605993340732525e-06,
+      "loss": 0.1035,
+      "step": 492
+    },
+    {
+      "epoch": 0.21896513435487452,
+      "grad_norm": 1.0639212306477968,
+      "learning_rate": 5.4716981132075475e-06,
+      "loss": 0.0798,
+      "step": 493
+    },
+    {
+      "epoch": 0.21940928270042195,
+      "grad_norm": 1.4130768874633084,
+      "learning_rate": 5.482796892341843e-06,
+      "loss": 0.1191,
+      "step": 494
+    },
+    {
+      "epoch": 0.21985343104596936,
+      "grad_norm": 1.1493987018132557,
+      "learning_rate": 5.4938956714761385e-06,
+      "loss": 0.0933,
+      "step": 495
+    },
+    {
+      "epoch": 0.22029757939151676,
+      "grad_norm": 1.5099978612143894,
+      "learning_rate": 5.5049944506104336e-06,
+      "loss": 0.1312,
+      "step": 496
+    },
+    {
+      "epoch": 0.2207417277370642,
+      "grad_norm": 1.7153041600616725,
+      "learning_rate": 5.516093229744729e-06,
+      "loss": 0.1631,
+      "step": 497
+    },
+    {
+      "epoch": 0.2211858760826116,
+      "grad_norm": 0.9681698333927948,
+      "learning_rate": 5.5271920088790245e-06,
+      "loss": 0.0811,
+      "step": 498
+    },
+    {
+      "epoch": 0.22163002442815902,
+      "grad_norm": 1.2720779336432644,
+      "learning_rate": 5.538290788013319e-06,
+      "loss": 0.1305,
+      "step": 499
+    },
+    {
+      "epoch": 0.22207417277370642,
+      "grad_norm": 1.2223556006920575,
+      "learning_rate": 5.549389567147614e-06,
+      "loss": 0.0942,
+      "step": 500
+    },
+    {
+      "epoch": 0.22251832111925382,
+      "grad_norm": 1.0558717161626225,
+      "learning_rate": 5.560488346281909e-06,
+      "loss": 0.1077,
+      "step": 501
+    },
+    {
+      "epoch": 0.22296246946480125,
+      "grad_norm": 1.1170017223897397,
+      "learning_rate": 5.571587125416205e-06,
+      "loss": 0.1309,
+      "step": 502
+    },
+    {
+      "epoch": 0.22340661781034865,
+      "grad_norm": 0.9964488397086242,
+      "learning_rate": 5.5826859045505e-06,
+      "loss": 0.0834,
+      "step": 503
+    },
+    {
+      "epoch": 0.22385076615589608,
+      "grad_norm": 1.2815734363162072,
+      "learning_rate": 5.593784683684795e-06,
+      "loss": 0.0901,
+      "step": 504
+    },
+    {
+      "epoch": 0.22429491450144348,
+      "grad_norm": 1.0732922534576572,
+      "learning_rate": 5.604883462819091e-06,
+      "loss": 0.0918,
+      "step": 505
+    },
+    {
+      "epoch": 0.22473906284699088,
+      "grad_norm": 1.0346342912485391,
+      "learning_rate": 5.615982241953386e-06,
+      "loss": 0.0983,
+      "step": 506
+    },
+    {
+      "epoch": 0.2251832111925383,
+      "grad_norm": 1.2353611959723767,
+      "learning_rate": 5.62708102108768e-06,
+      "loss": 0.1066,
+      "step": 507
+    },
+    {
+      "epoch": 0.22562735953808571,
+      "grad_norm": 1.5571341771119367,
+      "learning_rate": 5.638179800221976e-06,
+      "loss": 0.1197,
+      "step": 508
+    },
+    {
+      "epoch": 0.22607150788363314,
+      "grad_norm": 1.8593279138829277,
+      "learning_rate": 5.649278579356271e-06,
+      "loss": 0.1539,
+      "step": 509
+    },
+    {
+      "epoch": 0.22651565622918055,
+      "grad_norm": 1.3210670897027246,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.0972,
+      "step": 510
+    },
+    {
+      "epoch": 0.22695980457472795,
+      "grad_norm": 1.555152024387198,
+      "learning_rate": 5.671476137624862e-06,
+      "loss": 0.1017,
+      "step": 511
+    },
+    {
+      "epoch": 0.22740395292027538,
+      "grad_norm": 1.1784666679583122,
+      "learning_rate": 5.682574916759157e-06,
+      "loss": 0.0896,
+      "step": 512
+    },
+    {
+      "epoch": 0.22784810126582278,
+      "grad_norm": 0.897552288214822,
+      "learning_rate": 5.693673695893452e-06,
+      "loss": 0.0748,
+      "step": 513
+    },
+    {
+      "epoch": 0.2282922496113702,
+      "grad_norm": 0.9984908709528371,
+      "learning_rate": 5.704772475027748e-06,
+      "loss": 0.0962,
+      "step": 514
+    },
+    {
+      "epoch": 0.2287363979569176,
+      "grad_norm": 1.1522533946932125,
+      "learning_rate": 5.715871254162043e-06,
+      "loss": 0.1175,
+      "step": 515
+    },
+    {
+      "epoch": 0.229180546302465,
+      "grad_norm": 1.0543599181977559,
+      "learning_rate": 5.7269700332963374e-06,
+      "loss": 0.1386,
+      "step": 516
+    },
+    {
+      "epoch": 0.22962469464801244,
+      "grad_norm": 0.8788825627877698,
+      "learning_rate": 5.7380688124306325e-06,
+      "loss": 0.0947,
+      "step": 517
+    },
+    {
+      "epoch": 0.23006884299355984,
+      "grad_norm": 0.8916267274537737,
+      "learning_rate": 5.7491675915649284e-06,
+      "loss": 0.0875,
+      "step": 518
+    },
+    {
+      "epoch": 0.23051299133910727,
+      "grad_norm": 1.3741271628041778,
+      "learning_rate": 5.7602663706992235e-06,
+      "loss": 0.1431,
+      "step": 519
+    },
+    {
+      "epoch": 0.23095713968465467,
+      "grad_norm": 1.158157014144474,
+      "learning_rate": 5.7713651498335186e-06,
+      "loss": 0.1079,
+      "step": 520
+    },
+    {
+      "epoch": 0.2314012880302021,
+      "grad_norm": 0.9669483042814996,
+      "learning_rate": 5.7824639289678145e-06,
+      "loss": 0.0756,
+      "step": 521
+    },
+    {
+      "epoch": 0.2318454363757495,
+      "grad_norm": 0.9858439751900236,
+      "learning_rate": 5.7935627081021095e-06,
+      "loss": 0.1091,
+      "step": 522
+    },
+    {
+      "epoch": 0.2322895847212969,
+      "grad_norm": 1.0356934556615558,
+      "learning_rate": 5.804661487236405e-06,
+      "loss": 0.0899,
+      "step": 523
+    },
+    {
+      "epoch": 0.23273373306684433,
+      "grad_norm": 3.900766522495628,
+      "learning_rate": 5.8157602663707005e-06,
+      "loss": 0.0926,
+      "step": 524
+    },
+    {
+      "epoch": 0.23317788141239174,
+      "grad_norm": 1.1976168770257003,
+      "learning_rate": 5.826859045504995e-06,
+      "loss": 0.0963,
+      "step": 525
+    },
+    {
+      "epoch": 0.23362202975793916,
+      "grad_norm": 1.2441372325560522,
+      "learning_rate": 5.83795782463929e-06,
+      "loss": 0.1105,
+      "step": 526
+    },
+    {
+      "epoch": 0.23406617810348657,
+      "grad_norm": 1.379627337877928,
+      "learning_rate": 5.849056603773585e-06,
+      "loss": 0.0851,
+      "step": 527
+    },
+    {
+      "epoch": 0.23451032644903397,
+      "grad_norm": 1.0553207077914868,
+      "learning_rate": 5.860155382907881e-06,
+      "loss": 0.0876,
+      "step": 528
+    },
+    {
+      "epoch": 0.2349544747945814,
+      "grad_norm": 1.0117503811303656,
+      "learning_rate": 5.871254162042176e-06,
+      "loss": 0.0925,
+      "step": 529
+    },
+    {
+      "epoch": 0.2353986231401288,
+      "grad_norm": 1.2155239899129966,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.1177,
+      "step": 530
+    },
+    {
+      "epoch": 0.23584277148567623,
+      "grad_norm": 1.0134886033118073,
+      "learning_rate": 5.893451720310767e-06,
+      "loss": 0.0868,
+      "step": 531
+    },
+    {
+      "epoch": 0.23628691983122363,
+      "grad_norm": 1.2415034220382457,
+      "learning_rate": 5.904550499445062e-06,
+      "loss": 0.1016,
+      "step": 532
+    },
+    {
+      "epoch": 0.23673106817677103,
+      "grad_norm": 0.9266174346376681,
+      "learning_rate": 5.915649278579356e-06,
+      "loss": 0.1001,
+      "step": 533
+    },
+    {
+      "epoch": 0.23717521652231846,
+      "grad_norm": 1.3120341529200943,
+      "learning_rate": 5.926748057713651e-06,
+      "loss": 0.1248,
+      "step": 534
+    },
+    {
+      "epoch": 0.23761936486786586,
+      "grad_norm": 1.426598559638218,
+      "learning_rate": 5.937846836847947e-06,
+      "loss": 0.0932,
+      "step": 535
+    },
+    {
+      "epoch": 0.2380635132134133,
+      "grad_norm": 1.0182331505869047,
+      "learning_rate": 5.948945615982242e-06,
+      "loss": 0.1009,
+      "step": 536
+    },
+    {
+      "epoch": 0.2385076615589607,
+      "grad_norm": 1.4554113324351339,
+      "learning_rate": 5.960044395116537e-06,
+      "loss": 0.0998,
+      "step": 537
+    },
+    {
+      "epoch": 0.2389518099045081,
+      "grad_norm": 1.6220655384523863,
+      "learning_rate": 5.971143174250833e-06,
+      "loss": 0.105,
+      "step": 538
+    },
+    {
+      "epoch": 0.23939595825005552,
+      "grad_norm": 1.9400783770338148,
+      "learning_rate": 5.982241953385128e-06,
+      "loss": 0.1524,
+      "step": 539
+    },
+    {
+      "epoch": 0.23984010659560293,
+      "grad_norm": 1.4635091103881346,
+      "learning_rate": 5.993340732519424e-06,
+      "loss": 0.1233,
+      "step": 540
+    },
+    {
+      "epoch": 0.24028425494115035,
+      "grad_norm": 1.2937114895643234,
+      "learning_rate": 6.004439511653719e-06,
+      "loss": 0.1169,
+      "step": 541
+    },
+    {
+      "epoch": 0.24072840328669776,
+      "grad_norm": 1.2221129883336008,
+      "learning_rate": 6.015538290788013e-06,
+      "loss": 0.1049,
+      "step": 542
+    },
+    {
+      "epoch": 0.24117255163224516,
+      "grad_norm": 1.0735565355706922,
+      "learning_rate": 6.0266370699223085e-06,
+      "loss": 0.0886,
+      "step": 543
+    },
+    {
+      "epoch": 0.2416166999777926,
+      "grad_norm": 1.3261214209776406,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.1245,
+      "step": 544
+    },
+    {
+      "epoch": 0.24206084832334,
+      "grad_norm": 1.2122447330179282,
+      "learning_rate": 6.0488346281908995e-06,
+      "loss": 0.0824,
+      "step": 545
+    },
+    {
+      "epoch": 0.24250499666888742,
+      "grad_norm": 1.1700065718816541,
+      "learning_rate": 6.0599334073251945e-06,
+      "loss": 0.1062,
+      "step": 546
+    },
+    {
+      "epoch": 0.24294914501443482,
+      "grad_norm": 1.0126080671178852,
+      "learning_rate": 6.0710321864594904e-06,
+      "loss": 0.0876,
+      "step": 547
+    },
+    {
+      "epoch": 0.24339329335998222,
+      "grad_norm": 1.185450234336463,
+      "learning_rate": 6.0821309655937855e-06,
+      "loss": 0.1057,
+      "step": 548
+    },
+    {
+      "epoch": 0.24383744170552965,
+      "grad_norm": 0.94702847545811,
+      "learning_rate": 6.0932297447280806e-06,
+      "loss": 0.0704,
+      "step": 549
+    },
+    {
+      "epoch": 0.24428159005107705,
+      "grad_norm": 1.1538838071376252,
+      "learning_rate": 6.1043285238623765e-06,
+      "loss": 0.1062,
+      "step": 550
+    },
+    {
+      "epoch": 0.24472573839662448,
+      "grad_norm": 1.2787034775849095,
+      "learning_rate": 6.115427302996671e-06,
+      "loss": 0.1053,
+      "step": 551
+    },
+    {
+      "epoch": 0.24516988674217188,
+      "grad_norm": 2.111386576675479,
+      "learning_rate": 6.126526082130966e-06,
+      "loss": 0.1423,
+      "step": 552
+    },
+    {
+      "epoch": 0.24561403508771928,
+      "grad_norm": 0.8396851122657492,
+      "learning_rate": 6.137624861265261e-06,
+      "loss": 0.0857,
+      "step": 553
+    },
+    {
+      "epoch": 0.24605818343326671,
+      "grad_norm": 1.222223712186514,
+      "learning_rate": 6.148723640399557e-06,
+      "loss": 0.1241,
+      "step": 554
+    },
+    {
+      "epoch": 0.24650233177881412,
+      "grad_norm": 1.3542921033853719,
+      "learning_rate": 6.159822419533852e-06,
+      "loss": 0.1113,
+      "step": 555
+    },
+    {
+      "epoch": 0.24694648012436154,
+      "grad_norm": 0.921783149485065,
+      "learning_rate": 6.170921198668147e-06,
+      "loss": 0.0954,
+      "step": 556
+    },
+    {
+      "epoch": 0.24739062846990895,
+      "grad_norm": 1.170580173743532,
+      "learning_rate": 6.182019977802443e-06,
+      "loss": 0.0994,
+      "step": 557
+    },
+    {
+      "epoch": 0.24783477681545638,
+      "grad_norm": 1.4092773954441822,
+      "learning_rate": 6.193118756936738e-06,
+      "loss": 0.0931,
+      "step": 558
+    },
+    {
+      "epoch": 0.24827892516100378,
+      "grad_norm": 1.0651408845659296,
+      "learning_rate": 6.204217536071032e-06,
+      "loss": 0.1083,
+      "step": 559
+    },
+    {
+      "epoch": 0.24872307350655118,
+      "grad_norm": 1.2143455853200855,
+      "learning_rate": 6.215316315205327e-06,
+      "loss": 0.0952,
+      "step": 560
+    },
+    {
+      "epoch": 0.2491672218520986,
+      "grad_norm": 2.64991339168012,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 0.1145,
+      "step": 561
+    },
+    {
+      "epoch": 0.249611370197646,
+      "grad_norm": 0.7270989844733312,
+      "learning_rate": 6.237513873473918e-06,
+      "loss": 0.0861,
+      "step": 562
+    },
+    {
+      "epoch": 0.2500555185431934,
+      "grad_norm": 1.1431426285342126,
+      "learning_rate": 6.248612652608213e-06,
+      "loss": 0.1162,
+      "step": 563
+    },
+    {
+      "epoch": 0.2504996668887408,
+      "grad_norm": 0.860711327757446,
+      "learning_rate": 6.259711431742509e-06,
+      "loss": 0.0935,
+      "step": 564
+    },
+    {
+      "epoch": 0.25094381523428827,
+      "grad_norm": 0.8073003696766194,
+      "learning_rate": 6.270810210876804e-06,
+      "loss": 0.1079,
+      "step": 565
+    },
+    {
+      "epoch": 0.25138796357983567,
+      "grad_norm": 1.1101487411387352,
+      "learning_rate": 6.281908990011099e-06,
+      "loss": 0.1174,
+      "step": 566
+    },
+    {
+      "epoch": 0.2518321119253831,
+      "grad_norm": 1.0549286990788267,
+      "learning_rate": 6.293007769145395e-06,
+      "loss": 0.1051,
+      "step": 567
+    },
+    {
+      "epoch": 0.2522762602709305,
+      "grad_norm": 1.2019614422517741,
+      "learning_rate": 6.304106548279689e-06,
+      "loss": 0.1063,
+      "step": 568
+    },
+    {
+      "epoch": 0.2527204086164779,
+      "grad_norm": 0.9611079197173829,
+      "learning_rate": 6.3152053274139845e-06,
+      "loss": 0.0974,
+      "step": 569
+    },
+    {
+      "epoch": 0.25316455696202533,
+      "grad_norm": 1.1961495715533894,
+      "learning_rate": 6.32630410654828e-06,
+      "loss": 0.0991,
+      "step": 570
+    },
+    {
+      "epoch": 0.25360870530757273,
+      "grad_norm": 1.0576689858819426,
+      "learning_rate": 6.3374028856825754e-06,
+      "loss": 0.0953,
+      "step": 571
+    },
+    {
+      "epoch": 0.25405285365312014,
+      "grad_norm": 1.2656186745282128,
+      "learning_rate": 6.3485016648168705e-06,
+      "loss": 0.1083,
+      "step": 572
+    },
+    {
+      "epoch": 0.25449700199866754,
+      "grad_norm": 1.1776166049192853,
+      "learning_rate": 6.359600443951166e-06,
+      "loss": 0.1118,
+      "step": 573
+    },
+    {
+      "epoch": 0.254941150344215,
+      "grad_norm": 0.8664088781289958,
+      "learning_rate": 6.3706992230854615e-06,
+      "loss": 0.0863,
+      "step": 574
+    },
+    {
+      "epoch": 0.2553852986897624,
+      "grad_norm": 0.9925662850080711,
+      "learning_rate": 6.3817980022197565e-06,
+      "loss": 0.1058,
+      "step": 575
+    },
+    {
+      "epoch": 0.2558294470353098,
+      "grad_norm": 1.0606625159824246,
+      "learning_rate": 6.3928967813540525e-06,
+      "loss": 0.1075,
+      "step": 576
+    },
+    {
+      "epoch": 0.2562735953808572,
+      "grad_norm": 0.9900660508183841,
+      "learning_rate": 6.403995560488347e-06,
+      "loss": 0.0829,
+      "step": 577
+    },
+    {
+      "epoch": 0.2567177437264046,
+      "grad_norm": 0.8582749721826495,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.0914,
+      "step": 578
+    },
+    {
+      "epoch": 0.25716189207195206,
+      "grad_norm": 1.1617033243365338,
+      "learning_rate": 6.426193118756937e-06,
+      "loss": 0.1116,
+      "step": 579
+    },
+    {
+      "epoch": 0.25760604041749946,
+      "grad_norm": 1.0362694278635598,
+      "learning_rate": 6.437291897891233e-06,
+      "loss": 0.1145,
+      "step": 580
+    },
+    {
+      "epoch": 0.25805018876304686,
+      "grad_norm": 0.8723418719473582,
+      "learning_rate": 6.448390677025528e-06,
+      "loss": 0.0885,
+      "step": 581
+    },
+    {
+      "epoch": 0.25849433710859426,
+      "grad_norm": 1.0870215330646826,
+      "learning_rate": 6.459489456159823e-06,
+      "loss": 0.0894,
+      "step": 582
+    },
+    {
+      "epoch": 0.25893848545414166,
+      "grad_norm": 1.2407360085708667,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.0812,
+      "step": 583
+    },
+    {
+      "epoch": 0.2593826337996891,
+      "grad_norm": 1.8006078173259896,
+      "learning_rate": 6.481687014428414e-06,
+      "loss": 0.1429,
+      "step": 584
+    },
+    {
+      "epoch": 0.2598267821452365,
+      "grad_norm": 1.4748519282887915,
+      "learning_rate": 6.492785793562708e-06,
+      "loss": 0.0867,
+      "step": 585
+    },
+    {
+      "epoch": 0.2602709304907839,
+      "grad_norm": 1.0424639232757755,
+      "learning_rate": 6.503884572697003e-06,
+      "loss": 0.1146,
+      "step": 586
+    },
+    {
+      "epoch": 0.2607150788363313,
+      "grad_norm": 1.2786124769190668,
+      "learning_rate": 6.514983351831299e-06,
+      "loss": 0.1192,
+      "step": 587
+    },
+    {
+      "epoch": 0.26115922718187873,
+      "grad_norm": 0.9288875670879717,
+      "learning_rate": 6.526082130965594e-06,
+      "loss": 0.101,
+      "step": 588
+    },
+    {
+      "epoch": 0.2616033755274262,
+      "grad_norm": 0.6884103445991109,
+      "learning_rate": 6.537180910099889e-06,
+      "loss": 0.0743,
+      "step": 589
+    },
+    {
+      "epoch": 0.2620475238729736,
+      "grad_norm": 1.0193990545252207,
+      "learning_rate": 6.548279689234185e-06,
+      "loss": 0.0826,
+      "step": 590
+    },
+    {
+      "epoch": 0.262491672218521,
+      "grad_norm": 1.2098967104098979,
+      "learning_rate": 6.55937846836848e-06,
+      "loss": 0.116,
+      "step": 591
+    },
+    {
+      "epoch": 0.2629358205640684,
+      "grad_norm": 0.86346939462905,
+      "learning_rate": 6.570477247502775e-06,
+      "loss": 0.0833,
+      "step": 592
+    },
+    {
+      "epoch": 0.2633799689096158,
+      "grad_norm": 0.8577523664034169,
+      "learning_rate": 6.581576026637071e-06,
+      "loss": 0.0802,
+      "step": 593
+    },
+    {
+      "epoch": 0.26382411725516325,
+      "grad_norm": 0.7836334305793592,
+      "learning_rate": 6.592674805771365e-06,
+      "loss": 0.079,
+      "step": 594
+    },
+    {
+      "epoch": 0.26426826560071065,
+      "grad_norm": 0.9292792710410797,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.0841,
+      "step": 595
+    },
+    {
+      "epoch": 0.26471241394625805,
+      "grad_norm": 0.949730314196452,
+      "learning_rate": 6.614872364039956e-06,
+      "loss": 0.0779,
+      "step": 596
+    },
+    {
+      "epoch": 0.26515656229180545,
+      "grad_norm": 1.067206699117055,
+      "learning_rate": 6.625971143174251e-06,
+      "loss": 0.0813,
+      "step": 597
+    },
+    {
+      "epoch": 0.26560071063735285,
+      "grad_norm": 1.11737459145968,
+      "learning_rate": 6.6370699223085465e-06,
+      "loss": 0.0855,
+      "step": 598
+    },
+    {
+      "epoch": 0.2660448589829003,
+      "grad_norm": 1.3191268336017996,
+      "learning_rate": 6.648168701442842e-06,
+      "loss": 0.117,
+      "step": 599
+    },
+    {
+      "epoch": 0.2664890073284477,
+      "grad_norm": 1.0102499350799081,
+      "learning_rate": 6.6592674805771374e-06,
+      "loss": 0.0878,
+      "step": 600
+    },
+    {
+      "epoch": 0.2669331556739951,
+      "grad_norm": 0.878524576308141,
+      "learning_rate": 6.6703662597114325e-06,
+      "loss": 0.0886,
+      "step": 601
+    },
+    {
+      "epoch": 0.2673773040195425,
+      "grad_norm": 1.0385994736274717,
+      "learning_rate": 6.6814650388457284e-06,
+      "loss": 0.0895,
+      "step": 602
+    },
+    {
+      "epoch": 0.2678214523650899,
+      "grad_norm": 2.1746856925927305,
+      "learning_rate": 6.692563817980023e-06,
+      "loss": 0.1218,
+      "step": 603
+    },
+    {
+      "epoch": 0.2682656007106374,
+      "grad_norm": 1.1284346260514344,
+      "learning_rate": 6.703662597114318e-06,
+      "loss": 0.0984,
+      "step": 604
+    },
+    {
+      "epoch": 0.2687097490561848,
+      "grad_norm": 0.8795466679323811,
+      "learning_rate": 6.714761376248613e-06,
+      "loss": 0.1076,
+      "step": 605
+    },
+    {
+      "epoch": 0.2691538974017322,
+      "grad_norm": 0.7889851290279695,
+      "learning_rate": 6.725860155382909e-06,
+      "loss": 0.0809,
+      "step": 606
+    },
+    {
+      "epoch": 0.2695980457472796,
+      "grad_norm": 1.1495901760021123,
+      "learning_rate": 6.736958934517204e-06,
+      "loss": 0.1339,
+      "step": 607
+    },
+    {
+      "epoch": 0.270042194092827,
+      "grad_norm": 0.9404581532626506,
+      "learning_rate": 6.748057713651499e-06,
+      "loss": 0.0835,
+      "step": 608
+    },
+    {
+      "epoch": 0.27048634243837444,
+      "grad_norm": 1.1614393549481714,
+      "learning_rate": 6.759156492785795e-06,
+      "loss": 0.1022,
+      "step": 609
+    },
+    {
+      "epoch": 0.27093049078392184,
+      "grad_norm": 0.7805043342922547,
+      "learning_rate": 6.77025527192009e-06,
+      "loss": 0.0872,
+      "step": 610
+    },
+    {
+      "epoch": 0.27137463912946924,
+      "grad_norm": 1.1270320700845156,
+      "learning_rate": 6.781354051054384e-06,
+      "loss": 0.106,
+      "step": 611
+    },
+    {
+      "epoch": 0.27181878747501664,
+      "grad_norm": 0.8511881872825724,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.077,
+      "step": 612
+    },
+    {
+      "epoch": 0.27226293582056404,
+      "grad_norm": 0.8653445546881817,
+      "learning_rate": 6.803551609322975e-06,
+      "loss": 0.0703,
+      "step": 613
+    },
+    {
+      "epoch": 0.2727070841661115,
+      "grad_norm": 0.9249325812470924,
+      "learning_rate": 6.81465038845727e-06,
+      "loss": 0.1239,
+      "step": 614
+    },
+    {
+      "epoch": 0.2731512325116589,
+      "grad_norm": 1.2035733700107387,
+      "learning_rate": 6.825749167591565e-06,
+      "loss": 0.0803,
+      "step": 615
+    },
+    {
+      "epoch": 0.2735953808572063,
+      "grad_norm": 1.058685859950284,
+      "learning_rate": 6.836847946725861e-06,
+      "loss": 0.101,
+      "step": 616
+    },
+    {
+      "epoch": 0.2740395292027537,
+      "grad_norm": 0.9291079298055421,
+      "learning_rate": 6.847946725860156e-06,
+      "loss": 0.0907,
+      "step": 617
+    },
+    {
+      "epoch": 0.2744836775483011,
+      "grad_norm": 1.1511507883949077,
+      "learning_rate": 6.859045504994451e-06,
+      "loss": 0.1077,
+      "step": 618
+    },
+    {
+      "epoch": 0.27492782589384857,
+      "grad_norm": 1.0786240033108858,
+      "learning_rate": 6.870144284128747e-06,
+      "loss": 0.1251,
+      "step": 619
+    },
+    {
+      "epoch": 0.27537197423939597,
+      "grad_norm": 0.9285679879223824,
+      "learning_rate": 6.881243063263041e-06,
+      "loss": 0.079,
+      "step": 620
+    },
+    {
+      "epoch": 0.27581612258494337,
+      "grad_norm": 0.8117578679372189,
+      "learning_rate": 6.892341842397336e-06,
+      "loss": 0.0772,
+      "step": 621
+    },
+    {
+      "epoch": 0.27626027093049077,
+      "grad_norm": 1.0138147773665356,
+      "learning_rate": 6.9034406215316315e-06,
+      "loss": 0.1238,
+      "step": 622
+    },
+    {
+      "epoch": 0.27670441927603817,
+      "grad_norm": 0.988212212065645,
+      "learning_rate": 6.914539400665927e-06,
+      "loss": 0.0726,
+      "step": 623
+    },
+    {
+      "epoch": 0.27714856762158563,
+      "grad_norm": 1.2358456192045688,
+      "learning_rate": 6.9256381798002224e-06,
+      "loss": 0.1125,
+      "step": 624
+    },
+    {
+      "epoch": 0.27759271596713303,
+      "grad_norm": 1.2887749392067072,
+      "learning_rate": 6.9367369589345175e-06,
+      "loss": 0.1003,
+      "step": 625
+    },
+    {
+      "epoch": 0.27803686431268043,
+      "grad_norm": 0.9864604958886025,
+      "learning_rate": 6.947835738068813e-06,
+      "loss": 0.0831,
+      "step": 626
+    },
+    {
+      "epoch": 0.27848101265822783,
+      "grad_norm": 1.088865911862852,
+      "learning_rate": 6.9589345172031085e-06,
+      "loss": 0.101,
+      "step": 627
+    },
+    {
+      "epoch": 0.27892516100377523,
+      "grad_norm": 1.409906763744686,
+      "learning_rate": 6.970033296337404e-06,
+      "loss": 0.104,
+      "step": 628
+    },
+    {
+      "epoch": 0.2793693093493227,
+      "grad_norm": 1.2302985416041377,
+      "learning_rate": 6.981132075471699e-06,
+      "loss": 0.0969,
+      "step": 629
+    },
+    {
+      "epoch": 0.2798134576948701,
+      "grad_norm": 1.2642567912432223,
+      "learning_rate": 6.992230854605994e-06,
+      "loss": 0.1116,
+      "step": 630
+    },
+    {
+      "epoch": 0.2802576060404175,
+      "grad_norm": 1.2962548485310728,
+      "learning_rate": 7.003329633740289e-06,
+      "loss": 0.1144,
+      "step": 631
+    },
+    {
+      "epoch": 0.2807017543859649,
+      "grad_norm": 0.9021212064599634,
+      "learning_rate": 7.014428412874585e-06,
+      "loss": 0.0794,
+      "step": 632
+    },
+    {
+      "epoch": 0.2811459027315123,
+      "grad_norm": 0.9822662924030562,
+      "learning_rate": 7.02552719200888e-06,
+      "loss": 0.081,
+      "step": 633
+    },
+    {
+      "epoch": 0.28159005107705976,
+      "grad_norm": 0.8568051773621045,
+      "learning_rate": 7.036625971143175e-06,
+      "loss": 0.0838,
+      "step": 634
+    },
+    {
+      "epoch": 0.28203419942260716,
+      "grad_norm": 0.9134125872525329,
+      "learning_rate": 7.047724750277471e-06,
+      "loss": 0.0895,
+      "step": 635
+    },
+    {
+      "epoch": 0.28247834776815456,
+      "grad_norm": 1.0203574319172648,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.0872,
+      "step": 636
+    },
+    {
+      "epoch": 0.28292249611370196,
+      "grad_norm": 0.8320232571139462,
+      "learning_rate": 7.06992230854606e-06,
+      "loss": 0.0778,
+      "step": 637
+    },
+    {
+      "epoch": 0.28336664445924936,
+      "grad_norm": 1.1019251570363913,
+      "learning_rate": 7.081021087680355e-06,
+      "loss": 0.101,
+      "step": 638
+    },
+    {
+      "epoch": 0.2838107928047968,
+      "grad_norm": 0.9010932622870097,
+      "learning_rate": 7.092119866814651e-06,
+      "loss": 0.0852,
+      "step": 639
+    },
+    {
+      "epoch": 0.2842549411503442,
+      "grad_norm": 1.0180000788163226,
+      "learning_rate": 7.103218645948946e-06,
+      "loss": 0.0955,
+      "step": 640
+    },
+    {
+      "epoch": 0.2846990894958916,
+      "grad_norm": 0.9251835556617967,
+      "learning_rate": 7.114317425083241e-06,
+      "loss": 0.0871,
+      "step": 641
+    },
+    {
+      "epoch": 0.285143237841439,
+      "grad_norm": 0.8901698566560228,
+      "learning_rate": 7.125416204217537e-06,
+      "loss": 0.0706,
+      "step": 642
+    },
+    {
+      "epoch": 0.2855873861869865,
+      "grad_norm": 1.0154369539307455,
+      "learning_rate": 7.136514983351832e-06,
+      "loss": 0.0887,
+      "step": 643
+    },
+    {
+      "epoch": 0.2860315345325339,
+      "grad_norm": 1.321020159204264,
+      "learning_rate": 7.147613762486127e-06,
+      "loss": 0.0865,
+      "step": 644
+    },
+    {
+      "epoch": 0.2864756828780813,
+      "grad_norm": 2.535826739420158,
+      "learning_rate": 7.158712541620423e-06,
+      "loss": 0.1096,
+      "step": 645
+    },
+    {
+      "epoch": 0.2869198312236287,
+      "grad_norm": 1.0227251731125295,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.0977,
+      "step": 646
+    },
+    {
+      "epoch": 0.2873639795691761,
+      "grad_norm": 0.8089367869895305,
+      "learning_rate": 7.180910099889012e-06,
+      "loss": 0.0792,
+      "step": 647
+    },
+    {
+      "epoch": 0.28780812791472354,
+      "grad_norm": 0.9374423150392824,
+      "learning_rate": 7.1920088790233074e-06,
+      "loss": 0.1055,
+      "step": 648
+    },
+    {
+      "epoch": 0.28825227626027095,
+      "grad_norm": 1.07643087941822,
+      "learning_rate": 7.203107658157603e-06,
+      "loss": 0.0957,
+      "step": 649
+    },
+    {
+      "epoch": 0.28869642460581835,
+      "grad_norm": 0.9052355958500774,
+      "learning_rate": 7.214206437291898e-06,
+      "loss": 0.097,
+      "step": 650
+    },
+    {
+      "epoch": 0.28914057295136575,
+      "grad_norm": 1.1103580563722126,
+      "learning_rate": 7.2253052164261935e-06,
+      "loss": 0.1111,
+      "step": 651
+    },
+    {
+      "epoch": 0.28958472129691315,
+      "grad_norm": 1.7388144567675092,
+      "learning_rate": 7.236403995560489e-06,
+      "loss": 0.0939,
+      "step": 652
+    },
+    {
+      "epoch": 0.2900288696424606,
+      "grad_norm": 1.078770189866491,
+      "learning_rate": 7.2475027746947845e-06,
+      "loss": 0.0845,
+      "step": 653
+    },
+    {
+      "epoch": 0.290473017988008,
+      "grad_norm": 1.2137588838688103,
+      "learning_rate": 7.2586015538290795e-06,
+      "loss": 0.0868,
+      "step": 654
+    },
+    {
+      "epoch": 0.2909171663335554,
+      "grad_norm": 1.3727496776393755,
+      "learning_rate": 7.269700332963375e-06,
+      "loss": 0.0922,
+      "step": 655
+    },
+    {
+      "epoch": 0.2913613146791028,
+      "grad_norm": 1.2519592442230572,
+      "learning_rate": 7.28079911209767e-06,
+      "loss": 0.126,
+      "step": 656
+    },
+    {
+      "epoch": 0.2918054630246502,
+      "grad_norm": 0.8673663222128764,
+      "learning_rate": 7.291897891231965e-06,
+      "loss": 0.0778,
+      "step": 657
+    },
+    {
+      "epoch": 0.29224961137019767,
+      "grad_norm": 1.1245127772074948,
+      "learning_rate": 7.302996670366261e-06,
+      "loss": 0.094,
+      "step": 658
+    },
+    {
+      "epoch": 0.29269375971574507,
+      "grad_norm": 1.156956668230168,
+      "learning_rate": 7.314095449500556e-06,
+      "loss": 0.108,
+      "step": 659
+    },
+    {
+      "epoch": 0.2931379080612925,
+      "grad_norm": 0.8322393629574484,
+      "learning_rate": 7.325194228634851e-06,
+      "loss": 0.0933,
+      "step": 660
+    },
+    {
+      "epoch": 0.2935820564068399,
+      "grad_norm": 1.1303584309903236,
+      "learning_rate": 7.336293007769147e-06,
+      "loss": 0.1144,
+      "step": 661
+    },
+    {
+      "epoch": 0.2940262047523873,
+      "grad_norm": 1.1516682415849495,
+      "learning_rate": 7.347391786903442e-06,
+      "loss": 0.1161,
+      "step": 662
+    },
+    {
+      "epoch": 0.29447035309793473,
+      "grad_norm": 0.6573879684754841,
+      "learning_rate": 7.358490566037736e-06,
+      "loss": 0.0743,
+      "step": 663
+    },
+    {
+      "epoch": 0.29491450144348214,
+      "grad_norm": 0.7675390269253073,
+      "learning_rate": 7.369589345172031e-06,
+      "loss": 0.0733,
+      "step": 664
+    },
+    {
+      "epoch": 0.29535864978902954,
+      "grad_norm": 0.9852104843013336,
+      "learning_rate": 7.380688124306327e-06,
+      "loss": 0.1057,
+      "step": 665
+    },
+    {
+      "epoch": 0.29580279813457694,
+      "grad_norm": 1.3140910564234118,
+      "learning_rate": 7.391786903440622e-06,
+      "loss": 0.1013,
+      "step": 666
+    },
+    {
+      "epoch": 0.29624694648012434,
+      "grad_norm": 1.1588394898060206,
+      "learning_rate": 7.402885682574917e-06,
+      "loss": 0.1089,
+      "step": 667
+    },
+    {
+      "epoch": 0.2966910948256718,
+      "grad_norm": 1.1857494935699013,
+      "learning_rate": 7.413984461709213e-06,
+      "loss": 0.0865,
+      "step": 668
+    },
+    {
+      "epoch": 0.2971352431712192,
+      "grad_norm": 0.861410639016016,
+      "learning_rate": 7.425083240843508e-06,
+      "loss": 0.0771,
+      "step": 669
+    },
+    {
+      "epoch": 0.2975793915167666,
+      "grad_norm": 1.293845000506648,
+      "learning_rate": 7.436182019977803e-06,
+      "loss": 0.1212,
+      "step": 670
+    },
+    {
+      "epoch": 0.298023539862314,
+      "grad_norm": 1.0407435776743181,
+      "learning_rate": 7.447280799112099e-06,
+      "loss": 0.0948,
+      "step": 671
+    },
+    {
+      "epoch": 0.2984676882078614,
+      "grad_norm": 0.9347120808526156,
+      "learning_rate": 7.458379578246393e-06,
+      "loss": 0.0779,
+      "step": 672
+    },
+    {
+      "epoch": 0.29891183655340886,
+      "grad_norm": 0.9205867276903889,
+      "learning_rate": 7.469478357380688e-06,
+      "loss": 0.0838,
+      "step": 673
+    },
+    {
+      "epoch": 0.29935598489895626,
+      "grad_norm": 1.388402805970494,
+      "learning_rate": 7.480577136514983e-06,
+      "loss": 0.0924,
+      "step": 674
+    },
+    {
+      "epoch": 0.29980013324450366,
+      "grad_norm": 1.6172485445367795,
+      "learning_rate": 7.491675915649279e-06,
+      "loss": 0.1132,
+      "step": 675
+    },
+    {
+      "epoch": 0.30024428159005107,
+      "grad_norm": 0.9561370366379224,
+      "learning_rate": 7.502774694783574e-06,
+      "loss": 0.1032,
+      "step": 676
+    },
+    {
+      "epoch": 0.30068842993559847,
+      "grad_norm": 1.0171256621828135,
+      "learning_rate": 7.5138734739178694e-06,
+      "loss": 0.0763,
+      "step": 677
+    },
+    {
+      "epoch": 0.3011325782811459,
+      "grad_norm": 1.0399008828044272,
+      "learning_rate": 7.524972253052165e-06,
+      "loss": 0.0763,
+      "step": 678
+    },
+    {
+      "epoch": 0.3015767266266933,
+      "grad_norm": 1.433840414738971,
+      "learning_rate": 7.5360710321864604e-06,
+      "loss": 0.1028,
+      "step": 679
+    },
+    {
+      "epoch": 0.3020208749722407,
+      "grad_norm": 1.0023463068044798,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.0899,
+      "step": 680
+    },
+    {
+      "epoch": 0.30246502331778813,
+      "grad_norm": 0.8983226693149738,
+      "learning_rate": 7.55826859045505e-06,
+      "loss": 0.0906,
+      "step": 681
+    },
+    {
+      "epoch": 0.30290917166333553,
+      "grad_norm": 1.0458814923796536,
+      "learning_rate": 7.569367369589346e-06,
+      "loss": 0.087,
+      "step": 682
+    },
+    {
+      "epoch": 0.303353320008883,
+      "grad_norm": 1.317009232453199,
+      "learning_rate": 7.580466148723641e-06,
+      "loss": 0.1185,
+      "step": 683
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.9515944992821098,
+      "learning_rate": 7.591564927857937e-06,
+      "loss": 0.0784,
+      "step": 684
+    },
+    {
+      "epoch": 0.3042416166999778,
+      "grad_norm": 0.8794173564696354,
+      "learning_rate": 7.602663706992232e-06,
+      "loss": 0.0689,
+      "step": 685
+    },
+    {
+      "epoch": 0.3046857650455252,
+      "grad_norm": 1.3164826313969946,
+      "learning_rate": 7.613762486126527e-06,
+      "loss": 0.109,
+      "step": 686
+    },
+    {
+      "epoch": 0.3051299133910726,
+      "grad_norm": 1.208021624972249,
+      "learning_rate": 7.624861265260823e-06,
+      "loss": 0.1006,
+      "step": 687
+    },
+    {
+      "epoch": 0.30557406173662005,
+      "grad_norm": 0.9049457634328961,
+      "learning_rate": 7.635960044395118e-06,
+      "loss": 0.0898,
+      "step": 688
+    },
+    {
+      "epoch": 0.30601821008216745,
+      "grad_norm": 1.715228836512271,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.0954,
+      "step": 689
+    },
+    {
+      "epoch": 0.30646235842771485,
+      "grad_norm": 0.8243753589934399,
+      "learning_rate": 7.658157602663708e-06,
+      "loss": 0.0775,
+      "step": 690
+    },
+    {
+      "epoch": 0.30690650677326226,
+      "grad_norm": 1.322320542517841,
+      "learning_rate": 7.669256381798003e-06,
+      "loss": 0.1016,
+      "step": 691
+    },
+    {
+      "epoch": 0.30735065511880966,
+      "grad_norm": 0.9618609396789024,
+      "learning_rate": 7.680355160932298e-06,
+      "loss": 0.0973,
+      "step": 692
+    },
+    {
+      "epoch": 0.3077948034643571,
+      "grad_norm": 0.9164782000859841,
+      "learning_rate": 7.691453940066593e-06,
+      "loss": 0.0625,
+      "step": 693
+    },
+    {
+      "epoch": 0.3082389518099045,
+      "grad_norm": 0.9162081625393108,
+      "learning_rate": 7.702552719200888e-06,
+      "loss": 0.089,
+      "step": 694
+    },
+    {
+      "epoch": 0.3086831001554519,
+      "grad_norm": 1.0274536861399304,
+      "learning_rate": 7.713651498335183e-06,
+      "loss": 0.0938,
+      "step": 695
+    },
+    {
+      "epoch": 0.3091272485009993,
+      "grad_norm": 1.2217184648988348,
+      "learning_rate": 7.72475027746948e-06,
+      "loss": 0.0944,
+      "step": 696
+    },
+    {
+      "epoch": 0.3095713968465467,
+      "grad_norm": 1.2610843560867633,
+      "learning_rate": 7.735849056603775e-06,
+      "loss": 0.1108,
+      "step": 697
+    },
+    {
+      "epoch": 0.3100155451920942,
+      "grad_norm": 0.9157424175535596,
+      "learning_rate": 7.746947835738068e-06,
+      "loss": 0.0955,
+      "step": 698
+    },
+    {
+      "epoch": 0.3104596935376416,
+      "grad_norm": 0.8026621451546831,
+      "learning_rate": 7.758046614872365e-06,
+      "loss": 0.0659,
+      "step": 699
+    },
+    {
+      "epoch": 0.310903841883189,
+      "grad_norm": 0.8085721007510808,
+      "learning_rate": 7.76914539400666e-06,
+      "loss": 0.0677,
+      "step": 700
+    },
+    {
+      "epoch": 0.3113479902287364,
+      "grad_norm": 0.9769745515486552,
+      "learning_rate": 7.780244173140955e-06,
+      "loss": 0.0896,
+      "step": 701
+    },
+    {
+      "epoch": 0.3117921385742838,
+      "grad_norm": 1.0969240019260509,
+      "learning_rate": 7.79134295227525e-06,
+      "loss": 0.1038,
+      "step": 702
+    },
+    {
+      "epoch": 0.31223628691983124,
+      "grad_norm": 0.6785759710369018,
+      "learning_rate": 7.802441731409545e-06,
+      "loss": 0.0559,
+      "step": 703
+    },
+    {
+      "epoch": 0.31268043526537864,
+      "grad_norm": 0.9042685928894448,
+      "learning_rate": 7.81354051054384e-06,
+      "loss": 0.0843,
+      "step": 704
+    },
+    {
+      "epoch": 0.31312458361092604,
+      "grad_norm": 0.8411568659072639,
+      "learning_rate": 7.824639289678137e-06,
+      "loss": 0.1247,
+      "step": 705
+    },
+    {
+      "epoch": 0.31356873195647345,
+      "grad_norm": 0.9154910517915918,
+      "learning_rate": 7.835738068812432e-06,
+      "loss": 0.0964,
+      "step": 706
+    },
+    {
+      "epoch": 0.3140128803020209,
+      "grad_norm": 1.31944775400599,
+      "learning_rate": 7.846836847946726e-06,
+      "loss": 0.0986,
+      "step": 707
+    },
+    {
+      "epoch": 0.3144570286475683,
+      "grad_norm": 1.0518670866955138,
+      "learning_rate": 7.85793562708102e-06,
+      "loss": 0.095,
+      "step": 708
+    },
+    {
+      "epoch": 0.3149011769931157,
+      "grad_norm": 1.580342102818051,
+      "learning_rate": 7.869034406215318e-06,
+      "loss": 0.0979,
+      "step": 709
+    },
+    {
+      "epoch": 0.3153453253386631,
+      "grad_norm": 0.9955202031527373,
+      "learning_rate": 7.880133185349613e-06,
+      "loss": 0.0941,
+      "step": 710
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.9719600353134535,
+      "learning_rate": 7.891231964483908e-06,
+      "loss": 0.0929,
+      "step": 711
+    },
+    {
+      "epoch": 0.31623362202975797,
+      "grad_norm": 1.1433817316569035,
+      "learning_rate": 7.902330743618203e-06,
+      "loss": 0.0771,
+      "step": 712
+    },
+    {
+      "epoch": 0.31667777037530537,
+      "grad_norm": 1.3256882881979881,
+      "learning_rate": 7.913429522752498e-06,
+      "loss": 0.0926,
+      "step": 713
+    },
+    {
+      "epoch": 0.31712191872085277,
+      "grad_norm": 1.0742097800389574,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.111,
+      "step": 714
+    },
+    {
+      "epoch": 0.31756606706640017,
+      "grad_norm": 0.8425793522673473,
+      "learning_rate": 7.935627081021088e-06,
+      "loss": 0.08,
+      "step": 715
+    },
+    {
+      "epoch": 0.31801021541194757,
+      "grad_norm": 1.2662850654524707,
+      "learning_rate": 7.946725860155383e-06,
+      "loss": 0.1285,
+      "step": 716
+    },
+    {
+      "epoch": 0.31845436375749503,
+      "grad_norm": 1.0639958089530621,
+      "learning_rate": 7.957824639289678e-06,
+      "loss": 0.0911,
+      "step": 717
+    },
+    {
+      "epoch": 0.31889851210304243,
+      "grad_norm": 0.9354143320657179,
+      "learning_rate": 7.968923418423973e-06,
+      "loss": 0.0937,
+      "step": 718
+    },
+    {
+      "epoch": 0.31934266044858983,
+      "grad_norm": 0.864109073064804,
+      "learning_rate": 7.98002219755827e-06,
+      "loss": 0.07,
+      "step": 719
+    },
+    {
+      "epoch": 0.31978680879413723,
+      "grad_norm": 0.9458708669966496,
+      "learning_rate": 7.991120976692565e-06,
+      "loss": 0.0906,
+      "step": 720
+    },
+    {
+      "epoch": 0.32023095713968464,
+      "grad_norm": 0.9008487720151787,
+      "learning_rate": 8.00221975582686e-06,
+      "loss": 0.0707,
+      "step": 721
+    },
+    {
+      "epoch": 0.3206751054852321,
+      "grad_norm": 1.395224062553485,
+      "learning_rate": 8.013318534961155e-06,
+      "loss": 0.1097,
+      "step": 722
+    },
+    {
+      "epoch": 0.3211192538307795,
+      "grad_norm": 1.1228102547899144,
+      "learning_rate": 8.02441731409545e-06,
+      "loss": 0.103,
+      "step": 723
+    },
+    {
+      "epoch": 0.3215634021763269,
+      "grad_norm": 0.8489026661679754,
+      "learning_rate": 8.035516093229745e-06,
+      "loss": 0.0737,
+      "step": 724
+    },
+    {
+      "epoch": 0.3220075505218743,
+      "grad_norm": 1.116811038077338,
+      "learning_rate": 8.04661487236404e-06,
+      "loss": 0.0861,
+      "step": 725
+    },
+    {
+      "epoch": 0.3224516988674217,
+      "grad_norm": 1.29632952775896,
+      "learning_rate": 8.057713651498335e-06,
+      "loss": 0.1178,
+      "step": 726
+    },
+    {
+      "epoch": 0.32289584721296916,
+      "grad_norm": 0.9812002112145898,
+      "learning_rate": 8.06881243063263e-06,
+      "loss": 0.0832,
+      "step": 727
+    },
+    {
+      "epoch": 0.32333999555851656,
+      "grad_norm": 1.4524326080678385,
+      "learning_rate": 8.079911209766927e-06,
+      "loss": 0.1157,
+      "step": 728
+    },
+    {
+      "epoch": 0.32378414390406396,
+      "grad_norm": 1.004750311195787,
+      "learning_rate": 8.091009988901222e-06,
+      "loss": 0.0817,
+      "step": 729
+    },
+    {
+      "epoch": 0.32422829224961136,
+      "grad_norm": 0.9522750516409738,
+      "learning_rate": 8.102108768035517e-06,
+      "loss": 0.0731,
+      "step": 730
+    },
+    {
+      "epoch": 0.32467244059515876,
+      "grad_norm": 1.0736358930231333,
+      "learning_rate": 8.113207547169812e-06,
+      "loss": 0.0994,
+      "step": 731
+    },
+    {
+      "epoch": 0.3251165889407062,
+      "grad_norm": 1.8191859683022853,
+      "learning_rate": 8.124306326304107e-06,
+      "loss": 0.0937,
+      "step": 732
+    },
+    {
+      "epoch": 0.3255607372862536,
+      "grad_norm": 1.5487234761111215,
+      "learning_rate": 8.135405105438403e-06,
+      "loss": 0.094,
+      "step": 733
+    },
+    {
+      "epoch": 0.326004885631801,
+      "grad_norm": 1.0058113971822433,
+      "learning_rate": 8.146503884572698e-06,
+      "loss": 0.0887,
+      "step": 734
+    },
+    {
+      "epoch": 0.3264490339773484,
+      "grad_norm": 2.7718590597147963,
+      "learning_rate": 8.157602663706993e-06,
+      "loss": 0.1335,
+      "step": 735
+    },
+    {
+      "epoch": 0.3268931823228958,
+      "grad_norm": 2.05367085918422,
+      "learning_rate": 8.168701442841288e-06,
+      "loss": 0.1106,
+      "step": 736
+    },
+    {
+      "epoch": 0.3273373306684433,
+      "grad_norm": 1.115500020772734,
+      "learning_rate": 8.179800221975583e-06,
+      "loss": 0.0722,
+      "step": 737
+    },
+    {
+      "epoch": 0.3277814790139907,
+      "grad_norm": 1.1750795388816762,
+      "learning_rate": 8.19089900110988e-06,
+      "loss": 0.0625,
+      "step": 738
+    },
+    {
+      "epoch": 0.3282256273595381,
+      "grad_norm": 0.9103834127535441,
+      "learning_rate": 8.201997780244175e-06,
+      "loss": 0.0796,
+      "step": 739
+    },
+    {
+      "epoch": 0.3286697757050855,
+      "grad_norm": 0.9850210452729228,
+      "learning_rate": 8.21309655937847e-06,
+      "loss": 0.0658,
+      "step": 740
+    },
+    {
+      "epoch": 0.3291139240506329,
+      "grad_norm": 0.599291679194688,
+      "learning_rate": 8.224195338512763e-06,
+      "loss": 0.0528,
+      "step": 741
+    },
+    {
+      "epoch": 0.32955807239618035,
+      "grad_norm": 1.1762184481853932,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.101,
+      "step": 742
+    },
+    {
+      "epoch": 0.33000222074172775,
+      "grad_norm": 1.4905112857899239,
+      "learning_rate": 8.246392896781355e-06,
+      "loss": 0.1072,
+      "step": 743
+    },
+    {
+      "epoch": 0.33044636908727515,
+      "grad_norm": 0.9022874687789386,
+      "learning_rate": 8.25749167591565e-06,
+      "loss": 0.0948,
+      "step": 744
+    },
+    {
+      "epoch": 0.33089051743282255,
+      "grad_norm": 0.7870509929638926,
+      "learning_rate": 8.268590455049945e-06,
+      "loss": 0.0968,
+      "step": 745
+    },
+    {
+      "epoch": 0.33133466577836995,
+      "grad_norm": 0.9927750763334663,
+      "learning_rate": 8.27968923418424e-06,
+      "loss": 0.0944,
+      "step": 746
+    },
+    {
+      "epoch": 0.3317788141239174,
+      "grad_norm": 0.6875980968846429,
+      "learning_rate": 8.290788013318535e-06,
+      "loss": 0.0636,
+      "step": 747
+    },
+    {
+      "epoch": 0.3322229624694648,
+      "grad_norm": 0.8442026433419401,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.0831,
+      "step": 748
+    },
+    {
+      "epoch": 0.3326671108150122,
+      "grad_norm": 0.9412453312595589,
+      "learning_rate": 8.312985571587127e-06,
+      "loss": 0.0776,
+      "step": 749
+    },
+    {
+      "epoch": 0.3331112591605596,
+      "grad_norm": 0.9364463887563221,
+      "learning_rate": 8.32408435072142e-06,
+      "loss": 0.0887,
+      "step": 750
+    },
+    {
+      "epoch": 0.333555407506107,
+      "grad_norm": 0.7227541011442813,
+      "learning_rate": 8.335183129855715e-06,
+      "loss": 0.0857,
+      "step": 751
+    },
+    {
+      "epoch": 0.3339995558516545,
+      "grad_norm": 0.9456038972922441,
+      "learning_rate": 8.346281908990012e-06,
+      "loss": 0.0831,
+      "step": 752
+    },
+    {
+      "epoch": 0.3344437041972019,
+      "grad_norm": 0.8779308403041339,
+      "learning_rate": 8.357380688124307e-06,
+      "loss": 0.1193,
+      "step": 753
+    },
+    {
+      "epoch": 0.3348878525427493,
+      "grad_norm": 0.9670324236179916,
+      "learning_rate": 8.368479467258602e-06,
+      "loss": 0.0701,
+      "step": 754
+    },
+    {
+      "epoch": 0.3353320008882967,
+      "grad_norm": 0.661653951539699,
+      "learning_rate": 8.379578246392897e-06,
+      "loss": 0.0601,
+      "step": 755
+    },
+    {
+      "epoch": 0.3357761492338441,
+      "grad_norm": 0.7315432781058985,
+      "learning_rate": 8.390677025527192e-06,
+      "loss": 0.0931,
+      "step": 756
+    },
+    {
+      "epoch": 0.33622029757939154,
+      "grad_norm": 0.821857058746102,
+      "learning_rate": 8.40177580466149e-06,
+      "loss": 0.0853,
+      "step": 757
+    },
+    {
+      "epoch": 0.33666444592493894,
+      "grad_norm": 0.8172880203687719,
+      "learning_rate": 8.412874583795784e-06,
+      "loss": 0.1047,
+      "step": 758
+    },
+    {
+      "epoch": 0.33710859427048634,
+      "grad_norm": 1.0390709540622176,
+      "learning_rate": 8.423973362930078e-06,
+      "loss": 0.0954,
+      "step": 759
+    },
+    {
+      "epoch": 0.33755274261603374,
+      "grad_norm": 0.854833081193649,
+      "learning_rate": 8.435072142064373e-06,
+      "loss": 0.07,
+      "step": 760
+    },
+    {
+      "epoch": 0.33799689096158114,
+      "grad_norm": 1.0343320589862304,
+      "learning_rate": 8.44617092119867e-06,
+      "loss": 0.1111,
+      "step": 761
+    },
+    {
+      "epoch": 0.3384410393071286,
+      "grad_norm": 0.9843875498635718,
+      "learning_rate": 8.457269700332965e-06,
+      "loss": 0.0885,
+      "step": 762
+    },
+    {
+      "epoch": 0.338885187652676,
+      "grad_norm": 1.301338158343054,
+      "learning_rate": 8.46836847946726e-06,
+      "loss": 0.0978,
+      "step": 763
+    },
+    {
+      "epoch": 0.3393293359982234,
+      "grad_norm": 1.8891259339303466,
+      "learning_rate": 8.479467258601555e-06,
+      "loss": 0.1068,
+      "step": 764
+    },
+    {
+      "epoch": 0.3397734843437708,
+      "grad_norm": 0.8759783579336403,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.0702,
+      "step": 765
+    },
+    {
+      "epoch": 0.3402176326893182,
+      "grad_norm": 1.0288011027453567,
+      "learning_rate": 8.501664816870145e-06,
+      "loss": 0.1032,
+      "step": 766
+    },
+    {
+      "epoch": 0.34066178103486566,
+      "grad_norm": 1.0389974353449243,
+      "learning_rate": 8.51276359600444e-06,
+      "loss": 0.0876,
+      "step": 767
+    },
+    {
+      "epoch": 0.34110592938041306,
+      "grad_norm": 0.8136721936839324,
+      "learning_rate": 8.523862375138735e-06,
+      "loss": 0.0825,
+      "step": 768
+    },
+    {
+      "epoch": 0.34155007772596047,
+      "grad_norm": 0.7788745832938482,
+      "learning_rate": 8.53496115427303e-06,
+      "loss": 0.0841,
+      "step": 769
+    },
+    {
+      "epoch": 0.34199422607150787,
+      "grad_norm": 0.7903184210262955,
+      "learning_rate": 8.546059933407325e-06,
+      "loss": 0.0798,
+      "step": 770
+    },
+    {
+      "epoch": 0.34243837441705527,
+      "grad_norm": 0.9106664043393771,
+      "learning_rate": 8.557158712541622e-06,
+      "loss": 0.073,
+      "step": 771
+    },
+    {
+      "epoch": 0.3428825227626027,
+      "grad_norm": 1.129345042296519,
+      "learning_rate": 8.568257491675917e-06,
+      "loss": 0.1122,
+      "step": 772
+    },
+    {
+      "epoch": 0.3433266711081501,
+      "grad_norm": 1.0693131682338626,
+      "learning_rate": 8.579356270810212e-06,
+      "loss": 0.0963,
+      "step": 773
+    },
+    {
+      "epoch": 0.34377081945369753,
+      "grad_norm": 0.9775496598799794,
+      "learning_rate": 8.590455049944507e-06,
+      "loss": 0.0917,
+      "step": 774
+    },
+    {
+      "epoch": 0.34421496779924493,
+      "grad_norm": 1.734237792325461,
+      "learning_rate": 8.601553829078802e-06,
+      "loss": 0.0964,
+      "step": 775
+    },
+    {
+      "epoch": 0.3446591161447924,
+      "grad_norm": 0.8386031594806228,
+      "learning_rate": 8.612652608213097e-06,
+      "loss": 0.0808,
+      "step": 776
+    },
+    {
+      "epoch": 0.3451032644903398,
+      "grad_norm": 1.0606813630969132,
+      "learning_rate": 8.623751387347392e-06,
+      "loss": 0.0689,
+      "step": 777
+    },
+    {
+      "epoch": 0.3455474128358872,
+      "grad_norm": 0.9348314863875357,
+      "learning_rate": 8.634850166481687e-06,
+      "loss": 0.0732,
+      "step": 778
+    },
+    {
+      "epoch": 0.3459915611814346,
+      "grad_norm": 1.3147018531805297,
+      "learning_rate": 8.645948945615982e-06,
+      "loss": 0.1498,
+      "step": 779
+    },
+    {
+      "epoch": 0.346435709526982,
+      "grad_norm": 0.6686475142401406,
+      "learning_rate": 8.657047724750277e-06,
+      "loss": 0.0643,
+      "step": 780
+    },
+    {
+      "epoch": 0.34687985787252945,
+      "grad_norm": 1.4294668088170326,
+      "learning_rate": 8.668146503884574e-06,
+      "loss": 0.0973,
+      "step": 781
+    },
+    {
+      "epoch": 0.34732400621807685,
+      "grad_norm": 1.0334036928446257,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 0.0907,
+      "step": 782
+    },
+    {
+      "epoch": 0.34776815456362425,
+      "grad_norm": 0.787480188698369,
+      "learning_rate": 8.690344062153164e-06,
+      "loss": 0.0965,
+      "step": 783
+    },
+    {
+      "epoch": 0.34821230290917166,
+      "grad_norm": 1.1443565433110983,
+      "learning_rate": 8.70144284128746e-06,
+      "loss": 0.1128,
+      "step": 784
+    },
+    {
+      "epoch": 0.34865645125471906,
+      "grad_norm": 1.4871900340672362,
+      "learning_rate": 8.712541620421754e-06,
+      "loss": 0.1402,
+      "step": 785
+    },
+    {
+      "epoch": 0.3491005996002665,
+      "grad_norm": 1.2517565722671151,
+      "learning_rate": 8.72364039955605e-06,
+      "loss": 0.0937,
+      "step": 786
+    },
+    {
+      "epoch": 0.3495447479458139,
+      "grad_norm": 1.2066836215604748,
+      "learning_rate": 8.734739178690345e-06,
+      "loss": 0.1142,
+      "step": 787
+    },
+    {
+      "epoch": 0.3499888962913613,
+      "grad_norm": 0.9107920976880283,
+      "learning_rate": 8.74583795782464e-06,
+      "loss": 0.0885,
+      "step": 788
+    },
+    {
+      "epoch": 0.3504330446369087,
+      "grad_norm": 0.9724494230585177,
+      "learning_rate": 8.756936736958935e-06,
+      "loss": 0.0957,
+      "step": 789
+    },
+    {
+      "epoch": 0.3508771929824561,
+      "grad_norm": 0.8845187547684428,
+      "learning_rate": 8.768035516093231e-06,
+      "loss": 0.0713,
+      "step": 790
+    },
+    {
+      "epoch": 0.3513213413280036,
+      "grad_norm": 0.9555294870332015,
+      "learning_rate": 8.779134295227527e-06,
+      "loss": 0.0709,
+      "step": 791
+    },
+    {
+      "epoch": 0.351765489673551,
+      "grad_norm": 0.9971954202922753,
+      "learning_rate": 8.790233074361822e-06,
+      "loss": 0.086,
+      "step": 792
+    },
+    {
+      "epoch": 0.3522096380190984,
+      "grad_norm": 0.8799552486323357,
+      "learning_rate": 8.801331853496115e-06,
+      "loss": 0.0878,
+      "step": 793
+    },
+    {
+      "epoch": 0.3526537863646458,
+      "grad_norm": 1.1586612806863954,
+      "learning_rate": 8.812430632630412e-06,
+      "loss": 0.1017,
+      "step": 794
+    },
+    {
+      "epoch": 0.3530979347101932,
+      "grad_norm": 1.0618777970371058,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.0932,
+      "step": 795
+    },
+    {
+      "epoch": 0.35354208305574064,
+      "grad_norm": 1.0717541227428644,
+      "learning_rate": 8.834628190899002e-06,
+      "loss": 0.095,
+      "step": 796
+    },
+    {
+      "epoch": 0.35398623140128804,
+      "grad_norm": 1.1705694879290995,
+      "learning_rate": 8.845726970033297e-06,
+      "loss": 0.1118,
+      "step": 797
+    },
+    {
+      "epoch": 0.35443037974683544,
+      "grad_norm": 1.1829108047712253,
+      "learning_rate": 8.856825749167592e-06,
+      "loss": 0.1028,
+      "step": 798
+    },
+    {
+      "epoch": 0.35487452809238285,
+      "grad_norm": 0.7445292910176993,
+      "learning_rate": 8.867924528301887e-06,
+      "loss": 0.0639,
+      "step": 799
+    },
+    {
+      "epoch": 0.35531867643793025,
+      "grad_norm": 1.0475116637450934,
+      "learning_rate": 8.879023307436184e-06,
+      "loss": 0.0706,
+      "step": 800
+    },
+    {
+      "epoch": 0.3557628247834777,
+      "grad_norm": 1.1028250244486228,
+      "learning_rate": 8.890122086570479e-06,
+      "loss": 0.0819,
+      "step": 801
+    },
+    {
+      "epoch": 0.3562069731290251,
+      "grad_norm": 1.0807445455171478,
+      "learning_rate": 8.901220865704772e-06,
+      "loss": 0.0888,
+      "step": 802
+    },
+    {
+      "epoch": 0.3566511214745725,
+      "grad_norm": 1.0684985644998497,
+      "learning_rate": 8.912319644839067e-06,
+      "loss": 0.0822,
+      "step": 803
+    },
+    {
+      "epoch": 0.3570952698201199,
+      "grad_norm": 1.3191397870753145,
+      "learning_rate": 8.923418423973364e-06,
+      "loss": 0.1026,
+      "step": 804
+    },
+    {
+      "epoch": 0.3575394181656673,
+      "grad_norm": 1.0262852726585219,
+      "learning_rate": 8.93451720310766e-06,
+      "loss": 0.0776,
+      "step": 805
+    },
+    {
+      "epoch": 0.35798356651121477,
+      "grad_norm": 1.0652219347423726,
+      "learning_rate": 8.945615982241954e-06,
+      "loss": 0.0907,
+      "step": 806
+    },
+    {
+      "epoch": 0.35842771485676217,
+      "grad_norm": 1.0116135773749395,
+      "learning_rate": 8.95671476137625e-06,
+      "loss": 0.0959,
+      "step": 807
+    },
+    {
+      "epoch": 0.35887186320230957,
+      "grad_norm": 0.7818161469235702,
+      "learning_rate": 8.967813540510544e-06,
+      "loss": 0.0687,
+      "step": 808
+    },
+    {
+      "epoch": 0.359316011547857,
+      "grad_norm": 1.2323283633104487,
+      "learning_rate": 8.97891231964484e-06,
+      "loss": 0.1036,
+      "step": 809
+    },
+    {
+      "epoch": 0.3597601598934044,
+      "grad_norm": 1.0483052311963612,
+      "learning_rate": 8.990011098779136e-06,
+      "loss": 0.0899,
+      "step": 810
+    },
+    {
+      "epoch": 0.36020430823895183,
+      "grad_norm": 0.8215378781764873,
+      "learning_rate": 9.00110987791343e-06,
+      "loss": 0.089,
+      "step": 811
+    },
+    {
+      "epoch": 0.36064845658449923,
+      "grad_norm": 1.2325336486761531,
+      "learning_rate": 9.012208657047725e-06,
+      "loss": 0.1338,
+      "step": 812
+    },
+    {
+      "epoch": 0.36109260493004663,
+      "grad_norm": 1.0281228699327587,
+      "learning_rate": 9.02330743618202e-06,
+      "loss": 0.0928,
+      "step": 813
+    },
+    {
+      "epoch": 0.36153675327559404,
+      "grad_norm": 0.9331667240788679,
+      "learning_rate": 9.034406215316316e-06,
+      "loss": 0.0795,
+      "step": 814
+    },
+    {
+      "epoch": 0.36198090162114144,
+      "grad_norm": 0.9152493219085753,
+      "learning_rate": 9.045504994450612e-06,
+      "loss": 0.1015,
+      "step": 815
+    },
+    {
+      "epoch": 0.3624250499666889,
+      "grad_norm": 0.8370218994947535,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 0.0873,
+      "step": 816
+    },
+    {
+      "epoch": 0.3628691983122363,
+      "grad_norm": 0.7774818767814435,
+      "learning_rate": 9.067702552719202e-06,
+      "loss": 0.0822,
+      "step": 817
+    },
+    {
+      "epoch": 0.3633133466577837,
+      "grad_norm": 1.6384223947850474,
+      "learning_rate": 9.078801331853497e-06,
+      "loss": 0.081,
+      "step": 818
+    },
+    {
+      "epoch": 0.3637574950033311,
+      "grad_norm": 0.991313719663541,
+      "learning_rate": 9.089900110987792e-06,
+      "loss": 0.0837,
+      "step": 819
+    },
+    {
+      "epoch": 0.3642016433488785,
+      "grad_norm": 0.8596136661125425,
+      "learning_rate": 9.100998890122087e-06,
+      "loss": 0.0918,
+      "step": 820
+    },
+    {
+      "epoch": 0.36464579169442596,
+      "grad_norm": 1.0812055701714487,
+      "learning_rate": 9.112097669256382e-06,
+      "loss": 0.0825,
+      "step": 821
+    },
+    {
+      "epoch": 0.36508994003997336,
+      "grad_norm": 0.8070207504890635,
+      "learning_rate": 9.123196448390677e-06,
+      "loss": 0.0843,
+      "step": 822
+    },
+    {
+      "epoch": 0.36553408838552076,
+      "grad_norm": 1.1679552400138797,
+      "learning_rate": 9.134295227524974e-06,
+      "loss": 0.0914,
+      "step": 823
+    },
+    {
+      "epoch": 0.36597823673106816,
+      "grad_norm": 0.9627855966278458,
+      "learning_rate": 9.145394006659269e-06,
+      "loss": 0.08,
+      "step": 824
+    },
+    {
+      "epoch": 0.36642238507661556,
+      "grad_norm": 0.8861606234396192,
+      "learning_rate": 9.156492785793564e-06,
+      "loss": 0.0921,
+      "step": 825
+    },
+    {
+      "epoch": 0.366866533422163,
+      "grad_norm": 1.0771512576751088,
+      "learning_rate": 9.167591564927859e-06,
+      "loss": 0.0981,
+      "step": 826
+    },
+    {
+      "epoch": 0.3673106817677104,
+      "grad_norm": 0.7667307978284066,
+      "learning_rate": 9.178690344062154e-06,
+      "loss": 0.0763,
+      "step": 827
+    },
+    {
+      "epoch": 0.3677548301132578,
+      "grad_norm": 0.8789921021859833,
+      "learning_rate": 9.189789123196449e-06,
+      "loss": 0.0969,
+      "step": 828
+    },
+    {
+      "epoch": 0.3681989784588052,
+      "grad_norm": 1.100607694945035,
+      "learning_rate": 9.200887902330744e-06,
+      "loss": 0.0945,
+      "step": 829
+    },
+    {
+      "epoch": 0.3686431268043526,
+      "grad_norm": 0.8889918245566973,
+      "learning_rate": 9.21198668146504e-06,
+      "loss": 0.1027,
+      "step": 830
+    },
+    {
+      "epoch": 0.3690872751499001,
+      "grad_norm": 0.9919838528287025,
+      "learning_rate": 9.223085460599334e-06,
+      "loss": 0.0876,
+      "step": 831
+    },
+    {
+      "epoch": 0.3695314234954475,
+      "grad_norm": 0.7322878388650527,
+      "learning_rate": 9.23418423973363e-06,
+      "loss": 0.0827,
+      "step": 832
+    },
+    {
+      "epoch": 0.3699755718409949,
+      "grad_norm": 0.8334148632374379,
+      "learning_rate": 9.245283018867926e-06,
+      "loss": 0.0836,
+      "step": 833
+    },
+    {
+      "epoch": 0.3704197201865423,
+      "grad_norm": 1.2813561868661674,
+      "learning_rate": 9.256381798002221e-06,
+      "loss": 0.0889,
+      "step": 834
+    },
+    {
+      "epoch": 0.3708638685320897,
+      "grad_norm": 1.339839760620212,
+      "learning_rate": 9.267480577136516e-06,
+      "loss": 0.1145,
+      "step": 835
+    },
+    {
+      "epoch": 0.37130801687763715,
+      "grad_norm": 0.9154079084300466,
+      "learning_rate": 9.278579356270811e-06,
+      "loss": 0.0816,
+      "step": 836
+    },
+    {
+      "epoch": 0.37175216522318455,
+      "grad_norm": 0.6790049839601768,
+      "learning_rate": 9.289678135405106e-06,
+      "loss": 0.052,
+      "step": 837
+    },
+    {
+      "epoch": 0.37219631356873195,
+      "grad_norm": 1.2902350399135438,
+      "learning_rate": 9.300776914539401e-06,
+      "loss": 0.1205,
+      "step": 838
+    },
+    {
+      "epoch": 0.37264046191427935,
+      "grad_norm": 0.6777630897746993,
+      "learning_rate": 9.311875693673697e-06,
+      "loss": 0.0626,
+      "step": 839
+    },
+    {
+      "epoch": 0.37308461025982675,
+      "grad_norm": 0.974153136606222,
+      "learning_rate": 9.322974472807992e-06,
+      "loss": 0.0872,
+      "step": 840
+    },
+    {
+      "epoch": 0.3735287586053742,
+      "grad_norm": 0.8802239921866019,
+      "learning_rate": 9.334073251942287e-06,
+      "loss": 0.0705,
+      "step": 841
+    },
+    {
+      "epoch": 0.3739729069509216,
+      "grad_norm": 0.8051312485970141,
+      "learning_rate": 9.345172031076582e-06,
+      "loss": 0.0878,
+      "step": 842
+    },
+    {
+      "epoch": 0.374417055296469,
+      "grad_norm": 1.160749633212372,
+      "learning_rate": 9.356270810210878e-06,
+      "loss": 0.1112,
+      "step": 843
+    },
+    {
+      "epoch": 0.3748612036420164,
+      "grad_norm": 0.8346210460548523,
+      "learning_rate": 9.367369589345174e-06,
+      "loss": 0.1035,
+      "step": 844
+    },
+    {
+      "epoch": 0.3753053519875639,
+      "grad_norm": 1.0464653523849117,
+      "learning_rate": 9.378468368479467e-06,
+      "loss": 0.0901,
+      "step": 845
+    },
+    {
+      "epoch": 0.3757495003331113,
+      "grad_norm": 0.7515018279210384,
+      "learning_rate": 9.389567147613764e-06,
+      "loss": 0.0618,
+      "step": 846
+    },
+    {
+      "epoch": 0.3761936486786587,
+      "grad_norm": 1.1240058131240411,
+      "learning_rate": 9.400665926748059e-06,
+      "loss": 0.0682,
+      "step": 847
+    },
+    {
+      "epoch": 0.3766377970242061,
+      "grad_norm": 0.8330871505734428,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.0815,
+      "step": 848
+    },
+    {
+      "epoch": 0.3770819453697535,
+      "grad_norm": 0.6597157022162015,
+      "learning_rate": 9.422863485016649e-06,
+      "loss": 0.0705,
+      "step": 849
+    },
+    {
+      "epoch": 0.37752609371530094,
+      "grad_norm": 0.870783524840563,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.0756,
+      "step": 850
+    },
+    {
+      "epoch": 0.37797024206084834,
+      "grad_norm": 0.7718200428803087,
+      "learning_rate": 9.445061043285239e-06,
+      "loss": 0.0688,
+      "step": 851
+    },
+    {
+      "epoch": 0.37841439040639574,
+      "grad_norm": 1.0565515531045717,
+      "learning_rate": 9.456159822419536e-06,
+      "loss": 0.096,
+      "step": 852
+    },
+    {
+      "epoch": 0.37885853875194314,
+      "grad_norm": 0.9116267321999116,
+      "learning_rate": 9.46725860155383e-06,
+      "loss": 0.1219,
+      "step": 853
+    },
+    {
+      "epoch": 0.37930268709749054,
+      "grad_norm": 0.614891198569231,
+      "learning_rate": 9.478357380688124e-06,
+      "loss": 0.0618,
+      "step": 854
+    },
+    {
+      "epoch": 0.379746835443038,
+      "grad_norm": 0.6996810655829085,
+      "learning_rate": 9.48945615982242e-06,
+      "loss": 0.0733,
+      "step": 855
+    },
+    {
+      "epoch": 0.3801909837885854,
+      "grad_norm": 0.8742171165547441,
+      "learning_rate": 9.500554938956716e-06,
+      "loss": 0.0943,
+      "step": 856
+    },
+    {
+      "epoch": 0.3806351321341328,
+      "grad_norm": 0.6941177838819981,
+      "learning_rate": 9.511653718091011e-06,
+      "loss": 0.0979,
+      "step": 857
+    },
+    {
+      "epoch": 0.3810792804796802,
+      "grad_norm": 0.9912966659857255,
+      "learning_rate": 9.522752497225306e-06,
+      "loss": 0.1146,
+      "step": 858
+    },
+    {
+      "epoch": 0.3815234288252276,
+      "grad_norm": 0.8052784568872317,
+      "learning_rate": 9.533851276359601e-06,
+      "loss": 0.0765,
+      "step": 859
+    },
+    {
+      "epoch": 0.38196757717077506,
+      "grad_norm": 0.6650776701454606,
+      "learning_rate": 9.544950055493896e-06,
+      "loss": 0.0598,
+      "step": 860
+    },
+    {
+      "epoch": 0.38241172551632246,
+      "grad_norm": 0.9207926874686408,
+      "learning_rate": 9.556048834628191e-06,
+      "loss": 0.0812,
+      "step": 861
+    },
+    {
+      "epoch": 0.38285587386186987,
+      "grad_norm": 0.9371867517767639,
+      "learning_rate": 9.567147613762488e-06,
+      "loss": 0.1016,
+      "step": 862
+    },
+    {
+      "epoch": 0.38330002220741727,
+      "grad_norm": 0.8381824885212333,
+      "learning_rate": 9.578246392896782e-06,
+      "loss": 0.0975,
+      "step": 863
+    },
+    {
+      "epoch": 0.38374417055296467,
+      "grad_norm": 0.8006129424637977,
+      "learning_rate": 9.589345172031077e-06,
+      "loss": 0.0768,
+      "step": 864
+    },
+    {
+      "epoch": 0.3841883188985121,
+      "grad_norm": 1.5462086828607524,
+      "learning_rate": 9.600443951165372e-06,
+      "loss": 0.0973,
+      "step": 865
+    },
+    {
+      "epoch": 0.38463246724405953,
+      "grad_norm": 0.9098699947791947,
+      "learning_rate": 9.611542730299668e-06,
+      "loss": 0.0862,
+      "step": 866
+    },
+    {
+      "epoch": 0.38507661558960693,
+      "grad_norm": 0.6758969469373516,
+      "learning_rate": 9.622641509433963e-06,
+      "loss": 0.0926,
+      "step": 867
+    },
+    {
+      "epoch": 0.38552076393515433,
+      "grad_norm": 0.8013932877888278,
+      "learning_rate": 9.633740288568259e-06,
+      "loss": 0.0788,
+      "step": 868
+    },
+    {
+      "epoch": 0.38596491228070173,
+      "grad_norm": 0.8330625416886887,
+      "learning_rate": 9.644839067702554e-06,
+      "loss": 0.0908,
+      "step": 869
+    },
+    {
+      "epoch": 0.3864090606262492,
+      "grad_norm": 0.8653830292054833,
+      "learning_rate": 9.655937846836849e-06,
+      "loss": 0.0845,
+      "step": 870
+    },
+    {
+      "epoch": 0.3868532089717966,
+      "grad_norm": 0.9667174949251367,
+      "learning_rate": 9.667036625971144e-06,
+      "loss": 0.0661,
+      "step": 871
+    },
+    {
+      "epoch": 0.387297357317344,
+      "grad_norm": 0.9162313384818003,
+      "learning_rate": 9.678135405105439e-06,
+      "loss": 0.0839,
+      "step": 872
+    },
+    {
+      "epoch": 0.3877415056628914,
+      "grad_norm": 1.05316579924942,
+      "learning_rate": 9.689234184239734e-06,
+      "loss": 0.0884,
+      "step": 873
+    },
+    {
+      "epoch": 0.3881856540084388,
+      "grad_norm": 0.6947987167903426,
+      "learning_rate": 9.700332963374029e-06,
+      "loss": 0.0676,
+      "step": 874
+    },
+    {
+      "epoch": 0.38862980235398625,
+      "grad_norm": 0.7637218765216399,
+      "learning_rate": 9.711431742508326e-06,
+      "loss": 0.0855,
+      "step": 875
+    },
+    {
+      "epoch": 0.38907395069953365,
+      "grad_norm": 0.8049041095871476,
+      "learning_rate": 9.72253052164262e-06,
+      "loss": 0.0783,
+      "step": 876
+    },
+    {
+      "epoch": 0.38951809904508106,
+      "grad_norm": 1.2616209833129097,
+      "learning_rate": 9.733629300776916e-06,
+      "loss": 0.0994,
+      "step": 877
+    },
+    {
+      "epoch": 0.38996224739062846,
+      "grad_norm": 0.8501327438099852,
+      "learning_rate": 9.744728079911211e-06,
+      "loss": 0.0865,
+      "step": 878
+    },
+    {
+      "epoch": 0.39040639573617586,
+      "grad_norm": 0.8980289128819791,
+      "learning_rate": 9.755826859045506e-06,
+      "loss": 0.0644,
+      "step": 879
+    },
+    {
+      "epoch": 0.3908505440817233,
+      "grad_norm": 0.7464217413483473,
+      "learning_rate": 9.766925638179801e-06,
+      "loss": 0.0751,
+      "step": 880
+    },
+    {
+      "epoch": 0.3912946924272707,
+      "grad_norm": 1.0405400428578027,
+      "learning_rate": 9.778024417314096e-06,
+      "loss": 0.0862,
+      "step": 881
+    },
+    {
+      "epoch": 0.3917388407728181,
+      "grad_norm": 0.7773928325897406,
+      "learning_rate": 9.789123196448391e-06,
+      "loss": 0.0877,
+      "step": 882
+    },
+    {
+      "epoch": 0.3921829891183655,
+      "grad_norm": 0.6831878151686481,
+      "learning_rate": 9.800221975582686e-06,
+      "loss": 0.0613,
+      "step": 883
+    },
+    {
+      "epoch": 0.3926271374639129,
+      "grad_norm": 0.8414451878698297,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.0742,
+      "step": 884
+    },
+    {
+      "epoch": 0.3930712858094604,
+      "grad_norm": 0.8370630047631802,
+      "learning_rate": 9.822419533851278e-06,
+      "loss": 0.0788,
+      "step": 885
+    },
+    {
+      "epoch": 0.3935154341550078,
+      "grad_norm": 0.66195315447052,
+      "learning_rate": 9.833518312985573e-06,
+      "loss": 0.0673,
+      "step": 886
+    },
+    {
+      "epoch": 0.3939595825005552,
+      "grad_norm": 0.5443847344778766,
+      "learning_rate": 9.844617092119868e-06,
+      "loss": 0.065,
+      "step": 887
+    },
+    {
+      "epoch": 0.3944037308461026,
+      "grad_norm": 0.7436541471597082,
+      "learning_rate": 9.855715871254163e-06,
+      "loss": 0.0718,
+      "step": 888
+    },
+    {
+      "epoch": 0.39484787919165,
+      "grad_norm": 0.9048716378979127,
+      "learning_rate": 9.866814650388458e-06,
+      "loss": 0.0647,
+      "step": 889
+    },
+    {
+      "epoch": 0.39529202753719744,
+      "grad_norm": 0.887466655199119,
+      "learning_rate": 9.877913429522753e-06,
+      "loss": 0.0742,
+      "step": 890
+    },
+    {
+      "epoch": 0.39573617588274485,
+      "grad_norm": 0.5558633344968636,
+      "learning_rate": 9.889012208657048e-06,
+      "loss": 0.0547,
+      "step": 891
+    },
+    {
+      "epoch": 0.39618032422829225,
+      "grad_norm": 0.6342481865802312,
+      "learning_rate": 9.900110987791344e-06,
+      "loss": 0.0652,
+      "step": 892
+    },
+    {
+      "epoch": 0.39662447257383965,
+      "grad_norm": 0.8609031233478771,
+      "learning_rate": 9.911209766925639e-06,
+      "loss": 0.0837,
+      "step": 893
+    },
+    {
+      "epoch": 0.39706862091938705,
+      "grad_norm": 0.8888916230177222,
+      "learning_rate": 9.922308546059934e-06,
+      "loss": 0.0892,
+      "step": 894
+    },
+    {
+      "epoch": 0.3975127692649345,
+      "grad_norm": 0.8338604612800079,
+      "learning_rate": 9.93340732519423e-06,
+      "loss": 0.101,
+      "step": 895
+    },
+    {
+      "epoch": 0.3979569176104819,
+      "grad_norm": 0.689341320326161,
+      "learning_rate": 9.944506104328525e-06,
+      "loss": 0.0778,
+      "step": 896
+    },
+    {
+      "epoch": 0.3984010659560293,
+      "grad_norm": 0.7629364933455176,
+      "learning_rate": 9.955604883462819e-06,
+      "loss": 0.0702,
+      "step": 897
+    },
+    {
+      "epoch": 0.3988452143015767,
+      "grad_norm": 0.7205533738637083,
+      "learning_rate": 9.966703662597114e-06,
+      "loss": 0.0712,
+      "step": 898
+    },
+    {
+      "epoch": 0.3992893626471241,
+      "grad_norm": 1.1452213067609647,
+      "learning_rate": 9.97780244173141e-06,
+      "loss": 0.0868,
+      "step": 899
+    },
+    {
+      "epoch": 0.39973351099267157,
+      "grad_norm": 1.7583678128183422,
+      "learning_rate": 9.988901220865706e-06,
+      "loss": 0.0866,
+      "step": 900
+    },
+    {
+      "epoch": 0.40017765933821897,
+      "grad_norm": 1.0237065065243787,
+      "learning_rate": 1e-05,
+      "loss": 0.0966,
+      "step": 901
+    },
+    {
+      "epoch": 0.4006218076837664,
+      "grad_norm": 0.9116552545055456,
+      "learning_rate": 9.999999624207532e-06,
+      "loss": 0.0752,
+      "step": 902
+    },
+    {
+      "epoch": 0.4010659560293138,
+      "grad_norm": 0.9745084022934988,
+      "learning_rate": 9.999998496830188e-06,
+      "loss": 0.0816,
+      "step": 903
+    },
+    {
+      "epoch": 0.4015101043748612,
+      "grad_norm": 0.8729574496188063,
+      "learning_rate": 9.999996617868132e-06,
+      "loss": 0.096,
+      "step": 904
+    },
+    {
+      "epoch": 0.40195425272040863,
+      "grad_norm": 0.8195978110162689,
+      "learning_rate": 9.999993987321651e-06,
+      "loss": 0.0848,
+      "step": 905
+    },
+    {
+      "epoch": 0.40239840106595604,
+      "grad_norm": 1.0642057405652703,
+      "learning_rate": 9.999990605191136e-06,
+      "loss": 0.0734,
+      "step": 906
+    },
+    {
+      "epoch": 0.40284254941150344,
+      "grad_norm": 1.055742691312144,
+      "learning_rate": 9.9999864714771e-06,
+      "loss": 0.1099,
+      "step": 907
+    },
+    {
+      "epoch": 0.40328669775705084,
+      "grad_norm": 1.2115140501137451,
+      "learning_rate": 9.999981586180161e-06,
+      "loss": 0.0742,
+      "step": 908
+    },
+    {
+      "epoch": 0.4037308461025983,
+      "grad_norm": 0.899219389432472,
+      "learning_rate": 9.999975949301057e-06,
+      "loss": 0.078,
+      "step": 909
+    },
+    {
+      "epoch": 0.4041749944481457,
+      "grad_norm": 0.8631042166204541,
+      "learning_rate": 9.99996956084063e-06,
+      "loss": 0.0721,
+      "step": 910
+    },
+    {
+      "epoch": 0.4046191427936931,
+      "grad_norm": 0.90232933367849,
+      "learning_rate": 9.999962420799846e-06,
+      "loss": 0.0855,
+      "step": 911
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.9832129529060516,
+      "learning_rate": 9.999954529179773e-06,
+      "loss": 0.0884,
+      "step": 912
+    },
+    {
+      "epoch": 0.4055074394847879,
+      "grad_norm": 0.6541793575769728,
+      "learning_rate": 9.999945885981603e-06,
+      "loss": 0.0649,
+      "step": 913
+    },
+    {
+      "epoch": 0.40595158783033536,
+      "grad_norm": 0.7878273299354217,
+      "learning_rate": 9.999936491206631e-06,
+      "loss": 0.0806,
+      "step": 914
+    },
+    {
+      "epoch": 0.40639573617588276,
+      "grad_norm": 1.1901788044990722,
+      "learning_rate": 9.99992634485627e-06,
+      "loss": 0.0926,
+      "step": 915
+    },
+    {
+      "epoch": 0.40683988452143016,
+      "grad_norm": 1.14952509907395,
+      "learning_rate": 9.999915446932045e-06,
+      "loss": 0.1071,
+      "step": 916
+    },
+    {
+      "epoch": 0.40728403286697756,
+      "grad_norm": 0.7798635545238589,
+      "learning_rate": 9.999903797435596e-06,
+      "loss": 0.0771,
+      "step": 917
+    },
+    {
+      "epoch": 0.40772818121252496,
+      "grad_norm": 1.0917620513168562,
+      "learning_rate": 9.999891396368672e-06,
+      "loss": 0.0949,
+      "step": 918
+    },
+    {
+      "epoch": 0.4081723295580724,
+      "grad_norm": 1.0952645685085969,
+      "learning_rate": 9.999878243733138e-06,
+      "loss": 0.0817,
+      "step": 919
+    },
+    {
+      "epoch": 0.4086164779036198,
+      "grad_norm": 0.8880652433088206,
+      "learning_rate": 9.99986433953097e-06,
+      "loss": 0.0776,
+      "step": 920
+    },
+    {
+      "epoch": 0.4090606262491672,
+      "grad_norm": 0.8333025562604276,
+      "learning_rate": 9.99984968376426e-06,
+      "loss": 0.0956,
+      "step": 921
+    },
+    {
+      "epoch": 0.4095047745947146,
+      "grad_norm": 1.0999527872907782,
+      "learning_rate": 9.99983427643521e-06,
+      "loss": 0.0782,
+      "step": 922
+    },
+    {
+      "epoch": 0.40994892294026203,
+      "grad_norm": 1.0065680465940938,
+      "learning_rate": 9.999818117546135e-06,
+      "loss": 0.0862,
+      "step": 923
+    },
+    {
+      "epoch": 0.4103930712858095,
+      "grad_norm": 0.7446846870231809,
+      "learning_rate": 9.999801207099464e-06,
+      "loss": 0.0693,
+      "step": 924
+    },
+    {
+      "epoch": 0.4108372196313569,
+      "grad_norm": 0.7664119308596476,
+      "learning_rate": 9.99978354509774e-06,
+      "loss": 0.062,
+      "step": 925
+    },
+    {
+      "epoch": 0.4112813679769043,
+      "grad_norm": 0.6216063626642743,
+      "learning_rate": 9.99976513154362e-06,
+      "loss": 0.0855,
+      "step": 926
+    },
+    {
+      "epoch": 0.4117255163224517,
+      "grad_norm": 0.7583401455167236,
+      "learning_rate": 9.99974596643987e-06,
+      "loss": 0.0902,
+      "step": 927
+    },
+    {
+      "epoch": 0.4121696646679991,
+      "grad_norm": 0.8236694673839551,
+      "learning_rate": 9.999726049789367e-06,
+      "loss": 0.094,
+      "step": 928
+    },
+    {
+      "epoch": 0.41261381301354655,
+      "grad_norm": 1.0278078628460734,
+      "learning_rate": 9.999705381595111e-06,
+      "loss": 0.1162,
+      "step": 929
+    },
+    {
+      "epoch": 0.41305796135909395,
+      "grad_norm": 0.9106208454624181,
+      "learning_rate": 9.999683961860205e-06,
+      "loss": 0.0832,
+      "step": 930
+    },
+    {
+      "epoch": 0.41350210970464135,
+      "grad_norm": 0.9583162458878111,
+      "learning_rate": 9.99966179058787e-06,
+      "loss": 0.0622,
+      "step": 931
+    },
+    {
+      "epoch": 0.41394625805018875,
+      "grad_norm": 1.02310895201938,
+      "learning_rate": 9.999638867781437e-06,
+      "loss": 0.0747,
+      "step": 932
+    },
+    {
+      "epoch": 0.41439040639573615,
+      "grad_norm": 0.9574965389717746,
+      "learning_rate": 9.999615193444354e-06,
+      "loss": 0.0826,
+      "step": 933
+    },
+    {
+      "epoch": 0.4148345547412836,
+      "grad_norm": 0.6820969947085748,
+      "learning_rate": 9.99959076758018e-06,
+      "loss": 0.0794,
+      "step": 934
+    },
+    {
+      "epoch": 0.415278703086831,
+      "grad_norm": 0.8271166638389166,
+      "learning_rate": 9.999565590192584e-06,
+      "loss": 0.0677,
+      "step": 935
+    },
+    {
+      "epoch": 0.4157228514323784,
+      "grad_norm": 0.7913919823378441,
+      "learning_rate": 9.999539661285354e-06,
+      "loss": 0.0899,
+      "step": 936
+    },
+    {
+      "epoch": 0.4161669997779258,
+      "grad_norm": 0.6933241808880691,
+      "learning_rate": 9.999512980862382e-06,
+      "loss": 0.0762,
+      "step": 937
+    },
+    {
+      "epoch": 0.4166111481234732,
+      "grad_norm": 1.0464472300779635,
+      "learning_rate": 9.999485548927686e-06,
+      "loss": 0.0879,
+      "step": 938
+    },
+    {
+      "epoch": 0.4170552964690207,
+      "grad_norm": 0.9713981902628005,
+      "learning_rate": 9.999457365485383e-06,
+      "loss": 0.0859,
+      "step": 939
+    },
+    {
+      "epoch": 0.4174994448145681,
+      "grad_norm": 0.7157769729378113,
+      "learning_rate": 9.999428430539713e-06,
+      "loss": 0.063,
+      "step": 940
+    },
+    {
+      "epoch": 0.4179435931601155,
+      "grad_norm": 1.3110820742809257,
+      "learning_rate": 9.999398744095024e-06,
+      "loss": 0.1083,
+      "step": 941
+    },
+    {
+      "epoch": 0.4183877415056629,
+      "grad_norm": 0.9278738003250622,
+      "learning_rate": 9.999368306155778e-06,
+      "loss": 0.0682,
+      "step": 942
+    },
+    {
+      "epoch": 0.4188318898512103,
+      "grad_norm": 1.081752257204933,
+      "learning_rate": 9.999337116726555e-06,
+      "loss": 0.082,
+      "step": 943
+    },
+    {
+      "epoch": 0.41927603819675774,
+      "grad_norm": 0.7655308857156353,
+      "learning_rate": 9.999305175812035e-06,
+      "loss": 0.0824,
+      "step": 944
+    },
+    {
+      "epoch": 0.41972018654230514,
+      "grad_norm": 0.843733636393823,
+      "learning_rate": 9.999272483417027e-06,
+      "loss": 0.0583,
+      "step": 945
+    },
+    {
+      "epoch": 0.42016433488785254,
+      "grad_norm": 1.0473704718025856,
+      "learning_rate": 9.99923903954644e-06,
+      "loss": 0.0893,
+      "step": 946
+    },
+    {
+      "epoch": 0.42060848323339994,
+      "grad_norm": 0.5121181388849327,
+      "learning_rate": 9.999204844205304e-06,
+      "loss": 0.0599,
+      "step": 947
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.9066795499623488,
+      "learning_rate": 9.999169897398757e-06,
+      "loss": 0.1344,
+      "step": 948
+    },
+    {
+      "epoch": 0.4214967799244948,
+      "grad_norm": 0.910793967335512,
+      "learning_rate": 9.999134199132054e-06,
+      "loss": 0.0669,
+      "step": 949
+    },
+    {
+      "epoch": 0.4219409282700422,
+      "grad_norm": 0.5923909271714048,
+      "learning_rate": 9.999097749410561e-06,
+      "loss": 0.0739,
+      "step": 950
+    },
+    {
+      "epoch": 0.4223850766155896,
+      "grad_norm": 0.8626478233669734,
+      "learning_rate": 9.999060548239757e-06,
+      "loss": 0.085,
+      "step": 951
+    },
+    {
+      "epoch": 0.422829224961137,
+      "grad_norm": 1.2157299829969963,
+      "learning_rate": 9.999022595625233e-06,
+      "loss": 0.0927,
+      "step": 952
+    },
+    {
+      "epoch": 0.4232733733066844,
+      "grad_norm": 0.8683032499306118,
+      "learning_rate": 9.998983891572693e-06,
+      "loss": 0.0857,
+      "step": 953
+    },
+    {
+      "epoch": 0.42371752165223187,
+      "grad_norm": 0.7958109904631795,
+      "learning_rate": 9.998944436087956e-06,
+      "loss": 0.0848,
+      "step": 954
+    },
+    {
+      "epoch": 0.42416166999777927,
+      "grad_norm": 0.8911128877165261,
+      "learning_rate": 9.998904229176955e-06,
+      "loss": 0.0684,
+      "step": 955
+    },
+    {
+      "epoch": 0.42460581834332667,
+      "grad_norm": 0.8480526510619464,
+      "learning_rate": 9.998863270845731e-06,
+      "loss": 0.0819,
+      "step": 956
+    },
+    {
+      "epoch": 0.42504996668887407,
+      "grad_norm": 1.1640161965795854,
+      "learning_rate": 9.99882156110044e-06,
+      "loss": 0.1114,
+      "step": 957
+    },
+    {
+      "epoch": 0.42549411503442147,
+      "grad_norm": 0.7232203452809441,
+      "learning_rate": 9.998779099947356e-06,
+      "loss": 0.0598,
+      "step": 958
+    },
+    {
+      "epoch": 0.42593826337996893,
+      "grad_norm": 0.9706452342240306,
+      "learning_rate": 9.998735887392858e-06,
+      "loss": 0.083,
+      "step": 959
+    },
+    {
+      "epoch": 0.42638241172551633,
+      "grad_norm": 0.9335641087127532,
+      "learning_rate": 9.998691923443442e-06,
+      "loss": 0.0866,
+      "step": 960
+    },
+    {
+      "epoch": 0.42682656007106373,
+      "grad_norm": 0.6482813505555395,
+      "learning_rate": 9.998647208105717e-06,
+      "loss": 0.0792,
+      "step": 961
+    },
+    {
+      "epoch": 0.42727070841661113,
+      "grad_norm": 1.0124751766766096,
+      "learning_rate": 9.998601741386404e-06,
+      "loss": 0.0981,
+      "step": 962
+    },
+    {
+      "epoch": 0.42771485676215854,
+      "grad_norm": 0.8713994807005938,
+      "learning_rate": 9.998555523292338e-06,
+      "loss": 0.112,
+      "step": 963
+    },
+    {
+      "epoch": 0.428159005107706,
+      "grad_norm": 0.7341323526220324,
+      "learning_rate": 9.998508553830468e-06,
+      "loss": 0.079,
+      "step": 964
+    },
+    {
+      "epoch": 0.4286031534532534,
+      "grad_norm": 0.7729393765262612,
+      "learning_rate": 9.99846083300785e-06,
+      "loss": 0.0783,
+      "step": 965
+    },
+    {
+      "epoch": 0.4290473017988008,
+      "grad_norm": 0.9156169394277074,
+      "learning_rate": 9.99841236083166e-06,
+      "loss": 0.0806,
+      "step": 966
+    },
+    {
+      "epoch": 0.4294914501443482,
+      "grad_norm": 0.8551663881257657,
+      "learning_rate": 9.998363137309187e-06,
+      "loss": 0.0618,
+      "step": 967
+    },
+    {
+      "epoch": 0.4299355984898956,
+      "grad_norm": 0.5370934826053735,
+      "learning_rate": 9.998313162447824e-06,
+      "loss": 0.0652,
+      "step": 968
+    },
+    {
+      "epoch": 0.43037974683544306,
+      "grad_norm": 0.7190678146601934,
+      "learning_rate": 9.998262436255087e-06,
+      "loss": 0.0574,
+      "step": 969
+    },
+    {
+      "epoch": 0.43082389518099046,
+      "grad_norm": 0.790706849866094,
+      "learning_rate": 9.998210958738601e-06,
+      "loss": 0.0805,
+      "step": 970
+    },
+    {
+      "epoch": 0.43126804352653786,
+      "grad_norm": 0.7463621831379682,
+      "learning_rate": 9.998158729906102e-06,
+      "loss": 0.0712,
+      "step": 971
+    },
+    {
+      "epoch": 0.43171219187208526,
+      "grad_norm": 1.0561578156156401,
+      "learning_rate": 9.998105749765444e-06,
+      "loss": 0.0901,
+      "step": 972
+    },
+    {
+      "epoch": 0.43215634021763266,
+      "grad_norm": 0.7180209621862901,
+      "learning_rate": 9.998052018324586e-06,
+      "loss": 0.0687,
+      "step": 973
+    },
+    {
+      "epoch": 0.4326004885631801,
+      "grad_norm": 0.8430240030491031,
+      "learning_rate": 9.99799753559161e-06,
+      "loss": 0.0763,
+      "step": 974
+    },
+    {
+      "epoch": 0.4330446369087275,
+      "grad_norm": 0.6157610719315397,
+      "learning_rate": 9.997942301574701e-06,
+      "loss": 0.0679,
+      "step": 975
+    },
+    {
+      "epoch": 0.4334887852542749,
+      "grad_norm": 0.8284454213431681,
+      "learning_rate": 9.997886316282167e-06,
+      "loss": 0.0901,
+      "step": 976
+    },
+    {
+      "epoch": 0.4339329335998223,
+      "grad_norm": 0.6138908846381919,
+      "learning_rate": 9.997829579722418e-06,
+      "loss": 0.068,
+      "step": 977
+    },
+    {
+      "epoch": 0.4343770819453698,
+      "grad_norm": 0.6977507297268527,
+      "learning_rate": 9.997772091903984e-06,
+      "loss": 0.0719,
+      "step": 978
+    },
+    {
+      "epoch": 0.4348212302909172,
+      "grad_norm": 0.672879000533416,
+      "learning_rate": 9.997713852835509e-06,
+      "loss": 0.0859,
+      "step": 979
+    },
+    {
+      "epoch": 0.4352653786364646,
+      "grad_norm": 1.0953915290421439,
+      "learning_rate": 9.997654862525746e-06,
+      "loss": 0.0984,
+      "step": 980
+    },
+    {
+      "epoch": 0.435709526982012,
+      "grad_norm": 0.7214317597556656,
+      "learning_rate": 9.997595120983561e-06,
+      "loss": 0.0788,
+      "step": 981
+    },
+    {
+      "epoch": 0.4361536753275594,
+      "grad_norm": 0.6240169495789859,
+      "learning_rate": 9.997534628217935e-06,
+      "loss": 0.0533,
+      "step": 982
+    },
+    {
+      "epoch": 0.43659782367310684,
+      "grad_norm": 1.0463824152225416,
+      "learning_rate": 9.997473384237962e-06,
+      "loss": 0.1256,
+      "step": 983
+    },
+    {
+      "epoch": 0.43704197201865425,
+      "grad_norm": 0.6075177918870873,
+      "learning_rate": 9.997411389052846e-06,
+      "loss": 0.0907,
+      "step": 984
+    },
+    {
+      "epoch": 0.43748612036420165,
+      "grad_norm": 0.8030490613184201,
+      "learning_rate": 9.997348642671906e-06,
+      "loss": 0.075,
+      "step": 985
+    },
+    {
+      "epoch": 0.43793026870974905,
+      "grad_norm": 0.6431009442762712,
+      "learning_rate": 9.997285145104578e-06,
+      "loss": 0.0681,
+      "step": 986
+    },
+    {
+      "epoch": 0.43837441705529645,
+      "grad_norm": 0.7359898616321635,
+      "learning_rate": 9.997220896360402e-06,
+      "loss": 0.0668,
+      "step": 987
+    },
+    {
+      "epoch": 0.4388185654008439,
+      "grad_norm": 0.9059235668805204,
+      "learning_rate": 9.997155896449037e-06,
+      "loss": 0.1006,
+      "step": 988
+    },
+    {
+      "epoch": 0.4392627137463913,
+      "grad_norm": 0.5647469272415054,
+      "learning_rate": 9.997090145380253e-06,
+      "loss": 0.0699,
+      "step": 989
+    },
+    {
+      "epoch": 0.4397068620919387,
+      "grad_norm": 0.7349393202965936,
+      "learning_rate": 9.997023643163937e-06,
+      "loss": 0.0731,
+      "step": 990
+    },
+    {
+      "epoch": 0.4401510104374861,
+      "grad_norm": 0.6226852410738429,
+      "learning_rate": 9.996956389810082e-06,
+      "loss": 0.0663,
+      "step": 991
+    },
+    {
+      "epoch": 0.4405951587830335,
+      "grad_norm": 0.6263169289849984,
+      "learning_rate": 9.996888385328798e-06,
+      "loss": 0.0733,
+      "step": 992
+    },
+    {
+      "epoch": 0.44103930712858097,
+      "grad_norm": 0.7290496435666083,
+      "learning_rate": 9.996819629730305e-06,
+      "loss": 0.0664,
+      "step": 993
+    },
+    {
+      "epoch": 0.4414834554741284,
+      "grad_norm": 0.6744456519687402,
+      "learning_rate": 9.996750123024943e-06,
+      "loss": 0.0637,
+      "step": 994
+    },
+    {
+      "epoch": 0.4419276038196758,
+      "grad_norm": 0.9231205047824739,
+      "learning_rate": 9.996679865223157e-06,
+      "loss": 0.0598,
+      "step": 995
+    },
+    {
+      "epoch": 0.4423717521652232,
+      "grad_norm": 0.6917241860836084,
+      "learning_rate": 9.99660885633551e-06,
+      "loss": 0.07,
+      "step": 996
+    },
+    {
+      "epoch": 0.4428159005107706,
+      "grad_norm": 1.0027314643568717,
+      "learning_rate": 9.996537096372672e-06,
+      "loss": 0.0998,
+      "step": 997
+    },
+    {
+      "epoch": 0.44326004885631803,
+      "grad_norm": 0.7262325240442806,
+      "learning_rate": 9.996464585345433e-06,
+      "loss": 0.0703,
+      "step": 998
+    },
+    {
+      "epoch": 0.44370419720186544,
+      "grad_norm": 0.7680770645535525,
+      "learning_rate": 9.996391323264693e-06,
+      "loss": 0.0884,
+      "step": 999
+    },
+    {
+      "epoch": 0.44414834554741284,
+      "grad_norm": 0.7559490106951665,
+      "learning_rate": 9.996317310141462e-06,
+      "loss": 0.0885,
+      "step": 1000
+    },
+    {
+      "epoch": 0.44459249389296024,
+      "grad_norm": 0.648213091365164,
+      "learning_rate": 9.996242545986868e-06,
+      "loss": 0.0657,
+      "step": 1001
+    },
+    {
+      "epoch": 0.44503664223850764,
+      "grad_norm": 0.941859001821441,
+      "learning_rate": 9.996167030812146e-06,
+      "loss": 0.0771,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4454807905840551,
+      "grad_norm": 0.9336722754874243,
+      "learning_rate": 9.996090764628649e-06,
+      "loss": 0.062,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4459249389296025,
+      "grad_norm": 0.6689537581560581,
+      "learning_rate": 9.996013747447844e-06,
+      "loss": 0.0676,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4463690872751499,
+      "grad_norm": 0.6868609219124203,
+      "learning_rate": 9.995935979281304e-06,
+      "loss": 0.0698,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4468132356206973,
+      "grad_norm": 0.7598513445760847,
+      "learning_rate": 9.995857460140719e-06,
+      "loss": 0.0663,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4472573839662447,
+      "grad_norm": 0.8425846771969293,
+      "learning_rate": 9.995778190037893e-06,
+      "loss": 0.0709,
+      "step": 1007
+    },
+    {
+      "epoch": 0.44770153231179216,
+      "grad_norm": 0.7973802310676283,
+      "learning_rate": 9.995698168984743e-06,
+      "loss": 0.0669,
+      "step": 1008
+    },
+    {
+      "epoch": 0.44814568065733956,
+      "grad_norm": 0.8057530371483355,
+      "learning_rate": 9.995617396993297e-06,
+      "loss": 0.105,
+      "step": 1009
+    },
+    {
+      "epoch": 0.44858982900288696,
+      "grad_norm": 0.698450418542813,
+      "learning_rate": 9.995535874075692e-06,
+      "loss": 0.0886,
+      "step": 1010
+    },
+    {
+      "epoch": 0.44903397734843437,
+      "grad_norm": 0.6685872712573397,
+      "learning_rate": 9.99545360024419e-06,
+      "loss": 0.097,
+      "step": 1011
+    },
+    {
+      "epoch": 0.44947812569398177,
+      "grad_norm": 0.7945297170711836,
+      "learning_rate": 9.995370575511151e-06,
+      "loss": 0.0841,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4499222740395292,
+      "grad_norm": 0.6520316868875402,
+      "learning_rate": 9.99528679988906e-06,
+      "loss": 0.0787,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4503664223850766,
+      "grad_norm": 0.572444172084648,
+      "learning_rate": 9.995202273390505e-06,
+      "loss": 0.0594,
+      "step": 1014
+    },
+    {
+      "epoch": 0.450810570730624,
+      "grad_norm": 0.6063370258668184,
+      "learning_rate": 9.995116996028197e-06,
+      "loss": 0.0683,
+      "step": 1015
+    },
+    {
+      "epoch": 0.45125471907617143,
+      "grad_norm": 0.7022055879284854,
+      "learning_rate": 9.995030967814952e-06,
+      "loss": 0.0828,
+      "step": 1016
+    },
+    {
+      "epoch": 0.45169886742171883,
+      "grad_norm": 0.6787805656005202,
+      "learning_rate": 9.994944188763701e-06,
+      "loss": 0.0886,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4521430157672663,
+      "grad_norm": 0.6743594234424202,
+      "learning_rate": 9.994856658887491e-06,
+      "loss": 0.0858,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4525871641128137,
+      "grad_norm": 0.701751582175111,
+      "learning_rate": 9.994768378199476e-06,
+      "loss": 0.0877,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4530313124583611,
+      "grad_norm": 0.7777579093243342,
+      "learning_rate": 9.994679346712927e-06,
+      "loss": 0.0809,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4534754608039085,
+      "grad_norm": 0.7407021400306011,
+      "learning_rate": 9.994589564441229e-06,
+      "loss": 0.0667,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4539196091494559,
+      "grad_norm": 0.8640591667177118,
+      "learning_rate": 9.994499031397874e-06,
+      "loss": 0.0779,
+      "step": 1022
+    },
+    {
+      "epoch": 0.45436375749500335,
+      "grad_norm": 0.741626283806636,
+      "learning_rate": 9.994407747596474e-06,
+      "loss": 0.0598,
+      "step": 1023
+    },
+    {
+      "epoch": 0.45480790584055075,
+      "grad_norm": 0.7426285442565452,
+      "learning_rate": 9.994315713050749e-06,
+      "loss": 0.0763,
+      "step": 1024
+    },
+    {
+      "epoch": 0.45525205418609815,
+      "grad_norm": 1.1459706806907142,
+      "learning_rate": 9.994222927774535e-06,
+      "loss": 0.0928,
+      "step": 1025
+    },
+    {
+      "epoch": 0.45569620253164556,
+      "grad_norm": 0.9102603719521771,
+      "learning_rate": 9.994129391781777e-06,
+      "loss": 0.0695,
+      "step": 1026
+    },
+    {
+      "epoch": 0.45614035087719296,
+      "grad_norm": 0.8406791604735616,
+      "learning_rate": 9.994035105086536e-06,
+      "loss": 0.0655,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4565844992227404,
+      "grad_norm": 0.6764230445649027,
+      "learning_rate": 9.993940067702985e-06,
+      "loss": 0.0566,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4570286475682878,
+      "grad_norm": 1.1241952343326866,
+      "learning_rate": 9.993844279645411e-06,
+      "loss": 0.1138,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4574727959138352,
+      "grad_norm": 0.8203260077691763,
+      "learning_rate": 9.993747740928207e-06,
+      "loss": 0.0639,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4579169442593826,
+      "grad_norm": 0.9349985900242503,
+      "learning_rate": 9.993650451565892e-06,
+      "loss": 0.078,
+      "step": 1031
+    },
+    {
+      "epoch": 0.45836109260493,
+      "grad_norm": 0.9978670067027727,
+      "learning_rate": 9.993552411573088e-06,
+      "loss": 0.0952,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4588052409504775,
+      "grad_norm": 1.1441082866740984,
+      "learning_rate": 9.993453620964529e-06,
+      "loss": 0.0897,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4592493892960249,
+      "grad_norm": 1.0696466066343144,
+      "learning_rate": 9.993354079755066e-06,
+      "loss": 0.0903,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4596935376415723,
+      "grad_norm": 0.6531078727573577,
+      "learning_rate": 9.993253787959664e-06,
+      "loss": 0.0673,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4601376859871197,
+      "grad_norm": 1.4198972391798237,
+      "learning_rate": 9.993152745593398e-06,
+      "loss": 0.083,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4605818343326671,
+      "grad_norm": 1.070478952947314,
+      "learning_rate": 9.993050952671453e-06,
+      "loss": 0.0687,
+      "step": 1037
+    },
+    {
+      "epoch": 0.46102598267821454,
+      "grad_norm": 0.7417628375073222,
+      "learning_rate": 9.992948409209134e-06,
+      "loss": 0.0802,
+      "step": 1038
+    },
+    {
+      "epoch": 0.46147013102376194,
+      "grad_norm": 0.6201081139409721,
+      "learning_rate": 9.992845115221855e-06,
+      "loss": 0.0612,
+      "step": 1039
+    },
+    {
+      "epoch": 0.46191427936930934,
+      "grad_norm": 0.8344957050457038,
+      "learning_rate": 9.992741070725137e-06,
+      "loss": 0.0897,
+      "step": 1040
+    },
+    {
+      "epoch": 0.46235842771485675,
+      "grad_norm": 0.5628453325687238,
+      "learning_rate": 9.992636275734629e-06,
+      "loss": 0.069,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4628025760604042,
+      "grad_norm": 0.6969650084016832,
+      "learning_rate": 9.992530730266078e-06,
+      "loss": 0.073,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4632467244059516,
+      "grad_norm": 1.0860876745728663,
+      "learning_rate": 9.992424434335348e-06,
+      "loss": 0.0921,
+      "step": 1043
+    },
+    {
+      "epoch": 0.463690872751499,
+      "grad_norm": 0.5795193360907794,
+      "learning_rate": 9.99231738795842e-06,
+      "loss": 0.0504,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4641350210970464,
+      "grad_norm": 0.6583352279848995,
+      "learning_rate": 9.992209591151386e-06,
+      "loss": 0.0831,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4645791694425938,
+      "grad_norm": 0.5908927184246742,
+      "learning_rate": 9.992101043930444e-06,
+      "loss": 0.0996,
+      "step": 1046
+    },
+    {
+      "epoch": 0.46502331778814127,
+      "grad_norm": 0.7629134106102797,
+      "learning_rate": 9.991991746311916e-06,
+      "loss": 0.0595,
+      "step": 1047
+    },
+    {
+      "epoch": 0.46546746613368867,
+      "grad_norm": 0.8529582668397104,
+      "learning_rate": 9.991881698312229e-06,
+      "loss": 0.076,
+      "step": 1048
+    },
+    {
+      "epoch": 0.46591161447923607,
+      "grad_norm": 0.6804037477070127,
+      "learning_rate": 9.991770899947925e-06,
+      "loss": 0.075,
+      "step": 1049
+    },
+    {
+      "epoch": 0.46635576282478347,
+      "grad_norm": 0.8612466864991256,
+      "learning_rate": 9.991659351235662e-06,
+      "loss": 0.0796,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4667999111703309,
+      "grad_norm": 0.6453565757322578,
+      "learning_rate": 9.991547052192203e-06,
+      "loss": 0.0625,
+      "step": 1051
+    },
+    {
+      "epoch": 0.46724405951587833,
+      "grad_norm": 1.7133852156999947,
+      "learning_rate": 9.99143400283443e-06,
+      "loss": 0.097,
+      "step": 1052
+    },
+    {
+      "epoch": 0.46768820786142573,
+      "grad_norm": 0.7043109079290237,
+      "learning_rate": 9.991320203179338e-06,
+      "loss": 0.0556,
+      "step": 1053
+    },
+    {
+      "epoch": 0.46813235620697313,
+      "grad_norm": 0.9119065164949084,
+      "learning_rate": 9.991205653244032e-06,
+      "loss": 0.0618,
+      "step": 1054
+    },
+    {
+      "epoch": 0.46857650455252053,
+      "grad_norm": 0.5740417394473022,
+      "learning_rate": 9.991090353045729e-06,
+      "loss": 0.0492,
+      "step": 1055
+    },
+    {
+      "epoch": 0.46902065289806794,
+      "grad_norm": 0.8813556475853607,
+      "learning_rate": 9.990974302601762e-06,
+      "loss": 0.0603,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4694648012436154,
+      "grad_norm": 0.6951634771023119,
+      "learning_rate": 9.990857501929577e-06,
+      "loss": 0.0641,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4699089495891628,
+      "grad_norm": 0.9703272601786429,
+      "learning_rate": 9.990739951046729e-06,
+      "loss": 0.1064,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4703530979347102,
+      "grad_norm": 1.1076310196170094,
+      "learning_rate": 9.99062164997089e-06,
+      "loss": 0.077,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4707972462802576,
+      "grad_norm": 0.813678875992106,
+      "learning_rate": 9.990502598719837e-06,
+      "loss": 0.0579,
+      "step": 1060
+    },
+    {
+      "epoch": 0.471241394625805,
+      "grad_norm": 0.8445949631628347,
+      "learning_rate": 9.990382797311474e-06,
+      "loss": 0.0922,
+      "step": 1061
+    },
+    {
+      "epoch": 0.47168554297135246,
+      "grad_norm": 0.6432477880049939,
+      "learning_rate": 9.990262245763802e-06,
+      "loss": 0.066,
+      "step": 1062
+    },
+    {
+      "epoch": 0.47212969131689986,
+      "grad_norm": 1.0190163753358932,
+      "learning_rate": 9.990140944094946e-06,
+      "loss": 0.1059,
+      "step": 1063
+    },
+    {
+      "epoch": 0.47257383966244726,
+      "grad_norm": 0.9736504358272349,
+      "learning_rate": 9.990018892323138e-06,
+      "loss": 0.0716,
+      "step": 1064
+    },
+    {
+      "epoch": 0.47301798800799466,
+      "grad_norm": 0.8551545600086438,
+      "learning_rate": 9.989896090466725e-06,
+      "loss": 0.0721,
+      "step": 1065
+    },
+    {
+      "epoch": 0.47346213635354206,
+      "grad_norm": 0.7846245356864615,
+      "learning_rate": 9.989772538544167e-06,
+      "loss": 0.0837,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4739062846990895,
+      "grad_norm": 0.6875499187875691,
+      "learning_rate": 9.989648236574035e-06,
+      "loss": 0.0604,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4743504330446369,
+      "grad_norm": 1.010871584074772,
+      "learning_rate": 9.989523184575013e-06,
+      "loss": 0.0836,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4747945813901843,
+      "grad_norm": 0.8312720125551346,
+      "learning_rate": 9.989397382565898e-06,
+      "loss": 0.0712,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4752387297357317,
+      "grad_norm": 0.96789104638543,
+      "learning_rate": 9.989270830565603e-06,
+      "loss": 0.0784,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4756828780812791,
+      "grad_norm": 0.7404878797473026,
+      "learning_rate": 9.989143528593149e-06,
+      "loss": 0.0872,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4761270264268266,
+      "grad_norm": 0.6280381360800792,
+      "learning_rate": 9.98901547666767e-06,
+      "loss": 0.0661,
+      "step": 1072
+    },
+    {
+      "epoch": 0.476571174772374,
+      "grad_norm": 0.8234412511968835,
+      "learning_rate": 9.988886674808418e-06,
+      "loss": 0.0621,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4770153231179214,
+      "grad_norm": 0.5856951522424453,
+      "learning_rate": 9.988757123034753e-06,
+      "loss": 0.0657,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4774594714634688,
+      "grad_norm": 0.9863728037623454,
+      "learning_rate": 9.988626821366147e-06,
+      "loss": 0.0787,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4779036198090162,
+      "grad_norm": 0.8043968430169456,
+      "learning_rate": 9.988495769822188e-06,
+      "loss": 0.0712,
+      "step": 1076
+    },
+    {
+      "epoch": 0.47834776815456365,
+      "grad_norm": 0.7330127963179667,
+      "learning_rate": 9.988363968422577e-06,
+      "loss": 0.0749,
+      "step": 1077
+    },
+    {
+      "epoch": 0.47879191650011105,
+      "grad_norm": 0.8101029322049607,
+      "learning_rate": 9.988231417187122e-06,
+      "loss": 0.0901,
+      "step": 1078
+    },
+    {
+      "epoch": 0.47923606484565845,
+      "grad_norm": 0.9467235647843469,
+      "learning_rate": 9.98809811613575e-06,
+      "loss": 0.0923,
+      "step": 1079
+    },
+    {
+      "epoch": 0.47968021319120585,
+      "grad_norm": 0.7955200227969057,
+      "learning_rate": 9.9879640652885e-06,
+      "loss": 0.0802,
+      "step": 1080
+    },
+    {
+      "epoch": 0.48012436153675325,
+      "grad_norm": 1.0074849821784144,
+      "learning_rate": 9.987829264665518e-06,
+      "loss": 0.1319,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4805685098823007,
+      "grad_norm": 0.680242996169655,
+      "learning_rate": 9.98769371428707e-06,
+      "loss": 0.0659,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4810126582278481,
+      "grad_norm": 1.1307628357255757,
+      "learning_rate": 9.98755741417353e-06,
+      "loss": 0.0962,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4814568065733955,
+      "grad_norm": 0.7637117418782801,
+      "learning_rate": 9.987420364345388e-06,
+      "loss": 0.0596,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4819009549189429,
+      "grad_norm": 0.8434389453780773,
+      "learning_rate": 9.987282564823242e-06,
+      "loss": 0.0832,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4823451032644903,
+      "grad_norm": 1.3802825505619392,
+      "learning_rate": 9.98714401562781e-06,
+      "loss": 0.1092,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4827892516100378,
+      "grad_norm": 0.7526559659659204,
+      "learning_rate": 9.987004716779914e-06,
+      "loss": 0.0649,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4832333999555852,
+      "grad_norm": 0.8564554305307791,
+      "learning_rate": 9.986864668300494e-06,
+      "loss": 0.0745,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4836775483011326,
+      "grad_norm": 0.6121214143253134,
+      "learning_rate": 9.986723870210605e-06,
+      "loss": 0.0648,
+      "step": 1089
+    },
+    {
+      "epoch": 0.48412169664668,
+      "grad_norm": 1.2737874840599017,
+      "learning_rate": 9.986582322531406e-06,
+      "loss": 0.1167,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4845658449922274,
+      "grad_norm": 1.1300452790006459,
+      "learning_rate": 9.986440025284177e-06,
+      "loss": 0.1004,
+      "step": 1091
+    },
+    {
+      "epoch": 0.48500999333777484,
+      "grad_norm": 0.695366891600787,
+      "learning_rate": 9.986296978490308e-06,
+      "loss": 0.0631,
+      "step": 1092
+    },
+    {
+      "epoch": 0.48545414168332224,
+      "grad_norm": 0.907315227077863,
+      "learning_rate": 9.9861531821713e-06,
+      "loss": 0.0975,
+      "step": 1093
+    },
+    {
+      "epoch": 0.48589829002886964,
+      "grad_norm": 0.8452739013500794,
+      "learning_rate": 9.986008636348771e-06,
+      "loss": 0.0725,
+      "step": 1094
+    },
+    {
+      "epoch": 0.48634243837441704,
+      "grad_norm": 0.7641747681242534,
+      "learning_rate": 9.985863341044444e-06,
+      "loss": 0.0745,
+      "step": 1095
+    },
+    {
+      "epoch": 0.48678658671996444,
+      "grad_norm": 1.0767424752885635,
+      "learning_rate": 9.985717296280165e-06,
+      "loss": 0.1343,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4872307350655119,
+      "grad_norm": 1.082068622065411,
+      "learning_rate": 9.985570502077881e-06,
+      "loss": 0.0709,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4876748834110593,
+      "grad_norm": 1.1385183378270511,
+      "learning_rate": 9.98542295845966e-06,
+      "loss": 0.1136,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4881190317566067,
+      "grad_norm": 0.8790710582646716,
+      "learning_rate": 9.985274665447682e-06,
+      "loss": 0.0796,
+      "step": 1099
+    },
+    {
+      "epoch": 0.4885631801021541,
+      "grad_norm": 0.7645639481923898,
+      "learning_rate": 9.985125623064238e-06,
+      "loss": 0.1039,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4890073284477015,
+      "grad_norm": 0.9695248420140831,
+      "learning_rate": 9.98497583133173e-06,
+      "loss": 0.0667,
+      "step": 1101
+    },
+    {
+      "epoch": 0.48945147679324896,
+      "grad_norm": 0.8791125306416613,
+      "learning_rate": 9.984825290272673e-06,
+      "loss": 0.0703,
+      "step": 1102
+    },
+    {
+      "epoch": 0.48989562513879636,
+      "grad_norm": 0.6608187335865532,
+      "learning_rate": 9.984673999909698e-06,
+      "loss": 0.0607,
+      "step": 1103
+    },
+    {
+      "epoch": 0.49033977348434377,
+      "grad_norm": 0.7238746077559932,
+      "learning_rate": 9.984521960265545e-06,
+      "loss": 0.0582,
+      "step": 1104
+    },
+    {
+      "epoch": 0.49078392182989117,
+      "grad_norm": 0.7543190655536871,
+      "learning_rate": 9.98436917136307e-06,
+      "loss": 0.075,
+      "step": 1105
+    },
+    {
+      "epoch": 0.49122807017543857,
+      "grad_norm": 0.9831510577501872,
+      "learning_rate": 9.98421563322524e-06,
+      "loss": 0.0832,
+      "step": 1106
+    },
+    {
+      "epoch": 0.491672218520986,
+      "grad_norm": 1.0953392860217965,
+      "learning_rate": 9.984061345875133e-06,
+      "loss": 0.1055,
+      "step": 1107
+    },
+    {
+      "epoch": 0.49211636686653343,
+      "grad_norm": 1.0004839390300868,
+      "learning_rate": 9.983906309335942e-06,
+      "loss": 0.0882,
+      "step": 1108
+    },
+    {
+      "epoch": 0.49256051521208083,
+      "grad_norm": 0.7389464810035322,
+      "learning_rate": 9.98375052363097e-06,
+      "loss": 0.0608,
+      "step": 1109
+    },
+    {
+      "epoch": 0.49300466355762823,
+      "grad_norm": 0.6469655367019491,
+      "learning_rate": 9.983593988783634e-06,
+      "loss": 0.0687,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4934488119031757,
+      "grad_norm": 0.7833813336297293,
+      "learning_rate": 9.983436704817466e-06,
+      "loss": 0.0902,
+      "step": 1111
+    },
+    {
+      "epoch": 0.4938929602487231,
+      "grad_norm": 1.1676871219264449,
+      "learning_rate": 9.983278671756107e-06,
+      "loss": 0.0741,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4943371085942705,
+      "grad_norm": 0.5948813388699934,
+      "learning_rate": 9.983119889623314e-06,
+      "loss": 0.0584,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4947812569398179,
+      "grad_norm": 1.0932192784955042,
+      "learning_rate": 9.982960358442952e-06,
+      "loss": 0.0814,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4952254052853653,
+      "grad_norm": 0.9576191702524788,
+      "learning_rate": 9.982800078239004e-06,
+      "loss": 0.0939,
+      "step": 1115
+    },
+    {
+      "epoch": 0.49566955363091275,
+      "grad_norm": 0.8091517385774816,
+      "learning_rate": 9.982639049035559e-06,
+      "loss": 0.0894,
+      "step": 1116
+    },
+    {
+      "epoch": 0.49611370197646015,
+      "grad_norm": 1.1318080237230808,
+      "learning_rate": 9.982477270856827e-06,
+      "loss": 0.0743,
+      "step": 1117
+    },
+    {
+      "epoch": 0.49655785032200755,
+      "grad_norm": 1.2019606615618712,
+      "learning_rate": 9.982314743727121e-06,
+      "loss": 0.107,
+      "step": 1118
+    },
+    {
+      "epoch": 0.49700199866755496,
+      "grad_norm": 0.9414473018138827,
+      "learning_rate": 9.982151467670876e-06,
+      "loss": 0.0746,
+      "step": 1119
+    },
+    {
+      "epoch": 0.49744614701310236,
+      "grad_norm": 0.7821900901787131,
+      "learning_rate": 9.981987442712634e-06,
+      "loss": 0.062,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4978902953586498,
+      "grad_norm": 0.8874834331056065,
+      "learning_rate": 9.981822668877048e-06,
+      "loss": 0.0964,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4983344437041972,
+      "grad_norm": 0.763549676147823,
+      "learning_rate": 9.98165714618889e-06,
+      "loss": 0.0683,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4987785920497446,
+      "grad_norm": 0.7237088556818837,
+      "learning_rate": 9.98149087467304e-06,
+      "loss": 0.0706,
+      "step": 1123
+    },
+    {
+      "epoch": 0.499222740395292,
+      "grad_norm": 0.5389621815014336,
+      "learning_rate": 9.98132385435449e-06,
+      "loss": 0.0739,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4996668887408394,
+      "grad_norm": 0.5780461463044186,
+      "learning_rate": 9.981156085258347e-06,
+      "loss": 0.0532,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5001110370863868,
+      "grad_norm": 0.9337676739733405,
+      "learning_rate": 9.980987567409829e-06,
+      "loss": 0.0728,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5005551854319342,
+      "grad_norm": 0.7121233826282255,
+      "learning_rate": 9.980818300834267e-06,
+      "loss": 0.0684,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5009993337774816,
+      "grad_norm": 0.7441687384799486,
+      "learning_rate": 9.980648285557106e-06,
+      "loss": 0.0692,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5014434821230291,
+      "grad_norm": 0.6423868990281854,
+      "learning_rate": 9.980477521603901e-06,
+      "loss": 0.0677,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5018876304685765,
+      "grad_norm": 0.5495871418316753,
+      "learning_rate": 9.98030600900032e-06,
+      "loss": 0.0728,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5023317788141239,
+      "grad_norm": 0.6241490380000925,
+      "learning_rate": 9.980133747772148e-06,
+      "loss": 0.0662,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5027759271596713,
+      "grad_norm": 0.7343652098087937,
+      "learning_rate": 9.979960737945273e-06,
+      "loss": 0.0678,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5032200755052187,
+      "grad_norm": 1.1015252671394224,
+      "learning_rate": 9.979786979545704e-06,
+      "loss": 0.0993,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5036642238507661,
+      "grad_norm": 0.8816390886875364,
+      "learning_rate": 9.979612472599563e-06,
+      "loss": 0.0839,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5041083721963135,
+      "grad_norm": 0.7833322958956014,
+      "learning_rate": 9.979437217133077e-06,
+      "loss": 0.082,
+      "step": 1135
+    },
+    {
+      "epoch": 0.504552520541861,
+      "grad_norm": 0.9520003222392113,
+      "learning_rate": 9.979261213172592e-06,
+      "loss": 0.0768,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5049966688874084,
+      "grad_norm": 0.6650156355994222,
+      "learning_rate": 9.979084460744563e-06,
+      "loss": 0.0594,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5054408172329558,
+      "grad_norm": 0.7702494870526182,
+      "learning_rate": 9.97890695987556e-06,
+      "loss": 0.0744,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5058849655785033,
+      "grad_norm": 0.6441998434508551,
+      "learning_rate": 9.978728710592265e-06,
+      "loss": 0.0639,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8638339516257324,
+      "learning_rate": 9.97854971292147e-06,
+      "loss": 0.0737,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5067732622695981,
+      "grad_norm": 0.6599486193027585,
+      "learning_rate": 9.978369966890082e-06,
+      "loss": 0.0677,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5072174106151455,
+      "grad_norm": 0.5684535397639588,
+      "learning_rate": 9.978189472525121e-06,
+      "loss": 0.0667,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5076615589606929,
+      "grad_norm": 0.7930545362878336,
+      "learning_rate": 9.978008229853717e-06,
+      "loss": 0.0768,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5081057073062403,
+      "grad_norm": 1.013824145483496,
+      "learning_rate": 9.977826238903116e-06,
+      "loss": 0.0878,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5085498556517877,
+      "grad_norm": 0.7663207754691962,
+      "learning_rate": 9.97764349970067e-06,
+      "loss": 0.0757,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5089940039973351,
+      "grad_norm": 0.7907421181118522,
+      "learning_rate": 9.977460012273854e-06,
+      "loss": 0.0784,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5094381523428825,
+      "grad_norm": 0.6096514964784717,
+      "learning_rate": 9.977275776650244e-06,
+      "loss": 0.0806,
+      "step": 1147
+    },
+    {
+      "epoch": 0.50988230068843,
+      "grad_norm": 0.5159120034000472,
+      "learning_rate": 9.977090792857536e-06,
+      "loss": 0.0598,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5103264490339774,
+      "grad_norm": 0.7102391658769992,
+      "learning_rate": 9.976905060923536e-06,
+      "loss": 0.077,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5107705973795248,
+      "grad_norm": 0.7376867178754969,
+      "learning_rate": 9.97671858087616e-06,
+      "loss": 0.0726,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5112147457250722,
+      "grad_norm": 0.7790796748709151,
+      "learning_rate": 9.976531352743445e-06,
+      "loss": 0.0806,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5116588940706196,
+      "grad_norm": 0.618833561044351,
+      "learning_rate": 9.97634337655353e-06,
+      "loss": 0.0654,
+      "step": 1152
+    },
+    {
+      "epoch": 0.512103042416167,
+      "grad_norm": 0.808662756842487,
+      "learning_rate": 9.976154652334673e-06,
+      "loss": 0.1081,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5125471907617144,
+      "grad_norm": 1.0180790521960883,
+      "learning_rate": 9.97596518011524e-06,
+      "loss": 0.1018,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5129913391072618,
+      "grad_norm": 0.638695304240369,
+      "learning_rate": 9.975774959923717e-06,
+      "loss": 0.0655,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5134354874528092,
+      "grad_norm": 0.7566614134333154,
+      "learning_rate": 9.975583991788691e-06,
+      "loss": 0.0658,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5138796357983566,
+      "grad_norm": 0.7034717435901088,
+      "learning_rate": 9.97539227573887e-06,
+      "loss": 0.0815,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5143237841439041,
+      "grad_norm": 0.7342548435589596,
+      "learning_rate": 9.975199811803073e-06,
+      "loss": 0.0755,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5147679324894515,
+      "grad_norm": 0.881634196160518,
+      "learning_rate": 9.975006600010233e-06,
+      "loss": 0.0649,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5152120808349989,
+      "grad_norm": 0.82644127591562,
+      "learning_rate": 9.97481264038939e-06,
+      "loss": 0.0589,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5156562291805463,
+      "grad_norm": 0.8462515915216339,
+      "learning_rate": 9.974617932969697e-06,
+      "loss": 0.0735,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5161003775260937,
+      "grad_norm": 0.6116708783372564,
+      "learning_rate": 9.974422477780426e-06,
+      "loss": 0.0593,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5165445258716411,
+      "grad_norm": 0.7646830820557489,
+      "learning_rate": 9.974226274850956e-06,
+      "loss": 0.0866,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5169886742171885,
+      "grad_norm": 0.6551362409279565,
+      "learning_rate": 9.97402932421078e-06,
+      "loss": 0.0742,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5174328225627359,
+      "grad_norm": 0.5323199794273931,
+      "learning_rate": 9.973831625889501e-06,
+      "loss": 0.0599,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5178769709082833,
+      "grad_norm": 0.694095375162108,
+      "learning_rate": 9.97363317991684e-06,
+      "loss": 0.0646,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5183211192538307,
+      "grad_norm": 0.9040017314033096,
+      "learning_rate": 9.973433986322625e-06,
+      "loss": 0.0755,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5187652675993782,
+      "grad_norm": 0.7059881945266824,
+      "learning_rate": 9.973234045136798e-06,
+      "loss": 0.0869,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5192094159449256,
+      "grad_norm": 0.7791362124843618,
+      "learning_rate": 9.973033356389412e-06,
+      "loss": 0.0912,
+      "step": 1169
+    },
+    {
+      "epoch": 0.519653564290473,
+      "grad_norm": 0.7295114970239928,
+      "learning_rate": 9.972831920110635e-06,
+      "loss": 0.0769,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5200977126360204,
+      "grad_norm": 0.5612528230166707,
+      "learning_rate": 9.972629736330748e-06,
+      "loss": 0.0679,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5205418609815678,
+      "grad_norm": 0.7254809001904728,
+      "learning_rate": 9.972426805080141e-06,
+      "loss": 0.0715,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5209860093271153,
+      "grad_norm": 0.7275192896210665,
+      "learning_rate": 9.97222312638932e-06,
+      "loss": 0.0841,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5214301576726627,
+      "grad_norm": 0.8132581046482065,
+      "learning_rate": 9.972018700288898e-06,
+      "loss": 0.0715,
+      "step": 1174
+    },
+    {
+      "epoch": 0.52187430601821,
+      "grad_norm": 0.9966008387669913,
+      "learning_rate": 9.971813526809606e-06,
+      "loss": 0.0844,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5223184543637575,
+      "grad_norm": 0.621117239882631,
+      "learning_rate": 9.971607605982285e-06,
+      "loss": 0.0572,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5227626027093049,
+      "grad_norm": 0.8375409793845516,
+      "learning_rate": 9.971400937837887e-06,
+      "loss": 0.0728,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5232067510548524,
+      "grad_norm": 0.9177354811192968,
+      "learning_rate": 9.97119352240748e-06,
+      "loss": 0.0878,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5236508994003998,
+      "grad_norm": 0.7412042248938593,
+      "learning_rate": 9.97098535972224e-06,
+      "loss": 0.0649,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5240950477459472,
+      "grad_norm": 0.75392664768525,
+      "learning_rate": 9.970776449813457e-06,
+      "loss": 0.0682,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5245391960914946,
+      "grad_norm": 0.7437991387327917,
+      "learning_rate": 9.970566792712537e-06,
+      "loss": 0.0646,
+      "step": 1181
+    },
+    {
+      "epoch": 0.524983344437042,
+      "grad_norm": 0.708247278848663,
+      "learning_rate": 9.970356388450992e-06,
+      "loss": 0.0781,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5254274927825894,
+      "grad_norm": 0.508933410042531,
+      "learning_rate": 9.97014523706045e-06,
+      "loss": 0.0535,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5258716411281368,
+      "grad_norm": 1.0021738372524498,
+      "learning_rate": 9.96993333857265e-06,
+      "loss": 0.0925,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.8285101542802605,
+      "learning_rate": 9.969720693019447e-06,
+      "loss": 0.0653,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5267599378192316,
+      "grad_norm": 0.6010536869971059,
+      "learning_rate": 9.9695073004328e-06,
+      "loss": 0.0601,
+      "step": 1186
+    },
+    {
+      "epoch": 0.527204086164779,
+      "grad_norm": 0.43111203380102636,
+      "learning_rate": 9.969293160844793e-06,
+      "loss": 0.0435,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5276482345103265,
+      "grad_norm": 0.7272050961589046,
+      "learning_rate": 9.969078274287607e-06,
+      "loss": 0.0582,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5280923828558739,
+      "grad_norm": 1.0434151274290433,
+      "learning_rate": 9.968862640793547e-06,
+      "loss": 0.0882,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5285365312014213,
+      "grad_norm": 0.6610761510472204,
+      "learning_rate": 9.968646260395027e-06,
+      "loss": 0.0701,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5289806795469687,
+      "grad_norm": 0.7703015492278306,
+      "learning_rate": 9.96842913312457e-06,
+      "loss": 0.0755,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5294248278925161,
+      "grad_norm": 1.1334603205830818,
+      "learning_rate": 9.968211259014817e-06,
+      "loss": 0.102,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5298689762380635,
+      "grad_norm": 0.5680812506173959,
+      "learning_rate": 9.967992638098517e-06,
+      "loss": 0.0592,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5303131245836109,
+      "grad_norm": 0.7128376996533422,
+      "learning_rate": 9.96777327040853e-06,
+      "loss": 0.0757,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5307572729291583,
+      "grad_norm": 0.6547153588619065,
+      "learning_rate": 9.967553155977833e-06,
+      "loss": 0.0513,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5312014212747057,
+      "grad_norm": 0.8276344454564539,
+      "learning_rate": 9.967332294839514e-06,
+      "loss": 0.0722,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5316455696202531,
+      "grad_norm": 0.5942789130977645,
+      "learning_rate": 9.967110687026769e-06,
+      "loss": 0.0735,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5320897179658006,
+      "grad_norm": 0.858420102793841,
+      "learning_rate": 9.966888332572913e-06,
+      "loss": 0.0888,
+      "step": 1198
+    },
+    {
+      "epoch": 0.532533866311348,
+      "grad_norm": 0.5306419562771065,
+      "learning_rate": 9.966665231511367e-06,
+      "loss": 0.0735,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5329780146568954,
+      "grad_norm": 0.8818696703835058,
+      "learning_rate": 9.96644138387567e-06,
+      "loss": 0.0836,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5334221630024428,
+      "grad_norm": 0.6953800899178525,
+      "learning_rate": 9.966216789699466e-06,
+      "loss": 0.0609,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5338663113479902,
+      "grad_norm": 0.6141704558915728,
+      "learning_rate": 9.965991449016517e-06,
+      "loss": 0.0593,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5343104596935376,
+      "grad_norm": 0.7784143890008246,
+      "learning_rate": 9.965765361860696e-06,
+      "loss": 0.0682,
+      "step": 1203
+    },
+    {
+      "epoch": 0.534754608039085,
+      "grad_norm": 0.7591572831776785,
+      "learning_rate": 9.965538528265986e-06,
+      "loss": 0.0713,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5351987563846324,
+      "grad_norm": 1.0112012318529477,
+      "learning_rate": 9.965310948266488e-06,
+      "loss": 0.0885,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5356429047301798,
+      "grad_norm": 0.540867345550289,
+      "learning_rate": 9.965082621896407e-06,
+      "loss": 0.0683,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5360870530757272,
+      "grad_norm": 0.6987948599480821,
+      "learning_rate": 9.964853549190067e-06,
+      "loss": 0.0623,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5365312014212748,
+      "grad_norm": 1.0949702611838528,
+      "learning_rate": 9.9646237301819e-06,
+      "loss": 0.0715,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5369753497668222,
+      "grad_norm": 0.5434822885792773,
+      "learning_rate": 9.964393164906452e-06,
+      "loss": 0.077,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5374194981123696,
+      "grad_norm": 0.785351221183312,
+      "learning_rate": 9.964161853398381e-06,
+      "loss": 0.07,
+      "step": 1210
+    },
+    {
+      "epoch": 0.537863646457917,
+      "grad_norm": 0.5762754098127977,
+      "learning_rate": 9.963929795692458e-06,
+      "loss": 0.0553,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5383077948034644,
+      "grad_norm": 0.7778031310614973,
+      "learning_rate": 9.963696991823563e-06,
+      "loss": 0.0671,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5387519431490118,
+      "grad_norm": 0.7236031416359869,
+      "learning_rate": 9.963463441826693e-06,
+      "loss": 0.0861,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5391960914945592,
+      "grad_norm": 0.8183548985521056,
+      "learning_rate": 9.963229145736952e-06,
+      "loss": 0.0905,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5396402398401066,
+      "grad_norm": 0.8238577196866539,
+      "learning_rate": 9.96299410358956e-06,
+      "loss": 0.0791,
+      "step": 1215
+    },
+    {
+      "epoch": 0.540084388185654,
+      "grad_norm": 0.620826351541747,
+      "learning_rate": 9.962758315419847e-06,
+      "loss": 0.0627,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5405285365312015,
+      "grad_norm": 0.8069644430296897,
+      "learning_rate": 9.962521781263259e-06,
+      "loss": 0.0846,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5409726848767489,
+      "grad_norm": 0.6551620736321753,
+      "learning_rate": 9.962284501155347e-06,
+      "loss": 0.0626,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5414168332222963,
+      "grad_norm": 0.7163185717239511,
+      "learning_rate": 9.96204647513178e-06,
+      "loss": 0.0742,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5418609815678437,
+      "grad_norm": 0.6464830809225138,
+      "learning_rate": 9.96180770322834e-06,
+      "loss": 0.0597,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5423051299133911,
+      "grad_norm": 0.4702533761313221,
+      "learning_rate": 9.961568185480912e-06,
+      "loss": 0.0663,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5427492782589385,
+      "grad_norm": 0.5377074866144479,
+      "learning_rate": 9.961327921925506e-06,
+      "loss": 0.0682,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5431934266044859,
+      "grad_norm": 0.6242182477426378,
+      "learning_rate": 9.961086912598232e-06,
+      "loss": 0.0558,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5436375749500333,
+      "grad_norm": 0.919842928693314,
+      "learning_rate": 9.960845157535324e-06,
+      "loss": 0.0979,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5440817232955807,
+      "grad_norm": 0.4865673297809684,
+      "learning_rate": 9.960602656773118e-06,
+      "loss": 0.0591,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5445258716411281,
+      "grad_norm": 0.5994196236588832,
+      "learning_rate": 9.960359410348066e-06,
+      "loss": 0.06,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5449700199866756,
+      "grad_norm": 0.6813026653368286,
+      "learning_rate": 9.960115418296734e-06,
+      "loss": 0.0563,
+      "step": 1227
+    },
+    {
+      "epoch": 0.545414168332223,
+      "grad_norm": 0.4200726079302436,
+      "learning_rate": 9.959870680655797e-06,
+      "loss": 0.0448,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5458583166777704,
+      "grad_norm": 1.3484707743369218,
+      "learning_rate": 9.959625197462042e-06,
+      "loss": 0.058,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5463024650233178,
+      "grad_norm": 0.91837601072641,
+      "learning_rate": 9.959378968752371e-06,
+      "loss": 0.1046,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5467466133688652,
+      "grad_norm": 0.7675630782674875,
+      "learning_rate": 9.959131994563795e-06,
+      "loss": 0.0738,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5471907617144126,
+      "grad_norm": 0.5684812066450396,
+      "learning_rate": 9.958884274933442e-06,
+      "loss": 0.0699,
+      "step": 1232
+    },
+    {
+      "epoch": 0.54763491005996,
+      "grad_norm": 0.6471523630810303,
+      "learning_rate": 9.958635809898544e-06,
+      "loss": 0.0713,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5480790584055074,
+      "grad_norm": 0.6967540553492192,
+      "learning_rate": 9.95838659949645e-06,
+      "loss": 0.0776,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5485232067510548,
+      "grad_norm": 0.6469245863206669,
+      "learning_rate": 9.958136643764624e-06,
+      "loss": 0.0764,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5489673550966022,
+      "grad_norm": 0.8210874368087515,
+      "learning_rate": 9.957885942740635e-06,
+      "loss": 0.0779,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5494115034421497,
+      "grad_norm": 0.6891702613627725,
+      "learning_rate": 9.957634496462169e-06,
+      "loss": 0.0575,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5498556517876971,
+      "grad_norm": 0.5415934981486168,
+      "learning_rate": 9.957382304967024e-06,
+      "loss": 0.0653,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5502998001332445,
+      "grad_norm": 0.606098328678092,
+      "learning_rate": 9.957129368293108e-06,
+      "loss": 0.0578,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5507439484787919,
+      "grad_norm": 0.5816870040354246,
+      "learning_rate": 9.95687568647844e-06,
+      "loss": 0.0543,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5511880968243393,
+      "grad_norm": 0.7853352727315064,
+      "learning_rate": 9.956621259561152e-06,
+      "loss": 0.0731,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5516322451698867,
+      "grad_norm": 0.6165368112917523,
+      "learning_rate": 9.956366087579492e-06,
+      "loss": 0.0688,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5520763935154341,
+      "grad_norm": 0.8107389760561104,
+      "learning_rate": 9.956110170571816e-06,
+      "loss": 0.0823,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5525205418609815,
+      "grad_norm": 0.6920576051280234,
+      "learning_rate": 9.95585350857659e-06,
+      "loss": 0.0661,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5529646902065289,
+      "grad_norm": 0.6032293972841873,
+      "learning_rate": 9.9555961016324e-06,
+      "loss": 0.0629,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5534088385520763,
+      "grad_norm": 0.7423430185463719,
+      "learning_rate": 9.955337949777931e-06,
+      "loss": 0.0781,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5538529868976239,
+      "grad_norm": 0.7004081423242674,
+      "learning_rate": 9.955079053051992e-06,
+      "loss": 0.0695,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5542971352431713,
+      "grad_norm": 0.6468930634930644,
+      "learning_rate": 9.9548194114935e-06,
+      "loss": 0.0549,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5547412835887187,
+      "grad_norm": 0.4810505823476616,
+      "learning_rate": 9.954559025141484e-06,
+      "loss": 0.0514,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5551854319342661,
+      "grad_norm": 0.802783907262409,
+      "learning_rate": 9.95429789403508e-06,
+      "loss": 0.1085,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5556295802798135,
+      "grad_norm": 0.5353100797059109,
+      "learning_rate": 9.954036018213548e-06,
+      "loss": 0.0448,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5560737286253609,
+      "grad_norm": 0.5946904373037035,
+      "learning_rate": 9.953773397716247e-06,
+      "loss": 0.0724,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5565178769709083,
+      "grad_norm": 1.0490812743358358,
+      "learning_rate": 9.953510032582652e-06,
+      "loss": 0.0675,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5569620253164557,
+      "grad_norm": 0.5087723175679687,
+      "learning_rate": 9.953245922852355e-06,
+      "loss": 0.0599,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5574061736620031,
+      "grad_norm": 0.5243033641589454,
+      "learning_rate": 9.952981068565055e-06,
+      "loss": 0.0569,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5578503220075505,
+      "grad_norm": 0.7607475927289213,
+      "learning_rate": 9.952715469760566e-06,
+      "loss": 0.0775,
+      "step": 1256
+    },
+    {
+      "epoch": 0.558294470353098,
+      "grad_norm": 0.8559494882081514,
+      "learning_rate": 9.952449126478808e-06,
+      "loss": 0.0866,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5587386186986454,
+      "grad_norm": 0.7656790089546245,
+      "learning_rate": 9.952182038759818e-06,
+      "loss": 0.0722,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5591827670441928,
+      "grad_norm": 0.5660284655001593,
+      "learning_rate": 9.951914206643744e-06,
+      "loss": 0.0564,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5596269153897402,
+      "grad_norm": 0.5835362013761695,
+      "learning_rate": 9.95164563017085e-06,
+      "loss": 0.0788,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5600710637352876,
+      "grad_norm": 0.6769019916334489,
+      "learning_rate": 9.951376309381502e-06,
+      "loss": 0.0594,
+      "step": 1261
+    },
+    {
+      "epoch": 0.560515212080835,
+      "grad_norm": 0.5253934078988655,
+      "learning_rate": 9.951106244316184e-06,
+      "loss": 0.053,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5609593604263824,
+      "grad_norm": 0.7300022918221285,
+      "learning_rate": 9.950835435015495e-06,
+      "loss": 0.0651,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5614035087719298,
+      "grad_norm": 0.5278089608881616,
+      "learning_rate": 9.95056388152014e-06,
+      "loss": 0.0555,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5618476571174772,
+      "grad_norm": 0.6301552291199896,
+      "learning_rate": 9.950291583870938e-06,
+      "loss": 0.0862,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5622918054630246,
+      "grad_norm": 0.7781648389786646,
+      "learning_rate": 9.950018542108818e-06,
+      "loss": 0.0699,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5627359538085721,
+      "grad_norm": 0.6889192633019509,
+      "learning_rate": 9.949744756274828e-06,
+      "loss": 0.059,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5631801021541195,
+      "grad_norm": 0.84114983491113,
+      "learning_rate": 9.94947022641012e-06,
+      "loss": 0.0855,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5636242504996669,
+      "grad_norm": 0.6070250007384902,
+      "learning_rate": 9.949194952555958e-06,
+      "loss": 0.0635,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5640683988452143,
+      "grad_norm": 0.8282717974485503,
+      "learning_rate": 9.948918934753724e-06,
+      "loss": 0.0606,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5645125471907617,
+      "grad_norm": 0.8327690010964344,
+      "learning_rate": 9.948642173044906e-06,
+      "loss": 0.0617,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5649566955363091,
+      "grad_norm": 0.6842942727397727,
+      "learning_rate": 9.948364667471106e-06,
+      "loss": 0.0795,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5654008438818565,
+      "grad_norm": 0.5302443265967214,
+      "learning_rate": 9.94808641807404e-06,
+      "loss": 0.0512,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5658449922274039,
+      "grad_norm": 0.6110171632056333,
+      "learning_rate": 9.94780742489553e-06,
+      "loss": 0.0574,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5662891405729513,
+      "grad_norm": 0.5491078913412093,
+      "learning_rate": 9.947527687977519e-06,
+      "loss": 0.0512,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5667332889184987,
+      "grad_norm": 0.6269674055110014,
+      "learning_rate": 9.94724720736205e-06,
+      "loss": 0.074,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5671774372640462,
+      "grad_norm": 0.7276246324646,
+      "learning_rate": 9.946965983091286e-06,
+      "loss": 0.1055,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5676215856095936,
+      "grad_norm": 0.7755784175713893,
+      "learning_rate": 9.946684015207501e-06,
+      "loss": 0.0878,
+      "step": 1278
+    },
+    {
+      "epoch": 0.568065733955141,
+      "grad_norm": 0.6682673092927641,
+      "learning_rate": 9.94640130375308e-06,
+      "loss": 0.0985,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5685098823006884,
+      "grad_norm": 0.71248575096518,
+      "learning_rate": 9.946117848770518e-06,
+      "loss": 0.0546,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5689540306462358,
+      "grad_norm": 0.6304803324875314,
+      "learning_rate": 9.945833650302423e-06,
+      "loss": 0.0645,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5693981789917832,
+      "grad_norm": 0.6709593851017193,
+      "learning_rate": 9.945548708391517e-06,
+      "loss": 0.0711,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5698423273373306,
+      "grad_norm": 0.6036373550762799,
+      "learning_rate": 9.94526302308063e-06,
+      "loss": 0.0766,
+      "step": 1283
+    },
+    {
+      "epoch": 0.570286475682878,
+      "grad_norm": 0.6055441845545156,
+      "learning_rate": 9.944976594412702e-06,
+      "loss": 0.0663,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5707306240284254,
+      "grad_norm": 1.1523396950548679,
+      "learning_rate": 9.944689422430794e-06,
+      "loss": 0.0876,
+      "step": 1285
+    },
+    {
+      "epoch": 0.571174772373973,
+      "grad_norm": 0.8498417793676747,
+      "learning_rate": 9.94440150717807e-06,
+      "loss": 0.0764,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5716189207195204,
+      "grad_norm": 0.5983413012513809,
+      "learning_rate": 9.944112848697809e-06,
+      "loss": 0.0564,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5720630690650678,
+      "grad_norm": 0.7260849111122668,
+      "learning_rate": 9.9438234470334e-06,
+      "loss": 0.0728,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5725072174106152,
+      "grad_norm": 1.0050849305541534,
+      "learning_rate": 9.943533302228346e-06,
+      "loss": 0.0711,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5729513657561626,
+      "grad_norm": 0.800396030085453,
+      "learning_rate": 9.943242414326263e-06,
+      "loss": 0.0724,
+      "step": 1290
+    },
+    {
+      "epoch": 0.57339551410171,
+      "grad_norm": 0.9224090719942251,
+      "learning_rate": 9.94295078337087e-06,
+      "loss": 0.1002,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5738396624472574,
+      "grad_norm": 0.8382142062568964,
+      "learning_rate": 9.942658409406012e-06,
+      "loss": 0.0756,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5742838107928048,
+      "grad_norm": 0.5876570577747893,
+      "learning_rate": 9.942365292475632e-06,
+      "loss": 0.0605,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5747279591383522,
+      "grad_norm": 0.9775545409759507,
+      "learning_rate": 9.942071432623794e-06,
+      "loss": 0.0786,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5751721074838996,
+      "grad_norm": 0.6598504226159471,
+      "learning_rate": 9.941776829894667e-06,
+      "loss": 0.0847,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5756162558294471,
+      "grad_norm": 0.7984485905609663,
+      "learning_rate": 9.941481484332537e-06,
+      "loss": 0.0699,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5760604041749945,
+      "grad_norm": 0.6762263367814108,
+      "learning_rate": 9.941185395981799e-06,
+      "loss": 0.0786,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5765045525205419,
+      "grad_norm": 0.7731297634993985,
+      "learning_rate": 9.940888564886959e-06,
+      "loss": 0.0673,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5769487008660893,
+      "grad_norm": 0.919743609676064,
+      "learning_rate": 9.940590991092639e-06,
+      "loss": 0.065,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5773928492116367,
+      "grad_norm": 0.6382780908912015,
+      "learning_rate": 9.940292674643564e-06,
+      "loss": 0.0578,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5778369975571841,
+      "grad_norm": 0.6620249742612588,
+      "learning_rate": 9.93999361558458e-06,
+      "loss": 0.0635,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5782811459027315,
+      "grad_norm": 0.7452135944776405,
+      "learning_rate": 9.93969381396064e-06,
+      "loss": 0.0806,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5787252942482789,
+      "grad_norm": 0.6403437438856001,
+      "learning_rate": 9.93939326981681e-06,
+      "loss": 0.0687,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5791694425938263,
+      "grad_norm": 1.1014917271306792,
+      "learning_rate": 9.939091983198266e-06,
+      "loss": 0.098,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5796135909393737,
+      "grad_norm": 1.0225638964559536,
+      "learning_rate": 9.938789954150296e-06,
+      "loss": 0.0644,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5800577392849212,
+      "grad_norm": 0.637383310530645,
+      "learning_rate": 9.9384871827183e-06,
+      "loss": 0.06,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5805018876304686,
+      "grad_norm": 0.6546930860338038,
+      "learning_rate": 9.93818366894779e-06,
+      "loss": 0.0681,
+      "step": 1307
+    },
+    {
+      "epoch": 0.580946035976016,
+      "grad_norm": 0.5341000978856885,
+      "learning_rate": 9.93787941288439e-06,
+      "loss": 0.0589,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5813901843215634,
+      "grad_norm": 0.5540761577564739,
+      "learning_rate": 9.937574414573834e-06,
+      "loss": 0.07,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5818343326671108,
+      "grad_norm": 0.6399098162524178,
+      "learning_rate": 9.937268674061968e-06,
+      "loss": 0.0681,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5822784810126582,
+      "grad_norm": 0.9788827379867293,
+      "learning_rate": 9.936962191394753e-06,
+      "loss": 0.0775,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5827226293582056,
+      "grad_norm": 0.5825969221377589,
+      "learning_rate": 9.936654966618255e-06,
+      "loss": 0.0662,
+      "step": 1312
+    },
+    {
+      "epoch": 0.583166777703753,
+      "grad_norm": 0.6990166610409554,
+      "learning_rate": 9.936346999778657e-06,
+      "loss": 0.0748,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5836109260493004,
+      "grad_norm": 0.7726036624010885,
+      "learning_rate": 9.93603829092225e-06,
+      "loss": 0.0806,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5840550743948478,
+      "grad_norm": 0.5892730393957506,
+      "learning_rate": 9.93572884009544e-06,
+      "loss": 0.0687,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5844992227403953,
+      "grad_norm": 0.7324361900970796,
+      "learning_rate": 9.935418647344741e-06,
+      "loss": 0.0722,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5849433710859427,
+      "grad_norm": 0.8954678705930711,
+      "learning_rate": 9.935107712716781e-06,
+      "loss": 0.0829,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5853875194314901,
+      "grad_norm": 0.5986143926701439,
+      "learning_rate": 9.9347960362583e-06,
+      "loss": 0.068,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5858316677770375,
+      "grad_norm": 0.6105531585151114,
+      "learning_rate": 9.934483618016148e-06,
+      "loss": 0.0719,
+      "step": 1319
+    },
+    {
+      "epoch": 0.586275816122585,
+      "grad_norm": 0.7814579585476528,
+      "learning_rate": 9.934170458037285e-06,
+      "loss": 0.0899,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5867199644681323,
+      "grad_norm": 0.6130620918486331,
+      "learning_rate": 9.933856556368785e-06,
+      "loss": 0.0663,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5871641128136798,
+      "grad_norm": 0.7911643262947309,
+      "learning_rate": 9.933541913057833e-06,
+      "loss": 0.0782,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5876082611592272,
+      "grad_norm": 0.7091228860294103,
+      "learning_rate": 9.933226528151725e-06,
+      "loss": 0.0637,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5880524095047746,
+      "grad_norm": 0.7999422262730185,
+      "learning_rate": 9.93291040169787e-06,
+      "loss": 0.0616,
+      "step": 1324
+    },
+    {
+      "epoch": 0.588496557850322,
+      "grad_norm": 0.7189040382281308,
+      "learning_rate": 9.932593533743786e-06,
+      "loss": 0.0602,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5889407061958695,
+      "grad_norm": 0.5837515830643734,
+      "learning_rate": 9.932275924337104e-06,
+      "loss": 0.0586,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5893848545414169,
+      "grad_norm": 0.8281667273984812,
+      "learning_rate": 9.931957573525566e-06,
+      "loss": 0.0648,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5898290028869643,
+      "grad_norm": 0.5286485536715172,
+      "learning_rate": 9.931638481357024e-06,
+      "loss": 0.0531,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5902731512325117,
+      "grad_norm": 0.6275875445640038,
+      "learning_rate": 9.931318647879445e-06,
+      "loss": 0.064,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5907172995780591,
+      "grad_norm": 0.8183874857273685,
+      "learning_rate": 9.930998073140905e-06,
+      "loss": 0.1023,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5911614479236065,
+      "grad_norm": 0.6557543608908961,
+      "learning_rate": 9.93067675718959e-06,
+      "loss": 0.0591,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5916055962691539,
+      "grad_norm": 0.7946472702361359,
+      "learning_rate": 9.930354700073803e-06,
+      "loss": 0.0783,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5920497446147013,
+      "grad_norm": 0.4755584849969953,
+      "learning_rate": 9.930031901841952e-06,
+      "loss": 0.065,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5924938929602487,
+      "grad_norm": 0.6292133058734393,
+      "learning_rate": 9.929708362542559e-06,
+      "loss": 0.065,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5929380413057961,
+      "grad_norm": 0.7194753693955848,
+      "learning_rate": 9.929384082224258e-06,
+      "loss": 0.0649,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5933821896513436,
+      "grad_norm": 0.8182929882444262,
+      "learning_rate": 9.929059060935795e-06,
+      "loss": 0.0735,
+      "step": 1336
+    },
+    {
+      "epoch": 0.593826337996891,
+      "grad_norm": 0.6452166028375979,
+      "learning_rate": 9.928733298726024e-06,
+      "loss": 0.0773,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5942704863424384,
+      "grad_norm": 0.6174785647804547,
+      "learning_rate": 9.928406795643913e-06,
+      "loss": 0.088,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5947146346879858,
+      "grad_norm": 0.9075257756258466,
+      "learning_rate": 9.928079551738542e-06,
+      "loss": 0.0966,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5951587830335332,
+      "grad_norm": 0.7930757535133779,
+      "learning_rate": 9.927751567059103e-06,
+      "loss": 0.0788,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5956029313790806,
+      "grad_norm": 0.7918355971015104,
+      "learning_rate": 9.927422841654894e-06,
+      "loss": 0.0732,
+      "step": 1341
+    },
+    {
+      "epoch": 0.596047079724628,
+      "grad_norm": 0.5344795042621896,
+      "learning_rate": 9.92709337557533e-06,
+      "loss": 0.0555,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5964912280701754,
+      "grad_norm": 0.7151020498309698,
+      "learning_rate": 9.926763168869935e-06,
+      "loss": 0.0699,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5969353764157228,
+      "grad_norm": 0.8228281453238034,
+      "learning_rate": 9.926432221588342e-06,
+      "loss": 0.0723,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5973795247612703,
+      "grad_norm": 0.6976750613205994,
+      "learning_rate": 9.926100533780304e-06,
+      "loss": 0.0877,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5978236731068177,
+      "grad_norm": 0.8796022908023586,
+      "learning_rate": 9.925768105495675e-06,
+      "loss": 0.0867,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5982678214523651,
+      "grad_norm": 0.6214669294617633,
+      "learning_rate": 9.925434936784426e-06,
+      "loss": 0.0724,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5987119697979125,
+      "grad_norm": 0.5493814047261378,
+      "learning_rate": 9.925101027696636e-06,
+      "loss": 0.0546,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5991561181434599,
+      "grad_norm": 0.6738873954295155,
+      "learning_rate": 9.924766378282499e-06,
+      "loss": 0.068,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5996002664890073,
+      "grad_norm": 0.662002450028893,
+      "learning_rate": 9.92443098859232e-06,
+      "loss": 0.0618,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6000444148345547,
+      "grad_norm": 0.5691844703394149,
+      "learning_rate": 9.92409485867651e-06,
+      "loss": 0.0675,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6004885631801021,
+      "grad_norm": 0.5107643154244839,
+      "learning_rate": 9.923757988585599e-06,
+      "loss": 0.0582,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6009327115256495,
+      "grad_norm": 0.4892180420795259,
+      "learning_rate": 9.923420378370221e-06,
+      "loss": 0.0707,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6013768598711969,
+      "grad_norm": 0.5128501684775312,
+      "learning_rate": 9.923082028081125e-06,
+      "loss": 0.0515,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6018210082167444,
+      "grad_norm": 0.6712578771133637,
+      "learning_rate": 9.922742937769172e-06,
+      "loss": 0.0668,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6022651565622918,
+      "grad_norm": 0.8489596826538364,
+      "learning_rate": 9.922403107485335e-06,
+      "loss": 0.0852,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6027093049078392,
+      "grad_norm": 0.7130581914563381,
+      "learning_rate": 9.922062537280692e-06,
+      "loss": 0.0877,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6031534532533867,
+      "grad_norm": 0.5999507766881641,
+      "learning_rate": 9.921721227206438e-06,
+      "loss": 0.0693,
+      "step": 1358
+    },
+    {
+      "epoch": 0.603597601598934,
+      "grad_norm": 0.5747773135899322,
+      "learning_rate": 9.92137917731388e-06,
+      "loss": 0.0556,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6040417499444815,
+      "grad_norm": 0.7771993716251702,
+      "learning_rate": 9.921036387654429e-06,
+      "loss": 0.0643,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6044858982900289,
+      "grad_norm": 1.307956020844314,
+      "learning_rate": 9.920692858279616e-06,
+      "loss": 0.0551,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6049300466355763,
+      "grad_norm": 0.6255379323591398,
+      "learning_rate": 9.92034858924108e-06,
+      "loss": 0.0636,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6053741949811237,
+      "grad_norm": 0.553962551501975,
+      "learning_rate": 9.92000358059057e-06,
+      "loss": 0.0814,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6058183433266711,
+      "grad_norm": 0.7454747633443178,
+      "learning_rate": 9.919657832379943e-06,
+      "loss": 0.0603,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6062624916722186,
+      "grad_norm": 0.9733260117481642,
+      "learning_rate": 9.919311344661174e-06,
+      "loss": 0.1153,
+      "step": 1365
+    },
+    {
+      "epoch": 0.606706640017766,
+      "grad_norm": 0.613902681985122,
+      "learning_rate": 9.918964117486346e-06,
+      "loss": 0.0578,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6071507883633134,
+      "grad_norm": 0.7376648326818326,
+      "learning_rate": 9.918616150907651e-06,
+      "loss": 0.066,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.6476167687641429,
+      "learning_rate": 9.918267444977398e-06,
+      "loss": 0.0467,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6080390850544082,
+      "grad_norm": 0.6720043569712078,
+      "learning_rate": 9.917917999747999e-06,
+      "loss": 0.0647,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6084832333999556,
+      "grad_norm": 0.6554718376719761,
+      "learning_rate": 9.917567815271986e-06,
+      "loss": 0.0652,
+      "step": 1370
+    },
+    {
+      "epoch": 0.608927381745503,
+      "grad_norm": 0.7668531453750762,
+      "learning_rate": 9.917216891601996e-06,
+      "loss": 0.07,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6093715300910504,
+      "grad_norm": 0.8088811507082688,
+      "learning_rate": 9.916865228790776e-06,
+      "loss": 0.0771,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6098156784365978,
+      "grad_norm": 0.687085496356637,
+      "learning_rate": 9.91651282689119e-06,
+      "loss": 0.0758,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6102598267821452,
+      "grad_norm": 0.7118263597209087,
+      "learning_rate": 9.916159685956208e-06,
+      "loss": 0.0704,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6107039751276927,
+      "grad_norm": 0.5722664877188374,
+      "learning_rate": 9.915805806038917e-06,
+      "loss": 0.0568,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6111481234732401,
+      "grad_norm": 0.5809145124340924,
+      "learning_rate": 9.915451187192507e-06,
+      "loss": 0.0508,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6115922718187875,
+      "grad_norm": 0.7936146709815162,
+      "learning_rate": 9.915095829470284e-06,
+      "loss": 0.0685,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6120364201643349,
+      "grad_norm": 0.6343789137617302,
+      "learning_rate": 9.914739732925665e-06,
+      "loss": 0.0727,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6124805685098823,
+      "grad_norm": 0.6049507318538985,
+      "learning_rate": 9.914382897612178e-06,
+      "loss": 0.0613,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6129247168554297,
+      "grad_norm": 0.7820325942346816,
+      "learning_rate": 9.91402532358346e-06,
+      "loss": 0.0631,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6133688652009771,
+      "grad_norm": 0.647686549326212,
+      "learning_rate": 9.913667010893261e-06,
+      "loss": 0.0593,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6138130135465245,
+      "grad_norm": 0.5699910207599692,
+      "learning_rate": 9.913307959595443e-06,
+      "loss": 0.0617,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6142571618920719,
+      "grad_norm": 0.8214987498163691,
+      "learning_rate": 9.912948169743977e-06,
+      "loss": 0.1062,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6147013102376193,
+      "grad_norm": 0.7190447874189771,
+      "learning_rate": 9.912587641392943e-06,
+      "loss": 0.0816,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6151454585831668,
+      "grad_norm": 0.4801110982287438,
+      "learning_rate": 9.912226374596536e-06,
+      "loss": 0.0531,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6155896069287142,
+      "grad_norm": 0.5777279849946639,
+      "learning_rate": 9.911864369409062e-06,
+      "loss": 0.0593,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6160337552742616,
+      "grad_norm": 0.6445627055234285,
+      "learning_rate": 9.911501625884934e-06,
+      "loss": 0.0562,
+      "step": 1387
+    },
+    {
+      "epoch": 0.616477903619809,
+      "grad_norm": 0.6402760221505619,
+      "learning_rate": 9.911138144078681e-06,
+      "loss": 0.0628,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6169220519653564,
+      "grad_norm": 0.5943075838810645,
+      "learning_rate": 9.910773924044937e-06,
+      "loss": 0.0694,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6173662003109038,
+      "grad_norm": 0.6887164265976037,
+      "learning_rate": 9.910408965838455e-06,
+      "loss": 0.0662,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6178103486564512,
+      "grad_norm": 0.7025269495789147,
+      "learning_rate": 9.91004326951409e-06,
+      "loss": 0.0596,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6182544970019986,
+      "grad_norm": 0.6968614560116964,
+      "learning_rate": 9.909676835126819e-06,
+      "loss": 0.08,
+      "step": 1392
+    },
+    {
+      "epoch": 0.618698645347546,
+      "grad_norm": 0.9599784608542535,
+      "learning_rate": 9.909309662731713e-06,
+      "loss": 0.0981,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6191427936930934,
+      "grad_norm": 0.5936127126118025,
+      "learning_rate": 9.908941752383974e-06,
+      "loss": 0.0502,
+      "step": 1394
+    },
+    {
+      "epoch": 0.619586942038641,
+      "grad_norm": 0.6281665293752687,
+      "learning_rate": 9.9085731041389e-06,
+      "loss": 0.0894,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6200310903841884,
+      "grad_norm": 0.7270920690197641,
+      "learning_rate": 9.908203718051907e-06,
+      "loss": 0.0772,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6204752387297358,
+      "grad_norm": 0.9689326115460888,
+      "learning_rate": 9.90783359417852e-06,
+      "loss": 0.0955,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6209193870752832,
+      "grad_norm": 0.5718494750591436,
+      "learning_rate": 9.907462732574373e-06,
+      "loss": 0.0586,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6213635354208306,
+      "grad_norm": 0.547133748335618,
+      "learning_rate": 9.907091133295214e-06,
+      "loss": 0.0546,
+      "step": 1399
+    },
+    {
+      "epoch": 0.621807683766378,
+      "grad_norm": 0.7492139302396027,
+      "learning_rate": 9.906718796396901e-06,
+      "loss": 0.0726,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6222518321119254,
+      "grad_norm": 1.1332966961373652,
+      "learning_rate": 9.906345721935402e-06,
+      "loss": 0.0837,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6226959804574728,
+      "grad_norm": 0.6506799230647055,
+      "learning_rate": 9.905971909966798e-06,
+      "loss": 0.0676,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6231401288030202,
+      "grad_norm": 0.7078271615105967,
+      "learning_rate": 9.905597360547276e-06,
+      "loss": 0.0763,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6235842771485676,
+      "grad_norm": 0.8061592052669874,
+      "learning_rate": 9.90522207373314e-06,
+      "loss": 0.0941,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6240284254941151,
+      "grad_norm": 0.7215188703499849,
+      "learning_rate": 9.904846049580804e-06,
+      "loss": 0.0913,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6244725738396625,
+      "grad_norm": 0.7648810068346831,
+      "learning_rate": 9.904469288146785e-06,
+      "loss": 0.103,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6249167221852099,
+      "grad_norm": 0.7848759400980109,
+      "learning_rate": 9.90409178948772e-06,
+      "loss": 0.0853,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6253608705307573,
+      "grad_norm": 0.5249169314876924,
+      "learning_rate": 9.903713553660352e-06,
+      "loss": 0.0539,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6258050188763047,
+      "grad_norm": 0.651696006803284,
+      "learning_rate": 9.90333458072154e-06,
+      "loss": 0.0725,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6262491672218521,
+      "grad_norm": 0.5581372640238701,
+      "learning_rate": 9.902954870728246e-06,
+      "loss": 0.0516,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6266933155673995,
+      "grad_norm": 0.46394395658623727,
+      "learning_rate": 9.902574423737547e-06,
+      "loss": 0.0543,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6271374639129469,
+      "grad_norm": 0.6101555748112776,
+      "learning_rate": 9.902193239806634e-06,
+      "loss": 0.0688,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6275816122584943,
+      "grad_norm": 0.6266576234326965,
+      "learning_rate": 9.901811318992802e-06,
+      "loss": 0.0619,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6280257606040418,
+      "grad_norm": 0.9093069033662159,
+      "learning_rate": 9.901428661353462e-06,
+      "loss": 0.1009,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6284699089495892,
+      "grad_norm": 0.796133446241698,
+      "learning_rate": 9.901045266946134e-06,
+      "loss": 0.0919,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6289140572951366,
+      "grad_norm": 0.75094830010195,
+      "learning_rate": 9.900661135828448e-06,
+      "loss": 0.0594,
+      "step": 1416
+    },
+    {
+      "epoch": 0.629358205640684,
+      "grad_norm": 0.8379496784608408,
+      "learning_rate": 9.900276268058147e-06,
+      "loss": 0.0659,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6298023539862314,
+      "grad_norm": 0.7794466397530133,
+      "learning_rate": 9.899890663693078e-06,
+      "loss": 0.086,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6302465023317788,
+      "grad_norm": 1.781154839834256,
+      "learning_rate": 9.899504322791212e-06,
+      "loss": 0.1089,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6306906506773262,
+      "grad_norm": 0.582797339406938,
+      "learning_rate": 9.899117245410615e-06,
+      "loss": 0.0449,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6311347990228736,
+      "grad_norm": 0.7463622614388931,
+      "learning_rate": 9.898729431609477e-06,
+      "loss": 0.0542,
+      "step": 1421
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.7697052470584549,
+      "learning_rate": 9.89834088144609e-06,
+      "loss": 0.0635,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6320230957139684,
+      "grad_norm": 0.900347396335065,
+      "learning_rate": 9.897951594978858e-06,
+      "loss": 0.0778,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6324672440595159,
+      "grad_norm": 0.6468712468526486,
+      "learning_rate": 9.897561572266301e-06,
+      "loss": 0.066,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6329113924050633,
+      "grad_norm": 0.4986782583243913,
+      "learning_rate": 9.897170813367045e-06,
+      "loss": 0.0662,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6333555407506107,
+      "grad_norm": 0.6014710971137512,
+      "learning_rate": 9.896779318339826e-06,
+      "loss": 0.0657,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6337996890961581,
+      "grad_norm": 0.8579597362562785,
+      "learning_rate": 9.896387087243496e-06,
+      "loss": 0.08,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6342438374417055,
+      "grad_norm": 0.9386752220131082,
+      "learning_rate": 9.89599412013701e-06,
+      "loss": 0.1233,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6346879857872529,
+      "grad_norm": 0.7426460359725069,
+      "learning_rate": 9.89560041707944e-06,
+      "loss": 0.0861,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6351321341328003,
+      "grad_norm": 0.5143125228492677,
+      "learning_rate": 9.895205978129966e-06,
+      "loss": 0.0446,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6355762824783477,
+      "grad_norm": 1.049177103786211,
+      "learning_rate": 9.894810803347878e-06,
+      "loss": 0.0912,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6360204308238951,
+      "grad_norm": 0.5972518436510755,
+      "learning_rate": 9.894414892792579e-06,
+      "loss": 0.0657,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6364645791694425,
+      "grad_norm": 0.4154671445638412,
+      "learning_rate": 9.894018246523577e-06,
+      "loss": 0.0532,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6369087275149901,
+      "grad_norm": 0.65372867549855,
+      "learning_rate": 9.893620864600501e-06,
+      "loss": 0.087,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6373528758605375,
+      "grad_norm": 0.7777909981283909,
+      "learning_rate": 9.89322274708308e-06,
+      "loss": 0.0795,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6377970242060849,
+      "grad_norm": 0.6281013351353801,
+      "learning_rate": 9.892823894031159e-06,
+      "loss": 0.0525,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6382411725516323,
+      "grad_norm": 0.5274870510804406,
+      "learning_rate": 9.89242430550469e-06,
+      "loss": 0.0701,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6386853208971797,
+      "grad_norm": 0.9840882810513909,
+      "learning_rate": 9.892023981563744e-06,
+      "loss": 0.0888,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6391294692427271,
+      "grad_norm": 0.7186121306665683,
+      "learning_rate": 9.89162292226849e-06,
+      "loss": 0.0666,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6395736175882745,
+      "grad_norm": 0.7619824705257715,
+      "learning_rate": 9.891221127679216e-06,
+      "loss": 0.0654,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6400177659338219,
+      "grad_norm": 0.6418832068583645,
+      "learning_rate": 9.89081859785632e-06,
+      "loss": 0.0706,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6404619142793693,
+      "grad_norm": 0.6934692957378497,
+      "learning_rate": 9.890415332860308e-06,
+      "loss": 0.0684,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6409060626249167,
+      "grad_norm": 0.6769661850243969,
+      "learning_rate": 9.8900113327518e-06,
+      "loss": 0.0684,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6413502109704642,
+      "grad_norm": 0.6158605594742834,
+      "learning_rate": 9.88960659759152e-06,
+      "loss": 0.0652,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6417943593160116,
+      "grad_norm": 0.5847451238116226,
+      "learning_rate": 9.88920112744031e-06,
+      "loss": 0.062,
+      "step": 1445
+    },
+    {
+      "epoch": 0.642238507661559,
+      "grad_norm": 0.8117410133012406,
+      "learning_rate": 9.888794922359116e-06,
+      "loss": 0.0611,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6426826560071064,
+      "grad_norm": 0.7705515566871968,
+      "learning_rate": 9.888387982408998e-06,
+      "loss": 0.0752,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6431268043526538,
+      "grad_norm": 0.777589892103545,
+      "learning_rate": 9.887980307651128e-06,
+      "loss": 0.0764,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6435709526982012,
+      "grad_norm": 0.8491119564862617,
+      "learning_rate": 9.887571898146787e-06,
+      "loss": 0.081,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6440151010437486,
+      "grad_norm": 0.7918875216411717,
+      "learning_rate": 9.887162753957362e-06,
+      "loss": 0.0626,
+      "step": 1450
+    },
+    {
+      "epoch": 0.644459249389296,
+      "grad_norm": 0.5756994501297381,
+      "learning_rate": 9.886752875144358e-06,
+      "loss": 0.0635,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6449033977348434,
+      "grad_norm": 0.6492149852575843,
+      "learning_rate": 9.886342261769387e-06,
+      "loss": 0.0632,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6453475460803908,
+      "grad_norm": 0.7503617012041023,
+      "learning_rate": 9.885930913894166e-06,
+      "loss": 0.0627,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6457916944259383,
+      "grad_norm": 0.6405303313329822,
+      "learning_rate": 9.885518831580533e-06,
+      "loss": 0.0782,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6462358427714857,
+      "grad_norm": 0.6250298341424497,
+      "learning_rate": 9.88510601489043e-06,
+      "loss": 0.0807,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6466799911170331,
+      "grad_norm": 0.7939161691732389,
+      "learning_rate": 9.88469246388591e-06,
+      "loss": 0.069,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6471241394625805,
+      "grad_norm": 0.7492359887972416,
+      "learning_rate": 9.884278178629134e-06,
+      "loss": 0.0854,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6475682878081279,
+      "grad_norm": 0.6085287559130933,
+      "learning_rate": 9.883863159182379e-06,
+      "loss": 0.0608,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6480124361536753,
+      "grad_norm": 0.6361755513609607,
+      "learning_rate": 9.883447405608032e-06,
+      "loss": 0.0433,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6484565844992227,
+      "grad_norm": 0.6431691982633025,
+      "learning_rate": 9.88303091796858e-06,
+      "loss": 0.0419,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6489007328447701,
+      "grad_norm": 0.8190085662166918,
+      "learning_rate": 9.882613696326634e-06,
+      "loss": 0.0767,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6493448811903175,
+      "grad_norm": 0.755548239297633,
+      "learning_rate": 9.882195740744911e-06,
+      "loss": 0.0923,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6497890295358649,
+      "grad_norm": 0.5719039887354458,
+      "learning_rate": 9.881777051286232e-06,
+      "loss": 0.0723,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6502331778814124,
+      "grad_norm": 0.7825393714474173,
+      "learning_rate": 9.881357628013535e-06,
+      "loss": 0.0835,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6506773262269598,
+      "grad_norm": 0.6610007380508074,
+      "learning_rate": 9.880937470989868e-06,
+      "loss": 0.0656,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6511214745725072,
+      "grad_norm": 0.6218987787310035,
+      "learning_rate": 9.880516580278386e-06,
+      "loss": 0.0594,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6515656229180546,
+      "grad_norm": 0.6871813560151103,
+      "learning_rate": 9.880094955942357e-06,
+      "loss": 0.0539,
+      "step": 1467
+    },
+    {
+      "epoch": 0.652009771263602,
+      "grad_norm": 0.6311698562321532,
+      "learning_rate": 9.879672598045156e-06,
+      "loss": 0.0806,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6524539196091494,
+      "grad_norm": 0.6693355850981739,
+      "learning_rate": 9.879249506650275e-06,
+      "loss": 0.0856,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6528980679546968,
+      "grad_norm": 0.7127562713682332,
+      "learning_rate": 9.878825681821306e-06,
+      "loss": 0.0685,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6533422163002442,
+      "grad_norm": 0.9425811868582009,
+      "learning_rate": 9.878401123621963e-06,
+      "loss": 0.0823,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6537863646457917,
+      "grad_norm": 0.6086374723526442,
+      "learning_rate": 9.87797583211606e-06,
+      "loss": 0.0614,
+      "step": 1472
+    },
+    {
+      "epoch": 0.654230512991339,
+      "grad_norm": 0.7763257358459474,
+      "learning_rate": 9.877549807367528e-06,
+      "loss": 0.0857,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6546746613368866,
+      "grad_norm": 0.539995990500211,
+      "learning_rate": 9.877123049440405e-06,
+      "loss": 0.0531,
+      "step": 1474
+    },
+    {
+      "epoch": 0.655118809682434,
+      "grad_norm": 0.5915202116072419,
+      "learning_rate": 9.876695558398838e-06,
+      "loss": 0.0663,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6555629580279814,
+      "grad_norm": 0.6814775856965638,
+      "learning_rate": 9.876267334307091e-06,
+      "loss": 0.0536,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6560071063735288,
+      "grad_norm": 0.6565534779979346,
+      "learning_rate": 9.875838377229528e-06,
+      "loss": 0.0854,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6564512547190762,
+      "grad_norm": 0.6074291178048928,
+      "learning_rate": 9.875408687230633e-06,
+      "loss": 0.0643,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6568954030646236,
+      "grad_norm": 0.6683735797478915,
+      "learning_rate": 9.874978264374991e-06,
+      "loss": 0.0657,
+      "step": 1479
+    },
+    {
+      "epoch": 0.657339551410171,
+      "grad_norm": 0.6802029779866509,
+      "learning_rate": 9.874547108727306e-06,
+      "loss": 0.0571,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6577836997557184,
+      "grad_norm": 0.7709030410876545,
+      "learning_rate": 9.874115220352386e-06,
+      "loss": 0.0515,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6582278481012658,
+      "grad_norm": 0.6029748648689206,
+      "learning_rate": 9.873682599315152e-06,
+      "loss": 0.054,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6586719964468133,
+      "grad_norm": 0.7218151069926805,
+      "learning_rate": 9.873249245680634e-06,
+      "loss": 0.0842,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6591161447923607,
+      "grad_norm": 0.7118725638175186,
+      "learning_rate": 9.872815159513972e-06,
+      "loss": 0.0783,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6595602931379081,
+      "grad_norm": 0.5278035850844903,
+      "learning_rate": 9.872380340880416e-06,
+      "loss": 0.0504,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6600044414834555,
+      "grad_norm": 0.7090158968533085,
+      "learning_rate": 9.87194478984533e-06,
+      "loss": 0.0518,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6604485898290029,
+      "grad_norm": 0.7140940579751807,
+      "learning_rate": 9.87150850647418e-06,
+      "loss": 0.0644,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6608927381745503,
+      "grad_norm": 0.5233702996222387,
+      "learning_rate": 9.87107149083255e-06,
+      "loss": 0.0484,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6613368865200977,
+      "grad_norm": 0.5255868844598361,
+      "learning_rate": 9.870633742986129e-06,
+      "loss": 0.0486,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6617810348656451,
+      "grad_norm": 0.7456554727487221,
+      "learning_rate": 9.870195263000719e-06,
+      "loss": 0.0926,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6622251832111925,
+      "grad_norm": 0.5094239000908518,
+      "learning_rate": 9.869756050942231e-06,
+      "loss": 0.0609,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6626693315567399,
+      "grad_norm": 0.5934076694430587,
+      "learning_rate": 9.869316106876687e-06,
+      "loss": 0.0815,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6631134799022874,
+      "grad_norm": 0.5800872547088418,
+      "learning_rate": 9.868875430870217e-06,
+      "loss": 0.0562,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6635576282478348,
+      "grad_norm": 0.6466806398918721,
+      "learning_rate": 9.86843402298906e-06,
+      "loss": 0.0576,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6640017765933822,
+      "grad_norm": 0.49149929770034256,
+      "learning_rate": 9.86799188329957e-06,
+      "loss": 0.0552,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6644459249389296,
+      "grad_norm": 0.6869307446643815,
+      "learning_rate": 9.867549011868208e-06,
+      "loss": 0.0689,
+      "step": 1496
+    },
+    {
+      "epoch": 0.664890073284477,
+      "grad_norm": 1.0182179632443422,
+      "learning_rate": 9.867105408761544e-06,
+      "loss": 0.0658,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6653342216300244,
+      "grad_norm": 0.5094314413955369,
+      "learning_rate": 9.866661074046258e-06,
+      "loss": 0.0509,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6657783699755718,
+      "grad_norm": 0.6982818166570782,
+      "learning_rate": 9.866216007789145e-06,
+      "loss": 0.0728,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6662225183211192,
+      "grad_norm": 0.7524936010719828,
+      "learning_rate": 9.8657702100571e-06,
+      "loss": 0.0911,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.7557138142113028,
+      "learning_rate": 9.86532368091714e-06,
+      "loss": 0.059,
+      "step": 1501
+    },
+    {
+      "epoch": 0.667110815012214,
+      "grad_norm": 0.6888932895497213,
+      "learning_rate": 9.864876420436383e-06,
+      "loss": 0.0614,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6675549633577615,
+      "grad_norm": 0.5506269374818886,
+      "learning_rate": 9.86442842868206e-06,
+      "loss": 0.0599,
+      "step": 1503
+    },
+    {
+      "epoch": 0.667999111703309,
+      "grad_norm": 0.8018485119174299,
+      "learning_rate": 9.86397970572151e-06,
+      "loss": 0.0875,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6684432600488563,
+      "grad_norm": 0.5631462958993229,
+      "learning_rate": 9.863530251622189e-06,
+      "loss": 0.0691,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6688874083944037,
+      "grad_norm": 0.8176186875229494,
+      "learning_rate": 9.863080066451653e-06,
+      "loss": 0.072,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6693315567399512,
+      "grad_norm": 0.7128371312290194,
+      "learning_rate": 9.862629150277574e-06,
+      "loss": 0.0773,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6697757050854986,
+      "grad_norm": 0.759379781839776,
+      "learning_rate": 9.86217750316773e-06,
+      "loss": 0.0729,
+      "step": 1508
+    },
+    {
+      "epoch": 0.670219853431046,
+      "grad_norm": 0.6634457926439065,
+      "learning_rate": 9.861725125190017e-06,
+      "loss": 0.0595,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6706640017765934,
+      "grad_norm": 0.7426433953184679,
+      "learning_rate": 9.861272016412429e-06,
+      "loss": 0.065,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6711081501221408,
+      "grad_norm": 0.6600032637907064,
+      "learning_rate": 9.86081817690308e-06,
+      "loss": 0.0625,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6715522984676882,
+      "grad_norm": 0.703748253196849,
+      "learning_rate": 9.860363606730185e-06,
+      "loss": 0.073,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6719964468132357,
+      "grad_norm": 0.8768454660219879,
+      "learning_rate": 9.85990830596208e-06,
+      "loss": 0.093,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6724405951587831,
+      "grad_norm": 0.5810307286903331,
+      "learning_rate": 9.859452274667199e-06,
+      "loss": 0.0664,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6728847435043305,
+      "grad_norm": 0.6467795840989637,
+      "learning_rate": 9.858995512914096e-06,
+      "loss": 0.0736,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6733288918498779,
+      "grad_norm": 0.8604239627870944,
+      "learning_rate": 9.858538020771424e-06,
+      "loss": 0.0819,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6737730401954253,
+      "grad_norm": 0.5072234228135453,
+      "learning_rate": 9.858079798307959e-06,
+      "loss": 0.0764,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6742171885409727,
+      "grad_norm": 0.8843203312811015,
+      "learning_rate": 9.857620845592573e-06,
+      "loss": 0.0941,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6746613368865201,
+      "grad_norm": 0.7766118997696602,
+      "learning_rate": 9.85716116269426e-06,
+      "loss": 0.0697,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6751054852320675,
+      "grad_norm": 0.5371610065691803,
+      "learning_rate": 9.856700749682114e-06,
+      "loss": 0.075,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6755496335776149,
+      "grad_norm": 0.5048461380926635,
+      "learning_rate": 9.856239606625345e-06,
+      "loss": 0.0584,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6759937819231623,
+      "grad_norm": 0.819913205588501,
+      "learning_rate": 9.855777733593269e-06,
+      "loss": 0.0748,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6764379302687098,
+      "grad_norm": 0.6492894969300076,
+      "learning_rate": 9.855315130655315e-06,
+      "loss": 0.0573,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6768820786142572,
+      "grad_norm": 1.8883752330861565,
+      "learning_rate": 9.854851797881018e-06,
+      "loss": 0.0586,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6773262269598046,
+      "grad_norm": 0.6674563588470771,
+      "learning_rate": 9.854387735340028e-06,
+      "loss": 0.0618,
+      "step": 1525
+    },
+    {
+      "epoch": 0.677770375305352,
+      "grad_norm": 0.5745105185416173,
+      "learning_rate": 9.853922943102099e-06,
+      "loss": 0.0486,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6782145236508994,
+      "grad_norm": 0.8823764363833241,
+      "learning_rate": 9.853457421237098e-06,
+      "loss": 0.0655,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6786586719964468,
+      "grad_norm": 0.7502418552024257,
+      "learning_rate": 9.852991169815002e-06,
+      "loss": 0.0697,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6791028203419942,
+      "grad_norm": 0.9371750933741902,
+      "learning_rate": 9.852524188905894e-06,
+      "loss": 0.0637,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6795469686875416,
+      "grad_norm": 0.4760980667666431,
+      "learning_rate": 9.85205647857997e-06,
+      "loss": 0.0507,
+      "step": 1530
+    },
+    {
+      "epoch": 0.679991117033089,
+      "grad_norm": 0.5957775022770642,
+      "learning_rate": 9.851588038907536e-06,
+      "loss": 0.0725,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6804352653786364,
+      "grad_norm": 0.8583868263839582,
+      "learning_rate": 9.851118869959006e-06,
+      "loss": 0.0926,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6808794137241839,
+      "grad_norm": 0.5065773305637178,
+      "learning_rate": 9.850648971804903e-06,
+      "loss": 0.0519,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6813235620697313,
+      "grad_norm": 0.6463242249375162,
+      "learning_rate": 9.850178344515861e-06,
+      "loss": 0.0495,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6817677104152787,
+      "grad_norm": 0.713200223511551,
+      "learning_rate": 9.849706988162626e-06,
+      "loss": 0.0667,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6822118587608261,
+      "grad_norm": 0.8106886906091921,
+      "learning_rate": 9.849234902816047e-06,
+      "loss": 0.078,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6826560071063735,
+      "grad_norm": 0.618211559505216,
+      "learning_rate": 9.848762088547089e-06,
+      "loss": 0.0499,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6831001554519209,
+      "grad_norm": 0.4683137101850568,
+      "learning_rate": 9.848288545426821e-06,
+      "loss": 0.0485,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6835443037974683,
+      "grad_norm": 0.6498884933818719,
+      "learning_rate": 9.847814273526428e-06,
+      "loss": 0.0717,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6839884521430157,
+      "grad_norm": 0.6112948857459086,
+      "learning_rate": 9.8473392729172e-06,
+      "loss": 0.0556,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6844326004885631,
+      "grad_norm": 0.6350627430979878,
+      "learning_rate": 9.846863543670536e-06,
+      "loss": 0.067,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6848767488341105,
+      "grad_norm": 1.1432261374175676,
+      "learning_rate": 9.846387085857949e-06,
+      "loss": 0.1195,
+      "step": 1542
+    },
+    {
+      "epoch": 0.685320897179658,
+      "grad_norm": 0.5445218211710267,
+      "learning_rate": 9.845909899551056e-06,
+      "loss": 0.0633,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6857650455252055,
+      "grad_norm": 0.7666433728201425,
+      "learning_rate": 9.845431984821588e-06,
+      "loss": 0.0652,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6862091938707529,
+      "grad_norm": 0.6445193247682808,
+      "learning_rate": 9.844953341741383e-06,
+      "loss": 0.044,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6866533422163003,
+      "grad_norm": 0.6165834169963341,
+      "learning_rate": 9.844473970382391e-06,
+      "loss": 0.07,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6870974905618477,
+      "grad_norm": 0.785019896367438,
+      "learning_rate": 9.843993870816665e-06,
+      "loss": 0.0692,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6875416389073951,
+      "grad_norm": 0.8097989024467772,
+      "learning_rate": 9.843513043116378e-06,
+      "loss": 0.0714,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6879857872529425,
+      "grad_norm": 0.7549105996829288,
+      "learning_rate": 9.843031487353803e-06,
+      "loss": 0.0642,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6884299355984899,
+      "grad_norm": 0.4886439517505401,
+      "learning_rate": 9.842549203601327e-06,
+      "loss": 0.0562,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6888740839440373,
+      "grad_norm": 0.7561494998266006,
+      "learning_rate": 9.842066191931442e-06,
+      "loss": 0.0643,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6893182322895848,
+      "grad_norm": 0.854720220061704,
+      "learning_rate": 9.84158245241676e-06,
+      "loss": 0.0522,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6897623806351322,
+      "grad_norm": 0.5346332465205076,
+      "learning_rate": 9.84109798512999e-06,
+      "loss": 0.0518,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6902065289806796,
+      "grad_norm": 0.508165162429757,
+      "learning_rate": 9.840612790143958e-06,
+      "loss": 0.0538,
+      "step": 1554
+    },
+    {
+      "epoch": 0.690650677326227,
+      "grad_norm": 0.4808558159762227,
+      "learning_rate": 9.840126867531594e-06,
+      "loss": 0.0604,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6910948256717744,
+      "grad_norm": 0.6036124527734282,
+      "learning_rate": 9.839640217365941e-06,
+      "loss": 0.0636,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6915389740173218,
+      "grad_norm": 0.5365369852245644,
+      "learning_rate": 9.839152839720157e-06,
+      "loss": 0.0571,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6919831223628692,
+      "grad_norm": 0.9675442829718458,
+      "learning_rate": 9.838664734667496e-06,
+      "loss": 0.0768,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6924272707084166,
+      "grad_norm": 0.6682601437620973,
+      "learning_rate": 9.83817590228133e-06,
+      "loss": 0.0682,
+      "step": 1559
+    },
+    {
+      "epoch": 0.692871419053964,
+      "grad_norm": 0.915552570084289,
+      "learning_rate": 9.83768634263514e-06,
+      "loss": 0.0689,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6933155673995114,
+      "grad_norm": 0.7294790821869619,
+      "learning_rate": 9.837196055802514e-06,
+      "loss": 0.0595,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6937597157450589,
+      "grad_norm": 0.6877843622659181,
+      "learning_rate": 9.836705041857153e-06,
+      "loss": 0.0894,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6942038640906063,
+      "grad_norm": 0.5609065701691318,
+      "learning_rate": 9.836213300872862e-06,
+      "loss": 0.0523,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6946480124361537,
+      "grad_norm": 0.6750341373878816,
+      "learning_rate": 9.83572083292356e-06,
+      "loss": 0.0723,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6950921607817011,
+      "grad_norm": 0.8179982672102986,
+      "learning_rate": 9.835227638083271e-06,
+      "loss": 0.0741,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6955363091272485,
+      "grad_norm": 0.514148745778486,
+      "learning_rate": 9.834733716426133e-06,
+      "loss": 0.048,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6959804574727959,
+      "grad_norm": 0.5288948126890644,
+      "learning_rate": 9.834239068026388e-06,
+      "loss": 0.0665,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6964246058183433,
+      "grad_norm": 0.5107938326869891,
+      "learning_rate": 9.833743692958392e-06,
+      "loss": 0.0482,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6968687541638907,
+      "grad_norm": 1.053299477751095,
+      "learning_rate": 9.83324759129661e-06,
+      "loss": 0.0726,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6973129025094381,
+      "grad_norm": 0.6266804348595715,
+      "learning_rate": 9.832750763115611e-06,
+      "loss": 0.0738,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6977570508549855,
+      "grad_norm": 0.7607794431765756,
+      "learning_rate": 9.83225320849008e-06,
+      "loss": 0.0739,
+      "step": 1571
+    },
+    {
+      "epoch": 0.698201199200533,
+      "grad_norm": 0.7778393457616919,
+      "learning_rate": 9.831754927494803e-06,
+      "loss": 0.0948,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6986453475460804,
+      "grad_norm": 0.7195786654461274,
+      "learning_rate": 9.831255920204685e-06,
+      "loss": 0.0646,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6990894958916278,
+      "grad_norm": 0.6366879939595091,
+      "learning_rate": 9.830756186694734e-06,
+      "loss": 0.0808,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6995336442371752,
+      "grad_norm": 0.478436966515515,
+      "learning_rate": 9.830255727040066e-06,
+      "loss": 0.059,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6999777925827226,
+      "grad_norm": 0.5378537626686002,
+      "learning_rate": 9.829754541315912e-06,
+      "loss": 0.0624,
+      "step": 1576
+    },
+    {
+      "epoch": 0.70042194092827,
+      "grad_norm": 0.5458661735869976,
+      "learning_rate": 9.829252629597607e-06,
+      "loss": 0.056,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7008660892738174,
+      "grad_norm": 0.7598746622748797,
+      "learning_rate": 9.828749991960598e-06,
+      "loss": 0.0742,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7013102376193648,
+      "grad_norm": 0.7874301906671837,
+      "learning_rate": 9.828246628480438e-06,
+      "loss": 0.1126,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7017543859649122,
+      "grad_norm": 0.5696760012747649,
+      "learning_rate": 9.827742539232791e-06,
+      "loss": 0.0622,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7021985343104596,
+      "grad_norm": 0.49299605250258527,
+      "learning_rate": 9.827237724293434e-06,
+      "loss": 0.0542,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7026426826560072,
+      "grad_norm": 0.7950299040696133,
+      "learning_rate": 9.826732183738246e-06,
+      "loss": 0.0736,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7030868310015546,
+      "grad_norm": 0.6701858553269311,
+      "learning_rate": 9.826225917643217e-06,
+      "loss": 0.0769,
+      "step": 1583
+    },
+    {
+      "epoch": 0.703530979347102,
+      "grad_norm": 0.7198204203508045,
+      "learning_rate": 9.825718926084451e-06,
+      "loss": 0.0607,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7039751276926494,
+      "grad_norm": 0.7221114160407107,
+      "learning_rate": 9.825211209138154e-06,
+      "loss": 0.0662,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7044192760381968,
+      "grad_norm": 0.5718930945068039,
+      "learning_rate": 9.82470276688065e-06,
+      "loss": 0.063,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7048634243837442,
+      "grad_norm": 0.7156464008159005,
+      "learning_rate": 9.824193599388358e-06,
+      "loss": 0.061,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7053075727292916,
+      "grad_norm": 0.4730100411783297,
+      "learning_rate": 9.823683706737824e-06,
+      "loss": 0.0538,
+      "step": 1588
+    },
+    {
+      "epoch": 0.705751721074839,
+      "grad_norm": 0.8151288747975431,
+      "learning_rate": 9.823173089005686e-06,
+      "loss": 0.0586,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7061958694203864,
+      "grad_norm": 0.6500001880192491,
+      "learning_rate": 9.822661746268702e-06,
+      "loss": 0.0558,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7066400177659338,
+      "grad_norm": 0.47387928526278555,
+      "learning_rate": 9.822149678603733e-06,
+      "loss": 0.0537,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7070841661114813,
+      "grad_norm": 0.5533123786666888,
+      "learning_rate": 9.821636886087755e-06,
+      "loss": 0.049,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7075283144570287,
+      "grad_norm": 0.5418212216604064,
+      "learning_rate": 9.82112336879785e-06,
+      "loss": 0.0443,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7079724628025761,
+      "grad_norm": 0.595384134286072,
+      "learning_rate": 9.820609126811202e-06,
+      "loss": 0.0701,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7084166111481235,
+      "grad_norm": 0.5686391716299167,
+      "learning_rate": 9.820094160205118e-06,
+      "loss": 0.0612,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.6501450869897375,
+      "learning_rate": 9.819578469057e-06,
+      "loss": 0.0571,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7093049078392183,
+      "grad_norm": 0.8251798873678969,
+      "learning_rate": 9.819062053444369e-06,
+      "loss": 0.0903,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7097490561847657,
+      "grad_norm": 0.659869482663948,
+      "learning_rate": 9.81854491344485e-06,
+      "loss": 0.077,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7101932045303131,
+      "grad_norm": 0.5233735880329885,
+      "learning_rate": 9.818027049136177e-06,
+      "loss": 0.0756,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7106373528758605,
+      "grad_norm": 0.5435283663423869,
+      "learning_rate": 9.817508460596195e-06,
+      "loss": 0.0517,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7110815012214079,
+      "grad_norm": 0.8462175785185146,
+      "learning_rate": 9.816989147902855e-06,
+      "loss": 0.0776,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7115256495669554,
+      "grad_norm": 0.5250146868792045,
+      "learning_rate": 9.816469111134221e-06,
+      "loss": 0.0636,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7119697979125028,
+      "grad_norm": 0.4561566850818785,
+      "learning_rate": 9.81594835036846e-06,
+      "loss": 0.0549,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7124139462580502,
+      "grad_norm": 0.639031123731633,
+      "learning_rate": 9.815426865683858e-06,
+      "loss": 0.0739,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7128580946035976,
+      "grad_norm": 0.4546814856440393,
+      "learning_rate": 9.814904657158793e-06,
+      "loss": 0.0536,
+      "step": 1605
+    },
+    {
+      "epoch": 0.713302242949145,
+      "grad_norm": 0.49725993065397267,
+      "learning_rate": 9.81438172487177e-06,
+      "loss": 0.0538,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7137463912946924,
+      "grad_norm": 0.7681224223760409,
+      "learning_rate": 9.813858068901391e-06,
+      "loss": 0.0738,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7141905396402398,
+      "grad_norm": 0.5475444111433433,
+      "learning_rate": 9.813333689326371e-06,
+      "loss": 0.0532,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7146346879857872,
+      "grad_norm": 0.5001622305576355,
+      "learning_rate": 9.812808586225533e-06,
+      "loss": 0.0504,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7150788363313346,
+      "grad_norm": 0.5966203326606518,
+      "learning_rate": 9.812282759677811e-06,
+      "loss": 0.0632,
+      "step": 1610
+    },
+    {
+      "epoch": 0.715522984676882,
+      "grad_norm": 0.5024269719688089,
+      "learning_rate": 9.811756209762242e-06,
+      "loss": 0.054,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7159671330224295,
+      "grad_norm": 0.5953895951996429,
+      "learning_rate": 9.811228936557977e-06,
+      "loss": 0.0687,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7164112813679769,
+      "grad_norm": 0.9253768611459666,
+      "learning_rate": 9.810700940144275e-06,
+      "loss": 0.0936,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7168554297135243,
+      "grad_norm": 0.40706001374840456,
+      "learning_rate": 9.810172220600503e-06,
+      "loss": 0.0501,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7172995780590717,
+      "grad_norm": 0.6702001680907567,
+      "learning_rate": 9.809642778006135e-06,
+      "loss": 0.0721,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7177437264046191,
+      "grad_norm": 0.5621892036366335,
+      "learning_rate": 9.809112612440757e-06,
+      "loss": 0.0624,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7181878747501665,
+      "grad_norm": 0.6550556626236272,
+      "learning_rate": 9.808581723984059e-06,
+      "loss": 0.064,
+      "step": 1617
+    },
+    {
+      "epoch": 0.718632023095714,
+      "grad_norm": 1.171906822226206,
+      "learning_rate": 9.808050112715845e-06,
+      "loss": 0.1172,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7190761714412613,
+      "grad_norm": 0.4455710256139167,
+      "learning_rate": 9.807517778716025e-06,
+      "loss": 0.045,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7195203197868087,
+      "grad_norm": 0.6254516596662227,
+      "learning_rate": 9.806984722064616e-06,
+      "loss": 0.0705,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7199644681323563,
+      "grad_norm": 0.5571332804687471,
+      "learning_rate": 9.806450942841747e-06,
+      "loss": 0.0502,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7204086164779037,
+      "grad_norm": 1.0191625308451748,
+      "learning_rate": 9.805916441127657e-06,
+      "loss": 0.0589,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7208527648234511,
+      "grad_norm": 0.7859502245676739,
+      "learning_rate": 9.805381217002684e-06,
+      "loss": 0.0431,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7212969131689985,
+      "grad_norm": 0.8065166441211357,
+      "learning_rate": 9.804845270547288e-06,
+      "loss": 0.0728,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7217410615145459,
+      "grad_norm": 0.7073921214139905,
+      "learning_rate": 9.804308601842026e-06,
+      "loss": 0.0753,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7221852098600933,
+      "grad_norm": 0.7787625206064858,
+      "learning_rate": 9.80377121096757e-06,
+      "loss": 0.0854,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7226293582056407,
+      "grad_norm": 1.148628838473968,
+      "learning_rate": 9.8032330980047e-06,
+      "loss": 0.092,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7230735065511881,
+      "grad_norm": 0.5694973855182914,
+      "learning_rate": 9.802694263034302e-06,
+      "loss": 0.0661,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7235176548967355,
+      "grad_norm": 0.6928354319078996,
+      "learning_rate": 9.802154706137372e-06,
+      "loss": 0.0588,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7239618032422829,
+      "grad_norm": 0.5747608874251084,
+      "learning_rate": 9.801614427395018e-06,
+      "loss": 0.073,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7244059515878304,
+      "grad_norm": 0.8245000270563847,
+      "learning_rate": 9.801073426888447e-06,
+      "loss": 0.0602,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7248500999333778,
+      "grad_norm": 0.6063953480963461,
+      "learning_rate": 9.800531704698986e-06,
+      "loss": 0.0658,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7252942482789252,
+      "grad_norm": 0.5240737295642546,
+      "learning_rate": 9.799989260908063e-06,
+      "loss": 0.0564,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7257383966244726,
+      "grad_norm": 0.7016224236011144,
+      "learning_rate": 9.799446095597216e-06,
+      "loss": 0.0646,
+      "step": 1634
+    },
+    {
+      "epoch": 0.72618254497002,
+      "grad_norm": 0.6626808866197511,
+      "learning_rate": 9.798902208848093e-06,
+      "loss": 0.0818,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7266266933155674,
+      "grad_norm": 0.6141836548174839,
+      "learning_rate": 9.79835760074245e-06,
+      "loss": 0.0508,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7270708416611148,
+      "grad_norm": 0.6868383771343689,
+      "learning_rate": 9.797812271362149e-06,
+      "loss": 0.0552,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7275149900066622,
+      "grad_norm": 0.4532209654763099,
+      "learning_rate": 9.79726622078916e-06,
+      "loss": 0.0451,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7279591383522096,
+      "grad_norm": 0.6808218136734143,
+      "learning_rate": 9.79671944910557e-06,
+      "loss": 0.0571,
+      "step": 1639
+    },
+    {
+      "epoch": 0.728403286697757,
+      "grad_norm": 0.7510893535708428,
+      "learning_rate": 9.796171956393566e-06,
+      "loss": 0.0669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7288474350433045,
+      "grad_norm": 0.41663397739980335,
+      "learning_rate": 9.79562374273544e-06,
+      "loss": 0.0441,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7292915833888519,
+      "grad_norm": 0.5592036648884667,
+      "learning_rate": 9.795074808213604e-06,
+      "loss": 0.0562,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7297357317343993,
+      "grad_norm": 0.495293158677455,
+      "learning_rate": 9.794525152910573e-06,
+      "loss": 0.0621,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7301798800799467,
+      "grad_norm": 0.5602886889611556,
+      "learning_rate": 9.793974776908963e-06,
+      "loss": 0.0531,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7306240284254941,
+      "grad_norm": 0.5944252332564274,
+      "learning_rate": 9.79342368029151e-06,
+      "loss": 0.0563,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7310681767710415,
+      "grad_norm": 0.482240455293522,
+      "learning_rate": 9.792871863141052e-06,
+      "loss": 0.0576,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7315123251165889,
+      "grad_norm": 0.5733741929298166,
+      "learning_rate": 9.792319325540537e-06,
+      "loss": 0.0684,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7319564734621363,
+      "grad_norm": 0.4985634447799434,
+      "learning_rate": 9.79176606757302e-06,
+      "loss": 0.0597,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7324006218076837,
+      "grad_norm": 0.685246073677029,
+      "learning_rate": 9.791212089321662e-06,
+      "loss": 0.0721,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7328447701532311,
+      "grad_norm": 0.7566932790234222,
+      "learning_rate": 9.790657390869742e-06,
+      "loss": 0.0665,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7332889184987786,
+      "grad_norm": 0.47289806502397586,
+      "learning_rate": 9.790101972300635e-06,
+      "loss": 0.0536,
+      "step": 1651
+    },
+    {
+      "epoch": 0.733733066844326,
+      "grad_norm": 0.9971658246699995,
+      "learning_rate": 9.789545833697833e-06,
+      "loss": 0.0959,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7341772151898734,
+      "grad_norm": 0.588355861783506,
+      "learning_rate": 9.788988975144933e-06,
+      "loss": 0.0556,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7346213635354208,
+      "grad_norm": 0.5671908470000354,
+      "learning_rate": 9.788431396725637e-06,
+      "loss": 0.0669,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7350655118809682,
+      "grad_norm": 0.758993148904425,
+      "learning_rate": 9.787873098523763e-06,
+      "loss": 0.0652,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7355096602265156,
+      "grad_norm": 0.8048604855196553,
+      "learning_rate": 9.787314080623229e-06,
+      "loss": 0.0698,
+      "step": 1656
+    },
+    {
+      "epoch": 0.735953808572063,
+      "grad_norm": 0.8623667194773064,
+      "learning_rate": 9.786754343108066e-06,
+      "loss": 0.0779,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7363979569176105,
+      "grad_norm": 0.61076619104796,
+      "learning_rate": 9.786193886062415e-06,
+      "loss": 0.0769,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.36227469684290603,
+      "learning_rate": 9.785632709570519e-06,
+      "loss": 0.0414,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7372862536087053,
+      "grad_norm": 4.048112931451507,
+      "learning_rate": 9.785070813716733e-06,
+      "loss": 0.0454,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7377304019542528,
+      "grad_norm": 0.6035767336706519,
+      "learning_rate": 9.784508198585519e-06,
+      "loss": 0.0633,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7381745502998002,
+      "grad_norm": 0.469714457030227,
+      "learning_rate": 9.783944864261448e-06,
+      "loss": 0.0472,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7386186986453476,
+      "grad_norm": 0.8129946899965498,
+      "learning_rate": 9.783380810829198e-06,
+      "loss": 0.0613,
+      "step": 1663
+    },
+    {
+      "epoch": 0.739062846990895,
+      "grad_norm": 0.6753504015110883,
+      "learning_rate": 9.782816038373556e-06,
+      "loss": 0.0902,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7395069953364424,
+      "grad_norm": 0.6384325521859789,
+      "learning_rate": 9.782250546979421e-06,
+      "loss": 0.074,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7399511436819898,
+      "grad_norm": 0.6956365636896437,
+      "learning_rate": 9.781684336731791e-06,
+      "loss": 0.0567,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7403952920275372,
+      "grad_norm": 0.8256720537662366,
+      "learning_rate": 9.781117407715779e-06,
+      "loss": 0.0651,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7408394403730846,
+      "grad_norm": 0.5355855493602046,
+      "learning_rate": 9.780549760016602e-06,
+      "loss": 0.0619,
+      "step": 1668
+    },
+    {
+      "epoch": 0.741283588718632,
+      "grad_norm": 0.5253416844707828,
+      "learning_rate": 9.77998139371959e-06,
+      "loss": 0.0592,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7417277370641794,
+      "grad_norm": 0.6458381116052228,
+      "learning_rate": 9.779412308910176e-06,
+      "loss": 0.0631,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7421718854097269,
+      "grad_norm": 0.6458810882415751,
+      "learning_rate": 9.778842505673906e-06,
+      "loss": 0.0721,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7426160337552743,
+      "grad_norm": 0.578667702719237,
+      "learning_rate": 9.778271984096427e-06,
+      "loss": 0.0622,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7430601821008217,
+      "grad_norm": 0.7294244004095107,
+      "learning_rate": 9.777700744263502e-06,
+      "loss": 0.0763,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7435043304463691,
+      "grad_norm": 0.4949092508929545,
+      "learning_rate": 9.777128786260995e-06,
+      "loss": 0.0491,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7439484787919165,
+      "grad_norm": 0.49019523675297216,
+      "learning_rate": 9.776556110174882e-06,
+      "loss": 0.0487,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7443926271374639,
+      "grad_norm": 0.6147792916360124,
+      "learning_rate": 9.775982716091245e-06,
+      "loss": 0.0468,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7448367754830113,
+      "grad_norm": 0.7728423947263149,
+      "learning_rate": 9.775408604096276e-06,
+      "loss": 0.0843,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7452809238285587,
+      "grad_norm": 0.46460994551018925,
+      "learning_rate": 9.774833774276278e-06,
+      "loss": 0.0475,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7457250721741061,
+      "grad_norm": 0.6020989880539805,
+      "learning_rate": 9.77425822671765e-06,
+      "loss": 0.0548,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7461692205196535,
+      "grad_norm": 0.4185492378453488,
+      "learning_rate": 9.77368196150691e-06,
+      "loss": 0.0697,
+      "step": 1680
+    },
+    {
+      "epoch": 0.746613368865201,
+      "grad_norm": 0.47026038790260477,
+      "learning_rate": 9.77310497873068e-06,
+      "loss": 0.0577,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7470575172107484,
+      "grad_norm": 0.7242331312035356,
+      "learning_rate": 9.772527278475694e-06,
+      "loss": 0.0646,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7475016655562958,
+      "grad_norm": 0.5498639479583922,
+      "learning_rate": 9.771948860828783e-06,
+      "loss": 0.0768,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7479458139018432,
+      "grad_norm": 0.5132952000374971,
+      "learning_rate": 9.7713697258769e-06,
+      "loss": 0.0749,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7483899622473906,
+      "grad_norm": 0.567848025082148,
+      "learning_rate": 9.770789873707095e-06,
+      "loss": 0.0853,
+      "step": 1685
+    },
+    {
+      "epoch": 0.748834110592938,
+      "grad_norm": 0.5686689909367685,
+      "learning_rate": 9.770209304406531e-06,
+      "loss": 0.0628,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7492782589384854,
+      "grad_norm": 0.6034469126916269,
+      "learning_rate": 9.769628018062477e-06,
+      "loss": 0.0479,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7497224072840328,
+      "grad_norm": 0.43723219237091576,
+      "learning_rate": 9.769046014762307e-06,
+      "loss": 0.0654,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7501665556295802,
+      "grad_norm": 0.6855260472494596,
+      "learning_rate": 9.76846329459351e-06,
+      "loss": 0.0651,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7506107039751277,
+      "grad_norm": 0.722060394293234,
+      "learning_rate": 9.767879857643681e-06,
+      "loss": 0.0552,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7510548523206751,
+      "grad_norm": 0.463106704754472,
+      "learning_rate": 9.767295704000514e-06,
+      "loss": 0.0534,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7514990006662225,
+      "grad_norm": 0.6415358905679467,
+      "learning_rate": 9.766710833751823e-06,
+      "loss": 0.0806,
+      "step": 1692
+    },
+    {
+      "epoch": 0.75194314901177,
+      "grad_norm": 0.5352784756952503,
+      "learning_rate": 9.76612524698552e-06,
+      "loss": 0.0558,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7523872973573174,
+      "grad_norm": 1.1814652854922993,
+      "learning_rate": 9.76553894378963e-06,
+      "loss": 0.0847,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7528314457028648,
+      "grad_norm": 0.6986298733118028,
+      "learning_rate": 9.764951924252284e-06,
+      "loss": 0.05,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7532755940484122,
+      "grad_norm": 0.5447554709895742,
+      "learning_rate": 9.764364188461723e-06,
+      "loss": 0.0485,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7537197423939596,
+      "grad_norm": 0.47846902684606657,
+      "learning_rate": 9.76377573650629e-06,
+      "loss": 0.0551,
+      "step": 1697
+    },
+    {
+      "epoch": 0.754163890739507,
+      "grad_norm": 0.8115375903439791,
+      "learning_rate": 9.763186568474443e-06,
+      "loss": 0.0701,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7546080390850544,
+      "grad_norm": 0.5954397107201211,
+      "learning_rate": 9.762596684454742e-06,
+      "loss": 0.0474,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7550521874306019,
+      "grad_norm": 0.9043348007985654,
+      "learning_rate": 9.762006084535857e-06,
+      "loss": 0.0752,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7554963357761493,
+      "grad_norm": 0.5876955603814114,
+      "learning_rate": 9.761414768806566e-06,
+      "loss": 0.058,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7559404841216967,
+      "grad_norm": 0.6215494346347864,
+      "learning_rate": 9.76082273735575e-06,
+      "loss": 0.074,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7563846324672441,
+      "grad_norm": 0.6691646984262034,
+      "learning_rate": 9.760229990272407e-06,
+      "loss": 0.0752,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7568287808127915,
+      "grad_norm": 0.6530103461059469,
+      "learning_rate": 9.759636527645633e-06,
+      "loss": 0.0512,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7572729291583389,
+      "grad_norm": 0.5915232768461202,
+      "learning_rate": 9.759042349564638e-06,
+      "loss": 0.0505,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7577170775038863,
+      "grad_norm": 0.6211496225849954,
+      "learning_rate": 9.758447456118734e-06,
+      "loss": 0.0527,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7581612258494337,
+      "grad_norm": 0.9060279971096812,
+      "learning_rate": 9.757851847397349e-06,
+      "loss": 0.0687,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7586053741949811,
+      "grad_norm": 1.0250865769852973,
+      "learning_rate": 9.757255523490006e-06,
+      "loss": 0.0935,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7590495225405285,
+      "grad_norm": 0.546436812313742,
+      "learning_rate": 9.756658484486348e-06,
+      "loss": 0.0667,
+      "step": 1709
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 0.5515990500251661,
+      "learning_rate": 9.756060730476117e-06,
+      "loss": 0.0529,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7599378192316234,
+      "grad_norm": 0.618132276932801,
+      "learning_rate": 9.755462261549167e-06,
+      "loss": 0.0506,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7603819675771708,
+      "grad_norm": 0.6625277428767956,
+      "learning_rate": 9.754863077795459e-06,
+      "loss": 0.0718,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7608261159227182,
+      "grad_norm": 0.45483688763929425,
+      "learning_rate": 9.754263179305058e-06,
+      "loss": 0.0446,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7612702642682656,
+      "grad_norm": 0.6438763676440094,
+      "learning_rate": 9.753662566168142e-06,
+      "loss": 0.0856,
+      "step": 1714
+    },
+    {
+      "epoch": 0.761714412613813,
+      "grad_norm": 0.6595899132325235,
+      "learning_rate": 9.75306123847499e-06,
+      "loss": 0.0725,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7621585609593604,
+      "grad_norm": 0.6533315196440966,
+      "learning_rate": 9.752459196315996e-06,
+      "loss": 0.0607,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7626027093049078,
+      "grad_norm": 0.5435323069050279,
+      "learning_rate": 9.751856439781653e-06,
+      "loss": 0.0511,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7630468576504552,
+      "grad_norm": 0.5454956836386448,
+      "learning_rate": 9.751252968962567e-06,
+      "loss": 0.0828,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7634910059960026,
+      "grad_norm": 0.5715320834454507,
+      "learning_rate": 9.75064878394945e-06,
+      "loss": 0.0518,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7639351543415501,
+      "grad_norm": 0.6744811382532548,
+      "learning_rate": 9.750043884833121e-06,
+      "loss": 0.0508,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7643793026870975,
+      "grad_norm": 0.6108688890165651,
+      "learning_rate": 9.749438271704508e-06,
+      "loss": 0.0615,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7648234510326449,
+      "grad_norm": 0.5674222482544424,
+      "learning_rate": 9.748831944654643e-06,
+      "loss": 0.0644,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7652675993781923,
+      "grad_norm": 0.7356158997613679,
+      "learning_rate": 9.74822490377467e-06,
+      "loss": 0.0733,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7657117477237397,
+      "grad_norm": 0.4849462211495647,
+      "learning_rate": 9.747617149155834e-06,
+      "loss": 0.0443,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7661558960692871,
+      "grad_norm": 0.831191855315457,
+      "learning_rate": 9.747008680889493e-06,
+      "loss": 0.0659,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7666000444148345,
+      "grad_norm": 0.6526473429491929,
+      "learning_rate": 9.746399499067109e-06,
+      "loss": 0.064,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7670441927603819,
+      "grad_norm": 0.5803840610292479,
+      "learning_rate": 9.745789603780254e-06,
+      "loss": 0.053,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7674883411059293,
+      "grad_norm": 0.6610162442549882,
+      "learning_rate": 9.745178995120604e-06,
+      "loss": 0.0626,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7679324894514767,
+      "grad_norm": 0.4876204778498726,
+      "learning_rate": 9.744567673179946e-06,
+      "loss": 0.0522,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7683766377970243,
+      "grad_norm": 0.5622038449381063,
+      "learning_rate": 9.743955638050169e-06,
+      "loss": 0.0524,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7688207861425717,
+      "grad_norm": 0.577658138102478,
+      "learning_rate": 9.743342889823273e-06,
+      "loss": 0.0559,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7692649344881191,
+      "grad_norm": 0.6597870381577758,
+      "learning_rate": 9.742729428591368e-06,
+      "loss": 0.0603,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7697090828336665,
+      "grad_norm": 0.8344370622440579,
+      "learning_rate": 9.742115254446665e-06,
+      "loss": 0.0836,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7701532311792139,
+      "grad_norm": 0.5622150768145531,
+      "learning_rate": 9.741500367481481e-06,
+      "loss": 0.0826,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7705973795247613,
+      "grad_norm": 0.6330801870134145,
+      "learning_rate": 9.740884767788253e-06,
+      "loss": 0.0569,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7710415278703087,
+      "grad_norm": 0.5484349976426807,
+      "learning_rate": 9.740268455459507e-06,
+      "loss": 0.0759,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7714856762158561,
+      "grad_norm": 0.6003363970239456,
+      "learning_rate": 9.739651430587891e-06,
+      "loss": 0.0559,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7719298245614035,
+      "grad_norm": 0.6514870858876586,
+      "learning_rate": 9.739033693266152e-06,
+      "loss": 0.0583,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7723739729069509,
+      "grad_norm": 0.7128176043400777,
+      "learning_rate": 9.738415243587146e-06,
+      "loss": 0.0615,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7728181212524984,
+      "grad_norm": 0.4997932884849887,
+      "learning_rate": 9.737796081643838e-06,
+      "loss": 0.0572,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7732622695980458,
+      "grad_norm": 0.6809941857723402,
+      "learning_rate": 9.737176207529296e-06,
+      "loss": 0.0579,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7737064179435932,
+      "grad_norm": 0.579421035714199,
+      "learning_rate": 9.736555621336701e-06,
+      "loss": 0.0553,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7741505662891406,
+      "grad_norm": 0.49198429179195113,
+      "learning_rate": 9.735934323159337e-06,
+      "loss": 0.046,
+      "step": 1743
+    },
+    {
+      "epoch": 0.774594714634688,
+      "grad_norm": 0.6373312482523877,
+      "learning_rate": 9.735312313090593e-06,
+      "loss": 0.0645,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7750388629802354,
+      "grad_norm": 0.49553655680906583,
+      "learning_rate": 9.734689591223971e-06,
+      "loss": 0.0523,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7754830113257828,
+      "grad_norm": 0.7575640777298455,
+      "learning_rate": 9.734066157653075e-06,
+      "loss": 0.1155,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7759271596713302,
+      "grad_norm": 0.3941642367752884,
+      "learning_rate": 9.733442012471617e-06,
+      "loss": 0.0494,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7763713080168776,
+      "grad_norm": 0.6007166723222428,
+      "learning_rate": 9.732817155773417e-06,
+      "loss": 0.0809,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7768154563624251,
+      "grad_norm": 0.6048168057227016,
+      "learning_rate": 9.732191587652402e-06,
+      "loss": 0.0827,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7772596047079725,
+      "grad_norm": 0.5258636663852201,
+      "learning_rate": 9.731565308202607e-06,
+      "loss": 0.0532,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7777037530535199,
+      "grad_norm": 0.5183503036370198,
+      "learning_rate": 9.73093831751817e-06,
+      "loss": 0.0597,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7781479013990673,
+      "grad_norm": 0.4539057898000694,
+      "learning_rate": 9.73031061569334e-06,
+      "loss": 0.0449,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7785920497446147,
+      "grad_norm": 0.43734440001807745,
+      "learning_rate": 9.72968220282247e-06,
+      "loss": 0.0496,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7790361980901621,
+      "grad_norm": 0.4621698388749695,
+      "learning_rate": 9.729053079000021e-06,
+      "loss": 0.0446,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7794803464357095,
+      "grad_norm": 0.8795540562598588,
+      "learning_rate": 9.728423244320561e-06,
+      "loss": 0.0756,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7799244947812569,
+      "grad_norm": 0.517878545048104,
+      "learning_rate": 9.727792698878767e-06,
+      "loss": 0.0577,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7803686431268043,
+      "grad_norm": 0.5755116631603886,
+      "learning_rate": 9.72716144276942e-06,
+      "loss": 0.0507,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7808127914723517,
+      "grad_norm": 0.501557195914397,
+      "learning_rate": 9.726529476087406e-06,
+      "loss": 0.0695,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7812569398178992,
+      "grad_norm": 0.6162038871090194,
+      "learning_rate": 9.725896798927724e-06,
+      "loss": 0.073,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7817010881634466,
+      "grad_norm": 0.44599102883855124,
+      "learning_rate": 9.725263411385471e-06,
+      "loss": 0.046,
+      "step": 1760
+    },
+    {
+      "epoch": 0.782145236508994,
+      "grad_norm": 0.5089144227251173,
+      "learning_rate": 9.724629313555862e-06,
+      "loss": 0.0566,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7825893848545414,
+      "grad_norm": 0.5159558058324487,
+      "learning_rate": 9.723994505534209e-06,
+      "loss": 0.062,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7830335332000888,
+      "grad_norm": 0.7445098154181274,
+      "learning_rate": 9.723358987415933e-06,
+      "loss": 0.0774,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7834776815456362,
+      "grad_norm": 0.5480997952522682,
+      "learning_rate": 9.722722759296568e-06,
+      "loss": 0.0446,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7839218298911836,
+      "grad_norm": 0.6171588111915909,
+      "learning_rate": 9.722085821271747e-06,
+      "loss": 0.0695,
+      "step": 1765
+    },
+    {
+      "epoch": 0.784365978236731,
+      "grad_norm": 0.5440136789733718,
+      "learning_rate": 9.721448173437212e-06,
+      "loss": 0.0509,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7848101265822784,
+      "grad_norm": 0.5213635915956231,
+      "learning_rate": 9.720809815888814e-06,
+      "loss": 0.0543,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7852542749278258,
+      "grad_norm": 0.5190117014644481,
+      "learning_rate": 9.720170748722507e-06,
+      "loss": 0.0625,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7856984232733734,
+      "grad_norm": 0.4927607718599878,
+      "learning_rate": 9.719530972034356e-06,
+      "loss": 0.0609,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7861425716189208,
+      "grad_norm": 0.48217603174270474,
+      "learning_rate": 9.718890485920529e-06,
+      "loss": 0.0496,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7865867199644682,
+      "grad_norm": 0.5424722566403071,
+      "learning_rate": 9.7182492904773e-06,
+      "loss": 0.0553,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7870308683100156,
+      "grad_norm": 0.549948589003628,
+      "learning_rate": 9.717607385801055e-06,
+      "loss": 0.0673,
+      "step": 1772
+    },
+    {
+      "epoch": 0.787475016655563,
+      "grad_norm": 0.5315442944470616,
+      "learning_rate": 9.716964771988281e-06,
+      "loss": 0.0696,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7879191650011104,
+      "grad_norm": 0.6429809896985884,
+      "learning_rate": 9.716321449135578e-06,
+      "loss": 0.0789,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7883633133466578,
+      "grad_norm": 0.7545678487097474,
+      "learning_rate": 9.715677417339641e-06,
+      "loss": 0.0791,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7888074616922052,
+      "grad_norm": 0.5539973797183765,
+      "learning_rate": 9.715032676697285e-06,
+      "loss": 0.059,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7892516100377526,
+      "grad_norm": 0.5317979991080277,
+      "learning_rate": 9.714387227305422e-06,
+      "loss": 0.0622,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7896957583833,
+      "grad_norm": 0.5252592441778533,
+      "learning_rate": 9.713741069261076e-06,
+      "loss": 0.0463,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7901399067288475,
+      "grad_norm": 0.4965411657812668,
+      "learning_rate": 9.713094202661374e-06,
+      "loss": 0.0498,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7905840550743949,
+      "grad_norm": 0.7856085130172293,
+      "learning_rate": 9.712446627603553e-06,
+      "loss": 0.0732,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7910282034199423,
+      "grad_norm": 0.6433472150837076,
+      "learning_rate": 9.711798344184952e-06,
+      "loss": 0.0536,
+      "step": 1781
+    },
+    {
+      "epoch": 0.7914723517654897,
+      "grad_norm": 0.5926660288471911,
+      "learning_rate": 9.711149352503022e-06,
+      "loss": 0.0611,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7919165001110371,
+      "grad_norm": 0.5407209189318974,
+      "learning_rate": 9.710499652655313e-06,
+      "loss": 0.0464,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7923606484565845,
+      "grad_norm": 1.2981488377904062,
+      "learning_rate": 9.709849244739493e-06,
+      "loss": 0.111,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7928047968021319,
+      "grad_norm": 0.6752401202019049,
+      "learning_rate": 9.709198128853323e-06,
+      "loss": 0.0662,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7932489451476793,
+      "grad_norm": 0.5128430129414299,
+      "learning_rate": 9.708546305094679e-06,
+      "loss": 0.0892,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7936930934932267,
+      "grad_norm": 0.5031597058072875,
+      "learning_rate": 9.707893773561541e-06,
+      "loss": 0.0552,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7941372418387741,
+      "grad_norm": 0.7529404051762868,
+      "learning_rate": 9.707240534351995e-06,
+      "loss": 0.0829,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7945813901843216,
+      "grad_norm": 0.549999308663066,
+      "learning_rate": 9.706586587564236e-06,
+      "loss": 0.0456,
+      "step": 1789
+    },
+    {
+      "epoch": 0.795025538529869,
+      "grad_norm": 0.5507988655050372,
+      "learning_rate": 9.705931933296563e-06,
+      "loss": 0.0565,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7954696868754164,
+      "grad_norm": 0.5120352810106473,
+      "learning_rate": 9.705276571647377e-06,
+      "loss": 0.0529,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7959138352209638,
+      "grad_norm": 0.6321538206444037,
+      "learning_rate": 9.704620502715196e-06,
+      "loss": 0.0604,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7963579835665112,
+      "grad_norm": 0.5978551293263742,
+      "learning_rate": 9.703963726598636e-06,
+      "loss": 0.0615,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7968021319120586,
+      "grad_norm": 0.6845514683107613,
+      "learning_rate": 9.70330624339642e-06,
+      "loss": 0.0759,
+      "step": 1794
+    },
+    {
+      "epoch": 0.797246280257606,
+      "grad_norm": 0.5162860917319786,
+      "learning_rate": 9.702648053207381e-06,
+      "loss": 0.0606,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7976904286031534,
+      "grad_norm": 0.5757276725457067,
+      "learning_rate": 9.701989156130459e-06,
+      "loss": 0.0494,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7981345769487008,
+      "grad_norm": 0.7240460732805346,
+      "learning_rate": 9.70132955226469e-06,
+      "loss": 0.0681,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7985787252942482,
+      "grad_norm": 0.6852456458554386,
+      "learning_rate": 9.700669241709229e-06,
+      "loss": 0.0696,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7990228736397957,
+      "grad_norm": 0.7728747951665925,
+      "learning_rate": 9.70000822456333e-06,
+      "loss": 0.075,
+      "step": 1799
+    },
+    {
+      "epoch": 0.7994670219853431,
+      "grad_norm": 0.44864271076401085,
+      "learning_rate": 9.699346500926357e-06,
+      "loss": 0.0396,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7999111703308905,
+      "grad_norm": 0.45736945189860095,
+      "learning_rate": 9.698684070897774e-06,
+      "loss": 0.0528,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8003553186764379,
+      "grad_norm": 0.7755479737213115,
+      "learning_rate": 9.69802093457716e-06,
+      "loss": 0.0604,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8007994670219853,
+      "grad_norm": 0.5300657753961209,
+      "learning_rate": 9.697357092064196e-06,
+      "loss": 0.0675,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8012436153675327,
+      "grad_norm": 0.5613880914239765,
+      "learning_rate": 9.696692543458666e-06,
+      "loss": 0.0565,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8016877637130801,
+      "grad_norm": 1.5622355700759611,
+      "learning_rate": 9.696027288860463e-06,
+      "loss": 0.0748,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8021319120586275,
+      "grad_norm": 0.6809476301245401,
+      "learning_rate": 9.695361328369588e-06,
+      "loss": 0.077,
+      "step": 1806
+    },
+    {
+      "epoch": 0.802576060404175,
+      "grad_norm": 0.573004020361602,
+      "learning_rate": 9.694694662086143e-06,
+      "loss": 0.0688,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8030202087497224,
+      "grad_norm": 0.48241041676332075,
+      "learning_rate": 9.694027290110344e-06,
+      "loss": 0.046,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8034643570952699,
+      "grad_norm": 0.726848223856282,
+      "learning_rate": 9.693359212542504e-06,
+      "loss": 0.0554,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8039085054408173,
+      "grad_norm": 0.5939220725933183,
+      "learning_rate": 9.692690429483049e-06,
+      "loss": 0.0599,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8043526537863647,
+      "grad_norm": 0.8703575105729704,
+      "learning_rate": 9.692020941032508e-06,
+      "loss": 0.0697,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8047968021319121,
+      "grad_norm": 0.6745036941332984,
+      "learning_rate": 9.691350747291514e-06,
+      "loss": 0.0622,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8052409504774595,
+      "grad_norm": 0.5924055978232626,
+      "learning_rate": 9.690679848360811e-06,
+      "loss": 0.068,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8056850988230069,
+      "grad_norm": 0.6624345407427882,
+      "learning_rate": 9.690008244341247e-06,
+      "loss": 0.0671,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8061292471685543,
+      "grad_norm": 0.760605230819851,
+      "learning_rate": 9.689335935333775e-06,
+      "loss": 0.0703,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8065733955141017,
+      "grad_norm": 0.6543633249820602,
+      "learning_rate": 9.688662921439454e-06,
+      "loss": 0.0537,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8070175438596491,
+      "grad_norm": 0.6888300517583433,
+      "learning_rate": 9.687989202759448e-06,
+      "loss": 0.0495,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8074616922051966,
+      "grad_norm": 0.686773842932207,
+      "learning_rate": 9.68731477939503e-06,
+      "loss": 0.0659,
+      "step": 1818
+    },
+    {
+      "epoch": 0.807905840550744,
+      "grad_norm": 0.5300847493623394,
+      "learning_rate": 9.686639651447578e-06,
+      "loss": 0.0518,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8083499888962914,
+      "grad_norm": 0.5159319768487426,
+      "learning_rate": 9.685963819018575e-06,
+      "loss": 0.0515,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8087941372418388,
+      "grad_norm": 0.7950101365777946,
+      "learning_rate": 9.685287282209607e-06,
+      "loss": 0.0728,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8092382855873862,
+      "grad_norm": 1.1412965984185115,
+      "learning_rate": 9.684610041122375e-06,
+      "loss": 0.0802,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8096824339329336,
+      "grad_norm": 0.45513195688472363,
+      "learning_rate": 9.683932095858673e-06,
+      "loss": 0.0615,
+      "step": 1823
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.7127474010386736,
+      "learning_rate": 9.683253446520412e-06,
+      "loss": 0.0617,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8105707306240284,
+      "grad_norm": 0.5146350672182237,
+      "learning_rate": 9.682574093209603e-06,
+      "loss": 0.0821,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8110148789695758,
+      "grad_norm": 0.5927766784074162,
+      "learning_rate": 9.681894036028365e-06,
+      "loss": 0.0616,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8114590273151232,
+      "grad_norm": 0.6070773128654839,
+      "learning_rate": 9.681213275078922e-06,
+      "loss": 0.0615,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8119031756606707,
+      "grad_norm": 0.532740364018843,
+      "learning_rate": 9.680531810463606e-06,
+      "loss": 0.0572,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8123473240062181,
+      "grad_norm": 0.5406785460306237,
+      "learning_rate": 9.679849642284846e-06,
+      "loss": 0.0553,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8127914723517655,
+      "grad_norm": 0.7113722607782379,
+      "learning_rate": 9.679166770645193e-06,
+      "loss": 0.0588,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8132356206973129,
+      "grad_norm": 0.5228564923020365,
+      "learning_rate": 9.678483195647286e-06,
+      "loss": 0.0762,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8136797690428603,
+      "grad_norm": 0.3998194117652634,
+      "learning_rate": 9.67779891739388e-06,
+      "loss": 0.045,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8141239173884077,
+      "grad_norm": 0.46502028820157165,
+      "learning_rate": 9.677113935987839e-06,
+      "loss": 0.0508,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8145680657339551,
+      "grad_norm": 0.6755084418924148,
+      "learning_rate": 9.67642825153212e-06,
+      "loss": 0.0935,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8150122140795025,
+      "grad_norm": 0.6820496339271076,
+      "learning_rate": 9.675741864129797e-06,
+      "loss": 0.0648,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8154563624250499,
+      "grad_norm": 0.51959592407304,
+      "learning_rate": 9.675054773884045e-06,
+      "loss": 0.0804,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8159005107705973,
+      "grad_norm": 0.4703704489029568,
+      "learning_rate": 9.674366980898145e-06,
+      "loss": 0.0639,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8163446591161448,
+      "grad_norm": 0.4874780271160304,
+      "learning_rate": 9.673678485275484e-06,
+      "loss": 0.0508,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8167888074616922,
+      "grad_norm": 0.44148176077100115,
+      "learning_rate": 9.672989287119555e-06,
+      "loss": 0.0588,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8172329558072396,
+      "grad_norm": 0.7005816504280763,
+      "learning_rate": 9.672299386533956e-06,
+      "loss": 0.074,
+      "step": 1840
+    },
+    {
+      "epoch": 0.817677104152787,
+      "grad_norm": 0.6435780803327446,
+      "learning_rate": 9.67160878362239e-06,
+      "loss": 0.064,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8181212524983345,
+      "grad_norm": 0.6068596716742378,
+      "learning_rate": 9.670917478488669e-06,
+      "loss": 0.0626,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8185654008438819,
+      "grad_norm": 0.48495659068419156,
+      "learning_rate": 9.670225471236703e-06,
+      "loss": 0.0566,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8190095491894293,
+      "grad_norm": 0.5732730166801319,
+      "learning_rate": 9.669532761970518e-06,
+      "loss": 0.0594,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8194536975349767,
+      "grad_norm": 0.6406816449651614,
+      "learning_rate": 9.668839350794236e-06,
+      "loss": 0.0602,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8198978458805241,
+      "grad_norm": 0.7660699462244367,
+      "learning_rate": 9.66814523781209e-06,
+      "loss": 0.1004,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8203419942260715,
+      "grad_norm": 0.6805404383061566,
+      "learning_rate": 9.667450423128417e-06,
+      "loss": 0.0727,
+      "step": 1847
+    },
+    {
+      "epoch": 0.820786142571619,
+      "grad_norm": 1.8549554381012934,
+      "learning_rate": 9.666754906847659e-06,
+      "loss": 0.0612,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8212302909171664,
+      "grad_norm": 0.9016082075933314,
+      "learning_rate": 9.666058689074364e-06,
+      "loss": 0.0792,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8216744392627138,
+      "grad_norm": 0.5403087387935102,
+      "learning_rate": 9.665361769913187e-06,
+      "loss": 0.045,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8221185876082612,
+      "grad_norm": 0.45933630900068073,
+      "learning_rate": 9.664664149468885e-06,
+      "loss": 0.046,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8225627359538086,
+      "grad_norm": 0.6057928280686538,
+      "learning_rate": 9.663965827846321e-06,
+      "loss": 0.053,
+      "step": 1852
+    },
+    {
+      "epoch": 0.823006884299356,
+      "grad_norm": 0.7862585104580586,
+      "learning_rate": 9.663266805150468e-06,
+      "loss": 0.0706,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8234510326449034,
+      "grad_norm": 0.59770705811337,
+      "learning_rate": 9.662567081486398e-06,
+      "loss": 0.0568,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8238951809904508,
+      "grad_norm": 0.5457681124884074,
+      "learning_rate": 9.661866656959293e-06,
+      "loss": 0.0534,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8243393293359982,
+      "grad_norm": 0.7281386032100584,
+      "learning_rate": 9.661165531674438e-06,
+      "loss": 0.081,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8247834776815456,
+      "grad_norm": 0.5330397841966769,
+      "learning_rate": 9.660463705737224e-06,
+      "loss": 0.0657,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8252276260270931,
+      "grad_norm": 0.5773185928330131,
+      "learning_rate": 9.65976117925315e-06,
+      "loss": 0.0657,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8256717743726405,
+      "grad_norm": 0.6769338412517576,
+      "learning_rate": 9.659057952327812e-06,
+      "loss": 0.0713,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8261159227181879,
+      "grad_norm": 0.5999087119449273,
+      "learning_rate": 9.65835402506692e-06,
+      "loss": 0.0776,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8265600710637353,
+      "grad_norm": 0.46186334132300766,
+      "learning_rate": 9.657649397576289e-06,
+      "loss": 0.0435,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8270042194092827,
+      "grad_norm": 0.6132380100477,
+      "learning_rate": 9.656944069961832e-06,
+      "loss": 0.0503,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8274483677548301,
+      "grad_norm": 0.511934760130446,
+      "learning_rate": 9.656238042329575e-06,
+      "loss": 0.047,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8278925161003775,
+      "grad_norm": 0.9622008544679613,
+      "learning_rate": 9.655531314785643e-06,
+      "loss": 0.0727,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8283366644459249,
+      "grad_norm": 0.41625404724105025,
+      "learning_rate": 9.654823887436272e-06,
+      "loss": 0.0452,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8287808127914723,
+      "grad_norm": 0.6534224883169892,
+      "learning_rate": 9.6541157603878e-06,
+      "loss": 0.0812,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8292249611370197,
+      "grad_norm": 0.3831024637275559,
+      "learning_rate": 9.653406933746667e-06,
+      "loss": 0.0406,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8296691094825672,
+      "grad_norm": 0.5750420337273457,
+      "learning_rate": 9.652697407619425e-06,
+      "loss": 0.0655,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8301132578281146,
+      "grad_norm": 0.5340996381251997,
+      "learning_rate": 9.651987182112727e-06,
+      "loss": 0.0684,
+      "step": 1869
+    },
+    {
+      "epoch": 0.830557406173662,
+      "grad_norm": 0.5216822693020573,
+      "learning_rate": 9.651276257333334e-06,
+      "loss": 0.0506,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8310015545192094,
+      "grad_norm": 0.6549627952655241,
+      "learning_rate": 9.650564633388106e-06,
+      "loss": 0.0746,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8314457028647568,
+      "grad_norm": 0.41543419779685914,
+      "learning_rate": 9.649852310384017e-06,
+      "loss": 0.0418,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8318898512103042,
+      "grad_norm": 0.8710457342050398,
+      "learning_rate": 9.649139288428136e-06,
+      "loss": 0.0725,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8323339995558516,
+      "grad_norm": 0.7280421909189481,
+      "learning_rate": 9.648425567627646e-06,
+      "loss": 0.0834,
+      "step": 1874
+    },
+    {
+      "epoch": 0.832778147901399,
+      "grad_norm": 0.48421338065941566,
+      "learning_rate": 9.647711148089829e-06,
+      "loss": 0.051,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8332222962469464,
+      "grad_norm": 0.8228707840085564,
+      "learning_rate": 9.646996029922078e-06,
+      "loss": 0.0899,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8336664445924938,
+      "grad_norm": 0.44548368481732253,
+      "learning_rate": 9.646280213231882e-06,
+      "loss": 0.0459,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8341105929380414,
+      "grad_norm": 0.77874935532614,
+      "learning_rate": 9.645563698126846e-06,
+      "loss": 0.0874,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8345547412835888,
+      "grad_norm": 0.6153818855236423,
+      "learning_rate": 9.64484648471467e-06,
+      "loss": 0.0538,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8349988896291362,
+      "grad_norm": 0.6341254163576385,
+      "learning_rate": 9.644128573103166e-06,
+      "loss": 0.0794,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8354430379746836,
+      "grad_norm": 1.0916751108937903,
+      "learning_rate": 9.643409963400247e-06,
+      "loss": 0.0775,
+      "step": 1881
+    },
+    {
+      "epoch": 0.835887186320231,
+      "grad_norm": 0.5090370659080116,
+      "learning_rate": 9.642690655713935e-06,
+      "loss": 0.0515,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8363313346657784,
+      "grad_norm": 0.8125756753691618,
+      "learning_rate": 9.641970650152351e-06,
+      "loss": 0.0856,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8367754830113258,
+      "grad_norm": 0.7917848803419996,
+      "learning_rate": 9.641249946823722e-06,
+      "loss": 0.0789,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8372196313568732,
+      "grad_norm": 0.6341635191576055,
+      "learning_rate": 9.640528545836388e-06,
+      "loss": 0.0689,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8376637797024206,
+      "grad_norm": 0.45285744978069686,
+      "learning_rate": 9.639806447298786e-06,
+      "loss": 0.0502,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8381079280479681,
+      "grad_norm": 0.5918006498840634,
+      "learning_rate": 9.639083651319455e-06,
+      "loss": 0.077,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8385520763935155,
+      "grad_norm": 0.5566719058184162,
+      "learning_rate": 9.638360158007049e-06,
+      "loss": 0.0518,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8389962247390629,
+      "grad_norm": 0.5775755465954959,
+      "learning_rate": 9.637635967470317e-06,
+      "loss": 0.0583,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8394403730846103,
+      "grad_norm": 0.5105748469342627,
+      "learning_rate": 9.636911079818121e-06,
+      "loss": 0.0547,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8398845214301577,
+      "grad_norm": 0.7153133794510262,
+      "learning_rate": 9.636185495159423e-06,
+      "loss": 0.0598,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8403286697757051,
+      "grad_norm": 0.580672148124565,
+      "learning_rate": 9.63545921360329e-06,
+      "loss": 0.0601,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8407728181212525,
+      "grad_norm": 0.43212755100892075,
+      "learning_rate": 9.634732235258895e-06,
+      "loss": 0.0501,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8412169664667999,
+      "grad_norm": 0.6336655954017367,
+      "learning_rate": 9.634004560235513e-06,
+      "loss": 0.0742,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8416611148123473,
+      "grad_norm": 0.6609701387334967,
+      "learning_rate": 9.633276188642529e-06,
+      "loss": 0.0579,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.6932382794686757,
+      "learning_rate": 9.632547120589426e-06,
+      "loss": 0.0581,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8425494115034422,
+      "grad_norm": 0.5148690171452629,
+      "learning_rate": 9.631817356185799e-06,
+      "loss": 0.0507,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8429935598489896,
+      "grad_norm": 0.5764411558916347,
+      "learning_rate": 9.631086895541343e-06,
+      "loss": 0.0485,
+      "step": 1898
+    },
+    {
+      "epoch": 0.843437708194537,
+      "grad_norm": 0.713479181836388,
+      "learning_rate": 9.630355738765859e-06,
+      "loss": 0.0811,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8438818565400844,
+      "grad_norm": 0.6095181273242816,
+      "learning_rate": 9.62962388596925e-06,
+      "loss": 0.051,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8443260048856318,
+      "grad_norm": 0.6746333568364197,
+      "learning_rate": 9.628891337261527e-06,
+      "loss": 0.051,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8447701532311792,
+      "grad_norm": 0.5956567675202917,
+      "learning_rate": 9.628158092752807e-06,
+      "loss": 0.0544,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8452143015767266,
+      "grad_norm": 0.9239154351205555,
+      "learning_rate": 9.627424152553305e-06,
+      "loss": 0.0801,
+      "step": 1903
+    },
+    {
+      "epoch": 0.845658449922274,
+      "grad_norm": 0.8434016325954864,
+      "learning_rate": 9.626689516773348e-06,
+      "loss": 0.0597,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8461025982678214,
+      "grad_norm": 0.6102688127848747,
+      "learning_rate": 9.625954185523361e-06,
+      "loss": 0.0751,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8465467466133688,
+      "grad_norm": 0.7017970233762413,
+      "learning_rate": 9.62521815891388e-06,
+      "loss": 0.0665,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8469908949589163,
+      "grad_norm": 0.6370603868487351,
+      "learning_rate": 9.624481437055542e-06,
+      "loss": 0.0584,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8474350433044637,
+      "grad_norm": 0.8073786270061732,
+      "learning_rate": 9.623744020059086e-06,
+      "loss": 0.0739,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8478791916500111,
+      "grad_norm": 0.6765473850641361,
+      "learning_rate": 9.623005908035362e-06,
+      "loss": 0.0578,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8483233399955585,
+      "grad_norm": 0.7723618661951789,
+      "learning_rate": 9.622267101095318e-06,
+      "loss": 0.0477,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8487674883411059,
+      "grad_norm": 0.5605923323935563,
+      "learning_rate": 9.621527599350008e-06,
+      "loss": 0.0603,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8492116366866533,
+      "grad_norm": 0.7755772246757074,
+      "learning_rate": 9.620787402910597e-06,
+      "loss": 0.0983,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8496557850322007,
+      "grad_norm": 0.7813757821925909,
+      "learning_rate": 9.620046511888343e-06,
+      "loss": 0.0499,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8500999333777481,
+      "grad_norm": 0.8027788261683163,
+      "learning_rate": 9.619304926394619e-06,
+      "loss": 0.0588,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8505440817232955,
+      "grad_norm": 0.8601044017181758,
+      "learning_rate": 9.618562646540897e-06,
+      "loss": 0.0638,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8509882300688429,
+      "grad_norm": 0.5716901663276188,
+      "learning_rate": 9.617819672438754e-06,
+      "loss": 0.056,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8514323784143905,
+      "grad_norm": 0.7888611302795855,
+      "learning_rate": 9.617076004199868e-06,
+      "loss": 0.08,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8518765267599379,
+      "grad_norm": 0.5937752892253898,
+      "learning_rate": 9.616331641936031e-06,
+      "loss": 0.0545,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8523206751054853,
+      "grad_norm": 0.6861875565077891,
+      "learning_rate": 9.61558658575913e-06,
+      "loss": 0.055,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8527648234510327,
+      "grad_norm": 0.8458011501127822,
+      "learning_rate": 9.614840835781159e-06,
+      "loss": 0.1023,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8532089717965801,
+      "grad_norm": 0.5027378978279742,
+      "learning_rate": 9.614094392114218e-06,
+      "loss": 0.0518,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8536531201421275,
+      "grad_norm": 0.5758297033096836,
+      "learning_rate": 9.613347254870511e-06,
+      "loss": 0.0476,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8540972684876749,
+      "grad_norm": 0.8773520125115069,
+      "learning_rate": 9.612599424162344e-06,
+      "loss": 0.0858,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8545414168332223,
+      "grad_norm": 0.5342112118627843,
+      "learning_rate": 9.61185090010213e-06,
+      "loss": 0.0532,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8549855651787697,
+      "grad_norm": 0.617753551302806,
+      "learning_rate": 9.611101682802383e-06,
+      "loss": 0.0664,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8554297135243171,
+      "grad_norm": 0.6816206410848374,
+      "learning_rate": 9.610351772375724e-06,
+      "loss": 0.0919,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8558738618698646,
+      "grad_norm": 0.636349806575952,
+      "learning_rate": 9.609601168934878e-06,
+      "loss": 0.0544,
+      "step": 1927
+    },
+    {
+      "epoch": 0.856318010215412,
+      "grad_norm": 0.5095901638429818,
+      "learning_rate": 9.608849872592674e-06,
+      "loss": 0.0584,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8567621585609594,
+      "grad_norm": 0.5240679834795968,
+      "learning_rate": 9.608097883462043e-06,
+      "loss": 0.0553,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8572063069065068,
+      "grad_norm": 0.6099517865434585,
+      "learning_rate": 9.60734520165602e-06,
+      "loss": 0.0657,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8576504552520542,
+      "grad_norm": 0.555945227169772,
+      "learning_rate": 9.60659182728775e-06,
+      "loss": 0.0737,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8580946035976016,
+      "grad_norm": 0.5341529189989662,
+      "learning_rate": 9.605837760470476e-06,
+      "loss": 0.0494,
+      "step": 1932
+    },
+    {
+      "epoch": 0.858538751943149,
+      "grad_norm": 0.5439610693418829,
+      "learning_rate": 9.605083001317547e-06,
+      "loss": 0.0566,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8589829002886964,
+      "grad_norm": 0.6077814464368465,
+      "learning_rate": 9.604327549942415e-06,
+      "loss": 0.0615,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8594270486342438,
+      "grad_norm": 0.7793330895236799,
+      "learning_rate": 9.603571406458641e-06,
+      "loss": 0.0669,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8598711969797912,
+      "grad_norm": 0.5859164596399333,
+      "learning_rate": 9.60281457097988e-06,
+      "loss": 0.0554,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8603153453253387,
+      "grad_norm": 0.5913578923102047,
+      "learning_rate": 9.602057043619903e-06,
+      "loss": 0.0819,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8607594936708861,
+      "grad_norm": 0.3937188804578706,
+      "learning_rate": 9.601298824492577e-06,
+      "loss": 0.0455,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8612036420164335,
+      "grad_norm": 0.5687600158251599,
+      "learning_rate": 9.600539913711876e-06,
+      "loss": 0.0703,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8616477903619809,
+      "grad_norm": 0.7038982027083199,
+      "learning_rate": 9.599780311391876e-06,
+      "loss": 0.0559,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8620919387075283,
+      "grad_norm": 0.5434923734263241,
+      "learning_rate": 9.599020017646758e-06,
+      "loss": 0.059,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8625360870530757,
+      "grad_norm": 0.6288012353236909,
+      "learning_rate": 9.59825903259081e-06,
+      "loss": 0.0664,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8629802353986231,
+      "grad_norm": 0.7037145137062264,
+      "learning_rate": 9.597497356338415e-06,
+      "loss": 0.069,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8634243837441705,
+      "grad_norm": 0.5095647439957363,
+      "learning_rate": 9.59673498900407e-06,
+      "loss": 0.0618,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8638685320897179,
+      "grad_norm": 0.6494185255511562,
+      "learning_rate": 9.595971930702372e-06,
+      "loss": 0.0658,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8643126804352653,
+      "grad_norm": 0.5454132982093647,
+      "learning_rate": 9.595208181548022e-06,
+      "loss": 0.0591,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8647568287808128,
+      "grad_norm": 0.7893076144618603,
+      "learning_rate": 9.594443741655823e-06,
+      "loss": 0.0582,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8652009771263602,
+      "grad_norm": 0.7430564668455467,
+      "learning_rate": 9.593678611140683e-06,
+      "loss": 0.0836,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8656451254719076,
+      "grad_norm": 0.6526222088648067,
+      "learning_rate": 9.592912790117614e-06,
+      "loss": 0.0612,
+      "step": 1949
+    },
+    {
+      "epoch": 0.866089273817455,
+      "grad_norm": 0.5060704004253027,
+      "learning_rate": 9.592146278701734e-06,
+      "loss": 0.0528,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8665334221630024,
+      "grad_norm": 0.7447569328611817,
+      "learning_rate": 9.591379077008263e-06,
+      "loss": 0.0657,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8669775705085498,
+      "grad_norm": 0.6783080197900007,
+      "learning_rate": 9.590611185152521e-06,
+      "loss": 0.0748,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8674217188540972,
+      "grad_norm": 0.7292228470690477,
+      "learning_rate": 9.589842603249935e-06,
+      "loss": 0.0626,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8678658671996446,
+      "grad_norm": 0.5752723662054593,
+      "learning_rate": 9.58907333141604e-06,
+      "loss": 0.0562,
+      "step": 1954
+    },
+    {
+      "epoch": 0.868310015545192,
+      "grad_norm": 0.5789129306578911,
+      "learning_rate": 9.588303369766469e-06,
+      "loss": 0.0523,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8687541638907396,
+      "grad_norm": 0.5972967849185621,
+      "learning_rate": 9.58753271841696e-06,
+      "loss": 0.0638,
+      "step": 1956
+    },
+    {
+      "epoch": 0.869198312236287,
+      "grad_norm": 0.9160241760382546,
+      "learning_rate": 9.586761377483355e-06,
+      "loss": 0.083,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8696424605818344,
+      "grad_norm": 0.6452629891186096,
+      "learning_rate": 9.585989347081599e-06,
+      "loss": 0.0765,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8700866089273818,
+      "grad_norm": 0.5181341158162395,
+      "learning_rate": 9.58521662732774e-06,
+      "loss": 0.063,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8705307572729292,
+      "grad_norm": 0.5208504155693663,
+      "learning_rate": 9.584443218337935e-06,
+      "loss": 0.0609,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8709749056184766,
+      "grad_norm": 0.5883552043198282,
+      "learning_rate": 9.583669120228439e-06,
+      "loss": 0.0644,
+      "step": 1961
+    },
+    {
+      "epoch": 0.871419053964024,
+      "grad_norm": 0.5457449002927042,
+      "learning_rate": 9.582894333115608e-06,
+      "loss": 0.0567,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8718632023095714,
+      "grad_norm": 0.6679359172890208,
+      "learning_rate": 9.58211885711591e-06,
+      "loss": 0.0652,
+      "step": 1963
+    },
+    {
+      "epoch": 0.8723073506551188,
+      "grad_norm": 0.5085227180546535,
+      "learning_rate": 9.581342692345913e-06,
+      "loss": 0.0734,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8727514990006662,
+      "grad_norm": 0.63235227685218,
+      "learning_rate": 9.580565838922285e-06,
+      "loss": 0.0501,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8731956473462137,
+      "grad_norm": 0.5297922964410929,
+      "learning_rate": 9.579788296961801e-06,
+      "loss": 0.0525,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8736397956917611,
+      "grad_norm": 0.5775438947377388,
+      "learning_rate": 9.57901006658134e-06,
+      "loss": 0.0544,
+      "step": 1967
+    },
+    {
+      "epoch": 0.8740839440373085,
+      "grad_norm": 0.708123483636742,
+      "learning_rate": 9.57823114789788e-06,
+      "loss": 0.0617,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8745280923828559,
+      "grad_norm": 0.5577152090781191,
+      "learning_rate": 9.577451541028509e-06,
+      "loss": 0.0583,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8749722407284033,
+      "grad_norm": 0.6520169618466134,
+      "learning_rate": 9.576671246090415e-06,
+      "loss": 0.0566,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8754163890739507,
+      "grad_norm": 0.49841809419135175,
+      "learning_rate": 9.575890263200887e-06,
+      "loss": 0.0497,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8758605374194981,
+      "grad_norm": 0.5197540662913069,
+      "learning_rate": 9.575108592477322e-06,
+      "loss": 0.0555,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8763046857650455,
+      "grad_norm": 0.5124464300264364,
+      "learning_rate": 9.57432623403722e-06,
+      "loss": 0.0477,
+      "step": 1973
+    },
+    {
+      "epoch": 0.8767488341105929,
+      "grad_norm": 0.5779510703301827,
+      "learning_rate": 9.57354318799818e-06,
+      "loss": 0.0685,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8771929824561403,
+      "grad_norm": 0.5097058321667625,
+      "learning_rate": 9.572759454477907e-06,
+      "loss": 0.0488,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8776371308016878,
+      "grad_norm": 0.8063636142109883,
+      "learning_rate": 9.57197503359421e-06,
+      "loss": 0.0544,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8780812791472352,
+      "grad_norm": 0.4428063296481966,
+      "learning_rate": 9.571189925465002e-06,
+      "loss": 0.0533,
+      "step": 1977
+    },
+    {
+      "epoch": 0.8785254274927826,
+      "grad_norm": 0.4839122950764424,
+      "learning_rate": 9.570404130208297e-06,
+      "loss": 0.0479,
+      "step": 1978
+    },
+    {
+      "epoch": 0.87896957583833,
+      "grad_norm": 0.7120182407032511,
+      "learning_rate": 9.569617647942214e-06,
+      "loss": 0.0611,
+      "step": 1979
+    },
+    {
+      "epoch": 0.8794137241838774,
+      "grad_norm": 0.792336377588857,
+      "learning_rate": 9.568830478784975e-06,
+      "loss": 0.0618,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8798578725294248,
+      "grad_norm": 0.7252084523924553,
+      "learning_rate": 9.568042622854902e-06,
+      "loss": 0.075,
+      "step": 1981
+    },
+    {
+      "epoch": 0.8803020208749722,
+      "grad_norm": 0.5594674149617587,
+      "learning_rate": 9.567254080270427e-06,
+      "loss": 0.0619,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8807461692205196,
+      "grad_norm": 0.7969483073047114,
+      "learning_rate": 9.566464851150078e-06,
+      "loss": 0.0626,
+      "step": 1983
+    },
+    {
+      "epoch": 0.881190317566067,
+      "grad_norm": 0.760520710752294,
+      "learning_rate": 9.565674935612495e-06,
+      "loss": 0.0584,
+      "step": 1984
+    },
+    {
+      "epoch": 0.8816344659116144,
+      "grad_norm": 0.626615448212967,
+      "learning_rate": 9.564884333776408e-06,
+      "loss": 0.0596,
+      "step": 1985
+    },
+    {
+      "epoch": 0.8820786142571619,
+      "grad_norm": 0.6020936472164464,
+      "learning_rate": 9.564093045760663e-06,
+      "loss": 0.0623,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8825227626027093,
+      "grad_norm": 0.7034061959408842,
+      "learning_rate": 9.563301071684203e-06,
+      "loss": 0.0549,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8829669109482567,
+      "grad_norm": 0.5695284311311956,
+      "learning_rate": 9.562508411666077e-06,
+      "loss": 0.0633,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8834110592938041,
+      "grad_norm": 0.9573516672168828,
+      "learning_rate": 9.56171506582543e-06,
+      "loss": 0.0653,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8838552076393515,
+      "grad_norm": 0.5805029234381518,
+      "learning_rate": 9.56092103428152e-06,
+      "loss": 0.0607,
+      "step": 1990
+    },
+    {
+      "epoch": 0.884299355984899,
+      "grad_norm": 0.9854040215753008,
+      "learning_rate": 9.560126317153702e-06,
+      "loss": 0.0792,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8847435043304464,
+      "grad_norm": 0.7570653217798166,
+      "learning_rate": 9.559330914561435e-06,
+      "loss": 0.0574,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8851876526759938,
+      "grad_norm": 0.6760651246150844,
+      "learning_rate": 9.558534826624281e-06,
+      "loss": 0.0586,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8856318010215412,
+      "grad_norm": 0.7446222818408107,
+      "learning_rate": 9.55773805346191e-06,
+      "loss": 0.0903,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8860759493670886,
+      "grad_norm": 0.4971467452542264,
+      "learning_rate": 9.556940595194085e-06,
+      "loss": 0.0495,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8865200977126361,
+      "grad_norm": 0.48127766478143535,
+      "learning_rate": 9.55614245194068e-06,
+      "loss": 0.0476,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8869642460581835,
+      "grad_norm": 0.5264335771790009,
+      "learning_rate": 9.555343623821669e-06,
+      "loss": 0.0656,
+      "step": 1997
+    },
+    {
+      "epoch": 0.8874083944037309,
+      "grad_norm": 1.0687153694098346,
+      "learning_rate": 9.554544110957128e-06,
+      "loss": 0.1082,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8878525427492783,
+      "grad_norm": 0.6068726497280238,
+      "learning_rate": 9.553743913467241e-06,
+      "loss": 0.0613,
+      "step": 1999
+    },
+    {
+      "epoch": 0.8882966910948257,
+      "grad_norm": 0.7498237482418018,
+      "learning_rate": 9.552943031472289e-06,
+      "loss": 0.0771,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8887408394403731,
+      "grad_norm": 0.6023014910820245,
+      "learning_rate": 9.552141465092659e-06,
+      "loss": 0.0602,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8891849877859205,
+      "grad_norm": 0.6380409827625847,
+      "learning_rate": 9.551339214448838e-06,
+      "loss": 0.0619,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8896291361314679,
+      "grad_norm": 0.9507144923151234,
+      "learning_rate": 9.55053627966142e-06,
+      "loss": 0.0869,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8900732844770153,
+      "grad_norm": 0.5426047915412493,
+      "learning_rate": 9.5497326608511e-06,
+      "loss": 0.0571,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8905174328225627,
+      "grad_norm": 0.44148111117293815,
+      "learning_rate": 9.548928358138672e-06,
+      "loss": 0.0508,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8909615811681102,
+      "grad_norm": 0.6815953588544817,
+      "learning_rate": 9.548123371645042e-06,
+      "loss": 0.0648,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8914057295136576,
+      "grad_norm": 0.7228672896480205,
+      "learning_rate": 9.547317701491207e-06,
+      "loss": 0.0751,
+      "step": 2007
+    },
+    {
+      "epoch": 0.891849877859205,
+      "grad_norm": 0.5288810936079249,
+      "learning_rate": 9.546511347798278e-06,
+      "loss": 0.0718,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8922940262047524,
+      "grad_norm": 0.6326908645656705,
+      "learning_rate": 9.545704310687462e-06,
+      "loss": 0.0643,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8927381745502998,
+      "grad_norm": 0.715169955221305,
+      "learning_rate": 9.54489659028007e-06,
+      "loss": 0.0551,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8931823228958472,
+      "grad_norm": 0.5887866106129909,
+      "learning_rate": 9.544088186697515e-06,
+      "loss": 0.0537,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8936264712413946,
+      "grad_norm": 0.9378252587331583,
+      "learning_rate": 9.543279100061316e-06,
+      "loss": 0.0823,
+      "step": 2012
+    },
+    {
+      "epoch": 0.894070619586942,
+      "grad_norm": 0.5493178853844625,
+      "learning_rate": 9.542469330493092e-06,
+      "loss": 0.0518,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8945147679324894,
+      "grad_norm": 1.0489392256223762,
+      "learning_rate": 9.541658878114564e-06,
+      "loss": 0.0951,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8949589162780368,
+      "grad_norm": 0.9003525371835955,
+      "learning_rate": 9.540847743047556e-06,
+      "loss": 0.0823,
+      "step": 2015
+    },
+    {
+      "epoch": 0.8954030646235843,
+      "grad_norm": 0.6349100893476417,
+      "learning_rate": 9.540035925413997e-06,
+      "loss": 0.0687,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8958472129691317,
+      "grad_norm": 0.6722685205867144,
+      "learning_rate": 9.539223425335919e-06,
+      "loss": 0.0584,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8962913613146791,
+      "grad_norm": 0.5752288899218686,
+      "learning_rate": 9.53841024293545e-06,
+      "loss": 0.0615,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8967355096602265,
+      "grad_norm": 0.6641589129330268,
+      "learning_rate": 9.537596378334827e-06,
+      "loss": 0.0599,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8971796580057739,
+      "grad_norm": 0.6171698716517034,
+      "learning_rate": 9.53678183165639e-06,
+      "loss": 0.0614,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8976238063513213,
+      "grad_norm": 0.46805323991408604,
+      "learning_rate": 9.535966603022578e-06,
+      "loss": 0.0497,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8980679546968687,
+      "grad_norm": 0.5561153083603769,
+      "learning_rate": 9.53515069255593e-06,
+      "loss": 0.0549,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8985121030424161,
+      "grad_norm": 0.8567619540393755,
+      "learning_rate": 9.534334100379095e-06,
+      "loss": 0.087,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8989562513879635,
+      "grad_norm": 0.7529977371586785,
+      "learning_rate": 9.533516826614822e-06,
+      "loss": 0.071,
+      "step": 2024
+    },
+    {
+      "epoch": 0.899400399733511,
+      "grad_norm": 0.47166342806955874,
+      "learning_rate": 9.532698871385957e-06,
+      "loss": 0.0598,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8998445480790584,
+      "grad_norm": 0.5143678735436241,
+      "learning_rate": 9.531880234815454e-06,
+      "loss": 0.0564,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9002886964246058,
+      "grad_norm": 0.7879319949686112,
+      "learning_rate": 9.53106091702637e-06,
+      "loss": 0.0721,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9007328447701533,
+      "grad_norm": 0.9434208137805631,
+      "learning_rate": 9.53024091814186e-06,
+      "loss": 0.0777,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9011769931157007,
+      "grad_norm": 0.5393835319630947,
+      "learning_rate": 9.529420238285185e-06,
+      "loss": 0.0729,
+      "step": 2029
+    },
+    {
+      "epoch": 0.901621141461248,
+      "grad_norm": 1.1442231822471376,
+      "learning_rate": 9.528598877579707e-06,
+      "loss": 0.0577,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9020652898067955,
+      "grad_norm": 0.6771073887283316,
+      "learning_rate": 9.52777683614889e-06,
+      "loss": 0.0524,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9025094381523429,
+      "grad_norm": 0.4601735486943438,
+      "learning_rate": 9.5269541141163e-06,
+      "loss": 0.049,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9029535864978903,
+      "grad_norm": 0.857950337442236,
+      "learning_rate": 9.526130711605609e-06,
+      "loss": 0.0729,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9033977348434377,
+      "grad_norm": 0.6334716311616548,
+      "learning_rate": 9.525306628740585e-06,
+      "loss": 0.0557,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9038418831889852,
+      "grad_norm": 0.5606743700857357,
+      "learning_rate": 9.524481865645105e-06,
+      "loss": 0.0529,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9042860315345326,
+      "grad_norm": 0.7009413752452551,
+      "learning_rate": 9.523656422443142e-06,
+      "loss": 0.0705,
+      "step": 2036
+    },
+    {
+      "epoch": 0.90473017988008,
+      "grad_norm": 0.5676919166524141,
+      "learning_rate": 9.522830299258773e-06,
+      "loss": 0.0642,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9051743282256274,
+      "grad_norm": 0.5909379090037243,
+      "learning_rate": 9.522003496216184e-06,
+      "loss": 0.0611,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9056184765711748,
+      "grad_norm": 0.6412971030112977,
+      "learning_rate": 9.521176013439652e-06,
+      "loss": 0.0668,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9060626249167222,
+      "grad_norm": 0.6697892166854362,
+      "learning_rate": 9.520347851053567e-06,
+      "loss": 0.0589,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9065067732622696,
+      "grad_norm": 0.4472682187482963,
+      "learning_rate": 9.51951900918241e-06,
+      "loss": 0.0468,
+      "step": 2041
+    },
+    {
+      "epoch": 0.906950921607817,
+      "grad_norm": 0.6790472985515921,
+      "learning_rate": 9.518689487950772e-06,
+      "loss": 0.064,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9073950699533644,
+      "grad_norm": 0.4982007942752425,
+      "learning_rate": 9.517859287483347e-06,
+      "loss": 0.0555,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9078392182989118,
+      "grad_norm": 0.4637933782350014,
+      "learning_rate": 9.517028407904925e-06,
+      "loss": 0.0565,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9082833666444593,
+      "grad_norm": 0.5832414481867781,
+      "learning_rate": 9.516196849340402e-06,
+      "loss": 0.0742,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9087275149900067,
+      "grad_norm": 0.5576423922621532,
+      "learning_rate": 9.515364611914777e-06,
+      "loss": 0.0628,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9091716633355541,
+      "grad_norm": 0.5229540500090215,
+      "learning_rate": 9.514531695753146e-06,
+      "loss": 0.0517,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9096158116811015,
+      "grad_norm": 0.6260810656444592,
+      "learning_rate": 9.513698100980715e-06,
+      "loss": 0.0828,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9100599600266489,
+      "grad_norm": 0.4707579422417085,
+      "learning_rate": 9.512863827722785e-06,
+      "loss": 0.0593,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9105041083721963,
+      "grad_norm": 0.5601794044808186,
+      "learning_rate": 9.51202887610476e-06,
+      "loss": 0.063,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9109482567177437,
+      "grad_norm": 0.6023400251451128,
+      "learning_rate": 9.51119324625215e-06,
+      "loss": 0.0686,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5886804844960863,
+      "learning_rate": 9.510356938290562e-06,
+      "loss": 0.0618,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9118365534088385,
+      "grad_norm": 0.5729230838893967,
+      "learning_rate": 9.509519952345709e-06,
+      "loss": 0.0594,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9122807017543859,
+      "grad_norm": 0.5913444178469247,
+      "learning_rate": 9.508682288543405e-06,
+      "loss": 0.0544,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9127248500999334,
+      "grad_norm": 0.5581180816842339,
+      "learning_rate": 9.507843947009562e-06,
+      "loss": 0.0489,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9131689984454808,
+      "grad_norm": 0.9592080607831761,
+      "learning_rate": 9.507004927870202e-06,
+      "loss": 0.0674,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9136131467910282,
+      "grad_norm": 0.48343586169312186,
+      "learning_rate": 9.506165231251438e-06,
+      "loss": 0.0536,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9140572951365756,
+      "grad_norm": 0.5238333120956294,
+      "learning_rate": 9.505324857279494e-06,
+      "loss": 0.0521,
+      "step": 2058
+    },
+    {
+      "epoch": 0.914501443482123,
+      "grad_norm": 0.4166519007433999,
+      "learning_rate": 9.504483806080694e-06,
+      "loss": 0.0586,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9149455918276704,
+      "grad_norm": 0.6437419685785489,
+      "learning_rate": 9.503642077781457e-06,
+      "loss": 0.0767,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9153897401732178,
+      "grad_norm": 0.6570037258452504,
+      "learning_rate": 9.502799672508314e-06,
+      "loss": 0.0587,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9158338885187652,
+      "grad_norm": 0.6138821258144063,
+      "learning_rate": 9.501956590387891e-06,
+      "loss": 0.0736,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9162780368643126,
+      "grad_norm": 0.6641641647032757,
+      "learning_rate": 9.501112831546917e-06,
+      "loss": 0.0606,
+      "step": 2063
+    },
+    {
+      "epoch": 0.91672218520986,
+      "grad_norm": 0.4693758498407545,
+      "learning_rate": 9.500268396112224e-06,
+      "loss": 0.056,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9171663335554076,
+      "grad_norm": 0.7149305856179645,
+      "learning_rate": 9.499423284210745e-06,
+      "loss": 0.0627,
+      "step": 2065
+    },
+    {
+      "epoch": 0.917610481900955,
+      "grad_norm": 0.522841296737736,
+      "learning_rate": 9.498577495969515e-06,
+      "loss": 0.0722,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9180546302465024,
+      "grad_norm": 0.568329647769687,
+      "learning_rate": 9.497731031515669e-06,
+      "loss": 0.0546,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9184987785920498,
+      "grad_norm": 0.7376883712368818,
+      "learning_rate": 9.496883890976445e-06,
+      "loss": 0.0743,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9189429269375972,
+      "grad_norm": 0.8322448418666403,
+      "learning_rate": 9.496036074479184e-06,
+      "loss": 0.0505,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9193870752831446,
+      "grad_norm": 0.47540090764910864,
+      "learning_rate": 9.495187582151328e-06,
+      "loss": 0.0503,
+      "step": 2070
+    },
+    {
+      "epoch": 0.919831223628692,
+      "grad_norm": 0.5694137714424392,
+      "learning_rate": 9.494338414120419e-06,
+      "loss": 0.0552,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9202753719742394,
+      "grad_norm": 0.4737570623495173,
+      "learning_rate": 9.493488570514099e-06,
+      "loss": 0.0546,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9207195203197868,
+      "grad_norm": 0.5080926402077233,
+      "learning_rate": 9.492638051460116e-06,
+      "loss": 0.0634,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9211636686653342,
+      "grad_norm": 0.5930821340624887,
+      "learning_rate": 9.491786857086318e-06,
+      "loss": 0.0558,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9216078170108817,
+      "grad_norm": 0.6489494784499551,
+      "learning_rate": 9.490934987520653e-06,
+      "loss": 0.0574,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9220519653564291,
+      "grad_norm": 0.4274889745604907,
+      "learning_rate": 9.490082442891171e-06,
+      "loss": 0.0484,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9224961137019765,
+      "grad_norm": 0.7132092664683398,
+      "learning_rate": 9.489229223326027e-06,
+      "loss": 0.0506,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9229402620475239,
+      "grad_norm": 0.7098655395611968,
+      "learning_rate": 9.48837532895347e-06,
+      "loss": 0.0693,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9233844103930713,
+      "grad_norm": 0.6125346051926611,
+      "learning_rate": 9.487520759901858e-06,
+      "loss": 0.0665,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9238285587386187,
+      "grad_norm": 0.5007229228967727,
+      "learning_rate": 9.486665516299646e-06,
+      "loss": 0.0763,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9242727070841661,
+      "grad_norm": 0.6228392832149312,
+      "learning_rate": 9.485809598275391e-06,
+      "loss": 0.0602,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9247168554297135,
+      "grad_norm": 0.5392941744359442,
+      "learning_rate": 9.484953005957753e-06,
+      "loss": 0.0621,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9251610037752609,
+      "grad_norm": 0.6769648255970269,
+      "learning_rate": 9.484095739475492e-06,
+      "loss": 0.0829,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9256051521208084,
+      "grad_norm": 1.0001370312264637,
+      "learning_rate": 9.48323779895747e-06,
+      "loss": 0.0616,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9260493004663558,
+      "grad_norm": 0.6960955983895617,
+      "learning_rate": 9.482379184532652e-06,
+      "loss": 0.0701,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9264934488119032,
+      "grad_norm": 0.5402499843217174,
+      "learning_rate": 9.481519896330098e-06,
+      "loss": 0.0557,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9269375971574506,
+      "grad_norm": 0.5511208018996919,
+      "learning_rate": 9.480659934478975e-06,
+      "loss": 0.0613,
+      "step": 2087
+    },
+    {
+      "epoch": 0.927381745502998,
+      "grad_norm": 0.5938957138905275,
+      "learning_rate": 9.479799299108553e-06,
+      "loss": 0.0808,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9278258938485454,
+      "grad_norm": 0.7161437453790136,
+      "learning_rate": 9.478937990348196e-06,
+      "loss": 0.0606,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9282700421940928,
+      "grad_norm": 0.5468536840129006,
+      "learning_rate": 9.478076008327377e-06,
+      "loss": 0.0513,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9287141905396402,
+      "grad_norm": 0.9707762113595154,
+      "learning_rate": 9.477213353175663e-06,
+      "loss": 0.0869,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9291583388851876,
+      "grad_norm": 0.7985679050710678,
+      "learning_rate": 9.476350025022728e-06,
+      "loss": 0.0794,
+      "step": 2092
+    },
+    {
+      "epoch": 0.929602487230735,
+      "grad_norm": 0.6127936826464601,
+      "learning_rate": 9.475486023998345e-06,
+      "loss": 0.055,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9300466355762825,
+      "grad_norm": 0.6334292089301367,
+      "learning_rate": 9.474621350232387e-06,
+      "loss": 0.0801,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9304907839218299,
+      "grad_norm": 0.5293672478661852,
+      "learning_rate": 9.47375600385483e-06,
+      "loss": 0.0463,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9309349322673773,
+      "grad_norm": 0.6942808305366424,
+      "learning_rate": 9.47288998499575e-06,
+      "loss": 0.0772,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9313790806129247,
+      "grad_norm": 0.801229643208364,
+      "learning_rate": 9.472023293785322e-06,
+      "loss": 0.0838,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9318232289584721,
+      "grad_norm": 0.5539233738084255,
+      "learning_rate": 9.471155930353829e-06,
+      "loss": 0.0477,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9322673773040195,
+      "grad_norm": 0.5083333963598577,
+      "learning_rate": 9.470287894831648e-06,
+      "loss": 0.0493,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9327115256495669,
+      "grad_norm": 0.5714592576450407,
+      "learning_rate": 9.469419187349258e-06,
+      "loss": 0.0518,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9331556739951143,
+      "grad_norm": 0.5821855293618007,
+      "learning_rate": 9.468549808037241e-06,
+      "loss": 0.0584,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9335998223406617,
+      "grad_norm": 0.5808549016034106,
+      "learning_rate": 9.467679757026283e-06,
+      "loss": 0.0641,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9340439706862091,
+      "grad_norm": 0.5185841811549224,
+      "learning_rate": 9.466809034447165e-06,
+      "loss": 0.0484,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9344881190317567,
+      "grad_norm": 0.46447107395363035,
+      "learning_rate": 9.46593764043077e-06,
+      "loss": 0.0538,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9349322673773041,
+      "grad_norm": 0.5957468835643678,
+      "learning_rate": 9.465065575108084e-06,
+      "loss": 0.0687,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9353764157228515,
+      "grad_norm": 0.6722403187627196,
+      "learning_rate": 9.464192838610195e-06,
+      "loss": 0.0633,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9358205640683989,
+      "grad_norm": 0.7367314225600079,
+      "learning_rate": 9.463319431068289e-06,
+      "loss": 0.0688,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9362647124139463,
+      "grad_norm": 0.5272343499456363,
+      "learning_rate": 9.462445352613654e-06,
+      "loss": 0.0598,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9367088607594937,
+      "grad_norm": 0.5251390084394981,
+      "learning_rate": 9.461570603377678e-06,
+      "loss": 0.0461,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9371530091050411,
+      "grad_norm": 0.45228924863036196,
+      "learning_rate": 9.460695183491852e-06,
+      "loss": 0.0551,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9375971574505885,
+      "grad_norm": 0.5609307631212084,
+      "learning_rate": 9.459819093087765e-06,
+      "loss": 0.0582,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9380413057961359,
+      "grad_norm": 0.6083990162855326,
+      "learning_rate": 9.45894233229711e-06,
+      "loss": 0.0727,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9384854541416833,
+      "grad_norm": 0.6031459810485845,
+      "learning_rate": 9.458064901251679e-06,
+      "loss": 0.055,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9389296024872308,
+      "grad_norm": 0.5698111663947624,
+      "learning_rate": 9.457186800083363e-06,
+      "loss": 0.0542,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9393737508327782,
+      "grad_norm": 0.5620094721834581,
+      "learning_rate": 9.456308028924157e-06,
+      "loss": 0.0647,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9398178991783256,
+      "grad_norm": 0.47201186226610237,
+      "learning_rate": 9.455428587906154e-06,
+      "loss": 0.0587,
+      "step": 2116
+    },
+    {
+      "epoch": 0.940262047523873,
+      "grad_norm": 0.5090071386100868,
+      "learning_rate": 9.45454847716155e-06,
+      "loss": 0.0485,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9407061958694204,
+      "grad_norm": 0.6100904698718481,
+      "learning_rate": 9.453667696822644e-06,
+      "loss": 0.055,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9411503442149678,
+      "grad_norm": 0.5793242183604661,
+      "learning_rate": 9.452786247021825e-06,
+      "loss": 0.0633,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9415944925605152,
+      "grad_norm": 0.6603367019430398,
+      "learning_rate": 9.451904127891593e-06,
+      "loss": 0.062,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9420386409060626,
+      "grad_norm": 0.48144950864097125,
+      "learning_rate": 9.451021339564549e-06,
+      "loss": 0.0541,
+      "step": 2121
+    },
+    {
+      "epoch": 0.94248278925161,
+      "grad_norm": 0.8619120091253565,
+      "learning_rate": 9.450137882173385e-06,
+      "loss": 0.0622,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9429269375971574,
+      "grad_norm": 0.574402948879313,
+      "learning_rate": 9.449253755850902e-06,
+      "loss": 0.0579,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9433710859427049,
+      "grad_norm": 0.5040572133776431,
+      "learning_rate": 9.448368960730002e-06,
+      "loss": 0.0471,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9438152342882523,
+      "grad_norm": 0.8130523767663201,
+      "learning_rate": 9.447483496943682e-06,
+      "loss": 0.0607,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9442593826337997,
+      "grad_norm": 0.6488075902611455,
+      "learning_rate": 9.446597364625043e-06,
+      "loss": 0.0495,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9447035309793471,
+      "grad_norm": 0.516081567488049,
+      "learning_rate": 9.445710563907286e-06,
+      "loss": 0.0597,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9451476793248945,
+      "grad_norm": 0.6477844164341606,
+      "learning_rate": 9.444823094923712e-06,
+      "loss": 0.0581,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9455918276704419,
+      "grad_norm": 0.5960225920848716,
+      "learning_rate": 9.44393495780772e-06,
+      "loss": 0.0693,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9460359760159893,
+      "grad_norm": 0.6610698243004944,
+      "learning_rate": 9.443046152692818e-06,
+      "loss": 0.0602,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9464801243615367,
+      "grad_norm": 0.501365231071439,
+      "learning_rate": 9.442156679712604e-06,
+      "loss": 0.0507,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9469242727070841,
+      "grad_norm": 0.5333891976903705,
+      "learning_rate": 9.441266539000782e-06,
+      "loss": 0.0551,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.7140797441741807,
+      "learning_rate": 9.440375730691154e-06,
+      "loss": 0.0559,
+      "step": 2133
+    },
+    {
+      "epoch": 0.947812569398179,
+      "grad_norm": 1.0212073230522318,
+      "learning_rate": 9.439484254917626e-06,
+      "loss": 0.0601,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9482567177437264,
+      "grad_norm": 0.6547086006006103,
+      "learning_rate": 9.4385921118142e-06,
+      "loss": 0.0678,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9487008660892738,
+      "grad_norm": 0.5668847306270909,
+      "learning_rate": 9.437699301514983e-06,
+      "loss": 0.0562,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9491450144348212,
+      "grad_norm": 0.535851688994489,
+      "learning_rate": 9.436805824154175e-06,
+      "loss": 0.0536,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9495891627803686,
+      "grad_norm": 0.6725155861850945,
+      "learning_rate": 9.435911679866085e-06,
+      "loss": 0.0636,
+      "step": 2138
+    },
+    {
+      "epoch": 0.950033311125916,
+      "grad_norm": 0.472128745022765,
+      "learning_rate": 9.435016868785117e-06,
+      "loss": 0.0514,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9504774594714634,
+      "grad_norm": 0.5611219746408826,
+      "learning_rate": 9.434121391045775e-06,
+      "loss": 0.057,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9509216078170108,
+      "grad_norm": 0.5197846459934121,
+      "learning_rate": 9.433225246782664e-06,
+      "loss": 0.0497,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9513657561625583,
+      "grad_norm": 0.6901510307382841,
+      "learning_rate": 9.432328436130493e-06,
+      "loss": 0.0588,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9518099045081057,
+      "grad_norm": 0.6595671077813987,
+      "learning_rate": 9.431430959224067e-06,
+      "loss": 0.0852,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9522540528536532,
+      "grad_norm": 0.5676059741434758,
+      "learning_rate": 9.43053281619829e-06,
+      "loss": 0.0507,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9526982011992006,
+      "grad_norm": 0.4855036199663289,
+      "learning_rate": 9.429634007188169e-06,
+      "loss": 0.0474,
+      "step": 2145
+    },
+    {
+      "epoch": 0.953142349544748,
+      "grad_norm": 0.732709155194895,
+      "learning_rate": 9.42873453232881e-06,
+      "loss": 0.0619,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9535864978902954,
+      "grad_norm": 0.6306866348324036,
+      "learning_rate": 9.42783439175542e-06,
+      "loss": 0.0772,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9540306462358428,
+      "grad_norm": 0.6444304070837995,
+      "learning_rate": 9.426933585603304e-06,
+      "loss": 0.0565,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9544747945813902,
+      "grad_norm": 0.5421189880885259,
+      "learning_rate": 9.42603211400787e-06,
+      "loss": 0.0655,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9549189429269376,
+      "grad_norm": 0.5350689023537523,
+      "learning_rate": 9.425129977104626e-06,
+      "loss": 0.0598,
+      "step": 2150
+    },
+    {
+      "epoch": 0.955363091272485,
+      "grad_norm": 0.3832349941062708,
+      "learning_rate": 9.424227175029175e-06,
+      "loss": 0.0474,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9558072396180324,
+      "grad_norm": 0.5764500588860708,
+      "learning_rate": 9.423323707917226e-06,
+      "loss": 0.0673,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9562513879635799,
+      "grad_norm": 0.41134986778955357,
+      "learning_rate": 9.422419575904584e-06,
+      "loss": 0.0449,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9566955363091273,
+      "grad_norm": 0.40411844058020524,
+      "learning_rate": 9.421514779127156e-06,
+      "loss": 0.0457,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9571396846546747,
+      "grad_norm": 0.5724570461924485,
+      "learning_rate": 9.420609317720948e-06,
+      "loss": 0.064,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9575838330002221,
+      "grad_norm": 0.5481138081401853,
+      "learning_rate": 9.419703191822067e-06,
+      "loss": 0.0512,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9580279813457695,
+      "grad_norm": 0.5977006719684248,
+      "learning_rate": 9.418796401566719e-06,
+      "loss": 0.0452,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9584721296913169,
+      "grad_norm": 0.5349085172455206,
+      "learning_rate": 9.417888947091208e-06,
+      "loss": 0.065,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9589162780368643,
+      "grad_norm": 0.6787928669434861,
+      "learning_rate": 9.416980828531944e-06,
+      "loss": 0.0677,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9593604263824117,
+      "grad_norm": 0.709282953075048,
+      "learning_rate": 9.416072046025429e-06,
+      "loss": 0.0729,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9598045747279591,
+      "grad_norm": 0.6790718166354139,
+      "learning_rate": 9.415162599708268e-06,
+      "loss": 0.0699,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9602487230735065,
+      "grad_norm": 0.545039186094067,
+      "learning_rate": 9.414252489717168e-06,
+      "loss": 0.0594,
+      "step": 2162
+    },
+    {
+      "epoch": 0.960692871419054,
+      "grad_norm": 0.7438799552703795,
+      "learning_rate": 9.413341716188934e-06,
+      "loss": 0.0653,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9611370197646014,
+      "grad_norm": 0.5419774566213018,
+      "learning_rate": 9.412430279260473e-06,
+      "loss": 0.0452,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9615811681101488,
+      "grad_norm": 0.5417611334038369,
+      "learning_rate": 9.411518179068785e-06,
+      "loss": 0.0695,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9620253164556962,
+      "grad_norm": 0.556851360606536,
+      "learning_rate": 9.410605415750977e-06,
+      "loss": 0.0612,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9624694648012436,
+      "grad_norm": 0.8789606927293152,
+      "learning_rate": 9.40969198944425e-06,
+      "loss": 0.0604,
+      "step": 2167
+    },
+    {
+      "epoch": 0.962913613146791,
+      "grad_norm": 0.5933413334023052,
+      "learning_rate": 9.40877790028591e-06,
+      "loss": 0.0623,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9633577614923384,
+      "grad_norm": 0.5946398594923591,
+      "learning_rate": 9.407863148413361e-06,
+      "loss": 0.0419,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9638019098378858,
+      "grad_norm": 0.48221233179240436,
+      "learning_rate": 9.406947733964103e-06,
+      "loss": 0.0546,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9642460581834332,
+      "grad_norm": 0.6525276948819981,
+      "learning_rate": 9.40603165707574e-06,
+      "loss": 0.0526,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9646902065289806,
+      "grad_norm": 0.7205786693472772,
+      "learning_rate": 9.405114917885973e-06,
+      "loss": 0.0684,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9651343548745281,
+      "grad_norm": 0.5044988796095775,
+      "learning_rate": 9.404197516532605e-06,
+      "loss": 0.0478,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9655785032200755,
+      "grad_norm": 0.44346462918671525,
+      "learning_rate": 9.403279453153536e-06,
+      "loss": 0.0446,
+      "step": 2174
+    },
+    {
+      "epoch": 0.966022651565623,
+      "grad_norm": 0.5173823021640699,
+      "learning_rate": 9.402360727886766e-06,
+      "loss": 0.0623,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9664667999111703,
+      "grad_norm": 0.5687155650294274,
+      "learning_rate": 9.401441340870397e-06,
+      "loss": 0.0611,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9669109482567178,
+      "grad_norm": 0.6467987726672418,
+      "learning_rate": 9.400521292242626e-06,
+      "loss": 0.0529,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9673550966022652,
+      "grad_norm": 0.5130913991837206,
+      "learning_rate": 9.399600582141752e-06,
+      "loss": 0.0599,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9677992449478126,
+      "grad_norm": 0.6670998717556929,
+      "learning_rate": 9.398679210706176e-06,
+      "loss": 0.0682,
+      "step": 2179
+    },
+    {
+      "epoch": 0.96824339329336,
+      "grad_norm": 0.7694218770583813,
+      "learning_rate": 9.397757178074392e-06,
+      "loss": 0.0686,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9686875416389074,
+      "grad_norm": 0.7755428807155127,
+      "learning_rate": 9.396834484385e-06,
+      "loss": 0.0574,
+      "step": 2181
+    },
+    {
+      "epoch": 0.9691316899844548,
+      "grad_norm": 0.44478529742795087,
+      "learning_rate": 9.395911129776699e-06,
+      "loss": 0.0429,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9695758383300023,
+      "grad_norm": 0.7140201406983687,
+      "learning_rate": 9.394987114388278e-06,
+      "loss": 0.0711,
+      "step": 2183
+    },
+    {
+      "epoch": 0.9700199866755497,
+      "grad_norm": 0.8530149060800487,
+      "learning_rate": 9.394062438358637e-06,
+      "loss": 0.0529,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9704641350210971,
+      "grad_norm": 0.6524072835443322,
+      "learning_rate": 9.39313710182677e-06,
+      "loss": 0.0624,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9709082833666445,
+      "grad_norm": 0.4290712533966452,
+      "learning_rate": 9.39221110493177e-06,
+      "loss": 0.041,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9713524317121919,
+      "grad_norm": 0.5324670553025179,
+      "learning_rate": 9.39128444781283e-06,
+      "loss": 0.0477,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9717965800577393,
+      "grad_norm": 0.5901470773440277,
+      "learning_rate": 9.390357130609243e-06,
+      "loss": 0.0568,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9722407284032867,
+      "grad_norm": 0.7720115526671916,
+      "learning_rate": 9.3894291534604e-06,
+      "loss": 0.0559,
+      "step": 2189
+    },
+    {
+      "epoch": 0.9726848767488341,
+      "grad_norm": 0.6618879002055548,
+      "learning_rate": 9.38850051650579e-06,
+      "loss": 0.0789,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9731290250943815,
+      "grad_norm": 0.6242983609779144,
+      "learning_rate": 9.387571219885008e-06,
+      "loss": 0.0709,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9735731734399289,
+      "grad_norm": 0.5488182994392276,
+      "learning_rate": 9.386641263737736e-06,
+      "loss": 0.0483,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9740173217854764,
+      "grad_norm": 0.5678755050400883,
+      "learning_rate": 9.38571064820377e-06,
+      "loss": 0.0639,
+      "step": 2193
+    },
+    {
+      "epoch": 0.9744614701310238,
+      "grad_norm": 0.7972177120971958,
+      "learning_rate": 9.384779373422992e-06,
+      "loss": 0.0688,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9749056184765712,
+      "grad_norm": 0.4181144019615441,
+      "learning_rate": 9.38384743953539e-06,
+      "loss": 0.0434,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9753497668221186,
+      "grad_norm": 1.0045462377103442,
+      "learning_rate": 9.382914846681049e-06,
+      "loss": 0.0626,
+      "step": 2196
+    },
+    {
+      "epoch": 0.975793915167666,
+      "grad_norm": 0.5663797823359984,
+      "learning_rate": 9.381981595000153e-06,
+      "loss": 0.0536,
+      "step": 2197
+    },
+    {
+      "epoch": 0.9762380635132134,
+      "grad_norm": 0.5230033029520286,
+      "learning_rate": 9.381047684632986e-06,
+      "loss": 0.0431,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9766822118587608,
+      "grad_norm": 0.5941306490458375,
+      "learning_rate": 9.380113115719933e-06,
+      "loss": 0.0514,
+      "step": 2199
+    },
+    {
+      "epoch": 0.9771263602043082,
+      "grad_norm": 0.7074167144721983,
+      "learning_rate": 9.379177888401473e-06,
+      "loss": 0.0516,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9775705085498556,
+      "grad_norm": 0.8936942623677783,
+      "learning_rate": 9.378242002818186e-06,
+      "loss": 0.0634,
+      "step": 2201
+    },
+    {
+      "epoch": 0.978014656895403,
+      "grad_norm": 0.8015956656924698,
+      "learning_rate": 9.377305459110754e-06,
+      "loss": 0.0701,
+      "step": 2202
+    },
+    {
+      "epoch": 0.9784588052409505,
+      "grad_norm": 0.5888281751564788,
+      "learning_rate": 9.376368257419955e-06,
+      "loss": 0.0735,
+      "step": 2203
+    },
+    {
+      "epoch": 0.9789029535864979,
+      "grad_norm": 0.6871960760484549,
+      "learning_rate": 9.375430397886661e-06,
+      "loss": 0.0528,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9793471019320453,
+      "grad_norm": 0.5915896411171675,
+      "learning_rate": 9.374491880651856e-06,
+      "loss": 0.0577,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9797912502775927,
+      "grad_norm": 0.6268683607318559,
+      "learning_rate": 9.373552705856612e-06,
+      "loss": 0.0511,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9802353986231401,
+      "grad_norm": 0.5959527967947893,
+      "learning_rate": 9.372612873642101e-06,
+      "loss": 0.0577,
+      "step": 2207
+    },
+    {
+      "epoch": 0.9806795469686875,
+      "grad_norm": 0.6376605781281672,
+      "learning_rate": 9.3716723841496e-06,
+      "loss": 0.056,
+      "step": 2208
+    },
+    {
+      "epoch": 0.9811236953142349,
+      "grad_norm": 0.6655725615122635,
+      "learning_rate": 9.370731237520476e-06,
+      "loss": 0.0495,
+      "step": 2209
+    },
+    {
+      "epoch": 0.9815678436597823,
+      "grad_norm": 0.6178758449973094,
+      "learning_rate": 9.369789433896201e-06,
+      "loss": 0.0785,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9820119920053297,
+      "grad_norm": 0.6203954478751535,
+      "learning_rate": 9.368846973418347e-06,
+      "loss": 0.0541,
+      "step": 2211
+    },
+    {
+      "epoch": 0.9824561403508771,
+      "grad_norm": 0.5844524216400101,
+      "learning_rate": 9.367903856228575e-06,
+      "loss": 0.0484,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9829002886964247,
+      "grad_norm": 1.3376432627550732,
+      "learning_rate": 9.366960082468658e-06,
+      "loss": 0.0523,
+      "step": 2213
+    },
+    {
+      "epoch": 0.983344437041972,
+      "grad_norm": 0.7650415999284177,
+      "learning_rate": 9.36601565228046e-06,
+      "loss": 0.0758,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9837885853875195,
+      "grad_norm": 0.6031770466123566,
+      "learning_rate": 9.365070565805941e-06,
+      "loss": 0.0552,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9842327337330669,
+      "grad_norm": 0.5317856643301976,
+      "learning_rate": 9.364124823187169e-06,
+      "loss": 0.0495,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9846768820786143,
+      "grad_norm": 0.697207175537678,
+      "learning_rate": 9.363178424566302e-06,
+      "loss": 0.0575,
+      "step": 2217
+    },
+    {
+      "epoch": 0.9851210304241617,
+      "grad_norm": 0.6901112062939687,
+      "learning_rate": 9.3622313700856e-06,
+      "loss": 0.0558,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9855651787697091,
+      "grad_norm": 0.7725648574354754,
+      "learning_rate": 9.361283659887421e-06,
+      "loss": 0.0526,
+      "step": 2219
+    },
+    {
+      "epoch": 0.9860093271152565,
+      "grad_norm": 0.4919891771117441,
+      "learning_rate": 9.360335294114222e-06,
+      "loss": 0.0517,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9864534754608039,
+      "grad_norm": 0.604042614778997,
+      "learning_rate": 9.359386272908561e-06,
+      "loss": 0.0654,
+      "step": 2221
+    },
+    {
+      "epoch": 0.9868976238063514,
+      "grad_norm": 0.533434108631547,
+      "learning_rate": 9.35843659641309e-06,
+      "loss": 0.0587,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9873417721518988,
+      "grad_norm": 0.627831211955756,
+      "learning_rate": 9.35748626477056e-06,
+      "loss": 0.0577,
+      "step": 2223
+    },
+    {
+      "epoch": 0.9877859204974462,
+      "grad_norm": 0.48609069405336236,
+      "learning_rate": 9.356535278123826e-06,
+      "loss": 0.0625,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9882300688429936,
+      "grad_norm": 0.5441221898111536,
+      "learning_rate": 9.355583636615832e-06,
+      "loss": 0.0671,
+      "step": 2225
+    },
+    {
+      "epoch": 0.988674217188541,
+      "grad_norm": 0.6653694627323021,
+      "learning_rate": 9.354631340389633e-06,
+      "loss": 0.0678,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9891183655340884,
+      "grad_norm": 0.49687550533635927,
+      "learning_rate": 9.353678389588367e-06,
+      "loss": 0.048,
+      "step": 2227
+    },
+    {
+      "epoch": 0.9895625138796358,
+      "grad_norm": 0.5456376905073873,
+      "learning_rate": 9.352724784355286e-06,
+      "loss": 0.0763,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9900066622251832,
+      "grad_norm": 0.5234550844694897,
+      "learning_rate": 9.35177052483373e-06,
+      "loss": 0.0573,
+      "step": 2229
+    },
+    {
+      "epoch": 0.9904508105707306,
+      "grad_norm": 0.3992910106511535,
+      "learning_rate": 9.35081561116714e-06,
+      "loss": 0.0422,
+      "step": 2230
+    },
+    {
+      "epoch": 0.990894958916278,
+      "grad_norm": 0.4091987827719479,
+      "learning_rate": 9.349860043499056e-06,
+      "loss": 0.0474,
+      "step": 2231
+    },
+    {
+      "epoch": 0.9913391072618255,
+      "grad_norm": 0.5342639072522524,
+      "learning_rate": 9.348903821973114e-06,
+      "loss": 0.0663,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9917832556073729,
+      "grad_norm": 0.4467060129332395,
+      "learning_rate": 9.347946946733055e-06,
+      "loss": 0.0443,
+      "step": 2233
+    },
+    {
+      "epoch": 0.9922274039529203,
+      "grad_norm": 0.4523315857358909,
+      "learning_rate": 9.346989417922712e-06,
+      "loss": 0.0415,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9926715522984677,
+      "grad_norm": 0.525064951465678,
+      "learning_rate": 9.346031235686014e-06,
+      "loss": 0.0663,
+      "step": 2235
+    },
+    {
+      "epoch": 0.9931157006440151,
+      "grad_norm": 0.5615837318533068,
+      "learning_rate": 9.345072400166999e-06,
+      "loss": 0.0558,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9935598489895625,
+      "grad_norm": 0.5550096636418779,
+      "learning_rate": 9.34411291150979e-06,
+      "loss": 0.0699,
+      "step": 2237
+    },
+    {
+      "epoch": 0.9940039973351099,
+      "grad_norm": 0.9242044346261947,
+      "learning_rate": 9.343152769858616e-06,
+      "loss": 0.0623,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9944481456806573,
+      "grad_norm": 0.6358712068894604,
+      "learning_rate": 9.342191975357806e-06,
+      "loss": 0.0675,
+      "step": 2239
+    },
+    {
+      "epoch": 0.9948922940262047,
+      "grad_norm": 0.5189758212655892,
+      "learning_rate": 9.34123052815178e-06,
+      "loss": 0.0446,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9953364423717521,
+      "grad_norm": 0.4493321450297416,
+      "learning_rate": 9.340268428385062e-06,
+      "loss": 0.0416,
+      "step": 2241
+    },
+    {
+      "epoch": 0.9957805907172996,
+      "grad_norm": 0.4052216095743849,
+      "learning_rate": 9.339305676202268e-06,
+      "loss": 0.0524,
+      "step": 2242
+    },
+    {
+      "epoch": 0.996224739062847,
+      "grad_norm": 0.4257914645053657,
+      "learning_rate": 9.338342271748122e-06,
+      "loss": 0.0439,
+      "step": 2243
+    },
+    {
+      "epoch": 0.9966688874083944,
+      "grad_norm": 0.5443368549896095,
+      "learning_rate": 9.337378215167436e-06,
+      "loss": 0.0494,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9971130357539418,
+      "grad_norm": 0.48744281644400594,
+      "learning_rate": 9.336413506605123e-06,
+      "loss": 0.0692,
+      "step": 2245
+    },
+    {
+      "epoch": 0.9975571840994892,
+      "grad_norm": 0.730043582195284,
+      "learning_rate": 9.335448146206201e-06,
+      "loss": 0.0613,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9980013324450366,
+      "grad_norm": 0.6666928317513323,
+      "learning_rate": 9.334482134115774e-06,
+      "loss": 0.0626,
+      "step": 2247
+    },
+    {
+      "epoch": 0.998445480790584,
+      "grad_norm": 1.1582138795192063,
+      "learning_rate": 9.333515470479052e-06,
+      "loss": 0.0706,
+      "step": 2248
+    },
+    {
+      "epoch": 0.9988896291361314,
+      "grad_norm": 0.41845565011499725,
+      "learning_rate": 9.332548155441341e-06,
+      "loss": 0.0428,
+      "step": 2249
+    },
+    {
+      "epoch": 0.9993337774816788,
+      "grad_norm": 0.49853102857677845,
+      "learning_rate": 9.331580189148047e-06,
+      "loss": 0.0498,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9997779258272262,
+      "grad_norm": 0.469460122533607,
+      "learning_rate": 9.330611571744668e-06,
+      "loss": 0.0584,
+      "step": 2251
+    },
+    {
+      "epoch": 0.9997779258272262,
+      "eval_loss": 0.06305240094661713,
+      "eval_runtime": 420.7122,
+      "eval_samples_per_second": 36.048,
+      "eval_steps_per_second": 1.127,
+      "step": 2251
+    },
+    {
+      "epoch": 1.0002220741727736,
+      "grad_norm": 0.8383200378207877,
+      "learning_rate": 9.329642303376806e-06,
+      "loss": 0.0866,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0006662225183212,
+      "grad_norm": 0.4162547630354714,
+      "learning_rate": 9.328672384190158e-06,
+      "loss": 0.044,
+      "step": 2253
+    },
+    {
+      "epoch": 1.0011103708638684,
+      "grad_norm": 0.5811219466400265,
+      "learning_rate": 9.327701814330521e-06,
+      "loss": 0.0489,
+      "step": 2254
+    },
+    {
+      "epoch": 1.001554519209416,
+      "grad_norm": 0.4943074740960733,
+      "learning_rate": 9.326730593943784e-06,
+      "loss": 0.0532,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0019986675549633,
+      "grad_norm": 0.5854940362247197,
+      "learning_rate": 9.325758723175942e-06,
+      "loss": 0.0447,
+      "step": 2256
+    },
+    {
+      "epoch": 1.0024428159005108,
+      "grad_norm": 0.6734583854523404,
+      "learning_rate": 9.324786202173082e-06,
+      "loss": 0.0565,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0028869642460583,
+      "grad_norm": 0.626997176192022,
+      "learning_rate": 9.32381303108139e-06,
+      "loss": 0.0554,
+      "step": 2258
+    },
+    {
+      "epoch": 1.0033311125916056,
+      "grad_norm": 0.7899011567544635,
+      "learning_rate": 9.322839210047152e-06,
+      "loss": 0.0541,
+      "step": 2259
+    },
+    {
+      "epoch": 1.003775260937153,
+      "grad_norm": 0.44777931003199795,
+      "learning_rate": 9.321864739216747e-06,
+      "loss": 0.0475,
+      "step": 2260
+    },
+    {
+      "epoch": 1.0042194092827004,
+      "grad_norm": 0.7176524264984108,
+      "learning_rate": 9.320889618736657e-06,
+      "loss": 0.0657,
+      "step": 2261
+    },
+    {
+      "epoch": 1.0046635576282479,
+      "grad_norm": 0.6297899790013282,
+      "learning_rate": 9.319913848753457e-06,
+      "loss": 0.0548,
+      "step": 2262
+    },
+    {
+      "epoch": 1.0051077059737952,
+      "grad_norm": 0.6829050036731806,
+      "learning_rate": 9.318937429413823e-06,
+      "loss": 0.0677,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0055518543193427,
+      "grad_norm": 0.6099066772979853,
+      "learning_rate": 9.31796036086453e-06,
+      "loss": 0.0674,
+      "step": 2264
+    },
+    {
+      "epoch": 1.00599600266489,
+      "grad_norm": 0.5220221509404842,
+      "learning_rate": 9.316982643252444e-06,
+      "loss": 0.0489,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0064401510104375,
+      "grad_norm": 0.6319552346310404,
+      "learning_rate": 9.316004276724533e-06,
+      "loss": 0.0566,
+      "step": 2266
+    },
+    {
+      "epoch": 1.006884299355985,
+      "grad_norm": 0.75308545956684,
+      "learning_rate": 9.315025261427864e-06,
+      "loss": 0.0611,
+      "step": 2267
+    },
+    {
+      "epoch": 1.0073284477015323,
+      "grad_norm": 0.5854190329106237,
+      "learning_rate": 9.314045597509598e-06,
+      "loss": 0.0589,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0077725960470798,
+      "grad_norm": 0.4592027406322817,
+      "learning_rate": 9.313065285116997e-06,
+      "loss": 0.0363,
+      "step": 2269
+    },
+    {
+      "epoch": 1.008216744392627,
+      "grad_norm": 0.7355227136248891,
+      "learning_rate": 9.312084324397416e-06,
+      "loss": 0.0666,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0086608927381746,
+      "grad_norm": 0.6811364248406817,
+      "learning_rate": 9.311102715498312e-06,
+      "loss": 0.0664,
+      "step": 2271
+    },
+    {
+      "epoch": 1.009105041083722,
+      "grad_norm": 0.4859582181876178,
+      "learning_rate": 9.310120458567238e-06,
+      "loss": 0.0455,
+      "step": 2272
+    },
+    {
+      "epoch": 1.0095491894292694,
+      "grad_norm": 0.6126182761710498,
+      "learning_rate": 9.309137553751843e-06,
+      "loss": 0.0568,
+      "step": 2273
+    },
+    {
+      "epoch": 1.0099933377748167,
+      "grad_norm": 0.6903438432741064,
+      "learning_rate": 9.308154001199874e-06,
+      "loss": 0.0546,
+      "step": 2274
+    },
+    {
+      "epoch": 1.0104374861203642,
+      "grad_norm": 0.5254671203092518,
+      "learning_rate": 9.307169801059175e-06,
+      "loss": 0.0518,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0108816344659115,
+      "grad_norm": 0.502322384644144,
+      "learning_rate": 9.30618495347769e-06,
+      "loss": 0.0528,
+      "step": 2276
+    },
+    {
+      "epoch": 1.011325782811459,
+      "grad_norm": 0.4570165148396037,
+      "learning_rate": 9.305199458603456e-06,
+      "loss": 0.0423,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0117699311570065,
+      "grad_norm": 0.6475436488464652,
+      "learning_rate": 9.304213316584612e-06,
+      "loss": 0.0539,
+      "step": 2278
+    },
+    {
+      "epoch": 1.0122140795025538,
+      "grad_norm": 0.5830527566752789,
+      "learning_rate": 9.30322652756939e-06,
+      "loss": 0.0533,
+      "step": 2279
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.5218861392382439,
+      "learning_rate": 9.302239091706121e-06,
+      "loss": 0.0478,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0131023761936486,
+      "grad_norm": 0.7938697599205536,
+      "learning_rate": 9.301251009143236e-06,
+      "loss": 0.0913,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0135465245391961,
+      "grad_norm": 0.5550504645736817,
+      "learning_rate": 9.300262280029257e-06,
+      "loss": 0.0644,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0139906728847434,
+      "grad_norm": 0.594954378218846,
+      "learning_rate": 9.29927290451281e-06,
+      "loss": 0.0541,
+      "step": 2283
+    },
+    {
+      "epoch": 1.014434821230291,
+      "grad_norm": 0.9298600097346652,
+      "learning_rate": 9.298282882742612e-06,
+      "loss": 0.0518,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0148789695758382,
+      "grad_norm": 0.46811932472048295,
+      "learning_rate": 9.297292214867484e-06,
+      "loss": 0.0455,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0153231179213857,
+      "grad_norm": 0.5871427629727887,
+      "learning_rate": 9.296300901036337e-06,
+      "loss": 0.0572,
+      "step": 2286
+    },
+    {
+      "epoch": 1.0157672662669333,
+      "grad_norm": 0.6945899846686857,
+      "learning_rate": 9.295308941398183e-06,
+      "loss": 0.061,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0162114146124805,
+      "grad_norm": 0.640279460698154,
+      "learning_rate": 9.294316336102132e-06,
+      "loss": 0.0599,
+      "step": 2288
+    },
+    {
+      "epoch": 1.016655562958028,
+      "grad_norm": 0.31136909660088496,
+      "learning_rate": 9.293323085297386e-06,
+      "loss": 0.026,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0170997113035753,
+      "grad_norm": 0.4114152449104369,
+      "learning_rate": 9.29232918913325e-06,
+      "loss": 0.0379,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0175438596491229,
+      "grad_norm": 0.41732022602175256,
+      "learning_rate": 9.291334647759122e-06,
+      "loss": 0.0422,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0179880079946702,
+      "grad_norm": 0.762539369842014,
+      "learning_rate": 9.2903394613245e-06,
+      "loss": 0.0597,
+      "step": 2292
+    },
+    {
+      "epoch": 1.0184321563402177,
+      "grad_norm": 0.9415453759336739,
+      "learning_rate": 9.289343629978978e-06,
+      "loss": 0.0739,
+      "step": 2293
+    },
+    {
+      "epoch": 1.018876304685765,
+      "grad_norm": 0.5123579743591409,
+      "learning_rate": 9.288347153872245e-06,
+      "loss": 0.0571,
+      "step": 2294
+    },
+    {
+      "epoch": 1.0193204530313125,
+      "grad_norm": 0.458862217726016,
+      "learning_rate": 9.287350033154088e-06,
+      "loss": 0.0468,
+      "step": 2295
+    },
+    {
+      "epoch": 1.01976460137686,
+      "grad_norm": 1.3873099747217557,
+      "learning_rate": 9.28635226797439e-06,
+      "loss": 0.0745,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0202087497224073,
+      "grad_norm": 0.5997981500021724,
+      "learning_rate": 9.285353858483138e-06,
+      "loss": 0.0506,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0206528980679548,
+      "grad_norm": 0.6396330710132598,
+      "learning_rate": 9.284354804830403e-06,
+      "loss": 0.0518,
+      "step": 2298
+    },
+    {
+      "epoch": 1.021097046413502,
+      "grad_norm": 0.8119329100233045,
+      "learning_rate": 9.283355107166361e-06,
+      "loss": 0.088,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0215411947590496,
+      "grad_norm": 0.6001947150070646,
+      "learning_rate": 9.282354765641286e-06,
+      "loss": 0.0503,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0219853431045969,
+      "grad_norm": 0.5257172181736586,
+      "learning_rate": 9.281353780405546e-06,
+      "loss": 0.0632,
+      "step": 2301
+    },
+    {
+      "epoch": 1.0224294914501444,
+      "grad_norm": 0.4904406999367498,
+      "learning_rate": 9.280352151609604e-06,
+      "loss": 0.0412,
+      "step": 2302
+    },
+    {
+      "epoch": 1.0228736397956917,
+      "grad_norm": 0.6023825108536248,
+      "learning_rate": 9.279349879404024e-06,
+      "loss": 0.0501,
+      "step": 2303
+    },
+    {
+      "epoch": 1.0233177881412392,
+      "grad_norm": 0.5713607373831051,
+      "learning_rate": 9.278346963939464e-06,
+      "loss": 0.0648,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0237619364867865,
+      "grad_norm": 0.5117889653709818,
+      "learning_rate": 9.27734340536668e-06,
+      "loss": 0.0595,
+      "step": 2305
+    },
+    {
+      "epoch": 1.024206084832334,
+      "grad_norm": 0.5957895920503053,
+      "learning_rate": 9.27633920383652e-06,
+      "loss": 0.0588,
+      "step": 2306
+    },
+    {
+      "epoch": 1.0246502331778815,
+      "grad_norm": 0.6119278145978992,
+      "learning_rate": 9.275334359499936e-06,
+      "loss": 0.0615,
+      "step": 2307
+    },
+    {
+      "epoch": 1.0250943815234288,
+      "grad_norm": 0.5747304102651929,
+      "learning_rate": 9.274328872507973e-06,
+      "loss": 0.0609,
+      "step": 2308
+    },
+    {
+      "epoch": 1.0255385298689763,
+      "grad_norm": 0.5890004950126706,
+      "learning_rate": 9.273322743011775e-06,
+      "loss": 0.054,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0259826782145236,
+      "grad_norm": 0.46856707487309196,
+      "learning_rate": 9.272315971162573e-06,
+      "loss": 0.0438,
+      "step": 2310
+    },
+    {
+      "epoch": 1.0264268265600711,
+      "grad_norm": 0.5705380710124494,
+      "learning_rate": 9.27130855711171e-06,
+      "loss": 0.0477,
+      "step": 2311
+    },
+    {
+      "epoch": 1.0268709749056184,
+      "grad_norm": 0.7064165326935903,
+      "learning_rate": 9.270300501010612e-06,
+      "loss": 0.0568,
+      "step": 2312
+    },
+    {
+      "epoch": 1.027315123251166,
+      "grad_norm": 0.5416989623481593,
+      "learning_rate": 9.26929180301081e-06,
+      "loss": 0.0682,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0277592715967132,
+      "grad_norm": 0.49073246688028643,
+      "learning_rate": 9.268282463263928e-06,
+      "loss": 0.0629,
+      "step": 2314
+    },
+    {
+      "epoch": 1.0282034199422607,
+      "grad_norm": 0.47098174668190423,
+      "learning_rate": 9.267272481921686e-06,
+      "loss": 0.0492,
+      "step": 2315
+    },
+    {
+      "epoch": 1.0286475682878082,
+      "grad_norm": 0.3813955943581627,
+      "learning_rate": 9.266261859135901e-06,
+      "loss": 0.0308,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0290917166333555,
+      "grad_norm": 0.6582567910792767,
+      "learning_rate": 9.265250595058486e-06,
+      "loss": 0.0573,
+      "step": 2317
+    },
+    {
+      "epoch": 1.029535864978903,
+      "grad_norm": 0.5953836977171739,
+      "learning_rate": 9.264238689841456e-06,
+      "loss": 0.0631,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0299800133244503,
+      "grad_norm": 0.6292012725822909,
+      "learning_rate": 9.263226143636912e-06,
+      "loss": 0.0691,
+      "step": 2319
+    },
+    {
+      "epoch": 1.0304241616699978,
+      "grad_norm": 0.47225699541890964,
+      "learning_rate": 9.262212956597059e-06,
+      "loss": 0.0563,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0308683100155451,
+      "grad_norm": 0.48390119987242786,
+      "learning_rate": 9.261199128874197e-06,
+      "loss": 0.054,
+      "step": 2321
+    },
+    {
+      "epoch": 1.0313124583610926,
+      "grad_norm": 0.5985733647635482,
+      "learning_rate": 9.26018466062072e-06,
+      "loss": 0.0582,
+      "step": 2322
+    },
+    {
+      "epoch": 1.03175660670664,
+      "grad_norm": 0.5869595699898266,
+      "learning_rate": 9.259169551989121e-06,
+      "loss": 0.0562,
+      "step": 2323
+    },
+    {
+      "epoch": 1.0322007550521874,
+      "grad_norm": 0.5529638354373985,
+      "learning_rate": 9.258153803131989e-06,
+      "loss": 0.055,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0326449033977347,
+      "grad_norm": 0.39012786867402377,
+      "learning_rate": 9.257137414202006e-06,
+      "loss": 0.0389,
+      "step": 2325
+    },
+    {
+      "epoch": 1.0330890517432822,
+      "grad_norm": 0.5320314663484899,
+      "learning_rate": 9.256120385351953e-06,
+      "loss": 0.0593,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0335332000888298,
+      "grad_norm": 0.47550720714560507,
+      "learning_rate": 9.255102716734709e-06,
+      "loss": 0.0383,
+      "step": 2327
+    },
+    {
+      "epoch": 1.033977348434377,
+      "grad_norm": 0.6123732206520373,
+      "learning_rate": 9.254084408503243e-06,
+      "loss": 0.0733,
+      "step": 2328
+    },
+    {
+      "epoch": 1.0344214967799246,
+      "grad_norm": 0.6340880346930579,
+      "learning_rate": 9.253065460810627e-06,
+      "loss": 0.0703,
+      "step": 2329
+    },
+    {
+      "epoch": 1.0348656451254719,
+      "grad_norm": 0.4558646564237513,
+      "learning_rate": 9.252045873810026e-06,
+      "loss": 0.0389,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0353097934710194,
+      "grad_norm": 0.573613972583976,
+      "learning_rate": 9.251025647654698e-06,
+      "loss": 0.0578,
+      "step": 2331
+    },
+    {
+      "epoch": 1.0357539418165667,
+      "grad_norm": 0.45869687127989117,
+      "learning_rate": 9.250004782498006e-06,
+      "loss": 0.0486,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0361980901621142,
+      "grad_norm": 0.700597406034261,
+      "learning_rate": 9.248983278493399e-06,
+      "loss": 0.0486,
+      "step": 2333
+    },
+    {
+      "epoch": 1.0366422385076615,
+      "grad_norm": 0.5940754007816903,
+      "learning_rate": 9.247961135794428e-06,
+      "loss": 0.0487,
+      "step": 2334
+    },
+    {
+      "epoch": 1.037086386853209,
+      "grad_norm": 0.7212210545701735,
+      "learning_rate": 9.246938354554737e-06,
+      "loss": 0.051,
+      "step": 2335
+    },
+    {
+      "epoch": 1.0375305351987565,
+      "grad_norm": 0.41449723705228086,
+      "learning_rate": 9.245914934928068e-06,
+      "loss": 0.0441,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0379746835443038,
+      "grad_norm": 0.9446091827668867,
+      "learning_rate": 9.24489087706826e-06,
+      "loss": 0.0941,
+      "step": 2337
+    },
+    {
+      "epoch": 1.0384188318898513,
+      "grad_norm": 0.43472127189546134,
+      "learning_rate": 9.243866181129246e-06,
+      "loss": 0.0444,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0388629802353986,
+      "grad_norm": 0.8234990673910981,
+      "learning_rate": 9.242840847265053e-06,
+      "loss": 0.0449,
+      "step": 2339
+    },
+    {
+      "epoch": 1.039307128580946,
+      "grad_norm": 0.5137093258500119,
+      "learning_rate": 9.241814875629806e-06,
+      "loss": 0.047,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0397512769264934,
+      "grad_norm": 0.46164431011916773,
+      "learning_rate": 9.24078826637773e-06,
+      "loss": 0.0567,
+      "step": 2341
+    },
+    {
+      "epoch": 1.040195425272041,
+      "grad_norm": 0.5917045565013841,
+      "learning_rate": 9.239761019663139e-06,
+      "loss": 0.0529,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0406395736175882,
+      "grad_norm": 0.49372320773080497,
+      "learning_rate": 9.238733135640445e-06,
+      "loss": 0.0503,
+      "step": 2343
+    },
+    {
+      "epoch": 1.0410837219631357,
+      "grad_norm": 0.47687223129773526,
+      "learning_rate": 9.237704614464157e-06,
+      "loss": 0.0538,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0415278703086832,
+      "grad_norm": 0.5267358173749689,
+      "learning_rate": 9.236675456288879e-06,
+      "loss": 0.0533,
+      "step": 2345
+    },
+    {
+      "epoch": 1.0419720186542305,
+      "grad_norm": 0.5562866454702318,
+      "learning_rate": 9.235645661269313e-06,
+      "loss": 0.0617,
+      "step": 2346
+    },
+    {
+      "epoch": 1.042416166999778,
+      "grad_norm": 0.48761824525349823,
+      "learning_rate": 9.234615229560251e-06,
+      "loss": 0.0646,
+      "step": 2347
+    },
+    {
+      "epoch": 1.0428603153453253,
+      "grad_norm": 0.5413755626896984,
+      "learning_rate": 9.233584161316588e-06,
+      "loss": 0.0485,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0433044636908728,
+      "grad_norm": 0.48592846139974233,
+      "learning_rate": 9.232552456693308e-06,
+      "loss": 0.0466,
+      "step": 2349
+    },
+    {
+      "epoch": 1.04374861203642,
+      "grad_norm": 0.3666523977412306,
+      "learning_rate": 9.231520115845495e-06,
+      "loss": 0.0415,
+      "step": 2350
+    },
+    {
+      "epoch": 1.0441927603819676,
+      "grad_norm": 0.5041685160042538,
+      "learning_rate": 9.23048713892833e-06,
+      "loss": 0.0461,
+      "step": 2351
+    },
+    {
+      "epoch": 1.044636908727515,
+      "grad_norm": 0.4557622908913518,
+      "learning_rate": 9.229453526097085e-06,
+      "loss": 0.0444,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0450810570730624,
+      "grad_norm": 0.7840502309472853,
+      "learning_rate": 9.228419277507126e-06,
+      "loss": 0.0628,
+      "step": 2353
+    },
+    {
+      "epoch": 1.0455252054186097,
+      "grad_norm": 0.5291018953826038,
+      "learning_rate": 9.227384393313924e-06,
+      "loss": 0.0496,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0459693537641572,
+      "grad_norm": 0.49003402431309595,
+      "learning_rate": 9.226348873673036e-06,
+      "loss": 0.0549,
+      "step": 2355
+    },
+    {
+      "epoch": 1.0464135021097047,
+      "grad_norm": 0.5208623353433574,
+      "learning_rate": 9.22531271874012e-06,
+      "loss": 0.0473,
+      "step": 2356
+    },
+    {
+      "epoch": 1.046857650455252,
+      "grad_norm": 0.6890800835128176,
+      "learning_rate": 9.224275928670925e-06,
+      "loss": 0.0554,
+      "step": 2357
+    },
+    {
+      "epoch": 1.0473017988007995,
+      "grad_norm": 0.6619670076578327,
+      "learning_rate": 9.223238503621302e-06,
+      "loss": 0.0577,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0477459471463468,
+      "grad_norm": 0.5128511282981787,
+      "learning_rate": 9.22220044374719e-06,
+      "loss": 0.0406,
+      "step": 2359
+    },
+    {
+      "epoch": 1.0481900954918943,
+      "grad_norm": 0.521752374455369,
+      "learning_rate": 9.221161749204629e-06,
+      "loss": 0.05,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0486342438374416,
+      "grad_norm": 0.6524832106347492,
+      "learning_rate": 9.220122420149753e-06,
+      "loss": 0.0548,
+      "step": 2361
+    },
+    {
+      "epoch": 1.0490783921829892,
+      "grad_norm": 0.49101446033388607,
+      "learning_rate": 9.219082456738788e-06,
+      "loss": 0.0392,
+      "step": 2362
+    },
+    {
+      "epoch": 1.0495225405285364,
+      "grad_norm": 0.40291059778432986,
+      "learning_rate": 9.218041859128062e-06,
+      "loss": 0.0403,
+      "step": 2363
+    },
+    {
+      "epoch": 1.049966688874084,
+      "grad_norm": 0.52475244536887,
+      "learning_rate": 9.217000627473993e-06,
+      "loss": 0.0518,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0504108372196312,
+      "grad_norm": 0.6932318744101922,
+      "learning_rate": 9.215958761933093e-06,
+      "loss": 0.0586,
+      "step": 2365
+    },
+    {
+      "epoch": 1.0508549855651788,
+      "grad_norm": 0.4220631830774893,
+      "learning_rate": 9.214916262661977e-06,
+      "loss": 0.045,
+      "step": 2366
+    },
+    {
+      "epoch": 1.0512991339107263,
+      "grad_norm": 0.4512747724107325,
+      "learning_rate": 9.213873129817346e-06,
+      "loss": 0.0543,
+      "step": 2367
+    },
+    {
+      "epoch": 1.0517432822562736,
+      "grad_norm": 0.4570731167347105,
+      "learning_rate": 9.212829363556003e-06,
+      "loss": 0.0594,
+      "step": 2368
+    },
+    {
+      "epoch": 1.052187430601821,
+      "grad_norm": 0.508942846260648,
+      "learning_rate": 9.211784964034842e-06,
+      "loss": 0.0685,
+      "step": 2369
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.5865147734769778,
+      "learning_rate": 9.210739931410857e-06,
+      "loss": 0.064,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0530757272929159,
+      "grad_norm": 0.6404273704289678,
+      "learning_rate": 9.209694265841132e-06,
+      "loss": 0.0449,
+      "step": 2371
+    },
+    {
+      "epoch": 1.0535198756384632,
+      "grad_norm": 0.5024824890374885,
+      "learning_rate": 9.208647967482849e-06,
+      "loss": 0.053,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0539640239840107,
+      "grad_norm": 0.5079871293280772,
+      "learning_rate": 9.207601036493284e-06,
+      "loss": 0.0504,
+      "step": 2373
+    },
+    {
+      "epoch": 1.054408172329558,
+      "grad_norm": 0.5005289716599005,
+      "learning_rate": 9.206553473029807e-06,
+      "loss": 0.0469,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0548523206751055,
+      "grad_norm": 0.6348636511274667,
+      "learning_rate": 9.205505277249888e-06,
+      "loss": 0.0664,
+      "step": 2375
+    },
+    {
+      "epoch": 1.055296469020653,
+      "grad_norm": 0.7124285353701557,
+      "learning_rate": 9.204456449311086e-06,
+      "loss": 0.0811,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0557406173662003,
+      "grad_norm": 0.7155216157670515,
+      "learning_rate": 9.203406989371058e-06,
+      "loss": 0.068,
+      "step": 2377
+    },
+    {
+      "epoch": 1.0561847657117478,
+      "grad_norm": 0.6903207059209383,
+      "learning_rate": 9.202356897587556e-06,
+      "loss": 0.0625,
+      "step": 2378
+    },
+    {
+      "epoch": 1.056628914057295,
+      "grad_norm": 0.4483522509553745,
+      "learning_rate": 9.201306174118428e-06,
+      "loss": 0.0461,
+      "step": 2379
+    },
+    {
+      "epoch": 1.0570730624028426,
+      "grad_norm": 0.6352875132111704,
+      "learning_rate": 9.200254819121612e-06,
+      "loss": 0.0838,
+      "step": 2380
+    },
+    {
+      "epoch": 1.05751721074839,
+      "grad_norm": 0.600462738349843,
+      "learning_rate": 9.19920283275515e-06,
+      "loss": 0.0498,
+      "step": 2381
+    },
+    {
+      "epoch": 1.0579613590939374,
+      "grad_norm": 0.513630635347009,
+      "learning_rate": 9.198150215177168e-06,
+      "loss": 0.0557,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0584055074394847,
+      "grad_norm": 0.674636182160706,
+      "learning_rate": 9.197096966545896e-06,
+      "loss": 0.0608,
+      "step": 2383
+    },
+    {
+      "epoch": 1.0588496557850322,
+      "grad_norm": 0.7211986590126924,
+      "learning_rate": 9.196043087019651e-06,
+      "loss": 0.0675,
+      "step": 2384
+    },
+    {
+      "epoch": 1.0592938041305797,
+      "grad_norm": 0.48469937945694536,
+      "learning_rate": 9.194988576756855e-06,
+      "loss": 0.0533,
+      "step": 2385
+    },
+    {
+      "epoch": 1.059737952476127,
+      "grad_norm": 0.6223548159808464,
+      "learning_rate": 9.193933435916013e-06,
+      "loss": 0.0609,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0601821008216745,
+      "grad_norm": 0.4702625375198161,
+      "learning_rate": 9.192877664655736e-06,
+      "loss": 0.0629,
+      "step": 2387
+    },
+    {
+      "epoch": 1.0606262491672218,
+      "grad_norm": 0.6156516266579465,
+      "learning_rate": 9.191821263134718e-06,
+      "loss": 0.0519,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0610703975127693,
+      "grad_norm": 0.41634294802260347,
+      "learning_rate": 9.19076423151176e-06,
+      "loss": 0.0339,
+      "step": 2389
+    },
+    {
+      "epoch": 1.0615145458583166,
+      "grad_norm": 0.5676744907352943,
+      "learning_rate": 9.189706569945749e-06,
+      "loss": 0.0672,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0619586942038641,
+      "grad_norm": 0.5506701658376884,
+      "learning_rate": 9.188648278595669e-06,
+      "loss": 0.0549,
+      "step": 2391
+    },
+    {
+      "epoch": 1.0624028425494114,
+      "grad_norm": 0.5548831925430109,
+      "learning_rate": 9.187589357620602e-06,
+      "loss": 0.0495,
+      "step": 2392
+    },
+    {
+      "epoch": 1.062846990894959,
+      "grad_norm": 0.5733351694159566,
+      "learning_rate": 9.186529807179715e-06,
+      "loss": 0.0413,
+      "step": 2393
+    },
+    {
+      "epoch": 1.0632911392405062,
+      "grad_norm": 0.4263224419709196,
+      "learning_rate": 9.185469627432287e-06,
+      "loss": 0.0376,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0637352875860537,
+      "grad_norm": 0.5107765087622207,
+      "learning_rate": 9.184408818537673e-06,
+      "loss": 0.0441,
+      "step": 2395
+    },
+    {
+      "epoch": 1.0641794359316012,
+      "grad_norm": 0.4468370290354222,
+      "learning_rate": 9.183347380655332e-06,
+      "loss": 0.0393,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0646235842771485,
+      "grad_norm": 0.4890359715306359,
+      "learning_rate": 9.182285313944818e-06,
+      "loss": 0.0481,
+      "step": 2397
+    },
+    {
+      "epoch": 1.065067732622696,
+      "grad_norm": 0.45195434165164133,
+      "learning_rate": 9.181222618565777e-06,
+      "loss": 0.0509,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0655118809682433,
+      "grad_norm": 0.49289142398585184,
+      "learning_rate": 9.180159294677948e-06,
+      "loss": 0.0515,
+      "step": 2399
+    },
+    {
+      "epoch": 1.0659560293137909,
+      "grad_norm": 0.42385971921718624,
+      "learning_rate": 9.179095342441171e-06,
+      "loss": 0.0448,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0664001776593381,
+      "grad_norm": 0.6172728890244491,
+      "learning_rate": 9.178030762015372e-06,
+      "loss": 0.0401,
+      "step": 2401
+    },
+    {
+      "epoch": 1.0668443260048857,
+      "grad_norm": 0.5217625747110314,
+      "learning_rate": 9.176965553560578e-06,
+      "loss": 0.0578,
+      "step": 2402
+    },
+    {
+      "epoch": 1.067288474350433,
+      "grad_norm": 0.4881611280755445,
+      "learning_rate": 9.175899717236907e-06,
+      "loss": 0.0437,
+      "step": 2403
+    },
+    {
+      "epoch": 1.0677326226959805,
+      "grad_norm": 0.5243251412540129,
+      "learning_rate": 9.174833253204571e-06,
+      "loss": 0.0408,
+      "step": 2404
+    },
+    {
+      "epoch": 1.068176771041528,
+      "grad_norm": 0.4299956266707952,
+      "learning_rate": 9.17376616162388e-06,
+      "loss": 0.0416,
+      "step": 2405
+    },
+    {
+      "epoch": 1.0686209193870753,
+      "grad_norm": 0.5888767684513153,
+      "learning_rate": 9.172698442655236e-06,
+      "loss": 0.064,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0690650677326228,
+      "grad_norm": 0.4022150666774181,
+      "learning_rate": 9.171630096459134e-06,
+      "loss": 0.0416,
+      "step": 2407
+    },
+    {
+      "epoch": 1.06950921607817,
+      "grad_norm": 0.9093917263815293,
+      "learning_rate": 9.170561123196165e-06,
+      "loss": 0.0573,
+      "step": 2408
+    },
+    {
+      "epoch": 1.0699533644237176,
+      "grad_norm": 0.5641484333920734,
+      "learning_rate": 9.169491523027012e-06,
+      "loss": 0.0527,
+      "step": 2409
+    },
+    {
+      "epoch": 1.0703975127692649,
+      "grad_norm": 0.5614858136344199,
+      "learning_rate": 9.168421296112457e-06,
+      "loss": 0.0525,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0708416611148124,
+      "grad_norm": 0.5361742414415347,
+      "learning_rate": 9.167350442613371e-06,
+      "loss": 0.0512,
+      "step": 2411
+    },
+    {
+      "epoch": 1.0712858094603597,
+      "grad_norm": 0.4065744899599026,
+      "learning_rate": 9.166278962690724e-06,
+      "loss": 0.0352,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0717299578059072,
+      "grad_norm": 0.7190511616030503,
+      "learning_rate": 9.165206856505577e-06,
+      "loss": 0.0711,
+      "step": 2413
+    },
+    {
+      "epoch": 1.0721741061514547,
+      "grad_norm": 0.5266458384013354,
+      "learning_rate": 9.164134124219085e-06,
+      "loss": 0.0436,
+      "step": 2414
+    },
+    {
+      "epoch": 1.072618254497002,
+      "grad_norm": 0.5599669781296026,
+      "learning_rate": 9.163060765992495e-06,
+      "loss": 0.0759,
+      "step": 2415
+    },
+    {
+      "epoch": 1.0730624028425495,
+      "grad_norm": 0.5128109774041522,
+      "learning_rate": 9.161986781987156e-06,
+      "loss": 0.0455,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0735065511880968,
+      "grad_norm": 0.4078193083238796,
+      "learning_rate": 9.160912172364503e-06,
+      "loss": 0.0403,
+      "step": 2417
+    },
+    {
+      "epoch": 1.0739506995336443,
+      "grad_norm": 0.6961390181093496,
+      "learning_rate": 9.15983693728607e-06,
+      "loss": 0.0757,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0743948478791916,
+      "grad_norm": 0.518134789703249,
+      "learning_rate": 9.158761076913481e-06,
+      "loss": 0.0458,
+      "step": 2419
+    },
+    {
+      "epoch": 1.074838996224739,
+      "grad_norm": 0.6466366059180623,
+      "learning_rate": 9.157684591408458e-06,
+      "loss": 0.0639,
+      "step": 2420
+    },
+    {
+      "epoch": 1.0752831445702864,
+      "grad_norm": 0.45726102190092227,
+      "learning_rate": 9.156607480932813e-06,
+      "loss": 0.0485,
+      "step": 2421
+    },
+    {
+      "epoch": 1.075727292915834,
+      "grad_norm": 0.47111677703640187,
+      "learning_rate": 9.155529745648457e-06,
+      "loss": 0.0576,
+      "step": 2422
+    },
+    {
+      "epoch": 1.0761714412613812,
+      "grad_norm": 0.4424777927187194,
+      "learning_rate": 9.154451385717387e-06,
+      "loss": 0.0415,
+      "step": 2423
+    },
+    {
+      "epoch": 1.0766155896069287,
+      "grad_norm": 0.5637865585741092,
+      "learning_rate": 9.153372401301706e-06,
+      "loss": 0.0569,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0770597379524762,
+      "grad_norm": 0.6193090479661898,
+      "learning_rate": 9.152292792563596e-06,
+      "loss": 0.0541,
+      "step": 2425
+    },
+    {
+      "epoch": 1.0775038862980235,
+      "grad_norm": 0.5604767145835976,
+      "learning_rate": 9.151212559665345e-06,
+      "loss": 0.0487,
+      "step": 2426
+    },
+    {
+      "epoch": 1.077948034643571,
+      "grad_norm": 0.614420431496054,
+      "learning_rate": 9.150131702769332e-06,
+      "loss": 0.0543,
+      "step": 2427
+    },
+    {
+      "epoch": 1.0783921829891183,
+      "grad_norm": 0.500155979890897,
+      "learning_rate": 9.149050222038024e-06,
+      "loss": 0.0543,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0788363313346658,
+      "grad_norm": 0.6046519445448296,
+      "learning_rate": 9.147968117633988e-06,
+      "loss": 0.0669,
+      "step": 2429
+    },
+    {
+      "epoch": 1.0792804796802131,
+      "grad_norm": 0.5730375661279963,
+      "learning_rate": 9.14688538971988e-06,
+      "loss": 0.0495,
+      "step": 2430
+    },
+    {
+      "epoch": 1.0797246280257606,
+      "grad_norm": 0.482430110901872,
+      "learning_rate": 9.145802038458457e-06,
+      "loss": 0.0484,
+      "step": 2431
+    },
+    {
+      "epoch": 1.080168776371308,
+      "grad_norm": 0.49413095926471534,
+      "learning_rate": 9.144718064012562e-06,
+      "loss": 0.0482,
+      "step": 2432
+    },
+    {
+      "epoch": 1.0806129247168554,
+      "grad_norm": 0.6274129207366034,
+      "learning_rate": 9.143633466545136e-06,
+      "loss": 0.0609,
+      "step": 2433
+    },
+    {
+      "epoch": 1.0810570730624027,
+      "grad_norm": 0.4754419676708797,
+      "learning_rate": 9.142548246219212e-06,
+      "loss": 0.0539,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0815012214079502,
+      "grad_norm": 0.5129065641839461,
+      "learning_rate": 9.141462403197917e-06,
+      "loss": 0.0502,
+      "step": 2435
+    },
+    {
+      "epoch": 1.0819453697534978,
+      "grad_norm": 0.5492083841021022,
+      "learning_rate": 9.14037593764447e-06,
+      "loss": 0.0547,
+      "step": 2436
+    },
+    {
+      "epoch": 1.082389518099045,
+      "grad_norm": 0.5638343322048809,
+      "learning_rate": 9.139288849722188e-06,
+      "loss": 0.0552,
+      "step": 2437
+    },
+    {
+      "epoch": 1.0828336664445926,
+      "grad_norm": 0.48234420470469497,
+      "learning_rate": 9.138201139594478e-06,
+      "loss": 0.0509,
+      "step": 2438
+    },
+    {
+      "epoch": 1.0832778147901398,
+      "grad_norm": 0.48233545504362385,
+      "learning_rate": 9.137112807424842e-06,
+      "loss": 0.0618,
+      "step": 2439
+    },
+    {
+      "epoch": 1.0837219631356874,
+      "grad_norm": 0.6597245369594125,
+      "learning_rate": 9.136023853376872e-06,
+      "loss": 0.0505,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0841661114812347,
+      "grad_norm": 0.8640144400788339,
+      "learning_rate": 9.134934277614258e-06,
+      "loss": 0.0552,
+      "step": 2441
+    },
+    {
+      "epoch": 1.0846102598267822,
+      "grad_norm": 0.5151117640182895,
+      "learning_rate": 9.133844080300783e-06,
+      "loss": 0.0552,
+      "step": 2442
+    },
+    {
+      "epoch": 1.0850544081723295,
+      "grad_norm": 0.5553941913181785,
+      "learning_rate": 9.13275326160032e-06,
+      "loss": 0.068,
+      "step": 2443
+    },
+    {
+      "epoch": 1.085498556517877,
+      "grad_norm": 1.2548272109304157,
+      "learning_rate": 9.131661821676839e-06,
+      "loss": 0.0615,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0859427048634245,
+      "grad_norm": 0.7349991566989792,
+      "learning_rate": 9.130569760694402e-06,
+      "loss": 0.0592,
+      "step": 2445
+    },
+    {
+      "epoch": 1.0863868532089718,
+      "grad_norm": 0.4235071999153545,
+      "learning_rate": 9.129477078817165e-06,
+      "loss": 0.0399,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0868310015545193,
+      "grad_norm": 0.5132669145700011,
+      "learning_rate": 9.128383776209372e-06,
+      "loss": 0.051,
+      "step": 2447
+    },
+    {
+      "epoch": 1.0872751499000666,
+      "grad_norm": 1.6678578417052072,
+      "learning_rate": 9.127289853035371e-06,
+      "loss": 0.068,
+      "step": 2448
+    },
+    {
+      "epoch": 1.087719298245614,
+      "grad_norm": 0.3444388637101336,
+      "learning_rate": 9.126195309459593e-06,
+      "loss": 0.0326,
+      "step": 2449
+    },
+    {
+      "epoch": 1.0881634465911614,
+      "grad_norm": 0.38355662180503935,
+      "learning_rate": 9.12510014564657e-06,
+      "loss": 0.0552,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0886075949367089,
+      "grad_norm": 0.5490923753617978,
+      "learning_rate": 9.124004361760921e-06,
+      "loss": 0.0598,
+      "step": 2451
+    },
+    {
+      "epoch": 1.0890517432822562,
+      "grad_norm": 0.49443123554529045,
+      "learning_rate": 9.122907957967363e-06,
+      "loss": 0.0545,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0894958916278037,
+      "grad_norm": 0.47265541042225623,
+      "learning_rate": 9.121810934430702e-06,
+      "loss": 0.0523,
+      "step": 2453
+    },
+    {
+      "epoch": 1.0899400399733512,
+      "grad_norm": 0.6134212234858121,
+      "learning_rate": 9.12071329131584e-06,
+      "loss": 0.0669,
+      "step": 2454
+    },
+    {
+      "epoch": 1.0903841883188985,
+      "grad_norm": 0.5102377020582352,
+      "learning_rate": 9.119615028787771e-06,
+      "loss": 0.0488,
+      "step": 2455
+    },
+    {
+      "epoch": 1.090828336664446,
+      "grad_norm": 0.5477789256953114,
+      "learning_rate": 9.118516147011585e-06,
+      "loss": 0.0609,
+      "step": 2456
+    },
+    {
+      "epoch": 1.0912724850099933,
+      "grad_norm": 0.5014262210362811,
+      "learning_rate": 9.117416646152459e-06,
+      "loss": 0.0455,
+      "step": 2457
+    },
+    {
+      "epoch": 1.0917166333555408,
+      "grad_norm": 0.5068151485473731,
+      "learning_rate": 9.11631652637567e-06,
+      "loss": 0.0558,
+      "step": 2458
+    },
+    {
+      "epoch": 1.092160781701088,
+      "grad_norm": 0.5967678649273153,
+      "learning_rate": 9.115215787846583e-06,
+      "loss": 0.0576,
+      "step": 2459
+    },
+    {
+      "epoch": 1.0926049300466356,
+      "grad_norm": 0.5382239114996434,
+      "learning_rate": 9.114114430730656e-06,
+      "loss": 0.0568,
+      "step": 2460
+    },
+    {
+      "epoch": 1.093049078392183,
+      "grad_norm": 0.46675820455299727,
+      "learning_rate": 9.113012455193444e-06,
+      "loss": 0.0437,
+      "step": 2461
+    },
+    {
+      "epoch": 1.0934932267377304,
+      "grad_norm": 0.6129912443014283,
+      "learning_rate": 9.111909861400594e-06,
+      "loss": 0.0584,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0939373750832777,
+      "grad_norm": 0.49745724599799956,
+      "learning_rate": 9.110806649517841e-06,
+      "loss": 0.0541,
+      "step": 2463
+    },
+    {
+      "epoch": 1.0943815234288252,
+      "grad_norm": 0.5223020387615885,
+      "learning_rate": 9.109702819711018e-06,
+      "loss": 0.0579,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0948256717743727,
+      "grad_norm": 0.522897777305782,
+      "learning_rate": 9.108598372146052e-06,
+      "loss": 0.065,
+      "step": 2465
+    },
+    {
+      "epoch": 1.09526982011992,
+      "grad_norm": 0.4978718424633595,
+      "learning_rate": 9.107493306988955e-06,
+      "loss": 0.057,
+      "step": 2466
+    },
+    {
+      "epoch": 1.0957139684654675,
+      "grad_norm": 0.41889427668506446,
+      "learning_rate": 9.10638762440584e-06,
+      "loss": 0.0546,
+      "step": 2467
+    },
+    {
+      "epoch": 1.0961581168110148,
+      "grad_norm": 0.4889531194168305,
+      "learning_rate": 9.10528132456291e-06,
+      "loss": 0.0475,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0966022651565623,
+      "grad_norm": 0.49561727802236016,
+      "learning_rate": 9.10417440762646e-06,
+      "loss": 0.0624,
+      "step": 2469
+    },
+    {
+      "epoch": 1.0970464135021096,
+      "grad_norm": 0.5035282177847037,
+      "learning_rate": 9.10306687376288e-06,
+      "loss": 0.0613,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0974905618476571,
+      "grad_norm": 0.44693886713635184,
+      "learning_rate": 9.101958723138651e-06,
+      "loss": 0.0505,
+      "step": 2471
+    },
+    {
+      "epoch": 1.0979347101932044,
+      "grad_norm": 0.6676580453833789,
+      "learning_rate": 9.100849955920344e-06,
+      "loss": 0.0637,
+      "step": 2472
+    },
+    {
+      "epoch": 1.098378858538752,
+      "grad_norm": 0.47501330783440915,
+      "learning_rate": 9.099740572274627e-06,
+      "loss": 0.0391,
+      "step": 2473
+    },
+    {
+      "epoch": 1.0988230068842995,
+      "grad_norm": 0.6923283723141113,
+      "learning_rate": 9.098630572368262e-06,
+      "loss": 0.0518,
+      "step": 2474
+    },
+    {
+      "epoch": 1.0992671552298467,
+      "grad_norm": 0.47645749570109275,
+      "learning_rate": 9.097519956368096e-06,
+      "loss": 0.0273,
+      "step": 2475
+    },
+    {
+      "epoch": 1.0997113035753943,
+      "grad_norm": 0.8125207380279094,
+      "learning_rate": 9.096408724441078e-06,
+      "loss": 0.062,
+      "step": 2476
+    },
+    {
+      "epoch": 1.1001554519209416,
+      "grad_norm": 0.6208955890687268,
+      "learning_rate": 9.09529687675424e-06,
+      "loss": 0.0399,
+      "step": 2477
+    },
+    {
+      "epoch": 1.100599600266489,
+      "grad_norm": 0.6086645282708443,
+      "learning_rate": 9.094184413474716e-06,
+      "loss": 0.0538,
+      "step": 2478
+    },
+    {
+      "epoch": 1.1010437486120364,
+      "grad_norm": 0.5638622958089239,
+      "learning_rate": 9.093071334769727e-06,
+      "loss": 0.087,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1014878969575839,
+      "grad_norm": 0.5860122134432946,
+      "learning_rate": 9.091957640806585e-06,
+      "loss": 0.0514,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1019320453031312,
+      "grad_norm": 0.5483193797542968,
+      "learning_rate": 9.090843331752704e-06,
+      "loss": 0.0837,
+      "step": 2481
+    },
+    {
+      "epoch": 1.1023761936486787,
+      "grad_norm": 0.671106831086069,
+      "learning_rate": 9.089728407775576e-06,
+      "loss": 0.0533,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1028203419942262,
+      "grad_norm": 0.471025449998095,
+      "learning_rate": 9.088612869042794e-06,
+      "loss": 0.0694,
+      "step": 2483
+    },
+    {
+      "epoch": 1.1032644903397735,
+      "grad_norm": 0.46180112116306377,
+      "learning_rate": 9.087496715722049e-06,
+      "loss": 0.0426,
+      "step": 2484
+    },
+    {
+      "epoch": 1.103708638685321,
+      "grad_norm": 0.49860403863116065,
+      "learning_rate": 9.08637994798111e-06,
+      "loss": 0.0583,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1041527870308683,
+      "grad_norm": 0.5897260124865925,
+      "learning_rate": 9.08526256598785e-06,
+      "loss": 0.0648,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1045969353764158,
+      "grad_norm": 0.453592540271068,
+      "learning_rate": 9.084144569910229e-06,
+      "loss": 0.0455,
+      "step": 2487
+    },
+    {
+      "epoch": 1.105041083721963,
+      "grad_norm": 0.6113711566277047,
+      "learning_rate": 9.083025959916302e-06,
+      "loss": 0.0556,
+      "step": 2488
+    },
+    {
+      "epoch": 1.1054852320675106,
+      "grad_norm": 0.4026753684401193,
+      "learning_rate": 9.081906736174217e-06,
+      "loss": 0.0347,
+      "step": 2489
+    },
+    {
+      "epoch": 1.1059293804130579,
+      "grad_norm": 0.7535472922225546,
+      "learning_rate": 9.080786898852207e-06,
+      "loss": 0.0561,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1063735287586054,
+      "grad_norm": 0.7712189521196268,
+      "learning_rate": 9.079666448118607e-06,
+      "loss": 0.0815,
+      "step": 2491
+    },
+    {
+      "epoch": 1.1068176771041527,
+      "grad_norm": 0.549393105917364,
+      "learning_rate": 9.07854538414184e-06,
+      "loss": 0.0411,
+      "step": 2492
+    },
+    {
+      "epoch": 1.1072618254497002,
+      "grad_norm": 0.8270414561268318,
+      "learning_rate": 9.077423707090418e-06,
+      "loss": 0.0697,
+      "step": 2493
+    },
+    {
+      "epoch": 1.1077059737952477,
+      "grad_norm": 0.4463853554583335,
+      "learning_rate": 9.07630141713295e-06,
+      "loss": 0.0426,
+      "step": 2494
+    },
+    {
+      "epoch": 1.108150122140795,
+      "grad_norm": 0.6269526052059413,
+      "learning_rate": 9.075178514438133e-06,
+      "loss": 0.059,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1085942704863425,
+      "grad_norm": 0.7332597224037389,
+      "learning_rate": 9.074054999174762e-06,
+      "loss": 0.0647,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1090384188318898,
+      "grad_norm": 0.6649774981712752,
+      "learning_rate": 9.072930871511718e-06,
+      "loss": 0.0751,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1094825671774373,
+      "grad_norm": 0.8146175851807935,
+      "learning_rate": 9.071806131617976e-06,
+      "loss": 0.0591,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1099267155229846,
+      "grad_norm": 0.6708407619412956,
+      "learning_rate": 9.070680779662606e-06,
+      "loss": 0.0552,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1103708638685321,
+      "grad_norm": 0.6510598528132346,
+      "learning_rate": 9.069554815814765e-06,
+      "loss": 0.0547,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1108150122140794,
+      "grad_norm": 0.7337961607233396,
+      "learning_rate": 9.068428240243705e-06,
+      "loss": 0.08,
+      "step": 2501
+    },
+    {
+      "epoch": 1.111259160559627,
+      "grad_norm": 0.9471208974140737,
+      "learning_rate": 9.067301053118773e-06,
+      "loss": 0.0575,
+      "step": 2502
+    },
+    {
+      "epoch": 1.1117033089051742,
+      "grad_norm": 0.6247228777438872,
+      "learning_rate": 9.066173254609399e-06,
+      "loss": 0.0593,
+      "step": 2503
+    },
+    {
+      "epoch": 1.1121474572507217,
+      "grad_norm": 0.4119422798089505,
+      "learning_rate": 9.065044844885111e-06,
+      "loss": 0.0369,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1125916055962692,
+      "grad_norm": 0.5707033305872198,
+      "learning_rate": 9.063915824115531e-06,
+      "loss": 0.0586,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1130357539418165,
+      "grad_norm": 0.5864819346852046,
+      "learning_rate": 9.062786192470372e-06,
+      "loss": 0.0541,
+      "step": 2506
+    },
+    {
+      "epoch": 1.113479902287364,
+      "grad_norm": 0.4941216006043351,
+      "learning_rate": 9.06165595011943e-06,
+      "loss": 0.0467,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.7179600112901418,
+      "learning_rate": 9.060525097232603e-06,
+      "loss": 0.061,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1143681989784588,
+      "grad_norm": 0.536437166711899,
+      "learning_rate": 9.059393633979881e-06,
+      "loss": 0.0514,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1148123473240061,
+      "grad_norm": 0.5200228688873401,
+      "learning_rate": 9.058261560531337e-06,
+      "loss": 0.0489,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1152564956695536,
+      "grad_norm": 0.8639507676709893,
+      "learning_rate": 9.057128877057141e-06,
+      "loss": 0.0777,
+      "step": 2511
+    },
+    {
+      "epoch": 1.1157006440151012,
+      "grad_norm": 0.7459642047335504,
+      "learning_rate": 9.055995583727559e-06,
+      "loss": 0.0462,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1161447923606485,
+      "grad_norm": 0.43557336969374605,
+      "learning_rate": 9.05486168071294e-06,
+      "loss": 0.0411,
+      "step": 2513
+    },
+    {
+      "epoch": 1.116588940706196,
+      "grad_norm": 0.4693246881167572,
+      "learning_rate": 9.05372716818373e-06,
+      "loss": 0.0317,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1170330890517433,
+      "grad_norm": 0.5801314945806954,
+      "learning_rate": 9.052592046310466e-06,
+      "loss": 0.0504,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1174772373972908,
+      "grad_norm": 0.5303403405055431,
+      "learning_rate": 9.051456315263775e-06,
+      "loss": 0.0417,
+      "step": 2516
+    },
+    {
+      "epoch": 1.117921385742838,
+      "grad_norm": 0.4682906045403869,
+      "learning_rate": 9.05031997521438e-06,
+      "loss": 0.0441,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1183655340883856,
+      "grad_norm": 0.5610438227567353,
+      "learning_rate": 9.049183026333089e-06,
+      "loss": 0.054,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1188096824339329,
+      "grad_norm": 0.5309141601127093,
+      "learning_rate": 9.048045468790805e-06,
+      "loss": 0.0599,
+      "step": 2519
+    },
+    {
+      "epoch": 1.1192538307794804,
+      "grad_norm": 0.4715253267557383,
+      "learning_rate": 9.04690730275852e-06,
+      "loss": 0.0501,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1196979791250277,
+      "grad_norm": 0.42642161615167634,
+      "learning_rate": 9.045768528407326e-06,
+      "loss": 0.0352,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1201421274705752,
+      "grad_norm": 0.8452679234650169,
+      "learning_rate": 9.044629145908397e-06,
+      "loss": 0.0792,
+      "step": 2522
+    },
+    {
+      "epoch": 1.1205862758161227,
+      "grad_norm": 0.5024078744040577,
+      "learning_rate": 9.043489155433e-06,
+      "loss": 0.0578,
+      "step": 2523
+    },
+    {
+      "epoch": 1.12103042416167,
+      "grad_norm": 0.6039439853816904,
+      "learning_rate": 9.042348557152495e-06,
+      "loss": 0.0591,
+      "step": 2524
+    },
+    {
+      "epoch": 1.1214745725072175,
+      "grad_norm": 0.5200721258931258,
+      "learning_rate": 9.041207351238336e-06,
+      "loss": 0.0589,
+      "step": 2525
+    },
+    {
+      "epoch": 1.1219187208527648,
+      "grad_norm": 0.6201687763530818,
+      "learning_rate": 9.040065537862063e-06,
+      "loss": 0.0564,
+      "step": 2526
+    },
+    {
+      "epoch": 1.1223628691983123,
+      "grad_norm": 0.4816887693066457,
+      "learning_rate": 9.038923117195313e-06,
+      "loss": 0.047,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1228070175438596,
+      "grad_norm": 0.4930745832447622,
+      "learning_rate": 9.037780089409807e-06,
+      "loss": 0.0437,
+      "step": 2528
+    },
+    {
+      "epoch": 1.123251165889407,
+      "grad_norm": 0.4493292408106485,
+      "learning_rate": 9.036636454677363e-06,
+      "loss": 0.0587,
+      "step": 2529
+    },
+    {
+      "epoch": 1.1236953142349544,
+      "grad_norm": 0.5080635531655728,
+      "learning_rate": 9.035492213169892e-06,
+      "loss": 0.0536,
+      "step": 2530
+    },
+    {
+      "epoch": 1.124139462580502,
+      "grad_norm": 0.3142735249050316,
+      "learning_rate": 9.034347365059389e-06,
+      "loss": 0.0324,
+      "step": 2531
+    },
+    {
+      "epoch": 1.1245836109260492,
+      "grad_norm": 0.4572661512872438,
+      "learning_rate": 9.033201910517944e-06,
+      "loss": 0.0484,
+      "step": 2532
+    },
+    {
+      "epoch": 1.1250277592715967,
+      "grad_norm": 0.5858908999181529,
+      "learning_rate": 9.032055849717743e-06,
+      "loss": 0.041,
+      "step": 2533
+    },
+    {
+      "epoch": 1.1254719076171442,
+      "grad_norm": 0.5641689055994421,
+      "learning_rate": 9.030909182831052e-06,
+      "loss": 0.062,
+      "step": 2534
+    },
+    {
+      "epoch": 1.1259160559626915,
+      "grad_norm": 0.48230195194298414,
+      "learning_rate": 9.02976191003024e-06,
+      "loss": 0.055,
+      "step": 2535
+    },
+    {
+      "epoch": 1.126360204308239,
+      "grad_norm": 0.45378491026345646,
+      "learning_rate": 9.028614031487757e-06,
+      "loss": 0.0485,
+      "step": 2536
+    },
+    {
+      "epoch": 1.1268043526537863,
+      "grad_norm": 0.5351056809427763,
+      "learning_rate": 9.027465547376154e-06,
+      "loss": 0.0554,
+      "step": 2537
+    },
+    {
+      "epoch": 1.1272485009993338,
+      "grad_norm": 0.44870767793480254,
+      "learning_rate": 9.02631645786806e-06,
+      "loss": 0.0443,
+      "step": 2538
+    },
+    {
+      "epoch": 1.1276926493448811,
+      "grad_norm": 1.1371656771905172,
+      "learning_rate": 9.02516676313621e-06,
+      "loss": 0.0572,
+      "step": 2539
+    },
+    {
+      "epoch": 1.1281367976904286,
+      "grad_norm": 0.5747790388905519,
+      "learning_rate": 9.02401646335342e-06,
+      "loss": 0.0662,
+      "step": 2540
+    },
+    {
+      "epoch": 1.1285809460359761,
+      "grad_norm": 0.5477714109841172,
+      "learning_rate": 9.022865558692599e-06,
+      "loss": 0.0629,
+      "step": 2541
+    },
+    {
+      "epoch": 1.1290250943815234,
+      "grad_norm": 0.7201777798009189,
+      "learning_rate": 9.021714049326749e-06,
+      "loss": 0.0642,
+      "step": 2542
+    },
+    {
+      "epoch": 1.1294692427270707,
+      "grad_norm": 0.5084161768229669,
+      "learning_rate": 9.02056193542896e-06,
+      "loss": 0.0413,
+      "step": 2543
+    },
+    {
+      "epoch": 1.1299133910726182,
+      "grad_norm": 0.5818759215205692,
+      "learning_rate": 9.019409217172414e-06,
+      "loss": 0.0458,
+      "step": 2544
+    },
+    {
+      "epoch": 1.1303575394181657,
+      "grad_norm": 0.4060737061879425,
+      "learning_rate": 9.018255894730384e-06,
+      "loss": 0.0417,
+      "step": 2545
+    },
+    {
+      "epoch": 1.130801687763713,
+      "grad_norm": 0.37654168507528063,
+      "learning_rate": 9.017101968276237e-06,
+      "loss": 0.0356,
+      "step": 2546
+    },
+    {
+      "epoch": 1.1312458361092605,
+      "grad_norm": 0.5269591862707016,
+      "learning_rate": 9.015947437983423e-06,
+      "loss": 0.0502,
+      "step": 2547
+    },
+    {
+      "epoch": 1.1316899844548078,
+      "grad_norm": 0.6087429306966435,
+      "learning_rate": 9.014792304025492e-06,
+      "loss": 0.0549,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1321341328003554,
+      "grad_norm": 0.47161508183894374,
+      "learning_rate": 9.013636566576078e-06,
+      "loss": 0.046,
+      "step": 2549
+    },
+    {
+      "epoch": 1.1325782811459026,
+      "grad_norm": 0.5566854251234638,
+      "learning_rate": 9.012480225808908e-06,
+      "loss": 0.0582,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1330224294914502,
+      "grad_norm": 0.7226965671519505,
+      "learning_rate": 9.0113232818978e-06,
+      "loss": 0.0535,
+      "step": 2551
+    },
+    {
+      "epoch": 1.1334665778369977,
+      "grad_norm": 0.528548310886907,
+      "learning_rate": 9.010165735016663e-06,
+      "loss": 0.0471,
+      "step": 2552
+    },
+    {
+      "epoch": 1.133910726182545,
+      "grad_norm": 0.6224718792548548,
+      "learning_rate": 9.009007585339493e-06,
+      "loss": 0.0411,
+      "step": 2553
+    },
+    {
+      "epoch": 1.1343548745280925,
+      "grad_norm": 0.46637655730340244,
+      "learning_rate": 9.007848833040385e-06,
+      "loss": 0.0382,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1347990228736398,
+      "grad_norm": 0.6524984653497812,
+      "learning_rate": 9.006689478293513e-06,
+      "loss": 0.0649,
+      "step": 2555
+    },
+    {
+      "epoch": 1.1352431712191873,
+      "grad_norm": 0.42573749158130464,
+      "learning_rate": 9.005529521273152e-06,
+      "loss": 0.0333,
+      "step": 2556
+    },
+    {
+      "epoch": 1.1356873195647346,
+      "grad_norm": 0.4720058401308709,
+      "learning_rate": 9.004368962153662e-06,
+      "loss": 0.0454,
+      "step": 2557
+    },
+    {
+      "epoch": 1.136131467910282,
+      "grad_norm": 0.4623215710338066,
+      "learning_rate": 9.003207801109495e-06,
+      "loss": 0.0413,
+      "step": 2558
+    },
+    {
+      "epoch": 1.1365756162558294,
+      "grad_norm": 0.6101080407770559,
+      "learning_rate": 9.002046038315192e-06,
+      "loss": 0.0534,
+      "step": 2559
+    },
+    {
+      "epoch": 1.1370197646013769,
+      "grad_norm": 0.5040799125891366,
+      "learning_rate": 9.000883673945387e-06,
+      "loss": 0.0392,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1374639129469242,
+      "grad_norm": 0.6363733147612941,
+      "learning_rate": 8.999720708174802e-06,
+      "loss": 0.0628,
+      "step": 2561
+    },
+    {
+      "epoch": 1.1379080612924717,
+      "grad_norm": 0.5620665765918448,
+      "learning_rate": 8.998557141178252e-06,
+      "loss": 0.0524,
+      "step": 2562
+    },
+    {
+      "epoch": 1.1383522096380192,
+      "grad_norm": 0.48743987280774526,
+      "learning_rate": 8.99739297313064e-06,
+      "loss": 0.0502,
+      "step": 2563
+    },
+    {
+      "epoch": 1.1387963579835665,
+      "grad_norm": 0.4714981818194369,
+      "learning_rate": 8.99622820420696e-06,
+      "loss": 0.0477,
+      "step": 2564
+    },
+    {
+      "epoch": 1.139240506329114,
+      "grad_norm": 0.466559878027,
+      "learning_rate": 8.995062834582297e-06,
+      "loss": 0.0585,
+      "step": 2565
+    },
+    {
+      "epoch": 1.1396846546746613,
+      "grad_norm": 0.5806057329454937,
+      "learning_rate": 8.993896864431825e-06,
+      "loss": 0.0653,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1401288030202088,
+      "grad_norm": 0.565404505122367,
+      "learning_rate": 8.992730293930812e-06,
+      "loss": 0.0375,
+      "step": 2567
+    },
+    {
+      "epoch": 1.140572951365756,
+      "grad_norm": 0.49869252300833783,
+      "learning_rate": 8.99156312325461e-06,
+      "loss": 0.0412,
+      "step": 2568
+    },
+    {
+      "epoch": 1.1410170997113036,
+      "grad_norm": 0.4533506243280891,
+      "learning_rate": 8.990395352578665e-06,
+      "loss": 0.0423,
+      "step": 2569
+    },
+    {
+      "epoch": 1.141461248056851,
+      "grad_norm": 0.5186554476274421,
+      "learning_rate": 8.989226982078513e-06,
+      "loss": 0.0486,
+      "step": 2570
+    },
+    {
+      "epoch": 1.1419053964023984,
+      "grad_norm": 0.5971608007231468,
+      "learning_rate": 8.988058011929781e-06,
+      "loss": 0.063,
+      "step": 2571
+    },
+    {
+      "epoch": 1.1423495447479457,
+      "grad_norm": 0.6848430211876612,
+      "learning_rate": 8.986888442308187e-06,
+      "loss": 0.0815,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1427936930934932,
+      "grad_norm": 0.5213622343694703,
+      "learning_rate": 8.985718273389532e-06,
+      "loss": 0.0426,
+      "step": 2573
+    },
+    {
+      "epoch": 1.1432378414390407,
+      "grad_norm": 0.4678564391636386,
+      "learning_rate": 8.984547505349714e-06,
+      "loss": 0.0461,
+      "step": 2574
+    },
+    {
+      "epoch": 1.143681989784588,
+      "grad_norm": 0.6562307201524248,
+      "learning_rate": 8.983376138364723e-06,
+      "loss": 0.0666,
+      "step": 2575
+    },
+    {
+      "epoch": 1.1441261381301355,
+      "grad_norm": 0.7770057155172067,
+      "learning_rate": 8.982204172610632e-06,
+      "loss": 0.0761,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1445702864756828,
+      "grad_norm": 0.5346372112341888,
+      "learning_rate": 8.981031608263608e-06,
+      "loss": 0.0742,
+      "step": 2577
+    },
+    {
+      "epoch": 1.1450144348212303,
+      "grad_norm": 0.549185701186776,
+      "learning_rate": 8.979858445499908e-06,
+      "loss": 0.0583,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1454585831667776,
+      "grad_norm": 0.44014613661158175,
+      "learning_rate": 8.978684684495875e-06,
+      "loss": 0.0422,
+      "step": 2579
+    },
+    {
+      "epoch": 1.1459027315123251,
+      "grad_norm": 0.43518841056839336,
+      "learning_rate": 8.97751032542795e-06,
+      "loss": 0.0513,
+      "step": 2580
+    },
+    {
+      "epoch": 1.1463468798578726,
+      "grad_norm": 0.3840746634100509,
+      "learning_rate": 8.976335368472657e-06,
+      "loss": 0.0295,
+      "step": 2581
+    },
+    {
+      "epoch": 1.14679102820342,
+      "grad_norm": 0.7385893846304249,
+      "learning_rate": 8.97515981380661e-06,
+      "loss": 0.0794,
+      "step": 2582
+    },
+    {
+      "epoch": 1.1472351765489675,
+      "grad_norm": 0.38857897583910317,
+      "learning_rate": 8.97398366160652e-06,
+      "loss": 0.0361,
+      "step": 2583
+    },
+    {
+      "epoch": 1.1476793248945147,
+      "grad_norm": 0.6486029298679971,
+      "learning_rate": 8.972806912049178e-06,
+      "loss": 0.0732,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1481234732400623,
+      "grad_norm": 0.4395809321084259,
+      "learning_rate": 8.971629565311471e-06,
+      "loss": 0.0426,
+      "step": 2585
+    },
+    {
+      "epoch": 1.1485676215856095,
+      "grad_norm": 0.42320652941064185,
+      "learning_rate": 8.970451621570376e-06,
+      "loss": 0.0476,
+      "step": 2586
+    },
+    {
+      "epoch": 1.149011769931157,
+      "grad_norm": 0.3926981271234761,
+      "learning_rate": 8.969273081002954e-06,
+      "loss": 0.0408,
+      "step": 2587
+    },
+    {
+      "epoch": 1.1494559182767043,
+      "grad_norm": 0.6068945137716437,
+      "learning_rate": 8.96809394378636e-06,
+      "loss": 0.0448,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1499000666222519,
+      "grad_norm": 0.46200186631768236,
+      "learning_rate": 8.966914210097843e-06,
+      "loss": 0.0587,
+      "step": 2589
+    },
+    {
+      "epoch": 1.1503442149677992,
+      "grad_norm": 0.8019976203155184,
+      "learning_rate": 8.965733880114734e-06,
+      "loss": 0.0768,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1507883633133467,
+      "grad_norm": 0.4272911771923715,
+      "learning_rate": 8.964552954014455e-06,
+      "loss": 0.0555,
+      "step": 2591
+    },
+    {
+      "epoch": 1.1512325116588942,
+      "grad_norm": 0.46591985962506655,
+      "learning_rate": 8.963371431974521e-06,
+      "loss": 0.0491,
+      "step": 2592
+    },
+    {
+      "epoch": 1.1516766600044415,
+      "grad_norm": 0.34427825623736114,
+      "learning_rate": 8.962189314172537e-06,
+      "loss": 0.0291,
+      "step": 2593
+    },
+    {
+      "epoch": 1.152120808349989,
+      "grad_norm": 0.3941777289275942,
+      "learning_rate": 8.961006600786191e-06,
+      "loss": 0.0386,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1525649566955363,
+      "grad_norm": 0.7828627909758169,
+      "learning_rate": 8.959823291993268e-06,
+      "loss": 0.0644,
+      "step": 2595
+    },
+    {
+      "epoch": 1.1530091050410838,
+      "grad_norm": 0.44448667498264577,
+      "learning_rate": 8.95863938797164e-06,
+      "loss": 0.0437,
+      "step": 2596
+    },
+    {
+      "epoch": 1.153453253386631,
+      "grad_norm": 0.44346487742364565,
+      "learning_rate": 8.957454888899264e-06,
+      "loss": 0.0462,
+      "step": 2597
+    },
+    {
+      "epoch": 1.1538974017321786,
+      "grad_norm": 0.4503855208067358,
+      "learning_rate": 8.956269794954195e-06,
+      "loss": 0.0467,
+      "step": 2598
+    },
+    {
+      "epoch": 1.1543415500777259,
+      "grad_norm": 0.50075395685955,
+      "learning_rate": 8.95508410631457e-06,
+      "loss": 0.0513,
+      "step": 2599
+    },
+    {
+      "epoch": 1.1547856984232734,
+      "grad_norm": 1.5111182491065092,
+      "learning_rate": 8.953897823158618e-06,
+      "loss": 0.0422,
+      "step": 2600
+    },
+    {
+      "epoch": 1.1552298467688207,
+      "grad_norm": 0.4837059174435265,
+      "learning_rate": 8.95271094566466e-06,
+      "loss": 0.0484,
+      "step": 2601
+    },
+    {
+      "epoch": 1.1556739951143682,
+      "grad_norm": 0.9602960350821099,
+      "learning_rate": 8.9515234740111e-06,
+      "loss": 0.0596,
+      "step": 2602
+    },
+    {
+      "epoch": 1.1561181434599157,
+      "grad_norm": 0.4674309097106074,
+      "learning_rate": 8.950335408376438e-06,
+      "loss": 0.0388,
+      "step": 2603
+    },
+    {
+      "epoch": 1.156562291805463,
+      "grad_norm": 0.5494197343743635,
+      "learning_rate": 8.949146748939259e-06,
+      "loss": 0.0516,
+      "step": 2604
+    },
+    {
+      "epoch": 1.1570064401510105,
+      "grad_norm": 0.603296345877419,
+      "learning_rate": 8.94795749587824e-06,
+      "loss": 0.0648,
+      "step": 2605
+    },
+    {
+      "epoch": 1.1574505884965578,
+      "grad_norm": 0.6840688549848888,
+      "learning_rate": 8.946767649372144e-06,
+      "loss": 0.0633,
+      "step": 2606
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.7710851984691011,
+      "learning_rate": 8.945577209599829e-06,
+      "loss": 0.0557,
+      "step": 2607
+    },
+    {
+      "epoch": 1.1583388851876526,
+      "grad_norm": 0.5783757044901674,
+      "learning_rate": 8.944386176740233e-06,
+      "loss": 0.0552,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1587830335332001,
+      "grad_norm": 0.4854270839314596,
+      "learning_rate": 8.943194550972392e-06,
+      "loss": 0.0531,
+      "step": 2609
+    },
+    {
+      "epoch": 1.1592271818787476,
+      "grad_norm": 0.5109780619188811,
+      "learning_rate": 8.942002332475428e-06,
+      "loss": 0.0512,
+      "step": 2610
+    },
+    {
+      "epoch": 1.159671330224295,
+      "grad_norm": 0.541497267825657,
+      "learning_rate": 8.940809521428551e-06,
+      "loss": 0.0622,
+      "step": 2611
+    },
+    {
+      "epoch": 1.1601154785698422,
+      "grad_norm": 0.6325619808955779,
+      "learning_rate": 8.939616118011058e-06,
+      "loss": 0.0577,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1605596269153897,
+      "grad_norm": 0.5836195560024343,
+      "learning_rate": 8.938422122402342e-06,
+      "loss": 0.0608,
+      "step": 2613
+    },
+    {
+      "epoch": 1.1610037752609372,
+      "grad_norm": 0.5121529339351947,
+      "learning_rate": 8.937227534781878e-06,
+      "loss": 0.0496,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1614479236064845,
+      "grad_norm": 0.5247406850098297,
+      "learning_rate": 8.936032355329233e-06,
+      "loss": 0.0509,
+      "step": 2615
+    },
+    {
+      "epoch": 1.161892071952032,
+      "grad_norm": 0.43915296270490234,
+      "learning_rate": 8.934836584224065e-06,
+      "loss": 0.0483,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1623362202975793,
+      "grad_norm": 0.4849661298577939,
+      "learning_rate": 8.933640221646116e-06,
+      "loss": 0.0534,
+      "step": 2617
+    },
+    {
+      "epoch": 1.1627803686431268,
+      "grad_norm": 0.5053186137211374,
+      "learning_rate": 8.932443267775221e-06,
+      "loss": 0.0508,
+      "step": 2618
+    },
+    {
+      "epoch": 1.1632245169886741,
+      "grad_norm": 0.5361080174409868,
+      "learning_rate": 8.931245722791305e-06,
+      "loss": 0.0609,
+      "step": 2619
+    },
+    {
+      "epoch": 1.1636686653342216,
+      "grad_norm": 0.40985225869641795,
+      "learning_rate": 8.930047586874373e-06,
+      "loss": 0.0458,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1641128136797692,
+      "grad_norm": 0.48841596818152816,
+      "learning_rate": 8.928848860204531e-06,
+      "loss": 0.0782,
+      "step": 2621
+    },
+    {
+      "epoch": 1.1645569620253164,
+      "grad_norm": 0.5219141696813542,
+      "learning_rate": 8.927649542961965e-06,
+      "loss": 0.0498,
+      "step": 2622
+    },
+    {
+      "epoch": 1.165001110370864,
+      "grad_norm": 0.6061661462461002,
+      "learning_rate": 8.926449635326954e-06,
+      "loss": 0.0586,
+      "step": 2623
+    },
+    {
+      "epoch": 1.1654452587164112,
+      "grad_norm": 0.4500746397097711,
+      "learning_rate": 8.925249137479864e-06,
+      "loss": 0.0399,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1658894070619588,
+      "grad_norm": 0.5211157612826388,
+      "learning_rate": 8.92404804960115e-06,
+      "loss": 0.05,
+      "step": 2625
+    },
+    {
+      "epoch": 1.166333555407506,
+      "grad_norm": 0.4903699164468352,
+      "learning_rate": 8.922846371871355e-06,
+      "loss": 0.0496,
+      "step": 2626
+    },
+    {
+      "epoch": 1.1667777037530536,
+      "grad_norm": 0.7464483529673437,
+      "learning_rate": 8.921644104471114e-06,
+      "loss": 0.0601,
+      "step": 2627
+    },
+    {
+      "epoch": 1.1672218520986009,
+      "grad_norm": 0.4872521353158615,
+      "learning_rate": 8.920441247581148e-06,
+      "loss": 0.0525,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1676660004441484,
+      "grad_norm": 0.5674275180118361,
+      "learning_rate": 8.919237801382265e-06,
+      "loss": 0.0552,
+      "step": 2629
+    },
+    {
+      "epoch": 1.1681101487896957,
+      "grad_norm": 0.8950605367456311,
+      "learning_rate": 8.918033766055364e-06,
+      "loss": 0.0602,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1685542971352432,
+      "grad_norm": 0.6732326138636984,
+      "learning_rate": 8.916829141781432e-06,
+      "loss": 0.0558,
+      "step": 2631
+    },
+    {
+      "epoch": 1.1689984454807907,
+      "grad_norm": 0.48994316207093613,
+      "learning_rate": 8.915623928741546e-06,
+      "loss": 0.0628,
+      "step": 2632
+    },
+    {
+      "epoch": 1.169442593826338,
+      "grad_norm": 0.4212795487662145,
+      "learning_rate": 8.914418127116867e-06,
+      "loss": 0.0508,
+      "step": 2633
+    },
+    {
+      "epoch": 1.1698867421718855,
+      "grad_norm": 0.49361391712276814,
+      "learning_rate": 8.91321173708865e-06,
+      "loss": 0.0638,
+      "step": 2634
+    },
+    {
+      "epoch": 1.1703308905174328,
+      "grad_norm": 0.7512476603132378,
+      "learning_rate": 8.912004758838235e-06,
+      "loss": 0.0839,
+      "step": 2635
+    },
+    {
+      "epoch": 1.1707750388629803,
+      "grad_norm": 0.505025169980782,
+      "learning_rate": 8.910797192547051e-06,
+      "loss": 0.0457,
+      "step": 2636
+    },
+    {
+      "epoch": 1.1712191872085276,
+      "grad_norm": 0.5671039324048706,
+      "learning_rate": 8.909589038396617e-06,
+      "loss": 0.044,
+      "step": 2637
+    },
+    {
+      "epoch": 1.171663335554075,
+      "grad_norm": 0.4745487457505175,
+      "learning_rate": 8.908380296568537e-06,
+      "loss": 0.0479,
+      "step": 2638
+    },
+    {
+      "epoch": 1.1721074838996224,
+      "grad_norm": 0.8949118984931155,
+      "learning_rate": 8.907170967244508e-06,
+      "loss": 0.0576,
+      "step": 2639
+    },
+    {
+      "epoch": 1.17255163224517,
+      "grad_norm": 0.4520757913913456,
+      "learning_rate": 8.905961050606311e-06,
+      "loss": 0.0467,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1729957805907172,
+      "grad_norm": 0.4749545091509921,
+      "learning_rate": 8.904750546835817e-06,
+      "loss": 0.0446,
+      "step": 2641
+    },
+    {
+      "epoch": 1.1734399289362647,
+      "grad_norm": 0.522016267548402,
+      "learning_rate": 8.903539456114988e-06,
+      "loss": 0.053,
+      "step": 2642
+    },
+    {
+      "epoch": 1.1738840772818122,
+      "grad_norm": 0.530579514310406,
+      "learning_rate": 8.902327778625865e-06,
+      "loss": 0.0437,
+      "step": 2643
+    },
+    {
+      "epoch": 1.1743282256273595,
+      "grad_norm": 0.6048157018439171,
+      "learning_rate": 8.90111551455059e-06,
+      "loss": 0.0435,
+      "step": 2644
+    },
+    {
+      "epoch": 1.174772373972907,
+      "grad_norm": 0.5906854069033922,
+      "learning_rate": 8.899902664071384e-06,
+      "loss": 0.0626,
+      "step": 2645
+    },
+    {
+      "epoch": 1.1752165223184543,
+      "grad_norm": 0.47030730215576705,
+      "learning_rate": 8.898689227370563e-06,
+      "loss": 0.0386,
+      "step": 2646
+    },
+    {
+      "epoch": 1.1756606706640018,
+      "grad_norm": 0.6351029367838139,
+      "learning_rate": 8.897475204630521e-06,
+      "loss": 0.0591,
+      "step": 2647
+    },
+    {
+      "epoch": 1.176104819009549,
+      "grad_norm": 0.6410082214075851,
+      "learning_rate": 8.89626059603375e-06,
+      "loss": 0.0579,
+      "step": 2648
+    },
+    {
+      "epoch": 1.1765489673550966,
+      "grad_norm": 0.5406985095669288,
+      "learning_rate": 8.895045401762825e-06,
+      "loss": 0.0568,
+      "step": 2649
+    },
+    {
+      "epoch": 1.1769931157006441,
+      "grad_norm": 0.7480895085443751,
+      "learning_rate": 8.893829622000412e-06,
+      "loss": 0.0707,
+      "step": 2650
+    },
+    {
+      "epoch": 1.1774372640461914,
+      "grad_norm": 0.45890411967998634,
+      "learning_rate": 8.892613256929261e-06,
+      "loss": 0.0416,
+      "step": 2651
+    },
+    {
+      "epoch": 1.177881412391739,
+      "grad_norm": 0.396661981002553,
+      "learning_rate": 8.891396306732214e-06,
+      "loss": 0.0423,
+      "step": 2652
+    },
+    {
+      "epoch": 1.1783255607372862,
+      "grad_norm": 0.5069251937753831,
+      "learning_rate": 8.890178771592198e-06,
+      "loss": 0.0487,
+      "step": 2653
+    },
+    {
+      "epoch": 1.1787697090828337,
+      "grad_norm": 0.7785700220445251,
+      "learning_rate": 8.888960651692231e-06,
+      "loss": 0.0719,
+      "step": 2654
+    },
+    {
+      "epoch": 1.179213857428381,
+      "grad_norm": 0.5673896965611613,
+      "learning_rate": 8.887741947215415e-06,
+      "loss": 0.0556,
+      "step": 2655
+    },
+    {
+      "epoch": 1.1796580057739285,
+      "grad_norm": 0.542725413232058,
+      "learning_rate": 8.886522658344944e-06,
+      "loss": 0.0493,
+      "step": 2656
+    },
+    {
+      "epoch": 1.1801021541194758,
+      "grad_norm": 0.6266382123491776,
+      "learning_rate": 8.885302785264098e-06,
+      "loss": 0.0383,
+      "step": 2657
+    },
+    {
+      "epoch": 1.1805463024650233,
+      "grad_norm": 0.46581618376933115,
+      "learning_rate": 8.884082328156243e-06,
+      "loss": 0.0419,
+      "step": 2658
+    },
+    {
+      "epoch": 1.1809904508105706,
+      "grad_norm": 0.5581446013530325,
+      "learning_rate": 8.882861287204836e-06,
+      "loss": 0.0498,
+      "step": 2659
+    },
+    {
+      "epoch": 1.1814345991561181,
+      "grad_norm": 0.42956852547953145,
+      "learning_rate": 8.881639662593417e-06,
+      "loss": 0.0381,
+      "step": 2660
+    },
+    {
+      "epoch": 1.1818787475016657,
+      "grad_norm": 0.45803495008537815,
+      "learning_rate": 8.880417454505622e-06,
+      "loss": 0.0492,
+      "step": 2661
+    },
+    {
+      "epoch": 1.182322895847213,
+      "grad_norm": 0.6914169208326956,
+      "learning_rate": 8.879194663125164e-06,
+      "loss": 0.0517,
+      "step": 2662
+    },
+    {
+      "epoch": 1.1827670441927605,
+      "grad_norm": 0.6395374873974822,
+      "learning_rate": 8.877971288635853e-06,
+      "loss": 0.0631,
+      "step": 2663
+    },
+    {
+      "epoch": 1.1832111925383078,
+      "grad_norm": 0.558846270469075,
+      "learning_rate": 8.876747331221583e-06,
+      "loss": 0.058,
+      "step": 2664
+    },
+    {
+      "epoch": 1.1836553408838553,
+      "grad_norm": 0.49754542881845054,
+      "learning_rate": 8.875522791066333e-06,
+      "loss": 0.0616,
+      "step": 2665
+    },
+    {
+      "epoch": 1.1840994892294026,
+      "grad_norm": 0.641582498045218,
+      "learning_rate": 8.874297668354175e-06,
+      "loss": 0.0432,
+      "step": 2666
+    },
+    {
+      "epoch": 1.18454363757495,
+      "grad_norm": 0.5357146614905643,
+      "learning_rate": 8.873071963269265e-06,
+      "loss": 0.0581,
+      "step": 2667
+    },
+    {
+      "epoch": 1.1849877859204974,
+      "grad_norm": 0.3864891516746441,
+      "learning_rate": 8.871845675995847e-06,
+      "loss": 0.0397,
+      "step": 2668
+    },
+    {
+      "epoch": 1.1854319342660449,
+      "grad_norm": 0.5817776113032109,
+      "learning_rate": 8.870618806718252e-06,
+      "loss": 0.0548,
+      "step": 2669
+    },
+    {
+      "epoch": 1.1858760826115922,
+      "grad_norm": 0.6451637896241947,
+      "learning_rate": 8.8693913556209e-06,
+      "loss": 0.0555,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1863202309571397,
+      "grad_norm": 0.522938279572616,
+      "learning_rate": 8.868163322888298e-06,
+      "loss": 0.0429,
+      "step": 2671
+    },
+    {
+      "epoch": 1.1867643793026872,
+      "grad_norm": 0.6059080038510308,
+      "learning_rate": 8.86693470870504e-06,
+      "loss": 0.056,
+      "step": 2672
+    },
+    {
+      "epoch": 1.1872085276482345,
+      "grad_norm": 0.5024058690626159,
+      "learning_rate": 8.865705513255807e-06,
+      "loss": 0.0371,
+      "step": 2673
+    },
+    {
+      "epoch": 1.187652675993782,
+      "grad_norm": 0.5033221412559168,
+      "learning_rate": 8.864475736725369e-06,
+      "loss": 0.06,
+      "step": 2674
+    },
+    {
+      "epoch": 1.1880968243393293,
+      "grad_norm": 0.4156977033942008,
+      "learning_rate": 8.863245379298582e-06,
+      "loss": 0.0378,
+      "step": 2675
+    },
+    {
+      "epoch": 1.1885409726848768,
+      "grad_norm": 0.8144903668245238,
+      "learning_rate": 8.86201444116039e-06,
+      "loss": 0.0614,
+      "step": 2676
+    },
+    {
+      "epoch": 1.188985121030424,
+      "grad_norm": 0.4226975303790829,
+      "learning_rate": 8.860782922495821e-06,
+      "loss": 0.046,
+      "step": 2677
+    },
+    {
+      "epoch": 1.1894292693759716,
+      "grad_norm": 0.52956763181697,
+      "learning_rate": 8.859550823489997e-06,
+      "loss": 0.0384,
+      "step": 2678
+    },
+    {
+      "epoch": 1.189873417721519,
+      "grad_norm": 0.5489791027942442,
+      "learning_rate": 8.858318144328123e-06,
+      "loss": 0.0742,
+      "step": 2679
+    },
+    {
+      "epoch": 1.1903175660670664,
+      "grad_norm": 1.0552083022753025,
+      "learning_rate": 8.85708488519549e-06,
+      "loss": 0.0446,
+      "step": 2680
+    },
+    {
+      "epoch": 1.1907617144126137,
+      "grad_norm": 0.5078924313476717,
+      "learning_rate": 8.855851046277478e-06,
+      "loss": 0.054,
+      "step": 2681
+    },
+    {
+      "epoch": 1.1912058627581612,
+      "grad_norm": 0.5148608895352339,
+      "learning_rate": 8.854616627759553e-06,
+      "loss": 0.0523,
+      "step": 2682
+    },
+    {
+      "epoch": 1.1916500111037087,
+      "grad_norm": 0.7282958984855765,
+      "learning_rate": 8.853381629827272e-06,
+      "loss": 0.0506,
+      "step": 2683
+    },
+    {
+      "epoch": 1.192094159449256,
+      "grad_norm": 0.5448500153068538,
+      "learning_rate": 8.852146052666275e-06,
+      "loss": 0.0404,
+      "step": 2684
+    },
+    {
+      "epoch": 1.1925383077948035,
+      "grad_norm": 0.40034874918632524,
+      "learning_rate": 8.850909896462288e-06,
+      "loss": 0.0351,
+      "step": 2685
+    },
+    {
+      "epoch": 1.1929824561403508,
+      "grad_norm": 0.6365377232810542,
+      "learning_rate": 8.849673161401129e-06,
+      "loss": 0.0616,
+      "step": 2686
+    },
+    {
+      "epoch": 1.1934266044858983,
+      "grad_norm": 0.5874533737153114,
+      "learning_rate": 8.848435847668699e-06,
+      "loss": 0.06,
+      "step": 2687
+    },
+    {
+      "epoch": 1.1938707528314456,
+      "grad_norm": 0.4075740080944545,
+      "learning_rate": 8.847197955450988e-06,
+      "loss": 0.0399,
+      "step": 2688
+    },
+    {
+      "epoch": 1.1943149011769931,
+      "grad_norm": 0.4560997072221143,
+      "learning_rate": 8.845959484934073e-06,
+      "loss": 0.0482,
+      "step": 2689
+    },
+    {
+      "epoch": 1.1947590495225406,
+      "grad_norm": 0.7271393081297982,
+      "learning_rate": 8.844720436304113e-06,
+      "loss": 0.0694,
+      "step": 2690
+    },
+    {
+      "epoch": 1.195203197868088,
+      "grad_norm": 0.5135039124857559,
+      "learning_rate": 8.843480809747363e-06,
+      "loss": 0.064,
+      "step": 2691
+    },
+    {
+      "epoch": 1.1956473462136354,
+      "grad_norm": 0.5417817389406259,
+      "learning_rate": 8.842240605450158e-06,
+      "loss": 0.0816,
+      "step": 2692
+    },
+    {
+      "epoch": 1.1960914945591827,
+      "grad_norm": 0.40641657577367263,
+      "learning_rate": 8.840999823598921e-06,
+      "loss": 0.0378,
+      "step": 2693
+    },
+    {
+      "epoch": 1.1965356429047302,
+      "grad_norm": 0.7137594073963416,
+      "learning_rate": 8.839758464380163e-06,
+      "loss": 0.063,
+      "step": 2694
+    },
+    {
+      "epoch": 1.1969797912502775,
+      "grad_norm": 0.6152412711374633,
+      "learning_rate": 8.838516527980483e-06,
+      "loss": 0.0403,
+      "step": 2695
+    },
+    {
+      "epoch": 1.197423939595825,
+      "grad_norm": 0.5995872223176858,
+      "learning_rate": 8.837274014586564e-06,
+      "loss": 0.0461,
+      "step": 2696
+    },
+    {
+      "epoch": 1.1978680879413723,
+      "grad_norm": 0.5251664478854555,
+      "learning_rate": 8.836030924385175e-06,
+      "loss": 0.0558,
+      "step": 2697
+    },
+    {
+      "epoch": 1.1983122362869199,
+      "grad_norm": 0.444869598867083,
+      "learning_rate": 8.834787257563178e-06,
+      "loss": 0.0583,
+      "step": 2698
+    },
+    {
+      "epoch": 1.1987563846324671,
+      "grad_norm": 0.5140725378841854,
+      "learning_rate": 8.833543014307513e-06,
+      "loss": 0.0513,
+      "step": 2699
+    },
+    {
+      "epoch": 1.1992005329780147,
+      "grad_norm": 0.5779472328853184,
+      "learning_rate": 8.83229819480521e-06,
+      "loss": 0.0653,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1996446813235622,
+      "grad_norm": 0.5908242045279332,
+      "learning_rate": 8.831052799243394e-06,
+      "loss": 0.0565,
+      "step": 2701
+    },
+    {
+      "epoch": 1.2000888296691095,
+      "grad_norm": 0.9242559999496844,
+      "learning_rate": 8.82980682780926e-06,
+      "loss": 0.0628,
+      "step": 2702
+    },
+    {
+      "epoch": 1.200532978014657,
+      "grad_norm": 0.4108528589278202,
+      "learning_rate": 8.828560280690104e-06,
+      "loss": 0.0438,
+      "step": 2703
+    },
+    {
+      "epoch": 1.2009771263602043,
+      "grad_norm": 0.4411744291440838,
+      "learning_rate": 8.827313158073304e-06,
+      "loss": 0.0567,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2014212747057518,
+      "grad_norm": 0.5260335018456415,
+      "learning_rate": 8.826065460146318e-06,
+      "loss": 0.0579,
+      "step": 2705
+    },
+    {
+      "epoch": 1.201865423051299,
+      "grad_norm": 0.6891291032685649,
+      "learning_rate": 8.824817187096702e-06,
+      "loss": 0.056,
+      "step": 2706
+    },
+    {
+      "epoch": 1.2023095713968466,
+      "grad_norm": 0.7678244223744226,
+      "learning_rate": 8.823568339112089e-06,
+      "loss": 0.0577,
+      "step": 2707
+    },
+    {
+      "epoch": 1.2027537197423939,
+      "grad_norm": 0.605196561971164,
+      "learning_rate": 8.822318916380207e-06,
+      "loss": 0.0511,
+      "step": 2708
+    },
+    {
+      "epoch": 1.2031978680879414,
+      "grad_norm": 0.5123811082634186,
+      "learning_rate": 8.821068919088858e-06,
+      "loss": 0.0578,
+      "step": 2709
+    },
+    {
+      "epoch": 1.2036420164334887,
+      "grad_norm": 0.4168805885973522,
+      "learning_rate": 8.819818347425943e-06,
+      "loss": 0.0402,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2040861647790362,
+      "grad_norm": 0.4322765343829467,
+      "learning_rate": 8.818567201579444e-06,
+      "loss": 0.0404,
+      "step": 2711
+    },
+    {
+      "epoch": 1.2045303131245837,
+      "grad_norm": 0.6086948606935972,
+      "learning_rate": 8.817315481737428e-06,
+      "loss": 0.0574,
+      "step": 2712
+    },
+    {
+      "epoch": 1.204974461470131,
+      "grad_norm": 0.5388644202630685,
+      "learning_rate": 8.816063188088049e-06,
+      "loss": 0.0526,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2054186098156785,
+      "grad_norm": 0.4833033812042681,
+      "learning_rate": 8.814810320819551e-06,
+      "loss": 0.0522,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2058627581612258,
+      "grad_norm": 0.9567007955598438,
+      "learning_rate": 8.81355688012026e-06,
+      "loss": 0.0926,
+      "step": 2715
+    },
+    {
+      "epoch": 1.2063069065067733,
+      "grad_norm": 0.6610578357230338,
+      "learning_rate": 8.812302866178586e-06,
+      "loss": 0.0508,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2067510548523206,
+      "grad_norm": 0.5170485323761623,
+      "learning_rate": 8.811048279183034e-06,
+      "loss": 0.0482,
+      "step": 2717
+    },
+    {
+      "epoch": 1.207195203197868,
+      "grad_norm": 0.49170885647162377,
+      "learning_rate": 8.809793119322188e-06,
+      "loss": 0.0452,
+      "step": 2718
+    },
+    {
+      "epoch": 1.2076393515434156,
+      "grad_norm": 0.6569021320261341,
+      "learning_rate": 8.808537386784717e-06,
+      "loss": 0.0602,
+      "step": 2719
+    },
+    {
+      "epoch": 1.208083499888963,
+      "grad_norm": 0.5861436832219081,
+      "learning_rate": 8.807281081759382e-06,
+      "loss": 0.0503,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2085276482345104,
+      "grad_norm": 0.6362413163945917,
+      "learning_rate": 8.806024204435024e-06,
+      "loss": 0.0619,
+      "step": 2721
+    },
+    {
+      "epoch": 1.2089717965800577,
+      "grad_norm": 0.8112427507399114,
+      "learning_rate": 8.804766755000577e-06,
+      "loss": 0.0878,
+      "step": 2722
+    },
+    {
+      "epoch": 1.2094159449256052,
+      "grad_norm": 0.6213886119436266,
+      "learning_rate": 8.803508733645056e-06,
+      "loss": 0.0391,
+      "step": 2723
+    },
+    {
+      "epoch": 1.2098600932711525,
+      "grad_norm": 0.4701764036600453,
+      "learning_rate": 8.80225014055756e-06,
+      "loss": 0.044,
+      "step": 2724
+    },
+    {
+      "epoch": 1.2103042416167,
+      "grad_norm": 0.6176542835674168,
+      "learning_rate": 8.80099097592728e-06,
+      "loss": 0.0717,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2107483899622473,
+      "grad_norm": 0.5676221760627296,
+      "learning_rate": 8.799731239943488e-06,
+      "loss": 0.0485,
+      "step": 2726
+    },
+    {
+      "epoch": 1.2111925383077948,
+      "grad_norm": 0.7416139133479585,
+      "learning_rate": 8.798470932795545e-06,
+      "loss": 0.0848,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2116366866533421,
+      "grad_norm": 0.8518948114813076,
+      "learning_rate": 8.797210054672897e-06,
+      "loss": 0.0603,
+      "step": 2728
+    },
+    {
+      "epoch": 1.2120808349988896,
+      "grad_norm": 0.5447845702100781,
+      "learning_rate": 8.795948605765071e-06,
+      "loss": 0.0503,
+      "step": 2729
+    },
+    {
+      "epoch": 1.2125249833444371,
+      "grad_norm": 0.6041323757362492,
+      "learning_rate": 8.794686586261692e-06,
+      "loss": 0.0606,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2129691316899844,
+      "grad_norm": 0.3720241663572847,
+      "learning_rate": 8.793423996352458e-06,
+      "loss": 0.0396,
+      "step": 2731
+    },
+    {
+      "epoch": 1.213413280035532,
+      "grad_norm": 0.6614032276078902,
+      "learning_rate": 8.792160836227156e-06,
+      "loss": 0.0778,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2138574283810792,
+      "grad_norm": 0.4371895131814461,
+      "learning_rate": 8.790897106075665e-06,
+      "loss": 0.0464,
+      "step": 2733
+    },
+    {
+      "epoch": 1.2143015767266268,
+      "grad_norm": 0.42112583525881137,
+      "learning_rate": 8.78963280608794e-06,
+      "loss": 0.0283,
+      "step": 2734
+    },
+    {
+      "epoch": 1.214745725072174,
+      "grad_norm": 0.8563223231466682,
+      "learning_rate": 8.788367936454033e-06,
+      "loss": 0.0588,
+      "step": 2735
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.5558136427910825,
+      "learning_rate": 8.78710249736407e-06,
+      "loss": 0.0438,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2156340217632688,
+      "grad_norm": 0.5340710064178852,
+      "learning_rate": 8.78583648900827e-06,
+      "loss": 0.0517,
+      "step": 2737
+    },
+    {
+      "epoch": 1.2160781701088164,
+      "grad_norm": 0.4855046400709726,
+      "learning_rate": 8.784569911576937e-06,
+      "loss": 0.0429,
+      "step": 2738
+    },
+    {
+      "epoch": 1.2165223184543636,
+      "grad_norm": 0.5862019061765111,
+      "learning_rate": 8.783302765260456e-06,
+      "loss": 0.0549,
+      "step": 2739
+    },
+    {
+      "epoch": 1.2169664667999112,
+      "grad_norm": 0.754950173363004,
+      "learning_rate": 8.782035050249302e-06,
+      "loss": 0.051,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2174106151454587,
+      "grad_norm": 0.5310823633824483,
+      "learning_rate": 8.780766766734037e-06,
+      "loss": 0.0535,
+      "step": 2741
+    },
+    {
+      "epoch": 1.217854763491006,
+      "grad_norm": 0.515619945566938,
+      "learning_rate": 8.779497914905302e-06,
+      "loss": 0.0518,
+      "step": 2742
+    },
+    {
+      "epoch": 1.2182989118365535,
+      "grad_norm": 0.5419625287890412,
+      "learning_rate": 8.778228494953826e-06,
+      "loss": 0.0629,
+      "step": 2743
+    },
+    {
+      "epoch": 1.2187430601821008,
+      "grad_norm": 0.41887351869316514,
+      "learning_rate": 8.776958507070427e-06,
+      "loss": 0.0436,
+      "step": 2744
+    },
+    {
+      "epoch": 1.2191872085276483,
+      "grad_norm": 0.5475369085076727,
+      "learning_rate": 8.775687951446007e-06,
+      "loss": 0.043,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2196313568731956,
+      "grad_norm": 0.6765114475250111,
+      "learning_rate": 8.774416828271548e-06,
+      "loss": 0.0542,
+      "step": 2746
+    },
+    {
+      "epoch": 1.220075505218743,
+      "grad_norm": 0.678214320662665,
+      "learning_rate": 8.773145137738125e-06,
+      "loss": 0.0815,
+      "step": 2747
+    },
+    {
+      "epoch": 1.2205196535642906,
+      "grad_norm": 0.49324022002296203,
+      "learning_rate": 8.771872880036893e-06,
+      "loss": 0.0524,
+      "step": 2748
+    },
+    {
+      "epoch": 1.2209638019098379,
+      "grad_norm": 0.5958708788497837,
+      "learning_rate": 8.770600055359094e-06,
+      "loss": 0.0456,
+      "step": 2749
+    },
+    {
+      "epoch": 1.2214079502553852,
+      "grad_norm": 0.4999935592497019,
+      "learning_rate": 8.769326663896056e-06,
+      "loss": 0.0409,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2218520986009327,
+      "grad_norm": 0.43642259806336103,
+      "learning_rate": 8.76805270583919e-06,
+      "loss": 0.05,
+      "step": 2751
+    },
+    {
+      "epoch": 1.2222962469464802,
+      "grad_norm": 0.6350373365457326,
+      "learning_rate": 8.766778181379993e-06,
+      "loss": 0.046,
+      "step": 2752
+    },
+    {
+      "epoch": 1.2227403952920275,
+      "grad_norm": 0.49370518082751286,
+      "learning_rate": 8.765503090710052e-06,
+      "loss": 0.0504,
+      "step": 2753
+    },
+    {
+      "epoch": 1.223184543637575,
+      "grad_norm": 0.6617812231173426,
+      "learning_rate": 8.76422743402103e-06,
+      "loss": 0.0524,
+      "step": 2754
+    },
+    {
+      "epoch": 1.2236286919831223,
+      "grad_norm": 0.46946114196024546,
+      "learning_rate": 8.762951211504682e-06,
+      "loss": 0.0459,
+      "step": 2755
+    },
+    {
+      "epoch": 1.2240728403286698,
+      "grad_norm": 0.49081430568764717,
+      "learning_rate": 8.761674423352844e-06,
+      "loss": 0.0438,
+      "step": 2756
+    },
+    {
+      "epoch": 1.224516988674217,
+      "grad_norm": 0.7508698890744291,
+      "learning_rate": 8.760397069757443e-06,
+      "loss": 0.0535,
+      "step": 2757
+    },
+    {
+      "epoch": 1.2249611370197646,
+      "grad_norm": 0.7095498367244547,
+      "learning_rate": 8.759119150910482e-06,
+      "loss": 0.0439,
+      "step": 2758
+    },
+    {
+      "epoch": 1.2254052853653121,
+      "grad_norm": 0.5423777599181153,
+      "learning_rate": 8.757840667004059e-06,
+      "loss": 0.0584,
+      "step": 2759
+    },
+    {
+      "epoch": 1.2258494337108594,
+      "grad_norm": 0.38826805829970257,
+      "learning_rate": 8.756561618230348e-06,
+      "loss": 0.0372,
+      "step": 2760
+    },
+    {
+      "epoch": 1.226293582056407,
+      "grad_norm": 0.4214243171446597,
+      "learning_rate": 8.755282004781613e-06,
+      "loss": 0.0387,
+      "step": 2761
+    },
+    {
+      "epoch": 1.2267377304019542,
+      "grad_norm": 0.6110973962790421,
+      "learning_rate": 8.754001826850201e-06,
+      "loss": 0.0504,
+      "step": 2762
+    },
+    {
+      "epoch": 1.2271818787475017,
+      "grad_norm": 0.5706371024485902,
+      "learning_rate": 8.752721084628545e-06,
+      "loss": 0.0447,
+      "step": 2763
+    },
+    {
+      "epoch": 1.227626027093049,
+      "grad_norm": 0.7667925876568604,
+      "learning_rate": 8.751439778309162e-06,
+      "loss": 0.0661,
+      "step": 2764
+    },
+    {
+      "epoch": 1.2280701754385965,
+      "grad_norm": 0.4947295202945595,
+      "learning_rate": 8.750157908084655e-06,
+      "loss": 0.0481,
+      "step": 2765
+    },
+    {
+      "epoch": 1.2285143237841438,
+      "grad_norm": 0.4844603657944363,
+      "learning_rate": 8.74887547414771e-06,
+      "loss": 0.0428,
+      "step": 2766
+    },
+    {
+      "epoch": 1.2289584721296913,
+      "grad_norm": 0.5262436629587928,
+      "learning_rate": 8.747592476691102e-06,
+      "loss": 0.0572,
+      "step": 2767
+    },
+    {
+      "epoch": 1.2294026204752386,
+      "grad_norm": 0.48653491041855407,
+      "learning_rate": 8.746308915907681e-06,
+      "loss": 0.0466,
+      "step": 2768
+    },
+    {
+      "epoch": 1.2298467688207861,
+      "grad_norm": 0.3598264709368313,
+      "learning_rate": 8.745024791990392e-06,
+      "loss": 0.036,
+      "step": 2769
+    },
+    {
+      "epoch": 1.2302909171663337,
+      "grad_norm": 0.4281231861639907,
+      "learning_rate": 8.74374010513226e-06,
+      "loss": 0.0375,
+      "step": 2770
+    },
+    {
+      "epoch": 1.230735065511881,
+      "grad_norm": 0.6227157372551906,
+      "learning_rate": 8.742454855526396e-06,
+      "loss": 0.06,
+      "step": 2771
+    },
+    {
+      "epoch": 1.2311792138574285,
+      "grad_norm": 0.5721086967364898,
+      "learning_rate": 8.741169043365994e-06,
+      "loss": 0.0517,
+      "step": 2772
+    },
+    {
+      "epoch": 1.2316233622029757,
+      "grad_norm": 0.4118993047204808,
+      "learning_rate": 8.739882668844332e-06,
+      "loss": 0.0433,
+      "step": 2773
+    },
+    {
+      "epoch": 1.2320675105485233,
+      "grad_norm": 0.7343040841582446,
+      "learning_rate": 8.738595732154776e-06,
+      "loss": 0.0525,
+      "step": 2774
+    },
+    {
+      "epoch": 1.2325116588940705,
+      "grad_norm": 0.4304569080551511,
+      "learning_rate": 8.737308233490775e-06,
+      "loss": 0.0548,
+      "step": 2775
+    },
+    {
+      "epoch": 1.232955807239618,
+      "grad_norm": 0.5483425130829543,
+      "learning_rate": 8.736020173045858e-06,
+      "loss": 0.0547,
+      "step": 2776
+    },
+    {
+      "epoch": 1.2333999555851654,
+      "grad_norm": 0.45127669817592087,
+      "learning_rate": 8.734731551013648e-06,
+      "loss": 0.0458,
+      "step": 2777
+    },
+    {
+      "epoch": 1.2338441039307129,
+      "grad_norm": 0.5076179219488122,
+      "learning_rate": 8.733442367587842e-06,
+      "loss": 0.0495,
+      "step": 2778
+    },
+    {
+      "epoch": 1.2342882522762602,
+      "grad_norm": 0.4061357633325624,
+      "learning_rate": 8.732152622962229e-06,
+      "loss": 0.038,
+      "step": 2779
+    },
+    {
+      "epoch": 1.2347324006218077,
+      "grad_norm": 0.6061025558148905,
+      "learning_rate": 8.730862317330678e-06,
+      "loss": 0.0552,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2351765489673552,
+      "grad_norm": 0.7935186621762401,
+      "learning_rate": 8.729571450887145e-06,
+      "loss": 0.0497,
+      "step": 2781
+    },
+    {
+      "epoch": 1.2356206973129025,
+      "grad_norm": 0.4698985875136965,
+      "learning_rate": 8.728280023825667e-06,
+      "loss": 0.0449,
+      "step": 2782
+    },
+    {
+      "epoch": 1.23606484565845,
+      "grad_norm": 0.5845370985288897,
+      "learning_rate": 8.726988036340372e-06,
+      "loss": 0.0558,
+      "step": 2783
+    },
+    {
+      "epoch": 1.2365089940039973,
+      "grad_norm": 0.40081844727764104,
+      "learning_rate": 8.725695488625463e-06,
+      "loss": 0.0309,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2369531423495448,
+      "grad_norm": 0.565816281763518,
+      "learning_rate": 8.724402380875234e-06,
+      "loss": 0.0527,
+      "step": 2785
+    },
+    {
+      "epoch": 1.237397290695092,
+      "grad_norm": 0.4908483433280354,
+      "learning_rate": 8.72310871328406e-06,
+      "loss": 0.0406,
+      "step": 2786
+    },
+    {
+      "epoch": 1.2378414390406396,
+      "grad_norm": 0.5334199926157142,
+      "learning_rate": 8.7218144860464e-06,
+      "loss": 0.0432,
+      "step": 2787
+    },
+    {
+      "epoch": 1.238285587386187,
+      "grad_norm": 0.43486952273189144,
+      "learning_rate": 8.720519699356804e-06,
+      "loss": 0.0449,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2387297357317344,
+      "grad_norm": 0.5049916355349909,
+      "learning_rate": 8.719224353409895e-06,
+      "loss": 0.056,
+      "step": 2789
+    },
+    {
+      "epoch": 1.239173884077282,
+      "grad_norm": 0.44547159070564635,
+      "learning_rate": 8.717928448400387e-06,
+      "loss": 0.0444,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2396180324228292,
+      "grad_norm": 0.4328408997299911,
+      "learning_rate": 8.716631984523076e-06,
+      "loss": 0.0461,
+      "step": 2791
+    },
+    {
+      "epoch": 1.2400621807683767,
+      "grad_norm": 0.42920188220498545,
+      "learning_rate": 8.715334961972844e-06,
+      "loss": 0.0477,
+      "step": 2792
+    },
+    {
+      "epoch": 1.240506329113924,
+      "grad_norm": 0.45226945207712443,
+      "learning_rate": 8.714037380944655e-06,
+      "loss": 0.0445,
+      "step": 2793
+    },
+    {
+      "epoch": 1.2409504774594715,
+      "grad_norm": 0.6142043461351961,
+      "learning_rate": 8.712739241633557e-06,
+      "loss": 0.0526,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2413946258050188,
+      "grad_norm": 0.49524623051926364,
+      "learning_rate": 8.711440544234681e-06,
+      "loss": 0.0448,
+      "step": 2795
+    },
+    {
+      "epoch": 1.2418387741505663,
+      "grad_norm": 0.527589137144962,
+      "learning_rate": 8.710141288943247e-06,
+      "loss": 0.0681,
+      "step": 2796
+    },
+    {
+      "epoch": 1.2422829224961136,
+      "grad_norm": 0.45949156992557444,
+      "learning_rate": 8.708841475954551e-06,
+      "loss": 0.0597,
+      "step": 2797
+    },
+    {
+      "epoch": 1.2427270708416611,
+      "grad_norm": 0.5807053135952894,
+      "learning_rate": 8.707541105463982e-06,
+      "loss": 0.0602,
+      "step": 2798
+    },
+    {
+      "epoch": 1.2431712191872086,
+      "grad_norm": 0.47202242550842133,
+      "learning_rate": 8.706240177667003e-06,
+      "loss": 0.043,
+      "step": 2799
+    },
+    {
+      "epoch": 1.243615367532756,
+      "grad_norm": 0.5195621264387825,
+      "learning_rate": 8.704938692759166e-06,
+      "loss": 0.0384,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2440595158783034,
+      "grad_norm": 0.40308041507113895,
+      "learning_rate": 8.703636650936108e-06,
+      "loss": 0.0352,
+      "step": 2801
+    },
+    {
+      "epoch": 1.2445036642238507,
+      "grad_norm": 0.5663764923556683,
+      "learning_rate": 8.70233405239355e-06,
+      "loss": 0.0487,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2449478125693982,
+      "grad_norm": 0.6290221419877656,
+      "learning_rate": 8.70103089732729e-06,
+      "loss": 0.0518,
+      "step": 2803
+    },
+    {
+      "epoch": 1.2453919609149455,
+      "grad_norm": 0.4173665256358516,
+      "learning_rate": 8.699727185933215e-06,
+      "loss": 0.0352,
+      "step": 2804
+    },
+    {
+      "epoch": 1.245836109260493,
+      "grad_norm": 0.8204810767617772,
+      "learning_rate": 8.698422918407299e-06,
+      "loss": 0.0487,
+      "step": 2805
+    },
+    {
+      "epoch": 1.2462802576060403,
+      "grad_norm": 0.9850006429206544,
+      "learning_rate": 8.697118094945593e-06,
+      "loss": 0.0865,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2467244059515878,
+      "grad_norm": 0.48735159482648766,
+      "learning_rate": 8.695812715744235e-06,
+      "loss": 0.0358,
+      "step": 2807
+    },
+    {
+      "epoch": 1.2471685542971351,
+      "grad_norm": 0.5010917630261117,
+      "learning_rate": 8.694506780999444e-06,
+      "loss": 0.0477,
+      "step": 2808
+    },
+    {
+      "epoch": 1.2476127026426826,
+      "grad_norm": 0.4869736837475569,
+      "learning_rate": 8.693200290907525e-06,
+      "loss": 0.0484,
+      "step": 2809
+    },
+    {
+      "epoch": 1.2480568509882302,
+      "grad_norm": 0.3824351180128345,
+      "learning_rate": 8.691893245664867e-06,
+      "loss": 0.0352,
+      "step": 2810
+    },
+    {
+      "epoch": 1.2485009993337775,
+      "grad_norm": 0.5417070708808928,
+      "learning_rate": 8.690585645467937e-06,
+      "loss": 0.0447,
+      "step": 2811
+    },
+    {
+      "epoch": 1.248945147679325,
+      "grad_norm": 0.5086804337705053,
+      "learning_rate": 8.689277490513295e-06,
+      "loss": 0.05,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2493892960248723,
+      "grad_norm": 0.4431550852814911,
+      "learning_rate": 8.687968780997576e-06,
+      "loss": 0.0635,
+      "step": 2813
+    },
+    {
+      "epoch": 1.2498334443704198,
+      "grad_norm": 0.3783726536150246,
+      "learning_rate": 8.686659517117501e-06,
+      "loss": 0.0384,
+      "step": 2814
+    },
+    {
+      "epoch": 1.250277592715967,
+      "grad_norm": 0.5331217000478606,
+      "learning_rate": 8.685349699069875e-06,
+      "loss": 0.0409,
+      "step": 2815
+    },
+    {
+      "epoch": 1.2507217410615146,
+      "grad_norm": 0.38074858661235217,
+      "learning_rate": 8.684039327051586e-06,
+      "loss": 0.038,
+      "step": 2816
+    },
+    {
+      "epoch": 1.251165889407062,
+      "grad_norm": 0.6121662418691846,
+      "learning_rate": 8.682728401259606e-06,
+      "loss": 0.0556,
+      "step": 2817
+    },
+    {
+      "epoch": 1.2516100377526094,
+      "grad_norm": 0.8655438293508221,
+      "learning_rate": 8.681416921890988e-06,
+      "loss": 0.0559,
+      "step": 2818
+    },
+    {
+      "epoch": 1.2520541860981567,
+      "grad_norm": 0.5646419752199546,
+      "learning_rate": 8.680104889142871e-06,
+      "loss": 0.0546,
+      "step": 2819
+    },
+    {
+      "epoch": 1.2524983344437042,
+      "grad_norm": 0.47189997209382956,
+      "learning_rate": 8.678792303212474e-06,
+      "loss": 0.038,
+      "step": 2820
+    },
+    {
+      "epoch": 1.2529424827892517,
+      "grad_norm": 0.5347473900252707,
+      "learning_rate": 8.677479164297102e-06,
+      "loss": 0.0506,
+      "step": 2821
+    },
+    {
+      "epoch": 1.253386631134799,
+      "grad_norm": 0.6752833223029766,
+      "learning_rate": 8.676165472594145e-06,
+      "loss": 0.0553,
+      "step": 2822
+    },
+    {
+      "epoch": 1.2538307794803465,
+      "grad_norm": 0.6290436949126601,
+      "learning_rate": 8.674851228301066e-06,
+      "loss": 0.0628,
+      "step": 2823
+    },
+    {
+      "epoch": 1.2542749278258938,
+      "grad_norm": 0.5190239487706246,
+      "learning_rate": 8.673536431615426e-06,
+      "loss": 0.053,
+      "step": 2824
+    },
+    {
+      "epoch": 1.2547190761714413,
+      "grad_norm": 0.5039759155119382,
+      "learning_rate": 8.672221082734857e-06,
+      "loss": 0.0538,
+      "step": 2825
+    },
+    {
+      "epoch": 1.2551632245169886,
+      "grad_norm": 0.5007524426018746,
+      "learning_rate": 8.670905181857078e-06,
+      "loss": 0.0379,
+      "step": 2826
+    },
+    {
+      "epoch": 1.255607372862536,
+      "grad_norm": 0.4471624712255033,
+      "learning_rate": 8.669588729179895e-06,
+      "loss": 0.0425,
+      "step": 2827
+    },
+    {
+      "epoch": 1.2560515212080836,
+      "grad_norm": 0.6483988178781206,
+      "learning_rate": 8.668271724901188e-06,
+      "loss": 0.0667,
+      "step": 2828
+    },
+    {
+      "epoch": 1.256495669553631,
+      "grad_norm": 0.7074590212609329,
+      "learning_rate": 8.666954169218929e-06,
+      "loss": 0.0644,
+      "step": 2829
+    },
+    {
+      "epoch": 1.2569398178991782,
+      "grad_norm": 0.4799067719368653,
+      "learning_rate": 8.665636062331166e-06,
+      "loss": 0.0516,
+      "step": 2830
+    },
+    {
+      "epoch": 1.2573839662447257,
+      "grad_norm": 0.4519642992082907,
+      "learning_rate": 8.664317404436036e-06,
+      "loss": 0.0388,
+      "step": 2831
+    },
+    {
+      "epoch": 1.2578281145902732,
+      "grad_norm": 1.9317339882104076,
+      "learning_rate": 8.662998195731755e-06,
+      "loss": 0.0445,
+      "step": 2832
+    },
+    {
+      "epoch": 1.2582722629358205,
+      "grad_norm": 0.9801917673057263,
+      "learning_rate": 8.661678436416621e-06,
+      "loss": 0.0609,
+      "step": 2833
+    },
+    {
+      "epoch": 1.258716411281368,
+      "grad_norm": 0.4629508087617577,
+      "learning_rate": 8.660358126689015e-06,
+      "loss": 0.046,
+      "step": 2834
+    },
+    {
+      "epoch": 1.2591605596269153,
+      "grad_norm": 0.55871143511484,
+      "learning_rate": 8.659037266747405e-06,
+      "loss": 0.0669,
+      "step": 2835
+    },
+    {
+      "epoch": 1.2596047079724628,
+      "grad_norm": 0.4373848923476879,
+      "learning_rate": 8.65771585679034e-06,
+      "loss": 0.0365,
+      "step": 2836
+    },
+    {
+      "epoch": 1.2600488563180101,
+      "grad_norm": 0.4620726535649706,
+      "learning_rate": 8.656393897016446e-06,
+      "loss": 0.0508,
+      "step": 2837
+    },
+    {
+      "epoch": 1.2604930046635576,
+      "grad_norm": 0.7076034863902437,
+      "learning_rate": 8.655071387624439e-06,
+      "loss": 0.066,
+      "step": 2838
+    },
+    {
+      "epoch": 1.2609371530091051,
+      "grad_norm": 0.7170929705949399,
+      "learning_rate": 8.653748328813112e-06,
+      "loss": 0.0663,
+      "step": 2839
+    },
+    {
+      "epoch": 1.2613813013546524,
+      "grad_norm": 0.5278820360047365,
+      "learning_rate": 8.652424720781346e-06,
+      "loss": 0.0703,
+      "step": 2840
+    },
+    {
+      "epoch": 1.2618254497002,
+      "grad_norm": 0.6581330490442558,
+      "learning_rate": 8.6511005637281e-06,
+      "loss": 0.0475,
+      "step": 2841
+    },
+    {
+      "epoch": 1.2622695980457472,
+      "grad_norm": 0.5785660781607422,
+      "learning_rate": 8.649775857852419e-06,
+      "loss": 0.0539,
+      "step": 2842
+    },
+    {
+      "epoch": 1.2627137463912947,
+      "grad_norm": 0.7193486412721348,
+      "learning_rate": 8.648450603353427e-06,
+      "loss": 0.0527,
+      "step": 2843
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.4879073594292654,
+      "learning_rate": 8.647124800430332e-06,
+      "loss": 0.0462,
+      "step": 2844
+    },
+    {
+      "epoch": 1.2636020430823895,
+      "grad_norm": 0.6283549482710774,
+      "learning_rate": 8.645798449282427e-06,
+      "loss": 0.047,
+      "step": 2845
+    },
+    {
+      "epoch": 1.264046191427937,
+      "grad_norm": 0.590026590546146,
+      "learning_rate": 8.644471550109084e-06,
+      "loss": 0.0417,
+      "step": 2846
+    },
+    {
+      "epoch": 1.2644903397734844,
+      "grad_norm": 0.5727280121631517,
+      "learning_rate": 8.643144103109757e-06,
+      "loss": 0.0461,
+      "step": 2847
+    },
+    {
+      "epoch": 1.2649344881190316,
+      "grad_norm": 0.5188962100480214,
+      "learning_rate": 8.641816108483987e-06,
+      "loss": 0.0528,
+      "step": 2848
+    },
+    {
+      "epoch": 1.2653786364645792,
+      "grad_norm": 0.6762837917534186,
+      "learning_rate": 8.64048756643139e-06,
+      "loss": 0.0551,
+      "step": 2849
+    },
+    {
+      "epoch": 1.2658227848101267,
+      "grad_norm": 0.6063186060719946,
+      "learning_rate": 8.639158477151673e-06,
+      "loss": 0.0529,
+      "step": 2850
+    },
+    {
+      "epoch": 1.266266933155674,
+      "grad_norm": 0.8416294217857067,
+      "learning_rate": 8.637828840844615e-06,
+      "loss": 0.0605,
+      "step": 2851
+    },
+    {
+      "epoch": 1.2667110815012215,
+      "grad_norm": 0.42460216144695934,
+      "learning_rate": 8.636498657710091e-06,
+      "loss": 0.0406,
+      "step": 2852
+    },
+    {
+      "epoch": 1.2671552298467688,
+      "grad_norm": 0.45348809902081694,
+      "learning_rate": 8.635167927948041e-06,
+      "loss": 0.0423,
+      "step": 2853
+    },
+    {
+      "epoch": 1.2675993781923163,
+      "grad_norm": 0.5925945869143667,
+      "learning_rate": 8.633836651758502e-06,
+      "loss": 0.0422,
+      "step": 2854
+    },
+    {
+      "epoch": 1.2680435265378636,
+      "grad_norm": 0.474702929089884,
+      "learning_rate": 8.632504829341588e-06,
+      "loss": 0.0365,
+      "step": 2855
+    },
+    {
+      "epoch": 1.268487674883411,
+      "grad_norm": 0.8276698250673459,
+      "learning_rate": 8.63117246089749e-06,
+      "loss": 0.0477,
+      "step": 2856
+    },
+    {
+      "epoch": 1.2689318232289586,
+      "grad_norm": 0.4891929607601022,
+      "learning_rate": 8.62983954662649e-06,
+      "loss": 0.044,
+      "step": 2857
+    },
+    {
+      "epoch": 1.2693759715745059,
+      "grad_norm": 0.6525774912230711,
+      "learning_rate": 8.628506086728947e-06,
+      "loss": 0.0651,
+      "step": 2858
+    },
+    {
+      "epoch": 1.2698201199200532,
+      "grad_norm": 0.6594678233990844,
+      "learning_rate": 8.6271720814053e-06,
+      "loss": 0.0381,
+      "step": 2859
+    },
+    {
+      "epoch": 1.2702642682656007,
+      "grad_norm": 0.3241646392485162,
+      "learning_rate": 8.625837530856074e-06,
+      "loss": 0.0309,
+      "step": 2860
+    },
+    {
+      "epoch": 1.2707084166111482,
+      "grad_norm": 0.7122663241070569,
+      "learning_rate": 8.624502435281875e-06,
+      "loss": 0.0617,
+      "step": 2861
+    },
+    {
+      "epoch": 1.2711525649566955,
+      "grad_norm": 0.7005320306757822,
+      "learning_rate": 8.623166794883393e-06,
+      "loss": 0.0581,
+      "step": 2862
+    },
+    {
+      "epoch": 1.271596713302243,
+      "grad_norm": 0.5328754180731429,
+      "learning_rate": 8.621830609861392e-06,
+      "loss": 0.0493,
+      "step": 2863
+    },
+    {
+      "epoch": 1.2720408616477903,
+      "grad_norm": 0.6269919214433372,
+      "learning_rate": 8.620493880416727e-06,
+      "loss": 0.0516,
+      "step": 2864
+    },
+    {
+      "epoch": 1.2724850099933378,
+      "grad_norm": 0.5844430878482636,
+      "learning_rate": 8.619156606750329e-06,
+      "loss": 0.0725,
+      "step": 2865
+    },
+    {
+      "epoch": 1.272929158338885,
+      "grad_norm": 0.4737556741912441,
+      "learning_rate": 8.617818789063217e-06,
+      "loss": 0.0421,
+      "step": 2866
+    },
+    {
+      "epoch": 1.2733733066844326,
+      "grad_norm": 0.7524484026464038,
+      "learning_rate": 8.616480427556484e-06,
+      "loss": 0.0708,
+      "step": 2867
+    },
+    {
+      "epoch": 1.2738174550299801,
+      "grad_norm": 0.8150405412731492,
+      "learning_rate": 8.61514152243131e-06,
+      "loss": 0.0704,
+      "step": 2868
+    },
+    {
+      "epoch": 1.2742616033755274,
+      "grad_norm": 0.6289008429543912,
+      "learning_rate": 8.613802073888953e-06,
+      "loss": 0.0544,
+      "step": 2869
+    },
+    {
+      "epoch": 1.274705751721075,
+      "grad_norm": 0.49267955313551814,
+      "learning_rate": 8.612462082130758e-06,
+      "loss": 0.0423,
+      "step": 2870
+    },
+    {
+      "epoch": 1.2751499000666222,
+      "grad_norm": 0.8023016461317273,
+      "learning_rate": 8.611121547358146e-06,
+      "loss": 0.0581,
+      "step": 2871
+    },
+    {
+      "epoch": 1.2755940484121697,
+      "grad_norm": 0.5971368904341509,
+      "learning_rate": 8.609780469772623e-06,
+      "loss": 0.0555,
+      "step": 2872
+    },
+    {
+      "epoch": 1.276038196757717,
+      "grad_norm": 0.497140537686175,
+      "learning_rate": 8.608438849575777e-06,
+      "loss": 0.0511,
+      "step": 2873
+    },
+    {
+      "epoch": 1.2764823451032645,
+      "grad_norm": 0.5696597646442083,
+      "learning_rate": 8.607096686969274e-06,
+      "loss": 0.0592,
+      "step": 2874
+    },
+    {
+      "epoch": 1.276926493448812,
+      "grad_norm": 0.7494228951484337,
+      "learning_rate": 8.605753982154865e-06,
+      "loss": 0.0486,
+      "step": 2875
+    },
+    {
+      "epoch": 1.2773706417943593,
+      "grad_norm": 0.5764143358267766,
+      "learning_rate": 8.604410735334383e-06,
+      "loss": 0.0551,
+      "step": 2876
+    },
+    {
+      "epoch": 1.2778147901399066,
+      "grad_norm": 0.6011073566193142,
+      "learning_rate": 8.603066946709739e-06,
+      "loss": 0.0567,
+      "step": 2877
+    },
+    {
+      "epoch": 1.2782589384854541,
+      "grad_norm": 0.44537454001664567,
+      "learning_rate": 8.601722616482927e-06,
+      "loss": 0.0473,
+      "step": 2878
+    },
+    {
+      "epoch": 1.2787030868310016,
+      "grad_norm": 0.45558471841390763,
+      "learning_rate": 8.600377744856024e-06,
+      "loss": 0.0577,
+      "step": 2879
+    },
+    {
+      "epoch": 1.279147235176549,
+      "grad_norm": 0.5192384143571988,
+      "learning_rate": 8.599032332031185e-06,
+      "loss": 0.0549,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2795913835220964,
+      "grad_norm": 0.641251450966757,
+      "learning_rate": 8.59768637821065e-06,
+      "loss": 0.0599,
+      "step": 2881
+    },
+    {
+      "epoch": 1.2800355318676437,
+      "grad_norm": 0.5747913363539975,
+      "learning_rate": 8.596339883596738e-06,
+      "loss": 0.0546,
+      "step": 2882
+    },
+    {
+      "epoch": 1.2804796802131913,
+      "grad_norm": 0.5416056259419014,
+      "learning_rate": 8.594992848391852e-06,
+      "loss": 0.0536,
+      "step": 2883
+    },
+    {
+      "epoch": 1.2809238285587385,
+      "grad_norm": 0.6028308551659004,
+      "learning_rate": 8.59364527279847e-06,
+      "loss": 0.0699,
+      "step": 2884
+    },
+    {
+      "epoch": 1.281367976904286,
+      "grad_norm": 0.4368044045668833,
+      "learning_rate": 8.59229715701916e-06,
+      "loss": 0.0396,
+      "step": 2885
+    },
+    {
+      "epoch": 1.2818121252498336,
+      "grad_norm": 0.5104164604694876,
+      "learning_rate": 8.590948501256564e-06,
+      "loss": 0.04,
+      "step": 2886
+    },
+    {
+      "epoch": 1.2822562735953809,
+      "grad_norm": 0.7612641229651903,
+      "learning_rate": 8.58959930571341e-06,
+      "loss": 0.0624,
+      "step": 2887
+    },
+    {
+      "epoch": 1.2827004219409281,
+      "grad_norm": 0.5035114097597291,
+      "learning_rate": 8.588249570592502e-06,
+      "loss": 0.0437,
+      "step": 2888
+    },
+    {
+      "epoch": 1.2831445702864757,
+      "grad_norm": 0.499768070561634,
+      "learning_rate": 8.586899296096731e-06,
+      "loss": 0.0378,
+      "step": 2889
+    },
+    {
+      "epoch": 1.2835887186320232,
+      "grad_norm": 0.4925272487243439,
+      "learning_rate": 8.585548482429064e-06,
+      "loss": 0.0406,
+      "step": 2890
+    },
+    {
+      "epoch": 1.2840328669775705,
+      "grad_norm": 0.5079455242566149,
+      "learning_rate": 8.584197129792553e-06,
+      "loss": 0.049,
+      "step": 2891
+    },
+    {
+      "epoch": 1.284477015323118,
+      "grad_norm": 0.5999702910432165,
+      "learning_rate": 8.58284523839033e-06,
+      "loss": 0.0525,
+      "step": 2892
+    },
+    {
+      "epoch": 1.2849211636686653,
+      "grad_norm": 0.5762429387843919,
+      "learning_rate": 8.581492808425604e-06,
+      "loss": 0.0412,
+      "step": 2893
+    },
+    {
+      "epoch": 1.2853653120142128,
+      "grad_norm": 0.5170549640705353,
+      "learning_rate": 8.58013984010167e-06,
+      "loss": 0.0426,
+      "step": 2894
+    },
+    {
+      "epoch": 1.28580946035976,
+      "grad_norm": 0.44644738551782165,
+      "learning_rate": 8.578786333621902e-06,
+      "loss": 0.0449,
+      "step": 2895
+    },
+    {
+      "epoch": 1.2862536087053076,
+      "grad_norm": 0.49787403784527023,
+      "learning_rate": 8.577432289189755e-06,
+      "loss": 0.053,
+      "step": 2896
+    },
+    {
+      "epoch": 1.286697757050855,
+      "grad_norm": 0.4201343576902461,
+      "learning_rate": 8.576077707008766e-06,
+      "loss": 0.0548,
+      "step": 2897
+    },
+    {
+      "epoch": 1.2871419053964024,
+      "grad_norm": 0.43223595023539274,
+      "learning_rate": 8.57472258728255e-06,
+      "loss": 0.0414,
+      "step": 2898
+    },
+    {
+      "epoch": 1.2875860537419497,
+      "grad_norm": 0.45140599595731223,
+      "learning_rate": 8.573366930214807e-06,
+      "loss": 0.0512,
+      "step": 2899
+    },
+    {
+      "epoch": 1.2880302020874972,
+      "grad_norm": 0.7864322313017508,
+      "learning_rate": 8.57201073600931e-06,
+      "loss": 0.0611,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2884743504330447,
+      "grad_norm": 0.4533939072001686,
+      "learning_rate": 8.570654004869924e-06,
+      "loss": 0.0455,
+      "step": 2901
+    },
+    {
+      "epoch": 1.288918498778592,
+      "grad_norm": 0.5822538472138225,
+      "learning_rate": 8.569296737000586e-06,
+      "loss": 0.0689,
+      "step": 2902
+    },
+    {
+      "epoch": 1.2893626471241395,
+      "grad_norm": 0.6558805640817998,
+      "learning_rate": 8.567938932605315e-06,
+      "loss": 0.0716,
+      "step": 2903
+    },
+    {
+      "epoch": 1.2898067954696868,
+      "grad_norm": 0.6962929137245736,
+      "learning_rate": 8.566580591888216e-06,
+      "loss": 0.0456,
+      "step": 2904
+    },
+    {
+      "epoch": 1.2902509438152343,
+      "grad_norm": 0.5885420821356683,
+      "learning_rate": 8.565221715053467e-06,
+      "loss": 0.0519,
+      "step": 2905
+    },
+    {
+      "epoch": 1.2906950921607816,
+      "grad_norm": 0.5367742959382944,
+      "learning_rate": 8.563862302305333e-06,
+      "loss": 0.0385,
+      "step": 2906
+    },
+    {
+      "epoch": 1.2911392405063291,
+      "grad_norm": 0.7323581959894143,
+      "learning_rate": 8.562502353848155e-06,
+      "loss": 0.0659,
+      "step": 2907
+    },
+    {
+      "epoch": 1.2915833888518766,
+      "grad_norm": 0.5051887556869058,
+      "learning_rate": 8.561141869886356e-06,
+      "loss": 0.0559,
+      "step": 2908
+    },
+    {
+      "epoch": 1.292027537197424,
+      "grad_norm": 0.5491304399843213,
+      "learning_rate": 8.55978085062444e-06,
+      "loss": 0.0639,
+      "step": 2909
+    },
+    {
+      "epoch": 1.2924716855429714,
+      "grad_norm": 0.8492250674519668,
+      "learning_rate": 8.558419296266995e-06,
+      "loss": 0.0636,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2929158338885187,
+      "grad_norm": 0.3882329610931263,
+      "learning_rate": 8.557057207018681e-06,
+      "loss": 0.0487,
+      "step": 2911
+    },
+    {
+      "epoch": 1.2933599822340662,
+      "grad_norm": 0.625233318674613,
+      "learning_rate": 8.555694583084244e-06,
+      "loss": 0.0465,
+      "step": 2912
+    },
+    {
+      "epoch": 1.2938041305796135,
+      "grad_norm": 0.4026058166802596,
+      "learning_rate": 8.554331424668511e-06,
+      "loss": 0.0354,
+      "step": 2913
+    },
+    {
+      "epoch": 1.294248278925161,
+      "grad_norm": 0.5634837852049973,
+      "learning_rate": 8.552967731976388e-06,
+      "loss": 0.0541,
+      "step": 2914
+    },
+    {
+      "epoch": 1.2946924272707085,
+      "grad_norm": 0.48751917705852116,
+      "learning_rate": 8.551603505212862e-06,
+      "loss": 0.0322,
+      "step": 2915
+    },
+    {
+      "epoch": 1.2951365756162558,
+      "grad_norm": 0.5156212240116493,
+      "learning_rate": 8.550238744582997e-06,
+      "loss": 0.0489,
+      "step": 2916
+    },
+    {
+      "epoch": 1.2955807239618031,
+      "grad_norm": 0.6223211171347353,
+      "learning_rate": 8.548873450291939e-06,
+      "loss": 0.0457,
+      "step": 2917
+    },
+    {
+      "epoch": 1.2960248723073506,
+      "grad_norm": 0.47672895894969763,
+      "learning_rate": 8.547507622544916e-06,
+      "loss": 0.0463,
+      "step": 2918
+    },
+    {
+      "epoch": 1.2964690206528982,
+      "grad_norm": 0.4421458887458943,
+      "learning_rate": 8.546141261547238e-06,
+      "loss": 0.045,
+      "step": 2919
+    },
+    {
+      "epoch": 1.2969131689984454,
+      "grad_norm": 0.848623824582431,
+      "learning_rate": 8.544774367504291e-06,
+      "loss": 0.0605,
+      "step": 2920
+    },
+    {
+      "epoch": 1.297357317343993,
+      "grad_norm": 0.554513822402812,
+      "learning_rate": 8.54340694062154e-06,
+      "loss": 0.0518,
+      "step": 2921
+    },
+    {
+      "epoch": 1.2978014656895402,
+      "grad_norm": 0.5649747315400963,
+      "learning_rate": 8.542038981104532e-06,
+      "loss": 0.0565,
+      "step": 2922
+    },
+    {
+      "epoch": 1.2982456140350878,
+      "grad_norm": 0.42021656791448186,
+      "learning_rate": 8.540670489158899e-06,
+      "loss": 0.0426,
+      "step": 2923
+    },
+    {
+      "epoch": 1.298689762380635,
+      "grad_norm": 0.6402654468272049,
+      "learning_rate": 8.539301464990345e-06,
+      "loss": 0.0543,
+      "step": 2924
+    },
+    {
+      "epoch": 1.2991339107261826,
+      "grad_norm": 0.3618130143994902,
+      "learning_rate": 8.53793190880466e-06,
+      "loss": 0.038,
+      "step": 2925
+    },
+    {
+      "epoch": 1.29957805907173,
+      "grad_norm": 0.3384505689514398,
+      "learning_rate": 8.536561820807707e-06,
+      "loss": 0.0346,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3000222074172774,
+      "grad_norm": 0.46188012553966973,
+      "learning_rate": 8.535191201205439e-06,
+      "loss": 0.0514,
+      "step": 2927
+    },
+    {
+      "epoch": 1.3004663557628247,
+      "grad_norm": 0.5300445785864025,
+      "learning_rate": 8.533820050203881e-06,
+      "loss": 0.0483,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3009105041083722,
+      "grad_norm": 0.5536409349327417,
+      "learning_rate": 8.532448368009139e-06,
+      "loss": 0.0489,
+      "step": 2929
+    },
+    {
+      "epoch": 1.3013546524539197,
+      "grad_norm": 0.371658656980648,
+      "learning_rate": 8.531076154827402e-06,
+      "loss": 0.0445,
+      "step": 2930
+    },
+    {
+      "epoch": 1.301798800799467,
+      "grad_norm": 0.5129545567634131,
+      "learning_rate": 8.529703410864938e-06,
+      "loss": 0.0341,
+      "step": 2931
+    },
+    {
+      "epoch": 1.3022429491450145,
+      "grad_norm": 0.6790917670823785,
+      "learning_rate": 8.52833013632809e-06,
+      "loss": 0.0604,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3026870974905618,
+      "grad_norm": 0.4758905378740349,
+      "learning_rate": 8.526956331423289e-06,
+      "loss": 0.0458,
+      "step": 2933
+    },
+    {
+      "epoch": 1.3031312458361093,
+      "grad_norm": 0.5428940197749453,
+      "learning_rate": 8.525581996357036e-06,
+      "loss": 0.047,
+      "step": 2934
+    },
+    {
+      "epoch": 1.3035753941816566,
+      "grad_norm": 0.5546360540394489,
+      "learning_rate": 8.52420713133592e-06,
+      "loss": 0.0585,
+      "step": 2935
+    },
+    {
+      "epoch": 1.304019542527204,
+      "grad_norm": 0.8256368560390267,
+      "learning_rate": 8.522831736566607e-06,
+      "loss": 0.0592,
+      "step": 2936
+    },
+    {
+      "epoch": 1.3044636908727516,
+      "grad_norm": 0.6807172008845478,
+      "learning_rate": 8.521455812255843e-06,
+      "loss": 0.0727,
+      "step": 2937
+    },
+    {
+      "epoch": 1.304907839218299,
+      "grad_norm": 0.5972239265610148,
+      "learning_rate": 8.52007935861045e-06,
+      "loss": 0.0531,
+      "step": 2938
+    },
+    {
+      "epoch": 1.3053519875638464,
+      "grad_norm": 0.5756047785651232,
+      "learning_rate": 8.518702375837335e-06,
+      "loss": 0.0484,
+      "step": 2939
+    },
+    {
+      "epoch": 1.3057961359093937,
+      "grad_norm": 0.5254079002689842,
+      "learning_rate": 8.51732486414348e-06,
+      "loss": 0.0551,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3062402842549412,
+      "grad_norm": 0.5057686654115676,
+      "learning_rate": 8.515946823735948e-06,
+      "loss": 0.0555,
+      "step": 2941
+    },
+    {
+      "epoch": 1.3066844326004885,
+      "grad_norm": 0.5425819759814885,
+      "learning_rate": 8.514568254821884e-06,
+      "loss": 0.0461,
+      "step": 2942
+    },
+    {
+      "epoch": 1.307128580946036,
+      "grad_norm": 0.4919842279881122,
+      "learning_rate": 8.51318915760851e-06,
+      "loss": 0.0351,
+      "step": 2943
+    },
+    {
+      "epoch": 1.3075727292915835,
+      "grad_norm": 0.5651189893610779,
+      "learning_rate": 8.511809532303126e-06,
+      "loss": 0.0485,
+      "step": 2944
+    },
+    {
+      "epoch": 1.3080168776371308,
+      "grad_norm": 0.5135664399251805,
+      "learning_rate": 8.510429379113114e-06,
+      "loss": 0.0345,
+      "step": 2945
+    },
+    {
+      "epoch": 1.308461025982678,
+      "grad_norm": 0.698576525158577,
+      "learning_rate": 8.509048698245934e-06,
+      "loss": 0.0541,
+      "step": 2946
+    },
+    {
+      "epoch": 1.3089051743282256,
+      "grad_norm": 0.5914910850001974,
+      "learning_rate": 8.507667489909126e-06,
+      "loss": 0.0422,
+      "step": 2947
+    },
+    {
+      "epoch": 1.3093493226737731,
+      "grad_norm": 0.39464116666974663,
+      "learning_rate": 8.506285754310311e-06,
+      "loss": 0.0395,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3097934710193204,
+      "grad_norm": 0.41501372577846396,
+      "learning_rate": 8.504903491657185e-06,
+      "loss": 0.0392,
+      "step": 2949
+    },
+    {
+      "epoch": 1.310237619364868,
+      "grad_norm": 0.5710008833310989,
+      "learning_rate": 8.503520702157527e-06,
+      "loss": 0.0486,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3106817677104152,
+      "grad_norm": 0.912615051335769,
+      "learning_rate": 8.502137386019191e-06,
+      "loss": 0.0636,
+      "step": 2951
+    },
+    {
+      "epoch": 1.3111259160559627,
+      "grad_norm": 0.543540441011416,
+      "learning_rate": 8.500753543450118e-06,
+      "loss": 0.062,
+      "step": 2952
+    },
+    {
+      "epoch": 1.31157006440151,
+      "grad_norm": 0.39754298101965646,
+      "learning_rate": 8.499369174658318e-06,
+      "loss": 0.0383,
+      "step": 2953
+    },
+    {
+      "epoch": 1.3120142127470575,
+      "grad_norm": 0.5539227325166909,
+      "learning_rate": 8.497984279851888e-06,
+      "loss": 0.0499,
+      "step": 2954
+    },
+    {
+      "epoch": 1.312458361092605,
+      "grad_norm": 0.599023615011808,
+      "learning_rate": 8.496598859238997e-06,
+      "loss": 0.0472,
+      "step": 2955
+    },
+    {
+      "epoch": 1.3129025094381523,
+      "grad_norm": 0.4220558992968304,
+      "learning_rate": 8.495212913027906e-06,
+      "loss": 0.0446,
+      "step": 2956
+    },
+    {
+      "epoch": 1.3133466577836996,
+      "grad_norm": 0.49593231969327095,
+      "learning_rate": 8.493826441426937e-06,
+      "loss": 0.0636,
+      "step": 2957
+    },
+    {
+      "epoch": 1.3137908061292471,
+      "grad_norm": 0.5382980257808322,
+      "learning_rate": 8.492439444644506e-06,
+      "loss": 0.0462,
+      "step": 2958
+    },
+    {
+      "epoch": 1.3142349544747947,
+      "grad_norm": 0.4925896725925621,
+      "learning_rate": 8.4910519228891e-06,
+      "loss": 0.0602,
+      "step": 2959
+    },
+    {
+      "epoch": 1.314679102820342,
+      "grad_norm": 1.1877590532157762,
+      "learning_rate": 8.489663876369288e-06,
+      "loss": 0.0723,
+      "step": 2960
+    },
+    {
+      "epoch": 1.3151232511658895,
+      "grad_norm": 0.6920312294903902,
+      "learning_rate": 8.488275305293715e-06,
+      "loss": 0.0531,
+      "step": 2961
+    },
+    {
+      "epoch": 1.3155673995114368,
+      "grad_norm": 0.36761567736761985,
+      "learning_rate": 8.486886209871108e-06,
+      "loss": 0.0397,
+      "step": 2962
+    },
+    {
+      "epoch": 1.3160115478569843,
+      "grad_norm": 0.55057942024177,
+      "learning_rate": 8.485496590310274e-06,
+      "loss": 0.0591,
+      "step": 2963
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.5710852175463684,
+      "learning_rate": 8.484106446820094e-06,
+      "loss": 0.0471,
+      "step": 2964
+    },
+    {
+      "epoch": 1.316899844548079,
+      "grad_norm": 0.5828688111433763,
+      "learning_rate": 8.482715779609526e-06,
+      "loss": 0.0551,
+      "step": 2965
+    },
+    {
+      "epoch": 1.3173439928936266,
+      "grad_norm": 0.531963458189058,
+      "learning_rate": 8.481324588887619e-06,
+      "loss": 0.0504,
+      "step": 2966
+    },
+    {
+      "epoch": 1.3177881412391739,
+      "grad_norm": 0.9382025539874739,
+      "learning_rate": 8.47993287486349e-06,
+      "loss": 0.0778,
+      "step": 2967
+    },
+    {
+      "epoch": 1.3182322895847212,
+      "grad_norm": 0.5014993856992925,
+      "learning_rate": 8.478540637746334e-06,
+      "loss": 0.0635,
+      "step": 2968
+    },
+    {
+      "epoch": 1.3186764379302687,
+      "grad_norm": 0.396172698958811,
+      "learning_rate": 8.477147877745431e-06,
+      "loss": 0.0378,
+      "step": 2969
+    },
+    {
+      "epoch": 1.3191205862758162,
+      "grad_norm": 0.5285271119946755,
+      "learning_rate": 8.475754595070134e-06,
+      "loss": 0.0544,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3195647346213635,
+      "grad_norm": 0.4032933677589254,
+      "learning_rate": 8.474360789929881e-06,
+      "loss": 0.0431,
+      "step": 2971
+    },
+    {
+      "epoch": 1.320008882966911,
+      "grad_norm": 0.45143103762565806,
+      "learning_rate": 8.47296646253418e-06,
+      "loss": 0.0509,
+      "step": 2972
+    },
+    {
+      "epoch": 1.3204530313124583,
+      "grad_norm": 0.5252154450521924,
+      "learning_rate": 8.471571613092626e-06,
+      "loss": 0.0594,
+      "step": 2973
+    },
+    {
+      "epoch": 1.3208971796580058,
+      "grad_norm": 0.34642633471428613,
+      "learning_rate": 8.470176241814886e-06,
+      "loss": 0.027,
+      "step": 2974
+    },
+    {
+      "epoch": 1.321341328003553,
+      "grad_norm": 0.4784318360661084,
+      "learning_rate": 8.46878034891071e-06,
+      "loss": 0.0444,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3217854763491006,
+      "grad_norm": 0.580781905109618,
+      "learning_rate": 8.467383934589923e-06,
+      "loss": 0.043,
+      "step": 2976
+    },
+    {
+      "epoch": 1.322229624694648,
+      "grad_norm": 0.46709485088977415,
+      "learning_rate": 8.465986999062427e-06,
+      "loss": 0.0485,
+      "step": 2977
+    },
+    {
+      "epoch": 1.3226737730401954,
+      "grad_norm": 0.7439128060207424,
+      "learning_rate": 8.464589542538213e-06,
+      "loss": 0.0566,
+      "step": 2978
+    },
+    {
+      "epoch": 1.323117921385743,
+      "grad_norm": 0.5532502253043584,
+      "learning_rate": 8.463191565227336e-06,
+      "loss": 0.0486,
+      "step": 2979
+    },
+    {
+      "epoch": 1.3235620697312902,
+      "grad_norm": 0.46900654723685586,
+      "learning_rate": 8.461793067339936e-06,
+      "loss": 0.0384,
+      "step": 2980
+    },
+    {
+      "epoch": 1.3240062180768377,
+      "grad_norm": 0.4715987465734639,
+      "learning_rate": 8.460394049086232e-06,
+      "loss": 0.0467,
+      "step": 2981
+    },
+    {
+      "epoch": 1.324450366422385,
+      "grad_norm": 0.4464995773383881,
+      "learning_rate": 8.458994510676523e-06,
+      "loss": 0.0392,
+      "step": 2982
+    },
+    {
+      "epoch": 1.3248945147679325,
+      "grad_norm": 0.5120257592364733,
+      "learning_rate": 8.457594452321178e-06,
+      "loss": 0.0468,
+      "step": 2983
+    },
+    {
+      "epoch": 1.32533866311348,
+      "grad_norm": 0.5011205376177407,
+      "learning_rate": 8.456193874230656e-06,
+      "loss": 0.0479,
+      "step": 2984
+    },
+    {
+      "epoch": 1.3257828114590273,
+      "grad_norm": 0.4942109449168538,
+      "learning_rate": 8.454792776615482e-06,
+      "loss": 0.0456,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3262269598045746,
+      "grad_norm": 0.5223005309985502,
+      "learning_rate": 8.453391159686268e-06,
+      "loss": 0.0578,
+      "step": 2986
+    },
+    {
+      "epoch": 1.3266711081501221,
+      "grad_norm": 0.4751851347230112,
+      "learning_rate": 8.4519890236537e-06,
+      "loss": 0.0496,
+      "step": 2987
+    },
+    {
+      "epoch": 1.3271152564956696,
+      "grad_norm": 0.46876537416517666,
+      "learning_rate": 8.450586368728541e-06,
+      "loss": 0.0556,
+      "step": 2988
+    },
+    {
+      "epoch": 1.327559404841217,
+      "grad_norm": 0.5342196778194671,
+      "learning_rate": 8.449183195121638e-06,
+      "loss": 0.0462,
+      "step": 2989
+    },
+    {
+      "epoch": 1.3280035531867644,
+      "grad_norm": 0.7891053866526658,
+      "learning_rate": 8.447779503043907e-06,
+      "loss": 0.0695,
+      "step": 2990
+    },
+    {
+      "epoch": 1.3284477015323117,
+      "grad_norm": 0.4837197404803745,
+      "learning_rate": 8.44637529270635e-06,
+      "loss": 0.0416,
+      "step": 2991
+    },
+    {
+      "epoch": 1.3288918498778592,
+      "grad_norm": 0.4822831093408252,
+      "learning_rate": 8.444970564320044e-06,
+      "loss": 0.0492,
+      "step": 2992
+    },
+    {
+      "epoch": 1.3293359982234065,
+      "grad_norm": 1.1468745158193232,
+      "learning_rate": 8.443565318096141e-06,
+      "loss": 0.0829,
+      "step": 2993
+    },
+    {
+      "epoch": 1.329780146568954,
+      "grad_norm": 0.3846877899730632,
+      "learning_rate": 8.442159554245875e-06,
+      "loss": 0.0383,
+      "step": 2994
+    },
+    {
+      "epoch": 1.3302242949145016,
+      "grad_norm": 0.48249463397520415,
+      "learning_rate": 8.440753272980555e-06,
+      "loss": 0.0487,
+      "step": 2995
+    },
+    {
+      "epoch": 1.3306684432600488,
+      "grad_norm": 0.5300221622862935,
+      "learning_rate": 8.439346474511572e-06,
+      "loss": 0.0506,
+      "step": 2996
+    },
+    {
+      "epoch": 1.3311125916055961,
+      "grad_norm": 0.6183171953601485,
+      "learning_rate": 8.437939159050388e-06,
+      "loss": 0.063,
+      "step": 2997
+    },
+    {
+      "epoch": 1.3315567399511437,
+      "grad_norm": 0.5988476945915722,
+      "learning_rate": 8.43653132680855e-06,
+      "loss": 0.0625,
+      "step": 2998
+    },
+    {
+      "epoch": 1.3320008882966912,
+      "grad_norm": 0.5556515413344068,
+      "learning_rate": 8.435122977997675e-06,
+      "loss": 0.0491,
+      "step": 2999
+    },
+    {
+      "epoch": 1.3324450366422385,
+      "grad_norm": 0.4154893093213718,
+      "learning_rate": 8.433714112829464e-06,
+      "loss": 0.0431,
+      "step": 3000
+    },
+    {
+      "epoch": 1.332889184987786,
+      "grad_norm": 0.6020512275698489,
+      "learning_rate": 8.432304731515695e-06,
+      "loss": 0.0596,
+      "step": 3001
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.41304989339620046,
+      "learning_rate": 8.430894834268218e-06,
+      "loss": 0.0436,
+      "step": 3002
+    },
+    {
+      "epoch": 1.3337774816788808,
+      "grad_norm": 0.5204666086414845,
+      "learning_rate": 8.429484421298968e-06,
+      "loss": 0.0581,
+      "step": 3003
+    },
+    {
+      "epoch": 1.334221630024428,
+      "grad_norm": 0.5429684889982698,
+      "learning_rate": 8.428073492819953e-06,
+      "loss": 0.0404,
+      "step": 3004
+    },
+    {
+      "epoch": 1.3346657783699756,
+      "grad_norm": 0.4188357634818928,
+      "learning_rate": 8.426662049043258e-06,
+      "loss": 0.0422,
+      "step": 3005
+    },
+    {
+      "epoch": 1.335109926715523,
+      "grad_norm": 0.40403921511813046,
+      "learning_rate": 8.42525009018105e-06,
+      "loss": 0.0376,
+      "step": 3006
+    },
+    {
+      "epoch": 1.3355540750610704,
+      "grad_norm": 0.40629767747303225,
+      "learning_rate": 8.423837616445568e-06,
+      "loss": 0.0404,
+      "step": 3007
+    },
+    {
+      "epoch": 1.335998223406618,
+      "grad_norm": 0.4159409808633564,
+      "learning_rate": 8.42242462804913e-06,
+      "loss": 0.0394,
+      "step": 3008
+    },
+    {
+      "epoch": 1.3364423717521652,
+      "grad_norm": 0.7041981820609958,
+      "learning_rate": 8.421011125204134e-06,
+      "loss": 0.0693,
+      "step": 3009
+    },
+    {
+      "epoch": 1.3368865200977127,
+      "grad_norm": 0.7951485702237924,
+      "learning_rate": 8.419597108123054e-06,
+      "loss": 0.0612,
+      "step": 3010
+    },
+    {
+      "epoch": 1.33733066844326,
+      "grad_norm": 0.3990758518161739,
+      "learning_rate": 8.418182577018438e-06,
+      "loss": 0.0511,
+      "step": 3011
+    },
+    {
+      "epoch": 1.3377748167888075,
+      "grad_norm": 0.9209585349240309,
+      "learning_rate": 8.416767532102918e-06,
+      "loss": 0.0416,
+      "step": 3012
+    },
+    {
+      "epoch": 1.338218965134355,
+      "grad_norm": 0.5459049872521075,
+      "learning_rate": 8.415351973589197e-06,
+      "loss": 0.0434,
+      "step": 3013
+    },
+    {
+      "epoch": 1.3386631134799023,
+      "grad_norm": 0.5057701190286418,
+      "learning_rate": 8.413935901690057e-06,
+      "loss": 0.0434,
+      "step": 3014
+    },
+    {
+      "epoch": 1.3391072618254496,
+      "grad_norm": 0.48056478669043734,
+      "learning_rate": 8.412519316618359e-06,
+      "loss": 0.0502,
+      "step": 3015
+    },
+    {
+      "epoch": 1.339551410170997,
+      "grad_norm": 0.45090544113428555,
+      "learning_rate": 8.411102218587039e-06,
+      "loss": 0.0419,
+      "step": 3016
+    },
+    {
+      "epoch": 1.3399955585165446,
+      "grad_norm": 0.4066569281524885,
+      "learning_rate": 8.40968460780911e-06,
+      "loss": 0.0368,
+      "step": 3017
+    },
+    {
+      "epoch": 1.340439706862092,
+      "grad_norm": 0.48568103858702,
+      "learning_rate": 8.408266484497664e-06,
+      "loss": 0.0416,
+      "step": 3018
+    },
+    {
+      "epoch": 1.3408838552076394,
+      "grad_norm": 0.5224793245797686,
+      "learning_rate": 8.406847848865871e-06,
+      "loss": 0.0573,
+      "step": 3019
+    },
+    {
+      "epoch": 1.3413280035531867,
+      "grad_norm": 0.5080673094422519,
+      "learning_rate": 8.405428701126973e-06,
+      "loss": 0.0496,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3417721518987342,
+      "grad_norm": 0.5515622788542163,
+      "learning_rate": 8.404009041494292e-06,
+      "loss": 0.054,
+      "step": 3021
+    },
+    {
+      "epoch": 1.3422163002442815,
+      "grad_norm": 0.4992441781647522,
+      "learning_rate": 8.40258887018123e-06,
+      "loss": 0.0383,
+      "step": 3022
+    },
+    {
+      "epoch": 1.342660448589829,
+      "grad_norm": 0.5853495583783108,
+      "learning_rate": 8.40116818740126e-06,
+      "loss": 0.0419,
+      "step": 3023
+    },
+    {
+      "epoch": 1.3431045969353765,
+      "grad_norm": 0.47598229779455054,
+      "learning_rate": 8.399746993367936e-06,
+      "loss": 0.0342,
+      "step": 3024
+    },
+    {
+      "epoch": 1.3435487452809238,
+      "grad_norm": 0.46492013802658844,
+      "learning_rate": 8.398325288294886e-06,
+      "loss": 0.043,
+      "step": 3025
+    },
+    {
+      "epoch": 1.3439928936264711,
+      "grad_norm": 0.4287825450323698,
+      "learning_rate": 8.396903072395819e-06,
+      "loss": 0.0401,
+      "step": 3026
+    },
+    {
+      "epoch": 1.3444370419720186,
+      "grad_norm": 0.4586017747854982,
+      "learning_rate": 8.395480345884516e-06,
+      "loss": 0.0431,
+      "step": 3027
+    },
+    {
+      "epoch": 1.3448811903175661,
+      "grad_norm": 0.494703530250107,
+      "learning_rate": 8.39405710897484e-06,
+      "loss": 0.0462,
+      "step": 3028
+    },
+    {
+      "epoch": 1.3453253386631134,
+      "grad_norm": 0.5340552133978543,
+      "learning_rate": 8.392633361880724e-06,
+      "loss": 0.0567,
+      "step": 3029
+    },
+    {
+      "epoch": 1.345769487008661,
+      "grad_norm": 0.49912393023299867,
+      "learning_rate": 8.391209104816183e-06,
+      "loss": 0.0501,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3462136353542082,
+      "grad_norm": 0.6779311071781129,
+      "learning_rate": 8.389784337995306e-06,
+      "loss": 0.0682,
+      "step": 3031
+    },
+    {
+      "epoch": 1.3466577836997558,
+      "grad_norm": 0.47717232493661027,
+      "learning_rate": 8.388359061632262e-06,
+      "loss": 0.043,
+      "step": 3032
+    },
+    {
+      "epoch": 1.347101932045303,
+      "grad_norm": 0.47326694390163015,
+      "learning_rate": 8.386933275941294e-06,
+      "loss": 0.0456,
+      "step": 3033
+    },
+    {
+      "epoch": 1.3475460803908506,
+      "grad_norm": 0.6662990359563103,
+      "learning_rate": 8.385506981136717e-06,
+      "loss": 0.0487,
+      "step": 3034
+    },
+    {
+      "epoch": 1.347990228736398,
+      "grad_norm": 0.718683780503916,
+      "learning_rate": 8.384080177432933e-06,
+      "loss": 0.0394,
+      "step": 3035
+    },
+    {
+      "epoch": 1.3484343770819454,
+      "grad_norm": 0.5289766475406907,
+      "learning_rate": 8.382652865044414e-06,
+      "loss": 0.0553,
+      "step": 3036
+    },
+    {
+      "epoch": 1.3488785254274926,
+      "grad_norm": 0.6441199621383914,
+      "learning_rate": 8.381225044185708e-06,
+      "loss": 0.0522,
+      "step": 3037
+    },
+    {
+      "epoch": 1.3493226737730402,
+      "grad_norm": 0.9166796511578611,
+      "learning_rate": 8.37979671507144e-06,
+      "loss": 0.0458,
+      "step": 3038
+    },
+    {
+      "epoch": 1.3497668221185877,
+      "grad_norm": 0.531028146563615,
+      "learning_rate": 8.378367877916313e-06,
+      "loss": 0.0522,
+      "step": 3039
+    },
+    {
+      "epoch": 1.350210970464135,
+      "grad_norm": 0.46742487240112696,
+      "learning_rate": 8.376938532935106e-06,
+      "loss": 0.0379,
+      "step": 3040
+    },
+    {
+      "epoch": 1.3506551188096825,
+      "grad_norm": 0.5443821001104316,
+      "learning_rate": 8.375508680342674e-06,
+      "loss": 0.0439,
+      "step": 3041
+    },
+    {
+      "epoch": 1.3510992671552298,
+      "grad_norm": 0.5727922217841197,
+      "learning_rate": 8.374078320353944e-06,
+      "loss": 0.0424,
+      "step": 3042
+    },
+    {
+      "epoch": 1.3515434155007773,
+      "grad_norm": 0.49458279067850347,
+      "learning_rate": 8.37264745318393e-06,
+      "loss": 0.0506,
+      "step": 3043
+    },
+    {
+      "epoch": 1.3519875638463246,
+      "grad_norm": 1.3831663223083857,
+      "learning_rate": 8.371216079047713e-06,
+      "loss": 0.0392,
+      "step": 3044
+    },
+    {
+      "epoch": 1.352431712191872,
+      "grad_norm": 0.4619345954223979,
+      "learning_rate": 8.369784198160451e-06,
+      "loss": 0.0413,
+      "step": 3045
+    },
+    {
+      "epoch": 1.3528758605374196,
+      "grad_norm": 0.4847101460239981,
+      "learning_rate": 8.368351810737383e-06,
+      "loss": 0.0389,
+      "step": 3046
+    },
+    {
+      "epoch": 1.3533200088829669,
+      "grad_norm": 0.5592056103162423,
+      "learning_rate": 8.366918916993817e-06,
+      "loss": 0.0527,
+      "step": 3047
+    },
+    {
+      "epoch": 1.3537641572285144,
+      "grad_norm": 0.3872649705998668,
+      "learning_rate": 8.365485517145145e-06,
+      "loss": 0.0324,
+      "step": 3048
+    },
+    {
+      "epoch": 1.3542083055740617,
+      "grad_norm": 0.4866737656887245,
+      "learning_rate": 8.364051611406829e-06,
+      "loss": 0.0438,
+      "step": 3049
+    },
+    {
+      "epoch": 1.3546524539196092,
+      "grad_norm": 0.969360380725499,
+      "learning_rate": 8.362617199994413e-06,
+      "loss": 0.0471,
+      "step": 3050
+    },
+    {
+      "epoch": 1.3550966022651565,
+      "grad_norm": 0.582815842887394,
+      "learning_rate": 8.36118228312351e-06,
+      "loss": 0.042,
+      "step": 3051
+    },
+    {
+      "epoch": 1.355540750610704,
+      "grad_norm": 0.721376890360811,
+      "learning_rate": 8.359746861009812e-06,
+      "loss": 0.0625,
+      "step": 3052
+    },
+    {
+      "epoch": 1.3559848989562515,
+      "grad_norm": 0.6319029678683107,
+      "learning_rate": 8.358310933869091e-06,
+      "loss": 0.0645,
+      "step": 3053
+    },
+    {
+      "epoch": 1.3564290473017988,
+      "grad_norm": 2.7049677243291828,
+      "learning_rate": 8.356874501917188e-06,
+      "loss": 0.0531,
+      "step": 3054
+    },
+    {
+      "epoch": 1.356873195647346,
+      "grad_norm": 0.3779085536381152,
+      "learning_rate": 8.355437565370022e-06,
+      "loss": 0.0356,
+      "step": 3055
+    },
+    {
+      "epoch": 1.3573173439928936,
+      "grad_norm": 0.7424758490934648,
+      "learning_rate": 8.354000124443594e-06,
+      "loss": 0.0524,
+      "step": 3056
+    },
+    {
+      "epoch": 1.3577614923384411,
+      "grad_norm": 0.6071475225743952,
+      "learning_rate": 8.352562179353971e-06,
+      "loss": 0.0476,
+      "step": 3057
+    },
+    {
+      "epoch": 1.3582056406839884,
+      "grad_norm": 0.4823292349978002,
+      "learning_rate": 8.351123730317303e-06,
+      "loss": 0.037,
+      "step": 3058
+    },
+    {
+      "epoch": 1.358649789029536,
+      "grad_norm": 0.42940088227364875,
+      "learning_rate": 8.349684777549813e-06,
+      "loss": 0.032,
+      "step": 3059
+    },
+    {
+      "epoch": 1.3590939373750832,
+      "grad_norm": 0.6291618703403149,
+      "learning_rate": 8.348245321267798e-06,
+      "loss": 0.0542,
+      "step": 3060
+    },
+    {
+      "epoch": 1.3595380857206307,
+      "grad_norm": 0.46443218939790437,
+      "learning_rate": 8.346805361687637e-06,
+      "loss": 0.0482,
+      "step": 3061
+    },
+    {
+      "epoch": 1.359982234066178,
+      "grad_norm": 0.44266144811788105,
+      "learning_rate": 8.345364899025776e-06,
+      "loss": 0.0508,
+      "step": 3062
+    },
+    {
+      "epoch": 1.3604263824117255,
+      "grad_norm": 0.5931758192805282,
+      "learning_rate": 8.343923933498742e-06,
+      "loss": 0.052,
+      "step": 3063
+    },
+    {
+      "epoch": 1.360870530757273,
+      "grad_norm": 0.4228622965921645,
+      "learning_rate": 8.342482465323141e-06,
+      "loss": 0.0342,
+      "step": 3064
+    },
+    {
+      "epoch": 1.3613146791028203,
+      "grad_norm": 0.6573207468033229,
+      "learning_rate": 8.341040494715644e-06,
+      "loss": 0.054,
+      "step": 3065
+    },
+    {
+      "epoch": 1.3617588274483676,
+      "grad_norm": 0.694757132531818,
+      "learning_rate": 8.339598021893007e-06,
+      "loss": 0.0573,
+      "step": 3066
+    },
+    {
+      "epoch": 1.3622029757939151,
+      "grad_norm": 0.5168331449919007,
+      "learning_rate": 8.338155047072058e-06,
+      "loss": 0.0476,
+      "step": 3067
+    },
+    {
+      "epoch": 1.3626471241394627,
+      "grad_norm": 0.5294718006665416,
+      "learning_rate": 8.336711570469698e-06,
+      "loss": 0.0536,
+      "step": 3068
+    },
+    {
+      "epoch": 1.36309127248501,
+      "grad_norm": 0.5902958008882144,
+      "learning_rate": 8.33526759230291e-06,
+      "loss": 0.0543,
+      "step": 3069
+    },
+    {
+      "epoch": 1.3635354208305575,
+      "grad_norm": 0.5123768514504031,
+      "learning_rate": 8.333823112788747e-06,
+      "loss": 0.055,
+      "step": 3070
+    },
+    {
+      "epoch": 1.3639795691761047,
+      "grad_norm": 0.4051444968388966,
+      "learning_rate": 8.332378132144336e-06,
+      "loss": 0.0408,
+      "step": 3071
+    },
+    {
+      "epoch": 1.3644237175216523,
+      "grad_norm": 0.45045135771966455,
+      "learning_rate": 8.330932650586887e-06,
+      "loss": 0.0449,
+      "step": 3072
+    },
+    {
+      "epoch": 1.3648678658671995,
+      "grad_norm": 0.6922855447849937,
+      "learning_rate": 8.329486668333677e-06,
+      "loss": 0.052,
+      "step": 3073
+    },
+    {
+      "epoch": 1.365312014212747,
+      "grad_norm": 0.3651988449178566,
+      "learning_rate": 8.328040185602063e-06,
+      "loss": 0.0383,
+      "step": 3074
+    },
+    {
+      "epoch": 1.3657561625582946,
+      "grad_norm": 0.6637690048102727,
+      "learning_rate": 8.326593202609475e-06,
+      "loss": 0.0571,
+      "step": 3075
+    },
+    {
+      "epoch": 1.3662003109038419,
+      "grad_norm": 0.5428561517455449,
+      "learning_rate": 8.325145719573419e-06,
+      "loss": 0.0384,
+      "step": 3076
+    },
+    {
+      "epoch": 1.3666444592493894,
+      "grad_norm": 0.7398928249310872,
+      "learning_rate": 8.323697736711478e-06,
+      "loss": 0.0535,
+      "step": 3077
+    },
+    {
+      "epoch": 1.3670886075949367,
+      "grad_norm": 0.6968067508458905,
+      "learning_rate": 8.322249254241309e-06,
+      "loss": 0.0624,
+      "step": 3078
+    },
+    {
+      "epoch": 1.3675327559404842,
+      "grad_norm": 0.42153823690397146,
+      "learning_rate": 8.320800272380639e-06,
+      "loss": 0.036,
+      "step": 3079
+    },
+    {
+      "epoch": 1.3679769042860315,
+      "grad_norm": 0.7070448290401131,
+      "learning_rate": 8.319350791347279e-06,
+      "loss": 0.0682,
+      "step": 3080
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.5028115971086855,
+      "learning_rate": 8.31790081135911e-06,
+      "loss": 0.0543,
+      "step": 3081
+    },
+    {
+      "epoch": 1.3688652009771265,
+      "grad_norm": 0.5307360164451863,
+      "learning_rate": 8.316450332634084e-06,
+      "loss": 0.0535,
+      "step": 3082
+    },
+    {
+      "epoch": 1.3693093493226738,
+      "grad_norm": 0.6501199496749419,
+      "learning_rate": 8.31499935539024e-06,
+      "loss": 0.0461,
+      "step": 3083
+    },
+    {
+      "epoch": 1.369753497668221,
+      "grad_norm": 0.5797446911962002,
+      "learning_rate": 8.313547879845682e-06,
+      "loss": 0.0472,
+      "step": 3084
+    },
+    {
+      "epoch": 1.3701976460137686,
+      "grad_norm": 0.4268924633424178,
+      "learning_rate": 8.312095906218588e-06,
+      "loss": 0.0447,
+      "step": 3085
+    },
+    {
+      "epoch": 1.370641794359316,
+      "grad_norm": 0.6304839984425172,
+      "learning_rate": 8.310643434727216e-06,
+      "loss": 0.0625,
+      "step": 3086
+    },
+    {
+      "epoch": 1.3710859427048634,
+      "grad_norm": 0.5665653851714997,
+      "learning_rate": 8.3091904655899e-06,
+      "loss": 0.0539,
+      "step": 3087
+    },
+    {
+      "epoch": 1.371530091050411,
+      "grad_norm": 0.40002359226927725,
+      "learning_rate": 8.307736999025043e-06,
+      "loss": 0.0329,
+      "step": 3088
+    },
+    {
+      "epoch": 1.3719742393959582,
+      "grad_norm": 0.7513234692327438,
+      "learning_rate": 8.306283035251125e-06,
+      "loss": 0.0648,
+      "step": 3089
+    },
+    {
+      "epoch": 1.3724183877415057,
+      "grad_norm": 0.5469597650300881,
+      "learning_rate": 8.304828574486704e-06,
+      "loss": 0.0432,
+      "step": 3090
+    },
+    {
+      "epoch": 1.372862536087053,
+      "grad_norm": 0.48401783163541084,
+      "learning_rate": 8.303373616950408e-06,
+      "loss": 0.0457,
+      "step": 3091
+    },
+    {
+      "epoch": 1.3733066844326005,
+      "grad_norm": 0.4692359706390245,
+      "learning_rate": 8.301918162860944e-06,
+      "loss": 0.0422,
+      "step": 3092
+    },
+    {
+      "epoch": 1.373750832778148,
+      "grad_norm": 0.5378714608611251,
+      "learning_rate": 8.30046221243709e-06,
+      "loss": 0.0513,
+      "step": 3093
+    },
+    {
+      "epoch": 1.3741949811236953,
+      "grad_norm": 0.5625724873607021,
+      "learning_rate": 8.2990057658977e-06,
+      "loss": 0.0532,
+      "step": 3094
+    },
+    {
+      "epoch": 1.3746391294692426,
+      "grad_norm": 0.4899279935109507,
+      "learning_rate": 8.297548823461704e-06,
+      "loss": 0.0454,
+      "step": 3095
+    },
+    {
+      "epoch": 1.3750832778147901,
+      "grad_norm": 0.7726706942428412,
+      "learning_rate": 8.296091385348104e-06,
+      "loss": 0.0697,
+      "step": 3096
+    },
+    {
+      "epoch": 1.3755274261603376,
+      "grad_norm": 0.5012946731170276,
+      "learning_rate": 8.294633451775977e-06,
+      "loss": 0.0378,
+      "step": 3097
+    },
+    {
+      "epoch": 1.375971574505885,
+      "grad_norm": 0.48151498654702213,
+      "learning_rate": 8.293175022964476e-06,
+      "loss": 0.0434,
+      "step": 3098
+    },
+    {
+      "epoch": 1.3764157228514324,
+      "grad_norm": 0.4130737267146749,
+      "learning_rate": 8.291716099132829e-06,
+      "loss": 0.0359,
+      "step": 3099
+    },
+    {
+      "epoch": 1.3768598711969797,
+      "grad_norm": 0.4681903495603353,
+      "learning_rate": 8.290256680500336e-06,
+      "loss": 0.0398,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3773040195425272,
+      "grad_norm": 0.6272554974390412,
+      "learning_rate": 8.28879676728637e-06,
+      "loss": 0.0627,
+      "step": 3101
+    },
+    {
+      "epoch": 1.3777481678880745,
+      "grad_norm": 0.5051341999455976,
+      "learning_rate": 8.287336359710386e-06,
+      "loss": 0.047,
+      "step": 3102
+    },
+    {
+      "epoch": 1.378192316233622,
+      "grad_norm": 0.5262863424826024,
+      "learning_rate": 8.285875457991903e-06,
+      "loss": 0.047,
+      "step": 3103
+    },
+    {
+      "epoch": 1.3786364645791696,
+      "grad_norm": 0.5493114786417674,
+      "learning_rate": 8.284414062350524e-06,
+      "loss": 0.0477,
+      "step": 3104
+    },
+    {
+      "epoch": 1.3790806129247168,
+      "grad_norm": 0.529779249166117,
+      "learning_rate": 8.282952173005916e-06,
+      "loss": 0.0489,
+      "step": 3105
+    },
+    {
+      "epoch": 1.3795247612702641,
+      "grad_norm": 0.5169327523467377,
+      "learning_rate": 8.28148979017783e-06,
+      "loss": 0.0568,
+      "step": 3106
+    },
+    {
+      "epoch": 1.3799689096158116,
+      "grad_norm": 0.6639124677133846,
+      "learning_rate": 8.280026914086086e-06,
+      "loss": 0.0831,
+      "step": 3107
+    },
+    {
+      "epoch": 1.3804130579613592,
+      "grad_norm": 0.6604146215653776,
+      "learning_rate": 8.278563544950579e-06,
+      "loss": 0.0688,
+      "step": 3108
+    },
+    {
+      "epoch": 1.3808572063069064,
+      "grad_norm": 0.499500034008347,
+      "learning_rate": 8.277099682991276e-06,
+      "loss": 0.0466,
+      "step": 3109
+    },
+    {
+      "epoch": 1.381301354652454,
+      "grad_norm": 0.5463437639449784,
+      "learning_rate": 8.275635328428226e-06,
+      "loss": 0.0576,
+      "step": 3110
+    },
+    {
+      "epoch": 1.3817455029980013,
+      "grad_norm": 0.43757681367045054,
+      "learning_rate": 8.274170481481541e-06,
+      "loss": 0.0348,
+      "step": 3111
+    },
+    {
+      "epoch": 1.3821896513435488,
+      "grad_norm": 0.5431219597419497,
+      "learning_rate": 8.272705142371414e-06,
+      "loss": 0.0592,
+      "step": 3112
+    },
+    {
+      "epoch": 1.382633799689096,
+      "grad_norm": 0.6091624737010345,
+      "learning_rate": 8.271239311318111e-06,
+      "loss": 0.0645,
+      "step": 3113
+    },
+    {
+      "epoch": 1.3830779480346436,
+      "grad_norm": 0.46664330067036014,
+      "learning_rate": 8.269772988541971e-06,
+      "loss": 0.0293,
+      "step": 3114
+    },
+    {
+      "epoch": 1.383522096380191,
+      "grad_norm": 0.39018350334607316,
+      "learning_rate": 8.268306174263407e-06,
+      "loss": 0.0488,
+      "step": 3115
+    },
+    {
+      "epoch": 1.3839662447257384,
+      "grad_norm": 0.7360437552444796,
+      "learning_rate": 8.266838868702904e-06,
+      "loss": 0.0539,
+      "step": 3116
+    },
+    {
+      "epoch": 1.3844103930712859,
+      "grad_norm": 0.3523843914894874,
+      "learning_rate": 8.265371072081028e-06,
+      "loss": 0.033,
+      "step": 3117
+    },
+    {
+      "epoch": 1.3848545414168332,
+      "grad_norm": 0.5841378028235703,
+      "learning_rate": 8.263902784618409e-06,
+      "loss": 0.0512,
+      "step": 3118
+    },
+    {
+      "epoch": 1.3852986897623807,
+      "grad_norm": 0.6604716728640524,
+      "learning_rate": 8.262434006535759e-06,
+      "loss": 0.0841,
+      "step": 3119
+    },
+    {
+      "epoch": 1.385742838107928,
+      "grad_norm": 0.43203367132780174,
+      "learning_rate": 8.260964738053859e-06,
+      "loss": 0.0474,
+      "step": 3120
+    },
+    {
+      "epoch": 1.3861869864534755,
+      "grad_norm": 0.48844306169229995,
+      "learning_rate": 8.259494979393563e-06,
+      "loss": 0.0481,
+      "step": 3121
+    },
+    {
+      "epoch": 1.386631134799023,
+      "grad_norm": 0.3906618714847538,
+      "learning_rate": 8.258024730775805e-06,
+      "loss": 0.0398,
+      "step": 3122
+    },
+    {
+      "epoch": 1.3870752831445703,
+      "grad_norm": 0.47997636522094667,
+      "learning_rate": 8.256553992421583e-06,
+      "loss": 0.0554,
+      "step": 3123
+    },
+    {
+      "epoch": 1.3875194314901176,
+      "grad_norm": 0.6434789607226818,
+      "learning_rate": 8.255082764551978e-06,
+      "loss": 0.0484,
+      "step": 3124
+    },
+    {
+      "epoch": 1.387963579835665,
+      "grad_norm": 0.3887987489990569,
+      "learning_rate": 8.25361104738814e-06,
+      "loss": 0.0464,
+      "step": 3125
+    },
+    {
+      "epoch": 1.3884077281812126,
+      "grad_norm": 0.487796373914711,
+      "learning_rate": 8.252138841151292e-06,
+      "loss": 0.0441,
+      "step": 3126
+    },
+    {
+      "epoch": 1.38885187652676,
+      "grad_norm": 0.6398270077190703,
+      "learning_rate": 8.250666146062732e-06,
+      "loss": 0.0419,
+      "step": 3127
+    },
+    {
+      "epoch": 1.3892960248723074,
+      "grad_norm": 0.4422350214805325,
+      "learning_rate": 8.249192962343829e-06,
+      "loss": 0.041,
+      "step": 3128
+    },
+    {
+      "epoch": 1.3897401732178547,
+      "grad_norm": 0.5252093775770235,
+      "learning_rate": 8.247719290216032e-06,
+      "loss": 0.0479,
+      "step": 3129
+    },
+    {
+      "epoch": 1.3901843215634022,
+      "grad_norm": 0.6735609599734196,
+      "learning_rate": 8.246245129900856e-06,
+      "loss": 0.038,
+      "step": 3130
+    },
+    {
+      "epoch": 1.3906284699089495,
+      "grad_norm": 0.5147679810299035,
+      "learning_rate": 8.244770481619892e-06,
+      "loss": 0.05,
+      "step": 3131
+    },
+    {
+      "epoch": 1.391072618254497,
+      "grad_norm": 0.4260669615511085,
+      "learning_rate": 8.243295345594807e-06,
+      "loss": 0.0351,
+      "step": 3132
+    },
+    {
+      "epoch": 1.3915167666000445,
+      "grad_norm": 0.5948196620222996,
+      "learning_rate": 8.241819722047337e-06,
+      "loss": 0.0532,
+      "step": 3133
+    },
+    {
+      "epoch": 1.3919609149455918,
+      "grad_norm": 0.4954498969782017,
+      "learning_rate": 8.240343611199294e-06,
+      "loss": 0.0395,
+      "step": 3134
+    },
+    {
+      "epoch": 1.3924050632911391,
+      "grad_norm": 0.5748384966095534,
+      "learning_rate": 8.238867013272562e-06,
+      "loss": 0.051,
+      "step": 3135
+    },
+    {
+      "epoch": 1.3928492116366866,
+      "grad_norm": 0.47868756581223826,
+      "learning_rate": 8.237389928489099e-06,
+      "loss": 0.0406,
+      "step": 3136
+    },
+    {
+      "epoch": 1.3932933599822341,
+      "grad_norm": 0.42581407271928307,
+      "learning_rate": 8.235912357070938e-06,
+      "loss": 0.0337,
+      "step": 3137
+    },
+    {
+      "epoch": 1.3937375083277814,
+      "grad_norm": 0.6781537407713069,
+      "learning_rate": 8.234434299240179e-06,
+      "loss": 0.0483,
+      "step": 3138
+    },
+    {
+      "epoch": 1.394181656673329,
+      "grad_norm": 0.598623626498893,
+      "learning_rate": 8.232955755219002e-06,
+      "loss": 0.0498,
+      "step": 3139
+    },
+    {
+      "epoch": 1.3946258050188762,
+      "grad_norm": 0.6505669269254222,
+      "learning_rate": 8.231476725229659e-06,
+      "loss": 0.0391,
+      "step": 3140
+    },
+    {
+      "epoch": 1.3950699533644237,
+      "grad_norm": 0.5150042095978116,
+      "learning_rate": 8.229997209494468e-06,
+      "loss": 0.0538,
+      "step": 3141
+    },
+    {
+      "epoch": 1.395514101709971,
+      "grad_norm": 0.48445621459488397,
+      "learning_rate": 8.228517208235829e-06,
+      "loss": 0.0485,
+      "step": 3142
+    },
+    {
+      "epoch": 1.3959582500555185,
+      "grad_norm": 0.42479311306640655,
+      "learning_rate": 8.22703672167621e-06,
+      "loss": 0.0454,
+      "step": 3143
+    },
+    {
+      "epoch": 1.396402398401066,
+      "grad_norm": 0.4968133842399203,
+      "learning_rate": 8.225555750038157e-06,
+      "loss": 0.0548,
+      "step": 3144
+    },
+    {
+      "epoch": 1.3968465467466133,
+      "grad_norm": 0.839408513781402,
+      "learning_rate": 8.22407429354428e-06,
+      "loss": 0.0508,
+      "step": 3145
+    },
+    {
+      "epoch": 1.3972906950921609,
+      "grad_norm": 0.4593510375216177,
+      "learning_rate": 8.222592352417268e-06,
+      "loss": 0.0484,
+      "step": 3146
+    },
+    {
+      "epoch": 1.3977348434377082,
+      "grad_norm": 0.41788586732862426,
+      "learning_rate": 8.221109926879885e-06,
+      "loss": 0.0394,
+      "step": 3147
+    },
+    {
+      "epoch": 1.3981789917832557,
+      "grad_norm": 0.6011495923589089,
+      "learning_rate": 8.219627017154962e-06,
+      "loss": 0.0538,
+      "step": 3148
+    },
+    {
+      "epoch": 1.398623140128803,
+      "grad_norm": 0.483716139104255,
+      "learning_rate": 8.218143623465407e-06,
+      "loss": 0.0479,
+      "step": 3149
+    },
+    {
+      "epoch": 1.3990672884743505,
+      "grad_norm": 0.7567570451553496,
+      "learning_rate": 8.216659746034199e-06,
+      "loss": 0.059,
+      "step": 3150
+    },
+    {
+      "epoch": 1.399511436819898,
+      "grad_norm": 0.48553150567848463,
+      "learning_rate": 8.215175385084389e-06,
+      "loss": 0.0617,
+      "step": 3151
+    },
+    {
+      "epoch": 1.3999555851654453,
+      "grad_norm": 0.47459878094993846,
+      "learning_rate": 8.2136905408391e-06,
+      "loss": 0.0527,
+      "step": 3152
+    },
+    {
+      "epoch": 1.4003997335109926,
+      "grad_norm": 0.5833223937440162,
+      "learning_rate": 8.212205213521535e-06,
+      "loss": 0.0626,
+      "step": 3153
+    },
+    {
+      "epoch": 1.40084388185654,
+      "grad_norm": 0.35626979864365443,
+      "learning_rate": 8.210719403354961e-06,
+      "loss": 0.033,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4012880302020876,
+      "grad_norm": 0.7430849859689593,
+      "learning_rate": 8.209233110562719e-06,
+      "loss": 0.0645,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4017321785476349,
+      "grad_norm": 0.45738357167910687,
+      "learning_rate": 8.207746335368223e-06,
+      "loss": 0.0378,
+      "step": 3156
+    },
+    {
+      "epoch": 1.4021763268931824,
+      "grad_norm": 0.5265228817791134,
+      "learning_rate": 8.206259077994966e-06,
+      "loss": 0.0653,
+      "step": 3157
+    },
+    {
+      "epoch": 1.4026204752387297,
+      "grad_norm": 0.5945877517368485,
+      "learning_rate": 8.204771338666504e-06,
+      "loss": 0.0517,
+      "step": 3158
+    },
+    {
+      "epoch": 1.4030646235842772,
+      "grad_norm": 0.48633704182422527,
+      "learning_rate": 8.20328311760647e-06,
+      "loss": 0.0467,
+      "step": 3159
+    },
+    {
+      "epoch": 1.4035087719298245,
+      "grad_norm": 0.45125218858977967,
+      "learning_rate": 8.201794415038569e-06,
+      "loss": 0.0436,
+      "step": 3160
+    },
+    {
+      "epoch": 1.403952920275372,
+      "grad_norm": 0.454560780546378,
+      "learning_rate": 8.200305231186578e-06,
+      "loss": 0.0566,
+      "step": 3161
+    },
+    {
+      "epoch": 1.4043970686209195,
+      "grad_norm": 0.4442970232995022,
+      "learning_rate": 8.198815566274346e-06,
+      "loss": 0.0343,
+      "step": 3162
+    },
+    {
+      "epoch": 1.4048412169664668,
+      "grad_norm": 0.4600734220829904,
+      "learning_rate": 8.197325420525797e-06,
+      "loss": 0.0471,
+      "step": 3163
+    },
+    {
+      "epoch": 1.405285365312014,
+      "grad_norm": 0.46601308549255893,
+      "learning_rate": 8.195834794164925e-06,
+      "loss": 0.0427,
+      "step": 3164
+    },
+    {
+      "epoch": 1.4057295136575616,
+      "grad_norm": 0.5996621159738204,
+      "learning_rate": 8.194343687415795e-06,
+      "loss": 0.0457,
+      "step": 3165
+    },
+    {
+      "epoch": 1.4061736620031091,
+      "grad_norm": 0.9069132714321548,
+      "learning_rate": 8.192852100502547e-06,
+      "loss": 0.0678,
+      "step": 3166
+    },
+    {
+      "epoch": 1.4066178103486564,
+      "grad_norm": 0.43166301548877434,
+      "learning_rate": 8.191360033649392e-06,
+      "loss": 0.0352,
+      "step": 3167
+    },
+    {
+      "epoch": 1.407061958694204,
+      "grad_norm": 0.5216023574393313,
+      "learning_rate": 8.18986748708061e-06,
+      "loss": 0.0475,
+      "step": 3168
+    },
+    {
+      "epoch": 1.4075061070397512,
+      "grad_norm": 0.5468879202276986,
+      "learning_rate": 8.18837446102056e-06,
+      "loss": 0.0547,
+      "step": 3169
+    },
+    {
+      "epoch": 1.4079502553852987,
+      "grad_norm": 0.4122751122216154,
+      "learning_rate": 8.186880955693667e-06,
+      "loss": 0.0357,
+      "step": 3170
+    },
+    {
+      "epoch": 1.408394403730846,
+      "grad_norm": 0.49605981026080986,
+      "learning_rate": 8.18538697132443e-06,
+      "loss": 0.0593,
+      "step": 3171
+    },
+    {
+      "epoch": 1.4088385520763935,
+      "grad_norm": 0.7328755570855009,
+      "learning_rate": 8.183892508137423e-06,
+      "loss": 0.0691,
+      "step": 3172
+    },
+    {
+      "epoch": 1.409282700421941,
+      "grad_norm": 0.5835998191367716,
+      "learning_rate": 8.182397566357286e-06,
+      "loss": 0.0621,
+      "step": 3173
+    },
+    {
+      "epoch": 1.4097268487674883,
+      "grad_norm": 0.45687160374129326,
+      "learning_rate": 8.180902146208734e-06,
+      "loss": 0.0529,
+      "step": 3174
+    },
+    {
+      "epoch": 1.4101709971130356,
+      "grad_norm": 0.4464006299057523,
+      "learning_rate": 8.179406247916555e-06,
+      "loss": 0.0462,
+      "step": 3175
+    },
+    {
+      "epoch": 1.4106151454585831,
+      "grad_norm": 0.5104992313269215,
+      "learning_rate": 8.17790987170561e-06,
+      "loss": 0.0397,
+      "step": 3176
+    },
+    {
+      "epoch": 1.4110592938041306,
+      "grad_norm": 0.4500241032324194,
+      "learning_rate": 8.176413017800828e-06,
+      "loss": 0.0588,
+      "step": 3177
+    },
+    {
+      "epoch": 1.411503442149678,
+      "grad_norm": 0.9698569462217295,
+      "learning_rate": 8.174915686427211e-06,
+      "loss": 0.0593,
+      "step": 3178
+    },
+    {
+      "epoch": 1.4119475904952254,
+      "grad_norm": 1.9467722797227018,
+      "learning_rate": 8.173417877809835e-06,
+      "loss": 0.0657,
+      "step": 3179
+    },
+    {
+      "epoch": 1.4123917388407727,
+      "grad_norm": 0.8619854581857347,
+      "learning_rate": 8.171919592173843e-06,
+      "loss": 0.0616,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4128358871863202,
+      "grad_norm": 2.1957744086497324,
+      "learning_rate": 8.170420829744458e-06,
+      "loss": 0.0441,
+      "step": 3181
+    },
+    {
+      "epoch": 1.4132800355318675,
+      "grad_norm": 0.437679909069633,
+      "learning_rate": 8.168921590746964e-06,
+      "loss": 0.0448,
+      "step": 3182
+    },
+    {
+      "epoch": 1.413724183877415,
+      "grad_norm": 0.44545740554831986,
+      "learning_rate": 8.167421875406725e-06,
+      "loss": 0.0534,
+      "step": 3183
+    },
+    {
+      "epoch": 1.4141683322229626,
+      "grad_norm": 0.6661022181292509,
+      "learning_rate": 8.165921683949172e-06,
+      "loss": 0.0752,
+      "step": 3184
+    },
+    {
+      "epoch": 1.4146124805685099,
+      "grad_norm": 0.5379954758436954,
+      "learning_rate": 8.164421016599811e-06,
+      "loss": 0.0465,
+      "step": 3185
+    },
+    {
+      "epoch": 1.4150566289140574,
+      "grad_norm": 0.7649578135185098,
+      "learning_rate": 8.162919873584216e-06,
+      "loss": 0.0655,
+      "step": 3186
+    },
+    {
+      "epoch": 1.4155007772596047,
+      "grad_norm": 0.6155469523069211,
+      "learning_rate": 8.161418255128037e-06,
+      "loss": 0.058,
+      "step": 3187
+    },
+    {
+      "epoch": 1.4159449256051522,
+      "grad_norm": 0.43980966134894045,
+      "learning_rate": 8.15991616145699e-06,
+      "loss": 0.046,
+      "step": 3188
+    },
+    {
+      "epoch": 1.4163890739506995,
+      "grad_norm": 0.3772085957077231,
+      "learning_rate": 8.158413592796867e-06,
+      "loss": 0.0407,
+      "step": 3189
+    },
+    {
+      "epoch": 1.416833222296247,
+      "grad_norm": 0.4916035633372075,
+      "learning_rate": 8.156910549373529e-06,
+      "loss": 0.0509,
+      "step": 3190
+    },
+    {
+      "epoch": 1.4172773706417945,
+      "grad_norm": 0.44586244165360156,
+      "learning_rate": 8.15540703141291e-06,
+      "loss": 0.0421,
+      "step": 3191
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6752971156056389,
+      "learning_rate": 8.153903039141011e-06,
+      "loss": 0.0394,
+      "step": 3192
+    },
+    {
+      "epoch": 1.418165667332889,
+      "grad_norm": 0.5021744951340577,
+      "learning_rate": 8.15239857278391e-06,
+      "loss": 0.0446,
+      "step": 3193
+    },
+    {
+      "epoch": 1.4186098156784366,
+      "grad_norm": 0.4057752198360003,
+      "learning_rate": 8.150893632567755e-06,
+      "loss": 0.0413,
+      "step": 3194
+    },
+    {
+      "epoch": 1.419053964023984,
+      "grad_norm": 0.44646591292416776,
+      "learning_rate": 8.149388218718763e-06,
+      "loss": 0.033,
+      "step": 3195
+    },
+    {
+      "epoch": 1.4194981123695314,
+      "grad_norm": 0.5430849460817241,
+      "learning_rate": 8.147882331463221e-06,
+      "loss": 0.0476,
+      "step": 3196
+    },
+    {
+      "epoch": 1.419942260715079,
+      "grad_norm": 0.5748975413489712,
+      "learning_rate": 8.146375971027492e-06,
+      "loss": 0.0422,
+      "step": 3197
+    },
+    {
+      "epoch": 1.4203864090606262,
+      "grad_norm": 0.8569726405019049,
+      "learning_rate": 8.144869137638008e-06,
+      "loss": 0.0586,
+      "step": 3198
+    },
+    {
+      "epoch": 1.4208305574061737,
+      "grad_norm": 0.48111798749968265,
+      "learning_rate": 8.14336183152127e-06,
+      "loss": 0.0511,
+      "step": 3199
+    },
+    {
+      "epoch": 1.421274705751721,
+      "grad_norm": 0.4447988259403195,
+      "learning_rate": 8.141854052903853e-06,
+      "loss": 0.0409,
+      "step": 3200
+    },
+    {
+      "epoch": 1.4217188540972685,
+      "grad_norm": 0.42318494315184574,
+      "learning_rate": 8.1403458020124e-06,
+      "loss": 0.0393,
+      "step": 3201
+    },
+    {
+      "epoch": 1.422163002442816,
+      "grad_norm": 0.527946330855177,
+      "learning_rate": 8.138837079073628e-06,
+      "loss": 0.0544,
+      "step": 3202
+    },
+    {
+      "epoch": 1.4226071507883633,
+      "grad_norm": 0.5168740647207917,
+      "learning_rate": 8.137327884314323e-06,
+      "loss": 0.042,
+      "step": 3203
+    },
+    {
+      "epoch": 1.4230512991339106,
+      "grad_norm": 0.44556932550096595,
+      "learning_rate": 8.135818217961344e-06,
+      "loss": 0.043,
+      "step": 3204
+    },
+    {
+      "epoch": 1.423495447479458,
+      "grad_norm": 0.4521168782601511,
+      "learning_rate": 8.13430808024162e-06,
+      "loss": 0.0616,
+      "step": 3205
+    },
+    {
+      "epoch": 1.4239395958250056,
+      "grad_norm": 0.998651913471337,
+      "learning_rate": 8.132797471382148e-06,
+      "loss": 0.0857,
+      "step": 3206
+    },
+    {
+      "epoch": 1.424383744170553,
+      "grad_norm": 0.4845757133637773,
+      "learning_rate": 8.131286391609996e-06,
+      "loss": 0.0316,
+      "step": 3207
+    },
+    {
+      "epoch": 1.4248278925161004,
+      "grad_norm": 0.5115534012315782,
+      "learning_rate": 8.129774841152311e-06,
+      "loss": 0.0442,
+      "step": 3208
+    },
+    {
+      "epoch": 1.4252720408616477,
+      "grad_norm": 0.4658784277879884,
+      "learning_rate": 8.128262820236302e-06,
+      "loss": 0.0392,
+      "step": 3209
+    },
+    {
+      "epoch": 1.4257161892071952,
+      "grad_norm": 0.7323138806429754,
+      "learning_rate": 8.12675032908925e-06,
+      "loss": 0.0476,
+      "step": 3210
+    },
+    {
+      "epoch": 1.4261603375527425,
+      "grad_norm": 0.7544916173168683,
+      "learning_rate": 8.125237367938511e-06,
+      "loss": 0.0539,
+      "step": 3211
+    },
+    {
+      "epoch": 1.42660448589829,
+      "grad_norm": 0.62197595498055,
+      "learning_rate": 8.123723937011507e-06,
+      "loss": 0.0526,
+      "step": 3212
+    },
+    {
+      "epoch": 1.4270486342438375,
+      "grad_norm": 0.411592343530701,
+      "learning_rate": 8.12221003653573e-06,
+      "loss": 0.0422,
+      "step": 3213
+    },
+    {
+      "epoch": 1.4274927825893848,
+      "grad_norm": 0.5467878759617523,
+      "learning_rate": 8.12069566673875e-06,
+      "loss": 0.0486,
+      "step": 3214
+    },
+    {
+      "epoch": 1.4279369309349323,
+      "grad_norm": 0.3951319611407401,
+      "learning_rate": 8.119180827848199e-06,
+      "loss": 0.0372,
+      "step": 3215
+    },
+    {
+      "epoch": 1.4283810792804796,
+      "grad_norm": 0.5355923646902133,
+      "learning_rate": 8.117665520091783e-06,
+      "loss": 0.0583,
+      "step": 3216
+    },
+    {
+      "epoch": 1.4288252276260272,
+      "grad_norm": 0.49870365746589324,
+      "learning_rate": 8.11614974369728e-06,
+      "loss": 0.0521,
+      "step": 3217
+    },
+    {
+      "epoch": 1.4292693759715744,
+      "grad_norm": 0.5463670187827259,
+      "learning_rate": 8.114633498892537e-06,
+      "loss": 0.0567,
+      "step": 3218
+    },
+    {
+      "epoch": 1.429713524317122,
+      "grad_norm": 0.5699638456777965,
+      "learning_rate": 8.11311678590547e-06,
+      "loss": 0.05,
+      "step": 3219
+    },
+    {
+      "epoch": 1.4301576726626695,
+      "grad_norm": 0.3789965811347938,
+      "learning_rate": 8.11159960496407e-06,
+      "loss": 0.0405,
+      "step": 3220
+    },
+    {
+      "epoch": 1.4306018210082168,
+      "grad_norm": 0.4496180621274656,
+      "learning_rate": 8.11008195629639e-06,
+      "loss": 0.0414,
+      "step": 3221
+    },
+    {
+      "epoch": 1.431045969353764,
+      "grad_norm": 0.6660820533473273,
+      "learning_rate": 8.10856384013056e-06,
+      "loss": 0.0519,
+      "step": 3222
+    },
+    {
+      "epoch": 1.4314901176993116,
+      "grad_norm": 0.42475615332155653,
+      "learning_rate": 8.107045256694782e-06,
+      "loss": 0.0507,
+      "step": 3223
+    },
+    {
+      "epoch": 1.431934266044859,
+      "grad_norm": 0.531367471068279,
+      "learning_rate": 8.105526206217322e-06,
+      "loss": 0.0438,
+      "step": 3224
+    },
+    {
+      "epoch": 1.4323784143904064,
+      "grad_norm": 0.6173452877900699,
+      "learning_rate": 8.104006688926518e-06,
+      "loss": 0.044,
+      "step": 3225
+    },
+    {
+      "epoch": 1.4328225627359539,
+      "grad_norm": 0.5047091534195701,
+      "learning_rate": 8.102486705050782e-06,
+      "loss": 0.0448,
+      "step": 3226
+    },
+    {
+      "epoch": 1.4332667110815012,
+      "grad_norm": 0.756671059089655,
+      "learning_rate": 8.100966254818591e-06,
+      "loss": 0.0472,
+      "step": 3227
+    },
+    {
+      "epoch": 1.4337108594270487,
+      "grad_norm": 0.6679241833390342,
+      "learning_rate": 8.099445338458496e-06,
+      "loss": 0.0428,
+      "step": 3228
+    },
+    {
+      "epoch": 1.434155007772596,
+      "grad_norm": 0.40517368896131173,
+      "learning_rate": 8.097923956199118e-06,
+      "loss": 0.0366,
+      "step": 3229
+    },
+    {
+      "epoch": 1.4345991561181435,
+      "grad_norm": 0.4490587246876595,
+      "learning_rate": 8.096402108269144e-06,
+      "loss": 0.0356,
+      "step": 3230
+    },
+    {
+      "epoch": 1.435043304463691,
+      "grad_norm": 0.5508426411635136,
+      "learning_rate": 8.094879794897333e-06,
+      "loss": 0.0499,
+      "step": 3231
+    },
+    {
+      "epoch": 1.4354874528092383,
+      "grad_norm": 0.39803990157367924,
+      "learning_rate": 8.093357016312518e-06,
+      "loss": 0.039,
+      "step": 3232
+    },
+    {
+      "epoch": 1.4359316011547856,
+      "grad_norm": 0.3911481714996281,
+      "learning_rate": 8.091833772743595e-06,
+      "loss": 0.0408,
+      "step": 3233
+    },
+    {
+      "epoch": 1.436375749500333,
+      "grad_norm": 0.43202710834535546,
+      "learning_rate": 8.090310064419536e-06,
+      "loss": 0.0517,
+      "step": 3234
+    },
+    {
+      "epoch": 1.4368198978458806,
+      "grad_norm": 0.63394524089963,
+      "learning_rate": 8.088785891569379e-06,
+      "loss": 0.0498,
+      "step": 3235
+    },
+    {
+      "epoch": 1.437264046191428,
+      "grad_norm": 0.6014431551871622,
+      "learning_rate": 8.087261254422232e-06,
+      "loss": 0.0553,
+      "step": 3236
+    },
+    {
+      "epoch": 1.4377081945369754,
+      "grad_norm": 0.6136207381555616,
+      "learning_rate": 8.085736153207277e-06,
+      "loss": 0.0594,
+      "step": 3237
+    },
+    {
+      "epoch": 1.4381523428825227,
+      "grad_norm": 0.5952470946539382,
+      "learning_rate": 8.08421058815376e-06,
+      "loss": 0.0769,
+      "step": 3238
+    },
+    {
+      "epoch": 1.4385964912280702,
+      "grad_norm": 0.5189933118580958,
+      "learning_rate": 8.082684559490999e-06,
+      "loss": 0.0554,
+      "step": 3239
+    },
+    {
+      "epoch": 1.4390406395736175,
+      "grad_norm": 0.4135426057188439,
+      "learning_rate": 8.081158067448385e-06,
+      "loss": 0.0396,
+      "step": 3240
+    },
+    {
+      "epoch": 1.439484787919165,
+      "grad_norm": 0.5443217663228829,
+      "learning_rate": 8.079631112255372e-06,
+      "loss": 0.044,
+      "step": 3241
+    },
+    {
+      "epoch": 1.4399289362647125,
+      "grad_norm": 0.5826766711918354,
+      "learning_rate": 8.078103694141487e-06,
+      "loss": 0.0446,
+      "step": 3242
+    },
+    {
+      "epoch": 1.4403730846102598,
+      "grad_norm": 0.648859138960752,
+      "learning_rate": 8.076575813336333e-06,
+      "loss": 0.0609,
+      "step": 3243
+    },
+    {
+      "epoch": 1.440817232955807,
+      "grad_norm": 0.4409431909915432,
+      "learning_rate": 8.07504747006957e-06,
+      "loss": 0.0531,
+      "step": 3244
+    },
+    {
+      "epoch": 1.4412613813013546,
+      "grad_norm": 0.3721919303810988,
+      "learning_rate": 8.073518664570938e-06,
+      "loss": 0.0386,
+      "step": 3245
+    },
+    {
+      "epoch": 1.4417055296469021,
+      "grad_norm": 0.43841622793791346,
+      "learning_rate": 8.07198939707024e-06,
+      "loss": 0.0493,
+      "step": 3246
+    },
+    {
+      "epoch": 1.4421496779924494,
+      "grad_norm": 0.5931350944063895,
+      "learning_rate": 8.070459667797351e-06,
+      "loss": 0.0634,
+      "step": 3247
+    },
+    {
+      "epoch": 1.442593826337997,
+      "grad_norm": 0.41937120746821677,
+      "learning_rate": 8.068929476982217e-06,
+      "loss": 0.0423,
+      "step": 3248
+    },
+    {
+      "epoch": 1.4430379746835442,
+      "grad_norm": 0.6050623457059973,
+      "learning_rate": 8.067398824854851e-06,
+      "loss": 0.07,
+      "step": 3249
+    },
+    {
+      "epoch": 1.4434821230290917,
+      "grad_norm": 0.36263255781004144,
+      "learning_rate": 8.065867711645334e-06,
+      "loss": 0.0488,
+      "step": 3250
+    },
+    {
+      "epoch": 1.443926271374639,
+      "grad_norm": 0.5452720169189981,
+      "learning_rate": 8.064336137583821e-06,
+      "loss": 0.0592,
+      "step": 3251
+    },
+    {
+      "epoch": 1.4443704197201865,
+      "grad_norm": 0.5648377578907434,
+      "learning_rate": 8.062804102900532e-06,
+      "loss": 0.045,
+      "step": 3252
+    },
+    {
+      "epoch": 1.444814568065734,
+      "grad_norm": 0.5203009129121725,
+      "learning_rate": 8.061271607825758e-06,
+      "loss": 0.0455,
+      "step": 3253
+    },
+    {
+      "epoch": 1.4452587164112813,
+      "grad_norm": 0.5388011169643779,
+      "learning_rate": 8.059738652589862e-06,
+      "loss": 0.0563,
+      "step": 3254
+    },
+    {
+      "epoch": 1.4457028647568289,
+      "grad_norm": 0.41394935328254073,
+      "learning_rate": 8.058205237423266e-06,
+      "loss": 0.0523,
+      "step": 3255
+    },
+    {
+      "epoch": 1.4461470131023761,
+      "grad_norm": 0.921759574953431,
+      "learning_rate": 8.056671362556476e-06,
+      "loss": 0.0356,
+      "step": 3256
+    },
+    {
+      "epoch": 1.4465911614479237,
+      "grad_norm": 0.8754840871870988,
+      "learning_rate": 8.055137028220058e-06,
+      "loss": 0.0615,
+      "step": 3257
+    },
+    {
+      "epoch": 1.447035309793471,
+      "grad_norm": 0.4878410786722894,
+      "learning_rate": 8.053602234644644e-06,
+      "loss": 0.0415,
+      "step": 3258
+    },
+    {
+      "epoch": 1.4474794581390185,
+      "grad_norm": 0.6534435017529012,
+      "learning_rate": 8.052066982060945e-06,
+      "loss": 0.0579,
+      "step": 3259
+    },
+    {
+      "epoch": 1.447923606484566,
+      "grad_norm": 0.9319839406156598,
+      "learning_rate": 8.050531270699731e-06,
+      "loss": 0.0484,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4483677548301133,
+      "grad_norm": 0.4011114975003397,
+      "learning_rate": 8.048995100791847e-06,
+      "loss": 0.0372,
+      "step": 3261
+    },
+    {
+      "epoch": 1.4488119031756606,
+      "grad_norm": 0.5604426931227581,
+      "learning_rate": 8.047458472568208e-06,
+      "loss": 0.0496,
+      "step": 3262
+    },
+    {
+      "epoch": 1.449256051521208,
+      "grad_norm": 0.3856602729544308,
+      "learning_rate": 8.045921386259792e-06,
+      "loss": 0.0368,
+      "step": 3263
+    },
+    {
+      "epoch": 1.4497001998667556,
+      "grad_norm": 0.37413170967076825,
+      "learning_rate": 8.044383842097651e-06,
+      "loss": 0.0426,
+      "step": 3264
+    },
+    {
+      "epoch": 1.4501443482123029,
+      "grad_norm": 0.4129802274607094,
+      "learning_rate": 8.042845840312903e-06,
+      "loss": 0.0398,
+      "step": 3265
+    },
+    {
+      "epoch": 1.4505884965578504,
+      "grad_norm": 0.8418363464980803,
+      "learning_rate": 8.041307381136738e-06,
+      "loss": 0.0656,
+      "step": 3266
+    },
+    {
+      "epoch": 1.4510326449033977,
+      "grad_norm": 0.5365298547106184,
+      "learning_rate": 8.039768464800408e-06,
+      "loss": 0.048,
+      "step": 3267
+    },
+    {
+      "epoch": 1.4514767932489452,
+      "grad_norm": 1.0687462882240213,
+      "learning_rate": 8.038229091535244e-06,
+      "loss": 0.0503,
+      "step": 3268
+    },
+    {
+      "epoch": 1.4519209415944925,
+      "grad_norm": 0.49239171264959003,
+      "learning_rate": 8.036689261572636e-06,
+      "loss": 0.0563,
+      "step": 3269
+    },
+    {
+      "epoch": 1.45236508994004,
+      "grad_norm": 0.5011792630285886,
+      "learning_rate": 8.035148975144046e-06,
+      "loss": 0.0396,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4528092382855875,
+      "grad_norm": 0.49603707035563677,
+      "learning_rate": 8.033608232481009e-06,
+      "loss": 0.0501,
+      "step": 3271
+    },
+    {
+      "epoch": 1.4532533866311348,
+      "grad_norm": 0.5939987053954954,
+      "learning_rate": 8.032067033815123e-06,
+      "loss": 0.0458,
+      "step": 3272
+    },
+    {
+      "epoch": 1.453697534976682,
+      "grad_norm": 0.5812600520648006,
+      "learning_rate": 8.030525379378053e-06,
+      "loss": 0.0447,
+      "step": 3273
+    },
+    {
+      "epoch": 1.4541416833222296,
+      "grad_norm": 0.6480809819716409,
+      "learning_rate": 8.028983269401542e-06,
+      "loss": 0.0608,
+      "step": 3274
+    },
+    {
+      "epoch": 1.454585831667777,
+      "grad_norm": 0.5090480104564535,
+      "learning_rate": 8.027440704117391e-06,
+      "loss": 0.0559,
+      "step": 3275
+    },
+    {
+      "epoch": 1.4550299800133244,
+      "grad_norm": 0.6602628376919341,
+      "learning_rate": 8.025897683757473e-06,
+      "loss": 0.0587,
+      "step": 3276
+    },
+    {
+      "epoch": 1.455474128358872,
+      "grad_norm": 0.5252379759239632,
+      "learning_rate": 8.024354208553735e-06,
+      "loss": 0.061,
+      "step": 3277
+    },
+    {
+      "epoch": 1.4559182767044192,
+      "grad_norm": 0.4725375990607396,
+      "learning_rate": 8.022810278738185e-06,
+      "loss": 0.0408,
+      "step": 3278
+    },
+    {
+      "epoch": 1.4563624250499667,
+      "grad_norm": 0.5401711443633535,
+      "learning_rate": 8.021265894542898e-06,
+      "loss": 0.0667,
+      "step": 3279
+    },
+    {
+      "epoch": 1.456806573395514,
+      "grad_norm": 0.5352342030628493,
+      "learning_rate": 8.019721056200027e-06,
+      "loss": 0.0458,
+      "step": 3280
+    },
+    {
+      "epoch": 1.4572507217410615,
+      "grad_norm": 0.4210073842896914,
+      "learning_rate": 8.018175763941784e-06,
+      "loss": 0.0437,
+      "step": 3281
+    },
+    {
+      "epoch": 1.457694870086609,
+      "grad_norm": 0.3904193074228819,
+      "learning_rate": 8.016630018000457e-06,
+      "loss": 0.0456,
+      "step": 3282
+    },
+    {
+      "epoch": 1.4581390184321563,
+      "grad_norm": 0.4133159146388692,
+      "learning_rate": 8.015083818608393e-06,
+      "loss": 0.0385,
+      "step": 3283
+    },
+    {
+      "epoch": 1.4585831667777038,
+      "grad_norm": 0.5275288891932939,
+      "learning_rate": 8.013537165998014e-06,
+      "loss": 0.0477,
+      "step": 3284
+    },
+    {
+      "epoch": 1.4590273151232511,
+      "grad_norm": 0.43540224877298483,
+      "learning_rate": 8.011990060401806e-06,
+      "loss": 0.0433,
+      "step": 3285
+    },
+    {
+      "epoch": 1.4594714634687986,
+      "grad_norm": 0.476963740475049,
+      "learning_rate": 8.010442502052329e-06,
+      "loss": 0.0549,
+      "step": 3286
+    },
+    {
+      "epoch": 1.459915611814346,
+      "grad_norm": 0.4927774183701049,
+      "learning_rate": 8.008894491182205e-06,
+      "loss": 0.0464,
+      "step": 3287
+    },
+    {
+      "epoch": 1.4603597601598934,
+      "grad_norm": 0.4716288282663322,
+      "learning_rate": 8.007346028024125e-06,
+      "loss": 0.0437,
+      "step": 3288
+    },
+    {
+      "epoch": 1.460803908505441,
+      "grad_norm": 0.5277979981560662,
+      "learning_rate": 8.005797112810854e-06,
+      "loss": 0.0452,
+      "step": 3289
+    },
+    {
+      "epoch": 1.4612480568509882,
+      "grad_norm": 0.6486682818717127,
+      "learning_rate": 8.004247745775216e-06,
+      "loss": 0.0636,
+      "step": 3290
+    },
+    {
+      "epoch": 1.4616922051965355,
+      "grad_norm": 0.693590181694553,
+      "learning_rate": 8.00269792715011e-06,
+      "loss": 0.0392,
+      "step": 3291
+    },
+    {
+      "epoch": 1.462136353542083,
+      "grad_norm": 0.4973813779816927,
+      "learning_rate": 8.001147657168497e-06,
+      "loss": 0.0523,
+      "step": 3292
+    },
+    {
+      "epoch": 1.4625805018876306,
+      "grad_norm": 0.42880946574568146,
+      "learning_rate": 7.99959693606341e-06,
+      "loss": 0.0365,
+      "step": 3293
+    },
+    {
+      "epoch": 1.4630246502331778,
+      "grad_norm": 0.4647071982633707,
+      "learning_rate": 7.99804576406795e-06,
+      "loss": 0.0473,
+      "step": 3294
+    },
+    {
+      "epoch": 1.4634687985787254,
+      "grad_norm": 0.5564296867063627,
+      "learning_rate": 7.996494141415284e-06,
+      "loss": 0.0654,
+      "step": 3295
+    },
+    {
+      "epoch": 1.4639129469242727,
+      "grad_norm": 0.4602454467326126,
+      "learning_rate": 7.994942068338647e-06,
+      "loss": 0.0375,
+      "step": 3296
+    },
+    {
+      "epoch": 1.4643570952698202,
+      "grad_norm": 0.44240678049891874,
+      "learning_rate": 7.993389545071341e-06,
+      "loss": 0.0538,
+      "step": 3297
+    },
+    {
+      "epoch": 1.4648012436153675,
+      "grad_norm": 0.5956552998865925,
+      "learning_rate": 7.991836571846739e-06,
+      "loss": 0.0483,
+      "step": 3298
+    },
+    {
+      "epoch": 1.465245391960915,
+      "grad_norm": 0.8272941029925692,
+      "learning_rate": 7.990283148898277e-06,
+      "loss": 0.0621,
+      "step": 3299
+    },
+    {
+      "epoch": 1.4656895403064625,
+      "grad_norm": 0.521288053624164,
+      "learning_rate": 7.988729276459463e-06,
+      "loss": 0.0389,
+      "step": 3300
+    },
+    {
+      "epoch": 1.4661336886520098,
+      "grad_norm": 0.3918246593118355,
+      "learning_rate": 7.987174954763867e-06,
+      "loss": 0.0351,
+      "step": 3301
+    },
+    {
+      "epoch": 1.466577836997557,
+      "grad_norm": 0.5695857738579163,
+      "learning_rate": 7.985620184045133e-06,
+      "loss": 0.051,
+      "step": 3302
+    },
+    {
+      "epoch": 1.4670219853431046,
+      "grad_norm": 0.5461168059678839,
+      "learning_rate": 7.98406496453697e-06,
+      "loss": 0.045,
+      "step": 3303
+    },
+    {
+      "epoch": 1.467466133688652,
+      "grad_norm": 0.45316446756220935,
+      "learning_rate": 7.982509296473151e-06,
+      "loss": 0.0376,
+      "step": 3304
+    },
+    {
+      "epoch": 1.4679102820341994,
+      "grad_norm": 0.5116686497553603,
+      "learning_rate": 7.98095318008752e-06,
+      "loss": 0.0523,
+      "step": 3305
+    },
+    {
+      "epoch": 1.4683544303797469,
+      "grad_norm": 0.5042517448525269,
+      "learning_rate": 7.97939661561399e-06,
+      "loss": 0.0473,
+      "step": 3306
+    },
+    {
+      "epoch": 1.4687985787252942,
+      "grad_norm": 0.5555029036352702,
+      "learning_rate": 7.977839603286537e-06,
+      "loss": 0.053,
+      "step": 3307
+    },
+    {
+      "epoch": 1.4692427270708417,
+      "grad_norm": 0.3410097103369562,
+      "learning_rate": 7.976282143339207e-06,
+      "loss": 0.0293,
+      "step": 3308
+    },
+    {
+      "epoch": 1.469686875416389,
+      "grad_norm": 0.6134381443649801,
+      "learning_rate": 7.974724236006113e-06,
+      "loss": 0.0528,
+      "step": 3309
+    },
+    {
+      "epoch": 1.4701310237619365,
+      "grad_norm": 0.4165723941603629,
+      "learning_rate": 7.973165881521435e-06,
+      "loss": 0.041,
+      "step": 3310
+    },
+    {
+      "epoch": 1.470575172107484,
+      "grad_norm": 0.4357045233722827,
+      "learning_rate": 7.971607080119418e-06,
+      "loss": 0.0499,
+      "step": 3311
+    },
+    {
+      "epoch": 1.4710193204530313,
+      "grad_norm": 0.5469760548909597,
+      "learning_rate": 7.97004783203438e-06,
+      "loss": 0.0436,
+      "step": 3312
+    },
+    {
+      "epoch": 1.4714634687985786,
+      "grad_norm": 0.4196612097876158,
+      "learning_rate": 7.968488137500699e-06,
+      "loss": 0.0474,
+      "step": 3313
+    },
+    {
+      "epoch": 1.471907617144126,
+      "grad_norm": 0.4878242528103949,
+      "learning_rate": 7.966927996752824e-06,
+      "loss": 0.0407,
+      "step": 3314
+    },
+    {
+      "epoch": 1.4723517654896736,
+      "grad_norm": 0.4654212754221821,
+      "learning_rate": 7.965367410025275e-06,
+      "loss": 0.0276,
+      "step": 3315
+    },
+    {
+      "epoch": 1.472795913835221,
+      "grad_norm": 0.4496002039864948,
+      "learning_rate": 7.96380637755263e-06,
+      "loss": 0.043,
+      "step": 3316
+    },
+    {
+      "epoch": 1.4732400621807684,
+      "grad_norm": 0.843662279100629,
+      "learning_rate": 7.96224489956954e-06,
+      "loss": 0.0507,
+      "step": 3317
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.32395601492358206,
+      "learning_rate": 7.960682976310721e-06,
+      "loss": 0.0356,
+      "step": 3318
+    },
+    {
+      "epoch": 1.4741283588718632,
+      "grad_norm": 0.6598994798652713,
+      "learning_rate": 7.959120608010959e-06,
+      "loss": 0.0532,
+      "step": 3319
+    },
+    {
+      "epoch": 1.4745725072174105,
+      "grad_norm": 0.6135947839100228,
+      "learning_rate": 7.957557794905104e-06,
+      "loss": 0.0568,
+      "step": 3320
+    },
+    {
+      "epoch": 1.475016655562958,
+      "grad_norm": 0.5386416254084523,
+      "learning_rate": 7.955994537228068e-06,
+      "loss": 0.0524,
+      "step": 3321
+    },
+    {
+      "epoch": 1.4754608039085055,
+      "grad_norm": 0.41178054802832886,
+      "learning_rate": 7.954430835214844e-06,
+      "loss": 0.0478,
+      "step": 3322
+    },
+    {
+      "epoch": 1.4759049522540528,
+      "grad_norm": 0.553965746901051,
+      "learning_rate": 7.952866689100476e-06,
+      "loss": 0.0497,
+      "step": 3323
+    },
+    {
+      "epoch": 1.4763491005996003,
+      "grad_norm": 0.5002040941699343,
+      "learning_rate": 7.951302099120087e-06,
+      "loss": 0.0479,
+      "step": 3324
+    },
+    {
+      "epoch": 1.4767932489451476,
+      "grad_norm": 0.5774527696932492,
+      "learning_rate": 7.949737065508856e-06,
+      "loss": 0.0447,
+      "step": 3325
+    },
+    {
+      "epoch": 1.4772373972906951,
+      "grad_norm": 0.5834169178186834,
+      "learning_rate": 7.948171588502036e-06,
+      "loss": 0.0442,
+      "step": 3326
+    },
+    {
+      "epoch": 1.4776815456362424,
+      "grad_norm": 0.48069558429150877,
+      "learning_rate": 7.946605668334947e-06,
+      "loss": 0.0509,
+      "step": 3327
+    },
+    {
+      "epoch": 1.47812569398179,
+      "grad_norm": 0.6916569805639394,
+      "learning_rate": 7.945039305242972e-06,
+      "loss": 0.0543,
+      "step": 3328
+    },
+    {
+      "epoch": 1.4785698423273375,
+      "grad_norm": 0.45596286134877817,
+      "learning_rate": 7.943472499461562e-06,
+      "loss": 0.0463,
+      "step": 3329
+    },
+    {
+      "epoch": 1.4790139906728847,
+      "grad_norm": 0.7085391238735931,
+      "learning_rate": 7.941905251226235e-06,
+      "loss": 0.0643,
+      "step": 3330
+    },
+    {
+      "epoch": 1.479458139018432,
+      "grad_norm": 0.878574496425455,
+      "learning_rate": 7.940337560772573e-06,
+      "loss": 0.0623,
+      "step": 3331
+    },
+    {
+      "epoch": 1.4799022873639796,
+      "grad_norm": 0.3602452580945399,
+      "learning_rate": 7.93876942833623e-06,
+      "loss": 0.0325,
+      "step": 3332
+    },
+    {
+      "epoch": 1.480346435709527,
+      "grad_norm": 0.5694019432619376,
+      "learning_rate": 7.937200854152917e-06,
+      "loss": 0.0527,
+      "step": 3333
+    },
+    {
+      "epoch": 1.4807905840550744,
+      "grad_norm": 0.46397062738751477,
+      "learning_rate": 7.935631838458426e-06,
+      "loss": 0.0435,
+      "step": 3334
+    },
+    {
+      "epoch": 1.4812347324006219,
+      "grad_norm": 0.38977994271678623,
+      "learning_rate": 7.9340623814886e-06,
+      "loss": 0.0346,
+      "step": 3335
+    },
+    {
+      "epoch": 1.4816788807461692,
+      "grad_norm": 0.5991214620269069,
+      "learning_rate": 7.932492483479358e-06,
+      "loss": 0.0519,
+      "step": 3336
+    },
+    {
+      "epoch": 1.4821230290917167,
+      "grad_norm": 0.42521594572783633,
+      "learning_rate": 7.930922144666679e-06,
+      "loss": 0.0407,
+      "step": 3337
+    },
+    {
+      "epoch": 1.482567177437264,
+      "grad_norm": 0.6068474747842975,
+      "learning_rate": 7.929351365286614e-06,
+      "loss": 0.0569,
+      "step": 3338
+    },
+    {
+      "epoch": 1.4830113257828115,
+      "grad_norm": 0.3673998638748444,
+      "learning_rate": 7.927780145575281e-06,
+      "loss": 0.0398,
+      "step": 3339
+    },
+    {
+      "epoch": 1.483455474128359,
+      "grad_norm": 0.6378602388998965,
+      "learning_rate": 7.926208485768856e-06,
+      "loss": 0.0445,
+      "step": 3340
+    },
+    {
+      "epoch": 1.4838996224739063,
+      "grad_norm": 0.4635352438588869,
+      "learning_rate": 7.924636386103588e-06,
+      "loss": 0.0472,
+      "step": 3341
+    },
+    {
+      "epoch": 1.4843437708194536,
+      "grad_norm": 0.43741344875712923,
+      "learning_rate": 7.923063846815791e-06,
+      "loss": 0.0365,
+      "step": 3342
+    },
+    {
+      "epoch": 1.484787919165001,
+      "grad_norm": 0.409274641921781,
+      "learning_rate": 7.921490868141843e-06,
+      "loss": 0.0437,
+      "step": 3343
+    },
+    {
+      "epoch": 1.4852320675105486,
+      "grad_norm": 0.44864312363339953,
+      "learning_rate": 7.91991745031819e-06,
+      "loss": 0.0365,
+      "step": 3344
+    },
+    {
+      "epoch": 1.4856762158560959,
+      "grad_norm": 0.7767506869882325,
+      "learning_rate": 7.918343593581344e-06,
+      "loss": 0.055,
+      "step": 3345
+    },
+    {
+      "epoch": 1.4861203642016434,
+      "grad_norm": 0.663893547353565,
+      "learning_rate": 7.916769298167881e-06,
+      "loss": 0.0539,
+      "step": 3346
+    },
+    {
+      "epoch": 1.4865645125471907,
+      "grad_norm": 0.44134070690444965,
+      "learning_rate": 7.915194564314446e-06,
+      "loss": 0.0451,
+      "step": 3347
+    },
+    {
+      "epoch": 1.4870086608927382,
+      "grad_norm": 0.5644812574628542,
+      "learning_rate": 7.913619392257748e-06,
+      "loss": 0.052,
+      "step": 3348
+    },
+    {
+      "epoch": 1.4874528092382855,
+      "grad_norm": 0.4293480800359804,
+      "learning_rate": 7.912043782234562e-06,
+      "loss": 0.0375,
+      "step": 3349
+    },
+    {
+      "epoch": 1.487896957583833,
+      "grad_norm": 0.4208900641552744,
+      "learning_rate": 7.910467734481726e-06,
+      "loss": 0.0263,
+      "step": 3350
+    },
+    {
+      "epoch": 1.4883411059293805,
+      "grad_norm": 0.5686191171590155,
+      "learning_rate": 7.90889124923615e-06,
+      "loss": 0.0552,
+      "step": 3351
+    },
+    {
+      "epoch": 1.4887852542749278,
+      "grad_norm": 0.45995768523750874,
+      "learning_rate": 7.907314326734807e-06,
+      "loss": 0.0392,
+      "step": 3352
+    },
+    {
+      "epoch": 1.4892294026204753,
+      "grad_norm": 0.6197595746880575,
+      "learning_rate": 7.905736967214735e-06,
+      "loss": 0.055,
+      "step": 3353
+    },
+    {
+      "epoch": 1.4896735509660226,
+      "grad_norm": 0.6356786842165147,
+      "learning_rate": 7.904159170913035e-06,
+      "loss": 0.046,
+      "step": 3354
+    },
+    {
+      "epoch": 1.4901176993115701,
+      "grad_norm": 0.4601859054406351,
+      "learning_rate": 7.902580938066878e-06,
+      "loss": 0.0414,
+      "step": 3355
+    },
+    {
+      "epoch": 1.4905618476571174,
+      "grad_norm": 0.36553812505176564,
+      "learning_rate": 7.901002268913501e-06,
+      "loss": 0.0322,
+      "step": 3356
+    },
+    {
+      "epoch": 1.491005996002665,
+      "grad_norm": 0.615226798089253,
+      "learning_rate": 7.899423163690204e-06,
+      "loss": 0.0549,
+      "step": 3357
+    },
+    {
+      "epoch": 1.4914501443482124,
+      "grad_norm": 0.4095361485944277,
+      "learning_rate": 7.897843622634352e-06,
+      "loss": 0.0387,
+      "step": 3358
+    },
+    {
+      "epoch": 1.4918942926937597,
+      "grad_norm": 0.5006702378564764,
+      "learning_rate": 7.896263645983378e-06,
+      "loss": 0.0517,
+      "step": 3359
+    },
+    {
+      "epoch": 1.492338441039307,
+      "grad_norm": 0.6244366583571412,
+      "learning_rate": 7.89468323397478e-06,
+      "loss": 0.0496,
+      "step": 3360
+    },
+    {
+      "epoch": 1.4927825893848545,
+      "grad_norm": 0.5367025221002972,
+      "learning_rate": 7.893102386846118e-06,
+      "loss": 0.0436,
+      "step": 3361
+    },
+    {
+      "epoch": 1.493226737730402,
+      "grad_norm": 0.4508013926132327,
+      "learning_rate": 7.891521104835023e-06,
+      "loss": 0.0344,
+      "step": 3362
+    },
+    {
+      "epoch": 1.4936708860759493,
+      "grad_norm": 0.6393470972778723,
+      "learning_rate": 7.889939388179188e-06,
+      "loss": 0.0526,
+      "step": 3363
+    },
+    {
+      "epoch": 1.4941150344214968,
+      "grad_norm": 0.472940776234741,
+      "learning_rate": 7.888357237116372e-06,
+      "loss": 0.0431,
+      "step": 3364
+    },
+    {
+      "epoch": 1.4945591827670441,
+      "grad_norm": 0.4993044626375099,
+      "learning_rate": 7.886774651884397e-06,
+      "loss": 0.0541,
+      "step": 3365
+    },
+    {
+      "epoch": 1.4950033311125916,
+      "grad_norm": 0.4991379094630535,
+      "learning_rate": 7.885191632721156e-06,
+      "loss": 0.0423,
+      "step": 3366
+    },
+    {
+      "epoch": 1.495447479458139,
+      "grad_norm": 0.6104766035319223,
+      "learning_rate": 7.8836081798646e-06,
+      "loss": 0.0489,
+      "step": 3367
+    },
+    {
+      "epoch": 1.4958916278036865,
+      "grad_norm": 0.582209332395513,
+      "learning_rate": 7.882024293552752e-06,
+      "loss": 0.0481,
+      "step": 3368
+    },
+    {
+      "epoch": 1.496335776149234,
+      "grad_norm": 1.2878711444761508,
+      "learning_rate": 7.880439974023694e-06,
+      "loss": 0.0895,
+      "step": 3369
+    },
+    {
+      "epoch": 1.4967799244947813,
+      "grad_norm": 0.47752757971222215,
+      "learning_rate": 7.87885522151558e-06,
+      "loss": 0.0556,
+      "step": 3370
+    },
+    {
+      "epoch": 1.4972240728403285,
+      "grad_norm": 0.34960537720365853,
+      "learning_rate": 7.877270036266622e-06,
+      "loss": 0.0361,
+      "step": 3371
+    },
+    {
+      "epoch": 1.497668221185876,
+      "grad_norm": 0.6171844191638863,
+      "learning_rate": 7.875684418515101e-06,
+      "loss": 0.0413,
+      "step": 3372
+    },
+    {
+      "epoch": 1.4981123695314236,
+      "grad_norm": 0.5888605884776164,
+      "learning_rate": 7.874098368499362e-06,
+      "loss": 0.0433,
+      "step": 3373
+    },
+    {
+      "epoch": 1.4985565178769709,
+      "grad_norm": 0.32860735680481784,
+      "learning_rate": 7.872511886457816e-06,
+      "loss": 0.029,
+      "step": 3374
+    },
+    {
+      "epoch": 1.4990006662225184,
+      "grad_norm": 0.4838256059081808,
+      "learning_rate": 7.87092497262894e-06,
+      "loss": 0.0428,
+      "step": 3375
+    },
+    {
+      "epoch": 1.4994448145680657,
+      "grad_norm": 0.5961233345859986,
+      "learning_rate": 7.86933762725127e-06,
+      "loss": 0.0568,
+      "step": 3376
+    },
+    {
+      "epoch": 1.4998889629136132,
+      "grad_norm": 0.4985166006472107,
+      "learning_rate": 7.867749850563414e-06,
+      "loss": 0.0539,
+      "step": 3377
+    },
+    {
+      "epoch": 1.5003331112591605,
+      "grad_norm": 0.8288419788268757,
+      "learning_rate": 7.86616164280404e-06,
+      "loss": 0.0376,
+      "step": 3378
+    },
+    {
+      "epoch": 1.500777259604708,
+      "grad_norm": 0.5437911475680088,
+      "learning_rate": 7.864573004211884e-06,
+      "loss": 0.0629,
+      "step": 3379
+    },
+    {
+      "epoch": 1.5012214079502555,
+      "grad_norm": 0.5173272400012955,
+      "learning_rate": 7.862983935025745e-06,
+      "loss": 0.0496,
+      "step": 3380
+    },
+    {
+      "epoch": 1.5016655562958028,
+      "grad_norm": 0.41368737678867235,
+      "learning_rate": 7.861394435484488e-06,
+      "loss": 0.0428,
+      "step": 3381
+    },
+    {
+      "epoch": 1.50210970464135,
+      "grad_norm": 0.382287625598364,
+      "learning_rate": 7.85980450582704e-06,
+      "loss": 0.0362,
+      "step": 3382
+    },
+    {
+      "epoch": 1.5025538529868976,
+      "grad_norm": 0.7914574257051636,
+      "learning_rate": 7.858214146292394e-06,
+      "loss": 0.0723,
+      "step": 3383
+    },
+    {
+      "epoch": 1.502998001332445,
+      "grad_norm": 0.600368806153385,
+      "learning_rate": 7.85662335711961e-06,
+      "loss": 0.0426,
+      "step": 3384
+    },
+    {
+      "epoch": 1.5034421496779924,
+      "grad_norm": 0.516289015563682,
+      "learning_rate": 7.855032138547811e-06,
+      "loss": 0.0401,
+      "step": 3385
+    },
+    {
+      "epoch": 1.50388629802354,
+      "grad_norm": 0.4579999933485943,
+      "learning_rate": 7.853440490816182e-06,
+      "loss": 0.0372,
+      "step": 3386
+    },
+    {
+      "epoch": 1.5043304463690874,
+      "grad_norm": 0.4924686770621345,
+      "learning_rate": 7.851848414163976e-06,
+      "loss": 0.0457,
+      "step": 3387
+    },
+    {
+      "epoch": 1.5047745947146347,
+      "grad_norm": 0.5624519083004446,
+      "learning_rate": 7.850255908830508e-06,
+      "loss": 0.0431,
+      "step": 3388
+    },
+    {
+      "epoch": 1.505218743060182,
+      "grad_norm": 0.46856624335860086,
+      "learning_rate": 7.848662975055161e-06,
+      "loss": 0.0374,
+      "step": 3389
+    },
+    {
+      "epoch": 1.5056628914057295,
+      "grad_norm": 0.6866568127731671,
+      "learning_rate": 7.847069613077377e-06,
+      "loss": 0.088,
+      "step": 3390
+    },
+    {
+      "epoch": 1.506107039751277,
+      "grad_norm": 0.5263852322647139,
+      "learning_rate": 7.845475823136669e-06,
+      "loss": 0.0475,
+      "step": 3391
+    },
+    {
+      "epoch": 1.5065511880968243,
+      "grad_norm": 0.6378963066155895,
+      "learning_rate": 7.843881605472606e-06,
+      "loss": 0.0579,
+      "step": 3392
+    },
+    {
+      "epoch": 1.5069953364423716,
+      "grad_norm": 0.8470563699341165,
+      "learning_rate": 7.84228696032483e-06,
+      "loss": 0.0685,
+      "step": 3393
+    },
+    {
+      "epoch": 1.5074394847879191,
+      "grad_norm": 0.5121438548331829,
+      "learning_rate": 7.840691887933042e-06,
+      "loss": 0.0482,
+      "step": 3394
+    },
+    {
+      "epoch": 1.5078836331334666,
+      "grad_norm": 0.46761061686619343,
+      "learning_rate": 7.839096388537008e-06,
+      "loss": 0.053,
+      "step": 3395
+    },
+    {
+      "epoch": 1.508327781479014,
+      "grad_norm": 0.568716582017773,
+      "learning_rate": 7.837500462376559e-06,
+      "loss": 0.0446,
+      "step": 3396
+    },
+    {
+      "epoch": 1.5087719298245614,
+      "grad_norm": 0.427488840303415,
+      "learning_rate": 7.83590410969159e-06,
+      "loss": 0.045,
+      "step": 3397
+    },
+    {
+      "epoch": 1.509216078170109,
+      "grad_norm": 0.5336130953887138,
+      "learning_rate": 7.834307330722059e-06,
+      "loss": 0.0501,
+      "step": 3398
+    },
+    {
+      "epoch": 1.5096602265156562,
+      "grad_norm": 0.6643085395902935,
+      "learning_rate": 7.832710125707991e-06,
+      "loss": 0.0714,
+      "step": 3399
+    },
+    {
+      "epoch": 1.5101043748612035,
+      "grad_norm": 0.7517073070810807,
+      "learning_rate": 7.831112494889472e-06,
+      "loss": 0.0574,
+      "step": 3400
+    },
+    {
+      "epoch": 1.510548523206751,
+      "grad_norm": 0.6793225641094393,
+      "learning_rate": 7.829514438506651e-06,
+      "loss": 0.054,
+      "step": 3401
+    },
+    {
+      "epoch": 1.5109926715522985,
+      "grad_norm": 0.4505747895513854,
+      "learning_rate": 7.827915956799745e-06,
+      "loss": 0.0417,
+      "step": 3402
+    },
+    {
+      "epoch": 1.5114368198978458,
+      "grad_norm": 0.7701556350457059,
+      "learning_rate": 7.826317050009035e-06,
+      "loss": 0.046,
+      "step": 3403
+    },
+    {
+      "epoch": 1.5118809682433931,
+      "grad_norm": 0.4711452448196391,
+      "learning_rate": 7.82471771837486e-06,
+      "loss": 0.0393,
+      "step": 3404
+    },
+    {
+      "epoch": 1.5123251165889409,
+      "grad_norm": 0.5955356460688755,
+      "learning_rate": 7.823117962137628e-06,
+      "loss": 0.0475,
+      "step": 3405
+    },
+    {
+      "epoch": 1.5127692649344882,
+      "grad_norm": 0.47383670753602875,
+      "learning_rate": 7.821517781537811e-06,
+      "loss": 0.0471,
+      "step": 3406
+    },
+    {
+      "epoch": 1.5132134132800354,
+      "grad_norm": 0.36906733260563584,
+      "learning_rate": 7.819917176815942e-06,
+      "loss": 0.0424,
+      "step": 3407
+    },
+    {
+      "epoch": 1.513657561625583,
+      "grad_norm": 0.5026639235608937,
+      "learning_rate": 7.818316148212619e-06,
+      "loss": 0.0478,
+      "step": 3408
+    },
+    {
+      "epoch": 1.5141017099711305,
+      "grad_norm": 0.5632301347650163,
+      "learning_rate": 7.816714695968503e-06,
+      "loss": 0.0582,
+      "step": 3409
+    },
+    {
+      "epoch": 1.5145458583166778,
+      "grad_norm": 0.5601350342899486,
+      "learning_rate": 7.815112820324322e-06,
+      "loss": 0.052,
+      "step": 3410
+    },
+    {
+      "epoch": 1.514990006662225,
+      "grad_norm": 0.44545554824483924,
+      "learning_rate": 7.813510521520864e-06,
+      "loss": 0.0437,
+      "step": 3411
+    },
+    {
+      "epoch": 1.5154341550077726,
+      "grad_norm": 0.5321000678999631,
+      "learning_rate": 7.811907799798981e-06,
+      "loss": 0.0446,
+      "step": 3412
+    },
+    {
+      "epoch": 1.51587830335332,
+      "grad_norm": 0.6123306795917254,
+      "learning_rate": 7.81030465539959e-06,
+      "loss": 0.0472,
+      "step": 3413
+    },
+    {
+      "epoch": 1.5163224516988674,
+      "grad_norm": 0.7182097530608341,
+      "learning_rate": 7.808701088563669e-06,
+      "loss": 0.0652,
+      "step": 3414
+    },
+    {
+      "epoch": 1.5167666000444149,
+      "grad_norm": 0.6740358195253988,
+      "learning_rate": 7.807097099532264e-06,
+      "loss": 0.0446,
+      "step": 3415
+    },
+    {
+      "epoch": 1.5172107483899624,
+      "grad_norm": 0.43545100676950116,
+      "learning_rate": 7.805492688546481e-06,
+      "loss": 0.0327,
+      "step": 3416
+    },
+    {
+      "epoch": 1.5176548967355097,
+      "grad_norm": 0.4493644872097168,
+      "learning_rate": 7.80388785584749e-06,
+      "loss": 0.0401,
+      "step": 3417
+    },
+    {
+      "epoch": 1.518099045081057,
+      "grad_norm": 0.4270665812232271,
+      "learning_rate": 7.802282601676522e-06,
+      "loss": 0.0517,
+      "step": 3418
+    },
+    {
+      "epoch": 1.5185431934266045,
+      "grad_norm": 0.48691036016457884,
+      "learning_rate": 7.800676926274881e-06,
+      "loss": 0.0398,
+      "step": 3419
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.5623140589479151,
+      "learning_rate": 7.79907082988392e-06,
+      "loss": 0.0482,
+      "step": 3420
+    },
+    {
+      "epoch": 1.5194314901176993,
+      "grad_norm": 0.6102149116754874,
+      "learning_rate": 7.797464312745067e-06,
+      "loss": 0.0541,
+      "step": 3421
+    },
+    {
+      "epoch": 1.5198756384632466,
+      "grad_norm": 0.5623605927044153,
+      "learning_rate": 7.795857375099806e-06,
+      "loss": 0.05,
+      "step": 3422
+    },
+    {
+      "epoch": 1.520319786808794,
+      "grad_norm": 0.4402109113654038,
+      "learning_rate": 7.794250017189689e-06,
+      "loss": 0.046,
+      "step": 3423
+    },
+    {
+      "epoch": 1.5207639351543416,
+      "grad_norm": 0.48101120541164943,
+      "learning_rate": 7.792642239256327e-06,
+      "loss": 0.0461,
+      "step": 3424
+    },
+    {
+      "epoch": 1.521208083499889,
+      "grad_norm": 0.5296531840354222,
+      "learning_rate": 7.791034041541398e-06,
+      "loss": 0.0473,
+      "step": 3425
+    },
+    {
+      "epoch": 1.5216522318454364,
+      "grad_norm": 0.37685452582544965,
+      "learning_rate": 7.78942542428664e-06,
+      "loss": 0.0317,
+      "step": 3426
+    },
+    {
+      "epoch": 1.522096380190984,
+      "grad_norm": 0.3991510986249673,
+      "learning_rate": 7.78781638773386e-06,
+      "loss": 0.0347,
+      "step": 3427
+    },
+    {
+      "epoch": 1.5225405285365312,
+      "grad_norm": 0.7554838302963066,
+      "learning_rate": 7.786206932124918e-06,
+      "loss": 0.0586,
+      "step": 3428
+    },
+    {
+      "epoch": 1.5229846768820785,
+      "grad_norm": 0.42032863329599834,
+      "learning_rate": 7.784597057701745e-06,
+      "loss": 0.0252,
+      "step": 3429
+    },
+    {
+      "epoch": 1.523428825227626,
+      "grad_norm": 0.5551436730522737,
+      "learning_rate": 7.782986764706334e-06,
+      "loss": 0.0563,
+      "step": 3430
+    },
+    {
+      "epoch": 1.5238729735731735,
+      "grad_norm": 0.6187970153795604,
+      "learning_rate": 7.781376053380735e-06,
+      "loss": 0.0592,
+      "step": 3431
+    },
+    {
+      "epoch": 1.5243171219187208,
+      "grad_norm": 0.5501189193555737,
+      "learning_rate": 7.779764923967069e-06,
+      "loss": 0.0556,
+      "step": 3432
+    },
+    {
+      "epoch": 1.524761270264268,
+      "grad_norm": 0.44533955935715447,
+      "learning_rate": 7.778153376707513e-06,
+      "loss": 0.0507,
+      "step": 3433
+    },
+    {
+      "epoch": 1.5252054186098158,
+      "grad_norm": 0.4340430082400607,
+      "learning_rate": 7.776541411844315e-06,
+      "loss": 0.0385,
+      "step": 3434
+    },
+    {
+      "epoch": 1.5256495669553631,
+      "grad_norm": 0.5599425444294093,
+      "learning_rate": 7.774929029619775e-06,
+      "loss": 0.0454,
+      "step": 3435
+    },
+    {
+      "epoch": 1.5260937153009104,
+      "grad_norm": 0.3908201446163151,
+      "learning_rate": 7.773316230276267e-06,
+      "loss": 0.041,
+      "step": 3436
+    },
+    {
+      "epoch": 1.526537863646458,
+      "grad_norm": 0.6432918655836172,
+      "learning_rate": 7.771703014056217e-06,
+      "loss": 0.0664,
+      "step": 3437
+    },
+    {
+      "epoch": 1.5269820119920055,
+      "grad_norm": 0.4341973490141778,
+      "learning_rate": 7.770089381202121e-06,
+      "loss": 0.0536,
+      "step": 3438
+    },
+    {
+      "epoch": 1.5274261603375527,
+      "grad_norm": 0.5232821809072614,
+      "learning_rate": 7.768475331956537e-06,
+      "loss": 0.0506,
+      "step": 3439
+    },
+    {
+      "epoch": 1.5278703086831,
+      "grad_norm": 0.49831548658076835,
+      "learning_rate": 7.76686086656208e-06,
+      "loss": 0.0497,
+      "step": 3440
+    },
+    {
+      "epoch": 1.5283144570286475,
+      "grad_norm": 0.6350239638046895,
+      "learning_rate": 7.765245985261436e-06,
+      "loss": 0.0603,
+      "step": 3441
+    },
+    {
+      "epoch": 1.528758605374195,
+      "grad_norm": 0.8205513486577842,
+      "learning_rate": 7.763630688297347e-06,
+      "loss": 0.0555,
+      "step": 3442
+    },
+    {
+      "epoch": 1.5292027537197423,
+      "grad_norm": 0.5773711065726657,
+      "learning_rate": 7.76201497591262e-06,
+      "loss": 0.0564,
+      "step": 3443
+    },
+    {
+      "epoch": 1.5296469020652899,
+      "grad_norm": 0.42024558526878264,
+      "learning_rate": 7.760398848350121e-06,
+      "loss": 0.0327,
+      "step": 3444
+    },
+    {
+      "epoch": 1.5300910504108374,
+      "grad_norm": 0.5673807060962988,
+      "learning_rate": 7.758782305852787e-06,
+      "loss": 0.0551,
+      "step": 3445
+    },
+    {
+      "epoch": 1.5305351987563847,
+      "grad_norm": 0.4401980723100094,
+      "learning_rate": 7.757165348663606e-06,
+      "loss": 0.0341,
+      "step": 3446
+    },
+    {
+      "epoch": 1.530979347101932,
+      "grad_norm": 0.4134655160444097,
+      "learning_rate": 7.755547977025641e-06,
+      "loss": 0.0396,
+      "step": 3447
+    },
+    {
+      "epoch": 1.5314234954474795,
+      "grad_norm": 0.7907504517928898,
+      "learning_rate": 7.753930191182005e-06,
+      "loss": 0.072,
+      "step": 3448
+    },
+    {
+      "epoch": 1.531867643793027,
+      "grad_norm": 0.7425759733481122,
+      "learning_rate": 7.752311991375878e-06,
+      "loss": 0.0584,
+      "step": 3449
+    },
+    {
+      "epoch": 1.5323117921385743,
+      "grad_norm": 0.3701818263547308,
+      "learning_rate": 7.750693377850506e-06,
+      "loss": 0.0311,
+      "step": 3450
+    },
+    {
+      "epoch": 1.5327559404841216,
+      "grad_norm": 0.4194874739178355,
+      "learning_rate": 7.749074350849196e-06,
+      "loss": 0.0299,
+      "step": 3451
+    },
+    {
+      "epoch": 1.533200088829669,
+      "grad_norm": 0.399280041395756,
+      "learning_rate": 7.747454910615309e-06,
+      "loss": 0.0333,
+      "step": 3452
+    },
+    {
+      "epoch": 1.5336442371752166,
+      "grad_norm": 0.5116821227256184,
+      "learning_rate": 7.74583505739228e-06,
+      "loss": 0.053,
+      "step": 3453
+    },
+    {
+      "epoch": 1.5340883855207639,
+      "grad_norm": 0.6704114941806725,
+      "learning_rate": 7.744214791423597e-06,
+      "loss": 0.0559,
+      "step": 3454
+    },
+    {
+      "epoch": 1.5345325338663114,
+      "grad_norm": 0.4031097936225357,
+      "learning_rate": 7.742594112952816e-06,
+      "loss": 0.04,
+      "step": 3455
+    },
+    {
+      "epoch": 1.534976682211859,
+      "grad_norm": 0.39875784908896506,
+      "learning_rate": 7.74097302222355e-06,
+      "loss": 0.0359,
+      "step": 3456
+    },
+    {
+      "epoch": 1.5354208305574062,
+      "grad_norm": 0.49710136867375654,
+      "learning_rate": 7.739351519479479e-06,
+      "loss": 0.0433,
+      "step": 3457
+    },
+    {
+      "epoch": 1.5358649789029535,
+      "grad_norm": 0.587479752061736,
+      "learning_rate": 7.73772960496434e-06,
+      "loss": 0.0482,
+      "step": 3458
+    },
+    {
+      "epoch": 1.536309127248501,
+      "grad_norm": 0.9958134431642607,
+      "learning_rate": 7.736107278921937e-06,
+      "loss": 0.0604,
+      "step": 3459
+    },
+    {
+      "epoch": 1.5367532755940485,
+      "grad_norm": 0.5924386484163253,
+      "learning_rate": 7.73448454159613e-06,
+      "loss": 0.0643,
+      "step": 3460
+    },
+    {
+      "epoch": 1.5371974239395958,
+      "grad_norm": 0.4469041409309243,
+      "learning_rate": 7.732861393230845e-06,
+      "loss": 0.0377,
+      "step": 3461
+    },
+    {
+      "epoch": 1.537641572285143,
+      "grad_norm": 0.6971338285035604,
+      "learning_rate": 7.731237834070071e-06,
+      "loss": 0.0525,
+      "step": 3462
+    },
+    {
+      "epoch": 1.5380857206306906,
+      "grad_norm": 0.377836838682399,
+      "learning_rate": 7.729613864357854e-06,
+      "loss": 0.037,
+      "step": 3463
+    },
+    {
+      "epoch": 1.5385298689762381,
+      "grad_norm": 0.3909300894809345,
+      "learning_rate": 7.727989484338306e-06,
+      "loss": 0.0327,
+      "step": 3464
+    },
+    {
+      "epoch": 1.5389740173217854,
+      "grad_norm": 0.44535118870735724,
+      "learning_rate": 7.726364694255598e-06,
+      "loss": 0.0451,
+      "step": 3465
+    },
+    {
+      "epoch": 1.539418165667333,
+      "grad_norm": 0.4310182821670584,
+      "learning_rate": 7.724739494353963e-06,
+      "loss": 0.043,
+      "step": 3466
+    },
+    {
+      "epoch": 1.5398623140128804,
+      "grad_norm": 0.5072110735404695,
+      "learning_rate": 7.723113884877698e-06,
+      "loss": 0.0409,
+      "step": 3467
+    },
+    {
+      "epoch": 1.5403064623584277,
+      "grad_norm": 0.4243921084057989,
+      "learning_rate": 7.721487866071158e-06,
+      "loss": 0.0577,
+      "step": 3468
+    },
+    {
+      "epoch": 1.540750610703975,
+      "grad_norm": 0.6225210766153572,
+      "learning_rate": 7.71986143817876e-06,
+      "loss": 0.041,
+      "step": 3469
+    },
+    {
+      "epoch": 1.5411947590495225,
+      "grad_norm": 0.7207947800360287,
+      "learning_rate": 7.718234601444987e-06,
+      "loss": 0.0525,
+      "step": 3470
+    },
+    {
+      "epoch": 1.54163890739507,
+      "grad_norm": 0.6448910865817588,
+      "learning_rate": 7.716607356114378e-06,
+      "loss": 0.0691,
+      "step": 3471
+    },
+    {
+      "epoch": 1.5420830557406173,
+      "grad_norm": 0.4192180148347075,
+      "learning_rate": 7.714979702431537e-06,
+      "loss": 0.0401,
+      "step": 3472
+    },
+    {
+      "epoch": 1.5425272040861646,
+      "grad_norm": 0.5353520298031071,
+      "learning_rate": 7.713351640641127e-06,
+      "loss": 0.047,
+      "step": 3473
+    },
+    {
+      "epoch": 1.5429713524317124,
+      "grad_norm": 0.4566196176860585,
+      "learning_rate": 7.711723170987875e-06,
+      "loss": 0.0377,
+      "step": 3474
+    },
+    {
+      "epoch": 1.5434155007772596,
+      "grad_norm": 0.8744543565589462,
+      "learning_rate": 7.710094293716563e-06,
+      "loss": 0.0421,
+      "step": 3475
+    },
+    {
+      "epoch": 1.543859649122807,
+      "grad_norm": 0.4450101049065203,
+      "learning_rate": 7.708465009072046e-06,
+      "loss": 0.0368,
+      "step": 3476
+    },
+    {
+      "epoch": 1.5443037974683544,
+      "grad_norm": 0.3880510280013894,
+      "learning_rate": 7.706835317299228e-06,
+      "loss": 0.042,
+      "step": 3477
+    },
+    {
+      "epoch": 1.544747945813902,
+      "grad_norm": 0.5627812490127366,
+      "learning_rate": 7.705205218643079e-06,
+      "loss": 0.0485,
+      "step": 3478
+    },
+    {
+      "epoch": 1.5451920941594492,
+      "grad_norm": 0.5226084645577708,
+      "learning_rate": 7.703574713348633e-06,
+      "loss": 0.0392,
+      "step": 3479
+    },
+    {
+      "epoch": 1.5456362425049965,
+      "grad_norm": 0.646405817519741,
+      "learning_rate": 7.701943801660983e-06,
+      "loss": 0.0679,
+      "step": 3480
+    },
+    {
+      "epoch": 1.546080390850544,
+      "grad_norm": 0.4165107020464423,
+      "learning_rate": 7.700312483825281e-06,
+      "loss": 0.0398,
+      "step": 3481
+    },
+    {
+      "epoch": 1.5465245391960916,
+      "grad_norm": 0.4015766878565599,
+      "learning_rate": 7.698680760086743e-06,
+      "loss": 0.0397,
+      "step": 3482
+    },
+    {
+      "epoch": 1.5469686875416389,
+      "grad_norm": 0.5170446946635605,
+      "learning_rate": 7.697048630690642e-06,
+      "loss": 0.0437,
+      "step": 3483
+    },
+    {
+      "epoch": 1.5474128358871864,
+      "grad_norm": 0.6516919138739381,
+      "learning_rate": 7.69541609588232e-06,
+      "loss": 0.0487,
+      "step": 3484
+    },
+    {
+      "epoch": 1.5478569842327339,
+      "grad_norm": 0.3866667271027062,
+      "learning_rate": 7.69378315590717e-06,
+      "loss": 0.0455,
+      "step": 3485
+    },
+    {
+      "epoch": 1.5483011325782812,
+      "grad_norm": 0.4947363891869916,
+      "learning_rate": 7.692149811010651e-06,
+      "loss": 0.0455,
+      "step": 3486
+    },
+    {
+      "epoch": 1.5487452809238285,
+      "grad_norm": 0.3642222145788099,
+      "learning_rate": 7.690516061438287e-06,
+      "loss": 0.0302,
+      "step": 3487
+    },
+    {
+      "epoch": 1.549189429269376,
+      "grad_norm": 0.49548313003009337,
+      "learning_rate": 7.688881907435653e-06,
+      "loss": 0.0491,
+      "step": 3488
+    },
+    {
+      "epoch": 1.5496335776149235,
+      "grad_norm": 0.5051267185591125,
+      "learning_rate": 7.687247349248393e-06,
+      "loss": 0.037,
+      "step": 3489
+    },
+    {
+      "epoch": 1.5500777259604708,
+      "grad_norm": 0.35479785899347316,
+      "learning_rate": 7.685612387122206e-06,
+      "loss": 0.0393,
+      "step": 3490
+    },
+    {
+      "epoch": 1.550521874306018,
+      "grad_norm": 0.3491505040065738,
+      "learning_rate": 7.68397702130286e-06,
+      "loss": 0.0391,
+      "step": 3491
+    },
+    {
+      "epoch": 1.5509660226515656,
+      "grad_norm": 0.4064600432185325,
+      "learning_rate": 7.682341252036171e-06,
+      "loss": 0.0369,
+      "step": 3492
+    },
+    {
+      "epoch": 1.551410170997113,
+      "grad_norm": 0.5118602299189453,
+      "learning_rate": 7.68070507956803e-06,
+      "loss": 0.0455,
+      "step": 3493
+    },
+    {
+      "epoch": 1.5518543193426604,
+      "grad_norm": 0.4565926867283545,
+      "learning_rate": 7.679068504144378e-06,
+      "loss": 0.038,
+      "step": 3494
+    },
+    {
+      "epoch": 1.552298467688208,
+      "grad_norm": 0.7287584197582313,
+      "learning_rate": 7.677431526011218e-06,
+      "loss": 0.0569,
+      "step": 3495
+    },
+    {
+      "epoch": 1.5527426160337554,
+      "grad_norm": 0.5994194003053249,
+      "learning_rate": 7.67579414541462e-06,
+      "loss": 0.0652,
+      "step": 3496
+    },
+    {
+      "epoch": 1.5531867643793027,
+      "grad_norm": 0.826708798968826,
+      "learning_rate": 7.674156362600708e-06,
+      "loss": 0.0788,
+      "step": 3497
+    },
+    {
+      "epoch": 1.55363091272485,
+      "grad_norm": 0.6304316452414537,
+      "learning_rate": 7.672518177815669e-06,
+      "loss": 0.0447,
+      "step": 3498
+    },
+    {
+      "epoch": 1.5540750610703975,
+      "grad_norm": 0.5354755353841119,
+      "learning_rate": 7.67087959130575e-06,
+      "loss": 0.0398,
+      "step": 3499
+    },
+    {
+      "epoch": 1.554519209415945,
+      "grad_norm": 0.5926105062715095,
+      "learning_rate": 7.669240603317257e-06,
+      "loss": 0.0495,
+      "step": 3500
+    },
+    {
+      "epoch": 1.5549633577614923,
+      "grad_norm": 0.4138530773015362,
+      "learning_rate": 7.66760121409656e-06,
+      "loss": 0.0354,
+      "step": 3501
+    },
+    {
+      "epoch": 1.5554075061070396,
+      "grad_norm": 0.44256364969603434,
+      "learning_rate": 7.665961423890085e-06,
+      "loss": 0.045,
+      "step": 3502
+    },
+    {
+      "epoch": 1.5558516544525873,
+      "grad_norm": 0.47643325623563626,
+      "learning_rate": 7.664321232944321e-06,
+      "loss": 0.0403,
+      "step": 3503
+    },
+    {
+      "epoch": 1.5562958027981346,
+      "grad_norm": 0.3731544040854046,
+      "learning_rate": 7.662680641505817e-06,
+      "loss": 0.0412,
+      "step": 3504
+    },
+    {
+      "epoch": 1.556739951143682,
+      "grad_norm": 0.8553804228205281,
+      "learning_rate": 7.661039649821183e-06,
+      "loss": 0.0688,
+      "step": 3505
+    },
+    {
+      "epoch": 1.5571840994892294,
+      "grad_norm": 0.5134256859909045,
+      "learning_rate": 7.659398258137085e-06,
+      "loss": 0.044,
+      "step": 3506
+    },
+    {
+      "epoch": 1.557628247834777,
+      "grad_norm": 0.5647254301702445,
+      "learning_rate": 7.657756466700252e-06,
+      "loss": 0.0437,
+      "step": 3507
+    },
+    {
+      "epoch": 1.5580723961803242,
+      "grad_norm": 0.44604429848268956,
+      "learning_rate": 7.656114275757477e-06,
+      "loss": 0.0399,
+      "step": 3508
+    },
+    {
+      "epoch": 1.5585165445258715,
+      "grad_norm": 0.45991000971025053,
+      "learning_rate": 7.654471685555606e-06,
+      "loss": 0.0484,
+      "step": 3509
+    },
+    {
+      "epoch": 1.558960692871419,
+      "grad_norm": 0.5298662733782437,
+      "learning_rate": 7.65282869634155e-06,
+      "loss": 0.061,
+      "step": 3510
+    },
+    {
+      "epoch": 1.5594048412169665,
+      "grad_norm": 0.49816414515287183,
+      "learning_rate": 7.651185308362276e-06,
+      "loss": 0.0499,
+      "step": 3511
+    },
+    {
+      "epoch": 1.5598489895625138,
+      "grad_norm": 0.4052222432440109,
+      "learning_rate": 7.649541521864816e-06,
+      "loss": 0.0451,
+      "step": 3512
+    },
+    {
+      "epoch": 1.5602931379080613,
+      "grad_norm": 0.5042113116702226,
+      "learning_rate": 7.647897337096257e-06,
+      "loss": 0.0457,
+      "step": 3513
+    },
+    {
+      "epoch": 1.5607372862536089,
+      "grad_norm": 0.7798340643811736,
+      "learning_rate": 7.646252754303746e-06,
+      "loss": 0.1019,
+      "step": 3514
+    },
+    {
+      "epoch": 1.5611814345991561,
+      "grad_norm": 0.4972991596269239,
+      "learning_rate": 7.644607773734496e-06,
+      "loss": 0.0578,
+      "step": 3515
+    },
+    {
+      "epoch": 1.5616255829447034,
+      "grad_norm": 0.5179606694497955,
+      "learning_rate": 7.642962395635773e-06,
+      "loss": 0.0484,
+      "step": 3516
+    },
+    {
+      "epoch": 1.562069731290251,
+      "grad_norm": 0.4239628046426599,
+      "learning_rate": 7.641316620254907e-06,
+      "loss": 0.0401,
+      "step": 3517
+    },
+    {
+      "epoch": 1.5625138796357985,
+      "grad_norm": 0.41910897244128137,
+      "learning_rate": 7.639670447839284e-06,
+      "loss": 0.0431,
+      "step": 3518
+    },
+    {
+      "epoch": 1.5629580279813458,
+      "grad_norm": 0.6150053023433038,
+      "learning_rate": 7.638023878636353e-06,
+      "loss": 0.0401,
+      "step": 3519
+    },
+    {
+      "epoch": 1.563402176326893,
+      "grad_norm": 0.4401882181013756,
+      "learning_rate": 7.63637691289362e-06,
+      "loss": 0.0557,
+      "step": 3520
+    },
+    {
+      "epoch": 1.5638463246724406,
+      "grad_norm": 0.4753196454260942,
+      "learning_rate": 7.634729550858652e-06,
+      "loss": 0.0393,
+      "step": 3521
+    },
+    {
+      "epoch": 1.564290473017988,
+      "grad_norm": 0.36301949275943124,
+      "learning_rate": 7.633081792779079e-06,
+      "loss": 0.0352,
+      "step": 3522
+    },
+    {
+      "epoch": 1.5647346213635354,
+      "grad_norm": 0.6312670336926534,
+      "learning_rate": 7.631433638902583e-06,
+      "loss": 0.0458,
+      "step": 3523
+    },
+    {
+      "epoch": 1.5651787697090829,
+      "grad_norm": 0.4853790667151454,
+      "learning_rate": 7.629785089476912e-06,
+      "loss": 0.0465,
+      "step": 3524
+    },
+    {
+      "epoch": 1.5656229180546304,
+      "grad_norm": 0.5377885843949731,
+      "learning_rate": 7.628136144749867e-06,
+      "loss": 0.0432,
+      "step": 3525
+    },
+    {
+      "epoch": 1.5660670664001777,
+      "grad_norm": 0.4560714160433749,
+      "learning_rate": 7.626486804969316e-06,
+      "loss": 0.0443,
+      "step": 3526
+    },
+    {
+      "epoch": 1.566511214745725,
+      "grad_norm": 0.40682805145370315,
+      "learning_rate": 7.624837070383183e-06,
+      "loss": 0.0353,
+      "step": 3527
+    },
+    {
+      "epoch": 1.5669553630912725,
+      "grad_norm": 0.4791132302623183,
+      "learning_rate": 7.6231869412394495e-06,
+      "loss": 0.0502,
+      "step": 3528
+    },
+    {
+      "epoch": 1.56739951143682,
+      "grad_norm": 0.4343747460657118,
+      "learning_rate": 7.621536417786159e-06,
+      "loss": 0.0395,
+      "step": 3529
+    },
+    {
+      "epoch": 1.5678436597823673,
+      "grad_norm": 0.492794572243227,
+      "learning_rate": 7.619885500271413e-06,
+      "loss": 0.0361,
+      "step": 3530
+    },
+    {
+      "epoch": 1.5682878081279146,
+      "grad_norm": 0.3645392422465027,
+      "learning_rate": 7.618234188943372e-06,
+      "loss": 0.0389,
+      "step": 3531
+    },
+    {
+      "epoch": 1.568731956473462,
+      "grad_norm": 0.43518151556556955,
+      "learning_rate": 7.616582484050256e-06,
+      "loss": 0.037,
+      "step": 3532
+    },
+    {
+      "epoch": 1.5691761048190096,
+      "grad_norm": 0.5607869474661537,
+      "learning_rate": 7.614930385840345e-06,
+      "loss": 0.0494,
+      "step": 3533
+    },
+    {
+      "epoch": 1.5696202531645569,
+      "grad_norm": 0.3841819768936594,
+      "learning_rate": 7.613277894561978e-06,
+      "loss": 0.0426,
+      "step": 3534
+    },
+    {
+      "epoch": 1.5700644015101044,
+      "grad_norm": 0.3817054851480292,
+      "learning_rate": 7.611625010463549e-06,
+      "loss": 0.045,
+      "step": 3535
+    },
+    {
+      "epoch": 1.570508549855652,
+      "grad_norm": 0.8877604841062007,
+      "learning_rate": 7.60997173379352e-06,
+      "loss": 0.0511,
+      "step": 3536
+    },
+    {
+      "epoch": 1.5709526982011992,
+      "grad_norm": 0.5367152034215165,
+      "learning_rate": 7.608318064800403e-06,
+      "loss": 0.0437,
+      "step": 3537
+    },
+    {
+      "epoch": 1.5713968465467465,
+      "grad_norm": 0.37970223029486955,
+      "learning_rate": 7.606664003732771e-06,
+      "loss": 0.0426,
+      "step": 3538
+    },
+    {
+      "epoch": 1.571840994892294,
+      "grad_norm": 0.5586015168302153,
+      "learning_rate": 7.605009550839263e-06,
+      "loss": 0.0443,
+      "step": 3539
+    },
+    {
+      "epoch": 1.5722851432378415,
+      "grad_norm": 0.5314909266145166,
+      "learning_rate": 7.603354706368567e-06,
+      "loss": 0.0482,
+      "step": 3540
+    },
+    {
+      "epoch": 1.5727292915833888,
+      "grad_norm": 0.6999799538645216,
+      "learning_rate": 7.601699470569434e-06,
+      "loss": 0.0481,
+      "step": 3541
+    },
+    {
+      "epoch": 1.573173439928936,
+      "grad_norm": 0.6037904990943627,
+      "learning_rate": 7.600043843690677e-06,
+      "loss": 0.0563,
+      "step": 3542
+    },
+    {
+      "epoch": 1.5736175882744838,
+      "grad_norm": 0.47456339272048825,
+      "learning_rate": 7.5983878259811625e-06,
+      "loss": 0.069,
+      "step": 3543
+    },
+    {
+      "epoch": 1.5740617366200311,
+      "grad_norm": 0.9776402251476092,
+      "learning_rate": 7.59673141768982e-06,
+      "loss": 0.0634,
+      "step": 3544
+    },
+    {
+      "epoch": 1.5745058849655784,
+      "grad_norm": 0.430567824394842,
+      "learning_rate": 7.595074619065635e-06,
+      "loss": 0.0389,
+      "step": 3545
+    },
+    {
+      "epoch": 1.574950033311126,
+      "grad_norm": 0.592046861388675,
+      "learning_rate": 7.593417430357649e-06,
+      "loss": 0.0569,
+      "step": 3546
+    },
+    {
+      "epoch": 1.5753941816566734,
+      "grad_norm": 0.528657961671009,
+      "learning_rate": 7.591759851814972e-06,
+      "loss": 0.0477,
+      "step": 3547
+    },
+    {
+      "epoch": 1.5758383300022207,
+      "grad_norm": 0.6683751656639313,
+      "learning_rate": 7.590101883686761e-06,
+      "loss": 0.0567,
+      "step": 3548
+    },
+    {
+      "epoch": 1.576282478347768,
+      "grad_norm": 0.6152824586906789,
+      "learning_rate": 7.58844352622224e-06,
+      "loss": 0.0514,
+      "step": 3549
+    },
+    {
+      "epoch": 1.5767266266933155,
+      "grad_norm": 0.5175459666893117,
+      "learning_rate": 7.5867847796706865e-06,
+      "loss": 0.0466,
+      "step": 3550
+    },
+    {
+      "epoch": 1.577170775038863,
+      "grad_norm": 0.46509447514156393,
+      "learning_rate": 7.585125644281439e-06,
+      "loss": 0.0356,
+      "step": 3551
+    },
+    {
+      "epoch": 1.5776149233844103,
+      "grad_norm": 0.42445822473522876,
+      "learning_rate": 7.583466120303893e-06,
+      "loss": 0.0368,
+      "step": 3552
+    },
+    {
+      "epoch": 1.5780590717299579,
+      "grad_norm": 0.5278835867879181,
+      "learning_rate": 7.581806207987504e-06,
+      "loss": 0.0412,
+      "step": 3553
+    },
+    {
+      "epoch": 1.5785032200755054,
+      "grad_norm": 0.49560125396493776,
+      "learning_rate": 7.5801459075817865e-06,
+      "loss": 0.0435,
+      "step": 3554
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.5441296297037692,
+      "learning_rate": 7.578485219336307e-06,
+      "loss": 0.0408,
+      "step": 3555
+    },
+    {
+      "epoch": 1.5793915167666,
+      "grad_norm": 0.5569362734290482,
+      "learning_rate": 7.5768241435007e-06,
+      "loss": 0.0508,
+      "step": 3556
+    },
+    {
+      "epoch": 1.5798356651121475,
+      "grad_norm": 0.688395754837478,
+      "learning_rate": 7.57516268032465e-06,
+      "loss": 0.0575,
+      "step": 3557
+    },
+    {
+      "epoch": 1.580279813457695,
+      "grad_norm": 0.5366509995471049,
+      "learning_rate": 7.573500830057907e-06,
+      "loss": 0.0393,
+      "step": 3558
+    },
+    {
+      "epoch": 1.5807239618032423,
+      "grad_norm": 0.5521349728712537,
+      "learning_rate": 7.571838592950271e-06,
+      "loss": 0.0464,
+      "step": 3559
+    },
+    {
+      "epoch": 1.5811681101487896,
+      "grad_norm": 0.3794603822110102,
+      "learning_rate": 7.570175969251609e-06,
+      "loss": 0.0271,
+      "step": 3560
+    },
+    {
+      "epoch": 1.581612258494337,
+      "grad_norm": 0.954806853873049,
+      "learning_rate": 7.568512959211838e-06,
+      "loss": 0.0798,
+      "step": 3561
+    },
+    {
+      "epoch": 1.5820564068398846,
+      "grad_norm": 0.3774549347222219,
+      "learning_rate": 7.566849563080938e-06,
+      "loss": 0.034,
+      "step": 3562
+    },
+    {
+      "epoch": 1.5825005551854319,
+      "grad_norm": 0.5717245905532253,
+      "learning_rate": 7.565185781108944e-06,
+      "loss": 0.0499,
+      "step": 3563
+    },
+    {
+      "epoch": 1.5829447035309794,
+      "grad_norm": 0.5372548568959261,
+      "learning_rate": 7.563521613545954e-06,
+      "loss": 0.0543,
+      "step": 3564
+    },
+    {
+      "epoch": 1.583388851876527,
+      "grad_norm": 0.6132605143548788,
+      "learning_rate": 7.56185706064212e-06,
+      "loss": 0.0513,
+      "step": 3565
+    },
+    {
+      "epoch": 1.5838330002220742,
+      "grad_norm": 0.4784620611087414,
+      "learning_rate": 7.560192122647647e-06,
+      "loss": 0.0507,
+      "step": 3566
+    },
+    {
+      "epoch": 1.5842771485676215,
+      "grad_norm": 0.46891721716697665,
+      "learning_rate": 7.558526799812812e-06,
+      "loss": 0.0447,
+      "step": 3567
+    },
+    {
+      "epoch": 1.584721296913169,
+      "grad_norm": 0.6007245405053189,
+      "learning_rate": 7.556861092387937e-06,
+      "loss": 0.0461,
+      "step": 3568
+    },
+    {
+      "epoch": 1.5851654452587165,
+      "grad_norm": 0.6832271146214529,
+      "learning_rate": 7.555195000623404e-06,
+      "loss": 0.0615,
+      "step": 3569
+    },
+    {
+      "epoch": 1.5856095936042638,
+      "grad_norm": 0.628507751994035,
+      "learning_rate": 7.553528524769658e-06,
+      "loss": 0.0511,
+      "step": 3570
+    },
+    {
+      "epoch": 1.586053741949811,
+      "grad_norm": 0.47038947814194393,
+      "learning_rate": 7.551861665077199e-06,
+      "loss": 0.0503,
+      "step": 3571
+    },
+    {
+      "epoch": 1.5864978902953588,
+      "grad_norm": 0.5039990895794063,
+      "learning_rate": 7.550194421796583e-06,
+      "loss": 0.0757,
+      "step": 3572
+    },
+    {
+      "epoch": 1.586942038640906,
+      "grad_norm": 0.43304671963904556,
+      "learning_rate": 7.548526795178424e-06,
+      "loss": 0.0435,
+      "step": 3573
+    },
+    {
+      "epoch": 1.5873861869864534,
+      "grad_norm": 0.35567796030147675,
+      "learning_rate": 7.546858785473397e-06,
+      "loss": 0.0411,
+      "step": 3574
+    },
+    {
+      "epoch": 1.587830335332001,
+      "grad_norm": 0.5082433548982926,
+      "learning_rate": 7.54519039293223e-06,
+      "loss": 0.0476,
+      "step": 3575
+    },
+    {
+      "epoch": 1.5882744836775484,
+      "grad_norm": 0.4853723595468299,
+      "learning_rate": 7.543521617805711e-06,
+      "loss": 0.0403,
+      "step": 3576
+    },
+    {
+      "epoch": 1.5887186320230957,
+      "grad_norm": 0.4349996299659683,
+      "learning_rate": 7.541852460344687e-06,
+      "loss": 0.0444,
+      "step": 3577
+    },
+    {
+      "epoch": 1.589162780368643,
+      "grad_norm": 0.3738858521887595,
+      "learning_rate": 7.540182920800061e-06,
+      "loss": 0.0342,
+      "step": 3578
+    },
+    {
+      "epoch": 1.5896069287141905,
+      "grad_norm": 0.3860111205972228,
+      "learning_rate": 7.5385129994227916e-06,
+      "loss": 0.0326,
+      "step": 3579
+    },
+    {
+      "epoch": 1.590051077059738,
+      "grad_norm": 0.4607949804916419,
+      "learning_rate": 7.536842696463894e-06,
+      "loss": 0.0406,
+      "step": 3580
+    },
+    {
+      "epoch": 1.5904952254052853,
+      "grad_norm": 0.4090438356252611,
+      "learning_rate": 7.535172012174447e-06,
+      "loss": 0.0435,
+      "step": 3581
+    },
+    {
+      "epoch": 1.5909393737508328,
+      "grad_norm": 0.44992279283684267,
+      "learning_rate": 7.533500946805583e-06,
+      "loss": 0.0433,
+      "step": 3582
+    },
+    {
+      "epoch": 1.5913835220963803,
+      "grad_norm": 0.5608893338456544,
+      "learning_rate": 7.531829500608489e-06,
+      "loss": 0.0461,
+      "step": 3583
+    },
+    {
+      "epoch": 1.5918276704419276,
+      "grad_norm": 0.46587297637504893,
+      "learning_rate": 7.530157673834413e-06,
+      "loss": 0.0462,
+      "step": 3584
+    },
+    {
+      "epoch": 1.592271818787475,
+      "grad_norm": 0.4650339559246697,
+      "learning_rate": 7.528485466734658e-06,
+      "loss": 0.0463,
+      "step": 3585
+    },
+    {
+      "epoch": 1.5927159671330224,
+      "grad_norm": 0.6227086084109071,
+      "learning_rate": 7.526812879560586e-06,
+      "loss": 0.0579,
+      "step": 3586
+    },
+    {
+      "epoch": 1.59316011547857,
+      "grad_norm": 0.38020073620262274,
+      "learning_rate": 7.525139912563616e-06,
+      "loss": 0.0356,
+      "step": 3587
+    },
+    {
+      "epoch": 1.5936042638241172,
+      "grad_norm": 0.4538280786991795,
+      "learning_rate": 7.523466565995224e-06,
+      "loss": 0.041,
+      "step": 3588
+    },
+    {
+      "epoch": 1.5940484121696645,
+      "grad_norm": 0.41382625396603623,
+      "learning_rate": 7.521792840106937e-06,
+      "loss": 0.0309,
+      "step": 3589
+    },
+    {
+      "epoch": 1.594492560515212,
+      "grad_norm": 0.6086317820490929,
+      "learning_rate": 7.52011873515035e-06,
+      "loss": 0.055,
+      "step": 3590
+    },
+    {
+      "epoch": 1.5949367088607596,
+      "grad_norm": 0.5518315774986915,
+      "learning_rate": 7.518444251377108e-06,
+      "loss": 0.0448,
+      "step": 3591
+    },
+    {
+      "epoch": 1.5953808572063068,
+      "grad_norm": 0.46458719141632254,
+      "learning_rate": 7.516769389038915e-06,
+      "loss": 0.0375,
+      "step": 3592
+    },
+    {
+      "epoch": 1.5958250055518544,
+      "grad_norm": 0.6946531905139686,
+      "learning_rate": 7.515094148387529e-06,
+      "loss": 0.0637,
+      "step": 3593
+    },
+    {
+      "epoch": 1.5962691538974019,
+      "grad_norm": 0.47113343903410826,
+      "learning_rate": 7.51341852967477e-06,
+      "loss": 0.0403,
+      "step": 3594
+    },
+    {
+      "epoch": 1.5967133022429492,
+      "grad_norm": 0.34877248157599594,
+      "learning_rate": 7.511742533152509e-06,
+      "loss": 0.035,
+      "step": 3595
+    },
+    {
+      "epoch": 1.5971574505884965,
+      "grad_norm": 0.6597113801136277,
+      "learning_rate": 7.51006615907268e-06,
+      "loss": 0.0419,
+      "step": 3596
+    },
+    {
+      "epoch": 1.597601598934044,
+      "grad_norm": 0.6377364313042297,
+      "learning_rate": 7.508389407687267e-06,
+      "loss": 0.0503,
+      "step": 3597
+    },
+    {
+      "epoch": 1.5980457472795915,
+      "grad_norm": 0.45492565688398423,
+      "learning_rate": 7.506712279248316e-06,
+      "loss": 0.0399,
+      "step": 3598
+    },
+    {
+      "epoch": 1.5984898956251388,
+      "grad_norm": 0.4963351358417883,
+      "learning_rate": 7.5050347740079285e-06,
+      "loss": 0.0404,
+      "step": 3599
+    },
+    {
+      "epoch": 1.598934043970686,
+      "grad_norm": 0.3821109498168256,
+      "learning_rate": 7.503356892218261e-06,
+      "loss": 0.0341,
+      "step": 3600
+    },
+    {
+      "epoch": 1.5993781923162336,
+      "grad_norm": 0.5218817014308058,
+      "learning_rate": 7.501678634131528e-06,
+      "loss": 0.0457,
+      "step": 3601
+    },
+    {
+      "epoch": 1.599822340661781,
+      "grad_norm": 0.48140971211980904,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0416,
+      "step": 3602
+    },
+    {
+      "epoch": 1.6002664890073284,
+      "grad_norm": 0.6724090547593772,
+      "learning_rate": 7.498320990076006e-06,
+      "loss": 0.0407,
+      "step": 3603
+    },
+    {
+      "epoch": 1.6007106373528759,
+      "grad_norm": 0.4673511339888186,
+      "learning_rate": 7.496641604611926e-06,
+      "loss": 0.0445,
+      "step": 3604
+    },
+    {
+      "epoch": 1.6011547856984234,
+      "grad_norm": 0.4736546778146936,
+      "learning_rate": 7.494961843860204e-06,
+      "loss": 0.0457,
+      "step": 3605
+    },
+    {
+      "epoch": 1.6015989340439707,
+      "grad_norm": 0.46683144779872127,
+      "learning_rate": 7.4932817080733345e-06,
+      "loss": 0.0534,
+      "step": 3606
+    },
+    {
+      "epoch": 1.602043082389518,
+      "grad_norm": 0.38959987746176267,
+      "learning_rate": 7.491601197503871e-06,
+      "loss": 0.037,
+      "step": 3607
+    },
+    {
+      "epoch": 1.6024872307350655,
+      "grad_norm": 0.5453813196476248,
+      "learning_rate": 7.489920312404422e-06,
+      "loss": 0.0524,
+      "step": 3608
+    },
+    {
+      "epoch": 1.602931379080613,
+      "grad_norm": 0.5800473949841606,
+      "learning_rate": 7.488239053027653e-06,
+      "loss": 0.0575,
+      "step": 3609
+    },
+    {
+      "epoch": 1.6033755274261603,
+      "grad_norm": 0.5289683983594908,
+      "learning_rate": 7.486557419626288e-06,
+      "loss": 0.0548,
+      "step": 3610
+    },
+    {
+      "epoch": 1.6038196757717076,
+      "grad_norm": 0.6070529095018132,
+      "learning_rate": 7.484875412453102e-06,
+      "loss": 0.0412,
+      "step": 3611
+    },
+    {
+      "epoch": 1.6042638241172553,
+      "grad_norm": 0.589551206830852,
+      "learning_rate": 7.483193031760932e-06,
+      "loss": 0.0548,
+      "step": 3612
+    },
+    {
+      "epoch": 1.6047079724628026,
+      "grad_norm": 0.4017721125553618,
+      "learning_rate": 7.481510277802667e-06,
+      "loss": 0.0294,
+      "step": 3613
+    },
+    {
+      "epoch": 1.60515212080835,
+      "grad_norm": 0.45650083282856746,
+      "learning_rate": 7.479827150831254e-06,
+      "loss": 0.0368,
+      "step": 3614
+    },
+    {
+      "epoch": 1.6055962691538974,
+      "grad_norm": 0.3377636070281279,
+      "learning_rate": 7.478143651099694e-06,
+      "loss": 0.0361,
+      "step": 3615
+    },
+    {
+      "epoch": 1.606040417499445,
+      "grad_norm": 0.4206873654117964,
+      "learning_rate": 7.4764597788610496e-06,
+      "loss": 0.0373,
+      "step": 3616
+    },
+    {
+      "epoch": 1.6064845658449922,
+      "grad_norm": 0.5196617152819853,
+      "learning_rate": 7.47477553436843e-06,
+      "loss": 0.0535,
+      "step": 3617
+    },
+    {
+      "epoch": 1.6069287141905395,
+      "grad_norm": 0.499359501911801,
+      "learning_rate": 7.47309091787501e-06,
+      "loss": 0.0414,
+      "step": 3618
+    },
+    {
+      "epoch": 1.607372862536087,
+      "grad_norm": 0.5994533032039974,
+      "learning_rate": 7.471405929634014e-06,
+      "loss": 0.0521,
+      "step": 3619
+    },
+    {
+      "epoch": 1.6078170108816345,
+      "grad_norm": 0.3309953500833457,
+      "learning_rate": 7.469720569898725e-06,
+      "loss": 0.0331,
+      "step": 3620
+    },
+    {
+      "epoch": 1.6082611592271818,
+      "grad_norm": 0.4509086857710446,
+      "learning_rate": 7.468034838922482e-06,
+      "loss": 0.0458,
+      "step": 3621
+    },
+    {
+      "epoch": 1.6087053075727293,
+      "grad_norm": 0.45137595767758343,
+      "learning_rate": 7.4663487369586776e-06,
+      "loss": 0.0469,
+      "step": 3622
+    },
+    {
+      "epoch": 1.6091494559182768,
+      "grad_norm": 0.4696695504363821,
+      "learning_rate": 7.464662264260761e-06,
+      "loss": 0.0452,
+      "step": 3623
+    },
+    {
+      "epoch": 1.6095936042638241,
+      "grad_norm": 0.4596849970347727,
+      "learning_rate": 7.46297542108224e-06,
+      "loss": 0.0438,
+      "step": 3624
+    },
+    {
+      "epoch": 1.6100377526093714,
+      "grad_norm": 0.5614676347080366,
+      "learning_rate": 7.4612882076766744e-06,
+      "loss": 0.0599,
+      "step": 3625
+    },
+    {
+      "epoch": 1.610481900954919,
+      "grad_norm": 0.3947391918010033,
+      "learning_rate": 7.459600624297681e-06,
+      "loss": 0.0347,
+      "step": 3626
+    },
+    {
+      "epoch": 1.6109260493004665,
+      "grad_norm": 0.6133457502317237,
+      "learning_rate": 7.4579126711989326e-06,
+      "loss": 0.053,
+      "step": 3627
+    },
+    {
+      "epoch": 1.6113701976460137,
+      "grad_norm": 0.45537205293561567,
+      "learning_rate": 7.456224348634158e-06,
+      "loss": 0.0436,
+      "step": 3628
+    },
+    {
+      "epoch": 1.611814345991561,
+      "grad_norm": 0.46644204998070415,
+      "learning_rate": 7.454535656857138e-06,
+      "loss": 0.0492,
+      "step": 3629
+    },
+    {
+      "epoch": 1.6122584943371085,
+      "grad_norm": 0.5228600296837344,
+      "learning_rate": 7.4528465961217145e-06,
+      "loss": 0.0521,
+      "step": 3630
+    },
+    {
+      "epoch": 1.612702642682656,
+      "grad_norm": 0.5557514684322911,
+      "learning_rate": 7.451157166681781e-06,
+      "loss": 0.0526,
+      "step": 3631
+    },
+    {
+      "epoch": 1.6131467910282034,
+      "grad_norm": 0.4758338953909275,
+      "learning_rate": 7.449467368791287e-06,
+      "loss": 0.0417,
+      "step": 3632
+    },
+    {
+      "epoch": 1.6135909393737509,
+      "grad_norm": 0.5619608884775461,
+      "learning_rate": 7.4477772027042395e-06,
+      "loss": 0.0582,
+      "step": 3633
+    },
+    {
+      "epoch": 1.6140350877192984,
+      "grad_norm": 0.46236286351850703,
+      "learning_rate": 7.4460866686746966e-06,
+      "loss": 0.0384,
+      "step": 3634
+    },
+    {
+      "epoch": 1.6144792360648457,
+      "grad_norm": 0.44941172754104425,
+      "learning_rate": 7.444395766956776e-06,
+      "loss": 0.0451,
+      "step": 3635
+    },
+    {
+      "epoch": 1.614923384410393,
+      "grad_norm": 0.5343367958828964,
+      "learning_rate": 7.4427044978046496e-06,
+      "loss": 0.0539,
+      "step": 3636
+    },
+    {
+      "epoch": 1.6153675327559405,
+      "grad_norm": 0.7114764542018213,
+      "learning_rate": 7.4410128614725406e-06,
+      "loss": 0.0632,
+      "step": 3637
+    },
+    {
+      "epoch": 1.615811681101488,
+      "grad_norm": 0.33754698523568316,
+      "learning_rate": 7.439320858214736e-06,
+      "loss": 0.0432,
+      "step": 3638
+    },
+    {
+      "epoch": 1.6162558294470353,
+      "grad_norm": 0.43643183273515707,
+      "learning_rate": 7.437628488285568e-06,
+      "loss": 0.0379,
+      "step": 3639
+    },
+    {
+      "epoch": 1.6166999777925826,
+      "grad_norm": 0.44390629309704815,
+      "learning_rate": 7.435935751939429e-06,
+      "loss": 0.0446,
+      "step": 3640
+    },
+    {
+      "epoch": 1.6171441261381303,
+      "grad_norm": 0.4354396645312695,
+      "learning_rate": 7.4342426494307695e-06,
+      "loss": 0.043,
+      "step": 3641
+    },
+    {
+      "epoch": 1.6175882744836776,
+      "grad_norm": 0.44873911764176055,
+      "learning_rate": 7.432549181014088e-06,
+      "loss": 0.0302,
+      "step": 3642
+    },
+    {
+      "epoch": 1.6180324228292249,
+      "grad_norm": 0.41154412795322054,
+      "learning_rate": 7.430855346943942e-06,
+      "loss": 0.0395,
+      "step": 3643
+    },
+    {
+      "epoch": 1.6184765711747724,
+      "grad_norm": 0.5287832714190103,
+      "learning_rate": 7.4291611474749455e-06,
+      "loss": 0.0473,
+      "step": 3644
+    },
+    {
+      "epoch": 1.61892071952032,
+      "grad_norm": 0.4007407767223763,
+      "learning_rate": 7.427466582861765e-06,
+      "loss": 0.0378,
+      "step": 3645
+    },
+    {
+      "epoch": 1.6193648678658672,
+      "grad_norm": 0.58929177581027,
+      "learning_rate": 7.42577165335912e-06,
+      "loss": 0.0551,
+      "step": 3646
+    },
+    {
+      "epoch": 1.6198090162114145,
+      "grad_norm": 0.3940542333112033,
+      "learning_rate": 7.42407635922179e-06,
+      "loss": 0.0294,
+      "step": 3647
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.3862139275428533,
+      "learning_rate": 7.4223807007046045e-06,
+      "loss": 0.0318,
+      "step": 3648
+    },
+    {
+      "epoch": 1.6206973129025095,
+      "grad_norm": 0.4233634000349271,
+      "learning_rate": 7.4206846780624505e-06,
+      "loss": 0.0351,
+      "step": 3649
+    },
+    {
+      "epoch": 1.6211414612480568,
+      "grad_norm": 0.595696689600085,
+      "learning_rate": 7.418988291550271e-06,
+      "loss": 0.05,
+      "step": 3650
+    },
+    {
+      "epoch": 1.6215856095936043,
+      "grad_norm": 0.45676187183693634,
+      "learning_rate": 7.417291541423057e-06,
+      "loss": 0.0408,
+      "step": 3651
+    },
+    {
+      "epoch": 1.6220297579391518,
+      "grad_norm": 0.5662085190370286,
+      "learning_rate": 7.415594427935864e-06,
+      "loss": 0.0449,
+      "step": 3652
+    },
+    {
+      "epoch": 1.6224739062846991,
+      "grad_norm": 0.5893188179834489,
+      "learning_rate": 7.4138969513437945e-06,
+      "loss": 0.0514,
+      "step": 3653
+    },
+    {
+      "epoch": 1.6229180546302464,
+      "grad_norm": 1.1475511733222958,
+      "learning_rate": 7.412199111902007e-06,
+      "loss": 0.0539,
+      "step": 3654
+    },
+    {
+      "epoch": 1.623362202975794,
+      "grad_norm": 0.4595814551148835,
+      "learning_rate": 7.410500909865718e-06,
+      "loss": 0.0422,
+      "step": 3655
+    },
+    {
+      "epoch": 1.6238063513213414,
+      "grad_norm": 0.4905543614968632,
+      "learning_rate": 7.408802345490194e-06,
+      "loss": 0.0426,
+      "step": 3656
+    },
+    {
+      "epoch": 1.6242504996668887,
+      "grad_norm": 0.44546697092666765,
+      "learning_rate": 7.407103419030759e-06,
+      "loss": 0.0543,
+      "step": 3657
+    },
+    {
+      "epoch": 1.624694648012436,
+      "grad_norm": 0.4740804743160003,
+      "learning_rate": 7.405404130742793e-06,
+      "loss": 0.0493,
+      "step": 3658
+    },
+    {
+      "epoch": 1.6251387963579835,
+      "grad_norm": 1.5473311509032195,
+      "learning_rate": 7.4037044808817224e-06,
+      "loss": 0.0513,
+      "step": 3659
+    },
+    {
+      "epoch": 1.625582944703531,
+      "grad_norm": 0.6098239500874039,
+      "learning_rate": 7.402004469703038e-06,
+      "loss": 0.0367,
+      "step": 3660
+    },
+    {
+      "epoch": 1.6260270930490783,
+      "grad_norm": 0.6586494701114967,
+      "learning_rate": 7.4003040974622784e-06,
+      "loss": 0.0585,
+      "step": 3661
+    },
+    {
+      "epoch": 1.6264712413946258,
+      "grad_norm": 0.45627803806709566,
+      "learning_rate": 7.39860336441504e-06,
+      "loss": 0.048,
+      "step": 3662
+    },
+    {
+      "epoch": 1.6269153897401734,
+      "grad_norm": 0.5771583864203911,
+      "learning_rate": 7.3969022708169695e-06,
+      "loss": 0.0378,
+      "step": 3663
+    },
+    {
+      "epoch": 1.6273595380857206,
+      "grad_norm": 0.4931171394680339,
+      "learning_rate": 7.395200816923774e-06,
+      "loss": 0.033,
+      "step": 3664
+    },
+    {
+      "epoch": 1.627803686431268,
+      "grad_norm": 0.7206307962976604,
+      "learning_rate": 7.393499002991206e-06,
+      "loss": 0.0541,
+      "step": 3665
+    },
+    {
+      "epoch": 1.6282478347768155,
+      "grad_norm": 0.4622361606059328,
+      "learning_rate": 7.3917968292750785e-06,
+      "loss": 0.0415,
+      "step": 3666
+    },
+    {
+      "epoch": 1.628691983122363,
+      "grad_norm": 0.7660149085068831,
+      "learning_rate": 7.390094296031259e-06,
+      "loss": 0.0627,
+      "step": 3667
+    },
+    {
+      "epoch": 1.6291361314679103,
+      "grad_norm": 0.7776923661917167,
+      "learning_rate": 7.3883914035156666e-06,
+      "loss": 0.0396,
+      "step": 3668
+    },
+    {
+      "epoch": 1.6295802798134575,
+      "grad_norm": 0.6556574717804856,
+      "learning_rate": 7.386688151984275e-06,
+      "loss": 0.0449,
+      "step": 3669
+    },
+    {
+      "epoch": 1.630024428159005,
+      "grad_norm": 0.6486302576198958,
+      "learning_rate": 7.384984541693111e-06,
+      "loss": 0.0456,
+      "step": 3670
+    },
+    {
+      "epoch": 1.6304685765045526,
+      "grad_norm": 0.38753606074247743,
+      "learning_rate": 7.383280572898256e-06,
+      "loss": 0.0374,
+      "step": 3671
+    },
+    {
+      "epoch": 1.6309127248500999,
+      "grad_norm": 0.4446665040074296,
+      "learning_rate": 7.381576245855847e-06,
+      "loss": 0.0399,
+      "step": 3672
+    },
+    {
+      "epoch": 1.6313568731956474,
+      "grad_norm": 0.6808941373857945,
+      "learning_rate": 7.379871560822071e-06,
+      "loss": 0.0523,
+      "step": 3673
+    },
+    {
+      "epoch": 1.6318010215411949,
+      "grad_norm": 0.6878067318911092,
+      "learning_rate": 7.378166518053174e-06,
+      "loss": 0.0557,
+      "step": 3674
+    },
+    {
+      "epoch": 1.6322451698867422,
+      "grad_norm": 0.4460005003836504,
+      "learning_rate": 7.37646111780545e-06,
+      "loss": 0.0338,
+      "step": 3675
+    },
+    {
+      "epoch": 1.6326893182322895,
+      "grad_norm": 0.392173745509544,
+      "learning_rate": 7.374755360335253e-06,
+      "loss": 0.0358,
+      "step": 3676
+    },
+    {
+      "epoch": 1.633133466577837,
+      "grad_norm": 0.4942681170970006,
+      "learning_rate": 7.3730492458989825e-06,
+      "loss": 0.0451,
+      "step": 3677
+    },
+    {
+      "epoch": 1.6335776149233845,
+      "grad_norm": 0.35125391907425363,
+      "learning_rate": 7.371342774753101e-06,
+      "loss": 0.0376,
+      "step": 3678
+    },
+    {
+      "epoch": 1.6340217632689318,
+      "grad_norm": 0.5143098431682592,
+      "learning_rate": 7.369635947154119e-06,
+      "loss": 0.0469,
+      "step": 3679
+    },
+    {
+      "epoch": 1.634465911614479,
+      "grad_norm": 0.5196302676867172,
+      "learning_rate": 7.3679287633585995e-06,
+      "loss": 0.0436,
+      "step": 3680
+    },
+    {
+      "epoch": 1.6349100599600268,
+      "grad_norm": 0.5148856987270073,
+      "learning_rate": 7.366221223623163e-06,
+      "loss": 0.0429,
+      "step": 3681
+    },
+    {
+      "epoch": 1.635354208305574,
+      "grad_norm": 0.42764696599583774,
+      "learning_rate": 7.3645133282044835e-06,
+      "loss": 0.0333,
+      "step": 3682
+    },
+    {
+      "epoch": 1.6357983566511214,
+      "grad_norm": 0.3662627068077827,
+      "learning_rate": 7.362805077359283e-06,
+      "loss": 0.0312,
+      "step": 3683
+    },
+    {
+      "epoch": 1.636242504996669,
+      "grad_norm": 0.5318925951291891,
+      "learning_rate": 7.361096471344341e-06,
+      "loss": 0.0392,
+      "step": 3684
+    },
+    {
+      "epoch": 1.6366866533422164,
+      "grad_norm": 0.5101772714710058,
+      "learning_rate": 7.359387510416494e-06,
+      "loss": 0.0444,
+      "step": 3685
+    },
+    {
+      "epoch": 1.6371308016877637,
+      "grad_norm": 0.6838175180628748,
+      "learning_rate": 7.357678194832623e-06,
+      "loss": 0.0443,
+      "step": 3686
+    },
+    {
+      "epoch": 1.637574950033311,
+      "grad_norm": 0.4298262192631913,
+      "learning_rate": 7.355968524849671e-06,
+      "loss": 0.0397,
+      "step": 3687
+    },
+    {
+      "epoch": 1.6380190983788585,
+      "grad_norm": 0.5227260046796232,
+      "learning_rate": 7.354258500724627e-06,
+      "loss": 0.0428,
+      "step": 3688
+    },
+    {
+      "epoch": 1.638463246724406,
+      "grad_norm": 0.3848693505144518,
+      "learning_rate": 7.352548122714541e-06,
+      "loss": 0.038,
+      "step": 3689
+    },
+    {
+      "epoch": 1.6389073950699533,
+      "grad_norm": 0.35605605527782813,
+      "learning_rate": 7.350837391076509e-06,
+      "loss": 0.0385,
+      "step": 3690
+    },
+    {
+      "epoch": 1.6393515434155008,
+      "grad_norm": 0.5496746740887447,
+      "learning_rate": 7.349126306067681e-06,
+      "loss": 0.0399,
+      "step": 3691
+    },
+    {
+      "epoch": 1.6397956917610483,
+      "grad_norm": 0.3577570105037349,
+      "learning_rate": 7.347414867945266e-06,
+      "loss": 0.035,
+      "step": 3692
+    },
+    {
+      "epoch": 1.6402398401065956,
+      "grad_norm": 0.4964615292337297,
+      "learning_rate": 7.345703076966522e-06,
+      "loss": 0.0517,
+      "step": 3693
+    },
+    {
+      "epoch": 1.640683988452143,
+      "grad_norm": 0.7390592067058361,
+      "learning_rate": 7.343990933388757e-06,
+      "loss": 0.0367,
+      "step": 3694
+    },
+    {
+      "epoch": 1.6411281367976904,
+      "grad_norm": 0.3720816990200118,
+      "learning_rate": 7.342278437469338e-06,
+      "loss": 0.0385,
+      "step": 3695
+    },
+    {
+      "epoch": 1.641572285143238,
+      "grad_norm": 0.4417091992699088,
+      "learning_rate": 7.340565589465681e-06,
+      "loss": 0.0496,
+      "step": 3696
+    },
+    {
+      "epoch": 1.6420164334887852,
+      "grad_norm": 0.4387119629881648,
+      "learning_rate": 7.338852389635258e-06,
+      "loss": 0.0482,
+      "step": 3697
+    },
+    {
+      "epoch": 1.6424605818343325,
+      "grad_norm": 0.5237427797642873,
+      "learning_rate": 7.337138838235589e-06,
+      "loss": 0.0397,
+      "step": 3698
+    },
+    {
+      "epoch": 1.64290473017988,
+      "grad_norm": 0.5126897468237226,
+      "learning_rate": 7.335424935524254e-06,
+      "loss": 0.0354,
+      "step": 3699
+    },
+    {
+      "epoch": 1.6433488785254275,
+      "grad_norm": 0.5632224593772709,
+      "learning_rate": 7.333710681758876e-06,
+      "loss": 0.0515,
+      "step": 3700
+    },
+    {
+      "epoch": 1.6437930268709748,
+      "grad_norm": 0.37651906930453466,
+      "learning_rate": 7.331996077197141e-06,
+      "loss": 0.0263,
+      "step": 3701
+    },
+    {
+      "epoch": 1.6442371752165224,
+      "grad_norm": 0.403180672879644,
+      "learning_rate": 7.330281122096783e-06,
+      "loss": 0.0361,
+      "step": 3702
+    },
+    {
+      "epoch": 1.6446813235620699,
+      "grad_norm": 0.3870458503286661,
+      "learning_rate": 7.328565816715587e-06,
+      "loss": 0.0391,
+      "step": 3703
+    },
+    {
+      "epoch": 1.6451254719076172,
+      "grad_norm": 0.46910403259693756,
+      "learning_rate": 7.326850161311394e-06,
+      "loss": 0.0539,
+      "step": 3704
+    },
+    {
+      "epoch": 1.6455696202531644,
+      "grad_norm": 0.6277176707566176,
+      "learning_rate": 7.325134156142093e-06,
+      "loss": 0.0434,
+      "step": 3705
+    },
+    {
+      "epoch": 1.646013768598712,
+      "grad_norm": 0.5311345861511363,
+      "learning_rate": 7.323417801465633e-06,
+      "loss": 0.0528,
+      "step": 3706
+    },
+    {
+      "epoch": 1.6464579169442595,
+      "grad_norm": 0.49558316497497373,
+      "learning_rate": 7.32170109754001e-06,
+      "loss": 0.0518,
+      "step": 3707
+    },
+    {
+      "epoch": 1.6469020652898068,
+      "grad_norm": 0.4215235777284794,
+      "learning_rate": 7.319984044623274e-06,
+      "loss": 0.0374,
+      "step": 3708
+    },
+    {
+      "epoch": 1.647346213635354,
+      "grad_norm": 0.7018323227064814,
+      "learning_rate": 7.3182666429735236e-06,
+      "loss": 0.0589,
+      "step": 3709
+    },
+    {
+      "epoch": 1.6477903619809018,
+      "grad_norm": 0.3811938009083553,
+      "learning_rate": 7.316548892848919e-06,
+      "loss": 0.0336,
+      "step": 3710
+    },
+    {
+      "epoch": 1.648234510326449,
+      "grad_norm": 0.48162066324285185,
+      "learning_rate": 7.314830794507664e-06,
+      "loss": 0.0392,
+      "step": 3711
+    },
+    {
+      "epoch": 1.6486786586719964,
+      "grad_norm": 0.7241447798938966,
+      "learning_rate": 7.313112348208017e-06,
+      "loss": 0.0705,
+      "step": 3712
+    },
+    {
+      "epoch": 1.6491228070175439,
+      "grad_norm": 0.4040868157055559,
+      "learning_rate": 7.311393554208292e-06,
+      "loss": 0.0417,
+      "step": 3713
+    },
+    {
+      "epoch": 1.6495669553630914,
+      "grad_norm": 0.43212705890325137,
+      "learning_rate": 7.3096744127668515e-06,
+      "loss": 0.0516,
+      "step": 3714
+    },
+    {
+      "epoch": 1.6500111037086387,
+      "grad_norm": 0.7227761525435585,
+      "learning_rate": 7.307954924142113e-06,
+      "loss": 0.0427,
+      "step": 3715
+    },
+    {
+      "epoch": 1.650455252054186,
+      "grad_norm": 0.6454421002922318,
+      "learning_rate": 7.306235088592545e-06,
+      "loss": 0.0709,
+      "step": 3716
+    },
+    {
+      "epoch": 1.6508994003997335,
+      "grad_norm": 0.5263009072366842,
+      "learning_rate": 7.304514906376665e-06,
+      "loss": 0.0354,
+      "step": 3717
+    },
+    {
+      "epoch": 1.651343548745281,
+      "grad_norm": 0.4475608856046002,
+      "learning_rate": 7.3027943777530504e-06,
+      "loss": 0.0376,
+      "step": 3718
+    },
+    {
+      "epoch": 1.6517876970908283,
+      "grad_norm": 0.4187029855449,
+      "learning_rate": 7.301073502980321e-06,
+      "loss": 0.032,
+      "step": 3719
+    },
+    {
+      "epoch": 1.6522318454363758,
+      "grad_norm": 0.4688452297403804,
+      "learning_rate": 7.299352282317156e-06,
+      "loss": 0.0328,
+      "step": 3720
+    },
+    {
+      "epoch": 1.6526759937819233,
+      "grad_norm": 0.4905958290493516,
+      "learning_rate": 7.297630716022285e-06,
+      "loss": 0.0408,
+      "step": 3721
+    },
+    {
+      "epoch": 1.6531201421274706,
+      "grad_norm": 0.5196371163361964,
+      "learning_rate": 7.295908804354486e-06,
+      "loss": 0.0518,
+      "step": 3722
+    },
+    {
+      "epoch": 1.653564290473018,
+      "grad_norm": 0.5745303077003567,
+      "learning_rate": 7.294186547572593e-06,
+      "loss": 0.0525,
+      "step": 3723
+    },
+    {
+      "epoch": 1.6540084388185654,
+      "grad_norm": 0.4232397957508058,
+      "learning_rate": 7.292463945935492e-06,
+      "loss": 0.046,
+      "step": 3724
+    },
+    {
+      "epoch": 1.654452587164113,
+      "grad_norm": 0.5715502169683124,
+      "learning_rate": 7.290740999702117e-06,
+      "loss": 0.0502,
+      "step": 3725
+    },
+    {
+      "epoch": 1.6548967355096602,
+      "grad_norm": 0.45041445198482544,
+      "learning_rate": 7.289017709131456e-06,
+      "loss": 0.0385,
+      "step": 3726
+    },
+    {
+      "epoch": 1.6553408838552075,
+      "grad_norm": 0.5898309113830352,
+      "learning_rate": 7.287294074482551e-06,
+      "loss": 0.0586,
+      "step": 3727
+    },
+    {
+      "epoch": 1.655785032200755,
+      "grad_norm": 0.42537829603624505,
+      "learning_rate": 7.285570096014491e-06,
+      "loss": 0.0382,
+      "step": 3728
+    },
+    {
+      "epoch": 1.6562291805463025,
+      "grad_norm": 0.5068748041039001,
+      "learning_rate": 7.283845773986421e-06,
+      "loss": 0.0424,
+      "step": 3729
+    },
+    {
+      "epoch": 1.6566733288918498,
+      "grad_norm": 0.49158429307244256,
+      "learning_rate": 7.2821211086575365e-06,
+      "loss": 0.0413,
+      "step": 3730
+    },
+    {
+      "epoch": 1.6571174772373973,
+      "grad_norm": 0.5434741115681563,
+      "learning_rate": 7.280396100287082e-06,
+      "loss": 0.0415,
+      "step": 3731
+    },
+    {
+      "epoch": 1.6575616255829448,
+      "grad_norm": 0.54047842476067,
+      "learning_rate": 7.278670749134356e-06,
+      "loss": 0.0337,
+      "step": 3732
+    },
+    {
+      "epoch": 1.6580057739284921,
+      "grad_norm": 0.7610613678272631,
+      "learning_rate": 7.276945055458709e-06,
+      "loss": 0.0513,
+      "step": 3733
+    },
+    {
+      "epoch": 1.6584499222740394,
+      "grad_norm": 0.43103079664602795,
+      "learning_rate": 7.275219019519542e-06,
+      "loss": 0.0365,
+      "step": 3734
+    },
+    {
+      "epoch": 1.658894070619587,
+      "grad_norm": 0.6707297267218786,
+      "learning_rate": 7.2734926415763074e-06,
+      "loss": 0.0425,
+      "step": 3735
+    },
+    {
+      "epoch": 1.6593382189651344,
+      "grad_norm": 0.3906372094311339,
+      "learning_rate": 7.271765921888507e-06,
+      "loss": 0.0379,
+      "step": 3736
+    },
+    {
+      "epoch": 1.6597823673106817,
+      "grad_norm": 0.8497034117571121,
+      "learning_rate": 7.2700388607157e-06,
+      "loss": 0.0592,
+      "step": 3737
+    },
+    {
+      "epoch": 1.660226515656229,
+      "grad_norm": 0.4906869751133543,
+      "learning_rate": 7.268311458317491e-06,
+      "loss": 0.0448,
+      "step": 3738
+    },
+    {
+      "epoch": 1.6606706640017765,
+      "grad_norm": 0.46268372537996755,
+      "learning_rate": 7.266583714953536e-06,
+      "loss": 0.0363,
+      "step": 3739
+    },
+    {
+      "epoch": 1.661114812347324,
+      "grad_norm": 0.3974000929871205,
+      "learning_rate": 7.2648556308835476e-06,
+      "loss": 0.0375,
+      "step": 3740
+    },
+    {
+      "epoch": 1.6615589606928713,
+      "grad_norm": 0.3782320844424881,
+      "learning_rate": 7.263127206367285e-06,
+      "loss": 0.0327,
+      "step": 3741
+    },
+    {
+      "epoch": 1.6620031090384189,
+      "grad_norm": 0.7073636631807192,
+      "learning_rate": 7.2613984416645586e-06,
+      "loss": 0.0549,
+      "step": 3742
+    },
+    {
+      "epoch": 1.6624472573839664,
+      "grad_norm": 0.48125971182332594,
+      "learning_rate": 7.2596693370352325e-06,
+      "loss": 0.0342,
+      "step": 3743
+    },
+    {
+      "epoch": 1.6628914057295137,
+      "grad_norm": 0.4622696042568772,
+      "learning_rate": 7.257939892739221e-06,
+      "loss": 0.0366,
+      "step": 3744
+    },
+    {
+      "epoch": 1.663335554075061,
+      "grad_norm": 0.4506482837249478,
+      "learning_rate": 7.256210109036485e-06,
+      "loss": 0.0399,
+      "step": 3745
+    },
+    {
+      "epoch": 1.6637797024206085,
+      "grad_norm": 0.468794971821978,
+      "learning_rate": 7.254479986187045e-06,
+      "loss": 0.0486,
+      "step": 3746
+    },
+    {
+      "epoch": 1.664223850766156,
+      "grad_norm": 0.461349555479564,
+      "learning_rate": 7.252749524450967e-06,
+      "loss": 0.0399,
+      "step": 3747
+    },
+    {
+      "epoch": 1.6646679991117033,
+      "grad_norm": 0.35487507978196975,
+      "learning_rate": 7.251018724088367e-06,
+      "loss": 0.0399,
+      "step": 3748
+    },
+    {
+      "epoch": 1.6651121474572506,
+      "grad_norm": 0.5665408950529203,
+      "learning_rate": 7.249287585359416e-06,
+      "loss": 0.0515,
+      "step": 3749
+    },
+    {
+      "epoch": 1.6655562958027983,
+      "grad_norm": 0.47765729176525834,
+      "learning_rate": 7.24755610852433e-06,
+      "loss": 0.0524,
+      "step": 3750
+    },
+    {
+      "epoch": 1.6660004441483456,
+      "grad_norm": 0.49560492659442207,
+      "learning_rate": 7.245824293843382e-06,
+      "loss": 0.0474,
+      "step": 3751
+    },
+    {
+      "epoch": 1.6664445924938929,
+      "grad_norm": 0.47858236763124845,
+      "learning_rate": 7.244092141576895e-06,
+      "loss": 0.0404,
+      "step": 3752
+    },
+    {
+      "epoch": 1.6668887408394404,
+      "grad_norm": 0.5242443798051569,
+      "learning_rate": 7.2423596519852354e-06,
+      "loss": 0.0564,
+      "step": 3753
+    },
+    {
+      "epoch": 1.667332889184988,
+      "grad_norm": 0.4276118939925525,
+      "learning_rate": 7.240626825328832e-06,
+      "loss": 0.0396,
+      "step": 3754
+    },
+    {
+      "epoch": 1.6677770375305352,
+      "grad_norm": 0.365991499711862,
+      "learning_rate": 7.238893661868154e-06,
+      "loss": 0.0373,
+      "step": 3755
+    },
+    {
+      "epoch": 1.6682211858760825,
+      "grad_norm": 0.614415895237667,
+      "learning_rate": 7.237160161863725e-06,
+      "loss": 0.0528,
+      "step": 3756
+    },
+    {
+      "epoch": 1.66866533422163,
+      "grad_norm": 0.40303741888400607,
+      "learning_rate": 7.235426325576123e-06,
+      "loss": 0.0407,
+      "step": 3757
+    },
+    {
+      "epoch": 1.6691094825671775,
+      "grad_norm": 0.5735579153246894,
+      "learning_rate": 7.23369215326597e-06,
+      "loss": 0.047,
+      "step": 3758
+    },
+    {
+      "epoch": 1.6695536309127248,
+      "grad_norm": 0.6025767202418845,
+      "learning_rate": 7.231957645193943e-06,
+      "loss": 0.0375,
+      "step": 3759
+    },
+    {
+      "epoch": 1.6699977792582723,
+      "grad_norm": 0.4473170515938267,
+      "learning_rate": 7.2302228016207666e-06,
+      "loss": 0.0483,
+      "step": 3760
+    },
+    {
+      "epoch": 1.6704419276038198,
+      "grad_norm": 0.677052589914415,
+      "learning_rate": 7.2284876228072195e-06,
+      "loss": 0.0635,
+      "step": 3761
+    },
+    {
+      "epoch": 1.6708860759493671,
+      "grad_norm": 0.5710811815695489,
+      "learning_rate": 7.226752109014127e-06,
+      "loss": 0.0464,
+      "step": 3762
+    },
+    {
+      "epoch": 1.6713302242949144,
+      "grad_norm": 0.36528232704433305,
+      "learning_rate": 7.225016260502366e-06,
+      "loss": 0.0372,
+      "step": 3763
+    },
+    {
+      "epoch": 1.671774372640462,
+      "grad_norm": 0.5011404570603847,
+      "learning_rate": 7.223280077532866e-06,
+      "loss": 0.0361,
+      "step": 3764
+    },
+    {
+      "epoch": 1.6722185209860094,
+      "grad_norm": 0.9792675536360905,
+      "learning_rate": 7.221543560366602e-06,
+      "loss": 0.0695,
+      "step": 3765
+    },
+    {
+      "epoch": 1.6726626693315567,
+      "grad_norm": 0.40690296889128674,
+      "learning_rate": 7.219806709264605e-06,
+      "loss": 0.0434,
+      "step": 3766
+    },
+    {
+      "epoch": 1.673106817677104,
+      "grad_norm": 0.454187985377748,
+      "learning_rate": 7.21806952448795e-06,
+      "loss": 0.0353,
+      "step": 3767
+    },
+    {
+      "epoch": 1.6735509660226515,
+      "grad_norm": 0.41286435056484855,
+      "learning_rate": 7.216332006297769e-06,
+      "loss": 0.0379,
+      "step": 3768
+    },
+    {
+      "epoch": 1.673995114368199,
+      "grad_norm": 0.6022194735287347,
+      "learning_rate": 7.2145941549552364e-06,
+      "loss": 0.0491,
+      "step": 3769
+    },
+    {
+      "epoch": 1.6744392627137463,
+      "grad_norm": 0.368180906342434,
+      "learning_rate": 7.212855970721584e-06,
+      "loss": 0.0302,
+      "step": 3770
+    },
+    {
+      "epoch": 1.6748834110592938,
+      "grad_norm": 0.3923410824291727,
+      "learning_rate": 7.211117453858088e-06,
+      "loss": 0.0434,
+      "step": 3771
+    },
+    {
+      "epoch": 1.6753275594048413,
+      "grad_norm": 0.5009695065668602,
+      "learning_rate": 7.209378604626081e-06,
+      "loss": 0.042,
+      "step": 3772
+    },
+    {
+      "epoch": 1.6757717077503886,
+      "grad_norm": 0.41349339679074837,
+      "learning_rate": 7.207639423286938e-06,
+      "loss": 0.0363,
+      "step": 3773
+    },
+    {
+      "epoch": 1.676215856095936,
+      "grad_norm": 0.363113317588028,
+      "learning_rate": 7.205899910102087e-06,
+      "loss": 0.0502,
+      "step": 3774
+    },
+    {
+      "epoch": 1.6766600044414834,
+      "grad_norm": 0.5667120483543412,
+      "learning_rate": 7.204160065333009e-06,
+      "loss": 0.047,
+      "step": 3775
+    },
+    {
+      "epoch": 1.677104152787031,
+      "grad_norm": 0.6591252875236339,
+      "learning_rate": 7.202419889241231e-06,
+      "loss": 0.0622,
+      "step": 3776
+    },
+    {
+      "epoch": 1.6775483011325782,
+      "grad_norm": 0.4627599954413904,
+      "learning_rate": 7.2006793820883315e-06,
+      "loss": 0.0294,
+      "step": 3777
+    },
+    {
+      "epoch": 1.6779924494781255,
+      "grad_norm": 0.35617793994157315,
+      "learning_rate": 7.198938544135936e-06,
+      "loss": 0.0302,
+      "step": 3778
+    },
+    {
+      "epoch": 1.6784365978236733,
+      "grad_norm": 0.4097862631071463,
+      "learning_rate": 7.197197375645724e-06,
+      "loss": 0.0414,
+      "step": 3779
+    },
+    {
+      "epoch": 1.6788807461692206,
+      "grad_norm": 0.675818069401739,
+      "learning_rate": 7.195455876879425e-06,
+      "loss": 0.0386,
+      "step": 3780
+    },
+    {
+      "epoch": 1.6793248945147679,
+      "grad_norm": 0.5308299724023986,
+      "learning_rate": 7.193714048098812e-06,
+      "loss": 0.0425,
+      "step": 3781
+    },
+    {
+      "epoch": 1.6797690428603154,
+      "grad_norm": 0.5374963949654195,
+      "learning_rate": 7.191971889565713e-06,
+      "loss": 0.0381,
+      "step": 3782
+    },
+    {
+      "epoch": 1.6802131912058629,
+      "grad_norm": 0.5783394947551381,
+      "learning_rate": 7.190229401542004e-06,
+      "loss": 0.0568,
+      "step": 3783
+    },
+    {
+      "epoch": 1.6806573395514102,
+      "grad_norm": 0.5082898930645872,
+      "learning_rate": 7.18848658428961e-06,
+      "loss": 0.0421,
+      "step": 3784
+    },
+    {
+      "epoch": 1.6811014878969575,
+      "grad_norm": 0.4025027461987578,
+      "learning_rate": 7.186743438070507e-06,
+      "loss": 0.0311,
+      "step": 3785
+    },
+    {
+      "epoch": 1.681545636242505,
+      "grad_norm": 0.3763513501474927,
+      "learning_rate": 7.1849999631467194e-06,
+      "loss": 0.0334,
+      "step": 3786
+    },
+    {
+      "epoch": 1.6819897845880525,
+      "grad_norm": 0.4272449780002536,
+      "learning_rate": 7.183256159780321e-06,
+      "loss": 0.0461,
+      "step": 3787
+    },
+    {
+      "epoch": 1.6824339329335998,
+      "grad_norm": 0.6764426293492224,
+      "learning_rate": 7.181512028233433e-06,
+      "loss": 0.0494,
+      "step": 3788
+    },
+    {
+      "epoch": 1.6828780812791473,
+      "grad_norm": 0.4493107523320909,
+      "learning_rate": 7.17976756876823e-06,
+      "loss": 0.0407,
+      "step": 3789
+    },
+    {
+      "epoch": 1.6833222296246948,
+      "grad_norm": 0.5311611910450604,
+      "learning_rate": 7.178022781646936e-06,
+      "loss": 0.0446,
+      "step": 3790
+    },
+    {
+      "epoch": 1.683766377970242,
+      "grad_norm": 0.49447252037240147,
+      "learning_rate": 7.176277667131817e-06,
+      "loss": 0.0418,
+      "step": 3791
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.5692208839762621,
+      "learning_rate": 7.1745322254851966e-06,
+      "loss": 0.0492,
+      "step": 3792
+    },
+    {
+      "epoch": 1.684654674661337,
+      "grad_norm": 0.5169615968797148,
+      "learning_rate": 7.172786456969445e-06,
+      "loss": 0.0529,
+      "step": 3793
+    },
+    {
+      "epoch": 1.6850988230068844,
+      "grad_norm": 0.393440341612259,
+      "learning_rate": 7.171040361846979e-06,
+      "loss": 0.0467,
+      "step": 3794
+    },
+    {
+      "epoch": 1.6855429713524317,
+      "grad_norm": 0.4049011862637367,
+      "learning_rate": 7.1692939403802676e-06,
+      "loss": 0.0395,
+      "step": 3795
+    },
+    {
+      "epoch": 1.685987119697979,
+      "grad_norm": 0.33756471256165815,
+      "learning_rate": 7.167547192831827e-06,
+      "loss": 0.0321,
+      "step": 3796
+    },
+    {
+      "epoch": 1.6864312680435265,
+      "grad_norm": 0.4670323249605969,
+      "learning_rate": 7.1658001194642225e-06,
+      "loss": 0.0433,
+      "step": 3797
+    },
+    {
+      "epoch": 1.686875416389074,
+      "grad_norm": 0.47348531343201694,
+      "learning_rate": 7.16405272054007e-06,
+      "loss": 0.0391,
+      "step": 3798
+    },
+    {
+      "epoch": 1.6873195647346213,
+      "grad_norm": 0.5297792409513463,
+      "learning_rate": 7.1623049963220325e-06,
+      "loss": 0.0396,
+      "step": 3799
+    },
+    {
+      "epoch": 1.6877637130801688,
+      "grad_norm": 0.6022176716427052,
+      "learning_rate": 7.160556947072823e-06,
+      "loss": 0.0457,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6882078614257163,
+      "grad_norm": 0.4342556651872498,
+      "learning_rate": 7.158808573055205e-06,
+      "loss": 0.0406,
+      "step": 3801
+    },
+    {
+      "epoch": 1.6886520097712636,
+      "grad_norm": 0.571370547347438,
+      "learning_rate": 7.157059874531982e-06,
+      "loss": 0.039,
+      "step": 3802
+    },
+    {
+      "epoch": 1.689096158116811,
+      "grad_norm": 0.6752816385308873,
+      "learning_rate": 7.155310851766022e-06,
+      "loss": 0.0518,
+      "step": 3803
+    },
+    {
+      "epoch": 1.6895403064623584,
+      "grad_norm": 0.4681922042318178,
+      "learning_rate": 7.153561505020228e-06,
+      "loss": 0.0487,
+      "step": 3804
+    },
+    {
+      "epoch": 1.689984454807906,
+      "grad_norm": 0.4722352367968629,
+      "learning_rate": 7.151811834557556e-06,
+      "loss": 0.0506,
+      "step": 3805
+    },
+    {
+      "epoch": 1.6904286031534532,
+      "grad_norm": 0.46550029686119654,
+      "learning_rate": 7.150061840641012e-06,
+      "loss": 0.042,
+      "step": 3806
+    },
+    {
+      "epoch": 1.6908727514990005,
+      "grad_norm": 0.46380378977431014,
+      "learning_rate": 7.148311523533652e-06,
+      "loss": 0.051,
+      "step": 3807
+    },
+    {
+      "epoch": 1.691316899844548,
+      "grad_norm": 0.40338431598829894,
+      "learning_rate": 7.146560883498575e-06,
+      "loss": 0.0358,
+      "step": 3808
+    },
+    {
+      "epoch": 1.6917610481900955,
+      "grad_norm": 0.5997068568205779,
+      "learning_rate": 7.144809920798934e-06,
+      "loss": 0.0549,
+      "step": 3809
+    },
+    {
+      "epoch": 1.6922051965356428,
+      "grad_norm": 0.746314445865342,
+      "learning_rate": 7.143058635697928e-06,
+      "loss": 0.0572,
+      "step": 3810
+    },
+    {
+      "epoch": 1.6926493448811903,
+      "grad_norm": 0.6126953828556693,
+      "learning_rate": 7.141307028458805e-06,
+      "loss": 0.0581,
+      "step": 3811
+    },
+    {
+      "epoch": 1.6930934932267379,
+      "grad_norm": 0.45183518843270887,
+      "learning_rate": 7.13955509934486e-06,
+      "loss": 0.0527,
+      "step": 3812
+    },
+    {
+      "epoch": 1.6935376415722851,
+      "grad_norm": 0.40741342750683424,
+      "learning_rate": 7.137802848619442e-06,
+      "loss": 0.0448,
+      "step": 3813
+    },
+    {
+      "epoch": 1.6939817899178324,
+      "grad_norm": 0.409139232347241,
+      "learning_rate": 7.136050276545937e-06,
+      "loss": 0.0381,
+      "step": 3814
+    },
+    {
+      "epoch": 1.69442593826338,
+      "grad_norm": 0.532572512466246,
+      "learning_rate": 7.134297383387794e-06,
+      "loss": 0.0541,
+      "step": 3815
+    },
+    {
+      "epoch": 1.6948700866089275,
+      "grad_norm": 0.38463725645897007,
+      "learning_rate": 7.1325441694084955e-06,
+      "loss": 0.0413,
+      "step": 3816
+    },
+    {
+      "epoch": 1.6953142349544748,
+      "grad_norm": 1.088465520533885,
+      "learning_rate": 7.130790634871585e-06,
+      "loss": 0.052,
+      "step": 3817
+    },
+    {
+      "epoch": 1.695758383300022,
+      "grad_norm": 0.629929584219581,
+      "learning_rate": 7.129036780040646e-06,
+      "loss": 0.0567,
+      "step": 3818
+    },
+    {
+      "epoch": 1.6962025316455698,
+      "grad_norm": 0.63053218226347,
+      "learning_rate": 7.127282605179311e-06,
+      "loss": 0.0532,
+      "step": 3819
+    },
+    {
+      "epoch": 1.696646679991117,
+      "grad_norm": 0.4433559445013171,
+      "learning_rate": 7.125528110551266e-06,
+      "loss": 0.0463,
+      "step": 3820
+    },
+    {
+      "epoch": 1.6970908283366644,
+      "grad_norm": 0.42229460674035696,
+      "learning_rate": 7.12377329642024e-06,
+      "loss": 0.0397,
+      "step": 3821
+    },
+    {
+      "epoch": 1.6975349766822119,
+      "grad_norm": 0.49873647457140674,
+      "learning_rate": 7.122018163050011e-06,
+      "loss": 0.0545,
+      "step": 3822
+    },
+    {
+      "epoch": 1.6979791250277594,
+      "grad_norm": 0.6060966583878594,
+      "learning_rate": 7.1202627107044035e-06,
+      "loss": 0.0457,
+      "step": 3823
+    },
+    {
+      "epoch": 1.6984232733733067,
+      "grad_norm": 0.603057650967115,
+      "learning_rate": 7.118506939647295e-06,
+      "loss": 0.0464,
+      "step": 3824
+    },
+    {
+      "epoch": 1.698867421718854,
+      "grad_norm": 0.6194476035587843,
+      "learning_rate": 7.116750850142606e-06,
+      "loss": 0.0457,
+      "step": 3825
+    },
+    {
+      "epoch": 1.6993115700644015,
+      "grad_norm": 0.5096734124949287,
+      "learning_rate": 7.114994442454306e-06,
+      "loss": 0.045,
+      "step": 3826
+    },
+    {
+      "epoch": 1.699755718409949,
+      "grad_norm": 0.5365035460325245,
+      "learning_rate": 7.113237716846416e-06,
+      "loss": 0.0623,
+      "step": 3827
+    },
+    {
+      "epoch": 1.7001998667554963,
+      "grad_norm": 0.4418379177592074,
+      "learning_rate": 7.111480673582998e-06,
+      "loss": 0.0434,
+      "step": 3828
+    },
+    {
+      "epoch": 1.7006440151010438,
+      "grad_norm": 0.5486324781269635,
+      "learning_rate": 7.1097233129281674e-06,
+      "loss": 0.0515,
+      "step": 3829
+    },
+    {
+      "epoch": 1.7010881634465913,
+      "grad_norm": 0.46832429223654365,
+      "learning_rate": 7.107965635146085e-06,
+      "loss": 0.036,
+      "step": 3830
+    },
+    {
+      "epoch": 1.7015323117921386,
+      "grad_norm": 0.4428576565727785,
+      "learning_rate": 7.106207640500959e-06,
+      "loss": 0.0431,
+      "step": 3831
+    },
+    {
+      "epoch": 1.7019764601376859,
+      "grad_norm": 0.44260932357845195,
+      "learning_rate": 7.104449329257047e-06,
+      "loss": 0.0488,
+      "step": 3832
+    },
+    {
+      "epoch": 1.7024206084832334,
+      "grad_norm": 0.6480909417028274,
+      "learning_rate": 7.10269070167865e-06,
+      "loss": 0.0538,
+      "step": 3833
+    },
+    {
+      "epoch": 1.702864756828781,
+      "grad_norm": 0.5907643196584621,
+      "learning_rate": 7.100931758030126e-06,
+      "loss": 0.0667,
+      "step": 3834
+    },
+    {
+      "epoch": 1.7033089051743282,
+      "grad_norm": 0.530721353895675,
+      "learning_rate": 7.0991724985758694e-06,
+      "loss": 0.0355,
+      "step": 3835
+    },
+    {
+      "epoch": 1.7037530535198755,
+      "grad_norm": 0.536328694373341,
+      "learning_rate": 7.0974129235803256e-06,
+      "loss": 0.0492,
+      "step": 3836
+    },
+    {
+      "epoch": 1.704197201865423,
+      "grad_norm": 0.47794780473543447,
+      "learning_rate": 7.095653033307992e-06,
+      "loss": 0.0394,
+      "step": 3837
+    },
+    {
+      "epoch": 1.7046413502109705,
+      "grad_norm": 0.5788739767339988,
+      "learning_rate": 7.093892828023408e-06,
+      "loss": 0.0684,
+      "step": 3838
+    },
+    {
+      "epoch": 1.7050854985565178,
+      "grad_norm": 0.4591656439375814,
+      "learning_rate": 7.092132307991163e-06,
+      "loss": 0.0426,
+      "step": 3839
+    },
+    {
+      "epoch": 1.7055296469020653,
+      "grad_norm": 0.4376969955300339,
+      "learning_rate": 7.090371473475894e-06,
+      "loss": 0.0424,
+      "step": 3840
+    },
+    {
+      "epoch": 1.7059737952476128,
+      "grad_norm": 0.43964430921337727,
+      "learning_rate": 7.088610324742282e-06,
+      "loss": 0.0348,
+      "step": 3841
+    },
+    {
+      "epoch": 1.7064179435931601,
+      "grad_norm": 0.40903365943544095,
+      "learning_rate": 7.086848862055059e-06,
+      "loss": 0.0317,
+      "step": 3842
+    },
+    {
+      "epoch": 1.7068620919387074,
+      "grad_norm": 0.5320572070600705,
+      "learning_rate": 7.085087085679003e-06,
+      "loss": 0.0493,
+      "step": 3843
+    },
+    {
+      "epoch": 1.707306240284255,
+      "grad_norm": 0.4182055882812196,
+      "learning_rate": 7.0833249958789396e-06,
+      "loss": 0.0413,
+      "step": 3844
+    },
+    {
+      "epoch": 1.7077503886298024,
+      "grad_norm": 0.5174573980801822,
+      "learning_rate": 7.081562592919737e-06,
+      "loss": 0.0468,
+      "step": 3845
+    },
+    {
+      "epoch": 1.7081945369753497,
+      "grad_norm": 0.36708254507339966,
+      "learning_rate": 7.07979987706632e-06,
+      "loss": 0.0392,
+      "step": 3846
+    },
+    {
+      "epoch": 1.708638685320897,
+      "grad_norm": 0.42988860867301804,
+      "learning_rate": 7.078036848583651e-06,
+      "loss": 0.0438,
+      "step": 3847
+    },
+    {
+      "epoch": 1.7090828336664448,
+      "grad_norm": 0.8624108371914289,
+      "learning_rate": 7.076273507736744e-06,
+      "loss": 0.0564,
+      "step": 3848
+    },
+    {
+      "epoch": 1.709526982011992,
+      "grad_norm": 0.3825294027975058,
+      "learning_rate": 7.074509854790659e-06,
+      "loss": 0.042,
+      "step": 3849
+    },
+    {
+      "epoch": 1.7099711303575393,
+      "grad_norm": 0.7845280836992656,
+      "learning_rate": 7.072745890010502e-06,
+      "loss": 0.0513,
+      "step": 3850
+    },
+    {
+      "epoch": 1.7104152787030868,
+      "grad_norm": 0.511895797137735,
+      "learning_rate": 7.070981613661429e-06,
+      "loss": 0.042,
+      "step": 3851
+    },
+    {
+      "epoch": 1.7108594270486344,
+      "grad_norm": 0.39864960740423,
+      "learning_rate": 7.06921702600864e-06,
+      "loss": 0.0435,
+      "step": 3852
+    },
+    {
+      "epoch": 1.7113035753941817,
+      "grad_norm": 0.3728345248612773,
+      "learning_rate": 7.067452127317381e-06,
+      "loss": 0.0414,
+      "step": 3853
+    },
+    {
+      "epoch": 1.711747723739729,
+      "grad_norm": 0.4146176183523404,
+      "learning_rate": 7.065686917852948e-06,
+      "loss": 0.0394,
+      "step": 3854
+    },
+    {
+      "epoch": 1.7121918720852765,
+      "grad_norm": 0.4550263965386846,
+      "learning_rate": 7.063921397880682e-06,
+      "loss": 0.0353,
+      "step": 3855
+    },
+    {
+      "epoch": 1.712636020430824,
+      "grad_norm": 0.6273234491307014,
+      "learning_rate": 7.062155567665969e-06,
+      "loss": 0.0336,
+      "step": 3856
+    },
+    {
+      "epoch": 1.7130801687763713,
+      "grad_norm": 0.5739885387932523,
+      "learning_rate": 7.0603894274742445e-06,
+      "loss": 0.0463,
+      "step": 3857
+    },
+    {
+      "epoch": 1.7135243171219188,
+      "grad_norm": 0.46562620789463843,
+      "learning_rate": 7.05862297757099e-06,
+      "loss": 0.0453,
+      "step": 3858
+    },
+    {
+      "epoch": 1.7139684654674663,
+      "grad_norm": 0.5078681314633765,
+      "learning_rate": 7.056856218221731e-06,
+      "loss": 0.0472,
+      "step": 3859
+    },
+    {
+      "epoch": 1.7144126138130136,
+      "grad_norm": 0.44379024742354006,
+      "learning_rate": 7.055089149692044e-06,
+      "loss": 0.0362,
+      "step": 3860
+    },
+    {
+      "epoch": 1.7148567621585609,
+      "grad_norm": 0.6765029010883838,
+      "learning_rate": 7.053321772247546e-06,
+      "loss": 0.0605,
+      "step": 3861
+    },
+    {
+      "epoch": 1.7153009105041084,
+      "grad_norm": 0.5667840551717872,
+      "learning_rate": 7.051554086153907e-06,
+      "loss": 0.052,
+      "step": 3862
+    },
+    {
+      "epoch": 1.715745058849656,
+      "grad_norm": 0.4505940404872499,
+      "learning_rate": 7.049786091676838e-06,
+      "loss": 0.0437,
+      "step": 3863
+    },
+    {
+      "epoch": 1.7161892071952032,
+      "grad_norm": 0.3740985912782501,
+      "learning_rate": 7.0480177890821e-06,
+      "loss": 0.034,
+      "step": 3864
+    },
+    {
+      "epoch": 1.7166333555407505,
+      "grad_norm": 0.5770251755986612,
+      "learning_rate": 7.046249178635499e-06,
+      "loss": 0.0502,
+      "step": 3865
+    },
+    {
+      "epoch": 1.717077503886298,
+      "grad_norm": 0.5237695494745064,
+      "learning_rate": 7.044480260602888e-06,
+      "loss": 0.0565,
+      "step": 3866
+    },
+    {
+      "epoch": 1.7175216522318455,
+      "grad_norm": 0.49695229712892924,
+      "learning_rate": 7.042711035250162e-06,
+      "loss": 0.0526,
+      "step": 3867
+    },
+    {
+      "epoch": 1.7179658005773928,
+      "grad_norm": 0.7710570644630359,
+      "learning_rate": 7.0409415028432685e-06,
+      "loss": 0.0353,
+      "step": 3868
+    },
+    {
+      "epoch": 1.7184099489229403,
+      "grad_norm": 0.5569345846072523,
+      "learning_rate": 7.0391716636481976e-06,
+      "loss": 0.0463,
+      "step": 3869
+    },
+    {
+      "epoch": 1.7188540972684878,
+      "grad_norm": 0.46227610895953286,
+      "learning_rate": 7.037401517930986e-06,
+      "loss": 0.0518,
+      "step": 3870
+    },
+    {
+      "epoch": 1.719298245614035,
+      "grad_norm": 0.3855965081893482,
+      "learning_rate": 7.035631065957718e-06,
+      "loss": 0.0345,
+      "step": 3871
+    },
+    {
+      "epoch": 1.7197423939595824,
+      "grad_norm": 0.4363418780122806,
+      "learning_rate": 7.03386030799452e-06,
+      "loss": 0.0492,
+      "step": 3872
+    },
+    {
+      "epoch": 1.72018654230513,
+      "grad_norm": 0.4427385688478732,
+      "learning_rate": 7.03208924430757e-06,
+      "loss": 0.0463,
+      "step": 3873
+    },
+    {
+      "epoch": 1.7206306906506774,
+      "grad_norm": 0.5897859219840801,
+      "learning_rate": 7.030317875163086e-06,
+      "loss": 0.0472,
+      "step": 3874
+    },
+    {
+      "epoch": 1.7210748389962247,
+      "grad_norm": 0.44136482686712686,
+      "learning_rate": 7.0285462008273365e-06,
+      "loss": 0.0339,
+      "step": 3875
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.4341786806553805,
+      "learning_rate": 7.026774221566634e-06,
+      "loss": 0.0479,
+      "step": 3876
+    },
+    {
+      "epoch": 1.7219631356873195,
+      "grad_norm": 0.4859029892336248,
+      "learning_rate": 7.0250019376473375e-06,
+      "loss": 0.0481,
+      "step": 3877
+    },
+    {
+      "epoch": 1.722407284032867,
+      "grad_norm": 0.554393064292429,
+      "learning_rate": 7.0232293493358515e-06,
+      "loss": 0.0423,
+      "step": 3878
+    },
+    {
+      "epoch": 1.7228514323784143,
+      "grad_norm": 0.4954355417965089,
+      "learning_rate": 7.021456456898624e-06,
+      "loss": 0.0529,
+      "step": 3879
+    },
+    {
+      "epoch": 1.7232955807239618,
+      "grad_norm": 0.39348052805418005,
+      "learning_rate": 7.019683260602155e-06,
+      "loss": 0.04,
+      "step": 3880
+    },
+    {
+      "epoch": 1.7237397290695093,
+      "grad_norm": 0.4876831116193254,
+      "learning_rate": 7.017909760712982e-06,
+      "loss": 0.0416,
+      "step": 3881
+    },
+    {
+      "epoch": 1.7241838774150566,
+      "grad_norm": 0.5390273774730024,
+      "learning_rate": 7.016135957497693e-06,
+      "loss": 0.0685,
+      "step": 3882
+    },
+    {
+      "epoch": 1.724628025760604,
+      "grad_norm": 0.553831689747292,
+      "learning_rate": 7.014361851222923e-06,
+      "loss": 0.0575,
+      "step": 3883
+    },
+    {
+      "epoch": 1.7250721741061514,
+      "grad_norm": 0.5211453167042258,
+      "learning_rate": 7.012587442155349e-06,
+      "loss": 0.0508,
+      "step": 3884
+    },
+    {
+      "epoch": 1.725516322451699,
+      "grad_norm": 0.5052516566700862,
+      "learning_rate": 7.010812730561691e-06,
+      "loss": 0.0411,
+      "step": 3885
+    },
+    {
+      "epoch": 1.7259604707972462,
+      "grad_norm": 0.5073428394585604,
+      "learning_rate": 7.009037716708725e-06,
+      "loss": 0.0426,
+      "step": 3886
+    },
+    {
+      "epoch": 1.7264046191427935,
+      "grad_norm": 0.48746152511684293,
+      "learning_rate": 7.007262400863262e-06,
+      "loss": 0.043,
+      "step": 3887
+    },
+    {
+      "epoch": 1.7268487674883413,
+      "grad_norm": 0.48800631902973207,
+      "learning_rate": 7.005486783292164e-06,
+      "loss": 0.0416,
+      "step": 3888
+    },
+    {
+      "epoch": 1.7272929158338886,
+      "grad_norm": 0.5824704615616569,
+      "learning_rate": 7.003710864262333e-06,
+      "loss": 0.0485,
+      "step": 3889
+    },
+    {
+      "epoch": 1.7277370641794358,
+      "grad_norm": 0.4562507795280803,
+      "learning_rate": 7.0019346440407225e-06,
+      "loss": 0.0392,
+      "step": 3890
+    },
+    {
+      "epoch": 1.7281812125249834,
+      "grad_norm": 0.5660300341993364,
+      "learning_rate": 7.000158122894329e-06,
+      "loss": 0.0443,
+      "step": 3891
+    },
+    {
+      "epoch": 1.7286253608705309,
+      "grad_norm": 1.2237037452955262,
+      "learning_rate": 6.9983813010901925e-06,
+      "loss": 0.0446,
+      "step": 3892
+    },
+    {
+      "epoch": 1.7290695092160782,
+      "grad_norm": 0.3834611348340635,
+      "learning_rate": 6.996604178895398e-06,
+      "loss": 0.0321,
+      "step": 3893
+    },
+    {
+      "epoch": 1.7295136575616255,
+      "grad_norm": 0.500508115884937,
+      "learning_rate": 6.994826756577082e-06,
+      "loss": 0.0547,
+      "step": 3894
+    },
+    {
+      "epoch": 1.729957805907173,
+      "grad_norm": 0.41016179375129264,
+      "learning_rate": 6.993049034402417e-06,
+      "loss": 0.0299,
+      "step": 3895
+    },
+    {
+      "epoch": 1.7304019542527205,
+      "grad_norm": 0.5423253130879615,
+      "learning_rate": 6.991271012638626e-06,
+      "loss": 0.0616,
+      "step": 3896
+    },
+    {
+      "epoch": 1.7308461025982678,
+      "grad_norm": 0.5282418101161326,
+      "learning_rate": 6.9894926915529774e-06,
+      "loss": 0.0457,
+      "step": 3897
+    },
+    {
+      "epoch": 1.7312902509438153,
+      "grad_norm": 0.4989861140225304,
+      "learning_rate": 6.987714071412781e-06,
+      "loss": 0.0408,
+      "step": 3898
+    },
+    {
+      "epoch": 1.7317343992893628,
+      "grad_norm": 0.48460144374543296,
+      "learning_rate": 6.985935152485392e-06,
+      "loss": 0.0508,
+      "step": 3899
+    },
+    {
+      "epoch": 1.73217854763491,
+      "grad_norm": 0.37082720341713776,
+      "learning_rate": 6.984155935038217e-06,
+      "loss": 0.0376,
+      "step": 3900
+    },
+    {
+      "epoch": 1.7326226959804574,
+      "grad_norm": 0.6794555649441637,
+      "learning_rate": 6.9823764193387e-06,
+      "loss": 0.0568,
+      "step": 3901
+    },
+    {
+      "epoch": 1.7330668443260049,
+      "grad_norm": 0.5551808325017253,
+      "learning_rate": 6.980596605654332e-06,
+      "loss": 0.0502,
+      "step": 3902
+    },
+    {
+      "epoch": 1.7335109926715524,
+      "grad_norm": 0.5306956016396489,
+      "learning_rate": 6.9788164942526495e-06,
+      "loss": 0.0432,
+      "step": 3903
+    },
+    {
+      "epoch": 1.7339551410170997,
+      "grad_norm": 0.6362458412949452,
+      "learning_rate": 6.977036085401234e-06,
+      "loss": 0.0486,
+      "step": 3904
+    },
+    {
+      "epoch": 1.734399289362647,
+      "grad_norm": 0.29485239085618,
+      "learning_rate": 6.9752553793677105e-06,
+      "loss": 0.0245,
+      "step": 3905
+    },
+    {
+      "epoch": 1.7348434377081945,
+      "grad_norm": 0.5487150575741024,
+      "learning_rate": 6.9734743764197485e-06,
+      "loss": 0.0463,
+      "step": 3906
+    },
+    {
+      "epoch": 1.735287586053742,
+      "grad_norm": 0.7871819167162935,
+      "learning_rate": 6.9716930768250655e-06,
+      "loss": 0.0524,
+      "step": 3907
+    },
+    {
+      "epoch": 1.7357317343992893,
+      "grad_norm": 0.5219429848305476,
+      "learning_rate": 6.9699114808514215e-06,
+      "loss": 0.0549,
+      "step": 3908
+    },
+    {
+      "epoch": 1.7361758827448368,
+      "grad_norm": 0.4635849498371789,
+      "learning_rate": 6.968129588766617e-06,
+      "loss": 0.0604,
+      "step": 3909
+    },
+    {
+      "epoch": 1.7366200310903843,
+      "grad_norm": 0.43616273854041554,
+      "learning_rate": 6.966347400838502e-06,
+      "loss": 0.0373,
+      "step": 3910
+    },
+    {
+      "epoch": 1.7370641794359316,
+      "grad_norm": 0.593451901083156,
+      "learning_rate": 6.964564917334973e-06,
+      "loss": 0.0527,
+      "step": 3911
+    },
+    {
+      "epoch": 1.737508327781479,
+      "grad_norm": 0.6448036909846457,
+      "learning_rate": 6.962782138523963e-06,
+      "loss": 0.048,
+      "step": 3912
+    },
+    {
+      "epoch": 1.7379524761270264,
+      "grad_norm": 0.6507383892986597,
+      "learning_rate": 6.960999064673455e-06,
+      "loss": 0.0405,
+      "step": 3913
+    },
+    {
+      "epoch": 1.738396624472574,
+      "grad_norm": 0.3700619002880566,
+      "learning_rate": 6.959215696051478e-06,
+      "loss": 0.0375,
+      "step": 3914
+    },
+    {
+      "epoch": 1.7388407728181212,
+      "grad_norm": 0.5636095366670266,
+      "learning_rate": 6.957432032926099e-06,
+      "loss": 0.0473,
+      "step": 3915
+    },
+    {
+      "epoch": 1.7392849211636685,
+      "grad_norm": 0.31284491700710365,
+      "learning_rate": 6.955648075565435e-06,
+      "loss": 0.0262,
+      "step": 3916
+    },
+    {
+      "epoch": 1.7397290695092162,
+      "grad_norm": 0.88892656741503,
+      "learning_rate": 6.953863824237644e-06,
+      "loss": 0.0439,
+      "step": 3917
+    },
+    {
+      "epoch": 1.7401732178547635,
+      "grad_norm": 0.3940650039553802,
+      "learning_rate": 6.952079279210931e-06,
+      "loss": 0.0336,
+      "step": 3918
+    },
+    {
+      "epoch": 1.7406173662003108,
+      "grad_norm": 0.43415099426990134,
+      "learning_rate": 6.950294440753542e-06,
+      "loss": 0.0372,
+      "step": 3919
+    },
+    {
+      "epoch": 1.7410615145458583,
+      "grad_norm": 0.43498781141215753,
+      "learning_rate": 6.948509309133769e-06,
+      "loss": 0.0471,
+      "step": 3920
+    },
+    {
+      "epoch": 1.7415056628914058,
+      "grad_norm": 0.3673155534955715,
+      "learning_rate": 6.9467238846199465e-06,
+      "loss": 0.0329,
+      "step": 3921
+    },
+    {
+      "epoch": 1.7419498112369531,
+      "grad_norm": 0.3522859182784973,
+      "learning_rate": 6.944938167480456e-06,
+      "loss": 0.0337,
+      "step": 3922
+    },
+    {
+      "epoch": 1.7423939595825004,
+      "grad_norm": 0.40635886553673484,
+      "learning_rate": 6.943152157983719e-06,
+      "loss": 0.0429,
+      "step": 3923
+    },
+    {
+      "epoch": 1.742838107928048,
+      "grad_norm": 0.8451883765559832,
+      "learning_rate": 6.941365856398205e-06,
+      "loss": 0.0668,
+      "step": 3924
+    },
+    {
+      "epoch": 1.7432822562735955,
+      "grad_norm": 0.5600822140198348,
+      "learning_rate": 6.939579262992426e-06,
+      "loss": 0.0451,
+      "step": 3925
+    },
+    {
+      "epoch": 1.7437264046191427,
+      "grad_norm": 1.2545344075254676,
+      "learning_rate": 6.937792378034936e-06,
+      "loss": 0.0743,
+      "step": 3926
+    },
+    {
+      "epoch": 1.7441705529646903,
+      "grad_norm": 1.1071420667016871,
+      "learning_rate": 6.936005201794331e-06,
+      "loss": 0.073,
+      "step": 3927
+    },
+    {
+      "epoch": 1.7446147013102378,
+      "grad_norm": 0.4227881516723295,
+      "learning_rate": 6.93421773453926e-06,
+      "loss": 0.043,
+      "step": 3928
+    },
+    {
+      "epoch": 1.745058849655785,
+      "grad_norm": 0.46137549848480547,
+      "learning_rate": 6.932429976538407e-06,
+      "loss": 0.0413,
+      "step": 3929
+    },
+    {
+      "epoch": 1.7455029980013324,
+      "grad_norm": 0.5679012650124624,
+      "learning_rate": 6.930641928060501e-06,
+      "loss": 0.0331,
+      "step": 3930
+    },
+    {
+      "epoch": 1.7459471463468799,
+      "grad_norm": 0.5321986376581693,
+      "learning_rate": 6.928853589374318e-06,
+      "loss": 0.0467,
+      "step": 3931
+    },
+    {
+      "epoch": 1.7463912946924274,
+      "grad_norm": 0.6219676953962676,
+      "learning_rate": 6.927064960748675e-06,
+      "loss": 0.0479,
+      "step": 3932
+    },
+    {
+      "epoch": 1.7468354430379747,
+      "grad_norm": 0.5001191355293207,
+      "learning_rate": 6.925276042452433e-06,
+      "loss": 0.0535,
+      "step": 3933
+    },
+    {
+      "epoch": 1.747279591383522,
+      "grad_norm": 0.4563948076269455,
+      "learning_rate": 6.923486834754498e-06,
+      "loss": 0.0427,
+      "step": 3934
+    },
+    {
+      "epoch": 1.7477237397290695,
+      "grad_norm": 0.6974643048935755,
+      "learning_rate": 6.9216973379238175e-06,
+      "loss": 0.0485,
+      "step": 3935
+    },
+    {
+      "epoch": 1.748167888074617,
+      "grad_norm": 0.9428504067985796,
+      "learning_rate": 6.9199075522293815e-06,
+      "loss": 0.0612,
+      "step": 3936
+    },
+    {
+      "epoch": 1.7486120364201643,
+      "grad_norm": 0.8097896480889225,
+      "learning_rate": 6.918117477940227e-06,
+      "loss": 0.0467,
+      "step": 3937
+    },
+    {
+      "epoch": 1.7490561847657118,
+      "grad_norm": 0.5120773769465646,
+      "learning_rate": 6.916327115325434e-06,
+      "loss": 0.0554,
+      "step": 3938
+    },
+    {
+      "epoch": 1.7495003331112593,
+      "grad_norm": 0.49704552104919647,
+      "learning_rate": 6.914536464654123e-06,
+      "loss": 0.0429,
+      "step": 3939
+    },
+    {
+      "epoch": 1.7499444814568066,
+      "grad_norm": 0.4966530008001101,
+      "learning_rate": 6.912745526195457e-06,
+      "loss": 0.0416,
+      "step": 3940
+    },
+    {
+      "epoch": 1.7503886298023539,
+      "grad_norm": 0.6069162585490371,
+      "learning_rate": 6.910954300218648e-06,
+      "loss": 0.0407,
+      "step": 3941
+    },
+    {
+      "epoch": 1.7508327781479014,
+      "grad_norm": 0.5065891450556588,
+      "learning_rate": 6.9091627869929456e-06,
+      "loss": 0.0522,
+      "step": 3942
+    },
+    {
+      "epoch": 1.751276926493449,
+      "grad_norm": 0.5821716098410316,
+      "learning_rate": 6.907370986787647e-06,
+      "loss": 0.0588,
+      "step": 3943
+    },
+    {
+      "epoch": 1.7517210748389962,
+      "grad_norm": 0.4854527206747191,
+      "learning_rate": 6.905578899872085e-06,
+      "loss": 0.0347,
+      "step": 3944
+    },
+    {
+      "epoch": 1.7521652231845435,
+      "grad_norm": 0.6099586438595753,
+      "learning_rate": 6.903786526515648e-06,
+      "loss": 0.057,
+      "step": 3945
+    },
+    {
+      "epoch": 1.752609371530091,
+      "grad_norm": 0.39621484939213947,
+      "learning_rate": 6.901993866987755e-06,
+      "loss": 0.036,
+      "step": 3946
+    },
+    {
+      "epoch": 1.7530535198756385,
+      "grad_norm": 0.5825955749635414,
+      "learning_rate": 6.9002009215578736e-06,
+      "loss": 0.0371,
+      "step": 3947
+    },
+    {
+      "epoch": 1.7534976682211858,
+      "grad_norm": 0.5518034617506287,
+      "learning_rate": 6.898407690495516e-06,
+      "loss": 0.0535,
+      "step": 3948
+    },
+    {
+      "epoch": 1.7539418165667333,
+      "grad_norm": 0.7169225712774631,
+      "learning_rate": 6.896614174070234e-06,
+      "loss": 0.065,
+      "step": 3949
+    },
+    {
+      "epoch": 1.7543859649122808,
+      "grad_norm": 0.487246075580528,
+      "learning_rate": 6.894820372551624e-06,
+      "loss": 0.0425,
+      "step": 3950
+    },
+    {
+      "epoch": 1.7548301132578281,
+      "grad_norm": 0.5762490980340829,
+      "learning_rate": 6.893026286209324e-06,
+      "loss": 0.0579,
+      "step": 3951
+    },
+    {
+      "epoch": 1.7552742616033754,
+      "grad_norm": 0.40568305353842643,
+      "learning_rate": 6.891231915313017e-06,
+      "loss": 0.0447,
+      "step": 3952
+    },
+    {
+      "epoch": 1.755718409948923,
+      "grad_norm": 1.1277438855550295,
+      "learning_rate": 6.889437260132426e-06,
+      "loss": 0.0646,
+      "step": 3953
+    },
+    {
+      "epoch": 1.7561625582944704,
+      "grad_norm": 1.0348468689646162,
+      "learning_rate": 6.887642320937319e-06,
+      "loss": 0.0477,
+      "step": 3954
+    },
+    {
+      "epoch": 1.7566067066400177,
+      "grad_norm": 0.4516096937991291,
+      "learning_rate": 6.885847097997507e-06,
+      "loss": 0.0441,
+      "step": 3955
+    },
+    {
+      "epoch": 1.757050854985565,
+      "grad_norm": 0.5215834296906894,
+      "learning_rate": 6.884051591582838e-06,
+      "loss": 0.0413,
+      "step": 3956
+    },
+    {
+      "epoch": 1.7574950033311127,
+      "grad_norm": 0.4942925713506997,
+      "learning_rate": 6.882255801963215e-06,
+      "loss": 0.0313,
+      "step": 3957
+    },
+    {
+      "epoch": 1.75793915167666,
+      "grad_norm": 0.5511920080731652,
+      "learning_rate": 6.8804597294085676e-06,
+      "loss": 0.0468,
+      "step": 3958
+    },
+    {
+      "epoch": 1.7583833000222073,
+      "grad_norm": 0.8005773203813454,
+      "learning_rate": 6.87866337418888e-06,
+      "loss": 0.0723,
+      "step": 3959
+    },
+    {
+      "epoch": 1.7588274483677548,
+      "grad_norm": 0.6193224773136193,
+      "learning_rate": 6.876866736574175e-06,
+      "loss": 0.0458,
+      "step": 3960
+    },
+    {
+      "epoch": 1.7592715967133024,
+      "grad_norm": 0.7463153621358136,
+      "learning_rate": 6.875069816834517e-06,
+      "loss": 0.0452,
+      "step": 3961
+    },
+    {
+      "epoch": 1.7597157450588496,
+      "grad_norm": 0.6653689016758167,
+      "learning_rate": 6.873272615240013e-06,
+      "loss": 0.0379,
+      "step": 3962
+    },
+    {
+      "epoch": 1.760159893404397,
+      "grad_norm": 0.46506318034990934,
+      "learning_rate": 6.871475132060814e-06,
+      "loss": 0.0405,
+      "step": 3963
+    },
+    {
+      "epoch": 1.7606040417499444,
+      "grad_norm": 0.4803172632231074,
+      "learning_rate": 6.8696773675671125e-06,
+      "loss": 0.0494,
+      "step": 3964
+    },
+    {
+      "epoch": 1.761048190095492,
+      "grad_norm": 0.5683228177486831,
+      "learning_rate": 6.8678793220291406e-06,
+      "loss": 0.0474,
+      "step": 3965
+    },
+    {
+      "epoch": 1.7614923384410393,
+      "grad_norm": 0.6405917034428388,
+      "learning_rate": 6.866080995717179e-06,
+      "loss": 0.0444,
+      "step": 3966
+    },
+    {
+      "epoch": 1.7619364867865868,
+      "grad_norm": 0.469723345316827,
+      "learning_rate": 6.864282388901544e-06,
+      "loss": 0.0456,
+      "step": 3967
+    },
+    {
+      "epoch": 1.7623806351321343,
+      "grad_norm": 0.3596729988674669,
+      "learning_rate": 6.862483501852597e-06,
+      "loss": 0.0273,
+      "step": 3968
+    },
+    {
+      "epoch": 1.7628247834776816,
+      "grad_norm": 0.4095733039990315,
+      "learning_rate": 6.8606843348407416e-06,
+      "loss": 0.0423,
+      "step": 3969
+    },
+    {
+      "epoch": 1.7632689318232289,
+      "grad_norm": 0.5817103130183384,
+      "learning_rate": 6.858884888136423e-06,
+      "loss": 0.061,
+      "step": 3970
+    },
+    {
+      "epoch": 1.7637130801687764,
+      "grad_norm": 0.5515400819468078,
+      "learning_rate": 6.85708516201013e-06,
+      "loss": 0.0493,
+      "step": 3971
+    },
+    {
+      "epoch": 1.7641572285143239,
+      "grad_norm": 0.4010509115861224,
+      "learning_rate": 6.855285156732389e-06,
+      "loss": 0.0384,
+      "step": 3972
+    },
+    {
+      "epoch": 1.7646013768598712,
+      "grad_norm": 0.48651107038259084,
+      "learning_rate": 6.853484872573773e-06,
+      "loss": 0.0417,
+      "step": 3973
+    },
+    {
+      "epoch": 1.7650455252054185,
+      "grad_norm": 0.6804998963023164,
+      "learning_rate": 6.851684309804898e-06,
+      "loss": 0.0511,
+      "step": 3974
+    },
+    {
+      "epoch": 1.765489673550966,
+      "grad_norm": 0.5378576658955843,
+      "learning_rate": 6.849883468696414e-06,
+      "loss": 0.0466,
+      "step": 3975
+    },
+    {
+      "epoch": 1.7659338218965135,
+      "grad_norm": 0.6080913375351253,
+      "learning_rate": 6.848082349519021e-06,
+      "loss": 0.0504,
+      "step": 3976
+    },
+    {
+      "epoch": 1.7663779702420608,
+      "grad_norm": 0.4317853459587004,
+      "learning_rate": 6.846280952543459e-06,
+      "loss": 0.0419,
+      "step": 3977
+    },
+    {
+      "epoch": 1.7668221185876083,
+      "grad_norm": 0.44881873306480574,
+      "learning_rate": 6.844479278040506e-06,
+      "loss": 0.0373,
+      "step": 3978
+    },
+    {
+      "epoch": 1.7672662669331558,
+      "grad_norm": 0.38885972043532946,
+      "learning_rate": 6.842677326280984e-06,
+      "loss": 0.0396,
+      "step": 3979
+    },
+    {
+      "epoch": 1.767710415278703,
+      "grad_norm": 0.4524735899424509,
+      "learning_rate": 6.840875097535761e-06,
+      "loss": 0.0448,
+      "step": 3980
+    },
+    {
+      "epoch": 1.7681545636242504,
+      "grad_norm": 0.455210687429983,
+      "learning_rate": 6.8390725920757374e-06,
+      "loss": 0.0382,
+      "step": 3981
+    },
+    {
+      "epoch": 1.768598711969798,
+      "grad_norm": 0.398331712609469,
+      "learning_rate": 6.837269810171864e-06,
+      "loss": 0.0414,
+      "step": 3982
+    },
+    {
+      "epoch": 1.7690428603153454,
+      "grad_norm": 0.37522143210177494,
+      "learning_rate": 6.835466752095129e-06,
+      "loss": 0.0386,
+      "step": 3983
+    },
+    {
+      "epoch": 1.7694870086608927,
+      "grad_norm": 0.6174104941670406,
+      "learning_rate": 6.833663418116561e-06,
+      "loss": 0.0389,
+      "step": 3984
+    },
+    {
+      "epoch": 1.76993115700644,
+      "grad_norm": 0.6659454494298849,
+      "learning_rate": 6.831859808507233e-06,
+      "loss": 0.0447,
+      "step": 3985
+    },
+    {
+      "epoch": 1.7703753053519877,
+      "grad_norm": 0.47498410446306544,
+      "learning_rate": 6.830055923538258e-06,
+      "loss": 0.0417,
+      "step": 3986
+    },
+    {
+      "epoch": 1.770819453697535,
+      "grad_norm": 0.4396131477174768,
+      "learning_rate": 6.82825176348079e-06,
+      "loss": 0.0372,
+      "step": 3987
+    },
+    {
+      "epoch": 1.7712636020430823,
+      "grad_norm": 0.576388987980822,
+      "learning_rate": 6.826447328606026e-06,
+      "loss": 0.0432,
+      "step": 3988
+    },
+    {
+      "epoch": 1.7717077503886298,
+      "grad_norm": 0.5115559440177404,
+      "learning_rate": 6.8246426191852025e-06,
+      "loss": 0.046,
+      "step": 3989
+    },
+    {
+      "epoch": 1.7721518987341773,
+      "grad_norm": 0.3933392068642798,
+      "learning_rate": 6.822837635489597e-06,
+      "loss": 0.0424,
+      "step": 3990
+    },
+    {
+      "epoch": 1.7725960470797246,
+      "grad_norm": 0.48762159715898146,
+      "learning_rate": 6.821032377790533e-06,
+      "loss": 0.0428,
+      "step": 3991
+    },
+    {
+      "epoch": 1.773040195425272,
+      "grad_norm": 0.5275360991096351,
+      "learning_rate": 6.819226846359366e-06,
+      "loss": 0.0437,
+      "step": 3992
+    },
+    {
+      "epoch": 1.7734843437708194,
+      "grad_norm": 0.5998959208861094,
+      "learning_rate": 6.817421041467501e-06,
+      "loss": 0.0493,
+      "step": 3993
+    },
+    {
+      "epoch": 1.773928492116367,
+      "grad_norm": 0.558883606638067,
+      "learning_rate": 6.815614963386383e-06,
+      "loss": 0.0443,
+      "step": 3994
+    },
+    {
+      "epoch": 1.7743726404619142,
+      "grad_norm": 0.530805377008147,
+      "learning_rate": 6.813808612387493e-06,
+      "loss": 0.0493,
+      "step": 3995
+    },
+    {
+      "epoch": 1.7748167888074617,
+      "grad_norm": 0.44577841042032773,
+      "learning_rate": 6.812001988742356e-06,
+      "loss": 0.0423,
+      "step": 3996
+    },
+    {
+      "epoch": 1.7752609371530093,
+      "grad_norm": 0.4208112565964139,
+      "learning_rate": 6.81019509272254e-06,
+      "loss": 0.0383,
+      "step": 3997
+    },
+    {
+      "epoch": 1.7757050854985565,
+      "grad_norm": 0.5652594178574358,
+      "learning_rate": 6.808387924599653e-06,
+      "loss": 0.0469,
+      "step": 3998
+    },
+    {
+      "epoch": 1.7761492338441038,
+      "grad_norm": 0.47567339447317186,
+      "learning_rate": 6.806580484645342e-06,
+      "loss": 0.0536,
+      "step": 3999
+    },
+    {
+      "epoch": 1.7765933821896513,
+      "grad_norm": 0.5061651547732329,
+      "learning_rate": 6.804772773131294e-06,
+      "loss": 0.0477,
+      "step": 4000
+    },
+    {
+      "epoch": 1.7770375305351989,
+      "grad_norm": 0.3302628327205031,
+      "learning_rate": 6.80296479032924e-06,
+      "loss": 0.0287,
+      "step": 4001
+    },
+    {
+      "epoch": 1.7774816788807462,
+      "grad_norm": 0.5657117208796698,
+      "learning_rate": 6.801156536510953e-06,
+      "loss": 0.063,
+      "step": 4002
+    },
+    {
+      "epoch": 1.7779258272262934,
+      "grad_norm": 0.40290409532344185,
+      "learning_rate": 6.799348011948242e-06,
+      "loss": 0.0328,
+      "step": 4003
+    },
+    {
+      "epoch": 1.778369975571841,
+      "grad_norm": 0.4303302869190356,
+      "learning_rate": 6.797539216912958e-06,
+      "loss": 0.0389,
+      "step": 4004
+    },
+    {
+      "epoch": 1.7788141239173885,
+      "grad_norm": 0.4915190826163001,
+      "learning_rate": 6.795730151676996e-06,
+      "loss": 0.0506,
+      "step": 4005
+    },
+    {
+      "epoch": 1.7792582722629358,
+      "grad_norm": 0.4472370244553756,
+      "learning_rate": 6.793920816512287e-06,
+      "loss": 0.0479,
+      "step": 4006
+    },
+    {
+      "epoch": 1.7797024206084833,
+      "grad_norm": 0.46823865267198,
+      "learning_rate": 6.792111211690807e-06,
+      "loss": 0.04,
+      "step": 4007
+    },
+    {
+      "epoch": 1.7801465689540308,
+      "grad_norm": 0.713302627768962,
+      "learning_rate": 6.790301337484569e-06,
+      "loss": 0.0467,
+      "step": 4008
+    },
+    {
+      "epoch": 1.780590717299578,
+      "grad_norm": 0.544988693204241,
+      "learning_rate": 6.788491194165629e-06,
+      "loss": 0.0597,
+      "step": 4009
+    },
+    {
+      "epoch": 1.7810348656451254,
+      "grad_norm": 0.5488358962167847,
+      "learning_rate": 6.786680782006079e-06,
+      "loss": 0.0454,
+      "step": 4010
+    },
+    {
+      "epoch": 1.7814790139906729,
+      "grad_norm": 0.45588667933120275,
+      "learning_rate": 6.784870101278058e-06,
+      "loss": 0.0458,
+      "step": 4011
+    },
+    {
+      "epoch": 1.7819231623362204,
+      "grad_norm": 0.41134367023706303,
+      "learning_rate": 6.783059152253743e-06,
+      "loss": 0.0423,
+      "step": 4012
+    },
+    {
+      "epoch": 1.7823673106817677,
+      "grad_norm": 0.5379695606986187,
+      "learning_rate": 6.7812479352053465e-06,
+      "loss": 0.0434,
+      "step": 4013
+    },
+    {
+      "epoch": 1.782811459027315,
+      "grad_norm": 0.390627468228112,
+      "learning_rate": 6.779436450405127e-06,
+      "loss": 0.0459,
+      "step": 4014
+    },
+    {
+      "epoch": 1.7832556073728625,
+      "grad_norm": 0.7523391862616232,
+      "learning_rate": 6.7776246981253835e-06,
+      "loss": 0.0578,
+      "step": 4015
+    },
+    {
+      "epoch": 1.78369975571841,
+      "grad_norm": 0.4670417748350829,
+      "learning_rate": 6.775812678638449e-06,
+      "loss": 0.0391,
+      "step": 4016
+    },
+    {
+      "epoch": 1.7841439040639573,
+      "grad_norm": 0.6629585111671369,
+      "learning_rate": 6.7740003922167045e-06,
+      "loss": 0.0525,
+      "step": 4017
+    },
+    {
+      "epoch": 1.7845880524095048,
+      "grad_norm": 0.5468251468649823,
+      "learning_rate": 6.7721878391325655e-06,
+      "loss": 0.048,
+      "step": 4018
+    },
+    {
+      "epoch": 1.7850322007550523,
+      "grad_norm": 0.6407194911078166,
+      "learning_rate": 6.770375019658491e-06,
+      "loss": 0.0342,
+      "step": 4019
+    },
+    {
+      "epoch": 1.7854763491005996,
+      "grad_norm": 0.3555576600475007,
+      "learning_rate": 6.7685619340669775e-06,
+      "loss": 0.0424,
+      "step": 4020
+    },
+    {
+      "epoch": 1.785920497446147,
+      "grad_norm": 0.5655823535701915,
+      "learning_rate": 6.766748582630561e-06,
+      "loss": 0.0498,
+      "step": 4021
+    },
+    {
+      "epoch": 1.7863646457916944,
+      "grad_norm": 0.5446372824826297,
+      "learning_rate": 6.764934965621823e-06,
+      "loss": 0.0326,
+      "step": 4022
+    },
+    {
+      "epoch": 1.786808794137242,
+      "grad_norm": 0.7635864213363933,
+      "learning_rate": 6.763121083313378e-06,
+      "loss": 0.0469,
+      "step": 4023
+    },
+    {
+      "epoch": 1.7872529424827892,
+      "grad_norm": 0.5831230436540045,
+      "learning_rate": 6.761306935977883e-06,
+      "loss": 0.0437,
+      "step": 4024
+    },
+    {
+      "epoch": 1.7876970908283365,
+      "grad_norm": 0.4755466748257561,
+      "learning_rate": 6.759492523888036e-06,
+      "loss": 0.0349,
+      "step": 4025
+    },
+    {
+      "epoch": 1.7881412391738842,
+      "grad_norm": 0.3810710893901648,
+      "learning_rate": 6.757677847316576e-06,
+      "loss": 0.0304,
+      "step": 4026
+    },
+    {
+      "epoch": 1.7885853875194315,
+      "grad_norm": 0.5635909990676548,
+      "learning_rate": 6.755862906536276e-06,
+      "loss": 0.0509,
+      "step": 4027
+    },
+    {
+      "epoch": 1.7890295358649788,
+      "grad_norm": 0.44071811791526233,
+      "learning_rate": 6.754047701819954e-06,
+      "loss": 0.0409,
+      "step": 4028
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.6047548580207962,
+      "learning_rate": 6.752232233440469e-06,
+      "loss": 0.0472,
+      "step": 4029
+    },
+    {
+      "epoch": 1.7899178325560738,
+      "grad_norm": 0.4265056649643106,
+      "learning_rate": 6.750416501670712e-06,
+      "loss": 0.0341,
+      "step": 4030
+    },
+    {
+      "epoch": 1.7903619809016211,
+      "grad_norm": 0.4750323545017985,
+      "learning_rate": 6.74860050678362e-06,
+      "loss": 0.0399,
+      "step": 4031
+    },
+    {
+      "epoch": 1.7908061292471684,
+      "grad_norm": 0.43275311316068676,
+      "learning_rate": 6.74678424905217e-06,
+      "loss": 0.0381,
+      "step": 4032
+    },
+    {
+      "epoch": 1.791250277592716,
+      "grad_norm": 0.4591066133415168,
+      "learning_rate": 6.744967728749374e-06,
+      "loss": 0.049,
+      "step": 4033
+    },
+    {
+      "epoch": 1.7916944259382634,
+      "grad_norm": 0.46035832875677174,
+      "learning_rate": 6.743150946148286e-06,
+      "loss": 0.037,
+      "step": 4034
+    },
+    {
+      "epoch": 1.7921385742838107,
+      "grad_norm": 0.44089681593883423,
+      "learning_rate": 6.7413339015219995e-06,
+      "loss": 0.0489,
+      "step": 4035
+    },
+    {
+      "epoch": 1.7925827226293582,
+      "grad_norm": 0.3967713328470784,
+      "learning_rate": 6.739516595143649e-06,
+      "loss": 0.0348,
+      "step": 4036
+    },
+    {
+      "epoch": 1.7930268709749058,
+      "grad_norm": 0.47307408803909806,
+      "learning_rate": 6.737699027286404e-06,
+      "loss": 0.0488,
+      "step": 4037
+    },
+    {
+      "epoch": 1.793471019320453,
+      "grad_norm": 0.6980297064306341,
+      "learning_rate": 6.735881198223476e-06,
+      "loss": 0.0603,
+      "step": 4038
+    },
+    {
+      "epoch": 1.7939151676660003,
+      "grad_norm": 0.641654879554228,
+      "learning_rate": 6.734063108228118e-06,
+      "loss": 0.046,
+      "step": 4039
+    },
+    {
+      "epoch": 1.7943593160115479,
+      "grad_norm": 0.43889131829448813,
+      "learning_rate": 6.732244757573619e-06,
+      "loss": 0.0417,
+      "step": 4040
+    },
+    {
+      "epoch": 1.7948034643570954,
+      "grad_norm": 0.5791145146187624,
+      "learning_rate": 6.730426146533304e-06,
+      "loss": 0.0629,
+      "step": 4041
+    },
+    {
+      "epoch": 1.7952476127026427,
+      "grad_norm": 0.3191065387044477,
+      "learning_rate": 6.728607275380548e-06,
+      "loss": 0.034,
+      "step": 4042
+    },
+    {
+      "epoch": 1.79569176104819,
+      "grad_norm": 0.5785674908181686,
+      "learning_rate": 6.726788144388754e-06,
+      "loss": 0.0495,
+      "step": 4043
+    },
+    {
+      "epoch": 1.7961359093937375,
+      "grad_norm": 0.5420845224176172,
+      "learning_rate": 6.724968753831367e-06,
+      "loss": 0.0568,
+      "step": 4044
+    },
+    {
+      "epoch": 1.796580057739285,
+      "grad_norm": 0.4088444696114397,
+      "learning_rate": 6.723149103981874e-06,
+      "loss": 0.045,
+      "step": 4045
+    },
+    {
+      "epoch": 1.7970242060848323,
+      "grad_norm": 0.6392237685246868,
+      "learning_rate": 6.721329195113802e-06,
+      "loss": 0.0554,
+      "step": 4046
+    },
+    {
+      "epoch": 1.7974683544303798,
+      "grad_norm": 0.3977912647423139,
+      "learning_rate": 6.7195090275007104e-06,
+      "loss": 0.0361,
+      "step": 4047
+    },
+    {
+      "epoch": 1.7979125027759273,
+      "grad_norm": 0.36038011427697747,
+      "learning_rate": 6.717688601416201e-06,
+      "loss": 0.0406,
+      "step": 4048
+    },
+    {
+      "epoch": 1.7983566511214746,
+      "grad_norm": 0.5212415789499063,
+      "learning_rate": 6.715867917133919e-06,
+      "loss": 0.0422,
+      "step": 4049
+    },
+    {
+      "epoch": 1.7988007994670219,
+      "grad_norm": 0.49377763847355416,
+      "learning_rate": 6.714046974927539e-06,
+      "loss": 0.043,
+      "step": 4050
+    },
+    {
+      "epoch": 1.7992449478125694,
+      "grad_norm": 0.47840704302357145,
+      "learning_rate": 6.712225775070784e-06,
+      "loss": 0.044,
+      "step": 4051
+    },
+    {
+      "epoch": 1.799689096158117,
+      "grad_norm": 0.4112067824225565,
+      "learning_rate": 6.71040431783741e-06,
+      "loss": 0.0354,
+      "step": 4052
+    },
+    {
+      "epoch": 1.8001332445036642,
+      "grad_norm": 0.3668140961966663,
+      "learning_rate": 6.70858260350121e-06,
+      "loss": 0.0369,
+      "step": 4053
+    },
+    {
+      "epoch": 1.8005773928492115,
+      "grad_norm": 0.6061145451464093,
+      "learning_rate": 6.706760632336023e-06,
+      "loss": 0.0529,
+      "step": 4054
+    },
+    {
+      "epoch": 1.8010215411947592,
+      "grad_norm": 0.804408929113384,
+      "learning_rate": 6.704938404615718e-06,
+      "loss": 0.0695,
+      "step": 4055
+    },
+    {
+      "epoch": 1.8014656895403065,
+      "grad_norm": 0.44516753707686785,
+      "learning_rate": 6.703115920614212e-06,
+      "loss": 0.041,
+      "step": 4056
+    },
+    {
+      "epoch": 1.8019098378858538,
+      "grad_norm": 0.6944007446846528,
+      "learning_rate": 6.701293180605451e-06,
+      "loss": 0.063,
+      "step": 4057
+    },
+    {
+      "epoch": 1.8023539862314013,
+      "grad_norm": 0.6199411354460107,
+      "learning_rate": 6.699470184863423e-06,
+      "loss": 0.0478,
+      "step": 4058
+    },
+    {
+      "epoch": 1.8027981345769488,
+      "grad_norm": 0.6947222002075307,
+      "learning_rate": 6.6976469336621595e-06,
+      "loss": 0.0425,
+      "step": 4059
+    },
+    {
+      "epoch": 1.803242282922496,
+      "grad_norm": 0.5292009929059116,
+      "learning_rate": 6.6958234272757235e-06,
+      "loss": 0.0575,
+      "step": 4060
+    },
+    {
+      "epoch": 1.8036864312680434,
+      "grad_norm": 0.438201912001024,
+      "learning_rate": 6.6939996659782194e-06,
+      "loss": 0.0395,
+      "step": 4061
+    },
+    {
+      "epoch": 1.804130579613591,
+      "grad_norm": 0.5407542517674611,
+      "learning_rate": 6.692175650043789e-06,
+      "loss": 0.0403,
+      "step": 4062
+    },
+    {
+      "epoch": 1.8045747279591384,
+      "grad_norm": 0.5631536015470302,
+      "learning_rate": 6.690351379746613e-06,
+      "loss": 0.0421,
+      "step": 4063
+    },
+    {
+      "epoch": 1.8050188763046857,
+      "grad_norm": 0.5913174820930391,
+      "learning_rate": 6.6885268553609115e-06,
+      "loss": 0.0265,
+      "step": 4064
+    },
+    {
+      "epoch": 1.8054630246502332,
+      "grad_norm": 0.576629329530581,
+      "learning_rate": 6.68670207716094e-06,
+      "loss": 0.0393,
+      "step": 4065
+    },
+    {
+      "epoch": 1.8059071729957807,
+      "grad_norm": 0.47244671586292697,
+      "learning_rate": 6.6848770454209955e-06,
+      "loss": 0.043,
+      "step": 4066
+    },
+    {
+      "epoch": 1.806351321341328,
+      "grad_norm": 0.42376459190231364,
+      "learning_rate": 6.683051760415409e-06,
+      "loss": 0.0349,
+      "step": 4067
+    },
+    {
+      "epoch": 1.8067954696868753,
+      "grad_norm": 0.5089448244604045,
+      "learning_rate": 6.681226222418553e-06,
+      "loss": 0.0367,
+      "step": 4068
+    },
+    {
+      "epoch": 1.8072396180324228,
+      "grad_norm": 0.7287865087611148,
+      "learning_rate": 6.679400431704837e-06,
+      "loss": 0.0474,
+      "step": 4069
+    },
+    {
+      "epoch": 1.8076837663779703,
+      "grad_norm": 0.4089819648317366,
+      "learning_rate": 6.677574388548706e-06,
+      "loss": 0.0273,
+      "step": 4070
+    },
+    {
+      "epoch": 1.8081279147235176,
+      "grad_norm": 0.3655870554565193,
+      "learning_rate": 6.67574809322465e-06,
+      "loss": 0.0366,
+      "step": 4071
+    },
+    {
+      "epoch": 1.808572063069065,
+      "grad_norm": 0.4524590952965907,
+      "learning_rate": 6.6739215460071885e-06,
+      "loss": 0.0388,
+      "step": 4072
+    },
+    {
+      "epoch": 1.8090162114146124,
+      "grad_norm": 0.8966024978124474,
+      "learning_rate": 6.672094747170883e-06,
+      "loss": 0.0566,
+      "step": 4073
+    },
+    {
+      "epoch": 1.80946035976016,
+      "grad_norm": 0.3591914341706326,
+      "learning_rate": 6.670267696990335e-06,
+      "loss": 0.0318,
+      "step": 4074
+    },
+    {
+      "epoch": 1.8099045081057072,
+      "grad_norm": 0.44339585259060615,
+      "learning_rate": 6.668440395740178e-06,
+      "loss": 0.0363,
+      "step": 4075
+    },
+    {
+      "epoch": 1.8103486564512548,
+      "grad_norm": 0.3901591376438945,
+      "learning_rate": 6.666612843695087e-06,
+      "loss": 0.0485,
+      "step": 4076
+    },
+    {
+      "epoch": 1.8107928047968023,
+      "grad_norm": 0.4015830587763988,
+      "learning_rate": 6.664785041129777e-06,
+      "loss": 0.0383,
+      "step": 4077
+    },
+    {
+      "epoch": 1.8112369531423496,
+      "grad_norm": 0.5367818518459079,
+      "learning_rate": 6.662956988318994e-06,
+      "loss": 0.0456,
+      "step": 4078
+    },
+    {
+      "epoch": 1.8116811014878968,
+      "grad_norm": 0.40853437415084287,
+      "learning_rate": 6.661128685537526e-06,
+      "loss": 0.0322,
+      "step": 4079
+    },
+    {
+      "epoch": 1.8121252498334444,
+      "grad_norm": 0.6527784454525124,
+      "learning_rate": 6.659300133060201e-06,
+      "loss": 0.0507,
+      "step": 4080
+    },
+    {
+      "epoch": 1.8125693981789919,
+      "grad_norm": 0.5074533806315775,
+      "learning_rate": 6.657471331161878e-06,
+      "loss": 0.0432,
+      "step": 4081
+    },
+    {
+      "epoch": 1.8130135465245392,
+      "grad_norm": 0.4472089329933828,
+      "learning_rate": 6.65564228011746e-06,
+      "loss": 0.0599,
+      "step": 4082
+    },
+    {
+      "epoch": 1.8134576948700865,
+      "grad_norm": 0.47876435411805834,
+      "learning_rate": 6.653812980201882e-06,
+      "loss": 0.0478,
+      "step": 4083
+    },
+    {
+      "epoch": 1.813901843215634,
+      "grad_norm": 0.44056500378250846,
+      "learning_rate": 6.651983431690119e-06,
+      "loss": 0.0434,
+      "step": 4084
+    },
+    {
+      "epoch": 1.8143459915611815,
+      "grad_norm": 0.42532030681624505,
+      "learning_rate": 6.650153634857183e-06,
+      "loss": 0.0372,
+      "step": 4085
+    },
+    {
+      "epoch": 1.8147901399067288,
+      "grad_norm": 0.38789115474665614,
+      "learning_rate": 6.648323589978128e-06,
+      "loss": 0.0332,
+      "step": 4086
+    },
+    {
+      "epoch": 1.8152342882522763,
+      "grad_norm": 0.5287105455383381,
+      "learning_rate": 6.646493297328034e-06,
+      "loss": 0.0483,
+      "step": 4087
+    },
+    {
+      "epoch": 1.8156784365978238,
+      "grad_norm": 2.7184134061804475,
+      "learning_rate": 6.6446627571820295e-06,
+      "loss": 0.05,
+      "step": 4088
+    },
+    {
+      "epoch": 1.816122584943371,
+      "grad_norm": 0.48923408476238167,
+      "learning_rate": 6.642831969815275e-06,
+      "loss": 0.0524,
+      "step": 4089
+    },
+    {
+      "epoch": 1.8165667332889184,
+      "grad_norm": 0.4086306503799156,
+      "learning_rate": 6.641000935502968e-06,
+      "loss": 0.0369,
+      "step": 4090
+    },
+    {
+      "epoch": 1.817010881634466,
+      "grad_norm": 0.42279116752419027,
+      "learning_rate": 6.639169654520345e-06,
+      "loss": 0.0384,
+      "step": 4091
+    },
+    {
+      "epoch": 1.8174550299800134,
+      "grad_norm": 0.47809993483091023,
+      "learning_rate": 6.637338127142678e-06,
+      "loss": 0.0557,
+      "step": 4092
+    },
+    {
+      "epoch": 1.8178991783255607,
+      "grad_norm": 0.43894142864567726,
+      "learning_rate": 6.635506353645277e-06,
+      "loss": 0.0442,
+      "step": 4093
+    },
+    {
+      "epoch": 1.818343326671108,
+      "grad_norm": 0.4258410799351809,
+      "learning_rate": 6.633674334303489e-06,
+      "loss": 0.0453,
+      "step": 4094
+    },
+    {
+      "epoch": 1.8187874750166557,
+      "grad_norm": 0.4223073933724936,
+      "learning_rate": 6.631842069392698e-06,
+      "loss": 0.0425,
+      "step": 4095
+    },
+    {
+      "epoch": 1.819231623362203,
+      "grad_norm": 0.379351123137964,
+      "learning_rate": 6.630009559188323e-06,
+      "loss": 0.0333,
+      "step": 4096
+    },
+    {
+      "epoch": 1.8196757717077503,
+      "grad_norm": 0.38768367796199465,
+      "learning_rate": 6.628176803965823e-06,
+      "loss": 0.0391,
+      "step": 4097
+    },
+    {
+      "epoch": 1.8201199200532978,
+      "grad_norm": 0.34224780916225656,
+      "learning_rate": 6.62634380400069e-06,
+      "loss": 0.0341,
+      "step": 4098
+    },
+    {
+      "epoch": 1.8205640683988453,
+      "grad_norm": 0.42049665765180966,
+      "learning_rate": 6.624510559568458e-06,
+      "loss": 0.0332,
+      "step": 4099
+    },
+    {
+      "epoch": 1.8210082167443926,
+      "grad_norm": 0.4650798805753388,
+      "learning_rate": 6.622677070944692e-06,
+      "loss": 0.0522,
+      "step": 4100
+    },
+    {
+      "epoch": 1.82145236508994,
+      "grad_norm": 0.7403412237132078,
+      "learning_rate": 6.6208433384049974e-06,
+      "loss": 0.0468,
+      "step": 4101
+    },
+    {
+      "epoch": 1.8218965134354874,
+      "grad_norm": 0.5003084175625752,
+      "learning_rate": 6.619009362225017e-06,
+      "loss": 0.0369,
+      "step": 4102
+    },
+    {
+      "epoch": 1.822340661781035,
+      "grad_norm": 0.4709276604503224,
+      "learning_rate": 6.617175142680426e-06,
+      "loss": 0.0373,
+      "step": 4103
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.405634995248283,
+      "learning_rate": 6.615340680046941e-06,
+      "loss": 0.0402,
+      "step": 4104
+    },
+    {
+      "epoch": 1.8232289584721297,
+      "grad_norm": 0.48277401824042093,
+      "learning_rate": 6.613505974600313e-06,
+      "loss": 0.0427,
+      "step": 4105
+    },
+    {
+      "epoch": 1.8236731068176772,
+      "grad_norm": 0.5146304089304894,
+      "learning_rate": 6.611671026616328e-06,
+      "loss": 0.0449,
+      "step": 4106
+    },
+    {
+      "epoch": 1.8241172551632245,
+      "grad_norm": 0.5014137439041658,
+      "learning_rate": 6.609835836370808e-06,
+      "loss": 0.0462,
+      "step": 4107
+    },
+    {
+      "epoch": 1.8245614035087718,
+      "grad_norm": 0.5653319356301473,
+      "learning_rate": 6.6080004041396176e-06,
+      "loss": 0.0385,
+      "step": 4108
+    },
+    {
+      "epoch": 1.8250055518543193,
+      "grad_norm": 0.43607877589233934,
+      "learning_rate": 6.60616473019865e-06,
+      "loss": 0.0366,
+      "step": 4109
+    },
+    {
+      "epoch": 1.8254497001998669,
+      "grad_norm": 0.34512361350227927,
+      "learning_rate": 6.6043288148238405e-06,
+      "loss": 0.0327,
+      "step": 4110
+    },
+    {
+      "epoch": 1.8258938485454141,
+      "grad_norm": 0.46278297105670835,
+      "learning_rate": 6.6024926582911576e-06,
+      "loss": 0.0412,
+      "step": 4111
+    },
+    {
+      "epoch": 1.8263379968909614,
+      "grad_norm": 0.4906820801194115,
+      "learning_rate": 6.600656260876605e-06,
+      "loss": 0.0381,
+      "step": 4112
+    },
+    {
+      "epoch": 1.826782145236509,
+      "grad_norm": 0.48071884316968305,
+      "learning_rate": 6.598819622856227e-06,
+      "loss": 0.0409,
+      "step": 4113
+    },
+    {
+      "epoch": 1.8272262935820565,
+      "grad_norm": 0.47730730807604993,
+      "learning_rate": 6.596982744506101e-06,
+      "loss": 0.0557,
+      "step": 4114
+    },
+    {
+      "epoch": 1.8276704419276038,
+      "grad_norm": 0.3790889364580573,
+      "learning_rate": 6.595145626102339e-06,
+      "loss": 0.0404,
+      "step": 4115
+    },
+    {
+      "epoch": 1.8281145902731513,
+      "grad_norm": 0.463399073158891,
+      "learning_rate": 6.593308267921095e-06,
+      "loss": 0.0411,
+      "step": 4116
+    },
+    {
+      "epoch": 1.8285587386186988,
+      "grad_norm": 0.501176246654533,
+      "learning_rate": 6.59147067023855e-06,
+      "loss": 0.0506,
+      "step": 4117
+    },
+    {
+      "epoch": 1.829002886964246,
+      "grad_norm": 0.46578549229661426,
+      "learning_rate": 6.58963283333093e-06,
+      "loss": 0.0454,
+      "step": 4118
+    },
+    {
+      "epoch": 1.8294470353097934,
+      "grad_norm": 0.4499464900379103,
+      "learning_rate": 6.587794757474493e-06,
+      "loss": 0.0328,
+      "step": 4119
+    },
+    {
+      "epoch": 1.8298911836553409,
+      "grad_norm": 0.4379709545832753,
+      "learning_rate": 6.585956442945531e-06,
+      "loss": 0.0342,
+      "step": 4120
+    },
+    {
+      "epoch": 1.8303353320008884,
+      "grad_norm": 0.5260425346926577,
+      "learning_rate": 6.584117890020374e-06,
+      "loss": 0.0439,
+      "step": 4121
+    },
+    {
+      "epoch": 1.8307794803464357,
+      "grad_norm": 0.426360295901579,
+      "learning_rate": 6.5822790989753905e-06,
+      "loss": 0.0411,
+      "step": 4122
+    },
+    {
+      "epoch": 1.831223628691983,
+      "grad_norm": 0.45007999975919516,
+      "learning_rate": 6.5804400700869806e-06,
+      "loss": 0.0443,
+      "step": 4123
+    },
+    {
+      "epoch": 1.8316677770375307,
+      "grad_norm": 0.59501919115864,
+      "learning_rate": 6.578600803631579e-06,
+      "loss": 0.0454,
+      "step": 4124
+    },
+    {
+      "epoch": 1.832111925383078,
+      "grad_norm": 0.3671794598732458,
+      "learning_rate": 6.5767612998856625e-06,
+      "loss": 0.0393,
+      "step": 4125
+    },
+    {
+      "epoch": 1.8325560737286253,
+      "grad_norm": 0.36069972183791427,
+      "learning_rate": 6.574921559125737e-06,
+      "loss": 0.0375,
+      "step": 4126
+    },
+    {
+      "epoch": 1.8330002220741728,
+      "grad_norm": 0.47931723820550676,
+      "learning_rate": 6.573081581628349e-06,
+      "loss": 0.0398,
+      "step": 4127
+    },
+    {
+      "epoch": 1.8334443704197203,
+      "grad_norm": 0.4393540376137144,
+      "learning_rate": 6.571241367670077e-06,
+      "loss": 0.048,
+      "step": 4128
+    },
+    {
+      "epoch": 1.8338885187652676,
+      "grad_norm": 0.5502773042166398,
+      "learning_rate": 6.569400917527536e-06,
+      "loss": 0.0391,
+      "step": 4129
+    },
+    {
+      "epoch": 1.8343326671108149,
+      "grad_norm": 0.42922588665741507,
+      "learning_rate": 6.567560231477379e-06,
+      "loss": 0.0371,
+      "step": 4130
+    },
+    {
+      "epoch": 1.8347768154563624,
+      "grad_norm": 0.4243115655078441,
+      "learning_rate": 6.56571930979629e-06,
+      "loss": 0.0464,
+      "step": 4131
+    },
+    {
+      "epoch": 1.83522096380191,
+      "grad_norm": 0.5637304292180152,
+      "learning_rate": 6.563878152760992e-06,
+      "loss": 0.0587,
+      "step": 4132
+    },
+    {
+      "epoch": 1.8356651121474572,
+      "grad_norm": 0.5080427318268489,
+      "learning_rate": 6.562036760648242e-06,
+      "loss": 0.0445,
+      "step": 4133
+    },
+    {
+      "epoch": 1.8361092604930047,
+      "grad_norm": 0.5069784829710062,
+      "learning_rate": 6.560195133734833e-06,
+      "loss": 0.0329,
+      "step": 4134
+    },
+    {
+      "epoch": 1.8365534088385522,
+      "grad_norm": 0.3415574151612511,
+      "learning_rate": 6.55835327229759e-06,
+      "loss": 0.0292,
+      "step": 4135
+    },
+    {
+      "epoch": 1.8369975571840995,
+      "grad_norm": 0.4077726290203496,
+      "learning_rate": 6.556511176613381e-06,
+      "loss": 0.0347,
+      "step": 4136
+    },
+    {
+      "epoch": 1.8374417055296468,
+      "grad_norm": 0.5240761531044744,
+      "learning_rate": 6.554668846959102e-06,
+      "loss": 0.0454,
+      "step": 4137
+    },
+    {
+      "epoch": 1.8378858538751943,
+      "grad_norm": 0.5670009106024317,
+      "learning_rate": 6.552826283611684e-06,
+      "loss": 0.0488,
+      "step": 4138
+    },
+    {
+      "epoch": 1.8383300022207418,
+      "grad_norm": 0.5071828765865913,
+      "learning_rate": 6.5509834868480994e-06,
+      "loss": 0.0464,
+      "step": 4139
+    },
+    {
+      "epoch": 1.8387741505662891,
+      "grad_norm": 0.6942163546904666,
+      "learning_rate": 6.54914045694535e-06,
+      "loss": 0.0558,
+      "step": 4140
+    },
+    {
+      "epoch": 1.8392182989118364,
+      "grad_norm": 0.42462008853388156,
+      "learning_rate": 6.547297194180473e-06,
+      "loss": 0.028,
+      "step": 4141
+    },
+    {
+      "epoch": 1.839662447257384,
+      "grad_norm": 0.3935408254645297,
+      "learning_rate": 6.545453698830545e-06,
+      "loss": 0.028,
+      "step": 4142
+    },
+    {
+      "epoch": 1.8401065956029314,
+      "grad_norm": 0.43222089529601715,
+      "learning_rate": 6.543609971172673e-06,
+      "loss": 0.0439,
+      "step": 4143
+    },
+    {
+      "epoch": 1.8405507439484787,
+      "grad_norm": 0.8970440945065747,
+      "learning_rate": 6.541766011484001e-06,
+      "loss": 0.0569,
+      "step": 4144
+    },
+    {
+      "epoch": 1.8409948922940262,
+      "grad_norm": 0.43018474254953293,
+      "learning_rate": 6.539921820041708e-06,
+      "loss": 0.0429,
+      "step": 4145
+    },
+    {
+      "epoch": 1.8414390406395738,
+      "grad_norm": 0.5272129102858879,
+      "learning_rate": 6.538077397123006e-06,
+      "loss": 0.042,
+      "step": 4146
+    },
+    {
+      "epoch": 1.841883188985121,
+      "grad_norm": 0.5767831742144044,
+      "learning_rate": 6.536232743005144e-06,
+      "loss": 0.0505,
+      "step": 4147
+    },
+    {
+      "epoch": 1.8423273373306683,
+      "grad_norm": 0.5455867077197153,
+      "learning_rate": 6.534387857965405e-06,
+      "loss": 0.0499,
+      "step": 4148
+    },
+    {
+      "epoch": 1.8427714856762158,
+      "grad_norm": 0.4392562319330063,
+      "learning_rate": 6.532542742281105e-06,
+      "loss": 0.0404,
+      "step": 4149
+    },
+    {
+      "epoch": 1.8432156340217634,
+      "grad_norm": 0.41150977239123576,
+      "learning_rate": 6.5306973962296e-06,
+      "loss": 0.0388,
+      "step": 4150
+    },
+    {
+      "epoch": 1.8436597823673107,
+      "grad_norm": 1.313911582342841,
+      "learning_rate": 6.528851820088273e-06,
+      "loss": 0.0595,
+      "step": 4151
+    },
+    {
+      "epoch": 1.844103930712858,
+      "grad_norm": 0.3322454559510484,
+      "learning_rate": 6.527006014134546e-06,
+      "loss": 0.0274,
+      "step": 4152
+    },
+    {
+      "epoch": 1.8445480790584055,
+      "grad_norm": 0.44783811513765753,
+      "learning_rate": 6.525159978645876e-06,
+      "loss": 0.0391,
+      "step": 4153
+    },
+    {
+      "epoch": 1.844992227403953,
+      "grad_norm": 0.5214447787650061,
+      "learning_rate": 6.523313713899755e-06,
+      "loss": 0.0396,
+      "step": 4154
+    },
+    {
+      "epoch": 1.8454363757495003,
+      "grad_norm": 0.45172874563459914,
+      "learning_rate": 6.521467220173705e-06,
+      "loss": 0.0356,
+      "step": 4155
+    },
+    {
+      "epoch": 1.8458805240950478,
+      "grad_norm": 0.3652608177780825,
+      "learning_rate": 6.519620497745286e-06,
+      "loss": 0.0413,
+      "step": 4156
+    },
+    {
+      "epoch": 1.8463246724405953,
+      "grad_norm": 0.435616251575995,
+      "learning_rate": 6.5177735468920935e-06,
+      "loss": 0.0444,
+      "step": 4157
+    },
+    {
+      "epoch": 1.8467688207861426,
+      "grad_norm": 0.7590654646546372,
+      "learning_rate": 6.515926367891754e-06,
+      "loss": 0.049,
+      "step": 4158
+    },
+    {
+      "epoch": 1.8472129691316899,
+      "grad_norm": 0.45565692087369364,
+      "learning_rate": 6.51407896102193e-06,
+      "loss": 0.0377,
+      "step": 4159
+    },
+    {
+      "epoch": 1.8476571174772374,
+      "grad_norm": 1.8021394181856782,
+      "learning_rate": 6.512231326560319e-06,
+      "loss": 0.0632,
+      "step": 4160
+    },
+    {
+      "epoch": 1.8481012658227849,
+      "grad_norm": 0.37814866423712995,
+      "learning_rate": 6.510383464784651e-06,
+      "loss": 0.0357,
+      "step": 4161
+    },
+    {
+      "epoch": 1.8485454141683322,
+      "grad_norm": 0.5013395214860348,
+      "learning_rate": 6.508535375972691e-06,
+      "loss": 0.0359,
+      "step": 4162
+    },
+    {
+      "epoch": 1.8489895625138795,
+      "grad_norm": 0.6460169691477774,
+      "learning_rate": 6.506687060402238e-06,
+      "loss": 0.055,
+      "step": 4163
+    },
+    {
+      "epoch": 1.8494337108594272,
+      "grad_norm": 0.731410974029027,
+      "learning_rate": 6.504838518351127e-06,
+      "loss": 0.0418,
+      "step": 4164
+    },
+    {
+      "epoch": 1.8498778592049745,
+      "grad_norm": 0.4318554461256705,
+      "learning_rate": 6.502989750097224e-06,
+      "loss": 0.0423,
+      "step": 4165
+    },
+    {
+      "epoch": 1.8503220075505218,
+      "grad_norm": 0.4065429352215067,
+      "learning_rate": 6.501140755918428e-06,
+      "loss": 0.0325,
+      "step": 4166
+    },
+    {
+      "epoch": 1.8507661558960693,
+      "grad_norm": 0.5268522126273875,
+      "learning_rate": 6.499291536092679e-06,
+      "loss": 0.052,
+      "step": 4167
+    },
+    {
+      "epoch": 1.8512103042416168,
+      "grad_norm": 0.4963236453076724,
+      "learning_rate": 6.497442090897943e-06,
+      "loss": 0.0507,
+      "step": 4168
+    },
+    {
+      "epoch": 1.851654452587164,
+      "grad_norm": 0.5716468049405105,
+      "learning_rate": 6.495592420612224e-06,
+      "loss": 0.0557,
+      "step": 4169
+    },
+    {
+      "epoch": 1.8520986009327114,
+      "grad_norm": 0.36399318359325455,
+      "learning_rate": 6.493742525513556e-06,
+      "loss": 0.0318,
+      "step": 4170
+    },
+    {
+      "epoch": 1.852542749278259,
+      "grad_norm": 0.4368593597512668,
+      "learning_rate": 6.491892405880015e-06,
+      "loss": 0.036,
+      "step": 4171
+    },
+    {
+      "epoch": 1.8529868976238064,
+      "grad_norm": 0.490968844938355,
+      "learning_rate": 6.490042061989701e-06,
+      "loss": 0.0384,
+      "step": 4172
+    },
+    {
+      "epoch": 1.8534310459693537,
+      "grad_norm": 0.4895671851812533,
+      "learning_rate": 6.4881914941207545e-06,
+      "loss": 0.0526,
+      "step": 4173
+    },
+    {
+      "epoch": 1.8538751943149012,
+      "grad_norm": 0.4582350620997356,
+      "learning_rate": 6.486340702551347e-06,
+      "loss": 0.044,
+      "step": 4174
+    },
+    {
+      "epoch": 1.8543193426604487,
+      "grad_norm": 0.4289059463290211,
+      "learning_rate": 6.484489687559682e-06,
+      "loss": 0.0379,
+      "step": 4175
+    },
+    {
+      "epoch": 1.854763491005996,
+      "grad_norm": 0.3956879154243665,
+      "learning_rate": 6.4826384494240006e-06,
+      "loss": 0.0324,
+      "step": 4176
+    },
+    {
+      "epoch": 1.8552076393515433,
+      "grad_norm": 0.4900266012746407,
+      "learning_rate": 6.480786988422575e-06,
+      "loss": 0.0472,
+      "step": 4177
+    },
+    {
+      "epoch": 1.8556517876970908,
+      "grad_norm": 0.4376815416787356,
+      "learning_rate": 6.47893530483371e-06,
+      "loss": 0.0631,
+      "step": 4178
+    },
+    {
+      "epoch": 1.8560959360426383,
+      "grad_norm": 0.3452667188092541,
+      "learning_rate": 6.4770833989357464e-06,
+      "loss": 0.031,
+      "step": 4179
+    },
+    {
+      "epoch": 1.8565400843881856,
+      "grad_norm": 0.4575028834398297,
+      "learning_rate": 6.4752312710070565e-06,
+      "loss": 0.0357,
+      "step": 4180
+    },
+    {
+      "epoch": 1.856984232733733,
+      "grad_norm": 0.35802922814056004,
+      "learning_rate": 6.4733789213260465e-06,
+      "loss": 0.0363,
+      "step": 4181
+    },
+    {
+      "epoch": 1.8574283810792804,
+      "grad_norm": 0.4610516053299678,
+      "learning_rate": 6.471526350171158e-06,
+      "loss": 0.0421,
+      "step": 4182
+    },
+    {
+      "epoch": 1.857872529424828,
+      "grad_norm": 0.41994717663617964,
+      "learning_rate": 6.46967355782086e-06,
+      "loss": 0.0427,
+      "step": 4183
+    },
+    {
+      "epoch": 1.8583166777703752,
+      "grad_norm": 0.41511781298154615,
+      "learning_rate": 6.4678205445536615e-06,
+      "loss": 0.0388,
+      "step": 4184
+    },
+    {
+      "epoch": 1.8587608261159227,
+      "grad_norm": 0.4226703337206278,
+      "learning_rate": 6.465967310648103e-06,
+      "loss": 0.0363,
+      "step": 4185
+    },
+    {
+      "epoch": 1.8592049744614703,
+      "grad_norm": 0.3756285974893726,
+      "learning_rate": 6.464113856382752e-06,
+      "loss": 0.0431,
+      "step": 4186
+    },
+    {
+      "epoch": 1.8596491228070176,
+      "grad_norm": 0.7660609289210849,
+      "learning_rate": 6.46226018203622e-06,
+      "loss": 0.0548,
+      "step": 4187
+    },
+    {
+      "epoch": 1.8600932711525648,
+      "grad_norm": 0.44778181358321073,
+      "learning_rate": 6.460406287887142e-06,
+      "loss": 0.0402,
+      "step": 4188
+    },
+    {
+      "epoch": 1.8605374194981124,
+      "grad_norm": 0.4653083552128945,
+      "learning_rate": 6.4585521742141924e-06,
+      "loss": 0.0429,
+      "step": 4189
+    },
+    {
+      "epoch": 1.8609815678436599,
+      "grad_norm": 0.4092361015012088,
+      "learning_rate": 6.456697841296072e-06,
+      "loss": 0.0538,
+      "step": 4190
+    },
+    {
+      "epoch": 1.8614257161892072,
+      "grad_norm": 0.3943066511337006,
+      "learning_rate": 6.4548432894115236e-06,
+      "loss": 0.0422,
+      "step": 4191
+    },
+    {
+      "epoch": 1.8618698645347544,
+      "grad_norm": 0.574897072037143,
+      "learning_rate": 6.452988518839314e-06,
+      "loss": 0.0638,
+      "step": 4192
+    },
+    {
+      "epoch": 1.8623140128803022,
+      "grad_norm": 0.3525774963677566,
+      "learning_rate": 6.451133529858249e-06,
+      "loss": 0.0308,
+      "step": 4193
+    },
+    {
+      "epoch": 1.8627581612258495,
+      "grad_norm": 0.47268074934306253,
+      "learning_rate": 6.449278322747164e-06,
+      "loss": 0.0443,
+      "step": 4194
+    },
+    {
+      "epoch": 1.8632023095713968,
+      "grad_norm": 0.49332043887118054,
+      "learning_rate": 6.447422897784927e-06,
+      "loss": 0.0486,
+      "step": 4195
+    },
+    {
+      "epoch": 1.8636464579169443,
+      "grad_norm": 0.4599615865532288,
+      "learning_rate": 6.445567255250442e-06,
+      "loss": 0.0374,
+      "step": 4196
+    },
+    {
+      "epoch": 1.8640906062624918,
+      "grad_norm": 0.6216566278556651,
+      "learning_rate": 6.443711395422641e-06,
+      "loss": 0.0458,
+      "step": 4197
+    },
+    {
+      "epoch": 1.864534754608039,
+      "grad_norm": 0.45476603291439793,
+      "learning_rate": 6.4418553185804946e-06,
+      "loss": 0.0411,
+      "step": 4198
+    },
+    {
+      "epoch": 1.8649789029535864,
+      "grad_norm": 0.5568340998927974,
+      "learning_rate": 6.4399990250030005e-06,
+      "loss": 0.0469,
+      "step": 4199
+    },
+    {
+      "epoch": 1.8654230512991339,
+      "grad_norm": 0.5792755550634929,
+      "learning_rate": 6.438142514969192e-06,
+      "loss": 0.0486,
+      "step": 4200
+    },
+    {
+      "epoch": 1.8658671996446814,
+      "grad_norm": 0.5484197721468562,
+      "learning_rate": 6.436285788758133e-06,
+      "loss": 0.0373,
+      "step": 4201
+    },
+    {
+      "epoch": 1.8663113479902287,
+      "grad_norm": 0.408467206743211,
+      "learning_rate": 6.434428846648923e-06,
+      "loss": 0.0345,
+      "step": 4202
+    },
+    {
+      "epoch": 1.8667554963357762,
+      "grad_norm": 0.4273190617477224,
+      "learning_rate": 6.43257168892069e-06,
+      "loss": 0.0521,
+      "step": 4203
+    },
+    {
+      "epoch": 1.8671996446813237,
+      "grad_norm": 0.913822616436895,
+      "learning_rate": 6.430714315852595e-06,
+      "loss": 0.0731,
+      "step": 4204
+    },
+    {
+      "epoch": 1.867643793026871,
+      "grad_norm": 0.3654209870999856,
+      "learning_rate": 6.428856727723838e-06,
+      "loss": 0.0426,
+      "step": 4205
+    },
+    {
+      "epoch": 1.8680879413724183,
+      "grad_norm": 0.35677094964790335,
+      "learning_rate": 6.426998924813641e-06,
+      "loss": 0.0354,
+      "step": 4206
+    },
+    {
+      "epoch": 1.8685320897179658,
+      "grad_norm": 0.7442434568098797,
+      "learning_rate": 6.425140907401266e-06,
+      "loss": 0.057,
+      "step": 4207
+    },
+    {
+      "epoch": 1.8689762380635133,
+      "grad_norm": 0.5792617957834036,
+      "learning_rate": 6.423282675766002e-06,
+      "loss": 0.0409,
+      "step": 4208
+    },
+    {
+      "epoch": 1.8694203864090606,
+      "grad_norm": 0.7180220670929347,
+      "learning_rate": 6.4214242301871766e-06,
+      "loss": 0.055,
+      "step": 4209
+    },
+    {
+      "epoch": 1.869864534754608,
+      "grad_norm": 0.369340852755107,
+      "learning_rate": 6.4195655709441425e-06,
+      "loss": 0.0365,
+      "step": 4210
+    },
+    {
+      "epoch": 1.8703086831001554,
+      "grad_norm": 0.5907561592268787,
+      "learning_rate": 6.41770669831629e-06,
+      "loss": 0.05,
+      "step": 4211
+    },
+    {
+      "epoch": 1.870752831445703,
+      "grad_norm": 0.4737879055532281,
+      "learning_rate": 6.415847612583036e-06,
+      "loss": 0.0387,
+      "step": 4212
+    },
+    {
+      "epoch": 1.8711969797912502,
+      "grad_norm": 0.45358931580739376,
+      "learning_rate": 6.413988314023837e-06,
+      "loss": 0.0455,
+      "step": 4213
+    },
+    {
+      "epoch": 1.8716411281367977,
+      "grad_norm": 0.396008499582763,
+      "learning_rate": 6.412128802918174e-06,
+      "loss": 0.0428,
+      "step": 4214
+    },
+    {
+      "epoch": 1.8720852764823452,
+      "grad_norm": 0.446819083411039,
+      "learning_rate": 6.410269079545563e-06,
+      "loss": 0.0353,
+      "step": 4215
+    },
+    {
+      "epoch": 1.8725294248278925,
+      "grad_norm": 0.5360694926018593,
+      "learning_rate": 6.408409144185555e-06,
+      "loss": 0.0394,
+      "step": 4216
+    },
+    {
+      "epoch": 1.8729735731734398,
+      "grad_norm": 0.419696086205574,
+      "learning_rate": 6.406548997117728e-06,
+      "loss": 0.0442,
+      "step": 4217
+    },
+    {
+      "epoch": 1.8734177215189873,
+      "grad_norm": 0.5541434243753425,
+      "learning_rate": 6.404688638621691e-06,
+      "loss": 0.0362,
+      "step": 4218
+    },
+    {
+      "epoch": 1.8738618698645348,
+      "grad_norm": 0.5424748341749609,
+      "learning_rate": 6.402828068977092e-06,
+      "loss": 0.0385,
+      "step": 4219
+    },
+    {
+      "epoch": 1.8743060182100821,
+      "grad_norm": 0.5755082545471814,
+      "learning_rate": 6.400967288463604e-06,
+      "loss": 0.0408,
+      "step": 4220
+    },
+    {
+      "epoch": 1.8747501665556294,
+      "grad_norm": 0.5534620278515894,
+      "learning_rate": 6.399106297360934e-06,
+      "loss": 0.0394,
+      "step": 4221
+    },
+    {
+      "epoch": 1.875194314901177,
+      "grad_norm": 0.43339133128599205,
+      "learning_rate": 6.397245095948822e-06,
+      "loss": 0.0447,
+      "step": 4222
+    },
+    {
+      "epoch": 1.8756384632467245,
+      "grad_norm": 0.4242062003649521,
+      "learning_rate": 6.395383684507036e-06,
+      "loss": 0.0367,
+      "step": 4223
+    },
+    {
+      "epoch": 1.8760826115922717,
+      "grad_norm": 0.429738984394203,
+      "learning_rate": 6.393522063315379e-06,
+      "loss": 0.0434,
+      "step": 4224
+    },
+    {
+      "epoch": 1.8765267599378193,
+      "grad_norm": 0.5905398248805794,
+      "learning_rate": 6.391660232653685e-06,
+      "loss": 0.0387,
+      "step": 4225
+    },
+    {
+      "epoch": 1.8769709082833668,
+      "grad_norm": 0.46069485833459556,
+      "learning_rate": 6.389798192801816e-06,
+      "loss": 0.042,
+      "step": 4226
+    },
+    {
+      "epoch": 1.877415056628914,
+      "grad_norm": 0.3636972167781559,
+      "learning_rate": 6.387935944039672e-06,
+      "loss": 0.0355,
+      "step": 4227
+    },
+    {
+      "epoch": 1.8778592049744613,
+      "grad_norm": 0.4783149493763749,
+      "learning_rate": 6.3860734866471775e-06,
+      "loss": 0.0404,
+      "step": 4228
+    },
+    {
+      "epoch": 1.8783033533200089,
+      "grad_norm": 0.3543783240099781,
+      "learning_rate": 6.384210820904292e-06,
+      "loss": 0.0291,
+      "step": 4229
+    },
+    {
+      "epoch": 1.8787475016655564,
+      "grad_norm": 0.4823964461174852,
+      "learning_rate": 6.382347947091008e-06,
+      "loss": 0.0454,
+      "step": 4230
+    },
+    {
+      "epoch": 1.8791916500111037,
+      "grad_norm": 0.7938854928835402,
+      "learning_rate": 6.380484865487346e-06,
+      "loss": 0.0651,
+      "step": 4231
+    },
+    {
+      "epoch": 1.879635798356651,
+      "grad_norm": 0.43202860148998995,
+      "learning_rate": 6.378621576373356e-06,
+      "loss": 0.0373,
+      "step": 4232
+    },
+    {
+      "epoch": 1.8800799467021987,
+      "grad_norm": 0.40297846856972874,
+      "learning_rate": 6.376758080029126e-06,
+      "loss": 0.0399,
+      "step": 4233
+    },
+    {
+      "epoch": 1.880524095047746,
+      "grad_norm": 0.6777187231862498,
+      "learning_rate": 6.37489437673477e-06,
+      "loss": 0.0451,
+      "step": 4234
+    },
+    {
+      "epoch": 1.8809682433932933,
+      "grad_norm": 0.449724408023701,
+      "learning_rate": 6.3730304667704315e-06,
+      "loss": 0.0364,
+      "step": 4235
+    },
+    {
+      "epoch": 1.8814123917388408,
+      "grad_norm": 0.508560023808873,
+      "learning_rate": 6.371166350416293e-06,
+      "loss": 0.0357,
+      "step": 4236
+    },
+    {
+      "epoch": 1.8818565400843883,
+      "grad_norm": 0.45097496833863476,
+      "learning_rate": 6.369302027952559e-06,
+      "loss": 0.0314,
+      "step": 4237
+    },
+    {
+      "epoch": 1.8823006884299356,
+      "grad_norm": 0.3904261571231129,
+      "learning_rate": 6.36743749965947e-06,
+      "loss": 0.0417,
+      "step": 4238
+    },
+    {
+      "epoch": 1.8827448367754829,
+      "grad_norm": 0.3804238828895874,
+      "learning_rate": 6.365572765817295e-06,
+      "loss": 0.039,
+      "step": 4239
+    },
+    {
+      "epoch": 1.8831889851210304,
+      "grad_norm": 0.5067555487119486,
+      "learning_rate": 6.363707826706336e-06,
+      "loss": 0.0382,
+      "step": 4240
+    },
+    {
+      "epoch": 1.883633133466578,
+      "grad_norm": 0.4732566932753566,
+      "learning_rate": 6.3618426826069265e-06,
+      "loss": 0.0471,
+      "step": 4241
+    },
+    {
+      "epoch": 1.8840772818121252,
+      "grad_norm": 0.4900514966702779,
+      "learning_rate": 6.359977333799429e-06,
+      "loss": 0.0441,
+      "step": 4242
+    },
+    {
+      "epoch": 1.8845214301576727,
+      "grad_norm": 0.5196575689420362,
+      "learning_rate": 6.358111780564233e-06,
+      "loss": 0.0411,
+      "step": 4243
+    },
+    {
+      "epoch": 1.8849655785032202,
+      "grad_norm": 0.43773037282352034,
+      "learning_rate": 6.35624602318177e-06,
+      "loss": 0.0402,
+      "step": 4244
+    },
+    {
+      "epoch": 1.8854097268487675,
+      "grad_norm": 0.589548376792247,
+      "learning_rate": 6.354380061932489e-06,
+      "loss": 0.0381,
+      "step": 4245
+    },
+    {
+      "epoch": 1.8858538751943148,
+      "grad_norm": 0.5624692048798179,
+      "learning_rate": 6.352513897096878e-06,
+      "loss": 0.0394,
+      "step": 4246
+    },
+    {
+      "epoch": 1.8862980235398623,
+      "grad_norm": 0.48689251252375637,
+      "learning_rate": 6.3506475289554534e-06,
+      "loss": 0.0524,
+      "step": 4247
+    },
+    {
+      "epoch": 1.8867421718854098,
+      "grad_norm": 0.8189635814115437,
+      "learning_rate": 6.3487809577887625e-06,
+      "loss": 0.0587,
+      "step": 4248
+    },
+    {
+      "epoch": 1.8871863202309571,
+      "grad_norm": 0.43746947984188156,
+      "learning_rate": 6.346914183877379e-06,
+      "loss": 0.0316,
+      "step": 4249
+    },
+    {
+      "epoch": 1.8876304685765044,
+      "grad_norm": 0.5538154749300077,
+      "learning_rate": 6.345047207501916e-06,
+      "loss": 0.0471,
+      "step": 4250
+    },
+    {
+      "epoch": 1.888074616922052,
+      "grad_norm": 0.4360679802124053,
+      "learning_rate": 6.34318002894301e-06,
+      "loss": 0.0371,
+      "step": 4251
+    },
+    {
+      "epoch": 1.8885187652675994,
+      "grad_norm": 0.5075276918535925,
+      "learning_rate": 6.341312648481328e-06,
+      "loss": 0.0417,
+      "step": 4252
+    },
+    {
+      "epoch": 1.8889629136131467,
+      "grad_norm": 0.49156392064774324,
+      "learning_rate": 6.339445066397569e-06,
+      "loss": 0.0423,
+      "step": 4253
+    },
+    {
+      "epoch": 1.8894070619586942,
+      "grad_norm": 0.852671422019365,
+      "learning_rate": 6.337577282972465e-06,
+      "loss": 0.0524,
+      "step": 4254
+    },
+    {
+      "epoch": 1.8898512103042417,
+      "grad_norm": 0.37799180602912125,
+      "learning_rate": 6.335709298486773e-06,
+      "loss": 0.0303,
+      "step": 4255
+    },
+    {
+      "epoch": 1.890295358649789,
+      "grad_norm": 0.4549492374420902,
+      "learning_rate": 6.333841113221283e-06,
+      "loss": 0.0376,
+      "step": 4256
+    },
+    {
+      "epoch": 1.8907395069953363,
+      "grad_norm": 0.4682663299692366,
+      "learning_rate": 6.331972727456816e-06,
+      "loss": 0.0264,
+      "step": 4257
+    },
+    {
+      "epoch": 1.8911836553408838,
+      "grad_norm": 0.3907535099519968,
+      "learning_rate": 6.330104141474223e-06,
+      "loss": 0.0382,
+      "step": 4258
+    },
+    {
+      "epoch": 1.8916278036864314,
+      "grad_norm": 0.43935294276918296,
+      "learning_rate": 6.328235355554382e-06,
+      "loss": 0.0362,
+      "step": 4259
+    },
+    {
+      "epoch": 1.8920719520319786,
+      "grad_norm": 0.576541252998893,
+      "learning_rate": 6.326366369978204e-06,
+      "loss": 0.0481,
+      "step": 4260
+    },
+    {
+      "epoch": 1.892516100377526,
+      "grad_norm": 0.43778261815885045,
+      "learning_rate": 6.324497185026631e-06,
+      "loss": 0.034,
+      "step": 4261
+    },
+    {
+      "epoch": 1.8929602487230737,
+      "grad_norm": 0.44466828672780784,
+      "learning_rate": 6.3226278009806315e-06,
+      "loss": 0.0307,
+      "step": 4262
+    },
+    {
+      "epoch": 1.893404397068621,
+      "grad_norm": 0.41339084625105765,
+      "learning_rate": 6.320758218121205e-06,
+      "loss": 0.0443,
+      "step": 4263
+    },
+    {
+      "epoch": 1.8938485454141682,
+      "grad_norm": 0.4481199734504317,
+      "learning_rate": 6.318888436729382e-06,
+      "loss": 0.0354,
+      "step": 4264
+    },
+    {
+      "epoch": 1.8942926937597158,
+      "grad_norm": 0.6150629213528874,
+      "learning_rate": 6.317018457086226e-06,
+      "loss": 0.046,
+      "step": 4265
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.8066271182355852,
+      "learning_rate": 6.31514827947282e-06,
+      "loss": 0.0425,
+      "step": 4266
+    },
+    {
+      "epoch": 1.8951809904508106,
+      "grad_norm": 0.5827323877974867,
+      "learning_rate": 6.31327790417029e-06,
+      "loss": 0.0457,
+      "step": 4267
+    },
+    {
+      "epoch": 1.8956251387963579,
+      "grad_norm": 0.4984272032963798,
+      "learning_rate": 6.311407331459781e-06,
+      "loss": 0.03,
+      "step": 4268
+    },
+    {
+      "epoch": 1.8960692871419054,
+      "grad_norm": 0.5018245857439331,
+      "learning_rate": 6.309536561622474e-06,
+      "loss": 0.0399,
+      "step": 4269
+    },
+    {
+      "epoch": 1.8965134354874529,
+      "grad_norm": 0.46552046167173733,
+      "learning_rate": 6.307665594939575e-06,
+      "loss": 0.0379,
+      "step": 4270
+    },
+    {
+      "epoch": 1.8969575838330002,
+      "grad_norm": 0.48037243813958513,
+      "learning_rate": 6.3057944316923246e-06,
+      "loss": 0.0293,
+      "step": 4271
+    },
+    {
+      "epoch": 1.8974017321785477,
+      "grad_norm": 0.4527119332615495,
+      "learning_rate": 6.30392307216199e-06,
+      "loss": 0.0386,
+      "step": 4272
+    },
+    {
+      "epoch": 1.8978458805240952,
+      "grad_norm": 0.419684782365093,
+      "learning_rate": 6.3020515166298665e-06,
+      "loss": 0.0317,
+      "step": 4273
+    },
+    {
+      "epoch": 1.8982900288696425,
+      "grad_norm": 0.418987789446061,
+      "learning_rate": 6.300179765377283e-06,
+      "loss": 0.0361,
+      "step": 4274
+    },
+    {
+      "epoch": 1.8987341772151898,
+      "grad_norm": 0.58612758503126,
+      "learning_rate": 6.298307818685595e-06,
+      "loss": 0.0416,
+      "step": 4275
+    },
+    {
+      "epoch": 1.8991783255607373,
+      "grad_norm": 0.501397444674289,
+      "learning_rate": 6.296435676836188e-06,
+      "loss": 0.0443,
+      "step": 4276
+    },
+    {
+      "epoch": 1.8996224739062848,
+      "grad_norm": 0.4267458008276168,
+      "learning_rate": 6.294563340110474e-06,
+      "loss": 0.0465,
+      "step": 4277
+    },
+    {
+      "epoch": 1.900066622251832,
+      "grad_norm": 0.5455194479999508,
+      "learning_rate": 6.292690808789901e-06,
+      "loss": 0.055,
+      "step": 4278
+    },
+    {
+      "epoch": 1.9005107705973794,
+      "grad_norm": 0.678438654630706,
+      "learning_rate": 6.290818083155941e-06,
+      "loss": 0.0559,
+      "step": 4279
+    },
+    {
+      "epoch": 1.900954918942927,
+      "grad_norm": 0.5367006488443361,
+      "learning_rate": 6.288945163490093e-06,
+      "loss": 0.045,
+      "step": 4280
+    },
+    {
+      "epoch": 1.9013990672884744,
+      "grad_norm": 0.41451518482226474,
+      "learning_rate": 6.287072050073894e-06,
+      "loss": 0.0384,
+      "step": 4281
+    },
+    {
+      "epoch": 1.9018432156340217,
+      "grad_norm": 0.594043909692931,
+      "learning_rate": 6.2851987431889025e-06,
+      "loss": 0.0414,
+      "step": 4282
+    },
+    {
+      "epoch": 1.9022873639795692,
+      "grad_norm": 0.4491119239025248,
+      "learning_rate": 6.2833252431167066e-06,
+      "loss": 0.0393,
+      "step": 4283
+    },
+    {
+      "epoch": 1.9027315123251167,
+      "grad_norm": 0.39577109884937645,
+      "learning_rate": 6.2814515501389275e-06,
+      "loss": 0.035,
+      "step": 4284
+    },
+    {
+      "epoch": 1.903175660670664,
+      "grad_norm": 0.3708469471452283,
+      "learning_rate": 6.279577664537213e-06,
+      "loss": 0.026,
+      "step": 4285
+    },
+    {
+      "epoch": 1.9036198090162113,
+      "grad_norm": 0.4178266369288517,
+      "learning_rate": 6.2777035865932375e-06,
+      "loss": 0.0334,
+      "step": 4286
+    },
+    {
+      "epoch": 1.9040639573617588,
+      "grad_norm": 0.5196821555520535,
+      "learning_rate": 6.275829316588711e-06,
+      "loss": 0.0454,
+      "step": 4287
+    },
+    {
+      "epoch": 1.9045081057073063,
+      "grad_norm": 0.4709231790248879,
+      "learning_rate": 6.273954854805364e-06,
+      "loss": 0.0395,
+      "step": 4288
+    },
+    {
+      "epoch": 1.9049522540528536,
+      "grad_norm": 0.47040632233388197,
+      "learning_rate": 6.2720802015249615e-06,
+      "loss": 0.0459,
+      "step": 4289
+    },
+    {
+      "epoch": 1.905396402398401,
+      "grad_norm": 0.5132518236758158,
+      "learning_rate": 6.2702053570292976e-06,
+      "loss": 0.0453,
+      "step": 4290
+    },
+    {
+      "epoch": 1.9058405507439484,
+      "grad_norm": 0.4975445699296196,
+      "learning_rate": 6.26833032160019e-06,
+      "loss": 0.0384,
+      "step": 4291
+    },
+    {
+      "epoch": 1.906284699089496,
+      "grad_norm": 0.42703967829123596,
+      "learning_rate": 6.26645509551949e-06,
+      "loss": 0.036,
+      "step": 4292
+    },
+    {
+      "epoch": 1.9067288474350432,
+      "grad_norm": 0.4865047041410074,
+      "learning_rate": 6.264579679069077e-06,
+      "loss": 0.0371,
+      "step": 4293
+    },
+    {
+      "epoch": 1.9071729957805907,
+      "grad_norm": 0.5254303510911112,
+      "learning_rate": 6.262704072530856e-06,
+      "loss": 0.0501,
+      "step": 4294
+    },
+    {
+      "epoch": 1.9076171441261383,
+      "grad_norm": 0.36509386500252816,
+      "learning_rate": 6.260828276186762e-06,
+      "loss": 0.0366,
+      "step": 4295
+    },
+    {
+      "epoch": 1.9080612924716855,
+      "grad_norm": 0.4317077297037696,
+      "learning_rate": 6.258952290318763e-06,
+      "loss": 0.0367,
+      "step": 4296
+    },
+    {
+      "epoch": 1.9085054408172328,
+      "grad_norm": 0.4444631290624624,
+      "learning_rate": 6.257076115208847e-06,
+      "loss": 0.0366,
+      "step": 4297
+    },
+    {
+      "epoch": 1.9089495891627803,
+      "grad_norm": 0.3303398456720475,
+      "learning_rate": 6.255199751139036e-06,
+      "loss": 0.03,
+      "step": 4298
+    },
+    {
+      "epoch": 1.9093937375083279,
+      "grad_norm": 0.33258553726398404,
+      "learning_rate": 6.253323198391383e-06,
+      "loss": 0.0345,
+      "step": 4299
+    },
+    {
+      "epoch": 1.9098378858538752,
+      "grad_norm": 0.6445178538911288,
+      "learning_rate": 6.251446457247961e-06,
+      "loss": 0.0444,
+      "step": 4300
+    },
+    {
+      "epoch": 1.9102820341994224,
+      "grad_norm": 0.503486774604937,
+      "learning_rate": 6.249569527990878e-06,
+      "loss": 0.0508,
+      "step": 4301
+    },
+    {
+      "epoch": 1.9107261825449702,
+      "grad_norm": 0.532598812939261,
+      "learning_rate": 6.247692410902271e-06,
+      "loss": 0.0366,
+      "step": 4302
+    },
+    {
+      "epoch": 1.9111703308905175,
+      "grad_norm": 0.5689719802618134,
+      "learning_rate": 6.245815106264297e-06,
+      "loss": 0.0508,
+      "step": 4303
+    },
+    {
+      "epoch": 1.9116144792360648,
+      "grad_norm": 0.44795775453400327,
+      "learning_rate": 6.243937614359152e-06,
+      "loss": 0.0425,
+      "step": 4304
+    },
+    {
+      "epoch": 1.9120586275816123,
+      "grad_norm": 0.63806779442234,
+      "learning_rate": 6.242059935469051e-06,
+      "loss": 0.0409,
+      "step": 4305
+    },
+    {
+      "epoch": 1.9125027759271598,
+      "grad_norm": 0.41950353971072635,
+      "learning_rate": 6.240182069876244e-06,
+      "loss": 0.0344,
+      "step": 4306
+    },
+    {
+      "epoch": 1.912946924272707,
+      "grad_norm": 0.599261124567421,
+      "learning_rate": 6.238304017863005e-06,
+      "loss": 0.0415,
+      "step": 4307
+    },
+    {
+      "epoch": 1.9133910726182544,
+      "grad_norm": 0.364585724464603,
+      "learning_rate": 6.236425779711637e-06,
+      "loss": 0.0346,
+      "step": 4308
+    },
+    {
+      "epoch": 1.9138352209638019,
+      "grad_norm": 0.6254535972915019,
+      "learning_rate": 6.23454735570447e-06,
+      "loss": 0.0509,
+      "step": 4309
+    },
+    {
+      "epoch": 1.9142793693093494,
+      "grad_norm": 0.41487059504531737,
+      "learning_rate": 6.232668746123865e-06,
+      "loss": 0.0331,
+      "step": 4310
+    },
+    {
+      "epoch": 1.9147235176548967,
+      "grad_norm": 0.5104472659733067,
+      "learning_rate": 6.230789951252208e-06,
+      "loss": 0.0349,
+      "step": 4311
+    },
+    {
+      "epoch": 1.9151676660004442,
+      "grad_norm": 0.40794899664296186,
+      "learning_rate": 6.228910971371913e-06,
+      "loss": 0.0408,
+      "step": 4312
+    },
+    {
+      "epoch": 1.9156118143459917,
+      "grad_norm": 0.5437919242408334,
+      "learning_rate": 6.227031806765424e-06,
+      "loss": 0.0387,
+      "step": 4313
+    },
+    {
+      "epoch": 1.916055962691539,
+      "grad_norm": 0.4898004885001543,
+      "learning_rate": 6.225152457715211e-06,
+      "loss": 0.0501,
+      "step": 4314
+    },
+    {
+      "epoch": 1.9165001110370863,
+      "grad_norm": 0.32765369475691036,
+      "learning_rate": 6.223272924503773e-06,
+      "loss": 0.0313,
+      "step": 4315
+    },
+    {
+      "epoch": 1.9169442593826338,
+      "grad_norm": 1.1036222737179877,
+      "learning_rate": 6.221393207413634e-06,
+      "loss": 0.0555,
+      "step": 4316
+    },
+    {
+      "epoch": 1.9173884077281813,
+      "grad_norm": 0.5313402238818737,
+      "learning_rate": 6.219513306727347e-06,
+      "loss": 0.0435,
+      "step": 4317
+    },
+    {
+      "epoch": 1.9178325560737286,
+      "grad_norm": 0.44083561782231834,
+      "learning_rate": 6.217633222727495e-06,
+      "loss": 0.0375,
+      "step": 4318
+    },
+    {
+      "epoch": 1.918276704419276,
+      "grad_norm": 0.9040287143232594,
+      "learning_rate": 6.215752955696686e-06,
+      "loss": 0.0741,
+      "step": 4319
+    },
+    {
+      "epoch": 1.9187208527648234,
+      "grad_norm": 0.406508099143207,
+      "learning_rate": 6.213872505917554e-06,
+      "loss": 0.0309,
+      "step": 4320
+    },
+    {
+      "epoch": 1.919165001110371,
+      "grad_norm": 0.4001138707894935,
+      "learning_rate": 6.2119918736727666e-06,
+      "loss": 0.0292,
+      "step": 4321
+    },
+    {
+      "epoch": 1.9196091494559182,
+      "grad_norm": 0.37528739888488877,
+      "learning_rate": 6.210111059245011e-06,
+      "loss": 0.0323,
+      "step": 4322
+    },
+    {
+      "epoch": 1.9200532978014657,
+      "grad_norm": 0.4611567927734023,
+      "learning_rate": 6.2082300629170065e-06,
+      "loss": 0.0369,
+      "step": 4323
+    },
+    {
+      "epoch": 1.9204974461470132,
+      "grad_norm": 0.4615724924310186,
+      "learning_rate": 6.2063488849715e-06,
+      "loss": 0.0468,
+      "step": 4324
+    },
+    {
+      "epoch": 1.9209415944925605,
+      "grad_norm": 0.36847012924343564,
+      "learning_rate": 6.204467525691265e-06,
+      "loss": 0.0403,
+      "step": 4325
+    },
+    {
+      "epoch": 1.9213857428381078,
+      "grad_norm": 0.7462560129733484,
+      "learning_rate": 6.202585985359099e-06,
+      "loss": 0.0463,
+      "step": 4326
+    },
+    {
+      "epoch": 1.9218298911836553,
+      "grad_norm": 0.4111595385170657,
+      "learning_rate": 6.200704264257832e-06,
+      "loss": 0.0402,
+      "step": 4327
+    },
+    {
+      "epoch": 1.9222740395292028,
+      "grad_norm": 0.5886660338070746,
+      "learning_rate": 6.198822362670316e-06,
+      "loss": 0.0409,
+      "step": 4328
+    },
+    {
+      "epoch": 1.9227181878747501,
+      "grad_norm": 0.551351371253542,
+      "learning_rate": 6.196940280879436e-06,
+      "loss": 0.0371,
+      "step": 4329
+    },
+    {
+      "epoch": 1.9231623362202974,
+      "grad_norm": 0.5211858751187458,
+      "learning_rate": 6.1950580191681e-06,
+      "loss": 0.043,
+      "step": 4330
+    },
+    {
+      "epoch": 1.9236064845658452,
+      "grad_norm": 0.4258998039644192,
+      "learning_rate": 6.193175577819242e-06,
+      "loss": 0.0353,
+      "step": 4331
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.4043175914509727,
+      "learning_rate": 6.191292957115825e-06,
+      "loss": 0.0285,
+      "step": 4332
+    },
+    {
+      "epoch": 1.9244947812569397,
+      "grad_norm": 0.4563276066272647,
+      "learning_rate": 6.1894101573408425e-06,
+      "loss": 0.0371,
+      "step": 4333
+    },
+    {
+      "epoch": 1.9249389296024872,
+      "grad_norm": 0.54090150421601,
+      "learning_rate": 6.1875271787773075e-06,
+      "loss": 0.0519,
+      "step": 4334
+    },
+    {
+      "epoch": 1.9253830779480348,
+      "grad_norm": 0.3562700559697815,
+      "learning_rate": 6.185644021708266e-06,
+      "loss": 0.0267,
+      "step": 4335
+    },
+    {
+      "epoch": 1.925827226293582,
+      "grad_norm": 1.2138310918842605,
+      "learning_rate": 6.183760686416785e-06,
+      "loss": 0.0483,
+      "step": 4336
+    },
+    {
+      "epoch": 1.9262713746391293,
+      "grad_norm": 0.4024294145624342,
+      "learning_rate": 6.181877173185966e-06,
+      "loss": 0.0311,
+      "step": 4337
+    },
+    {
+      "epoch": 1.9267155229846769,
+      "grad_norm": 0.6790926459631366,
+      "learning_rate": 6.1799934822989315e-06,
+      "loss": 0.0494,
+      "step": 4338
+    },
+    {
+      "epoch": 1.9271596713302244,
+      "grad_norm": 0.48577305067511145,
+      "learning_rate": 6.178109614038832e-06,
+      "loss": 0.0412,
+      "step": 4339
+    },
+    {
+      "epoch": 1.9276038196757717,
+      "grad_norm": 0.4461768395787564,
+      "learning_rate": 6.176225568688844e-06,
+      "loss": 0.0337,
+      "step": 4340
+    },
+    {
+      "epoch": 1.9280479680213192,
+      "grad_norm": 0.38001980047585654,
+      "learning_rate": 6.174341346532173e-06,
+      "loss": 0.0404,
+      "step": 4341
+    },
+    {
+      "epoch": 1.9284921163668667,
+      "grad_norm": 0.4252479347353827,
+      "learning_rate": 6.1724569478520495e-06,
+      "loss": 0.0452,
+      "step": 4342
+    },
+    {
+      "epoch": 1.928936264712414,
+      "grad_norm": 0.3799614809822163,
+      "learning_rate": 6.1705723729317295e-06,
+      "loss": 0.0401,
+      "step": 4343
+    },
+    {
+      "epoch": 1.9293804130579613,
+      "grad_norm": 0.8795175854169174,
+      "learning_rate": 6.168687622054497e-06,
+      "loss": 0.0419,
+      "step": 4344
+    },
+    {
+      "epoch": 1.9298245614035088,
+      "grad_norm": 0.35580314936274277,
+      "learning_rate": 6.1668026955036645e-06,
+      "loss": 0.0275,
+      "step": 4345
+    },
+    {
+      "epoch": 1.9302687097490563,
+      "grad_norm": 0.5402722431604882,
+      "learning_rate": 6.1649175935625635e-06,
+      "loss": 0.0459,
+      "step": 4346
+    },
+    {
+      "epoch": 1.9307128580946036,
+      "grad_norm": 0.6557570062151068,
+      "learning_rate": 6.1630323165145615e-06,
+      "loss": 0.0463,
+      "step": 4347
+    },
+    {
+      "epoch": 1.9311570064401509,
+      "grad_norm": 0.45047259263477546,
+      "learning_rate": 6.161146864643045e-06,
+      "loss": 0.0408,
+      "step": 4348
+    },
+    {
+      "epoch": 1.9316011547856984,
+      "grad_norm": 0.4701310559005135,
+      "learning_rate": 6.159261238231431e-06,
+      "loss": 0.0379,
+      "step": 4349
+    },
+    {
+      "epoch": 1.932045303131246,
+      "grad_norm": 0.4022795195574756,
+      "learning_rate": 6.15737543756316e-06,
+      "loss": 0.0372,
+      "step": 4350
+    },
+    {
+      "epoch": 1.9324894514767932,
+      "grad_norm": 0.4838913493812182,
+      "learning_rate": 6.1554894629217e-06,
+      "loss": 0.0512,
+      "step": 4351
+    },
+    {
+      "epoch": 1.9329335998223407,
+      "grad_norm": 1.2278585350906228,
+      "learning_rate": 6.153603314590547e-06,
+      "loss": 0.0864,
+      "step": 4352
+    },
+    {
+      "epoch": 1.9333777481678882,
+      "grad_norm": 0.354715018531944,
+      "learning_rate": 6.1517169928532185e-06,
+      "loss": 0.0324,
+      "step": 4353
+    },
+    {
+      "epoch": 1.9338218965134355,
+      "grad_norm": 0.3643440217338777,
+      "learning_rate": 6.149830497993261e-06,
+      "loss": 0.0346,
+      "step": 4354
+    },
+    {
+      "epoch": 1.9342660448589828,
+      "grad_norm": 0.5368481909179409,
+      "learning_rate": 6.147943830294248e-06,
+      "loss": 0.0406,
+      "step": 4355
+    },
+    {
+      "epoch": 1.9347101932045303,
+      "grad_norm": 0.3847206541326022,
+      "learning_rate": 6.146056990039777e-06,
+      "loss": 0.0291,
+      "step": 4356
+    },
+    {
+      "epoch": 1.9351543415500778,
+      "grad_norm": 0.4195486386731195,
+      "learning_rate": 6.1441699775134724e-06,
+      "loss": 0.037,
+      "step": 4357
+    },
+    {
+      "epoch": 1.935598489895625,
+      "grad_norm": 0.4387846462865948,
+      "learning_rate": 6.142282792998985e-06,
+      "loss": 0.0336,
+      "step": 4358
+    },
+    {
+      "epoch": 1.9360426382411724,
+      "grad_norm": 0.33683241345246556,
+      "learning_rate": 6.14039543677999e-06,
+      "loss": 0.0357,
+      "step": 4359
+    },
+    {
+      "epoch": 1.93648678658672,
+      "grad_norm": 0.39849510173786257,
+      "learning_rate": 6.138507909140187e-06,
+      "loss": 0.0384,
+      "step": 4360
+    },
+    {
+      "epoch": 1.9369309349322674,
+      "grad_norm": 0.4846553810738608,
+      "learning_rate": 6.136620210363307e-06,
+      "loss": 0.0488,
+      "step": 4361
+    },
+    {
+      "epoch": 1.9373750832778147,
+      "grad_norm": 0.44711186795351665,
+      "learning_rate": 6.1347323407331e-06,
+      "loss": 0.0403,
+      "step": 4362
+    },
+    {
+      "epoch": 1.9378192316233622,
+      "grad_norm": 0.546459442186708,
+      "learning_rate": 6.132844300533348e-06,
+      "loss": 0.0442,
+      "step": 4363
+    },
+    {
+      "epoch": 1.9382633799689097,
+      "grad_norm": 0.6727853387902581,
+      "learning_rate": 6.130956090047852e-06,
+      "loss": 0.0542,
+      "step": 4364
+    },
+    {
+      "epoch": 1.938707528314457,
+      "grad_norm": 0.5269547149952615,
+      "learning_rate": 6.129067709560445e-06,
+      "loss": 0.0553,
+      "step": 4365
+    },
+    {
+      "epoch": 1.9391516766600043,
+      "grad_norm": 1.3382161278121019,
+      "learning_rate": 6.127179159354985e-06,
+      "loss": 0.0426,
+      "step": 4366
+    },
+    {
+      "epoch": 1.9395958250055518,
+      "grad_norm": 0.4304150369100395,
+      "learning_rate": 6.125290439715346e-06,
+      "loss": 0.0373,
+      "step": 4367
+    },
+    {
+      "epoch": 1.9400399733510993,
+      "grad_norm": 0.43883673281798163,
+      "learning_rate": 6.12340155092544e-06,
+      "loss": 0.0377,
+      "step": 4368
+    },
+    {
+      "epoch": 1.9404841216966466,
+      "grad_norm": 0.5188193338070988,
+      "learning_rate": 6.121512493269197e-06,
+      "loss": 0.0371,
+      "step": 4369
+    },
+    {
+      "epoch": 1.9409282700421941,
+      "grad_norm": 0.505484234164325,
+      "learning_rate": 6.119623267030576e-06,
+      "loss": 0.0402,
+      "step": 4370
+    },
+    {
+      "epoch": 1.9413724183877417,
+      "grad_norm": 0.5176850701235032,
+      "learning_rate": 6.1177338724935576e-06,
+      "loss": 0.0466,
+      "step": 4371
+    },
+    {
+      "epoch": 1.941816566733289,
+      "grad_norm": 0.4703785260705156,
+      "learning_rate": 6.115844309942153e-06,
+      "loss": 0.04,
+      "step": 4372
+    },
+    {
+      "epoch": 1.9422607150788362,
+      "grad_norm": 0.43937571329979475,
+      "learning_rate": 6.1139545796603925e-06,
+      "loss": 0.0395,
+      "step": 4373
+    },
+    {
+      "epoch": 1.9427048634243838,
+      "grad_norm": 0.5671825068350916,
+      "learning_rate": 6.112064681932335e-06,
+      "loss": 0.0445,
+      "step": 4374
+    },
+    {
+      "epoch": 1.9431490117699313,
+      "grad_norm": 0.38466158842057985,
+      "learning_rate": 6.110174617042066e-06,
+      "loss": 0.033,
+      "step": 4375
+    },
+    {
+      "epoch": 1.9435931601154786,
+      "grad_norm": 0.46940056697212595,
+      "learning_rate": 6.108284385273695e-06,
+      "loss": 0.0448,
+      "step": 4376
+    },
+    {
+      "epoch": 1.9440373084610258,
+      "grad_norm": 0.4747699143093753,
+      "learning_rate": 6.106393986911353e-06,
+      "loss": 0.0459,
+      "step": 4377
+    },
+    {
+      "epoch": 1.9444814568065734,
+      "grad_norm": 0.4922037733238623,
+      "learning_rate": 6.1045034222392e-06,
+      "loss": 0.0462,
+      "step": 4378
+    },
+    {
+      "epoch": 1.9449256051521209,
+      "grad_norm": 0.48074571927831655,
+      "learning_rate": 6.102612691541422e-06,
+      "loss": 0.0433,
+      "step": 4379
+    },
+    {
+      "epoch": 1.9453697534976682,
+      "grad_norm": 0.45838348922641425,
+      "learning_rate": 6.1007217951022244e-06,
+      "loss": 0.0567,
+      "step": 4380
+    },
+    {
+      "epoch": 1.9458139018432157,
+      "grad_norm": 0.39557432428184086,
+      "learning_rate": 6.098830733205844e-06,
+      "loss": 0.0318,
+      "step": 4381
+    },
+    {
+      "epoch": 1.9462580501887632,
+      "grad_norm": 0.4177848324601465,
+      "learning_rate": 6.096939506136539e-06,
+      "loss": 0.0475,
+      "step": 4382
+    },
+    {
+      "epoch": 1.9467021985343105,
+      "grad_norm": 0.4875377634328118,
+      "learning_rate": 6.095048114178591e-06,
+      "loss": 0.0385,
+      "step": 4383
+    },
+    {
+      "epoch": 1.9471463468798578,
+      "grad_norm": 0.3458943442785933,
+      "learning_rate": 6.093156557616311e-06,
+      "loss": 0.0259,
+      "step": 4384
+    },
+    {
+      "epoch": 1.9475904952254053,
+      "grad_norm": 0.48182367581929,
+      "learning_rate": 6.09126483673403e-06,
+      "loss": 0.0516,
+      "step": 4385
+    },
+    {
+      "epoch": 1.9480346435709528,
+      "grad_norm": 0.4073510855592602,
+      "learning_rate": 6.089372951816108e-06,
+      "loss": 0.0313,
+      "step": 4386
+    },
+    {
+      "epoch": 1.9484787919165,
+      "grad_norm": 0.48547952748532,
+      "learning_rate": 6.087480903146926e-06,
+      "loss": 0.0445,
+      "step": 4387
+    },
+    {
+      "epoch": 1.9489229402620474,
+      "grad_norm": 0.5136489604538488,
+      "learning_rate": 6.085588691010888e-06,
+      "loss": 0.0452,
+      "step": 4388
+    },
+    {
+      "epoch": 1.9493670886075949,
+      "grad_norm": 0.37826853557517365,
+      "learning_rate": 6.0836963156924335e-06,
+      "loss": 0.0356,
+      "step": 4389
+    },
+    {
+      "epoch": 1.9498112369531424,
+      "grad_norm": 0.4912572577888434,
+      "learning_rate": 6.081803777476012e-06,
+      "loss": 0.0462,
+      "step": 4390
+    },
+    {
+      "epoch": 1.9502553852986897,
+      "grad_norm": 0.4860207069904635,
+      "learning_rate": 6.079911076646106e-06,
+      "loss": 0.0425,
+      "step": 4391
+    },
+    {
+      "epoch": 1.9506995336442372,
+      "grad_norm": 0.3704904721214982,
+      "learning_rate": 6.07801821348722e-06,
+      "loss": 0.032,
+      "step": 4392
+    },
+    {
+      "epoch": 1.9511436819897847,
+      "grad_norm": 0.4740957603042776,
+      "learning_rate": 6.076125188283885e-06,
+      "loss": 0.0447,
+      "step": 4393
+    },
+    {
+      "epoch": 1.951587830335332,
+      "grad_norm": 0.4597613542477622,
+      "learning_rate": 6.074232001320654e-06,
+      "loss": 0.0508,
+      "step": 4394
+    },
+    {
+      "epoch": 1.9520319786808793,
+      "grad_norm": 0.35666990194120035,
+      "learning_rate": 6.072338652882105e-06,
+      "loss": 0.036,
+      "step": 4395
+    },
+    {
+      "epoch": 1.9524761270264268,
+      "grad_norm": 0.5521794806193095,
+      "learning_rate": 6.070445143252842e-06,
+      "loss": 0.0555,
+      "step": 4396
+    },
+    {
+      "epoch": 1.9529202753719743,
+      "grad_norm": 0.4402940395866301,
+      "learning_rate": 6.0685514727174885e-06,
+      "loss": 0.0365,
+      "step": 4397
+    },
+    {
+      "epoch": 1.9533644237175216,
+      "grad_norm": 0.45905507047472427,
+      "learning_rate": 6.066657641560697e-06,
+      "loss": 0.0371,
+      "step": 4398
+    },
+    {
+      "epoch": 1.953808572063069,
+      "grad_norm": 0.5661699028159852,
+      "learning_rate": 6.064763650067145e-06,
+      "loss": 0.0382,
+      "step": 4399
+    },
+    {
+      "epoch": 1.9542527204086166,
+      "grad_norm": 0.35700483978034786,
+      "learning_rate": 6.062869498521527e-06,
+      "loss": 0.0366,
+      "step": 4400
+    },
+    {
+      "epoch": 1.954696868754164,
+      "grad_norm": 0.43424661857402086,
+      "learning_rate": 6.060975187208569e-06,
+      "loss": 0.0489,
+      "step": 4401
+    },
+    {
+      "epoch": 1.9551410170997112,
+      "grad_norm": 0.4129343066796859,
+      "learning_rate": 6.059080716413016e-06,
+      "loss": 0.0389,
+      "step": 4402
+    },
+    {
+      "epoch": 1.9555851654452587,
+      "grad_norm": 0.5034728724416178,
+      "learning_rate": 6.057186086419643e-06,
+      "loss": 0.043,
+      "step": 4403
+    },
+    {
+      "epoch": 1.9560293137908062,
+      "grad_norm": 0.36382177248066033,
+      "learning_rate": 6.055291297513243e-06,
+      "loss": 0.0305,
+      "step": 4404
+    },
+    {
+      "epoch": 1.9564734621363535,
+      "grad_norm": 0.42171053505933703,
+      "learning_rate": 6.053396349978632e-06,
+      "loss": 0.0537,
+      "step": 4405
+    },
+    {
+      "epoch": 1.9569176104819008,
+      "grad_norm": 0.42351346045820926,
+      "learning_rate": 6.0515012441006574e-06,
+      "loss": 0.0444,
+      "step": 4406
+    },
+    {
+      "epoch": 1.9573617588274483,
+      "grad_norm": 0.3951815066981233,
+      "learning_rate": 6.0496059801641835e-06,
+      "loss": 0.0329,
+      "step": 4407
+    },
+    {
+      "epoch": 1.9578059071729959,
+      "grad_norm": 0.6939750072874196,
+      "learning_rate": 6.047710558454102e-06,
+      "loss": 0.0495,
+      "step": 4408
+    },
+    {
+      "epoch": 1.9582500555185431,
+      "grad_norm": 0.47139234544641123,
+      "learning_rate": 6.0458149792553245e-06,
+      "loss": 0.0338,
+      "step": 4409
+    },
+    {
+      "epoch": 1.9586942038640907,
+      "grad_norm": 0.4743764736759579,
+      "learning_rate": 6.043919242852792e-06,
+      "loss": 0.0401,
+      "step": 4410
+    },
+    {
+      "epoch": 1.9591383522096382,
+      "grad_norm": 0.7527833386589416,
+      "learning_rate": 6.042023349531463e-06,
+      "loss": 0.046,
+      "step": 4411
+    },
+    {
+      "epoch": 1.9595825005551855,
+      "grad_norm": 0.4434725389118798,
+      "learning_rate": 6.040127299576324e-06,
+      "loss": 0.0366,
+      "step": 4412
+    },
+    {
+      "epoch": 1.9600266489007327,
+      "grad_norm": 0.5310028240539822,
+      "learning_rate": 6.038231093272383e-06,
+      "loss": 0.0475,
+      "step": 4413
+    },
+    {
+      "epoch": 1.9604707972462803,
+      "grad_norm": 0.6870302944712058,
+      "learning_rate": 6.036334730904672e-06,
+      "loss": 0.0401,
+      "step": 4414
+    },
+    {
+      "epoch": 1.9609149455918278,
+      "grad_norm": 0.3754707520324698,
+      "learning_rate": 6.034438212758249e-06,
+      "loss": 0.0344,
+      "step": 4415
+    },
+    {
+      "epoch": 1.961359093937375,
+      "grad_norm": 0.7273648929204632,
+      "learning_rate": 6.032541539118188e-06,
+      "loss": 0.059,
+      "step": 4416
+    },
+    {
+      "epoch": 1.9618032422829224,
+      "grad_norm": 0.5563803052374947,
+      "learning_rate": 6.030644710269595e-06,
+      "loss": 0.0337,
+      "step": 4417
+    },
+    {
+      "epoch": 1.9622473906284699,
+      "grad_norm": 0.5196612260430283,
+      "learning_rate": 6.028747726497594e-06,
+      "loss": 0.043,
+      "step": 4418
+    },
+    {
+      "epoch": 1.9626915389740174,
+      "grad_norm": 0.37874941124980105,
+      "learning_rate": 6.026850588087334e-06,
+      "loss": 0.0438,
+      "step": 4419
+    },
+    {
+      "epoch": 1.9631356873195647,
+      "grad_norm": 0.3791041958907497,
+      "learning_rate": 6.024953295323987e-06,
+      "loss": 0.0318,
+      "step": 4420
+    },
+    {
+      "epoch": 1.9635798356651122,
+      "grad_norm": 0.7223736573889502,
+      "learning_rate": 6.02305584849275e-06,
+      "loss": 0.0406,
+      "step": 4421
+    },
+    {
+      "epoch": 1.9640239840106597,
+      "grad_norm": 0.5182650699628792,
+      "learning_rate": 6.02115824787884e-06,
+      "loss": 0.0292,
+      "step": 4422
+    },
+    {
+      "epoch": 1.964468132356207,
+      "grad_norm": 0.42934775755171317,
+      "learning_rate": 6.019260493767499e-06,
+      "loss": 0.0328,
+      "step": 4423
+    },
+    {
+      "epoch": 1.9649122807017543,
+      "grad_norm": 0.5027721140555967,
+      "learning_rate": 6.0173625864439924e-06,
+      "loss": 0.0328,
+      "step": 4424
+    },
+    {
+      "epoch": 1.9653564290473018,
+      "grad_norm": 0.4239046281164633,
+      "learning_rate": 6.015464526193605e-06,
+      "loss": 0.0402,
+      "step": 4425
+    },
+    {
+      "epoch": 1.9658005773928493,
+      "grad_norm": 0.3685503956667686,
+      "learning_rate": 6.013566313301651e-06,
+      "loss": 0.0356,
+      "step": 4426
+    },
+    {
+      "epoch": 1.9662447257383966,
+      "grad_norm": 0.5335432976601984,
+      "learning_rate": 6.011667948053462e-06,
+      "loss": 0.0488,
+      "step": 4427
+    },
+    {
+      "epoch": 1.9666888740839439,
+      "grad_norm": 0.41452010755189883,
+      "learning_rate": 6.009769430734395e-06,
+      "loss": 0.0458,
+      "step": 4428
+    },
+    {
+      "epoch": 1.9671330224294914,
+      "grad_norm": 0.4362710704789863,
+      "learning_rate": 6.007870761629831e-06,
+      "loss": 0.0399,
+      "step": 4429
+    },
+    {
+      "epoch": 1.967577170775039,
+      "grad_norm": 0.40405492276821936,
+      "learning_rate": 6.005971941025171e-06,
+      "loss": 0.0338,
+      "step": 4430
+    },
+    {
+      "epoch": 1.9680213191205862,
+      "grad_norm": 0.602695490290635,
+      "learning_rate": 6.004072969205838e-06,
+      "loss": 0.0684,
+      "step": 4431
+    },
+    {
+      "epoch": 1.9684654674661337,
+      "grad_norm": 0.47106411491289435,
+      "learning_rate": 6.002173846457282e-06,
+      "loss": 0.0402,
+      "step": 4432
+    },
+    {
+      "epoch": 1.9689096158116812,
+      "grad_norm": 0.5941226728121359,
+      "learning_rate": 6.0002745730649725e-06,
+      "loss": 0.0376,
+      "step": 4433
+    },
+    {
+      "epoch": 1.9693537641572285,
+      "grad_norm": 0.3952404366832551,
+      "learning_rate": 5.998375149314404e-06,
+      "loss": 0.0375,
+      "step": 4434
+    },
+    {
+      "epoch": 1.9697979125027758,
+      "grad_norm": 0.5414636177351856,
+      "learning_rate": 5.996475575491091e-06,
+      "loss": 0.036,
+      "step": 4435
+    },
+    {
+      "epoch": 1.9702420608483233,
+      "grad_norm": 0.35141134241433086,
+      "learning_rate": 5.994575851880571e-06,
+      "loss": 0.0299,
+      "step": 4436
+    },
+    {
+      "epoch": 1.9706862091938708,
+      "grad_norm": 0.5469528061005677,
+      "learning_rate": 5.992675978768406e-06,
+      "loss": 0.0312,
+      "step": 4437
+    },
+    {
+      "epoch": 1.9711303575394181,
+      "grad_norm": 0.4485140591176533,
+      "learning_rate": 5.99077595644018e-06,
+      "loss": 0.0397,
+      "step": 4438
+    },
+    {
+      "epoch": 1.9715745058849656,
+      "grad_norm": 0.4559475041999291,
+      "learning_rate": 5.988875785181496e-06,
+      "loss": 0.0399,
+      "step": 4439
+    },
+    {
+      "epoch": 1.9720186542305131,
+      "grad_norm": 0.5184191786649656,
+      "learning_rate": 5.986975465277983e-06,
+      "loss": 0.042,
+      "step": 4440
+    },
+    {
+      "epoch": 1.9724628025760604,
+      "grad_norm": 0.5018306786878316,
+      "learning_rate": 5.9850749970152935e-06,
+      "loss": 0.0433,
+      "step": 4441
+    },
+    {
+      "epoch": 1.9729069509216077,
+      "grad_norm": 0.41853899287083135,
+      "learning_rate": 5.983174380679096e-06,
+      "loss": 0.0311,
+      "step": 4442
+    },
+    {
+      "epoch": 1.9733510992671552,
+      "grad_norm": 0.3979151331311135,
+      "learning_rate": 5.98127361655509e-06,
+      "loss": 0.0393,
+      "step": 4443
+    },
+    {
+      "epoch": 1.9737952476127028,
+      "grad_norm": 0.4773579735193138,
+      "learning_rate": 5.979372704928991e-06,
+      "loss": 0.0404,
+      "step": 4444
+    },
+    {
+      "epoch": 1.97423939595825,
+      "grad_norm": 0.5742352016540867,
+      "learning_rate": 5.977471646086535e-06,
+      "loss": 0.0371,
+      "step": 4445
+    },
+    {
+      "epoch": 1.9746835443037973,
+      "grad_norm": 0.4462263069511724,
+      "learning_rate": 5.97557044031349e-06,
+      "loss": 0.0327,
+      "step": 4446
+    },
+    {
+      "epoch": 1.9751276926493448,
+      "grad_norm": 0.6732019962659918,
+      "learning_rate": 5.973669087895633e-06,
+      "loss": 0.0596,
+      "step": 4447
+    },
+    {
+      "epoch": 1.9755718409948924,
+      "grad_norm": 0.485770657972991,
+      "learning_rate": 5.971767589118772e-06,
+      "loss": 0.0337,
+      "step": 4448
+    },
+    {
+      "epoch": 1.9760159893404396,
+      "grad_norm": 0.342670431757156,
+      "learning_rate": 5.969865944268737e-06,
+      "loss": 0.0303,
+      "step": 4449
+    },
+    {
+      "epoch": 1.9764601376859872,
+      "grad_norm": 0.4112414603189444,
+      "learning_rate": 5.9679641536313734e-06,
+      "loss": 0.038,
+      "step": 4450
+    },
+    {
+      "epoch": 1.9769042860315347,
+      "grad_norm": 0.6127929425160972,
+      "learning_rate": 5.9660622174925564e-06,
+      "loss": 0.0537,
+      "step": 4451
+    },
+    {
+      "epoch": 1.977348434377082,
+      "grad_norm": 0.33024844150086474,
+      "learning_rate": 5.964160136138177e-06,
+      "loss": 0.0353,
+      "step": 4452
+    },
+    {
+      "epoch": 1.9777925827226293,
+      "grad_norm": 0.5529535095882723,
+      "learning_rate": 5.96225790985415e-06,
+      "loss": 0.0442,
+      "step": 4453
+    },
+    {
+      "epoch": 1.9782367310681768,
+      "grad_norm": 0.4225770450854345,
+      "learning_rate": 5.960355538926414e-06,
+      "loss": 0.0494,
+      "step": 4454
+    },
+    {
+      "epoch": 1.9786808794137243,
+      "grad_norm": 0.5282007627303459,
+      "learning_rate": 5.958453023640928e-06,
+      "loss": 0.0416,
+      "step": 4455
+    },
+    {
+      "epoch": 1.9791250277592716,
+      "grad_norm": 0.3898204372093461,
+      "learning_rate": 5.956550364283671e-06,
+      "loss": 0.0373,
+      "step": 4456
+    },
+    {
+      "epoch": 1.9795691761048189,
+      "grad_norm": 0.6543021886189491,
+      "learning_rate": 5.954647561140643e-06,
+      "loss": 0.0406,
+      "step": 4457
+    },
+    {
+      "epoch": 1.9800133244503664,
+      "grad_norm": 0.7003740987141756,
+      "learning_rate": 5.952744614497872e-06,
+      "loss": 0.0437,
+      "step": 4458
+    },
+    {
+      "epoch": 1.9804574727959139,
+      "grad_norm": 0.4937974164168222,
+      "learning_rate": 5.9508415246414e-06,
+      "loss": 0.0539,
+      "step": 4459
+    },
+    {
+      "epoch": 1.9809016211414612,
+      "grad_norm": 0.46863912061469276,
+      "learning_rate": 5.948938291857296e-06,
+      "loss": 0.0503,
+      "step": 4460
+    },
+    {
+      "epoch": 1.9813457694870087,
+      "grad_norm": 0.5235494039593215,
+      "learning_rate": 5.947034916431646e-06,
+      "loss": 0.0462,
+      "step": 4461
+    },
+    {
+      "epoch": 1.9817899178325562,
+      "grad_norm": 0.4097032398279417,
+      "learning_rate": 5.945131398650561e-06,
+      "loss": 0.038,
+      "step": 4462
+    },
+    {
+      "epoch": 1.9822340661781035,
+      "grad_norm": 0.4528672743260559,
+      "learning_rate": 5.943227738800172e-06,
+      "loss": 0.0558,
+      "step": 4463
+    },
+    {
+      "epoch": 1.9826782145236508,
+      "grad_norm": 0.5204018448947473,
+      "learning_rate": 5.941323937166632e-06,
+      "loss": 0.0522,
+      "step": 4464
+    },
+    {
+      "epoch": 1.9831223628691983,
+      "grad_norm": 0.3540421279343773,
+      "learning_rate": 5.939419994036113e-06,
+      "loss": 0.038,
+      "step": 4465
+    },
+    {
+      "epoch": 1.9835665112147458,
+      "grad_norm": 0.6263363230917521,
+      "learning_rate": 5.937515909694811e-06,
+      "loss": 0.0569,
+      "step": 4466
+    },
+    {
+      "epoch": 1.984010659560293,
+      "grad_norm": 0.4116502316940924,
+      "learning_rate": 5.9356116844289426e-06,
+      "loss": 0.0413,
+      "step": 4467
+    },
+    {
+      "epoch": 1.9844548079058404,
+      "grad_norm": 0.44077017956645004,
+      "learning_rate": 5.933707318524744e-06,
+      "loss": 0.0333,
+      "step": 4468
+    },
+    {
+      "epoch": 1.9848989562513881,
+      "grad_norm": 0.3729944832338479,
+      "learning_rate": 5.931802812268476e-06,
+      "loss": 0.0327,
+      "step": 4469
+    },
+    {
+      "epoch": 1.9853431045969354,
+      "grad_norm": 0.40626287554685864,
+      "learning_rate": 5.929898165946416e-06,
+      "loss": 0.0363,
+      "step": 4470
+    },
+    {
+      "epoch": 1.9857872529424827,
+      "grad_norm": 0.5950032125541065,
+      "learning_rate": 5.927993379844864e-06,
+      "loss": 0.0426,
+      "step": 4471
+    },
+    {
+      "epoch": 1.9862314012880302,
+      "grad_norm": 0.40799770820647496,
+      "learning_rate": 5.9260884542501455e-06,
+      "loss": 0.044,
+      "step": 4472
+    },
+    {
+      "epoch": 1.9866755496335777,
+      "grad_norm": 0.41140378050861237,
+      "learning_rate": 5.9241833894486e-06,
+      "loss": 0.0482,
+      "step": 4473
+    },
+    {
+      "epoch": 1.987119697979125,
+      "grad_norm": 0.40378379008973436,
+      "learning_rate": 5.922278185726591e-06,
+      "loss": 0.0327,
+      "step": 4474
+    },
+    {
+      "epoch": 1.9875638463246723,
+      "grad_norm": 0.41453258443290436,
+      "learning_rate": 5.920372843370504e-06,
+      "loss": 0.0416,
+      "step": 4475
+    },
+    {
+      "epoch": 1.9880079946702198,
+      "grad_norm": 0.5109670847768875,
+      "learning_rate": 5.9184673626667455e-06,
+      "loss": 0.039,
+      "step": 4476
+    },
+    {
+      "epoch": 1.9884521430157673,
+      "grad_norm": 0.5686294228064248,
+      "learning_rate": 5.9165617439017395e-06,
+      "loss": 0.0534,
+      "step": 4477
+    },
+    {
+      "epoch": 1.9888962913613146,
+      "grad_norm": 0.4064180941136422,
+      "learning_rate": 5.914655987361934e-06,
+      "loss": 0.0328,
+      "step": 4478
+    },
+    {
+      "epoch": 1.9893404397068621,
+      "grad_norm": 0.5070558821658653,
+      "learning_rate": 5.912750093333796e-06,
+      "loss": 0.0542,
+      "step": 4479
+    },
+    {
+      "epoch": 1.9897845880524097,
+      "grad_norm": 0.3530913079623509,
+      "learning_rate": 5.910844062103814e-06,
+      "loss": 0.0395,
+      "step": 4480
+    },
+    {
+      "epoch": 1.990228736397957,
+      "grad_norm": 0.43523429839148015,
+      "learning_rate": 5.908937893958497e-06,
+      "loss": 0.0366,
+      "step": 4481
+    },
+    {
+      "epoch": 1.9906728847435042,
+      "grad_norm": 0.4229033433291945,
+      "learning_rate": 5.907031589184374e-06,
+      "loss": 0.0383,
+      "step": 4482
+    },
+    {
+      "epoch": 1.9911170330890517,
+      "grad_norm": 0.39083478145330547,
+      "learning_rate": 5.905125148067997e-06,
+      "loss": 0.0371,
+      "step": 4483
+    },
+    {
+      "epoch": 1.9915611814345993,
+      "grad_norm": 0.5025271192866929,
+      "learning_rate": 5.9032185708959354e-06,
+      "loss": 0.0395,
+      "step": 4484
+    },
+    {
+      "epoch": 1.9920053297801465,
+      "grad_norm": 0.349650340955071,
+      "learning_rate": 5.901311857954777e-06,
+      "loss": 0.0266,
+      "step": 4485
+    },
+    {
+      "epoch": 1.9924494781256938,
+      "grad_norm": 0.5279876104400976,
+      "learning_rate": 5.899405009531136e-06,
+      "loss": 0.0402,
+      "step": 4486
+    },
+    {
+      "epoch": 1.9928936264712414,
+      "grad_norm": 0.6233845443742896,
+      "learning_rate": 5.897498025911645e-06,
+      "loss": 0.048,
+      "step": 4487
+    },
+    {
+      "epoch": 1.9933377748167889,
+      "grad_norm": 0.49026655627892524,
+      "learning_rate": 5.8955909073829555e-06,
+      "loss": 0.0412,
+      "step": 4488
+    },
+    {
+      "epoch": 1.9937819231623362,
+      "grad_norm": 0.5000915892039338,
+      "learning_rate": 5.893683654231737e-06,
+      "loss": 0.0394,
+      "step": 4489
+    },
+    {
+      "epoch": 1.9942260715078837,
+      "grad_norm": 0.5000990386217609,
+      "learning_rate": 5.891776266744686e-06,
+      "loss": 0.0355,
+      "step": 4490
+    },
+    {
+      "epoch": 1.9946702198534312,
+      "grad_norm": 0.39359957737650036,
+      "learning_rate": 5.889868745208514e-06,
+      "loss": 0.0304,
+      "step": 4491
+    },
+    {
+      "epoch": 1.9951143681989785,
+      "grad_norm": 0.9651263441707514,
+      "learning_rate": 5.8879610899099505e-06,
+      "loss": 0.0529,
+      "step": 4492
+    },
+    {
+      "epoch": 1.9955585165445258,
+      "grad_norm": 0.32907807731556993,
+      "learning_rate": 5.886053301135755e-06,
+      "loss": 0.0322,
+      "step": 4493
+    },
+    {
+      "epoch": 1.9960026648900733,
+      "grad_norm": 0.4231846959633054,
+      "learning_rate": 5.8841453791726944e-06,
+      "loss": 0.0362,
+      "step": 4494
+    },
+    {
+      "epoch": 1.9964468132356208,
+      "grad_norm": 0.4958581720016726,
+      "learning_rate": 5.882237324307564e-06,
+      "loss": 0.0304,
+      "step": 4495
+    },
+    {
+      "epoch": 1.996890961581168,
+      "grad_norm": 0.44600019470214586,
+      "learning_rate": 5.880329136827178e-06,
+      "loss": 0.0405,
+      "step": 4496
+    },
+    {
+      "epoch": 1.9973351099267154,
+      "grad_norm": 0.5659602027375302,
+      "learning_rate": 5.878420817018369e-06,
+      "loss": 0.0472,
+      "step": 4497
+    },
+    {
+      "epoch": 1.9977792582722629,
+      "grad_norm": 1.0132374448664145,
+      "learning_rate": 5.87651236516799e-06,
+      "loss": 0.0706,
+      "step": 4498
+    },
+    {
+      "epoch": 1.9982234066178104,
+      "grad_norm": 0.4914853042492375,
+      "learning_rate": 5.874603781562911e-06,
+      "loss": 0.0422,
+      "step": 4499
+    },
+    {
+      "epoch": 1.9986675549633577,
+      "grad_norm": 0.40024702771181153,
+      "learning_rate": 5.872695066490028e-06,
+      "loss": 0.0344,
+      "step": 4500
+    },
+    {
+      "epoch": 1.9991117033089052,
+      "grad_norm": 0.49874140037011355,
+      "learning_rate": 5.870786220236253e-06,
+      "loss": 0.0417,
+      "step": 4501
+    },
+    {
+      "epoch": 1.9995558516544527,
+      "grad_norm": 0.47988916644310126,
+      "learning_rate": 5.868877243088515e-06,
+      "loss": 0.0441,
+      "step": 4502
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.904943869460001,
+      "learning_rate": 5.866968135333769e-06,
+      "loss": 0.0502,
+      "step": 4503
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.04465880244970322,
+      "eval_runtime": 403.5178,
+      "eval_samples_per_second": 37.584,
+      "eval_steps_per_second": 1.175,
+      "step": 4503
+    },
+    {
+      "epoch": 2.0004441483455473,
+      "grad_norm": 0.4375728669928517,
+      "learning_rate": 5.8650588972589865e-06,
+      "loss": 0.0263,
+      "step": 4504
+    },
+    {
+      "epoch": 2.000888296691095,
+      "grad_norm": 0.9215657711078544,
+      "learning_rate": 5.863149529151154e-06,
+      "loss": 0.0492,
+      "step": 4505
+    },
+    {
+      "epoch": 2.0013324450366423,
+      "grad_norm": 0.5271072063506441,
+      "learning_rate": 5.8612400312972865e-06,
+      "loss": 0.0508,
+      "step": 4506
+    },
+    {
+      "epoch": 2.0017765933821896,
+      "grad_norm": 0.48411386914362037,
+      "learning_rate": 5.859330403984413e-06,
+      "loss": 0.0371,
+      "step": 4507
+    },
+    {
+      "epoch": 2.002220741727737,
+      "grad_norm": 0.4017233429265068,
+      "learning_rate": 5.85742064749958e-06,
+      "loss": 0.0418,
+      "step": 4508
+    },
+    {
+      "epoch": 2.0026648900732846,
+      "grad_norm": 0.4120119737365717,
+      "learning_rate": 5.85551076212986e-06,
+      "loss": 0.046,
+      "step": 4509
+    },
+    {
+      "epoch": 2.003109038418832,
+      "grad_norm": 0.4366707287416164,
+      "learning_rate": 5.8536007481623406e-06,
+      "loss": 0.0443,
+      "step": 4510
+    },
+    {
+      "epoch": 2.003553186764379,
+      "grad_norm": 0.7145508695971371,
+      "learning_rate": 5.851690605884127e-06,
+      "loss": 0.0532,
+      "step": 4511
+    },
+    {
+      "epoch": 2.0039973351099265,
+      "grad_norm": 0.5011936156881229,
+      "learning_rate": 5.84978033558235e-06,
+      "loss": 0.0354,
+      "step": 4512
+    },
+    {
+      "epoch": 2.0044414834554742,
+      "grad_norm": 0.36581304940892057,
+      "learning_rate": 5.847869937544151e-06,
+      "loss": 0.0294,
+      "step": 4513
+    },
+    {
+      "epoch": 2.0048856318010215,
+      "grad_norm": 0.37996248718965503,
+      "learning_rate": 5.845959412056699e-06,
+      "loss": 0.0286,
+      "step": 4514
+    },
+    {
+      "epoch": 2.005329780146569,
+      "grad_norm": 0.5097034078222489,
+      "learning_rate": 5.844048759407177e-06,
+      "loss": 0.0414,
+      "step": 4515
+    },
+    {
+      "epoch": 2.0057739284921166,
+      "grad_norm": 0.4963020471009325,
+      "learning_rate": 5.842137979882786e-06,
+      "loss": 0.0459,
+      "step": 4516
+    },
+    {
+      "epoch": 2.006218076837664,
+      "grad_norm": 0.6161319347902735,
+      "learning_rate": 5.840227073770754e-06,
+      "loss": 0.0629,
+      "step": 4517
+    },
+    {
+      "epoch": 2.006662225183211,
+      "grad_norm": 0.3511248203446856,
+      "learning_rate": 5.838316041358319e-06,
+      "loss": 0.0295,
+      "step": 4518
+    },
+    {
+      "epoch": 2.0071063735287584,
+      "grad_norm": 0.37830736549229443,
+      "learning_rate": 5.836404882932744e-06,
+      "loss": 0.0319,
+      "step": 4519
+    },
+    {
+      "epoch": 2.007550521874306,
+      "grad_norm": 0.6333845141434062,
+      "learning_rate": 5.8344935987813045e-06,
+      "loss": 0.0379,
+      "step": 4520
+    },
+    {
+      "epoch": 2.0079946702198535,
+      "grad_norm": 0.5166443794796353,
+      "learning_rate": 5.832582189191304e-06,
+      "loss": 0.0346,
+      "step": 4521
+    },
+    {
+      "epoch": 2.0084388185654007,
+      "grad_norm": 0.4317191740162663,
+      "learning_rate": 5.8306706544500544e-06,
+      "loss": 0.0319,
+      "step": 4522
+    },
+    {
+      "epoch": 2.0088829669109485,
+      "grad_norm": 0.4388602499538576,
+      "learning_rate": 5.828758994844896e-06,
+      "loss": 0.0377,
+      "step": 4523
+    },
+    {
+      "epoch": 2.0093271152564958,
+      "grad_norm": 0.5946947521373457,
+      "learning_rate": 5.826847210663184e-06,
+      "loss": 0.0421,
+      "step": 4524
+    },
+    {
+      "epoch": 2.009771263602043,
+      "grad_norm": 0.3821952135869883,
+      "learning_rate": 5.8249353021922895e-06,
+      "loss": 0.0303,
+      "step": 4525
+    },
+    {
+      "epoch": 2.0102154119475903,
+      "grad_norm": 0.5519322372289749,
+      "learning_rate": 5.823023269719606e-06,
+      "loss": 0.0513,
+      "step": 4526
+    },
+    {
+      "epoch": 2.010659560293138,
+      "grad_norm": 0.6313520947120921,
+      "learning_rate": 5.821111113532545e-06,
+      "loss": 0.0484,
+      "step": 4527
+    },
+    {
+      "epoch": 2.0111037086386854,
+      "grad_norm": 0.5368562306090153,
+      "learning_rate": 5.819198833918533e-06,
+      "loss": 0.051,
+      "step": 4528
+    },
+    {
+      "epoch": 2.0115478569842327,
+      "grad_norm": 0.39624325698963303,
+      "learning_rate": 5.817286431165024e-06,
+      "loss": 0.0333,
+      "step": 4529
+    },
+    {
+      "epoch": 2.01199200532978,
+      "grad_norm": 0.411233856841024,
+      "learning_rate": 5.815373905559478e-06,
+      "loss": 0.0312,
+      "step": 4530
+    },
+    {
+      "epoch": 2.0124361536753277,
+      "grad_norm": 0.3742461123258978,
+      "learning_rate": 5.813461257389384e-06,
+      "loss": 0.0268,
+      "step": 4531
+    },
+    {
+      "epoch": 2.012880302020875,
+      "grad_norm": 0.3749818385865246,
+      "learning_rate": 5.811548486942246e-06,
+      "loss": 0.0346,
+      "step": 4532
+    },
+    {
+      "epoch": 2.0133244503664223,
+      "grad_norm": 0.34116601308623506,
+      "learning_rate": 5.809635594505585e-06,
+      "loss": 0.0251,
+      "step": 4533
+    },
+    {
+      "epoch": 2.01376859871197,
+      "grad_norm": 0.5178521236398543,
+      "learning_rate": 5.807722580366939e-06,
+      "loss": 0.0381,
+      "step": 4534
+    },
+    {
+      "epoch": 2.0142127470575173,
+      "grad_norm": 0.32707032102652794,
+      "learning_rate": 5.805809444813869e-06,
+      "loss": 0.0275,
+      "step": 4535
+    },
+    {
+      "epoch": 2.0146568954030646,
+      "grad_norm": 0.4024749636693741,
+      "learning_rate": 5.80389618813395e-06,
+      "loss": 0.0365,
+      "step": 4536
+    },
+    {
+      "epoch": 2.015101043748612,
+      "grad_norm": 0.3506397945464826,
+      "learning_rate": 5.8019828106147805e-06,
+      "loss": 0.0252,
+      "step": 4537
+    },
+    {
+      "epoch": 2.0155451920941596,
+      "grad_norm": 0.5822851446048461,
+      "learning_rate": 5.80006931254397e-06,
+      "loss": 0.0369,
+      "step": 4538
+    },
+    {
+      "epoch": 2.015989340439707,
+      "grad_norm": 0.44067833638988113,
+      "learning_rate": 5.798155694209151e-06,
+      "loss": 0.0344,
+      "step": 4539
+    },
+    {
+      "epoch": 2.016433488785254,
+      "grad_norm": 0.4460004586157456,
+      "learning_rate": 5.796241955897972e-06,
+      "loss": 0.0381,
+      "step": 4540
+    },
+    {
+      "epoch": 2.0168776371308015,
+      "grad_norm": 0.3328661208832795,
+      "learning_rate": 5.7943280978981034e-06,
+      "loss": 0.0284,
+      "step": 4541
+    },
+    {
+      "epoch": 2.017321785476349,
+      "grad_norm": 0.48357435513498903,
+      "learning_rate": 5.792414120497227e-06,
+      "loss": 0.0361,
+      "step": 4542
+    },
+    {
+      "epoch": 2.0177659338218965,
+      "grad_norm": 0.4866683235779641,
+      "learning_rate": 5.790500023983049e-06,
+      "loss": 0.0359,
+      "step": 4543
+    },
+    {
+      "epoch": 2.018210082167444,
+      "grad_norm": 0.39359993621118305,
+      "learning_rate": 5.788585808643287e-06,
+      "loss": 0.0281,
+      "step": 4544
+    },
+    {
+      "epoch": 2.0186542305129915,
+      "grad_norm": 0.4058975435223384,
+      "learning_rate": 5.786671474765683e-06,
+      "loss": 0.0314,
+      "step": 4545
+    },
+    {
+      "epoch": 2.019098378858539,
+      "grad_norm": 0.38121689511830104,
+      "learning_rate": 5.784757022637993e-06,
+      "loss": 0.0291,
+      "step": 4546
+    },
+    {
+      "epoch": 2.019542527204086,
+      "grad_norm": 0.4508375438456433,
+      "learning_rate": 5.782842452547992e-06,
+      "loss": 0.0334,
+      "step": 4547
+    },
+    {
+      "epoch": 2.0199866755496334,
+      "grad_norm": 0.42673347363830266,
+      "learning_rate": 5.780927764783473e-06,
+      "loss": 0.0249,
+      "step": 4548
+    },
+    {
+      "epoch": 2.020430823895181,
+      "grad_norm": 0.43669165432209434,
+      "learning_rate": 5.779012959632244e-06,
+      "loss": 0.0358,
+      "step": 4549
+    },
+    {
+      "epoch": 2.0208749722407284,
+      "grad_norm": 0.4022820994220209,
+      "learning_rate": 5.777098037382135e-06,
+      "loss": 0.0293,
+      "step": 4550
+    },
+    {
+      "epoch": 2.0213191205862757,
+      "grad_norm": 0.42387266543810437,
+      "learning_rate": 5.77518299832099e-06,
+      "loss": 0.0384,
+      "step": 4551
+    },
+    {
+      "epoch": 2.021763268931823,
+      "grad_norm": 0.4585217445539665,
+      "learning_rate": 5.7732678427366725e-06,
+      "loss": 0.0349,
+      "step": 4552
+    },
+    {
+      "epoch": 2.0222074172773707,
+      "grad_norm": 0.44358027335422784,
+      "learning_rate": 5.771352570917062e-06,
+      "loss": 0.0332,
+      "step": 4553
+    },
+    {
+      "epoch": 2.022651565622918,
+      "grad_norm": 0.43507443913798416,
+      "learning_rate": 5.769437183150057e-06,
+      "loss": 0.0347,
+      "step": 4554
+    },
+    {
+      "epoch": 2.0230957139684653,
+      "grad_norm": 0.37205140444853446,
+      "learning_rate": 5.767521679723574e-06,
+      "loss": 0.029,
+      "step": 4555
+    },
+    {
+      "epoch": 2.023539862314013,
+      "grad_norm": 0.3603917915607534,
+      "learning_rate": 5.765606060925545e-06,
+      "loss": 0.0309,
+      "step": 4556
+    },
+    {
+      "epoch": 2.0239840106595604,
+      "grad_norm": 0.5400074623803027,
+      "learning_rate": 5.763690327043919e-06,
+      "loss": 0.0459,
+      "step": 4557
+    },
+    {
+      "epoch": 2.0244281590051076,
+      "grad_norm": 0.39760305609807395,
+      "learning_rate": 5.761774478366664e-06,
+      "loss": 0.0338,
+      "step": 4558
+    },
+    {
+      "epoch": 2.024872307350655,
+      "grad_norm": 0.44010489846206996,
+      "learning_rate": 5.759858515181763e-06,
+      "loss": 0.0305,
+      "step": 4559
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.38439690133076265,
+      "learning_rate": 5.757942437777222e-06,
+      "loss": 0.0281,
+      "step": 4560
+    },
+    {
+      "epoch": 2.02576060404175,
+      "grad_norm": 0.36756874447130394,
+      "learning_rate": 5.756026246441056e-06,
+      "loss": 0.0276,
+      "step": 4561
+    },
+    {
+      "epoch": 2.0262047523872972,
+      "grad_norm": 0.3543127710037651,
+      "learning_rate": 5.754109941461302e-06,
+      "loss": 0.0304,
+      "step": 4562
+    },
+    {
+      "epoch": 2.026648900732845,
+      "grad_norm": 0.4789471062184959,
+      "learning_rate": 5.7521935231260166e-06,
+      "loss": 0.0361,
+      "step": 4563
+    },
+    {
+      "epoch": 2.0270930490783923,
+      "grad_norm": 0.4658261130750723,
+      "learning_rate": 5.7502769917232635e-06,
+      "loss": 0.0297,
+      "step": 4564
+    },
+    {
+      "epoch": 2.0275371974239396,
+      "grad_norm": 0.37364282464180315,
+      "learning_rate": 5.748360347541136e-06,
+      "loss": 0.0301,
+      "step": 4565
+    },
+    {
+      "epoch": 2.027981345769487,
+      "grad_norm": 0.5624254377961666,
+      "learning_rate": 5.746443590867735e-06,
+      "loss": 0.0375,
+      "step": 4566
+    },
+    {
+      "epoch": 2.0284254941150346,
+      "grad_norm": 0.4305222164333765,
+      "learning_rate": 5.7445267219911815e-06,
+      "loss": 0.033,
+      "step": 4567
+    },
+    {
+      "epoch": 2.028869642460582,
+      "grad_norm": 0.669501618559028,
+      "learning_rate": 5.742609741199615e-06,
+      "loss": 0.0497,
+      "step": 4568
+    },
+    {
+      "epoch": 2.029313790806129,
+      "grad_norm": 0.38137453136293886,
+      "learning_rate": 5.740692648781191e-06,
+      "loss": 0.0377,
+      "step": 4569
+    },
+    {
+      "epoch": 2.0297579391516765,
+      "grad_norm": 0.4610447553595522,
+      "learning_rate": 5.738775445024078e-06,
+      "loss": 0.0394,
+      "step": 4570
+    },
+    {
+      "epoch": 2.030202087497224,
+      "grad_norm": 0.46477609561023486,
+      "learning_rate": 5.736858130216468e-06,
+      "loss": 0.044,
+      "step": 4571
+    },
+    {
+      "epoch": 2.0306462358427715,
+      "grad_norm": 0.34915412378821464,
+      "learning_rate": 5.7349407046465625e-06,
+      "loss": 0.0234,
+      "step": 4572
+    },
+    {
+      "epoch": 2.0310903841883188,
+      "grad_norm": 0.47023014530194523,
+      "learning_rate": 5.733023168602584e-06,
+      "loss": 0.0427,
+      "step": 4573
+    },
+    {
+      "epoch": 2.0315345325338665,
+      "grad_norm": 0.5052588063022365,
+      "learning_rate": 5.731105522372773e-06,
+      "loss": 0.0406,
+      "step": 4574
+    },
+    {
+      "epoch": 2.031978680879414,
+      "grad_norm": 0.5345856053201472,
+      "learning_rate": 5.729187766245382e-06,
+      "loss": 0.0347,
+      "step": 4575
+    },
+    {
+      "epoch": 2.032422829224961,
+      "grad_norm": 0.41581297859158317,
+      "learning_rate": 5.727269900508682e-06,
+      "loss": 0.0366,
+      "step": 4576
+    },
+    {
+      "epoch": 2.0328669775705084,
+      "grad_norm": 0.5119888393853211,
+      "learning_rate": 5.725351925450964e-06,
+      "loss": 0.0372,
+      "step": 4577
+    },
+    {
+      "epoch": 2.033311125916056,
+      "grad_norm": 0.4247836793476431,
+      "learning_rate": 5.723433841360528e-06,
+      "loss": 0.0259,
+      "step": 4578
+    },
+    {
+      "epoch": 2.0337552742616034,
+      "grad_norm": 0.3959142952189294,
+      "learning_rate": 5.721515648525698e-06,
+      "loss": 0.0358,
+      "step": 4579
+    },
+    {
+      "epoch": 2.0341994226071507,
+      "grad_norm": 0.4826767145510797,
+      "learning_rate": 5.719597347234809e-06,
+      "loss": 0.0404,
+      "step": 4580
+    },
+    {
+      "epoch": 2.034643570952698,
+      "grad_norm": 0.4091839747164647,
+      "learning_rate": 5.7176789377762155e-06,
+      "loss": 0.0421,
+      "step": 4581
+    },
+    {
+      "epoch": 2.0350877192982457,
+      "grad_norm": 0.41330812118932564,
+      "learning_rate": 5.715760420438284e-06,
+      "loss": 0.025,
+      "step": 4582
+    },
+    {
+      "epoch": 2.035531867643793,
+      "grad_norm": 0.3786001612535725,
+      "learning_rate": 5.713841795509405e-06,
+      "loss": 0.0304,
+      "step": 4583
+    },
+    {
+      "epoch": 2.0359760159893403,
+      "grad_norm": 0.38932682295296067,
+      "learning_rate": 5.711923063277979e-06,
+      "loss": 0.033,
+      "step": 4584
+    },
+    {
+      "epoch": 2.036420164334888,
+      "grad_norm": 0.3925654682001415,
+      "learning_rate": 5.710004224032421e-06,
+      "loss": 0.035,
+      "step": 4585
+    },
+    {
+      "epoch": 2.0368643126804353,
+      "grad_norm": 0.40837083418560577,
+      "learning_rate": 5.708085278061167e-06,
+      "loss": 0.0232,
+      "step": 4586
+    },
+    {
+      "epoch": 2.0373084610259826,
+      "grad_norm": 0.4396699075372347,
+      "learning_rate": 5.706166225652669e-06,
+      "loss": 0.0404,
+      "step": 4587
+    },
+    {
+      "epoch": 2.03775260937153,
+      "grad_norm": 0.486841383129979,
+      "learning_rate": 5.704247067095391e-06,
+      "loss": 0.0325,
+      "step": 4588
+    },
+    {
+      "epoch": 2.0381967577170776,
+      "grad_norm": 0.4503221450108852,
+      "learning_rate": 5.702327802677815e-06,
+      "loss": 0.0346,
+      "step": 4589
+    },
+    {
+      "epoch": 2.038640906062625,
+      "grad_norm": 0.4565221863051935,
+      "learning_rate": 5.70040843268844e-06,
+      "loss": 0.0404,
+      "step": 4590
+    },
+    {
+      "epoch": 2.0390850544081722,
+      "grad_norm": 0.39766027067422877,
+      "learning_rate": 5.698488957415782e-06,
+      "loss": 0.0325,
+      "step": 4591
+    },
+    {
+      "epoch": 2.03952920275372,
+      "grad_norm": 0.469121812114123,
+      "learning_rate": 5.6965693771483654e-06,
+      "loss": 0.0361,
+      "step": 4592
+    },
+    {
+      "epoch": 2.0399733510992673,
+      "grad_norm": 0.34294205911608633,
+      "learning_rate": 5.6946496921747394e-06,
+      "loss": 0.0274,
+      "step": 4593
+    },
+    {
+      "epoch": 2.0404174994448145,
+      "grad_norm": 0.45197263089243245,
+      "learning_rate": 5.692729902783467e-06,
+      "loss": 0.0357,
+      "step": 4594
+    },
+    {
+      "epoch": 2.040861647790362,
+      "grad_norm": 0.4143752687268431,
+      "learning_rate": 5.6908100092631215e-06,
+      "loss": 0.0325,
+      "step": 4595
+    },
+    {
+      "epoch": 2.0413057961359096,
+      "grad_norm": 0.5126979012787386,
+      "learning_rate": 5.688890011902295e-06,
+      "loss": 0.0402,
+      "step": 4596
+    },
+    {
+      "epoch": 2.041749944481457,
+      "grad_norm": 0.38129445526321254,
+      "learning_rate": 5.686969910989599e-06,
+      "loss": 0.0247,
+      "step": 4597
+    },
+    {
+      "epoch": 2.042194092827004,
+      "grad_norm": 0.43439418379805417,
+      "learning_rate": 5.685049706813657e-06,
+      "loss": 0.0334,
+      "step": 4598
+    },
+    {
+      "epoch": 2.0426382411725514,
+      "grad_norm": 0.3528242448836029,
+      "learning_rate": 5.683129399663105e-06,
+      "loss": 0.0237,
+      "step": 4599
+    },
+    {
+      "epoch": 2.043082389518099,
+      "grad_norm": 0.4979008424907728,
+      "learning_rate": 5.681208989826601e-06,
+      "loss": 0.0346,
+      "step": 4600
+    },
+    {
+      "epoch": 2.0435265378636465,
+      "grad_norm": 0.4344255344249799,
+      "learning_rate": 5.679288477592815e-06,
+      "loss": 0.0288,
+      "step": 4601
+    },
+    {
+      "epoch": 2.0439706862091938,
+      "grad_norm": 0.5875305767400001,
+      "learning_rate": 5.67736786325043e-06,
+      "loss": 0.0336,
+      "step": 4602
+    },
+    {
+      "epoch": 2.0444148345547415,
+      "grad_norm": 0.6007666135225987,
+      "learning_rate": 5.675447147088148e-06,
+      "loss": 0.0337,
+      "step": 4603
+    },
+    {
+      "epoch": 2.044858982900289,
+      "grad_norm": 0.3501303227446166,
+      "learning_rate": 5.673526329394688e-06,
+      "loss": 0.0316,
+      "step": 4604
+    },
+    {
+      "epoch": 2.045303131245836,
+      "grad_norm": 0.5466639051174087,
+      "learning_rate": 5.6716054104587784e-06,
+      "loss": 0.0513,
+      "step": 4605
+    },
+    {
+      "epoch": 2.0457472795913834,
+      "grad_norm": 0.3453198669378243,
+      "learning_rate": 5.669684390569167e-06,
+      "loss": 0.0292,
+      "step": 4606
+    },
+    {
+      "epoch": 2.046191427936931,
+      "grad_norm": 0.3855523349827862,
+      "learning_rate": 5.667763270014616e-06,
+      "loss": 0.0274,
+      "step": 4607
+    },
+    {
+      "epoch": 2.0466355762824784,
+      "grad_norm": 0.34309904151982246,
+      "learning_rate": 5.665842049083902e-06,
+      "loss": 0.0264,
+      "step": 4608
+    },
+    {
+      "epoch": 2.0470797246280257,
+      "grad_norm": 0.7952938969316026,
+      "learning_rate": 5.6639207280658194e-06,
+      "loss": 0.0534,
+      "step": 4609
+    },
+    {
+      "epoch": 2.047523872973573,
+      "grad_norm": 0.3602547692842435,
+      "learning_rate": 5.6619993072491694e-06,
+      "loss": 0.027,
+      "step": 4610
+    },
+    {
+      "epoch": 2.0479680213191207,
+      "grad_norm": 0.33009638203411035,
+      "learning_rate": 5.6600777869227805e-06,
+      "loss": 0.0267,
+      "step": 4611
+    },
+    {
+      "epoch": 2.048412169664668,
+      "grad_norm": 0.43117489579425794,
+      "learning_rate": 5.658156167375488e-06,
+      "loss": 0.038,
+      "step": 4612
+    },
+    {
+      "epoch": 2.0488563180102153,
+      "grad_norm": 0.7172674302994839,
+      "learning_rate": 5.656234448896142e-06,
+      "loss": 0.0461,
+      "step": 4613
+    },
+    {
+      "epoch": 2.049300466355763,
+      "grad_norm": 0.4507173154420471,
+      "learning_rate": 5.654312631773612e-06,
+      "loss": 0.027,
+      "step": 4614
+    },
+    {
+      "epoch": 2.0497446147013103,
+      "grad_norm": 0.6398319387217463,
+      "learning_rate": 5.652390716296778e-06,
+      "loss": 0.041,
+      "step": 4615
+    },
+    {
+      "epoch": 2.0501887630468576,
+      "grad_norm": 0.43458565414603195,
+      "learning_rate": 5.650468702754537e-06,
+      "loss": 0.0475,
+      "step": 4616
+    },
+    {
+      "epoch": 2.050632911392405,
+      "grad_norm": 0.4724564626628951,
+      "learning_rate": 5.6485465914358005e-06,
+      "loss": 0.039,
+      "step": 4617
+    },
+    {
+      "epoch": 2.0510770597379526,
+      "grad_norm": 0.4813725140216776,
+      "learning_rate": 5.646624382629495e-06,
+      "loss": 0.0442,
+      "step": 4618
+    },
+    {
+      "epoch": 2.0515212080835,
+      "grad_norm": 0.422057899941897,
+      "learning_rate": 5.64470207662456e-06,
+      "loss": 0.0354,
+      "step": 4619
+    },
+    {
+      "epoch": 2.051965356429047,
+      "grad_norm": 0.5520244840521104,
+      "learning_rate": 5.6427796737099515e-06,
+      "loss": 0.0393,
+      "step": 4620
+    },
+    {
+      "epoch": 2.0524095047745945,
+      "grad_norm": 0.5717238340997483,
+      "learning_rate": 5.64085717417464e-06,
+      "loss": 0.041,
+      "step": 4621
+    },
+    {
+      "epoch": 2.0528536531201422,
+      "grad_norm": 0.4162602609657127,
+      "learning_rate": 5.638934578307608e-06,
+      "loss": 0.0233,
+      "step": 4622
+    },
+    {
+      "epoch": 2.0532978014656895,
+      "grad_norm": 0.40689474682922516,
+      "learning_rate": 5.637011886397854e-06,
+      "loss": 0.0479,
+      "step": 4623
+    },
+    {
+      "epoch": 2.053741949811237,
+      "grad_norm": 0.46291048789069594,
+      "learning_rate": 5.635089098734394e-06,
+      "loss": 0.0345,
+      "step": 4624
+    },
+    {
+      "epoch": 2.0541860981567845,
+      "grad_norm": 0.6061112924495777,
+      "learning_rate": 5.633166215606254e-06,
+      "loss": 0.0519,
+      "step": 4625
+    },
+    {
+      "epoch": 2.054630246502332,
+      "grad_norm": 0.37385156175737194,
+      "learning_rate": 5.631243237302478e-06,
+      "loss": 0.0432,
+      "step": 4626
+    },
+    {
+      "epoch": 2.055074394847879,
+      "grad_norm": 0.32050349885999035,
+      "learning_rate": 5.629320164112116e-06,
+      "loss": 0.031,
+      "step": 4627
+    },
+    {
+      "epoch": 2.0555185431934264,
+      "grad_norm": 0.44064173532980155,
+      "learning_rate": 5.627396996324247e-06,
+      "loss": 0.0255,
+      "step": 4628
+    },
+    {
+      "epoch": 2.055962691538974,
+      "grad_norm": 0.35873848998988794,
+      "learning_rate": 5.625473734227952e-06,
+      "loss": 0.0352,
+      "step": 4629
+    },
+    {
+      "epoch": 2.0564068398845214,
+      "grad_norm": 0.6158412215543486,
+      "learning_rate": 5.623550378112328e-06,
+      "loss": 0.0442,
+      "step": 4630
+    },
+    {
+      "epoch": 2.0568509882300687,
+      "grad_norm": 0.4813854059047201,
+      "learning_rate": 5.621626928266489e-06,
+      "loss": 0.0478,
+      "step": 4631
+    },
+    {
+      "epoch": 2.0572951365756165,
+      "grad_norm": 0.4790652853444873,
+      "learning_rate": 5.619703384979566e-06,
+      "loss": 0.0463,
+      "step": 4632
+    },
+    {
+      "epoch": 2.0577392849211638,
+      "grad_norm": 0.41738179511720797,
+      "learning_rate": 5.617779748540695e-06,
+      "loss": 0.0338,
+      "step": 4633
+    },
+    {
+      "epoch": 2.058183433266711,
+      "grad_norm": 0.41477600567937595,
+      "learning_rate": 5.615856019239034e-06,
+      "loss": 0.0361,
+      "step": 4634
+    },
+    {
+      "epoch": 2.0586275816122583,
+      "grad_norm": 0.4967975613891594,
+      "learning_rate": 5.613932197363753e-06,
+      "loss": 0.0333,
+      "step": 4635
+    },
+    {
+      "epoch": 2.059071729957806,
+      "grad_norm": 0.49458901207002304,
+      "learning_rate": 5.612008283204033e-06,
+      "loss": 0.0427,
+      "step": 4636
+    },
+    {
+      "epoch": 2.0595158783033534,
+      "grad_norm": 0.42427512577649273,
+      "learning_rate": 5.610084277049071e-06,
+      "loss": 0.0354,
+      "step": 4637
+    },
+    {
+      "epoch": 2.0599600266489007,
+      "grad_norm": 0.43245137715403886,
+      "learning_rate": 5.608160179188079e-06,
+      "loss": 0.0326,
+      "step": 4638
+    },
+    {
+      "epoch": 2.060404174994448,
+      "grad_norm": 0.4874634691089771,
+      "learning_rate": 5.6062359899102815e-06,
+      "loss": 0.0371,
+      "step": 4639
+    },
+    {
+      "epoch": 2.0608483233399957,
+      "grad_norm": 0.5004987254382077,
+      "learning_rate": 5.604311709504917e-06,
+      "loss": 0.0397,
+      "step": 4640
+    },
+    {
+      "epoch": 2.061292471685543,
+      "grad_norm": 0.4751004755592279,
+      "learning_rate": 5.602387338261236e-06,
+      "loss": 0.0414,
+      "step": 4641
+    },
+    {
+      "epoch": 2.0617366200310903,
+      "grad_norm": 0.5299506551694105,
+      "learning_rate": 5.600462876468506e-06,
+      "loss": 0.0368,
+      "step": 4642
+    },
+    {
+      "epoch": 2.062180768376638,
+      "grad_norm": 0.5312690498368814,
+      "learning_rate": 5.598538324416007e-06,
+      "loss": 0.0449,
+      "step": 4643
+    },
+    {
+      "epoch": 2.0626249167221853,
+      "grad_norm": 0.38593169971896735,
+      "learning_rate": 5.5966136823930286e-06,
+      "loss": 0.0281,
+      "step": 4644
+    },
+    {
+      "epoch": 2.0630690650677326,
+      "grad_norm": 0.3958642054782115,
+      "learning_rate": 5.594688950688879e-06,
+      "loss": 0.034,
+      "step": 4645
+    },
+    {
+      "epoch": 2.06351321341328,
+      "grad_norm": 0.40570614021956714,
+      "learning_rate": 5.592764129592879e-06,
+      "loss": 0.0323,
+      "step": 4646
+    },
+    {
+      "epoch": 2.0639573617588276,
+      "grad_norm": 0.45390531771906856,
+      "learning_rate": 5.590839219394361e-06,
+      "loss": 0.0254,
+      "step": 4647
+    },
+    {
+      "epoch": 2.064401510104375,
+      "grad_norm": 0.39108513738732803,
+      "learning_rate": 5.58891422038267e-06,
+      "loss": 0.0278,
+      "step": 4648
+    },
+    {
+      "epoch": 2.064845658449922,
+      "grad_norm": 0.5223612545370472,
+      "learning_rate": 5.58698913284717e-06,
+      "loss": 0.0367,
+      "step": 4649
+    },
+    {
+      "epoch": 2.0652898067954695,
+      "grad_norm": 0.48567914535043527,
+      "learning_rate": 5.585063957077231e-06,
+      "loss": 0.0303,
+      "step": 4650
+    },
+    {
+      "epoch": 2.065733955141017,
+      "grad_norm": 0.3022843285079369,
+      "learning_rate": 5.583138693362241e-06,
+      "loss": 0.0268,
+      "step": 4651
+    },
+    {
+      "epoch": 2.0661781034865645,
+      "grad_norm": 0.3626425665807493,
+      "learning_rate": 5.5812133419916e-06,
+      "loss": 0.0282,
+      "step": 4652
+    },
+    {
+      "epoch": 2.066622251832112,
+      "grad_norm": 0.4001866579471354,
+      "learning_rate": 5.5792879032547205e-06,
+      "loss": 0.0346,
+      "step": 4653
+    },
+    {
+      "epoch": 2.0670664001776595,
+      "grad_norm": 0.4654295875105734,
+      "learning_rate": 5.577362377441029e-06,
+      "loss": 0.0348,
+      "step": 4654
+    },
+    {
+      "epoch": 2.067510548523207,
+      "grad_norm": 0.39677357569841354,
+      "learning_rate": 5.5754367648399644e-06,
+      "loss": 0.0274,
+      "step": 4655
+    },
+    {
+      "epoch": 2.067954696868754,
+      "grad_norm": 0.4143564936380847,
+      "learning_rate": 5.5735110657409775e-06,
+      "loss": 0.0427,
+      "step": 4656
+    },
+    {
+      "epoch": 2.0683988452143014,
+      "grad_norm": 0.3681145367593844,
+      "learning_rate": 5.571585280433537e-06,
+      "loss": 0.0321,
+      "step": 4657
+    },
+    {
+      "epoch": 2.068842993559849,
+      "grad_norm": 0.46007574827334824,
+      "learning_rate": 5.569659409207119e-06,
+      "loss": 0.0324,
+      "step": 4658
+    },
+    {
+      "epoch": 2.0692871419053964,
+      "grad_norm": 0.4377445107919845,
+      "learning_rate": 5.567733452351214e-06,
+      "loss": 0.0364,
+      "step": 4659
+    },
+    {
+      "epoch": 2.0697312902509437,
+      "grad_norm": 0.5314428707092943,
+      "learning_rate": 5.565807410155329e-06,
+      "loss": 0.039,
+      "step": 4660
+    },
+    {
+      "epoch": 2.0701754385964914,
+      "grad_norm": 0.4445997114256765,
+      "learning_rate": 5.563881282908976e-06,
+      "loss": 0.0451,
+      "step": 4661
+    },
+    {
+      "epoch": 2.0706195869420387,
+      "grad_norm": 0.41675191455829885,
+      "learning_rate": 5.561955070901689e-06,
+      "loss": 0.032,
+      "step": 4662
+    },
+    {
+      "epoch": 2.071063735287586,
+      "grad_norm": 0.49984516013690966,
+      "learning_rate": 5.56002877442301e-06,
+      "loss": 0.0379,
+      "step": 4663
+    },
+    {
+      "epoch": 2.0715078836331333,
+      "grad_norm": 0.48011634971976885,
+      "learning_rate": 5.558102393762491e-06,
+      "loss": 0.0289,
+      "step": 4664
+    },
+    {
+      "epoch": 2.071952031978681,
+      "grad_norm": 0.44249526000612904,
+      "learning_rate": 5.556175929209703e-06,
+      "loss": 0.0254,
+      "step": 4665
+    },
+    {
+      "epoch": 2.0723961803242283,
+      "grad_norm": 0.40597832272770745,
+      "learning_rate": 5.554249381054224e-06,
+      "loss": 0.0269,
+      "step": 4666
+    },
+    {
+      "epoch": 2.0728403286697756,
+      "grad_norm": 0.45148588795717426,
+      "learning_rate": 5.552322749585649e-06,
+      "loss": 0.0388,
+      "step": 4667
+    },
+    {
+      "epoch": 2.073284477015323,
+      "grad_norm": 0.598994856507402,
+      "learning_rate": 5.550396035093582e-06,
+      "loss": 0.0423,
+      "step": 4668
+    },
+    {
+      "epoch": 2.0737286253608707,
+      "grad_norm": 0.5181098142664673,
+      "learning_rate": 5.548469237867642e-06,
+      "loss": 0.0433,
+      "step": 4669
+    },
+    {
+      "epoch": 2.074172773706418,
+      "grad_norm": 0.5101400816921863,
+      "learning_rate": 5.546542358197458e-06,
+      "loss": 0.0463,
+      "step": 4670
+    },
+    {
+      "epoch": 2.0746169220519652,
+      "grad_norm": 0.5101937458580871,
+      "learning_rate": 5.544615396372673e-06,
+      "loss": 0.0341,
+      "step": 4671
+    },
+    {
+      "epoch": 2.075061070397513,
+      "grad_norm": 0.5392434065378148,
+      "learning_rate": 5.542688352682944e-06,
+      "loss": 0.0451,
+      "step": 4672
+    },
+    {
+      "epoch": 2.0755052187430603,
+      "grad_norm": 0.6897926097007651,
+      "learning_rate": 5.540761227417934e-06,
+      "loss": 0.0355,
+      "step": 4673
+    },
+    {
+      "epoch": 2.0759493670886076,
+      "grad_norm": 0.28051948597694654,
+      "learning_rate": 5.53883402086733e-06,
+      "loss": 0.0243,
+      "step": 4674
+    },
+    {
+      "epoch": 2.076393515434155,
+      "grad_norm": 0.793052180191732,
+      "learning_rate": 5.536906733320816e-06,
+      "loss": 0.0563,
+      "step": 4675
+    },
+    {
+      "epoch": 2.0768376637797026,
+      "grad_norm": 0.4215318597715739,
+      "learning_rate": 5.5349793650681006e-06,
+      "loss": 0.0354,
+      "step": 4676
+    },
+    {
+      "epoch": 2.07728181212525,
+      "grad_norm": 0.34425051368314813,
+      "learning_rate": 5.533051916398899e-06,
+      "loss": 0.0295,
+      "step": 4677
+    },
+    {
+      "epoch": 2.077725960470797,
+      "grad_norm": 0.6685404197751216,
+      "learning_rate": 5.531124387602938e-06,
+      "loss": 0.0399,
+      "step": 4678
+    },
+    {
+      "epoch": 2.0781701088163445,
+      "grad_norm": 0.31141908709280325,
+      "learning_rate": 5.529196778969961e-06,
+      "loss": 0.0271,
+      "step": 4679
+    },
+    {
+      "epoch": 2.078614257161892,
+      "grad_norm": 0.4606806580907262,
+      "learning_rate": 5.527269090789718e-06,
+      "loss": 0.0394,
+      "step": 4680
+    },
+    {
+      "epoch": 2.0790584055074395,
+      "grad_norm": 0.4252999601959849,
+      "learning_rate": 5.525341323351975e-06,
+      "loss": 0.0394,
+      "step": 4681
+    },
+    {
+      "epoch": 2.0795025538529868,
+      "grad_norm": 0.5409739246288477,
+      "learning_rate": 5.5234134769465065e-06,
+      "loss": 0.0521,
+      "step": 4682
+    },
+    {
+      "epoch": 2.0799467021985345,
+      "grad_norm": 0.4948477331543779,
+      "learning_rate": 5.5214855518631005e-06,
+      "loss": 0.0485,
+      "step": 4683
+    },
+    {
+      "epoch": 2.080390850544082,
+      "grad_norm": 0.3543340984392336,
+      "learning_rate": 5.519557548391557e-06,
+      "loss": 0.0242,
+      "step": 4684
+    },
+    {
+      "epoch": 2.080834998889629,
+      "grad_norm": 0.3751926442365401,
+      "learning_rate": 5.517629466821691e-06,
+      "loss": 0.036,
+      "step": 4685
+    },
+    {
+      "epoch": 2.0812791472351764,
+      "grad_norm": 0.4117924266399266,
+      "learning_rate": 5.515701307443321e-06,
+      "loss": 0.0485,
+      "step": 4686
+    },
+    {
+      "epoch": 2.081723295580724,
+      "grad_norm": 0.47251260035381365,
+      "learning_rate": 5.513773070546284e-06,
+      "loss": 0.0419,
+      "step": 4687
+    },
+    {
+      "epoch": 2.0821674439262714,
+      "grad_norm": 0.6076861002224307,
+      "learning_rate": 5.5118447564204295e-06,
+      "loss": 0.0391,
+      "step": 4688
+    },
+    {
+      "epoch": 2.0826115922718187,
+      "grad_norm": 0.4792646073001442,
+      "learning_rate": 5.50991636535561e-06,
+      "loss": 0.0295,
+      "step": 4689
+    },
+    {
+      "epoch": 2.0830557406173664,
+      "grad_norm": 0.32440405189073857,
+      "learning_rate": 5.5079878976417e-06,
+      "loss": 0.022,
+      "step": 4690
+    },
+    {
+      "epoch": 2.0834998889629137,
+      "grad_norm": 0.4221580236556979,
+      "learning_rate": 5.506059353568581e-06,
+      "loss": 0.0263,
+      "step": 4691
+    },
+    {
+      "epoch": 2.083944037308461,
+      "grad_norm": 0.35561395384075745,
+      "learning_rate": 5.504130733426145e-06,
+      "loss": 0.0273,
+      "step": 4692
+    },
+    {
+      "epoch": 2.0843881856540083,
+      "grad_norm": 0.4935918433503542,
+      "learning_rate": 5.502202037504293e-06,
+      "loss": 0.0328,
+      "step": 4693
+    },
+    {
+      "epoch": 2.084832333999556,
+      "grad_norm": 0.439030655630297,
+      "learning_rate": 5.500273266092947e-06,
+      "loss": 0.0278,
+      "step": 4694
+    },
+    {
+      "epoch": 2.0852764823451033,
+      "grad_norm": 0.5492104667996629,
+      "learning_rate": 5.49834441948203e-06,
+      "loss": 0.0477,
+      "step": 4695
+    },
+    {
+      "epoch": 2.0857206306906506,
+      "grad_norm": 0.48997887857295236,
+      "learning_rate": 5.496415497961482e-06,
+      "loss": 0.036,
+      "step": 4696
+    },
+    {
+      "epoch": 2.086164779036198,
+      "grad_norm": 0.9251501688761363,
+      "learning_rate": 5.49448650182125e-06,
+      "loss": 0.0543,
+      "step": 4697
+    },
+    {
+      "epoch": 2.0866089273817456,
+      "grad_norm": 0.44096687951153934,
+      "learning_rate": 5.492557431351298e-06,
+      "loss": 0.0372,
+      "step": 4698
+    },
+    {
+      "epoch": 2.087053075727293,
+      "grad_norm": 0.48261463177180974,
+      "learning_rate": 5.4906282868415974e-06,
+      "loss": 0.0323,
+      "step": 4699
+    },
+    {
+      "epoch": 2.08749722407284,
+      "grad_norm": 0.44612947746443893,
+      "learning_rate": 5.488699068582129e-06,
+      "loss": 0.0379,
+      "step": 4700
+    },
+    {
+      "epoch": 2.087941372418388,
+      "grad_norm": 0.6019236481517227,
+      "learning_rate": 5.486769776862891e-06,
+      "loss": 0.0392,
+      "step": 4701
+    },
+    {
+      "epoch": 2.0883855207639352,
+      "grad_norm": 0.41037964449549674,
+      "learning_rate": 5.484840411973888e-06,
+      "loss": 0.045,
+      "step": 4702
+    },
+    {
+      "epoch": 2.0888296691094825,
+      "grad_norm": 0.6139112370737255,
+      "learning_rate": 5.482910974205133e-06,
+      "loss": 0.0434,
+      "step": 4703
+    },
+    {
+      "epoch": 2.08927381745503,
+      "grad_norm": 0.47897436644713304,
+      "learning_rate": 5.480981463846655e-06,
+      "loss": 0.0407,
+      "step": 4704
+    },
+    {
+      "epoch": 2.0897179658005776,
+      "grad_norm": 0.4257961843862624,
+      "learning_rate": 5.479051881188494e-06,
+      "loss": 0.0402,
+      "step": 4705
+    },
+    {
+      "epoch": 2.090162114146125,
+      "grad_norm": 0.4458277075765739,
+      "learning_rate": 5.477122226520698e-06,
+      "loss": 0.0335,
+      "step": 4706
+    },
+    {
+      "epoch": 2.090606262491672,
+      "grad_norm": 0.5283109987423077,
+      "learning_rate": 5.475192500133324e-06,
+      "loss": 0.0321,
+      "step": 4707
+    },
+    {
+      "epoch": 2.0910504108372194,
+      "grad_norm": 0.6005581697681983,
+      "learning_rate": 5.473262702316447e-06,
+      "loss": 0.0369,
+      "step": 4708
+    },
+    {
+      "epoch": 2.091494559182767,
+      "grad_norm": 0.40161109082786944,
+      "learning_rate": 5.471332833360147e-06,
+      "loss": 0.0289,
+      "step": 4709
+    },
+    {
+      "epoch": 2.0919387075283145,
+      "grad_norm": 0.4371088945729004,
+      "learning_rate": 5.4694028935545126e-06,
+      "loss": 0.0343,
+      "step": 4710
+    },
+    {
+      "epoch": 2.0923828558738617,
+      "grad_norm": 1.1703243713580254,
+      "learning_rate": 5.467472883189653e-06,
+      "loss": 0.0507,
+      "step": 4711
+    },
+    {
+      "epoch": 2.0928270042194095,
+      "grad_norm": 0.40364633007175293,
+      "learning_rate": 5.465542802555677e-06,
+      "loss": 0.0295,
+      "step": 4712
+    },
+    {
+      "epoch": 2.0932711525649568,
+      "grad_norm": 0.46666883180291846,
+      "learning_rate": 5.4636126519427095e-06,
+      "loss": 0.0412,
+      "step": 4713
+    },
+    {
+      "epoch": 2.093715300910504,
+      "grad_norm": 0.4145958391325037,
+      "learning_rate": 5.461682431640885e-06,
+      "loss": 0.0319,
+      "step": 4714
+    },
+    {
+      "epoch": 2.0941594492560514,
+      "grad_norm": 0.3217656626768428,
+      "learning_rate": 5.459752141940347e-06,
+      "loss": 0.0318,
+      "step": 4715
+    },
+    {
+      "epoch": 2.094603597601599,
+      "grad_norm": 0.5490857014126957,
+      "learning_rate": 5.457821783131254e-06,
+      "loss": 0.046,
+      "step": 4716
+    },
+    {
+      "epoch": 2.0950477459471464,
+      "grad_norm": 0.4425046183099989,
+      "learning_rate": 5.455891355503768e-06,
+      "loss": 0.0299,
+      "step": 4717
+    },
+    {
+      "epoch": 2.0954918942926937,
+      "grad_norm": 0.6349791543930909,
+      "learning_rate": 5.453960859348069e-06,
+      "loss": 0.0565,
+      "step": 4718
+    },
+    {
+      "epoch": 2.095936042638241,
+      "grad_norm": 0.4072328187100098,
+      "learning_rate": 5.4520302949543415e-06,
+      "loss": 0.0349,
+      "step": 4719
+    },
+    {
+      "epoch": 2.0963801909837887,
+      "grad_norm": 0.3811118913667539,
+      "learning_rate": 5.450099662612781e-06,
+      "loss": 0.0305,
+      "step": 4720
+    },
+    {
+      "epoch": 2.096824339329336,
+      "grad_norm": 0.4774255455908721,
+      "learning_rate": 5.448168962613596e-06,
+      "loss": 0.0368,
+      "step": 4721
+    },
+    {
+      "epoch": 2.0972684876748833,
+      "grad_norm": 0.32813215206793156,
+      "learning_rate": 5.446238195247003e-06,
+      "loss": 0.0321,
+      "step": 4722
+    },
+    {
+      "epoch": 2.097712636020431,
+      "grad_norm": 0.5512782021205551,
+      "learning_rate": 5.44430736080323e-06,
+      "loss": 0.0387,
+      "step": 4723
+    },
+    {
+      "epoch": 2.0981567843659783,
+      "grad_norm": 0.5835802051010996,
+      "learning_rate": 5.44237645957251e-06,
+      "loss": 0.0354,
+      "step": 4724
+    },
+    {
+      "epoch": 2.0986009327115256,
+      "grad_norm": 0.3806252215797266,
+      "learning_rate": 5.440445491845095e-06,
+      "loss": 0.0363,
+      "step": 4725
+    },
+    {
+      "epoch": 2.099045081057073,
+      "grad_norm": 0.42251102762378473,
+      "learning_rate": 5.438514457911241e-06,
+      "loss": 0.0386,
+      "step": 4726
+    },
+    {
+      "epoch": 2.0994892294026206,
+      "grad_norm": 0.35748316520246215,
+      "learning_rate": 5.436583358061215e-06,
+      "loss": 0.0282,
+      "step": 4727
+    },
+    {
+      "epoch": 2.099933377748168,
+      "grad_norm": 0.46907671848937027,
+      "learning_rate": 5.434652192585294e-06,
+      "loss": 0.0411,
+      "step": 4728
+    },
+    {
+      "epoch": 2.100377526093715,
+      "grad_norm": 0.41478959673973786,
+      "learning_rate": 5.432720961773765e-06,
+      "loss": 0.0323,
+      "step": 4729
+    },
+    {
+      "epoch": 2.1008216744392625,
+      "grad_norm": 0.6441994593730457,
+      "learning_rate": 5.430789665916925e-06,
+      "loss": 0.0477,
+      "step": 4730
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.7936242646685359,
+      "learning_rate": 5.428858305305079e-06,
+      "loss": 0.0438,
+      "step": 4731
+    },
+    {
+      "epoch": 2.1017099711303575,
+      "grad_norm": 0.4542564531841097,
+      "learning_rate": 5.426926880228547e-06,
+      "loss": 0.039,
+      "step": 4732
+    },
+    {
+      "epoch": 2.102154119475905,
+      "grad_norm": 0.43879296755150776,
+      "learning_rate": 5.424995390977651e-06,
+      "loss": 0.024,
+      "step": 4733
+    },
+    {
+      "epoch": 2.1025982678214525,
+      "grad_norm": 0.4720404728255635,
+      "learning_rate": 5.423063837842728e-06,
+      "loss": 0.0366,
+      "step": 4734
+    },
+    {
+      "epoch": 2.103042416167,
+      "grad_norm": 0.3210152022294204,
+      "learning_rate": 5.421132221114124e-06,
+      "loss": 0.0287,
+      "step": 4735
+    },
+    {
+      "epoch": 2.103486564512547,
+      "grad_norm": 0.8156671041659774,
+      "learning_rate": 5.419200541082194e-06,
+      "loss": 0.0423,
+      "step": 4736
+    },
+    {
+      "epoch": 2.1039307128580944,
+      "grad_norm": 0.5417603660166322,
+      "learning_rate": 5.417268798037303e-06,
+      "loss": 0.0388,
+      "step": 4737
+    },
+    {
+      "epoch": 2.104374861203642,
+      "grad_norm": 0.46205560016247954,
+      "learning_rate": 5.415336992269821e-06,
+      "loss": 0.0363,
+      "step": 4738
+    },
+    {
+      "epoch": 2.1048190095491894,
+      "grad_norm": 0.3348835489607936,
+      "learning_rate": 5.413405124070134e-06,
+      "loss": 0.0206,
+      "step": 4739
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.46193255668613026,
+      "learning_rate": 5.411473193728636e-06,
+      "loss": 0.0295,
+      "step": 4740
+    },
+    {
+      "epoch": 2.1057073062402845,
+      "grad_norm": 0.4346087883354204,
+      "learning_rate": 5.409541201535727e-06,
+      "loss": 0.039,
+      "step": 4741
+    },
+    {
+      "epoch": 2.1061514545858318,
+      "grad_norm": 0.46994212871132296,
+      "learning_rate": 5.407609147781816e-06,
+      "loss": 0.0374,
+      "step": 4742
+    },
+    {
+      "epoch": 2.106595602931379,
+      "grad_norm": 0.4605573678630878,
+      "learning_rate": 5.405677032757329e-06,
+      "loss": 0.0371,
+      "step": 4743
+    },
+    {
+      "epoch": 2.1070397512769263,
+      "grad_norm": 0.6425519607388686,
+      "learning_rate": 5.403744856752691e-06,
+      "loss": 0.0337,
+      "step": 4744
+    },
+    {
+      "epoch": 2.107483899622474,
+      "grad_norm": 0.34739298130141844,
+      "learning_rate": 5.401812620058343e-06,
+      "loss": 0.0269,
+      "step": 4745
+    },
+    {
+      "epoch": 2.1079280479680214,
+      "grad_norm": 0.4490417625063517,
+      "learning_rate": 5.399880322964733e-06,
+      "loss": 0.0372,
+      "step": 4746
+    },
+    {
+      "epoch": 2.1083721963135686,
+      "grad_norm": 0.6489786195301251,
+      "learning_rate": 5.397947965762317e-06,
+      "loss": 0.0509,
+      "step": 4747
+    },
+    {
+      "epoch": 2.108816344659116,
+      "grad_norm": 0.326521120100658,
+      "learning_rate": 5.396015548741562e-06,
+      "loss": 0.0288,
+      "step": 4748
+    },
+    {
+      "epoch": 2.1092604930046637,
+      "grad_norm": 0.4078407477455278,
+      "learning_rate": 5.394083072192944e-06,
+      "loss": 0.0382,
+      "step": 4749
+    },
+    {
+      "epoch": 2.109704641350211,
+      "grad_norm": 0.5177065744163067,
+      "learning_rate": 5.392150536406945e-06,
+      "loss": 0.0449,
+      "step": 4750
+    },
+    {
+      "epoch": 2.1101487896957583,
+      "grad_norm": 0.33995975799200945,
+      "learning_rate": 5.39021794167406e-06,
+      "loss": 0.0264,
+      "step": 4751
+    },
+    {
+      "epoch": 2.110592938041306,
+      "grad_norm": 0.42003616424190965,
+      "learning_rate": 5.388285288284787e-06,
+      "loss": 0.0372,
+      "step": 4752
+    },
+    {
+      "epoch": 2.1110370863868533,
+      "grad_norm": 0.3472432854029145,
+      "learning_rate": 5.386352576529641e-06,
+      "loss": 0.0307,
+      "step": 4753
+    },
+    {
+      "epoch": 2.1114812347324006,
+      "grad_norm": 0.4130591691843907,
+      "learning_rate": 5.384419806699141e-06,
+      "loss": 0.0276,
+      "step": 4754
+    },
+    {
+      "epoch": 2.111925383077948,
+      "grad_norm": 0.4591952950693401,
+      "learning_rate": 5.382486979083812e-06,
+      "loss": 0.0363,
+      "step": 4755
+    },
+    {
+      "epoch": 2.1123695314234956,
+      "grad_norm": 0.3857971037693101,
+      "learning_rate": 5.380554093974193e-06,
+      "loss": 0.0359,
+      "step": 4756
+    },
+    {
+      "epoch": 2.112813679769043,
+      "grad_norm": 0.37782187453237637,
+      "learning_rate": 5.37862115166083e-06,
+      "loss": 0.0358,
+      "step": 4757
+    },
+    {
+      "epoch": 2.11325782811459,
+      "grad_norm": 0.3395886410665437,
+      "learning_rate": 5.376688152434275e-06,
+      "loss": 0.0243,
+      "step": 4758
+    },
+    {
+      "epoch": 2.1137019764601375,
+      "grad_norm": 0.4470609752189812,
+      "learning_rate": 5.374755096585093e-06,
+      "loss": 0.0246,
+      "step": 4759
+    },
+    {
+      "epoch": 2.114146124805685,
+      "grad_norm": 0.4371584818717651,
+      "learning_rate": 5.372821984403854e-06,
+      "loss": 0.028,
+      "step": 4760
+    },
+    {
+      "epoch": 2.1145902731512325,
+      "grad_norm": 0.49305971027495465,
+      "learning_rate": 5.370888816181138e-06,
+      "loss": 0.0362,
+      "step": 4761
+    },
+    {
+      "epoch": 2.11503442149678,
+      "grad_norm": 0.4175857188422132,
+      "learning_rate": 5.368955592207531e-06,
+      "loss": 0.033,
+      "step": 4762
+    },
+    {
+      "epoch": 2.1154785698423275,
+      "grad_norm": 0.5139037264567338,
+      "learning_rate": 5.367022312773633e-06,
+      "loss": 0.0323,
+      "step": 4763
+    },
+    {
+      "epoch": 2.115922718187875,
+      "grad_norm": 0.8385325013800425,
+      "learning_rate": 5.365088978170045e-06,
+      "loss": 0.0387,
+      "step": 4764
+    },
+    {
+      "epoch": 2.116366866533422,
+      "grad_norm": 0.4295996848147971,
+      "learning_rate": 5.363155588687383e-06,
+      "loss": 0.0297,
+      "step": 4765
+    },
+    {
+      "epoch": 2.1168110148789694,
+      "grad_norm": 0.5951452068654114,
+      "learning_rate": 5.361222144616267e-06,
+      "loss": 0.0395,
+      "step": 4766
+    },
+    {
+      "epoch": 2.117255163224517,
+      "grad_norm": 0.3679591157867813,
+      "learning_rate": 5.359288646247326e-06,
+      "loss": 0.0324,
+      "step": 4767
+    },
+    {
+      "epoch": 2.1176993115700644,
+      "grad_norm": 0.351749833643755,
+      "learning_rate": 5.357355093871199e-06,
+      "loss": 0.0262,
+      "step": 4768
+    },
+    {
+      "epoch": 2.1181434599156117,
+      "grad_norm": 0.4550233701632367,
+      "learning_rate": 5.355421487778529e-06,
+      "loss": 0.0405,
+      "step": 4769
+    },
+    {
+      "epoch": 2.1185876082611594,
+      "grad_norm": 0.5196049007489856,
+      "learning_rate": 5.353487828259973e-06,
+      "loss": 0.0354,
+      "step": 4770
+    },
+    {
+      "epoch": 2.1190317566067067,
+      "grad_norm": 0.5003818333286508,
+      "learning_rate": 5.351554115606194e-06,
+      "loss": 0.0346,
+      "step": 4771
+    },
+    {
+      "epoch": 2.119475904952254,
+      "grad_norm": 0.48260320470669443,
+      "learning_rate": 5.349620350107857e-06,
+      "loss": 0.0363,
+      "step": 4772
+    },
+    {
+      "epoch": 2.1199200532978013,
+      "grad_norm": 0.5154065971255504,
+      "learning_rate": 5.347686532055643e-06,
+      "loss": 0.037,
+      "step": 4773
+    },
+    {
+      "epoch": 2.120364201643349,
+      "grad_norm": 0.5641510479972823,
+      "learning_rate": 5.345752661740236e-06,
+      "loss": 0.0418,
+      "step": 4774
+    },
+    {
+      "epoch": 2.1208083499888963,
+      "grad_norm": 0.3842072493781642,
+      "learning_rate": 5.343818739452332e-06,
+      "loss": 0.0289,
+      "step": 4775
+    },
+    {
+      "epoch": 2.1212524983344436,
+      "grad_norm": 0.48007816649878726,
+      "learning_rate": 5.34188476548263e-06,
+      "loss": 0.0382,
+      "step": 4776
+    },
+    {
+      "epoch": 2.121696646679991,
+      "grad_norm": 0.403727363474673,
+      "learning_rate": 5.339950740121842e-06,
+      "loss": 0.0355,
+      "step": 4777
+    },
+    {
+      "epoch": 2.1221407950255387,
+      "grad_norm": 0.5060901595070586,
+      "learning_rate": 5.338016663660681e-06,
+      "loss": 0.0372,
+      "step": 4778
+    },
+    {
+      "epoch": 2.122584943371086,
+      "grad_norm": 0.991711640962585,
+      "learning_rate": 5.336082536389875e-06,
+      "loss": 0.0575,
+      "step": 4779
+    },
+    {
+      "epoch": 2.1230290917166332,
+      "grad_norm": 0.6280177866580242,
+      "learning_rate": 5.334148358600154e-06,
+      "loss": 0.0524,
+      "step": 4780
+    },
+    {
+      "epoch": 2.123473240062181,
+      "grad_norm": 0.38877559204447126,
+      "learning_rate": 5.332214130582259e-06,
+      "loss": 0.0252,
+      "step": 4781
+    },
+    {
+      "epoch": 2.1239173884077283,
+      "grad_norm": 0.5153735769890214,
+      "learning_rate": 5.330279852626936e-06,
+      "loss": 0.0399,
+      "step": 4782
+    },
+    {
+      "epoch": 2.1243615367532755,
+      "grad_norm": 0.4835493880223582,
+      "learning_rate": 5.32834552502494e-06,
+      "loss": 0.0398,
+      "step": 4783
+    },
+    {
+      "epoch": 2.124805685098823,
+      "grad_norm": 0.5440156606024599,
+      "learning_rate": 5.326411148067036e-06,
+      "loss": 0.0359,
+      "step": 4784
+    },
+    {
+      "epoch": 2.1252498334443706,
+      "grad_norm": 0.47437512421776024,
+      "learning_rate": 5.324476722043991e-06,
+      "loss": 0.0357,
+      "step": 4785
+    },
+    {
+      "epoch": 2.125693981789918,
+      "grad_norm": 0.4015404790313753,
+      "learning_rate": 5.322542247246583e-06,
+      "loss": 0.0259,
+      "step": 4786
+    },
+    {
+      "epoch": 2.126138130135465,
+      "grad_norm": 0.3327411043647303,
+      "learning_rate": 5.320607723965594e-06,
+      "loss": 0.0288,
+      "step": 4787
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.4112477949684857,
+      "learning_rate": 5.318673152491821e-06,
+      "loss": 0.0306,
+      "step": 4788
+    },
+    {
+      "epoch": 2.12702642682656,
+      "grad_norm": 0.4352341389187628,
+      "learning_rate": 5.316738533116058e-06,
+      "loss": 0.0294,
+      "step": 4789
+    },
+    {
+      "epoch": 2.1274705751721075,
+      "grad_norm": 0.5247331669668354,
+      "learning_rate": 5.314803866129114e-06,
+      "loss": 0.0455,
+      "step": 4790
+    },
+    {
+      "epoch": 2.1279147235176548,
+      "grad_norm": 0.5740961919642005,
+      "learning_rate": 5.3128691518218015e-06,
+      "loss": 0.0313,
+      "step": 4791
+    },
+    {
+      "epoch": 2.1283588718632025,
+      "grad_norm": 0.3931985647767843,
+      "learning_rate": 5.310934390484939e-06,
+      "loss": 0.0303,
+      "step": 4792
+    },
+    {
+      "epoch": 2.12880302020875,
+      "grad_norm": 0.4665172928488037,
+      "learning_rate": 5.308999582409357e-06,
+      "loss": 0.0324,
+      "step": 4793
+    },
+    {
+      "epoch": 2.129247168554297,
+      "grad_norm": 0.394722347425358,
+      "learning_rate": 5.307064727885889e-06,
+      "loss": 0.0386,
+      "step": 4794
+    },
+    {
+      "epoch": 2.1296913168998444,
+      "grad_norm": 0.3880325853085057,
+      "learning_rate": 5.305129827205375e-06,
+      "loss": 0.0305,
+      "step": 4795
+    },
+    {
+      "epoch": 2.130135465245392,
+      "grad_norm": 0.33971233702286885,
+      "learning_rate": 5.303194880658668e-06,
+      "loss": 0.0336,
+      "step": 4796
+    },
+    {
+      "epoch": 2.1305796135909394,
+      "grad_norm": 0.5540350723444334,
+      "learning_rate": 5.301259888536616e-06,
+      "loss": 0.0369,
+      "step": 4797
+    },
+    {
+      "epoch": 2.1310237619364867,
+      "grad_norm": 0.562001973004502,
+      "learning_rate": 5.299324851130086e-06,
+      "loss": 0.0463,
+      "step": 4798
+    },
+    {
+      "epoch": 2.1314679102820344,
+      "grad_norm": 0.40052051170910113,
+      "learning_rate": 5.297389768729949e-06,
+      "loss": 0.0343,
+      "step": 4799
+    },
+    {
+      "epoch": 2.1319120586275817,
+      "grad_norm": 0.37582805444109113,
+      "learning_rate": 5.295454641627076e-06,
+      "loss": 0.0279,
+      "step": 4800
+    },
+    {
+      "epoch": 2.132356206973129,
+      "grad_norm": 0.556183127195747,
+      "learning_rate": 5.293519470112351e-06,
+      "loss": 0.055,
+      "step": 4801
+    },
+    {
+      "epoch": 2.1328003553186763,
+      "grad_norm": 0.43600853079083646,
+      "learning_rate": 5.2915842544766645e-06,
+      "loss": 0.0341,
+      "step": 4802
+    },
+    {
+      "epoch": 2.133244503664224,
+      "grad_norm": 0.43024160171177206,
+      "learning_rate": 5.289648995010912e-06,
+      "loss": 0.0311,
+      "step": 4803
+    },
+    {
+      "epoch": 2.1336886520097713,
+      "grad_norm": 0.4112510401014529,
+      "learning_rate": 5.287713692005993e-06,
+      "loss": 0.0319,
+      "step": 4804
+    },
+    {
+      "epoch": 2.1341328003553186,
+      "grad_norm": 0.4594812201613338,
+      "learning_rate": 5.285778345752821e-06,
+      "loss": 0.0372,
+      "step": 4805
+    },
+    {
+      "epoch": 2.134576948700866,
+      "grad_norm": 0.3657072390383706,
+      "learning_rate": 5.2838429565423074e-06,
+      "loss": 0.0283,
+      "step": 4806
+    },
+    {
+      "epoch": 2.1350210970464136,
+      "grad_norm": 0.6140456097709794,
+      "learning_rate": 5.281907524665377e-06,
+      "loss": 0.0454,
+      "step": 4807
+    },
+    {
+      "epoch": 2.135465245391961,
+      "grad_norm": 0.39045919917216204,
+      "learning_rate": 5.279972050412957e-06,
+      "loss": 0.0378,
+      "step": 4808
+    },
+    {
+      "epoch": 2.135909393737508,
+      "grad_norm": 0.3927097404516126,
+      "learning_rate": 5.278036534075981e-06,
+      "loss": 0.0333,
+      "step": 4809
+    },
+    {
+      "epoch": 2.136353542083056,
+      "grad_norm": 0.5754764647926375,
+      "learning_rate": 5.276100975945393e-06,
+      "loss": 0.0327,
+      "step": 4810
+    },
+    {
+      "epoch": 2.1367976904286032,
+      "grad_norm": 0.9692035662201081,
+      "learning_rate": 5.274165376312136e-06,
+      "loss": 0.0433,
+      "step": 4811
+    },
+    {
+      "epoch": 2.1372418387741505,
+      "grad_norm": 0.3898618853951006,
+      "learning_rate": 5.272229735467166e-06,
+      "loss": 0.0301,
+      "step": 4812
+    },
+    {
+      "epoch": 2.137685987119698,
+      "grad_norm": 0.5041899434970386,
+      "learning_rate": 5.270294053701442e-06,
+      "loss": 0.0293,
+      "step": 4813
+    },
+    {
+      "epoch": 2.1381301354652456,
+      "grad_norm": 0.4402127129653806,
+      "learning_rate": 5.268358331305931e-06,
+      "loss": 0.0321,
+      "step": 4814
+    },
+    {
+      "epoch": 2.138574283810793,
+      "grad_norm": 0.45642467547731175,
+      "learning_rate": 5.266422568571604e-06,
+      "loss": 0.0359,
+      "step": 4815
+    },
+    {
+      "epoch": 2.13901843215634,
+      "grad_norm": 0.3935008736888044,
+      "learning_rate": 5.264486765789439e-06,
+      "loss": 0.034,
+      "step": 4816
+    },
+    {
+      "epoch": 2.1394625805018874,
+      "grad_norm": 0.334289910641262,
+      "learning_rate": 5.262550923250421e-06,
+      "loss": 0.0235,
+      "step": 4817
+    },
+    {
+      "epoch": 2.139906728847435,
+      "grad_norm": 0.45188823216686197,
+      "learning_rate": 5.260615041245538e-06,
+      "loss": 0.0293,
+      "step": 4818
+    },
+    {
+      "epoch": 2.1403508771929824,
+      "grad_norm": 0.5579134180377057,
+      "learning_rate": 5.25867912006579e-06,
+      "loss": 0.0428,
+      "step": 4819
+    },
+    {
+      "epoch": 2.1407950255385297,
+      "grad_norm": 0.33710370134564305,
+      "learning_rate": 5.256743160002174e-06,
+      "loss": 0.0265,
+      "step": 4820
+    },
+    {
+      "epoch": 2.1412391738840775,
+      "grad_norm": 0.5496665050765797,
+      "learning_rate": 5.254807161345699e-06,
+      "loss": 0.0482,
+      "step": 4821
+    },
+    {
+      "epoch": 2.1416833222296248,
+      "grad_norm": 0.41971477565058973,
+      "learning_rate": 5.2528711243873795e-06,
+      "loss": 0.032,
+      "step": 4822
+    },
+    {
+      "epoch": 2.142127470575172,
+      "grad_norm": 0.4604575178968341,
+      "learning_rate": 5.2509350494182365e-06,
+      "loss": 0.0366,
+      "step": 4823
+    },
+    {
+      "epoch": 2.1425716189207193,
+      "grad_norm": 0.3714626261982935,
+      "learning_rate": 5.2489989367292916e-06,
+      "loss": 0.0256,
+      "step": 4824
+    },
+    {
+      "epoch": 2.143015767266267,
+      "grad_norm": 0.40647254628304297,
+      "learning_rate": 5.247062786611575e-06,
+      "loss": 0.034,
+      "step": 4825
+    },
+    {
+      "epoch": 2.1434599156118144,
+      "grad_norm": 0.534798459704081,
+      "learning_rate": 5.245126599356126e-06,
+      "loss": 0.0344,
+      "step": 4826
+    },
+    {
+      "epoch": 2.1439040639573617,
+      "grad_norm": 0.3767967274585948,
+      "learning_rate": 5.243190375253987e-06,
+      "loss": 0.0388,
+      "step": 4827
+    },
+    {
+      "epoch": 2.1443482123029094,
+      "grad_norm": 0.43416732973374494,
+      "learning_rate": 5.241254114596201e-06,
+      "loss": 0.036,
+      "step": 4828
+    },
+    {
+      "epoch": 2.1447923606484567,
+      "grad_norm": 0.37510655690708544,
+      "learning_rate": 5.2393178176738246e-06,
+      "loss": 0.0339,
+      "step": 4829
+    },
+    {
+      "epoch": 2.145236508994004,
+      "grad_norm": 0.509295659153276,
+      "learning_rate": 5.237381484777914e-06,
+      "loss": 0.0449,
+      "step": 4830
+    },
+    {
+      "epoch": 2.1456806573395513,
+      "grad_norm": 0.43018107465023353,
+      "learning_rate": 5.235445116199536e-06,
+      "loss": 0.0386,
+      "step": 4831
+    },
+    {
+      "epoch": 2.146124805685099,
+      "grad_norm": 0.532799480118876,
+      "learning_rate": 5.2335087122297545e-06,
+      "loss": 0.0536,
+      "step": 4832
+    },
+    {
+      "epoch": 2.1465689540306463,
+      "grad_norm": 0.3898351411048676,
+      "learning_rate": 5.231572273159649e-06,
+      "loss": 0.0328,
+      "step": 4833
+    },
+    {
+      "epoch": 2.1470131023761936,
+      "grad_norm": 0.34769775883890974,
+      "learning_rate": 5.229635799280298e-06,
+      "loss": 0.0297,
+      "step": 4834
+    },
+    {
+      "epoch": 2.147457250721741,
+      "grad_norm": 0.4711513587854737,
+      "learning_rate": 5.2276992908827825e-06,
+      "loss": 0.0422,
+      "step": 4835
+    },
+    {
+      "epoch": 2.1479013990672886,
+      "grad_norm": 0.405896587319854,
+      "learning_rate": 5.2257627482581985e-06,
+      "loss": 0.033,
+      "step": 4836
+    },
+    {
+      "epoch": 2.148345547412836,
+      "grad_norm": 0.3745089869031137,
+      "learning_rate": 5.2238261716976375e-06,
+      "loss": 0.0246,
+      "step": 4837
+    },
+    {
+      "epoch": 2.148789695758383,
+      "grad_norm": 0.40688545389600017,
+      "learning_rate": 5.2218895614922e-06,
+      "loss": 0.044,
+      "step": 4838
+    },
+    {
+      "epoch": 2.1492338441039305,
+      "grad_norm": 0.48659815613858365,
+      "learning_rate": 5.219952917932993e-06,
+      "loss": 0.0412,
+      "step": 4839
+    },
+    {
+      "epoch": 2.149677992449478,
+      "grad_norm": 0.5300858396288847,
+      "learning_rate": 5.218016241311126e-06,
+      "loss": 0.0441,
+      "step": 4840
+    },
+    {
+      "epoch": 2.1501221407950255,
+      "grad_norm": 0.5970494126272853,
+      "learning_rate": 5.216079531917714e-06,
+      "loss": 0.0377,
+      "step": 4841
+    },
+    {
+      "epoch": 2.150566289140573,
+      "grad_norm": 0.4023665494232216,
+      "learning_rate": 5.2141427900438765e-06,
+      "loss": 0.0261,
+      "step": 4842
+    },
+    {
+      "epoch": 2.1510104374861205,
+      "grad_norm": 0.3360515019130085,
+      "learning_rate": 5.212206015980742e-06,
+      "loss": 0.0232,
+      "step": 4843
+    },
+    {
+      "epoch": 2.151454585831668,
+      "grad_norm": 0.37156982937954675,
+      "learning_rate": 5.210269210019438e-06,
+      "loss": 0.0313,
+      "step": 4844
+    },
+    {
+      "epoch": 2.151898734177215,
+      "grad_norm": 0.43881581226491895,
+      "learning_rate": 5.2083323724511e-06,
+      "loss": 0.0387,
+      "step": 4845
+    },
+    {
+      "epoch": 2.1523428825227624,
+      "grad_norm": 0.5133014074927194,
+      "learning_rate": 5.206395503566867e-06,
+      "loss": 0.033,
+      "step": 4846
+    },
+    {
+      "epoch": 2.15278703086831,
+      "grad_norm": 0.7393084909457972,
+      "learning_rate": 5.204458603657885e-06,
+      "loss": 0.0451,
+      "step": 4847
+    },
+    {
+      "epoch": 2.1532311792138574,
+      "grad_norm": 0.5276996020099548,
+      "learning_rate": 5.2025216730153016e-06,
+      "loss": 0.0244,
+      "step": 4848
+    },
+    {
+      "epoch": 2.1536753275594047,
+      "grad_norm": 0.47830343576785345,
+      "learning_rate": 5.200584711930267e-06,
+      "loss": 0.0387,
+      "step": 4849
+    },
+    {
+      "epoch": 2.1541194759049525,
+      "grad_norm": 0.5578861068890496,
+      "learning_rate": 5.198647720693948e-06,
+      "loss": 0.0421,
+      "step": 4850
+    },
+    {
+      "epoch": 2.1545636242504997,
+      "grad_norm": 0.3280166182279637,
+      "learning_rate": 5.1967106995975e-06,
+      "loss": 0.0233,
+      "step": 4851
+    },
+    {
+      "epoch": 2.155007772596047,
+      "grad_norm": 0.5452793800296186,
+      "learning_rate": 5.194773648932092e-06,
+      "loss": 0.0403,
+      "step": 4852
+    },
+    {
+      "epoch": 2.1554519209415943,
+      "grad_norm": 0.5130821544913388,
+      "learning_rate": 5.192836568988895e-06,
+      "loss": 0.0307,
+      "step": 4853
+    },
+    {
+      "epoch": 2.155896069287142,
+      "grad_norm": 0.33770672647996586,
+      "learning_rate": 5.190899460059088e-06,
+      "loss": 0.0232,
+      "step": 4854
+    },
+    {
+      "epoch": 2.1563402176326893,
+      "grad_norm": 0.6783231395478261,
+      "learning_rate": 5.188962322433848e-06,
+      "loss": 0.0359,
+      "step": 4855
+    },
+    {
+      "epoch": 2.1567843659782366,
+      "grad_norm": 0.49312348081568724,
+      "learning_rate": 5.187025156404361e-06,
+      "loss": 0.051,
+      "step": 4856
+    },
+    {
+      "epoch": 2.1572285143237844,
+      "grad_norm": 0.5188160057025217,
+      "learning_rate": 5.185087962261817e-06,
+      "loss": 0.037,
+      "step": 4857
+    },
+    {
+      "epoch": 2.1576726626693317,
+      "grad_norm": 0.4979475713257613,
+      "learning_rate": 5.183150740297407e-06,
+      "loss": 0.0457,
+      "step": 4858
+    },
+    {
+      "epoch": 2.158116811014879,
+      "grad_norm": 0.8466081637820392,
+      "learning_rate": 5.181213490802329e-06,
+      "loss": 0.0467,
+      "step": 4859
+    },
+    {
+      "epoch": 2.1585609593604262,
+      "grad_norm": 0.4708804574175826,
+      "learning_rate": 5.179276214067788e-06,
+      "loss": 0.0404,
+      "step": 4860
+    },
+    {
+      "epoch": 2.159005107705974,
+      "grad_norm": 0.3951062024143596,
+      "learning_rate": 5.1773389103849835e-06,
+      "loss": 0.0262,
+      "step": 4861
+    },
+    {
+      "epoch": 2.1594492560515213,
+      "grad_norm": 0.5862820412725568,
+      "learning_rate": 5.175401580045131e-06,
+      "loss": 0.0471,
+      "step": 4862
+    },
+    {
+      "epoch": 2.1598934043970686,
+      "grad_norm": 0.3583235113891142,
+      "learning_rate": 5.173464223339438e-06,
+      "loss": 0.0261,
+      "step": 4863
+    },
+    {
+      "epoch": 2.160337552742616,
+      "grad_norm": 0.7139304922452345,
+      "learning_rate": 5.171526840559129e-06,
+      "loss": 0.0581,
+      "step": 4864
+    },
+    {
+      "epoch": 2.1607817010881636,
+      "grad_norm": 0.41880853503590465,
+      "learning_rate": 5.169589431995421e-06,
+      "loss": 0.0305,
+      "step": 4865
+    },
+    {
+      "epoch": 2.161225849433711,
+      "grad_norm": 0.974797863480534,
+      "learning_rate": 5.16765199793954e-06,
+      "loss": 0.0476,
+      "step": 4866
+    },
+    {
+      "epoch": 2.161669997779258,
+      "grad_norm": 0.6058306397735719,
+      "learning_rate": 5.165714538682716e-06,
+      "loss": 0.0603,
+      "step": 4867
+    },
+    {
+      "epoch": 2.1621141461248055,
+      "grad_norm": 0.6017608849107138,
+      "learning_rate": 5.163777054516182e-06,
+      "loss": 0.0415,
+      "step": 4868
+    },
+    {
+      "epoch": 2.162558294470353,
+      "grad_norm": 1.5666330014886598,
+      "learning_rate": 5.161839545731175e-06,
+      "loss": 0.0424,
+      "step": 4869
+    },
+    {
+      "epoch": 2.1630024428159005,
+      "grad_norm": 0.4625370484241095,
+      "learning_rate": 5.159902012618933e-06,
+      "loss": 0.0337,
+      "step": 4870
+    },
+    {
+      "epoch": 2.1634465911614478,
+      "grad_norm": 0.6306574373250589,
+      "learning_rate": 5.1579644554707054e-06,
+      "loss": 0.057,
+      "step": 4871
+    },
+    {
+      "epoch": 2.1638907395069955,
+      "grad_norm": 0.9235120056329936,
+      "learning_rate": 5.156026874577735e-06,
+      "loss": 0.0561,
+      "step": 4872
+    },
+    {
+      "epoch": 2.164334887852543,
+      "grad_norm": 0.4148788359576288,
+      "learning_rate": 5.154089270231275e-06,
+      "loss": 0.0365,
+      "step": 4873
+    },
+    {
+      "epoch": 2.16477903619809,
+      "grad_norm": 0.5384050767235271,
+      "learning_rate": 5.152151642722582e-06,
+      "loss": 0.0397,
+      "step": 4874
+    },
+    {
+      "epoch": 2.1652231845436374,
+      "grad_norm": 0.42398179308700884,
+      "learning_rate": 5.15021399234291e-06,
+      "loss": 0.0244,
+      "step": 4875
+    },
+    {
+      "epoch": 2.165667332889185,
+      "grad_norm": 0.4428879862840884,
+      "learning_rate": 5.148276319383525e-06,
+      "loss": 0.0338,
+      "step": 4876
+    },
+    {
+      "epoch": 2.1661114812347324,
+      "grad_norm": 0.5151742619891833,
+      "learning_rate": 5.146338624135689e-06,
+      "loss": 0.0429,
+      "step": 4877
+    },
+    {
+      "epoch": 2.1665556295802797,
+      "grad_norm": 0.5906956484118873,
+      "learning_rate": 5.144400906890672e-06,
+      "loss": 0.0486,
+      "step": 4878
+    },
+    {
+      "epoch": 2.1669997779258274,
+      "grad_norm": 0.5089563089876176,
+      "learning_rate": 5.142463167939748e-06,
+      "loss": 0.0336,
+      "step": 4879
+    },
+    {
+      "epoch": 2.1674439262713747,
+      "grad_norm": 0.4990120356759535,
+      "learning_rate": 5.140525407574187e-06,
+      "loss": 0.0397,
+      "step": 4880
+    },
+    {
+      "epoch": 2.167888074616922,
+      "grad_norm": 0.47055574764670566,
+      "learning_rate": 5.138587626085271e-06,
+      "loss": 0.0411,
+      "step": 4881
+    },
+    {
+      "epoch": 2.1683322229624693,
+      "grad_norm": 0.49103425451011984,
+      "learning_rate": 5.136649823764281e-06,
+      "loss": 0.0422,
+      "step": 4882
+    },
+    {
+      "epoch": 2.168776371308017,
+      "grad_norm": 0.44674321466653016,
+      "learning_rate": 5.1347120009025005e-06,
+      "loss": 0.0312,
+      "step": 4883
+    },
+    {
+      "epoch": 2.1692205196535643,
+      "grad_norm": 0.2849178715545674,
+      "learning_rate": 5.132774157791218e-06,
+      "loss": 0.0226,
+      "step": 4884
+    },
+    {
+      "epoch": 2.1696646679991116,
+      "grad_norm": 0.46152868939886915,
+      "learning_rate": 5.130836294721726e-06,
+      "loss": 0.0364,
+      "step": 4885
+    },
+    {
+      "epoch": 2.170108816344659,
+      "grad_norm": 0.43325439116729864,
+      "learning_rate": 5.128898411985315e-06,
+      "loss": 0.0288,
+      "step": 4886
+    },
+    {
+      "epoch": 2.1705529646902066,
+      "grad_norm": 0.39479226787289406,
+      "learning_rate": 5.1269605098732825e-06,
+      "loss": 0.0251,
+      "step": 4887
+    },
+    {
+      "epoch": 2.170997113035754,
+      "grad_norm": 0.41019425261041265,
+      "learning_rate": 5.12502258867693e-06,
+      "loss": 0.031,
+      "step": 4888
+    },
+    {
+      "epoch": 2.1714412613813012,
+      "grad_norm": 0.508850109174841,
+      "learning_rate": 5.123084648687557e-06,
+      "loss": 0.0474,
+      "step": 4889
+    },
+    {
+      "epoch": 2.171885409726849,
+      "grad_norm": 0.42681095953850656,
+      "learning_rate": 5.121146690196472e-06,
+      "loss": 0.0244,
+      "step": 4890
+    },
+    {
+      "epoch": 2.1723295580723962,
+      "grad_norm": 0.32346458488317453,
+      "learning_rate": 5.1192087134949804e-06,
+      "loss": 0.0264,
+      "step": 4891
+    },
+    {
+      "epoch": 2.1727737064179435,
+      "grad_norm": 0.40304790268945045,
+      "learning_rate": 5.1172707188743955e-06,
+      "loss": 0.0264,
+      "step": 4892
+    },
+    {
+      "epoch": 2.173217854763491,
+      "grad_norm": 0.5433823527011491,
+      "learning_rate": 5.115332706626028e-06,
+      "loss": 0.0459,
+      "step": 4893
+    },
+    {
+      "epoch": 2.1736620031090386,
+      "grad_norm": 0.33519824313183216,
+      "learning_rate": 5.113394677041197e-06,
+      "loss": 0.0289,
+      "step": 4894
+    },
+    {
+      "epoch": 2.174106151454586,
+      "grad_norm": 0.399864916190364,
+      "learning_rate": 5.111456630411218e-06,
+      "loss": 0.026,
+      "step": 4895
+    },
+    {
+      "epoch": 2.174550299800133,
+      "grad_norm": 0.49817314831150256,
+      "learning_rate": 5.109518567027416e-06,
+      "loss": 0.0341,
+      "step": 4896
+    },
+    {
+      "epoch": 2.1749944481456804,
+      "grad_norm": 0.41814385729810516,
+      "learning_rate": 5.107580487181112e-06,
+      "loss": 0.0315,
+      "step": 4897
+    },
+    {
+      "epoch": 2.175438596491228,
+      "grad_norm": 0.44895548729597046,
+      "learning_rate": 5.105642391163633e-06,
+      "loss": 0.0403,
+      "step": 4898
+    },
+    {
+      "epoch": 2.1758827448367755,
+      "grad_norm": 0.3688472653021723,
+      "learning_rate": 5.10370427926631e-06,
+      "loss": 0.0345,
+      "step": 4899
+    },
+    {
+      "epoch": 2.1763268931823228,
+      "grad_norm": 0.4629343116213356,
+      "learning_rate": 5.1017661517804694e-06,
+      "loss": 0.0353,
+      "step": 4900
+    },
+    {
+      "epoch": 2.1767710415278705,
+      "grad_norm": 0.4559917864661212,
+      "learning_rate": 5.099828008997448e-06,
+      "loss": 0.0398,
+      "step": 4901
+    },
+    {
+      "epoch": 2.1772151898734178,
+      "grad_norm": 0.45480161362407634,
+      "learning_rate": 5.097889851208583e-06,
+      "loss": 0.035,
+      "step": 4902
+    },
+    {
+      "epoch": 2.177659338218965,
+      "grad_norm": 0.42473449296712756,
+      "learning_rate": 5.0959516787052085e-06,
+      "loss": 0.0317,
+      "step": 4903
+    },
+    {
+      "epoch": 2.1781034865645124,
+      "grad_norm": 0.4952860166933172,
+      "learning_rate": 5.094013491778668e-06,
+      "loss": 0.0443,
+      "step": 4904
+    },
+    {
+      "epoch": 2.17854763491006,
+      "grad_norm": 0.4001101623153066,
+      "learning_rate": 5.092075290720302e-06,
+      "loss": 0.0298,
+      "step": 4905
+    },
+    {
+      "epoch": 2.1789917832556074,
+      "grad_norm": 0.5225430247317463,
+      "learning_rate": 5.0901370758214565e-06,
+      "loss": 0.0398,
+      "step": 4906
+    },
+    {
+      "epoch": 2.1794359316011547,
+      "grad_norm": 0.6146833193819635,
+      "learning_rate": 5.088198847373477e-06,
+      "loss": 0.0464,
+      "step": 4907
+    },
+    {
+      "epoch": 2.1798800799467024,
+      "grad_norm": 0.47802919024440194,
+      "learning_rate": 5.086260605667712e-06,
+      "loss": 0.037,
+      "step": 4908
+    },
+    {
+      "epoch": 2.1803242282922497,
+      "grad_norm": 0.5368694025591854,
+      "learning_rate": 5.084322350995512e-06,
+      "loss": 0.0488,
+      "step": 4909
+    },
+    {
+      "epoch": 2.180768376637797,
+      "grad_norm": 0.32415608827454506,
+      "learning_rate": 5.0823840836482316e-06,
+      "loss": 0.0233,
+      "step": 4910
+    },
+    {
+      "epoch": 2.1812125249833443,
+      "grad_norm": 0.4761380556003533,
+      "learning_rate": 5.080445803917225e-06,
+      "loss": 0.0471,
+      "step": 4911
+    },
+    {
+      "epoch": 2.181656673328892,
+      "grad_norm": 0.3515304190735633,
+      "learning_rate": 5.078507512093844e-06,
+      "loss": 0.0259,
+      "step": 4912
+    },
+    {
+      "epoch": 2.1821008216744393,
+      "grad_norm": 0.6155251528359741,
+      "learning_rate": 5.076569208469454e-06,
+      "loss": 0.0474,
+      "step": 4913
+    },
+    {
+      "epoch": 2.1825449700199866,
+      "grad_norm": 0.4580217790122366,
+      "learning_rate": 5.0746308933354105e-06,
+      "loss": 0.0438,
+      "step": 4914
+    },
+    {
+      "epoch": 2.182989118365534,
+      "grad_norm": 0.408141050974567,
+      "learning_rate": 5.072692566983074e-06,
+      "loss": 0.0376,
+      "step": 4915
+    },
+    {
+      "epoch": 2.1834332667110816,
+      "grad_norm": 0.4622430926137514,
+      "learning_rate": 5.070754229703811e-06,
+      "loss": 0.0325,
+      "step": 4916
+    },
+    {
+      "epoch": 2.183877415056629,
+      "grad_norm": 0.49891927431787836,
+      "learning_rate": 5.068815881788986e-06,
+      "loss": 0.0383,
+      "step": 4917
+    },
+    {
+      "epoch": 2.184321563402176,
+      "grad_norm": 0.3944468340370449,
+      "learning_rate": 5.0668775235299636e-06,
+      "loss": 0.0364,
+      "step": 4918
+    },
+    {
+      "epoch": 2.184765711747724,
+      "grad_norm": 0.5143753892518631,
+      "learning_rate": 5.064939155218115e-06,
+      "loss": 0.0327,
+      "step": 4919
+    },
+    {
+      "epoch": 2.1852098600932712,
+      "grad_norm": 0.4613419832071711,
+      "learning_rate": 5.0630007771448064e-06,
+      "loss": 0.0365,
+      "step": 4920
+    },
+    {
+      "epoch": 2.1856540084388185,
+      "grad_norm": 0.42353949789442463,
+      "learning_rate": 5.061062389601413e-06,
+      "loss": 0.0305,
+      "step": 4921
+    },
+    {
+      "epoch": 2.186098156784366,
+      "grad_norm": 0.4615529011211128,
+      "learning_rate": 5.059123992879303e-06,
+      "loss": 0.0363,
+      "step": 4922
+    },
+    {
+      "epoch": 2.1865423051299135,
+      "grad_norm": 0.38154251403565953,
+      "learning_rate": 5.057185587269854e-06,
+      "loss": 0.0237,
+      "step": 4923
+    },
+    {
+      "epoch": 2.186986453475461,
+      "grad_norm": 0.45850860345485767,
+      "learning_rate": 5.05524717306444e-06,
+      "loss": 0.0309,
+      "step": 4924
+    },
+    {
+      "epoch": 2.187430601821008,
+      "grad_norm": 0.3296860610369256,
+      "learning_rate": 5.053308750554437e-06,
+      "loss": 0.0311,
+      "step": 4925
+    },
+    {
+      "epoch": 2.1878747501665554,
+      "grad_norm": 0.3804383024783399,
+      "learning_rate": 5.051370320031221e-06,
+      "loss": 0.0337,
+      "step": 4926
+    },
+    {
+      "epoch": 2.188318898512103,
+      "grad_norm": 0.44377911056623215,
+      "learning_rate": 5.049431881786176e-06,
+      "loss": 0.0451,
+      "step": 4927
+    },
+    {
+      "epoch": 2.1887630468576504,
+      "grad_norm": 0.3897615063112148,
+      "learning_rate": 5.04749343611068e-06,
+      "loss": 0.0264,
+      "step": 4928
+    },
+    {
+      "epoch": 2.1892071952031977,
+      "grad_norm": 0.857708991433585,
+      "learning_rate": 5.045554983296111e-06,
+      "loss": 0.0421,
+      "step": 4929
+    },
+    {
+      "epoch": 2.1896513435487455,
+      "grad_norm": 0.4460864492390492,
+      "learning_rate": 5.043616523633856e-06,
+      "loss": 0.0348,
+      "step": 4930
+    },
+    {
+      "epoch": 2.1900954918942928,
+      "grad_norm": 0.46033992290883,
+      "learning_rate": 5.0416780574152976e-06,
+      "loss": 0.0391,
+      "step": 4931
+    },
+    {
+      "epoch": 2.19053964023984,
+      "grad_norm": 0.36797085648755834,
+      "learning_rate": 5.0397395849318165e-06,
+      "loss": 0.0379,
+      "step": 4932
+    },
+    {
+      "epoch": 2.1909837885853873,
+      "grad_norm": 0.3675945320964035,
+      "learning_rate": 5.0378011064748025e-06,
+      "loss": 0.0252,
+      "step": 4933
+    },
+    {
+      "epoch": 2.191427936930935,
+      "grad_norm": 0.6163310761797228,
+      "learning_rate": 5.035862622335641e-06,
+      "loss": 0.0477,
+      "step": 4934
+    },
+    {
+      "epoch": 2.1918720852764824,
+      "grad_norm": 0.5755380695413213,
+      "learning_rate": 5.0339241328057164e-06,
+      "loss": 0.0321,
+      "step": 4935
+    },
+    {
+      "epoch": 2.1923162336220297,
+      "grad_norm": 0.7365554315970313,
+      "learning_rate": 5.0319856381764175e-06,
+      "loss": 0.0421,
+      "step": 4936
+    },
+    {
+      "epoch": 2.1927603819675774,
+      "grad_norm": 0.3786316850609603,
+      "learning_rate": 5.030047138739136e-06,
+      "loss": 0.0274,
+      "step": 4937
+    },
+    {
+      "epoch": 2.1932045303131247,
+      "grad_norm": 0.46503560614578676,
+      "learning_rate": 5.028108634785258e-06,
+      "loss": 0.0329,
+      "step": 4938
+    },
+    {
+      "epoch": 2.193648678658672,
+      "grad_norm": 0.39835692305695053,
+      "learning_rate": 5.0261701266061746e-06,
+      "loss": 0.0267,
+      "step": 4939
+    },
+    {
+      "epoch": 2.1940928270042193,
+      "grad_norm": 0.3883823936267887,
+      "learning_rate": 5.024231614493277e-06,
+      "loss": 0.0284,
+      "step": 4940
+    },
+    {
+      "epoch": 2.194536975349767,
+      "grad_norm": 0.38466420585767425,
+      "learning_rate": 5.022293098737957e-06,
+      "loss": 0.0302,
+      "step": 4941
+    },
+    {
+      "epoch": 2.1949811236953143,
+      "grad_norm": 0.5034584168375414,
+      "learning_rate": 5.0203545796316044e-06,
+      "loss": 0.05,
+      "step": 4942
+    },
+    {
+      "epoch": 2.1954252720408616,
+      "grad_norm": 0.5196054438026021,
+      "learning_rate": 5.0184160574656125e-06,
+      "loss": 0.0348,
+      "step": 4943
+    },
+    {
+      "epoch": 2.195869420386409,
+      "grad_norm": 0.4645537912636916,
+      "learning_rate": 5.0164775325313755e-06,
+      "loss": 0.0376,
+      "step": 4944
+    },
+    {
+      "epoch": 2.1963135687319566,
+      "grad_norm": 0.4894769238899238,
+      "learning_rate": 5.0145390051202846e-06,
+      "loss": 0.0357,
+      "step": 4945
+    },
+    {
+      "epoch": 2.196757717077504,
+      "grad_norm": 0.3225076116998955,
+      "learning_rate": 5.012600475523733e-06,
+      "loss": 0.0268,
+      "step": 4946
+    },
+    {
+      "epoch": 2.197201865423051,
+      "grad_norm": 0.46007489982886224,
+      "learning_rate": 5.010661944033118e-06,
+      "loss": 0.0406,
+      "step": 4947
+    },
+    {
+      "epoch": 2.197646013768599,
+      "grad_norm": 0.6157542604470476,
+      "learning_rate": 5.008723410939832e-06,
+      "loss": 0.0329,
+      "step": 4948
+    },
+    {
+      "epoch": 2.198090162114146,
+      "grad_norm": 0.41686085632858955,
+      "learning_rate": 5.006784876535268e-06,
+      "loss": 0.0416,
+      "step": 4949
+    },
+    {
+      "epoch": 2.1985343104596935,
+      "grad_norm": 0.5468745101745486,
+      "learning_rate": 5.004846341110822e-06,
+      "loss": 0.0499,
+      "step": 4950
+    },
+    {
+      "epoch": 2.198978458805241,
+      "grad_norm": 0.43554432758066686,
+      "learning_rate": 5.002907804957889e-06,
+      "loss": 0.0353,
+      "step": 4951
+    },
+    {
+      "epoch": 2.1994226071507885,
+      "grad_norm": 0.4925752472694445,
+      "learning_rate": 5.000969268367862e-06,
+      "loss": 0.0367,
+      "step": 4952
+    },
+    {
+      "epoch": 2.199866755496336,
+      "grad_norm": 0.4262017675003274,
+      "learning_rate": 4.999030731632139e-06,
+      "loss": 0.0444,
+      "step": 4953
+    },
+    {
+      "epoch": 2.200310903841883,
+      "grad_norm": 0.3899535716527509,
+      "learning_rate": 4.997092195042113e-06,
+      "loss": 0.0299,
+      "step": 4954
+    },
+    {
+      "epoch": 2.2007550521874304,
+      "grad_norm": 0.4534373375430896,
+      "learning_rate": 4.995153658889181e-06,
+      "loss": 0.0346,
+      "step": 4955
+    },
+    {
+      "epoch": 2.201199200532978,
+      "grad_norm": 0.32786374874254637,
+      "learning_rate": 4.993215123464734e-06,
+      "loss": 0.0292,
+      "step": 4956
+    },
+    {
+      "epoch": 2.2016433488785254,
+      "grad_norm": 0.48869610889723925,
+      "learning_rate": 4.991276589060169e-06,
+      "loss": 0.0435,
+      "step": 4957
+    },
+    {
+      "epoch": 2.2020874972240727,
+      "grad_norm": 0.41141706707817044,
+      "learning_rate": 4.989338055966883e-06,
+      "loss": 0.0317,
+      "step": 4958
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.35799145890163087,
+      "learning_rate": 4.987399524476268e-06,
+      "loss": 0.0343,
+      "step": 4959
+    },
+    {
+      "epoch": 2.2029757939151677,
+      "grad_norm": 0.39541269197255546,
+      "learning_rate": 4.985460994879717e-06,
+      "loss": 0.0303,
+      "step": 4960
+    },
+    {
+      "epoch": 2.203419942260715,
+      "grad_norm": 0.4446866574845067,
+      "learning_rate": 4.983522467468627e-06,
+      "loss": 0.0458,
+      "step": 4961
+    },
+    {
+      "epoch": 2.2038640906062623,
+      "grad_norm": 0.44841266243980166,
+      "learning_rate": 4.981583942534388e-06,
+      "loss": 0.0401,
+      "step": 4962
+    },
+    {
+      "epoch": 2.20430823895181,
+      "grad_norm": 0.40609549354200053,
+      "learning_rate": 4.979645420368397e-06,
+      "loss": 0.0318,
+      "step": 4963
+    },
+    {
+      "epoch": 2.2047523872973573,
+      "grad_norm": 0.4655683020461079,
+      "learning_rate": 4.977706901262045e-06,
+      "loss": 0.0318,
+      "step": 4964
+    },
+    {
+      "epoch": 2.2051965356429046,
+      "grad_norm": 0.4105935741895686,
+      "learning_rate": 4.975768385506725e-06,
+      "loss": 0.0365,
+      "step": 4965
+    },
+    {
+      "epoch": 2.2056406839884524,
+      "grad_norm": 0.4840996358141117,
+      "learning_rate": 4.973829873393827e-06,
+      "loss": 0.0288,
+      "step": 4966
+    },
+    {
+      "epoch": 2.2060848323339997,
+      "grad_norm": 0.43687024211685477,
+      "learning_rate": 4.971891365214743e-06,
+      "loss": 0.0284,
+      "step": 4967
+    },
+    {
+      "epoch": 2.206528980679547,
+      "grad_norm": 0.5919063970733193,
+      "learning_rate": 4.969952861260865e-06,
+      "loss": 0.0316,
+      "step": 4968
+    },
+    {
+      "epoch": 2.2069731290250942,
+      "grad_norm": 0.34222056631687486,
+      "learning_rate": 4.968014361823583e-06,
+      "loss": 0.0269,
+      "step": 4969
+    },
+    {
+      "epoch": 2.207417277370642,
+      "grad_norm": 0.577057583701676,
+      "learning_rate": 4.966075867194285e-06,
+      "loss": 0.0391,
+      "step": 4970
+    },
+    {
+      "epoch": 2.2078614257161893,
+      "grad_norm": 0.45130320193766277,
+      "learning_rate": 4.964137377664362e-06,
+      "loss": 0.0358,
+      "step": 4971
+    },
+    {
+      "epoch": 2.2083055740617366,
+      "grad_norm": 0.37843493642246295,
+      "learning_rate": 4.9621988935252e-06,
+      "loss": 0.0356,
+      "step": 4972
+    },
+    {
+      "epoch": 2.208749722407284,
+      "grad_norm": 0.37308118817776387,
+      "learning_rate": 4.9602604150681835e-06,
+      "loss": 0.0244,
+      "step": 4973
+    },
+    {
+      "epoch": 2.2091938707528316,
+      "grad_norm": 0.5336868565377078,
+      "learning_rate": 4.958321942584703e-06,
+      "loss": 0.0488,
+      "step": 4974
+    },
+    {
+      "epoch": 2.209638019098379,
+      "grad_norm": 0.3742330589994002,
+      "learning_rate": 4.956383476366145e-06,
+      "loss": 0.0309,
+      "step": 4975
+    },
+    {
+      "epoch": 2.210082167443926,
+      "grad_norm": 0.46020603183712594,
+      "learning_rate": 4.95444501670389e-06,
+      "loss": 0.0323,
+      "step": 4976
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.8475632066917597,
+      "learning_rate": 4.9525065638893226e-06,
+      "loss": 0.0465,
+      "step": 4977
+    },
+    {
+      "epoch": 2.210970464135021,
+      "grad_norm": 0.5219077605882717,
+      "learning_rate": 4.950568118213825e-06,
+      "loss": 0.0393,
+      "step": 4978
+    },
+    {
+      "epoch": 2.2114146124805685,
+      "grad_norm": 0.37860034305038076,
+      "learning_rate": 4.948629679968778e-06,
+      "loss": 0.0366,
+      "step": 4979
+    },
+    {
+      "epoch": 2.2118587608261158,
+      "grad_norm": 0.6128046042996554,
+      "learning_rate": 4.946691249445565e-06,
+      "loss": 0.04,
+      "step": 4980
+    },
+    {
+      "epoch": 2.2123029091716635,
+      "grad_norm": 0.4839071616267329,
+      "learning_rate": 4.944752826935562e-06,
+      "loss": 0.035,
+      "step": 4981
+    },
+    {
+      "epoch": 2.212747057517211,
+      "grad_norm": 0.45065394012575055,
+      "learning_rate": 4.942814412730147e-06,
+      "loss": 0.0267,
+      "step": 4982
+    },
+    {
+      "epoch": 2.213191205862758,
+      "grad_norm": 0.5986492760384308,
+      "learning_rate": 4.940876007120699e-06,
+      "loss": 0.0289,
+      "step": 4983
+    },
+    {
+      "epoch": 2.2136353542083054,
+      "grad_norm": 0.5082807449831283,
+      "learning_rate": 4.938937610398588e-06,
+      "loss": 0.034,
+      "step": 4984
+    },
+    {
+      "epoch": 2.214079502553853,
+      "grad_norm": 0.46534139162341914,
+      "learning_rate": 4.9369992228551935e-06,
+      "loss": 0.0323,
+      "step": 4985
+    },
+    {
+      "epoch": 2.2145236508994004,
+      "grad_norm": 0.4943061938684834,
+      "learning_rate": 4.935060844781886e-06,
+      "loss": 0.0353,
+      "step": 4986
+    },
+    {
+      "epoch": 2.2149677992449477,
+      "grad_norm": 0.4336318300801724,
+      "learning_rate": 4.933122476470038e-06,
+      "loss": 0.0275,
+      "step": 4987
+    },
+    {
+      "epoch": 2.2154119475904954,
+      "grad_norm": 0.4360854695437589,
+      "learning_rate": 4.931184118211016e-06,
+      "loss": 0.0349,
+      "step": 4988
+    },
+    {
+      "epoch": 2.2158560959360427,
+      "grad_norm": 0.6620193372024554,
+      "learning_rate": 4.929245770296191e-06,
+      "loss": 0.0483,
+      "step": 4989
+    },
+    {
+      "epoch": 2.21630024428159,
+      "grad_norm": 0.48994703833403463,
+      "learning_rate": 4.927307433016927e-06,
+      "loss": 0.0432,
+      "step": 4990
+    },
+    {
+      "epoch": 2.2167443926271373,
+      "grad_norm": 0.4622061500888628,
+      "learning_rate": 4.925369106664591e-06,
+      "loss": 0.0387,
+      "step": 4991
+    },
+    {
+      "epoch": 2.217188540972685,
+      "grad_norm": 0.4624898194765989,
+      "learning_rate": 4.923430791530547e-06,
+      "loss": 0.0196,
+      "step": 4992
+    },
+    {
+      "epoch": 2.2176326893182323,
+      "grad_norm": 0.3453109472117995,
+      "learning_rate": 4.9214924879061565e-06,
+      "loss": 0.0286,
+      "step": 4993
+    },
+    {
+      "epoch": 2.2180768376637796,
+      "grad_norm": 0.41871877730955476,
+      "learning_rate": 4.919554196082778e-06,
+      "loss": 0.0397,
+      "step": 4994
+    },
+    {
+      "epoch": 2.2185209860093273,
+      "grad_norm": 0.42587370028059823,
+      "learning_rate": 4.91761591635177e-06,
+      "loss": 0.0408,
+      "step": 4995
+    },
+    {
+      "epoch": 2.2189651343548746,
+      "grad_norm": 0.36571656649505513,
+      "learning_rate": 4.9156776490044875e-06,
+      "loss": 0.0335,
+      "step": 4996
+    },
+    {
+      "epoch": 2.219409282700422,
+      "grad_norm": 0.4229412174299105,
+      "learning_rate": 4.91373939433229e-06,
+      "loss": 0.0396,
+      "step": 4997
+    },
+    {
+      "epoch": 2.219853431045969,
+      "grad_norm": 0.3904502538208118,
+      "learning_rate": 4.911801152626525e-06,
+      "loss": 0.0352,
+      "step": 4998
+    },
+    {
+      "epoch": 2.220297579391517,
+      "grad_norm": 0.7376575867139856,
+      "learning_rate": 4.909862924178545e-06,
+      "loss": 0.04,
+      "step": 4999
+    },
+    {
+      "epoch": 2.2207417277370642,
+      "grad_norm": 0.45886434872272697,
+      "learning_rate": 4.9079247092797e-06,
+      "loss": 0.0375,
+      "step": 5000
+    },
+    {
+      "epoch": 2.2211858760826115,
+      "grad_norm": 0.38398725710621795,
+      "learning_rate": 4.905986508221333e-06,
+      "loss": 0.034,
+      "step": 5001
+    },
+    {
+      "epoch": 2.221630024428159,
+      "grad_norm": 0.4363686982956202,
+      "learning_rate": 4.904048321294791e-06,
+      "loss": 0.0395,
+      "step": 5002
+    },
+    {
+      "epoch": 2.2220741727737066,
+      "grad_norm": 0.47908384837669443,
+      "learning_rate": 4.9021101487914185e-06,
+      "loss": 0.0451,
+      "step": 5003
+    },
+    {
+      "epoch": 2.222518321119254,
+      "grad_norm": 0.49404577583056375,
+      "learning_rate": 4.900171991002553e-06,
+      "loss": 0.0382,
+      "step": 5004
+    },
+    {
+      "epoch": 2.222962469464801,
+      "grad_norm": 0.4254636742864719,
+      "learning_rate": 4.898233848219532e-06,
+      "loss": 0.0385,
+      "step": 5005
+    },
+    {
+      "epoch": 2.2234066178103484,
+      "grad_norm": 0.5219692883962996,
+      "learning_rate": 4.896295720733694e-06,
+      "loss": 0.04,
+      "step": 5006
+    },
+    {
+      "epoch": 2.223850766155896,
+      "grad_norm": 0.8849555604502084,
+      "learning_rate": 4.894357608836368e-06,
+      "loss": 0.0357,
+      "step": 5007
+    },
+    {
+      "epoch": 2.2242949145014435,
+      "grad_norm": 0.772611179602613,
+      "learning_rate": 4.89241951281889e-06,
+      "loss": 0.0418,
+      "step": 5008
+    },
+    {
+      "epoch": 2.2247390628469907,
+      "grad_norm": 0.5532053379610344,
+      "learning_rate": 4.890481432972586e-06,
+      "loss": 0.0352,
+      "step": 5009
+    },
+    {
+      "epoch": 2.2251832111925385,
+      "grad_norm": 0.4003154527748429,
+      "learning_rate": 4.8885433695887836e-06,
+      "loss": 0.032,
+      "step": 5010
+    },
+    {
+      "epoch": 2.2256273595380858,
+      "grad_norm": 0.45204263035502396,
+      "learning_rate": 4.886605322958806e-06,
+      "loss": 0.0303,
+      "step": 5011
+    },
+    {
+      "epoch": 2.226071507883633,
+      "grad_norm": 0.3440942170404326,
+      "learning_rate": 4.884667293373973e-06,
+      "loss": 0.022,
+      "step": 5012
+    },
+    {
+      "epoch": 2.2265156562291804,
+      "grad_norm": 0.7116582097517625,
+      "learning_rate": 4.882729281125605e-06,
+      "loss": 0.0523,
+      "step": 5013
+    },
+    {
+      "epoch": 2.226959804574728,
+      "grad_norm": 0.4053030773514669,
+      "learning_rate": 4.88079128650502e-06,
+      "loss": 0.027,
+      "step": 5014
+    },
+    {
+      "epoch": 2.2274039529202754,
+      "grad_norm": 0.6656245142980673,
+      "learning_rate": 4.878853309803529e-06,
+      "loss": 0.0462,
+      "step": 5015
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.38760410447982047,
+      "learning_rate": 4.876915351312444e-06,
+      "loss": 0.0265,
+      "step": 5016
+    },
+    {
+      "epoch": 2.2282922496113704,
+      "grad_norm": 0.407666066744564,
+      "learning_rate": 4.874977411323073e-06,
+      "loss": 0.0354,
+      "step": 5017
+    },
+    {
+      "epoch": 2.2287363979569177,
+      "grad_norm": 0.483920298386755,
+      "learning_rate": 4.873039490126718e-06,
+      "loss": 0.0367,
+      "step": 5018
+    },
+    {
+      "epoch": 2.229180546302465,
+      "grad_norm": 0.45353775305411304,
+      "learning_rate": 4.871101588014686e-06,
+      "loss": 0.0358,
+      "step": 5019
+    },
+    {
+      "epoch": 2.2296246946480123,
+      "grad_norm": 0.4339146999685688,
+      "learning_rate": 4.869163705278276e-06,
+      "loss": 0.0326,
+      "step": 5020
+    },
+    {
+      "epoch": 2.23006884299356,
+      "grad_norm": 1.02624057870213,
+      "learning_rate": 4.867225842208783e-06,
+      "loss": 0.0598,
+      "step": 5021
+    },
+    {
+      "epoch": 2.2305129913391073,
+      "grad_norm": 0.44283161622309714,
+      "learning_rate": 4.8652879990975e-06,
+      "loss": 0.0378,
+      "step": 5022
+    },
+    {
+      "epoch": 2.2309571396846546,
+      "grad_norm": 0.44468206860911264,
+      "learning_rate": 4.863350176235721e-06,
+      "loss": 0.0435,
+      "step": 5023
+    },
+    {
+      "epoch": 2.2314012880302023,
+      "grad_norm": 0.4317246743183953,
+      "learning_rate": 4.861412373914729e-06,
+      "loss": 0.0339,
+      "step": 5024
+    },
+    {
+      "epoch": 2.2318454363757496,
+      "grad_norm": 1.9956444643720284,
+      "learning_rate": 4.8594745924258144e-06,
+      "loss": 0.0611,
+      "step": 5025
+    },
+    {
+      "epoch": 2.232289584721297,
+      "grad_norm": 0.5268379152925193,
+      "learning_rate": 4.857536832060255e-06,
+      "loss": 0.0408,
+      "step": 5026
+    },
+    {
+      "epoch": 2.232733733066844,
+      "grad_norm": 0.4937129380739942,
+      "learning_rate": 4.85559909310933e-06,
+      "loss": 0.0523,
+      "step": 5027
+    },
+    {
+      "epoch": 2.233177881412392,
+      "grad_norm": 0.42104370790377565,
+      "learning_rate": 4.853661375864313e-06,
+      "loss": 0.0226,
+      "step": 5028
+    },
+    {
+      "epoch": 2.233622029757939,
+      "grad_norm": 0.38358747557836487,
+      "learning_rate": 4.851723680616477e-06,
+      "loss": 0.0288,
+      "step": 5029
+    },
+    {
+      "epoch": 2.2340661781034865,
+      "grad_norm": 0.35278248179931115,
+      "learning_rate": 4.84978600765709e-06,
+      "loss": 0.0273,
+      "step": 5030
+    },
+    {
+      "epoch": 2.234510326449034,
+      "grad_norm": 0.5070929341551949,
+      "learning_rate": 4.84784835727742e-06,
+      "loss": 0.0421,
+      "step": 5031
+    },
+    {
+      "epoch": 2.2349544747945815,
+      "grad_norm": 0.4414918243805681,
+      "learning_rate": 4.845910729768726e-06,
+      "loss": 0.0299,
+      "step": 5032
+    },
+    {
+      "epoch": 2.235398623140129,
+      "grad_norm": 0.4272744129564785,
+      "learning_rate": 4.843973125422266e-06,
+      "loss": 0.0377,
+      "step": 5033
+    },
+    {
+      "epoch": 2.235842771485676,
+      "grad_norm": 0.4042780850426815,
+      "learning_rate": 4.842035544529296e-06,
+      "loss": 0.0273,
+      "step": 5034
+    },
+    {
+      "epoch": 2.2362869198312234,
+      "grad_norm": 0.384529185725243,
+      "learning_rate": 4.8400979873810675e-06,
+      "loss": 0.0297,
+      "step": 5035
+    },
+    {
+      "epoch": 2.236731068176771,
+      "grad_norm": 0.5143165626824721,
+      "learning_rate": 4.838160454268827e-06,
+      "loss": 0.0378,
+      "step": 5036
+    },
+    {
+      "epoch": 2.2371752165223184,
+      "grad_norm": 0.48737479457986954,
+      "learning_rate": 4.8362229454838185e-06,
+      "loss": 0.0343,
+      "step": 5037
+    },
+    {
+      "epoch": 2.2376193648678657,
+      "grad_norm": 0.4465431723189805,
+      "learning_rate": 4.834285461317286e-06,
+      "loss": 0.0367,
+      "step": 5038
+    },
+    {
+      "epoch": 2.2380635132134135,
+      "grad_norm": 0.5195671115110356,
+      "learning_rate": 4.832348002060461e-06,
+      "loss": 0.0396,
+      "step": 5039
+    },
+    {
+      "epoch": 2.2385076615589607,
+      "grad_norm": 0.46391257174900896,
+      "learning_rate": 4.830410568004581e-06,
+      "loss": 0.036,
+      "step": 5040
+    },
+    {
+      "epoch": 2.238951809904508,
+      "grad_norm": 0.5290347592358237,
+      "learning_rate": 4.8284731594408715e-06,
+      "loss": 0.0448,
+      "step": 5041
+    },
+    {
+      "epoch": 2.2393959582500553,
+      "grad_norm": 0.4737259034687236,
+      "learning_rate": 4.826535776660562e-06,
+      "loss": 0.0328,
+      "step": 5042
+    },
+    {
+      "epoch": 2.239840106595603,
+      "grad_norm": 0.6929115316947699,
+      "learning_rate": 4.824598419954871e-06,
+      "loss": 0.0473,
+      "step": 5043
+    },
+    {
+      "epoch": 2.2402842549411504,
+      "grad_norm": 0.6258064825937694,
+      "learning_rate": 4.822661089615017e-06,
+      "loss": 0.0594,
+      "step": 5044
+    },
+    {
+      "epoch": 2.2407284032866976,
+      "grad_norm": 0.4537831813620375,
+      "learning_rate": 4.8207237859322144e-06,
+      "loss": 0.0413,
+      "step": 5045
+    },
+    {
+      "epoch": 2.2411725516322454,
+      "grad_norm": 0.37664633790712776,
+      "learning_rate": 4.818786509197672e-06,
+      "loss": 0.0303,
+      "step": 5046
+    },
+    {
+      "epoch": 2.2416166999777927,
+      "grad_norm": 0.3886976621007685,
+      "learning_rate": 4.816849259702594e-06,
+      "loss": 0.0311,
+      "step": 5047
+    },
+    {
+      "epoch": 2.24206084832334,
+      "grad_norm": 0.4486972497289991,
+      "learning_rate": 4.814912037738185e-06,
+      "loss": 0.0442,
+      "step": 5048
+    },
+    {
+      "epoch": 2.2425049966688873,
+      "grad_norm": 0.35002545439343274,
+      "learning_rate": 4.812974843595641e-06,
+      "loss": 0.0318,
+      "step": 5049
+    },
+    {
+      "epoch": 2.242949145014435,
+      "grad_norm": 0.48300556886418516,
+      "learning_rate": 4.811037677566154e-06,
+      "loss": 0.0395,
+      "step": 5050
+    },
+    {
+      "epoch": 2.2433932933599823,
+      "grad_norm": 0.45349978147674774,
+      "learning_rate": 4.8091005399409145e-06,
+      "loss": 0.0371,
+      "step": 5051
+    },
+    {
+      "epoch": 2.2438374417055296,
+      "grad_norm": 0.4518413356936792,
+      "learning_rate": 4.807163431011107e-06,
+      "loss": 0.0311,
+      "step": 5052
+    },
+    {
+      "epoch": 2.244281590051077,
+      "grad_norm": 0.3777032140049169,
+      "learning_rate": 4.80522635106791e-06,
+      "loss": 0.0292,
+      "step": 5053
+    },
+    {
+      "epoch": 2.2447257383966246,
+      "grad_norm": 0.4369243999495971,
+      "learning_rate": 4.8032893004025016e-06,
+      "loss": 0.0264,
+      "step": 5054
+    },
+    {
+      "epoch": 2.245169886742172,
+      "grad_norm": 0.6447452509815467,
+      "learning_rate": 4.801352279306054e-06,
+      "loss": 0.0427,
+      "step": 5055
+    },
+    {
+      "epoch": 2.245614035087719,
+      "grad_norm": 0.5030519012852677,
+      "learning_rate": 4.799415288069733e-06,
+      "loss": 0.0381,
+      "step": 5056
+    },
+    {
+      "epoch": 2.246058183433267,
+      "grad_norm": 0.4348315875534344,
+      "learning_rate": 4.797478326984702e-06,
+      "loss": 0.0235,
+      "step": 5057
+    },
+    {
+      "epoch": 2.246502331778814,
+      "grad_norm": 0.47850023846789697,
+      "learning_rate": 4.795541396342116e-06,
+      "loss": 0.036,
+      "step": 5058
+    },
+    {
+      "epoch": 2.2469464801243615,
+      "grad_norm": 0.5227923023136899,
+      "learning_rate": 4.793604496433133e-06,
+      "loss": 0.0382,
+      "step": 5059
+    },
+    {
+      "epoch": 2.247390628469909,
+      "grad_norm": 0.5404616232699001,
+      "learning_rate": 4.791667627548902e-06,
+      "loss": 0.0426,
+      "step": 5060
+    },
+    {
+      "epoch": 2.2478347768154565,
+      "grad_norm": 0.5031844080280006,
+      "learning_rate": 4.7897307899805624e-06,
+      "loss": 0.0296,
+      "step": 5061
+    },
+    {
+      "epoch": 2.248278925161004,
+      "grad_norm": 0.557884313184048,
+      "learning_rate": 4.78779398401926e-06,
+      "loss": 0.0324,
+      "step": 5062
+    },
+    {
+      "epoch": 2.248723073506551,
+      "grad_norm": 0.46196149242671997,
+      "learning_rate": 4.785857209956124e-06,
+      "loss": 0.031,
+      "step": 5063
+    },
+    {
+      "epoch": 2.2491672218520984,
+      "grad_norm": 0.40473001188667446,
+      "learning_rate": 4.783920468082288e-06,
+      "loss": 0.0341,
+      "step": 5064
+    },
+    {
+      "epoch": 2.249611370197646,
+      "grad_norm": 0.558756976897473,
+      "learning_rate": 4.781983758688876e-06,
+      "loss": 0.0257,
+      "step": 5065
+    },
+    {
+      "epoch": 2.2500555185431934,
+      "grad_norm": 0.4215840569154271,
+      "learning_rate": 4.780047082067009e-06,
+      "loss": 0.0402,
+      "step": 5066
+    },
+    {
+      "epoch": 2.2504996668887407,
+      "grad_norm": 0.41373536715916187,
+      "learning_rate": 4.778110438507801e-06,
+      "loss": 0.0352,
+      "step": 5067
+    },
+    {
+      "epoch": 2.2509438152342884,
+      "grad_norm": 0.41096920941351195,
+      "learning_rate": 4.776173828302365e-06,
+      "loss": 0.0411,
+      "step": 5068
+    },
+    {
+      "epoch": 2.2513879635798357,
+      "grad_norm": 0.4730245858374104,
+      "learning_rate": 4.774237251741805e-06,
+      "loss": 0.0327,
+      "step": 5069
+    },
+    {
+      "epoch": 2.251832111925383,
+      "grad_norm": 0.38391334338943445,
+      "learning_rate": 4.7723007091172175e-06,
+      "loss": 0.0368,
+      "step": 5070
+    },
+    {
+      "epoch": 2.2522762602709303,
+      "grad_norm": 0.34213411038422287,
+      "learning_rate": 4.770364200719703e-06,
+      "loss": 0.0288,
+      "step": 5071
+    },
+    {
+      "epoch": 2.252720408616478,
+      "grad_norm": 0.3470528030999554,
+      "learning_rate": 4.7684277268403515e-06,
+      "loss": 0.0218,
+      "step": 5072
+    },
+    {
+      "epoch": 2.2531645569620253,
+      "grad_norm": 0.48424365825920085,
+      "learning_rate": 4.766491287770246e-06,
+      "loss": 0.047,
+      "step": 5073
+    },
+    {
+      "epoch": 2.2536087053075726,
+      "grad_norm": 0.44693843285607443,
+      "learning_rate": 4.7645548838004665e-06,
+      "loss": 0.0314,
+      "step": 5074
+    },
+    {
+      "epoch": 2.2540528536531204,
+      "grad_norm": 0.43485150995828376,
+      "learning_rate": 4.762618515222085e-06,
+      "loss": 0.0342,
+      "step": 5075
+    },
+    {
+      "epoch": 2.2544970019986676,
+      "grad_norm": 0.4279292260890883,
+      "learning_rate": 4.760682182326176e-06,
+      "loss": 0.0407,
+      "step": 5076
+    },
+    {
+      "epoch": 2.254941150344215,
+      "grad_norm": 0.4726345938038404,
+      "learning_rate": 4.7587458854038e-06,
+      "loss": 0.0366,
+      "step": 5077
+    },
+    {
+      "epoch": 2.2553852986897622,
+      "grad_norm": 0.4982236299340504,
+      "learning_rate": 4.756809624746015e-06,
+      "loss": 0.0351,
+      "step": 5078
+    },
+    {
+      "epoch": 2.25582944703531,
+      "grad_norm": 0.416498003591409,
+      "learning_rate": 4.754873400643875e-06,
+      "loss": 0.0345,
+      "step": 5079
+    },
+    {
+      "epoch": 2.2562735953808573,
+      "grad_norm": 0.5784538425140838,
+      "learning_rate": 4.7529372133884265e-06,
+      "loss": 0.0351,
+      "step": 5080
+    },
+    {
+      "epoch": 2.2567177437264045,
+      "grad_norm": 0.4069962515786798,
+      "learning_rate": 4.75100106327071e-06,
+      "loss": 0.0287,
+      "step": 5081
+    },
+    {
+      "epoch": 2.2571618920719523,
+      "grad_norm": 0.44000239841542776,
+      "learning_rate": 4.749064950581765e-06,
+      "loss": 0.0318,
+      "step": 5082
+    },
+    {
+      "epoch": 2.2576060404174996,
+      "grad_norm": 0.4673614175152977,
+      "learning_rate": 4.747128875612621e-06,
+      "loss": 0.0298,
+      "step": 5083
+    },
+    {
+      "epoch": 2.258050188763047,
+      "grad_norm": 0.577011350942782,
+      "learning_rate": 4.745192838654304e-06,
+      "loss": 0.0388,
+      "step": 5084
+    },
+    {
+      "epoch": 2.258494337108594,
+      "grad_norm": 0.40314614436836343,
+      "learning_rate": 4.743256839997828e-06,
+      "loss": 0.0318,
+      "step": 5085
+    },
+    {
+      "epoch": 2.2589384854541414,
+      "grad_norm": 0.3926832668709334,
+      "learning_rate": 4.741320879934213e-06,
+      "loss": 0.0361,
+      "step": 5086
+    },
+    {
+      "epoch": 2.259382633799689,
+      "grad_norm": 0.5101559666731487,
+      "learning_rate": 4.739384958754461e-06,
+      "loss": 0.0451,
+      "step": 5087
+    },
+    {
+      "epoch": 2.2598267821452365,
+      "grad_norm": 0.35566284403310655,
+      "learning_rate": 4.73744907674958e-06,
+      "loss": 0.0228,
+      "step": 5088
+    },
+    {
+      "epoch": 2.2602709304907838,
+      "grad_norm": 0.5266642057409271,
+      "learning_rate": 4.7355132342105615e-06,
+      "loss": 0.0387,
+      "step": 5089
+    },
+    {
+      "epoch": 2.2607150788363315,
+      "grad_norm": 0.4021115546007318,
+      "learning_rate": 4.733577431428398e-06,
+      "loss": 0.0313,
+      "step": 5090
+    },
+    {
+      "epoch": 2.261159227181879,
+      "grad_norm": 0.48902419392318525,
+      "learning_rate": 4.73164166869407e-06,
+      "loss": 0.0465,
+      "step": 5091
+    },
+    {
+      "epoch": 2.261603375527426,
+      "grad_norm": 0.4831670488263336,
+      "learning_rate": 4.729705946298557e-06,
+      "loss": 0.0298,
+      "step": 5092
+    },
+    {
+      "epoch": 2.2620475238729734,
+      "grad_norm": 1.0045520985152046,
+      "learning_rate": 4.727770264532835e-06,
+      "loss": 0.0386,
+      "step": 5093
+    },
+    {
+      "epoch": 2.262491672218521,
+      "grad_norm": 0.5424294421568248,
+      "learning_rate": 4.725834623687866e-06,
+      "loss": 0.0448,
+      "step": 5094
+    },
+    {
+      "epoch": 2.2629358205640684,
+      "grad_norm": 0.37090231243704935,
+      "learning_rate": 4.723899024054609e-06,
+      "loss": 0.0276,
+      "step": 5095
+    },
+    {
+      "epoch": 2.2633799689096157,
+      "grad_norm": 0.41587596025980633,
+      "learning_rate": 4.7219634659240195e-06,
+      "loss": 0.037,
+      "step": 5096
+    },
+    {
+      "epoch": 2.2638241172551634,
+      "grad_norm": 0.4579599445978751,
+      "learning_rate": 4.720027949587046e-06,
+      "loss": 0.0355,
+      "step": 5097
+    },
+    {
+      "epoch": 2.2642682656007107,
+      "grad_norm": 0.33053691402038254,
+      "learning_rate": 4.718092475334623e-06,
+      "loss": 0.0237,
+      "step": 5098
+    },
+    {
+      "epoch": 2.264712413946258,
+      "grad_norm": 0.4734298805101396,
+      "learning_rate": 4.716157043457692e-06,
+      "loss": 0.0406,
+      "step": 5099
+    },
+    {
+      "epoch": 2.2651565622918053,
+      "grad_norm": 0.495566512713766,
+      "learning_rate": 4.71422165424718e-06,
+      "loss": 0.0446,
+      "step": 5100
+    },
+    {
+      "epoch": 2.265600710637353,
+      "grad_norm": 0.5845335631241523,
+      "learning_rate": 4.712286307994008e-06,
+      "loss": 0.0483,
+      "step": 5101
+    },
+    {
+      "epoch": 2.2660448589829003,
+      "grad_norm": 0.5632138786502331,
+      "learning_rate": 4.71035100498909e-06,
+      "loss": 0.0301,
+      "step": 5102
+    },
+    {
+      "epoch": 2.2664890073284476,
+      "grad_norm": 0.4679864832864739,
+      "learning_rate": 4.708415745523338e-06,
+      "loss": 0.0282,
+      "step": 5103
+    },
+    {
+      "epoch": 2.2669331556739953,
+      "grad_norm": 0.3347447433418278,
+      "learning_rate": 4.70648052988765e-06,
+      "loss": 0.0284,
+      "step": 5104
+    },
+    {
+      "epoch": 2.2673773040195426,
+      "grad_norm": 0.3922788691977759,
+      "learning_rate": 4.704545358372926e-06,
+      "loss": 0.0376,
+      "step": 5105
+    },
+    {
+      "epoch": 2.26782145236509,
+      "grad_norm": 0.574421309924243,
+      "learning_rate": 4.702610231270053e-06,
+      "loss": 0.0357,
+      "step": 5106
+    },
+    {
+      "epoch": 2.268265600710637,
+      "grad_norm": 0.3970980176428026,
+      "learning_rate": 4.7006751488699145e-06,
+      "loss": 0.0352,
+      "step": 5107
+    },
+    {
+      "epoch": 2.268709749056185,
+      "grad_norm": 0.48320849905313684,
+      "learning_rate": 4.698740111463386e-06,
+      "loss": 0.0482,
+      "step": 5108
+    },
+    {
+      "epoch": 2.2691538974017322,
+      "grad_norm": 0.48488451433259316,
+      "learning_rate": 4.696805119341334e-06,
+      "loss": 0.0323,
+      "step": 5109
+    },
+    {
+      "epoch": 2.2695980457472795,
+      "grad_norm": 0.38461440123857293,
+      "learning_rate": 4.694870172794625e-06,
+      "loss": 0.0256,
+      "step": 5110
+    },
+    {
+      "epoch": 2.270042194092827,
+      "grad_norm": 0.5557330047140479,
+      "learning_rate": 4.692935272114113e-06,
+      "loss": 0.0303,
+      "step": 5111
+    },
+    {
+      "epoch": 2.2704863424383745,
+      "grad_norm": 0.4256236131352179,
+      "learning_rate": 4.6910004175906435e-06,
+      "loss": 0.0233,
+      "step": 5112
+    },
+    {
+      "epoch": 2.270930490783922,
+      "grad_norm": 0.46988918438082106,
+      "learning_rate": 4.689065609515062e-06,
+      "loss": 0.0363,
+      "step": 5113
+    },
+    {
+      "epoch": 2.271374639129469,
+      "grad_norm": 0.49373029930466805,
+      "learning_rate": 4.687130848178202e-06,
+      "loss": 0.0483,
+      "step": 5114
+    },
+    {
+      "epoch": 2.2718187874750164,
+      "grad_norm": 0.34972548300588746,
+      "learning_rate": 4.685196133870887e-06,
+      "loss": 0.0255,
+      "step": 5115
+    },
+    {
+      "epoch": 2.272262935820564,
+      "grad_norm": 0.581571662658351,
+      "learning_rate": 4.683261466883942e-06,
+      "loss": 0.0414,
+      "step": 5116
+    },
+    {
+      "epoch": 2.2727070841661114,
+      "grad_norm": 0.44401771745072516,
+      "learning_rate": 4.681326847508181e-06,
+      "loss": 0.032,
+      "step": 5117
+    },
+    {
+      "epoch": 2.2731512325116587,
+      "grad_norm": 0.5278060546522586,
+      "learning_rate": 4.6793922760344065e-06,
+      "loss": 0.0418,
+      "step": 5118
+    },
+    {
+      "epoch": 2.2735953808572065,
+      "grad_norm": 0.5206864377402288,
+      "learning_rate": 4.6774577527534195e-06,
+      "loss": 0.0388,
+      "step": 5119
+    },
+    {
+      "epoch": 2.2740395292027538,
+      "grad_norm": 0.4082716881176943,
+      "learning_rate": 4.675523277956011e-06,
+      "loss": 0.0283,
+      "step": 5120
+    },
+    {
+      "epoch": 2.274483677548301,
+      "grad_norm": 0.4188100723565611,
+      "learning_rate": 4.673588851932964e-06,
+      "loss": 0.0377,
+      "step": 5121
+    },
+    {
+      "epoch": 2.2749278258938483,
+      "grad_norm": 0.32957447472088847,
+      "learning_rate": 4.671654474975061e-06,
+      "loss": 0.0284,
+      "step": 5122
+    },
+    {
+      "epoch": 2.275371974239396,
+      "grad_norm": 0.5325173700148402,
+      "learning_rate": 4.669720147373065e-06,
+      "loss": 0.0304,
+      "step": 5123
+    },
+    {
+      "epoch": 2.2758161225849434,
+      "grad_norm": 0.7259350026566369,
+      "learning_rate": 4.667785869417744e-06,
+      "loss": 0.0484,
+      "step": 5124
+    },
+    {
+      "epoch": 2.2762602709304907,
+      "grad_norm": 0.34263569388234766,
+      "learning_rate": 4.6658516413998486e-06,
+      "loss": 0.0302,
+      "step": 5125
+    },
+    {
+      "epoch": 2.2767044192760384,
+      "grad_norm": 0.4050161385405534,
+      "learning_rate": 4.663917463610128e-06,
+      "loss": 0.0327,
+      "step": 5126
+    },
+    {
+      "epoch": 2.2771485676215857,
+      "grad_norm": 0.4351043792797754,
+      "learning_rate": 4.661983336339319e-06,
+      "loss": 0.0347,
+      "step": 5127
+    },
+    {
+      "epoch": 2.277592715967133,
+      "grad_norm": 0.4544801574632252,
+      "learning_rate": 4.66004925987816e-06,
+      "loss": 0.0335,
+      "step": 5128
+    },
+    {
+      "epoch": 2.2780368643126803,
+      "grad_norm": 0.5298789728156417,
+      "learning_rate": 4.6581152345173714e-06,
+      "loss": 0.0481,
+      "step": 5129
+    },
+    {
+      "epoch": 2.278481012658228,
+      "grad_norm": 0.42879788700967447,
+      "learning_rate": 4.656181260547669e-06,
+      "loss": 0.0306,
+      "step": 5130
+    },
+    {
+      "epoch": 2.2789251610037753,
+      "grad_norm": 0.432457604882885,
+      "learning_rate": 4.654247338259766e-06,
+      "loss": 0.0348,
+      "step": 5131
+    },
+    {
+      "epoch": 2.2793693093493226,
+      "grad_norm": 0.4238026290724064,
+      "learning_rate": 4.652313467944358e-06,
+      "loss": 0.0362,
+      "step": 5132
+    },
+    {
+      "epoch": 2.2798134576948703,
+      "grad_norm": 0.37078187284702663,
+      "learning_rate": 4.650379649892145e-06,
+      "loss": 0.0225,
+      "step": 5133
+    },
+    {
+      "epoch": 2.2802576060404176,
+      "grad_norm": 0.541283465006368,
+      "learning_rate": 4.648445884393808e-06,
+      "loss": 0.0427,
+      "step": 5134
+    },
+    {
+      "epoch": 2.280701754385965,
+      "grad_norm": 0.5756809781245248,
+      "learning_rate": 4.646512171740028e-06,
+      "loss": 0.0416,
+      "step": 5135
+    },
+    {
+      "epoch": 2.281145902731512,
+      "grad_norm": 0.6172689341906578,
+      "learning_rate": 4.6445785122214715e-06,
+      "loss": 0.0468,
+      "step": 5136
+    },
+    {
+      "epoch": 2.28159005107706,
+      "grad_norm": 0.4664467318852636,
+      "learning_rate": 4.6426449061288035e-06,
+      "loss": 0.0394,
+      "step": 5137
+    },
+    {
+      "epoch": 2.282034199422607,
+      "grad_norm": 0.3708061416037573,
+      "learning_rate": 4.640711353752675e-06,
+      "loss": 0.0308,
+      "step": 5138
+    },
+    {
+      "epoch": 2.2824783477681545,
+      "grad_norm": 0.4312428309303351,
+      "learning_rate": 4.638777855383735e-06,
+      "loss": 0.0361,
+      "step": 5139
+    },
+    {
+      "epoch": 2.282922496113702,
+      "grad_norm": 0.3963057474327594,
+      "learning_rate": 4.636844411312618e-06,
+      "loss": 0.0244,
+      "step": 5140
+    },
+    {
+      "epoch": 2.2833666444592495,
+      "grad_norm": 0.48035729602544386,
+      "learning_rate": 4.634911021829956e-06,
+      "loss": 0.0388,
+      "step": 5141
+    },
+    {
+      "epoch": 2.283810792804797,
+      "grad_norm": 0.3951503991589514,
+      "learning_rate": 4.63297768722637e-06,
+      "loss": 0.0311,
+      "step": 5142
+    },
+    {
+      "epoch": 2.284254941150344,
+      "grad_norm": 0.3422700971686053,
+      "learning_rate": 4.6310444077924705e-06,
+      "loss": 0.0244,
+      "step": 5143
+    },
+    {
+      "epoch": 2.2846990894958914,
+      "grad_norm": 0.6947651223374093,
+      "learning_rate": 4.629111183818863e-06,
+      "loss": 0.0351,
+      "step": 5144
+    },
+    {
+      "epoch": 2.285143237841439,
+      "grad_norm": 0.4898604396865276,
+      "learning_rate": 4.627178015596147e-06,
+      "loss": 0.0369,
+      "step": 5145
+    },
+    {
+      "epoch": 2.2855873861869864,
+      "grad_norm": 0.384272758272605,
+      "learning_rate": 4.625244903414908e-06,
+      "loss": 0.0378,
+      "step": 5146
+    },
+    {
+      "epoch": 2.2860315345325337,
+      "grad_norm": 0.6383973512520991,
+      "learning_rate": 4.623311847565725e-06,
+      "loss": 0.0371,
+      "step": 5147
+    },
+    {
+      "epoch": 2.2864756828780815,
+      "grad_norm": 0.43116354217379305,
+      "learning_rate": 4.621378848339172e-06,
+      "loss": 0.0385,
+      "step": 5148
+    },
+    {
+      "epoch": 2.2869198312236287,
+      "grad_norm": 0.37732659096182064,
+      "learning_rate": 4.619445906025807e-06,
+      "loss": 0.0302,
+      "step": 5149
+    },
+    {
+      "epoch": 2.287363979569176,
+      "grad_norm": 0.38619197439251224,
+      "learning_rate": 4.6175130209161894e-06,
+      "loss": 0.0309,
+      "step": 5150
+    },
+    {
+      "epoch": 2.2878081279147233,
+      "grad_norm": 0.3996619270485742,
+      "learning_rate": 4.615580193300861e-06,
+      "loss": 0.0342,
+      "step": 5151
+    },
+    {
+      "epoch": 2.288252276260271,
+      "grad_norm": 0.4975335885399077,
+      "learning_rate": 4.613647423470361e-06,
+      "loss": 0.0295,
+      "step": 5152
+    },
+    {
+      "epoch": 2.2886964246058183,
+      "grad_norm": 0.3800465801231585,
+      "learning_rate": 4.611714711715215e-06,
+      "loss": 0.0287,
+      "step": 5153
+    },
+    {
+      "epoch": 2.2891405729513656,
+      "grad_norm": 0.4187996600512099,
+      "learning_rate": 4.609782058325944e-06,
+      "loss": 0.0357,
+      "step": 5154
+    },
+    {
+      "epoch": 2.2895847212969134,
+      "grad_norm": 0.6627247371766289,
+      "learning_rate": 4.607849463593056e-06,
+      "loss": 0.0319,
+      "step": 5155
+    },
+    {
+      "epoch": 2.2900288696424607,
+      "grad_norm": 0.44696110840918063,
+      "learning_rate": 4.6059169278070576e-06,
+      "loss": 0.0439,
+      "step": 5156
+    },
+    {
+      "epoch": 2.290473017988008,
+      "grad_norm": 0.43597000773880085,
+      "learning_rate": 4.603984451258439e-06,
+      "loss": 0.0309,
+      "step": 5157
+    },
+    {
+      "epoch": 2.2909171663335552,
+      "grad_norm": 0.3416433662316786,
+      "learning_rate": 4.602052034237684e-06,
+      "loss": 0.0244,
+      "step": 5158
+    },
+    {
+      "epoch": 2.291361314679103,
+      "grad_norm": 0.4086413859197415,
+      "learning_rate": 4.600119677035269e-06,
+      "loss": 0.0357,
+      "step": 5159
+    },
+    {
+      "epoch": 2.2918054630246503,
+      "grad_norm": 0.4905561101374712,
+      "learning_rate": 4.598187379941659e-06,
+      "loss": 0.036,
+      "step": 5160
+    },
+    {
+      "epoch": 2.2922496113701976,
+      "grad_norm": 0.47425089286645844,
+      "learning_rate": 4.59625514324731e-06,
+      "loss": 0.0344,
+      "step": 5161
+    },
+    {
+      "epoch": 2.2926937597157453,
+      "grad_norm": 0.4768903592454644,
+      "learning_rate": 4.594322967242673e-06,
+      "loss": 0.0445,
+      "step": 5162
+    },
+    {
+      "epoch": 2.2931379080612926,
+      "grad_norm": 0.6604194357601725,
+      "learning_rate": 4.592390852218185e-06,
+      "loss": 0.0353,
+      "step": 5163
+    },
+    {
+      "epoch": 2.29358205640684,
+      "grad_norm": 0.6477706525687013,
+      "learning_rate": 4.590458798464275e-06,
+      "loss": 0.0346,
+      "step": 5164
+    },
+    {
+      "epoch": 2.294026204752387,
+      "grad_norm": 0.4782427676732749,
+      "learning_rate": 4.588526806271366e-06,
+      "loss": 0.0332,
+      "step": 5165
+    },
+    {
+      "epoch": 2.294470353097935,
+      "grad_norm": 0.33598954590538455,
+      "learning_rate": 4.5865948759298656e-06,
+      "loss": 0.0242,
+      "step": 5166
+    },
+    {
+      "epoch": 2.294914501443482,
+      "grad_norm": 0.5314954081213854,
+      "learning_rate": 4.58466300773018e-06,
+      "loss": 0.0375,
+      "step": 5167
+    },
+    {
+      "epoch": 2.2953586497890295,
+      "grad_norm": 0.45504963601939913,
+      "learning_rate": 4.582731201962699e-06,
+      "loss": 0.0325,
+      "step": 5168
+    },
+    {
+      "epoch": 2.2958027981345768,
+      "grad_norm": 0.5020378540227112,
+      "learning_rate": 4.5807994589178066e-06,
+      "loss": 0.0397,
+      "step": 5169
+    },
+    {
+      "epoch": 2.2962469464801245,
+      "grad_norm": 0.44086322018058155,
+      "learning_rate": 4.578867778885877e-06,
+      "loss": 0.0385,
+      "step": 5170
+    },
+    {
+      "epoch": 2.296691094825672,
+      "grad_norm": 0.4190415887827377,
+      "learning_rate": 4.5769361621572735e-06,
+      "loss": 0.0351,
+      "step": 5171
+    },
+    {
+      "epoch": 2.297135243171219,
+      "grad_norm": 0.40356235040543037,
+      "learning_rate": 4.575004609022349e-06,
+      "loss": 0.0384,
+      "step": 5172
+    },
+    {
+      "epoch": 2.2975793915167664,
+      "grad_norm": 0.5214338427099589,
+      "learning_rate": 4.573073119771455e-06,
+      "loss": 0.0414,
+      "step": 5173
+    },
+    {
+      "epoch": 2.298023539862314,
+      "grad_norm": 0.6042586649916282,
+      "learning_rate": 4.571141694694922e-06,
+      "loss": 0.0398,
+      "step": 5174
+    },
+    {
+      "epoch": 2.2984676882078614,
+      "grad_norm": 0.5451011557742893,
+      "learning_rate": 4.569210334083077e-06,
+      "loss": 0.0284,
+      "step": 5175
+    },
+    {
+      "epoch": 2.2989118365534087,
+      "grad_norm": 0.6693775358695208,
+      "learning_rate": 4.567279038226237e-06,
+      "loss": 0.0393,
+      "step": 5176
+    },
+    {
+      "epoch": 2.2993559848989564,
+      "grad_norm": 0.4565914726092485,
+      "learning_rate": 4.565347807414709e-06,
+      "loss": 0.0369,
+      "step": 5177
+    },
+    {
+      "epoch": 2.2998001332445037,
+      "grad_norm": 0.3909149508886464,
+      "learning_rate": 4.563416641938786e-06,
+      "loss": 0.0312,
+      "step": 5178
+    },
+    {
+      "epoch": 2.300244281590051,
+      "grad_norm": 0.48805937276628275,
+      "learning_rate": 4.5614855420887595e-06,
+      "loss": 0.0417,
+      "step": 5179
+    },
+    {
+      "epoch": 2.3006884299355983,
+      "grad_norm": 0.47062502463527334,
+      "learning_rate": 4.559554508154906e-06,
+      "loss": 0.0359,
+      "step": 5180
+    },
+    {
+      "epoch": 2.301132578281146,
+      "grad_norm": 0.41598990819931514,
+      "learning_rate": 4.557623540427492e-06,
+      "loss": 0.0298,
+      "step": 5181
+    },
+    {
+      "epoch": 2.3015767266266933,
+      "grad_norm": 0.5301634737897044,
+      "learning_rate": 4.555692639196774e-06,
+      "loss": 0.0419,
+      "step": 5182
+    },
+    {
+      "epoch": 2.3020208749722406,
+      "grad_norm": 0.3965912885980032,
+      "learning_rate": 4.553761804752997e-06,
+      "loss": 0.0268,
+      "step": 5183
+    },
+    {
+      "epoch": 2.3024650233177884,
+      "grad_norm": 0.3853380797199742,
+      "learning_rate": 4.551831037386405e-06,
+      "loss": 0.0296,
+      "step": 5184
+    },
+    {
+      "epoch": 2.3029091716633356,
+      "grad_norm": 0.4626560137060772,
+      "learning_rate": 4.54990033738722e-06,
+      "loss": 0.0347,
+      "step": 5185
+    },
+    {
+      "epoch": 2.303353320008883,
+      "grad_norm": 0.4662767664074809,
+      "learning_rate": 4.54796970504566e-06,
+      "loss": 0.0382,
+      "step": 5186
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.5212735203918284,
+      "learning_rate": 4.546039140651932e-06,
+      "loss": 0.0347,
+      "step": 5187
+    },
+    {
+      "epoch": 2.304241616699978,
+      "grad_norm": 0.6660381074871294,
+      "learning_rate": 4.544108644496232e-06,
+      "loss": 0.0414,
+      "step": 5188
+    },
+    {
+      "epoch": 2.3046857650455252,
+      "grad_norm": 0.49995652041639127,
+      "learning_rate": 4.542178216868746e-06,
+      "loss": 0.0394,
+      "step": 5189
+    },
+    {
+      "epoch": 2.3051299133910725,
+      "grad_norm": 0.38693338487127227,
+      "learning_rate": 4.540247858059654e-06,
+      "loss": 0.0244,
+      "step": 5190
+    },
+    {
+      "epoch": 2.3055740617366203,
+      "grad_norm": 0.42802350164666275,
+      "learning_rate": 4.538317568359117e-06,
+      "loss": 0.0234,
+      "step": 5191
+    },
+    {
+      "epoch": 2.3060182100821676,
+      "grad_norm": 0.33959690635237333,
+      "learning_rate": 4.536387348057292e-06,
+      "loss": 0.0273,
+      "step": 5192
+    },
+    {
+      "epoch": 2.306462358427715,
+      "grad_norm": 0.560656839558957,
+      "learning_rate": 4.5344571974443255e-06,
+      "loss": 0.0361,
+      "step": 5193
+    },
+    {
+      "epoch": 2.306906506773262,
+      "grad_norm": 0.3618868056091977,
+      "learning_rate": 4.5325271168103496e-06,
+      "loss": 0.0289,
+      "step": 5194
+    },
+    {
+      "epoch": 2.3073506551188094,
+      "grad_norm": 0.38869936597852134,
+      "learning_rate": 4.530597106445487e-06,
+      "loss": 0.0335,
+      "step": 5195
+    },
+    {
+      "epoch": 2.307794803464357,
+      "grad_norm": 0.451335091229613,
+      "learning_rate": 4.528667166639855e-06,
+      "loss": 0.0293,
+      "step": 5196
+    },
+    {
+      "epoch": 2.3082389518099045,
+      "grad_norm": 0.3756798518286319,
+      "learning_rate": 4.526737297683554e-06,
+      "loss": 0.0307,
+      "step": 5197
+    },
+    {
+      "epoch": 2.3086831001554518,
+      "grad_norm": 0.41638755620692586,
+      "learning_rate": 4.524807499866678e-06,
+      "loss": 0.0313,
+      "step": 5198
+    },
+    {
+      "epoch": 2.3091272485009995,
+      "grad_norm": 0.49512523947166825,
+      "learning_rate": 4.522877773479305e-06,
+      "loss": 0.0341,
+      "step": 5199
+    },
+    {
+      "epoch": 2.3095713968465468,
+      "grad_norm": 0.38110036222157506,
+      "learning_rate": 4.520948118811508e-06,
+      "loss": 0.0301,
+      "step": 5200
+    },
+    {
+      "epoch": 2.310015545192094,
+      "grad_norm": 0.4176831056196918,
+      "learning_rate": 4.519018536153346e-06,
+      "loss": 0.0325,
+      "step": 5201
+    },
+    {
+      "epoch": 2.3104596935376414,
+      "grad_norm": 0.37358674774847195,
+      "learning_rate": 4.517089025794869e-06,
+      "loss": 0.0223,
+      "step": 5202
+    },
+    {
+      "epoch": 2.310903841883189,
+      "grad_norm": 0.5722574491295905,
+      "learning_rate": 4.515159588026114e-06,
+      "loss": 0.0385,
+      "step": 5203
+    },
+    {
+      "epoch": 2.3113479902287364,
+      "grad_norm": 0.3452449788849875,
+      "learning_rate": 4.51323022313711e-06,
+      "loss": 0.0239,
+      "step": 5204
+    },
+    {
+      "epoch": 2.3117921385742837,
+      "grad_norm": 0.5594235169002529,
+      "learning_rate": 4.511300931417872e-06,
+      "loss": 0.0378,
+      "step": 5205
+    },
+    {
+      "epoch": 2.3122362869198314,
+      "grad_norm": 0.4629383899400429,
+      "learning_rate": 4.509371713158404e-06,
+      "loss": 0.0465,
+      "step": 5206
+    },
+    {
+      "epoch": 2.3126804352653787,
+      "grad_norm": 0.344536109032876,
+      "learning_rate": 4.507442568648702e-06,
+      "loss": 0.0214,
+      "step": 5207
+    },
+    {
+      "epoch": 2.313124583610926,
+      "grad_norm": 0.5327665523608216,
+      "learning_rate": 4.505513498178752e-06,
+      "loss": 0.0474,
+      "step": 5208
+    },
+    {
+      "epoch": 2.3135687319564733,
+      "grad_norm": 0.43197164521390413,
+      "learning_rate": 4.503584502038521e-06,
+      "loss": 0.0307,
+      "step": 5209
+    },
+    {
+      "epoch": 2.314012880302021,
+      "grad_norm": 0.46940506061763837,
+      "learning_rate": 4.501655580517972e-06,
+      "loss": 0.0405,
+      "step": 5210
+    },
+    {
+      "epoch": 2.3144570286475683,
+      "grad_norm": 0.3043189475892609,
+      "learning_rate": 4.499726733907056e-06,
+      "loss": 0.0208,
+      "step": 5211
+    },
+    {
+      "epoch": 2.3149011769931156,
+      "grad_norm": 0.5162978477725472,
+      "learning_rate": 4.497797962495707e-06,
+      "loss": 0.0417,
+      "step": 5212
+    },
+    {
+      "epoch": 2.3153453253386633,
+      "grad_norm": 0.4166068328105937,
+      "learning_rate": 4.495869266573857e-06,
+      "loss": 0.0363,
+      "step": 5213
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.5916403656140393,
+      "learning_rate": 4.49394064643142e-06,
+      "loss": 0.0401,
+      "step": 5214
+    },
+    {
+      "epoch": 2.316233622029758,
+      "grad_norm": 0.6022617123765319,
+      "learning_rate": 4.492012102358301e-06,
+      "loss": 0.0307,
+      "step": 5215
+    },
+    {
+      "epoch": 2.316677770375305,
+      "grad_norm": 0.4369169634282751,
+      "learning_rate": 4.490083634644391e-06,
+      "loss": 0.042,
+      "step": 5216
+    },
+    {
+      "epoch": 2.317121918720853,
+      "grad_norm": 0.5201884422005028,
+      "learning_rate": 4.488155243579574e-06,
+      "loss": 0.0353,
+      "step": 5217
+    },
+    {
+      "epoch": 2.3175660670664002,
+      "grad_norm": 0.4192988414508763,
+      "learning_rate": 4.486226929453716e-06,
+      "loss": 0.0347,
+      "step": 5218
+    },
+    {
+      "epoch": 2.3180102154119475,
+      "grad_norm": 0.47826969893341476,
+      "learning_rate": 4.4842986925566805e-06,
+      "loss": 0.0412,
+      "step": 5219
+    },
+    {
+      "epoch": 2.3184543637574953,
+      "grad_norm": 0.4411896998427744,
+      "learning_rate": 4.482370533178311e-06,
+      "loss": 0.0473,
+      "step": 5220
+    },
+    {
+      "epoch": 2.3188985121030425,
+      "grad_norm": 0.7154366399859141,
+      "learning_rate": 4.4804424516084435e-06,
+      "loss": 0.0318,
+      "step": 5221
+    },
+    {
+      "epoch": 2.31934266044859,
+      "grad_norm": 0.4486899236378927,
+      "learning_rate": 4.478514448136901e-06,
+      "loss": 0.03,
+      "step": 5222
+    },
+    {
+      "epoch": 2.319786808794137,
+      "grad_norm": 0.6696629896177303,
+      "learning_rate": 4.476586523053494e-06,
+      "loss": 0.0447,
+      "step": 5223
+    },
+    {
+      "epoch": 2.3202309571396844,
+      "grad_norm": 0.36764590834428945,
+      "learning_rate": 4.474658676648025e-06,
+      "loss": 0.0317,
+      "step": 5224
+    },
+    {
+      "epoch": 2.320675105485232,
+      "grad_norm": 0.4074350155089678,
+      "learning_rate": 4.4727309092102825e-06,
+      "loss": 0.0338,
+      "step": 5225
+    },
+    {
+      "epoch": 2.3211192538307794,
+      "grad_norm": 0.5410060660419875,
+      "learning_rate": 4.47080322103004e-06,
+      "loss": 0.0406,
+      "step": 5226
+    },
+    {
+      "epoch": 2.3215634021763267,
+      "grad_norm": 0.3869245785320258,
+      "learning_rate": 4.4688756123970625e-06,
+      "loss": 0.025,
+      "step": 5227
+    },
+    {
+      "epoch": 2.3220075505218745,
+      "grad_norm": 0.3306613006808203,
+      "learning_rate": 4.466948083601103e-06,
+      "loss": 0.0216,
+      "step": 5228
+    },
+    {
+      "epoch": 2.3224516988674218,
+      "grad_norm": 0.3885322170673341,
+      "learning_rate": 4.4650206349319e-06,
+      "loss": 0.0334,
+      "step": 5229
+    },
+    {
+      "epoch": 2.322895847212969,
+      "grad_norm": 0.45727532111911046,
+      "learning_rate": 4.463093266679185e-06,
+      "loss": 0.0314,
+      "step": 5230
+    },
+    {
+      "epoch": 2.3233399955585163,
+      "grad_norm": 0.40470162685593614,
+      "learning_rate": 4.4611659791326726e-06,
+      "loss": 0.0225,
+      "step": 5231
+    },
+    {
+      "epoch": 2.323784143904064,
+      "grad_norm": 0.3709631261080152,
+      "learning_rate": 4.459238772582067e-06,
+      "loss": 0.0281,
+      "step": 5232
+    },
+    {
+      "epoch": 2.3242282922496114,
+      "grad_norm": 0.5000330379973269,
+      "learning_rate": 4.457311647317058e-06,
+      "loss": 0.0408,
+      "step": 5233
+    },
+    {
+      "epoch": 2.3246724405951587,
+      "grad_norm": 0.40795990787328895,
+      "learning_rate": 4.4553846036273294e-06,
+      "loss": 0.0304,
+      "step": 5234
+    },
+    {
+      "epoch": 2.3251165889407064,
+      "grad_norm": 0.41652506953547214,
+      "learning_rate": 4.453457641802542e-06,
+      "loss": 0.0283,
+      "step": 5235
+    },
+    {
+      "epoch": 2.3255607372862537,
+      "grad_norm": 0.422544506737885,
+      "learning_rate": 4.451530762132359e-06,
+      "loss": 0.0247,
+      "step": 5236
+    },
+    {
+      "epoch": 2.326004885631801,
+      "grad_norm": 0.42941673106868544,
+      "learning_rate": 4.4496039649064185e-06,
+      "loss": 0.0271,
+      "step": 5237
+    },
+    {
+      "epoch": 2.3264490339773483,
+      "grad_norm": 0.5518213591251218,
+      "learning_rate": 4.4476772504143525e-06,
+      "loss": 0.0283,
+      "step": 5238
+    },
+    {
+      "epoch": 2.326893182322896,
+      "grad_norm": 0.3905942457168205,
+      "learning_rate": 4.445750618945778e-06,
+      "loss": 0.0338,
+      "step": 5239
+    },
+    {
+      "epoch": 2.3273373306684433,
+      "grad_norm": 0.5264154088819042,
+      "learning_rate": 4.443824070790298e-06,
+      "loss": 0.0373,
+      "step": 5240
+    },
+    {
+      "epoch": 2.3277814790139906,
+      "grad_norm": 0.4859563195815842,
+      "learning_rate": 4.4418976062375095e-06,
+      "loss": 0.0472,
+      "step": 5241
+    },
+    {
+      "epoch": 2.3282256273595383,
+      "grad_norm": 0.43082619257296717,
+      "learning_rate": 4.439971225576992e-06,
+      "loss": 0.0321,
+      "step": 5242
+    },
+    {
+      "epoch": 2.3286697757050856,
+      "grad_norm": 0.4148209810054243,
+      "learning_rate": 4.438044929098312e-06,
+      "loss": 0.027,
+      "step": 5243
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.3906796792959702,
+      "learning_rate": 4.436118717091025e-06,
+      "loss": 0.0302,
+      "step": 5244
+    },
+    {
+      "epoch": 2.32955807239618,
+      "grad_norm": 0.5074169188006878,
+      "learning_rate": 4.434192589844674e-06,
+      "loss": 0.033,
+      "step": 5245
+    },
+    {
+      "epoch": 2.330002220741728,
+      "grad_norm": 0.5700168097501661,
+      "learning_rate": 4.432266547648786e-06,
+      "loss": 0.0386,
+      "step": 5246
+    },
+    {
+      "epoch": 2.330446369087275,
+      "grad_norm": 0.7002894794054985,
+      "learning_rate": 4.430340590792883e-06,
+      "loss": 0.0452,
+      "step": 5247
+    },
+    {
+      "epoch": 2.3308905174328225,
+      "grad_norm": 0.43587113620478884,
+      "learning_rate": 4.428414719566464e-06,
+      "loss": 0.0455,
+      "step": 5248
+    },
+    {
+      "epoch": 2.33133466577837,
+      "grad_norm": 0.4414200442172319,
+      "learning_rate": 4.426488934259023e-06,
+      "loss": 0.038,
+      "step": 5249
+    },
+    {
+      "epoch": 2.3317788141239175,
+      "grad_norm": 0.3603980656114873,
+      "learning_rate": 4.424563235160039e-06,
+      "loss": 0.0284,
+      "step": 5250
+    },
+    {
+      "epoch": 2.332222962469465,
+      "grad_norm": 0.3831150016755091,
+      "learning_rate": 4.422637622558973e-06,
+      "loss": 0.0302,
+      "step": 5251
+    },
+    {
+      "epoch": 2.332667110815012,
+      "grad_norm": 0.3992416324305334,
+      "learning_rate": 4.42071209674528e-06,
+      "loss": 0.0324,
+      "step": 5252
+    },
+    {
+      "epoch": 2.3331112591605594,
+      "grad_norm": 0.4795655074267854,
+      "learning_rate": 4.4187866580084005e-06,
+      "loss": 0.0456,
+      "step": 5253
+    },
+    {
+      "epoch": 2.333555407506107,
+      "grad_norm": 0.3484662534276821,
+      "learning_rate": 4.41686130663776e-06,
+      "loss": 0.0223,
+      "step": 5254
+    },
+    {
+      "epoch": 2.3339995558516544,
+      "grad_norm": 0.4234114697000452,
+      "learning_rate": 4.4149360429227695e-06,
+      "loss": 0.0296,
+      "step": 5255
+    },
+    {
+      "epoch": 2.3344437041972017,
+      "grad_norm": 0.4632049279941834,
+      "learning_rate": 4.4130108671528315e-06,
+      "loss": 0.0264,
+      "step": 5256
+    },
+    {
+      "epoch": 2.3348878525427494,
+      "grad_norm": 0.5054504530004318,
+      "learning_rate": 4.41108577961733e-06,
+      "loss": 0.0423,
+      "step": 5257
+    },
+    {
+      "epoch": 2.3353320008882967,
+      "grad_norm": 0.4132515605214912,
+      "learning_rate": 4.40916078060564e-06,
+      "loss": 0.026,
+      "step": 5258
+    },
+    {
+      "epoch": 2.335776149233844,
+      "grad_norm": 0.4886399938932667,
+      "learning_rate": 4.407235870407122e-06,
+      "loss": 0.0378,
+      "step": 5259
+    },
+    {
+      "epoch": 2.3362202975793913,
+      "grad_norm": 0.48169156288056764,
+      "learning_rate": 4.4053110493111226e-06,
+      "loss": 0.0366,
+      "step": 5260
+    },
+    {
+      "epoch": 2.336664445924939,
+      "grad_norm": 0.5264104312524703,
+      "learning_rate": 4.403386317606972e-06,
+      "loss": 0.0427,
+      "step": 5261
+    },
+    {
+      "epoch": 2.3371085942704863,
+      "grad_norm": 0.3726652002849068,
+      "learning_rate": 4.4014616755839955e-06,
+      "loss": 0.026,
+      "step": 5262
+    },
+    {
+      "epoch": 2.3375527426160336,
+      "grad_norm": 0.531223484070124,
+      "learning_rate": 4.399537123531494e-06,
+      "loss": 0.0389,
+      "step": 5263
+    },
+    {
+      "epoch": 2.3379968909615814,
+      "grad_norm": 0.5377515081195285,
+      "learning_rate": 4.3976126617387645e-06,
+      "loss": 0.0466,
+      "step": 5264
+    },
+    {
+      "epoch": 2.3384410393071287,
+      "grad_norm": 0.4256267011064791,
+      "learning_rate": 4.395688290495084e-06,
+      "loss": 0.0381,
+      "step": 5265
+    },
+    {
+      "epoch": 2.338885187652676,
+      "grad_norm": 0.4792927506755932,
+      "learning_rate": 4.393764010089719e-06,
+      "loss": 0.046,
+      "step": 5266
+    },
+    {
+      "epoch": 2.3393293359982232,
+      "grad_norm": 0.5281025771868061,
+      "learning_rate": 4.391839820811923e-06,
+      "loss": 0.0396,
+      "step": 5267
+    },
+    {
+      "epoch": 2.339773484343771,
+      "grad_norm": 0.4893018865311666,
+      "learning_rate": 4.389915722950931e-06,
+      "loss": 0.0368,
+      "step": 5268
+    },
+    {
+      "epoch": 2.3402176326893183,
+      "grad_norm": 0.4896668042908599,
+      "learning_rate": 4.387991716795968e-06,
+      "loss": 0.0406,
+      "step": 5269
+    },
+    {
+      "epoch": 2.3406617810348656,
+      "grad_norm": 0.39529432125848185,
+      "learning_rate": 4.386067802636249e-06,
+      "loss": 0.0323,
+      "step": 5270
+    },
+    {
+      "epoch": 2.3411059293804133,
+      "grad_norm": 0.38393804306481244,
+      "learning_rate": 4.384143980760968e-06,
+      "loss": 0.0321,
+      "step": 5271
+    },
+    {
+      "epoch": 2.3415500777259606,
+      "grad_norm": 0.4030478157940605,
+      "learning_rate": 4.382220251459306e-06,
+      "loss": 0.0348,
+      "step": 5272
+    },
+    {
+      "epoch": 2.341994226071508,
+      "grad_norm": 0.46335758617758316,
+      "learning_rate": 4.380296615020437e-06,
+      "loss": 0.0481,
+      "step": 5273
+    },
+    {
+      "epoch": 2.342438374417055,
+      "grad_norm": 0.39799741674310996,
+      "learning_rate": 4.3783730717335124e-06,
+      "loss": 0.0365,
+      "step": 5274
+    },
+    {
+      "epoch": 2.342882522762603,
+      "grad_norm": 0.6313177165283153,
+      "learning_rate": 4.376449621887674e-06,
+      "loss": 0.0357,
+      "step": 5275
+    },
+    {
+      "epoch": 2.34332667110815,
+      "grad_norm": 0.46331907567288616,
+      "learning_rate": 4.37452626577205e-06,
+      "loss": 0.0404,
+      "step": 5276
+    },
+    {
+      "epoch": 2.3437708194536975,
+      "grad_norm": 0.45182075889541223,
+      "learning_rate": 4.372603003675755e-06,
+      "loss": 0.0273,
+      "step": 5277
+    },
+    {
+      "epoch": 2.3442149677992448,
+      "grad_norm": 0.45396241594109277,
+      "learning_rate": 4.370679835887885e-06,
+      "loss": 0.0334,
+      "step": 5278
+    },
+    {
+      "epoch": 2.3446591161447925,
+      "grad_norm": 0.4008935982764722,
+      "learning_rate": 4.368756762697525e-06,
+      "loss": 0.0295,
+      "step": 5279
+    },
+    {
+      "epoch": 2.34510326449034,
+      "grad_norm": 0.4497120560175991,
+      "learning_rate": 4.366833784393746e-06,
+      "loss": 0.034,
+      "step": 5280
+    },
+    {
+      "epoch": 2.345547412835887,
+      "grad_norm": 0.5274538552861234,
+      "learning_rate": 4.364910901265607e-06,
+      "loss": 0.0358,
+      "step": 5281
+    },
+    {
+      "epoch": 2.3459915611814344,
+      "grad_norm": 0.5012732757402258,
+      "learning_rate": 4.362988113602147e-06,
+      "loss": 0.0351,
+      "step": 5282
+    },
+    {
+      "epoch": 2.346435709526982,
+      "grad_norm": 0.41702907826337277,
+      "learning_rate": 4.361065421692394e-06,
+      "loss": 0.0311,
+      "step": 5283
+    },
+    {
+      "epoch": 2.3468798578725294,
+      "grad_norm": 0.4723862662210319,
+      "learning_rate": 4.3591428258253634e-06,
+      "loss": 0.0283,
+      "step": 5284
+    },
+    {
+      "epoch": 2.3473240062180767,
+      "grad_norm": 0.4642888220640226,
+      "learning_rate": 4.35722032629005e-06,
+      "loss": 0.0341,
+      "step": 5285
+    },
+    {
+      "epoch": 2.3477681545636244,
+      "grad_norm": 0.5515201989991365,
+      "learning_rate": 4.35529792337544e-06,
+      "loss": 0.0426,
+      "step": 5286
+    },
+    {
+      "epoch": 2.3482123029091717,
+      "grad_norm": 0.42814364810008887,
+      "learning_rate": 4.353375617370506e-06,
+      "loss": 0.0386,
+      "step": 5287
+    },
+    {
+      "epoch": 2.348656451254719,
+      "grad_norm": 0.422713836528967,
+      "learning_rate": 4.3514534085642e-06,
+      "loss": 0.0477,
+      "step": 5288
+    },
+    {
+      "epoch": 2.3491005996002663,
+      "grad_norm": 0.388857891163739,
+      "learning_rate": 4.349531297245464e-06,
+      "loss": 0.0284,
+      "step": 5289
+    },
+    {
+      "epoch": 2.349544747945814,
+      "grad_norm": 0.33430013810360537,
+      "learning_rate": 4.347609283703224e-06,
+      "loss": 0.0222,
+      "step": 5290
+    },
+    {
+      "epoch": 2.3499888962913613,
+      "grad_norm": 0.4550663748035299,
+      "learning_rate": 4.345687368226391e-06,
+      "loss": 0.0484,
+      "step": 5291
+    },
+    {
+      "epoch": 2.3504330446369086,
+      "grad_norm": 0.4904858360330206,
+      "learning_rate": 4.343765551103859e-06,
+      "loss": 0.0431,
+      "step": 5292
+    },
+    {
+      "epoch": 2.3508771929824563,
+      "grad_norm": 0.418865690563144,
+      "learning_rate": 4.3418438326245134e-06,
+      "loss": 0.0264,
+      "step": 5293
+    },
+    {
+      "epoch": 2.3513213413280036,
+      "grad_norm": 0.3313326958383252,
+      "learning_rate": 4.33992221307722e-06,
+      "loss": 0.0359,
+      "step": 5294
+    },
+    {
+      "epoch": 2.351765489673551,
+      "grad_norm": 0.38342759888274064,
+      "learning_rate": 4.338000692750832e-06,
+      "loss": 0.03,
+      "step": 5295
+    },
+    {
+      "epoch": 2.352209638019098,
+      "grad_norm": 0.6016424768870177,
+      "learning_rate": 4.336079271934184e-06,
+      "loss": 0.0432,
+      "step": 5296
+    },
+    {
+      "epoch": 2.352653786364646,
+      "grad_norm": 0.5554548071020693,
+      "learning_rate": 4.334157950916098e-06,
+      "loss": 0.0659,
+      "step": 5297
+    },
+    {
+      "epoch": 2.3530979347101932,
+      "grad_norm": 0.5014661231015702,
+      "learning_rate": 4.332236729985385e-06,
+      "loss": 0.0343,
+      "step": 5298
+    },
+    {
+      "epoch": 2.3535420830557405,
+      "grad_norm": 0.5146747988924121,
+      "learning_rate": 4.330315609430835e-06,
+      "loss": 0.0351,
+      "step": 5299
+    },
+    {
+      "epoch": 2.3539862314012883,
+      "grad_norm": 0.3876173155044,
+      "learning_rate": 4.328394589541223e-06,
+      "loss": 0.0282,
+      "step": 5300
+    },
+    {
+      "epoch": 2.3544303797468356,
+      "grad_norm": 0.414824658223231,
+      "learning_rate": 4.326473670605315e-06,
+      "loss": 0.0323,
+      "step": 5301
+    },
+    {
+      "epoch": 2.354874528092383,
+      "grad_norm": 0.47728404717752354,
+      "learning_rate": 4.324552852911854e-06,
+      "loss": 0.0459,
+      "step": 5302
+    },
+    {
+      "epoch": 2.35531867643793,
+      "grad_norm": 0.40142086992053616,
+      "learning_rate": 4.322632136749572e-06,
+      "loss": 0.0329,
+      "step": 5303
+    },
+    {
+      "epoch": 2.355762824783478,
+      "grad_norm": 0.4704980131084727,
+      "learning_rate": 4.3207115224071874e-06,
+      "loss": 0.0366,
+      "step": 5304
+    },
+    {
+      "epoch": 2.356206973129025,
+      "grad_norm": 0.402777301290612,
+      "learning_rate": 4.318791010173401e-06,
+      "loss": 0.0319,
+      "step": 5305
+    },
+    {
+      "epoch": 2.3566511214745725,
+      "grad_norm": 0.41278158938376935,
+      "learning_rate": 4.316870600336896e-06,
+      "loss": 0.0217,
+      "step": 5306
+    },
+    {
+      "epoch": 2.3570952698201197,
+      "grad_norm": 0.5124207452989102,
+      "learning_rate": 4.314950293186346e-06,
+      "loss": 0.0398,
+      "step": 5307
+    },
+    {
+      "epoch": 2.3575394181656675,
+      "grad_norm": 0.4244779016373727,
+      "learning_rate": 4.3130300890104035e-06,
+      "loss": 0.0263,
+      "step": 5308
+    },
+    {
+      "epoch": 2.3579835665112148,
+      "grad_norm": 0.6692076129286184,
+      "learning_rate": 4.311109988097706e-06,
+      "loss": 0.0347,
+      "step": 5309
+    },
+    {
+      "epoch": 2.358427714856762,
+      "grad_norm": 0.5345819569804751,
+      "learning_rate": 4.30918999073688e-06,
+      "loss": 0.0406,
+      "step": 5310
+    },
+    {
+      "epoch": 2.3588718632023093,
+      "grad_norm": 0.4416731105459865,
+      "learning_rate": 4.307270097216535e-06,
+      "loss": 0.0478,
+      "step": 5311
+    },
+    {
+      "epoch": 2.359316011547857,
+      "grad_norm": 0.4789720007194078,
+      "learning_rate": 4.305350307825261e-06,
+      "loss": 0.0347,
+      "step": 5312
+    },
+    {
+      "epoch": 2.3597601598934044,
+      "grad_norm": 0.39908333172506805,
+      "learning_rate": 4.303430622851635e-06,
+      "loss": 0.0346,
+      "step": 5313
+    },
+    {
+      "epoch": 2.3602043082389517,
+      "grad_norm": 0.49076992885563064,
+      "learning_rate": 4.301511042584219e-06,
+      "loss": 0.035,
+      "step": 5314
+    },
+    {
+      "epoch": 2.3606484565844994,
+      "grad_norm": 0.3862322602452697,
+      "learning_rate": 4.29959156731156e-06,
+      "loss": 0.0409,
+      "step": 5315
+    },
+    {
+      "epoch": 2.3610926049300467,
+      "grad_norm": 0.4148195535388891,
+      "learning_rate": 4.297672197322186e-06,
+      "loss": 0.0382,
+      "step": 5316
+    },
+    {
+      "epoch": 2.361536753275594,
+      "grad_norm": 0.5373449885131869,
+      "learning_rate": 4.29575293290461e-06,
+      "loss": 0.0371,
+      "step": 5317
+    },
+    {
+      "epoch": 2.3619809016211413,
+      "grad_norm": 0.41088601906859734,
+      "learning_rate": 4.293833774347333e-06,
+      "loss": 0.0297,
+      "step": 5318
+    },
+    {
+      "epoch": 2.362425049966689,
+      "grad_norm": 1.1416110662239514,
+      "learning_rate": 4.291914721938835e-06,
+      "loss": 0.0382,
+      "step": 5319
+    },
+    {
+      "epoch": 2.3628691983122363,
+      "grad_norm": 0.4504483504568535,
+      "learning_rate": 4.289995775967581e-06,
+      "loss": 0.0297,
+      "step": 5320
+    },
+    {
+      "epoch": 2.3633133466577836,
+      "grad_norm": 0.38320982061969394,
+      "learning_rate": 4.2880769367220234e-06,
+      "loss": 0.024,
+      "step": 5321
+    },
+    {
+      "epoch": 2.3637574950033313,
+      "grad_norm": 0.3866521874991146,
+      "learning_rate": 4.2861582044905966e-06,
+      "loss": 0.0264,
+      "step": 5322
+    },
+    {
+      "epoch": 2.3642016433488786,
+      "grad_norm": 0.5381622631772952,
+      "learning_rate": 4.284239579561718e-06,
+      "loss": 0.0514,
+      "step": 5323
+    },
+    {
+      "epoch": 2.364645791694426,
+      "grad_norm": 0.4285272602850889,
+      "learning_rate": 4.282321062223788e-06,
+      "loss": 0.0365,
+      "step": 5324
+    },
+    {
+      "epoch": 2.365089940039973,
+      "grad_norm": 0.5903609551581849,
+      "learning_rate": 4.280402652765194e-06,
+      "loss": 0.0428,
+      "step": 5325
+    },
+    {
+      "epoch": 2.365534088385521,
+      "grad_norm": 0.6051857396994949,
+      "learning_rate": 4.278484351474303e-06,
+      "loss": 0.0336,
+      "step": 5326
+    },
+    {
+      "epoch": 2.365978236731068,
+      "grad_norm": 0.569315522539356,
+      "learning_rate": 4.2765661586394736e-06,
+      "loss": 0.0334,
+      "step": 5327
+    },
+    {
+      "epoch": 2.3664223850766155,
+      "grad_norm": 0.6976731675123818,
+      "learning_rate": 4.2746480745490385e-06,
+      "loss": 0.0367,
+      "step": 5328
+    },
+    {
+      "epoch": 2.3668665334221632,
+      "grad_norm": 0.5611945388042203,
+      "learning_rate": 4.272730099491319e-06,
+      "loss": 0.0407,
+      "step": 5329
+    },
+    {
+      "epoch": 2.3673106817677105,
+      "grad_norm": 0.49730153273797245,
+      "learning_rate": 4.27081223375462e-06,
+      "loss": 0.038,
+      "step": 5330
+    },
+    {
+      "epoch": 2.367754830113258,
+      "grad_norm": 0.5313463154313218,
+      "learning_rate": 4.268894477627229e-06,
+      "loss": 0.0435,
+      "step": 5331
+    },
+    {
+      "epoch": 2.368198978458805,
+      "grad_norm": 0.48983566744126533,
+      "learning_rate": 4.2669768313974155e-06,
+      "loss": 0.0373,
+      "step": 5332
+    },
+    {
+      "epoch": 2.3686431268043524,
+      "grad_norm": 0.5179274904316602,
+      "learning_rate": 4.265059295353439e-06,
+      "loss": 0.0445,
+      "step": 5333
+    },
+    {
+      "epoch": 2.3690872751499,
+      "grad_norm": 0.5588093911942073,
+      "learning_rate": 4.2631418697835335e-06,
+      "loss": 0.0351,
+      "step": 5334
+    },
+    {
+      "epoch": 2.3695314234954474,
+      "grad_norm": 0.3586724840114786,
+      "learning_rate": 4.261224554975923e-06,
+      "loss": 0.0214,
+      "step": 5335
+    },
+    {
+      "epoch": 2.3699755718409947,
+      "grad_norm": 0.418317991811463,
+      "learning_rate": 4.259307351218812e-06,
+      "loss": 0.0301,
+      "step": 5336
+    },
+    {
+      "epoch": 2.3704197201865425,
+      "grad_norm": 0.33540637763336467,
+      "learning_rate": 4.2573902588003844e-06,
+      "loss": 0.023,
+      "step": 5337
+    },
+    {
+      "epoch": 2.3708638685320897,
+      "grad_norm": 0.45375734251758604,
+      "learning_rate": 4.2554732780088185e-06,
+      "loss": 0.0394,
+      "step": 5338
+    },
+    {
+      "epoch": 2.371308016877637,
+      "grad_norm": 0.5337079104144298,
+      "learning_rate": 4.253556409132267e-06,
+      "loss": 0.0441,
+      "step": 5339
+    },
+    {
+      "epoch": 2.3717521652231843,
+      "grad_norm": 0.5739232408785224,
+      "learning_rate": 4.251639652458866e-06,
+      "loss": 0.0417,
+      "step": 5340
+    },
+    {
+      "epoch": 2.372196313568732,
+      "grad_norm": 0.6130343197365772,
+      "learning_rate": 4.249723008276737e-06,
+      "loss": 0.0387,
+      "step": 5341
+    },
+    {
+      "epoch": 2.3726404619142794,
+      "grad_norm": 0.4104117521503695,
+      "learning_rate": 4.247806476873987e-06,
+      "loss": 0.033,
+      "step": 5342
+    },
+    {
+      "epoch": 2.3730846102598266,
+      "grad_norm": 0.48677001689089716,
+      "learning_rate": 4.245890058538697e-06,
+      "loss": 0.0377,
+      "step": 5343
+    },
+    {
+      "epoch": 2.3735287586053744,
+      "grad_norm": 0.3714192923602575,
+      "learning_rate": 4.2439737535589455e-06,
+      "loss": 0.0294,
+      "step": 5344
+    },
+    {
+      "epoch": 2.3739729069509217,
+      "grad_norm": 0.40375546238165533,
+      "learning_rate": 4.2420575622227786e-06,
+      "loss": 0.0314,
+      "step": 5345
+    },
+    {
+      "epoch": 2.374417055296469,
+      "grad_norm": 0.7948951370956706,
+      "learning_rate": 4.240141484818238e-06,
+      "loss": 0.0393,
+      "step": 5346
+    },
+    {
+      "epoch": 2.3748612036420162,
+      "grad_norm": 0.7705250912114215,
+      "learning_rate": 4.238225521633339e-06,
+      "loss": 0.0446,
+      "step": 5347
+    },
+    {
+      "epoch": 2.375305351987564,
+      "grad_norm": 0.5226003605622348,
+      "learning_rate": 4.2363096729560824e-06,
+      "loss": 0.0397,
+      "step": 5348
+    },
+    {
+      "epoch": 2.3757495003331113,
+      "grad_norm": 0.37536659049033705,
+      "learning_rate": 4.234393939074456e-06,
+      "loss": 0.0264,
+      "step": 5349
+    },
+    {
+      "epoch": 2.3761936486786586,
+      "grad_norm": 0.35662293585999577,
+      "learning_rate": 4.2324783202764265e-06,
+      "loss": 0.0294,
+      "step": 5350
+    },
+    {
+      "epoch": 2.3766377970242063,
+      "grad_norm": 0.5673307864129802,
+      "learning_rate": 4.230562816849944e-06,
+      "loss": 0.0471,
+      "step": 5351
+    },
+    {
+      "epoch": 2.3770819453697536,
+      "grad_norm": 0.42418386406735814,
+      "learning_rate": 4.228647429082939e-06,
+      "loss": 0.0306,
+      "step": 5352
+    },
+    {
+      "epoch": 2.377526093715301,
+      "grad_norm": 0.5746067334007909,
+      "learning_rate": 4.22673215726333e-06,
+      "loss": 0.031,
+      "step": 5353
+    },
+    {
+      "epoch": 2.377970242060848,
+      "grad_norm": 0.3492096810181002,
+      "learning_rate": 4.224817001679011e-06,
+      "loss": 0.021,
+      "step": 5354
+    },
+    {
+      "epoch": 2.378414390406396,
+      "grad_norm": 0.5532015823221401,
+      "learning_rate": 4.222901962617867e-06,
+      "loss": 0.0432,
+      "step": 5355
+    },
+    {
+      "epoch": 2.378858538751943,
+      "grad_norm": 0.5532405005608227,
+      "learning_rate": 4.220987040367757e-06,
+      "loss": 0.0407,
+      "step": 5356
+    },
+    {
+      "epoch": 2.3793026870974905,
+      "grad_norm": 0.35180832498879067,
+      "learning_rate": 4.219072235216529e-06,
+      "loss": 0.028,
+      "step": 5357
+    },
+    {
+      "epoch": 2.379746835443038,
+      "grad_norm": 0.7448501299403424,
+      "learning_rate": 4.2171575474520084e-06,
+      "loss": 0.0328,
+      "step": 5358
+    },
+    {
+      "epoch": 2.3801909837885855,
+      "grad_norm": 0.4554992316854567,
+      "learning_rate": 4.215242977362009e-06,
+      "loss": 0.0301,
+      "step": 5359
+    },
+    {
+      "epoch": 2.380635132134133,
+      "grad_norm": 0.5712009195638152,
+      "learning_rate": 4.213328525234317e-06,
+      "loss": 0.0344,
+      "step": 5360
+    },
+    {
+      "epoch": 2.38107928047968,
+      "grad_norm": 0.46746160638085626,
+      "learning_rate": 4.211414191356714e-06,
+      "loss": 0.0387,
+      "step": 5361
+    },
+    {
+      "epoch": 2.3815234288252274,
+      "grad_norm": 0.4318040553120997,
+      "learning_rate": 4.209499976016953e-06,
+      "loss": 0.0359,
+      "step": 5362
+    },
+    {
+      "epoch": 2.381967577170775,
+      "grad_norm": 0.4278461319811184,
+      "learning_rate": 4.2075858795027745e-06,
+      "loss": 0.0368,
+      "step": 5363
+    },
+    {
+      "epoch": 2.3824117255163224,
+      "grad_norm": 0.32663597876713696,
+      "learning_rate": 4.205671902101899e-06,
+      "loss": 0.0256,
+      "step": 5364
+    },
+    {
+      "epoch": 2.3828558738618697,
+      "grad_norm": 0.389353159520307,
+      "learning_rate": 4.203758044102029e-06,
+      "loss": 0.0259,
+      "step": 5365
+    },
+    {
+      "epoch": 2.3833000222074174,
+      "grad_norm": 0.5653053154481631,
+      "learning_rate": 4.2018443057908495e-06,
+      "loss": 0.047,
+      "step": 5366
+    },
+    {
+      "epoch": 2.3837441705529647,
+      "grad_norm": 0.34982509990328564,
+      "learning_rate": 4.199930687456031e-06,
+      "loss": 0.0267,
+      "step": 5367
+    },
+    {
+      "epoch": 2.384188318898512,
+      "grad_norm": 0.41740396260225204,
+      "learning_rate": 4.198017189385221e-06,
+      "loss": 0.0293,
+      "step": 5368
+    },
+    {
+      "epoch": 2.3846324672440593,
+      "grad_norm": 0.38322501636096945,
+      "learning_rate": 4.1961038118660504e-06,
+      "loss": 0.0329,
+      "step": 5369
+    },
+    {
+      "epoch": 2.385076615589607,
+      "grad_norm": 0.45223291597674525,
+      "learning_rate": 4.194190555186133e-06,
+      "loss": 0.0347,
+      "step": 5370
+    },
+    {
+      "epoch": 2.3855207639351543,
+      "grad_norm": 0.5113109910932524,
+      "learning_rate": 4.1922774196330614e-06,
+      "loss": 0.037,
+      "step": 5371
+    },
+    {
+      "epoch": 2.3859649122807016,
+      "grad_norm": 0.4161439641942477,
+      "learning_rate": 4.190364405494417e-06,
+      "loss": 0.032,
+      "step": 5372
+    },
+    {
+      "epoch": 2.3864090606262494,
+      "grad_norm": 0.3241401849079966,
+      "learning_rate": 4.1884515130577545e-06,
+      "loss": 0.0277,
+      "step": 5373
+    },
+    {
+      "epoch": 2.3868532089717966,
+      "grad_norm": 0.41604888730983397,
+      "learning_rate": 4.1865387426106165e-06,
+      "loss": 0.0342,
+      "step": 5374
+    },
+    {
+      "epoch": 2.387297357317344,
+      "grad_norm": 0.4441137282989938,
+      "learning_rate": 4.184626094440524e-06,
+      "loss": 0.0284,
+      "step": 5375
+    },
+    {
+      "epoch": 2.3877415056628912,
+      "grad_norm": 0.5249221470565395,
+      "learning_rate": 4.182713568834979e-06,
+      "loss": 0.0363,
+      "step": 5376
+    },
+    {
+      "epoch": 2.388185654008439,
+      "grad_norm": 0.5338153166350271,
+      "learning_rate": 4.180801166081466e-06,
+      "loss": 0.0505,
+      "step": 5377
+    },
+    {
+      "epoch": 2.3886298023539863,
+      "grad_norm": 0.6473388727476899,
+      "learning_rate": 4.178888886467457e-06,
+      "loss": 0.0426,
+      "step": 5378
+    },
+    {
+      "epoch": 2.3890739506995335,
+      "grad_norm": 0.362371921356082,
+      "learning_rate": 4.176976730280396e-06,
+      "loss": 0.0367,
+      "step": 5379
+    },
+    {
+      "epoch": 2.3895180990450813,
+      "grad_norm": 0.401397401331949,
+      "learning_rate": 4.175064697807712e-06,
+      "loss": 0.0315,
+      "step": 5380
+    },
+    {
+      "epoch": 2.3899622473906286,
+      "grad_norm": 0.4918164976839154,
+      "learning_rate": 4.173152789336818e-06,
+      "loss": 0.0377,
+      "step": 5381
+    },
+    {
+      "epoch": 2.390406395736176,
+      "grad_norm": 0.4428239856696088,
+      "learning_rate": 4.171241005155105e-06,
+      "loss": 0.0424,
+      "step": 5382
+    },
+    {
+      "epoch": 2.390850544081723,
+      "grad_norm": 0.5071830732288508,
+      "learning_rate": 4.169329345549945e-06,
+      "loss": 0.0343,
+      "step": 5383
+    },
+    {
+      "epoch": 2.391294692427271,
+      "grad_norm": 0.5154632693129214,
+      "learning_rate": 4.167417810808698e-06,
+      "loss": 0.0342,
+      "step": 5384
+    },
+    {
+      "epoch": 2.391738840772818,
+      "grad_norm": 0.6827121614415474,
+      "learning_rate": 4.165506401218697e-06,
+      "loss": 0.0537,
+      "step": 5385
+    },
+    {
+      "epoch": 2.3921829891183655,
+      "grad_norm": 0.5360898324290861,
+      "learning_rate": 4.163595117067258e-06,
+      "loss": 0.0415,
+      "step": 5386
+    },
+    {
+      "epoch": 2.3926271374639128,
+      "grad_norm": 0.4268060401836661,
+      "learning_rate": 4.1616839586416825e-06,
+      "loss": 0.0295,
+      "step": 5387
+    },
+    {
+      "epoch": 2.3930712858094605,
+      "grad_norm": 0.6427383929847603,
+      "learning_rate": 4.159772926229247e-06,
+      "loss": 0.0312,
+      "step": 5388
+    },
+    {
+      "epoch": 2.393515434155008,
+      "grad_norm": 0.5325554425749491,
+      "learning_rate": 4.1578620201172144e-06,
+      "loss": 0.0354,
+      "step": 5389
+    },
+    {
+      "epoch": 2.393959582500555,
+      "grad_norm": 0.43938556377430205,
+      "learning_rate": 4.155951240592825e-06,
+      "loss": 0.0372,
+      "step": 5390
+    },
+    {
+      "epoch": 2.3944037308461024,
+      "grad_norm": 0.5152442233098135,
+      "learning_rate": 4.154040587943303e-06,
+      "loss": 0.0402,
+      "step": 5391
+    },
+    {
+      "epoch": 2.39484787919165,
+      "grad_norm": 0.42703323082019723,
+      "learning_rate": 4.1521300624558516e-06,
+      "loss": 0.0373,
+      "step": 5392
+    },
+    {
+      "epoch": 2.3952920275371974,
+      "grad_norm": 0.402722970762066,
+      "learning_rate": 4.150219664417653e-06,
+      "loss": 0.0311,
+      "step": 5393
+    },
+    {
+      "epoch": 2.3957361758827447,
+      "grad_norm": 0.6214432623647976,
+      "learning_rate": 4.148309394115872e-06,
+      "loss": 0.0364,
+      "step": 5394
+    },
+    {
+      "epoch": 2.3961803242282924,
+      "grad_norm": 0.44276990850604514,
+      "learning_rate": 4.14639925183766e-06,
+      "loss": 0.0341,
+      "step": 5395
+    },
+    {
+      "epoch": 2.3966244725738397,
+      "grad_norm": 0.4294877304529703,
+      "learning_rate": 4.144489237870141e-06,
+      "loss": 0.0298,
+      "step": 5396
+    },
+    {
+      "epoch": 2.397068620919387,
+      "grad_norm": 0.46553831317336336,
+      "learning_rate": 4.142579352500421e-06,
+      "loss": 0.0318,
+      "step": 5397
+    },
+    {
+      "epoch": 2.3975127692649343,
+      "grad_norm": 0.4319310212143851,
+      "learning_rate": 4.14066959601559e-06,
+      "loss": 0.0412,
+      "step": 5398
+    },
+    {
+      "epoch": 2.397956917610482,
+      "grad_norm": 0.6999580389169097,
+      "learning_rate": 4.138759968702716e-06,
+      "loss": 0.0496,
+      "step": 5399
+    },
+    {
+      "epoch": 2.3984010659560293,
+      "grad_norm": 0.6261420455213397,
+      "learning_rate": 4.1368504708488476e-06,
+      "loss": 0.0387,
+      "step": 5400
+    },
+    {
+      "epoch": 2.3988452143015766,
+      "grad_norm": 0.352239243573041,
+      "learning_rate": 4.134941102741016e-06,
+      "loss": 0.0286,
+      "step": 5401
+    },
+    {
+      "epoch": 2.3992893626471243,
+      "grad_norm": 0.4855267366461917,
+      "learning_rate": 4.133031864666232e-06,
+      "loss": 0.042,
+      "step": 5402
+    },
+    {
+      "epoch": 2.3997335109926716,
+      "grad_norm": 0.48074181838542007,
+      "learning_rate": 4.1311227569114855e-06,
+      "loss": 0.0516,
+      "step": 5403
+    },
+    {
+      "epoch": 2.400177659338219,
+      "grad_norm": 0.4770686273769997,
+      "learning_rate": 4.12921377976375e-06,
+      "loss": 0.0378,
+      "step": 5404
+    },
+    {
+      "epoch": 2.400621807683766,
+      "grad_norm": 0.5247613321525735,
+      "learning_rate": 4.127304933509972e-06,
+      "loss": 0.0414,
+      "step": 5405
+    },
+    {
+      "epoch": 2.401065956029314,
+      "grad_norm": 0.4623711119275284,
+      "learning_rate": 4.125396218437089e-06,
+      "loss": 0.0589,
+      "step": 5406
+    },
+    {
+      "epoch": 2.4015101043748612,
+      "grad_norm": 0.49698176360744994,
+      "learning_rate": 4.123487634832011e-06,
+      "loss": 0.0456,
+      "step": 5407
+    },
+    {
+      "epoch": 2.4019542527204085,
+      "grad_norm": 0.4203922139028562,
+      "learning_rate": 4.121579182981632e-06,
+      "loss": 0.0312,
+      "step": 5408
+    },
+    {
+      "epoch": 2.4023984010659563,
+      "grad_norm": 0.3669217976875289,
+      "learning_rate": 4.119670863172824e-06,
+      "loss": 0.0289,
+      "step": 5409
+    },
+    {
+      "epoch": 2.4028425494115035,
+      "grad_norm": 0.38804870930885565,
+      "learning_rate": 4.117762675692437e-06,
+      "loss": 0.0251,
+      "step": 5410
+    },
+    {
+      "epoch": 2.403286697757051,
+      "grad_norm": 0.45435459736787537,
+      "learning_rate": 4.115854620827306e-06,
+      "loss": 0.0257,
+      "step": 5411
+    },
+    {
+      "epoch": 2.403730846102598,
+      "grad_norm": 0.46297467981750196,
+      "learning_rate": 4.1139466988642475e-06,
+      "loss": 0.0433,
+      "step": 5412
+    },
+    {
+      "epoch": 2.404174994448146,
+      "grad_norm": 0.4824849195132302,
+      "learning_rate": 4.11203891009005e-06,
+      "loss": 0.0484,
+      "step": 5413
+    },
+    {
+      "epoch": 2.404619142793693,
+      "grad_norm": 0.45065972785277525,
+      "learning_rate": 4.110131254791489e-06,
+      "loss": 0.028,
+      "step": 5414
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.5431313239180846,
+      "learning_rate": 4.108223733255316e-06,
+      "loss": 0.0442,
+      "step": 5415
+    },
+    {
+      "epoch": 2.4055074394847877,
+      "grad_norm": 0.46821071422858296,
+      "learning_rate": 4.106316345768265e-06,
+      "loss": 0.027,
+      "step": 5416
+    },
+    {
+      "epoch": 2.4059515878303355,
+      "grad_norm": 0.4155972801004647,
+      "learning_rate": 4.104409092617047e-06,
+      "loss": 0.033,
+      "step": 5417
+    },
+    {
+      "epoch": 2.4063957361758828,
+      "grad_norm": 0.6893775471251186,
+      "learning_rate": 4.1025019740883556e-06,
+      "loss": 0.0504,
+      "step": 5418
+    },
+    {
+      "epoch": 2.40683988452143,
+      "grad_norm": 0.4083036792334355,
+      "learning_rate": 4.100594990468865e-06,
+      "loss": 0.0333,
+      "step": 5419
+    },
+    {
+      "epoch": 2.4072840328669773,
+      "grad_norm": 0.4849564201026396,
+      "learning_rate": 4.0986881420452254e-06,
+      "loss": 0.0338,
+      "step": 5420
+    },
+    {
+      "epoch": 2.407728181212525,
+      "grad_norm": 0.4268462015661056,
+      "learning_rate": 4.096781429104068e-06,
+      "loss": 0.0341,
+      "step": 5421
+    },
+    {
+      "epoch": 2.4081723295580724,
+      "grad_norm": 0.4834546635420465,
+      "learning_rate": 4.094874851932002e-06,
+      "loss": 0.035,
+      "step": 5422
+    },
+    {
+      "epoch": 2.4086164779036197,
+      "grad_norm": 0.46585094752698974,
+      "learning_rate": 4.092968410815625e-06,
+      "loss": 0.0288,
+      "step": 5423
+    },
+    {
+      "epoch": 2.4090606262491674,
+      "grad_norm": 0.6920655509720604,
+      "learning_rate": 4.091062106041504e-06,
+      "loss": 0.0296,
+      "step": 5424
+    },
+    {
+      "epoch": 2.4095047745947147,
+      "grad_norm": 0.5102739827979796,
+      "learning_rate": 4.089155937896187e-06,
+      "loss": 0.0316,
+      "step": 5425
+    },
+    {
+      "epoch": 2.409948922940262,
+      "grad_norm": 0.36415630770650137,
+      "learning_rate": 4.087249906666206e-06,
+      "loss": 0.0296,
+      "step": 5426
+    },
+    {
+      "epoch": 2.4103930712858093,
+      "grad_norm": 0.3263196544116076,
+      "learning_rate": 4.085344012638067e-06,
+      "loss": 0.0273,
+      "step": 5427
+    },
+    {
+      "epoch": 2.410837219631357,
+      "grad_norm": 0.5467141941835238,
+      "learning_rate": 4.083438256098261e-06,
+      "loss": 0.0368,
+      "step": 5428
+    },
+    {
+      "epoch": 2.4112813679769043,
+      "grad_norm": 0.5887360709523032,
+      "learning_rate": 4.081532637333255e-06,
+      "loss": 0.0389,
+      "step": 5429
+    },
+    {
+      "epoch": 2.4117255163224516,
+      "grad_norm": 0.4074807509483992,
+      "learning_rate": 4.079627156629497e-06,
+      "loss": 0.0239,
+      "step": 5430
+    },
+    {
+      "epoch": 2.4121696646679993,
+      "grad_norm": 0.3424789553984037,
+      "learning_rate": 4.07772181427341e-06,
+      "loss": 0.0301,
+      "step": 5431
+    },
+    {
+      "epoch": 2.4126138130135466,
+      "grad_norm": 0.607202231589749,
+      "learning_rate": 4.075816610551402e-06,
+      "loss": 0.0496,
+      "step": 5432
+    },
+    {
+      "epoch": 2.413057961359094,
+      "grad_norm": 0.47864092786503787,
+      "learning_rate": 4.073911545749857e-06,
+      "loss": 0.0316,
+      "step": 5433
+    },
+    {
+      "epoch": 2.413502109704641,
+      "grad_norm": 0.5034557440074656,
+      "learning_rate": 4.072006620155136e-06,
+      "loss": 0.026,
+      "step": 5434
+    },
+    {
+      "epoch": 2.413946258050189,
+      "grad_norm": 0.36403294815194714,
+      "learning_rate": 4.070101834053585e-06,
+      "loss": 0.0304,
+      "step": 5435
+    },
+    {
+      "epoch": 2.414390406395736,
+      "grad_norm": 0.4715313044352068,
+      "learning_rate": 4.068197187731526e-06,
+      "loss": 0.0369,
+      "step": 5436
+    },
+    {
+      "epoch": 2.4148345547412835,
+      "grad_norm": 0.27768669661494494,
+      "learning_rate": 4.066292681475257e-06,
+      "loss": 0.0174,
+      "step": 5437
+    },
+    {
+      "epoch": 2.4152787030868312,
+      "grad_norm": 0.6691350505746541,
+      "learning_rate": 4.064388315571059e-06,
+      "loss": 0.0318,
+      "step": 5438
+    },
+    {
+      "epoch": 2.4157228514323785,
+      "grad_norm": 0.4180829624879483,
+      "learning_rate": 4.062484090305191e-06,
+      "loss": 0.0416,
+      "step": 5439
+    },
+    {
+      "epoch": 2.416166999777926,
+      "grad_norm": 0.3034927493418548,
+      "learning_rate": 4.060580005963888e-06,
+      "loss": 0.0249,
+      "step": 5440
+    },
+    {
+      "epoch": 2.416611148123473,
+      "grad_norm": 0.42896525131188157,
+      "learning_rate": 4.05867606283337e-06,
+      "loss": 0.0345,
+      "step": 5441
+    },
+    {
+      "epoch": 2.417055296469021,
+      "grad_norm": 0.356634005220255,
+      "learning_rate": 4.0567722611998285e-06,
+      "loss": 0.0244,
+      "step": 5442
+    },
+    {
+      "epoch": 2.417499444814568,
+      "grad_norm": 0.49362452123065936,
+      "learning_rate": 4.054868601349441e-06,
+      "loss": 0.0349,
+      "step": 5443
+    },
+    {
+      "epoch": 2.4179435931601154,
+      "grad_norm": 0.3476206524561192,
+      "learning_rate": 4.052965083568356e-06,
+      "loss": 0.0262,
+      "step": 5444
+    },
+    {
+      "epoch": 2.4183877415056627,
+      "grad_norm": 0.4046428177539252,
+      "learning_rate": 4.051061708142705e-06,
+      "loss": 0.0308,
+      "step": 5445
+    },
+    {
+      "epoch": 2.4188318898512104,
+      "grad_norm": 0.6886851057569472,
+      "learning_rate": 4.0491584753586e-06,
+      "loss": 0.0489,
+      "step": 5446
+    },
+    {
+      "epoch": 2.4192760381967577,
+      "grad_norm": 0.4422724381498167,
+      "learning_rate": 4.047255385502129e-06,
+      "loss": 0.0338,
+      "step": 5447
+    },
+    {
+      "epoch": 2.419720186542305,
+      "grad_norm": 0.35168629623661846,
+      "learning_rate": 4.045352438859359e-06,
+      "loss": 0.0202,
+      "step": 5448
+    },
+    {
+      "epoch": 2.4201643348878523,
+      "grad_norm": 0.449273667282489,
+      "learning_rate": 4.043449635716332e-06,
+      "loss": 0.0396,
+      "step": 5449
+    },
+    {
+      "epoch": 2.4206084832334,
+      "grad_norm": 0.5131783172439474,
+      "learning_rate": 4.0415469763590745e-06,
+      "loss": 0.0416,
+      "step": 5450
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.46807114802560923,
+      "learning_rate": 4.0396444610735865e-06,
+      "loss": 0.0439,
+      "step": 5451
+    },
+    {
+      "epoch": 2.4214967799244946,
+      "grad_norm": 0.6426317590879459,
+      "learning_rate": 4.037742090145851e-06,
+      "loss": 0.05,
+      "step": 5452
+    },
+    {
+      "epoch": 2.4219409282700424,
+      "grad_norm": 0.5241231307427713,
+      "learning_rate": 4.0358398638618245e-06,
+      "loss": 0.0279,
+      "step": 5453
+    },
+    {
+      "epoch": 2.4223850766155897,
+      "grad_norm": 0.3603709645668666,
+      "learning_rate": 4.033937782507445e-06,
+      "loss": 0.0323,
+      "step": 5454
+    },
+    {
+      "epoch": 2.422829224961137,
+      "grad_norm": 0.4940232401715231,
+      "learning_rate": 4.032035846368627e-06,
+      "loss": 0.0316,
+      "step": 5455
+    },
+    {
+      "epoch": 2.4232733733066842,
+      "grad_norm": 0.4547687513915308,
+      "learning_rate": 4.030134055731266e-06,
+      "loss": 0.0398,
+      "step": 5456
+    },
+    {
+      "epoch": 2.423717521652232,
+      "grad_norm": 0.42013634054715737,
+      "learning_rate": 4.028232410881228e-06,
+      "loss": 0.0218,
+      "step": 5457
+    },
+    {
+      "epoch": 2.4241616699977793,
+      "grad_norm": 0.44875489093858906,
+      "learning_rate": 4.026330912104369e-06,
+      "loss": 0.0447,
+      "step": 5458
+    },
+    {
+      "epoch": 2.4246058183433266,
+      "grad_norm": 0.827939430580337,
+      "learning_rate": 4.024429559686513e-06,
+      "loss": 0.0678,
+      "step": 5459
+    },
+    {
+      "epoch": 2.4250499666888743,
+      "grad_norm": 0.7485263459812835,
+      "learning_rate": 4.022528353913466e-06,
+      "loss": 0.0442,
+      "step": 5460
+    },
+    {
+      "epoch": 2.4254941150344216,
+      "grad_norm": 0.3550612167644366,
+      "learning_rate": 4.020627295071012e-06,
+      "loss": 0.0334,
+      "step": 5461
+    },
+    {
+      "epoch": 2.425938263379969,
+      "grad_norm": 0.5079952410434588,
+      "learning_rate": 4.018726383444911e-06,
+      "loss": 0.0424,
+      "step": 5462
+    },
+    {
+      "epoch": 2.426382411725516,
+      "grad_norm": 0.3588389906173018,
+      "learning_rate": 4.016825619320904e-06,
+      "loss": 0.0256,
+      "step": 5463
+    },
+    {
+      "epoch": 2.426826560071064,
+      "grad_norm": 0.37557127358260195,
+      "learning_rate": 4.014925002984708e-06,
+      "loss": 0.027,
+      "step": 5464
+    },
+    {
+      "epoch": 2.427270708416611,
+      "grad_norm": 0.3878661608771635,
+      "learning_rate": 4.013024534722018e-06,
+      "loss": 0.0331,
+      "step": 5465
+    },
+    {
+      "epoch": 2.4277148567621585,
+      "grad_norm": 0.5433675019773777,
+      "learning_rate": 4.011124214818506e-06,
+      "loss": 0.032,
+      "step": 5466
+    },
+    {
+      "epoch": 2.428159005107706,
+      "grad_norm": 0.3899751784915578,
+      "learning_rate": 4.0092240435598225e-06,
+      "loss": 0.0356,
+      "step": 5467
+    },
+    {
+      "epoch": 2.4286031534532535,
+      "grad_norm": 0.4327878275713323,
+      "learning_rate": 4.007324021231594e-06,
+      "loss": 0.0319,
+      "step": 5468
+    },
+    {
+      "epoch": 2.429047301798801,
+      "grad_norm": 0.49698866083093113,
+      "learning_rate": 4.00542414811943e-06,
+      "loss": 0.0305,
+      "step": 5469
+    },
+    {
+      "epoch": 2.429491450144348,
+      "grad_norm": 0.4026428626742401,
+      "learning_rate": 4.00352442450891e-06,
+      "loss": 0.0348,
+      "step": 5470
+    },
+    {
+      "epoch": 2.4299355984898954,
+      "grad_norm": 0.48636219393391805,
+      "learning_rate": 4.001624850685598e-06,
+      "loss": 0.0346,
+      "step": 5471
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.4112453328110524,
+      "learning_rate": 3.999725426935029e-06,
+      "loss": 0.0428,
+      "step": 5472
+    },
+    {
+      "epoch": 2.4308238951809904,
+      "grad_norm": 0.5337831095831886,
+      "learning_rate": 3.99782615354272e-06,
+      "loss": 0.0442,
+      "step": 5473
+    },
+    {
+      "epoch": 2.4312680435265377,
+      "grad_norm": 0.37305930397141557,
+      "learning_rate": 3.995927030794163e-06,
+      "loss": 0.0357,
+      "step": 5474
+    },
+    {
+      "epoch": 2.4317121918720854,
+      "grad_norm": 0.4326214227306239,
+      "learning_rate": 3.994028058974832e-06,
+      "loss": 0.0294,
+      "step": 5475
+    },
+    {
+      "epoch": 2.4321563402176327,
+      "grad_norm": 0.35769209966498045,
+      "learning_rate": 3.992129238370171e-06,
+      "loss": 0.033,
+      "step": 5476
+    },
+    {
+      "epoch": 2.43260048856318,
+      "grad_norm": 0.5540164347858187,
+      "learning_rate": 3.9902305692656056e-06,
+      "loss": 0.0396,
+      "step": 5477
+    },
+    {
+      "epoch": 2.4330446369087273,
+      "grad_norm": 0.6878308326999009,
+      "learning_rate": 3.98833205194654e-06,
+      "loss": 0.0291,
+      "step": 5478
+    },
+    {
+      "epoch": 2.433488785254275,
+      "grad_norm": 0.6656069204526499,
+      "learning_rate": 3.98643368669835e-06,
+      "loss": 0.0377,
+      "step": 5479
+    },
+    {
+      "epoch": 2.4339329335998223,
+      "grad_norm": 0.36440064361824326,
+      "learning_rate": 3.984535473806395e-06,
+      "loss": 0.031,
+      "step": 5480
+    },
+    {
+      "epoch": 2.4343770819453696,
+      "grad_norm": 0.46753874608394186,
+      "learning_rate": 3.98263741355601e-06,
+      "loss": 0.0296,
+      "step": 5481
+    },
+    {
+      "epoch": 2.4348212302909173,
+      "grad_norm": 0.41167122039004567,
+      "learning_rate": 3.980739506232503e-06,
+      "loss": 0.0308,
+      "step": 5482
+    },
+    {
+      "epoch": 2.4352653786364646,
+      "grad_norm": 0.4469078274911405,
+      "learning_rate": 3.978841752121161e-06,
+      "loss": 0.0355,
+      "step": 5483
+    },
+    {
+      "epoch": 2.435709526982012,
+      "grad_norm": 0.39826295605317574,
+      "learning_rate": 3.976944151507251e-06,
+      "loss": 0.0288,
+      "step": 5484
+    },
+    {
+      "epoch": 2.436153675327559,
+      "grad_norm": 0.4535357297321569,
+      "learning_rate": 3.975046704676014e-06,
+      "loss": 0.0286,
+      "step": 5485
+    },
+    {
+      "epoch": 2.436597823673107,
+      "grad_norm": 0.4396407655662091,
+      "learning_rate": 3.973149411912668e-06,
+      "loss": 0.0305,
+      "step": 5486
+    },
+    {
+      "epoch": 2.4370419720186542,
+      "grad_norm": 0.3507364407629326,
+      "learning_rate": 3.971252273502407e-06,
+      "loss": 0.0248,
+      "step": 5487
+    },
+    {
+      "epoch": 2.4374861203642015,
+      "grad_norm": 0.3450849459589068,
+      "learning_rate": 3.969355289730407e-06,
+      "loss": 0.0206,
+      "step": 5488
+    },
+    {
+      "epoch": 2.4379302687097493,
+      "grad_norm": 0.503140049702594,
+      "learning_rate": 3.967458460881815e-06,
+      "loss": 0.04,
+      "step": 5489
+    },
+    {
+      "epoch": 2.4383744170552966,
+      "grad_norm": 0.5640472250485181,
+      "learning_rate": 3.965561787241754e-06,
+      "loss": 0.0404,
+      "step": 5490
+    },
+    {
+      "epoch": 2.438818565400844,
+      "grad_norm": 0.6056306520006238,
+      "learning_rate": 3.963665269095328e-06,
+      "loss": 0.0404,
+      "step": 5491
+    },
+    {
+      "epoch": 2.439262713746391,
+      "grad_norm": 0.4258100611472551,
+      "learning_rate": 3.961768906727618e-06,
+      "loss": 0.028,
+      "step": 5492
+    },
+    {
+      "epoch": 2.439706862091939,
+      "grad_norm": 0.38264655994704755,
+      "learning_rate": 3.959872700423678e-06,
+      "loss": 0.0256,
+      "step": 5493
+    },
+    {
+      "epoch": 2.440151010437486,
+      "grad_norm": 0.389861248826636,
+      "learning_rate": 3.957976650468539e-06,
+      "loss": 0.0333,
+      "step": 5494
+    },
+    {
+      "epoch": 2.4405951587830335,
+      "grad_norm": 0.6710261752550767,
+      "learning_rate": 3.956080757147211e-06,
+      "loss": 0.0288,
+      "step": 5495
+    },
+    {
+      "epoch": 2.441039307128581,
+      "grad_norm": 0.533539192209703,
+      "learning_rate": 3.9541850207446754e-06,
+      "loss": 0.029,
+      "step": 5496
+    },
+    {
+      "epoch": 2.4414834554741285,
+      "grad_norm": 0.7167183091153162,
+      "learning_rate": 3.9522894415459e-06,
+      "loss": 0.0473,
+      "step": 5497
+    },
+    {
+      "epoch": 2.4419276038196758,
+      "grad_norm": 0.322074964494659,
+      "learning_rate": 3.950394019835817e-06,
+      "loss": 0.0258,
+      "step": 5498
+    },
+    {
+      "epoch": 2.442371752165223,
+      "grad_norm": 0.3900552673739537,
+      "learning_rate": 3.948498755899344e-06,
+      "loss": 0.0287,
+      "step": 5499
+    },
+    {
+      "epoch": 2.4428159005107704,
+      "grad_norm": 0.4466770793345454,
+      "learning_rate": 3.94660365002137e-06,
+      "loss": 0.0292,
+      "step": 5500
+    },
+    {
+      "epoch": 2.443260048856318,
+      "grad_norm": 0.5418862551373604,
+      "learning_rate": 3.94470870248676e-06,
+      "loss": 0.0434,
+      "step": 5501
+    },
+    {
+      "epoch": 2.4437041972018654,
+      "grad_norm": 0.5170924051250978,
+      "learning_rate": 3.942813913580358e-06,
+      "loss": 0.044,
+      "step": 5502
+    },
+    {
+      "epoch": 2.4441483455474127,
+      "grad_norm": 0.5438284418911795,
+      "learning_rate": 3.940919283586985e-06,
+      "loss": 0.0448,
+      "step": 5503
+    },
+    {
+      "epoch": 2.4445924938929604,
+      "grad_norm": 0.5198158479237611,
+      "learning_rate": 3.9390248127914325e-06,
+      "loss": 0.0528,
+      "step": 5504
+    },
+    {
+      "epoch": 2.4450366422385077,
+      "grad_norm": 0.520564892893801,
+      "learning_rate": 3.937130501478475e-06,
+      "loss": 0.0488,
+      "step": 5505
+    },
+    {
+      "epoch": 2.445480790584055,
+      "grad_norm": 0.46307199806507215,
+      "learning_rate": 3.935236349932858e-06,
+      "loss": 0.0334,
+      "step": 5506
+    },
+    {
+      "epoch": 2.4459249389296023,
+      "grad_norm": 0.5703039507859231,
+      "learning_rate": 3.933342358439304e-06,
+      "loss": 0.0388,
+      "step": 5507
+    },
+    {
+      "epoch": 2.44636908727515,
+      "grad_norm": 0.480094036827002,
+      "learning_rate": 3.931448527282512e-06,
+      "loss": 0.0383,
+      "step": 5508
+    },
+    {
+      "epoch": 2.4468132356206973,
+      "grad_norm": 0.5416338368617786,
+      "learning_rate": 3.9295548567471595e-06,
+      "loss": 0.044,
+      "step": 5509
+    },
+    {
+      "epoch": 2.4472573839662446,
+      "grad_norm": 0.3789299998421564,
+      "learning_rate": 3.927661347117896e-06,
+      "loss": 0.0308,
+      "step": 5510
+    },
+    {
+      "epoch": 2.4477015323117923,
+      "grad_norm": 0.42767639025063464,
+      "learning_rate": 3.925767998679347e-06,
+      "loss": 0.032,
+      "step": 5511
+    },
+    {
+      "epoch": 2.4481456806573396,
+      "grad_norm": 0.3723173022056738,
+      "learning_rate": 3.923874811716116e-06,
+      "loss": 0.0299,
+      "step": 5512
+    },
+    {
+      "epoch": 2.448589829002887,
+      "grad_norm": 0.647876325917259,
+      "learning_rate": 3.92198178651278e-06,
+      "loss": 0.0334,
+      "step": 5513
+    },
+    {
+      "epoch": 2.449033977348434,
+      "grad_norm": 0.3651110551254141,
+      "learning_rate": 3.920088923353895e-06,
+      "loss": 0.0404,
+      "step": 5514
+    },
+    {
+      "epoch": 2.449478125693982,
+      "grad_norm": 0.43323931222905593,
+      "learning_rate": 3.918196222523989e-06,
+      "loss": 0.0458,
+      "step": 5515
+    },
+    {
+      "epoch": 2.4499222740395292,
+      "grad_norm": 0.4066604237779864,
+      "learning_rate": 3.916303684307568e-06,
+      "loss": 0.0291,
+      "step": 5516
+    },
+    {
+      "epoch": 2.4503664223850765,
+      "grad_norm": 0.4166876666433253,
+      "learning_rate": 3.914411308989113e-06,
+      "loss": 0.027,
+      "step": 5517
+    },
+    {
+      "epoch": 2.4508105707306242,
+      "grad_norm": 0.45791540033658323,
+      "learning_rate": 3.9125190968530766e-06,
+      "loss": 0.0271,
+      "step": 5518
+    },
+    {
+      "epoch": 2.4512547190761715,
+      "grad_norm": 0.4364649579290464,
+      "learning_rate": 3.910627048183893e-06,
+      "loss": 0.0325,
+      "step": 5519
+    },
+    {
+      "epoch": 2.451698867421719,
+      "grad_norm": 0.34758907020261137,
+      "learning_rate": 3.908735163265971e-06,
+      "loss": 0.0225,
+      "step": 5520
+    },
+    {
+      "epoch": 2.452143015767266,
+      "grad_norm": 0.386095366133868,
+      "learning_rate": 3.906843442383691e-06,
+      "loss": 0.0253,
+      "step": 5521
+    },
+    {
+      "epoch": 2.452587164112814,
+      "grad_norm": 0.5285696022742041,
+      "learning_rate": 3.90495188582141e-06,
+      "loss": 0.0288,
+      "step": 5522
+    },
+    {
+      "epoch": 2.453031312458361,
+      "grad_norm": 0.5685349400918212,
+      "learning_rate": 3.903060493863463e-06,
+      "loss": 0.0384,
+      "step": 5523
+    },
+    {
+      "epoch": 2.4534754608039084,
+      "grad_norm": 0.33536529902813805,
+      "learning_rate": 3.901169266794158e-06,
+      "loss": 0.0256,
+      "step": 5524
+    },
+    {
+      "epoch": 2.4539196091494557,
+      "grad_norm": 0.3367209757616414,
+      "learning_rate": 3.899278204897777e-06,
+      "loss": 0.0308,
+      "step": 5525
+    },
+    {
+      "epoch": 2.4543637574950035,
+      "grad_norm": 0.35301257937407804,
+      "learning_rate": 3.89738730845858e-06,
+      "loss": 0.0326,
+      "step": 5526
+    },
+    {
+      "epoch": 2.4548079058405508,
+      "grad_norm": 0.44674618292196533,
+      "learning_rate": 3.895496577760802e-06,
+      "loss": 0.0378,
+      "step": 5527
+    },
+    {
+      "epoch": 2.455252054186098,
+      "grad_norm": 0.33090638899042374,
+      "learning_rate": 3.893606013088649e-06,
+      "loss": 0.0287,
+      "step": 5528
+    },
+    {
+      "epoch": 2.4556962025316453,
+      "grad_norm": 0.6629654415706786,
+      "learning_rate": 3.8917156147263075e-06,
+      "loss": 0.0361,
+      "step": 5529
+    },
+    {
+      "epoch": 2.456140350877193,
+      "grad_norm": 0.4137805309428058,
+      "learning_rate": 3.889825382957935e-06,
+      "loss": 0.0381,
+      "step": 5530
+    },
+    {
+      "epoch": 2.4565844992227404,
+      "grad_norm": 0.48510449046292115,
+      "learning_rate": 3.887935318067665e-06,
+      "loss": 0.0369,
+      "step": 5531
+    },
+    {
+      "epoch": 2.4570286475682876,
+      "grad_norm": 0.4301148483997256,
+      "learning_rate": 3.886045420339608e-06,
+      "loss": 0.0288,
+      "step": 5532
+    },
+    {
+      "epoch": 2.4574727959138354,
+      "grad_norm": 0.42246466717722975,
+      "learning_rate": 3.884155690057849e-06,
+      "loss": 0.0301,
+      "step": 5533
+    },
+    {
+      "epoch": 2.4579169442593827,
+      "grad_norm": 0.4345894945232888,
+      "learning_rate": 3.882266127506444e-06,
+      "loss": 0.0327,
+      "step": 5534
+    },
+    {
+      "epoch": 2.45836109260493,
+      "grad_norm": 0.47014262773033944,
+      "learning_rate": 3.880376732969427e-06,
+      "loss": 0.0406,
+      "step": 5535
+    },
+    {
+      "epoch": 2.4588052409504773,
+      "grad_norm": 0.5529245797017762,
+      "learning_rate": 3.8784875067308035e-06,
+      "loss": 0.0428,
+      "step": 5536
+    },
+    {
+      "epoch": 2.459249389296025,
+      "grad_norm": 0.5853637583441337,
+      "learning_rate": 3.876598449074561e-06,
+      "loss": 0.0484,
+      "step": 5537
+    },
+    {
+      "epoch": 2.4596935376415723,
+      "grad_norm": 0.4308021015437979,
+      "learning_rate": 3.874709560284655e-06,
+      "loss": 0.0331,
+      "step": 5538
+    },
+    {
+      "epoch": 2.4601376859871196,
+      "grad_norm": 0.41105828222307955,
+      "learning_rate": 3.872820840645017e-06,
+      "loss": 0.0291,
+      "step": 5539
+    },
+    {
+      "epoch": 2.4605818343326673,
+      "grad_norm": 0.4669061473028007,
+      "learning_rate": 3.8709322904395556e-06,
+      "loss": 0.0275,
+      "step": 5540
+    },
+    {
+      "epoch": 2.4610259826782146,
+      "grad_norm": 0.4419324556460313,
+      "learning_rate": 3.869043909952149e-06,
+      "loss": 0.046,
+      "step": 5541
+    },
+    {
+      "epoch": 2.461470131023762,
+      "grad_norm": 0.4298170656814399,
+      "learning_rate": 3.867155699466653e-06,
+      "loss": 0.0376,
+      "step": 5542
+    },
+    {
+      "epoch": 2.461914279369309,
+      "grad_norm": 0.9256029510191909,
+      "learning_rate": 3.865267659266901e-06,
+      "loss": 0.0365,
+      "step": 5543
+    },
+    {
+      "epoch": 2.462358427714857,
+      "grad_norm": 0.4535797528392373,
+      "learning_rate": 3.863379789636696e-06,
+      "loss": 0.0302,
+      "step": 5544
+    },
+    {
+      "epoch": 2.462802576060404,
+      "grad_norm": 0.5482760330729667,
+      "learning_rate": 3.861492090859816e-06,
+      "loss": 0.0365,
+      "step": 5545
+    },
+    {
+      "epoch": 2.4632467244059515,
+      "grad_norm": 0.36157460865248275,
+      "learning_rate": 3.8596045632200126e-06,
+      "loss": 0.0266,
+      "step": 5546
+    },
+    {
+      "epoch": 2.4636908727514992,
+      "grad_norm": 0.4873814379100533,
+      "learning_rate": 3.857717207001017e-06,
+      "loss": 0.037,
+      "step": 5547
+    },
+    {
+      "epoch": 2.4641350210970465,
+      "grad_norm": 0.7370873255745318,
+      "learning_rate": 3.855830022486528e-06,
+      "loss": 0.0415,
+      "step": 5548
+    },
+    {
+      "epoch": 2.464579169442594,
+      "grad_norm": 0.493249372834125,
+      "learning_rate": 3.853943009960225e-06,
+      "loss": 0.0386,
+      "step": 5549
+    },
+    {
+      "epoch": 2.465023317788141,
+      "grad_norm": 0.6919203726918821,
+      "learning_rate": 3.852056169705753e-06,
+      "loss": 0.0389,
+      "step": 5550
+    },
+    {
+      "epoch": 2.465467466133689,
+      "grad_norm": 0.4116990565345076,
+      "learning_rate": 3.850169502006741e-06,
+      "loss": 0.0302,
+      "step": 5551
+    },
+    {
+      "epoch": 2.465911614479236,
+      "grad_norm": 0.34358727354238183,
+      "learning_rate": 3.848283007146784e-06,
+      "loss": 0.0325,
+      "step": 5552
+    },
+    {
+      "epoch": 2.4663557628247834,
+      "grad_norm": 0.35614401203480484,
+      "learning_rate": 3.846396685409455e-06,
+      "loss": 0.0322,
+      "step": 5553
+    },
+    {
+      "epoch": 2.4667999111703307,
+      "grad_norm": 0.25982925770375787,
+      "learning_rate": 3.8445105370782995e-06,
+      "loss": 0.0262,
+      "step": 5554
+    },
+    {
+      "epoch": 2.4672440595158784,
+      "grad_norm": 0.3510865325198374,
+      "learning_rate": 3.842624562436841e-06,
+      "loss": 0.0386,
+      "step": 5555
+    },
+    {
+      "epoch": 2.4676882078614257,
+      "grad_norm": 0.4049390820424808,
+      "learning_rate": 3.8407387617685696e-06,
+      "loss": 0.0247,
+      "step": 5556
+    },
+    {
+      "epoch": 2.468132356206973,
+      "grad_norm": 0.42859540412907965,
+      "learning_rate": 3.838853135356956e-06,
+      "loss": 0.0368,
+      "step": 5557
+    },
+    {
+      "epoch": 2.4685765045525203,
+      "grad_norm": 0.35853923916250774,
+      "learning_rate": 3.836967683485441e-06,
+      "loss": 0.0329,
+      "step": 5558
+    },
+    {
+      "epoch": 2.469020652898068,
+      "grad_norm": 0.4278309028616722,
+      "learning_rate": 3.835082406437437e-06,
+      "loss": 0.0362,
+      "step": 5559
+    },
+    {
+      "epoch": 2.4694648012436153,
+      "grad_norm": 1.0923622589898494,
+      "learning_rate": 3.833197304496336e-06,
+      "loss": 0.044,
+      "step": 5560
+    },
+    {
+      "epoch": 2.4699089495891626,
+      "grad_norm": 0.37654012560879,
+      "learning_rate": 3.8313123779455035e-06,
+      "loss": 0.0262,
+      "step": 5561
+    },
+    {
+      "epoch": 2.4703530979347104,
+      "grad_norm": 0.41898160595248646,
+      "learning_rate": 3.829427627068272e-06,
+      "loss": 0.0344,
+      "step": 5562
+    },
+    {
+      "epoch": 2.4707972462802577,
+      "grad_norm": 0.3844261706628325,
+      "learning_rate": 3.827543052147952e-06,
+      "loss": 0.0273,
+      "step": 5563
+    },
+    {
+      "epoch": 2.471241394625805,
+      "grad_norm": 0.4489319056107684,
+      "learning_rate": 3.8256586534678285e-06,
+      "loss": 0.0318,
+      "step": 5564
+    },
+    {
+      "epoch": 2.4716855429713522,
+      "grad_norm": 0.6833288857085646,
+      "learning_rate": 3.8237744313111565e-06,
+      "loss": 0.0404,
+      "step": 5565
+    },
+    {
+      "epoch": 2.4721296913169,
+      "grad_norm": 0.430664114118146,
+      "learning_rate": 3.82189038596117e-06,
+      "loss": 0.0352,
+      "step": 5566
+    },
+    {
+      "epoch": 2.4725738396624473,
+      "grad_norm": 0.4030254224585019,
+      "learning_rate": 3.820006517701069e-06,
+      "loss": 0.0321,
+      "step": 5567
+    },
+    {
+      "epoch": 2.4730179880079945,
+      "grad_norm": 0.5685844819493571,
+      "learning_rate": 3.8181228268140354e-06,
+      "loss": 0.0507,
+      "step": 5568
+    },
+    {
+      "epoch": 2.4734621363535423,
+      "grad_norm": 0.44264389289592554,
+      "learning_rate": 3.816239313583217e-06,
+      "loss": 0.0376,
+      "step": 5569
+    },
+    {
+      "epoch": 2.4739062846990896,
+      "grad_norm": 0.3377904339220037,
+      "learning_rate": 3.814355978291736e-06,
+      "loss": 0.0327,
+      "step": 5570
+    },
+    {
+      "epoch": 2.474350433044637,
+      "grad_norm": 0.630683589245786,
+      "learning_rate": 3.8124728212226938e-06,
+      "loss": 0.0471,
+      "step": 5571
+    },
+    {
+      "epoch": 2.474794581390184,
+      "grad_norm": 0.42602963126310434,
+      "learning_rate": 3.810589842659159e-06,
+      "loss": 0.0297,
+      "step": 5572
+    },
+    {
+      "epoch": 2.475238729735732,
+      "grad_norm": 0.4633136103256966,
+      "learning_rate": 3.808707042884176e-06,
+      "loss": 0.028,
+      "step": 5573
+    },
+    {
+      "epoch": 2.475682878081279,
+      "grad_norm": 0.5782431299481687,
+      "learning_rate": 3.8068244221807606e-06,
+      "loss": 0.0383,
+      "step": 5574
+    },
+    {
+      "epoch": 2.4761270264268265,
+      "grad_norm": 0.5372901069709559,
+      "learning_rate": 3.8049419808319033e-06,
+      "loss": 0.036,
+      "step": 5575
+    },
+    {
+      "epoch": 2.476571174772374,
+      "grad_norm": 0.2611922701250809,
+      "learning_rate": 3.8030597191205643e-06,
+      "loss": 0.0217,
+      "step": 5576
+    },
+    {
+      "epoch": 2.4770153231179215,
+      "grad_norm": 0.36585043972090653,
+      "learning_rate": 3.8011776373296837e-06,
+      "loss": 0.0286,
+      "step": 5577
+    },
+    {
+      "epoch": 2.477459471463469,
+      "grad_norm": 0.5157568238144473,
+      "learning_rate": 3.79929573574217e-06,
+      "loss": 0.0343,
+      "step": 5578
+    },
+    {
+      "epoch": 2.477903619809016,
+      "grad_norm": 0.4610943022253674,
+      "learning_rate": 3.797414014640903e-06,
+      "loss": 0.0318,
+      "step": 5579
+    },
+    {
+      "epoch": 2.478347768154564,
+      "grad_norm": 0.47373180317876196,
+      "learning_rate": 3.795532474308737e-06,
+      "loss": 0.0427,
+      "step": 5580
+    },
+    {
+      "epoch": 2.478791916500111,
+      "grad_norm": 0.4145736062776239,
+      "learning_rate": 3.7936511150285014e-06,
+      "loss": 0.038,
+      "step": 5581
+    },
+    {
+      "epoch": 2.4792360648456584,
+      "grad_norm": 0.5254698644896055,
+      "learning_rate": 3.7917699370829935e-06,
+      "loss": 0.0349,
+      "step": 5582
+    },
+    {
+      "epoch": 2.4796802131912057,
+      "grad_norm": 0.4171576347283061,
+      "learning_rate": 3.789888940754991e-06,
+      "loss": 0.0303,
+      "step": 5583
+    },
+    {
+      "epoch": 2.4801243615367534,
+      "grad_norm": 0.39892945172204647,
+      "learning_rate": 3.788008126327235e-06,
+      "loss": 0.041,
+      "step": 5584
+    },
+    {
+      "epoch": 2.4805685098823007,
+      "grad_norm": 0.4156599218368941,
+      "learning_rate": 3.7861274940824473e-06,
+      "loss": 0.0308,
+      "step": 5585
+    },
+    {
+      "epoch": 2.481012658227848,
+      "grad_norm": 0.3949693643427857,
+      "learning_rate": 3.784247044303317e-06,
+      "loss": 0.0311,
+      "step": 5586
+    },
+    {
+      "epoch": 2.4814568065733953,
+      "grad_norm": 0.3619897883910703,
+      "learning_rate": 3.782366777272506e-06,
+      "loss": 0.0304,
+      "step": 5587
+    },
+    {
+      "epoch": 2.481900954918943,
+      "grad_norm": 0.3797535629139442,
+      "learning_rate": 3.7804866932726535e-06,
+      "loss": 0.0265,
+      "step": 5588
+    },
+    {
+      "epoch": 2.4823451032644903,
+      "grad_norm": 0.4514704220606347,
+      "learning_rate": 3.778606792586368e-06,
+      "loss": 0.0242,
+      "step": 5589
+    },
+    {
+      "epoch": 2.4827892516100376,
+      "grad_norm": 0.5885462347335189,
+      "learning_rate": 3.7767270754962294e-06,
+      "loss": 0.0462,
+      "step": 5590
+    },
+    {
+      "epoch": 2.4832333999555853,
+      "grad_norm": 0.46737727413834057,
+      "learning_rate": 3.7748475422847896e-06,
+      "loss": 0.0476,
+      "step": 5591
+    },
+    {
+      "epoch": 2.4836775483011326,
+      "grad_norm": 0.34347582817428995,
+      "learning_rate": 3.7729681932345776e-06,
+      "loss": 0.0309,
+      "step": 5592
+    },
+    {
+      "epoch": 2.48412169664668,
+      "grad_norm": 0.5416760213809743,
+      "learning_rate": 3.771089028628087e-06,
+      "loss": 0.0301,
+      "step": 5593
+    },
+    {
+      "epoch": 2.484565844992227,
+      "grad_norm": 0.5279294508428536,
+      "learning_rate": 3.7692100487477936e-06,
+      "loss": 0.0338,
+      "step": 5594
+    },
+    {
+      "epoch": 2.485009993337775,
+      "grad_norm": 0.40771734537444654,
+      "learning_rate": 3.7673312538761362e-06,
+      "loss": 0.0244,
+      "step": 5595
+    },
+    {
+      "epoch": 2.4854541416833222,
+      "grad_norm": 0.3225283085841615,
+      "learning_rate": 3.765452644295532e-06,
+      "loss": 0.0252,
+      "step": 5596
+    },
+    {
+      "epoch": 2.4858982900288695,
+      "grad_norm": 0.4343781070994849,
+      "learning_rate": 3.7635742202883664e-06,
+      "loss": 0.0301,
+      "step": 5597
+    },
+    {
+      "epoch": 2.4863424383744173,
+      "grad_norm": 0.48736918945639,
+      "learning_rate": 3.761695982136997e-06,
+      "loss": 0.0283,
+      "step": 5598
+    },
+    {
+      "epoch": 2.4867865867199646,
+      "grad_norm": 0.4628216568506521,
+      "learning_rate": 3.759817930123756e-06,
+      "loss": 0.0325,
+      "step": 5599
+    },
+    {
+      "epoch": 2.487230735065512,
+      "grad_norm": 0.4979234883742358,
+      "learning_rate": 3.75794006453095e-06,
+      "loss": 0.046,
+      "step": 5600
+    },
+    {
+      "epoch": 2.487674883411059,
+      "grad_norm": 0.5295486655610855,
+      "learning_rate": 3.7560623856408496e-06,
+      "loss": 0.0421,
+      "step": 5601
+    },
+    {
+      "epoch": 2.488119031756607,
+      "grad_norm": 0.3550831132877616,
+      "learning_rate": 3.7541848937357037e-06,
+      "loss": 0.0282,
+      "step": 5602
+    },
+    {
+      "epoch": 2.488563180102154,
+      "grad_norm": 0.4820776586497428,
+      "learning_rate": 3.7523075890977323e-06,
+      "loss": 0.0327,
+      "step": 5603
+    },
+    {
+      "epoch": 2.4890073284477015,
+      "grad_norm": 0.36069709664433147,
+      "learning_rate": 3.7504304720091227e-06,
+      "loss": 0.0282,
+      "step": 5604
+    },
+    {
+      "epoch": 2.489451476793249,
+      "grad_norm": 0.3378590317401163,
+      "learning_rate": 3.7485535427520393e-06,
+      "loss": 0.0327,
+      "step": 5605
+    },
+    {
+      "epoch": 2.4898956251387965,
+      "grad_norm": 0.5101551267438247,
+      "learning_rate": 3.7466768016086187e-06,
+      "loss": 0.0426,
+      "step": 5606
+    },
+    {
+      "epoch": 2.4903397734843438,
+      "grad_norm": 0.4554277064950327,
+      "learning_rate": 3.7448002488609647e-06,
+      "loss": 0.0363,
+      "step": 5607
+    },
+    {
+      "epoch": 2.490783921829891,
+      "grad_norm": 0.5190026980108126,
+      "learning_rate": 3.7429238847911555e-06,
+      "loss": 0.0303,
+      "step": 5608
+    },
+    {
+      "epoch": 2.4912280701754383,
+      "grad_norm": 0.4937954972214749,
+      "learning_rate": 3.7410477096812402e-06,
+      "loss": 0.0489,
+      "step": 5609
+    },
+    {
+      "epoch": 2.491672218520986,
+      "grad_norm": 0.36659365050297416,
+      "learning_rate": 3.7391717238132386e-06,
+      "loss": 0.0266,
+      "step": 5610
+    },
+    {
+      "epoch": 2.4921163668665334,
+      "grad_norm": 0.391793755436005,
+      "learning_rate": 3.737295927469146e-06,
+      "loss": 0.0288,
+      "step": 5611
+    },
+    {
+      "epoch": 2.4925605152120807,
+      "grad_norm": 0.386017966668196,
+      "learning_rate": 3.7354203209309246e-06,
+      "loss": 0.0347,
+      "step": 5612
+    },
+    {
+      "epoch": 2.4930046635576284,
+      "grad_norm": 0.30330961970277515,
+      "learning_rate": 3.733544904480512e-06,
+      "loss": 0.0248,
+      "step": 5613
+    },
+    {
+      "epoch": 2.4934488119031757,
+      "grad_norm": 0.4407907387683592,
+      "learning_rate": 3.7316696783998124e-06,
+      "loss": 0.0347,
+      "step": 5614
+    },
+    {
+      "epoch": 2.493892960248723,
+      "grad_norm": 0.37646444767265064,
+      "learning_rate": 3.7297946429707045e-06,
+      "loss": 0.0328,
+      "step": 5615
+    },
+    {
+      "epoch": 2.4943371085942703,
+      "grad_norm": 0.4575522855234119,
+      "learning_rate": 3.727919798475038e-06,
+      "loss": 0.0352,
+      "step": 5616
+    },
+    {
+      "epoch": 2.494781256939818,
+      "grad_norm": 0.35695162304512706,
+      "learning_rate": 3.7260451451946365e-06,
+      "loss": 0.0316,
+      "step": 5617
+    },
+    {
+      "epoch": 2.4952254052853653,
+      "grad_norm": 0.43360958825611745,
+      "learning_rate": 3.724170683411291e-06,
+      "loss": 0.0332,
+      "step": 5618
+    },
+    {
+      "epoch": 2.4956695536309126,
+      "grad_norm": 0.4262372003773419,
+      "learning_rate": 3.722296413406763e-06,
+      "loss": 0.0255,
+      "step": 5619
+    },
+    {
+      "epoch": 2.4961137019764603,
+      "grad_norm": 0.43363995486396417,
+      "learning_rate": 3.7204223354627894e-06,
+      "loss": 0.031,
+      "step": 5620
+    },
+    {
+      "epoch": 2.4965578503220076,
+      "grad_norm": 0.47249288635203707,
+      "learning_rate": 3.718548449861074e-06,
+      "loss": 0.0331,
+      "step": 5621
+    },
+    {
+      "epoch": 2.497001998667555,
+      "grad_norm": 0.4319106936736054,
+      "learning_rate": 3.716674756883295e-06,
+      "loss": 0.0392,
+      "step": 5622
+    },
+    {
+      "epoch": 2.497446147013102,
+      "grad_norm": 0.5119693456166591,
+      "learning_rate": 3.714801256811099e-06,
+      "loss": 0.0409,
+      "step": 5623
+    },
+    {
+      "epoch": 2.49789029535865,
+      "grad_norm": 0.40570657622630646,
+      "learning_rate": 3.712927949926108e-06,
+      "loss": 0.0266,
+      "step": 5624
+    },
+    {
+      "epoch": 2.498334443704197,
+      "grad_norm": 0.40821652834924144,
+      "learning_rate": 3.7110548365099075e-06,
+      "loss": 0.0402,
+      "step": 5625
+    },
+    {
+      "epoch": 2.4987785920497445,
+      "grad_norm": 0.3940755701591239,
+      "learning_rate": 3.7091819168440624e-06,
+      "loss": 0.0315,
+      "step": 5626
+    },
+    {
+      "epoch": 2.4992227403952922,
+      "grad_norm": 0.495948418850682,
+      "learning_rate": 3.7073091912101002e-06,
+      "loss": 0.0326,
+      "step": 5627
+    },
+    {
+      "epoch": 2.4996668887408395,
+      "grad_norm": 0.4435967840620372,
+      "learning_rate": 3.705436659889527e-06,
+      "loss": 0.0368,
+      "step": 5628
+    },
+    {
+      "epoch": 2.500111037086387,
+      "grad_norm": 0.32392962929044805,
+      "learning_rate": 3.7035643231638135e-06,
+      "loss": 0.029,
+      "step": 5629
+    },
+    {
+      "epoch": 2.500555185431934,
+      "grad_norm": 0.6179012687291638,
+      "learning_rate": 3.7016921813144063e-06,
+      "loss": 0.0382,
+      "step": 5630
+    },
+    {
+      "epoch": 2.5009993337774814,
+      "grad_norm": 0.45744910004367445,
+      "learning_rate": 3.6998202346227183e-06,
+      "loss": 0.0418,
+      "step": 5631
+    },
+    {
+      "epoch": 2.501443482123029,
+      "grad_norm": 0.5527549347471298,
+      "learning_rate": 3.697948483370135e-06,
+      "loss": 0.0257,
+      "step": 5632
+    },
+    {
+      "epoch": 2.5018876304685764,
+      "grad_norm": 0.5434768947853562,
+      "learning_rate": 3.696076927838011e-06,
+      "loss": 0.0389,
+      "step": 5633
+    },
+    {
+      "epoch": 2.502331778814124,
+      "grad_norm": 0.3410765055084389,
+      "learning_rate": 3.6942055683076767e-06,
+      "loss": 0.0299,
+      "step": 5634
+    },
+    {
+      "epoch": 2.5027759271596715,
+      "grad_norm": 0.7036711081682665,
+      "learning_rate": 3.692334405060427e-06,
+      "loss": 0.0405,
+      "step": 5635
+    },
+    {
+      "epoch": 2.5032200755052187,
+      "grad_norm": 0.3864467247613326,
+      "learning_rate": 3.6904634383775283e-06,
+      "loss": 0.0293,
+      "step": 5636
+    },
+    {
+      "epoch": 2.503664223850766,
+      "grad_norm": 0.5152666553526902,
+      "learning_rate": 3.6885926685402213e-06,
+      "loss": 0.042,
+      "step": 5637
+    },
+    {
+      "epoch": 2.5041083721963133,
+      "grad_norm": 0.3913873319828441,
+      "learning_rate": 3.6867220958297132e-06,
+      "loss": 0.0341,
+      "step": 5638
+    },
+    {
+      "epoch": 2.504552520541861,
+      "grad_norm": 0.3712133677859356,
+      "learning_rate": 3.6848517205271805e-06,
+      "loss": 0.0257,
+      "step": 5639
+    },
+    {
+      "epoch": 2.5049966688874084,
+      "grad_norm": 0.42947535443731244,
+      "learning_rate": 3.682981542913776e-06,
+      "loss": 0.0369,
+      "step": 5640
+    },
+    {
+      "epoch": 2.5054408172329556,
+      "grad_norm": 0.5274708204343582,
+      "learning_rate": 3.6811115632706185e-06,
+      "loss": 0.0324,
+      "step": 5641
+    },
+    {
+      "epoch": 2.5058849655785034,
+      "grad_norm": 0.4291181112678621,
+      "learning_rate": 3.6792417818787972e-06,
+      "loss": 0.0293,
+      "step": 5642
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.42231381186795797,
+      "learning_rate": 3.677372199019371e-06,
+      "loss": 0.0366,
+      "step": 5643
+    },
+    {
+      "epoch": 2.506773262269598,
+      "grad_norm": 0.4789995350453658,
+      "learning_rate": 3.6755028149733697e-06,
+      "loss": 0.0327,
+      "step": 5644
+    },
+    {
+      "epoch": 2.5072174106151452,
+      "grad_norm": 0.40689270267790856,
+      "learning_rate": 3.6736336300217964e-06,
+      "loss": 0.0237,
+      "step": 5645
+    },
+    {
+      "epoch": 2.507661558960693,
+      "grad_norm": 0.5913689239490344,
+      "learning_rate": 3.6717646444456196e-06,
+      "loss": 0.0366,
+      "step": 5646
+    },
+    {
+      "epoch": 2.5081057073062403,
+      "grad_norm": 0.3914862469261757,
+      "learning_rate": 3.669895858525778e-06,
+      "loss": 0.025,
+      "step": 5647
+    },
+    {
+      "epoch": 2.5085498556517876,
+      "grad_norm": 0.42201187516388383,
+      "learning_rate": 3.6680272725431854e-06,
+      "loss": 0.0336,
+      "step": 5648
+    },
+    {
+      "epoch": 2.5089940039973353,
+      "grad_norm": 0.42314579939123975,
+      "learning_rate": 3.6661588867787183e-06,
+      "loss": 0.0368,
+      "step": 5649
+    },
+    {
+      "epoch": 2.5094381523428826,
+      "grad_norm": 0.46190362474830615,
+      "learning_rate": 3.664290701513229e-06,
+      "loss": 0.0366,
+      "step": 5650
+    },
+    {
+      "epoch": 2.50988230068843,
+      "grad_norm": 0.5151193711763113,
+      "learning_rate": 3.662422717027536e-06,
+      "loss": 0.0343,
+      "step": 5651
+    },
+    {
+      "epoch": 2.510326449033977,
+      "grad_norm": 0.4683557027001918,
+      "learning_rate": 3.6605549336024327e-06,
+      "loss": 0.0502,
+      "step": 5652
+    },
+    {
+      "epoch": 2.510770597379525,
+      "grad_norm": 0.3966005086612889,
+      "learning_rate": 3.658687351518674e-06,
+      "loss": 0.0307,
+      "step": 5653
+    },
+    {
+      "epoch": 2.511214745725072,
+      "grad_norm": 0.4901749101545214,
+      "learning_rate": 3.656819971056992e-06,
+      "loss": 0.0514,
+      "step": 5654
+    },
+    {
+      "epoch": 2.5116588940706195,
+      "grad_norm": 0.44088737564599295,
+      "learning_rate": 3.654952792498086e-06,
+      "loss": 0.035,
+      "step": 5655
+    },
+    {
+      "epoch": 2.512103042416167,
+      "grad_norm": 0.637540920302593,
+      "learning_rate": 3.653085816122621e-06,
+      "loss": 0.0433,
+      "step": 5656
+    },
+    {
+      "epoch": 2.5125471907617145,
+      "grad_norm": 0.3989139686742009,
+      "learning_rate": 3.651219042211239e-06,
+      "loss": 0.0266,
+      "step": 5657
+    },
+    {
+      "epoch": 2.512991339107262,
+      "grad_norm": 0.49156765417598675,
+      "learning_rate": 3.649352471044548e-06,
+      "loss": 0.0232,
+      "step": 5658
+    },
+    {
+      "epoch": 2.513435487452809,
+      "grad_norm": 0.3796366437427131,
+      "learning_rate": 3.647486102903124e-06,
+      "loss": 0.0254,
+      "step": 5659
+    },
+    {
+      "epoch": 2.5138796357983564,
+      "grad_norm": 0.4428861844260197,
+      "learning_rate": 3.6456199380675128e-06,
+      "loss": 0.0336,
+      "step": 5660
+    },
+    {
+      "epoch": 2.514323784143904,
+      "grad_norm": 0.4660014764332606,
+      "learning_rate": 3.6437539768182305e-06,
+      "loss": 0.0379,
+      "step": 5661
+    },
+    {
+      "epoch": 2.5147679324894514,
+      "grad_norm": 0.48208527317355315,
+      "learning_rate": 3.6418882194357662e-06,
+      "loss": 0.0309,
+      "step": 5662
+    },
+    {
+      "epoch": 2.515212080834999,
+      "grad_norm": 0.46984872075734047,
+      "learning_rate": 3.6400226662005733e-06,
+      "loss": 0.0281,
+      "step": 5663
+    },
+    {
+      "epoch": 2.5156562291805464,
+      "grad_norm": 0.3674696428042144,
+      "learning_rate": 3.638157317393074e-06,
+      "loss": 0.0254,
+      "step": 5664
+    },
+    {
+      "epoch": 2.5161003775260937,
+      "grad_norm": 0.3790734362034619,
+      "learning_rate": 3.636292173293665e-06,
+      "loss": 0.0322,
+      "step": 5665
+    },
+    {
+      "epoch": 2.516544525871641,
+      "grad_norm": 0.36816453960949835,
+      "learning_rate": 3.634427234182708e-06,
+      "loss": 0.0339,
+      "step": 5666
+    },
+    {
+      "epoch": 2.5169886742171883,
+      "grad_norm": 0.310536315755628,
+      "learning_rate": 3.632562500340532e-06,
+      "loss": 0.0204,
+      "step": 5667
+    },
+    {
+      "epoch": 2.517432822562736,
+      "grad_norm": 0.5208883625261909,
+      "learning_rate": 3.6306979720474424e-06,
+      "loss": 0.0385,
+      "step": 5668
+    },
+    {
+      "epoch": 2.5178769709082833,
+      "grad_norm": 0.4448683196893281,
+      "learning_rate": 3.6288336495837085e-06,
+      "loss": 0.0357,
+      "step": 5669
+    },
+    {
+      "epoch": 2.5183211192538306,
+      "grad_norm": 0.44092160751797316,
+      "learning_rate": 3.6269695332295697e-06,
+      "loss": 0.0344,
+      "step": 5670
+    },
+    {
+      "epoch": 2.5187652675993784,
+      "grad_norm": 0.45635940839778255,
+      "learning_rate": 3.6251056232652327e-06,
+      "loss": 0.0305,
+      "step": 5671
+    },
+    {
+      "epoch": 2.5192094159449256,
+      "grad_norm": 0.5533263657116287,
+      "learning_rate": 3.6232419199708764e-06,
+      "loss": 0.0284,
+      "step": 5672
+    },
+    {
+      "epoch": 2.519653564290473,
+      "grad_norm": 0.37835546841851936,
+      "learning_rate": 3.6213784236266447e-06,
+      "loss": 0.0326,
+      "step": 5673
+    },
+    {
+      "epoch": 2.5200977126360202,
+      "grad_norm": 0.4665241718562087,
+      "learning_rate": 3.6195151345126556e-06,
+      "loss": 0.036,
+      "step": 5674
+    },
+    {
+      "epoch": 2.520541860981568,
+      "grad_norm": 0.3945669545233313,
+      "learning_rate": 3.6176520529089932e-06,
+      "loss": 0.0339,
+      "step": 5675
+    },
+    {
+      "epoch": 2.5209860093271153,
+      "grad_norm": 0.37055474601271443,
+      "learning_rate": 3.6157891790957096e-06,
+      "loss": 0.0339,
+      "step": 5676
+    },
+    {
+      "epoch": 2.5214301576726625,
+      "grad_norm": 0.492060267079228,
+      "learning_rate": 3.6139265133528246e-06,
+      "loss": 0.0316,
+      "step": 5677
+    },
+    {
+      "epoch": 2.5218743060182103,
+      "grad_norm": 0.5701337130019367,
+      "learning_rate": 3.612064055960331e-06,
+      "loss": 0.0464,
+      "step": 5678
+    },
+    {
+      "epoch": 2.5223184543637576,
+      "grad_norm": 0.39863450477642115,
+      "learning_rate": 3.6102018071981846e-06,
+      "loss": 0.0334,
+      "step": 5679
+    },
+    {
+      "epoch": 2.522762602709305,
+      "grad_norm": 0.515203363747466,
+      "learning_rate": 3.6083397673463172e-06,
+      "loss": 0.0486,
+      "step": 5680
+    },
+    {
+      "epoch": 2.523206751054852,
+      "grad_norm": 0.3893233450260606,
+      "learning_rate": 3.606477936684622e-06,
+      "loss": 0.0359,
+      "step": 5681
+    },
+    {
+      "epoch": 2.5236508994004,
+      "grad_norm": 0.45713551996484814,
+      "learning_rate": 3.6046163154929657e-06,
+      "loss": 0.0374,
+      "step": 5682
+    },
+    {
+      "epoch": 2.524095047745947,
+      "grad_norm": 0.5641320893715582,
+      "learning_rate": 3.6027549040511806e-06,
+      "loss": 0.058,
+      "step": 5683
+    },
+    {
+      "epoch": 2.5245391960914945,
+      "grad_norm": 0.3265494892390024,
+      "learning_rate": 3.600893702639067e-06,
+      "loss": 0.023,
+      "step": 5684
+    },
+    {
+      "epoch": 2.524983344437042,
+      "grad_norm": 0.41936797741330234,
+      "learning_rate": 3.5990327115363967e-06,
+      "loss": 0.0256,
+      "step": 5685
+    },
+    {
+      "epoch": 2.5254274927825895,
+      "grad_norm": 0.46801276167069494,
+      "learning_rate": 3.5971719310229093e-06,
+      "loss": 0.0254,
+      "step": 5686
+    },
+    {
+      "epoch": 2.525871641128137,
+      "grad_norm": 0.49189838872219027,
+      "learning_rate": 3.595311361378311e-06,
+      "loss": 0.0364,
+      "step": 5687
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.5586944397026717,
+      "learning_rate": 3.593451002882275e-06,
+      "loss": 0.0335,
+      "step": 5688
+    },
+    {
+      "epoch": 2.5267599378192314,
+      "grad_norm": 0.4422926900473203,
+      "learning_rate": 3.5915908558144476e-06,
+      "loss": 0.0298,
+      "step": 5689
+    },
+    {
+      "epoch": 2.527204086164779,
+      "grad_norm": 0.3434949432587989,
+      "learning_rate": 3.5897309204544375e-06,
+      "loss": 0.0336,
+      "step": 5690
+    },
+    {
+      "epoch": 2.5276482345103264,
+      "grad_norm": 0.3621563955807786,
+      "learning_rate": 3.587871197081828e-06,
+      "loss": 0.0254,
+      "step": 5691
+    },
+    {
+      "epoch": 2.528092382855874,
+      "grad_norm": 0.4132257794043078,
+      "learning_rate": 3.586011685976164e-06,
+      "loss": 0.0332,
+      "step": 5692
+    },
+    {
+      "epoch": 2.5285365312014214,
+      "grad_norm": 0.5011026660419655,
+      "learning_rate": 3.5841523874169648e-06,
+      "loss": 0.0345,
+      "step": 5693
+    },
+    {
+      "epoch": 2.5289806795469687,
+      "grad_norm": 0.45248007625680975,
+      "learning_rate": 3.582293301683713e-06,
+      "loss": 0.0293,
+      "step": 5694
+    },
+    {
+      "epoch": 2.529424827892516,
+      "grad_norm": 0.7885398440569542,
+      "learning_rate": 3.580434429055859e-06,
+      "loss": 0.032,
+      "step": 5695
+    },
+    {
+      "epoch": 2.5298689762380633,
+      "grad_norm": 0.8455065643807562,
+      "learning_rate": 3.578575769812824e-06,
+      "loss": 0.0566,
+      "step": 5696
+    },
+    {
+      "epoch": 2.530313124583611,
+      "grad_norm": 0.3581406871410735,
+      "learning_rate": 3.576717324233998e-06,
+      "loss": 0.0283,
+      "step": 5697
+    },
+    {
+      "epoch": 2.5307572729291583,
+      "grad_norm": 0.37566727714871373,
+      "learning_rate": 3.5748590925987347e-06,
+      "loss": 0.0274,
+      "step": 5698
+    },
+    {
+      "epoch": 2.5312014212747056,
+      "grad_norm": 0.40595777932831095,
+      "learning_rate": 3.5730010751863605e-06,
+      "loss": 0.029,
+      "step": 5699
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.47532996991241144,
+      "learning_rate": 3.571143272276164e-06,
+      "loss": 0.0532,
+      "step": 5700
+    },
+    {
+      "epoch": 2.5320897179658006,
+      "grad_norm": 0.44327227035484773,
+      "learning_rate": 3.5692856841474045e-06,
+      "loss": 0.0355,
+      "step": 5701
+    },
+    {
+      "epoch": 2.532533866311348,
+      "grad_norm": 0.34290979724264187,
+      "learning_rate": 3.5674283110793105e-06,
+      "loss": 0.0219,
+      "step": 5702
+    },
+    {
+      "epoch": 2.532978014656895,
+      "grad_norm": 0.6517538589913022,
+      "learning_rate": 3.5655711533510783e-06,
+      "loss": 0.0354,
+      "step": 5703
+    },
+    {
+      "epoch": 2.533422163002443,
+      "grad_norm": 0.40477955297885887,
+      "learning_rate": 3.5637142112418684e-06,
+      "loss": 0.0292,
+      "step": 5704
+    },
+    {
+      "epoch": 2.5338663113479902,
+      "grad_norm": 0.47637177118935925,
+      "learning_rate": 3.5618574850308095e-06,
+      "loss": 0.0256,
+      "step": 5705
+    },
+    {
+      "epoch": 2.5343104596935375,
+      "grad_norm": 0.4050526950368258,
+      "learning_rate": 3.560000974997001e-06,
+      "loss": 0.0279,
+      "step": 5706
+    },
+    {
+      "epoch": 2.5347546080390853,
+      "grad_norm": 0.4510491005204953,
+      "learning_rate": 3.5581446814195054e-06,
+      "loss": 0.039,
+      "step": 5707
+    },
+    {
+      "epoch": 2.5351987563846325,
+      "grad_norm": 0.43059426602406603,
+      "learning_rate": 3.556288604577359e-06,
+      "loss": 0.0344,
+      "step": 5708
+    },
+    {
+      "epoch": 2.53564290473018,
+      "grad_norm": 0.38353556571368813,
+      "learning_rate": 3.5544327447495598e-06,
+      "loss": 0.033,
+      "step": 5709
+    },
+    {
+      "epoch": 2.536087053075727,
+      "grad_norm": 0.3857238286700737,
+      "learning_rate": 3.5525771022150746e-06,
+      "loss": 0.0282,
+      "step": 5710
+    },
+    {
+      "epoch": 2.536531201421275,
+      "grad_norm": 0.48091169542814494,
+      "learning_rate": 3.5507216772528392e-06,
+      "loss": 0.0354,
+      "step": 5711
+    },
+    {
+      "epoch": 2.536975349766822,
+      "grad_norm": 1.0810667528006974,
+      "learning_rate": 3.548866470141753e-06,
+      "loss": 0.0411,
+      "step": 5712
+    },
+    {
+      "epoch": 2.5374194981123694,
+      "grad_norm": 0.37822644450804105,
+      "learning_rate": 3.547011481160686e-06,
+      "loss": 0.0287,
+      "step": 5713
+    },
+    {
+      "epoch": 2.537863646457917,
+      "grad_norm": 0.44547431389156866,
+      "learning_rate": 3.5451567105884777e-06,
+      "loss": 0.0446,
+      "step": 5714
+    },
+    {
+      "epoch": 2.5383077948034645,
+      "grad_norm": 0.33403093136486245,
+      "learning_rate": 3.543302158703929e-06,
+      "loss": 0.0304,
+      "step": 5715
+    },
+    {
+      "epoch": 2.5387519431490118,
+      "grad_norm": 0.3853720547985273,
+      "learning_rate": 3.5414478257858097e-06,
+      "loss": 0.0316,
+      "step": 5716
+    },
+    {
+      "epoch": 2.539196091494559,
+      "grad_norm": 0.4078086186517431,
+      "learning_rate": 3.53959371211286e-06,
+      "loss": 0.0359,
+      "step": 5717
+    },
+    {
+      "epoch": 2.5396402398401063,
+      "grad_norm": 0.3855600056546902,
+      "learning_rate": 3.5377398179637807e-06,
+      "loss": 0.0201,
+      "step": 5718
+    },
+    {
+      "epoch": 2.540084388185654,
+      "grad_norm": 0.3712283572301996,
+      "learning_rate": 3.5358861436172487e-06,
+      "loss": 0.0327,
+      "step": 5719
+    },
+    {
+      "epoch": 2.5405285365312014,
+      "grad_norm": 0.4121457515255565,
+      "learning_rate": 3.5340326893518993e-06,
+      "loss": 0.0337,
+      "step": 5720
+    },
+    {
+      "epoch": 2.540972684876749,
+      "grad_norm": 0.4589358591248568,
+      "learning_rate": 3.5321794554463397e-06,
+      "loss": 0.0282,
+      "step": 5721
+    },
+    {
+      "epoch": 2.5414168332222964,
+      "grad_norm": 0.41283412784679185,
+      "learning_rate": 3.530326442179142e-06,
+      "loss": 0.0433,
+      "step": 5722
+    },
+    {
+      "epoch": 2.5418609815678437,
+      "grad_norm": 0.3832687248919424,
+      "learning_rate": 3.5284736498288452e-06,
+      "loss": 0.0275,
+      "step": 5723
+    },
+    {
+      "epoch": 2.542305129913391,
+      "grad_norm": 0.4017937297419123,
+      "learning_rate": 3.526621078673954e-06,
+      "loss": 0.0301,
+      "step": 5724
+    },
+    {
+      "epoch": 2.5427492782589383,
+      "grad_norm": 0.3799788724726438,
+      "learning_rate": 3.5247687289929443e-06,
+      "loss": 0.0352,
+      "step": 5725
+    },
+    {
+      "epoch": 2.543193426604486,
+      "grad_norm": 0.3381062594883136,
+      "learning_rate": 3.5229166010642544e-06,
+      "loss": 0.0237,
+      "step": 5726
+    },
+    {
+      "epoch": 2.5436375749500333,
+      "grad_norm": 0.9370246012561935,
+      "learning_rate": 3.521064695166292e-06,
+      "loss": 0.0417,
+      "step": 5727
+    },
+    {
+      "epoch": 2.5440817232955806,
+      "grad_norm": 0.5170089647125808,
+      "learning_rate": 3.5192130115774283e-06,
+      "loss": 0.0379,
+      "step": 5728
+    },
+    {
+      "epoch": 2.5445258716411283,
+      "grad_norm": 0.5478100844231081,
+      "learning_rate": 3.5173615505760015e-06,
+      "loss": 0.0327,
+      "step": 5729
+    },
+    {
+      "epoch": 2.5449700199866756,
+      "grad_norm": 0.3263956967060353,
+      "learning_rate": 3.5155103124403184e-06,
+      "loss": 0.0235,
+      "step": 5730
+    },
+    {
+      "epoch": 2.545414168332223,
+      "grad_norm": 0.4151765481340016,
+      "learning_rate": 3.513659297448655e-06,
+      "loss": 0.0315,
+      "step": 5731
+    },
+    {
+      "epoch": 2.54585831667777,
+      "grad_norm": 0.4249849360391111,
+      "learning_rate": 3.511808505879247e-06,
+      "loss": 0.037,
+      "step": 5732
+    },
+    {
+      "epoch": 2.546302465023318,
+      "grad_norm": 0.430816357070103,
+      "learning_rate": 3.5099579380103e-06,
+      "loss": 0.0313,
+      "step": 5733
+    },
+    {
+      "epoch": 2.546746613368865,
+      "grad_norm": 0.36068218836576676,
+      "learning_rate": 3.508107594119987e-06,
+      "loss": 0.0248,
+      "step": 5734
+    },
+    {
+      "epoch": 2.5471907617144125,
+      "grad_norm": 0.5603158518242672,
+      "learning_rate": 3.506257474486444e-06,
+      "loss": 0.0407,
+      "step": 5735
+    },
+    {
+      "epoch": 2.5476349100599602,
+      "grad_norm": 0.457105631896688,
+      "learning_rate": 3.5044075793877784e-06,
+      "loss": 0.0285,
+      "step": 5736
+    },
+    {
+      "epoch": 2.5480790584055075,
+      "grad_norm": 0.4513751440943639,
+      "learning_rate": 3.5025579091020584e-06,
+      "loss": 0.0295,
+      "step": 5737
+    },
+    {
+      "epoch": 2.548523206751055,
+      "grad_norm": 0.4467104682114999,
+      "learning_rate": 3.500708463907323e-06,
+      "loss": 0.0297,
+      "step": 5738
+    },
+    {
+      "epoch": 2.548967355096602,
+      "grad_norm": 0.3775580565507929,
+      "learning_rate": 3.498859244081573e-06,
+      "loss": 0.0302,
+      "step": 5739
+    },
+    {
+      "epoch": 2.54941150344215,
+      "grad_norm": 0.4736436006509606,
+      "learning_rate": 3.4970102499027787e-06,
+      "loss": 0.0324,
+      "step": 5740
+    },
+    {
+      "epoch": 2.549855651787697,
+      "grad_norm": 0.38970978243484516,
+      "learning_rate": 3.4951614816488733e-06,
+      "loss": 0.0239,
+      "step": 5741
+    },
+    {
+      "epoch": 2.5502998001332444,
+      "grad_norm": 0.3858729170075839,
+      "learning_rate": 3.4933129395977627e-06,
+      "loss": 0.0282,
+      "step": 5742
+    },
+    {
+      "epoch": 2.550743948478792,
+      "grad_norm": 0.46868923999899664,
+      "learning_rate": 3.491464624027311e-06,
+      "loss": 0.0309,
+      "step": 5743
+    },
+    {
+      "epoch": 2.5511880968243394,
+      "grad_norm": 0.42676283372780777,
+      "learning_rate": 3.489616535215351e-06,
+      "loss": 0.0288,
+      "step": 5744
+    },
+    {
+      "epoch": 2.5516322451698867,
+      "grad_norm": 0.47537306808660484,
+      "learning_rate": 3.487768673439684e-06,
+      "loss": 0.0409,
+      "step": 5745
+    },
+    {
+      "epoch": 2.552076393515434,
+      "grad_norm": 0.5509640851909691,
+      "learning_rate": 3.4859210389780717e-06,
+      "loss": 0.0537,
+      "step": 5746
+    },
+    {
+      "epoch": 2.5525205418609813,
+      "grad_norm": 0.4532699668327261,
+      "learning_rate": 3.484073632108248e-06,
+      "loss": 0.0374,
+      "step": 5747
+    },
+    {
+      "epoch": 2.552964690206529,
+      "grad_norm": 0.30162200474796635,
+      "learning_rate": 3.4822264531079074e-06,
+      "loss": 0.0346,
+      "step": 5748
+    },
+    {
+      "epoch": 2.5534088385520763,
+      "grad_norm": 0.41586878305661834,
+      "learning_rate": 3.4803795022547152e-06,
+      "loss": 0.0385,
+      "step": 5749
+    },
+    {
+      "epoch": 2.553852986897624,
+      "grad_norm": 0.5284306714224316,
+      "learning_rate": 3.478532779826297e-06,
+      "loss": 0.0329,
+      "step": 5750
+    },
+    {
+      "epoch": 2.5542971352431714,
+      "grad_norm": 0.3711615788486314,
+      "learning_rate": 3.476686286100247e-06,
+      "loss": 0.0277,
+      "step": 5751
+    },
+    {
+      "epoch": 2.5547412835887187,
+      "grad_norm": 0.4470649615235087,
+      "learning_rate": 3.4748400213541233e-06,
+      "loss": 0.0463,
+      "step": 5752
+    },
+    {
+      "epoch": 2.555185431934266,
+      "grad_norm": 0.40717240259190046,
+      "learning_rate": 3.4729939858654548e-06,
+      "loss": 0.0296,
+      "step": 5753
+    },
+    {
+      "epoch": 2.5556295802798132,
+      "grad_norm": 0.4846910790786083,
+      "learning_rate": 3.471148179911728e-06,
+      "loss": 0.0404,
+      "step": 5754
+    },
+    {
+      "epoch": 2.556073728625361,
+      "grad_norm": 0.37627232321927323,
+      "learning_rate": 3.4693026037704012e-06,
+      "loss": 0.0334,
+      "step": 5755
+    },
+    {
+      "epoch": 2.5565178769709083,
+      "grad_norm": 0.34629139309877494,
+      "learning_rate": 3.467457257718896e-06,
+      "loss": 0.0272,
+      "step": 5756
+    },
+    {
+      "epoch": 2.5569620253164556,
+      "grad_norm": 0.3421995312642631,
+      "learning_rate": 3.4656121420345968e-06,
+      "loss": 0.0317,
+      "step": 5757
+    },
+    {
+      "epoch": 2.5574061736620033,
+      "grad_norm": 0.44077802939094335,
+      "learning_rate": 3.463767256994856e-06,
+      "loss": 0.0549,
+      "step": 5758
+    },
+    {
+      "epoch": 2.5578503220075506,
+      "grad_norm": 0.3954224399718614,
+      "learning_rate": 3.461922602876995e-06,
+      "loss": 0.0334,
+      "step": 5759
+    },
+    {
+      "epoch": 2.558294470353098,
+      "grad_norm": 0.41169694619074404,
+      "learning_rate": 3.460078179958294e-06,
+      "loss": 0.0341,
+      "step": 5760
+    },
+    {
+      "epoch": 2.558738618698645,
+      "grad_norm": 0.4812985244340657,
+      "learning_rate": 3.458233988516e-06,
+      "loss": 0.0407,
+      "step": 5761
+    },
+    {
+      "epoch": 2.559182767044193,
+      "grad_norm": 0.3331989432004887,
+      "learning_rate": 3.4563900288273287e-06,
+      "loss": 0.0247,
+      "step": 5762
+    },
+    {
+      "epoch": 2.55962691538974,
+      "grad_norm": 0.3885261825147752,
+      "learning_rate": 3.454546301169458e-06,
+      "loss": 0.0294,
+      "step": 5763
+    },
+    {
+      "epoch": 2.5600710637352875,
+      "grad_norm": 0.43455546109798127,
+      "learning_rate": 3.4527028058195276e-06,
+      "loss": 0.0276,
+      "step": 5764
+    },
+    {
+      "epoch": 2.560515212080835,
+      "grad_norm": 0.41019774224604005,
+      "learning_rate": 3.4508595430546516e-06,
+      "loss": 0.0376,
+      "step": 5765
+    },
+    {
+      "epoch": 2.5609593604263825,
+      "grad_norm": 0.5340205931151287,
+      "learning_rate": 3.4490165131519027e-06,
+      "loss": 0.0363,
+      "step": 5766
+    },
+    {
+      "epoch": 2.56140350877193,
+      "grad_norm": 0.32981175279651753,
+      "learning_rate": 3.4471737163883178e-06,
+      "loss": 0.0341,
+      "step": 5767
+    },
+    {
+      "epoch": 2.561847657117477,
+      "grad_norm": 0.4191900996748185,
+      "learning_rate": 3.4453311530409008e-06,
+      "loss": 0.0336,
+      "step": 5768
+    },
+    {
+      "epoch": 2.5622918054630244,
+      "grad_norm": 0.4533154388512584,
+      "learning_rate": 3.4434888233866205e-06,
+      "loss": 0.0379,
+      "step": 5769
+    },
+    {
+      "epoch": 2.562735953808572,
+      "grad_norm": 0.4083356757917847,
+      "learning_rate": 3.4416467277024097e-06,
+      "loss": 0.0256,
+      "step": 5770
+    },
+    {
+      "epoch": 2.5631801021541194,
+      "grad_norm": 0.46425748132873407,
+      "learning_rate": 3.4398048662651693e-06,
+      "loss": 0.0455,
+      "step": 5771
+    },
+    {
+      "epoch": 2.563624250499667,
+      "grad_norm": 0.3043869275862561,
+      "learning_rate": 3.4379632393517593e-06,
+      "loss": 0.022,
+      "step": 5772
+    },
+    {
+      "epoch": 2.5640683988452144,
+      "grad_norm": 0.7475044005307006,
+      "learning_rate": 3.43612184723901e-06,
+      "loss": 0.0502,
+      "step": 5773
+    },
+    {
+      "epoch": 2.5645125471907617,
+      "grad_norm": 0.32220506593319287,
+      "learning_rate": 3.4342806902037118e-06,
+      "loss": 0.0271,
+      "step": 5774
+    },
+    {
+      "epoch": 2.564956695536309,
+      "grad_norm": 0.36616059681338226,
+      "learning_rate": 3.4324397685226217e-06,
+      "loss": 0.0327,
+      "step": 5775
+    },
+    {
+      "epoch": 2.5654008438818563,
+      "grad_norm": 0.4598617313466593,
+      "learning_rate": 3.4305990824724645e-06,
+      "loss": 0.0255,
+      "step": 5776
+    },
+    {
+      "epoch": 2.565844992227404,
+      "grad_norm": 0.4169024524406747,
+      "learning_rate": 3.428758632329925e-06,
+      "loss": 0.0261,
+      "step": 5777
+    },
+    {
+      "epoch": 2.5662891405729513,
+      "grad_norm": 0.46153415454886376,
+      "learning_rate": 3.426918418371652e-06,
+      "loss": 0.0313,
+      "step": 5778
+    },
+    {
+      "epoch": 2.5667332889184986,
+      "grad_norm": 0.5358677284520018,
+      "learning_rate": 3.4250784408742644e-06,
+      "loss": 0.045,
+      "step": 5779
+    },
+    {
+      "epoch": 2.5671774372640463,
+      "grad_norm": 0.4437829933438522,
+      "learning_rate": 3.4232387001143396e-06,
+      "loss": 0.0423,
+      "step": 5780
+    },
+    {
+      "epoch": 2.5676215856095936,
+      "grad_norm": 0.39105850673238723,
+      "learning_rate": 3.4213991963684212e-06,
+      "loss": 0.0353,
+      "step": 5781
+    },
+    {
+      "epoch": 2.568065733955141,
+      "grad_norm": 0.5049529429986211,
+      "learning_rate": 3.419559929913021e-06,
+      "loss": 0.0345,
+      "step": 5782
+    },
+    {
+      "epoch": 2.568509882300688,
+      "grad_norm": 0.4292519880189004,
+      "learning_rate": 3.4177209010246104e-06,
+      "loss": 0.0372,
+      "step": 5783
+    },
+    {
+      "epoch": 2.568954030646236,
+      "grad_norm": 0.373978887044069,
+      "learning_rate": 3.415882109979627e-06,
+      "loss": 0.0428,
+      "step": 5784
+    },
+    {
+      "epoch": 2.5693981789917832,
+      "grad_norm": 0.500647171885995,
+      "learning_rate": 3.4140435570544708e-06,
+      "loss": 0.036,
+      "step": 5785
+    },
+    {
+      "epoch": 2.5698423273373305,
+      "grad_norm": 0.5585745761796531,
+      "learning_rate": 3.4122052425255097e-06,
+      "loss": 0.0371,
+      "step": 5786
+    },
+    {
+      "epoch": 2.5702864756828783,
+      "grad_norm": 0.40623789381422887,
+      "learning_rate": 3.4103671666690706e-06,
+      "loss": 0.0346,
+      "step": 5787
+    },
+    {
+      "epoch": 2.5707306240284256,
+      "grad_norm": 0.5204708462096757,
+      "learning_rate": 3.4085293297614513e-06,
+      "loss": 0.0341,
+      "step": 5788
+    },
+    {
+      "epoch": 2.571174772373973,
+      "grad_norm": 0.38514191058110864,
+      "learning_rate": 3.406691732078907e-06,
+      "loss": 0.0247,
+      "step": 5789
+    },
+    {
+      "epoch": 2.57161892071952,
+      "grad_norm": 0.47415946915349233,
+      "learning_rate": 3.4048543738976624e-06,
+      "loss": 0.03,
+      "step": 5790
+    },
+    {
+      "epoch": 2.572063069065068,
+      "grad_norm": 0.35195512662300593,
+      "learning_rate": 3.4030172554939022e-06,
+      "loss": 0.0179,
+      "step": 5791
+    },
+    {
+      "epoch": 2.572507217410615,
+      "grad_norm": 0.455360916505429,
+      "learning_rate": 3.401180377143774e-06,
+      "loss": 0.0355,
+      "step": 5792
+    },
+    {
+      "epoch": 2.5729513657561625,
+      "grad_norm": 0.6594611426447926,
+      "learning_rate": 3.399343739123395e-06,
+      "loss": 0.04,
+      "step": 5793
+    },
+    {
+      "epoch": 2.57339551410171,
+      "grad_norm": 0.30764478468780626,
+      "learning_rate": 3.3975073417088445e-06,
+      "loss": 0.024,
+      "step": 5794
+    },
+    {
+      "epoch": 2.5738396624472575,
+      "grad_norm": 0.5139517791311652,
+      "learning_rate": 3.3956711851761603e-06,
+      "loss": 0.0386,
+      "step": 5795
+    },
+    {
+      "epoch": 2.5742838107928048,
+      "grad_norm": 0.44276110456272694,
+      "learning_rate": 3.393835269801351e-06,
+      "loss": 0.0276,
+      "step": 5796
+    },
+    {
+      "epoch": 2.574727959138352,
+      "grad_norm": 0.6084624820775043,
+      "learning_rate": 3.3919995958603845e-06,
+      "loss": 0.0349,
+      "step": 5797
+    },
+    {
+      "epoch": 2.5751721074838994,
+      "grad_norm": 0.6187475898080996,
+      "learning_rate": 3.3901641636291925e-06,
+      "loss": 0.0369,
+      "step": 5798
+    },
+    {
+      "epoch": 2.575616255829447,
+      "grad_norm": 0.4166403983258688,
+      "learning_rate": 3.388328973383673e-06,
+      "loss": 0.028,
+      "step": 5799
+    },
+    {
+      "epoch": 2.5760604041749944,
+      "grad_norm": 0.37895711615417893,
+      "learning_rate": 3.3864940253996885e-06,
+      "loss": 0.0309,
+      "step": 5800
+    },
+    {
+      "epoch": 2.576504552520542,
+      "grad_norm": 0.4165330638891502,
+      "learning_rate": 3.3846593199530598e-06,
+      "loss": 0.0324,
+      "step": 5801
+    },
+    {
+      "epoch": 2.5769487008660894,
+      "grad_norm": 0.3850803256916831,
+      "learning_rate": 3.3828248573195744e-06,
+      "loss": 0.0243,
+      "step": 5802
+    },
+    {
+      "epoch": 2.5773928492116367,
+      "grad_norm": 0.4535252370266945,
+      "learning_rate": 3.3809906377749853e-06,
+      "loss": 0.026,
+      "step": 5803
+    },
+    {
+      "epoch": 2.577836997557184,
+      "grad_norm": 0.4473597091096733,
+      "learning_rate": 3.3791566615950034e-06,
+      "loss": 0.0446,
+      "step": 5804
+    },
+    {
+      "epoch": 2.5782811459027313,
+      "grad_norm": 0.4461365886821962,
+      "learning_rate": 3.37732292905531e-06,
+      "loss": 0.0353,
+      "step": 5805
+    },
+    {
+      "epoch": 2.578725294248279,
+      "grad_norm": 0.4044232712398919,
+      "learning_rate": 3.375489440431544e-06,
+      "loss": 0.0356,
+      "step": 5806
+    },
+    {
+      "epoch": 2.5791694425938263,
+      "grad_norm": 0.4100077927626587,
+      "learning_rate": 3.373656195999312e-06,
+      "loss": 0.0288,
+      "step": 5807
+    },
+    {
+      "epoch": 2.5796135909393736,
+      "grad_norm": 0.49681714136189903,
+      "learning_rate": 3.3718231960341807e-06,
+      "loss": 0.0392,
+      "step": 5808
+    },
+    {
+      "epoch": 2.5800577392849213,
+      "grad_norm": 0.3508735282517198,
+      "learning_rate": 3.3699904408116778e-06,
+      "loss": 0.0266,
+      "step": 5809
+    },
+    {
+      "epoch": 2.5805018876304686,
+      "grad_norm": 0.4685055266479743,
+      "learning_rate": 3.368157930607303e-06,
+      "loss": 0.0322,
+      "step": 5810
+    },
+    {
+      "epoch": 2.580946035976016,
+      "grad_norm": 0.3392092135667048,
+      "learning_rate": 3.3663256656965115e-06,
+      "loss": 0.0266,
+      "step": 5811
+    },
+    {
+      "epoch": 2.581390184321563,
+      "grad_norm": 0.3673851977352648,
+      "learning_rate": 3.364493646354724e-06,
+      "loss": 0.0291,
+      "step": 5812
+    },
+    {
+      "epoch": 2.581834332667111,
+      "grad_norm": 0.4826180223086163,
+      "learning_rate": 3.3626618728573233e-06,
+      "loss": 0.0366,
+      "step": 5813
+    },
+    {
+      "epoch": 2.5822784810126582,
+      "grad_norm": 0.46379192115203366,
+      "learning_rate": 3.3608303454796578e-06,
+      "loss": 0.0359,
+      "step": 5814
+    },
+    {
+      "epoch": 2.5827226293582055,
+      "grad_norm": 0.33507634737295205,
+      "learning_rate": 3.3589990644970325e-06,
+      "loss": 0.0293,
+      "step": 5815
+    },
+    {
+      "epoch": 2.5831667777037532,
+      "grad_norm": 0.394019845804789,
+      "learning_rate": 3.3571680301847265e-06,
+      "loss": 0.0287,
+      "step": 5816
+    },
+    {
+      "epoch": 2.5836109260493005,
+      "grad_norm": 0.4267750495828387,
+      "learning_rate": 3.355337242817972e-06,
+      "loss": 0.0225,
+      "step": 5817
+    },
+    {
+      "epoch": 2.584055074394848,
+      "grad_norm": 0.40408538263164145,
+      "learning_rate": 3.3535067026719683e-06,
+      "loss": 0.0354,
+      "step": 5818
+    },
+    {
+      "epoch": 2.584499222740395,
+      "grad_norm": 0.3264537051893323,
+      "learning_rate": 3.3516764100218744e-06,
+      "loss": 0.0184,
+      "step": 5819
+    },
+    {
+      "epoch": 2.584943371085943,
+      "grad_norm": 0.5823548169064077,
+      "learning_rate": 3.3498463651428183e-06,
+      "loss": 0.0555,
+      "step": 5820
+    },
+    {
+      "epoch": 2.58538751943149,
+      "grad_norm": 0.6018406880891104,
+      "learning_rate": 3.348016568309882e-06,
+      "loss": 0.0323,
+      "step": 5821
+    },
+    {
+      "epoch": 2.5858316677770374,
+      "grad_norm": 0.4499473097806689,
+      "learning_rate": 3.3461870197981205e-06,
+      "loss": 0.0313,
+      "step": 5822
+    },
+    {
+      "epoch": 2.586275816122585,
+      "grad_norm": 0.4573988840299094,
+      "learning_rate": 3.3443577198825416e-06,
+      "loss": 0.0217,
+      "step": 5823
+    },
+    {
+      "epoch": 2.5867199644681325,
+      "grad_norm": 0.47710512867182214,
+      "learning_rate": 3.342528668838123e-06,
+      "loss": 0.0305,
+      "step": 5824
+    },
+    {
+      "epoch": 2.5871641128136798,
+      "grad_norm": 0.3333521539172945,
+      "learning_rate": 3.3406998669398015e-06,
+      "loss": 0.0276,
+      "step": 5825
+    },
+    {
+      "epoch": 2.587608261159227,
+      "grad_norm": 0.4557037415155258,
+      "learning_rate": 3.338871314462474e-06,
+      "loss": 0.0283,
+      "step": 5826
+    },
+    {
+      "epoch": 2.5880524095047743,
+      "grad_norm": 0.4444177366240355,
+      "learning_rate": 3.337043011681007e-06,
+      "loss": 0.0328,
+      "step": 5827
+    },
+    {
+      "epoch": 2.588496557850322,
+      "grad_norm": 0.3980372800704896,
+      "learning_rate": 3.335214958870225e-06,
+      "loss": 0.0312,
+      "step": 5828
+    },
+    {
+      "epoch": 2.5889407061958694,
+      "grad_norm": 0.38163158317470414,
+      "learning_rate": 3.333387156304914e-06,
+      "loss": 0.021,
+      "step": 5829
+    },
+    {
+      "epoch": 2.589384854541417,
+      "grad_norm": 0.3444340204979025,
+      "learning_rate": 3.3315596042598235e-06,
+      "loss": 0.0283,
+      "step": 5830
+    },
+    {
+      "epoch": 2.5898290028869644,
+      "grad_norm": 0.4610922606737187,
+      "learning_rate": 3.3297323030096672e-06,
+      "loss": 0.032,
+      "step": 5831
+    },
+    {
+      "epoch": 2.5902731512325117,
+      "grad_norm": 0.41870835356631375,
+      "learning_rate": 3.327905252829117e-06,
+      "loss": 0.0448,
+      "step": 5832
+    },
+    {
+      "epoch": 2.590717299578059,
+      "grad_norm": 0.4779584466171409,
+      "learning_rate": 3.326078453992813e-06,
+      "loss": 0.0242,
+      "step": 5833
+    },
+    {
+      "epoch": 2.5911614479236063,
+      "grad_norm": 0.5325628313786069,
+      "learning_rate": 3.324251906775351e-06,
+      "loss": 0.0395,
+      "step": 5834
+    },
+    {
+      "epoch": 2.591605596269154,
+      "grad_norm": 0.358759607760571,
+      "learning_rate": 3.3224256114512953e-06,
+      "loss": 0.0316,
+      "step": 5835
+    },
+    {
+      "epoch": 2.5920497446147013,
+      "grad_norm": 0.43363197580686524,
+      "learning_rate": 3.3205995682951666e-06,
+      "loss": 0.0334,
+      "step": 5836
+    },
+    {
+      "epoch": 2.5924938929602486,
+      "grad_norm": 0.4445842523123569,
+      "learning_rate": 3.31877377758145e-06,
+      "loss": 0.033,
+      "step": 5837
+    },
+    {
+      "epoch": 2.5929380413057963,
+      "grad_norm": 0.603779398622737,
+      "learning_rate": 3.316948239584592e-06,
+      "loss": 0.0449,
+      "step": 5838
+    },
+    {
+      "epoch": 2.5933821896513436,
+      "grad_norm": 0.4392799290696822,
+      "learning_rate": 3.3151229545790066e-06,
+      "loss": 0.0384,
+      "step": 5839
+    },
+    {
+      "epoch": 2.593826337996891,
+      "grad_norm": 0.38613834204137615,
+      "learning_rate": 3.3132979228390615e-06,
+      "loss": 0.0248,
+      "step": 5840
+    },
+    {
+      "epoch": 2.594270486342438,
+      "grad_norm": 0.31483135068991924,
+      "learning_rate": 3.3114731446390897e-06,
+      "loss": 0.0204,
+      "step": 5841
+    },
+    {
+      "epoch": 2.594714634687986,
+      "grad_norm": 0.42067924470792134,
+      "learning_rate": 3.3096486202533884e-06,
+      "loss": 0.0312,
+      "step": 5842
+    },
+    {
+      "epoch": 2.595158783033533,
+      "grad_norm": 0.4747651071435795,
+      "learning_rate": 3.3078243499562126e-06,
+      "loss": 0.0325,
+      "step": 5843
+    },
+    {
+      "epoch": 2.5956029313790805,
+      "grad_norm": 0.3567548194704181,
+      "learning_rate": 3.3060003340217822e-06,
+      "loss": 0.0241,
+      "step": 5844
+    },
+    {
+      "epoch": 2.5960470797246282,
+      "grad_norm": 0.3574932064483991,
+      "learning_rate": 3.3041765727242773e-06,
+      "loss": 0.0302,
+      "step": 5845
+    },
+    {
+      "epoch": 2.5964912280701755,
+      "grad_norm": 0.8739424998669374,
+      "learning_rate": 3.302353066337842e-06,
+      "loss": 0.0458,
+      "step": 5846
+    },
+    {
+      "epoch": 2.596935376415723,
+      "grad_norm": 0.4781713811806308,
+      "learning_rate": 3.300529815136577e-06,
+      "loss": 0.0328,
+      "step": 5847
+    },
+    {
+      "epoch": 2.59737952476127,
+      "grad_norm": 0.4686560316809108,
+      "learning_rate": 3.2987068193945515e-06,
+      "loss": 0.0392,
+      "step": 5848
+    },
+    {
+      "epoch": 2.597823673106818,
+      "grad_norm": 0.7682784843583998,
+      "learning_rate": 3.296884079385789e-06,
+      "loss": 0.0432,
+      "step": 5849
+    },
+    {
+      "epoch": 2.598267821452365,
+      "grad_norm": 0.43912852313892203,
+      "learning_rate": 3.2950615953842816e-06,
+      "loss": 0.0357,
+      "step": 5850
+    },
+    {
+      "epoch": 2.5987119697979124,
+      "grad_norm": 0.4363862214055466,
+      "learning_rate": 3.293239367663978e-06,
+      "loss": 0.0355,
+      "step": 5851
+    },
+    {
+      "epoch": 2.59915611814346,
+      "grad_norm": 0.4194478678628829,
+      "learning_rate": 3.2914173964987905e-06,
+      "loss": 0.0376,
+      "step": 5852
+    },
+    {
+      "epoch": 2.5996002664890074,
+      "grad_norm": 0.3939580447848396,
+      "learning_rate": 3.289595682162593e-06,
+      "loss": 0.0351,
+      "step": 5853
+    },
+    {
+      "epoch": 2.6000444148345547,
+      "grad_norm": 0.3563672060439901,
+      "learning_rate": 3.2877742249292174e-06,
+      "loss": 0.0264,
+      "step": 5854
+    },
+    {
+      "epoch": 2.600488563180102,
+      "grad_norm": 0.45772110095911156,
+      "learning_rate": 3.2859530250724604e-06,
+      "loss": 0.0286,
+      "step": 5855
+    },
+    {
+      "epoch": 2.6009327115256493,
+      "grad_norm": 0.3788943790281068,
+      "learning_rate": 3.284132082866083e-06,
+      "loss": 0.0231,
+      "step": 5856
+    },
+    {
+      "epoch": 2.601376859871197,
+      "grad_norm": 0.42056256980059764,
+      "learning_rate": 3.2823113985837996e-06,
+      "loss": 0.0301,
+      "step": 5857
+    },
+    {
+      "epoch": 2.6018210082167443,
+      "grad_norm": 0.4731371391351043,
+      "learning_rate": 3.2804909724992917e-06,
+      "loss": 0.0421,
+      "step": 5858
+    },
+    {
+      "epoch": 2.602265156562292,
+      "grad_norm": 0.4171211336356501,
+      "learning_rate": 3.2786708048862e-06,
+      "loss": 0.0267,
+      "step": 5859
+    },
+    {
+      "epoch": 2.6027093049078394,
+      "grad_norm": 0.33335158042812324,
+      "learning_rate": 3.276850896018128e-06,
+      "loss": 0.0287,
+      "step": 5860
+    },
+    {
+      "epoch": 2.6031534532533867,
+      "grad_norm": 0.6871125152697936,
+      "learning_rate": 3.2750312461686346e-06,
+      "loss": 0.0253,
+      "step": 5861
+    },
+    {
+      "epoch": 2.603597601598934,
+      "grad_norm": 0.38868974447379917,
+      "learning_rate": 3.273211855611248e-06,
+      "loss": 0.0384,
+      "step": 5862
+    },
+    {
+      "epoch": 2.6040417499444812,
+      "grad_norm": 0.41427271665417736,
+      "learning_rate": 3.271392724619454e-06,
+      "loss": 0.0283,
+      "step": 5863
+    },
+    {
+      "epoch": 2.604485898290029,
+      "grad_norm": 0.47635109894278754,
+      "learning_rate": 3.2695738534666964e-06,
+      "loss": 0.0307,
+      "step": 5864
+    },
+    {
+      "epoch": 2.6049300466355763,
+      "grad_norm": 0.43305461847503646,
+      "learning_rate": 3.2677552424263836e-06,
+      "loss": 0.0323,
+      "step": 5865
+    },
+    {
+      "epoch": 2.6053741949811235,
+      "grad_norm": 0.5925027302243593,
+      "learning_rate": 3.2659368917718813e-06,
+      "loss": 0.0349,
+      "step": 5866
+    },
+    {
+      "epoch": 2.6058183433266713,
+      "grad_norm": 0.5670868601287159,
+      "learning_rate": 3.264118801776524e-06,
+      "loss": 0.0272,
+      "step": 5867
+    },
+    {
+      "epoch": 2.6062624916722186,
+      "grad_norm": 0.4639294675991596,
+      "learning_rate": 3.262300972713598e-06,
+      "loss": 0.0281,
+      "step": 5868
+    },
+    {
+      "epoch": 2.606706640017766,
+      "grad_norm": 0.5113196767933541,
+      "learning_rate": 3.2604834048563527e-06,
+      "loss": 0.0258,
+      "step": 5869
+    },
+    {
+      "epoch": 2.607150788363313,
+      "grad_norm": 0.37055184296693755,
+      "learning_rate": 3.2586660984780017e-06,
+      "loss": 0.0292,
+      "step": 5870
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.5340070208292783,
+      "learning_rate": 3.256849053851716e-06,
+      "loss": 0.0362,
+      "step": 5871
+    },
+    {
+      "epoch": 2.608039085054408,
+      "grad_norm": 0.6461515771963295,
+      "learning_rate": 3.2550322712506265e-06,
+      "loss": 0.0378,
+      "step": 5872
+    },
+    {
+      "epoch": 2.6084832333999555,
+      "grad_norm": 0.43751867677053585,
+      "learning_rate": 3.2532157509478313e-06,
+      "loss": 0.0297,
+      "step": 5873
+    },
+    {
+      "epoch": 2.608927381745503,
+      "grad_norm": 0.4731032790298419,
+      "learning_rate": 3.2513994932163806e-06,
+      "loss": 0.0426,
+      "step": 5874
+    },
+    {
+      "epoch": 2.6093715300910505,
+      "grad_norm": 0.40443116847842614,
+      "learning_rate": 3.2495834983292894e-06,
+      "loss": 0.0315,
+      "step": 5875
+    },
+    {
+      "epoch": 2.609815678436598,
+      "grad_norm": 0.4867165911113075,
+      "learning_rate": 3.2477677665595333e-06,
+      "loss": 0.0319,
+      "step": 5876
+    },
+    {
+      "epoch": 2.610259826782145,
+      "grad_norm": 0.42932234052521123,
+      "learning_rate": 3.2459522981800473e-06,
+      "loss": 0.0465,
+      "step": 5877
+    },
+    {
+      "epoch": 2.610703975127693,
+      "grad_norm": 0.35081568597263124,
+      "learning_rate": 3.244137093463725e-06,
+      "loss": 0.0194,
+      "step": 5878
+    },
+    {
+      "epoch": 2.61114812347324,
+      "grad_norm": 0.5690602488132523,
+      "learning_rate": 3.2423221526834253e-06,
+      "loss": 0.0381,
+      "step": 5879
+    },
+    {
+      "epoch": 2.6115922718187874,
+      "grad_norm": 0.36352934356142463,
+      "learning_rate": 3.2405074761119648e-06,
+      "loss": 0.0282,
+      "step": 5880
+    },
+    {
+      "epoch": 2.612036420164335,
+      "grad_norm": 0.3774759920653436,
+      "learning_rate": 3.2386930640221193e-06,
+      "loss": 0.0344,
+      "step": 5881
+    },
+    {
+      "epoch": 2.6124805685098824,
+      "grad_norm": 0.5371544305687926,
+      "learning_rate": 3.2368789166866244e-06,
+      "loss": 0.0497,
+      "step": 5882
+    },
+    {
+      "epoch": 2.6129247168554297,
+      "grad_norm": 0.38943009927535477,
+      "learning_rate": 3.2350650343781775e-06,
+      "loss": 0.0227,
+      "step": 5883
+    },
+    {
+      "epoch": 2.613368865200977,
+      "grad_norm": 0.4501911900625063,
+      "learning_rate": 3.2332514173694396e-06,
+      "loss": 0.0395,
+      "step": 5884
+    },
+    {
+      "epoch": 2.6138130135465243,
+      "grad_norm": 0.4234499617743984,
+      "learning_rate": 3.2314380659330246e-06,
+      "loss": 0.0301,
+      "step": 5885
+    },
+    {
+      "epoch": 2.614257161892072,
+      "grad_norm": 0.4318219863554864,
+      "learning_rate": 3.22962498034151e-06,
+      "loss": 0.0294,
+      "step": 5886
+    },
+    {
+      "epoch": 2.6147013102376193,
+      "grad_norm": 0.503946757308322,
+      "learning_rate": 3.227812160867436e-06,
+      "loss": 0.0394,
+      "step": 5887
+    },
+    {
+      "epoch": 2.615145458583167,
+      "grad_norm": 0.3890770124231559,
+      "learning_rate": 3.2259996077832976e-06,
+      "loss": 0.0267,
+      "step": 5888
+    },
+    {
+      "epoch": 2.6155896069287143,
+      "grad_norm": 0.3697580136655012,
+      "learning_rate": 3.2241873213615514e-06,
+      "loss": 0.034,
+      "step": 5889
+    },
+    {
+      "epoch": 2.6160337552742616,
+      "grad_norm": 0.5649235090654835,
+      "learning_rate": 3.2223753018746186e-06,
+      "loss": 0.0272,
+      "step": 5890
+    },
+    {
+      "epoch": 2.616477903619809,
+      "grad_norm": 0.3708345448824024,
+      "learning_rate": 3.220563549594874e-06,
+      "loss": 0.035,
+      "step": 5891
+    },
+    {
+      "epoch": 2.616922051965356,
+      "grad_norm": 0.3432122520668072,
+      "learning_rate": 3.2187520647946547e-06,
+      "loss": 0.0283,
+      "step": 5892
+    },
+    {
+      "epoch": 2.617366200310904,
+      "grad_norm": 0.42022838385345057,
+      "learning_rate": 3.2169408477462594e-06,
+      "loss": 0.0364,
+      "step": 5893
+    },
+    {
+      "epoch": 2.6178103486564512,
+      "grad_norm": 0.44719094760718237,
+      "learning_rate": 3.2151298987219437e-06,
+      "loss": 0.0325,
+      "step": 5894
+    },
+    {
+      "epoch": 2.6182544970019985,
+      "grad_norm": 0.35618319174657215,
+      "learning_rate": 3.2133192179939215e-06,
+      "loss": 0.0256,
+      "step": 5895
+    },
+    {
+      "epoch": 2.6186986453475463,
+      "grad_norm": 0.3925862521483163,
+      "learning_rate": 3.2115088058343725e-06,
+      "loss": 0.0282,
+      "step": 5896
+    },
+    {
+      "epoch": 2.6191427936930936,
+      "grad_norm": 0.45474329229543986,
+      "learning_rate": 3.209698662515432e-06,
+      "loss": 0.0296,
+      "step": 5897
+    },
+    {
+      "epoch": 2.619586942038641,
+      "grad_norm": 0.3953757491585112,
+      "learning_rate": 3.2078887883091948e-06,
+      "loss": 0.0422,
+      "step": 5898
+    },
+    {
+      "epoch": 2.620031090384188,
+      "grad_norm": 0.36604722451208876,
+      "learning_rate": 3.2060791834877136e-06,
+      "loss": 0.029,
+      "step": 5899
+    },
+    {
+      "epoch": 2.620475238729736,
+      "grad_norm": 0.32800664474353197,
+      "learning_rate": 3.204269848323004e-06,
+      "loss": 0.0295,
+      "step": 5900
+    },
+    {
+      "epoch": 2.620919387075283,
+      "grad_norm": 0.5458287273191036,
+      "learning_rate": 3.2024607830870424e-06,
+      "loss": 0.0475,
+      "step": 5901
+    },
+    {
+      "epoch": 2.6213635354208304,
+      "grad_norm": 0.43819732935782096,
+      "learning_rate": 3.2006519880517597e-06,
+      "loss": 0.0452,
+      "step": 5902
+    },
+    {
+      "epoch": 2.621807683766378,
+      "grad_norm": 0.36142275590555617,
+      "learning_rate": 3.1988434634890476e-06,
+      "loss": 0.0277,
+      "step": 5903
+    },
+    {
+      "epoch": 2.6222518321119255,
+      "grad_norm": 0.4477755001983317,
+      "learning_rate": 3.197035209670761e-06,
+      "loss": 0.0294,
+      "step": 5904
+    },
+    {
+      "epoch": 2.6226959804574728,
+      "grad_norm": 0.46965829732778464,
+      "learning_rate": 3.1952272268687083e-06,
+      "loss": 0.0336,
+      "step": 5905
+    },
+    {
+      "epoch": 2.62314012880302,
+      "grad_norm": 0.39994107522556216,
+      "learning_rate": 3.19341951535466e-06,
+      "loss": 0.0214,
+      "step": 5906
+    },
+    {
+      "epoch": 2.6235842771485673,
+      "grad_norm": 0.4238276596672287,
+      "learning_rate": 3.1916120754003475e-06,
+      "loss": 0.0269,
+      "step": 5907
+    },
+    {
+      "epoch": 2.624028425494115,
+      "grad_norm": 1.8594679605411257,
+      "learning_rate": 3.1898049072774605e-06,
+      "loss": 0.0367,
+      "step": 5908
+    },
+    {
+      "epoch": 2.6244725738396624,
+      "grad_norm": 0.31000445284347417,
+      "learning_rate": 3.1879980112576457e-06,
+      "loss": 0.0248,
+      "step": 5909
+    },
+    {
+      "epoch": 2.62491672218521,
+      "grad_norm": 0.40069203774397927,
+      "learning_rate": 3.1861913876125093e-06,
+      "loss": 0.0314,
+      "step": 5910
+    },
+    {
+      "epoch": 2.6253608705307574,
+      "grad_norm": 0.39410237773498447,
+      "learning_rate": 3.1843850366136198e-06,
+      "loss": 0.0293,
+      "step": 5911
+    },
+    {
+      "epoch": 2.6258050188763047,
+      "grad_norm": 0.3268432306409947,
+      "learning_rate": 3.182578958532499e-06,
+      "loss": 0.0266,
+      "step": 5912
+    },
+    {
+      "epoch": 2.626249167221852,
+      "grad_norm": 0.45108591338261844,
+      "learning_rate": 3.180773153640635e-06,
+      "loss": 0.0341,
+      "step": 5913
+    },
+    {
+      "epoch": 2.6266933155673993,
+      "grad_norm": 0.43982934139295143,
+      "learning_rate": 3.178967622209469e-06,
+      "loss": 0.0328,
+      "step": 5914
+    },
+    {
+      "epoch": 2.627137463912947,
+      "grad_norm": 0.34491560520514963,
+      "learning_rate": 3.177162364510404e-06,
+      "loss": 0.0243,
+      "step": 5915
+    },
+    {
+      "epoch": 2.6275816122584943,
+      "grad_norm": 0.404112232349346,
+      "learning_rate": 3.175357380814799e-06,
+      "loss": 0.0233,
+      "step": 5916
+    },
+    {
+      "epoch": 2.628025760604042,
+      "grad_norm": 0.4687180614976695,
+      "learning_rate": 3.1735526713939757e-06,
+      "loss": 0.0337,
+      "step": 5917
+    },
+    {
+      "epoch": 2.6284699089495893,
+      "grad_norm": 0.37747833327236135,
+      "learning_rate": 3.1717482365192106e-06,
+      "loss": 0.0263,
+      "step": 5918
+    },
+    {
+      "epoch": 2.6289140572951366,
+      "grad_norm": 0.38430475009185994,
+      "learning_rate": 3.1699440764617432e-06,
+      "loss": 0.0398,
+      "step": 5919
+    },
+    {
+      "epoch": 2.629358205640684,
+      "grad_norm": 0.39716398492985894,
+      "learning_rate": 3.1681401914927678e-06,
+      "loss": 0.0324,
+      "step": 5920
+    },
+    {
+      "epoch": 2.629802353986231,
+      "grad_norm": 0.40985834454742587,
+      "learning_rate": 3.1663365818834406e-06,
+      "loss": 0.0365,
+      "step": 5921
+    },
+    {
+      "epoch": 2.630246502331779,
+      "grad_norm": 0.32571897096876656,
+      "learning_rate": 3.1645332479048734e-06,
+      "loss": 0.03,
+      "step": 5922
+    },
+    {
+      "epoch": 2.630690650677326,
+      "grad_norm": 0.41303965455433855,
+      "learning_rate": 3.1627301898281364e-06,
+      "loss": 0.0256,
+      "step": 5923
+    },
+    {
+      "epoch": 2.6311347990228735,
+      "grad_norm": 0.4244488216427459,
+      "learning_rate": 3.1609274079242625e-06,
+      "loss": 0.0383,
+      "step": 5924
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.5719404451011891,
+      "learning_rate": 3.159124902464241e-06,
+      "loss": 0.0549,
+      "step": 5925
+    },
+    {
+      "epoch": 2.6320230957139685,
+      "grad_norm": 0.4271035366652883,
+      "learning_rate": 3.1573226737190164e-06,
+      "loss": 0.0365,
+      "step": 5926
+    },
+    {
+      "epoch": 2.632467244059516,
+      "grad_norm": 0.3779054549587856,
+      "learning_rate": 3.155520721959496e-06,
+      "loss": 0.0338,
+      "step": 5927
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.3266233056591113,
+      "learning_rate": 3.1537190474565437e-06,
+      "loss": 0.0219,
+      "step": 5928
+    },
+    {
+      "epoch": 2.633355540750611,
+      "grad_norm": 0.6026007410470748,
+      "learning_rate": 3.151917650480979e-06,
+      "loss": 0.0593,
+      "step": 5929
+    },
+    {
+      "epoch": 2.633799689096158,
+      "grad_norm": 0.47698196221281425,
+      "learning_rate": 3.1501165313035877e-06,
+      "loss": 0.0393,
+      "step": 5930
+    },
+    {
+      "epoch": 2.6342438374417054,
+      "grad_norm": 0.47225489877722016,
+      "learning_rate": 3.148315690195104e-06,
+      "loss": 0.036,
+      "step": 5931
+    },
+    {
+      "epoch": 2.634687985787253,
+      "grad_norm": 0.32874590760012345,
+      "learning_rate": 3.146515127426228e-06,
+      "loss": 0.021,
+      "step": 5932
+    },
+    {
+      "epoch": 2.6351321341328005,
+      "grad_norm": 0.4293763106482794,
+      "learning_rate": 3.144714843267613e-06,
+      "loss": 0.0334,
+      "step": 5933
+    },
+    {
+      "epoch": 2.6355762824783477,
+      "grad_norm": 0.34918767908800713,
+      "learning_rate": 3.142914837989873e-06,
+      "loss": 0.0235,
+      "step": 5934
+    },
+    {
+      "epoch": 2.636020430823895,
+      "grad_norm": 0.514273957879937,
+      "learning_rate": 3.1411151118635774e-06,
+      "loss": 0.0335,
+      "step": 5935
+    },
+    {
+      "epoch": 2.6364645791694423,
+      "grad_norm": 0.42893295828776457,
+      "learning_rate": 3.1393156651592597e-06,
+      "loss": 0.0271,
+      "step": 5936
+    },
+    {
+      "epoch": 2.63690872751499,
+      "grad_norm": 0.5880664186723723,
+      "learning_rate": 3.137516498147405e-06,
+      "loss": 0.0346,
+      "step": 5937
+    },
+    {
+      "epoch": 2.6373528758605373,
+      "grad_norm": 0.4554614929699438,
+      "learning_rate": 3.1357176110984578e-06,
+      "loss": 0.0466,
+      "step": 5938
+    },
+    {
+      "epoch": 2.637797024206085,
+      "grad_norm": 0.4626649260029378,
+      "learning_rate": 3.1339190042828227e-06,
+      "loss": 0.0325,
+      "step": 5939
+    },
+    {
+      "epoch": 2.6382411725516324,
+      "grad_norm": 0.385619247407723,
+      "learning_rate": 3.132120677970859e-06,
+      "loss": 0.0317,
+      "step": 5940
+    },
+    {
+      "epoch": 2.6386853208971797,
+      "grad_norm": 0.772804476594651,
+      "learning_rate": 3.1303226324328896e-06,
+      "loss": 0.041,
+      "step": 5941
+    },
+    {
+      "epoch": 2.639129469242727,
+      "grad_norm": 0.7245537321538326,
+      "learning_rate": 3.1285248679391866e-06,
+      "loss": 0.0335,
+      "step": 5942
+    },
+    {
+      "epoch": 2.6395736175882742,
+      "grad_norm": 0.4668910845302412,
+      "learning_rate": 3.1267273847599888e-06,
+      "loss": 0.0282,
+      "step": 5943
+    },
+    {
+      "epoch": 2.640017765933822,
+      "grad_norm": 0.4437588379794779,
+      "learning_rate": 3.1249301831654842e-06,
+      "loss": 0.0248,
+      "step": 5944
+    },
+    {
+      "epoch": 2.6404619142793693,
+      "grad_norm": 0.391809580171727,
+      "learning_rate": 3.123133263425827e-06,
+      "loss": 0.0218,
+      "step": 5945
+    },
+    {
+      "epoch": 2.6409060626249166,
+      "grad_norm": 0.4573419094850795,
+      "learning_rate": 3.1213366258111207e-06,
+      "loss": 0.0377,
+      "step": 5946
+    },
+    {
+      "epoch": 2.6413502109704643,
+      "grad_norm": 0.3567944286524997,
+      "learning_rate": 3.1195402705914337e-06,
+      "loss": 0.0244,
+      "step": 5947
+    },
+    {
+      "epoch": 2.6417943593160116,
+      "grad_norm": 0.5003131202018632,
+      "learning_rate": 3.1177441980367873e-06,
+      "loss": 0.0389,
+      "step": 5948
+    },
+    {
+      "epoch": 2.642238507661559,
+      "grad_norm": 0.5664593011176533,
+      "learning_rate": 3.115948408417162e-06,
+      "loss": 0.027,
+      "step": 5949
+    },
+    {
+      "epoch": 2.642682656007106,
+      "grad_norm": 0.4398029849585822,
+      "learning_rate": 3.1141529020024964e-06,
+      "loss": 0.0306,
+      "step": 5950
+    },
+    {
+      "epoch": 2.643126804352654,
+      "grad_norm": 0.3797625962528218,
+      "learning_rate": 3.1123576790626825e-06,
+      "loss": 0.0266,
+      "step": 5951
+    },
+    {
+      "epoch": 2.643570952698201,
+      "grad_norm": 0.41053216679804944,
+      "learning_rate": 3.1105627398675743e-06,
+      "loss": 0.0352,
+      "step": 5952
+    },
+    {
+      "epoch": 2.6440151010437485,
+      "grad_norm": 0.3836633925963885,
+      "learning_rate": 3.1087680846869844e-06,
+      "loss": 0.0324,
+      "step": 5953
+    },
+    {
+      "epoch": 2.644459249389296,
+      "grad_norm": 0.439056518784003,
+      "learning_rate": 3.1069737137906776e-06,
+      "loss": 0.042,
+      "step": 5954
+    },
+    {
+      "epoch": 2.6449033977348435,
+      "grad_norm": 0.45357743660503874,
+      "learning_rate": 3.1051796274483776e-06,
+      "loss": 0.0343,
+      "step": 5955
+    },
+    {
+      "epoch": 2.645347546080391,
+      "grad_norm": 0.38982067943155413,
+      "learning_rate": 3.1033858259297677e-06,
+      "loss": 0.0374,
+      "step": 5956
+    },
+    {
+      "epoch": 2.645791694425938,
+      "grad_norm": 0.36926658548126234,
+      "learning_rate": 3.1015923095044844e-06,
+      "loss": 0.0307,
+      "step": 5957
+    },
+    {
+      "epoch": 2.646235842771486,
+      "grad_norm": 0.5321168036178672,
+      "learning_rate": 3.0997990784421273e-06,
+      "loss": 0.0571,
+      "step": 5958
+    },
+    {
+      "epoch": 2.646679991117033,
+      "grad_norm": 0.46761980336540304,
+      "learning_rate": 3.0980061330122463e-06,
+      "loss": 0.0398,
+      "step": 5959
+    },
+    {
+      "epoch": 2.6471241394625804,
+      "grad_norm": 0.41367845862555414,
+      "learning_rate": 3.096213473484354e-06,
+      "loss": 0.0211,
+      "step": 5960
+    },
+    {
+      "epoch": 2.647568287808128,
+      "grad_norm": 0.42099626188495787,
+      "learning_rate": 3.094421100127916e-06,
+      "loss": 0.0365,
+      "step": 5961
+    },
+    {
+      "epoch": 2.6480124361536754,
+      "grad_norm": 0.4995405826324344,
+      "learning_rate": 3.092629013212356e-06,
+      "loss": 0.0394,
+      "step": 5962
+    },
+    {
+      "epoch": 2.6484565844992227,
+      "grad_norm": 0.4212882639660684,
+      "learning_rate": 3.090837213007054e-06,
+      "loss": 0.0357,
+      "step": 5963
+    },
+    {
+      "epoch": 2.64890073284477,
+      "grad_norm": 0.4419706138728259,
+      "learning_rate": 3.0890456997813534e-06,
+      "loss": 0.0382,
+      "step": 5964
+    },
+    {
+      "epoch": 2.6493448811903173,
+      "grad_norm": 0.5281379267202536,
+      "learning_rate": 3.087254473804544e-06,
+      "loss": 0.0442,
+      "step": 5965
+    },
+    {
+      "epoch": 2.649789029535865,
+      "grad_norm": 0.9881997927190076,
+      "learning_rate": 3.0854635353458795e-06,
+      "loss": 0.0348,
+      "step": 5966
+    },
+    {
+      "epoch": 2.6502331778814123,
+      "grad_norm": 0.4051851073824622,
+      "learning_rate": 3.083672884674568e-06,
+      "loss": 0.0328,
+      "step": 5967
+    },
+    {
+      "epoch": 2.65067732622696,
+      "grad_norm": 0.43295642683770935,
+      "learning_rate": 3.081882522059774e-06,
+      "loss": 0.0445,
+      "step": 5968
+    },
+    {
+      "epoch": 2.6511214745725074,
+      "grad_norm": 0.7752250617010704,
+      "learning_rate": 3.0800924477706185e-06,
+      "loss": 0.0366,
+      "step": 5969
+    },
+    {
+      "epoch": 2.6515656229180546,
+      "grad_norm": 0.551166830656724,
+      "learning_rate": 3.0783026620761846e-06,
+      "loss": 0.0313,
+      "step": 5970
+    },
+    {
+      "epoch": 2.652009771263602,
+      "grad_norm": 0.35152614881285826,
+      "learning_rate": 3.076513165245504e-06,
+      "loss": 0.0279,
+      "step": 5971
+    },
+    {
+      "epoch": 2.6524539196091492,
+      "grad_norm": 0.39759843537124434,
+      "learning_rate": 3.0747239575475674e-06,
+      "loss": 0.0363,
+      "step": 5972
+    },
+    {
+      "epoch": 2.652898067954697,
+      "grad_norm": 0.4918571550335759,
+      "learning_rate": 3.072935039251327e-06,
+      "loss": 0.03,
+      "step": 5973
+    },
+    {
+      "epoch": 2.6533422163002442,
+      "grad_norm": 0.3903119645882834,
+      "learning_rate": 3.071146410625682e-06,
+      "loss": 0.0262,
+      "step": 5974
+    },
+    {
+      "epoch": 2.6537863646457915,
+      "grad_norm": 0.3990290251739691,
+      "learning_rate": 3.0693580719395e-06,
+      "loss": 0.0457,
+      "step": 5975
+    },
+    {
+      "epoch": 2.6542305129913393,
+      "grad_norm": 0.4382406546867813,
+      "learning_rate": 3.067570023461594e-06,
+      "loss": 0.0345,
+      "step": 5976
+    },
+    {
+      "epoch": 2.6546746613368866,
+      "grad_norm": 0.4165975886070993,
+      "learning_rate": 3.065782265460741e-06,
+      "loss": 0.0273,
+      "step": 5977
+    },
+    {
+      "epoch": 2.655118809682434,
+      "grad_norm": 0.5153364937600957,
+      "learning_rate": 3.06399479820567e-06,
+      "loss": 0.0386,
+      "step": 5978
+    },
+    {
+      "epoch": 2.655562958027981,
+      "grad_norm": 0.5238135987655357,
+      "learning_rate": 3.062207621965067e-06,
+      "loss": 0.0352,
+      "step": 5979
+    },
+    {
+      "epoch": 2.656007106373529,
+      "grad_norm": 0.427195897063526,
+      "learning_rate": 3.0604207370075743e-06,
+      "loss": 0.0252,
+      "step": 5980
+    },
+    {
+      "epoch": 2.656451254719076,
+      "grad_norm": 0.46584491776786274,
+      "learning_rate": 3.0586341436017954e-06,
+      "loss": 0.0476,
+      "step": 5981
+    },
+    {
+      "epoch": 2.6568954030646235,
+      "grad_norm": 0.43759944048446714,
+      "learning_rate": 3.056847842016282e-06,
+      "loss": 0.0273,
+      "step": 5982
+    },
+    {
+      "epoch": 2.657339551410171,
+      "grad_norm": 0.47962982437840673,
+      "learning_rate": 3.0550618325195457e-06,
+      "loss": 0.0292,
+      "step": 5983
+    },
+    {
+      "epoch": 2.6577836997557185,
+      "grad_norm": 0.40526954332365667,
+      "learning_rate": 3.053276115380055e-06,
+      "loss": 0.0249,
+      "step": 5984
+    },
+    {
+      "epoch": 2.6582278481012658,
+      "grad_norm": 0.5052423442608693,
+      "learning_rate": 3.0514906908662346e-06,
+      "loss": 0.0247,
+      "step": 5985
+    },
+    {
+      "epoch": 2.658671996446813,
+      "grad_norm": 0.5146431419630214,
+      "learning_rate": 3.0497055592464596e-06,
+      "loss": 0.0381,
+      "step": 5986
+    },
+    {
+      "epoch": 2.659116144792361,
+      "grad_norm": 0.40107763434385424,
+      "learning_rate": 3.04792072078907e-06,
+      "loss": 0.0305,
+      "step": 5987
+    },
+    {
+      "epoch": 2.659560293137908,
+      "grad_norm": 0.5005148653434598,
+      "learning_rate": 3.046136175762357e-06,
+      "loss": 0.0308,
+      "step": 5988
+    },
+    {
+      "epoch": 2.6600044414834554,
+      "grad_norm": 0.5429487251371317,
+      "learning_rate": 3.0443519244345666e-06,
+      "loss": 0.0359,
+      "step": 5989
+    },
+    {
+      "epoch": 2.660448589829003,
+      "grad_norm": 0.5062800108359077,
+      "learning_rate": 3.0425679670739026e-06,
+      "loss": 0.0409,
+      "step": 5990
+    },
+    {
+      "epoch": 2.6608927381745504,
+      "grad_norm": 0.37298776061190037,
+      "learning_rate": 3.040784303948523e-06,
+      "loss": 0.0314,
+      "step": 5991
+    },
+    {
+      "epoch": 2.6613368865200977,
+      "grad_norm": 0.6131690894387388,
+      "learning_rate": 3.0390009353265458e-06,
+      "loss": 0.0276,
+      "step": 5992
+    },
+    {
+      "epoch": 2.661781034865645,
+      "grad_norm": 0.4287293989906706,
+      "learning_rate": 3.0372178614760382e-06,
+      "loss": 0.038,
+      "step": 5993
+    },
+    {
+      "epoch": 2.6622251832111923,
+      "grad_norm": 0.494407979190499,
+      "learning_rate": 3.035435082665029e-06,
+      "loss": 0.0328,
+      "step": 5994
+    },
+    {
+      "epoch": 2.66266933155674,
+      "grad_norm": 0.4258763284446213,
+      "learning_rate": 3.033652599161499e-06,
+      "loss": 0.0268,
+      "step": 5995
+    },
+    {
+      "epoch": 2.6631134799022873,
+      "grad_norm": 0.4705172448160009,
+      "learning_rate": 3.0318704112333847e-06,
+      "loss": 0.0314,
+      "step": 5996
+    },
+    {
+      "epoch": 2.663557628247835,
+      "grad_norm": 0.4749657917164724,
+      "learning_rate": 3.0300885191485797e-06,
+      "loss": 0.0363,
+      "step": 5997
+    },
+    {
+      "epoch": 2.6640017765933823,
+      "grad_norm": 0.7383043422980394,
+      "learning_rate": 3.0283069231749344e-06,
+      "loss": 0.0361,
+      "step": 5998
+    },
+    {
+      "epoch": 2.6644459249389296,
+      "grad_norm": 0.38814202685884,
+      "learning_rate": 3.026525623580252e-06,
+      "loss": 0.026,
+      "step": 5999
+    },
+    {
+      "epoch": 2.664890073284477,
+      "grad_norm": 0.3764976664025451,
+      "learning_rate": 3.0247446206322916e-06,
+      "loss": 0.0321,
+      "step": 6000
+    },
+    {
+      "epoch": 2.665334221630024,
+      "grad_norm": 0.45096374436929443,
+      "learning_rate": 3.0229639145987687e-06,
+      "loss": 0.034,
+      "step": 6001
+    },
+    {
+      "epoch": 2.665778369975572,
+      "grad_norm": 0.36848556078596323,
+      "learning_rate": 3.021183505747354e-06,
+      "loss": 0.031,
+      "step": 6002
+    },
+    {
+      "epoch": 2.6662225183211192,
+      "grad_norm": 0.3699352216981277,
+      "learning_rate": 3.0194033943456696e-06,
+      "loss": 0.028,
+      "step": 6003
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.381719196592165,
+      "learning_rate": 3.0176235806613008e-06,
+      "loss": 0.0245,
+      "step": 6004
+    },
+    {
+      "epoch": 2.6671108150122143,
+      "grad_norm": 0.4602734928315702,
+      "learning_rate": 3.0158440649617836e-06,
+      "loss": 0.039,
+      "step": 6005
+    },
+    {
+      "epoch": 2.6675549633577615,
+      "grad_norm": 0.34646717181854414,
+      "learning_rate": 3.014064847514609e-06,
+      "loss": 0.0248,
+      "step": 6006
+    },
+    {
+      "epoch": 2.667999111703309,
+      "grad_norm": 0.41630402405704753,
+      "learning_rate": 3.0122859285872214e-06,
+      "loss": 0.0364,
+      "step": 6007
+    },
+    {
+      "epoch": 2.668443260048856,
+      "grad_norm": 0.39925464642353814,
+      "learning_rate": 3.010507308447025e-06,
+      "loss": 0.0286,
+      "step": 6008
+    },
+    {
+      "epoch": 2.668887408394404,
+      "grad_norm": 0.5709915993898592,
+      "learning_rate": 3.0087289873613746e-06,
+      "loss": 0.0409,
+      "step": 6009
+    },
+    {
+      "epoch": 2.669331556739951,
+      "grad_norm": 0.5153833490415598,
+      "learning_rate": 3.0069509655975835e-06,
+      "loss": 0.0388,
+      "step": 6010
+    },
+    {
+      "epoch": 2.6697757050854984,
+      "grad_norm": 0.37873158972984705,
+      "learning_rate": 3.0051732434229185e-06,
+      "loss": 0.0254,
+      "step": 6011
+    },
+    {
+      "epoch": 2.670219853431046,
+      "grad_norm": 0.4404773311355675,
+      "learning_rate": 3.003395821104602e-06,
+      "loss": 0.0308,
+      "step": 6012
+    },
+    {
+      "epoch": 2.6706640017765935,
+      "grad_norm": 0.5123588004903873,
+      "learning_rate": 3.001618698909809e-06,
+      "loss": 0.032,
+      "step": 6013
+    },
+    {
+      "epoch": 2.6711081501221408,
+      "grad_norm": 0.41512949784084213,
+      "learning_rate": 2.999841877105672e-06,
+      "loss": 0.0277,
+      "step": 6014
+    },
+    {
+      "epoch": 2.671552298467688,
+      "grad_norm": 0.4025208152025212,
+      "learning_rate": 2.9980653559592775e-06,
+      "loss": 0.0313,
+      "step": 6015
+    },
+    {
+      "epoch": 2.671996446813236,
+      "grad_norm": 0.5344242242777247,
+      "learning_rate": 2.996289135737668e-06,
+      "loss": 0.0351,
+      "step": 6016
+    },
+    {
+      "epoch": 2.672440595158783,
+      "grad_norm": 0.4264946062568879,
+      "learning_rate": 2.994513216707838e-06,
+      "loss": 0.0263,
+      "step": 6017
+    },
+    {
+      "epoch": 2.6728847435043304,
+      "grad_norm": 0.3811830713887272,
+      "learning_rate": 2.992737599136739e-06,
+      "loss": 0.0268,
+      "step": 6018
+    },
+    {
+      "epoch": 2.673328891849878,
+      "grad_norm": 0.386704376318305,
+      "learning_rate": 2.9909622832912767e-06,
+      "loss": 0.0315,
+      "step": 6019
+    },
+    {
+      "epoch": 2.6737730401954254,
+      "grad_norm": 0.4136667792416568,
+      "learning_rate": 2.989187269438308e-06,
+      "loss": 0.0307,
+      "step": 6020
+    },
+    {
+      "epoch": 2.6742171885409727,
+      "grad_norm": 0.4174591961736545,
+      "learning_rate": 2.987412557844653e-06,
+      "loss": 0.0319,
+      "step": 6021
+    },
+    {
+      "epoch": 2.67466133688652,
+      "grad_norm": 0.42669426474826655,
+      "learning_rate": 2.985638148777078e-06,
+      "loss": 0.0288,
+      "step": 6022
+    },
+    {
+      "epoch": 2.6751054852320673,
+      "grad_norm": 0.3748882818930551,
+      "learning_rate": 2.983864042502308e-06,
+      "loss": 0.0263,
+      "step": 6023
+    },
+    {
+      "epoch": 2.675549633577615,
+      "grad_norm": 0.4706253013201654,
+      "learning_rate": 2.9820902392870197e-06,
+      "loss": 0.0284,
+      "step": 6024
+    },
+    {
+      "epoch": 2.6759937819231623,
+      "grad_norm": 0.4662239351305156,
+      "learning_rate": 2.980316739397847e-06,
+      "loss": 0.0358,
+      "step": 6025
+    },
+    {
+      "epoch": 2.67643793026871,
+      "grad_norm": 0.314532858998878,
+      "learning_rate": 2.9785435431013755e-06,
+      "loss": 0.0232,
+      "step": 6026
+    },
+    {
+      "epoch": 2.6768820786142573,
+      "grad_norm": 0.5337532314358315,
+      "learning_rate": 2.97677065066415e-06,
+      "loss": 0.0333,
+      "step": 6027
+    },
+    {
+      "epoch": 2.6773262269598046,
+      "grad_norm": 0.5956371023489798,
+      "learning_rate": 2.9749980623526633e-06,
+      "loss": 0.0381,
+      "step": 6028
+    },
+    {
+      "epoch": 2.677770375305352,
+      "grad_norm": 0.3325969522067864,
+      "learning_rate": 2.9732257784333673e-06,
+      "loss": 0.0249,
+      "step": 6029
+    },
+    {
+      "epoch": 2.678214523650899,
+      "grad_norm": 0.44226896551023326,
+      "learning_rate": 2.9714537991726656e-06,
+      "loss": 0.0321,
+      "step": 6030
+    },
+    {
+      "epoch": 2.678658671996447,
+      "grad_norm": 0.407289289939863,
+      "learning_rate": 2.9696821248369152e-06,
+      "loss": 0.0294,
+      "step": 6031
+    },
+    {
+      "epoch": 2.679102820341994,
+      "grad_norm": 0.42539794841254347,
+      "learning_rate": 2.9679107556924314e-06,
+      "loss": 0.0312,
+      "step": 6032
+    },
+    {
+      "epoch": 2.6795469686875415,
+      "grad_norm": 0.3452430100807996,
+      "learning_rate": 2.966139692005481e-06,
+      "loss": 0.0311,
+      "step": 6033
+    },
+    {
+      "epoch": 2.6799911170330892,
+      "grad_norm": 0.47114526962385006,
+      "learning_rate": 2.9643689340422844e-06,
+      "loss": 0.0351,
+      "step": 6034
+    },
+    {
+      "epoch": 2.6804352653786365,
+      "grad_norm": 0.3056012283706184,
+      "learning_rate": 2.962598482069015e-06,
+      "loss": 0.0254,
+      "step": 6035
+    },
+    {
+      "epoch": 2.680879413724184,
+      "grad_norm": 0.39834900420012637,
+      "learning_rate": 2.960828336351804e-06,
+      "loss": 0.0329,
+      "step": 6036
+    },
+    {
+      "epoch": 2.681323562069731,
+      "grad_norm": 0.4921282055431307,
+      "learning_rate": 2.9590584971567327e-06,
+      "loss": 0.0378,
+      "step": 6037
+    },
+    {
+      "epoch": 2.681767710415279,
+      "grad_norm": 0.401942181093132,
+      "learning_rate": 2.957288964749839e-06,
+      "loss": 0.0282,
+      "step": 6038
+    },
+    {
+      "epoch": 2.682211858760826,
+      "grad_norm": 0.5998754085660062,
+      "learning_rate": 2.955519739397114e-06,
+      "loss": 0.04,
+      "step": 6039
+    },
+    {
+      "epoch": 2.6826560071063734,
+      "grad_norm": 0.5229301227786665,
+      "learning_rate": 2.9537508213645026e-06,
+      "loss": 0.0439,
+      "step": 6040
+    },
+    {
+      "epoch": 2.683100155451921,
+      "grad_norm": 0.4131624612623256,
+      "learning_rate": 2.9519822109179007e-06,
+      "loss": 0.0324,
+      "step": 6041
+    },
+    {
+      "epoch": 2.6835443037974684,
+      "grad_norm": 0.3861923836427622,
+      "learning_rate": 2.950213908323164e-06,
+      "loss": 0.0277,
+      "step": 6042
+    },
+    {
+      "epoch": 2.6839884521430157,
+      "grad_norm": 0.4648268575218616,
+      "learning_rate": 2.948445913846094e-06,
+      "loss": 0.0289,
+      "step": 6043
+    },
+    {
+      "epoch": 2.684432600488563,
+      "grad_norm": 0.4732042527119129,
+      "learning_rate": 2.9466782277524554e-06,
+      "loss": 0.0309,
+      "step": 6044
+    },
+    {
+      "epoch": 2.6848767488341103,
+      "grad_norm": 0.3982899300782419,
+      "learning_rate": 2.944910850307958e-06,
+      "loss": 0.0314,
+      "step": 6045
+    },
+    {
+      "epoch": 2.685320897179658,
+      "grad_norm": 0.4153109085037015,
+      "learning_rate": 2.9431437817782705e-06,
+      "loss": 0.0339,
+      "step": 6046
+    },
+    {
+      "epoch": 2.6857650455252053,
+      "grad_norm": 0.43324892108850316,
+      "learning_rate": 2.9413770224290126e-06,
+      "loss": 0.0311,
+      "step": 6047
+    },
+    {
+      "epoch": 2.686209193870753,
+      "grad_norm": 0.28406951308335854,
+      "learning_rate": 2.9396105725257563e-06,
+      "loss": 0.0231,
+      "step": 6048
+    },
+    {
+      "epoch": 2.6866533422163004,
+      "grad_norm": 0.39298370689744644,
+      "learning_rate": 2.9378444323340316e-06,
+      "loss": 0.0304,
+      "step": 6049
+    },
+    {
+      "epoch": 2.6870974905618477,
+      "grad_norm": 0.4485933834660012,
+      "learning_rate": 2.9360786021193192e-06,
+      "loss": 0.0285,
+      "step": 6050
+    },
+    {
+      "epoch": 2.687541638907395,
+      "grad_norm": 0.35520292176022183,
+      "learning_rate": 2.934313082147053e-06,
+      "loss": 0.0226,
+      "step": 6051
+    },
+    {
+      "epoch": 2.6879857872529422,
+      "grad_norm": 0.44687364826159626,
+      "learning_rate": 2.93254787268262e-06,
+      "loss": 0.0357,
+      "step": 6052
+    },
+    {
+      "epoch": 2.68842993559849,
+      "grad_norm": 0.5271674075113904,
+      "learning_rate": 2.930782973991362e-06,
+      "loss": 0.0377,
+      "step": 6053
+    },
+    {
+      "epoch": 2.6888740839440373,
+      "grad_norm": 0.4626973474509376,
+      "learning_rate": 2.929018386338571e-06,
+      "loss": 0.0502,
+      "step": 6054
+    },
+    {
+      "epoch": 2.689318232289585,
+      "grad_norm": 0.4073720832807278,
+      "learning_rate": 2.927254109989499e-06,
+      "loss": 0.0328,
+      "step": 6055
+    },
+    {
+      "epoch": 2.6897623806351323,
+      "grad_norm": 0.38757290739790723,
+      "learning_rate": 2.9254901452093424e-06,
+      "loss": 0.031,
+      "step": 6056
+    },
+    {
+      "epoch": 2.6902065289806796,
+      "grad_norm": 0.3981460707866334,
+      "learning_rate": 2.923726492263258e-06,
+      "loss": 0.0329,
+      "step": 6057
+    },
+    {
+      "epoch": 2.690650677326227,
+      "grad_norm": 0.34586475207518175,
+      "learning_rate": 2.9219631514163514e-06,
+      "loss": 0.0229,
+      "step": 6058
+    },
+    {
+      "epoch": 2.691094825671774,
+      "grad_norm": 0.3814494157202918,
+      "learning_rate": 2.9202001229336817e-06,
+      "loss": 0.0252,
+      "step": 6059
+    },
+    {
+      "epoch": 2.691538974017322,
+      "grad_norm": 0.5320910876205124,
+      "learning_rate": 2.9184374070802633e-06,
+      "loss": 0.0528,
+      "step": 6060
+    },
+    {
+      "epoch": 2.691983122362869,
+      "grad_norm": 0.5320769049550915,
+      "learning_rate": 2.916675004121062e-06,
+      "loss": 0.0374,
+      "step": 6061
+    },
+    {
+      "epoch": 2.6924272707084165,
+      "grad_norm": 0.46814435114343156,
+      "learning_rate": 2.9149129143209974e-06,
+      "loss": 0.0366,
+      "step": 6062
+    },
+    {
+      "epoch": 2.692871419053964,
+      "grad_norm": 0.46820101387242424,
+      "learning_rate": 2.9131511379449428e-06,
+      "loss": 0.0398,
+      "step": 6063
+    },
+    {
+      "epoch": 2.6933155673995115,
+      "grad_norm": 0.42299234767365035,
+      "learning_rate": 2.9113896752577205e-06,
+      "loss": 0.0306,
+      "step": 6064
+    },
+    {
+      "epoch": 2.693759715745059,
+      "grad_norm": 0.3540454635472545,
+      "learning_rate": 2.9096285265241063e-06,
+      "loss": 0.0253,
+      "step": 6065
+    },
+    {
+      "epoch": 2.694203864090606,
+      "grad_norm": 0.47552195678010667,
+      "learning_rate": 2.9078676920088378e-06,
+      "loss": 0.0304,
+      "step": 6066
+    },
+    {
+      "epoch": 2.694648012436154,
+      "grad_norm": 0.41330422015572915,
+      "learning_rate": 2.9061071719765933e-06,
+      "loss": 0.0334,
+      "step": 6067
+    },
+    {
+      "epoch": 2.695092160781701,
+      "grad_norm": 0.3903849704575043,
+      "learning_rate": 2.9043469666920088e-06,
+      "loss": 0.0261,
+      "step": 6068
+    },
+    {
+      "epoch": 2.6955363091272484,
+      "grad_norm": 0.37706353026097844,
+      "learning_rate": 2.902587076419676e-06,
+      "loss": 0.0294,
+      "step": 6069
+    },
+    {
+      "epoch": 2.695980457472796,
+      "grad_norm": 0.3862061240373804,
+      "learning_rate": 2.900827501424133e-06,
+      "loss": 0.0354,
+      "step": 6070
+    },
+    {
+      "epoch": 2.6964246058183434,
+      "grad_norm": 0.4432184785361431,
+      "learning_rate": 2.899068241969876e-06,
+      "loss": 0.032,
+      "step": 6071
+    },
+    {
+      "epoch": 2.6968687541638907,
+      "grad_norm": 0.4814495852537153,
+      "learning_rate": 2.8973092983213493e-06,
+      "loss": 0.0335,
+      "step": 6072
+    },
+    {
+      "epoch": 2.697312902509438,
+      "grad_norm": 0.444449262745695,
+      "learning_rate": 2.8955506707429545e-06,
+      "loss": 0.0428,
+      "step": 6073
+    },
+    {
+      "epoch": 2.6977570508549853,
+      "grad_norm": 0.38128392307642117,
+      "learning_rate": 2.8937923594990435e-06,
+      "loss": 0.0253,
+      "step": 6074
+    },
+    {
+      "epoch": 2.698201199200533,
+      "grad_norm": 0.4767935660247025,
+      "learning_rate": 2.8920343648539174e-06,
+      "loss": 0.0261,
+      "step": 6075
+    },
+    {
+      "epoch": 2.6986453475460803,
+      "grad_norm": 0.5621830464805478,
+      "learning_rate": 2.8902766870718347e-06,
+      "loss": 0.0478,
+      "step": 6076
+    },
+    {
+      "epoch": 2.699089495891628,
+      "grad_norm": 0.34049032210297114,
+      "learning_rate": 2.8885193264170036e-06,
+      "loss": 0.0294,
+      "step": 6077
+    },
+    {
+      "epoch": 2.6995336442371753,
+      "grad_norm": 0.4750456467312549,
+      "learning_rate": 2.886762283153586e-06,
+      "loss": 0.0371,
+      "step": 6078
+    },
+    {
+      "epoch": 2.6999777925827226,
+      "grad_norm": 0.4893300140355996,
+      "learning_rate": 2.885005557545694e-06,
+      "loss": 0.0285,
+      "step": 6079
+    },
+    {
+      "epoch": 2.70042194092827,
+      "grad_norm": 0.4397921048743652,
+      "learning_rate": 2.8832491498573965e-06,
+      "loss": 0.0365,
+      "step": 6080
+    },
+    {
+      "epoch": 2.700866089273817,
+      "grad_norm": 0.4948308694304273,
+      "learning_rate": 2.8814930603527067e-06,
+      "loss": 0.0318,
+      "step": 6081
+    },
+    {
+      "epoch": 2.701310237619365,
+      "grad_norm": 0.4221767743708031,
+      "learning_rate": 2.8797372892955978e-06,
+      "loss": 0.0355,
+      "step": 6082
+    },
+    {
+      "epoch": 2.7017543859649122,
+      "grad_norm": 0.548027854837134,
+      "learning_rate": 2.877981836949991e-06,
+      "loss": 0.0286,
+      "step": 6083
+    },
+    {
+      "epoch": 2.7021985343104595,
+      "grad_norm": 0.50981889703332,
+      "learning_rate": 2.8762267035797607e-06,
+      "loss": 0.0346,
+      "step": 6084
+    },
+    {
+      "epoch": 2.7026426826560073,
+      "grad_norm": 0.49304824695259686,
+      "learning_rate": 2.8744718894487345e-06,
+      "loss": 0.0392,
+      "step": 6085
+    },
+    {
+      "epoch": 2.7030868310015546,
+      "grad_norm": 0.33965710652598363,
+      "learning_rate": 2.8727173948206905e-06,
+      "loss": 0.0211,
+      "step": 6086
+    },
+    {
+      "epoch": 2.703530979347102,
+      "grad_norm": 0.4906097802841706,
+      "learning_rate": 2.870963219959357e-06,
+      "loss": 0.0339,
+      "step": 6087
+    },
+    {
+      "epoch": 2.703975127692649,
+      "grad_norm": 0.45131838531292917,
+      "learning_rate": 2.869209365128417e-06,
+      "loss": 0.0327,
+      "step": 6088
+    },
+    {
+      "epoch": 2.704419276038197,
+      "grad_norm": 0.44310343452513123,
+      "learning_rate": 2.8674558305915057e-06,
+      "loss": 0.0413,
+      "step": 6089
+    },
+    {
+      "epoch": 2.704863424383744,
+      "grad_norm": 0.4694661668866994,
+      "learning_rate": 2.865702616612208e-06,
+      "loss": 0.0303,
+      "step": 6090
+    },
+    {
+      "epoch": 2.7053075727292915,
+      "grad_norm": 0.42182727027273254,
+      "learning_rate": 2.8639497234540646e-06,
+      "loss": 0.0322,
+      "step": 6091
+    },
+    {
+      "epoch": 2.705751721074839,
+      "grad_norm": 0.34911280190910793,
+      "learning_rate": 2.862197151380561e-06,
+      "loss": 0.0297,
+      "step": 6092
+    },
+    {
+      "epoch": 2.7061958694203865,
+      "grad_norm": 0.5070815814024743,
+      "learning_rate": 2.8604449006551406e-06,
+      "loss": 0.0535,
+      "step": 6093
+    },
+    {
+      "epoch": 2.7066400177659338,
+      "grad_norm": 0.6056329049163846,
+      "learning_rate": 2.8586929715411963e-06,
+      "loss": 0.0589,
+      "step": 6094
+    },
+    {
+      "epoch": 2.707084166111481,
+      "grad_norm": 0.42386614097773856,
+      "learning_rate": 2.8569413643020725e-06,
+      "loss": 0.0337,
+      "step": 6095
+    },
+    {
+      "epoch": 2.707528314457029,
+      "grad_norm": 0.5481706485098858,
+      "learning_rate": 2.855190079201067e-06,
+      "loss": 0.0346,
+      "step": 6096
+    },
+    {
+      "epoch": 2.707972462802576,
+      "grad_norm": 0.4317878891723994,
+      "learning_rate": 2.8534391165014275e-06,
+      "loss": 0.0447,
+      "step": 6097
+    },
+    {
+      "epoch": 2.7084166111481234,
+      "grad_norm": 0.4182536684547791,
+      "learning_rate": 2.8516884764663512e-06,
+      "loss": 0.0381,
+      "step": 6098
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.39160951294185736,
+      "learning_rate": 2.849938159358989e-06,
+      "loss": 0.0261,
+      "step": 6099
+    },
+    {
+      "epoch": 2.7093049078392184,
+      "grad_norm": 0.3551951087739855,
+      "learning_rate": 2.848188165442446e-06,
+      "loss": 0.0293,
+      "step": 6100
+    },
+    {
+      "epoch": 2.7097490561847657,
+      "grad_norm": 0.4415214034913667,
+      "learning_rate": 2.846438494979774e-06,
+      "loss": 0.0374,
+      "step": 6101
+    },
+    {
+      "epoch": 2.710193204530313,
+      "grad_norm": 0.6570425809867534,
+      "learning_rate": 2.844689148233979e-06,
+      "loss": 0.0493,
+      "step": 6102
+    },
+    {
+      "epoch": 2.7106373528758603,
+      "grad_norm": 0.42365442096686196,
+      "learning_rate": 2.842940125468019e-06,
+      "loss": 0.0382,
+      "step": 6103
+    },
+    {
+      "epoch": 2.711081501221408,
+      "grad_norm": 0.4801471324952032,
+      "learning_rate": 2.8411914269447984e-06,
+      "loss": 0.0252,
+      "step": 6104
+    },
+    {
+      "epoch": 2.7115256495669553,
+      "grad_norm": 0.3451437825598062,
+      "learning_rate": 2.8394430529271777e-06,
+      "loss": 0.0299,
+      "step": 6105
+    },
+    {
+      "epoch": 2.711969797912503,
+      "grad_norm": 0.5219358142266073,
+      "learning_rate": 2.8376950036779683e-06,
+      "loss": 0.0295,
+      "step": 6106
+    },
+    {
+      "epoch": 2.7124139462580503,
+      "grad_norm": 0.3693933163750994,
+      "learning_rate": 2.8359472794599307e-06,
+      "loss": 0.0277,
+      "step": 6107
+    },
+    {
+      "epoch": 2.7128580946035976,
+      "grad_norm": 0.4033904609462883,
+      "learning_rate": 2.8341998805357796e-06,
+      "loss": 0.0265,
+      "step": 6108
+    },
+    {
+      "epoch": 2.713302242949145,
+      "grad_norm": 0.4144874366737247,
+      "learning_rate": 2.832452807168175e-06,
+      "loss": 0.0268,
+      "step": 6109
+    },
+    {
+      "epoch": 2.713746391294692,
+      "grad_norm": 0.5838661054136332,
+      "learning_rate": 2.8307060596197337e-06,
+      "loss": 0.0384,
+      "step": 6110
+    },
+    {
+      "epoch": 2.71419053964024,
+      "grad_norm": 0.5420801857113182,
+      "learning_rate": 2.8289596381530214e-06,
+      "loss": 0.0305,
+      "step": 6111
+    },
+    {
+      "epoch": 2.714634687985787,
+      "grad_norm": 0.42444475722005587,
+      "learning_rate": 2.8272135430305558e-06,
+      "loss": 0.041,
+      "step": 6112
+    },
+    {
+      "epoch": 2.7150788363313345,
+      "grad_norm": 0.36578048462525586,
+      "learning_rate": 2.825467774514803e-06,
+      "loss": 0.0224,
+      "step": 6113
+    },
+    {
+      "epoch": 2.7155229846768822,
+      "grad_norm": 0.33007914592544324,
+      "learning_rate": 2.823722332868185e-06,
+      "loss": 0.0252,
+      "step": 6114
+    },
+    {
+      "epoch": 2.7159671330224295,
+      "grad_norm": 0.5564364593461901,
+      "learning_rate": 2.821977218353067e-06,
+      "loss": 0.0304,
+      "step": 6115
+    },
+    {
+      "epoch": 2.716411281367977,
+      "grad_norm": 0.539249473006622,
+      "learning_rate": 2.820232431231771e-06,
+      "loss": 0.0417,
+      "step": 6116
+    },
+    {
+      "epoch": 2.716855429713524,
+      "grad_norm": 0.39638502515608776,
+      "learning_rate": 2.818487971766568e-06,
+      "loss": 0.0276,
+      "step": 6117
+    },
+    {
+      "epoch": 2.717299578059072,
+      "grad_norm": 0.3923969070351993,
+      "learning_rate": 2.816743840219681e-06,
+      "loss": 0.0345,
+      "step": 6118
+    },
+    {
+      "epoch": 2.717743726404619,
+      "grad_norm": 0.7090345157048296,
+      "learning_rate": 2.8150000368532826e-06,
+      "loss": 0.0459,
+      "step": 6119
+    },
+    {
+      "epoch": 2.7181878747501664,
+      "grad_norm": 0.49769141697815394,
+      "learning_rate": 2.8132565619294943e-06,
+      "loss": 0.0312,
+      "step": 6120
+    },
+    {
+      "epoch": 2.718632023095714,
+      "grad_norm": 0.3567290862892367,
+      "learning_rate": 2.8115134157103906e-06,
+      "loss": 0.0427,
+      "step": 6121
+    },
+    {
+      "epoch": 2.7190761714412615,
+      "grad_norm": 0.41308801601010114,
+      "learning_rate": 2.809770598457997e-06,
+      "loss": 0.0308,
+      "step": 6122
+    },
+    {
+      "epoch": 2.7195203197868087,
+      "grad_norm": 0.42911807998450446,
+      "learning_rate": 2.8080281104342875e-06,
+      "loss": 0.0389,
+      "step": 6123
+    },
+    {
+      "epoch": 2.719964468132356,
+      "grad_norm": 0.3668652303738415,
+      "learning_rate": 2.8062859519011885e-06,
+      "loss": 0.031,
+      "step": 6124
+    },
+    {
+      "epoch": 2.7204086164779038,
+      "grad_norm": 0.4729128881363199,
+      "learning_rate": 2.8045441231205773e-06,
+      "loss": 0.0359,
+      "step": 6125
+    },
+    {
+      "epoch": 2.720852764823451,
+      "grad_norm": 0.49961860986096596,
+      "learning_rate": 2.802802624354276e-06,
+      "loss": 0.0284,
+      "step": 6126
+    },
+    {
+      "epoch": 2.7212969131689984,
+      "grad_norm": 0.38704332861461993,
+      "learning_rate": 2.8010614558640653e-06,
+      "loss": 0.0306,
+      "step": 6127
+    },
+    {
+      "epoch": 2.721741061514546,
+      "grad_norm": 0.36762472796404744,
+      "learning_rate": 2.7993206179116706e-06,
+      "loss": 0.0297,
+      "step": 6128
+    },
+    {
+      "epoch": 2.7221852098600934,
+      "grad_norm": 0.40237829108230383,
+      "learning_rate": 2.79758011075877e-06,
+      "loss": 0.0379,
+      "step": 6129
+    },
+    {
+      "epoch": 2.7226293582056407,
+      "grad_norm": 0.39422460053368863,
+      "learning_rate": 2.7958399346669916e-06,
+      "loss": 0.0324,
+      "step": 6130
+    },
+    {
+      "epoch": 2.723073506551188,
+      "grad_norm": 0.4944769619389698,
+      "learning_rate": 2.7941000898979153e-06,
+      "loss": 0.0365,
+      "step": 6131
+    },
+    {
+      "epoch": 2.7235176548967353,
+      "grad_norm": 0.43625796235808223,
+      "learning_rate": 2.7923605767130644e-06,
+      "loss": 0.0385,
+      "step": 6132
+    },
+    {
+      "epoch": 2.723961803242283,
+      "grad_norm": 0.39500536045009,
+      "learning_rate": 2.790621395373921e-06,
+      "loss": 0.0247,
+      "step": 6133
+    },
+    {
+      "epoch": 2.7244059515878303,
+      "grad_norm": 0.4784483734678597,
+      "learning_rate": 2.7888825461419124e-06,
+      "loss": 0.0226,
+      "step": 6134
+    },
+    {
+      "epoch": 2.724850099933378,
+      "grad_norm": 0.5298205495609496,
+      "learning_rate": 2.7871440292784167e-06,
+      "loss": 0.0407,
+      "step": 6135
+    },
+    {
+      "epoch": 2.7252942482789253,
+      "grad_norm": 0.46575202716742126,
+      "learning_rate": 2.7854058450447657e-06,
+      "loss": 0.0412,
+      "step": 6136
+    },
+    {
+      "epoch": 2.7257383966244726,
+      "grad_norm": 0.4738047535073062,
+      "learning_rate": 2.783667993702234e-06,
+      "loss": 0.0385,
+      "step": 6137
+    },
+    {
+      "epoch": 2.72618254497002,
+      "grad_norm": 0.3869731324535632,
+      "learning_rate": 2.7819304755120514e-06,
+      "loss": 0.0295,
+      "step": 6138
+    },
+    {
+      "epoch": 2.726626693315567,
+      "grad_norm": 0.3652035596631365,
+      "learning_rate": 2.7801932907353966e-06,
+      "loss": 0.0295,
+      "step": 6139
+    },
+    {
+      "epoch": 2.727070841661115,
+      "grad_norm": 0.41533452786136,
+      "learning_rate": 2.778456439633398e-06,
+      "loss": 0.0242,
+      "step": 6140
+    },
+    {
+      "epoch": 2.727514990006662,
+      "grad_norm": 0.35015673870621206,
+      "learning_rate": 2.776719922467135e-06,
+      "loss": 0.0313,
+      "step": 6141
+    },
+    {
+      "epoch": 2.7279591383522095,
+      "grad_norm": 0.48649742529511003,
+      "learning_rate": 2.7749837394976353e-06,
+      "loss": 0.0403,
+      "step": 6142
+    },
+    {
+      "epoch": 2.7284032866977572,
+      "grad_norm": 0.4158217744500075,
+      "learning_rate": 2.773247890985874e-06,
+      "loss": 0.0294,
+      "step": 6143
+    },
+    {
+      "epoch": 2.7288474350433045,
+      "grad_norm": 0.3886529547809444,
+      "learning_rate": 2.7715123771927817e-06,
+      "loss": 0.0285,
+      "step": 6144
+    },
+    {
+      "epoch": 2.729291583388852,
+      "grad_norm": 0.48618253354322,
+      "learning_rate": 2.7697771983792334e-06,
+      "loss": 0.0381,
+      "step": 6145
+    },
+    {
+      "epoch": 2.729735731734399,
+      "grad_norm": 0.4239190944164018,
+      "learning_rate": 2.7680423548060574e-06,
+      "loss": 0.028,
+      "step": 6146
+    },
+    {
+      "epoch": 2.730179880079947,
+      "grad_norm": 0.6790565769570187,
+      "learning_rate": 2.766307846734032e-06,
+      "loss": 0.0344,
+      "step": 6147
+    },
+    {
+      "epoch": 2.730624028425494,
+      "grad_norm": 0.4083045746295625,
+      "learning_rate": 2.764573674423879e-06,
+      "loss": 0.0292,
+      "step": 6148
+    },
+    {
+      "epoch": 2.7310681767710414,
+      "grad_norm": 0.39741625214319276,
+      "learning_rate": 2.7628398381362765e-06,
+      "loss": 0.0369,
+      "step": 6149
+    },
+    {
+      "epoch": 2.731512325116589,
+      "grad_norm": 0.3364754032489686,
+      "learning_rate": 2.7611063381318483e-06,
+      "loss": 0.0223,
+      "step": 6150
+    },
+    {
+      "epoch": 2.7319564734621364,
+      "grad_norm": 0.33915812547001584,
+      "learning_rate": 2.7593731746711695e-06,
+      "loss": 0.0276,
+      "step": 6151
+    },
+    {
+      "epoch": 2.7324006218076837,
+      "grad_norm": 0.3730496797762164,
+      "learning_rate": 2.757640348014764e-06,
+      "loss": 0.0274,
+      "step": 6152
+    },
+    {
+      "epoch": 2.732844770153231,
+      "grad_norm": 0.3751953423117338,
+      "learning_rate": 2.755907858423108e-06,
+      "loss": 0.0363,
+      "step": 6153
+    },
+    {
+      "epoch": 2.7332889184987788,
+      "grad_norm": 0.42372086556075383,
+      "learning_rate": 2.754175706156619e-06,
+      "loss": 0.0244,
+      "step": 6154
+    },
+    {
+      "epoch": 2.733733066844326,
+      "grad_norm": 0.3686003953358143,
+      "learning_rate": 2.7524438914756714e-06,
+      "loss": 0.0227,
+      "step": 6155
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.38850199266907187,
+      "learning_rate": 2.750712414640588e-06,
+      "loss": 0.0318,
+      "step": 6156
+    },
+    {
+      "epoch": 2.734621363535421,
+      "grad_norm": 0.4212753598660164,
+      "learning_rate": 2.748981275911633e-06,
+      "loss": 0.027,
+      "step": 6157
+    },
+    {
+      "epoch": 2.7350655118809684,
+      "grad_norm": 0.3950070118539609,
+      "learning_rate": 2.747250475549033e-06,
+      "loss": 0.0347,
+      "step": 6158
+    },
+    {
+      "epoch": 2.7355096602265156,
+      "grad_norm": 0.5989665019227255,
+      "learning_rate": 2.745520013812956e-06,
+      "loss": 0.0413,
+      "step": 6159
+    },
+    {
+      "epoch": 2.735953808572063,
+      "grad_norm": 0.4076839235525957,
+      "learning_rate": 2.743789890963516e-06,
+      "loss": 0.0239,
+      "step": 6160
+    },
+    {
+      "epoch": 2.7363979569176102,
+      "grad_norm": 0.4896550081248881,
+      "learning_rate": 2.742060107260781e-06,
+      "loss": 0.0284,
+      "step": 6161
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.5961993037262533,
+      "learning_rate": 2.740330662964768e-06,
+      "loss": 0.0399,
+      "step": 6162
+    },
+    {
+      "epoch": 2.7372862536087053,
+      "grad_norm": 0.3528802270631205,
+      "learning_rate": 2.7386015583354414e-06,
+      "loss": 0.0271,
+      "step": 6163
+    },
+    {
+      "epoch": 2.737730401954253,
+      "grad_norm": 0.38484518351915703,
+      "learning_rate": 2.736872793632717e-06,
+      "loss": 0.0369,
+      "step": 6164
+    },
+    {
+      "epoch": 2.7381745502998003,
+      "grad_norm": 0.4319515735643185,
+      "learning_rate": 2.7351443691164537e-06,
+      "loss": 0.0228,
+      "step": 6165
+    },
+    {
+      "epoch": 2.7386186986453476,
+      "grad_norm": 0.42098419937374343,
+      "learning_rate": 2.7334162850464645e-06,
+      "loss": 0.0292,
+      "step": 6166
+    },
+    {
+      "epoch": 2.739062846990895,
+      "grad_norm": 0.35177937154537675,
+      "learning_rate": 2.7316885416825123e-06,
+      "loss": 0.0218,
+      "step": 6167
+    },
+    {
+      "epoch": 2.739506995336442,
+      "grad_norm": 0.41712238277892294,
+      "learning_rate": 2.7299611392843005e-06,
+      "loss": 0.0368,
+      "step": 6168
+    },
+    {
+      "epoch": 2.73995114368199,
+      "grad_norm": 0.39108614184700036,
+      "learning_rate": 2.7282340781114926e-06,
+      "loss": 0.0261,
+      "step": 6169
+    },
+    {
+      "epoch": 2.740395292027537,
+      "grad_norm": 0.526808362535383,
+      "learning_rate": 2.726507358423695e-06,
+      "loss": 0.0267,
+      "step": 6170
+    },
+    {
+      "epoch": 2.7408394403730845,
+      "grad_norm": 0.3749087607755556,
+      "learning_rate": 2.7247809804804593e-06,
+      "loss": 0.0198,
+      "step": 6171
+    },
+    {
+      "epoch": 2.741283588718632,
+      "grad_norm": 0.41805354629010627,
+      "learning_rate": 2.723054944541292e-06,
+      "loss": 0.0358,
+      "step": 6172
+    },
+    {
+      "epoch": 2.7417277370641795,
+      "grad_norm": 0.37604430696616314,
+      "learning_rate": 2.721329250865646e-06,
+      "loss": 0.0273,
+      "step": 6173
+    },
+    {
+      "epoch": 2.742171885409727,
+      "grad_norm": 0.4611005258922287,
+      "learning_rate": 2.719603899712919e-06,
+      "loss": 0.031,
+      "step": 6174
+    },
+    {
+      "epoch": 2.742616033755274,
+      "grad_norm": 0.41841947895126935,
+      "learning_rate": 2.7178788913424635e-06,
+      "loss": 0.0315,
+      "step": 6175
+    },
+    {
+      "epoch": 2.743060182100822,
+      "grad_norm": 0.41801694735137535,
+      "learning_rate": 2.7161542260135797e-06,
+      "loss": 0.0336,
+      "step": 6176
+    },
+    {
+      "epoch": 2.743504330446369,
+      "grad_norm": 0.4765481557698237,
+      "learning_rate": 2.7144299039855105e-06,
+      "loss": 0.0375,
+      "step": 6177
+    },
+    {
+      "epoch": 2.7439484787919164,
+      "grad_norm": 0.3580598459132636,
+      "learning_rate": 2.7127059255174504e-06,
+      "loss": 0.0286,
+      "step": 6178
+    },
+    {
+      "epoch": 2.744392627137464,
+      "grad_norm": 0.4101523804697921,
+      "learning_rate": 2.7109822908685445e-06,
+      "loss": 0.0274,
+      "step": 6179
+    },
+    {
+      "epoch": 2.7448367754830114,
+      "grad_norm": 0.4715955568359928,
+      "learning_rate": 2.7092590002978837e-06,
+      "loss": 0.0345,
+      "step": 6180
+    },
+    {
+      "epoch": 2.7452809238285587,
+      "grad_norm": 0.6089181044007468,
+      "learning_rate": 2.70753605406451e-06,
+      "loss": 0.0343,
+      "step": 6181
+    },
+    {
+      "epoch": 2.745725072174106,
+      "grad_norm": 0.5806324207660223,
+      "learning_rate": 2.7058134524274083e-06,
+      "loss": 0.0346,
+      "step": 6182
+    },
+    {
+      "epoch": 2.7461692205196533,
+      "grad_norm": 0.40045495221538185,
+      "learning_rate": 2.7040911956455153e-06,
+      "loss": 0.0278,
+      "step": 6183
+    },
+    {
+      "epoch": 2.746613368865201,
+      "grad_norm": 0.45888251149836196,
+      "learning_rate": 2.702369283977718e-06,
+      "loss": 0.0321,
+      "step": 6184
+    },
+    {
+      "epoch": 2.7470575172107483,
+      "grad_norm": 0.39535983779938544,
+      "learning_rate": 2.7006477176828443e-06,
+      "loss": 0.0291,
+      "step": 6185
+    },
+    {
+      "epoch": 2.747501665556296,
+      "grad_norm": 0.40307288228268773,
+      "learning_rate": 2.6989264970196795e-06,
+      "loss": 0.031,
+      "step": 6186
+    },
+    {
+      "epoch": 2.7479458139018433,
+      "grad_norm": 0.42537830184812087,
+      "learning_rate": 2.697205622246952e-06,
+      "loss": 0.0379,
+      "step": 6187
+    },
+    {
+      "epoch": 2.7483899622473906,
+      "grad_norm": 0.7384045353118717,
+      "learning_rate": 2.6954850936233357e-06,
+      "loss": 0.0388,
+      "step": 6188
+    },
+    {
+      "epoch": 2.748834110592938,
+      "grad_norm": 0.49237325688101374,
+      "learning_rate": 2.693764911407456e-06,
+      "loss": 0.0357,
+      "step": 6189
+    },
+    {
+      "epoch": 2.749278258938485,
+      "grad_norm": 0.43303507384385986,
+      "learning_rate": 2.6920450758578885e-06,
+      "loss": 0.0305,
+      "step": 6190
+    },
+    {
+      "epoch": 2.749722407284033,
+      "grad_norm": 0.5198323318463377,
+      "learning_rate": 2.690325587233148e-06,
+      "loss": 0.042,
+      "step": 6191
+    },
+    {
+      "epoch": 2.7501665556295802,
+      "grad_norm": 0.5579062089534479,
+      "learning_rate": 2.6886064457917094e-06,
+      "loss": 0.0442,
+      "step": 6192
+    },
+    {
+      "epoch": 2.750610703975128,
+      "grad_norm": 0.2992794413549194,
+      "learning_rate": 2.6868876517919845e-06,
+      "loss": 0.0214,
+      "step": 6193
+    },
+    {
+      "epoch": 2.7510548523206753,
+      "grad_norm": 0.5197412003207367,
+      "learning_rate": 2.6851692054923385e-06,
+      "loss": 0.0447,
+      "step": 6194
+    },
+    {
+      "epoch": 2.7514990006662225,
+      "grad_norm": 0.45629204802490403,
+      "learning_rate": 2.6834511071510823e-06,
+      "loss": 0.0335,
+      "step": 6195
+    },
+    {
+      "epoch": 2.75194314901177,
+      "grad_norm": 0.3939710962212484,
+      "learning_rate": 2.681733357026476e-06,
+      "loss": 0.0334,
+      "step": 6196
+    },
+    {
+      "epoch": 2.752387297357317,
+      "grad_norm": 0.4395100384242003,
+      "learning_rate": 2.680015955376727e-06,
+      "loss": 0.0269,
+      "step": 6197
+    },
+    {
+      "epoch": 2.752831445702865,
+      "grad_norm": 0.35669081976582817,
+      "learning_rate": 2.6782989024599913e-06,
+      "loss": 0.0239,
+      "step": 6198
+    },
+    {
+      "epoch": 2.753275594048412,
+      "grad_norm": 0.42101464193049454,
+      "learning_rate": 2.6765821985343676e-06,
+      "loss": 0.0297,
+      "step": 6199
+    },
+    {
+      "epoch": 2.7537197423939594,
+      "grad_norm": 0.4443116484682141,
+      "learning_rate": 2.6748658438579075e-06,
+      "loss": 0.0367,
+      "step": 6200
+    },
+    {
+      "epoch": 2.754163890739507,
+      "grad_norm": 0.3576485677615678,
+      "learning_rate": 2.6731498386886094e-06,
+      "loss": 0.0269,
+      "step": 6201
+    },
+    {
+      "epoch": 2.7546080390850545,
+      "grad_norm": 0.8029786113817223,
+      "learning_rate": 2.6714341832844137e-06,
+      "loss": 0.0366,
+      "step": 6202
+    },
+    {
+      "epoch": 2.7550521874306018,
+      "grad_norm": 0.4869826644782611,
+      "learning_rate": 2.6697188779032173e-06,
+      "loss": 0.0293,
+      "step": 6203
+    },
+    {
+      "epoch": 2.755496335776149,
+      "grad_norm": 0.43061987045181044,
+      "learning_rate": 2.6680039228028603e-06,
+      "loss": 0.0352,
+      "step": 6204
+    },
+    {
+      "epoch": 2.755940484121697,
+      "grad_norm": 0.3632393699444563,
+      "learning_rate": 2.6662893182411255e-06,
+      "loss": 0.0256,
+      "step": 6205
+    },
+    {
+      "epoch": 2.756384632467244,
+      "grad_norm": 0.37775571779218736,
+      "learning_rate": 2.6645750644757484e-06,
+      "loss": 0.0342,
+      "step": 6206
+    },
+    {
+      "epoch": 2.7568287808127914,
+      "grad_norm": 0.31327481469952234,
+      "learning_rate": 2.6628611617644133e-06,
+      "loss": 0.0209,
+      "step": 6207
+    },
+    {
+      "epoch": 2.757272929158339,
+      "grad_norm": 0.5059332837930003,
+      "learning_rate": 2.6611476103647425e-06,
+      "loss": 0.0286,
+      "step": 6208
+    },
+    {
+      "epoch": 2.7577170775038864,
+      "grad_norm": 0.39890166964039236,
+      "learning_rate": 2.6594344105343207e-06,
+      "loss": 0.0335,
+      "step": 6209
+    },
+    {
+      "epoch": 2.7581612258494337,
+      "grad_norm": 0.40337347616568126,
+      "learning_rate": 2.657721562530664e-06,
+      "loss": 0.0206,
+      "step": 6210
+    },
+    {
+      "epoch": 2.758605374194981,
+      "grad_norm": 0.44935773733297946,
+      "learning_rate": 2.656009066611244e-06,
+      "loss": 0.03,
+      "step": 6211
+    },
+    {
+      "epoch": 2.7590495225405283,
+      "grad_norm": 0.417217650015585,
+      "learning_rate": 2.654296923033481e-06,
+      "loss": 0.0354,
+      "step": 6212
+    },
+    {
+      "epoch": 2.759493670886076,
+      "grad_norm": 0.39156228698655793,
+      "learning_rate": 2.652585132054734e-06,
+      "loss": 0.0301,
+      "step": 6213
+    },
+    {
+      "epoch": 2.7599378192316233,
+      "grad_norm": 0.4181623279832991,
+      "learning_rate": 2.6508736939323187e-06,
+      "loss": 0.0416,
+      "step": 6214
+    },
+    {
+      "epoch": 2.760381967577171,
+      "grad_norm": 0.38946142096359226,
+      "learning_rate": 2.649162608923493e-06,
+      "loss": 0.022,
+      "step": 6215
+    },
+    {
+      "epoch": 2.7608261159227183,
+      "grad_norm": 0.28459219563440136,
+      "learning_rate": 2.6474518772854606e-06,
+      "loss": 0.0211,
+      "step": 6216
+    },
+    {
+      "epoch": 2.7612702642682656,
+      "grad_norm": 0.39516529374066633,
+      "learning_rate": 2.6457414992753728e-06,
+      "loss": 0.031,
+      "step": 6217
+    },
+    {
+      "epoch": 2.761714412613813,
+      "grad_norm": 0.6507735193198333,
+      "learning_rate": 2.6440314751503314e-06,
+      "loss": 0.0405,
+      "step": 6218
+    },
+    {
+      "epoch": 2.76215856095936,
+      "grad_norm": 0.39863052089616485,
+      "learning_rate": 2.6423218051673766e-06,
+      "loss": 0.0257,
+      "step": 6219
+    },
+    {
+      "epoch": 2.762602709304908,
+      "grad_norm": 0.45747750899314005,
+      "learning_rate": 2.6406124895835084e-06,
+      "loss": 0.0309,
+      "step": 6220
+    },
+    {
+      "epoch": 2.763046857650455,
+      "grad_norm": 0.39469825591698227,
+      "learning_rate": 2.6389035286556598e-06,
+      "loss": 0.0301,
+      "step": 6221
+    },
+    {
+      "epoch": 2.7634910059960025,
+      "grad_norm": 0.4493000737157054,
+      "learning_rate": 2.637194922640719e-06,
+      "loss": 0.0336,
+      "step": 6222
+    },
+    {
+      "epoch": 2.7639351543415502,
+      "grad_norm": 0.3609277682343467,
+      "learning_rate": 2.6354866717955186e-06,
+      "loss": 0.0303,
+      "step": 6223
+    },
+    {
+      "epoch": 2.7643793026870975,
+      "grad_norm": 0.3650445917615703,
+      "learning_rate": 2.6337787763768384e-06,
+      "loss": 0.0242,
+      "step": 6224
+    },
+    {
+      "epoch": 2.764823451032645,
+      "grad_norm": 0.48546314397013385,
+      "learning_rate": 2.6320712366414005e-06,
+      "loss": 0.0294,
+      "step": 6225
+    },
+    {
+      "epoch": 2.765267599378192,
+      "grad_norm": 0.32486540563930044,
+      "learning_rate": 2.6303640528458834e-06,
+      "loss": 0.0223,
+      "step": 6226
+    },
+    {
+      "epoch": 2.76571174772374,
+      "grad_norm": 0.3434852926148082,
+      "learning_rate": 2.6286572252469e-06,
+      "loss": 0.0316,
+      "step": 6227
+    },
+    {
+      "epoch": 2.766155896069287,
+      "grad_norm": 0.7142093600067161,
+      "learning_rate": 2.626950754101018e-06,
+      "loss": 0.056,
+      "step": 6228
+    },
+    {
+      "epoch": 2.7666000444148344,
+      "grad_norm": 0.6028670535609708,
+      "learning_rate": 2.6252446396647503e-06,
+      "loss": 0.0323,
+      "step": 6229
+    },
+    {
+      "epoch": 2.767044192760382,
+      "grad_norm": 0.42675838339191474,
+      "learning_rate": 2.6235388821945497e-06,
+      "loss": 0.0297,
+      "step": 6230
+    },
+    {
+      "epoch": 2.7674883411059295,
+      "grad_norm": 0.7192216736608809,
+      "learning_rate": 2.621833481946826e-06,
+      "loss": 0.0357,
+      "step": 6231
+    },
+    {
+      "epoch": 2.7679324894514767,
+      "grad_norm": 0.4766794360092108,
+      "learning_rate": 2.6201284391779303e-06,
+      "loss": 0.0331,
+      "step": 6232
+    },
+    {
+      "epoch": 2.768376637797024,
+      "grad_norm": 0.4250368407242259,
+      "learning_rate": 2.618423754144155e-06,
+      "loss": 0.0321,
+      "step": 6233
+    },
+    {
+      "epoch": 2.7688207861425718,
+      "grad_norm": 0.5212475478820174,
+      "learning_rate": 2.616719427101745e-06,
+      "loss": 0.0324,
+      "step": 6234
+    },
+    {
+      "epoch": 2.769264934488119,
+      "grad_norm": 0.3008339483542143,
+      "learning_rate": 2.6150154583068922e-06,
+      "loss": 0.0226,
+      "step": 6235
+    },
+    {
+      "epoch": 2.7697090828336663,
+      "grad_norm": 0.3816969610714372,
+      "learning_rate": 2.613311848015725e-06,
+      "loss": 0.0259,
+      "step": 6236
+    },
+    {
+      "epoch": 2.770153231179214,
+      "grad_norm": 0.38596290627054197,
+      "learning_rate": 2.611608596484335e-06,
+      "loss": 0.0288,
+      "step": 6237
+    },
+    {
+      "epoch": 2.7705973795247614,
+      "grad_norm": 0.4303882315778446,
+      "learning_rate": 2.609905703968742e-06,
+      "loss": 0.0223,
+      "step": 6238
+    },
+    {
+      "epoch": 2.7710415278703087,
+      "grad_norm": 0.4053342617572821,
+      "learning_rate": 2.6082031707249223e-06,
+      "loss": 0.03,
+      "step": 6239
+    },
+    {
+      "epoch": 2.771485676215856,
+      "grad_norm": 0.7823903528274498,
+      "learning_rate": 2.6065009970087974e-06,
+      "loss": 0.0324,
+      "step": 6240
+    },
+    {
+      "epoch": 2.7719298245614032,
+      "grad_norm": 0.42522914693804564,
+      "learning_rate": 2.6047991830762297e-06,
+      "loss": 0.038,
+      "step": 6241
+    },
+    {
+      "epoch": 2.772373972906951,
+      "grad_norm": 0.3532987358681675,
+      "learning_rate": 2.60309772918303e-06,
+      "loss": 0.0235,
+      "step": 6242
+    },
+    {
+      "epoch": 2.7728181212524983,
+      "grad_norm": 0.33902634374584634,
+      "learning_rate": 2.6013966355849618e-06,
+      "loss": 0.0327,
+      "step": 6243
+    },
+    {
+      "epoch": 2.773262269598046,
+      "grad_norm": 0.44149160136100024,
+      "learning_rate": 2.5996959025377224e-06,
+      "loss": 0.0492,
+      "step": 6244
+    },
+    {
+      "epoch": 2.7737064179435933,
+      "grad_norm": 0.48103215081730055,
+      "learning_rate": 2.597995530296963e-06,
+      "loss": 0.0267,
+      "step": 6245
+    },
+    {
+      "epoch": 2.7741505662891406,
+      "grad_norm": 0.3235934136296388,
+      "learning_rate": 2.5962955191182792e-06,
+      "loss": 0.0183,
+      "step": 6246
+    },
+    {
+      "epoch": 2.774594714634688,
+      "grad_norm": 0.3462677290866656,
+      "learning_rate": 2.59459586925721e-06,
+      "loss": 0.0272,
+      "step": 6247
+    },
+    {
+      "epoch": 2.775038862980235,
+      "grad_norm": 0.32294840719861506,
+      "learning_rate": 2.592896580969242e-06,
+      "loss": 0.0281,
+      "step": 6248
+    },
+    {
+      "epoch": 2.775483011325783,
+      "grad_norm": 0.3499956168537934,
+      "learning_rate": 2.591197654509807e-06,
+      "loss": 0.0309,
+      "step": 6249
+    },
+    {
+      "epoch": 2.77592715967133,
+      "grad_norm": 0.40008372827541966,
+      "learning_rate": 2.5894990901342833e-06,
+      "loss": 0.0268,
+      "step": 6250
+    },
+    {
+      "epoch": 2.7763713080168775,
+      "grad_norm": 0.40878589097636386,
+      "learning_rate": 2.587800888097993e-06,
+      "loss": 0.0313,
+      "step": 6251
+    },
+    {
+      "epoch": 2.776815456362425,
+      "grad_norm": 0.7688538094871912,
+      "learning_rate": 2.5861030486562084e-06,
+      "loss": 0.0479,
+      "step": 6252
+    },
+    {
+      "epoch": 2.7772596047079725,
+      "grad_norm": 0.40809425302097857,
+      "learning_rate": 2.5844055720641357e-06,
+      "loss": 0.0311,
+      "step": 6253
+    },
+    {
+      "epoch": 2.77770375305352,
+      "grad_norm": 0.37450235777366314,
+      "learning_rate": 2.5827084585769436e-06,
+      "loss": 0.0286,
+      "step": 6254
+    },
+    {
+      "epoch": 2.778147901399067,
+      "grad_norm": 0.3661400531713744,
+      "learning_rate": 2.581011708449731e-06,
+      "loss": 0.0271,
+      "step": 6255
+    },
+    {
+      "epoch": 2.778592049744615,
+      "grad_norm": 0.37170138943829045,
+      "learning_rate": 2.57931532193755e-06,
+      "loss": 0.0346,
+      "step": 6256
+    },
+    {
+      "epoch": 2.779036198090162,
+      "grad_norm": 0.30626642522716074,
+      "learning_rate": 2.577619299295398e-06,
+      "loss": 0.026,
+      "step": 6257
+    },
+    {
+      "epoch": 2.7794803464357094,
+      "grad_norm": 0.3007717640940463,
+      "learning_rate": 2.5759236407782128e-06,
+      "loss": 0.0247,
+      "step": 6258
+    },
+    {
+      "epoch": 2.779924494781257,
+      "grad_norm": 0.5105689813889596,
+      "learning_rate": 2.5742283466408803e-06,
+      "loss": 0.0477,
+      "step": 6259
+    },
+    {
+      "epoch": 2.7803686431268044,
+      "grad_norm": 0.49116305516555464,
+      "learning_rate": 2.572533417138237e-06,
+      "loss": 0.0257,
+      "step": 6260
+    },
+    {
+      "epoch": 2.7808127914723517,
+      "grad_norm": 0.356325832140254,
+      "learning_rate": 2.570838852525055e-06,
+      "loss": 0.0263,
+      "step": 6261
+    },
+    {
+      "epoch": 2.781256939817899,
+      "grad_norm": 0.4435162531660302,
+      "learning_rate": 2.569144653056058e-06,
+      "loss": 0.0435,
+      "step": 6262
+    },
+    {
+      "epoch": 2.7817010881634467,
+      "grad_norm": 0.5419805855499744,
+      "learning_rate": 2.5674508189859147e-06,
+      "loss": 0.0314,
+      "step": 6263
+    },
+    {
+      "epoch": 2.782145236508994,
+      "grad_norm": 0.3656656264231573,
+      "learning_rate": 2.565757350569233e-06,
+      "loss": 0.0241,
+      "step": 6264
+    },
+    {
+      "epoch": 2.7825893848545413,
+      "grad_norm": 0.43315585203736967,
+      "learning_rate": 2.5640642480605722e-06,
+      "loss": 0.0372,
+      "step": 6265
+    },
+    {
+      "epoch": 2.783033533200089,
+      "grad_norm": 0.6118620035354352,
+      "learning_rate": 2.5623715117144337e-06,
+      "loss": 0.0389,
+      "step": 6266
+    },
+    {
+      "epoch": 2.7834776815456364,
+      "grad_norm": 1.003007529298919,
+      "learning_rate": 2.5606791417852655e-06,
+      "loss": 0.0489,
+      "step": 6267
+    },
+    {
+      "epoch": 2.7839218298911836,
+      "grad_norm": 0.335632805390222,
+      "learning_rate": 2.558987138527461e-06,
+      "loss": 0.0298,
+      "step": 6268
+    },
+    {
+      "epoch": 2.784365978236731,
+      "grad_norm": 0.49218374523384917,
+      "learning_rate": 2.5572955021953525e-06,
+      "loss": 0.0504,
+      "step": 6269
+    },
+    {
+      "epoch": 2.7848101265822782,
+      "grad_norm": 0.3662564117457857,
+      "learning_rate": 2.555604233043224e-06,
+      "loss": 0.024,
+      "step": 6270
+    },
+    {
+      "epoch": 2.785254274927826,
+      "grad_norm": 0.4410048606518427,
+      "learning_rate": 2.553913331325305e-06,
+      "loss": 0.0308,
+      "step": 6271
+    },
+    {
+      "epoch": 2.7856984232733732,
+      "grad_norm": 0.6748277215653389,
+      "learning_rate": 2.5522227972957626e-06,
+      "loss": 0.0353,
+      "step": 6272
+    },
+    {
+      "epoch": 2.786142571618921,
+      "grad_norm": 0.4988731713313121,
+      "learning_rate": 2.550532631208713e-06,
+      "loss": 0.0262,
+      "step": 6273
+    },
+    {
+      "epoch": 2.7865867199644683,
+      "grad_norm": 0.3910346837842304,
+      "learning_rate": 2.5488428333182213e-06,
+      "loss": 0.0369,
+      "step": 6274
+    },
+    {
+      "epoch": 2.7870308683100156,
+      "grad_norm": 0.5189622685954451,
+      "learning_rate": 2.5471534038782876e-06,
+      "loss": 0.0365,
+      "step": 6275
+    },
+    {
+      "epoch": 2.787475016655563,
+      "grad_norm": 0.4108944491644082,
+      "learning_rate": 2.545464343142862e-06,
+      "loss": 0.0322,
+      "step": 6276
+    },
+    {
+      "epoch": 2.78791916500111,
+      "grad_norm": 0.3860699292761844,
+      "learning_rate": 2.543775651365844e-06,
+      "loss": 0.0232,
+      "step": 6277
+    },
+    {
+      "epoch": 2.788363313346658,
+      "grad_norm": 0.47557488338460535,
+      "learning_rate": 2.5420873288010682e-06,
+      "loss": 0.0387,
+      "step": 6278
+    },
+    {
+      "epoch": 2.788807461692205,
+      "grad_norm": 0.3839192710393702,
+      "learning_rate": 2.5403993757023193e-06,
+      "loss": 0.0261,
+      "step": 6279
+    },
+    {
+      "epoch": 2.7892516100377525,
+      "grad_norm": 0.4089956846684774,
+      "learning_rate": 2.538711792323328e-06,
+      "loss": 0.0393,
+      "step": 6280
+    },
+    {
+      "epoch": 2.7896957583833,
+      "grad_norm": 0.3534578778564665,
+      "learning_rate": 2.5370245789177615e-06,
+      "loss": 0.0234,
+      "step": 6281
+    },
+    {
+      "epoch": 2.7901399067288475,
+      "grad_norm": 0.38317967580423545,
+      "learning_rate": 2.53533773573924e-06,
+      "loss": 0.0264,
+      "step": 6282
+    },
+    {
+      "epoch": 2.7905840550743948,
+      "grad_norm": 0.42936934153083767,
+      "learning_rate": 2.533651263041324e-06,
+      "loss": 0.0393,
+      "step": 6283
+    },
+    {
+      "epoch": 2.791028203419942,
+      "grad_norm": 0.554120368450601,
+      "learning_rate": 2.5319651610775194e-06,
+      "loss": 0.0317,
+      "step": 6284
+    },
+    {
+      "epoch": 2.79147235176549,
+      "grad_norm": 0.4081200054049489,
+      "learning_rate": 2.5302794301012766e-06,
+      "loss": 0.0297,
+      "step": 6285
+    },
+    {
+      "epoch": 2.791916500111037,
+      "grad_norm": 0.7070187464980049,
+      "learning_rate": 2.528594070365988e-06,
+      "loss": 0.0356,
+      "step": 6286
+    },
+    {
+      "epoch": 2.7923606484565844,
+      "grad_norm": 0.45395646778800836,
+      "learning_rate": 2.52690908212499e-06,
+      "loss": 0.0344,
+      "step": 6287
+    },
+    {
+      "epoch": 2.792804796802132,
+      "grad_norm": 0.7713383991827947,
+      "learning_rate": 2.525224465631571e-06,
+      "loss": 0.0323,
+      "step": 6288
+    },
+    {
+      "epoch": 2.7932489451476794,
+      "grad_norm": 0.43262436851175007,
+      "learning_rate": 2.5235402211389525e-06,
+      "loss": 0.0311,
+      "step": 6289
+    },
+    {
+      "epoch": 2.7936930934932267,
+      "grad_norm": 0.9524822174469669,
+      "learning_rate": 2.5218563489003062e-06,
+      "loss": 0.0307,
+      "step": 6290
+    },
+    {
+      "epoch": 2.794137241838774,
+      "grad_norm": 0.35257313740001833,
+      "learning_rate": 2.520172849168749e-06,
+      "loss": 0.0321,
+      "step": 6291
+    },
+    {
+      "epoch": 2.7945813901843217,
+      "grad_norm": 0.4825645734393702,
+      "learning_rate": 2.518489722197335e-06,
+      "loss": 0.0374,
+      "step": 6292
+    },
+    {
+      "epoch": 2.795025538529869,
+      "grad_norm": 0.7297573078351898,
+      "learning_rate": 2.51680696823907e-06,
+      "loss": 0.041,
+      "step": 6293
+    },
+    {
+      "epoch": 2.7954696868754163,
+      "grad_norm": 0.4555687972584318,
+      "learning_rate": 2.5151245875468993e-06,
+      "loss": 0.0386,
+      "step": 6294
+    },
+    {
+      "epoch": 2.795913835220964,
+      "grad_norm": 0.4970797678676715,
+      "learning_rate": 2.5134425803737137e-06,
+      "loss": 0.0418,
+      "step": 6295
+    },
+    {
+      "epoch": 2.7963579835665113,
+      "grad_norm": 0.6816809012387476,
+      "learning_rate": 2.511760946972348e-06,
+      "loss": 0.0409,
+      "step": 6296
+    },
+    {
+      "epoch": 2.7968021319120586,
+      "grad_norm": 0.5264337163942395,
+      "learning_rate": 2.5100796875955815e-06,
+      "loss": 0.0473,
+      "step": 6297
+    },
+    {
+      "epoch": 2.797246280257606,
+      "grad_norm": 0.5625709988931173,
+      "learning_rate": 2.508398802496132e-06,
+      "loss": 0.0372,
+      "step": 6298
+    },
+    {
+      "epoch": 2.797690428603153,
+      "grad_norm": 0.5237928838120754,
+      "learning_rate": 2.5067182919266676e-06,
+      "loss": 0.0311,
+      "step": 6299
+    },
+    {
+      "epoch": 2.798134576948701,
+      "grad_norm": 0.4778467003244999,
+      "learning_rate": 2.5050381561397974e-06,
+      "loss": 0.0429,
+      "step": 6300
+    },
+    {
+      "epoch": 2.7985787252942482,
+      "grad_norm": 0.43069407371279944,
+      "learning_rate": 2.503358395388074e-06,
+      "loss": 0.0435,
+      "step": 6301
+    },
+    {
+      "epoch": 2.799022873639796,
+      "grad_norm": 0.35842408526989356,
+      "learning_rate": 2.501679009923997e-06,
+      "loss": 0.0279,
+      "step": 6302
+    },
+    {
+      "epoch": 2.7994670219853433,
+      "grad_norm": 0.3642207986414275,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0311,
+      "step": 6303
+    },
+    {
+      "epoch": 2.7999111703308905,
+      "grad_norm": 0.655911709523861,
+      "learning_rate": 2.498321365868471e-06,
+      "loss": 0.036,
+      "step": 6304
+    },
+    {
+      "epoch": 2.800355318676438,
+      "grad_norm": 0.3959150499907849,
+      "learning_rate": 2.49664310778174e-06,
+      "loss": 0.0253,
+      "step": 6305
+    },
+    {
+      "epoch": 2.800799467021985,
+      "grad_norm": 0.5170912208789272,
+      "learning_rate": 2.4949652259920727e-06,
+      "loss": 0.0279,
+      "step": 6306
+    },
+    {
+      "epoch": 2.801243615367533,
+      "grad_norm": 0.4192727714282743,
+      "learning_rate": 2.4932877207516844e-06,
+      "loss": 0.0288,
+      "step": 6307
+    },
+    {
+      "epoch": 2.80168776371308,
+      "grad_norm": 0.4736817072181479,
+      "learning_rate": 2.4916105923127355e-06,
+      "loss": 0.0395,
+      "step": 6308
+    },
+    {
+      "epoch": 2.8021319120586274,
+      "grad_norm": 0.4164208273038199,
+      "learning_rate": 2.489933840927323e-06,
+      "loss": 0.025,
+      "step": 6309
+    },
+    {
+      "epoch": 2.802576060404175,
+      "grad_norm": 0.4237935178364351,
+      "learning_rate": 2.4882574668474925e-06,
+      "loss": 0.0326,
+      "step": 6310
+    },
+    {
+      "epoch": 2.8030202087497225,
+      "grad_norm": 0.3766334421419762,
+      "learning_rate": 2.486581470325232e-06,
+      "loss": 0.0228,
+      "step": 6311
+    },
+    {
+      "epoch": 2.8034643570952698,
+      "grad_norm": 0.41806136620537815,
+      "learning_rate": 2.484905851612471e-06,
+      "loss": 0.0364,
+      "step": 6312
+    },
+    {
+      "epoch": 2.803908505440817,
+      "grad_norm": 0.5567970380481517,
+      "learning_rate": 2.4832306109610877e-06,
+      "loss": 0.0575,
+      "step": 6313
+    },
+    {
+      "epoch": 2.804352653786365,
+      "grad_norm": 0.38489435628632473,
+      "learning_rate": 2.4815557486228937e-06,
+      "loss": 0.0243,
+      "step": 6314
+    },
+    {
+      "epoch": 2.804796802131912,
+      "grad_norm": 0.33130559466084525,
+      "learning_rate": 2.479881264849651e-06,
+      "loss": 0.0237,
+      "step": 6315
+    },
+    {
+      "epoch": 2.8052409504774594,
+      "grad_norm": 1.680243538004499,
+      "learning_rate": 2.478207159893064e-06,
+      "loss": 0.032,
+      "step": 6316
+    },
+    {
+      "epoch": 2.805685098823007,
+      "grad_norm": 0.48052932859070213,
+      "learning_rate": 2.476533434004779e-06,
+      "loss": 0.0311,
+      "step": 6317
+    },
+    {
+      "epoch": 2.8061292471685544,
+      "grad_norm": 0.36896833958182274,
+      "learning_rate": 2.474860087436384e-06,
+      "loss": 0.0261,
+      "step": 6318
+    },
+    {
+      "epoch": 2.8065733955141017,
+      "grad_norm": 0.633026238867653,
+      "learning_rate": 2.4731871204394155e-06,
+      "loss": 0.048,
+      "step": 6319
+    },
+    {
+      "epoch": 2.807017543859649,
+      "grad_norm": 0.4418796227317967,
+      "learning_rate": 2.4715145332653433e-06,
+      "loss": 0.0359,
+      "step": 6320
+    },
+    {
+      "epoch": 2.8074616922051967,
+      "grad_norm": 0.4502178155781612,
+      "learning_rate": 2.4698423261655887e-06,
+      "loss": 0.0306,
+      "step": 6321
+    },
+    {
+      "epoch": 2.807905840550744,
+      "grad_norm": 0.4566900893012358,
+      "learning_rate": 2.468170499391512e-06,
+      "loss": 0.0362,
+      "step": 6322
+    },
+    {
+      "epoch": 2.8083499888962913,
+      "grad_norm": 0.38715597816758157,
+      "learning_rate": 2.4664990531944176e-06,
+      "loss": 0.0404,
+      "step": 6323
+    },
+    {
+      "epoch": 2.808794137241839,
+      "grad_norm": 0.43966470207668296,
+      "learning_rate": 2.4648279878255523e-06,
+      "loss": 0.0258,
+      "step": 6324
+    },
+    {
+      "epoch": 2.8092382855873863,
+      "grad_norm": 0.3596530155746306,
+      "learning_rate": 2.4631573035361073e-06,
+      "loss": 0.0294,
+      "step": 6325
+    },
+    {
+      "epoch": 2.8096824339329336,
+      "grad_norm": 0.44583220946986085,
+      "learning_rate": 2.4614870005772105e-06,
+      "loss": 0.0223,
+      "step": 6326
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.5911442996714825,
+      "learning_rate": 2.45981707919994e-06,
+      "loss": 0.0393,
+      "step": 6327
+    },
+    {
+      "epoch": 2.810570730624028,
+      "grad_norm": 0.5581799888907024,
+      "learning_rate": 2.458147539655313e-06,
+      "loss": 0.0326,
+      "step": 6328
+    },
+    {
+      "epoch": 2.811014878969576,
+      "grad_norm": 0.39311733798131593,
+      "learning_rate": 2.4564783821942884e-06,
+      "loss": 0.029,
+      "step": 6329
+    },
+    {
+      "epoch": 2.811459027315123,
+      "grad_norm": 0.4450796457627704,
+      "learning_rate": 2.454809607067772e-06,
+      "loss": 0.0369,
+      "step": 6330
+    },
+    {
+      "epoch": 2.811903175660671,
+      "grad_norm": 0.43590500853966985,
+      "learning_rate": 2.4531412145266055e-06,
+      "loss": 0.0386,
+      "step": 6331
+    },
+    {
+      "epoch": 2.8123473240062182,
+      "grad_norm": 0.457972822762452,
+      "learning_rate": 2.4514732048215774e-06,
+      "loss": 0.042,
+      "step": 6332
+    },
+    {
+      "epoch": 2.8127914723517655,
+      "grad_norm": 0.32284714510978635,
+      "learning_rate": 2.4498055782034187e-06,
+      "loss": 0.0334,
+      "step": 6333
+    },
+    {
+      "epoch": 2.813235620697313,
+      "grad_norm": 0.36527473928195653,
+      "learning_rate": 2.4481383349228016e-06,
+      "loss": 0.0253,
+      "step": 6334
+    },
+    {
+      "epoch": 2.81367976904286,
+      "grad_norm": 0.3860105760601358,
+      "learning_rate": 2.446471475230342e-06,
+      "loss": 0.0391,
+      "step": 6335
+    },
+    {
+      "epoch": 2.814123917388408,
+      "grad_norm": 0.41855117247115486,
+      "learning_rate": 2.4448049993765975e-06,
+      "loss": 0.0296,
+      "step": 6336
+    },
+    {
+      "epoch": 2.814568065733955,
+      "grad_norm": 0.4146118074577428,
+      "learning_rate": 2.4431389076120657e-06,
+      "loss": 0.0325,
+      "step": 6337
+    },
+    {
+      "epoch": 2.8150122140795024,
+      "grad_norm": 0.3990029724898874,
+      "learning_rate": 2.4414732001871892e-06,
+      "loss": 0.0337,
+      "step": 6338
+    },
+    {
+      "epoch": 2.81545636242505,
+      "grad_norm": 0.5638606230897282,
+      "learning_rate": 2.4398078773523526e-06,
+      "loss": 0.0445,
+      "step": 6339
+    },
+    {
+      "epoch": 2.8159005107705974,
+      "grad_norm": 0.45902483366087526,
+      "learning_rate": 2.438142939357882e-06,
+      "loss": 0.0417,
+      "step": 6340
+    },
+    {
+      "epoch": 2.8163446591161447,
+      "grad_norm": 0.3183617092675609,
+      "learning_rate": 2.4364783864540482e-06,
+      "loss": 0.0252,
+      "step": 6341
+    },
+    {
+      "epoch": 2.816788807461692,
+      "grad_norm": 0.3554967849352285,
+      "learning_rate": 2.434814218891057e-06,
+      "loss": 0.0253,
+      "step": 6342
+    },
+    {
+      "epoch": 2.8172329558072398,
+      "grad_norm": 0.39321716328224376,
+      "learning_rate": 2.433150436919064e-06,
+      "loss": 0.0272,
+      "step": 6343
+    },
+    {
+      "epoch": 2.817677104152787,
+      "grad_norm": 0.5605795416471407,
+      "learning_rate": 2.4314870407881637e-06,
+      "loss": 0.0348,
+      "step": 6344
+    },
+    {
+      "epoch": 2.8181212524983343,
+      "grad_norm": 0.3732190806583137,
+      "learning_rate": 2.4298240307483923e-06,
+      "loss": 0.0233,
+      "step": 6345
+    },
+    {
+      "epoch": 2.818565400843882,
+      "grad_norm": 0.32751466948341074,
+      "learning_rate": 2.4281614070497282e-06,
+      "loss": 0.0284,
+      "step": 6346
+    },
+    {
+      "epoch": 2.8190095491894294,
+      "grad_norm": 0.38489431013084585,
+      "learning_rate": 2.4264991699420953e-06,
+      "loss": 0.0235,
+      "step": 6347
+    },
+    {
+      "epoch": 2.8194536975349767,
+      "grad_norm": 0.5207677415942126,
+      "learning_rate": 2.4248373196753512e-06,
+      "loss": 0.0362,
+      "step": 6348
+    },
+    {
+      "epoch": 2.819897845880524,
+      "grad_norm": 0.39019062549891126,
+      "learning_rate": 2.423175856499302e-06,
+      "loss": 0.0221,
+      "step": 6349
+    },
+    {
+      "epoch": 2.8203419942260712,
+      "grad_norm": 0.4152809637919385,
+      "learning_rate": 2.4215147806636942e-06,
+      "loss": 0.0314,
+      "step": 6350
+    },
+    {
+      "epoch": 2.820786142571619,
+      "grad_norm": 0.4000109524825699,
+      "learning_rate": 2.4198540924182156e-06,
+      "loss": 0.0307,
+      "step": 6351
+    },
+    {
+      "epoch": 2.8212302909171663,
+      "grad_norm": 0.4388997785747502,
+      "learning_rate": 2.4181937920124966e-06,
+      "loss": 0.0298,
+      "step": 6352
+    },
+    {
+      "epoch": 2.821674439262714,
+      "grad_norm": 0.4019919408619044,
+      "learning_rate": 2.4165338796961093e-06,
+      "loss": 0.033,
+      "step": 6353
+    },
+    {
+      "epoch": 2.8221185876082613,
+      "grad_norm": 0.570406566396068,
+      "learning_rate": 2.414874355718563e-06,
+      "loss": 0.0379,
+      "step": 6354
+    },
+    {
+      "epoch": 2.8225627359538086,
+      "grad_norm": 0.4407385335769093,
+      "learning_rate": 2.413215220329315e-06,
+      "loss": 0.0282,
+      "step": 6355
+    },
+    {
+      "epoch": 2.823006884299356,
+      "grad_norm": 0.38569267409608843,
+      "learning_rate": 2.411556473777761e-06,
+      "loss": 0.0282,
+      "step": 6356
+    },
+    {
+      "epoch": 2.823451032644903,
+      "grad_norm": 0.4606495839160357,
+      "learning_rate": 2.4098981163132395e-06,
+      "loss": 0.0342,
+      "step": 6357
+    },
+    {
+      "epoch": 2.823895180990451,
+      "grad_norm": 0.43263213722768684,
+      "learning_rate": 2.4082401481850306e-06,
+      "loss": 0.0369,
+      "step": 6358
+    },
+    {
+      "epoch": 2.824339329335998,
+      "grad_norm": 0.5551215018246358,
+      "learning_rate": 2.4065825696423522e-06,
+      "loss": 0.0327,
+      "step": 6359
+    },
+    {
+      "epoch": 2.8247834776815455,
+      "grad_norm": 0.3806554615954976,
+      "learning_rate": 2.4049253809343678e-06,
+      "loss": 0.0333,
+      "step": 6360
+    },
+    {
+      "epoch": 2.825227626027093,
+      "grad_norm": 0.4594230517569721,
+      "learning_rate": 2.4032685823101814e-06,
+      "loss": 0.0302,
+      "step": 6361
+    },
+    {
+      "epoch": 2.8256717743726405,
+      "grad_norm": 0.49282566341591966,
+      "learning_rate": 2.4016121740188375e-06,
+      "loss": 0.0283,
+      "step": 6362
+    },
+    {
+      "epoch": 2.826115922718188,
+      "grad_norm": 0.37575563744052526,
+      "learning_rate": 2.3999561563093234e-06,
+      "loss": 0.0349,
+      "step": 6363
+    },
+    {
+      "epoch": 2.826560071063735,
+      "grad_norm": 0.4760513100430366,
+      "learning_rate": 2.3983005294305673e-06,
+      "loss": 0.0374,
+      "step": 6364
+    },
+    {
+      "epoch": 2.827004219409283,
+      "grad_norm": 0.395278083467666,
+      "learning_rate": 2.396645293631435e-06,
+      "loss": 0.0248,
+      "step": 6365
+    },
+    {
+      "epoch": 2.82744836775483,
+      "grad_norm": 0.4876360922819094,
+      "learning_rate": 2.3949904491607384e-06,
+      "loss": 0.0342,
+      "step": 6366
+    },
+    {
+      "epoch": 2.8278925161003774,
+      "grad_norm": 0.40885930978989254,
+      "learning_rate": 2.393335996267229e-06,
+      "loss": 0.0325,
+      "step": 6367
+    },
+    {
+      "epoch": 2.828336664445925,
+      "grad_norm": 0.44935957035165386,
+      "learning_rate": 2.3916819351995984e-06,
+      "loss": 0.0273,
+      "step": 6368
+    },
+    {
+      "epoch": 2.8287808127914724,
+      "grad_norm": 1.506178960229344,
+      "learning_rate": 2.3900282662064806e-06,
+      "loss": 0.059,
+      "step": 6369
+    },
+    {
+      "epoch": 2.8292249611370197,
+      "grad_norm": 0.30959880923138633,
+      "learning_rate": 2.3883749895364523e-06,
+      "loss": 0.0217,
+      "step": 6370
+    },
+    {
+      "epoch": 2.829669109482567,
+      "grad_norm": 0.7823253783794836,
+      "learning_rate": 2.3867221054380244e-06,
+      "loss": 0.0329,
+      "step": 6371
+    },
+    {
+      "epoch": 2.8301132578281147,
+      "grad_norm": 0.4806013956209637,
+      "learning_rate": 2.3850696141596563e-06,
+      "loss": 0.0455,
+      "step": 6372
+    },
+    {
+      "epoch": 2.830557406173662,
+      "grad_norm": 0.6896449670646502,
+      "learning_rate": 2.3834175159497446e-06,
+      "loss": 0.0562,
+      "step": 6373
+    },
+    {
+      "epoch": 2.8310015545192093,
+      "grad_norm": 0.4709885726846783,
+      "learning_rate": 2.3817658110566288e-06,
+      "loss": 0.0326,
+      "step": 6374
+    },
+    {
+      "epoch": 2.831445702864757,
+      "grad_norm": 0.4631209021676538,
+      "learning_rate": 2.380114499728589e-06,
+      "loss": 0.0265,
+      "step": 6375
+    },
+    {
+      "epoch": 2.8318898512103043,
+      "grad_norm": 0.4360372773811446,
+      "learning_rate": 2.3784635822138424e-06,
+      "loss": 0.0409,
+      "step": 6376
+    },
+    {
+      "epoch": 2.8323339995558516,
+      "grad_norm": 0.36688129563252986,
+      "learning_rate": 2.3768130587605513e-06,
+      "loss": 0.0248,
+      "step": 6377
+    },
+    {
+      "epoch": 2.832778147901399,
+      "grad_norm": 0.37437289245252986,
+      "learning_rate": 2.3751629296168177e-06,
+      "loss": 0.0305,
+      "step": 6378
+    },
+    {
+      "epoch": 2.833222296246946,
+      "grad_norm": 0.531778233535108,
+      "learning_rate": 2.3735131950306845e-06,
+      "loss": 0.0352,
+      "step": 6379
+    },
+    {
+      "epoch": 2.833666444592494,
+      "grad_norm": 0.4416928778863043,
+      "learning_rate": 2.371863855250134e-06,
+      "loss": 0.0356,
+      "step": 6380
+    },
+    {
+      "epoch": 2.8341105929380412,
+      "grad_norm": 0.3539473400154312,
+      "learning_rate": 2.3702149105230914e-06,
+      "loss": 0.0329,
+      "step": 6381
+    },
+    {
+      "epoch": 2.834554741283589,
+      "grad_norm": 0.8474469374112829,
+      "learning_rate": 2.3685663610974193e-06,
+      "loss": 0.0339,
+      "step": 6382
+    },
+    {
+      "epoch": 2.8349988896291363,
+      "grad_norm": 0.42747029305832396,
+      "learning_rate": 2.3669182072209225e-06,
+      "loss": 0.0381,
+      "step": 6383
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.7608488814785124,
+      "learning_rate": 2.3652704491413477e-06,
+      "loss": 0.032,
+      "step": 6384
+    },
+    {
+      "epoch": 2.835887186320231,
+      "grad_norm": 0.388700958157769,
+      "learning_rate": 2.3636230871063803e-06,
+      "loss": 0.0273,
+      "step": 6385
+    },
+    {
+      "epoch": 2.836331334665778,
+      "grad_norm": 0.5022022192841024,
+      "learning_rate": 2.3619761213636496e-06,
+      "loss": 0.0437,
+      "step": 6386
+    },
+    {
+      "epoch": 2.836775483011326,
+      "grad_norm": 0.34657365126440165,
+      "learning_rate": 2.360329552160718e-06,
+      "loss": 0.0246,
+      "step": 6387
+    },
+    {
+      "epoch": 2.837219631356873,
+      "grad_norm": 0.3939450249882695,
+      "learning_rate": 2.358683379745094e-06,
+      "loss": 0.0267,
+      "step": 6388
+    },
+    {
+      "epoch": 2.8376637797024205,
+      "grad_norm": 0.41677776471298256,
+      "learning_rate": 2.357037604364229e-06,
+      "loss": 0.0324,
+      "step": 6389
+    },
+    {
+      "epoch": 2.838107928047968,
+      "grad_norm": 0.3562978131516035,
+      "learning_rate": 2.3553922262655045e-06,
+      "loss": 0.0289,
+      "step": 6390
+    },
+    {
+      "epoch": 2.8385520763935155,
+      "grad_norm": 0.3800097306965494,
+      "learning_rate": 2.3537472456962536e-06,
+      "loss": 0.0269,
+      "step": 6391
+    },
+    {
+      "epoch": 2.8389962247390628,
+      "grad_norm": 0.3927488750912727,
+      "learning_rate": 2.3521026629037456e-06,
+      "loss": 0.032,
+      "step": 6392
+    },
+    {
+      "epoch": 2.83944037308461,
+      "grad_norm": 0.6796389696750893,
+      "learning_rate": 2.3504584781351857e-06,
+      "loss": 0.035,
+      "step": 6393
+    },
+    {
+      "epoch": 2.839884521430158,
+      "grad_norm": 0.4312865368452842,
+      "learning_rate": 2.3488146916377246e-06,
+      "loss": 0.024,
+      "step": 6394
+    },
+    {
+      "epoch": 2.840328669775705,
+      "grad_norm": 0.5503469902463891,
+      "learning_rate": 2.3471713036584507e-06,
+      "loss": 0.0387,
+      "step": 6395
+    },
+    {
+      "epoch": 2.8407728181212524,
+      "grad_norm": 0.39537225808331916,
+      "learning_rate": 2.345528314444394e-06,
+      "loss": 0.0365,
+      "step": 6396
+    },
+    {
+      "epoch": 2.8412169664668,
+      "grad_norm": 0.4828036261450407,
+      "learning_rate": 2.343885724242523e-06,
+      "loss": 0.0359,
+      "step": 6397
+    },
+    {
+      "epoch": 2.8416611148123474,
+      "grad_norm": 0.3405572301380993,
+      "learning_rate": 2.342243533299749e-06,
+      "loss": 0.0296,
+      "step": 6398
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3471570803326242,
+      "learning_rate": 2.3406017418629173e-06,
+      "loss": 0.0295,
+      "step": 6399
+    },
+    {
+      "epoch": 2.842549411503442,
+      "grad_norm": 0.30534728377719983,
+      "learning_rate": 2.3389603501788187e-06,
+      "loss": 0.0161,
+      "step": 6400
+    },
+    {
+      "epoch": 2.8429935598489897,
+      "grad_norm": 0.4261866679259168,
+      "learning_rate": 2.3373193584941833e-06,
+      "loss": 0.0213,
+      "step": 6401
+    },
+    {
+      "epoch": 2.843437708194537,
+      "grad_norm": 0.5685316006056542,
+      "learning_rate": 2.335678767055679e-06,
+      "loss": 0.0434,
+      "step": 6402
+    },
+    {
+      "epoch": 2.8438818565400843,
+      "grad_norm": 0.49631869614829277,
+      "learning_rate": 2.334038576109917e-06,
+      "loss": 0.0356,
+      "step": 6403
+    },
+    {
+      "epoch": 2.844326004885632,
+      "grad_norm": 0.5515165392461051,
+      "learning_rate": 2.332398785903442e-06,
+      "loss": 0.0363,
+      "step": 6404
+    },
+    {
+      "epoch": 2.8447701532311793,
+      "grad_norm": 0.532355259631848,
+      "learning_rate": 2.330759396682744e-06,
+      "loss": 0.0399,
+      "step": 6405
+    },
+    {
+      "epoch": 2.8452143015767266,
+      "grad_norm": 0.5194908272702733,
+      "learning_rate": 2.329120408694253e-06,
+      "loss": 0.0419,
+      "step": 6406
+    },
+    {
+      "epoch": 2.845658449922274,
+      "grad_norm": 0.4352160286424336,
+      "learning_rate": 2.327481822184331e-06,
+      "loss": 0.0323,
+      "step": 6407
+    },
+    {
+      "epoch": 2.846102598267821,
+      "grad_norm": 0.40556462446399455,
+      "learning_rate": 2.3258436373992914e-06,
+      "loss": 0.0295,
+      "step": 6408
+    },
+    {
+      "epoch": 2.846546746613369,
+      "grad_norm": 0.48488189817066807,
+      "learning_rate": 2.3242058545853806e-06,
+      "loss": 0.0323,
+      "step": 6409
+    },
+    {
+      "epoch": 2.846990894958916,
+      "grad_norm": 0.7693171033124235,
+      "learning_rate": 2.322568473988782e-06,
+      "loss": 0.0304,
+      "step": 6410
+    },
+    {
+      "epoch": 2.847435043304464,
+      "grad_norm": 0.6057507153145041,
+      "learning_rate": 2.3209314958556232e-06,
+      "loss": 0.0262,
+      "step": 6411
+    },
+    {
+      "epoch": 2.8478791916500112,
+      "grad_norm": 0.4479104368079513,
+      "learning_rate": 2.319294920431972e-06,
+      "loss": 0.0363,
+      "step": 6412
+    },
+    {
+      "epoch": 2.8483233399955585,
+      "grad_norm": 0.5011576633143359,
+      "learning_rate": 2.317658747963828e-06,
+      "loss": 0.0415,
+      "step": 6413
+    },
+    {
+      "epoch": 2.848767488341106,
+      "grad_norm": 0.43395301260402036,
+      "learning_rate": 2.316022978697143e-06,
+      "loss": 0.0305,
+      "step": 6414
+    },
+    {
+      "epoch": 2.849211636686653,
+      "grad_norm": 0.4161396554904535,
+      "learning_rate": 2.314387612877795e-06,
+      "loss": 0.031,
+      "step": 6415
+    },
+    {
+      "epoch": 2.849655785032201,
+      "grad_norm": 0.453771160365058,
+      "learning_rate": 2.312752650751609e-06,
+      "loss": 0.0328,
+      "step": 6416
+    },
+    {
+      "epoch": 2.850099933377748,
+      "grad_norm": 0.4057605385720707,
+      "learning_rate": 2.3111180925643477e-06,
+      "loss": 0.0308,
+      "step": 6417
+    },
+    {
+      "epoch": 2.8505440817232954,
+      "grad_norm": 0.4469537485385078,
+      "learning_rate": 2.309483938561714e-06,
+      "loss": 0.0345,
+      "step": 6418
+    },
+    {
+      "epoch": 2.850988230068843,
+      "grad_norm": 0.3300098185481985,
+      "learning_rate": 2.3078501889893477e-06,
+      "loss": 0.028,
+      "step": 6419
+    },
+    {
+      "epoch": 2.8514323784143905,
+      "grad_norm": 0.36465566926853343,
+      "learning_rate": 2.3062168440928324e-06,
+      "loss": 0.0299,
+      "step": 6420
+    },
+    {
+      "epoch": 2.8518765267599377,
+      "grad_norm": 0.39053368090966234,
+      "learning_rate": 2.304583904117682e-06,
+      "loss": 0.029,
+      "step": 6421
+    },
+    {
+      "epoch": 2.852320675105485,
+      "grad_norm": 0.3264096837914728,
+      "learning_rate": 2.302951369309358e-06,
+      "loss": 0.0245,
+      "step": 6422
+    },
+    {
+      "epoch": 2.8527648234510328,
+      "grad_norm": 0.4328229716057836,
+      "learning_rate": 2.30131923991326e-06,
+      "loss": 0.0383,
+      "step": 6423
+    },
+    {
+      "epoch": 2.85320897179658,
+      "grad_norm": 0.4077375134882527,
+      "learning_rate": 2.2996875161747194e-06,
+      "loss": 0.038,
+      "step": 6424
+    },
+    {
+      "epoch": 2.8536531201421274,
+      "grad_norm": 0.4094918043040497,
+      "learning_rate": 2.298056198339017e-06,
+      "loss": 0.0298,
+      "step": 6425
+    },
+    {
+      "epoch": 2.854097268487675,
+      "grad_norm": 0.32420798744454565,
+      "learning_rate": 2.296425286651368e-06,
+      "loss": 0.0335,
+      "step": 6426
+    },
+    {
+      "epoch": 2.8545414168332224,
+      "grad_norm": 0.4249764666477243,
+      "learning_rate": 2.294794781356922e-06,
+      "loss": 0.0373,
+      "step": 6427
+    },
+    {
+      "epoch": 2.8549855651787697,
+      "grad_norm": 0.8704460304111548,
+      "learning_rate": 2.293164682700774e-06,
+      "loss": 0.0456,
+      "step": 6428
+    },
+    {
+      "epoch": 2.855429713524317,
+      "grad_norm": 0.5067415787866943,
+      "learning_rate": 2.2915349909279573e-06,
+      "loss": 0.0348,
+      "step": 6429
+    },
+    {
+      "epoch": 2.8558738618698647,
+      "grad_norm": 0.38730263257871017,
+      "learning_rate": 2.2899057062834363e-06,
+      "loss": 0.0508,
+      "step": 6430
+    },
+    {
+      "epoch": 2.856318010215412,
+      "grad_norm": 0.426646849291928,
+      "learning_rate": 2.2882768290121277e-06,
+      "loss": 0.03,
+      "step": 6431
+    },
+    {
+      "epoch": 2.8567621585609593,
+      "grad_norm": 0.4433931975798466,
+      "learning_rate": 2.286648359358874e-06,
+      "loss": 0.0361,
+      "step": 6432
+    },
+    {
+      "epoch": 2.857206306906507,
+      "grad_norm": 0.41545375813689084,
+      "learning_rate": 2.2850202975684637e-06,
+      "loss": 0.0302,
+      "step": 6433
+    },
+    {
+      "epoch": 2.8576504552520543,
+      "grad_norm": 0.3777035808530584,
+      "learning_rate": 2.283392643885624e-06,
+      "loss": 0.0287,
+      "step": 6434
+    },
+    {
+      "epoch": 2.8580946035976016,
+      "grad_norm": 0.7240655817753611,
+      "learning_rate": 2.2817653985550132e-06,
+      "loss": 0.0336,
+      "step": 6435
+    },
+    {
+      "epoch": 2.858538751943149,
+      "grad_norm": 0.3261261166151642,
+      "learning_rate": 2.2801385618212395e-06,
+      "loss": 0.0244,
+      "step": 6436
+    },
+    {
+      "epoch": 2.858982900288696,
+      "grad_norm": 0.365601220986991,
+      "learning_rate": 2.2785121339288446e-06,
+      "loss": 0.0281,
+      "step": 6437
+    },
+    {
+      "epoch": 2.859427048634244,
+      "grad_norm": 0.4479492062462993,
+      "learning_rate": 2.276886115122304e-06,
+      "loss": 0.0345,
+      "step": 6438
+    },
+    {
+      "epoch": 2.859871196979791,
+      "grad_norm": 0.3824493182929994,
+      "learning_rate": 2.2752605056460374e-06,
+      "loss": 0.0263,
+      "step": 6439
+    },
+    {
+      "epoch": 2.860315345325339,
+      "grad_norm": 0.444006507921262,
+      "learning_rate": 2.2736353057444045e-06,
+      "loss": 0.0286,
+      "step": 6440
+    },
+    {
+      "epoch": 2.8607594936708862,
+      "grad_norm": 0.4135565678132424,
+      "learning_rate": 2.272010515661694e-06,
+      "loss": 0.031,
+      "step": 6441
+    },
+    {
+      "epoch": 2.8612036420164335,
+      "grad_norm": 0.4704071890757161,
+      "learning_rate": 2.2703861356421476e-06,
+      "loss": 0.0374,
+      "step": 6442
+    },
+    {
+      "epoch": 2.861647790361981,
+      "grad_norm": 0.5654612362616459,
+      "learning_rate": 2.268762165929931e-06,
+      "loss": 0.0298,
+      "step": 6443
+    },
+    {
+      "epoch": 2.862091938707528,
+      "grad_norm": 0.4525755398497248,
+      "learning_rate": 2.267138606769156e-06,
+      "loss": 0.0373,
+      "step": 6444
+    },
+    {
+      "epoch": 2.862536087053076,
+      "grad_norm": 0.3792079187603561,
+      "learning_rate": 2.2655154584038718e-06,
+      "loss": 0.0274,
+      "step": 6445
+    },
+    {
+      "epoch": 2.862980235398623,
+      "grad_norm": 0.6472477866875229,
+      "learning_rate": 2.263892721078067e-06,
+      "loss": 0.0306,
+      "step": 6446
+    },
+    {
+      "epoch": 2.8634243837441704,
+      "grad_norm": 0.4724763504374727,
+      "learning_rate": 2.2622703950356607e-06,
+      "loss": 0.0395,
+      "step": 6447
+    },
+    {
+      "epoch": 2.863868532089718,
+      "grad_norm": 0.5178919025923505,
+      "learning_rate": 2.2606484805205235e-06,
+      "loss": 0.0329,
+      "step": 6448
+    },
+    {
+      "epoch": 2.8643126804352654,
+      "grad_norm": 0.4061558909889993,
+      "learning_rate": 2.2590269777764516e-06,
+      "loss": 0.0299,
+      "step": 6449
+    },
+    {
+      "epoch": 2.8647568287808127,
+      "grad_norm": 0.39901609043961084,
+      "learning_rate": 2.257405887047186e-06,
+      "loss": 0.0328,
+      "step": 6450
+    },
+    {
+      "epoch": 2.86520097712636,
+      "grad_norm": 0.49608870962855456,
+      "learning_rate": 2.2557852085764053e-06,
+      "loss": 0.0303,
+      "step": 6451
+    },
+    {
+      "epoch": 2.8656451254719078,
+      "grad_norm": 0.38259419645750875,
+      "learning_rate": 2.254164942607721e-06,
+      "loss": 0.0219,
+      "step": 6452
+    },
+    {
+      "epoch": 2.866089273817455,
+      "grad_norm": 0.42520263669940206,
+      "learning_rate": 2.2525450893846906e-06,
+      "loss": 0.0248,
+      "step": 6453
+    },
+    {
+      "epoch": 2.8665334221630023,
+      "grad_norm": 0.343086064868294,
+      "learning_rate": 2.2509256491508063e-06,
+      "loss": 0.0226,
+      "step": 6454
+    },
+    {
+      "epoch": 2.86697757050855,
+      "grad_norm": 0.48422523578882104,
+      "learning_rate": 2.249306622149494e-06,
+      "loss": 0.0336,
+      "step": 6455
+    },
+    {
+      "epoch": 2.8674217188540974,
+      "grad_norm": 0.40460550415553925,
+      "learning_rate": 2.2476880086241225e-06,
+      "loss": 0.0325,
+      "step": 6456
+    },
+    {
+      "epoch": 2.8678658671996446,
+      "grad_norm": 0.38849375983156137,
+      "learning_rate": 2.2460698088179985e-06,
+      "loss": 0.034,
+      "step": 6457
+    },
+    {
+      "epoch": 2.868310015545192,
+      "grad_norm": 0.3659863472898491,
+      "learning_rate": 2.24445202297436e-06,
+      "loss": 0.0311,
+      "step": 6458
+    },
+    {
+      "epoch": 2.8687541638907397,
+      "grad_norm": 0.32032734176784455,
+      "learning_rate": 2.242834651336394e-06,
+      "loss": 0.0304,
+      "step": 6459
+    },
+    {
+      "epoch": 2.869198312236287,
+      "grad_norm": 0.2895901124804986,
+      "learning_rate": 2.2412176941472146e-06,
+      "loss": 0.0252,
+      "step": 6460
+    },
+    {
+      "epoch": 2.8696424605818343,
+      "grad_norm": 0.4544732858743473,
+      "learning_rate": 2.2396011516498794e-06,
+      "loss": 0.0378,
+      "step": 6461
+    },
+    {
+      "epoch": 2.870086608927382,
+      "grad_norm": 0.33869349353641764,
+      "learning_rate": 2.2379850240873836e-06,
+      "loss": 0.0243,
+      "step": 6462
+    },
+    {
+      "epoch": 2.8705307572729293,
+      "grad_norm": 0.39779581383793367,
+      "learning_rate": 2.2363693117026554e-06,
+      "loss": 0.0328,
+      "step": 6463
+    },
+    {
+      "epoch": 2.8709749056184766,
+      "grad_norm": 0.5359509284893266,
+      "learning_rate": 2.2347540147385636e-06,
+      "loss": 0.0276,
+      "step": 6464
+    },
+    {
+      "epoch": 2.871419053964024,
+      "grad_norm": 0.3901564758848709,
+      "learning_rate": 2.2331391334379205e-06,
+      "loss": 0.0286,
+      "step": 6465
+    },
+    {
+      "epoch": 2.871863202309571,
+      "grad_norm": 0.4664762506585298,
+      "learning_rate": 2.231524668043465e-06,
+      "loss": 0.0421,
+      "step": 6466
+    },
+    {
+      "epoch": 2.872307350655119,
+      "grad_norm": 0.4098655625284004,
+      "learning_rate": 2.229910618797879e-06,
+      "loss": 0.0376,
+      "step": 6467
+    },
+    {
+      "epoch": 2.872751499000666,
+      "grad_norm": 0.4583633277386958,
+      "learning_rate": 2.228296985943785e-06,
+      "loss": 0.0383,
+      "step": 6468
+    },
+    {
+      "epoch": 2.873195647346214,
+      "grad_norm": 0.37685547831341093,
+      "learning_rate": 2.226683769723734e-06,
+      "loss": 0.028,
+      "step": 6469
+    },
+    {
+      "epoch": 2.873639795691761,
+      "grad_norm": 0.400587688868656,
+      "learning_rate": 2.225070970380224e-06,
+      "loss": 0.0337,
+      "step": 6470
+    },
+    {
+      "epoch": 2.8740839440373085,
+      "grad_norm": 0.3702484950984152,
+      "learning_rate": 2.2234585881556864e-06,
+      "loss": 0.0274,
+      "step": 6471
+    },
+    {
+      "epoch": 2.874528092382856,
+      "grad_norm": 0.4674975669326644,
+      "learning_rate": 2.2218466232924867e-06,
+      "loss": 0.03,
+      "step": 6472
+    },
+    {
+      "epoch": 2.874972240728403,
+      "grad_norm": 0.37515826872200714,
+      "learning_rate": 2.2202350760329328e-06,
+      "loss": 0.0257,
+      "step": 6473
+    },
+    {
+      "epoch": 2.875416389073951,
+      "grad_norm": 0.3257808540366288,
+      "learning_rate": 2.2186239466192676e-06,
+      "loss": 0.0219,
+      "step": 6474
+    },
+    {
+      "epoch": 2.875860537419498,
+      "grad_norm": 0.4753925237563355,
+      "learning_rate": 2.2170132352936675e-06,
+      "loss": 0.0434,
+      "step": 6475
+    },
+    {
+      "epoch": 2.8763046857650454,
+      "grad_norm": 0.44581196717512095,
+      "learning_rate": 2.2154029422982563e-06,
+      "loss": 0.0371,
+      "step": 6476
+    },
+    {
+      "epoch": 2.876748834110593,
+      "grad_norm": 0.4263527151388236,
+      "learning_rate": 2.2137930678750835e-06,
+      "loss": 0.0376,
+      "step": 6477
+    },
+    {
+      "epoch": 2.8771929824561404,
+      "grad_norm": 0.46476168558248493,
+      "learning_rate": 2.2121836122661416e-06,
+      "loss": 0.034,
+      "step": 6478
+    },
+    {
+      "epoch": 2.8776371308016877,
+      "grad_norm": 0.34003139533262217,
+      "learning_rate": 2.2105745757133612e-06,
+      "loss": 0.0294,
+      "step": 6479
+    },
+    {
+      "epoch": 2.878081279147235,
+      "grad_norm": 0.49849907569693336,
+      "learning_rate": 2.2089659584586047e-06,
+      "loss": 0.0377,
+      "step": 6480
+    },
+    {
+      "epoch": 2.8785254274927827,
+      "grad_norm": 0.5382448752849822,
+      "learning_rate": 2.2073577607436737e-06,
+      "loss": 0.0443,
+      "step": 6481
+    },
+    {
+      "epoch": 2.87896957583833,
+      "grad_norm": 0.3879946256129708,
+      "learning_rate": 2.2057499828103142e-06,
+      "loss": 0.0327,
+      "step": 6482
+    },
+    {
+      "epoch": 2.8794137241838773,
+      "grad_norm": 0.5104922320743165,
+      "learning_rate": 2.2041426249001955e-06,
+      "loss": 0.0362,
+      "step": 6483
+    },
+    {
+      "epoch": 2.879857872529425,
+      "grad_norm": 0.4930647183156878,
+      "learning_rate": 2.2025356872549345e-06,
+      "loss": 0.0256,
+      "step": 6484
+    },
+    {
+      "epoch": 2.8803020208749723,
+      "grad_norm": 0.3966997553776206,
+      "learning_rate": 2.2009291701160817e-06,
+      "loss": 0.0271,
+      "step": 6485
+    },
+    {
+      "epoch": 2.8807461692205196,
+      "grad_norm": 0.4962925195644924,
+      "learning_rate": 2.1993230737251216e-06,
+      "loss": 0.0371,
+      "step": 6486
+    },
+    {
+      "epoch": 2.881190317566067,
+      "grad_norm": 0.42642379725740076,
+      "learning_rate": 2.197717398323477e-06,
+      "loss": 0.0263,
+      "step": 6487
+    },
+    {
+      "epoch": 2.881634465911614,
+      "grad_norm": 0.4141444359146802,
+      "learning_rate": 2.1961121441525113e-06,
+      "loss": 0.0344,
+      "step": 6488
+    },
+    {
+      "epoch": 2.882078614257162,
+      "grad_norm": 0.5311512207572496,
+      "learning_rate": 2.19450731145352e-06,
+      "loss": 0.0364,
+      "step": 6489
+    },
+    {
+      "epoch": 2.8825227626027092,
+      "grad_norm": 0.4267919708323801,
+      "learning_rate": 2.192902900467736e-06,
+      "loss": 0.0274,
+      "step": 6490
+    },
+    {
+      "epoch": 2.882966910948257,
+      "grad_norm": 0.31959536449480613,
+      "learning_rate": 2.1912989114363326e-06,
+      "loss": 0.0242,
+      "step": 6491
+    },
+    {
+      "epoch": 2.8834110592938043,
+      "grad_norm": 0.44718270123585074,
+      "learning_rate": 2.1896953446004104e-06,
+      "loss": 0.0397,
+      "step": 6492
+    },
+    {
+      "epoch": 2.8838552076393515,
+      "grad_norm": 0.4584663064416918,
+      "learning_rate": 2.1880922002010208e-06,
+      "loss": 0.0321,
+      "step": 6493
+    },
+    {
+      "epoch": 2.884299355984899,
+      "grad_norm": 0.6111455889547676,
+      "learning_rate": 2.186489478479137e-06,
+      "loss": 0.0423,
+      "step": 6494
+    },
+    {
+      "epoch": 2.884743504330446,
+      "grad_norm": 0.5635891360279379,
+      "learning_rate": 2.1848871796756784e-06,
+      "loss": 0.0559,
+      "step": 6495
+    },
+    {
+      "epoch": 2.885187652675994,
+      "grad_norm": 0.49224878038120795,
+      "learning_rate": 2.183285304031498e-06,
+      "loss": 0.0381,
+      "step": 6496
+    },
+    {
+      "epoch": 2.885631801021541,
+      "grad_norm": 0.412800271356595,
+      "learning_rate": 2.1816838517873834e-06,
+      "loss": 0.0281,
+      "step": 6497
+    },
+    {
+      "epoch": 2.8860759493670884,
+      "grad_norm": 0.42819522838748797,
+      "learning_rate": 2.1800828231840583e-06,
+      "loss": 0.0349,
+      "step": 6498
+    },
+    {
+      "epoch": 2.886520097712636,
+      "grad_norm": 0.3172148118997228,
+      "learning_rate": 2.178482218462191e-06,
+      "loss": 0.0198,
+      "step": 6499
+    },
+    {
+      "epoch": 2.8869642460581835,
+      "grad_norm": 0.6336802661233113,
+      "learning_rate": 2.176882037862373e-06,
+      "loss": 0.0527,
+      "step": 6500
+    },
+    {
+      "epoch": 2.8874083944037308,
+      "grad_norm": 0.6279273972202338,
+      "learning_rate": 2.1752822816251405e-06,
+      "loss": 0.0366,
+      "step": 6501
+    },
+    {
+      "epoch": 2.887852542749278,
+      "grad_norm": 0.6022931994329644,
+      "learning_rate": 2.173682949990968e-06,
+      "loss": 0.0306,
+      "step": 6502
+    },
+    {
+      "epoch": 2.888296691094826,
+      "grad_norm": 0.5362069754281357,
+      "learning_rate": 2.172084043200256e-06,
+      "loss": 0.0342,
+      "step": 6503
+    },
+    {
+      "epoch": 2.888740839440373,
+      "grad_norm": 0.7929642629941647,
+      "learning_rate": 2.17048556149335e-06,
+      "loss": 0.0333,
+      "step": 6504
+    },
+    {
+      "epoch": 2.8891849877859204,
+      "grad_norm": 0.33575039821832753,
+      "learning_rate": 2.16888750511053e-06,
+      "loss": 0.0242,
+      "step": 6505
+    },
+    {
+      "epoch": 2.889629136131468,
+      "grad_norm": 0.3741108552741579,
+      "learning_rate": 2.1672898742920094e-06,
+      "loss": 0.0273,
+      "step": 6506
+    },
+    {
+      "epoch": 2.8900732844770154,
+      "grad_norm": 0.3774493813914148,
+      "learning_rate": 2.1656926692779423e-06,
+      "loss": 0.0313,
+      "step": 6507
+    },
+    {
+      "epoch": 2.8905174328225627,
+      "grad_norm": 0.9718206954815033,
+      "learning_rate": 2.1640958903084118e-06,
+      "loss": 0.0401,
+      "step": 6508
+    },
+    {
+      "epoch": 2.89096158116811,
+      "grad_norm": 0.3694310192990195,
+      "learning_rate": 2.1624995376234403e-06,
+      "loss": 0.0348,
+      "step": 6509
+    },
+    {
+      "epoch": 2.8914057295136577,
+      "grad_norm": 0.6520572022397477,
+      "learning_rate": 2.1609036114629933e-06,
+      "loss": 0.0402,
+      "step": 6510
+    },
+    {
+      "epoch": 2.891849877859205,
+      "grad_norm": 0.4215772661238576,
+      "learning_rate": 2.159308112066959e-06,
+      "loss": 0.0293,
+      "step": 6511
+    },
+    {
+      "epoch": 2.8922940262047523,
+      "grad_norm": 0.38467024272056793,
+      "learning_rate": 2.1577130396751705e-06,
+      "loss": 0.0325,
+      "step": 6512
+    },
+    {
+      "epoch": 2.8927381745503,
+      "grad_norm": 0.38039523075207066,
+      "learning_rate": 2.1561183945273958e-06,
+      "loss": 0.029,
+      "step": 6513
+    },
+    {
+      "epoch": 2.8931823228958473,
+      "grad_norm": 0.35012157636808233,
+      "learning_rate": 2.154524176863334e-06,
+      "loss": 0.0298,
+      "step": 6514
+    },
+    {
+      "epoch": 2.8936264712413946,
+      "grad_norm": 0.32604790956444796,
+      "learning_rate": 2.1529303869226244e-06,
+      "loss": 0.0223,
+      "step": 6515
+    },
+    {
+      "epoch": 2.894070619586942,
+      "grad_norm": 0.6062038168527448,
+      "learning_rate": 2.151337024944841e-06,
+      "loss": 0.0328,
+      "step": 6516
+    },
+    {
+      "epoch": 2.894514767932489,
+      "grad_norm": 0.4616383089873512,
+      "learning_rate": 2.149744091169493e-06,
+      "loss": 0.0355,
+      "step": 6517
+    },
+    {
+      "epoch": 2.894958916278037,
+      "grad_norm": 0.31498629669392003,
+      "learning_rate": 2.1481515858360254e-06,
+      "loss": 0.0223,
+      "step": 6518
+    },
+    {
+      "epoch": 2.895403064623584,
+      "grad_norm": 0.3418552975452249,
+      "learning_rate": 2.1465595091838204e-06,
+      "loss": 0.0284,
+      "step": 6519
+    },
+    {
+      "epoch": 2.895847212969132,
+      "grad_norm": 0.6349150689262983,
+      "learning_rate": 2.144967861452191e-06,
+      "loss": 0.0288,
+      "step": 6520
+    },
+    {
+      "epoch": 2.8962913613146792,
+      "grad_norm": 0.3672718765769208,
+      "learning_rate": 2.143376642880391e-06,
+      "loss": 0.028,
+      "step": 6521
+    },
+    {
+      "epoch": 2.8967355096602265,
+      "grad_norm": 0.6329188736427485,
+      "learning_rate": 2.141785853707607e-06,
+      "loss": 0.0419,
+      "step": 6522
+    },
+    {
+      "epoch": 2.897179658005774,
+      "grad_norm": 0.35536621143471536,
+      "learning_rate": 2.1401954941729614e-06,
+      "loss": 0.0276,
+      "step": 6523
+    },
+    {
+      "epoch": 2.897623806351321,
+      "grad_norm": 0.4803884960432198,
+      "learning_rate": 2.1386055645155144e-06,
+      "loss": 0.0334,
+      "step": 6524
+    },
+    {
+      "epoch": 2.898067954696869,
+      "grad_norm": 0.7351089421454166,
+      "learning_rate": 2.137016064974256e-06,
+      "loss": 0.034,
+      "step": 6525
+    },
+    {
+      "epoch": 2.898512103042416,
+      "grad_norm": 0.35740753737071096,
+      "learning_rate": 2.135426995788115e-06,
+      "loss": 0.0255,
+      "step": 6526
+    },
+    {
+      "epoch": 2.8989562513879634,
+      "grad_norm": 0.40072320482835744,
+      "learning_rate": 2.133838357195961e-06,
+      "loss": 0.0369,
+      "step": 6527
+    },
+    {
+      "epoch": 2.899400399733511,
+      "grad_norm": 0.3671291054027193,
+      "learning_rate": 2.1322501494365873e-06,
+      "loss": 0.0248,
+      "step": 6528
+    },
+    {
+      "epoch": 2.8998445480790584,
+      "grad_norm": 0.46498735962046045,
+      "learning_rate": 2.1306623727487306e-06,
+      "loss": 0.0336,
+      "step": 6529
+    },
+    {
+      "epoch": 2.9002886964246057,
+      "grad_norm": 0.37582439553874314,
+      "learning_rate": 2.1290750273710625e-06,
+      "loss": 0.0311,
+      "step": 6530
+    },
+    {
+      "epoch": 2.900732844770153,
+      "grad_norm": 0.3769841744886795,
+      "learning_rate": 2.127488113542185e-06,
+      "loss": 0.0219,
+      "step": 6531
+    },
+    {
+      "epoch": 2.9011769931157008,
+      "grad_norm": 0.5408699565947783,
+      "learning_rate": 2.1259016315006388e-06,
+      "loss": 0.0369,
+      "step": 6532
+    },
+    {
+      "epoch": 2.901621141461248,
+      "grad_norm": 0.41566157060894093,
+      "learning_rate": 2.1243155814849003e-06,
+      "loss": 0.0322,
+      "step": 6533
+    },
+    {
+      "epoch": 2.9020652898067953,
+      "grad_norm": 0.4200346887421879,
+      "learning_rate": 2.1227299637333793e-06,
+      "loss": 0.0265,
+      "step": 6534
+    },
+    {
+      "epoch": 2.902509438152343,
+      "grad_norm": 0.481827142730724,
+      "learning_rate": 2.1211447784844223e-06,
+      "loss": 0.0332,
+      "step": 6535
+    },
+    {
+      "epoch": 2.9029535864978904,
+      "grad_norm": 0.41218410599797384,
+      "learning_rate": 2.1195600259763064e-06,
+      "loss": 0.0344,
+      "step": 6536
+    },
+    {
+      "epoch": 2.9033977348434377,
+      "grad_norm": 0.5125400394558322,
+      "learning_rate": 2.1179757064472495e-06,
+      "loss": 0.0311,
+      "step": 6537
+    },
+    {
+      "epoch": 2.903841883188985,
+      "grad_norm": 0.5243375032164893,
+      "learning_rate": 2.1163918201354005e-06,
+      "loss": 0.0442,
+      "step": 6538
+    },
+    {
+      "epoch": 2.9042860315345327,
+      "grad_norm": 0.36105640514776777,
+      "learning_rate": 2.114808367278845e-06,
+      "loss": 0.0315,
+      "step": 6539
+    },
+    {
+      "epoch": 2.90473017988008,
+      "grad_norm": 0.37491287513386307,
+      "learning_rate": 2.113225348115603e-06,
+      "loss": 0.0273,
+      "step": 6540
+    },
+    {
+      "epoch": 2.9051743282256273,
+      "grad_norm": 0.3675658951021628,
+      "learning_rate": 2.11164276288363e-06,
+      "loss": 0.0249,
+      "step": 6541
+    },
+    {
+      "epoch": 2.905618476571175,
+      "grad_norm": 0.47441864691687574,
+      "learning_rate": 2.110060611820813e-06,
+      "loss": 0.0439,
+      "step": 6542
+    },
+    {
+      "epoch": 2.9060626249167223,
+      "grad_norm": 0.4331767301461994,
+      "learning_rate": 2.1084788951649753e-06,
+      "loss": 0.0308,
+      "step": 6543
+    },
+    {
+      "epoch": 2.9065067732622696,
+      "grad_norm": 0.32889369683548264,
+      "learning_rate": 2.106897613153882e-06,
+      "loss": 0.0269,
+      "step": 6544
+    },
+    {
+      "epoch": 2.906950921607817,
+      "grad_norm": 0.46027509577610654,
+      "learning_rate": 2.105316766025221e-06,
+      "loss": 0.0309,
+      "step": 6545
+    },
+    {
+      "epoch": 2.907395069953364,
+      "grad_norm": 0.5494880014843939,
+      "learning_rate": 2.1037363540166224e-06,
+      "loss": 0.0273,
+      "step": 6546
+    },
+    {
+      "epoch": 2.907839218298912,
+      "grad_norm": 0.4011778001737706,
+      "learning_rate": 2.1021563773656493e-06,
+      "loss": 0.0352,
+      "step": 6547
+    },
+    {
+      "epoch": 2.908283366644459,
+      "grad_norm": 0.45266951475580575,
+      "learning_rate": 2.1005768363097977e-06,
+      "loss": 0.0301,
+      "step": 6548
+    },
+    {
+      "epoch": 2.908727514990007,
+      "grad_norm": 0.49868301394140047,
+      "learning_rate": 2.0989977310865e-06,
+      "loss": 0.0339,
+      "step": 6549
+    },
+    {
+      "epoch": 2.909171663335554,
+      "grad_norm": 0.6954722852910884,
+      "learning_rate": 2.0974190619331224e-06,
+      "loss": 0.0349,
+      "step": 6550
+    },
+    {
+      "epoch": 2.9096158116811015,
+      "grad_norm": 0.4383145753772345,
+      "learning_rate": 2.0958408290869662e-06,
+      "loss": 0.0356,
+      "step": 6551
+    },
+    {
+      "epoch": 2.910059960026649,
+      "grad_norm": 0.3704293568574071,
+      "learning_rate": 2.0942630327852687e-06,
+      "loss": 0.0303,
+      "step": 6552
+    },
+    {
+      "epoch": 2.910504108372196,
+      "grad_norm": 0.3498231235044318,
+      "learning_rate": 2.092685673265195e-06,
+      "loss": 0.0318,
+      "step": 6553
+    },
+    {
+      "epoch": 2.910948256717744,
+      "grad_norm": 0.36641412131018597,
+      "learning_rate": 2.0911087507638513e-06,
+      "loss": 0.0237,
+      "step": 6554
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.4005634976837491,
+      "learning_rate": 2.0895322655182754e-06,
+      "loss": 0.0327,
+      "step": 6555
+    },
+    {
+      "epoch": 2.9118365534088384,
+      "grad_norm": 0.36560498009667564,
+      "learning_rate": 2.0879562177654404e-06,
+      "loss": 0.0247,
+      "step": 6556
+    },
+    {
+      "epoch": 2.912280701754386,
+      "grad_norm": 0.38952731591083495,
+      "learning_rate": 2.0863806077422534e-06,
+      "loss": 0.027,
+      "step": 6557
+    },
+    {
+      "epoch": 2.9127248500999334,
+      "grad_norm": 0.4138708556738211,
+      "learning_rate": 2.0848054356855557e-06,
+      "loss": 0.0304,
+      "step": 6558
+    },
+    {
+      "epoch": 2.9131689984454807,
+      "grad_norm": 0.5073707851794668,
+      "learning_rate": 2.08323070183212e-06,
+      "loss": 0.0353,
+      "step": 6559
+    },
+    {
+      "epoch": 2.913613146791028,
+      "grad_norm": 0.5942930274632409,
+      "learning_rate": 2.081656406418658e-06,
+      "loss": 0.0331,
+      "step": 6560
+    },
+    {
+      "epoch": 2.9140572951365757,
+      "grad_norm": 0.35640079528674945,
+      "learning_rate": 2.080082549681811e-06,
+      "loss": 0.0279,
+      "step": 6561
+    },
+    {
+      "epoch": 2.914501443482123,
+      "grad_norm": 0.3814193495027854,
+      "learning_rate": 2.0785091318581577e-06,
+      "loss": 0.0305,
+      "step": 6562
+    },
+    {
+      "epoch": 2.9149455918276703,
+      "grad_norm": 0.358555755728585,
+      "learning_rate": 2.076936153184211e-06,
+      "loss": 0.0239,
+      "step": 6563
+    },
+    {
+      "epoch": 2.915389740173218,
+      "grad_norm": 0.4250693357842293,
+      "learning_rate": 2.0753636138964134e-06,
+      "loss": 0.0286,
+      "step": 6564
+    },
+    {
+      "epoch": 2.9158338885187653,
+      "grad_norm": 0.6221059510930096,
+      "learning_rate": 2.0737915142311454e-06,
+      "loss": 0.0441,
+      "step": 6565
+    },
+    {
+      "epoch": 2.9162780368643126,
+      "grad_norm": 0.38815256704717155,
+      "learning_rate": 2.07221985442472e-06,
+      "loss": 0.0292,
+      "step": 6566
+    },
+    {
+      "epoch": 2.91672218520986,
+      "grad_norm": 0.420249928229787,
+      "learning_rate": 2.0706486347133853e-06,
+      "loss": 0.0289,
+      "step": 6567
+    },
+    {
+      "epoch": 2.9171663335554077,
+      "grad_norm": 0.39867842058238456,
+      "learning_rate": 2.0690778553333215e-06,
+      "loss": 0.0399,
+      "step": 6568
+    },
+    {
+      "epoch": 2.917610481900955,
+      "grad_norm": 0.35700246538754193,
+      "learning_rate": 2.0675075165206456e-06,
+      "loss": 0.0328,
+      "step": 6569
+    },
+    {
+      "epoch": 2.9180546302465022,
+      "grad_norm": 0.39862639265555505,
+      "learning_rate": 2.0659376185114024e-06,
+      "loss": 0.0282,
+      "step": 6570
+    },
+    {
+      "epoch": 2.91849877859205,
+      "grad_norm": 0.3874386180786859,
+      "learning_rate": 2.064368161541576e-06,
+      "loss": 0.0269,
+      "step": 6571
+    },
+    {
+      "epoch": 2.9189429269375973,
+      "grad_norm": 0.35734528243866814,
+      "learning_rate": 2.0627991458470826e-06,
+      "loss": 0.0277,
+      "step": 6572
+    },
+    {
+      "epoch": 2.9193870752831446,
+      "grad_norm": 0.43326587014764417,
+      "learning_rate": 2.061230571663772e-06,
+      "loss": 0.0287,
+      "step": 6573
+    },
+    {
+      "epoch": 2.919831223628692,
+      "grad_norm": 0.442123671529554,
+      "learning_rate": 2.0596624392274277e-06,
+      "loss": 0.0348,
+      "step": 6574
+    },
+    {
+      "epoch": 2.920275371974239,
+      "grad_norm": 0.41360740653208006,
+      "learning_rate": 2.058094748773768e-06,
+      "loss": 0.0245,
+      "step": 6575
+    },
+    {
+      "epoch": 2.920719520319787,
+      "grad_norm": 0.4521622442444811,
+      "learning_rate": 2.05652750053844e-06,
+      "loss": 0.0272,
+      "step": 6576
+    },
+    {
+      "epoch": 2.921163668665334,
+      "grad_norm": 0.6652238384193798,
+      "learning_rate": 2.0549606947570295e-06,
+      "loss": 0.0403,
+      "step": 6577
+    },
+    {
+      "epoch": 2.921607817010882,
+      "grad_norm": 0.33561551996607286,
+      "learning_rate": 2.053394331665054e-06,
+      "loss": 0.023,
+      "step": 6578
+    },
+    {
+      "epoch": 2.922051965356429,
+      "grad_norm": 0.4533033136781839,
+      "learning_rate": 2.051828411497964e-06,
+      "loss": 0.0418,
+      "step": 6579
+    },
+    {
+      "epoch": 2.9224961137019765,
+      "grad_norm": 0.2933903835866281,
+      "learning_rate": 2.0502629344911475e-06,
+      "loss": 0.0204,
+      "step": 6580
+    },
+    {
+      "epoch": 2.9229402620475238,
+      "grad_norm": 0.4807368532427772,
+      "learning_rate": 2.0486979008799164e-06,
+      "loss": 0.0388,
+      "step": 6581
+    },
+    {
+      "epoch": 2.923384410393071,
+      "grad_norm": 0.39312159469638,
+      "learning_rate": 2.047133310899525e-06,
+      "loss": 0.0352,
+      "step": 6582
+    },
+    {
+      "epoch": 2.923828558738619,
+      "grad_norm": 0.49386287194544953,
+      "learning_rate": 2.045569164785157e-06,
+      "loss": 0.0318,
+      "step": 6583
+    },
+    {
+      "epoch": 2.924272707084166,
+      "grad_norm": 0.3909685186859985,
+      "learning_rate": 2.044005462771931e-06,
+      "loss": 0.0244,
+      "step": 6584
+    },
+    {
+      "epoch": 2.9247168554297134,
+      "grad_norm": 0.4648834902668276,
+      "learning_rate": 2.0424422050948976e-06,
+      "loss": 0.0387,
+      "step": 6585
+    },
+    {
+      "epoch": 2.925161003775261,
+      "grad_norm": 0.36487456799403395,
+      "learning_rate": 2.0408793919890424e-06,
+      "loss": 0.0288,
+      "step": 6586
+    },
+    {
+      "epoch": 2.9256051521208084,
+      "grad_norm": 0.5349577863858896,
+      "learning_rate": 2.0393170236892795e-06,
+      "loss": 0.0426,
+      "step": 6587
+    },
+    {
+      "epoch": 2.9260493004663557,
+      "grad_norm": 0.35432394100233194,
+      "learning_rate": 2.0377551004304613e-06,
+      "loss": 0.0304,
+      "step": 6588
+    },
+    {
+      "epoch": 2.926493448811903,
+      "grad_norm": 0.5133489099915194,
+      "learning_rate": 2.036193622447371e-06,
+      "loss": 0.0436,
+      "step": 6589
+    },
+    {
+      "epoch": 2.9269375971574507,
+      "grad_norm": 0.40489974505235427,
+      "learning_rate": 2.034632589974726e-06,
+      "loss": 0.0319,
+      "step": 6590
+    },
+    {
+      "epoch": 2.927381745502998,
+      "grad_norm": 0.3994253955209103,
+      "learning_rate": 2.033072003247175e-06,
+      "loss": 0.0308,
+      "step": 6591
+    },
+    {
+      "epoch": 2.9278258938485453,
+      "grad_norm": 0.34532953088972845,
+      "learning_rate": 2.0315118624993035e-06,
+      "loss": 0.0261,
+      "step": 6592
+    },
+    {
+      "epoch": 2.928270042194093,
+      "grad_norm": 0.35520607631530465,
+      "learning_rate": 2.0299521679656225e-06,
+      "loss": 0.0228,
+      "step": 6593
+    },
+    {
+      "epoch": 2.9287141905396403,
+      "grad_norm": 0.32280141712366317,
+      "learning_rate": 2.0283929198805837e-06,
+      "loss": 0.0189,
+      "step": 6594
+    },
+    {
+      "epoch": 2.9291583388851876,
+      "grad_norm": 0.43110913730957606,
+      "learning_rate": 2.0268341184785674e-06,
+      "loss": 0.0378,
+      "step": 6595
+    },
+    {
+      "epoch": 2.929602487230735,
+      "grad_norm": 0.354328605582052,
+      "learning_rate": 2.025275763993888e-06,
+      "loss": 0.0313,
+      "step": 6596
+    },
+    {
+      "epoch": 2.9300466355762826,
+      "grad_norm": 0.3880427101102111,
+      "learning_rate": 2.023717856660795e-06,
+      "loss": 0.0283,
+      "step": 6597
+    },
+    {
+      "epoch": 2.93049078392183,
+      "grad_norm": 0.37092577723252107,
+      "learning_rate": 2.0221603967134645e-06,
+      "loss": 0.0309,
+      "step": 6598
+    },
+    {
+      "epoch": 2.9309349322673772,
+      "grad_norm": 0.3853764249821867,
+      "learning_rate": 2.0206033843860113e-06,
+      "loss": 0.036,
+      "step": 6599
+    },
+    {
+      "epoch": 2.931379080612925,
+      "grad_norm": 0.32195066732125943,
+      "learning_rate": 2.0190468199124804e-06,
+      "loss": 0.0266,
+      "step": 6600
+    },
+    {
+      "epoch": 2.9318232289584722,
+      "grad_norm": 0.377934707171494,
+      "learning_rate": 2.01749070352685e-06,
+      "loss": 0.0323,
+      "step": 6601
+    },
+    {
+      "epoch": 2.9322673773040195,
+      "grad_norm": 0.4090601636781386,
+      "learning_rate": 2.0159350354630307e-06,
+      "loss": 0.0275,
+      "step": 6602
+    },
+    {
+      "epoch": 2.932711525649567,
+      "grad_norm": 0.4773326358669816,
+      "learning_rate": 2.0143798159548677e-06,
+      "loss": 0.0331,
+      "step": 6603
+    },
+    {
+      "epoch": 2.933155673995114,
+      "grad_norm": 0.4210797373291252,
+      "learning_rate": 2.0128250452361334e-06,
+      "loss": 0.0362,
+      "step": 6604
+    },
+    {
+      "epoch": 2.933599822340662,
+      "grad_norm": 0.4248736186586538,
+      "learning_rate": 2.0112707235405386e-06,
+      "loss": 0.0309,
+      "step": 6605
+    },
+    {
+      "epoch": 2.934043970686209,
+      "grad_norm": 0.3339641218919188,
+      "learning_rate": 2.0097168511017234e-06,
+      "loss": 0.0187,
+      "step": 6606
+    },
+    {
+      "epoch": 2.934488119031757,
+      "grad_norm": 0.5167791996296286,
+      "learning_rate": 2.0081634281532613e-06,
+      "loss": 0.0315,
+      "step": 6607
+    },
+    {
+      "epoch": 2.934932267377304,
+      "grad_norm": 0.4991803228092949,
+      "learning_rate": 2.0066104549286602e-06,
+      "loss": 0.0436,
+      "step": 6608
+    },
+    {
+      "epoch": 2.9353764157228515,
+      "grad_norm": 0.3779855137295941,
+      "learning_rate": 2.005057931661355e-06,
+      "loss": 0.0273,
+      "step": 6609
+    },
+    {
+      "epoch": 2.9358205640683988,
+      "grad_norm": 0.5036801733855913,
+      "learning_rate": 2.0035058585847173e-06,
+      "loss": 0.0328,
+      "step": 6610
+    },
+    {
+      "epoch": 2.936264712413946,
+      "grad_norm": 0.3544894089098337,
+      "learning_rate": 2.001954235932051e-06,
+      "loss": 0.0235,
+      "step": 6611
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.42819328292612396,
+      "learning_rate": 2.0004030639365907e-06,
+      "loss": 0.0308,
+      "step": 6612
+    },
+    {
+      "epoch": 2.937153009105041,
+      "grad_norm": 0.39806520440496346,
+      "learning_rate": 1.9988523428315045e-06,
+      "loss": 0.0277,
+      "step": 6613
+    },
+    {
+      "epoch": 2.9375971574505884,
+      "grad_norm": 0.41174193415314053,
+      "learning_rate": 1.997302072849893e-06,
+      "loss": 0.0282,
+      "step": 6614
+    },
+    {
+      "epoch": 2.938041305796136,
+      "grad_norm": 0.4414650832704903,
+      "learning_rate": 1.995752254224786e-06,
+      "loss": 0.0303,
+      "step": 6615
+    },
+    {
+      "epoch": 2.9384854541416834,
+      "grad_norm": 0.4432623492720395,
+      "learning_rate": 1.994202887189148e-06,
+      "loss": 0.0253,
+      "step": 6616
+    },
+    {
+      "epoch": 2.9389296024872307,
+      "grad_norm": 0.4800911306612613,
+      "learning_rate": 1.9926539719758747e-06,
+      "loss": 0.0388,
+      "step": 6617
+    },
+    {
+      "epoch": 2.939373750832778,
+      "grad_norm": 0.4059304636548806,
+      "learning_rate": 1.9911055088177967e-06,
+      "loss": 0.0276,
+      "step": 6618
+    },
+    {
+      "epoch": 2.9398178991783257,
+      "grad_norm": 0.5632995645750524,
+      "learning_rate": 1.9895574979476717e-06,
+      "loss": 0.0364,
+      "step": 6619
+    },
+    {
+      "epoch": 2.940262047523873,
+      "grad_norm": 0.3730668889292687,
+      "learning_rate": 1.9880099395981954e-06,
+      "loss": 0.0266,
+      "step": 6620
+    },
+    {
+      "epoch": 2.9407061958694203,
+      "grad_norm": 0.4942132883415099,
+      "learning_rate": 1.986462834001989e-06,
+      "loss": 0.0424,
+      "step": 6621
+    },
+    {
+      "epoch": 2.941150344214968,
+      "grad_norm": 0.35051082450586374,
+      "learning_rate": 1.984916181391609e-06,
+      "loss": 0.029,
+      "step": 6622
+    },
+    {
+      "epoch": 2.9415944925605153,
+      "grad_norm": 0.49757968739541286,
+      "learning_rate": 1.983369981999544e-06,
+      "loss": 0.0391,
+      "step": 6623
+    },
+    {
+      "epoch": 2.9420386409060626,
+      "grad_norm": 0.3866532839130346,
+      "learning_rate": 1.9818242360582145e-06,
+      "loss": 0.0196,
+      "step": 6624
+    },
+    {
+      "epoch": 2.94248278925161,
+      "grad_norm": 0.3312297948056805,
+      "learning_rate": 1.980278943799974e-06,
+      "loss": 0.024,
+      "step": 6625
+    },
+    {
+      "epoch": 2.942926937597157,
+      "grad_norm": 0.369854753089198,
+      "learning_rate": 1.978734105457103e-06,
+      "loss": 0.0261,
+      "step": 6626
+    },
+    {
+      "epoch": 2.943371085942705,
+      "grad_norm": 0.40104725649008754,
+      "learning_rate": 1.9771897212618172e-06,
+      "loss": 0.0338,
+      "step": 6627
+    },
+    {
+      "epoch": 2.943815234288252,
+      "grad_norm": 0.34952669002936676,
+      "learning_rate": 1.9756457914462677e-06,
+      "loss": 0.0245,
+      "step": 6628
+    },
+    {
+      "epoch": 2.9442593826338,
+      "grad_norm": 0.41129129029247236,
+      "learning_rate": 1.9741023162425265e-06,
+      "loss": 0.0313,
+      "step": 6629
+    },
+    {
+      "epoch": 2.9447035309793472,
+      "grad_norm": 0.5211003073625236,
+      "learning_rate": 1.9725592958826102e-06,
+      "loss": 0.0421,
+      "step": 6630
+    },
+    {
+      "epoch": 2.9451476793248945,
+      "grad_norm": 0.3984417865901208,
+      "learning_rate": 1.9710167305984607e-06,
+      "loss": 0.0375,
+      "step": 6631
+    },
+    {
+      "epoch": 2.945591827670442,
+      "grad_norm": 0.30941970478869846,
+      "learning_rate": 1.9694746206219477e-06,
+      "loss": 0.0216,
+      "step": 6632
+    },
+    {
+      "epoch": 2.946035976015989,
+      "grad_norm": 0.4229304742479391,
+      "learning_rate": 1.9679329661848795e-06,
+      "loss": 0.034,
+      "step": 6633
+    },
+    {
+      "epoch": 2.946480124361537,
+      "grad_norm": 0.476591153079035,
+      "learning_rate": 1.966391767518992e-06,
+      "loss": 0.0516,
+      "step": 6634
+    },
+    {
+      "epoch": 2.946924272707084,
+      "grad_norm": 0.6821582454050819,
+      "learning_rate": 1.9648510248559546e-06,
+      "loss": 0.034,
+      "step": 6635
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.43772389004233875,
+      "learning_rate": 1.9633107384273668e-06,
+      "loss": 0.0296,
+      "step": 6636
+    },
+    {
+      "epoch": 2.947812569398179,
+      "grad_norm": 0.49311057599048846,
+      "learning_rate": 1.9617709084647584e-06,
+      "loss": 0.0471,
+      "step": 6637
+    },
+    {
+      "epoch": 2.9482567177437264,
+      "grad_norm": 0.48536816799978794,
+      "learning_rate": 1.9602315351995928e-06,
+      "loss": 0.0393,
+      "step": 6638
+    },
+    {
+      "epoch": 2.9487008660892737,
+      "grad_norm": 0.5141630455505187,
+      "learning_rate": 1.958692618863264e-06,
+      "loss": 0.0342,
+      "step": 6639
+    },
+    {
+      "epoch": 2.949145014434821,
+      "grad_norm": 0.4586328384368867,
+      "learning_rate": 1.9571541596870974e-06,
+      "loss": 0.0391,
+      "step": 6640
+    },
+    {
+      "epoch": 2.9495891627803688,
+      "grad_norm": 0.4178954318106239,
+      "learning_rate": 1.9556161579023493e-06,
+      "loss": 0.0315,
+      "step": 6641
+    },
+    {
+      "epoch": 2.950033311125916,
+      "grad_norm": 0.4366977616184424,
+      "learning_rate": 1.9540786137402097e-06,
+      "loss": 0.0325,
+      "step": 6642
+    },
+    {
+      "epoch": 2.9504774594714633,
+      "grad_norm": 0.35044020018653177,
+      "learning_rate": 1.952541527431794e-06,
+      "loss": 0.0238,
+      "step": 6643
+    },
+    {
+      "epoch": 2.950921607817011,
+      "grad_norm": 0.36474391941516165,
+      "learning_rate": 1.951004899208154e-06,
+      "loss": 0.0284,
+      "step": 6644
+    },
+    {
+      "epoch": 2.9513657561625584,
+      "grad_norm": 0.3808970857069157,
+      "learning_rate": 1.9494687293002724e-06,
+      "loss": 0.0262,
+      "step": 6645
+    },
+    {
+      "epoch": 2.9518099045081057,
+      "grad_norm": 0.49087581676544356,
+      "learning_rate": 1.947933017939057e-06,
+      "loss": 0.0334,
+      "step": 6646
+    },
+    {
+      "epoch": 2.952254052853653,
+      "grad_norm": 0.4067523864649502,
+      "learning_rate": 1.946397765355356e-06,
+      "loss": 0.0224,
+      "step": 6647
+    },
+    {
+      "epoch": 2.9526982011992007,
+      "grad_norm": 0.5385094196478863,
+      "learning_rate": 1.9448629717799444e-06,
+      "loss": 0.0274,
+      "step": 6648
+    },
+    {
+      "epoch": 2.953142349544748,
+      "grad_norm": 0.4534794423284923,
+      "learning_rate": 1.9433286374435243e-06,
+      "loss": 0.0307,
+      "step": 6649
+    },
+    {
+      "epoch": 2.9535864978902953,
+      "grad_norm": 0.42491711765749635,
+      "learning_rate": 1.9417947625767338e-06,
+      "loss": 0.0282,
+      "step": 6650
+    },
+    {
+      "epoch": 2.954030646235843,
+      "grad_norm": 0.43191372654900584,
+      "learning_rate": 1.9402613474101418e-06,
+      "loss": 0.0284,
+      "step": 6651
+    },
+    {
+      "epoch": 2.9544747945813903,
+      "grad_norm": 0.4659070261041793,
+      "learning_rate": 1.9387283921742417e-06,
+      "loss": 0.0348,
+      "step": 6652
+    },
+    {
+      "epoch": 2.9549189429269376,
+      "grad_norm": 0.3521658502328422,
+      "learning_rate": 1.9371958970994697e-06,
+      "loss": 0.0264,
+      "step": 6653
+    },
+    {
+      "epoch": 2.955363091272485,
+      "grad_norm": 0.3744257822368085,
+      "learning_rate": 1.935663862416181e-06,
+      "loss": 0.0233,
+      "step": 6654
+    },
+    {
+      "epoch": 2.955807239618032,
+      "grad_norm": 0.3533483955176666,
+      "learning_rate": 1.934132288354667e-06,
+      "loss": 0.0223,
+      "step": 6655
+    },
+    {
+      "epoch": 2.95625138796358,
+      "grad_norm": 0.48990756810689967,
+      "learning_rate": 1.9326011751451523e-06,
+      "loss": 0.0296,
+      "step": 6656
+    },
+    {
+      "epoch": 2.956695536309127,
+      "grad_norm": 0.4155057113271156,
+      "learning_rate": 1.9310705230177834e-06,
+      "loss": 0.032,
+      "step": 6657
+    },
+    {
+      "epoch": 2.957139684654675,
+      "grad_norm": 0.46467269039209746,
+      "learning_rate": 1.9295403322026485e-06,
+      "loss": 0.0302,
+      "step": 6658
+    },
+    {
+      "epoch": 2.957583833000222,
+      "grad_norm": 0.436904391432258,
+      "learning_rate": 1.928010602929762e-06,
+      "loss": 0.0249,
+      "step": 6659
+    },
+    {
+      "epoch": 2.9580279813457695,
+      "grad_norm": 0.4078298528030376,
+      "learning_rate": 1.9264813354290635e-06,
+      "loss": 0.0339,
+      "step": 6660
+    },
+    {
+      "epoch": 2.958472129691317,
+      "grad_norm": 0.43921352541689956,
+      "learning_rate": 1.92495252993043e-06,
+      "loss": 0.0293,
+      "step": 6661
+    },
+    {
+      "epoch": 2.958916278036864,
+      "grad_norm": 0.47221505355593735,
+      "learning_rate": 1.9234241866636693e-06,
+      "loss": 0.0267,
+      "step": 6662
+    },
+    {
+      "epoch": 2.959360426382412,
+      "grad_norm": 0.3695229541487737,
+      "learning_rate": 1.9218963058585117e-06,
+      "loss": 0.0273,
+      "step": 6663
+    },
+    {
+      "epoch": 2.959804574727959,
+      "grad_norm": 0.3129170332680405,
+      "learning_rate": 1.9203688877446285e-06,
+      "loss": 0.0177,
+      "step": 6664
+    },
+    {
+      "epoch": 2.9602487230735064,
+      "grad_norm": 0.3892558088874507,
+      "learning_rate": 1.9188419325516177e-06,
+      "loss": 0.0381,
+      "step": 6665
+    },
+    {
+      "epoch": 2.960692871419054,
+      "grad_norm": 0.3716168146474431,
+      "learning_rate": 1.9173154405090024e-06,
+      "loss": 0.0222,
+      "step": 6666
+    },
+    {
+      "epoch": 2.9611370197646014,
+      "grad_norm": 0.4147701146563668,
+      "learning_rate": 1.9157894118462416e-06,
+      "loss": 0.0307,
+      "step": 6667
+    },
+    {
+      "epoch": 2.9615811681101487,
+      "grad_norm": 0.39499526563472726,
+      "learning_rate": 1.9142638467927254e-06,
+      "loss": 0.0365,
+      "step": 6668
+    },
+    {
+      "epoch": 2.962025316455696,
+      "grad_norm": 0.3874596806249992,
+      "learning_rate": 1.9127387455777673e-06,
+      "loss": 0.0321,
+      "step": 6669
+    },
+    {
+      "epoch": 2.9624694648012437,
+      "grad_norm": 0.42027304319891273,
+      "learning_rate": 1.911214108430623e-06,
+      "loss": 0.033,
+      "step": 6670
+    },
+    {
+      "epoch": 2.962913613146791,
+      "grad_norm": 0.44739921526779486,
+      "learning_rate": 1.9096899355804655e-06,
+      "loss": 0.0275,
+      "step": 6671
+    },
+    {
+      "epoch": 2.9633577614923383,
+      "grad_norm": 0.37802801428083804,
+      "learning_rate": 1.9081662272564055e-06,
+      "loss": 0.0319,
+      "step": 6672
+    },
+    {
+      "epoch": 2.963801909837886,
+      "grad_norm": 0.38442939573511425,
+      "learning_rate": 1.9066429836874844e-06,
+      "loss": 0.0295,
+      "step": 6673
+    },
+    {
+      "epoch": 2.9642460581834333,
+      "grad_norm": 0.6866561926311151,
+      "learning_rate": 1.9051202051026669e-06,
+      "loss": 0.0381,
+      "step": 6674
+    },
+    {
+      "epoch": 2.9646902065289806,
+      "grad_norm": 0.39471908589991933,
+      "learning_rate": 1.9035978917308568e-06,
+      "loss": 0.0282,
+      "step": 6675
+    },
+    {
+      "epoch": 2.965134354874528,
+      "grad_norm": 0.3683316118105389,
+      "learning_rate": 1.902076043800884e-06,
+      "loss": 0.0237,
+      "step": 6676
+    },
+    {
+      "epoch": 2.9655785032200757,
+      "grad_norm": 0.37626637800602947,
+      "learning_rate": 1.9005546615415044e-06,
+      "loss": 0.0241,
+      "step": 6677
+    },
+    {
+      "epoch": 2.966022651565623,
+      "grad_norm": 0.3496930484588171,
+      "learning_rate": 1.8990337451814095e-06,
+      "loss": 0.0286,
+      "step": 6678
+    },
+    {
+      "epoch": 2.9664667999111702,
+      "grad_norm": 0.41344810193656406,
+      "learning_rate": 1.897513294949221e-06,
+      "loss": 0.0273,
+      "step": 6679
+    },
+    {
+      "epoch": 2.966910948256718,
+      "grad_norm": 0.378120591478157,
+      "learning_rate": 1.895993311073483e-06,
+      "loss": 0.0294,
+      "step": 6680
+    },
+    {
+      "epoch": 2.9673550966022653,
+      "grad_norm": 0.33697535541119333,
+      "learning_rate": 1.8944737937826813e-06,
+      "loss": 0.0273,
+      "step": 6681
+    },
+    {
+      "epoch": 2.9677992449478126,
+      "grad_norm": 0.3663439761387446,
+      "learning_rate": 1.8929547433052202e-06,
+      "loss": 0.032,
+      "step": 6682
+    },
+    {
+      "epoch": 2.96824339329336,
+      "grad_norm": 0.5047952130541636,
+      "learning_rate": 1.8914361598694408e-06,
+      "loss": 0.0339,
+      "step": 6683
+    },
+    {
+      "epoch": 2.968687541638907,
+      "grad_norm": 0.4039062494421022,
+      "learning_rate": 1.8899180437036119e-06,
+      "loss": 0.0375,
+      "step": 6684
+    },
+    {
+      "epoch": 2.969131689984455,
+      "grad_norm": 0.3025200589569862,
+      "learning_rate": 1.8884003950359337e-06,
+      "loss": 0.0262,
+      "step": 6685
+    },
+    {
+      "epoch": 2.969575838330002,
+      "grad_norm": 0.44229043636076315,
+      "learning_rate": 1.8868832140945297e-06,
+      "loss": 0.028,
+      "step": 6686
+    },
+    {
+      "epoch": 2.97001998667555,
+      "grad_norm": 0.4708612100607534,
+      "learning_rate": 1.8853665011074645e-06,
+      "loss": 0.027,
+      "step": 6687
+    },
+    {
+      "epoch": 2.970464135021097,
+      "grad_norm": 0.3722598107484293,
+      "learning_rate": 1.8838502563027212e-06,
+      "loss": 0.0253,
+      "step": 6688
+    },
+    {
+      "epoch": 2.9709082833666445,
+      "grad_norm": 0.6036467003416066,
+      "learning_rate": 1.8823344799082177e-06,
+      "loss": 0.0284,
+      "step": 6689
+    },
+    {
+      "epoch": 2.9713524317121918,
+      "grad_norm": 0.34918427015083503,
+      "learning_rate": 1.8808191721518043e-06,
+      "loss": 0.0294,
+      "step": 6690
+    },
+    {
+      "epoch": 2.971796580057739,
+      "grad_norm": 0.42849383719138034,
+      "learning_rate": 1.879304333261251e-06,
+      "loss": 0.0278,
+      "step": 6691
+    },
+    {
+      "epoch": 2.972240728403287,
+      "grad_norm": 0.48072735895909585,
+      "learning_rate": 1.87778996346427e-06,
+      "loss": 0.0417,
+      "step": 6692
+    },
+    {
+      "epoch": 2.972684876748834,
+      "grad_norm": 0.29126197319559194,
+      "learning_rate": 1.8762760629884958e-06,
+      "loss": 0.0243,
+      "step": 6693
+    },
+    {
+      "epoch": 2.9731290250943814,
+      "grad_norm": 0.3610203458255818,
+      "learning_rate": 1.8747626320614904e-06,
+      "loss": 0.0269,
+      "step": 6694
+    },
+    {
+      "epoch": 2.973573173439929,
+      "grad_norm": 0.4145287000272817,
+      "learning_rate": 1.87324967091075e-06,
+      "loss": 0.0461,
+      "step": 6695
+    },
+    {
+      "epoch": 2.9740173217854764,
+      "grad_norm": 0.3579965634087584,
+      "learning_rate": 1.8717371797637002e-06,
+      "loss": 0.0296,
+      "step": 6696
+    },
+    {
+      "epoch": 2.9744614701310237,
+      "grad_norm": 0.41883841344207373,
+      "learning_rate": 1.8702251588476889e-06,
+      "loss": 0.0314,
+      "step": 6697
+    },
+    {
+      "epoch": 2.974905618476571,
+      "grad_norm": 0.5313573343030277,
+      "learning_rate": 1.868713608390005e-06,
+      "loss": 0.0408,
+      "step": 6698
+    },
+    {
+      "epoch": 2.9753497668221187,
+      "grad_norm": 0.39733118203820383,
+      "learning_rate": 1.8672025286178546e-06,
+      "loss": 0.0357,
+      "step": 6699
+    },
+    {
+      "epoch": 2.975793915167666,
+      "grad_norm": 0.4305571327880009,
+      "learning_rate": 1.8656919197583816e-06,
+      "loss": 0.0287,
+      "step": 6700
+    },
+    {
+      "epoch": 2.9762380635132133,
+      "grad_norm": 0.38007221498778115,
+      "learning_rate": 1.8641817820386576e-06,
+      "loss": 0.0256,
+      "step": 6701
+    },
+    {
+      "epoch": 2.976682211858761,
+      "grad_norm": 0.35438367987841246,
+      "learning_rate": 1.862672115685678e-06,
+      "loss": 0.0318,
+      "step": 6702
+    },
+    {
+      "epoch": 2.9771263602043083,
+      "grad_norm": 0.3521742278005089,
+      "learning_rate": 1.861162920926372e-06,
+      "loss": 0.0246,
+      "step": 6703
+    },
+    {
+      "epoch": 2.9775705085498556,
+      "grad_norm": 0.4587000972835763,
+      "learning_rate": 1.8596541979876016e-06,
+      "loss": 0.0403,
+      "step": 6704
+    },
+    {
+      "epoch": 2.978014656895403,
+      "grad_norm": 0.4391686978547565,
+      "learning_rate": 1.8581459470961488e-06,
+      "loss": 0.033,
+      "step": 6705
+    },
+    {
+      "epoch": 2.9784588052409506,
+      "grad_norm": 0.8681641943800276,
+      "learning_rate": 1.856638168478731e-06,
+      "loss": 0.0271,
+      "step": 6706
+    },
+    {
+      "epoch": 2.978902953586498,
+      "grad_norm": 0.3412875403024399,
+      "learning_rate": 1.8551308623619945e-06,
+      "loss": 0.0278,
+      "step": 6707
+    },
+    {
+      "epoch": 2.979347101932045,
+      "grad_norm": 0.39261569640506333,
+      "learning_rate": 1.8536240289725078e-06,
+      "loss": 0.035,
+      "step": 6708
+    },
+    {
+      "epoch": 2.979791250277593,
+      "grad_norm": 0.3999387719956585,
+      "learning_rate": 1.8521176685367804e-06,
+      "loss": 0.0287,
+      "step": 6709
+    },
+    {
+      "epoch": 2.9802353986231402,
+      "grad_norm": 0.3873753971954603,
+      "learning_rate": 1.850611781281239e-06,
+      "loss": 0.0322,
+      "step": 6710
+    },
+    {
+      "epoch": 2.9806795469686875,
+      "grad_norm": 0.4380866274093617,
+      "learning_rate": 1.8491063674322457e-06,
+      "loss": 0.033,
+      "step": 6711
+    },
+    {
+      "epoch": 2.981123695314235,
+      "grad_norm": 0.4171386719486352,
+      "learning_rate": 1.8476014272160896e-06,
+      "loss": 0.0354,
+      "step": 6712
+    },
+    {
+      "epoch": 2.981567843659782,
+      "grad_norm": 0.37606765896917305,
+      "learning_rate": 1.8460969608589913e-06,
+      "loss": 0.0309,
+      "step": 6713
+    },
+    {
+      "epoch": 2.98201199200533,
+      "grad_norm": 0.33316278037626945,
+      "learning_rate": 1.8445929685870912e-06,
+      "loss": 0.0281,
+      "step": 6714
+    },
+    {
+      "epoch": 2.982456140350877,
+      "grad_norm": 0.4295124088150545,
+      "learning_rate": 1.8430894506264724e-06,
+      "loss": 0.0281,
+      "step": 6715
+    },
+    {
+      "epoch": 2.982900288696425,
+      "grad_norm": 0.6219155251106016,
+      "learning_rate": 1.8415864072031335e-06,
+      "loss": 0.0412,
+      "step": 6716
+    },
+    {
+      "epoch": 2.983344437041972,
+      "grad_norm": 1.0649791746325505,
+      "learning_rate": 1.8400838385430104e-06,
+      "loss": 0.0296,
+      "step": 6717
+    },
+    {
+      "epoch": 2.9837885853875195,
+      "grad_norm": 0.4321630566554828,
+      "learning_rate": 1.838581744871965e-06,
+      "loss": 0.0349,
+      "step": 6718
+    },
+    {
+      "epoch": 2.9842327337330667,
+      "grad_norm": 0.3860040470778094,
+      "learning_rate": 1.8370801264157857e-06,
+      "loss": 0.029,
+      "step": 6719
+    },
+    {
+      "epoch": 2.984676882078614,
+      "grad_norm": 0.38156983645011616,
+      "learning_rate": 1.8355789834001898e-06,
+      "loss": 0.0284,
+      "step": 6720
+    },
+    {
+      "epoch": 2.9851210304241618,
+      "grad_norm": 0.5127281233435997,
+      "learning_rate": 1.8340783160508297e-06,
+      "loss": 0.0355,
+      "step": 6721
+    },
+    {
+      "epoch": 2.985565178769709,
+      "grad_norm": 0.3767235771133534,
+      "learning_rate": 1.8325781245932772e-06,
+      "loss": 0.0297,
+      "step": 6722
+    },
+    {
+      "epoch": 2.9860093271152564,
+      "grad_norm": 0.4559507813085918,
+      "learning_rate": 1.8310784092530376e-06,
+      "loss": 0.0362,
+      "step": 6723
+    },
+    {
+      "epoch": 2.986453475460804,
+      "grad_norm": 0.36189720669347597,
+      "learning_rate": 1.8295791702555455e-06,
+      "loss": 0.0294,
+      "step": 6724
+    },
+    {
+      "epoch": 2.9868976238063514,
+      "grad_norm": 0.3457736093542848,
+      "learning_rate": 1.8280804078261577e-06,
+      "loss": 0.0305,
+      "step": 6725
+    },
+    {
+      "epoch": 2.9873417721518987,
+      "grad_norm": 0.5911976787943448,
+      "learning_rate": 1.826582122190167e-06,
+      "loss": 0.0368,
+      "step": 6726
+    },
+    {
+      "epoch": 2.987785920497446,
+      "grad_norm": 0.4315279094342813,
+      "learning_rate": 1.8250843135727898e-06,
+      "loss": 0.0365,
+      "step": 6727
+    },
+    {
+      "epoch": 2.9882300688429937,
+      "grad_norm": 0.3782115314248128,
+      "learning_rate": 1.8235869821991726e-06,
+      "loss": 0.0297,
+      "step": 6728
+    },
+    {
+      "epoch": 2.988674217188541,
+      "grad_norm": 0.42412903793940915,
+      "learning_rate": 1.8220901282943915e-06,
+      "loss": 0.0348,
+      "step": 6729
+    },
+    {
+      "epoch": 2.9891183655340883,
+      "grad_norm": 0.42286509801094985,
+      "learning_rate": 1.820593752083446e-06,
+      "loss": 0.0356,
+      "step": 6730
+    },
+    {
+      "epoch": 2.989562513879636,
+      "grad_norm": 0.48807857838384455,
+      "learning_rate": 1.8190978537912662e-06,
+      "loss": 0.0441,
+      "step": 6731
+    },
+    {
+      "epoch": 2.9900066622251833,
+      "grad_norm": 0.4660013011382873,
+      "learning_rate": 1.8176024336427167e-06,
+      "loss": 0.0356,
+      "step": 6732
+    },
+    {
+      "epoch": 2.9904508105707306,
+      "grad_norm": 0.3894973608346122,
+      "learning_rate": 1.8161074918625792e-06,
+      "loss": 0.0234,
+      "step": 6733
+    },
+    {
+      "epoch": 2.990894958916278,
+      "grad_norm": 0.43148032392899044,
+      "learning_rate": 1.8146130286755704e-06,
+      "loss": 0.0316,
+      "step": 6734
+    },
+    {
+      "epoch": 2.9913391072618256,
+      "grad_norm": 0.4541609973903636,
+      "learning_rate": 1.8131190443063357e-06,
+      "loss": 0.0228,
+      "step": 6735
+    },
+    {
+      "epoch": 2.991783255607373,
+      "grad_norm": 0.37931640734503375,
+      "learning_rate": 1.8116255389794418e-06,
+      "loss": 0.03,
+      "step": 6736
+    },
+    {
+      "epoch": 2.99222740395292,
+      "grad_norm": 0.38246202047737776,
+      "learning_rate": 1.8101325129193897e-06,
+      "loss": 0.0432,
+      "step": 6737
+    },
+    {
+      "epoch": 2.992671552298468,
+      "grad_norm": 0.42501044694023266,
+      "learning_rate": 1.8086399663506099e-06,
+      "loss": 0.0378,
+      "step": 6738
+    },
+    {
+      "epoch": 2.993115700644015,
+      "grad_norm": 0.4349574825819462,
+      "learning_rate": 1.8071478994974534e-06,
+      "loss": 0.022,
+      "step": 6739
+    },
+    {
+      "epoch": 2.9935598489895625,
+      "grad_norm": 0.36182919079127596,
+      "learning_rate": 1.8056563125842046e-06,
+      "loss": 0.0281,
+      "step": 6740
+    },
+    {
+      "epoch": 2.99400399733511,
+      "grad_norm": 0.3297563936891292,
+      "learning_rate": 1.8041652058350768e-06,
+      "loss": 0.0258,
+      "step": 6741
+    },
+    {
+      "epoch": 2.994448145680657,
+      "grad_norm": 0.7003299863726014,
+      "learning_rate": 1.802674579474204e-06,
+      "loss": 0.0351,
+      "step": 6742
+    },
+    {
+      "epoch": 2.994892294026205,
+      "grad_norm": 0.6139326715543403,
+      "learning_rate": 1.801184433725655e-06,
+      "loss": 0.0379,
+      "step": 6743
+    },
+    {
+      "epoch": 2.995336442371752,
+      "grad_norm": 0.5062057494487118,
+      "learning_rate": 1.7996947688134241e-06,
+      "loss": 0.0315,
+      "step": 6744
+    },
+    {
+      "epoch": 2.9957805907173,
+      "grad_norm": 0.7604714501213958,
+      "learning_rate": 1.7982055849614327e-06,
+      "loss": 0.0435,
+      "step": 6745
+    },
+    {
+      "epoch": 2.996224739062847,
+      "grad_norm": 0.4500952174598657,
+      "learning_rate": 1.7967168823935333e-06,
+      "loss": 0.0364,
+      "step": 6746
+    },
+    {
+      "epoch": 2.9966688874083944,
+      "grad_norm": 0.6302920187503901,
+      "learning_rate": 1.7952286613334986e-06,
+      "loss": 0.0319,
+      "step": 6747
+    },
+    {
+      "epoch": 2.9971130357539417,
+      "grad_norm": 0.36539112186296896,
+      "learning_rate": 1.793740922005034e-06,
+      "loss": 0.0308,
+      "step": 6748
+    },
+    {
+      "epoch": 2.997557184099489,
+      "grad_norm": 0.43381962390876766,
+      "learning_rate": 1.7922536646317767e-06,
+      "loss": 0.0377,
+      "step": 6749
+    },
+    {
+      "epoch": 2.9980013324450367,
+      "grad_norm": 0.33856021759003035,
+      "learning_rate": 1.7907668894372826e-06,
+      "loss": 0.0239,
+      "step": 6750
+    },
+    {
+      "epoch": 2.998445480790584,
+      "grad_norm": 0.37584653604138824,
+      "learning_rate": 1.78928059664504e-06,
+      "loss": 0.024,
+      "step": 6751
+    },
+    {
+      "epoch": 2.9988896291361313,
+      "grad_norm": 0.361912705997716,
+      "learning_rate": 1.7877947864784662e-06,
+      "loss": 0.0261,
+      "step": 6752
+    },
+    {
+      "epoch": 2.999333777481679,
+      "grad_norm": 0.35751593403707377,
+      "learning_rate": 1.7863094591609003e-06,
+      "loss": 0.0254,
+      "step": 6753
+    },
+    {
+      "epoch": 2.9997779258272264,
+      "grad_norm": 0.4724255587714232,
+      "learning_rate": 1.7848246149156134e-06,
+      "loss": 0.0421,
+      "step": 6754
+    },
+    {
+      "epoch": 2.9997779258272264,
+      "eval_loss": 0.03672339767217636,
+      "eval_runtime": 403.6019,
+      "eval_samples_per_second": 37.577,
+      "eval_steps_per_second": 1.174,
+      "step": 6754
+    },
+    {
+      "epoch": 3.0002220741727736,
+      "grad_norm": 0.3269520644923097,
+      "learning_rate": 1.783340253965803e-06,
+      "loss": 0.0215,
+      "step": 6755
+    },
+    {
+      "epoch": 3.000666222518321,
+      "grad_norm": 0.3360624322653449,
+      "learning_rate": 1.7818563765345942e-06,
+      "loss": 0.0284,
+      "step": 6756
+    },
+    {
+      "epoch": 3.0011103708638687,
+      "grad_norm": 0.3740872446254905,
+      "learning_rate": 1.7803729828450405e-06,
+      "loss": 0.0207,
+      "step": 6757
+    },
+    {
+      "epoch": 3.001554519209416,
+      "grad_norm": 0.29900746203751005,
+      "learning_rate": 1.7788900731201174e-06,
+      "loss": 0.0206,
+      "step": 6758
+    },
+    {
+      "epoch": 3.0019986675549633,
+      "grad_norm": 0.4119590022541271,
+      "learning_rate": 1.7774076475827335e-06,
+      "loss": 0.0239,
+      "step": 6759
+    },
+    {
+      "epoch": 3.002442815900511,
+      "grad_norm": 0.4073542383308494,
+      "learning_rate": 1.7759257064557229e-06,
+      "loss": 0.022,
+      "step": 6760
+    },
+    {
+      "epoch": 3.0028869642460583,
+      "grad_norm": 0.6159273406264103,
+      "learning_rate": 1.7744442499618453e-06,
+      "loss": 0.032,
+      "step": 6761
+    },
+    {
+      "epoch": 3.0033311125916056,
+      "grad_norm": 0.8514717794851285,
+      "learning_rate": 1.77296327832379e-06,
+      "loss": 0.0421,
+      "step": 6762
+    },
+    {
+      "epoch": 3.003775260937153,
+      "grad_norm": 0.3400576483770952,
+      "learning_rate": 1.7714827917641737e-06,
+      "loss": 0.0209,
+      "step": 6763
+    },
+    {
+      "epoch": 3.0042194092827006,
+      "grad_norm": 0.38696151832577724,
+      "learning_rate": 1.7700027905055344e-06,
+      "loss": 0.0232,
+      "step": 6764
+    },
+    {
+      "epoch": 3.004663557628248,
+      "grad_norm": 0.40863063034907293,
+      "learning_rate": 1.7685232747703424e-06,
+      "loss": 0.0246,
+      "step": 6765
+    },
+    {
+      "epoch": 3.005107705973795,
+      "grad_norm": 0.39214643564722806,
+      "learning_rate": 1.7670442447809989e-06,
+      "loss": 0.0283,
+      "step": 6766
+    },
+    {
+      "epoch": 3.0055518543193425,
+      "grad_norm": 0.5937092447872463,
+      "learning_rate": 1.7655657007598216e-06,
+      "loss": 0.0269,
+      "step": 6767
+    },
+    {
+      "epoch": 3.00599600266489,
+      "grad_norm": 0.43120240703534374,
+      "learning_rate": 1.7640876429290633e-06,
+      "loss": 0.0238,
+      "step": 6768
+    },
+    {
+      "epoch": 3.0064401510104375,
+      "grad_norm": 0.405113495539343,
+      "learning_rate": 1.7626100715109018e-06,
+      "loss": 0.0206,
+      "step": 6769
+    },
+    {
+      "epoch": 3.006884299355985,
+      "grad_norm": 0.35078016785761906,
+      "learning_rate": 1.761132986727439e-06,
+      "loss": 0.0207,
+      "step": 6770
+    },
+    {
+      "epoch": 3.0073284477015325,
+      "grad_norm": 0.36832682307040443,
+      "learning_rate": 1.7596563888007073e-06,
+      "loss": 0.0246,
+      "step": 6771
+    },
+    {
+      "epoch": 3.00777259604708,
+      "grad_norm": 0.38464153090067194,
+      "learning_rate": 1.7581802779526642e-06,
+      "loss": 0.0187,
+      "step": 6772
+    },
+    {
+      "epoch": 3.008216744392627,
+      "grad_norm": 0.44328209442836686,
+      "learning_rate": 1.7567046544051935e-06,
+      "loss": 0.0254,
+      "step": 6773
+    },
+    {
+      "epoch": 3.0086608927381744,
+      "grad_norm": 0.4311223508530591,
+      "learning_rate": 1.7552295183801093e-06,
+      "loss": 0.0282,
+      "step": 6774
+    },
+    {
+      "epoch": 3.009105041083722,
+      "grad_norm": 0.4187784150098612,
+      "learning_rate": 1.7537548700991463e-06,
+      "loss": 0.026,
+      "step": 6775
+    },
+    {
+      "epoch": 3.0095491894292694,
+      "grad_norm": 0.49752308362720904,
+      "learning_rate": 1.75228070978397e-06,
+      "loss": 0.0333,
+      "step": 6776
+    },
+    {
+      "epoch": 3.0099933377748167,
+      "grad_norm": 0.3661955272882175,
+      "learning_rate": 1.750807037656172e-06,
+      "loss": 0.0263,
+      "step": 6777
+    },
+    {
+      "epoch": 3.010437486120364,
+      "grad_norm": 0.3951032549450352,
+      "learning_rate": 1.7493338539372701e-06,
+      "loss": 0.0245,
+      "step": 6778
+    },
+    {
+      "epoch": 3.0108816344659117,
+      "grad_norm": 0.4288937494263082,
+      "learning_rate": 1.7478611588487098e-06,
+      "loss": 0.0224,
+      "step": 6779
+    },
+    {
+      "epoch": 3.011325782811459,
+      "grad_norm": 1.0043569602354414,
+      "learning_rate": 1.7463889526118628e-06,
+      "loss": 0.0321,
+      "step": 6780
+    },
+    {
+      "epoch": 3.0117699311570063,
+      "grad_norm": 0.345117052755733,
+      "learning_rate": 1.7449172354480236e-06,
+      "loss": 0.0222,
+      "step": 6781
+    },
+    {
+      "epoch": 3.012214079502554,
+      "grad_norm": 0.37263032859522277,
+      "learning_rate": 1.7434460075784183e-06,
+      "loss": 0.0371,
+      "step": 6782
+    },
+    {
+      "epoch": 3.0126582278481013,
+      "grad_norm": 0.42797870753405914,
+      "learning_rate": 1.741975269224197e-06,
+      "loss": 0.0275,
+      "step": 6783
+    },
+    {
+      "epoch": 3.0131023761936486,
+      "grad_norm": 0.4487558433311281,
+      "learning_rate": 1.7405050206064372e-06,
+      "loss": 0.0278,
+      "step": 6784
+    },
+    {
+      "epoch": 3.013546524539196,
+      "grad_norm": 0.3104547304379723,
+      "learning_rate": 1.739035261946142e-06,
+      "loss": 0.0162,
+      "step": 6785
+    },
+    {
+      "epoch": 3.0139906728847436,
+      "grad_norm": 0.4274226742898415,
+      "learning_rate": 1.7375659934642425e-06,
+      "loss": 0.0232,
+      "step": 6786
+    },
+    {
+      "epoch": 3.014434821230291,
+      "grad_norm": 0.4378508355204737,
+      "learning_rate": 1.7360972153815919e-06,
+      "loss": 0.0263,
+      "step": 6787
+    },
+    {
+      "epoch": 3.0148789695758382,
+      "grad_norm": 0.35914322608698723,
+      "learning_rate": 1.7346289279189732e-06,
+      "loss": 0.0236,
+      "step": 6788
+    },
+    {
+      "epoch": 3.015323117921386,
+      "grad_norm": 0.3717761462102794,
+      "learning_rate": 1.7331611312970965e-06,
+      "loss": 0.0223,
+      "step": 6789
+    },
+    {
+      "epoch": 3.0157672662669333,
+      "grad_norm": 0.49078555792657397,
+      "learning_rate": 1.7316938257365945e-06,
+      "loss": 0.0479,
+      "step": 6790
+    },
+    {
+      "epoch": 3.0162114146124805,
+      "grad_norm": 0.5303389635543914,
+      "learning_rate": 1.7302270114580316e-06,
+      "loss": 0.0257,
+      "step": 6791
+    },
+    {
+      "epoch": 3.016655562958028,
+      "grad_norm": 0.5963610689982617,
+      "learning_rate": 1.7287606886818914e-06,
+      "loss": 0.025,
+      "step": 6792
+    },
+    {
+      "epoch": 3.0170997113035756,
+      "grad_norm": 0.3358156185785528,
+      "learning_rate": 1.7272948576285874e-06,
+      "loss": 0.0259,
+      "step": 6793
+    },
+    {
+      "epoch": 3.017543859649123,
+      "grad_norm": 0.44451083436086486,
+      "learning_rate": 1.7258295185184604e-06,
+      "loss": 0.0371,
+      "step": 6794
+    },
+    {
+      "epoch": 3.01798800799467,
+      "grad_norm": 0.8409978484969116,
+      "learning_rate": 1.7243646715717754e-06,
+      "loss": 0.0198,
+      "step": 6795
+    },
+    {
+      "epoch": 3.0184321563402174,
+      "grad_norm": 0.40598942285053524,
+      "learning_rate": 1.7229003170087232e-06,
+      "loss": 0.0253,
+      "step": 6796
+    },
+    {
+      "epoch": 3.018876304685765,
+      "grad_norm": 0.40045735020677964,
+      "learning_rate": 1.7214364550494235e-06,
+      "loss": 0.0305,
+      "step": 6797
+    },
+    {
+      "epoch": 3.0193204530313125,
+      "grad_norm": 0.4205212377847487,
+      "learning_rate": 1.7199730859139157e-06,
+      "loss": 0.0233,
+      "step": 6798
+    },
+    {
+      "epoch": 3.0197646013768598,
+      "grad_norm": 0.39200127688818526,
+      "learning_rate": 1.7185102098221713e-06,
+      "loss": 0.0225,
+      "step": 6799
+    },
+    {
+      "epoch": 3.0202087497224075,
+      "grad_norm": 0.4741877419032876,
+      "learning_rate": 1.717047826994085e-06,
+      "loss": 0.0262,
+      "step": 6800
+    },
+    {
+      "epoch": 3.020652898067955,
+      "grad_norm": 0.40147810647504373,
+      "learning_rate": 1.7155859376494776e-06,
+      "loss": 0.0267,
+      "step": 6801
+    },
+    {
+      "epoch": 3.021097046413502,
+      "grad_norm": 0.4280335792767891,
+      "learning_rate": 1.7141245420080982e-06,
+      "loss": 0.0242,
+      "step": 6802
+    },
+    {
+      "epoch": 3.0215411947590494,
+      "grad_norm": 0.34267631232424817,
+      "learning_rate": 1.7126636402896158e-06,
+      "loss": 0.0228,
+      "step": 6803
+    },
+    {
+      "epoch": 3.021985343104597,
+      "grad_norm": 0.517542942077693,
+      "learning_rate": 1.7112032327136296e-06,
+      "loss": 0.0289,
+      "step": 6804
+    },
+    {
+      "epoch": 3.0224294914501444,
+      "grad_norm": 0.48842613500647986,
+      "learning_rate": 1.7097433194996654e-06,
+      "loss": 0.0362,
+      "step": 6805
+    },
+    {
+      "epoch": 3.0228736397956917,
+      "grad_norm": 0.5122441524272605,
+      "learning_rate": 1.7082839008671714e-06,
+      "loss": 0.0327,
+      "step": 6806
+    },
+    {
+      "epoch": 3.023317788141239,
+      "grad_norm": 0.3656414000536619,
+      "learning_rate": 1.706824977035524e-06,
+      "loss": 0.0239,
+      "step": 6807
+    },
+    {
+      "epoch": 3.0237619364867867,
+      "grad_norm": 0.35844316810361615,
+      "learning_rate": 1.705366548224025e-06,
+      "loss": 0.022,
+      "step": 6808
+    },
+    {
+      "epoch": 3.024206084832334,
+      "grad_norm": 0.5208246273651607,
+      "learning_rate": 1.7039086146518986e-06,
+      "loss": 0.0396,
+      "step": 6809
+    },
+    {
+      "epoch": 3.0246502331778813,
+      "grad_norm": 0.35888427376762055,
+      "learning_rate": 1.7024511765382978e-06,
+      "loss": 0.0218,
+      "step": 6810
+    },
+    {
+      "epoch": 3.025094381523429,
+      "grad_norm": 0.4996038911163621,
+      "learning_rate": 1.7009942341023012e-06,
+      "loss": 0.031,
+      "step": 6811
+    },
+    {
+      "epoch": 3.0255385298689763,
+      "grad_norm": 0.4715140989011031,
+      "learning_rate": 1.699537787562911e-06,
+      "loss": 0.027,
+      "step": 6812
+    },
+    {
+      "epoch": 3.0259826782145236,
+      "grad_norm": 0.4700851815071929,
+      "learning_rate": 1.6980818371390567e-06,
+      "loss": 0.0352,
+      "step": 6813
+    },
+    {
+      "epoch": 3.026426826560071,
+      "grad_norm": 0.42536546436391975,
+      "learning_rate": 1.6966263830495939e-06,
+      "loss": 0.0307,
+      "step": 6814
+    },
+    {
+      "epoch": 3.0268709749056186,
+      "grad_norm": 0.3928810675523391,
+      "learning_rate": 1.6951714255132985e-06,
+      "loss": 0.0317,
+      "step": 6815
+    },
+    {
+      "epoch": 3.027315123251166,
+      "grad_norm": 0.36328069775398497,
+      "learning_rate": 1.6937169647488765e-06,
+      "loss": 0.0225,
+      "step": 6816
+    },
+    {
+      "epoch": 3.027759271596713,
+      "grad_norm": 0.4176283792585422,
+      "learning_rate": 1.6922630009749592e-06,
+      "loss": 0.0335,
+      "step": 6817
+    },
+    {
+      "epoch": 3.0282034199422605,
+      "grad_norm": 0.45941453677809496,
+      "learning_rate": 1.6908095344101016e-06,
+      "loss": 0.0307,
+      "step": 6818
+    },
+    {
+      "epoch": 3.0286475682878082,
+      "grad_norm": 0.4022633883354449,
+      "learning_rate": 1.6893565652727857e-06,
+      "loss": 0.0285,
+      "step": 6819
+    },
+    {
+      "epoch": 3.0290917166333555,
+      "grad_norm": 0.46606194138773704,
+      "learning_rate": 1.687904093781414e-06,
+      "loss": 0.0346,
+      "step": 6820
+    },
+    {
+      "epoch": 3.029535864978903,
+      "grad_norm": 0.3781526947765934,
+      "learning_rate": 1.68645212015432e-06,
+      "loss": 0.0217,
+      "step": 6821
+    },
+    {
+      "epoch": 3.0299800133244505,
+      "grad_norm": 0.40815913991276614,
+      "learning_rate": 1.68500064460976e-06,
+      "loss": 0.0303,
+      "step": 6822
+    },
+    {
+      "epoch": 3.030424161669998,
+      "grad_norm": 0.42153363035695735,
+      "learning_rate": 1.6835496673659145e-06,
+      "loss": 0.0329,
+      "step": 6823
+    },
+    {
+      "epoch": 3.030868310015545,
+      "grad_norm": 0.5257269502253524,
+      "learning_rate": 1.6820991886408911e-06,
+      "loss": 0.0342,
+      "step": 6824
+    },
+    {
+      "epoch": 3.0313124583610924,
+      "grad_norm": 0.36375569874164343,
+      "learning_rate": 1.6806492086527226e-06,
+      "loss": 0.0268,
+      "step": 6825
+    },
+    {
+      "epoch": 3.03175660670664,
+      "grad_norm": 0.4517518524381332,
+      "learning_rate": 1.6791997276193623e-06,
+      "loss": 0.0291,
+      "step": 6826
+    },
+    {
+      "epoch": 3.0322007550521874,
+      "grad_norm": 0.41358501413398874,
+      "learning_rate": 1.6777507457586933e-06,
+      "loss": 0.0263,
+      "step": 6827
+    },
+    {
+      "epoch": 3.0326449033977347,
+      "grad_norm": 0.4371773593425646,
+      "learning_rate": 1.6763022632885223e-06,
+      "loss": 0.026,
+      "step": 6828
+    },
+    {
+      "epoch": 3.0330890517432825,
+      "grad_norm": 0.41878585193753987,
+      "learning_rate": 1.674854280426581e-06,
+      "loss": 0.0289,
+      "step": 6829
+    },
+    {
+      "epoch": 3.0335332000888298,
+      "grad_norm": 0.6072899498645479,
+      "learning_rate": 1.6734067973905272e-06,
+      "loss": 0.046,
+      "step": 6830
+    },
+    {
+      "epoch": 3.033977348434377,
+      "grad_norm": 0.37768879943900463,
+      "learning_rate": 1.6719598143979392e-06,
+      "loss": 0.0175,
+      "step": 6831
+    },
+    {
+      "epoch": 3.0344214967799243,
+      "grad_norm": 0.42243370237101113,
+      "learning_rate": 1.6705133316663247e-06,
+      "loss": 0.0318,
+      "step": 6832
+    },
+    {
+      "epoch": 3.034865645125472,
+      "grad_norm": 0.4272111914281022,
+      "learning_rate": 1.6690673494131143e-06,
+      "loss": 0.0318,
+      "step": 6833
+    },
+    {
+      "epoch": 3.0353097934710194,
+      "grad_norm": 0.555542458187917,
+      "learning_rate": 1.6676218678556637e-06,
+      "loss": 0.0333,
+      "step": 6834
+    },
+    {
+      "epoch": 3.0357539418165667,
+      "grad_norm": 0.5174806483008364,
+      "learning_rate": 1.6661768872112544e-06,
+      "loss": 0.0257,
+      "step": 6835
+    },
+    {
+      "epoch": 3.036198090162114,
+      "grad_norm": 0.40261446781293897,
+      "learning_rate": 1.6647324076970917e-06,
+      "loss": 0.0271,
+      "step": 6836
+    },
+    {
+      "epoch": 3.0366422385076617,
+      "grad_norm": 0.28515858656757614,
+      "learning_rate": 1.663288429530303e-06,
+      "loss": 0.0139,
+      "step": 6837
+    },
+    {
+      "epoch": 3.037086386853209,
+      "grad_norm": 0.3801745026945722,
+      "learning_rate": 1.661844952927944e-06,
+      "loss": 0.0278,
+      "step": 6838
+    },
+    {
+      "epoch": 3.0375305351987563,
+      "grad_norm": 0.5653418682210891,
+      "learning_rate": 1.660401978106994e-06,
+      "loss": 0.0356,
+      "step": 6839
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.4351497414805404,
+      "learning_rate": 1.6589595052843567e-06,
+      "loss": 0.0255,
+      "step": 6840
+    },
+    {
+      "epoch": 3.0384188318898513,
+      "grad_norm": 0.36924524660599933,
+      "learning_rate": 1.6575175346768597e-06,
+      "loss": 0.0213,
+      "step": 6841
+    },
+    {
+      "epoch": 3.0388629802353986,
+      "grad_norm": 0.36560298465093416,
+      "learning_rate": 1.6560760665012581e-06,
+      "loss": 0.0291,
+      "step": 6842
+    },
+    {
+      "epoch": 3.039307128580946,
+      "grad_norm": 0.5231650605635862,
+      "learning_rate": 1.6546351009742252e-06,
+      "loss": 0.0416,
+      "step": 6843
+    },
+    {
+      "epoch": 3.0397512769264936,
+      "grad_norm": 0.40002607438188575,
+      "learning_rate": 1.6531946383123647e-06,
+      "loss": 0.0279,
+      "step": 6844
+    },
+    {
+      "epoch": 3.040195425272041,
+      "grad_norm": 0.3820335905823042,
+      "learning_rate": 1.6517546787322019e-06,
+      "loss": 0.0332,
+      "step": 6845
+    },
+    {
+      "epoch": 3.040639573617588,
+      "grad_norm": 0.4052075541407361,
+      "learning_rate": 1.6503152224501883e-06,
+      "loss": 0.0273,
+      "step": 6846
+    },
+    {
+      "epoch": 3.0410837219631355,
+      "grad_norm": 0.40616006159309487,
+      "learning_rate": 1.6488762696826992e-06,
+      "loss": 0.029,
+      "step": 6847
+    },
+    {
+      "epoch": 3.041527870308683,
+      "grad_norm": 0.41149108296545567,
+      "learning_rate": 1.6474378206460306e-06,
+      "loss": 0.0239,
+      "step": 6848
+    },
+    {
+      "epoch": 3.0419720186542305,
+      "grad_norm": 0.35312483842889825,
+      "learning_rate": 1.6459998755564078e-06,
+      "loss": 0.0265,
+      "step": 6849
+    },
+    {
+      "epoch": 3.042416166999778,
+      "grad_norm": 0.44423084298706683,
+      "learning_rate": 1.64456243462998e-06,
+      "loss": 0.0214,
+      "step": 6850
+    },
+    {
+      "epoch": 3.0428603153453255,
+      "grad_norm": 0.3441262864593607,
+      "learning_rate": 1.6431254980828137e-06,
+      "loss": 0.0261,
+      "step": 6851
+    },
+    {
+      "epoch": 3.043304463690873,
+      "grad_norm": 0.5333130750896956,
+      "learning_rate": 1.6416890661309098e-06,
+      "loss": 0.0345,
+      "step": 6852
+    },
+    {
+      "epoch": 3.04374861203642,
+      "grad_norm": 0.32916273546434566,
+      "learning_rate": 1.6402531389901894e-06,
+      "loss": 0.0196,
+      "step": 6853
+    },
+    {
+      "epoch": 3.0441927603819674,
+      "grad_norm": 0.4575672416898921,
+      "learning_rate": 1.6388177168764919e-06,
+      "loss": 0.0316,
+      "step": 6854
+    },
+    {
+      "epoch": 3.044636908727515,
+      "grad_norm": 0.4231274981211455,
+      "learning_rate": 1.6373828000055886e-06,
+      "loss": 0.0267,
+      "step": 6855
+    },
+    {
+      "epoch": 3.0450810570730624,
+      "grad_norm": 0.35267922234888455,
+      "learning_rate": 1.6359483885931709e-06,
+      "loss": 0.0224,
+      "step": 6856
+    },
+    {
+      "epoch": 3.0455252054186097,
+      "grad_norm": 0.4270485321761482,
+      "learning_rate": 1.634514482854856e-06,
+      "loss": 0.0215,
+      "step": 6857
+    },
+    {
+      "epoch": 3.045969353764157,
+      "grad_norm": 0.45838498705619496,
+      "learning_rate": 1.6330810830061833e-06,
+      "loss": 0.0259,
+      "step": 6858
+    },
+    {
+      "epoch": 3.0464135021097047,
+      "grad_norm": 0.39622697822045655,
+      "learning_rate": 1.6316481892626202e-06,
+      "loss": 0.0227,
+      "step": 6859
+    },
+    {
+      "epoch": 3.046857650455252,
+      "grad_norm": 0.4767421948555728,
+      "learning_rate": 1.6302158018395504e-06,
+      "loss": 0.021,
+      "step": 6860
+    },
+    {
+      "epoch": 3.0473017988007993,
+      "grad_norm": 0.4108572908804955,
+      "learning_rate": 1.6287839209522883e-06,
+      "loss": 0.0243,
+      "step": 6861
+    },
+    {
+      "epoch": 3.047745947146347,
+      "grad_norm": 0.3361055439360816,
+      "learning_rate": 1.62735254681607e-06,
+      "loss": 0.016,
+      "step": 6862
+    },
+    {
+      "epoch": 3.0481900954918943,
+      "grad_norm": 0.4890820081997742,
+      "learning_rate": 1.6259216796460553e-06,
+      "loss": 0.0332,
+      "step": 6863
+    },
+    {
+      "epoch": 3.0486342438374416,
+      "grad_norm": 0.3446651596488856,
+      "learning_rate": 1.6244913196573291e-06,
+      "loss": 0.0231,
+      "step": 6864
+    },
+    {
+      "epoch": 3.049078392182989,
+      "grad_norm": 0.4912264351287514,
+      "learning_rate": 1.623061467064896e-06,
+      "loss": 0.0272,
+      "step": 6865
+    },
+    {
+      "epoch": 3.0495225405285367,
+      "grad_norm": 0.4925885654573531,
+      "learning_rate": 1.6216321220836885e-06,
+      "loss": 0.0248,
+      "step": 6866
+    },
+    {
+      "epoch": 3.049966688874084,
+      "grad_norm": 0.3824243589770029,
+      "learning_rate": 1.6202032849285626e-06,
+      "loss": 0.0225,
+      "step": 6867
+    },
+    {
+      "epoch": 3.0504108372196312,
+      "grad_norm": 0.36019367540666763,
+      "learning_rate": 1.618774955814293e-06,
+      "loss": 0.0179,
+      "step": 6868
+    },
+    {
+      "epoch": 3.050854985565179,
+      "grad_norm": 0.5143104427412662,
+      "learning_rate": 1.6173471349555858e-06,
+      "loss": 0.0376,
+      "step": 6869
+    },
+    {
+      "epoch": 3.0512991339107263,
+      "grad_norm": 0.5934473311861377,
+      "learning_rate": 1.6159198225670676e-06,
+      "loss": 0.0448,
+      "step": 6870
+    },
+    {
+      "epoch": 3.0517432822562736,
+      "grad_norm": 0.45164863726044957,
+      "learning_rate": 1.6144930188632835e-06,
+      "loss": 0.0278,
+      "step": 6871
+    },
+    {
+      "epoch": 3.052187430601821,
+      "grad_norm": 0.5798490797388325,
+      "learning_rate": 1.6130667240587083e-06,
+      "loss": 0.0297,
+      "step": 6872
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.38501590071391395,
+      "learning_rate": 1.6116409383677383e-06,
+      "loss": 0.0244,
+      "step": 6873
+    },
+    {
+      "epoch": 3.053075727292916,
+      "grad_norm": 0.4626822552615348,
+      "learning_rate": 1.6102156620046937e-06,
+      "loss": 0.0234,
+      "step": 6874
+    },
+    {
+      "epoch": 3.053519875638463,
+      "grad_norm": 0.4709097565849538,
+      "learning_rate": 1.6087908951838193e-06,
+      "loss": 0.0325,
+      "step": 6875
+    },
+    {
+      "epoch": 3.0539640239840105,
+      "grad_norm": 0.38512614079510565,
+      "learning_rate": 1.6073666381192777e-06,
+      "loss": 0.0185,
+      "step": 6876
+    },
+    {
+      "epoch": 3.054408172329558,
+      "grad_norm": 0.3981638223693639,
+      "learning_rate": 1.6059428910251617e-06,
+      "loss": 0.0297,
+      "step": 6877
+    },
+    {
+      "epoch": 3.0548523206751055,
+      "grad_norm": 0.40972645998312435,
+      "learning_rate": 1.604519654115484e-06,
+      "loss": 0.0217,
+      "step": 6878
+    },
+    {
+      "epoch": 3.0552964690206528,
+      "grad_norm": 0.44788376402765917,
+      "learning_rate": 1.6030969276041813e-06,
+      "loss": 0.0255,
+      "step": 6879
+    },
+    {
+      "epoch": 3.0557406173662005,
+      "grad_norm": 0.45493435461264287,
+      "learning_rate": 1.6016747117051135e-06,
+      "loss": 0.0367,
+      "step": 6880
+    },
+    {
+      "epoch": 3.056184765711748,
+      "grad_norm": 0.5367527315236188,
+      "learning_rate": 1.6002530066320659e-06,
+      "loss": 0.0386,
+      "step": 6881
+    },
+    {
+      "epoch": 3.056628914057295,
+      "grad_norm": 0.36716772288492844,
+      "learning_rate": 1.5988318125987412e-06,
+      "loss": 0.0243,
+      "step": 6882
+    },
+    {
+      "epoch": 3.0570730624028424,
+      "grad_norm": 0.39700109646152437,
+      "learning_rate": 1.597411129818771e-06,
+      "loss": 0.0331,
+      "step": 6883
+    },
+    {
+      "epoch": 3.05751721074839,
+      "grad_norm": 0.4200578523175907,
+      "learning_rate": 1.5959909585057099e-06,
+      "loss": 0.0254,
+      "step": 6884
+    },
+    {
+      "epoch": 3.0579613590939374,
+      "grad_norm": 0.5911121451638379,
+      "learning_rate": 1.5945712988730278e-06,
+      "loss": 0.0418,
+      "step": 6885
+    },
+    {
+      "epoch": 3.0584055074394847,
+      "grad_norm": 0.3316421396306147,
+      "learning_rate": 1.5931521511341292e-06,
+      "loss": 0.0266,
+      "step": 6886
+    },
+    {
+      "epoch": 3.058849655785032,
+      "grad_norm": 0.38444781154877616,
+      "learning_rate": 1.5917335155023368e-06,
+      "loss": 0.018,
+      "step": 6887
+    },
+    {
+      "epoch": 3.0592938041305797,
+      "grad_norm": 0.38538195654260493,
+      "learning_rate": 1.590315392190891e-06,
+      "loss": 0.0223,
+      "step": 6888
+    },
+    {
+      "epoch": 3.059737952476127,
+      "grad_norm": 0.3188424830635555,
+      "learning_rate": 1.5888977814129625e-06,
+      "loss": 0.0162,
+      "step": 6889
+    },
+    {
+      "epoch": 3.0601821008216743,
+      "grad_norm": 0.5604048534911154,
+      "learning_rate": 1.5874806833816436e-06,
+      "loss": 0.0451,
+      "step": 6890
+    },
+    {
+      "epoch": 3.060626249167222,
+      "grad_norm": 0.4396556865933855,
+      "learning_rate": 1.5860640983099435e-06,
+      "loss": 0.027,
+      "step": 6891
+    },
+    {
+      "epoch": 3.0610703975127693,
+      "grad_norm": 0.5338475568401356,
+      "learning_rate": 1.584648026410805e-06,
+      "loss": 0.0428,
+      "step": 6892
+    },
+    {
+      "epoch": 3.0615145458583166,
+      "grad_norm": 0.41364738868710765,
+      "learning_rate": 1.583232467897083e-06,
+      "loss": 0.0255,
+      "step": 6893
+    },
+    {
+      "epoch": 3.061958694203864,
+      "grad_norm": 0.44128675402771145,
+      "learning_rate": 1.581817422981562e-06,
+      "loss": 0.0266,
+      "step": 6894
+    },
+    {
+      "epoch": 3.0624028425494116,
+      "grad_norm": 0.40811877248427353,
+      "learning_rate": 1.5804028918769488e-06,
+      "loss": 0.0202,
+      "step": 6895
+    },
+    {
+      "epoch": 3.062846990894959,
+      "grad_norm": 0.38420681148774133,
+      "learning_rate": 1.5789888747958666e-06,
+      "loss": 0.0299,
+      "step": 6896
+    },
+    {
+      "epoch": 3.0632911392405062,
+      "grad_norm": 0.3556137019009215,
+      "learning_rate": 1.5775753719508708e-06,
+      "loss": 0.021,
+      "step": 6897
+    },
+    {
+      "epoch": 3.063735287586054,
+      "grad_norm": 0.2882748081505788,
+      "learning_rate": 1.5761623835544348e-06,
+      "loss": 0.0258,
+      "step": 6898
+    },
+    {
+      "epoch": 3.0641794359316012,
+      "grad_norm": 0.3787886197229445,
+      "learning_rate": 1.5747499098189524e-06,
+      "loss": 0.0281,
+      "step": 6899
+    },
+    {
+      "epoch": 3.0646235842771485,
+      "grad_norm": 0.3354175757689207,
+      "learning_rate": 1.5733379509567426e-06,
+      "loss": 0.0234,
+      "step": 6900
+    },
+    {
+      "epoch": 3.065067732622696,
+      "grad_norm": 0.3209071922045546,
+      "learning_rate": 1.5719265071800498e-06,
+      "loss": 0.0182,
+      "step": 6901
+    },
+    {
+      "epoch": 3.0655118809682436,
+      "grad_norm": 0.43501967745626874,
+      "learning_rate": 1.5705155787010324e-06,
+      "loss": 0.0353,
+      "step": 6902
+    },
+    {
+      "epoch": 3.065956029313791,
+      "grad_norm": 0.4929319158649113,
+      "learning_rate": 1.5691051657317835e-06,
+      "loss": 0.0271,
+      "step": 6903
+    },
+    {
+      "epoch": 3.066400177659338,
+      "grad_norm": 0.3584810561885387,
+      "learning_rate": 1.5676952684843072e-06,
+      "loss": 0.0258,
+      "step": 6904
+    },
+    {
+      "epoch": 3.0668443260048854,
+      "grad_norm": 0.4638657269405188,
+      "learning_rate": 1.5662858871705366e-06,
+      "loss": 0.0236,
+      "step": 6905
+    },
+    {
+      "epoch": 3.067288474350433,
+      "grad_norm": 0.5065060680688717,
+      "learning_rate": 1.5648770220023263e-06,
+      "loss": 0.0287,
+      "step": 6906
+    },
+    {
+      "epoch": 3.0677326226959805,
+      "grad_norm": 0.4859275697928332,
+      "learning_rate": 1.5634686731914533e-06,
+      "loss": 0.0434,
+      "step": 6907
+    },
+    {
+      "epoch": 3.0681767710415278,
+      "grad_norm": 0.38551762262809786,
+      "learning_rate": 1.562060840949612e-06,
+      "loss": 0.0265,
+      "step": 6908
+    },
+    {
+      "epoch": 3.0686209193870755,
+      "grad_norm": 0.6596634602131578,
+      "learning_rate": 1.5606535254884297e-06,
+      "loss": 0.0343,
+      "step": 6909
+    },
+    {
+      "epoch": 3.0690650677326228,
+      "grad_norm": 0.40308217569658306,
+      "learning_rate": 1.5592467270194456e-06,
+      "loss": 0.0306,
+      "step": 6910
+    },
+    {
+      "epoch": 3.06950921607817,
+      "grad_norm": 0.6298585513540671,
+      "learning_rate": 1.5578404457541264e-06,
+      "loss": 0.0276,
+      "step": 6911
+    },
+    {
+      "epoch": 3.0699533644237174,
+      "grad_norm": 0.6519358046715888,
+      "learning_rate": 1.5564346819038616e-06,
+      "loss": 0.0335,
+      "step": 6912
+    },
+    {
+      "epoch": 3.070397512769265,
+      "grad_norm": 0.4424893479760814,
+      "learning_rate": 1.5550294356799573e-06,
+      "loss": 0.0249,
+      "step": 6913
+    },
+    {
+      "epoch": 3.0708416611148124,
+      "grad_norm": 0.44065297318246016,
+      "learning_rate": 1.55362470729365e-06,
+      "loss": 0.031,
+      "step": 6914
+    },
+    {
+      "epoch": 3.0712858094603597,
+      "grad_norm": 0.3171014816611921,
+      "learning_rate": 1.5522204969560945e-06,
+      "loss": 0.0228,
+      "step": 6915
+    },
+    {
+      "epoch": 3.071729957805907,
+      "grad_norm": 0.48982852285102907,
+      "learning_rate": 1.5508168048783645e-06,
+      "loss": 0.0243,
+      "step": 6916
+    },
+    {
+      "epoch": 3.0721741061514547,
+      "grad_norm": 0.36638227513318433,
+      "learning_rate": 1.5494136312714598e-06,
+      "loss": 0.0209,
+      "step": 6917
+    },
+    {
+      "epoch": 3.072618254497002,
+      "grad_norm": 0.40243990074588076,
+      "learning_rate": 1.5480109763463031e-06,
+      "loss": 0.0323,
+      "step": 6918
+    },
+    {
+      "epoch": 3.0730624028425493,
+      "grad_norm": 0.6138342320939462,
+      "learning_rate": 1.5466088403137326e-06,
+      "loss": 0.0303,
+      "step": 6919
+    },
+    {
+      "epoch": 3.073506551188097,
+      "grad_norm": 0.41638032443840284,
+      "learning_rate": 1.5452072233845194e-06,
+      "loss": 0.0302,
+      "step": 6920
+    },
+    {
+      "epoch": 3.0739506995336443,
+      "grad_norm": 0.3958771196139635,
+      "learning_rate": 1.5438061257693459e-06,
+      "loss": 0.025,
+      "step": 6921
+    },
+    {
+      "epoch": 3.0743948478791916,
+      "grad_norm": 0.409201101395247,
+      "learning_rate": 1.5424055476788219e-06,
+      "loss": 0.0281,
+      "step": 6922
+    },
+    {
+      "epoch": 3.074838996224739,
+      "grad_norm": 0.36270140865155653,
+      "learning_rate": 1.54100548932348e-06,
+      "loss": 0.0211,
+      "step": 6923
+    },
+    {
+      "epoch": 3.0752831445702866,
+      "grad_norm": 0.3939239641466573,
+      "learning_rate": 1.5396059509137694e-06,
+      "loss": 0.0221,
+      "step": 6924
+    },
+    {
+      "epoch": 3.075727292915834,
+      "grad_norm": 0.3910124202096161,
+      "learning_rate": 1.5382069326600645e-06,
+      "loss": 0.0247,
+      "step": 6925
+    },
+    {
+      "epoch": 3.076171441261381,
+      "grad_norm": 0.4045068362861821,
+      "learning_rate": 1.536808434772667e-06,
+      "loss": 0.0252,
+      "step": 6926
+    },
+    {
+      "epoch": 3.076615589606929,
+      "grad_norm": 0.48842679698944363,
+      "learning_rate": 1.5354104574617889e-06,
+      "loss": 0.0272,
+      "step": 6927
+    },
+    {
+      "epoch": 3.0770597379524762,
+      "grad_norm": 0.5568048405499652,
+      "learning_rate": 1.5340130009375725e-06,
+      "loss": 0.0271,
+      "step": 6928
+    },
+    {
+      "epoch": 3.0775038862980235,
+      "grad_norm": 0.35344319652310796,
+      "learning_rate": 1.5326160654100803e-06,
+      "loss": 0.0203,
+      "step": 6929
+    },
+    {
+      "epoch": 3.077948034643571,
+      "grad_norm": 0.41134278821618964,
+      "learning_rate": 1.5312196510892907e-06,
+      "loss": 0.0203,
+      "step": 6930
+    },
+    {
+      "epoch": 3.0783921829891185,
+      "grad_norm": 0.43439941775477803,
+      "learning_rate": 1.529823758185115e-06,
+      "loss": 0.0307,
+      "step": 6931
+    },
+    {
+      "epoch": 3.078836331334666,
+      "grad_norm": 0.36710103440734276,
+      "learning_rate": 1.5284283869073753e-06,
+      "loss": 0.0226,
+      "step": 6932
+    },
+    {
+      "epoch": 3.079280479680213,
+      "grad_norm": 0.3795190028300577,
+      "learning_rate": 1.5270335374658202e-06,
+      "loss": 0.0193,
+      "step": 6933
+    },
+    {
+      "epoch": 3.0797246280257604,
+      "grad_norm": 0.416086597648837,
+      "learning_rate": 1.5256392100701201e-06,
+      "loss": 0.0341,
+      "step": 6934
+    },
+    {
+      "epoch": 3.080168776371308,
+      "grad_norm": 0.47903570547844193,
+      "learning_rate": 1.5242454049298672e-06,
+      "loss": 0.0326,
+      "step": 6935
+    },
+    {
+      "epoch": 3.0806129247168554,
+      "grad_norm": 0.41924576417637155,
+      "learning_rate": 1.5228521222545694e-06,
+      "loss": 0.0232,
+      "step": 6936
+    },
+    {
+      "epoch": 3.0810570730624027,
+      "grad_norm": 0.4086739351412735,
+      "learning_rate": 1.5214593622536677e-06,
+      "loss": 0.0191,
+      "step": 6937
+    },
+    {
+      "epoch": 3.0815012214079505,
+      "grad_norm": 0.42090655404898863,
+      "learning_rate": 1.5200671251365118e-06,
+      "loss": 0.0234,
+      "step": 6938
+    },
+    {
+      "epoch": 3.0819453697534978,
+      "grad_norm": 0.4491694516637582,
+      "learning_rate": 1.5186754111123814e-06,
+      "loss": 0.0309,
+      "step": 6939
+    },
+    {
+      "epoch": 3.082389518099045,
+      "grad_norm": 0.3593852978556601,
+      "learning_rate": 1.5172842203904752e-06,
+      "loss": 0.0219,
+      "step": 6940
+    },
+    {
+      "epoch": 3.0828336664445923,
+      "grad_norm": 0.4142188627731481,
+      "learning_rate": 1.5158935531799102e-06,
+      "loss": 0.0231,
+      "step": 6941
+    },
+    {
+      "epoch": 3.08327781479014,
+      "grad_norm": 0.3433910778382263,
+      "learning_rate": 1.5145034096897271e-06,
+      "loss": 0.0232,
+      "step": 6942
+    },
+    {
+      "epoch": 3.0837219631356874,
+      "grad_norm": 0.42040065985676595,
+      "learning_rate": 1.5131137901288928e-06,
+      "loss": 0.0274,
+      "step": 6943
+    },
+    {
+      "epoch": 3.0841661114812347,
+      "grad_norm": 0.37811344114245793,
+      "learning_rate": 1.5117246947062864e-06,
+      "loss": 0.0278,
+      "step": 6944
+    },
+    {
+      "epoch": 3.084610259826782,
+      "grad_norm": 0.6332645566079246,
+      "learning_rate": 1.5103361236307135e-06,
+      "loss": 0.0199,
+      "step": 6945
+    },
+    {
+      "epoch": 3.0850544081723297,
+      "grad_norm": 0.3367889244804879,
+      "learning_rate": 1.5089480771109021e-06,
+      "loss": 0.0221,
+      "step": 6946
+    },
+    {
+      "epoch": 3.085498556517877,
+      "grad_norm": 0.45979021485293203,
+      "learning_rate": 1.507560555355494e-06,
+      "loss": 0.0259,
+      "step": 6947
+    },
+    {
+      "epoch": 3.0859427048634243,
+      "grad_norm": 0.3846307657242651,
+      "learning_rate": 1.5061735585730636e-06,
+      "loss": 0.0252,
+      "step": 6948
+    },
+    {
+      "epoch": 3.086386853208972,
+      "grad_norm": 0.36534660664962226,
+      "learning_rate": 1.504787086972096e-06,
+      "loss": 0.0173,
+      "step": 6949
+    },
+    {
+      "epoch": 3.0868310015545193,
+      "grad_norm": 0.6011247419581762,
+      "learning_rate": 1.5034011407610021e-06,
+      "loss": 0.0399,
+      "step": 6950
+    },
+    {
+      "epoch": 3.0872751499000666,
+      "grad_norm": 0.4220887506749395,
+      "learning_rate": 1.502015720148115e-06,
+      "loss": 0.0265,
+      "step": 6951
+    },
+    {
+      "epoch": 3.087719298245614,
+      "grad_norm": 0.383622304930661,
+      "learning_rate": 1.5006308253416846e-06,
+      "loss": 0.0235,
+      "step": 6952
+    },
+    {
+      "epoch": 3.0881634465911616,
+      "grad_norm": 0.48234757617698376,
+      "learning_rate": 1.4992464565498831e-06,
+      "loss": 0.0354,
+      "step": 6953
+    },
+    {
+      "epoch": 3.088607594936709,
+      "grad_norm": 0.4645377835459482,
+      "learning_rate": 1.4978626139808094e-06,
+      "loss": 0.0332,
+      "step": 6954
+    },
+    {
+      "epoch": 3.089051743282256,
+      "grad_norm": 0.4061976190338302,
+      "learning_rate": 1.4964792978424746e-06,
+      "loss": 0.0231,
+      "step": 6955
+    },
+    {
+      "epoch": 3.089495891627804,
+      "grad_norm": 0.5482133012861865,
+      "learning_rate": 1.495096508342816e-06,
+      "loss": 0.0369,
+      "step": 6956
+    },
+    {
+      "epoch": 3.089940039973351,
+      "grad_norm": 0.3661650923331457,
+      "learning_rate": 1.4937142456896907e-06,
+      "loss": 0.0231,
+      "step": 6957
+    },
+    {
+      "epoch": 3.0903841883188985,
+      "grad_norm": 0.4124097316559942,
+      "learning_rate": 1.4923325100908749e-06,
+      "loss": 0.026,
+      "step": 6958
+    },
+    {
+      "epoch": 3.090828336664446,
+      "grad_norm": 0.5506003037235153,
+      "learning_rate": 1.490951301754066e-06,
+      "loss": 0.0272,
+      "step": 6959
+    },
+    {
+      "epoch": 3.0912724850099935,
+      "grad_norm": 0.39939980904986866,
+      "learning_rate": 1.4895706208868876e-06,
+      "loss": 0.028,
+      "step": 6960
+    },
+    {
+      "epoch": 3.091716633355541,
+      "grad_norm": 0.46904750344961466,
+      "learning_rate": 1.4881904676968756e-06,
+      "loss": 0.0258,
+      "step": 6961
+    },
+    {
+      "epoch": 3.092160781701088,
+      "grad_norm": 0.46041949493275547,
+      "learning_rate": 1.4868108423914913e-06,
+      "loss": 0.0284,
+      "step": 6962
+    },
+    {
+      "epoch": 3.0926049300466354,
+      "grad_norm": 0.3681186282877754,
+      "learning_rate": 1.4854317451781175e-06,
+      "loss": 0.0232,
+      "step": 6963
+    },
+    {
+      "epoch": 3.093049078392183,
+      "grad_norm": 0.4723461677240559,
+      "learning_rate": 1.4840531762640524e-06,
+      "loss": 0.0277,
+      "step": 6964
+    },
+    {
+      "epoch": 3.0934932267377304,
+      "grad_norm": 0.38175424189288976,
+      "learning_rate": 1.4826751358565211e-06,
+      "loss": 0.019,
+      "step": 6965
+    },
+    {
+      "epoch": 3.0939373750832777,
+      "grad_norm": 0.41086502374434375,
+      "learning_rate": 1.4812976241626659e-06,
+      "loss": 0.0287,
+      "step": 6966
+    },
+    {
+      "epoch": 3.0943815234288254,
+      "grad_norm": 0.39348852384191785,
+      "learning_rate": 1.4799206413895494e-06,
+      "loss": 0.026,
+      "step": 6967
+    },
+    {
+      "epoch": 3.0948256717743727,
+      "grad_norm": 0.4897251060835583,
+      "learning_rate": 1.4785441877441587e-06,
+      "loss": 0.0283,
+      "step": 6968
+    },
+    {
+      "epoch": 3.09526982011992,
+      "grad_norm": 0.396902884934688,
+      "learning_rate": 1.4771682634333933e-06,
+      "loss": 0.0241,
+      "step": 6969
+    },
+    {
+      "epoch": 3.0957139684654673,
+      "grad_norm": 0.6024718135186303,
+      "learning_rate": 1.4757928686640788e-06,
+      "loss": 0.0324,
+      "step": 6970
+    },
+    {
+      "epoch": 3.096158116811015,
+      "grad_norm": 0.3483981306238304,
+      "learning_rate": 1.4744180036429656e-06,
+      "loss": 0.0203,
+      "step": 6971
+    },
+    {
+      "epoch": 3.0966022651565623,
+      "grad_norm": 0.45268776356014967,
+      "learning_rate": 1.4730436685767135e-06,
+      "loss": 0.0302,
+      "step": 6972
+    },
+    {
+      "epoch": 3.0970464135021096,
+      "grad_norm": 0.5030795577403537,
+      "learning_rate": 1.4716698636719107e-06,
+      "loss": 0.0291,
+      "step": 6973
+    },
+    {
+      "epoch": 3.097490561847657,
+      "grad_norm": 0.4396401120191148,
+      "learning_rate": 1.470296589135065e-06,
+      "loss": 0.0342,
+      "step": 6974
+    },
+    {
+      "epoch": 3.0979347101932047,
+      "grad_norm": 0.3979132301675201,
+      "learning_rate": 1.4689238451725995e-06,
+      "loss": 0.0201,
+      "step": 6975
+    },
+    {
+      "epoch": 3.098378858538752,
+      "grad_norm": 0.3787151125083963,
+      "learning_rate": 1.4675516319908629e-06,
+      "loss": 0.0257,
+      "step": 6976
+    },
+    {
+      "epoch": 3.0988230068842992,
+      "grad_norm": 0.37709012703404926,
+      "learning_rate": 1.466179949796121e-06,
+      "loss": 0.031,
+      "step": 6977
+    },
+    {
+      "epoch": 3.099267155229847,
+      "grad_norm": 0.46313083842641917,
+      "learning_rate": 1.4648087987945625e-06,
+      "loss": 0.0187,
+      "step": 6978
+    },
+    {
+      "epoch": 3.0997113035753943,
+      "grad_norm": 0.41171533054686754,
+      "learning_rate": 1.4634381791922936e-06,
+      "loss": 0.0265,
+      "step": 6979
+    },
+    {
+      "epoch": 3.1001554519209416,
+      "grad_norm": 0.35930958286366377,
+      "learning_rate": 1.4620680911953433e-06,
+      "loss": 0.0287,
+      "step": 6980
+    },
+    {
+      "epoch": 3.100599600266489,
+      "grad_norm": 0.5934414847112055,
+      "learning_rate": 1.460698535009657e-06,
+      "loss": 0.021,
+      "step": 6981
+    },
+    {
+      "epoch": 3.1010437486120366,
+      "grad_norm": 0.4183538172458836,
+      "learning_rate": 1.4593295108411027e-06,
+      "loss": 0.0239,
+      "step": 6982
+    },
+    {
+      "epoch": 3.101487896957584,
+      "grad_norm": 0.3854427898688108,
+      "learning_rate": 1.4579610188954685e-06,
+      "loss": 0.0274,
+      "step": 6983
+    },
+    {
+      "epoch": 3.101932045303131,
+      "grad_norm": 0.4711207189434762,
+      "learning_rate": 1.4565930593784616e-06,
+      "loss": 0.0282,
+      "step": 6984
+    },
+    {
+      "epoch": 3.1023761936486784,
+      "grad_norm": 0.4616599969953312,
+      "learning_rate": 1.455225632495712e-06,
+      "loss": 0.029,
+      "step": 6985
+    },
+    {
+      "epoch": 3.102820341994226,
+      "grad_norm": 0.4076291882386512,
+      "learning_rate": 1.453858738452763e-06,
+      "loss": 0.0326,
+      "step": 6986
+    },
+    {
+      "epoch": 3.1032644903397735,
+      "grad_norm": 0.31683467511767754,
+      "learning_rate": 1.4524923774550825e-06,
+      "loss": 0.0176,
+      "step": 6987
+    },
+    {
+      "epoch": 3.1037086386853208,
+      "grad_norm": 0.39424949712455315,
+      "learning_rate": 1.4511265497080624e-06,
+      "loss": 0.0302,
+      "step": 6988
+    },
+    {
+      "epoch": 3.1041527870308685,
+      "grad_norm": 0.4103257196866521,
+      "learning_rate": 1.4497612554170054e-06,
+      "loss": 0.0273,
+      "step": 6989
+    },
+    {
+      "epoch": 3.104596935376416,
+      "grad_norm": 0.39987357613853897,
+      "learning_rate": 1.4483964947871392e-06,
+      "loss": 0.0255,
+      "step": 6990
+    },
+    {
+      "epoch": 3.105041083721963,
+      "grad_norm": 0.41100905630593115,
+      "learning_rate": 1.4470322680236132e-06,
+      "loss": 0.0302,
+      "step": 6991
+    },
+    {
+      "epoch": 3.1054852320675104,
+      "grad_norm": 0.4415871306778436,
+      "learning_rate": 1.4456685753314898e-06,
+      "loss": 0.0248,
+      "step": 6992
+    },
+    {
+      "epoch": 3.105929380413058,
+      "grad_norm": 0.3985527776293508,
+      "learning_rate": 1.4443054169157566e-06,
+      "loss": 0.024,
+      "step": 6993
+    },
+    {
+      "epoch": 3.1063735287586054,
+      "grad_norm": 0.5025368665991552,
+      "learning_rate": 1.4429427929813205e-06,
+      "loss": 0.0327,
+      "step": 6994
+    },
+    {
+      "epoch": 3.1068176771041527,
+      "grad_norm": 0.543210885126374,
+      "learning_rate": 1.4415807037330065e-06,
+      "loss": 0.0244,
+      "step": 6995
+    },
+    {
+      "epoch": 3.1072618254497,
+      "grad_norm": 0.5795921769157866,
+      "learning_rate": 1.4402191493755614e-06,
+      "loss": 0.0319,
+      "step": 6996
+    },
+    {
+      "epoch": 3.1077059737952477,
+      "grad_norm": 0.33181453971757086,
+      "learning_rate": 1.4388581301136463e-06,
+      "loss": 0.0239,
+      "step": 6997
+    },
+    {
+      "epoch": 3.108150122140795,
+      "grad_norm": 0.35538456672100055,
+      "learning_rate": 1.4374976461518475e-06,
+      "loss": 0.0202,
+      "step": 6998
+    },
+    {
+      "epoch": 3.1085942704863423,
+      "grad_norm": 0.30394370356689626,
+      "learning_rate": 1.436137697694669e-06,
+      "loss": 0.0148,
+      "step": 6999
+    },
+    {
+      "epoch": 3.10903841883189,
+      "grad_norm": 0.5358782537926752,
+      "learning_rate": 1.4347782849465335e-06,
+      "loss": 0.0361,
+      "step": 7000
+    },
+    {
+      "epoch": 3.1094825671774373,
+      "grad_norm": 0.45423657841045967,
+      "learning_rate": 1.4334194081117853e-06,
+      "loss": 0.0296,
+      "step": 7001
+    },
+    {
+      "epoch": 3.1099267155229846,
+      "grad_norm": 0.391451201654507,
+      "learning_rate": 1.4320610673946862e-06,
+      "loss": 0.0219,
+      "step": 7002
+    },
+    {
+      "epoch": 3.110370863868532,
+      "grad_norm": 0.3695407430957425,
+      "learning_rate": 1.4307032629994162e-06,
+      "loss": 0.0275,
+      "step": 7003
+    },
+    {
+      "epoch": 3.1108150122140796,
+      "grad_norm": 0.3522320469665786,
+      "learning_rate": 1.4293459951300775e-06,
+      "loss": 0.0191,
+      "step": 7004
+    },
+    {
+      "epoch": 3.111259160559627,
+      "grad_norm": 0.39441008031342695,
+      "learning_rate": 1.4279892639906906e-06,
+      "loss": 0.022,
+      "step": 7005
+    },
+    {
+      "epoch": 3.111703308905174,
+      "grad_norm": 0.3604951087918798,
+      "learning_rate": 1.4266330697851955e-06,
+      "loss": 0.0291,
+      "step": 7006
+    },
+    {
+      "epoch": 3.112147457250722,
+      "grad_norm": 0.3966852264525251,
+      "learning_rate": 1.4252774127174502e-06,
+      "loss": 0.0336,
+      "step": 7007
+    },
+    {
+      "epoch": 3.1125916055962692,
+      "grad_norm": 0.4920706934740084,
+      "learning_rate": 1.4239222929912354e-06,
+      "loss": 0.0339,
+      "step": 7008
+    },
+    {
+      "epoch": 3.1130357539418165,
+      "grad_norm": 0.452968414032015,
+      "learning_rate": 1.422567710810246e-06,
+      "loss": 0.0334,
+      "step": 7009
+    },
+    {
+      "epoch": 3.113479902287364,
+      "grad_norm": 0.34946018464557227,
+      "learning_rate": 1.421213666378099e-06,
+      "loss": 0.0259,
+      "step": 7010
+    },
+    {
+      "epoch": 3.1139240506329116,
+      "grad_norm": 0.3785739486603564,
+      "learning_rate": 1.419860159898331e-06,
+      "loss": 0.0248,
+      "step": 7011
+    },
+    {
+      "epoch": 3.114368198978459,
+      "grad_norm": 0.48985400850238237,
+      "learning_rate": 1.418507191574397e-06,
+      "loss": 0.0245,
+      "step": 7012
+    },
+    {
+      "epoch": 3.114812347324006,
+      "grad_norm": 0.37256560600518596,
+      "learning_rate": 1.4171547616096726e-06,
+      "loss": 0.0263,
+      "step": 7013
+    },
+    {
+      "epoch": 3.1152564956695534,
+      "grad_norm": 0.4745149660869735,
+      "learning_rate": 1.4158028702074478e-06,
+      "loss": 0.0315,
+      "step": 7014
+    },
+    {
+      "epoch": 3.115700644015101,
+      "grad_norm": 0.4272234241381395,
+      "learning_rate": 1.4144515175709366e-06,
+      "loss": 0.0359,
+      "step": 7015
+    },
+    {
+      "epoch": 3.1161447923606485,
+      "grad_norm": 0.4717714680449159,
+      "learning_rate": 1.4131007039032702e-06,
+      "loss": 0.0247,
+      "step": 7016
+    },
+    {
+      "epoch": 3.1165889407061957,
+      "grad_norm": 0.3661175444111234,
+      "learning_rate": 1.4117504294074985e-06,
+      "loss": 0.0199,
+      "step": 7017
+    },
+    {
+      "epoch": 3.1170330890517435,
+      "grad_norm": 0.39156431702822175,
+      "learning_rate": 1.4104006942865911e-06,
+      "loss": 0.0225,
+      "step": 7018
+    },
+    {
+      "epoch": 3.1174772373972908,
+      "grad_norm": 0.4841602294941045,
+      "learning_rate": 1.4090514987434372e-06,
+      "loss": 0.0224,
+      "step": 7019
+    },
+    {
+      "epoch": 3.117921385742838,
+      "grad_norm": 0.3409883421268121,
+      "learning_rate": 1.4077028429808415e-06,
+      "loss": 0.0184,
+      "step": 7020
+    },
+    {
+      "epoch": 3.1183655340883853,
+      "grad_norm": 0.4571971842074867,
+      "learning_rate": 1.4063547272015305e-06,
+      "loss": 0.0334,
+      "step": 7021
+    },
+    {
+      "epoch": 3.118809682433933,
+      "grad_norm": 0.4859572155144659,
+      "learning_rate": 1.4050071516081499e-06,
+      "loss": 0.0236,
+      "step": 7022
+    },
+    {
+      "epoch": 3.1192538307794804,
+      "grad_norm": 0.36054003070051327,
+      "learning_rate": 1.4036601164032626e-06,
+      "loss": 0.0244,
+      "step": 7023
+    },
+    {
+      "epoch": 3.1196979791250277,
+      "grad_norm": 0.4164627314286096,
+      "learning_rate": 1.4023136217893518e-06,
+      "loss": 0.0276,
+      "step": 7024
+    },
+    {
+      "epoch": 3.120142127470575,
+      "grad_norm": 0.5468105573468193,
+      "learning_rate": 1.4009676679688167e-06,
+      "loss": 0.0344,
+      "step": 7025
+    },
+    {
+      "epoch": 3.1205862758161227,
+      "grad_norm": 0.5699602944440265,
+      "learning_rate": 1.399622255143978e-06,
+      "loss": 0.0227,
+      "step": 7026
+    },
+    {
+      "epoch": 3.12103042416167,
+      "grad_norm": 0.3763296846537518,
+      "learning_rate": 1.3982773835170738e-06,
+      "loss": 0.0233,
+      "step": 7027
+    },
+    {
+      "epoch": 3.1214745725072173,
+      "grad_norm": 0.3809901917507222,
+      "learning_rate": 1.396933053290262e-06,
+      "loss": 0.0261,
+      "step": 7028
+    },
+    {
+      "epoch": 3.121918720852765,
+      "grad_norm": 0.3942636268702358,
+      "learning_rate": 1.3955892646656172e-06,
+      "loss": 0.0244,
+      "step": 7029
+    },
+    {
+      "epoch": 3.1223628691983123,
+      "grad_norm": 0.4034606683568663,
+      "learning_rate": 1.3942460178451357e-06,
+      "loss": 0.0267,
+      "step": 7030
+    },
+    {
+      "epoch": 3.1228070175438596,
+      "grad_norm": 0.4219843417449712,
+      "learning_rate": 1.3929033130307273e-06,
+      "loss": 0.0291,
+      "step": 7031
+    },
+    {
+      "epoch": 3.123251165889407,
+      "grad_norm": 0.39268086756994935,
+      "learning_rate": 1.3915611504242248e-06,
+      "loss": 0.0242,
+      "step": 7032
+    },
+    {
+      "epoch": 3.1236953142349546,
+      "grad_norm": 0.41199677628842846,
+      "learning_rate": 1.390219530227378e-06,
+      "loss": 0.0284,
+      "step": 7033
+    },
+    {
+      "epoch": 3.124139462580502,
+      "grad_norm": 0.39148197640312005,
+      "learning_rate": 1.3888784526418552e-06,
+      "loss": 0.0268,
+      "step": 7034
+    },
+    {
+      "epoch": 3.124583610926049,
+      "grad_norm": 0.6462152373026261,
+      "learning_rate": 1.3875379178692433e-06,
+      "loss": 0.042,
+      "step": 7035
+    },
+    {
+      "epoch": 3.125027759271597,
+      "grad_norm": 0.5120068554721335,
+      "learning_rate": 1.3861979261110493e-06,
+      "loss": 0.0352,
+      "step": 7036
+    },
+    {
+      "epoch": 3.125471907617144,
+      "grad_norm": 0.5096494527748306,
+      "learning_rate": 1.3848584775686923e-06,
+      "loss": 0.0329,
+      "step": 7037
+    },
+    {
+      "epoch": 3.1259160559626915,
+      "grad_norm": 0.49019540876863293,
+      "learning_rate": 1.3835195724435175e-06,
+      "loss": 0.0274,
+      "step": 7038
+    },
+    {
+      "epoch": 3.126360204308239,
+      "grad_norm": 0.4378287851338807,
+      "learning_rate": 1.3821812109367838e-06,
+      "loss": 0.0261,
+      "step": 7039
+    },
+    {
+      "epoch": 3.1268043526537865,
+      "grad_norm": 0.4665401569139363,
+      "learning_rate": 1.38084339324967e-06,
+      "loss": 0.0244,
+      "step": 7040
+    },
+    {
+      "epoch": 3.127248500999334,
+      "grad_norm": 0.4839568704489146,
+      "learning_rate": 1.3795061195832749e-06,
+      "loss": 0.0185,
+      "step": 7041
+    },
+    {
+      "epoch": 3.127692649344881,
+      "grad_norm": 0.38633411841281007,
+      "learning_rate": 1.3781693901386094e-06,
+      "loss": 0.0206,
+      "step": 7042
+    },
+    {
+      "epoch": 3.1281367976904284,
+      "grad_norm": 0.3852498984003089,
+      "learning_rate": 1.3768332051166089e-06,
+      "loss": 0.023,
+      "step": 7043
+    },
+    {
+      "epoch": 3.128580946035976,
+      "grad_norm": 0.3527312929509418,
+      "learning_rate": 1.3754975647181245e-06,
+      "loss": 0.0272,
+      "step": 7044
+    },
+    {
+      "epoch": 3.1290250943815234,
+      "grad_norm": 0.4078162247196801,
+      "learning_rate": 1.374162469143926e-06,
+      "loss": 0.0184,
+      "step": 7045
+    },
+    {
+      "epoch": 3.1294692427270707,
+      "grad_norm": 0.6947202101134006,
+      "learning_rate": 1.3728279185947002e-06,
+      "loss": 0.0287,
+      "step": 7046
+    },
+    {
+      "epoch": 3.1299133910726185,
+      "grad_norm": 0.3650330793591096,
+      "learning_rate": 1.3714939132710547e-06,
+      "loss": 0.0203,
+      "step": 7047
+    },
+    {
+      "epoch": 3.1303575394181657,
+      "grad_norm": 0.3894198333673228,
+      "learning_rate": 1.3701604533735102e-06,
+      "loss": 0.0383,
+      "step": 7048
+    },
+    {
+      "epoch": 3.130801687763713,
+      "grad_norm": 0.3626357108946466,
+      "learning_rate": 1.3688275391025096e-06,
+      "loss": 0.019,
+      "step": 7049
+    },
+    {
+      "epoch": 3.1312458361092603,
+      "grad_norm": 0.421447173128161,
+      "learning_rate": 1.3674951706584134e-06,
+      "loss": 0.0182,
+      "step": 7050
+    },
+    {
+      "epoch": 3.131689984454808,
+      "grad_norm": 0.4627577920703621,
+      "learning_rate": 1.3661633482414977e-06,
+      "loss": 0.0265,
+      "step": 7051
+    },
+    {
+      "epoch": 3.1321341328003554,
+      "grad_norm": 0.43244182202980463,
+      "learning_rate": 1.3648320720519592e-06,
+      "loss": 0.0196,
+      "step": 7052
+    },
+    {
+      "epoch": 3.1325782811459026,
+      "grad_norm": 0.4592125815303857,
+      "learning_rate": 1.3635013422899124e-06,
+      "loss": 0.0298,
+      "step": 7053
+    },
+    {
+      "epoch": 3.13302242949145,
+      "grad_norm": 0.44770730468191844,
+      "learning_rate": 1.3621711591553854e-06,
+      "loss": 0.0347,
+      "step": 7054
+    },
+    {
+      "epoch": 3.1334665778369977,
+      "grad_norm": 0.4417156773267622,
+      "learning_rate": 1.3608415228483291e-06,
+      "loss": 0.0286,
+      "step": 7055
+    },
+    {
+      "epoch": 3.133910726182545,
+      "grad_norm": 0.33340946983600855,
+      "learning_rate": 1.3595124335686104e-06,
+      "loss": 0.0239,
+      "step": 7056
+    },
+    {
+      "epoch": 3.1343548745280922,
+      "grad_norm": 0.45018385116758614,
+      "learning_rate": 1.3581838915160145e-06,
+      "loss": 0.0307,
+      "step": 7057
+    },
+    {
+      "epoch": 3.13479902287364,
+      "grad_norm": 0.49692015647969867,
+      "learning_rate": 1.3568558968902445e-06,
+      "loss": 0.0291,
+      "step": 7058
+    },
+    {
+      "epoch": 3.1352431712191873,
+      "grad_norm": 0.43270232873861747,
+      "learning_rate": 1.3555284498909183e-06,
+      "loss": 0.031,
+      "step": 7059
+    },
+    {
+      "epoch": 3.1356873195647346,
+      "grad_norm": 0.40820462560013654,
+      "learning_rate": 1.3542015507175743e-06,
+      "loss": 0.0235,
+      "step": 7060
+    },
+    {
+      "epoch": 3.136131467910282,
+      "grad_norm": 0.4972923121321135,
+      "learning_rate": 1.3528751995696688e-06,
+      "loss": 0.051,
+      "step": 7061
+    },
+    {
+      "epoch": 3.1365756162558296,
+      "grad_norm": 0.41578377070138484,
+      "learning_rate": 1.3515493966465743e-06,
+      "loss": 0.028,
+      "step": 7062
+    },
+    {
+      "epoch": 3.137019764601377,
+      "grad_norm": 0.4491202264177913,
+      "learning_rate": 1.350224142147582e-06,
+      "loss": 0.0172,
+      "step": 7063
+    },
+    {
+      "epoch": 3.137463912946924,
+      "grad_norm": 0.4586826960329923,
+      "learning_rate": 1.3488994362719016e-06,
+      "loss": 0.0244,
+      "step": 7064
+    },
+    {
+      "epoch": 3.137908061292472,
+      "grad_norm": 0.3366156664630001,
+      "learning_rate": 1.3475752792186559e-06,
+      "loss": 0.0177,
+      "step": 7065
+    },
+    {
+      "epoch": 3.138352209638019,
+      "grad_norm": 0.4038403979608836,
+      "learning_rate": 1.3462516711868894e-06,
+      "loss": 0.0283,
+      "step": 7066
+    },
+    {
+      "epoch": 3.1387963579835665,
+      "grad_norm": 0.36057350319064324,
+      "learning_rate": 1.3449286123755628e-06,
+      "loss": 0.0282,
+      "step": 7067
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.4323404311148623,
+      "learning_rate": 1.343606102983555e-06,
+      "loss": 0.0285,
+      "step": 7068
+    },
+    {
+      "epoch": 3.1396846546746615,
+      "grad_norm": 0.42881049252651987,
+      "learning_rate": 1.3422841432096623e-06,
+      "loss": 0.0299,
+      "step": 7069
+    },
+    {
+      "epoch": 3.140128803020209,
+      "grad_norm": 0.2902742598750208,
+      "learning_rate": 1.3409627332525954e-06,
+      "loss": 0.0181,
+      "step": 7070
+    },
+    {
+      "epoch": 3.140572951365756,
+      "grad_norm": 0.4433992896224678,
+      "learning_rate": 1.3396418733109856e-06,
+      "loss": 0.0257,
+      "step": 7071
+    },
+    {
+      "epoch": 3.1410170997113034,
+      "grad_norm": 0.6894028385625418,
+      "learning_rate": 1.3383215635833829e-06,
+      "loss": 0.0304,
+      "step": 7072
+    },
+    {
+      "epoch": 3.141461248056851,
+      "grad_norm": 0.5245286511325312,
+      "learning_rate": 1.337001804268247e-06,
+      "loss": 0.0269,
+      "step": 7073
+    },
+    {
+      "epoch": 3.1419053964023984,
+      "grad_norm": 0.5000099376315333,
+      "learning_rate": 1.3356825955639645e-06,
+      "loss": 0.0308,
+      "step": 7074
+    },
+    {
+      "epoch": 3.1423495447479457,
+      "grad_norm": 0.6132659503324829,
+      "learning_rate": 1.3343639376688355e-06,
+      "loss": 0.032,
+      "step": 7075
+    },
+    {
+      "epoch": 3.1427936930934934,
+      "grad_norm": 0.4546873288387578,
+      "learning_rate": 1.3330458307810734e-06,
+      "loss": 0.0246,
+      "step": 7076
+    },
+    {
+      "epoch": 3.1432378414390407,
+      "grad_norm": 0.385185952241093,
+      "learning_rate": 1.3317282750988137e-06,
+      "loss": 0.0184,
+      "step": 7077
+    },
+    {
+      "epoch": 3.143681989784588,
+      "grad_norm": 0.39438378600816204,
+      "learning_rate": 1.3304112708201073e-06,
+      "loss": 0.0279,
+      "step": 7078
+    },
+    {
+      "epoch": 3.1441261381301353,
+      "grad_norm": 0.4806029181761233,
+      "learning_rate": 1.329094818142922e-06,
+      "loss": 0.0241,
+      "step": 7079
+    },
+    {
+      "epoch": 3.144570286475683,
+      "grad_norm": 0.5228838911668552,
+      "learning_rate": 1.327778917265144e-06,
+      "loss": 0.0274,
+      "step": 7080
+    },
+    {
+      "epoch": 3.1450144348212303,
+      "grad_norm": 0.46531223651019554,
+      "learning_rate": 1.3264635683845755e-06,
+      "loss": 0.0241,
+      "step": 7081
+    },
+    {
+      "epoch": 3.1454585831667776,
+      "grad_norm": 0.32902026350154623,
+      "learning_rate": 1.3251487716989341e-06,
+      "loss": 0.0169,
+      "step": 7082
+    },
+    {
+      "epoch": 3.145902731512325,
+      "grad_norm": 0.3895325085859838,
+      "learning_rate": 1.3238345274058572e-06,
+      "loss": 0.0219,
+      "step": 7083
+    },
+    {
+      "epoch": 3.1463468798578726,
+      "grad_norm": 0.4481405965406686,
+      "learning_rate": 1.322520835702898e-06,
+      "loss": 0.025,
+      "step": 7084
+    },
+    {
+      "epoch": 3.14679102820342,
+      "grad_norm": 0.5128203909715668,
+      "learning_rate": 1.3212076967875265e-06,
+      "loss": 0.0348,
+      "step": 7085
+    },
+    {
+      "epoch": 3.1472351765489672,
+      "grad_norm": 0.46447445611495575,
+      "learning_rate": 1.3198951108571312e-06,
+      "loss": 0.0313,
+      "step": 7086
+    },
+    {
+      "epoch": 3.147679324894515,
+      "grad_norm": 0.29983798874260054,
+      "learning_rate": 1.3185830781090136e-06,
+      "loss": 0.0142,
+      "step": 7087
+    },
+    {
+      "epoch": 3.1481234732400623,
+      "grad_norm": 0.39996284685169026,
+      "learning_rate": 1.3172715987403955e-06,
+      "loss": 0.0279,
+      "step": 7088
+    },
+    {
+      "epoch": 3.1485676215856095,
+      "grad_norm": 0.44271681407903213,
+      "learning_rate": 1.3159606729484165e-06,
+      "loss": 0.0309,
+      "step": 7089
+    },
+    {
+      "epoch": 3.149011769931157,
+      "grad_norm": 0.39836027232808446,
+      "learning_rate": 1.3146503009301258e-06,
+      "loss": 0.0258,
+      "step": 7090
+    },
+    {
+      "epoch": 3.1494559182767046,
+      "grad_norm": 0.38922372733595656,
+      "learning_rate": 1.3133404828824998e-06,
+      "loss": 0.0249,
+      "step": 7091
+    },
+    {
+      "epoch": 3.149900066622252,
+      "grad_norm": 0.4629939120495913,
+      "learning_rate": 1.3120312190024265e-06,
+      "loss": 0.0177,
+      "step": 7092
+    },
+    {
+      "epoch": 3.150344214967799,
+      "grad_norm": 0.5175669897186087,
+      "learning_rate": 1.3107225094867066e-06,
+      "loss": 0.0352,
+      "step": 7093
+    },
+    {
+      "epoch": 3.150788363313347,
+      "grad_norm": 0.4487683854548024,
+      "learning_rate": 1.3094143545320636e-06,
+      "loss": 0.0309,
+      "step": 7094
+    },
+    {
+      "epoch": 3.151232511658894,
+      "grad_norm": 0.4853986633083335,
+      "learning_rate": 1.3081067543351351e-06,
+      "loss": 0.0351,
+      "step": 7095
+    },
+    {
+      "epoch": 3.1516766600044415,
+      "grad_norm": 0.4346141380035006,
+      "learning_rate": 1.3067997090924755e-06,
+      "loss": 0.0275,
+      "step": 7096
+    },
+    {
+      "epoch": 3.1521208083499888,
+      "grad_norm": 0.7565904341486527,
+      "learning_rate": 1.305493219000558e-06,
+      "loss": 0.0331,
+      "step": 7097
+    },
+    {
+      "epoch": 3.1525649566955365,
+      "grad_norm": 0.4041064888863333,
+      "learning_rate": 1.3041872842557669e-06,
+      "loss": 0.0303,
+      "step": 7098
+    },
+    {
+      "epoch": 3.153009105041084,
+      "grad_norm": 0.39758421079788286,
+      "learning_rate": 1.3028819050544078e-06,
+      "loss": 0.02,
+      "step": 7099
+    },
+    {
+      "epoch": 3.153453253386631,
+      "grad_norm": 0.43003387120599507,
+      "learning_rate": 1.3015770815927009e-06,
+      "loss": 0.034,
+      "step": 7100
+    },
+    {
+      "epoch": 3.1538974017321784,
+      "grad_norm": 0.4278201479573488,
+      "learning_rate": 1.3002728140667847e-06,
+      "loss": 0.0233,
+      "step": 7101
+    },
+    {
+      "epoch": 3.154341550077726,
+      "grad_norm": 0.37161324406542584,
+      "learning_rate": 1.2989691026727114e-06,
+      "loss": 0.0197,
+      "step": 7102
+    },
+    {
+      "epoch": 3.1547856984232734,
+      "grad_norm": 0.49261598275828133,
+      "learning_rate": 1.2976659476064528e-06,
+      "loss": 0.0332,
+      "step": 7103
+    },
+    {
+      "epoch": 3.1552298467688207,
+      "grad_norm": 0.41968097121084996,
+      "learning_rate": 1.2963633490638927e-06,
+      "loss": 0.0266,
+      "step": 7104
+    },
+    {
+      "epoch": 3.155673995114368,
+      "grad_norm": 0.37929111645310976,
+      "learning_rate": 1.2950613072408352e-06,
+      "loss": 0.0206,
+      "step": 7105
+    },
+    {
+      "epoch": 3.1561181434599157,
+      "grad_norm": 0.4490037141777807,
+      "learning_rate": 1.2937598223330006e-06,
+      "loss": 0.0335,
+      "step": 7106
+    },
+    {
+      "epoch": 3.156562291805463,
+      "grad_norm": 0.6757819685260033,
+      "learning_rate": 1.2924588945360195e-06,
+      "loss": 0.0387,
+      "step": 7107
+    },
+    {
+      "epoch": 3.1570064401510103,
+      "grad_norm": 0.38030068064549954,
+      "learning_rate": 1.2911585240454483e-06,
+      "loss": 0.0225,
+      "step": 7108
+    },
+    {
+      "epoch": 3.157450588496558,
+      "grad_norm": 0.3303111905779037,
+      "learning_rate": 1.2898587110567546e-06,
+      "loss": 0.0158,
+      "step": 7109
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.36066815697274596,
+      "learning_rate": 1.2885594557653197e-06,
+      "loss": 0.0271,
+      "step": 7110
+    },
+    {
+      "epoch": 3.1583388851876526,
+      "grad_norm": 0.5116023254049137,
+      "learning_rate": 1.2872607583664443e-06,
+      "loss": 0.0332,
+      "step": 7111
+    },
+    {
+      "epoch": 3.1587830335332,
+      "grad_norm": 0.4098679802743982,
+      "learning_rate": 1.2859626190553459e-06,
+      "loss": 0.0206,
+      "step": 7112
+    },
+    {
+      "epoch": 3.1592271818787476,
+      "grad_norm": 0.3664260571550095,
+      "learning_rate": 1.2846650380271563e-06,
+      "loss": 0.0215,
+      "step": 7113
+    },
+    {
+      "epoch": 3.159671330224295,
+      "grad_norm": 0.4801826025425368,
+      "learning_rate": 1.283368015476925e-06,
+      "loss": 0.0207,
+      "step": 7114
+    },
+    {
+      "epoch": 3.160115478569842,
+      "grad_norm": 0.43450502438961686,
+      "learning_rate": 1.2820715515996146e-06,
+      "loss": 0.02,
+      "step": 7115
+    },
+    {
+      "epoch": 3.16055962691539,
+      "grad_norm": 0.4768531731139893,
+      "learning_rate": 1.280775646590106e-06,
+      "loss": 0.0321,
+      "step": 7116
+    },
+    {
+      "epoch": 3.1610037752609372,
+      "grad_norm": 0.3425693376069199,
+      "learning_rate": 1.2794803006431984e-06,
+      "loss": 0.0188,
+      "step": 7117
+    },
+    {
+      "epoch": 3.1614479236064845,
+      "grad_norm": 0.5214149879054594,
+      "learning_rate": 1.2781855139535988e-06,
+      "loss": 0.0262,
+      "step": 7118
+    },
+    {
+      "epoch": 3.161892071952032,
+      "grad_norm": 0.40187878997324955,
+      "learning_rate": 1.2768912867159406e-06,
+      "loss": 0.0361,
+      "step": 7119
+    },
+    {
+      "epoch": 3.1623362202975795,
+      "grad_norm": 0.5083763707824389,
+      "learning_rate": 1.2755976191247682e-06,
+      "loss": 0.021,
+      "step": 7120
+    },
+    {
+      "epoch": 3.162780368643127,
+      "grad_norm": 0.4055192408506625,
+      "learning_rate": 1.2743045113745385e-06,
+      "loss": 0.0227,
+      "step": 7121
+    },
+    {
+      "epoch": 3.163224516988674,
+      "grad_norm": 0.4094976103066204,
+      "learning_rate": 1.2730119636596288e-06,
+      "loss": 0.0281,
+      "step": 7122
+    },
+    {
+      "epoch": 3.163668665334222,
+      "grad_norm": 0.45282087571650875,
+      "learning_rate": 1.2717199761743336e-06,
+      "loss": 0.0251,
+      "step": 7123
+    },
+    {
+      "epoch": 3.164112813679769,
+      "grad_norm": 0.4299624187324527,
+      "learning_rate": 1.2704285491128553e-06,
+      "loss": 0.0283,
+      "step": 7124
+    },
+    {
+      "epoch": 3.1645569620253164,
+      "grad_norm": 0.4009907594057322,
+      "learning_rate": 1.2691376826693235e-06,
+      "loss": 0.0281,
+      "step": 7125
+    },
+    {
+      "epoch": 3.1650011103708637,
+      "grad_norm": 0.36139049623487757,
+      "learning_rate": 1.2678473770377726e-06,
+      "loss": 0.0227,
+      "step": 7126
+    },
+    {
+      "epoch": 3.1654452587164115,
+      "grad_norm": 0.3911090099172185,
+      "learning_rate": 1.2665576324121587e-06,
+      "loss": 0.019,
+      "step": 7127
+    },
+    {
+      "epoch": 3.1658894070619588,
+      "grad_norm": 0.4537050187193512,
+      "learning_rate": 1.2652684489863532e-06,
+      "loss": 0.0171,
+      "step": 7128
+    },
+    {
+      "epoch": 3.166333555407506,
+      "grad_norm": 0.4737431638279446,
+      "learning_rate": 1.2639798269541432e-06,
+      "loss": 0.0262,
+      "step": 7129
+    },
+    {
+      "epoch": 3.1667777037530533,
+      "grad_norm": 0.5356720104817232,
+      "learning_rate": 1.2626917665092265e-06,
+      "loss": 0.0367,
+      "step": 7130
+    },
+    {
+      "epoch": 3.167221852098601,
+      "grad_norm": 0.41884008779805304,
+      "learning_rate": 1.2614042678452254e-06,
+      "loss": 0.0277,
+      "step": 7131
+    },
+    {
+      "epoch": 3.1676660004441484,
+      "grad_norm": 0.3813088744442865,
+      "learning_rate": 1.260117331155669e-06,
+      "loss": 0.0241,
+      "step": 7132
+    },
+    {
+      "epoch": 3.1681101487896957,
+      "grad_norm": 0.4824659904614872,
+      "learning_rate": 1.258830956634008e-06,
+      "loss": 0.0266,
+      "step": 7133
+    },
+    {
+      "epoch": 3.168554297135243,
+      "grad_norm": 0.3933046865818541,
+      "learning_rate": 1.2575451444736065e-06,
+      "loss": 0.0241,
+      "step": 7134
+    },
+    {
+      "epoch": 3.1689984454807907,
+      "grad_norm": 0.4675792699661695,
+      "learning_rate": 1.25625989486774e-06,
+      "loss": 0.0231,
+      "step": 7135
+    },
+    {
+      "epoch": 3.169442593826338,
+      "grad_norm": 0.47235339239033924,
+      "learning_rate": 1.2549752080096078e-06,
+      "loss": 0.0228,
+      "step": 7136
+    },
+    {
+      "epoch": 3.1698867421718853,
+      "grad_norm": 0.5375969536724705,
+      "learning_rate": 1.2536910840923205e-06,
+      "loss": 0.0221,
+      "step": 7137
+    },
+    {
+      "epoch": 3.170330890517433,
+      "grad_norm": 0.4497621301881044,
+      "learning_rate": 1.2524075233089e-06,
+      "loss": 0.0343,
+      "step": 7138
+    },
+    {
+      "epoch": 3.1707750388629803,
+      "grad_norm": 0.292034357924613,
+      "learning_rate": 1.251124525852289e-06,
+      "loss": 0.0126,
+      "step": 7139
+    },
+    {
+      "epoch": 3.1712191872085276,
+      "grad_norm": 0.35964553866038385,
+      "learning_rate": 1.2498420919153464e-06,
+      "loss": 0.0275,
+      "step": 7140
+    },
+    {
+      "epoch": 3.171663335554075,
+      "grad_norm": 0.35000228782660847,
+      "learning_rate": 1.2485602216908378e-06,
+      "loss": 0.0175,
+      "step": 7141
+    },
+    {
+      "epoch": 3.1721074838996226,
+      "grad_norm": 0.3682428993293255,
+      "learning_rate": 1.2472789153714572e-06,
+      "loss": 0.0239,
+      "step": 7142
+    },
+    {
+      "epoch": 3.17255163224517,
+      "grad_norm": 0.5552689170116238,
+      "learning_rate": 1.245998173149801e-06,
+      "loss": 0.0296,
+      "step": 7143
+    },
+    {
+      "epoch": 3.172995780590717,
+      "grad_norm": 0.4582090480824963,
+      "learning_rate": 1.244717995218389e-06,
+      "loss": 0.0331,
+      "step": 7144
+    },
+    {
+      "epoch": 3.173439928936265,
+      "grad_norm": 0.5212899725680629,
+      "learning_rate": 1.2434383817696548e-06,
+      "loss": 0.0258,
+      "step": 7145
+    },
+    {
+      "epoch": 3.173884077281812,
+      "grad_norm": 0.381695700080771,
+      "learning_rate": 1.2421593329959437e-06,
+      "loss": 0.0243,
+      "step": 7146
+    },
+    {
+      "epoch": 3.1743282256273595,
+      "grad_norm": 0.3974320066485413,
+      "learning_rate": 1.2408808490895176e-06,
+      "loss": 0.0296,
+      "step": 7147
+    },
+    {
+      "epoch": 3.174772373972907,
+      "grad_norm": 0.37908843626651684,
+      "learning_rate": 1.2396029302425589e-06,
+      "loss": 0.0233,
+      "step": 7148
+    },
+    {
+      "epoch": 3.1752165223184545,
+      "grad_norm": 0.3916584808786519,
+      "learning_rate": 1.2383255766471564e-06,
+      "loss": 0.0231,
+      "step": 7149
+    },
+    {
+      "epoch": 3.175660670664002,
+      "grad_norm": 0.4921323330468778,
+      "learning_rate": 1.2370487884953198e-06,
+      "loss": 0.0282,
+      "step": 7150
+    },
+    {
+      "epoch": 3.176104819009549,
+      "grad_norm": 0.4474378513318365,
+      "learning_rate": 1.2357725659789727e-06,
+      "loss": 0.0316,
+      "step": 7151
+    },
+    {
+      "epoch": 3.1765489673550964,
+      "grad_norm": 0.507768145335404,
+      "learning_rate": 1.234496909289949e-06,
+      "loss": 0.0314,
+      "step": 7152
+    },
+    {
+      "epoch": 3.176993115700644,
+      "grad_norm": 0.43929762431880826,
+      "learning_rate": 1.2332218186200062e-06,
+      "loss": 0.036,
+      "step": 7153
+    },
+    {
+      "epoch": 3.1774372640461914,
+      "grad_norm": 0.45050762765015684,
+      "learning_rate": 1.2319472941608118e-06,
+      "loss": 0.0243,
+      "step": 7154
+    },
+    {
+      "epoch": 3.1778814123917387,
+      "grad_norm": 0.35294587274588873,
+      "learning_rate": 1.2306733361039457e-06,
+      "loss": 0.0214,
+      "step": 7155
+    },
+    {
+      "epoch": 3.1783255607372864,
+      "grad_norm": 0.41208302752606807,
+      "learning_rate": 1.2293999446409067e-06,
+      "loss": 0.0261,
+      "step": 7156
+    },
+    {
+      "epoch": 3.1787697090828337,
+      "grad_norm": 0.5041315813475921,
+      "learning_rate": 1.228127119963109e-06,
+      "loss": 0.0285,
+      "step": 7157
+    },
+    {
+      "epoch": 3.179213857428381,
+      "grad_norm": 0.44300292068122016,
+      "learning_rate": 1.2268548622618753e-06,
+      "loss": 0.0297,
+      "step": 7158
+    },
+    {
+      "epoch": 3.1796580057739283,
+      "grad_norm": 0.45740073144316085,
+      "learning_rate": 1.2255831717284528e-06,
+      "loss": 0.0279,
+      "step": 7159
+    },
+    {
+      "epoch": 3.180102154119476,
+      "grad_norm": 0.3604803514556297,
+      "learning_rate": 1.2243120485539944e-06,
+      "loss": 0.0223,
+      "step": 7160
+    },
+    {
+      "epoch": 3.1805463024650233,
+      "grad_norm": 0.3321847278445319,
+      "learning_rate": 1.223041492929573e-06,
+      "loss": 0.0165,
+      "step": 7161
+    },
+    {
+      "epoch": 3.1809904508105706,
+      "grad_norm": 0.35593201085972326,
+      "learning_rate": 1.221771505046176e-06,
+      "loss": 0.0214,
+      "step": 7162
+    },
+    {
+      "epoch": 3.181434599156118,
+      "grad_norm": 0.48430071345607734,
+      "learning_rate": 1.2205020850947009e-06,
+      "loss": 0.0324,
+      "step": 7163
+    },
+    {
+      "epoch": 3.1818787475016657,
+      "grad_norm": 0.3786369491408754,
+      "learning_rate": 1.219233233265964e-06,
+      "loss": 0.0196,
+      "step": 7164
+    },
+    {
+      "epoch": 3.182322895847213,
+      "grad_norm": 0.422396003313245,
+      "learning_rate": 1.2179649497506984e-06,
+      "loss": 0.0235,
+      "step": 7165
+    },
+    {
+      "epoch": 3.1827670441927602,
+      "grad_norm": 0.5088819210251279,
+      "learning_rate": 1.216697234739545e-06,
+      "loss": 0.0245,
+      "step": 7166
+    },
+    {
+      "epoch": 3.183211192538308,
+      "grad_norm": 0.44593024209496107,
+      "learning_rate": 1.2154300884230647e-06,
+      "loss": 0.0248,
+      "step": 7167
+    },
+    {
+      "epoch": 3.1836553408838553,
+      "grad_norm": 0.44151282819218973,
+      "learning_rate": 1.2141635109917322e-06,
+      "loss": 0.0243,
+      "step": 7168
+    },
+    {
+      "epoch": 3.1840994892294026,
+      "grad_norm": 0.5179310919609202,
+      "learning_rate": 1.2128975026359308e-06,
+      "loss": 0.0303,
+      "step": 7169
+    },
+    {
+      "epoch": 3.18454363757495,
+      "grad_norm": 0.47378558079444577,
+      "learning_rate": 1.2116320635459694e-06,
+      "loss": 0.0308,
+      "step": 7170
+    },
+    {
+      "epoch": 3.1849877859204976,
+      "grad_norm": 0.3483756788167636,
+      "learning_rate": 1.2103671939120603e-06,
+      "loss": 0.0205,
+      "step": 7171
+    },
+    {
+      "epoch": 3.185431934266045,
+      "grad_norm": 0.43627278086001237,
+      "learning_rate": 1.2091028939243372e-06,
+      "loss": 0.0348,
+      "step": 7172
+    },
+    {
+      "epoch": 3.185876082611592,
+      "grad_norm": 0.49521087150532456,
+      "learning_rate": 1.207839163772845e-06,
+      "loss": 0.0312,
+      "step": 7173
+    },
+    {
+      "epoch": 3.18632023095714,
+      "grad_norm": 0.4856268760629859,
+      "learning_rate": 1.206576003647545e-06,
+      "loss": 0.02,
+      "step": 7174
+    },
+    {
+      "epoch": 3.186764379302687,
+      "grad_norm": 0.39637861480749204,
+      "learning_rate": 1.2053134137383082e-06,
+      "loss": 0.0255,
+      "step": 7175
+    },
+    {
+      "epoch": 3.1872085276482345,
+      "grad_norm": 0.43819489132093886,
+      "learning_rate": 1.2040513942349285e-06,
+      "loss": 0.0197,
+      "step": 7176
+    },
+    {
+      "epoch": 3.1876526759937818,
+      "grad_norm": 0.4216663259343465,
+      "learning_rate": 1.2027899453271046e-06,
+      "loss": 0.023,
+      "step": 7177
+    },
+    {
+      "epoch": 3.1880968243393295,
+      "grad_norm": 0.36509628207338957,
+      "learning_rate": 1.2015290672044555e-06,
+      "loss": 0.0213,
+      "step": 7178
+    },
+    {
+      "epoch": 3.188540972684877,
+      "grad_norm": 0.5214855632257979,
+      "learning_rate": 1.2002687600565138e-06,
+      "loss": 0.0326,
+      "step": 7179
+    },
+    {
+      "epoch": 3.188985121030424,
+      "grad_norm": 0.4141851354865393,
+      "learning_rate": 1.199009024072722e-06,
+      "loss": 0.0224,
+      "step": 7180
+    },
+    {
+      "epoch": 3.1894292693759714,
+      "grad_norm": 0.33059616966272004,
+      "learning_rate": 1.1977498594424404e-06,
+      "loss": 0.0197,
+      "step": 7181
+    },
+    {
+      "epoch": 3.189873417721519,
+      "grad_norm": 0.5492371676948207,
+      "learning_rate": 1.196491266354946e-06,
+      "loss": 0.0251,
+      "step": 7182
+    },
+    {
+      "epoch": 3.1903175660670664,
+      "grad_norm": 0.4321665285793292,
+      "learning_rate": 1.1952332449994236e-06,
+      "loss": 0.0254,
+      "step": 7183
+    },
+    {
+      "epoch": 3.1907617144126137,
+      "grad_norm": 0.5421889351399802,
+      "learning_rate": 1.1939757955649762e-06,
+      "loss": 0.0205,
+      "step": 7184
+    },
+    {
+      "epoch": 3.1912058627581614,
+      "grad_norm": 0.3637124855912809,
+      "learning_rate": 1.1927189182406207e-06,
+      "loss": 0.028,
+      "step": 7185
+    },
+    {
+      "epoch": 3.1916500111037087,
+      "grad_norm": 0.4272124427935157,
+      "learning_rate": 1.191462613215284e-06,
+      "loss": 0.0245,
+      "step": 7186
+    },
+    {
+      "epoch": 3.192094159449256,
+      "grad_norm": 0.5695918619211989,
+      "learning_rate": 1.190206880677815e-06,
+      "loss": 0.0275,
+      "step": 7187
+    },
+    {
+      "epoch": 3.1925383077948033,
+      "grad_norm": 0.47921802084665843,
+      "learning_rate": 1.188951720816967e-06,
+      "loss": 0.0327,
+      "step": 7188
+    },
+    {
+      "epoch": 3.192982456140351,
+      "grad_norm": 0.47676512936038173,
+      "learning_rate": 1.1876971338214144e-06,
+      "loss": 0.0237,
+      "step": 7189
+    },
+    {
+      "epoch": 3.1934266044858983,
+      "grad_norm": 0.32414238199829704,
+      "learning_rate": 1.1864431198797433e-06,
+      "loss": 0.0223,
+      "step": 7190
+    },
+    {
+      "epoch": 3.1938707528314456,
+      "grad_norm": 0.37243806607221175,
+      "learning_rate": 1.1851896791804507e-06,
+      "loss": 0.0226,
+      "step": 7191
+    },
+    {
+      "epoch": 3.194314901176993,
+      "grad_norm": 0.45790234199301505,
+      "learning_rate": 1.1839368119119504e-06,
+      "loss": 0.0366,
+      "step": 7192
+    },
+    {
+      "epoch": 3.1947590495225406,
+      "grad_norm": 0.3980186449229834,
+      "learning_rate": 1.182684518262574e-06,
+      "loss": 0.0285,
+      "step": 7193
+    },
+    {
+      "epoch": 3.195203197868088,
+      "grad_norm": 0.37521542650727085,
+      "learning_rate": 1.1814327984205576e-06,
+      "loss": 0.0262,
+      "step": 7194
+    },
+    {
+      "epoch": 3.195647346213635,
+      "grad_norm": 0.5161650488172221,
+      "learning_rate": 1.1801816525740578e-06,
+      "loss": 0.0314,
+      "step": 7195
+    },
+    {
+      "epoch": 3.196091494559183,
+      "grad_norm": 0.4308857816197324,
+      "learning_rate": 1.1789310809111444e-06,
+      "loss": 0.0214,
+      "step": 7196
+    },
+    {
+      "epoch": 3.1965356429047302,
+      "grad_norm": 0.34195463841205975,
+      "learning_rate": 1.1776810836197965e-06,
+      "loss": 0.0148,
+      "step": 7197
+    },
+    {
+      "epoch": 3.1969797912502775,
+      "grad_norm": 0.40193187911181305,
+      "learning_rate": 1.1764316608879122e-06,
+      "loss": 0.0246,
+      "step": 7198
+    },
+    {
+      "epoch": 3.197423939595825,
+      "grad_norm": 0.45090018760439604,
+      "learning_rate": 1.1751828129033e-06,
+      "loss": 0.0332,
+      "step": 7199
+    },
+    {
+      "epoch": 3.1978680879413726,
+      "grad_norm": 0.45570245011519855,
+      "learning_rate": 1.1739345398536834e-06,
+      "loss": 0.0287,
+      "step": 7200
+    },
+    {
+      "epoch": 3.19831223628692,
+      "grad_norm": 0.3951478023748843,
+      "learning_rate": 1.1726868419266985e-06,
+      "loss": 0.022,
+      "step": 7201
+    },
+    {
+      "epoch": 3.198756384632467,
+      "grad_norm": 0.45157628259237137,
+      "learning_rate": 1.1714397193098975e-06,
+      "loss": 0.0254,
+      "step": 7202
+    },
+    {
+      "epoch": 3.199200532978015,
+      "grad_norm": 0.577449890397031,
+      "learning_rate": 1.1701931721907417e-06,
+      "loss": 0.0475,
+      "step": 7203
+    },
+    {
+      "epoch": 3.199644681323562,
+      "grad_norm": 0.3377142483011571,
+      "learning_rate": 1.1689472007566082e-06,
+      "loss": 0.0233,
+      "step": 7204
+    },
+    {
+      "epoch": 3.2000888296691095,
+      "grad_norm": 0.5632120326062124,
+      "learning_rate": 1.1677018051947898e-06,
+      "loss": 0.0306,
+      "step": 7205
+    },
+    {
+      "epoch": 3.2005329780146567,
+      "grad_norm": 0.3154600517639766,
+      "learning_rate": 1.1664569856924885e-06,
+      "loss": 0.0185,
+      "step": 7206
+    },
+    {
+      "epoch": 3.2009771263602045,
+      "grad_norm": 0.3451810701202185,
+      "learning_rate": 1.1652127424368248e-06,
+      "loss": 0.0188,
+      "step": 7207
+    },
+    {
+      "epoch": 3.2014212747057518,
+      "grad_norm": 0.4315345094850551,
+      "learning_rate": 1.1639690756148258e-06,
+      "loss": 0.0247,
+      "step": 7208
+    },
+    {
+      "epoch": 3.201865423051299,
+      "grad_norm": 0.46106139760465975,
+      "learning_rate": 1.162725985413436e-06,
+      "loss": 0.0373,
+      "step": 7209
+    },
+    {
+      "epoch": 3.2023095713968464,
+      "grad_norm": 0.35725860550074096,
+      "learning_rate": 1.1614834720195173e-06,
+      "loss": 0.0225,
+      "step": 7210
+    },
+    {
+      "epoch": 3.202753719742394,
+      "grad_norm": 0.4457350062298839,
+      "learning_rate": 1.1602415356198366e-06,
+      "loss": 0.0257,
+      "step": 7211
+    },
+    {
+      "epoch": 3.2031978680879414,
+      "grad_norm": 0.414436238208724,
+      "learning_rate": 1.1590001764010795e-06,
+      "loss": 0.0233,
+      "step": 7212
+    },
+    {
+      "epoch": 3.2036420164334887,
+      "grad_norm": 0.5552444496185882,
+      "learning_rate": 1.1577593945498439e-06,
+      "loss": 0.0267,
+      "step": 7213
+    },
+    {
+      "epoch": 3.2040861647790364,
+      "grad_norm": 0.9572529892460224,
+      "learning_rate": 1.156519190252638e-06,
+      "loss": 0.0208,
+      "step": 7214
+    },
+    {
+      "epoch": 3.2045303131245837,
+      "grad_norm": 0.4000797957812149,
+      "learning_rate": 1.1552795636958874e-06,
+      "loss": 0.021,
+      "step": 7215
+    },
+    {
+      "epoch": 3.204974461470131,
+      "grad_norm": 0.44923949008034364,
+      "learning_rate": 1.154040515065929e-06,
+      "loss": 0.0257,
+      "step": 7216
+    },
+    {
+      "epoch": 3.2054186098156783,
+      "grad_norm": 0.3620392807393453,
+      "learning_rate": 1.1528020445490122e-06,
+      "loss": 0.0168,
+      "step": 7217
+    },
+    {
+      "epoch": 3.205862758161226,
+      "grad_norm": 0.40007534369795833,
+      "learning_rate": 1.1515641523313026e-06,
+      "loss": 0.019,
+      "step": 7218
+    },
+    {
+      "epoch": 3.2063069065067733,
+      "grad_norm": 0.48298838416688933,
+      "learning_rate": 1.1503268385988726e-06,
+      "loss": 0.0289,
+      "step": 7219
+    },
+    {
+      "epoch": 3.2067510548523206,
+      "grad_norm": 0.42002715971416216,
+      "learning_rate": 1.1490901035377127e-06,
+      "loss": 0.0218,
+      "step": 7220
+    },
+    {
+      "epoch": 3.207195203197868,
+      "grad_norm": 0.3286114150705751,
+      "learning_rate": 1.147853947333727e-06,
+      "loss": 0.0197,
+      "step": 7221
+    },
+    {
+      "epoch": 3.2076393515434156,
+      "grad_norm": 0.4609222221897965,
+      "learning_rate": 1.1466183701727285e-06,
+      "loss": 0.0324,
+      "step": 7222
+    },
+    {
+      "epoch": 3.208083499888963,
+      "grad_norm": 0.685447809015932,
+      "learning_rate": 1.1453833722404467e-06,
+      "loss": 0.0324,
+      "step": 7223
+    },
+    {
+      "epoch": 3.20852764823451,
+      "grad_norm": 0.3436945045836064,
+      "learning_rate": 1.1441489537225242e-06,
+      "loss": 0.0157,
+      "step": 7224
+    },
+    {
+      "epoch": 3.208971796580058,
+      "grad_norm": 0.5032138454546828,
+      "learning_rate": 1.142915114804512e-06,
+      "loss": 0.0305,
+      "step": 7225
+    },
+    {
+      "epoch": 3.2094159449256052,
+      "grad_norm": 0.4731526252440794,
+      "learning_rate": 1.1416818556718766e-06,
+      "loss": 0.0213,
+      "step": 7226
+    },
+    {
+      "epoch": 3.2098600932711525,
+      "grad_norm": 0.4103989387772799,
+      "learning_rate": 1.1404491765100028e-06,
+      "loss": 0.0233,
+      "step": 7227
+    },
+    {
+      "epoch": 3.2103042416167,
+      "grad_norm": 0.4494816164510247,
+      "learning_rate": 1.1392170775041788e-06,
+      "loss": 0.0297,
+      "step": 7228
+    },
+    {
+      "epoch": 3.2107483899622475,
+      "grad_norm": 0.49371762049111106,
+      "learning_rate": 1.1379855588396111e-06,
+      "loss": 0.0352,
+      "step": 7229
+    },
+    {
+      "epoch": 3.211192538307795,
+      "grad_norm": 0.5061710158808476,
+      "learning_rate": 1.1367546207014197e-06,
+      "loss": 0.029,
+      "step": 7230
+    },
+    {
+      "epoch": 3.211636686653342,
+      "grad_norm": 0.8145152873713564,
+      "learning_rate": 1.1355242632746322e-06,
+      "loss": 0.0262,
+      "step": 7231
+    },
+    {
+      "epoch": 3.21208083499889,
+      "grad_norm": 0.3714708449077717,
+      "learning_rate": 1.134294486744194e-06,
+      "loss": 0.0195,
+      "step": 7232
+    },
+    {
+      "epoch": 3.212524983344437,
+      "grad_norm": 0.3746783694058875,
+      "learning_rate": 1.1330652912949614e-06,
+      "loss": 0.0315,
+      "step": 7233
+    },
+    {
+      "epoch": 3.2129691316899844,
+      "grad_norm": 0.3552761658812224,
+      "learning_rate": 1.131836677111703e-06,
+      "loss": 0.0195,
+      "step": 7234
+    },
+    {
+      "epoch": 3.2134132800355317,
+      "grad_norm": 0.425007652814607,
+      "learning_rate": 1.130608644379102e-06,
+      "loss": 0.025,
+      "step": 7235
+    },
+    {
+      "epoch": 3.2138574283810795,
+      "grad_norm": 0.43005151453900625,
+      "learning_rate": 1.12938119328175e-06,
+      "loss": 0.0209,
+      "step": 7236
+    },
+    {
+      "epoch": 3.2143015767266268,
+      "grad_norm": 0.4595931551885754,
+      "learning_rate": 1.1281543240041553e-06,
+      "loss": 0.0299,
+      "step": 7237
+    },
+    {
+      "epoch": 3.214745725072174,
+      "grad_norm": 0.46371024210057155,
+      "learning_rate": 1.1269280367307366e-06,
+      "loss": 0.0323,
+      "step": 7238
+    },
+    {
+      "epoch": 3.2151898734177213,
+      "grad_norm": 0.6313936018053397,
+      "learning_rate": 1.125702331645826e-06,
+      "loss": 0.045,
+      "step": 7239
+    },
+    {
+      "epoch": 3.215634021763269,
+      "grad_norm": 0.4072252776090633,
+      "learning_rate": 1.1244772089336676e-06,
+      "loss": 0.0201,
+      "step": 7240
+    },
+    {
+      "epoch": 3.2160781701088164,
+      "grad_norm": 0.4916818117136901,
+      "learning_rate": 1.1232526687784196e-06,
+      "loss": 0.0321,
+      "step": 7241
+    },
+    {
+      "epoch": 3.2165223184543636,
+      "grad_norm": 0.4020576743291894,
+      "learning_rate": 1.1220287113641487e-06,
+      "loss": 0.0235,
+      "step": 7242
+    },
+    {
+      "epoch": 3.216966466799911,
+      "grad_norm": 0.3816616945761622,
+      "learning_rate": 1.1208053368748379e-06,
+      "loss": 0.0195,
+      "step": 7243
+    },
+    {
+      "epoch": 3.2174106151454587,
+      "grad_norm": 0.7096021175300753,
+      "learning_rate": 1.1195825454943805e-06,
+      "loss": 0.0277,
+      "step": 7244
+    },
+    {
+      "epoch": 3.217854763491006,
+      "grad_norm": 0.4537143505718799,
+      "learning_rate": 1.1183603374065832e-06,
+      "loss": 0.0317,
+      "step": 7245
+    },
+    {
+      "epoch": 3.2182989118365533,
+      "grad_norm": 0.3124829853038914,
+      "learning_rate": 1.1171387127951667e-06,
+      "loss": 0.0134,
+      "step": 7246
+    },
+    {
+      "epoch": 3.218743060182101,
+      "grad_norm": 0.3523041765470259,
+      "learning_rate": 1.1159176718437581e-06,
+      "loss": 0.0197,
+      "step": 7247
+    },
+    {
+      "epoch": 3.2191872085276483,
+      "grad_norm": 0.46781406748141774,
+      "learning_rate": 1.114697214735903e-06,
+      "loss": 0.03,
+      "step": 7248
+    },
+    {
+      "epoch": 3.2196313568731956,
+      "grad_norm": 0.45497498562705085,
+      "learning_rate": 1.113477341655056e-06,
+      "loss": 0.0326,
+      "step": 7249
+    },
+    {
+      "epoch": 3.220075505218743,
+      "grad_norm": 0.34355725383797536,
+      "learning_rate": 1.1122580527845844e-06,
+      "loss": 0.0199,
+      "step": 7250
+    },
+    {
+      "epoch": 3.2205196535642906,
+      "grad_norm": 0.4222581331896639,
+      "learning_rate": 1.1110393483077697e-06,
+      "loss": 0.0248,
+      "step": 7251
+    },
+    {
+      "epoch": 3.220963801909838,
+      "grad_norm": 0.3466800241459592,
+      "learning_rate": 1.1098212284078037e-06,
+      "loss": 0.0262,
+      "step": 7252
+    },
+    {
+      "epoch": 3.221407950255385,
+      "grad_norm": 0.4157855377746249,
+      "learning_rate": 1.108603693267788e-06,
+      "loss": 0.0256,
+      "step": 7253
+    },
+    {
+      "epoch": 3.221852098600933,
+      "grad_norm": 0.383389548238552,
+      "learning_rate": 1.1073867430707409e-06,
+      "loss": 0.024,
+      "step": 7254
+    },
+    {
+      "epoch": 3.22229624694648,
+      "grad_norm": 0.30999218839566933,
+      "learning_rate": 1.1061703779995903e-06,
+      "loss": 0.0132,
+      "step": 7255
+    },
+    {
+      "epoch": 3.2227403952920275,
+      "grad_norm": 0.4028830849965787,
+      "learning_rate": 1.1049545982371763e-06,
+      "loss": 0.0237,
+      "step": 7256
+    },
+    {
+      "epoch": 3.223184543637575,
+      "grad_norm": 0.399211499707528,
+      "learning_rate": 1.1037394039662514e-06,
+      "loss": 0.025,
+      "step": 7257
+    },
+    {
+      "epoch": 3.2236286919831225,
+      "grad_norm": 0.4718629067386526,
+      "learning_rate": 1.1025247953694812e-06,
+      "loss": 0.022,
+      "step": 7258
+    },
+    {
+      "epoch": 3.22407284032867,
+      "grad_norm": 0.37912161530644367,
+      "learning_rate": 1.1013107726294398e-06,
+      "loss": 0.0218,
+      "step": 7259
+    },
+    {
+      "epoch": 3.224516988674217,
+      "grad_norm": 0.34533958399869225,
+      "learning_rate": 1.100097335928616e-06,
+      "loss": 0.0165,
+      "step": 7260
+    },
+    {
+      "epoch": 3.224961137019765,
+      "grad_norm": 0.5551950864778356,
+      "learning_rate": 1.0988844854494108e-06,
+      "loss": 0.0351,
+      "step": 7261
+    },
+    {
+      "epoch": 3.225405285365312,
+      "grad_norm": 0.5024454449508828,
+      "learning_rate": 1.0976722213741353e-06,
+      "loss": 0.0264,
+      "step": 7262
+    },
+    {
+      "epoch": 3.2258494337108594,
+      "grad_norm": 0.44064209832820683,
+      "learning_rate": 1.0964605438850157e-06,
+      "loss": 0.0277,
+      "step": 7263
+    },
+    {
+      "epoch": 3.2262935820564067,
+      "grad_norm": 0.3886834550548472,
+      "learning_rate": 1.0952494531641845e-06,
+      "loss": 0.0254,
+      "step": 7264
+    },
+    {
+      "epoch": 3.2267377304019544,
+      "grad_norm": 0.44369145821666817,
+      "learning_rate": 1.0940389493936903e-06,
+      "loss": 0.0305,
+      "step": 7265
+    },
+    {
+      "epoch": 3.2271818787475017,
+      "grad_norm": 0.5083410379311883,
+      "learning_rate": 1.092829032755493e-06,
+      "loss": 0.0234,
+      "step": 7266
+    },
+    {
+      "epoch": 3.227626027093049,
+      "grad_norm": 0.5641232827648482,
+      "learning_rate": 1.091619703431463e-06,
+      "loss": 0.033,
+      "step": 7267
+    },
+    {
+      "epoch": 3.2280701754385963,
+      "grad_norm": 0.49450571188792775,
+      "learning_rate": 1.0904109616033837e-06,
+      "loss": 0.0259,
+      "step": 7268
+    },
+    {
+      "epoch": 3.228514323784144,
+      "grad_norm": 0.40475539158874035,
+      "learning_rate": 1.0892028074529504e-06,
+      "loss": 0.0304,
+      "step": 7269
+    },
+    {
+      "epoch": 3.2289584721296913,
+      "grad_norm": 0.3822868070952512,
+      "learning_rate": 1.0879952411617668e-06,
+      "loss": 0.0258,
+      "step": 7270
+    },
+    {
+      "epoch": 3.2294026204752386,
+      "grad_norm": 0.32534348009223135,
+      "learning_rate": 1.0867882629113512e-06,
+      "loss": 0.0165,
+      "step": 7271
+    },
+    {
+      "epoch": 3.229846768820786,
+      "grad_norm": 0.44997009638741614,
+      "learning_rate": 1.085581872883134e-06,
+      "loss": 0.0288,
+      "step": 7272
+    },
+    {
+      "epoch": 3.2302909171663337,
+      "grad_norm": 0.3775740023239124,
+      "learning_rate": 1.0843760712584557e-06,
+      "loss": 0.0191,
+      "step": 7273
+    },
+    {
+      "epoch": 3.230735065511881,
+      "grad_norm": 0.36873567439006616,
+      "learning_rate": 1.0831708582185684e-06,
+      "loss": 0.0215,
+      "step": 7274
+    },
+    {
+      "epoch": 3.2311792138574282,
+      "grad_norm": 0.41830159166514674,
+      "learning_rate": 1.081966233944638e-06,
+      "loss": 0.0349,
+      "step": 7275
+    },
+    {
+      "epoch": 3.231623362202976,
+      "grad_norm": 0.3474261761721994,
+      "learning_rate": 1.0807621986177369e-06,
+      "loss": 0.0197,
+      "step": 7276
+    },
+    {
+      "epoch": 3.2320675105485233,
+      "grad_norm": 0.4726466217876324,
+      "learning_rate": 1.0795587524188532e-06,
+      "loss": 0.0257,
+      "step": 7277
+    },
+    {
+      "epoch": 3.2325116588940705,
+      "grad_norm": 0.5853092492971105,
+      "learning_rate": 1.0783558955288864e-06,
+      "loss": 0.033,
+      "step": 7278
+    },
+    {
+      "epoch": 3.232955807239618,
+      "grad_norm": 0.4076799037255477,
+      "learning_rate": 1.0771536281286454e-06,
+      "loss": 0.0251,
+      "step": 7279
+    },
+    {
+      "epoch": 3.2333999555851656,
+      "grad_norm": 0.3975030901670652,
+      "learning_rate": 1.0759519503988525e-06,
+      "loss": 0.0223,
+      "step": 7280
+    },
+    {
+      "epoch": 3.233844103930713,
+      "grad_norm": 0.40051231781743457,
+      "learning_rate": 1.0747508625201387e-06,
+      "loss": 0.0247,
+      "step": 7281
+    },
+    {
+      "epoch": 3.23428825227626,
+      "grad_norm": 0.44390849962202206,
+      "learning_rate": 1.0735503646730483e-06,
+      "loss": 0.0281,
+      "step": 7282
+    },
+    {
+      "epoch": 3.234732400621808,
+      "grad_norm": 0.4052200206226338,
+      "learning_rate": 1.0723504570380367e-06,
+      "loss": 0.0286,
+      "step": 7283
+    },
+    {
+      "epoch": 3.235176548967355,
+      "grad_norm": 0.37902885647278484,
+      "learning_rate": 1.0711511397954706e-06,
+      "loss": 0.0205,
+      "step": 7284
+    },
+    {
+      "epoch": 3.2356206973129025,
+      "grad_norm": 0.3395460769637531,
+      "learning_rate": 1.0699524131256273e-06,
+      "loss": 0.0201,
+      "step": 7285
+    },
+    {
+      "epoch": 3.2360648456584498,
+      "grad_norm": 0.3790836643740825,
+      "learning_rate": 1.0687542772086978e-06,
+      "loss": 0.0213,
+      "step": 7286
+    },
+    {
+      "epoch": 3.2365089940039975,
+      "grad_norm": 0.3344270224539923,
+      "learning_rate": 1.0675567322247794e-06,
+      "loss": 0.0188,
+      "step": 7287
+    },
+    {
+      "epoch": 3.236953142349545,
+      "grad_norm": 0.49193821672434174,
+      "learning_rate": 1.0663597783538843e-06,
+      "loss": 0.0258,
+      "step": 7288
+    },
+    {
+      "epoch": 3.237397290695092,
+      "grad_norm": 0.40685832395170973,
+      "learning_rate": 1.0651634157759361e-06,
+      "loss": 0.0266,
+      "step": 7289
+    },
+    {
+      "epoch": 3.2378414390406394,
+      "grad_norm": 0.4090407200808684,
+      "learning_rate": 1.063967644670767e-06,
+      "loss": 0.0264,
+      "step": 7290
+    },
+    {
+      "epoch": 3.238285587386187,
+      "grad_norm": 0.3883951968082906,
+      "learning_rate": 1.0627724652181237e-06,
+      "loss": 0.033,
+      "step": 7291
+    },
+    {
+      "epoch": 3.2387297357317344,
+      "grad_norm": 0.43888578494036506,
+      "learning_rate": 1.06157787759766e-06,
+      "loss": 0.0291,
+      "step": 7292
+    },
+    {
+      "epoch": 3.2391738840772817,
+      "grad_norm": 0.5336092270346524,
+      "learning_rate": 1.0603838819889429e-06,
+      "loss": 0.031,
+      "step": 7293
+    },
+    {
+      "epoch": 3.2396180324228294,
+      "grad_norm": 0.4556708038719475,
+      "learning_rate": 1.0591904785714507e-06,
+      "loss": 0.0189,
+      "step": 7294
+    },
+    {
+      "epoch": 3.2400621807683767,
+      "grad_norm": 0.42245566909222265,
+      "learning_rate": 1.0579976675245724e-06,
+      "loss": 0.0241,
+      "step": 7295
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.4528542240509129,
+      "learning_rate": 1.0568054490276075e-06,
+      "loss": 0.0251,
+      "step": 7296
+    },
+    {
+      "epoch": 3.2409504774594713,
+      "grad_norm": 0.4541061461339724,
+      "learning_rate": 1.0556138232597684e-06,
+      "loss": 0.0201,
+      "step": 7297
+    },
+    {
+      "epoch": 3.241394625805019,
+      "grad_norm": 0.3790673698790023,
+      "learning_rate": 1.054422790400173e-06,
+      "loss": 0.019,
+      "step": 7298
+    },
+    {
+      "epoch": 3.2418387741505663,
+      "grad_norm": 0.41427572592093276,
+      "learning_rate": 1.0532323506278564e-06,
+      "loss": 0.0302,
+      "step": 7299
+    },
+    {
+      "epoch": 3.2422829224961136,
+      "grad_norm": 0.4164805864455333,
+      "learning_rate": 1.0520425041217613e-06,
+      "loss": 0.0374,
+      "step": 7300
+    },
+    {
+      "epoch": 3.242727070841661,
+      "grad_norm": 0.5061586285788595,
+      "learning_rate": 1.0508532510607421e-06,
+      "loss": 0.0369,
+      "step": 7301
+    },
+    {
+      "epoch": 3.2431712191872086,
+      "grad_norm": 0.51039712429684,
+      "learning_rate": 1.049664591623563e-06,
+      "loss": 0.0256,
+      "step": 7302
+    },
+    {
+      "epoch": 3.243615367532756,
+      "grad_norm": 0.3147661792422783,
+      "learning_rate": 1.0484765259889024e-06,
+      "loss": 0.0178,
+      "step": 7303
+    },
+    {
+      "epoch": 3.244059515878303,
+      "grad_norm": 0.34967275161446504,
+      "learning_rate": 1.0472890543353425e-06,
+      "loss": 0.022,
+      "step": 7304
+    },
+    {
+      "epoch": 3.244503664223851,
+      "grad_norm": 0.425569161792133,
+      "learning_rate": 1.0461021768413827e-06,
+      "loss": 0.0327,
+      "step": 7305
+    },
+    {
+      "epoch": 3.2449478125693982,
+      "grad_norm": 0.3260603432721134,
+      "learning_rate": 1.0449158936854308e-06,
+      "loss": 0.019,
+      "step": 7306
+    },
+    {
+      "epoch": 3.2453919609149455,
+      "grad_norm": 0.40232295144807395,
+      "learning_rate": 1.0437302050458053e-06,
+      "loss": 0.0343,
+      "step": 7307
+    },
+    {
+      "epoch": 3.245836109260493,
+      "grad_norm": 0.435190410566104,
+      "learning_rate": 1.0425451111007368e-06,
+      "loss": 0.0371,
+      "step": 7308
+    },
+    {
+      "epoch": 3.2462802576060406,
+      "grad_norm": 0.36886684345005827,
+      "learning_rate": 1.0413606120283616e-06,
+      "loss": 0.0159,
+      "step": 7309
+    },
+    {
+      "epoch": 3.246724405951588,
+      "grad_norm": 0.6185604106742186,
+      "learning_rate": 1.040176708006732e-06,
+      "loss": 0.0317,
+      "step": 7310
+    },
+    {
+      "epoch": 3.247168554297135,
+      "grad_norm": 0.4287929879516384,
+      "learning_rate": 1.0389933992138106e-06,
+      "loss": 0.0285,
+      "step": 7311
+    },
+    {
+      "epoch": 3.247612702642683,
+      "grad_norm": 0.5172218558312743,
+      "learning_rate": 1.0378106858274639e-06,
+      "loss": 0.0269,
+      "step": 7312
+    },
+    {
+      "epoch": 3.24805685098823,
+      "grad_norm": 0.47500334875701583,
+      "learning_rate": 1.036628568025479e-06,
+      "loss": 0.0291,
+      "step": 7313
+    },
+    {
+      "epoch": 3.2485009993337775,
+      "grad_norm": 0.36315045000352464,
+      "learning_rate": 1.035447045985547e-06,
+      "loss": 0.0205,
+      "step": 7314
+    },
+    {
+      "epoch": 3.2489451476793247,
+      "grad_norm": 0.3702839335927883,
+      "learning_rate": 1.0342661198852689e-06,
+      "loss": 0.0272,
+      "step": 7315
+    },
+    {
+      "epoch": 3.2493892960248725,
+      "grad_norm": 0.4172702979013527,
+      "learning_rate": 1.0330857899021584e-06,
+      "loss": 0.0288,
+      "step": 7316
+    },
+    {
+      "epoch": 3.2498334443704198,
+      "grad_norm": 0.4003396546345466,
+      "learning_rate": 1.03190605621364e-06,
+      "loss": 0.0249,
+      "step": 7317
+    },
+    {
+      "epoch": 3.250277592715967,
+      "grad_norm": 0.36419349273758006,
+      "learning_rate": 1.0307269189970482e-06,
+      "loss": 0.0227,
+      "step": 7318
+    },
+    {
+      "epoch": 3.250721741061515,
+      "grad_norm": 0.46337961532186456,
+      "learning_rate": 1.0295483784296274e-06,
+      "loss": 0.0292,
+      "step": 7319
+    },
+    {
+      "epoch": 3.251165889407062,
+      "grad_norm": 0.4844395758839405,
+      "learning_rate": 1.0283704346885303e-06,
+      "loss": 0.0315,
+      "step": 7320
+    },
+    {
+      "epoch": 3.2516100377526094,
+      "grad_norm": 0.43388879742046443,
+      "learning_rate": 1.027193087950823e-06,
+      "loss": 0.0253,
+      "step": 7321
+    },
+    {
+      "epoch": 3.2520541860981567,
+      "grad_norm": 0.43253571684648157,
+      "learning_rate": 1.0260163383934807e-06,
+      "loss": 0.0292,
+      "step": 7322
+    },
+    {
+      "epoch": 3.252498334443704,
+      "grad_norm": 0.37106610809103213,
+      "learning_rate": 1.0248401861933888e-06,
+      "loss": 0.0208,
+      "step": 7323
+    },
+    {
+      "epoch": 3.2529424827892517,
+      "grad_norm": 0.35259048739158694,
+      "learning_rate": 1.0236646315273436e-06,
+      "loss": 0.0223,
+      "step": 7324
+    },
+    {
+      "epoch": 3.253386631134799,
+      "grad_norm": 0.5151899180549023,
+      "learning_rate": 1.0224896745720513e-06,
+      "loss": 0.0313,
+      "step": 7325
+    },
+    {
+      "epoch": 3.2538307794803463,
+      "grad_norm": 0.4063218873557718,
+      "learning_rate": 1.0213153155041255e-06,
+      "loss": 0.0291,
+      "step": 7326
+    },
+    {
+      "epoch": 3.254274927825894,
+      "grad_norm": 0.6251543341151611,
+      "learning_rate": 1.0201415545000941e-06,
+      "loss": 0.0267,
+      "step": 7327
+    },
+    {
+      "epoch": 3.2547190761714413,
+      "grad_norm": 0.4171212598426182,
+      "learning_rate": 1.0189683917363947e-06,
+      "loss": 0.0227,
+      "step": 7328
+    },
+    {
+      "epoch": 3.2551632245169886,
+      "grad_norm": 0.49598266444833977,
+      "learning_rate": 1.0177958273893684e-06,
+      "loss": 0.0239,
+      "step": 7329
+    },
+    {
+      "epoch": 3.255607372862536,
+      "grad_norm": 0.3899965486282789,
+      "learning_rate": 1.016623861635277e-06,
+      "loss": 0.0209,
+      "step": 7330
+    },
+    {
+      "epoch": 3.2560515212080836,
+      "grad_norm": 0.43687191924955027,
+      "learning_rate": 1.0154524946502864e-06,
+      "loss": 0.0272,
+      "step": 7331
+    },
+    {
+      "epoch": 3.256495669553631,
+      "grad_norm": 0.4063747702825569,
+      "learning_rate": 1.01428172661047e-06,
+      "loss": 0.0255,
+      "step": 7332
+    },
+    {
+      "epoch": 3.256939817899178,
+      "grad_norm": 0.48223387404929335,
+      "learning_rate": 1.0131115576918154e-06,
+      "loss": 0.031,
+      "step": 7333
+    },
+    {
+      "epoch": 3.257383966244726,
+      "grad_norm": 0.39589428565081675,
+      "learning_rate": 1.011941988070219e-06,
+      "loss": 0.0222,
+      "step": 7334
+    },
+    {
+      "epoch": 3.257828114590273,
+      "grad_norm": 0.4100082037653643,
+      "learning_rate": 1.0107730179214875e-06,
+      "loss": 0.0359,
+      "step": 7335
+    },
+    {
+      "epoch": 3.2582722629358205,
+      "grad_norm": 0.7048819567321716,
+      "learning_rate": 1.0096046474213378e-06,
+      "loss": 0.0239,
+      "step": 7336
+    },
+    {
+      "epoch": 3.258716411281368,
+      "grad_norm": 0.4205899532427439,
+      "learning_rate": 1.008436876745393e-06,
+      "loss": 0.0329,
+      "step": 7337
+    },
+    {
+      "epoch": 3.2591605596269155,
+      "grad_norm": 0.3747110714323454,
+      "learning_rate": 1.00726970606919e-06,
+      "loss": 0.0217,
+      "step": 7338
+    },
+    {
+      "epoch": 3.259604707972463,
+      "grad_norm": 0.4970634544493451,
+      "learning_rate": 1.0061031355681766e-06,
+      "loss": 0.029,
+      "step": 7339
+    },
+    {
+      "epoch": 3.26004885631801,
+      "grad_norm": 0.35455430411205807,
+      "learning_rate": 1.0049371654177036e-06,
+      "loss": 0.0241,
+      "step": 7340
+    },
+    {
+      "epoch": 3.260493004663558,
+      "grad_norm": 0.38458418543297923,
+      "learning_rate": 1.0037717957930404e-06,
+      "loss": 0.0264,
+      "step": 7341
+    },
+    {
+      "epoch": 3.260937153009105,
+      "grad_norm": 0.44567315711458705,
+      "learning_rate": 1.0026070268693616e-06,
+      "loss": 0.033,
+      "step": 7342
+    },
+    {
+      "epoch": 3.2613813013546524,
+      "grad_norm": 0.6191592999982571,
+      "learning_rate": 1.0014428588217495e-06,
+      "loss": 0.0433,
+      "step": 7343
+    },
+    {
+      "epoch": 3.2618254497001997,
+      "grad_norm": 0.31811859198013076,
+      "learning_rate": 1.0002792918251991e-06,
+      "loss": 0.0178,
+      "step": 7344
+    },
+    {
+      "epoch": 3.2622695980457475,
+      "grad_norm": 0.4500830212917085,
+      "learning_rate": 9.991163260546154e-07,
+      "loss": 0.0283,
+      "step": 7345
+    },
+    {
+      "epoch": 3.2627137463912947,
+      "grad_norm": 0.45351163208898665,
+      "learning_rate": 9.979539616848088e-07,
+      "loss": 0.0347,
+      "step": 7346
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.35335672660967254,
+      "learning_rate": 9.96792198890506e-07,
+      "loss": 0.0188,
+      "step": 7347
+    },
+    {
+      "epoch": 3.2636020430823893,
+      "grad_norm": 0.44507187430087025,
+      "learning_rate": 9.956310378463397e-07,
+      "loss": 0.028,
+      "step": 7348
+    },
+    {
+      "epoch": 3.264046191427937,
+      "grad_norm": 0.5621390992373146,
+      "learning_rate": 9.94470478726849e-07,
+      "loss": 0.0318,
+      "step": 7349
+    },
+    {
+      "epoch": 3.2644903397734844,
+      "grad_norm": 0.4752468135015428,
+      "learning_rate": 9.933105217064876e-07,
+      "loss": 0.0301,
+      "step": 7350
+    },
+    {
+      "epoch": 3.2649344881190316,
+      "grad_norm": 0.4866383052323718,
+      "learning_rate": 9.921511669596169e-07,
+      "loss": 0.0258,
+      "step": 7351
+    },
+    {
+      "epoch": 3.265378636464579,
+      "grad_norm": 0.4858680526762395,
+      "learning_rate": 9.909924146605065e-07,
+      "loss": 0.0337,
+      "step": 7352
+    },
+    {
+      "epoch": 3.2658227848101267,
+      "grad_norm": 0.39515608234465965,
+      "learning_rate": 9.898342649833392e-07,
+      "loss": 0.0192,
+      "step": 7353
+    },
+    {
+      "epoch": 3.266266933155674,
+      "grad_norm": 0.36506881604203717,
+      "learning_rate": 9.88676718102201e-07,
+      "loss": 0.021,
+      "step": 7354
+    },
+    {
+      "epoch": 3.2667110815012212,
+      "grad_norm": 0.4187792232437751,
+      "learning_rate": 9.87519774191093e-07,
+      "loss": 0.0331,
+      "step": 7355
+    },
+    {
+      "epoch": 3.267155229846769,
+      "grad_norm": 0.46579982125230246,
+      "learning_rate": 9.863634334239241e-07,
+      "loss": 0.0231,
+      "step": 7356
+    },
+    {
+      "epoch": 3.2675993781923163,
+      "grad_norm": 0.38694241705258164,
+      "learning_rate": 9.852076959745082e-07,
+      "loss": 0.0259,
+      "step": 7357
+    },
+    {
+      "epoch": 3.2680435265378636,
+      "grad_norm": 0.46212859105639015,
+      "learning_rate": 9.840525620165763e-07,
+      "loss": 0.0317,
+      "step": 7358
+    },
+    {
+      "epoch": 3.268487674883411,
+      "grad_norm": 0.5473555001130638,
+      "learning_rate": 9.828980317237652e-07,
+      "loss": 0.0396,
+      "step": 7359
+    },
+    {
+      "epoch": 3.2689318232289586,
+      "grad_norm": 0.36341029826160076,
+      "learning_rate": 9.817441052696164e-07,
+      "loss": 0.0185,
+      "step": 7360
+    },
+    {
+      "epoch": 3.269375971574506,
+      "grad_norm": 0.44926471264353585,
+      "learning_rate": 9.805907828275874e-07,
+      "loss": 0.0315,
+      "step": 7361
+    },
+    {
+      "epoch": 3.269820119920053,
+      "grad_norm": 0.3017207630293574,
+      "learning_rate": 9.794380645710428e-07,
+      "loss": 0.017,
+      "step": 7362
+    },
+    {
+      "epoch": 3.270264268265601,
+      "grad_norm": 0.4066309090272054,
+      "learning_rate": 9.782859506732517e-07,
+      "loss": 0.0252,
+      "step": 7363
+    },
+    {
+      "epoch": 3.270708416611148,
+      "grad_norm": 0.45822398666676095,
+      "learning_rate": 9.771344413074018e-07,
+      "loss": 0.0289,
+      "step": 7364
+    },
+    {
+      "epoch": 3.2711525649566955,
+      "grad_norm": 0.45754692936117375,
+      "learning_rate": 9.75983536646581e-07,
+      "loss": 0.0204,
+      "step": 7365
+    },
+    {
+      "epoch": 3.2715967133022428,
+      "grad_norm": 0.4003872658949929,
+      "learning_rate": 9.748332368637903e-07,
+      "loss": 0.0239,
+      "step": 7366
+    },
+    {
+      "epoch": 3.2720408616477905,
+      "grad_norm": 0.40491759770498775,
+      "learning_rate": 9.736835421319397e-07,
+      "loss": 0.016,
+      "step": 7367
+    },
+    {
+      "epoch": 3.272485009993338,
+      "grad_norm": 0.40672999266918314,
+      "learning_rate": 9.725344526238495e-07,
+      "loss": 0.0275,
+      "step": 7368
+    },
+    {
+      "epoch": 3.272929158338885,
+      "grad_norm": 0.4720762654089572,
+      "learning_rate": 9.713859685122428e-07,
+      "loss": 0.0267,
+      "step": 7369
+    },
+    {
+      "epoch": 3.273373306684433,
+      "grad_norm": 0.41020011208140905,
+      "learning_rate": 9.702380899697621e-07,
+      "loss": 0.0231,
+      "step": 7370
+    },
+    {
+      "epoch": 3.27381745502998,
+      "grad_norm": 0.51306227756855,
+      "learning_rate": 9.69090817168949e-07,
+      "loss": 0.0213,
+      "step": 7371
+    },
+    {
+      "epoch": 3.2742616033755274,
+      "grad_norm": 0.5514386921409069,
+      "learning_rate": 9.67944150282259e-07,
+      "loss": 0.0388,
+      "step": 7372
+    },
+    {
+      "epoch": 3.2747057517210747,
+      "grad_norm": 0.47876201361463355,
+      "learning_rate": 9.667980894820572e-07,
+      "loss": 0.0402,
+      "step": 7373
+    },
+    {
+      "epoch": 3.2751499000666224,
+      "grad_norm": 0.35596410172080417,
+      "learning_rate": 9.65652634940612e-07,
+      "loss": 0.0211,
+      "step": 7374
+    },
+    {
+      "epoch": 3.2755940484121697,
+      "grad_norm": 0.5330920908251411,
+      "learning_rate": 9.64507786830109e-07,
+      "loss": 0.0312,
+      "step": 7375
+    },
+    {
+      "epoch": 3.276038196757717,
+      "grad_norm": 0.44331071434194597,
+      "learning_rate": 9.633635453226376e-07,
+      "loss": 0.0241,
+      "step": 7376
+    },
+    {
+      "epoch": 3.2764823451032643,
+      "grad_norm": 0.4178851244614469,
+      "learning_rate": 9.622199105901947e-07,
+      "loss": 0.0251,
+      "step": 7377
+    },
+    {
+      "epoch": 3.276926493448812,
+      "grad_norm": 0.4925055411838903,
+      "learning_rate": 9.610768828046891e-07,
+      "loss": 0.0312,
+      "step": 7378
+    },
+    {
+      "epoch": 3.2773706417943593,
+      "grad_norm": 0.3674042441682116,
+      "learning_rate": 9.59934462137938e-07,
+      "loss": 0.0176,
+      "step": 7379
+    },
+    {
+      "epoch": 3.2778147901399066,
+      "grad_norm": 0.41837211074854475,
+      "learning_rate": 9.58792648761664e-07,
+      "loss": 0.0272,
+      "step": 7380
+    },
+    {
+      "epoch": 3.278258938485454,
+      "grad_norm": 0.35983457178846384,
+      "learning_rate": 9.576514428475058e-07,
+      "loss": 0.0182,
+      "step": 7381
+    },
+    {
+      "epoch": 3.2787030868310016,
+      "grad_norm": 0.5213483263717448,
+      "learning_rate": 9.565108445670013e-07,
+      "loss": 0.0285,
+      "step": 7382
+    },
+    {
+      "epoch": 3.279147235176549,
+      "grad_norm": 0.3798413510246776,
+      "learning_rate": 9.55370854091604e-07,
+      "loss": 0.0199,
+      "step": 7383
+    },
+    {
+      "epoch": 3.2795913835220962,
+      "grad_norm": 0.40316029556208016,
+      "learning_rate": 9.542314715926753e-07,
+      "loss": 0.023,
+      "step": 7384
+    },
+    {
+      "epoch": 3.280035531867644,
+      "grad_norm": 0.3279127936611918,
+      "learning_rate": 9.5309269724148e-07,
+      "loss": 0.0197,
+      "step": 7385
+    },
+    {
+      "epoch": 3.2804796802131913,
+      "grad_norm": 0.3077370020937381,
+      "learning_rate": 9.519545312091966e-07,
+      "loss": 0.0186,
+      "step": 7386
+    },
+    {
+      "epoch": 3.2809238285587385,
+      "grad_norm": 0.29637415446133053,
+      "learning_rate": 9.508169736669137e-07,
+      "loss": 0.0149,
+      "step": 7387
+    },
+    {
+      "epoch": 3.281367976904286,
+      "grad_norm": 0.36806133901057364,
+      "learning_rate": 9.496800247856219e-07,
+      "loss": 0.0255,
+      "step": 7388
+    },
+    {
+      "epoch": 3.2818121252498336,
+      "grad_norm": 0.42318209672526275,
+      "learning_rate": 9.485436847362257e-07,
+      "loss": 0.0368,
+      "step": 7389
+    },
+    {
+      "epoch": 3.282256273595381,
+      "grad_norm": 0.4481117630581104,
+      "learning_rate": 9.474079536895365e-07,
+      "loss": 0.0323,
+      "step": 7390
+    },
+    {
+      "epoch": 3.282700421940928,
+      "grad_norm": 0.4094546536057865,
+      "learning_rate": 9.462728318162712e-07,
+      "loss": 0.0216,
+      "step": 7391
+    },
+    {
+      "epoch": 3.283144570286476,
+      "grad_norm": 0.4833214117317346,
+      "learning_rate": 9.451383192870623e-07,
+      "loss": 0.0314,
+      "step": 7392
+    },
+    {
+      "epoch": 3.283588718632023,
+      "grad_norm": 0.49172490663115703,
+      "learning_rate": 9.440044162724432e-07,
+      "loss": 0.0229,
+      "step": 7393
+    },
+    {
+      "epoch": 3.2840328669775705,
+      "grad_norm": 0.3837206768061979,
+      "learning_rate": 9.428711229428594e-07,
+      "loss": 0.0281,
+      "step": 7394
+    },
+    {
+      "epoch": 3.2844770153231178,
+      "grad_norm": 0.38763854772417794,
+      "learning_rate": 9.417384394686646e-07,
+      "loss": 0.0256,
+      "step": 7395
+    },
+    {
+      "epoch": 3.2849211636686655,
+      "grad_norm": 0.4292993654748122,
+      "learning_rate": 9.406063660201214e-07,
+      "loss": 0.0178,
+      "step": 7396
+    },
+    {
+      "epoch": 3.285365312014213,
+      "grad_norm": 0.4608324783734576,
+      "learning_rate": 9.394749027673955e-07,
+      "loss": 0.03,
+      "step": 7397
+    },
+    {
+      "epoch": 3.28580946035976,
+      "grad_norm": 0.4116770755294882,
+      "learning_rate": 9.383440498805712e-07,
+      "loss": 0.0266,
+      "step": 7398
+    },
+    {
+      "epoch": 3.286253608705308,
+      "grad_norm": 0.4151574100082846,
+      "learning_rate": 9.3721380752963e-07,
+      "loss": 0.0316,
+      "step": 7399
+    },
+    {
+      "epoch": 3.286697757050855,
+      "grad_norm": 0.5201578859027275,
+      "learning_rate": 9.36084175884468e-07,
+      "loss": 0.0273,
+      "step": 7400
+    },
+    {
+      "epoch": 3.2871419053964024,
+      "grad_norm": 0.517408046194756,
+      "learning_rate": 9.3495515511489e-07,
+      "loss": 0.039,
+      "step": 7401
+    },
+    {
+      "epoch": 3.2875860537419497,
+      "grad_norm": 0.4065347285711147,
+      "learning_rate": 9.338267453906036e-07,
+      "loss": 0.0193,
+      "step": 7402
+    },
+    {
+      "epoch": 3.2880302020874974,
+      "grad_norm": 0.35637207302672474,
+      "learning_rate": 9.326989468812281e-07,
+      "loss": 0.0252,
+      "step": 7403
+    },
+    {
+      "epoch": 3.2884743504330447,
+      "grad_norm": 0.3942806613898594,
+      "learning_rate": 9.315717597562951e-07,
+      "loss": 0.0224,
+      "step": 7404
+    },
+    {
+      "epoch": 3.288918498778592,
+      "grad_norm": 0.38601773480071766,
+      "learning_rate": 9.304451841852358e-07,
+      "loss": 0.0291,
+      "step": 7405
+    },
+    {
+      "epoch": 3.2893626471241393,
+      "grad_norm": 0.4412874940460958,
+      "learning_rate": 9.293192203373952e-07,
+      "loss": 0.0223,
+      "step": 7406
+    },
+    {
+      "epoch": 3.289806795469687,
+      "grad_norm": 0.36047506716707345,
+      "learning_rate": 9.281938683820258e-07,
+      "loss": 0.0196,
+      "step": 7407
+    },
+    {
+      "epoch": 3.2902509438152343,
+      "grad_norm": 0.38105181354876544,
+      "learning_rate": 9.270691284882826e-07,
+      "loss": 0.0215,
+      "step": 7408
+    },
+    {
+      "epoch": 3.2906950921607816,
+      "grad_norm": 0.5134105812079635,
+      "learning_rate": 9.259450008252396e-07,
+      "loss": 0.0407,
+      "step": 7409
+    },
+    {
+      "epoch": 3.291139240506329,
+      "grad_norm": 0.3562398576839381,
+      "learning_rate": 9.248214855618676e-07,
+      "loss": 0.0267,
+      "step": 7410
+    },
+    {
+      "epoch": 3.2915833888518766,
+      "grad_norm": 0.40663195402797286,
+      "learning_rate": 9.236985828670519e-07,
+      "loss": 0.0187,
+      "step": 7411
+    },
+    {
+      "epoch": 3.292027537197424,
+      "grad_norm": 0.7855995128015365,
+      "learning_rate": 9.225762929095844e-07,
+      "loss": 0.0372,
+      "step": 7412
+    },
+    {
+      "epoch": 3.292471685542971,
+      "grad_norm": 0.4097302898181189,
+      "learning_rate": 9.214546158581622e-07,
+      "loss": 0.0252,
+      "step": 7413
+    },
+    {
+      "epoch": 3.292915833888519,
+      "grad_norm": 0.4179153487359494,
+      "learning_rate": 9.203335518813922e-07,
+      "loss": 0.0285,
+      "step": 7414
+    },
+    {
+      "epoch": 3.2933599822340662,
+      "grad_norm": 0.5431971211223403,
+      "learning_rate": 9.192131011477934e-07,
+      "loss": 0.0303,
+      "step": 7415
+    },
+    {
+      "epoch": 3.2938041305796135,
+      "grad_norm": 0.5065980372967815,
+      "learning_rate": 9.180932638257845e-07,
+      "loss": 0.028,
+      "step": 7416
+    },
+    {
+      "epoch": 3.294248278925161,
+      "grad_norm": 0.38822134135830527,
+      "learning_rate": 9.169740400836974e-07,
+      "loss": 0.029,
+      "step": 7417
+    },
+    {
+      "epoch": 3.2946924272707085,
+      "grad_norm": 0.3332433458833639,
+      "learning_rate": 9.158554300897727e-07,
+      "loss": 0.0174,
+      "step": 7418
+    },
+    {
+      "epoch": 3.295136575616256,
+      "grad_norm": 0.3544713822711631,
+      "learning_rate": 9.147374340121523e-07,
+      "loss": 0.022,
+      "step": 7419
+    },
+    {
+      "epoch": 3.295580723961803,
+      "grad_norm": 0.32602497933389224,
+      "learning_rate": 9.13620052018892e-07,
+      "loss": 0.0205,
+      "step": 7420
+    },
+    {
+      "epoch": 3.296024872307351,
+      "grad_norm": 0.45839560888184433,
+      "learning_rate": 9.125032842779535e-07,
+      "loss": 0.0285,
+      "step": 7421
+    },
+    {
+      "epoch": 3.296469020652898,
+      "grad_norm": 0.4451229775575657,
+      "learning_rate": 9.113871309572059e-07,
+      "loss": 0.0314,
+      "step": 7422
+    },
+    {
+      "epoch": 3.2969131689984454,
+      "grad_norm": 0.4294843373554615,
+      "learning_rate": 9.10271592224426e-07,
+      "loss": 0.0279,
+      "step": 7423
+    },
+    {
+      "epoch": 3.2973573173439927,
+      "grad_norm": 0.49025642020365773,
+      "learning_rate": 9.091566682472991e-07,
+      "loss": 0.0319,
+      "step": 7424
+    },
+    {
+      "epoch": 3.2978014656895405,
+      "grad_norm": 0.39446663575000246,
+      "learning_rate": 9.08042359193414e-07,
+      "loss": 0.0264,
+      "step": 7425
+    },
+    {
+      "epoch": 3.2982456140350878,
+      "grad_norm": 0.3073338148065069,
+      "learning_rate": 9.06928665230275e-07,
+      "loss": 0.022,
+      "step": 7426
+    },
+    {
+      "epoch": 3.298689762380635,
+      "grad_norm": 0.5363424487989449,
+      "learning_rate": 9.058155865252854e-07,
+      "loss": 0.0369,
+      "step": 7427
+    },
+    {
+      "epoch": 3.299133910726183,
+      "grad_norm": 0.4241449124561232,
+      "learning_rate": 9.047031232457609e-07,
+      "loss": 0.0284,
+      "step": 7428
+    },
+    {
+      "epoch": 3.29957805907173,
+      "grad_norm": 0.5039477021882938,
+      "learning_rate": 9.035912755589254e-07,
+      "loss": 0.0296,
+      "step": 7429
+    },
+    {
+      "epoch": 3.3000222074172774,
+      "grad_norm": 0.6325313716948877,
+      "learning_rate": 9.024800436319059e-07,
+      "loss": 0.0356,
+      "step": 7430
+    },
+    {
+      "epoch": 3.3004663557628247,
+      "grad_norm": 0.35326550695737485,
+      "learning_rate": 9.013694276317392e-07,
+      "loss": 0.0293,
+      "step": 7431
+    },
+    {
+      "epoch": 3.3009105041083724,
+      "grad_norm": 0.3943172052588995,
+      "learning_rate": 9.002594277253735e-07,
+      "loss": 0.0256,
+      "step": 7432
+    },
+    {
+      "epoch": 3.3013546524539197,
+      "grad_norm": 0.3559437962740476,
+      "learning_rate": 8.991500440796569e-07,
+      "loss": 0.0261,
+      "step": 7433
+    },
+    {
+      "epoch": 3.301798800799467,
+      "grad_norm": 0.3905214750787488,
+      "learning_rate": 8.9804127686135e-07,
+      "loss": 0.0166,
+      "step": 7434
+    },
+    {
+      "epoch": 3.3022429491450143,
+      "grad_norm": 0.5100939910295724,
+      "learning_rate": 8.969331262371206e-07,
+      "loss": 0.0305,
+      "step": 7435
+    },
+    {
+      "epoch": 3.302687097490562,
+      "grad_norm": 0.4081830294338292,
+      "learning_rate": 8.958255923735404e-07,
+      "loss": 0.0229,
+      "step": 7436
+    },
+    {
+      "epoch": 3.3031312458361093,
+      "grad_norm": 0.41704083656049834,
+      "learning_rate": 8.947186754370907e-07,
+      "loss": 0.0281,
+      "step": 7437
+    },
+    {
+      "epoch": 3.3035753941816566,
+      "grad_norm": 0.4789522896492798,
+      "learning_rate": 8.936123755941611e-07,
+      "loss": 0.0309,
+      "step": 7438
+    },
+    {
+      "epoch": 3.304019542527204,
+      "grad_norm": 0.5252768558443547,
+      "learning_rate": 8.925066930110465e-07,
+      "loss": 0.0265,
+      "step": 7439
+    },
+    {
+      "epoch": 3.3044636908727516,
+      "grad_norm": 0.34510316412181224,
+      "learning_rate": 8.914016278539516e-07,
+      "loss": 0.019,
+      "step": 7440
+    },
+    {
+      "epoch": 3.304907839218299,
+      "grad_norm": 0.5857701871825775,
+      "learning_rate": 8.902971802889832e-07,
+      "loss": 0.0387,
+      "step": 7441
+    },
+    {
+      "epoch": 3.305351987563846,
+      "grad_norm": 0.5033019496461394,
+      "learning_rate": 8.891933504821604e-07,
+      "loss": 0.0275,
+      "step": 7442
+    },
+    {
+      "epoch": 3.305796135909394,
+      "grad_norm": 0.3999151889282477,
+      "learning_rate": 8.880901385994079e-07,
+      "loss": 0.0233,
+      "step": 7443
+    },
+    {
+      "epoch": 3.306240284254941,
+      "grad_norm": 0.45097536087731294,
+      "learning_rate": 8.869875448065563e-07,
+      "loss": 0.0281,
+      "step": 7444
+    },
+    {
+      "epoch": 3.3066844326004885,
+      "grad_norm": 0.4146176937666059,
+      "learning_rate": 8.858855692693446e-07,
+      "loss": 0.0294,
+      "step": 7445
+    },
+    {
+      "epoch": 3.307128580946036,
+      "grad_norm": 0.32341587117342246,
+      "learning_rate": 8.847842121534195e-07,
+      "loss": 0.0254,
+      "step": 7446
+    },
+    {
+      "epoch": 3.3075727292915835,
+      "grad_norm": 0.3794842748021583,
+      "learning_rate": 8.836834736243316e-07,
+      "loss": 0.0219,
+      "step": 7447
+    },
+    {
+      "epoch": 3.308016877637131,
+      "grad_norm": 0.5092740919087244,
+      "learning_rate": 8.825833538475403e-07,
+      "loss": 0.0279,
+      "step": 7448
+    },
+    {
+      "epoch": 3.308461025982678,
+      "grad_norm": 0.3833821462873475,
+      "learning_rate": 8.814838529884162e-07,
+      "loss": 0.0263,
+      "step": 7449
+    },
+    {
+      "epoch": 3.308905174328226,
+      "grad_norm": 0.7887453397226605,
+      "learning_rate": 8.803849712122292e-07,
+      "loss": 0.0208,
+      "step": 7450
+    },
+    {
+      "epoch": 3.309349322673773,
+      "grad_norm": 0.3907962831950114,
+      "learning_rate": 8.792867086841605e-07,
+      "loss": 0.0206,
+      "step": 7451
+    },
+    {
+      "epoch": 3.3097934710193204,
+      "grad_norm": 0.3760419094116477,
+      "learning_rate": 8.781890655692998e-07,
+      "loss": 0.022,
+      "step": 7452
+    },
+    {
+      "epoch": 3.3102376193648677,
+      "grad_norm": 0.32255052704787457,
+      "learning_rate": 8.770920420326384e-07,
+      "loss": 0.0181,
+      "step": 7453
+    },
+    {
+      "epoch": 3.3106817677104154,
+      "grad_norm": 0.6578667511681319,
+      "learning_rate": 8.759956382390794e-07,
+      "loss": 0.0254,
+      "step": 7454
+    },
+    {
+      "epoch": 3.3111259160559627,
+      "grad_norm": 0.42423201463173366,
+      "learning_rate": 8.748998543534304e-07,
+      "loss": 0.0331,
+      "step": 7455
+    },
+    {
+      "epoch": 3.31157006440151,
+      "grad_norm": 0.4695388522211957,
+      "learning_rate": 8.738046905404069e-07,
+      "loss": 0.04,
+      "step": 7456
+    },
+    {
+      "epoch": 3.3120142127470578,
+      "grad_norm": 0.488606567568898,
+      "learning_rate": 8.72710146964631e-07,
+      "loss": 0.0274,
+      "step": 7457
+    },
+    {
+      "epoch": 3.312458361092605,
+      "grad_norm": 0.5406502377024667,
+      "learning_rate": 8.716162237906289e-07,
+      "loss": 0.028,
+      "step": 7458
+    },
+    {
+      "epoch": 3.3129025094381523,
+      "grad_norm": 0.5699097094725243,
+      "learning_rate": 8.705229211828376e-07,
+      "loss": 0.03,
+      "step": 7459
+    },
+    {
+      "epoch": 3.3133466577836996,
+      "grad_norm": 0.44717354800350295,
+      "learning_rate": 8.694302393055992e-07,
+      "loss": 0.0208,
+      "step": 7460
+    },
+    {
+      "epoch": 3.313790806129247,
+      "grad_norm": 0.4253321381991377,
+      "learning_rate": 8.683381783231615e-07,
+      "loss": 0.0234,
+      "step": 7461
+    },
+    {
+      "epoch": 3.3142349544747947,
+      "grad_norm": 0.3858599803822746,
+      "learning_rate": 8.672467383996802e-07,
+      "loss": 0.0281,
+      "step": 7462
+    },
+    {
+      "epoch": 3.314679102820342,
+      "grad_norm": 1.0071452454502907,
+      "learning_rate": 8.661559196992186e-07,
+      "loss": 0.029,
+      "step": 7463
+    },
+    {
+      "epoch": 3.3151232511658892,
+      "grad_norm": 0.5059569170522437,
+      "learning_rate": 8.650657223857428e-07,
+      "loss": 0.0304,
+      "step": 7464
+    },
+    {
+      "epoch": 3.315567399511437,
+      "grad_norm": 0.4605836239110675,
+      "learning_rate": 8.639761466231294e-07,
+      "loss": 0.0374,
+      "step": 7465
+    },
+    {
+      "epoch": 3.3160115478569843,
+      "grad_norm": 0.5663222302112452,
+      "learning_rate": 8.628871925751598e-07,
+      "loss": 0.029,
+      "step": 7466
+    },
+    {
+      "epoch": 3.3164556962025316,
+      "grad_norm": 0.4770838107638845,
+      "learning_rate": 8.617988604055222e-07,
+      "loss": 0.029,
+      "step": 7467
+    },
+    {
+      "epoch": 3.316899844548079,
+      "grad_norm": 0.5186866939183644,
+      "learning_rate": 8.607111502778121e-07,
+      "loss": 0.0367,
+      "step": 7468
+    },
+    {
+      "epoch": 3.3173439928936266,
+      "grad_norm": 0.43011652502356673,
+      "learning_rate": 8.596240623555313e-07,
+      "loss": 0.0276,
+      "step": 7469
+    },
+    {
+      "epoch": 3.317788141239174,
+      "grad_norm": 0.39270558043084164,
+      "learning_rate": 8.585375968020854e-07,
+      "loss": 0.0244,
+      "step": 7470
+    },
+    {
+      "epoch": 3.318232289584721,
+      "grad_norm": 0.3922390336468883,
+      "learning_rate": 8.574517537807897e-07,
+      "loss": 0.0247,
+      "step": 7471
+    },
+    {
+      "epoch": 3.318676437930269,
+      "grad_norm": 0.4905605263072924,
+      "learning_rate": 8.563665334548654e-07,
+      "loss": 0.0281,
+      "step": 7472
+    },
+    {
+      "epoch": 3.319120586275816,
+      "grad_norm": 0.4695814471855208,
+      "learning_rate": 8.552819359874387e-07,
+      "loss": 0.0222,
+      "step": 7473
+    },
+    {
+      "epoch": 3.3195647346213635,
+      "grad_norm": 0.49486185686147216,
+      "learning_rate": 8.541979615415446e-07,
+      "loss": 0.0304,
+      "step": 7474
+    },
+    {
+      "epoch": 3.3200088829669108,
+      "grad_norm": 0.40290781496927697,
+      "learning_rate": 8.531146102801208e-07,
+      "loss": 0.0221,
+      "step": 7475
+    },
+    {
+      "epoch": 3.3204530313124585,
+      "grad_norm": 0.45538047230870293,
+      "learning_rate": 8.520318823660146e-07,
+      "loss": 0.0257,
+      "step": 7476
+    },
+    {
+      "epoch": 3.320897179658006,
+      "grad_norm": 0.4425830829496296,
+      "learning_rate": 8.50949777961978e-07,
+      "loss": 0.027,
+      "step": 7477
+    },
+    {
+      "epoch": 3.321341328003553,
+      "grad_norm": 0.32164145024703766,
+      "learning_rate": 8.498682972306693e-07,
+      "loss": 0.0171,
+      "step": 7478
+    },
+    {
+      "epoch": 3.321785476349101,
+      "grad_norm": 0.33243355641729805,
+      "learning_rate": 8.487874403346547e-07,
+      "loss": 0.0171,
+      "step": 7479
+    },
+    {
+      "epoch": 3.322229624694648,
+      "grad_norm": 0.4774020513789674,
+      "learning_rate": 8.477072074364051e-07,
+      "loss": 0.0242,
+      "step": 7480
+    },
+    {
+      "epoch": 3.3226737730401954,
+      "grad_norm": 0.7842160045286404,
+      "learning_rate": 8.466275986982963e-07,
+      "loss": 0.0482,
+      "step": 7481
+    },
+    {
+      "epoch": 3.3231179213857427,
+      "grad_norm": 0.4462293368851325,
+      "learning_rate": 8.455486142826135e-07,
+      "loss": 0.0206,
+      "step": 7482
+    },
+    {
+      "epoch": 3.3235620697312904,
+      "grad_norm": 0.3371098785501344,
+      "learning_rate": 8.444702543515454e-07,
+      "loss": 0.016,
+      "step": 7483
+    },
+    {
+      "epoch": 3.3240062180768377,
+      "grad_norm": 0.4689818865427122,
+      "learning_rate": 8.433925190671876e-07,
+      "loss": 0.0294,
+      "step": 7484
+    },
+    {
+      "epoch": 3.324450366422385,
+      "grad_norm": 0.46126294870782475,
+      "learning_rate": 8.423154085915447e-07,
+      "loss": 0.0405,
+      "step": 7485
+    },
+    {
+      "epoch": 3.3248945147679323,
+      "grad_norm": 0.4269127171229397,
+      "learning_rate": 8.412389230865209e-07,
+      "loss": 0.0215,
+      "step": 7486
+    },
+    {
+      "epoch": 3.32533866311348,
+      "grad_norm": 0.3601691103633931,
+      "learning_rate": 8.401630627139317e-07,
+      "loss": 0.0283,
+      "step": 7487
+    },
+    {
+      "epoch": 3.3257828114590273,
+      "grad_norm": 0.3308167717075761,
+      "learning_rate": 8.39087827635498e-07,
+      "loss": 0.0203,
+      "step": 7488
+    },
+    {
+      "epoch": 3.3262269598045746,
+      "grad_norm": 0.4504744849786426,
+      "learning_rate": 8.380132180128453e-07,
+      "loss": 0.0308,
+      "step": 7489
+    },
+    {
+      "epoch": 3.326671108150122,
+      "grad_norm": 0.43274682572551426,
+      "learning_rate": 8.369392340075056e-07,
+      "loss": 0.0192,
+      "step": 7490
+    },
+    {
+      "epoch": 3.3271152564956696,
+      "grad_norm": 0.38752988514350534,
+      "learning_rate": 8.358658757809179e-07,
+      "loss": 0.0326,
+      "step": 7491
+    },
+    {
+      "epoch": 3.327559404841217,
+      "grad_norm": 0.3966193669552188,
+      "learning_rate": 8.347931434944245e-07,
+      "loss": 0.0218,
+      "step": 7492
+    },
+    {
+      "epoch": 3.328003553186764,
+      "grad_norm": 0.37005902190632034,
+      "learning_rate": 8.337210373092763e-07,
+      "loss": 0.02,
+      "step": 7493
+    },
+    {
+      "epoch": 3.328447701532312,
+      "grad_norm": 0.3982123833521082,
+      "learning_rate": 8.326495573866284e-07,
+      "loss": 0.0216,
+      "step": 7494
+    },
+    {
+      "epoch": 3.3288918498778592,
+      "grad_norm": 0.41678676657338065,
+      "learning_rate": 8.315787038875434e-07,
+      "loss": 0.0284,
+      "step": 7495
+    },
+    {
+      "epoch": 3.3293359982234065,
+      "grad_norm": 0.36841312927749376,
+      "learning_rate": 8.305084769729882e-07,
+      "loss": 0.0294,
+      "step": 7496
+    },
+    {
+      "epoch": 3.329780146568954,
+      "grad_norm": 0.42365099250088206,
+      "learning_rate": 8.294388768038375e-07,
+      "loss": 0.0294,
+      "step": 7497
+    },
+    {
+      "epoch": 3.3302242949145016,
+      "grad_norm": 0.40673867597662916,
+      "learning_rate": 8.283699035408677e-07,
+      "loss": 0.0235,
+      "step": 7498
+    },
+    {
+      "epoch": 3.330668443260049,
+      "grad_norm": 0.40817718954362553,
+      "learning_rate": 8.273015573447646e-07,
+      "loss": 0.0229,
+      "step": 7499
+    },
+    {
+      "epoch": 3.331112591605596,
+      "grad_norm": 0.47791402955813106,
+      "learning_rate": 8.262338383761199e-07,
+      "loss": 0.0306,
+      "step": 7500
+    },
+    {
+      "epoch": 3.331556739951144,
+      "grad_norm": 0.3730312585666912,
+      "learning_rate": 8.251667467954289e-07,
+      "loss": 0.0241,
+      "step": 7501
+    },
+    {
+      "epoch": 3.332000888296691,
+      "grad_norm": 0.5006016285563061,
+      "learning_rate": 8.241002827630945e-07,
+      "loss": 0.0284,
+      "step": 7502
+    },
+    {
+      "epoch": 3.3324450366422385,
+      "grad_norm": 0.5039257261200133,
+      "learning_rate": 8.230344464394236e-07,
+      "loss": 0.0234,
+      "step": 7503
+    },
+    {
+      "epoch": 3.3328891849877857,
+      "grad_norm": 0.38046904572274926,
+      "learning_rate": 8.219692379846289e-07,
+      "loss": 0.033,
+      "step": 7504
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.3757296330009624,
+      "learning_rate": 8.209046575588303e-07,
+      "loss": 0.0291,
+      "step": 7505
+    },
+    {
+      "epoch": 3.3337774816788808,
+      "grad_norm": 0.380415197644764,
+      "learning_rate": 8.198407053220519e-07,
+      "loss": 0.0219,
+      "step": 7506
+    },
+    {
+      "epoch": 3.334221630024428,
+      "grad_norm": 0.39647003849848267,
+      "learning_rate": 8.187773814342242e-07,
+      "loss": 0.0307,
+      "step": 7507
+    },
+    {
+      "epoch": 3.334665778369976,
+      "grad_norm": 0.48671927289549416,
+      "learning_rate": 8.177146860551838e-07,
+      "loss": 0.0289,
+      "step": 7508
+    },
+    {
+      "epoch": 3.335109926715523,
+      "grad_norm": 0.4169191245545791,
+      "learning_rate": 8.166526193446695e-07,
+      "loss": 0.0286,
+      "step": 7509
+    },
+    {
+      "epoch": 3.3355540750610704,
+      "grad_norm": 0.6209631358998482,
+      "learning_rate": 8.155911814623291e-07,
+      "loss": 0.0268,
+      "step": 7510
+    },
+    {
+      "epoch": 3.3359982234066177,
+      "grad_norm": 0.4569957343588699,
+      "learning_rate": 8.145303725677145e-07,
+      "loss": 0.025,
+      "step": 7511
+    },
+    {
+      "epoch": 3.3364423717521654,
+      "grad_norm": 0.45123275529280976,
+      "learning_rate": 8.134701928202843e-07,
+      "loss": 0.0207,
+      "step": 7512
+    },
+    {
+      "epoch": 3.3368865200977127,
+      "grad_norm": 0.3644402406540846,
+      "learning_rate": 8.124106423794015e-07,
+      "loss": 0.0156,
+      "step": 7513
+    },
+    {
+      "epoch": 3.33733066844326,
+      "grad_norm": 0.6935267382582291,
+      "learning_rate": 8.113517214043326e-07,
+      "loss": 0.042,
+      "step": 7514
+    },
+    {
+      "epoch": 3.3377748167888073,
+      "grad_norm": 0.40069923978491,
+      "learning_rate": 8.102934300542531e-07,
+      "loss": 0.0282,
+      "step": 7515
+    },
+    {
+      "epoch": 3.338218965134355,
+      "grad_norm": 0.4259846167113701,
+      "learning_rate": 8.092357684882413e-07,
+      "loss": 0.0289,
+      "step": 7516
+    },
+    {
+      "epoch": 3.3386631134799023,
+      "grad_norm": 0.4192011063222261,
+      "learning_rate": 8.081787368652822e-07,
+      "loss": 0.023,
+      "step": 7517
+    },
+    {
+      "epoch": 3.3391072618254496,
+      "grad_norm": 0.35804521930200645,
+      "learning_rate": 8.071223353442658e-07,
+      "loss": 0.0294,
+      "step": 7518
+    },
+    {
+      "epoch": 3.339551410170997,
+      "grad_norm": 0.3873707899082061,
+      "learning_rate": 8.060665640839882e-07,
+      "loss": 0.0228,
+      "step": 7519
+    },
+    {
+      "epoch": 3.3399955585165446,
+      "grad_norm": 0.33653268211511883,
+      "learning_rate": 8.050114232431472e-07,
+      "loss": 0.0229,
+      "step": 7520
+    },
+    {
+      "epoch": 3.340439706862092,
+      "grad_norm": 0.4424670871666316,
+      "learning_rate": 8.039569129803493e-07,
+      "loss": 0.0273,
+      "step": 7521
+    },
+    {
+      "epoch": 3.340883855207639,
+      "grad_norm": 0.36754810698365975,
+      "learning_rate": 8.029030334541061e-07,
+      "loss": 0.0212,
+      "step": 7522
+    },
+    {
+      "epoch": 3.341328003553187,
+      "grad_norm": 0.40011975008051315,
+      "learning_rate": 8.01849784822833e-07,
+      "loss": 0.0213,
+      "step": 7523
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.4091440045335287,
+      "learning_rate": 8.007971672448511e-07,
+      "loss": 0.0326,
+      "step": 7524
+    },
+    {
+      "epoch": 3.3422163002442815,
+      "grad_norm": 0.3979067624482782,
+      "learning_rate": 7.997451808783884e-07,
+      "loss": 0.0292,
+      "step": 7525
+    },
+    {
+      "epoch": 3.342660448589829,
+      "grad_norm": 0.4350141704240594,
+      "learning_rate": 7.986938258815741e-07,
+      "loss": 0.0294,
+      "step": 7526
+    },
+    {
+      "epoch": 3.3431045969353765,
+      "grad_norm": 0.3288238940788284,
+      "learning_rate": 7.976431024124448e-07,
+      "loss": 0.02,
+      "step": 7527
+    },
+    {
+      "epoch": 3.343548745280924,
+      "grad_norm": 0.40499639306856394,
+      "learning_rate": 7.965930106289432e-07,
+      "loss": 0.0249,
+      "step": 7528
+    },
+    {
+      "epoch": 3.343992893626471,
+      "grad_norm": 0.4511936733531736,
+      "learning_rate": 7.955435506889154e-07,
+      "loss": 0.0225,
+      "step": 7529
+    },
+    {
+      "epoch": 3.344437041972019,
+      "grad_norm": 0.32996474391647085,
+      "learning_rate": 7.944947227501143e-07,
+      "loss": 0.019,
+      "step": 7530
+    },
+    {
+      "epoch": 3.344881190317566,
+      "grad_norm": 0.33989523301533825,
+      "learning_rate": 7.934465269701941e-07,
+      "loss": 0.0281,
+      "step": 7531
+    },
+    {
+      "epoch": 3.3453253386631134,
+      "grad_norm": 0.5424696702981813,
+      "learning_rate": 7.923989635067181e-07,
+      "loss": 0.0341,
+      "step": 7532
+    },
+    {
+      "epoch": 3.3457694870086607,
+      "grad_norm": 0.5354455170131127,
+      "learning_rate": 7.913520325171537e-07,
+      "loss": 0.0332,
+      "step": 7533
+    },
+    {
+      "epoch": 3.3462136353542085,
+      "grad_norm": 0.5188210769342587,
+      "learning_rate": 7.903057341588683e-07,
+      "loss": 0.0346,
+      "step": 7534
+    },
+    {
+      "epoch": 3.3466577836997558,
+      "grad_norm": 0.32324444187136564,
+      "learning_rate": 7.892600685891433e-07,
+      "loss": 0.0202,
+      "step": 7535
+    },
+    {
+      "epoch": 3.347101932045303,
+      "grad_norm": 0.38834278141098566,
+      "learning_rate": 7.882150359651586e-07,
+      "loss": 0.0233,
+      "step": 7536
+    },
+    {
+      "epoch": 3.3475460803908508,
+      "grad_norm": 0.44587329880583815,
+      "learning_rate": 7.871706364439985e-07,
+      "loss": 0.0285,
+      "step": 7537
+    },
+    {
+      "epoch": 3.347990228736398,
+      "grad_norm": 0.38089139300391106,
+      "learning_rate": 7.861268701826552e-07,
+      "loss": 0.0264,
+      "step": 7538
+    },
+    {
+      "epoch": 3.3484343770819454,
+      "grad_norm": 0.3825436500451749,
+      "learning_rate": 7.850837373380244e-07,
+      "loss": 0.0206,
+      "step": 7539
+    },
+    {
+      "epoch": 3.3488785254274926,
+      "grad_norm": 0.3827903172923659,
+      "learning_rate": 7.840412380669071e-07,
+      "loss": 0.0221,
+      "step": 7540
+    },
+    {
+      "epoch": 3.3493226737730404,
+      "grad_norm": 0.38625911378052546,
+      "learning_rate": 7.829993725260082e-07,
+      "loss": 0.0192,
+      "step": 7541
+    },
+    {
+      "epoch": 3.3497668221185877,
+      "grad_norm": 0.37721835784709495,
+      "learning_rate": 7.81958140871939e-07,
+      "loss": 0.0291,
+      "step": 7542
+    },
+    {
+      "epoch": 3.350210970464135,
+      "grad_norm": 0.39810907293172026,
+      "learning_rate": 7.809175432612126e-07,
+      "loss": 0.0239,
+      "step": 7543
+    },
+    {
+      "epoch": 3.3506551188096823,
+      "grad_norm": 1.1010267945006187,
+      "learning_rate": 7.798775798502484e-07,
+      "loss": 0.0314,
+      "step": 7544
+    },
+    {
+      "epoch": 3.35109926715523,
+      "grad_norm": 0.4353628491116733,
+      "learning_rate": 7.788382507953718e-07,
+      "loss": 0.022,
+      "step": 7545
+    },
+    {
+      "epoch": 3.3515434155007773,
+      "grad_norm": 0.392235080799356,
+      "learning_rate": 7.777995562528107e-07,
+      "loss": 0.0187,
+      "step": 7546
+    },
+    {
+      "epoch": 3.3519875638463246,
+      "grad_norm": 0.4931466449389555,
+      "learning_rate": 7.767614963787007e-07,
+      "loss": 0.0176,
+      "step": 7547
+    },
+    {
+      "epoch": 3.352431712191872,
+      "grad_norm": 0.39853671183474193,
+      "learning_rate": 7.757240713290764e-07,
+      "loss": 0.0259,
+      "step": 7548
+    },
+    {
+      "epoch": 3.3528758605374196,
+      "grad_norm": 0.4566981980981673,
+      "learning_rate": 7.746872812598821e-07,
+      "loss": 0.0255,
+      "step": 7549
+    },
+    {
+      "epoch": 3.353320008882967,
+      "grad_norm": 0.4075307870306547,
+      "learning_rate": 7.736511263269664e-07,
+      "loss": 0.0259,
+      "step": 7550
+    },
+    {
+      "epoch": 3.353764157228514,
+      "grad_norm": 0.44508580022829525,
+      "learning_rate": 7.726156066860769e-07,
+      "loss": 0.0221,
+      "step": 7551
+    },
+    {
+      "epoch": 3.354208305574062,
+      "grad_norm": 0.5633393812829065,
+      "learning_rate": 7.715807224928734e-07,
+      "loss": 0.0295,
+      "step": 7552
+    },
+    {
+      "epoch": 3.354652453919609,
+      "grad_norm": 0.4540973249501273,
+      "learning_rate": 7.705464739029172e-07,
+      "loss": 0.0262,
+      "step": 7553
+    },
+    {
+      "epoch": 3.3550966022651565,
+      "grad_norm": 0.5423667462933897,
+      "learning_rate": 7.695128610716707e-07,
+      "loss": 0.0373,
+      "step": 7554
+    },
+    {
+      "epoch": 3.355540750610704,
+      "grad_norm": 0.41533591822222976,
+      "learning_rate": 7.684798841545043e-07,
+      "loss": 0.02,
+      "step": 7555
+    },
+    {
+      "epoch": 3.3559848989562515,
+      "grad_norm": 0.37206716629030406,
+      "learning_rate": 7.674475433066925e-07,
+      "loss": 0.027,
+      "step": 7556
+    },
+    {
+      "epoch": 3.356429047301799,
+      "grad_norm": 0.489116896622887,
+      "learning_rate": 7.664158386834131e-07,
+      "loss": 0.0276,
+      "step": 7557
+    },
+    {
+      "epoch": 3.356873195647346,
+      "grad_norm": 0.3415873437320046,
+      "learning_rate": 7.653847704397504e-07,
+      "loss": 0.0199,
+      "step": 7558
+    },
+    {
+      "epoch": 3.357317343992894,
+      "grad_norm": 0.4918207209258021,
+      "learning_rate": 7.643543387306896e-07,
+      "loss": 0.0232,
+      "step": 7559
+    },
+    {
+      "epoch": 3.357761492338441,
+      "grad_norm": 0.4284932846117368,
+      "learning_rate": 7.63324543711122e-07,
+      "loss": 0.0304,
+      "step": 7560
+    },
+    {
+      "epoch": 3.3582056406839884,
+      "grad_norm": 0.3697104694086724,
+      "learning_rate": 7.622953855358456e-07,
+      "loss": 0.0206,
+      "step": 7561
+    },
+    {
+      "epoch": 3.3586497890295357,
+      "grad_norm": 0.3548831840036858,
+      "learning_rate": 7.612668643595561e-07,
+      "loss": 0.0241,
+      "step": 7562
+    },
+    {
+      "epoch": 3.3590939373750834,
+      "grad_norm": 0.3814303904037155,
+      "learning_rate": 7.60238980336862e-07,
+      "loss": 0.0285,
+      "step": 7563
+    },
+    {
+      "epoch": 3.3595380857206307,
+      "grad_norm": 0.40972342736425,
+      "learning_rate": 7.592117336222709e-07,
+      "loss": 0.0231,
+      "step": 7564
+    },
+    {
+      "epoch": 3.359982234066178,
+      "grad_norm": 0.4313904313100148,
+      "learning_rate": 7.581851243701938e-07,
+      "loss": 0.0272,
+      "step": 7565
+    },
+    {
+      "epoch": 3.3604263824117258,
+      "grad_norm": 0.42249063117216307,
+      "learning_rate": 7.571591527349481e-07,
+      "loss": 0.0245,
+      "step": 7566
+    },
+    {
+      "epoch": 3.360870530757273,
+      "grad_norm": 0.6077788918648797,
+      "learning_rate": 7.561338188707562e-07,
+      "loss": 0.0245,
+      "step": 7567
+    },
+    {
+      "epoch": 3.3613146791028203,
+      "grad_norm": 0.3504403283983759,
+      "learning_rate": 7.551091229317398e-07,
+      "loss": 0.0233,
+      "step": 7568
+    },
+    {
+      "epoch": 3.3617588274483676,
+      "grad_norm": 0.5216233787292206,
+      "learning_rate": 7.540850650719317e-07,
+      "loss": 0.0323,
+      "step": 7569
+    },
+    {
+      "epoch": 3.3622029757939154,
+      "grad_norm": 0.350140151698931,
+      "learning_rate": 7.530616454452644e-07,
+      "loss": 0.0177,
+      "step": 7570
+    },
+    {
+      "epoch": 3.3626471241394627,
+      "grad_norm": 0.4407132954191557,
+      "learning_rate": 7.520388642055737e-07,
+      "loss": 0.0315,
+      "step": 7571
+    },
+    {
+      "epoch": 3.36309127248501,
+      "grad_norm": 0.39660867426600405,
+      "learning_rate": 7.510167215066022e-07,
+      "loss": 0.0219,
+      "step": 7572
+    },
+    {
+      "epoch": 3.3635354208305572,
+      "grad_norm": 0.4580507155342704,
+      "learning_rate": 7.499952175019947e-07,
+      "loss": 0.0289,
+      "step": 7573
+    },
+    {
+      "epoch": 3.363979569176105,
+      "grad_norm": 0.3682289326212527,
+      "learning_rate": 7.489743523453013e-07,
+      "loss": 0.0239,
+      "step": 7574
+    },
+    {
+      "epoch": 3.3644237175216523,
+      "grad_norm": 0.488168671743432,
+      "learning_rate": 7.479541261899758e-07,
+      "loss": 0.0257,
+      "step": 7575
+    },
+    {
+      "epoch": 3.3648678658671995,
+      "grad_norm": 0.5131286442085937,
+      "learning_rate": 7.469345391893739e-07,
+      "loss": 0.0251,
+      "step": 7576
+    },
+    {
+      "epoch": 3.365312014212747,
+      "grad_norm": 0.5086234654496984,
+      "learning_rate": 7.459155914967581e-07,
+      "loss": 0.0312,
+      "step": 7577
+    },
+    {
+      "epoch": 3.3657561625582946,
+      "grad_norm": 1.2159009232304825,
+      "learning_rate": 7.448972832652939e-07,
+      "loss": 0.034,
+      "step": 7578
+    },
+    {
+      "epoch": 3.366200310903842,
+      "grad_norm": 0.42264429800124254,
+      "learning_rate": 7.438796146480471e-07,
+      "loss": 0.0366,
+      "step": 7579
+    },
+    {
+      "epoch": 3.366644459249389,
+      "grad_norm": 0.474511735916579,
+      "learning_rate": 7.428625857979943e-07,
+      "loss": 0.0282,
+      "step": 7580
+    },
+    {
+      "epoch": 3.367088607594937,
+      "grad_norm": 0.48689953790784196,
+      "learning_rate": 7.418461968680124e-07,
+      "loss": 0.0303,
+      "step": 7581
+    },
+    {
+      "epoch": 3.367532755940484,
+      "grad_norm": 0.34637453409423774,
+      "learning_rate": 7.408304480108791e-07,
+      "loss": 0.0213,
+      "step": 7582
+    },
+    {
+      "epoch": 3.3679769042860315,
+      "grad_norm": 0.4489470189672078,
+      "learning_rate": 7.398153393792801e-07,
+      "loss": 0.0314,
+      "step": 7583
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.38625164102595216,
+      "learning_rate": 7.388008711258049e-07,
+      "loss": 0.0205,
+      "step": 7584
+    },
+    {
+      "epoch": 3.3688652009771265,
+      "grad_norm": 0.38974030926942727,
+      "learning_rate": 7.37787043402941e-07,
+      "loss": 0.0263,
+      "step": 7585
+    },
+    {
+      "epoch": 3.369309349322674,
+      "grad_norm": 0.43774082705869743,
+      "learning_rate": 7.367738563630894e-07,
+      "loss": 0.0218,
+      "step": 7586
+    },
+    {
+      "epoch": 3.369753497668221,
+      "grad_norm": 0.6108985899563969,
+      "learning_rate": 7.357613101585459e-07,
+      "loss": 0.0311,
+      "step": 7587
+    },
+    {
+      "epoch": 3.370197646013769,
+      "grad_norm": 0.46815029644071443,
+      "learning_rate": 7.347494049415139e-07,
+      "loss": 0.031,
+      "step": 7588
+    },
+    {
+      "epoch": 3.370641794359316,
+      "grad_norm": 0.38759059737412377,
+      "learning_rate": 7.337381408641004e-07,
+      "loss": 0.0201,
+      "step": 7589
+    },
+    {
+      "epoch": 3.3710859427048634,
+      "grad_norm": 0.44352496378452017,
+      "learning_rate": 7.327275180783156e-07,
+      "loss": 0.0321,
+      "step": 7590
+    },
+    {
+      "epoch": 3.3715300910504107,
+      "grad_norm": 0.48865244874942404,
+      "learning_rate": 7.317175367360729e-07,
+      "loss": 0.024,
+      "step": 7591
+    },
+    {
+      "epoch": 3.3719742393959584,
+      "grad_norm": 0.4173965077203295,
+      "learning_rate": 7.30708196989191e-07,
+      "loss": 0.0208,
+      "step": 7592
+    },
+    {
+      "epoch": 3.3724183877415057,
+      "grad_norm": 0.4555470735195906,
+      "learning_rate": 7.296994989893885e-07,
+      "loss": 0.0217,
+      "step": 7593
+    },
+    {
+      "epoch": 3.372862536087053,
+      "grad_norm": 0.3350515129151666,
+      "learning_rate": 7.286914428882913e-07,
+      "loss": 0.021,
+      "step": 7594
+    },
+    {
+      "epoch": 3.3733066844326007,
+      "grad_norm": 0.310648697981219,
+      "learning_rate": 7.276840288374281e-07,
+      "loss": 0.0257,
+      "step": 7595
+    },
+    {
+      "epoch": 3.373750832778148,
+      "grad_norm": 0.5863731801123462,
+      "learning_rate": 7.266772569882269e-07,
+      "loss": 0.0311,
+      "step": 7596
+    },
+    {
+      "epoch": 3.3741949811236953,
+      "grad_norm": 0.31868586904538343,
+      "learning_rate": 7.256711274920264e-07,
+      "loss": 0.0194,
+      "step": 7597
+    },
+    {
+      "epoch": 3.3746391294692426,
+      "grad_norm": 0.3360320574193569,
+      "learning_rate": 7.246656405000646e-07,
+      "loss": 0.0219,
+      "step": 7598
+    },
+    {
+      "epoch": 3.37508327781479,
+      "grad_norm": 0.4187991374923001,
+      "learning_rate": 7.236607961634812e-07,
+      "loss": 0.0266,
+      "step": 7599
+    },
+    {
+      "epoch": 3.3755274261603376,
+      "grad_norm": 0.3828524070852837,
+      "learning_rate": 7.22656594633322e-07,
+      "loss": 0.026,
+      "step": 7600
+    },
+    {
+      "epoch": 3.375971574505885,
+      "grad_norm": 0.4895737837652768,
+      "learning_rate": 7.216530360605379e-07,
+      "loss": 0.0413,
+      "step": 7601
+    },
+    {
+      "epoch": 3.376415722851432,
+      "grad_norm": 0.5857587373369051,
+      "learning_rate": 7.206501205959759e-07,
+      "loss": 0.0219,
+      "step": 7602
+    },
+    {
+      "epoch": 3.37685987119698,
+      "grad_norm": 0.383441583877617,
+      "learning_rate": 7.196478483903968e-07,
+      "loss": 0.0196,
+      "step": 7603
+    },
+    {
+      "epoch": 3.3773040195425272,
+      "grad_norm": 0.33141940449265894,
+      "learning_rate": 7.186462195944555e-07,
+      "loss": 0.0217,
+      "step": 7604
+    },
+    {
+      "epoch": 3.3777481678880745,
+      "grad_norm": 0.37300245379755964,
+      "learning_rate": 7.176452343587148e-07,
+      "loss": 0.0216,
+      "step": 7605
+    },
+    {
+      "epoch": 3.378192316233622,
+      "grad_norm": 0.33936395007321735,
+      "learning_rate": 7.166448928336411e-07,
+      "loss": 0.0222,
+      "step": 7606
+    },
+    {
+      "epoch": 3.3786364645791696,
+      "grad_norm": 0.5246214544898155,
+      "learning_rate": 7.156451951696003e-07,
+      "loss": 0.0235,
+      "step": 7607
+    },
+    {
+      "epoch": 3.379080612924717,
+      "grad_norm": 0.3948135131762046,
+      "learning_rate": 7.146461415168637e-07,
+      "loss": 0.0231,
+      "step": 7608
+    },
+    {
+      "epoch": 3.379524761270264,
+      "grad_norm": 0.37493268321499074,
+      "learning_rate": 7.136477320256102e-07,
+      "loss": 0.0328,
+      "step": 7609
+    },
+    {
+      "epoch": 3.379968909615812,
+      "grad_norm": 0.38860406022862326,
+      "learning_rate": 7.126499668459135e-07,
+      "loss": 0.0215,
+      "step": 7610
+    },
+    {
+      "epoch": 3.380413057961359,
+      "grad_norm": 0.443708157737675,
+      "learning_rate": 7.116528461277561e-07,
+      "loss": 0.0272,
+      "step": 7611
+    },
+    {
+      "epoch": 3.3808572063069064,
+      "grad_norm": 0.4409205615918148,
+      "learning_rate": 7.106563700210234e-07,
+      "loss": 0.0199,
+      "step": 7612
+    },
+    {
+      "epoch": 3.3813013546524537,
+      "grad_norm": 0.37530972086273984,
+      "learning_rate": 7.096605386754995e-07,
+      "loss": 0.0291,
+      "step": 7613
+    },
+    {
+      "epoch": 3.3817455029980015,
+      "grad_norm": 0.4774461806256775,
+      "learning_rate": 7.086653522408788e-07,
+      "loss": 0.0274,
+      "step": 7614
+    },
+    {
+      "epoch": 3.3821896513435488,
+      "grad_norm": 0.4294747376572903,
+      "learning_rate": 7.076708108667512e-07,
+      "loss": 0.024,
+      "step": 7615
+    },
+    {
+      "epoch": 3.382633799689096,
+      "grad_norm": 0.41100245317627226,
+      "learning_rate": 7.066769147026154e-07,
+      "loss": 0.0287,
+      "step": 7616
+    },
+    {
+      "epoch": 3.383077948034644,
+      "grad_norm": 0.3045061371856552,
+      "learning_rate": 7.056836638978698e-07,
+      "loss": 0.0198,
+      "step": 7617
+    },
+    {
+      "epoch": 3.383522096380191,
+      "grad_norm": 0.3136165513411521,
+      "learning_rate": 7.046910586018186e-07,
+      "loss": 0.0145,
+      "step": 7618
+    },
+    {
+      "epoch": 3.3839662447257384,
+      "grad_norm": 0.443848462635569,
+      "learning_rate": 7.036990989636628e-07,
+      "loss": 0.0296,
+      "step": 7619
+    },
+    {
+      "epoch": 3.3844103930712857,
+      "grad_norm": 0.4329635290692527,
+      "learning_rate": 7.027077851325164e-07,
+      "loss": 0.0286,
+      "step": 7620
+    },
+    {
+      "epoch": 3.3848545414168334,
+      "grad_norm": 0.4425986943763542,
+      "learning_rate": 7.017171172573872e-07,
+      "loss": 0.0319,
+      "step": 7621
+    },
+    {
+      "epoch": 3.3852986897623807,
+      "grad_norm": 0.3486098761277207,
+      "learning_rate": 7.007270954871903e-07,
+      "loss": 0.0269,
+      "step": 7622
+    },
+    {
+      "epoch": 3.385742838107928,
+      "grad_norm": 0.35813998328060914,
+      "learning_rate": 6.997377199707439e-07,
+      "loss": 0.0212,
+      "step": 7623
+    },
+    {
+      "epoch": 3.3861869864534753,
+      "grad_norm": 0.38415273327037347,
+      "learning_rate": 6.987489908567663e-07,
+      "loss": 0.0283,
+      "step": 7624
+    },
+    {
+      "epoch": 3.386631134799023,
+      "grad_norm": 0.42667107713144753,
+      "learning_rate": 6.977609082938791e-07,
+      "loss": 0.0364,
+      "step": 7625
+    },
+    {
+      "epoch": 3.3870752831445703,
+      "grad_norm": 0.4255004286761174,
+      "learning_rate": 6.967734724306119e-07,
+      "loss": 0.0374,
+      "step": 7626
+    },
+    {
+      "epoch": 3.3875194314901176,
+      "grad_norm": 0.3765149927746781,
+      "learning_rate": 6.957866834153898e-07,
+      "loss": 0.028,
+      "step": 7627
+    },
+    {
+      "epoch": 3.387963579835665,
+      "grad_norm": 0.5330001693156503,
+      "learning_rate": 6.948005413965448e-07,
+      "loss": 0.0379,
+      "step": 7628
+    },
+    {
+      "epoch": 3.3884077281812126,
+      "grad_norm": 0.5059080740347456,
+      "learning_rate": 6.938150465223126e-07,
+      "loss": 0.0244,
+      "step": 7629
+    },
+    {
+      "epoch": 3.38885187652676,
+      "grad_norm": 0.43454836222929405,
+      "learning_rate": 6.928301989408253e-07,
+      "loss": 0.0323,
+      "step": 7630
+    },
+    {
+      "epoch": 3.389296024872307,
+      "grad_norm": 0.3908471485523718,
+      "learning_rate": 6.918459988001281e-07,
+      "loss": 0.0312,
+      "step": 7631
+    },
+    {
+      "epoch": 3.389740173217855,
+      "grad_norm": 0.4420676727706239,
+      "learning_rate": 6.908624462481584e-07,
+      "loss": 0.026,
+      "step": 7632
+    },
+    {
+      "epoch": 3.390184321563402,
+      "grad_norm": 0.35748857241333987,
+      "learning_rate": 6.898795414327624e-07,
+      "loss": 0.0212,
+      "step": 7633
+    },
+    {
+      "epoch": 3.3906284699089495,
+      "grad_norm": 0.3032775147821817,
+      "learning_rate": 6.888972845016889e-07,
+      "loss": 0.0255,
+      "step": 7634
+    },
+    {
+      "epoch": 3.391072618254497,
+      "grad_norm": 0.3982622931840567,
+      "learning_rate": 6.879156756025851e-07,
+      "loss": 0.0281,
+      "step": 7635
+    },
+    {
+      "epoch": 3.3915167666000445,
+      "grad_norm": 0.5203243077622092,
+      "learning_rate": 6.869347148830035e-07,
+      "loss": 0.0249,
+      "step": 7636
+    },
+    {
+      "epoch": 3.391960914945592,
+      "grad_norm": 0.5073777888896281,
+      "learning_rate": 6.85954402490403e-07,
+      "loss": 0.0408,
+      "step": 7637
+    },
+    {
+      "epoch": 3.392405063291139,
+      "grad_norm": 0.6112073451304307,
+      "learning_rate": 6.849747385721373e-07,
+      "loss": 0.0262,
+      "step": 7638
+    },
+    {
+      "epoch": 3.392849211636687,
+      "grad_norm": 0.4334382118989557,
+      "learning_rate": 6.839957232754679e-07,
+      "loss": 0.0247,
+      "step": 7639
+    },
+    {
+      "epoch": 3.393293359982234,
+      "grad_norm": 0.37151198090225157,
+      "learning_rate": 6.830173567475584e-07,
+      "loss": 0.0251,
+      "step": 7640
+    },
+    {
+      "epoch": 3.3937375083277814,
+      "grad_norm": 0.47432508657043243,
+      "learning_rate": 6.820396391354722e-07,
+      "loss": 0.0361,
+      "step": 7641
+    },
+    {
+      "epoch": 3.3941816566733287,
+      "grad_norm": 0.39488273115178324,
+      "learning_rate": 6.810625705861762e-07,
+      "loss": 0.0226,
+      "step": 7642
+    },
+    {
+      "epoch": 3.3946258050188765,
+      "grad_norm": 0.5607339834034196,
+      "learning_rate": 6.80086151246544e-07,
+      "loss": 0.0403,
+      "step": 7643
+    },
+    {
+      "epoch": 3.3950699533644237,
+      "grad_norm": 0.4287176433981967,
+      "learning_rate": 6.791103812633443e-07,
+      "loss": 0.026,
+      "step": 7644
+    },
+    {
+      "epoch": 3.395514101709971,
+      "grad_norm": 0.4160527081898972,
+      "learning_rate": 6.781352607832536e-07,
+      "loss": 0.0248,
+      "step": 7645
+    },
+    {
+      "epoch": 3.3959582500555188,
+      "grad_norm": 0.31792810620229006,
+      "learning_rate": 6.771607899528504e-07,
+      "loss": 0.0179,
+      "step": 7646
+    },
+    {
+      "epoch": 3.396402398401066,
+      "grad_norm": 0.5611086221445625,
+      "learning_rate": 6.761869689186101e-07,
+      "loss": 0.0266,
+      "step": 7647
+    },
+    {
+      "epoch": 3.3968465467466133,
+      "grad_norm": 0.5050890369850448,
+      "learning_rate": 6.752137978269191e-07,
+      "loss": 0.0362,
+      "step": 7648
+    },
+    {
+      "epoch": 3.3972906950921606,
+      "grad_norm": 0.3650794330064718,
+      "learning_rate": 6.742412768240586e-07,
+      "loss": 0.0259,
+      "step": 7649
+    },
+    {
+      "epoch": 3.3977348434377084,
+      "grad_norm": 0.38912734986088593,
+      "learning_rate": 6.732694060562162e-07,
+      "loss": 0.0222,
+      "step": 7650
+    },
+    {
+      "epoch": 3.3981789917832557,
+      "grad_norm": 0.3696530901709022,
+      "learning_rate": 6.722981856694811e-07,
+      "loss": 0.0242,
+      "step": 7651
+    },
+    {
+      "epoch": 3.398623140128803,
+      "grad_norm": 0.35581748069826435,
+      "learning_rate": 6.713276158098425e-07,
+      "loss": 0.0203,
+      "step": 7652
+    },
+    {
+      "epoch": 3.3990672884743502,
+      "grad_norm": 0.3267819000046619,
+      "learning_rate": 6.703576966231939e-07,
+      "loss": 0.0177,
+      "step": 7653
+    },
+    {
+      "epoch": 3.399511436819898,
+      "grad_norm": 1.0254709261428898,
+      "learning_rate": 6.693884282553332e-07,
+      "loss": 0.0272,
+      "step": 7654
+    },
+    {
+      "epoch": 3.3999555851654453,
+      "grad_norm": 0.5090905796653779,
+      "learning_rate": 6.684198108519546e-07,
+      "loss": 0.026,
+      "step": 7655
+    },
+    {
+      "epoch": 3.4003997335109926,
+      "grad_norm": 0.47438772486404496,
+      "learning_rate": 6.674518445586592e-07,
+      "loss": 0.0293,
+      "step": 7656
+    },
+    {
+      "epoch": 3.40084388185654,
+      "grad_norm": 0.508053594793761,
+      "learning_rate": 6.664845295209499e-07,
+      "loss": 0.0397,
+      "step": 7657
+    },
+    {
+      "epoch": 3.4012880302020876,
+      "grad_norm": 0.361105004393522,
+      "learning_rate": 6.655178658842282e-07,
+      "loss": 0.0197,
+      "step": 7658
+    },
+    {
+      "epoch": 3.401732178547635,
+      "grad_norm": 0.36194311552740505,
+      "learning_rate": 6.645518537938012e-07,
+      "loss": 0.0233,
+      "step": 7659
+    },
+    {
+      "epoch": 3.402176326893182,
+      "grad_norm": 0.4835513612825343,
+      "learning_rate": 6.635864933948771e-07,
+      "loss": 0.0243,
+      "step": 7660
+    },
+    {
+      "epoch": 3.40262047523873,
+      "grad_norm": 0.40734978570499325,
+      "learning_rate": 6.626217848325656e-07,
+      "loss": 0.0235,
+      "step": 7661
+    },
+    {
+      "epoch": 3.403064623584277,
+      "grad_norm": 0.48301835002180094,
+      "learning_rate": 6.616577282518794e-07,
+      "loss": 0.0307,
+      "step": 7662
+    },
+    {
+      "epoch": 3.4035087719298245,
+      "grad_norm": 0.4251997392123568,
+      "learning_rate": 6.606943237977331e-07,
+      "loss": 0.0193,
+      "step": 7663
+    },
+    {
+      "epoch": 3.4039529202753718,
+      "grad_norm": 0.45591767603385597,
+      "learning_rate": 6.597315716149394e-07,
+      "loss": 0.0348,
+      "step": 7664
+    },
+    {
+      "epoch": 3.4043970686209195,
+      "grad_norm": 0.38978320299224184,
+      "learning_rate": 6.587694718482213e-07,
+      "loss": 0.024,
+      "step": 7665
+    },
+    {
+      "epoch": 3.404841216966467,
+      "grad_norm": 0.36098640085183586,
+      "learning_rate": 6.578080246421947e-07,
+      "loss": 0.0231,
+      "step": 7666
+    },
+    {
+      "epoch": 3.405285365312014,
+      "grad_norm": 0.42381127870689095,
+      "learning_rate": 6.568472301413836e-07,
+      "loss": 0.0293,
+      "step": 7667
+    },
+    {
+      "epoch": 3.405729513657562,
+      "grad_norm": 0.36978104460309597,
+      "learning_rate": 6.558870884902119e-07,
+      "loss": 0.0228,
+      "step": 7668
+    },
+    {
+      "epoch": 3.406173662003109,
+      "grad_norm": 0.5467829172657439,
+      "learning_rate": 6.549275998330029e-07,
+      "loss": 0.027,
+      "step": 7669
+    },
+    {
+      "epoch": 3.4066178103486564,
+      "grad_norm": 0.42439207241368543,
+      "learning_rate": 6.539687643139847e-07,
+      "loss": 0.0243,
+      "step": 7670
+    },
+    {
+      "epoch": 3.4070619586942037,
+      "grad_norm": 0.34272251460913966,
+      "learning_rate": 6.530105820772897e-07,
+      "loss": 0.0151,
+      "step": 7671
+    },
+    {
+      "epoch": 3.4075061070397514,
+      "grad_norm": 0.4182566186357329,
+      "learning_rate": 6.52053053266945e-07,
+      "loss": 0.0205,
+      "step": 7672
+    },
+    {
+      "epoch": 3.4079502553852987,
+      "grad_norm": 0.350434375510393,
+      "learning_rate": 6.51096178026886e-07,
+      "loss": 0.0237,
+      "step": 7673
+    },
+    {
+      "epoch": 3.408394403730846,
+      "grad_norm": 0.36842396127077415,
+      "learning_rate": 6.50139956500947e-07,
+      "loss": 0.0248,
+      "step": 7674
+    },
+    {
+      "epoch": 3.4088385520763937,
+      "grad_norm": 0.3776577384417059,
+      "learning_rate": 6.491843888328625e-07,
+      "loss": 0.026,
+      "step": 7675
+    },
+    {
+      "epoch": 3.409282700421941,
+      "grad_norm": 0.4739319012198071,
+      "learning_rate": 6.482294751662721e-07,
+      "loss": 0.0353,
+      "step": 7676
+    },
+    {
+      "epoch": 3.4097268487674883,
+      "grad_norm": 0.35847335034757755,
+      "learning_rate": 6.472752156447148e-07,
+      "loss": 0.0198,
+      "step": 7677
+    },
+    {
+      "epoch": 3.4101709971130356,
+      "grad_norm": 0.41470003345548434,
+      "learning_rate": 6.463216104116327e-07,
+      "loss": 0.0206,
+      "step": 7678
+    },
+    {
+      "epoch": 3.4106151454585834,
+      "grad_norm": 0.41774456358583356,
+      "learning_rate": 6.453686596103697e-07,
+      "loss": 0.0296,
+      "step": 7679
+    },
+    {
+      "epoch": 3.4110592938041306,
+      "grad_norm": 0.4014386567309541,
+      "learning_rate": 6.444163633841688e-07,
+      "loss": 0.0283,
+      "step": 7680
+    },
+    {
+      "epoch": 3.411503442149678,
+      "grad_norm": 0.5677232388670659,
+      "learning_rate": 6.434647218761764e-07,
+      "loss": 0.0399,
+      "step": 7681
+    },
+    {
+      "epoch": 3.4119475904952252,
+      "grad_norm": 0.38057897363042437,
+      "learning_rate": 6.425137352294408e-07,
+      "loss": 0.0241,
+      "step": 7682
+    },
+    {
+      "epoch": 3.412391738840773,
+      "grad_norm": 0.4520213263061807,
+      "learning_rate": 6.415634035869117e-07,
+      "loss": 0.0242,
+      "step": 7683
+    },
+    {
+      "epoch": 3.4128358871863202,
+      "grad_norm": 0.4445568386773207,
+      "learning_rate": 6.406137270914404e-07,
+      "loss": 0.0327,
+      "step": 7684
+    },
+    {
+      "epoch": 3.4132800355318675,
+      "grad_norm": 0.48242749215653374,
+      "learning_rate": 6.396647058857792e-07,
+      "loss": 0.031,
+      "step": 7685
+    },
+    {
+      "epoch": 3.413724183877415,
+      "grad_norm": 0.3708344900625483,
+      "learning_rate": 6.387163401125812e-07,
+      "loss": 0.0233,
+      "step": 7686
+    },
+    {
+      "epoch": 3.4141683322229626,
+      "grad_norm": 0.3782593306434555,
+      "learning_rate": 6.377686299144025e-07,
+      "loss": 0.0269,
+      "step": 7687
+    },
+    {
+      "epoch": 3.41461248056851,
+      "grad_norm": 0.43908345913471114,
+      "learning_rate": 6.368215754337004e-07,
+      "loss": 0.0307,
+      "step": 7688
+    },
+    {
+      "epoch": 3.415056628914057,
+      "grad_norm": 0.5407716737755903,
+      "learning_rate": 6.358751768128324e-07,
+      "loss": 0.0259,
+      "step": 7689
+    },
+    {
+      "epoch": 3.415500777259605,
+      "grad_norm": 0.3714076929525467,
+      "learning_rate": 6.349294341940593e-07,
+      "loss": 0.0182,
+      "step": 7690
+    },
+    {
+      "epoch": 3.415944925605152,
+      "grad_norm": 0.5168319772596983,
+      "learning_rate": 6.339843477195423e-07,
+      "loss": 0.0405,
+      "step": 7691
+    },
+    {
+      "epoch": 3.4163890739506995,
+      "grad_norm": 0.470888647389716,
+      "learning_rate": 6.330399175313429e-07,
+      "loss": 0.0326,
+      "step": 7692
+    },
+    {
+      "epoch": 3.4168332222962468,
+      "grad_norm": 0.3446195227420253,
+      "learning_rate": 6.320961437714257e-07,
+      "loss": 0.0136,
+      "step": 7693
+    },
+    {
+      "epoch": 3.4172773706417945,
+      "grad_norm": 0.37511059576115724,
+      "learning_rate": 6.311530265816551e-07,
+      "loss": 0.0336,
+      "step": 7694
+    },
+    {
+      "epoch": 3.4177215189873418,
+      "grad_norm": 0.362187597907496,
+      "learning_rate": 6.302105661037988e-07,
+      "loss": 0.0212,
+      "step": 7695
+    },
+    {
+      "epoch": 3.418165667332889,
+      "grad_norm": 0.4435954280766343,
+      "learning_rate": 6.292687624795257e-07,
+      "loss": 0.0315,
+      "step": 7696
+    },
+    {
+      "epoch": 3.418609815678437,
+      "grad_norm": 0.36818994844478536,
+      "learning_rate": 6.283276158504015e-07,
+      "loss": 0.0226,
+      "step": 7697
+    },
+    {
+      "epoch": 3.419053964023984,
+      "grad_norm": 0.35373090512917094,
+      "learning_rate": 6.27387126357899e-07,
+      "loss": 0.0165,
+      "step": 7698
+    },
+    {
+      "epoch": 3.4194981123695314,
+      "grad_norm": 0.364744423870996,
+      "learning_rate": 6.264472941433886e-07,
+      "loss": 0.0159,
+      "step": 7699
+    },
+    {
+      "epoch": 3.4199422607150787,
+      "grad_norm": 0.43254863668472154,
+      "learning_rate": 6.255081193481438e-07,
+      "loss": 0.0269,
+      "step": 7700
+    },
+    {
+      "epoch": 3.4203864090606264,
+      "grad_norm": 0.3606692968436507,
+      "learning_rate": 6.24569602113338e-07,
+      "loss": 0.0304,
+      "step": 7701
+    },
+    {
+      "epoch": 3.4208305574061737,
+      "grad_norm": 0.5007533786139542,
+      "learning_rate": 6.236317425800481e-07,
+      "loss": 0.0287,
+      "step": 7702
+    },
+    {
+      "epoch": 3.421274705751721,
+      "grad_norm": 0.41340230947723855,
+      "learning_rate": 6.226945408892477e-07,
+      "loss": 0.0199,
+      "step": 7703
+    },
+    {
+      "epoch": 3.4217188540972687,
+      "grad_norm": 0.365085912872857,
+      "learning_rate": 6.21757997181815e-07,
+      "loss": 0.021,
+      "step": 7704
+    },
+    {
+      "epoch": 3.422163002442816,
+      "grad_norm": 0.5548854003945028,
+      "learning_rate": 6.208221115985285e-07,
+      "loss": 0.0336,
+      "step": 7705
+    },
+    {
+      "epoch": 3.4226071507883633,
+      "grad_norm": 0.5898794387281542,
+      "learning_rate": 6.198868842800681e-07,
+      "loss": 0.0329,
+      "step": 7706
+    },
+    {
+      "epoch": 3.4230512991339106,
+      "grad_norm": 0.670696139265725,
+      "learning_rate": 6.189523153670152e-07,
+      "loss": 0.0411,
+      "step": 7707
+    },
+    {
+      "epoch": 3.4234954474794583,
+      "grad_norm": 0.44881661557818314,
+      "learning_rate": 6.180184049998489e-07,
+      "loss": 0.0294,
+      "step": 7708
+    },
+    {
+      "epoch": 3.4239395958250056,
+      "grad_norm": 0.41059208128468105,
+      "learning_rate": 6.170851533189537e-07,
+      "loss": 0.0221,
+      "step": 7709
+    },
+    {
+      "epoch": 3.424383744170553,
+      "grad_norm": 0.4264078976239932,
+      "learning_rate": 6.161525604646124e-07,
+      "loss": 0.0247,
+      "step": 7710
+    },
+    {
+      "epoch": 3.4248278925161,
+      "grad_norm": 0.4696315553933537,
+      "learning_rate": 6.152206265770095e-07,
+      "loss": 0.026,
+      "step": 7711
+    },
+    {
+      "epoch": 3.425272040861648,
+      "grad_norm": 0.45378823926056083,
+      "learning_rate": 6.142893517962312e-07,
+      "loss": 0.0235,
+      "step": 7712
+    },
+    {
+      "epoch": 3.4257161892071952,
+      "grad_norm": 0.4121833782140256,
+      "learning_rate": 6.133587362622645e-07,
+      "loss": 0.0272,
+      "step": 7713
+    },
+    {
+      "epoch": 3.4261603375527425,
+      "grad_norm": 0.3934369295423382,
+      "learning_rate": 6.124287801149942e-07,
+      "loss": 0.0254,
+      "step": 7714
+    },
+    {
+      "epoch": 3.42660448589829,
+      "grad_norm": 0.5182408476406597,
+      "learning_rate": 6.114994834942106e-07,
+      "loss": 0.0311,
+      "step": 7715
+    },
+    {
+      "epoch": 3.4270486342438375,
+      "grad_norm": 0.443783825192744,
+      "learning_rate": 6.105708465396021e-07,
+      "loss": 0.0245,
+      "step": 7716
+    },
+    {
+      "epoch": 3.427492782589385,
+      "grad_norm": 0.47129625335261466,
+      "learning_rate": 6.096428693907591e-07,
+      "loss": 0.0254,
+      "step": 7717
+    },
+    {
+      "epoch": 3.427936930934932,
+      "grad_norm": 0.38995292895347056,
+      "learning_rate": 6.087155521871713e-07,
+      "loss": 0.028,
+      "step": 7718
+    },
+    {
+      "epoch": 3.42838107928048,
+      "grad_norm": 0.38715826339897036,
+      "learning_rate": 6.077888950682326e-07,
+      "loss": 0.0228,
+      "step": 7719
+    },
+    {
+      "epoch": 3.428825227626027,
+      "grad_norm": 0.4108315391155564,
+      "learning_rate": 6.068628981732322e-07,
+      "loss": 0.0235,
+      "step": 7720
+    },
+    {
+      "epoch": 3.4292693759715744,
+      "grad_norm": 0.3702934691579192,
+      "learning_rate": 6.059375616413643e-07,
+      "loss": 0.0211,
+      "step": 7721
+    },
+    {
+      "epoch": 3.4297135243171217,
+      "grad_norm": 0.3963620092994827,
+      "learning_rate": 6.050128856117232e-07,
+      "loss": 0.019,
+      "step": 7722
+    },
+    {
+      "epoch": 3.4301576726626695,
+      "grad_norm": 0.37150524249459593,
+      "learning_rate": 6.040888702233033e-07,
+      "loss": 0.0224,
+      "step": 7723
+    },
+    {
+      "epoch": 3.4306018210082168,
+      "grad_norm": 0.44821126190431915,
+      "learning_rate": 6.031655156150007e-07,
+      "loss": 0.0305,
+      "step": 7724
+    },
+    {
+      "epoch": 3.431045969353764,
+      "grad_norm": 0.35283254586034607,
+      "learning_rate": 6.022428219256087e-07,
+      "loss": 0.024,
+      "step": 7725
+    },
+    {
+      "epoch": 3.431490117699312,
+      "grad_norm": 0.4935930259013348,
+      "learning_rate": 6.013207892938261e-07,
+      "loss": 0.0457,
+      "step": 7726
+    },
+    {
+      "epoch": 3.431934266044859,
+      "grad_norm": 0.45369029101077796,
+      "learning_rate": 6.003994178582489e-07,
+      "loss": 0.0243,
+      "step": 7727
+    },
+    {
+      "epoch": 3.4323784143904064,
+      "grad_norm": 0.48244218877941225,
+      "learning_rate": 5.994787077573754e-07,
+      "loss": 0.0231,
+      "step": 7728
+    },
+    {
+      "epoch": 3.4328225627359537,
+      "grad_norm": 0.3599007846172956,
+      "learning_rate": 5.985586591296044e-07,
+      "loss": 0.0182,
+      "step": 7729
+    },
+    {
+      "epoch": 3.4332667110815014,
+      "grad_norm": 0.40382161309843,
+      "learning_rate": 5.976392721132351e-07,
+      "loss": 0.0152,
+      "step": 7730
+    },
+    {
+      "epoch": 3.4337108594270487,
+      "grad_norm": 0.44806877408277146,
+      "learning_rate": 5.967205468464648e-07,
+      "loss": 0.026,
+      "step": 7731
+    },
+    {
+      "epoch": 3.434155007772596,
+      "grad_norm": 0.3566806457323869,
+      "learning_rate": 5.958024834673953e-07,
+      "loss": 0.0192,
+      "step": 7732
+    },
+    {
+      "epoch": 3.4345991561181437,
+      "grad_norm": 0.4251189473202698,
+      "learning_rate": 5.948850821140267e-07,
+      "loss": 0.0218,
+      "step": 7733
+    },
+    {
+      "epoch": 3.435043304463691,
+      "grad_norm": 0.38225003080929293,
+      "learning_rate": 5.939683429242604e-07,
+      "loss": 0.0217,
+      "step": 7734
+    },
+    {
+      "epoch": 3.4354874528092383,
+      "grad_norm": 0.34970498765009345,
+      "learning_rate": 5.930522660358973e-07,
+      "loss": 0.0167,
+      "step": 7735
+    },
+    {
+      "epoch": 3.4359316011547856,
+      "grad_norm": 0.41748998353130423,
+      "learning_rate": 5.921368515866405e-07,
+      "loss": 0.0314,
+      "step": 7736
+    },
+    {
+      "epoch": 3.436375749500333,
+      "grad_norm": 0.5166214998053038,
+      "learning_rate": 5.912220997140905e-07,
+      "loss": 0.0327,
+      "step": 7737
+    },
+    {
+      "epoch": 3.4368198978458806,
+      "grad_norm": 0.4503364511144795,
+      "learning_rate": 5.903080105557507e-07,
+      "loss": 0.0305,
+      "step": 7738
+    },
+    {
+      "epoch": 3.437264046191428,
+      "grad_norm": 0.38794068759388,
+      "learning_rate": 5.893945842490245e-07,
+      "loss": 0.0234,
+      "step": 7739
+    },
+    {
+      "epoch": 3.437708194536975,
+      "grad_norm": 0.4805056088393158,
+      "learning_rate": 5.884818209312159e-07,
+      "loss": 0.0298,
+      "step": 7740
+    },
+    {
+      "epoch": 3.438152342882523,
+      "grad_norm": 0.5076559758304271,
+      "learning_rate": 5.875697207395286e-07,
+      "loss": 0.035,
+      "step": 7741
+    },
+    {
+      "epoch": 3.43859649122807,
+      "grad_norm": 0.42338511678046004,
+      "learning_rate": 5.866582838110657e-07,
+      "loss": 0.0363,
+      "step": 7742
+    },
+    {
+      "epoch": 3.4390406395736175,
+      "grad_norm": 0.36808395417139134,
+      "learning_rate": 5.857475102828325e-07,
+      "loss": 0.0188,
+      "step": 7743
+    },
+    {
+      "epoch": 3.439484787919165,
+      "grad_norm": 0.3177380221983695,
+      "learning_rate": 5.848374002917329e-07,
+      "loss": 0.0193,
+      "step": 7744
+    },
+    {
+      "epoch": 3.4399289362647125,
+      "grad_norm": 0.3914364144553906,
+      "learning_rate": 5.839279539745729e-07,
+      "loss": 0.0223,
+      "step": 7745
+    },
+    {
+      "epoch": 3.44037308461026,
+      "grad_norm": 0.402651938874195,
+      "learning_rate": 5.830191714680578e-07,
+      "loss": 0.0199,
+      "step": 7746
+    },
+    {
+      "epoch": 3.440817232955807,
+      "grad_norm": 0.3522412455115687,
+      "learning_rate": 5.821110529087932e-07,
+      "loss": 0.0247,
+      "step": 7747
+    },
+    {
+      "epoch": 3.441261381301355,
+      "grad_norm": 0.4021635163074727,
+      "learning_rate": 5.812035984332832e-07,
+      "loss": 0.0229,
+      "step": 7748
+    },
+    {
+      "epoch": 3.441705529646902,
+      "grad_norm": 0.4251094753016676,
+      "learning_rate": 5.802968081779342e-07,
+      "loss": 0.024,
+      "step": 7749
+    },
+    {
+      "epoch": 3.4421496779924494,
+      "grad_norm": 0.3662362137440131,
+      "learning_rate": 5.79390682279053e-07,
+      "loss": 0.0247,
+      "step": 7750
+    },
+    {
+      "epoch": 3.4425938263379967,
+      "grad_norm": 0.4585215842279442,
+      "learning_rate": 5.784852208728453e-07,
+      "loss": 0.0295,
+      "step": 7751
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.5141537675737569,
+      "learning_rate": 5.775804240954181e-07,
+      "loss": 0.0318,
+      "step": 7752
+    },
+    {
+      "epoch": 3.4434821230290917,
+      "grad_norm": 0.5180769012624266,
+      "learning_rate": 5.766762920827762e-07,
+      "loss": 0.0255,
+      "step": 7753
+    },
+    {
+      "epoch": 3.443926271374639,
+      "grad_norm": 0.4003620998147694,
+      "learning_rate": 5.757728249708261e-07,
+      "loss": 0.0271,
+      "step": 7754
+    },
+    {
+      "epoch": 3.4443704197201868,
+      "grad_norm": 0.3418440737923465,
+      "learning_rate": 5.748700228953758e-07,
+      "loss": 0.0184,
+      "step": 7755
+    },
+    {
+      "epoch": 3.444814568065734,
+      "grad_norm": 0.687451428270991,
+      "learning_rate": 5.739678859921299e-07,
+      "loss": 0.0362,
+      "step": 7756
+    },
+    {
+      "epoch": 3.4452587164112813,
+      "grad_norm": 0.5024182127797775,
+      "learning_rate": 5.730664143966969e-07,
+      "loss": 0.0278,
+      "step": 7757
+    },
+    {
+      "epoch": 3.4457028647568286,
+      "grad_norm": 0.38303277714696055,
+      "learning_rate": 5.721656082445825e-07,
+      "loss": 0.0193,
+      "step": 7758
+    },
+    {
+      "epoch": 3.4461470131023764,
+      "grad_norm": 0.47510936774303797,
+      "learning_rate": 5.712654676711921e-07,
+      "loss": 0.0246,
+      "step": 7759
+    },
+    {
+      "epoch": 3.4465911614479237,
+      "grad_norm": 0.3741311721785513,
+      "learning_rate": 5.703659928118333e-07,
+      "loss": 0.0181,
+      "step": 7760
+    },
+    {
+      "epoch": 3.447035309793471,
+      "grad_norm": 0.4576763768736328,
+      "learning_rate": 5.694671838017119e-07,
+      "loss": 0.0274,
+      "step": 7761
+    },
+    {
+      "epoch": 3.4474794581390182,
+      "grad_norm": 0.36322150440607,
+      "learning_rate": 5.685690407759342e-07,
+      "loss": 0.0198,
+      "step": 7762
+    },
+    {
+      "epoch": 3.447923606484566,
+      "grad_norm": 0.5039426250866698,
+      "learning_rate": 5.676715638695063e-07,
+      "loss": 0.0299,
+      "step": 7763
+    },
+    {
+      "epoch": 3.4483677548301133,
+      "grad_norm": 0.499760373842697,
+      "learning_rate": 5.667747532173362e-07,
+      "loss": 0.0244,
+      "step": 7764
+    },
+    {
+      "epoch": 3.4488119031756606,
+      "grad_norm": 0.42349126436586376,
+      "learning_rate": 5.658786089542262e-07,
+      "loss": 0.0265,
+      "step": 7765
+    },
+    {
+      "epoch": 3.449256051521208,
+      "grad_norm": 0.5133192715069692,
+      "learning_rate": 5.649831312148845e-07,
+      "loss": 0.0381,
+      "step": 7766
+    },
+    {
+      "epoch": 3.4497001998667556,
+      "grad_norm": 0.4265674001652618,
+      "learning_rate": 5.640883201339154e-07,
+      "loss": 0.0254,
+      "step": 7767
+    },
+    {
+      "epoch": 3.450144348212303,
+      "grad_norm": 0.5615805590965591,
+      "learning_rate": 5.631941758458254e-07,
+      "loss": 0.028,
+      "step": 7768
+    },
+    {
+      "epoch": 3.45058849655785,
+      "grad_norm": 0.6650175323233352,
+      "learning_rate": 5.623006984850193e-07,
+      "loss": 0.0313,
+      "step": 7769
+    },
+    {
+      "epoch": 3.451032644903398,
+      "grad_norm": 0.4291032020769662,
+      "learning_rate": 5.61407888185801e-07,
+      "loss": 0.031,
+      "step": 7770
+    },
+    {
+      "epoch": 3.451476793248945,
+      "grad_norm": 0.438818710394981,
+      "learning_rate": 5.60515745082375e-07,
+      "loss": 0.0271,
+      "step": 7771
+    },
+    {
+      "epoch": 3.4519209415944925,
+      "grad_norm": 0.4109123087663869,
+      "learning_rate": 5.596242693088478e-07,
+      "loss": 0.022,
+      "step": 7772
+    },
+    {
+      "epoch": 3.4523650899400398,
+      "grad_norm": 0.5366106862877935,
+      "learning_rate": 5.587334609992195e-07,
+      "loss": 0.04,
+      "step": 7773
+    },
+    {
+      "epoch": 3.4528092382855875,
+      "grad_norm": 0.37769149540290886,
+      "learning_rate": 5.578433202873967e-07,
+      "loss": 0.0292,
+      "step": 7774
+    },
+    {
+      "epoch": 3.453253386631135,
+      "grad_norm": 0.4649378802579783,
+      "learning_rate": 5.569538473071834e-07,
+      "loss": 0.0227,
+      "step": 7775
+    },
+    {
+      "epoch": 3.453697534976682,
+      "grad_norm": 0.48743583603374796,
+      "learning_rate": 5.560650421922798e-07,
+      "loss": 0.0315,
+      "step": 7776
+    },
+    {
+      "epoch": 3.45414168332223,
+      "grad_norm": 0.5101610442676535,
+      "learning_rate": 5.551769050762895e-07,
+      "loss": 0.03,
+      "step": 7777
+    },
+    {
+      "epoch": 3.454585831667777,
+      "grad_norm": 0.3279417409470981,
+      "learning_rate": 5.542894360927148e-07,
+      "loss": 0.0171,
+      "step": 7778
+    },
+    {
+      "epoch": 3.4550299800133244,
+      "grad_norm": 0.36215562405473767,
+      "learning_rate": 5.534026353749572e-07,
+      "loss": 0.0249,
+      "step": 7779
+    },
+    {
+      "epoch": 3.4554741283588717,
+      "grad_norm": 0.33716482541449544,
+      "learning_rate": 5.52516503056319e-07,
+      "loss": 0.0205,
+      "step": 7780
+    },
+    {
+      "epoch": 3.4559182767044194,
+      "grad_norm": 0.3903797663782476,
+      "learning_rate": 5.516310392699991e-07,
+      "loss": 0.0192,
+      "step": 7781
+    },
+    {
+      "epoch": 3.4563624250499667,
+      "grad_norm": 0.4369763600207604,
+      "learning_rate": 5.507462441490985e-07,
+      "loss": 0.0254,
+      "step": 7782
+    },
+    {
+      "epoch": 3.456806573395514,
+      "grad_norm": 0.4196928785980343,
+      "learning_rate": 5.498621178266167e-07,
+      "loss": 0.0242,
+      "step": 7783
+    },
+    {
+      "epoch": 3.4572507217410617,
+      "grad_norm": 0.36290486329076327,
+      "learning_rate": 5.489786604354535e-07,
+      "loss": 0.0284,
+      "step": 7784
+    },
+    {
+      "epoch": 3.457694870086609,
+      "grad_norm": 0.4926206504376682,
+      "learning_rate": 5.480958721084074e-07,
+      "loss": 0.0281,
+      "step": 7785
+    },
+    {
+      "epoch": 3.4581390184321563,
+      "grad_norm": 0.4384128683223243,
+      "learning_rate": 5.472137529781768e-07,
+      "loss": 0.0281,
+      "step": 7786
+    },
+    {
+      "epoch": 3.4585831667777036,
+      "grad_norm": 0.42664316505284405,
+      "learning_rate": 5.463323031773581e-07,
+      "loss": 0.0286,
+      "step": 7787
+    },
+    {
+      "epoch": 3.4590273151232513,
+      "grad_norm": 0.47490376622622354,
+      "learning_rate": 5.454515228384493e-07,
+      "loss": 0.0208,
+      "step": 7788
+    },
+    {
+      "epoch": 3.4594714634687986,
+      "grad_norm": 0.40714987915665685,
+      "learning_rate": 5.445714120938467e-07,
+      "loss": 0.0321,
+      "step": 7789
+    },
+    {
+      "epoch": 3.459915611814346,
+      "grad_norm": 0.3507834802550716,
+      "learning_rate": 5.436919710758432e-07,
+      "loss": 0.0225,
+      "step": 7790
+    },
+    {
+      "epoch": 3.460359760159893,
+      "grad_norm": 0.3642385320392474,
+      "learning_rate": 5.42813199916637e-07,
+      "loss": 0.025,
+      "step": 7791
+    },
+    {
+      "epoch": 3.460803908505441,
+      "grad_norm": 0.3676585444584738,
+      "learning_rate": 5.419350987483224e-07,
+      "loss": 0.026,
+      "step": 7792
+    },
+    {
+      "epoch": 3.4612480568509882,
+      "grad_norm": 0.42929474369532705,
+      "learning_rate": 5.410576677028906e-07,
+      "loss": 0.0316,
+      "step": 7793
+    },
+    {
+      "epoch": 3.4616922051965355,
+      "grad_norm": 0.45518167323880976,
+      "learning_rate": 5.401809069122354e-07,
+      "loss": 0.0284,
+      "step": 7794
+    },
+    {
+      "epoch": 3.462136353542083,
+      "grad_norm": 0.4983281791130265,
+      "learning_rate": 5.393048165081493e-07,
+      "loss": 0.0411,
+      "step": 7795
+    },
+    {
+      "epoch": 3.4625805018876306,
+      "grad_norm": 0.419339473620287,
+      "learning_rate": 5.384293966223231e-07,
+      "loss": 0.0215,
+      "step": 7796
+    },
+    {
+      "epoch": 3.463024650233178,
+      "grad_norm": 0.3599252115295799,
+      "learning_rate": 5.37554647386348e-07,
+      "loss": 0.0254,
+      "step": 7797
+    },
+    {
+      "epoch": 3.463468798578725,
+      "grad_norm": 0.35977012849103657,
+      "learning_rate": 5.366805689317129e-07,
+      "loss": 0.0293,
+      "step": 7798
+    },
+    {
+      "epoch": 3.463912946924273,
+      "grad_norm": 0.5206144817249008,
+      "learning_rate": 5.358071613898064e-07,
+      "loss": 0.0272,
+      "step": 7799
+    },
+    {
+      "epoch": 3.46435709526982,
+      "grad_norm": 0.2890681249275682,
+      "learning_rate": 5.349344248919175e-07,
+      "loss": 0.0199,
+      "step": 7800
+    },
+    {
+      "epoch": 3.4648012436153675,
+      "grad_norm": 0.4535682999458182,
+      "learning_rate": 5.340623595692313e-07,
+      "loss": 0.0312,
+      "step": 7801
+    },
+    {
+      "epoch": 3.4652453919609147,
+      "grad_norm": 0.4712735624226538,
+      "learning_rate": 5.331909655528361e-07,
+      "loss": 0.0294,
+      "step": 7802
+    },
+    {
+      "epoch": 3.4656895403064625,
+      "grad_norm": 0.39969281186884054,
+      "learning_rate": 5.323202429737179e-07,
+      "loss": 0.0214,
+      "step": 7803
+    },
+    {
+      "epoch": 3.4661336886520098,
+      "grad_norm": 0.40892930316739623,
+      "learning_rate": 5.31450191962759e-07,
+      "loss": 0.0273,
+      "step": 7804
+    },
+    {
+      "epoch": 3.466577836997557,
+      "grad_norm": 0.40420963258096243,
+      "learning_rate": 5.305808126507433e-07,
+      "loss": 0.0257,
+      "step": 7805
+    },
+    {
+      "epoch": 3.467021985343105,
+      "grad_norm": 0.4298719054004784,
+      "learning_rate": 5.297121051683546e-07,
+      "loss": 0.0202,
+      "step": 7806
+    },
+    {
+      "epoch": 3.467466133688652,
+      "grad_norm": 0.4156335033725289,
+      "learning_rate": 5.288440696461716e-07,
+      "loss": 0.0206,
+      "step": 7807
+    },
+    {
+      "epoch": 3.4679102820341994,
+      "grad_norm": 0.6442580410348813,
+      "learning_rate": 5.279767062146784e-07,
+      "loss": 0.0335,
+      "step": 7808
+    },
+    {
+      "epoch": 3.4683544303797467,
+      "grad_norm": 0.42219725211126974,
+      "learning_rate": 5.271100150042518e-07,
+      "loss": 0.0263,
+      "step": 7809
+    },
+    {
+      "epoch": 3.4687985787252944,
+      "grad_norm": 0.39548539897151663,
+      "learning_rate": 5.262439961451709e-07,
+      "loss": 0.0249,
+      "step": 7810
+    },
+    {
+      "epoch": 3.4692427270708417,
+      "grad_norm": 0.3080627076539832,
+      "learning_rate": 5.253786497676134e-07,
+      "loss": 0.0189,
+      "step": 7811
+    },
+    {
+      "epoch": 3.469686875416389,
+      "grad_norm": 0.435674321335036,
+      "learning_rate": 5.245139760016549e-07,
+      "loss": 0.026,
+      "step": 7812
+    },
+    {
+      "epoch": 3.4701310237619367,
+      "grad_norm": 0.3308167388953746,
+      "learning_rate": 5.236499749772716e-07,
+      "loss": 0.0192,
+      "step": 7813
+    },
+    {
+      "epoch": 3.470575172107484,
+      "grad_norm": 0.411730156267874,
+      "learning_rate": 5.227866468243376e-07,
+      "loss": 0.0278,
+      "step": 7814
+    },
+    {
+      "epoch": 3.4710193204530313,
+      "grad_norm": 0.4770932092793088,
+      "learning_rate": 5.219239916726243e-07,
+      "loss": 0.0356,
+      "step": 7815
+    },
+    {
+      "epoch": 3.4714634687985786,
+      "grad_norm": 0.5082942621753359,
+      "learning_rate": 5.210620096518044e-07,
+      "loss": 0.0335,
+      "step": 7816
+    },
+    {
+      "epoch": 3.4719076171441263,
+      "grad_norm": 0.5849666623331924,
+      "learning_rate": 5.202007008914489e-07,
+      "loss": 0.0298,
+      "step": 7817
+    },
+    {
+      "epoch": 3.4723517654896736,
+      "grad_norm": 0.5043080758971297,
+      "learning_rate": 5.193400655210251e-07,
+      "loss": 0.0257,
+      "step": 7818
+    },
+    {
+      "epoch": 3.472795913835221,
+      "grad_norm": 0.35362037908035154,
+      "learning_rate": 5.184801036699033e-07,
+      "loss": 0.0195,
+      "step": 7819
+    },
+    {
+      "epoch": 3.473240062180768,
+      "grad_norm": 0.3793466003982311,
+      "learning_rate": 5.176208154673502e-07,
+      "loss": 0.0198,
+      "step": 7820
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.3602963289642836,
+      "learning_rate": 5.167622010425305e-07,
+      "loss": 0.0241,
+      "step": 7821
+    },
+    {
+      "epoch": 3.474128358871863,
+      "grad_norm": 0.4045415464931093,
+      "learning_rate": 5.159042605245085e-07,
+      "loss": 0.0308,
+      "step": 7822
+    },
+    {
+      "epoch": 3.4745725072174105,
+      "grad_norm": 0.3631349056889571,
+      "learning_rate": 5.150469940422487e-07,
+      "loss": 0.0199,
+      "step": 7823
+    },
+    {
+      "epoch": 3.475016655562958,
+      "grad_norm": 0.47233718849299045,
+      "learning_rate": 5.141904017246097e-07,
+      "loss": 0.0192,
+      "step": 7824
+    },
+    {
+      "epoch": 3.4754608039085055,
+      "grad_norm": 0.4948999022011918,
+      "learning_rate": 5.133344837003557e-07,
+      "loss": 0.0354,
+      "step": 7825
+    },
+    {
+      "epoch": 3.475904952254053,
+      "grad_norm": 0.37036897500750726,
+      "learning_rate": 5.124792400981432e-07,
+      "loss": 0.0236,
+      "step": 7826
+    },
+    {
+      "epoch": 3.4763491005996,
+      "grad_norm": 0.33159665048430614,
+      "learning_rate": 5.116246710465306e-07,
+      "loss": 0.0243,
+      "step": 7827
+    },
+    {
+      "epoch": 3.476793248945148,
+      "grad_norm": 0.4934189727153532,
+      "learning_rate": 5.10770776673975e-07,
+      "loss": 0.0221,
+      "step": 7828
+    },
+    {
+      "epoch": 3.477237397290695,
+      "grad_norm": 0.40096875281363703,
+      "learning_rate": 5.099175571088283e-07,
+      "loss": 0.0224,
+      "step": 7829
+    },
+    {
+      "epoch": 3.4776815456362424,
+      "grad_norm": 0.5694968604223205,
+      "learning_rate": 5.090650124793472e-07,
+      "loss": 0.0344,
+      "step": 7830
+    },
+    {
+      "epoch": 3.4781256939817897,
+      "grad_norm": 0.347016380477264,
+      "learning_rate": 5.082131429136833e-07,
+      "loss": 0.0246,
+      "step": 7831
+    },
+    {
+      "epoch": 3.4785698423273375,
+      "grad_norm": 0.3939110962708003,
+      "learning_rate": 5.073619485398845e-07,
+      "loss": 0.0216,
+      "step": 7832
+    },
+    {
+      "epoch": 3.4790139906728847,
+      "grad_norm": 0.5312038258908143,
+      "learning_rate": 5.065114294859019e-07,
+      "loss": 0.026,
+      "step": 7833
+    },
+    {
+      "epoch": 3.479458139018432,
+      "grad_norm": 0.3820612120014884,
+      "learning_rate": 5.056615858795838e-07,
+      "loss": 0.0231,
+      "step": 7834
+    },
+    {
+      "epoch": 3.4799022873639798,
+      "grad_norm": 0.5500948375085558,
+      "learning_rate": 5.048124178486724e-07,
+      "loss": 0.0371,
+      "step": 7835
+    },
+    {
+      "epoch": 3.480346435709527,
+      "grad_norm": 0.4385367139000315,
+      "learning_rate": 5.039639255208156e-07,
+      "loss": 0.032,
+      "step": 7836
+    },
+    {
+      "epoch": 3.4807905840550744,
+      "grad_norm": 0.3825085947680331,
+      "learning_rate": 5.031161090235559e-07,
+      "loss": 0.0214,
+      "step": 7837
+    },
+    {
+      "epoch": 3.4812347324006216,
+      "grad_norm": 0.42377290578412,
+      "learning_rate": 5.022689684843329e-07,
+      "loss": 0.0253,
+      "step": 7838
+    },
+    {
+      "epoch": 3.4816788807461694,
+      "grad_norm": 0.3200977768252115,
+      "learning_rate": 5.014225040304871e-07,
+      "loss": 0.0172,
+      "step": 7839
+    },
+    {
+      "epoch": 3.4821230290917167,
+      "grad_norm": 0.4075927728120571,
+      "learning_rate": 5.005767157892572e-07,
+      "loss": 0.0234,
+      "step": 7840
+    },
+    {
+      "epoch": 3.482567177437264,
+      "grad_norm": 0.494712633880696,
+      "learning_rate": 4.99731603887777e-07,
+      "loss": 0.0308,
+      "step": 7841
+    },
+    {
+      "epoch": 3.4830113257828117,
+      "grad_norm": 0.3951757300545286,
+      "learning_rate": 4.98887168453085e-07,
+      "loss": 0.0249,
+      "step": 7842
+    },
+    {
+      "epoch": 3.483455474128359,
+      "grad_norm": 0.3887321324351885,
+      "learning_rate": 4.980434096121106e-07,
+      "loss": 0.0299,
+      "step": 7843
+    },
+    {
+      "epoch": 3.4838996224739063,
+      "grad_norm": 0.403812873928378,
+      "learning_rate": 4.97200327491687e-07,
+      "loss": 0.0219,
+      "step": 7844
+    },
+    {
+      "epoch": 3.4843437708194536,
+      "grad_norm": 0.4454713027127471,
+      "learning_rate": 4.963579222185444e-07,
+      "loss": 0.0247,
+      "step": 7845
+    },
+    {
+      "epoch": 3.4847879191650013,
+      "grad_norm": 0.41374791040066133,
+      "learning_rate": 4.955161939193087e-07,
+      "loss": 0.0213,
+      "step": 7846
+    },
+    {
+      "epoch": 3.4852320675105486,
+      "grad_norm": 0.460381233413127,
+      "learning_rate": 4.946751427205054e-07,
+      "loss": 0.0359,
+      "step": 7847
+    },
+    {
+      "epoch": 3.485676215856096,
+      "grad_norm": 0.5274863598727876,
+      "learning_rate": 4.938347687485629e-07,
+      "loss": 0.0249,
+      "step": 7848
+    },
+    {
+      "epoch": 3.486120364201643,
+      "grad_norm": 0.3376353042657825,
+      "learning_rate": 4.929950721297993e-07,
+      "loss": 0.0197,
+      "step": 7849
+    },
+    {
+      "epoch": 3.486564512547191,
+      "grad_norm": 0.4896159029658331,
+      "learning_rate": 4.921560529904374e-07,
+      "loss": 0.0291,
+      "step": 7850
+    },
+    {
+      "epoch": 3.487008660892738,
+      "grad_norm": 0.36649328899036365,
+      "learning_rate": 4.913177114565964e-07,
+      "loss": 0.0198,
+      "step": 7851
+    },
+    {
+      "epoch": 3.4874528092382855,
+      "grad_norm": 0.46757175677374957,
+      "learning_rate": 4.90480047654291e-07,
+      "loss": 0.0359,
+      "step": 7852
+    },
+    {
+      "epoch": 3.487896957583833,
+      "grad_norm": 0.4027846982490894,
+      "learning_rate": 4.896430617094389e-07,
+      "loss": 0.0284,
+      "step": 7853
+    },
+    {
+      "epoch": 3.4883411059293805,
+      "grad_norm": 0.41506130883345627,
+      "learning_rate": 4.888067537478519e-07,
+      "loss": 0.0253,
+      "step": 7854
+    },
+    {
+      "epoch": 3.488785254274928,
+      "grad_norm": 0.4526519663181836,
+      "learning_rate": 4.879711238952412e-07,
+      "loss": 0.0327,
+      "step": 7855
+    },
+    {
+      "epoch": 3.489229402620475,
+      "grad_norm": 0.3779652676361818,
+      "learning_rate": 4.871361722772166e-07,
+      "loss": 0.0336,
+      "step": 7856
+    },
+    {
+      "epoch": 3.489673550966023,
+      "grad_norm": 0.6066583572691198,
+      "learning_rate": 4.86301899019287e-07,
+      "loss": 0.0277,
+      "step": 7857
+    },
+    {
+      "epoch": 3.49011769931157,
+      "grad_norm": 0.516592567980747,
+      "learning_rate": 4.854683042468538e-07,
+      "loss": 0.0345,
+      "step": 7858
+    },
+    {
+      "epoch": 3.4905618476571174,
+      "grad_norm": 0.3300726184397244,
+      "learning_rate": 4.84635388085225e-07,
+      "loss": 0.0208,
+      "step": 7859
+    },
+    {
+      "epoch": 3.4910059960026647,
+      "grad_norm": 0.4075507938842562,
+      "learning_rate": 4.838031506595992e-07,
+      "loss": 0.0253,
+      "step": 7860
+    },
+    {
+      "epoch": 3.4914501443482124,
+      "grad_norm": 0.41601787060330486,
+      "learning_rate": 4.829715920950761e-07,
+      "loss": 0.0237,
+      "step": 7861
+    },
+    {
+      "epoch": 3.4918942926937597,
+      "grad_norm": 0.5863384888723406,
+      "learning_rate": 4.821407125166549e-07,
+      "loss": 0.0264,
+      "step": 7862
+    },
+    {
+      "epoch": 3.492338441039307,
+      "grad_norm": 0.47433808650477144,
+      "learning_rate": 4.81310512049229e-07,
+      "loss": 0.0285,
+      "step": 7863
+    },
+    {
+      "epoch": 3.4927825893848548,
+      "grad_norm": 0.4906763276486267,
+      "learning_rate": 4.804809908175911e-07,
+      "loss": 0.0251,
+      "step": 7864
+    },
+    {
+      "epoch": 3.493226737730402,
+      "grad_norm": 0.46826469445803337,
+      "learning_rate": 4.796521489464351e-07,
+      "loss": 0.0269,
+      "step": 7865
+    },
+    {
+      "epoch": 3.4936708860759493,
+      "grad_norm": 0.3869063106968133,
+      "learning_rate": 4.788239865603478e-07,
+      "loss": 0.0304,
+      "step": 7866
+    },
+    {
+      "epoch": 3.4941150344214966,
+      "grad_norm": 0.4155671863362175,
+      "learning_rate": 4.779965037838164e-07,
+      "loss": 0.0222,
+      "step": 7867
+    },
+    {
+      "epoch": 3.4945591827670444,
+      "grad_norm": 0.40399166433294564,
+      "learning_rate": 4.771697007412268e-07,
+      "loss": 0.0296,
+      "step": 7868
+    },
+    {
+      "epoch": 3.4950033311125916,
+      "grad_norm": 0.4132725467300106,
+      "learning_rate": 4.763435775568592e-07,
+      "loss": 0.0263,
+      "step": 7869
+    },
+    {
+      "epoch": 3.495447479458139,
+      "grad_norm": 0.40246185290766967,
+      "learning_rate": 4.7551813435489703e-07,
+      "loss": 0.0256,
+      "step": 7870
+    },
+    {
+      "epoch": 3.4958916278036867,
+      "grad_norm": 0.4363439376254432,
+      "learning_rate": 4.746933712594154e-07,
+      "loss": 0.0303,
+      "step": 7871
+    },
+    {
+      "epoch": 3.496335776149234,
+      "grad_norm": 0.346578595904209,
+      "learning_rate": 4.7386928839439183e-07,
+      "loss": 0.0216,
+      "step": 7872
+    },
+    {
+      "epoch": 3.4967799244947813,
+      "grad_norm": 0.40218292971261804,
+      "learning_rate": 4.7304588588370113e-07,
+      "loss": 0.022,
+      "step": 7873
+    },
+    {
+      "epoch": 3.4972240728403285,
+      "grad_norm": 0.3613752874610863,
+      "learning_rate": 4.722231638511121e-07,
+      "loss": 0.0188,
+      "step": 7874
+    },
+    {
+      "epoch": 3.497668221185876,
+      "grad_norm": 0.4305489342659995,
+      "learning_rate": 4.7140112242029356e-07,
+      "loss": 0.0244,
+      "step": 7875
+    },
+    {
+      "epoch": 3.4981123695314236,
+      "grad_norm": 0.3319846218251371,
+      "learning_rate": 4.7057976171481614e-07,
+      "loss": 0.0159,
+      "step": 7876
+    },
+    {
+      "epoch": 3.498556517876971,
+      "grad_norm": 0.3685184737755979,
+      "learning_rate": 4.69759081858141e-07,
+      "loss": 0.0248,
+      "step": 7877
+    },
+    {
+      "epoch": 3.499000666222518,
+      "grad_norm": 0.5404043718315673,
+      "learning_rate": 4.689390829736312e-07,
+      "loss": 0.031,
+      "step": 7878
+    },
+    {
+      "epoch": 3.499444814568066,
+      "grad_norm": 0.3833802780855093,
+      "learning_rate": 4.681197651845476e-07,
+      "loss": 0.0156,
+      "step": 7879
+    },
+    {
+      "epoch": 3.499888962913613,
+      "grad_norm": 0.387890807698906,
+      "learning_rate": 4.6730112861404497e-07,
+      "loss": 0.028,
+      "step": 7880
+    },
+    {
+      "epoch": 3.5003331112591605,
+      "grad_norm": 0.5278897289398781,
+      "learning_rate": 4.6648317338518045e-07,
+      "loss": 0.0305,
+      "step": 7881
+    },
+    {
+      "epoch": 3.5007772596047078,
+      "grad_norm": 0.429242176272356,
+      "learning_rate": 4.656658996209057e-07,
+      "loss": 0.0299,
+      "step": 7882
+    },
+    {
+      "epoch": 3.5012214079502555,
+      "grad_norm": 0.4369670295816076,
+      "learning_rate": 4.6484930744407074e-07,
+      "loss": 0.0327,
+      "step": 7883
+    },
+    {
+      "epoch": 3.501665556295803,
+      "grad_norm": 0.4901187441061219,
+      "learning_rate": 4.6403339697742413e-07,
+      "loss": 0.0228,
+      "step": 7884
+    },
+    {
+      "epoch": 3.50210970464135,
+      "grad_norm": 0.3420486455934448,
+      "learning_rate": 4.63218168343611e-07,
+      "loss": 0.0216,
+      "step": 7885
+    },
+    {
+      "epoch": 3.502553852986898,
+      "grad_norm": 0.40939983156980164,
+      "learning_rate": 4.624036216651723e-07,
+      "loss": 0.0263,
+      "step": 7886
+    },
+    {
+      "epoch": 3.502998001332445,
+      "grad_norm": 0.3910656718501315,
+      "learning_rate": 4.615897570645511e-07,
+      "loss": 0.0183,
+      "step": 7887
+    },
+    {
+      "epoch": 3.5034421496779924,
+      "grad_norm": 0.4314655959506559,
+      "learning_rate": 4.6077657466408245e-07,
+      "loss": 0.0312,
+      "step": 7888
+    },
+    {
+      "epoch": 3.5038862980235397,
+      "grad_norm": 0.3568409941841216,
+      "learning_rate": 4.599640745860029e-07,
+      "loss": 0.0271,
+      "step": 7889
+    },
+    {
+      "epoch": 3.5043304463690874,
+      "grad_norm": 0.4093953752586413,
+      "learning_rate": 4.5915225695244536e-07,
+      "loss": 0.0241,
+      "step": 7890
+    },
+    {
+      "epoch": 3.5047745947146347,
+      "grad_norm": 0.3977398657560647,
+      "learning_rate": 4.583411218854383e-07,
+      "loss": 0.0237,
+      "step": 7891
+    },
+    {
+      "epoch": 3.505218743060182,
+      "grad_norm": 0.4275747909847334,
+      "learning_rate": 4.575306695069087e-07,
+      "loss": 0.0301,
+      "step": 7892
+    },
+    {
+      "epoch": 3.5056628914057297,
+      "grad_norm": 0.5388719609229746,
+      "learning_rate": 4.567208999386852e-07,
+      "loss": 0.035,
+      "step": 7893
+    },
+    {
+      "epoch": 3.506107039751277,
+      "grad_norm": 0.36096697143744105,
+      "learning_rate": 4.5591181330248534e-07,
+      "loss": 0.0207,
+      "step": 7894
+    },
+    {
+      "epoch": 3.5065511880968243,
+      "grad_norm": 0.4142429440699439,
+      "learning_rate": 4.5510340971993086e-07,
+      "loss": 0.0186,
+      "step": 7895
+    },
+    {
+      "epoch": 3.5069953364423716,
+      "grad_norm": 0.4393049642057595,
+      "learning_rate": 4.542956893125394e-07,
+      "loss": 0.025,
+      "step": 7896
+    },
+    {
+      "epoch": 3.507439484787919,
+      "grad_norm": 0.6384180216184727,
+      "learning_rate": 4.534886522017229e-07,
+      "loss": 0.0266,
+      "step": 7897
+    },
+    {
+      "epoch": 3.5078836331334666,
+      "grad_norm": 0.43433435288410693,
+      "learning_rate": 4.526822985087931e-07,
+      "loss": 0.031,
+      "step": 7898
+    },
+    {
+      "epoch": 3.508327781479014,
+      "grad_norm": 0.4028630240642223,
+      "learning_rate": 4.5187662835495974e-07,
+      "loss": 0.0225,
+      "step": 7899
+    },
+    {
+      "epoch": 3.5087719298245617,
+      "grad_norm": 0.3939689206368112,
+      "learning_rate": 4.510716418613281e-07,
+      "loss": 0.0224,
+      "step": 7900
+    },
+    {
+      "epoch": 3.509216078170109,
+      "grad_norm": 0.515126016444997,
+      "learning_rate": 4.502673391489026e-07,
+      "loss": 0.029,
+      "step": 7901
+    },
+    {
+      "epoch": 3.5096602265156562,
+      "grad_norm": 0.4945699098085951,
+      "learning_rate": 4.4946372033858157e-07,
+      "loss": 0.028,
+      "step": 7902
+    },
+    {
+      "epoch": 3.5101043748612035,
+      "grad_norm": 0.43373798757249604,
+      "learning_rate": 4.486607855511627e-07,
+      "loss": 0.0272,
+      "step": 7903
+    },
+    {
+      "epoch": 3.510548523206751,
+      "grad_norm": 0.40314045697996204,
+      "learning_rate": 4.4785853490734277e-07,
+      "loss": 0.027,
+      "step": 7904
+    },
+    {
+      "epoch": 3.5109926715522985,
+      "grad_norm": 0.4288828381204982,
+      "learning_rate": 4.470569685277115e-07,
+      "loss": 0.021,
+      "step": 7905
+    },
+    {
+      "epoch": 3.511436819897846,
+      "grad_norm": 0.47380514557902603,
+      "learning_rate": 4.462560865327592e-07,
+      "loss": 0.0352,
+      "step": 7906
+    },
+    {
+      "epoch": 3.511880968243393,
+      "grad_norm": 0.6718664840187395,
+      "learning_rate": 4.454558890428728e-07,
+      "loss": 0.037,
+      "step": 7907
+    },
+    {
+      "epoch": 3.512325116588941,
+      "grad_norm": 0.42187384792878063,
+      "learning_rate": 4.446563761783329e-07,
+      "loss": 0.0265,
+      "step": 7908
+    },
+    {
+      "epoch": 3.512769264934488,
+      "grad_norm": 0.4115565734686187,
+      "learning_rate": 4.43857548059321e-07,
+      "loss": 0.0292,
+      "step": 7909
+    },
+    {
+      "epoch": 3.5132134132800354,
+      "grad_norm": 0.5047831437385433,
+      "learning_rate": 4.430594048059167e-07,
+      "loss": 0.0283,
+      "step": 7910
+    },
+    {
+      "epoch": 3.5136575616255827,
+      "grad_norm": 0.4497401686326858,
+      "learning_rate": 4.422619465380917e-07,
+      "loss": 0.0257,
+      "step": 7911
+    },
+    {
+      "epoch": 3.5141017099711305,
+      "grad_norm": 0.41364265915281817,
+      "learning_rate": 4.4146517337571857e-07,
+      "loss": 0.0219,
+      "step": 7912
+    },
+    {
+      "epoch": 3.5145458583166778,
+      "grad_norm": 0.5685124202462428,
+      "learning_rate": 4.4066908543856704e-07,
+      "loss": 0.0421,
+      "step": 7913
+    },
+    {
+      "epoch": 3.514990006662225,
+      "grad_norm": 0.31669420534645987,
+      "learning_rate": 4.3987368284630015e-07,
+      "loss": 0.0161,
+      "step": 7914
+    },
+    {
+      "epoch": 3.515434155007773,
+      "grad_norm": 0.3911190961806719,
+      "learning_rate": 4.3907896571848187e-07,
+      "loss": 0.0295,
+      "step": 7915
+    },
+    {
+      "epoch": 3.51587830335332,
+      "grad_norm": 0.49151230231883697,
+      "learning_rate": 4.382849341745715e-07,
+      "loss": 0.0323,
+      "step": 7916
+    },
+    {
+      "epoch": 3.5163224516988674,
+      "grad_norm": 0.3734463542020498,
+      "learning_rate": 4.3749158833392535e-07,
+      "loss": 0.0233,
+      "step": 7917
+    },
+    {
+      "epoch": 3.5167666000444147,
+      "grad_norm": 0.4234162479441925,
+      "learning_rate": 4.366989283157985e-07,
+      "loss": 0.0221,
+      "step": 7918
+    },
+    {
+      "epoch": 3.5172107483899624,
+      "grad_norm": 0.36581058726451665,
+      "learning_rate": 4.3590695423933795e-07,
+      "loss": 0.0232,
+      "step": 7919
+    },
+    {
+      "epoch": 3.5176548967355097,
+      "grad_norm": 0.41877616966330095,
+      "learning_rate": 4.3511566622359224e-07,
+      "loss": 0.032,
+      "step": 7920
+    },
+    {
+      "epoch": 3.518099045081057,
+      "grad_norm": 0.381478457217342,
+      "learning_rate": 4.3432506438750745e-07,
+      "loss": 0.0198,
+      "step": 7921
+    },
+    {
+      "epoch": 3.5185431934266047,
+      "grad_norm": 0.5312067893718723,
+      "learning_rate": 4.335351488499218e-07,
+      "loss": 0.0341,
+      "step": 7922
+    },
+    {
+      "epoch": 3.518987341772152,
+      "grad_norm": 0.32613641832900486,
+      "learning_rate": 4.327459197295736e-07,
+      "loss": 0.0153,
+      "step": 7923
+    },
+    {
+      "epoch": 3.5194314901176993,
+      "grad_norm": 0.4186436124064849,
+      "learning_rate": 4.319573771450991e-07,
+      "loss": 0.0239,
+      "step": 7924
+    },
+    {
+      "epoch": 3.5198756384632466,
+      "grad_norm": 0.5213492953580083,
+      "learning_rate": 4.3116952121502686e-07,
+      "loss": 0.0245,
+      "step": 7925
+    },
+    {
+      "epoch": 3.520319786808794,
+      "grad_norm": 0.4470067109711501,
+      "learning_rate": 4.303823520577871e-07,
+      "loss": 0.029,
+      "step": 7926
+    },
+    {
+      "epoch": 3.5207639351543416,
+      "grad_norm": 0.44926277605141984,
+      "learning_rate": 4.295958697917035e-07,
+      "loss": 0.0251,
+      "step": 7927
+    },
+    {
+      "epoch": 3.521208083499889,
+      "grad_norm": 0.4428984780991236,
+      "learning_rate": 4.288100745349988e-07,
+      "loss": 0.0283,
+      "step": 7928
+    },
+    {
+      "epoch": 3.5216522318454366,
+      "grad_norm": 0.4649455292363394,
+      "learning_rate": 4.2802496640579115e-07,
+      "loss": 0.0261,
+      "step": 7929
+    },
+    {
+      "epoch": 3.522096380190984,
+      "grad_norm": 0.4546600325025468,
+      "learning_rate": 4.2724054552209515e-07,
+      "loss": 0.0272,
+      "step": 7930
+    },
+    {
+      "epoch": 3.522540528536531,
+      "grad_norm": 0.44460414877352883,
+      "learning_rate": 4.2645681200182197e-07,
+      "loss": 0.02,
+      "step": 7931
+    },
+    {
+      "epoch": 3.5229846768820785,
+      "grad_norm": 0.3712603269325949,
+      "learning_rate": 4.256737659627813e-07,
+      "loss": 0.0252,
+      "step": 7932
+    },
+    {
+      "epoch": 3.523428825227626,
+      "grad_norm": 0.4419446907695471,
+      "learning_rate": 4.248914075226779e-07,
+      "loss": 0.0296,
+      "step": 7933
+    },
+    {
+      "epoch": 3.5238729735731735,
+      "grad_norm": 0.4371622849898781,
+      "learning_rate": 4.2410973679911317e-07,
+      "loss": 0.0193,
+      "step": 7934
+    },
+    {
+      "epoch": 3.524317121918721,
+      "grad_norm": 0.42263243970502656,
+      "learning_rate": 4.2332875390958707e-07,
+      "loss": 0.0284,
+      "step": 7935
+    },
+    {
+      "epoch": 3.524761270264268,
+      "grad_norm": 0.4360454559382049,
+      "learning_rate": 4.225484589714918e-07,
+      "loss": 0.0267,
+      "step": 7936
+    },
+    {
+      "epoch": 3.525205418609816,
+      "grad_norm": 0.4221730269522906,
+      "learning_rate": 4.2176885210212127e-07,
+      "loss": 0.0283,
+      "step": 7937
+    },
+    {
+      "epoch": 3.525649566955363,
+      "grad_norm": 0.34554558425959747,
+      "learning_rate": 4.209899334186623e-07,
+      "loss": 0.015,
+      "step": 7938
+    },
+    {
+      "epoch": 3.5260937153009104,
+      "grad_norm": 0.3944057611136553,
+      "learning_rate": 4.2021170303820025e-07,
+      "loss": 0.0247,
+      "step": 7939
+    },
+    {
+      "epoch": 3.5265378636464577,
+      "grad_norm": 0.3950267732155638,
+      "learning_rate": 4.1943416107771585e-07,
+      "loss": 0.0237,
+      "step": 7940
+    },
+    {
+      "epoch": 3.5269820119920055,
+      "grad_norm": 0.5276030208132706,
+      "learning_rate": 4.186573076540884e-07,
+      "loss": 0.0293,
+      "step": 7941
+    },
+    {
+      "epoch": 3.5274261603375527,
+      "grad_norm": 0.42130456420073215,
+      "learning_rate": 4.178811428840901e-07,
+      "loss": 0.0221,
+      "step": 7942
+    },
+    {
+      "epoch": 3.5278703086831,
+      "grad_norm": 0.4967582028112826,
+      "learning_rate": 4.1710566688439314e-07,
+      "loss": 0.0264,
+      "step": 7943
+    },
+    {
+      "epoch": 3.5283144570286478,
+      "grad_norm": 0.38558337467937864,
+      "learning_rate": 4.163308797715637e-07,
+      "loss": 0.0276,
+      "step": 7944
+    },
+    {
+      "epoch": 3.528758605374195,
+      "grad_norm": 0.5044853945425457,
+      "learning_rate": 4.155567816620659e-07,
+      "loss": 0.0204,
+      "step": 7945
+    },
+    {
+      "epoch": 3.5292027537197423,
+      "grad_norm": 0.8869805733234495,
+      "learning_rate": 4.147833726722611e-07,
+      "loss": 0.0291,
+      "step": 7946
+    },
+    {
+      "epoch": 3.5296469020652896,
+      "grad_norm": 0.3536249761317266,
+      "learning_rate": 4.140106529184035e-07,
+      "loss": 0.0235,
+      "step": 7947
+    },
+    {
+      "epoch": 3.5300910504108374,
+      "grad_norm": 0.4154592460581955,
+      "learning_rate": 4.1323862251664684e-07,
+      "loss": 0.0233,
+      "step": 7948
+    },
+    {
+      "epoch": 3.5305351987563847,
+      "grad_norm": 0.3504099993535625,
+      "learning_rate": 4.1246728158304107e-07,
+      "loss": 0.0187,
+      "step": 7949
+    },
+    {
+      "epoch": 3.530979347101932,
+      "grad_norm": 0.4437614260756121,
+      "learning_rate": 4.1169663023353124e-07,
+      "loss": 0.0262,
+      "step": 7950
+    },
+    {
+      "epoch": 3.5314234954474797,
+      "grad_norm": 0.3600199231026137,
+      "learning_rate": 4.109266685839597e-07,
+      "loss": 0.02,
+      "step": 7951
+    },
+    {
+      "epoch": 3.531867643793027,
+      "grad_norm": 0.3877433229131516,
+      "learning_rate": 4.101573967500655e-07,
+      "loss": 0.0236,
+      "step": 7952
+    },
+    {
+      "epoch": 3.5323117921385743,
+      "grad_norm": 0.39474586031469927,
+      "learning_rate": 4.0938881484748116e-07,
+      "loss": 0.0261,
+      "step": 7953
+    },
+    {
+      "epoch": 3.5327559404841216,
+      "grad_norm": 0.4148159595077121,
+      "learning_rate": 4.086209229917387e-07,
+      "loss": 0.0255,
+      "step": 7954
+    },
+    {
+      "epoch": 3.533200088829669,
+      "grad_norm": 0.47070012710839254,
+      "learning_rate": 4.0785372129826586e-07,
+      "loss": 0.0318,
+      "step": 7955
+    },
+    {
+      "epoch": 3.5336442371752166,
+      "grad_norm": 0.3680844624737353,
+      "learning_rate": 4.0708720988238584e-07,
+      "loss": 0.0166,
+      "step": 7956
+    },
+    {
+      "epoch": 3.534088385520764,
+      "grad_norm": 0.3875786535924789,
+      "learning_rate": 4.063213888593176e-07,
+      "loss": 0.0256,
+      "step": 7957
+    },
+    {
+      "epoch": 3.5345325338663116,
+      "grad_norm": 0.39292909474931037,
+      "learning_rate": 4.0555625834417857e-07,
+      "loss": 0.0244,
+      "step": 7958
+    },
+    {
+      "epoch": 3.534976682211859,
+      "grad_norm": 0.4238019487678821,
+      "learning_rate": 4.047918184519789e-07,
+      "loss": 0.0265,
+      "step": 7959
+    },
+    {
+      "epoch": 3.535420830557406,
+      "grad_norm": 0.3868503914395472,
+      "learning_rate": 4.040280692976278e-07,
+      "loss": 0.025,
+      "step": 7960
+    },
+    {
+      "epoch": 3.5358649789029535,
+      "grad_norm": 0.5061016125162293,
+      "learning_rate": 4.032650109959302e-07,
+      "loss": 0.0305,
+      "step": 7961
+    },
+    {
+      "epoch": 3.5363091272485008,
+      "grad_norm": 0.44915195593836177,
+      "learning_rate": 4.0250264366158643e-07,
+      "loss": 0.0227,
+      "step": 7962
+    },
+    {
+      "epoch": 3.5367532755940485,
+      "grad_norm": 0.3701090299363028,
+      "learning_rate": 4.017409674091932e-07,
+      "loss": 0.0275,
+      "step": 7963
+    },
+    {
+      "epoch": 3.537197423939596,
+      "grad_norm": 0.33071866294982566,
+      "learning_rate": 4.009799823532434e-07,
+      "loss": 0.0166,
+      "step": 7964
+    },
+    {
+      "epoch": 3.537641572285143,
+      "grad_norm": 0.3869700046276894,
+      "learning_rate": 4.0021968860812556e-07,
+      "loss": 0.025,
+      "step": 7965
+    },
+    {
+      "epoch": 3.538085720630691,
+      "grad_norm": 0.34946266928394254,
+      "learning_rate": 3.994600862881248e-07,
+      "loss": 0.0205,
+      "step": 7966
+    },
+    {
+      "epoch": 3.538529868976238,
+      "grad_norm": 0.391173218273872,
+      "learning_rate": 3.9870117550742273e-07,
+      "loss": 0.0263,
+      "step": 7967
+    },
+    {
+      "epoch": 3.5389740173217854,
+      "grad_norm": 0.3665430028174577,
+      "learning_rate": 3.9794295638009683e-07,
+      "loss": 0.022,
+      "step": 7968
+    },
+    {
+      "epoch": 3.5394181656673327,
+      "grad_norm": 0.3931083199354571,
+      "learning_rate": 3.971854290201205e-07,
+      "loss": 0.0233,
+      "step": 7969
+    },
+    {
+      "epoch": 3.5398623140128804,
+      "grad_norm": 0.4886210610765913,
+      "learning_rate": 3.964285935413609e-07,
+      "loss": 0.0213,
+      "step": 7970
+    },
+    {
+      "epoch": 3.5403064623584277,
+      "grad_norm": 0.3576029073891739,
+      "learning_rate": 3.9567245005758537e-07,
+      "loss": 0.023,
+      "step": 7971
+    },
+    {
+      "epoch": 3.540750610703975,
+      "grad_norm": 0.4994177352645637,
+      "learning_rate": 3.9491699868245414e-07,
+      "loss": 0.0419,
+      "step": 7972
+    },
+    {
+      "epoch": 3.5411947590495227,
+      "grad_norm": 0.43622296138249816,
+      "learning_rate": 3.941622395295247e-07,
+      "loss": 0.0318,
+      "step": 7973
+    },
+    {
+      "epoch": 3.54163890739507,
+      "grad_norm": 0.4309663417901952,
+      "learning_rate": 3.934081727122513e-07,
+      "loss": 0.0342,
+      "step": 7974
+    },
+    {
+      "epoch": 3.5420830557406173,
+      "grad_norm": 0.4349122435505889,
+      "learning_rate": 3.9265479834398103e-07,
+      "loss": 0.0286,
+      "step": 7975
+    },
+    {
+      "epoch": 3.5425272040861646,
+      "grad_norm": 0.4420907520326438,
+      "learning_rate": 3.919021165379594e-07,
+      "loss": 0.0192,
+      "step": 7976
+    },
+    {
+      "epoch": 3.5429713524317124,
+      "grad_norm": 0.40134602793452784,
+      "learning_rate": 3.911501274073276e-07,
+      "loss": 0.0222,
+      "step": 7977
+    },
+    {
+      "epoch": 3.5434155007772596,
+      "grad_norm": 0.4162128270891882,
+      "learning_rate": 3.9039883106512243e-07,
+      "loss": 0.0263,
+      "step": 7978
+    },
+    {
+      "epoch": 3.543859649122807,
+      "grad_norm": 0.43122424865043063,
+      "learning_rate": 3.8964822762427633e-07,
+      "loss": 0.0251,
+      "step": 7979
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.351905898436228,
+      "learning_rate": 3.888983171976185e-07,
+      "loss": 0.0225,
+      "step": 7980
+    },
+    {
+      "epoch": 3.544747945813902,
+      "grad_norm": 0.37615773316072754,
+      "learning_rate": 3.8814909989787155e-07,
+      "loss": 0.02,
+      "step": 7981
+    },
+    {
+      "epoch": 3.5451920941594492,
+      "grad_norm": 0.42395585895532933,
+      "learning_rate": 3.87400575837657e-07,
+      "loss": 0.0202,
+      "step": 7982
+    },
+    {
+      "epoch": 3.5456362425049965,
+      "grad_norm": 0.3990970301083849,
+      "learning_rate": 3.8665274512948994e-07,
+      "loss": 0.0201,
+      "step": 7983
+    },
+    {
+      "epoch": 3.546080390850544,
+      "grad_norm": 0.5764101771289384,
+      "learning_rate": 3.859056078857826e-07,
+      "loss": 0.0381,
+      "step": 7984
+    },
+    {
+      "epoch": 3.5465245391960916,
+      "grad_norm": 0.3820488932414435,
+      "learning_rate": 3.851591642188418e-07,
+      "loss": 0.0242,
+      "step": 7985
+    },
+    {
+      "epoch": 3.546968687541639,
+      "grad_norm": 0.3459706819898099,
+      "learning_rate": 3.8441341424087233e-07,
+      "loss": 0.0238,
+      "step": 7986
+    },
+    {
+      "epoch": 3.5474128358871866,
+      "grad_norm": 0.4227398045664523,
+      "learning_rate": 3.836683580639705e-07,
+      "loss": 0.0244,
+      "step": 7987
+    },
+    {
+      "epoch": 3.547856984232734,
+      "grad_norm": 0.4309060643690936,
+      "learning_rate": 3.829239958001324e-07,
+      "loss": 0.0194,
+      "step": 7988
+    },
+    {
+      "epoch": 3.548301132578281,
+      "grad_norm": 0.42116836328880336,
+      "learning_rate": 3.8218032756124844e-07,
+      "loss": 0.0204,
+      "step": 7989
+    },
+    {
+      "epoch": 3.5487452809238285,
+      "grad_norm": 0.37821657931686165,
+      "learning_rate": 3.814373534591037e-07,
+      "loss": 0.0194,
+      "step": 7990
+    },
+    {
+      "epoch": 3.5491894292693758,
+      "grad_norm": 0.42284081107970184,
+      "learning_rate": 3.8069507360538163e-07,
+      "loss": 0.0221,
+      "step": 7991
+    },
+    {
+      "epoch": 3.5496335776149235,
+      "grad_norm": 0.4662610237314549,
+      "learning_rate": 3.799534881116573e-07,
+      "loss": 0.0276,
+      "step": 7992
+    },
+    {
+      "epoch": 3.5500777259604708,
+      "grad_norm": 0.45003555474688933,
+      "learning_rate": 3.7921259708940503e-07,
+      "loss": 0.0274,
+      "step": 7993
+    },
+    {
+      "epoch": 3.550521874306018,
+      "grad_norm": 0.5571173920439147,
+      "learning_rate": 3.7847240064999233e-07,
+      "loss": 0.0294,
+      "step": 7994
+    },
+    {
+      "epoch": 3.550966022651566,
+      "grad_norm": 0.4728510369440864,
+      "learning_rate": 3.7773289890468414e-07,
+      "loss": 0.026,
+      "step": 7995
+    },
+    {
+      "epoch": 3.551410170997113,
+      "grad_norm": 0.3553733406102519,
+      "learning_rate": 3.7699409196463977e-07,
+      "loss": 0.0175,
+      "step": 7996
+    },
+    {
+      "epoch": 3.5518543193426604,
+      "grad_norm": 0.4021957600650995,
+      "learning_rate": 3.762559799409149e-07,
+      "loss": 0.0268,
+      "step": 7997
+    },
+    {
+      "epoch": 3.5522984676882077,
+      "grad_norm": 0.5246882857506893,
+      "learning_rate": 3.7551856294445967e-07,
+      "loss": 0.0243,
+      "step": 7998
+    },
+    {
+      "epoch": 3.5527426160337554,
+      "grad_norm": 0.4805020243101749,
+      "learning_rate": 3.7478184108612036e-07,
+      "loss": 0.0185,
+      "step": 7999
+    },
+    {
+      "epoch": 3.5531867643793027,
+      "grad_norm": 0.49276293248029474,
+      "learning_rate": 3.74045814476639e-07,
+      "loss": 0.0322,
+      "step": 8000
+    },
+    {
+      "epoch": 3.55363091272485,
+      "grad_norm": 0.5382419991792012,
+      "learning_rate": 3.733104832266532e-07,
+      "loss": 0.0273,
+      "step": 8001
+    },
+    {
+      "epoch": 3.5540750610703977,
+      "grad_norm": 0.4236743982252115,
+      "learning_rate": 3.7257584744669615e-07,
+      "loss": 0.024,
+      "step": 8002
+    },
+    {
+      "epoch": 3.554519209415945,
+      "grad_norm": 0.3569656738216824,
+      "learning_rate": 3.718419072471946e-07,
+      "loss": 0.0206,
+      "step": 8003
+    },
+    {
+      "epoch": 3.5549633577614923,
+      "grad_norm": 0.6957355605647527,
+      "learning_rate": 3.7110866273847356e-07,
+      "loss": 0.0286,
+      "step": 8004
+    },
+    {
+      "epoch": 3.5554075061070396,
+      "grad_norm": 0.39146983502282784,
+      "learning_rate": 3.70376114030751e-07,
+      "loss": 0.0211,
+      "step": 8005
+    },
+    {
+      "epoch": 3.5558516544525873,
+      "grad_norm": 0.4183758252296795,
+      "learning_rate": 3.696442612341422e-07,
+      "loss": 0.023,
+      "step": 8006
+    },
+    {
+      "epoch": 3.5562958027981346,
+      "grad_norm": 0.4440464207564376,
+      "learning_rate": 3.6891310445865693e-07,
+      "loss": 0.0284,
+      "step": 8007
+    },
+    {
+      "epoch": 3.556739951143682,
+      "grad_norm": 0.3721283097773306,
+      "learning_rate": 3.681826438142011e-07,
+      "loss": 0.0228,
+      "step": 8008
+    },
+    {
+      "epoch": 3.5571840994892296,
+      "grad_norm": 0.4139366078303485,
+      "learning_rate": 3.6745287941057417e-07,
+      "loss": 0.0229,
+      "step": 8009
+    },
+    {
+      "epoch": 3.557628247834777,
+      "grad_norm": 0.5055064614278328,
+      "learning_rate": 3.6672381135747284e-07,
+      "loss": 0.0278,
+      "step": 8010
+    },
+    {
+      "epoch": 3.5580723961803242,
+      "grad_norm": 0.44529281390449127,
+      "learning_rate": 3.6599543976448884e-07,
+      "loss": 0.0261,
+      "step": 8011
+    },
+    {
+      "epoch": 3.5585165445258715,
+      "grad_norm": 0.4907105184886079,
+      "learning_rate": 3.6526776474110627e-07,
+      "loss": 0.0244,
+      "step": 8012
+    },
+    {
+      "epoch": 3.558960692871419,
+      "grad_norm": 0.4951645075676849,
+      "learning_rate": 3.645407863967104e-07,
+      "loss": 0.037,
+      "step": 8013
+    },
+    {
+      "epoch": 3.5594048412169665,
+      "grad_norm": 0.37863500443452736,
+      "learning_rate": 3.6381450484057777e-07,
+      "loss": 0.0252,
+      "step": 8014
+    },
+    {
+      "epoch": 3.559848989562514,
+      "grad_norm": 0.39735984529923846,
+      "learning_rate": 3.630889201818788e-07,
+      "loss": 0.0206,
+      "step": 8015
+    },
+    {
+      "epoch": 3.5602931379080616,
+      "grad_norm": 0.4346987177337589,
+      "learning_rate": 3.623640325296829e-07,
+      "loss": 0.0257,
+      "step": 8016
+    },
+    {
+      "epoch": 3.560737286253609,
+      "grad_norm": 0.33730436829577654,
+      "learning_rate": 3.616398419929523e-07,
+      "loss": 0.0189,
+      "step": 8017
+    },
+    {
+      "epoch": 3.561181434599156,
+      "grad_norm": 0.3665779162327361,
+      "learning_rate": 3.6091634868054557e-07,
+      "loss": 0.0189,
+      "step": 8018
+    },
+    {
+      "epoch": 3.5616255829447034,
+      "grad_norm": 0.3886589674388881,
+      "learning_rate": 3.601935527012168e-07,
+      "loss": 0.0225,
+      "step": 8019
+    },
+    {
+      "epoch": 3.5620697312902507,
+      "grad_norm": 0.46005381668223383,
+      "learning_rate": 3.594714541636124e-07,
+      "loss": 0.0251,
+      "step": 8020
+    },
+    {
+      "epoch": 3.5625138796357985,
+      "grad_norm": 0.437136127998239,
+      "learning_rate": 3.5875005317627776e-07,
+      "loss": 0.0336,
+      "step": 8021
+    },
+    {
+      "epoch": 3.5629580279813458,
+      "grad_norm": 0.41144565522854637,
+      "learning_rate": 3.580293498476517e-07,
+      "loss": 0.0321,
+      "step": 8022
+    },
+    {
+      "epoch": 3.563402176326893,
+      "grad_norm": 0.4402714548710384,
+      "learning_rate": 3.573093442860659e-07,
+      "loss": 0.0218,
+      "step": 8023
+    },
+    {
+      "epoch": 3.563846324672441,
+      "grad_norm": 0.46313910275690856,
+      "learning_rate": 3.565900365997521e-07,
+      "loss": 0.038,
+      "step": 8024
+    },
+    {
+      "epoch": 3.564290473017988,
+      "grad_norm": 0.4585923557925007,
+      "learning_rate": 3.558714268968344e-07,
+      "loss": 0.0242,
+      "step": 8025
+    },
+    {
+      "epoch": 3.5647346213635354,
+      "grad_norm": 0.40511430211397,
+      "learning_rate": 3.5515351528533024e-07,
+      "loss": 0.0217,
+      "step": 8026
+    },
+    {
+      "epoch": 3.5651787697090827,
+      "grad_norm": 0.4897536707816401,
+      "learning_rate": 3.5443630187315504e-07,
+      "loss": 0.0293,
+      "step": 8027
+    },
+    {
+      "epoch": 3.5656229180546304,
+      "grad_norm": 0.3947409838204125,
+      "learning_rate": 3.537197867681191e-07,
+      "loss": 0.0231,
+      "step": 8028
+    },
+    {
+      "epoch": 3.5660670664001777,
+      "grad_norm": 0.42936961733669526,
+      "learning_rate": 3.5300397007792364e-07,
+      "loss": 0.0266,
+      "step": 8029
+    },
+    {
+      "epoch": 3.566511214745725,
+      "grad_norm": 0.3453771987941547,
+      "learning_rate": 3.5228885191017084e-07,
+      "loss": 0.0181,
+      "step": 8030
+    },
+    {
+      "epoch": 3.5669553630912727,
+      "grad_norm": 0.38903220671574995,
+      "learning_rate": 3.515744323723558e-07,
+      "loss": 0.0254,
+      "step": 8031
+    },
+    {
+      "epoch": 3.56739951143682,
+      "grad_norm": 0.39781787732509344,
+      "learning_rate": 3.508607115718654e-07,
+      "loss": 0.0181,
+      "step": 8032
+    },
+    {
+      "epoch": 3.5678436597823673,
+      "grad_norm": 0.42441099532371085,
+      "learning_rate": 3.50147689615985e-07,
+      "loss": 0.0222,
+      "step": 8033
+    },
+    {
+      "epoch": 3.5682878081279146,
+      "grad_norm": 0.38842537901922425,
+      "learning_rate": 3.494353666118938e-07,
+      "loss": 0.0378,
+      "step": 8034
+    },
+    {
+      "epoch": 3.568731956473462,
+      "grad_norm": 0.35803044170961684,
+      "learning_rate": 3.4872374266666674e-07,
+      "loss": 0.0241,
+      "step": 8035
+    },
+    {
+      "epoch": 3.5691761048190096,
+      "grad_norm": 0.41680398082335124,
+      "learning_rate": 3.4801281788727326e-07,
+      "loss": 0.0226,
+      "step": 8036
+    },
+    {
+      "epoch": 3.569620253164557,
+      "grad_norm": 0.4978684696151685,
+      "learning_rate": 3.4730259238057563e-07,
+      "loss": 0.0265,
+      "step": 8037
+    },
+    {
+      "epoch": 3.5700644015101046,
+      "grad_norm": 0.526050521323082,
+      "learning_rate": 3.46593066253334e-07,
+      "loss": 0.0241,
+      "step": 8038
+    },
+    {
+      "epoch": 3.570508549855652,
+      "grad_norm": 0.4100317457053481,
+      "learning_rate": 3.4588423961220306e-07,
+      "loss": 0.0244,
+      "step": 8039
+    },
+    {
+      "epoch": 3.570952698201199,
+      "grad_norm": 0.4007810456784438,
+      "learning_rate": 3.4517611256372875e-07,
+      "loss": 0.0207,
+      "step": 8040
+    },
+    {
+      "epoch": 3.5713968465467465,
+      "grad_norm": 0.44507883967406275,
+      "learning_rate": 3.444686852143575e-07,
+      "loss": 0.0203,
+      "step": 8041
+    },
+    {
+      "epoch": 3.571840994892294,
+      "grad_norm": 0.4110302975726491,
+      "learning_rate": 3.4376195767042706e-07,
+      "loss": 0.0196,
+      "step": 8042
+    },
+    {
+      "epoch": 3.5722851432378415,
+      "grad_norm": 0.42428375307760136,
+      "learning_rate": 3.4305593003816917e-07,
+      "loss": 0.0239,
+      "step": 8043
+    },
+    {
+      "epoch": 3.572729291583389,
+      "grad_norm": 0.4068665526914318,
+      "learning_rate": 3.423506024237122e-07,
+      "loss": 0.0251,
+      "step": 8044
+    },
+    {
+      "epoch": 3.573173439928936,
+      "grad_norm": 0.5621905349071596,
+      "learning_rate": 3.416459749330808e-07,
+      "loss": 0.0436,
+      "step": 8045
+    },
+    {
+      "epoch": 3.573617588274484,
+      "grad_norm": 0.4013053666322752,
+      "learning_rate": 3.409420476721892e-07,
+      "loss": 0.0233,
+      "step": 8046
+    },
+    {
+      "epoch": 3.574061736620031,
+      "grad_norm": 0.40409509056363074,
+      "learning_rate": 3.4023882074685266e-07,
+      "loss": 0.0196,
+      "step": 8047
+    },
+    {
+      "epoch": 3.5745058849655784,
+      "grad_norm": 0.401583653285129,
+      "learning_rate": 3.3953629426277666e-07,
+      "loss": 0.0254,
+      "step": 8048
+    },
+    {
+      "epoch": 3.5749500333111257,
+      "grad_norm": 0.39247780188844755,
+      "learning_rate": 3.3883446832556286e-07,
+      "loss": 0.0226,
+      "step": 8049
+    },
+    {
+      "epoch": 3.5753941816566734,
+      "grad_norm": 0.4595707850818669,
+      "learning_rate": 3.381333430407074e-07,
+      "loss": 0.0331,
+      "step": 8050
+    },
+    {
+      "epoch": 3.5758383300022207,
+      "grad_norm": 0.38262273059202667,
+      "learning_rate": 3.3743291851360215e-07,
+      "loss": 0.032,
+      "step": 8051
+    },
+    {
+      "epoch": 3.576282478347768,
+      "grad_norm": 0.5184355648640612,
+      "learning_rate": 3.3673319484953224e-07,
+      "loss": 0.0535,
+      "step": 8052
+    },
+    {
+      "epoch": 3.5767266266933158,
+      "grad_norm": 0.4837920320850102,
+      "learning_rate": 3.3603417215367916e-07,
+      "loss": 0.028,
+      "step": 8053
+    },
+    {
+      "epoch": 3.577170775038863,
+      "grad_norm": 0.4493839583313447,
+      "learning_rate": 3.3533585053111604e-07,
+      "loss": 0.0239,
+      "step": 8054
+    },
+    {
+      "epoch": 3.5776149233844103,
+      "grad_norm": 0.45676379041549287,
+      "learning_rate": 3.346382300868134e-07,
+      "loss": 0.0183,
+      "step": 8055
+    },
+    {
+      "epoch": 3.5780590717299576,
+      "grad_norm": 0.42408641826733806,
+      "learning_rate": 3.339413109256362e-07,
+      "loss": 0.019,
+      "step": 8056
+    },
+    {
+      "epoch": 3.5785032200755054,
+      "grad_norm": 0.41032911192121174,
+      "learning_rate": 3.3324509315234066e-07,
+      "loss": 0.0243,
+      "step": 8057
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.3781126839349984,
+      "learning_rate": 3.325495768715831e-07,
+      "loss": 0.023,
+      "step": 8058
+    },
+    {
+      "epoch": 3.5793915167666,
+      "grad_norm": 0.48422608200816336,
+      "learning_rate": 3.318547621879109e-07,
+      "loss": 0.0209,
+      "step": 8059
+    },
+    {
+      "epoch": 3.5798356651121477,
+      "grad_norm": 0.5803863155007879,
+      "learning_rate": 3.311606492057651e-07,
+      "loss": 0.03,
+      "step": 8060
+    },
+    {
+      "epoch": 3.580279813457695,
+      "grad_norm": 0.4295397381772821,
+      "learning_rate": 3.304672380294832e-07,
+      "loss": 0.0276,
+      "step": 8061
+    },
+    {
+      "epoch": 3.5807239618032423,
+      "grad_norm": 0.4359738260640305,
+      "learning_rate": 3.2977452876329806e-07,
+      "loss": 0.0226,
+      "step": 8062
+    },
+    {
+      "epoch": 3.5811681101487896,
+      "grad_norm": 0.6578716400312555,
+      "learning_rate": 3.290825215113325e-07,
+      "loss": 0.0317,
+      "step": 8063
+    },
+    {
+      "epoch": 3.581612258494337,
+      "grad_norm": 0.31354942464141744,
+      "learning_rate": 3.2839121637761095e-07,
+      "loss": 0.021,
+      "step": 8064
+    },
+    {
+      "epoch": 3.5820564068398846,
+      "grad_norm": 0.3962803838449033,
+      "learning_rate": 3.277006134660454e-07,
+      "loss": 0.0279,
+      "step": 8065
+    },
+    {
+      "epoch": 3.582500555185432,
+      "grad_norm": 0.4214624266956511,
+      "learning_rate": 3.270107128804462e-07,
+      "loss": 0.0344,
+      "step": 8066
+    },
+    {
+      "epoch": 3.5829447035309796,
+      "grad_norm": 0.3602571631127595,
+      "learning_rate": 3.26321514724518e-07,
+      "loss": 0.0228,
+      "step": 8067
+    },
+    {
+      "epoch": 3.583388851876527,
+      "grad_norm": 0.37827106337184596,
+      "learning_rate": 3.2563301910185585e-07,
+      "loss": 0.0229,
+      "step": 8068
+    },
+    {
+      "epoch": 3.583833000222074,
+      "grad_norm": 0.3969145699088532,
+      "learning_rate": 3.249452261159558e-07,
+      "loss": 0.0216,
+      "step": 8069
+    },
+    {
+      "epoch": 3.5842771485676215,
+      "grad_norm": 0.3539428309291324,
+      "learning_rate": 3.242581358702046e-07,
+      "loss": 0.0219,
+      "step": 8070
+    },
+    {
+      "epoch": 3.5847212969131688,
+      "grad_norm": 0.46162564562225794,
+      "learning_rate": 3.235717484678813e-07,
+      "loss": 0.0219,
+      "step": 8071
+    },
+    {
+      "epoch": 3.5851654452587165,
+      "grad_norm": 0.395257543151881,
+      "learning_rate": 3.2288606401216283e-07,
+      "loss": 0.0283,
+      "step": 8072
+    },
+    {
+      "epoch": 3.585609593604264,
+      "grad_norm": 0.3543800684377955,
+      "learning_rate": 3.2220108260612e-07,
+      "loss": 0.022,
+      "step": 8073
+    },
+    {
+      "epoch": 3.586053741949811,
+      "grad_norm": 0.3956501467663245,
+      "learning_rate": 3.2151680435271504e-07,
+      "loss": 0.0227,
+      "step": 8074
+    },
+    {
+      "epoch": 3.586497890295359,
+      "grad_norm": 0.6908876097601867,
+      "learning_rate": 3.208332293548094e-07,
+      "loss": 0.0402,
+      "step": 8075
+    },
+    {
+      "epoch": 3.586942038640906,
+      "grad_norm": 0.41210563747436485,
+      "learning_rate": 3.2015035771515377e-07,
+      "loss": 0.0207,
+      "step": 8076
+    },
+    {
+      "epoch": 3.5873861869864534,
+      "grad_norm": 0.4210761069284605,
+      "learning_rate": 3.1946818953639604e-07,
+      "loss": 0.0199,
+      "step": 8077
+    },
+    {
+      "epoch": 3.5878303353320007,
+      "grad_norm": 0.562680504099952,
+      "learning_rate": 3.1878672492107796e-07,
+      "loss": 0.0333,
+      "step": 8078
+    },
+    {
+      "epoch": 3.5882744836775484,
+      "grad_norm": 0.45181012070367105,
+      "learning_rate": 3.181059639716355e-07,
+      "loss": 0.0292,
+      "step": 8079
+    },
+    {
+      "epoch": 3.5887186320230957,
+      "grad_norm": 0.3560150149469517,
+      "learning_rate": 3.1742590679039675e-07,
+      "loss": 0.0263,
+      "step": 8080
+    },
+    {
+      "epoch": 3.589162780368643,
+      "grad_norm": 0.2727495861254254,
+      "learning_rate": 3.167465534795888e-07,
+      "loss": 0.0142,
+      "step": 8081
+    },
+    {
+      "epoch": 3.5896069287141907,
+      "grad_norm": 0.3723555834846713,
+      "learning_rate": 3.1606790414132784e-07,
+      "loss": 0.0204,
+      "step": 8082
+    },
+    {
+      "epoch": 3.590051077059738,
+      "grad_norm": 0.3414027162319298,
+      "learning_rate": 3.153899588776266e-07,
+      "loss": 0.0229,
+      "step": 8083
+    },
+    {
+      "epoch": 3.5904952254052853,
+      "grad_norm": 0.42338168106939433,
+      "learning_rate": 3.147127177903936e-07,
+      "loss": 0.0232,
+      "step": 8084
+    },
+    {
+      "epoch": 3.5909393737508326,
+      "grad_norm": 0.41145971331036213,
+      "learning_rate": 3.1403618098142683e-07,
+      "loss": 0.0196,
+      "step": 8085
+    },
+    {
+      "epoch": 3.5913835220963803,
+      "grad_norm": 0.4400619534919001,
+      "learning_rate": 3.133603485524217e-07,
+      "loss": 0.0305,
+      "step": 8086
+    },
+    {
+      "epoch": 3.5918276704419276,
+      "grad_norm": 0.3445253112185815,
+      "learning_rate": 3.126852206049702e-07,
+      "loss": 0.0184,
+      "step": 8087
+    },
+    {
+      "epoch": 3.592271818787475,
+      "grad_norm": 0.4780681637104892,
+      "learning_rate": 3.1201079724055284e-07,
+      "loss": 0.0217,
+      "step": 8088
+    },
+    {
+      "epoch": 3.5927159671330227,
+      "grad_norm": 0.3835860710733381,
+      "learning_rate": 3.113370785605474e-07,
+      "loss": 0.0177,
+      "step": 8089
+    },
+    {
+      "epoch": 3.59316011547857,
+      "grad_norm": 0.36022249138379353,
+      "learning_rate": 3.106640646662268e-07,
+      "loss": 0.0229,
+      "step": 8090
+    },
+    {
+      "epoch": 3.5936042638241172,
+      "grad_norm": 0.3513194805512465,
+      "learning_rate": 3.099917556587534e-07,
+      "loss": 0.0183,
+      "step": 8091
+    },
+    {
+      "epoch": 3.5940484121696645,
+      "grad_norm": 0.44275629565228397,
+      "learning_rate": 3.0932015163918973e-07,
+      "loss": 0.0295,
+      "step": 8092
+    },
+    {
+      "epoch": 3.594492560515212,
+      "grad_norm": 0.4202067622758764,
+      "learning_rate": 3.0864925270848725e-07,
+      "loss": 0.022,
+      "step": 8093
+    },
+    {
+      "epoch": 3.5949367088607596,
+      "grad_norm": 0.35964770608777313,
+      "learning_rate": 3.079790589674947e-07,
+      "loss": 0.019,
+      "step": 8094
+    },
+    {
+      "epoch": 3.595380857206307,
+      "grad_norm": 0.33143293773682697,
+      "learning_rate": 3.073095705169532e-07,
+      "loss": 0.0174,
+      "step": 8095
+    },
+    {
+      "epoch": 3.5958250055518546,
+      "grad_norm": 0.35370082577225564,
+      "learning_rate": 3.066407874574978e-07,
+      "loss": 0.0245,
+      "step": 8096
+    },
+    {
+      "epoch": 3.596269153897402,
+      "grad_norm": 0.46712833979581136,
+      "learning_rate": 3.05972709889657e-07,
+      "loss": 0.0248,
+      "step": 8097
+    },
+    {
+      "epoch": 3.596713302242949,
+      "grad_norm": 0.38025100976241183,
+      "learning_rate": 3.0530533791385765e-07,
+      "loss": 0.0244,
+      "step": 8098
+    },
+    {
+      "epoch": 3.5971574505884965,
+      "grad_norm": 0.342220120184616,
+      "learning_rate": 3.0463867163041396e-07,
+      "loss": 0.0155,
+      "step": 8099
+    },
+    {
+      "epoch": 3.5976015989340437,
+      "grad_norm": 0.36933627653900136,
+      "learning_rate": 3.0397271113953796e-07,
+      "loss": 0.0209,
+      "step": 8100
+    },
+    {
+      "epoch": 3.5980457472795915,
+      "grad_norm": 0.3882041575523787,
+      "learning_rate": 3.0330745654133576e-07,
+      "loss": 0.0289,
+      "step": 8101
+    },
+    {
+      "epoch": 3.5984898956251388,
+      "grad_norm": 0.39071440323579304,
+      "learning_rate": 3.026429079358051e-07,
+      "loss": 0.0225,
+      "step": 8102
+    },
+    {
+      "epoch": 3.598934043970686,
+      "grad_norm": 0.5224321778667499,
+      "learning_rate": 3.0197906542283996e-07,
+      "loss": 0.0366,
+      "step": 8103
+    },
+    {
+      "epoch": 3.599378192316234,
+      "grad_norm": 0.463268142143758,
+      "learning_rate": 3.013159291022261e-07,
+      "loss": 0.0254,
+      "step": 8104
+    },
+    {
+      "epoch": 3.599822340661781,
+      "grad_norm": 0.4758792312691063,
+      "learning_rate": 3.006534990736448e-07,
+      "loss": 0.0177,
+      "step": 8105
+    },
+    {
+      "epoch": 3.6002664890073284,
+      "grad_norm": 0.3801891704764608,
+      "learning_rate": 2.99991775436671e-07,
+      "loss": 0.0285,
+      "step": 8106
+    },
+    {
+      "epoch": 3.6007106373528757,
+      "grad_norm": 0.47832635752425606,
+      "learning_rate": 2.993307582907728e-07,
+      "loss": 0.0393,
+      "step": 8107
+    },
+    {
+      "epoch": 3.6011547856984234,
+      "grad_norm": 0.4174887456808177,
+      "learning_rate": 2.9867044773531083e-07,
+      "loss": 0.0304,
+      "step": 8108
+    },
+    {
+      "epoch": 3.6015989340439707,
+      "grad_norm": 0.4160783686882648,
+      "learning_rate": 2.9801084386954337e-07,
+      "loss": 0.0224,
+      "step": 8109
+    },
+    {
+      "epoch": 3.602043082389518,
+      "grad_norm": 0.3728837062303793,
+      "learning_rate": 2.9735194679261835e-07,
+      "loss": 0.0212,
+      "step": 8110
+    },
+    {
+      "epoch": 3.6024872307350657,
+      "grad_norm": 0.7022594125847725,
+      "learning_rate": 2.966937566035799e-07,
+      "loss": 0.0301,
+      "step": 8111
+    },
+    {
+      "epoch": 3.602931379080613,
+      "grad_norm": 0.4052906698509884,
+      "learning_rate": 2.9603627340136553e-07,
+      "loss": 0.0293,
+      "step": 8112
+    },
+    {
+      "epoch": 3.6033755274261603,
+      "grad_norm": 0.3921299918104816,
+      "learning_rate": 2.953794972848051e-07,
+      "loss": 0.025,
+      "step": 8113
+    },
+    {
+      "epoch": 3.6038196757717076,
+      "grad_norm": 0.406918068376974,
+      "learning_rate": 2.947234283526229e-07,
+      "loss": 0.0278,
+      "step": 8114
+    },
+    {
+      "epoch": 3.6042638241172553,
+      "grad_norm": 0.4175532634627423,
+      "learning_rate": 2.940680667034396e-07,
+      "loss": 0.0318,
+      "step": 8115
+    },
+    {
+      "epoch": 3.6047079724628026,
+      "grad_norm": 0.43540375334253767,
+      "learning_rate": 2.934134124357646e-07,
+      "loss": 0.0268,
+      "step": 8116
+    },
+    {
+      "epoch": 3.60515212080835,
+      "grad_norm": 0.4718399188563787,
+      "learning_rate": 2.927594656480054e-07,
+      "loss": 0.0319,
+      "step": 8117
+    },
+    {
+      "epoch": 3.6055962691538976,
+      "grad_norm": 0.4746986163908478,
+      "learning_rate": 2.921062264384605e-07,
+      "loss": 0.0239,
+      "step": 8118
+    },
+    {
+      "epoch": 3.606040417499445,
+      "grad_norm": 0.43178883447784155,
+      "learning_rate": 2.914536949053226e-07,
+      "loss": 0.0252,
+      "step": 8119
+    },
+    {
+      "epoch": 3.606484565844992,
+      "grad_norm": 0.44193392412448756,
+      "learning_rate": 2.908018711466787e-07,
+      "loss": 0.0293,
+      "step": 8120
+    },
+    {
+      "epoch": 3.6069287141905395,
+      "grad_norm": 0.361976350777308,
+      "learning_rate": 2.901507552605087e-07,
+      "loss": 0.0155,
+      "step": 8121
+    },
+    {
+      "epoch": 3.607372862536087,
+      "grad_norm": 0.4277695581236822,
+      "learning_rate": 2.895003473446861e-07,
+      "loss": 0.0286,
+      "step": 8122
+    },
+    {
+      "epoch": 3.6078170108816345,
+      "grad_norm": 0.4459278029555017,
+      "learning_rate": 2.8885064749697987e-07,
+      "loss": 0.028,
+      "step": 8123
+    },
+    {
+      "epoch": 3.608261159227182,
+      "grad_norm": 0.566646999620044,
+      "learning_rate": 2.882016558150491e-07,
+      "loss": 0.0364,
+      "step": 8124
+    },
+    {
+      "epoch": 3.6087053075727296,
+      "grad_norm": 0.4313671970565694,
+      "learning_rate": 2.87553372396448e-07,
+      "loss": 0.031,
+      "step": 8125
+    },
+    {
+      "epoch": 3.609149455918277,
+      "grad_norm": 0.40292145142350155,
+      "learning_rate": 2.869057973386269e-07,
+      "loss": 0.0272,
+      "step": 8126
+    },
+    {
+      "epoch": 3.609593604263824,
+      "grad_norm": 0.41179464266273536,
+      "learning_rate": 2.8625893073892577e-07,
+      "loss": 0.0208,
+      "step": 8127
+    },
+    {
+      "epoch": 3.6100377526093714,
+      "grad_norm": 0.4344939010412826,
+      "learning_rate": 2.85612772694579e-07,
+      "loss": 0.0246,
+      "step": 8128
+    },
+    {
+      "epoch": 3.6104819009549187,
+      "grad_norm": 0.4503615177393324,
+      "learning_rate": 2.8496732330271726e-07,
+      "loss": 0.0318,
+      "step": 8129
+    },
+    {
+      "epoch": 3.6109260493004665,
+      "grad_norm": 0.49743560082851446,
+      "learning_rate": 2.8432258266036016e-07,
+      "loss": 0.0242,
+      "step": 8130
+    },
+    {
+      "epoch": 3.6113701976460137,
+      "grad_norm": 0.418557316697792,
+      "learning_rate": 2.8367855086442353e-07,
+      "loss": 0.0279,
+      "step": 8131
+    },
+    {
+      "epoch": 3.611814345991561,
+      "grad_norm": 0.3450331986237824,
+      "learning_rate": 2.830352280117188e-07,
+      "loss": 0.022,
+      "step": 8132
+    },
+    {
+      "epoch": 3.6122584943371088,
+      "grad_norm": 0.44492026164748794,
+      "learning_rate": 2.8239261419894526e-07,
+      "loss": 0.0323,
+      "step": 8133
+    },
+    {
+      "epoch": 3.612702642682656,
+      "grad_norm": 0.39841298833512895,
+      "learning_rate": 2.8175070952270014e-07,
+      "loss": 0.0285,
+      "step": 8134
+    },
+    {
+      "epoch": 3.6131467910282034,
+      "grad_norm": 0.43301418955026805,
+      "learning_rate": 2.811095140794734e-07,
+      "loss": 0.0274,
+      "step": 8135
+    },
+    {
+      "epoch": 3.6135909393737506,
+      "grad_norm": 0.5096476202086357,
+      "learning_rate": 2.804690279656458e-07,
+      "loss": 0.051,
+      "step": 8136
+    },
+    {
+      "epoch": 3.6140350877192984,
+      "grad_norm": 0.3293367472926153,
+      "learning_rate": 2.7982925127749416e-07,
+      "loss": 0.0227,
+      "step": 8137
+    },
+    {
+      "epoch": 3.6144792360648457,
+      "grad_norm": 0.37776627801259316,
+      "learning_rate": 2.791901841111877e-07,
+      "loss": 0.0254,
+      "step": 8138
+    },
+    {
+      "epoch": 3.614923384410393,
+      "grad_norm": 0.40922593503138716,
+      "learning_rate": 2.78551826562789e-07,
+      "loss": 0.045,
+      "step": 8139
+    },
+    {
+      "epoch": 3.6153675327559407,
+      "grad_norm": 0.42201659981374173,
+      "learning_rate": 2.779141787282547e-07,
+      "loss": 0.0251,
+      "step": 8140
+    },
+    {
+      "epoch": 3.615811681101488,
+      "grad_norm": 0.4053495089948873,
+      "learning_rate": 2.7727724070343296e-07,
+      "loss": 0.023,
+      "step": 8141
+    },
+    {
+      "epoch": 3.6162558294470353,
+      "grad_norm": 0.4556874573880963,
+      "learning_rate": 2.7664101258406626e-07,
+      "loss": 0.0227,
+      "step": 8142
+    },
+    {
+      "epoch": 3.6166999777925826,
+      "grad_norm": 0.3450822468016936,
+      "learning_rate": 2.7600549446579306e-07,
+      "loss": 0.018,
+      "step": 8143
+    },
+    {
+      "epoch": 3.6171441261381303,
+      "grad_norm": 0.35001380471776244,
+      "learning_rate": 2.753706864441391e-07,
+      "loss": 0.0232,
+      "step": 8144
+    },
+    {
+      "epoch": 3.6175882744836776,
+      "grad_norm": 0.364015851517707,
+      "learning_rate": 2.7473658861452923e-07,
+      "loss": 0.0185,
+      "step": 8145
+    },
+    {
+      "epoch": 3.618032422829225,
+      "grad_norm": 0.4713893596705623,
+      "learning_rate": 2.741032010722788e-07,
+      "loss": 0.0337,
+      "step": 8146
+    },
+    {
+      "epoch": 3.6184765711747726,
+      "grad_norm": 0.3440041200074425,
+      "learning_rate": 2.734705239125951e-07,
+      "loss": 0.0197,
+      "step": 8147
+    },
+    {
+      "epoch": 3.61892071952032,
+      "grad_norm": 0.41811195093373094,
+      "learning_rate": 2.728385572305814e-07,
+      "loss": 0.0278,
+      "step": 8148
+    },
+    {
+      "epoch": 3.619364867865867,
+      "grad_norm": 0.339333589639759,
+      "learning_rate": 2.7220730112123337e-07,
+      "loss": 0.0187,
+      "step": 8149
+    },
+    {
+      "epoch": 3.6198090162114145,
+      "grad_norm": 0.7462572452071613,
+      "learning_rate": 2.715767556794391e-07,
+      "loss": 0.0425,
+      "step": 8150
+    },
+    {
+      "epoch": 3.620253164556962,
+      "grad_norm": 0.5298727144349012,
+      "learning_rate": 2.7094692099997986e-07,
+      "loss": 0.0308,
+      "step": 8151
+    },
+    {
+      "epoch": 3.6206973129025095,
+      "grad_norm": 0.411102053477688,
+      "learning_rate": 2.7031779717753223e-07,
+      "loss": 0.0249,
+      "step": 8152
+    },
+    {
+      "epoch": 3.621141461248057,
+      "grad_norm": 0.4354411151955604,
+      "learning_rate": 2.696893843066617e-07,
+      "loss": 0.0307,
+      "step": 8153
+    },
+    {
+      "epoch": 3.6215856095936045,
+      "grad_norm": 0.3792115584655087,
+      "learning_rate": 2.6906168248183095e-07,
+      "loss": 0.0248,
+      "step": 8154
+    },
+    {
+      "epoch": 3.622029757939152,
+      "grad_norm": 0.3457882220586487,
+      "learning_rate": 2.68434691797394e-07,
+      "loss": 0.0197,
+      "step": 8155
+    },
+    {
+      "epoch": 3.622473906284699,
+      "grad_norm": 0.5361251704803487,
+      "learning_rate": 2.6780841234759826e-07,
+      "loss": 0.0382,
+      "step": 8156
+    },
+    {
+      "epoch": 3.6229180546302464,
+      "grad_norm": 0.45121861235515354,
+      "learning_rate": 2.6718284422658447e-07,
+      "loss": 0.0226,
+      "step": 8157
+    },
+    {
+      "epoch": 3.6233622029757937,
+      "grad_norm": 0.3726443210546779,
+      "learning_rate": 2.665579875283847e-07,
+      "loss": 0.0223,
+      "step": 8158
+    },
+    {
+      "epoch": 3.6238063513213414,
+      "grad_norm": 0.39308466518304386,
+      "learning_rate": 2.6593384234692597e-07,
+      "loss": 0.0259,
+      "step": 8159
+    },
+    {
+      "epoch": 3.6242504996668887,
+      "grad_norm": 0.3787552610454571,
+      "learning_rate": 2.6531040877602997e-07,
+      "loss": 0.0194,
+      "step": 8160
+    },
+    {
+      "epoch": 3.624694648012436,
+      "grad_norm": 0.41262964342629177,
+      "learning_rate": 2.646876869094073e-07,
+      "loss": 0.0197,
+      "step": 8161
+    },
+    {
+      "epoch": 3.6251387963579838,
+      "grad_norm": 0.34296275500584317,
+      "learning_rate": 2.640656768406641e-07,
+      "loss": 0.0215,
+      "step": 8162
+    },
+    {
+      "epoch": 3.625582944703531,
+      "grad_norm": 0.37447434249834716,
+      "learning_rate": 2.634443786632995e-07,
+      "loss": 0.0251,
+      "step": 8163
+    },
+    {
+      "epoch": 3.6260270930490783,
+      "grad_norm": 0.44671265535557425,
+      "learning_rate": 2.628237924707044e-07,
+      "loss": 0.0251,
+      "step": 8164
+    },
+    {
+      "epoch": 3.6264712413946256,
+      "grad_norm": 0.39092522082367537,
+      "learning_rate": 2.622039183561642e-07,
+      "loss": 0.0245,
+      "step": 8165
+    },
+    {
+      "epoch": 3.6269153897401734,
+      "grad_norm": 0.3971038919445469,
+      "learning_rate": 2.6158475641285544e-07,
+      "loss": 0.0297,
+      "step": 8166
+    },
+    {
+      "epoch": 3.6273595380857206,
+      "grad_norm": 0.7481153045495028,
+      "learning_rate": 2.609663067338497e-07,
+      "loss": 0.0379,
+      "step": 8167
+    },
+    {
+      "epoch": 3.627803686431268,
+      "grad_norm": 0.3705770719942717,
+      "learning_rate": 2.6034856941211104e-07,
+      "loss": 0.0238,
+      "step": 8168
+    },
+    {
+      "epoch": 3.6282478347768157,
+      "grad_norm": 0.42989268956281285,
+      "learning_rate": 2.597315445404941e-07,
+      "loss": 0.0264,
+      "step": 8169
+    },
+    {
+      "epoch": 3.628691983122363,
+      "grad_norm": 0.3871931353713456,
+      "learning_rate": 2.5911523221174963e-07,
+      "loss": 0.024,
+      "step": 8170
+    },
+    {
+      "epoch": 3.6291361314679103,
+      "grad_norm": 0.5499381201238311,
+      "learning_rate": 2.584996325185185e-07,
+      "loss": 0.036,
+      "step": 8171
+    },
+    {
+      "epoch": 3.6295802798134575,
+      "grad_norm": 0.42718459557037247,
+      "learning_rate": 2.5788474555333675e-07,
+      "loss": 0.0284,
+      "step": 8172
+    },
+    {
+      "epoch": 3.630024428159005,
+      "grad_norm": 0.45150263461678425,
+      "learning_rate": 2.5727057140863266e-07,
+      "loss": 0.0424,
+      "step": 8173
+    },
+    {
+      "epoch": 3.6304685765045526,
+      "grad_norm": 0.4604751454319313,
+      "learning_rate": 2.566571101767268e-07,
+      "loss": 0.0229,
+      "step": 8174
+    },
+    {
+      "epoch": 3.6309127248501,
+      "grad_norm": 0.40368285144270716,
+      "learning_rate": 2.5604436194983204e-07,
+      "loss": 0.0241,
+      "step": 8175
+    },
+    {
+      "epoch": 3.6313568731956476,
+      "grad_norm": 0.35454391047270123,
+      "learning_rate": 2.554323268200559e-07,
+      "loss": 0.0194,
+      "step": 8176
+    },
+    {
+      "epoch": 3.631801021541195,
+      "grad_norm": 0.4841769515138133,
+      "learning_rate": 2.548210048793964e-07,
+      "loss": 0.0253,
+      "step": 8177
+    },
+    {
+      "epoch": 3.632245169886742,
+      "grad_norm": 0.34838665066054253,
+      "learning_rate": 2.5421039621974677e-07,
+      "loss": 0.0219,
+      "step": 8178
+    },
+    {
+      "epoch": 3.6326893182322895,
+      "grad_norm": 0.40000850001012683,
+      "learning_rate": 2.5360050093289123e-07,
+      "loss": 0.0281,
+      "step": 8179
+    },
+    {
+      "epoch": 3.6331334665778368,
+      "grad_norm": 0.3910550154100924,
+      "learning_rate": 2.529913191105088e-07,
+      "loss": 0.0223,
+      "step": 8180
+    },
+    {
+      "epoch": 3.6335776149233845,
+      "grad_norm": 0.48803942739172496,
+      "learning_rate": 2.523828508441672e-07,
+      "loss": 0.025,
+      "step": 8181
+    },
+    {
+      "epoch": 3.634021763268932,
+      "grad_norm": 0.5754863723730677,
+      "learning_rate": 2.5177509622533183e-07,
+      "loss": 0.0406,
+      "step": 8182
+    },
+    {
+      "epoch": 3.634465911614479,
+      "grad_norm": 0.4286654639108965,
+      "learning_rate": 2.511680553453572e-07,
+      "loss": 0.0216,
+      "step": 8183
+    },
+    {
+      "epoch": 3.634910059960027,
+      "grad_norm": 0.35950293556792007,
+      "learning_rate": 2.5056172829549254e-07,
+      "loss": 0.0331,
+      "step": 8184
+    },
+    {
+      "epoch": 3.635354208305574,
+      "grad_norm": 0.40214970790890675,
+      "learning_rate": 2.4995611516688003e-07,
+      "loss": 0.0233,
+      "step": 8185
+    },
+    {
+      "epoch": 3.6357983566511214,
+      "grad_norm": 0.36548024593908557,
+      "learning_rate": 2.4935121605055125e-07,
+      "loss": 0.0261,
+      "step": 8186
+    },
+    {
+      "epoch": 3.6362425049966687,
+      "grad_norm": 0.5386358859158857,
+      "learning_rate": 2.487470310374346e-07,
+      "loss": 0.0215,
+      "step": 8187
+    },
+    {
+      "epoch": 3.6366866533422164,
+      "grad_norm": 0.38408368355582256,
+      "learning_rate": 2.481435602183485e-07,
+      "loss": 0.0227,
+      "step": 8188
+    },
+    {
+      "epoch": 3.6371308016877637,
+      "grad_norm": 0.3671311848219369,
+      "learning_rate": 2.475408036840055e-07,
+      "loss": 0.0203,
+      "step": 8189
+    },
+    {
+      "epoch": 3.637574950033311,
+      "grad_norm": 0.46572660180603004,
+      "learning_rate": 2.469387615250096e-07,
+      "loss": 0.0274,
+      "step": 8190
+    },
+    {
+      "epoch": 3.6380190983788587,
+      "grad_norm": 0.4911123345635677,
+      "learning_rate": 2.4633743383185917e-07,
+      "loss": 0.0252,
+      "step": 8191
+    },
+    {
+      "epoch": 3.638463246724406,
+      "grad_norm": 0.3574418971645557,
+      "learning_rate": 2.4573682069494234e-07,
+      "loss": 0.0205,
+      "step": 8192
+    },
+    {
+      "epoch": 3.6389073950699533,
+      "grad_norm": 0.4154556222921757,
+      "learning_rate": 2.451369222045419e-07,
+      "loss": 0.0248,
+      "step": 8193
+    },
+    {
+      "epoch": 3.6393515434155006,
+      "grad_norm": 0.3834951658389965,
+      "learning_rate": 2.445377384508335e-07,
+      "loss": 0.0228,
+      "step": 8194
+    },
+    {
+      "epoch": 3.6397956917610483,
+      "grad_norm": 0.3958357851929866,
+      "learning_rate": 2.4393926952388405e-07,
+      "loss": 0.0195,
+      "step": 8195
+    },
+    {
+      "epoch": 3.6402398401065956,
+      "grad_norm": 0.481708316236042,
+      "learning_rate": 2.433415155136543e-07,
+      "loss": 0.0245,
+      "step": 8196
+    },
+    {
+      "epoch": 3.640683988452143,
+      "grad_norm": 0.45974860157380204,
+      "learning_rate": 2.427444765099951e-07,
+      "loss": 0.0228,
+      "step": 8197
+    },
+    {
+      "epoch": 3.6411281367976907,
+      "grad_norm": 0.4731819261447121,
+      "learning_rate": 2.4214815260265367e-07,
+      "loss": 0.0498,
+      "step": 8198
+    },
+    {
+      "epoch": 3.641572285143238,
+      "grad_norm": 0.40140095766392536,
+      "learning_rate": 2.4155254388126605e-07,
+      "loss": 0.0309,
+      "step": 8199
+    },
+    {
+      "epoch": 3.6420164334887852,
+      "grad_norm": 0.34222283638473855,
+      "learning_rate": 2.4095765043536335e-07,
+      "loss": 0.0196,
+      "step": 8200
+    },
+    {
+      "epoch": 3.6424605818343325,
+      "grad_norm": 0.4008139911989502,
+      "learning_rate": 2.403634723543674e-07,
+      "loss": 0.0163,
+      "step": 8201
+    },
+    {
+      "epoch": 3.64290473017988,
+      "grad_norm": 0.4485823633640726,
+      "learning_rate": 2.3977000972759454e-07,
+      "loss": 0.0247,
+      "step": 8202
+    },
+    {
+      "epoch": 3.6433488785254275,
+      "grad_norm": 0.32021601023059565,
+      "learning_rate": 2.391772626442507e-07,
+      "loss": 0.0162,
+      "step": 8203
+    },
+    {
+      "epoch": 3.643793026870975,
+      "grad_norm": 0.43897160516812744,
+      "learning_rate": 2.385852311934367e-07,
+      "loss": 0.0192,
+      "step": 8204
+    },
+    {
+      "epoch": 3.6442371752165226,
+      "grad_norm": 0.35383635594441887,
+      "learning_rate": 2.379939154641442e-07,
+      "loss": 0.0224,
+      "step": 8205
+    },
+    {
+      "epoch": 3.64468132356207,
+      "grad_norm": 0.4656360521009951,
+      "learning_rate": 2.3740331554525875e-07,
+      "loss": 0.0281,
+      "step": 8206
+    },
+    {
+      "epoch": 3.645125471907617,
+      "grad_norm": 0.3752654330901412,
+      "learning_rate": 2.3681343152555768e-07,
+      "loss": 0.021,
+      "step": 8207
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.43380693965119627,
+      "learning_rate": 2.3622426349371064e-07,
+      "loss": 0.0233,
+      "step": 8208
+    },
+    {
+      "epoch": 3.6460137685987117,
+      "grad_norm": 0.3468026081865424,
+      "learning_rate": 2.3563581153827897e-07,
+      "loss": 0.0208,
+      "step": 8209
+    },
+    {
+      "epoch": 3.6464579169442595,
+      "grad_norm": 0.3963696177580075,
+      "learning_rate": 2.3504807574771638e-07,
+      "loss": 0.0234,
+      "step": 8210
+    },
+    {
+      "epoch": 3.6469020652898068,
+      "grad_norm": 0.3493278085805506,
+      "learning_rate": 2.3446105621037108e-07,
+      "loss": 0.018,
+      "step": 8211
+    },
+    {
+      "epoch": 3.647346213635354,
+      "grad_norm": 0.448389498368257,
+      "learning_rate": 2.3387475301448138e-07,
+      "loss": 0.0262,
+      "step": 8212
+    },
+    {
+      "epoch": 3.647790361980902,
+      "grad_norm": 0.4639744403941518,
+      "learning_rate": 2.33289166248179e-07,
+      "loss": 0.0299,
+      "step": 8213
+    },
+    {
+      "epoch": 3.648234510326449,
+      "grad_norm": 0.34846046299317557,
+      "learning_rate": 2.327042959994863e-07,
+      "loss": 0.0197,
+      "step": 8214
+    },
+    {
+      "epoch": 3.6486786586719964,
+      "grad_norm": 0.35001828631869614,
+      "learning_rate": 2.3212014235632074e-07,
+      "loss": 0.0184,
+      "step": 8215
+    },
+    {
+      "epoch": 3.6491228070175437,
+      "grad_norm": 0.3539191157147139,
+      "learning_rate": 2.3153670540648932e-07,
+      "loss": 0.0202,
+      "step": 8216
+    },
+    {
+      "epoch": 3.6495669553630914,
+      "grad_norm": 0.39166688802077126,
+      "learning_rate": 2.3095398523769353e-07,
+      "loss": 0.0237,
+      "step": 8217
+    },
+    {
+      "epoch": 3.6500111037086387,
+      "grad_norm": 0.41163896619179474,
+      "learning_rate": 2.3037198193752553e-07,
+      "loss": 0.0254,
+      "step": 8218
+    },
+    {
+      "epoch": 3.650455252054186,
+      "grad_norm": 0.43685004953046314,
+      "learning_rate": 2.2979069559347088e-07,
+      "loss": 0.0237,
+      "step": 8219
+    },
+    {
+      "epoch": 3.6508994003997337,
+      "grad_norm": 0.5309766077496497,
+      "learning_rate": 2.292101262929064e-07,
+      "loss": 0.0292,
+      "step": 8220
+    },
+    {
+      "epoch": 3.651343548745281,
+      "grad_norm": 0.41452507271756617,
+      "learning_rate": 2.2863027412310056e-07,
+      "loss": 0.0292,
+      "step": 8221
+    },
+    {
+      "epoch": 3.6517876970908283,
+      "grad_norm": 0.44136007566907587,
+      "learning_rate": 2.2805113917121647e-07,
+      "loss": 0.0284,
+      "step": 8222
+    },
+    {
+      "epoch": 3.6522318454363756,
+      "grad_norm": 0.3476979126592021,
+      "learning_rate": 2.274727215243072e-07,
+      "loss": 0.0197,
+      "step": 8223
+    },
+    {
+      "epoch": 3.6526759937819233,
+      "grad_norm": 0.38817929599514056,
+      "learning_rate": 2.2689502126931938e-07,
+      "loss": 0.0224,
+      "step": 8224
+    },
+    {
+      "epoch": 3.6531201421274706,
+      "grad_norm": 0.4746782142221425,
+      "learning_rate": 2.2631803849309076e-07,
+      "loss": 0.0352,
+      "step": 8225
+    },
+    {
+      "epoch": 3.653564290473018,
+      "grad_norm": 0.40968095773887653,
+      "learning_rate": 2.2574177328235137e-07,
+      "loss": 0.0276,
+      "step": 8226
+    },
+    {
+      "epoch": 3.6540084388185656,
+      "grad_norm": 0.41401682518721233,
+      "learning_rate": 2.2516622572372416e-07,
+      "loss": 0.0283,
+      "step": 8227
+    },
+    {
+      "epoch": 3.654452587164113,
+      "grad_norm": 0.4117127298389683,
+      "learning_rate": 2.2459139590372325e-07,
+      "loss": 0.0228,
+      "step": 8228
+    },
+    {
+      "epoch": 3.65489673550966,
+      "grad_norm": 0.4311069896164078,
+      "learning_rate": 2.240172839087551e-07,
+      "loss": 0.0281,
+      "step": 8229
+    },
+    {
+      "epoch": 3.6553408838552075,
+      "grad_norm": 0.4327836235259352,
+      "learning_rate": 2.2344388982512012e-07,
+      "loss": 0.0228,
+      "step": 8230
+    },
+    {
+      "epoch": 3.655785032200755,
+      "grad_norm": 0.4573606402349546,
+      "learning_rate": 2.2287121373900712e-07,
+      "loss": 0.0319,
+      "step": 8231
+    },
+    {
+      "epoch": 3.6562291805463025,
+      "grad_norm": 0.4345901205357684,
+      "learning_rate": 2.2229925573650001e-07,
+      "loss": 0.0269,
+      "step": 8232
+    },
+    {
+      "epoch": 3.65667332889185,
+      "grad_norm": 0.3825914316334593,
+      "learning_rate": 2.2172801590357395e-07,
+      "loss": 0.0282,
+      "step": 8233
+    },
+    {
+      "epoch": 3.6571174772373976,
+      "grad_norm": 0.31366037192320456,
+      "learning_rate": 2.2115749432609524e-07,
+      "loss": 0.0167,
+      "step": 8234
+    },
+    {
+      "epoch": 3.657561625582945,
+      "grad_norm": 0.3788275375636668,
+      "learning_rate": 2.205876910898236e-07,
+      "loss": 0.0306,
+      "step": 8235
+    },
+    {
+      "epoch": 3.658005773928492,
+      "grad_norm": 0.38419523963619157,
+      "learning_rate": 2.2001860628041106e-07,
+      "loss": 0.0299,
+      "step": 8236
+    },
+    {
+      "epoch": 3.6584499222740394,
+      "grad_norm": 0.4550088476883181,
+      "learning_rate": 2.1945023998339865e-07,
+      "loss": 0.0347,
+      "step": 8237
+    },
+    {
+      "epoch": 3.6588940706195867,
+      "grad_norm": 0.597248639500098,
+      "learning_rate": 2.1888259228422248e-07,
+      "loss": 0.0194,
+      "step": 8238
+    },
+    {
+      "epoch": 3.6593382189651344,
+      "grad_norm": 0.5188207400851287,
+      "learning_rate": 2.1831566326820986e-07,
+      "loss": 0.0276,
+      "step": 8239
+    },
+    {
+      "epoch": 3.6597823673106817,
+      "grad_norm": 0.32831002856910196,
+      "learning_rate": 2.177494530205798e-07,
+      "loss": 0.0245,
+      "step": 8240
+    },
+    {
+      "epoch": 3.660226515656229,
+      "grad_norm": 0.3836254147289484,
+      "learning_rate": 2.1718396162644319e-07,
+      "loss": 0.0212,
+      "step": 8241
+    },
+    {
+      "epoch": 3.6606706640017768,
+      "grad_norm": 0.4548723138202921,
+      "learning_rate": 2.1661918917080304e-07,
+      "loss": 0.0253,
+      "step": 8242
+    },
+    {
+      "epoch": 3.661114812347324,
+      "grad_norm": 0.3684605663304441,
+      "learning_rate": 2.1605513573855375e-07,
+      "loss": 0.0178,
+      "step": 8243
+    },
+    {
+      "epoch": 3.6615589606928713,
+      "grad_norm": 0.3733701839690256,
+      "learning_rate": 2.1549180141448356e-07,
+      "loss": 0.0194,
+      "step": 8244
+    },
+    {
+      "epoch": 3.6620031090384186,
+      "grad_norm": 0.4105795750947663,
+      "learning_rate": 2.1492918628326864e-07,
+      "loss": 0.0148,
+      "step": 8245
+    },
+    {
+      "epoch": 3.6624472573839664,
+      "grad_norm": 0.41761384153375,
+      "learning_rate": 2.143672904294819e-07,
+      "loss": 0.0293,
+      "step": 8246
+    },
+    {
+      "epoch": 3.6628914057295137,
+      "grad_norm": 0.3728653241577783,
+      "learning_rate": 2.1380611393758576e-07,
+      "loss": 0.0223,
+      "step": 8247
+    },
+    {
+      "epoch": 3.663335554075061,
+      "grad_norm": 0.3955291650313776,
+      "learning_rate": 2.1324565689193332e-07,
+      "loss": 0.025,
+      "step": 8248
+    },
+    {
+      "epoch": 3.6637797024206087,
+      "grad_norm": 0.5976258676447095,
+      "learning_rate": 2.1268591937677164e-07,
+      "loss": 0.0328,
+      "step": 8249
+    },
+    {
+      "epoch": 3.664223850766156,
+      "grad_norm": 0.7950146359972061,
+      "learning_rate": 2.1212690147623894e-07,
+      "loss": 0.0295,
+      "step": 8250
+    },
+    {
+      "epoch": 3.6646679991117033,
+      "grad_norm": 0.5530704513836777,
+      "learning_rate": 2.1156860327436302e-07,
+      "loss": 0.0302,
+      "step": 8251
+    },
+    {
+      "epoch": 3.6651121474572506,
+      "grad_norm": 0.29807912775738393,
+      "learning_rate": 2.1101102485506842e-07,
+      "loss": 0.0178,
+      "step": 8252
+    },
+    {
+      "epoch": 3.6655562958027983,
+      "grad_norm": 0.4517710106710957,
+      "learning_rate": 2.1045416630216808e-07,
+      "loss": 0.0242,
+      "step": 8253
+    },
+    {
+      "epoch": 3.6660004441483456,
+      "grad_norm": 0.4254157446074955,
+      "learning_rate": 2.0989802769936563e-07,
+      "loss": 0.0268,
+      "step": 8254
+    },
+    {
+      "epoch": 3.666444592493893,
+      "grad_norm": 0.3693301648290188,
+      "learning_rate": 2.0934260913025973e-07,
+      "loss": 0.0266,
+      "step": 8255
+    },
+    {
+      "epoch": 3.6668887408394406,
+      "grad_norm": 0.4631155691374193,
+      "learning_rate": 2.0878791067833805e-07,
+      "loss": 0.0293,
+      "step": 8256
+    },
+    {
+      "epoch": 3.667332889184988,
+      "grad_norm": 0.3868441750288522,
+      "learning_rate": 2.0823393242698275e-07,
+      "loss": 0.0261,
+      "step": 8257
+    },
+    {
+      "epoch": 3.667777037530535,
+      "grad_norm": 0.430361544922243,
+      "learning_rate": 2.0768067445946506e-07,
+      "loss": 0.0247,
+      "step": 8258
+    },
+    {
+      "epoch": 3.6682211858760825,
+      "grad_norm": 0.4116233802052011,
+      "learning_rate": 2.0712813685894894e-07,
+      "loss": 0.0315,
+      "step": 8259
+    },
+    {
+      "epoch": 3.6686653342216298,
+      "grad_norm": 0.3788192226900461,
+      "learning_rate": 2.0657631970849078e-07,
+      "loss": 0.0229,
+      "step": 8260
+    },
+    {
+      "epoch": 3.6691094825671775,
+      "grad_norm": 0.37557385002997257,
+      "learning_rate": 2.0602522309103813e-07,
+      "loss": 0.0236,
+      "step": 8261
+    },
+    {
+      "epoch": 3.669553630912725,
+      "grad_norm": 0.3805674438167674,
+      "learning_rate": 2.054748470894291e-07,
+      "loss": 0.0252,
+      "step": 8262
+    },
+    {
+      "epoch": 3.6699977792582725,
+      "grad_norm": 0.3690281782315337,
+      "learning_rate": 2.0492519178639536e-07,
+      "loss": 0.0161,
+      "step": 8263
+    },
+    {
+      "epoch": 3.67044192760382,
+      "grad_norm": 0.7116450161368049,
+      "learning_rate": 2.0437625726456024e-07,
+      "loss": 0.0292,
+      "step": 8264
+    },
+    {
+      "epoch": 3.670886075949367,
+      "grad_norm": 0.5939366820148858,
+      "learning_rate": 2.0382804360643603e-07,
+      "loss": 0.0382,
+      "step": 8265
+    },
+    {
+      "epoch": 3.6713302242949144,
+      "grad_norm": 0.4269858268062977,
+      "learning_rate": 2.0328055089443023e-07,
+      "loss": 0.0236,
+      "step": 8266
+    },
+    {
+      "epoch": 3.6717743726404617,
+      "grad_norm": 0.43163831702859096,
+      "learning_rate": 2.027337792108397e-07,
+      "loss": 0.0244,
+      "step": 8267
+    },
+    {
+      "epoch": 3.6722185209860094,
+      "grad_norm": 0.5025109334901467,
+      "learning_rate": 2.0218772863785263e-07,
+      "loss": 0.0296,
+      "step": 8268
+    },
+    {
+      "epoch": 3.6726626693315567,
+      "grad_norm": 0.6343391256263229,
+      "learning_rate": 2.016423992575517e-07,
+      "loss": 0.0248,
+      "step": 8269
+    },
+    {
+      "epoch": 3.673106817677104,
+      "grad_norm": 0.4000275239829487,
+      "learning_rate": 2.0109779115190742e-07,
+      "loss": 0.0198,
+      "step": 8270
+    },
+    {
+      "epoch": 3.6735509660226517,
+      "grad_norm": 0.3303791647606292,
+      "learning_rate": 2.0055390440278376e-07,
+      "loss": 0.0258,
+      "step": 8271
+    },
+    {
+      "epoch": 3.673995114368199,
+      "grad_norm": 0.36510639486980745,
+      "learning_rate": 2.0001073909193702e-07,
+      "loss": 0.0173,
+      "step": 8272
+    },
+    {
+      "epoch": 3.6744392627137463,
+      "grad_norm": 0.4445660217933993,
+      "learning_rate": 1.9946829530101408e-07,
+      "loss": 0.0279,
+      "step": 8273
+    },
+    {
+      "epoch": 3.6748834110592936,
+      "grad_norm": 0.3323295635321309,
+      "learning_rate": 1.989265731115525e-07,
+      "loss": 0.0211,
+      "step": 8274
+    },
+    {
+      "epoch": 3.6753275594048413,
+      "grad_norm": 0.4082377088884518,
+      "learning_rate": 1.983855726049838e-07,
+      "loss": 0.0281,
+      "step": 8275
+    },
+    {
+      "epoch": 3.6757717077503886,
+      "grad_norm": 0.5538219755733254,
+      "learning_rate": 1.9784529386262798e-07,
+      "loss": 0.0375,
+      "step": 8276
+    },
+    {
+      "epoch": 3.676215856095936,
+      "grad_norm": 0.4054313962604177,
+      "learning_rate": 1.9730573696569888e-07,
+      "loss": 0.0221,
+      "step": 8277
+    },
+    {
+      "epoch": 3.6766600044414837,
+      "grad_norm": 0.338502488617252,
+      "learning_rate": 1.9676690199530169e-07,
+      "loss": 0.0231,
+      "step": 8278
+    },
+    {
+      "epoch": 3.677104152787031,
+      "grad_norm": 0.4016411596672554,
+      "learning_rate": 1.9622878903243104e-07,
+      "loss": 0.024,
+      "step": 8279
+    },
+    {
+      "epoch": 3.6775483011325782,
+      "grad_norm": 0.4045075471165877,
+      "learning_rate": 1.95691398157975e-07,
+      "loss": 0.0179,
+      "step": 8280
+    },
+    {
+      "epoch": 3.6779924494781255,
+      "grad_norm": 0.40898342731485937,
+      "learning_rate": 1.9515472945271396e-07,
+      "loss": 0.0337,
+      "step": 8281
+    },
+    {
+      "epoch": 3.6784365978236733,
+      "grad_norm": 0.3742680249776215,
+      "learning_rate": 1.946187829973162e-07,
+      "loss": 0.0181,
+      "step": 8282
+    },
+    {
+      "epoch": 3.6788807461692206,
+      "grad_norm": 0.38742125913942227,
+      "learning_rate": 1.9408355887234443e-07,
+      "loss": 0.0267,
+      "step": 8283
+    },
+    {
+      "epoch": 3.679324894514768,
+      "grad_norm": 0.31803662853945275,
+      "learning_rate": 1.9354905715825323e-07,
+      "loss": 0.0106,
+      "step": 8284
+    },
+    {
+      "epoch": 3.6797690428603156,
+      "grad_norm": 0.3842318920408417,
+      "learning_rate": 1.9301527793538445e-07,
+      "loss": 0.0208,
+      "step": 8285
+    },
+    {
+      "epoch": 3.680213191205863,
+      "grad_norm": 0.34972059951342765,
+      "learning_rate": 1.9248222128397663e-07,
+      "loss": 0.0209,
+      "step": 8286
+    },
+    {
+      "epoch": 3.68065733955141,
+      "grad_norm": 0.4945754865360583,
+      "learning_rate": 1.9194988728415632e-07,
+      "loss": 0.0285,
+      "step": 8287
+    },
+    {
+      "epoch": 3.6811014878969575,
+      "grad_norm": 0.4665722830251576,
+      "learning_rate": 1.9141827601594221e-07,
+      "loss": 0.0228,
+      "step": 8288
+    },
+    {
+      "epoch": 3.6815456362425047,
+      "grad_norm": 0.33231118424653433,
+      "learning_rate": 1.908873875592454e-07,
+      "loss": 0.0215,
+      "step": 8289
+    },
+    {
+      "epoch": 3.6819897845880525,
+      "grad_norm": 0.41202362800291603,
+      "learning_rate": 1.9035722199386542e-07,
+      "loss": 0.0219,
+      "step": 8290
+    },
+    {
+      "epoch": 3.6824339329335998,
+      "grad_norm": 0.4209126047122881,
+      "learning_rate": 1.8982777939949736e-07,
+      "loss": 0.0356,
+      "step": 8291
+    },
+    {
+      "epoch": 3.6828780812791475,
+      "grad_norm": 0.35749851244891667,
+      "learning_rate": 1.8929905985572484e-07,
+      "loss": 0.0194,
+      "step": 8292
+    },
+    {
+      "epoch": 3.683322229624695,
+      "grad_norm": 0.38704831620698,
+      "learning_rate": 1.8877106344202312e-07,
+      "loss": 0.0329,
+      "step": 8293
+    },
+    {
+      "epoch": 3.683766377970242,
+      "grad_norm": 0.526720346124437,
+      "learning_rate": 1.8824379023775874e-07,
+      "loss": 0.0357,
+      "step": 8294
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.42113199922195677,
+      "learning_rate": 1.877172403221905e-07,
+      "loss": 0.0315,
+      "step": 8295
+    },
+    {
+      "epoch": 3.6846546746613367,
+      "grad_norm": 0.3078574734152821,
+      "learning_rate": 1.871914137744668e-07,
+      "loss": 0.0165,
+      "step": 8296
+    },
+    {
+      "epoch": 3.6850988230068844,
+      "grad_norm": 0.4123152188226272,
+      "learning_rate": 1.866663106736294e-07,
+      "loss": 0.0401,
+      "step": 8297
+    },
+    {
+      "epoch": 3.6855429713524317,
+      "grad_norm": 0.3318217390451302,
+      "learning_rate": 1.8614193109860955e-07,
+      "loss": 0.0228,
+      "step": 8298
+    },
+    {
+      "epoch": 3.685987119697979,
+      "grad_norm": 0.4377759250920508,
+      "learning_rate": 1.8561827512823095e-07,
+      "loss": 0.0271,
+      "step": 8299
+    },
+    {
+      "epoch": 3.6864312680435267,
+      "grad_norm": 0.8207219751826093,
+      "learning_rate": 1.8509534284120721e-07,
+      "loss": 0.0344,
+      "step": 8300
+    },
+    {
+      "epoch": 3.686875416389074,
+      "grad_norm": 0.8375758066775824,
+      "learning_rate": 1.84573134316145e-07,
+      "loss": 0.0367,
+      "step": 8301
+    },
+    {
+      "epoch": 3.6873195647346213,
+      "grad_norm": 0.36428841642403553,
+      "learning_rate": 1.840516496315392e-07,
+      "loss": 0.0217,
+      "step": 8302
+    },
+    {
+      "epoch": 3.6877637130801686,
+      "grad_norm": 0.5076188683241221,
+      "learning_rate": 1.8353088886578053e-07,
+      "loss": 0.0294,
+      "step": 8303
+    },
+    {
+      "epoch": 3.6882078614257163,
+      "grad_norm": 0.6798347292105588,
+      "learning_rate": 1.830108520971463e-07,
+      "loss": 0.0405,
+      "step": 8304
+    },
+    {
+      "epoch": 3.6886520097712636,
+      "grad_norm": 0.3568052667311966,
+      "learning_rate": 1.8249153940380738e-07,
+      "loss": 0.0279,
+      "step": 8305
+    },
+    {
+      "epoch": 3.689096158116811,
+      "grad_norm": 0.4023381862191607,
+      "learning_rate": 1.8197295086382515e-07,
+      "loss": 0.0271,
+      "step": 8306
+    },
+    {
+      "epoch": 3.6895403064623586,
+      "grad_norm": 0.3720480087573231,
+      "learning_rate": 1.8145508655515177e-07,
+      "loss": 0.0199,
+      "step": 8307
+    },
+    {
+      "epoch": 3.689984454807906,
+      "grad_norm": 0.4144873947838932,
+      "learning_rate": 1.8093794655563214e-07,
+      "loss": 0.0215,
+      "step": 8308
+    },
+    {
+      "epoch": 3.6904286031534532,
+      "grad_norm": 0.39511282629465616,
+      "learning_rate": 1.804215309430013e-07,
+      "loss": 0.0175,
+      "step": 8309
+    },
+    {
+      "epoch": 3.6908727514990005,
+      "grad_norm": 0.5527099360678772,
+      "learning_rate": 1.799058397948844e-07,
+      "loss": 0.0381,
+      "step": 8310
+    },
+    {
+      "epoch": 3.691316899844548,
+      "grad_norm": 0.46007077628363413,
+      "learning_rate": 1.7939087318879833e-07,
+      "loss": 0.0333,
+      "step": 8311
+    },
+    {
+      "epoch": 3.6917610481900955,
+      "grad_norm": 0.43813513461840425,
+      "learning_rate": 1.788766312021528e-07,
+      "loss": 0.0298,
+      "step": 8312
+    },
+    {
+      "epoch": 3.692205196535643,
+      "grad_norm": 0.4101350408780889,
+      "learning_rate": 1.7836311391224494e-07,
+      "loss": 0.0259,
+      "step": 8313
+    },
+    {
+      "epoch": 3.6926493448811906,
+      "grad_norm": 0.2992536206219798,
+      "learning_rate": 1.7785032139626734e-07,
+      "loss": 0.0185,
+      "step": 8314
+    },
+    {
+      "epoch": 3.693093493226738,
+      "grad_norm": 0.365121204268502,
+      "learning_rate": 1.7733825373129954e-07,
+      "loss": 0.0216,
+      "step": 8315
+    },
+    {
+      "epoch": 3.693537641572285,
+      "grad_norm": 0.416511886265595,
+      "learning_rate": 1.7682691099431548e-07,
+      "loss": 0.0372,
+      "step": 8316
+    },
+    {
+      "epoch": 3.6939817899178324,
+      "grad_norm": 0.3845252122577115,
+      "learning_rate": 1.763162932621787e-07,
+      "loss": 0.0257,
+      "step": 8317
+    },
+    {
+      "epoch": 3.6944259382633797,
+      "grad_norm": 0.4233868381729098,
+      "learning_rate": 1.7580640061164223e-07,
+      "loss": 0.0294,
+      "step": 8318
+    },
+    {
+      "epoch": 3.6948700866089275,
+      "grad_norm": 0.462463687883106,
+      "learning_rate": 1.7529723311935198e-07,
+      "loss": 0.0364,
+      "step": 8319
+    },
+    {
+      "epoch": 3.6953142349544748,
+      "grad_norm": 0.43549161369245626,
+      "learning_rate": 1.7478879086184564e-07,
+      "loss": 0.0263,
+      "step": 8320
+    },
+    {
+      "epoch": 3.695758383300022,
+      "grad_norm": 0.39333187210295867,
+      "learning_rate": 1.742810739155504e-07,
+      "loss": 0.0246,
+      "step": 8321
+    },
+    {
+      "epoch": 3.6962025316455698,
+      "grad_norm": 0.35014124457318635,
+      "learning_rate": 1.737740823567835e-07,
+      "loss": 0.0218,
+      "step": 8322
+    },
+    {
+      "epoch": 3.696646679991117,
+      "grad_norm": 0.42978622961517393,
+      "learning_rate": 1.7326781626175627e-07,
+      "loss": 0.0257,
+      "step": 8323
+    },
+    {
+      "epoch": 3.6970908283366644,
+      "grad_norm": 0.4118470726335261,
+      "learning_rate": 1.727622757065678e-07,
+      "loss": 0.0261,
+      "step": 8324
+    },
+    {
+      "epoch": 3.6975349766822116,
+      "grad_norm": 0.471980787223015,
+      "learning_rate": 1.7225746076720894e-07,
+      "loss": 0.0226,
+      "step": 8325
+    },
+    {
+      "epoch": 3.6979791250277594,
+      "grad_norm": 0.4565879663220365,
+      "learning_rate": 1.717533715195635e-07,
+      "loss": 0.0248,
+      "step": 8326
+    },
+    {
+      "epoch": 3.6984232733733067,
+      "grad_norm": 0.3976292965837061,
+      "learning_rate": 1.712500080394036e-07,
+      "loss": 0.0324,
+      "step": 8327
+    },
+    {
+      "epoch": 3.698867421718854,
+      "grad_norm": 0.4111827692476157,
+      "learning_rate": 1.7074737040239375e-07,
+      "loss": 0.0248,
+      "step": 8328
+    },
+    {
+      "epoch": 3.6993115700644017,
+      "grad_norm": 0.36585319212766676,
+      "learning_rate": 1.7024545868408903e-07,
+      "loss": 0.0209,
+      "step": 8329
+    },
+    {
+      "epoch": 3.699755718409949,
+      "grad_norm": 0.47005920699211934,
+      "learning_rate": 1.6974427295993412e-07,
+      "loss": 0.0203,
+      "step": 8330
+    },
+    {
+      "epoch": 3.7001998667554963,
+      "grad_norm": 0.36932538349452104,
+      "learning_rate": 1.6924381330526817e-07,
+      "loss": 0.0212,
+      "step": 8331
+    },
+    {
+      "epoch": 3.7006440151010436,
+      "grad_norm": 0.446723580160887,
+      "learning_rate": 1.6874407979531604e-07,
+      "loss": 0.0247,
+      "step": 8332
+    },
+    {
+      "epoch": 3.7010881634465913,
+      "grad_norm": 0.445687716290222,
+      "learning_rate": 1.682450725051976e-07,
+      "loss": 0.0251,
+      "step": 8333
+    },
+    {
+      "epoch": 3.7015323117921386,
+      "grad_norm": 0.3855963805554051,
+      "learning_rate": 1.677467915099229e-07,
+      "loss": 0.0161,
+      "step": 8334
+    },
+    {
+      "epoch": 3.701976460137686,
+      "grad_norm": 0.39452015734075696,
+      "learning_rate": 1.6724923688439033e-07,
+      "loss": 0.0238,
+      "step": 8335
+    },
+    {
+      "epoch": 3.7024206084832336,
+      "grad_norm": 0.45003118141173504,
+      "learning_rate": 1.667524087033906e-07,
+      "loss": 0.0264,
+      "step": 8336
+    },
+    {
+      "epoch": 3.702864756828781,
+      "grad_norm": 0.42598757489225364,
+      "learning_rate": 1.6625630704160788e-07,
+      "loss": 0.0336,
+      "step": 8337
+    },
+    {
+      "epoch": 3.703308905174328,
+      "grad_norm": 0.38079855634441273,
+      "learning_rate": 1.6576093197361253e-07,
+      "loss": 0.0286,
+      "step": 8338
+    },
+    {
+      "epoch": 3.7037530535198755,
+      "grad_norm": 0.3983365488271041,
+      "learning_rate": 1.652662835738683e-07,
+      "loss": 0.0205,
+      "step": 8339
+    },
+    {
+      "epoch": 3.704197201865423,
+      "grad_norm": 0.29963760935621175,
+      "learning_rate": 1.6477236191673018e-07,
+      "loss": 0.0189,
+      "step": 8340
+    },
+    {
+      "epoch": 3.7046413502109705,
+      "grad_norm": 0.771557737936414,
+      "learning_rate": 1.6427916707644153e-07,
+      "loss": 0.0319,
+      "step": 8341
+    },
+    {
+      "epoch": 3.705085498556518,
+      "grad_norm": 0.4894556058318927,
+      "learning_rate": 1.6378669912713862e-07,
+      "loss": 0.0316,
+      "step": 8342
+    },
+    {
+      "epoch": 3.7055296469020655,
+      "grad_norm": 0.3795606721666376,
+      "learning_rate": 1.6329495814284778e-07,
+      "loss": 0.0196,
+      "step": 8343
+    },
+    {
+      "epoch": 3.705973795247613,
+      "grad_norm": 0.3778626203357572,
+      "learning_rate": 1.62803944197486e-07,
+      "loss": 0.0266,
+      "step": 8344
+    },
+    {
+      "epoch": 3.70641794359316,
+      "grad_norm": 0.4426801167498142,
+      "learning_rate": 1.6231365736486093e-07,
+      "loss": 0.0288,
+      "step": 8345
+    },
+    {
+      "epoch": 3.7068620919387074,
+      "grad_norm": 0.35264417606088994,
+      "learning_rate": 1.6182409771867137e-07,
+      "loss": 0.0189,
+      "step": 8346
+    },
+    {
+      "epoch": 3.7073062402842547,
+      "grad_norm": 0.38892860903541526,
+      "learning_rate": 1.6133526533250566e-07,
+      "loss": 0.0318,
+      "step": 8347
+    },
+    {
+      "epoch": 3.7077503886298024,
+      "grad_norm": 0.3867272658310112,
+      "learning_rate": 1.6084716027984503e-07,
+      "loss": 0.0246,
+      "step": 8348
+    },
+    {
+      "epoch": 3.7081945369753497,
+      "grad_norm": 0.6221413345427776,
+      "learning_rate": 1.6035978263405804e-07,
+      "loss": 0.0333,
+      "step": 8349
+    },
+    {
+      "epoch": 3.708638685320897,
+      "grad_norm": 0.5154361170774239,
+      "learning_rate": 1.5987313246840718e-07,
+      "loss": 0.0342,
+      "step": 8350
+    },
+    {
+      "epoch": 3.7090828336664448,
+      "grad_norm": 0.3524314183478963,
+      "learning_rate": 1.593872098560445e-07,
+      "loss": 0.0211,
+      "step": 8351
+    },
+    {
+      "epoch": 3.709526982011992,
+      "grad_norm": 0.38476894824992575,
+      "learning_rate": 1.58902014870011e-07,
+      "loss": 0.0211,
+      "step": 8352
+    },
+    {
+      "epoch": 3.7099711303575393,
+      "grad_norm": 0.4245656298307411,
+      "learning_rate": 1.5841754758324058e-07,
+      "loss": 0.0209,
+      "step": 8353
+    },
+    {
+      "epoch": 3.7104152787030866,
+      "grad_norm": 0.3820971674731869,
+      "learning_rate": 1.579338080685572e-07,
+      "loss": 0.0221,
+      "step": 8354
+    },
+    {
+      "epoch": 3.7108594270486344,
+      "grad_norm": 0.37340680941918303,
+      "learning_rate": 1.5745079639867488e-07,
+      "loss": 0.0297,
+      "step": 8355
+    },
+    {
+      "epoch": 3.7113035753941817,
+      "grad_norm": 0.29722253097229934,
+      "learning_rate": 1.5696851264619785e-07,
+      "loss": 0.0178,
+      "step": 8356
+    },
+    {
+      "epoch": 3.711747723739729,
+      "grad_norm": 0.4252952177272931,
+      "learning_rate": 1.5648695688362304e-07,
+      "loss": 0.0257,
+      "step": 8357
+    },
+    {
+      "epoch": 3.7121918720852767,
+      "grad_norm": 0.49778617697788585,
+      "learning_rate": 1.560061291833348e-07,
+      "loss": 0.0377,
+      "step": 8358
+    },
+    {
+      "epoch": 3.712636020430824,
+      "grad_norm": 0.45908962801821235,
+      "learning_rate": 1.5552602961761033e-07,
+      "loss": 0.0236,
+      "step": 8359
+    },
+    {
+      "epoch": 3.7130801687763713,
+      "grad_norm": 0.35236790788170025,
+      "learning_rate": 1.5504665825861687e-07,
+      "loss": 0.018,
+      "step": 8360
+    },
+    {
+      "epoch": 3.7135243171219185,
+      "grad_norm": 0.516154033831261,
+      "learning_rate": 1.5456801517841236e-07,
+      "loss": 0.0338,
+      "step": 8361
+    },
+    {
+      "epoch": 3.7139684654674663,
+      "grad_norm": 0.2906873978074194,
+      "learning_rate": 1.540901004489448e-07,
+      "loss": 0.0257,
+      "step": 8362
+    },
+    {
+      "epoch": 3.7144126138130136,
+      "grad_norm": 0.363688530548789,
+      "learning_rate": 1.5361291414205226e-07,
+      "loss": 0.0175,
+      "step": 8363
+    },
+    {
+      "epoch": 3.714856762158561,
+      "grad_norm": 0.4720314053715725,
+      "learning_rate": 1.5313645632946407e-07,
+      "loss": 0.0406,
+      "step": 8364
+    },
+    {
+      "epoch": 3.7153009105041086,
+      "grad_norm": 0.36832914963768065,
+      "learning_rate": 1.5266072708280177e-07,
+      "loss": 0.0232,
+      "step": 8365
+    },
+    {
+      "epoch": 3.715745058849656,
+      "grad_norm": 0.40219479844398387,
+      "learning_rate": 1.5218572647357265e-07,
+      "loss": 0.0325,
+      "step": 8366
+    },
+    {
+      "epoch": 3.716189207195203,
+      "grad_norm": 0.3686119581029194,
+      "learning_rate": 1.517114545731796e-07,
+      "loss": 0.019,
+      "step": 8367
+    },
+    {
+      "epoch": 3.7166333555407505,
+      "grad_norm": 0.4442003891466178,
+      "learning_rate": 1.5123791145291332e-07,
+      "loss": 0.0275,
+      "step": 8368
+    },
+    {
+      "epoch": 3.7170775038862978,
+      "grad_norm": 0.36490090535494746,
+      "learning_rate": 1.5076509718395416e-07,
+      "loss": 0.0155,
+      "step": 8369
+    },
+    {
+      "epoch": 3.7175216522318455,
+      "grad_norm": 0.4009221914723727,
+      "learning_rate": 1.502930118373752e-07,
+      "loss": 0.0223,
+      "step": 8370
+    },
+    {
+      "epoch": 3.717965800577393,
+      "grad_norm": 0.43768126067136603,
+      "learning_rate": 1.4982165548413862e-07,
+      "loss": 0.0192,
+      "step": 8371
+    },
+    {
+      "epoch": 3.7184099489229405,
+      "grad_norm": 0.5261889966594541,
+      "learning_rate": 1.4935102819509717e-07,
+      "loss": 0.0353,
+      "step": 8372
+    },
+    {
+      "epoch": 3.718854097268488,
+      "grad_norm": 0.34594837099051023,
+      "learning_rate": 1.488811300409948e-07,
+      "loss": 0.0188,
+      "step": 8373
+    },
+    {
+      "epoch": 3.719298245614035,
+      "grad_norm": 0.5250244707730272,
+      "learning_rate": 1.4841196109246448e-07,
+      "loss": 0.0355,
+      "step": 8374
+    },
+    {
+      "epoch": 3.7197423939595824,
+      "grad_norm": 0.3336300196833185,
+      "learning_rate": 1.4794352142003088e-07,
+      "loss": 0.0199,
+      "step": 8375
+    },
+    {
+      "epoch": 3.7201865423051297,
+      "grad_norm": 0.37689522491294797,
+      "learning_rate": 1.4747581109410713e-07,
+      "loss": 0.0174,
+      "step": 8376
+    },
+    {
+      "epoch": 3.7206306906506774,
+      "grad_norm": 0.48406735678739193,
+      "learning_rate": 1.4700883018499979e-07,
+      "loss": 0.025,
+      "step": 8377
+    },
+    {
+      "epoch": 3.7210748389962247,
+      "grad_norm": 0.45821951845296044,
+      "learning_rate": 1.4654257876290267e-07,
+      "loss": 0.0392,
+      "step": 8378
+    },
+    {
+      "epoch": 3.721518987341772,
+      "grad_norm": 0.43841923148082784,
+      "learning_rate": 1.4607705689790197e-07,
+      "loss": 0.0213,
+      "step": 8379
+    },
+    {
+      "epoch": 3.7219631356873197,
+      "grad_norm": 0.38180624149856224,
+      "learning_rate": 1.4561226465997337e-07,
+      "loss": 0.0244,
+      "step": 8380
+    },
+    {
+      "epoch": 3.722407284032867,
+      "grad_norm": 0.38238991365264746,
+      "learning_rate": 1.4514820211898263e-07,
+      "loss": 0.0235,
+      "step": 8381
+    },
+    {
+      "epoch": 3.7228514323784143,
+      "grad_norm": 0.4147995047117634,
+      "learning_rate": 1.4468486934468728e-07,
+      "loss": 0.0244,
+      "step": 8382
+    },
+    {
+      "epoch": 3.7232955807239616,
+      "grad_norm": 0.37825138010304116,
+      "learning_rate": 1.442222664067333e-07,
+      "loss": 0.0204,
+      "step": 8383
+    },
+    {
+      "epoch": 3.7237397290695093,
+      "grad_norm": 0.43713370136688856,
+      "learning_rate": 1.437603933746573e-07,
+      "loss": 0.0283,
+      "step": 8384
+    },
+    {
+      "epoch": 3.7241838774150566,
+      "grad_norm": 0.3959829514177559,
+      "learning_rate": 1.4329925031788815e-07,
+      "loss": 0.024,
+      "step": 8385
+    },
+    {
+      "epoch": 3.724628025760604,
+      "grad_norm": 0.36795641365580717,
+      "learning_rate": 1.4283883730574212e-07,
+      "loss": 0.0242,
+      "step": 8386
+    },
+    {
+      "epoch": 3.7250721741061517,
+      "grad_norm": 0.33923664126749403,
+      "learning_rate": 1.4237915440742768e-07,
+      "loss": 0.0259,
+      "step": 8387
+    },
+    {
+      "epoch": 3.725516322451699,
+      "grad_norm": 0.43212191542717354,
+      "learning_rate": 1.4192020169204292e-07,
+      "loss": 0.0205,
+      "step": 8388
+    },
+    {
+      "epoch": 3.7259604707972462,
+      "grad_norm": 0.5186925399242536,
+      "learning_rate": 1.4146197922857597e-07,
+      "loss": 0.032,
+      "step": 8389
+    },
+    {
+      "epoch": 3.7264046191427935,
+      "grad_norm": 0.3400385668731632,
+      "learning_rate": 1.410044870859062e-07,
+      "loss": 0.0224,
+      "step": 8390
+    },
+    {
+      "epoch": 3.7268487674883413,
+      "grad_norm": 0.4227136704820241,
+      "learning_rate": 1.4054772533280137e-07,
+      "loss": 0.0217,
+      "step": 8391
+    },
+    {
+      "epoch": 3.7272929158338886,
+      "grad_norm": 0.4011138415574072,
+      "learning_rate": 1.4009169403792154e-07,
+      "loss": 0.025,
+      "step": 8392
+    },
+    {
+      "epoch": 3.727737064179436,
+      "grad_norm": 0.3003293847098222,
+      "learning_rate": 1.396363932698147e-07,
+      "loss": 0.0174,
+      "step": 8393
+    },
+    {
+      "epoch": 3.7281812125249836,
+      "grad_norm": 0.4465399668144551,
+      "learning_rate": 1.3918182309692164e-07,
+      "loss": 0.0361,
+      "step": 8394
+    },
+    {
+      "epoch": 3.728625360870531,
+      "grad_norm": 0.3571377988241023,
+      "learning_rate": 1.3872798358757155e-07,
+      "loss": 0.0186,
+      "step": 8395
+    },
+    {
+      "epoch": 3.729069509216078,
+      "grad_norm": 0.3618570693279145,
+      "learning_rate": 1.3827487480998437e-07,
+      "loss": 0.0204,
+      "step": 8396
+    },
+    {
+      "epoch": 3.7295136575616255,
+      "grad_norm": 0.4415246137854366,
+      "learning_rate": 1.3782249683226946e-07,
+      "loss": 0.0215,
+      "step": 8397
+    },
+    {
+      "epoch": 3.7299578059071727,
+      "grad_norm": 0.3988226147588891,
+      "learning_rate": 1.373708497224263e-07,
+      "loss": 0.025,
+      "step": 8398
+    },
+    {
+      "epoch": 3.7304019542527205,
+      "grad_norm": 0.3588668524798711,
+      "learning_rate": 1.3691993354834733e-07,
+      "loss": 0.017,
+      "step": 8399
+    },
+    {
+      "epoch": 3.7308461025982678,
+      "grad_norm": 0.4766338898375096,
+      "learning_rate": 1.3646974837781102e-07,
+      "loss": 0.0246,
+      "step": 8400
+    },
+    {
+      "epoch": 3.7312902509438155,
+      "grad_norm": 0.4346113992522893,
+      "learning_rate": 1.3602029427848885e-07,
+      "loss": 0.0224,
+      "step": 8401
+    },
+    {
+      "epoch": 3.731734399289363,
+      "grad_norm": 0.4308704837627075,
+      "learning_rate": 1.355715713179412e-07,
+      "loss": 0.0228,
+      "step": 8402
+    },
+    {
+      "epoch": 3.73217854763491,
+      "grad_norm": 0.3138469968218578,
+      "learning_rate": 1.35123579563618e-07,
+      "loss": 0.0211,
+      "step": 8403
+    },
+    {
+      "epoch": 3.7326226959804574,
+      "grad_norm": 0.3603696166551371,
+      "learning_rate": 1.346763190828604e-07,
+      "loss": 0.0202,
+      "step": 8404
+    },
+    {
+      "epoch": 3.7330668443260047,
+      "grad_norm": 0.472833085205998,
+      "learning_rate": 1.3422978994290014e-07,
+      "loss": 0.032,
+      "step": 8405
+    },
+    {
+      "epoch": 3.7335109926715524,
+      "grad_norm": 0.5373078797879084,
+      "learning_rate": 1.3378399221085691e-07,
+      "loss": 0.0331,
+      "step": 8406
+    },
+    {
+      "epoch": 3.7339551410170997,
+      "grad_norm": 0.32376977842916665,
+      "learning_rate": 1.3333892595374265e-07,
+      "loss": 0.0193,
+      "step": 8407
+    },
+    {
+      "epoch": 3.734399289362647,
+      "grad_norm": 0.5384681977899171,
+      "learning_rate": 1.3289459123845772e-07,
+      "loss": 0.0439,
+      "step": 8408
+    },
+    {
+      "epoch": 3.7348434377081947,
+      "grad_norm": 0.360367652411676,
+      "learning_rate": 1.3245098813179315e-07,
+      "loss": 0.0259,
+      "step": 8409
+    },
+    {
+      "epoch": 3.735287586053742,
+      "grad_norm": 0.34043662992938706,
+      "learning_rate": 1.3200811670043057e-07,
+      "loss": 0.0156,
+      "step": 8410
+    },
+    {
+      "epoch": 3.7357317343992893,
+      "grad_norm": 0.46845973568867966,
+      "learning_rate": 1.3156597701094065e-07,
+      "loss": 0.0223,
+      "step": 8411
+    },
+    {
+      "epoch": 3.7361758827448366,
+      "grad_norm": 0.4170724326037999,
+      "learning_rate": 1.3112456912978467e-07,
+      "loss": 0.0183,
+      "step": 8412
+    },
+    {
+      "epoch": 3.7366200310903843,
+      "grad_norm": 0.6062425923257145,
+      "learning_rate": 1.3068389312331398e-07,
+      "loss": 0.0326,
+      "step": 8413
+    },
+    {
+      "epoch": 3.7370641794359316,
+      "grad_norm": 0.32063888344376806,
+      "learning_rate": 1.3024394905776893e-07,
+      "loss": 0.0155,
+      "step": 8414
+    },
+    {
+      "epoch": 3.737508327781479,
+      "grad_norm": 0.4839068579509274,
+      "learning_rate": 1.298047369992811e-07,
+      "loss": 0.0285,
+      "step": 8415
+    },
+    {
+      "epoch": 3.7379524761270266,
+      "grad_norm": 0.4686148975397421,
+      "learning_rate": 1.2936625701387152e-07,
+      "loss": 0.0291,
+      "step": 8416
+    },
+    {
+      "epoch": 3.738396624472574,
+      "grad_norm": 0.40247010635972097,
+      "learning_rate": 1.289285091674508e-07,
+      "loss": 0.0248,
+      "step": 8417
+    },
+    {
+      "epoch": 3.738840772818121,
+      "grad_norm": 0.3788928148883821,
+      "learning_rate": 1.2849149352582135e-07,
+      "loss": 0.0277,
+      "step": 8418
+    },
+    {
+      "epoch": 3.7392849211636685,
+      "grad_norm": 0.45807438580819443,
+      "learning_rate": 1.280552101546717e-07,
+      "loss": 0.0311,
+      "step": 8419
+    },
+    {
+      "epoch": 3.7397290695092162,
+      "grad_norm": 0.513273904365022,
+      "learning_rate": 1.2761965911958385e-07,
+      "loss": 0.0245,
+      "step": 8420
+    },
+    {
+      "epoch": 3.7401732178547635,
+      "grad_norm": 0.39246829057407184,
+      "learning_rate": 1.2718484048602876e-07,
+      "loss": 0.0231,
+      "step": 8421
+    },
+    {
+      "epoch": 3.740617366200311,
+      "grad_norm": 0.41649329925637596,
+      "learning_rate": 1.267507543193669e-07,
+      "loss": 0.0199,
+      "step": 8422
+    },
+    {
+      "epoch": 3.7410615145458586,
+      "grad_norm": 0.42862616530797626,
+      "learning_rate": 1.2631740068484888e-07,
+      "loss": 0.0307,
+      "step": 8423
+    },
+    {
+      "epoch": 3.741505662891406,
+      "grad_norm": 0.35644680831072795,
+      "learning_rate": 1.258847796476148e-07,
+      "loss": 0.0241,
+      "step": 8424
+    },
+    {
+      "epoch": 3.741949811236953,
+      "grad_norm": 0.32721758016106417,
+      "learning_rate": 1.2545289127269488e-07,
+      "loss": 0.0235,
+      "step": 8425
+    },
+    {
+      "epoch": 3.7423939595825004,
+      "grad_norm": 0.40082921797531834,
+      "learning_rate": 1.2502173562500995e-07,
+      "loss": 0.021,
+      "step": 8426
+    },
+    {
+      "epoch": 3.7428381079280477,
+      "grad_norm": 0.39626883986681716,
+      "learning_rate": 1.2459131276936876e-07,
+      "loss": 0.0242,
+      "step": 8427
+    },
+    {
+      "epoch": 3.7432822562735955,
+      "grad_norm": 0.5073039426598518,
+      "learning_rate": 1.241616227704723e-07,
+      "loss": 0.0278,
+      "step": 8428
+    },
+    {
+      "epoch": 3.7437264046191427,
+      "grad_norm": 0.3603567472819875,
+      "learning_rate": 1.2373266569290997e-07,
+      "loss": 0.032,
+      "step": 8429
+    },
+    {
+      "epoch": 3.7441705529646905,
+      "grad_norm": 0.4679950484890861,
+      "learning_rate": 1.2330444160116196e-07,
+      "loss": 0.0208,
+      "step": 8430
+    },
+    {
+      "epoch": 3.7446147013102378,
+      "grad_norm": 0.468095341335435,
+      "learning_rate": 1.2287695055959615e-07,
+      "loss": 0.0223,
+      "step": 8431
+    },
+    {
+      "epoch": 3.745058849655785,
+      "grad_norm": 0.4819559753911326,
+      "learning_rate": 1.2245019263247283e-07,
+      "loss": 0.0262,
+      "step": 8432
+    },
+    {
+      "epoch": 3.7455029980013324,
+      "grad_norm": 0.4954170402461803,
+      "learning_rate": 1.2202416788394067e-07,
+      "loss": 0.0333,
+      "step": 8433
+    },
+    {
+      "epoch": 3.7459471463468796,
+      "grad_norm": 0.6174858065052228,
+      "learning_rate": 1.215988763780379e-07,
+      "loss": 0.0313,
+      "step": 8434
+    },
+    {
+      "epoch": 3.7463912946924274,
+      "grad_norm": 0.42561144433663495,
+      "learning_rate": 1.2117431817869453e-07,
+      "loss": 0.0277,
+      "step": 8435
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.3709807838740829,
+      "learning_rate": 1.207504933497272e-07,
+      "loss": 0.0243,
+      "step": 8436
+    },
+    {
+      "epoch": 3.747279591383522,
+      "grad_norm": 0.3000902649747397,
+      "learning_rate": 1.2032740195484448e-07,
+      "loss": 0.0147,
+      "step": 8437
+    },
+    {
+      "epoch": 3.7477237397290697,
+      "grad_norm": 0.5581005324860493,
+      "learning_rate": 1.1990504405764492e-07,
+      "loss": 0.0302,
+      "step": 8438
+    },
+    {
+      "epoch": 3.748167888074617,
+      "grad_norm": 0.5633312132178465,
+      "learning_rate": 1.1948341972161492e-07,
+      "loss": 0.0309,
+      "step": 8439
+    },
+    {
+      "epoch": 3.7486120364201643,
+      "grad_norm": 0.36629488791856035,
+      "learning_rate": 1.1906252901013271e-07,
+      "loss": 0.0238,
+      "step": 8440
+    },
+    {
+      "epoch": 3.7490561847657116,
+      "grad_norm": 0.49491041444267353,
+      "learning_rate": 1.1864237198646544e-07,
+      "loss": 0.0298,
+      "step": 8441
+    },
+    {
+      "epoch": 3.7495003331112593,
+      "grad_norm": 0.564678297749189,
+      "learning_rate": 1.1822294871376928e-07,
+      "loss": 0.0305,
+      "step": 8442
+    },
+    {
+      "epoch": 3.7499444814568066,
+      "grad_norm": 0.40055447684951356,
+      "learning_rate": 1.1780425925509043e-07,
+      "loss": 0.0272,
+      "step": 8443
+    },
+    {
+      "epoch": 3.750388629802354,
+      "grad_norm": 0.470018676790432,
+      "learning_rate": 1.1738630367336579e-07,
+      "loss": 0.032,
+      "step": 8444
+    },
+    {
+      "epoch": 3.7508327781479016,
+      "grad_norm": 0.4384932264169123,
+      "learning_rate": 1.1696908203142066e-07,
+      "loss": 0.0256,
+      "step": 8445
+    },
+    {
+      "epoch": 3.751276926493449,
+      "grad_norm": 0.4653823127550869,
+      "learning_rate": 1.1655259439197042e-07,
+      "loss": 0.03,
+      "step": 8446
+    },
+    {
+      "epoch": 3.751721074838996,
+      "grad_norm": 0.41947700796617987,
+      "learning_rate": 1.1613684081762111e-07,
+      "loss": 0.0236,
+      "step": 8447
+    },
+    {
+      "epoch": 3.7521652231845435,
+      "grad_norm": 0.38536838427097736,
+      "learning_rate": 1.1572182137086662e-07,
+      "loss": 0.0242,
+      "step": 8448
+    },
+    {
+      "epoch": 3.7526093715300908,
+      "grad_norm": 0.341291576535971,
+      "learning_rate": 1.1530753611409151e-07,
+      "loss": 0.0229,
+      "step": 8449
+    },
+    {
+      "epoch": 3.7530535198756385,
+      "grad_norm": 0.3792987155782225,
+      "learning_rate": 1.1489398510957039e-07,
+      "loss": 0.0237,
+      "step": 8450
+    },
+    {
+      "epoch": 3.753497668221186,
+      "grad_norm": 0.37854108977812184,
+      "learning_rate": 1.1448116841946688e-07,
+      "loss": 0.0287,
+      "step": 8451
+    },
+    {
+      "epoch": 3.7539418165667335,
+      "grad_norm": 0.46875753276752063,
+      "learning_rate": 1.1406908610583467e-07,
+      "loss": 0.0311,
+      "step": 8452
+    },
+    {
+      "epoch": 3.754385964912281,
+      "grad_norm": 0.4064331207182262,
+      "learning_rate": 1.1365773823061532e-07,
+      "loss": 0.0283,
+      "step": 8453
+    },
+    {
+      "epoch": 3.754830113257828,
+      "grad_norm": 0.4002763856777381,
+      "learning_rate": 1.1324712485564271e-07,
+      "loss": 0.024,
+      "step": 8454
+    },
+    {
+      "epoch": 3.7552742616033754,
+      "grad_norm": 0.45827533835175305,
+      "learning_rate": 1.1283724604263857e-07,
+      "loss": 0.0246,
+      "step": 8455
+    },
+    {
+      "epoch": 3.7557184099489227,
+      "grad_norm": 0.39394686171582316,
+      "learning_rate": 1.1242810185321473e-07,
+      "loss": 0.03,
+      "step": 8456
+    },
+    {
+      "epoch": 3.7561625582944704,
+      "grad_norm": 0.42301627464562236,
+      "learning_rate": 1.1201969234887256e-07,
+      "loss": 0.0321,
+      "step": 8457
+    },
+    {
+      "epoch": 3.7566067066400177,
+      "grad_norm": 0.5331607328416476,
+      "learning_rate": 1.1161201759100349e-07,
+      "loss": 0.0279,
+      "step": 8458
+    },
+    {
+      "epoch": 3.757050854985565,
+      "grad_norm": 0.44317622469926854,
+      "learning_rate": 1.1120507764088684e-07,
+      "loss": 0.0449,
+      "step": 8459
+    },
+    {
+      "epoch": 3.7574950033311127,
+      "grad_norm": 0.3625167956575573,
+      "learning_rate": 1.1079887255969257e-07,
+      "loss": 0.0194,
+      "step": 8460
+    },
+    {
+      "epoch": 3.75793915167666,
+      "grad_norm": 0.38235614382527433,
+      "learning_rate": 1.1039340240848129e-07,
+      "loss": 0.0286,
+      "step": 8461
+    },
+    {
+      "epoch": 3.7583833000222073,
+      "grad_norm": 0.4781732717681245,
+      "learning_rate": 1.0998866724820145e-07,
+      "loss": 0.0275,
+      "step": 8462
+    },
+    {
+      "epoch": 3.7588274483677546,
+      "grad_norm": 0.36277030824808554,
+      "learning_rate": 1.0958466713969218e-07,
+      "loss": 0.02,
+      "step": 8463
+    },
+    {
+      "epoch": 3.7592715967133024,
+      "grad_norm": 0.4310996548128226,
+      "learning_rate": 1.09181402143681e-07,
+      "loss": 0.0271,
+      "step": 8464
+    },
+    {
+      "epoch": 3.7597157450588496,
+      "grad_norm": 0.44245437066064086,
+      "learning_rate": 1.0877887232078499e-07,
+      "loss": 0.0221,
+      "step": 8465
+    },
+    {
+      "epoch": 3.760159893404397,
+      "grad_norm": 0.45765152233325074,
+      "learning_rate": 1.0837707773151185e-07,
+      "loss": 0.0229,
+      "step": 8466
+    },
+    {
+      "epoch": 3.7606040417499447,
+      "grad_norm": 0.5026222951099458,
+      "learning_rate": 1.0797601843625827e-07,
+      "loss": 0.0253,
+      "step": 8467
+    },
+    {
+      "epoch": 3.761048190095492,
+      "grad_norm": 0.5808566690208385,
+      "learning_rate": 1.0757569449530991e-07,
+      "loss": 0.0229,
+      "step": 8468
+    },
+    {
+      "epoch": 3.7614923384410393,
+      "grad_norm": 0.36877164338905205,
+      "learning_rate": 1.0717610596884309e-07,
+      "loss": 0.0285,
+      "step": 8469
+    },
+    {
+      "epoch": 3.7619364867865865,
+      "grad_norm": 0.3700848691513958,
+      "learning_rate": 1.0677725291692143e-07,
+      "loss": 0.0204,
+      "step": 8470
+    },
+    {
+      "epoch": 3.7623806351321343,
+      "grad_norm": 0.3125009535356397,
+      "learning_rate": 1.0637913539950029e-07,
+      "loss": 0.0181,
+      "step": 8471
+    },
+    {
+      "epoch": 3.7628247834776816,
+      "grad_norm": 0.4808509442055829,
+      "learning_rate": 1.0598175347642293e-07,
+      "loss": 0.028,
+      "step": 8472
+    },
+    {
+      "epoch": 3.763268931823229,
+      "grad_norm": 0.4442187709817902,
+      "learning_rate": 1.0558510720742265e-07,
+      "loss": 0.0316,
+      "step": 8473
+    },
+    {
+      "epoch": 3.7637130801687766,
+      "grad_norm": 0.5324409399106215,
+      "learning_rate": 1.0518919665212235e-07,
+      "loss": 0.0319,
+      "step": 8474
+    },
+    {
+      "epoch": 3.764157228514324,
+      "grad_norm": 0.42314697759763437,
+      "learning_rate": 1.0479402187003496e-07,
+      "loss": 0.0222,
+      "step": 8475
+    },
+    {
+      "epoch": 3.764601376859871,
+      "grad_norm": 0.378666602483765,
+      "learning_rate": 1.0439958292056074e-07,
+      "loss": 0.029,
+      "step": 8476
+    },
+    {
+      "epoch": 3.7650455252054185,
+      "grad_norm": 0.5003419916114011,
+      "learning_rate": 1.040058798629906e-07,
+      "loss": 0.0259,
+      "step": 8477
+    },
+    {
+      "epoch": 3.7654896735509658,
+      "grad_norm": 0.5187802895889871,
+      "learning_rate": 1.0361291275650498e-07,
+      "loss": 0.0412,
+      "step": 8478
+    },
+    {
+      "epoch": 3.7659338218965135,
+      "grad_norm": 0.41358816649715524,
+      "learning_rate": 1.0322068166017386e-07,
+      "loss": 0.0325,
+      "step": 8479
+    },
+    {
+      "epoch": 3.766377970242061,
+      "grad_norm": 0.3554104231788337,
+      "learning_rate": 1.0282918663295616e-07,
+      "loss": 0.0266,
+      "step": 8480
+    },
+    {
+      "epoch": 3.7668221185876085,
+      "grad_norm": 0.433977374572574,
+      "learning_rate": 1.0243842773369983e-07,
+      "loss": 0.0262,
+      "step": 8481
+    },
+    {
+      "epoch": 3.767266266933156,
+      "grad_norm": 0.3050089012609615,
+      "learning_rate": 1.0204840502114288e-07,
+      "loss": 0.016,
+      "step": 8482
+    },
+    {
+      "epoch": 3.767710415278703,
+      "grad_norm": 0.35035155995481493,
+      "learning_rate": 1.0165911855391286e-07,
+      "loss": 0.022,
+      "step": 8483
+    },
+    {
+      "epoch": 3.7681545636242504,
+      "grad_norm": 0.3852773325944005,
+      "learning_rate": 1.0127056839052462e-07,
+      "loss": 0.017,
+      "step": 8484
+    },
+    {
+      "epoch": 3.7685987119697977,
+      "grad_norm": 0.43026798549784784,
+      "learning_rate": 1.0088275458938535e-07,
+      "loss": 0.0246,
+      "step": 8485
+    },
+    {
+      "epoch": 3.7690428603153454,
+      "grad_norm": 0.49903822216529625,
+      "learning_rate": 1.004956772087895e-07,
+      "loss": 0.0278,
+      "step": 8486
+    },
+    {
+      "epoch": 3.7694870086608927,
+      "grad_norm": 0.30382121537402157,
+      "learning_rate": 1.0010933630692166e-07,
+      "loss": 0.0168,
+      "step": 8487
+    },
+    {
+      "epoch": 3.76993115700644,
+      "grad_norm": 0.41575029483477577,
+      "learning_rate": 9.972373194185481e-08,
+      "loss": 0.0273,
+      "step": 8488
+    },
+    {
+      "epoch": 3.7703753053519877,
+      "grad_norm": 0.425652535729631,
+      "learning_rate": 9.933886417155258e-08,
+      "loss": 0.0258,
+      "step": 8489
+    },
+    {
+      "epoch": 3.770819453697535,
+      "grad_norm": 0.4620847686387116,
+      "learning_rate": 9.895473305386593e-08,
+      "loss": 0.024,
+      "step": 8490
+    },
+    {
+      "epoch": 3.7712636020430823,
+      "grad_norm": 0.4272116851275043,
+      "learning_rate": 9.857133864653812e-08,
+      "loss": 0.0246,
+      "step": 8491
+    },
+    {
+      "epoch": 3.7717077503886296,
+      "grad_norm": 0.3869470039777947,
+      "learning_rate": 9.818868100719803e-08,
+      "loss": 0.0266,
+      "step": 8492
+    },
+    {
+      "epoch": 3.7721518987341773,
+      "grad_norm": 0.38019015167069964,
+      "learning_rate": 9.780676019336632e-08,
+      "loss": 0.0235,
+      "step": 8493
+    },
+    {
+      "epoch": 3.7725960470797246,
+      "grad_norm": 0.6659937799546494,
+      "learning_rate": 9.742557626245264e-08,
+      "loss": 0.0336,
+      "step": 8494
+    },
+    {
+      "epoch": 3.773040195425272,
+      "grad_norm": 0.5490763929102886,
+      "learning_rate": 9.704512927175502e-08,
+      "loss": 0.0325,
+      "step": 8495
+    },
+    {
+      "epoch": 3.7734843437708196,
+      "grad_norm": 0.5157099798547079,
+      "learning_rate": 9.666541927846107e-08,
+      "loss": 0.0408,
+      "step": 8496
+    },
+    {
+      "epoch": 3.773928492116367,
+      "grad_norm": 0.6662930057806212,
+      "learning_rate": 9.62864463396479e-08,
+      "loss": 0.0181,
+      "step": 8497
+    },
+    {
+      "epoch": 3.7743726404619142,
+      "grad_norm": 0.36908564910307085,
+      "learning_rate": 9.590821051228105e-08,
+      "loss": 0.0196,
+      "step": 8498
+    },
+    {
+      "epoch": 3.7748167888074615,
+      "grad_norm": 0.34743147522098267,
+      "learning_rate": 9.553071185321616e-08,
+      "loss": 0.0177,
+      "step": 8499
+    },
+    {
+      "epoch": 3.7752609371530093,
+      "grad_norm": 0.4868308303940559,
+      "learning_rate": 9.515395041919839e-08,
+      "loss": 0.0292,
+      "step": 8500
+    },
+    {
+      "epoch": 3.7757050854985565,
+      "grad_norm": 0.4939860503734759,
+      "learning_rate": 9.477792626685966e-08,
+      "loss": 0.026,
+      "step": 8501
+    },
+    {
+      "epoch": 3.776149233844104,
+      "grad_norm": 0.31320876766724953,
+      "learning_rate": 9.440263945272365e-08,
+      "loss": 0.018,
+      "step": 8502
+    },
+    {
+      "epoch": 3.7765933821896516,
+      "grad_norm": 0.36925404461957756,
+      "learning_rate": 9.402809003320357e-08,
+      "loss": 0.022,
+      "step": 8503
+    },
+    {
+      "epoch": 3.777037530535199,
+      "grad_norm": 0.5099202980127734,
+      "learning_rate": 9.365427806459826e-08,
+      "loss": 0.0219,
+      "step": 8504
+    },
+    {
+      "epoch": 3.777481678880746,
+      "grad_norm": 0.4048961589102498,
+      "learning_rate": 9.32812036031e-08,
+      "loss": 0.02,
+      "step": 8505
+    },
+    {
+      "epoch": 3.7779258272262934,
+      "grad_norm": 0.3828893218593422,
+      "learning_rate": 9.290886670478727e-08,
+      "loss": 0.0329,
+      "step": 8506
+    },
+    {
+      "epoch": 3.7783699755718407,
+      "grad_norm": 0.47619859186660657,
+      "learning_rate": 9.253726742562808e-08,
+      "loss": 0.0339,
+      "step": 8507
+    },
+    {
+      "epoch": 3.7788141239173885,
+      "grad_norm": 0.3817560157892672,
+      "learning_rate": 9.216640582148218e-08,
+      "loss": 0.0245,
+      "step": 8508
+    },
+    {
+      "epoch": 3.7792582722629358,
+      "grad_norm": 0.44955998588034224,
+      "learning_rate": 9.179628194809387e-08,
+      "loss": 0.036,
+      "step": 8509
+    },
+    {
+      "epoch": 3.7797024206084835,
+      "grad_norm": 0.39409371816193695,
+      "learning_rate": 9.142689586110032e-08,
+      "loss": 0.0243,
+      "step": 8510
+    },
+    {
+      "epoch": 3.780146568954031,
+      "grad_norm": 0.3329695501761576,
+      "learning_rate": 9.105824761602711e-08,
+      "loss": 0.0177,
+      "step": 8511
+    },
+    {
+      "epoch": 3.780590717299578,
+      "grad_norm": 0.41637622163142424,
+      "learning_rate": 9.069033726828657e-08,
+      "loss": 0.0243,
+      "step": 8512
+    },
+    {
+      "epoch": 3.7810348656451254,
+      "grad_norm": 0.513878740141547,
+      "learning_rate": 9.032316487318338e-08,
+      "loss": 0.0422,
+      "step": 8513
+    },
+    {
+      "epoch": 3.7814790139906727,
+      "grad_norm": 0.3779536127482641,
+      "learning_rate": 8.995673048591002e-08,
+      "loss": 0.0243,
+      "step": 8514
+    },
+    {
+      "epoch": 3.7819231623362204,
+      "grad_norm": 0.40656693503772895,
+      "learning_rate": 8.959103416154635e-08,
+      "loss": 0.0204,
+      "step": 8515
+    },
+    {
+      "epoch": 3.7823673106817677,
+      "grad_norm": 0.37125663642968887,
+      "learning_rate": 8.922607595506339e-08,
+      "loss": 0.0222,
+      "step": 8516
+    },
+    {
+      "epoch": 3.782811459027315,
+      "grad_norm": 0.4060879405476042,
+      "learning_rate": 8.886185592132113e-08,
+      "loss": 0.0213,
+      "step": 8517
+    },
+    {
+      "epoch": 3.7832556073728627,
+      "grad_norm": 0.4488358277812707,
+      "learning_rate": 8.849837411506745e-08,
+      "loss": 0.0344,
+      "step": 8518
+    },
+    {
+      "epoch": 3.78369975571841,
+      "grad_norm": 0.4066455322610377,
+      "learning_rate": 8.813563059093977e-08,
+      "loss": 0.0242,
+      "step": 8519
+    },
+    {
+      "epoch": 3.7841439040639573,
+      "grad_norm": 0.4551394122387141,
+      "learning_rate": 8.777362540346501e-08,
+      "loss": 0.0293,
+      "step": 8520
+    },
+    {
+      "epoch": 3.7845880524095046,
+      "grad_norm": 0.42457221786596655,
+      "learning_rate": 8.741235860705855e-08,
+      "loss": 0.0259,
+      "step": 8521
+    },
+    {
+      "epoch": 3.7850322007550523,
+      "grad_norm": 0.3043935300931366,
+      "learning_rate": 8.70518302560247e-08,
+      "loss": 0.0182,
+      "step": 8522
+    },
+    {
+      "epoch": 3.7854763491005996,
+      "grad_norm": 0.4462173753124753,
+      "learning_rate": 8.669204040455737e-08,
+      "loss": 0.0265,
+      "step": 8523
+    },
+    {
+      "epoch": 3.785920497446147,
+      "grad_norm": 0.41601964127930957,
+      "learning_rate": 8.633298910673826e-08,
+      "loss": 0.0288,
+      "step": 8524
+    },
+    {
+      "epoch": 3.7863646457916946,
+      "grad_norm": 0.2813068890343098,
+      "learning_rate": 8.597467641654034e-08,
+      "loss": 0.0212,
+      "step": 8525
+    },
+    {
+      "epoch": 3.786808794137242,
+      "grad_norm": 0.43623034836188324,
+      "learning_rate": 8.561710238782272e-08,
+      "loss": 0.0249,
+      "step": 8526
+    },
+    {
+      "epoch": 3.787252942482789,
+      "grad_norm": 0.5751855788252586,
+      "learning_rate": 8.526026707433577e-08,
+      "loss": 0.0413,
+      "step": 8527
+    },
+    {
+      "epoch": 3.7876970908283365,
+      "grad_norm": 0.3868182612344314,
+      "learning_rate": 8.490417052971766e-08,
+      "loss": 0.0281,
+      "step": 8528
+    },
+    {
+      "epoch": 3.7881412391738842,
+      "grad_norm": 0.356372518499826,
+      "learning_rate": 8.45488128074945e-08,
+      "loss": 0.022,
+      "step": 8529
+    },
+    {
+      "epoch": 3.7885853875194315,
+      "grad_norm": 0.32262084504669236,
+      "learning_rate": 8.419419396108464e-08,
+      "loss": 0.0168,
+      "step": 8530
+    },
+    {
+      "epoch": 3.789029535864979,
+      "grad_norm": 0.4043798709101995,
+      "learning_rate": 8.384031404379211e-08,
+      "loss": 0.024,
+      "step": 8531
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.41478103237109665,
+      "learning_rate": 8.34871731088116e-08,
+      "loss": 0.027,
+      "step": 8532
+    },
+    {
+      "epoch": 3.789917832556074,
+      "grad_norm": 0.4196449951874737,
+      "learning_rate": 8.313477120922563e-08,
+      "loss": 0.0209,
+      "step": 8533
+    },
+    {
+      "epoch": 3.790361980901621,
+      "grad_norm": 0.4317330818971433,
+      "learning_rate": 8.278310839800685e-08,
+      "loss": 0.0284,
+      "step": 8534
+    },
+    {
+      "epoch": 3.7908061292471684,
+      "grad_norm": 0.44713201608661013,
+      "learning_rate": 8.243218472801461e-08,
+      "loss": 0.0438,
+      "step": 8535
+    },
+    {
+      "epoch": 3.7912502775927157,
+      "grad_norm": 0.4371441171214935,
+      "learning_rate": 8.208200025200119e-08,
+      "loss": 0.0201,
+      "step": 8536
+    },
+    {
+      "epoch": 3.7916944259382634,
+      "grad_norm": 0.5043700394965174,
+      "learning_rate": 8.173255502260336e-08,
+      "loss": 0.0285,
+      "step": 8537
+    },
+    {
+      "epoch": 3.7921385742838107,
+      "grad_norm": 0.3893482313023411,
+      "learning_rate": 8.138384909234964e-08,
+      "loss": 0.0258,
+      "step": 8538
+    },
+    {
+      "epoch": 3.7925827226293585,
+      "grad_norm": 0.3403931667558864,
+      "learning_rate": 8.103588251365534e-08,
+      "loss": 0.0199,
+      "step": 8539
+    },
+    {
+      "epoch": 3.7930268709749058,
+      "grad_norm": 0.4249904672408077,
+      "learning_rate": 8.068865533882752e-08,
+      "loss": 0.0221,
+      "step": 8540
+    },
+    {
+      "epoch": 3.793471019320453,
+      "grad_norm": 0.4067407706983934,
+      "learning_rate": 8.034216762005831e-08,
+      "loss": 0.0374,
+      "step": 8541
+    },
+    {
+      "epoch": 3.7939151676660003,
+      "grad_norm": 0.41570749972762994,
+      "learning_rate": 7.99964194094327e-08,
+      "loss": 0.024,
+      "step": 8542
+    },
+    {
+      "epoch": 3.7943593160115476,
+      "grad_norm": 0.48499039450946113,
+      "learning_rate": 7.96514107589208e-08,
+      "loss": 0.0341,
+      "step": 8543
+    },
+    {
+      "epoch": 3.7948034643570954,
+      "grad_norm": 0.40373614464609714,
+      "learning_rate": 7.93071417203839e-08,
+      "loss": 0.0268,
+      "step": 8544
+    },
+    {
+      "epoch": 3.7952476127026427,
+      "grad_norm": 0.35452268236632206,
+      "learning_rate": 7.896361234557226e-08,
+      "loss": 0.0192,
+      "step": 8545
+    },
+    {
+      "epoch": 3.79569176104819,
+      "grad_norm": 0.4666753180876556,
+      "learning_rate": 7.862082268612237e-08,
+      "loss": 0.0342,
+      "step": 8546
+    },
+    {
+      "epoch": 3.7961359093937377,
+      "grad_norm": 0.39409303578975985,
+      "learning_rate": 7.8278772793563e-08,
+      "loss": 0.0244,
+      "step": 8547
+    },
+    {
+      "epoch": 3.796580057739285,
+      "grad_norm": 0.543942317931089,
+      "learning_rate": 7.793746271930968e-08,
+      "loss": 0.0232,
+      "step": 8548
+    },
+    {
+      "epoch": 3.7970242060848323,
+      "grad_norm": 0.4331154507830711,
+      "learning_rate": 7.759689251466695e-08,
+      "loss": 0.0314,
+      "step": 8549
+    },
+    {
+      "epoch": 3.7974683544303796,
+      "grad_norm": 0.4001739441704062,
+      "learning_rate": 7.72570622308283e-08,
+      "loss": 0.0201,
+      "step": 8550
+    },
+    {
+      "epoch": 3.7979125027759273,
+      "grad_norm": 0.4891526654283636,
+      "learning_rate": 7.691797191887618e-08,
+      "loss": 0.027,
+      "step": 8551
+    },
+    {
+      "epoch": 3.7983566511214746,
+      "grad_norm": 0.40920185175638346,
+      "learning_rate": 7.657962162978038e-08,
+      "loss": 0.0264,
+      "step": 8552
+    },
+    {
+      "epoch": 3.798800799467022,
+      "grad_norm": 0.4710270896413282,
+      "learning_rate": 7.624201141440301e-08,
+      "loss": 0.0295,
+      "step": 8553
+    },
+    {
+      "epoch": 3.7992449478125696,
+      "grad_norm": 0.3793884463205298,
+      "learning_rate": 7.59051413234907e-08,
+      "loss": 0.0162,
+      "step": 8554
+    },
+    {
+      "epoch": 3.799689096158117,
+      "grad_norm": 0.2877108795069413,
+      "learning_rate": 7.556901140768125e-08,
+      "loss": 0.0134,
+      "step": 8555
+    },
+    {
+      "epoch": 3.800133244503664,
+      "grad_norm": 0.3350085177589909,
+      "learning_rate": 7.523362171750148e-08,
+      "loss": 0.0139,
+      "step": 8556
+    },
+    {
+      "epoch": 3.8005773928492115,
+      "grad_norm": 0.4349812650446206,
+      "learning_rate": 7.489897230336496e-08,
+      "loss": 0.0276,
+      "step": 8557
+    },
+    {
+      "epoch": 3.801021541194759,
+      "grad_norm": 0.3673667656791951,
+      "learning_rate": 7.456506321557533e-08,
+      "loss": 0.0212,
+      "step": 8558
+    },
+    {
+      "epoch": 3.8014656895403065,
+      "grad_norm": 0.4327339896844779,
+      "learning_rate": 7.423189450432633e-08,
+      "loss": 0.0234,
+      "step": 8559
+    },
+    {
+      "epoch": 3.801909837885854,
+      "grad_norm": 0.3414463099110009,
+      "learning_rate": 7.389946621969679e-08,
+      "loss": 0.02,
+      "step": 8560
+    },
+    {
+      "epoch": 3.8023539862314015,
+      "grad_norm": 0.35401161038931783,
+      "learning_rate": 7.356777841165786e-08,
+      "loss": 0.0236,
+      "step": 8561
+    },
+    {
+      "epoch": 3.802798134576949,
+      "grad_norm": 0.3318458174565355,
+      "learning_rate": 7.32368311300674e-08,
+      "loss": 0.0228,
+      "step": 8562
+    },
+    {
+      "epoch": 3.803242282922496,
+      "grad_norm": 0.4407338624172669,
+      "learning_rate": 7.290662442467178e-08,
+      "loss": 0.0243,
+      "step": 8563
+    },
+    {
+      "epoch": 3.8036864312680434,
+      "grad_norm": 0.3688164392706434,
+      "learning_rate": 7.257715834510737e-08,
+      "loss": 0.0194,
+      "step": 8564
+    },
+    {
+      "epoch": 3.8041305796135907,
+      "grad_norm": 0.4497260712320215,
+      "learning_rate": 7.224843294089844e-08,
+      "loss": 0.0287,
+      "step": 8565
+    },
+    {
+      "epoch": 3.8045747279591384,
+      "grad_norm": 0.33236407987313205,
+      "learning_rate": 7.192044826145772e-08,
+      "loss": 0.0191,
+      "step": 8566
+    },
+    {
+      "epoch": 3.8050188763046857,
+      "grad_norm": 0.37564613529503493,
+      "learning_rate": 7.159320435608741e-08,
+      "loss": 0.0161,
+      "step": 8567
+    },
+    {
+      "epoch": 3.8054630246502335,
+      "grad_norm": 0.3522232268407622,
+      "learning_rate": 7.126670127397705e-08,
+      "loss": 0.0215,
+      "step": 8568
+    },
+    {
+      "epoch": 3.8059071729957807,
+      "grad_norm": 0.429595537532537,
+      "learning_rate": 7.094093906420629e-08,
+      "loss": 0.0288,
+      "step": 8569
+    },
+    {
+      "epoch": 3.806351321341328,
+      "grad_norm": 0.4178495453889158,
+      "learning_rate": 7.061591777574261e-08,
+      "loss": 0.0273,
+      "step": 8570
+    },
+    {
+      "epoch": 3.8067954696868753,
+      "grad_norm": 0.34370415504576896,
+      "learning_rate": 7.029163745744194e-08,
+      "loss": 0.0198,
+      "step": 8571
+    },
+    {
+      "epoch": 3.8072396180324226,
+      "grad_norm": 0.6738541750686097,
+      "learning_rate": 6.996809815804917e-08,
+      "loss": 0.0216,
+      "step": 8572
+    },
+    {
+      "epoch": 3.8076837663779703,
+      "grad_norm": 0.3104036826233125,
+      "learning_rate": 6.964529992619817e-08,
+      "loss": 0.0158,
+      "step": 8573
+    },
+    {
+      "epoch": 3.8081279147235176,
+      "grad_norm": 0.3875694214420725,
+      "learning_rate": 6.932324281041014e-08,
+      "loss": 0.0282,
+      "step": 8574
+    },
+    {
+      "epoch": 3.808572063069065,
+      "grad_norm": 0.7370247616254277,
+      "learning_rate": 6.900192685909635e-08,
+      "loss": 0.0427,
+      "step": 8575
+    },
+    {
+      "epoch": 3.8090162114146127,
+      "grad_norm": 0.3713027373129603,
+      "learning_rate": 6.868135212055649e-08,
+      "loss": 0.0273,
+      "step": 8576
+    },
+    {
+      "epoch": 3.80946035976016,
+      "grad_norm": 0.4118005174789893,
+      "learning_rate": 6.836151864297702e-08,
+      "loss": 0.0256,
+      "step": 8577
+    },
+    {
+      "epoch": 3.8099045081057072,
+      "grad_norm": 0.47276389861446366,
+      "learning_rate": 6.80424264744356e-08,
+      "loss": 0.0345,
+      "step": 8578
+    },
+    {
+      "epoch": 3.8103486564512545,
+      "grad_norm": 0.49450924911427574,
+      "learning_rate": 6.772407566289718e-08,
+      "loss": 0.0299,
+      "step": 8579
+    },
+    {
+      "epoch": 3.8107928047968023,
+      "grad_norm": 0.3914626594412097,
+      "learning_rate": 6.740646625621461e-08,
+      "loss": 0.0244,
+      "step": 8580
+    },
+    {
+      "epoch": 3.8112369531423496,
+      "grad_norm": 0.3930999395745575,
+      "learning_rate": 6.708959830213024e-08,
+      "loss": 0.0359,
+      "step": 8581
+    },
+    {
+      "epoch": 3.811681101487897,
+      "grad_norm": 0.49573525692180653,
+      "learning_rate": 6.677347184827487e-08,
+      "loss": 0.0319,
+      "step": 8582
+    },
+    {
+      "epoch": 3.8121252498334446,
+      "grad_norm": 0.49861370512502395,
+      "learning_rate": 6.645808694216715e-08,
+      "loss": 0.0256,
+      "step": 8583
+    },
+    {
+      "epoch": 3.812569398178992,
+      "grad_norm": 0.4401777666781036,
+      "learning_rate": 6.614344363121583e-08,
+      "loss": 0.0223,
+      "step": 8584
+    },
+    {
+      "epoch": 3.813013546524539,
+      "grad_norm": 0.36977940669747483,
+      "learning_rate": 6.582954196271641e-08,
+      "loss": 0.0214,
+      "step": 8585
+    },
+    {
+      "epoch": 3.8134576948700865,
+      "grad_norm": 0.4927485327079211,
+      "learning_rate": 6.55163819838528e-08,
+      "loss": 0.0287,
+      "step": 8586
+    },
+    {
+      "epoch": 3.8139018432156337,
+      "grad_norm": 0.37546001627014713,
+      "learning_rate": 6.520396374170013e-08,
+      "loss": 0.0229,
+      "step": 8587
+    },
+    {
+      "epoch": 3.8143459915611815,
+      "grad_norm": 0.4196889003684699,
+      "learning_rate": 6.489228728321917e-08,
+      "loss": 0.0276,
+      "step": 8588
+    },
+    {
+      "epoch": 3.8147901399067288,
+      "grad_norm": 0.4820349858082287,
+      "learning_rate": 6.458135265525967e-08,
+      "loss": 0.0274,
+      "step": 8589
+    },
+    {
+      "epoch": 3.8152342882522765,
+      "grad_norm": 0.3797443254174065,
+      "learning_rate": 6.427115990456201e-08,
+      "loss": 0.0237,
+      "step": 8590
+    },
+    {
+      "epoch": 3.815678436597824,
+      "grad_norm": 0.4390761052826011,
+      "learning_rate": 6.396170907775167e-08,
+      "loss": 0.03,
+      "step": 8591
+    },
+    {
+      "epoch": 3.816122584943371,
+      "grad_norm": 0.32997954286919473,
+      "learning_rate": 6.365300022134479e-08,
+      "loss": 0.0242,
+      "step": 8592
+    },
+    {
+      "epoch": 3.8165667332889184,
+      "grad_norm": 0.45875132604087615,
+      "learning_rate": 6.334503338174646e-08,
+      "loss": 0.0269,
+      "step": 8593
+    },
+    {
+      "epoch": 3.8170108816344657,
+      "grad_norm": 0.48148563808244804,
+      "learning_rate": 6.303780860524855e-08,
+      "loss": 0.0273,
+      "step": 8594
+    },
+    {
+      "epoch": 3.8174550299800134,
+      "grad_norm": 0.4353447524714429,
+      "learning_rate": 6.273132593803189e-08,
+      "loss": 0.0195,
+      "step": 8595
+    },
+    {
+      "epoch": 3.8178991783255607,
+      "grad_norm": 0.38567398492319127,
+      "learning_rate": 6.242558542616739e-08,
+      "loss": 0.0193,
+      "step": 8596
+    },
+    {
+      "epoch": 3.818343326671108,
+      "grad_norm": 0.4173237164922603,
+      "learning_rate": 6.212058711561165e-08,
+      "loss": 0.0278,
+      "step": 8597
+    },
+    {
+      "epoch": 3.8187874750166557,
+      "grad_norm": 0.3154210166282003,
+      "learning_rate": 6.18163310522113e-08,
+      "loss": 0.0189,
+      "step": 8598
+    },
+    {
+      "epoch": 3.819231623362203,
+      "grad_norm": 0.5134503050749415,
+      "learning_rate": 6.151281728170144e-08,
+      "loss": 0.0289,
+      "step": 8599
+    },
+    {
+      "epoch": 3.8196757717077503,
+      "grad_norm": 0.46640864477037924,
+      "learning_rate": 6.121004584970558e-08,
+      "loss": 0.0308,
+      "step": 8600
+    },
+    {
+      "epoch": 3.8201199200532976,
+      "grad_norm": 0.36566694826665086,
+      "learning_rate": 6.090801680173563e-08,
+      "loss": 0.022,
+      "step": 8601
+    },
+    {
+      "epoch": 3.8205640683988453,
+      "grad_norm": 0.39260388522181044,
+      "learning_rate": 6.060673018319085e-08,
+      "loss": 0.033,
+      "step": 8602
+    },
+    {
+      "epoch": 3.8210082167443926,
+      "grad_norm": 0.32628501291346773,
+      "learning_rate": 6.030618603935945e-08,
+      "loss": 0.0195,
+      "step": 8603
+    },
+    {
+      "epoch": 3.82145236508994,
+      "grad_norm": 0.5032358609244834,
+      "learning_rate": 6.000638441542029e-08,
+      "loss": 0.0367,
+      "step": 8604
+    },
+    {
+      "epoch": 3.8218965134354876,
+      "grad_norm": 0.49029558220343267,
+      "learning_rate": 5.970732535643675e-08,
+      "loss": 0.0357,
+      "step": 8605
+    },
+    {
+      "epoch": 3.822340661781035,
+      "grad_norm": 0.4041938487036102,
+      "learning_rate": 5.94090089073629e-08,
+      "loss": 0.0192,
+      "step": 8606
+    },
+    {
+      "epoch": 3.8227848101265822,
+      "grad_norm": 0.35662580614791967,
+      "learning_rate": 5.911143511304174e-08,
+      "loss": 0.0223,
+      "step": 8607
+    },
+    {
+      "epoch": 3.8232289584721295,
+      "grad_norm": 0.42334321858420404,
+      "learning_rate": 5.8814604018202494e-08,
+      "loss": 0.0333,
+      "step": 8608
+    },
+    {
+      "epoch": 3.8236731068176772,
+      "grad_norm": 0.3454601103621082,
+      "learning_rate": 5.851851566746392e-08,
+      "loss": 0.0184,
+      "step": 8609
+    },
+    {
+      "epoch": 3.8241172551632245,
+      "grad_norm": 0.39337857078220745,
+      "learning_rate": 5.8223170105333734e-08,
+      "loss": 0.0228,
+      "step": 8610
+    },
+    {
+      "epoch": 3.824561403508772,
+      "grad_norm": 0.3888924660914583,
+      "learning_rate": 5.792856737620756e-08,
+      "loss": 0.0222,
+      "step": 8611
+    },
+    {
+      "epoch": 3.8250055518543196,
+      "grad_norm": 0.3633339990171769,
+      "learning_rate": 5.763470752436884e-08,
+      "loss": 0.0178,
+      "step": 8612
+    },
+    {
+      "epoch": 3.825449700199867,
+      "grad_norm": 0.5069652365071539,
+      "learning_rate": 5.734159059398947e-08,
+      "loss": 0.0311,
+      "step": 8613
+    },
+    {
+      "epoch": 3.825893848545414,
+      "grad_norm": 0.36481775204860845,
+      "learning_rate": 5.7049216629129764e-08,
+      "loss": 0.0214,
+      "step": 8614
+    },
+    {
+      "epoch": 3.8263379968909614,
+      "grad_norm": 0.4066225715814007,
+      "learning_rate": 5.6757585673739014e-08,
+      "loss": 0.0196,
+      "step": 8615
+    },
+    {
+      "epoch": 3.8267821452365087,
+      "grad_norm": 0.38508300233088066,
+      "learning_rate": 5.6466697771654365e-08,
+      "loss": 0.0359,
+      "step": 8616
+    },
+    {
+      "epoch": 3.8272262935820565,
+      "grad_norm": 0.3857136200508219,
+      "learning_rate": 5.617655296660085e-08,
+      "loss": 0.0295,
+      "step": 8617
+    },
+    {
+      "epoch": 3.8276704419276038,
+      "grad_norm": 0.38361748458889544,
+      "learning_rate": 5.5887151302192465e-08,
+      "loss": 0.0225,
+      "step": 8618
+    },
+    {
+      "epoch": 3.8281145902731515,
+      "grad_norm": 0.4136696628077852,
+      "learning_rate": 5.5598492821931083e-08,
+      "loss": 0.0259,
+      "step": 8619
+    },
+    {
+      "epoch": 3.8285587386186988,
+      "grad_norm": 0.4157902458106096,
+      "learning_rate": 5.531057756920644e-08,
+      "loss": 0.0299,
+      "step": 8620
+    },
+    {
+      "epoch": 3.829002886964246,
+      "grad_norm": 0.41824752196070275,
+      "learning_rate": 5.502340558729835e-08,
+      "loss": 0.0276,
+      "step": 8621
+    },
+    {
+      "epoch": 3.8294470353097934,
+      "grad_norm": 0.3542732476447781,
+      "learning_rate": 5.4736976919372295e-08,
+      "loss": 0.0214,
+      "step": 8622
+    },
+    {
+      "epoch": 3.8298911836553406,
+      "grad_norm": 0.47614032589088634,
+      "learning_rate": 5.445129160848384e-08,
+      "loss": 0.0271,
+      "step": 8623
+    },
+    {
+      "epoch": 3.8303353320008884,
+      "grad_norm": 0.38508281181821147,
+      "learning_rate": 5.416634969757695e-08,
+      "loss": 0.0184,
+      "step": 8624
+    },
+    {
+      "epoch": 3.8307794803464357,
+      "grad_norm": 0.3910260501124829,
+      "learning_rate": 5.388215122948237e-08,
+      "loss": 0.0259,
+      "step": 8625
+    },
+    {
+      "epoch": 3.831223628691983,
+      "grad_norm": 0.31939094873163437,
+      "learning_rate": 5.359869624692038e-08,
+      "loss": 0.0197,
+      "step": 8626
+    },
+    {
+      "epoch": 3.8316677770375307,
+      "grad_norm": 0.30534186293474436,
+      "learning_rate": 5.331598479249911e-08,
+      "loss": 0.0201,
+      "step": 8627
+    },
+    {
+      "epoch": 3.832111925383078,
+      "grad_norm": 0.33277037895962464,
+      "learning_rate": 5.303401690871457e-08,
+      "loss": 0.0203,
+      "step": 8628
+    },
+    {
+      "epoch": 3.8325560737286253,
+      "grad_norm": 0.46652287015112265,
+      "learning_rate": 5.275279263795175e-08,
+      "loss": 0.0282,
+      "step": 8629
+    },
+    {
+      "epoch": 3.8330002220741726,
+      "grad_norm": 0.3848369302003942,
+      "learning_rate": 5.2472312022483486e-08,
+      "loss": 0.0258,
+      "step": 8630
+    },
+    {
+      "epoch": 3.8334443704197203,
+      "grad_norm": 0.4448218074585927,
+      "learning_rate": 5.2192575104469956e-08,
+      "loss": 0.0254,
+      "step": 8631
+    },
+    {
+      "epoch": 3.8338885187652676,
+      "grad_norm": 0.384660032948373,
+      "learning_rate": 5.1913581925960853e-08,
+      "loss": 0.0298,
+      "step": 8632
+    },
+    {
+      "epoch": 3.834332667110815,
+      "grad_norm": 0.4204621984497172,
+      "learning_rate": 5.16353325288943e-08,
+      "loss": 0.0242,
+      "step": 8633
+    },
+    {
+      "epoch": 3.8347768154563626,
+      "grad_norm": 0.46188483366801475,
+      "learning_rate": 5.135782695509461e-08,
+      "loss": 0.0231,
+      "step": 8634
+    },
+    {
+      "epoch": 3.83522096380191,
+      "grad_norm": 0.31629549729913453,
+      "learning_rate": 5.1081065246277314e-08,
+      "loss": 0.0158,
+      "step": 8635
+    },
+    {
+      "epoch": 3.835665112147457,
+      "grad_norm": 0.4353107611550551,
+      "learning_rate": 5.0805047444042467e-08,
+      "loss": 0.0287,
+      "step": 8636
+    },
+    {
+      "epoch": 3.8361092604930045,
+      "grad_norm": 0.40299033179334315,
+      "learning_rate": 5.0529773589881315e-08,
+      "loss": 0.0195,
+      "step": 8637
+    },
+    {
+      "epoch": 3.8365534088385522,
+      "grad_norm": 0.39085359021822447,
+      "learning_rate": 5.0255243725171876e-08,
+      "loss": 0.0279,
+      "step": 8638
+    },
+    {
+      "epoch": 3.8369975571840995,
+      "grad_norm": 0.48519972906519826,
+      "learning_rate": 4.998145789118114e-08,
+      "loss": 0.0333,
+      "step": 8639
+    },
+    {
+      "epoch": 3.837441705529647,
+      "grad_norm": 0.3517351658799901,
+      "learning_rate": 4.970841612906285e-08,
+      "loss": 0.0224,
+      "step": 8640
+    },
+    {
+      "epoch": 3.8378858538751945,
+      "grad_norm": 0.4190846054803923,
+      "learning_rate": 4.943611847986085e-08,
+      "loss": 0.0234,
+      "step": 8641
+    },
+    {
+      "epoch": 3.838330002220742,
+      "grad_norm": 0.35500926856677334,
+      "learning_rate": 4.9164564984505723e-08,
+      "loss": 0.0198,
+      "step": 8642
+    },
+    {
+      "epoch": 3.838774150566289,
+      "grad_norm": 0.34433663873966086,
+      "learning_rate": 4.889375568381594e-08,
+      "loss": 0.0174,
+      "step": 8643
+    },
+    {
+      "epoch": 3.8392182989118364,
+      "grad_norm": 0.38443528004632466,
+      "learning_rate": 4.8623690618499474e-08,
+      "loss": 0.0229,
+      "step": 8644
+    },
+    {
+      "epoch": 3.8396624472573837,
+      "grad_norm": 0.6274521829246321,
+      "learning_rate": 4.835436982915165e-08,
+      "loss": 0.0344,
+      "step": 8645
+    },
+    {
+      "epoch": 3.8401065956029314,
+      "grad_norm": 0.6266172275143739,
+      "learning_rate": 4.808579335625563e-08,
+      "loss": 0.0327,
+      "step": 8646
+    },
+    {
+      "epoch": 3.8405507439484787,
+      "grad_norm": 0.34058549807195543,
+      "learning_rate": 4.7817961240183567e-08,
+      "loss": 0.0224,
+      "step": 8647
+    },
+    {
+      "epoch": 3.8409948922940265,
+      "grad_norm": 0.4390835521149124,
+      "learning_rate": 4.7550873521194364e-08,
+      "loss": 0.0195,
+      "step": 8648
+    },
+    {
+      "epoch": 3.8414390406395738,
+      "grad_norm": 0.3842688631078936,
+      "learning_rate": 4.728453023943591e-08,
+      "loss": 0.0231,
+      "step": 8649
+    },
+    {
+      "epoch": 3.841883188985121,
+      "grad_norm": 0.4692045722646363,
+      "learning_rate": 4.701893143494507e-08,
+      "loss": 0.0324,
+      "step": 8650
+    },
+    {
+      "epoch": 3.8423273373306683,
+      "grad_norm": 0.45747006524072764,
+      "learning_rate": 4.675407714764491e-08,
+      "loss": 0.0308,
+      "step": 8651
+    },
+    {
+      "epoch": 3.8427714856762156,
+      "grad_norm": 0.3514460328352755,
+      "learning_rate": 4.648996741734857e-08,
+      "loss": 0.0207,
+      "step": 8652
+    },
+    {
+      "epoch": 3.8432156340217634,
+      "grad_norm": 0.3912542737417533,
+      "learning_rate": 4.622660228375486e-08,
+      "loss": 0.0219,
+      "step": 8653
+    },
+    {
+      "epoch": 3.8436597823673107,
+      "grad_norm": 0.3136962524397564,
+      "learning_rate": 4.596398178645323e-08,
+      "loss": 0.0178,
+      "step": 8654
+    },
+    {
+      "epoch": 3.844103930712858,
+      "grad_norm": 0.35110936427287714,
+      "learning_rate": 4.5702105964919305e-08,
+      "loss": 0.0163,
+      "step": 8655
+    },
+    {
+      "epoch": 3.8445480790584057,
+      "grad_norm": 0.3924180108841284,
+      "learning_rate": 4.5440974858517174e-08,
+      "loss": 0.0185,
+      "step": 8656
+    },
+    {
+      "epoch": 3.844992227403953,
+      "grad_norm": 0.41446900223517336,
+      "learning_rate": 4.5180588506500424e-08,
+      "loss": 0.0294,
+      "step": 8657
+    },
+    {
+      "epoch": 3.8454363757495003,
+      "grad_norm": 0.3695057701384476,
+      "learning_rate": 4.492094694800886e-08,
+      "loss": 0.0269,
+      "step": 8658
+    },
+    {
+      "epoch": 3.8458805240950475,
+      "grad_norm": 0.3819021189729295,
+      "learning_rate": 4.4662050222070707e-08,
+      "loss": 0.0242,
+      "step": 8659
+    },
+    {
+      "epoch": 3.8463246724405953,
+      "grad_norm": 0.4011303876812642,
+      "learning_rate": 4.440389836760317e-08,
+      "loss": 0.028,
+      "step": 8660
+    },
+    {
+      "epoch": 3.8467688207861426,
+      "grad_norm": 0.42320733559927765,
+      "learning_rate": 4.414649142341021e-08,
+      "loss": 0.023,
+      "step": 8661
+    },
+    {
+      "epoch": 3.84721296913169,
+      "grad_norm": 0.4205022645129385,
+      "learning_rate": 4.388982942818476e-08,
+      "loss": 0.0261,
+      "step": 8662
+    },
+    {
+      "epoch": 3.8476571174772376,
+      "grad_norm": 0.3312900714579485,
+      "learning_rate": 4.363391242050819e-08,
+      "loss": 0.0191,
+      "step": 8663
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.40420539853558396,
+      "learning_rate": 4.3378740438848045e-08,
+      "loss": 0.0317,
+      "step": 8664
+    },
+    {
+      "epoch": 3.848545414168332,
+      "grad_norm": 0.41550220010293626,
+      "learning_rate": 4.312431352156143e-08,
+      "loss": 0.0255,
+      "step": 8665
+    },
+    {
+      "epoch": 3.8489895625138795,
+      "grad_norm": 0.34141480859605966,
+      "learning_rate": 4.287063170689332e-08,
+      "loss": 0.0219,
+      "step": 8666
+    },
+    {
+      "epoch": 3.849433710859427,
+      "grad_norm": 0.38457339309127786,
+      "learning_rate": 4.261769503297597e-08,
+      "loss": 0.0241,
+      "step": 8667
+    },
+    {
+      "epoch": 3.8498778592049745,
+      "grad_norm": 0.3865401623267944,
+      "learning_rate": 4.236550353783009e-08,
+      "loss": 0.0177,
+      "step": 8668
+    },
+    {
+      "epoch": 3.850322007550522,
+      "grad_norm": 0.40767308229409815,
+      "learning_rate": 4.211405725936535e-08,
+      "loss": 0.0255,
+      "step": 8669
+    },
+    {
+      "epoch": 3.8507661558960695,
+      "grad_norm": 0.3151055767858847,
+      "learning_rate": 4.186335623537707e-08,
+      "loss": 0.0233,
+      "step": 8670
+    },
+    {
+      "epoch": 3.851210304241617,
+      "grad_norm": 0.3928367577202992,
+      "learning_rate": 4.1613400503550114e-08,
+      "loss": 0.0212,
+      "step": 8671
+    },
+    {
+      "epoch": 3.851654452587164,
+      "grad_norm": 0.4221600614877391,
+      "learning_rate": 4.13641901014572e-08,
+      "loss": 0.034,
+      "step": 8672
+    },
+    {
+      "epoch": 3.8520986009327114,
+      "grad_norm": 0.3485554952889989,
+      "learning_rate": 4.1115725066559476e-08,
+      "loss": 0.0189,
+      "step": 8673
+    },
+    {
+      "epoch": 3.8525427492782587,
+      "grad_norm": 0.4184669816003682,
+      "learning_rate": 4.086800543620484e-08,
+      "loss": 0.0265,
+      "step": 8674
+    },
+    {
+      "epoch": 3.8529868976238064,
+      "grad_norm": 0.40633919115050987,
+      "learning_rate": 4.062103124763017e-08,
+      "loss": 0.0178,
+      "step": 8675
+    },
+    {
+      "epoch": 3.8534310459693537,
+      "grad_norm": 0.5255893189755798,
+      "learning_rate": 4.0374802537959114e-08,
+      "loss": 0.0255,
+      "step": 8676
+    },
+    {
+      "epoch": 3.8538751943149014,
+      "grad_norm": 0.4396123179486424,
+      "learning_rate": 4.012931934420483e-08,
+      "loss": 0.0317,
+      "step": 8677
+    },
+    {
+      "epoch": 3.8543193426604487,
+      "grad_norm": 0.3657385642630762,
+      "learning_rate": 3.9884581703267254e-08,
+      "loss": 0.022,
+      "step": 8678
+    },
+    {
+      "epoch": 3.854763491005996,
+      "grad_norm": 0.34626474417977715,
+      "learning_rate": 3.964058965193473e-08,
+      "loss": 0.022,
+      "step": 8679
+    },
+    {
+      "epoch": 3.8552076393515433,
+      "grad_norm": 0.38562426634481894,
+      "learning_rate": 3.939734322688349e-08,
+      "loss": 0.0338,
+      "step": 8680
+    },
+    {
+      "epoch": 3.8556517876970906,
+      "grad_norm": 0.4035987782299242,
+      "learning_rate": 3.9154842464677045e-08,
+      "loss": 0.0236,
+      "step": 8681
+    },
+    {
+      "epoch": 3.8560959360426383,
+      "grad_norm": 0.4159647047153014,
+      "learning_rate": 3.8913087401767914e-08,
+      "loss": 0.024,
+      "step": 8682
+    },
+    {
+      "epoch": 3.8565400843881856,
+      "grad_norm": 0.5090881621984651,
+      "learning_rate": 3.867207807449591e-08,
+      "loss": 0.03,
+      "step": 8683
+    },
+    {
+      "epoch": 3.856984232733733,
+      "grad_norm": 0.3724391004939136,
+      "learning_rate": 3.843181451908928e-08,
+      "loss": 0.0222,
+      "step": 8684
+    },
+    {
+      "epoch": 3.8574283810792807,
+      "grad_norm": 0.3994107927570375,
+      "learning_rate": 3.8192296771663026e-08,
+      "loss": 0.0365,
+      "step": 8685
+    },
+    {
+      "epoch": 3.857872529424828,
+      "grad_norm": 0.534388923736043,
+      "learning_rate": 3.795352486822057e-08,
+      "loss": 0.0286,
+      "step": 8686
+    },
+    {
+      "epoch": 3.8583166777703752,
+      "grad_norm": 0.3902311138190663,
+      "learning_rate": 3.7715498844653755e-08,
+      "loss": 0.0332,
+      "step": 8687
+    },
+    {
+      "epoch": 3.8587608261159225,
+      "grad_norm": 0.6121843349896643,
+      "learning_rate": 3.7478218736742286e-08,
+      "loss": 0.0354,
+      "step": 8688
+    },
+    {
+      "epoch": 3.8592049744614703,
+      "grad_norm": 0.3370534483485381,
+      "learning_rate": 3.724168458015265e-08,
+      "loss": 0.0221,
+      "step": 8689
+    },
+    {
+      "epoch": 3.8596491228070176,
+      "grad_norm": 0.450573063405829,
+      "learning_rate": 3.700589641044083e-08,
+      "loss": 0.0225,
+      "step": 8690
+    },
+    {
+      "epoch": 3.860093271152565,
+      "grad_norm": 0.49278309629574163,
+      "learning_rate": 3.677085426304905e-08,
+      "loss": 0.029,
+      "step": 8691
+    },
+    {
+      "epoch": 3.8605374194981126,
+      "grad_norm": 0.4055229539649227,
+      "learning_rate": 3.6536558173308476e-08,
+      "loss": 0.025,
+      "step": 8692
+    },
+    {
+      "epoch": 3.86098156784366,
+      "grad_norm": 0.3771030172546445,
+      "learning_rate": 3.630300817643762e-08,
+      "loss": 0.0196,
+      "step": 8693
+    },
+    {
+      "epoch": 3.861425716189207,
+      "grad_norm": 0.4460409041626926,
+      "learning_rate": 3.607020430754338e-08,
+      "loss": 0.0263,
+      "step": 8694
+    },
+    {
+      "epoch": 3.8618698645347544,
+      "grad_norm": 0.36455946204318906,
+      "learning_rate": 3.583814660161944e-08,
+      "loss": 0.0254,
+      "step": 8695
+    },
+    {
+      "epoch": 3.862314012880302,
+      "grad_norm": 0.45869268016899095,
+      "learning_rate": 3.5606835093548456e-08,
+      "loss": 0.0364,
+      "step": 8696
+    },
+    {
+      "epoch": 3.8627581612258495,
+      "grad_norm": 0.3785302300726189,
+      "learning_rate": 3.537626981810094e-08,
+      "loss": 0.0203,
+      "step": 8697
+    },
+    {
+      "epoch": 3.8632023095713968,
+      "grad_norm": 0.41102132869975794,
+      "learning_rate": 3.514645080993362e-08,
+      "loss": 0.0304,
+      "step": 8698
+    },
+    {
+      "epoch": 3.8636464579169445,
+      "grad_norm": 0.3842024608267308,
+      "learning_rate": 3.49173781035933e-08,
+      "loss": 0.0217,
+      "step": 8699
+    },
+    {
+      "epoch": 3.864090606262492,
+      "grad_norm": 0.5453458258267285,
+      "learning_rate": 3.4689051733513e-08,
+      "loss": 0.0423,
+      "step": 8700
+    },
+    {
+      "epoch": 3.864534754608039,
+      "grad_norm": 0.4282646993236422,
+      "learning_rate": 3.446147173401415e-08,
+      "loss": 0.0283,
+      "step": 8701
+    },
+    {
+      "epoch": 3.8649789029535864,
+      "grad_norm": 0.3536636171621975,
+      "learning_rate": 3.4234638139306055e-08,
+      "loss": 0.0239,
+      "step": 8702
+    },
+    {
+      "epoch": 3.8654230512991337,
+      "grad_norm": 0.32067352971181584,
+      "learning_rate": 3.4008550983484766e-08,
+      "loss": 0.019,
+      "step": 8703
+    },
+    {
+      "epoch": 3.8658671996446814,
+      "grad_norm": 0.42623278208292326,
+      "learning_rate": 3.378321030053644e-08,
+      "loss": 0.0235,
+      "step": 8704
+    },
+    {
+      "epoch": 3.8663113479902287,
+      "grad_norm": 0.4469975288314765,
+      "learning_rate": 3.355861612433231e-08,
+      "loss": 0.0309,
+      "step": 8705
+    },
+    {
+      "epoch": 3.8667554963357764,
+      "grad_norm": 0.4959636073318917,
+      "learning_rate": 3.3334768488633706e-08,
+      "loss": 0.0284,
+      "step": 8706
+    },
+    {
+      "epoch": 3.8671996446813237,
+      "grad_norm": 0.503087622712729,
+      "learning_rate": 3.31116674270876e-08,
+      "loss": 0.0256,
+      "step": 8707
+    },
+    {
+      "epoch": 3.867643793026871,
+      "grad_norm": 0.36447403400661404,
+      "learning_rate": 3.2889312973231616e-08,
+      "loss": 0.0187,
+      "step": 8708
+    },
+    {
+      "epoch": 3.8680879413724183,
+      "grad_norm": 0.39131073302761393,
+      "learning_rate": 3.266770516048734e-08,
+      "loss": 0.0227,
+      "step": 8709
+    },
+    {
+      "epoch": 3.8685320897179656,
+      "grad_norm": 0.39335266998944496,
+      "learning_rate": 3.2446844022167576e-08,
+      "loss": 0.0302,
+      "step": 8710
+    },
+    {
+      "epoch": 3.8689762380635133,
+      "grad_norm": 0.3587677625718898,
+      "learning_rate": 3.2226729591471326e-08,
+      "loss": 0.0196,
+      "step": 8711
+    },
+    {
+      "epoch": 3.8694203864090606,
+      "grad_norm": 0.360223764784636,
+      "learning_rate": 3.2007361901485455e-08,
+      "loss": 0.0202,
+      "step": 8712
+    },
+    {
+      "epoch": 3.869864534754608,
+      "grad_norm": 0.45416360753066903,
+      "learning_rate": 3.1788740985184144e-08,
+      "loss": 0.0296,
+      "step": 8713
+    },
+    {
+      "epoch": 3.8703086831001556,
+      "grad_norm": 0.4540255704066019,
+      "learning_rate": 3.1570866875430536e-08,
+      "loss": 0.03,
+      "step": 8714
+    },
+    {
+      "epoch": 3.870752831445703,
+      "grad_norm": 0.39674557183309156,
+      "learning_rate": 3.135373960497401e-08,
+      "loss": 0.0246,
+      "step": 8715
+    },
+    {
+      "epoch": 3.87119697979125,
+      "grad_norm": 0.4615034431538654,
+      "learning_rate": 3.113735920645344e-08,
+      "loss": 0.0209,
+      "step": 8716
+    },
+    {
+      "epoch": 3.8716411281367975,
+      "grad_norm": 0.40129516372174695,
+      "learning_rate": 3.092172571239338e-08,
+      "loss": 0.0239,
+      "step": 8717
+    },
+    {
+      "epoch": 3.8720852764823452,
+      "grad_norm": 0.3795932466717671,
+      "learning_rate": 3.070683915520845e-08,
+      "loss": 0.0275,
+      "step": 8718
+    },
+    {
+      "epoch": 3.8725294248278925,
+      "grad_norm": 0.5210559447705239,
+      "learning_rate": 3.049269956719891e-08,
+      "loss": 0.0261,
+      "step": 8719
+    },
+    {
+      "epoch": 3.87297357317344,
+      "grad_norm": 0.4437613175714331,
+      "learning_rate": 3.0279306980554034e-08,
+      "loss": 0.0295,
+      "step": 8720
+    },
+    {
+      "epoch": 3.8734177215189876,
+      "grad_norm": 0.3896516822182538,
+      "learning_rate": 3.006666142734982e-08,
+      "loss": 0.0246,
+      "step": 8721
+    },
+    {
+      "epoch": 3.873861869864535,
+      "grad_norm": 0.43115789241542724,
+      "learning_rate": 2.9854762939551254e-08,
+      "loss": 0.0314,
+      "step": 8722
+    },
+    {
+      "epoch": 3.874306018210082,
+      "grad_norm": 0.34436582503878505,
+      "learning_rate": 2.9643611549008967e-08,
+      "loss": 0.0219,
+      "step": 8723
+    },
+    {
+      "epoch": 3.8747501665556294,
+      "grad_norm": 0.430010546023126,
+      "learning_rate": 2.9433207287464238e-08,
+      "loss": 0.0267,
+      "step": 8724
+    },
+    {
+      "epoch": 3.8751943149011767,
+      "grad_norm": 0.28855368882245647,
+      "learning_rate": 2.9223550186543435e-08,
+      "loss": 0.0137,
+      "step": 8725
+    },
+    {
+      "epoch": 3.8756384632467245,
+      "grad_norm": 0.3534494715975773,
+      "learning_rate": 2.9014640277761353e-08,
+      "loss": 0.0274,
+      "step": 8726
+    },
+    {
+      "epoch": 3.8760826115922717,
+      "grad_norm": 0.3093760532300058,
+      "learning_rate": 2.8806477592521755e-08,
+      "loss": 0.015,
+      "step": 8727
+    },
+    {
+      "epoch": 3.8765267599378195,
+      "grad_norm": 0.6497895297842237,
+      "learning_rate": 2.8599062162114056e-08,
+      "loss": 0.037,
+      "step": 8728
+    },
+    {
+      "epoch": 3.8769709082833668,
+      "grad_norm": 0.5014893102635929,
+      "learning_rate": 2.8392394017716095e-08,
+      "loss": 0.0344,
+      "step": 8729
+    },
+    {
+      "epoch": 3.877415056628914,
+      "grad_norm": 0.3936192272437833,
+      "learning_rate": 2.8186473190395246e-08,
+      "loss": 0.0303,
+      "step": 8730
+    },
+    {
+      "epoch": 3.8778592049744613,
+      "grad_norm": 0.4010118354626057,
+      "learning_rate": 2.798129971110286e-08,
+      "loss": 0.0307,
+      "step": 8731
+    },
+    {
+      "epoch": 3.8783033533200086,
+      "grad_norm": 0.3784503093468852,
+      "learning_rate": 2.7776873610681486e-08,
+      "loss": 0.02,
+      "step": 8732
+    },
+    {
+      "epoch": 3.8787475016655564,
+      "grad_norm": 0.4230263125616594,
+      "learning_rate": 2.7573194919859325e-08,
+      "loss": 0.0224,
+      "step": 8733
+    },
+    {
+      "epoch": 3.8791916500111037,
+      "grad_norm": 0.4140738675854264,
+      "learning_rate": 2.737026366925244e-08,
+      "loss": 0.0258,
+      "step": 8734
+    },
+    {
+      "epoch": 3.879635798356651,
+      "grad_norm": 0.42420423811982005,
+      "learning_rate": 2.716807988936532e-08,
+      "loss": 0.022,
+      "step": 8735
+    },
+    {
+      "epoch": 3.8800799467021987,
+      "grad_norm": 0.3980720494346127,
+      "learning_rate": 2.696664361058976e-08,
+      "loss": 0.0269,
+      "step": 8736
+    },
+    {
+      "epoch": 3.880524095047746,
+      "grad_norm": 0.3946254166839832,
+      "learning_rate": 2.6765954863204323e-08,
+      "loss": 0.0288,
+      "step": 8737
+    },
+    {
+      "epoch": 3.8809682433932933,
+      "grad_norm": 0.3788227216815116,
+      "learning_rate": 2.6566013677376545e-08,
+      "loss": 0.0238,
+      "step": 8738
+    },
+    {
+      "epoch": 3.8814123917388406,
+      "grad_norm": 0.34261939630726157,
+      "learning_rate": 2.6366820083160715e-08,
+      "loss": 0.0145,
+      "step": 8739
+    },
+    {
+      "epoch": 3.8818565400843883,
+      "grad_norm": 0.42749605488948866,
+      "learning_rate": 2.6168374110498995e-08,
+      "loss": 0.0257,
+      "step": 8740
+    },
+    {
+      "epoch": 3.8823006884299356,
+      "grad_norm": 0.4042334016798025,
+      "learning_rate": 2.5970675789220855e-08,
+      "loss": 0.0186,
+      "step": 8741
+    },
+    {
+      "epoch": 3.882744836775483,
+      "grad_norm": 0.39831255423745965,
+      "learning_rate": 2.577372514904475e-08,
+      "loss": 0.0209,
+      "step": 8742
+    },
+    {
+      "epoch": 3.8831889851210306,
+      "grad_norm": 0.42360207246789,
+      "learning_rate": 2.5577522219575324e-08,
+      "loss": 0.0275,
+      "step": 8743
+    },
+    {
+      "epoch": 3.883633133466578,
+      "grad_norm": 0.48707455874974,
+      "learning_rate": 2.5382067030304546e-08,
+      "loss": 0.0282,
+      "step": 8744
+    },
+    {
+      "epoch": 3.884077281812125,
+      "grad_norm": 0.4723400265193081,
+      "learning_rate": 2.5187359610612805e-08,
+      "loss": 0.0382,
+      "step": 8745
+    },
+    {
+      "epoch": 3.8845214301576725,
+      "grad_norm": 0.6222325293211652,
+      "learning_rate": 2.499339998976835e-08,
+      "loss": 0.0342,
+      "step": 8746
+    },
+    {
+      "epoch": 3.88496557850322,
+      "grad_norm": 0.35612361294226674,
+      "learning_rate": 2.4800188196926757e-08,
+      "loss": 0.0183,
+      "step": 8747
+    },
+    {
+      "epoch": 3.8854097268487675,
+      "grad_norm": 0.42945931063185067,
+      "learning_rate": 2.4607724261130893e-08,
+      "loss": 0.0249,
+      "step": 8748
+    },
+    {
+      "epoch": 3.885853875194315,
+      "grad_norm": 0.3845905954635827,
+      "learning_rate": 2.441600821131096e-08,
+      "loss": 0.0219,
+      "step": 8749
+    },
+    {
+      "epoch": 3.8862980235398625,
+      "grad_norm": 0.3836148595852599,
+      "learning_rate": 2.422504007628501e-08,
+      "loss": 0.0207,
+      "step": 8750
+    },
+    {
+      "epoch": 3.88674217188541,
+      "grad_norm": 0.3404491277084874,
+      "learning_rate": 2.4034819884759532e-08,
+      "loss": 0.0198,
+      "step": 8751
+    },
+    {
+      "epoch": 3.887186320230957,
+      "grad_norm": 0.3619930083510835,
+      "learning_rate": 2.3845347665327202e-08,
+      "loss": 0.0191,
+      "step": 8752
+    },
+    {
+      "epoch": 3.8876304685765044,
+      "grad_norm": 0.4286874704038195,
+      "learning_rate": 2.3656623446469684e-08,
+      "loss": 0.0299,
+      "step": 8753
+    },
+    {
+      "epoch": 3.8880746169220517,
+      "grad_norm": 0.3418186033416821,
+      "learning_rate": 2.3468647256554845e-08,
+      "loss": 0.0201,
+      "step": 8754
+    },
+    {
+      "epoch": 3.8885187652675994,
+      "grad_norm": 0.4273211578376309,
+      "learning_rate": 2.3281419123838966e-08,
+      "loss": 0.0254,
+      "step": 8755
+    },
+    {
+      "epoch": 3.8889629136131467,
+      "grad_norm": 0.46117222518267137,
+      "learning_rate": 2.3094939076465095e-08,
+      "loss": 0.0235,
+      "step": 8756
+    },
+    {
+      "epoch": 3.8894070619586945,
+      "grad_norm": 0.30961316233144026,
+      "learning_rate": 2.2909207142464695e-08,
+      "loss": 0.0178,
+      "step": 8757
+    },
+    {
+      "epoch": 3.8898512103042417,
+      "grad_norm": 0.4313089227592287,
+      "learning_rate": 2.2724223349756547e-08,
+      "loss": 0.0334,
+      "step": 8758
+    },
+    {
+      "epoch": 3.890295358649789,
+      "grad_norm": 0.45892558729550764,
+      "learning_rate": 2.253998772614674e-08,
+      "loss": 0.0358,
+      "step": 8759
+    },
+    {
+      "epoch": 3.8907395069953363,
+      "grad_norm": 0.4320146677307238,
+      "learning_rate": 2.235650029932923e-08,
+      "loss": 0.0275,
+      "step": 8760
+    },
+    {
+      "epoch": 3.8911836553408836,
+      "grad_norm": 0.47469416867571373,
+      "learning_rate": 2.2173761096884737e-08,
+      "loss": 0.0313,
+      "step": 8761
+    },
+    {
+      "epoch": 3.8916278036864314,
+      "grad_norm": 0.38809247080958487,
+      "learning_rate": 2.1991770146282953e-08,
+      "loss": 0.0254,
+      "step": 8762
+    },
+    {
+      "epoch": 3.8920719520319786,
+      "grad_norm": 0.36883556562596115,
+      "learning_rate": 2.181052747487922e-08,
+      "loss": 0.0223,
+      "step": 8763
+    },
+    {
+      "epoch": 3.892516100377526,
+      "grad_norm": 0.4140272360800753,
+      "learning_rate": 2.1630033109918403e-08,
+      "loss": 0.0305,
+      "step": 8764
+    },
+    {
+      "epoch": 3.8929602487230737,
+      "grad_norm": 0.4360012174993987,
+      "learning_rate": 2.1450287078531028e-08,
+      "loss": 0.0247,
+      "step": 8765
+    },
+    {
+      "epoch": 3.893404397068621,
+      "grad_norm": 0.46749106481514396,
+      "learning_rate": 2.127128940773604e-08,
+      "loss": 0.0336,
+      "step": 8766
+    },
+    {
+      "epoch": 3.8938485454141682,
+      "grad_norm": 0.48628664662774723,
+      "learning_rate": 2.1093040124440246e-08,
+      "loss": 0.0364,
+      "step": 8767
+    },
+    {
+      "epoch": 3.8942926937597155,
+      "grad_norm": 0.4655918973858269,
+      "learning_rate": 2.091553925543721e-08,
+      "loss": 0.029,
+      "step": 8768
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.4385695873892939,
+      "learning_rate": 2.073878682740893e-08,
+      "loss": 0.0332,
+      "step": 8769
+    },
+    {
+      "epoch": 3.8951809904508106,
+      "grad_norm": 0.42920683680763916,
+      "learning_rate": 2.056278286692359e-08,
+      "loss": 0.0234,
+      "step": 8770
+    },
+    {
+      "epoch": 3.895625138796358,
+      "grad_norm": 0.44077294633964265,
+      "learning_rate": 2.0387527400437812e-08,
+      "loss": 0.0171,
+      "step": 8771
+    },
+    {
+      "epoch": 3.8960692871419056,
+      "grad_norm": 0.4448703366882169,
+      "learning_rate": 2.0213020454295517e-08,
+      "loss": 0.0241,
+      "step": 8772
+    },
+    {
+      "epoch": 3.896513435487453,
+      "grad_norm": 0.3962409747625572,
+      "learning_rate": 2.003926205472795e-08,
+      "loss": 0.0303,
+      "step": 8773
+    },
+    {
+      "epoch": 3.896957583833,
+      "grad_norm": 0.45182968264253826,
+      "learning_rate": 1.986625222785421e-08,
+      "loss": 0.0517,
+      "step": 8774
+    },
+    {
+      "epoch": 3.8974017321785475,
+      "grad_norm": 0.3768583215614216,
+      "learning_rate": 1.9693990999680167e-08,
+      "loss": 0.0243,
+      "step": 8775
+    },
+    {
+      "epoch": 3.897845880524095,
+      "grad_norm": 0.37480588059892506,
+      "learning_rate": 1.952247839610011e-08,
+      "loss": 0.0232,
+      "step": 8776
+    },
+    {
+      "epoch": 3.8982900288696425,
+      "grad_norm": 0.43502480214398537,
+      "learning_rate": 1.9351714442895077e-08,
+      "loss": 0.0243,
+      "step": 8777
+    },
+    {
+      "epoch": 3.8987341772151898,
+      "grad_norm": 0.33331146266153816,
+      "learning_rate": 1.918169916573398e-08,
+      "loss": 0.0223,
+      "step": 8778
+    },
+    {
+      "epoch": 3.8991783255607375,
+      "grad_norm": 0.3283297894096133,
+      "learning_rate": 1.9012432590172493e-08,
+      "loss": 0.0182,
+      "step": 8779
+    },
+    {
+      "epoch": 3.899622473906285,
+      "grad_norm": 0.4613996841914982,
+      "learning_rate": 1.8843914741654146e-08,
+      "loss": 0.0264,
+      "step": 8780
+    },
+    {
+      "epoch": 3.900066622251832,
+      "grad_norm": 0.3728645160784641,
+      "learning_rate": 1.8676145645511456e-08,
+      "loss": 0.0219,
+      "step": 8781
+    },
+    {
+      "epoch": 3.9005107705973794,
+      "grad_norm": 0.367364355087082,
+      "learning_rate": 1.850912532696092e-08,
+      "loss": 0.0169,
+      "step": 8782
+    },
+    {
+      "epoch": 3.9009549189429267,
+      "grad_norm": 0.43782054130287257,
+      "learning_rate": 1.8342853811110227e-08,
+      "loss": 0.0244,
+      "step": 8783
+    },
+    {
+      "epoch": 3.9013990672884744,
+      "grad_norm": 0.46621513501531153,
+      "learning_rate": 1.817733112295217e-08,
+      "loss": 0.0237,
+      "step": 8784
+    },
+    {
+      "epoch": 3.9018432156340217,
+      "grad_norm": 0.3948393185866851,
+      "learning_rate": 1.8012557287367394e-08,
+      "loss": 0.0197,
+      "step": 8785
+    },
+    {
+      "epoch": 3.9022873639795694,
+      "grad_norm": 0.5251739552887689,
+      "learning_rate": 1.7848532329124978e-08,
+      "loss": 0.0376,
+      "step": 8786
+    },
+    {
+      "epoch": 3.9027315123251167,
+      "grad_norm": 0.3262674897809267,
+      "learning_rate": 1.7685256272879646e-08,
+      "loss": 0.0189,
+      "step": 8787
+    },
+    {
+      "epoch": 3.903175660670664,
+      "grad_norm": 0.32400288553912415,
+      "learning_rate": 1.7522729143174545e-08,
+      "loss": 0.0185,
+      "step": 8788
+    },
+    {
+      "epoch": 3.9036198090162113,
+      "grad_norm": 0.36601512988828766,
+      "learning_rate": 1.7360950964441236e-08,
+      "loss": 0.0196,
+      "step": 8789
+    },
+    {
+      "epoch": 3.9040639573617586,
+      "grad_norm": 0.5081663386558348,
+      "learning_rate": 1.7199921760997494e-08,
+      "loss": 0.0312,
+      "step": 8790
+    },
+    {
+      "epoch": 3.9045081057073063,
+      "grad_norm": 0.717337078541349,
+      "learning_rate": 1.7039641557048402e-08,
+      "loss": 0.035,
+      "step": 8791
+    },
+    {
+      "epoch": 3.9049522540528536,
+      "grad_norm": 0.44298957493772145,
+      "learning_rate": 1.6880110376686353e-08,
+      "loss": 0.0368,
+      "step": 8792
+    },
+    {
+      "epoch": 3.905396402398401,
+      "grad_norm": 0.34392561304252184,
+      "learning_rate": 1.672132824389272e-08,
+      "loss": 0.0178,
+      "step": 8793
+    },
+    {
+      "epoch": 3.9058405507439486,
+      "grad_norm": 0.42295010360189583,
+      "learning_rate": 1.6563295182534524e-08,
+      "loss": 0.0251,
+      "step": 8794
+    },
+    {
+      "epoch": 3.906284699089496,
+      "grad_norm": 0.3790549559366509,
+      "learning_rate": 1.6406011216366647e-08,
+      "loss": 0.0235,
+      "step": 8795
+    },
+    {
+      "epoch": 3.9067288474350432,
+      "grad_norm": 0.41052130644928436,
+      "learning_rate": 1.6249476369031845e-08,
+      "loss": 0.0186,
+      "step": 8796
+    },
+    {
+      "epoch": 3.9071729957805905,
+      "grad_norm": 0.4367038611052701,
+      "learning_rate": 1.6093690664059635e-08,
+      "loss": 0.0339,
+      "step": 8797
+    },
+    {
+      "epoch": 3.9076171441261383,
+      "grad_norm": 0.37822957010835845,
+      "learning_rate": 1.5938654124867394e-08,
+      "loss": 0.0215,
+      "step": 8798
+    },
+    {
+      "epoch": 3.9080612924716855,
+      "grad_norm": 0.3875394125371318,
+      "learning_rate": 1.5784366774760362e-08,
+      "loss": 0.0217,
+      "step": 8799
+    },
+    {
+      "epoch": 3.908505440817233,
+      "grad_norm": 0.4043786874203682,
+      "learning_rate": 1.563082863692944e-08,
+      "loss": 0.0236,
+      "step": 8800
+    },
+    {
+      "epoch": 3.9089495891627806,
+      "grad_norm": 0.4592008857137564,
+      "learning_rate": 1.5478039734455053e-08,
+      "loss": 0.0229,
+      "step": 8801
+    },
+    {
+      "epoch": 3.909393737508328,
+      "grad_norm": 0.29589640326829375,
+      "learning_rate": 1.5326000090303272e-08,
+      "loss": 0.0162,
+      "step": 8802
+    },
+    {
+      "epoch": 3.909837885853875,
+      "grad_norm": 0.42576417588302456,
+      "learning_rate": 1.5174709727328595e-08,
+      "loss": 0.0359,
+      "step": 8803
+    },
+    {
+      "epoch": 3.9102820341994224,
+      "grad_norm": 0.4093180434069703,
+      "learning_rate": 1.5024168668272275e-08,
+      "loss": 0.0264,
+      "step": 8804
+    },
+    {
+      "epoch": 3.91072618254497,
+      "grad_norm": 0.46276973652514886,
+      "learning_rate": 1.4874376935763434e-08,
+      "loss": 0.0289,
+      "step": 8805
+    },
+    {
+      "epoch": 3.9111703308905175,
+      "grad_norm": 0.4688883489908628,
+      "learning_rate": 1.4725334552318504e-08,
+      "loss": 0.0181,
+      "step": 8806
+    },
+    {
+      "epoch": 3.9116144792360648,
+      "grad_norm": 0.3039807302776599,
+      "learning_rate": 1.4577041540340676e-08,
+      "loss": 0.015,
+      "step": 8807
+    },
+    {
+      "epoch": 3.9120586275816125,
+      "grad_norm": 0.3741536582807147,
+      "learning_rate": 1.442949792212045e-08,
+      "loss": 0.0235,
+      "step": 8808
+    },
+    {
+      "epoch": 3.91250277592716,
+      "grad_norm": 0.5513005564638584,
+      "learning_rate": 1.428270371983731e-08,
+      "loss": 0.0238,
+      "step": 8809
+    },
+    {
+      "epoch": 3.912946924272707,
+      "grad_norm": 0.37263633834750676,
+      "learning_rate": 1.4136658955556381e-08,
+      "loss": 0.0249,
+      "step": 8810
+    },
+    {
+      "epoch": 3.9133910726182544,
+      "grad_norm": 0.36432340872278923,
+      "learning_rate": 1.3991363651230106e-08,
+      "loss": 0.0207,
+      "step": 8811
+    },
+    {
+      "epoch": 3.9138352209638017,
+      "grad_norm": 0.49111074770917923,
+      "learning_rate": 1.38468178286999e-08,
+      "loss": 0.0296,
+      "step": 8812
+    },
+    {
+      "epoch": 3.9142793693093494,
+      "grad_norm": 0.36763184203730237,
+      "learning_rate": 1.3703021509692827e-08,
+      "loss": 0.0262,
+      "step": 8813
+    },
+    {
+      "epoch": 3.9147235176548967,
+      "grad_norm": 0.4095976827275787,
+      "learning_rate": 1.3559974715823266e-08,
+      "loss": 0.0297,
+      "step": 8814
+    },
+    {
+      "epoch": 3.9151676660004444,
+      "grad_norm": 0.4248033682883231,
+      "learning_rate": 1.3417677468595125e-08,
+      "loss": 0.025,
+      "step": 8815
+    },
+    {
+      "epoch": 3.9156118143459917,
+      "grad_norm": 0.4638888518779016,
+      "learning_rate": 1.3276129789397407e-08,
+      "loss": 0.0225,
+      "step": 8816
+    },
+    {
+      "epoch": 3.916055962691539,
+      "grad_norm": 0.5594291449958435,
+      "learning_rate": 1.3135331699506426e-08,
+      "loss": 0.0285,
+      "step": 8817
+    },
+    {
+      "epoch": 3.9165001110370863,
+      "grad_norm": 0.5046033957125147,
+      "learning_rate": 1.2995283220087473e-08,
+      "loss": 0.0329,
+      "step": 8818
+    },
+    {
+      "epoch": 3.9169442593826336,
+      "grad_norm": 0.4673687167446593,
+      "learning_rate": 1.2855984372191488e-08,
+      "loss": 0.0256,
+      "step": 8819
+    },
+    {
+      "epoch": 3.9173884077281813,
+      "grad_norm": 0.37484792767681735,
+      "learning_rate": 1.2717435176758386e-08,
+      "loss": 0.0271,
+      "step": 8820
+    },
+    {
+      "epoch": 3.9178325560737286,
+      "grad_norm": 0.5235629211435013,
+      "learning_rate": 1.2579635654613176e-08,
+      "loss": 0.0353,
+      "step": 8821
+    },
+    {
+      "epoch": 3.918276704419276,
+      "grad_norm": 0.37938171555742584,
+      "learning_rate": 1.24425858264704e-08,
+      "loss": 0.0236,
+      "step": 8822
+    },
+    {
+      "epoch": 3.9187208527648236,
+      "grad_norm": 0.31241390505019695,
+      "learning_rate": 1.2306285712931354e-08,
+      "loss": 0.0175,
+      "step": 8823
+    },
+    {
+      "epoch": 3.919165001110371,
+      "grad_norm": 0.3890372627975678,
+      "learning_rate": 1.2170735334482986e-08,
+      "loss": 0.0323,
+      "step": 8824
+    },
+    {
+      "epoch": 3.919609149455918,
+      "grad_norm": 0.5060461363110739,
+      "learning_rate": 1.2035934711501773e-08,
+      "loss": 0.0213,
+      "step": 8825
+    },
+    {
+      "epoch": 3.9200532978014655,
+      "grad_norm": 0.4210215040945434,
+      "learning_rate": 1.1901883864250396e-08,
+      "loss": 0.0201,
+      "step": 8826
+    },
+    {
+      "epoch": 3.9204974461470132,
+      "grad_norm": 0.3987058322406445,
+      "learning_rate": 1.1768582812878848e-08,
+      "loss": 0.0265,
+      "step": 8827
+    },
+    {
+      "epoch": 3.9209415944925605,
+      "grad_norm": 0.3908166480296464,
+      "learning_rate": 1.1636031577424434e-08,
+      "loss": 0.0204,
+      "step": 8828
+    },
+    {
+      "epoch": 3.921385742838108,
+      "grad_norm": 0.4484182047406765,
+      "learning_rate": 1.150423017781177e-08,
+      "loss": 0.0262,
+      "step": 8829
+    },
+    {
+      "epoch": 3.9218298911836555,
+      "grad_norm": 0.39560743278914223,
+      "learning_rate": 1.1373178633853344e-08,
+      "loss": 0.0301,
+      "step": 8830
+    },
+    {
+      "epoch": 3.922274039529203,
+      "grad_norm": 0.44427517578613046,
+      "learning_rate": 1.124287696524784e-08,
+      "loss": 0.033,
+      "step": 8831
+    },
+    {
+      "epoch": 3.92271818787475,
+      "grad_norm": 0.48713519459485133,
+      "learning_rate": 1.111332519158237e-08,
+      "loss": 0.0293,
+      "step": 8832
+    },
+    {
+      "epoch": 3.9231623362202974,
+      "grad_norm": 0.38920760124214965,
+      "learning_rate": 1.0984523332330244e-08,
+      "loss": 0.0313,
+      "step": 8833
+    },
+    {
+      "epoch": 3.923606484565845,
+      "grad_norm": 0.39034349333664137,
+      "learning_rate": 1.0856471406852642e-08,
+      "loss": 0.0227,
+      "step": 8834
+    },
+    {
+      "epoch": 3.9240506329113924,
+      "grad_norm": 0.5491500358691043,
+      "learning_rate": 1.0729169434398613e-08,
+      "loss": 0.046,
+      "step": 8835
+    },
+    {
+      "epoch": 3.9244947812569397,
+      "grad_norm": 0.38151095276050573,
+      "learning_rate": 1.0602617434102846e-08,
+      "loss": 0.0248,
+      "step": 8836
+    },
+    {
+      "epoch": 3.9249389296024875,
+      "grad_norm": 0.4149228267505079,
+      "learning_rate": 1.0476815424989018e-08,
+      "loss": 0.0212,
+      "step": 8837
+    },
+    {
+      "epoch": 3.9253830779480348,
+      "grad_norm": 0.4455456792947203,
+      "learning_rate": 1.0351763425966999e-08,
+      "loss": 0.0276,
+      "step": 8838
+    },
+    {
+      "epoch": 3.925827226293582,
+      "grad_norm": 0.38613196310803816,
+      "learning_rate": 1.022746145583453e-08,
+      "loss": 0.0224,
+      "step": 8839
+    },
+    {
+      "epoch": 3.9262713746391293,
+      "grad_norm": 0.4192361066154168,
+      "learning_rate": 1.0103909533275557e-08,
+      "loss": 0.0299,
+      "step": 8840
+    },
+    {
+      "epoch": 3.9267155229846766,
+      "grad_norm": 0.6267500325799473,
+      "learning_rate": 9.981107676862444e-09,
+      "loss": 0.0255,
+      "step": 8841
+    },
+    {
+      "epoch": 3.9271596713302244,
+      "grad_norm": 0.43799023094491873,
+      "learning_rate": 9.859055905054871e-09,
+      "loss": 0.0318,
+      "step": 8842
+    },
+    {
+      "epoch": 3.9276038196757717,
+      "grad_norm": 0.40925366502801463,
+      "learning_rate": 9.737754236198716e-09,
+      "loss": 0.0237,
+      "step": 8843
+    },
+    {
+      "epoch": 3.9280479680213194,
+      "grad_norm": 0.34421339563935327,
+      "learning_rate": 9.617202688527727e-09,
+      "loss": 0.02,
+      "step": 8844
+    },
+    {
+      "epoch": 3.9284921163668667,
+      "grad_norm": 0.36452089076221383,
+      "learning_rate": 9.497401280162966e-09,
+      "loss": 0.0272,
+      "step": 8845
+    },
+    {
+      "epoch": 3.928936264712414,
+      "grad_norm": 0.37821446185307267,
+      "learning_rate": 9.378350029112248e-09,
+      "loss": 0.0187,
+      "step": 8846
+    },
+    {
+      "epoch": 3.9293804130579613,
+      "grad_norm": 0.4769435842108248,
+      "learning_rate": 9.260048953271817e-09,
+      "loss": 0.0304,
+      "step": 8847
+    },
+    {
+      "epoch": 3.9298245614035086,
+      "grad_norm": 0.42416826190385143,
+      "learning_rate": 9.142498070424111e-09,
+      "loss": 0.0252,
+      "step": 8848
+    },
+    {
+      "epoch": 3.9302687097490563,
+      "grad_norm": 0.4780719986413718,
+      "learning_rate": 9.02569739823833e-09,
+      "loss": 0.0243,
+      "step": 8849
+    },
+    {
+      "epoch": 3.9307128580946036,
+      "grad_norm": 0.5064893232422892,
+      "learning_rate": 8.9096469542721e-09,
+      "loss": 0.029,
+      "step": 8850
+    },
+    {
+      "epoch": 3.931157006440151,
+      "grad_norm": 0.472928807421496,
+      "learning_rate": 8.794346755969795e-09,
+      "loss": 0.0234,
+      "step": 8851
+    },
+    {
+      "epoch": 3.9316011547856986,
+      "grad_norm": 0.40013223823456623,
+      "learning_rate": 8.679796820663111e-09,
+      "loss": 0.0344,
+      "step": 8852
+    },
+    {
+      "epoch": 3.932045303131246,
+      "grad_norm": 0.40854096697741643,
+      "learning_rate": 8.565997165570494e-09,
+      "loss": 0.0317,
+      "step": 8853
+    },
+    {
+      "epoch": 3.932489451476793,
+      "grad_norm": 0.6774113697293435,
+      "learning_rate": 8.452947807798261e-09,
+      "loss": 0.0265,
+      "step": 8854
+    },
+    {
+      "epoch": 3.9329335998223405,
+      "grad_norm": 0.47756331700457766,
+      "learning_rate": 8.340648764339487e-09,
+      "loss": 0.0245,
+      "step": 8855
+    },
+    {
+      "epoch": 3.933377748167888,
+      "grad_norm": 0.408015599055906,
+      "learning_rate": 8.229100052074557e-09,
+      "loss": 0.0266,
+      "step": 8856
+    },
+    {
+      "epoch": 3.9338218965134355,
+      "grad_norm": 0.3723159509813085,
+      "learning_rate": 8.118301687771169e-09,
+      "loss": 0.0244,
+      "step": 8857
+    },
+    {
+      "epoch": 3.934266044858983,
+      "grad_norm": 0.44846859119809684,
+      "learning_rate": 8.008253688084888e-09,
+      "loss": 0.0199,
+      "step": 8858
+    },
+    {
+      "epoch": 3.9347101932045305,
+      "grad_norm": 0.3532724419502754,
+      "learning_rate": 7.898956069556375e-09,
+      "loss": 0.0167,
+      "step": 8859
+    },
+    {
+      "epoch": 3.935154341550078,
+      "grad_norm": 0.38377139505288993,
+      "learning_rate": 7.790408848616371e-09,
+      "loss": 0.025,
+      "step": 8860
+    },
+    {
+      "epoch": 3.935598489895625,
+      "grad_norm": 0.47706384193903933,
+      "learning_rate": 7.682612041580161e-09,
+      "loss": 0.0298,
+      "step": 8861
+    },
+    {
+      "epoch": 3.9360426382411724,
+      "grad_norm": 0.47799572814835295,
+      "learning_rate": 7.575565664652562e-09,
+      "loss": 0.0232,
+      "step": 8862
+    },
+    {
+      "epoch": 3.9364867865867197,
+      "grad_norm": 0.42569464791407546,
+      "learning_rate": 7.469269733923478e-09,
+      "loss": 0.0284,
+      "step": 8863
+    },
+    {
+      "epoch": 3.9369309349322674,
+      "grad_norm": 0.49128556583785543,
+      "learning_rate": 7.363724265371796e-09,
+      "loss": 0.0322,
+      "step": 8864
+    },
+    {
+      "epoch": 3.9373750832778147,
+      "grad_norm": 0.4426824222148885,
+      "learning_rate": 7.258929274862048e-09,
+      "loss": 0.0244,
+      "step": 8865
+    },
+    {
+      "epoch": 3.9378192316233624,
+      "grad_norm": 0.4220498782198706,
+      "learning_rate": 7.154884778147187e-09,
+      "loss": 0.0381,
+      "step": 8866
+    },
+    {
+      "epoch": 3.9382633799689097,
+      "grad_norm": 0.40573724780385634,
+      "learning_rate": 7.051590790866925e-09,
+      "loss": 0.0208,
+      "step": 8867
+    },
+    {
+      "epoch": 3.938707528314457,
+      "grad_norm": 0.36192582224558784,
+      "learning_rate": 6.949047328547731e-09,
+      "loss": 0.0186,
+      "step": 8868
+    },
+    {
+      "epoch": 3.9391516766600043,
+      "grad_norm": 0.44081181805446706,
+      "learning_rate": 6.847254406603943e-09,
+      "loss": 0.0305,
+      "step": 8869
+    },
+    {
+      "epoch": 3.9395958250055516,
+      "grad_norm": 0.47409619445972684,
+      "learning_rate": 6.746212040336653e-09,
+      "loss": 0.0276,
+      "step": 8870
+    },
+    {
+      "epoch": 3.9400399733510993,
+      "grad_norm": 0.3907822464278477,
+      "learning_rate": 6.645920244934267e-09,
+      "loss": 0.0265,
+      "step": 8871
+    },
+    {
+      "epoch": 3.9404841216966466,
+      "grad_norm": 0.3877502095250313,
+      "learning_rate": 6.546379035472505e-09,
+      "loss": 0.0196,
+      "step": 8872
+    },
+    {
+      "epoch": 3.9409282700421944,
+      "grad_norm": 0.36359890877779866,
+      "learning_rate": 6.447588426913287e-09,
+      "loss": 0.0207,
+      "step": 8873
+    },
+    {
+      "epoch": 3.9413724183877417,
+      "grad_norm": 0.4046244907134214,
+      "learning_rate": 6.349548434108066e-09,
+      "loss": 0.0245,
+      "step": 8874
+    },
+    {
+      "epoch": 3.941816566733289,
+      "grad_norm": 0.30688221794365883,
+      "learning_rate": 6.252259071792277e-09,
+      "loss": 0.0252,
+      "step": 8875
+    },
+    {
+      "epoch": 3.9422607150788362,
+      "grad_norm": 0.3706992283845886,
+      "learning_rate": 6.155720354590888e-09,
+      "loss": 0.0179,
+      "step": 8876
+    },
+    {
+      "epoch": 3.9427048634243835,
+      "grad_norm": 0.3606911516685271,
+      "learning_rate": 6.059932297015625e-09,
+      "loss": 0.0155,
+      "step": 8877
+    },
+    {
+      "epoch": 3.9431490117699313,
+      "grad_norm": 0.3448866184147424,
+      "learning_rate": 5.964894913464969e-09,
+      "loss": 0.0238,
+      "step": 8878
+    },
+    {
+      "epoch": 3.9435931601154786,
+      "grad_norm": 0.42743435978316485,
+      "learning_rate": 5.8706082182241605e-09,
+      "loss": 0.019,
+      "step": 8879
+    },
+    {
+      "epoch": 3.944037308461026,
+      "grad_norm": 0.38374298902655224,
+      "learning_rate": 5.777072225466307e-09,
+      "loss": 0.0232,
+      "step": 8880
+    },
+    {
+      "epoch": 3.9444814568065736,
+      "grad_norm": 0.3737396272194393,
+      "learning_rate": 5.684286949251272e-09,
+      "loss": 0.0218,
+      "step": 8881
+    },
+    {
+      "epoch": 3.944925605152121,
+      "grad_norm": 0.3717422178901769,
+      "learning_rate": 5.592252403526788e-09,
+      "loss": 0.018,
+      "step": 8882
+    },
+    {
+      "epoch": 3.945369753497668,
+      "grad_norm": 0.4530028394270273,
+      "learning_rate": 5.500968602126788e-09,
+      "loss": 0.0311,
+      "step": 8883
+    },
+    {
+      "epoch": 3.9458139018432155,
+      "grad_norm": 0.34242197949439085,
+      "learning_rate": 5.410435558773075e-09,
+      "loss": 0.0229,
+      "step": 8884
+    },
+    {
+      "epoch": 3.946258050188763,
+      "grad_norm": 0.39126976030692906,
+      "learning_rate": 5.3206532870742065e-09,
+      "loss": 0.0253,
+      "step": 8885
+    },
+    {
+      "epoch": 3.9467021985343105,
+      "grad_norm": 0.43259511059270805,
+      "learning_rate": 5.231621800525499e-09,
+      "loss": 0.0222,
+      "step": 8886
+    },
+    {
+      "epoch": 3.9471463468798578,
+      "grad_norm": 0.5231951089873627,
+      "learning_rate": 5.143341112510691e-09,
+      "loss": 0.0432,
+      "step": 8887
+    },
+    {
+      "epoch": 3.9475904952254055,
+      "grad_norm": 0.4699803462882175,
+      "learning_rate": 5.055811236299724e-09,
+      "loss": 0.0242,
+      "step": 8888
+    },
+    {
+      "epoch": 3.948034643570953,
+      "grad_norm": 0.4043702599111712,
+      "learning_rate": 4.969032185049294e-09,
+      "loss": 0.0367,
+      "step": 8889
+    },
+    {
+      "epoch": 3.9484787919165,
+      "grad_norm": 0.41944855374085116,
+      "learning_rate": 4.883003971803968e-09,
+      "loss": 0.0236,
+      "step": 8890
+    },
+    {
+      "epoch": 3.9489229402620474,
+      "grad_norm": 0.40457657146672565,
+      "learning_rate": 4.797726609495623e-09,
+      "loss": 0.0188,
+      "step": 8891
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.4052288455507649,
+      "learning_rate": 4.7132001109423396e-09,
+      "loss": 0.0266,
+      "step": 8892
+    },
+    {
+      "epoch": 3.9498112369531424,
+      "grad_norm": 0.5765436543403538,
+      "learning_rate": 4.629424488850065e-09,
+      "loss": 0.0362,
+      "step": 8893
+    },
+    {
+      "epoch": 3.9502553852986897,
+      "grad_norm": 0.4676308344783614,
+      "learning_rate": 4.546399755812059e-09,
+      "loss": 0.0249,
+      "step": 8894
+    },
+    {
+      "epoch": 3.9506995336442374,
+      "grad_norm": 0.28808412670077804,
+      "learning_rate": 4.4641259243077825e-09,
+      "loss": 0.0137,
+      "step": 8895
+    },
+    {
+      "epoch": 3.9511436819897847,
+      "grad_norm": 0.47058698890390266,
+      "learning_rate": 4.382603006705121e-09,
+      "loss": 0.0319,
+      "step": 8896
+    },
+    {
+      "epoch": 3.951587830335332,
+      "grad_norm": 0.41572017055736116,
+      "learning_rate": 4.301831015257607e-09,
+      "loss": 0.0221,
+      "step": 8897
+    },
+    {
+      "epoch": 3.9520319786808793,
+      "grad_norm": 0.39457081647577197,
+      "learning_rate": 4.221809962107193e-09,
+      "loss": 0.0248,
+      "step": 8898
+    },
+    {
+      "epoch": 3.9524761270264266,
+      "grad_norm": 0.5337090298271753,
+      "learning_rate": 4.142539859282035e-09,
+      "loss": 0.036,
+      "step": 8899
+    },
+    {
+      "epoch": 3.9529202753719743,
+      "grad_norm": 0.4292452210463301,
+      "learning_rate": 4.064020718698158e-09,
+      "loss": 0.0276,
+      "step": 8900
+    },
+    {
+      "epoch": 3.9533644237175216,
+      "grad_norm": 0.35441583703246865,
+      "learning_rate": 3.986252552157788e-09,
+      "loss": 0.0228,
+      "step": 8901
+    },
+    {
+      "epoch": 3.953808572063069,
+      "grad_norm": 0.4031250072105655,
+      "learning_rate": 3.909235371351017e-09,
+      "loss": 0.028,
+      "step": 8902
+    },
+    {
+      "epoch": 3.9542527204086166,
+      "grad_norm": 0.32842995453695917,
+      "learning_rate": 3.832969187855251e-09,
+      "loss": 0.0244,
+      "step": 8903
+    },
+    {
+      "epoch": 3.954696868754164,
+      "grad_norm": 0.4192351275506548,
+      "learning_rate": 3.757454013134099e-09,
+      "loss": 0.0215,
+      "step": 8904
+    },
+    {
+      "epoch": 3.955141017099711,
+      "grad_norm": 0.4517856702968858,
+      "learning_rate": 3.682689858539035e-09,
+      "loss": 0.0308,
+      "step": 8905
+    },
+    {
+      "epoch": 3.9555851654452585,
+      "grad_norm": 0.4000805272591098,
+      "learning_rate": 3.608676735308292e-09,
+      "loss": 0.0257,
+      "step": 8906
+    },
+    {
+      "epoch": 3.9560293137908062,
+      "grad_norm": 0.39861614075630786,
+      "learning_rate": 3.5354146545668597e-09,
+      "loss": 0.0268,
+      "step": 8907
+    },
+    {
+      "epoch": 3.9564734621363535,
+      "grad_norm": 0.3761331007710691,
+      "learning_rate": 3.462903627328151e-09,
+      "loss": 0.0212,
+      "step": 8908
+    },
+    {
+      "epoch": 3.956917610481901,
+      "grad_norm": 0.552448221226994,
+      "learning_rate": 3.3911436644912256e-09,
+      "loss": 0.0272,
+      "step": 8909
+    },
+    {
+      "epoch": 3.9573617588274486,
+      "grad_norm": 0.40174202534071046,
+      "learning_rate": 3.3201347768430093e-09,
+      "loss": 0.0289,
+      "step": 8910
+    },
+    {
+      "epoch": 3.957805907172996,
+      "grad_norm": 0.3916127999245288,
+      "learning_rate": 3.249876975057187e-09,
+      "loss": 0.0251,
+      "step": 8911
+    },
+    {
+      "epoch": 3.958250055518543,
+      "grad_norm": 0.40161063795102814,
+      "learning_rate": 3.1803702696947547e-09,
+      "loss": 0.0201,
+      "step": 8912
+    },
+    {
+      "epoch": 3.9586942038640904,
+      "grad_norm": 0.4175398836504673,
+      "learning_rate": 3.111614671204022e-09,
+      "loss": 0.0236,
+      "step": 8913
+    },
+    {
+      "epoch": 3.959138352209638,
+      "grad_norm": 0.5695134220357339,
+      "learning_rate": 3.043610189919499e-09,
+      "loss": 0.0289,
+      "step": 8914
+    },
+    {
+      "epoch": 3.9595825005551855,
+      "grad_norm": 0.47624104829660213,
+      "learning_rate": 2.97635683606412e-09,
+      "loss": 0.0319,
+      "step": 8915
+    },
+    {
+      "epoch": 3.9600266489007327,
+      "grad_norm": 0.4611136348425102,
+      "learning_rate": 2.909854619747021e-09,
+      "loss": 0.0207,
+      "step": 8916
+    },
+    {
+      "epoch": 3.9604707972462805,
+      "grad_norm": 0.49713796334737675,
+      "learning_rate": 2.8441035509640947e-09,
+      "loss": 0.0273,
+      "step": 8917
+    },
+    {
+      "epoch": 3.9609149455918278,
+      "grad_norm": 0.5029210915242077,
+      "learning_rate": 2.7791036395996563e-09,
+      "loss": 0.0413,
+      "step": 8918
+    },
+    {
+      "epoch": 3.961359093937375,
+      "grad_norm": 0.4809399615159702,
+      "learning_rate": 2.7148548954236687e-09,
+      "loss": 0.0326,
+      "step": 8919
+    },
+    {
+      "epoch": 3.9618032422829224,
+      "grad_norm": 0.4107926474040777,
+      "learning_rate": 2.6513573280939618e-09,
+      "loss": 0.0243,
+      "step": 8920
+    },
+    {
+      "epoch": 3.9622473906284696,
+      "grad_norm": 0.39038109738475957,
+      "learning_rate": 2.5886109471551233e-09,
+      "loss": 0.0251,
+      "step": 8921
+    },
+    {
+      "epoch": 3.9626915389740174,
+      "grad_norm": 0.328940131759137,
+      "learning_rate": 2.526615762039608e-09,
+      "loss": 0.0163,
+      "step": 8922
+    },
+    {
+      "epoch": 3.9631356873195647,
+      "grad_norm": 0.4003085651392341,
+      "learning_rate": 2.465371782066073e-09,
+      "loss": 0.0236,
+      "step": 8923
+    },
+    {
+      "epoch": 3.9635798356651124,
+      "grad_norm": 0.4454242880030863,
+      "learning_rate": 2.4048790164404866e-09,
+      "loss": 0.0232,
+      "step": 8924
+    },
+    {
+      "epoch": 3.9640239840106597,
+      "grad_norm": 0.34629533945323293,
+      "learning_rate": 2.3451374742555764e-09,
+      "loss": 0.0177,
+      "step": 8925
+    },
+    {
+      "epoch": 3.964468132356207,
+      "grad_norm": 0.4133618406412272,
+      "learning_rate": 2.2861471644919363e-09,
+      "loss": 0.0191,
+      "step": 8926
+    },
+    {
+      "epoch": 3.9649122807017543,
+      "grad_norm": 0.42649199370055857,
+      "learning_rate": 2.2279080960163625e-09,
+      "loss": 0.027,
+      "step": 8927
+    },
+    {
+      "epoch": 3.9653564290473016,
+      "grad_norm": 0.3956675634414063,
+      "learning_rate": 2.170420277584073e-09,
+      "loss": 0.0263,
+      "step": 8928
+    },
+    {
+      "epoch": 3.9658005773928493,
+      "grad_norm": 0.4346162397069076,
+      "learning_rate": 2.1136837178353797e-09,
+      "loss": 0.0258,
+      "step": 8929
+    },
+    {
+      "epoch": 3.9662447257383966,
+      "grad_norm": 0.4508730863193407,
+      "learning_rate": 2.05769842529957e-09,
+      "loss": 0.0355,
+      "step": 8930
+    },
+    {
+      "epoch": 3.966688874083944,
+      "grad_norm": 0.44677554596928787,
+      "learning_rate": 2.002464408392135e-09,
+      "loss": 0.0245,
+      "step": 8931
+    },
+    {
+      "epoch": 3.9671330224294916,
+      "grad_norm": 0.3494473501914632,
+      "learning_rate": 1.9479816754147672e-09,
+      "loss": 0.0231,
+      "step": 8932
+    },
+    {
+      "epoch": 3.967577170775039,
+      "grad_norm": 0.3733325402809653,
+      "learning_rate": 1.894250234558137e-09,
+      "loss": 0.0177,
+      "step": 8933
+    },
+    {
+      "epoch": 3.968021319120586,
+      "grad_norm": 0.42732301014072877,
+      "learning_rate": 1.8412700938985618e-09,
+      "loss": 0.0245,
+      "step": 8934
+    },
+    {
+      "epoch": 3.9684654674661335,
+      "grad_norm": 0.5850533107887496,
+      "learning_rate": 1.7890412614002262e-09,
+      "loss": 0.0173,
+      "step": 8935
+    },
+    {
+      "epoch": 3.9689096158116812,
+      "grad_norm": 0.42327533255156274,
+      "learning_rate": 1.7375637449135174e-09,
+      "loss": 0.0217,
+      "step": 8936
+    },
+    {
+      "epoch": 3.9693537641572285,
+      "grad_norm": 0.45340718199873464,
+      "learning_rate": 1.68683755217669e-09,
+      "loss": 0.0329,
+      "step": 8937
+    },
+    {
+      "epoch": 3.969797912502776,
+      "grad_norm": 0.43589126986170634,
+      "learning_rate": 1.6368626908147556e-09,
+      "loss": 0.0294,
+      "step": 8938
+    },
+    {
+      "epoch": 3.9702420608483235,
+      "grad_norm": 0.37544409700894205,
+      "learning_rate": 1.5876391683400383e-09,
+      "loss": 0.0356,
+      "step": 8939
+    },
+    {
+      "epoch": 3.970686209193871,
+      "grad_norm": 0.3711146055897954,
+      "learning_rate": 1.5391669921505093e-09,
+      "loss": 0.0304,
+      "step": 8940
+    },
+    {
+      "epoch": 3.971130357539418,
+      "grad_norm": 0.40437917019760156,
+      "learning_rate": 1.4914461695336723e-09,
+      "loss": 0.0186,
+      "step": 8941
+    },
+    {
+      "epoch": 3.9715745058849654,
+      "grad_norm": 0.36049022928808616,
+      "learning_rate": 1.4444767076626787e-09,
+      "loss": 0.0221,
+      "step": 8942
+    },
+    {
+      "epoch": 3.972018654230513,
+      "grad_norm": 0.49140822525474664,
+      "learning_rate": 1.3982586135968813e-09,
+      "loss": 0.0267,
+      "step": 8943
+    },
+    {
+      "epoch": 3.9724628025760604,
+      "grad_norm": 0.5302909084020938,
+      "learning_rate": 1.3527918942840556e-09,
+      "loss": 0.041,
+      "step": 8944
+    },
+    {
+      "epoch": 3.9729069509216077,
+      "grad_norm": 0.5476906455774254,
+      "learning_rate": 1.3080765565592902e-09,
+      "loss": 0.0273,
+      "step": 8945
+    },
+    {
+      "epoch": 3.9733510992671555,
+      "grad_norm": 0.3333489521628196,
+      "learning_rate": 1.2641126071433197e-09,
+      "loss": 0.0268,
+      "step": 8946
+    },
+    {
+      "epoch": 3.9737952476127028,
+      "grad_norm": 0.39064699714492984,
+      "learning_rate": 1.2209000526447469e-09,
+      "loss": 0.0204,
+      "step": 8947
+    },
+    {
+      "epoch": 3.97423939595825,
+      "grad_norm": 0.5569247363465006,
+      "learning_rate": 1.1784388995594864e-09,
+      "loss": 0.0321,
+      "step": 8948
+    },
+    {
+      "epoch": 3.9746835443037973,
+      "grad_norm": 0.4766119680419059,
+      "learning_rate": 1.1367291542702107e-09,
+      "loss": 0.0284,
+      "step": 8949
+    },
+    {
+      "epoch": 3.9751276926493446,
+      "grad_norm": 0.5034524006899211,
+      "learning_rate": 1.0957708230457942e-09,
+      "loss": 0.0271,
+      "step": 8950
+    },
+    {
+      "epoch": 3.9755718409948924,
+      "grad_norm": 0.3381689403401387,
+      "learning_rate": 1.0555639120440887e-09,
+      "loss": 0.0201,
+      "step": 8951
+    },
+    {
+      "epoch": 3.9760159893404396,
+      "grad_norm": 0.46151159178049117,
+      "learning_rate": 1.0161084273080378e-09,
+      "loss": 0.0296,
+      "step": 8952
+    },
+    {
+      "epoch": 3.9764601376859874,
+      "grad_norm": 0.4521959026876067,
+      "learning_rate": 9.774043747690087e-10,
+      "loss": 0.0311,
+      "step": 8953
+    },
+    {
+      "epoch": 3.9769042860315347,
+      "grad_norm": 0.48398453476246417,
+      "learning_rate": 9.394517602445697e-10,
+      "loss": 0.0309,
+      "step": 8954
+    },
+    {
+      "epoch": 3.977348434377082,
+      "grad_norm": 0.3856389581537802,
+      "learning_rate": 9.022505894396017e-10,
+      "loss": 0.0216,
+      "step": 8955
+    },
+    {
+      "epoch": 3.9777925827226293,
+      "grad_norm": 0.3450747091674961,
+      "learning_rate": 8.658008679462981e-10,
+      "loss": 0.0157,
+      "step": 8956
+    },
+    {
+      "epoch": 3.9782367310681765,
+      "grad_norm": 0.3798481275168741,
+      "learning_rate": 8.301026012436098e-10,
+      "loss": 0.0197,
+      "step": 8957
+    },
+    {
+      "epoch": 3.9786808794137243,
+      "grad_norm": 0.4175802111530846,
+      "learning_rate": 7.951557946972444e-10,
+      "loss": 0.0254,
+      "step": 8958
+    },
+    {
+      "epoch": 3.9791250277592716,
+      "grad_norm": 0.3948056525475575,
+      "learning_rate": 7.609604535613324e-10,
+      "loss": 0.0198,
+      "step": 8959
+    },
+    {
+      "epoch": 3.979569176104819,
+      "grad_norm": 0.5080884160142413,
+      "learning_rate": 7.275165829745412e-10,
+      "loss": 0.0204,
+      "step": 8960
+    },
+    {
+      "epoch": 3.9800133244503666,
+      "grad_norm": 0.3995310763869156,
+      "learning_rate": 6.948241879650708e-10,
+      "loss": 0.0296,
+      "step": 8961
+    },
+    {
+      "epoch": 3.980457472795914,
+      "grad_norm": 0.37430290093081664,
+      "learning_rate": 6.628832734467683e-10,
+      "loss": 0.0238,
+      "step": 8962
+    },
+    {
+      "epoch": 3.980901621141461,
+      "grad_norm": 0.43982737443124764,
+      "learning_rate": 6.316938442213483e-10,
+      "loss": 0.0209,
+      "step": 8963
+    },
+    {
+      "epoch": 3.9813457694870085,
+      "grad_norm": 0.3775182755482849,
+      "learning_rate": 6.012559049761723e-10,
+      "loss": 0.025,
+      "step": 8964
+    },
+    {
+      "epoch": 3.981789917832556,
+      "grad_norm": 0.39039801224690585,
+      "learning_rate": 5.715694602875799e-10,
+      "loss": 0.0207,
+      "step": 8965
+    },
+    {
+      "epoch": 3.9822340661781035,
+      "grad_norm": 0.4199762038802494,
+      "learning_rate": 5.426345146175571e-10,
+      "loss": 0.0191,
+      "step": 8966
+    },
+    {
+      "epoch": 3.982678214523651,
+      "grad_norm": 0.38333717972878917,
+      "learning_rate": 5.144510723154028e-10,
+      "loss": 0.0284,
+      "step": 8967
+    },
+    {
+      "epoch": 3.9831223628691985,
+      "grad_norm": 0.4104443185350045,
+      "learning_rate": 4.87019137617728e-10,
+      "loss": 0.0189,
+      "step": 8968
+    },
+    {
+      "epoch": 3.983566511214746,
+      "grad_norm": 0.5433078848500579,
+      "learning_rate": 4.60338714647901e-10,
+      "loss": 0.0302,
+      "step": 8969
+    },
+    {
+      "epoch": 3.984010659560293,
+      "grad_norm": 0.3909125068421461,
+      "learning_rate": 4.3440980741660254e-10,
+      "loss": 0.0238,
+      "step": 8970
+    },
+    {
+      "epoch": 3.9844548079058404,
+      "grad_norm": 0.41307839626778975,
+      "learning_rate": 4.092324198212705e-10,
+      "loss": 0.0235,
+      "step": 8971
+    },
+    {
+      "epoch": 3.984898956251388,
+      "grad_norm": 0.321331298647483,
+      "learning_rate": 3.848065556461e-10,
+      "loss": 0.0236,
+      "step": 8972
+    },
+    {
+      "epoch": 3.9853431045969354,
+      "grad_norm": 0.3312110185029968,
+      "learning_rate": 3.6113221856370896e-10,
+      "loss": 0.0219,
+      "step": 8973
+    },
+    {
+      "epoch": 3.9857872529424827,
+      "grad_norm": 0.4192103414685381,
+      "learning_rate": 3.382094121318069e-10,
+      "loss": 0.0266,
+      "step": 8974
+    },
+    {
+      "epoch": 3.9862314012880304,
+      "grad_norm": 0.35796707903705355,
+      "learning_rate": 3.160381397965262e-10,
+      "loss": 0.0177,
+      "step": 8975
+    },
+    {
+      "epoch": 3.9866755496335777,
+      "grad_norm": 0.3908253098336717,
+      "learning_rate": 2.946184048902012e-10,
+      "loss": 0.0235,
+      "step": 8976
+    },
+    {
+      "epoch": 3.987119697979125,
+      "grad_norm": 0.42457467356965617,
+      "learning_rate": 2.7395021063303385e-10,
+      "loss": 0.0397,
+      "step": 8977
+    },
+    {
+      "epoch": 3.9875638463246723,
+      "grad_norm": 0.4884888876304642,
+      "learning_rate": 2.540335601319832e-10,
+      "loss": 0.0228,
+      "step": 8978
+    },
+    {
+      "epoch": 3.9880079946702196,
+      "grad_norm": 0.45397167330183835,
+      "learning_rate": 2.348684563802106e-10,
+      "loss": 0.0215,
+      "step": 8979
+    },
+    {
+      "epoch": 3.9884521430157673,
+      "grad_norm": 0.48150914332233963,
+      "learning_rate": 2.1645490225929989e-10,
+      "loss": 0.0207,
+      "step": 8980
+    },
+    {
+      "epoch": 3.9888962913613146,
+      "grad_norm": 0.518549165106476,
+      "learning_rate": 1.9879290053592682e-10,
+      "loss": 0.0276,
+      "step": 8981
+    },
+    {
+      "epoch": 3.9893404397068624,
+      "grad_norm": 0.3673159296164483,
+      "learning_rate": 1.8188245386629998e-10,
+      "loss": 0.0188,
+      "step": 8982
+    },
+    {
+      "epoch": 3.9897845880524097,
+      "grad_norm": 0.34298096212254114,
+      "learning_rate": 1.657235647917199e-10,
+      "loss": 0.0169,
+      "step": 8983
+    },
+    {
+      "epoch": 3.990228736397957,
+      "grad_norm": 0.4555639884295353,
+      "learning_rate": 1.5031623574135456e-10,
+      "loss": 0.0288,
+      "step": 8984
+    },
+    {
+      "epoch": 3.9906728847435042,
+      "grad_norm": 0.3767875220229666,
+      "learning_rate": 1.3566046903057405e-10,
+      "loss": 0.0245,
+      "step": 8985
+    },
+    {
+      "epoch": 3.9911170330890515,
+      "grad_norm": 0.3435105743106647,
+      "learning_rate": 1.217562668631711e-10,
+      "loss": 0.0192,
+      "step": 8986
+    },
+    {
+      "epoch": 3.9915611814345993,
+      "grad_norm": 0.5264624768016674,
+      "learning_rate": 1.0860363132914053e-10,
+      "loss": 0.0215,
+      "step": 8987
+    },
+    {
+      "epoch": 3.9920053297801465,
+      "grad_norm": 0.4654928703557639,
+      "learning_rate": 9.620256440467934e-11,
+      "loss": 0.0253,
+      "step": 8988
+    },
+    {
+      "epoch": 3.992449478125694,
+      "grad_norm": 0.42855294671291005,
+      "learning_rate": 8.455306795496221e-11,
+      "loss": 0.0322,
+      "step": 8989
+    },
+    {
+      "epoch": 3.9928936264712416,
+      "grad_norm": 0.33395128737336627,
+      "learning_rate": 7.365514373081084e-11,
+      "loss": 0.0159,
+      "step": 8990
+    },
+    {
+      "epoch": 3.993337774816789,
+      "grad_norm": 0.7884624510448948,
+      "learning_rate": 6.35087933698042e-11,
+      "loss": 0.0429,
+      "step": 8991
+    },
+    {
+      "epoch": 3.993781923162336,
+      "grad_norm": 0.4471065505434232,
+      "learning_rate": 5.411401839738872e-11,
+      "loss": 0.032,
+      "step": 8992
+    },
+    {
+      "epoch": 3.9942260715078834,
+      "grad_norm": 0.3926302149338509,
+      "learning_rate": 4.547082022632321e-11,
+      "loss": 0.0269,
+      "step": 8993
+    },
+    {
+      "epoch": 3.994670219853431,
+      "grad_norm": 0.35432194327597677,
+      "learning_rate": 3.7579200155013483e-11,
+      "loss": 0.024,
+      "step": 8994
+    },
+    {
+      "epoch": 3.9951143681989785,
+      "grad_norm": 0.39044461483214893,
+      "learning_rate": 3.043915937028796e-11,
+      "loss": 0.025,
+      "step": 8995
+    },
+    {
+      "epoch": 3.9955585165445258,
+      "grad_norm": 0.3981648535488869,
+      "learning_rate": 2.4050698944622087e-11,
+      "loss": 0.0221,
+      "step": 8996
+    },
+    {
+      "epoch": 3.9960026648900735,
+      "grad_norm": 0.42669487275827683,
+      "learning_rate": 1.8413819839468993e-11,
+      "loss": 0.0231,
+      "step": 8997
+    },
+    {
+      "epoch": 3.996446813235621,
+      "grad_norm": 0.33538547874165714,
+      "learning_rate": 1.3528522901373741e-11,
+      "loss": 0.0253,
+      "step": 8998
+    },
+    {
+      "epoch": 3.996890961581168,
+      "grad_norm": 0.4935415446009399,
+      "learning_rate": 9.394808864748861e-12,
+      "loss": 0.0315,
+      "step": 8999
+    },
+    {
+      "epoch": 3.9973351099267154,
+      "grad_norm": 0.37018534448771545,
+      "learning_rate": 6.012678351319245e-12,
+      "loss": 0.0243,
+      "step": 9000
+    },
+    {
+      "epoch": 3.9977792582722627,
+      "grad_norm": 0.38372140221372397,
+      "learning_rate": 3.3821318690119286e-12,
+      "loss": 0.021,
+      "step": 9001
+    },
+    {
+      "epoch": 3.9982234066178104,
+      "grad_norm": 0.3907968845917238,
+      "learning_rate": 1.503169813621419e-12,
+      "loss": 0.0222,
+      "step": 9002
+    },
+    {
+      "epoch": 3.9986675549633577,
+      "grad_norm": 0.4131506866019968,
+      "learning_rate": 3.757924676994762e-13,
+      "loss": 0.0311,
+      "step": 9003
+    },
+    {
+      "epoch": 3.9991117033089054,
+      "grad_norm": 0.47497534756496,
+      "learning_rate": 0.0,
+      "loss": 0.0269,
+      "step": 9004
+    },
+    {
+      "epoch": 3.9991117033089054,
+      "eval_loss": 0.03634560480713844,
+      "eval_runtime": 402.6359,
+      "eval_samples_per_second": 37.667,
+      "eval_steps_per_second": 1.177,
+      "step": 9004
+    },
+    {
+      "epoch": 3.9991117033089054,
+      "step": 9004,
+      "total_flos": 1010387024977920.0,
+      "train_loss": 0.05092642861352228,
+      "train_runtime": 98184.5209,
+      "train_samples_per_second": 11.738,
+      "train_steps_per_second": 0.092
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 9004,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "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": 1010387024977920.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}