diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,83538 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 11928,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00025150905432595576,
+      "grad_norm": 5.952406406402588,
+      "learning_rate": 8.382229673093043e-09,
+      "loss": 0.8367,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005030181086519115,
+      "grad_norm": 6.153665542602539,
+      "learning_rate": 1.6764459346186086e-08,
+      "loss": 0.8394,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007545271629778672,
+      "grad_norm": 5.872622966766357,
+      "learning_rate": 2.5146689019279132e-08,
+      "loss": 0.8047,
+      "step": 3
+    },
+    {
+      "epoch": 0.001006036217303823,
+      "grad_norm": 6.132771968841553,
+      "learning_rate": 3.352891869237217e-08,
+      "loss": 0.8308,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012575452716297787,
+      "grad_norm": 6.068148612976074,
+      "learning_rate": 4.191114836546522e-08,
+      "loss": 0.7871,
+      "step": 5
+    },
+    {
+      "epoch": 0.0015090543259557343,
+      "grad_norm": 5.759831428527832,
+      "learning_rate": 5.0293378038558264e-08,
+      "loss": 0.8009,
+      "step": 6
+    },
+    {
+      "epoch": 0.0017605633802816902,
+      "grad_norm": 6.06758975982666,
+      "learning_rate": 5.8675607711651307e-08,
+      "loss": 0.8295,
+      "step": 7
+    },
+    {
+      "epoch": 0.002012072434607646,
+      "grad_norm": 6.002471923828125,
+      "learning_rate": 6.705783738474434e-08,
+      "loss": 0.8253,
+      "step": 8
+    },
+    {
+      "epoch": 0.0022635814889336017,
+      "grad_norm": 6.0556440353393555,
+      "learning_rate": 7.544006705783739e-08,
+      "loss": 0.8259,
+      "step": 9
+    },
+    {
+      "epoch": 0.0025150905432595573,
+      "grad_norm": 5.968860626220703,
+      "learning_rate": 8.382229673093044e-08,
+      "loss": 0.8368,
+      "step": 10
+    },
+    {
+      "epoch": 0.002766599597585513,
+      "grad_norm": 5.972998142242432,
+      "learning_rate": 9.220452640402346e-08,
+      "loss": 0.8222,
+      "step": 11
+    },
+    {
+      "epoch": 0.0030181086519114686,
+      "grad_norm": 5.935672760009766,
+      "learning_rate": 1.0058675607711653e-07,
+      "loss": 0.8099,
+      "step": 12
+    },
+    {
+      "epoch": 0.0032696177062374247,
+      "grad_norm": 6.067246913909912,
+      "learning_rate": 1.0896898575020955e-07,
+      "loss": 0.832,
+      "step": 13
+    },
+    {
+      "epoch": 0.0035211267605633804,
+      "grad_norm": 6.090888977050781,
+      "learning_rate": 1.1735121542330261e-07,
+      "loss": 0.8276,
+      "step": 14
+    },
+    {
+      "epoch": 0.003772635814889336,
+      "grad_norm": 5.8601837158203125,
+      "learning_rate": 1.2573344509639564e-07,
+      "loss": 0.8273,
+      "step": 15
+    },
+    {
+      "epoch": 0.004024144869215292,
+      "grad_norm": 6.011853218078613,
+      "learning_rate": 1.3411567476948869e-07,
+      "loss": 0.8194,
+      "step": 16
+    },
+    {
+      "epoch": 0.004275653923541248,
+      "grad_norm": 5.5620808601379395,
+      "learning_rate": 1.4249790444258174e-07,
+      "loss": 0.791,
+      "step": 17
+    },
+    {
+      "epoch": 0.004527162977867203,
+      "grad_norm": 5.906956672668457,
+      "learning_rate": 1.5088013411567478e-07,
+      "loss": 0.8452,
+      "step": 18
+    },
+    {
+      "epoch": 0.004778672032193159,
+      "grad_norm": 5.789212226867676,
+      "learning_rate": 1.592623637887678e-07,
+      "loss": 0.8181,
+      "step": 19
+    },
+    {
+      "epoch": 0.005030181086519115,
+      "grad_norm": 5.891463756561279,
+      "learning_rate": 1.6764459346186088e-07,
+      "loss": 0.8255,
+      "step": 20
+    },
+    {
+      "epoch": 0.00528169014084507,
+      "grad_norm": 5.891812801361084,
+      "learning_rate": 1.7602682313495393e-07,
+      "loss": 0.8251,
+      "step": 21
+    },
+    {
+      "epoch": 0.005533199195171026,
+      "grad_norm": 5.88001823425293,
+      "learning_rate": 1.8440905280804693e-07,
+      "loss": 0.8264,
+      "step": 22
+    },
+    {
+      "epoch": 0.005784708249496982,
+      "grad_norm": 5.622618675231934,
+      "learning_rate": 1.9279128248113998e-07,
+      "loss": 0.7947,
+      "step": 23
+    },
+    {
+      "epoch": 0.006036217303822937,
+      "grad_norm": 5.888658046722412,
+      "learning_rate": 2.0117351215423305e-07,
+      "loss": 0.843,
+      "step": 24
+    },
+    {
+      "epoch": 0.006287726358148894,
+      "grad_norm": 5.512774467468262,
+      "learning_rate": 2.095557418273261e-07,
+      "loss": 0.7908,
+      "step": 25
+    },
+    {
+      "epoch": 0.006539235412474849,
+      "grad_norm": 5.594388484954834,
+      "learning_rate": 2.179379715004191e-07,
+      "loss": 0.8071,
+      "step": 26
+    },
+    {
+      "epoch": 0.006790744466800805,
+      "grad_norm": 5.4865617752075195,
+      "learning_rate": 2.2632020117351218e-07,
+      "loss": 0.7797,
+      "step": 27
+    },
+    {
+      "epoch": 0.007042253521126761,
+      "grad_norm": 5.4088969230651855,
+      "learning_rate": 2.3470243084660523e-07,
+      "loss": 0.7925,
+      "step": 28
+    },
+    {
+      "epoch": 0.007293762575452716,
+      "grad_norm": 5.587332248687744,
+      "learning_rate": 2.430846605196983e-07,
+      "loss": 0.7799,
+      "step": 29
+    },
+    {
+      "epoch": 0.007545271629778672,
+      "grad_norm": 5.692927837371826,
+      "learning_rate": 2.5146689019279127e-07,
+      "loss": 0.8255,
+      "step": 30
+    },
+    {
+      "epoch": 0.007796780684104628,
+      "grad_norm": 5.381591796875,
+      "learning_rate": 2.598491198658843e-07,
+      "loss": 0.8173,
+      "step": 31
+    },
+    {
+      "epoch": 0.008048289738430584,
+      "grad_norm": 5.045378684997559,
+      "learning_rate": 2.6823134953897737e-07,
+      "loss": 0.7774,
+      "step": 32
+    },
+    {
+      "epoch": 0.00829979879275654,
+      "grad_norm": 4.887575626373291,
+      "learning_rate": 2.766135792120704e-07,
+      "loss": 0.7928,
+      "step": 33
+    },
+    {
+      "epoch": 0.008551307847082495,
+      "grad_norm": 4.738620281219482,
+      "learning_rate": 2.8499580888516347e-07,
+      "loss": 0.768,
+      "step": 34
+    },
+    {
+      "epoch": 0.008802816901408451,
+      "grad_norm": 4.525857448577881,
+      "learning_rate": 2.933780385582565e-07,
+      "loss": 0.7588,
+      "step": 35
+    },
+    {
+      "epoch": 0.009054325955734407,
+      "grad_norm": 4.540877342224121,
+      "learning_rate": 3.0176026823134957e-07,
+      "loss": 0.7746,
+      "step": 36
+    },
+    {
+      "epoch": 0.009305835010060362,
+      "grad_norm": 4.293127059936523,
+      "learning_rate": 3.101424979044426e-07,
+      "loss": 0.7452,
+      "step": 37
+    },
+    {
+      "epoch": 0.009557344064386318,
+      "grad_norm": 4.619227409362793,
+      "learning_rate": 3.185247275775356e-07,
+      "loss": 0.764,
+      "step": 38
+    },
+    {
+      "epoch": 0.009808853118712274,
+      "grad_norm": 4.370224475860596,
+      "learning_rate": 3.269069572506287e-07,
+      "loss": 0.7549,
+      "step": 39
+    },
+    {
+      "epoch": 0.01006036217303823,
+      "grad_norm": 4.469829082489014,
+      "learning_rate": 3.3528918692372177e-07,
+      "loss": 0.7507,
+      "step": 40
+    },
+    {
+      "epoch": 0.010311871227364185,
+      "grad_norm": 4.509310722351074,
+      "learning_rate": 3.4367141659681476e-07,
+      "loss": 0.753,
+      "step": 41
+    },
+    {
+      "epoch": 0.01056338028169014,
+      "grad_norm": 4.393963813781738,
+      "learning_rate": 3.5205364626990787e-07,
+      "loss": 0.7567,
+      "step": 42
+    },
+    {
+      "epoch": 0.010814889336016096,
+      "grad_norm": 4.315878391265869,
+      "learning_rate": 3.6043587594300086e-07,
+      "loss": 0.7331,
+      "step": 43
+    },
+    {
+      "epoch": 0.011066398390342052,
+      "grad_norm": 3.5298051834106445,
+      "learning_rate": 3.6881810561609386e-07,
+      "loss": 0.7346,
+      "step": 44
+    },
+    {
+      "epoch": 0.011317907444668008,
+      "grad_norm": 2.8748891353607178,
+      "learning_rate": 3.7720033528918696e-07,
+      "loss": 0.6869,
+      "step": 45
+    },
+    {
+      "epoch": 0.011569416498993963,
+      "grad_norm": 2.6105713844299316,
+      "learning_rate": 3.8558256496227996e-07,
+      "loss": 0.7293,
+      "step": 46
+    },
+    {
+      "epoch": 0.011820925553319919,
+      "grad_norm": 2.637141704559326,
+      "learning_rate": 3.9396479463537306e-07,
+      "loss": 0.7522,
+      "step": 47
+    },
+    {
+      "epoch": 0.012072434607645875,
+      "grad_norm": 2.4601848125457764,
+      "learning_rate": 4.023470243084661e-07,
+      "loss": 0.7251,
+      "step": 48
+    },
+    {
+      "epoch": 0.01232394366197183,
+      "grad_norm": 2.4165666103363037,
+      "learning_rate": 4.107292539815591e-07,
+      "loss": 0.6959,
+      "step": 49
+    },
+    {
+      "epoch": 0.012575452716297788,
+      "grad_norm": 2.4845943450927734,
+      "learning_rate": 4.191114836546522e-07,
+      "loss": 0.7036,
+      "step": 50
+    },
+    {
+      "epoch": 0.012826961770623743,
+      "grad_norm": 2.3925764560699463,
+      "learning_rate": 4.274937133277452e-07,
+      "loss": 0.719,
+      "step": 51
+    },
+    {
+      "epoch": 0.013078470824949699,
+      "grad_norm": 2.3824775218963623,
+      "learning_rate": 4.358759430008382e-07,
+      "loss": 0.673,
+      "step": 52
+    },
+    {
+      "epoch": 0.013329979879275655,
+      "grad_norm": 2.228027820587158,
+      "learning_rate": 4.442581726739313e-07,
+      "loss": 0.7147,
+      "step": 53
+    },
+    {
+      "epoch": 0.01358148893360161,
+      "grad_norm": 2.108171224594116,
+      "learning_rate": 4.5264040234702435e-07,
+      "loss": 0.712,
+      "step": 54
+    },
+    {
+      "epoch": 0.013832997987927566,
+      "grad_norm": 1.9733554124832153,
+      "learning_rate": 4.610226320201174e-07,
+      "loss": 0.6986,
+      "step": 55
+    },
+    {
+      "epoch": 0.014084507042253521,
+      "grad_norm": 1.9992328882217407,
+      "learning_rate": 4.6940486169321045e-07,
+      "loss": 0.6853,
+      "step": 56
+    },
+    {
+      "epoch": 0.014336016096579477,
+      "grad_norm": 1.991721749305725,
+      "learning_rate": 4.777870913663035e-07,
+      "loss": 0.6935,
+      "step": 57
+    },
+    {
+      "epoch": 0.014587525150905433,
+      "grad_norm": 1.8876029253005981,
+      "learning_rate": 4.861693210393966e-07,
+      "loss": 0.686,
+      "step": 58
+    },
+    {
+      "epoch": 0.014839034205231388,
+      "grad_norm": 1.6275702714920044,
+      "learning_rate": 4.945515507124896e-07,
+      "loss": 0.6991,
+      "step": 59
+    },
+    {
+      "epoch": 0.015090543259557344,
+      "grad_norm": 1.4490710496902466,
+      "learning_rate": 5.029337803855825e-07,
+      "loss": 0.7004,
+      "step": 60
+    },
+    {
+      "epoch": 0.0153420523138833,
+      "grad_norm": 1.453107476234436,
+      "learning_rate": 5.113160100586757e-07,
+      "loss": 0.6786,
+      "step": 61
+    },
+    {
+      "epoch": 0.015593561368209255,
+      "grad_norm": 1.5337774753570557,
+      "learning_rate": 5.196982397317686e-07,
+      "loss": 0.6438,
+      "step": 62
+    },
+    {
+      "epoch": 0.01584507042253521,
+      "grad_norm": 1.4964603185653687,
+      "learning_rate": 5.280804694048618e-07,
+      "loss": 0.6416,
+      "step": 63
+    },
+    {
+      "epoch": 0.01609657947686117,
+      "grad_norm": 1.6416826248168945,
+      "learning_rate": 5.364626990779547e-07,
+      "loss": 0.6863,
+      "step": 64
+    },
+    {
+      "epoch": 0.016348088531187122,
+      "grad_norm": 1.6651757955551147,
+      "learning_rate": 5.448449287510478e-07,
+      "loss": 0.6805,
+      "step": 65
+    },
+    {
+      "epoch": 0.01659959758551308,
+      "grad_norm": 1.5026111602783203,
+      "learning_rate": 5.532271584241408e-07,
+      "loss": 0.6329,
+      "step": 66
+    },
+    {
+      "epoch": 0.016851106639839034,
+      "grad_norm": 1.5439717769622803,
+      "learning_rate": 5.616093880972339e-07,
+      "loss": 0.6221,
+      "step": 67
+    },
+    {
+      "epoch": 0.01710261569416499,
+      "grad_norm": 1.5488256216049194,
+      "learning_rate": 5.699916177703269e-07,
+      "loss": 0.657,
+      "step": 68
+    },
+    {
+      "epoch": 0.017354124748490945,
+      "grad_norm": 1.4593708515167236,
+      "learning_rate": 5.7837384744342e-07,
+      "loss": 0.6557,
+      "step": 69
+    },
+    {
+      "epoch": 0.017605633802816902,
+      "grad_norm": 1.368078589439392,
+      "learning_rate": 5.86756077116513e-07,
+      "loss": 0.6378,
+      "step": 70
+    },
+    {
+      "epoch": 0.017857142857142856,
+      "grad_norm": 1.2166882753372192,
+      "learning_rate": 5.951383067896061e-07,
+      "loss": 0.6182,
+      "step": 71
+    },
+    {
+      "epoch": 0.018108651911468814,
+      "grad_norm": 1.1713978052139282,
+      "learning_rate": 6.035205364626991e-07,
+      "loss": 0.6465,
+      "step": 72
+    },
+    {
+      "epoch": 0.018360160965794767,
+      "grad_norm": 1.0981969833374023,
+      "learning_rate": 6.119027661357922e-07,
+      "loss": 0.6062,
+      "step": 73
+    },
+    {
+      "epoch": 0.018611670020120725,
+      "grad_norm": 1.0137531757354736,
+      "learning_rate": 6.202849958088852e-07,
+      "loss": 0.6123,
+      "step": 74
+    },
+    {
+      "epoch": 0.01886317907444668,
+      "grad_norm": 1.0327494144439697,
+      "learning_rate": 6.286672254819783e-07,
+      "loss": 0.6315,
+      "step": 75
+    },
+    {
+      "epoch": 0.019114688128772636,
+      "grad_norm": 0.9916748404502869,
+      "learning_rate": 6.370494551550712e-07,
+      "loss": 0.6504,
+      "step": 76
+    },
+    {
+      "epoch": 0.01936619718309859,
+      "grad_norm": 0.9969479441642761,
+      "learning_rate": 6.454316848281643e-07,
+      "loss": 0.6598,
+      "step": 77
+    },
+    {
+      "epoch": 0.019617706237424547,
+      "grad_norm": 0.9439221620559692,
+      "learning_rate": 6.538139145012574e-07,
+      "loss": 0.6136,
+      "step": 78
+    },
+    {
+      "epoch": 0.0198692152917505,
+      "grad_norm": 0.9918829202651978,
+      "learning_rate": 6.621961441743505e-07,
+      "loss": 0.6273,
+      "step": 79
+    },
+    {
+      "epoch": 0.02012072434607646,
+      "grad_norm": 0.8537020683288574,
+      "learning_rate": 6.705783738474435e-07,
+      "loss": 0.6015,
+      "step": 80
+    },
+    {
+      "epoch": 0.020372233400402416,
+      "grad_norm": 0.8386716246604919,
+      "learning_rate": 6.789606035205365e-07,
+      "loss": 0.6091,
+      "step": 81
+    },
+    {
+      "epoch": 0.02062374245472837,
+      "grad_norm": 0.8146318793296814,
+      "learning_rate": 6.873428331936295e-07,
+      "loss": 0.6225,
+      "step": 82
+    },
+    {
+      "epoch": 0.020875251509054327,
+      "grad_norm": 0.7194071412086487,
+      "learning_rate": 6.957250628667227e-07,
+      "loss": 0.6023,
+      "step": 83
+    },
+    {
+      "epoch": 0.02112676056338028,
+      "grad_norm": 0.672926127910614,
+      "learning_rate": 7.041072925398157e-07,
+      "loss": 0.572,
+      "step": 84
+    },
+    {
+      "epoch": 0.02137826961770624,
+      "grad_norm": 0.729033887386322,
+      "learning_rate": 7.124895222129087e-07,
+      "loss": 0.6156,
+      "step": 85
+    },
+    {
+      "epoch": 0.021629778672032193,
+      "grad_norm": 0.6650065779685974,
+      "learning_rate": 7.208717518860017e-07,
+      "loss": 0.5969,
+      "step": 86
+    },
+    {
+      "epoch": 0.02188128772635815,
+      "grad_norm": 0.7430527806282043,
+      "learning_rate": 7.292539815590948e-07,
+      "loss": 0.583,
+      "step": 87
+    },
+    {
+      "epoch": 0.022132796780684104,
+      "grad_norm": 0.7079744338989258,
+      "learning_rate": 7.376362112321877e-07,
+      "loss": 0.5836,
+      "step": 88
+    },
+    {
+      "epoch": 0.02238430583501006,
+      "grad_norm": 0.6756497025489807,
+      "learning_rate": 7.460184409052809e-07,
+      "loss": 0.5745,
+      "step": 89
+    },
+    {
+      "epoch": 0.022635814889336015,
+      "grad_norm": 0.655807614326477,
+      "learning_rate": 7.544006705783739e-07,
+      "loss": 0.5946,
+      "step": 90
+    },
+    {
+      "epoch": 0.022887323943661973,
+      "grad_norm": 0.5830637216567993,
+      "learning_rate": 7.62782900251467e-07,
+      "loss": 0.5515,
+      "step": 91
+    },
+    {
+      "epoch": 0.023138832997987926,
+      "grad_norm": 0.6083541512489319,
+      "learning_rate": 7.711651299245599e-07,
+      "loss": 0.6012,
+      "step": 92
+    },
+    {
+      "epoch": 0.023390342052313884,
+      "grad_norm": 0.5965361595153809,
+      "learning_rate": 7.79547359597653e-07,
+      "loss": 0.6039,
+      "step": 93
+    },
+    {
+      "epoch": 0.023641851106639838,
+      "grad_norm": 0.5302997827529907,
+      "learning_rate": 7.879295892707461e-07,
+      "loss": 0.5848,
+      "step": 94
+    },
+    {
+      "epoch": 0.023893360160965795,
+      "grad_norm": 0.6046608090400696,
+      "learning_rate": 7.963118189438392e-07,
+      "loss": 0.5951,
+      "step": 95
+    },
+    {
+      "epoch": 0.02414486921529175,
+      "grad_norm": 0.5611419677734375,
+      "learning_rate": 8.046940486169322e-07,
+      "loss": 0.545,
+      "step": 96
+    },
+    {
+      "epoch": 0.024396378269617706,
+      "grad_norm": 0.5234445929527283,
+      "learning_rate": 8.130762782900252e-07,
+      "loss": 0.5226,
+      "step": 97
+    },
+    {
+      "epoch": 0.02464788732394366,
+      "grad_norm": 0.564545750617981,
+      "learning_rate": 8.214585079631182e-07,
+      "loss": 0.5596,
+      "step": 98
+    },
+    {
+      "epoch": 0.024899396378269618,
+      "grad_norm": 0.5597972273826599,
+      "learning_rate": 8.298407376362114e-07,
+      "loss": 0.5828,
+      "step": 99
+    },
+    {
+      "epoch": 0.025150905432595575,
+      "grad_norm": 0.5264250040054321,
+      "learning_rate": 8.382229673093044e-07,
+      "loss": 0.5531,
+      "step": 100
+    },
+    {
+      "epoch": 0.02540241448692153,
+      "grad_norm": 0.5365298390388489,
+      "learning_rate": 8.466051969823974e-07,
+      "loss": 0.5539,
+      "step": 101
+    },
+    {
+      "epoch": 0.025653923541247486,
+      "grad_norm": 0.5420358180999756,
+      "learning_rate": 8.549874266554904e-07,
+      "loss": 0.5675,
+      "step": 102
+    },
+    {
+      "epoch": 0.02590543259557344,
+      "grad_norm": 0.48730742931365967,
+      "learning_rate": 8.633696563285835e-07,
+      "loss": 0.5321,
+      "step": 103
+    },
+    {
+      "epoch": 0.026156941649899398,
+      "grad_norm": 0.47938722372055054,
+      "learning_rate": 8.717518860016764e-07,
+      "loss": 0.5236,
+      "step": 104
+    },
+    {
+      "epoch": 0.02640845070422535,
+      "grad_norm": 0.4556962549686432,
+      "learning_rate": 8.801341156747697e-07,
+      "loss": 0.5348,
+      "step": 105
+    },
+    {
+      "epoch": 0.02665995975855131,
+      "grad_norm": 0.5305030941963196,
+      "learning_rate": 8.885163453478626e-07,
+      "loss": 0.5567,
+      "step": 106
+    },
+    {
+      "epoch": 0.026911468812877263,
+      "grad_norm": 0.5030955672264099,
+      "learning_rate": 8.968985750209557e-07,
+      "loss": 0.5235,
+      "step": 107
+    },
+    {
+      "epoch": 0.02716297786720322,
+      "grad_norm": 0.45537424087524414,
+      "learning_rate": 9.052808046940487e-07,
+      "loss": 0.5289,
+      "step": 108
+    },
+    {
+      "epoch": 0.027414486921529174,
+      "grad_norm": 0.4804135859012604,
+      "learning_rate": 9.136630343671417e-07,
+      "loss": 0.5481,
+      "step": 109
+    },
+    {
+      "epoch": 0.02766599597585513,
+      "grad_norm": 0.4467982351779938,
+      "learning_rate": 9.220452640402348e-07,
+      "loss": 0.5437,
+      "step": 110
+    },
+    {
+      "epoch": 0.027917505030181086,
+      "grad_norm": 0.5055573582649231,
+      "learning_rate": 9.304274937133279e-07,
+      "loss": 0.5313,
+      "step": 111
+    },
+    {
+      "epoch": 0.028169014084507043,
+      "grad_norm": 0.5437252521514893,
+      "learning_rate": 9.388097233864209e-07,
+      "loss": 0.538,
+      "step": 112
+    },
+    {
+      "epoch": 0.028420523138832997,
+      "grad_norm": 0.44438403844833374,
+      "learning_rate": 9.471919530595138e-07,
+      "loss": 0.5469,
+      "step": 113
+    },
+    {
+      "epoch": 0.028672032193158954,
+      "grad_norm": 0.45675307512283325,
+      "learning_rate": 9.55574182732607e-07,
+      "loss": 0.5007,
+      "step": 114
+    },
+    {
+      "epoch": 0.028923541247484908,
+      "grad_norm": 0.44954171776771545,
+      "learning_rate": 9.639564124056998e-07,
+      "loss": 0.5704,
+      "step": 115
+    },
+    {
+      "epoch": 0.029175050301810865,
+      "grad_norm": 0.4356625974178314,
+      "learning_rate": 9.72338642078793e-07,
+      "loss": 0.5339,
+      "step": 116
+    },
+    {
+      "epoch": 0.02942655935613682,
+      "grad_norm": 0.44852906465530396,
+      "learning_rate": 9.807208717518862e-07,
+      "loss": 0.5344,
+      "step": 117
+    },
+    {
+      "epoch": 0.029678068410462777,
+      "grad_norm": 0.5255141258239746,
+      "learning_rate": 9.891031014249792e-07,
+      "loss": 0.5457,
+      "step": 118
+    },
+    {
+      "epoch": 0.02992957746478873,
+      "grad_norm": 0.4613071084022522,
+      "learning_rate": 9.97485331098072e-07,
+      "loss": 0.566,
+      "step": 119
+    },
+    {
+      "epoch": 0.030181086519114688,
+      "grad_norm": 0.4343860149383545,
+      "learning_rate": 1.005867560771165e-06,
+      "loss": 0.5314,
+      "step": 120
+    },
+    {
+      "epoch": 0.030432595573440645,
+      "grad_norm": 0.4519299864768982,
+      "learning_rate": 1.0142497904442584e-06,
+      "loss": 0.5543,
+      "step": 121
+    },
+    {
+      "epoch": 0.0306841046277666,
+      "grad_norm": 0.45609790086746216,
+      "learning_rate": 1.0226320201173514e-06,
+      "loss": 0.5292,
+      "step": 122
+    },
+    {
+      "epoch": 0.030935613682092557,
+      "grad_norm": 0.449398010969162,
+      "learning_rate": 1.0310142497904444e-06,
+      "loss": 0.5476,
+      "step": 123
+    },
+    {
+      "epoch": 0.03118712273641851,
+      "grad_norm": 0.45544400811195374,
+      "learning_rate": 1.0393964794635373e-06,
+      "loss": 0.5264,
+      "step": 124
+    },
+    {
+      "epoch": 0.031438631790744465,
+      "grad_norm": 0.4554864466190338,
+      "learning_rate": 1.0477787091366303e-06,
+      "loss": 0.5316,
+      "step": 125
+    },
+    {
+      "epoch": 0.03169014084507042,
+      "grad_norm": 0.42402562499046326,
+      "learning_rate": 1.0561609388097236e-06,
+      "loss": 0.5422,
+      "step": 126
+    },
+    {
+      "epoch": 0.03194164989939638,
+      "grad_norm": 0.47918954491615295,
+      "learning_rate": 1.0645431684828166e-06,
+      "loss": 0.5639,
+      "step": 127
+    },
+    {
+      "epoch": 0.03219315895372234,
+      "grad_norm": 0.44865623116493225,
+      "learning_rate": 1.0729253981559095e-06,
+      "loss": 0.5292,
+      "step": 128
+    },
+    {
+      "epoch": 0.03244466800804829,
+      "grad_norm": 0.4049345850944519,
+      "learning_rate": 1.0813076278290025e-06,
+      "loss": 0.5318,
+      "step": 129
+    },
+    {
+      "epoch": 0.032696177062374245,
+      "grad_norm": 0.4720096290111542,
+      "learning_rate": 1.0896898575020956e-06,
+      "loss": 0.5059,
+      "step": 130
+    },
+    {
+      "epoch": 0.0329476861167002,
+      "grad_norm": 0.4590936005115509,
+      "learning_rate": 1.0980720871751886e-06,
+      "loss": 0.4943,
+      "step": 131
+    },
+    {
+      "epoch": 0.03319919517102616,
+      "grad_norm": 0.4278355538845062,
+      "learning_rate": 1.1064543168482817e-06,
+      "loss": 0.5273,
+      "step": 132
+    },
+    {
+      "epoch": 0.03345070422535211,
+      "grad_norm": 0.4701170325279236,
+      "learning_rate": 1.1148365465213747e-06,
+      "loss": 0.5292,
+      "step": 133
+    },
+    {
+      "epoch": 0.03370221327967807,
+      "grad_norm": 0.46045827865600586,
+      "learning_rate": 1.1232187761944678e-06,
+      "loss": 0.5208,
+      "step": 134
+    },
+    {
+      "epoch": 0.033953722334004025,
+      "grad_norm": 0.4302027225494385,
+      "learning_rate": 1.1316010058675608e-06,
+      "loss": 0.5287,
+      "step": 135
+    },
+    {
+      "epoch": 0.03420523138832998,
+      "grad_norm": 0.510970413684845,
+      "learning_rate": 1.1399832355406539e-06,
+      "loss": 0.5514,
+      "step": 136
+    },
+    {
+      "epoch": 0.03445674044265593,
+      "grad_norm": 0.4021950364112854,
+      "learning_rate": 1.148365465213747e-06,
+      "loss": 0.5356,
+      "step": 137
+    },
+    {
+      "epoch": 0.03470824949698189,
+      "grad_norm": 0.4260016083717346,
+      "learning_rate": 1.15674769488684e-06,
+      "loss": 0.5668,
+      "step": 138
+    },
+    {
+      "epoch": 0.03495975855130785,
+      "grad_norm": 0.4197461009025574,
+      "learning_rate": 1.165129924559933e-06,
+      "loss": 0.5274,
+      "step": 139
+    },
+    {
+      "epoch": 0.035211267605633804,
+      "grad_norm": 0.4209143817424774,
+      "learning_rate": 1.173512154233026e-06,
+      "loss": 0.5337,
+      "step": 140
+    },
+    {
+      "epoch": 0.03546277665995976,
+      "grad_norm": 0.4232824742794037,
+      "learning_rate": 1.1818943839061191e-06,
+      "loss": 0.5264,
+      "step": 141
+    },
+    {
+      "epoch": 0.03571428571428571,
+      "grad_norm": 0.45739907026290894,
+      "learning_rate": 1.1902766135792122e-06,
+      "loss": 0.5232,
+      "step": 142
+    },
+    {
+      "epoch": 0.03596579476861167,
+      "grad_norm": 0.4441063702106476,
+      "learning_rate": 1.1986588432523052e-06,
+      "loss": 0.5409,
+      "step": 143
+    },
+    {
+      "epoch": 0.03621730382293763,
+      "grad_norm": 0.43283092975616455,
+      "learning_rate": 1.2070410729253983e-06,
+      "loss": 0.5167,
+      "step": 144
+    },
+    {
+      "epoch": 0.036468812877263584,
+      "grad_norm": 0.4424374997615814,
+      "learning_rate": 1.2154233025984913e-06,
+      "loss": 0.5366,
+      "step": 145
+    },
+    {
+      "epoch": 0.036720321931589535,
+      "grad_norm": 0.5044413805007935,
+      "learning_rate": 1.2238055322715844e-06,
+      "loss": 0.5398,
+      "step": 146
+    },
+    {
+      "epoch": 0.03697183098591549,
+      "grad_norm": 0.46940240263938904,
+      "learning_rate": 1.2321877619446772e-06,
+      "loss": 0.499,
+      "step": 147
+    },
+    {
+      "epoch": 0.03722334004024145,
+      "grad_norm": 0.44987642765045166,
+      "learning_rate": 1.2405699916177705e-06,
+      "loss": 0.5116,
+      "step": 148
+    },
+    {
+      "epoch": 0.03747484909456741,
+      "grad_norm": 0.4250459671020508,
+      "learning_rate": 1.2489522212908635e-06,
+      "loss": 0.5055,
+      "step": 149
+    },
+    {
+      "epoch": 0.03772635814889336,
+      "grad_norm": 0.4617857038974762,
+      "learning_rate": 1.2573344509639566e-06,
+      "loss": 0.5347,
+      "step": 150
+    },
+    {
+      "epoch": 0.037977867203219315,
+      "grad_norm": 0.48853859305381775,
+      "learning_rate": 1.2657166806370496e-06,
+      "loss": 0.5067,
+      "step": 151
+    },
+    {
+      "epoch": 0.03822937625754527,
+      "grad_norm": 0.4387054145336151,
+      "learning_rate": 1.2740989103101425e-06,
+      "loss": 0.5091,
+      "step": 152
+    },
+    {
+      "epoch": 0.03848088531187123,
+      "grad_norm": 0.41969773173332214,
+      "learning_rate": 1.2824811399832357e-06,
+      "loss": 0.495,
+      "step": 153
+    },
+    {
+      "epoch": 0.03873239436619718,
+      "grad_norm": 0.43173882365226746,
+      "learning_rate": 1.2908633696563286e-06,
+      "loss": 0.4991,
+      "step": 154
+    },
+    {
+      "epoch": 0.03898390342052314,
+      "grad_norm": 0.4111224412918091,
+      "learning_rate": 1.2992455993294218e-06,
+      "loss": 0.5335,
+      "step": 155
+    },
+    {
+      "epoch": 0.039235412474849095,
+      "grad_norm": 0.40816769003868103,
+      "learning_rate": 1.3076278290025149e-06,
+      "loss": 0.5024,
+      "step": 156
+    },
+    {
+      "epoch": 0.03948692152917505,
+      "grad_norm": 0.41955387592315674,
+      "learning_rate": 1.3160100586756077e-06,
+      "loss": 0.5378,
+      "step": 157
+    },
+    {
+      "epoch": 0.039738430583501,
+      "grad_norm": 0.4666993021965027,
+      "learning_rate": 1.324392288348701e-06,
+      "loss": 0.5112,
+      "step": 158
+    },
+    {
+      "epoch": 0.03998993963782696,
+      "grad_norm": 0.41977325081825256,
+      "learning_rate": 1.3327745180217938e-06,
+      "loss": 0.5198,
+      "step": 159
+    },
+    {
+      "epoch": 0.04024144869215292,
+      "grad_norm": 0.4063010811805725,
+      "learning_rate": 1.341156747694887e-06,
+      "loss": 0.4893,
+      "step": 160
+    },
+    {
+      "epoch": 0.040492957746478875,
+      "grad_norm": 0.42728182673454285,
+      "learning_rate": 1.3495389773679801e-06,
+      "loss": 0.5049,
+      "step": 161
+    },
+    {
+      "epoch": 0.04074446680080483,
+      "grad_norm": 0.4450185298919678,
+      "learning_rate": 1.357921207041073e-06,
+      "loss": 0.5015,
+      "step": 162
+    },
+    {
+      "epoch": 0.04099597585513078,
+      "grad_norm": 0.3846238851547241,
+      "learning_rate": 1.3663034367141662e-06,
+      "loss": 0.515,
+      "step": 163
+    },
+    {
+      "epoch": 0.04124748490945674,
+      "grad_norm": 0.4138552248477936,
+      "learning_rate": 1.374685666387259e-06,
+      "loss": 0.5249,
+      "step": 164
+    },
+    {
+      "epoch": 0.0414989939637827,
+      "grad_norm": 0.4613569378852844,
+      "learning_rate": 1.383067896060352e-06,
+      "loss": 0.5409,
+      "step": 165
+    },
+    {
+      "epoch": 0.041750503018108655,
+      "grad_norm": 0.40616121888160706,
+      "learning_rate": 1.3914501257334454e-06,
+      "loss": 0.503,
+      "step": 166
+    },
+    {
+      "epoch": 0.042002012072434605,
+      "grad_norm": 0.4463784098625183,
+      "learning_rate": 1.3998323554065382e-06,
+      "loss": 0.5177,
+      "step": 167
+    },
+    {
+      "epoch": 0.04225352112676056,
+      "grad_norm": 0.40528833866119385,
+      "learning_rate": 1.4082145850796315e-06,
+      "loss": 0.4995,
+      "step": 168
+    },
+    {
+      "epoch": 0.04250503018108652,
+      "grad_norm": 0.41246384382247925,
+      "learning_rate": 1.4165968147527243e-06,
+      "loss": 0.5055,
+      "step": 169
+    },
+    {
+      "epoch": 0.04275653923541248,
+      "grad_norm": 0.4155355989933014,
+      "learning_rate": 1.4249790444258174e-06,
+      "loss": 0.5046,
+      "step": 170
+    },
+    {
+      "epoch": 0.04300804828973843,
+      "grad_norm": 0.4172171354293823,
+      "learning_rate": 1.4333612740989102e-06,
+      "loss": 0.4889,
+      "step": 171
+    },
+    {
+      "epoch": 0.043259557344064385,
+      "grad_norm": 0.40003055334091187,
+      "learning_rate": 1.4417435037720035e-06,
+      "loss": 0.4965,
+      "step": 172
+    },
+    {
+      "epoch": 0.04351106639839034,
+      "grad_norm": 0.48593074083328247,
+      "learning_rate": 1.4501257334450967e-06,
+      "loss": 0.4937,
+      "step": 173
+    },
+    {
+      "epoch": 0.0437625754527163,
+      "grad_norm": 0.41993674635887146,
+      "learning_rate": 1.4585079631181895e-06,
+      "loss": 0.4663,
+      "step": 174
+    },
+    {
+      "epoch": 0.04401408450704225,
+      "grad_norm": 0.40402930974960327,
+      "learning_rate": 1.4668901927912826e-06,
+      "loss": 0.5113,
+      "step": 175
+    },
+    {
+      "epoch": 0.04426559356136821,
+      "grad_norm": 0.4269852340221405,
+      "learning_rate": 1.4752724224643754e-06,
+      "loss": 0.5034,
+      "step": 176
+    },
+    {
+      "epoch": 0.044517102615694165,
+      "grad_norm": 0.41009941697120667,
+      "learning_rate": 1.4836546521374687e-06,
+      "loss": 0.51,
+      "step": 177
+    },
+    {
+      "epoch": 0.04476861167002012,
+      "grad_norm": 0.40047401189804077,
+      "learning_rate": 1.4920368818105617e-06,
+      "loss": 0.4719,
+      "step": 178
+    },
+    {
+      "epoch": 0.04502012072434608,
+      "grad_norm": 0.389114648103714,
+      "learning_rate": 1.5004191114836548e-06,
+      "loss": 0.4971,
+      "step": 179
+    },
+    {
+      "epoch": 0.04527162977867203,
+      "grad_norm": 0.41636624932289124,
+      "learning_rate": 1.5088013411567478e-06,
+      "loss": 0.4907,
+      "step": 180
+    },
+    {
+      "epoch": 0.04552313883299799,
+      "grad_norm": 0.41309916973114014,
+      "learning_rate": 1.5171835708298407e-06,
+      "loss": 0.509,
+      "step": 181
+    },
+    {
+      "epoch": 0.045774647887323945,
+      "grad_norm": 0.4215018153190613,
+      "learning_rate": 1.525565800502934e-06,
+      "loss": 0.4977,
+      "step": 182
+    },
+    {
+      "epoch": 0.0460261569416499,
+      "grad_norm": 0.4160129427909851,
+      "learning_rate": 1.533948030176027e-06,
+      "loss": 0.5073,
+      "step": 183
+    },
+    {
+      "epoch": 0.04627766599597585,
+      "grad_norm": 0.492434024810791,
+      "learning_rate": 1.5423302598491198e-06,
+      "loss": 0.4979,
+      "step": 184
+    },
+    {
+      "epoch": 0.04652917505030181,
+      "grad_norm": 0.3985655605792999,
+      "learning_rate": 1.550712489522213e-06,
+      "loss": 0.5291,
+      "step": 185
+    },
+    {
+      "epoch": 0.04678068410462777,
+      "grad_norm": 0.4607956111431122,
+      "learning_rate": 1.559094719195306e-06,
+      "loss": 0.5082,
+      "step": 186
+    },
+    {
+      "epoch": 0.047032193158953725,
+      "grad_norm": 0.4652690887451172,
+      "learning_rate": 1.5674769488683992e-06,
+      "loss": 0.5108,
+      "step": 187
+    },
+    {
+      "epoch": 0.047283702213279676,
+      "grad_norm": 0.43926411867141724,
+      "learning_rate": 1.5758591785414922e-06,
+      "loss": 0.5274,
+      "step": 188
+    },
+    {
+      "epoch": 0.04753521126760563,
+      "grad_norm": 0.4484438896179199,
+      "learning_rate": 1.584241408214585e-06,
+      "loss": 0.5397,
+      "step": 189
+    },
+    {
+      "epoch": 0.04778672032193159,
+      "grad_norm": 0.45970824360847473,
+      "learning_rate": 1.5926236378876783e-06,
+      "loss": 0.5249,
+      "step": 190
+    },
+    {
+      "epoch": 0.04803822937625755,
+      "grad_norm": 0.44468578696250916,
+      "learning_rate": 1.6010058675607712e-06,
+      "loss": 0.4686,
+      "step": 191
+    },
+    {
+      "epoch": 0.0482897384305835,
+      "grad_norm": 0.4506581127643585,
+      "learning_rate": 1.6093880972338644e-06,
+      "loss": 0.4777,
+      "step": 192
+    },
+    {
+      "epoch": 0.048541247484909456,
+      "grad_norm": 0.39604270458221436,
+      "learning_rate": 1.6177703269069575e-06,
+      "loss": 0.4839,
+      "step": 193
+    },
+    {
+      "epoch": 0.04879275653923541,
+      "grad_norm": 0.41182881593704224,
+      "learning_rate": 1.6261525565800503e-06,
+      "loss": 0.4888,
+      "step": 194
+    },
+    {
+      "epoch": 0.04904426559356137,
+      "grad_norm": 0.44034087657928467,
+      "learning_rate": 1.6345347862531436e-06,
+      "loss": 0.547,
+      "step": 195
+    },
+    {
+      "epoch": 0.04929577464788732,
+      "grad_norm": 0.453494131565094,
+      "learning_rate": 1.6429170159262364e-06,
+      "loss": 0.4995,
+      "step": 196
+    },
+    {
+      "epoch": 0.04954728370221328,
+      "grad_norm": 0.41761818528175354,
+      "learning_rate": 1.6512992455993297e-06,
+      "loss": 0.4797,
+      "step": 197
+    },
+    {
+      "epoch": 0.049798792756539235,
+      "grad_norm": 0.4482254385948181,
+      "learning_rate": 1.6596814752724227e-06,
+      "loss": 0.4607,
+      "step": 198
+    },
+    {
+      "epoch": 0.05005030181086519,
+      "grad_norm": 0.41836273670196533,
+      "learning_rate": 1.6680637049455156e-06,
+      "loss": 0.4702,
+      "step": 199
+    },
+    {
+      "epoch": 0.05030181086519115,
+      "grad_norm": 0.46722131967544556,
+      "learning_rate": 1.6764459346186088e-06,
+      "loss": 0.5276,
+      "step": 200
+    },
+    {
+      "epoch": 0.0505533199195171,
+      "grad_norm": 0.43717604875564575,
+      "learning_rate": 1.6848281642917017e-06,
+      "loss": 0.4681,
+      "step": 201
+    },
+    {
+      "epoch": 0.05080482897384306,
+      "grad_norm": 0.3926369249820709,
+      "learning_rate": 1.6932103939647947e-06,
+      "loss": 0.5017,
+      "step": 202
+    },
+    {
+      "epoch": 0.051056338028169015,
+      "grad_norm": 0.4253596365451813,
+      "learning_rate": 1.7015926236378878e-06,
+      "loss": 0.507,
+      "step": 203
+    },
+    {
+      "epoch": 0.05130784708249497,
+      "grad_norm": 0.4381323456764221,
+      "learning_rate": 1.7099748533109808e-06,
+      "loss": 0.4898,
+      "step": 204
+    },
+    {
+      "epoch": 0.05155935613682092,
+      "grad_norm": 0.4161489009857178,
+      "learning_rate": 1.718357082984074e-06,
+      "loss": 0.4943,
+      "step": 205
+    },
+    {
+      "epoch": 0.05181086519114688,
+      "grad_norm": 0.4436219036579132,
+      "learning_rate": 1.726739312657167e-06,
+      "loss": 0.4665,
+      "step": 206
+    },
+    {
+      "epoch": 0.05206237424547284,
+      "grad_norm": 0.43012455105781555,
+      "learning_rate": 1.73512154233026e-06,
+      "loss": 0.515,
+      "step": 207
+    },
+    {
+      "epoch": 0.052313883299798795,
+      "grad_norm": 0.4141829311847687,
+      "learning_rate": 1.7435037720033528e-06,
+      "loss": 0.5038,
+      "step": 208
+    },
+    {
+      "epoch": 0.052565392354124746,
+      "grad_norm": 0.41762787103652954,
+      "learning_rate": 1.751886001676446e-06,
+      "loss": 0.5124,
+      "step": 209
+    },
+    {
+      "epoch": 0.0528169014084507,
+      "grad_norm": 0.39720189571380615,
+      "learning_rate": 1.7602682313495393e-06,
+      "loss": 0.4578,
+      "step": 210
+    },
+    {
+      "epoch": 0.05306841046277666,
+      "grad_norm": 0.41066601872444153,
+      "learning_rate": 1.7686504610226322e-06,
+      "loss": 0.4708,
+      "step": 211
+    },
+    {
+      "epoch": 0.05331991951710262,
+      "grad_norm": 0.43149372935295105,
+      "learning_rate": 1.7770326906957252e-06,
+      "loss": 0.4894,
+      "step": 212
+    },
+    {
+      "epoch": 0.05357142857142857,
+      "grad_norm": 0.42833876609802246,
+      "learning_rate": 1.785414920368818e-06,
+      "loss": 0.4971,
+      "step": 213
+    },
+    {
+      "epoch": 0.053822937625754526,
+      "grad_norm": 0.43520480394363403,
+      "learning_rate": 1.7937971500419113e-06,
+      "loss": 0.518,
+      "step": 214
+    },
+    {
+      "epoch": 0.05407444668008048,
+      "grad_norm": 0.432537317276001,
+      "learning_rate": 1.8021793797150044e-06,
+      "loss": 0.4551,
+      "step": 215
+    },
+    {
+      "epoch": 0.05432595573440644,
+      "grad_norm": 0.4245152771472931,
+      "learning_rate": 1.8105616093880974e-06,
+      "loss": 0.4727,
+      "step": 216
+    },
+    {
+      "epoch": 0.05457746478873239,
+      "grad_norm": 0.4602232873439789,
+      "learning_rate": 1.8189438390611905e-06,
+      "loss": 0.4905,
+      "step": 217
+    },
+    {
+      "epoch": 0.05482897384305835,
+      "grad_norm": 0.4128243625164032,
+      "learning_rate": 1.8273260687342833e-06,
+      "loss": 0.4866,
+      "step": 218
+    },
+    {
+      "epoch": 0.055080482897384306,
+      "grad_norm": 0.44259390234947205,
+      "learning_rate": 1.8357082984073766e-06,
+      "loss": 0.4894,
+      "step": 219
+    },
+    {
+      "epoch": 0.05533199195171026,
+      "grad_norm": 0.46606069803237915,
+      "learning_rate": 1.8440905280804696e-06,
+      "loss": 0.5172,
+      "step": 220
+    },
+    {
+      "epoch": 0.05558350100603622,
+      "grad_norm": 0.4518153667449951,
+      "learning_rate": 1.8524727577535625e-06,
+      "loss": 0.5033,
+      "step": 221
+    },
+    {
+      "epoch": 0.05583501006036217,
+      "grad_norm": 0.4819628894329071,
+      "learning_rate": 1.8608549874266557e-06,
+      "loss": 0.488,
+      "step": 222
+    },
+    {
+      "epoch": 0.05608651911468813,
+      "grad_norm": 0.44667211174964905,
+      "learning_rate": 1.8692372170997485e-06,
+      "loss": 0.5062,
+      "step": 223
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 0.4287307560443878,
+      "learning_rate": 1.8776194467728418e-06,
+      "loss": 0.4432,
+      "step": 224
+    },
+    {
+      "epoch": 0.05658953722334004,
+      "grad_norm": 0.46748512983322144,
+      "learning_rate": 1.8860016764459349e-06,
+      "loss": 0.4407,
+      "step": 225
+    },
+    {
+      "epoch": 0.056841046277665994,
+      "grad_norm": 0.44293415546417236,
+      "learning_rate": 1.8943839061190277e-06,
+      "loss": 0.4634,
+      "step": 226
+    },
+    {
+      "epoch": 0.05709255533199195,
+      "grad_norm": 0.41884273290634155,
+      "learning_rate": 1.902766135792121e-06,
+      "loss": 0.4978,
+      "step": 227
+    },
+    {
+      "epoch": 0.05734406438631791,
+      "grad_norm": 0.4945652186870575,
+      "learning_rate": 1.911148365465214e-06,
+      "loss": 0.493,
+      "step": 228
+    },
+    {
+      "epoch": 0.057595573440643866,
+      "grad_norm": 0.4238063097000122,
+      "learning_rate": 1.919530595138307e-06,
+      "loss": 0.4558,
+      "step": 229
+    },
+    {
+      "epoch": 0.057847082494969816,
+      "grad_norm": 0.39406704902648926,
+      "learning_rate": 1.9279128248113997e-06,
+      "loss": 0.479,
+      "step": 230
+    },
+    {
+      "epoch": 0.058098591549295774,
+      "grad_norm": 0.4454141855239868,
+      "learning_rate": 1.936295054484493e-06,
+      "loss": 0.4817,
+      "step": 231
+    },
+    {
+      "epoch": 0.05835010060362173,
+      "grad_norm": 0.45866918563842773,
+      "learning_rate": 1.944677284157586e-06,
+      "loss": 0.461,
+      "step": 232
+    },
+    {
+      "epoch": 0.05860160965794769,
+      "grad_norm": 0.41936853528022766,
+      "learning_rate": 1.953059513830679e-06,
+      "loss": 0.4745,
+      "step": 233
+    },
+    {
+      "epoch": 0.05885311871227364,
+      "grad_norm": 0.4199956953525543,
+      "learning_rate": 1.9614417435037723e-06,
+      "loss": 0.486,
+      "step": 234
+    },
+    {
+      "epoch": 0.059104627766599596,
+      "grad_norm": 0.4122017025947571,
+      "learning_rate": 1.969823973176865e-06,
+      "loss": 0.5059,
+      "step": 235
+    },
+    {
+      "epoch": 0.059356136820925554,
+      "grad_norm": 0.4136897027492523,
+      "learning_rate": 1.9782062028499584e-06,
+      "loss": 0.4689,
+      "step": 236
+    },
+    {
+      "epoch": 0.05960764587525151,
+      "grad_norm": 0.44036155939102173,
+      "learning_rate": 1.9865884325230515e-06,
+      "loss": 0.4885,
+      "step": 237
+    },
+    {
+      "epoch": 0.05985915492957746,
+      "grad_norm": 0.39697256684303284,
+      "learning_rate": 1.994970662196144e-06,
+      "loss": 0.4923,
+      "step": 238
+    },
+    {
+      "epoch": 0.06011066398390342,
+      "grad_norm": 0.3980274200439453,
+      "learning_rate": 2.0033528918692376e-06,
+      "loss": 0.4702,
+      "step": 239
+    },
+    {
+      "epoch": 0.060362173038229376,
+      "grad_norm": 0.4443422257900238,
+      "learning_rate": 2.01173512154233e-06,
+      "loss": 0.4845,
+      "step": 240
+    },
+    {
+      "epoch": 0.060613682092555334,
+      "grad_norm": 0.39754951000213623,
+      "learning_rate": 2.0201173512154237e-06,
+      "loss": 0.4652,
+      "step": 241
+    },
+    {
+      "epoch": 0.06086519114688129,
+      "grad_norm": 0.41801464557647705,
+      "learning_rate": 2.0284995808885167e-06,
+      "loss": 0.4821,
+      "step": 242
+    },
+    {
+      "epoch": 0.06111670020120724,
+      "grad_norm": 0.45718705654144287,
+      "learning_rate": 2.0368818105616093e-06,
+      "loss": 0.5005,
+      "step": 243
+    },
+    {
+      "epoch": 0.0613682092555332,
+      "grad_norm": 0.4293982684612274,
+      "learning_rate": 2.045264040234703e-06,
+      "loss": 0.5131,
+      "step": 244
+    },
+    {
+      "epoch": 0.061619718309859156,
+      "grad_norm": 0.4417547285556793,
+      "learning_rate": 2.0536462699077954e-06,
+      "loss": 0.4852,
+      "step": 245
+    },
+    {
+      "epoch": 0.061871227364185113,
+      "grad_norm": 0.43375566601753235,
+      "learning_rate": 2.062028499580889e-06,
+      "loss": 0.474,
+      "step": 246
+    },
+    {
+      "epoch": 0.062122736418511064,
+      "grad_norm": 0.4546149969100952,
+      "learning_rate": 2.070410729253982e-06,
+      "loss": 0.4667,
+      "step": 247
+    },
+    {
+      "epoch": 0.06237424547283702,
+      "grad_norm": 0.4586324095726013,
+      "learning_rate": 2.0787929589270746e-06,
+      "loss": 0.514,
+      "step": 248
+    },
+    {
+      "epoch": 0.06262575452716297,
+      "grad_norm": 0.40271812677383423,
+      "learning_rate": 2.087175188600168e-06,
+      "loss": 0.4615,
+      "step": 249
+    },
+    {
+      "epoch": 0.06287726358148893,
+      "grad_norm": 0.44942378997802734,
+      "learning_rate": 2.0955574182732607e-06,
+      "loss": 0.4823,
+      "step": 250
+    },
+    {
+      "epoch": 0.06312877263581489,
+      "grad_norm": 0.3924196660518646,
+      "learning_rate": 2.1039396479463537e-06,
+      "loss": 0.4828,
+      "step": 251
+    },
+    {
+      "epoch": 0.06338028169014084,
+      "grad_norm": 0.4237384796142578,
+      "learning_rate": 2.112321877619447e-06,
+      "loss": 0.4931,
+      "step": 252
+    },
+    {
+      "epoch": 0.0636317907444668,
+      "grad_norm": 0.40384024381637573,
+      "learning_rate": 2.12070410729254e-06,
+      "loss": 0.473,
+      "step": 253
+    },
+    {
+      "epoch": 0.06388329979879276,
+      "grad_norm": 0.4171174466609955,
+      "learning_rate": 2.1290863369656333e-06,
+      "loss": 0.4971,
+      "step": 254
+    },
+    {
+      "epoch": 0.06413480885311872,
+      "grad_norm": 0.43529245257377625,
+      "learning_rate": 2.137468566638726e-06,
+      "loss": 0.4433,
+      "step": 255
+    },
+    {
+      "epoch": 0.06438631790744467,
+      "grad_norm": 0.46728262305259705,
+      "learning_rate": 2.145850796311819e-06,
+      "loss": 0.4561,
+      "step": 256
+    },
+    {
+      "epoch": 0.06463782696177062,
+      "grad_norm": 0.5025231242179871,
+      "learning_rate": 2.1542330259849124e-06,
+      "loss": 0.4916,
+      "step": 257
+    },
+    {
+      "epoch": 0.06488933601609657,
+      "grad_norm": 0.3811725378036499,
+      "learning_rate": 2.162615255658005e-06,
+      "loss": 0.4842,
+      "step": 258
+    },
+    {
+      "epoch": 0.06514084507042253,
+      "grad_norm": 0.4305613040924072,
+      "learning_rate": 2.1709974853310985e-06,
+      "loss": 0.4609,
+      "step": 259
+    },
+    {
+      "epoch": 0.06539235412474849,
+      "grad_norm": 0.42544025182724,
+      "learning_rate": 2.179379715004191e-06,
+      "loss": 0.4745,
+      "step": 260
+    },
+    {
+      "epoch": 0.06564386317907445,
+      "grad_norm": 0.42461109161376953,
+      "learning_rate": 2.1877619446772842e-06,
+      "loss": 0.4945,
+      "step": 261
+    },
+    {
+      "epoch": 0.0658953722334004,
+      "grad_norm": 0.42680662870407104,
+      "learning_rate": 2.1961441743503773e-06,
+      "loss": 0.4918,
+      "step": 262
+    },
+    {
+      "epoch": 0.06614688128772636,
+      "grad_norm": 0.42169326543807983,
+      "learning_rate": 2.2045264040234703e-06,
+      "loss": 0.4724,
+      "step": 263
+    },
+    {
+      "epoch": 0.06639839034205232,
+      "grad_norm": 0.4097406268119812,
+      "learning_rate": 2.2129086336965634e-06,
+      "loss": 0.4682,
+      "step": 264
+    },
+    {
+      "epoch": 0.06664989939637828,
+      "grad_norm": 0.5153217911720276,
+      "learning_rate": 2.2212908633696564e-06,
+      "loss": 0.4742,
+      "step": 265
+    },
+    {
+      "epoch": 0.06690140845070422,
+      "grad_norm": 0.4730076491832733,
+      "learning_rate": 2.2296730930427495e-06,
+      "loss": 0.4757,
+      "step": 266
+    },
+    {
+      "epoch": 0.06715291750503018,
+      "grad_norm": 0.48586782813072205,
+      "learning_rate": 2.2380553227158425e-06,
+      "loss": 0.4851,
+      "step": 267
+    },
+    {
+      "epoch": 0.06740442655935613,
+      "grad_norm": 0.4455511271953583,
+      "learning_rate": 2.2464375523889356e-06,
+      "loss": 0.4448,
+      "step": 268
+    },
+    {
+      "epoch": 0.06765593561368209,
+      "grad_norm": 0.5135384798049927,
+      "learning_rate": 2.2548197820620286e-06,
+      "loss": 0.4944,
+      "step": 269
+    },
+    {
+      "epoch": 0.06790744466800805,
+      "grad_norm": 0.4233526289463043,
+      "learning_rate": 2.2632020117351217e-06,
+      "loss": 0.4523,
+      "step": 270
+    },
+    {
+      "epoch": 0.068158953722334,
+      "grad_norm": 0.4129721522331238,
+      "learning_rate": 2.2715842414082147e-06,
+      "loss": 0.4809,
+      "step": 271
+    },
+    {
+      "epoch": 0.06841046277665996,
+      "grad_norm": 0.4507489800453186,
+      "learning_rate": 2.2799664710813078e-06,
+      "loss": 0.4521,
+      "step": 272
+    },
+    {
+      "epoch": 0.06866197183098592,
+      "grad_norm": 0.5152677297592163,
+      "learning_rate": 2.288348700754401e-06,
+      "loss": 0.4878,
+      "step": 273
+    },
+    {
+      "epoch": 0.06891348088531186,
+      "grad_norm": 0.4466034471988678,
+      "learning_rate": 2.296730930427494e-06,
+      "loss": 0.4493,
+      "step": 274
+    },
+    {
+      "epoch": 0.06916498993963782,
+      "grad_norm": 0.4332417845726013,
+      "learning_rate": 2.305113160100587e-06,
+      "loss": 0.4904,
+      "step": 275
+    },
+    {
+      "epoch": 0.06941649899396378,
+      "grad_norm": 0.4686703085899353,
+      "learning_rate": 2.31349538977368e-06,
+      "loss": 0.5085,
+      "step": 276
+    },
+    {
+      "epoch": 0.06966800804828974,
+      "grad_norm": 0.46179959177970886,
+      "learning_rate": 2.321877619446773e-06,
+      "loss": 0.4841,
+      "step": 277
+    },
+    {
+      "epoch": 0.0699195171026157,
+      "grad_norm": 0.43604379892349243,
+      "learning_rate": 2.330259849119866e-06,
+      "loss": 0.4666,
+      "step": 278
+    },
+    {
+      "epoch": 0.07017102615694165,
+      "grad_norm": 0.4303692877292633,
+      "learning_rate": 2.338642078792959e-06,
+      "loss": 0.4745,
+      "step": 279
+    },
+    {
+      "epoch": 0.07042253521126761,
+      "grad_norm": 0.46457597613334656,
+      "learning_rate": 2.347024308466052e-06,
+      "loss": 0.5055,
+      "step": 280
+    },
+    {
+      "epoch": 0.07067404426559357,
+      "grad_norm": 0.3907102942466736,
+      "learning_rate": 2.355406538139145e-06,
+      "loss": 0.4719,
+      "step": 281
+    },
+    {
+      "epoch": 0.07092555331991952,
+      "grad_norm": 0.4557718336582184,
+      "learning_rate": 2.3637887678122383e-06,
+      "loss": 0.4796,
+      "step": 282
+    },
+    {
+      "epoch": 0.07117706237424547,
+      "grad_norm": 0.4310013949871063,
+      "learning_rate": 2.3721709974853313e-06,
+      "loss": 0.4753,
+      "step": 283
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 0.4076017737388611,
+      "learning_rate": 2.3805532271584244e-06,
+      "loss": 0.4697,
+      "step": 284
+    },
+    {
+      "epoch": 0.07168008048289738,
+      "grad_norm": 0.4143841564655304,
+      "learning_rate": 2.3889354568315174e-06,
+      "loss": 0.4839,
+      "step": 285
+    },
+    {
+      "epoch": 0.07193158953722334,
+      "grad_norm": 0.4165581166744232,
+      "learning_rate": 2.3973176865046105e-06,
+      "loss": 0.4777,
+      "step": 286
+    },
+    {
+      "epoch": 0.0721830985915493,
+      "grad_norm": 0.3928240239620209,
+      "learning_rate": 2.4056999161777035e-06,
+      "loss": 0.4581,
+      "step": 287
+    },
+    {
+      "epoch": 0.07243460764587525,
+      "grad_norm": 0.45718878507614136,
+      "learning_rate": 2.4140821458507966e-06,
+      "loss": 0.4856,
+      "step": 288
+    },
+    {
+      "epoch": 0.07268611670020121,
+      "grad_norm": 0.40149986743927,
+      "learning_rate": 2.4224643755238896e-06,
+      "loss": 0.4889,
+      "step": 289
+    },
+    {
+      "epoch": 0.07293762575452717,
+      "grad_norm": 0.40798652172088623,
+      "learning_rate": 2.4308466051969827e-06,
+      "loss": 0.4864,
+      "step": 290
+    },
+    {
+      "epoch": 0.07318913480885311,
+      "grad_norm": 0.4065083861351013,
+      "learning_rate": 2.4392288348700757e-06,
+      "loss": 0.4788,
+      "step": 291
+    },
+    {
+      "epoch": 0.07344064386317907,
+      "grad_norm": 0.4597596824169159,
+      "learning_rate": 2.4476110645431688e-06,
+      "loss": 0.4578,
+      "step": 292
+    },
+    {
+      "epoch": 0.07369215291750503,
+      "grad_norm": 0.4461594521999359,
+      "learning_rate": 2.455993294216262e-06,
+      "loss": 0.4879,
+      "step": 293
+    },
+    {
+      "epoch": 0.07394366197183098,
+      "grad_norm": 0.3977876901626587,
+      "learning_rate": 2.4643755238893544e-06,
+      "loss": 0.4687,
+      "step": 294
+    },
+    {
+      "epoch": 0.07419517102615694,
+      "grad_norm": 0.4070913791656494,
+      "learning_rate": 2.472757753562448e-06,
+      "loss": 0.4712,
+      "step": 295
+    },
+    {
+      "epoch": 0.0744466800804829,
+      "grad_norm": 0.495626300573349,
+      "learning_rate": 2.481139983235541e-06,
+      "loss": 0.4682,
+      "step": 296
+    },
+    {
+      "epoch": 0.07469818913480886,
+      "grad_norm": 0.42833268642425537,
+      "learning_rate": 2.489522212908634e-06,
+      "loss": 0.4579,
+      "step": 297
+    },
+    {
+      "epoch": 0.07494969818913481,
+      "grad_norm": 0.4342649579048157,
+      "learning_rate": 2.497904442581727e-06,
+      "loss": 0.4299,
+      "step": 298
+    },
+    {
+      "epoch": 0.07520120724346077,
+      "grad_norm": 0.48605650663375854,
+      "learning_rate": 2.50628667225482e-06,
+      "loss": 0.4821,
+      "step": 299
+    },
+    {
+      "epoch": 0.07545271629778671,
+      "grad_norm": 0.45871275663375854,
+      "learning_rate": 2.514668901927913e-06,
+      "loss": 0.4783,
+      "step": 300
+    },
+    {
+      "epoch": 0.07570422535211267,
+      "grad_norm": 0.42935168743133545,
+      "learning_rate": 2.5230511316010058e-06,
+      "loss": 0.4676,
+      "step": 301
+    },
+    {
+      "epoch": 0.07595573440643863,
+      "grad_norm": 0.429830938577652,
+      "learning_rate": 2.5314333612740992e-06,
+      "loss": 0.4846,
+      "step": 302
+    },
+    {
+      "epoch": 0.07620724346076459,
+      "grad_norm": 0.4596365988254547,
+      "learning_rate": 2.5398155909471923e-06,
+      "loss": 0.47,
+      "step": 303
+    },
+    {
+      "epoch": 0.07645875251509054,
+      "grad_norm": 0.49432510137557983,
+      "learning_rate": 2.548197820620285e-06,
+      "loss": 0.4693,
+      "step": 304
+    },
+    {
+      "epoch": 0.0767102615694165,
+      "grad_norm": 0.44141435623168945,
+      "learning_rate": 2.5565800502933784e-06,
+      "loss": 0.4631,
+      "step": 305
+    },
+    {
+      "epoch": 0.07696177062374246,
+      "grad_norm": 0.43977904319763184,
+      "learning_rate": 2.5649622799664714e-06,
+      "loss": 0.4513,
+      "step": 306
+    },
+    {
+      "epoch": 0.07721327967806842,
+      "grad_norm": 0.4588465988636017,
+      "learning_rate": 2.573344509639564e-06,
+      "loss": 0.4577,
+      "step": 307
+    },
+    {
+      "epoch": 0.07746478873239436,
+      "grad_norm": 0.454364538192749,
+      "learning_rate": 2.581726739312657e-06,
+      "loss": 0.4682,
+      "step": 308
+    },
+    {
+      "epoch": 0.07771629778672032,
+      "grad_norm": 0.4163168668746948,
+      "learning_rate": 2.5901089689857506e-06,
+      "loss": 0.4905,
+      "step": 309
+    },
+    {
+      "epoch": 0.07796780684104627,
+      "grad_norm": 0.4823428988456726,
+      "learning_rate": 2.5984911986588436e-06,
+      "loss": 0.4737,
+      "step": 310
+    },
+    {
+      "epoch": 0.07821931589537223,
+      "grad_norm": 0.46709927916526794,
+      "learning_rate": 2.6068734283319363e-06,
+      "loss": 0.4498,
+      "step": 311
+    },
+    {
+      "epoch": 0.07847082494969819,
+      "grad_norm": 0.45121198892593384,
+      "learning_rate": 2.6152556580050297e-06,
+      "loss": 0.4373,
+      "step": 312
+    },
+    {
+      "epoch": 0.07872233400402415,
+      "grad_norm": 0.41081297397613525,
+      "learning_rate": 2.623637887678123e-06,
+      "loss": 0.4669,
+      "step": 313
+    },
+    {
+      "epoch": 0.0789738430583501,
+      "grad_norm": 0.4702714681625366,
+      "learning_rate": 2.6320201173512154e-06,
+      "loss": 0.45,
+      "step": 314
+    },
+    {
+      "epoch": 0.07922535211267606,
+      "grad_norm": 0.4380156099796295,
+      "learning_rate": 2.6404023470243085e-06,
+      "loss": 0.4857,
+      "step": 315
+    },
+    {
+      "epoch": 0.079476861167002,
+      "grad_norm": 0.4465760290622711,
+      "learning_rate": 2.648784576697402e-06,
+      "loss": 0.495,
+      "step": 316
+    },
+    {
+      "epoch": 0.07972837022132796,
+      "grad_norm": 0.4858304262161255,
+      "learning_rate": 2.6571668063704946e-06,
+      "loss": 0.4662,
+      "step": 317
+    },
+    {
+      "epoch": 0.07997987927565392,
+      "grad_norm": 0.44153428077697754,
+      "learning_rate": 2.6655490360435876e-06,
+      "loss": 0.4861,
+      "step": 318
+    },
+    {
+      "epoch": 0.08023138832997988,
+      "grad_norm": 0.44172754883766174,
+      "learning_rate": 2.673931265716681e-06,
+      "loss": 0.4725,
+      "step": 319
+    },
+    {
+      "epoch": 0.08048289738430583,
+      "grad_norm": 0.4511873126029968,
+      "learning_rate": 2.682313495389774e-06,
+      "loss": 0.4707,
+      "step": 320
+    },
+    {
+      "epoch": 0.08073440643863179,
+      "grad_norm": 0.4447844326496124,
+      "learning_rate": 2.6906957250628668e-06,
+      "loss": 0.4294,
+      "step": 321
+    },
+    {
+      "epoch": 0.08098591549295775,
+      "grad_norm": 0.4373539686203003,
+      "learning_rate": 2.6990779547359602e-06,
+      "loss": 0.4257,
+      "step": 322
+    },
+    {
+      "epoch": 0.08123742454728371,
+      "grad_norm": 0.4205957055091858,
+      "learning_rate": 2.7074601844090533e-06,
+      "loss": 0.4782,
+      "step": 323
+    },
+    {
+      "epoch": 0.08148893360160966,
+      "grad_norm": 0.4675666093826294,
+      "learning_rate": 2.715842414082146e-06,
+      "loss": 0.4852,
+      "step": 324
+    },
+    {
+      "epoch": 0.08174044265593561,
+      "grad_norm": 0.4324727952480316,
+      "learning_rate": 2.724224643755239e-06,
+      "loss": 0.4666,
+      "step": 325
+    },
+    {
+      "epoch": 0.08199195171026157,
+      "grad_norm": 0.49073663353919983,
+      "learning_rate": 2.7326068734283324e-06,
+      "loss": 0.4565,
+      "step": 326
+    },
+    {
+      "epoch": 0.08224346076458752,
+      "grad_norm": 0.38460773229599,
+      "learning_rate": 2.740989103101425e-06,
+      "loss": 0.4608,
+      "step": 327
+    },
+    {
+      "epoch": 0.08249496981891348,
+      "grad_norm": 0.4604051113128662,
+      "learning_rate": 2.749371332774518e-06,
+      "loss": 0.5168,
+      "step": 328
+    },
+    {
+      "epoch": 0.08274647887323944,
+      "grad_norm": 0.4031774401664734,
+      "learning_rate": 2.7577535624476116e-06,
+      "loss": 0.4719,
+      "step": 329
+    },
+    {
+      "epoch": 0.0829979879275654,
+      "grad_norm": 0.40108442306518555,
+      "learning_rate": 2.766135792120704e-06,
+      "loss": 0.4358,
+      "step": 330
+    },
+    {
+      "epoch": 0.08324949698189135,
+      "grad_norm": 0.533218502998352,
+      "learning_rate": 2.7745180217937973e-06,
+      "loss": 0.4492,
+      "step": 331
+    },
+    {
+      "epoch": 0.08350100603621731,
+      "grad_norm": 0.43500781059265137,
+      "learning_rate": 2.7829002514668907e-06,
+      "loss": 0.4561,
+      "step": 332
+    },
+    {
+      "epoch": 0.08375251509054325,
+      "grad_norm": 0.47511211037635803,
+      "learning_rate": 2.7912824811399838e-06,
+      "loss": 0.4575,
+      "step": 333
+    },
+    {
+      "epoch": 0.08400402414486921,
+      "grad_norm": 0.41863763332366943,
+      "learning_rate": 2.7996647108130764e-06,
+      "loss": 0.4525,
+      "step": 334
+    },
+    {
+      "epoch": 0.08425553319919517,
+      "grad_norm": 0.4359985589981079,
+      "learning_rate": 2.8080469404861695e-06,
+      "loss": 0.4666,
+      "step": 335
+    },
+    {
+      "epoch": 0.08450704225352113,
+      "grad_norm": 0.44898366928100586,
+      "learning_rate": 2.816429170159263e-06,
+      "loss": 0.4736,
+      "step": 336
+    },
+    {
+      "epoch": 0.08475855130784708,
+      "grad_norm": 0.44528570771217346,
+      "learning_rate": 2.8248113998323556e-06,
+      "loss": 0.4477,
+      "step": 337
+    },
+    {
+      "epoch": 0.08501006036217304,
+      "grad_norm": 0.4499231278896332,
+      "learning_rate": 2.8331936295054486e-06,
+      "loss": 0.4482,
+      "step": 338
+    },
+    {
+      "epoch": 0.085261569416499,
+      "grad_norm": 0.4471151828765869,
+      "learning_rate": 2.841575859178542e-06,
+      "loss": 0.4732,
+      "step": 339
+    },
+    {
+      "epoch": 0.08551307847082495,
+      "grad_norm": 0.42974570393562317,
+      "learning_rate": 2.8499580888516347e-06,
+      "loss": 0.4857,
+      "step": 340
+    },
+    {
+      "epoch": 0.08576458752515091,
+      "grad_norm": 0.40349292755126953,
+      "learning_rate": 2.8583403185247278e-06,
+      "loss": 0.4637,
+      "step": 341
+    },
+    {
+      "epoch": 0.08601609657947686,
+      "grad_norm": 0.38529425859451294,
+      "learning_rate": 2.8667225481978204e-06,
+      "loss": 0.4581,
+      "step": 342
+    },
+    {
+      "epoch": 0.08626760563380281,
+      "grad_norm": 0.4391642212867737,
+      "learning_rate": 2.875104777870914e-06,
+      "loss": 0.4549,
+      "step": 343
+    },
+    {
+      "epoch": 0.08651911468812877,
+      "grad_norm": 0.43910202383995056,
+      "learning_rate": 2.883487007544007e-06,
+      "loss": 0.45,
+      "step": 344
+    },
+    {
+      "epoch": 0.08677062374245473,
+      "grad_norm": 0.4166280925273895,
+      "learning_rate": 2.8918692372171e-06,
+      "loss": 0.4512,
+      "step": 345
+    },
+    {
+      "epoch": 0.08702213279678069,
+      "grad_norm": 0.40169304609298706,
+      "learning_rate": 2.9002514668901934e-06,
+      "loss": 0.4629,
+      "step": 346
+    },
+    {
+      "epoch": 0.08727364185110664,
+      "grad_norm": 0.45463281869888306,
+      "learning_rate": 2.908633696563286e-06,
+      "loss": 0.4465,
+      "step": 347
+    },
+    {
+      "epoch": 0.0875251509054326,
+      "grad_norm": 0.4328685700893402,
+      "learning_rate": 2.917015926236379e-06,
+      "loss": 0.4352,
+      "step": 348
+    },
+    {
+      "epoch": 0.08777665995975856,
+      "grad_norm": 0.4192608892917633,
+      "learning_rate": 2.9253981559094726e-06,
+      "loss": 0.483,
+      "step": 349
+    },
+    {
+      "epoch": 0.0880281690140845,
+      "grad_norm": 0.4864928126335144,
+      "learning_rate": 2.933780385582565e-06,
+      "loss": 0.464,
+      "step": 350
+    },
+    {
+      "epoch": 0.08827967806841046,
+      "grad_norm": 0.4200999140739441,
+      "learning_rate": 2.9421626152556582e-06,
+      "loss": 0.4781,
+      "step": 351
+    },
+    {
+      "epoch": 0.08853118712273642,
+      "grad_norm": 0.430940181016922,
+      "learning_rate": 2.950544844928751e-06,
+      "loss": 0.5017,
+      "step": 352
+    },
+    {
+      "epoch": 0.08878269617706237,
+      "grad_norm": 0.4381304085254669,
+      "learning_rate": 2.9589270746018443e-06,
+      "loss": 0.4565,
+      "step": 353
+    },
+    {
+      "epoch": 0.08903420523138833,
+      "grad_norm": 0.4690648019313812,
+      "learning_rate": 2.9673093042749374e-06,
+      "loss": 0.4408,
+      "step": 354
+    },
+    {
+      "epoch": 0.08928571428571429,
+      "grad_norm": 0.45738598704338074,
+      "learning_rate": 2.97569153394803e-06,
+      "loss": 0.4732,
+      "step": 355
+    },
+    {
+      "epoch": 0.08953722334004025,
+      "grad_norm": 0.4283290505409241,
+      "learning_rate": 2.9840737636211235e-06,
+      "loss": 0.4854,
+      "step": 356
+    },
+    {
+      "epoch": 0.0897887323943662,
+      "grad_norm": 0.4096241891384125,
+      "learning_rate": 2.9924559932942165e-06,
+      "loss": 0.4874,
+      "step": 357
+    },
+    {
+      "epoch": 0.09004024144869216,
+      "grad_norm": 0.47754043340682983,
+      "learning_rate": 3.0008382229673096e-06,
+      "loss": 0.4878,
+      "step": 358
+    },
+    {
+      "epoch": 0.0902917505030181,
+      "grad_norm": 0.43061840534210205,
+      "learning_rate": 3.009220452640403e-06,
+      "loss": 0.4499,
+      "step": 359
+    },
+    {
+      "epoch": 0.09054325955734406,
+      "grad_norm": 0.4733719825744629,
+      "learning_rate": 3.0176026823134957e-06,
+      "loss": 0.4817,
+      "step": 360
+    },
+    {
+      "epoch": 0.09079476861167002,
+      "grad_norm": 0.44821831583976746,
+      "learning_rate": 3.0259849119865887e-06,
+      "loss": 0.4783,
+      "step": 361
+    },
+    {
+      "epoch": 0.09104627766599598,
+      "grad_norm": 0.4314734935760498,
+      "learning_rate": 3.0343671416596814e-06,
+      "loss": 0.4892,
+      "step": 362
+    },
+    {
+      "epoch": 0.09129778672032193,
+      "grad_norm": 0.4240245223045349,
+      "learning_rate": 3.042749371332775e-06,
+      "loss": 0.448,
+      "step": 363
+    },
+    {
+      "epoch": 0.09154929577464789,
+      "grad_norm": 0.41133949160575867,
+      "learning_rate": 3.051131601005868e-06,
+      "loss": 0.4519,
+      "step": 364
+    },
+    {
+      "epoch": 0.09180080482897385,
+      "grad_norm": 0.4324391782283783,
+      "learning_rate": 3.0595138306789605e-06,
+      "loss": 0.4609,
+      "step": 365
+    },
+    {
+      "epoch": 0.0920523138832998,
+      "grad_norm": 0.3948531150817871,
+      "learning_rate": 3.067896060352054e-06,
+      "loss": 0.4525,
+      "step": 366
+    },
+    {
+      "epoch": 0.09230382293762575,
+      "grad_norm": 0.4043181836605072,
+      "learning_rate": 3.076278290025147e-06,
+      "loss": 0.4684,
+      "step": 367
+    },
+    {
+      "epoch": 0.0925553319919517,
+      "grad_norm": 0.49464017152786255,
+      "learning_rate": 3.0846605196982397e-06,
+      "loss": 0.4863,
+      "step": 368
+    },
+    {
+      "epoch": 0.09280684104627766,
+      "grad_norm": 0.4050739109516144,
+      "learning_rate": 3.0930427493713327e-06,
+      "loss": 0.4399,
+      "step": 369
+    },
+    {
+      "epoch": 0.09305835010060362,
+      "grad_norm": 0.46191224455833435,
+      "learning_rate": 3.101424979044426e-06,
+      "loss": 0.4729,
+      "step": 370
+    },
+    {
+      "epoch": 0.09330985915492958,
+      "grad_norm": 0.4379456043243408,
+      "learning_rate": 3.1098072087175192e-06,
+      "loss": 0.4765,
+      "step": 371
+    },
+    {
+      "epoch": 0.09356136820925554,
+      "grad_norm": 0.443185031414032,
+      "learning_rate": 3.118189438390612e-06,
+      "loss": 0.4497,
+      "step": 372
+    },
+    {
+      "epoch": 0.09381287726358149,
+      "grad_norm": 0.4520972967147827,
+      "learning_rate": 3.1265716680637053e-06,
+      "loss": 0.4766,
+      "step": 373
+    },
+    {
+      "epoch": 0.09406438631790745,
+      "grad_norm": 0.4627498984336853,
+      "learning_rate": 3.1349538977367984e-06,
+      "loss": 0.4559,
+      "step": 374
+    },
+    {
+      "epoch": 0.0943158953722334,
+      "grad_norm": 0.42422908544540405,
+      "learning_rate": 3.143336127409891e-06,
+      "loss": 0.442,
+      "step": 375
+    },
+    {
+      "epoch": 0.09456740442655935,
+      "grad_norm": 0.4159485399723053,
+      "learning_rate": 3.1517183570829845e-06,
+      "loss": 0.4841,
+      "step": 376
+    },
+    {
+      "epoch": 0.09481891348088531,
+      "grad_norm": 0.3887682855129242,
+      "learning_rate": 3.1601005867560775e-06,
+      "loss": 0.4615,
+      "step": 377
+    },
+    {
+      "epoch": 0.09507042253521127,
+      "grad_norm": 0.4549719989299774,
+      "learning_rate": 3.16848281642917e-06,
+      "loss": 0.4507,
+      "step": 378
+    },
+    {
+      "epoch": 0.09532193158953722,
+      "grad_norm": 0.4081493020057678,
+      "learning_rate": 3.176865046102263e-06,
+      "loss": 0.4658,
+      "step": 379
+    },
+    {
+      "epoch": 0.09557344064386318,
+      "grad_norm": 0.4539159834384918,
+      "learning_rate": 3.1852472757753567e-06,
+      "loss": 0.4618,
+      "step": 380
+    },
+    {
+      "epoch": 0.09582494969818914,
+      "grad_norm": 0.4574434459209442,
+      "learning_rate": 3.1936295054484493e-06,
+      "loss": 0.4952,
+      "step": 381
+    },
+    {
+      "epoch": 0.0960764587525151,
+      "grad_norm": 0.4301118552684784,
+      "learning_rate": 3.2020117351215424e-06,
+      "loss": 0.4533,
+      "step": 382
+    },
+    {
+      "epoch": 0.09632796780684105,
+      "grad_norm": 0.422446072101593,
+      "learning_rate": 3.210393964794636e-06,
+      "loss": 0.4402,
+      "step": 383
+    },
+    {
+      "epoch": 0.096579476861167,
+      "grad_norm": 0.43976905941963196,
+      "learning_rate": 3.218776194467729e-06,
+      "loss": 0.4586,
+      "step": 384
+    },
+    {
+      "epoch": 0.09683098591549295,
+      "grad_norm": 0.4008273482322693,
+      "learning_rate": 3.2271584241408215e-06,
+      "loss": 0.4547,
+      "step": 385
+    },
+    {
+      "epoch": 0.09708249496981891,
+      "grad_norm": 0.3887328803539276,
+      "learning_rate": 3.235540653813915e-06,
+      "loss": 0.4525,
+      "step": 386
+    },
+    {
+      "epoch": 0.09733400402414487,
+      "grad_norm": 0.4399072229862213,
+      "learning_rate": 3.243922883487008e-06,
+      "loss": 0.4558,
+      "step": 387
+    },
+    {
+      "epoch": 0.09758551307847083,
+      "grad_norm": 0.42582908272743225,
+      "learning_rate": 3.2523051131601007e-06,
+      "loss": 0.4602,
+      "step": 388
+    },
+    {
+      "epoch": 0.09783702213279678,
+      "grad_norm": 0.41464588046073914,
+      "learning_rate": 3.2606873428331937e-06,
+      "loss": 0.48,
+      "step": 389
+    },
+    {
+      "epoch": 0.09808853118712274,
+      "grad_norm": 0.4072147309780121,
+      "learning_rate": 3.269069572506287e-06,
+      "loss": 0.461,
+      "step": 390
+    },
+    {
+      "epoch": 0.0983400402414487,
+      "grad_norm": 0.39760878682136536,
+      "learning_rate": 3.27745180217938e-06,
+      "loss": 0.4555,
+      "step": 391
+    },
+    {
+      "epoch": 0.09859154929577464,
+      "grad_norm": 0.42994043231010437,
+      "learning_rate": 3.285834031852473e-06,
+      "loss": 0.4648,
+      "step": 392
+    },
+    {
+      "epoch": 0.0988430583501006,
+      "grad_norm": 0.4006013870239258,
+      "learning_rate": 3.2942162615255663e-06,
+      "loss": 0.4379,
+      "step": 393
+    },
+    {
+      "epoch": 0.09909456740442656,
+      "grad_norm": 0.44781169295310974,
+      "learning_rate": 3.3025984911986594e-06,
+      "loss": 0.4628,
+      "step": 394
+    },
+    {
+      "epoch": 0.09934607645875251,
+      "grad_norm": 0.45750489830970764,
+      "learning_rate": 3.310980720871752e-06,
+      "loss": 0.4313,
+      "step": 395
+    },
+    {
+      "epoch": 0.09959758551307847,
+      "grad_norm": 0.4136182963848114,
+      "learning_rate": 3.3193629505448455e-06,
+      "loss": 0.4463,
+      "step": 396
+    },
+    {
+      "epoch": 0.09984909456740443,
+      "grad_norm": 0.45176711678504944,
+      "learning_rate": 3.3277451802179385e-06,
+      "loss": 0.4713,
+      "step": 397
+    },
+    {
+      "epoch": 0.10010060362173039,
+      "grad_norm": 0.4363192021846771,
+      "learning_rate": 3.336127409891031e-06,
+      "loss": 0.4316,
+      "step": 398
+    },
+    {
+      "epoch": 0.10035211267605634,
+      "grad_norm": 0.4392482340335846,
+      "learning_rate": 3.344509639564124e-06,
+      "loss": 0.4388,
+      "step": 399
+    },
+    {
+      "epoch": 0.1006036217303823,
+      "grad_norm": 0.43276289105415344,
+      "learning_rate": 3.3528918692372177e-06,
+      "loss": 0.481,
+      "step": 400
+    },
+    {
+      "epoch": 0.10085513078470824,
+      "grad_norm": 0.40245917439460754,
+      "learning_rate": 3.3612740989103103e-06,
+      "loss": 0.4321,
+      "step": 401
+    },
+    {
+      "epoch": 0.1011066398390342,
+      "grad_norm": 0.4259881377220154,
+      "learning_rate": 3.3696563285834033e-06,
+      "loss": 0.4465,
+      "step": 402
+    },
+    {
+      "epoch": 0.10135814889336016,
+      "grad_norm": 0.452953964471817,
+      "learning_rate": 3.378038558256497e-06,
+      "loss": 0.4421,
+      "step": 403
+    },
+    {
+      "epoch": 0.10160965794768612,
+      "grad_norm": 0.39972519874572754,
+      "learning_rate": 3.3864207879295894e-06,
+      "loss": 0.431,
+      "step": 404
+    },
+    {
+      "epoch": 0.10186116700201207,
+      "grad_norm": 0.48080217838287354,
+      "learning_rate": 3.3948030176026825e-06,
+      "loss": 0.4602,
+      "step": 405
+    },
+    {
+      "epoch": 0.10211267605633803,
+      "grad_norm": 0.3960705101490021,
+      "learning_rate": 3.4031852472757755e-06,
+      "loss": 0.4666,
+      "step": 406
+    },
+    {
+      "epoch": 0.10236418511066399,
+      "grad_norm": 0.40246954560279846,
+      "learning_rate": 3.411567476948869e-06,
+      "loss": 0.4266,
+      "step": 407
+    },
+    {
+      "epoch": 0.10261569416498995,
+      "grad_norm": 0.43389272689819336,
+      "learning_rate": 3.4199497066219616e-06,
+      "loss": 0.4464,
+      "step": 408
+    },
+    {
+      "epoch": 0.10286720321931589,
+      "grad_norm": 0.3892342746257782,
+      "learning_rate": 3.4283319362950547e-06,
+      "loss": 0.4403,
+      "step": 409
+    },
+    {
+      "epoch": 0.10311871227364185,
+      "grad_norm": 0.45472291111946106,
+      "learning_rate": 3.436714165968148e-06,
+      "loss": 0.479,
+      "step": 410
+    },
+    {
+      "epoch": 0.1033702213279678,
+      "grad_norm": 0.4327657222747803,
+      "learning_rate": 3.445096395641241e-06,
+      "loss": 0.4696,
+      "step": 411
+    },
+    {
+      "epoch": 0.10362173038229376,
+      "grad_norm": 0.44970017671585083,
+      "learning_rate": 3.453478625314334e-06,
+      "loss": 0.4425,
+      "step": 412
+    },
+    {
+      "epoch": 0.10387323943661972,
+      "grad_norm": 0.4211669862270355,
+      "learning_rate": 3.4618608549874273e-06,
+      "loss": 0.4674,
+      "step": 413
+    },
+    {
+      "epoch": 0.10412474849094568,
+      "grad_norm": 0.3977251946926117,
+      "learning_rate": 3.47024308466052e-06,
+      "loss": 0.4427,
+      "step": 414
+    },
+    {
+      "epoch": 0.10437625754527163,
+      "grad_norm": 0.41054680943489075,
+      "learning_rate": 3.478625314333613e-06,
+      "loss": 0.4539,
+      "step": 415
+    },
+    {
+      "epoch": 0.10462776659959759,
+      "grad_norm": 0.42676669359207153,
+      "learning_rate": 3.4870075440067056e-06,
+      "loss": 0.4568,
+      "step": 416
+    },
+    {
+      "epoch": 0.10487927565392353,
+      "grad_norm": 0.4475442171096802,
+      "learning_rate": 3.495389773679799e-06,
+      "loss": 0.4512,
+      "step": 417
+    },
+    {
+      "epoch": 0.10513078470824949,
+      "grad_norm": 0.406112402677536,
+      "learning_rate": 3.503772003352892e-06,
+      "loss": 0.3903,
+      "step": 418
+    },
+    {
+      "epoch": 0.10538229376257545,
+      "grad_norm": 0.4194413125514984,
+      "learning_rate": 3.512154233025985e-06,
+      "loss": 0.4447,
+      "step": 419
+    },
+    {
+      "epoch": 0.1056338028169014,
+      "grad_norm": 0.4342004954814911,
+      "learning_rate": 3.5205364626990787e-06,
+      "loss": 0.4612,
+      "step": 420
+    },
+    {
+      "epoch": 0.10588531187122736,
+      "grad_norm": 0.408413290977478,
+      "learning_rate": 3.5289186923721713e-06,
+      "loss": 0.4595,
+      "step": 421
+    },
+    {
+      "epoch": 0.10613682092555332,
+      "grad_norm": 0.45207443833351135,
+      "learning_rate": 3.5373009220452643e-06,
+      "loss": 0.4918,
+      "step": 422
+    },
+    {
+      "epoch": 0.10638832997987928,
+      "grad_norm": 0.4605708718299866,
+      "learning_rate": 3.545683151718358e-06,
+      "loss": 0.4277,
+      "step": 423
+    },
+    {
+      "epoch": 0.10663983903420524,
+      "grad_norm": 0.45539170503616333,
+      "learning_rate": 3.5540653813914504e-06,
+      "loss": 0.4672,
+      "step": 424
+    },
+    {
+      "epoch": 0.1068913480885312,
+      "grad_norm": 0.4451138973236084,
+      "learning_rate": 3.5624476110645435e-06,
+      "loss": 0.4436,
+      "step": 425
+    },
+    {
+      "epoch": 0.10714285714285714,
+      "grad_norm": 0.4522078335285187,
+      "learning_rate": 3.570829840737636e-06,
+      "loss": 0.4035,
+      "step": 426
+    },
+    {
+      "epoch": 0.1073943661971831,
+      "grad_norm": 0.48479580879211426,
+      "learning_rate": 3.5792120704107296e-06,
+      "loss": 0.4759,
+      "step": 427
+    },
+    {
+      "epoch": 0.10764587525150905,
+      "grad_norm": 0.44372591376304626,
+      "learning_rate": 3.5875943000838226e-06,
+      "loss": 0.4588,
+      "step": 428
+    },
+    {
+      "epoch": 0.10789738430583501,
+      "grad_norm": 0.42548373341560364,
+      "learning_rate": 3.5959765297569153e-06,
+      "loss": 0.4506,
+      "step": 429
+    },
+    {
+      "epoch": 0.10814889336016097,
+      "grad_norm": 0.49597272276878357,
+      "learning_rate": 3.6043587594300087e-06,
+      "loss": 0.4831,
+      "step": 430
+    },
+    {
+      "epoch": 0.10840040241448692,
+      "grad_norm": 0.44530630111694336,
+      "learning_rate": 3.6127409891031018e-06,
+      "loss": 0.4139,
+      "step": 431
+    },
+    {
+      "epoch": 0.10865191146881288,
+      "grad_norm": 0.40483203530311584,
+      "learning_rate": 3.621123218776195e-06,
+      "loss": 0.4462,
+      "step": 432
+    },
+    {
+      "epoch": 0.10890342052313884,
+      "grad_norm": 0.38956066966056824,
+      "learning_rate": 3.6295054484492875e-06,
+      "loss": 0.4374,
+      "step": 433
+    },
+    {
+      "epoch": 0.10915492957746478,
+      "grad_norm": 0.46306437253952026,
+      "learning_rate": 3.637887678122381e-06,
+      "loss": 0.4532,
+      "step": 434
+    },
+    {
+      "epoch": 0.10940643863179074,
+      "grad_norm": 0.43688806891441345,
+      "learning_rate": 3.646269907795474e-06,
+      "loss": 0.4223,
+      "step": 435
+    },
+    {
+      "epoch": 0.1096579476861167,
+      "grad_norm": 0.43097516894340515,
+      "learning_rate": 3.6546521374685666e-06,
+      "loss": 0.4335,
+      "step": 436
+    },
+    {
+      "epoch": 0.10990945674044265,
+      "grad_norm": 0.4253983199596405,
+      "learning_rate": 3.66303436714166e-06,
+      "loss": 0.439,
+      "step": 437
+    },
+    {
+      "epoch": 0.11016096579476861,
+      "grad_norm": 0.4934920072555542,
+      "learning_rate": 3.671416596814753e-06,
+      "loss": 0.4308,
+      "step": 438
+    },
+    {
+      "epoch": 0.11041247484909457,
+      "grad_norm": 0.42577728629112244,
+      "learning_rate": 3.6797988264878458e-06,
+      "loss": 0.4365,
+      "step": 439
+    },
+    {
+      "epoch": 0.11066398390342053,
+      "grad_norm": 0.3953479826450348,
+      "learning_rate": 3.6881810561609392e-06,
+      "loss": 0.4392,
+      "step": 440
+    },
+    {
+      "epoch": 0.11091549295774648,
+      "grad_norm": 0.4147731363773346,
+      "learning_rate": 3.6965632858340323e-06,
+      "loss": 0.4588,
+      "step": 441
+    },
+    {
+      "epoch": 0.11116700201207244,
+      "grad_norm": 0.529985249042511,
+      "learning_rate": 3.704945515507125e-06,
+      "loss": 0.4484,
+      "step": 442
+    },
+    {
+      "epoch": 0.11141851106639838,
+      "grad_norm": 0.4451605975627899,
+      "learning_rate": 3.713327745180218e-06,
+      "loss": 0.4692,
+      "step": 443
+    },
+    {
+      "epoch": 0.11167002012072434,
+      "grad_norm": 0.46244457364082336,
+      "learning_rate": 3.7217099748533114e-06,
+      "loss": 0.4871,
+      "step": 444
+    },
+    {
+      "epoch": 0.1119215291750503,
+      "grad_norm": 0.39799419045448303,
+      "learning_rate": 3.7300922045264045e-06,
+      "loss": 0.4153,
+      "step": 445
+    },
+    {
+      "epoch": 0.11217303822937626,
+      "grad_norm": 0.48522502183914185,
+      "learning_rate": 3.738474434199497e-06,
+      "loss": 0.4595,
+      "step": 446
+    },
+    {
+      "epoch": 0.11242454728370221,
+      "grad_norm": 0.4234901964664459,
+      "learning_rate": 3.7468566638725906e-06,
+      "loss": 0.4008,
+      "step": 447
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 0.4186321198940277,
+      "learning_rate": 3.7552388935456836e-06,
+      "loss": 0.4629,
+      "step": 448
+    },
+    {
+      "epoch": 0.11292756539235413,
+      "grad_norm": 0.4625905156135559,
+      "learning_rate": 3.7636211232187762e-06,
+      "loss": 0.4222,
+      "step": 449
+    },
+    {
+      "epoch": 0.11317907444668009,
+      "grad_norm": 0.44527822732925415,
+      "learning_rate": 3.7720033528918697e-06,
+      "loss": 0.4633,
+      "step": 450
+    },
+    {
+      "epoch": 0.11343058350100603,
+      "grad_norm": 0.47749683260917664,
+      "learning_rate": 3.7803855825649628e-06,
+      "loss": 0.4123,
+      "step": 451
+    },
+    {
+      "epoch": 0.11368209255533199,
+      "grad_norm": 0.4436284601688385,
+      "learning_rate": 3.7887678122380554e-06,
+      "loss": 0.4425,
+      "step": 452
+    },
+    {
+      "epoch": 0.11393360160965794,
+      "grad_norm": 0.41336312890052795,
+      "learning_rate": 3.7971500419111484e-06,
+      "loss": 0.4376,
+      "step": 453
+    },
+    {
+      "epoch": 0.1141851106639839,
+      "grad_norm": 0.43879643082618713,
+      "learning_rate": 3.805532271584242e-06,
+      "loss": 0.4315,
+      "step": 454
+    },
+    {
+      "epoch": 0.11443661971830986,
+      "grad_norm": 0.44419604539871216,
+      "learning_rate": 3.8139145012573345e-06,
+      "loss": 0.4351,
+      "step": 455
+    },
+    {
+      "epoch": 0.11468812877263582,
+      "grad_norm": 0.46073997020721436,
+      "learning_rate": 3.822296730930428e-06,
+      "loss": 0.4417,
+      "step": 456
+    },
+    {
+      "epoch": 0.11493963782696177,
+      "grad_norm": 0.538664698600769,
+      "learning_rate": 3.8306789606035215e-06,
+      "loss": 0.4817,
+      "step": 457
+    },
+    {
+      "epoch": 0.11519114688128773,
+      "grad_norm": 0.49703216552734375,
+      "learning_rate": 3.839061190276614e-06,
+      "loss": 0.4625,
+      "step": 458
+    },
+    {
+      "epoch": 0.11544265593561369,
+      "grad_norm": 0.4681941568851471,
+      "learning_rate": 3.847443419949707e-06,
+      "loss": 0.4454,
+      "step": 459
+    },
+    {
+      "epoch": 0.11569416498993963,
+      "grad_norm": 0.512788712978363,
+      "learning_rate": 3.855825649622799e-06,
+      "loss": 0.4434,
+      "step": 460
+    },
+    {
+      "epoch": 0.11594567404426559,
+      "grad_norm": 0.5022100210189819,
+      "learning_rate": 3.864207879295893e-06,
+      "loss": 0.4626,
+      "step": 461
+    },
+    {
+      "epoch": 0.11619718309859155,
+      "grad_norm": 0.48100098967552185,
+      "learning_rate": 3.872590108968986e-06,
+      "loss": 0.4431,
+      "step": 462
+    },
+    {
+      "epoch": 0.1164486921529175,
+      "grad_norm": 0.4972802698612213,
+      "learning_rate": 3.880972338642079e-06,
+      "loss": 0.4199,
+      "step": 463
+    },
+    {
+      "epoch": 0.11670020120724346,
+      "grad_norm": 0.492680162191391,
+      "learning_rate": 3.889354568315172e-06,
+      "loss": 0.4546,
+      "step": 464
+    },
+    {
+      "epoch": 0.11695171026156942,
+      "grad_norm": 0.4608025848865509,
+      "learning_rate": 3.897736797988265e-06,
+      "loss": 0.4149,
+      "step": 465
+    },
+    {
+      "epoch": 0.11720321931589538,
+      "grad_norm": 0.4747462868690491,
+      "learning_rate": 3.906119027661358e-06,
+      "loss": 0.4254,
+      "step": 466
+    },
+    {
+      "epoch": 0.11745472837022133,
+      "grad_norm": 0.5062370300292969,
+      "learning_rate": 3.914501257334451e-06,
+      "loss": 0.4295,
+      "step": 467
+    },
+    {
+      "epoch": 0.11770623742454728,
+      "grad_norm": 0.4820232093334198,
+      "learning_rate": 3.922883487007545e-06,
+      "loss": 0.4812,
+      "step": 468
+    },
+    {
+      "epoch": 0.11795774647887323,
+      "grad_norm": 0.5073891282081604,
+      "learning_rate": 3.931265716680637e-06,
+      "loss": 0.4456,
+      "step": 469
+    },
+    {
+      "epoch": 0.11820925553319919,
+      "grad_norm": 0.4702168107032776,
+      "learning_rate": 3.93964794635373e-06,
+      "loss": 0.4479,
+      "step": 470
+    },
+    {
+      "epoch": 0.11846076458752515,
+      "grad_norm": 0.4862816035747528,
+      "learning_rate": 3.948030176026823e-06,
+      "loss": 0.4485,
+      "step": 471
+    },
+    {
+      "epoch": 0.11871227364185111,
+      "grad_norm": 0.49707740545272827,
+      "learning_rate": 3.956412405699917e-06,
+      "loss": 0.4292,
+      "step": 472
+    },
+    {
+      "epoch": 0.11896378269617706,
+      "grad_norm": 0.4554164707660675,
+      "learning_rate": 3.9647946353730094e-06,
+      "loss": 0.4791,
+      "step": 473
+    },
+    {
+      "epoch": 0.11921529175050302,
+      "grad_norm": 0.4359682500362396,
+      "learning_rate": 3.973176865046103e-06,
+      "loss": 0.465,
+      "step": 474
+    },
+    {
+      "epoch": 0.11946680080482898,
+      "grad_norm": 0.47159573435783386,
+      "learning_rate": 3.9815590947191955e-06,
+      "loss": 0.46,
+      "step": 475
+    },
+    {
+      "epoch": 0.11971830985915492,
+      "grad_norm": 0.47286924719810486,
+      "learning_rate": 3.989941324392288e-06,
+      "loss": 0.4639,
+      "step": 476
+    },
+    {
+      "epoch": 0.11996981891348088,
+      "grad_norm": 0.46829286217689514,
+      "learning_rate": 3.998323554065382e-06,
+      "loss": 0.4678,
+      "step": 477
+    },
+    {
+      "epoch": 0.12022132796780684,
+      "grad_norm": 0.44564419984817505,
+      "learning_rate": 4.006705783738475e-06,
+      "loss": 0.4401,
+      "step": 478
+    },
+    {
+      "epoch": 0.1204728370221328,
+      "grad_norm": 0.4393386244773865,
+      "learning_rate": 4.015088013411568e-06,
+      "loss": 0.46,
+      "step": 479
+    },
+    {
+      "epoch": 0.12072434607645875,
+      "grad_norm": 0.4447605311870575,
+      "learning_rate": 4.02347024308466e-06,
+      "loss": 0.4711,
+      "step": 480
+    },
+    {
+      "epoch": 0.12097585513078471,
+      "grad_norm": 0.4248807430267334,
+      "learning_rate": 4.031852472757754e-06,
+      "loss": 0.4239,
+      "step": 481
+    },
+    {
+      "epoch": 0.12122736418511067,
+      "grad_norm": 0.4450969398021698,
+      "learning_rate": 4.040234702430847e-06,
+      "loss": 0.4406,
+      "step": 482
+    },
+    {
+      "epoch": 0.12147887323943662,
+      "grad_norm": 0.5100280046463013,
+      "learning_rate": 4.04861693210394e-06,
+      "loss": 0.4406,
+      "step": 483
+    },
+    {
+      "epoch": 0.12173038229376258,
+      "grad_norm": 0.47610798478126526,
+      "learning_rate": 4.056999161777033e-06,
+      "loss": 0.4112,
+      "step": 484
+    },
+    {
+      "epoch": 0.12198189134808853,
+      "grad_norm": 0.43737804889678955,
+      "learning_rate": 4.065381391450126e-06,
+      "loss": 0.4757,
+      "step": 485
+    },
+    {
+      "epoch": 0.12223340040241448,
+      "grad_norm": 0.4593689441680908,
+      "learning_rate": 4.073763621123219e-06,
+      "loss": 0.4445,
+      "step": 486
+    },
+    {
+      "epoch": 0.12248490945674044,
+      "grad_norm": 0.46989330649375916,
+      "learning_rate": 4.082145850796312e-06,
+      "loss": 0.4559,
+      "step": 487
+    },
+    {
+      "epoch": 0.1227364185110664,
+      "grad_norm": 0.48121753334999084,
+      "learning_rate": 4.090528080469406e-06,
+      "loss": 0.4492,
+      "step": 488
+    },
+    {
+      "epoch": 0.12298792756539235,
+      "grad_norm": 0.4634334444999695,
+      "learning_rate": 4.098910310142498e-06,
+      "loss": 0.4346,
+      "step": 489
+    },
+    {
+      "epoch": 0.12323943661971831,
+      "grad_norm": 0.4351527988910675,
+      "learning_rate": 4.107292539815591e-06,
+      "loss": 0.4539,
+      "step": 490
+    },
+    {
+      "epoch": 0.12349094567404427,
+      "grad_norm": 0.47944945096969604,
+      "learning_rate": 4.115674769488684e-06,
+      "loss": 0.4655,
+      "step": 491
+    },
+    {
+      "epoch": 0.12374245472837023,
+      "grad_norm": 0.45442450046539307,
+      "learning_rate": 4.124056999161778e-06,
+      "loss": 0.4277,
+      "step": 492
+    },
+    {
+      "epoch": 0.12399396378269617,
+      "grad_norm": 0.4729836583137512,
+      "learning_rate": 4.1324392288348704e-06,
+      "loss": 0.4299,
+      "step": 493
+    },
+    {
+      "epoch": 0.12424547283702213,
+      "grad_norm": 0.40985509753227234,
+      "learning_rate": 4.140821458507964e-06,
+      "loss": 0.393,
+      "step": 494
+    },
+    {
+      "epoch": 0.12449698189134809,
+      "grad_norm": 0.49194350838661194,
+      "learning_rate": 4.1492036881810565e-06,
+      "loss": 0.4523,
+      "step": 495
+    },
+    {
+      "epoch": 0.12474849094567404,
+      "grad_norm": 0.47581902146339417,
+      "learning_rate": 4.157585917854149e-06,
+      "loss": 0.4494,
+      "step": 496
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.4293755888938904,
+      "learning_rate": 4.165968147527243e-06,
+      "loss": 0.4087,
+      "step": 497
+    },
+    {
+      "epoch": 0.12525150905432594,
+      "grad_norm": 0.41463249921798706,
+      "learning_rate": 4.174350377200336e-06,
+      "loss": 0.4436,
+      "step": 498
+    },
+    {
+      "epoch": 0.12550301810865191,
+      "grad_norm": 0.4227687120437622,
+      "learning_rate": 4.182732606873429e-06,
+      "loss": 0.4367,
+      "step": 499
+    },
+    {
+      "epoch": 0.12575452716297786,
+      "grad_norm": 0.4326210916042328,
+      "learning_rate": 4.191114836546521e-06,
+      "loss": 0.4265,
+      "step": 500
+    },
+    {
+      "epoch": 0.12600603621730383,
+      "grad_norm": 0.4279535412788391,
+      "learning_rate": 4.199497066219615e-06,
+      "loss": 0.4702,
+      "step": 501
+    },
+    {
+      "epoch": 0.12625754527162977,
+      "grad_norm": 0.4595158100128174,
+      "learning_rate": 4.2078792958927074e-06,
+      "loss": 0.4425,
+      "step": 502
+    },
+    {
+      "epoch": 0.12650905432595574,
+      "grad_norm": 0.41721752285957336,
+      "learning_rate": 4.216261525565801e-06,
+      "loss": 0.4491,
+      "step": 503
+    },
+    {
+      "epoch": 0.1267605633802817,
+      "grad_norm": 0.43518510460853577,
+      "learning_rate": 4.224643755238894e-06,
+      "loss": 0.4387,
+      "step": 504
+    },
+    {
+      "epoch": 0.12701207243460766,
+      "grad_norm": 0.4557619094848633,
+      "learning_rate": 4.233025984911987e-06,
+      "loss": 0.4852,
+      "step": 505
+    },
+    {
+      "epoch": 0.1272635814889336,
+      "grad_norm": 0.4129253327846527,
+      "learning_rate": 4.24140821458508e-06,
+      "loss": 0.4359,
+      "step": 506
+    },
+    {
+      "epoch": 0.12751509054325955,
+      "grad_norm": 0.4084147810935974,
+      "learning_rate": 4.249790444258173e-06,
+      "loss": 0.4316,
+      "step": 507
+    },
+    {
+      "epoch": 0.12776659959758552,
+      "grad_norm": 0.4409344792366028,
+      "learning_rate": 4.258172673931267e-06,
+      "loss": 0.4375,
+      "step": 508
+    },
+    {
+      "epoch": 0.12801810865191146,
+      "grad_norm": 0.4991019368171692,
+      "learning_rate": 4.266554903604359e-06,
+      "loss": 0.4551,
+      "step": 509
+    },
+    {
+      "epoch": 0.12826961770623743,
+      "grad_norm": 0.4926038682460785,
+      "learning_rate": 4.274937133277452e-06,
+      "loss": 0.4643,
+      "step": 510
+    },
+    {
+      "epoch": 0.12852112676056338,
+      "grad_norm": 0.46267032623291016,
+      "learning_rate": 4.283319362950545e-06,
+      "loss": 0.4648,
+      "step": 511
+    },
+    {
+      "epoch": 0.12877263581488935,
+      "grad_norm": 0.45014211535453796,
+      "learning_rate": 4.291701592623638e-06,
+      "loss": 0.4574,
+      "step": 512
+    },
+    {
+      "epoch": 0.1290241448692153,
+      "grad_norm": 0.5043582320213318,
+      "learning_rate": 4.300083822296731e-06,
+      "loss": 0.463,
+      "step": 513
+    },
+    {
+      "epoch": 0.12927565392354123,
+      "grad_norm": 0.41695088148117065,
+      "learning_rate": 4.308466051969825e-06,
+      "loss": 0.4247,
+      "step": 514
+    },
+    {
+      "epoch": 0.1295271629778672,
+      "grad_norm": 0.4031003713607788,
+      "learning_rate": 4.3168482816429175e-06,
+      "loss": 0.4563,
+      "step": 515
+    },
+    {
+      "epoch": 0.12977867203219315,
+      "grad_norm": 0.4944654405117035,
+      "learning_rate": 4.32523051131601e-06,
+      "loss": 0.4323,
+      "step": 516
+    },
+    {
+      "epoch": 0.13003018108651912,
+      "grad_norm": 0.5555236339569092,
+      "learning_rate": 4.333612740989104e-06,
+      "loss": 0.4263,
+      "step": 517
+    },
+    {
+      "epoch": 0.13028169014084506,
+      "grad_norm": 0.4785892963409424,
+      "learning_rate": 4.341994970662197e-06,
+      "loss": 0.4331,
+      "step": 518
+    },
+    {
+      "epoch": 0.13053319919517103,
+      "grad_norm": 0.4259514808654785,
+      "learning_rate": 4.35037720033529e-06,
+      "loss": 0.4495,
+      "step": 519
+    },
+    {
+      "epoch": 0.13078470824949698,
+      "grad_norm": 0.4405849575996399,
+      "learning_rate": 4.358759430008382e-06,
+      "loss": 0.4482,
+      "step": 520
+    },
+    {
+      "epoch": 0.13103621730382295,
+      "grad_norm": 0.4837403893470764,
+      "learning_rate": 4.367141659681476e-06,
+      "loss": 0.4452,
+      "step": 521
+    },
+    {
+      "epoch": 0.1312877263581489,
+      "grad_norm": 0.44248178601264954,
+      "learning_rate": 4.3755238893545684e-06,
+      "loss": 0.4755,
+      "step": 522
+    },
+    {
+      "epoch": 0.13153923541247484,
+      "grad_norm": 0.4511367678642273,
+      "learning_rate": 4.383906119027662e-06,
+      "loss": 0.4325,
+      "step": 523
+    },
+    {
+      "epoch": 0.1317907444668008,
+      "grad_norm": 0.4366004168987274,
+      "learning_rate": 4.3922883487007545e-06,
+      "loss": 0.4221,
+      "step": 524
+    },
+    {
+      "epoch": 0.13204225352112675,
+      "grad_norm": 0.4173247218132019,
+      "learning_rate": 4.400670578373848e-06,
+      "loss": 0.4719,
+      "step": 525
+    },
+    {
+      "epoch": 0.13229376257545272,
+      "grad_norm": 0.4634314179420471,
+      "learning_rate": 4.409052808046941e-06,
+      "loss": 0.4307,
+      "step": 526
+    },
+    {
+      "epoch": 0.13254527162977867,
+      "grad_norm": 0.464970201253891,
+      "learning_rate": 4.417435037720033e-06,
+      "loss": 0.4247,
+      "step": 527
+    },
+    {
+      "epoch": 0.13279678068410464,
+      "grad_norm": 0.49737635254859924,
+      "learning_rate": 4.425817267393127e-06,
+      "loss": 0.4293,
+      "step": 528
+    },
+    {
+      "epoch": 0.13304828973843058,
+      "grad_norm": 0.4311039447784424,
+      "learning_rate": 4.43419949706622e-06,
+      "loss": 0.4513,
+      "step": 529
+    },
+    {
+      "epoch": 0.13329979879275655,
+      "grad_norm": 0.46590545773506165,
+      "learning_rate": 4.442581726739313e-06,
+      "loss": 0.4163,
+      "step": 530
+    },
+    {
+      "epoch": 0.1335513078470825,
+      "grad_norm": 0.4511418640613556,
+      "learning_rate": 4.450963956412406e-06,
+      "loss": 0.4741,
+      "step": 531
+    },
+    {
+      "epoch": 0.13380281690140844,
+      "grad_norm": 0.4536955952644348,
+      "learning_rate": 4.459346186085499e-06,
+      "loss": 0.4525,
+      "step": 532
+    },
+    {
+      "epoch": 0.1340543259557344,
+      "grad_norm": 0.4565694332122803,
+      "learning_rate": 4.467728415758592e-06,
+      "loss": 0.4405,
+      "step": 533
+    },
+    {
+      "epoch": 0.13430583501006035,
+      "grad_norm": 0.4305288791656494,
+      "learning_rate": 4.476110645431685e-06,
+      "loss": 0.4585,
+      "step": 534
+    },
+    {
+      "epoch": 0.13455734406438632,
+      "grad_norm": 0.4427644610404968,
+      "learning_rate": 4.4844928751047785e-06,
+      "loss": 0.4338,
+      "step": 535
+    },
+    {
+      "epoch": 0.13480885311871227,
+      "grad_norm": 0.4909707307815552,
+      "learning_rate": 4.492875104777871e-06,
+      "loss": 0.4331,
+      "step": 536
+    },
+    {
+      "epoch": 0.13506036217303824,
+      "grad_norm": 0.5035997629165649,
+      "learning_rate": 4.501257334450964e-06,
+      "loss": 0.4287,
+      "step": 537
+    },
+    {
+      "epoch": 0.13531187122736418,
+      "grad_norm": 0.45080435276031494,
+      "learning_rate": 4.509639564124057e-06,
+      "loss": 0.452,
+      "step": 538
+    },
+    {
+      "epoch": 0.13556338028169015,
+      "grad_norm": 0.4062177836894989,
+      "learning_rate": 4.518021793797151e-06,
+      "loss": 0.421,
+      "step": 539
+    },
+    {
+      "epoch": 0.1358148893360161,
+      "grad_norm": 0.47915151715278625,
+      "learning_rate": 4.526404023470243e-06,
+      "loss": 0.4678,
+      "step": 540
+    },
+    {
+      "epoch": 0.13606639839034204,
+      "grad_norm": 0.4649512469768524,
+      "learning_rate": 4.534786253143337e-06,
+      "loss": 0.4532,
+      "step": 541
+    },
+    {
+      "epoch": 0.136317907444668,
+      "grad_norm": 0.5145220756530762,
+      "learning_rate": 4.5431684828164294e-06,
+      "loss": 0.4529,
+      "step": 542
+    },
+    {
+      "epoch": 0.13656941649899396,
+      "grad_norm": 0.39938247203826904,
+      "learning_rate": 4.551550712489523e-06,
+      "loss": 0.4457,
+      "step": 543
+    },
+    {
+      "epoch": 0.13682092555331993,
+      "grad_norm": 0.5199249386787415,
+      "learning_rate": 4.5599329421626155e-06,
+      "loss": 0.442,
+      "step": 544
+    },
+    {
+      "epoch": 0.13707243460764587,
+      "grad_norm": 0.43543845415115356,
+      "learning_rate": 4.568315171835709e-06,
+      "loss": 0.4312,
+      "step": 545
+    },
+    {
+      "epoch": 0.13732394366197184,
+      "grad_norm": 0.44674405455589294,
+      "learning_rate": 4.576697401508802e-06,
+      "loss": 0.4456,
+      "step": 546
+    },
+    {
+      "epoch": 0.13757545271629779,
+      "grad_norm": 0.45393526554107666,
+      "learning_rate": 4.585079631181894e-06,
+      "loss": 0.4537,
+      "step": 547
+    },
+    {
+      "epoch": 0.13782696177062373,
+      "grad_norm": 0.43489089608192444,
+      "learning_rate": 4.593461860854988e-06,
+      "loss": 0.398,
+      "step": 548
+    },
+    {
+      "epoch": 0.1380784708249497,
+      "grad_norm": 0.41286396980285645,
+      "learning_rate": 4.601844090528081e-06,
+      "loss": 0.4434,
+      "step": 549
+    },
+    {
+      "epoch": 0.13832997987927564,
+      "grad_norm": 0.44615626335144043,
+      "learning_rate": 4.610226320201174e-06,
+      "loss": 0.4542,
+      "step": 550
+    },
+    {
+      "epoch": 0.13858148893360162,
+      "grad_norm": 0.4532471001148224,
+      "learning_rate": 4.6186085498742664e-06,
+      "loss": 0.462,
+      "step": 551
+    },
+    {
+      "epoch": 0.13883299798792756,
+      "grad_norm": 0.42889168858528137,
+      "learning_rate": 4.62699077954736e-06,
+      "loss": 0.4172,
+      "step": 552
+    },
+    {
+      "epoch": 0.13908450704225353,
+      "grad_norm": 0.46256425976753235,
+      "learning_rate": 4.6353730092204525e-06,
+      "loss": 0.4216,
+      "step": 553
+    },
+    {
+      "epoch": 0.13933601609657947,
+      "grad_norm": 0.4532136023044586,
+      "learning_rate": 4.643755238893546e-06,
+      "loss": 0.4269,
+      "step": 554
+    },
+    {
+      "epoch": 0.13958752515090544,
+      "grad_norm": 0.43736207485198975,
+      "learning_rate": 4.6521374685666395e-06,
+      "loss": 0.4318,
+      "step": 555
+    },
+    {
+      "epoch": 0.1398390342052314,
+      "grad_norm": 0.40762320160865784,
+      "learning_rate": 4.660519698239732e-06,
+      "loss": 0.4601,
+      "step": 556
+    },
+    {
+      "epoch": 0.14009054325955733,
+      "grad_norm": 0.4910624027252197,
+      "learning_rate": 4.668901927912825e-06,
+      "loss": 0.4335,
+      "step": 557
+    },
+    {
+      "epoch": 0.1403420523138833,
+      "grad_norm": 0.4497995972633362,
+      "learning_rate": 4.677284157585918e-06,
+      "loss": 0.4311,
+      "step": 558
+    },
+    {
+      "epoch": 0.14059356136820925,
+      "grad_norm": 0.4066920876502991,
+      "learning_rate": 4.685666387259012e-06,
+      "loss": 0.4364,
+      "step": 559
+    },
+    {
+      "epoch": 0.14084507042253522,
+      "grad_norm": 0.4641585350036621,
+      "learning_rate": 4.694048616932104e-06,
+      "loss": 0.4406,
+      "step": 560
+    },
+    {
+      "epoch": 0.14109657947686116,
+      "grad_norm": 0.46958115696907043,
+      "learning_rate": 4.702430846605197e-06,
+      "loss": 0.4608,
+      "step": 561
+    },
+    {
+      "epoch": 0.14134808853118713,
+      "grad_norm": 0.45457375049591064,
+      "learning_rate": 4.71081307627829e-06,
+      "loss": 0.4342,
+      "step": 562
+    },
+    {
+      "epoch": 0.14159959758551308,
+      "grad_norm": 0.4222368597984314,
+      "learning_rate": 4.719195305951383e-06,
+      "loss": 0.4364,
+      "step": 563
+    },
+    {
+      "epoch": 0.14185110663983905,
+      "grad_norm": 0.4284665882587433,
+      "learning_rate": 4.7275775356244765e-06,
+      "loss": 0.4101,
+      "step": 564
+    },
+    {
+      "epoch": 0.142102615694165,
+      "grad_norm": 0.44402772188186646,
+      "learning_rate": 4.73595976529757e-06,
+      "loss": 0.3907,
+      "step": 565
+    },
+    {
+      "epoch": 0.14235412474849093,
+      "grad_norm": 0.4303772747516632,
+      "learning_rate": 4.744341994970663e-06,
+      "loss": 0.4507,
+      "step": 566
+    },
+    {
+      "epoch": 0.1426056338028169,
+      "grad_norm": 0.4649859070777893,
+      "learning_rate": 4.752724224643755e-06,
+      "loss": 0.4351,
+      "step": 567
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.4507344365119934,
+      "learning_rate": 4.761106454316849e-06,
+      "loss": 0.4358,
+      "step": 568
+    },
+    {
+      "epoch": 0.14310865191146882,
+      "grad_norm": 0.41727519035339355,
+      "learning_rate": 4.769488683989942e-06,
+      "loss": 0.4509,
+      "step": 569
+    },
+    {
+      "epoch": 0.14336016096579476,
+      "grad_norm": 0.4347597658634186,
+      "learning_rate": 4.777870913663035e-06,
+      "loss": 0.4301,
+      "step": 570
+    },
+    {
+      "epoch": 0.14361167002012074,
+      "grad_norm": 0.46765145659446716,
+      "learning_rate": 4.7862531433361274e-06,
+      "loss": 0.4384,
+      "step": 571
+    },
+    {
+      "epoch": 0.14386317907444668,
+      "grad_norm": 0.46579083800315857,
+      "learning_rate": 4.794635373009221e-06,
+      "loss": 0.4452,
+      "step": 572
+    },
+    {
+      "epoch": 0.14411468812877262,
+      "grad_norm": 0.4726623296737671,
+      "learning_rate": 4.8030176026823135e-06,
+      "loss": 0.4457,
+      "step": 573
+    },
+    {
+      "epoch": 0.1443661971830986,
+      "grad_norm": 0.4563959240913391,
+      "learning_rate": 4.811399832355407e-06,
+      "loss": 0.4354,
+      "step": 574
+    },
+    {
+      "epoch": 0.14461770623742454,
+      "grad_norm": 0.4445574879646301,
+      "learning_rate": 4.8197820620285005e-06,
+      "loss": 0.447,
+      "step": 575
+    },
+    {
+      "epoch": 0.1448692152917505,
+      "grad_norm": 0.43714749813079834,
+      "learning_rate": 4.828164291701593e-06,
+      "loss": 0.4659,
+      "step": 576
+    },
+    {
+      "epoch": 0.14512072434607645,
+      "grad_norm": 0.4672676920890808,
+      "learning_rate": 4.836546521374686e-06,
+      "loss": 0.3998,
+      "step": 577
+    },
+    {
+      "epoch": 0.14537223340040242,
+      "grad_norm": 0.4701994061470032,
+      "learning_rate": 4.844928751047779e-06,
+      "loss": 0.4256,
+      "step": 578
+    },
+    {
+      "epoch": 0.14562374245472837,
+      "grad_norm": 0.44352054595947266,
+      "learning_rate": 4.853310980720873e-06,
+      "loss": 0.4475,
+      "step": 579
+    },
+    {
+      "epoch": 0.14587525150905434,
+      "grad_norm": 0.44086214900016785,
+      "learning_rate": 4.861693210393965e-06,
+      "loss": 0.4671,
+      "step": 580
+    },
+    {
+      "epoch": 0.14612676056338028,
+      "grad_norm": 0.4606599509716034,
+      "learning_rate": 4.870075440067058e-06,
+      "loss": 0.4333,
+      "step": 581
+    },
+    {
+      "epoch": 0.14637826961770622,
+      "grad_norm": 0.42309969663619995,
+      "learning_rate": 4.878457669740151e-06,
+      "loss": 0.4494,
+      "step": 582
+    },
+    {
+      "epoch": 0.1466297786720322,
+      "grad_norm": 0.48204946517944336,
+      "learning_rate": 4.886839899413244e-06,
+      "loss": 0.4518,
+      "step": 583
+    },
+    {
+      "epoch": 0.14688128772635814,
+      "grad_norm": 0.4570547044277191,
+      "learning_rate": 4.8952221290863375e-06,
+      "loss": 0.46,
+      "step": 584
+    },
+    {
+      "epoch": 0.1471327967806841,
+      "grad_norm": 0.4400968551635742,
+      "learning_rate": 4.903604358759431e-06,
+      "loss": 0.415,
+      "step": 585
+    },
+    {
+      "epoch": 0.14738430583501005,
+      "grad_norm": 0.48668766021728516,
+      "learning_rate": 4.911986588432524e-06,
+      "loss": 0.4209,
+      "step": 586
+    },
+    {
+      "epoch": 0.14763581488933603,
+      "grad_norm": 0.4713693857192993,
+      "learning_rate": 4.920368818105616e-06,
+      "loss": 0.4527,
+      "step": 587
+    },
+    {
+      "epoch": 0.14788732394366197,
+      "grad_norm": 0.4478401839733124,
+      "learning_rate": 4.928751047778709e-06,
+      "loss": 0.4484,
+      "step": 588
+    },
+    {
+      "epoch": 0.14813883299798794,
+      "grad_norm": 0.46621429920196533,
+      "learning_rate": 4.937133277451802e-06,
+      "loss": 0.4337,
+      "step": 589
+    },
+    {
+      "epoch": 0.14839034205231388,
+      "grad_norm": 0.4544297456741333,
+      "learning_rate": 4.945515507124896e-06,
+      "loss": 0.4596,
+      "step": 590
+    },
+    {
+      "epoch": 0.14864185110663983,
+      "grad_norm": 0.4180624186992645,
+      "learning_rate": 4.9538977367979884e-06,
+      "loss": 0.4492,
+      "step": 591
+    },
+    {
+      "epoch": 0.1488933601609658,
+      "grad_norm": 0.4281826615333557,
+      "learning_rate": 4.962279966471082e-06,
+      "loss": 0.4149,
+      "step": 592
+    },
+    {
+      "epoch": 0.14914486921529174,
+      "grad_norm": 0.4696996808052063,
+      "learning_rate": 4.9706621961441745e-06,
+      "loss": 0.4426,
+      "step": 593
+    },
+    {
+      "epoch": 0.1493963782696177,
+      "grad_norm": 0.42568251490592957,
+      "learning_rate": 4.979044425817268e-06,
+      "loss": 0.4312,
+      "step": 594
+    },
+    {
+      "epoch": 0.14964788732394366,
+      "grad_norm": 0.45928239822387695,
+      "learning_rate": 4.9874266554903615e-06,
+      "loss": 0.4318,
+      "step": 595
+    },
+    {
+      "epoch": 0.14989939637826963,
+      "grad_norm": 0.45567813515663147,
+      "learning_rate": 4.995808885163454e-06,
+      "loss": 0.457,
+      "step": 596
+    },
+    {
+      "epoch": 0.15015090543259557,
+      "grad_norm": 0.42233970761299133,
+      "learning_rate": 5.0041911148365476e-06,
+      "loss": 0.4411,
+      "step": 597
+    },
+    {
+      "epoch": 0.15040241448692154,
+      "grad_norm": 0.5426906943321228,
+      "learning_rate": 5.01257334450964e-06,
+      "loss": 0.4288,
+      "step": 598
+    },
+    {
+      "epoch": 0.1506539235412475,
+      "grad_norm": 0.43617966771125793,
+      "learning_rate": 5.020955574182733e-06,
+      "loss": 0.4557,
+      "step": 599
+    },
+    {
+      "epoch": 0.15090543259557343,
+      "grad_norm": 0.42684197425842285,
+      "learning_rate": 5.029337803855826e-06,
+      "loss": 0.4727,
+      "step": 600
+    },
+    {
+      "epoch": 0.1511569416498994,
+      "grad_norm": 0.4799075126647949,
+      "learning_rate": 5.037720033528919e-06,
+      "loss": 0.4247,
+      "step": 601
+    },
+    {
+      "epoch": 0.15140845070422534,
+      "grad_norm": 0.47558286786079407,
+      "learning_rate": 5.0461022632020115e-06,
+      "loss": 0.4396,
+      "step": 602
+    },
+    {
+      "epoch": 0.15165995975855132,
+      "grad_norm": 0.6897980570793152,
+      "learning_rate": 5.054484492875105e-06,
+      "loss": 0.4226,
+      "step": 603
+    },
+    {
+      "epoch": 0.15191146881287726,
+      "grad_norm": 0.49807223677635193,
+      "learning_rate": 5.0628667225481985e-06,
+      "loss": 0.4555,
+      "step": 604
+    },
+    {
+      "epoch": 0.15216297786720323,
+      "grad_norm": 0.4609929025173187,
+      "learning_rate": 5.071248952221292e-06,
+      "loss": 0.44,
+      "step": 605
+    },
+    {
+      "epoch": 0.15241448692152917,
+      "grad_norm": 0.42947277426719666,
+      "learning_rate": 5.079631181894385e-06,
+      "loss": 0.4162,
+      "step": 606
+    },
+    {
+      "epoch": 0.15266599597585512,
+      "grad_norm": 0.4348846971988678,
+      "learning_rate": 5.088013411567477e-06,
+      "loss": 0.4073,
+      "step": 607
+    },
+    {
+      "epoch": 0.1529175050301811,
+      "grad_norm": 0.5616949200630188,
+      "learning_rate": 5.09639564124057e-06,
+      "loss": 0.4516,
+      "step": 608
+    },
+    {
+      "epoch": 0.15316901408450703,
+      "grad_norm": 0.4498026371002197,
+      "learning_rate": 5.104777870913663e-06,
+      "loss": 0.3877,
+      "step": 609
+    },
+    {
+      "epoch": 0.153420523138833,
+      "grad_norm": 0.4951980412006378,
+      "learning_rate": 5.113160100586757e-06,
+      "loss": 0.4378,
+      "step": 610
+    },
+    {
+      "epoch": 0.15367203219315895,
+      "grad_norm": 0.5059778094291687,
+      "learning_rate": 5.12154233025985e-06,
+      "loss": 0.4475,
+      "step": 611
+    },
+    {
+      "epoch": 0.15392354124748492,
+      "grad_norm": 0.4149338901042938,
+      "learning_rate": 5.129924559932943e-06,
+      "loss": 0.451,
+      "step": 612
+    },
+    {
+      "epoch": 0.15417505030181086,
+      "grad_norm": 0.4755608141422272,
+      "learning_rate": 5.1383067896060355e-06,
+      "loss": 0.4458,
+      "step": 613
+    },
+    {
+      "epoch": 0.15442655935613683,
+      "grad_norm": 0.4765431582927704,
+      "learning_rate": 5.146689019279128e-06,
+      "loss": 0.4514,
+      "step": 614
+    },
+    {
+      "epoch": 0.15467806841046278,
+      "grad_norm": 0.445700079202652,
+      "learning_rate": 5.155071248952222e-06,
+      "loss": 0.4362,
+      "step": 615
+    },
+    {
+      "epoch": 0.15492957746478872,
+      "grad_norm": 0.46473878622055054,
+      "learning_rate": 5.163453478625314e-06,
+      "loss": 0.4453,
+      "step": 616
+    },
+    {
+      "epoch": 0.1551810865191147,
+      "grad_norm": 0.471645712852478,
+      "learning_rate": 5.1718357082984086e-06,
+      "loss": 0.4708,
+      "step": 617
+    },
+    {
+      "epoch": 0.15543259557344064,
+      "grad_norm": 0.4333302676677704,
+      "learning_rate": 5.180217937971501e-06,
+      "loss": 0.436,
+      "step": 618
+    },
+    {
+      "epoch": 0.1556841046277666,
+      "grad_norm": 0.414048433303833,
+      "learning_rate": 5.188600167644594e-06,
+      "loss": 0.4313,
+      "step": 619
+    },
+    {
+      "epoch": 0.15593561368209255,
+      "grad_norm": 0.4397493302822113,
+      "learning_rate": 5.196982397317687e-06,
+      "loss": 0.4555,
+      "step": 620
+    },
+    {
+      "epoch": 0.15618712273641852,
+      "grad_norm": 0.43406859040260315,
+      "learning_rate": 5.20536462699078e-06,
+      "loss": 0.4276,
+      "step": 621
+    },
+    {
+      "epoch": 0.15643863179074446,
+      "grad_norm": 0.39688730239868164,
+      "learning_rate": 5.2137468566638725e-06,
+      "loss": 0.4435,
+      "step": 622
+    },
+    {
+      "epoch": 0.15669014084507044,
+      "grad_norm": 0.4431197941303253,
+      "learning_rate": 5.222129086336965e-06,
+      "loss": 0.4676,
+      "step": 623
+    },
+    {
+      "epoch": 0.15694164989939638,
+      "grad_norm": 0.3907865881919861,
+      "learning_rate": 5.2305113160100595e-06,
+      "loss": 0.4142,
+      "step": 624
+    },
+    {
+      "epoch": 0.15719315895372232,
+      "grad_norm": 0.48109179735183716,
+      "learning_rate": 5.238893545683152e-06,
+      "loss": 0.4503,
+      "step": 625
+    },
+    {
+      "epoch": 0.1574446680080483,
+      "grad_norm": 0.42568618059158325,
+      "learning_rate": 5.247275775356246e-06,
+      "loss": 0.4298,
+      "step": 626
+    },
+    {
+      "epoch": 0.15769617706237424,
+      "grad_norm": 0.4349900186061859,
+      "learning_rate": 5.255658005029338e-06,
+      "loss": 0.4306,
+      "step": 627
+    },
+    {
+      "epoch": 0.1579476861167002,
+      "grad_norm": 0.45168644189834595,
+      "learning_rate": 5.264040234702431e-06,
+      "loss": 0.4328,
+      "step": 628
+    },
+    {
+      "epoch": 0.15819919517102615,
+      "grad_norm": 0.4641101360321045,
+      "learning_rate": 5.272422464375524e-06,
+      "loss": 0.4293,
+      "step": 629
+    },
+    {
+      "epoch": 0.15845070422535212,
+      "grad_norm": 0.450825035572052,
+      "learning_rate": 5.280804694048617e-06,
+      "loss": 0.4415,
+      "step": 630
+    },
+    {
+      "epoch": 0.15870221327967807,
+      "grad_norm": 0.4262521266937256,
+      "learning_rate": 5.289186923721711e-06,
+      "loss": 0.4678,
+      "step": 631
+    },
+    {
+      "epoch": 0.158953722334004,
+      "grad_norm": 0.4274531304836273,
+      "learning_rate": 5.297569153394804e-06,
+      "loss": 0.4443,
+      "step": 632
+    },
+    {
+      "epoch": 0.15920523138832998,
+      "grad_norm": 0.46715742349624634,
+      "learning_rate": 5.3059513830678965e-06,
+      "loss": 0.4056,
+      "step": 633
+    },
+    {
+      "epoch": 0.15945674044265593,
+      "grad_norm": 0.4867947995662689,
+      "learning_rate": 5.314333612740989e-06,
+      "loss": 0.4539,
+      "step": 634
+    },
+    {
+      "epoch": 0.1597082494969819,
+      "grad_norm": 0.44424811005592346,
+      "learning_rate": 5.322715842414083e-06,
+      "loss": 0.4558,
+      "step": 635
+    },
+    {
+      "epoch": 0.15995975855130784,
+      "grad_norm": 0.4673738181591034,
+      "learning_rate": 5.331098072087175e-06,
+      "loss": 0.4272,
+      "step": 636
+    },
+    {
+      "epoch": 0.1602112676056338,
+      "grad_norm": 0.4130132794380188,
+      "learning_rate": 5.3394803017602695e-06,
+      "loss": 0.4446,
+      "step": 637
+    },
+    {
+      "epoch": 0.16046277665995975,
+      "grad_norm": 0.43853166699409485,
+      "learning_rate": 5.347862531433362e-06,
+      "loss": 0.4472,
+      "step": 638
+    },
+    {
+      "epoch": 0.16071428571428573,
+      "grad_norm": 0.4459279477596283,
+      "learning_rate": 5.356244761106455e-06,
+      "loss": 0.4636,
+      "step": 639
+    },
+    {
+      "epoch": 0.16096579476861167,
+      "grad_norm": 0.4516228437423706,
+      "learning_rate": 5.364626990779548e-06,
+      "loss": 0.4081,
+      "step": 640
+    },
+    {
+      "epoch": 0.1612173038229376,
+      "grad_norm": 0.41011399030685425,
+      "learning_rate": 5.373009220452641e-06,
+      "loss": 0.4401,
+      "step": 641
+    },
+    {
+      "epoch": 0.16146881287726358,
+      "grad_norm": 0.3807559311389923,
+      "learning_rate": 5.3813914501257335e-06,
+      "loss": 0.4051,
+      "step": 642
+    },
+    {
+      "epoch": 0.16172032193158953,
+      "grad_norm": 0.41934090852737427,
+      "learning_rate": 5.389773679798826e-06,
+      "loss": 0.4234,
+      "step": 643
+    },
+    {
+      "epoch": 0.1619718309859155,
+      "grad_norm": 0.39974445104599,
+      "learning_rate": 5.3981559094719205e-06,
+      "loss": 0.421,
+      "step": 644
+    },
+    {
+      "epoch": 0.16222334004024144,
+      "grad_norm": 0.4237842261791229,
+      "learning_rate": 5.406538139145013e-06,
+      "loss": 0.4301,
+      "step": 645
+    },
+    {
+      "epoch": 0.16247484909456741,
+      "grad_norm": 0.4234157204627991,
+      "learning_rate": 5.4149203688181066e-06,
+      "loss": 0.414,
+      "step": 646
+    },
+    {
+      "epoch": 0.16272635814889336,
+      "grad_norm": 0.4288012385368347,
+      "learning_rate": 5.423302598491199e-06,
+      "loss": 0.4019,
+      "step": 647
+    },
+    {
+      "epoch": 0.16297786720321933,
+      "grad_norm": 0.4307977855205536,
+      "learning_rate": 5.431684828164292e-06,
+      "loss": 0.4164,
+      "step": 648
+    },
+    {
+      "epoch": 0.16322937625754527,
+      "grad_norm": 0.4072883725166321,
+      "learning_rate": 5.4400670578373844e-06,
+      "loss": 0.4237,
+      "step": 649
+    },
+    {
+      "epoch": 0.16348088531187122,
+      "grad_norm": 0.44802358746528625,
+      "learning_rate": 5.448449287510478e-06,
+      "loss": 0.4257,
+      "step": 650
+    },
+    {
+      "epoch": 0.1637323943661972,
+      "grad_norm": 0.44640132784843445,
+      "learning_rate": 5.456831517183571e-06,
+      "loss": 0.4128,
+      "step": 651
+    },
+    {
+      "epoch": 0.16398390342052313,
+      "grad_norm": 0.45102569460868835,
+      "learning_rate": 5.465213746856665e-06,
+      "loss": 0.4205,
+      "step": 652
+    },
+    {
+      "epoch": 0.1642354124748491,
+      "grad_norm": 0.4356350004673004,
+      "learning_rate": 5.4735959765297575e-06,
+      "loss": 0.4444,
+      "step": 653
+    },
+    {
+      "epoch": 0.16448692152917505,
+      "grad_norm": 0.4442964196205139,
+      "learning_rate": 5.48197820620285e-06,
+      "loss": 0.4406,
+      "step": 654
+    },
+    {
+      "epoch": 0.16473843058350102,
+      "grad_norm": 0.5060700178146362,
+      "learning_rate": 5.490360435875944e-06,
+      "loss": 0.4425,
+      "step": 655
+    },
+    {
+      "epoch": 0.16498993963782696,
+      "grad_norm": 0.45222505927085876,
+      "learning_rate": 5.498742665549036e-06,
+      "loss": 0.4296,
+      "step": 656
+    },
+    {
+      "epoch": 0.16524144869215293,
+      "grad_norm": 0.47446000576019287,
+      "learning_rate": 5.507124895222129e-06,
+      "loss": 0.4272,
+      "step": 657
+    },
+    {
+      "epoch": 0.16549295774647887,
+      "grad_norm": 0.4703046977519989,
+      "learning_rate": 5.515507124895223e-06,
+      "loss": 0.4174,
+      "step": 658
+    },
+    {
+      "epoch": 0.16574446680080482,
+      "grad_norm": 0.48616233468055725,
+      "learning_rate": 5.523889354568316e-06,
+      "loss": 0.4248,
+      "step": 659
+    },
+    {
+      "epoch": 0.1659959758551308,
+      "grad_norm": 0.4568127989768982,
+      "learning_rate": 5.532271584241408e-06,
+      "loss": 0.4402,
+      "step": 660
+    },
+    {
+      "epoch": 0.16624748490945673,
+      "grad_norm": 0.44677838683128357,
+      "learning_rate": 5.540653813914502e-06,
+      "loss": 0.4516,
+      "step": 661
+    },
+    {
+      "epoch": 0.1664989939637827,
+      "grad_norm": 0.42325931787490845,
+      "learning_rate": 5.5490360435875945e-06,
+      "loss": 0.4393,
+      "step": 662
+    },
+    {
+      "epoch": 0.16675050301810865,
+      "grad_norm": 0.4119207561016083,
+      "learning_rate": 5.557418273260687e-06,
+      "loss": 0.3994,
+      "step": 663
+    },
+    {
+      "epoch": 0.16700201207243462,
+      "grad_norm": 0.45278915762901306,
+      "learning_rate": 5.5658005029337815e-06,
+      "loss": 0.4034,
+      "step": 664
+    },
+    {
+      "epoch": 0.16725352112676056,
+      "grad_norm": 0.5348659157752991,
+      "learning_rate": 5.574182732606874e-06,
+      "loss": 0.4471,
+      "step": 665
+    },
+    {
+      "epoch": 0.1675050301810865,
+      "grad_norm": 0.44384071230888367,
+      "learning_rate": 5.5825649622799676e-06,
+      "loss": 0.4334,
+      "step": 666
+    },
+    {
+      "epoch": 0.16775653923541248,
+      "grad_norm": 0.4300418496131897,
+      "learning_rate": 5.59094719195306e-06,
+      "loss": 0.4094,
+      "step": 667
+    },
+    {
+      "epoch": 0.16800804828973842,
+      "grad_norm": 0.4564829170703888,
+      "learning_rate": 5.599329421626153e-06,
+      "loss": 0.4138,
+      "step": 668
+    },
+    {
+      "epoch": 0.1682595573440644,
+      "grad_norm": 0.5108182430267334,
+      "learning_rate": 5.6077116512992454e-06,
+      "loss": 0.4135,
+      "step": 669
+    },
+    {
+      "epoch": 0.16851106639839034,
+      "grad_norm": 0.4513753652572632,
+      "learning_rate": 5.616093880972339e-06,
+      "loss": 0.456,
+      "step": 670
+    },
+    {
+      "epoch": 0.1687625754527163,
+      "grad_norm": 0.4487459659576416,
+      "learning_rate": 5.624476110645432e-06,
+      "loss": 0.4374,
+      "step": 671
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 0.48467063903808594,
+      "learning_rate": 5.632858340318526e-06,
+      "loss": 0.4575,
+      "step": 672
+    },
+    {
+      "epoch": 0.16926559356136822,
+      "grad_norm": 0.47948938608169556,
+      "learning_rate": 5.6412405699916185e-06,
+      "loss": 0.4399,
+      "step": 673
+    },
+    {
+      "epoch": 0.16951710261569417,
+      "grad_norm": 0.4448109269142151,
+      "learning_rate": 5.649622799664711e-06,
+      "loss": 0.4395,
+      "step": 674
+    },
+    {
+      "epoch": 0.1697686116700201,
+      "grad_norm": 0.44593971967697144,
+      "learning_rate": 5.658005029337804e-06,
+      "loss": 0.4382,
+      "step": 675
+    },
+    {
+      "epoch": 0.17002012072434608,
+      "grad_norm": 0.4504484236240387,
+      "learning_rate": 5.666387259010897e-06,
+      "loss": 0.4252,
+      "step": 676
+    },
+    {
+      "epoch": 0.17027162977867202,
+      "grad_norm": 0.5128602981567383,
+      "learning_rate": 5.67476948868399e-06,
+      "loss": 0.4444,
+      "step": 677
+    },
+    {
+      "epoch": 0.170523138832998,
+      "grad_norm": 0.4349728226661682,
+      "learning_rate": 5.683151718357084e-06,
+      "loss": 0.4152,
+      "step": 678
+    },
+    {
+      "epoch": 0.17077464788732394,
+      "grad_norm": 0.4326830804347992,
+      "learning_rate": 5.691533948030177e-06,
+      "loss": 0.4416,
+      "step": 679
+    },
+    {
+      "epoch": 0.1710261569416499,
+      "grad_norm": 0.42152220010757446,
+      "learning_rate": 5.699916177703269e-06,
+      "loss": 0.4363,
+      "step": 680
+    },
+    {
+      "epoch": 0.17127766599597585,
+      "grad_norm": 0.4485848844051361,
+      "learning_rate": 5.708298407376363e-06,
+      "loss": 0.4467,
+      "step": 681
+    },
+    {
+      "epoch": 0.17152917505030182,
+      "grad_norm": 0.42400580644607544,
+      "learning_rate": 5.7166806370494555e-06,
+      "loss": 0.4293,
+      "step": 682
+    },
+    {
+      "epoch": 0.17178068410462777,
+      "grad_norm": 0.4600382447242737,
+      "learning_rate": 5.725062866722548e-06,
+      "loss": 0.4293,
+      "step": 683
+    },
+    {
+      "epoch": 0.1720321931589537,
+      "grad_norm": 0.5220768451690674,
+      "learning_rate": 5.733445096395641e-06,
+      "loss": 0.3961,
+      "step": 684
+    },
+    {
+      "epoch": 0.17228370221327968,
+      "grad_norm": 0.5387700796127319,
+      "learning_rate": 5.741827326068735e-06,
+      "loss": 0.468,
+      "step": 685
+    },
+    {
+      "epoch": 0.17253521126760563,
+      "grad_norm": 0.4986119270324707,
+      "learning_rate": 5.750209555741828e-06,
+      "loss": 0.4452,
+      "step": 686
+    },
+    {
+      "epoch": 0.1727867203219316,
+      "grad_norm": 0.4661785066127777,
+      "learning_rate": 5.758591785414921e-06,
+      "loss": 0.4413,
+      "step": 687
+    },
+    {
+      "epoch": 0.17303822937625754,
+      "grad_norm": 0.4665091335773468,
+      "learning_rate": 5.766974015088014e-06,
+      "loss": 0.4327,
+      "step": 688
+    },
+    {
+      "epoch": 0.1732897384305835,
+      "grad_norm": 0.46862325072288513,
+      "learning_rate": 5.7753562447611064e-06,
+      "loss": 0.4422,
+      "step": 689
+    },
+    {
+      "epoch": 0.17354124748490946,
+      "grad_norm": 0.47405147552490234,
+      "learning_rate": 5.7837384744342e-06,
+      "loss": 0.4377,
+      "step": 690
+    },
+    {
+      "epoch": 0.1737927565392354,
+      "grad_norm": 0.49012836813926697,
+      "learning_rate": 5.792120704107293e-06,
+      "loss": 0.45,
+      "step": 691
+    },
+    {
+      "epoch": 0.17404426559356137,
+      "grad_norm": 0.49610865116119385,
+      "learning_rate": 5.800502933780387e-06,
+      "loss": 0.405,
+      "step": 692
+    },
+    {
+      "epoch": 0.1742957746478873,
+      "grad_norm": 0.5310260653495789,
+      "learning_rate": 5.8088851634534795e-06,
+      "loss": 0.4148,
+      "step": 693
+    },
+    {
+      "epoch": 0.17454728370221329,
+      "grad_norm": 0.4385521411895752,
+      "learning_rate": 5.817267393126572e-06,
+      "loss": 0.3932,
+      "step": 694
+    },
+    {
+      "epoch": 0.17479879275653923,
+      "grad_norm": 0.44616925716400146,
+      "learning_rate": 5.825649622799665e-06,
+      "loss": 0.4482,
+      "step": 695
+    },
+    {
+      "epoch": 0.1750503018108652,
+      "grad_norm": 0.5454944372177124,
+      "learning_rate": 5.834031852472758e-06,
+      "loss": 0.4579,
+      "step": 696
+    },
+    {
+      "epoch": 0.17530181086519114,
+      "grad_norm": 0.4732443690299988,
+      "learning_rate": 5.842414082145851e-06,
+      "loss": 0.4198,
+      "step": 697
+    },
+    {
+      "epoch": 0.17555331991951711,
+      "grad_norm": 0.5170097947120667,
+      "learning_rate": 5.850796311818945e-06,
+      "loss": 0.4551,
+      "step": 698
+    },
+    {
+      "epoch": 0.17580482897384306,
+      "grad_norm": 0.4894830584526062,
+      "learning_rate": 5.859178541492038e-06,
+      "loss": 0.411,
+      "step": 699
+    },
+    {
+      "epoch": 0.176056338028169,
+      "grad_norm": 0.47865331172943115,
+      "learning_rate": 5.86756077116513e-06,
+      "loss": 0.4162,
+      "step": 700
+    },
+    {
+      "epoch": 0.17630784708249497,
+      "grad_norm": 0.5109310150146484,
+      "learning_rate": 5.875943000838223e-06,
+      "loss": 0.4255,
+      "step": 701
+    },
+    {
+      "epoch": 0.17655935613682092,
+      "grad_norm": 0.4590068459510803,
+      "learning_rate": 5.8843252305113165e-06,
+      "loss": 0.4096,
+      "step": 702
+    },
+    {
+      "epoch": 0.1768108651911469,
+      "grad_norm": 0.5353825092315674,
+      "learning_rate": 5.892707460184409e-06,
+      "loss": 0.4477,
+      "step": 703
+    },
+    {
+      "epoch": 0.17706237424547283,
+      "grad_norm": 0.4406241178512573,
+      "learning_rate": 5.901089689857502e-06,
+      "loss": 0.4037,
+      "step": 704
+    },
+    {
+      "epoch": 0.1773138832997988,
+      "grad_norm": 0.4536992609500885,
+      "learning_rate": 5.909471919530596e-06,
+      "loss": 0.4397,
+      "step": 705
+    },
+    {
+      "epoch": 0.17756539235412475,
+      "grad_norm": 0.4446880519390106,
+      "learning_rate": 5.917854149203689e-06,
+      "loss": 0.4454,
+      "step": 706
+    },
+    {
+      "epoch": 0.17781690140845072,
+      "grad_norm": 0.4549560546875,
+      "learning_rate": 5.926236378876782e-06,
+      "loss": 0.4199,
+      "step": 707
+    },
+    {
+      "epoch": 0.17806841046277666,
+      "grad_norm": 0.4807390570640564,
+      "learning_rate": 5.934618608549875e-06,
+      "loss": 0.4336,
+      "step": 708
+    },
+    {
+      "epoch": 0.1783199195171026,
+      "grad_norm": 0.4760938882827759,
+      "learning_rate": 5.943000838222967e-06,
+      "loss": 0.4095,
+      "step": 709
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 0.4646598696708679,
+      "learning_rate": 5.95138306789606e-06,
+      "loss": 0.3987,
+      "step": 710
+    },
+    {
+      "epoch": 0.17882293762575452,
+      "grad_norm": 0.5018036961555481,
+      "learning_rate": 5.9597652975691535e-06,
+      "loss": 0.4004,
+      "step": 711
+    },
+    {
+      "epoch": 0.1790744466800805,
+      "grad_norm": 0.4650225341320038,
+      "learning_rate": 5.968147527242247e-06,
+      "loss": 0.4507,
+      "step": 712
+    },
+    {
+      "epoch": 0.17932595573440643,
+      "grad_norm": 0.44860950112342834,
+      "learning_rate": 5.9765297569153405e-06,
+      "loss": 0.44,
+      "step": 713
+    },
+    {
+      "epoch": 0.1795774647887324,
+      "grad_norm": 0.46424388885498047,
+      "learning_rate": 5.984911986588433e-06,
+      "loss": 0.4355,
+      "step": 714
+    },
+    {
+      "epoch": 0.17982897384305835,
+      "grad_norm": 0.49789562821388245,
+      "learning_rate": 5.993294216261526e-06,
+      "loss": 0.4256,
+      "step": 715
+    },
+    {
+      "epoch": 0.18008048289738432,
+      "grad_norm": 0.4016691744327545,
+      "learning_rate": 6.001676445934619e-06,
+      "loss": 0.4454,
+      "step": 716
+    },
+    {
+      "epoch": 0.18033199195171026,
+      "grad_norm": 0.47032442688941956,
+      "learning_rate": 6.010058675607712e-06,
+      "loss": 0.4023,
+      "step": 717
+    },
+    {
+      "epoch": 0.1805835010060362,
+      "grad_norm": 0.4652996063232422,
+      "learning_rate": 6.018440905280806e-06,
+      "loss": 0.4393,
+      "step": 718
+    },
+    {
+      "epoch": 0.18083501006036218,
+      "grad_norm": 0.4808005392551422,
+      "learning_rate": 6.026823134953899e-06,
+      "loss": 0.4102,
+      "step": 719
+    },
+    {
+      "epoch": 0.18108651911468812,
+      "grad_norm": 0.42760953307151794,
+      "learning_rate": 6.035205364626991e-06,
+      "loss": 0.4416,
+      "step": 720
+    },
+    {
+      "epoch": 0.1813380281690141,
+      "grad_norm": 0.4750773012638092,
+      "learning_rate": 6.043587594300084e-06,
+      "loss": 0.4321,
+      "step": 721
+    },
+    {
+      "epoch": 0.18158953722334004,
+      "grad_norm": 0.4393182098865509,
+      "learning_rate": 6.0519698239731775e-06,
+      "loss": 0.4264,
+      "step": 722
+    },
+    {
+      "epoch": 0.181841046277666,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 6.06035205364627e-06,
+      "loss": 0.4255,
+      "step": 723
+    },
+    {
+      "epoch": 0.18209255533199195,
+      "grad_norm": 0.4402623772621155,
+      "learning_rate": 6.068734283319363e-06,
+      "loss": 0.4308,
+      "step": 724
+    },
+    {
+      "epoch": 0.1823440643863179,
+      "grad_norm": 0.4287785589694977,
+      "learning_rate": 6.077116512992457e-06,
+      "loss": 0.4413,
+      "step": 725
+    },
+    {
+      "epoch": 0.18259557344064387,
+      "grad_norm": 0.49795329570770264,
+      "learning_rate": 6.08549874266555e-06,
+      "loss": 0.4398,
+      "step": 726
+    },
+    {
+      "epoch": 0.1828470824949698,
+      "grad_norm": 0.4231499433517456,
+      "learning_rate": 6.093880972338643e-06,
+      "loss": 0.4207,
+      "step": 727
+    },
+    {
+      "epoch": 0.18309859154929578,
+      "grad_norm": 0.4155465066432953,
+      "learning_rate": 6.102263202011736e-06,
+      "loss": 0.4195,
+      "step": 728
+    },
+    {
+      "epoch": 0.18335010060362172,
+      "grad_norm": 0.4875541925430298,
+      "learning_rate": 6.110645431684828e-06,
+      "loss": 0.4675,
+      "step": 729
+    },
+    {
+      "epoch": 0.1836016096579477,
+      "grad_norm": 0.46131083369255066,
+      "learning_rate": 6.119027661357921e-06,
+      "loss": 0.4296,
+      "step": 730
+    },
+    {
+      "epoch": 0.18385311871227364,
+      "grad_norm": 0.46306151151657104,
+      "learning_rate": 6.1274098910310145e-06,
+      "loss": 0.4315,
+      "step": 731
+    },
+    {
+      "epoch": 0.1841046277665996,
+      "grad_norm": 0.4583171308040619,
+      "learning_rate": 6.135792120704108e-06,
+      "loss": 0.3973,
+      "step": 732
+    },
+    {
+      "epoch": 0.18435613682092555,
+      "grad_norm": 0.4396132528781891,
+      "learning_rate": 6.1441743503772014e-06,
+      "loss": 0.4166,
+      "step": 733
+    },
+    {
+      "epoch": 0.1846076458752515,
+      "grad_norm": 0.45620885491371155,
+      "learning_rate": 6.152556580050294e-06,
+      "loss": 0.4134,
+      "step": 734
+    },
+    {
+      "epoch": 0.18485915492957747,
+      "grad_norm": 0.48990169167518616,
+      "learning_rate": 6.160938809723387e-06,
+      "loss": 0.4615,
+      "step": 735
+    },
+    {
+      "epoch": 0.1851106639839034,
+      "grad_norm": 0.4548690617084503,
+      "learning_rate": 6.169321039396479e-06,
+      "loss": 0.4019,
+      "step": 736
+    },
+    {
+      "epoch": 0.18536217303822938,
+      "grad_norm": 0.4465884864330292,
+      "learning_rate": 6.177703269069573e-06,
+      "loss": 0.4115,
+      "step": 737
+    },
+    {
+      "epoch": 0.18561368209255533,
+      "grad_norm": 0.5192622542381287,
+      "learning_rate": 6.1860854987426654e-06,
+      "loss": 0.4071,
+      "step": 738
+    },
+    {
+      "epoch": 0.1858651911468813,
+      "grad_norm": 0.5081340074539185,
+      "learning_rate": 6.19446772841576e-06,
+      "loss": 0.4464,
+      "step": 739
+    },
+    {
+      "epoch": 0.18611670020120724,
+      "grad_norm": 0.6187810301780701,
+      "learning_rate": 6.202849958088852e-06,
+      "loss": 0.4386,
+      "step": 740
+    },
+    {
+      "epoch": 0.1863682092555332,
+      "grad_norm": 0.45214521884918213,
+      "learning_rate": 6.211232187761945e-06,
+      "loss": 0.4626,
+      "step": 741
+    },
+    {
+      "epoch": 0.18661971830985916,
+      "grad_norm": 0.5023926496505737,
+      "learning_rate": 6.2196144174350385e-06,
+      "loss": 0.4768,
+      "step": 742
+    },
+    {
+      "epoch": 0.1868712273641851,
+      "grad_norm": 0.49080634117126465,
+      "learning_rate": 6.227996647108131e-06,
+      "loss": 0.4381,
+      "step": 743
+    },
+    {
+      "epoch": 0.18712273641851107,
+      "grad_norm": 0.5521237850189209,
+      "learning_rate": 6.236378876781224e-06,
+      "loss": 0.4458,
+      "step": 744
+    },
+    {
+      "epoch": 0.18737424547283701,
+      "grad_norm": 0.46638408303260803,
+      "learning_rate": 6.244761106454318e-06,
+      "loss": 0.4433,
+      "step": 745
+    },
+    {
+      "epoch": 0.18762575452716299,
+      "grad_norm": 0.5203446745872498,
+      "learning_rate": 6.253143336127411e-06,
+      "loss": 0.4255,
+      "step": 746
+    },
+    {
+      "epoch": 0.18787726358148893,
+      "grad_norm": 0.5413115620613098,
+      "learning_rate": 6.261525565800503e-06,
+      "loss": 0.4302,
+      "step": 747
+    },
+    {
+      "epoch": 0.1881287726358149,
+      "grad_norm": 0.5347134470939636,
+      "learning_rate": 6.269907795473597e-06,
+      "loss": 0.4414,
+      "step": 748
+    },
+    {
+      "epoch": 0.18838028169014084,
+      "grad_norm": 0.43149664998054504,
+      "learning_rate": 6.278290025146689e-06,
+      "loss": 0.4167,
+      "step": 749
+    },
+    {
+      "epoch": 0.1886317907444668,
+      "grad_norm": 0.522585391998291,
+      "learning_rate": 6.286672254819782e-06,
+      "loss": 0.4172,
+      "step": 750
+    },
+    {
+      "epoch": 0.18888329979879276,
+      "grad_norm": 0.5557539463043213,
+      "learning_rate": 6.2950544844928755e-06,
+      "loss": 0.436,
+      "step": 751
+    },
+    {
+      "epoch": 0.1891348088531187,
+      "grad_norm": 0.47711122035980225,
+      "learning_rate": 6.303436714165969e-06,
+      "loss": 0.4602,
+      "step": 752
+    },
+    {
+      "epoch": 0.18938631790744467,
+      "grad_norm": 0.5225254893302917,
+      "learning_rate": 6.3118189438390624e-06,
+      "loss": 0.438,
+      "step": 753
+    },
+    {
+      "epoch": 0.18963782696177062,
+      "grad_norm": 0.4202134907245636,
+      "learning_rate": 6.320201173512155e-06,
+      "loss": 0.4453,
+      "step": 754
+    },
+    {
+      "epoch": 0.1898893360160966,
+      "grad_norm": 0.48253220319747925,
+      "learning_rate": 6.328583403185248e-06,
+      "loss": 0.4296,
+      "step": 755
+    },
+    {
+      "epoch": 0.19014084507042253,
+      "grad_norm": 0.525970995426178,
+      "learning_rate": 6.33696563285834e-06,
+      "loss": 0.4248,
+      "step": 756
+    },
+    {
+      "epoch": 0.1903923541247485,
+      "grad_norm": 0.49865517020225525,
+      "learning_rate": 6.345347862531434e-06,
+      "loss": 0.4083,
+      "step": 757
+    },
+    {
+      "epoch": 0.19064386317907445,
+      "grad_norm": 0.5443746447563171,
+      "learning_rate": 6.353730092204526e-06,
+      "loss": 0.4326,
+      "step": 758
+    },
+    {
+      "epoch": 0.1908953722334004,
+      "grad_norm": 0.5639643669128418,
+      "learning_rate": 6.362112321877621e-06,
+      "loss": 0.4212,
+      "step": 759
+    },
+    {
+      "epoch": 0.19114688128772636,
+      "grad_norm": 0.4444926381111145,
+      "learning_rate": 6.370494551550713e-06,
+      "loss": 0.407,
+      "step": 760
+    },
+    {
+      "epoch": 0.1913983903420523,
+      "grad_norm": 0.5601978898048401,
+      "learning_rate": 6.378876781223806e-06,
+      "loss": 0.4492,
+      "step": 761
+    },
+    {
+      "epoch": 0.19164989939637828,
+      "grad_norm": 0.5076690912246704,
+      "learning_rate": 6.387259010896899e-06,
+      "loss": 0.4241,
+      "step": 762
+    },
+    {
+      "epoch": 0.19190140845070422,
+      "grad_norm": 0.45043453574180603,
+      "learning_rate": 6.395641240569992e-06,
+      "loss": 0.4345,
+      "step": 763
+    },
+    {
+      "epoch": 0.1921529175050302,
+      "grad_norm": 0.413898229598999,
+      "learning_rate": 6.404023470243085e-06,
+      "loss": 0.4396,
+      "step": 764
+    },
+    {
+      "epoch": 0.19240442655935613,
+      "grad_norm": 0.43958157300949097,
+      "learning_rate": 6.412405699916177e-06,
+      "loss": 0.4294,
+      "step": 765
+    },
+    {
+      "epoch": 0.1926559356136821,
+      "grad_norm": 0.46111080050468445,
+      "learning_rate": 6.420787929589272e-06,
+      "loss": 0.4508,
+      "step": 766
+    },
+    {
+      "epoch": 0.19290744466800805,
+      "grad_norm": 0.4380728602409363,
+      "learning_rate": 6.429170159262364e-06,
+      "loss": 0.4308,
+      "step": 767
+    },
+    {
+      "epoch": 0.193158953722334,
+      "grad_norm": 0.47174325585365295,
+      "learning_rate": 6.437552388935458e-06,
+      "loss": 0.4047,
+      "step": 768
+    },
+    {
+      "epoch": 0.19341046277665996,
+      "grad_norm": 0.5069427490234375,
+      "learning_rate": 6.44593461860855e-06,
+      "loss": 0.4623,
+      "step": 769
+    },
+    {
+      "epoch": 0.1936619718309859,
+      "grad_norm": 0.47728949785232544,
+      "learning_rate": 6.454316848281643e-06,
+      "loss": 0.4226,
+      "step": 770
+    },
+    {
+      "epoch": 0.19391348088531188,
+      "grad_norm": 0.4369402229785919,
+      "learning_rate": 6.462699077954736e-06,
+      "loss": 0.4355,
+      "step": 771
+    },
+    {
+      "epoch": 0.19416498993963782,
+      "grad_norm": 0.49592912197113037,
+      "learning_rate": 6.47108130762783e-06,
+      "loss": 0.4504,
+      "step": 772
+    },
+    {
+      "epoch": 0.1944164989939638,
+      "grad_norm": 0.4935101270675659,
+      "learning_rate": 6.479463537300923e-06,
+      "loss": 0.4434,
+      "step": 773
+    },
+    {
+      "epoch": 0.19466800804828974,
+      "grad_norm": 0.4729015827178955,
+      "learning_rate": 6.487845766974016e-06,
+      "loss": 0.4066,
+      "step": 774
+    },
+    {
+      "epoch": 0.19491951710261568,
+      "grad_norm": 0.5029162168502808,
+      "learning_rate": 6.496227996647109e-06,
+      "loss": 0.4262,
+      "step": 775
+    },
+    {
+      "epoch": 0.19517102615694165,
+      "grad_norm": 0.6650319695472717,
+      "learning_rate": 6.504610226320201e-06,
+      "loss": 0.434,
+      "step": 776
+    },
+    {
+      "epoch": 0.1954225352112676,
+      "grad_norm": 0.464813768863678,
+      "learning_rate": 6.512992455993295e-06,
+      "loss": 0.4368,
+      "step": 777
+    },
+    {
+      "epoch": 0.19567404426559357,
+      "grad_norm": 0.5334668755531311,
+      "learning_rate": 6.521374685666387e-06,
+      "loss": 0.4279,
+      "step": 778
+    },
+    {
+      "epoch": 0.1959255533199195,
+      "grad_norm": 0.6014412641525269,
+      "learning_rate": 6.529756915339482e-06,
+      "loss": 0.4236,
+      "step": 779
+    },
+    {
+      "epoch": 0.19617706237424548,
+      "grad_norm": 0.48810648918151855,
+      "learning_rate": 6.538139145012574e-06,
+      "loss": 0.4528,
+      "step": 780
+    },
+    {
+      "epoch": 0.19642857142857142,
+      "grad_norm": 0.5004664063453674,
+      "learning_rate": 6.546521374685667e-06,
+      "loss": 0.4335,
+      "step": 781
+    },
+    {
+      "epoch": 0.1966800804828974,
+      "grad_norm": 0.5810479521751404,
+      "learning_rate": 6.55490360435876e-06,
+      "loss": 0.4348,
+      "step": 782
+    },
+    {
+      "epoch": 0.19693158953722334,
+      "grad_norm": 0.5041718482971191,
+      "learning_rate": 6.563285834031853e-06,
+      "loss": 0.4341,
+      "step": 783
+    },
+    {
+      "epoch": 0.19718309859154928,
+      "grad_norm": 0.498522013425827,
+      "learning_rate": 6.571668063704946e-06,
+      "loss": 0.4269,
+      "step": 784
+    },
+    {
+      "epoch": 0.19743460764587525,
+      "grad_norm": 0.45353996753692627,
+      "learning_rate": 6.580050293378038e-06,
+      "loss": 0.4199,
+      "step": 785
+    },
+    {
+      "epoch": 0.1976861167002012,
+      "grad_norm": 0.45865288376808167,
+      "learning_rate": 6.588432523051133e-06,
+      "loss": 0.4335,
+      "step": 786
+    },
+    {
+      "epoch": 0.19793762575452717,
+      "grad_norm": 0.4978802800178528,
+      "learning_rate": 6.596814752724225e-06,
+      "loss": 0.424,
+      "step": 787
+    },
+    {
+      "epoch": 0.1981891348088531,
+      "grad_norm": 0.45598986744880676,
+      "learning_rate": 6.605196982397319e-06,
+      "loss": 0.4375,
+      "step": 788
+    },
+    {
+      "epoch": 0.19844064386317908,
+      "grad_norm": 0.448933869600296,
+      "learning_rate": 6.613579212070411e-06,
+      "loss": 0.4242,
+      "step": 789
+    },
+    {
+      "epoch": 0.19869215291750503,
+      "grad_norm": 0.4256827235221863,
+      "learning_rate": 6.621961441743504e-06,
+      "loss": 0.4396,
+      "step": 790
+    },
+    {
+      "epoch": 0.198943661971831,
+      "grad_norm": 0.4087207317352295,
+      "learning_rate": 6.630343671416597e-06,
+      "loss": 0.3984,
+      "step": 791
+    },
+    {
+      "epoch": 0.19919517102615694,
+      "grad_norm": 0.5045828819274902,
+      "learning_rate": 6.638725901089691e-06,
+      "loss": 0.4333,
+      "step": 792
+    },
+    {
+      "epoch": 0.19944668008048289,
+      "grad_norm": 0.4137192368507385,
+      "learning_rate": 6.6471081307627836e-06,
+      "loss": 0.4048,
+      "step": 793
+    },
+    {
+      "epoch": 0.19969818913480886,
+      "grad_norm": 0.43266329169273376,
+      "learning_rate": 6.655490360435877e-06,
+      "loss": 0.4228,
+      "step": 794
+    },
+    {
+      "epoch": 0.1999496981891348,
+      "grad_norm": 0.43893131613731384,
+      "learning_rate": 6.66387259010897e-06,
+      "loss": 0.4284,
+      "step": 795
+    },
+    {
+      "epoch": 0.20020120724346077,
+      "grad_norm": 0.4277397096157074,
+      "learning_rate": 6.672254819782062e-06,
+      "loss": 0.4379,
+      "step": 796
+    },
+    {
+      "epoch": 0.20045271629778671,
+      "grad_norm": 0.4732038974761963,
+      "learning_rate": 6.680637049455155e-06,
+      "loss": 0.4393,
+      "step": 797
+    },
+    {
+      "epoch": 0.2007042253521127,
+      "grad_norm": 0.40471675992012024,
+      "learning_rate": 6.689019279128248e-06,
+      "loss": 0.4128,
+      "step": 798
+    },
+    {
+      "epoch": 0.20095573440643863,
+      "grad_norm": 0.47634950280189514,
+      "learning_rate": 6.697401508801342e-06,
+      "loss": 0.4163,
+      "step": 799
+    },
+    {
+      "epoch": 0.2012072434607646,
+      "grad_norm": 0.44929447770118713,
+      "learning_rate": 6.705783738474435e-06,
+      "loss": 0.4569,
+      "step": 800
+    },
+    {
+      "epoch": 0.20145875251509054,
+      "grad_norm": 0.40945762395858765,
+      "learning_rate": 6.714165968147528e-06,
+      "loss": 0.4021,
+      "step": 801
+    },
+    {
+      "epoch": 0.2017102615694165,
+      "grad_norm": 0.4612343907356262,
+      "learning_rate": 6.722548197820621e-06,
+      "loss": 0.4204,
+      "step": 802
+    },
+    {
+      "epoch": 0.20196177062374246,
+      "grad_norm": 0.40259966254234314,
+      "learning_rate": 6.730930427493714e-06,
+      "loss": 0.4317,
+      "step": 803
+    },
+    {
+      "epoch": 0.2022132796780684,
+      "grad_norm": 0.4034963548183441,
+      "learning_rate": 6.739312657166807e-06,
+      "loss": 0.4112,
+      "step": 804
+    },
+    {
+      "epoch": 0.20246478873239437,
+      "grad_norm": 0.4619353413581848,
+      "learning_rate": 6.747694886839899e-06,
+      "loss": 0.4341,
+      "step": 805
+    },
+    {
+      "epoch": 0.20271629778672032,
+      "grad_norm": 0.40814584493637085,
+      "learning_rate": 6.756077116512994e-06,
+      "loss": 0.4438,
+      "step": 806
+    },
+    {
+      "epoch": 0.2029678068410463,
+      "grad_norm": 0.47560909390449524,
+      "learning_rate": 6.764459346186086e-06,
+      "loss": 0.4254,
+      "step": 807
+    },
+    {
+      "epoch": 0.20321931589537223,
+      "grad_norm": 0.43871134519577026,
+      "learning_rate": 6.772841575859179e-06,
+      "loss": 0.4295,
+      "step": 808
+    },
+    {
+      "epoch": 0.20347082494969818,
+      "grad_norm": 0.49143895506858826,
+      "learning_rate": 6.781223805532272e-06,
+      "loss": 0.4294,
+      "step": 809
+    },
+    {
+      "epoch": 0.20372233400402415,
+      "grad_norm": 0.4383901059627533,
+      "learning_rate": 6.789606035205365e-06,
+      "loss": 0.4183,
+      "step": 810
+    },
+    {
+      "epoch": 0.2039738430583501,
+      "grad_norm": 0.4648225009441376,
+      "learning_rate": 6.797988264878458e-06,
+      "loss": 0.4105,
+      "step": 811
+    },
+    {
+      "epoch": 0.20422535211267606,
+      "grad_norm": 0.4313584566116333,
+      "learning_rate": 6.806370494551551e-06,
+      "loss": 0.4132,
+      "step": 812
+    },
+    {
+      "epoch": 0.204476861167002,
+      "grad_norm": 0.4867895543575287,
+      "learning_rate": 6.8147527242246446e-06,
+      "loss": 0.4601,
+      "step": 813
+    },
+    {
+      "epoch": 0.20472837022132798,
+      "grad_norm": 0.4108380675315857,
+      "learning_rate": 6.823134953897738e-06,
+      "loss": 0.4437,
+      "step": 814
+    },
+    {
+      "epoch": 0.20497987927565392,
+      "grad_norm": 0.46014007925987244,
+      "learning_rate": 6.831517183570831e-06,
+      "loss": 0.4335,
+      "step": 815
+    },
+    {
+      "epoch": 0.2052313883299799,
+      "grad_norm": 0.42462071776390076,
+      "learning_rate": 6.839899413243923e-06,
+      "loss": 0.4316,
+      "step": 816
+    },
+    {
+      "epoch": 0.20548289738430583,
+      "grad_norm": 0.4779759645462036,
+      "learning_rate": 6.848281642917016e-06,
+      "loss": 0.3952,
+      "step": 817
+    },
+    {
+      "epoch": 0.20573440643863178,
+      "grad_norm": 0.4665524363517761,
+      "learning_rate": 6.856663872590109e-06,
+      "loss": 0.4114,
+      "step": 818
+    },
+    {
+      "epoch": 0.20598591549295775,
+      "grad_norm": 0.5013046264648438,
+      "learning_rate": 6.865046102263203e-06,
+      "loss": 0.4086,
+      "step": 819
+    },
+    {
+      "epoch": 0.2062374245472837,
+      "grad_norm": 0.4999219477176666,
+      "learning_rate": 6.873428331936296e-06,
+      "loss": 0.433,
+      "step": 820
+    },
+    {
+      "epoch": 0.20648893360160966,
+      "grad_norm": 0.46619290113449097,
+      "learning_rate": 6.881810561609389e-06,
+      "loss": 0.4102,
+      "step": 821
+    },
+    {
+      "epoch": 0.2067404426559356,
+      "grad_norm": 0.4653915464878082,
+      "learning_rate": 6.890192791282482e-06,
+      "loss": 0.42,
+      "step": 822
+    },
+    {
+      "epoch": 0.20699195171026158,
+      "grad_norm": 0.4532758295536041,
+      "learning_rate": 6.898575020955574e-06,
+      "loss": 0.4216,
+      "step": 823
+    },
+    {
+      "epoch": 0.20724346076458752,
+      "grad_norm": 0.45343294739723206,
+      "learning_rate": 6.906957250628668e-06,
+      "loss": 0.4259,
+      "step": 824
+    },
+    {
+      "epoch": 0.2074949698189135,
+      "grad_norm": 0.4827394485473633,
+      "learning_rate": 6.91533948030176e-06,
+      "loss": 0.4252,
+      "step": 825
+    },
+    {
+      "epoch": 0.20774647887323944,
+      "grad_norm": 0.48318737745285034,
+      "learning_rate": 6.923721709974855e-06,
+      "loss": 0.4241,
+      "step": 826
+    },
+    {
+      "epoch": 0.20799798792756538,
+      "grad_norm": 0.4679957330226898,
+      "learning_rate": 6.932103939647947e-06,
+      "loss": 0.4154,
+      "step": 827
+    },
+    {
+      "epoch": 0.20824949698189135,
+      "grad_norm": 0.43381795287132263,
+      "learning_rate": 6.94048616932104e-06,
+      "loss": 0.4172,
+      "step": 828
+    },
+    {
+      "epoch": 0.2085010060362173,
+      "grad_norm": 0.45988166332244873,
+      "learning_rate": 6.948868398994133e-06,
+      "loss": 0.4147,
+      "step": 829
+    },
+    {
+      "epoch": 0.20875251509054327,
+      "grad_norm": 0.400395005941391,
+      "learning_rate": 6.957250628667226e-06,
+      "loss": 0.4308,
+      "step": 830
+    },
+    {
+      "epoch": 0.2090040241448692,
+      "grad_norm": 0.4453587532043457,
+      "learning_rate": 6.965632858340319e-06,
+      "loss": 0.4018,
+      "step": 831
+    },
+    {
+      "epoch": 0.20925553319919518,
+      "grad_norm": 0.4165091812610626,
+      "learning_rate": 6.974015088013411e-06,
+      "loss": 0.4022,
+      "step": 832
+    },
+    {
+      "epoch": 0.20950704225352113,
+      "grad_norm": 0.49082446098327637,
+      "learning_rate": 6.9823973176865055e-06,
+      "loss": 0.4378,
+      "step": 833
+    },
+    {
+      "epoch": 0.20975855130784707,
+      "grad_norm": 0.4576517641544342,
+      "learning_rate": 6.990779547359598e-06,
+      "loss": 0.4107,
+      "step": 834
+    },
+    {
+      "epoch": 0.21001006036217304,
+      "grad_norm": 0.43755435943603516,
+      "learning_rate": 6.999161777032692e-06,
+      "loss": 0.4226,
+      "step": 835
+    },
+    {
+      "epoch": 0.21026156941649898,
+      "grad_norm": 0.45865172147750854,
+      "learning_rate": 7.007544006705784e-06,
+      "loss": 0.4089,
+      "step": 836
+    },
+    {
+      "epoch": 0.21051307847082495,
+      "grad_norm": 0.5721135139465332,
+      "learning_rate": 7.015926236378877e-06,
+      "loss": 0.4173,
+      "step": 837
+    },
+    {
+      "epoch": 0.2107645875251509,
+      "grad_norm": 0.43587735295295715,
+      "learning_rate": 7.02430846605197e-06,
+      "loss": 0.4121,
+      "step": 838
+    },
+    {
+      "epoch": 0.21101609657947687,
+      "grad_norm": 0.5028086304664612,
+      "learning_rate": 7.032690695725063e-06,
+      "loss": 0.4339,
+      "step": 839
+    },
+    {
+      "epoch": 0.2112676056338028,
+      "grad_norm": 0.5480897426605225,
+      "learning_rate": 7.041072925398157e-06,
+      "loss": 0.422,
+      "step": 840
+    },
+    {
+      "epoch": 0.21151911468812878,
+      "grad_norm": 0.4621444642543793,
+      "learning_rate": 7.04945515507125e-06,
+      "loss": 0.396,
+      "step": 841
+    },
+    {
+      "epoch": 0.21177062374245473,
+      "grad_norm": 0.4964749813079834,
+      "learning_rate": 7.0578373847443426e-06,
+      "loss": 0.4597,
+      "step": 842
+    },
+    {
+      "epoch": 0.21202213279678067,
+      "grad_norm": 0.48463791608810425,
+      "learning_rate": 7.066219614417435e-06,
+      "loss": 0.421,
+      "step": 843
+    },
+    {
+      "epoch": 0.21227364185110664,
+      "grad_norm": 0.4578920900821686,
+      "learning_rate": 7.074601844090529e-06,
+      "loss": 0.4552,
+      "step": 844
+    },
+    {
+      "epoch": 0.21252515090543259,
+      "grad_norm": 0.5604377388954163,
+      "learning_rate": 7.082984073763621e-06,
+      "loss": 0.4275,
+      "step": 845
+    },
+    {
+      "epoch": 0.21277665995975856,
+      "grad_norm": 0.424990177154541,
+      "learning_rate": 7.091366303436716e-06,
+      "loss": 0.4237,
+      "step": 846
+    },
+    {
+      "epoch": 0.2130281690140845,
+      "grad_norm": 0.5051470398902893,
+      "learning_rate": 7.099748533109808e-06,
+      "loss": 0.4197,
+      "step": 847
+    },
+    {
+      "epoch": 0.21327967806841047,
+      "grad_norm": 0.5595173239707947,
+      "learning_rate": 7.108130762782901e-06,
+      "loss": 0.4045,
+      "step": 848
+    },
+    {
+      "epoch": 0.21353118712273642,
+      "grad_norm": 0.43428587913513184,
+      "learning_rate": 7.116512992455994e-06,
+      "loss": 0.416,
+      "step": 849
+    },
+    {
+      "epoch": 0.2137826961770624,
+      "grad_norm": 0.5479362607002258,
+      "learning_rate": 7.124895222129087e-06,
+      "loss": 0.4212,
+      "step": 850
+    },
+    {
+      "epoch": 0.21403420523138833,
+      "grad_norm": 0.4589390754699707,
+      "learning_rate": 7.13327745180218e-06,
+      "loss": 0.4336,
+      "step": 851
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.4126336872577667,
+      "learning_rate": 7.141659681475272e-06,
+      "loss": 0.4356,
+      "step": 852
+    },
+    {
+      "epoch": 0.21453722334004025,
+      "grad_norm": 0.5102934241294861,
+      "learning_rate": 7.1500419111483665e-06,
+      "loss": 0.4416,
+      "step": 853
+    },
+    {
+      "epoch": 0.2147887323943662,
+      "grad_norm": 0.4380166232585907,
+      "learning_rate": 7.158424140821459e-06,
+      "loss": 0.4087,
+      "step": 854
+    },
+    {
+      "epoch": 0.21504024144869216,
+      "grad_norm": 0.4341869652271271,
+      "learning_rate": 7.166806370494553e-06,
+      "loss": 0.4436,
+      "step": 855
+    },
+    {
+      "epoch": 0.2152917505030181,
+      "grad_norm": 0.42801982164382935,
+      "learning_rate": 7.175188600167645e-06,
+      "loss": 0.4048,
+      "step": 856
+    },
+    {
+      "epoch": 0.21554325955734407,
+      "grad_norm": 0.46920889616012573,
+      "learning_rate": 7.183570829840738e-06,
+      "loss": 0.4304,
+      "step": 857
+    },
+    {
+      "epoch": 0.21579476861167002,
+      "grad_norm": 0.4114355444908142,
+      "learning_rate": 7.1919530595138305e-06,
+      "loss": 0.4258,
+      "step": 858
+    },
+    {
+      "epoch": 0.216046277665996,
+      "grad_norm": 0.47151094675064087,
+      "learning_rate": 7.200335289186924e-06,
+      "loss": 0.3901,
+      "step": 859
+    },
+    {
+      "epoch": 0.21629778672032193,
+      "grad_norm": 0.4899015724658966,
+      "learning_rate": 7.2087175188600175e-06,
+      "loss": 0.4144,
+      "step": 860
+    },
+    {
+      "epoch": 0.21654929577464788,
+      "grad_norm": 0.4297754764556885,
+      "learning_rate": 7.217099748533111e-06,
+      "loss": 0.4041,
+      "step": 861
+    },
+    {
+      "epoch": 0.21680080482897385,
+      "grad_norm": 0.5144082307815552,
+      "learning_rate": 7.2254819782062036e-06,
+      "loss": 0.4448,
+      "step": 862
+    },
+    {
+      "epoch": 0.2170523138832998,
+      "grad_norm": 0.4766099750995636,
+      "learning_rate": 7.233864207879296e-06,
+      "loss": 0.3988,
+      "step": 863
+    },
+    {
+      "epoch": 0.21730382293762576,
+      "grad_norm": 0.46279922127723694,
+      "learning_rate": 7.24224643755239e-06,
+      "loss": 0.3987,
+      "step": 864
+    },
+    {
+      "epoch": 0.2175553319919517,
+      "grad_norm": 0.5055639743804932,
+      "learning_rate": 7.250628667225482e-06,
+      "loss": 0.4175,
+      "step": 865
+    },
+    {
+      "epoch": 0.21780684104627768,
+      "grad_norm": 0.5239527821540833,
+      "learning_rate": 7.259010896898575e-06,
+      "loss": 0.4355,
+      "step": 866
+    },
+    {
+      "epoch": 0.21805835010060362,
+      "grad_norm": 0.4184260964393616,
+      "learning_rate": 7.267393126571669e-06,
+      "loss": 0.4454,
+      "step": 867
+    },
+    {
+      "epoch": 0.21830985915492956,
+      "grad_norm": 0.5631728768348694,
+      "learning_rate": 7.275775356244762e-06,
+      "loss": 0.4466,
+      "step": 868
+    },
+    {
+      "epoch": 0.21856136820925554,
+      "grad_norm": 0.4764256775379181,
+      "learning_rate": 7.2841575859178545e-06,
+      "loss": 0.4278,
+      "step": 869
+    },
+    {
+      "epoch": 0.21881287726358148,
+      "grad_norm": 0.4316968619823456,
+      "learning_rate": 7.292539815590948e-06,
+      "loss": 0.4642,
+      "step": 870
+    },
+    {
+      "epoch": 0.21906438631790745,
+      "grad_norm": 0.4983559846878052,
+      "learning_rate": 7.300922045264041e-06,
+      "loss": 0.4296,
+      "step": 871
+    },
+    {
+      "epoch": 0.2193158953722334,
+      "grad_norm": 0.49155017733573914,
+      "learning_rate": 7.309304274937133e-06,
+      "loss": 0.452,
+      "step": 872
+    },
+    {
+      "epoch": 0.21956740442655936,
+      "grad_norm": 0.5096904635429382,
+      "learning_rate": 7.3176865046102275e-06,
+      "loss": 0.4112,
+      "step": 873
+    },
+    {
+      "epoch": 0.2198189134808853,
+      "grad_norm": 0.41947141289711,
+      "learning_rate": 7.32606873428332e-06,
+      "loss": 0.426,
+      "step": 874
+    },
+    {
+      "epoch": 0.22007042253521128,
+      "grad_norm": 0.49421125650405884,
+      "learning_rate": 7.334450963956414e-06,
+      "loss": 0.4014,
+      "step": 875
+    },
+    {
+      "epoch": 0.22032193158953722,
+      "grad_norm": 0.532922089099884,
+      "learning_rate": 7.342833193629506e-06,
+      "loss": 0.4217,
+      "step": 876
+    },
+    {
+      "epoch": 0.22057344064386317,
+      "grad_norm": 0.4467262923717499,
+      "learning_rate": 7.351215423302599e-06,
+      "loss": 0.4348,
+      "step": 877
+    },
+    {
+      "epoch": 0.22082494969818914,
+      "grad_norm": 0.4734607934951782,
+      "learning_rate": 7.3595976529756915e-06,
+      "loss": 0.432,
+      "step": 878
+    },
+    {
+      "epoch": 0.22107645875251508,
+      "grad_norm": 0.4447082579135895,
+      "learning_rate": 7.367979882648785e-06,
+      "loss": 0.4272,
+      "step": 879
+    },
+    {
+      "epoch": 0.22132796780684105,
+      "grad_norm": 0.4721175730228424,
+      "learning_rate": 7.3763621123218785e-06,
+      "loss": 0.4091,
+      "step": 880
+    },
+    {
+      "epoch": 0.221579476861167,
+      "grad_norm": 0.45364779233932495,
+      "learning_rate": 7.384744341994972e-06,
+      "loss": 0.422,
+      "step": 881
+    },
+    {
+      "epoch": 0.22183098591549297,
+      "grad_norm": 0.5278758406639099,
+      "learning_rate": 7.3931265716680646e-06,
+      "loss": 0.4253,
+      "step": 882
+    },
+    {
+      "epoch": 0.2220824949698189,
+      "grad_norm": 0.4425363540649414,
+      "learning_rate": 7.401508801341157e-06,
+      "loss": 0.4227,
+      "step": 883
+    },
+    {
+      "epoch": 0.22233400402414488,
+      "grad_norm": 0.4397623836994171,
+      "learning_rate": 7.40989103101425e-06,
+      "loss": 0.4274,
+      "step": 884
+    },
+    {
+      "epoch": 0.22258551307847083,
+      "grad_norm": 0.5451933741569519,
+      "learning_rate": 7.418273260687343e-06,
+      "loss": 0.4258,
+      "step": 885
+    },
+    {
+      "epoch": 0.22283702213279677,
+      "grad_norm": 0.5051050186157227,
+      "learning_rate": 7.426655490360436e-06,
+      "loss": 0.4037,
+      "step": 886
+    },
+    {
+      "epoch": 0.22308853118712274,
+      "grad_norm": 0.4557664096355438,
+      "learning_rate": 7.43503772003353e-06,
+      "loss": 0.4222,
+      "step": 887
+    },
+    {
+      "epoch": 0.22334004024144868,
+      "grad_norm": 0.5401166677474976,
+      "learning_rate": 7.443419949706623e-06,
+      "loss": 0.4399,
+      "step": 888
+    },
+    {
+      "epoch": 0.22359154929577466,
+      "grad_norm": 0.48601412773132324,
+      "learning_rate": 7.4518021793797155e-06,
+      "loss": 0.4292,
+      "step": 889
+    },
+    {
+      "epoch": 0.2238430583501006,
+      "grad_norm": 0.5143569707870483,
+      "learning_rate": 7.460184409052809e-06,
+      "loss": 0.4331,
+      "step": 890
+    },
+    {
+      "epoch": 0.22409456740442657,
+      "grad_norm": 0.4917342960834503,
+      "learning_rate": 7.4685666387259016e-06,
+      "loss": 0.4079,
+      "step": 891
+    },
+    {
+      "epoch": 0.2243460764587525,
+      "grad_norm": 0.5398588180541992,
+      "learning_rate": 7.476948868398994e-06,
+      "loss": 0.4244,
+      "step": 892
+    },
+    {
+      "epoch": 0.22459758551307846,
+      "grad_norm": 0.4740265905857086,
+      "learning_rate": 7.485331098072087e-06,
+      "loss": 0.4131,
+      "step": 893
+    },
+    {
+      "epoch": 0.22484909456740443,
+      "grad_norm": 0.5333375334739685,
+      "learning_rate": 7.493713327745181e-06,
+      "loss": 0.3979,
+      "step": 894
+    },
+    {
+      "epoch": 0.22510060362173037,
+      "grad_norm": 0.47554007172584534,
+      "learning_rate": 7.502095557418274e-06,
+      "loss": 0.4184,
+      "step": 895
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 0.5044271945953369,
+      "learning_rate": 7.510477787091367e-06,
+      "loss": 0.4374,
+      "step": 896
+    },
+    {
+      "epoch": 0.2256036217303823,
+      "grad_norm": 0.4042213261127472,
+      "learning_rate": 7.51886001676446e-06,
+      "loss": 0.4121,
+      "step": 897
+    },
+    {
+      "epoch": 0.22585513078470826,
+      "grad_norm": 0.47659948468208313,
+      "learning_rate": 7.5272422464375525e-06,
+      "loss": 0.411,
+      "step": 898
+    },
+    {
+      "epoch": 0.2261066398390342,
+      "grad_norm": 0.48039042949676514,
+      "learning_rate": 7.535624476110646e-06,
+      "loss": 0.4294,
+      "step": 899
+    },
+    {
+      "epoch": 0.22635814889336017,
+      "grad_norm": 0.468188613653183,
+      "learning_rate": 7.5440067057837394e-06,
+      "loss": 0.4076,
+      "step": 900
+    },
+    {
+      "epoch": 0.22660965794768612,
+      "grad_norm": 0.47702232003211975,
+      "learning_rate": 7.552388935456833e-06,
+      "loss": 0.423,
+      "step": 901
+    },
+    {
+      "epoch": 0.22686116700201206,
+      "grad_norm": 0.4733107089996338,
+      "learning_rate": 7.5607711651299255e-06,
+      "loss": 0.4356,
+      "step": 902
+    },
+    {
+      "epoch": 0.22711267605633803,
+      "grad_norm": 0.4408296048641205,
+      "learning_rate": 7.569153394803018e-06,
+      "loss": 0.4132,
+      "step": 903
+    },
+    {
+      "epoch": 0.22736418511066397,
+      "grad_norm": 0.4489647150039673,
+      "learning_rate": 7.577535624476111e-06,
+      "loss": 0.4259,
+      "step": 904
+    },
+    {
+      "epoch": 0.22761569416498995,
+      "grad_norm": 0.42070379853248596,
+      "learning_rate": 7.585917854149204e-06,
+      "loss": 0.4326,
+      "step": 905
+    },
+    {
+      "epoch": 0.2278672032193159,
+      "grad_norm": 0.43029969930648804,
+      "learning_rate": 7.594300083822297e-06,
+      "loss": 0.4053,
+      "step": 906
+    },
+    {
+      "epoch": 0.22811871227364186,
+      "grad_norm": 0.38450947403907776,
+      "learning_rate": 7.602682313495391e-06,
+      "loss": 0.4256,
+      "step": 907
+    },
+    {
+      "epoch": 0.2283702213279678,
+      "grad_norm": 0.45817065238952637,
+      "learning_rate": 7.611064543168484e-06,
+      "loss": 0.4427,
+      "step": 908
+    },
+    {
+      "epoch": 0.22862173038229378,
+      "grad_norm": 0.44824421405792236,
+      "learning_rate": 7.6194467728415765e-06,
+      "loss": 0.4002,
+      "step": 909
+    },
+    {
+      "epoch": 0.22887323943661972,
+      "grad_norm": 0.4120684862136841,
+      "learning_rate": 7.627829002514669e-06,
+      "loss": 0.4054,
+      "step": 910
+    },
+    {
+      "epoch": 0.22912474849094566,
+      "grad_norm": 0.4775068759918213,
+      "learning_rate": 7.636211232187762e-06,
+      "loss": 0.4149,
+      "step": 911
+    },
+    {
+      "epoch": 0.22937625754527163,
+      "grad_norm": 0.5041594505310059,
+      "learning_rate": 7.644593461860856e-06,
+      "loss": 0.4529,
+      "step": 912
+    },
+    {
+      "epoch": 0.22962776659959758,
+      "grad_norm": 0.4362272620201111,
+      "learning_rate": 7.652975691533949e-06,
+      "loss": 0.4144,
+      "step": 913
+    },
+    {
+      "epoch": 0.22987927565392355,
+      "grad_norm": 0.4660468101501465,
+      "learning_rate": 7.661357921207043e-06,
+      "loss": 0.4065,
+      "step": 914
+    },
+    {
+      "epoch": 0.2301307847082495,
+      "grad_norm": 0.4543038606643677,
+      "learning_rate": 7.669740150880136e-06,
+      "loss": 0.4356,
+      "step": 915
+    },
+    {
+      "epoch": 0.23038229376257546,
+      "grad_norm": 0.4837052524089813,
+      "learning_rate": 7.678122380553228e-06,
+      "loss": 0.4341,
+      "step": 916
+    },
+    {
+      "epoch": 0.2306338028169014,
+      "grad_norm": 0.47712114453315735,
+      "learning_rate": 7.686504610226321e-06,
+      "loss": 0.4127,
+      "step": 917
+    },
+    {
+      "epoch": 0.23088531187122738,
+      "grad_norm": 0.44308313727378845,
+      "learning_rate": 7.694886839899413e-06,
+      "loss": 0.4173,
+      "step": 918
+    },
+    {
+      "epoch": 0.23113682092555332,
+      "grad_norm": 0.4565160274505615,
+      "learning_rate": 7.703269069572506e-06,
+      "loss": 0.4009,
+      "step": 919
+    },
+    {
+      "epoch": 0.23138832997987926,
+      "grad_norm": 0.4511704742908478,
+      "learning_rate": 7.711651299245599e-06,
+      "loss": 0.4337,
+      "step": 920
+    },
+    {
+      "epoch": 0.23163983903420524,
+      "grad_norm": 0.46274664998054504,
+      "learning_rate": 7.720033528918693e-06,
+      "loss": 0.4184,
+      "step": 921
+    },
+    {
+      "epoch": 0.23189134808853118,
+      "grad_norm": 0.4714046120643616,
+      "learning_rate": 7.728415758591786e-06,
+      "loss": 0.4313,
+      "step": 922
+    },
+    {
+      "epoch": 0.23214285714285715,
+      "grad_norm": 0.46814224123954773,
+      "learning_rate": 7.73679798826488e-06,
+      "loss": 0.4318,
+      "step": 923
+    },
+    {
+      "epoch": 0.2323943661971831,
+      "grad_norm": 0.47005271911621094,
+      "learning_rate": 7.745180217937973e-06,
+      "loss": 0.4237,
+      "step": 924
+    },
+    {
+      "epoch": 0.23264587525150907,
+      "grad_norm": 0.48916304111480713,
+      "learning_rate": 7.753562447611065e-06,
+      "loss": 0.3893,
+      "step": 925
+    },
+    {
+      "epoch": 0.232897384305835,
+      "grad_norm": 0.4641698896884918,
+      "learning_rate": 7.761944677284158e-06,
+      "loss": 0.4322,
+      "step": 926
+    },
+    {
+      "epoch": 0.23314889336016095,
+      "grad_norm": 0.44672325253486633,
+      "learning_rate": 7.770326906957252e-06,
+      "loss": 0.4581,
+      "step": 927
+    },
+    {
+      "epoch": 0.23340040241448692,
+      "grad_norm": 0.43729367852211,
+      "learning_rate": 7.778709136630345e-06,
+      "loss": 0.4139,
+      "step": 928
+    },
+    {
+      "epoch": 0.23365191146881287,
+      "grad_norm": 0.45290911197662354,
+      "learning_rate": 7.787091366303437e-06,
+      "loss": 0.418,
+      "step": 929
+    },
+    {
+      "epoch": 0.23390342052313884,
+      "grad_norm": 0.42791497707366943,
+      "learning_rate": 7.79547359597653e-06,
+      "loss": 0.4188,
+      "step": 930
+    },
+    {
+      "epoch": 0.23415492957746478,
+      "grad_norm": 0.5161230564117432,
+      "learning_rate": 7.803855825649623e-06,
+      "loss": 0.4236,
+      "step": 931
+    },
+    {
+      "epoch": 0.23440643863179075,
+      "grad_norm": 0.4312325119972229,
+      "learning_rate": 7.812238055322715e-06,
+      "loss": 0.4313,
+      "step": 932
+    },
+    {
+      "epoch": 0.2346579476861167,
+      "grad_norm": 0.47641661763191223,
+      "learning_rate": 7.82062028499581e-06,
+      "loss": 0.4152,
+      "step": 933
+    },
+    {
+      "epoch": 0.23490945674044267,
+      "grad_norm": 0.4813452363014221,
+      "learning_rate": 7.829002514668902e-06,
+      "loss": 0.4111,
+      "step": 934
+    },
+    {
+      "epoch": 0.2351609657947686,
+      "grad_norm": 0.49906906485557556,
+      "learning_rate": 7.837384744341997e-06,
+      "loss": 0.3846,
+      "step": 935
+    },
+    {
+      "epoch": 0.23541247484909456,
+      "grad_norm": 0.48748451471328735,
+      "learning_rate": 7.84576697401509e-06,
+      "loss": 0.4329,
+      "step": 936
+    },
+    {
+      "epoch": 0.23566398390342053,
+      "grad_norm": 0.5034842491149902,
+      "learning_rate": 7.854149203688182e-06,
+      "loss": 0.4054,
+      "step": 937
+    },
+    {
+      "epoch": 0.23591549295774647,
+      "grad_norm": 0.45687928795814514,
+      "learning_rate": 7.862531433361274e-06,
+      "loss": 0.4466,
+      "step": 938
+    },
+    {
+      "epoch": 0.23616700201207244,
+      "grad_norm": 0.4433618485927582,
+      "learning_rate": 7.870913663034367e-06,
+      "loss": 0.3867,
+      "step": 939
+    },
+    {
+      "epoch": 0.23641851106639838,
+      "grad_norm": 0.4428368806838989,
+      "learning_rate": 7.87929589270746e-06,
+      "loss": 0.3938,
+      "step": 940
+    },
+    {
+      "epoch": 0.23667002012072436,
+      "grad_norm": 0.49069783091545105,
+      "learning_rate": 7.887678122380554e-06,
+      "loss": 0.3964,
+      "step": 941
+    },
+    {
+      "epoch": 0.2369215291750503,
+      "grad_norm": 0.45439690351486206,
+      "learning_rate": 7.896060352053647e-06,
+      "loss": 0.4115,
+      "step": 942
+    },
+    {
+      "epoch": 0.23717303822937627,
+      "grad_norm": 0.4886208772659302,
+      "learning_rate": 7.90444258172674e-06,
+      "loss": 0.4415,
+      "step": 943
+    },
+    {
+      "epoch": 0.23742454728370221,
+      "grad_norm": 0.47720471024513245,
+      "learning_rate": 7.912824811399834e-06,
+      "loss": 0.3896,
+      "step": 944
+    },
+    {
+      "epoch": 0.23767605633802816,
+      "grad_norm": 0.47643229365348816,
+      "learning_rate": 7.921207041072926e-06,
+      "loss": 0.4226,
+      "step": 945
+    },
+    {
+      "epoch": 0.23792756539235413,
+      "grad_norm": 0.4700299799442291,
+      "learning_rate": 7.929589270746019e-06,
+      "loss": 0.4165,
+      "step": 946
+    },
+    {
+      "epoch": 0.23817907444668007,
+      "grad_norm": 0.41711124777793884,
+      "learning_rate": 7.937971500419113e-06,
+      "loss": 0.4345,
+      "step": 947
+    },
+    {
+      "epoch": 0.23843058350100604,
+      "grad_norm": 0.45945024490356445,
+      "learning_rate": 7.946353730092206e-06,
+      "loss": 0.4379,
+      "step": 948
+    },
+    {
+      "epoch": 0.238682092555332,
+      "grad_norm": 0.4340158998966217,
+      "learning_rate": 7.954735959765298e-06,
+      "loss": 0.391,
+      "step": 949
+    },
+    {
+      "epoch": 0.23893360160965796,
+      "grad_norm": 0.4647141396999359,
+      "learning_rate": 7.963118189438391e-06,
+      "loss": 0.4351,
+      "step": 950
+    },
+    {
+      "epoch": 0.2391851106639839,
+      "grad_norm": 0.43337324261665344,
+      "learning_rate": 7.971500419111484e-06,
+      "loss": 0.4201,
+      "step": 951
+    },
+    {
+      "epoch": 0.23943661971830985,
+      "grad_norm": 0.45250189304351807,
+      "learning_rate": 7.979882648784576e-06,
+      "loss": 0.4328,
+      "step": 952
+    },
+    {
+      "epoch": 0.23968812877263582,
+      "grad_norm": 0.47799962759017944,
+      "learning_rate": 7.98826487845767e-06,
+      "loss": 0.4089,
+      "step": 953
+    },
+    {
+      "epoch": 0.23993963782696176,
+      "grad_norm": 0.4971179962158203,
+      "learning_rate": 7.996647108130763e-06,
+      "loss": 0.4312,
+      "step": 954
+    },
+    {
+      "epoch": 0.24019114688128773,
+      "grad_norm": 0.4583699703216553,
+      "learning_rate": 8.005029337803858e-06,
+      "loss": 0.4179,
+      "step": 955
+    },
+    {
+      "epoch": 0.24044265593561368,
+      "grad_norm": 0.3827430009841919,
+      "learning_rate": 8.01341156747695e-06,
+      "loss": 0.4026,
+      "step": 956
+    },
+    {
+      "epoch": 0.24069416498993965,
+      "grad_norm": 0.4374185800552368,
+      "learning_rate": 8.021793797150043e-06,
+      "loss": 0.4196,
+      "step": 957
+    },
+    {
+      "epoch": 0.2409456740442656,
+      "grad_norm": 0.4754588305950165,
+      "learning_rate": 8.030176026823135e-06,
+      "loss": 0.4214,
+      "step": 958
+    },
+    {
+      "epoch": 0.24119718309859156,
+      "grad_norm": 0.44250696897506714,
+      "learning_rate": 8.038558256496228e-06,
+      "loss": 0.3991,
+      "step": 959
+    },
+    {
+      "epoch": 0.2414486921529175,
+      "grad_norm": 0.5119554400444031,
+      "learning_rate": 8.04694048616932e-06,
+      "loss": 0.428,
+      "step": 960
+    },
+    {
+      "epoch": 0.24170020120724345,
+      "grad_norm": 0.5002796649932861,
+      "learning_rate": 8.055322715842415e-06,
+      "loss": 0.4342,
+      "step": 961
+    },
+    {
+      "epoch": 0.24195171026156942,
+      "grad_norm": 0.470553457736969,
+      "learning_rate": 8.063704945515508e-06,
+      "loss": 0.3983,
+      "step": 962
+    },
+    {
+      "epoch": 0.24220321931589536,
+      "grad_norm": 0.4792780876159668,
+      "learning_rate": 8.0720871751886e-06,
+      "loss": 0.4156,
+      "step": 963
+    },
+    {
+      "epoch": 0.24245472837022133,
+      "grad_norm": 0.44463014602661133,
+      "learning_rate": 8.080469404861695e-06,
+      "loss": 0.4143,
+      "step": 964
+    },
+    {
+      "epoch": 0.24270623742454728,
+      "grad_norm": 0.4440299868583679,
+      "learning_rate": 8.088851634534787e-06,
+      "loss": 0.3994,
+      "step": 965
+    },
+    {
+      "epoch": 0.24295774647887325,
+      "grad_norm": 0.45312440395355225,
+      "learning_rate": 8.09723386420788e-06,
+      "loss": 0.4236,
+      "step": 966
+    },
+    {
+      "epoch": 0.2432092555331992,
+      "grad_norm": 0.4270833134651184,
+      "learning_rate": 8.105616093880972e-06,
+      "loss": 0.4166,
+      "step": 967
+    },
+    {
+      "epoch": 0.24346076458752516,
+      "grad_norm": 0.46889373660087585,
+      "learning_rate": 8.113998323554067e-06,
+      "loss": 0.4467,
+      "step": 968
+    },
+    {
+      "epoch": 0.2437122736418511,
+      "grad_norm": 0.43110471963882446,
+      "learning_rate": 8.12238055322716e-06,
+      "loss": 0.4207,
+      "step": 969
+    },
+    {
+      "epoch": 0.24396378269617705,
+      "grad_norm": 0.500971257686615,
+      "learning_rate": 8.130762782900252e-06,
+      "loss": 0.4096,
+      "step": 970
+    },
+    {
+      "epoch": 0.24421529175050302,
+      "grad_norm": 0.43988287448883057,
+      "learning_rate": 8.139145012573345e-06,
+      "loss": 0.4236,
+      "step": 971
+    },
+    {
+      "epoch": 0.24446680080482897,
+      "grad_norm": 0.43207594752311707,
+      "learning_rate": 8.147527242246437e-06,
+      "loss": 0.4139,
+      "step": 972
+    },
+    {
+      "epoch": 0.24471830985915494,
+      "grad_norm": 0.6875976920127869,
+      "learning_rate": 8.155909471919532e-06,
+      "loss": 0.4108,
+      "step": 973
+    },
+    {
+      "epoch": 0.24496981891348088,
+      "grad_norm": 0.4743395447731018,
+      "learning_rate": 8.164291701592624e-06,
+      "loss": 0.423,
+      "step": 974
+    },
+    {
+      "epoch": 0.24522132796780685,
+      "grad_norm": 0.45225653052330017,
+      "learning_rate": 8.172673931265719e-06,
+      "loss": 0.4037,
+      "step": 975
+    },
+    {
+      "epoch": 0.2454728370221328,
+      "grad_norm": 0.45756688714027405,
+      "learning_rate": 8.181056160938811e-06,
+      "loss": 0.4419,
+      "step": 976
+    },
+    {
+      "epoch": 0.24572434607645877,
+      "grad_norm": 0.48021066188812256,
+      "learning_rate": 8.189438390611904e-06,
+      "loss": 0.4449,
+      "step": 977
+    },
+    {
+      "epoch": 0.2459758551307847,
+      "grad_norm": 0.44351431727409363,
+      "learning_rate": 8.197820620284996e-06,
+      "loss": 0.4226,
+      "step": 978
+    },
+    {
+      "epoch": 0.24622736418511065,
+      "grad_norm": 0.4529029428958893,
+      "learning_rate": 8.206202849958089e-06,
+      "loss": 0.4384,
+      "step": 979
+    },
+    {
+      "epoch": 0.24647887323943662,
+      "grad_norm": 0.5040020942687988,
+      "learning_rate": 8.214585079631182e-06,
+      "loss": 0.4464,
+      "step": 980
+    },
+    {
+      "epoch": 0.24673038229376257,
+      "grad_norm": 0.45902392268180847,
+      "learning_rate": 8.222967309304276e-06,
+      "loss": 0.3818,
+      "step": 981
+    },
+    {
+      "epoch": 0.24698189134808854,
+      "grad_norm": 0.4509660601615906,
+      "learning_rate": 8.231349538977369e-06,
+      "loss": 0.3941,
+      "step": 982
+    },
+    {
+      "epoch": 0.24723340040241448,
+      "grad_norm": 0.38173332810401917,
+      "learning_rate": 8.239731768650461e-06,
+      "loss": 0.4172,
+      "step": 983
+    },
+    {
+      "epoch": 0.24748490945674045,
+      "grad_norm": 0.45830610394477844,
+      "learning_rate": 8.248113998323556e-06,
+      "loss": 0.407,
+      "step": 984
+    },
+    {
+      "epoch": 0.2477364185110664,
+      "grad_norm": 0.49054500460624695,
+      "learning_rate": 8.256496227996648e-06,
+      "loss": 0.4212,
+      "step": 985
+    },
+    {
+      "epoch": 0.24798792756539234,
+      "grad_norm": 0.4405127763748169,
+      "learning_rate": 8.264878457669741e-06,
+      "loss": 0.4381,
+      "step": 986
+    },
+    {
+      "epoch": 0.2482394366197183,
+      "grad_norm": 0.4861133098602295,
+      "learning_rate": 8.273260687342833e-06,
+      "loss": 0.4282,
+      "step": 987
+    },
+    {
+      "epoch": 0.24849094567404426,
+      "grad_norm": 0.39729857444763184,
+      "learning_rate": 8.281642917015928e-06,
+      "loss": 0.4062,
+      "step": 988
+    },
+    {
+      "epoch": 0.24874245472837023,
+      "grad_norm": 0.48022037744522095,
+      "learning_rate": 8.29002514668902e-06,
+      "loss": 0.425,
+      "step": 989
+    },
+    {
+      "epoch": 0.24899396378269617,
+      "grad_norm": 0.46410584449768066,
+      "learning_rate": 8.298407376362113e-06,
+      "loss": 0.3949,
+      "step": 990
+    },
+    {
+      "epoch": 0.24924547283702214,
+      "grad_norm": 0.48868152499198914,
+      "learning_rate": 8.306789606035206e-06,
+      "loss": 0.4166,
+      "step": 991
+    },
+    {
+      "epoch": 0.24949698189134809,
+      "grad_norm": 0.46738943457603455,
+      "learning_rate": 8.315171835708298e-06,
+      "loss": 0.4221,
+      "step": 992
+    },
+    {
+      "epoch": 0.24974849094567406,
+      "grad_norm": 0.42502716183662415,
+      "learning_rate": 8.323554065381391e-06,
+      "loss": 0.4006,
+      "step": 993
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41025519371032715,
+      "learning_rate": 8.331936295054485e-06,
+      "loss": 0.4027,
+      "step": 994
+    },
+    {
+      "epoch": 0.25025150905432597,
+      "grad_norm": 0.46180248260498047,
+      "learning_rate": 8.340318524727578e-06,
+      "loss": 0.4375,
+      "step": 995
+    },
+    {
+      "epoch": 0.2505030181086519,
+      "grad_norm": 0.4456992447376251,
+      "learning_rate": 8.348700754400672e-06,
+      "loss": 0.4348,
+      "step": 996
+    },
+    {
+      "epoch": 0.25075452716297786,
+      "grad_norm": 0.4408442974090576,
+      "learning_rate": 8.357082984073765e-06,
+      "loss": 0.4062,
+      "step": 997
+    },
+    {
+      "epoch": 0.25100603621730383,
+      "grad_norm": 0.49061262607574463,
+      "learning_rate": 8.365465213746857e-06,
+      "loss": 0.4509,
+      "step": 998
+    },
+    {
+      "epoch": 0.2512575452716298,
+      "grad_norm": 0.5399837493896484,
+      "learning_rate": 8.37384744341995e-06,
+      "loss": 0.4313,
+      "step": 999
+    },
+    {
+      "epoch": 0.2515090543259557,
+      "grad_norm": 0.5015750527381897,
+      "learning_rate": 8.382229673093043e-06,
+      "loss": 0.4199,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2517605633802817,
+      "grad_norm": 0.550068199634552,
+      "learning_rate": 8.390611902766137e-06,
+      "loss": 0.4691,
+      "step": 1001
+    },
+    {
+      "epoch": 0.25201207243460766,
+      "grad_norm": 0.4467526376247406,
+      "learning_rate": 8.39899413243923e-06,
+      "loss": 0.4382,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2522635814889336,
+      "grad_norm": 0.5115616321563721,
+      "learning_rate": 8.407376362112322e-06,
+      "loss": 0.4377,
+      "step": 1003
+    },
+    {
+      "epoch": 0.25251509054325955,
+      "grad_norm": 0.4908027946949005,
+      "learning_rate": 8.415758591785415e-06,
+      "loss": 0.4417,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2527665995975855,
+      "grad_norm": 0.5211309790611267,
+      "learning_rate": 8.42414082145851e-06,
+      "loss": 0.4143,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2530181086519115,
+      "grad_norm": 0.5352211594581604,
+      "learning_rate": 8.432523051131602e-06,
+      "loss": 0.3923,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2532696177062374,
+      "grad_norm": 0.511669397354126,
+      "learning_rate": 8.440905280804694e-06,
+      "loss": 0.4113,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2535211267605634,
+      "grad_norm": 0.5156270861625671,
+      "learning_rate": 8.449287510477789e-06,
+      "loss": 0.4233,
+      "step": 1008
+    },
+    {
+      "epoch": 0.25377263581488935,
+      "grad_norm": 0.49817922711372375,
+      "learning_rate": 8.457669740150881e-06,
+      "loss": 0.3877,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2540241448692153,
+      "grad_norm": 0.4637084901332855,
+      "learning_rate": 8.466051969823974e-06,
+      "loss": 0.4413,
+      "step": 1010
+    },
+    {
+      "epoch": 0.25427565392354123,
+      "grad_norm": 0.5281439423561096,
+      "learning_rate": 8.474434199497067e-06,
+      "loss": 0.4063,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2545271629778672,
+      "grad_norm": 0.45103931427001953,
+      "learning_rate": 8.48281642917016e-06,
+      "loss": 0.4049,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2547786720321932,
+      "grad_norm": 0.47929713129997253,
+      "learning_rate": 8.491198658843252e-06,
+      "loss": 0.413,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2550301810865191,
+      "grad_norm": 0.4654606580734253,
+      "learning_rate": 8.499580888516346e-06,
+      "loss": 0.4246,
+      "step": 1014
+    },
+    {
+      "epoch": 0.25528169014084506,
+      "grad_norm": 0.4492524266242981,
+      "learning_rate": 8.507963118189439e-06,
+      "loss": 0.4328,
+      "step": 1015
+    },
+    {
+      "epoch": 0.25553319919517103,
+      "grad_norm": 0.5223946571350098,
+      "learning_rate": 8.516345347862533e-06,
+      "loss": 0.3964,
+      "step": 1016
+    },
+    {
+      "epoch": 0.255784708249497,
+      "grad_norm": 0.5574830770492554,
+      "learning_rate": 8.524727577535626e-06,
+      "loss": 0.4041,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2560362173038229,
+      "grad_norm": 0.5438414216041565,
+      "learning_rate": 8.533109807208718e-06,
+      "loss": 0.422,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2562877263581489,
+      "grad_norm": 0.49950259923934937,
+      "learning_rate": 8.541492036881811e-06,
+      "loss": 0.4054,
+      "step": 1019
+    },
+    {
+      "epoch": 0.25653923541247486,
+      "grad_norm": 0.5470491051673889,
+      "learning_rate": 8.549874266554904e-06,
+      "loss": 0.4237,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2567907444668008,
+      "grad_norm": 0.4732471704483032,
+      "learning_rate": 8.558256496227996e-06,
+      "loss": 0.4189,
+      "step": 1021
+    },
+    {
+      "epoch": 0.25704225352112675,
+      "grad_norm": 0.4920424818992615,
+      "learning_rate": 8.56663872590109e-06,
+      "loss": 0.4443,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2572937625754527,
+      "grad_norm": 0.4606372117996216,
+      "learning_rate": 8.575020955574183e-06,
+      "loss": 0.4187,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2575452716297787,
+      "grad_norm": 0.5176842212677002,
+      "learning_rate": 8.583403185247276e-06,
+      "loss": 0.4096,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2577967806841046,
+      "grad_norm": 0.4598933458328247,
+      "learning_rate": 8.59178541492037e-06,
+      "loss": 0.4443,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2580482897384306,
+      "grad_norm": 0.4918026030063629,
+      "learning_rate": 8.600167644593463e-06,
+      "loss": 0.4208,
+      "step": 1026
+    },
+    {
+      "epoch": 0.25829979879275655,
+      "grad_norm": 0.46681898832321167,
+      "learning_rate": 8.608549874266555e-06,
+      "loss": 0.4168,
+      "step": 1027
+    },
+    {
+      "epoch": 0.25855130784708247,
+      "grad_norm": 0.4654182493686676,
+      "learning_rate": 8.61693210393965e-06,
+      "loss": 0.4186,
+      "step": 1028
+    },
+    {
+      "epoch": 0.25880281690140844,
+      "grad_norm": 0.5116965770721436,
+      "learning_rate": 8.625314333612742e-06,
+      "loss": 0.43,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2590543259557344,
+      "grad_norm": 0.4793979823589325,
+      "learning_rate": 8.633696563285835e-06,
+      "loss": 0.4371,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2593058350100604,
+      "grad_norm": 0.49715420603752136,
+      "learning_rate": 8.642078792958928e-06,
+      "loss": 0.4197,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2595573440643863,
+      "grad_norm": 0.43482688069343567,
+      "learning_rate": 8.65046102263202e-06,
+      "loss": 0.4,
+      "step": 1032
+    },
+    {
+      "epoch": 0.25980885311871227,
+      "grad_norm": 0.5626974701881409,
+      "learning_rate": 8.658843252305113e-06,
+      "loss": 0.4498,
+      "step": 1033
+    },
+    {
+      "epoch": 0.26006036217303824,
+      "grad_norm": 0.4670769274234772,
+      "learning_rate": 8.667225481978207e-06,
+      "loss": 0.4038,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2603118712273642,
+      "grad_norm": 0.4788265824317932,
+      "learning_rate": 8.6756077116513e-06,
+      "loss": 0.4186,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2605633802816901,
+      "grad_norm": 0.5014175772666931,
+      "learning_rate": 8.683989941324394e-06,
+      "loss": 0.4175,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2608148893360161,
+      "grad_norm": 0.4646727740764618,
+      "learning_rate": 8.692372170997487e-06,
+      "loss": 0.4149,
+      "step": 1037
+    },
+    {
+      "epoch": 0.26106639839034207,
+      "grad_norm": 0.4624409079551697,
+      "learning_rate": 8.70075440067058e-06,
+      "loss": 0.376,
+      "step": 1038
+    },
+    {
+      "epoch": 0.261317907444668,
+      "grad_norm": 0.4742669463157654,
+      "learning_rate": 8.709136630343672e-06,
+      "loss": 0.4228,
+      "step": 1039
+    },
+    {
+      "epoch": 0.26156941649899396,
+      "grad_norm": 0.551210880279541,
+      "learning_rate": 8.717518860016765e-06,
+      "loss": 0.4292,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2618209255533199,
+      "grad_norm": 0.48132550716400146,
+      "learning_rate": 8.725901089689857e-06,
+      "loss": 0.4313,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2620724346076459,
+      "grad_norm": 0.5049678683280945,
+      "learning_rate": 8.734283319362952e-06,
+      "loss": 0.4193,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2623239436619718,
+      "grad_norm": 0.5476779341697693,
+      "learning_rate": 8.742665549036044e-06,
+      "loss": 0.4106,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2625754527162978,
+      "grad_norm": 0.4799102246761322,
+      "learning_rate": 8.751047778709137e-06,
+      "loss": 0.4198,
+      "step": 1044
+    },
+    {
+      "epoch": 0.26282696177062376,
+      "grad_norm": 0.46002301573753357,
+      "learning_rate": 8.75943000838223e-06,
+      "loss": 0.4007,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2630784708249497,
+      "grad_norm": 0.5139147043228149,
+      "learning_rate": 8.767812238055324e-06,
+      "loss": 0.399,
+      "step": 1046
+    },
+    {
+      "epoch": 0.26332997987927564,
+      "grad_norm": 0.4995320439338684,
+      "learning_rate": 8.776194467728416e-06,
+      "loss": 0.4379,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2635814889336016,
+      "grad_norm": 0.4521893560886383,
+      "learning_rate": 8.784576697401509e-06,
+      "loss": 0.4338,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2638329979879276,
+      "grad_norm": 0.4362622797489166,
+      "learning_rate": 8.792958927074603e-06,
+      "loss": 0.4217,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2640845070422535,
+      "grad_norm": 0.4953138828277588,
+      "learning_rate": 8.801341156747696e-06,
+      "loss": 0.4167,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2643360160965795,
+      "grad_norm": 0.4723548889160156,
+      "learning_rate": 8.809723386420789e-06,
+      "loss": 0.4537,
+      "step": 1051
+    },
+    {
+      "epoch": 0.26458752515090544,
+      "grad_norm": 0.5056288242340088,
+      "learning_rate": 8.818105616093881e-06,
+      "loss": 0.4412,
+      "step": 1052
+    },
+    {
+      "epoch": 0.26483903420523136,
+      "grad_norm": 0.5119395852088928,
+      "learning_rate": 8.826487845766974e-06,
+      "loss": 0.4184,
+      "step": 1053
+    },
+    {
+      "epoch": 0.26509054325955733,
+      "grad_norm": 0.45162543654441833,
+      "learning_rate": 8.834870075440067e-06,
+      "loss": 0.4135,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2653420523138833,
+      "grad_norm": 0.5046412348747253,
+      "learning_rate": 8.84325230511316e-06,
+      "loss": 0.3995,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2655935613682093,
+      "grad_norm": 0.42225155234336853,
+      "learning_rate": 8.851634534786253e-06,
+      "loss": 0.4316,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2658450704225352,
+      "grad_norm": 0.5072396397590637,
+      "learning_rate": 8.860016764459348e-06,
+      "loss": 0.4268,
+      "step": 1057
+    },
+    {
+      "epoch": 0.26609657947686116,
+      "grad_norm": 0.4895479083061218,
+      "learning_rate": 8.86839899413244e-06,
+      "loss": 0.4227,
+      "step": 1058
+    },
+    {
+      "epoch": 0.26634808853118713,
+      "grad_norm": 0.42159146070480347,
+      "learning_rate": 8.876781223805533e-06,
+      "loss": 0.4053,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2665995975855131,
+      "grad_norm": 0.4871593713760376,
+      "learning_rate": 8.885163453478626e-06,
+      "loss": 0.4027,
+      "step": 1060
+    },
+    {
+      "epoch": 0.266851106639839,
+      "grad_norm": 0.4512350559234619,
+      "learning_rate": 8.893545683151718e-06,
+      "loss": 0.4113,
+      "step": 1061
+    },
+    {
+      "epoch": 0.267102615694165,
+      "grad_norm": 0.4652332365512848,
+      "learning_rate": 8.901927912824813e-06,
+      "loss": 0.4308,
+      "step": 1062
+    },
+    {
+      "epoch": 0.26735412474849096,
+      "grad_norm": 0.4900280833244324,
+      "learning_rate": 8.910310142497905e-06,
+      "loss": 0.399,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2676056338028169,
+      "grad_norm": 0.42869365215301514,
+      "learning_rate": 8.918692372170998e-06,
+      "loss": 0.4127,
+      "step": 1064
+    },
+    {
+      "epoch": 0.26785714285714285,
+      "grad_norm": 0.4814034700393677,
+      "learning_rate": 8.92707460184409e-06,
+      "loss": 0.4181,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2681086519114688,
+      "grad_norm": 0.47565895318984985,
+      "learning_rate": 8.935456831517185e-06,
+      "loss": 0.4279,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2683601609657948,
+      "grad_norm": 0.4539470374584198,
+      "learning_rate": 8.943839061190277e-06,
+      "loss": 0.4044,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2686116700201207,
+      "grad_norm": 0.44740232825279236,
+      "learning_rate": 8.95222129086337e-06,
+      "loss": 0.4206,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2688631790744467,
+      "grad_norm": 0.46885591745376587,
+      "learning_rate": 8.960603520536464e-06,
+      "loss": 0.3829,
+      "step": 1069
+    },
+    {
+      "epoch": 0.26911468812877265,
+      "grad_norm": 0.5142282247543335,
+      "learning_rate": 8.968985750209557e-06,
+      "loss": 0.4,
+      "step": 1070
+    },
+    {
+      "epoch": 0.26936619718309857,
+      "grad_norm": 0.406148225069046,
+      "learning_rate": 8.97736797988265e-06,
+      "loss": 0.3931,
+      "step": 1071
+    },
+    {
+      "epoch": 0.26961770623742454,
+      "grad_norm": 0.449266254901886,
+      "learning_rate": 8.985750209555742e-06,
+      "loss": 0.3975,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2698692152917505,
+      "grad_norm": 0.530730664730072,
+      "learning_rate": 8.994132439228835e-06,
+      "loss": 0.4193,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2701207243460765,
+      "grad_norm": 0.495766282081604,
+      "learning_rate": 9.002514668901928e-06,
+      "loss": 0.4453,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2703722334004024,
+      "grad_norm": 0.43788525462150574,
+      "learning_rate": 9.010896898575022e-06,
+      "loss": 0.4432,
+      "step": 1075
+    },
+    {
+      "epoch": 0.27062374245472837,
+      "grad_norm": 0.5969604253768921,
+      "learning_rate": 9.019279128248114e-06,
+      "loss": 0.4155,
+      "step": 1076
+    },
+    {
+      "epoch": 0.27087525150905434,
+      "grad_norm": 0.5051907896995544,
+      "learning_rate": 9.027661357921209e-06,
+      "loss": 0.4363,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2711267605633803,
+      "grad_norm": 0.48711901903152466,
+      "learning_rate": 9.036043587594301e-06,
+      "loss": 0.4054,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2713782696177062,
+      "grad_norm": 0.745858907699585,
+      "learning_rate": 9.044425817267394e-06,
+      "loss": 0.436,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2716297786720322,
+      "grad_norm": 0.558805525302887,
+      "learning_rate": 9.052808046940487e-06,
+      "loss": 0.3979,
+      "step": 1080
+    },
+    {
+      "epoch": 0.27188128772635817,
+      "grad_norm": 0.5273854732513428,
+      "learning_rate": 9.06119027661358e-06,
+      "loss": 0.4198,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2721327967806841,
+      "grad_norm": 0.5657176375389099,
+      "learning_rate": 9.069572506286674e-06,
+      "loss": 0.4124,
+      "step": 1082
+    },
+    {
+      "epoch": 0.27238430583501005,
+      "grad_norm": 0.46423089504241943,
+      "learning_rate": 9.077954735959766e-06,
+      "loss": 0.426,
+      "step": 1083
+    },
+    {
+      "epoch": 0.272635814889336,
+      "grad_norm": 0.5339906811714172,
+      "learning_rate": 9.086336965632859e-06,
+      "loss": 0.4062,
+      "step": 1084
+    },
+    {
+      "epoch": 0.272887323943662,
+      "grad_norm": 0.46117010712623596,
+      "learning_rate": 9.094719195305951e-06,
+      "loss": 0.4059,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2731388329979879,
+      "grad_norm": 0.4512513279914856,
+      "learning_rate": 9.103101424979046e-06,
+      "loss": 0.4119,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2733903420523139,
+      "grad_norm": 0.5174707770347595,
+      "learning_rate": 9.111483654652138e-06,
+      "loss": 0.436,
+      "step": 1087
+    },
+    {
+      "epoch": 0.27364185110663986,
+      "grad_norm": 0.4645453989505768,
+      "learning_rate": 9.119865884325231e-06,
+      "loss": 0.4332,
+      "step": 1088
+    },
+    {
+      "epoch": 0.27389336016096577,
+      "grad_norm": 0.5541778206825256,
+      "learning_rate": 9.128248113998325e-06,
+      "loss": 0.4114,
+      "step": 1089
+    },
+    {
+      "epoch": 0.27414486921529174,
+      "grad_norm": 0.47235891222953796,
+      "learning_rate": 9.136630343671418e-06,
+      "loss": 0.4452,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2743963782696177,
+      "grad_norm": 0.4705214500427246,
+      "learning_rate": 9.14501257334451e-06,
+      "loss": 0.4028,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2746478873239437,
+      "grad_norm": 0.4457109272480011,
+      "learning_rate": 9.153394803017603e-06,
+      "loss": 0.3873,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2748993963782696,
+      "grad_norm": 0.4580100178718567,
+      "learning_rate": 9.161777032690696e-06,
+      "loss": 0.4188,
+      "step": 1093
+    },
+    {
+      "epoch": 0.27515090543259557,
+      "grad_norm": 0.50628662109375,
+      "learning_rate": 9.170159262363788e-06,
+      "loss": 0.4683,
+      "step": 1094
+    },
+    {
+      "epoch": 0.27540241448692154,
+      "grad_norm": 0.5623934864997864,
+      "learning_rate": 9.178541492036883e-06,
+      "loss": 0.4115,
+      "step": 1095
+    },
+    {
+      "epoch": 0.27565392354124746,
+      "grad_norm": 0.45169952511787415,
+      "learning_rate": 9.186923721709975e-06,
+      "loss": 0.4382,
+      "step": 1096
+    },
+    {
+      "epoch": 0.27590543259557343,
+      "grad_norm": 0.4219382107257843,
+      "learning_rate": 9.19530595138307e-06,
+      "loss": 0.3918,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2761569416498994,
+      "grad_norm": 0.47983744740486145,
+      "learning_rate": 9.203688181056162e-06,
+      "loss": 0.4257,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2764084507042254,
+      "grad_norm": 0.462584525346756,
+      "learning_rate": 9.212070410729255e-06,
+      "loss": 0.4289,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2766599597585513,
+      "grad_norm": 0.4568127393722534,
+      "learning_rate": 9.220452640402348e-06,
+      "loss": 0.4181,
+      "step": 1100
+    },
+    {
+      "epoch": 0.27691146881287726,
+      "grad_norm": 0.42234259843826294,
+      "learning_rate": 9.22883487007544e-06,
+      "loss": 0.4076,
+      "step": 1101
+    },
+    {
+      "epoch": 0.27716297786720323,
+      "grad_norm": 0.5175244212150574,
+      "learning_rate": 9.237217099748533e-06,
+      "loss": 0.4207,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2774144869215292,
+      "grad_norm": 0.4390292763710022,
+      "learning_rate": 9.245599329421627e-06,
+      "loss": 0.4413,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2776659959758551,
+      "grad_norm": 0.5344337821006775,
+      "learning_rate": 9.25398155909472e-06,
+      "loss": 0.4002,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2779175050301811,
+      "grad_norm": 0.5546042919158936,
+      "learning_rate": 9.262363788767812e-06,
+      "loss": 0.3978,
+      "step": 1105
+    },
+    {
+      "epoch": 0.27816901408450706,
+      "grad_norm": 0.45198217034339905,
+      "learning_rate": 9.270746018440905e-06,
+      "loss": 0.4228,
+      "step": 1106
+    },
+    {
+      "epoch": 0.278420523138833,
+      "grad_norm": 0.5244150757789612,
+      "learning_rate": 9.279128248114e-06,
+      "loss": 0.4018,
+      "step": 1107
+    },
+    {
+      "epoch": 0.27867203219315895,
+      "grad_norm": 0.44966578483581543,
+      "learning_rate": 9.287510477787092e-06,
+      "loss": 0.388,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2789235412474849,
+      "grad_norm": 0.5236296653747559,
+      "learning_rate": 9.295892707460186e-06,
+      "loss": 0.4182,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2791750503018109,
+      "grad_norm": 0.46590518951416016,
+      "learning_rate": 9.304274937133279e-06,
+      "loss": 0.4121,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2794265593561368,
+      "grad_norm": 0.4703758955001831,
+      "learning_rate": 9.312657166806372e-06,
+      "loss": 0.4208,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2796780684104628,
+      "grad_norm": 0.431341290473938,
+      "learning_rate": 9.321039396479464e-06,
+      "loss": 0.4101,
+      "step": 1112
+    },
+    {
+      "epoch": 0.27992957746478875,
+      "grad_norm": 0.5059375762939453,
+      "learning_rate": 9.329421626152557e-06,
+      "loss": 0.4478,
+      "step": 1113
+    },
+    {
+      "epoch": 0.28018108651911466,
+      "grad_norm": 0.4744166433811188,
+      "learning_rate": 9.33780385582565e-06,
+      "loss": 0.4076,
+      "step": 1114
+    },
+    {
+      "epoch": 0.28043259557344064,
+      "grad_norm": 0.4455539584159851,
+      "learning_rate": 9.346186085498742e-06,
+      "loss": 0.4235,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2806841046277666,
+      "grad_norm": 0.5029807090759277,
+      "learning_rate": 9.354568315171836e-06,
+      "loss": 0.4215,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2809356136820926,
+      "grad_norm": 0.4714094400405884,
+      "learning_rate": 9.362950544844929e-06,
+      "loss": 0.3958,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2811871227364185,
+      "grad_norm": 0.4298758804798126,
+      "learning_rate": 9.371332774518023e-06,
+      "loss": 0.4283,
+      "step": 1118
+    },
+    {
+      "epoch": 0.28143863179074446,
+      "grad_norm": 0.4799964427947998,
+      "learning_rate": 9.379715004191116e-06,
+      "loss": 0.4207,
+      "step": 1119
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 0.5128260850906372,
+      "learning_rate": 9.388097233864209e-06,
+      "loss": 0.4336,
+      "step": 1120
+    },
+    {
+      "epoch": 0.28194164989939635,
+      "grad_norm": 0.44608020782470703,
+      "learning_rate": 9.396479463537301e-06,
+      "loss": 0.4154,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2821931589537223,
+      "grad_norm": 0.4927178621292114,
+      "learning_rate": 9.404861693210394e-06,
+      "loss": 0.4369,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2824446680080483,
+      "grad_norm": 0.42505180835723877,
+      "learning_rate": 9.413243922883488e-06,
+      "loss": 0.4089,
+      "step": 1123
+    },
+    {
+      "epoch": 0.28269617706237427,
+      "grad_norm": 0.4519820213317871,
+      "learning_rate": 9.42162615255658e-06,
+      "loss": 0.4242,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2829476861167002,
+      "grad_norm": 0.4712681770324707,
+      "learning_rate": 9.430008382229673e-06,
+      "loss": 0.4362,
+      "step": 1125
+    },
+    {
+      "epoch": 0.28319919517102615,
+      "grad_norm": 0.45173585414886475,
+      "learning_rate": 9.438390611902766e-06,
+      "loss": 0.4332,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2834507042253521,
+      "grad_norm": 0.4184447228908539,
+      "learning_rate": 9.44677284157586e-06,
+      "loss": 0.4312,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2837022132796781,
+      "grad_norm": 0.4533083736896515,
+      "learning_rate": 9.455155071248953e-06,
+      "loss": 0.4127,
+      "step": 1128
+    },
+    {
+      "epoch": 0.283953722334004,
+      "grad_norm": 0.4317009449005127,
+      "learning_rate": 9.463537300922047e-06,
+      "loss": 0.4158,
+      "step": 1129
+    },
+    {
+      "epoch": 0.28420523138833,
+      "grad_norm": 0.4305717945098877,
+      "learning_rate": 9.47191953059514e-06,
+      "loss": 0.4266,
+      "step": 1130
+    },
+    {
+      "epoch": 0.28445674044265595,
+      "grad_norm": 0.4272315204143524,
+      "learning_rate": 9.480301760268233e-06,
+      "loss": 0.4434,
+      "step": 1131
+    },
+    {
+      "epoch": 0.28470824949698187,
+      "grad_norm": 0.4827929735183716,
+      "learning_rate": 9.488683989941325e-06,
+      "loss": 0.4404,
+      "step": 1132
+    },
+    {
+      "epoch": 0.28495975855130784,
+      "grad_norm": 0.4054642617702484,
+      "learning_rate": 9.497066219614418e-06,
+      "loss": 0.41,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2852112676056338,
+      "grad_norm": 0.4518846869468689,
+      "learning_rate": 9.50544844928751e-06,
+      "loss": 0.4361,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2854627766599598,
+      "grad_norm": 0.4325503706932068,
+      "learning_rate": 9.513830678960603e-06,
+      "loss": 0.399,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.4354625642299652,
+      "learning_rate": 9.522212908633697e-06,
+      "loss": 0.4186,
+      "step": 1136
+    },
+    {
+      "epoch": 0.28596579476861167,
+      "grad_norm": 0.43451640009880066,
+      "learning_rate": 9.53059513830679e-06,
+      "loss": 0.4315,
+      "step": 1137
+    },
+    {
+      "epoch": 0.28621730382293764,
+      "grad_norm": 0.4460737109184265,
+      "learning_rate": 9.538977367979884e-06,
+      "loss": 0.3852,
+      "step": 1138
+    },
+    {
+      "epoch": 0.28646881287726356,
+      "grad_norm": 0.44246113300323486,
+      "learning_rate": 9.547359597652977e-06,
+      "loss": 0.4176,
+      "step": 1139
+    },
+    {
+      "epoch": 0.28672032193158953,
+      "grad_norm": 0.4565284550189972,
+      "learning_rate": 9.55574182732607e-06,
+      "loss": 0.4202,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2869718309859155,
+      "grad_norm": 0.46833330392837524,
+      "learning_rate": 9.564124056999162e-06,
+      "loss": 0.4173,
+      "step": 1141
+    },
+    {
+      "epoch": 0.28722334004024147,
+      "grad_norm": 0.5017228722572327,
+      "learning_rate": 9.572506286672255e-06,
+      "loss": 0.3843,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2874748490945674,
+      "grad_norm": 0.5037913918495178,
+      "learning_rate": 9.58088851634535e-06,
+      "loss": 0.4185,
+      "step": 1143
+    },
+    {
+      "epoch": 0.28772635814889336,
+      "grad_norm": 0.5422639846801758,
+      "learning_rate": 9.589270746018442e-06,
+      "loss": 0.4084,
+      "step": 1144
+    },
+    {
+      "epoch": 0.28797786720321933,
+      "grad_norm": 0.4788282513618469,
+      "learning_rate": 9.597652975691534e-06,
+      "loss": 0.3975,
+      "step": 1145
+    },
+    {
+      "epoch": 0.28822937625754524,
+      "grad_norm": 0.5363534092903137,
+      "learning_rate": 9.606035205364627e-06,
+      "loss": 0.4217,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2884808853118712,
+      "grad_norm": 0.44976770877838135,
+      "learning_rate": 9.614417435037721e-06,
+      "loss": 0.3822,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2887323943661972,
+      "grad_norm": 0.49459314346313477,
+      "learning_rate": 9.622799664710814e-06,
+      "loss": 0.4358,
+      "step": 1148
+    },
+    {
+      "epoch": 0.28898390342052316,
+      "grad_norm": 0.45928260684013367,
+      "learning_rate": 9.631181894383907e-06,
+      "loss": 0.4268,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2892354124748491,
+      "grad_norm": 0.43904656171798706,
+      "learning_rate": 9.639564124057001e-06,
+      "loss": 0.4202,
+      "step": 1150
+    },
+    {
+      "epoch": 0.28948692152917505,
+      "grad_norm": 0.45102381706237793,
+      "learning_rate": 9.647946353730094e-06,
+      "loss": 0.4087,
+      "step": 1151
+    },
+    {
+      "epoch": 0.289738430583501,
+      "grad_norm": 0.4579281210899353,
+      "learning_rate": 9.656328583403186e-06,
+      "loss": 0.4153,
+      "step": 1152
+    },
+    {
+      "epoch": 0.289989939637827,
+      "grad_norm": 0.4361988604068756,
+      "learning_rate": 9.664710813076279e-06,
+      "loss": 0.4393,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2902414486921529,
+      "grad_norm": 0.43042805790901184,
+      "learning_rate": 9.673093042749371e-06,
+      "loss": 0.4031,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2904929577464789,
+      "grad_norm": 0.5227105021476746,
+      "learning_rate": 9.681475272422464e-06,
+      "loss": 0.433,
+      "step": 1155
+    },
+    {
+      "epoch": 0.29074446680080485,
+      "grad_norm": 0.48663586378097534,
+      "learning_rate": 9.689857502095558e-06,
+      "loss": 0.3878,
+      "step": 1156
+    },
+    {
+      "epoch": 0.29099597585513076,
+      "grad_norm": 0.5375685691833496,
+      "learning_rate": 9.698239731768651e-06,
+      "loss": 0.4254,
+      "step": 1157
+    },
+    {
+      "epoch": 0.29124748490945673,
+      "grad_norm": 0.5156546235084534,
+      "learning_rate": 9.706621961441745e-06,
+      "loss": 0.414,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2914989939637827,
+      "grad_norm": 0.4819703698158264,
+      "learning_rate": 9.715004191114838e-06,
+      "loss": 0.4194,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2917505030181087,
+      "grad_norm": 0.4525262117385864,
+      "learning_rate": 9.72338642078793e-06,
+      "loss": 0.421,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2920020120724346,
+      "grad_norm": 0.5090615153312683,
+      "learning_rate": 9.731768650461023e-06,
+      "loss": 0.4238,
+      "step": 1161
+    },
+    {
+      "epoch": 0.29225352112676056,
+      "grad_norm": 0.42741072177886963,
+      "learning_rate": 9.740150880134116e-06,
+      "loss": 0.3865,
+      "step": 1162
+    },
+    {
+      "epoch": 0.29250503018108653,
+      "grad_norm": 0.5380824208259583,
+      "learning_rate": 9.74853310980721e-06,
+      "loss": 0.4206,
+      "step": 1163
+    },
+    {
+      "epoch": 0.29275653923541245,
+      "grad_norm": 0.44021573662757874,
+      "learning_rate": 9.756915339480303e-06,
+      "loss": 0.4082,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2930080482897384,
+      "grad_norm": 0.4981013834476471,
+      "learning_rate": 9.765297569153395e-06,
+      "loss": 0.4137,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2932595573440644,
+      "grad_norm": 0.5061078071594238,
+      "learning_rate": 9.773679798826488e-06,
+      "loss": 0.404,
+      "step": 1166
+    },
+    {
+      "epoch": 0.29351106639839036,
+      "grad_norm": 0.4524628818035126,
+      "learning_rate": 9.78206202849958e-06,
+      "loss": 0.4101,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2937625754527163,
+      "grad_norm": 0.5405965447425842,
+      "learning_rate": 9.790444258172675e-06,
+      "loss": 0.4407,
+      "step": 1168
+    },
+    {
+      "epoch": 0.29401408450704225,
+      "grad_norm": 0.43946367502212524,
+      "learning_rate": 9.798826487845768e-06,
+      "loss": 0.4234,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2942655935613682,
+      "grad_norm": 0.52604079246521,
+      "learning_rate": 9.807208717518862e-06,
+      "loss": 0.4164,
+      "step": 1170
+    },
+    {
+      "epoch": 0.29451710261569414,
+      "grad_norm": 0.46760255098342896,
+      "learning_rate": 9.815590947191955e-06,
+      "loss": 0.4186,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2947686116700201,
+      "grad_norm": 0.5770349502563477,
+      "learning_rate": 9.823973176865047e-06,
+      "loss": 0.3987,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2950201207243461,
+      "grad_norm": 0.5591742992401123,
+      "learning_rate": 9.83235540653814e-06,
+      "loss": 0.4338,
+      "step": 1173
+    },
+    {
+      "epoch": 0.29527162977867205,
+      "grad_norm": 0.5872971415519714,
+      "learning_rate": 9.840737636211232e-06,
+      "loss": 0.4077,
+      "step": 1174
+    },
+    {
+      "epoch": 0.29552313883299797,
+      "grad_norm": 0.5050575733184814,
+      "learning_rate": 9.849119865884325e-06,
+      "loss": 0.4163,
+      "step": 1175
+    },
+    {
+      "epoch": 0.29577464788732394,
+      "grad_norm": 0.5279958844184875,
+      "learning_rate": 9.857502095557418e-06,
+      "loss": 0.4493,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2960261569416499,
+      "grad_norm": 0.4451628625392914,
+      "learning_rate": 9.865884325230512e-06,
+      "loss": 0.4077,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2962776659959759,
+      "grad_norm": 0.42817094922065735,
+      "learning_rate": 9.874266554903605e-06,
+      "loss": 0.3958,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2965291750503018,
+      "grad_norm": 0.5152884125709534,
+      "learning_rate": 9.882648784576699e-06,
+      "loss": 0.3644,
+      "step": 1179
+    },
+    {
+      "epoch": 0.29678068410462777,
+      "grad_norm": 0.5255314707756042,
+      "learning_rate": 9.891031014249792e-06,
+      "loss": 0.4299,
+      "step": 1180
+    },
+    {
+      "epoch": 0.29703219315895374,
+      "grad_norm": 0.45311322808265686,
+      "learning_rate": 9.899413243922884e-06,
+      "loss": 0.4234,
+      "step": 1181
+    },
+    {
+      "epoch": 0.29728370221327965,
+      "grad_norm": 0.5598365068435669,
+      "learning_rate": 9.907795473595977e-06,
+      "loss": 0.4224,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2975352112676056,
+      "grad_norm": 0.45655763149261475,
+      "learning_rate": 9.916177703269071e-06,
+      "loss": 0.4283,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2977867203219316,
+      "grad_norm": 0.5361334085464478,
+      "learning_rate": 9.924559932942164e-06,
+      "loss": 0.4216,
+      "step": 1184
+    },
+    {
+      "epoch": 0.29803822937625757,
+      "grad_norm": 0.45937347412109375,
+      "learning_rate": 9.932942162615256e-06,
+      "loss": 0.4372,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2982897384305835,
+      "grad_norm": 0.5063474178314209,
+      "learning_rate": 9.941324392288349e-06,
+      "loss": 0.434,
+      "step": 1186
+    },
+    {
+      "epoch": 0.29854124748490946,
+      "grad_norm": 0.5321951508522034,
+      "learning_rate": 9.949706621961442e-06,
+      "loss": 0.4464,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2987927565392354,
+      "grad_norm": 0.4343951344490051,
+      "learning_rate": 9.958088851634536e-06,
+      "loss": 0.4008,
+      "step": 1188
+    },
+    {
+      "epoch": 0.29904426559356134,
+      "grad_norm": 0.48022282123565674,
+      "learning_rate": 9.966471081307629e-06,
+      "loss": 0.4137,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2992957746478873,
+      "grad_norm": 0.4919980466365814,
+      "learning_rate": 9.974853310980723e-06,
+      "loss": 0.4176,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2995472837022133,
+      "grad_norm": 0.4395059049129486,
+      "learning_rate": 9.983235540653816e-06,
+      "loss": 0.3917,
+      "step": 1191
+    },
+    {
+      "epoch": 0.29979879275653926,
+      "grad_norm": 0.4927704930305481,
+      "learning_rate": 9.991617770326908e-06,
+      "loss": 0.3653,
+      "step": 1192
+    },
+    {
+      "epoch": 0.30005030181086517,
+      "grad_norm": 0.458484411239624,
+      "learning_rate": 1e-05,
+      "loss": 0.4349,
+      "step": 1193
+    },
+    {
+      "epoch": 0.30030181086519114,
+      "grad_norm": 0.44261494278907776,
+      "learning_rate": 9.999999785890641e-06,
+      "loss": 0.4293,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3005533199195171,
+      "grad_norm": 0.5019251108169556,
+      "learning_rate": 9.999999143562579e-06,
+      "loss": 0.423,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3008048289738431,
+      "grad_norm": 0.42558160424232483,
+      "learning_rate": 9.99999807301587e-06,
+      "loss": 0.3981,
+      "step": 1196
+    },
+    {
+      "epoch": 0.301056338028169,
+      "grad_norm": 0.44284799695014954,
+      "learning_rate": 9.999996574250606e-06,
+      "loss": 0.4336,
+      "step": 1197
+    },
+    {
+      "epoch": 0.301307847082495,
+      "grad_norm": 0.45541757345199585,
+      "learning_rate": 9.999994647266916e-06,
+      "loss": 0.4049,
+      "step": 1198
+    },
+    {
+      "epoch": 0.30155935613682094,
+      "grad_norm": 0.5015101432800293,
+      "learning_rate": 9.999992292064964e-06,
+      "loss": 0.3797,
+      "step": 1199
+    },
+    {
+      "epoch": 0.30181086519114686,
+      "grad_norm": 0.4215429723262787,
+      "learning_rate": 9.999989508644953e-06,
+      "loss": 0.4165,
+      "step": 1200
+    },
+    {
+      "epoch": 0.30206237424547283,
+      "grad_norm": 0.43915969133377075,
+      "learning_rate": 9.999986297007118e-06,
+      "loss": 0.4326,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3023138832997988,
+      "grad_norm": 0.46269261837005615,
+      "learning_rate": 9.999982657151738e-06,
+      "loss": 0.4243,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3025653923541248,
+      "grad_norm": 0.4217459261417389,
+      "learning_rate": 9.999978589079125e-06,
+      "loss": 0.4055,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3028169014084507,
+      "grad_norm": 0.5113700032234192,
+      "learning_rate": 9.999974092789623e-06,
+      "loss": 0.4091,
+      "step": 1204
+    },
+    {
+      "epoch": 0.30306841046277666,
+      "grad_norm": 0.4101250171661377,
+      "learning_rate": 9.999969168283621e-06,
+      "loss": 0.401,
+      "step": 1205
+    },
+    {
+      "epoch": 0.30331991951710263,
+      "grad_norm": 0.5615075826644897,
+      "learning_rate": 9.999963815561538e-06,
+      "loss": 0.4226,
+      "step": 1206
+    },
+    {
+      "epoch": 0.30357142857142855,
+      "grad_norm": 0.4130154550075531,
+      "learning_rate": 9.999958034623836e-06,
+      "loss": 0.4115,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3038229376257545,
+      "grad_norm": 0.4956052899360657,
+      "learning_rate": 9.999951825471005e-06,
+      "loss": 0.4305,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3040744466800805,
+      "grad_norm": 0.5325097441673279,
+      "learning_rate": 9.999945188103582e-06,
+      "loss": 0.4063,
+      "step": 1209
+    },
+    {
+      "epoch": 0.30432595573440646,
+      "grad_norm": 0.45363694429397583,
+      "learning_rate": 9.999938122522133e-06,
+      "loss": 0.4244,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3045774647887324,
+      "grad_norm": 0.41715431213378906,
+      "learning_rate": 9.999930628727264e-06,
+      "loss": 0.3687,
+      "step": 1211
+    },
+    {
+      "epoch": 0.30482897384305835,
+      "grad_norm": 0.49758753180503845,
+      "learning_rate": 9.999922706719614e-06,
+      "loss": 0.3923,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3050804828973843,
+      "grad_norm": 0.4848545491695404,
+      "learning_rate": 9.999914356499864e-06,
+      "loss": 0.4266,
+      "step": 1213
+    },
+    {
+      "epoch": 0.30533199195171024,
+      "grad_norm": 0.3956749439239502,
+      "learning_rate": 9.99990557806873e-06,
+      "loss": 0.3876,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3055835010060362,
+      "grad_norm": 0.4752289354801178,
+      "learning_rate": 9.999896371426962e-06,
+      "loss": 0.439,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3058350100603622,
+      "grad_norm": 0.4210512936115265,
+      "learning_rate": 9.999886736575349e-06,
+      "loss": 0.4215,
+      "step": 1216
+    },
+    {
+      "epoch": 0.30608651911468815,
+      "grad_norm": 0.44492197036743164,
+      "learning_rate": 9.999876673514718e-06,
+      "loss": 0.436,
+      "step": 1217
+    },
+    {
+      "epoch": 0.30633802816901406,
+      "grad_norm": 0.41508495807647705,
+      "learning_rate": 9.999866182245926e-06,
+      "loss": 0.3962,
+      "step": 1218
+    },
+    {
+      "epoch": 0.30658953722334004,
+      "grad_norm": 0.5030407905578613,
+      "learning_rate": 9.999855262769875e-06,
+      "loss": 0.4369,
+      "step": 1219
+    },
+    {
+      "epoch": 0.306841046277666,
+      "grad_norm": 0.43382754921913147,
+      "learning_rate": 9.9998439150875e-06,
+      "loss": 0.4056,
+      "step": 1220
+    },
+    {
+      "epoch": 0.307092555331992,
+      "grad_norm": 0.4502340853214264,
+      "learning_rate": 9.999832139199775e-06,
+      "loss": 0.4028,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3073440643863179,
+      "grad_norm": 0.46800076961517334,
+      "learning_rate": 9.999819935107705e-06,
+      "loss": 0.401,
+      "step": 1222
+    },
+    {
+      "epoch": 0.30759557344064387,
+      "grad_norm": 0.47623687982559204,
+      "learning_rate": 9.999807302812335e-06,
+      "loss": 0.394,
+      "step": 1223
+    },
+    {
+      "epoch": 0.30784708249496984,
+      "grad_norm": 0.4112638831138611,
+      "learning_rate": 9.99979424231475e-06,
+      "loss": 0.407,
+      "step": 1224
+    },
+    {
+      "epoch": 0.30809859154929575,
+      "grad_norm": 0.5141077041625977,
+      "learning_rate": 9.999780753616064e-06,
+      "loss": 0.4244,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3083501006036217,
+      "grad_norm": 0.4799872636795044,
+      "learning_rate": 9.999766836717437e-06,
+      "loss": 0.4207,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3086016096579477,
+      "grad_norm": 0.5113214254379272,
+      "learning_rate": 9.999752491620058e-06,
+      "loss": 0.4062,
+      "step": 1227
+    },
+    {
+      "epoch": 0.30885311871227367,
+      "grad_norm": 0.5039216876029968,
+      "learning_rate": 9.999737718325157e-06,
+      "loss": 0.42,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3091046277665996,
+      "grad_norm": 0.44573184847831726,
+      "learning_rate": 9.999722516833999e-06,
+      "loss": 0.4212,
+      "step": 1229
+    },
+    {
+      "epoch": 0.30935613682092555,
+      "grad_norm": 0.4311496615409851,
+      "learning_rate": 9.999706887147884e-06,
+      "loss": 0.4031,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3096076458752515,
+      "grad_norm": 0.5166363716125488,
+      "learning_rate": 9.999690829268154e-06,
+      "loss": 0.4214,
+      "step": 1231
+    },
+    {
+      "epoch": 0.30985915492957744,
+      "grad_norm": 0.44504567980766296,
+      "learning_rate": 9.999674343196182e-06,
+      "loss": 0.3918,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3101106639839034,
+      "grad_norm": 0.44314002990722656,
+      "learning_rate": 9.99965742893338e-06,
+      "loss": 0.4428,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3103621730382294,
+      "grad_norm": 0.4878178536891937,
+      "learning_rate": 9.999640086481198e-06,
+      "loss": 0.3875,
+      "step": 1234
+    },
+    {
+      "epoch": 0.31061368209255535,
+      "grad_norm": 0.45717740058898926,
+      "learning_rate": 9.99962231584112e-06,
+      "loss": 0.3977,
+      "step": 1235
+    },
+    {
+      "epoch": 0.31086519114688127,
+      "grad_norm": 0.4271807372570038,
+      "learning_rate": 9.999604117014667e-06,
+      "loss": 0.4027,
+      "step": 1236
+    },
+    {
+      "epoch": 0.31111670020120724,
+      "grad_norm": 0.4877811074256897,
+      "learning_rate": 9.999585490003399e-06,
+      "loss": 0.3856,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3113682092555332,
+      "grad_norm": 0.5068530440330505,
+      "learning_rate": 9.999566434808912e-06,
+      "loss": 0.3895,
+      "step": 1238
+    },
+    {
+      "epoch": 0.31161971830985913,
+      "grad_norm": 0.4381490647792816,
+      "learning_rate": 9.999546951432837e-06,
+      "loss": 0.4222,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3118712273641851,
+      "grad_norm": 0.44648122787475586,
+      "learning_rate": 9.999527039876843e-06,
+      "loss": 0.4092,
+      "step": 1240
+    },
+    {
+      "epoch": 0.31212273641851107,
+      "grad_norm": 0.5271798968315125,
+      "learning_rate": 9.999506700142633e-06,
+      "loss": 0.4265,
+      "step": 1241
+    },
+    {
+      "epoch": 0.31237424547283704,
+      "grad_norm": 0.467507928609848,
+      "learning_rate": 9.999485932231951e-06,
+      "loss": 0.4255,
+      "step": 1242
+    },
+    {
+      "epoch": 0.31262575452716296,
+      "grad_norm": 0.43331214785575867,
+      "learning_rate": 9.999464736146578e-06,
+      "loss": 0.4022,
+      "step": 1243
+    },
+    {
+      "epoch": 0.31287726358148893,
+      "grad_norm": 0.511960506439209,
+      "learning_rate": 9.999443111888325e-06,
+      "loss": 0.4229,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3131287726358149,
+      "grad_norm": 0.4495149850845337,
+      "learning_rate": 9.999421059459047e-06,
+      "loss": 0.4109,
+      "step": 1245
+    },
+    {
+      "epoch": 0.31338028169014087,
+      "grad_norm": 0.4889768362045288,
+      "learning_rate": 9.999398578860631e-06,
+      "loss": 0.3932,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3136317907444668,
+      "grad_norm": 0.4641020596027374,
+      "learning_rate": 9.999375670095003e-06,
+      "loss": 0.4236,
+      "step": 1247
+    },
+    {
+      "epoch": 0.31388329979879276,
+      "grad_norm": 0.47060105204582214,
+      "learning_rate": 9.999352333164125e-06,
+      "loss": 0.418,
+      "step": 1248
+    },
+    {
+      "epoch": 0.31413480885311873,
+      "grad_norm": 0.4797796308994293,
+      "learning_rate": 9.999328568069994e-06,
+      "loss": 0.4168,
+      "step": 1249
+    },
+    {
+      "epoch": 0.31438631790744465,
+      "grad_norm": 0.474357932806015,
+      "learning_rate": 9.99930437481465e-06,
+      "loss": 0.4113,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3146378269617706,
+      "grad_norm": 0.4692329466342926,
+      "learning_rate": 9.99927975340016e-06,
+      "loss": 0.3967,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3148893360160966,
+      "grad_norm": 0.5016288161277771,
+      "learning_rate": 9.999254703828634e-06,
+      "loss": 0.427,
+      "step": 1252
+    },
+    {
+      "epoch": 0.31514084507042256,
+      "grad_norm": 0.42063063383102417,
+      "learning_rate": 9.999229226102218e-06,
+      "loss": 0.423,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3153923541247485,
+      "grad_norm": 0.4207392632961273,
+      "learning_rate": 9.999203320223095e-06,
+      "loss": 0.4237,
+      "step": 1254
+    },
+    {
+      "epoch": 0.31564386317907445,
+      "grad_norm": 0.5132213830947876,
+      "learning_rate": 9.999176986193481e-06,
+      "loss": 0.3826,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3158953722334004,
+      "grad_norm": 0.4629036784172058,
+      "learning_rate": 9.999150224015634e-06,
+      "loss": 0.3976,
+      "step": 1256
+    },
+    {
+      "epoch": 0.31614688128772633,
+      "grad_norm": 0.45483675599098206,
+      "learning_rate": 9.999123033691844e-06,
+      "loss": 0.4291,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3163983903420523,
+      "grad_norm": 0.45758309960365295,
+      "learning_rate": 9.999095415224443e-06,
+      "loss": 0.4088,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3166498993963783,
+      "grad_norm": 0.49029773473739624,
+      "learning_rate": 9.999067368615791e-06,
+      "loss": 0.4092,
+      "step": 1259
+    },
+    {
+      "epoch": 0.31690140845070425,
+      "grad_norm": 0.4661732017993927,
+      "learning_rate": 9.999038893868293e-06,
+      "loss": 0.4094,
+      "step": 1260
+    },
+    {
+      "epoch": 0.31715291750503016,
+      "grad_norm": 0.43760454654693604,
+      "learning_rate": 9.999009990984389e-06,
+      "loss": 0.3962,
+      "step": 1261
+    },
+    {
+      "epoch": 0.31740442655935613,
+      "grad_norm": 0.46413469314575195,
+      "learning_rate": 9.998980659966553e-06,
+      "loss": 0.4161,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3176559356136821,
+      "grad_norm": 0.46015673875808716,
+      "learning_rate": 9.998950900817297e-06,
+      "loss": 0.3893,
+      "step": 1263
+    },
+    {
+      "epoch": 0.317907444668008,
+      "grad_norm": 0.44548308849334717,
+      "learning_rate": 9.998920713539169e-06,
+      "loss": 0.4252,
+      "step": 1264
+    },
+    {
+      "epoch": 0.318158953722334,
+      "grad_norm": 0.452181875705719,
+      "learning_rate": 9.998890098134757e-06,
+      "loss": 0.4226,
+      "step": 1265
+    },
+    {
+      "epoch": 0.31841046277665996,
+      "grad_norm": 0.5392069220542908,
+      "learning_rate": 9.998859054606677e-06,
+      "loss": 0.4124,
+      "step": 1266
+    },
+    {
+      "epoch": 0.31866197183098594,
+      "grad_norm": 0.43742406368255615,
+      "learning_rate": 9.998827582957596e-06,
+      "loss": 0.4125,
+      "step": 1267
+    },
+    {
+      "epoch": 0.31891348088531185,
+      "grad_norm": 0.4227537512779236,
+      "learning_rate": 9.998795683190202e-06,
+      "loss": 0.4211,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3191649899396378,
+      "grad_norm": 0.4836387038230896,
+      "learning_rate": 9.998763355307232e-06,
+      "loss": 0.4342,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3194164989939638,
+      "grad_norm": 0.48729145526885986,
+      "learning_rate": 9.998730599311452e-06,
+      "loss": 0.4045,
+      "step": 1270
+    },
+    {
+      "epoch": 0.31966800804828976,
+      "grad_norm": 0.46449029445648193,
+      "learning_rate": 9.998697415205667e-06,
+      "loss": 0.4175,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3199195171026157,
+      "grad_norm": 0.4990949332714081,
+      "learning_rate": 9.998663802992723e-06,
+      "loss": 0.435,
+      "step": 1272
+    },
+    {
+      "epoch": 0.32017102615694165,
+      "grad_norm": 0.42860570549964905,
+      "learning_rate": 9.998629762675493e-06,
+      "loss": 0.4114,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3204225352112676,
+      "grad_norm": 0.5099340081214905,
+      "learning_rate": 9.998595294256897e-06,
+      "loss": 0.4395,
+      "step": 1274
+    },
+    {
+      "epoch": 0.32067404426559354,
+      "grad_norm": 0.5020102262496948,
+      "learning_rate": 9.998560397739885e-06,
+      "loss": 0.3879,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3209255533199195,
+      "grad_norm": 0.41706791520118713,
+      "learning_rate": 9.998525073127445e-06,
+      "loss": 0.4166,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3211770623742455,
+      "grad_norm": 0.43367356061935425,
+      "learning_rate": 9.998489320422604e-06,
+      "loss": 0.4152,
+      "step": 1277
+    },
+    {
+      "epoch": 0.32142857142857145,
+      "grad_norm": 0.3915824294090271,
+      "learning_rate": 9.998453139628422e-06,
+      "loss": 0.393,
+      "step": 1278
+    },
+    {
+      "epoch": 0.32168008048289737,
+      "grad_norm": 0.49054497480392456,
+      "learning_rate": 9.998416530748e-06,
+      "loss": 0.4235,
+      "step": 1279
+    },
+    {
+      "epoch": 0.32193158953722334,
+      "grad_norm": 0.4301794469356537,
+      "learning_rate": 9.998379493784472e-06,
+      "loss": 0.4096,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3221830985915493,
+      "grad_norm": 0.5517282485961914,
+      "learning_rate": 9.998342028741009e-06,
+      "loss": 0.395,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3224346076458752,
+      "grad_norm": 0.47454625368118286,
+      "learning_rate": 9.998304135620822e-06,
+      "loss": 0.4091,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3226861167002012,
+      "grad_norm": 0.44431060552597046,
+      "learning_rate": 9.998265814427156e-06,
+      "loss": 0.4341,
+      "step": 1283
+    },
+    {
+      "epoch": 0.32293762575452717,
+      "grad_norm": 0.5420903563499451,
+      "learning_rate": 9.998227065163289e-06,
+      "loss": 0.3937,
+      "step": 1284
+    },
+    {
+      "epoch": 0.32318913480885314,
+      "grad_norm": 0.49979931116104126,
+      "learning_rate": 9.998187887832544e-06,
+      "loss": 0.4236,
+      "step": 1285
+    },
+    {
+      "epoch": 0.32344064386317906,
+      "grad_norm": 0.488714337348938,
+      "learning_rate": 9.998148282438276e-06,
+      "loss": 0.3862,
+      "step": 1286
+    },
+    {
+      "epoch": 0.323692152917505,
+      "grad_norm": 0.5104209184646606,
+      "learning_rate": 9.998108248983875e-06,
+      "loss": 0.4263,
+      "step": 1287
+    },
+    {
+      "epoch": 0.323943661971831,
+      "grad_norm": 0.5148292779922485,
+      "learning_rate": 9.998067787472772e-06,
+      "loss": 0.3773,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3241951710261569,
+      "grad_norm": 0.46651163697242737,
+      "learning_rate": 9.998026897908429e-06,
+      "loss": 0.4054,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3244466800804829,
+      "grad_norm": 0.5032572150230408,
+      "learning_rate": 9.99798558029435e-06,
+      "loss": 0.3935,
+      "step": 1290
+    },
+    {
+      "epoch": 0.32469818913480886,
+      "grad_norm": 0.4591807425022125,
+      "learning_rate": 9.997943834634074e-06,
+      "loss": 0.3915,
+      "step": 1291
+    },
+    {
+      "epoch": 0.32494969818913483,
+      "grad_norm": 0.4753274917602539,
+      "learning_rate": 9.997901660931175e-06,
+      "loss": 0.4095,
+      "step": 1292
+    },
+    {
+      "epoch": 0.32520120724346074,
+      "grad_norm": 0.4749658703804016,
+      "learning_rate": 9.997859059189266e-06,
+      "loss": 0.384,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3254527162977867,
+      "grad_norm": 0.4379113018512726,
+      "learning_rate": 9.997816029411996e-06,
+      "loss": 0.4158,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3257042253521127,
+      "grad_norm": 0.5521072745323181,
+      "learning_rate": 9.997772571603047e-06,
+      "loss": 0.4081,
+      "step": 1295
+    },
+    {
+      "epoch": 0.32595573440643866,
+      "grad_norm": 0.45863693952560425,
+      "learning_rate": 9.997728685766144e-06,
+      "loss": 0.4219,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3262072434607646,
+      "grad_norm": 0.5103247165679932,
+      "learning_rate": 9.997684371905046e-06,
+      "loss": 0.428,
+      "step": 1297
+    },
+    {
+      "epoch": 0.32645875251509054,
+      "grad_norm": 0.4744894802570343,
+      "learning_rate": 9.997639630023546e-06,
+      "loss": 0.4249,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3267102615694165,
+      "grad_norm": 0.47997725009918213,
+      "learning_rate": 9.997594460125477e-06,
+      "loss": 0.4097,
+      "step": 1299
+    },
+    {
+      "epoch": 0.32696177062374243,
+      "grad_norm": 0.48586708307266235,
+      "learning_rate": 9.997548862214708e-06,
+      "loss": 0.4036,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3272132796780684,
+      "grad_norm": 0.4447525441646576,
+      "learning_rate": 9.99750283629514e-06,
+      "loss": 0.3932,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3274647887323944,
+      "grad_norm": 0.4103879928588867,
+      "learning_rate": 9.997456382370723e-06,
+      "loss": 0.4181,
+      "step": 1302
+    },
+    {
+      "epoch": 0.32771629778672035,
+      "grad_norm": 0.4686889052391052,
+      "learning_rate": 9.99740950044543e-06,
+      "loss": 0.4132,
+      "step": 1303
+    },
+    {
+      "epoch": 0.32796780684104626,
+      "grad_norm": 0.41833746433258057,
+      "learning_rate": 9.997362190523275e-06,
+      "loss": 0.4269,
+      "step": 1304
+    },
+    {
+      "epoch": 0.32821931589537223,
+      "grad_norm": 0.4362890124320984,
+      "learning_rate": 9.997314452608313e-06,
+      "loss": 0.4028,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3284708249496982,
+      "grad_norm": 0.47095364332199097,
+      "learning_rate": 9.99726628670463e-06,
+      "loss": 0.4157,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3287223340040241,
+      "grad_norm": 0.44647061824798584,
+      "learning_rate": 9.997217692816355e-06,
+      "loss": 0.3963,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3289738430583501,
+      "grad_norm": 0.44647330045700073,
+      "learning_rate": 9.997168670947644e-06,
+      "loss": 0.3993,
+      "step": 1308
+    },
+    {
+      "epoch": 0.32922535211267606,
+      "grad_norm": 0.46429726481437683,
+      "learning_rate": 9.9971192211027e-06,
+      "loss": 0.4067,
+      "step": 1309
+    },
+    {
+      "epoch": 0.32947686116700203,
+      "grad_norm": 0.465006560087204,
+      "learning_rate": 9.997069343285757e-06,
+      "loss": 0.3886,
+      "step": 1310
+    },
+    {
+      "epoch": 0.32972837022132795,
+      "grad_norm": 0.4865345358848572,
+      "learning_rate": 9.997019037501086e-06,
+      "loss": 0.4055,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3299798792756539,
+      "grad_norm": 0.46681225299835205,
+      "learning_rate": 9.996968303752996e-06,
+      "loss": 0.4287,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3302313883299799,
+      "grad_norm": 0.5172387361526489,
+      "learning_rate": 9.996917142045832e-06,
+      "loss": 0.4436,
+      "step": 1313
+    },
+    {
+      "epoch": 0.33048289738430586,
+      "grad_norm": 0.3964807689189911,
+      "learning_rate": 9.996865552383975e-06,
+      "loss": 0.4051,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3307344064386318,
+      "grad_norm": 0.41110795736312866,
+      "learning_rate": 9.996813534771843e-06,
+      "loss": 0.3752,
+      "step": 1315
+    },
+    {
+      "epoch": 0.33098591549295775,
+      "grad_norm": 0.4638274908065796,
+      "learning_rate": 9.996761089213891e-06,
+      "loss": 0.4208,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3312374245472837,
+      "grad_norm": 0.4156520962715149,
+      "learning_rate": 9.996708215714613e-06,
+      "loss": 0.4042,
+      "step": 1317
+    },
+    {
+      "epoch": 0.33148893360160964,
+      "grad_norm": 0.42104753851890564,
+      "learning_rate": 9.996654914278535e-06,
+      "loss": 0.4499,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3317404426559356,
+      "grad_norm": 0.42806974053382874,
+      "learning_rate": 9.996601184910223e-06,
+      "loss": 0.3979,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3319919517102616,
+      "grad_norm": 0.38782554864883423,
+      "learning_rate": 9.996547027614279e-06,
+      "loss": 0.392,
+      "step": 1320
+    },
+    {
+      "epoch": 0.33224346076458755,
+      "grad_norm": 0.4238336682319641,
+      "learning_rate": 9.996492442395338e-06,
+      "loss": 0.4193,
+      "step": 1321
+    },
+    {
+      "epoch": 0.33249496981891347,
+      "grad_norm": 0.4696759879589081,
+      "learning_rate": 9.996437429258079e-06,
+      "loss": 0.4137,
+      "step": 1322
+    },
+    {
+      "epoch": 0.33274647887323944,
+      "grad_norm": 0.4905252158641815,
+      "learning_rate": 9.996381988207211e-06,
+      "loss": 0.4354,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3329979879275654,
+      "grad_norm": 0.4017488360404968,
+      "learning_rate": 9.996326119247484e-06,
+      "loss": 0.4061,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3332494969818913,
+      "grad_norm": 0.4732036292552948,
+      "learning_rate": 9.99626982238368e-06,
+      "loss": 0.4296,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3335010060362173,
+      "grad_norm": 0.4144299626350403,
+      "learning_rate": 9.996213097620623e-06,
+      "loss": 0.4169,
+      "step": 1326
+    },
+    {
+      "epoch": 0.33375251509054327,
+      "grad_norm": 0.45625752210617065,
+      "learning_rate": 9.996155944963173e-06,
+      "loss": 0.4019,
+      "step": 1327
+    },
+    {
+      "epoch": 0.33400402414486924,
+      "grad_norm": 0.3766085207462311,
+      "learning_rate": 9.996098364416219e-06,
+      "loss": 0.3802,
+      "step": 1328
+    },
+    {
+      "epoch": 0.33425553319919515,
+      "grad_norm": 0.40613994002342224,
+      "learning_rate": 9.996040355984697e-06,
+      "loss": 0.3963,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3345070422535211,
+      "grad_norm": 0.4915297031402588,
+      "learning_rate": 9.995981919673571e-06,
+      "loss": 0.4075,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3347585513078471,
+      "grad_norm": 0.40623193979263306,
+      "learning_rate": 9.995923055487853e-06,
+      "loss": 0.392,
+      "step": 1331
+    },
+    {
+      "epoch": 0.335010060362173,
+      "grad_norm": 0.47242090106010437,
+      "learning_rate": 9.995863763432577e-06,
+      "loss": 0.3792,
+      "step": 1332
+    },
+    {
+      "epoch": 0.335261569416499,
+      "grad_norm": 0.39874276518821716,
+      "learning_rate": 9.995804043512824e-06,
+      "loss": 0.3989,
+      "step": 1333
+    },
+    {
+      "epoch": 0.33551307847082495,
+      "grad_norm": 0.4230663776397705,
+      "learning_rate": 9.995743895733707e-06,
+      "loss": 0.4004,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3357645875251509,
+      "grad_norm": 0.4566534161567688,
+      "learning_rate": 9.99568332010038e-06,
+      "loss": 0.4478,
+      "step": 1335
+    },
+    {
+      "epoch": 0.33601609657947684,
+      "grad_norm": 0.48702409863471985,
+      "learning_rate": 9.99562231661803e-06,
+      "loss": 0.4014,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3362676056338028,
+      "grad_norm": 0.4124425947666168,
+      "learning_rate": 9.995560885291879e-06,
+      "loss": 0.4075,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3365191146881288,
+      "grad_norm": 0.4736267626285553,
+      "learning_rate": 9.995499026127194e-06,
+      "loss": 0.381,
+      "step": 1338
+    },
+    {
+      "epoch": 0.33677062374245476,
+      "grad_norm": 0.41389134526252747,
+      "learning_rate": 9.995436739129267e-06,
+      "loss": 0.4299,
+      "step": 1339
+    },
+    {
+      "epoch": 0.33702213279678067,
+      "grad_norm": 0.49148082733154297,
+      "learning_rate": 9.995374024303433e-06,
+      "loss": 0.4016,
+      "step": 1340
+    },
+    {
+      "epoch": 0.33727364185110664,
+      "grad_norm": 0.4507843554019928,
+      "learning_rate": 9.995310881655066e-06,
+      "loss": 0.4305,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3375251509054326,
+      "grad_norm": 0.41378965973854065,
+      "learning_rate": 9.995247311189573e-06,
+      "loss": 0.3896,
+      "step": 1342
+    },
+    {
+      "epoch": 0.33777665995975853,
+      "grad_norm": 0.511529803276062,
+      "learning_rate": 9.9951833129124e-06,
+      "loss": 0.4367,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 0.4993535578250885,
+      "learning_rate": 9.995118886829023e-06,
+      "loss": 0.4217,
+      "step": 1344
+    },
+    {
+      "epoch": 0.33827967806841047,
+      "grad_norm": 0.4538334012031555,
+      "learning_rate": 9.995054032944963e-06,
+      "loss": 0.4088,
+      "step": 1345
+    },
+    {
+      "epoch": 0.33853118712273644,
+      "grad_norm": 0.42592087388038635,
+      "learning_rate": 9.994988751265775e-06,
+      "loss": 0.4142,
+      "step": 1346
+    },
+    {
+      "epoch": 0.33878269617706236,
+      "grad_norm": 0.4514321982860565,
+      "learning_rate": 9.994923041797049e-06,
+      "loss": 0.4318,
+      "step": 1347
+    },
+    {
+      "epoch": 0.33903420523138833,
+      "grad_norm": 0.43856939673423767,
+      "learning_rate": 9.994856904544412e-06,
+      "loss": 0.4195,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3392857142857143,
+      "grad_norm": 0.4376979470252991,
+      "learning_rate": 9.99479033951353e-06,
+      "loss": 0.391,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3395372233400402,
+      "grad_norm": 0.43217065930366516,
+      "learning_rate": 9.9947233467101e-06,
+      "loss": 0.4241,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3397887323943662,
+      "grad_norm": 0.44194620847702026,
+      "learning_rate": 9.994655926139864e-06,
+      "loss": 0.401,
+      "step": 1351
+    },
+    {
+      "epoch": 0.34004024144869216,
+      "grad_norm": 0.439430296421051,
+      "learning_rate": 9.994588077808595e-06,
+      "loss": 0.4185,
+      "step": 1352
+    },
+    {
+      "epoch": 0.34029175050301813,
+      "grad_norm": 0.4721353054046631,
+      "learning_rate": 9.994519801722103e-06,
+      "loss": 0.4053,
+      "step": 1353
+    },
+    {
+      "epoch": 0.34054325955734405,
+      "grad_norm": 0.4690324068069458,
+      "learning_rate": 9.994451097886236e-06,
+      "loss": 0.4294,
+      "step": 1354
+    },
+    {
+      "epoch": 0.34079476861167,
+      "grad_norm": 0.4453498125076294,
+      "learning_rate": 9.994381966306877e-06,
+      "loss": 0.4371,
+      "step": 1355
+    },
+    {
+      "epoch": 0.341046277665996,
+      "grad_norm": 0.43481093645095825,
+      "learning_rate": 9.994312406989947e-06,
+      "loss": 0.4068,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3412977867203219,
+      "grad_norm": 0.45696157217025757,
+      "learning_rate": 9.994242419941403e-06,
+      "loss": 0.4137,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3415492957746479,
+      "grad_norm": 0.5058369636535645,
+      "learning_rate": 9.994172005167242e-06,
+      "loss": 0.4273,
+      "step": 1358
+    },
+    {
+      "epoch": 0.34180080482897385,
+      "grad_norm": 0.43768513202667236,
+      "learning_rate": 9.994101162673491e-06,
+      "loss": 0.3701,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3420523138832998,
+      "grad_norm": 0.447859525680542,
+      "learning_rate": 9.994029892466218e-06,
+      "loss": 0.4028,
+      "step": 1360
+    },
+    {
+      "epoch": 0.34230382293762573,
+      "grad_norm": 0.5294939279556274,
+      "learning_rate": 9.993958194551528e-06,
+      "loss": 0.4043,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3425553319919517,
+      "grad_norm": 0.4057393968105316,
+      "learning_rate": 9.993886068935559e-06,
+      "loss": 0.4308,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3428068410462777,
+      "grad_norm": 0.4801323413848877,
+      "learning_rate": 9.993813515624491e-06,
+      "loss": 0.3747,
+      "step": 1363
+    },
+    {
+      "epoch": 0.34305835010060365,
+      "grad_norm": 0.519101083278656,
+      "learning_rate": 9.993740534624536e-06,
+      "loss": 0.3993,
+      "step": 1364
+    },
+    {
+      "epoch": 0.34330985915492956,
+      "grad_norm": 0.44538334012031555,
+      "learning_rate": 9.993667125941946e-06,
+      "loss": 0.3992,
+      "step": 1365
+    },
+    {
+      "epoch": 0.34356136820925554,
+      "grad_norm": 0.5548309087753296,
+      "learning_rate": 9.993593289583005e-06,
+      "loss": 0.4186,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3438128772635815,
+      "grad_norm": 0.46345213055610657,
+      "learning_rate": 9.993519025554041e-06,
+      "loss": 0.3935,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3440643863179074,
+      "grad_norm": 0.5251536965370178,
+      "learning_rate": 9.993444333861411e-06,
+      "loss": 0.4056,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3443158953722334,
+      "grad_norm": 0.5084334015846252,
+      "learning_rate": 9.993369214511512e-06,
+      "loss": 0.4039,
+      "step": 1369
+    },
+    {
+      "epoch": 0.34456740442655936,
+      "grad_norm": 0.5444664359092712,
+      "learning_rate": 9.99329366751078e-06,
+      "loss": 0.3857,
+      "step": 1370
+    },
+    {
+      "epoch": 0.34481891348088534,
+      "grad_norm": 0.4715754985809326,
+      "learning_rate": 9.993217692865683e-06,
+      "loss": 0.4328,
+      "step": 1371
+    },
+    {
+      "epoch": 0.34507042253521125,
+      "grad_norm": 0.5068243741989136,
+      "learning_rate": 9.993141290582726e-06,
+      "loss": 0.379,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3453219315895372,
+      "grad_norm": 0.5259790420532227,
+      "learning_rate": 9.993064460668456e-06,
+      "loss": 0.4116,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3455734406438632,
+      "grad_norm": 0.46770498156547546,
+      "learning_rate": 9.992987203129451e-06,
+      "loss": 0.4302,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3458249496981891,
+      "grad_norm": 0.47269558906555176,
+      "learning_rate": 9.99290951797233e-06,
+      "loss": 0.4102,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3460764587525151,
+      "grad_norm": 0.43284550309181213,
+      "learning_rate": 9.99283140520374e-06,
+      "loss": 0.4029,
+      "step": 1376
+    },
+    {
+      "epoch": 0.34632796780684105,
+      "grad_norm": 0.49771684408187866,
+      "learning_rate": 9.992752864830379e-06,
+      "loss": 0.414,
+      "step": 1377
+    },
+    {
+      "epoch": 0.346579476861167,
+      "grad_norm": 0.4438658356666565,
+      "learning_rate": 9.992673896858969e-06,
+      "loss": 0.4105,
+      "step": 1378
+    },
+    {
+      "epoch": 0.34683098591549294,
+      "grad_norm": 0.46066081523895264,
+      "learning_rate": 9.992594501296272e-06,
+      "loss": 0.4012,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3470824949698189,
+      "grad_norm": 0.562817394733429,
+      "learning_rate": 9.992514678149092e-06,
+      "loss": 0.424,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3473340040241449,
+      "grad_norm": 0.5239977240562439,
+      "learning_rate": 9.992434427424261e-06,
+      "loss": 0.3936,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3475855130784708,
+      "grad_norm": 0.4289341866970062,
+      "learning_rate": 9.992353749128653e-06,
+      "loss": 0.4092,
+      "step": 1382
+    },
+    {
+      "epoch": 0.34783702213279677,
+      "grad_norm": 0.565528154373169,
+      "learning_rate": 9.992272643269181e-06,
+      "loss": 0.4239,
+      "step": 1383
+    },
+    {
+      "epoch": 0.34808853118712274,
+      "grad_norm": 0.44149690866470337,
+      "learning_rate": 9.992191109852788e-06,
+      "loss": 0.4068,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3483400402414487,
+      "grad_norm": 0.4838801920413971,
+      "learning_rate": 9.992109148886457e-06,
+      "loss": 0.4237,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3485915492957746,
+      "grad_norm": 0.5003640651702881,
+      "learning_rate": 9.992026760377207e-06,
+      "loss": 0.3963,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3488430583501006,
+      "grad_norm": 0.4707315266132355,
+      "learning_rate": 9.991943944332097e-06,
+      "loss": 0.4227,
+      "step": 1387
+    },
+    {
+      "epoch": 0.34909456740442657,
+      "grad_norm": 0.4761320650577545,
+      "learning_rate": 9.991860700758217e-06,
+      "loss": 0.4195,
+      "step": 1388
+    },
+    {
+      "epoch": 0.34934607645875254,
+      "grad_norm": 0.39643487334251404,
+      "learning_rate": 9.991777029662698e-06,
+      "loss": 0.4183,
+      "step": 1389
+    },
+    {
+      "epoch": 0.34959758551307846,
+      "grad_norm": 0.4800592064857483,
+      "learning_rate": 9.991692931052703e-06,
+      "loss": 0.4301,
+      "step": 1390
+    },
+    {
+      "epoch": 0.34984909456740443,
+      "grad_norm": 0.42438507080078125,
+      "learning_rate": 9.991608404935435e-06,
+      "loss": 0.4011,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3501006036217304,
+      "grad_norm": 0.3981070816516876,
+      "learning_rate": 9.991523451318137e-06,
+      "loss": 0.4184,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3503521126760563,
+      "grad_norm": 0.4492408037185669,
+      "learning_rate": 9.991438070208082e-06,
+      "loss": 0.4009,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3506036217303823,
+      "grad_norm": 0.41902977228164673,
+      "learning_rate": 9.991352261612583e-06,
+      "loss": 0.4174,
+      "step": 1394
+    },
+    {
+      "epoch": 0.35085513078470826,
+      "grad_norm": 0.4707157611846924,
+      "learning_rate": 9.991266025538988e-06,
+      "loss": 0.4226,
+      "step": 1395
+    },
+    {
+      "epoch": 0.35110663983903423,
+      "grad_norm": 0.4216676950454712,
+      "learning_rate": 9.991179361994683e-06,
+      "loss": 0.4397,
+      "step": 1396
+    },
+    {
+      "epoch": 0.35135814889336014,
+      "grad_norm": 0.47134819626808167,
+      "learning_rate": 9.99109227098709e-06,
+      "loss": 0.3967,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3516096579476861,
+      "grad_norm": 0.4079870283603668,
+      "learning_rate": 9.991004752523668e-06,
+      "loss": 0.394,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3518611670020121,
+      "grad_norm": 0.46213197708129883,
+      "learning_rate": 9.990916806611915e-06,
+      "loss": 0.413,
+      "step": 1399
+    },
+    {
+      "epoch": 0.352112676056338,
+      "grad_norm": 0.40814223885536194,
+      "learning_rate": 9.99082843325936e-06,
+      "loss": 0.4259,
+      "step": 1400
+    },
+    {
+      "epoch": 0.352364185110664,
+      "grad_norm": 0.39051973819732666,
+      "learning_rate": 9.99073963247357e-06,
+      "loss": 0.4197,
+      "step": 1401
+    },
+    {
+      "epoch": 0.35261569416498995,
+      "grad_norm": 0.4494209885597229,
+      "learning_rate": 9.990650404262152e-06,
+      "loss": 0.3972,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3528672032193159,
+      "grad_norm": 0.43199771642684937,
+      "learning_rate": 9.99056074863275e-06,
+      "loss": 0.4034,
+      "step": 1403
+    },
+    {
+      "epoch": 0.35311871227364183,
+      "grad_norm": 0.46416175365448,
+      "learning_rate": 9.99047066559304e-06,
+      "loss": 0.4253,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3533702213279678,
+      "grad_norm": 0.49905675649642944,
+      "learning_rate": 9.990380155150739e-06,
+      "loss": 0.4184,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3536217303822938,
+      "grad_norm": 0.41556206345558167,
+      "learning_rate": 9.990289217313597e-06,
+      "loss": 0.3867,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3538732394366197,
+      "grad_norm": 0.47178298234939575,
+      "learning_rate": 9.990197852089403e-06,
+      "loss": 0.4356,
+      "step": 1407
+    },
+    {
+      "epoch": 0.35412474849094566,
+      "grad_norm": 0.4404408037662506,
+      "learning_rate": 9.99010605948598e-06,
+      "loss": 0.4127,
+      "step": 1408
+    },
+    {
+      "epoch": 0.35437625754527163,
+      "grad_norm": 0.4431535601615906,
+      "learning_rate": 9.990013839511193e-06,
+      "loss": 0.3965,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3546277665995976,
+      "grad_norm": 0.4630604684352875,
+      "learning_rate": 9.989921192172936e-06,
+      "loss": 0.4251,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3548792756539235,
+      "grad_norm": 0.4784621298313141,
+      "learning_rate": 9.989828117479149e-06,
+      "loss": 0.4314,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3551307847082495,
+      "grad_norm": 0.4682464301586151,
+      "learning_rate": 9.989734615437797e-06,
+      "loss": 0.3869,
+      "step": 1412
+    },
+    {
+      "epoch": 0.35538229376257546,
+      "grad_norm": 0.4220459759235382,
+      "learning_rate": 9.989640686056891e-06,
+      "loss": 0.4142,
+      "step": 1413
+    },
+    {
+      "epoch": 0.35563380281690143,
+      "grad_norm": 0.4825068712234497,
+      "learning_rate": 9.989546329344474e-06,
+      "loss": 0.3999,
+      "step": 1414
+    },
+    {
+      "epoch": 0.35588531187122735,
+      "grad_norm": 0.47478923201560974,
+      "learning_rate": 9.989451545308633e-06,
+      "loss": 0.4274,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3561368209255533,
+      "grad_norm": 0.37347909808158875,
+      "learning_rate": 9.989356333957477e-06,
+      "loss": 0.4146,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3563883299798793,
+      "grad_norm": 0.484246164560318,
+      "learning_rate": 9.989260695299165e-06,
+      "loss": 0.3937,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3566398390342052,
+      "grad_norm": 0.47754985094070435,
+      "learning_rate": 9.989164629341889e-06,
+      "loss": 0.4186,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3568913480885312,
+      "grad_norm": 0.45537054538726807,
+      "learning_rate": 9.989068136093873e-06,
+      "loss": 0.4436,
+      "step": 1419
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.4763503074645996,
+      "learning_rate": 9.988971215563383e-06,
+      "loss": 0.4262,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3573943661971831,
+      "grad_norm": 0.41354265809059143,
+      "learning_rate": 9.98887386775872e-06,
+      "loss": 0.4016,
+      "step": 1421
+    },
+    {
+      "epoch": 0.35764587525150904,
+      "grad_norm": 0.4630916118621826,
+      "learning_rate": 9.988776092688221e-06,
+      "loss": 0.438,
+      "step": 1422
+    },
+    {
+      "epoch": 0.357897384305835,
+      "grad_norm": 0.40816864371299744,
+      "learning_rate": 9.988677890360258e-06,
+      "loss": 0.4204,
+      "step": 1423
+    },
+    {
+      "epoch": 0.358148893360161,
+      "grad_norm": 0.4877932071685791,
+      "learning_rate": 9.988579260783242e-06,
+      "loss": 0.3938,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3584004024144869,
+      "grad_norm": 0.4669157862663269,
+      "learning_rate": 9.988480203965623e-06,
+      "loss": 0.4211,
+      "step": 1425
+    },
+    {
+      "epoch": 0.35865191146881287,
+      "grad_norm": 0.40130600333213806,
+      "learning_rate": 9.988380719915881e-06,
+      "loss": 0.4003,
+      "step": 1426
+    },
+    {
+      "epoch": 0.35890342052313884,
+      "grad_norm": 0.4261634647846222,
+      "learning_rate": 9.988280808642538e-06,
+      "loss": 0.4073,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3591549295774648,
+      "grad_norm": 0.4852161705493927,
+      "learning_rate": 9.988180470154149e-06,
+      "loss": 0.4191,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3594064386317907,
+      "grad_norm": 0.4650234878063202,
+      "learning_rate": 9.98807970445931e-06,
+      "loss": 0.4394,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3596579476861167,
+      "grad_norm": 0.4383230209350586,
+      "learning_rate": 9.987978511566651e-06,
+      "loss": 0.3868,
+      "step": 1430
+    },
+    {
+      "epoch": 0.35990945674044267,
+      "grad_norm": 0.4849243760108948,
+      "learning_rate": 9.987876891484836e-06,
+      "loss": 0.4096,
+      "step": 1431
+    },
+    {
+      "epoch": 0.36016096579476864,
+      "grad_norm": 0.40414345264434814,
+      "learning_rate": 9.987774844222568e-06,
+      "loss": 0.4105,
+      "step": 1432
+    },
+    {
+      "epoch": 0.36041247484909456,
+      "grad_norm": 0.4532756507396698,
+      "learning_rate": 9.987672369788589e-06,
+      "loss": 0.4123,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3606639839034205,
+      "grad_norm": 0.46058785915374756,
+      "learning_rate": 9.987569468191674e-06,
+      "loss": 0.4172,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3609154929577465,
+      "grad_norm": 0.448282390832901,
+      "learning_rate": 9.987466139440636e-06,
+      "loss": 0.393,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3611670020120724,
+      "grad_norm": 0.4629223644733429,
+      "learning_rate": 9.987362383544326e-06,
+      "loss": 0.4136,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3614185110663984,
+      "grad_norm": 0.44559571146965027,
+      "learning_rate": 9.987258200511627e-06,
+      "loss": 0.4002,
+      "step": 1437
+    },
+    {
+      "epoch": 0.36167002012072436,
+      "grad_norm": 0.42130130529403687,
+      "learning_rate": 9.987153590351466e-06,
+      "loss": 0.4134,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3619215291750503,
+      "grad_norm": 0.4000697731971741,
+      "learning_rate": 9.987048553072796e-06,
+      "loss": 0.3989,
+      "step": 1439
+    },
+    {
+      "epoch": 0.36217303822937624,
+      "grad_norm": 0.42225921154022217,
+      "learning_rate": 9.986943088684619e-06,
+      "loss": 0.4023,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3624245472837022,
+      "grad_norm": 0.39401090145111084,
+      "learning_rate": 9.986837197195964e-06,
+      "loss": 0.4141,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3626760563380282,
+      "grad_norm": 0.41057607531547546,
+      "learning_rate": 9.9867308786159e-06,
+      "loss": 0.4282,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3629275653923541,
+      "grad_norm": 0.40181800723075867,
+      "learning_rate": 9.986624132953533e-06,
+      "loss": 0.3689,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3631790744466801,
+      "grad_norm": 0.4301682710647583,
+      "learning_rate": 9.986516960218005e-06,
+      "loss": 0.4181,
+      "step": 1444
+    },
+    {
+      "epoch": 0.36343058350100604,
+      "grad_norm": 0.40248462557792664,
+      "learning_rate": 9.986409360418497e-06,
+      "loss": 0.4252,
+      "step": 1445
+    },
+    {
+      "epoch": 0.363682092555332,
+      "grad_norm": 0.451557993888855,
+      "learning_rate": 9.98630133356422e-06,
+      "loss": 0.4201,
+      "step": 1446
+    },
+    {
+      "epoch": 0.36393360160965793,
+      "grad_norm": 0.4034324586391449,
+      "learning_rate": 9.986192879664428e-06,
+      "loss": 0.4109,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3641851106639839,
+      "grad_norm": 0.41685450077056885,
+      "learning_rate": 9.98608399872841e-06,
+      "loss": 0.4119,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3644366197183099,
+      "grad_norm": 0.4460361897945404,
+      "learning_rate": 9.985974690765492e-06,
+      "loss": 0.4037,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3646881287726358,
+      "grad_norm": 0.42670106887817383,
+      "learning_rate": 9.985864955785032e-06,
+      "loss": 0.3833,
+      "step": 1450
+    },
+    {
+      "epoch": 0.36493963782696176,
+      "grad_norm": 0.4803377091884613,
+      "learning_rate": 9.98575479379643e-06,
+      "loss": 0.3959,
+      "step": 1451
+    },
+    {
+      "epoch": 0.36519114688128773,
+      "grad_norm": 0.4145619869232178,
+      "learning_rate": 9.98564420480912e-06,
+      "loss": 0.3964,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3654426559356137,
+      "grad_norm": 0.45921090245246887,
+      "learning_rate": 9.985533188832575e-06,
+      "loss": 0.3993,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3656941649899396,
+      "grad_norm": 0.39491912722587585,
+      "learning_rate": 9.985421745876302e-06,
+      "loss": 0.3956,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3659456740442656,
+      "grad_norm": 0.4830361306667328,
+      "learning_rate": 9.985309875949844e-06,
+      "loss": 0.4028,
+      "step": 1455
+    },
+    {
+      "epoch": 0.36619718309859156,
+      "grad_norm": 0.42572900652885437,
+      "learning_rate": 9.985197579062784e-06,
+      "loss": 0.422,
+      "step": 1456
+    },
+    {
+      "epoch": 0.36644869215291753,
+      "grad_norm": 0.4604795277118683,
+      "learning_rate": 9.98508485522474e-06,
+      "loss": 0.4295,
+      "step": 1457
+    },
+    {
+      "epoch": 0.36670020120724345,
+      "grad_norm": 0.4539150297641754,
+      "learning_rate": 9.984971704445363e-06,
+      "loss": 0.3922,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3669517102615694,
+      "grad_norm": 0.4076423645019531,
+      "learning_rate": 9.984858126734345e-06,
+      "loss": 0.412,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3672032193158954,
+      "grad_norm": 0.4578339457511902,
+      "learning_rate": 9.984744122101415e-06,
+      "loss": 0.4124,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3674547283702213,
+      "grad_norm": 0.43359407782554626,
+      "learning_rate": 9.984629690556336e-06,
+      "loss": 0.4002,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3677062374245473,
+      "grad_norm": 0.4007089138031006,
+      "learning_rate": 9.984514832108905e-06,
+      "loss": 0.413,
+      "step": 1462
+    },
+    {
+      "epoch": 0.36795774647887325,
+      "grad_norm": 0.4894416928291321,
+      "learning_rate": 9.984399546768964e-06,
+      "loss": 0.4383,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3682092555331992,
+      "grad_norm": 0.45550626516342163,
+      "learning_rate": 9.984283834546383e-06,
+      "loss": 0.4012,
+      "step": 1464
+    },
+    {
+      "epoch": 0.36846076458752514,
+      "grad_norm": 0.4249536991119385,
+      "learning_rate": 9.984167695451075e-06,
+      "loss": 0.4238,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3687122736418511,
+      "grad_norm": 0.43250423669815063,
+      "learning_rate": 9.984051129492982e-06,
+      "loss": 0.4172,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3689637826961771,
+      "grad_norm": 0.42905688285827637,
+      "learning_rate": 9.983934136682092e-06,
+      "loss": 0.4134,
+      "step": 1467
+    },
+    {
+      "epoch": 0.369215291750503,
+      "grad_norm": 0.4656728208065033,
+      "learning_rate": 9.983816717028421e-06,
+      "loss": 0.4207,
+      "step": 1468
+    },
+    {
+      "epoch": 0.36946680080482897,
+      "grad_norm": 0.3909688889980316,
+      "learning_rate": 9.983698870542028e-06,
+      "loss": 0.4257,
+      "step": 1469
+    },
+    {
+      "epoch": 0.36971830985915494,
+      "grad_norm": 0.49250149726867676,
+      "learning_rate": 9.983580597233005e-06,
+      "loss": 0.3796,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3699698189134809,
+      "grad_norm": 0.42419931292533875,
+      "learning_rate": 9.98346189711148e-06,
+      "loss": 0.4042,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3702213279678068,
+      "grad_norm": 0.41211193799972534,
+      "learning_rate": 9.98334277018762e-06,
+      "loss": 0.3932,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3704728370221328,
+      "grad_norm": 0.41742807626724243,
+      "learning_rate": 9.983223216471627e-06,
+      "loss": 0.3787,
+      "step": 1473
+    },
+    {
+      "epoch": 0.37072434607645877,
+      "grad_norm": 0.42407408356666565,
+      "learning_rate": 9.98310323597374e-06,
+      "loss": 0.4117,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3709758551307847,
+      "grad_norm": 0.469777375459671,
+      "learning_rate": 9.982982828704237e-06,
+      "loss": 0.4397,
+      "step": 1475
+    },
+    {
+      "epoch": 0.37122736418511065,
+      "grad_norm": 0.4012710154056549,
+      "learning_rate": 9.982861994673427e-06,
+      "loss": 0.3935,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3714788732394366,
+      "grad_norm": 0.5193198323249817,
+      "learning_rate": 9.982740733891661e-06,
+      "loss": 0.3994,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3717303822937626,
+      "grad_norm": 0.4547984004020691,
+      "learning_rate": 9.982619046369321e-06,
+      "loss": 0.3832,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3719818913480885,
+      "grad_norm": 0.43693944811820984,
+      "learning_rate": 9.982496932116835e-06,
+      "loss": 0.3778,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3722334004024145,
+      "grad_norm": 0.4877944588661194,
+      "learning_rate": 9.982374391144653e-06,
+      "loss": 0.4425,
+      "step": 1480
+    },
+    {
+      "epoch": 0.37248490945674045,
+      "grad_norm": 0.4946407973766327,
+      "learning_rate": 9.982251423463275e-06,
+      "loss": 0.407,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3727364185110664,
+      "grad_norm": 0.4377669394016266,
+      "learning_rate": 9.98212802908323e-06,
+      "loss": 0.3936,
+      "step": 1482
+    },
+    {
+      "epoch": 0.37298792756539234,
+      "grad_norm": 0.42440077662467957,
+      "learning_rate": 9.982004208015091e-06,
+      "loss": 0.4186,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3732394366197183,
+      "grad_norm": 0.5275130867958069,
+      "learning_rate": 9.981879960269458e-06,
+      "loss": 0.4257,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3734909456740443,
+      "grad_norm": 0.5148937106132507,
+      "learning_rate": 9.98175528585697e-06,
+      "loss": 0.398,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3737424547283702,
+      "grad_norm": 0.44241297245025635,
+      "learning_rate": 9.981630184788311e-06,
+      "loss": 0.4011,
+      "step": 1486
+    },
+    {
+      "epoch": 0.37399396378269617,
+      "grad_norm": 0.4566717743873596,
+      "learning_rate": 9.98150465707419e-06,
+      "loss": 0.4122,
+      "step": 1487
+    },
+    {
+      "epoch": 0.37424547283702214,
+      "grad_norm": 0.5034930109977722,
+      "learning_rate": 9.981378702725359e-06,
+      "loss": 0.3888,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3744969818913481,
+      "grad_norm": 0.41092514991760254,
+      "learning_rate": 9.981252321752606e-06,
+      "loss": 0.4067,
+      "step": 1489
+    },
+    {
+      "epoch": 0.37474849094567403,
+      "grad_norm": 0.47386881709098816,
+      "learning_rate": 9.981125514166755e-06,
+      "loss": 0.4153,
+      "step": 1490
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.4686683714389801,
+      "learning_rate": 9.980998279978664e-06,
+      "loss": 0.4279,
+      "step": 1491
+    },
+    {
+      "epoch": 0.37525150905432597,
+      "grad_norm": 0.4030835032463074,
+      "learning_rate": 9.980870619199232e-06,
+      "loss": 0.4043,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3755030181086519,
+      "grad_norm": 0.39192453026771545,
+      "learning_rate": 9.980742531839393e-06,
+      "loss": 0.3521,
+      "step": 1493
+    },
+    {
+      "epoch": 0.37575452716297786,
+      "grad_norm": 0.42179104685783386,
+      "learning_rate": 9.980614017910112e-06,
+      "loss": 0.3976,
+      "step": 1494
+    },
+    {
+      "epoch": 0.37600603621730383,
+      "grad_norm": 0.4499700665473938,
+      "learning_rate": 9.980485077422404e-06,
+      "loss": 0.4033,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3762575452716298,
+      "grad_norm": 0.40622663497924805,
+      "learning_rate": 9.980355710387304e-06,
+      "loss": 0.3914,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3765090543259557,
+      "grad_norm": 0.441628098487854,
+      "learning_rate": 9.980225916815894e-06,
+      "loss": 0.41,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3767605633802817,
+      "grad_norm": 0.4532158672809601,
+      "learning_rate": 9.980095696719291e-06,
+      "loss": 0.4123,
+      "step": 1498
+    },
+    {
+      "epoch": 0.37701207243460766,
+      "grad_norm": 0.3944168984889984,
+      "learning_rate": 9.979965050108648e-06,
+      "loss": 0.3929,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3772635814889336,
+      "grad_norm": 0.41683652997016907,
+      "learning_rate": 9.979833976995153e-06,
+      "loss": 0.3946,
+      "step": 1500
+    },
+    {
+      "epoch": 0.37751509054325955,
+      "grad_norm": 0.4310116469860077,
+      "learning_rate": 9.97970247739003e-06,
+      "loss": 0.4022,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3777665995975855,
+      "grad_norm": 0.44122573733329773,
+      "learning_rate": 9.979570551304543e-06,
+      "loss": 0.403,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3780181086519115,
+      "grad_norm": 0.4259469211101532,
+      "learning_rate": 9.979438198749991e-06,
+      "loss": 0.3935,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3782696177062374,
+      "grad_norm": 0.4257745146751404,
+      "learning_rate": 9.979305419737709e-06,
+      "loss": 0.4151,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3785211267605634,
+      "grad_norm": 0.44491034746170044,
+      "learning_rate": 9.979172214279067e-06,
+      "loss": 0.4164,
+      "step": 1505
+    },
+    {
+      "epoch": 0.37877263581488935,
+      "grad_norm": 0.40862441062927246,
+      "learning_rate": 9.979038582385475e-06,
+      "loss": 0.3876,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3790241448692153,
+      "grad_norm": 0.3930971026420593,
+      "learning_rate": 9.978904524068378e-06,
+      "loss": 0.3938,
+      "step": 1507
+    },
+    {
+      "epoch": 0.37927565392354123,
+      "grad_norm": 0.3884839117527008,
+      "learning_rate": 9.978770039339256e-06,
+      "loss": 0.3696,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3795271629778672,
+      "grad_norm": 0.43522050976753235,
+      "learning_rate": 9.978635128209626e-06,
+      "loss": 0.4352,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3797786720321932,
+      "grad_norm": 0.3984873294830322,
+      "learning_rate": 9.978499790691045e-06,
+      "loss": 0.3988,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3800301810865191,
+      "grad_norm": 0.4236599802970886,
+      "learning_rate": 9.978364026795102e-06,
+      "loss": 0.4121,
+      "step": 1511
+    },
+    {
+      "epoch": 0.38028169014084506,
+      "grad_norm": 0.40510380268096924,
+      "learning_rate": 9.978227836533424e-06,
+      "loss": 0.416,
+      "step": 1512
+    },
+    {
+      "epoch": 0.38053319919517103,
+      "grad_norm": 0.5215635299682617,
+      "learning_rate": 9.978091219917675e-06,
+      "loss": 0.419,
+      "step": 1513
+    },
+    {
+      "epoch": 0.380784708249497,
+      "grad_norm": 0.4424278140068054,
+      "learning_rate": 9.977954176959558e-06,
+      "loss": 0.4188,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3810362173038229,
+      "grad_norm": 0.40322160720825195,
+      "learning_rate": 9.977816707670806e-06,
+      "loss": 0.4067,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3812877263581489,
+      "grad_norm": 0.4065972864627838,
+      "learning_rate": 9.977678812063195e-06,
+      "loss": 0.3698,
+      "step": 1516
+    },
+    {
+      "epoch": 0.38153923541247486,
+      "grad_norm": 0.47318071126937866,
+      "learning_rate": 9.977540490148534e-06,
+      "loss": 0.3954,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3817907444668008,
+      "grad_norm": 0.4260292947292328,
+      "learning_rate": 9.977401741938667e-06,
+      "loss": 0.4078,
+      "step": 1518
+    },
+    {
+      "epoch": 0.38204225352112675,
+      "grad_norm": 0.5069632530212402,
+      "learning_rate": 9.977262567445482e-06,
+      "loss": 0.4108,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3822937625754527,
+      "grad_norm": 0.42806974053382874,
+      "learning_rate": 9.977122966680896e-06,
+      "loss": 0.4037,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3825452716297787,
+      "grad_norm": 0.42634284496307373,
+      "learning_rate": 9.976982939656866e-06,
+      "loss": 0.3896,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3827967806841046,
+      "grad_norm": 0.4124469757080078,
+      "learning_rate": 9.976842486385379e-06,
+      "loss": 0.3939,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3830482897384306,
+      "grad_norm": 0.43972286581993103,
+      "learning_rate": 9.976701606878471e-06,
+      "loss": 0.4143,
+      "step": 1523
+    },
+    {
+      "epoch": 0.38329979879275655,
+      "grad_norm": 0.44509053230285645,
+      "learning_rate": 9.976560301148203e-06,
+      "loss": 0.4016,
+      "step": 1524
+    },
+    {
+      "epoch": 0.38355130784708247,
+      "grad_norm": 0.4027416408061981,
+      "learning_rate": 9.976418569206678e-06,
+      "loss": 0.3824,
+      "step": 1525
+    },
+    {
+      "epoch": 0.38380281690140844,
+      "grad_norm": 0.4699016511440277,
+      "learning_rate": 9.976276411066037e-06,
+      "loss": 0.391,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3840543259557344,
+      "grad_norm": 0.44244781136512756,
+      "learning_rate": 9.976133826738452e-06,
+      "loss": 0.3873,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3843058350100604,
+      "grad_norm": 0.4619574546813965,
+      "learning_rate": 9.975990816236135e-06,
+      "loss": 0.4172,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3845573440643863,
+      "grad_norm": 0.45410341024398804,
+      "learning_rate": 9.975847379571336e-06,
+      "loss": 0.4099,
+      "step": 1529
+    },
+    {
+      "epoch": 0.38480885311871227,
+      "grad_norm": 0.4313986003398895,
+      "learning_rate": 9.975703516756334e-06,
+      "loss": 0.4118,
+      "step": 1530
+    },
+    {
+      "epoch": 0.38506036217303824,
+      "grad_norm": 0.4216327965259552,
+      "learning_rate": 9.975559227803458e-06,
+      "loss": 0.3965,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3853118712273642,
+      "grad_norm": 0.46033933758735657,
+      "learning_rate": 9.975414512725058e-06,
+      "loss": 0.3729,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3855633802816901,
+      "grad_norm": 0.42646700143814087,
+      "learning_rate": 9.97526937153353e-06,
+      "loss": 0.3818,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3858148893360161,
+      "grad_norm": 0.4267144203186035,
+      "learning_rate": 9.975123804241309e-06,
+      "loss": 0.4088,
+      "step": 1534
+    },
+    {
+      "epoch": 0.38606639839034207,
+      "grad_norm": 0.46120572090148926,
+      "learning_rate": 9.974977810860858e-06,
+      "loss": 0.4,
+      "step": 1535
+    },
+    {
+      "epoch": 0.386317907444668,
+      "grad_norm": 0.41420215368270874,
+      "learning_rate": 9.97483139140468e-06,
+      "loss": 0.4145,
+      "step": 1536
+    },
+    {
+      "epoch": 0.38656941649899396,
+      "grad_norm": 0.514680802822113,
+      "learning_rate": 9.974684545885315e-06,
+      "loss": 0.4021,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3868209255533199,
+      "grad_norm": 0.4189431369304657,
+      "learning_rate": 9.97453727431534e-06,
+      "loss": 0.3973,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3870724346076459,
+      "grad_norm": 0.47960418462753296,
+      "learning_rate": 9.97438957670737e-06,
+      "loss": 0.3942,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3873239436619718,
+      "grad_norm": 0.49763473868370056,
+      "learning_rate": 9.974241453074051e-06,
+      "loss": 0.3891,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3875754527162978,
+      "grad_norm": 0.48309236764907837,
+      "learning_rate": 9.974092903428072e-06,
+      "loss": 0.416,
+      "step": 1541
+    },
+    {
+      "epoch": 0.38782696177062376,
+      "grad_norm": 0.48040205240249634,
+      "learning_rate": 9.973943927782152e-06,
+      "loss": 0.3942,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3880784708249497,
+      "grad_norm": 0.4268017113208771,
+      "learning_rate": 9.973794526149051e-06,
+      "loss": 0.4178,
+      "step": 1543
+    },
+    {
+      "epoch": 0.38832997987927564,
+      "grad_norm": 0.4728141725063324,
+      "learning_rate": 9.973644698541567e-06,
+      "loss": 0.3855,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3885814889336016,
+      "grad_norm": 0.46177420020103455,
+      "learning_rate": 9.973494444972527e-06,
+      "loss": 0.4203,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3888329979879276,
+      "grad_norm": 0.47535744309425354,
+      "learning_rate": 9.973343765454803e-06,
+      "loss": 0.4344,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3890845070422535,
+      "grad_norm": 0.4919487535953522,
+      "learning_rate": 9.973192660001299e-06,
+      "loss": 0.429,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3893360160965795,
+      "grad_norm": 0.40322598814964294,
+      "learning_rate": 9.973041128624956e-06,
+      "loss": 0.4181,
+      "step": 1548
+    },
+    {
+      "epoch": 0.38958752515090544,
+      "grad_norm": 0.538712739944458,
+      "learning_rate": 9.972889171338752e-06,
+      "loss": 0.3997,
+      "step": 1549
+    },
+    {
+      "epoch": 0.38983903420523136,
+      "grad_norm": 0.43118852376937866,
+      "learning_rate": 9.9727367881557e-06,
+      "loss": 0.4037,
+      "step": 1550
+    },
+    {
+      "epoch": 0.39009054325955733,
+      "grad_norm": 0.4399496912956238,
+      "learning_rate": 9.97258397908885e-06,
+      "loss": 0.4109,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3903420523138833,
+      "grad_norm": 0.42758503556251526,
+      "learning_rate": 9.972430744151292e-06,
+      "loss": 0.401,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3905935613682093,
+      "grad_norm": 0.4662076234817505,
+      "learning_rate": 9.97227708335615e-06,
+      "loss": 0.4129,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3908450704225352,
+      "grad_norm": 0.38940244913101196,
+      "learning_rate": 9.97212299671658e-06,
+      "loss": 0.3821,
+      "step": 1554
+    },
+    {
+      "epoch": 0.39109657947686116,
+      "grad_norm": 0.4017338156700134,
+      "learning_rate": 9.97196848424578e-06,
+      "loss": 0.4161,
+      "step": 1555
+    },
+    {
+      "epoch": 0.39134808853118713,
+      "grad_norm": 0.4648374319076538,
+      "learning_rate": 9.971813545956986e-06,
+      "loss": 0.3979,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3915995975855131,
+      "grad_norm": 0.3893650472164154,
+      "learning_rate": 9.971658181863464e-06,
+      "loss": 0.3949,
+      "step": 1557
+    },
+    {
+      "epoch": 0.391851106639839,
+      "grad_norm": 0.47235551476478577,
+      "learning_rate": 9.971502391978523e-06,
+      "loss": 0.4114,
+      "step": 1558
+    },
+    {
+      "epoch": 0.392102615694165,
+      "grad_norm": 0.5151132941246033,
+      "learning_rate": 9.971346176315501e-06,
+      "loss": 0.4216,
+      "step": 1559
+    },
+    {
+      "epoch": 0.39235412474849096,
+      "grad_norm": 0.36058980226516724,
+      "learning_rate": 9.971189534887781e-06,
+      "loss": 0.3828,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3926056338028169,
+      "grad_norm": 0.5667259693145752,
+      "learning_rate": 9.971032467708779e-06,
+      "loss": 0.3984,
+      "step": 1561
+    },
+    {
+      "epoch": 0.39285714285714285,
+      "grad_norm": 0.4725882112979889,
+      "learning_rate": 9.970874974791942e-06,
+      "loss": 0.4021,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3931086519114688,
+      "grad_norm": 0.4235588014125824,
+      "learning_rate": 9.970717056150764e-06,
+      "loss": 0.3842,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3933601609657948,
+      "grad_norm": 0.5390638709068298,
+      "learning_rate": 9.970558711798763e-06,
+      "loss": 0.4046,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3936116700201207,
+      "grad_norm": 0.5337787866592407,
+      "learning_rate": 9.970399941749507e-06,
+      "loss": 0.3996,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3938631790744467,
+      "grad_norm": 0.4832894504070282,
+      "learning_rate": 9.970240746016588e-06,
+      "loss": 0.3957,
+      "step": 1566
+    },
+    {
+      "epoch": 0.39411468812877265,
+      "grad_norm": 0.5507562160491943,
+      "learning_rate": 9.970081124613647e-06,
+      "loss": 0.395,
+      "step": 1567
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 0.4837891459465027,
+      "learning_rate": 9.969921077554347e-06,
+      "loss": 0.3878,
+      "step": 1568
+    },
+    {
+      "epoch": 0.39461770623742454,
+      "grad_norm": 0.4195319414138794,
+      "learning_rate": 9.969760604852399e-06,
+      "loss": 0.4046,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3948692152917505,
+      "grad_norm": 0.5062696933746338,
+      "learning_rate": 9.969599706521546e-06,
+      "loss": 0.4181,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3951207243460765,
+      "grad_norm": 0.5001570582389832,
+      "learning_rate": 9.969438382575569e-06,
+      "loss": 0.3919,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3953722334004024,
+      "grad_norm": 0.47448521852493286,
+      "learning_rate": 9.969276633028281e-06,
+      "loss": 0.3904,
+      "step": 1572
+    },
+    {
+      "epoch": 0.39562374245472837,
+      "grad_norm": 0.5753097534179688,
+      "learning_rate": 9.96911445789354e-06,
+      "loss": 0.4031,
+      "step": 1573
+    },
+    {
+      "epoch": 0.39587525150905434,
+      "grad_norm": 0.4995958209037781,
+      "learning_rate": 9.96895185718523e-06,
+      "loss": 0.4172,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3961267605633803,
+      "grad_norm": 0.42421847581863403,
+      "learning_rate": 9.96878883091728e-06,
+      "loss": 0.3958,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3963782696177062,
+      "grad_norm": 0.42944973707199097,
+      "learning_rate": 9.968625379103651e-06,
+      "loss": 0.4098,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3966297786720322,
+      "grad_norm": 0.5947927832603455,
+      "learning_rate": 9.968461501758343e-06,
+      "loss": 0.4095,
+      "step": 1577
+    },
+    {
+      "epoch": 0.39688128772635817,
+      "grad_norm": 0.43222859501838684,
+      "learning_rate": 9.968297198895388e-06,
+      "loss": 0.4088,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3971327967806841,
+      "grad_norm": 0.4837366044521332,
+      "learning_rate": 9.968132470528862e-06,
+      "loss": 0.3992,
+      "step": 1579
+    },
+    {
+      "epoch": 0.39738430583501005,
+      "grad_norm": 0.46640220284461975,
+      "learning_rate": 9.967967316672869e-06,
+      "loss": 0.3845,
+      "step": 1580
+    },
+    {
+      "epoch": 0.397635814889336,
+      "grad_norm": 0.39366641640663147,
+      "learning_rate": 9.967801737341556e-06,
+      "loss": 0.4074,
+      "step": 1581
+    },
+    {
+      "epoch": 0.397887323943662,
+      "grad_norm": 0.4720052182674408,
+      "learning_rate": 9.9676357325491e-06,
+      "loss": 0.4037,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3981388329979879,
+      "grad_norm": 0.43749743700027466,
+      "learning_rate": 9.967469302309722e-06,
+      "loss": 0.4145,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3983903420523139,
+      "grad_norm": 0.39652055501937866,
+      "learning_rate": 9.967302446637677e-06,
+      "loss": 0.4018,
+      "step": 1584
+    },
+    {
+      "epoch": 0.39864185110663986,
+      "grad_norm": 0.4925057291984558,
+      "learning_rate": 9.96713516554725e-06,
+      "loss": 0.4028,
+      "step": 1585
+    },
+    {
+      "epoch": 0.39889336016096577,
+      "grad_norm": 0.448711633682251,
+      "learning_rate": 9.966967459052771e-06,
+      "loss": 0.4107,
+      "step": 1586
+    },
+    {
+      "epoch": 0.39914486921529174,
+      "grad_norm": 0.3823001980781555,
+      "learning_rate": 9.966799327168603e-06,
+      "loss": 0.4128,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3993963782696177,
+      "grad_norm": 0.47937512397766113,
+      "learning_rate": 9.966630769909145e-06,
+      "loss": 0.4312,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3996478873239437,
+      "grad_norm": 0.4800701439380646,
+      "learning_rate": 9.966461787288832e-06,
+      "loss": 0.4116,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3998993963782696,
+      "grad_norm": 0.4551905691623688,
+      "learning_rate": 9.966292379322138e-06,
+      "loss": 0.4165,
+      "step": 1590
+    },
+    {
+      "epoch": 0.40015090543259557,
+      "grad_norm": 0.4783093333244324,
+      "learning_rate": 9.966122546023568e-06,
+      "loss": 0.3833,
+      "step": 1591
+    },
+    {
+      "epoch": 0.40040241448692154,
+      "grad_norm": 0.4122217893600464,
+      "learning_rate": 9.965952287407674e-06,
+      "loss": 0.3951,
+      "step": 1592
+    },
+    {
+      "epoch": 0.40065392354124746,
+      "grad_norm": 0.4596366584300995,
+      "learning_rate": 9.965781603489032e-06,
+      "loss": 0.3803,
+      "step": 1593
+    },
+    {
+      "epoch": 0.40090543259557343,
+      "grad_norm": 0.4768621027469635,
+      "learning_rate": 9.96561049428226e-06,
+      "loss": 0.3958,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4011569416498994,
+      "grad_norm": 0.45015203952789307,
+      "learning_rate": 9.965438959802015e-06,
+      "loss": 0.3932,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4014084507042254,
+      "grad_norm": 0.44677606225013733,
+      "learning_rate": 9.965267000062986e-06,
+      "loss": 0.4062,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4016599597585513,
+      "grad_norm": 0.45594921708106995,
+      "learning_rate": 9.965094615079902e-06,
+      "loss": 0.3921,
+      "step": 1597
+    },
+    {
+      "epoch": 0.40191146881287726,
+      "grad_norm": 0.44499877095222473,
+      "learning_rate": 9.964921804867524e-06,
+      "loss": 0.4086,
+      "step": 1598
+    },
+    {
+      "epoch": 0.40216297786720323,
+      "grad_norm": 0.45410072803497314,
+      "learning_rate": 9.964748569440656e-06,
+      "loss": 0.4115,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4024144869215292,
+      "grad_norm": 0.42579391598701477,
+      "learning_rate": 9.964574908814131e-06,
+      "loss": 0.4072,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4026659959758551,
+      "grad_norm": 0.42585766315460205,
+      "learning_rate": 9.964400823002825e-06,
+      "loss": 0.3703,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4029175050301811,
+      "grad_norm": 0.4478361904621124,
+      "learning_rate": 9.964226312021645e-06,
+      "loss": 0.4314,
+      "step": 1602
+    },
+    {
+      "epoch": 0.40316901408450706,
+      "grad_norm": 0.3969224691390991,
+      "learning_rate": 9.964051375885537e-06,
+      "loss": 0.3911,
+      "step": 1603
+    },
+    {
+      "epoch": 0.403420523138833,
+      "grad_norm": 0.4182923138141632,
+      "learning_rate": 9.963876014609484e-06,
+      "loss": 0.4009,
+      "step": 1604
+    },
+    {
+      "epoch": 0.40367203219315895,
+      "grad_norm": 0.43487611413002014,
+      "learning_rate": 9.963700228208503e-06,
+      "loss": 0.3828,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4039235412474849,
+      "grad_norm": 0.4310348927974701,
+      "learning_rate": 9.963524016697651e-06,
+      "loss": 0.4142,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4041750503018109,
+      "grad_norm": 0.4118841588497162,
+      "learning_rate": 9.96334738009202e-06,
+      "loss": 0.4051,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4044265593561368,
+      "grad_norm": 0.4341716170310974,
+      "learning_rate": 9.963170318406737e-06,
+      "loss": 0.4171,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4046780684104628,
+      "grad_norm": 0.4583278298377991,
+      "learning_rate": 9.962992831656964e-06,
+      "loss": 0.4026,
+      "step": 1609
+    },
+    {
+      "epoch": 0.40492957746478875,
+      "grad_norm": 0.38871103525161743,
+      "learning_rate": 9.962814919857903e-06,
+      "loss": 0.3959,
+      "step": 1610
+    },
+    {
+      "epoch": 0.40518108651911466,
+      "grad_norm": 0.4146328866481781,
+      "learning_rate": 9.962636583024792e-06,
+      "loss": 0.3994,
+      "step": 1611
+    },
+    {
+      "epoch": 0.40543259557344064,
+      "grad_norm": 0.371489554643631,
+      "learning_rate": 9.962457821172903e-06,
+      "loss": 0.3671,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4056841046277666,
+      "grad_norm": 0.4303867816925049,
+      "learning_rate": 9.962278634317549e-06,
+      "loss": 0.3833,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4059356136820926,
+      "grad_norm": 0.3755389451980591,
+      "learning_rate": 9.962099022474072e-06,
+      "loss": 0.4061,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4061871227364185,
+      "grad_norm": 0.4392898678779602,
+      "learning_rate": 9.961918985657857e-06,
+      "loss": 0.4042,
+      "step": 1615
+    },
+    {
+      "epoch": 0.40643863179074446,
+      "grad_norm": 0.4400614798069,
+      "learning_rate": 9.961738523884322e-06,
+      "loss": 0.4054,
+      "step": 1616
+    },
+    {
+      "epoch": 0.40669014084507044,
+      "grad_norm": 0.4476352035999298,
+      "learning_rate": 9.961557637168924e-06,
+      "loss": 0.416,
+      "step": 1617
+    },
+    {
+      "epoch": 0.40694164989939635,
+      "grad_norm": 0.4568561613559723,
+      "learning_rate": 9.961376325527152e-06,
+      "loss": 0.3866,
+      "step": 1618
+    },
+    {
+      "epoch": 0.4071931589537223,
+      "grad_norm": 0.40012219548225403,
+      "learning_rate": 9.961194588974538e-06,
+      "loss": 0.3919,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4074446680080483,
+      "grad_norm": 0.4288276135921478,
+      "learning_rate": 9.961012427526644e-06,
+      "loss": 0.3801,
+      "step": 1620
+    },
+    {
+      "epoch": 0.40769617706237427,
+      "grad_norm": 0.4316503703594208,
+      "learning_rate": 9.960829841199071e-06,
+      "loss": 0.4102,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4079476861167002,
+      "grad_norm": 0.4526152014732361,
+      "learning_rate": 9.960646830007456e-06,
+      "loss": 0.4437,
+      "step": 1622
+    },
+    {
+      "epoch": 0.40819919517102615,
+      "grad_norm": 0.4108959138393402,
+      "learning_rate": 9.960463393967476e-06,
+      "loss": 0.3752,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4084507042253521,
+      "grad_norm": 0.4516206383705139,
+      "learning_rate": 9.960279533094838e-06,
+      "loss": 0.4061,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4087022132796781,
+      "grad_norm": 0.46021711826324463,
+      "learning_rate": 9.96009524740529e-06,
+      "loss": 0.4302,
+      "step": 1625
+    },
+    {
+      "epoch": 0.408953722334004,
+      "grad_norm": 0.3884812295436859,
+      "learning_rate": 9.959910536914614e-06,
+      "loss": 0.429,
+      "step": 1626
+    },
+    {
+      "epoch": 0.40920523138833,
+      "grad_norm": 0.4301206171512604,
+      "learning_rate": 9.95972540163863e-06,
+      "loss": 0.4074,
+      "step": 1627
+    },
+    {
+      "epoch": 0.40945674044265595,
+      "grad_norm": 0.4203178286552429,
+      "learning_rate": 9.959539841593192e-06,
+      "loss": 0.3939,
+      "step": 1628
+    },
+    {
+      "epoch": 0.40970824949698187,
+      "grad_norm": 0.4511706233024597,
+      "learning_rate": 9.959353856794194e-06,
+      "loss": 0.4382,
+      "step": 1629
+    },
+    {
+      "epoch": 0.40995975855130784,
+      "grad_norm": 0.42685285210609436,
+      "learning_rate": 9.959167447257563e-06,
+      "loss": 0.3805,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4102112676056338,
+      "grad_norm": 0.4132843315601349,
+      "learning_rate": 9.958980612999265e-06,
+      "loss": 0.3986,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4104627766599598,
+      "grad_norm": 0.3776272237300873,
+      "learning_rate": 9.9587933540353e-06,
+      "loss": 0.3944,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4107142857142857,
+      "grad_norm": 0.5198211073875427,
+      "learning_rate": 9.958605670381709e-06,
+      "loss": 0.3953,
+      "step": 1633
+    },
+    {
+      "epoch": 0.41096579476861167,
+      "grad_norm": 0.4831434488296509,
+      "learning_rate": 9.958417562054561e-06,
+      "loss": 0.4026,
+      "step": 1634
+    },
+    {
+      "epoch": 0.41121730382293764,
+      "grad_norm": 0.43184301257133484,
+      "learning_rate": 9.958229029069969e-06,
+      "loss": 0.4047,
+      "step": 1635
+    },
+    {
+      "epoch": 0.41146881287726356,
+      "grad_norm": 0.46622034907341003,
+      "learning_rate": 9.95804007144408e-06,
+      "loss": 0.3992,
+      "step": 1636
+    },
+    {
+      "epoch": 0.41172032193158953,
+      "grad_norm": 0.4247199296951294,
+      "learning_rate": 9.957850689193075e-06,
+      "loss": 0.4125,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4119718309859155,
+      "grad_norm": 0.4305685758590698,
+      "learning_rate": 9.957660882333176e-06,
+      "loss": 0.3811,
+      "step": 1638
+    },
+    {
+      "epoch": 0.41222334004024147,
+      "grad_norm": 0.45092910528182983,
+      "learning_rate": 9.957470650880636e-06,
+      "loss": 0.3988,
+      "step": 1639
+    },
+    {
+      "epoch": 0.4124748490945674,
+      "grad_norm": 0.40285438299179077,
+      "learning_rate": 9.957279994851751e-06,
+      "loss": 0.3985,
+      "step": 1640
+    },
+    {
+      "epoch": 0.41272635814889336,
+      "grad_norm": 0.3911204934120178,
+      "learning_rate": 9.957088914262844e-06,
+      "loss": 0.4334,
+      "step": 1641
+    },
+    {
+      "epoch": 0.41297786720321933,
+      "grad_norm": 0.4495960474014282,
+      "learning_rate": 9.956897409130286e-06,
+      "loss": 0.4241,
+      "step": 1642
+    },
+    {
+      "epoch": 0.41322937625754524,
+      "grad_norm": 0.4449693560600281,
+      "learning_rate": 9.956705479470473e-06,
+      "loss": 0.4062,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4134808853118712,
+      "grad_norm": 0.4287528097629547,
+      "learning_rate": 9.956513125299847e-06,
+      "loss": 0.3907,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4137323943661972,
+      "grad_norm": 0.40359970927238464,
+      "learning_rate": 9.956320346634877e-06,
+      "loss": 0.4134,
+      "step": 1645
+    },
+    {
+      "epoch": 0.41398390342052316,
+      "grad_norm": 0.48502546548843384,
+      "learning_rate": 9.956127143492077e-06,
+      "loss": 0.3796,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4142354124748491,
+      "grad_norm": 0.4318941533565521,
+      "learning_rate": 9.955933515887992e-06,
+      "loss": 0.4273,
+      "step": 1647
+    },
+    {
+      "epoch": 0.41448692152917505,
+      "grad_norm": 0.4548565447330475,
+      "learning_rate": 9.955739463839206e-06,
+      "loss": 0.4136,
+      "step": 1648
+    },
+    {
+      "epoch": 0.414738430583501,
+      "grad_norm": 0.4548591077327728,
+      "learning_rate": 9.955544987362339e-06,
+      "loss": 0.4114,
+      "step": 1649
+    },
+    {
+      "epoch": 0.414989939637827,
+      "grad_norm": 0.4239116311073303,
+      "learning_rate": 9.955350086474045e-06,
+      "loss": 0.4106,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4152414486921529,
+      "grad_norm": 0.40276509523391724,
+      "learning_rate": 9.955154761191017e-06,
+      "loss": 0.3762,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4154929577464789,
+      "grad_norm": 0.513320803642273,
+      "learning_rate": 9.954959011529982e-06,
+      "loss": 0.4079,
+      "step": 1652
+    },
+    {
+      "epoch": 0.41574446680080485,
+      "grad_norm": 0.42079389095306396,
+      "learning_rate": 9.954762837507705e-06,
+      "loss": 0.3759,
+      "step": 1653
+    },
+    {
+      "epoch": 0.41599597585513076,
+      "grad_norm": 0.4646183252334595,
+      "learning_rate": 9.95456623914099e-06,
+      "loss": 0.4014,
+      "step": 1654
+    },
+    {
+      "epoch": 0.41624748490945673,
+      "grad_norm": 0.49453848600387573,
+      "learning_rate": 9.954369216446672e-06,
+      "loss": 0.4085,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4164989939637827,
+      "grad_norm": 0.4750482738018036,
+      "learning_rate": 9.954171769441625e-06,
+      "loss": 0.3955,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4167505030181087,
+      "grad_norm": 0.4437958598136902,
+      "learning_rate": 9.953973898142759e-06,
+      "loss": 0.3681,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4170020120724346,
+      "grad_norm": 0.4792766571044922,
+      "learning_rate": 9.953775602567019e-06,
+      "loss": 0.3918,
+      "step": 1658
+    },
+    {
+      "epoch": 0.41725352112676056,
+      "grad_norm": 0.5086737275123596,
+      "learning_rate": 9.95357688273139e-06,
+      "loss": 0.3897,
+      "step": 1659
+    },
+    {
+      "epoch": 0.41750503018108653,
+      "grad_norm": 0.4676213264465332,
+      "learning_rate": 9.953377738652892e-06,
+      "loss": 0.3902,
+      "step": 1660
+    },
+    {
+      "epoch": 0.41775653923541245,
+      "grad_norm": 0.5371580719947815,
+      "learning_rate": 9.953178170348578e-06,
+      "loss": 0.4179,
+      "step": 1661
+    },
+    {
+      "epoch": 0.4180080482897384,
+      "grad_norm": 0.4958730638027191,
+      "learning_rate": 9.952978177835542e-06,
+      "loss": 0.3896,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4182595573440644,
+      "grad_norm": 0.43639904260635376,
+      "learning_rate": 9.952777761130909e-06,
+      "loss": 0.3949,
+      "step": 1663
+    },
+    {
+      "epoch": 0.41851106639839036,
+      "grad_norm": 0.43316468596458435,
+      "learning_rate": 9.952576920251845e-06,
+      "loss": 0.3852,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4187625754527163,
+      "grad_norm": 0.3840535879135132,
+      "learning_rate": 9.952375655215551e-06,
+      "loss": 0.3757,
+      "step": 1665
+    },
+    {
+      "epoch": 0.41901408450704225,
+      "grad_norm": 0.4622785151004791,
+      "learning_rate": 9.952173966039266e-06,
+      "loss": 0.3856,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4192655935613682,
+      "grad_norm": 0.4158206880092621,
+      "learning_rate": 9.951971852740262e-06,
+      "loss": 0.3824,
+      "step": 1667
+    },
+    {
+      "epoch": 0.41951710261569414,
+      "grad_norm": 0.42178934812545776,
+      "learning_rate": 9.951769315335843e-06,
+      "loss": 0.3969,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4197686116700201,
+      "grad_norm": 0.4934524595737457,
+      "learning_rate": 9.951566353843365e-06,
+      "loss": 0.3818,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4200201207243461,
+      "grad_norm": 0.4420875608921051,
+      "learning_rate": 9.951362968280205e-06,
+      "loss": 0.4343,
+      "step": 1670
+    },
+    {
+      "epoch": 0.42027162977867205,
+      "grad_norm": 0.42826712131500244,
+      "learning_rate": 9.951159158663782e-06,
+      "loss": 0.3899,
+      "step": 1671
+    },
+    {
+      "epoch": 0.42052313883299797,
+      "grad_norm": 0.4212586283683777,
+      "learning_rate": 9.950954925011552e-06,
+      "loss": 0.3819,
+      "step": 1672
+    },
+    {
+      "epoch": 0.42077464788732394,
+      "grad_norm": 0.5084717869758606,
+      "learning_rate": 9.950750267341004e-06,
+      "loss": 0.4106,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4210261569416499,
+      "grad_norm": 0.4205119013786316,
+      "learning_rate": 9.950545185669668e-06,
+      "loss": 0.3926,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4212776659959759,
+      "grad_norm": 0.4944150447845459,
+      "learning_rate": 9.95033968001511e-06,
+      "loss": 0.4076,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4215291750503018,
+      "grad_norm": 0.4203665852546692,
+      "learning_rate": 9.950133750394926e-06,
+      "loss": 0.3952,
+      "step": 1676
+    },
+    {
+      "epoch": 0.42178068410462777,
+      "grad_norm": 0.41072648763656616,
+      "learning_rate": 9.949927396826753e-06,
+      "loss": 0.3923,
+      "step": 1677
+    },
+    {
+      "epoch": 0.42203219315895374,
+      "grad_norm": 0.4945848882198334,
+      "learning_rate": 9.949720619328265e-06,
+      "loss": 0.4096,
+      "step": 1678
+    },
+    {
+      "epoch": 0.42228370221327965,
+      "grad_norm": 0.3855612576007843,
+      "learning_rate": 9.949513417917173e-06,
+      "loss": 0.3916,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4225352112676056,
+      "grad_norm": 0.38849636912345886,
+      "learning_rate": 9.94930579261122e-06,
+      "loss": 0.3744,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4227867203219316,
+      "grad_norm": 0.45562243461608887,
+      "learning_rate": 9.949097743428188e-06,
+      "loss": 0.3882,
+      "step": 1681
+    },
+    {
+      "epoch": 0.42303822937625757,
+      "grad_norm": 0.38708576560020447,
+      "learning_rate": 9.948889270385897e-06,
+      "loss": 0.4077,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4232897384305835,
+      "grad_norm": 0.4899725317955017,
+      "learning_rate": 9.948680373502199e-06,
+      "loss": 0.4271,
+      "step": 1683
+    },
+    {
+      "epoch": 0.42354124748490946,
+      "grad_norm": 0.4218765199184418,
+      "learning_rate": 9.948471052794988e-06,
+      "loss": 0.4052,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4237927565392354,
+      "grad_norm": 0.45039162039756775,
+      "learning_rate": 9.948261308282187e-06,
+      "loss": 0.398,
+      "step": 1685
+    },
+    {
+      "epoch": 0.42404426559356134,
+      "grad_norm": 0.3818122148513794,
+      "learning_rate": 9.94805113998176e-06,
+      "loss": 0.4196,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4242957746478873,
+      "grad_norm": 0.4184150695800781,
+      "learning_rate": 9.94784054791171e-06,
+      "loss": 0.3955,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4245472837022133,
+      "grad_norm": 0.4112408459186554,
+      "learning_rate": 9.94762953209007e-06,
+      "loss": 0.4208,
+      "step": 1688
+    },
+    {
+      "epoch": 0.42479879275653926,
+      "grad_norm": 0.44503697752952576,
+      "learning_rate": 9.947418092534912e-06,
+      "loss": 0.3858,
+      "step": 1689
+    },
+    {
+      "epoch": 0.42505030181086517,
+      "grad_norm": 0.42923927307128906,
+      "learning_rate": 9.947206229264346e-06,
+      "loss": 0.3939,
+      "step": 1690
+    },
+    {
+      "epoch": 0.42530181086519114,
+      "grad_norm": 0.4138977825641632,
+      "learning_rate": 9.946993942296517e-06,
+      "loss": 0.3967,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4255533199195171,
+      "grad_norm": 0.4376116991043091,
+      "learning_rate": 9.946781231649605e-06,
+      "loss": 0.3883,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4258048289738431,
+      "grad_norm": 0.3793240487575531,
+      "learning_rate": 9.946568097341827e-06,
+      "loss": 0.388,
+      "step": 1693
+    },
+    {
+      "epoch": 0.426056338028169,
+      "grad_norm": 0.4271707832813263,
+      "learning_rate": 9.946354539391436e-06,
+      "loss": 0.3985,
+      "step": 1694
+    },
+    {
+      "epoch": 0.426307847082495,
+      "grad_norm": 0.396963894367218,
+      "learning_rate": 9.946140557816724e-06,
+      "loss": 0.3966,
+      "step": 1695
+    },
+    {
+      "epoch": 0.42655935613682094,
+      "grad_norm": 0.39994847774505615,
+      "learning_rate": 9.945926152636017e-06,
+      "loss": 0.3922,
+      "step": 1696
+    },
+    {
+      "epoch": 0.42681086519114686,
+      "grad_norm": 0.38306140899658203,
+      "learning_rate": 9.945711323867674e-06,
+      "loss": 0.406,
+      "step": 1697
+    },
+    {
+      "epoch": 0.42706237424547283,
+      "grad_norm": 0.4517843723297119,
+      "learning_rate": 9.945496071530098e-06,
+      "loss": 0.3871,
+      "step": 1698
+    },
+    {
+      "epoch": 0.4273138832997988,
+      "grad_norm": 0.40096959471702576,
+      "learning_rate": 9.945280395641724e-06,
+      "loss": 0.4098,
+      "step": 1699
+    },
+    {
+      "epoch": 0.4275653923541248,
+      "grad_norm": 0.42314156889915466,
+      "learning_rate": 9.945064296221019e-06,
+      "loss": 0.3858,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4278169014084507,
+      "grad_norm": 0.4481252431869507,
+      "learning_rate": 9.944847773286495e-06,
+      "loss": 0.4273,
+      "step": 1701
+    },
+    {
+      "epoch": 0.42806841046277666,
+      "grad_norm": 0.4047831892967224,
+      "learning_rate": 9.944630826856694e-06,
+      "loss": 0.3892,
+      "step": 1702
+    },
+    {
+      "epoch": 0.42831991951710263,
+      "grad_norm": 0.40450939536094666,
+      "learning_rate": 9.944413456950195e-06,
+      "loss": 0.4199,
+      "step": 1703
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.3987005352973938,
+      "learning_rate": 9.944195663585616e-06,
+      "loss": 0.4254,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4288229376257545,
+      "grad_norm": 0.426527738571167,
+      "learning_rate": 9.94397744678161e-06,
+      "loss": 0.3992,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4290744466800805,
+      "grad_norm": 0.4253356158733368,
+      "learning_rate": 9.943758806556864e-06,
+      "loss": 0.3876,
+      "step": 1706
+    },
+    {
+      "epoch": 0.42932595573440646,
+      "grad_norm": 0.4413727819919586,
+      "learning_rate": 9.943539742930105e-06,
+      "loss": 0.4213,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4295774647887324,
+      "grad_norm": 0.4230106472969055,
+      "learning_rate": 9.943320255920093e-06,
+      "loss": 0.4092,
+      "step": 1708
+    },
+    {
+      "epoch": 0.42982897384305835,
+      "grad_norm": 0.4080696702003479,
+      "learning_rate": 9.943100345545627e-06,
+      "loss": 0.3789,
+      "step": 1709
+    },
+    {
+      "epoch": 0.4300804828973843,
+      "grad_norm": 0.4206973612308502,
+      "learning_rate": 9.94288001182554e-06,
+      "loss": 0.4128,
+      "step": 1710
+    },
+    {
+      "epoch": 0.43033199195171024,
+      "grad_norm": 0.41314277052879333,
+      "learning_rate": 9.942659254778703e-06,
+      "loss": 0.3883,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4305835010060362,
+      "grad_norm": 0.42901384830474854,
+      "learning_rate": 9.942438074424024e-06,
+      "loss": 0.3917,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4308350100603622,
+      "grad_norm": 0.39158564805984497,
+      "learning_rate": 9.942216470780441e-06,
+      "loss": 0.3778,
+      "step": 1713
+    },
+    {
+      "epoch": 0.43108651911468815,
+      "grad_norm": 0.423117995262146,
+      "learning_rate": 9.941994443866936e-06,
+      "loss": 0.408,
+      "step": 1714
+    },
+    {
+      "epoch": 0.43133802816901406,
+      "grad_norm": 0.44455990195274353,
+      "learning_rate": 9.941771993702524e-06,
+      "loss": 0.3849,
+      "step": 1715
+    },
+    {
+      "epoch": 0.43158953722334004,
+      "grad_norm": 0.41482892632484436,
+      "learning_rate": 9.941549120306257e-06,
+      "loss": 0.4121,
+      "step": 1716
+    },
+    {
+      "epoch": 0.431841046277666,
+      "grad_norm": 0.47499921917915344,
+      "learning_rate": 9.941325823697221e-06,
+      "loss": 0.3987,
+      "step": 1717
+    },
+    {
+      "epoch": 0.432092555331992,
+      "grad_norm": 0.40756213665008545,
+      "learning_rate": 9.941102103894541e-06,
+      "loss": 0.4043,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4323440643863179,
+      "grad_norm": 0.42999786138534546,
+      "learning_rate": 9.94087796091738e-06,
+      "loss": 0.4134,
+      "step": 1719
+    },
+    {
+      "epoch": 0.43259557344064387,
+      "grad_norm": 0.469237744808197,
+      "learning_rate": 9.940653394784929e-06,
+      "loss": 0.3982,
+      "step": 1720
+    },
+    {
+      "epoch": 0.43284708249496984,
+      "grad_norm": 0.3991852402687073,
+      "learning_rate": 9.940428405516423e-06,
+      "loss": 0.3709,
+      "step": 1721
+    },
+    {
+      "epoch": 0.43309859154929575,
+      "grad_norm": 0.42819303274154663,
+      "learning_rate": 9.940202993131132e-06,
+      "loss": 0.3955,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4333501006036217,
+      "grad_norm": 0.40518656373023987,
+      "learning_rate": 9.939977157648361e-06,
+      "loss": 0.3797,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4336016096579477,
+      "grad_norm": 0.38709619641304016,
+      "learning_rate": 9.939750899087451e-06,
+      "loss": 0.4136,
+      "step": 1724
+    },
+    {
+      "epoch": 0.43385311871227367,
+      "grad_norm": 0.400952011346817,
+      "learning_rate": 9.939524217467779e-06,
+      "loss": 0.3917,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4341046277665996,
+      "grad_norm": 0.4235305190086365,
+      "learning_rate": 9.93929711280876e-06,
+      "loss": 0.3936,
+      "step": 1726
+    },
+    {
+      "epoch": 0.43435613682092555,
+      "grad_norm": 0.4117744266986847,
+      "learning_rate": 9.939069585129841e-06,
+      "loss": 0.4083,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4346076458752515,
+      "grad_norm": 0.37095022201538086,
+      "learning_rate": 9.938841634450513e-06,
+      "loss": 0.3841,
+      "step": 1728
+    },
+    {
+      "epoch": 0.43485915492957744,
+      "grad_norm": 0.4193072021007538,
+      "learning_rate": 9.938613260790295e-06,
+      "loss": 0.4061,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4351106639839034,
+      "grad_norm": 0.45826607942581177,
+      "learning_rate": 9.938384464168748e-06,
+      "loss": 0.3937,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4353621730382294,
+      "grad_norm": 0.3635403513908386,
+      "learning_rate": 9.938155244605467e-06,
+      "loss": 0.4176,
+      "step": 1731
+    },
+    {
+      "epoch": 0.43561368209255535,
+      "grad_norm": 0.47581997513771057,
+      "learning_rate": 9.937925602120083e-06,
+      "loss": 0.4053,
+      "step": 1732
+    },
+    {
+      "epoch": 0.43586519114688127,
+      "grad_norm": 0.4900721311569214,
+      "learning_rate": 9.937695536732259e-06,
+      "loss": 0.3965,
+      "step": 1733
+    },
+    {
+      "epoch": 0.43611670020120724,
+      "grad_norm": 0.4279528856277466,
+      "learning_rate": 9.937465048461705e-06,
+      "loss": 0.4019,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4363682092555332,
+      "grad_norm": 0.42261025309562683,
+      "learning_rate": 9.937234137328157e-06,
+      "loss": 0.4054,
+      "step": 1735
+    },
+    {
+      "epoch": 0.43661971830985913,
+      "grad_norm": 0.4319506287574768,
+      "learning_rate": 9.937002803351394e-06,
+      "loss": 0.3924,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4368712273641851,
+      "grad_norm": 0.4078814387321472,
+      "learning_rate": 9.936771046551225e-06,
+      "loss": 0.3855,
+      "step": 1737
+    },
+    {
+      "epoch": 0.43712273641851107,
+      "grad_norm": 0.4161885678768158,
+      "learning_rate": 9.936538866947501e-06,
+      "loss": 0.4115,
+      "step": 1738
+    },
+    {
+      "epoch": 0.43737424547283704,
+      "grad_norm": 0.3567904233932495,
+      "learning_rate": 9.936306264560107e-06,
+      "loss": 0.3882,
+      "step": 1739
+    },
+    {
+      "epoch": 0.43762575452716296,
+      "grad_norm": 0.43193182349205017,
+      "learning_rate": 9.93607323940896e-06,
+      "loss": 0.3699,
+      "step": 1740
+    },
+    {
+      "epoch": 0.43787726358148893,
+      "grad_norm": 0.3843153715133667,
+      "learning_rate": 9.935839791514024e-06,
+      "loss": 0.4166,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4381287726358149,
+      "grad_norm": 0.3638061583042145,
+      "learning_rate": 9.935605920895286e-06,
+      "loss": 0.3891,
+      "step": 1742
+    },
+    {
+      "epoch": 0.43838028169014087,
+      "grad_norm": 0.3796941041946411,
+      "learning_rate": 9.93537162757278e-06,
+      "loss": 0.3806,
+      "step": 1743
+    },
+    {
+      "epoch": 0.4386317907444668,
+      "grad_norm": 0.3639376163482666,
+      "learning_rate": 9.935136911566566e-06,
+      "loss": 0.3785,
+      "step": 1744
+    },
+    {
+      "epoch": 0.43888329979879276,
+      "grad_norm": 0.4310847520828247,
+      "learning_rate": 9.93490177289675e-06,
+      "loss": 0.4118,
+      "step": 1745
+    },
+    {
+      "epoch": 0.43913480885311873,
+      "grad_norm": 0.4827233552932739,
+      "learning_rate": 9.934666211583472e-06,
+      "loss": 0.4192,
+      "step": 1746
+    },
+    {
+      "epoch": 0.43938631790744465,
+      "grad_norm": 0.36181917786598206,
+      "learning_rate": 9.934430227646904e-06,
+      "loss": 0.4225,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4396378269617706,
+      "grad_norm": 0.4338870346546173,
+      "learning_rate": 9.934193821107256e-06,
+      "loss": 0.3857,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4398893360160966,
+      "grad_norm": 0.454122930765152,
+      "learning_rate": 9.933956991984775e-06,
+      "loss": 0.4203,
+      "step": 1749
+    },
+    {
+      "epoch": 0.44014084507042256,
+      "grad_norm": 0.440476655960083,
+      "learning_rate": 9.933719740299745e-06,
+      "loss": 0.411,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4403923541247485,
+      "grad_norm": 0.3782346546649933,
+      "learning_rate": 9.933482066072485e-06,
+      "loss": 0.383,
+      "step": 1751
+    },
+    {
+      "epoch": 0.44064386317907445,
+      "grad_norm": 0.4182790219783783,
+      "learning_rate": 9.93324396932335e-06,
+      "loss": 0.3782,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4408953722334004,
+      "grad_norm": 0.40436455607414246,
+      "learning_rate": 9.93300545007273e-06,
+      "loss": 0.3773,
+      "step": 1753
+    },
+    {
+      "epoch": 0.44114688128772633,
+      "grad_norm": 0.448367714881897,
+      "learning_rate": 9.932766508341054e-06,
+      "loss": 0.4067,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4413983903420523,
+      "grad_norm": 0.39983075857162476,
+      "learning_rate": 9.932527144148788e-06,
+      "loss": 0.3907,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4416498993963783,
+      "grad_norm": 0.4192653000354767,
+      "learning_rate": 9.932287357516428e-06,
+      "loss": 0.3743,
+      "step": 1756
+    },
+    {
+      "epoch": 0.44190140845070425,
+      "grad_norm": 0.4609915614128113,
+      "learning_rate": 9.932047148464512e-06,
+      "loss": 0.3671,
+      "step": 1757
+    },
+    {
+      "epoch": 0.44215291750503016,
+      "grad_norm": 0.4244169294834137,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.3848,
+      "step": 1758
+    },
+    {
+      "epoch": 0.44240442655935613,
+      "grad_norm": 0.4056912064552307,
+      "learning_rate": 9.931565463184339e-06,
+      "loss": 0.4049,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4426559356136821,
+      "grad_norm": 0.5338891744613647,
+      "learning_rate": 9.931323986997334e-06,
+      "loss": 0.3858,
+      "step": 1760
+    },
+    {
+      "epoch": 0.442907444668008,
+      "grad_norm": 0.4530712962150574,
+      "learning_rate": 9.93108208847328e-06,
+      "loss": 0.4024,
+      "step": 1761
+    },
+    {
+      "epoch": 0.443158953722334,
+      "grad_norm": 0.43134480714797974,
+      "learning_rate": 9.930839767632895e-06,
+      "loss": 0.4158,
+      "step": 1762
+    },
+    {
+      "epoch": 0.44341046277665996,
+      "grad_norm": 0.475333571434021,
+      "learning_rate": 9.930597024496933e-06,
+      "loss": 0.4289,
+      "step": 1763
+    },
+    {
+      "epoch": 0.44366197183098594,
+      "grad_norm": 0.46597325801849365,
+      "learning_rate": 9.930353859086177e-06,
+      "loss": 0.409,
+      "step": 1764
+    },
+    {
+      "epoch": 0.44391348088531185,
+      "grad_norm": 0.3584541976451874,
+      "learning_rate": 9.93011027142146e-06,
+      "loss": 0.3936,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4441649899396378,
+      "grad_norm": 0.4492575526237488,
+      "learning_rate": 9.92986626152364e-06,
+      "loss": 0.4067,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4444164989939638,
+      "grad_norm": 0.4518819749355316,
+      "learning_rate": 9.929621829413616e-06,
+      "loss": 0.3997,
+      "step": 1767
+    },
+    {
+      "epoch": 0.44466800804828976,
+      "grad_norm": 0.39923927187919617,
+      "learning_rate": 9.929376975112324e-06,
+      "loss": 0.4011,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4449195171026157,
+      "grad_norm": 0.4502282440662384,
+      "learning_rate": 9.92913169864073e-06,
+      "loss": 0.4245,
+      "step": 1769
+    },
+    {
+      "epoch": 0.44517102615694165,
+      "grad_norm": 0.49698328971862793,
+      "learning_rate": 9.928886000019842e-06,
+      "loss": 0.409,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4454225352112676,
+      "grad_norm": 0.43445685505867004,
+      "learning_rate": 9.928639879270705e-06,
+      "loss": 0.3875,
+      "step": 1771
+    },
+    {
+      "epoch": 0.44567404426559354,
+      "grad_norm": 0.43353793025016785,
+      "learning_rate": 9.928393336414394e-06,
+      "loss": 0.4041,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4459255533199195,
+      "grad_norm": 0.5283091068267822,
+      "learning_rate": 9.928146371472027e-06,
+      "loss": 0.409,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4461770623742455,
+      "grad_norm": 0.3890385925769806,
+      "learning_rate": 9.927898984464753e-06,
+      "loss": 0.3992,
+      "step": 1774
+    },
+    {
+      "epoch": 0.44642857142857145,
+      "grad_norm": 0.36789190769195557,
+      "learning_rate": 9.92765117541376e-06,
+      "loss": 0.3825,
+      "step": 1775
+    },
+    {
+      "epoch": 0.44668008048289737,
+      "grad_norm": 0.4484959840774536,
+      "learning_rate": 9.927402944340271e-06,
+      "loss": 0.3804,
+      "step": 1776
+    },
+    {
+      "epoch": 0.44693158953722334,
+      "grad_norm": 0.4116009473800659,
+      "learning_rate": 9.927154291265546e-06,
+      "loss": 0.4022,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4471830985915493,
+      "grad_norm": 0.4052729308605194,
+      "learning_rate": 9.92690521621088e-06,
+      "loss": 0.4163,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4474346076458752,
+      "grad_norm": 0.4312780797481537,
+      "learning_rate": 9.926655719197604e-06,
+      "loss": 0.4167,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4476861167002012,
+      "grad_norm": 0.38401511311531067,
+      "learning_rate": 9.926405800247088e-06,
+      "loss": 0.3918,
+      "step": 1780
+    },
+    {
+      "epoch": 0.44793762575452717,
+      "grad_norm": 0.4729456305503845,
+      "learning_rate": 9.926155459380733e-06,
+      "loss": 0.416,
+      "step": 1781
+    },
+    {
+      "epoch": 0.44818913480885314,
+      "grad_norm": 0.43422985076904297,
+      "learning_rate": 9.925904696619983e-06,
+      "loss": 0.3793,
+      "step": 1782
+    },
+    {
+      "epoch": 0.44844064386317906,
+      "grad_norm": 0.3729407489299774,
+      "learning_rate": 9.92565351198631e-06,
+      "loss": 0.369,
+      "step": 1783
+    },
+    {
+      "epoch": 0.448692152917505,
+      "grad_norm": 0.3949672281742096,
+      "learning_rate": 9.92540190550123e-06,
+      "loss": 0.3661,
+      "step": 1784
+    },
+    {
+      "epoch": 0.448943661971831,
+      "grad_norm": 0.5043959617614746,
+      "learning_rate": 9.92514987718629e-06,
+      "loss": 0.388,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4491951710261569,
+      "grad_norm": 0.4054137170314789,
+      "learning_rate": 9.924897427063074e-06,
+      "loss": 0.3962,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4494466800804829,
+      "grad_norm": 0.426899790763855,
+      "learning_rate": 9.924644555153203e-06,
+      "loss": 0.4045,
+      "step": 1787
+    },
+    {
+      "epoch": 0.44969818913480886,
+      "grad_norm": 0.4646405577659607,
+      "learning_rate": 9.924391261478334e-06,
+      "loss": 0.3934,
+      "step": 1788
+    },
+    {
+      "epoch": 0.44994969818913483,
+      "grad_norm": 0.3873588442802429,
+      "learning_rate": 9.924137546060162e-06,
+      "loss": 0.4043,
+      "step": 1789
+    },
+    {
+      "epoch": 0.45020120724346074,
+      "grad_norm": 0.5013946890830994,
+      "learning_rate": 9.923883408920414e-06,
+      "loss": 0.3877,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4504527162977867,
+      "grad_norm": 0.4803258776664734,
+      "learning_rate": 9.923628850080856e-06,
+      "loss": 0.4154,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 0.47052329778671265,
+      "learning_rate": 9.923373869563288e-06,
+      "loss": 0.4065,
+      "step": 1792
+    },
+    {
+      "epoch": 0.45095573440643866,
+      "grad_norm": 0.4442056119441986,
+      "learning_rate": 9.92311846738955e-06,
+      "loss": 0.3974,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4512072434607646,
+      "grad_norm": 0.4476431906223297,
+      "learning_rate": 9.922862643581512e-06,
+      "loss": 0.4098,
+      "step": 1794
+    },
+    {
+      "epoch": 0.45145875251509054,
+      "grad_norm": 0.41717463731765747,
+      "learning_rate": 9.922606398161088e-06,
+      "loss": 0.4118,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4517102615694165,
+      "grad_norm": 0.4191329777240753,
+      "learning_rate": 9.922349731150221e-06,
+      "loss": 0.403,
+      "step": 1796
+    },
+    {
+      "epoch": 0.45196177062374243,
+      "grad_norm": 0.41227200627326965,
+      "learning_rate": 9.922092642570894e-06,
+      "loss": 0.4162,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4522132796780684,
+      "grad_norm": 0.3862850069999695,
+      "learning_rate": 9.921835132445124e-06,
+      "loss": 0.4071,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4524647887323944,
+      "grad_norm": 0.42549851536750793,
+      "learning_rate": 9.921577200794968e-06,
+      "loss": 0.4289,
+      "step": 1799
+    },
+    {
+      "epoch": 0.45271629778672035,
+      "grad_norm": 0.40862515568733215,
+      "learning_rate": 9.921318847642511e-06,
+      "loss": 0.4009,
+      "step": 1800
+    },
+    {
+      "epoch": 0.45296780684104626,
+      "grad_norm": 0.3994438350200653,
+      "learning_rate": 9.921060073009884e-06,
+      "loss": 0.3995,
+      "step": 1801
+    },
+    {
+      "epoch": 0.45321931589537223,
+      "grad_norm": 0.45960313081741333,
+      "learning_rate": 9.920800876919248e-06,
+      "loss": 0.4031,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4534708249496982,
+      "grad_norm": 0.42091697454452515,
+      "learning_rate": 9.9205412593928e-06,
+      "loss": 0.398,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4537223340040241,
+      "grad_norm": 0.45224541425704956,
+      "learning_rate": 9.920281220452776e-06,
+      "loss": 0.3865,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4539738430583501,
+      "grad_norm": 0.45991694927215576,
+      "learning_rate": 9.920020760121447e-06,
+      "loss": 0.3965,
+      "step": 1805
+    },
+    {
+      "epoch": 0.45422535211267606,
+      "grad_norm": 0.433249294757843,
+      "learning_rate": 9.919759878421121e-06,
+      "loss": 0.405,
+      "step": 1806
+    },
+    {
+      "epoch": 0.45447686116700203,
+      "grad_norm": 0.43937093019485474,
+      "learning_rate": 9.919498575374138e-06,
+      "loss": 0.4068,
+      "step": 1807
+    },
+    {
+      "epoch": 0.45472837022132795,
+      "grad_norm": 0.5110582709312439,
+      "learning_rate": 9.919236851002879e-06,
+      "loss": 0.4061,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4549798792756539,
+      "grad_norm": 0.4061850607395172,
+      "learning_rate": 9.918974705329756e-06,
+      "loss": 0.3686,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4552313883299799,
+      "grad_norm": 0.49449342489242554,
+      "learning_rate": 9.918712138377226e-06,
+      "loss": 0.3984,
+      "step": 1810
+    },
+    {
+      "epoch": 0.45548289738430586,
+      "grad_norm": 0.44511350989341736,
+      "learning_rate": 9.91844915016777e-06,
+      "loss": 0.3846,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4557344064386318,
+      "grad_norm": 0.3825312852859497,
+      "learning_rate": 9.918185740723916e-06,
+      "loss": 0.4109,
+      "step": 1812
+    },
+    {
+      "epoch": 0.45598591549295775,
+      "grad_norm": 0.44606679677963257,
+      "learning_rate": 9.91792191006822e-06,
+      "loss": 0.3899,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4562374245472837,
+      "grad_norm": 0.42780601978302,
+      "learning_rate": 9.917657658223278e-06,
+      "loss": 0.409,
+      "step": 1814
+    },
+    {
+      "epoch": 0.45648893360160964,
+      "grad_norm": 0.44149789214134216,
+      "learning_rate": 9.917392985211725e-06,
+      "loss": 0.4126,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4567404426559356,
+      "grad_norm": 0.40314993262290955,
+      "learning_rate": 9.917127891056223e-06,
+      "loss": 0.3824,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4569919517102616,
+      "grad_norm": 0.42890751361846924,
+      "learning_rate": 9.916862375779482e-06,
+      "loss": 0.4023,
+      "step": 1817
+    },
+    {
+      "epoch": 0.45724346076458755,
+      "grad_norm": 0.39626947045326233,
+      "learning_rate": 9.916596439404235e-06,
+      "loss": 0.4113,
+      "step": 1818
+    },
+    {
+      "epoch": 0.45749496981891347,
+      "grad_norm": 0.5269667506217957,
+      "learning_rate": 9.916330081953262e-06,
+      "loss": 0.3958,
+      "step": 1819
+    },
+    {
+      "epoch": 0.45774647887323944,
+      "grad_norm": 0.4348817765712738,
+      "learning_rate": 9.916063303449374e-06,
+      "loss": 0.373,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4579979879275654,
+      "grad_norm": 0.36105191707611084,
+      "learning_rate": 9.91579610391542e-06,
+      "loss": 0.3845,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4582494969818913,
+      "grad_norm": 0.5355477333068848,
+      "learning_rate": 9.915528483374283e-06,
+      "loss": 0.4226,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4585010060362173,
+      "grad_norm": 0.45753616094589233,
+      "learning_rate": 9.915260441848883e-06,
+      "loss": 0.4067,
+      "step": 1823
+    },
+    {
+      "epoch": 0.45875251509054327,
+      "grad_norm": 0.41938281059265137,
+      "learning_rate": 9.914991979362173e-06,
+      "loss": 0.4038,
+      "step": 1824
+    },
+    {
+      "epoch": 0.45900402414486924,
+      "grad_norm": 0.41249436140060425,
+      "learning_rate": 9.91472309593715e-06,
+      "loss": 0.3903,
+      "step": 1825
+    },
+    {
+      "epoch": 0.45925553319919515,
+      "grad_norm": 0.40385809540748596,
+      "learning_rate": 9.914453791596841e-06,
+      "loss": 0.4094,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4595070422535211,
+      "grad_norm": 0.4012143909931183,
+      "learning_rate": 9.914184066364308e-06,
+      "loss": 0.3886,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4597585513078471,
+      "grad_norm": 0.41770312190055847,
+      "learning_rate": 9.913913920262654e-06,
+      "loss": 0.3926,
+      "step": 1828
+    },
+    {
+      "epoch": 0.460010060362173,
+      "grad_norm": 0.3697265684604645,
+      "learning_rate": 9.913643353315015e-06,
+      "loss": 0.3827,
+      "step": 1829
+    },
+    {
+      "epoch": 0.460261569416499,
+      "grad_norm": 0.4404751658439636,
+      "learning_rate": 9.91337236554456e-06,
+      "loss": 0.3875,
+      "step": 1830
+    },
+    {
+      "epoch": 0.46051307847082495,
+      "grad_norm": 0.43419405817985535,
+      "learning_rate": 9.9131009569745e-06,
+      "loss": 0.4164,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4607645875251509,
+      "grad_norm": 0.4106455445289612,
+      "learning_rate": 9.91282912762808e-06,
+      "loss": 0.4161,
+      "step": 1832
+    },
+    {
+      "epoch": 0.46101609657947684,
+      "grad_norm": 0.48274555802345276,
+      "learning_rate": 9.912556877528582e-06,
+      "loss": 0.3943,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4612676056338028,
+      "grad_norm": 0.39638254046440125,
+      "learning_rate": 9.912284206699317e-06,
+      "loss": 0.4098,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4615191146881288,
+      "grad_norm": 0.4625331461429596,
+      "learning_rate": 9.912011115163642e-06,
+      "loss": 0.3682,
+      "step": 1835
+    },
+    {
+      "epoch": 0.46177062374245476,
+      "grad_norm": 0.4894983172416687,
+      "learning_rate": 9.911737602944943e-06,
+      "loss": 0.3913,
+      "step": 1836
+    },
+    {
+      "epoch": 0.46202213279678067,
+      "grad_norm": 0.3602517247200012,
+      "learning_rate": 9.911463670066648e-06,
+      "loss": 0.3814,
+      "step": 1837
+    },
+    {
+      "epoch": 0.46227364185110664,
+      "grad_norm": 0.4734443128108978,
+      "learning_rate": 9.911189316552217e-06,
+      "loss": 0.3926,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4625251509054326,
+      "grad_norm": 0.4669021964073181,
+      "learning_rate": 9.910914542425143e-06,
+      "loss": 0.4003,
+      "step": 1839
+    },
+    {
+      "epoch": 0.46277665995975853,
+      "grad_norm": 0.3793289363384247,
+      "learning_rate": 9.91063934770896e-06,
+      "loss": 0.3894,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4630281690140845,
+      "grad_norm": 0.4388173818588257,
+      "learning_rate": 9.910363732427241e-06,
+      "loss": 0.4081,
+      "step": 1841
+    },
+    {
+      "epoch": 0.46327967806841047,
+      "grad_norm": 0.44316378235816956,
+      "learning_rate": 9.910087696603585e-06,
+      "loss": 0.3768,
+      "step": 1842
+    },
+    {
+      "epoch": 0.46353118712273644,
+      "grad_norm": 0.4063831567764282,
+      "learning_rate": 9.909811240261635e-06,
+      "loss": 0.3986,
+      "step": 1843
+    },
+    {
+      "epoch": 0.46378269617706236,
+      "grad_norm": 0.40213075280189514,
+      "learning_rate": 9.90953436342507e-06,
+      "loss": 0.4181,
+      "step": 1844
+    },
+    {
+      "epoch": 0.46403420523138833,
+      "grad_norm": 0.38924410939216614,
+      "learning_rate": 9.909257066117599e-06,
+      "loss": 0.4157,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4642857142857143,
+      "grad_norm": 0.3829781115055084,
+      "learning_rate": 9.908979348362974e-06,
+      "loss": 0.3814,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4645372233400402,
+      "grad_norm": 0.3905456066131592,
+      "learning_rate": 9.908701210184976e-06,
+      "loss": 0.3843,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4647887323943662,
+      "grad_norm": 0.3960774838924408,
+      "learning_rate": 9.90842265160743e-06,
+      "loss": 0.409,
+      "step": 1848
+    },
+    {
+      "epoch": 0.46504024144869216,
+      "grad_norm": 0.39505699276924133,
+      "learning_rate": 9.908143672654192e-06,
+      "loss": 0.3942,
+      "step": 1849
+    },
+    {
+      "epoch": 0.46529175050301813,
+      "grad_norm": 0.4073687791824341,
+      "learning_rate": 9.907864273349152e-06,
+      "loss": 0.4076,
+      "step": 1850
+    },
+    {
+      "epoch": 0.46554325955734405,
+      "grad_norm": 0.4546225368976593,
+      "learning_rate": 9.907584453716238e-06,
+      "loss": 0.3739,
+      "step": 1851
+    },
+    {
+      "epoch": 0.46579476861167,
+      "grad_norm": 0.4456266760826111,
+      "learning_rate": 9.907304213779422e-06,
+      "loss": 0.3742,
+      "step": 1852
+    },
+    {
+      "epoch": 0.466046277665996,
+      "grad_norm": 0.4133337438106537,
+      "learning_rate": 9.907023553562699e-06,
+      "loss": 0.3993,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4662977867203219,
+      "grad_norm": 0.4463947117328644,
+      "learning_rate": 9.906742473090105e-06,
+      "loss": 0.3906,
+      "step": 1854
+    },
+    {
+      "epoch": 0.4665492957746479,
+      "grad_norm": 0.4527071714401245,
+      "learning_rate": 9.906460972385715e-06,
+      "loss": 0.3838,
+      "step": 1855
+    },
+    {
+      "epoch": 0.46680080482897385,
+      "grad_norm": 0.42571839690208435,
+      "learning_rate": 9.906179051473638e-06,
+      "loss": 0.3812,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4670523138832998,
+      "grad_norm": 0.5096774697303772,
+      "learning_rate": 9.905896710378019e-06,
+      "loss": 0.4192,
+      "step": 1857
+    },
+    {
+      "epoch": 0.46730382293762573,
+      "grad_norm": 0.43383294343948364,
+      "learning_rate": 9.905613949123036e-06,
+      "loss": 0.3769,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4675553319919517,
+      "grad_norm": 0.45289137959480286,
+      "learning_rate": 9.90533076773291e-06,
+      "loss": 0.4123,
+      "step": 1859
+    },
+    {
+      "epoch": 0.4678068410462777,
+      "grad_norm": 0.4052902162075043,
+      "learning_rate": 9.905047166231889e-06,
+      "loss": 0.4346,
+      "step": 1860
+    },
+    {
+      "epoch": 0.46805835010060365,
+      "grad_norm": 0.41848742961883545,
+      "learning_rate": 9.904763144644265e-06,
+      "loss": 0.4022,
+      "step": 1861
+    },
+    {
+      "epoch": 0.46830985915492956,
+      "grad_norm": 0.4163335859775543,
+      "learning_rate": 9.904478702994362e-06,
+      "loss": 0.4254,
+      "step": 1862
+    },
+    {
+      "epoch": 0.46856136820925554,
+      "grad_norm": 0.36754798889160156,
+      "learning_rate": 9.90419384130654e-06,
+      "loss": 0.4084,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4688128772635815,
+      "grad_norm": 0.4322851002216339,
+      "learning_rate": 9.903908559605197e-06,
+      "loss": 0.3982,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4690643863179074,
+      "grad_norm": 0.41940197348594666,
+      "learning_rate": 9.903622857914766e-06,
+      "loss": 0.4128,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4693158953722334,
+      "grad_norm": 0.3755689859390259,
+      "learning_rate": 9.90333673625971e-06,
+      "loss": 0.4054,
+      "step": 1866
+    },
+    {
+      "epoch": 0.46956740442655936,
+      "grad_norm": 0.422498881816864,
+      "learning_rate": 9.903050194664541e-06,
+      "loss": 0.4145,
+      "step": 1867
+    },
+    {
+      "epoch": 0.46981891348088534,
+      "grad_norm": 0.3807189166545868,
+      "learning_rate": 9.902763233153796e-06,
+      "loss": 0.4154,
+      "step": 1868
+    },
+    {
+      "epoch": 0.47007042253521125,
+      "grad_norm": 0.4063322842121124,
+      "learning_rate": 9.90247585175205e-06,
+      "loss": 0.3964,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4703219315895372,
+      "grad_norm": 0.42195388674736023,
+      "learning_rate": 9.902188050483918e-06,
+      "loss": 0.4153,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4705734406438632,
+      "grad_norm": 0.44243067502975464,
+      "learning_rate": 9.901899829374048e-06,
+      "loss": 0.4107,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4708249496981891,
+      "grad_norm": 0.39202213287353516,
+      "learning_rate": 9.901611188447123e-06,
+      "loss": 0.3929,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4710764587525151,
+      "grad_norm": 0.4214573800563812,
+      "learning_rate": 9.901322127727864e-06,
+      "loss": 0.4043,
+      "step": 1873
+    },
+    {
+      "epoch": 0.47132796780684105,
+      "grad_norm": 0.4241698980331421,
+      "learning_rate": 9.901032647241028e-06,
+      "loss": 0.4463,
+      "step": 1874
+    },
+    {
+      "epoch": 0.471579476861167,
+      "grad_norm": 0.3987068831920624,
+      "learning_rate": 9.900742747011405e-06,
+      "loss": 0.4225,
+      "step": 1875
+    },
+    {
+      "epoch": 0.47183098591549294,
+      "grad_norm": 0.38595229387283325,
+      "learning_rate": 9.900452427063827e-06,
+      "loss": 0.401,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4720824949698189,
+      "grad_norm": 0.4303046464920044,
+      "learning_rate": 9.900161687423155e-06,
+      "loss": 0.424,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4723340040241449,
+      "grad_norm": 0.453784704208374,
+      "learning_rate": 9.89987052811429e-06,
+      "loss": 0.4054,
+      "step": 1878
+    },
+    {
+      "epoch": 0.4725855130784708,
+      "grad_norm": 0.41478443145751953,
+      "learning_rate": 9.899578949162167e-06,
+      "loss": 0.3934,
+      "step": 1879
+    },
+    {
+      "epoch": 0.47283702213279677,
+      "grad_norm": 0.4205837547779083,
+      "learning_rate": 9.899286950591758e-06,
+      "loss": 0.379,
+      "step": 1880
+    },
+    {
+      "epoch": 0.47308853118712274,
+      "grad_norm": 0.4482855498790741,
+      "learning_rate": 9.898994532428071e-06,
+      "loss": 0.3874,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4733400402414487,
+      "grad_norm": 0.3712460994720459,
+      "learning_rate": 9.898701694696154e-06,
+      "loss": 0.423,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4735915492957746,
+      "grad_norm": 0.4586082100868225,
+      "learning_rate": 9.89840843742108e-06,
+      "loss": 0.39,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4738430583501006,
+      "grad_norm": 0.3899551033973694,
+      "learning_rate": 9.898114760627968e-06,
+      "loss": 0.4005,
+      "step": 1884
+    },
+    {
+      "epoch": 0.47409456740442657,
+      "grad_norm": 0.39340347051620483,
+      "learning_rate": 9.89782066434197e-06,
+      "loss": 0.4229,
+      "step": 1885
+    },
+    {
+      "epoch": 0.47434607645875254,
+      "grad_norm": 0.4018033742904663,
+      "learning_rate": 9.897526148588272e-06,
+      "loss": 0.3884,
+      "step": 1886
+    },
+    {
+      "epoch": 0.47459758551307846,
+      "grad_norm": 0.38298699259757996,
+      "learning_rate": 9.8972312133921e-06,
+      "loss": 0.414,
+      "step": 1887
+    },
+    {
+      "epoch": 0.47484909456740443,
+      "grad_norm": 0.36675944924354553,
+      "learning_rate": 9.896935858778708e-06,
+      "loss": 0.3942,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4751006036217304,
+      "grad_norm": 0.4059602916240692,
+      "learning_rate": 9.896640084773399e-06,
+      "loss": 0.3728,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4753521126760563,
+      "grad_norm": 0.4235600233078003,
+      "learning_rate": 9.896343891401498e-06,
+      "loss": 0.3886,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4756036217303823,
+      "grad_norm": 0.3745618760585785,
+      "learning_rate": 9.896047278688375e-06,
+      "loss": 0.3832,
+      "step": 1891
+    },
+    {
+      "epoch": 0.47585513078470826,
+      "grad_norm": 0.41866034269332886,
+      "learning_rate": 9.89575024665943e-06,
+      "loss": 0.3678,
+      "step": 1892
+    },
+    {
+      "epoch": 0.47610663983903423,
+      "grad_norm": 0.39261144399642944,
+      "learning_rate": 9.895452795340106e-06,
+      "loss": 0.4036,
+      "step": 1893
+    },
+    {
+      "epoch": 0.47635814889336014,
+      "grad_norm": 0.43027186393737793,
+      "learning_rate": 9.895154924755875e-06,
+      "loss": 0.3915,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4766096579476861,
+      "grad_norm": 0.42222538590431213,
+      "learning_rate": 9.894856634932249e-06,
+      "loss": 0.3949,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4768611670020121,
+      "grad_norm": 0.3839363157749176,
+      "learning_rate": 9.894557925894775e-06,
+      "loss": 0.4005,
+      "step": 1896
+    },
+    {
+      "epoch": 0.477112676056338,
+      "grad_norm": 0.39426472783088684,
+      "learning_rate": 9.894258797669034e-06,
+      "loss": 0.3867,
+      "step": 1897
+    },
+    {
+      "epoch": 0.477364185110664,
+      "grad_norm": 0.3634321391582489,
+      "learning_rate": 9.893959250280646e-06,
+      "loss": 0.3862,
+      "step": 1898
+    },
+    {
+      "epoch": 0.47761569416498995,
+      "grad_norm": 0.4207385182380676,
+      "learning_rate": 9.893659283755264e-06,
+      "loss": 0.3887,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4778672032193159,
+      "grad_norm": 0.3956140875816345,
+      "learning_rate": 9.89335889811858e-06,
+      "loss": 0.397,
+      "step": 1900
+    },
+    {
+      "epoch": 0.47811871227364183,
+      "grad_norm": 0.37039822340011597,
+      "learning_rate": 9.893058093396318e-06,
+      "loss": 0.386,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4783702213279678,
+      "grad_norm": 0.38195037841796875,
+      "learning_rate": 9.892756869614242e-06,
+      "loss": 0.4047,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4786217303822938,
+      "grad_norm": 0.39862626791000366,
+      "learning_rate": 9.892455226798148e-06,
+      "loss": 0.4128,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4788732394366197,
+      "grad_norm": 0.3864774703979492,
+      "learning_rate": 9.892153164973873e-06,
+      "loss": 0.3959,
+      "step": 1904
+    },
+    {
+      "epoch": 0.47912474849094566,
+      "grad_norm": 0.40958353877067566,
+      "learning_rate": 9.891850684167284e-06,
+      "loss": 0.4151,
+      "step": 1905
+    },
+    {
+      "epoch": 0.47937625754527163,
+      "grad_norm": 0.3836343586444855,
+      "learning_rate": 9.891547784404285e-06,
+      "loss": 0.3945,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4796277665995976,
+      "grad_norm": 0.37921276688575745,
+      "learning_rate": 9.891244465710822e-06,
+      "loss": 0.3856,
+      "step": 1907
+    },
+    {
+      "epoch": 0.4798792756539235,
+      "grad_norm": 0.39044657349586487,
+      "learning_rate": 9.890940728112869e-06,
+      "loss": 0.3913,
+      "step": 1908
+    },
+    {
+      "epoch": 0.4801307847082495,
+      "grad_norm": 0.38842642307281494,
+      "learning_rate": 9.89063657163644e-06,
+      "loss": 0.378,
+      "step": 1909
+    },
+    {
+      "epoch": 0.48038229376257546,
+      "grad_norm": 0.4392659068107605,
+      "learning_rate": 9.890331996307585e-06,
+      "loss": 0.3937,
+      "step": 1910
+    },
+    {
+      "epoch": 0.48063380281690143,
+      "grad_norm": 0.41104656457901,
+      "learning_rate": 9.89002700215239e-06,
+      "loss": 0.3849,
+      "step": 1911
+    },
+    {
+      "epoch": 0.48088531187122735,
+      "grad_norm": 0.3936188519001007,
+      "learning_rate": 9.88972158919697e-06,
+      "loss": 0.4185,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4811368209255533,
+      "grad_norm": 0.39013659954071045,
+      "learning_rate": 9.88941575746749e-06,
+      "loss": 0.353,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4813883299798793,
+      "grad_norm": 0.3972854018211365,
+      "learning_rate": 9.889109506990137e-06,
+      "loss": 0.3933,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4816398390342052,
+      "grad_norm": 0.46189555525779724,
+      "learning_rate": 9.88880283779114e-06,
+      "loss": 0.4109,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4818913480885312,
+      "grad_norm": 0.391872763633728,
+      "learning_rate": 9.888495749896766e-06,
+      "loss": 0.4003,
+      "step": 1916
+    },
+    {
+      "epoch": 0.48214285714285715,
+      "grad_norm": 0.49157020449638367,
+      "learning_rate": 9.888188243333313e-06,
+      "loss": 0.3938,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4823943661971831,
+      "grad_norm": 0.6632839441299438,
+      "learning_rate": 9.887880318127116e-06,
+      "loss": 0.3952,
+      "step": 1918
+    },
+    {
+      "epoch": 0.48264587525150904,
+      "grad_norm": 0.42987990379333496,
+      "learning_rate": 9.88757197430455e-06,
+      "loss": 0.4032,
+      "step": 1919
+    },
+    {
+      "epoch": 0.482897384305835,
+      "grad_norm": 0.43732520937919617,
+      "learning_rate": 9.887263211892017e-06,
+      "loss": 0.4166,
+      "step": 1920
+    },
+    {
+      "epoch": 0.483148893360161,
+      "grad_norm": 0.45760130882263184,
+      "learning_rate": 9.886954030915969e-06,
+      "loss": 0.3951,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4834004024144869,
+      "grad_norm": 0.4097820520401001,
+      "learning_rate": 9.886644431402879e-06,
+      "loss": 0.422,
+      "step": 1922
+    },
+    {
+      "epoch": 0.48365191146881287,
+      "grad_norm": 0.47169190645217896,
+      "learning_rate": 9.886334413379263e-06,
+      "loss": 0.4127,
+      "step": 1923
+    },
+    {
+      "epoch": 0.48390342052313884,
+      "grad_norm": 0.45416510105133057,
+      "learning_rate": 9.886023976871678e-06,
+      "loss": 0.3942,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4841549295774648,
+      "grad_norm": 0.4500854015350342,
+      "learning_rate": 9.885713121906701e-06,
+      "loss": 0.3861,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4844064386317907,
+      "grad_norm": 0.49403610825538635,
+      "learning_rate": 9.885401848510962e-06,
+      "loss": 0.4131,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4846579476861167,
+      "grad_norm": 0.43086037039756775,
+      "learning_rate": 9.88509015671112e-06,
+      "loss": 0.3868,
+      "step": 1927
+    },
+    {
+      "epoch": 0.48490945674044267,
+      "grad_norm": 0.4836970269680023,
+      "learning_rate": 9.884778046533863e-06,
+      "loss": 0.3939,
+      "step": 1928
+    },
+    {
+      "epoch": 0.48516096579476864,
+      "grad_norm": 0.41725337505340576,
+      "learning_rate": 9.884465518005927e-06,
+      "loss": 0.4104,
+      "step": 1929
+    },
+    {
+      "epoch": 0.48541247484909456,
+      "grad_norm": 0.43721911311149597,
+      "learning_rate": 9.884152571154077e-06,
+      "loss": 0.4114,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4856639839034205,
+      "grad_norm": 0.39998695254325867,
+      "learning_rate": 9.883839206005115e-06,
+      "loss": 0.386,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4859154929577465,
+      "grad_norm": 0.41072046756744385,
+      "learning_rate": 9.883525422585877e-06,
+      "loss": 0.4145,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4861670020120724,
+      "grad_norm": 0.40265727043151855,
+      "learning_rate": 9.883211220923237e-06,
+      "loss": 0.4205,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4864185110663984,
+      "grad_norm": 0.4055761396884918,
+      "learning_rate": 9.882896601044107e-06,
+      "loss": 0.4011,
+      "step": 1934
+    },
+    {
+      "epoch": 0.48667002012072436,
+      "grad_norm": 0.43374502658843994,
+      "learning_rate": 9.882581562975431e-06,
+      "loss": 0.4031,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4869215291750503,
+      "grad_norm": 0.3941003978252411,
+      "learning_rate": 9.882266106744187e-06,
+      "loss": 0.3767,
+      "step": 1936
+    },
+    {
+      "epoch": 0.48717303822937624,
+      "grad_norm": 0.41482222080230713,
+      "learning_rate": 9.881950232377397e-06,
+      "loss": 0.3977,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4874245472837022,
+      "grad_norm": 0.39503028988838196,
+      "learning_rate": 9.88163393990211e-06,
+      "loss": 0.3972,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4876760563380282,
+      "grad_norm": 0.43991711735725403,
+      "learning_rate": 9.881317229345414e-06,
+      "loss": 0.375,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4879275653923541,
+      "grad_norm": 0.41871747374534607,
+      "learning_rate": 9.881000100734436e-06,
+      "loss": 0.4188,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4881790744466801,
+      "grad_norm": 0.4454898238182068,
+      "learning_rate": 9.880682554096334e-06,
+      "loss": 0.3924,
+      "step": 1941
+    },
+    {
+      "epoch": 0.48843058350100604,
+      "grad_norm": 0.4203794598579407,
+      "learning_rate": 9.880364589458306e-06,
+      "loss": 0.3826,
+      "step": 1942
+    },
+    {
+      "epoch": 0.488682092555332,
+      "grad_norm": 0.4890785813331604,
+      "learning_rate": 9.880046206847581e-06,
+      "loss": 0.4006,
+      "step": 1943
+    },
+    {
+      "epoch": 0.48893360160965793,
+      "grad_norm": 0.9945582151412964,
+      "learning_rate": 9.879727406291429e-06,
+      "loss": 0.3835,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4891851106639839,
+      "grad_norm": 0.44654008746147156,
+      "learning_rate": 9.879408187817153e-06,
+      "loss": 0.3997,
+      "step": 1945
+    },
+    {
+      "epoch": 0.4894366197183099,
+      "grad_norm": 0.5195226669311523,
+      "learning_rate": 9.879088551452088e-06,
+      "loss": 0.3954,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4896881287726358,
+      "grad_norm": 0.4593251645565033,
+      "learning_rate": 9.878768497223614e-06,
+      "loss": 0.4029,
+      "step": 1947
+    },
+    {
+      "epoch": 0.48993963782696176,
+      "grad_norm": 0.5262354612350464,
+      "learning_rate": 9.87844802515914e-06,
+      "loss": 0.3896,
+      "step": 1948
+    },
+    {
+      "epoch": 0.49019114688128773,
+      "grad_norm": 0.5133786797523499,
+      "learning_rate": 9.878127135286112e-06,
+      "loss": 0.3685,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4904426559356137,
+      "grad_norm": 0.43951818346977234,
+      "learning_rate": 9.877805827632013e-06,
+      "loss": 0.4082,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4906941649899396,
+      "grad_norm": 0.5165718197822571,
+      "learning_rate": 9.877484102224359e-06,
+      "loss": 0.4198,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4909456740442656,
+      "grad_norm": 0.47304123640060425,
+      "learning_rate": 9.877161959090704e-06,
+      "loss": 0.4279,
+      "step": 1952
+    },
+    {
+      "epoch": 0.49119718309859156,
+      "grad_norm": 0.41676169633865356,
+      "learning_rate": 9.87683939825864e-06,
+      "loss": 0.3885,
+      "step": 1953
+    },
+    {
+      "epoch": 0.49144869215291753,
+      "grad_norm": 0.4127172529697418,
+      "learning_rate": 9.876516419755793e-06,
+      "loss": 0.3738,
+      "step": 1954
+    },
+    {
+      "epoch": 0.49170020120724345,
+      "grad_norm": 0.4324229061603546,
+      "learning_rate": 9.87619302360982e-06,
+      "loss": 0.3703,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4919517102615694,
+      "grad_norm": 0.3876967430114746,
+      "learning_rate": 9.875869209848418e-06,
+      "loss": 0.4076,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4922032193158954,
+      "grad_norm": 0.4635095000267029,
+      "learning_rate": 9.875544978499326e-06,
+      "loss": 0.4025,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4924547283702213,
+      "grad_norm": 0.410577654838562,
+      "learning_rate": 9.875220329590304e-06,
+      "loss": 0.4046,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4927062374245473,
+      "grad_norm": 0.3948517441749573,
+      "learning_rate": 9.874895263149163e-06,
+      "loss": 0.4266,
+      "step": 1959
+    },
+    {
+      "epoch": 0.49295774647887325,
+      "grad_norm": 0.3900180757045746,
+      "learning_rate": 9.874569779203737e-06,
+      "loss": 0.3899,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4932092555331992,
+      "grad_norm": 0.3953598737716675,
+      "learning_rate": 9.874243877781906e-06,
+      "loss": 0.4044,
+      "step": 1961
+    },
+    {
+      "epoch": 0.49346076458752514,
+      "grad_norm": 0.3972851037979126,
+      "learning_rate": 9.87391755891158e-06,
+      "loss": 0.4096,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4937122736418511,
+      "grad_norm": 0.3685133457183838,
+      "learning_rate": 9.873590822620706e-06,
+      "loss": 0.4009,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4939637826961771,
+      "grad_norm": 0.5181507468223572,
+      "learning_rate": 9.873263668937268e-06,
+      "loss": 0.4115,
+      "step": 1964
+    },
+    {
+      "epoch": 0.494215291750503,
+      "grad_norm": 0.38723844289779663,
+      "learning_rate": 9.872936097889284e-06,
+      "loss": 0.3957,
+      "step": 1965
+    },
+    {
+      "epoch": 0.49446680080482897,
+      "grad_norm": 0.42989009618759155,
+      "learning_rate": 9.872608109504807e-06,
+      "loss": 0.402,
+      "step": 1966
+    },
+    {
+      "epoch": 0.49471830985915494,
+      "grad_norm": 0.4205054044723511,
+      "learning_rate": 9.872279703811929e-06,
+      "loss": 0.4242,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4949698189134809,
+      "grad_norm": 0.369878888130188,
+      "learning_rate": 9.871950880838774e-06,
+      "loss": 0.4199,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4952213279678068,
+      "grad_norm": 0.3752254843711853,
+      "learning_rate": 9.871621640613506e-06,
+      "loss": 0.3765,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4954728370221328,
+      "grad_norm": 0.4001399278640747,
+      "learning_rate": 9.871291983164322e-06,
+      "loss": 0.3697,
+      "step": 1970
+    },
+    {
+      "epoch": 0.49572434607645877,
+      "grad_norm": 0.38128310441970825,
+      "learning_rate": 9.870961908519454e-06,
+      "loss": 0.4113,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4959758551307847,
+      "grad_norm": 0.3907400071620941,
+      "learning_rate": 9.87063141670717e-06,
+      "loss": 0.404,
+      "step": 1972
+    },
+    {
+      "epoch": 0.49622736418511065,
+      "grad_norm": 0.3958117961883545,
+      "learning_rate": 9.870300507755777e-06,
+      "loss": 0.3943,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4964788732394366,
+      "grad_norm": 0.412076473236084,
+      "learning_rate": 9.869969181693613e-06,
+      "loss": 0.4207,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4967303822937626,
+      "grad_norm": 0.3858659863471985,
+      "learning_rate": 9.869637438549056e-06,
+      "loss": 0.396,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4969818913480885,
+      "grad_norm": 0.44061511754989624,
+      "learning_rate": 9.869305278350516e-06,
+      "loss": 0.4096,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4972334004024145,
+      "grad_norm": 0.400973379611969,
+      "learning_rate": 9.868972701126442e-06,
+      "loss": 0.382,
+      "step": 1977
+    },
+    {
+      "epoch": 0.49748490945674045,
+      "grad_norm": 0.430945485830307,
+      "learning_rate": 9.868639706905314e-06,
+      "loss": 0.3994,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4977364185110664,
+      "grad_norm": 0.36873745918273926,
+      "learning_rate": 9.868306295715656e-06,
+      "loss": 0.3846,
+      "step": 1979
+    },
+    {
+      "epoch": 0.49798792756539234,
+      "grad_norm": 0.3921675682067871,
+      "learning_rate": 9.86797246758602e-06,
+      "loss": 0.3977,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4982394366197183,
+      "grad_norm": 0.3892073333263397,
+      "learning_rate": 9.867638222544994e-06,
+      "loss": 0.383,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4984909456740443,
+      "grad_norm": 0.3832413852214813,
+      "learning_rate": 9.867303560621207e-06,
+      "loss": 0.3954,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4987424547283702,
+      "grad_norm": 0.3576008677482605,
+      "learning_rate": 9.86696848184332e-06,
+      "loss": 0.3985,
+      "step": 1983
+    },
+    {
+      "epoch": 0.49899396378269617,
+      "grad_norm": 0.44481638073921204,
+      "learning_rate": 9.86663298624003e-06,
+      "loss": 0.3919,
+      "step": 1984
+    },
+    {
+      "epoch": 0.49924547283702214,
+      "grad_norm": 0.37005627155303955,
+      "learning_rate": 9.86629707384007e-06,
+      "loss": 0.3772,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4994969818913481,
+      "grad_norm": 0.39425691962242126,
+      "learning_rate": 9.86596074467221e-06,
+      "loss": 0.4188,
+      "step": 1986
+    },
+    {
+      "epoch": 0.49974849094567403,
+      "grad_norm": 0.39733070135116577,
+      "learning_rate": 9.865623998765253e-06,
+      "loss": 0.4111,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40979063510894775,
+      "learning_rate": 9.865286836148039e-06,
+      "loss": 0.3874,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5002515090543259,
+      "grad_norm": 0.361208438873291,
+      "learning_rate": 9.864949256849445e-06,
+      "loss": 0.4006,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5005030181086519,
+      "grad_norm": 0.42786529660224915,
+      "learning_rate": 9.864611260898383e-06,
+      "loss": 0.397,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5007545271629779,
+      "grad_norm": 0.40797531604766846,
+      "learning_rate": 9.8642728483238e-06,
+      "loss": 0.4219,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5010060362173038,
+      "grad_norm": 0.3617155849933624,
+      "learning_rate": 9.863934019154676e-06,
+      "loss": 0.4213,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5012575452716298,
+      "grad_norm": 0.39363011717796326,
+      "learning_rate": 9.863594773420033e-06,
+      "loss": 0.391,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5015090543259557,
+      "grad_norm": 0.36764830350875854,
+      "learning_rate": 9.863255111148925e-06,
+      "loss": 0.393,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5017605633802817,
+      "grad_norm": 0.377768337726593,
+      "learning_rate": 9.862915032370441e-06,
+      "loss": 0.4024,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5020120724346077,
+      "grad_norm": 0.3849732577800751,
+      "learning_rate": 9.862574537113705e-06,
+      "loss": 0.3818,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5022635814889336,
+      "grad_norm": 0.399882048368454,
+      "learning_rate": 9.862233625407882e-06,
+      "loss": 0.3862,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5025150905432596,
+      "grad_norm": 0.439528226852417,
+      "learning_rate": 9.861892297282167e-06,
+      "loss": 0.3957,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5027665995975855,
+      "grad_norm": 0.40875646471977234,
+      "learning_rate": 9.86155055276579e-06,
+      "loss": 0.3947,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5030181086519114,
+      "grad_norm": 0.41643694043159485,
+      "learning_rate": 9.861208391888024e-06,
+      "loss": 0.4071,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5032696177062375,
+      "grad_norm": 0.47003522515296936,
+      "learning_rate": 9.860865814678172e-06,
+      "loss": 0.3874,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5035211267605634,
+      "grad_norm": 0.3645997643470764,
+      "learning_rate": 9.860522821165572e-06,
+      "loss": 0.3991,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5037726358148893,
+      "grad_norm": 0.3944624960422516,
+      "learning_rate": 9.860179411379598e-06,
+      "loss": 0.3663,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5040241448692153,
+      "grad_norm": 0.41858211159706116,
+      "learning_rate": 9.859835585349664e-06,
+      "loss": 0.4056,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5042756539235412,
+      "grad_norm": 0.42395514249801636,
+      "learning_rate": 9.859491343105215e-06,
+      "loss": 0.4058,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5045271629778671,
+      "grad_norm": 0.41400447487831116,
+      "learning_rate": 9.859146684675733e-06,
+      "loss": 0.4235,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5047786720321932,
+      "grad_norm": 0.3939945101737976,
+      "learning_rate": 9.858801610090736e-06,
+      "loss": 0.3848,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5050301810865191,
+      "grad_norm": 0.4896427094936371,
+      "learning_rate": 9.858456119379779e-06,
+      "loss": 0.3592,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5052816901408451,
+      "grad_norm": 0.43367308378219604,
+      "learning_rate": 9.858110212572448e-06,
+      "loss": 0.4206,
+      "step": 2009
+    },
+    {
+      "epoch": 0.505533199195171,
+      "grad_norm": 0.36704427003860474,
+      "learning_rate": 9.85776388969837e-06,
+      "loss": 0.4041,
+      "step": 2010
+    },
+    {
+      "epoch": 0.505784708249497,
+      "grad_norm": 0.5114539265632629,
+      "learning_rate": 9.857417150787206e-06,
+      "loss": 0.3805,
+      "step": 2011
+    },
+    {
+      "epoch": 0.506036217303823,
+      "grad_norm": 0.4381665885448456,
+      "learning_rate": 9.857069995868648e-06,
+      "loss": 0.4056,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5062877263581489,
+      "grad_norm": 0.450435072183609,
+      "learning_rate": 9.856722424972434e-06,
+      "loss": 0.4096,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5065392354124748,
+      "grad_norm": 0.44745296239852905,
+      "learning_rate": 9.856374438128327e-06,
+      "loss": 0.4042,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5067907444668008,
+      "grad_norm": 0.4272843301296234,
+      "learning_rate": 9.85602603536613e-06,
+      "loss": 0.3735,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 0.3728364109992981,
+      "learning_rate": 9.855677216715682e-06,
+      "loss": 0.4043,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5072937625754527,
+      "grad_norm": 0.4467531740665436,
+      "learning_rate": 9.855327982206859e-06,
+      "loss": 0.3911,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5075452716297787,
+      "grad_norm": 0.4462815225124359,
+      "learning_rate": 9.854978331869568e-06,
+      "loss": 0.4317,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5077967806841046,
+      "grad_norm": 0.38220712542533875,
+      "learning_rate": 9.854628265733755e-06,
+      "loss": 0.4053,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5080482897384306,
+      "grad_norm": 0.3900271952152252,
+      "learning_rate": 9.854277783829402e-06,
+      "loss": 0.3861,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5082997987927566,
+      "grad_norm": 0.4110308587551117,
+      "learning_rate": 9.853926886186527e-06,
+      "loss": 0.3917,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5085513078470825,
+      "grad_norm": 0.40754181146621704,
+      "learning_rate": 9.853575572835179e-06,
+      "loss": 0.4176,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5088028169014085,
+      "grad_norm": 0.4448073208332062,
+      "learning_rate": 9.853223843805445e-06,
+      "loss": 0.3992,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5090543259557344,
+      "grad_norm": 0.40429484844207764,
+      "learning_rate": 9.852871699127453e-06,
+      "loss": 0.4232,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5093058350100603,
+      "grad_norm": 0.4180743992328644,
+      "learning_rate": 9.852519138831358e-06,
+      "loss": 0.4074,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5095573440643864,
+      "grad_norm": 0.46965914964675903,
+      "learning_rate": 9.852166162947356e-06,
+      "loss": 0.3685,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5098088531187123,
+      "grad_norm": 0.400019109249115,
+      "learning_rate": 9.851812771505678e-06,
+      "loss": 0.3915,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5100603621730382,
+      "grad_norm": 0.39726099371910095,
+      "learning_rate": 9.851458964536589e-06,
+      "loss": 0.3435,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5103118712273642,
+      "grad_norm": 0.41831856966018677,
+      "learning_rate": 9.85110474207039e-06,
+      "loss": 0.382,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5105633802816901,
+      "grad_norm": 0.41029006242752075,
+      "learning_rate": 9.85075010413742e-06,
+      "loss": 0.4017,
+      "step": 2030
+    },
+    {
+      "epoch": 0.510814889336016,
+      "grad_norm": 0.42589080333709717,
+      "learning_rate": 9.850395050768047e-06,
+      "loss": 0.3784,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5110663983903421,
+      "grad_norm": 0.4287776052951813,
+      "learning_rate": 9.850039581992683e-06,
+      "loss": 0.4288,
+      "step": 2032
+    },
+    {
+      "epoch": 0.511317907444668,
+      "grad_norm": 0.38227567076683044,
+      "learning_rate": 9.84968369784177e-06,
+      "loss": 0.4244,
+      "step": 2033
+    },
+    {
+      "epoch": 0.511569416498994,
+      "grad_norm": 0.40670591592788696,
+      "learning_rate": 9.849327398345788e-06,
+      "loss": 0.414,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5118209255533199,
+      "grad_norm": 0.4272759258747101,
+      "learning_rate": 9.848970683535253e-06,
+      "loss": 0.3923,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5120724346076458,
+      "grad_norm": 0.5142744779586792,
+      "learning_rate": 9.84861355344071e-06,
+      "loss": 0.3805,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5123239436619719,
+      "grad_norm": 0.38634923100471497,
+      "learning_rate": 9.848256008092754e-06,
+      "loss": 0.3946,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5125754527162978,
+      "grad_norm": 0.4425947964191437,
+      "learning_rate": 9.847898047522e-06,
+      "loss": 0.3706,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5128269617706237,
+      "grad_norm": 0.4305839240550995,
+      "learning_rate": 9.847539671759105e-06,
+      "loss": 0.3941,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5130784708249497,
+      "grad_norm": 0.4277026653289795,
+      "learning_rate": 9.847180880834764e-06,
+      "loss": 0.3712,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5133299798792756,
+      "grad_norm": 0.3730026185512543,
+      "learning_rate": 9.846821674779705e-06,
+      "loss": 0.3905,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5135814889336016,
+      "grad_norm": 0.39186757802963257,
+      "learning_rate": 9.846462053624691e-06,
+      "loss": 0.4005,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5138329979879276,
+      "grad_norm": 0.39070603251457214,
+      "learning_rate": 9.846102017400523e-06,
+      "loss": 0.3957,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5140845070422535,
+      "grad_norm": 0.3925558924674988,
+      "learning_rate": 9.845741566138031e-06,
+      "loss": 0.3974,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5143360160965795,
+      "grad_norm": 0.4226996600627899,
+      "learning_rate": 9.84538069986809e-06,
+      "loss": 0.4011,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5145875251509054,
+      "grad_norm": 0.39153674244880676,
+      "learning_rate": 9.845019418621606e-06,
+      "loss": 0.3668,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5148390342052314,
+      "grad_norm": 0.4205423891544342,
+      "learning_rate": 9.844657722429518e-06,
+      "loss": 0.402,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5150905432595574,
+      "grad_norm": 0.4090724289417267,
+      "learning_rate": 9.844295611322804e-06,
+      "loss": 0.4084,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5153420523138833,
+      "grad_norm": 0.38471901416778564,
+      "learning_rate": 9.843933085332477e-06,
+      "loss": 0.3758,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5155935613682092,
+      "grad_norm": 0.3938809037208557,
+      "learning_rate": 9.843570144489585e-06,
+      "loss": 0.3913,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5158450704225352,
+      "grad_norm": 0.36798256635665894,
+      "learning_rate": 9.843206788825211e-06,
+      "loss": 0.4112,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5160965794768612,
+      "grad_norm": 0.3941897749900818,
+      "learning_rate": 9.842843018370475e-06,
+      "loss": 0.4002,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5163480885311871,
+      "grad_norm": 0.4066377580165863,
+      "learning_rate": 9.84247883315653e-06,
+      "loss": 0.4136,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5165995975855131,
+      "grad_norm": 0.4132152795791626,
+      "learning_rate": 9.84211423321457e-06,
+      "loss": 0.4202,
+      "step": 2054
+    },
+    {
+      "epoch": 0.516851106639839,
+      "grad_norm": 0.3840596377849579,
+      "learning_rate": 9.841749218575815e-06,
+      "loss": 0.4002,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5171026156941649,
+      "grad_norm": 0.4097898304462433,
+      "learning_rate": 9.841383789271533e-06,
+      "loss": 0.4072,
+      "step": 2056
+    },
+    {
+      "epoch": 0.517354124748491,
+      "grad_norm": 0.39788705110549927,
+      "learning_rate": 9.841017945333014e-06,
+      "loss": 0.4225,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5176056338028169,
+      "grad_norm": 0.3664261996746063,
+      "learning_rate": 9.840651686791593e-06,
+      "loss": 0.4118,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5178571428571429,
+      "grad_norm": 0.31850335001945496,
+      "learning_rate": 9.84028501367864e-06,
+      "loss": 0.3815,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5181086519114688,
+      "grad_norm": 0.4274592697620392,
+      "learning_rate": 9.839917926025555e-06,
+      "loss": 0.4198,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5183601609657947,
+      "grad_norm": 0.3651498854160309,
+      "learning_rate": 9.839550423863779e-06,
+      "loss": 0.3958,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5186116700201208,
+      "grad_norm": 0.35910671949386597,
+      "learning_rate": 9.839182507224786e-06,
+      "loss": 0.3888,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5188631790744467,
+      "grad_norm": 0.39101144671440125,
+      "learning_rate": 9.838814176140084e-06,
+      "loss": 0.4144,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5191146881287726,
+      "grad_norm": 0.34476083517074585,
+      "learning_rate": 9.838445430641219e-06,
+      "loss": 0.4075,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5193661971830986,
+      "grad_norm": 0.3631290793418884,
+      "learning_rate": 9.838076270759771e-06,
+      "loss": 0.3737,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5196177062374245,
+      "grad_norm": 0.38996070623397827,
+      "learning_rate": 9.83770669652736e-06,
+      "loss": 0.3986,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5198692152917505,
+      "grad_norm": 0.35395678877830505,
+      "learning_rate": 9.837336707975633e-06,
+      "loss": 0.384,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5201207243460765,
+      "grad_norm": 0.3823234438896179,
+      "learning_rate": 9.83696630513628e-06,
+      "loss": 0.3907,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5203722334004024,
+      "grad_norm": 0.40023064613342285,
+      "learning_rate": 9.836595488041022e-06,
+      "loss": 0.3846,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5206237424547284,
+      "grad_norm": 0.4110654592514038,
+      "learning_rate": 9.83622425672162e-06,
+      "loss": 0.4082,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5208752515090543,
+      "grad_norm": 0.3667696416378021,
+      "learning_rate": 9.835852611209865e-06,
+      "loss": 0.4005,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5211267605633803,
+      "grad_norm": 0.39130160212516785,
+      "learning_rate": 9.835480551537587e-06,
+      "loss": 0.412,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5213782696177063,
+      "grad_norm": 0.4086337685585022,
+      "learning_rate": 9.83510807773665e-06,
+      "loss": 0.4241,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5216297786720322,
+      "grad_norm": 0.3477274179458618,
+      "learning_rate": 9.834735189838954e-06,
+      "loss": 0.4019,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5218812877263581,
+      "grad_norm": 0.4180651605129242,
+      "learning_rate": 9.834361887876436e-06,
+      "loss": 0.4283,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5221327967806841,
+      "grad_norm": 0.4306004047393799,
+      "learning_rate": 9.833988171881066e-06,
+      "loss": 0.4239,
+      "step": 2076
+    },
+    {
+      "epoch": 0.52238430583501,
+      "grad_norm": 0.38910388946533203,
+      "learning_rate": 9.83361404188485e-06,
+      "loss": 0.3697,
+      "step": 2077
+    },
+    {
+      "epoch": 0.522635814889336,
+      "grad_norm": 0.49108606576919556,
+      "learning_rate": 9.83323949791983e-06,
+      "loss": 0.4168,
+      "step": 2078
+    },
+    {
+      "epoch": 0.522887323943662,
+      "grad_norm": 0.42532941699028015,
+      "learning_rate": 9.832864540018083e-06,
+      "loss": 0.4159,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5231388329979879,
+      "grad_norm": 0.3561785817146301,
+      "learning_rate": 9.832489168211723e-06,
+      "loss": 0.3945,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5233903420523138,
+      "grad_norm": 0.3978264629840851,
+      "learning_rate": 9.832113382532899e-06,
+      "loss": 0.4115,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5236418511066399,
+      "grad_norm": 0.38779211044311523,
+      "learning_rate": 9.831737183013792e-06,
+      "loss": 0.3627,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5238933601609658,
+      "grad_norm": 0.3900914192199707,
+      "learning_rate": 9.831360569686623e-06,
+      "loss": 0.4184,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5241448692152918,
+      "grad_norm": 0.44345182180404663,
+      "learning_rate": 9.830983542583647e-06,
+      "loss": 0.4029,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5243963782696177,
+      "grad_norm": 0.3975575864315033,
+      "learning_rate": 9.830606101737153e-06,
+      "loss": 0.3758,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5246478873239436,
+      "grad_norm": 0.4029177129268646,
+      "learning_rate": 9.830228247179465e-06,
+      "loss": 0.4035,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5248993963782697,
+      "grad_norm": 0.44459268450737,
+      "learning_rate": 9.829849978942948e-06,
+      "loss": 0.4014,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5251509054325956,
+      "grad_norm": 0.39327511191368103,
+      "learning_rate": 9.829471297059991e-06,
+      "loss": 0.391,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5254024144869215,
+      "grad_norm": 0.37261277437210083,
+      "learning_rate": 9.829092201563035e-06,
+      "loss": 0.4013,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5256539235412475,
+      "grad_norm": 0.4421737790107727,
+      "learning_rate": 9.828712692484541e-06,
+      "loss": 0.3947,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5259054325955734,
+      "grad_norm": 0.35906025767326355,
+      "learning_rate": 9.828332769857014e-06,
+      "loss": 0.3556,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5261569416498993,
+      "grad_norm": 0.3702889084815979,
+      "learning_rate": 9.82795243371299e-06,
+      "loss": 0.4143,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5264084507042254,
+      "grad_norm": 0.41510313749313354,
+      "learning_rate": 9.827571684085045e-06,
+      "loss": 0.4048,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5266599597585513,
+      "grad_norm": 0.3920029103755951,
+      "learning_rate": 9.827190521005786e-06,
+      "loss": 0.4078,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5269114688128773,
+      "grad_norm": 0.4151366651058197,
+      "learning_rate": 9.826808944507855e-06,
+      "loss": 0.4112,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5271629778672032,
+      "grad_norm": 0.40691617131233215,
+      "learning_rate": 9.826426954623937e-06,
+      "loss": 0.3947,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5274144869215291,
+      "grad_norm": 0.3902684152126312,
+      "learning_rate": 9.826044551386743e-06,
+      "loss": 0.3756,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5276659959758552,
+      "grad_norm": 0.436042457818985,
+      "learning_rate": 9.825661734829027e-06,
+      "loss": 0.3787,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5279175050301811,
+      "grad_norm": 0.40818825364112854,
+      "learning_rate": 9.825278504983571e-06,
+      "loss": 0.408,
+      "step": 2099
+    },
+    {
+      "epoch": 0.528169014084507,
+      "grad_norm": 0.4207872152328491,
+      "learning_rate": 9.824894861883198e-06,
+      "loss": 0.3696,
+      "step": 2100
+    },
+    {
+      "epoch": 0.528420523138833,
+      "grad_norm": 0.40610066056251526,
+      "learning_rate": 9.824510805560765e-06,
+      "loss": 0.3853,
+      "step": 2101
+    },
+    {
+      "epoch": 0.528672032193159,
+      "grad_norm": 0.4684671461582184,
+      "learning_rate": 9.824126336049164e-06,
+      "loss": 0.3864,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5289235412474849,
+      "grad_norm": 0.41243529319763184,
+      "learning_rate": 9.823741453381322e-06,
+      "loss": 0.4081,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5291750503018109,
+      "grad_norm": 0.3892309069633484,
+      "learning_rate": 9.8233561575902e-06,
+      "loss": 0.3777,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5294265593561368,
+      "grad_norm": 0.4003976285457611,
+      "learning_rate": 9.822970448708799e-06,
+      "loss": 0.3686,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5296780684104627,
+      "grad_norm": 0.38374191522598267,
+      "learning_rate": 9.822584326770152e-06,
+      "loss": 0.3977,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5299295774647887,
+      "grad_norm": 0.4218411147594452,
+      "learning_rate": 9.822197791807328e-06,
+      "loss": 0.3836,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5301810865191147,
+      "grad_norm": 0.4624592959880829,
+      "learning_rate": 9.821810843853428e-06,
+      "loss": 0.4268,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5304325955734407,
+      "grad_norm": 0.3895992040634155,
+      "learning_rate": 9.821423482941597e-06,
+      "loss": 0.3982,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5306841046277666,
+      "grad_norm": 0.4353783428668976,
+      "learning_rate": 9.821035709105006e-06,
+      "loss": 0.4131,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5309356136820925,
+      "grad_norm": 0.4408473074436188,
+      "learning_rate": 9.820647522376868e-06,
+      "loss": 0.3785,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5311871227364185,
+      "grad_norm": 0.39889922738075256,
+      "learning_rate": 9.820258922790427e-06,
+      "loss": 0.4074,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5314386317907445,
+      "grad_norm": 0.43087536096572876,
+      "learning_rate": 9.819869910378964e-06,
+      "loss": 0.3995,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5316901408450704,
+      "grad_norm": 0.3881208002567291,
+      "learning_rate": 9.819480485175797e-06,
+      "loss": 0.4023,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5319416498993964,
+      "grad_norm": 0.39766907691955566,
+      "learning_rate": 9.819090647214277e-06,
+      "loss": 0.42,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5321931589537223,
+      "grad_norm": 0.42352211475372314,
+      "learning_rate": 9.818700396527791e-06,
+      "loss": 0.3988,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5324446680080482,
+      "grad_norm": 0.48607903718948364,
+      "learning_rate": 9.818309733149762e-06,
+      "loss": 0.384,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5326961770623743,
+      "grad_norm": 0.38936519622802734,
+      "learning_rate": 9.817918657113648e-06,
+      "loss": 0.4082,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5329476861167002,
+      "grad_norm": 0.4536868929862976,
+      "learning_rate": 9.81752716845294e-06,
+      "loss": 0.399,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5331991951710262,
+      "grad_norm": 0.42506319284439087,
+      "learning_rate": 9.81713526720117e-06,
+      "loss": 0.3987,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5334507042253521,
+      "grad_norm": 0.42283493280410767,
+      "learning_rate": 9.8167429533919e-06,
+      "loss": 0.3804,
+      "step": 2121
+    },
+    {
+      "epoch": 0.533702213279678,
+      "grad_norm": 0.4490463137626648,
+      "learning_rate": 9.816350227058728e-06,
+      "loss": 0.4026,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5339537223340041,
+      "grad_norm": 0.4041505455970764,
+      "learning_rate": 9.815957088235293e-06,
+      "loss": 0.3712,
+      "step": 2123
+    },
+    {
+      "epoch": 0.53420523138833,
+      "grad_norm": 0.3451807200908661,
+      "learning_rate": 9.81556353695526e-06,
+      "loss": 0.376,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5344567404426559,
+      "grad_norm": 0.3851028382778168,
+      "learning_rate": 9.815169573252336e-06,
+      "loss": 0.3902,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5347082494969819,
+      "grad_norm": 0.38843706250190735,
+      "learning_rate": 9.814775197160262e-06,
+      "loss": 0.4208,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5349597585513078,
+      "grad_norm": 0.3881021738052368,
+      "learning_rate": 9.814380408712813e-06,
+      "loss": 0.4059,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5352112676056338,
+      "grad_norm": 0.33903956413269043,
+      "learning_rate": 9.813985207943802e-06,
+      "loss": 0.3739,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5354627766599598,
+      "grad_norm": 0.36936119198799133,
+      "learning_rate": 9.813589594887074e-06,
+      "loss": 0.416,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 0.38683557510375977,
+      "learning_rate": 9.81319356957651e-06,
+      "loss": 0.3924,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5359657947686117,
+      "grad_norm": 0.3721410036087036,
+      "learning_rate": 9.812797132046028e-06,
+      "loss": 0.4107,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5362173038229376,
+      "grad_norm": 0.34902095794677734,
+      "learning_rate": 9.812400282329579e-06,
+      "loss": 0.3693,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5364688128772636,
+      "grad_norm": 0.39966878294944763,
+      "learning_rate": 9.812003020461155e-06,
+      "loss": 0.3925,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5367203219315896,
+      "grad_norm": 0.36834654211997986,
+      "learning_rate": 9.811605346474775e-06,
+      "loss": 0.3942,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5369718309859155,
+      "grad_norm": 0.37292787432670593,
+      "learning_rate": 9.811207260404499e-06,
+      "loss": 0.4042,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5372233400402414,
+      "grad_norm": 0.35834237933158875,
+      "learning_rate": 9.810808762284419e-06,
+      "loss": 0.3952,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5374748490945674,
+      "grad_norm": 0.40317603945732117,
+      "learning_rate": 9.810409852148665e-06,
+      "loss": 0.4013,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5377263581488934,
+      "grad_norm": 0.4264310598373413,
+      "learning_rate": 9.8100105300314e-06,
+      "loss": 0.4076,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5379778672032193,
+      "grad_norm": 0.42888686060905457,
+      "learning_rate": 9.809610795966826e-06,
+      "loss": 0.3646,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5382293762575453,
+      "grad_norm": 0.42549654841423035,
+      "learning_rate": 9.809210649989175e-06,
+      "loss": 0.4074,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5384808853118712,
+      "grad_norm": 0.4057452976703644,
+      "learning_rate": 9.80881009213272e-06,
+      "loss": 0.3669,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5387323943661971,
+      "grad_norm": 0.41238072514533997,
+      "learning_rate": 9.808409122431764e-06,
+      "loss": 0.401,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5389839034205232,
+      "grad_norm": 0.42542120814323425,
+      "learning_rate": 9.808007740920647e-06,
+      "loss": 0.3928,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5392354124748491,
+      "grad_norm": 0.37695184350013733,
+      "learning_rate": 9.807605947633745e-06,
+      "loss": 0.3742,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5394869215291751,
+      "grad_norm": 0.4471421539783478,
+      "learning_rate": 9.807203742605472e-06,
+      "loss": 0.3786,
+      "step": 2145
+    },
+    {
+      "epoch": 0.539738430583501,
+      "grad_norm": 0.45084428787231445,
+      "learning_rate": 9.80680112587027e-06,
+      "loss": 0.3869,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5399899396378269,
+      "grad_norm": 0.37542101740837097,
+      "learning_rate": 9.806398097462624e-06,
+      "loss": 0.3566,
+      "step": 2147
+    },
+    {
+      "epoch": 0.540241448692153,
+      "grad_norm": 0.4376560151576996,
+      "learning_rate": 9.805994657417049e-06,
+      "loss": 0.4074,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5404929577464789,
+      "grad_norm": 0.5008549094200134,
+      "learning_rate": 9.8055908057681e-06,
+      "loss": 0.3917,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5407444668008048,
+      "grad_norm": 0.37436121702194214,
+      "learning_rate": 9.80518654255036e-06,
+      "loss": 0.3788,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5409959758551308,
+      "grad_norm": 0.3661990463733673,
+      "learning_rate": 9.804781867798454e-06,
+      "loss": 0.3834,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5412474849094567,
+      "grad_norm": 0.45389556884765625,
+      "learning_rate": 9.804376781547041e-06,
+      "loss": 0.3688,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5414989939637826,
+      "grad_norm": 0.3825472593307495,
+      "learning_rate": 9.80397128383081e-06,
+      "loss": 0.3752,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5417505030181087,
+      "grad_norm": 0.3918491005897522,
+      "learning_rate": 9.803565374684494e-06,
+      "loss": 0.391,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5420020120724346,
+      "grad_norm": 0.3703051507472992,
+      "learning_rate": 9.803159054142855e-06,
+      "loss": 0.3866,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5422535211267606,
+      "grad_norm": 0.4313543140888214,
+      "learning_rate": 9.802752322240692e-06,
+      "loss": 0.4013,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5425050301810865,
+      "grad_norm": 0.3883597254753113,
+      "learning_rate": 9.802345179012837e-06,
+      "loss": 0.4049,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5427565392354124,
+      "grad_norm": 0.37119433283805847,
+      "learning_rate": 9.801937624494161e-06,
+      "loss": 0.3798,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5430080482897385,
+      "grad_norm": 0.3962958753108978,
+      "learning_rate": 9.801529658719568e-06,
+      "loss": 0.4036,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5432595573440644,
+      "grad_norm": 0.37826666235923767,
+      "learning_rate": 9.801121281724e-06,
+      "loss": 0.3931,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5435110663983903,
+      "grad_norm": 0.41150638461112976,
+      "learning_rate": 9.800712493542428e-06,
+      "loss": 0.3994,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5437625754527163,
+      "grad_norm": 0.3833552896976471,
+      "learning_rate": 9.800303294209865e-06,
+      "loss": 0.4038,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5440140845070423,
+      "grad_norm": 0.35805314779281616,
+      "learning_rate": 9.799893683761355e-06,
+      "loss": 0.3862,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5442655935613682,
+      "grad_norm": 0.39856329560279846,
+      "learning_rate": 9.79948366223198e-06,
+      "loss": 0.3888,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5445171026156942,
+      "grad_norm": 0.42184582352638245,
+      "learning_rate": 9.799073229656853e-06,
+      "loss": 0.3901,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5447686116700201,
+      "grad_norm": 0.36999794840812683,
+      "learning_rate": 9.798662386071127e-06,
+      "loss": 0.3794,
+      "step": 2166
+    },
+    {
+      "epoch": 0.545020120724346,
+      "grad_norm": 0.36188700795173645,
+      "learning_rate": 9.79825113150999e-06,
+      "loss": 0.4028,
+      "step": 2167
+    },
+    {
+      "epoch": 0.545271629778672,
+      "grad_norm": 0.39296042919158936,
+      "learning_rate": 9.797839466008659e-06,
+      "loss": 0.3667,
+      "step": 2168
+    },
+    {
+      "epoch": 0.545523138832998,
+      "grad_norm": 0.38612136244773865,
+      "learning_rate": 9.797427389602393e-06,
+      "loss": 0.4027,
+      "step": 2169
+    },
+    {
+      "epoch": 0.545774647887324,
+      "grad_norm": 0.3653174042701721,
+      "learning_rate": 9.797014902326487e-06,
+      "loss": 0.3901,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5460261569416499,
+      "grad_norm": 0.37336331605911255,
+      "learning_rate": 9.796602004216261e-06,
+      "loss": 0.3656,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5462776659959758,
+      "grad_norm": 0.3571634888648987,
+      "learning_rate": 9.796188695307083e-06,
+      "loss": 0.3877,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5465291750503019,
+      "grad_norm": 0.3366375267505646,
+      "learning_rate": 9.795774975634347e-06,
+      "loss": 0.3793,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5467806841046278,
+      "grad_norm": 0.41412922739982605,
+      "learning_rate": 9.795360845233485e-06,
+      "loss": 0.395,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5470321931589537,
+      "grad_norm": 0.3812969923019409,
+      "learning_rate": 9.794946304139969e-06,
+      "loss": 0.4144,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5472837022132797,
+      "grad_norm": 0.3604353368282318,
+      "learning_rate": 9.794531352389298e-06,
+      "loss": 0.3636,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5475352112676056,
+      "grad_norm": 0.3878379166126251,
+      "learning_rate": 9.794115990017012e-06,
+      "loss": 0.4151,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5477867203219315,
+      "grad_norm": 0.3887823820114136,
+      "learning_rate": 9.793700217058683e-06,
+      "loss": 0.4009,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5480382293762576,
+      "grad_norm": 0.37774476408958435,
+      "learning_rate": 9.793284033549919e-06,
+      "loss": 0.4105,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5482897384305835,
+      "grad_norm": 0.37441837787628174,
+      "learning_rate": 9.792867439526366e-06,
+      "loss": 0.4019,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5485412474849095,
+      "grad_norm": 0.37095755338668823,
+      "learning_rate": 9.792450435023699e-06,
+      "loss": 0.3687,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5487927565392354,
+      "grad_norm": 0.40394505858421326,
+      "learning_rate": 9.792033020077634e-06,
+      "loss": 0.4008,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5490442655935613,
+      "grad_norm": 0.38541585206985474,
+      "learning_rate": 9.79161519472392e-06,
+      "loss": 0.4145,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5492957746478874,
+      "grad_norm": 0.3427443504333496,
+      "learning_rate": 9.79119695899834e-06,
+      "loss": 0.3838,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5495472837022133,
+      "grad_norm": 0.3911786675453186,
+      "learning_rate": 9.790778312936715e-06,
+      "loss": 0.3893,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5497987927565392,
+      "grad_norm": 0.36838027834892273,
+      "learning_rate": 9.790359256574899e-06,
+      "loss": 0.3763,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5500503018108652,
+      "grad_norm": 0.3471270501613617,
+      "learning_rate": 9.78993978994878e-06,
+      "loss": 0.3924,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5503018108651911,
+      "grad_norm": 0.33406156301498413,
+      "learning_rate": 9.789519913094286e-06,
+      "loss": 0.3775,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5505533199195171,
+      "grad_norm": 0.36162400245666504,
+      "learning_rate": 9.789099626047372e-06,
+      "loss": 0.3834,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5508048289738431,
+      "grad_norm": 0.4071558713912964,
+      "learning_rate": 9.788678928844036e-06,
+      "loss": 0.3924,
+      "step": 2190
+    },
+    {
+      "epoch": 0.551056338028169,
+      "grad_norm": 0.3942394256591797,
+      "learning_rate": 9.788257821520308e-06,
+      "loss": 0.3962,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5513078470824949,
+      "grad_norm": 0.41246795654296875,
+      "learning_rate": 9.787836304112253e-06,
+      "loss": 0.3822,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5515593561368209,
+      "grad_norm": 0.3739457130432129,
+      "learning_rate": 9.78741437665597e-06,
+      "loss": 0.3988,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5518108651911469,
+      "grad_norm": 0.3873920440673828,
+      "learning_rate": 9.786992039187598e-06,
+      "loss": 0.4142,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5520623742454729,
+      "grad_norm": 0.3712387979030609,
+      "learning_rate": 9.786569291743305e-06,
+      "loss": 0.3964,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5523138832997988,
+      "grad_norm": 0.4185509979724884,
+      "learning_rate": 9.786146134359294e-06,
+      "loss": 0.4275,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5525653923541247,
+      "grad_norm": 0.3870210647583008,
+      "learning_rate": 9.785722567071811e-06,
+      "loss": 0.4037,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5528169014084507,
+      "grad_norm": 0.39058247208595276,
+      "learning_rate": 9.785298589917128e-06,
+      "loss": 0.4133,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5530684104627767,
+      "grad_norm": 0.39365655183792114,
+      "learning_rate": 9.784874202931558e-06,
+      "loss": 0.4017,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5533199195171026,
+      "grad_norm": 0.44448891282081604,
+      "learning_rate": 9.784449406151448e-06,
+      "loss": 0.3863,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5535714285714286,
+      "grad_norm": 0.3912818133831024,
+      "learning_rate": 9.784024199613176e-06,
+      "loss": 0.3809,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5538229376257545,
+      "grad_norm": 0.3555707633495331,
+      "learning_rate": 9.783598583353161e-06,
+      "loss": 0.4174,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5540744466800804,
+      "grad_norm": 0.3489442765712738,
+      "learning_rate": 9.783172557407852e-06,
+      "loss": 0.4051,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5543259557344065,
+      "grad_norm": 0.4021318554878235,
+      "learning_rate": 9.78274612181374e-06,
+      "loss": 0.3971,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5545774647887324,
+      "grad_norm": 0.3958930969238281,
+      "learning_rate": 9.78231927660734e-06,
+      "loss": 0.408,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5548289738430584,
+      "grad_norm": 0.38725799322128296,
+      "learning_rate": 9.781892021825215e-06,
+      "loss": 0.4053,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5550804828973843,
+      "grad_norm": 0.3717680275440216,
+      "learning_rate": 9.781464357503951e-06,
+      "loss": 0.4101,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5553319919517102,
+      "grad_norm": 0.42192167043685913,
+      "learning_rate": 9.781036283680179e-06,
+      "loss": 0.3914,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5555835010060363,
+      "grad_norm": 0.3718080520629883,
+      "learning_rate": 9.78060780039056e-06,
+      "loss": 0.3633,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5558350100603622,
+      "grad_norm": 0.38646015524864197,
+      "learning_rate": 9.780178907671788e-06,
+      "loss": 0.3974,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5560865191146881,
+      "grad_norm": 0.36778998374938965,
+      "learning_rate": 9.7797496055606e-06,
+      "loss": 0.4049,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5563380281690141,
+      "grad_norm": 0.3933306336402893,
+      "learning_rate": 9.779319894093759e-06,
+      "loss": 0.4053,
+      "step": 2212
+    },
+    {
+      "epoch": 0.55658953722334,
+      "grad_norm": 0.39641377329826355,
+      "learning_rate": 9.778889773308069e-06,
+      "loss": 0.4054,
+      "step": 2213
+    },
+    {
+      "epoch": 0.556841046277666,
+      "grad_norm": 0.44454824924468994,
+      "learning_rate": 9.778459243240365e-06,
+      "loss": 0.4066,
+      "step": 2214
+    },
+    {
+      "epoch": 0.557092555331992,
+      "grad_norm": 0.3420102894306183,
+      "learning_rate": 9.778028303927522e-06,
+      "loss": 0.3604,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5573440643863179,
+      "grad_norm": 0.4014604091644287,
+      "learning_rate": 9.777596955406446e-06,
+      "loss": 0.4214,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5575955734406438,
+      "grad_norm": 0.416327565908432,
+      "learning_rate": 9.77716519771408e-06,
+      "loss": 0.4317,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5578470824949698,
+      "grad_norm": 0.43000099062919617,
+      "learning_rate": 9.7767330308874e-06,
+      "loss": 0.4061,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5580985915492958,
+      "grad_norm": 0.3848433494567871,
+      "learning_rate": 9.776300454963417e-06,
+      "loss": 0.3885,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5583501006036218,
+      "grad_norm": 0.3921235203742981,
+      "learning_rate": 9.775867469979184e-06,
+      "loss": 0.4042,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5586016096579477,
+      "grad_norm": 0.3721983730792999,
+      "learning_rate": 9.775434075971777e-06,
+      "loss": 0.3859,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5588531187122736,
+      "grad_norm": 0.38795045018196106,
+      "learning_rate": 9.77500027297832e-06,
+      "loss": 0.3851,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5591046277665996,
+      "grad_norm": 0.3403348922729492,
+      "learning_rate": 9.774566061035957e-06,
+      "loss": 0.3788,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5593561368209256,
+      "grad_norm": 0.4166750907897949,
+      "learning_rate": 9.774131440181884e-06,
+      "loss": 0.3765,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5596076458752515,
+      "grad_norm": 0.4601471722126007,
+      "learning_rate": 9.77369641045332e-06,
+      "loss": 0.4047,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5598591549295775,
+      "grad_norm": 0.3640338182449341,
+      "learning_rate": 9.77326097188752e-06,
+      "loss": 0.4057,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5601106639839034,
+      "grad_norm": 0.49202394485473633,
+      "learning_rate": 9.77282512452178e-06,
+      "loss": 0.3838,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5603621730382293,
+      "grad_norm": 0.4419405162334442,
+      "learning_rate": 9.77238886839343e-06,
+      "loss": 0.4005,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5606136820925554,
+      "grad_norm": 0.43457552790641785,
+      "learning_rate": 9.771952203539826e-06,
+      "loss": 0.3764,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5608651911468813,
+      "grad_norm": 0.41830188035964966,
+      "learning_rate": 9.77151512999837e-06,
+      "loss": 0.3886,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5611167002012073,
+      "grad_norm": 0.42050424218177795,
+      "learning_rate": 9.771077647806494e-06,
+      "loss": 0.4007,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5613682092555332,
+      "grad_norm": 0.3640553057193756,
+      "learning_rate": 9.770639757001665e-06,
+      "loss": 0.4145,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5616197183098591,
+      "grad_norm": 0.3788280487060547,
+      "learning_rate": 9.770201457621386e-06,
+      "loss": 0.3933,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5618712273641852,
+      "grad_norm": 0.4044826924800873,
+      "learning_rate": 9.769762749703194e-06,
+      "loss": 0.3897,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5621227364185111,
+      "grad_norm": 0.3986685574054718,
+      "learning_rate": 9.769323633284662e-06,
+      "loss": 0.3978,
+      "step": 2235
+    },
+    {
+      "epoch": 0.562374245472837,
+      "grad_norm": 0.3790046274662018,
+      "learning_rate": 9.768884108403399e-06,
+      "loss": 0.3759,
+      "step": 2236
+    },
+    {
+      "epoch": 0.562625754527163,
+      "grad_norm": 0.382379412651062,
+      "learning_rate": 9.768444175097043e-06,
+      "loss": 0.3848,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5628772635814889,
+      "grad_norm": 0.3763681650161743,
+      "learning_rate": 9.768003833403278e-06,
+      "loss": 0.3983,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5631287726358148,
+      "grad_norm": 0.3796675503253937,
+      "learning_rate": 9.767563083359812e-06,
+      "loss": 0.4085,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 0.40289655327796936,
+      "learning_rate": 9.767121925004393e-06,
+      "loss": 0.4021,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5636317907444668,
+      "grad_norm": 0.3628096878528595,
+      "learning_rate": 9.766680358374805e-06,
+      "loss": 0.3883,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5638832997987927,
+      "grad_norm": 0.3843015730381012,
+      "learning_rate": 9.766238383508863e-06,
+      "loss": 0.4002,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5641348088531187,
+      "grad_norm": 0.3902144432067871,
+      "learning_rate": 9.765796000444423e-06,
+      "loss": 0.3985,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5643863179074446,
+      "grad_norm": 0.3879857063293457,
+      "learning_rate": 9.76535320921937e-06,
+      "loss": 0.396,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5646378269617707,
+      "grad_norm": 0.38419634103775024,
+      "learning_rate": 9.764910009871626e-06,
+      "loss": 0.3882,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5648893360160966,
+      "grad_norm": 0.4009936451911926,
+      "learning_rate": 9.76446640243915e-06,
+      "loss": 0.3931,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5651408450704225,
+      "grad_norm": 0.35781916975975037,
+      "learning_rate": 9.764022386959931e-06,
+      "loss": 0.3602,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5653923541247485,
+      "grad_norm": 0.3593212366104126,
+      "learning_rate": 9.763577963472e-06,
+      "loss": 0.3984,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5656438631790744,
+      "grad_norm": 0.41306471824645996,
+      "learning_rate": 9.763133132013415e-06,
+      "loss": 0.4013,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5658953722334004,
+      "grad_norm": 0.4159875810146332,
+      "learning_rate": 9.762687892622278e-06,
+      "loss": 0.3889,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5661468812877264,
+      "grad_norm": 0.4286075532436371,
+      "learning_rate": 9.762242245336718e-06,
+      "loss": 0.418,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5663983903420523,
+      "grad_norm": 0.41156652569770813,
+      "learning_rate": 9.7617961901949e-06,
+      "loss": 0.4001,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5666498993963782,
+      "grad_norm": 0.4073083698749542,
+      "learning_rate": 9.76134972723503e-06,
+      "loss": 0.3782,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5669014084507042,
+      "grad_norm": 0.43342074751853943,
+      "learning_rate": 9.76090285649534e-06,
+      "loss": 0.4144,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5671529175050302,
+      "grad_norm": 0.37153980135917664,
+      "learning_rate": 9.760455578014107e-06,
+      "loss": 0.4146,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5674044265593562,
+      "grad_norm": 0.4138156473636627,
+      "learning_rate": 9.760007891829635e-06,
+      "loss": 0.411,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5676559356136821,
+      "grad_norm": 0.39025819301605225,
+      "learning_rate": 9.759559797980265e-06,
+      "loss": 0.3999,
+      "step": 2257
+    },
+    {
+      "epoch": 0.567907444668008,
+      "grad_norm": 0.3810308873653412,
+      "learning_rate": 9.759111296504374e-06,
+      "loss": 0.3973,
+      "step": 2258
+    },
+    {
+      "epoch": 0.568158953722334,
+      "grad_norm": 0.38749489188194275,
+      "learning_rate": 9.758662387440374e-06,
+      "loss": 0.3751,
+      "step": 2259
+    },
+    {
+      "epoch": 0.56841046277666,
+      "grad_norm": 0.42117783427238464,
+      "learning_rate": 9.75821307082671e-06,
+      "loss": 0.3943,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5686619718309859,
+      "grad_norm": 0.38864991068840027,
+      "learning_rate": 9.757763346701863e-06,
+      "loss": 0.3931,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5689134808853119,
+      "grad_norm": 0.4100189805030823,
+      "learning_rate": 9.757313215104352e-06,
+      "loss": 0.4236,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5691649899396378,
+      "grad_norm": 0.36280521750450134,
+      "learning_rate": 9.756862676072724e-06,
+      "loss": 0.3773,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5694164989939637,
+      "grad_norm": 0.4067562222480774,
+      "learning_rate": 9.756411729645567e-06,
+      "loss": 0.388,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5696680080482898,
+      "grad_norm": 0.3898506164550781,
+      "learning_rate": 9.755960375861502e-06,
+      "loss": 0.3675,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5699195171026157,
+      "grad_norm": 0.38783085346221924,
+      "learning_rate": 9.755508614759183e-06,
+      "loss": 0.4199,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5701710261569416,
+      "grad_norm": 0.3513777256011963,
+      "learning_rate": 9.755056446377302e-06,
+      "loss": 0.3937,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5704225352112676,
+      "grad_norm": 0.39856287837028503,
+      "learning_rate": 9.754603870754584e-06,
+      "loss": 0.407,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5706740442655935,
+      "grad_norm": 0.39470675587654114,
+      "learning_rate": 9.754150887929789e-06,
+      "loss": 0.3876,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5709255533199196,
+      "grad_norm": 0.43347692489624023,
+      "learning_rate": 9.753697497941713e-06,
+      "loss": 0.4072,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5711770623742455,
+      "grad_norm": 0.41265973448753357,
+      "learning_rate": 9.753243700829185e-06,
+      "loss": 0.3898,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.4033988416194916,
+      "learning_rate": 9.75278949663107e-06,
+      "loss": 0.3958,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5716800804828974,
+      "grad_norm": 0.4230313301086426,
+      "learning_rate": 9.752334885386268e-06,
+      "loss": 0.3657,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5719315895372233,
+      "grad_norm": 0.4359712600708008,
+      "learning_rate": 9.751879867133713e-06,
+      "loss": 0.4302,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5721830985915493,
+      "grad_norm": 0.3776133060455322,
+      "learning_rate": 9.751424441912376e-06,
+      "loss": 0.3786,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5724346076458753,
+      "grad_norm": 0.39019957184791565,
+      "learning_rate": 9.75096860976126e-06,
+      "loss": 0.3824,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5726861167002012,
+      "grad_norm": 0.38191524147987366,
+      "learning_rate": 9.750512370719404e-06,
+      "loss": 0.4106,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5729376257545271,
+      "grad_norm": 0.35566771030426025,
+      "learning_rate": 9.750055724825885e-06,
+      "loss": 0.387,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5731891348088531,
+      "grad_norm": 0.37126556038856506,
+      "learning_rate": 9.749598672119807e-06,
+      "loss": 0.4093,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5734406438631791,
+      "grad_norm": 0.3403175473213196,
+      "learning_rate": 9.749141212640317e-06,
+      "loss": 0.3993,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5736921529175051,
+      "grad_norm": 0.37190091609954834,
+      "learning_rate": 9.748683346426591e-06,
+      "loss": 0.3956,
+      "step": 2281
+    },
+    {
+      "epoch": 0.573943661971831,
+      "grad_norm": 0.3574543595314026,
+      "learning_rate": 9.748225073517845e-06,
+      "loss": 0.4014,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5741951710261569,
+      "grad_norm": 0.4083276689052582,
+      "learning_rate": 9.747766393953327e-06,
+      "loss": 0.424,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5744466800804829,
+      "grad_norm": 0.35851186513900757,
+      "learning_rate": 9.74730730777232e-06,
+      "loss": 0.3974,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5746981891348089,
+      "grad_norm": 0.3459572494029999,
+      "learning_rate": 9.74684781501414e-06,
+      "loss": 0.384,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5749496981891348,
+      "grad_norm": 0.38862690329551697,
+      "learning_rate": 9.746387915718139e-06,
+      "loss": 0.4025,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5752012072434608,
+      "grad_norm": 0.3741775155067444,
+      "learning_rate": 9.745927609923709e-06,
+      "loss": 0.3985,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5754527162977867,
+      "grad_norm": 0.33961108326911926,
+      "learning_rate": 9.745466897670268e-06,
+      "loss": 0.3606,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5757042253521126,
+      "grad_norm": 0.3567553460597992,
+      "learning_rate": 9.745005778997277e-06,
+      "loss": 0.3874,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5759557344064387,
+      "grad_norm": 0.4063310921192169,
+      "learning_rate": 9.744544253944223e-06,
+      "loss": 0.385,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5762072434607646,
+      "grad_norm": 0.3446536660194397,
+      "learning_rate": 9.744082322550637e-06,
+      "loss": 0.4224,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5764587525150905,
+      "grad_norm": 0.3623557984828949,
+      "learning_rate": 9.743619984856078e-06,
+      "loss": 0.3903,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5767102615694165,
+      "grad_norm": 0.37968283891677856,
+      "learning_rate": 9.743157240900145e-06,
+      "loss": 0.404,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5769617706237424,
+      "grad_norm": 0.37783414125442505,
+      "learning_rate": 9.742694090722466e-06,
+      "loss": 0.3811,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5772132796780685,
+      "grad_norm": 0.36868754029273987,
+      "learning_rate": 9.742230534362708e-06,
+      "loss": 0.3869,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5774647887323944,
+      "grad_norm": 0.4327180087566376,
+      "learning_rate": 9.741766571860573e-06,
+      "loss": 0.399,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5777162977867203,
+      "grad_norm": 0.4020644724369049,
+      "learning_rate": 9.741302203255796e-06,
+      "loss": 0.4062,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5779678068410463,
+      "grad_norm": 0.3767627477645874,
+      "learning_rate": 9.740837428588147e-06,
+      "loss": 0.3712,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5782193158953722,
+      "grad_norm": 0.428296834230423,
+      "learning_rate": 9.740372247897429e-06,
+      "loss": 0.3948,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5784708249496981,
+      "grad_norm": 0.3837886154651642,
+      "learning_rate": 9.739906661223485e-06,
+      "loss": 0.3894,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5787223340040242,
+      "grad_norm": 0.38341113924980164,
+      "learning_rate": 9.739440668606188e-06,
+      "loss": 0.3709,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5789738430583501,
+      "grad_norm": 0.4215813875198364,
+      "learning_rate": 9.738974270085447e-06,
+      "loss": 0.3978,
+      "step": 2302
+    },
+    {
+      "epoch": 0.579225352112676,
+      "grad_norm": 0.4352063834667206,
+      "learning_rate": 9.738507465701207e-06,
+      "loss": 0.3826,
+      "step": 2303
+    },
+    {
+      "epoch": 0.579476861167002,
+      "grad_norm": 0.36459967494010925,
+      "learning_rate": 9.738040255493446e-06,
+      "loss": 0.3813,
+      "step": 2304
+    },
+    {
+      "epoch": 0.579728370221328,
+      "grad_norm": 0.37163200974464417,
+      "learning_rate": 9.737572639502179e-06,
+      "loss": 0.3872,
+      "step": 2305
+    },
+    {
+      "epoch": 0.579979879275654,
+      "grad_norm": 0.4496228098869324,
+      "learning_rate": 9.737104617767454e-06,
+      "loss": 0.4004,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5802313883299799,
+      "grad_norm": 0.4434552490711212,
+      "learning_rate": 9.736636190329355e-06,
+      "loss": 0.3779,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5804828973843058,
+      "grad_norm": 0.3943467438220978,
+      "learning_rate": 9.736167357227995e-06,
+      "loss": 0.3992,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5807344064386318,
+      "grad_norm": 0.41185152530670166,
+      "learning_rate": 9.735698118503531e-06,
+      "loss": 0.3785,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5809859154929577,
+      "grad_norm": 0.4015391170978546,
+      "learning_rate": 9.735228474196152e-06,
+      "loss": 0.399,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5812374245472837,
+      "grad_norm": 0.43395721912384033,
+      "learning_rate": 9.734758424346075e-06,
+      "loss": 0.3948,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5814889336016097,
+      "grad_norm": 0.41467130184173584,
+      "learning_rate": 9.734287968993561e-06,
+      "loss": 0.4115,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5817404426559356,
+      "grad_norm": 0.4236811697483063,
+      "learning_rate": 9.7338171081789e-06,
+      "loss": 0.4043,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5819919517102615,
+      "grad_norm": 0.38736096024513245,
+      "learning_rate": 9.733345841942418e-06,
+      "loss": 0.3729,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5822434607645876,
+      "grad_norm": 0.37845367193222046,
+      "learning_rate": 9.732874170324479e-06,
+      "loss": 0.4106,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5824949698189135,
+      "grad_norm": 0.35245969891548157,
+      "learning_rate": 9.732402093365471e-06,
+      "loss": 0.4144,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5827464788732394,
+      "grad_norm": 0.39204463362693787,
+      "learning_rate": 9.731929611105833e-06,
+      "loss": 0.4053,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5829979879275654,
+      "grad_norm": 0.389548122882843,
+      "learning_rate": 9.731456723586026e-06,
+      "loss": 0.4192,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5832494969818913,
+      "grad_norm": 0.38718488812446594,
+      "learning_rate": 9.73098343084655e-06,
+      "loss": 0.3687,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5835010060362174,
+      "grad_norm": 0.409236341714859,
+      "learning_rate": 9.73050973292794e-06,
+      "loss": 0.3934,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5837525150905433,
+      "grad_norm": 0.41710254549980164,
+      "learning_rate": 9.730035629870766e-06,
+      "loss": 0.375,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5840040241448692,
+      "grad_norm": 0.421811580657959,
+      "learning_rate": 9.729561121715632e-06,
+      "loss": 0.3895,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5842555331991952,
+      "grad_norm": 0.39268958568573,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.3992,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5845070422535211,
+      "grad_norm": 0.3878178298473358,
+      "learning_rate": 9.728610890274068e-06,
+      "loss": 0.3971,
+      "step": 2324
+    },
+    {
+      "epoch": 0.584758551307847,
+      "grad_norm": 0.4158124029636383,
+      "learning_rate": 9.728135167069022e-06,
+      "loss": 0.3639,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5850100603621731,
+      "grad_norm": 0.3816595673561096,
+      "learning_rate": 9.727659038928778e-06,
+      "loss": 0.3709,
+      "step": 2326
+    },
+    {
+      "epoch": 0.585261569416499,
+      "grad_norm": 0.40008431673049927,
+      "learning_rate": 9.727182505894112e-06,
+      "loss": 0.3762,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5855130784708249,
+      "grad_norm": 0.40834906697273254,
+      "learning_rate": 9.72670556800584e-06,
+      "loss": 0.3712,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5857645875251509,
+      "grad_norm": 0.37313729524612427,
+      "learning_rate": 9.726228225304806e-06,
+      "loss": 0.3801,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5860160965794768,
+      "grad_norm": 0.4533138871192932,
+      "learning_rate": 9.72575047783189e-06,
+      "loss": 0.4047,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5862676056338029,
+      "grad_norm": 0.3437679708003998,
+      "learning_rate": 9.72527232562801e-06,
+      "loss": 0.4038,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5865191146881288,
+      "grad_norm": 0.45910757780075073,
+      "learning_rate": 9.724793768734117e-06,
+      "loss": 0.4086,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5867706237424547,
+      "grad_norm": 0.41133296489715576,
+      "learning_rate": 9.724314807191197e-06,
+      "loss": 0.386,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5870221327967807,
+      "grad_norm": 0.39628586173057556,
+      "learning_rate": 9.723835441040268e-06,
+      "loss": 0.4064,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5872736418511066,
+      "grad_norm": 0.4965059459209442,
+      "learning_rate": 9.723355670322385e-06,
+      "loss": 0.416,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5875251509054326,
+      "grad_norm": 0.4447380304336548,
+      "learning_rate": 9.722875495078638e-06,
+      "loss": 0.4007,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5877766599597586,
+      "grad_norm": 0.3534230887889862,
+      "learning_rate": 9.722394915350153e-06,
+      "loss": 0.3665,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5880281690140845,
+      "grad_norm": 0.5113065838813782,
+      "learning_rate": 9.721913931178084e-06,
+      "loss": 0.3697,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5882796780684104,
+      "grad_norm": 0.37465915083885193,
+      "learning_rate": 9.72143254260363e-06,
+      "loss": 0.3943,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5885311871227364,
+      "grad_norm": 0.4188118278980255,
+      "learning_rate": 9.720950749668013e-06,
+      "loss": 0.3736,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5887826961770624,
+      "grad_norm": 0.4283120036125183,
+      "learning_rate": 9.720468552412501e-06,
+      "loss": 0.3916,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5890342052313883,
+      "grad_norm": 0.4219818115234375,
+      "learning_rate": 9.719985950878386e-06,
+      "loss": 0.37,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5892857142857143,
+      "grad_norm": 0.41337502002716064,
+      "learning_rate": 9.719502945107004e-06,
+      "loss": 0.3874,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5895372233400402,
+      "grad_norm": 0.3993411660194397,
+      "learning_rate": 9.71901953513972e-06,
+      "loss": 0.3988,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5897887323943662,
+      "grad_norm": 0.4108680188655853,
+      "learning_rate": 9.718535721017936e-06,
+      "loss": 0.4016,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5900402414486922,
+      "grad_norm": 0.4107624888420105,
+      "learning_rate": 9.718051502783084e-06,
+      "loss": 0.3762,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5902917505030181,
+      "grad_norm": 0.4297145903110504,
+      "learning_rate": 9.717566880476639e-06,
+      "loss": 0.3873,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5905432595573441,
+      "grad_norm": 0.42795687913894653,
+      "learning_rate": 9.717081854140103e-06,
+      "loss": 0.4027,
+      "step": 2348
+    },
+    {
+      "epoch": 0.59079476861167,
+      "grad_norm": 0.42112451791763306,
+      "learning_rate": 9.716596423815016e-06,
+      "loss": 0.4161,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5910462776659959,
+      "grad_norm": 0.46127063035964966,
+      "learning_rate": 9.716110589542952e-06,
+      "loss": 0.3833,
+      "step": 2350
+    },
+    {
+      "epoch": 0.591297786720322,
+      "grad_norm": 0.4687446355819702,
+      "learning_rate": 9.71562435136552e-06,
+      "loss": 0.391,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5915492957746479,
+      "grad_norm": 0.4273424446582794,
+      "learning_rate": 9.715137709324363e-06,
+      "loss": 0.3899,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5918008048289738,
+      "grad_norm": 0.4016727805137634,
+      "learning_rate": 9.71465066346116e-06,
+      "loss": 0.3874,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5920523138832998,
+      "grad_norm": 0.4089299440383911,
+      "learning_rate": 9.714163213817621e-06,
+      "loss": 0.3946,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5923038229376257,
+      "grad_norm": 0.3954598903656006,
+      "learning_rate": 9.713675360435495e-06,
+      "loss": 0.393,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5925553319919518,
+      "grad_norm": 0.42446476221084595,
+      "learning_rate": 9.713187103356563e-06,
+      "loss": 0.3604,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5928068410462777,
+      "grad_norm": 0.461434543132782,
+      "learning_rate": 9.71269844262264e-06,
+      "loss": 0.3915,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5930583501006036,
+      "grad_norm": 0.3812006711959839,
+      "learning_rate": 9.712209378275581e-06,
+      "loss": 0.3874,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5933098591549296,
+      "grad_norm": 0.44694066047668457,
+      "learning_rate": 9.711719910357267e-06,
+      "loss": 0.3899,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5935613682092555,
+      "grad_norm": 0.39847058057785034,
+      "learning_rate": 9.711230038909619e-06,
+      "loss": 0.4042,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5938128772635815,
+      "grad_norm": 0.3936542570590973,
+      "learning_rate": 9.71073976397459e-06,
+      "loss": 0.4044,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5940643863179075,
+      "grad_norm": 0.38293659687042236,
+      "learning_rate": 9.710249085594171e-06,
+      "loss": 0.3467,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5943158953722334,
+      "grad_norm": 0.4120056927204132,
+      "learning_rate": 9.709758003810388e-06,
+      "loss": 0.4156,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5945674044265593,
+      "grad_norm": 0.40175196528434753,
+      "learning_rate": 9.709266518665293e-06,
+      "loss": 0.3832,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5948189134808853,
+      "grad_norm": 0.3934299051761627,
+      "learning_rate": 9.708774630200983e-06,
+      "loss": 0.3808,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5950704225352113,
+      "grad_norm": 0.3721473515033722,
+      "learning_rate": 9.708282338459582e-06,
+      "loss": 0.3742,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5953219315895373,
+      "grad_norm": 0.34081289172172546,
+      "learning_rate": 9.707789643483256e-06,
+      "loss": 0.4101,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5955734406438632,
+      "grad_norm": 0.36988210678100586,
+      "learning_rate": 9.707296545314197e-06,
+      "loss": 0.4075,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5958249496981891,
+      "grad_norm": 0.3822750747203827,
+      "learning_rate": 9.70680304399464e-06,
+      "loss": 0.3974,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5960764587525151,
+      "grad_norm": 0.3905336856842041,
+      "learning_rate": 9.706309139566847e-06,
+      "loss": 0.3861,
+      "step": 2370
+    },
+    {
+      "epoch": 0.596327967806841,
+      "grad_norm": 0.3887244760990143,
+      "learning_rate": 9.705814832073118e-06,
+      "loss": 0.3831,
+      "step": 2371
+    },
+    {
+      "epoch": 0.596579476861167,
+      "grad_norm": 0.38636234402656555,
+      "learning_rate": 9.705320121555789e-06,
+      "loss": 0.3969,
+      "step": 2372
+    },
+    {
+      "epoch": 0.596830985915493,
+      "grad_norm": 0.3817340135574341,
+      "learning_rate": 9.704825008057229e-06,
+      "loss": 0.3942,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5970824949698189,
+      "grad_norm": 0.41212648153305054,
+      "learning_rate": 9.704329491619837e-06,
+      "loss": 0.3865,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5973340040241448,
+      "grad_norm": 0.3645521402359009,
+      "learning_rate": 9.703833572286056e-06,
+      "loss": 0.389,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5975855130784709,
+      "grad_norm": 0.4224928319454193,
+      "learning_rate": 9.703337250098357e-06,
+      "loss": 0.3855,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5978370221327968,
+      "grad_norm": 0.40865787863731384,
+      "learning_rate": 9.702840525099247e-06,
+      "loss": 0.3671,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5980885311871227,
+      "grad_norm": 0.348753422498703,
+      "learning_rate": 9.702343397331266e-06,
+      "loss": 0.3635,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5983400402414487,
+      "grad_norm": 0.4527108669281006,
+      "learning_rate": 9.701845866836992e-06,
+      "loss": 0.4217,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5985915492957746,
+      "grad_norm": 0.4157269597053528,
+      "learning_rate": 9.70134793365903e-06,
+      "loss": 0.3824,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5988430583501007,
+      "grad_norm": 0.35614490509033203,
+      "learning_rate": 9.700849597840035e-06,
+      "loss": 0.4004,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5990945674044266,
+      "grad_norm": 0.41033607721328735,
+      "learning_rate": 9.700350859422675e-06,
+      "loss": 0.3927,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5993460764587525,
+      "grad_norm": 0.380023717880249,
+      "learning_rate": 9.699851718449672e-06,
+      "loss": 0.4076,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5995975855130785,
+      "grad_norm": 0.3968260586261749,
+      "learning_rate": 9.699352174963772e-06,
+      "loss": 0.3977,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5998490945674044,
+      "grad_norm": 0.3838002383708954,
+      "learning_rate": 9.698852229007756e-06,
+      "loss": 0.401,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6001006036217303,
+      "grad_norm": 0.4010978043079376,
+      "learning_rate": 9.698351880624444e-06,
+      "loss": 0.3975,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6003521126760564,
+      "grad_norm": 0.4017295241355896,
+      "learning_rate": 9.697851129856687e-06,
+      "loss": 0.3984,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6006036217303823,
+      "grad_norm": 0.4557097554206848,
+      "learning_rate": 9.697349976747366e-06,
+      "loss": 0.388,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6008551307847082,
+      "grad_norm": 0.3855472803115845,
+      "learning_rate": 9.696848421339409e-06,
+      "loss": 0.4105,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6011066398390342,
+      "grad_norm": 0.37822774052619934,
+      "learning_rate": 9.696346463675767e-06,
+      "loss": 0.3676,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6013581488933601,
+      "grad_norm": 0.41425397992134094,
+      "learning_rate": 9.695844103799432e-06,
+      "loss": 0.3996,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6016096579476862,
+      "grad_norm": 0.3729078769683838,
+      "learning_rate": 9.695341341753426e-06,
+      "loss": 0.3881,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6018611670020121,
+      "grad_norm": 0.41532474756240845,
+      "learning_rate": 9.69483817758081e-06,
+      "loss": 0.397,
+      "step": 2393
+    },
+    {
+      "epoch": 0.602112676056338,
+      "grad_norm": 0.389249324798584,
+      "learning_rate": 9.694334611324672e-06,
+      "loss": 0.3902,
+      "step": 2394
+    },
+    {
+      "epoch": 0.602364185110664,
+      "grad_norm": 0.3921426236629486,
+      "learning_rate": 9.693830643028142e-06,
+      "loss": 0.4204,
+      "step": 2395
+    },
+    {
+      "epoch": 0.60261569416499,
+      "grad_norm": 0.3730478584766388,
+      "learning_rate": 9.693326272734384e-06,
+      "loss": 0.384,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6028672032193159,
+      "grad_norm": 0.3994380831718445,
+      "learning_rate": 9.692821500486592e-06,
+      "loss": 0.4142,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6031187122736419,
+      "grad_norm": 0.3791671395301819,
+      "learning_rate": 9.692316326327995e-06,
+      "loss": 0.3905,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6033702213279678,
+      "grad_norm": 0.3484344780445099,
+      "learning_rate": 9.69181075030186e-06,
+      "loss": 0.4205,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6036217303822937,
+      "grad_norm": 0.3879513144493103,
+      "learning_rate": 9.691304772451487e-06,
+      "loss": 0.3693,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6038732394366197,
+      "grad_norm": 0.4011459946632385,
+      "learning_rate": 9.690798392820208e-06,
+      "loss": 0.4037,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6041247484909457,
+      "grad_norm": 0.3635562062263489,
+      "learning_rate": 9.690291611451394e-06,
+      "loss": 0.391,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6043762575452716,
+      "grad_norm": 0.37556272745132446,
+      "learning_rate": 9.689784428388444e-06,
+      "loss": 0.4034,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6046277665995976,
+      "grad_norm": 0.38869237899780273,
+      "learning_rate": 9.689276843674797e-06,
+      "loss": 0.3951,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6048792756539235,
+      "grad_norm": 0.41964560747146606,
+      "learning_rate": 9.688768857353925e-06,
+      "loss": 0.4044,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6051307847082495,
+      "grad_norm": 0.40390586853027344,
+      "learning_rate": 9.688260469469333e-06,
+      "loss": 0.4318,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6053822937625755,
+      "grad_norm": 0.38943222165107727,
+      "learning_rate": 9.687751680064562e-06,
+      "loss": 0.4171,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6056338028169014,
+      "grad_norm": 0.37889039516448975,
+      "learning_rate": 9.687242489183187e-06,
+      "loss": 0.3723,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6058853118712274,
+      "grad_norm": 0.3625709116458893,
+      "learning_rate": 9.686732896868814e-06,
+      "loss": 0.3664,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6061368209255533,
+      "grad_norm": 0.37902170419692993,
+      "learning_rate": 9.68622290316509e-06,
+      "loss": 0.3782,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6063883299798792,
+      "grad_norm": 0.35425660014152527,
+      "learning_rate": 9.68571250811569e-06,
+      "loss": 0.3889,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6066398390342053,
+      "grad_norm": 0.3910084068775177,
+      "learning_rate": 9.685201711764328e-06,
+      "loss": 0.3994,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6068913480885312,
+      "grad_norm": 0.3474578261375427,
+      "learning_rate": 9.68469051415475e-06,
+      "loss": 0.3665,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6071428571428571,
+      "grad_norm": 0.37786930799484253,
+      "learning_rate": 9.684178915330736e-06,
+      "loss": 0.4192,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6073943661971831,
+      "grad_norm": 0.34964871406555176,
+      "learning_rate": 9.683666915336102e-06,
+      "loss": 0.3848,
+      "step": 2415
+    },
+    {
+      "epoch": 0.607645875251509,
+      "grad_norm": 0.37684234976768494,
+      "learning_rate": 9.683154514214698e-06,
+      "loss": 0.3607,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6078973843058351,
+      "grad_norm": 0.41027969121932983,
+      "learning_rate": 9.68264171201041e-06,
+      "loss": 0.4109,
+      "step": 2417
+    },
+    {
+      "epoch": 0.608148893360161,
+      "grad_norm": 0.3731859028339386,
+      "learning_rate": 9.682128508767151e-06,
+      "loss": 0.4011,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6084004024144869,
+      "grad_norm": 0.3846946954727173,
+      "learning_rate": 9.681614904528877e-06,
+      "loss": 0.3984,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6086519114688129,
+      "grad_norm": 0.378128319978714,
+      "learning_rate": 9.681100899339574e-06,
+      "loss": 0.396,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6089034205231388,
+      "grad_norm": 0.39976274967193604,
+      "learning_rate": 9.680586493243265e-06,
+      "loss": 0.3737,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6091549295774648,
+      "grad_norm": 0.38559094071388245,
+      "learning_rate": 9.680071686284005e-06,
+      "loss": 0.385,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6094064386317908,
+      "grad_norm": 0.43395066261291504,
+      "learning_rate": 9.679556478505882e-06,
+      "loss": 0.3969,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6096579476861167,
+      "grad_norm": 0.3740188181400299,
+      "learning_rate": 9.679040869953023e-06,
+      "loss": 0.3783,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6099094567404426,
+      "grad_norm": 0.38674283027648926,
+      "learning_rate": 9.678524860669584e-06,
+      "loss": 0.4026,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6101609657947686,
+      "grad_norm": 0.399495929479599,
+      "learning_rate": 9.678008450699761e-06,
+      "loss": 0.3948,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6104124748490946,
+      "grad_norm": 0.390152245759964,
+      "learning_rate": 9.677491640087779e-06,
+      "loss": 0.3892,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6106639839034205,
+      "grad_norm": 0.36457139253616333,
+      "learning_rate": 9.6769744288779e-06,
+      "loss": 0.3923,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6109154929577465,
+      "grad_norm": 0.3784310519695282,
+      "learning_rate": 9.676456817114423e-06,
+      "loss": 0.3828,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6111670020120724,
+      "grad_norm": 0.41220057010650635,
+      "learning_rate": 9.675938804841673e-06,
+      "loss": 0.4132,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6114185110663984,
+      "grad_norm": 0.3676017224788666,
+      "learning_rate": 9.675420392104016e-06,
+      "loss": 0.3921,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6116700201207244,
+      "grad_norm": 0.3908953070640564,
+      "learning_rate": 9.674901578945853e-06,
+      "loss": 0.3877,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6119215291750503,
+      "grad_norm": 0.38477277755737305,
+      "learning_rate": 9.674382365411617e-06,
+      "loss": 0.3855,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6121730382293763,
+      "grad_norm": 0.40748539566993713,
+      "learning_rate": 9.673862751545773e-06,
+      "loss": 0.3879,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6124245472837022,
+      "grad_norm": 0.389136403799057,
+      "learning_rate": 9.673342737392824e-06,
+      "loss": 0.4344,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6126760563380281,
+      "grad_norm": 0.3874034583568573,
+      "learning_rate": 9.672822322997305e-06,
+      "loss": 0.4105,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6129275653923542,
+      "grad_norm": 0.3671362102031708,
+      "learning_rate": 9.672301508403788e-06,
+      "loss": 0.383,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6131790744466801,
+      "grad_norm": 0.35450032353401184,
+      "learning_rate": 9.671780293656876e-06,
+      "loss": 0.3807,
+      "step": 2438
+    },
+    {
+      "epoch": 0.613430583501006,
+      "grad_norm": 0.4165121018886566,
+      "learning_rate": 9.67125867880121e-06,
+      "loss": 0.3878,
+      "step": 2439
+    },
+    {
+      "epoch": 0.613682092555332,
+      "grad_norm": 0.40714791417121887,
+      "learning_rate": 9.67073666388146e-06,
+      "loss": 0.4076,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6139336016096579,
+      "grad_norm": 0.42955923080444336,
+      "learning_rate": 9.670214248942335e-06,
+      "loss": 0.4108,
+      "step": 2441
+    },
+    {
+      "epoch": 0.614185110663984,
+      "grad_norm": 0.38127392530441284,
+      "learning_rate": 9.669691434028576e-06,
+      "loss": 0.4032,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6144366197183099,
+      "grad_norm": 0.4305511713027954,
+      "learning_rate": 9.66916821918496e-06,
+      "loss": 0.3918,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6146881287726358,
+      "grad_norm": 0.45002248883247375,
+      "learning_rate": 9.668644604456297e-06,
+      "loss": 0.3847,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6149396378269618,
+      "grad_norm": 0.34926843643188477,
+      "learning_rate": 9.668120589887429e-06,
+      "loss": 0.3919,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6151911468812877,
+      "grad_norm": 0.39543116092681885,
+      "learning_rate": 9.667596175523237e-06,
+      "loss": 0.3568,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6154426559356136,
+      "grad_norm": 0.3891063630580902,
+      "learning_rate": 9.667071361408633e-06,
+      "loss": 0.3759,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6156941649899397,
+      "grad_norm": 0.3981350064277649,
+      "learning_rate": 9.666546147588563e-06,
+      "loss": 0.3784,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6159456740442656,
+      "grad_norm": 0.3683044910430908,
+      "learning_rate": 9.666020534108009e-06,
+      "loss": 0.3846,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6161971830985915,
+      "grad_norm": 0.423318088054657,
+      "learning_rate": 9.665494521011988e-06,
+      "loss": 0.3747,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6164486921529175,
+      "grad_norm": 0.3727003037929535,
+      "learning_rate": 9.664968108345549e-06,
+      "loss": 0.3517,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6167002012072434,
+      "grad_norm": 0.3899252414703369,
+      "learning_rate": 9.664441296153775e-06,
+      "loss": 0.3964,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6169517102615694,
+      "grad_norm": 0.415444940328598,
+      "learning_rate": 9.663914084481784e-06,
+      "loss": 0.3811,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6172032193158954,
+      "grad_norm": 0.4385893940925598,
+      "learning_rate": 9.66338647337473e-06,
+      "loss": 0.4241,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6174547283702213,
+      "grad_norm": 0.3797210454940796,
+      "learning_rate": 9.662858462877797e-06,
+      "loss": 0.3835,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6177062374245473,
+      "grad_norm": 0.4100744426250458,
+      "learning_rate": 9.662330053036208e-06,
+      "loss": 0.408,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6179577464788732,
+      "grad_norm": 0.3938503563404083,
+      "learning_rate": 9.66180124389522e-06,
+      "loss": 0.4077,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6182092555331992,
+      "grad_norm": 0.3984260857105255,
+      "learning_rate": 9.661272035500115e-06,
+      "loss": 0.3775,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6184607645875252,
+      "grad_norm": 0.39625781774520874,
+      "learning_rate": 9.660742427896224e-06,
+      "loss": 0.3969,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6187122736418511,
+      "grad_norm": 0.4334189295768738,
+      "learning_rate": 9.6602124211289e-06,
+      "loss": 0.3946,
+      "step": 2460
+    },
+    {
+      "epoch": 0.618963782696177,
+      "grad_norm": 0.38942191004753113,
+      "learning_rate": 9.65968201524354e-06,
+      "loss": 0.3788,
+      "step": 2461
+    },
+    {
+      "epoch": 0.619215291750503,
+      "grad_norm": 0.36435267329216003,
+      "learning_rate": 9.659151210285562e-06,
+      "loss": 0.3685,
+      "step": 2462
+    },
+    {
+      "epoch": 0.619466800804829,
+      "grad_norm": 0.4146406650543213,
+      "learning_rate": 9.658620006300432e-06,
+      "loss": 0.3682,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 0.41353920102119446,
+      "learning_rate": 9.658088403333642e-06,
+      "loss": 0.3967,
+      "step": 2464
+    },
+    {
+      "epoch": 0.6199698189134809,
+      "grad_norm": 0.41589221358299255,
+      "learning_rate": 9.657556401430723e-06,
+      "loss": 0.4066,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6202213279678068,
+      "grad_norm": 0.47424760460853577,
+      "learning_rate": 9.657024000637235e-06,
+      "loss": 0.3937,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6204728370221329,
+      "grad_norm": 0.3641650080680847,
+      "learning_rate": 9.656491200998774e-06,
+      "loss": 0.3926,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6207243460764588,
+      "grad_norm": 0.4509759247303009,
+      "learning_rate": 9.655958002560974e-06,
+      "loss": 0.4217,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6209758551307847,
+      "grad_norm": 0.442841112613678,
+      "learning_rate": 9.655424405369497e-06,
+      "loss": 0.3757,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6212273641851107,
+      "grad_norm": 0.3849235475063324,
+      "learning_rate": 9.654890409470047e-06,
+      "loss": 0.4041,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6214788732394366,
+      "grad_norm": 0.39690494537353516,
+      "learning_rate": 9.654356014908352e-06,
+      "loss": 0.3915,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6217303822937625,
+      "grad_norm": 0.4252414107322693,
+      "learning_rate": 9.653821221730183e-06,
+      "loss": 0.3808,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6219818913480886,
+      "grad_norm": 0.38544967770576477,
+      "learning_rate": 9.65328602998134e-06,
+      "loss": 0.3856,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6222334004024145,
+      "grad_norm": 0.3949087858200073,
+      "learning_rate": 9.65275043970766e-06,
+      "loss": 0.3938,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6224849094567404,
+      "grad_norm": 0.3579958975315094,
+      "learning_rate": 9.65221445095501e-06,
+      "loss": 0.4041,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6227364185110664,
+      "grad_norm": 0.40078824758529663,
+      "learning_rate": 9.6516780637693e-06,
+      "loss": 0.367,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6229879275653923,
+      "grad_norm": 0.39681002497673035,
+      "learning_rate": 9.651141278196462e-06,
+      "loss": 0.4074,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6232394366197183,
+      "grad_norm": 0.37867873907089233,
+      "learning_rate": 9.650604094282471e-06,
+      "loss": 0.3701,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6234909456740443,
+      "grad_norm": 0.43801194429397583,
+      "learning_rate": 9.650066512073336e-06,
+      "loss": 0.3802,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6237424547283702,
+      "grad_norm": 0.34844502806663513,
+      "learning_rate": 9.649528531615094e-06,
+      "loss": 0.3914,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6239939637826962,
+      "grad_norm": 0.3765834867954254,
+      "learning_rate": 9.64899015295382e-06,
+      "loss": 0.3964,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6242454728370221,
+      "grad_norm": 0.4189784526824951,
+      "learning_rate": 9.648451376135624e-06,
+      "loss": 0.3936,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6244969818913481,
+      "grad_norm": 0.3877945840358734,
+      "learning_rate": 9.647912201206646e-06,
+      "loss": 0.3809,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6247484909456741,
+      "grad_norm": 0.34489884972572327,
+      "learning_rate": 9.647372628213068e-06,
+      "loss": 0.3805,
+      "step": 2484
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.3784612715244293,
+      "learning_rate": 9.646832657201097e-06,
+      "loss": 0.3824,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6252515090543259,
+      "grad_norm": 0.3295539319515228,
+      "learning_rate": 9.646292288216978e-06,
+      "loss": 0.3783,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6255030181086519,
+      "grad_norm": 0.3725329339504242,
+      "learning_rate": 9.645751521306994e-06,
+      "loss": 0.4002,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6257545271629779,
+      "grad_norm": 0.386722207069397,
+      "learning_rate": 9.645210356517454e-06,
+      "loss": 0.3934,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6260060362173038,
+      "grad_norm": 0.3421211838722229,
+      "learning_rate": 9.64466879389471e-06,
+      "loss": 0.3854,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6262575452716298,
+      "grad_norm": 0.42979565262794495,
+      "learning_rate": 9.644126833485139e-06,
+      "loss": 0.3659,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6265090543259557,
+      "grad_norm": 0.36490967869758606,
+      "learning_rate": 9.643584475335157e-06,
+      "loss": 0.384,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6267605633802817,
+      "grad_norm": 0.4025133550167084,
+      "learning_rate": 9.643041719491218e-06,
+      "loss": 0.3844,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6270120724346077,
+      "grad_norm": 0.36443737149238586,
+      "learning_rate": 9.6424985659998e-06,
+      "loss": 0.3851,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6272635814889336,
+      "grad_norm": 0.33622828125953674,
+      "learning_rate": 9.641955014907425e-06,
+      "loss": 0.3756,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6275150905432596,
+      "grad_norm": 0.3600637912750244,
+      "learning_rate": 9.64141106626064e-06,
+      "loss": 0.3833,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6277665995975855,
+      "grad_norm": 0.37680599093437195,
+      "learning_rate": 9.640866720106037e-06,
+      "loss": 0.3976,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6280181086519114,
+      "grad_norm": 0.380525141954422,
+      "learning_rate": 9.64032197649023e-06,
+      "loss": 0.3714,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6282696177062375,
+      "grad_norm": 0.3835989534854889,
+      "learning_rate": 9.639776835459878e-06,
+      "loss": 0.4056,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6285211267605634,
+      "grad_norm": 0.46692678332328796,
+      "learning_rate": 9.639231297061663e-06,
+      "loss": 0.4207,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6287726358148893,
+      "grad_norm": 0.3480624258518219,
+      "learning_rate": 9.638685361342314e-06,
+      "loss": 0.4059,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6290241448692153,
+      "grad_norm": 0.4090098440647125,
+      "learning_rate": 9.63813902834858e-06,
+      "loss": 0.3965,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6292756539235412,
+      "grad_norm": 0.37328529357910156,
+      "learning_rate": 9.637592298127258e-06,
+      "loss": 0.3768,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6295271629778671,
+      "grad_norm": 0.43410345911979675,
+      "learning_rate": 9.637045170725165e-06,
+      "loss": 0.3961,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6297786720321932,
+      "grad_norm": 0.3536240756511688,
+      "learning_rate": 9.636497646189165e-06,
+      "loss": 0.3811,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6300301810865191,
+      "grad_norm": 0.35403186082839966,
+      "learning_rate": 9.635949724566147e-06,
+      "loss": 0.3764,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6302816901408451,
+      "grad_norm": 0.3755878508090973,
+      "learning_rate": 9.635401405903037e-06,
+      "loss": 0.378,
+      "step": 2506
+    },
+    {
+      "epoch": 0.630533199195171,
+      "grad_norm": 0.36756351590156555,
+      "learning_rate": 9.634852690246795e-06,
+      "loss": 0.387,
+      "step": 2507
+    },
+    {
+      "epoch": 0.630784708249497,
+      "grad_norm": 0.3477321267127991,
+      "learning_rate": 9.634303577644415e-06,
+      "loss": 0.3545,
+      "step": 2508
+    },
+    {
+      "epoch": 0.631036217303823,
+      "grad_norm": 0.33334577083587646,
+      "learning_rate": 9.633754068142928e-06,
+      "loss": 0.3817,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6312877263581489,
+      "grad_norm": 0.34079161286354065,
+      "learning_rate": 9.633204161789392e-06,
+      "loss": 0.367,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6315392354124748,
+      "grad_norm": 0.4175172746181488,
+      "learning_rate": 9.632653858630905e-06,
+      "loss": 0.3825,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6317907444668008,
+      "grad_norm": 0.3560580015182495,
+      "learning_rate": 9.632103158714596e-06,
+      "loss": 0.3525,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6320422535211268,
+      "grad_norm": 0.3939068913459778,
+      "learning_rate": 9.631552062087632e-06,
+      "loss": 0.4022,
+      "step": 2513
+    },
+    {
+      "epoch": 0.6322937625754527,
+      "grad_norm": 0.3403138518333435,
+      "learning_rate": 9.631000568797208e-06,
+      "loss": 0.4065,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6325452716297787,
+      "grad_norm": 0.3796042799949646,
+      "learning_rate": 9.630448678890556e-06,
+      "loss": 0.4052,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6327967806841046,
+      "grad_norm": 0.3852931559085846,
+      "learning_rate": 9.629896392414943e-06,
+      "loss": 0.3769,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6330482897384306,
+      "grad_norm": 0.3552147448062897,
+      "learning_rate": 9.62934370941767e-06,
+      "loss": 0.3913,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6332997987927566,
+      "grad_norm": 0.37440934777259827,
+      "learning_rate": 9.628790629946066e-06,
+      "loss": 0.387,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6335513078470825,
+      "grad_norm": 0.4127063751220703,
+      "learning_rate": 9.628237154047504e-06,
+      "loss": 0.3924,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6338028169014085,
+      "grad_norm": 0.3693302273750305,
+      "learning_rate": 9.627683281769384e-06,
+      "loss": 0.3824,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6340543259557344,
+      "grad_norm": 0.3535366654396057,
+      "learning_rate": 9.627129013159142e-06,
+      "loss": 0.3505,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6343058350100603,
+      "grad_norm": 0.43735992908477783,
+      "learning_rate": 9.626574348264246e-06,
+      "loss": 0.3739,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6345573440643864,
+      "grad_norm": 0.41980883479118347,
+      "learning_rate": 9.626019287132202e-06,
+      "loss": 0.3887,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6348088531187123,
+      "grad_norm": 0.3785690665245056,
+      "learning_rate": 9.625463829810547e-06,
+      "loss": 0.4176,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6350603621730382,
+      "grad_norm": 0.42933017015457153,
+      "learning_rate": 9.62490797634685e-06,
+      "loss": 0.3994,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6353118712273642,
+      "grad_norm": 0.3780410885810852,
+      "learning_rate": 9.624351726788719e-06,
+      "loss": 0.379,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6355633802816901,
+      "grad_norm": 0.36872223019599915,
+      "learning_rate": 9.623795081183794e-06,
+      "loss": 0.4074,
+      "step": 2527
+    },
+    {
+      "epoch": 0.635814889336016,
+      "grad_norm": 0.3785404562950134,
+      "learning_rate": 9.623238039579742e-06,
+      "loss": 0.3602,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6360663983903421,
+      "grad_norm": 0.3856281638145447,
+      "learning_rate": 9.622680602024278e-06,
+      "loss": 0.3694,
+      "step": 2529
+    },
+    {
+      "epoch": 0.636317907444668,
+      "grad_norm": 0.3590550124645233,
+      "learning_rate": 9.62212276856514e-06,
+      "loss": 0.3768,
+      "step": 2530
+    },
+    {
+      "epoch": 0.636569416498994,
+      "grad_norm": 0.4134266972541809,
+      "learning_rate": 9.621564539250103e-06,
+      "loss": 0.3803,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6368209255533199,
+      "grad_norm": 0.3633298873901367,
+      "learning_rate": 9.621005914126974e-06,
+      "loss": 0.3774,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6370724346076458,
+      "grad_norm": 0.4221361577510834,
+      "learning_rate": 9.6204468932436e-06,
+      "loss": 0.3943,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6373239436619719,
+      "grad_norm": 0.3500104546546936,
+      "learning_rate": 9.619887476647854e-06,
+      "loss": 0.3893,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6375754527162978,
+      "grad_norm": 0.365842342376709,
+      "learning_rate": 9.619327664387648e-06,
+      "loss": 0.3697,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6378269617706237,
+      "grad_norm": 0.36636513471603394,
+      "learning_rate": 9.618767456510924e-06,
+      "loss": 0.3848,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6380784708249497,
+      "grad_norm": 0.37152019143104553,
+      "learning_rate": 9.618206853065664e-06,
+      "loss": 0.377,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6383299798792756,
+      "grad_norm": 0.37324413657188416,
+      "learning_rate": 9.617645854099878e-06,
+      "loss": 0.3837,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6385814889336016,
+      "grad_norm": 0.3649263083934784,
+      "learning_rate": 9.617084459661612e-06,
+      "loss": 0.3921,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6388329979879276,
+      "grad_norm": 0.4074450135231018,
+      "learning_rate": 9.616522669798947e-06,
+      "loss": 0.4087,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6390845070422535,
+      "grad_norm": 0.3745105564594269,
+      "learning_rate": 9.615960484559995e-06,
+      "loss": 0.3926,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6393360160965795,
+      "grad_norm": 0.42575785517692566,
+      "learning_rate": 9.615397903992906e-06,
+      "loss": 0.3892,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6395875251509054,
+      "grad_norm": 0.3537648916244507,
+      "learning_rate": 9.61483492814586e-06,
+      "loss": 0.3679,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6398390342052314,
+      "grad_norm": 0.37703290581703186,
+      "learning_rate": 9.614271557067072e-06,
+      "loss": 0.3567,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6400905432595574,
+      "grad_norm": 0.40531283617019653,
+      "learning_rate": 9.613707790804794e-06,
+      "loss": 0.3992,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6403420523138833,
+      "grad_norm": 0.37500283122062683,
+      "learning_rate": 9.613143629407305e-06,
+      "loss": 0.3933,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6405935613682092,
+      "grad_norm": 0.41981109976768494,
+      "learning_rate": 9.612579072922926e-06,
+      "loss": 0.3811,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6408450704225352,
+      "grad_norm": 0.416825532913208,
+      "learning_rate": 9.612014121400003e-06,
+      "loss": 0.3978,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6410965794768612,
+      "grad_norm": 0.40258339047431946,
+      "learning_rate": 9.611448774886925e-06,
+      "loss": 0.3916,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6413480885311871,
+      "grad_norm": 0.44783157110214233,
+      "learning_rate": 9.610883033432107e-06,
+      "loss": 0.3962,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6415995975855131,
+      "grad_norm": 0.3728145360946655,
+      "learning_rate": 9.610316897084004e-06,
+      "loss": 0.3778,
+      "step": 2551
+    },
+    {
+      "epoch": 0.641851106639839,
+      "grad_norm": 0.3508928418159485,
+      "learning_rate": 9.6097503658911e-06,
+      "loss": 0.3749,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6421026156941649,
+      "grad_norm": 0.38925930857658386,
+      "learning_rate": 9.609183439901917e-06,
+      "loss": 0.3726,
+      "step": 2553
+    },
+    {
+      "epoch": 0.642354124748491,
+      "grad_norm": 0.3719974458217621,
+      "learning_rate": 9.608616119165007e-06,
+      "loss": 0.3923,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6426056338028169,
+      "grad_norm": 0.3723568618297577,
+      "learning_rate": 9.608048403728957e-06,
+      "loss": 0.3739,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.3961268365383148,
+      "learning_rate": 9.60748029364239e-06,
+      "loss": 0.3812,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6431086519114688,
+      "grad_norm": 0.38643985986709595,
+      "learning_rate": 9.60691178895396e-06,
+      "loss": 0.4085,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6433601609657947,
+      "grad_norm": 0.3590714931488037,
+      "learning_rate": 9.606342889712354e-06,
+      "loss": 0.3862,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6436116700201208,
+      "grad_norm": 0.35285237431526184,
+      "learning_rate": 9.605773595966298e-06,
+      "loss": 0.4006,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6438631790744467,
+      "grad_norm": 0.36502790451049805,
+      "learning_rate": 9.605203907764548e-06,
+      "loss": 0.3655,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6441146881287726,
+      "grad_norm": 0.3753395676612854,
+      "learning_rate": 9.604633825155894e-06,
+      "loss": 0.4009,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6443661971830986,
+      "grad_norm": 0.39320510625839233,
+      "learning_rate": 9.604063348189158e-06,
+      "loss": 0.3895,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6446177062374245,
+      "grad_norm": 0.320444256067276,
+      "learning_rate": 9.603492476913199e-06,
+      "loss": 0.3756,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6448692152917505,
+      "grad_norm": 0.349065899848938,
+      "learning_rate": 9.60292121137691e-06,
+      "loss": 0.4084,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6451207243460765,
+      "grad_norm": 0.34628018736839294,
+      "learning_rate": 9.602349551629213e-06,
+      "loss": 0.4003,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6453722334004024,
+      "grad_norm": 0.35905465483665466,
+      "learning_rate": 9.601777497719071e-06,
+      "loss": 0.3999,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6456237424547284,
+      "grad_norm": 0.3582354187965393,
+      "learning_rate": 9.601205049695473e-06,
+      "loss": 0.3676,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6458752515090543,
+      "grad_norm": 0.35198500752449036,
+      "learning_rate": 9.60063220760745e-06,
+      "loss": 0.4059,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6461267605633803,
+      "grad_norm": 0.3506782352924347,
+      "learning_rate": 9.600058971504058e-06,
+      "loss": 0.3826,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6463782696177063,
+      "grad_norm": 0.3575206398963928,
+      "learning_rate": 9.599485341434394e-06,
+      "loss": 0.3791,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6466297786720322,
+      "grad_norm": 0.3570882976055145,
+      "learning_rate": 9.598911317447583e-06,
+      "loss": 0.3798,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6468812877263581,
+      "grad_norm": 0.36667707562446594,
+      "learning_rate": 9.598336899592791e-06,
+      "loss": 0.3967,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6471327967806841,
+      "grad_norm": 0.31375589966773987,
+      "learning_rate": 9.597762087919209e-06,
+      "loss": 0.3925,
+      "step": 2573
+    },
+    {
+      "epoch": 0.64738430583501,
+      "grad_norm": 0.3781939446926117,
+      "learning_rate": 9.597186882476069e-06,
+      "loss": 0.3934,
+      "step": 2574
+    },
+    {
+      "epoch": 0.647635814889336,
+      "grad_norm": 0.36152127385139465,
+      "learning_rate": 9.59661128331263e-06,
+      "loss": 0.4066,
+      "step": 2575
+    },
+    {
+      "epoch": 0.647887323943662,
+      "grad_norm": 0.3301715552806854,
+      "learning_rate": 9.596035290478192e-06,
+      "loss": 0.3804,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6481388329979879,
+      "grad_norm": 0.4257327914237976,
+      "learning_rate": 9.595458904022086e-06,
+      "loss": 0.3964,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6483903420523138,
+      "grad_norm": 0.3644980490207672,
+      "learning_rate": 9.594882123993671e-06,
+      "loss": 0.3798,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6486418511066399,
+      "grad_norm": 0.3696765601634979,
+      "learning_rate": 9.59430495044235e-06,
+      "loss": 0.3466,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6488933601609658,
+      "grad_norm": 0.450061172246933,
+      "learning_rate": 9.593727383417551e-06,
+      "loss": 0.3997,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6491448692152918,
+      "grad_norm": 0.4249400794506073,
+      "learning_rate": 9.59314942296874e-06,
+      "loss": 0.3988,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6493963782696177,
+      "grad_norm": 0.44383350014686584,
+      "learning_rate": 9.592571069145415e-06,
+      "loss": 0.3948,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6496478873239436,
+      "grad_norm": 0.3797934353351593,
+      "learning_rate": 9.591992321997107e-06,
+      "loss": 0.3807,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6498993963782697,
+      "grad_norm": 0.40666332840919495,
+      "learning_rate": 9.591413181573388e-06,
+      "loss": 0.4152,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6501509054325956,
+      "grad_norm": 0.36471301317214966,
+      "learning_rate": 9.590833647923852e-06,
+      "loss": 0.3952,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6504024144869215,
+      "grad_norm": 0.40786394476890564,
+      "learning_rate": 9.590253721098135e-06,
+      "loss": 0.3957,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6506539235412475,
+      "grad_norm": 0.3736938238143921,
+      "learning_rate": 9.589673401145902e-06,
+      "loss": 0.3554,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6509054325955734,
+      "grad_norm": 0.3855222761631012,
+      "learning_rate": 9.589092688116855e-06,
+      "loss": 0.4014,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6511569416498993,
+      "grad_norm": 0.41750627756118774,
+      "learning_rate": 9.58851158206073e-06,
+      "loss": 0.3998,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6514084507042254,
+      "grad_norm": 0.4044592082500458,
+      "learning_rate": 9.587930083027293e-06,
+      "loss": 0.3571,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6516599597585513,
+      "grad_norm": 0.3685952126979828,
+      "learning_rate": 9.587348191066345e-06,
+      "loss": 0.3984,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6519114688128773,
+      "grad_norm": 0.3659170866012573,
+      "learning_rate": 9.586765906227727e-06,
+      "loss": 0.3794,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6521629778672032,
+      "grad_norm": 0.3769875466823578,
+      "learning_rate": 9.586183228561299e-06,
+      "loss": 0.3803,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6524144869215291,
+      "grad_norm": 0.40973907709121704,
+      "learning_rate": 9.58560015811697e-06,
+      "loss": 0.3876,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6526659959758552,
+      "grad_norm": 0.377444326877594,
+      "learning_rate": 9.585016694944676e-06,
+      "loss": 0.402,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6529175050301811,
+      "grad_norm": 0.3654913902282715,
+      "learning_rate": 9.584432839094387e-06,
+      "loss": 0.3613,
+      "step": 2596
+    },
+    {
+      "epoch": 0.653169014084507,
+      "grad_norm": 0.3855876326560974,
+      "learning_rate": 9.583848590616102e-06,
+      "loss": 0.3833,
+      "step": 2597
+    },
+    {
+      "epoch": 0.653420523138833,
+      "grad_norm": 0.40650826692581177,
+      "learning_rate": 9.583263949559864e-06,
+      "loss": 0.3993,
+      "step": 2598
+    },
+    {
+      "epoch": 0.653672032193159,
+      "grad_norm": 0.4067690670490265,
+      "learning_rate": 9.582678915975741e-06,
+      "loss": 0.3818,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6539235412474849,
+      "grad_norm": 0.38773733377456665,
+      "learning_rate": 9.582093489913838e-06,
+      "loss": 0.39,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6541750503018109,
+      "grad_norm": 0.3709178864955902,
+      "learning_rate": 9.581507671424293e-06,
+      "loss": 0.3899,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6544265593561368,
+      "grad_norm": 0.4474826157093048,
+      "learning_rate": 9.580921460557278e-06,
+      "loss": 0.406,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6546780684104627,
+      "grad_norm": 0.3681783974170685,
+      "learning_rate": 9.580334857362997e-06,
+      "loss": 0.393,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6549295774647887,
+      "grad_norm": 0.4004930555820465,
+      "learning_rate": 9.57974786189169e-06,
+      "loss": 0.4031,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6551810865191147,
+      "grad_norm": 0.48355597257614136,
+      "learning_rate": 9.579160474193632e-06,
+      "loss": 0.3869,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6554325955734407,
+      "grad_norm": 0.3602335751056671,
+      "learning_rate": 9.578572694319124e-06,
+      "loss": 0.4049,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6556841046277666,
+      "grad_norm": 0.40569397807121277,
+      "learning_rate": 9.577984522318508e-06,
+      "loss": 0.3723,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6559356136820925,
+      "grad_norm": 0.41683316230773926,
+      "learning_rate": 9.57739595824216e-06,
+      "loss": 0.3858,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6561871227364185,
+      "grad_norm": 0.3983708322048187,
+      "learning_rate": 9.576807002140483e-06,
+      "loss": 0.4024,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6564386317907445,
+      "grad_norm": 0.3469032347202301,
+      "learning_rate": 9.576217654063917e-06,
+      "loss": 0.3775,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6566901408450704,
+      "grad_norm": 0.4117089509963989,
+      "learning_rate": 9.57562791406294e-06,
+      "loss": 0.4051,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6569416498993964,
+      "grad_norm": 0.37733131647109985,
+      "learning_rate": 9.575037782188054e-06,
+      "loss": 0.3771,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6571931589537223,
+      "grad_norm": 0.3928118944168091,
+      "learning_rate": 9.574447258489808e-06,
+      "loss": 0.3604,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6574446680080482,
+      "grad_norm": 0.3928970396518707,
+      "learning_rate": 9.573856343018768e-06,
+      "loss": 0.3788,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6576961770623743,
+      "grad_norm": 0.402646541595459,
+      "learning_rate": 9.573265035825548e-06,
+      "loss": 0.3922,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6579476861167002,
+      "grad_norm": 0.3756962716579437,
+      "learning_rate": 9.572673336960787e-06,
+      "loss": 0.3839,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6581991951710262,
+      "grad_norm": 0.3633159399032593,
+      "learning_rate": 9.572081246475162e-06,
+      "loss": 0.3558,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6584507042253521,
+      "grad_norm": 0.39355260133743286,
+      "learning_rate": 9.571488764419381e-06,
+      "loss": 0.3855,
+      "step": 2618
+    },
+    {
+      "epoch": 0.658702213279678,
+      "grad_norm": 0.3599531054496765,
+      "learning_rate": 9.570895890844188e-06,
+      "loss": 0.3917,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6589537223340041,
+      "grad_norm": 0.3619001507759094,
+      "learning_rate": 9.570302625800353e-06,
+      "loss": 0.3906,
+      "step": 2620
+    },
+    {
+      "epoch": 0.65920523138833,
+      "grad_norm": 0.35282886028289795,
+      "learning_rate": 9.569708969338694e-06,
+      "loss": 0.3902,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6594567404426559,
+      "grad_norm": 0.36944806575775146,
+      "learning_rate": 9.569114921510048e-06,
+      "loss": 0.3758,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6597082494969819,
+      "grad_norm": 0.3732403516769409,
+      "learning_rate": 9.568520482365293e-06,
+      "loss": 0.3795,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6599597585513078,
+      "grad_norm": 0.35911738872528076,
+      "learning_rate": 9.56792565195534e-06,
+      "loss": 0.4018,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6602112676056338,
+      "grad_norm": 0.385262131690979,
+      "learning_rate": 9.567330430331133e-06,
+      "loss": 0.3562,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6604627766599598,
+      "grad_norm": 0.3803010582923889,
+      "learning_rate": 9.566734817543645e-06,
+      "loss": 0.3979,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6607142857142857,
+      "grad_norm": 0.4101577401161194,
+      "learning_rate": 9.566138813643891e-06,
+      "loss": 0.4253,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6609657947686117,
+      "grad_norm": 0.39751216769218445,
+      "learning_rate": 9.565542418682914e-06,
+      "loss": 0.3937,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6612173038229376,
+      "grad_norm": 0.3846282958984375,
+      "learning_rate": 9.564945632711789e-06,
+      "loss": 0.403,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6614688128772636,
+      "grad_norm": 0.3686662018299103,
+      "learning_rate": 9.56434845578163e-06,
+      "loss": 0.37,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6617203219315896,
+      "grad_norm": 0.43031734228134155,
+      "learning_rate": 9.56375088794358e-06,
+      "loss": 0.381,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6619718309859155,
+      "grad_norm": 0.36381101608276367,
+      "learning_rate": 9.563152929248817e-06,
+      "loss": 0.4052,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6622233400402414,
+      "grad_norm": 0.3557662069797516,
+      "learning_rate": 9.562554579748553e-06,
+      "loss": 0.4099,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6624748490945674,
+      "grad_norm": 0.3922789692878723,
+      "learning_rate": 9.561955839494032e-06,
+      "loss": 0.3999,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6627263581488934,
+      "grad_norm": 0.39800694584846497,
+      "learning_rate": 9.561356708536532e-06,
+      "loss": 0.4047,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6629778672032193,
+      "grad_norm": 0.35587430000305176,
+      "learning_rate": 9.560757186927367e-06,
+      "loss": 0.3867,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6632293762575453,
+      "grad_norm": 0.4054100811481476,
+      "learning_rate": 9.56015727471788e-06,
+      "loss": 0.3606,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6634808853118712,
+      "grad_norm": 0.3700953423976898,
+      "learning_rate": 9.559556971959452e-06,
+      "loss": 0.3743,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6637323943661971,
+      "grad_norm": 0.3898264169692993,
+      "learning_rate": 9.558956278703493e-06,
+      "loss": 0.3565,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6639839034205232,
+      "grad_norm": 0.4122966527938843,
+      "learning_rate": 9.55835519500145e-06,
+      "loss": 0.4195,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6642354124748491,
+      "grad_norm": 0.375590443611145,
+      "learning_rate": 9.557753720904801e-06,
+      "loss": 0.409,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6644869215291751,
+      "grad_norm": 0.3992207944393158,
+      "learning_rate": 9.557151856465059e-06,
+      "loss": 0.3778,
+      "step": 2642
+    },
+    {
+      "epoch": 0.664738430583501,
+      "grad_norm": 0.36784839630126953,
+      "learning_rate": 9.556549601733769e-06,
+      "loss": 0.3846,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6649899396378269,
+      "grad_norm": 0.3585883677005768,
+      "learning_rate": 9.555946956762513e-06,
+      "loss": 0.3747,
+      "step": 2644
+    },
+    {
+      "epoch": 0.665241448692153,
+      "grad_norm": 0.4158109128475189,
+      "learning_rate": 9.555343921602901e-06,
+      "loss": 0.3686,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6654929577464789,
+      "grad_norm": 0.41617828607559204,
+      "learning_rate": 9.55474049630658e-06,
+      "loss": 0.3826,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6657444668008048,
+      "grad_norm": 0.3927265703678131,
+      "learning_rate": 9.554136680925232e-06,
+      "loss": 0.4062,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6659959758551308,
+      "grad_norm": 0.3728967607021332,
+      "learning_rate": 9.553532475510565e-06,
+      "loss": 0.3673,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6662474849094567,
+      "grad_norm": 0.3700278103351593,
+      "learning_rate": 9.55292788011433e-06,
+      "loss": 0.3934,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6664989939637826,
+      "grad_norm": 0.4425957500934601,
+      "learning_rate": 9.552322894788306e-06,
+      "loss": 0.3672,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6667505030181087,
+      "grad_norm": 0.36857491731643677,
+      "learning_rate": 9.551717519584303e-06,
+      "loss": 0.3733,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6670020120724346,
+      "grad_norm": 0.4787845313549042,
+      "learning_rate": 9.551111754554172e-06,
+      "loss": 0.3793,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6672535211267606,
+      "grad_norm": 0.3845168948173523,
+      "learning_rate": 9.55050559974979e-06,
+      "loss": 0.3796,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6675050301810865,
+      "grad_norm": 0.38348764181137085,
+      "learning_rate": 9.54989905522307e-06,
+      "loss": 0.3925,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6677565392354124,
+      "grad_norm": 0.41745656728744507,
+      "learning_rate": 9.549292121025961e-06,
+      "loss": 0.3836,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6680080482897385,
+      "grad_norm": 0.37018293142318726,
+      "learning_rate": 9.548684797210444e-06,
+      "loss": 0.3806,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6682595573440644,
+      "grad_norm": 0.3973758816719055,
+      "learning_rate": 9.548077083828528e-06,
+      "loss": 0.3908,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6685110663983903,
+      "grad_norm": 0.3895440101623535,
+      "learning_rate": 9.547468980932263e-06,
+      "loss": 0.3907,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6687625754527163,
+      "grad_norm": 0.32289016246795654,
+      "learning_rate": 9.546860488573729e-06,
+      "loss": 0.3738,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6690140845070423,
+      "grad_norm": 0.4137352705001831,
+      "learning_rate": 9.54625160680504e-06,
+      "loss": 0.3909,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6692655935613682,
+      "grad_norm": 0.34959176182746887,
+      "learning_rate": 9.545642335678341e-06,
+      "loss": 0.3639,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6695171026156942,
+      "grad_norm": 0.36586570739746094,
+      "learning_rate": 9.545032675245814e-06,
+      "loss": 0.3729,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6697686116700201,
+      "grad_norm": 0.39214131236076355,
+      "learning_rate": 9.544422625559671e-06,
+      "loss": 0.3859,
+      "step": 2663
+    },
+    {
+      "epoch": 0.670020120724346,
+      "grad_norm": 0.35369038581848145,
+      "learning_rate": 9.543812186672161e-06,
+      "loss": 0.3796,
+      "step": 2664
+    },
+    {
+      "epoch": 0.670271629778672,
+      "grad_norm": 0.40630146861076355,
+      "learning_rate": 9.543201358635564e-06,
+      "loss": 0.3971,
+      "step": 2665
+    },
+    {
+      "epoch": 0.670523138832998,
+      "grad_norm": 0.34926897287368774,
+      "learning_rate": 9.54259014150219e-06,
+      "loss": 0.3889,
+      "step": 2666
+    },
+    {
+      "epoch": 0.670774647887324,
+      "grad_norm": 0.4089818298816681,
+      "learning_rate": 9.54197853532439e-06,
+      "loss": 0.392,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6710261569416499,
+      "grad_norm": 0.3891892433166504,
+      "learning_rate": 9.541366540154544e-06,
+      "loss": 0.4018,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6712776659959758,
+      "grad_norm": 0.3339185118675232,
+      "learning_rate": 9.540754156045064e-06,
+      "loss": 0.3712,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6715291750503019,
+      "grad_norm": 0.40440425276756287,
+      "learning_rate": 9.540141383048398e-06,
+      "loss": 0.4362,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6717806841046278,
+      "grad_norm": 0.39212754368782043,
+      "learning_rate": 9.539528221217026e-06,
+      "loss": 0.4054,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6720321931589537,
+      "grad_norm": 0.3525048494338989,
+      "learning_rate": 9.538914670603458e-06,
+      "loss": 0.3754,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6722837022132797,
+      "grad_norm": 0.37717682123184204,
+      "learning_rate": 9.538300731260247e-06,
+      "loss": 0.4042,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6725352112676056,
+      "grad_norm": 0.4224480986595154,
+      "learning_rate": 9.537686403239967e-06,
+      "loss": 0.359,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6727867203219315,
+      "grad_norm": 0.4100128412246704,
+      "learning_rate": 9.537071686595237e-06,
+      "loss": 0.3911,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6730382293762576,
+      "grad_norm": 0.34391623735427856,
+      "learning_rate": 9.536456581378699e-06,
+      "loss": 0.3899,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6732897384305835,
+      "grad_norm": 0.40277257561683655,
+      "learning_rate": 9.535841087643036e-06,
+      "loss": 0.3973,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6735412474849095,
+      "grad_norm": 0.3905733823776245,
+      "learning_rate": 9.535225205440958e-06,
+      "loss": 0.3931,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6737927565392354,
+      "grad_norm": 0.3948303461074829,
+      "learning_rate": 9.534608934825217e-06,
+      "loss": 0.3773,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6740442655935613,
+      "grad_norm": 0.4447704255580902,
+      "learning_rate": 9.533992275848587e-06,
+      "loss": 0.4014,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6742957746478874,
+      "grad_norm": 0.3908079266548157,
+      "learning_rate": 9.533375228563883e-06,
+      "loss": 0.394,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6745472837022133,
+      "grad_norm": 0.354306697845459,
+      "learning_rate": 9.532757793023952e-06,
+      "loss": 0.3703,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6747987927565392,
+      "grad_norm": 0.3971775472164154,
+      "learning_rate": 9.532139969281673e-06,
+      "loss": 0.3872,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6750503018108652,
+      "grad_norm": 0.3223625421524048,
+      "learning_rate": 9.531521757389957e-06,
+      "loss": 0.3702,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6753018108651911,
+      "grad_norm": 0.4071231186389923,
+      "learning_rate": 9.530903157401755e-06,
+      "loss": 0.4038,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6755533199195171,
+      "grad_norm": 0.3665800094604492,
+      "learning_rate": 9.530284169370039e-06,
+      "loss": 0.3836,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6758048289738431,
+      "grad_norm": 0.3697899878025055,
+      "learning_rate": 9.529664793347827e-06,
+      "loss": 0.383,
+      "step": 2687
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 0.4246238172054291,
+      "learning_rate": 9.529045029388162e-06,
+      "loss": 0.4008,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6763078470824949,
+      "grad_norm": 0.3734142482280731,
+      "learning_rate": 9.528424877544125e-06,
+      "loss": 0.3812,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6765593561368209,
+      "grad_norm": 0.35906723141670227,
+      "learning_rate": 9.527804337868827e-06,
+      "loss": 0.3821,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6768108651911469,
+      "grad_norm": 0.42187240719795227,
+      "learning_rate": 9.527183410415413e-06,
+      "loss": 0.3884,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6770623742454729,
+      "grad_norm": 0.3730980455875397,
+      "learning_rate": 9.526562095237061e-06,
+      "loss": 0.3747,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6773138832997988,
+      "grad_norm": 0.44730526208877563,
+      "learning_rate": 9.525940392386985e-06,
+      "loss": 0.3753,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6775653923541247,
+      "grad_norm": 0.38479888439178467,
+      "learning_rate": 9.525318301918427e-06,
+      "loss": 0.402,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6778169014084507,
+      "grad_norm": 0.40130650997161865,
+      "learning_rate": 9.524695823884669e-06,
+      "loss": 0.3775,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6780684104627767,
+      "grad_norm": 0.41748368740081787,
+      "learning_rate": 9.524072958339019e-06,
+      "loss": 0.4146,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6783199195171026,
+      "grad_norm": 0.3546696901321411,
+      "learning_rate": 9.523449705334821e-06,
+      "loss": 0.3782,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6785714285714286,
+      "grad_norm": 0.4278590679168701,
+      "learning_rate": 9.522826064925457e-06,
+      "loss": 0.4017,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6788229376257545,
+      "grad_norm": 0.3383933901786804,
+      "learning_rate": 9.522202037164333e-06,
+      "loss": 0.3774,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6790744466800804,
+      "grad_norm": 0.3658550977706909,
+      "learning_rate": 9.521577622104897e-06,
+      "loss": 0.3783,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6793259557344065,
+      "grad_norm": 0.40239429473876953,
+      "learning_rate": 9.520952819800624e-06,
+      "loss": 0.3896,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6795774647887324,
+      "grad_norm": 0.3836422562599182,
+      "learning_rate": 9.520327630305026e-06,
+      "loss": 0.3932,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6798289738430584,
+      "grad_norm": 0.3793487548828125,
+      "learning_rate": 9.519702053671643e-06,
+      "loss": 0.3851,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6800804828973843,
+      "grad_norm": 0.38546112179756165,
+      "learning_rate": 9.519076089954056e-06,
+      "loss": 0.3839,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6803319919517102,
+      "grad_norm": 0.38321325182914734,
+      "learning_rate": 9.518449739205873e-06,
+      "loss": 0.3793,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6805835010060363,
+      "grad_norm": 0.35967323184013367,
+      "learning_rate": 9.517823001480737e-06,
+      "loss": 0.3805,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6808350100603622,
+      "grad_norm": 0.3868929445743561,
+      "learning_rate": 9.517195876832324e-06,
+      "loss": 0.3666,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6810865191146881,
+      "grad_norm": 0.33470460772514343,
+      "learning_rate": 9.516568365314345e-06,
+      "loss": 0.3674,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6813380281690141,
+      "grad_norm": 0.37914395332336426,
+      "learning_rate": 9.51594046698054e-06,
+      "loss": 0.3757,
+      "step": 2709
+    },
+    {
+      "epoch": 0.68158953722334,
+      "grad_norm": 0.3739543557167053,
+      "learning_rate": 9.515312181884685e-06,
+      "loss": 0.3861,
+      "step": 2710
+    },
+    {
+      "epoch": 0.681841046277666,
+      "grad_norm": 0.3794519007205963,
+      "learning_rate": 9.514683510080592e-06,
+      "loss": 0.3961,
+      "step": 2711
+    },
+    {
+      "epoch": 0.682092555331992,
+      "grad_norm": 0.39454448223114014,
+      "learning_rate": 9.514054451622098e-06,
+      "loss": 0.4207,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6823440643863179,
+      "grad_norm": 0.38329020142555237,
+      "learning_rate": 9.51342500656308e-06,
+      "loss": 0.3901,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6825955734406438,
+      "grad_norm": 0.40806901454925537,
+      "learning_rate": 9.512795174957445e-06,
+      "loss": 0.3576,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6828470824949698,
+      "grad_norm": 0.4223353862762451,
+      "learning_rate": 9.512164956859138e-06,
+      "loss": 0.366,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6830985915492958,
+      "grad_norm": 0.4174785614013672,
+      "learning_rate": 9.511534352322128e-06,
+      "loss": 0.3953,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6833501006036218,
+      "grad_norm": 0.390242338180542,
+      "learning_rate": 9.510903361400426e-06,
+      "loss": 0.3757,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6836016096579477,
+      "grad_norm": 0.45529499650001526,
+      "learning_rate": 9.510271984148071e-06,
+      "loss": 0.3877,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6838531187122736,
+      "grad_norm": 0.38153916597366333,
+      "learning_rate": 9.509640220619136e-06,
+      "loss": 0.4038,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6841046277665996,
+      "grad_norm": 0.4252956211566925,
+      "learning_rate": 9.50900807086773e-06,
+      "loss": 0.3811,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6843561368209256,
+      "grad_norm": 0.3555954098701477,
+      "learning_rate": 9.50837553494799e-06,
+      "loss": 0.3916,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6846076458752515,
+      "grad_norm": 0.3798430562019348,
+      "learning_rate": 9.50774261291409e-06,
+      "loss": 0.3667,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6848591549295775,
+      "grad_norm": 0.3840332627296448,
+      "learning_rate": 9.507109304820234e-06,
+      "loss": 0.3989,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6851106639839034,
+      "grad_norm": 0.38104766607284546,
+      "learning_rate": 9.506475610720665e-06,
+      "loss": 0.3793,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6853621730382293,
+      "grad_norm": 0.4474826455116272,
+      "learning_rate": 9.505841530669652e-06,
+      "loss": 0.3764,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6856136820925554,
+      "grad_norm": 0.37356293201446533,
+      "learning_rate": 9.505207064721499e-06,
+      "loss": 0.4024,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6858651911468813,
+      "grad_norm": 0.4107537269592285,
+      "learning_rate": 9.504572212930544e-06,
+      "loss": 0.3951,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6861167002012073,
+      "grad_norm": 0.40174853801727295,
+      "learning_rate": 9.503936975351164e-06,
+      "loss": 0.3964,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6863682092555332,
+      "grad_norm": 0.3748626708984375,
+      "learning_rate": 9.503301352037756e-06,
+      "loss": 0.3891,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6866197183098591,
+      "grad_norm": 0.357077032327652,
+      "learning_rate": 9.50266534304476e-06,
+      "loss": 0.3828,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6868712273641852,
+      "grad_norm": 0.40127694606781006,
+      "learning_rate": 9.502028948426645e-06,
+      "loss": 0.3847,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6871227364185111,
+      "grad_norm": 0.3611276149749756,
+      "learning_rate": 9.501392168237914e-06,
+      "loss": 0.3901,
+      "step": 2732
+    },
+    {
+      "epoch": 0.687374245472837,
+      "grad_norm": 0.35712650418281555,
+      "learning_rate": 9.500755002533109e-06,
+      "loss": 0.3725,
+      "step": 2733
+    },
+    {
+      "epoch": 0.687625754527163,
+      "grad_norm": 0.38140082359313965,
+      "learning_rate": 9.50011745136679e-06,
+      "loss": 0.4035,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6878772635814889,
+      "grad_norm": 0.375500351190567,
+      "learning_rate": 9.499479514793568e-06,
+      "loss": 0.3949,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6881287726358148,
+      "grad_norm": 0.3318289816379547,
+      "learning_rate": 9.49884119286807e-06,
+      "loss": 0.3663,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6883802816901409,
+      "grad_norm": 0.4133455753326416,
+      "learning_rate": 9.498202485644972e-06,
+      "loss": 0.3892,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6886317907444668,
+      "grad_norm": 0.34787869453430176,
+      "learning_rate": 9.49756339317897e-06,
+      "loss": 0.4052,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6888832997987927,
+      "grad_norm": 0.3496398627758026,
+      "learning_rate": 9.4969239155248e-06,
+      "loss": 0.3995,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6891348088531187,
+      "grad_norm": 0.3664000630378723,
+      "learning_rate": 9.49628405273723e-06,
+      "loss": 0.3926,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6893863179074446,
+      "grad_norm": 0.38853469491004944,
+      "learning_rate": 9.49564380487106e-06,
+      "loss": 0.3909,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6896378269617707,
+      "grad_norm": 0.3786064684391022,
+      "learning_rate": 9.495003171981122e-06,
+      "loss": 0.392,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6898893360160966,
+      "grad_norm": 0.3976394534111023,
+      "learning_rate": 9.494362154122283e-06,
+      "loss": 0.3786,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6901408450704225,
+      "grad_norm": 0.36655136942863464,
+      "learning_rate": 9.493720751349442e-06,
+      "loss": 0.3733,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6903923541247485,
+      "grad_norm": 0.32457855343818665,
+      "learning_rate": 9.493078963717533e-06,
+      "loss": 0.3873,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6906438631790744,
+      "grad_norm": 0.3550087511539459,
+      "learning_rate": 9.492436791281516e-06,
+      "loss": 0.3992,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6908953722334004,
+      "grad_norm": 0.357652872800827,
+      "learning_rate": 9.491794234096396e-06,
+      "loss": 0.3859,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6911468812877264,
+      "grad_norm": 0.3854414224624634,
+      "learning_rate": 9.491151292217198e-06,
+      "loss": 0.4151,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6913983903420523,
+      "grad_norm": 0.34755367040634155,
+      "learning_rate": 9.490507965698988e-06,
+      "loss": 0.3522,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6916498993963782,
+      "grad_norm": 0.35060808062553406,
+      "learning_rate": 9.489864254596866e-06,
+      "loss": 0.3856,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6919014084507042,
+      "grad_norm": 0.3610481023788452,
+      "learning_rate": 9.489220158965957e-06,
+      "loss": 0.393,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6921529175050302,
+      "grad_norm": 0.32515424489974976,
+      "learning_rate": 9.488575678861426e-06,
+      "loss": 0.3973,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6924044265593562,
+      "grad_norm": 0.3888111412525177,
+      "learning_rate": 9.487930814338468e-06,
+      "loss": 0.3946,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6926559356136821,
+      "grad_norm": 0.3749738931655884,
+      "learning_rate": 9.487285565452313e-06,
+      "loss": 0.3777,
+      "step": 2754
+    },
+    {
+      "epoch": 0.692907444668008,
+      "grad_norm": 0.41394150257110596,
+      "learning_rate": 9.486639932258223e-06,
+      "loss": 0.3777,
+      "step": 2755
+    },
+    {
+      "epoch": 0.693158953722334,
+      "grad_norm": 0.35159727931022644,
+      "learning_rate": 9.485993914811488e-06,
+      "loss": 0.3798,
+      "step": 2756
+    },
+    {
+      "epoch": 0.69341046277666,
+      "grad_norm": 0.36780303716659546,
+      "learning_rate": 9.485347513167443e-06,
+      "loss": 0.382,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6936619718309859,
+      "grad_norm": 0.3663868308067322,
+      "learning_rate": 9.48470072738144e-06,
+      "loss": 0.3858,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6939134808853119,
+      "grad_norm": 0.3358771502971649,
+      "learning_rate": 9.484053557508876e-06,
+      "loss": 0.4216,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6941649899396378,
+      "grad_norm": 0.3996930718421936,
+      "learning_rate": 9.483406003605178e-06,
+      "loss": 0.3943,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6944164989939637,
+      "grad_norm": 0.41368430852890015,
+      "learning_rate": 9.482758065725805e-06,
+      "loss": 0.3836,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6946680080482898,
+      "grad_norm": 0.4018966555595398,
+      "learning_rate": 9.482109743926247e-06,
+      "loss": 0.3969,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6949195171026157,
+      "grad_norm": 0.3682190477848053,
+      "learning_rate": 9.481461038262027e-06,
+      "loss": 0.3671,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6951710261569416,
+      "grad_norm": 0.36615151166915894,
+      "learning_rate": 9.480811948788708e-06,
+      "loss": 0.378,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6954225352112676,
+      "grad_norm": 0.3735343813896179,
+      "learning_rate": 9.480162475561877e-06,
+      "loss": 0.3756,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6956740442655935,
+      "grad_norm": 0.37882131338119507,
+      "learning_rate": 9.479512618637156e-06,
+      "loss": 0.3736,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6959255533199196,
+      "grad_norm": 0.355418860912323,
+      "learning_rate": 9.478862378070204e-06,
+      "loss": 0.3856,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6961770623742455,
+      "grad_norm": 0.34449321031570435,
+      "learning_rate": 9.47821175391671e-06,
+      "loss": 0.3713,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6964285714285714,
+      "grad_norm": 0.39640021324157715,
+      "learning_rate": 9.477560746232394e-06,
+      "loss": 0.379,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6966800804828974,
+      "grad_norm": 0.3537764251232147,
+      "learning_rate": 9.476909355073012e-06,
+      "loss": 0.3901,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6969315895372233,
+      "grad_norm": 0.34988686442375183,
+      "learning_rate": 9.47625758049435e-06,
+      "loss": 0.3643,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6971830985915493,
+      "grad_norm": 0.39470940828323364,
+      "learning_rate": 9.47560542255223e-06,
+      "loss": 0.3841,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6974346076458753,
+      "grad_norm": 0.3538375496864319,
+      "learning_rate": 9.474952881302506e-06,
+      "loss": 0.3895,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6976861167002012,
+      "grad_norm": 0.34231036901474,
+      "learning_rate": 9.474299956801062e-06,
+      "loss": 0.385,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6979376257545271,
+      "grad_norm": 0.3452184796333313,
+      "learning_rate": 9.473646649103819e-06,
+      "loss": 0.3913,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6981891348088531,
+      "grad_norm": 0.3894374966621399,
+      "learning_rate": 9.472992958266725e-06,
+      "loss": 0.3922,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6984406438631791,
+      "grad_norm": 0.35543131828308105,
+      "learning_rate": 9.47233888434577e-06,
+      "loss": 0.373,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6986921529175051,
+      "grad_norm": 0.3584343194961548,
+      "learning_rate": 9.471684427396966e-06,
+      "loss": 0.3948,
+      "step": 2778
+    },
+    {
+      "epoch": 0.698943661971831,
+      "grad_norm": 0.35712918639183044,
+      "learning_rate": 9.471029587476367e-06,
+      "loss": 0.4008,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6991951710261569,
+      "grad_norm": 0.36257344484329224,
+      "learning_rate": 9.470374364640054e-06,
+      "loss": 0.4058,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6994466800804829,
+      "grad_norm": 0.37731629610061646,
+      "learning_rate": 9.469718758944144e-06,
+      "loss": 0.4053,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6996981891348089,
+      "grad_norm": 0.3586489260196686,
+      "learning_rate": 9.469062770444784e-06,
+      "loss": 0.402,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6999496981891348,
+      "grad_norm": 0.36410945653915405,
+      "learning_rate": 9.468406399198156e-06,
+      "loss": 0.3954,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7002012072434608,
+      "grad_norm": 0.40461426973342896,
+      "learning_rate": 9.467749645260475e-06,
+      "loss": 0.406,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7004527162977867,
+      "grad_norm": 0.36193108558654785,
+      "learning_rate": 9.467092508687987e-06,
+      "loss": 0.3755,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7007042253521126,
+      "grad_norm": 0.32516640424728394,
+      "learning_rate": 9.46643498953697e-06,
+      "loss": 0.3889,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7009557344064387,
+      "grad_norm": 0.3632727861404419,
+      "learning_rate": 9.46577708786374e-06,
+      "loss": 0.4103,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7012072434607646,
+      "grad_norm": 0.3422258496284485,
+      "learning_rate": 9.46511880372464e-06,
+      "loss": 0.3833,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7014587525150905,
+      "grad_norm": 0.386264830827713,
+      "learning_rate": 9.464460137176047e-06,
+      "loss": 0.3842,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7017102615694165,
+      "grad_norm": 0.33963924646377563,
+      "learning_rate": 9.463801088274374e-06,
+      "loss": 0.3707,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7019617706237424,
+      "grad_norm": 0.3681434094905853,
+      "learning_rate": 9.463141657076063e-06,
+      "loss": 0.3914,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7022132796780685,
+      "grad_norm": 0.3675888180732727,
+      "learning_rate": 9.46248184363759e-06,
+      "loss": 0.3785,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7024647887323944,
+      "grad_norm": 0.3703523278236389,
+      "learning_rate": 9.461821648015464e-06,
+      "loss": 0.4011,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7027162977867203,
+      "grad_norm": 0.34886541962623596,
+      "learning_rate": 9.461161070266226e-06,
+      "loss": 0.3953,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7029678068410463,
+      "grad_norm": 0.34929558634757996,
+      "learning_rate": 9.460500110446453e-06,
+      "loss": 0.4049,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7032193158953722,
+      "grad_norm": 0.3256225883960724,
+      "learning_rate": 9.459838768612751e-06,
+      "loss": 0.3611,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7034708249496981,
+      "grad_norm": 0.34201568365097046,
+      "learning_rate": 9.459177044821758e-06,
+      "loss": 0.3878,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7037223340040242,
+      "grad_norm": 0.32223188877105713,
+      "learning_rate": 9.458514939130148e-06,
+      "loss": 0.3879,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7039738430583501,
+      "grad_norm": 0.38301658630371094,
+      "learning_rate": 9.457852451594625e-06,
+      "loss": 0.3776,
+      "step": 2799
+    },
+    {
+      "epoch": 0.704225352112676,
+      "grad_norm": 0.3359746038913727,
+      "learning_rate": 9.457189582271928e-06,
+      "loss": 0.3815,
+      "step": 2800
+    },
+    {
+      "epoch": 0.704476861167002,
+      "grad_norm": 0.42329642176628113,
+      "learning_rate": 9.456526331218827e-06,
+      "loss": 0.3806,
+      "step": 2801
+    },
+    {
+      "epoch": 0.704728370221328,
+      "grad_norm": 0.43376052379608154,
+      "learning_rate": 9.455862698492127e-06,
+      "loss": 0.3921,
+      "step": 2802
+    },
+    {
+      "epoch": 0.704979879275654,
+      "grad_norm": 0.3609297275543213,
+      "learning_rate": 9.455198684148662e-06,
+      "loss": 0.3749,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7052313883299799,
+      "grad_norm": 0.3790986239910126,
+      "learning_rate": 9.454534288245302e-06,
+      "loss": 0.3872,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7054828973843058,
+      "grad_norm": 0.36633041501045227,
+      "learning_rate": 9.453869510838946e-06,
+      "loss": 0.4007,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7057344064386318,
+      "grad_norm": 0.4222126305103302,
+      "learning_rate": 9.45320435198653e-06,
+      "loss": 0.3949,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7059859154929577,
+      "grad_norm": 0.34727922081947327,
+      "learning_rate": 9.452538811745023e-06,
+      "loss": 0.3738,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7062374245472837,
+      "grad_norm": 0.371955543756485,
+      "learning_rate": 9.451872890171419e-06,
+      "loss": 0.3913,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7064889336016097,
+      "grad_norm": 0.3855418860912323,
+      "learning_rate": 9.451206587322754e-06,
+      "loss": 0.367,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7067404426559356,
+      "grad_norm": 0.3608242869377136,
+      "learning_rate": 9.450539903256091e-06,
+      "loss": 0.3805,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7069919517102615,
+      "grad_norm": 0.3822399079799652,
+      "learning_rate": 9.449872838028529e-06,
+      "loss": 0.3969,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7072434607645876,
+      "grad_norm": 0.35486891865730286,
+      "learning_rate": 9.449205391697196e-06,
+      "loss": 0.3947,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7074949698189135,
+      "grad_norm": 0.3677852749824524,
+      "learning_rate": 9.448537564319254e-06,
+      "loss": 0.3856,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7077464788732394,
+      "grad_norm": 0.36130937933921814,
+      "learning_rate": 9.447869355951901e-06,
+      "loss": 0.401,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7079979879275654,
+      "grad_norm": 0.35179662704467773,
+      "learning_rate": 9.447200766652363e-06,
+      "loss": 0.3734,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7082494969818913,
+      "grad_norm": 0.3460598289966583,
+      "learning_rate": 9.446531796477901e-06,
+      "loss": 0.3591,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7085010060362174,
+      "grad_norm": 0.34915661811828613,
+      "learning_rate": 9.445862445485808e-06,
+      "loss": 0.4048,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7087525150905433,
+      "grad_norm": 0.38304591178894043,
+      "learning_rate": 9.44519271373341e-06,
+      "loss": 0.3861,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7090040241448692,
+      "grad_norm": 0.4027135968208313,
+      "learning_rate": 9.444522601278065e-06,
+      "loss": 0.4149,
+      "step": 2819
+    },
+    {
+      "epoch": 0.7092555331991952,
+      "grad_norm": 0.3455667793750763,
+      "learning_rate": 9.443852108177164e-06,
+      "loss": 0.3671,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7095070422535211,
+      "grad_norm": 0.4052219092845917,
+      "learning_rate": 9.44318123448813e-06,
+      "loss": 0.3654,
+      "step": 2821
+    },
+    {
+      "epoch": 0.709758551307847,
+      "grad_norm": 0.36053210496902466,
+      "learning_rate": 9.44250998026842e-06,
+      "loss": 0.4084,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7100100603621731,
+      "grad_norm": 0.3680732846260071,
+      "learning_rate": 9.441838345575523e-06,
+      "loss": 0.3848,
+      "step": 2823
+    },
+    {
+      "epoch": 0.710261569416499,
+      "grad_norm": 0.36795133352279663,
+      "learning_rate": 9.441166330466959e-06,
+      "loss": 0.3666,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7105130784708249,
+      "grad_norm": 0.3703378438949585,
+      "learning_rate": 9.440493935000283e-06,
+      "loss": 0.3959,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7107645875251509,
+      "grad_norm": 0.37262654304504395,
+      "learning_rate": 9.439821159233083e-06,
+      "loss": 0.3584,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7110160965794768,
+      "grad_norm": 0.35950973629951477,
+      "learning_rate": 9.439148003222973e-06,
+      "loss": 0.373,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7112676056338029,
+      "grad_norm": 0.3705592751502991,
+      "learning_rate": 9.43847446702761e-06,
+      "loss": 0.4044,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7115191146881288,
+      "grad_norm": 0.3991115987300873,
+      "learning_rate": 9.437800550704674e-06,
+      "loss": 0.3979,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7117706237424547,
+      "grad_norm": 0.37263038754463196,
+      "learning_rate": 9.437126254311886e-06,
+      "loss": 0.3845,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7120221327967807,
+      "grad_norm": 0.389017790555954,
+      "learning_rate": 9.436451577906991e-06,
+      "loss": 0.392,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7122736418511066,
+      "grad_norm": 0.33222588896751404,
+      "learning_rate": 9.435776521547772e-06,
+      "loss": 0.3847,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7125251509054326,
+      "grad_norm": 0.342219740152359,
+      "learning_rate": 9.435101085292047e-06,
+      "loss": 0.3885,
+      "step": 2833
+    },
+    {
+      "epoch": 0.7127766599597586,
+      "grad_norm": 0.3644763231277466,
+      "learning_rate": 9.434425269197658e-06,
+      "loss": 0.3635,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7130281690140845,
+      "grad_norm": 0.37079551815986633,
+      "learning_rate": 9.433749073322487e-06,
+      "loss": 0.4046,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7132796780684104,
+      "grad_norm": 0.39011722803115845,
+      "learning_rate": 9.433072497724445e-06,
+      "loss": 0.3589,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7135311871227364,
+      "grad_norm": 0.36313149333000183,
+      "learning_rate": 9.432395542461476e-06,
+      "loss": 0.3964,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7137826961770624,
+      "grad_norm": 0.35538873076438904,
+      "learning_rate": 9.431718207591559e-06,
+      "loss": 0.3914,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7140342052313883,
+      "grad_norm": 0.3488622307777405,
+      "learning_rate": 9.431040493172702e-06,
+      "loss": 0.3841,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.4144222140312195,
+      "learning_rate": 9.430362399262947e-06,
+      "loss": 0.3963,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7145372233400402,
+      "grad_norm": 0.34440502524375916,
+      "learning_rate": 9.429683925920369e-06,
+      "loss": 0.3667,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7147887323943662,
+      "grad_norm": 0.33740341663360596,
+      "learning_rate": 9.429005073203075e-06,
+      "loss": 0.3964,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7150402414486922,
+      "grad_norm": 0.38044601678848267,
+      "learning_rate": 9.428325841169203e-06,
+      "loss": 0.4139,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7152917505030181,
+      "grad_norm": 0.39153364300727844,
+      "learning_rate": 9.427646229876927e-06,
+      "loss": 0.3849,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7155432595573441,
+      "grad_norm": 0.3845551908016205,
+      "learning_rate": 9.42696623938445e-06,
+      "loss": 0.4071,
+      "step": 2845
+    },
+    {
+      "epoch": 0.71579476861167,
+      "grad_norm": 0.32743361592292786,
+      "learning_rate": 9.426285869750012e-06,
+      "loss": 0.3933,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7160462776659959,
+      "grad_norm": 0.48608964681625366,
+      "learning_rate": 9.425605121031878e-06,
+      "loss": 0.4126,
+      "step": 2847
+    },
+    {
+      "epoch": 0.716297786720322,
+      "grad_norm": 0.42301592230796814,
+      "learning_rate": 9.424923993288352e-06,
+      "loss": 0.3774,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7165492957746479,
+      "grad_norm": 0.3551022708415985,
+      "learning_rate": 9.424242486577768e-06,
+      "loss": 0.3884,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7168008048289738,
+      "grad_norm": 0.4087948799133301,
+      "learning_rate": 9.423560600958493e-06,
+      "loss": 0.3816,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7170523138832998,
+      "grad_norm": 0.37452781200408936,
+      "learning_rate": 9.422878336488928e-06,
+      "loss": 0.3793,
+      "step": 2851
+    },
+    {
+      "epoch": 0.7173038229376257,
+      "grad_norm": 0.35576769709587097,
+      "learning_rate": 9.422195693227501e-06,
+      "loss": 0.3753,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7175553319919518,
+      "grad_norm": 0.3551032841205597,
+      "learning_rate": 9.42151267123268e-06,
+      "loss": 0.3923,
+      "step": 2853
+    },
+    {
+      "epoch": 0.7178068410462777,
+      "grad_norm": 0.37760961055755615,
+      "learning_rate": 9.420829270562956e-06,
+      "loss": 0.4031,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7180583501006036,
+      "grad_norm": 0.3662315011024475,
+      "learning_rate": 9.420145491276864e-06,
+      "loss": 0.3539,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7183098591549296,
+      "grad_norm": 0.3553292751312256,
+      "learning_rate": 9.419461333432965e-06,
+      "loss": 0.3812,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7185613682092555,
+      "grad_norm": 0.4659683108329773,
+      "learning_rate": 9.418776797089848e-06,
+      "loss": 0.3933,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7188128772635815,
+      "grad_norm": 0.37304118275642395,
+      "learning_rate": 9.418091882306141e-06,
+      "loss": 0.3758,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7190643863179075,
+      "grad_norm": 0.36666375398635864,
+      "learning_rate": 9.417406589140507e-06,
+      "loss": 0.4104,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7193158953722334,
+      "grad_norm": 0.42461302876472473,
+      "learning_rate": 9.416720917651631e-06,
+      "loss": 0.3893,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7195674044265593,
+      "grad_norm": 0.4254009425640106,
+      "learning_rate": 9.416034867898243e-06,
+      "loss": 0.3885,
+      "step": 2861
+    },
+    {
+      "epoch": 0.7198189134808853,
+      "grad_norm": 0.3596349060535431,
+      "learning_rate": 9.415348439939091e-06,
+      "loss": 0.3596,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7200704225352113,
+      "grad_norm": 0.3977017104625702,
+      "learning_rate": 9.41466163383297e-06,
+      "loss": 0.3848,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7203219315895373,
+      "grad_norm": 0.40862926840782166,
+      "learning_rate": 9.4139744496387e-06,
+      "loss": 0.4016,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7205734406438632,
+      "grad_norm": 0.3802616596221924,
+      "learning_rate": 9.413286887415128e-06,
+      "loss": 0.3909,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7208249496981891,
+      "grad_norm": 0.34951651096343994,
+      "learning_rate": 9.412598947221146e-06,
+      "loss": 0.3707,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7210764587525151,
+      "grad_norm": 0.36474528908729553,
+      "learning_rate": 9.411910629115667e-06,
+      "loss": 0.4124,
+      "step": 2867
+    },
+    {
+      "epoch": 0.721327967806841,
+      "grad_norm": 0.3925243020057678,
+      "learning_rate": 9.411221933157646e-06,
+      "loss": 0.3872,
+      "step": 2868
+    },
+    {
+      "epoch": 0.721579476861167,
+      "grad_norm": 0.3429969251155853,
+      "learning_rate": 9.41053285940606e-06,
+      "loss": 0.3867,
+      "step": 2869
+    },
+    {
+      "epoch": 0.721830985915493,
+      "grad_norm": 0.35117414593696594,
+      "learning_rate": 9.409843407919929e-06,
+      "loss": 0.3853,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7220824949698189,
+      "grad_norm": 0.35899850726127625,
+      "learning_rate": 9.409153578758298e-06,
+      "loss": 0.3932,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7223340040241448,
+      "grad_norm": 0.37349846959114075,
+      "learning_rate": 9.408463371980248e-06,
+      "loss": 0.3779,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7225855130784709,
+      "grad_norm": 0.3280394375324249,
+      "learning_rate": 9.407772787644887e-06,
+      "loss": 0.3697,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7228370221327968,
+      "grad_norm": 0.36341550946235657,
+      "learning_rate": 9.407081825811362e-06,
+      "loss": 0.3828,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7230885311871227,
+      "grad_norm": 0.36753761768341064,
+      "learning_rate": 9.40639048653885e-06,
+      "loss": 0.384,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7233400402414487,
+      "grad_norm": 0.3492330014705658,
+      "learning_rate": 9.405698769886557e-06,
+      "loss": 0.3842,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7235915492957746,
+      "grad_norm": 0.40695223212242126,
+      "learning_rate": 9.405006675913729e-06,
+      "loss": 0.4184,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7238430583501007,
+      "grad_norm": 0.342252641916275,
+      "learning_rate": 9.404314204679636e-06,
+      "loss": 0.3739,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7240945674044266,
+      "grad_norm": 0.39723700284957886,
+      "learning_rate": 9.403621356243584e-06,
+      "loss": 0.4288,
+      "step": 2879
+    },
+    {
+      "epoch": 0.7243460764587525,
+      "grad_norm": 0.3767862021923065,
+      "learning_rate": 9.402928130664913e-06,
+      "loss": 0.3883,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7245975855130785,
+      "grad_norm": 0.37945619225502014,
+      "learning_rate": 9.402234528002991e-06,
+      "loss": 0.406,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7248490945674044,
+      "grad_norm": 0.3551163375377655,
+      "learning_rate": 9.401540548317223e-06,
+      "loss": 0.3808,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7251006036217303,
+      "grad_norm": 0.3756506145000458,
+      "learning_rate": 9.400846191667043e-06,
+      "loss": 0.3665,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7253521126760564,
+      "grad_norm": 0.38300275802612305,
+      "learning_rate": 9.400151458111918e-06,
+      "loss": 0.3968,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7256036217303823,
+      "grad_norm": 0.37950319051742554,
+      "learning_rate": 9.399456347711348e-06,
+      "loss": 0.3843,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7258551307847082,
+      "grad_norm": 0.3413698673248291,
+      "learning_rate": 9.398760860524865e-06,
+      "loss": 0.3965,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7261066398390342,
+      "grad_norm": 0.37162476778030396,
+      "learning_rate": 9.398064996612032e-06,
+      "loss": 0.3786,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7263581488933601,
+      "grad_norm": 0.3777274787425995,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.4001,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7266096579476862,
+      "grad_norm": 0.3456476032733917,
+      "learning_rate": 9.396672138845737e-06,
+      "loss": 0.3708,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7268611670020121,
+      "grad_norm": 0.41536620259284973,
+      "learning_rate": 9.395975145111565e-06,
+      "loss": 0.3886,
+      "step": 2890
+    },
+    {
+      "epoch": 0.727112676056338,
+      "grad_norm": 0.3782283067703247,
+      "learning_rate": 9.395277774889621e-06,
+      "loss": 0.3711,
+      "step": 2891
+    },
+    {
+      "epoch": 0.727364185110664,
+      "grad_norm": 0.31303125619888306,
+      "learning_rate": 9.394580028239633e-06,
+      "loss": 0.366,
+      "step": 2892
+    },
+    {
+      "epoch": 0.72761569416499,
+      "grad_norm": 0.3784593641757965,
+      "learning_rate": 9.39388190522136e-06,
+      "loss": 0.4001,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7278672032193159,
+      "grad_norm": 0.43708619475364685,
+      "learning_rate": 9.393183405894589e-06,
+      "loss": 0.3977,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7281187122736419,
+      "grad_norm": 0.34523478150367737,
+      "learning_rate": 9.39248453031914e-06,
+      "loss": 0.4209,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7283702213279678,
+      "grad_norm": 0.37533217668533325,
+      "learning_rate": 9.391785278554875e-06,
+      "loss": 0.3937,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7286217303822937,
+      "grad_norm": 0.42460107803344727,
+      "learning_rate": 9.391085650661672e-06,
+      "loss": 0.3885,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7288732394366197,
+      "grad_norm": 0.3811191916465759,
+      "learning_rate": 9.390385646699457e-06,
+      "loss": 0.3899,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7291247484909457,
+      "grad_norm": 0.41132891178131104,
+      "learning_rate": 9.389685266728175e-06,
+      "loss": 0.3725,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7293762575452716,
+      "grad_norm": 0.4103216230869293,
+      "learning_rate": 9.388984510807812e-06,
+      "loss": 0.3838,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7296277665995976,
+      "grad_norm": 0.399691641330719,
+      "learning_rate": 9.388283378998382e-06,
+      "loss": 0.3846,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7298792756539235,
+      "grad_norm": 0.4203183650970459,
+      "learning_rate": 9.387581871359936e-06,
+      "loss": 0.3557,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7301307847082495,
+      "grad_norm": 0.3951526880264282,
+      "learning_rate": 9.386879987952549e-06,
+      "loss": 0.3983,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7303822937625755,
+      "grad_norm": 0.3411506414413452,
+      "learning_rate": 9.386177728836337e-06,
+      "loss": 0.3993,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7306338028169014,
+      "grad_norm": 0.3646346628665924,
+      "learning_rate": 9.385475094071442e-06,
+      "loss": 0.3693,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7308853118712274,
+      "grad_norm": 0.39517220854759216,
+      "learning_rate": 9.384772083718042e-06,
+      "loss": 0.3887,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7311368209255533,
+      "grad_norm": 0.36498531699180603,
+      "learning_rate": 9.384068697836342e-06,
+      "loss": 0.3819,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7313883299798792,
+      "grad_norm": 0.36910420656204224,
+      "learning_rate": 9.383364936486585e-06,
+      "loss": 0.3941,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7316398390342053,
+      "grad_norm": 0.3762173652648926,
+      "learning_rate": 9.382660799729044e-06,
+      "loss": 0.3762,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7318913480885312,
+      "grad_norm": 0.40564772486686707,
+      "learning_rate": 9.381956287624024e-06,
+      "loss": 0.3928,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7321428571428571,
+      "grad_norm": 0.354449063539505,
+      "learning_rate": 9.381251400231859e-06,
+      "loss": 0.4233,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 0.3970174789428711,
+      "learning_rate": 9.380546137612922e-06,
+      "loss": 0.4486,
+      "step": 2912
+    },
+    {
+      "epoch": 0.732645875251509,
+      "grad_norm": 0.3722879886627197,
+      "learning_rate": 9.379840499827612e-06,
+      "loss": 0.3981,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7328973843058351,
+      "grad_norm": 0.3905206024646759,
+      "learning_rate": 9.379134486936366e-06,
+      "loss": 0.3734,
+      "step": 2914
+    },
+    {
+      "epoch": 0.733148893360161,
+      "grad_norm": 0.3671495020389557,
+      "learning_rate": 9.378428098999645e-06,
+      "loss": 0.3537,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7334004024144869,
+      "grad_norm": 0.3310393691062927,
+      "learning_rate": 9.37772133607795e-06,
+      "loss": 0.3933,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7336519114688129,
+      "grad_norm": 0.3759323060512543,
+      "learning_rate": 9.377014198231807e-06,
+      "loss": 0.3828,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7339034205231388,
+      "grad_norm": 0.36852389574050903,
+      "learning_rate": 9.376306685521784e-06,
+      "loss": 0.3723,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7341549295774648,
+      "grad_norm": 0.3421872854232788,
+      "learning_rate": 9.375598798008468e-06,
+      "loss": 0.3778,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7344064386317908,
+      "grad_norm": 0.37425142526626587,
+      "learning_rate": 9.37489053575249e-06,
+      "loss": 0.3804,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7346579476861167,
+      "grad_norm": 0.3890654444694519,
+      "learning_rate": 9.374181898814508e-06,
+      "loss": 0.3768,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7349094567404426,
+      "grad_norm": 0.35745400190353394,
+      "learning_rate": 9.373472887255209e-06,
+      "loss": 0.3843,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7351609657947686,
+      "grad_norm": 0.3883761465549469,
+      "learning_rate": 9.372763501135319e-06,
+      "loss": 0.3916,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7354124748490946,
+      "grad_norm": 0.393292635679245,
+      "learning_rate": 9.37205374051559e-06,
+      "loss": 0.3977,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7356639839034205,
+      "grad_norm": 0.34699979424476624,
+      "learning_rate": 9.37134360545681e-06,
+      "loss": 0.383,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7359154929577465,
+      "grad_norm": 0.40010884404182434,
+      "learning_rate": 9.370633096019799e-06,
+      "loss": 0.3855,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7361670020120724,
+      "grad_norm": 0.36946627497673035,
+      "learning_rate": 9.369922212265403e-06,
+      "loss": 0.3852,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7364185110663984,
+      "grad_norm": 0.4117673635482788,
+      "learning_rate": 9.36921095425451e-06,
+      "loss": 0.3849,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7366700201207244,
+      "grad_norm": 0.36381223797798157,
+      "learning_rate": 9.368499322048031e-06,
+      "loss": 0.3921,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7369215291750503,
+      "grad_norm": 0.3884536027908325,
+      "learning_rate": 9.367787315706916e-06,
+      "loss": 0.3812,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7371730382293763,
+      "grad_norm": 0.35373517870903015,
+      "learning_rate": 9.36707493529214e-06,
+      "loss": 0.3656,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7374245472837022,
+      "grad_norm": 0.3612503111362457,
+      "learning_rate": 9.366362180864718e-06,
+      "loss": 0.3782,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7376760563380281,
+      "grad_norm": 0.37395453453063965,
+      "learning_rate": 9.36564905248569e-06,
+      "loss": 0.389,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7379275653923542,
+      "grad_norm": 0.3890708088874817,
+      "learning_rate": 9.364935550216133e-06,
+      "loss": 0.3922,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7381790744466801,
+      "grad_norm": 0.38554948568344116,
+      "learning_rate": 9.364221674117151e-06,
+      "loss": 0.4182,
+      "step": 2935
+    },
+    {
+      "epoch": 0.738430583501006,
+      "grad_norm": 0.3420090079307556,
+      "learning_rate": 9.363507424249887e-06,
+      "loss": 0.4097,
+      "step": 2936
+    },
+    {
+      "epoch": 0.738682092555332,
+      "grad_norm": 0.3459598124027252,
+      "learning_rate": 9.362792800675511e-06,
+      "loss": 0.3826,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7389336016096579,
+      "grad_norm": 0.35301342606544495,
+      "learning_rate": 9.362077803455223e-06,
+      "loss": 0.3986,
+      "step": 2938
+    },
+    {
+      "epoch": 0.739185110663984,
+      "grad_norm": 0.37958618998527527,
+      "learning_rate": 9.361362432650261e-06,
+      "loss": 0.4021,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7394366197183099,
+      "grad_norm": 0.362627238035202,
+      "learning_rate": 9.360646688321891e-06,
+      "loss": 0.3895,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7396881287726358,
+      "grad_norm": 0.37005671858787537,
+      "learning_rate": 9.359930570531412e-06,
+      "loss": 0.3918,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7399396378269618,
+      "grad_norm": 0.34973013401031494,
+      "learning_rate": 9.359214079340158e-06,
+      "loss": 0.3559,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7401911468812877,
+      "grad_norm": 0.4032422602176666,
+      "learning_rate": 9.358497214809485e-06,
+      "loss": 0.3907,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7404426559356136,
+      "grad_norm": 0.36071479320526123,
+      "learning_rate": 9.357779977000796e-06,
+      "loss": 0.4044,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7406941649899397,
+      "grad_norm": 0.4488549828529358,
+      "learning_rate": 9.357062365975511e-06,
+      "loss": 0.3727,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7409456740442656,
+      "grad_norm": 0.37173953652381897,
+      "learning_rate": 9.356344381795094e-06,
+      "loss": 0.3777,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7411971830985915,
+      "grad_norm": 0.3299272954463959,
+      "learning_rate": 9.355626024521035e-06,
+      "loss": 0.3997,
+      "step": 2947
+    },
+    {
+      "epoch": 0.7414486921529175,
+      "grad_norm": 0.33907902240753174,
+      "learning_rate": 9.354907294214853e-06,
+      "loss": 0.3877,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7417002012072434,
+      "grad_norm": 0.3761814832687378,
+      "learning_rate": 9.354188190938108e-06,
+      "loss": 0.3763,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7419517102615694,
+      "grad_norm": 0.37312766909599304,
+      "learning_rate": 9.353468714752381e-06,
+      "loss": 0.3922,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7422032193158954,
+      "grad_norm": 0.3855811059474945,
+      "learning_rate": 9.352748865719296e-06,
+      "loss": 0.3519,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7424547283702213,
+      "grad_norm": 0.3510308861732483,
+      "learning_rate": 9.352028643900502e-06,
+      "loss": 0.3622,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7427062374245473,
+      "grad_norm": 0.36363258957862854,
+      "learning_rate": 9.351308049357679e-06,
+      "loss": 0.403,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7429577464788732,
+      "grad_norm": 0.35569775104522705,
+      "learning_rate": 9.350587082152544e-06,
+      "loss": 0.3811,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7432092555331992,
+      "grad_norm": 0.32615721225738525,
+      "learning_rate": 9.349865742346842e-06,
+      "loss": 0.3984,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7434607645875252,
+      "grad_norm": 0.3774353563785553,
+      "learning_rate": 9.349144030002353e-06,
+      "loss": 0.379,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7437122736418511,
+      "grad_norm": 0.350462943315506,
+      "learning_rate": 9.348421945180885e-06,
+      "loss": 0.3865,
+      "step": 2957
+    },
+    {
+      "epoch": 0.743963782696177,
+      "grad_norm": 0.36871710419654846,
+      "learning_rate": 9.347699487944282e-06,
+      "loss": 0.3823,
+      "step": 2958
+    },
+    {
+      "epoch": 0.744215291750503,
+      "grad_norm": 0.35802212357521057,
+      "learning_rate": 9.346976658354417e-06,
+      "loss": 0.3924,
+      "step": 2959
+    },
+    {
+      "epoch": 0.744466800804829,
+      "grad_norm": 0.3417361080646515,
+      "learning_rate": 9.346253456473196e-06,
+      "loss": 0.3814,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7447183098591549,
+      "grad_norm": 0.38555172085762024,
+      "learning_rate": 9.345529882362554e-06,
+      "loss": 0.3986,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7449698189134809,
+      "grad_norm": 0.35595080256462097,
+      "learning_rate": 9.344805936084466e-06,
+      "loss": 0.3729,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7452213279678068,
+      "grad_norm": 0.36647212505340576,
+      "learning_rate": 9.344081617700929e-06,
+      "loss": 0.3817,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7454728370221329,
+      "grad_norm": 0.38703134655952454,
+      "learning_rate": 9.343356927273978e-06,
+      "loss": 0.3971,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7457243460764588,
+      "grad_norm": 0.3738643229007721,
+      "learning_rate": 9.342631864865678e-06,
+      "loss": 0.3807,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7459758551307847,
+      "grad_norm": 0.3684743046760559,
+      "learning_rate": 9.341906430538129e-06,
+      "loss": 0.3829,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7462273641851107,
+      "grad_norm": 0.37854307889938354,
+      "learning_rate": 9.341180624353454e-06,
+      "loss": 0.3737,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7464788732394366,
+      "grad_norm": 0.3924698829650879,
+      "learning_rate": 9.34045444637382e-06,
+      "loss": 0.3751,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7467303822937625,
+      "grad_norm": 0.3787640333175659,
+      "learning_rate": 9.339727896661413e-06,
+      "loss": 0.3736,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7469818913480886,
+      "grad_norm": 0.351458877325058,
+      "learning_rate": 9.339000975278463e-06,
+      "loss": 0.3986,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7472334004024145,
+      "grad_norm": 0.3286757469177246,
+      "learning_rate": 9.338273682287222e-06,
+      "loss": 0.3831,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7474849094567404,
+      "grad_norm": 0.3744126558303833,
+      "learning_rate": 9.337546017749981e-06,
+      "loss": 0.3868,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7477364185110664,
+      "grad_norm": 0.3369508981704712,
+      "learning_rate": 9.33681798172906e-06,
+      "loss": 0.3785,
+      "step": 2973
+    },
+    {
+      "epoch": 0.7479879275653923,
+      "grad_norm": 0.3627220690250397,
+      "learning_rate": 9.33608957428681e-06,
+      "loss": 0.3718,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7482394366197183,
+      "grad_norm": 0.36162272095680237,
+      "learning_rate": 9.335360795485615e-06,
+      "loss": 0.3863,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7484909456740443,
+      "grad_norm": 0.3735608160495758,
+      "learning_rate": 9.334631645387888e-06,
+      "loss": 0.3625,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7487424547283702,
+      "grad_norm": 0.34616851806640625,
+      "learning_rate": 9.33390212405608e-06,
+      "loss": 0.3572,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7489939637826962,
+      "grad_norm": 0.35577431321144104,
+      "learning_rate": 9.333172231552666e-06,
+      "loss": 0.3677,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7492454728370221,
+      "grad_norm": 0.3937419652938843,
+      "learning_rate": 9.332441967940161e-06,
+      "loss": 0.3848,
+      "step": 2979
+    },
+    {
+      "epoch": 0.7494969818913481,
+      "grad_norm": 0.3335328996181488,
+      "learning_rate": 9.331711333281101e-06,
+      "loss": 0.3998,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7497484909456741,
+      "grad_norm": 0.37114816904067993,
+      "learning_rate": 9.330980327638068e-06,
+      "loss": 0.3803,
+      "step": 2981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4506068229675293,
+      "learning_rate": 9.330248951073664e-06,
+      "loss": 0.4137,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7502515090543259,
+      "grad_norm": 0.3454316556453705,
+      "learning_rate": 9.329517203650526e-06,
+      "loss": 0.3696,
+      "step": 2983
+    },
+    {
+      "epoch": 0.7505030181086519,
+      "grad_norm": 0.38480788469314575,
+      "learning_rate": 9.328785085431326e-06,
+      "loss": 0.3937,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7507545271629779,
+      "grad_norm": 0.4474145472049713,
+      "learning_rate": 9.328052596478763e-06,
+      "loss": 0.3948,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7510060362173038,
+      "grad_norm": 0.3649021089076996,
+      "learning_rate": 9.327319736855574e-06,
+      "loss": 0.3645,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7512575452716298,
+      "grad_norm": 0.37888169288635254,
+      "learning_rate": 9.326586506624517e-06,
+      "loss": 0.3737,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7515090543259557,
+      "grad_norm": 0.38091379404067993,
+      "learning_rate": 9.325852905848396e-06,
+      "loss": 0.421,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7517605633802817,
+      "grad_norm": 0.38274237513542175,
+      "learning_rate": 9.325118934590036e-06,
+      "loss": 0.3658,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7520120724346077,
+      "grad_norm": 0.38675278425216675,
+      "learning_rate": 9.324384592912295e-06,
+      "loss": 0.3835,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7522635814889336,
+      "grad_norm": 0.3240179717540741,
+      "learning_rate": 9.323649880878069e-06,
+      "loss": 0.3706,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7525150905432596,
+      "grad_norm": 0.4843953549861908,
+      "learning_rate": 9.322914798550277e-06,
+      "loss": 0.4072,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7527665995975855,
+      "grad_norm": 0.40057262778282166,
+      "learning_rate": 9.32217934599188e-06,
+      "loss": 0.3835,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7530181086519114,
+      "grad_norm": 0.3491225838661194,
+      "learning_rate": 9.321443523265858e-06,
+      "loss": 0.3848,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7532696177062375,
+      "grad_norm": 0.4448799192905426,
+      "learning_rate": 9.320707330435235e-06,
+      "loss": 0.3585,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7535211267605634,
+      "grad_norm": 0.3571915328502655,
+      "learning_rate": 9.319970767563061e-06,
+      "loss": 0.4124,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7537726358148893,
+      "grad_norm": 0.3818497657775879,
+      "learning_rate": 9.319233834712413e-06,
+      "loss": 0.3861,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7540241448692153,
+      "grad_norm": 0.4860544204711914,
+      "learning_rate": 9.318496531946411e-06,
+      "loss": 0.3887,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7542756539235412,
+      "grad_norm": 0.3781794011592865,
+      "learning_rate": 9.317758859328194e-06,
+      "loss": 0.3826,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7545271629778671,
+      "grad_norm": 0.3716043531894684,
+      "learning_rate": 9.317020816920945e-06,
+      "loss": 0.365,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7547786720321932,
+      "grad_norm": 0.39104846119880676,
+      "learning_rate": 9.31628240478787e-06,
+      "loss": 0.3894,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7550301810865191,
+      "grad_norm": 0.3662312924861908,
+      "learning_rate": 9.31554362299221e-06,
+      "loss": 0.3853,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7552816901408451,
+      "grad_norm": 0.4046803414821625,
+      "learning_rate": 9.314804471597235e-06,
+      "loss": 0.3735,
+      "step": 3003
+    },
+    {
+      "epoch": 0.755533199195171,
+      "grad_norm": 0.36208009719848633,
+      "learning_rate": 9.314064950666252e-06,
+      "loss": 0.4042,
+      "step": 3004
+    },
+    {
+      "epoch": 0.755784708249497,
+      "grad_norm": 0.33379390835762024,
+      "learning_rate": 9.313325060262594e-06,
+      "loss": 0.3717,
+      "step": 3005
+    },
+    {
+      "epoch": 0.756036217303823,
+      "grad_norm": 0.3620927035808563,
+      "learning_rate": 9.312584800449629e-06,
+      "loss": 0.4094,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7562877263581489,
+      "grad_norm": 0.3750810921192169,
+      "learning_rate": 9.311844171290755e-06,
+      "loss": 0.3692,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7565392354124748,
+      "grad_norm": 0.35461145639419556,
+      "learning_rate": 9.311103172849404e-06,
+      "loss": 0.4094,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7567907444668008,
+      "grad_norm": 0.37598538398742676,
+      "learning_rate": 9.310361805189033e-06,
+      "loss": 0.3806,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7570422535211268,
+      "grad_norm": 0.4294974207878113,
+      "learning_rate": 9.309620068373143e-06,
+      "loss": 0.4,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7572937625754527,
+      "grad_norm": 0.3354276716709137,
+      "learning_rate": 9.308877962465251e-06,
+      "loss": 0.3908,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7575452716297787,
+      "grad_norm": 0.38359084725379944,
+      "learning_rate": 9.308135487528919e-06,
+      "loss": 0.3721,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7577967806841046,
+      "grad_norm": 0.37038716673851013,
+      "learning_rate": 9.307392643627736e-06,
+      "loss": 0.3704,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7580482897384306,
+      "grad_norm": 0.34647151827812195,
+      "learning_rate": 9.306649430825318e-06,
+      "loss": 0.3985,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7582997987927566,
+      "grad_norm": 0.3945516347885132,
+      "learning_rate": 9.30590584918532e-06,
+      "loss": 0.3832,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7585513078470825,
+      "grad_norm": 0.3935578763484955,
+      "learning_rate": 9.305161898771422e-06,
+      "loss": 0.3768,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7588028169014085,
+      "grad_norm": 0.37610089778900146,
+      "learning_rate": 9.304417579647343e-06,
+      "loss": 0.3763,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7590543259557344,
+      "grad_norm": 0.3979266881942749,
+      "learning_rate": 9.303672891876825e-06,
+      "loss": 0.4074,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7593058350100603,
+      "grad_norm": 0.36932116746902466,
+      "learning_rate": 9.302927835523647e-06,
+      "loss": 0.4137,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7595573440643864,
+      "grad_norm": 0.3783572018146515,
+      "learning_rate": 9.302182410651618e-06,
+      "loss": 0.3826,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7598088531187123,
+      "grad_norm": 0.34661227464675903,
+      "learning_rate": 9.301436617324584e-06,
+      "loss": 0.3745,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7600603621730382,
+      "grad_norm": 0.3569898307323456,
+      "learning_rate": 9.30069045560641e-06,
+      "loss": 0.3919,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7603118712273642,
+      "grad_norm": 0.43195685744285583,
+      "learning_rate": 9.299943925561004e-06,
+      "loss": 0.3983,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7605633802816901,
+      "grad_norm": 0.35552507638931274,
+      "learning_rate": 9.299197027252302e-06,
+      "loss": 0.3962,
+      "step": 3024
+    },
+    {
+      "epoch": 0.760814889336016,
+      "grad_norm": 0.3416507840156555,
+      "learning_rate": 9.29844976074427e-06,
+      "loss": 0.3906,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7610663983903421,
+      "grad_norm": 0.3881553113460541,
+      "learning_rate": 9.297702126100906e-06,
+      "loss": 0.3863,
+      "step": 3026
+    },
+    {
+      "epoch": 0.761317907444668,
+      "grad_norm": 0.3564288914203644,
+      "learning_rate": 9.296954123386243e-06,
+      "loss": 0.364,
+      "step": 3027
+    },
+    {
+      "epoch": 0.761569416498994,
+      "grad_norm": 0.3937084674835205,
+      "learning_rate": 9.29620575266434e-06,
+      "loss": 0.3769,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7618209255533199,
+      "grad_norm": 0.3818730115890503,
+      "learning_rate": 9.295457013999291e-06,
+      "loss": 0.3923,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7620724346076458,
+      "grad_norm": 0.4041476547718048,
+      "learning_rate": 9.294707907455223e-06,
+      "loss": 0.3904,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7623239436619719,
+      "grad_norm": 0.36864545941352844,
+      "learning_rate": 9.293958433096289e-06,
+      "loss": 0.419,
+      "step": 3031
+    },
+    {
+      "epoch": 0.7625754527162978,
+      "grad_norm": 0.37678250670433044,
+      "learning_rate": 9.293208590986676e-06,
+      "loss": 0.389,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7628269617706237,
+      "grad_norm": 0.35496804118156433,
+      "learning_rate": 9.292458381190608e-06,
+      "loss": 0.3996,
+      "step": 3033
+    },
+    {
+      "epoch": 0.7630784708249497,
+      "grad_norm": 0.3487403392791748,
+      "learning_rate": 9.291707803772332e-06,
+      "loss": 0.3565,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7633299798792756,
+      "grad_norm": 0.3654000461101532,
+      "learning_rate": 9.290956858796132e-06,
+      "loss": 0.3849,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7635814889336016,
+      "grad_norm": 0.3704080879688263,
+      "learning_rate": 9.29020554632632e-06,
+      "loss": 0.37,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7638329979879276,
+      "grad_norm": 0.3763080835342407,
+      "learning_rate": 9.289453866427245e-06,
+      "loss": 0.4109,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7640845070422535,
+      "grad_norm": 0.38172775506973267,
+      "learning_rate": 9.288701819163279e-06,
+      "loss": 0.3791,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7643360160965795,
+      "grad_norm": 0.34394824504852295,
+      "learning_rate": 9.287949404598833e-06,
+      "loss": 0.3799,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7645875251509054,
+      "grad_norm": 0.34752610325813293,
+      "learning_rate": 9.287196622798346e-06,
+      "loss": 0.374,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7648390342052314,
+      "grad_norm": 0.34505367279052734,
+      "learning_rate": 9.286443473826288e-06,
+      "loss": 0.3854,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7650905432595574,
+      "grad_norm": 0.3505760431289673,
+      "learning_rate": 9.285689957747163e-06,
+      "loss": 0.3881,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7653420523138833,
+      "grad_norm": 0.36895132064819336,
+      "learning_rate": 9.284936074625503e-06,
+      "loss": 0.3623,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7655935613682092,
+      "grad_norm": 0.3329405188560486,
+      "learning_rate": 9.284181824525877e-06,
+      "loss": 0.3676,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7658450704225352,
+      "grad_norm": 0.3532065749168396,
+      "learning_rate": 9.283427207512878e-06,
+      "loss": 0.3853,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7660965794768612,
+      "grad_norm": 0.3908108174800873,
+      "learning_rate": 9.282672223651137e-06,
+      "loss": 0.3951,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7663480885311871,
+      "grad_norm": 0.34189653396606445,
+      "learning_rate": 9.28191687300531e-06,
+      "loss": 0.3953,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7665995975855131,
+      "grad_norm": 0.357767790555954,
+      "learning_rate": 9.281161155640093e-06,
+      "loss": 0.402,
+      "step": 3048
+    },
+    {
+      "epoch": 0.766851106639839,
+      "grad_norm": 0.361261248588562,
+      "learning_rate": 9.280405071620204e-06,
+      "loss": 0.3652,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7671026156941649,
+      "grad_norm": 0.34182727336883545,
+      "learning_rate": 9.2796486210104e-06,
+      "loss": 0.3679,
+      "step": 3050
+    },
+    {
+      "epoch": 0.767354124748491,
+      "grad_norm": 0.3388752341270447,
+      "learning_rate": 9.278891803875466e-06,
+      "loss": 0.4031,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7676056338028169,
+      "grad_norm": 0.3388610780239105,
+      "learning_rate": 9.278134620280215e-06,
+      "loss": 0.3881,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7678571428571429,
+      "grad_norm": 0.32106414437294006,
+      "learning_rate": 9.277377070289498e-06,
+      "loss": 0.3708,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7681086519114688,
+      "grad_norm": 0.402169406414032,
+      "learning_rate": 9.276619153968197e-06,
+      "loss": 0.3991,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7683601609657947,
+      "grad_norm": 0.32417601346969604,
+      "learning_rate": 9.275860871381217e-06,
+      "loss": 0.3865,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7686116700201208,
+      "grad_norm": 0.3850818872451782,
+      "learning_rate": 9.275102222593503e-06,
+      "loss": 0.3899,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7688631790744467,
+      "grad_norm": 0.36075320839881897,
+      "learning_rate": 9.27434320767003e-06,
+      "loss": 0.3824,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7691146881287726,
+      "grad_norm": 0.37227320671081543,
+      "learning_rate": 9.2735838266758e-06,
+      "loss": 0.3926,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7693661971830986,
+      "grad_norm": 0.395997554063797,
+      "learning_rate": 9.272824079675854e-06,
+      "loss": 0.3692,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7696177062374245,
+      "grad_norm": 0.3681703507900238,
+      "learning_rate": 9.272063966735253e-06,
+      "loss": 0.4038,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7698692152917505,
+      "grad_norm": 0.3632868230342865,
+      "learning_rate": 9.2713034879191e-06,
+      "loss": 0.4239,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7701207243460765,
+      "grad_norm": 0.38830870389938354,
+      "learning_rate": 9.270542643292523e-06,
+      "loss": 0.3763,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7703722334004024,
+      "grad_norm": 0.35282665491104126,
+      "learning_rate": 9.269781432920688e-06,
+      "loss": 0.3827,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7706237424547284,
+      "grad_norm": 0.3645845949649811,
+      "learning_rate": 9.269019856868784e-06,
+      "loss": 0.3984,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7708752515090543,
+      "grad_norm": 0.35338932275772095,
+      "learning_rate": 9.268257915202037e-06,
+      "loss": 0.3992,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7711267605633803,
+      "grad_norm": 0.3768799304962158,
+      "learning_rate": 9.2674956079857e-06,
+      "loss": 0.3858,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7713782696177063,
+      "grad_norm": 0.3796723484992981,
+      "learning_rate": 9.26673293528506e-06,
+      "loss": 0.3858,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7716297786720322,
+      "grad_norm": 0.3658340275287628,
+      "learning_rate": 9.26596989716544e-06,
+      "loss": 0.3844,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7718812877263581,
+      "grad_norm": 0.41284385323524475,
+      "learning_rate": 9.265206493692185e-06,
+      "loss": 0.3697,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7721327967806841,
+      "grad_norm": 0.4010927379131317,
+      "learning_rate": 9.264442724930675e-06,
+      "loss": 0.3509,
+      "step": 3070
+    },
+    {
+      "epoch": 0.77238430583501,
+      "grad_norm": 0.37611401081085205,
+      "learning_rate": 9.263678590946326e-06,
+      "loss": 0.3877,
+      "step": 3071
+    },
+    {
+      "epoch": 0.772635814889336,
+      "grad_norm": 0.4208700656890869,
+      "learning_rate": 9.26291409180458e-06,
+      "loss": 0.3599,
+      "step": 3072
+    },
+    {
+      "epoch": 0.772887323943662,
+      "grad_norm": 0.3623879849910736,
+      "learning_rate": 9.262149227570908e-06,
+      "loss": 0.4031,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7731388329979879,
+      "grad_norm": 0.3639488220214844,
+      "learning_rate": 9.261383998310822e-06,
+      "loss": 0.3963,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7733903420523138,
+      "grad_norm": 0.4167322814464569,
+      "learning_rate": 9.260618404089853e-06,
+      "loss": 0.376,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7736418511066399,
+      "grad_norm": 0.33790600299835205,
+      "learning_rate": 9.259852444973573e-06,
+      "loss": 0.3803,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7738933601609658,
+      "grad_norm": 0.4101136028766632,
+      "learning_rate": 9.25908612102758e-06,
+      "loss": 0.3665,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7741448692152918,
+      "grad_norm": 0.3587418794631958,
+      "learning_rate": 9.258319432317506e-06,
+      "loss": 0.3875,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7743963782696177,
+      "grad_norm": 0.36111775040626526,
+      "learning_rate": 9.257552378909013e-06,
+      "loss": 0.3797,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7746478873239436,
+      "grad_norm": 0.4025454521179199,
+      "learning_rate": 9.256784960867793e-06,
+      "loss": 0.3968,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7748993963782697,
+      "grad_norm": 0.37682798504829407,
+      "learning_rate": 9.256017178259572e-06,
+      "loss": 0.3583,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7751509054325956,
+      "grad_norm": 0.40718135237693787,
+      "learning_rate": 9.255249031150106e-06,
+      "loss": 0.4151,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7754024144869215,
+      "grad_norm": 0.45130354166030884,
+      "learning_rate": 9.25448051960518e-06,
+      "loss": 0.3853,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7756539235412475,
+      "grad_norm": 0.3748539090156555,
+      "learning_rate": 9.253711643690612e-06,
+      "loss": 0.3902,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7759054325955734,
+      "grad_norm": 0.4091399610042572,
+      "learning_rate": 9.252942403472256e-06,
+      "loss": 0.3685,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7761569416498993,
+      "grad_norm": 0.5184744000434875,
+      "learning_rate": 9.252172799015989e-06,
+      "loss": 0.4028,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7764084507042254,
+      "grad_norm": 0.3658330738544464,
+      "learning_rate": 9.251402830387721e-06,
+      "loss": 0.398,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7766599597585513,
+      "grad_norm": 0.4008294641971588,
+      "learning_rate": 9.250632497653398e-06,
+      "loss": 0.392,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7769114688128773,
+      "grad_norm": 0.42256030440330505,
+      "learning_rate": 9.249861800878995e-06,
+      "loss": 0.3954,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7771629778672032,
+      "grad_norm": 0.38519713282585144,
+      "learning_rate": 9.249090740130515e-06,
+      "loss": 0.3897,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7774144869215291,
+      "grad_norm": 0.41861796379089355,
+      "learning_rate": 9.248319315473995e-06,
+      "loss": 0.3863,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7776659959758552,
+      "grad_norm": 0.41915813088417053,
+      "learning_rate": 9.247547526975505e-06,
+      "loss": 0.4048,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7779175050301811,
+      "grad_norm": 0.34398457407951355,
+      "learning_rate": 9.246775374701139e-06,
+      "loss": 0.3496,
+      "step": 3093
+    },
+    {
+      "epoch": 0.778169014084507,
+      "grad_norm": 0.4545997679233551,
+      "learning_rate": 9.246002858717031e-06,
+      "loss": 0.3996,
+      "step": 3094
+    },
+    {
+      "epoch": 0.778420523138833,
+      "grad_norm": 0.34864526987075806,
+      "learning_rate": 9.245229979089341e-06,
+      "loss": 0.3708,
+      "step": 3095
+    },
+    {
+      "epoch": 0.778672032193159,
+      "grad_norm": 0.32429444789886475,
+      "learning_rate": 9.244456735884261e-06,
+      "loss": 0.3773,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7789235412474849,
+      "grad_norm": 0.3772585093975067,
+      "learning_rate": 9.243683129168016e-06,
+      "loss": 0.3615,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7791750503018109,
+      "grad_norm": 0.4083939492702484,
+      "learning_rate": 9.242909159006858e-06,
+      "loss": 0.4064,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7794265593561368,
+      "grad_norm": 0.3335086703300476,
+      "learning_rate": 9.242134825467076e-06,
+      "loss": 0.346,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7796780684104627,
+      "grad_norm": 0.3571106493473053,
+      "learning_rate": 9.241360128614984e-06,
+      "loss": 0.3438,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7799295774647887,
+      "grad_norm": 0.3634457588195801,
+      "learning_rate": 9.24058506851693e-06,
+      "loss": 0.3714,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7801810865191147,
+      "grad_norm": 0.36663031578063965,
+      "learning_rate": 9.239809645239295e-06,
+      "loss": 0.4045,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7804325955734407,
+      "grad_norm": 0.35045841336250305,
+      "learning_rate": 9.239033858848487e-06,
+      "loss": 0.3919,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7806841046277666,
+      "grad_norm": 0.4044874310493469,
+      "learning_rate": 9.238257709410949e-06,
+      "loss": 0.3807,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7809356136820925,
+      "grad_norm": 0.3510379493236542,
+      "learning_rate": 9.237481196993152e-06,
+      "loss": 0.3711,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7811871227364185,
+      "grad_norm": 0.39423373341560364,
+      "learning_rate": 9.2367043216616e-06,
+      "loss": 0.3912,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7814386317907445,
+      "grad_norm": 0.3713517189025879,
+      "learning_rate": 9.23592708348283e-06,
+      "loss": 0.3774,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7816901408450704,
+      "grad_norm": 0.3517390787601471,
+      "learning_rate": 9.235149482523402e-06,
+      "loss": 0.3597,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7819416498993964,
+      "grad_norm": 0.38510939478874207,
+      "learning_rate": 9.234371518849918e-06,
+      "loss": 0.3465,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7821931589537223,
+      "grad_norm": 0.3652108609676361,
+      "learning_rate": 9.233593192529002e-06,
+      "loss": 0.3745,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7824446680080482,
+      "grad_norm": 0.39064309000968933,
+      "learning_rate": 9.232814503627316e-06,
+      "loss": 0.3825,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7826961770623743,
+      "grad_norm": 0.34675848484039307,
+      "learning_rate": 9.232035452211546e-06,
+      "loss": 0.3928,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7829476861167002,
+      "grad_norm": 0.36757782101631165,
+      "learning_rate": 9.231256038348418e-06,
+      "loss": 0.3966,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7831991951710262,
+      "grad_norm": 0.36838653683662415,
+      "learning_rate": 9.230476262104678e-06,
+      "loss": 0.3835,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7834507042253521,
+      "grad_norm": 0.33910053968429565,
+      "learning_rate": 9.229696123547114e-06,
+      "loss": 0.4,
+      "step": 3115
+    },
+    {
+      "epoch": 0.783702213279678,
+      "grad_norm": 0.3448108434677124,
+      "learning_rate": 9.228915622742536e-06,
+      "loss": 0.3819,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7839537223340041,
+      "grad_norm": 0.3653436303138733,
+      "learning_rate": 9.228134759757791e-06,
+      "loss": 0.3806,
+      "step": 3117
+    },
+    {
+      "epoch": 0.78420523138833,
+      "grad_norm": 0.33516430854797363,
+      "learning_rate": 9.227353534659758e-06,
+      "loss": 0.3677,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7844567404426559,
+      "grad_norm": 0.36953532695770264,
+      "learning_rate": 9.226571947515339e-06,
+      "loss": 0.3824,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7847082494969819,
+      "grad_norm": 0.37254080176353455,
+      "learning_rate": 9.225789998391473e-06,
+      "loss": 0.3841,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7849597585513078,
+      "grad_norm": 0.34900060296058655,
+      "learning_rate": 9.225007687355132e-06,
+      "loss": 0.3699,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7852112676056338,
+      "grad_norm": 0.42499294877052307,
+      "learning_rate": 9.224225014473312e-06,
+      "loss": 0.377,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7854627766599598,
+      "grad_norm": 0.4103010296821594,
+      "learning_rate": 9.223441979813049e-06,
+      "loss": 0.3935,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.34033411741256714,
+      "learning_rate": 9.222658583441399e-06,
+      "loss": 0.3798,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7859657947686117,
+      "grad_norm": 0.45133984088897705,
+      "learning_rate": 9.221874825425461e-06,
+      "loss": 0.3996,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7862173038229376,
+      "grad_norm": 0.3784712255001068,
+      "learning_rate": 9.221090705832353e-06,
+      "loss": 0.3937,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7864688128772636,
+      "grad_norm": 0.36274659633636475,
+      "learning_rate": 9.220306224729237e-06,
+      "loss": 0.3899,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7867203219315896,
+      "grad_norm": 0.36185523867607117,
+      "learning_rate": 9.219521382183291e-06,
+      "loss": 0.38,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7869718309859155,
+      "grad_norm": 0.3780527710914612,
+      "learning_rate": 9.21873617826174e-06,
+      "loss": 0.3743,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7872233400402414,
+      "grad_norm": 0.3563750088214874,
+      "learning_rate": 9.217950613031826e-06,
+      "loss": 0.3779,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7874748490945674,
+      "grad_norm": 0.33744367957115173,
+      "learning_rate": 9.21716468656083e-06,
+      "loss": 0.3847,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7877263581488934,
+      "grad_norm": 0.4224430322647095,
+      "learning_rate": 9.216378398916059e-06,
+      "loss": 0.4001,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7879778672032193,
+      "grad_norm": 0.3875563144683838,
+      "learning_rate": 9.215591750164856e-06,
+      "loss": 0.3813,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7882293762575453,
+      "grad_norm": 0.41393423080444336,
+      "learning_rate": 9.214804740374594e-06,
+      "loss": 0.3688,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7884808853118712,
+      "grad_norm": 0.39912310242652893,
+      "learning_rate": 9.214017369612672e-06,
+      "loss": 0.3786,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 0.3480337858200073,
+      "learning_rate": 9.213229637946526e-06,
+      "loss": 0.3602,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7889839034205232,
+      "grad_norm": 0.3411673605442047,
+      "learning_rate": 9.21244154544362e-06,
+      "loss": 0.3987,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7892354124748491,
+      "grad_norm": 0.3513990342617035,
+      "learning_rate": 9.211653092171447e-06,
+      "loss": 0.3716,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7894869215291751,
+      "grad_norm": 0.3667123317718506,
+      "learning_rate": 9.210864278197536e-06,
+      "loss": 0.3672,
+      "step": 3139
+    },
+    {
+      "epoch": 0.789738430583501,
+      "grad_norm": 0.3660683333873749,
+      "learning_rate": 9.210075103589443e-06,
+      "loss": 0.4099,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7899899396378269,
+      "grad_norm": 0.3497997224330902,
+      "learning_rate": 9.209285568414755e-06,
+      "loss": 0.3661,
+      "step": 3141
+    },
+    {
+      "epoch": 0.790241448692153,
+      "grad_norm": 0.3466975688934326,
+      "learning_rate": 9.20849567274109e-06,
+      "loss": 0.3909,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7904929577464789,
+      "grad_norm": 0.3633776009082794,
+      "learning_rate": 9.2077054166361e-06,
+      "loss": 0.3701,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7907444668008048,
+      "grad_norm": 0.36884036660194397,
+      "learning_rate": 9.206914800167463e-06,
+      "loss": 0.3917,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7909959758551308,
+      "grad_norm": 0.34670379757881165,
+      "learning_rate": 9.206123823402894e-06,
+      "loss": 0.4399,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7912474849094567,
+      "grad_norm": 0.3583228588104248,
+      "learning_rate": 9.205332486410133e-06,
+      "loss": 0.4055,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7914989939637826,
+      "grad_norm": 0.38067370653152466,
+      "learning_rate": 9.204540789256951e-06,
+      "loss": 0.4284,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7917505030181087,
+      "grad_norm": 0.3454825282096863,
+      "learning_rate": 9.203748732011154e-06,
+      "loss": 0.4033,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7920020120724346,
+      "grad_norm": 0.3998292088508606,
+      "learning_rate": 9.202956314740578e-06,
+      "loss": 0.3992,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7922535211267606,
+      "grad_norm": 0.3713396489620209,
+      "learning_rate": 9.202163537513088e-06,
+      "loss": 0.3693,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7925050301810865,
+      "grad_norm": 0.393930047750473,
+      "learning_rate": 9.201370400396578e-06,
+      "loss": 0.3837,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7927565392354124,
+      "grad_norm": 0.388494610786438,
+      "learning_rate": 9.200576903458978e-06,
+      "loss": 0.3694,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7930080482897385,
+      "grad_norm": 0.4377411901950836,
+      "learning_rate": 9.199783046768245e-06,
+      "loss": 0.4008,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7932595573440644,
+      "grad_norm": 0.3396592438220978,
+      "learning_rate": 9.198988830392365e-06,
+      "loss": 0.3754,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7935110663983903,
+      "grad_norm": 0.3519555926322937,
+      "learning_rate": 9.198194254399364e-06,
+      "loss": 0.3816,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7937625754527163,
+      "grad_norm": 0.368377149105072,
+      "learning_rate": 9.197399318857288e-06,
+      "loss": 0.371,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7940140845070423,
+      "grad_norm": 0.3477652370929718,
+      "learning_rate": 9.19660402383422e-06,
+      "loss": 0.3841,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7942655935613682,
+      "grad_norm": 0.37322860956192017,
+      "learning_rate": 9.19580836939827e-06,
+      "loss": 0.3886,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7945171026156942,
+      "grad_norm": 0.39691033959388733,
+      "learning_rate": 9.195012355617581e-06,
+      "loss": 0.373,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7947686116700201,
+      "grad_norm": 0.355160117149353,
+      "learning_rate": 9.194215982560328e-06,
+      "loss": 0.3933,
+      "step": 3160
+    },
+    {
+      "epoch": 0.795020120724346,
+      "grad_norm": 0.36236023902893066,
+      "learning_rate": 9.193419250294717e-06,
+      "loss": 0.4166,
+      "step": 3161
+    },
+    {
+      "epoch": 0.795271629778672,
+      "grad_norm": 0.35770562291145325,
+      "learning_rate": 9.192622158888979e-06,
+      "loss": 0.3911,
+      "step": 3162
+    },
+    {
+      "epoch": 0.795523138832998,
+      "grad_norm": 0.36751723289489746,
+      "learning_rate": 9.191824708411384e-06,
+      "loss": 0.3973,
+      "step": 3163
+    },
+    {
+      "epoch": 0.795774647887324,
+      "grad_norm": 0.365925133228302,
+      "learning_rate": 9.191026898930224e-06,
+      "loss": 0.3763,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7960261569416499,
+      "grad_norm": 0.3750317394733429,
+      "learning_rate": 9.190228730513832e-06,
+      "loss": 0.3783,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7962776659959758,
+      "grad_norm": 0.343177855014801,
+      "learning_rate": 9.189430203230562e-06,
+      "loss": 0.3744,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7965291750503019,
+      "grad_norm": 0.3460996150970459,
+      "learning_rate": 9.188631317148804e-06,
+      "loss": 0.4047,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7967806841046278,
+      "grad_norm": 0.38344576954841614,
+      "learning_rate": 9.187832072336977e-06,
+      "loss": 0.3893,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7970321931589537,
+      "grad_norm": 0.36154693365097046,
+      "learning_rate": 9.187032468863532e-06,
+      "loss": 0.3512,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7972837022132797,
+      "grad_norm": 0.38037383556365967,
+      "learning_rate": 9.186232506796952e-06,
+      "loss": 0.382,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7975352112676056,
+      "grad_norm": 0.33756163716316223,
+      "learning_rate": 9.185432186205744e-06,
+      "loss": 0.3945,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7977867203219315,
+      "grad_norm": 0.394203245639801,
+      "learning_rate": 9.184631507158456e-06,
+      "loss": 0.3817,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7980382293762576,
+      "grad_norm": 0.37978580594062805,
+      "learning_rate": 9.183830469723658e-06,
+      "loss": 0.3803,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7982897384305835,
+      "grad_norm": 0.36637356877326965,
+      "learning_rate": 9.183029073969953e-06,
+      "loss": 0.3997,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7985412474849095,
+      "grad_norm": 0.40284034609794617,
+      "learning_rate": 9.182227319965978e-06,
+      "loss": 0.3895,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7987927565392354,
+      "grad_norm": 0.34037232398986816,
+      "learning_rate": 9.181425207780396e-06,
+      "loss": 0.3971,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7990442655935613,
+      "grad_norm": 0.3634487986564636,
+      "learning_rate": 9.180622737481904e-06,
+      "loss": 0.3883,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7992957746478874,
+      "grad_norm": 0.3719540238380432,
+      "learning_rate": 9.179819909139228e-06,
+      "loss": 0.3946,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7995472837022133,
+      "grad_norm": 0.34387439489364624,
+      "learning_rate": 9.179016722821126e-06,
+      "loss": 0.3701,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7997987927565392,
+      "grad_norm": 0.41140124201774597,
+      "learning_rate": 9.178213178596386e-06,
+      "loss": 0.3822,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8000503018108652,
+      "grad_norm": 0.3771357238292694,
+      "learning_rate": 9.177409276533825e-06,
+      "loss": 0.4017,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8003018108651911,
+      "grad_norm": 0.37716689705848694,
+      "learning_rate": 9.176605016702294e-06,
+      "loss": 0.3574,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8005533199195171,
+      "grad_norm": 0.357952356338501,
+      "learning_rate": 9.175800399170673e-06,
+      "loss": 0.3653,
+      "step": 3183
+    },
+    {
+      "epoch": 0.8008048289738431,
+      "grad_norm": 0.3895781636238098,
+      "learning_rate": 9.17499542400787e-06,
+      "loss": 0.3879,
+      "step": 3184
+    },
+    {
+      "epoch": 0.801056338028169,
+      "grad_norm": 0.37393030524253845,
+      "learning_rate": 9.174190091282828e-06,
+      "loss": 0.3834,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8013078470824949,
+      "grad_norm": 0.4014933407306671,
+      "learning_rate": 9.173384401064519e-06,
+      "loss": 0.3581,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8015593561368209,
+      "grad_norm": 0.39869409799575806,
+      "learning_rate": 9.172578353421943e-06,
+      "loss": 0.3896,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8018108651911469,
+      "grad_norm": 0.45484480261802673,
+      "learning_rate": 9.171771948424138e-06,
+      "loss": 0.3831,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8020623742454729,
+      "grad_norm": 0.4879845380783081,
+      "learning_rate": 9.17096518614016e-06,
+      "loss": 0.3998,
+      "step": 3189
+    },
+    {
+      "epoch": 0.8023138832997988,
+      "grad_norm": 0.37346404790878296,
+      "learning_rate": 9.17015806663911e-06,
+      "loss": 0.3768,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8025653923541247,
+      "grad_norm": 0.38394322991371155,
+      "learning_rate": 9.169350589990109e-06,
+      "loss": 0.37,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8028169014084507,
+      "grad_norm": 0.42078500986099243,
+      "learning_rate": 9.168542756262313e-06,
+      "loss": 0.373,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8030684104627767,
+      "grad_norm": 0.37299197912216187,
+      "learning_rate": 9.16773456552491e-06,
+      "loss": 0.3727,
+      "step": 3193
+    },
+    {
+      "epoch": 0.8033199195171026,
+      "grad_norm": 0.36001479625701904,
+      "learning_rate": 9.166926017847113e-06,
+      "loss": 0.3708,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8035714285714286,
+      "grad_norm": 0.4601292014122009,
+      "learning_rate": 9.166117113298172e-06,
+      "loss": 0.3842,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8038229376257545,
+      "grad_norm": 0.3911070227622986,
+      "learning_rate": 9.165307851947362e-06,
+      "loss": 0.3644,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8040744466800804,
+      "grad_norm": 0.3885655701160431,
+      "learning_rate": 9.164498233863994e-06,
+      "loss": 0.3674,
+      "step": 3197
+    },
+    {
+      "epoch": 0.8043259557344065,
+      "grad_norm": 0.46075260639190674,
+      "learning_rate": 9.163688259117405e-06,
+      "loss": 0.3889,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8045774647887324,
+      "grad_norm": 0.4196132719516754,
+      "learning_rate": 9.162877927776963e-06,
+      "loss": 0.428,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8048289738430584,
+      "grad_norm": 0.38232311606407166,
+      "learning_rate": 9.162067239912072e-06,
+      "loss": 0.4044,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8050804828973843,
+      "grad_norm": 0.39964160323143005,
+      "learning_rate": 9.161256195592157e-06,
+      "loss": 0.3878,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8053319919517102,
+      "grad_norm": 0.3779206871986389,
+      "learning_rate": 9.160444794886682e-06,
+      "loss": 0.3822,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8055835010060363,
+      "grad_norm": 0.3651765286922455,
+      "learning_rate": 9.15963303786514e-06,
+      "loss": 0.3952,
+      "step": 3203
+    },
+    {
+      "epoch": 0.8058350100603622,
+      "grad_norm": 0.3684598207473755,
+      "learning_rate": 9.15882092459705e-06,
+      "loss": 0.3693,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8060865191146881,
+      "grad_norm": 0.35791802406311035,
+      "learning_rate": 9.158008455151965e-06,
+      "loss": 0.4058,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8063380281690141,
+      "grad_norm": 0.3659406900405884,
+      "learning_rate": 9.157195629599468e-06,
+      "loss": 0.3879,
+      "step": 3206
+    },
+    {
+      "epoch": 0.80658953722334,
+      "grad_norm": 0.3828873932361603,
+      "learning_rate": 9.156382448009173e-06,
+      "loss": 0.3876,
+      "step": 3207
+    },
+    {
+      "epoch": 0.806841046277666,
+      "grad_norm": 0.33085519075393677,
+      "learning_rate": 9.155568910450722e-06,
+      "loss": 0.3745,
+      "step": 3208
+    },
+    {
+      "epoch": 0.807092555331992,
+      "grad_norm": 0.35317882895469666,
+      "learning_rate": 9.154755016993794e-06,
+      "loss": 0.3587,
+      "step": 3209
+    },
+    {
+      "epoch": 0.8073440643863179,
+      "grad_norm": 0.37781715393066406,
+      "learning_rate": 9.15394076770809e-06,
+      "loss": 0.4005,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8075955734406438,
+      "grad_norm": 0.33460843563079834,
+      "learning_rate": 9.153126162663343e-06,
+      "loss": 0.3969,
+      "step": 3211
+    },
+    {
+      "epoch": 0.8078470824949698,
+      "grad_norm": 0.4079720377922058,
+      "learning_rate": 9.152311201929326e-06,
+      "loss": 0.3821,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8080985915492958,
+      "grad_norm": 0.3783186972141266,
+      "learning_rate": 9.15149588557583e-06,
+      "loss": 0.3985,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8083501006036218,
+      "grad_norm": 0.33954814076423645,
+      "learning_rate": 9.150680213672683e-06,
+      "loss": 0.4062,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8086016096579477,
+      "grad_norm": 0.4284415543079376,
+      "learning_rate": 9.149864186289743e-06,
+      "loss": 0.3759,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8088531187122736,
+      "grad_norm": 0.349122554063797,
+      "learning_rate": 9.149047803496896e-06,
+      "loss": 0.3752,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8091046277665996,
+      "grad_norm": 0.3756011426448822,
+      "learning_rate": 9.148231065364062e-06,
+      "loss": 0.3826,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8093561368209256,
+      "grad_norm": 0.42645543813705444,
+      "learning_rate": 9.147413971961187e-06,
+      "loss": 0.404,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8096076458752515,
+      "grad_norm": 0.35553669929504395,
+      "learning_rate": 9.146596523358252e-06,
+      "loss": 0.3921,
+      "step": 3219
+    },
+    {
+      "epoch": 0.8098591549295775,
+      "grad_norm": 0.38775989413261414,
+      "learning_rate": 9.145778719625266e-06,
+      "loss": 0.3986,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8101106639839034,
+      "grad_norm": 0.3600097596645355,
+      "learning_rate": 9.144960560832268e-06,
+      "loss": 0.385,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8103621730382293,
+      "grad_norm": 0.3468279540538788,
+      "learning_rate": 9.144142047049329e-06,
+      "loss": 0.3786,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8106136820925554,
+      "grad_norm": 0.3529747426509857,
+      "learning_rate": 9.14332317834655e-06,
+      "loss": 0.3729,
+      "step": 3223
+    },
+    {
+      "epoch": 0.8108651911468813,
+      "grad_norm": 0.31644824147224426,
+      "learning_rate": 9.14250395479406e-06,
+      "loss": 0.3905,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8111167002012073,
+      "grad_norm": 0.3824257254600525,
+      "learning_rate": 9.141684376462024e-06,
+      "loss": 0.389,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8113682092555332,
+      "grad_norm": 0.3192039132118225,
+      "learning_rate": 9.140864443420629e-06,
+      "loss": 0.3734,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8116197183098591,
+      "grad_norm": 0.3341809809207916,
+      "learning_rate": 9.140044155740102e-06,
+      "loss": 0.383,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8118712273641852,
+      "grad_norm": 0.33594390749931335,
+      "learning_rate": 9.139223513490692e-06,
+      "loss": 0.3706,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8121227364185111,
+      "grad_norm": 0.34297066926956177,
+      "learning_rate": 9.138402516742681e-06,
+      "loss": 0.3923,
+      "step": 3229
+    },
+    {
+      "epoch": 0.812374245472837,
+      "grad_norm": 0.32909277081489563,
+      "learning_rate": 9.137581165566388e-06,
+      "loss": 0.3715,
+      "step": 3230
+    },
+    {
+      "epoch": 0.812625754527163,
+      "grad_norm": 0.35775429010391235,
+      "learning_rate": 9.13675946003215e-06,
+      "loss": 0.3597,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8128772635814889,
+      "grad_norm": 0.3310528099536896,
+      "learning_rate": 9.135937400210345e-06,
+      "loss": 0.3929,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8131287726358148,
+      "grad_norm": 0.35094767808914185,
+      "learning_rate": 9.135114986171373e-06,
+      "loss": 0.4054,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8133802816901409,
+      "grad_norm": 0.38536086678504944,
+      "learning_rate": 9.134292217985675e-06,
+      "loss": 0.4061,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8136317907444668,
+      "grad_norm": 0.35083386301994324,
+      "learning_rate": 9.133469095723712e-06,
+      "loss": 0.3607,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8138832997987927,
+      "grad_norm": 0.34941115975379944,
+      "learning_rate": 9.13264561945598e-06,
+      "loss": 0.3522,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8141348088531187,
+      "grad_norm": 0.3351636826992035,
+      "learning_rate": 9.131821789253003e-06,
+      "loss": 0.3788,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8143863179074446,
+      "grad_norm": 0.3241204023361206,
+      "learning_rate": 9.130997605185338e-06,
+      "loss": 0.3929,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8146378269617707,
+      "grad_norm": 0.35494330525398254,
+      "learning_rate": 9.130173067323575e-06,
+      "loss": 0.4019,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8148893360160966,
+      "grad_norm": 0.3341788351535797,
+      "learning_rate": 9.129348175738324e-06,
+      "loss": 0.3821,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8151408450704225,
+      "grad_norm": 0.34449446201324463,
+      "learning_rate": 9.128522930500237e-06,
+      "loss": 0.3671,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8153923541247485,
+      "grad_norm": 0.31433650851249695,
+      "learning_rate": 9.127697331679988e-06,
+      "loss": 0.3803,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8156438631790744,
+      "grad_norm": 0.3512547016143799,
+      "learning_rate": 9.126871379348284e-06,
+      "loss": 0.3744,
+      "step": 3243
+    },
+    {
+      "epoch": 0.8158953722334004,
+      "grad_norm": 0.3199286162853241,
+      "learning_rate": 9.126045073575865e-06,
+      "loss": 0.4147,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8161468812877264,
+      "grad_norm": 0.37143221497535706,
+      "learning_rate": 9.125218414433498e-06,
+      "loss": 0.3738,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8163983903420523,
+      "grad_norm": 0.3939398229122162,
+      "learning_rate": 9.124391401991981e-06,
+      "loss": 0.3829,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8166498993963782,
+      "grad_norm": 0.34580913186073303,
+      "learning_rate": 9.123564036322143e-06,
+      "loss": 0.3956,
+      "step": 3247
+    },
+    {
+      "epoch": 0.8169014084507042,
+      "grad_norm": 0.32879477739334106,
+      "learning_rate": 9.122736317494842e-06,
+      "loss": 0.3746,
+      "step": 3248
+    },
+    {
+      "epoch": 0.8171529175050302,
+      "grad_norm": 0.375439316034317,
+      "learning_rate": 9.121908245580967e-06,
+      "loss": 0.3988,
+      "step": 3249
+    },
+    {
+      "epoch": 0.8174044265593562,
+      "grad_norm": 0.3741564452648163,
+      "learning_rate": 9.121079820651438e-06,
+      "loss": 0.4051,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8176559356136821,
+      "grad_norm": 0.3896329700946808,
+      "learning_rate": 9.120251042777203e-06,
+      "loss": 0.38,
+      "step": 3251
+    },
+    {
+      "epoch": 0.817907444668008,
+      "grad_norm": 0.38490694761276245,
+      "learning_rate": 9.119421912029243e-06,
+      "loss": 0.352,
+      "step": 3252
+    },
+    {
+      "epoch": 0.818158953722334,
+      "grad_norm": 0.3421596884727478,
+      "learning_rate": 9.118592428478565e-06,
+      "loss": 0.3867,
+      "step": 3253
+    },
+    {
+      "epoch": 0.81841046277666,
+      "grad_norm": 0.36744779348373413,
+      "learning_rate": 9.117762592196214e-06,
+      "loss": 0.3598,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8186619718309859,
+      "grad_norm": 0.33967792987823486,
+      "learning_rate": 9.116932403253257e-06,
+      "loss": 0.4086,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8189134808853119,
+      "grad_norm": 0.35060232877731323,
+      "learning_rate": 9.116101861720793e-06,
+      "loss": 0.3724,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8191649899396378,
+      "grad_norm": 0.3623863458633423,
+      "learning_rate": 9.115270967669958e-06,
+      "loss": 0.3924,
+      "step": 3257
+    },
+    {
+      "epoch": 0.8194164989939637,
+      "grad_norm": 0.3579019606113434,
+      "learning_rate": 9.114439721171909e-06,
+      "loss": 0.4078,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8196680080482898,
+      "grad_norm": 0.3488744795322418,
+      "learning_rate": 9.113608122297836e-06,
+      "loss": 0.4064,
+      "step": 3259
+    },
+    {
+      "epoch": 0.8199195171026157,
+      "grad_norm": 0.38803979754447937,
+      "learning_rate": 9.112776171118964e-06,
+      "loss": 0.4151,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8201710261569416,
+      "grad_norm": 0.3529983162879944,
+      "learning_rate": 9.11194386770654e-06,
+      "loss": 0.3722,
+      "step": 3261
+    },
+    {
+      "epoch": 0.8204225352112676,
+      "grad_norm": 0.39066073298454285,
+      "learning_rate": 9.111111212131851e-06,
+      "loss": 0.4216,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8206740442655935,
+      "grad_norm": 0.34698736667633057,
+      "learning_rate": 9.110278204466203e-06,
+      "loss": 0.3725,
+      "step": 3263
+    },
+    {
+      "epoch": 0.8209255533199196,
+      "grad_norm": 0.337932288646698,
+      "learning_rate": 9.109444844780942e-06,
+      "loss": 0.4123,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8211770623742455,
+      "grad_norm": 0.3363993465900421,
+      "learning_rate": 9.108611133147438e-06,
+      "loss": 0.3716,
+      "step": 3265
+    },
+    {
+      "epoch": 0.8214285714285714,
+      "grad_norm": 0.3734219968318939,
+      "learning_rate": 9.107777069637094e-06,
+      "loss": 0.3874,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8216800804828974,
+      "grad_norm": 0.32347917556762695,
+      "learning_rate": 9.106942654321343e-06,
+      "loss": 0.358,
+      "step": 3267
+    },
+    {
+      "epoch": 0.8219315895372233,
+      "grad_norm": 0.3709772229194641,
+      "learning_rate": 9.106107887271647e-06,
+      "loss": 0.3634,
+      "step": 3268
+    },
+    {
+      "epoch": 0.8221830985915493,
+      "grad_norm": 0.3945581018924713,
+      "learning_rate": 9.105272768559496e-06,
+      "loss": 0.379,
+      "step": 3269
+    },
+    {
+      "epoch": 0.8224346076458753,
+      "grad_norm": 0.33245036005973816,
+      "learning_rate": 9.104437298256416e-06,
+      "loss": 0.3605,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8226861167002012,
+      "grad_norm": 0.37672650814056396,
+      "learning_rate": 9.103601476433959e-06,
+      "loss": 0.389,
+      "step": 3271
+    },
+    {
+      "epoch": 0.8229376257545271,
+      "grad_norm": 0.37281307578086853,
+      "learning_rate": 9.102765303163708e-06,
+      "loss": 0.3872,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8231891348088531,
+      "grad_norm": 0.3488701581954956,
+      "learning_rate": 9.101928778517275e-06,
+      "loss": 0.3543,
+      "step": 3273
+    },
+    {
+      "epoch": 0.8234406438631791,
+      "grad_norm": 0.33811309933662415,
+      "learning_rate": 9.101091902566303e-06,
+      "loss": 0.3833,
+      "step": 3274
+    },
+    {
+      "epoch": 0.8236921529175051,
+      "grad_norm": 0.40357568860054016,
+      "learning_rate": 9.100254675382467e-06,
+      "loss": 0.3806,
+      "step": 3275
+    },
+    {
+      "epoch": 0.823943661971831,
+      "grad_norm": 0.39533787965774536,
+      "learning_rate": 9.099417097037468e-06,
+      "loss": 0.3767,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8241951710261569,
+      "grad_norm": 0.38113778829574585,
+      "learning_rate": 9.098579167603042e-06,
+      "loss": 0.3862,
+      "step": 3277
+    },
+    {
+      "epoch": 0.8244466800804829,
+      "grad_norm": 0.38164108991622925,
+      "learning_rate": 9.09774088715095e-06,
+      "loss": 0.3707,
+      "step": 3278
+    },
+    {
+      "epoch": 0.8246981891348089,
+      "grad_norm": 0.4421541094779968,
+      "learning_rate": 9.096902255752986e-06,
+      "loss": 0.3845,
+      "step": 3279
+    },
+    {
+      "epoch": 0.8249496981891348,
+      "grad_norm": 0.34088340401649475,
+      "learning_rate": 9.096063273480975e-06,
+      "loss": 0.3685,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8252012072434608,
+      "grad_norm": 0.44785863161087036,
+      "learning_rate": 9.09522394040677e-06,
+      "loss": 0.3757,
+      "step": 3281
+    },
+    {
+      "epoch": 0.8254527162977867,
+      "grad_norm": 0.4285958409309387,
+      "learning_rate": 9.094384256602252e-06,
+      "loss": 0.3549,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8257042253521126,
+      "grad_norm": 0.35675740242004395,
+      "learning_rate": 9.093544222139338e-06,
+      "loss": 0.364,
+      "step": 3283
+    },
+    {
+      "epoch": 0.8259557344064387,
+      "grad_norm": 0.3988295793533325,
+      "learning_rate": 9.09270383708997e-06,
+      "loss": 0.4011,
+      "step": 3284
+    },
+    {
+      "epoch": 0.8262072434607646,
+      "grad_norm": 0.4526250958442688,
+      "learning_rate": 9.091863101526124e-06,
+      "loss": 0.3882,
+      "step": 3285
+    },
+    {
+      "epoch": 0.8264587525150905,
+      "grad_norm": 0.416526734828949,
+      "learning_rate": 9.091022015519798e-06,
+      "loss": 0.4004,
+      "step": 3286
+    },
+    {
+      "epoch": 0.8267102615694165,
+      "grad_norm": 0.3291431665420532,
+      "learning_rate": 9.090180579143033e-06,
+      "loss": 0.3499,
+      "step": 3287
+    },
+    {
+      "epoch": 0.8269617706237424,
+      "grad_norm": 0.34957194328308105,
+      "learning_rate": 9.08933879246789e-06,
+      "loss": 0.36,
+      "step": 3288
+    },
+    {
+      "epoch": 0.8272132796780685,
+      "grad_norm": 0.35774147510528564,
+      "learning_rate": 9.08849665556646e-06,
+      "loss": 0.3683,
+      "step": 3289
+    },
+    {
+      "epoch": 0.8274647887323944,
+      "grad_norm": 0.34094980359077454,
+      "learning_rate": 9.087654168510871e-06,
+      "loss": 0.3622,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8277162977867203,
+      "grad_norm": 0.3316769599914551,
+      "learning_rate": 9.086811331373273e-06,
+      "loss": 0.3855,
+      "step": 3291
+    },
+    {
+      "epoch": 0.8279678068410463,
+      "grad_norm": 0.35548505187034607,
+      "learning_rate": 9.085968144225853e-06,
+      "loss": 0.3918,
+      "step": 3292
+    },
+    {
+      "epoch": 0.8282193158953722,
+      "grad_norm": 0.38808438181877136,
+      "learning_rate": 9.085124607140822e-06,
+      "loss": 0.3791,
+      "step": 3293
+    },
+    {
+      "epoch": 0.8284708249496981,
+      "grad_norm": 0.37164920568466187,
+      "learning_rate": 9.084280720190426e-06,
+      "loss": 0.4012,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8287223340040242,
+      "grad_norm": 0.35741862654685974,
+      "learning_rate": 9.083436483446937e-06,
+      "loss": 0.3659,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8289738430583501,
+      "grad_norm": 0.3638781011104584,
+      "learning_rate": 9.082591896982658e-06,
+      "loss": 0.3967,
+      "step": 3296
+    },
+    {
+      "epoch": 0.829225352112676,
+      "grad_norm": 0.42987996339797974,
+      "learning_rate": 9.081746960869926e-06,
+      "loss": 0.418,
+      "step": 3297
+    },
+    {
+      "epoch": 0.829476861167002,
+      "grad_norm": 0.36449959874153137,
+      "learning_rate": 9.0809016751811e-06,
+      "loss": 0.3814,
+      "step": 3298
+    },
+    {
+      "epoch": 0.829728370221328,
+      "grad_norm": 0.37069931626319885,
+      "learning_rate": 9.080056039988576e-06,
+      "loss": 0.3668,
+      "step": 3299
+    },
+    {
+      "epoch": 0.829979879275654,
+      "grad_norm": 0.3759606182575226,
+      "learning_rate": 9.079210055364777e-06,
+      "loss": 0.3879,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8302313883299799,
+      "grad_norm": 0.3514502942562103,
+      "learning_rate": 9.078363721382157e-06,
+      "loss": 0.3783,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8304828973843058,
+      "grad_norm": 0.3794296979904175,
+      "learning_rate": 9.077517038113197e-06,
+      "loss": 0.3802,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8307344064386318,
+      "grad_norm": 0.3764783442020416,
+      "learning_rate": 9.076670005630413e-06,
+      "loss": 0.3802,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8309859154929577,
+      "grad_norm": 0.31016576290130615,
+      "learning_rate": 9.075822624006345e-06,
+      "loss": 0.3664,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8312374245472837,
+      "grad_norm": 0.3612141013145447,
+      "learning_rate": 9.074974893313571e-06,
+      "loss": 0.3909,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8314889336016097,
+      "grad_norm": 0.34043174982070923,
+      "learning_rate": 9.074126813624687e-06,
+      "loss": 0.3947,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8317404426559356,
+      "grad_norm": 0.3125540614128113,
+      "learning_rate": 9.07327838501233e-06,
+      "loss": 0.3821,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8319919517102615,
+      "grad_norm": 0.34392425417900085,
+      "learning_rate": 9.072429607549161e-06,
+      "loss": 0.3785,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8322434607645876,
+      "grad_norm": 0.3548860251903534,
+      "learning_rate": 9.071580481307875e-06,
+      "loss": 0.394,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8324949698189135,
+      "grad_norm": 0.34213536977767944,
+      "learning_rate": 9.070731006361191e-06,
+      "loss": 0.3685,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8327464788732394,
+      "grad_norm": 0.40426772832870483,
+      "learning_rate": 9.069881182781864e-06,
+      "loss": 0.3613,
+      "step": 3311
+    },
+    {
+      "epoch": 0.8329979879275654,
+      "grad_norm": 0.34350600838661194,
+      "learning_rate": 9.069031010642673e-06,
+      "loss": 0.3672,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8332494969818913,
+      "grad_norm": 0.3403952717781067,
+      "learning_rate": 9.068180490016432e-06,
+      "loss": 0.3847,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8335010060362174,
+      "grad_norm": 0.4108610153198242,
+      "learning_rate": 9.067329620975983e-06,
+      "loss": 0.4024,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8337525150905433,
+      "grad_norm": 0.38602137565612793,
+      "learning_rate": 9.066478403594196e-06,
+      "loss": 0.3723,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8340040241448692,
+      "grad_norm": 0.36774957180023193,
+      "learning_rate": 9.065626837943977e-06,
+      "loss": 0.3747,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8342555331991952,
+      "grad_norm": 0.39249497652053833,
+      "learning_rate": 9.06477492409825e-06,
+      "loss": 0.3835,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8345070422535211,
+      "grad_norm": 0.3368522524833679,
+      "learning_rate": 9.063922662129981e-06,
+      "loss": 0.4076,
+      "step": 3318
+    },
+    {
+      "epoch": 0.834758551307847,
+      "grad_norm": 0.37408214807510376,
+      "learning_rate": 9.063070052112161e-06,
+      "loss": 0.398,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8350100603621731,
+      "grad_norm": 0.35431167483329773,
+      "learning_rate": 9.06221709411781e-06,
+      "loss": 0.3663,
+      "step": 3320
+    },
+    {
+      "epoch": 0.835261569416499,
+      "grad_norm": 0.3219449818134308,
+      "learning_rate": 9.061363788219975e-06,
+      "loss": 0.3794,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8355130784708249,
+      "grad_norm": 0.36822405457496643,
+      "learning_rate": 9.060510134491742e-06,
+      "loss": 0.3678,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8357645875251509,
+      "grad_norm": 0.32559409737586975,
+      "learning_rate": 9.059656133006216e-06,
+      "loss": 0.3461,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8360160965794768,
+      "grad_norm": 0.3630576431751251,
+      "learning_rate": 9.058801783836542e-06,
+      "loss": 0.3485,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8362676056338029,
+      "grad_norm": 0.3687560260295868,
+      "learning_rate": 9.057947087055885e-06,
+      "loss": 0.3806,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8365191146881288,
+      "grad_norm": 0.3402329981327057,
+      "learning_rate": 9.057092042737447e-06,
+      "loss": 0.3662,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8367706237424547,
+      "grad_norm": 0.34224575757980347,
+      "learning_rate": 9.056236650954457e-06,
+      "loss": 0.3971,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8370221327967807,
+      "grad_norm": 0.3538998067378998,
+      "learning_rate": 9.055380911780175e-06,
+      "loss": 0.3753,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8372736418511066,
+      "grad_norm": 0.3665867745876312,
+      "learning_rate": 9.054524825287885e-06,
+      "loss": 0.3894,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8375251509054326,
+      "grad_norm": 0.3492240309715271,
+      "learning_rate": 9.053668391550912e-06,
+      "loss": 0.385,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8377766599597586,
+      "grad_norm": 0.3423280417919159,
+      "learning_rate": 9.052811610642599e-06,
+      "loss": 0.3453,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8380281690140845,
+      "grad_norm": 0.4070049226284027,
+      "learning_rate": 9.051954482636327e-06,
+      "loss": 0.3987,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8382796780684104,
+      "grad_norm": 0.3565402328968048,
+      "learning_rate": 9.051097007605501e-06,
+      "loss": 0.3752,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8385311871227364,
+      "grad_norm": 0.31762558221817017,
+      "learning_rate": 9.050239185623562e-06,
+      "loss": 0.356,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8387826961770624,
+      "grad_norm": 0.36530783772468567,
+      "learning_rate": 9.049381016763973e-06,
+      "loss": 0.4123,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8390342052313883,
+      "grad_norm": 0.33472946286201477,
+      "learning_rate": 9.048522501100233e-06,
+      "loss": 0.3735,
+      "step": 3336
+    },
+    {
+      "epoch": 0.8392857142857143,
+      "grad_norm": 0.3728182315826416,
+      "learning_rate": 9.047663638705868e-06,
+      "loss": 0.3716,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8395372233400402,
+      "grad_norm": 0.394871324300766,
+      "learning_rate": 9.046804429654437e-06,
+      "loss": 0.4137,
+      "step": 3338
+    },
+    {
+      "epoch": 0.8397887323943662,
+      "grad_norm": 0.3549323081970215,
+      "learning_rate": 9.045944874019522e-06,
+      "loss": 0.4062,
+      "step": 3339
+    },
+    {
+      "epoch": 0.8400402414486922,
+      "grad_norm": 0.34717488288879395,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.3685,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8402917505030181,
+      "grad_norm": 0.34687772393226624,
+      "learning_rate": 9.044224723293734e-06,
+      "loss": 0.3554,
+      "step": 3341
+    },
+    {
+      "epoch": 0.8405432595573441,
+      "grad_norm": 0.4031140208244324,
+      "learning_rate": 9.043364128350183e-06,
+      "loss": 0.4139,
+      "step": 3342
+    },
+    {
+      "epoch": 0.84079476861167,
+      "grad_norm": 0.33681127429008484,
+      "learning_rate": 9.042503187117788e-06,
+      "loss": 0.3659,
+      "step": 3343
+    },
+    {
+      "epoch": 0.8410462776659959,
+      "grad_norm": 0.31947392225265503,
+      "learning_rate": 9.041641899670286e-06,
+      "loss": 0.3752,
+      "step": 3344
+    },
+    {
+      "epoch": 0.841297786720322,
+      "grad_norm": 0.4157194197177887,
+      "learning_rate": 9.04078026608144e-06,
+      "loss": 0.4241,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8415492957746479,
+      "grad_norm": 0.340215802192688,
+      "learning_rate": 9.039918286425042e-06,
+      "loss": 0.3933,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8418008048289738,
+      "grad_norm": 0.3351587951183319,
+      "learning_rate": 9.039055960774918e-06,
+      "loss": 0.3608,
+      "step": 3347
+    },
+    {
+      "epoch": 0.8420523138832998,
+      "grad_norm": 0.3640073537826538,
+      "learning_rate": 9.038193289204919e-06,
+      "loss": 0.3826,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8423038229376257,
+      "grad_norm": 0.37294724583625793,
+      "learning_rate": 9.037330271788927e-06,
+      "loss": 0.3644,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8425553319919518,
+      "grad_norm": 0.3982885777950287,
+      "learning_rate": 9.036466908600856e-06,
+      "loss": 0.3856,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8428068410462777,
+      "grad_norm": 0.3464895486831665,
+      "learning_rate": 9.035603199714645e-06,
+      "loss": 0.3804,
+      "step": 3351
+    },
+    {
+      "epoch": 0.8430583501006036,
+      "grad_norm": 0.34546592831611633,
+      "learning_rate": 9.034739145204266e-06,
+      "loss": 0.3949,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8433098591549296,
+      "grad_norm": 0.39847803115844727,
+      "learning_rate": 9.033874745143722e-06,
+      "loss": 0.3775,
+      "step": 3353
+    },
+    {
+      "epoch": 0.8435613682092555,
+      "grad_norm": 0.34165751934051514,
+      "learning_rate": 9.033009999607042e-06,
+      "loss": 0.3892,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8438128772635815,
+      "grad_norm": 0.35674750804901123,
+      "learning_rate": 9.032144908668284e-06,
+      "loss": 0.3703,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8440643863179075,
+      "grad_norm": 0.3607436418533325,
+      "learning_rate": 9.031279472401542e-06,
+      "loss": 0.402,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8443158953722334,
+      "grad_norm": 0.35292503237724304,
+      "learning_rate": 9.03041369088093e-06,
+      "loss": 0.3808,
+      "step": 3357
+    },
+    {
+      "epoch": 0.8445674044265593,
+      "grad_norm": 0.35224297642707825,
+      "learning_rate": 9.029547564180602e-06,
+      "loss": 0.3921,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8448189134808853,
+      "grad_norm": 0.3472045361995697,
+      "learning_rate": 9.028681092374733e-06,
+      "loss": 0.4116,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 0.3715015947818756,
+      "learning_rate": 9.027814275537533e-06,
+      "loss": 0.3823,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8453219315895373,
+      "grad_norm": 0.3859766721725464,
+      "learning_rate": 9.026947113743237e-06,
+      "loss": 0.3872,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8455734406438632,
+      "grad_norm": 0.36284056305885315,
+      "learning_rate": 9.026079607066112e-06,
+      "loss": 0.3781,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8458249496981891,
+      "grad_norm": 0.35753417015075684,
+      "learning_rate": 9.025211755580458e-06,
+      "loss": 0.3545,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8460764587525151,
+      "grad_norm": 0.4312700927257538,
+      "learning_rate": 9.024343559360597e-06,
+      "loss": 0.3925,
+      "step": 3364
+    },
+    {
+      "epoch": 0.846327967806841,
+      "grad_norm": 0.3738209903240204,
+      "learning_rate": 9.023475018480888e-06,
+      "loss": 0.3912,
+      "step": 3365
+    },
+    {
+      "epoch": 0.846579476861167,
+      "grad_norm": 0.3941510021686554,
+      "learning_rate": 9.022606133015713e-06,
+      "loss": 0.3617,
+      "step": 3366
+    },
+    {
+      "epoch": 0.846830985915493,
+      "grad_norm": 0.4269218146800995,
+      "learning_rate": 9.021736903039488e-06,
+      "loss": 0.3998,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8470824949698189,
+      "grad_norm": 0.35374465584754944,
+      "learning_rate": 9.020867328626659e-06,
+      "loss": 0.3785,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8473340040241448,
+      "grad_norm": 0.34942927956581116,
+      "learning_rate": 9.019997409851696e-06,
+      "loss": 0.374,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8475855130784709,
+      "grad_norm": 0.4133439064025879,
+      "learning_rate": 9.019127146789106e-06,
+      "loss": 0.4161,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8478370221327968,
+      "grad_norm": 0.38321179151535034,
+      "learning_rate": 9.018256539513417e-06,
+      "loss": 0.3675,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8480885311871227,
+      "grad_norm": 0.3283655345439911,
+      "learning_rate": 9.017385588099195e-06,
+      "loss": 0.3838,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8483400402414487,
+      "grad_norm": 0.3512178063392639,
+      "learning_rate": 9.016514292621027e-06,
+      "loss": 0.3827,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8485915492957746,
+      "grad_norm": 0.4433364272117615,
+      "learning_rate": 9.015642653153542e-06,
+      "loss": 0.3925,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8488430583501007,
+      "grad_norm": 0.3722871243953705,
+      "learning_rate": 9.014770669771383e-06,
+      "loss": 0.364,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8490945674044266,
+      "grad_norm": 0.3842884600162506,
+      "learning_rate": 9.013898342549233e-06,
+      "loss": 0.3998,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8493460764587525,
+      "grad_norm": 0.39675992727279663,
+      "learning_rate": 9.013025671561798e-06,
+      "loss": 0.4026,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8495975855130785,
+      "grad_norm": 0.34844744205474854,
+      "learning_rate": 9.012152656883824e-06,
+      "loss": 0.3794,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8498490945674044,
+      "grad_norm": 0.3376064598560333,
+      "learning_rate": 9.011279298590072e-06,
+      "loss": 0.3641,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8501006036217303,
+      "grad_norm": 0.3386366665363312,
+      "learning_rate": 9.010405596755345e-06,
+      "loss": 0.3722,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8503521126760564,
+      "grad_norm": 0.3699395954608917,
+      "learning_rate": 9.009531551454465e-06,
+      "loss": 0.39,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8506036217303823,
+      "grad_norm": 0.35013461112976074,
+      "learning_rate": 9.008657162762293e-06,
+      "loss": 0.3577,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8508551307847082,
+      "grad_norm": 0.3586879074573517,
+      "learning_rate": 9.007782430753712e-06,
+      "loss": 0.3754,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8511066398390342,
+      "grad_norm": 0.3715461790561676,
+      "learning_rate": 9.006907355503639e-06,
+      "loss": 0.3569,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8513581488933601,
+      "grad_norm": 0.4095889925956726,
+      "learning_rate": 9.006031937087018e-06,
+      "loss": 0.3823,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8516096579476862,
+      "grad_norm": 0.34555843472480774,
+      "learning_rate": 9.005156175578823e-06,
+      "loss": 0.4092,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8518611670020121,
+      "grad_norm": 0.3641471266746521,
+      "learning_rate": 9.004280071054058e-06,
+      "loss": 0.3735,
+      "step": 3387
+    },
+    {
+      "epoch": 0.852112676056338,
+      "grad_norm": 0.3723176419734955,
+      "learning_rate": 9.003403623587757e-06,
+      "loss": 0.3639,
+      "step": 3388
+    },
+    {
+      "epoch": 0.852364185110664,
+      "grad_norm": 0.3690603971481323,
+      "learning_rate": 9.002526833254979e-06,
+      "loss": 0.3734,
+      "step": 3389
+    },
+    {
+      "epoch": 0.85261569416499,
+      "grad_norm": 0.33589431643486023,
+      "learning_rate": 9.001649700130816e-06,
+      "loss": 0.3743,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8528672032193159,
+      "grad_norm": 0.3753211498260498,
+      "learning_rate": 9.000772224290393e-06,
+      "loss": 0.4133,
+      "step": 3391
+    },
+    {
+      "epoch": 0.8531187122736419,
+      "grad_norm": 0.34786948561668396,
+      "learning_rate": 8.999894405808857e-06,
+      "loss": 0.3791,
+      "step": 3392
+    },
+    {
+      "epoch": 0.8533702213279678,
+      "grad_norm": 0.3724614977836609,
+      "learning_rate": 8.99901624476139e-06,
+      "loss": 0.3819,
+      "step": 3393
+    },
+    {
+      "epoch": 0.8536217303822937,
+      "grad_norm": 0.34503018856048584,
+      "learning_rate": 8.998137741223196e-06,
+      "loss": 0.3858,
+      "step": 3394
+    },
+    {
+      "epoch": 0.8538732394366197,
+      "grad_norm": 0.37502291798591614,
+      "learning_rate": 8.99725889526952e-06,
+      "loss": 0.4026,
+      "step": 3395
+    },
+    {
+      "epoch": 0.8541247484909457,
+      "grad_norm": 0.3725481927394867,
+      "learning_rate": 8.996379706975624e-06,
+      "loss": 0.353,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8543762575452716,
+      "grad_norm": 0.3500644862651825,
+      "learning_rate": 8.995500176416809e-06,
+      "loss": 0.3909,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8546277665995976,
+      "grad_norm": 0.3630251884460449,
+      "learning_rate": 8.9946203036684e-06,
+      "loss": 0.3797,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8548792756539235,
+      "grad_norm": 0.3816511034965515,
+      "learning_rate": 8.99374008880575e-06,
+      "loss": 0.3622,
+      "step": 3399
+    },
+    {
+      "epoch": 0.8551307847082495,
+      "grad_norm": 0.37829670310020447,
+      "learning_rate": 8.992859531904247e-06,
+      "loss": 0.3791,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8553822937625755,
+      "grad_norm": 0.3598082363605499,
+      "learning_rate": 8.991978633039305e-06,
+      "loss": 0.3779,
+      "step": 3401
+    },
+    {
+      "epoch": 0.8556338028169014,
+      "grad_norm": 0.3571262061595917,
+      "learning_rate": 8.991097392286368e-06,
+      "loss": 0.3634,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8558853118712274,
+      "grad_norm": 0.38512668013572693,
+      "learning_rate": 8.990215809720905e-06,
+      "loss": 0.3692,
+      "step": 3403
+    },
+    {
+      "epoch": 0.8561368209255533,
+      "grad_norm": 0.35630372166633606,
+      "learning_rate": 8.989333885418423e-06,
+      "loss": 0.3878,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8563883299798792,
+      "grad_norm": 0.37692365050315857,
+      "learning_rate": 8.988451619454449e-06,
+      "loss": 0.3886,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8566398390342053,
+      "grad_norm": 0.3339127004146576,
+      "learning_rate": 8.987569011904547e-06,
+      "loss": 0.3821,
+      "step": 3406
+    },
+    {
+      "epoch": 0.8568913480885312,
+      "grad_norm": 0.3581376075744629,
+      "learning_rate": 8.986686062844303e-06,
+      "loss": 0.4086,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.39820095896720886,
+      "learning_rate": 8.985802772349341e-06,
+      "loss": 0.3902,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8573943661971831,
+      "grad_norm": 0.3543420135974884,
+      "learning_rate": 8.984919140495305e-06,
+      "loss": 0.3623,
+      "step": 3409
+    },
+    {
+      "epoch": 0.857645875251509,
+      "grad_norm": 0.3442290723323822,
+      "learning_rate": 8.984035167357874e-06,
+      "loss": 0.3704,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8578973843058351,
+      "grad_norm": 0.3457908034324646,
+      "learning_rate": 8.983150853012756e-06,
+      "loss": 0.3606,
+      "step": 3411
+    },
+    {
+      "epoch": 0.858148893360161,
+      "grad_norm": 0.36458566784858704,
+      "learning_rate": 8.982266197535685e-06,
+      "loss": 0.3841,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8584004024144869,
+      "grad_norm": 0.3184933364391327,
+      "learning_rate": 8.981381201002428e-06,
+      "loss": 0.3728,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8586519114688129,
+      "grad_norm": 0.3275623917579651,
+      "learning_rate": 8.98049586348878e-06,
+      "loss": 0.373,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8589034205231388,
+      "grad_norm": 0.35574376583099365,
+      "learning_rate": 8.979610185070562e-06,
+      "loss": 0.3598,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8591549295774648,
+      "grad_norm": 0.3995131850242615,
+      "learning_rate": 8.978724165823626e-06,
+      "loss": 0.3797,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8594064386317908,
+      "grad_norm": 0.37157928943634033,
+      "learning_rate": 8.977837805823856e-06,
+      "loss": 0.4016,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8596579476861167,
+      "grad_norm": 0.33222272992134094,
+      "learning_rate": 8.976951105147167e-06,
+      "loss": 0.3852,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8599094567404426,
+      "grad_norm": 0.37010085582733154,
+      "learning_rate": 8.976064063869493e-06,
+      "loss": 0.3865,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8601609657947686,
+      "grad_norm": 0.3240501582622528,
+      "learning_rate": 8.975176682066805e-06,
+      "loss": 0.4128,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8604124748490946,
+      "grad_norm": 0.33426669239997864,
+      "learning_rate": 8.974288959815105e-06,
+      "loss": 0.3636,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8606639839034205,
+      "grad_norm": 0.34247368574142456,
+      "learning_rate": 8.973400897190418e-06,
+      "loss": 0.3842,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8609154929577465,
+      "grad_norm": 0.343246191740036,
+      "learning_rate": 8.9725124942688e-06,
+      "loss": 0.3869,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8611670020120724,
+      "grad_norm": 0.3464204668998718,
+      "learning_rate": 8.971623751126341e-06,
+      "loss": 0.387,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8614185110663984,
+      "grad_norm": 0.3398580551147461,
+      "learning_rate": 8.970734667839155e-06,
+      "loss": 0.3775,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8616700201207244,
+      "grad_norm": 0.3092399835586548,
+      "learning_rate": 8.969845244483383e-06,
+      "loss": 0.3597,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8619215291750503,
+      "grad_norm": 0.37892070412635803,
+      "learning_rate": 8.968955481135202e-06,
+      "loss": 0.4005,
+      "step": 3427
+    },
+    {
+      "epoch": 0.8621730382293763,
+      "grad_norm": 0.33688947558403015,
+      "learning_rate": 8.968065377870814e-06,
+      "loss": 0.3724,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8624245472837022,
+      "grad_norm": 0.35349977016448975,
+      "learning_rate": 8.967174934766452e-06,
+      "loss": 0.3535,
+      "step": 3429
+    },
+    {
+      "epoch": 0.8626760563380281,
+      "grad_norm": 0.34343835711479187,
+      "learning_rate": 8.966284151898373e-06,
+      "loss": 0.3627,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8629275653923542,
+      "grad_norm": 0.34207355976104736,
+      "learning_rate": 8.965393029342871e-06,
+      "loss": 0.3745,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8631790744466801,
+      "grad_norm": 0.3593146502971649,
+      "learning_rate": 8.964501567176263e-06,
+      "loss": 0.3759,
+      "step": 3432
+    },
+    {
+      "epoch": 0.863430583501006,
+      "grad_norm": 0.37325525283813477,
+      "learning_rate": 8.963609765474897e-06,
+      "loss": 0.3783,
+      "step": 3433
+    },
+    {
+      "epoch": 0.863682092555332,
+      "grad_norm": 0.375765860080719,
+      "learning_rate": 8.96271762431515e-06,
+      "loss": 0.3963,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8639336016096579,
+      "grad_norm": 0.3585033714771271,
+      "learning_rate": 8.96182514377343e-06,
+      "loss": 0.3527,
+      "step": 3435
+    },
+    {
+      "epoch": 0.864185110663984,
+      "grad_norm": 0.42358458042144775,
+      "learning_rate": 8.960932323926172e-06,
+      "loss": 0.3856,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8644366197183099,
+      "grad_norm": 0.374779611825943,
+      "learning_rate": 8.96003916484984e-06,
+      "loss": 0.3643,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8646881287726358,
+      "grad_norm": 0.3398546874523163,
+      "learning_rate": 8.959145666620928e-06,
+      "loss": 0.368,
+      "step": 3438
+    },
+    {
+      "epoch": 0.8649396378269618,
+      "grad_norm": 0.36778420209884644,
+      "learning_rate": 8.958251829315957e-06,
+      "loss": 0.3526,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8651911468812877,
+      "grad_norm": 0.34406542778015137,
+      "learning_rate": 8.957357653011481e-06,
+      "loss": 0.3614,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8654426559356136,
+      "grad_norm": 0.3503788411617279,
+      "learning_rate": 8.956463137784077e-06,
+      "loss": 0.3757,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8656941649899397,
+      "grad_norm": 0.3319445550441742,
+      "learning_rate": 8.955568283710359e-06,
+      "loss": 0.3972,
+      "step": 3442
+    },
+    {
+      "epoch": 0.8659456740442656,
+      "grad_norm": 0.39669501781463623,
+      "learning_rate": 8.95467309086696e-06,
+      "loss": 0.3801,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8661971830985915,
+      "grad_norm": 0.3425920605659485,
+      "learning_rate": 8.953777559330554e-06,
+      "loss": 0.3563,
+      "step": 3444
+    },
+    {
+      "epoch": 0.8664486921529175,
+      "grad_norm": 0.335559606552124,
+      "learning_rate": 8.952881689177834e-06,
+      "loss": 0.3907,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8667002012072434,
+      "grad_norm": 0.3417853116989136,
+      "learning_rate": 8.951985480485528e-06,
+      "loss": 0.3652,
+      "step": 3446
+    },
+    {
+      "epoch": 0.8669517102615694,
+      "grad_norm": 0.3930087983608246,
+      "learning_rate": 8.951088933330387e-06,
+      "loss": 0.3849,
+      "step": 3447
+    },
+    {
+      "epoch": 0.8672032193158954,
+      "grad_norm": 0.3572435677051544,
+      "learning_rate": 8.950192047789198e-06,
+      "loss": 0.368,
+      "step": 3448
+    },
+    {
+      "epoch": 0.8674547283702213,
+      "grad_norm": 0.3223349153995514,
+      "learning_rate": 8.949294823938773e-06,
+      "loss": 0.4082,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8677062374245473,
+      "grad_norm": 0.40850120782852173,
+      "learning_rate": 8.94839726185595e-06,
+      "loss": 0.3569,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8679577464788732,
+      "grad_norm": 0.3549629747867584,
+      "learning_rate": 8.947499361617606e-06,
+      "loss": 0.3895,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8682092555331992,
+      "grad_norm": 0.333429753780365,
+      "learning_rate": 8.946601123300636e-06,
+      "loss": 0.3784,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8684607645875252,
+      "grad_norm": 0.34425967931747437,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.3644,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8687122736418511,
+      "grad_norm": 0.4264945089817047,
+      "learning_rate": 8.944803632738563e-06,
+      "loss": 0.3787,
+      "step": 3454
+    },
+    {
+      "epoch": 0.868963782696177,
+      "grad_norm": 0.35965225100517273,
+      "learning_rate": 8.943904380647406e-06,
+      "loss": 0.3696,
+      "step": 3455
+    },
+    {
+      "epoch": 0.869215291750503,
+      "grad_norm": 0.39319103956222534,
+      "learning_rate": 8.94300479078551e-06,
+      "loss": 0.3785,
+      "step": 3456
+    },
+    {
+      "epoch": 0.869466800804829,
+      "grad_norm": 0.3534758985042572,
+      "learning_rate": 8.942104863229923e-06,
+      "loss": 0.3616,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8697183098591549,
+      "grad_norm": 0.3524431586265564,
+      "learning_rate": 8.941204598057715e-06,
+      "loss": 0.3972,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8699698189134809,
+      "grad_norm": 0.36104217171669006,
+      "learning_rate": 8.940303995345988e-06,
+      "loss": 0.37,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8702213279678068,
+      "grad_norm": 0.39354491233825684,
+      "learning_rate": 8.939403055171877e-06,
+      "loss": 0.3858,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8704728370221329,
+      "grad_norm": 0.3502269983291626,
+      "learning_rate": 8.938501777612538e-06,
+      "loss": 0.3702,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8707243460764588,
+      "grad_norm": 0.3592958450317383,
+      "learning_rate": 8.937600162745159e-06,
+      "loss": 0.3917,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8709758551307847,
+      "grad_norm": 0.3925688862800598,
+      "learning_rate": 8.936698210646962e-06,
+      "loss": 0.3878,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8712273641851107,
+      "grad_norm": 0.35481512546539307,
+      "learning_rate": 8.93579592139519e-06,
+      "loss": 0.3642,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8714788732394366,
+      "grad_norm": 0.35621413588523865,
+      "learning_rate": 8.93489329506712e-06,
+      "loss": 0.408,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8717303822937625,
+      "grad_norm": 0.3342260420322418,
+      "learning_rate": 8.933990331740056e-06,
+      "loss": 0.3867,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8719818913480886,
+      "grad_norm": 0.3696461617946625,
+      "learning_rate": 8.933087031491332e-06,
+      "loss": 0.3788,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8722334004024145,
+      "grad_norm": 0.38092076778411865,
+      "learning_rate": 8.932183394398309e-06,
+      "loss": 0.375,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8724849094567404,
+      "grad_norm": 0.34763234853744507,
+      "learning_rate": 8.931279420538377e-06,
+      "loss": 0.3722,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8727364185110664,
+      "grad_norm": 0.3381394147872925,
+      "learning_rate": 8.930375109988956e-06,
+      "loss": 0.4076,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8729879275653923,
+      "grad_norm": 0.34451600909233093,
+      "learning_rate": 8.929470462827496e-06,
+      "loss": 0.3787,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8732394366197183,
+      "grad_norm": 0.33672693371772766,
+      "learning_rate": 8.928565479131473e-06,
+      "loss": 0.3889,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8734909456740443,
+      "grad_norm": 0.3693379759788513,
+      "learning_rate": 8.927660158978392e-06,
+      "loss": 0.3873,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8737424547283702,
+      "grad_norm": 0.32799556851387024,
+      "learning_rate": 8.926754502445794e-06,
+      "loss": 0.3619,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8739939637826962,
+      "grad_norm": 0.36521029472351074,
+      "learning_rate": 8.925848509611237e-06,
+      "loss": 0.3807,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8742454728370221,
+      "grad_norm": 0.3440057933330536,
+      "learning_rate": 8.924942180552315e-06,
+      "loss": 0.3797,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8744969818913481,
+      "grad_norm": 0.35602283477783203,
+      "learning_rate": 8.924035515346648e-06,
+      "loss": 0.3678,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8747484909456741,
+      "grad_norm": 0.34389829635620117,
+      "learning_rate": 8.923128514071888e-06,
+      "loss": 0.3683,
+      "step": 3478
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.37036365270614624,
+      "learning_rate": 8.922221176805715e-06,
+      "loss": 0.3992,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8752515090543259,
+      "grad_norm": 0.3461644649505615,
+      "learning_rate": 8.921313503625835e-06,
+      "loss": 0.3997,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8755030181086519,
+      "grad_norm": 0.35094374418258667,
+      "learning_rate": 8.920405494609986e-06,
+      "loss": 0.3968,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8757545271629779,
+      "grad_norm": 0.35670819878578186,
+      "learning_rate": 8.919497149835932e-06,
+      "loss": 0.3609,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8760060362173038,
+      "grad_norm": 0.37578085064888,
+      "learning_rate": 8.918588469381468e-06,
+      "loss": 0.3756,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8762575452716298,
+      "grad_norm": 0.35081595182418823,
+      "learning_rate": 8.917679453324415e-06,
+      "loss": 0.379,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8765090543259557,
+      "grad_norm": 0.3629753291606903,
+      "learning_rate": 8.916770101742627e-06,
+      "loss": 0.4066,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8767605633802817,
+      "grad_norm": 0.3397653102874756,
+      "learning_rate": 8.915860414713981e-06,
+      "loss": 0.4097,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8770120724346077,
+      "grad_norm": 0.3421812951564789,
+      "learning_rate": 8.91495039231639e-06,
+      "loss": 0.4011,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8772635814889336,
+      "grad_norm": 0.39413323998451233,
+      "learning_rate": 8.914040034627788e-06,
+      "loss": 0.4002,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8775150905432596,
+      "grad_norm": 0.37600401043891907,
+      "learning_rate": 8.913129341726144e-06,
+      "loss": 0.3507,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8777665995975855,
+      "grad_norm": 0.3659670352935791,
+      "learning_rate": 8.912218313689453e-06,
+      "loss": 0.3735,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8780181086519114,
+      "grad_norm": 0.3923790156841278,
+      "learning_rate": 8.911306950595737e-06,
+      "loss": 0.358,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8782696177062375,
+      "grad_norm": 0.37082934379577637,
+      "learning_rate": 8.910395252523053e-06,
+      "loss": 0.392,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8785211267605634,
+      "grad_norm": 0.4046441912651062,
+      "learning_rate": 8.909483219549475e-06,
+      "loss": 0.3697,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8787726358148893,
+      "grad_norm": 0.3740319311618805,
+      "learning_rate": 8.908570851753117e-06,
+      "loss": 0.3507,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8790241448692153,
+      "grad_norm": 0.368447870016098,
+      "learning_rate": 8.907658149212119e-06,
+      "loss": 0.3536,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8792756539235412,
+      "grad_norm": 0.3541277348995209,
+      "learning_rate": 8.906745112004646e-06,
+      "loss": 0.3628,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8795271629778671,
+      "grad_norm": 0.36320897936820984,
+      "learning_rate": 8.905831740208896e-06,
+      "loss": 0.3541,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8797786720321932,
+      "grad_norm": 0.3422798216342926,
+      "learning_rate": 8.904918033903091e-06,
+      "loss": 0.364,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8800301810865191,
+      "grad_norm": 0.35075971484184265,
+      "learning_rate": 8.904003993165487e-06,
+      "loss": 0.3883,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8802816901408451,
+      "grad_norm": 0.3443789780139923,
+      "learning_rate": 8.903089618074362e-06,
+      "loss": 0.3693,
+      "step": 3500
+    },
+    {
+      "epoch": 0.880533199195171,
+      "grad_norm": 0.3593481481075287,
+      "learning_rate": 8.902174908708032e-06,
+      "loss": 0.3749,
+      "step": 3501
+    },
+    {
+      "epoch": 0.880784708249497,
+      "grad_norm": 0.33803337812423706,
+      "learning_rate": 8.901259865144831e-06,
+      "loss": 0.3927,
+      "step": 3502
+    },
+    {
+      "epoch": 0.881036217303823,
+      "grad_norm": 0.4012623131275177,
+      "learning_rate": 8.900344487463128e-06,
+      "loss": 0.393,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8812877263581489,
+      "grad_norm": 0.35286056995391846,
+      "learning_rate": 8.899428775741321e-06,
+      "loss": 0.377,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8815392354124748,
+      "grad_norm": 0.3573596775531769,
+      "learning_rate": 8.898512730057835e-06,
+      "loss": 0.3808,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8817907444668008,
+      "grad_norm": 0.33580759167671204,
+      "learning_rate": 8.897596350491122e-06,
+      "loss": 0.3601,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8820422535211268,
+      "grad_norm": 0.32967039942741394,
+      "learning_rate": 8.896679637119665e-06,
+      "loss": 0.3586,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8822937625754527,
+      "grad_norm": 0.3663707971572876,
+      "learning_rate": 8.895762590021973e-06,
+      "loss": 0.3816,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8825452716297787,
+      "grad_norm": 0.33888426423072815,
+      "learning_rate": 8.89484520927659e-06,
+      "loss": 0.3904,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8827967806841046,
+      "grad_norm": 0.334952175617218,
+      "learning_rate": 8.893927494962078e-06,
+      "loss": 0.3941,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8830482897384306,
+      "grad_norm": 0.32887038588523865,
+      "learning_rate": 8.893009447157039e-06,
+      "loss": 0.3737,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8832997987927566,
+      "grad_norm": 0.32995015382766724,
+      "learning_rate": 8.892091065940093e-06,
+      "loss": 0.3787,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8835513078470825,
+      "grad_norm": 0.33268022537231445,
+      "learning_rate": 8.891172351389898e-06,
+      "loss": 0.3939,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8838028169014085,
+      "grad_norm": 0.34591397643089294,
+      "learning_rate": 8.890253303585133e-06,
+      "loss": 0.3809,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8840543259557344,
+      "grad_norm": 0.357835054397583,
+      "learning_rate": 8.88933392260451e-06,
+      "loss": 0.3565,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8843058350100603,
+      "grad_norm": 0.3453388214111328,
+      "learning_rate": 8.888414208526768e-06,
+      "loss": 0.3891,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8845573440643864,
+      "grad_norm": 0.35793086886405945,
+      "learning_rate": 8.887494161430676e-06,
+      "loss": 0.3938,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8848088531187123,
+      "grad_norm": 0.32767254114151,
+      "learning_rate": 8.886573781395028e-06,
+      "loss": 0.3872,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8850603621730382,
+      "grad_norm": 0.3634639084339142,
+      "learning_rate": 8.88565306849865e-06,
+      "loss": 0.4025,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8853118712273642,
+      "grad_norm": 0.34274405241012573,
+      "learning_rate": 8.884732022820396e-06,
+      "loss": 0.4144,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8855633802816901,
+      "grad_norm": 0.3598005771636963,
+      "learning_rate": 8.883810644439146e-06,
+      "loss": 0.3667,
+      "step": 3521
+    },
+    {
+      "epoch": 0.885814889336016,
+      "grad_norm": 0.3647115230560303,
+      "learning_rate": 8.882888933433813e-06,
+      "loss": 0.3797,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8860663983903421,
+      "grad_norm": 0.3279248774051666,
+      "learning_rate": 8.881966889883334e-06,
+      "loss": 0.3718,
+      "step": 3523
+    },
+    {
+      "epoch": 0.886317907444668,
+      "grad_norm": 0.39648541808128357,
+      "learning_rate": 8.881044513866675e-06,
+      "loss": 0.3783,
+      "step": 3524
+    },
+    {
+      "epoch": 0.886569416498994,
+      "grad_norm": 0.3662944734096527,
+      "learning_rate": 8.880121805462834e-06,
+      "loss": 0.3796,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8868209255533199,
+      "grad_norm": 0.3248366415500641,
+      "learning_rate": 8.879198764750834e-06,
+      "loss": 0.3804,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8870724346076458,
+      "grad_norm": 0.3559374511241913,
+      "learning_rate": 8.878275391809727e-06,
+      "loss": 0.3701,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8873239436619719,
+      "grad_norm": 0.37350761890411377,
+      "learning_rate": 8.877351686718596e-06,
+      "loss": 0.3769,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8875754527162978,
+      "grad_norm": 0.37377414107322693,
+      "learning_rate": 8.876427649556549e-06,
+      "loss": 0.4021,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8878269617706237,
+      "grad_norm": 0.3348924219608307,
+      "learning_rate": 8.875503280402727e-06,
+      "loss": 0.3881,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8880784708249497,
+      "grad_norm": 0.33580613136291504,
+      "learning_rate": 8.874578579336293e-06,
+      "loss": 0.3735,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8883299798792756,
+      "grad_norm": 0.35645267367362976,
+      "learning_rate": 8.873653546436442e-06,
+      "loss": 0.3988,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8885814889336016,
+      "grad_norm": 0.33889126777648926,
+      "learning_rate": 8.872728181782399e-06,
+      "loss": 0.3856,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8888329979879276,
+      "grad_norm": 0.329007089138031,
+      "learning_rate": 8.871802485453414e-06,
+      "loss": 0.37,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8890845070422535,
+      "grad_norm": 0.31005194783210754,
+      "learning_rate": 8.87087645752877e-06,
+      "loss": 0.3784,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8893360160965795,
+      "grad_norm": 0.3521806299686432,
+      "learning_rate": 8.86995009808777e-06,
+      "loss": 0.3909,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8895875251509054,
+      "grad_norm": 0.3143180310726166,
+      "learning_rate": 8.869023407209758e-06,
+      "loss": 0.4024,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8898390342052314,
+      "grad_norm": 0.32495957612991333,
+      "learning_rate": 8.868096384974094e-06,
+      "loss": 0.3827,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8900905432595574,
+      "grad_norm": 0.30812323093414307,
+      "learning_rate": 8.867169031460175e-06,
+      "loss": 0.362,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8903420523138833,
+      "grad_norm": 0.3465098440647125,
+      "learning_rate": 8.86624134674742e-06,
+      "loss": 0.3744,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8905935613682092,
+      "grad_norm": 0.3428845703601837,
+      "learning_rate": 8.86531333091528e-06,
+      "loss": 0.3918,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8908450704225352,
+      "grad_norm": 0.35687991976737976,
+      "learning_rate": 8.864384984043234e-06,
+      "loss": 0.3882,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8910965794768612,
+      "grad_norm": 0.37335455417633057,
+      "learning_rate": 8.863456306210793e-06,
+      "loss": 0.3702,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8913480885311871,
+      "grad_norm": 0.3432762622833252,
+      "learning_rate": 8.862527297497488e-06,
+      "loss": 0.3811,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8915995975855131,
+      "grad_norm": 0.3809764087200165,
+      "learning_rate": 8.861597957982881e-06,
+      "loss": 0.3995,
+      "step": 3545
+    },
+    {
+      "epoch": 0.891851106639839,
+      "grad_norm": 0.3548218309879303,
+      "learning_rate": 8.86066828774657e-06,
+      "loss": 0.3965,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8921026156941649,
+      "grad_norm": 0.3315393924713135,
+      "learning_rate": 8.859738286868172e-06,
+      "loss": 0.374,
+      "step": 3547
+    },
+    {
+      "epoch": 0.892354124748491,
+      "grad_norm": 0.35290321707725525,
+      "learning_rate": 8.858807955427335e-06,
+      "loss": 0.3589,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8926056338028169,
+      "grad_norm": 0.368217796087265,
+      "learning_rate": 8.857877293503739e-06,
+      "loss": 0.3713,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 0.3560136556625366,
+      "learning_rate": 8.856946301177085e-06,
+      "loss": 0.3697,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8931086519114688,
+      "grad_norm": 0.3632422089576721,
+      "learning_rate": 8.85601497852711e-06,
+      "loss": 0.3798,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8933601609657947,
+      "grad_norm": 0.3566484749317169,
+      "learning_rate": 8.855083325633578e-06,
+      "loss": 0.3761,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8936116700201208,
+      "grad_norm": 0.36059099435806274,
+      "learning_rate": 8.854151342576274e-06,
+      "loss": 0.3683,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8938631790744467,
+      "grad_norm": 0.36407434940338135,
+      "learning_rate": 8.853219029435019e-06,
+      "loss": 0.3793,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8941146881287726,
+      "grad_norm": 0.36072081327438354,
+      "learning_rate": 8.852286386289662e-06,
+      "loss": 0.3949,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8943661971830986,
+      "grad_norm": 0.3494257628917694,
+      "learning_rate": 8.851353413220073e-06,
+      "loss": 0.3772,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8946177062374245,
+      "grad_norm": 0.3464342951774597,
+      "learning_rate": 8.850420110306159e-06,
+      "loss": 0.4028,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8948692152917505,
+      "grad_norm": 0.3632872402667999,
+      "learning_rate": 8.849486477627851e-06,
+      "loss": 0.378,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8951207243460765,
+      "grad_norm": 0.3627418875694275,
+      "learning_rate": 8.848552515265108e-06,
+      "loss": 0.3887,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8953722334004024,
+      "grad_norm": 0.3480847477912903,
+      "learning_rate": 8.84761822329792e-06,
+      "loss": 0.3831,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8956237424547284,
+      "grad_norm": 0.3408527374267578,
+      "learning_rate": 8.8466836018063e-06,
+      "loss": 0.3716,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8958752515090543,
+      "grad_norm": 0.35451194643974304,
+      "learning_rate": 8.845748650870297e-06,
+      "loss": 0.3733,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8961267605633803,
+      "grad_norm": 0.3594439625740051,
+      "learning_rate": 8.844813370569978e-06,
+      "loss": 0.3822,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8963782696177063,
+      "grad_norm": 0.33832788467407227,
+      "learning_rate": 8.843877760985447e-06,
+      "loss": 0.3858,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8966297786720322,
+      "grad_norm": 0.33233413100242615,
+      "learning_rate": 8.842941822196835e-06,
+      "loss": 0.3995,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8968812877263581,
+      "grad_norm": 0.3875477612018585,
+      "learning_rate": 8.842005554284296e-06,
+      "loss": 0.3922,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8971327967806841,
+      "grad_norm": 0.32244715094566345,
+      "learning_rate": 8.841068957328018e-06,
+      "loss": 0.345,
+      "step": 3567
+    },
+    {
+      "epoch": 0.89738430583501,
+      "grad_norm": 0.36826932430267334,
+      "learning_rate": 8.84013203140821e-06,
+      "loss": 0.3665,
+      "step": 3568
+    },
+    {
+      "epoch": 0.897635814889336,
+      "grad_norm": 0.3594595789909363,
+      "learning_rate": 8.839194776605121e-06,
+      "loss": 0.3917,
+      "step": 3569
+    },
+    {
+      "epoch": 0.897887323943662,
+      "grad_norm": 0.31721267104148865,
+      "learning_rate": 8.838257192999016e-06,
+      "loss": 0.3672,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8981388329979879,
+      "grad_norm": 0.37041252851486206,
+      "learning_rate": 8.837319280670196e-06,
+      "loss": 0.3852,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8983903420523138,
+      "grad_norm": 0.3686506152153015,
+      "learning_rate": 8.836381039698983e-06,
+      "loss": 0.3628,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8986418511066399,
+      "grad_norm": 0.3326662480831146,
+      "learning_rate": 8.835442470165736e-06,
+      "loss": 0.3954,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8988933601609658,
+      "grad_norm": 0.35208654403686523,
+      "learning_rate": 8.834503572150835e-06,
+      "loss": 0.3983,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8991448692152918,
+      "grad_norm": 0.34213465452194214,
+      "learning_rate": 8.833564345734693e-06,
+      "loss": 0.3492,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8993963782696177,
+      "grad_norm": 0.3441331684589386,
+      "learning_rate": 8.832624790997747e-06,
+      "loss": 0.3914,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8996478873239436,
+      "grad_norm": 0.33155056834220886,
+      "learning_rate": 8.831684908020463e-06,
+      "loss": 0.4002,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8998993963782697,
+      "grad_norm": 0.36803004145622253,
+      "learning_rate": 8.83074469688334e-06,
+      "loss": 0.3868,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9001509054325956,
+      "grad_norm": 0.35357141494750977,
+      "learning_rate": 8.829804157666896e-06,
+      "loss": 0.3788,
+      "step": 3579
+    },
+    {
+      "epoch": 0.9004024144869215,
+      "grad_norm": 0.3572876751422882,
+      "learning_rate": 8.828863290451689e-06,
+      "loss": 0.414,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9006539235412475,
+      "grad_norm": 0.3837227523326874,
+      "learning_rate": 8.82792209531829e-06,
+      "loss": 0.3958,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9009054325955734,
+      "grad_norm": 0.3402830958366394,
+      "learning_rate": 8.826980572347314e-06,
+      "loss": 0.3599,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9011569416498993,
+      "grad_norm": 0.41772645711898804,
+      "learning_rate": 8.826038721619393e-06,
+      "loss": 0.3844,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 0.36479735374450684,
+      "learning_rate": 8.82509654321519e-06,
+      "loss": 0.3924,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9016599597585513,
+      "grad_norm": 0.3720943033695221,
+      "learning_rate": 8.824154037215399e-06,
+      "loss": 0.3861,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9019114688128773,
+      "grad_norm": 0.3853396475315094,
+      "learning_rate": 8.823211203700738e-06,
+      "loss": 0.4013,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9021629778672032,
+      "grad_norm": 0.3603013753890991,
+      "learning_rate": 8.822268042751956e-06,
+      "loss": 0.379,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9024144869215291,
+      "grad_norm": 0.3663939833641052,
+      "learning_rate": 8.821324554449826e-06,
+      "loss": 0.3767,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9026659959758552,
+      "grad_norm": 0.3621923327445984,
+      "learning_rate": 8.820380738875156e-06,
+      "loss": 0.3731,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9029175050301811,
+      "grad_norm": 0.333791583776474,
+      "learning_rate": 8.819436596108775e-06,
+      "loss": 0.384,
+      "step": 3590
+    },
+    {
+      "epoch": 0.903169014084507,
+      "grad_norm": 0.36841997504234314,
+      "learning_rate": 8.818492126231545e-06,
+      "loss": 0.3656,
+      "step": 3591
+    },
+    {
+      "epoch": 0.903420523138833,
+      "grad_norm": 0.35295945405960083,
+      "learning_rate": 8.817547329324352e-06,
+      "loss": 0.3935,
+      "step": 3592
+    },
+    {
+      "epoch": 0.903672032193159,
+      "grad_norm": 0.3341405987739563,
+      "learning_rate": 8.816602205468113e-06,
+      "loss": 0.3899,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9039235412474849,
+      "grad_norm": 0.32092586159706116,
+      "learning_rate": 8.815656754743772e-06,
+      "loss": 0.3715,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9041750503018109,
+      "grad_norm": 0.3715393841266632,
+      "learning_rate": 8.814710977232299e-06,
+      "loss": 0.3753,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9044265593561368,
+      "grad_norm": 0.35690295696258545,
+      "learning_rate": 8.813764873014697e-06,
+      "loss": 0.3926,
+      "step": 3596
+    },
+    {
+      "epoch": 0.9046780684104627,
+      "grad_norm": 0.3435974419116974,
+      "learning_rate": 8.812818442171994e-06,
+      "loss": 0.3801,
+      "step": 3597
+    },
+    {
+      "epoch": 0.9049295774647887,
+      "grad_norm": 0.3631943166255951,
+      "learning_rate": 8.811871684785242e-06,
+      "loss": 0.3926,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9051810865191147,
+      "grad_norm": 0.3382608890533447,
+      "learning_rate": 8.810924600935527e-06,
+      "loss": 0.3677,
+      "step": 3599
+    },
+    {
+      "epoch": 0.9054325955734407,
+      "grad_norm": 0.3581875264644623,
+      "learning_rate": 8.809977190703961e-06,
+      "loss": 0.367,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9056841046277666,
+      "grad_norm": 0.3534909784793854,
+      "learning_rate": 8.809029454171684e-06,
+      "loss": 0.424,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9059356136820925,
+      "grad_norm": 0.3113289475440979,
+      "learning_rate": 8.808081391419865e-06,
+      "loss": 0.3781,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9061871227364185,
+      "grad_norm": 0.39388152956962585,
+      "learning_rate": 8.807133002529697e-06,
+      "loss": 0.4189,
+      "step": 3603
+    },
+    {
+      "epoch": 0.9064386317907445,
+      "grad_norm": 0.3407596945762634,
+      "learning_rate": 8.806184287582404e-06,
+      "loss": 0.3558,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9066901408450704,
+      "grad_norm": 0.38331282138824463,
+      "learning_rate": 8.80523524665924e-06,
+      "loss": 0.3851,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9069416498993964,
+      "grad_norm": 0.3627070188522339,
+      "learning_rate": 8.804285879841481e-06,
+      "loss": 0.3898,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9071931589537223,
+      "grad_norm": 0.31576594710350037,
+      "learning_rate": 8.803336187210437e-06,
+      "loss": 0.3956,
+      "step": 3607
+    },
+    {
+      "epoch": 0.9074446680080482,
+      "grad_norm": 0.38593655824661255,
+      "learning_rate": 8.802386168847442e-06,
+      "loss": 0.3779,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9076961770623743,
+      "grad_norm": 0.36420145630836487,
+      "learning_rate": 8.80143582483386e-06,
+      "loss": 0.3473,
+      "step": 3609
+    },
+    {
+      "epoch": 0.9079476861167002,
+      "grad_norm": 0.3500615656375885,
+      "learning_rate": 8.800485155251079e-06,
+      "loss": 0.3624,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9081991951710262,
+      "grad_norm": 0.3793070614337921,
+      "learning_rate": 8.799534160180521e-06,
+      "loss": 0.3848,
+      "step": 3611
+    },
+    {
+      "epoch": 0.9084507042253521,
+      "grad_norm": 0.34537017345428467,
+      "learning_rate": 8.798582839703634e-06,
+      "loss": 0.3851,
+      "step": 3612
+    },
+    {
+      "epoch": 0.908702213279678,
+      "grad_norm": 0.3755810260772705,
+      "learning_rate": 8.797631193901888e-06,
+      "loss": 0.3578,
+      "step": 3613
+    },
+    {
+      "epoch": 0.9089537223340041,
+      "grad_norm": 0.36311477422714233,
+      "learning_rate": 8.79667922285679e-06,
+      "loss": 0.4059,
+      "step": 3614
+    },
+    {
+      "epoch": 0.90920523138833,
+      "grad_norm": 0.3528873920440674,
+      "learning_rate": 8.795726926649867e-06,
+      "loss": 0.3788,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9094567404426559,
+      "grad_norm": 0.3925495445728302,
+      "learning_rate": 8.794774305362679e-06,
+      "loss": 0.3891,
+      "step": 3616
+    },
+    {
+      "epoch": 0.9097082494969819,
+      "grad_norm": 0.3574492931365967,
+      "learning_rate": 8.793821359076814e-06,
+      "loss": 0.4097,
+      "step": 3617
+    },
+    {
+      "epoch": 0.9099597585513078,
+      "grad_norm": 0.3412817716598511,
+      "learning_rate": 8.79286808787388e-06,
+      "loss": 0.3801,
+      "step": 3618
+    },
+    {
+      "epoch": 0.9102112676056338,
+      "grad_norm": 0.38012850284576416,
+      "learning_rate": 8.791914491835525e-06,
+      "loss": 0.3801,
+      "step": 3619
+    },
+    {
+      "epoch": 0.9104627766599598,
+      "grad_norm": 0.3861580789089203,
+      "learning_rate": 8.790960571043416e-06,
+      "loss": 0.4157,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9107142857142857,
+      "grad_norm": 0.36277568340301514,
+      "learning_rate": 8.79000632557925e-06,
+      "loss": 0.3985,
+      "step": 3621
+    },
+    {
+      "epoch": 0.9109657947686117,
+      "grad_norm": 0.39433610439300537,
+      "learning_rate": 8.789051755524752e-06,
+      "loss": 0.4042,
+      "step": 3622
+    },
+    {
+      "epoch": 0.9112173038229376,
+      "grad_norm": 0.34599095582962036,
+      "learning_rate": 8.788096860961674e-06,
+      "loss": 0.3646,
+      "step": 3623
+    },
+    {
+      "epoch": 0.9114688128772636,
+      "grad_norm": 0.35807907581329346,
+      "learning_rate": 8.7871416419718e-06,
+      "loss": 0.3742,
+      "step": 3624
+    },
+    {
+      "epoch": 0.9117203219315896,
+      "grad_norm": 0.4068434536457062,
+      "learning_rate": 8.786186098636935e-06,
+      "loss": 0.3949,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9119718309859155,
+      "grad_norm": 0.34174662828445435,
+      "learning_rate": 8.78523023103892e-06,
+      "loss": 0.3641,
+      "step": 3626
+    },
+    {
+      "epoch": 0.9122233400402414,
+      "grad_norm": 0.35758674144744873,
+      "learning_rate": 8.784274039259613e-06,
+      "loss": 0.3822,
+      "step": 3627
+    },
+    {
+      "epoch": 0.9124748490945674,
+      "grad_norm": 0.3813161849975586,
+      "learning_rate": 8.78331752338091e-06,
+      "loss": 0.398,
+      "step": 3628
+    },
+    {
+      "epoch": 0.9127263581488934,
+      "grad_norm": 0.37230536341667175,
+      "learning_rate": 8.78236068348473e-06,
+      "loss": 0.3727,
+      "step": 3629
+    },
+    {
+      "epoch": 0.9129778672032193,
+      "grad_norm": 0.3861142098903656,
+      "learning_rate": 8.781403519653018e-06,
+      "loss": 0.3595,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9132293762575453,
+      "grad_norm": 0.35886576771736145,
+      "learning_rate": 8.780446031967753e-06,
+      "loss": 0.3457,
+      "step": 3631
+    },
+    {
+      "epoch": 0.9134808853118712,
+      "grad_norm": 0.4137284755706787,
+      "learning_rate": 8.779488220510935e-06,
+      "loss": 0.393,
+      "step": 3632
+    },
+    {
+      "epoch": 0.9137323943661971,
+      "grad_norm": 0.3399023115634918,
+      "learning_rate": 8.778530085364595e-06,
+      "loss": 0.3529,
+      "step": 3633
+    },
+    {
+      "epoch": 0.9139839034205232,
+      "grad_norm": 0.35265326499938965,
+      "learning_rate": 8.777571626610793e-06,
+      "loss": 0.3613,
+      "step": 3634
+    },
+    {
+      "epoch": 0.9142354124748491,
+      "grad_norm": 0.38770735263824463,
+      "learning_rate": 8.776612844331611e-06,
+      "loss": 0.3949,
+      "step": 3635
+    },
+    {
+      "epoch": 0.9144869215291751,
+      "grad_norm": 0.33681389689445496,
+      "learning_rate": 8.775653738609167e-06,
+      "loss": 0.3873,
+      "step": 3636
+    },
+    {
+      "epoch": 0.914738430583501,
+      "grad_norm": 0.36732831597328186,
+      "learning_rate": 8.7746943095256e-06,
+      "loss": 0.3771,
+      "step": 3637
+    },
+    {
+      "epoch": 0.9149899396378269,
+      "grad_norm": 0.35177841782569885,
+      "learning_rate": 8.77373455716308e-06,
+      "loss": 0.3578,
+      "step": 3638
+    },
+    {
+      "epoch": 0.915241448692153,
+      "grad_norm": 0.342661589384079,
+      "learning_rate": 8.772774481603805e-06,
+      "loss": 0.3895,
+      "step": 3639
+    },
+    {
+      "epoch": 0.9154929577464789,
+      "grad_norm": 0.3410513699054718,
+      "learning_rate": 8.771814082929997e-06,
+      "loss": 0.369,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9157444668008048,
+      "grad_norm": 0.3464511036872864,
+      "learning_rate": 8.77085336122391e-06,
+      "loss": 0.3705,
+      "step": 3641
+    },
+    {
+      "epoch": 0.9159959758551308,
+      "grad_norm": 0.35824257135391235,
+      "learning_rate": 8.76989231656782e-06,
+      "loss": 0.3974,
+      "step": 3642
+    },
+    {
+      "epoch": 0.9162474849094567,
+      "grad_norm": 0.3505759835243225,
+      "learning_rate": 8.768930949044041e-06,
+      "loss": 0.3966,
+      "step": 3643
+    },
+    {
+      "epoch": 0.9164989939637826,
+      "grad_norm": 0.347253680229187,
+      "learning_rate": 8.767969258734903e-06,
+      "loss": 0.4126,
+      "step": 3644
+    },
+    {
+      "epoch": 0.9167505030181087,
+      "grad_norm": 0.36433228850364685,
+      "learning_rate": 8.767007245722769e-06,
+      "loss": 0.3819,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9170020120724346,
+      "grad_norm": 0.34045785665512085,
+      "learning_rate": 8.766044910090033e-06,
+      "loss": 0.3647,
+      "step": 3646
+    },
+    {
+      "epoch": 0.9172535211267606,
+      "grad_norm": 0.38475969433784485,
+      "learning_rate": 8.76508225191911e-06,
+      "loss": 0.4221,
+      "step": 3647
+    },
+    {
+      "epoch": 0.9175050301810865,
+      "grad_norm": 0.3489360809326172,
+      "learning_rate": 8.764119271292446e-06,
+      "loss": 0.3675,
+      "step": 3648
+    },
+    {
+      "epoch": 0.9177565392354124,
+      "grad_norm": 0.33245396614074707,
+      "learning_rate": 8.763155968292517e-06,
+      "loss": 0.3854,
+      "step": 3649
+    },
+    {
+      "epoch": 0.9180080482897385,
+      "grad_norm": 0.3598446547985077,
+      "learning_rate": 8.762192343001818e-06,
+      "loss": 0.3766,
+      "step": 3650
+    },
+    {
+      "epoch": 0.9182595573440644,
+      "grad_norm": 0.3599037528038025,
+      "learning_rate": 8.761228395502883e-06,
+      "loss": 0.3813,
+      "step": 3651
+    },
+    {
+      "epoch": 0.9185110663983903,
+      "grad_norm": 0.3316107392311096,
+      "learning_rate": 8.760264125878266e-06,
+      "loss": 0.383,
+      "step": 3652
+    },
+    {
+      "epoch": 0.9187625754527163,
+      "grad_norm": 0.3711565136909485,
+      "learning_rate": 8.75929953421055e-06,
+      "loss": 0.3962,
+      "step": 3653
+    },
+    {
+      "epoch": 0.9190140845070423,
+      "grad_norm": 0.34968239068984985,
+      "learning_rate": 8.758334620582346e-06,
+      "loss": 0.3993,
+      "step": 3654
+    },
+    {
+      "epoch": 0.9192655935613682,
+      "grad_norm": 0.327274888753891,
+      "learning_rate": 8.757369385076296e-06,
+      "loss": 0.4015,
+      "step": 3655
+    },
+    {
+      "epoch": 0.9195171026156942,
+      "grad_norm": 0.38763725757598877,
+      "learning_rate": 8.756403827775063e-06,
+      "loss": 0.3636,
+      "step": 3656
+    },
+    {
+      "epoch": 0.9197686116700201,
+      "grad_norm": 0.3284519612789154,
+      "learning_rate": 8.755437948761344e-06,
+      "loss": 0.3446,
+      "step": 3657
+    },
+    {
+      "epoch": 0.920020120724346,
+      "grad_norm": 0.34053710103034973,
+      "learning_rate": 8.754471748117857e-06,
+      "loss": 0.3627,
+      "step": 3658
+    },
+    {
+      "epoch": 0.920271629778672,
+      "grad_norm": 0.3225838243961334,
+      "learning_rate": 8.753505225927352e-06,
+      "loss": 0.3911,
+      "step": 3659
+    },
+    {
+      "epoch": 0.920523138832998,
+      "grad_norm": 0.36908066272735596,
+      "learning_rate": 8.752538382272608e-06,
+      "loss": 0.3626,
+      "step": 3660
+    },
+    {
+      "epoch": 0.920774647887324,
+      "grad_norm": 0.3289790749549866,
+      "learning_rate": 8.751571217236426e-06,
+      "loss": 0.4104,
+      "step": 3661
+    },
+    {
+      "epoch": 0.9210261569416499,
+      "grad_norm": 0.35275307297706604,
+      "learning_rate": 8.75060373090164e-06,
+      "loss": 0.3679,
+      "step": 3662
+    },
+    {
+      "epoch": 0.9212776659959758,
+      "grad_norm": 0.381730318069458,
+      "learning_rate": 8.749635923351108e-06,
+      "loss": 0.3927,
+      "step": 3663
+    },
+    {
+      "epoch": 0.9215291750503019,
+      "grad_norm": 0.36701148748397827,
+      "learning_rate": 8.748667794667715e-06,
+      "loss": 0.3881,
+      "step": 3664
+    },
+    {
+      "epoch": 0.9217806841046278,
+      "grad_norm": 0.3680243492126465,
+      "learning_rate": 8.74769934493438e-06,
+      "loss": 0.3805,
+      "step": 3665
+    },
+    {
+      "epoch": 0.9220321931589537,
+      "grad_norm": 0.344265878200531,
+      "learning_rate": 8.74673057423404e-06,
+      "loss": 0.3879,
+      "step": 3666
+    },
+    {
+      "epoch": 0.9222837022132797,
+      "grad_norm": 0.3326249122619629,
+      "learning_rate": 8.745761482649667e-06,
+      "loss": 0.3726,
+      "step": 3667
+    },
+    {
+      "epoch": 0.9225352112676056,
+      "grad_norm": 0.3622082769870758,
+      "learning_rate": 8.744792070264254e-06,
+      "loss": 0.3893,
+      "step": 3668
+    },
+    {
+      "epoch": 0.9227867203219315,
+      "grad_norm": 0.4040569067001343,
+      "learning_rate": 8.743822337160829e-06,
+      "loss": 0.3559,
+      "step": 3669
+    },
+    {
+      "epoch": 0.9230382293762576,
+      "grad_norm": 0.34916210174560547,
+      "learning_rate": 8.742852283422443e-06,
+      "loss": 0.3637,
+      "step": 3670
+    },
+    {
+      "epoch": 0.9232897384305835,
+      "grad_norm": 0.34636667370796204,
+      "learning_rate": 8.741881909132171e-06,
+      "loss": 0.3768,
+      "step": 3671
+    },
+    {
+      "epoch": 0.9235412474849095,
+      "grad_norm": 0.344542533159256,
+      "learning_rate": 8.740911214373125e-06,
+      "loss": 0.3727,
+      "step": 3672
+    },
+    {
+      "epoch": 0.9237927565392354,
+      "grad_norm": 0.37723875045776367,
+      "learning_rate": 8.739940199228436e-06,
+      "loss": 0.3802,
+      "step": 3673
+    },
+    {
+      "epoch": 0.9240442655935613,
+      "grad_norm": 0.3496474027633667,
+      "learning_rate": 8.738968863781267e-06,
+      "loss": 0.3946,
+      "step": 3674
+    },
+    {
+      "epoch": 0.9242957746478874,
+      "grad_norm": 0.31747496128082275,
+      "learning_rate": 8.737997208114806e-06,
+      "loss": 0.3364,
+      "step": 3675
+    },
+    {
+      "epoch": 0.9245472837022133,
+      "grad_norm": 0.36128780245780945,
+      "learning_rate": 8.737025232312267e-06,
+      "loss": 0.3646,
+      "step": 3676
+    },
+    {
+      "epoch": 0.9247987927565392,
+      "grad_norm": 0.3360286056995392,
+      "learning_rate": 8.736052936456897e-06,
+      "loss": 0.3604,
+      "step": 3677
+    },
+    {
+      "epoch": 0.9250503018108652,
+      "grad_norm": 0.37256118655204773,
+      "learning_rate": 8.735080320631966e-06,
+      "loss": 0.3742,
+      "step": 3678
+    },
+    {
+      "epoch": 0.9253018108651911,
+      "grad_norm": 0.3662985861301422,
+      "learning_rate": 8.734107384920771e-06,
+      "loss": 0.4107,
+      "step": 3679
+    },
+    {
+      "epoch": 0.9255533199195171,
+      "grad_norm": 0.37539082765579224,
+      "learning_rate": 8.733134129406638e-06,
+      "loss": 0.3769,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9258048289738431,
+      "grad_norm": 0.379936546087265,
+      "learning_rate": 8.732160554172923e-06,
+      "loss": 0.3548,
+      "step": 3681
+    },
+    {
+      "epoch": 0.926056338028169,
+      "grad_norm": 0.3667657673358917,
+      "learning_rate": 8.731186659303004e-06,
+      "loss": 0.3804,
+      "step": 3682
+    },
+    {
+      "epoch": 0.9263078470824949,
+      "grad_norm": 0.374489963054657,
+      "learning_rate": 8.73021244488029e-06,
+      "loss": 0.3722,
+      "step": 3683
+    },
+    {
+      "epoch": 0.9265593561368209,
+      "grad_norm": 0.37897610664367676,
+      "learning_rate": 8.729237910988218e-06,
+      "loss": 0.4052,
+      "step": 3684
+    },
+    {
+      "epoch": 0.9268108651911469,
+      "grad_norm": 0.332736611366272,
+      "learning_rate": 8.728263057710247e-06,
+      "loss": 0.3906,
+      "step": 3685
+    },
+    {
+      "epoch": 0.9270623742454729,
+      "grad_norm": 0.37332481145858765,
+      "learning_rate": 8.72728788512987e-06,
+      "loss": 0.3862,
+      "step": 3686
+    },
+    {
+      "epoch": 0.9273138832997988,
+      "grad_norm": 0.33331403136253357,
+      "learning_rate": 8.726312393330602e-06,
+      "loss": 0.3965,
+      "step": 3687
+    },
+    {
+      "epoch": 0.9275653923541247,
+      "grad_norm": 0.3460955321788788,
+      "learning_rate": 8.72533658239599e-06,
+      "loss": 0.3946,
+      "step": 3688
+    },
+    {
+      "epoch": 0.9278169014084507,
+      "grad_norm": 0.3659730553627014,
+      "learning_rate": 8.724360452409606e-06,
+      "loss": 0.3875,
+      "step": 3689
+    },
+    {
+      "epoch": 0.9280684104627767,
+      "grad_norm": 0.33877822756767273,
+      "learning_rate": 8.723384003455049e-06,
+      "loss": 0.3761,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9283199195171026,
+      "grad_norm": 0.34681734442710876,
+      "learning_rate": 8.722407235615944e-06,
+      "loss": 0.3584,
+      "step": 3691
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 0.3557383716106415,
+      "learning_rate": 8.721430148975946e-06,
+      "loss": 0.39,
+      "step": 3692
+    },
+    {
+      "epoch": 0.9288229376257545,
+      "grad_norm": 0.3659672141075134,
+      "learning_rate": 8.72045274361874e-06,
+      "loss": 0.3935,
+      "step": 3693
+    },
+    {
+      "epoch": 0.9290744466800804,
+      "grad_norm": 0.3567541241645813,
+      "learning_rate": 8.71947501962803e-06,
+      "loss": 0.3621,
+      "step": 3694
+    },
+    {
+      "epoch": 0.9293259557344065,
+      "grad_norm": 0.3142370581626892,
+      "learning_rate": 8.718496977087554e-06,
+      "loss": 0.3623,
+      "step": 3695
+    },
+    {
+      "epoch": 0.9295774647887324,
+      "grad_norm": 0.3656524121761322,
+      "learning_rate": 8.717518616081073e-06,
+      "loss": 0.3803,
+      "step": 3696
+    },
+    {
+      "epoch": 0.9298289738430584,
+      "grad_norm": 0.3680150508880615,
+      "learning_rate": 8.716539936692381e-06,
+      "loss": 0.3862,
+      "step": 3697
+    },
+    {
+      "epoch": 0.9300804828973843,
+      "grad_norm": 0.32054194808006287,
+      "learning_rate": 8.715560939005293e-06,
+      "loss": 0.401,
+      "step": 3698
+    },
+    {
+      "epoch": 0.9303319919517102,
+      "grad_norm": 0.3696306049823761,
+      "learning_rate": 8.714581623103654e-06,
+      "loss": 0.3615,
+      "step": 3699
+    },
+    {
+      "epoch": 0.9305835010060363,
+      "grad_norm": 0.3776056170463562,
+      "learning_rate": 8.713601989071342e-06,
+      "loss": 0.3761,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9308350100603622,
+      "grad_norm": 0.3336465656757355,
+      "learning_rate": 8.712622036992248e-06,
+      "loss": 0.3806,
+      "step": 3701
+    },
+    {
+      "epoch": 0.9310865191146881,
+      "grad_norm": 0.3547345995903015,
+      "learning_rate": 8.711641766950302e-06,
+      "loss": 0.3758,
+      "step": 3702
+    },
+    {
+      "epoch": 0.9313380281690141,
+      "grad_norm": 0.3893287181854248,
+      "learning_rate": 8.710661179029461e-06,
+      "loss": 0.3892,
+      "step": 3703
+    },
+    {
+      "epoch": 0.93158953722334,
+      "grad_norm": 0.36249130964279175,
+      "learning_rate": 8.709680273313703e-06,
+      "loss": 0.3726,
+      "step": 3704
+    },
+    {
+      "epoch": 0.931841046277666,
+      "grad_norm": 0.35625818371772766,
+      "learning_rate": 8.708699049887038e-06,
+      "loss": 0.3782,
+      "step": 3705
+    },
+    {
+      "epoch": 0.932092555331992,
+      "grad_norm": 0.3533684313297272,
+      "learning_rate": 8.707717508833499e-06,
+      "loss": 0.4038,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9323440643863179,
+      "grad_norm": 0.3416036367416382,
+      "learning_rate": 8.706735650237153e-06,
+      "loss": 0.3865,
+      "step": 3707
+    },
+    {
+      "epoch": 0.9325955734406438,
+      "grad_norm": 0.3354741036891937,
+      "learning_rate": 8.705753474182085e-06,
+      "loss": 0.3464,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9328470824949698,
+      "grad_norm": 0.37271448969841003,
+      "learning_rate": 8.704770980752417e-06,
+      "loss": 0.3944,
+      "step": 3709
+    },
+    {
+      "epoch": 0.9330985915492958,
+      "grad_norm": 0.345333993434906,
+      "learning_rate": 8.703788170032293e-06,
+      "loss": 0.3661,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9333501006036218,
+      "grad_norm": 0.39637884497642517,
+      "learning_rate": 8.70280504210588e-06,
+      "loss": 0.3813,
+      "step": 3711
+    },
+    {
+      "epoch": 0.9336016096579477,
+      "grad_norm": 0.35547831654548645,
+      "learning_rate": 8.701821597057381e-06,
+      "loss": 0.4037,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9338531187122736,
+      "grad_norm": 0.3298320472240448,
+      "learning_rate": 8.700837834971021e-06,
+      "loss": 0.3871,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9341046277665996,
+      "grad_norm": 0.3607611358165741,
+      "learning_rate": 8.699853755931053e-06,
+      "loss": 0.3879,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9343561368209256,
+      "grad_norm": 0.3281857967376709,
+      "learning_rate": 8.698869360021755e-06,
+      "loss": 0.3806,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9346076458752515,
+      "grad_norm": 0.3486366271972656,
+      "learning_rate": 8.697884647327438e-06,
+      "loss": 0.3984,
+      "step": 3716
+    },
+    {
+      "epoch": 0.9348591549295775,
+      "grad_norm": 0.32354462146759033,
+      "learning_rate": 8.696899617932436e-06,
+      "loss": 0.3889,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9351106639839034,
+      "grad_norm": 0.3479534387588501,
+      "learning_rate": 8.69591427192111e-06,
+      "loss": 0.3921,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9353621730382293,
+      "grad_norm": 0.3232344090938568,
+      "learning_rate": 8.694928609377844e-06,
+      "loss": 0.3755,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9356136820925554,
+      "grad_norm": 0.35283714532852173,
+      "learning_rate": 8.69394263038706e-06,
+      "loss": 0.3604,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9358651911468813,
+      "grad_norm": 0.35240834951400757,
+      "learning_rate": 8.6929563350332e-06,
+      "loss": 0.3754,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9361167002012073,
+      "grad_norm": 0.34090203046798706,
+      "learning_rate": 8.691969723400732e-06,
+      "loss": 0.4053,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9363682092555332,
+      "grad_norm": 0.389652281999588,
+      "learning_rate": 8.690982795574155e-06,
+      "loss": 0.3918,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9366197183098591,
+      "grad_norm": 0.3855276107788086,
+      "learning_rate": 8.689995551637992e-06,
+      "loss": 0.3933,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9368712273641852,
+      "grad_norm": 0.3753201365470886,
+      "learning_rate": 8.689007991676795e-06,
+      "loss": 0.37,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9371227364185111,
+      "grad_norm": 0.34677380323410034,
+      "learning_rate": 8.68802011577514e-06,
+      "loss": 0.3988,
+      "step": 3726
+    },
+    {
+      "epoch": 0.937374245472837,
+      "grad_norm": 0.3636135458946228,
+      "learning_rate": 8.687031924017635e-06,
+      "loss": 0.3825,
+      "step": 3727
+    },
+    {
+      "epoch": 0.937625754527163,
+      "grad_norm": 0.3638603389263153,
+      "learning_rate": 8.686043416488913e-06,
+      "loss": 0.3959,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9378772635814889,
+      "grad_norm": 0.40067726373672485,
+      "learning_rate": 8.685054593273631e-06,
+      "loss": 0.3546,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9381287726358148,
+      "grad_norm": 0.34523940086364746,
+      "learning_rate": 8.684065454456478e-06,
+      "loss": 0.3723,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9383802816901409,
+      "grad_norm": 0.3261178135871887,
+      "learning_rate": 8.683076000122165e-06,
+      "loss": 0.3744,
+      "step": 3731
+    },
+    {
+      "epoch": 0.9386317907444668,
+      "grad_norm": 0.4006332457065582,
+      "learning_rate": 8.682086230355432e-06,
+      "loss": 0.357,
+      "step": 3732
+    },
+    {
+      "epoch": 0.9388832997987927,
+      "grad_norm": 0.33518972992897034,
+      "learning_rate": 8.681096145241053e-06,
+      "loss": 0.3654,
+      "step": 3733
+    },
+    {
+      "epoch": 0.9391348088531187,
+      "grad_norm": 0.3528534770011902,
+      "learning_rate": 8.680105744863817e-06,
+      "loss": 0.3879,
+      "step": 3734
+    },
+    {
+      "epoch": 0.9393863179074446,
+      "grad_norm": 0.3312220573425293,
+      "learning_rate": 8.679115029308543e-06,
+      "loss": 0.3973,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9396378269617707,
+      "grad_norm": 0.38453420996665955,
+      "learning_rate": 8.678123998660087e-06,
+      "loss": 0.3724,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9398893360160966,
+      "grad_norm": 0.3209126889705658,
+      "learning_rate": 8.677132653003318e-06,
+      "loss": 0.3678,
+      "step": 3737
+    },
+    {
+      "epoch": 0.9401408450704225,
+      "grad_norm": 0.39340800046920776,
+      "learning_rate": 8.676140992423143e-06,
+      "loss": 0.3451,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9403923541247485,
+      "grad_norm": 0.38696199655532837,
+      "learning_rate": 8.67514901700449e-06,
+      "loss": 0.4004,
+      "step": 3739
+    },
+    {
+      "epoch": 0.9406438631790744,
+      "grad_norm": 0.35889217257499695,
+      "learning_rate": 8.674156726832315e-06,
+      "loss": 0.3793,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9408953722334004,
+      "grad_norm": 0.403751015663147,
+      "learning_rate": 8.673164121991601e-06,
+      "loss": 0.4051,
+      "step": 3741
+    },
+    {
+      "epoch": 0.9411468812877264,
+      "grad_norm": 0.43993040919303894,
+      "learning_rate": 8.672171202567359e-06,
+      "loss": 0.3685,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9413983903420523,
+      "grad_norm": 0.3885231614112854,
+      "learning_rate": 8.671177968644628e-06,
+      "loss": 0.3809,
+      "step": 3743
+    },
+    {
+      "epoch": 0.9416498993963782,
+      "grad_norm": 0.3906707167625427,
+      "learning_rate": 8.67018442030847e-06,
+      "loss": 0.3877,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9419014084507042,
+      "grad_norm": 0.35688477754592896,
+      "learning_rate": 8.669190557643977e-06,
+      "loss": 0.3578,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9421529175050302,
+      "grad_norm": 0.37239667773246765,
+      "learning_rate": 8.668196380736267e-06,
+      "loss": 0.3765,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9424044265593562,
+      "grad_norm": 0.38432252407073975,
+      "learning_rate": 8.667201889670485e-06,
+      "loss": 0.375,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9426559356136821,
+      "grad_norm": 0.36130064725875854,
+      "learning_rate": 8.666207084531804e-06,
+      "loss": 0.3808,
+      "step": 3748
+    },
+    {
+      "epoch": 0.942907444668008,
+      "grad_norm": 0.34920212626457214,
+      "learning_rate": 8.665211965405422e-06,
+      "loss": 0.3584,
+      "step": 3749
+    },
+    {
+      "epoch": 0.943158953722334,
+      "grad_norm": 0.34862497448921204,
+      "learning_rate": 8.664216532376563e-06,
+      "loss": 0.3585,
+      "step": 3750
+    },
+    {
+      "epoch": 0.94341046277666,
+      "grad_norm": 0.35228458046913147,
+      "learning_rate": 8.663220785530485e-06,
+      "loss": 0.3725,
+      "step": 3751
+    },
+    {
+      "epoch": 0.9436619718309859,
+      "grad_norm": 0.3779107928276062,
+      "learning_rate": 8.662224724952459e-06,
+      "loss": 0.3687,
+      "step": 3752
+    },
+    {
+      "epoch": 0.9439134808853119,
+      "grad_norm": 0.36739468574523926,
+      "learning_rate": 8.661228350727798e-06,
+      "loss": 0.3429,
+      "step": 3753
+    },
+    {
+      "epoch": 0.9441649899396378,
+      "grad_norm": 0.381713330745697,
+      "learning_rate": 8.660231662941834e-06,
+      "loss": 0.3805,
+      "step": 3754
+    },
+    {
+      "epoch": 0.9444164989939637,
+      "grad_norm": 0.35245978832244873,
+      "learning_rate": 8.659234661679926e-06,
+      "loss": 0.3795,
+      "step": 3755
+    },
+    {
+      "epoch": 0.9446680080482898,
+      "grad_norm": 0.3956359028816223,
+      "learning_rate": 8.658237347027461e-06,
+      "loss": 0.3569,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9449195171026157,
+      "grad_norm": 0.3203680217266083,
+      "learning_rate": 8.657239719069854e-06,
+      "loss": 0.386,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9451710261569416,
+      "grad_norm": 0.3383410573005676,
+      "learning_rate": 8.656241777892544e-06,
+      "loss": 0.3816,
+      "step": 3758
+    },
+    {
+      "epoch": 0.9454225352112676,
+      "grad_norm": 0.3426474928855896,
+      "learning_rate": 8.655243523580998e-06,
+      "loss": 0.3848,
+      "step": 3759
+    },
+    {
+      "epoch": 0.9456740442655935,
+      "grad_norm": 0.3572126030921936,
+      "learning_rate": 8.654244956220713e-06,
+      "loss": 0.3666,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9459255533199196,
+      "grad_norm": 0.3556264340877533,
+      "learning_rate": 8.653246075897208e-06,
+      "loss": 0.4032,
+      "step": 3761
+    },
+    {
+      "epoch": 0.9461770623742455,
+      "grad_norm": 0.3631819486618042,
+      "learning_rate": 8.652246882696032e-06,
+      "loss": 0.3807,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9464285714285714,
+      "grad_norm": 0.36357030272483826,
+      "learning_rate": 8.651247376702756e-06,
+      "loss": 0.3523,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9466800804828974,
+      "grad_norm": 0.34328320622444153,
+      "learning_rate": 8.650247558002987e-06,
+      "loss": 0.3881,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9469315895372233,
+      "grad_norm": 0.349572092294693,
+      "learning_rate": 8.64924742668235e-06,
+      "loss": 0.3589,
+      "step": 3765
+    },
+    {
+      "epoch": 0.9471830985915493,
+      "grad_norm": 0.3555934429168701,
+      "learning_rate": 8.6482469828265e-06,
+      "loss": 0.3848,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9474346076458753,
+      "grad_norm": 0.32735902070999146,
+      "learning_rate": 8.64724622652112e-06,
+      "loss": 0.3698,
+      "step": 3767
+    },
+    {
+      "epoch": 0.9476861167002012,
+      "grad_norm": 0.35206887125968933,
+      "learning_rate": 8.646245157851918e-06,
+      "loss": 0.3755,
+      "step": 3768
+    },
+    {
+      "epoch": 0.9479376257545271,
+      "grad_norm": 0.34176838397979736,
+      "learning_rate": 8.645243776904629e-06,
+      "loss": 0.3976,
+      "step": 3769
+    },
+    {
+      "epoch": 0.9481891348088531,
+      "grad_norm": 0.3586902320384979,
+      "learning_rate": 8.644242083765014e-06,
+      "loss": 0.3693,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9484406438631791,
+      "grad_norm": 0.33086657524108887,
+      "learning_rate": 8.643240078518865e-06,
+      "loss": 0.3722,
+      "step": 3771
+    },
+    {
+      "epoch": 0.9486921529175051,
+      "grad_norm": 0.384689062833786,
+      "learning_rate": 8.642237761251992e-06,
+      "loss": 0.3484,
+      "step": 3772
+    },
+    {
+      "epoch": 0.948943661971831,
+      "grad_norm": 0.3711892366409302,
+      "learning_rate": 8.641235132050243e-06,
+      "loss": 0.3786,
+      "step": 3773
+    },
+    {
+      "epoch": 0.9491951710261569,
+      "grad_norm": 0.325150728225708,
+      "learning_rate": 8.640232190999484e-06,
+      "loss": 0.4007,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9494466800804829,
+      "grad_norm": 0.3467983603477478,
+      "learning_rate": 8.63922893818561e-06,
+      "loss": 0.367,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9496981891348089,
+      "grad_norm": 0.40407732129096985,
+      "learning_rate": 8.638225373694546e-06,
+      "loss": 0.374,
+      "step": 3776
+    },
+    {
+      "epoch": 0.9499496981891348,
+      "grad_norm": 0.3339010775089264,
+      "learning_rate": 8.637221497612238e-06,
+      "loss": 0.3786,
+      "step": 3777
+    },
+    {
+      "epoch": 0.9502012072434608,
+      "grad_norm": 0.33294668793678284,
+      "learning_rate": 8.636217310024664e-06,
+      "loss": 0.4051,
+      "step": 3778
+    },
+    {
+      "epoch": 0.9504527162977867,
+      "grad_norm": 0.32620373368263245,
+      "learning_rate": 8.635212811017826e-06,
+      "loss": 0.3619,
+      "step": 3779
+    },
+    {
+      "epoch": 0.9507042253521126,
+      "grad_norm": 0.3283163011074066,
+      "learning_rate": 8.634208000677751e-06,
+      "loss": 0.3663,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9509557344064387,
+      "grad_norm": 0.3436046838760376,
+      "learning_rate": 8.633202879090496e-06,
+      "loss": 0.3836,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9512072434607646,
+      "grad_norm": 0.3591076731681824,
+      "learning_rate": 8.632197446342145e-06,
+      "loss": 0.3852,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9514587525150905,
+      "grad_norm": 0.3302284777164459,
+      "learning_rate": 8.631191702518806e-06,
+      "loss": 0.4113,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9517102615694165,
+      "grad_norm": 0.3252635896205902,
+      "learning_rate": 8.630185647706614e-06,
+      "loss": 0.3768,
+      "step": 3784
+    },
+    {
+      "epoch": 0.9519617706237424,
+      "grad_norm": 0.3460747003555298,
+      "learning_rate": 8.629179281991732e-06,
+      "loss": 0.4131,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9522132796780685,
+      "grad_norm": 0.3621247112751007,
+      "learning_rate": 8.628172605460347e-06,
+      "loss": 0.393,
+      "step": 3786
+    },
+    {
+      "epoch": 0.9524647887323944,
+      "grad_norm": 0.3651982843875885,
+      "learning_rate": 8.627165618198676e-06,
+      "loss": 0.3871,
+      "step": 3787
+    },
+    {
+      "epoch": 0.9527162977867203,
+      "grad_norm": 0.3419380784034729,
+      "learning_rate": 8.626158320292963e-06,
+      "loss": 0.3815,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9529678068410463,
+      "grad_norm": 0.3466692268848419,
+      "learning_rate": 8.625150711829475e-06,
+      "loss": 0.3587,
+      "step": 3789
+    },
+    {
+      "epoch": 0.9532193158953722,
+      "grad_norm": 0.3754306435585022,
+      "learning_rate": 8.624142792894505e-06,
+      "loss": 0.4,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9534708249496981,
+      "grad_norm": 0.348619669675827,
+      "learning_rate": 8.62313456357438e-06,
+      "loss": 0.3605,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9537223340040242,
+      "grad_norm": 0.34513843059539795,
+      "learning_rate": 8.622126023955446e-06,
+      "loss": 0.3467,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9539738430583501,
+      "grad_norm": 0.3718356788158417,
+      "learning_rate": 8.621117174124076e-06,
+      "loss": 0.379,
+      "step": 3793
+    },
+    {
+      "epoch": 0.954225352112676,
+      "grad_norm": 0.38062798976898193,
+      "learning_rate": 8.620108014166674e-06,
+      "loss": 0.3619,
+      "step": 3794
+    },
+    {
+      "epoch": 0.954476861167002,
+      "grad_norm": 0.3671969473361969,
+      "learning_rate": 8.619098544169671e-06,
+      "loss": 0.3721,
+      "step": 3795
+    },
+    {
+      "epoch": 0.954728370221328,
+      "grad_norm": 0.40202799439430237,
+      "learning_rate": 8.618088764219514e-06,
+      "loss": 0.3779,
+      "step": 3796
+    },
+    {
+      "epoch": 0.954979879275654,
+      "grad_norm": 0.36368417739868164,
+      "learning_rate": 8.617078674402692e-06,
+      "loss": 0.3934,
+      "step": 3797
+    },
+    {
+      "epoch": 0.9552313883299799,
+      "grad_norm": 0.42696356773376465,
+      "learning_rate": 8.616068274805709e-06,
+      "loss": 0.3764,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9554828973843058,
+      "grad_norm": 0.3894241750240326,
+      "learning_rate": 8.615057565515102e-06,
+      "loss": 0.3658,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9557344064386318,
+      "grad_norm": 0.38357433676719666,
+      "learning_rate": 8.614046546617427e-06,
+      "loss": 0.3971,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9559859154929577,
+      "grad_norm": 0.3617189824581146,
+      "learning_rate": 8.613035218199276e-06,
+      "loss": 0.3809,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9562374245472837,
+      "grad_norm": 0.3896433711051941,
+      "learning_rate": 8.612023580347264e-06,
+      "loss": 0.3586,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9564889336016097,
+      "grad_norm": 0.42416492104530334,
+      "learning_rate": 8.611011633148027e-06,
+      "loss": 0.4036,
+      "step": 3803
+    },
+    {
+      "epoch": 0.9567404426559356,
+      "grad_norm": 0.35320156812667847,
+      "learning_rate": 8.609999376688235e-06,
+      "loss": 0.3765,
+      "step": 3804
+    },
+    {
+      "epoch": 0.9569919517102615,
+      "grad_norm": 0.37311264872550964,
+      "learning_rate": 8.60898681105458e-06,
+      "loss": 0.3923,
+      "step": 3805
+    },
+    {
+      "epoch": 0.9572434607645876,
+      "grad_norm": 0.3776063621044159,
+      "learning_rate": 8.607973936333782e-06,
+      "loss": 0.3784,
+      "step": 3806
+    },
+    {
+      "epoch": 0.9574949698189135,
+      "grad_norm": 0.34867435693740845,
+      "learning_rate": 8.606960752612587e-06,
+      "loss": 0.3768,
+      "step": 3807
+    },
+    {
+      "epoch": 0.9577464788732394,
+      "grad_norm": 0.3809923529624939,
+      "learning_rate": 8.60594725997777e-06,
+      "loss": 0.3824,
+      "step": 3808
+    },
+    {
+      "epoch": 0.9579979879275654,
+      "grad_norm": 0.352938711643219,
+      "learning_rate": 8.604933458516129e-06,
+      "loss": 0.403,
+      "step": 3809
+    },
+    {
+      "epoch": 0.9582494969818913,
+      "grad_norm": 0.3831866681575775,
+      "learning_rate": 8.603919348314487e-06,
+      "loss": 0.37,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9585010060362174,
+      "grad_norm": 0.3907982110977173,
+      "learning_rate": 8.602904929459702e-06,
+      "loss": 0.3907,
+      "step": 3811
+    },
+    {
+      "epoch": 0.9587525150905433,
+      "grad_norm": 0.3331465423107147,
+      "learning_rate": 8.601890202038648e-06,
+      "loss": 0.3946,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9590040241448692,
+      "grad_norm": 0.3363270163536072,
+      "learning_rate": 8.600875166138232e-06,
+      "loss": 0.375,
+      "step": 3813
+    },
+    {
+      "epoch": 0.9592555331991952,
+      "grad_norm": 0.34430885314941406,
+      "learning_rate": 8.599859821845386e-06,
+      "loss": 0.3892,
+      "step": 3814
+    },
+    {
+      "epoch": 0.9595070422535211,
+      "grad_norm": 0.35445672273635864,
+      "learning_rate": 8.598844169247064e-06,
+      "loss": 0.3733,
+      "step": 3815
+    },
+    {
+      "epoch": 0.959758551307847,
+      "grad_norm": 0.3446792960166931,
+      "learning_rate": 8.597828208430257e-06,
+      "loss": 0.3627,
+      "step": 3816
+    },
+    {
+      "epoch": 0.9600100603621731,
+      "grad_norm": 0.39067867398262024,
+      "learning_rate": 8.596811939481971e-06,
+      "loss": 0.3893,
+      "step": 3817
+    },
+    {
+      "epoch": 0.960261569416499,
+      "grad_norm": 0.373501181602478,
+      "learning_rate": 8.59579536248924e-06,
+      "loss": 0.3754,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9605130784708249,
+      "grad_norm": 0.33692559599876404,
+      "learning_rate": 8.594778477539136e-06,
+      "loss": 0.3615,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9607645875251509,
+      "grad_norm": 0.38633298873901367,
+      "learning_rate": 8.593761284718742e-06,
+      "loss": 0.3761,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9610160965794768,
+      "grad_norm": 0.37641310691833496,
+      "learning_rate": 8.592743784115178e-06,
+      "loss": 0.3895,
+      "step": 3821
+    },
+    {
+      "epoch": 0.9612676056338029,
+      "grad_norm": 0.34640181064605713,
+      "learning_rate": 8.591725975815584e-06,
+      "loss": 0.3954,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9615191146881288,
+      "grad_norm": 0.3677440583705902,
+      "learning_rate": 8.59070785990713e-06,
+      "loss": 0.3547,
+      "step": 3823
+    },
+    {
+      "epoch": 0.9617706237424547,
+      "grad_norm": 0.3272249102592468,
+      "learning_rate": 8.589689436477011e-06,
+      "loss": 0.3579,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9620221327967807,
+      "grad_norm": 0.3785829544067383,
+      "learning_rate": 8.58867070561245e-06,
+      "loss": 0.3946,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9622736418511066,
+      "grad_norm": 0.36952534317970276,
+      "learning_rate": 8.587651667400692e-06,
+      "loss": 0.3967,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9625251509054326,
+      "grad_norm": 0.3846037983894348,
+      "learning_rate": 8.586632321929013e-06,
+      "loss": 0.3898,
+      "step": 3827
+    },
+    {
+      "epoch": 0.9627766599597586,
+      "grad_norm": 0.38403981924057007,
+      "learning_rate": 8.585612669284715e-06,
+      "loss": 0.3846,
+      "step": 3828
+    },
+    {
+      "epoch": 0.9630281690140845,
+      "grad_norm": 0.3689204156398773,
+      "learning_rate": 8.584592709555125e-06,
+      "loss": 0.3439,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9632796780684104,
+      "grad_norm": 0.34798160195350647,
+      "learning_rate": 8.58357244282759e-06,
+      "loss": 0.372,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9635311871227364,
+      "grad_norm": 0.4079124629497528,
+      "learning_rate": 8.582551869189497e-06,
+      "loss": 0.3902,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9637826961770624,
+      "grad_norm": 0.4098033010959625,
+      "learning_rate": 8.581530988728249e-06,
+      "loss": 0.3796,
+      "step": 3832
+    },
+    {
+      "epoch": 0.9640342052313883,
+      "grad_norm": 0.3369835615158081,
+      "learning_rate": 8.580509801531276e-06,
+      "loss": 0.3445,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9642857142857143,
+      "grad_norm": 0.3969493806362152,
+      "learning_rate": 8.57948830768604e-06,
+      "loss": 0.3782,
+      "step": 3834
+    },
+    {
+      "epoch": 0.9645372233400402,
+      "grad_norm": 0.4460330903530121,
+      "learning_rate": 8.57846650728002e-06,
+      "loss": 0.3792,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9647887323943662,
+      "grad_norm": 0.3550911843776703,
+      "learning_rate": 8.577444400400733e-06,
+      "loss": 0.3809,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9650402414486922,
+      "grad_norm": 0.4082079827785492,
+      "learning_rate": 8.576421987135716e-06,
+      "loss": 0.4033,
+      "step": 3837
+    },
+    {
+      "epoch": 0.9652917505030181,
+      "grad_norm": 0.40458518266677856,
+      "learning_rate": 8.575399267572527e-06,
+      "loss": 0.3572,
+      "step": 3838
+    },
+    {
+      "epoch": 0.9655432595573441,
+      "grad_norm": 0.3640364110469818,
+      "learning_rate": 8.574376241798758e-06,
+      "loss": 0.3959,
+      "step": 3839
+    },
+    {
+      "epoch": 0.96579476861167,
+      "grad_norm": 0.3785410523414612,
+      "learning_rate": 8.573352909902027e-06,
+      "loss": 0.3834,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9660462776659959,
+      "grad_norm": 0.35065385699272156,
+      "learning_rate": 8.572329271969972e-06,
+      "loss": 0.3953,
+      "step": 3841
+    },
+    {
+      "epoch": 0.966297786720322,
+      "grad_norm": 0.3185063600540161,
+      "learning_rate": 8.571305328090264e-06,
+      "loss": 0.3784,
+      "step": 3842
+    },
+    {
+      "epoch": 0.9665492957746479,
+      "grad_norm": 0.3544987142086029,
+      "learning_rate": 8.570281078350598e-06,
+      "loss": 0.3713,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9668008048289738,
+      "grad_norm": 0.3804239332675934,
+      "learning_rate": 8.569256522838692e-06,
+      "loss": 0.3862,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9670523138832998,
+      "grad_norm": 0.32411083579063416,
+      "learning_rate": 8.568231661642294e-06,
+      "loss": 0.364,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9673038229376257,
+      "grad_norm": 0.3537362515926361,
+      "learning_rate": 8.567206494849178e-06,
+      "loss": 0.3677,
+      "step": 3846
+    },
+    {
+      "epoch": 0.9675553319919518,
+      "grad_norm": 0.3677324056625366,
+      "learning_rate": 8.56618102254714e-06,
+      "loss": 0.4157,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9678068410462777,
+      "grad_norm": 0.35382696986198425,
+      "learning_rate": 8.56515524482401e-06,
+      "loss": 0.3653,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9680583501006036,
+      "grad_norm": 0.35097870230674744,
+      "learning_rate": 8.564129161767636e-06,
+      "loss": 0.3956,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9683098591549296,
+      "grad_norm": 0.35212019085884094,
+      "learning_rate": 8.563102773465894e-06,
+      "loss": 0.3872,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9685613682092555,
+      "grad_norm": 0.36232250928878784,
+      "learning_rate": 8.562076080006693e-06,
+      "loss": 0.3922,
+      "step": 3851
+    },
+    {
+      "epoch": 0.9688128772635815,
+      "grad_norm": 0.33010706305503845,
+      "learning_rate": 8.561049081477958e-06,
+      "loss": 0.3726,
+      "step": 3852
+    },
+    {
+      "epoch": 0.9690643863179075,
+      "grad_norm": 0.3573167026042938,
+      "learning_rate": 8.56002177796765e-06,
+      "loss": 0.3629,
+      "step": 3853
+    },
+    {
+      "epoch": 0.9693158953722334,
+      "grad_norm": 0.372215211391449,
+      "learning_rate": 8.558994169563745e-06,
+      "loss": 0.3706,
+      "step": 3854
+    },
+    {
+      "epoch": 0.9695674044265593,
+      "grad_norm": 0.33711981773376465,
+      "learning_rate": 8.557966256354256e-06,
+      "loss": 0.3661,
+      "step": 3855
+    },
+    {
+      "epoch": 0.9698189134808853,
+      "grad_norm": 0.37138739228248596,
+      "learning_rate": 8.556938038427217e-06,
+      "loss": 0.3524,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9700704225352113,
+      "grad_norm": 0.34503763914108276,
+      "learning_rate": 8.555909515870683e-06,
+      "loss": 0.3523,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9703219315895373,
+      "grad_norm": 0.3223390281200409,
+      "learning_rate": 8.55488068877275e-06,
+      "loss": 0.3686,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9705734406438632,
+      "grad_norm": 0.3484128713607788,
+      "learning_rate": 8.553851557221521e-06,
+      "loss": 0.3785,
+      "step": 3859
+    },
+    {
+      "epoch": 0.9708249496981891,
+      "grad_norm": 0.36130252480506897,
+      "learning_rate": 8.552822121305139e-06,
+      "loss": 0.408,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9710764587525151,
+      "grad_norm": 0.3652491569519043,
+      "learning_rate": 8.551792381111771e-06,
+      "loss": 0.3706,
+      "step": 3861
+    },
+    {
+      "epoch": 0.971327967806841,
+      "grad_norm": 0.3538327217102051,
+      "learning_rate": 8.550762336729605e-06,
+      "loss": 0.3522,
+      "step": 3862
+    },
+    {
+      "epoch": 0.971579476861167,
+      "grad_norm": 0.35552993416786194,
+      "learning_rate": 8.549731988246858e-06,
+      "loss": 0.3924,
+      "step": 3863
+    },
+    {
+      "epoch": 0.971830985915493,
+      "grad_norm": 0.3610166013240814,
+      "learning_rate": 8.548701335751774e-06,
+      "loss": 0.3665,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9720824949698189,
+      "grad_norm": 0.368213951587677,
+      "learning_rate": 8.54767037933262e-06,
+      "loss": 0.3747,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9723340040241448,
+      "grad_norm": 0.3184708058834076,
+      "learning_rate": 8.546639119077693e-06,
+      "loss": 0.3717,
+      "step": 3866
+    },
+    {
+      "epoch": 0.9725855130784709,
+      "grad_norm": 0.35572683811187744,
+      "learning_rate": 8.545607555075313e-06,
+      "loss": 0.3767,
+      "step": 3867
+    },
+    {
+      "epoch": 0.9728370221327968,
+      "grad_norm": 0.33128035068511963,
+      "learning_rate": 8.544575687413826e-06,
+      "loss": 0.3522,
+      "step": 3868
+    },
+    {
+      "epoch": 0.9730885311871227,
+      "grad_norm": 0.3530205488204956,
+      "learning_rate": 8.543543516181607e-06,
+      "loss": 0.409,
+      "step": 3869
+    },
+    {
+      "epoch": 0.9733400402414487,
+      "grad_norm": 0.31648722290992737,
+      "learning_rate": 8.542511041467054e-06,
+      "loss": 0.3962,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9735915492957746,
+      "grad_norm": 0.33272168040275574,
+      "learning_rate": 8.541478263358594e-06,
+      "loss": 0.3799,
+      "step": 3871
+    },
+    {
+      "epoch": 0.9738430583501007,
+      "grad_norm": 0.3218432366847992,
+      "learning_rate": 8.540445181944673e-06,
+      "loss": 0.3903,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9740945674044266,
+      "grad_norm": 0.33532074093818665,
+      "learning_rate": 8.539411797313772e-06,
+      "loss": 0.3831,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9743460764587525,
+      "grad_norm": 0.336195170879364,
+      "learning_rate": 8.538378109554395e-06,
+      "loss": 0.3744,
+      "step": 3874
+    },
+    {
+      "epoch": 0.9745975855130785,
+      "grad_norm": 0.364101380109787,
+      "learning_rate": 8.537344118755067e-06,
+      "loss": 0.4106,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9748490945674044,
+      "grad_norm": 0.3507227301597595,
+      "learning_rate": 8.536309825004346e-06,
+      "loss": 0.3932,
+      "step": 3876
+    },
+    {
+      "epoch": 0.9751006036217303,
+      "grad_norm": 0.368144154548645,
+      "learning_rate": 8.53527522839081e-06,
+      "loss": 0.3728,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9753521126760564,
+      "grad_norm": 0.3607654273509979,
+      "learning_rate": 8.53424032900307e-06,
+      "loss": 0.379,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9756036217303823,
+      "grad_norm": 0.3326222598552704,
+      "learning_rate": 8.533205126929754e-06,
+      "loss": 0.3651,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9758551307847082,
+      "grad_norm": 0.3740299344062805,
+      "learning_rate": 8.532169622259524e-06,
+      "loss": 0.3597,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9761066398390342,
+      "grad_norm": 0.3337053060531616,
+      "learning_rate": 8.531133815081061e-06,
+      "loss": 0.4037,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9763581488933601,
+      "grad_norm": 0.35001134872436523,
+      "learning_rate": 8.530097705483078e-06,
+      "loss": 0.3737,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9766096579476862,
+      "grad_norm": 0.3661789000034332,
+      "learning_rate": 8.52906129355431e-06,
+      "loss": 0.3587,
+      "step": 3883
+    },
+    {
+      "epoch": 0.9768611670020121,
+      "grad_norm": 0.33241093158721924,
+      "learning_rate": 8.528024579383522e-06,
+      "loss": 0.377,
+      "step": 3884
+    },
+    {
+      "epoch": 0.977112676056338,
+      "grad_norm": 0.3616609573364258,
+      "learning_rate": 8.5269875630595e-06,
+      "loss": 0.3804,
+      "step": 3885
+    },
+    {
+      "epoch": 0.977364185110664,
+      "grad_norm": 0.3413762152194977,
+      "learning_rate": 8.525950244671056e-06,
+      "loss": 0.3983,
+      "step": 3886
+    },
+    {
+      "epoch": 0.97761569416499,
+      "grad_norm": 0.41827529668807983,
+      "learning_rate": 8.524912624307033e-06,
+      "loss": 0.3762,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9778672032193159,
+      "grad_norm": 0.36674749851226807,
+      "learning_rate": 8.523874702056296e-06,
+      "loss": 0.37,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9781187122736419,
+      "grad_norm": 0.3255644142627716,
+      "learning_rate": 8.522836478007734e-06,
+      "loss": 0.3585,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9783702213279678,
+      "grad_norm": 0.3610647916793823,
+      "learning_rate": 8.521797952250269e-06,
+      "loss": 0.3807,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9786217303822937,
+      "grad_norm": 0.37085309624671936,
+      "learning_rate": 8.52075912487284e-06,
+      "loss": 0.3336,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9788732394366197,
+      "grad_norm": 0.3353044390678406,
+      "learning_rate": 8.519719995964419e-06,
+      "loss": 0.3723,
+      "step": 3892
+    },
+    {
+      "epoch": 0.9791247484909457,
+      "grad_norm": 0.3177686631679535,
+      "learning_rate": 8.518680565614e-06,
+      "loss": 0.3736,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9793762575452716,
+      "grad_norm": 0.36374950408935547,
+      "learning_rate": 8.517640833910602e-06,
+      "loss": 0.4041,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9796277665995976,
+      "grad_norm": 0.3766506314277649,
+      "learning_rate": 8.516600800943273e-06,
+      "loss": 0.3796,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9798792756539235,
+      "grad_norm": 0.3212386667728424,
+      "learning_rate": 8.515560466801085e-06,
+      "loss": 0.3817,
+      "step": 3896
+    },
+    {
+      "epoch": 0.9801307847082495,
+      "grad_norm": 0.35744473338127136,
+      "learning_rate": 8.514519831573137e-06,
+      "loss": 0.395,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9803822937625755,
+      "grad_norm": 0.3569160997867584,
+      "learning_rate": 8.513478895348552e-06,
+      "loss": 0.3832,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9806338028169014,
+      "grad_norm": 0.3330937922000885,
+      "learning_rate": 8.512437658216479e-06,
+      "loss": 0.4121,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9808853118712274,
+      "grad_norm": 0.3199692368507385,
+      "learning_rate": 8.511396120266095e-06,
+      "loss": 0.3629,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9811368209255533,
+      "grad_norm": 0.35635682940483093,
+      "learning_rate": 8.510354281586601e-06,
+      "loss": 0.3798,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9813883299798792,
+      "grad_norm": 0.3268703520298004,
+      "learning_rate": 8.509312142267223e-06,
+      "loss": 0.378,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9816398390342053,
+      "grad_norm": 0.3214782774448395,
+      "learning_rate": 8.508269702397214e-06,
+      "loss": 0.3788,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9818913480885312,
+      "grad_norm": 0.3744148015975952,
+      "learning_rate": 8.507226962065852e-06,
+      "loss": 0.39,
+      "step": 3904
+    },
+    {
+      "epoch": 0.9821428571428571,
+      "grad_norm": 0.33143913745880127,
+      "learning_rate": 8.506183921362443e-06,
+      "loss": 0.3522,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9823943661971831,
+      "grad_norm": 0.3448927104473114,
+      "learning_rate": 8.505140580376317e-06,
+      "loss": 0.3731,
+      "step": 3906
+    },
+    {
+      "epoch": 0.982645875251509,
+      "grad_norm": 0.37213438749313354,
+      "learning_rate": 8.504096939196826e-06,
+      "loss": 0.3569,
+      "step": 3907
+    },
+    {
+      "epoch": 0.9828973843058351,
+      "grad_norm": 0.3362247943878174,
+      "learning_rate": 8.503052997913354e-06,
+      "loss": 0.3844,
+      "step": 3908
+    },
+    {
+      "epoch": 0.983148893360161,
+      "grad_norm": 0.35339227318763733,
+      "learning_rate": 8.50200875661531e-06,
+      "loss": 0.3926,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9834004024144869,
+      "grad_norm": 0.3829168975353241,
+      "learning_rate": 8.500964215392122e-06,
+      "loss": 0.3594,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9836519114688129,
+      "grad_norm": 0.338537335395813,
+      "learning_rate": 8.499919374333251e-06,
+      "loss": 0.3922,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9839034205231388,
+      "grad_norm": 0.3299337923526764,
+      "learning_rate": 8.498874233528183e-06,
+      "loss": 0.3835,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9841549295774648,
+      "grad_norm": 0.3599414527416229,
+      "learning_rate": 8.497828793066425e-06,
+      "loss": 0.3518,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9844064386317908,
+      "grad_norm": 0.3576641380786896,
+      "learning_rate": 8.496783053037512e-06,
+      "loss": 0.3741,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9846579476861167,
+      "grad_norm": 0.3295727074146271,
+      "learning_rate": 8.495737013531008e-06,
+      "loss": 0.3744,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9849094567404426,
+      "grad_norm": 0.3430532217025757,
+      "learning_rate": 8.494690674636497e-06,
+      "loss": 0.3722,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9851609657947686,
+      "grad_norm": 0.3892742395401001,
+      "learning_rate": 8.493644036443592e-06,
+      "loss": 0.3736,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9854124748490946,
+      "grad_norm": 0.3252136707305908,
+      "learning_rate": 8.492597099041932e-06,
+      "loss": 0.3948,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9856639839034205,
+      "grad_norm": 0.3606947660446167,
+      "learning_rate": 8.49154986252118e-06,
+      "loss": 0.3537,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9859154929577465,
+      "grad_norm": 0.34790658950805664,
+      "learning_rate": 8.490502326971026e-06,
+      "loss": 0.3874,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9861670020120724,
+      "grad_norm": 0.34207773208618164,
+      "learning_rate": 8.489454492481184e-06,
+      "loss": 0.4054,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9864185110663984,
+      "grad_norm": 0.32859066128730774,
+      "learning_rate": 8.488406359141395e-06,
+      "loss": 0.3512,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9866700201207244,
+      "grad_norm": 0.42876073718070984,
+      "learning_rate": 8.487357927041425e-06,
+      "loss": 0.3749,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9869215291750503,
+      "grad_norm": 0.3597780764102936,
+      "learning_rate": 8.486309196271063e-06,
+      "loss": 0.3671,
+      "step": 3924
+    },
+    {
+      "epoch": 0.9871730382293763,
+      "grad_norm": 0.3601861596107483,
+      "learning_rate": 8.485260166920131e-06,
+      "loss": 0.3696,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9874245472837022,
+      "grad_norm": 0.450916588306427,
+      "learning_rate": 8.484210839078467e-06,
+      "loss": 0.3995,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9876760563380281,
+      "grad_norm": 0.38571393489837646,
+      "learning_rate": 8.483161212835944e-06,
+      "loss": 0.3746,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9879275653923542,
+      "grad_norm": 0.32381007075309753,
+      "learning_rate": 8.482111288282452e-06,
+      "loss": 0.3762,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9881790744466801,
+      "grad_norm": 0.4630524516105652,
+      "learning_rate": 8.481061065507915e-06,
+      "loss": 0.3895,
+      "step": 3929
+    },
+    {
+      "epoch": 0.988430583501006,
+      "grad_norm": 0.3516670763492584,
+      "learning_rate": 8.480010544602274e-06,
+      "loss": 0.3775,
+      "step": 3930
+    },
+    {
+      "epoch": 0.988682092555332,
+      "grad_norm": 0.3213813006877899,
+      "learning_rate": 8.4789597256555e-06,
+      "loss": 0.3751,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9889336016096579,
+      "grad_norm": 0.37605127692222595,
+      "learning_rate": 8.47790860875759e-06,
+      "loss": 0.3625,
+      "step": 3932
+    },
+    {
+      "epoch": 0.989185110663984,
+      "grad_norm": 0.3587568700313568,
+      "learning_rate": 8.476857193998564e-06,
+      "loss": 0.372,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9894366197183099,
+      "grad_norm": 0.3404126763343811,
+      "learning_rate": 8.475805481468472e-06,
+      "loss": 0.3751,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9896881287726358,
+      "grad_norm": 0.3651595711708069,
+      "learning_rate": 8.474753471257385e-06,
+      "loss": 0.3587,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9899396378269618,
+      "grad_norm": 0.3151306211948395,
+      "learning_rate": 8.473701163455401e-06,
+      "loss": 0.3889,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9901911468812877,
+      "grad_norm": 0.327453076839447,
+      "learning_rate": 8.472648558152646e-06,
+      "loss": 0.3706,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9904426559356136,
+      "grad_norm": 0.3796255588531494,
+      "learning_rate": 8.471595655439263e-06,
+      "loss": 0.3789,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9906941649899397,
+      "grad_norm": 0.3350921869277954,
+      "learning_rate": 8.470542455405432e-06,
+      "loss": 0.3938,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9909456740442656,
+      "grad_norm": 0.3343561291694641,
+      "learning_rate": 8.469488958141352e-06,
+      "loss": 0.3646,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9911971830985915,
+      "grad_norm": 0.35207462310791016,
+      "learning_rate": 8.468435163737248e-06,
+      "loss": 0.3782,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9914486921529175,
+      "grad_norm": 0.32531505823135376,
+      "learning_rate": 8.46738107228337e-06,
+      "loss": 0.3897,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9917002012072434,
+      "grad_norm": 0.367502897977829,
+      "learning_rate": 8.466326683869994e-06,
+      "loss": 0.3878,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9919517102615694,
+      "grad_norm": 0.37151041626930237,
+      "learning_rate": 8.465271998587424e-06,
+      "loss": 0.3696,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9922032193158954,
+      "grad_norm": 0.3209855556488037,
+      "learning_rate": 8.464217016525985e-06,
+      "loss": 0.3534,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9924547283702213,
+      "grad_norm": 0.3292153477668762,
+      "learning_rate": 8.463161737776031e-06,
+      "loss": 0.3612,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9927062374245473,
+      "grad_norm": 0.35022595524787903,
+      "learning_rate": 8.46210616242794e-06,
+      "loss": 0.3583,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9929577464788732,
+      "grad_norm": 0.3790688216686249,
+      "learning_rate": 8.461050290572114e-06,
+      "loss": 0.3809,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9932092555331992,
+      "grad_norm": 0.3780459463596344,
+      "learning_rate": 8.459994122298985e-06,
+      "loss": 0.3793,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9934607645875252,
+      "grad_norm": 0.325181782245636,
+      "learning_rate": 8.458937657699004e-06,
+      "loss": 0.3724,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9937122736418511,
+      "grad_norm": 0.35834598541259766,
+      "learning_rate": 8.457880896862651e-06,
+      "loss": 0.3791,
+      "step": 3951
+    },
+    {
+      "epoch": 0.993963782696177,
+      "grad_norm": 0.37170320749282837,
+      "learning_rate": 8.456823839880433e-06,
+      "loss": 0.3852,
+      "step": 3952
+    },
+    {
+      "epoch": 0.994215291750503,
+      "grad_norm": 0.3402329385280609,
+      "learning_rate": 8.455766486842878e-06,
+      "loss": 0.3763,
+      "step": 3953
+    },
+    {
+      "epoch": 0.994466800804829,
+      "grad_norm": 0.3130355179309845,
+      "learning_rate": 8.454708837840543e-06,
+      "loss": 0.3747,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9947183098591549,
+      "grad_norm": 0.32022619247436523,
+      "learning_rate": 8.453650892964008e-06,
+      "loss": 0.3611,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9949698189134809,
+      "grad_norm": 0.35048943758010864,
+      "learning_rate": 8.45259265230388e-06,
+      "loss": 0.3663,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9952213279678068,
+      "grad_norm": 0.3296698033809662,
+      "learning_rate": 8.45153411595079e-06,
+      "loss": 0.3886,
+      "step": 3957
+    },
+    {
+      "epoch": 0.9954728370221329,
+      "grad_norm": 0.3470214605331421,
+      "learning_rate": 8.450475283995398e-06,
+      "loss": 0.411,
+      "step": 3958
+    },
+    {
+      "epoch": 0.9957243460764588,
+      "grad_norm": 0.3476480543613434,
+      "learning_rate": 8.449416156528383e-06,
+      "loss": 0.3892,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9959758551307847,
+      "grad_norm": 0.3552456796169281,
+      "learning_rate": 8.448356733640453e-06,
+      "loss": 0.4037,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9962273641851107,
+      "grad_norm": 0.38896414637565613,
+      "learning_rate": 8.447297015422342e-06,
+      "loss": 0.3919,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9964788732394366,
+      "grad_norm": 0.3473275899887085,
+      "learning_rate": 8.446237001964808e-06,
+      "loss": 0.3809,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9967303822937625,
+      "grad_norm": 0.3539365530014038,
+      "learning_rate": 8.445176693358634e-06,
+      "loss": 0.3513,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9969818913480886,
+      "grad_norm": 0.39693740010261536,
+      "learning_rate": 8.444116089694631e-06,
+      "loss": 0.3734,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9972334004024145,
+      "grad_norm": 0.3378576934337616,
+      "learning_rate": 8.443055191063629e-06,
+      "loss": 0.3644,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9974849094567404,
+      "grad_norm": 0.3627392053604126,
+      "learning_rate": 8.44199399755649e-06,
+      "loss": 0.3903,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9977364185110664,
+      "grad_norm": 0.38984745740890503,
+      "learning_rate": 8.440932509264099e-06,
+      "loss": 0.3932,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9979879275653923,
+      "grad_norm": 0.34223487973213196,
+      "learning_rate": 8.439870726277364e-06,
+      "loss": 0.3599,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9982394366197183,
+      "grad_norm": 0.33947303891181946,
+      "learning_rate": 8.438808648687223e-06,
+      "loss": 0.381,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9984909456740443,
+      "grad_norm": 0.3354572057723999,
+      "learning_rate": 8.437746276584631e-06,
+      "loss": 0.3726,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9987424547283702,
+      "grad_norm": 0.36442792415618896,
+      "learning_rate": 8.43668361006058e-06,
+      "loss": 0.3951,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9989939637826962,
+      "grad_norm": 0.3597300946712494,
+      "learning_rate": 8.435620649206076e-06,
+      "loss": 0.3476,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9992454728370221,
+      "grad_norm": 0.3294813632965088,
+      "learning_rate": 8.434557394112156e-06,
+      "loss": 0.3753,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9994969818913481,
+      "grad_norm": 0.32441380620002747,
+      "learning_rate": 8.433493844869883e-06,
+      "loss": 0.3672,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9997484909456741,
+      "grad_norm": 0.3753267526626587,
+      "learning_rate": 8.432430001570343e-06,
+      "loss": 0.377,
+      "step": 3975
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3612028658390045,
+      "learning_rate": 8.431365864304645e-06,
+      "loss": 0.3862,
+      "step": 3976
+    },
+    {
+      "epoch": 1.000251509054326,
+      "grad_norm": 0.3513164818286896,
+      "learning_rate": 8.430301433163927e-06,
+      "loss": 0.3648,
+      "step": 3977
+    },
+    {
+      "epoch": 1.0005030181086518,
+      "grad_norm": 0.3320719599723816,
+      "learning_rate": 8.42923670823935e-06,
+      "loss": 0.3366,
+      "step": 3978
+    },
+    {
+      "epoch": 1.0007545271629779,
+      "grad_norm": 0.3477346897125244,
+      "learning_rate": 8.428171689622105e-06,
+      "loss": 0.3598,
+      "step": 3979
+    },
+    {
+      "epoch": 1.0010060362173039,
+      "grad_norm": 0.3291724920272827,
+      "learning_rate": 8.4271063774034e-06,
+      "loss": 0.3622,
+      "step": 3980
+    },
+    {
+      "epoch": 1.0012575452716297,
+      "grad_norm": 0.3617793917655945,
+      "learning_rate": 8.426040771674475e-06,
+      "loss": 0.3601,
+      "step": 3981
+    },
+    {
+      "epoch": 1.0015090543259557,
+      "grad_norm": 0.33502376079559326,
+      "learning_rate": 8.42497487252659e-06,
+      "loss": 0.3577,
+      "step": 3982
+    },
+    {
+      "epoch": 1.0017605633802817,
+      "grad_norm": 0.35225436091423035,
+      "learning_rate": 8.423908680051035e-06,
+      "loss": 0.3543,
+      "step": 3983
+    },
+    {
+      "epoch": 1.0020120724346075,
+      "grad_norm": 0.3484959304332733,
+      "learning_rate": 8.42284219433912e-06,
+      "loss": 0.3501,
+      "step": 3984
+    },
+    {
+      "epoch": 1.0022635814889336,
+      "grad_norm": 0.348874568939209,
+      "learning_rate": 8.421775415482183e-06,
+      "loss": 0.3557,
+      "step": 3985
+    },
+    {
+      "epoch": 1.0025150905432596,
+      "grad_norm": 0.36964142322540283,
+      "learning_rate": 8.42070834357159e-06,
+      "loss": 0.3564,
+      "step": 3986
+    },
+    {
+      "epoch": 1.0027665995975854,
+      "grad_norm": 0.3703876733779907,
+      "learning_rate": 8.419640978698728e-06,
+      "loss": 0.3379,
+      "step": 3987
+    },
+    {
+      "epoch": 1.0030181086519114,
+      "grad_norm": 0.35077622532844543,
+      "learning_rate": 8.418573320955008e-06,
+      "loss": 0.3577,
+      "step": 3988
+    },
+    {
+      "epoch": 1.0032696177062375,
+      "grad_norm": 0.35275188088417053,
+      "learning_rate": 8.417505370431869e-06,
+      "loss": 0.3819,
+      "step": 3989
+    },
+    {
+      "epoch": 1.0035211267605635,
+      "grad_norm": 0.3845551311969757,
+      "learning_rate": 8.416437127220777e-06,
+      "loss": 0.3294,
+      "step": 3990
+    },
+    {
+      "epoch": 1.0037726358148893,
+      "grad_norm": 0.3270184397697449,
+      "learning_rate": 8.415368591413218e-06,
+      "loss": 0.3455,
+      "step": 3991
+    },
+    {
+      "epoch": 1.0040241448692153,
+      "grad_norm": 0.3642314374446869,
+      "learning_rate": 8.414299763100704e-06,
+      "loss": 0.3649,
+      "step": 3992
+    },
+    {
+      "epoch": 1.0042756539235413,
+      "grad_norm": 0.39639967679977417,
+      "learning_rate": 8.413230642374776e-06,
+      "loss": 0.3558,
+      "step": 3993
+    },
+    {
+      "epoch": 1.0045271629778671,
+      "grad_norm": 0.3734232485294342,
+      "learning_rate": 8.412161229326997e-06,
+      "loss": 0.3575,
+      "step": 3994
+    },
+    {
+      "epoch": 1.0047786720321932,
+      "grad_norm": 0.3293455243110657,
+      "learning_rate": 8.411091524048953e-06,
+      "loss": 0.369,
+      "step": 3995
+    },
+    {
+      "epoch": 1.0050301810865192,
+      "grad_norm": 0.3649202883243561,
+      "learning_rate": 8.410021526632262e-06,
+      "loss": 0.3557,
+      "step": 3996
+    },
+    {
+      "epoch": 1.005281690140845,
+      "grad_norm": 0.3437403440475464,
+      "learning_rate": 8.408951237168559e-06,
+      "loss": 0.3742,
+      "step": 3997
+    },
+    {
+      "epoch": 1.005533199195171,
+      "grad_norm": 0.34628748893737793,
+      "learning_rate": 8.40788065574951e-06,
+      "loss": 0.3851,
+      "step": 3998
+    },
+    {
+      "epoch": 1.005784708249497,
+      "grad_norm": 0.40169525146484375,
+      "learning_rate": 8.4068097824668e-06,
+      "loss": 0.3603,
+      "step": 3999
+    },
+    {
+      "epoch": 1.0060362173038229,
+      "grad_norm": 0.3196563422679901,
+      "learning_rate": 8.405738617412148e-06,
+      "loss": 0.3477,
+      "step": 4000
+    },
+    {
+      "epoch": 1.006287726358149,
+      "grad_norm": 0.3684934675693512,
+      "learning_rate": 8.404667160677289e-06,
+      "loss": 0.3876,
+      "step": 4001
+    },
+    {
+      "epoch": 1.006539235412475,
+      "grad_norm": 0.35352370142936707,
+      "learning_rate": 8.403595412353987e-06,
+      "loss": 0.3676,
+      "step": 4002
+    },
+    {
+      "epoch": 1.0067907444668007,
+      "grad_norm": 0.3421543836593628,
+      "learning_rate": 8.40252337253403e-06,
+      "loss": 0.3632,
+      "step": 4003
+    },
+    {
+      "epoch": 1.0070422535211268,
+      "grad_norm": 0.34123724699020386,
+      "learning_rate": 8.401451041309233e-06,
+      "loss": 0.3834,
+      "step": 4004
+    },
+    {
+      "epoch": 1.0072937625754528,
+      "grad_norm": 0.3148738741874695,
+      "learning_rate": 8.400378418771434e-06,
+      "loss": 0.3571,
+      "step": 4005
+    },
+    {
+      "epoch": 1.0075452716297786,
+      "grad_norm": 0.3324401378631592,
+      "learning_rate": 8.399305505012496e-06,
+      "loss": 0.3645,
+      "step": 4006
+    },
+    {
+      "epoch": 1.0077967806841046,
+      "grad_norm": 0.3198019564151764,
+      "learning_rate": 8.398232300124307e-06,
+      "loss": 0.3457,
+      "step": 4007
+    },
+    {
+      "epoch": 1.0080482897384306,
+      "grad_norm": 0.3236132264137268,
+      "learning_rate": 8.39715880419878e-06,
+      "loss": 0.3419,
+      "step": 4008
+    },
+    {
+      "epoch": 1.0082997987927564,
+      "grad_norm": 0.3504725396633148,
+      "learning_rate": 8.396085017327854e-06,
+      "loss": 0.3376,
+      "step": 4009
+    },
+    {
+      "epoch": 1.0085513078470825,
+      "grad_norm": 0.3179847002029419,
+      "learning_rate": 8.395010939603493e-06,
+      "loss": 0.3549,
+      "step": 4010
+    },
+    {
+      "epoch": 1.0088028169014085,
+      "grad_norm": 0.31894704699516296,
+      "learning_rate": 8.393936571117685e-06,
+      "loss": 0.3486,
+      "step": 4011
+    },
+    {
+      "epoch": 1.0090543259557343,
+      "grad_norm": 0.3283448815345764,
+      "learning_rate": 8.392861911962441e-06,
+      "loss": 0.3577,
+      "step": 4012
+    },
+    {
+      "epoch": 1.0093058350100603,
+      "grad_norm": 0.33308663964271545,
+      "learning_rate": 8.3917869622298e-06,
+      "loss": 0.3494,
+      "step": 4013
+    },
+    {
+      "epoch": 1.0095573440643864,
+      "grad_norm": 0.33545833826065063,
+      "learning_rate": 8.390711722011825e-06,
+      "loss": 0.3131,
+      "step": 4014
+    },
+    {
+      "epoch": 1.0098088531187124,
+      "grad_norm": 0.36120879650115967,
+      "learning_rate": 8.389636191400603e-06,
+      "loss": 0.3673,
+      "step": 4015
+    },
+    {
+      "epoch": 1.0100603621730382,
+      "grad_norm": 0.3518035411834717,
+      "learning_rate": 8.388560370488247e-06,
+      "loss": 0.3616,
+      "step": 4016
+    },
+    {
+      "epoch": 1.0103118712273642,
+      "grad_norm": 0.3019152879714966,
+      "learning_rate": 8.387484259366894e-06,
+      "loss": 0.3508,
+      "step": 4017
+    },
+    {
+      "epoch": 1.0105633802816902,
+      "grad_norm": 0.3448386788368225,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.3544,
+      "step": 4018
+    },
+    {
+      "epoch": 1.010814889336016,
+      "grad_norm": 0.32667118310928345,
+      "learning_rate": 8.38533116686587e-06,
+      "loss": 0.3429,
+      "step": 4019
+    },
+    {
+      "epoch": 1.011066398390342,
+      "grad_norm": 0.3479123115539551,
+      "learning_rate": 8.384254185670599e-06,
+      "loss": 0.3632,
+      "step": 4020
+    },
+    {
+      "epoch": 1.011317907444668,
+      "grad_norm": 0.3146592378616333,
+      "learning_rate": 8.383176914635127e-06,
+      "loss": 0.3615,
+      "step": 4021
+    },
+    {
+      "epoch": 1.011569416498994,
+      "grad_norm": 0.36215662956237793,
+      "learning_rate": 8.38209935385172e-06,
+      "loss": 0.3659,
+      "step": 4022
+    },
+    {
+      "epoch": 1.01182092555332,
+      "grad_norm": 0.3197457790374756,
+      "learning_rate": 8.381021503412659e-06,
+      "loss": 0.3269,
+      "step": 4023
+    },
+    {
+      "epoch": 1.012072434607646,
+      "grad_norm": 0.34255650639533997,
+      "learning_rate": 8.379943363410259e-06,
+      "loss": 0.3715,
+      "step": 4024
+    },
+    {
+      "epoch": 1.0123239436619718,
+      "grad_norm": 0.34382399916648865,
+      "learning_rate": 8.378864933936856e-06,
+      "loss": 0.3318,
+      "step": 4025
+    },
+    {
+      "epoch": 1.0125754527162978,
+      "grad_norm": 0.3288285434246063,
+      "learning_rate": 8.37778621508481e-06,
+      "loss": 0.3737,
+      "step": 4026
+    },
+    {
+      "epoch": 1.0128269617706238,
+      "grad_norm": 0.33756786584854126,
+      "learning_rate": 8.376707206946503e-06,
+      "loss": 0.3459,
+      "step": 4027
+    },
+    {
+      "epoch": 1.0130784708249496,
+      "grad_norm": 0.37219181656837463,
+      "learning_rate": 8.37562790961435e-06,
+      "loss": 0.3549,
+      "step": 4028
+    },
+    {
+      "epoch": 1.0133299798792756,
+      "grad_norm": 0.3869669735431671,
+      "learning_rate": 8.374548323180783e-06,
+      "loss": 0.3594,
+      "step": 4029
+    },
+    {
+      "epoch": 1.0135814889336017,
+      "grad_norm": 0.3502846360206604,
+      "learning_rate": 8.373468447738265e-06,
+      "loss": 0.3629,
+      "step": 4030
+    },
+    {
+      "epoch": 1.0138329979879275,
+      "grad_norm": 0.34061774611473083,
+      "learning_rate": 8.372388283379277e-06,
+      "loss": 0.3644,
+      "step": 4031
+    },
+    {
+      "epoch": 1.0140845070422535,
+      "grad_norm": 0.3984410762786865,
+      "learning_rate": 8.37130783019633e-06,
+      "loss": 0.3692,
+      "step": 4032
+    },
+    {
+      "epoch": 1.0143360160965795,
+      "grad_norm": 0.3561273515224457,
+      "learning_rate": 8.370227088281962e-06,
+      "loss": 0.3375,
+      "step": 4033
+    },
+    {
+      "epoch": 1.0145875251509053,
+      "grad_norm": 0.3067105710506439,
+      "learning_rate": 8.369146057728726e-06,
+      "loss": 0.3403,
+      "step": 4034
+    },
+    {
+      "epoch": 1.0148390342052314,
+      "grad_norm": 0.39280879497528076,
+      "learning_rate": 8.368064738629205e-06,
+      "loss": 0.3513,
+      "step": 4035
+    },
+    {
+      "epoch": 1.0150905432595574,
+      "grad_norm": 0.3620704114437103,
+      "learning_rate": 8.366983131076012e-06,
+      "loss": 0.3579,
+      "step": 4036
+    },
+    {
+      "epoch": 1.0153420523138832,
+      "grad_norm": 0.3495025038719177,
+      "learning_rate": 8.365901235161778e-06,
+      "loss": 0.3641,
+      "step": 4037
+    },
+    {
+      "epoch": 1.0155935613682092,
+      "grad_norm": 0.350424587726593,
+      "learning_rate": 8.36481905097916e-06,
+      "loss": 0.3567,
+      "step": 4038
+    },
+    {
+      "epoch": 1.0158450704225352,
+      "grad_norm": 0.31671619415283203,
+      "learning_rate": 8.363736578620838e-06,
+      "loss": 0.3795,
+      "step": 4039
+    },
+    {
+      "epoch": 1.0160965794768613,
+      "grad_norm": 0.3311213552951813,
+      "learning_rate": 8.362653818179524e-06,
+      "loss": 0.3523,
+      "step": 4040
+    },
+    {
+      "epoch": 1.016348088531187,
+      "grad_norm": 0.3370438814163208,
+      "learning_rate": 8.361570769747948e-06,
+      "loss": 0.3614,
+      "step": 4041
+    },
+    {
+      "epoch": 1.016599597585513,
+      "grad_norm": 0.33826106786727905,
+      "learning_rate": 8.360487433418863e-06,
+      "loss": 0.3289,
+      "step": 4042
+    },
+    {
+      "epoch": 1.0168511066398391,
+      "grad_norm": 0.37292927503585815,
+      "learning_rate": 8.359403809285054e-06,
+      "loss": 0.3549,
+      "step": 4043
+    },
+    {
+      "epoch": 1.017102615694165,
+      "grad_norm": 0.3336230516433716,
+      "learning_rate": 8.358319897439324e-06,
+      "loss": 0.3397,
+      "step": 4044
+    },
+    {
+      "epoch": 1.017354124748491,
+      "grad_norm": 0.3715561032295227,
+      "learning_rate": 8.357235697974506e-06,
+      "loss": 0.335,
+      "step": 4045
+    },
+    {
+      "epoch": 1.017605633802817,
+      "grad_norm": 0.3317098617553711,
+      "learning_rate": 8.356151210983451e-06,
+      "loss": 0.3557,
+      "step": 4046
+    },
+    {
+      "epoch": 1.0178571428571428,
+      "grad_norm": 0.3567894697189331,
+      "learning_rate": 8.355066436559042e-06,
+      "loss": 0.3431,
+      "step": 4047
+    },
+    {
+      "epoch": 1.0181086519114688,
+      "grad_norm": 0.31306058168411255,
+      "learning_rate": 8.353981374794184e-06,
+      "loss": 0.3529,
+      "step": 4048
+    },
+    {
+      "epoch": 1.0183601609657948,
+      "grad_norm": 0.3463604748249054,
+      "learning_rate": 8.3528960257818e-06,
+      "loss": 0.3559,
+      "step": 4049
+    },
+    {
+      "epoch": 1.0186116700201207,
+      "grad_norm": 0.36035671830177307,
+      "learning_rate": 8.35181038961485e-06,
+      "loss": 0.3697,
+      "step": 4050
+    },
+    {
+      "epoch": 1.0188631790744467,
+      "grad_norm": 0.31114959716796875,
+      "learning_rate": 8.350724466386309e-06,
+      "loss": 0.3163,
+      "step": 4051
+    },
+    {
+      "epoch": 1.0191146881287727,
+      "grad_norm": 0.33771616220474243,
+      "learning_rate": 8.349638256189178e-06,
+      "loss": 0.3651,
+      "step": 4052
+    },
+    {
+      "epoch": 1.0193661971830985,
+      "grad_norm": 0.38065779209136963,
+      "learning_rate": 8.348551759116485e-06,
+      "loss": 0.337,
+      "step": 4053
+    },
+    {
+      "epoch": 1.0196177062374245,
+      "grad_norm": 0.3252953290939331,
+      "learning_rate": 8.347464975261283e-06,
+      "loss": 0.3479,
+      "step": 4054
+    },
+    {
+      "epoch": 1.0198692152917506,
+      "grad_norm": 0.3428071439266205,
+      "learning_rate": 8.346377904716649e-06,
+      "loss": 0.3571,
+      "step": 4055
+    },
+    {
+      "epoch": 1.0201207243460764,
+      "grad_norm": 0.3688924312591553,
+      "learning_rate": 8.34529054757568e-06,
+      "loss": 0.3772,
+      "step": 4056
+    },
+    {
+      "epoch": 1.0203722334004024,
+      "grad_norm": 0.3667786121368408,
+      "learning_rate": 8.344202903931504e-06,
+      "loss": 0.359,
+      "step": 4057
+    },
+    {
+      "epoch": 1.0206237424547284,
+      "grad_norm": 0.3390733003616333,
+      "learning_rate": 8.343114973877273e-06,
+      "loss": 0.3513,
+      "step": 4058
+    },
+    {
+      "epoch": 1.0208752515090542,
+      "grad_norm": 0.34325966238975525,
+      "learning_rate": 8.342026757506156e-06,
+      "loss": 0.3834,
+      "step": 4059
+    },
+    {
+      "epoch": 1.0211267605633803,
+      "grad_norm": 0.37722912430763245,
+      "learning_rate": 8.340938254911358e-06,
+      "loss": 0.3481,
+      "step": 4060
+    },
+    {
+      "epoch": 1.0213782696177063,
+      "grad_norm": 0.33139869570732117,
+      "learning_rate": 8.339849466186096e-06,
+      "loss": 0.3473,
+      "step": 4061
+    },
+    {
+      "epoch": 1.021629778672032,
+      "grad_norm": 0.3419877886772156,
+      "learning_rate": 8.338760391423623e-06,
+      "loss": 0.3419,
+      "step": 4062
+    },
+    {
+      "epoch": 1.0218812877263581,
+      "grad_norm": 0.37569326162338257,
+      "learning_rate": 8.33767103071721e-06,
+      "loss": 0.3518,
+      "step": 4063
+    },
+    {
+      "epoch": 1.0221327967806841,
+      "grad_norm": 0.3753376603126526,
+      "learning_rate": 8.336581384160152e-06,
+      "loss": 0.3399,
+      "step": 4064
+    },
+    {
+      "epoch": 1.0223843058350102,
+      "grad_norm": 0.3511211574077606,
+      "learning_rate": 8.335491451845774e-06,
+      "loss": 0.3732,
+      "step": 4065
+    },
+    {
+      "epoch": 1.022635814889336,
+      "grad_norm": 0.3783870339393616,
+      "learning_rate": 8.334401233867418e-06,
+      "loss": 0.365,
+      "step": 4066
+    },
+    {
+      "epoch": 1.022887323943662,
+      "grad_norm": 0.41309934854507446,
+      "learning_rate": 8.333310730318457e-06,
+      "loss": 0.3703,
+      "step": 4067
+    },
+    {
+      "epoch": 1.023138832997988,
+      "grad_norm": 0.5328013300895691,
+      "learning_rate": 8.332219941292286e-06,
+      "loss": 0.3598,
+      "step": 4068
+    },
+    {
+      "epoch": 1.0233903420523138,
+      "grad_norm": 0.3765561282634735,
+      "learning_rate": 8.331128866882323e-06,
+      "loss": 0.3709,
+      "step": 4069
+    },
+    {
+      "epoch": 1.0236418511066399,
+      "grad_norm": 0.37244337797164917,
+      "learning_rate": 8.330037507182012e-06,
+      "loss": 0.3685,
+      "step": 4070
+    },
+    {
+      "epoch": 1.0238933601609659,
+      "grad_norm": 0.3581307530403137,
+      "learning_rate": 8.328945862284821e-06,
+      "loss": 0.364,
+      "step": 4071
+    },
+    {
+      "epoch": 1.0241448692152917,
+      "grad_norm": 0.32666563987731934,
+      "learning_rate": 8.327853932284242e-06,
+      "loss": 0.3368,
+      "step": 4072
+    },
+    {
+      "epoch": 1.0243963782696177,
+      "grad_norm": 0.31926366686820984,
+      "learning_rate": 8.326761717273793e-06,
+      "loss": 0.3595,
+      "step": 4073
+    },
+    {
+      "epoch": 1.0246478873239437,
+      "grad_norm": 0.3939271867275238,
+      "learning_rate": 8.325669217347017e-06,
+      "loss": 0.3444,
+      "step": 4074
+    },
+    {
+      "epoch": 1.0248993963782695,
+      "grad_norm": 0.33755072951316833,
+      "learning_rate": 8.324576432597476e-06,
+      "loss": 0.3422,
+      "step": 4075
+    },
+    {
+      "epoch": 1.0251509054325956,
+      "grad_norm": 0.3223980963230133,
+      "learning_rate": 8.32348336311876e-06,
+      "loss": 0.3244,
+      "step": 4076
+    },
+    {
+      "epoch": 1.0254024144869216,
+      "grad_norm": 0.3657818138599396,
+      "learning_rate": 8.322390009004488e-06,
+      "loss": 0.3487,
+      "step": 4077
+    },
+    {
+      "epoch": 1.0256539235412474,
+      "grad_norm": 0.3748267590999603,
+      "learning_rate": 8.321296370348297e-06,
+      "loss": 0.353,
+      "step": 4078
+    },
+    {
+      "epoch": 1.0259054325955734,
+      "grad_norm": 0.3817254304885864,
+      "learning_rate": 8.320202447243851e-06,
+      "loss": 0.3505,
+      "step": 4079
+    },
+    {
+      "epoch": 1.0261569416498995,
+      "grad_norm": 0.32929888367652893,
+      "learning_rate": 8.319108239784834e-06,
+      "loss": 0.3457,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0264084507042253,
+      "grad_norm": 0.377116322517395,
+      "learning_rate": 8.318013748064962e-06,
+      "loss": 0.349,
+      "step": 4081
+    },
+    {
+      "epoch": 1.0266599597585513,
+      "grad_norm": 0.34213367104530334,
+      "learning_rate": 8.316918972177968e-06,
+      "loss": 0.3627,
+      "step": 4082
+    },
+    {
+      "epoch": 1.0269114688128773,
+      "grad_norm": 0.3591276705265045,
+      "learning_rate": 8.315823912217615e-06,
+      "loss": 0.3497,
+      "step": 4083
+    },
+    {
+      "epoch": 1.0271629778672031,
+      "grad_norm": 0.33743128180503845,
+      "learning_rate": 8.314728568277691e-06,
+      "loss": 0.3589,
+      "step": 4084
+    },
+    {
+      "epoch": 1.0274144869215291,
+      "grad_norm": 0.34647324681282043,
+      "learning_rate": 8.313632940452e-06,
+      "loss": 0.3588,
+      "step": 4085
+    },
+    {
+      "epoch": 1.0276659959758552,
+      "grad_norm": 0.3492962718009949,
+      "learning_rate": 8.312537028834374e-06,
+      "loss": 0.3581,
+      "step": 4086
+    },
+    {
+      "epoch": 1.027917505030181,
+      "grad_norm": 0.3124542832374573,
+      "learning_rate": 8.311440833518678e-06,
+      "loss": 0.3418,
+      "step": 4087
+    },
+    {
+      "epoch": 1.028169014084507,
+      "grad_norm": 0.3951411545276642,
+      "learning_rate": 8.310344354598791e-06,
+      "loss": 0.3615,
+      "step": 4088
+    },
+    {
+      "epoch": 1.028420523138833,
+      "grad_norm": 0.3501111567020416,
+      "learning_rate": 8.30924759216862e-06,
+      "loss": 0.3331,
+      "step": 4089
+    },
+    {
+      "epoch": 1.028672032193159,
+      "grad_norm": 0.36356234550476074,
+      "learning_rate": 8.308150546322093e-06,
+      "loss": 0.3728,
+      "step": 4090
+    },
+    {
+      "epoch": 1.0289235412474849,
+      "grad_norm": 0.3362111747264862,
+      "learning_rate": 8.30705321715317e-06,
+      "loss": 0.3322,
+      "step": 4091
+    },
+    {
+      "epoch": 1.029175050301811,
+      "grad_norm": 0.3690354824066162,
+      "learning_rate": 8.305955604755827e-06,
+      "loss": 0.3943,
+      "step": 4092
+    },
+    {
+      "epoch": 1.029426559356137,
+      "grad_norm": 0.3664776682853699,
+      "learning_rate": 8.304857709224068e-06,
+      "loss": 0.3586,
+      "step": 4093
+    },
+    {
+      "epoch": 1.0296780684104627,
+      "grad_norm": 0.33852067589759827,
+      "learning_rate": 8.303759530651921e-06,
+      "loss": 0.3354,
+      "step": 4094
+    },
+    {
+      "epoch": 1.0299295774647887,
+      "grad_norm": 0.339894562959671,
+      "learning_rate": 8.30266106913344e-06,
+      "loss": 0.3487,
+      "step": 4095
+    },
+    {
+      "epoch": 1.0301810865191148,
+      "grad_norm": 0.34180840849876404,
+      "learning_rate": 8.301562324762698e-06,
+      "loss": 0.3661,
+      "step": 4096
+    },
+    {
+      "epoch": 1.0304325955734406,
+      "grad_norm": 0.35557109117507935,
+      "learning_rate": 8.300463297633798e-06,
+      "loss": 0.3469,
+      "step": 4097
+    },
+    {
+      "epoch": 1.0306841046277666,
+      "grad_norm": 0.32619279623031616,
+      "learning_rate": 8.299363987840864e-06,
+      "loss": 0.3601,
+      "step": 4098
+    },
+    {
+      "epoch": 1.0309356136820926,
+      "grad_norm": 0.32750555872917175,
+      "learning_rate": 8.298264395478046e-06,
+      "loss": 0.3436,
+      "step": 4099
+    },
+    {
+      "epoch": 1.0311871227364184,
+      "grad_norm": 0.3431573808193207,
+      "learning_rate": 8.297164520639515e-06,
+      "loss": 0.3605,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0314386317907445,
+      "grad_norm": 0.3561439514160156,
+      "learning_rate": 8.29606436341947e-06,
+      "loss": 0.3392,
+      "step": 4101
+    },
+    {
+      "epoch": 1.0316901408450705,
+      "grad_norm": 0.3567664921283722,
+      "learning_rate": 8.294963923912134e-06,
+      "loss": 0.3511,
+      "step": 4102
+    },
+    {
+      "epoch": 1.0319416498993963,
+      "grad_norm": 0.35723668336868286,
+      "learning_rate": 8.293863202211751e-06,
+      "loss": 0.3727,
+      "step": 4103
+    },
+    {
+      "epoch": 1.0321931589537223,
+      "grad_norm": 0.34615108370780945,
+      "learning_rate": 8.292762198412591e-06,
+      "loss": 0.3533,
+      "step": 4104
+    },
+    {
+      "epoch": 1.0324446680080483,
+      "grad_norm": 0.345207542181015,
+      "learning_rate": 8.291660912608948e-06,
+      "loss": 0.3307,
+      "step": 4105
+    },
+    {
+      "epoch": 1.0326961770623742,
+      "grad_norm": 0.3570762276649475,
+      "learning_rate": 8.290559344895139e-06,
+      "loss": 0.3474,
+      "step": 4106
+    },
+    {
+      "epoch": 1.0329476861167002,
+      "grad_norm": 0.3682233393192291,
+      "learning_rate": 8.28945749536551e-06,
+      "loss": 0.3729,
+      "step": 4107
+    },
+    {
+      "epoch": 1.0331991951710262,
+      "grad_norm": 0.4089745879173279,
+      "learning_rate": 8.288355364114423e-06,
+      "loss": 0.3803,
+      "step": 4108
+    },
+    {
+      "epoch": 1.033450704225352,
+      "grad_norm": 0.3189091682434082,
+      "learning_rate": 8.287252951236272e-06,
+      "loss": 0.367,
+      "step": 4109
+    },
+    {
+      "epoch": 1.033702213279678,
+      "grad_norm": 0.3654859662055969,
+      "learning_rate": 8.28615025682547e-06,
+      "loss": 0.3674,
+      "step": 4110
+    },
+    {
+      "epoch": 1.033953722334004,
+      "grad_norm": 0.38779398798942566,
+      "learning_rate": 8.285047280976458e-06,
+      "loss": 0.3746,
+      "step": 4111
+    },
+    {
+      "epoch": 1.0342052313883299,
+      "grad_norm": 0.3446475863456726,
+      "learning_rate": 8.283944023783697e-06,
+      "loss": 0.3528,
+      "step": 4112
+    },
+    {
+      "epoch": 1.034456740442656,
+      "grad_norm": 0.33927372097969055,
+      "learning_rate": 8.282840485341675e-06,
+      "loss": 0.3567,
+      "step": 4113
+    },
+    {
+      "epoch": 1.034708249496982,
+      "grad_norm": 0.35856184363365173,
+      "learning_rate": 8.281736665744902e-06,
+      "loss": 0.3442,
+      "step": 4114
+    },
+    {
+      "epoch": 1.034959758551308,
+      "grad_norm": 0.35313889384269714,
+      "learning_rate": 8.280632565087913e-06,
+      "loss": 0.3552,
+      "step": 4115
+    },
+    {
+      "epoch": 1.0352112676056338,
+      "grad_norm": 0.33305633068084717,
+      "learning_rate": 8.27952818346527e-06,
+      "loss": 0.3299,
+      "step": 4116
+    },
+    {
+      "epoch": 1.0354627766599598,
+      "grad_norm": 0.3259493112564087,
+      "learning_rate": 8.278423520971556e-06,
+      "loss": 0.3376,
+      "step": 4117
+    },
+    {
+      "epoch": 1.0357142857142858,
+      "grad_norm": 0.3438626527786255,
+      "learning_rate": 8.277318577701375e-06,
+      "loss": 0.3475,
+      "step": 4118
+    },
+    {
+      "epoch": 1.0359657947686116,
+      "grad_norm": 0.31181490421295166,
+      "learning_rate": 8.27621335374936e-06,
+      "loss": 0.3567,
+      "step": 4119
+    },
+    {
+      "epoch": 1.0362173038229376,
+      "grad_norm": 0.3249053359031677,
+      "learning_rate": 8.275107849210168e-06,
+      "loss": 0.358,
+      "step": 4120
+    },
+    {
+      "epoch": 1.0364688128772637,
+      "grad_norm": 0.38430964946746826,
+      "learning_rate": 8.27400206417848e-06,
+      "loss": 0.3461,
+      "step": 4121
+    },
+    {
+      "epoch": 1.0367203219315895,
+      "grad_norm": 0.3271859884262085,
+      "learning_rate": 8.272895998748996e-06,
+      "loss": 0.349,
+      "step": 4122
+    },
+    {
+      "epoch": 1.0369718309859155,
+      "grad_norm": 0.32328277826309204,
+      "learning_rate": 8.271789653016445e-06,
+      "loss": 0.3598,
+      "step": 4123
+    },
+    {
+      "epoch": 1.0372233400402415,
+      "grad_norm": 0.3417379856109619,
+      "learning_rate": 8.270683027075576e-06,
+      "loss": 0.3591,
+      "step": 4124
+    },
+    {
+      "epoch": 1.0374748490945673,
+      "grad_norm": 0.31911101937294006,
+      "learning_rate": 8.26957612102117e-06,
+      "loss": 0.3416,
+      "step": 4125
+    },
+    {
+      "epoch": 1.0377263581488934,
+      "grad_norm": 0.31615710258483887,
+      "learning_rate": 8.268468934948023e-06,
+      "loss": 0.3325,
+      "step": 4126
+    },
+    {
+      "epoch": 1.0379778672032194,
+      "grad_norm": 0.34855955839157104,
+      "learning_rate": 8.267361468950958e-06,
+      "loss": 0.3408,
+      "step": 4127
+    },
+    {
+      "epoch": 1.0382293762575452,
+      "grad_norm": 0.3437637984752655,
+      "learning_rate": 8.266253723124825e-06,
+      "loss": 0.3463,
+      "step": 4128
+    },
+    {
+      "epoch": 1.0384808853118712,
+      "grad_norm": 0.33804404735565186,
+      "learning_rate": 8.265145697564493e-06,
+      "loss": 0.3452,
+      "step": 4129
+    },
+    {
+      "epoch": 1.0387323943661972,
+      "grad_norm": 0.3620853126049042,
+      "learning_rate": 8.26403739236486e-06,
+      "loss": 0.364,
+      "step": 4130
+    },
+    {
+      "epoch": 1.038983903420523,
+      "grad_norm": 0.3575562834739685,
+      "learning_rate": 8.262928807620843e-06,
+      "loss": 0.3741,
+      "step": 4131
+    },
+    {
+      "epoch": 1.039235412474849,
+      "grad_norm": 0.34148597717285156,
+      "learning_rate": 8.261819943427387e-06,
+      "loss": 0.3194,
+      "step": 4132
+    },
+    {
+      "epoch": 1.039486921529175,
+      "grad_norm": 0.3127982020378113,
+      "learning_rate": 8.26071079987946e-06,
+      "loss": 0.3364,
+      "step": 4133
+    },
+    {
+      "epoch": 1.039738430583501,
+      "grad_norm": 0.3347800374031067,
+      "learning_rate": 8.25960137707205e-06,
+      "loss": 0.3603,
+      "step": 4134
+    },
+    {
+      "epoch": 1.039989939637827,
+      "grad_norm": 0.3456893861293793,
+      "learning_rate": 8.258491675100175e-06,
+      "loss": 0.3734,
+      "step": 4135
+    },
+    {
+      "epoch": 1.040241448692153,
+      "grad_norm": 0.3339012861251831,
+      "learning_rate": 8.257381694058873e-06,
+      "loss": 0.3563,
+      "step": 4136
+    },
+    {
+      "epoch": 1.040492957746479,
+      "grad_norm": 0.31846827268600464,
+      "learning_rate": 8.256271434043206e-06,
+      "loss": 0.3469,
+      "step": 4137
+    },
+    {
+      "epoch": 1.0407444668008048,
+      "grad_norm": 0.3521193861961365,
+      "learning_rate": 8.255160895148263e-06,
+      "loss": 0.3225,
+      "step": 4138
+    },
+    {
+      "epoch": 1.0409959758551308,
+      "grad_norm": 0.34375718235969543,
+      "learning_rate": 8.254050077469153e-06,
+      "loss": 0.3366,
+      "step": 4139
+    },
+    {
+      "epoch": 1.0412474849094568,
+      "grad_norm": 0.3343333899974823,
+      "learning_rate": 8.252938981101011e-06,
+      "loss": 0.3674,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0414989939637826,
+      "grad_norm": 0.32967597246170044,
+      "learning_rate": 8.251827606138996e-06,
+      "loss": 0.3581,
+      "step": 4141
+    },
+    {
+      "epoch": 1.0417505030181087,
+      "grad_norm": 0.3318265676498413,
+      "learning_rate": 8.25071595267829e-06,
+      "loss": 0.3474,
+      "step": 4142
+    },
+    {
+      "epoch": 1.0420020120724347,
+      "grad_norm": 0.29342010617256165,
+      "learning_rate": 8.249604020814099e-06,
+      "loss": 0.3322,
+      "step": 4143
+    },
+    {
+      "epoch": 1.0422535211267605,
+      "grad_norm": 0.3520354628562927,
+      "learning_rate": 8.248491810641655e-06,
+      "loss": 0.3457,
+      "step": 4144
+    },
+    {
+      "epoch": 1.0425050301810865,
+      "grad_norm": 0.3410925269126892,
+      "learning_rate": 8.247379322256206e-06,
+      "loss": 0.3429,
+      "step": 4145
+    },
+    {
+      "epoch": 1.0427565392354126,
+      "grad_norm": 0.31338703632354736,
+      "learning_rate": 8.246266555753036e-06,
+      "loss": 0.3464,
+      "step": 4146
+    },
+    {
+      "epoch": 1.0430080482897384,
+      "grad_norm": 0.31659606099128723,
+      "learning_rate": 8.245153511227443e-06,
+      "loss": 0.3355,
+      "step": 4147
+    },
+    {
+      "epoch": 1.0432595573440644,
+      "grad_norm": 0.34237006306648254,
+      "learning_rate": 8.244040188774755e-06,
+      "loss": 0.3729,
+      "step": 4148
+    },
+    {
+      "epoch": 1.0435110663983904,
+      "grad_norm": 0.3474802076816559,
+      "learning_rate": 8.24292658849032e-06,
+      "loss": 0.3643,
+      "step": 4149
+    },
+    {
+      "epoch": 1.0437625754527162,
+      "grad_norm": 0.31629714369773865,
+      "learning_rate": 8.241812710469507e-06,
+      "loss": 0.3424,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0440140845070423,
+      "grad_norm": 0.3564571440219879,
+      "learning_rate": 8.240698554807717e-06,
+      "loss": 0.333,
+      "step": 4151
+    },
+    {
+      "epoch": 1.0442655935613683,
+      "grad_norm": 0.36076509952545166,
+      "learning_rate": 8.239584121600371e-06,
+      "loss": 0.3744,
+      "step": 4152
+    },
+    {
+      "epoch": 1.044517102615694,
+      "grad_norm": 0.33399519324302673,
+      "learning_rate": 8.238469410942911e-06,
+      "loss": 0.3488,
+      "step": 4153
+    },
+    {
+      "epoch": 1.04476861167002,
+      "grad_norm": 0.35272693634033203,
+      "learning_rate": 8.237354422930807e-06,
+      "loss": 0.3651,
+      "step": 4154
+    },
+    {
+      "epoch": 1.0450201207243461,
+      "grad_norm": 0.3702928125858307,
+      "learning_rate": 8.236239157659548e-06,
+      "loss": 0.3624,
+      "step": 4155
+    },
+    {
+      "epoch": 1.045271629778672,
+      "grad_norm": 0.3613063395023346,
+      "learning_rate": 8.235123615224651e-06,
+      "loss": 0.3579,
+      "step": 4156
+    },
+    {
+      "epoch": 1.045523138832998,
+      "grad_norm": 0.3584921956062317,
+      "learning_rate": 8.234007795721657e-06,
+      "loss": 0.3474,
+      "step": 4157
+    },
+    {
+      "epoch": 1.045774647887324,
+      "grad_norm": 0.3665167987346649,
+      "learning_rate": 8.232891699246126e-06,
+      "loss": 0.3479,
+      "step": 4158
+    },
+    {
+      "epoch": 1.0460261569416498,
+      "grad_norm": 0.3354139029979706,
+      "learning_rate": 8.231775325893646e-06,
+      "loss": 0.3608,
+      "step": 4159
+    },
+    {
+      "epoch": 1.0462776659959758,
+      "grad_norm": 0.31492966413497925,
+      "learning_rate": 8.230658675759827e-06,
+      "loss": 0.3536,
+      "step": 4160
+    },
+    {
+      "epoch": 1.0465291750503019,
+      "grad_norm": 0.3596387505531311,
+      "learning_rate": 8.229541748940301e-06,
+      "loss": 0.3646,
+      "step": 4161
+    },
+    {
+      "epoch": 1.0467806841046277,
+      "grad_norm": 0.3706818222999573,
+      "learning_rate": 8.22842454553073e-06,
+      "loss": 0.355,
+      "step": 4162
+    },
+    {
+      "epoch": 1.0470321931589537,
+      "grad_norm": 0.37192225456237793,
+      "learning_rate": 8.227307065626796e-06,
+      "loss": 0.3815,
+      "step": 4163
+    },
+    {
+      "epoch": 1.0472837022132797,
+      "grad_norm": 0.3455519378185272,
+      "learning_rate": 8.2261893093242e-06,
+      "loss": 0.363,
+      "step": 4164
+    },
+    {
+      "epoch": 1.0475352112676057,
+      "grad_norm": 0.33832308650016785,
+      "learning_rate": 8.225071276718672e-06,
+      "loss": 0.3476,
+      "step": 4165
+    },
+    {
+      "epoch": 1.0477867203219315,
+      "grad_norm": 0.3629179000854492,
+      "learning_rate": 8.223952967905967e-06,
+      "loss": 0.3711,
+      "step": 4166
+    },
+    {
+      "epoch": 1.0480382293762576,
+      "grad_norm": 0.3548809587955475,
+      "learning_rate": 8.222834382981858e-06,
+      "loss": 0.3507,
+      "step": 4167
+    },
+    {
+      "epoch": 1.0482897384305836,
+      "grad_norm": 0.3626018464565277,
+      "learning_rate": 8.221715522042146e-06,
+      "loss": 0.3621,
+      "step": 4168
+    },
+    {
+      "epoch": 1.0485412474849094,
+      "grad_norm": 0.310674786567688,
+      "learning_rate": 8.220596385182654e-06,
+      "loss": 0.3744,
+      "step": 4169
+    },
+    {
+      "epoch": 1.0487927565392354,
+      "grad_norm": 0.35073432326316833,
+      "learning_rate": 8.219476972499229e-06,
+      "loss": 0.3453,
+      "step": 4170
+    },
+    {
+      "epoch": 1.0490442655935615,
+      "grad_norm": 0.3294939696788788,
+      "learning_rate": 8.218357284087745e-06,
+      "loss": 0.3357,
+      "step": 4171
+    },
+    {
+      "epoch": 1.0492957746478873,
+      "grad_norm": 0.33905303478240967,
+      "learning_rate": 8.217237320044092e-06,
+      "loss": 0.3369,
+      "step": 4172
+    },
+    {
+      "epoch": 1.0495472837022133,
+      "grad_norm": 0.34386754035949707,
+      "learning_rate": 8.216117080464189e-06,
+      "loss": 0.3506,
+      "step": 4173
+    },
+    {
+      "epoch": 1.0497987927565393,
+      "grad_norm": 0.3818106949329376,
+      "learning_rate": 8.214996565443979e-06,
+      "loss": 0.3677,
+      "step": 4174
+    },
+    {
+      "epoch": 1.0500503018108651,
+      "grad_norm": 0.3344820439815521,
+      "learning_rate": 8.213875775079426e-06,
+      "loss": 0.3711,
+      "step": 4175
+    },
+    {
+      "epoch": 1.0503018108651911,
+      "grad_norm": 0.3421842157840729,
+      "learning_rate": 8.212754709466519e-06,
+      "loss": 0.3573,
+      "step": 4176
+    },
+    {
+      "epoch": 1.0505533199195172,
+      "grad_norm": 0.37867775559425354,
+      "learning_rate": 8.211633368701268e-06,
+      "loss": 0.3564,
+      "step": 4177
+    },
+    {
+      "epoch": 1.050804828973843,
+      "grad_norm": 0.3431204855442047,
+      "learning_rate": 8.210511752879713e-06,
+      "loss": 0.3577,
+      "step": 4178
+    },
+    {
+      "epoch": 1.051056338028169,
+      "grad_norm": 0.32079920172691345,
+      "learning_rate": 8.209389862097912e-06,
+      "loss": 0.3484,
+      "step": 4179
+    },
+    {
+      "epoch": 1.051307847082495,
+      "grad_norm": 0.3572438955307007,
+      "learning_rate": 8.208267696451947e-06,
+      "loss": 0.382,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0515593561368208,
+      "grad_norm": 0.3327823579311371,
+      "learning_rate": 8.207145256037922e-06,
+      "loss": 0.332,
+      "step": 4181
+    },
+    {
+      "epoch": 1.0518108651911469,
+      "grad_norm": 0.35914427042007446,
+      "learning_rate": 8.206022540951972e-06,
+      "loss": 0.3365,
+      "step": 4182
+    },
+    {
+      "epoch": 1.0520623742454729,
+      "grad_norm": 0.35635095834732056,
+      "learning_rate": 8.204899551290246e-06,
+      "loss": 0.366,
+      "step": 4183
+    },
+    {
+      "epoch": 1.0523138832997987,
+      "grad_norm": 0.3339400291442871,
+      "learning_rate": 8.203776287148925e-06,
+      "loss": 0.342,
+      "step": 4184
+    },
+    {
+      "epoch": 1.0525653923541247,
+      "grad_norm": 0.3433733582496643,
+      "learning_rate": 8.202652748624208e-06,
+      "loss": 0.3418,
+      "step": 4185
+    },
+    {
+      "epoch": 1.0528169014084507,
+      "grad_norm": 0.34672051668167114,
+      "learning_rate": 8.201528935812318e-06,
+      "loss": 0.3415,
+      "step": 4186
+    },
+    {
+      "epoch": 1.0530684104627768,
+      "grad_norm": 0.3377094566822052,
+      "learning_rate": 8.200404848809504e-06,
+      "loss": 0.3579,
+      "step": 4187
+    },
+    {
+      "epoch": 1.0533199195171026,
+      "grad_norm": 0.3439509868621826,
+      "learning_rate": 8.199280487712035e-06,
+      "loss": 0.3485,
+      "step": 4188
+    },
+    {
+      "epoch": 1.0535714285714286,
+      "grad_norm": 0.3966297507286072,
+      "learning_rate": 8.198155852616208e-06,
+      "loss": 0.381,
+      "step": 4189
+    },
+    {
+      "epoch": 1.0538229376257546,
+      "grad_norm": 0.3165939450263977,
+      "learning_rate": 8.19703094361834e-06,
+      "loss": 0.3462,
+      "step": 4190
+    },
+    {
+      "epoch": 1.0540744466800804,
+      "grad_norm": 0.30367371439933777,
+      "learning_rate": 8.195905760814772e-06,
+      "loss": 0.3467,
+      "step": 4191
+    },
+    {
+      "epoch": 1.0543259557344065,
+      "grad_norm": 0.34282615780830383,
+      "learning_rate": 8.194780304301869e-06,
+      "loss": 0.3466,
+      "step": 4192
+    },
+    {
+      "epoch": 1.0545774647887325,
+      "grad_norm": 0.3848336338996887,
+      "learning_rate": 8.19365457417602e-06,
+      "loss": 0.3636,
+      "step": 4193
+    },
+    {
+      "epoch": 1.0548289738430583,
+      "grad_norm": 0.32148125767707825,
+      "learning_rate": 8.192528570533636e-06,
+      "loss": 0.3611,
+      "step": 4194
+    },
+    {
+      "epoch": 1.0550804828973843,
+      "grad_norm": 0.35009652376174927,
+      "learning_rate": 8.191402293471151e-06,
+      "loss": 0.3771,
+      "step": 4195
+    },
+    {
+      "epoch": 1.0553319919517103,
+      "grad_norm": 0.37090542912483215,
+      "learning_rate": 8.190275743085025e-06,
+      "loss": 0.3681,
+      "step": 4196
+    },
+    {
+      "epoch": 1.0555835010060362,
+      "grad_norm": 0.3471124470233917,
+      "learning_rate": 8.18914891947174e-06,
+      "loss": 0.3665,
+      "step": 4197
+    },
+    {
+      "epoch": 1.0558350100603622,
+      "grad_norm": 0.34268495440483093,
+      "learning_rate": 8.188021822727804e-06,
+      "loss": 0.3778,
+      "step": 4198
+    },
+    {
+      "epoch": 1.0560865191146882,
+      "grad_norm": 0.35175612568855286,
+      "learning_rate": 8.18689445294974e-06,
+      "loss": 0.3611,
+      "step": 4199
+    },
+    {
+      "epoch": 1.056338028169014,
+      "grad_norm": 0.3248383104801178,
+      "learning_rate": 8.185766810234106e-06,
+      "loss": 0.3624,
+      "step": 4200
+    },
+    {
+      "epoch": 1.05658953722334,
+      "grad_norm": 0.35728633403778076,
+      "learning_rate": 8.184638894677472e-06,
+      "loss": 0.345,
+      "step": 4201
+    },
+    {
+      "epoch": 1.056841046277666,
+      "grad_norm": 0.3417149782180786,
+      "learning_rate": 8.183510706376441e-06,
+      "loss": 0.3637,
+      "step": 4202
+    },
+    {
+      "epoch": 1.0570925553319919,
+      "grad_norm": 0.37575763463974,
+      "learning_rate": 8.182382245427634e-06,
+      "loss": 0.3408,
+      "step": 4203
+    },
+    {
+      "epoch": 1.057344064386318,
+      "grad_norm": 0.37389978766441345,
+      "learning_rate": 8.181253511927696e-06,
+      "loss": 0.3252,
+      "step": 4204
+    },
+    {
+      "epoch": 1.057595573440644,
+      "grad_norm": 0.35000425577163696,
+      "learning_rate": 8.180124505973299e-06,
+      "loss": 0.3451,
+      "step": 4205
+    },
+    {
+      "epoch": 1.0578470824949697,
+      "grad_norm": 0.3386198878288269,
+      "learning_rate": 8.17899522766113e-06,
+      "loss": 0.3179,
+      "step": 4206
+    },
+    {
+      "epoch": 1.0580985915492958,
+      "grad_norm": 0.3945431411266327,
+      "learning_rate": 8.177865677087908e-06,
+      "loss": 0.3231,
+      "step": 4207
+    },
+    {
+      "epoch": 1.0583501006036218,
+      "grad_norm": 0.34005576372146606,
+      "learning_rate": 8.176735854350373e-06,
+      "loss": 0.347,
+      "step": 4208
+    },
+    {
+      "epoch": 1.0586016096579476,
+      "grad_norm": 0.36175423860549927,
+      "learning_rate": 8.175605759545285e-06,
+      "loss": 0.3508,
+      "step": 4209
+    },
+    {
+      "epoch": 1.0588531187122736,
+      "grad_norm": 0.35928767919540405,
+      "learning_rate": 8.17447539276943e-06,
+      "loss": 0.3871,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0591046277665996,
+      "grad_norm": 0.4204372763633728,
+      "learning_rate": 8.173344754119615e-06,
+      "loss": 0.3536,
+      "step": 4211
+    },
+    {
+      "epoch": 1.0593561368209254,
+      "grad_norm": 0.3679214119911194,
+      "learning_rate": 8.172213843692676e-06,
+      "loss": 0.3841,
+      "step": 4212
+    },
+    {
+      "epoch": 1.0596076458752515,
+      "grad_norm": 0.3367496132850647,
+      "learning_rate": 8.171082661585468e-06,
+      "loss": 0.3806,
+      "step": 4213
+    },
+    {
+      "epoch": 1.0598591549295775,
+      "grad_norm": 0.3853965401649475,
+      "learning_rate": 8.169951207894866e-06,
+      "loss": 0.35,
+      "step": 4214
+    },
+    {
+      "epoch": 1.0601106639839035,
+      "grad_norm": 0.3903757929801941,
+      "learning_rate": 8.168819482717775e-06,
+      "loss": 0.3529,
+      "step": 4215
+    },
+    {
+      "epoch": 1.0603621730382293,
+      "grad_norm": 0.35560932755470276,
+      "learning_rate": 8.167687486151119e-06,
+      "loss": 0.3368,
+      "step": 4216
+    },
+    {
+      "epoch": 1.0606136820925554,
+      "grad_norm": 0.3805678188800812,
+      "learning_rate": 8.166555218291847e-06,
+      "loss": 0.332,
+      "step": 4217
+    },
+    {
+      "epoch": 1.0608651911468814,
+      "grad_norm": 0.32614269852638245,
+      "learning_rate": 8.16542267923693e-06,
+      "loss": 0.3387,
+      "step": 4218
+    },
+    {
+      "epoch": 1.0611167002012072,
+      "grad_norm": 0.35080480575561523,
+      "learning_rate": 8.164289869083365e-06,
+      "loss": 0.3367,
+      "step": 4219
+    },
+    {
+      "epoch": 1.0613682092555332,
+      "grad_norm": 0.3625401556491852,
+      "learning_rate": 8.163156787928169e-06,
+      "loss": 0.3508,
+      "step": 4220
+    },
+    {
+      "epoch": 1.0616197183098592,
+      "grad_norm": 0.4160093367099762,
+      "learning_rate": 8.162023435868381e-06,
+      "loss": 0.3577,
+      "step": 4221
+    },
+    {
+      "epoch": 1.061871227364185,
+      "grad_norm": 0.3521963059902191,
+      "learning_rate": 8.160889813001066e-06,
+      "loss": 0.3468,
+      "step": 4222
+    },
+    {
+      "epoch": 1.062122736418511,
+      "grad_norm": 0.37421825528144836,
+      "learning_rate": 8.159755919423315e-06,
+      "loss": 0.3796,
+      "step": 4223
+    },
+    {
+      "epoch": 1.062374245472837,
+      "grad_norm": 0.34013405442237854,
+      "learning_rate": 8.158621755232237e-06,
+      "loss": 0.3479,
+      "step": 4224
+    },
+    {
+      "epoch": 1.062625754527163,
+      "grad_norm": 0.376768559217453,
+      "learning_rate": 8.157487320524964e-06,
+      "loss": 0.3701,
+      "step": 4225
+    },
+    {
+      "epoch": 1.062877263581489,
+      "grad_norm": 0.36440804600715637,
+      "learning_rate": 8.156352615398658e-06,
+      "loss": 0.386,
+      "step": 4226
+    },
+    {
+      "epoch": 1.063128772635815,
+      "grad_norm": 0.3305298089981079,
+      "learning_rate": 8.155217639950494e-06,
+      "loss": 0.3242,
+      "step": 4227
+    },
+    {
+      "epoch": 1.0633802816901408,
+      "grad_norm": 0.33112478256225586,
+      "learning_rate": 8.154082394277678e-06,
+      "loss": 0.3429,
+      "step": 4228
+    },
+    {
+      "epoch": 1.0636317907444668,
+      "grad_norm": 0.3222111165523529,
+      "learning_rate": 8.15294687847744e-06,
+      "loss": 0.3478,
+      "step": 4229
+    },
+    {
+      "epoch": 1.0638832997987928,
+      "grad_norm": 0.3655967116355896,
+      "learning_rate": 8.151811092647024e-06,
+      "loss": 0.3619,
+      "step": 4230
+    },
+    {
+      "epoch": 1.0641348088531186,
+      "grad_norm": 0.3364644944667816,
+      "learning_rate": 8.150675036883705e-06,
+      "loss": 0.3434,
+      "step": 4231
+    },
+    {
+      "epoch": 1.0643863179074446,
+      "grad_norm": 0.34147587418556213,
+      "learning_rate": 8.149538711284782e-06,
+      "loss": 0.3561,
+      "step": 4232
+    },
+    {
+      "epoch": 1.0646378269617707,
+      "grad_norm": 0.33940380811691284,
+      "learning_rate": 8.14840211594757e-06,
+      "loss": 0.3664,
+      "step": 4233
+    },
+    {
+      "epoch": 1.0648893360160965,
+      "grad_norm": 0.3298323154449463,
+      "learning_rate": 8.147265250969415e-06,
+      "loss": 0.3362,
+      "step": 4234
+    },
+    {
+      "epoch": 1.0651408450704225,
+      "grad_norm": 0.34013256430625916,
+      "learning_rate": 8.146128116447679e-06,
+      "loss": 0.3753,
+      "step": 4235
+    },
+    {
+      "epoch": 1.0653923541247485,
+      "grad_norm": 0.3295930325984955,
+      "learning_rate": 8.144990712479753e-06,
+      "loss": 0.3482,
+      "step": 4236
+    },
+    {
+      "epoch": 1.0656438631790746,
+      "grad_norm": 0.33826765418052673,
+      "learning_rate": 8.143853039163046e-06,
+      "loss": 0.3402,
+      "step": 4237
+    },
+    {
+      "epoch": 1.0658953722334004,
+      "grad_norm": 0.35167795419692993,
+      "learning_rate": 8.142715096594994e-06,
+      "loss": 0.3993,
+      "step": 4238
+    },
+    {
+      "epoch": 1.0661468812877264,
+      "grad_norm": 0.35461968183517456,
+      "learning_rate": 8.141576884873054e-06,
+      "loss": 0.3738,
+      "step": 4239
+    },
+    {
+      "epoch": 1.0663983903420524,
+      "grad_norm": 0.3144620358943939,
+      "learning_rate": 8.140438404094711e-06,
+      "loss": 0.3594,
+      "step": 4240
+    },
+    {
+      "epoch": 1.0666498993963782,
+      "grad_norm": 0.33328554034233093,
+      "learning_rate": 8.139299654357462e-06,
+      "loss": 0.3361,
+      "step": 4241
+    },
+    {
+      "epoch": 1.0669014084507042,
+      "grad_norm": 0.4244604706764221,
+      "learning_rate": 8.138160635758839e-06,
+      "loss": 0.3409,
+      "step": 4242
+    },
+    {
+      "epoch": 1.0671529175050303,
+      "grad_norm": 0.336699515581131,
+      "learning_rate": 8.137021348396389e-06,
+      "loss": 0.3496,
+      "step": 4243
+    },
+    {
+      "epoch": 1.067404426559356,
+      "grad_norm": 0.3206990361213684,
+      "learning_rate": 8.135881792367686e-06,
+      "loss": 0.3369,
+      "step": 4244
+    },
+    {
+      "epoch": 1.067655935613682,
+      "grad_norm": 0.34667348861694336,
+      "learning_rate": 8.134741967770325e-06,
+      "loss": 0.3572,
+      "step": 4245
+    },
+    {
+      "epoch": 1.0679074446680081,
+      "grad_norm": 0.3545694351196289,
+      "learning_rate": 8.133601874701926e-06,
+      "loss": 0.3533,
+      "step": 4246
+    },
+    {
+      "epoch": 1.068158953722334,
+      "grad_norm": 0.33755382895469666,
+      "learning_rate": 8.13246151326013e-06,
+      "loss": 0.3517,
+      "step": 4247
+    },
+    {
+      "epoch": 1.06841046277666,
+      "grad_norm": 0.33736327290534973,
+      "learning_rate": 8.131320883542601e-06,
+      "loss": 0.3281,
+      "step": 4248
+    },
+    {
+      "epoch": 1.068661971830986,
+      "grad_norm": 0.3781490623950958,
+      "learning_rate": 8.13017998564703e-06,
+      "loss": 0.3733,
+      "step": 4249
+    },
+    {
+      "epoch": 1.0689134808853118,
+      "grad_norm": 0.35155022144317627,
+      "learning_rate": 8.129038819671122e-06,
+      "loss": 0.3716,
+      "step": 4250
+    },
+    {
+      "epoch": 1.0691649899396378,
+      "grad_norm": 0.36461761593818665,
+      "learning_rate": 8.127897385712616e-06,
+      "loss": 0.3367,
+      "step": 4251
+    },
+    {
+      "epoch": 1.0694164989939638,
+      "grad_norm": 0.3492373526096344,
+      "learning_rate": 8.126755683869267e-06,
+      "loss": 0.3376,
+      "step": 4252
+    },
+    {
+      "epoch": 1.0696680080482897,
+      "grad_norm": 0.3286789357662201,
+      "learning_rate": 8.125613714238855e-06,
+      "loss": 0.3334,
+      "step": 4253
+    },
+    {
+      "epoch": 1.0699195171026157,
+      "grad_norm": 0.33751797676086426,
+      "learning_rate": 8.124471476919183e-06,
+      "loss": 0.3599,
+      "step": 4254
+    },
+    {
+      "epoch": 1.0701710261569417,
+      "grad_norm": 0.37270689010620117,
+      "learning_rate": 8.123328972008075e-06,
+      "loss": 0.3753,
+      "step": 4255
+    },
+    {
+      "epoch": 1.0704225352112675,
+      "grad_norm": 0.3367967903614044,
+      "learning_rate": 8.122186199603378e-06,
+      "loss": 0.3578,
+      "step": 4256
+    },
+    {
+      "epoch": 1.0706740442655935,
+      "grad_norm": 0.3527868986129761,
+      "learning_rate": 8.121043159802969e-06,
+      "loss": 0.3491,
+      "step": 4257
+    },
+    {
+      "epoch": 1.0709255533199196,
+      "grad_norm": 0.3186793029308319,
+      "learning_rate": 8.119899852704736e-06,
+      "loss": 0.3345,
+      "step": 4258
+    },
+    {
+      "epoch": 1.0711770623742454,
+      "grad_norm": 0.34226033091545105,
+      "learning_rate": 8.1187562784066e-06,
+      "loss": 0.371,
+      "step": 4259
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.36788326501846313,
+      "learning_rate": 8.1176124370065e-06,
+      "loss": 0.3486,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0716800804828974,
+      "grad_norm": 0.33759674429893494,
+      "learning_rate": 8.116468328602397e-06,
+      "loss": 0.3845,
+      "step": 4261
+    },
+    {
+      "epoch": 1.0719315895372232,
+      "grad_norm": 0.32118430733680725,
+      "learning_rate": 8.115323953292278e-06,
+      "loss": 0.3362,
+      "step": 4262
+    },
+    {
+      "epoch": 1.0721830985915493,
+      "grad_norm": 0.3653360605239868,
+      "learning_rate": 8.114179311174154e-06,
+      "loss": 0.3502,
+      "step": 4263
+    },
+    {
+      "epoch": 1.0724346076458753,
+      "grad_norm": 0.32888534665107727,
+      "learning_rate": 8.113034402346052e-06,
+      "loss": 0.3454,
+      "step": 4264
+    },
+    {
+      "epoch": 1.0726861167002013,
+      "grad_norm": 0.3332807719707489,
+      "learning_rate": 8.11188922690603e-06,
+      "loss": 0.3384,
+      "step": 4265
+    },
+    {
+      "epoch": 1.0729376257545271,
+      "grad_norm": 0.36372309923171997,
+      "learning_rate": 8.110743784952162e-06,
+      "loss": 0.3606,
+      "step": 4266
+    },
+    {
+      "epoch": 1.0731891348088531,
+      "grad_norm": 0.3299432396888733,
+      "learning_rate": 8.10959807658255e-06,
+      "loss": 0.3564,
+      "step": 4267
+    },
+    {
+      "epoch": 1.0734406438631792,
+      "grad_norm": 0.3273921012878418,
+      "learning_rate": 8.108452101895317e-06,
+      "loss": 0.33,
+      "step": 4268
+    },
+    {
+      "epoch": 1.073692152917505,
+      "grad_norm": 0.3181513547897339,
+      "learning_rate": 8.107305860988608e-06,
+      "loss": 0.3416,
+      "step": 4269
+    },
+    {
+      "epoch": 1.073943661971831,
+      "grad_norm": 0.3415665924549103,
+      "learning_rate": 8.10615935396059e-06,
+      "loss": 0.353,
+      "step": 4270
+    },
+    {
+      "epoch": 1.074195171026157,
+      "grad_norm": 0.34259921312332153,
+      "learning_rate": 8.105012580909457e-06,
+      "loss": 0.373,
+      "step": 4271
+    },
+    {
+      "epoch": 1.0744466800804828,
+      "grad_norm": 0.3158281147480011,
+      "learning_rate": 8.103865541933421e-06,
+      "loss": 0.3496,
+      "step": 4272
+    },
+    {
+      "epoch": 1.0746981891348089,
+      "grad_norm": 0.32061338424682617,
+      "learning_rate": 8.102718237130718e-06,
+      "loss": 0.3632,
+      "step": 4273
+    },
+    {
+      "epoch": 1.0749496981891349,
+      "grad_norm": 0.35798168182373047,
+      "learning_rate": 8.101570666599608e-06,
+      "loss": 0.3756,
+      "step": 4274
+    },
+    {
+      "epoch": 1.0752012072434607,
+      "grad_norm": 0.30440935492515564,
+      "learning_rate": 8.100422830438376e-06,
+      "loss": 0.359,
+      "step": 4275
+    },
+    {
+      "epoch": 1.0754527162977867,
+      "grad_norm": 0.31603163480758667,
+      "learning_rate": 8.099274728745324e-06,
+      "loss": 0.339,
+      "step": 4276
+    },
+    {
+      "epoch": 1.0757042253521127,
+      "grad_norm": 0.3540377914905548,
+      "learning_rate": 8.09812636161878e-06,
+      "loss": 0.3457,
+      "step": 4277
+    },
+    {
+      "epoch": 1.0759557344064385,
+      "grad_norm": 0.33938080072402954,
+      "learning_rate": 8.096977729157096e-06,
+      "loss": 0.3666,
+      "step": 4278
+    },
+    {
+      "epoch": 1.0762072434607646,
+      "grad_norm": 0.36812707781791687,
+      "learning_rate": 8.095828831458643e-06,
+      "loss": 0.3479,
+      "step": 4279
+    },
+    {
+      "epoch": 1.0764587525150906,
+      "grad_norm": 0.33766672015190125,
+      "learning_rate": 8.094679668621818e-06,
+      "loss": 0.3474,
+      "step": 4280
+    },
+    {
+      "epoch": 1.0767102615694164,
+      "grad_norm": 0.3450581133365631,
+      "learning_rate": 8.09353024074504e-06,
+      "loss": 0.3536,
+      "step": 4281
+    },
+    {
+      "epoch": 1.0769617706237424,
+      "grad_norm": 0.37925583124160767,
+      "learning_rate": 8.09238054792675e-06,
+      "loss": 0.347,
+      "step": 4282
+    },
+    {
+      "epoch": 1.0772132796780685,
+      "grad_norm": 0.35549673438072205,
+      "learning_rate": 8.091230590265411e-06,
+      "loss": 0.3493,
+      "step": 4283
+    },
+    {
+      "epoch": 1.0774647887323943,
+      "grad_norm": 0.317950040102005,
+      "learning_rate": 8.090080367859512e-06,
+      "loss": 0.3477,
+      "step": 4284
+    },
+    {
+      "epoch": 1.0777162977867203,
+      "grad_norm": 0.3614874482154846,
+      "learning_rate": 8.08892988080756e-06,
+      "loss": 0.36,
+      "step": 4285
+    },
+    {
+      "epoch": 1.0779678068410463,
+      "grad_norm": 0.35449376702308655,
+      "learning_rate": 8.087779129208088e-06,
+      "loss": 0.371,
+      "step": 4286
+    },
+    {
+      "epoch": 1.0782193158953723,
+      "grad_norm": 0.32577499747276306,
+      "learning_rate": 8.086628113159651e-06,
+      "loss": 0.3402,
+      "step": 4287
+    },
+    {
+      "epoch": 1.0784708249496981,
+      "grad_norm": 0.3384754955768585,
+      "learning_rate": 8.085476832760828e-06,
+      "loss": 0.3697,
+      "step": 4288
+    },
+    {
+      "epoch": 1.0787223340040242,
+      "grad_norm": 0.3504764139652252,
+      "learning_rate": 8.084325288110215e-06,
+      "loss": 0.3467,
+      "step": 4289
+    },
+    {
+      "epoch": 1.0789738430583502,
+      "grad_norm": 0.3410959243774414,
+      "learning_rate": 8.083173479306436e-06,
+      "loss": 0.3749,
+      "step": 4290
+    },
+    {
+      "epoch": 1.079225352112676,
+      "grad_norm": 0.34386518597602844,
+      "learning_rate": 8.082021406448137e-06,
+      "loss": 0.3713,
+      "step": 4291
+    },
+    {
+      "epoch": 1.079476861167002,
+      "grad_norm": 0.3209969699382782,
+      "learning_rate": 8.080869069633987e-06,
+      "loss": 0.339,
+      "step": 4292
+    },
+    {
+      "epoch": 1.079728370221328,
+      "grad_norm": 0.3502729833126068,
+      "learning_rate": 8.079716468962673e-06,
+      "loss": 0.3597,
+      "step": 4293
+    },
+    {
+      "epoch": 1.0799798792756539,
+      "grad_norm": 0.3301171064376831,
+      "learning_rate": 8.078563604532911e-06,
+      "loss": 0.3534,
+      "step": 4294
+    },
+    {
+      "epoch": 1.08023138832998,
+      "grad_norm": 0.38163283467292786,
+      "learning_rate": 8.077410476443436e-06,
+      "loss": 0.3326,
+      "step": 4295
+    },
+    {
+      "epoch": 1.080482897384306,
+      "grad_norm": 0.30207228660583496,
+      "learning_rate": 8.076257084793007e-06,
+      "loss": 0.3437,
+      "step": 4296
+    },
+    {
+      "epoch": 1.0807344064386317,
+      "grad_norm": 0.34427884221076965,
+      "learning_rate": 8.075103429680402e-06,
+      "loss": 0.3593,
+      "step": 4297
+    },
+    {
+      "epoch": 1.0809859154929577,
+      "grad_norm": 0.40237826108932495,
+      "learning_rate": 8.073949511204426e-06,
+      "loss": 0.3564,
+      "step": 4298
+    },
+    {
+      "epoch": 1.0812374245472838,
+      "grad_norm": 0.34032902121543884,
+      "learning_rate": 8.072795329463907e-06,
+      "loss": 0.3585,
+      "step": 4299
+    },
+    {
+      "epoch": 1.0814889336016096,
+      "grad_norm": 0.3582513630390167,
+      "learning_rate": 8.07164088455769e-06,
+      "loss": 0.3272,
+      "step": 4300
+    },
+    {
+      "epoch": 1.0817404426559356,
+      "grad_norm": 0.3856683075428009,
+      "learning_rate": 8.070486176584647e-06,
+      "loss": 0.3603,
+      "step": 4301
+    },
+    {
+      "epoch": 1.0819919517102616,
+      "grad_norm": 0.3110501766204834,
+      "learning_rate": 8.069331205643671e-06,
+      "loss": 0.3748,
+      "step": 4302
+    },
+    {
+      "epoch": 1.0822434607645874,
+      "grad_norm": 0.3185417950153351,
+      "learning_rate": 8.068175971833679e-06,
+      "loss": 0.3331,
+      "step": 4303
+    },
+    {
+      "epoch": 1.0824949698189135,
+      "grad_norm": 0.33823883533477783,
+      "learning_rate": 8.06702047525361e-06,
+      "loss": 0.3492,
+      "step": 4304
+    },
+    {
+      "epoch": 1.0827464788732395,
+      "grad_norm": 0.3247283399105072,
+      "learning_rate": 8.065864716002426e-06,
+      "loss": 0.3648,
+      "step": 4305
+    },
+    {
+      "epoch": 1.0829979879275653,
+      "grad_norm": 0.35124877095222473,
+      "learning_rate": 8.064708694179107e-06,
+      "loss": 0.345,
+      "step": 4306
+    },
+    {
+      "epoch": 1.0832494969818913,
+      "grad_norm": 0.3573397696018219,
+      "learning_rate": 8.063552409882662e-06,
+      "loss": 0.3676,
+      "step": 4307
+    },
+    {
+      "epoch": 1.0835010060362174,
+      "grad_norm": 0.33024418354034424,
+      "learning_rate": 8.06239586321212e-06,
+      "loss": 0.3826,
+      "step": 4308
+    },
+    {
+      "epoch": 1.0837525150905432,
+      "grad_norm": 0.33846285939216614,
+      "learning_rate": 8.06123905426653e-06,
+      "loss": 0.3585,
+      "step": 4309
+    },
+    {
+      "epoch": 1.0840040241448692,
+      "grad_norm": 0.34866371750831604,
+      "learning_rate": 8.060081983144964e-06,
+      "loss": 0.3492,
+      "step": 4310
+    },
+    {
+      "epoch": 1.0842555331991952,
+      "grad_norm": 0.3329204320907593,
+      "learning_rate": 8.058924649946523e-06,
+      "loss": 0.3892,
+      "step": 4311
+    },
+    {
+      "epoch": 1.084507042253521,
+      "grad_norm": 0.32447880506515503,
+      "learning_rate": 8.05776705477032e-06,
+      "loss": 0.3566,
+      "step": 4312
+    },
+    {
+      "epoch": 1.084758551307847,
+      "grad_norm": 0.33247438073158264,
+      "learning_rate": 8.0566091977155e-06,
+      "loss": 0.3458,
+      "step": 4313
+    },
+    {
+      "epoch": 1.085010060362173,
+      "grad_norm": 0.3221612572669983,
+      "learning_rate": 8.055451078881221e-06,
+      "loss": 0.3629,
+      "step": 4314
+    },
+    {
+      "epoch": 1.085261569416499,
+      "grad_norm": 0.3234616816043854,
+      "learning_rate": 8.054292698366674e-06,
+      "loss": 0.3566,
+      "step": 4315
+    },
+    {
+      "epoch": 1.085513078470825,
+      "grad_norm": 0.3308125436306,
+      "learning_rate": 8.053134056271064e-06,
+      "loss": 0.3519,
+      "step": 4316
+    },
+    {
+      "epoch": 1.085764587525151,
+      "grad_norm": 0.3306364119052887,
+      "learning_rate": 8.051975152693623e-06,
+      "loss": 0.3253,
+      "step": 4317
+    },
+    {
+      "epoch": 1.086016096579477,
+      "grad_norm": 0.343723326921463,
+      "learning_rate": 8.050815987733604e-06,
+      "loss": 0.3455,
+      "step": 4318
+    },
+    {
+      "epoch": 1.0862676056338028,
+      "grad_norm": 0.36126473546028137,
+      "learning_rate": 8.049656561490282e-06,
+      "loss": 0.3846,
+      "step": 4319
+    },
+    {
+      "epoch": 1.0865191146881288,
+      "grad_norm": 0.3083570897579193,
+      "learning_rate": 8.048496874062953e-06,
+      "loss": 0.3608,
+      "step": 4320
+    },
+    {
+      "epoch": 1.0867706237424548,
+      "grad_norm": 0.33808434009552,
+      "learning_rate": 8.04733692555094e-06,
+      "loss": 0.3466,
+      "step": 4321
+    },
+    {
+      "epoch": 1.0870221327967806,
+      "grad_norm": 0.3496091067790985,
+      "learning_rate": 8.04617671605358e-06,
+      "loss": 0.3795,
+      "step": 4322
+    },
+    {
+      "epoch": 1.0872736418511066,
+      "grad_norm": 0.34945693612098694,
+      "learning_rate": 8.045016245670243e-06,
+      "loss": 0.354,
+      "step": 4323
+    },
+    {
+      "epoch": 1.0875251509054327,
+      "grad_norm": 0.3567672669887543,
+      "learning_rate": 8.043855514500314e-06,
+      "loss": 0.3537,
+      "step": 4324
+    },
+    {
+      "epoch": 1.0877766599597585,
+      "grad_norm": 0.35311269760131836,
+      "learning_rate": 8.042694522643202e-06,
+      "loss": 0.3536,
+      "step": 4325
+    },
+    {
+      "epoch": 1.0880281690140845,
+      "grad_norm": 0.33748674392700195,
+      "learning_rate": 8.041533270198341e-06,
+      "loss": 0.371,
+      "step": 4326
+    },
+    {
+      "epoch": 1.0882796780684105,
+      "grad_norm": 0.35377809405326843,
+      "learning_rate": 8.04037175726518e-06,
+      "loss": 0.3517,
+      "step": 4327
+    },
+    {
+      "epoch": 1.0885311871227363,
+      "grad_norm": 0.3475686311721802,
+      "learning_rate": 8.039209983943201e-06,
+      "loss": 0.3564,
+      "step": 4328
+    },
+    {
+      "epoch": 1.0887826961770624,
+      "grad_norm": 0.35228431224823,
+      "learning_rate": 8.0380479503319e-06,
+      "loss": 0.3494,
+      "step": 4329
+    },
+    {
+      "epoch": 1.0890342052313884,
+      "grad_norm": 0.31928402185440063,
+      "learning_rate": 8.036885656530797e-06,
+      "loss": 0.3457,
+      "step": 4330
+    },
+    {
+      "epoch": 1.0892857142857142,
+      "grad_norm": 0.3187348246574402,
+      "learning_rate": 8.035723102639437e-06,
+      "loss": 0.3289,
+      "step": 4331
+    },
+    {
+      "epoch": 1.0895372233400402,
+      "grad_norm": 0.3454899191856384,
+      "learning_rate": 8.034560288757386e-06,
+      "loss": 0.3655,
+      "step": 4332
+    },
+    {
+      "epoch": 1.0897887323943662,
+      "grad_norm": 0.3134077787399292,
+      "learning_rate": 8.033397214984226e-06,
+      "loss": 0.35,
+      "step": 4333
+    },
+    {
+      "epoch": 1.0900402414486923,
+      "grad_norm": 0.326066792011261,
+      "learning_rate": 8.032233881419576e-06,
+      "loss": 0.3629,
+      "step": 4334
+    },
+    {
+      "epoch": 1.090291750503018,
+      "grad_norm": 0.34841668605804443,
+      "learning_rate": 8.031070288163061e-06,
+      "loss": 0.3792,
+      "step": 4335
+    },
+    {
+      "epoch": 1.090543259557344,
+      "grad_norm": 0.3101913034915924,
+      "learning_rate": 8.029906435314339e-06,
+      "loss": 0.3148,
+      "step": 4336
+    },
+    {
+      "epoch": 1.0907947686116701,
+      "grad_norm": 0.36832088232040405,
+      "learning_rate": 8.028742322973085e-06,
+      "loss": 0.3647,
+      "step": 4337
+    },
+    {
+      "epoch": 1.091046277665996,
+      "grad_norm": 0.3566989004611969,
+      "learning_rate": 8.027577951238999e-06,
+      "loss": 0.3651,
+      "step": 4338
+    },
+    {
+      "epoch": 1.091297786720322,
+      "grad_norm": 0.3333961069583893,
+      "learning_rate": 8.026413320211804e-06,
+      "loss": 0.3356,
+      "step": 4339
+    },
+    {
+      "epoch": 1.091549295774648,
+      "grad_norm": 0.38320064544677734,
+      "learning_rate": 8.02524842999124e-06,
+      "loss": 0.3819,
+      "step": 4340
+    },
+    {
+      "epoch": 1.0918008048289738,
+      "grad_norm": 0.3310604989528656,
+      "learning_rate": 8.024083280677073e-06,
+      "loss": 0.3626,
+      "step": 4341
+    },
+    {
+      "epoch": 1.0920523138832998,
+      "grad_norm": 0.3529944121837616,
+      "learning_rate": 8.02291787236909e-06,
+      "loss": 0.3707,
+      "step": 4342
+    },
+    {
+      "epoch": 1.0923038229376258,
+      "grad_norm": 0.3804549276828766,
+      "learning_rate": 8.021752205167108e-06,
+      "loss": 0.3516,
+      "step": 4343
+    },
+    {
+      "epoch": 1.0925553319919517,
+      "grad_norm": 0.3501472473144531,
+      "learning_rate": 8.02058627917095e-06,
+      "loss": 0.3521,
+      "step": 4344
+    },
+    {
+      "epoch": 1.0928068410462777,
+      "grad_norm": 0.36724236607551575,
+      "learning_rate": 8.019420094480475e-06,
+      "loss": 0.3649,
+      "step": 4345
+    },
+    {
+      "epoch": 1.0930583501006037,
+      "grad_norm": 0.3527858555316925,
+      "learning_rate": 8.018253651195556e-06,
+      "loss": 0.36,
+      "step": 4346
+    },
+    {
+      "epoch": 1.0933098591549295,
+      "grad_norm": 0.37071692943573,
+      "learning_rate": 8.017086949416095e-06,
+      "loss": 0.3735,
+      "step": 4347
+    },
+    {
+      "epoch": 1.0935613682092555,
+      "grad_norm": 0.3491380512714386,
+      "learning_rate": 8.015919989242014e-06,
+      "loss": 0.3577,
+      "step": 4348
+    },
+    {
+      "epoch": 1.0938128772635816,
+      "grad_norm": 0.36504706740379333,
+      "learning_rate": 8.014752770773252e-06,
+      "loss": 0.3314,
+      "step": 4349
+    },
+    {
+      "epoch": 1.0940643863179074,
+      "grad_norm": 0.36621904373168945,
+      "learning_rate": 8.013585294109773e-06,
+      "loss": 0.3417,
+      "step": 4350
+    },
+    {
+      "epoch": 1.0943158953722334,
+      "grad_norm": 0.35412460565567017,
+      "learning_rate": 8.012417559351569e-06,
+      "loss": 0.3774,
+      "step": 4351
+    },
+    {
+      "epoch": 1.0945674044265594,
+      "grad_norm": 0.37957510352134705,
+      "learning_rate": 8.011249566598647e-06,
+      "loss": 0.384,
+      "step": 4352
+    },
+    {
+      "epoch": 1.0948189134808852,
+      "grad_norm": 0.34848281741142273,
+      "learning_rate": 8.010081315951037e-06,
+      "loss": 0.3393,
+      "step": 4353
+    },
+    {
+      "epoch": 1.0950704225352113,
+      "grad_norm": 0.33474472165107727,
+      "learning_rate": 8.008912807508794e-06,
+      "loss": 0.351,
+      "step": 4354
+    },
+    {
+      "epoch": 1.0953219315895373,
+      "grad_norm": 0.34298956394195557,
+      "learning_rate": 8.007744041371993e-06,
+      "loss": 0.337,
+      "step": 4355
+    },
+    {
+      "epoch": 1.095573440643863,
+      "grad_norm": 0.3466607332229614,
+      "learning_rate": 8.00657501764073e-06,
+      "loss": 0.3387,
+      "step": 4356
+    },
+    {
+      "epoch": 1.095824949698189,
+      "grad_norm": 0.3456454575061798,
+      "learning_rate": 8.005405736415127e-06,
+      "loss": 0.3355,
+      "step": 4357
+    },
+    {
+      "epoch": 1.0960764587525151,
+      "grad_norm": 0.3142288327217102,
+      "learning_rate": 8.004236197795323e-06,
+      "loss": 0.3811,
+      "step": 4358
+    },
+    {
+      "epoch": 1.096327967806841,
+      "grad_norm": 0.3510921597480774,
+      "learning_rate": 8.003066401881484e-06,
+      "loss": 0.3682,
+      "step": 4359
+    },
+    {
+      "epoch": 1.096579476861167,
+      "grad_norm": 0.37072980403900146,
+      "learning_rate": 8.001896348773795e-06,
+      "loss": 0.3559,
+      "step": 4360
+    },
+    {
+      "epoch": 1.096830985915493,
+      "grad_norm": 0.3439767062664032,
+      "learning_rate": 8.000726038572463e-06,
+      "loss": 0.3517,
+      "step": 4361
+    },
+    {
+      "epoch": 1.0970824949698188,
+      "grad_norm": 0.3286254405975342,
+      "learning_rate": 7.999555471377719e-06,
+      "loss": 0.3587,
+      "step": 4362
+    },
+    {
+      "epoch": 1.0973340040241448,
+      "grad_norm": 0.33705952763557434,
+      "learning_rate": 7.998384647289813e-06,
+      "loss": 0.3666,
+      "step": 4363
+    },
+    {
+      "epoch": 1.0975855130784709,
+      "grad_norm": 0.3337515890598297,
+      "learning_rate": 7.997213566409022e-06,
+      "loss": 0.3502,
+      "step": 4364
+    },
+    {
+      "epoch": 1.0978370221327969,
+      "grad_norm": 0.3719703257083893,
+      "learning_rate": 7.996042228835637e-06,
+      "loss": 0.3815,
+      "step": 4365
+    },
+    {
+      "epoch": 1.0980885311871227,
+      "grad_norm": 0.35766953229904175,
+      "learning_rate": 7.994870634669978e-06,
+      "loss": 0.3295,
+      "step": 4366
+    },
+    {
+      "epoch": 1.0983400402414487,
+      "grad_norm": 0.33565643429756165,
+      "learning_rate": 7.993698784012387e-06,
+      "loss": 0.3499,
+      "step": 4367
+    },
+    {
+      "epoch": 1.0985915492957747,
+      "grad_norm": 0.34716835618019104,
+      "learning_rate": 7.992526676963222e-06,
+      "loss": 0.3523,
+      "step": 4368
+    },
+    {
+      "epoch": 1.0988430583501005,
+      "grad_norm": 0.33831220865249634,
+      "learning_rate": 7.991354313622868e-06,
+      "loss": 0.3546,
+      "step": 4369
+    },
+    {
+      "epoch": 1.0990945674044266,
+      "grad_norm": 0.36244910955429077,
+      "learning_rate": 7.990181694091733e-06,
+      "loss": 0.3547,
+      "step": 4370
+    },
+    {
+      "epoch": 1.0993460764587526,
+      "grad_norm": 0.33830562233924866,
+      "learning_rate": 7.98900881847024e-06,
+      "loss": 0.366,
+      "step": 4371
+    },
+    {
+      "epoch": 1.0995975855130784,
+      "grad_norm": 0.35777151584625244,
+      "learning_rate": 7.987835686858845e-06,
+      "loss": 0.336,
+      "step": 4372
+    },
+    {
+      "epoch": 1.0998490945674044,
+      "grad_norm": 0.36557304859161377,
+      "learning_rate": 7.986662299358012e-06,
+      "loss": 0.3743,
+      "step": 4373
+    },
+    {
+      "epoch": 1.1001006036217305,
+      "grad_norm": 0.33524492383003235,
+      "learning_rate": 7.985488656068238e-06,
+      "loss": 0.3942,
+      "step": 4374
+    },
+    {
+      "epoch": 1.1003521126760563,
+      "grad_norm": 0.370383620262146,
+      "learning_rate": 7.984314757090036e-06,
+      "loss": 0.3503,
+      "step": 4375
+    },
+    {
+      "epoch": 1.1006036217303823,
+      "grad_norm": 0.31732454895973206,
+      "learning_rate": 7.983140602523949e-06,
+      "loss": 0.3528,
+      "step": 4376
+    },
+    {
+      "epoch": 1.1008551307847083,
+      "grad_norm": 0.3776906728744507,
+      "learning_rate": 7.981966192470529e-06,
+      "loss": 0.3587,
+      "step": 4377
+    },
+    {
+      "epoch": 1.1011066398390341,
+      "grad_norm": 0.30028223991394043,
+      "learning_rate": 7.980791527030361e-06,
+      "loss": 0.3388,
+      "step": 4378
+    },
+    {
+      "epoch": 1.1013581488933601,
+      "grad_norm": 0.32881900668144226,
+      "learning_rate": 7.979616606304045e-06,
+      "loss": 0.3689,
+      "step": 4379
+    },
+    {
+      "epoch": 1.1016096579476862,
+      "grad_norm": 0.4046640694141388,
+      "learning_rate": 7.978441430392208e-06,
+      "loss": 0.374,
+      "step": 4380
+    },
+    {
+      "epoch": 1.101861167002012,
+      "grad_norm": 0.3668253719806671,
+      "learning_rate": 7.977265999395496e-06,
+      "loss": 0.3649,
+      "step": 4381
+    },
+    {
+      "epoch": 1.102112676056338,
+      "grad_norm": 0.34378865361213684,
+      "learning_rate": 7.976090313414575e-06,
+      "loss": 0.3674,
+      "step": 4382
+    },
+    {
+      "epoch": 1.102364185110664,
+      "grad_norm": 0.3307862877845764,
+      "learning_rate": 7.974914372550139e-06,
+      "loss": 0.3359,
+      "step": 4383
+    },
+    {
+      "epoch": 1.10261569416499,
+      "grad_norm": 0.3454669713973999,
+      "learning_rate": 7.973738176902897e-06,
+      "loss": 0.3663,
+      "step": 4384
+    },
+    {
+      "epoch": 1.1028672032193159,
+      "grad_norm": 0.3528616428375244,
+      "learning_rate": 7.972561726573584e-06,
+      "loss": 0.3359,
+      "step": 4385
+    },
+    {
+      "epoch": 1.1031187122736419,
+      "grad_norm": 0.35522276163101196,
+      "learning_rate": 7.971385021662956e-06,
+      "loss": 0.3597,
+      "step": 4386
+    },
+    {
+      "epoch": 1.103370221327968,
+      "grad_norm": 0.3870282769203186,
+      "learning_rate": 7.970208062271791e-06,
+      "loss": 0.3485,
+      "step": 4387
+    },
+    {
+      "epoch": 1.1036217303822937,
+      "grad_norm": 0.3756462037563324,
+      "learning_rate": 7.969030848500886e-06,
+      "loss": 0.3455,
+      "step": 4388
+    },
+    {
+      "epoch": 1.1038732394366197,
+      "grad_norm": 0.34227368235588074,
+      "learning_rate": 7.967853380451062e-06,
+      "loss": 0.3508,
+      "step": 4389
+    },
+    {
+      "epoch": 1.1041247484909458,
+      "grad_norm": 0.4221192002296448,
+      "learning_rate": 7.966675658223162e-06,
+      "loss": 0.3851,
+      "step": 4390
+    },
+    {
+      "epoch": 1.1043762575452716,
+      "grad_norm": 0.31191956996917725,
+      "learning_rate": 7.965497681918052e-06,
+      "loss": 0.3239,
+      "step": 4391
+    },
+    {
+      "epoch": 1.1046277665995976,
+      "grad_norm": 0.32340946793556213,
+      "learning_rate": 7.96431945163662e-06,
+      "loss": 0.3637,
+      "step": 4392
+    },
+    {
+      "epoch": 1.1048792756539236,
+      "grad_norm": 0.3376328945159912,
+      "learning_rate": 7.96314096747977e-06,
+      "loss": 0.3395,
+      "step": 4393
+    },
+    {
+      "epoch": 1.1051307847082494,
+      "grad_norm": 0.33500152826309204,
+      "learning_rate": 7.961962229548433e-06,
+      "loss": 0.3576,
+      "step": 4394
+    },
+    {
+      "epoch": 1.1053822937625755,
+      "grad_norm": 0.3390989601612091,
+      "learning_rate": 7.960783237943561e-06,
+      "loss": 0.3521,
+      "step": 4395
+    },
+    {
+      "epoch": 1.1056338028169015,
+      "grad_norm": 0.3361843228340149,
+      "learning_rate": 7.959603992766127e-06,
+      "loss": 0.3493,
+      "step": 4396
+    },
+    {
+      "epoch": 1.1058853118712273,
+      "grad_norm": 0.3171842098236084,
+      "learning_rate": 7.958424494117128e-06,
+      "loss": 0.3438,
+      "step": 4397
+    },
+    {
+      "epoch": 1.1061368209255533,
+      "grad_norm": 0.2926395535469055,
+      "learning_rate": 7.957244742097579e-06,
+      "loss": 0.343,
+      "step": 4398
+    },
+    {
+      "epoch": 1.1063883299798793,
+      "grad_norm": 0.34122782945632935,
+      "learning_rate": 7.956064736808516e-06,
+      "loss": 0.345,
+      "step": 4399
+    },
+    {
+      "epoch": 1.1066398390342052,
+      "grad_norm": 0.3118756115436554,
+      "learning_rate": 7.954884478351003e-06,
+      "loss": 0.338,
+      "step": 4400
+    },
+    {
+      "epoch": 1.1068913480885312,
+      "grad_norm": 0.3336293399333954,
+      "learning_rate": 7.953703966826118e-06,
+      "loss": 0.3618,
+      "step": 4401
+    },
+    {
+      "epoch": 1.1071428571428572,
+      "grad_norm": 0.3180750906467438,
+      "learning_rate": 7.95252320233497e-06,
+      "loss": 0.3661,
+      "step": 4402
+    },
+    {
+      "epoch": 1.107394366197183,
+      "grad_norm": 0.3299802541732788,
+      "learning_rate": 7.951342184978678e-06,
+      "loss": 0.3472,
+      "step": 4403
+    },
+    {
+      "epoch": 1.107645875251509,
+      "grad_norm": 0.34347137808799744,
+      "learning_rate": 7.950160914858392e-06,
+      "loss": 0.3272,
+      "step": 4404
+    },
+    {
+      "epoch": 1.107897384305835,
+      "grad_norm": 0.36151158809661865,
+      "learning_rate": 7.94897939207528e-06,
+      "loss": 0.3307,
+      "step": 4405
+    },
+    {
+      "epoch": 1.1081488933601609,
+      "grad_norm": 0.323889821767807,
+      "learning_rate": 7.947797616730532e-06,
+      "loss": 0.3571,
+      "step": 4406
+    },
+    {
+      "epoch": 1.108400402414487,
+      "grad_norm": 0.33470696210861206,
+      "learning_rate": 7.94661558892536e-06,
+      "loss": 0.354,
+      "step": 4407
+    },
+    {
+      "epoch": 1.108651911468813,
+      "grad_norm": 0.33871525526046753,
+      "learning_rate": 7.945433308760998e-06,
+      "loss": 0.3722,
+      "step": 4408
+    },
+    {
+      "epoch": 1.1089034205231387,
+      "grad_norm": 0.3187379837036133,
+      "learning_rate": 7.944250776338696e-06,
+      "loss": 0.3618,
+      "step": 4409
+    },
+    {
+      "epoch": 1.1091549295774648,
+      "grad_norm": 0.3083096444606781,
+      "learning_rate": 7.943067991759736e-06,
+      "loss": 0.3508,
+      "step": 4410
+    },
+    {
+      "epoch": 1.1094064386317908,
+      "grad_norm": 0.33438965678215027,
+      "learning_rate": 7.941884955125416e-06,
+      "loss": 0.3632,
+      "step": 4411
+    },
+    {
+      "epoch": 1.1096579476861166,
+      "grad_norm": 0.35571396350860596,
+      "learning_rate": 7.940701666537051e-06,
+      "loss": 0.3458,
+      "step": 4412
+    },
+    {
+      "epoch": 1.1099094567404426,
+      "grad_norm": 0.3360385298728943,
+      "learning_rate": 7.939518126095986e-06,
+      "loss": 0.3572,
+      "step": 4413
+    },
+    {
+      "epoch": 1.1101609657947686,
+      "grad_norm": 0.3844273090362549,
+      "learning_rate": 7.938334333903584e-06,
+      "loss": 0.3569,
+      "step": 4414
+    },
+    {
+      "epoch": 1.1104124748490947,
+      "grad_norm": 0.3142733573913574,
+      "learning_rate": 7.937150290061228e-06,
+      "loss": 0.3575,
+      "step": 4415
+    },
+    {
+      "epoch": 1.1106639839034205,
+      "grad_norm": 0.32374638319015503,
+      "learning_rate": 7.935965994670325e-06,
+      "loss": 0.3405,
+      "step": 4416
+    },
+    {
+      "epoch": 1.1109154929577465,
+      "grad_norm": 0.33620190620422363,
+      "learning_rate": 7.9347814478323e-06,
+      "loss": 0.3691,
+      "step": 4417
+    },
+    {
+      "epoch": 1.1111670020120725,
+      "grad_norm": 0.3400149345397949,
+      "learning_rate": 7.933596649648606e-06,
+      "loss": 0.3382,
+      "step": 4418
+    },
+    {
+      "epoch": 1.1114185110663983,
+      "grad_norm": 0.34986597299575806,
+      "learning_rate": 7.932411600220712e-06,
+      "loss": 0.3667,
+      "step": 4419
+    },
+    {
+      "epoch": 1.1116700201207244,
+      "grad_norm": 0.34316298365592957,
+      "learning_rate": 7.931226299650108e-06,
+      "loss": 0.3592,
+      "step": 4420
+    },
+    {
+      "epoch": 1.1119215291750504,
+      "grad_norm": 0.362543523311615,
+      "learning_rate": 7.930040748038309e-06,
+      "loss": 0.3544,
+      "step": 4421
+    },
+    {
+      "epoch": 1.1121730382293762,
+      "grad_norm": 0.35389113426208496,
+      "learning_rate": 7.92885494548685e-06,
+      "loss": 0.3427,
+      "step": 4422
+    },
+    {
+      "epoch": 1.1124245472837022,
+      "grad_norm": 0.32142454385757446,
+      "learning_rate": 7.927668892097288e-06,
+      "loss": 0.354,
+      "step": 4423
+    },
+    {
+      "epoch": 1.1126760563380282,
+      "grad_norm": 0.3471391499042511,
+      "learning_rate": 7.926482587971202e-06,
+      "loss": 0.3538,
+      "step": 4424
+    },
+    {
+      "epoch": 1.112927565392354,
+      "grad_norm": 0.33723774552345276,
+      "learning_rate": 7.925296033210191e-06,
+      "loss": 0.3634,
+      "step": 4425
+    },
+    {
+      "epoch": 1.11317907444668,
+      "grad_norm": 0.3441479206085205,
+      "learning_rate": 7.924109227915872e-06,
+      "loss": 0.3669,
+      "step": 4426
+    },
+    {
+      "epoch": 1.113430583501006,
+      "grad_norm": 0.3241642713546753,
+      "learning_rate": 7.922922172189892e-06,
+      "loss": 0.3629,
+      "step": 4427
+    },
+    {
+      "epoch": 1.113682092555332,
+      "grad_norm": 0.35414087772369385,
+      "learning_rate": 7.921734866133917e-06,
+      "loss": 0.3546,
+      "step": 4428
+    },
+    {
+      "epoch": 1.113933601609658,
+      "grad_norm": 0.3270585834980011,
+      "learning_rate": 7.920547309849626e-06,
+      "loss": 0.3449,
+      "step": 4429
+    },
+    {
+      "epoch": 1.114185110663984,
+      "grad_norm": 0.3540771007537842,
+      "learning_rate": 7.91935950343873e-06,
+      "loss": 0.3508,
+      "step": 4430
+    },
+    {
+      "epoch": 1.1144366197183098,
+      "grad_norm": 0.3404322862625122,
+      "learning_rate": 7.918171447002955e-06,
+      "loss": 0.3358,
+      "step": 4431
+    },
+    {
+      "epoch": 1.1146881287726358,
+      "grad_norm": 0.335871160030365,
+      "learning_rate": 7.916983140644052e-06,
+      "loss": 0.345,
+      "step": 4432
+    },
+    {
+      "epoch": 1.1149396378269618,
+      "grad_norm": 0.351862370967865,
+      "learning_rate": 7.915794584463792e-06,
+      "loss": 0.3417,
+      "step": 4433
+    },
+    {
+      "epoch": 1.1151911468812878,
+      "grad_norm": 0.3961928188800812,
+      "learning_rate": 7.914605778563965e-06,
+      "loss": 0.3346,
+      "step": 4434
+    },
+    {
+      "epoch": 1.1154426559356136,
+      "grad_norm": 0.33614158630371094,
+      "learning_rate": 7.913416723046387e-06,
+      "loss": 0.3679,
+      "step": 4435
+    },
+    {
+      "epoch": 1.1156941649899397,
+      "grad_norm": 0.3272840082645416,
+      "learning_rate": 7.912227418012895e-06,
+      "loss": 0.3413,
+      "step": 4436
+    },
+    {
+      "epoch": 1.1159456740442657,
+      "grad_norm": 0.3329838812351227,
+      "learning_rate": 7.911037863565344e-06,
+      "loss": 0.3587,
+      "step": 4437
+    },
+    {
+      "epoch": 1.1161971830985915,
+      "grad_norm": 0.41790270805358887,
+      "learning_rate": 7.90984805980561e-06,
+      "loss": 0.3805,
+      "step": 4438
+    },
+    {
+      "epoch": 1.1164486921529175,
+      "grad_norm": 0.3353562355041504,
+      "learning_rate": 7.908658006835593e-06,
+      "loss": 0.3301,
+      "step": 4439
+    },
+    {
+      "epoch": 1.1167002012072436,
+      "grad_norm": 0.3343071937561035,
+      "learning_rate": 7.907467704757214e-06,
+      "loss": 0.3614,
+      "step": 4440
+    },
+    {
+      "epoch": 1.1169517102615694,
+      "grad_norm": 0.37369048595428467,
+      "learning_rate": 7.906277153672417e-06,
+      "loss": 0.3467,
+      "step": 4441
+    },
+    {
+      "epoch": 1.1172032193158954,
+      "grad_norm": 0.3621070981025696,
+      "learning_rate": 7.905086353683162e-06,
+      "loss": 0.3337,
+      "step": 4442
+    },
+    {
+      "epoch": 1.1174547283702214,
+      "grad_norm": 0.33832234144210815,
+      "learning_rate": 7.903895304891436e-06,
+      "loss": 0.3742,
+      "step": 4443
+    },
+    {
+      "epoch": 1.1177062374245472,
+      "grad_norm": 0.3197494447231293,
+      "learning_rate": 7.902704007399243e-06,
+      "loss": 0.3868,
+      "step": 4444
+    },
+    {
+      "epoch": 1.1179577464788732,
+      "grad_norm": 0.3825930058956146,
+      "learning_rate": 7.901512461308612e-06,
+      "loss": 0.3603,
+      "step": 4445
+    },
+    {
+      "epoch": 1.1182092555331993,
+      "grad_norm": 0.3661872446537018,
+      "learning_rate": 7.90032066672159e-06,
+      "loss": 0.3633,
+      "step": 4446
+    },
+    {
+      "epoch": 1.118460764587525,
+      "grad_norm": 0.31290680170059204,
+      "learning_rate": 7.899128623740246e-06,
+      "loss": 0.3503,
+      "step": 4447
+    },
+    {
+      "epoch": 1.118712273641851,
+      "grad_norm": 0.35227593779563904,
+      "learning_rate": 7.897936332466674e-06,
+      "loss": 0.3781,
+      "step": 4448
+    },
+    {
+      "epoch": 1.1189637826961771,
+      "grad_norm": 0.3571607768535614,
+      "learning_rate": 7.896743793002983e-06,
+      "loss": 0.355,
+      "step": 4449
+    },
+    {
+      "epoch": 1.119215291750503,
+      "grad_norm": 0.35419657826423645,
+      "learning_rate": 7.89555100545131e-06,
+      "loss": 0.3337,
+      "step": 4450
+    },
+    {
+      "epoch": 1.119466800804829,
+      "grad_norm": 0.31918078660964966,
+      "learning_rate": 7.894357969913807e-06,
+      "loss": 0.3707,
+      "step": 4451
+    },
+    {
+      "epoch": 1.119718309859155,
+      "grad_norm": 0.33282384276390076,
+      "learning_rate": 7.893164686492652e-06,
+      "loss": 0.3861,
+      "step": 4452
+    },
+    {
+      "epoch": 1.1199698189134808,
+      "grad_norm": 0.3873891234397888,
+      "learning_rate": 7.891971155290039e-06,
+      "loss": 0.3788,
+      "step": 4453
+    },
+    {
+      "epoch": 1.1202213279678068,
+      "grad_norm": 0.3210555613040924,
+      "learning_rate": 7.89077737640819e-06,
+      "loss": 0.3426,
+      "step": 4454
+    },
+    {
+      "epoch": 1.1204728370221329,
+      "grad_norm": 0.33169567584991455,
+      "learning_rate": 7.889583349949341e-06,
+      "loss": 0.3279,
+      "step": 4455
+    },
+    {
+      "epoch": 1.1207243460764587,
+      "grad_norm": 0.36757755279541016,
+      "learning_rate": 7.88838907601576e-06,
+      "loss": 0.3631,
+      "step": 4456
+    },
+    {
+      "epoch": 1.1209758551307847,
+      "grad_norm": 0.32751592993736267,
+      "learning_rate": 7.88719455470972e-06,
+      "loss": 0.3331,
+      "step": 4457
+    },
+    {
+      "epoch": 1.1212273641851107,
+      "grad_norm": 0.3210444152355194,
+      "learning_rate": 7.88599978613353e-06,
+      "loss": 0.3527,
+      "step": 4458
+    },
+    {
+      "epoch": 1.1214788732394365,
+      "grad_norm": 0.37147364020347595,
+      "learning_rate": 7.884804770389514e-06,
+      "loss": 0.3601,
+      "step": 4459
+    },
+    {
+      "epoch": 1.1217303822937625,
+      "grad_norm": 0.33347150683403015,
+      "learning_rate": 7.883609507580016e-06,
+      "loss": 0.336,
+      "step": 4460
+    },
+    {
+      "epoch": 1.1219818913480886,
+      "grad_norm": 0.38581255078315735,
+      "learning_rate": 7.882413997807404e-06,
+      "loss": 0.3488,
+      "step": 4461
+    },
+    {
+      "epoch": 1.1222334004024144,
+      "grad_norm": 0.3418939411640167,
+      "learning_rate": 7.881218241174064e-06,
+      "loss": 0.3565,
+      "step": 4462
+    },
+    {
+      "epoch": 1.1224849094567404,
+      "grad_norm": 0.3888770043849945,
+      "learning_rate": 7.880022237782407e-06,
+      "loss": 0.3497,
+      "step": 4463
+    },
+    {
+      "epoch": 1.1227364185110664,
+      "grad_norm": 0.3226841688156128,
+      "learning_rate": 7.878825987734864e-06,
+      "loss": 0.3558,
+      "step": 4464
+    },
+    {
+      "epoch": 1.1229879275653925,
+      "grad_norm": 0.3461357355117798,
+      "learning_rate": 7.877629491133884e-06,
+      "loss": 0.3294,
+      "step": 4465
+    },
+    {
+      "epoch": 1.1232394366197183,
+      "grad_norm": 0.3348862826824188,
+      "learning_rate": 7.876432748081939e-06,
+      "loss": 0.3626,
+      "step": 4466
+    },
+    {
+      "epoch": 1.1234909456740443,
+      "grad_norm": 0.3597047030925751,
+      "learning_rate": 7.875235758681527e-06,
+      "loss": 0.3706,
+      "step": 4467
+    },
+    {
+      "epoch": 1.1237424547283703,
+      "grad_norm": 0.3321612477302551,
+      "learning_rate": 7.874038523035157e-06,
+      "loss": 0.3608,
+      "step": 4468
+    },
+    {
+      "epoch": 1.1239939637826961,
+      "grad_norm": 0.3340033292770386,
+      "learning_rate": 7.872841041245369e-06,
+      "loss": 0.3429,
+      "step": 4469
+    },
+    {
+      "epoch": 1.1242454728370221,
+      "grad_norm": 0.34898126125335693,
+      "learning_rate": 7.871643313414718e-06,
+      "loss": 0.3439,
+      "step": 4470
+    },
+    {
+      "epoch": 1.1244969818913482,
+      "grad_norm": 0.33899804949760437,
+      "learning_rate": 7.870445339645783e-06,
+      "loss": 0.3356,
+      "step": 4471
+    },
+    {
+      "epoch": 1.124748490945674,
+      "grad_norm": 0.30220794677734375,
+      "learning_rate": 7.869247120041161e-06,
+      "loss": 0.3427,
+      "step": 4472
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.3090187907218933,
+      "learning_rate": 7.868048654703474e-06,
+      "loss": 0.344,
+      "step": 4473
+    },
+    {
+      "epoch": 1.125251509054326,
+      "grad_norm": 0.33787021040916443,
+      "learning_rate": 7.866849943735361e-06,
+      "loss": 0.3362,
+      "step": 4474
+    },
+    {
+      "epoch": 1.1255030181086518,
+      "grad_norm": 0.3474128842353821,
+      "learning_rate": 7.865650987239485e-06,
+      "loss": 0.3474,
+      "step": 4475
+    },
+    {
+      "epoch": 1.1257545271629779,
+      "grad_norm": 0.34391891956329346,
+      "learning_rate": 7.864451785318532e-06,
+      "loss": 0.3917,
+      "step": 4476
+    },
+    {
+      "epoch": 1.1260060362173039,
+      "grad_norm": 0.34758976101875305,
+      "learning_rate": 7.863252338075202e-06,
+      "loss": 0.3533,
+      "step": 4477
+    },
+    {
+      "epoch": 1.1262575452716297,
+      "grad_norm": 0.32106417417526245,
+      "learning_rate": 7.862052645612222e-06,
+      "loss": 0.3453,
+      "step": 4478
+    },
+    {
+      "epoch": 1.1265090543259557,
+      "grad_norm": 0.337090402841568,
+      "learning_rate": 7.860852708032337e-06,
+      "loss": 0.3541,
+      "step": 4479
+    },
+    {
+      "epoch": 1.1267605633802817,
+      "grad_norm": 0.3655794560909271,
+      "learning_rate": 7.859652525438314e-06,
+      "loss": 0.3466,
+      "step": 4480
+    },
+    {
+      "epoch": 1.1270120724346078,
+      "grad_norm": 0.32245180010795593,
+      "learning_rate": 7.858452097932945e-06,
+      "loss": 0.3464,
+      "step": 4481
+    },
+    {
+      "epoch": 1.1272635814889336,
+      "grad_norm": 0.32112833857536316,
+      "learning_rate": 7.857251425619034e-06,
+      "loss": 0.3705,
+      "step": 4482
+    },
+    {
+      "epoch": 1.1275150905432596,
+      "grad_norm": 0.3438814580440521,
+      "learning_rate": 7.856050508599413e-06,
+      "loss": 0.337,
+      "step": 4483
+    },
+    {
+      "epoch": 1.1277665995975856,
+      "grad_norm": 0.33978280425071716,
+      "learning_rate": 7.854849346976935e-06,
+      "loss": 0.3429,
+      "step": 4484
+    },
+    {
+      "epoch": 1.1280181086519114,
+      "grad_norm": 0.3462032079696655,
+      "learning_rate": 7.85364794085447e-06,
+      "loss": 0.3581,
+      "step": 4485
+    },
+    {
+      "epoch": 1.1282696177062375,
+      "grad_norm": 0.3374899923801422,
+      "learning_rate": 7.85244629033491e-06,
+      "loss": 0.3614,
+      "step": 4486
+    },
+    {
+      "epoch": 1.1285211267605635,
+      "grad_norm": 0.3162481486797333,
+      "learning_rate": 7.851244395521171e-06,
+      "loss": 0.3457,
+      "step": 4487
+    },
+    {
+      "epoch": 1.1287726358148893,
+      "grad_norm": 0.3583298921585083,
+      "learning_rate": 7.850042256516187e-06,
+      "loss": 0.3418,
+      "step": 4488
+    },
+    {
+      "epoch": 1.1290241448692153,
+      "grad_norm": 0.33638957142829895,
+      "learning_rate": 7.848839873422913e-06,
+      "loss": 0.3523,
+      "step": 4489
+    },
+    {
+      "epoch": 1.1292756539235413,
+      "grad_norm": 0.3432900905609131,
+      "learning_rate": 7.847637246344326e-06,
+      "loss": 0.3641,
+      "step": 4490
+    },
+    {
+      "epoch": 1.1295271629778671,
+      "grad_norm": 0.3538980185985565,
+      "learning_rate": 7.846434375383425e-06,
+      "loss": 0.347,
+      "step": 4491
+    },
+    {
+      "epoch": 1.1297786720321932,
+      "grad_norm": 0.3335748016834259,
+      "learning_rate": 7.845231260643226e-06,
+      "loss": 0.3426,
+      "step": 4492
+    },
+    {
+      "epoch": 1.1300301810865192,
+      "grad_norm": 0.3384160101413727,
+      "learning_rate": 7.84402790222677e-06,
+      "loss": 0.3483,
+      "step": 4493
+    },
+    {
+      "epoch": 1.130281690140845,
+      "grad_norm": 0.3617180585861206,
+      "learning_rate": 7.842824300237114e-06,
+      "loss": 0.3726,
+      "step": 4494
+    },
+    {
+      "epoch": 1.130533199195171,
+      "grad_norm": 0.3085126280784607,
+      "learning_rate": 7.841620454777344e-06,
+      "loss": 0.3614,
+      "step": 4495
+    },
+    {
+      "epoch": 1.130784708249497,
+      "grad_norm": 0.3673594892024994,
+      "learning_rate": 7.840416365950558e-06,
+      "loss": 0.3277,
+      "step": 4496
+    },
+    {
+      "epoch": 1.1310362173038229,
+      "grad_norm": 0.3403297960758209,
+      "learning_rate": 7.839212033859882e-06,
+      "loss": 0.36,
+      "step": 4497
+    },
+    {
+      "epoch": 1.131287726358149,
+      "grad_norm": 0.3315725028514862,
+      "learning_rate": 7.838007458608455e-06,
+      "loss": 0.3523,
+      "step": 4498
+    },
+    {
+      "epoch": 1.131539235412475,
+      "grad_norm": 0.3540287911891937,
+      "learning_rate": 7.836802640299442e-06,
+      "loss": 0.3584,
+      "step": 4499
+    },
+    {
+      "epoch": 1.1317907444668007,
+      "grad_norm": 0.32690343260765076,
+      "learning_rate": 7.835597579036031e-06,
+      "loss": 0.3393,
+      "step": 4500
+    },
+    {
+      "epoch": 1.1320422535211268,
+      "grad_norm": 0.37675777077674866,
+      "learning_rate": 7.83439227492143e-06,
+      "loss": 0.3682,
+      "step": 4501
+    },
+    {
+      "epoch": 1.1322937625754528,
+      "grad_norm": 0.35298097133636475,
+      "learning_rate": 7.833186728058859e-06,
+      "loss": 0.3618,
+      "step": 4502
+    },
+    {
+      "epoch": 1.1325452716297786,
+      "grad_norm": 0.33117878437042236,
+      "learning_rate": 7.831980938551572e-06,
+      "loss": 0.3414,
+      "step": 4503
+    },
+    {
+      "epoch": 1.1327967806841046,
+      "grad_norm": 0.32076895236968994,
+      "learning_rate": 7.83077490650283e-06,
+      "loss": 0.3558,
+      "step": 4504
+    },
+    {
+      "epoch": 1.1330482897384306,
+      "grad_norm": 0.3626091182231903,
+      "learning_rate": 7.829568632015932e-06,
+      "loss": 0.339,
+      "step": 4505
+    },
+    {
+      "epoch": 1.1332997987927564,
+      "grad_norm": 0.35690760612487793,
+      "learning_rate": 7.828362115194179e-06,
+      "loss": 0.349,
+      "step": 4506
+    },
+    {
+      "epoch": 1.1335513078470825,
+      "grad_norm": 0.32652193307876587,
+      "learning_rate": 7.827155356140905e-06,
+      "loss": 0.3566,
+      "step": 4507
+    },
+    {
+      "epoch": 1.1338028169014085,
+      "grad_norm": 0.3355286419391632,
+      "learning_rate": 7.825948354959464e-06,
+      "loss": 0.3516,
+      "step": 4508
+    },
+    {
+      "epoch": 1.1340543259557343,
+      "grad_norm": 0.35634520649909973,
+      "learning_rate": 7.824741111753223e-06,
+      "loss": 0.3398,
+      "step": 4509
+    },
+    {
+      "epoch": 1.1343058350100603,
+      "grad_norm": 0.35237765312194824,
+      "learning_rate": 7.823533626625577e-06,
+      "loss": 0.3559,
+      "step": 4510
+    },
+    {
+      "epoch": 1.1345573440643864,
+      "grad_norm": 0.3162923753261566,
+      "learning_rate": 7.822325899679941e-06,
+      "loss": 0.3601,
+      "step": 4511
+    },
+    {
+      "epoch": 1.1348088531187122,
+      "grad_norm": 0.33308616280555725,
+      "learning_rate": 7.821117931019749e-06,
+      "loss": 0.346,
+      "step": 4512
+    },
+    {
+      "epoch": 1.1350603621730382,
+      "grad_norm": 0.3171202540397644,
+      "learning_rate": 7.819909720748454e-06,
+      "loss": 0.3534,
+      "step": 4513
+    },
+    {
+      "epoch": 1.1353118712273642,
+      "grad_norm": 0.35061752796173096,
+      "learning_rate": 7.818701268969532e-06,
+      "loss": 0.3766,
+      "step": 4514
+    },
+    {
+      "epoch": 1.1355633802816902,
+      "grad_norm": 0.3866909444332123,
+      "learning_rate": 7.817492575786481e-06,
+      "loss": 0.3422,
+      "step": 4515
+    },
+    {
+      "epoch": 1.135814889336016,
+      "grad_norm": 0.3715068995952606,
+      "learning_rate": 7.816283641302815e-06,
+      "loss": 0.3729,
+      "step": 4516
+    },
+    {
+      "epoch": 1.136066398390342,
+      "grad_norm": 0.3531903922557831,
+      "learning_rate": 7.815074465622076e-06,
+      "loss": 0.3698,
+      "step": 4517
+    },
+    {
+      "epoch": 1.136317907444668,
+      "grad_norm": 0.4138471186161041,
+      "learning_rate": 7.81386504884782e-06,
+      "loss": 0.341,
+      "step": 4518
+    },
+    {
+      "epoch": 1.136569416498994,
+      "grad_norm": 0.350268691778183,
+      "learning_rate": 7.812655391083625e-06,
+      "loss": 0.3162,
+      "step": 4519
+    },
+    {
+      "epoch": 1.13682092555332,
+      "grad_norm": 0.36769789457321167,
+      "learning_rate": 7.811445492433091e-06,
+      "loss": 0.3529,
+      "step": 4520
+    },
+    {
+      "epoch": 1.137072434607646,
+      "grad_norm": 0.3521837890148163,
+      "learning_rate": 7.810235352999842e-06,
+      "loss": 0.3505,
+      "step": 4521
+    },
+    {
+      "epoch": 1.1373239436619718,
+      "grad_norm": 0.3225024342536926,
+      "learning_rate": 7.809024972887513e-06,
+      "loss": 0.3648,
+      "step": 4522
+    },
+    {
+      "epoch": 1.1375754527162978,
+      "grad_norm": 0.34223929047584534,
+      "learning_rate": 7.807814352199769e-06,
+      "loss": 0.3557,
+      "step": 4523
+    },
+    {
+      "epoch": 1.1378269617706238,
+      "grad_norm": 0.33616122603416443,
+      "learning_rate": 7.806603491040293e-06,
+      "loss": 0.348,
+      "step": 4524
+    },
+    {
+      "epoch": 1.1380784708249496,
+      "grad_norm": 0.36442843079566956,
+      "learning_rate": 7.805392389512785e-06,
+      "loss": 0.3607,
+      "step": 4525
+    },
+    {
+      "epoch": 1.1383299798792756,
+      "grad_norm": 0.3255125880241394,
+      "learning_rate": 7.80418104772097e-06,
+      "loss": 0.3719,
+      "step": 4526
+    },
+    {
+      "epoch": 1.1385814889336017,
+      "grad_norm": 0.31949490308761597,
+      "learning_rate": 7.802969465768588e-06,
+      "loss": 0.3558,
+      "step": 4527
+    },
+    {
+      "epoch": 1.1388329979879275,
+      "grad_norm": 0.32983100414276123,
+      "learning_rate": 7.801757643759408e-06,
+      "loss": 0.363,
+      "step": 4528
+    },
+    {
+      "epoch": 1.1390845070422535,
+      "grad_norm": 0.3546229898929596,
+      "learning_rate": 7.800545581797217e-06,
+      "loss": 0.3636,
+      "step": 4529
+    },
+    {
+      "epoch": 1.1393360160965795,
+      "grad_norm": 0.40820834040641785,
+      "learning_rate": 7.799333279985813e-06,
+      "loss": 0.3654,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1395875251509056,
+      "grad_norm": 0.35177260637283325,
+      "learning_rate": 7.798120738429028e-06,
+      "loss": 0.357,
+      "step": 4531
+    },
+    {
+      "epoch": 1.1398390342052314,
+      "grad_norm": 0.3217819929122925,
+      "learning_rate": 7.796907957230706e-06,
+      "loss": 0.3425,
+      "step": 4532
+    },
+    {
+      "epoch": 1.1400905432595574,
+      "grad_norm": 0.4682590961456299,
+      "learning_rate": 7.795694936494715e-06,
+      "loss": 0.3726,
+      "step": 4533
+    },
+    {
+      "epoch": 1.1403420523138834,
+      "grad_norm": 0.4092445373535156,
+      "learning_rate": 7.794481676324944e-06,
+      "loss": 0.3628,
+      "step": 4534
+    },
+    {
+      "epoch": 1.1405935613682092,
+      "grad_norm": 0.31070423126220703,
+      "learning_rate": 7.793268176825297e-06,
+      "loss": 0.3421,
+      "step": 4535
+    },
+    {
+      "epoch": 1.1408450704225352,
+      "grad_norm": 0.34520095586776733,
+      "learning_rate": 7.792054438099706e-06,
+      "loss": 0.36,
+      "step": 4536
+    },
+    {
+      "epoch": 1.1410965794768613,
+      "grad_norm": 0.3309139609336853,
+      "learning_rate": 7.790840460252121e-06,
+      "loss": 0.3451,
+      "step": 4537
+    },
+    {
+      "epoch": 1.141348088531187,
+      "grad_norm": 0.36794281005859375,
+      "learning_rate": 7.789626243386508e-06,
+      "loss": 0.3752,
+      "step": 4538
+    },
+    {
+      "epoch": 1.141599597585513,
+      "grad_norm": 0.3274729251861572,
+      "learning_rate": 7.78841178760686e-06,
+      "loss": 0.3556,
+      "step": 4539
+    },
+    {
+      "epoch": 1.1418511066398391,
+      "grad_norm": 0.32525065541267395,
+      "learning_rate": 7.787197093017186e-06,
+      "loss": 0.3666,
+      "step": 4540
+    },
+    {
+      "epoch": 1.142102615694165,
+      "grad_norm": 0.3399212062358856,
+      "learning_rate": 7.78598215972152e-06,
+      "loss": 0.3457,
+      "step": 4541
+    },
+    {
+      "epoch": 1.142354124748491,
+      "grad_norm": 0.3338754177093506,
+      "learning_rate": 7.784766987823908e-06,
+      "loss": 0.3819,
+      "step": 4542
+    },
+    {
+      "epoch": 1.142605633802817,
+      "grad_norm": 0.3345630168914795,
+      "learning_rate": 7.783551577428427e-06,
+      "loss": 0.3374,
+      "step": 4543
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.3454863727092743,
+      "learning_rate": 7.782335928639167e-06,
+      "loss": 0.3487,
+      "step": 4544
+    },
+    {
+      "epoch": 1.1431086519114688,
+      "grad_norm": 0.34839943051338196,
+      "learning_rate": 7.781120041560241e-06,
+      "loss": 0.3635,
+      "step": 4545
+    },
+    {
+      "epoch": 1.1433601609657948,
+      "grad_norm": 0.3082118034362793,
+      "learning_rate": 7.779903916295781e-06,
+      "loss": 0.3468,
+      "step": 4546
+    },
+    {
+      "epoch": 1.1436116700201207,
+      "grad_norm": 0.32561156153678894,
+      "learning_rate": 7.778687552949944e-06,
+      "loss": 0.3237,
+      "step": 4547
+    },
+    {
+      "epoch": 1.1438631790744467,
+      "grad_norm": 0.3549325466156006,
+      "learning_rate": 7.777470951626899e-06,
+      "loss": 0.3679,
+      "step": 4548
+    },
+    {
+      "epoch": 1.1441146881287727,
+      "grad_norm": 0.33792486786842346,
+      "learning_rate": 7.776254112430843e-06,
+      "loss": 0.3508,
+      "step": 4549
+    },
+    {
+      "epoch": 1.1443661971830985,
+      "grad_norm": 0.3282465934753418,
+      "learning_rate": 7.775037035465992e-06,
+      "loss": 0.3501,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1446177062374245,
+      "grad_norm": 0.3474048972129822,
+      "learning_rate": 7.773819720836577e-06,
+      "loss": 0.3573,
+      "step": 4551
+    },
+    {
+      "epoch": 1.1448692152917506,
+      "grad_norm": 0.36000293493270874,
+      "learning_rate": 7.77260216864686e-06,
+      "loss": 0.3291,
+      "step": 4552
+    },
+    {
+      "epoch": 1.1451207243460764,
+      "grad_norm": 0.3380999267101288,
+      "learning_rate": 7.771384379001107e-06,
+      "loss": 0.3461,
+      "step": 4553
+    },
+    {
+      "epoch": 1.1453722334004024,
+      "grad_norm": 0.3201408386230469,
+      "learning_rate": 7.770166352003622e-06,
+      "loss": 0.3633,
+      "step": 4554
+    },
+    {
+      "epoch": 1.1456237424547284,
+      "grad_norm": 0.3245110809803009,
+      "learning_rate": 7.768948087758721e-06,
+      "loss": 0.3486,
+      "step": 4555
+    },
+    {
+      "epoch": 1.1458752515090542,
+      "grad_norm": 0.33637115359306335,
+      "learning_rate": 7.767729586370738e-06,
+      "loss": 0.3535,
+      "step": 4556
+    },
+    {
+      "epoch": 1.1461267605633803,
+      "grad_norm": 0.33941903710365295,
+      "learning_rate": 7.76651084794403e-06,
+      "loss": 0.3553,
+      "step": 4557
+    },
+    {
+      "epoch": 1.1463782696177063,
+      "grad_norm": 0.3234095275402069,
+      "learning_rate": 7.765291872582975e-06,
+      "loss": 0.3674,
+      "step": 4558
+    },
+    {
+      "epoch": 1.146629778672032,
+      "grad_norm": 0.34868893027305603,
+      "learning_rate": 7.76407266039197e-06,
+      "loss": 0.349,
+      "step": 4559
+    },
+    {
+      "epoch": 1.1468812877263581,
+      "grad_norm": 0.3286401927471161,
+      "learning_rate": 7.762853211475436e-06,
+      "loss": 0.3362,
+      "step": 4560
+    },
+    {
+      "epoch": 1.1471327967806841,
+      "grad_norm": 0.3449994623661041,
+      "learning_rate": 7.761633525937808e-06,
+      "loss": 0.3408,
+      "step": 4561
+    },
+    {
+      "epoch": 1.14738430583501,
+      "grad_norm": 0.42614054679870605,
+      "learning_rate": 7.760413603883546e-06,
+      "loss": 0.3397,
+      "step": 4562
+    },
+    {
+      "epoch": 1.147635814889336,
+      "grad_norm": 0.34790942072868347,
+      "learning_rate": 7.759193445417126e-06,
+      "loss": 0.3705,
+      "step": 4563
+    },
+    {
+      "epoch": 1.147887323943662,
+      "grad_norm": 0.33942970633506775,
+      "learning_rate": 7.757973050643048e-06,
+      "loss": 0.3671,
+      "step": 4564
+    },
+    {
+      "epoch": 1.148138832997988,
+      "grad_norm": 0.34391435980796814,
+      "learning_rate": 7.756752419665833e-06,
+      "loss": 0.3341,
+      "step": 4565
+    },
+    {
+      "epoch": 1.1483903420523138,
+      "grad_norm": 0.3598215579986572,
+      "learning_rate": 7.755531552590021e-06,
+      "loss": 0.347,
+      "step": 4566
+    },
+    {
+      "epoch": 1.1486418511066399,
+      "grad_norm": 0.31856873631477356,
+      "learning_rate": 7.754310449520169e-06,
+      "loss": 0.3475,
+      "step": 4567
+    },
+    {
+      "epoch": 1.1488933601609659,
+      "grad_norm": 0.3364936411380768,
+      "learning_rate": 7.753089110560858e-06,
+      "loss": 0.3521,
+      "step": 4568
+    },
+    {
+      "epoch": 1.1491448692152917,
+      "grad_norm": 0.3098117709159851,
+      "learning_rate": 7.751867535816689e-06,
+      "loss": 0.3435,
+      "step": 4569
+    },
+    {
+      "epoch": 1.1493963782696177,
+      "grad_norm": 0.33782443404197693,
+      "learning_rate": 7.750645725392278e-06,
+      "loss": 0.3309,
+      "step": 4570
+    },
+    {
+      "epoch": 1.1496478873239437,
+      "grad_norm": 0.3181942403316498,
+      "learning_rate": 7.749423679392271e-06,
+      "loss": 0.3532,
+      "step": 4571
+    },
+    {
+      "epoch": 1.1498993963782695,
+      "grad_norm": 0.32375866174697876,
+      "learning_rate": 7.748201397921326e-06,
+      "loss": 0.3546,
+      "step": 4572
+    },
+    {
+      "epoch": 1.1501509054325956,
+      "grad_norm": 0.32787060737609863,
+      "learning_rate": 7.746978881084124e-06,
+      "loss": 0.3732,
+      "step": 4573
+    },
+    {
+      "epoch": 1.1504024144869216,
+      "grad_norm": 0.34005919098854065,
+      "learning_rate": 7.745756128985367e-06,
+      "loss": 0.3587,
+      "step": 4574
+    },
+    {
+      "epoch": 1.1506539235412474,
+      "grad_norm": 0.3388522267341614,
+      "learning_rate": 7.744533141729773e-06,
+      "loss": 0.3393,
+      "step": 4575
+    },
+    {
+      "epoch": 1.1509054325955734,
+      "grad_norm": 0.35405799746513367,
+      "learning_rate": 7.743309919422086e-06,
+      "loss": 0.3485,
+      "step": 4576
+    },
+    {
+      "epoch": 1.1511569416498995,
+      "grad_norm": 0.3198212683200836,
+      "learning_rate": 7.742086462167066e-06,
+      "loss": 0.3477,
+      "step": 4577
+    },
+    {
+      "epoch": 1.1514084507042253,
+      "grad_norm": 0.3119359612464905,
+      "learning_rate": 7.740862770069494e-06,
+      "loss": 0.3464,
+      "step": 4578
+    },
+    {
+      "epoch": 1.1516599597585513,
+      "grad_norm": 0.32057052850723267,
+      "learning_rate": 7.739638843234176e-06,
+      "loss": 0.3406,
+      "step": 4579
+    },
+    {
+      "epoch": 1.1519114688128773,
+      "grad_norm": 0.37748971581459045,
+      "learning_rate": 7.738414681765928e-06,
+      "loss": 0.3845,
+      "step": 4580
+    },
+    {
+      "epoch": 1.1521629778672033,
+      "grad_norm": 0.35111579298973083,
+      "learning_rate": 7.737190285769594e-06,
+      "loss": 0.3628,
+      "step": 4581
+    },
+    {
+      "epoch": 1.1524144869215291,
+      "grad_norm": 0.31299889087677,
+      "learning_rate": 7.735965655350034e-06,
+      "loss": 0.3171,
+      "step": 4582
+    },
+    {
+      "epoch": 1.1526659959758552,
+      "grad_norm": 0.3520580232143402,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.3537,
+      "step": 4583
+    },
+    {
+      "epoch": 1.1529175050301812,
+      "grad_norm": 0.3616025447845459,
+      "learning_rate": 7.733515691660795e-06,
+      "loss": 0.3483,
+      "step": 4584
+    },
+    {
+      "epoch": 1.153169014084507,
+      "grad_norm": 0.3729570508003235,
+      "learning_rate": 7.732290358600936e-06,
+      "loss": 0.3418,
+      "step": 4585
+    },
+    {
+      "epoch": 1.153420523138833,
+      "grad_norm": 0.3173588812351227,
+      "learning_rate": 7.731064791537501e-06,
+      "loss": 0.3572,
+      "step": 4586
+    },
+    {
+      "epoch": 1.153672032193159,
+      "grad_norm": 0.3116159439086914,
+      "learning_rate": 7.72983899057545e-06,
+      "loss": 0.3276,
+      "step": 4587
+    },
+    {
+      "epoch": 1.1539235412474849,
+      "grad_norm": 0.33578798174858093,
+      "learning_rate": 7.728612955819773e-06,
+      "loss": 0.3469,
+      "step": 4588
+    },
+    {
+      "epoch": 1.154175050301811,
+      "grad_norm": 0.35310569405555725,
+      "learning_rate": 7.727386687375461e-06,
+      "loss": 0.3701,
+      "step": 4589
+    },
+    {
+      "epoch": 1.154426559356137,
+      "grad_norm": 0.37188154458999634,
+      "learning_rate": 7.726160185347544e-06,
+      "loss": 0.4008,
+      "step": 4590
+    },
+    {
+      "epoch": 1.1546780684104627,
+      "grad_norm": 0.3462255597114563,
+      "learning_rate": 7.724933449841061e-06,
+      "loss": 0.34,
+      "step": 4591
+    },
+    {
+      "epoch": 1.1549295774647887,
+      "grad_norm": 0.31674739718437195,
+      "learning_rate": 7.723706480961078e-06,
+      "loss": 0.3469,
+      "step": 4592
+    },
+    {
+      "epoch": 1.1551810865191148,
+      "grad_norm": 0.3254792392253876,
+      "learning_rate": 7.722479278812672e-06,
+      "loss": 0.349,
+      "step": 4593
+    },
+    {
+      "epoch": 1.1554325955734406,
+      "grad_norm": 0.36106935143470764,
+      "learning_rate": 7.721251843500948e-06,
+      "loss": 0.347,
+      "step": 4594
+    },
+    {
+      "epoch": 1.1556841046277666,
+      "grad_norm": 0.3273780345916748,
+      "learning_rate": 7.720024175131027e-06,
+      "loss": 0.3775,
+      "step": 4595
+    },
+    {
+      "epoch": 1.1559356136820926,
+      "grad_norm": 0.31497684121131897,
+      "learning_rate": 7.71879627380805e-06,
+      "loss": 0.3642,
+      "step": 4596
+    },
+    {
+      "epoch": 1.1561871227364184,
+      "grad_norm": 0.360487163066864,
+      "learning_rate": 7.717568139637184e-06,
+      "loss": 0.3624,
+      "step": 4597
+    },
+    {
+      "epoch": 1.1564386317907445,
+      "grad_norm": 0.3374006748199463,
+      "learning_rate": 7.716339772723608e-06,
+      "loss": 0.3559,
+      "step": 4598
+    },
+    {
+      "epoch": 1.1566901408450705,
+      "grad_norm": 0.3451824486255646,
+      "learning_rate": 7.715111173172522e-06,
+      "loss": 0.3666,
+      "step": 4599
+    },
+    {
+      "epoch": 1.1569416498993963,
+      "grad_norm": 0.3282189667224884,
+      "learning_rate": 7.713882341089151e-06,
+      "loss": 0.3532,
+      "step": 4600
+    },
+    {
+      "epoch": 1.1571931589537223,
+      "grad_norm": 0.34830883145332336,
+      "learning_rate": 7.712653276578734e-06,
+      "loss": 0.3937,
+      "step": 4601
+    },
+    {
+      "epoch": 1.1574446680080483,
+      "grad_norm": 0.32040953636169434,
+      "learning_rate": 7.711423979746537e-06,
+      "loss": 0.3401,
+      "step": 4602
+    },
+    {
+      "epoch": 1.1576961770623742,
+      "grad_norm": 0.3415544927120209,
+      "learning_rate": 7.710194450697837e-06,
+      "loss": 0.3673,
+      "step": 4603
+    },
+    {
+      "epoch": 1.1579476861167002,
+      "grad_norm": 0.35662606358528137,
+      "learning_rate": 7.708964689537937e-06,
+      "loss": 0.3489,
+      "step": 4604
+    },
+    {
+      "epoch": 1.1581991951710262,
+      "grad_norm": 0.34960421919822693,
+      "learning_rate": 7.707734696372158e-06,
+      "loss": 0.3528,
+      "step": 4605
+    },
+    {
+      "epoch": 1.158450704225352,
+      "grad_norm": 0.30523577332496643,
+      "learning_rate": 7.706504471305843e-06,
+      "loss": 0.344,
+      "step": 4606
+    },
+    {
+      "epoch": 1.158702213279678,
+      "grad_norm": 0.33770856261253357,
+      "learning_rate": 7.70527401444435e-06,
+      "loss": 0.3737,
+      "step": 4607
+    },
+    {
+      "epoch": 1.158953722334004,
+      "grad_norm": 0.34141412377357483,
+      "learning_rate": 7.704043325893064e-06,
+      "loss": 0.3384,
+      "step": 4608
+    },
+    {
+      "epoch": 1.1592052313883299,
+      "grad_norm": 0.35698843002319336,
+      "learning_rate": 7.702812405757382e-06,
+      "loss": 0.3231,
+      "step": 4609
+    },
+    {
+      "epoch": 1.159456740442656,
+      "grad_norm": 0.37319764494895935,
+      "learning_rate": 7.701581254142728e-06,
+      "loss": 0.3508,
+      "step": 4610
+    },
+    {
+      "epoch": 1.159708249496982,
+      "grad_norm": 0.32721927762031555,
+      "learning_rate": 7.70034987115454e-06,
+      "loss": 0.3482,
+      "step": 4611
+    },
+    {
+      "epoch": 1.1599597585513077,
+      "grad_norm": 0.3475745618343353,
+      "learning_rate": 7.699118256898277e-06,
+      "loss": 0.3454,
+      "step": 4612
+    },
+    {
+      "epoch": 1.1602112676056338,
+      "grad_norm": 0.3759898245334625,
+      "learning_rate": 7.697886411479422e-06,
+      "loss": 0.3683,
+      "step": 4613
+    },
+    {
+      "epoch": 1.1604627766599598,
+      "grad_norm": 0.35182997584342957,
+      "learning_rate": 7.696654335003475e-06,
+      "loss": 0.3519,
+      "step": 4614
+    },
+    {
+      "epoch": 1.1607142857142858,
+      "grad_norm": 0.38022348284721375,
+      "learning_rate": 7.695422027575953e-06,
+      "loss": 0.3434,
+      "step": 4615
+    },
+    {
+      "epoch": 1.1609657947686116,
+      "grad_norm": 0.36011841893196106,
+      "learning_rate": 7.694189489302399e-06,
+      "loss": 0.3508,
+      "step": 4616
+    },
+    {
+      "epoch": 1.1612173038229376,
+      "grad_norm": 0.3434779942035675,
+      "learning_rate": 7.692956720288369e-06,
+      "loss": 0.3481,
+      "step": 4617
+    },
+    {
+      "epoch": 1.1614688128772637,
+      "grad_norm": 0.3736356198787689,
+      "learning_rate": 7.69172372063944e-06,
+      "loss": 0.3376,
+      "step": 4618
+    },
+    {
+      "epoch": 1.1617203219315895,
+      "grad_norm": 0.3600045144557953,
+      "learning_rate": 7.690490490461217e-06,
+      "loss": 0.3435,
+      "step": 4619
+    },
+    {
+      "epoch": 1.1619718309859155,
+      "grad_norm": 0.3338755965232849,
+      "learning_rate": 7.689257029859316e-06,
+      "loss": 0.3457,
+      "step": 4620
+    },
+    {
+      "epoch": 1.1622233400402415,
+      "grad_norm": 0.33322474360466003,
+      "learning_rate": 7.688023338939373e-06,
+      "loss": 0.3442,
+      "step": 4621
+    },
+    {
+      "epoch": 1.1624748490945673,
+      "grad_norm": 0.3295151889324188,
+      "learning_rate": 7.686789417807045e-06,
+      "loss": 0.329,
+      "step": 4622
+    },
+    {
+      "epoch": 1.1627263581488934,
+      "grad_norm": 0.36218923330307007,
+      "learning_rate": 7.685555266568014e-06,
+      "loss": 0.3676,
+      "step": 4623
+    },
+    {
+      "epoch": 1.1629778672032194,
+      "grad_norm": 0.3499116003513336,
+      "learning_rate": 7.684320885327976e-06,
+      "loss": 0.374,
+      "step": 4624
+    },
+    {
+      "epoch": 1.1632293762575452,
+      "grad_norm": 0.3172034025192261,
+      "learning_rate": 7.683086274192647e-06,
+      "loss": 0.3304,
+      "step": 4625
+    },
+    {
+      "epoch": 1.1634808853118712,
+      "grad_norm": 0.3143727481365204,
+      "learning_rate": 7.681851433267762e-06,
+      "loss": 0.35,
+      "step": 4626
+    },
+    {
+      "epoch": 1.1637323943661972,
+      "grad_norm": 0.3694257140159607,
+      "learning_rate": 7.680616362659082e-06,
+      "loss": 0.3583,
+      "step": 4627
+    },
+    {
+      "epoch": 1.163983903420523,
+      "grad_norm": 0.3408942222595215,
+      "learning_rate": 7.679381062472377e-06,
+      "loss": 0.3374,
+      "step": 4628
+    },
+    {
+      "epoch": 1.164235412474849,
+      "grad_norm": 0.29644355177879333,
+      "learning_rate": 7.678145532813448e-06,
+      "loss": 0.3226,
+      "step": 4629
+    },
+    {
+      "epoch": 1.164486921529175,
+      "grad_norm": 0.3752206265926361,
+      "learning_rate": 7.67690977378811e-06,
+      "loss": 0.3861,
+      "step": 4630
+    },
+    {
+      "epoch": 1.1647384305835011,
+      "grad_norm": 0.3474452495574951,
+      "learning_rate": 7.675673785502195e-06,
+      "loss": 0.3807,
+      "step": 4631
+    },
+    {
+      "epoch": 1.164989939637827,
+      "grad_norm": 0.3323638141155243,
+      "learning_rate": 7.674437568061559e-06,
+      "loss": 0.347,
+      "step": 4632
+    },
+    {
+      "epoch": 1.165241448692153,
+      "grad_norm": 0.322618305683136,
+      "learning_rate": 7.673201121572077e-06,
+      "loss": 0.3689,
+      "step": 4633
+    },
+    {
+      "epoch": 1.165492957746479,
+      "grad_norm": 0.3595317304134369,
+      "learning_rate": 7.671964446139643e-06,
+      "loss": 0.3701,
+      "step": 4634
+    },
+    {
+      "epoch": 1.1657444668008048,
+      "grad_norm": 0.33650046586990356,
+      "learning_rate": 7.67072754187017e-06,
+      "loss": 0.3491,
+      "step": 4635
+    },
+    {
+      "epoch": 1.1659959758551308,
+      "grad_norm": 0.3536091148853302,
+      "learning_rate": 7.66949040886959e-06,
+      "loss": 0.3488,
+      "step": 4636
+    },
+    {
+      "epoch": 1.1662474849094568,
+      "grad_norm": 0.3690861761569977,
+      "learning_rate": 7.668253047243856e-06,
+      "loss": 0.3511,
+      "step": 4637
+    },
+    {
+      "epoch": 1.1664989939637826,
+      "grad_norm": 0.33993828296661377,
+      "learning_rate": 7.667015457098944e-06,
+      "loss": 0.358,
+      "step": 4638
+    },
+    {
+      "epoch": 1.1667505030181087,
+      "grad_norm": 0.3575940728187561,
+      "learning_rate": 7.66577763854084e-06,
+      "loss": 0.3414,
+      "step": 4639
+    },
+    {
+      "epoch": 1.1670020120724347,
+      "grad_norm": 0.3681694567203522,
+      "learning_rate": 7.664539591675559e-06,
+      "loss": 0.3922,
+      "step": 4640
+    },
+    {
+      "epoch": 1.1672535211267605,
+      "grad_norm": 0.3579506576061249,
+      "learning_rate": 7.663301316609131e-06,
+      "loss": 0.341,
+      "step": 4641
+    },
+    {
+      "epoch": 1.1675050301810865,
+      "grad_norm": 0.3645816147327423,
+      "learning_rate": 7.662062813447608e-06,
+      "loss": 0.3502,
+      "step": 4642
+    },
+    {
+      "epoch": 1.1677565392354126,
+      "grad_norm": 0.34026220440864563,
+      "learning_rate": 7.660824082297057e-06,
+      "loss": 0.3656,
+      "step": 4643
+    },
+    {
+      "epoch": 1.1680080482897384,
+      "grad_norm": 0.3202768862247467,
+      "learning_rate": 7.659585123263571e-06,
+      "loss": 0.354,
+      "step": 4644
+    },
+    {
+      "epoch": 1.1682595573440644,
+      "grad_norm": 0.3285091519355774,
+      "learning_rate": 7.658345936453257e-06,
+      "loss": 0.3649,
+      "step": 4645
+    },
+    {
+      "epoch": 1.1685110663983904,
+      "grad_norm": 0.3168383538722992,
+      "learning_rate": 7.657106521972246e-06,
+      "loss": 0.3475,
+      "step": 4646
+    },
+    {
+      "epoch": 1.1687625754527162,
+      "grad_norm": 0.3457097113132477,
+      "learning_rate": 7.655866879926682e-06,
+      "loss": 0.3591,
+      "step": 4647
+    },
+    {
+      "epoch": 1.1690140845070423,
+      "grad_norm": 0.31571948528289795,
+      "learning_rate": 7.654627010422735e-06,
+      "loss": 0.3494,
+      "step": 4648
+    },
+    {
+      "epoch": 1.1692655935613683,
+      "grad_norm": 0.3494684398174286,
+      "learning_rate": 7.653386913566593e-06,
+      "loss": 0.3648,
+      "step": 4649
+    },
+    {
+      "epoch": 1.169517102615694,
+      "grad_norm": 0.3523896038532257,
+      "learning_rate": 7.65214658946446e-06,
+      "loss": 0.3614,
+      "step": 4650
+    },
+    {
+      "epoch": 1.16976861167002,
+      "grad_norm": 0.32975268363952637,
+      "learning_rate": 7.650906038222563e-06,
+      "loss": 0.3505,
+      "step": 4651
+    },
+    {
+      "epoch": 1.1700201207243461,
+      "grad_norm": 0.35798007249832153,
+      "learning_rate": 7.64966525994715e-06,
+      "loss": 0.355,
+      "step": 4652
+    },
+    {
+      "epoch": 1.170271629778672,
+      "grad_norm": 0.3435303568840027,
+      "learning_rate": 7.648424254744481e-06,
+      "loss": 0.3548,
+      "step": 4653
+    },
+    {
+      "epoch": 1.170523138832998,
+      "grad_norm": 0.3186478316783905,
+      "learning_rate": 7.647183022720846e-06,
+      "loss": 0.3469,
+      "step": 4654
+    },
+    {
+      "epoch": 1.170774647887324,
+      "grad_norm": 0.3599293529987335,
+      "learning_rate": 7.645941563982544e-06,
+      "loss": 0.3726,
+      "step": 4655
+    },
+    {
+      "epoch": 1.1710261569416498,
+      "grad_norm": 0.3607332706451416,
+      "learning_rate": 7.644699878635901e-06,
+      "loss": 0.339,
+      "step": 4656
+    },
+    {
+      "epoch": 1.1712776659959758,
+      "grad_norm": 0.3603340983390808,
+      "learning_rate": 7.643457966787258e-06,
+      "loss": 0.369,
+      "step": 4657
+    },
+    {
+      "epoch": 1.1715291750503019,
+      "grad_norm": 0.3286212980747223,
+      "learning_rate": 7.642215828542977e-06,
+      "loss": 0.3645,
+      "step": 4658
+    },
+    {
+      "epoch": 1.1717806841046277,
+      "grad_norm": 0.35398638248443604,
+      "learning_rate": 7.64097346400944e-06,
+      "loss": 0.3453,
+      "step": 4659
+    },
+    {
+      "epoch": 1.1720321931589537,
+      "grad_norm": 0.3555023968219757,
+      "learning_rate": 7.63973087329305e-06,
+      "loss": 0.3515,
+      "step": 4660
+    },
+    {
+      "epoch": 1.1722837022132797,
+      "grad_norm": 0.3314943015575409,
+      "learning_rate": 7.638488056500222e-06,
+      "loss": 0.3579,
+      "step": 4661
+    },
+    {
+      "epoch": 1.1725352112676055,
+      "grad_norm": 0.3424238860607147,
+      "learning_rate": 7.637245013737399e-06,
+      "loss": 0.3462,
+      "step": 4662
+    },
+    {
+      "epoch": 1.1727867203219315,
+      "grad_norm": 0.3617474138736725,
+      "learning_rate": 7.636001745111039e-06,
+      "loss": 0.3541,
+      "step": 4663
+    },
+    {
+      "epoch": 1.1730382293762576,
+      "grad_norm": 0.3585506081581116,
+      "learning_rate": 7.634758250727621e-06,
+      "loss": 0.3603,
+      "step": 4664
+    },
+    {
+      "epoch": 1.1732897384305836,
+      "grad_norm": 0.35511475801467896,
+      "learning_rate": 7.633514530693642e-06,
+      "loss": 0.3553,
+      "step": 4665
+    },
+    {
+      "epoch": 1.1735412474849094,
+      "grad_norm": 0.3293471038341522,
+      "learning_rate": 7.632270585115618e-06,
+      "loss": 0.3478,
+      "step": 4666
+    },
+    {
+      "epoch": 1.1737927565392354,
+      "grad_norm": 0.3938407599925995,
+      "learning_rate": 7.631026414100086e-06,
+      "loss": 0.3484,
+      "step": 4667
+    },
+    {
+      "epoch": 1.1740442655935615,
+      "grad_norm": 0.3182221055030823,
+      "learning_rate": 7.629782017753602e-06,
+      "loss": 0.3623,
+      "step": 4668
+    },
+    {
+      "epoch": 1.1742957746478873,
+      "grad_norm": 0.34111714363098145,
+      "learning_rate": 7.628537396182739e-06,
+      "loss": 0.3747,
+      "step": 4669
+    },
+    {
+      "epoch": 1.1745472837022133,
+      "grad_norm": 0.3676598072052002,
+      "learning_rate": 7.627292549494092e-06,
+      "loss": 0.3687,
+      "step": 4670
+    },
+    {
+      "epoch": 1.1747987927565393,
+      "grad_norm": 0.3540399670600891,
+      "learning_rate": 7.626047477794276e-06,
+      "loss": 0.3611,
+      "step": 4671
+    },
+    {
+      "epoch": 1.1750503018108651,
+      "grad_norm": 0.32811835408210754,
+      "learning_rate": 7.62480218118992e-06,
+      "loss": 0.3235,
+      "step": 4672
+    },
+    {
+      "epoch": 1.1753018108651911,
+      "grad_norm": 0.3579806089401245,
+      "learning_rate": 7.6235566597876786e-06,
+      "loss": 0.3429,
+      "step": 4673
+    },
+    {
+      "epoch": 1.1755533199195172,
+      "grad_norm": 0.3340808153152466,
+      "learning_rate": 7.622310913694222e-06,
+      "loss": 0.3554,
+      "step": 4674
+    },
+    {
+      "epoch": 1.175804828973843,
+      "grad_norm": 0.3324950933456421,
+      "learning_rate": 7.621064943016241e-06,
+      "loss": 0.3406,
+      "step": 4675
+    },
+    {
+      "epoch": 1.176056338028169,
+      "grad_norm": 0.3520958423614502,
+      "learning_rate": 7.6198187478604455e-06,
+      "loss": 0.3199,
+      "step": 4676
+    },
+    {
+      "epoch": 1.176307847082495,
+      "grad_norm": 0.3330570459365845,
+      "learning_rate": 7.618572328333565e-06,
+      "loss": 0.3447,
+      "step": 4677
+    },
+    {
+      "epoch": 1.1765593561368208,
+      "grad_norm": 0.3154620826244354,
+      "learning_rate": 7.617325684542344e-06,
+      "loss": 0.3549,
+      "step": 4678
+    },
+    {
+      "epoch": 1.1768108651911469,
+      "grad_norm": 0.29848769307136536,
+      "learning_rate": 7.6160788165935525e-06,
+      "loss": 0.3355,
+      "step": 4679
+    },
+    {
+      "epoch": 1.1770623742454729,
+      "grad_norm": 0.32562530040740967,
+      "learning_rate": 7.6148317245939766e-06,
+      "loss": 0.3314,
+      "step": 4680
+    },
+    {
+      "epoch": 1.177313883299799,
+      "grad_norm": 0.33761054277420044,
+      "learning_rate": 7.613584408650423e-06,
+      "loss": 0.3452,
+      "step": 4681
+    },
+    {
+      "epoch": 1.1775653923541247,
+      "grad_norm": 0.3247910737991333,
+      "learning_rate": 7.612336868869714e-06,
+      "loss": 0.3577,
+      "step": 4682
+    },
+    {
+      "epoch": 1.1778169014084507,
+      "grad_norm": 0.3061235845088959,
+      "learning_rate": 7.611089105358695e-06,
+      "loss": 0.3222,
+      "step": 4683
+    },
+    {
+      "epoch": 1.1780684104627768,
+      "grad_norm": 0.32039159536361694,
+      "learning_rate": 7.609841118224229e-06,
+      "loss": 0.3555,
+      "step": 4684
+    },
+    {
+      "epoch": 1.1783199195171026,
+      "grad_norm": 0.3344159722328186,
+      "learning_rate": 7.608592907573199e-06,
+      "loss": 0.3256,
+      "step": 4685
+    },
+    {
+      "epoch": 1.1785714285714286,
+      "grad_norm": 0.32991334795951843,
+      "learning_rate": 7.607344473512506e-06,
+      "loss": 0.3265,
+      "step": 4686
+    },
+    {
+      "epoch": 1.1788229376257546,
+      "grad_norm": 0.35629698634147644,
+      "learning_rate": 7.606095816149069e-06,
+      "loss": 0.3511,
+      "step": 4687
+    },
+    {
+      "epoch": 1.1790744466800804,
+      "grad_norm": 0.3231953978538513,
+      "learning_rate": 7.60484693558983e-06,
+      "loss": 0.3497,
+      "step": 4688
+    },
+    {
+      "epoch": 1.1793259557344065,
+      "grad_norm": 0.35513460636138916,
+      "learning_rate": 7.603597831941747e-06,
+      "loss": 0.3495,
+      "step": 4689
+    },
+    {
+      "epoch": 1.1795774647887325,
+      "grad_norm": 0.34873831272125244,
+      "learning_rate": 7.602348505311797e-06,
+      "loss": 0.3525,
+      "step": 4690
+    },
+    {
+      "epoch": 1.1798289738430583,
+      "grad_norm": 0.300342857837677,
+      "learning_rate": 7.601098955806978e-06,
+      "loss": 0.3704,
+      "step": 4691
+    },
+    {
+      "epoch": 1.1800804828973843,
+      "grad_norm": 0.3526897430419922,
+      "learning_rate": 7.5998491835343065e-06,
+      "loss": 0.3513,
+      "step": 4692
+    },
+    {
+      "epoch": 1.1803319919517103,
+      "grad_norm": 0.3581010103225708,
+      "learning_rate": 7.598599188600817e-06,
+      "loss": 0.3653,
+      "step": 4693
+    },
+    {
+      "epoch": 1.1805835010060362,
+      "grad_norm": 0.31560835242271423,
+      "learning_rate": 7.5973489711135625e-06,
+      "loss": 0.3527,
+      "step": 4694
+    },
+    {
+      "epoch": 1.1808350100603622,
+      "grad_norm": 0.3127973973751068,
+      "learning_rate": 7.596098531179619e-06,
+      "loss": 0.378,
+      "step": 4695
+    },
+    {
+      "epoch": 1.1810865191146882,
+      "grad_norm": 0.35242587327957153,
+      "learning_rate": 7.594847868906076e-06,
+      "loss": 0.3592,
+      "step": 4696
+    },
+    {
+      "epoch": 1.181338028169014,
+      "grad_norm": 0.33138447999954224,
+      "learning_rate": 7.593596984400048e-06,
+      "loss": 0.391,
+      "step": 4697
+    },
+    {
+      "epoch": 1.18158953722334,
+      "grad_norm": 0.3515087068080902,
+      "learning_rate": 7.592345877768663e-06,
+      "loss": 0.3641,
+      "step": 4698
+    },
+    {
+      "epoch": 1.181841046277666,
+      "grad_norm": 0.32551079988479614,
+      "learning_rate": 7.591094549119071e-06,
+      "loss": 0.3725,
+      "step": 4699
+    },
+    {
+      "epoch": 1.1820925553319919,
+      "grad_norm": 0.32543718814849854,
+      "learning_rate": 7.589842998558441e-06,
+      "loss": 0.3395,
+      "step": 4700
+    },
+    {
+      "epoch": 1.182344064386318,
+      "grad_norm": 0.33554258942604065,
+      "learning_rate": 7.58859122619396e-06,
+      "loss": 0.3503,
+      "step": 4701
+    },
+    {
+      "epoch": 1.182595573440644,
+      "grad_norm": 0.33632540702819824,
+      "learning_rate": 7.587339232132835e-06,
+      "loss": 0.3622,
+      "step": 4702
+    },
+    {
+      "epoch": 1.1828470824949697,
+      "grad_norm": 0.32022663950920105,
+      "learning_rate": 7.586087016482291e-06,
+      "loss": 0.3443,
+      "step": 4703
+    },
+    {
+      "epoch": 1.1830985915492958,
+      "grad_norm": 0.3406105637550354,
+      "learning_rate": 7.584834579349572e-06,
+      "loss": 0.3413,
+      "step": 4704
+    },
+    {
+      "epoch": 1.1833501006036218,
+      "grad_norm": 0.3901978135108948,
+      "learning_rate": 7.5835819208419425e-06,
+      "loss": 0.3665,
+      "step": 4705
+    },
+    {
+      "epoch": 1.1836016096579476,
+      "grad_norm": 0.3854331076145172,
+      "learning_rate": 7.5823290410666835e-06,
+      "loss": 0.3607,
+      "step": 4706
+    },
+    {
+      "epoch": 1.1838531187122736,
+      "grad_norm": 0.3867035210132599,
+      "learning_rate": 7.5810759401310975e-06,
+      "loss": 0.3537,
+      "step": 4707
+    },
+    {
+      "epoch": 1.1841046277665996,
+      "grad_norm": 0.34121641516685486,
+      "learning_rate": 7.579822618142505e-06,
+      "loss": 0.3496,
+      "step": 4708
+    },
+    {
+      "epoch": 1.1843561368209254,
+      "grad_norm": 0.3433249592781067,
+      "learning_rate": 7.578569075208244e-06,
+      "loss": 0.3553,
+      "step": 4709
+    },
+    {
+      "epoch": 1.1846076458752515,
+      "grad_norm": 0.34000903367996216,
+      "learning_rate": 7.577315311435674e-06,
+      "loss": 0.3435,
+      "step": 4710
+    },
+    {
+      "epoch": 1.1848591549295775,
+      "grad_norm": 0.3518748879432678,
+      "learning_rate": 7.5760613269321715e-06,
+      "loss": 0.382,
+      "step": 4711
+    },
+    {
+      "epoch": 1.1851106639839033,
+      "grad_norm": 0.30435121059417725,
+      "learning_rate": 7.574807121805131e-06,
+      "loss": 0.3234,
+      "step": 4712
+    },
+    {
+      "epoch": 1.1853621730382293,
+      "grad_norm": 0.33615031838417053,
+      "learning_rate": 7.573552696161969e-06,
+      "loss": 0.3528,
+      "step": 4713
+    },
+    {
+      "epoch": 1.1856136820925554,
+      "grad_norm": 0.3324757516384125,
+      "learning_rate": 7.572298050110118e-06,
+      "loss": 0.3725,
+      "step": 4714
+    },
+    {
+      "epoch": 1.1858651911468814,
+      "grad_norm": 0.37572944164276123,
+      "learning_rate": 7.571043183757032e-06,
+      "loss": 0.3328,
+      "step": 4715
+    },
+    {
+      "epoch": 1.1861167002012072,
+      "grad_norm": 0.31655603647232056,
+      "learning_rate": 7.56978809721018e-06,
+      "loss": 0.3458,
+      "step": 4716
+    },
+    {
+      "epoch": 1.1863682092555332,
+      "grad_norm": 0.33465415239334106,
+      "learning_rate": 7.568532790577057e-06,
+      "loss": 0.369,
+      "step": 4717
+    },
+    {
+      "epoch": 1.1866197183098592,
+      "grad_norm": 0.33323749899864197,
+      "learning_rate": 7.567277263965167e-06,
+      "loss": 0.3469,
+      "step": 4718
+    },
+    {
+      "epoch": 1.186871227364185,
+      "grad_norm": 0.31222718954086304,
+      "learning_rate": 7.566021517482041e-06,
+      "loss": 0.37,
+      "step": 4719
+    },
+    {
+      "epoch": 1.187122736418511,
+      "grad_norm": 0.3619459271430969,
+      "learning_rate": 7.5647655512352245e-06,
+      "loss": 0.3535,
+      "step": 4720
+    },
+    {
+      "epoch": 1.187374245472837,
+      "grad_norm": 0.3543090224266052,
+      "learning_rate": 7.563509365332285e-06,
+      "loss": 0.3578,
+      "step": 4721
+    },
+    {
+      "epoch": 1.187625754527163,
+      "grad_norm": 0.3527710437774658,
+      "learning_rate": 7.562252959880804e-06,
+      "loss": 0.356,
+      "step": 4722
+    },
+    {
+      "epoch": 1.187877263581489,
+      "grad_norm": 0.3505546748638153,
+      "learning_rate": 7.560996334988386e-06,
+      "loss": 0.3641,
+      "step": 4723
+    },
+    {
+      "epoch": 1.188128772635815,
+      "grad_norm": 0.3206879794597626,
+      "learning_rate": 7.5597394907626555e-06,
+      "loss": 0.361,
+      "step": 4724
+    },
+    {
+      "epoch": 1.1883802816901408,
+      "grad_norm": 0.3364570140838623,
+      "learning_rate": 7.55848242731125e-06,
+      "loss": 0.339,
+      "step": 4725
+    },
+    {
+      "epoch": 1.1886317907444668,
+      "grad_norm": 0.3259868919849396,
+      "learning_rate": 7.557225144741831e-06,
+      "loss": 0.3472,
+      "step": 4726
+    },
+    {
+      "epoch": 1.1888832997987928,
+      "grad_norm": 0.34462079405784607,
+      "learning_rate": 7.5559676431620745e-06,
+      "loss": 0.3461,
+      "step": 4727
+    },
+    {
+      "epoch": 1.1891348088531186,
+      "grad_norm": 0.34534603357315063,
+      "learning_rate": 7.554709922679681e-06,
+      "loss": 0.38,
+      "step": 4728
+    },
+    {
+      "epoch": 1.1893863179074446,
+      "grad_norm": 0.32316854596138,
+      "learning_rate": 7.553451983402364e-06,
+      "loss": 0.3291,
+      "step": 4729
+    },
+    {
+      "epoch": 1.1896378269617707,
+      "grad_norm": 0.34536492824554443,
+      "learning_rate": 7.552193825437861e-06,
+      "loss": 0.3607,
+      "step": 4730
+    },
+    {
+      "epoch": 1.1898893360160967,
+      "grad_norm": 0.34370651841163635,
+      "learning_rate": 7.550935448893921e-06,
+      "loss": 0.3472,
+      "step": 4731
+    },
+    {
+      "epoch": 1.1901408450704225,
+      "grad_norm": 0.3375087380409241,
+      "learning_rate": 7.54967685387832e-06,
+      "loss": 0.364,
+      "step": 4732
+    },
+    {
+      "epoch": 1.1903923541247485,
+      "grad_norm": 0.3438502848148346,
+      "learning_rate": 7.548418040498847e-06,
+      "loss": 0.3446,
+      "step": 4733
+    },
+    {
+      "epoch": 1.1906438631790746,
+      "grad_norm": 0.34919726848602295,
+      "learning_rate": 7.547159008863312e-06,
+      "loss": 0.3661,
+      "step": 4734
+    },
+    {
+      "epoch": 1.1908953722334004,
+      "grad_norm": 0.3322198987007141,
+      "learning_rate": 7.545899759079542e-06,
+      "loss": 0.3845,
+      "step": 4735
+    },
+    {
+      "epoch": 1.1911468812877264,
+      "grad_norm": 0.3598858416080475,
+      "learning_rate": 7.544640291255385e-06,
+      "loss": 0.3669,
+      "step": 4736
+    },
+    {
+      "epoch": 1.1913983903420524,
+      "grad_norm": 0.3647436797618866,
+      "learning_rate": 7.543380605498707e-06,
+      "loss": 0.3699,
+      "step": 4737
+    },
+    {
+      "epoch": 1.1916498993963782,
+      "grad_norm": 0.3301692605018616,
+      "learning_rate": 7.542120701917391e-06,
+      "loss": 0.3728,
+      "step": 4738
+    },
+    {
+      "epoch": 1.1919014084507042,
+      "grad_norm": 0.33648648858070374,
+      "learning_rate": 7.540860580619339e-06,
+      "loss": 0.3443,
+      "step": 4739
+    },
+    {
+      "epoch": 1.1921529175050303,
+      "grad_norm": 0.3432157337665558,
+      "learning_rate": 7.539600241712475e-06,
+      "loss": 0.3518,
+      "step": 4740
+    },
+    {
+      "epoch": 1.192404426559356,
+      "grad_norm": 0.32822272181510925,
+      "learning_rate": 7.538339685304737e-06,
+      "loss": 0.3609,
+      "step": 4741
+    },
+    {
+      "epoch": 1.192655935613682,
+      "grad_norm": 0.334031343460083,
+      "learning_rate": 7.537078911504087e-06,
+      "loss": 0.3529,
+      "step": 4742
+    },
+    {
+      "epoch": 1.1929074446680081,
+      "grad_norm": 0.3261743187904358,
+      "learning_rate": 7.5358179204184975e-06,
+      "loss": 0.3695,
+      "step": 4743
+    },
+    {
+      "epoch": 1.193158953722334,
+      "grad_norm": 0.3211628496646881,
+      "learning_rate": 7.53455671215597e-06,
+      "loss": 0.3529,
+      "step": 4744
+    },
+    {
+      "epoch": 1.19341046277666,
+      "grad_norm": 0.30509626865386963,
+      "learning_rate": 7.533295286824513e-06,
+      "loss": 0.3663,
+      "step": 4745
+    },
+    {
+      "epoch": 1.193661971830986,
+      "grad_norm": 0.3319106101989746,
+      "learning_rate": 7.532033644532166e-06,
+      "loss": 0.3598,
+      "step": 4746
+    },
+    {
+      "epoch": 1.1939134808853118,
+      "grad_norm": 0.34237736463546753,
+      "learning_rate": 7.530771785386976e-06,
+      "loss": 0.3661,
+      "step": 4747
+    },
+    {
+      "epoch": 1.1941649899396378,
+      "grad_norm": 0.3307172656059265,
+      "learning_rate": 7.5295097094970136e-06,
+      "loss": 0.3489,
+      "step": 4748
+    },
+    {
+      "epoch": 1.1944164989939638,
+      "grad_norm": 0.32638081908226013,
+      "learning_rate": 7.528247416970371e-06,
+      "loss": 0.3613,
+      "step": 4749
+    },
+    {
+      "epoch": 1.1946680080482897,
+      "grad_norm": 0.3481774628162384,
+      "learning_rate": 7.5269849079151535e-06,
+      "loss": 0.3557,
+      "step": 4750
+    },
+    {
+      "epoch": 1.1949195171026157,
+      "grad_norm": 0.3151308298110962,
+      "learning_rate": 7.525722182439488e-06,
+      "loss": 0.3375,
+      "step": 4751
+    },
+    {
+      "epoch": 1.1951710261569417,
+      "grad_norm": 0.32890570163726807,
+      "learning_rate": 7.524459240651518e-06,
+      "loss": 0.3387,
+      "step": 4752
+    },
+    {
+      "epoch": 1.1954225352112675,
+      "grad_norm": 0.3260030150413513,
+      "learning_rate": 7.523196082659408e-06,
+      "loss": 0.3428,
+      "step": 4753
+    },
+    {
+      "epoch": 1.1956740442655935,
+      "grad_norm": 0.3384077847003937,
+      "learning_rate": 7.521932708571338e-06,
+      "loss": 0.3585,
+      "step": 4754
+    },
+    {
+      "epoch": 1.1959255533199196,
+      "grad_norm": 0.3400633633136749,
+      "learning_rate": 7.520669118495507e-06,
+      "loss": 0.3382,
+      "step": 4755
+    },
+    {
+      "epoch": 1.1961770623742454,
+      "grad_norm": 0.3479855954647064,
+      "learning_rate": 7.519405312540137e-06,
+      "loss": 0.3566,
+      "step": 4756
+    },
+    {
+      "epoch": 1.1964285714285714,
+      "grad_norm": 0.3062722980976105,
+      "learning_rate": 7.518141290813463e-06,
+      "loss": 0.3361,
+      "step": 4757
+    },
+    {
+      "epoch": 1.1966800804828974,
+      "grad_norm": 0.33496060967445374,
+      "learning_rate": 7.5168770534237414e-06,
+      "loss": 0.3296,
+      "step": 4758
+    },
+    {
+      "epoch": 1.1969315895372232,
+      "grad_norm": 0.3257729709148407,
+      "learning_rate": 7.515612600479243e-06,
+      "loss": 0.3502,
+      "step": 4759
+    },
+    {
+      "epoch": 1.1971830985915493,
+      "grad_norm": 0.3281537890434265,
+      "learning_rate": 7.514347932088266e-06,
+      "loss": 0.3289,
+      "step": 4760
+    },
+    {
+      "epoch": 1.1974346076458753,
+      "grad_norm": 0.324022501707077,
+      "learning_rate": 7.513083048359117e-06,
+      "loss": 0.327,
+      "step": 4761
+    },
+    {
+      "epoch": 1.197686116700201,
+      "grad_norm": 0.3308146595954895,
+      "learning_rate": 7.511817949400126e-06,
+      "loss": 0.3437,
+      "step": 4762
+    },
+    {
+      "epoch": 1.1979376257545271,
+      "grad_norm": 0.36193880438804626,
+      "learning_rate": 7.510552635319643e-06,
+      "loss": 0.3438,
+      "step": 4763
+    },
+    {
+      "epoch": 1.1981891348088531,
+      "grad_norm": 0.3138233721256256,
+      "learning_rate": 7.509287106226033e-06,
+      "loss": 0.3297,
+      "step": 4764
+    },
+    {
+      "epoch": 1.1984406438631792,
+      "grad_norm": 0.3267454504966736,
+      "learning_rate": 7.50802136222768e-06,
+      "loss": 0.3373,
+      "step": 4765
+    },
+    {
+      "epoch": 1.198692152917505,
+      "grad_norm": 0.34192612767219543,
+      "learning_rate": 7.506755403432987e-06,
+      "loss": 0.366,
+      "step": 4766
+    },
+    {
+      "epoch": 1.198943661971831,
+      "grad_norm": 0.3394359350204468,
+      "learning_rate": 7.505489229950375e-06,
+      "loss": 0.3285,
+      "step": 4767
+    },
+    {
+      "epoch": 1.199195171026157,
+      "grad_norm": 0.3370779752731323,
+      "learning_rate": 7.504222841888287e-06,
+      "loss": 0.3262,
+      "step": 4768
+    },
+    {
+      "epoch": 1.1994466800804828,
+      "grad_norm": 0.33641576766967773,
+      "learning_rate": 7.502956239355178e-06,
+      "loss": 0.3301,
+      "step": 4769
+    },
+    {
+      "epoch": 1.1996981891348089,
+      "grad_norm": 0.33692795038223267,
+      "learning_rate": 7.5016894224595235e-06,
+      "loss": 0.3633,
+      "step": 4770
+    },
+    {
+      "epoch": 1.1999496981891349,
+      "grad_norm": 0.3135643005371094,
+      "learning_rate": 7.5004223913098224e-06,
+      "loss": 0.3528,
+      "step": 4771
+    },
+    {
+      "epoch": 1.2002012072434607,
+      "grad_norm": 0.37040475010871887,
+      "learning_rate": 7.4991551460145874e-06,
+      "loss": 0.3748,
+      "step": 4772
+    },
+    {
+      "epoch": 1.2004527162977867,
+      "grad_norm": 0.35265108942985535,
+      "learning_rate": 7.497887686682347e-06,
+      "loss": 0.3373,
+      "step": 4773
+    },
+    {
+      "epoch": 1.2007042253521127,
+      "grad_norm": 0.30251818895339966,
+      "learning_rate": 7.496620013421654e-06,
+      "loss": 0.348,
+      "step": 4774
+    },
+    {
+      "epoch": 1.2009557344064385,
+      "grad_norm": 0.37606722116470337,
+      "learning_rate": 7.495352126341074e-06,
+      "loss": 0.3535,
+      "step": 4775
+    },
+    {
+      "epoch": 1.2012072434607646,
+      "grad_norm": 0.3614901602268219,
+      "learning_rate": 7.4940840255491975e-06,
+      "loss": 0.3766,
+      "step": 4776
+    },
+    {
+      "epoch": 1.2014587525150906,
+      "grad_norm": 0.33923235535621643,
+      "learning_rate": 7.492815711154626e-06,
+      "loss": 0.3423,
+      "step": 4777
+    },
+    {
+      "epoch": 1.2017102615694164,
+      "grad_norm": 0.388129860162735,
+      "learning_rate": 7.4915471832659835e-06,
+      "loss": 0.372,
+      "step": 4778
+    },
+    {
+      "epoch": 1.2019617706237424,
+      "grad_norm": 0.3465493619441986,
+      "learning_rate": 7.490278441991911e-06,
+      "loss": 0.3534,
+      "step": 4779
+    },
+    {
+      "epoch": 1.2022132796780685,
+      "grad_norm": 0.3515889644622803,
+      "learning_rate": 7.489009487441071e-06,
+      "loss": 0.3313,
+      "step": 4780
+    },
+    {
+      "epoch": 1.2024647887323945,
+      "grad_norm": 0.375966340303421,
+      "learning_rate": 7.4877403197221385e-06,
+      "loss": 0.4004,
+      "step": 4781
+    },
+    {
+      "epoch": 1.2027162977867203,
+      "grad_norm": 0.3232581317424774,
+      "learning_rate": 7.48647093894381e-06,
+      "loss": 0.3247,
+      "step": 4782
+    },
+    {
+      "epoch": 1.2029678068410463,
+      "grad_norm": 0.3182356059551239,
+      "learning_rate": 7.485201345214803e-06,
+      "loss": 0.3493,
+      "step": 4783
+    },
+    {
+      "epoch": 1.2032193158953723,
+      "grad_norm": 0.340154230594635,
+      "learning_rate": 7.483931538643847e-06,
+      "loss": 0.3646,
+      "step": 4784
+    },
+    {
+      "epoch": 1.2034708249496981,
+      "grad_norm": 0.35035526752471924,
+      "learning_rate": 7.482661519339696e-06,
+      "loss": 0.3693,
+      "step": 4785
+    },
+    {
+      "epoch": 1.2037223340040242,
+      "grad_norm": 0.34506088495254517,
+      "learning_rate": 7.481391287411115e-06,
+      "loss": 0.343,
+      "step": 4786
+    },
+    {
+      "epoch": 1.2039738430583502,
+      "grad_norm": 0.34625160694122314,
+      "learning_rate": 7.480120842966895e-06,
+      "loss": 0.3267,
+      "step": 4787
+    },
+    {
+      "epoch": 1.204225352112676,
+      "grad_norm": 0.35969218611717224,
+      "learning_rate": 7.478850186115839e-06,
+      "loss": 0.3706,
+      "step": 4788
+    },
+    {
+      "epoch": 1.204476861167002,
+      "grad_norm": 0.3208048641681671,
+      "learning_rate": 7.477579316966773e-06,
+      "loss": 0.3509,
+      "step": 4789
+    },
+    {
+      "epoch": 1.204728370221328,
+      "grad_norm": 0.34545454382896423,
+      "learning_rate": 7.4763082356285385e-06,
+      "loss": 0.3524,
+      "step": 4790
+    },
+    {
+      "epoch": 1.2049798792756539,
+      "grad_norm": 0.37139612436294556,
+      "learning_rate": 7.4750369422099955e-06,
+      "loss": 0.3832,
+      "step": 4791
+    },
+    {
+      "epoch": 1.20523138832998,
+      "grad_norm": 0.3318977952003479,
+      "learning_rate": 7.473765436820021e-06,
+      "loss": 0.3348,
+      "step": 4792
+    },
+    {
+      "epoch": 1.205482897384306,
+      "grad_norm": 0.34390851855278015,
+      "learning_rate": 7.472493719567513e-06,
+      "loss": 0.3608,
+      "step": 4793
+    },
+    {
+      "epoch": 1.2057344064386317,
+      "grad_norm": 0.3598691523075104,
+      "learning_rate": 7.471221790561387e-06,
+      "loss": 0.3632,
+      "step": 4794
+    },
+    {
+      "epoch": 1.2059859154929577,
+      "grad_norm": 0.3379955291748047,
+      "learning_rate": 7.469949649910574e-06,
+      "loss": 0.3586,
+      "step": 4795
+    },
+    {
+      "epoch": 1.2062374245472838,
+      "grad_norm": 0.37648218870162964,
+      "learning_rate": 7.468677297724025e-06,
+      "loss": 0.3536,
+      "step": 4796
+    },
+    {
+      "epoch": 1.2064889336016096,
+      "grad_norm": 0.3806565999984741,
+      "learning_rate": 7.46740473411071e-06,
+      "loss": 0.3358,
+      "step": 4797
+    },
+    {
+      "epoch": 1.2067404426559356,
+      "grad_norm": 0.34232839941978455,
+      "learning_rate": 7.466131959179614e-06,
+      "loss": 0.3444,
+      "step": 4798
+    },
+    {
+      "epoch": 1.2069919517102616,
+      "grad_norm": 0.3516295254230499,
+      "learning_rate": 7.464858973039746e-06,
+      "loss": 0.3442,
+      "step": 4799
+    },
+    {
+      "epoch": 1.2072434607645874,
+      "grad_norm": 0.3554040491580963,
+      "learning_rate": 7.4635857758001266e-06,
+      "loss": 0.3613,
+      "step": 4800
+    },
+    {
+      "epoch": 1.2074949698189135,
+      "grad_norm": 0.3366535007953644,
+      "learning_rate": 7.4623123675697976e-06,
+      "loss": 0.3493,
+      "step": 4801
+    },
+    {
+      "epoch": 1.2077464788732395,
+      "grad_norm": 0.34996628761291504,
+      "learning_rate": 7.461038748457818e-06,
+      "loss": 0.3765,
+      "step": 4802
+    },
+    {
+      "epoch": 1.2079979879275653,
+      "grad_norm": 0.36713096499443054,
+      "learning_rate": 7.459764918573264e-06,
+      "loss": 0.3513,
+      "step": 4803
+    },
+    {
+      "epoch": 1.2082494969818913,
+      "grad_norm": 0.3152211308479309,
+      "learning_rate": 7.4584908780252365e-06,
+      "loss": 0.3413,
+      "step": 4804
+    },
+    {
+      "epoch": 1.2085010060362174,
+      "grad_norm": 0.33181753754615784,
+      "learning_rate": 7.457216626922843e-06,
+      "loss": 0.3507,
+      "step": 4805
+    },
+    {
+      "epoch": 1.2087525150905432,
+      "grad_norm": 0.32960614562034607,
+      "learning_rate": 7.45594216537522e-06,
+      "loss": 0.3585,
+      "step": 4806
+    },
+    {
+      "epoch": 1.2090040241448692,
+      "grad_norm": 0.3321514427661896,
+      "learning_rate": 7.454667493491513e-06,
+      "loss": 0.3329,
+      "step": 4807
+    },
+    {
+      "epoch": 1.2092555331991952,
+      "grad_norm": 0.33364543318748474,
+      "learning_rate": 7.453392611380892e-06,
+      "loss": 0.3285,
+      "step": 4808
+    },
+    {
+      "epoch": 1.209507042253521,
+      "grad_norm": 0.31527218222618103,
+      "learning_rate": 7.452117519152542e-06,
+      "loss": 0.3499,
+      "step": 4809
+    },
+    {
+      "epoch": 1.209758551307847,
+      "grad_norm": 0.31813716888427734,
+      "learning_rate": 7.450842216915667e-06,
+      "loss": 0.3498,
+      "step": 4810
+    },
+    {
+      "epoch": 1.210010060362173,
+      "grad_norm": 0.30781570076942444,
+      "learning_rate": 7.449566704779488e-06,
+      "loss": 0.3651,
+      "step": 4811
+    },
+    {
+      "epoch": 1.2102615694164989,
+      "grad_norm": 0.31299492716789246,
+      "learning_rate": 7.448290982853247e-06,
+      "loss": 0.3392,
+      "step": 4812
+    },
+    {
+      "epoch": 1.210513078470825,
+      "grad_norm": 0.3456888794898987,
+      "learning_rate": 7.4470150512461966e-06,
+      "loss": 0.3628,
+      "step": 4813
+    },
+    {
+      "epoch": 1.210764587525151,
+      "grad_norm": 0.3564990758895874,
+      "learning_rate": 7.445738910067618e-06,
+      "loss": 0.3645,
+      "step": 4814
+    },
+    {
+      "epoch": 1.211016096579477,
+      "grad_norm": 0.32649949193000793,
+      "learning_rate": 7.444462559426802e-06,
+      "loss": 0.3589,
+      "step": 4815
+    },
+    {
+      "epoch": 1.2112676056338028,
+      "grad_norm": 0.36671027541160583,
+      "learning_rate": 7.44318599943306e-06,
+      "loss": 0.3756,
+      "step": 4816
+    },
+    {
+      "epoch": 1.2115191146881288,
+      "grad_norm": 0.36808088421821594,
+      "learning_rate": 7.441909230195722e-06,
+      "loss": 0.3589,
+      "step": 4817
+    },
+    {
+      "epoch": 1.2117706237424548,
+      "grad_norm": 0.35420069098472595,
+      "learning_rate": 7.440632251824135e-06,
+      "loss": 0.332,
+      "step": 4818
+    },
+    {
+      "epoch": 1.2120221327967806,
+      "grad_norm": 0.3255870044231415,
+      "learning_rate": 7.4393550644276635e-06,
+      "loss": 0.3622,
+      "step": 4819
+    },
+    {
+      "epoch": 1.2122736418511066,
+      "grad_norm": 0.335366427898407,
+      "learning_rate": 7.438077668115692e-06,
+      "loss": 0.3587,
+      "step": 4820
+    },
+    {
+      "epoch": 1.2125251509054327,
+      "grad_norm": 0.3825639486312866,
+      "learning_rate": 7.43680006299762e-06,
+      "loss": 0.3776,
+      "step": 4821
+    },
+    {
+      "epoch": 1.2127766599597585,
+      "grad_norm": 0.3819475769996643,
+      "learning_rate": 7.435522249182868e-06,
+      "loss": 0.3309,
+      "step": 4822
+    },
+    {
+      "epoch": 1.2130281690140845,
+      "grad_norm": 0.3363218605518341,
+      "learning_rate": 7.434244226780871e-06,
+      "loss": 0.3539,
+      "step": 4823
+    },
+    {
+      "epoch": 1.2132796780684105,
+      "grad_norm": 0.4024132490158081,
+      "learning_rate": 7.432965995901085e-06,
+      "loss": 0.3538,
+      "step": 4824
+    },
+    {
+      "epoch": 1.2135311871227363,
+      "grad_norm": 0.35047945380210876,
+      "learning_rate": 7.4316875566529825e-06,
+      "loss": 0.3228,
+      "step": 4825
+    },
+    {
+      "epoch": 1.2137826961770624,
+      "grad_norm": 0.3095787763595581,
+      "learning_rate": 7.430408909146053e-06,
+      "loss": 0.3695,
+      "step": 4826
+    },
+    {
+      "epoch": 1.2140342052313884,
+      "grad_norm": 0.33355486392974854,
+      "learning_rate": 7.429130053489805e-06,
+      "loss": 0.3494,
+      "step": 4827
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 0.37087491154670715,
+      "learning_rate": 7.427850989793764e-06,
+      "loss": 0.3527,
+      "step": 4828
+    },
+    {
+      "epoch": 1.2145372233400402,
+      "grad_norm": 0.3255065977573395,
+      "learning_rate": 7.426571718167476e-06,
+      "loss": 0.3543,
+      "step": 4829
+    },
+    {
+      "epoch": 1.2147887323943662,
+      "grad_norm": 0.33966347575187683,
+      "learning_rate": 7.4252922387205e-06,
+      "loss": 0.3526,
+      "step": 4830
+    },
+    {
+      "epoch": 1.2150402414486923,
+      "grad_norm": 0.32921141386032104,
+      "learning_rate": 7.424012551562416e-06,
+      "loss": 0.3662,
+      "step": 4831
+    },
+    {
+      "epoch": 1.215291750503018,
+      "grad_norm": 0.3616410791873932,
+      "learning_rate": 7.422732656802821e-06,
+      "loss": 0.3537,
+      "step": 4832
+    },
+    {
+      "epoch": 1.215543259557344,
+      "grad_norm": 0.3477363586425781,
+      "learning_rate": 7.421452554551332e-06,
+      "loss": 0.3633,
+      "step": 4833
+    },
+    {
+      "epoch": 1.2157947686116701,
+      "grad_norm": 0.34547752141952515,
+      "learning_rate": 7.420172244917579e-06,
+      "loss": 0.3455,
+      "step": 4834
+    },
+    {
+      "epoch": 1.216046277665996,
+      "grad_norm": 0.36257532238960266,
+      "learning_rate": 7.418891728011214e-06,
+      "loss": 0.3381,
+      "step": 4835
+    },
+    {
+      "epoch": 1.216297786720322,
+      "grad_norm": 0.34526896476745605,
+      "learning_rate": 7.417611003941905e-06,
+      "loss": 0.3347,
+      "step": 4836
+    },
+    {
+      "epoch": 1.216549295774648,
+      "grad_norm": 0.35130295157432556,
+      "learning_rate": 7.416330072819338e-06,
+      "loss": 0.3579,
+      "step": 4837
+    },
+    {
+      "epoch": 1.2168008048289738,
+      "grad_norm": 0.3254670202732086,
+      "learning_rate": 7.415048934753217e-06,
+      "loss": 0.3455,
+      "step": 4838
+    },
+    {
+      "epoch": 1.2170523138832998,
+      "grad_norm": 0.3543333113193512,
+      "learning_rate": 7.413767589853264e-06,
+      "loss": 0.3465,
+      "step": 4839
+    },
+    {
+      "epoch": 1.2173038229376258,
+      "grad_norm": 0.3948500156402588,
+      "learning_rate": 7.412486038229217e-06,
+      "loss": 0.354,
+      "step": 4840
+    },
+    {
+      "epoch": 1.2175553319919517,
+      "grad_norm": 0.3444294035434723,
+      "learning_rate": 7.411204279990832e-06,
+      "loss": 0.3694,
+      "step": 4841
+    },
+    {
+      "epoch": 1.2178068410462777,
+      "grad_norm": 0.35417085886001587,
+      "learning_rate": 7.4099223152478865e-06,
+      "loss": 0.3593,
+      "step": 4842
+    },
+    {
+      "epoch": 1.2180583501006037,
+      "grad_norm": 0.35232633352279663,
+      "learning_rate": 7.408640144110171e-06,
+      "loss": 0.3407,
+      "step": 4843
+    },
+    {
+      "epoch": 1.2183098591549295,
+      "grad_norm": 0.35325491428375244,
+      "learning_rate": 7.407357766687495e-06,
+      "loss": 0.3534,
+      "step": 4844
+    },
+    {
+      "epoch": 1.2185613682092555,
+      "grad_norm": 0.31295645236968994,
+      "learning_rate": 7.406075183089686e-06,
+      "loss": 0.3598,
+      "step": 4845
+    },
+    {
+      "epoch": 1.2188128772635816,
+      "grad_norm": 0.3566473424434662,
+      "learning_rate": 7.40479239342659e-06,
+      "loss": 0.3381,
+      "step": 4846
+    },
+    {
+      "epoch": 1.2190643863179074,
+      "grad_norm": 0.36591169238090515,
+      "learning_rate": 7.403509397808071e-06,
+      "loss": 0.3401,
+      "step": 4847
+    },
+    {
+      "epoch": 1.2193158953722334,
+      "grad_norm": 0.33974406123161316,
+      "learning_rate": 7.402226196344008e-06,
+      "loss": 0.3635,
+      "step": 4848
+    },
+    {
+      "epoch": 1.2195674044265594,
+      "grad_norm": 0.33850404620170593,
+      "learning_rate": 7.400942789144299e-06,
+      "loss": 0.37,
+      "step": 4849
+    },
+    {
+      "epoch": 1.2198189134808852,
+      "grad_norm": 0.34571510553359985,
+      "learning_rate": 7.399659176318861e-06,
+      "loss": 0.3681,
+      "step": 4850
+    },
+    {
+      "epoch": 1.2200704225352113,
+      "grad_norm": 0.29631301760673523,
+      "learning_rate": 7.398375357977626e-06,
+      "loss": 0.3839,
+      "step": 4851
+    },
+    {
+      "epoch": 1.2203219315895373,
+      "grad_norm": 0.33825013041496277,
+      "learning_rate": 7.397091334230547e-06,
+      "loss": 0.3593,
+      "step": 4852
+    },
+    {
+      "epoch": 1.220573440643863,
+      "grad_norm": 0.3286561667919159,
+      "learning_rate": 7.395807105187591e-06,
+      "loss": 0.3398,
+      "step": 4853
+    },
+    {
+      "epoch": 1.220824949698189,
+      "grad_norm": 0.34450599551200867,
+      "learning_rate": 7.3945226709587434e-06,
+      "loss": 0.3356,
+      "step": 4854
+    },
+    {
+      "epoch": 1.2210764587525151,
+      "grad_norm": 0.3374083638191223,
+      "learning_rate": 7.393238031654011e-06,
+      "loss": 0.342,
+      "step": 4855
+    },
+    {
+      "epoch": 1.221327967806841,
+      "grad_norm": 0.3444235920906067,
+      "learning_rate": 7.391953187383411e-06,
+      "loss": 0.3222,
+      "step": 4856
+    },
+    {
+      "epoch": 1.221579476861167,
+      "grad_norm": 0.33307015895843506,
+      "learning_rate": 7.390668138256987e-06,
+      "loss": 0.3378,
+      "step": 4857
+    },
+    {
+      "epoch": 1.221830985915493,
+      "grad_norm": 0.362884521484375,
+      "learning_rate": 7.38938288438479e-06,
+      "loss": 0.3765,
+      "step": 4858
+    },
+    {
+      "epoch": 1.2220824949698188,
+      "grad_norm": 0.30631768703460693,
+      "learning_rate": 7.388097425876899e-06,
+      "loss": 0.3387,
+      "step": 4859
+    },
+    {
+      "epoch": 1.2223340040241448,
+      "grad_norm": 0.3113538920879364,
+      "learning_rate": 7.386811762843404e-06,
+      "loss": 0.3577,
+      "step": 4860
+    },
+    {
+      "epoch": 1.2225855130784709,
+      "grad_norm": 0.335956871509552,
+      "learning_rate": 7.385525895394412e-06,
+      "loss": 0.3272,
+      "step": 4861
+    },
+    {
+      "epoch": 1.2228370221327967,
+      "grad_norm": 0.3435831367969513,
+      "learning_rate": 7.384239823640052e-06,
+      "loss": 0.3509,
+      "step": 4862
+    },
+    {
+      "epoch": 1.2230885311871227,
+      "grad_norm": 0.31683871150016785,
+      "learning_rate": 7.382953547690465e-06,
+      "loss": 0.3362,
+      "step": 4863
+    },
+    {
+      "epoch": 1.2233400402414487,
+      "grad_norm": 0.33224907517433167,
+      "learning_rate": 7.381667067655815e-06,
+      "loss": 0.3766,
+      "step": 4864
+    },
+    {
+      "epoch": 1.2235915492957747,
+      "grad_norm": 0.34509238600730896,
+      "learning_rate": 7.380380383646282e-06,
+      "loss": 0.3812,
+      "step": 4865
+    },
+    {
+      "epoch": 1.2238430583501005,
+      "grad_norm": 0.36535876989364624,
+      "learning_rate": 7.379093495772059e-06,
+      "loss": 0.3212,
+      "step": 4866
+    },
+    {
+      "epoch": 1.2240945674044266,
+      "grad_norm": 0.3264424204826355,
+      "learning_rate": 7.377806404143362e-06,
+      "loss": 0.3563,
+      "step": 4867
+    },
+    {
+      "epoch": 1.2243460764587526,
+      "grad_norm": 0.332707017660141,
+      "learning_rate": 7.376519108870423e-06,
+      "loss": 0.3651,
+      "step": 4868
+    },
+    {
+      "epoch": 1.2245975855130784,
+      "grad_norm": 0.3452300429344177,
+      "learning_rate": 7.375231610063488e-06,
+      "loss": 0.3535,
+      "step": 4869
+    },
+    {
+      "epoch": 1.2248490945674044,
+      "grad_norm": 0.2937973439693451,
+      "learning_rate": 7.373943907832826e-06,
+      "loss": 0.354,
+      "step": 4870
+    },
+    {
+      "epoch": 1.2251006036217305,
+      "grad_norm": 0.33667927980422974,
+      "learning_rate": 7.37265600228872e-06,
+      "loss": 0.3762,
+      "step": 4871
+    },
+    {
+      "epoch": 1.2253521126760563,
+      "grad_norm": 0.32756221294403076,
+      "learning_rate": 7.37136789354147e-06,
+      "loss": 0.35,
+      "step": 4872
+    },
+    {
+      "epoch": 1.2256036217303823,
+      "grad_norm": 0.3230566680431366,
+      "learning_rate": 7.370079581701396e-06,
+      "loss": 0.3338,
+      "step": 4873
+    },
+    {
+      "epoch": 1.2258551307847083,
+      "grad_norm": 0.3240143358707428,
+      "learning_rate": 7.368791066878832e-06,
+      "loss": 0.3427,
+      "step": 4874
+    },
+    {
+      "epoch": 1.2261066398390341,
+      "grad_norm": 0.3279706835746765,
+      "learning_rate": 7.367502349184132e-06,
+      "loss": 0.3649,
+      "step": 4875
+    },
+    {
+      "epoch": 1.2263581488933601,
+      "grad_norm": 0.35657429695129395,
+      "learning_rate": 7.366213428727669e-06,
+      "loss": 0.3473,
+      "step": 4876
+    },
+    {
+      "epoch": 1.2266096579476862,
+      "grad_norm": 0.3800789713859558,
+      "learning_rate": 7.364924305619827e-06,
+      "loss": 0.3424,
+      "step": 4877
+    },
+    {
+      "epoch": 1.226861167002012,
+      "grad_norm": 0.33746159076690674,
+      "learning_rate": 7.363634979971012e-06,
+      "loss": 0.3587,
+      "step": 4878
+    },
+    {
+      "epoch": 1.227112676056338,
+      "grad_norm": 0.319251149892807,
+      "learning_rate": 7.362345451891649e-06,
+      "loss": 0.36,
+      "step": 4879
+    },
+    {
+      "epoch": 1.227364185110664,
+      "grad_norm": 0.3327750265598297,
+      "learning_rate": 7.361055721492177e-06,
+      "loss": 0.3533,
+      "step": 4880
+    },
+    {
+      "epoch": 1.22761569416499,
+      "grad_norm": 0.3240911364555359,
+      "learning_rate": 7.359765788883053e-06,
+      "loss": 0.3473,
+      "step": 4881
+    },
+    {
+      "epoch": 1.2278672032193159,
+      "grad_norm": 0.3320063054561615,
+      "learning_rate": 7.358475654174753e-06,
+      "loss": 0.3587,
+      "step": 4882
+    },
+    {
+      "epoch": 1.2281187122736419,
+      "grad_norm": 0.31429430842399597,
+      "learning_rate": 7.357185317477766e-06,
+      "loss": 0.3512,
+      "step": 4883
+    },
+    {
+      "epoch": 1.228370221327968,
+      "grad_norm": 0.33201098442077637,
+      "learning_rate": 7.355894778902605e-06,
+      "loss": 0.3469,
+      "step": 4884
+    },
+    {
+      "epoch": 1.2286217303822937,
+      "grad_norm": 0.31669896841049194,
+      "learning_rate": 7.354604038559794e-06,
+      "loss": 0.3443,
+      "step": 4885
+    },
+    {
+      "epoch": 1.2288732394366197,
+      "grad_norm": 0.34272029995918274,
+      "learning_rate": 7.3533130965598765e-06,
+      "loss": 0.3383,
+      "step": 4886
+    },
+    {
+      "epoch": 1.2291247484909458,
+      "grad_norm": 0.3486393988132477,
+      "learning_rate": 7.352021953013415e-06,
+      "loss": 0.3458,
+      "step": 4887
+    },
+    {
+      "epoch": 1.2293762575452716,
+      "grad_norm": 0.383306622505188,
+      "learning_rate": 7.350730608030987e-06,
+      "loss": 0.3301,
+      "step": 4888
+    },
+    {
+      "epoch": 1.2296277665995976,
+      "grad_norm": 0.3175544738769531,
+      "learning_rate": 7.3494390617231896e-06,
+      "loss": 0.3521,
+      "step": 4889
+    },
+    {
+      "epoch": 1.2298792756539236,
+      "grad_norm": 0.32851359248161316,
+      "learning_rate": 7.348147314200634e-06,
+      "loss": 0.3298,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2301307847082494,
+      "grad_norm": 0.35966333746910095,
+      "learning_rate": 7.346855365573951e-06,
+      "loss": 0.3386,
+      "step": 4891
+    },
+    {
+      "epoch": 1.2303822937625755,
+      "grad_norm": 0.34041905403137207,
+      "learning_rate": 7.345563215953787e-06,
+      "loss": 0.3561,
+      "step": 4892
+    },
+    {
+      "epoch": 1.2306338028169015,
+      "grad_norm": 0.3270442485809326,
+      "learning_rate": 7.3442708654508085e-06,
+      "loss": 0.3536,
+      "step": 4893
+    },
+    {
+      "epoch": 1.2308853118712273,
+      "grad_norm": 0.3366083800792694,
+      "learning_rate": 7.342978314175695e-06,
+      "loss": 0.3454,
+      "step": 4894
+    },
+    {
+      "epoch": 1.2311368209255533,
+      "grad_norm": 0.3429853320121765,
+      "learning_rate": 7.341685562239148e-06,
+      "loss": 0.3413,
+      "step": 4895
+    },
+    {
+      "epoch": 1.2313883299798793,
+      "grad_norm": 0.3033541738986969,
+      "learning_rate": 7.3403926097518805e-06,
+      "loss": 0.3329,
+      "step": 4896
+    },
+    {
+      "epoch": 1.2316398390342052,
+      "grad_norm": 0.33623993396759033,
+      "learning_rate": 7.339099456824628e-06,
+      "loss": 0.3463,
+      "step": 4897
+    },
+    {
+      "epoch": 1.2318913480885312,
+      "grad_norm": 0.32286033034324646,
+      "learning_rate": 7.3378061035681415e-06,
+      "loss": 0.344,
+      "step": 4898
+    },
+    {
+      "epoch": 1.2321428571428572,
+      "grad_norm": 0.3481849431991577,
+      "learning_rate": 7.336512550093186e-06,
+      "loss": 0.3445,
+      "step": 4899
+    },
+    {
+      "epoch": 1.232394366197183,
+      "grad_norm": 0.3121998906135559,
+      "learning_rate": 7.335218796510548e-06,
+      "loss": 0.3416,
+      "step": 4900
+    },
+    {
+      "epoch": 1.232645875251509,
+      "grad_norm": 0.3597795367240906,
+      "learning_rate": 7.333924842931031e-06,
+      "loss": 0.3588,
+      "step": 4901
+    },
+    {
+      "epoch": 1.232897384305835,
+      "grad_norm": 0.31854021549224854,
+      "learning_rate": 7.332630689465449e-06,
+      "loss": 0.3392,
+      "step": 4902
+    },
+    {
+      "epoch": 1.2331488933601609,
+      "grad_norm": 0.35786551237106323,
+      "learning_rate": 7.331336336224643e-06,
+      "loss": 0.3464,
+      "step": 4903
+    },
+    {
+      "epoch": 1.233400402414487,
+      "grad_norm": 0.36554989218711853,
+      "learning_rate": 7.330041783319466e-06,
+      "loss": 0.3524,
+      "step": 4904
+    },
+    {
+      "epoch": 1.233651911468813,
+      "grad_norm": 0.3338761627674103,
+      "learning_rate": 7.328747030860786e-06,
+      "loss": 0.3245,
+      "step": 4905
+    },
+    {
+      "epoch": 1.2339034205231387,
+      "grad_norm": 0.33866626024246216,
+      "learning_rate": 7.327452078959491e-06,
+      "loss": 0.3601,
+      "step": 4906
+    },
+    {
+      "epoch": 1.2341549295774648,
+      "grad_norm": 0.3417625427246094,
+      "learning_rate": 7.326156927726485e-06,
+      "loss": 0.3723,
+      "step": 4907
+    },
+    {
+      "epoch": 1.2344064386317908,
+      "grad_norm": 0.402842253446579,
+      "learning_rate": 7.324861577272693e-06,
+      "loss": 0.3539,
+      "step": 4908
+    },
+    {
+      "epoch": 1.2346579476861166,
+      "grad_norm": 0.3404982089996338,
+      "learning_rate": 7.323566027709049e-06,
+      "loss": 0.3507,
+      "step": 4909
+    },
+    {
+      "epoch": 1.2349094567404426,
+      "grad_norm": 0.32714393734931946,
+      "learning_rate": 7.322270279146512e-06,
+      "loss": 0.3506,
+      "step": 4910
+    },
+    {
+      "epoch": 1.2351609657947686,
+      "grad_norm": 0.35655197501182556,
+      "learning_rate": 7.320974331696053e-06,
+      "loss": 0.3603,
+      "step": 4911
+    },
+    {
+      "epoch": 1.2354124748490944,
+      "grad_norm": 0.3561979830265045,
+      "learning_rate": 7.319678185468662e-06,
+      "loss": 0.3486,
+      "step": 4912
+    },
+    {
+      "epoch": 1.2356639839034205,
+      "grad_norm": 0.36184290051460266,
+      "learning_rate": 7.318381840575347e-06,
+      "loss": 0.3567,
+      "step": 4913
+    },
+    {
+      "epoch": 1.2359154929577465,
+      "grad_norm": 0.30532655119895935,
+      "learning_rate": 7.317085297127131e-06,
+      "loss": 0.3518,
+      "step": 4914
+    },
+    {
+      "epoch": 1.2361670020120725,
+      "grad_norm": 0.3469163477420807,
+      "learning_rate": 7.315788555235055e-06,
+      "loss": 0.3704,
+      "step": 4915
+    },
+    {
+      "epoch": 1.2364185110663983,
+      "grad_norm": 0.34762290120124817,
+      "learning_rate": 7.314491615010178e-06,
+      "loss": 0.3404,
+      "step": 4916
+    },
+    {
+      "epoch": 1.2366700201207244,
+      "grad_norm": 0.33168843388557434,
+      "learning_rate": 7.313194476563572e-06,
+      "loss": 0.3181,
+      "step": 4917
+    },
+    {
+      "epoch": 1.2369215291750504,
+      "grad_norm": 0.34730586409568787,
+      "learning_rate": 7.311897140006331e-06,
+      "loss": 0.37,
+      "step": 4918
+    },
+    {
+      "epoch": 1.2371730382293762,
+      "grad_norm": 0.3283190131187439,
+      "learning_rate": 7.310599605449563e-06,
+      "loss": 0.3246,
+      "step": 4919
+    },
+    {
+      "epoch": 1.2374245472837022,
+      "grad_norm": 0.34569719433784485,
+      "learning_rate": 7.309301873004394e-06,
+      "loss": 0.3726,
+      "step": 4920
+    },
+    {
+      "epoch": 1.2376760563380282,
+      "grad_norm": 0.33751311898231506,
+      "learning_rate": 7.308003942781966e-06,
+      "loss": 0.3524,
+      "step": 4921
+    },
+    {
+      "epoch": 1.237927565392354,
+      "grad_norm": 0.31516730785369873,
+      "learning_rate": 7.30670581489344e-06,
+      "loss": 0.3436,
+      "step": 4922
+    },
+    {
+      "epoch": 1.23817907444668,
+      "grad_norm": 0.40124109387397766,
+      "learning_rate": 7.305407489449991e-06,
+      "loss": 0.3557,
+      "step": 4923
+    },
+    {
+      "epoch": 1.238430583501006,
+      "grad_norm": 0.31682252883911133,
+      "learning_rate": 7.3041089665628125e-06,
+      "loss": 0.3739,
+      "step": 4924
+    },
+    {
+      "epoch": 1.238682092555332,
+      "grad_norm": 0.32461637258529663,
+      "learning_rate": 7.302810246343117e-06,
+      "loss": 0.3655,
+      "step": 4925
+    },
+    {
+      "epoch": 1.238933601609658,
+      "grad_norm": 0.32863402366638184,
+      "learning_rate": 7.30151132890213e-06,
+      "loss": 0.3388,
+      "step": 4926
+    },
+    {
+      "epoch": 1.239185110663984,
+      "grad_norm": 0.35274699330329895,
+      "learning_rate": 7.300212214351095e-06,
+      "loss": 0.3431,
+      "step": 4927
+    },
+    {
+      "epoch": 1.2394366197183098,
+      "grad_norm": 0.343986839056015,
+      "learning_rate": 7.298912902801274e-06,
+      "loss": 0.3763,
+      "step": 4928
+    },
+    {
+      "epoch": 1.2396881287726358,
+      "grad_norm": 0.3199062943458557,
+      "learning_rate": 7.297613394363946e-06,
+      "loss": 0.3223,
+      "step": 4929
+    },
+    {
+      "epoch": 1.2399396378269618,
+      "grad_norm": 0.3335794508457184,
+      "learning_rate": 7.296313689150404e-06,
+      "loss": 0.37,
+      "step": 4930
+    },
+    {
+      "epoch": 1.2401911468812878,
+      "grad_norm": 0.365524560213089,
+      "learning_rate": 7.295013787271959e-06,
+      "loss": 0.3746,
+      "step": 4931
+    },
+    {
+      "epoch": 1.2404426559356136,
+      "grad_norm": 0.3475268483161926,
+      "learning_rate": 7.293713688839941e-06,
+      "loss": 0.3658,
+      "step": 4932
+    },
+    {
+      "epoch": 1.2406941649899397,
+      "grad_norm": 0.3196357786655426,
+      "learning_rate": 7.292413393965696e-06,
+      "loss": 0.334,
+      "step": 4933
+    },
+    {
+      "epoch": 1.2409456740442657,
+      "grad_norm": 0.3176928460597992,
+      "learning_rate": 7.291112902760584e-06,
+      "loss": 0.3439,
+      "step": 4934
+    },
+    {
+      "epoch": 1.2411971830985915,
+      "grad_norm": 0.3524039685726166,
+      "learning_rate": 7.289812215335987e-06,
+      "loss": 0.3795,
+      "step": 4935
+    },
+    {
+      "epoch": 1.2414486921529175,
+      "grad_norm": 0.3586216866970062,
+      "learning_rate": 7.288511331803296e-06,
+      "loss": 0.3812,
+      "step": 4936
+    },
+    {
+      "epoch": 1.2417002012072436,
+      "grad_norm": 0.31926751136779785,
+      "learning_rate": 7.2872102522739286e-06,
+      "loss": 0.3386,
+      "step": 4937
+    },
+    {
+      "epoch": 1.2419517102615694,
+      "grad_norm": 0.338290274143219,
+      "learning_rate": 7.285908976859312e-06,
+      "loss": 0.3553,
+      "step": 4938
+    },
+    {
+      "epoch": 1.2422032193158954,
+      "grad_norm": 0.3404878079891205,
+      "learning_rate": 7.284607505670891e-06,
+      "loss": 0.3486,
+      "step": 4939
+    },
+    {
+      "epoch": 1.2424547283702214,
+      "grad_norm": 0.3355543613433838,
+      "learning_rate": 7.283305838820129e-06,
+      "loss": 0.3815,
+      "step": 4940
+    },
+    {
+      "epoch": 1.2427062374245472,
+      "grad_norm": 0.3367030620574951,
+      "learning_rate": 7.282003976418508e-06,
+      "loss": 0.3753,
+      "step": 4941
+    },
+    {
+      "epoch": 1.2429577464788732,
+      "grad_norm": 0.3236370086669922,
+      "learning_rate": 7.280701918577521e-06,
+      "loss": 0.3493,
+      "step": 4942
+    },
+    {
+      "epoch": 1.2432092555331993,
+      "grad_norm": 0.3462064862251282,
+      "learning_rate": 7.279399665408684e-06,
+      "loss": 0.3639,
+      "step": 4943
+    },
+    {
+      "epoch": 1.243460764587525,
+      "grad_norm": 0.35062482953071594,
+      "learning_rate": 7.278097217023523e-06,
+      "loss": 0.3644,
+      "step": 4944
+    },
+    {
+      "epoch": 1.243712273641851,
+      "grad_norm": 0.38055381178855896,
+      "learning_rate": 7.27679457353359e-06,
+      "loss": 0.3517,
+      "step": 4945
+    },
+    {
+      "epoch": 1.2439637826961771,
+      "grad_norm": 0.36633816361427307,
+      "learning_rate": 7.275491735050444e-06,
+      "loss": 0.36,
+      "step": 4946
+    },
+    {
+      "epoch": 1.244215291750503,
+      "grad_norm": 0.3013075590133667,
+      "learning_rate": 7.274188701685666e-06,
+      "loss": 0.3564,
+      "step": 4947
+    },
+    {
+      "epoch": 1.244466800804829,
+      "grad_norm": 0.35343462228775024,
+      "learning_rate": 7.272885473550855e-06,
+      "loss": 0.3335,
+      "step": 4948
+    },
+    {
+      "epoch": 1.244718309859155,
+      "grad_norm": 0.337429404258728,
+      "learning_rate": 7.271582050757619e-06,
+      "loss": 0.3329,
+      "step": 4949
+    },
+    {
+      "epoch": 1.2449698189134808,
+      "grad_norm": 0.3367510735988617,
+      "learning_rate": 7.2702784334175945e-06,
+      "loss": 0.3195,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2452213279678068,
+      "grad_norm": 0.32875868678092957,
+      "learning_rate": 7.268974621642424e-06,
+      "loss": 0.3472,
+      "step": 4951
+    },
+    {
+      "epoch": 1.2454728370221329,
+      "grad_norm": 0.3070097863674164,
+      "learning_rate": 7.267670615543771e-06,
+      "loss": 0.3493,
+      "step": 4952
+    },
+    {
+      "epoch": 1.2457243460764587,
+      "grad_norm": 0.3756659924983978,
+      "learning_rate": 7.266366415233317e-06,
+      "loss": 0.3728,
+      "step": 4953
+    },
+    {
+      "epoch": 1.2459758551307847,
+      "grad_norm": 0.3073340356349945,
+      "learning_rate": 7.265062020822757e-06,
+      "loss": 0.3499,
+      "step": 4954
+    },
+    {
+      "epoch": 1.2462273641851107,
+      "grad_norm": 0.3308236300945282,
+      "learning_rate": 7.263757432423806e-06,
+      "loss": 0.385,
+      "step": 4955
+    },
+    {
+      "epoch": 1.2464788732394365,
+      "grad_norm": 0.3272673487663269,
+      "learning_rate": 7.262452650148193e-06,
+      "loss": 0.3626,
+      "step": 4956
+    },
+    {
+      "epoch": 1.2467303822937625,
+      "grad_norm": 0.33081069588661194,
+      "learning_rate": 7.2611476741076636e-06,
+      "loss": 0.3471,
+      "step": 4957
+    },
+    {
+      "epoch": 1.2469818913480886,
+      "grad_norm": 0.33813828229904175,
+      "learning_rate": 7.2598425044139835e-06,
+      "loss": 0.3423,
+      "step": 4958
+    },
+    {
+      "epoch": 1.2472334004024144,
+      "grad_norm": 0.319319486618042,
+      "learning_rate": 7.258537141178928e-06,
+      "loss": 0.3595,
+      "step": 4959
+    },
+    {
+      "epoch": 1.2474849094567404,
+      "grad_norm": 0.30126598477363586,
+      "learning_rate": 7.257231584514297e-06,
+      "loss": 0.3368,
+      "step": 4960
+    },
+    {
+      "epoch": 1.2477364185110664,
+      "grad_norm": 0.3376123309135437,
+      "learning_rate": 7.255925834531902e-06,
+      "loss": 0.34,
+      "step": 4961
+    },
+    {
+      "epoch": 1.2479879275653922,
+      "grad_norm": 0.3421916663646698,
+      "learning_rate": 7.254619891343572e-06,
+      "loss": 0.3507,
+      "step": 4962
+    },
+    {
+      "epoch": 1.2482394366197183,
+      "grad_norm": 0.29679960012435913,
+      "learning_rate": 7.253313755061153e-06,
+      "loss": 0.3465,
+      "step": 4963
+    },
+    {
+      "epoch": 1.2484909456740443,
+      "grad_norm": 0.3243803381919861,
+      "learning_rate": 7.252007425796508e-06,
+      "loss": 0.3479,
+      "step": 4964
+    },
+    {
+      "epoch": 1.2487424547283703,
+      "grad_norm": 0.3532167971134186,
+      "learning_rate": 7.2507009036615136e-06,
+      "loss": 0.346,
+      "step": 4965
+    },
+    {
+      "epoch": 1.2489939637826961,
+      "grad_norm": 0.30621767044067383,
+      "learning_rate": 7.24939418876807e-06,
+      "loss": 0.3443,
+      "step": 4966
+    },
+    {
+      "epoch": 1.2492454728370221,
+      "grad_norm": 0.3399735689163208,
+      "learning_rate": 7.248087281228085e-06,
+      "loss": 0.3302,
+      "step": 4967
+    },
+    {
+      "epoch": 1.2494969818913482,
+      "grad_norm": 0.34660354256629944,
+      "learning_rate": 7.246780181153489e-06,
+      "loss": 0.3406,
+      "step": 4968
+    },
+    {
+      "epoch": 1.249748490945674,
+      "grad_norm": 0.3369465470314026,
+      "learning_rate": 7.2454728886562255e-06,
+      "loss": 0.3625,
+      "step": 4969
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3431127369403839,
+      "learning_rate": 7.244165403848257e-06,
+      "loss": 0.3365,
+      "step": 4970
+    },
+    {
+      "epoch": 1.250251509054326,
+      "grad_norm": 0.3352718651294708,
+      "learning_rate": 7.242857726841561e-06,
+      "loss": 0.3753,
+      "step": 4971
+    },
+    {
+      "epoch": 1.2505030181086518,
+      "grad_norm": 0.3221033215522766,
+      "learning_rate": 7.2415498577481325e-06,
+      "loss": 0.342,
+      "step": 4972
+    },
+    {
+      "epoch": 1.2507545271629779,
+      "grad_norm": 0.33727914094924927,
+      "learning_rate": 7.240241796679981e-06,
+      "loss": 0.3419,
+      "step": 4973
+    },
+    {
+      "epoch": 1.2510060362173039,
+      "grad_norm": 0.3359692096710205,
+      "learning_rate": 7.238933543749135e-06,
+      "loss": 0.3671,
+      "step": 4974
+    },
+    {
+      "epoch": 1.25125754527163,
+      "grad_norm": 0.32565391063690186,
+      "learning_rate": 7.237625099067638e-06,
+      "loss": 0.3667,
+      "step": 4975
+    },
+    {
+      "epoch": 1.2515090543259557,
+      "grad_norm": 0.310983806848526,
+      "learning_rate": 7.236316462747548e-06,
+      "loss": 0.3488,
+      "step": 4976
+    },
+    {
+      "epoch": 1.2517605633802817,
+      "grad_norm": 0.3416151702404022,
+      "learning_rate": 7.235007634900945e-06,
+      "loss": 0.3496,
+      "step": 4977
+    },
+    {
+      "epoch": 1.2520120724346078,
+      "grad_norm": 0.3046923577785492,
+      "learning_rate": 7.23369861563992e-06,
+      "loss": 0.3299,
+      "step": 4978
+    },
+    {
+      "epoch": 1.2522635814889336,
+      "grad_norm": 0.3572548031806946,
+      "learning_rate": 7.232389405076583e-06,
+      "loss": 0.3538,
+      "step": 4979
+    },
+    {
+      "epoch": 1.2525150905432596,
+      "grad_norm": 0.32661741971969604,
+      "learning_rate": 7.231080003323059e-06,
+      "loss": 0.3715,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2527665995975856,
+      "grad_norm": 0.3313870131969452,
+      "learning_rate": 7.22977041049149e-06,
+      "loss": 0.3579,
+      "step": 4981
+    },
+    {
+      "epoch": 1.2530181086519114,
+      "grad_norm": 0.3333343267440796,
+      "learning_rate": 7.228460626694034e-06,
+      "loss": 0.318,
+      "step": 4982
+    },
+    {
+      "epoch": 1.2532696177062375,
+      "grad_norm": 0.3657362163066864,
+      "learning_rate": 7.227150652042868e-06,
+      "loss": 0.3484,
+      "step": 4983
+    },
+    {
+      "epoch": 1.2535211267605635,
+      "grad_norm": 0.3520519733428955,
+      "learning_rate": 7.225840486650181e-06,
+      "loss": 0.3439,
+      "step": 4984
+    },
+    {
+      "epoch": 1.2537726358148893,
+      "grad_norm": 0.3073820173740387,
+      "learning_rate": 7.2245301306281815e-06,
+      "loss": 0.3389,
+      "step": 4985
+    },
+    {
+      "epoch": 1.2540241448692153,
+      "grad_norm": 0.3221682608127594,
+      "learning_rate": 7.223219584089091e-06,
+      "loss": 0.3708,
+      "step": 4986
+    },
+    {
+      "epoch": 1.2542756539235413,
+      "grad_norm": 0.32095620036125183,
+      "learning_rate": 7.221908847145153e-06,
+      "loss": 0.3656,
+      "step": 4987
+    },
+    {
+      "epoch": 1.2545271629778671,
+      "grad_norm": 0.3564129173755646,
+      "learning_rate": 7.220597919908624e-06,
+      "loss": 0.3757,
+      "step": 4988
+    },
+    {
+      "epoch": 1.2547786720321932,
+      "grad_norm": 0.33230066299438477,
+      "learning_rate": 7.219286802491774e-06,
+      "loss": 0.3458,
+      "step": 4989
+    },
+    {
+      "epoch": 1.2550301810865192,
+      "grad_norm": 0.33365389704704285,
+      "learning_rate": 7.217975495006892e-06,
+      "loss": 0.3577,
+      "step": 4990
+    },
+    {
+      "epoch": 1.255281690140845,
+      "grad_norm": 0.3553404211997986,
+      "learning_rate": 7.216663997566285e-06,
+      "loss": 0.3845,
+      "step": 4991
+    },
+    {
+      "epoch": 1.255533199195171,
+      "grad_norm": 0.3103611171245575,
+      "learning_rate": 7.215352310282275e-06,
+      "loss": 0.3408,
+      "step": 4992
+    },
+    {
+      "epoch": 1.255784708249497,
+      "grad_norm": 0.34521424770355225,
+      "learning_rate": 7.2140404332671986e-06,
+      "loss": 0.342,
+      "step": 4993
+    },
+    {
+      "epoch": 1.2560362173038229,
+      "grad_norm": 0.31152811646461487,
+      "learning_rate": 7.212728366633411e-06,
+      "loss": 0.3621,
+      "step": 4994
+    },
+    {
+      "epoch": 1.256287726358149,
+      "grad_norm": 0.335000604391098,
+      "learning_rate": 7.211416110493279e-06,
+      "loss": 0.3201,
+      "step": 4995
+    },
+    {
+      "epoch": 1.256539235412475,
+      "grad_norm": 0.3560849130153656,
+      "learning_rate": 7.210103664959194e-06,
+      "loss": 0.373,
+      "step": 4996
+    },
+    {
+      "epoch": 1.2567907444668007,
+      "grad_norm": 0.357250452041626,
+      "learning_rate": 7.2087910301435545e-06,
+      "loss": 0.343,
+      "step": 4997
+    },
+    {
+      "epoch": 1.2570422535211268,
+      "grad_norm": 0.32552990317344666,
+      "learning_rate": 7.207478206158782e-06,
+      "loss": 0.3473,
+      "step": 4998
+    },
+    {
+      "epoch": 1.2572937625754528,
+      "grad_norm": 0.3105219006538391,
+      "learning_rate": 7.2061651931173115e-06,
+      "loss": 0.3472,
+      "step": 4999
+    },
+    {
+      "epoch": 1.2575452716297786,
+      "grad_norm": 0.3435421586036682,
+      "learning_rate": 7.204851991131594e-06,
+      "loss": 0.3495,
+      "step": 5000
+    },
+    {
+      "epoch": 1.2577967806841046,
+      "grad_norm": 0.34035179018974304,
+      "learning_rate": 7.203538600314096e-06,
+      "loss": 0.3422,
+      "step": 5001
+    },
+    {
+      "epoch": 1.2580482897384306,
+      "grad_norm": 0.3376530408859253,
+      "learning_rate": 7.2022250207773035e-06,
+      "loss": 0.3356,
+      "step": 5002
+    },
+    {
+      "epoch": 1.2582997987927564,
+      "grad_norm": 0.3429863750934601,
+      "learning_rate": 7.200911252633714e-06,
+      "loss": 0.3598,
+      "step": 5003
+    },
+    {
+      "epoch": 1.2585513078470825,
+      "grad_norm": 0.3484230935573578,
+      "learning_rate": 7.199597295995846e-06,
+      "loss": 0.3519,
+      "step": 5004
+    },
+    {
+      "epoch": 1.2588028169014085,
+      "grad_norm": 0.35319894552230835,
+      "learning_rate": 7.1982831509762294e-06,
+      "loss": 0.3632,
+      "step": 5005
+    },
+    {
+      "epoch": 1.2590543259557343,
+      "grad_norm": 0.3312159478664398,
+      "learning_rate": 7.196968817687413e-06,
+      "loss": 0.3324,
+      "step": 5006
+    },
+    {
+      "epoch": 1.2593058350100603,
+      "grad_norm": 0.34530702233314514,
+      "learning_rate": 7.195654296241963e-06,
+      "loss": 0.3721,
+      "step": 5007
+    },
+    {
+      "epoch": 1.2595573440643864,
+      "grad_norm": 0.3831932842731476,
+      "learning_rate": 7.194339586752457e-06,
+      "loss": 0.3432,
+      "step": 5008
+    },
+    {
+      "epoch": 1.2598088531187122,
+      "grad_norm": 0.3508305847644806,
+      "learning_rate": 7.193024689331493e-06,
+      "loss": 0.3323,
+      "step": 5009
+    },
+    {
+      "epoch": 1.2600603621730382,
+      "grad_norm": 0.33183974027633667,
+      "learning_rate": 7.1917096040916835e-06,
+      "loss": 0.3467,
+      "step": 5010
+    },
+    {
+      "epoch": 1.2603118712273642,
+      "grad_norm": 0.33060377836227417,
+      "learning_rate": 7.190394331145659e-06,
+      "loss": 0.3598,
+      "step": 5011
+    },
+    {
+      "epoch": 1.26056338028169,
+      "grad_norm": 0.3584049940109253,
+      "learning_rate": 7.189078870606063e-06,
+      "loss": 0.333,
+      "step": 5012
+    },
+    {
+      "epoch": 1.260814889336016,
+      "grad_norm": 0.35449668765068054,
+      "learning_rate": 7.187763222585556e-06,
+      "loss": 0.37,
+      "step": 5013
+    },
+    {
+      "epoch": 1.261066398390342,
+      "grad_norm": 0.34245765209198,
+      "learning_rate": 7.186447387196815e-06,
+      "loss": 0.3424,
+      "step": 5014
+    },
+    {
+      "epoch": 1.2613179074446679,
+      "grad_norm": 0.3323255777359009,
+      "learning_rate": 7.1851313645525356e-06,
+      "loss": 0.3614,
+      "step": 5015
+    },
+    {
+      "epoch": 1.261569416498994,
+      "grad_norm": 0.35160186886787415,
+      "learning_rate": 7.183815154765423e-06,
+      "loss": 0.3396,
+      "step": 5016
+    },
+    {
+      "epoch": 1.26182092555332,
+      "grad_norm": 0.31636881828308105,
+      "learning_rate": 7.182498757948207e-06,
+      "loss": 0.3336,
+      "step": 5017
+    },
+    {
+      "epoch": 1.262072434607646,
+      "grad_norm": 0.36350032687187195,
+      "learning_rate": 7.181182174213623e-06,
+      "loss": 0.3448,
+      "step": 5018
+    },
+    {
+      "epoch": 1.2623239436619718,
+      "grad_norm": 0.3364773690700531,
+      "learning_rate": 7.179865403674433e-06,
+      "loss": 0.3263,
+      "step": 5019
+    },
+    {
+      "epoch": 1.2625754527162978,
+      "grad_norm": 0.32822203636169434,
+      "learning_rate": 7.178548446443407e-06,
+      "loss": 0.3504,
+      "step": 5020
+    },
+    {
+      "epoch": 1.2628269617706238,
+      "grad_norm": 0.33104854822158813,
+      "learning_rate": 7.177231302633337e-06,
+      "loss": 0.3321,
+      "step": 5021
+    },
+    {
+      "epoch": 1.2630784708249496,
+      "grad_norm": 0.34756192564964294,
+      "learning_rate": 7.175913972357025e-06,
+      "loss": 0.3538,
+      "step": 5022
+    },
+    {
+      "epoch": 1.2633299798792756,
+      "grad_norm": 0.35056936740875244,
+      "learning_rate": 7.174596455727295e-06,
+      "loss": 0.352,
+      "step": 5023
+    },
+    {
+      "epoch": 1.2635814889336017,
+      "grad_norm": 0.3366295099258423,
+      "learning_rate": 7.173278752856983e-06,
+      "loss": 0.3545,
+      "step": 5024
+    },
+    {
+      "epoch": 1.2638329979879277,
+      "grad_norm": 0.3299401104450226,
+      "learning_rate": 7.171960863858941e-06,
+      "loss": 0.3385,
+      "step": 5025
+    },
+    {
+      "epoch": 1.2640845070422535,
+      "grad_norm": 0.3155493140220642,
+      "learning_rate": 7.17064278884604e-06,
+      "loss": 0.3304,
+      "step": 5026
+    },
+    {
+      "epoch": 1.2643360160965795,
+      "grad_norm": 0.3616369068622589,
+      "learning_rate": 7.169324527931162e-06,
+      "loss": 0.3593,
+      "step": 5027
+    },
+    {
+      "epoch": 1.2645875251509056,
+      "grad_norm": 0.3170899748802185,
+      "learning_rate": 7.16800608122721e-06,
+      "loss": 0.3077,
+      "step": 5028
+    },
+    {
+      "epoch": 1.2648390342052314,
+      "grad_norm": 0.3360782563686371,
+      "learning_rate": 7.1666874488471e-06,
+      "loss": 0.3558,
+      "step": 5029
+    },
+    {
+      "epoch": 1.2650905432595574,
+      "grad_norm": 0.32805049419403076,
+      "learning_rate": 7.165368630903766e-06,
+      "loss": 0.332,
+      "step": 5030
+    },
+    {
+      "epoch": 1.2653420523138834,
+      "grad_norm": 0.34620416164398193,
+      "learning_rate": 7.164049627510154e-06,
+      "loss": 0.3559,
+      "step": 5031
+    },
+    {
+      "epoch": 1.2655935613682092,
+      "grad_norm": 0.33738458156585693,
+      "learning_rate": 7.1627304387792285e-06,
+      "loss": 0.355,
+      "step": 5032
+    },
+    {
+      "epoch": 1.2658450704225352,
+      "grad_norm": 0.3080560266971588,
+      "learning_rate": 7.161411064823973e-06,
+      "loss": 0.3521,
+      "step": 5033
+    },
+    {
+      "epoch": 1.2660965794768613,
+      "grad_norm": 0.3708006739616394,
+      "learning_rate": 7.160091505757381e-06,
+      "loss": 0.384,
+      "step": 5034
+    },
+    {
+      "epoch": 1.266348088531187,
+      "grad_norm": 0.34174585342407227,
+      "learning_rate": 7.158771761692464e-06,
+      "loss": 0.3298,
+      "step": 5035
+    },
+    {
+      "epoch": 1.266599597585513,
+      "grad_norm": 0.3278071880340576,
+      "learning_rate": 7.157451832742253e-06,
+      "loss": 0.3747,
+      "step": 5036
+    },
+    {
+      "epoch": 1.2668511066398391,
+      "grad_norm": 0.3372129797935486,
+      "learning_rate": 7.156131719019789e-06,
+      "loss": 0.371,
+      "step": 5037
+    },
+    {
+      "epoch": 1.267102615694165,
+      "grad_norm": 0.3328630030155182,
+      "learning_rate": 7.15481142063813e-06,
+      "loss": 0.3402,
+      "step": 5038
+    },
+    {
+      "epoch": 1.267354124748491,
+      "grad_norm": 0.3106217086315155,
+      "learning_rate": 7.1534909377103555e-06,
+      "loss": 0.3318,
+      "step": 5039
+    },
+    {
+      "epoch": 1.267605633802817,
+      "grad_norm": 0.3225652575492859,
+      "learning_rate": 7.152170270349553e-06,
+      "loss": 0.3657,
+      "step": 5040
+    },
+    {
+      "epoch": 1.2678571428571428,
+      "grad_norm": 0.3484492003917694,
+      "learning_rate": 7.1508494186688305e-06,
+      "loss": 0.3852,
+      "step": 5041
+    },
+    {
+      "epoch": 1.2681086519114688,
+      "grad_norm": 0.3207867741584778,
+      "learning_rate": 7.149528382781312e-06,
+      "loss": 0.3607,
+      "step": 5042
+    },
+    {
+      "epoch": 1.2683601609657948,
+      "grad_norm": 0.32192859053611755,
+      "learning_rate": 7.148207162800135e-06,
+      "loss": 0.3463,
+      "step": 5043
+    },
+    {
+      "epoch": 1.2686116700201207,
+      "grad_norm": 0.31769081950187683,
+      "learning_rate": 7.146885758838453e-06,
+      "loss": 0.3437,
+      "step": 5044
+    },
+    {
+      "epoch": 1.2688631790744467,
+      "grad_norm": 0.3407571315765381,
+      "learning_rate": 7.145564171009437e-06,
+      "loss": 0.3505,
+      "step": 5045
+    },
+    {
+      "epoch": 1.2691146881287727,
+      "grad_norm": 0.32075372338294983,
+      "learning_rate": 7.144242399426272e-06,
+      "loss": 0.3455,
+      "step": 5046
+    },
+    {
+      "epoch": 1.2693661971830985,
+      "grad_norm": 0.30012497305870056,
+      "learning_rate": 7.14292044420216e-06,
+      "loss": 0.33,
+      "step": 5047
+    },
+    {
+      "epoch": 1.2696177062374245,
+      "grad_norm": 0.3274124264717102,
+      "learning_rate": 7.141598305450319e-06,
+      "loss": 0.3383,
+      "step": 5048
+    },
+    {
+      "epoch": 1.2698692152917506,
+      "grad_norm": 0.3339790403842926,
+      "learning_rate": 7.1402759832839795e-06,
+      "loss": 0.3598,
+      "step": 5049
+    },
+    {
+      "epoch": 1.2701207243460764,
+      "grad_norm": 0.33048734068870544,
+      "learning_rate": 7.138953477816393e-06,
+      "loss": 0.3401,
+      "step": 5050
+    },
+    {
+      "epoch": 1.2703722334004024,
+      "grad_norm": 0.3411872386932373,
+      "learning_rate": 7.137630789160821e-06,
+      "loss": 0.3694,
+      "step": 5051
+    },
+    {
+      "epoch": 1.2706237424547284,
+      "grad_norm": 0.3702620267868042,
+      "learning_rate": 7.136307917430545e-06,
+      "loss": 0.3455,
+      "step": 5052
+    },
+    {
+      "epoch": 1.2708752515090542,
+      "grad_norm": 0.3102932274341583,
+      "learning_rate": 7.1349848627388616e-06,
+      "loss": 0.3652,
+      "step": 5053
+    },
+    {
+      "epoch": 1.2711267605633803,
+      "grad_norm": 0.3049952983856201,
+      "learning_rate": 7.13366162519908e-06,
+      "loss": 0.3456,
+      "step": 5054
+    },
+    {
+      "epoch": 1.2713782696177063,
+      "grad_norm": 0.34085172414779663,
+      "learning_rate": 7.132338204924529e-06,
+      "loss": 0.3461,
+      "step": 5055
+    },
+    {
+      "epoch": 1.271629778672032,
+      "grad_norm": 0.33940356969833374,
+      "learning_rate": 7.131014602028551e-06,
+      "loss": 0.3462,
+      "step": 5056
+    },
+    {
+      "epoch": 1.2718812877263581,
+      "grad_norm": 0.3196280896663666,
+      "learning_rate": 7.129690816624504e-06,
+      "loss": 0.357,
+      "step": 5057
+    },
+    {
+      "epoch": 1.2721327967806841,
+      "grad_norm": 0.3302951157093048,
+      "learning_rate": 7.128366848825761e-06,
+      "loss": 0.3587,
+      "step": 5058
+    },
+    {
+      "epoch": 1.27238430583501,
+      "grad_norm": 0.3771762549877167,
+      "learning_rate": 7.127042698745712e-06,
+      "loss": 0.3846,
+      "step": 5059
+    },
+    {
+      "epoch": 1.272635814889336,
+      "grad_norm": 0.3365668058395386,
+      "learning_rate": 7.125718366497763e-06,
+      "loss": 0.3886,
+      "step": 5060
+    },
+    {
+      "epoch": 1.272887323943662,
+      "grad_norm": 0.299969345331192,
+      "learning_rate": 7.124393852195335e-06,
+      "loss": 0.3545,
+      "step": 5061
+    },
+    {
+      "epoch": 1.2731388329979878,
+      "grad_norm": 0.3523898422718048,
+      "learning_rate": 7.123069155951864e-06,
+      "loss": 0.3674,
+      "step": 5062
+    },
+    {
+      "epoch": 1.2733903420523138,
+      "grad_norm": 0.3061320185661316,
+      "learning_rate": 7.1217442778808e-06,
+      "loss": 0.3479,
+      "step": 5063
+    },
+    {
+      "epoch": 1.2736418511066399,
+      "grad_norm": 0.37146908044815063,
+      "learning_rate": 7.120419218095614e-06,
+      "loss": 0.3602,
+      "step": 5064
+    },
+    {
+      "epoch": 1.2738933601609657,
+      "grad_norm": 0.32229381799697876,
+      "learning_rate": 7.119093976709785e-06,
+      "loss": 0.3616,
+      "step": 5065
+    },
+    {
+      "epoch": 1.2741448692152917,
+      "grad_norm": 0.3241557776927948,
+      "learning_rate": 7.117768553836816e-06,
+      "loss": 0.3269,
+      "step": 5066
+    },
+    {
+      "epoch": 1.2743963782696177,
+      "grad_norm": 0.33732870221138,
+      "learning_rate": 7.1164429495902185e-06,
+      "loss": 0.3396,
+      "step": 5067
+    },
+    {
+      "epoch": 1.2746478873239437,
+      "grad_norm": 0.34197482466697693,
+      "learning_rate": 7.115117164083522e-06,
+      "loss": 0.3481,
+      "step": 5068
+    },
+    {
+      "epoch": 1.2748993963782695,
+      "grad_norm": 0.32349589467048645,
+      "learning_rate": 7.113791197430275e-06,
+      "loss": 0.3505,
+      "step": 5069
+    },
+    {
+      "epoch": 1.2751509054325956,
+      "grad_norm": 0.31883949041366577,
+      "learning_rate": 7.112465049744033e-06,
+      "loss": 0.305,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2754024144869216,
+      "grad_norm": 0.3327849507331848,
+      "learning_rate": 7.111138721138376e-06,
+      "loss": 0.3398,
+      "step": 5071
+    },
+    {
+      "epoch": 1.2756539235412474,
+      "grad_norm": 0.3398069143295288,
+      "learning_rate": 7.109812211726895e-06,
+      "loss": 0.3686,
+      "step": 5072
+    },
+    {
+      "epoch": 1.2759054325955734,
+      "grad_norm": 0.3131691515445709,
+      "learning_rate": 7.108485521623196e-06,
+      "loss": 0.3489,
+      "step": 5073
+    },
+    {
+      "epoch": 1.2761569416498995,
+      "grad_norm": 0.3182958662509918,
+      "learning_rate": 7.107158650940904e-06,
+      "loss": 0.349,
+      "step": 5074
+    },
+    {
+      "epoch": 1.2764084507042255,
+      "grad_norm": 0.2990667521953583,
+      "learning_rate": 7.105831599793655e-06,
+      "loss": 0.3483,
+      "step": 5075
+    },
+    {
+      "epoch": 1.2766599597585513,
+      "grad_norm": 0.3262629508972168,
+      "learning_rate": 7.104504368295105e-06,
+      "loss": 0.364,
+      "step": 5076
+    },
+    {
+      "epoch": 1.2769114688128773,
+      "grad_norm": 0.3283725082874298,
+      "learning_rate": 7.10317695655892e-06,
+      "loss": 0.356,
+      "step": 5077
+    },
+    {
+      "epoch": 1.2771629778672033,
+      "grad_norm": 0.31310954689979553,
+      "learning_rate": 7.101849364698786e-06,
+      "loss": 0.3504,
+      "step": 5078
+    },
+    {
+      "epoch": 1.2774144869215291,
+      "grad_norm": 0.3293515145778656,
+      "learning_rate": 7.100521592828405e-06,
+      "loss": 0.3362,
+      "step": 5079
+    },
+    {
+      "epoch": 1.2776659959758552,
+      "grad_norm": 0.31112322211265564,
+      "learning_rate": 7.0991936410614885e-06,
+      "loss": 0.3425,
+      "step": 5080
+    },
+    {
+      "epoch": 1.2779175050301812,
+      "grad_norm": 0.3442460298538208,
+      "learning_rate": 7.09786550951177e-06,
+      "loss": 0.3326,
+      "step": 5081
+    },
+    {
+      "epoch": 1.278169014084507,
+      "grad_norm": 0.3595294952392578,
+      "learning_rate": 7.096537198292994e-06,
+      "loss": 0.3586,
+      "step": 5082
+    },
+    {
+      "epoch": 1.278420523138833,
+      "grad_norm": 0.359210729598999,
+      "learning_rate": 7.0952087075189235e-06,
+      "loss": 0.382,
+      "step": 5083
+    },
+    {
+      "epoch": 1.278672032193159,
+      "grad_norm": 0.3556901216506958,
+      "learning_rate": 7.0938800373033355e-06,
+      "loss": 0.3755,
+      "step": 5084
+    },
+    {
+      "epoch": 1.2789235412474849,
+      "grad_norm": 0.3710108697414398,
+      "learning_rate": 7.0925511877600195e-06,
+      "loss": 0.3571,
+      "step": 5085
+    },
+    {
+      "epoch": 1.279175050301811,
+      "grad_norm": 0.3306417465209961,
+      "learning_rate": 7.091222159002786e-06,
+      "loss": 0.3409,
+      "step": 5086
+    },
+    {
+      "epoch": 1.279426559356137,
+      "grad_norm": 0.35231444239616394,
+      "learning_rate": 7.0898929511454585e-06,
+      "loss": 0.3652,
+      "step": 5087
+    },
+    {
+      "epoch": 1.2796780684104627,
+      "grad_norm": 0.35659119486808777,
+      "learning_rate": 7.088563564301874e-06,
+      "loss": 0.3515,
+      "step": 5088
+    },
+    {
+      "epoch": 1.2799295774647887,
+      "grad_norm": 0.35471397638320923,
+      "learning_rate": 7.0872339985858855e-06,
+      "loss": 0.3801,
+      "step": 5089
+    },
+    {
+      "epoch": 1.2801810865191148,
+      "grad_norm": 0.3614024221897125,
+      "learning_rate": 7.085904254111362e-06,
+      "loss": 0.3543,
+      "step": 5090
+    },
+    {
+      "epoch": 1.2804325955734406,
+      "grad_norm": 0.3554777503013611,
+      "learning_rate": 7.0845743309921896e-06,
+      "loss": 0.3349,
+      "step": 5091
+    },
+    {
+      "epoch": 1.2806841046277666,
+      "grad_norm": 0.33316969871520996,
+      "learning_rate": 7.083244229342266e-06,
+      "loss": 0.3471,
+      "step": 5092
+    },
+    {
+      "epoch": 1.2809356136820926,
+      "grad_norm": 0.3104395866394043,
+      "learning_rate": 7.081913949275508e-06,
+      "loss": 0.3418,
+      "step": 5093
+    },
+    {
+      "epoch": 1.2811871227364184,
+      "grad_norm": 0.3648764491081238,
+      "learning_rate": 7.080583490905845e-06,
+      "loss": 0.3543,
+      "step": 5094
+    },
+    {
+      "epoch": 1.2814386317907445,
+      "grad_norm": 0.36732056736946106,
+      "learning_rate": 7.07925285434722e-06,
+      "loss": 0.3516,
+      "step": 5095
+    },
+    {
+      "epoch": 1.2816901408450705,
+      "grad_norm": 0.31676825881004333,
+      "learning_rate": 7.077922039713596e-06,
+      "loss": 0.3175,
+      "step": 5096
+    },
+    {
+      "epoch": 1.2819416498993963,
+      "grad_norm": 0.35175764560699463,
+      "learning_rate": 7.07659104711895e-06,
+      "loss": 0.3602,
+      "step": 5097
+    },
+    {
+      "epoch": 1.2821931589537223,
+      "grad_norm": 0.3779788911342621,
+      "learning_rate": 7.075259876677272e-06,
+      "loss": 0.3549,
+      "step": 5098
+    },
+    {
+      "epoch": 1.2824446680080483,
+      "grad_norm": 0.35006478428840637,
+      "learning_rate": 7.073928528502569e-06,
+      "loss": 0.3774,
+      "step": 5099
+    },
+    {
+      "epoch": 1.2826961770623742,
+      "grad_norm": 0.3286049962043762,
+      "learning_rate": 7.07259700270886e-06,
+      "loss": 0.3319,
+      "step": 5100
+    },
+    {
+      "epoch": 1.2829476861167002,
+      "grad_norm": 0.3339432179927826,
+      "learning_rate": 7.071265299410185e-06,
+      "loss": 0.3614,
+      "step": 5101
+    },
+    {
+      "epoch": 1.2831991951710262,
+      "grad_norm": 0.34671711921691895,
+      "learning_rate": 7.069933418720594e-06,
+      "loss": 0.3644,
+      "step": 5102
+    },
+    {
+      "epoch": 1.283450704225352,
+      "grad_norm": 0.33326461911201477,
+      "learning_rate": 7.068601360754157e-06,
+      "loss": 0.3369,
+      "step": 5103
+    },
+    {
+      "epoch": 1.283702213279678,
+      "grad_norm": 0.3220560550689697,
+      "learning_rate": 7.0672691256249526e-06,
+      "loss": 0.3421,
+      "step": 5104
+    },
+    {
+      "epoch": 1.283953722334004,
+      "grad_norm": 0.3266070485115051,
+      "learning_rate": 7.065936713447081e-06,
+      "loss": 0.3523,
+      "step": 5105
+    },
+    {
+      "epoch": 1.2842052313883299,
+      "grad_norm": 0.33557188510894775,
+      "learning_rate": 7.064604124334655e-06,
+      "loss": 0.3568,
+      "step": 5106
+    },
+    {
+      "epoch": 1.284456740442656,
+      "grad_norm": 0.33787620067596436,
+      "learning_rate": 7.063271358401802e-06,
+      "loss": 0.3501,
+      "step": 5107
+    },
+    {
+      "epoch": 1.284708249496982,
+      "grad_norm": 0.31968948245048523,
+      "learning_rate": 7.061938415762664e-06,
+      "loss": 0.3282,
+      "step": 5108
+    },
+    {
+      "epoch": 1.2849597585513077,
+      "grad_norm": 0.309199720621109,
+      "learning_rate": 7.060605296531401e-06,
+      "loss": 0.3448,
+      "step": 5109
+    },
+    {
+      "epoch": 1.2852112676056338,
+      "grad_norm": 0.31305763125419617,
+      "learning_rate": 7.059272000822185e-06,
+      "loss": 0.3649,
+      "step": 5110
+    },
+    {
+      "epoch": 1.2854627766599598,
+      "grad_norm": 0.3350978195667267,
+      "learning_rate": 7.057938528749204e-06,
+      "loss": 0.3349,
+      "step": 5111
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.3685004413127899,
+      "learning_rate": 7.056604880426664e-06,
+      "loss": 0.3658,
+      "step": 5112
+    },
+    {
+      "epoch": 1.2859657947686116,
+      "grad_norm": 0.3085992634296417,
+      "learning_rate": 7.055271055968782e-06,
+      "loss": 0.3426,
+      "step": 5113
+    },
+    {
+      "epoch": 1.2862173038229376,
+      "grad_norm": 0.4011576175689697,
+      "learning_rate": 7.05393705548979e-06,
+      "loss": 0.3575,
+      "step": 5114
+    },
+    {
+      "epoch": 1.2864688128772634,
+      "grad_norm": 0.3109534680843353,
+      "learning_rate": 7.052602879103939e-06,
+      "loss": 0.3262,
+      "step": 5115
+    },
+    {
+      "epoch": 1.2867203219315895,
+      "grad_norm": 0.3395739197731018,
+      "learning_rate": 7.051268526925493e-06,
+      "loss": 0.3527,
+      "step": 5116
+    },
+    {
+      "epoch": 1.2869718309859155,
+      "grad_norm": 0.31294944882392883,
+      "learning_rate": 7.049933999068731e-06,
+      "loss": 0.3524,
+      "step": 5117
+    },
+    {
+      "epoch": 1.2872233400402415,
+      "grad_norm": 0.36359307169914246,
+      "learning_rate": 7.0485992956479466e-06,
+      "loss": 0.338,
+      "step": 5118
+    },
+    {
+      "epoch": 1.2874748490945673,
+      "grad_norm": 0.32687315344810486,
+      "learning_rate": 7.0472644167774474e-06,
+      "loss": 0.3405,
+      "step": 5119
+    },
+    {
+      "epoch": 1.2877263581488934,
+      "grad_norm": 0.28143706917762756,
+      "learning_rate": 7.045929362571559e-06,
+      "loss": 0.3903,
+      "step": 5120
+    },
+    {
+      "epoch": 1.2879778672032194,
+      "grad_norm": 0.32885175943374634,
+      "learning_rate": 7.04459413314462e-06,
+      "loss": 0.3788,
+      "step": 5121
+    },
+    {
+      "epoch": 1.2882293762575452,
+      "grad_norm": 0.31145739555358887,
+      "learning_rate": 7.0432587286109845e-06,
+      "loss": 0.3523,
+      "step": 5122
+    },
+    {
+      "epoch": 1.2884808853118712,
+      "grad_norm": 0.3306049108505249,
+      "learning_rate": 7.041923149085021e-06,
+      "loss": 0.3387,
+      "step": 5123
+    },
+    {
+      "epoch": 1.2887323943661972,
+      "grad_norm": 0.31860941648483276,
+      "learning_rate": 7.040587394681115e-06,
+      "loss": 0.3411,
+      "step": 5124
+    },
+    {
+      "epoch": 1.2889839034205233,
+      "grad_norm": 0.37950772047042847,
+      "learning_rate": 7.039251465513664e-06,
+      "loss": 0.3654,
+      "step": 5125
+    },
+    {
+      "epoch": 1.289235412474849,
+      "grad_norm": 0.3230748474597931,
+      "learning_rate": 7.037915361697082e-06,
+      "loss": 0.372,
+      "step": 5126
+    },
+    {
+      "epoch": 1.289486921529175,
+      "grad_norm": 0.3333132266998291,
+      "learning_rate": 7.036579083345799e-06,
+      "loss": 0.3265,
+      "step": 5127
+    },
+    {
+      "epoch": 1.2897384305835011,
+      "grad_norm": 0.3597933053970337,
+      "learning_rate": 7.035242630574257e-06,
+      "loss": 0.363,
+      "step": 5128
+    },
+    {
+      "epoch": 1.289989939637827,
+      "grad_norm": 0.3329184055328369,
+      "learning_rate": 7.033906003496917e-06,
+      "loss": 0.3719,
+      "step": 5129
+    },
+    {
+      "epoch": 1.290241448692153,
+      "grad_norm": 0.32776138186454773,
+      "learning_rate": 7.0325692022282535e-06,
+      "loss": 0.3565,
+      "step": 5130
+    },
+    {
+      "epoch": 1.290492957746479,
+      "grad_norm": 0.3282029628753662,
+      "learning_rate": 7.031232226882752e-06,
+      "loss": 0.345,
+      "step": 5131
+    },
+    {
+      "epoch": 1.2907444668008048,
+      "grad_norm": 0.36610522866249084,
+      "learning_rate": 7.029895077574918e-06,
+      "loss": 0.331,
+      "step": 5132
+    },
+    {
+      "epoch": 1.2909959758551308,
+      "grad_norm": 0.3236190974712372,
+      "learning_rate": 7.028557754419269e-06,
+      "loss": 0.3581,
+      "step": 5133
+    },
+    {
+      "epoch": 1.2912474849094568,
+      "grad_norm": 0.34408023953437805,
+      "learning_rate": 7.0272202575303395e-06,
+      "loss": 0.3365,
+      "step": 5134
+    },
+    {
+      "epoch": 1.2914989939637826,
+      "grad_norm": 0.32077619433403015,
+      "learning_rate": 7.025882587022676e-06,
+      "loss": 0.3443,
+      "step": 5135
+    },
+    {
+      "epoch": 1.2917505030181087,
+      "grad_norm": 0.3198366165161133,
+      "learning_rate": 7.024544743010845e-06,
+      "loss": 0.348,
+      "step": 5136
+    },
+    {
+      "epoch": 1.2920020120724347,
+      "grad_norm": 0.33954963088035583,
+      "learning_rate": 7.023206725609421e-06,
+      "loss": 0.3489,
+      "step": 5137
+    },
+    {
+      "epoch": 1.2922535211267605,
+      "grad_norm": 0.3308905065059662,
+      "learning_rate": 7.021868534932998e-06,
+      "loss": 0.367,
+      "step": 5138
+    },
+    {
+      "epoch": 1.2925050301810865,
+      "grad_norm": 0.3465730547904968,
+      "learning_rate": 7.0205301710961825e-06,
+      "loss": 0.3596,
+      "step": 5139
+    },
+    {
+      "epoch": 1.2927565392354126,
+      "grad_norm": 0.3291175067424774,
+      "learning_rate": 7.0191916342136e-06,
+      "loss": 0.3774,
+      "step": 5140
+    },
+    {
+      "epoch": 1.2930080482897384,
+      "grad_norm": 0.3781532645225525,
+      "learning_rate": 7.017852924399886e-06,
+      "loss": 0.3358,
+      "step": 5141
+    },
+    {
+      "epoch": 1.2932595573440644,
+      "grad_norm": 0.3416769504547119,
+      "learning_rate": 7.016514041769693e-06,
+      "loss": 0.34,
+      "step": 5142
+    },
+    {
+      "epoch": 1.2935110663983904,
+      "grad_norm": 0.31418657302856445,
+      "learning_rate": 7.015174986437687e-06,
+      "loss": 0.3507,
+      "step": 5143
+    },
+    {
+      "epoch": 1.2937625754527162,
+      "grad_norm": 0.394488662481308,
+      "learning_rate": 7.01383575851855e-06,
+      "loss": 0.3669,
+      "step": 5144
+    },
+    {
+      "epoch": 1.2940140845070423,
+      "grad_norm": 0.3305054008960724,
+      "learning_rate": 7.012496358126979e-06,
+      "loss": 0.37,
+      "step": 5145
+    },
+    {
+      "epoch": 1.2942655935613683,
+      "grad_norm": 0.33333835005760193,
+      "learning_rate": 7.011156785377686e-06,
+      "loss": 0.3599,
+      "step": 5146
+    },
+    {
+      "epoch": 1.294517102615694,
+      "grad_norm": 0.32395216822624207,
+      "learning_rate": 7.009817040385395e-06,
+      "loss": 0.3402,
+      "step": 5147
+    },
+    {
+      "epoch": 1.29476861167002,
+      "grad_norm": 0.35054081678390503,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.3328,
+      "step": 5148
+    },
+    {
+      "epoch": 1.2950201207243461,
+      "grad_norm": 0.32759636640548706,
+      "learning_rate": 7.007137034130801e-06,
+      "loss": 0.3594,
+      "step": 5149
+    },
+    {
+      "epoch": 1.295271629778672,
+      "grad_norm": 0.3191884160041809,
+      "learning_rate": 7.005796773098023e-06,
+      "loss": 0.3493,
+      "step": 5150
+    },
+    {
+      "epoch": 1.295523138832998,
+      "grad_norm": 0.33479100465774536,
+      "learning_rate": 7.0044563402813e-06,
+      "loss": 0.3783,
+      "step": 5151
+    },
+    {
+      "epoch": 1.295774647887324,
+      "grad_norm": 0.3412805199623108,
+      "learning_rate": 7.003115735795431e-06,
+      "loss": 0.3371,
+      "step": 5152
+    },
+    {
+      "epoch": 1.2960261569416498,
+      "grad_norm": 0.3714444041252136,
+      "learning_rate": 7.00177495975523e-06,
+      "loss": 0.3369,
+      "step": 5153
+    },
+    {
+      "epoch": 1.2962776659959758,
+      "grad_norm": 0.33331310749053955,
+      "learning_rate": 7.000434012275528e-06,
+      "loss": 0.3524,
+      "step": 5154
+    },
+    {
+      "epoch": 1.2965291750503019,
+      "grad_norm": 0.36643174290657043,
+      "learning_rate": 6.999092893471166e-06,
+      "loss": 0.3589,
+      "step": 5155
+    },
+    {
+      "epoch": 1.2967806841046277,
+      "grad_norm": 0.35542479157447815,
+      "learning_rate": 6.997751603457006e-06,
+      "loss": 0.3364,
+      "step": 5156
+    },
+    {
+      "epoch": 1.2970321931589537,
+      "grad_norm": 0.3719502389431,
+      "learning_rate": 6.996410142347918e-06,
+      "loss": 0.3643,
+      "step": 5157
+    },
+    {
+      "epoch": 1.2972837022132797,
+      "grad_norm": 0.3411712348461151,
+      "learning_rate": 6.995068510258791e-06,
+      "loss": 0.3813,
+      "step": 5158
+    },
+    {
+      "epoch": 1.2975352112676055,
+      "grad_norm": 0.3225451707839966,
+      "learning_rate": 6.993726707304527e-06,
+      "loss": 0.3343,
+      "step": 5159
+    },
+    {
+      "epoch": 1.2977867203219315,
+      "grad_norm": 0.3910596966743469,
+      "learning_rate": 6.992384733600044e-06,
+      "loss": 0.3578,
+      "step": 5160
+    },
+    {
+      "epoch": 1.2980382293762576,
+      "grad_norm": 0.3352847993373871,
+      "learning_rate": 6.991042589260271e-06,
+      "loss": 0.3369,
+      "step": 5161
+    },
+    {
+      "epoch": 1.2982897384305834,
+      "grad_norm": 0.3266632854938507,
+      "learning_rate": 6.989700274400157e-06,
+      "loss": 0.3751,
+      "step": 5162
+    },
+    {
+      "epoch": 1.2985412474849094,
+      "grad_norm": 0.36302119493484497,
+      "learning_rate": 6.988357789134662e-06,
+      "loss": 0.3549,
+      "step": 5163
+    },
+    {
+      "epoch": 1.2987927565392354,
+      "grad_norm": 0.3544050455093384,
+      "learning_rate": 6.987015133578763e-06,
+      "loss": 0.3655,
+      "step": 5164
+    },
+    {
+      "epoch": 1.2990442655935612,
+      "grad_norm": 0.3703245520591736,
+      "learning_rate": 6.985672307847447e-06,
+      "loss": 0.3511,
+      "step": 5165
+    },
+    {
+      "epoch": 1.2992957746478873,
+      "grad_norm": 0.3711421489715576,
+      "learning_rate": 6.98432931205572e-06,
+      "loss": 0.3394,
+      "step": 5166
+    },
+    {
+      "epoch": 1.2995472837022133,
+      "grad_norm": 0.37200912833213806,
+      "learning_rate": 6.982986146318602e-06,
+      "loss": 0.3642,
+      "step": 5167
+    },
+    {
+      "epoch": 1.2997987927565393,
+      "grad_norm": 0.35503825545310974,
+      "learning_rate": 6.981642810751126e-06,
+      "loss": 0.3262,
+      "step": 5168
+    },
+    {
+      "epoch": 1.3000503018108651,
+      "grad_norm": 0.35312941670417786,
+      "learning_rate": 6.980299305468341e-06,
+      "loss": 0.3773,
+      "step": 5169
+    },
+    {
+      "epoch": 1.3003018108651911,
+      "grad_norm": 0.35091546177864075,
+      "learning_rate": 6.978955630585309e-06,
+      "loss": 0.3409,
+      "step": 5170
+    },
+    {
+      "epoch": 1.3005533199195172,
+      "grad_norm": 0.30944791436195374,
+      "learning_rate": 6.9776117862171065e-06,
+      "loss": 0.3384,
+      "step": 5171
+    },
+    {
+      "epoch": 1.3008048289738432,
+      "grad_norm": 0.36516040563583374,
+      "learning_rate": 6.976267772478828e-06,
+      "loss": 0.3533,
+      "step": 5172
+    },
+    {
+      "epoch": 1.301056338028169,
+      "grad_norm": 0.34464800357818604,
+      "learning_rate": 6.974923589485577e-06,
+      "loss": 0.3366,
+      "step": 5173
+    },
+    {
+      "epoch": 1.301307847082495,
+      "grad_norm": 0.3472990095615387,
+      "learning_rate": 6.973579237352475e-06,
+      "loss": 0.3472,
+      "step": 5174
+    },
+    {
+      "epoch": 1.301559356136821,
+      "grad_norm": 0.33396056294441223,
+      "learning_rate": 6.97223471619466e-06,
+      "loss": 0.3405,
+      "step": 5175
+    },
+    {
+      "epoch": 1.3018108651911469,
+      "grad_norm": 0.31734970211982727,
+      "learning_rate": 6.970890026127278e-06,
+      "loss": 0.34,
+      "step": 5176
+    },
+    {
+      "epoch": 1.3020623742454729,
+      "grad_norm": 0.34466373920440674,
+      "learning_rate": 6.9695451672654965e-06,
+      "loss": 0.3438,
+      "step": 5177
+    },
+    {
+      "epoch": 1.302313883299799,
+      "grad_norm": 0.3734425902366638,
+      "learning_rate": 6.968200139724492e-06,
+      "loss": 0.3757,
+      "step": 5178
+    },
+    {
+      "epoch": 1.3025653923541247,
+      "grad_norm": 0.3214707374572754,
+      "learning_rate": 6.966854943619459e-06,
+      "loss": 0.3634,
+      "step": 5179
+    },
+    {
+      "epoch": 1.3028169014084507,
+      "grad_norm": 0.32335829734802246,
+      "learning_rate": 6.965509579065605e-06,
+      "loss": 0.3365,
+      "step": 5180
+    },
+    {
+      "epoch": 1.3030684104627768,
+      "grad_norm": 0.3490079641342163,
+      "learning_rate": 6.964164046178151e-06,
+      "loss": 0.3431,
+      "step": 5181
+    },
+    {
+      "epoch": 1.3033199195171026,
+      "grad_norm": 0.3411654531955719,
+      "learning_rate": 6.962818345072333e-06,
+      "loss": 0.365,
+      "step": 5182
+    },
+    {
+      "epoch": 1.3035714285714286,
+      "grad_norm": 0.3305014669895172,
+      "learning_rate": 6.961472475863406e-06,
+      "loss": 0.342,
+      "step": 5183
+    },
+    {
+      "epoch": 1.3038229376257546,
+      "grad_norm": 0.33585986495018005,
+      "learning_rate": 6.96012643866663e-06,
+      "loss": 0.336,
+      "step": 5184
+    },
+    {
+      "epoch": 1.3040744466800804,
+      "grad_norm": 0.3303409516811371,
+      "learning_rate": 6.958780233597289e-06,
+      "loss": 0.3481,
+      "step": 5185
+    },
+    {
+      "epoch": 1.3043259557344065,
+      "grad_norm": 0.362411230802536,
+      "learning_rate": 6.957433860770674e-06,
+      "loss": 0.3593,
+      "step": 5186
+    },
+    {
+      "epoch": 1.3045774647887325,
+      "grad_norm": 0.3479267656803131,
+      "learning_rate": 6.956087320302094e-06,
+      "loss": 0.3674,
+      "step": 5187
+    },
+    {
+      "epoch": 1.3048289738430583,
+      "grad_norm": 0.3524285554885864,
+      "learning_rate": 6.9547406123068724e-06,
+      "loss": 0.3559,
+      "step": 5188
+    },
+    {
+      "epoch": 1.3050804828973843,
+      "grad_norm": 0.3306998014450073,
+      "learning_rate": 6.953393736900346e-06,
+      "loss": 0.3547,
+      "step": 5189
+    },
+    {
+      "epoch": 1.3053319919517103,
+      "grad_norm": 0.35222721099853516,
+      "learning_rate": 6.9520466941978685e-06,
+      "loss": 0.3576,
+      "step": 5190
+    },
+    {
+      "epoch": 1.3055835010060362,
+      "grad_norm": 0.30763161182403564,
+      "learning_rate": 6.950699484314802e-06,
+      "loss": 0.359,
+      "step": 5191
+    },
+    {
+      "epoch": 1.3058350100603622,
+      "grad_norm": 0.3283224105834961,
+      "learning_rate": 6.949352107366528e-06,
+      "loss": 0.3612,
+      "step": 5192
+    },
+    {
+      "epoch": 1.3060865191146882,
+      "grad_norm": 0.32542482018470764,
+      "learning_rate": 6.9480045634684405e-06,
+      "loss": 0.3491,
+      "step": 5193
+    },
+    {
+      "epoch": 1.306338028169014,
+      "grad_norm": 0.3252429962158203,
+      "learning_rate": 6.94665685273595e-06,
+      "loss": 0.3556,
+      "step": 5194
+    },
+    {
+      "epoch": 1.30658953722334,
+      "grad_norm": 0.30148446559906006,
+      "learning_rate": 6.945308975284478e-06,
+      "loss": 0.347,
+      "step": 5195
+    },
+    {
+      "epoch": 1.306841046277666,
+      "grad_norm": 0.3339576721191406,
+      "learning_rate": 6.943960931229462e-06,
+      "loss": 0.3376,
+      "step": 5196
+    },
+    {
+      "epoch": 1.3070925553319919,
+      "grad_norm": 0.3315853476524353,
+      "learning_rate": 6.942612720686355e-06,
+      "loss": 0.3464,
+      "step": 5197
+    },
+    {
+      "epoch": 1.307344064386318,
+      "grad_norm": 0.3184394836425781,
+      "learning_rate": 6.9412643437706194e-06,
+      "loss": 0.3549,
+      "step": 5198
+    },
+    {
+      "epoch": 1.307595573440644,
+      "grad_norm": 0.34649431705474854,
+      "learning_rate": 6.939915800597738e-06,
+      "loss": 0.3625,
+      "step": 5199
+    },
+    {
+      "epoch": 1.3078470824949697,
+      "grad_norm": 0.34910690784454346,
+      "learning_rate": 6.938567091283205e-06,
+      "loss": 0.3417,
+      "step": 5200
+    },
+    {
+      "epoch": 1.3080985915492958,
+      "grad_norm": 0.34020349383354187,
+      "learning_rate": 6.937218215942527e-06,
+      "loss": 0.3706,
+      "step": 5201
+    },
+    {
+      "epoch": 1.3083501006036218,
+      "grad_norm": 0.30801448225975037,
+      "learning_rate": 6.935869174691229e-06,
+      "loss": 0.351,
+      "step": 5202
+    },
+    {
+      "epoch": 1.3086016096579476,
+      "grad_norm": 0.3194606900215149,
+      "learning_rate": 6.934519967644847e-06,
+      "loss": 0.3343,
+      "step": 5203
+    },
+    {
+      "epoch": 1.3088531187122736,
+      "grad_norm": 0.3294532001018524,
+      "learning_rate": 6.93317059491893e-06,
+      "loss": 0.3352,
+      "step": 5204
+    },
+    {
+      "epoch": 1.3091046277665996,
+      "grad_norm": 0.3065512180328369,
+      "learning_rate": 6.931821056629048e-06,
+      "loss": 0.362,
+      "step": 5205
+    },
+    {
+      "epoch": 1.3093561368209254,
+      "grad_norm": 0.31400826573371887,
+      "learning_rate": 6.930471352890777e-06,
+      "loss": 0.3369,
+      "step": 5206
+    },
+    {
+      "epoch": 1.3096076458752515,
+      "grad_norm": 0.32463496923446655,
+      "learning_rate": 6.9291214838197114e-06,
+      "loss": 0.3772,
+      "step": 5207
+    },
+    {
+      "epoch": 1.3098591549295775,
+      "grad_norm": 0.3142559230327606,
+      "learning_rate": 6.92777144953146e-06,
+      "loss": 0.3552,
+      "step": 5208
+    },
+    {
+      "epoch": 1.3101106639839033,
+      "grad_norm": 0.33447784185409546,
+      "learning_rate": 6.926421250141644e-06,
+      "loss": 0.372,
+      "step": 5209
+    },
+    {
+      "epoch": 1.3103621730382293,
+      "grad_norm": 0.34302520751953125,
+      "learning_rate": 6.925070885765899e-06,
+      "loss": 0.335,
+      "step": 5210
+    },
+    {
+      "epoch": 1.3106136820925554,
+      "grad_norm": 0.32417216897010803,
+      "learning_rate": 6.923720356519877e-06,
+      "loss": 0.3592,
+      "step": 5211
+    },
+    {
+      "epoch": 1.3108651911468812,
+      "grad_norm": 0.30292844772338867,
+      "learning_rate": 6.922369662519239e-06,
+      "loss": 0.338,
+      "step": 5212
+    },
+    {
+      "epoch": 1.3111167002012072,
+      "grad_norm": 0.32632318139076233,
+      "learning_rate": 6.921018803879667e-06,
+      "loss": 0.3547,
+      "step": 5213
+    },
+    {
+      "epoch": 1.3113682092555332,
+      "grad_norm": 0.3217123746871948,
+      "learning_rate": 6.919667780716852e-06,
+      "loss": 0.3661,
+      "step": 5214
+    },
+    {
+      "epoch": 1.311619718309859,
+      "grad_norm": 0.32875531911849976,
+      "learning_rate": 6.918316593146501e-06,
+      "loss": 0.3559,
+      "step": 5215
+    },
+    {
+      "epoch": 1.311871227364185,
+      "grad_norm": 0.33865031599998474,
+      "learning_rate": 6.916965241284335e-06,
+      "loss": 0.3564,
+      "step": 5216
+    },
+    {
+      "epoch": 1.312122736418511,
+      "grad_norm": 0.2909534275531769,
+      "learning_rate": 6.9156137252460885e-06,
+      "loss": 0.3394,
+      "step": 5217
+    },
+    {
+      "epoch": 1.312374245472837,
+      "grad_norm": 0.3667255640029907,
+      "learning_rate": 6.91426204514751e-06,
+      "loss": 0.355,
+      "step": 5218
+    },
+    {
+      "epoch": 1.312625754527163,
+      "grad_norm": 0.34232789278030396,
+      "learning_rate": 6.912910201104363e-06,
+      "loss": 0.343,
+      "step": 5219
+    },
+    {
+      "epoch": 1.312877263581489,
+      "grad_norm": 0.33543142676353455,
+      "learning_rate": 6.911558193232426e-06,
+      "loss": 0.3417,
+      "step": 5220
+    },
+    {
+      "epoch": 1.313128772635815,
+      "grad_norm": 0.3160308003425598,
+      "learning_rate": 6.910206021647487e-06,
+      "loss": 0.3491,
+      "step": 5221
+    },
+    {
+      "epoch": 1.313380281690141,
+      "grad_norm": 0.3830724060535431,
+      "learning_rate": 6.908853686465353e-06,
+      "loss": 0.3414,
+      "step": 5222
+    },
+    {
+      "epoch": 1.3136317907444668,
+      "grad_norm": 0.33467820286750793,
+      "learning_rate": 6.907501187801843e-06,
+      "loss": 0.3589,
+      "step": 5223
+    },
+    {
+      "epoch": 1.3138832997987928,
+      "grad_norm": 0.3496091961860657,
+      "learning_rate": 6.906148525772789e-06,
+      "loss": 0.3608,
+      "step": 5224
+    },
+    {
+      "epoch": 1.3141348088531188,
+      "grad_norm": 0.3401433229446411,
+      "learning_rate": 6.904795700494038e-06,
+      "loss": 0.3581,
+      "step": 5225
+    },
+    {
+      "epoch": 1.3143863179074446,
+      "grad_norm": 0.3567175567150116,
+      "learning_rate": 6.9034427120814505e-06,
+      "loss": 0.337,
+      "step": 5226
+    },
+    {
+      "epoch": 1.3146378269617707,
+      "grad_norm": 0.33114707469940186,
+      "learning_rate": 6.902089560650904e-06,
+      "loss": 0.3523,
+      "step": 5227
+    },
+    {
+      "epoch": 1.3148893360160967,
+      "grad_norm": 0.3312157988548279,
+      "learning_rate": 6.900736246318287e-06,
+      "loss": 0.353,
+      "step": 5228
+    },
+    {
+      "epoch": 1.3151408450704225,
+      "grad_norm": 0.3470541536808014,
+      "learning_rate": 6.899382769199501e-06,
+      "loss": 0.358,
+      "step": 5229
+    },
+    {
+      "epoch": 1.3153923541247485,
+      "grad_norm": 0.3177447021007538,
+      "learning_rate": 6.898029129410463e-06,
+      "loss": 0.3718,
+      "step": 5230
+    },
+    {
+      "epoch": 1.3156438631790746,
+      "grad_norm": 0.297389954328537,
+      "learning_rate": 6.896675327067104e-06,
+      "loss": 0.342,
+      "step": 5231
+    },
+    {
+      "epoch": 1.3158953722334004,
+      "grad_norm": 0.33685746788978577,
+      "learning_rate": 6.895321362285369e-06,
+      "loss": 0.3386,
+      "step": 5232
+    },
+    {
+      "epoch": 1.3161468812877264,
+      "grad_norm": 0.3213297724723816,
+      "learning_rate": 6.893967235181216e-06,
+      "loss": 0.3766,
+      "step": 5233
+    },
+    {
+      "epoch": 1.3163983903420524,
+      "grad_norm": 0.2946451008319855,
+      "learning_rate": 6.892612945870618e-06,
+      "loss": 0.3446,
+      "step": 5234
+    },
+    {
+      "epoch": 1.3166498993963782,
+      "grad_norm": 0.3347758650779724,
+      "learning_rate": 6.891258494469561e-06,
+      "loss": 0.3605,
+      "step": 5235
+    },
+    {
+      "epoch": 1.3169014084507042,
+      "grad_norm": 0.30589884519577026,
+      "learning_rate": 6.889903881094047e-06,
+      "loss": 0.3453,
+      "step": 5236
+    },
+    {
+      "epoch": 1.3171529175050303,
+      "grad_norm": 0.3220566511154175,
+      "learning_rate": 6.888549105860088e-06,
+      "loss": 0.3572,
+      "step": 5237
+    },
+    {
+      "epoch": 1.317404426559356,
+      "grad_norm": 0.3071398138999939,
+      "learning_rate": 6.887194168883713e-06,
+      "loss": 0.3602,
+      "step": 5238
+    },
+    {
+      "epoch": 1.317655935613682,
+      "grad_norm": 0.34958013892173767,
+      "learning_rate": 6.885839070280964e-06,
+      "loss": 0.3445,
+      "step": 5239
+    },
+    {
+      "epoch": 1.3179074446680081,
+      "grad_norm": 0.33036911487579346,
+      "learning_rate": 6.884483810167896e-06,
+      "loss": 0.3653,
+      "step": 5240
+    },
+    {
+      "epoch": 1.318158953722334,
+      "grad_norm": 0.3227456212043762,
+      "learning_rate": 6.883128388660578e-06,
+      "loss": 0.3573,
+      "step": 5241
+    },
+    {
+      "epoch": 1.31841046277666,
+      "grad_norm": 0.304959774017334,
+      "learning_rate": 6.881772805875095e-06,
+      "loss": 0.3421,
+      "step": 5242
+    },
+    {
+      "epoch": 1.318661971830986,
+      "grad_norm": 0.3372381627559662,
+      "learning_rate": 6.8804170619275445e-06,
+      "loss": 0.3234,
+      "step": 5243
+    },
+    {
+      "epoch": 1.3189134808853118,
+      "grad_norm": 0.3018898665904999,
+      "learning_rate": 6.879061156934035e-06,
+      "loss": 0.3658,
+      "step": 5244
+    },
+    {
+      "epoch": 1.3191649899396378,
+      "grad_norm": 0.3537410497665405,
+      "learning_rate": 6.877705091010693e-06,
+      "loss": 0.3786,
+      "step": 5245
+    },
+    {
+      "epoch": 1.3194164989939638,
+      "grad_norm": 0.3048062026500702,
+      "learning_rate": 6.8763488642736585e-06,
+      "loss": 0.3467,
+      "step": 5246
+    },
+    {
+      "epoch": 1.3196680080482897,
+      "grad_norm": 0.3114982545375824,
+      "learning_rate": 6.874992476839081e-06,
+      "loss": 0.3364,
+      "step": 5247
+    },
+    {
+      "epoch": 1.3199195171026157,
+      "grad_norm": 0.35760530829429626,
+      "learning_rate": 6.873635928823127e-06,
+      "loss": 0.3648,
+      "step": 5248
+    },
+    {
+      "epoch": 1.3201710261569417,
+      "grad_norm": 0.3398659825325012,
+      "learning_rate": 6.8722792203419775e-06,
+      "loss": 0.3682,
+      "step": 5249
+    },
+    {
+      "epoch": 1.3204225352112675,
+      "grad_norm": 0.29969504475593567,
+      "learning_rate": 6.8709223515118265e-06,
+      "loss": 0.3255,
+      "step": 5250
+    },
+    {
+      "epoch": 1.3206740442655935,
+      "grad_norm": 0.31497398018836975,
+      "learning_rate": 6.8695653224488805e-06,
+      "loss": 0.382,
+      "step": 5251
+    },
+    {
+      "epoch": 1.3209255533199196,
+      "grad_norm": 0.3121738135814667,
+      "learning_rate": 6.868208133269359e-06,
+      "loss": 0.3807,
+      "step": 5252
+    },
+    {
+      "epoch": 1.3211770623742454,
+      "grad_norm": 0.31807759404182434,
+      "learning_rate": 6.8668507840895005e-06,
+      "loss": 0.3508,
+      "step": 5253
+    },
+    {
+      "epoch": 1.3214285714285714,
+      "grad_norm": 0.3177856504917145,
+      "learning_rate": 6.86549327502555e-06,
+      "loss": 0.3825,
+      "step": 5254
+    },
+    {
+      "epoch": 1.3216800804828974,
+      "grad_norm": 0.3255070447921753,
+      "learning_rate": 6.864135606193771e-06,
+      "loss": 0.3538,
+      "step": 5255
+    },
+    {
+      "epoch": 1.3219315895372232,
+      "grad_norm": 0.33017322421073914,
+      "learning_rate": 6.862777777710441e-06,
+      "loss": 0.3318,
+      "step": 5256
+    },
+    {
+      "epoch": 1.3221830985915493,
+      "grad_norm": 0.3061206340789795,
+      "learning_rate": 6.861419789691845e-06,
+      "loss": 0.351,
+      "step": 5257
+    },
+    {
+      "epoch": 1.3224346076458753,
+      "grad_norm": 0.3208281993865967,
+      "learning_rate": 6.860061642254291e-06,
+      "loss": 0.3615,
+      "step": 5258
+    },
+    {
+      "epoch": 1.322686116700201,
+      "grad_norm": 0.31412699818611145,
+      "learning_rate": 6.858703335514093e-06,
+      "loss": 0.3814,
+      "step": 5259
+    },
+    {
+      "epoch": 1.3229376257545271,
+      "grad_norm": 0.3277878165245056,
+      "learning_rate": 6.857344869587583e-06,
+      "loss": 0.3539,
+      "step": 5260
+    },
+    {
+      "epoch": 1.3231891348088531,
+      "grad_norm": 0.34601694345474243,
+      "learning_rate": 6.855986244591104e-06,
+      "loss": 0.3516,
+      "step": 5261
+    },
+    {
+      "epoch": 1.323440643863179,
+      "grad_norm": 0.33867859840393066,
+      "learning_rate": 6.854627460641015e-06,
+      "loss": 0.3606,
+      "step": 5262
+    },
+    {
+      "epoch": 1.323692152917505,
+      "grad_norm": 0.3611641824245453,
+      "learning_rate": 6.8532685178536865e-06,
+      "loss": 0.3437,
+      "step": 5263
+    },
+    {
+      "epoch": 1.323943661971831,
+      "grad_norm": 0.3286331593990326,
+      "learning_rate": 6.851909416345502e-06,
+      "loss": 0.3623,
+      "step": 5264
+    },
+    {
+      "epoch": 1.3241951710261568,
+      "grad_norm": 0.3348291218280792,
+      "learning_rate": 6.850550156232862e-06,
+      "loss": 0.3634,
+      "step": 5265
+    },
+    {
+      "epoch": 1.3244466800804828,
+      "grad_norm": 0.3188924491405487,
+      "learning_rate": 6.849190737632179e-06,
+      "loss": 0.3462,
+      "step": 5266
+    },
+    {
+      "epoch": 1.3246981891348089,
+      "grad_norm": 0.35183659195899963,
+      "learning_rate": 6.847831160659877e-06,
+      "loss": 0.355,
+      "step": 5267
+    },
+    {
+      "epoch": 1.3249496981891349,
+      "grad_norm": 0.2944953739643097,
+      "learning_rate": 6.846471425432397e-06,
+      "loss": 0.3173,
+      "step": 5268
+    },
+    {
+      "epoch": 1.3252012072434607,
+      "grad_norm": 0.34920939803123474,
+      "learning_rate": 6.845111532066189e-06,
+      "loss": 0.3544,
+      "step": 5269
+    },
+    {
+      "epoch": 1.3254527162977867,
+      "grad_norm": 0.3055819571018219,
+      "learning_rate": 6.843751480677723e-06,
+      "loss": 0.3572,
+      "step": 5270
+    },
+    {
+      "epoch": 1.3257042253521127,
+      "grad_norm": 0.33443525433540344,
+      "learning_rate": 6.8423912713834765e-06,
+      "loss": 0.3686,
+      "step": 5271
+    },
+    {
+      "epoch": 1.3259557344064388,
+      "grad_norm": 0.3278621435165405,
+      "learning_rate": 6.841030904299943e-06,
+      "loss": 0.368,
+      "step": 5272
+    },
+    {
+      "epoch": 1.3262072434607646,
+      "grad_norm": 0.3702344000339508,
+      "learning_rate": 6.839670379543632e-06,
+      "loss": 0.3329,
+      "step": 5273
+    },
+    {
+      "epoch": 1.3264587525150906,
+      "grad_norm": 0.31350359320640564,
+      "learning_rate": 6.838309697231061e-06,
+      "loss": 0.3686,
+      "step": 5274
+    },
+    {
+      "epoch": 1.3267102615694166,
+      "grad_norm": 0.3589387834072113,
+      "learning_rate": 6.836948857478764e-06,
+      "loss": 0.3401,
+      "step": 5275
+    },
+    {
+      "epoch": 1.3269617706237424,
+      "grad_norm": 0.37206393480300903,
+      "learning_rate": 6.83558786040329e-06,
+      "loss": 0.3487,
+      "step": 5276
+    },
+    {
+      "epoch": 1.3272132796780685,
+      "grad_norm": 0.3218708336353302,
+      "learning_rate": 6.834226706121198e-06,
+      "loss": 0.3381,
+      "step": 5277
+    },
+    {
+      "epoch": 1.3274647887323945,
+      "grad_norm": 0.32571589946746826,
+      "learning_rate": 6.832865394749065e-06,
+      "loss": 0.349,
+      "step": 5278
+    },
+    {
+      "epoch": 1.3277162977867203,
+      "grad_norm": 0.3747791051864624,
+      "learning_rate": 6.831503926403476e-06,
+      "loss": 0.3726,
+      "step": 5279
+    },
+    {
+      "epoch": 1.3279678068410463,
+      "grad_norm": 0.33507394790649414,
+      "learning_rate": 6.830142301201035e-06,
+      "loss": 0.3544,
+      "step": 5280
+    },
+    {
+      "epoch": 1.3282193158953723,
+      "grad_norm": 0.35225021839141846,
+      "learning_rate": 6.8287805192583534e-06,
+      "loss": 0.3614,
+      "step": 5281
+    },
+    {
+      "epoch": 1.3284708249496981,
+      "grad_norm": 0.3295999765396118,
+      "learning_rate": 6.8274185806920625e-06,
+      "loss": 0.3604,
+      "step": 5282
+    },
+    {
+      "epoch": 1.3287223340040242,
+      "grad_norm": 0.3460395038127899,
+      "learning_rate": 6.826056485618803e-06,
+      "loss": 0.352,
+      "step": 5283
+    },
+    {
+      "epoch": 1.3289738430583502,
+      "grad_norm": 0.3388642370700836,
+      "learning_rate": 6.824694234155228e-06,
+      "loss": 0.3313,
+      "step": 5284
+    },
+    {
+      "epoch": 1.329225352112676,
+      "grad_norm": 0.3405665457248688,
+      "learning_rate": 6.823331826418008e-06,
+      "loss": 0.3362,
+      "step": 5285
+    },
+    {
+      "epoch": 1.329476861167002,
+      "grad_norm": 0.3302387297153473,
+      "learning_rate": 6.8219692625238236e-06,
+      "loss": 0.3406,
+      "step": 5286
+    },
+    {
+      "epoch": 1.329728370221328,
+      "grad_norm": 0.3334062695503235,
+      "learning_rate": 6.8206065425893695e-06,
+      "loss": 0.3437,
+      "step": 5287
+    },
+    {
+      "epoch": 1.3299798792756539,
+      "grad_norm": 0.3606247007846832,
+      "learning_rate": 6.819243666731356e-06,
+      "loss": 0.3348,
+      "step": 5288
+    },
+    {
+      "epoch": 1.33023138832998,
+      "grad_norm": 0.3403606712818146,
+      "learning_rate": 6.817880635066503e-06,
+      "loss": 0.3566,
+      "step": 5289
+    },
+    {
+      "epoch": 1.330482897384306,
+      "grad_norm": 0.29230329394340515,
+      "learning_rate": 6.816517447711546e-06,
+      "loss": 0.3524,
+      "step": 5290
+    },
+    {
+      "epoch": 1.3307344064386317,
+      "grad_norm": 0.35570812225341797,
+      "learning_rate": 6.815154104783233e-06,
+      "loss": 0.3324,
+      "step": 5291
+    },
+    {
+      "epoch": 1.3309859154929577,
+      "grad_norm": 0.3560788333415985,
+      "learning_rate": 6.813790606398327e-06,
+      "loss": 0.3826,
+      "step": 5292
+    },
+    {
+      "epoch": 1.3312374245472838,
+      "grad_norm": 0.34960126876831055,
+      "learning_rate": 6.8124269526736035e-06,
+      "loss": 0.3403,
+      "step": 5293
+    },
+    {
+      "epoch": 1.3314889336016096,
+      "grad_norm": 0.37958434224128723,
+      "learning_rate": 6.811063143725849e-06,
+      "loss": 0.3727,
+      "step": 5294
+    },
+    {
+      "epoch": 1.3317404426559356,
+      "grad_norm": 0.3619947135448456,
+      "learning_rate": 6.809699179671867e-06,
+      "loss": 0.3378,
+      "step": 5295
+    },
+    {
+      "epoch": 1.3319919517102616,
+      "grad_norm": 0.35475045442581177,
+      "learning_rate": 6.808335060628471e-06,
+      "loss": 0.3679,
+      "step": 5296
+    },
+    {
+      "epoch": 1.3322434607645874,
+      "grad_norm": 0.3520188331604004,
+      "learning_rate": 6.806970786712489e-06,
+      "loss": 0.3589,
+      "step": 5297
+    },
+    {
+      "epoch": 1.3324949698189135,
+      "grad_norm": 0.3598145842552185,
+      "learning_rate": 6.8056063580407636e-06,
+      "loss": 0.3664,
+      "step": 5298
+    },
+    {
+      "epoch": 1.3327464788732395,
+      "grad_norm": 0.3405790627002716,
+      "learning_rate": 6.804241774730152e-06,
+      "loss": 0.3671,
+      "step": 5299
+    },
+    {
+      "epoch": 1.3329979879275653,
+      "grad_norm": 0.33218327164649963,
+      "learning_rate": 6.802877036897518e-06,
+      "loss": 0.3589,
+      "step": 5300
+    },
+    {
+      "epoch": 1.3332494969818913,
+      "grad_norm": 0.33771365880966187,
+      "learning_rate": 6.801512144659745e-06,
+      "loss": 0.364,
+      "step": 5301
+    },
+    {
+      "epoch": 1.3335010060362174,
+      "grad_norm": 0.31353524327278137,
+      "learning_rate": 6.800147098133727e-06,
+      "loss": 0.3522,
+      "step": 5302
+    },
+    {
+      "epoch": 1.3337525150905432,
+      "grad_norm": 0.2938469350337982,
+      "learning_rate": 6.798781897436371e-06,
+      "loss": 0.3508,
+      "step": 5303
+    },
+    {
+      "epoch": 1.3340040241448692,
+      "grad_norm": 0.342989057302475,
+      "learning_rate": 6.7974165426845996e-06,
+      "loss": 0.3249,
+      "step": 5304
+    },
+    {
+      "epoch": 1.3342555331991952,
+      "grad_norm": 0.3309261202812195,
+      "learning_rate": 6.796051033995346e-06,
+      "loss": 0.3399,
+      "step": 5305
+    },
+    {
+      "epoch": 1.334507042253521,
+      "grad_norm": 0.3215269446372986,
+      "learning_rate": 6.7946853714855565e-06,
+      "loss": 0.3515,
+      "step": 5306
+    },
+    {
+      "epoch": 1.334758551307847,
+      "grad_norm": 0.33649203181266785,
+      "learning_rate": 6.793319555272192e-06,
+      "loss": 0.3635,
+      "step": 5307
+    },
+    {
+      "epoch": 1.335010060362173,
+      "grad_norm": 0.34016871452331543,
+      "learning_rate": 6.791953585472228e-06,
+      "loss": 0.3454,
+      "step": 5308
+    },
+    {
+      "epoch": 1.3352615694164989,
+      "grad_norm": 0.3312501609325409,
+      "learning_rate": 6.790587462202649e-06,
+      "loss": 0.3533,
+      "step": 5309
+    },
+    {
+      "epoch": 1.335513078470825,
+      "grad_norm": 0.33119526505470276,
+      "learning_rate": 6.789221185580456e-06,
+      "loss": 0.3511,
+      "step": 5310
+    },
+    {
+      "epoch": 1.335764587525151,
+      "grad_norm": 0.325152188539505,
+      "learning_rate": 6.7878547557226616e-06,
+      "loss": 0.3515,
+      "step": 5311
+    },
+    {
+      "epoch": 1.3360160965794767,
+      "grad_norm": 0.31850430369377136,
+      "learning_rate": 6.786488172746293e-06,
+      "loss": 0.3725,
+      "step": 5312
+    },
+    {
+      "epoch": 1.3362676056338028,
+      "grad_norm": 0.3241901397705078,
+      "learning_rate": 6.785121436768387e-06,
+      "loss": 0.3467,
+      "step": 5313
+    },
+    {
+      "epoch": 1.3365191146881288,
+      "grad_norm": 0.3087508976459503,
+      "learning_rate": 6.783754547905999e-06,
+      "loss": 0.3348,
+      "step": 5314
+    },
+    {
+      "epoch": 1.3367706237424548,
+      "grad_norm": 0.32791846990585327,
+      "learning_rate": 6.782387506276191e-06,
+      "loss": 0.3526,
+      "step": 5315
+    },
+    {
+      "epoch": 1.3370221327967806,
+      "grad_norm": 0.32451775670051575,
+      "learning_rate": 6.781020311996046e-06,
+      "loss": 0.3459,
+      "step": 5316
+    },
+    {
+      "epoch": 1.3372736418511066,
+      "grad_norm": 0.2936570644378662,
+      "learning_rate": 6.7796529651826525e-06,
+      "loss": 0.3415,
+      "step": 5317
+    },
+    {
+      "epoch": 1.3375251509054327,
+      "grad_norm": 0.3310757577419281,
+      "learning_rate": 6.778285465953116e-06,
+      "loss": 0.3567,
+      "step": 5318
+    },
+    {
+      "epoch": 1.3377766599597585,
+      "grad_norm": 0.3124331533908844,
+      "learning_rate": 6.776917814424555e-06,
+      "loss": 0.3571,
+      "step": 5319
+    },
+    {
+      "epoch": 1.3380281690140845,
+      "grad_norm": 0.3068905174732208,
+      "learning_rate": 6.775550010714099e-06,
+      "loss": 0.3553,
+      "step": 5320
+    },
+    {
+      "epoch": 1.3382796780684105,
+      "grad_norm": 0.31664854288101196,
+      "learning_rate": 6.774182054938893e-06,
+      "loss": 0.3275,
+      "step": 5321
+    },
+    {
+      "epoch": 1.3385311871227366,
+      "grad_norm": 0.3378714919090271,
+      "learning_rate": 6.772813947216092e-06,
+      "loss": 0.3547,
+      "step": 5322
+    },
+    {
+      "epoch": 1.3387826961770624,
+      "grad_norm": 0.3408490717411041,
+      "learning_rate": 6.771445687662868e-06,
+      "loss": 0.3551,
+      "step": 5323
+    },
+    {
+      "epoch": 1.3390342052313884,
+      "grad_norm": 0.3183440864086151,
+      "learning_rate": 6.770077276396402e-06,
+      "loss": 0.382,
+      "step": 5324
+    },
+    {
+      "epoch": 1.3392857142857144,
+      "grad_norm": 0.32740750908851624,
+      "learning_rate": 6.7687087135338915e-06,
+      "loss": 0.3654,
+      "step": 5325
+    },
+    {
+      "epoch": 1.3395372233400402,
+      "grad_norm": 0.3713664412498474,
+      "learning_rate": 6.7673399991925445e-06,
+      "loss": 0.3561,
+      "step": 5326
+    },
+    {
+      "epoch": 1.3397887323943662,
+      "grad_norm": 0.3032859265804291,
+      "learning_rate": 6.765971133489584e-06,
+      "loss": 0.3618,
+      "step": 5327
+    },
+    {
+      "epoch": 1.3400402414486923,
+      "grad_norm": 0.36480486392974854,
+      "learning_rate": 6.764602116542243e-06,
+      "loss": 0.3482,
+      "step": 5328
+    },
+    {
+      "epoch": 1.340291750503018,
+      "grad_norm": 0.3249639570713043,
+      "learning_rate": 6.763232948467769e-06,
+      "loss": 0.3664,
+      "step": 5329
+    },
+    {
+      "epoch": 1.340543259557344,
+      "grad_norm": 0.338945209980011,
+      "learning_rate": 6.761863629383425e-06,
+      "loss": 0.3656,
+      "step": 5330
+    },
+    {
+      "epoch": 1.3407947686116701,
+      "grad_norm": 0.31661343574523926,
+      "learning_rate": 6.760494159406483e-06,
+      "loss": 0.3439,
+      "step": 5331
+    },
+    {
+      "epoch": 1.341046277665996,
+      "grad_norm": 0.3573559522628784,
+      "learning_rate": 6.759124538654231e-06,
+      "loss": 0.3649,
+      "step": 5332
+    },
+    {
+      "epoch": 1.341297786720322,
+      "grad_norm": 0.3466895818710327,
+      "learning_rate": 6.757754767243966e-06,
+      "loss": 0.3554,
+      "step": 5333
+    },
+    {
+      "epoch": 1.341549295774648,
+      "grad_norm": 0.32429978251457214,
+      "learning_rate": 6.756384845293002e-06,
+      "loss": 0.3441,
+      "step": 5334
+    },
+    {
+      "epoch": 1.3418008048289738,
+      "grad_norm": 0.3625079393386841,
+      "learning_rate": 6.7550147729186635e-06,
+      "loss": 0.354,
+      "step": 5335
+    },
+    {
+      "epoch": 1.3420523138832998,
+      "grad_norm": 0.33437997102737427,
+      "learning_rate": 6.75364455023829e-06,
+      "loss": 0.3665,
+      "step": 5336
+    },
+    {
+      "epoch": 1.3423038229376258,
+      "grad_norm": 0.32188209891319275,
+      "learning_rate": 6.7522741773692305e-06,
+      "loss": 0.3652,
+      "step": 5337
+    },
+    {
+      "epoch": 1.3425553319919517,
+      "grad_norm": 0.3210011422634125,
+      "learning_rate": 6.75090365442885e-06,
+      "loss": 0.3763,
+      "step": 5338
+    },
+    {
+      "epoch": 1.3428068410462777,
+      "grad_norm": 0.32578200101852417,
+      "learning_rate": 6.749532981534526e-06,
+      "loss": 0.354,
+      "step": 5339
+    },
+    {
+      "epoch": 1.3430583501006037,
+      "grad_norm": 0.3230711817741394,
+      "learning_rate": 6.748162158803646e-06,
+      "loss": 0.3456,
+      "step": 5340
+    },
+    {
+      "epoch": 1.3433098591549295,
+      "grad_norm": 0.3201490342617035,
+      "learning_rate": 6.746791186353614e-06,
+      "loss": 0.3599,
+      "step": 5341
+    },
+    {
+      "epoch": 1.3435613682092555,
+      "grad_norm": 0.3256816267967224,
+      "learning_rate": 6.745420064301845e-06,
+      "loss": 0.3373,
+      "step": 5342
+    },
+    {
+      "epoch": 1.3438128772635816,
+      "grad_norm": 0.3285231590270996,
+      "learning_rate": 6.744048792765767e-06,
+      "loss": 0.3481,
+      "step": 5343
+    },
+    {
+      "epoch": 1.3440643863179074,
+      "grad_norm": 0.3109275698661804,
+      "learning_rate": 6.74267737186282e-06,
+      "loss": 0.3511,
+      "step": 5344
+    },
+    {
+      "epoch": 1.3443158953722334,
+      "grad_norm": 0.32050541043281555,
+      "learning_rate": 6.7413058017104585e-06,
+      "loss": 0.3641,
+      "step": 5345
+    },
+    {
+      "epoch": 1.3445674044265594,
+      "grad_norm": 0.3238290846347809,
+      "learning_rate": 6.739934082426149e-06,
+      "loss": 0.365,
+      "step": 5346
+    },
+    {
+      "epoch": 1.3448189134808852,
+      "grad_norm": 0.3411909341812134,
+      "learning_rate": 6.73856221412737e-06,
+      "loss": 0.3451,
+      "step": 5347
+    },
+    {
+      "epoch": 1.3450704225352113,
+      "grad_norm": 0.33590638637542725,
+      "learning_rate": 6.737190196931614e-06,
+      "loss": 0.3539,
+      "step": 5348
+    },
+    {
+      "epoch": 1.3453219315895373,
+      "grad_norm": 0.31567126512527466,
+      "learning_rate": 6.735818030956386e-06,
+      "loss": 0.3732,
+      "step": 5349
+    },
+    {
+      "epoch": 1.345573440643863,
+      "grad_norm": 0.3316844701766968,
+      "learning_rate": 6.734445716319202e-06,
+      "loss": 0.35,
+      "step": 5350
+    },
+    {
+      "epoch": 1.345824949698189,
+      "grad_norm": 0.2987160384654999,
+      "learning_rate": 6.733073253137593e-06,
+      "loss": 0.3509,
+      "step": 5351
+    },
+    {
+      "epoch": 1.3460764587525151,
+      "grad_norm": 0.3660013675689697,
+      "learning_rate": 6.731700641529103e-06,
+      "loss": 0.3718,
+      "step": 5352
+    },
+    {
+      "epoch": 1.346327967806841,
+      "grad_norm": 0.3536967635154724,
+      "learning_rate": 6.730327881611286e-06,
+      "loss": 0.37,
+      "step": 5353
+    },
+    {
+      "epoch": 1.346579476861167,
+      "grad_norm": 0.34046685695648193,
+      "learning_rate": 6.72895497350171e-06,
+      "loss": 0.3546,
+      "step": 5354
+    },
+    {
+      "epoch": 1.346830985915493,
+      "grad_norm": 0.3331921100616455,
+      "learning_rate": 6.727581917317958e-06,
+      "loss": 0.3619,
+      "step": 5355
+    },
+    {
+      "epoch": 1.3470824949698188,
+      "grad_norm": 0.3279572129249573,
+      "learning_rate": 6.726208713177622e-06,
+      "loss": 0.3664,
+      "step": 5356
+    },
+    {
+      "epoch": 1.3473340040241448,
+      "grad_norm": 0.3194010257720947,
+      "learning_rate": 6.7248353611983084e-06,
+      "loss": 0.3375,
+      "step": 5357
+    },
+    {
+      "epoch": 1.3475855130784709,
+      "grad_norm": 0.28898268938064575,
+      "learning_rate": 6.7234618614976375e-06,
+      "loss": 0.3384,
+      "step": 5358
+    },
+    {
+      "epoch": 1.3478370221327967,
+      "grad_norm": 0.33326342701911926,
+      "learning_rate": 6.72208821419324e-06,
+      "loss": 0.3417,
+      "step": 5359
+    },
+    {
+      "epoch": 1.3480885311871227,
+      "grad_norm": 0.36949121952056885,
+      "learning_rate": 6.7207144194027605e-06,
+      "loss": 0.379,
+      "step": 5360
+    },
+    {
+      "epoch": 1.3483400402414487,
+      "grad_norm": 0.3311922550201416,
+      "learning_rate": 6.719340477243854e-06,
+      "loss": 0.3505,
+      "step": 5361
+    },
+    {
+      "epoch": 1.3485915492957745,
+      "grad_norm": 0.30410414934158325,
+      "learning_rate": 6.717966387834194e-06,
+      "loss": 0.3436,
+      "step": 5362
+    },
+    {
+      "epoch": 1.3488430583501005,
+      "grad_norm": 0.3302040696144104,
+      "learning_rate": 6.716592151291459e-06,
+      "loss": 0.3396,
+      "step": 5363
+    },
+    {
+      "epoch": 1.3490945674044266,
+      "grad_norm": 0.3249216377735138,
+      "learning_rate": 6.715217767733346e-06,
+      "loss": 0.3363,
+      "step": 5364
+    },
+    {
+      "epoch": 1.3493460764587526,
+      "grad_norm": 0.30946671962738037,
+      "learning_rate": 6.713843237277562e-06,
+      "loss": 0.3726,
+      "step": 5365
+    },
+    {
+      "epoch": 1.3495975855130784,
+      "grad_norm": 0.3507625162601471,
+      "learning_rate": 6.712468560041825e-06,
+      "loss": 0.3618,
+      "step": 5366
+    },
+    {
+      "epoch": 1.3498490945674044,
+      "grad_norm": 0.3141408860683441,
+      "learning_rate": 6.711093736143869e-06,
+      "loss": 0.3566,
+      "step": 5367
+    },
+    {
+      "epoch": 1.3501006036217305,
+      "grad_norm": 0.31941407918930054,
+      "learning_rate": 6.7097187657014395e-06,
+      "loss": 0.3568,
+      "step": 5368
+    },
+    {
+      "epoch": 1.3503521126760563,
+      "grad_norm": 0.29959455132484436,
+      "learning_rate": 6.708343648832294e-06,
+      "loss": 0.3355,
+      "step": 5369
+    },
+    {
+      "epoch": 1.3506036217303823,
+      "grad_norm": 0.32126718759536743,
+      "learning_rate": 6.706968385654202e-06,
+      "loss": 0.3352,
+      "step": 5370
+    },
+    {
+      "epoch": 1.3508551307847083,
+      "grad_norm": 0.31553056836128235,
+      "learning_rate": 6.705592976284948e-06,
+      "loss": 0.3303,
+      "step": 5371
+    },
+    {
+      "epoch": 1.3511066398390343,
+      "grad_norm": 0.3731096088886261,
+      "learning_rate": 6.704217420842325e-06,
+      "loss": 0.3627,
+      "step": 5372
+    },
+    {
+      "epoch": 1.3513581488933601,
+      "grad_norm": 0.3548559248447418,
+      "learning_rate": 6.702841719444141e-06,
+      "loss": 0.3606,
+      "step": 5373
+    },
+    {
+      "epoch": 1.3516096579476862,
+      "grad_norm": 0.3334583342075348,
+      "learning_rate": 6.701465872208216e-06,
+      "loss": 0.3507,
+      "step": 5374
+    },
+    {
+      "epoch": 1.3518611670020122,
+      "grad_norm": 0.36840519309043884,
+      "learning_rate": 6.700089879252385e-06,
+      "loss": 0.3599,
+      "step": 5375
+    },
+    {
+      "epoch": 1.352112676056338,
+      "grad_norm": 0.33070552349090576,
+      "learning_rate": 6.69871374069449e-06,
+      "loss": 0.3544,
+      "step": 5376
+    },
+    {
+      "epoch": 1.352364185110664,
+      "grad_norm": 0.3327641487121582,
+      "learning_rate": 6.6973374566523904e-06,
+      "loss": 0.3508,
+      "step": 5377
+    },
+    {
+      "epoch": 1.35261569416499,
+      "grad_norm": 0.32900556921958923,
+      "learning_rate": 6.695961027243957e-06,
+      "loss": 0.3593,
+      "step": 5378
+    },
+    {
+      "epoch": 1.3528672032193159,
+      "grad_norm": 0.33715447783470154,
+      "learning_rate": 6.694584452587071e-06,
+      "loss": 0.3524,
+      "step": 5379
+    },
+    {
+      "epoch": 1.3531187122736419,
+      "grad_norm": 0.31155914068222046,
+      "learning_rate": 6.693207732799628e-06,
+      "loss": 0.3298,
+      "step": 5380
+    },
+    {
+      "epoch": 1.353370221327968,
+      "grad_norm": 0.3378448188304901,
+      "learning_rate": 6.691830867999536e-06,
+      "loss": 0.3213,
+      "step": 5381
+    },
+    {
+      "epoch": 1.3536217303822937,
+      "grad_norm": 0.31682777404785156,
+      "learning_rate": 6.690453858304713e-06,
+      "loss": 0.3306,
+      "step": 5382
+    },
+    {
+      "epoch": 1.3538732394366197,
+      "grad_norm": 0.3367248773574829,
+      "learning_rate": 6.689076703833093e-06,
+      "loss": 0.3662,
+      "step": 5383
+    },
+    {
+      "epoch": 1.3541247484909458,
+      "grad_norm": 0.3396073579788208,
+      "learning_rate": 6.687699404702621e-06,
+      "loss": 0.3576,
+      "step": 5384
+    },
+    {
+      "epoch": 1.3543762575452716,
+      "grad_norm": 0.33652400970458984,
+      "learning_rate": 6.686321961031252e-06,
+      "loss": 0.3369,
+      "step": 5385
+    },
+    {
+      "epoch": 1.3546277665995976,
+      "grad_norm": 0.33180204033851624,
+      "learning_rate": 6.684944372936958e-06,
+      "loss": 0.3447,
+      "step": 5386
+    },
+    {
+      "epoch": 1.3548792756539236,
+      "grad_norm": 0.32716917991638184,
+      "learning_rate": 6.6835666405377185e-06,
+      "loss": 0.352,
+      "step": 5387
+    },
+    {
+      "epoch": 1.3551307847082494,
+      "grad_norm": 0.33084622025489807,
+      "learning_rate": 6.682188763951528e-06,
+      "loss": 0.3252,
+      "step": 5388
+    },
+    {
+      "epoch": 1.3553822937625755,
+      "grad_norm": 0.3281020224094391,
+      "learning_rate": 6.680810743296394e-06,
+      "loss": 0.3865,
+      "step": 5389
+    },
+    {
+      "epoch": 1.3556338028169015,
+      "grad_norm": 0.3415309488773346,
+      "learning_rate": 6.6794325786903346e-06,
+      "loss": 0.3433,
+      "step": 5390
+    },
+    {
+      "epoch": 1.3558853118712273,
+      "grad_norm": 0.31689122319221497,
+      "learning_rate": 6.678054270251383e-06,
+      "loss": 0.3526,
+      "step": 5391
+    },
+    {
+      "epoch": 1.3561368209255533,
+      "grad_norm": 0.33708077669143677,
+      "learning_rate": 6.67667581809758e-06,
+      "loss": 0.326,
+      "step": 5392
+    },
+    {
+      "epoch": 1.3563883299798793,
+      "grad_norm": 0.3383912146091461,
+      "learning_rate": 6.6752972223469825e-06,
+      "loss": 0.3415,
+      "step": 5393
+    },
+    {
+      "epoch": 1.3566398390342052,
+      "grad_norm": 0.32202810049057007,
+      "learning_rate": 6.673918483117659e-06,
+      "loss": 0.3472,
+      "step": 5394
+    },
+    {
+      "epoch": 1.3568913480885312,
+      "grad_norm": 0.31360799074172974,
+      "learning_rate": 6.672539600527688e-06,
+      "loss": 0.3451,
+      "step": 5395
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.33597296476364136,
+      "learning_rate": 6.671160574695164e-06,
+      "loss": 0.3443,
+      "step": 5396
+    },
+    {
+      "epoch": 1.357394366197183,
+      "grad_norm": 0.35125431418418884,
+      "learning_rate": 6.669781405738193e-06,
+      "loss": 0.3528,
+      "step": 5397
+    },
+    {
+      "epoch": 1.357645875251509,
+      "grad_norm": 0.2997105121612549,
+      "learning_rate": 6.66840209377489e-06,
+      "loss": 0.3226,
+      "step": 5398
+    },
+    {
+      "epoch": 1.357897384305835,
+      "grad_norm": 0.3280561566352844,
+      "learning_rate": 6.667022638923384e-06,
+      "loss": 0.3512,
+      "step": 5399
+    },
+    {
+      "epoch": 1.3581488933601609,
+      "grad_norm": 0.3559575080871582,
+      "learning_rate": 6.665643041301818e-06,
+      "loss": 0.3466,
+      "step": 5400
+    },
+    {
+      "epoch": 1.358400402414487,
+      "grad_norm": 0.3210724890232086,
+      "learning_rate": 6.6642633010283464e-06,
+      "loss": 0.3509,
+      "step": 5401
+    },
+    {
+      "epoch": 1.358651911468813,
+      "grad_norm": 0.3236124813556671,
+      "learning_rate": 6.662883418221136e-06,
+      "loss": 0.3407,
+      "step": 5402
+    },
+    {
+      "epoch": 1.3589034205231387,
+      "grad_norm": 0.3242613971233368,
+      "learning_rate": 6.661503392998362e-06,
+      "loss": 0.3438,
+      "step": 5403
+    },
+    {
+      "epoch": 1.3591549295774648,
+      "grad_norm": 0.33517026901245117,
+      "learning_rate": 6.660123225478217e-06,
+      "loss": 0.3578,
+      "step": 5404
+    },
+    {
+      "epoch": 1.3594064386317908,
+      "grad_norm": 0.3171681761741638,
+      "learning_rate": 6.658742915778904e-06,
+      "loss": 0.339,
+      "step": 5405
+    },
+    {
+      "epoch": 1.3596579476861166,
+      "grad_norm": 0.3439024090766907,
+      "learning_rate": 6.6573624640186375e-06,
+      "loss": 0.3572,
+      "step": 5406
+    },
+    {
+      "epoch": 1.3599094567404426,
+      "grad_norm": 0.3413439691066742,
+      "learning_rate": 6.655981870315643e-06,
+      "loss": 0.3571,
+      "step": 5407
+    },
+    {
+      "epoch": 1.3601609657947686,
+      "grad_norm": 0.33437255024909973,
+      "learning_rate": 6.654601134788162e-06,
+      "loss": 0.3457,
+      "step": 5408
+    },
+    {
+      "epoch": 1.3604124748490944,
+      "grad_norm": 0.36181584000587463,
+      "learning_rate": 6.653220257554446e-06,
+      "loss": 0.3669,
+      "step": 5409
+    },
+    {
+      "epoch": 1.3606639839034205,
+      "grad_norm": 0.34758883714675903,
+      "learning_rate": 6.6518392387327545e-06,
+      "loss": 0.3664,
+      "step": 5410
+    },
+    {
+      "epoch": 1.3609154929577465,
+      "grad_norm": 0.3313416540622711,
+      "learning_rate": 6.650458078441368e-06,
+      "loss": 0.3483,
+      "step": 5411
+    },
+    {
+      "epoch": 1.3611670020120723,
+      "grad_norm": 0.3105969727039337,
+      "learning_rate": 6.649076776798573e-06,
+      "loss": 0.3585,
+      "step": 5412
+    },
+    {
+      "epoch": 1.3614185110663983,
+      "grad_norm": 0.3290698230266571,
+      "learning_rate": 6.647695333922668e-06,
+      "loss": 0.3359,
+      "step": 5413
+    },
+    {
+      "epoch": 1.3616700201207244,
+      "grad_norm": 0.3040936589241028,
+      "learning_rate": 6.646313749931965e-06,
+      "loss": 0.3348,
+      "step": 5414
+    },
+    {
+      "epoch": 1.3619215291750504,
+      "grad_norm": 0.3156396746635437,
+      "learning_rate": 6.6449320249447905e-06,
+      "loss": 0.3541,
+      "step": 5415
+    },
+    {
+      "epoch": 1.3621730382293762,
+      "grad_norm": 0.31559011340141296,
+      "learning_rate": 6.6435501590794775e-06,
+      "loss": 0.3304,
+      "step": 5416
+    },
+    {
+      "epoch": 1.3624245472837022,
+      "grad_norm": 0.30537793040275574,
+      "learning_rate": 6.642168152454375e-06,
+      "loss": 0.3714,
+      "step": 5417
+    },
+    {
+      "epoch": 1.3626760563380282,
+      "grad_norm": 0.3393206298351288,
+      "learning_rate": 6.640786005187844e-06,
+      "loss": 0.3662,
+      "step": 5418
+    },
+    {
+      "epoch": 1.362927565392354,
+      "grad_norm": 0.3266585171222687,
+      "learning_rate": 6.639403717398256e-06,
+      "loss": 0.369,
+      "step": 5419
+    },
+    {
+      "epoch": 1.36317907444668,
+      "grad_norm": 0.31395620107650757,
+      "learning_rate": 6.6380212892039954e-06,
+      "loss": 0.3556,
+      "step": 5420
+    },
+    {
+      "epoch": 1.363430583501006,
+      "grad_norm": 0.32085859775543213,
+      "learning_rate": 6.636638720723459e-06,
+      "loss": 0.3398,
+      "step": 5421
+    },
+    {
+      "epoch": 1.3636820925553321,
+      "grad_norm": 0.325742244720459,
+      "learning_rate": 6.635256012075056e-06,
+      "loss": 0.3437,
+      "step": 5422
+    },
+    {
+      "epoch": 1.363933601609658,
+      "grad_norm": 0.32642316818237305,
+      "learning_rate": 6.633873163377206e-06,
+      "loss": 0.3582,
+      "step": 5423
+    },
+    {
+      "epoch": 1.364185110663984,
+      "grad_norm": 0.3359222710132599,
+      "learning_rate": 6.63249017474834e-06,
+      "loss": 0.3304,
+      "step": 5424
+    },
+    {
+      "epoch": 1.36443661971831,
+      "grad_norm": 0.3578890562057495,
+      "learning_rate": 6.631107046306902e-06,
+      "loss": 0.3756,
+      "step": 5425
+    },
+    {
+      "epoch": 1.3646881287726358,
+      "grad_norm": 0.3200285732746124,
+      "learning_rate": 6.629723778171352e-06,
+      "loss": 0.3426,
+      "step": 5426
+    },
+    {
+      "epoch": 1.3649396378269618,
+      "grad_norm": 0.2880136966705322,
+      "learning_rate": 6.628340370460156e-06,
+      "loss": 0.3657,
+      "step": 5427
+    },
+    {
+      "epoch": 1.3651911468812878,
+      "grad_norm": 0.3454881012439728,
+      "learning_rate": 6.626956823291793e-06,
+      "loss": 0.3552,
+      "step": 5428
+    },
+    {
+      "epoch": 1.3654426559356136,
+      "grad_norm": 0.3579876124858856,
+      "learning_rate": 6.625573136784755e-06,
+      "loss": 0.3684,
+      "step": 5429
+    },
+    {
+      "epoch": 1.3656941649899397,
+      "grad_norm": 0.32106253504753113,
+      "learning_rate": 6.62418931105755e-06,
+      "loss": 0.379,
+      "step": 5430
+    },
+    {
+      "epoch": 1.3659456740442657,
+      "grad_norm": 0.34069597721099854,
+      "learning_rate": 6.6228053462286905e-06,
+      "loss": 0.3692,
+      "step": 5431
+    },
+    {
+      "epoch": 1.3661971830985915,
+      "grad_norm": 0.31019026041030884,
+      "learning_rate": 6.621421242416703e-06,
+      "loss": 0.3666,
+      "step": 5432
+    },
+    {
+      "epoch": 1.3664486921529175,
+      "grad_norm": 0.41305261850357056,
+      "learning_rate": 6.6200369997401325e-06,
+      "loss": 0.3689,
+      "step": 5433
+    },
+    {
+      "epoch": 1.3667002012072436,
+      "grad_norm": 0.3238224685192108,
+      "learning_rate": 6.618652618317527e-06,
+      "loss": 0.3469,
+      "step": 5434
+    },
+    {
+      "epoch": 1.3669517102615694,
+      "grad_norm": 0.3078828454017639,
+      "learning_rate": 6.617268098267451e-06,
+      "loss": 0.3354,
+      "step": 5435
+    },
+    {
+      "epoch": 1.3672032193158954,
+      "grad_norm": 0.29392170906066895,
+      "learning_rate": 6.615883439708481e-06,
+      "loss": 0.3254,
+      "step": 5436
+    },
+    {
+      "epoch": 1.3674547283702214,
+      "grad_norm": 0.34196245670318604,
+      "learning_rate": 6.6144986427592014e-06,
+      "loss": 0.3339,
+      "step": 5437
+    },
+    {
+      "epoch": 1.3677062374245472,
+      "grad_norm": 0.31988444924354553,
+      "learning_rate": 6.613113707538214e-06,
+      "loss": 0.3454,
+      "step": 5438
+    },
+    {
+      "epoch": 1.3679577464788732,
+      "grad_norm": 0.3208816349506378,
+      "learning_rate": 6.6117286341641305e-06,
+      "loss": 0.3436,
+      "step": 5439
+    },
+    {
+      "epoch": 1.3682092555331993,
+      "grad_norm": 0.30724069476127625,
+      "learning_rate": 6.610343422755572e-06,
+      "loss": 0.3624,
+      "step": 5440
+    },
+    {
+      "epoch": 1.368460764587525,
+      "grad_norm": 0.34737128019332886,
+      "learning_rate": 6.608958073431173e-06,
+      "loss": 0.3271,
+      "step": 5441
+    },
+    {
+      "epoch": 1.368712273641851,
+      "grad_norm": 0.3552662134170532,
+      "learning_rate": 6.607572586309581e-06,
+      "loss": 0.3809,
+      "step": 5442
+    },
+    {
+      "epoch": 1.3689637826961771,
+      "grad_norm": 0.2979397475719452,
+      "learning_rate": 6.606186961509452e-06,
+      "loss": 0.3306,
+      "step": 5443
+    },
+    {
+      "epoch": 1.369215291750503,
+      "grad_norm": 0.3502536118030548,
+      "learning_rate": 6.6048011991494595e-06,
+      "loss": 0.3469,
+      "step": 5444
+    },
+    {
+      "epoch": 1.369466800804829,
+      "grad_norm": 0.33274149894714355,
+      "learning_rate": 6.603415299348284e-06,
+      "loss": 0.3341,
+      "step": 5445
+    },
+    {
+      "epoch": 1.369718309859155,
+      "grad_norm": 0.3190041482448578,
+      "learning_rate": 6.60202926222462e-06,
+      "loss": 0.3562,
+      "step": 5446
+    },
+    {
+      "epoch": 1.3699698189134808,
+      "grad_norm": 0.31601500511169434,
+      "learning_rate": 6.600643087897171e-06,
+      "loss": 0.356,
+      "step": 5447
+    },
+    {
+      "epoch": 1.3702213279678068,
+      "grad_norm": 0.3234473764896393,
+      "learning_rate": 6.599256776484655e-06,
+      "loss": 0.3756,
+      "step": 5448
+    },
+    {
+      "epoch": 1.3704728370221329,
+      "grad_norm": 0.33312302827835083,
+      "learning_rate": 6.597870328105801e-06,
+      "loss": 0.3666,
+      "step": 5449
+    },
+    {
+      "epoch": 1.3707243460764587,
+      "grad_norm": 0.32716864347457886,
+      "learning_rate": 6.59648374287935e-06,
+      "loss": 0.3525,
+      "step": 5450
+    },
+    {
+      "epoch": 1.3709758551307847,
+      "grad_norm": 0.315767765045166,
+      "learning_rate": 6.595097020924054e-06,
+      "loss": 0.3242,
+      "step": 5451
+    },
+    {
+      "epoch": 1.3712273641851107,
+      "grad_norm": 0.3259473741054535,
+      "learning_rate": 6.593710162358676e-06,
+      "loss": 0.3392,
+      "step": 5452
+    },
+    {
+      "epoch": 1.3714788732394365,
+      "grad_norm": 0.3474818170070648,
+      "learning_rate": 6.592323167301994e-06,
+      "loss": 0.3662,
+      "step": 5453
+    },
+    {
+      "epoch": 1.3717303822937625,
+      "grad_norm": 0.3318646252155304,
+      "learning_rate": 6.590936035872792e-06,
+      "loss": 0.3754,
+      "step": 5454
+    },
+    {
+      "epoch": 1.3719818913480886,
+      "grad_norm": 0.3492967486381531,
+      "learning_rate": 6.589548768189875e-06,
+      "loss": 0.3491,
+      "step": 5455
+    },
+    {
+      "epoch": 1.3722334004024144,
+      "grad_norm": 0.33331018686294556,
+      "learning_rate": 6.588161364372047e-06,
+      "loss": 0.3543,
+      "step": 5456
+    },
+    {
+      "epoch": 1.3724849094567404,
+      "grad_norm": 0.3557076156139374,
+      "learning_rate": 6.586773824538136e-06,
+      "loss": 0.3616,
+      "step": 5457
+    },
+    {
+      "epoch": 1.3727364185110664,
+      "grad_norm": 0.34410160779953003,
+      "learning_rate": 6.585386148806974e-06,
+      "loss": 0.3588,
+      "step": 5458
+    },
+    {
+      "epoch": 1.3729879275653922,
+      "grad_norm": 0.32868391275405884,
+      "learning_rate": 6.5839983372974045e-06,
+      "loss": 0.3557,
+      "step": 5459
+    },
+    {
+      "epoch": 1.3732394366197183,
+      "grad_norm": 0.29760217666625977,
+      "learning_rate": 6.5826103901282875e-06,
+      "loss": 0.3299,
+      "step": 5460
+    },
+    {
+      "epoch": 1.3734909456740443,
+      "grad_norm": 0.33605876564979553,
+      "learning_rate": 6.581222307418492e-06,
+      "loss": 0.3432,
+      "step": 5461
+    },
+    {
+      "epoch": 1.37374245472837,
+      "grad_norm": 0.30532124638557434,
+      "learning_rate": 6.579834089286898e-06,
+      "loss": 0.3592,
+      "step": 5462
+    },
+    {
+      "epoch": 1.3739939637826961,
+      "grad_norm": 0.30395251512527466,
+      "learning_rate": 6.578445735852397e-06,
+      "loss": 0.341,
+      "step": 5463
+    },
+    {
+      "epoch": 1.3742454728370221,
+      "grad_norm": 0.32410097122192383,
+      "learning_rate": 6.5770572472338935e-06,
+      "loss": 0.3545,
+      "step": 5464
+    },
+    {
+      "epoch": 1.3744969818913482,
+      "grad_norm": 0.32119905948638916,
+      "learning_rate": 6.575668623550302e-06,
+      "loss": 0.346,
+      "step": 5465
+    },
+    {
+      "epoch": 1.374748490945674,
+      "grad_norm": 0.30863913893699646,
+      "learning_rate": 6.574279864920552e-06,
+      "loss": 0.363,
+      "step": 5466
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.3214772939682007,
+      "learning_rate": 6.572890971463579e-06,
+      "loss": 0.3546,
+      "step": 5467
+    },
+    {
+      "epoch": 1.375251509054326,
+      "grad_norm": 0.30236709117889404,
+      "learning_rate": 6.571501943298335e-06,
+      "loss": 0.3459,
+      "step": 5468
+    },
+    {
+      "epoch": 1.3755030181086518,
+      "grad_norm": 0.35171201825141907,
+      "learning_rate": 6.57011278054378e-06,
+      "loss": 0.3439,
+      "step": 5469
+    },
+    {
+      "epoch": 1.3757545271629779,
+      "grad_norm": 0.3353353440761566,
+      "learning_rate": 6.568723483318889e-06,
+      "loss": 0.3565,
+      "step": 5470
+    },
+    {
+      "epoch": 1.3760060362173039,
+      "grad_norm": 0.3299080431461334,
+      "learning_rate": 6.567334051742645e-06,
+      "loss": 0.3654,
+      "step": 5471
+    },
+    {
+      "epoch": 1.37625754527163,
+      "grad_norm": 0.3008752167224884,
+      "learning_rate": 6.565944485934046e-06,
+      "loss": 0.3346,
+      "step": 5472
+    },
+    {
+      "epoch": 1.3765090543259557,
+      "grad_norm": 0.33441752195358276,
+      "learning_rate": 6.564554786012096e-06,
+      "loss": 0.3622,
+      "step": 5473
+    },
+    {
+      "epoch": 1.3767605633802817,
+      "grad_norm": 0.326872855424881,
+      "learning_rate": 6.563164952095818e-06,
+      "loss": 0.374,
+      "step": 5474
+    },
+    {
+      "epoch": 1.3770120724346078,
+      "grad_norm": 0.30647537112236023,
+      "learning_rate": 6.561774984304241e-06,
+      "loss": 0.3631,
+      "step": 5475
+    },
+    {
+      "epoch": 1.3772635814889336,
+      "grad_norm": 0.2820424735546112,
+      "learning_rate": 6.560384882756406e-06,
+      "loss": 0.3572,
+      "step": 5476
+    },
+    {
+      "epoch": 1.3775150905432596,
+      "grad_norm": 0.3259304463863373,
+      "learning_rate": 6.558994647571369e-06,
+      "loss": 0.3525,
+      "step": 5477
+    },
+    {
+      "epoch": 1.3777665995975856,
+      "grad_norm": 0.3217025399208069,
+      "learning_rate": 6.557604278868193e-06,
+      "loss": 0.3713,
+      "step": 5478
+    },
+    {
+      "epoch": 1.3780181086519114,
+      "grad_norm": 0.336544394493103,
+      "learning_rate": 6.5562137767659516e-06,
+      "loss": 0.3471,
+      "step": 5479
+    },
+    {
+      "epoch": 1.3782696177062375,
+      "grad_norm": 0.3470653295516968,
+      "learning_rate": 6.554823141383739e-06,
+      "loss": 0.3517,
+      "step": 5480
+    },
+    {
+      "epoch": 1.3785211267605635,
+      "grad_norm": 0.37579667568206787,
+      "learning_rate": 6.553432372840651e-06,
+      "loss": 0.3326,
+      "step": 5481
+    },
+    {
+      "epoch": 1.3787726358148893,
+      "grad_norm": 0.31945011019706726,
+      "learning_rate": 6.552041471255799e-06,
+      "loss": 0.343,
+      "step": 5482
+    },
+    {
+      "epoch": 1.3790241448692153,
+      "grad_norm": 0.3295787274837494,
+      "learning_rate": 6.550650436748304e-06,
+      "loss": 0.3246,
+      "step": 5483
+    },
+    {
+      "epoch": 1.3792756539235413,
+      "grad_norm": 0.392974317073822,
+      "learning_rate": 6.5492592694373e-06,
+      "loss": 0.3482,
+      "step": 5484
+    },
+    {
+      "epoch": 1.3795271629778671,
+      "grad_norm": 0.33396854996681213,
+      "learning_rate": 6.547867969441931e-06,
+      "loss": 0.3562,
+      "step": 5485
+    },
+    {
+      "epoch": 1.3797786720321932,
+      "grad_norm": 0.3102302551269531,
+      "learning_rate": 6.546476536881355e-06,
+      "loss": 0.3471,
+      "step": 5486
+    },
+    {
+      "epoch": 1.3800301810865192,
+      "grad_norm": 0.36257678270339966,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.3628,
+      "step": 5487
+    },
+    {
+      "epoch": 1.380281690140845,
+      "grad_norm": 0.3609892725944519,
+      "learning_rate": 6.543693274541259e-06,
+      "loss": 0.353,
+      "step": 5488
+    },
+    {
+      "epoch": 1.380533199195171,
+      "grad_norm": 0.3334091603755951,
+      "learning_rate": 6.542301445000108e-06,
+      "loss": 0.3528,
+      "step": 5489
+    },
+    {
+      "epoch": 1.380784708249497,
+      "grad_norm": 0.3559829890727997,
+      "learning_rate": 6.540909483370488e-06,
+      "loss": 0.342,
+      "step": 5490
+    },
+    {
+      "epoch": 1.3810362173038229,
+      "grad_norm": 0.35456085205078125,
+      "learning_rate": 6.539517389771609e-06,
+      "loss": 0.3803,
+      "step": 5491
+    },
+    {
+      "epoch": 1.381287726358149,
+      "grad_norm": 0.3310755491256714,
+      "learning_rate": 6.538125164322699e-06,
+      "loss": 0.3475,
+      "step": 5492
+    },
+    {
+      "epoch": 1.381539235412475,
+      "grad_norm": 0.35925358533859253,
+      "learning_rate": 6.536732807142989e-06,
+      "loss": 0.37,
+      "step": 5493
+    },
+    {
+      "epoch": 1.3817907444668007,
+      "grad_norm": 0.3170955777168274,
+      "learning_rate": 6.535340318351729e-06,
+      "loss": 0.3237,
+      "step": 5494
+    },
+    {
+      "epoch": 1.3820422535211268,
+      "grad_norm": 0.31947341561317444,
+      "learning_rate": 6.5339476980681746e-06,
+      "loss": 0.3396,
+      "step": 5495
+    },
+    {
+      "epoch": 1.3822937625754528,
+      "grad_norm": 0.3393300473690033,
+      "learning_rate": 6.532554946411598e-06,
+      "loss": 0.3373,
+      "step": 5496
+    },
+    {
+      "epoch": 1.3825452716297786,
+      "grad_norm": 0.3222068250179291,
+      "learning_rate": 6.531162063501275e-06,
+      "loss": 0.3653,
+      "step": 5497
+    },
+    {
+      "epoch": 1.3827967806841046,
+      "grad_norm": 0.3433172404766083,
+      "learning_rate": 6.529769049456502e-06,
+      "loss": 0.3526,
+      "step": 5498
+    },
+    {
+      "epoch": 1.3830482897384306,
+      "grad_norm": 0.31900420784950256,
+      "learning_rate": 6.52837590439658e-06,
+      "loss": 0.3523,
+      "step": 5499
+    },
+    {
+      "epoch": 1.3832997987927564,
+      "grad_norm": 0.3333498537540436,
+      "learning_rate": 6.5269826284408235e-06,
+      "loss": 0.3579,
+      "step": 5500
+    },
+    {
+      "epoch": 1.3835513078470825,
+      "grad_norm": 0.3352772891521454,
+      "learning_rate": 6.525589221708558e-06,
+      "loss": 0.3495,
+      "step": 5501
+    },
+    {
+      "epoch": 1.3838028169014085,
+      "grad_norm": 0.35834312438964844,
+      "learning_rate": 6.524195684319119e-06,
+      "loss": 0.3236,
+      "step": 5502
+    },
+    {
+      "epoch": 1.3840543259557343,
+      "grad_norm": 0.3257136940956116,
+      "learning_rate": 6.522802016391856e-06,
+      "loss": 0.3437,
+      "step": 5503
+    },
+    {
+      "epoch": 1.3843058350100603,
+      "grad_norm": 0.337438702583313,
+      "learning_rate": 6.521408218046126e-06,
+      "loss": 0.3477,
+      "step": 5504
+    },
+    {
+      "epoch": 1.3845573440643864,
+      "grad_norm": 0.33436325192451477,
+      "learning_rate": 6.5200142894012995e-06,
+      "loss": 0.3485,
+      "step": 5505
+    },
+    {
+      "epoch": 1.3848088531187122,
+      "grad_norm": 0.3181125521659851,
+      "learning_rate": 6.51862023057676e-06,
+      "loss": 0.3766,
+      "step": 5506
+    },
+    {
+      "epoch": 1.3850603621730382,
+      "grad_norm": 0.3215303421020508,
+      "learning_rate": 6.517226041691897e-06,
+      "loss": 0.3572,
+      "step": 5507
+    },
+    {
+      "epoch": 1.3853118712273642,
+      "grad_norm": 0.3357185125350952,
+      "learning_rate": 6.515831722866115e-06,
+      "loss": 0.343,
+      "step": 5508
+    },
+    {
+      "epoch": 1.38556338028169,
+      "grad_norm": 0.31294816732406616,
+      "learning_rate": 6.514437274218829e-06,
+      "loss": 0.3519,
+      "step": 5509
+    },
+    {
+      "epoch": 1.385814889336016,
+      "grad_norm": 0.30923759937286377,
+      "learning_rate": 6.513042695869465e-06,
+      "loss": 0.3427,
+      "step": 5510
+    },
+    {
+      "epoch": 1.386066398390342,
+      "grad_norm": 0.3177286982536316,
+      "learning_rate": 6.51164798793746e-06,
+      "loss": 0.3483,
+      "step": 5511
+    },
+    {
+      "epoch": 1.3863179074446679,
+      "grad_norm": 0.30468153953552246,
+      "learning_rate": 6.510253150542262e-06,
+      "loss": 0.3639,
+      "step": 5512
+    },
+    {
+      "epoch": 1.386569416498994,
+      "grad_norm": 0.33087295293807983,
+      "learning_rate": 6.508858183803328e-06,
+      "loss": 0.352,
+      "step": 5513
+    },
+    {
+      "epoch": 1.38682092555332,
+      "grad_norm": 0.32209211587905884,
+      "learning_rate": 6.507463087840133e-06,
+      "loss": 0.3568,
+      "step": 5514
+    },
+    {
+      "epoch": 1.387072434607646,
+      "grad_norm": 0.3076551854610443,
+      "learning_rate": 6.506067862772153e-06,
+      "loss": 0.3604,
+      "step": 5515
+    },
+    {
+      "epoch": 1.3873239436619718,
+      "grad_norm": 0.3314887285232544,
+      "learning_rate": 6.504672508718882e-06,
+      "loss": 0.3678,
+      "step": 5516
+    },
+    {
+      "epoch": 1.3875754527162978,
+      "grad_norm": 0.31186097860336304,
+      "learning_rate": 6.503277025799825e-06,
+      "loss": 0.3576,
+      "step": 5517
+    },
+    {
+      "epoch": 1.3878269617706238,
+      "grad_norm": 0.29338905215263367,
+      "learning_rate": 6.501881414134495e-06,
+      "loss": 0.3615,
+      "step": 5518
+    },
+    {
+      "epoch": 1.3880784708249496,
+      "grad_norm": 0.3048270046710968,
+      "learning_rate": 6.500485673842417e-06,
+      "loss": 0.3395,
+      "step": 5519
+    },
+    {
+      "epoch": 1.3883299798792756,
+      "grad_norm": 0.30701351165771484,
+      "learning_rate": 6.499089805043129e-06,
+      "loss": 0.3739,
+      "step": 5520
+    },
+    {
+      "epoch": 1.3885814889336017,
+      "grad_norm": 0.3141977787017822,
+      "learning_rate": 6.497693807856177e-06,
+      "loss": 0.3709,
+      "step": 5521
+    },
+    {
+      "epoch": 1.3888329979879277,
+      "grad_norm": 0.3316255509853363,
+      "learning_rate": 6.496297682401121e-06,
+      "loss": 0.3554,
+      "step": 5522
+    },
+    {
+      "epoch": 1.3890845070422535,
+      "grad_norm": 0.3405584990978241,
+      "learning_rate": 6.494901428797526e-06,
+      "loss": 0.3521,
+      "step": 5523
+    },
+    {
+      "epoch": 1.3893360160965795,
+      "grad_norm": 0.34003138542175293,
+      "learning_rate": 6.493505047164978e-06,
+      "loss": 0.3322,
+      "step": 5524
+    },
+    {
+      "epoch": 1.3895875251509056,
+      "grad_norm": 0.32138487696647644,
+      "learning_rate": 6.492108537623067e-06,
+      "loss": 0.3329,
+      "step": 5525
+    },
+    {
+      "epoch": 1.3898390342052314,
+      "grad_norm": 0.328475683927536,
+      "learning_rate": 6.490711900291393e-06,
+      "loss": 0.349,
+      "step": 5526
+    },
+    {
+      "epoch": 1.3900905432595574,
+      "grad_norm": 0.3196926414966583,
+      "learning_rate": 6.489315135289571e-06,
+      "loss": 0.3406,
+      "step": 5527
+    },
+    {
+      "epoch": 1.3903420523138834,
+      "grad_norm": 0.32950088381767273,
+      "learning_rate": 6.487918242737225e-06,
+      "loss": 0.3643,
+      "step": 5528
+    },
+    {
+      "epoch": 1.3905935613682092,
+      "grad_norm": 0.32452213764190674,
+      "learning_rate": 6.4865212227539895e-06,
+      "loss": 0.3614,
+      "step": 5529
+    },
+    {
+      "epoch": 1.3908450704225352,
+      "grad_norm": 0.3157281279563904,
+      "learning_rate": 6.4851240754595104e-06,
+      "loss": 0.3499,
+      "step": 5530
+    },
+    {
+      "epoch": 1.3910965794768613,
+      "grad_norm": 0.3089020252227783,
+      "learning_rate": 6.483726800973447e-06,
+      "loss": 0.351,
+      "step": 5531
+    },
+    {
+      "epoch": 1.391348088531187,
+      "grad_norm": 0.32950559258461,
+      "learning_rate": 6.482329399415463e-06,
+      "loss": 0.3476,
+      "step": 5532
+    },
+    {
+      "epoch": 1.391599597585513,
+      "grad_norm": 0.31183376908302307,
+      "learning_rate": 6.48093187090524e-06,
+      "loss": 0.3315,
+      "step": 5533
+    },
+    {
+      "epoch": 1.3918511066398391,
+      "grad_norm": 0.34056970477104187,
+      "learning_rate": 6.4795342155624685e-06,
+      "loss": 0.3579,
+      "step": 5534
+    },
+    {
+      "epoch": 1.392102615694165,
+      "grad_norm": 0.29875314235687256,
+      "learning_rate": 6.478136433506846e-06,
+      "loss": 0.3608,
+      "step": 5535
+    },
+    {
+      "epoch": 1.392354124748491,
+      "grad_norm": 0.32909777760505676,
+      "learning_rate": 6.4767385248580865e-06,
+      "loss": 0.3757,
+      "step": 5536
+    },
+    {
+      "epoch": 1.392605633802817,
+      "grad_norm": 0.33318114280700684,
+      "learning_rate": 6.4753404897359085e-06,
+      "loss": 0.3523,
+      "step": 5537
+    },
+    {
+      "epoch": 1.3928571428571428,
+      "grad_norm": 0.32883554697036743,
+      "learning_rate": 6.473942328260049e-06,
+      "loss": 0.3536,
+      "step": 5538
+    },
+    {
+      "epoch": 1.3931086519114688,
+      "grad_norm": 0.3212123215198517,
+      "learning_rate": 6.4725440405502495e-06,
+      "loss": 0.3598,
+      "step": 5539
+    },
+    {
+      "epoch": 1.3933601609657948,
+      "grad_norm": 0.3189206123352051,
+      "learning_rate": 6.471145626726265e-06,
+      "loss": 0.3533,
+      "step": 5540
+    },
+    {
+      "epoch": 1.3936116700201207,
+      "grad_norm": 0.3086141347885132,
+      "learning_rate": 6.469747086907862e-06,
+      "loss": 0.3797,
+      "step": 5541
+    },
+    {
+      "epoch": 1.3938631790744467,
+      "grad_norm": 0.32901155948638916,
+      "learning_rate": 6.468348421214814e-06,
+      "loss": 0.3595,
+      "step": 5542
+    },
+    {
+      "epoch": 1.3941146881287727,
+      "grad_norm": 0.33005937933921814,
+      "learning_rate": 6.466949629766911e-06,
+      "loss": 0.3243,
+      "step": 5543
+    },
+    {
+      "epoch": 1.3943661971830985,
+      "grad_norm": 0.3070705831050873,
+      "learning_rate": 6.465550712683949e-06,
+      "loss": 0.3423,
+      "step": 5544
+    },
+    {
+      "epoch": 1.3946177062374245,
+      "grad_norm": 0.32530298829078674,
+      "learning_rate": 6.464151670085738e-06,
+      "loss": 0.3656,
+      "step": 5545
+    },
+    {
+      "epoch": 1.3948692152917506,
+      "grad_norm": 0.31196725368499756,
+      "learning_rate": 6.4627525020920946e-06,
+      "loss": 0.375,
+      "step": 5546
+    },
+    {
+      "epoch": 1.3951207243460764,
+      "grad_norm": 0.32940876483917236,
+      "learning_rate": 6.461353208822851e-06,
+      "loss": 0.3649,
+      "step": 5547
+    },
+    {
+      "epoch": 1.3953722334004024,
+      "grad_norm": 0.31415605545043945,
+      "learning_rate": 6.459953790397847e-06,
+      "loss": 0.3396,
+      "step": 5548
+    },
+    {
+      "epoch": 1.3956237424547284,
+      "grad_norm": 0.34104838967323303,
+      "learning_rate": 6.458554246936934e-06,
+      "loss": 0.3251,
+      "step": 5549
+    },
+    {
+      "epoch": 1.3958752515090542,
+      "grad_norm": 0.33823341131210327,
+      "learning_rate": 6.457154578559975e-06,
+      "loss": 0.3594,
+      "step": 5550
+    },
+    {
+      "epoch": 1.3961267605633803,
+      "grad_norm": 0.34172114729881287,
+      "learning_rate": 6.455754785386843e-06,
+      "loss": 0.3679,
+      "step": 5551
+    },
+    {
+      "epoch": 1.3963782696177063,
+      "grad_norm": 0.3444267213344574,
+      "learning_rate": 6.454354867537418e-06,
+      "loss": 0.366,
+      "step": 5552
+    },
+    {
+      "epoch": 1.396629778672032,
+      "grad_norm": 0.32145577669143677,
+      "learning_rate": 6.452954825131599e-06,
+      "loss": 0.3459,
+      "step": 5553
+    },
+    {
+      "epoch": 1.3968812877263581,
+      "grad_norm": 0.3104569911956787,
+      "learning_rate": 6.4515546582892895e-06,
+      "loss": 0.338,
+      "step": 5554
+    },
+    {
+      "epoch": 1.3971327967806841,
+      "grad_norm": 0.3085978031158447,
+      "learning_rate": 6.450154367130403e-06,
+      "loss": 0.3378,
+      "step": 5555
+    },
+    {
+      "epoch": 1.39738430583501,
+      "grad_norm": 0.33246156573295593,
+      "learning_rate": 6.448753951774869e-06,
+      "loss": 0.34,
+      "step": 5556
+    },
+    {
+      "epoch": 1.397635814889336,
+      "grad_norm": 0.3080597519874573,
+      "learning_rate": 6.447353412342621e-06,
+      "loss": 0.3697,
+      "step": 5557
+    },
+    {
+      "epoch": 1.397887323943662,
+      "grad_norm": 0.3250534236431122,
+      "learning_rate": 6.445952748953607e-06,
+      "loss": 0.3371,
+      "step": 5558
+    },
+    {
+      "epoch": 1.3981388329979878,
+      "grad_norm": 0.3486895263195038,
+      "learning_rate": 6.4445519617277874e-06,
+      "loss": 0.3384,
+      "step": 5559
+    },
+    {
+      "epoch": 1.3983903420523138,
+      "grad_norm": 0.3366147577762604,
+      "learning_rate": 6.443151050785129e-06,
+      "loss": 0.3304,
+      "step": 5560
+    },
+    {
+      "epoch": 1.3986418511066399,
+      "grad_norm": 0.3010328412055969,
+      "learning_rate": 6.4417500162456114e-06,
+      "loss": 0.34,
+      "step": 5561
+    },
+    {
+      "epoch": 1.3988933601609657,
+      "grad_norm": 0.32444706559181213,
+      "learning_rate": 6.440348858229224e-06,
+      "loss": 0.3565,
+      "step": 5562
+    },
+    {
+      "epoch": 1.3991448692152917,
+      "grad_norm": 0.3689126968383789,
+      "learning_rate": 6.4389475768559675e-06,
+      "loss": 0.3513,
+      "step": 5563
+    },
+    {
+      "epoch": 1.3993963782696177,
+      "grad_norm": 0.3049902319908142,
+      "learning_rate": 6.437546172245855e-06,
+      "loss": 0.3215,
+      "step": 5564
+    },
+    {
+      "epoch": 1.3996478873239437,
+      "grad_norm": 0.3056788146495819,
+      "learning_rate": 6.436144644518905e-06,
+      "loss": 0.3364,
+      "step": 5565
+    },
+    {
+      "epoch": 1.3998993963782695,
+      "grad_norm": 0.33238446712493896,
+      "learning_rate": 6.434742993795149e-06,
+      "loss": 0.3486,
+      "step": 5566
+    },
+    {
+      "epoch": 1.4001509054325956,
+      "grad_norm": 0.3054902255535126,
+      "learning_rate": 6.433341220194633e-06,
+      "loss": 0.3234,
+      "step": 5567
+    },
+    {
+      "epoch": 1.4004024144869216,
+      "grad_norm": 0.3176266551017761,
+      "learning_rate": 6.431939323837409e-06,
+      "loss": 0.3447,
+      "step": 5568
+    },
+    {
+      "epoch": 1.4006539235412474,
+      "grad_norm": 0.3124770522117615,
+      "learning_rate": 6.430537304843539e-06,
+      "loss": 0.3541,
+      "step": 5569
+    },
+    {
+      "epoch": 1.4009054325955734,
+      "grad_norm": 0.2936578094959259,
+      "learning_rate": 6.429135163333099e-06,
+      "loss": 0.3353,
+      "step": 5570
+    },
+    {
+      "epoch": 1.4011569416498995,
+      "grad_norm": 0.32754024863243103,
+      "learning_rate": 6.427732899426172e-06,
+      "loss": 0.3473,
+      "step": 5571
+    },
+    {
+      "epoch": 1.4014084507042255,
+      "grad_norm": 0.31696975231170654,
+      "learning_rate": 6.426330513242855e-06,
+      "loss": 0.3386,
+      "step": 5572
+    },
+    {
+      "epoch": 1.4016599597585513,
+      "grad_norm": 0.33291855454444885,
+      "learning_rate": 6.424928004903252e-06,
+      "loss": 0.3386,
+      "step": 5573
+    },
+    {
+      "epoch": 1.4019114688128773,
+      "grad_norm": 0.3217375874519348,
+      "learning_rate": 6.423525374527479e-06,
+      "loss": 0.3495,
+      "step": 5574
+    },
+    {
+      "epoch": 1.4021629778672033,
+      "grad_norm": 0.2930941879749298,
+      "learning_rate": 6.422122622235665e-06,
+      "loss": 0.357,
+      "step": 5575
+    },
+    {
+      "epoch": 1.4024144869215291,
+      "grad_norm": 0.3463757336139679,
+      "learning_rate": 6.420719748147943e-06,
+      "loss": 0.3652,
+      "step": 5576
+    },
+    {
+      "epoch": 1.4026659959758552,
+      "grad_norm": 0.4066574275493622,
+      "learning_rate": 6.419316752384464e-06,
+      "loss": 0.3555,
+      "step": 5577
+    },
+    {
+      "epoch": 1.4029175050301812,
+      "grad_norm": 0.33143872022628784,
+      "learning_rate": 6.417913635065385e-06,
+      "loss": 0.3418,
+      "step": 5578
+    },
+    {
+      "epoch": 1.403169014084507,
+      "grad_norm": 0.30597761273384094,
+      "learning_rate": 6.4165103963108724e-06,
+      "loss": 0.3428,
+      "step": 5579
+    },
+    {
+      "epoch": 1.403420523138833,
+      "grad_norm": 0.33264756202697754,
+      "learning_rate": 6.415107036241106e-06,
+      "loss": 0.3635,
+      "step": 5580
+    },
+    {
+      "epoch": 1.403672032193159,
+      "grad_norm": 0.32489314675331116,
+      "learning_rate": 6.413703554976276e-06,
+      "loss": 0.3375,
+      "step": 5581
+    },
+    {
+      "epoch": 1.4039235412474849,
+      "grad_norm": 0.3370649218559265,
+      "learning_rate": 6.41229995263658e-06,
+      "loss": 0.3336,
+      "step": 5582
+    },
+    {
+      "epoch": 1.404175050301811,
+      "grad_norm": 0.34233397245407104,
+      "learning_rate": 6.410896229342228e-06,
+      "loss": 0.3634,
+      "step": 5583
+    },
+    {
+      "epoch": 1.404426559356137,
+      "grad_norm": 0.34717515110969543,
+      "learning_rate": 6.40949238521344e-06,
+      "loss": 0.3228,
+      "step": 5584
+    },
+    {
+      "epoch": 1.4046780684104627,
+      "grad_norm": 0.30195194482803345,
+      "learning_rate": 6.408088420370448e-06,
+      "loss": 0.3488,
+      "step": 5585
+    },
+    {
+      "epoch": 1.4049295774647887,
+      "grad_norm": 0.33940741419792175,
+      "learning_rate": 6.406684334933491e-06,
+      "loss": 0.3741,
+      "step": 5586
+    },
+    {
+      "epoch": 1.4051810865191148,
+      "grad_norm": 0.3448837101459503,
+      "learning_rate": 6.405280129022821e-06,
+      "loss": 0.3687,
+      "step": 5587
+    },
+    {
+      "epoch": 1.4054325955734406,
+      "grad_norm": 0.33431434631347656,
+      "learning_rate": 6.4038758027587e-06,
+      "loss": 0.3487,
+      "step": 5588
+    },
+    {
+      "epoch": 1.4056841046277666,
+      "grad_norm": 0.32833728194236755,
+      "learning_rate": 6.402471356261399e-06,
+      "loss": 0.3436,
+      "step": 5589
+    },
+    {
+      "epoch": 1.4059356136820926,
+      "grad_norm": 0.3193908929824829,
+      "learning_rate": 6.4010667896512e-06,
+      "loss": 0.3325,
+      "step": 5590
+    },
+    {
+      "epoch": 1.4061871227364184,
+      "grad_norm": 0.36166292428970337,
+      "learning_rate": 6.399662103048396e-06,
+      "loss": 0.3701,
+      "step": 5591
+    },
+    {
+      "epoch": 1.4064386317907445,
+      "grad_norm": 0.3254150152206421,
+      "learning_rate": 6.398257296573288e-06,
+      "loss": 0.3524,
+      "step": 5592
+    },
+    {
+      "epoch": 1.4066901408450705,
+      "grad_norm": 0.30729982256889343,
+      "learning_rate": 6.396852370346191e-06,
+      "loss": 0.3713,
+      "step": 5593
+    },
+    {
+      "epoch": 1.4069416498993963,
+      "grad_norm": 0.3343053162097931,
+      "learning_rate": 6.395447324487427e-06,
+      "loss": 0.3501,
+      "step": 5594
+    },
+    {
+      "epoch": 1.4071931589537223,
+      "grad_norm": 0.3478783369064331,
+      "learning_rate": 6.394042159117329e-06,
+      "loss": 0.3196,
+      "step": 5595
+    },
+    {
+      "epoch": 1.4074446680080483,
+      "grad_norm": 0.30152082443237305,
+      "learning_rate": 6.392636874356242e-06,
+      "loss": 0.3359,
+      "step": 5596
+    },
+    {
+      "epoch": 1.4076961770623742,
+      "grad_norm": 0.3307269513607025,
+      "learning_rate": 6.391231470324517e-06,
+      "loss": 0.3744,
+      "step": 5597
+    },
+    {
+      "epoch": 1.4079476861167002,
+      "grad_norm": 0.33799752593040466,
+      "learning_rate": 6.389825947142523e-06,
+      "loss": 0.3632,
+      "step": 5598
+    },
+    {
+      "epoch": 1.4081991951710262,
+      "grad_norm": 0.31849145889282227,
+      "learning_rate": 6.38842030493063e-06,
+      "loss": 0.3573,
+      "step": 5599
+    },
+    {
+      "epoch": 1.408450704225352,
+      "grad_norm": 0.3179837763309479,
+      "learning_rate": 6.387014543809224e-06,
+      "loss": 0.3578,
+      "step": 5600
+    },
+    {
+      "epoch": 1.408702213279678,
+      "grad_norm": 0.3246977627277374,
+      "learning_rate": 6.3856086638986995e-06,
+      "loss": 0.3751,
+      "step": 5601
+    },
+    {
+      "epoch": 1.408953722334004,
+      "grad_norm": 0.35665377974510193,
+      "learning_rate": 6.38420266531946e-06,
+      "loss": 0.3697,
+      "step": 5602
+    },
+    {
+      "epoch": 1.4092052313883299,
+      "grad_norm": 0.3288874924182892,
+      "learning_rate": 6.382796548191923e-06,
+      "loss": 0.3492,
+      "step": 5603
+    },
+    {
+      "epoch": 1.409456740442656,
+      "grad_norm": 0.3103671073913574,
+      "learning_rate": 6.381390312636513e-06,
+      "loss": 0.3453,
+      "step": 5604
+    },
+    {
+      "epoch": 1.409708249496982,
+      "grad_norm": 0.3277508616447449,
+      "learning_rate": 6.379983958773663e-06,
+      "loss": 0.348,
+      "step": 5605
+    },
+    {
+      "epoch": 1.4099597585513077,
+      "grad_norm": 0.2977486848831177,
+      "learning_rate": 6.378577486723821e-06,
+      "loss": 0.3576,
+      "step": 5606
+    },
+    {
+      "epoch": 1.4102112676056338,
+      "grad_norm": 0.30424800515174866,
+      "learning_rate": 6.377170896607442e-06,
+      "loss": 0.3413,
+      "step": 5607
+    },
+    {
+      "epoch": 1.4104627766599598,
+      "grad_norm": 0.35188329219818115,
+      "learning_rate": 6.3757641885449904e-06,
+      "loss": 0.3816,
+      "step": 5608
+    },
+    {
+      "epoch": 1.4107142857142856,
+      "grad_norm": 0.299070805311203,
+      "learning_rate": 6.374357362656944e-06,
+      "loss": 0.3635,
+      "step": 5609
+    },
+    {
+      "epoch": 1.4109657947686116,
+      "grad_norm": 0.32047006487846375,
+      "learning_rate": 6.372950419063787e-06,
+      "loss": 0.3641,
+      "step": 5610
+    },
+    {
+      "epoch": 1.4112173038229376,
+      "grad_norm": 0.32744547724723816,
+      "learning_rate": 6.3715433578860155e-06,
+      "loss": 0.358,
+      "step": 5611
+    },
+    {
+      "epoch": 1.4114688128772634,
+      "grad_norm": 0.34040847420692444,
+      "learning_rate": 6.3701361792441355e-06,
+      "loss": 0.3533,
+      "step": 5612
+    },
+    {
+      "epoch": 1.4117203219315895,
+      "grad_norm": 0.34646251797676086,
+      "learning_rate": 6.368728883258664e-06,
+      "loss": 0.3391,
+      "step": 5613
+    },
+    {
+      "epoch": 1.4119718309859155,
+      "grad_norm": 0.3157348930835724,
+      "learning_rate": 6.367321470050125e-06,
+      "loss": 0.349,
+      "step": 5614
+    },
+    {
+      "epoch": 1.4122233400402415,
+      "grad_norm": 0.3335605263710022,
+      "learning_rate": 6.365913939739057e-06,
+      "loss": 0.3467,
+      "step": 5615
+    },
+    {
+      "epoch": 1.4124748490945673,
+      "grad_norm": 0.35618653893470764,
+      "learning_rate": 6.364506292446005e-06,
+      "loss": 0.383,
+      "step": 5616
+    },
+    {
+      "epoch": 1.4127263581488934,
+      "grad_norm": 0.3262132406234741,
+      "learning_rate": 6.363098528291525e-06,
+      "loss": 0.3624,
+      "step": 5617
+    },
+    {
+      "epoch": 1.4129778672032194,
+      "grad_norm": 0.3189094662666321,
+      "learning_rate": 6.361690647396184e-06,
+      "loss": 0.3665,
+      "step": 5618
+    },
+    {
+      "epoch": 1.4132293762575452,
+      "grad_norm": 0.29632070660591125,
+      "learning_rate": 6.3602826498805585e-06,
+      "loss": 0.3518,
+      "step": 5619
+    },
+    {
+      "epoch": 1.4134808853118712,
+      "grad_norm": 0.35020795464515686,
+      "learning_rate": 6.358874535865233e-06,
+      "loss": 0.342,
+      "step": 5620
+    },
+    {
+      "epoch": 1.4137323943661972,
+      "grad_norm": 0.322587788105011,
+      "learning_rate": 6.357466305470805e-06,
+      "loss": 0.3361,
+      "step": 5621
+    },
+    {
+      "epoch": 1.4139839034205233,
+      "grad_norm": 0.31895244121551514,
+      "learning_rate": 6.356057958817879e-06,
+      "loss": 0.3561,
+      "step": 5622
+    },
+    {
+      "epoch": 1.414235412474849,
+      "grad_norm": 0.36562541127204895,
+      "learning_rate": 6.354649496027075e-06,
+      "loss": 0.3231,
+      "step": 5623
+    },
+    {
+      "epoch": 1.414486921529175,
+      "grad_norm": 0.3357252776622772,
+      "learning_rate": 6.353240917219014e-06,
+      "loss": 0.3615,
+      "step": 5624
+    },
+    {
+      "epoch": 1.4147384305835011,
+      "grad_norm": 0.3266746699810028,
+      "learning_rate": 6.351832222514335e-06,
+      "loss": 0.3515,
+      "step": 5625
+    },
+    {
+      "epoch": 1.414989939637827,
+      "grad_norm": 0.317125529050827,
+      "learning_rate": 6.350423412033683e-06,
+      "loss": 0.3172,
+      "step": 5626
+    },
+    {
+      "epoch": 1.415241448692153,
+      "grad_norm": 0.3371025621891022,
+      "learning_rate": 6.349014485897714e-06,
+      "loss": 0.351,
+      "step": 5627
+    },
+    {
+      "epoch": 1.415492957746479,
+      "grad_norm": 0.33615246415138245,
+      "learning_rate": 6.347605444227093e-06,
+      "loss": 0.3431,
+      "step": 5628
+    },
+    {
+      "epoch": 1.4157444668008048,
+      "grad_norm": 0.36137500405311584,
+      "learning_rate": 6.346196287142497e-06,
+      "loss": 0.3449,
+      "step": 5629
+    },
+    {
+      "epoch": 1.4159959758551308,
+      "grad_norm": 0.37134072184562683,
+      "learning_rate": 6.344787014764611e-06,
+      "loss": 0.3563,
+      "step": 5630
+    },
+    {
+      "epoch": 1.4162474849094568,
+      "grad_norm": 0.3492390811443329,
+      "learning_rate": 6.34337762721413e-06,
+      "loss": 0.3607,
+      "step": 5631
+    },
+    {
+      "epoch": 1.4164989939637826,
+      "grad_norm": 0.3468017578125,
+      "learning_rate": 6.341968124611759e-06,
+      "loss": 0.3554,
+      "step": 5632
+    },
+    {
+      "epoch": 1.4167505030181087,
+      "grad_norm": 0.2991659939289093,
+      "learning_rate": 6.340558507078215e-06,
+      "loss": 0.3202,
+      "step": 5633
+    },
+    {
+      "epoch": 1.4170020120724347,
+      "grad_norm": 0.3461068272590637,
+      "learning_rate": 6.339148774734221e-06,
+      "loss": 0.35,
+      "step": 5634
+    },
+    {
+      "epoch": 1.4172535211267605,
+      "grad_norm": 0.3426593542098999,
+      "learning_rate": 6.33773892770051e-06,
+      "loss": 0.3411,
+      "step": 5635
+    },
+    {
+      "epoch": 1.4175050301810865,
+      "grad_norm": 0.3247561752796173,
+      "learning_rate": 6.3363289660978315e-06,
+      "loss": 0.3484,
+      "step": 5636
+    },
+    {
+      "epoch": 1.4177565392354126,
+      "grad_norm": 0.3241268992424011,
+      "learning_rate": 6.334918890046935e-06,
+      "loss": 0.3471,
+      "step": 5637
+    },
+    {
+      "epoch": 1.4180080482897384,
+      "grad_norm": 0.36099836230278015,
+      "learning_rate": 6.333508699668587e-06,
+      "loss": 0.3545,
+      "step": 5638
+    },
+    {
+      "epoch": 1.4182595573440644,
+      "grad_norm": 0.31838127970695496,
+      "learning_rate": 6.332098395083562e-06,
+      "loss": 0.3412,
+      "step": 5639
+    },
+    {
+      "epoch": 1.4185110663983904,
+      "grad_norm": 0.3607694208621979,
+      "learning_rate": 6.330687976412642e-06,
+      "loss": 0.3937,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4187625754527162,
+      "grad_norm": 0.3097631335258484,
+      "learning_rate": 6.329277443776623e-06,
+      "loss": 0.3405,
+      "step": 5641
+    },
+    {
+      "epoch": 1.4190140845070423,
+      "grad_norm": 0.3108190894126892,
+      "learning_rate": 6.327866797296306e-06,
+      "loss": 0.3474,
+      "step": 5642
+    },
+    {
+      "epoch": 1.4192655935613683,
+      "grad_norm": 0.32645246386528015,
+      "learning_rate": 6.326456037092505e-06,
+      "loss": 0.3367,
+      "step": 5643
+    },
+    {
+      "epoch": 1.419517102615694,
+      "grad_norm": 0.3206358850002289,
+      "learning_rate": 6.325045163286043e-06,
+      "loss": 0.3654,
+      "step": 5644
+    },
+    {
+      "epoch": 1.41976861167002,
+      "grad_norm": 0.3293161988258362,
+      "learning_rate": 6.323634175997753e-06,
+      "loss": 0.3539,
+      "step": 5645
+    },
+    {
+      "epoch": 1.4200201207243461,
+      "grad_norm": 0.3224540948867798,
+      "learning_rate": 6.322223075348475e-06,
+      "loss": 0.3403,
+      "step": 5646
+    },
+    {
+      "epoch": 1.420271629778672,
+      "grad_norm": 0.3604409694671631,
+      "learning_rate": 6.320811861459063e-06,
+      "loss": 0.3479,
+      "step": 5647
+    },
+    {
+      "epoch": 1.420523138832998,
+      "grad_norm": 0.337251216173172,
+      "learning_rate": 6.319400534450378e-06,
+      "loss": 0.3347,
+      "step": 5648
+    },
+    {
+      "epoch": 1.420774647887324,
+      "grad_norm": 0.2904793322086334,
+      "learning_rate": 6.317989094443291e-06,
+      "loss": 0.3252,
+      "step": 5649
+    },
+    {
+      "epoch": 1.4210261569416498,
+      "grad_norm": 0.31111976504325867,
+      "learning_rate": 6.316577541558683e-06,
+      "loss": 0.3448,
+      "step": 5650
+    },
+    {
+      "epoch": 1.4212776659959758,
+      "grad_norm": 0.3352508544921875,
+      "learning_rate": 6.315165875917446e-06,
+      "loss": 0.3272,
+      "step": 5651
+    },
+    {
+      "epoch": 1.4215291750503019,
+      "grad_norm": 0.33469197154045105,
+      "learning_rate": 6.313754097640479e-06,
+      "loss": 0.3502,
+      "step": 5652
+    },
+    {
+      "epoch": 1.4217806841046277,
+      "grad_norm": 0.3249984383583069,
+      "learning_rate": 6.312342206848693e-06,
+      "loss": 0.3638,
+      "step": 5653
+    },
+    {
+      "epoch": 1.4220321931589537,
+      "grad_norm": 0.3012884259223938,
+      "learning_rate": 6.310930203663006e-06,
+      "loss": 0.3517,
+      "step": 5654
+    },
+    {
+      "epoch": 1.4222837022132797,
+      "grad_norm": 0.32274380326271057,
+      "learning_rate": 6.309518088204349e-06,
+      "loss": 0.3494,
+      "step": 5655
+    },
+    {
+      "epoch": 1.4225352112676055,
+      "grad_norm": 0.3095703721046448,
+      "learning_rate": 6.3081058605936594e-06,
+      "loss": 0.3247,
+      "step": 5656
+    },
+    {
+      "epoch": 1.4227867203219315,
+      "grad_norm": 0.3313826024532318,
+      "learning_rate": 6.3066935209518875e-06,
+      "loss": 0.3503,
+      "step": 5657
+    },
+    {
+      "epoch": 1.4230382293762576,
+      "grad_norm": 0.3489868938922882,
+      "learning_rate": 6.305281069399989e-06,
+      "loss": 0.349,
+      "step": 5658
+    },
+    {
+      "epoch": 1.4232897384305834,
+      "grad_norm": 0.2928728759288788,
+      "learning_rate": 6.303868506058933e-06,
+      "loss": 0.3532,
+      "step": 5659
+    },
+    {
+      "epoch": 1.4235412474849094,
+      "grad_norm": 0.3247617781162262,
+      "learning_rate": 6.302455831049696e-06,
+      "loss": 0.3393,
+      "step": 5660
+    },
+    {
+      "epoch": 1.4237927565392354,
+      "grad_norm": 0.32944563031196594,
+      "learning_rate": 6.301043044493266e-06,
+      "loss": 0.3453,
+      "step": 5661
+    },
+    {
+      "epoch": 1.4240442655935612,
+      "grad_norm": 0.31871670484542847,
+      "learning_rate": 6.299630146510638e-06,
+      "loss": 0.3624,
+      "step": 5662
+    },
+    {
+      "epoch": 1.4242957746478873,
+      "grad_norm": 0.34367579221725464,
+      "learning_rate": 6.2982171372228196e-06,
+      "loss": 0.3523,
+      "step": 5663
+    },
+    {
+      "epoch": 1.4245472837022133,
+      "grad_norm": 0.3165755569934845,
+      "learning_rate": 6.296804016750824e-06,
+      "loss": 0.3379,
+      "step": 5664
+    },
+    {
+      "epoch": 1.4247987927565393,
+      "grad_norm": 0.31686344742774963,
+      "learning_rate": 6.295390785215677e-06,
+      "loss": 0.3231,
+      "step": 5665
+    },
+    {
+      "epoch": 1.4250503018108651,
+      "grad_norm": 0.3904785215854645,
+      "learning_rate": 6.293977442738414e-06,
+      "loss": 0.3522,
+      "step": 5666
+    },
+    {
+      "epoch": 1.4253018108651911,
+      "grad_norm": 0.336385041475296,
+      "learning_rate": 6.292563989440077e-06,
+      "loss": 0.358,
+      "step": 5667
+    },
+    {
+      "epoch": 1.4255533199195172,
+      "grad_norm": 0.3112278878688812,
+      "learning_rate": 6.291150425441721e-06,
+      "loss": 0.3522,
+      "step": 5668
+    },
+    {
+      "epoch": 1.4258048289738432,
+      "grad_norm": 0.3297823667526245,
+      "learning_rate": 6.289736750864409e-06,
+      "loss": 0.3437,
+      "step": 5669
+    },
+    {
+      "epoch": 1.426056338028169,
+      "grad_norm": 0.3268705904483795,
+      "learning_rate": 6.288322965829212e-06,
+      "loss": 0.3288,
+      "step": 5670
+    },
+    {
+      "epoch": 1.426307847082495,
+      "grad_norm": 0.3283686339855194,
+      "learning_rate": 6.286909070457213e-06,
+      "loss": 0.3501,
+      "step": 5671
+    },
+    {
+      "epoch": 1.426559356136821,
+      "grad_norm": 0.3172062635421753,
+      "learning_rate": 6.285495064869503e-06,
+      "loss": 0.3426,
+      "step": 5672
+    },
+    {
+      "epoch": 1.4268108651911469,
+      "grad_norm": 0.35733988881111145,
+      "learning_rate": 6.284080949187183e-06,
+      "loss": 0.3113,
+      "step": 5673
+    },
+    {
+      "epoch": 1.4270623742454729,
+      "grad_norm": 0.32833603024482727,
+      "learning_rate": 6.282666723531363e-06,
+      "loss": 0.345,
+      "step": 5674
+    },
+    {
+      "epoch": 1.427313883299799,
+      "grad_norm": 0.3474726378917694,
+      "learning_rate": 6.281252388023162e-06,
+      "loss": 0.351,
+      "step": 5675
+    },
+    {
+      "epoch": 1.4275653923541247,
+      "grad_norm": 0.3398433029651642,
+      "learning_rate": 6.279837942783711e-06,
+      "loss": 0.3523,
+      "step": 5676
+    },
+    {
+      "epoch": 1.4278169014084507,
+      "grad_norm": 0.35440170764923096,
+      "learning_rate": 6.278423387934145e-06,
+      "loss": 0.3421,
+      "step": 5677
+    },
+    {
+      "epoch": 1.4280684104627768,
+      "grad_norm": 0.3336888551712036,
+      "learning_rate": 6.277008723595615e-06,
+      "loss": 0.3525,
+      "step": 5678
+    },
+    {
+      "epoch": 1.4283199195171026,
+      "grad_norm": 0.2921485900878906,
+      "learning_rate": 6.275593949889276e-06,
+      "loss": 0.3576,
+      "step": 5679
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.3030543029308319,
+      "learning_rate": 6.274179066936294e-06,
+      "loss": 0.364,
+      "step": 5680
+    },
+    {
+      "epoch": 1.4288229376257546,
+      "grad_norm": 0.328914612531662,
+      "learning_rate": 6.272764074857848e-06,
+      "loss": 0.3624,
+      "step": 5681
+    },
+    {
+      "epoch": 1.4290744466800804,
+      "grad_norm": 0.3531835973262787,
+      "learning_rate": 6.27134897377512e-06,
+      "loss": 0.3486,
+      "step": 5682
+    },
+    {
+      "epoch": 1.4293259557344065,
+      "grad_norm": 0.3383164703845978,
+      "learning_rate": 6.269933763809306e-06,
+      "loss": 0.3471,
+      "step": 5683
+    },
+    {
+      "epoch": 1.4295774647887325,
+      "grad_norm": 0.31009918451309204,
+      "learning_rate": 6.268518445081611e-06,
+      "loss": 0.3755,
+      "step": 5684
+    },
+    {
+      "epoch": 1.4298289738430583,
+      "grad_norm": 0.29445788264274597,
+      "learning_rate": 6.2671030177132466e-06,
+      "loss": 0.3177,
+      "step": 5685
+    },
+    {
+      "epoch": 1.4300804828973843,
+      "grad_norm": 0.33102452754974365,
+      "learning_rate": 6.265687481825435e-06,
+      "loss": 0.3525,
+      "step": 5686
+    },
+    {
+      "epoch": 1.4303319919517103,
+      "grad_norm": 0.31349411606788635,
+      "learning_rate": 6.26427183753941e-06,
+      "loss": 0.3394,
+      "step": 5687
+    },
+    {
+      "epoch": 1.4305835010060362,
+      "grad_norm": 0.30602481961250305,
+      "learning_rate": 6.262856084976411e-06,
+      "loss": 0.3581,
+      "step": 5688
+    },
+    {
+      "epoch": 1.4308350100603622,
+      "grad_norm": 0.31970512866973877,
+      "learning_rate": 6.261440224257688e-06,
+      "loss": 0.3677,
+      "step": 5689
+    },
+    {
+      "epoch": 1.4310865191146882,
+      "grad_norm": 0.29605525732040405,
+      "learning_rate": 6.260024255504502e-06,
+      "loss": 0.3446,
+      "step": 5690
+    },
+    {
+      "epoch": 1.431338028169014,
+      "grad_norm": 0.3755524754524231,
+      "learning_rate": 6.258608178838122e-06,
+      "loss": 0.3658,
+      "step": 5691
+    },
+    {
+      "epoch": 1.43158953722334,
+      "grad_norm": 0.31317082047462463,
+      "learning_rate": 6.2571919943798235e-06,
+      "loss": 0.3382,
+      "step": 5692
+    },
+    {
+      "epoch": 1.431841046277666,
+      "grad_norm": 0.32691314816474915,
+      "learning_rate": 6.255775702250895e-06,
+      "loss": 0.349,
+      "step": 5693
+    },
+    {
+      "epoch": 1.4320925553319919,
+      "grad_norm": 0.31760266423225403,
+      "learning_rate": 6.254359302572635e-06,
+      "loss": 0.3833,
+      "step": 5694
+    },
+    {
+      "epoch": 1.432344064386318,
+      "grad_norm": 0.3521325886249542,
+      "learning_rate": 6.252942795466348e-06,
+      "loss": 0.3629,
+      "step": 5695
+    },
+    {
+      "epoch": 1.432595573440644,
+      "grad_norm": 0.30917102098464966,
+      "learning_rate": 6.251526181053349e-06,
+      "loss": 0.3508,
+      "step": 5696
+    },
+    {
+      "epoch": 1.4328470824949697,
+      "grad_norm": 0.3195309340953827,
+      "learning_rate": 6.250109459454963e-06,
+      "loss": 0.3367,
+      "step": 5697
+    },
+    {
+      "epoch": 1.4330985915492958,
+      "grad_norm": 0.32293376326560974,
+      "learning_rate": 6.248692630792521e-06,
+      "loss": 0.3347,
+      "step": 5698
+    },
+    {
+      "epoch": 1.4333501006036218,
+      "grad_norm": 0.34250807762145996,
+      "learning_rate": 6.247275695187368e-06,
+      "loss": 0.3414,
+      "step": 5699
+    },
+    {
+      "epoch": 1.4336016096579476,
+      "grad_norm": 0.3318396806716919,
+      "learning_rate": 6.245858652760854e-06,
+      "loss": 0.336,
+      "step": 5700
+    },
+    {
+      "epoch": 1.4338531187122736,
+      "grad_norm": 0.34146398305892944,
+      "learning_rate": 6.244441503634341e-06,
+      "loss": 0.357,
+      "step": 5701
+    },
+    {
+      "epoch": 1.4341046277665996,
+      "grad_norm": 0.2918093204498291,
+      "learning_rate": 6.243024247929198e-06,
+      "loss": 0.3461,
+      "step": 5702
+    },
+    {
+      "epoch": 1.4343561368209254,
+      "grad_norm": 0.32614874839782715,
+      "learning_rate": 6.2416068857668045e-06,
+      "loss": 0.3522,
+      "step": 5703
+    },
+    {
+      "epoch": 1.4346076458752515,
+      "grad_norm": 0.3218587040901184,
+      "learning_rate": 6.240189417268548e-06,
+      "loss": 0.3755,
+      "step": 5704
+    },
+    {
+      "epoch": 1.4348591549295775,
+      "grad_norm": 0.3028290271759033,
+      "learning_rate": 6.238771842555828e-06,
+      "loss": 0.3393,
+      "step": 5705
+    },
+    {
+      "epoch": 1.4351106639839033,
+      "grad_norm": 0.2967846393585205,
+      "learning_rate": 6.2373541617500475e-06,
+      "loss": 0.3598,
+      "step": 5706
+    },
+    {
+      "epoch": 1.4353621730382293,
+      "grad_norm": 0.3133407235145569,
+      "learning_rate": 6.235936374972626e-06,
+      "loss": 0.3657,
+      "step": 5707
+    },
+    {
+      "epoch": 1.4356136820925554,
+      "grad_norm": 0.30290549993515015,
+      "learning_rate": 6.234518482344985e-06,
+      "loss": 0.34,
+      "step": 5708
+    },
+    {
+      "epoch": 1.4358651911468812,
+      "grad_norm": 0.2992716133594513,
+      "learning_rate": 6.233100483988559e-06,
+      "loss": 0.3348,
+      "step": 5709
+    },
+    {
+      "epoch": 1.4361167002012072,
+      "grad_norm": 0.34080713987350464,
+      "learning_rate": 6.231682380024792e-06,
+      "loss": 0.3632,
+      "step": 5710
+    },
+    {
+      "epoch": 1.4363682092555332,
+      "grad_norm": 0.3630107641220093,
+      "learning_rate": 6.230264170575133e-06,
+      "loss": 0.3345,
+      "step": 5711
+    },
+    {
+      "epoch": 1.436619718309859,
+      "grad_norm": 0.31429681181907654,
+      "learning_rate": 6.228845855761044e-06,
+      "loss": 0.3585,
+      "step": 5712
+    },
+    {
+      "epoch": 1.436871227364185,
+      "grad_norm": 0.298191100358963,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.3379,
+      "step": 5713
+    },
+    {
+      "epoch": 1.437122736418511,
+      "grad_norm": 0.33849817514419556,
+      "learning_rate": 6.226008910525466e-06,
+      "loss": 0.3389,
+      "step": 5714
+    },
+    {
+      "epoch": 1.437374245472837,
+      "grad_norm": 0.3714556396007538,
+      "learning_rate": 6.224590280346944e-06,
+      "loss": 0.3591,
+      "step": 5715
+    },
+    {
+      "epoch": 1.437625754527163,
+      "grad_norm": 0.3296235501766205,
+      "learning_rate": 6.223171545289925e-06,
+      "loss": 0.3273,
+      "step": 5716
+    },
+    {
+      "epoch": 1.437877263581489,
+      "grad_norm": 0.34104108810424805,
+      "learning_rate": 6.221752705475915e-06,
+      "loss": 0.3745,
+      "step": 5717
+    },
+    {
+      "epoch": 1.438128772635815,
+      "grad_norm": 0.3797450065612793,
+      "learning_rate": 6.22033376102643e-06,
+      "loss": 0.3646,
+      "step": 5718
+    },
+    {
+      "epoch": 1.438380281690141,
+      "grad_norm": 0.3067338764667511,
+      "learning_rate": 6.218914712062992e-06,
+      "loss": 0.3526,
+      "step": 5719
+    },
+    {
+      "epoch": 1.4386317907444668,
+      "grad_norm": 0.30397993326187134,
+      "learning_rate": 6.217495558707135e-06,
+      "loss": 0.3417,
+      "step": 5720
+    },
+    {
+      "epoch": 1.4388832997987928,
+      "grad_norm": 0.34097644686698914,
+      "learning_rate": 6.2160763010803995e-06,
+      "loss": 0.3868,
+      "step": 5721
+    },
+    {
+      "epoch": 1.4391348088531188,
+      "grad_norm": 0.3221248686313629,
+      "learning_rate": 6.214656939304337e-06,
+      "loss": 0.3695,
+      "step": 5722
+    },
+    {
+      "epoch": 1.4393863179074446,
+      "grad_norm": 0.30336886644363403,
+      "learning_rate": 6.213237473500505e-06,
+      "loss": 0.3425,
+      "step": 5723
+    },
+    {
+      "epoch": 1.4396378269617707,
+      "grad_norm": 0.30754998326301575,
+      "learning_rate": 6.2118179037904755e-06,
+      "loss": 0.3354,
+      "step": 5724
+    },
+    {
+      "epoch": 1.4398893360160967,
+      "grad_norm": 0.35792577266693115,
+      "learning_rate": 6.2103982302958225e-06,
+      "loss": 0.3458,
+      "step": 5725
+    },
+    {
+      "epoch": 1.4401408450704225,
+      "grad_norm": 0.3202516436576843,
+      "learning_rate": 6.208978453138134e-06,
+      "loss": 0.3495,
+      "step": 5726
+    },
+    {
+      "epoch": 1.4403923541247485,
+      "grad_norm": 0.30896228551864624,
+      "learning_rate": 6.207558572439003e-06,
+      "loss": 0.3359,
+      "step": 5727
+    },
+    {
+      "epoch": 1.4406438631790746,
+      "grad_norm": 0.3637436330318451,
+      "learning_rate": 6.2061385883200365e-06,
+      "loss": 0.3406,
+      "step": 5728
+    },
+    {
+      "epoch": 1.4408953722334004,
+      "grad_norm": 0.32914838194847107,
+      "learning_rate": 6.204718500902845e-06,
+      "loss": 0.3735,
+      "step": 5729
+    },
+    {
+      "epoch": 1.4411468812877264,
+      "grad_norm": 0.3156082332134247,
+      "learning_rate": 6.20329831030905e-06,
+      "loss": 0.3455,
+      "step": 5730
+    },
+    {
+      "epoch": 1.4413983903420524,
+      "grad_norm": 0.35359394550323486,
+      "learning_rate": 6.201878016660282e-06,
+      "loss": 0.3381,
+      "step": 5731
+    },
+    {
+      "epoch": 1.4416498993963782,
+      "grad_norm": 0.3604298532009125,
+      "learning_rate": 6.200457620078182e-06,
+      "loss": 0.3768,
+      "step": 5732
+    },
+    {
+      "epoch": 1.4419014084507042,
+      "grad_norm": 0.31656792759895325,
+      "learning_rate": 6.199037120684396e-06,
+      "loss": 0.3786,
+      "step": 5733
+    },
+    {
+      "epoch": 1.4421529175050303,
+      "grad_norm": 0.32002219557762146,
+      "learning_rate": 6.1976165186005825e-06,
+      "loss": 0.3784,
+      "step": 5734
+    },
+    {
+      "epoch": 1.442404426559356,
+      "grad_norm": 0.33488380908966064,
+      "learning_rate": 6.196195813948406e-06,
+      "loss": 0.3552,
+      "step": 5735
+    },
+    {
+      "epoch": 1.442655935613682,
+      "grad_norm": 0.33732855319976807,
+      "learning_rate": 6.194775006849541e-06,
+      "loss": 0.3446,
+      "step": 5736
+    },
+    {
+      "epoch": 1.4429074446680081,
+      "grad_norm": 0.27075111865997314,
+      "learning_rate": 6.19335409742567e-06,
+      "loss": 0.3357,
+      "step": 5737
+    },
+    {
+      "epoch": 1.443158953722334,
+      "grad_norm": 0.3128473162651062,
+      "learning_rate": 6.191933085798488e-06,
+      "loss": 0.3443,
+      "step": 5738
+    },
+    {
+      "epoch": 1.44341046277666,
+      "grad_norm": 0.3088710606098175,
+      "learning_rate": 6.190511972089694e-06,
+      "loss": 0.3581,
+      "step": 5739
+    },
+    {
+      "epoch": 1.443661971830986,
+      "grad_norm": 0.29833412170410156,
+      "learning_rate": 6.189090756420997e-06,
+      "loss": 0.3203,
+      "step": 5740
+    },
+    {
+      "epoch": 1.4439134808853118,
+      "grad_norm": 0.318301260471344,
+      "learning_rate": 6.187669438914116e-06,
+      "loss": 0.3681,
+      "step": 5741
+    },
+    {
+      "epoch": 1.4441649899396378,
+      "grad_norm": 0.3490622341632843,
+      "learning_rate": 6.186248019690777e-06,
+      "loss": 0.3474,
+      "step": 5742
+    },
+    {
+      "epoch": 1.4444164989939638,
+      "grad_norm": 0.33258891105651855,
+      "learning_rate": 6.1848264988727165e-06,
+      "loss": 0.3604,
+      "step": 5743
+    },
+    {
+      "epoch": 1.4446680080482897,
+      "grad_norm": 0.354710191488266,
+      "learning_rate": 6.183404876581679e-06,
+      "loss": 0.3331,
+      "step": 5744
+    },
+    {
+      "epoch": 1.4449195171026157,
+      "grad_norm": 0.3129853904247284,
+      "learning_rate": 6.181983152939417e-06,
+      "loss": 0.3661,
+      "step": 5745
+    },
+    {
+      "epoch": 1.4451710261569417,
+      "grad_norm": 0.3277970850467682,
+      "learning_rate": 6.180561328067692e-06,
+      "loss": 0.3596,
+      "step": 5746
+    },
+    {
+      "epoch": 1.4454225352112675,
+      "grad_norm": 0.3437343239784241,
+      "learning_rate": 6.179139402088275e-06,
+      "loss": 0.3577,
+      "step": 5747
+    },
+    {
+      "epoch": 1.4456740442655935,
+      "grad_norm": 0.3663479685783386,
+      "learning_rate": 6.1777173751229445e-06,
+      "loss": 0.3585,
+      "step": 5748
+    },
+    {
+      "epoch": 1.4459255533199196,
+      "grad_norm": 0.3225582242012024,
+      "learning_rate": 6.176295247293489e-06,
+      "loss": 0.347,
+      "step": 5749
+    },
+    {
+      "epoch": 1.4461770623742454,
+      "grad_norm": 0.33006182312965393,
+      "learning_rate": 6.174873018721705e-06,
+      "loss": 0.3493,
+      "step": 5750
+    },
+    {
+      "epoch": 1.4464285714285714,
+      "grad_norm": 0.33908021450042725,
+      "learning_rate": 6.173450689529397e-06,
+      "loss": 0.3456,
+      "step": 5751
+    },
+    {
+      "epoch": 1.4466800804828974,
+      "grad_norm": 0.35512205958366394,
+      "learning_rate": 6.172028259838378e-06,
+      "loss": 0.3777,
+      "step": 5752
+    },
+    {
+      "epoch": 1.4469315895372232,
+      "grad_norm": 0.3256780207157135,
+      "learning_rate": 6.17060572977047e-06,
+      "loss": 0.3644,
+      "step": 5753
+    },
+    {
+      "epoch": 1.4471830985915493,
+      "grad_norm": 0.39678511023521423,
+      "learning_rate": 6.169183099447505e-06,
+      "loss": 0.3304,
+      "step": 5754
+    },
+    {
+      "epoch": 1.4474346076458753,
+      "grad_norm": 0.3706040680408478,
+      "learning_rate": 6.167760368991322e-06,
+      "loss": 0.3491,
+      "step": 5755
+    },
+    {
+      "epoch": 1.447686116700201,
+      "grad_norm": 0.3017033040523529,
+      "learning_rate": 6.16633753852377e-06,
+      "loss": 0.3393,
+      "step": 5756
+    },
+    {
+      "epoch": 1.4479376257545271,
+      "grad_norm": 0.32443055510520935,
+      "learning_rate": 6.164914608166703e-06,
+      "loss": 0.3369,
+      "step": 5757
+    },
+    {
+      "epoch": 1.4481891348088531,
+      "grad_norm": 0.33810749650001526,
+      "learning_rate": 6.163491578041988e-06,
+      "loss": 0.3254,
+      "step": 5758
+    },
+    {
+      "epoch": 1.448440643863179,
+      "grad_norm": 0.3356841504573822,
+      "learning_rate": 6.1620684482714975e-06,
+      "loss": 0.3624,
+      "step": 5759
+    },
+    {
+      "epoch": 1.448692152917505,
+      "grad_norm": 0.32188060879707336,
+      "learning_rate": 6.160645218977115e-06,
+      "loss": 0.327,
+      "step": 5760
+    },
+    {
+      "epoch": 1.448943661971831,
+      "grad_norm": 0.30660948157310486,
+      "learning_rate": 6.159221890280731e-06,
+      "loss": 0.3529,
+      "step": 5761
+    },
+    {
+      "epoch": 1.4491951710261568,
+      "grad_norm": 0.3086193799972534,
+      "learning_rate": 6.157798462304243e-06,
+      "loss": 0.3633,
+      "step": 5762
+    },
+    {
+      "epoch": 1.4494466800804828,
+      "grad_norm": 0.33312803506851196,
+      "learning_rate": 6.15637493516956e-06,
+      "loss": 0.3652,
+      "step": 5763
+    },
+    {
+      "epoch": 1.4496981891348089,
+      "grad_norm": 0.3129318058490753,
+      "learning_rate": 6.154951308998599e-06,
+      "loss": 0.378,
+      "step": 5764
+    },
+    {
+      "epoch": 1.4499496981891349,
+      "grad_norm": 0.31855911016464233,
+      "learning_rate": 6.153527583913284e-06,
+      "loss": 0.338,
+      "step": 5765
+    },
+    {
+      "epoch": 1.4502012072434607,
+      "grad_norm": 0.3277183473110199,
+      "learning_rate": 6.152103760035546e-06,
+      "loss": 0.3527,
+      "step": 5766
+    },
+    {
+      "epoch": 1.4504527162977867,
+      "grad_norm": 0.34277212619781494,
+      "learning_rate": 6.15067983748733e-06,
+      "loss": 0.3351,
+      "step": 5767
+    },
+    {
+      "epoch": 1.4507042253521127,
+      "grad_norm": 0.3247811198234558,
+      "learning_rate": 6.149255816390585e-06,
+      "loss": 0.3464,
+      "step": 5768
+    },
+    {
+      "epoch": 1.4509557344064388,
+      "grad_norm": 0.3350554406642914,
+      "learning_rate": 6.147831696867266e-06,
+      "loss": 0.3549,
+      "step": 5769
+    },
+    {
+      "epoch": 1.4512072434607646,
+      "grad_norm": 0.33898022770881653,
+      "learning_rate": 6.146407479039345e-06,
+      "loss": 0.3493,
+      "step": 5770
+    },
+    {
+      "epoch": 1.4514587525150906,
+      "grad_norm": 0.3372337818145752,
+      "learning_rate": 6.144983163028796e-06,
+      "loss": 0.3463,
+      "step": 5771
+    },
+    {
+      "epoch": 1.4517102615694166,
+      "grad_norm": 0.36416393518447876,
+      "learning_rate": 6.143558748957601e-06,
+      "loss": 0.3475,
+      "step": 5772
+    },
+    {
+      "epoch": 1.4519617706237424,
+      "grad_norm": 0.33351224660873413,
+      "learning_rate": 6.142134236947755e-06,
+      "loss": 0.358,
+      "step": 5773
+    },
+    {
+      "epoch": 1.4522132796780685,
+      "grad_norm": 0.29332223534584045,
+      "learning_rate": 6.140709627121255e-06,
+      "loss": 0.3688,
+      "step": 5774
+    },
+    {
+      "epoch": 1.4524647887323945,
+      "grad_norm": 0.34419727325439453,
+      "learning_rate": 6.1392849196001125e-06,
+      "loss": 0.3338,
+      "step": 5775
+    },
+    {
+      "epoch": 1.4527162977867203,
+      "grad_norm": 0.32147595286369324,
+      "learning_rate": 6.137860114506343e-06,
+      "loss": 0.3517,
+      "step": 5776
+    },
+    {
+      "epoch": 1.4529678068410463,
+      "grad_norm": 0.3297555148601532,
+      "learning_rate": 6.136435211961974e-06,
+      "loss": 0.3182,
+      "step": 5777
+    },
+    {
+      "epoch": 1.4532193158953723,
+      "grad_norm": 0.3482721745967865,
+      "learning_rate": 6.135010212089038e-06,
+      "loss": 0.3632,
+      "step": 5778
+    },
+    {
+      "epoch": 1.4534708249496981,
+      "grad_norm": 0.32314664125442505,
+      "learning_rate": 6.133585115009579e-06,
+      "loss": 0.3837,
+      "step": 5779
+    },
+    {
+      "epoch": 1.4537223340040242,
+      "grad_norm": 0.33287930488586426,
+      "learning_rate": 6.132159920845645e-06,
+      "loss": 0.3742,
+      "step": 5780
+    },
+    {
+      "epoch": 1.4539738430583502,
+      "grad_norm": 0.32902443408966064,
+      "learning_rate": 6.1307346297192984e-06,
+      "loss": 0.3698,
+      "step": 5781
+    },
+    {
+      "epoch": 1.454225352112676,
+      "grad_norm": 0.3330136835575104,
+      "learning_rate": 6.129309241752603e-06,
+      "loss": 0.358,
+      "step": 5782
+    },
+    {
+      "epoch": 1.454476861167002,
+      "grad_norm": 0.3234924376010895,
+      "learning_rate": 6.127883757067636e-06,
+      "loss": 0.3546,
+      "step": 5783
+    },
+    {
+      "epoch": 1.454728370221328,
+      "grad_norm": 0.3510439693927765,
+      "learning_rate": 6.126458175786483e-06,
+      "loss": 0.3506,
+      "step": 5784
+    },
+    {
+      "epoch": 1.4549798792756539,
+      "grad_norm": 0.32338663935661316,
+      "learning_rate": 6.125032498031234e-06,
+      "loss": 0.3674,
+      "step": 5785
+    },
+    {
+      "epoch": 1.45523138832998,
+      "grad_norm": 0.3300342857837677,
+      "learning_rate": 6.1236067239239885e-06,
+      "loss": 0.3535,
+      "step": 5786
+    },
+    {
+      "epoch": 1.455482897384306,
+      "grad_norm": 0.31873607635498047,
+      "learning_rate": 6.122180853586857e-06,
+      "loss": 0.3526,
+      "step": 5787
+    },
+    {
+      "epoch": 1.4557344064386317,
+      "grad_norm": 0.3264380991458893,
+      "learning_rate": 6.120754887141955e-06,
+      "loss": 0.3741,
+      "step": 5788
+    },
+    {
+      "epoch": 1.4559859154929577,
+      "grad_norm": 0.36225321888923645,
+      "learning_rate": 6.119328824711409e-06,
+      "loss": 0.3473,
+      "step": 5789
+    },
+    {
+      "epoch": 1.4562374245472838,
+      "grad_norm": 0.3539520800113678,
+      "learning_rate": 6.117902666417352e-06,
+      "loss": 0.3592,
+      "step": 5790
+    },
+    {
+      "epoch": 1.4564889336016096,
+      "grad_norm": 0.3481525182723999,
+      "learning_rate": 6.116476412381926e-06,
+      "loss": 0.364,
+      "step": 5791
+    },
+    {
+      "epoch": 1.4567404426559356,
+      "grad_norm": 0.3597382605075836,
+      "learning_rate": 6.115050062727278e-06,
+      "loss": 0.3502,
+      "step": 5792
+    },
+    {
+      "epoch": 1.4569919517102616,
+      "grad_norm": 0.3276865482330322,
+      "learning_rate": 6.113623617575568e-06,
+      "loss": 0.3589,
+      "step": 5793
+    },
+    {
+      "epoch": 1.4572434607645874,
+      "grad_norm": 0.35708415508270264,
+      "learning_rate": 6.112197077048963e-06,
+      "loss": 0.374,
+      "step": 5794
+    },
+    {
+      "epoch": 1.4574949698189135,
+      "grad_norm": 0.33358344435691833,
+      "learning_rate": 6.110770441269636e-06,
+      "loss": 0.3551,
+      "step": 5795
+    },
+    {
+      "epoch": 1.4577464788732395,
+      "grad_norm": 0.3249559998512268,
+      "learning_rate": 6.10934371035977e-06,
+      "loss": 0.3376,
+      "step": 5796
+    },
+    {
+      "epoch": 1.4579979879275653,
+      "grad_norm": 0.35899099707603455,
+      "learning_rate": 6.1079168844415535e-06,
+      "loss": 0.3644,
+      "step": 5797
+    },
+    {
+      "epoch": 1.4582494969818913,
+      "grad_norm": 0.33779412508010864,
+      "learning_rate": 6.106489963637189e-06,
+      "loss": 0.3545,
+      "step": 5798
+    },
+    {
+      "epoch": 1.4585010060362174,
+      "grad_norm": 0.3385300040245056,
+      "learning_rate": 6.105062948068881e-06,
+      "loss": 0.3381,
+      "step": 5799
+    },
+    {
+      "epoch": 1.4587525150905432,
+      "grad_norm": 0.3292633295059204,
+      "learning_rate": 6.103635837858844e-06,
+      "loss": 0.3323,
+      "step": 5800
+    },
+    {
+      "epoch": 1.4590040241448692,
+      "grad_norm": 0.3348362445831299,
+      "learning_rate": 6.1022086331293005e-06,
+      "loss": 0.3462,
+      "step": 5801
+    },
+    {
+      "epoch": 1.4592555331991952,
+      "grad_norm": 0.3152986764907837,
+      "learning_rate": 6.100781334002485e-06,
+      "loss": 0.3209,
+      "step": 5802
+    },
+    {
+      "epoch": 1.459507042253521,
+      "grad_norm": 0.3165169656276703,
+      "learning_rate": 6.099353940600634e-06,
+      "loss": 0.336,
+      "step": 5803
+    },
+    {
+      "epoch": 1.459758551307847,
+      "grad_norm": 0.36417675018310547,
+      "learning_rate": 6.097926453045996e-06,
+      "loss": 0.3508,
+      "step": 5804
+    },
+    {
+      "epoch": 1.460010060362173,
+      "grad_norm": 0.3556537926197052,
+      "learning_rate": 6.0964988714608255e-06,
+      "loss": 0.3644,
+      "step": 5805
+    },
+    {
+      "epoch": 1.4602615694164989,
+      "grad_norm": 0.3152537941932678,
+      "learning_rate": 6.0950711959673854e-06,
+      "loss": 0.3538,
+      "step": 5806
+    },
+    {
+      "epoch": 1.460513078470825,
+      "grad_norm": 0.3294949531555176,
+      "learning_rate": 6.093643426687949e-06,
+      "loss": 0.3665,
+      "step": 5807
+    },
+    {
+      "epoch": 1.460764587525151,
+      "grad_norm": 0.33261191844940186,
+      "learning_rate": 6.092215563744797e-06,
+      "loss": 0.3564,
+      "step": 5808
+    },
+    {
+      "epoch": 1.4610160965794767,
+      "grad_norm": 0.36204931139945984,
+      "learning_rate": 6.0907876072602126e-06,
+      "loss": 0.3704,
+      "step": 5809
+    },
+    {
+      "epoch": 1.4612676056338028,
+      "grad_norm": 0.34487199783325195,
+      "learning_rate": 6.0893595573564935e-06,
+      "loss": 0.3626,
+      "step": 5810
+    },
+    {
+      "epoch": 1.4615191146881288,
+      "grad_norm": 0.3406592607498169,
+      "learning_rate": 6.0879314141559434e-06,
+      "loss": 0.3619,
+      "step": 5811
+    },
+    {
+      "epoch": 1.4617706237424548,
+      "grad_norm": 0.3268572986125946,
+      "learning_rate": 6.086503177780874e-06,
+      "loss": 0.3383,
+      "step": 5812
+    },
+    {
+      "epoch": 1.4620221327967806,
+      "grad_norm": 0.33901724219322205,
+      "learning_rate": 6.085074848353604e-06,
+      "loss": 0.3209,
+      "step": 5813
+    },
+    {
+      "epoch": 1.4622736418511066,
+      "grad_norm": 0.2882567346096039,
+      "learning_rate": 6.083646425996462e-06,
+      "loss": 0.3496,
+      "step": 5814
+    },
+    {
+      "epoch": 1.4625251509054327,
+      "grad_norm": 0.30257371068000793,
+      "learning_rate": 6.082217910831784e-06,
+      "loss": 0.3685,
+      "step": 5815
+    },
+    {
+      "epoch": 1.4627766599597585,
+      "grad_norm": 0.34829580783843994,
+      "learning_rate": 6.080789302981911e-06,
+      "loss": 0.356,
+      "step": 5816
+    },
+    {
+      "epoch": 1.4630281690140845,
+      "grad_norm": 0.32623910903930664,
+      "learning_rate": 6.079360602569196e-06,
+      "loss": 0.3614,
+      "step": 5817
+    },
+    {
+      "epoch": 1.4632796780684105,
+      "grad_norm": 0.3189256191253662,
+      "learning_rate": 6.0779318097159965e-06,
+      "loss": 0.3708,
+      "step": 5818
+    },
+    {
+      "epoch": 1.4635311871227366,
+      "grad_norm": 0.32534220814704895,
+      "learning_rate": 6.076502924544683e-06,
+      "loss": 0.3314,
+      "step": 5819
+    },
+    {
+      "epoch": 1.4637826961770624,
+      "grad_norm": 0.3330930471420288,
+      "learning_rate": 6.075073947177628e-06,
+      "loss": 0.3503,
+      "step": 5820
+    },
+    {
+      "epoch": 1.4640342052313884,
+      "grad_norm": 0.32736390829086304,
+      "learning_rate": 6.073644877737215e-06,
+      "loss": 0.3712,
+      "step": 5821
+    },
+    {
+      "epoch": 1.4642857142857144,
+      "grad_norm": 0.3611101806163788,
+      "learning_rate": 6.072215716345835e-06,
+      "loss": 0.3876,
+      "step": 5822
+    },
+    {
+      "epoch": 1.4645372233400402,
+      "grad_norm": 0.33069029450416565,
+      "learning_rate": 6.070786463125885e-06,
+      "loss": 0.3498,
+      "step": 5823
+    },
+    {
+      "epoch": 1.4647887323943662,
+      "grad_norm": 0.32250040769577026,
+      "learning_rate": 6.069357118199775e-06,
+      "loss": 0.3651,
+      "step": 5824
+    },
+    {
+      "epoch": 1.4650402414486923,
+      "grad_norm": 0.3161625564098358,
+      "learning_rate": 6.067927681689917e-06,
+      "loss": 0.367,
+      "step": 5825
+    },
+    {
+      "epoch": 1.465291750503018,
+      "grad_norm": 0.3699602484703064,
+      "learning_rate": 6.066498153718735e-06,
+      "loss": 0.342,
+      "step": 5826
+    },
+    {
+      "epoch": 1.465543259557344,
+      "grad_norm": 0.33579832315444946,
+      "learning_rate": 6.065068534408657e-06,
+      "loss": 0.3465,
+      "step": 5827
+    },
+    {
+      "epoch": 1.4657947686116701,
+      "grad_norm": 0.3177899718284607,
+      "learning_rate": 6.063638823882123e-06,
+      "loss": 0.3512,
+      "step": 5828
+    },
+    {
+      "epoch": 1.466046277665996,
+      "grad_norm": 0.31834790110588074,
+      "learning_rate": 6.062209022261577e-06,
+      "loss": 0.3652,
+      "step": 5829
+    },
+    {
+      "epoch": 1.466297786720322,
+      "grad_norm": 0.30503425002098083,
+      "learning_rate": 6.060779129669474e-06,
+      "loss": 0.3319,
+      "step": 5830
+    },
+    {
+      "epoch": 1.466549295774648,
+      "grad_norm": 0.32431966066360474,
+      "learning_rate": 6.059349146228275e-06,
+      "loss": 0.35,
+      "step": 5831
+    },
+    {
+      "epoch": 1.4668008048289738,
+      "grad_norm": 0.3199666440486908,
+      "learning_rate": 6.057919072060448e-06,
+      "loss": 0.3537,
+      "step": 5832
+    },
+    {
+      "epoch": 1.4670523138832998,
+      "grad_norm": 0.3279325067996979,
+      "learning_rate": 6.0564889072884715e-06,
+      "loss": 0.3408,
+      "step": 5833
+    },
+    {
+      "epoch": 1.4673038229376258,
+      "grad_norm": 0.3161599338054657,
+      "learning_rate": 6.0550586520348285e-06,
+      "loss": 0.3576,
+      "step": 5834
+    },
+    {
+      "epoch": 1.4675553319919517,
+      "grad_norm": 0.32007861137390137,
+      "learning_rate": 6.053628306422014e-06,
+      "loss": 0.3436,
+      "step": 5835
+    },
+    {
+      "epoch": 1.4678068410462777,
+      "grad_norm": 0.3110705316066742,
+      "learning_rate": 6.052197870572525e-06,
+      "loss": 0.3723,
+      "step": 5836
+    },
+    {
+      "epoch": 1.4680583501006037,
+      "grad_norm": 0.33387455344200134,
+      "learning_rate": 6.0507673446088726e-06,
+      "loss": 0.3686,
+      "step": 5837
+    },
+    {
+      "epoch": 1.4683098591549295,
+      "grad_norm": 0.3183239996433258,
+      "learning_rate": 6.049336728653569e-06,
+      "loss": 0.367,
+      "step": 5838
+    },
+    {
+      "epoch": 1.4685613682092555,
+      "grad_norm": 0.3348864018917084,
+      "learning_rate": 6.0479060228291396e-06,
+      "loss": 0.3625,
+      "step": 5839
+    },
+    {
+      "epoch": 1.4688128772635816,
+      "grad_norm": 0.3361799120903015,
+      "learning_rate": 6.046475227258115e-06,
+      "loss": 0.3475,
+      "step": 5840
+    },
+    {
+      "epoch": 1.4690643863179074,
+      "grad_norm": 0.3200930655002594,
+      "learning_rate": 6.045044342063034e-06,
+      "loss": 0.3711,
+      "step": 5841
+    },
+    {
+      "epoch": 1.4693158953722334,
+      "grad_norm": 0.2997252345085144,
+      "learning_rate": 6.043613367366444e-06,
+      "loss": 0.3506,
+      "step": 5842
+    },
+    {
+      "epoch": 1.4695674044265594,
+      "grad_norm": 0.3714944124221802,
+      "learning_rate": 6.042182303290898e-06,
+      "loss": 0.3711,
+      "step": 5843
+    },
+    {
+      "epoch": 1.4698189134808852,
+      "grad_norm": 0.32264792919158936,
+      "learning_rate": 6.040751149958955e-06,
+      "loss": 0.3467,
+      "step": 5844
+    },
+    {
+      "epoch": 1.4700704225352113,
+      "grad_norm": 0.3230764865875244,
+      "learning_rate": 6.0393199074931886e-06,
+      "loss": 0.3556,
+      "step": 5845
+    },
+    {
+      "epoch": 1.4703219315895373,
+      "grad_norm": 0.31362685561180115,
+      "learning_rate": 6.037888576016174e-06,
+      "loss": 0.3421,
+      "step": 5846
+    },
+    {
+      "epoch": 1.470573440643863,
+      "grad_norm": 0.3665870130062103,
+      "learning_rate": 6.036457155650496e-06,
+      "loss": 0.3703,
+      "step": 5847
+    },
+    {
+      "epoch": 1.470824949698189,
+      "grad_norm": 0.327773779630661,
+      "learning_rate": 6.035025646518747e-06,
+      "loss": 0.357,
+      "step": 5848
+    },
+    {
+      "epoch": 1.4710764587525151,
+      "grad_norm": 0.32785624265670776,
+      "learning_rate": 6.033594048743525e-06,
+      "loss": 0.3694,
+      "step": 5849
+    },
+    {
+      "epoch": 1.471327967806841,
+      "grad_norm": 0.3332252502441406,
+      "learning_rate": 6.03216236244744e-06,
+      "loss": 0.3334,
+      "step": 5850
+    },
+    {
+      "epoch": 1.471579476861167,
+      "grad_norm": 0.3057732582092285,
+      "learning_rate": 6.030730587753106e-06,
+      "loss": 0.37,
+      "step": 5851
+    },
+    {
+      "epoch": 1.471830985915493,
+      "grad_norm": 0.28898757696151733,
+      "learning_rate": 6.0292987247831455e-06,
+      "loss": 0.3595,
+      "step": 5852
+    },
+    {
+      "epoch": 1.4720824949698188,
+      "grad_norm": 0.3072353005409241,
+      "learning_rate": 6.0278667736601885e-06,
+      "loss": 0.3623,
+      "step": 5853
+    },
+    {
+      "epoch": 1.4723340040241448,
+      "grad_norm": 0.32753127813339233,
+      "learning_rate": 6.026434734506872e-06,
+      "loss": 0.3604,
+      "step": 5854
+    },
+    {
+      "epoch": 1.4725855130784709,
+      "grad_norm": 0.33749762177467346,
+      "learning_rate": 6.025002607445842e-06,
+      "loss": 0.3586,
+      "step": 5855
+    },
+    {
+      "epoch": 1.4728370221327967,
+      "grad_norm": 0.3170812427997589,
+      "learning_rate": 6.023570392599751e-06,
+      "loss": 0.3351,
+      "step": 5856
+    },
+    {
+      "epoch": 1.4730885311871227,
+      "grad_norm": 0.3415549099445343,
+      "learning_rate": 6.02213809009126e-06,
+      "loss": 0.334,
+      "step": 5857
+    },
+    {
+      "epoch": 1.4733400402414487,
+      "grad_norm": 0.32296910881996155,
+      "learning_rate": 6.020705700043036e-06,
+      "loss": 0.3375,
+      "step": 5858
+    },
+    {
+      "epoch": 1.4735915492957745,
+      "grad_norm": 0.3227507770061493,
+      "learning_rate": 6.019273222577754e-06,
+      "loss": 0.3523,
+      "step": 5859
+    },
+    {
+      "epoch": 1.4738430583501005,
+      "grad_norm": 0.3127809166908264,
+      "learning_rate": 6.017840657818097e-06,
+      "loss": 0.3279,
+      "step": 5860
+    },
+    {
+      "epoch": 1.4740945674044266,
+      "grad_norm": 0.3176402449607849,
+      "learning_rate": 6.016408005886756e-06,
+      "loss": 0.3604,
+      "step": 5861
+    },
+    {
+      "epoch": 1.4743460764587526,
+      "grad_norm": 0.3369061350822449,
+      "learning_rate": 6.014975266906427e-06,
+      "loss": 0.3406,
+      "step": 5862
+    },
+    {
+      "epoch": 1.4745975855130784,
+      "grad_norm": 0.3053755462169647,
+      "learning_rate": 6.0135424409998156e-06,
+      "loss": 0.3587,
+      "step": 5863
+    },
+    {
+      "epoch": 1.4748490945674044,
+      "grad_norm": 0.30855026841163635,
+      "learning_rate": 6.0121095282896356e-06,
+      "loss": 0.3404,
+      "step": 5864
+    },
+    {
+      "epoch": 1.4751006036217305,
+      "grad_norm": 0.31946873664855957,
+      "learning_rate": 6.010676528898606e-06,
+      "loss": 0.334,
+      "step": 5865
+    },
+    {
+      "epoch": 1.4753521126760563,
+      "grad_norm": 0.34600701928138733,
+      "learning_rate": 6.009243442949454e-06,
+      "loss": 0.3573,
+      "step": 5866
+    },
+    {
+      "epoch": 1.4756036217303823,
+      "grad_norm": 0.32230931520462036,
+      "learning_rate": 6.007810270564916e-06,
+      "loss": 0.3549,
+      "step": 5867
+    },
+    {
+      "epoch": 1.4758551307847083,
+      "grad_norm": 0.3308599293231964,
+      "learning_rate": 6.006377011867732e-06,
+      "loss": 0.3662,
+      "step": 5868
+    },
+    {
+      "epoch": 1.4761066398390343,
+      "grad_norm": 0.3071610629558563,
+      "learning_rate": 6.004943666980654e-06,
+      "loss": 0.334,
+      "step": 5869
+    },
+    {
+      "epoch": 1.4763581488933601,
+      "grad_norm": 0.30614978075027466,
+      "learning_rate": 6.003510236026436e-06,
+      "loss": 0.3623,
+      "step": 5870
+    },
+    {
+      "epoch": 1.4766096579476862,
+      "grad_norm": 0.325398325920105,
+      "learning_rate": 6.0020767191278465e-06,
+      "loss": 0.3395,
+      "step": 5871
+    },
+    {
+      "epoch": 1.4768611670020122,
+      "grad_norm": 0.3325035572052002,
+      "learning_rate": 6.000643116407654e-06,
+      "loss": 0.3651,
+      "step": 5872
+    },
+    {
+      "epoch": 1.477112676056338,
+      "grad_norm": 0.30189478397369385,
+      "learning_rate": 5.999209427988638e-06,
+      "loss": 0.3646,
+      "step": 5873
+    },
+    {
+      "epoch": 1.477364185110664,
+      "grad_norm": 0.3175562918186188,
+      "learning_rate": 5.997775653993586e-06,
+      "loss": 0.3582,
+      "step": 5874
+    },
+    {
+      "epoch": 1.47761569416499,
+      "grad_norm": 0.3137342929840088,
+      "learning_rate": 5.996341794545292e-06,
+      "loss": 0.3407,
+      "step": 5875
+    },
+    {
+      "epoch": 1.4778672032193159,
+      "grad_norm": 0.3235321342945099,
+      "learning_rate": 5.9949078497665555e-06,
+      "loss": 0.3353,
+      "step": 5876
+    },
+    {
+      "epoch": 1.4781187122736419,
+      "grad_norm": 0.3290002942085266,
+      "learning_rate": 5.993473819780185e-06,
+      "loss": 0.3676,
+      "step": 5877
+    },
+    {
+      "epoch": 1.478370221327968,
+      "grad_norm": 0.3220723271369934,
+      "learning_rate": 5.992039704708998e-06,
+      "loss": 0.356,
+      "step": 5878
+    },
+    {
+      "epoch": 1.4786217303822937,
+      "grad_norm": 0.33870404958724976,
+      "learning_rate": 5.9906055046758174e-06,
+      "loss": 0.3572,
+      "step": 5879
+    },
+    {
+      "epoch": 1.4788732394366197,
+      "grad_norm": 0.3206806778907776,
+      "learning_rate": 5.989171219803471e-06,
+      "loss": 0.3593,
+      "step": 5880
+    },
+    {
+      "epoch": 1.4791247484909458,
+      "grad_norm": 0.35171839594841003,
+      "learning_rate": 5.987736850214798e-06,
+      "loss": 0.357,
+      "step": 5881
+    },
+    {
+      "epoch": 1.4793762575452716,
+      "grad_norm": 0.3402028977870941,
+      "learning_rate": 5.986302396032644e-06,
+      "loss": 0.3724,
+      "step": 5882
+    },
+    {
+      "epoch": 1.4796277665995976,
+      "grad_norm": 0.35204577445983887,
+      "learning_rate": 5.98486785737986e-06,
+      "loss": 0.35,
+      "step": 5883
+    },
+    {
+      "epoch": 1.4798792756539236,
+      "grad_norm": 0.3249172568321228,
+      "learning_rate": 5.983433234379306e-06,
+      "loss": 0.3657,
+      "step": 5884
+    },
+    {
+      "epoch": 1.4801307847082494,
+      "grad_norm": 0.3084418475627899,
+      "learning_rate": 5.9819985271538465e-06,
+      "loss": 0.3527,
+      "step": 5885
+    },
+    {
+      "epoch": 1.4803822937625755,
+      "grad_norm": 0.3349623382091522,
+      "learning_rate": 5.980563735826355e-06,
+      "loss": 0.3399,
+      "step": 5886
+    },
+    {
+      "epoch": 1.4806338028169015,
+      "grad_norm": 0.31369122862815857,
+      "learning_rate": 5.9791288605197175e-06,
+      "loss": 0.3405,
+      "step": 5887
+    },
+    {
+      "epoch": 1.4808853118712273,
+      "grad_norm": 0.3203195035457611,
+      "learning_rate": 5.9776939013568145e-06,
+      "loss": 0.3463,
+      "step": 5888
+    },
+    {
+      "epoch": 1.4811368209255533,
+      "grad_norm": 0.31735894083976746,
+      "learning_rate": 5.976258858460548e-06,
+      "loss": 0.3336,
+      "step": 5889
+    },
+    {
+      "epoch": 1.4813883299798793,
+      "grad_norm": 0.3369232416152954,
+      "learning_rate": 5.974823731953817e-06,
+      "loss": 0.3262,
+      "step": 5890
+    },
+    {
+      "epoch": 1.4816398390342052,
+      "grad_norm": 0.31136730313301086,
+      "learning_rate": 5.973388521959532e-06,
+      "loss": 0.3543,
+      "step": 5891
+    },
+    {
+      "epoch": 1.4818913480885312,
+      "grad_norm": 0.3076231777667999,
+      "learning_rate": 5.97195322860061e-06,
+      "loss": 0.3416,
+      "step": 5892
+    },
+    {
+      "epoch": 1.4821428571428572,
+      "grad_norm": 0.33634480834007263,
+      "learning_rate": 5.970517851999974e-06,
+      "loss": 0.345,
+      "step": 5893
+    },
+    {
+      "epoch": 1.482394366197183,
+      "grad_norm": 0.32039153575897217,
+      "learning_rate": 5.969082392280557e-06,
+      "loss": 0.3473,
+      "step": 5894
+    },
+    {
+      "epoch": 1.482645875251509,
+      "grad_norm": 0.3011680543422699,
+      "learning_rate": 5.967646849565294e-06,
+      "loss": 0.343,
+      "step": 5895
+    },
+    {
+      "epoch": 1.482897384305835,
+      "grad_norm": 0.33874598145484924,
+      "learning_rate": 5.966211223977132e-06,
+      "loss": 0.36,
+      "step": 5896
+    },
+    {
+      "epoch": 1.4831488933601609,
+      "grad_norm": 0.3423505127429962,
+      "learning_rate": 5.964775515639023e-06,
+      "loss": 0.3646,
+      "step": 5897
+    },
+    {
+      "epoch": 1.483400402414487,
+      "grad_norm": 0.33873575925827026,
+      "learning_rate": 5.963339724673928e-06,
+      "loss": 0.3458,
+      "step": 5898
+    },
+    {
+      "epoch": 1.483651911468813,
+      "grad_norm": 0.3190835416316986,
+      "learning_rate": 5.96190385120481e-06,
+      "loss": 0.3767,
+      "step": 5899
+    },
+    {
+      "epoch": 1.4839034205231387,
+      "grad_norm": 0.33720406889915466,
+      "learning_rate": 5.960467895354646e-06,
+      "loss": 0.3535,
+      "step": 5900
+    },
+    {
+      "epoch": 1.4841549295774648,
+      "grad_norm": 0.3207988440990448,
+      "learning_rate": 5.959031857246415e-06,
+      "loss": 0.3518,
+      "step": 5901
+    },
+    {
+      "epoch": 1.4844064386317908,
+      "grad_norm": 0.3778199255466461,
+      "learning_rate": 5.9575957370031065e-06,
+      "loss": 0.3516,
+      "step": 5902
+    },
+    {
+      "epoch": 1.4846579476861166,
+      "grad_norm": 0.3221379220485687,
+      "learning_rate": 5.956159534747713e-06,
+      "loss": 0.3216,
+      "step": 5903
+    },
+    {
+      "epoch": 1.4849094567404426,
+      "grad_norm": 0.296059787273407,
+      "learning_rate": 5.954723250603237e-06,
+      "loss": 0.3198,
+      "step": 5904
+    },
+    {
+      "epoch": 1.4851609657947686,
+      "grad_norm": 0.394718736410141,
+      "learning_rate": 5.953286884692688e-06,
+      "loss": 0.3623,
+      "step": 5905
+    },
+    {
+      "epoch": 1.4854124748490944,
+      "grad_norm": 0.3284577429294586,
+      "learning_rate": 5.9518504371390805e-06,
+      "loss": 0.3453,
+      "step": 5906
+    },
+    {
+      "epoch": 1.4856639839034205,
+      "grad_norm": 0.3131723999977112,
+      "learning_rate": 5.950413908065437e-06,
+      "loss": 0.3722,
+      "step": 5907
+    },
+    {
+      "epoch": 1.4859154929577465,
+      "grad_norm": 0.3282751142978668,
+      "learning_rate": 5.94897729759479e-06,
+      "loss": 0.3348,
+      "step": 5908
+    },
+    {
+      "epoch": 1.4861670020120723,
+      "grad_norm": 0.3348959684371948,
+      "learning_rate": 5.947540605850173e-06,
+      "loss": 0.3454,
+      "step": 5909
+    },
+    {
+      "epoch": 1.4864185110663983,
+      "grad_norm": 0.34094908833503723,
+      "learning_rate": 5.946103832954631e-06,
+      "loss": 0.3677,
+      "step": 5910
+    },
+    {
+      "epoch": 1.4866700201207244,
+      "grad_norm": 0.35626813769340515,
+      "learning_rate": 5.944666979031215e-06,
+      "loss": 0.3592,
+      "step": 5911
+    },
+    {
+      "epoch": 1.4869215291750504,
+      "grad_norm": 0.3323785960674286,
+      "learning_rate": 5.943230044202981e-06,
+      "loss": 0.3671,
+      "step": 5912
+    },
+    {
+      "epoch": 1.4871730382293762,
+      "grad_norm": 0.32840991020202637,
+      "learning_rate": 5.941793028592996e-06,
+      "loss": 0.3488,
+      "step": 5913
+    },
+    {
+      "epoch": 1.4874245472837022,
+      "grad_norm": 0.3254261314868927,
+      "learning_rate": 5.940355932324329e-06,
+      "loss": 0.343,
+      "step": 5914
+    },
+    {
+      "epoch": 1.4876760563380282,
+      "grad_norm": 0.3155461847782135,
+      "learning_rate": 5.938918755520059e-06,
+      "loss": 0.3398,
+      "step": 5915
+    },
+    {
+      "epoch": 1.487927565392354,
+      "grad_norm": 0.3168039619922638,
+      "learning_rate": 5.9374814983032724e-06,
+      "loss": 0.3664,
+      "step": 5916
+    },
+    {
+      "epoch": 1.48817907444668,
+      "grad_norm": 0.344101220369339,
+      "learning_rate": 5.9360441607970585e-06,
+      "loss": 0.3417,
+      "step": 5917
+    },
+    {
+      "epoch": 1.488430583501006,
+      "grad_norm": 0.3618612587451935,
+      "learning_rate": 5.93460674312452e-06,
+      "loss": 0.3684,
+      "step": 5918
+    },
+    {
+      "epoch": 1.4886820925553321,
+      "grad_norm": 0.32880085706710815,
+      "learning_rate": 5.933169245408761e-06,
+      "loss": 0.3381,
+      "step": 5919
+    },
+    {
+      "epoch": 1.488933601609658,
+      "grad_norm": 0.3657226860523224,
+      "learning_rate": 5.931731667772893e-06,
+      "loss": 0.3754,
+      "step": 5920
+    },
+    {
+      "epoch": 1.489185110663984,
+      "grad_norm": 0.3417291045188904,
+      "learning_rate": 5.930294010340035e-06,
+      "loss": 0.3586,
+      "step": 5921
+    },
+    {
+      "epoch": 1.48943661971831,
+      "grad_norm": 0.3205607235431671,
+      "learning_rate": 5.928856273233316e-06,
+      "loss": 0.3415,
+      "step": 5922
+    },
+    {
+      "epoch": 1.4896881287726358,
+      "grad_norm": 0.35127079486846924,
+      "learning_rate": 5.92741845657587e-06,
+      "loss": 0.3403,
+      "step": 5923
+    },
+    {
+      "epoch": 1.4899396378269618,
+      "grad_norm": 0.3201601207256317,
+      "learning_rate": 5.925980560490834e-06,
+      "loss": 0.3326,
+      "step": 5924
+    },
+    {
+      "epoch": 1.4901911468812878,
+      "grad_norm": 0.3274860978126526,
+      "learning_rate": 5.924542585101356e-06,
+      "loss": 0.3409,
+      "step": 5925
+    },
+    {
+      "epoch": 1.4904426559356136,
+      "grad_norm": 0.3243807852268219,
+      "learning_rate": 5.923104530530589e-06,
+      "loss": 0.3353,
+      "step": 5926
+    },
+    {
+      "epoch": 1.4906941649899397,
+      "grad_norm": 0.3296162486076355,
+      "learning_rate": 5.921666396901694e-06,
+      "loss": 0.3653,
+      "step": 5927
+    },
+    {
+      "epoch": 1.4909456740442657,
+      "grad_norm": 0.31820058822631836,
+      "learning_rate": 5.9202281843378385e-06,
+      "loss": 0.343,
+      "step": 5928
+    },
+    {
+      "epoch": 1.4911971830985915,
+      "grad_norm": 0.3398142457008362,
+      "learning_rate": 5.918789892962196e-06,
+      "loss": 0.3509,
+      "step": 5929
+    },
+    {
+      "epoch": 1.4914486921529175,
+      "grad_norm": 0.32767853140830994,
+      "learning_rate": 5.917351522897946e-06,
+      "loss": 0.3578,
+      "step": 5930
+    },
+    {
+      "epoch": 1.4917002012072436,
+      "grad_norm": 0.30963650345802307,
+      "learning_rate": 5.915913074268277e-06,
+      "loss": 0.3416,
+      "step": 5931
+    },
+    {
+      "epoch": 1.4919517102615694,
+      "grad_norm": 0.3078908920288086,
+      "learning_rate": 5.914474547196384e-06,
+      "loss": 0.3493,
+      "step": 5932
+    },
+    {
+      "epoch": 1.4922032193158954,
+      "grad_norm": 0.32465776801109314,
+      "learning_rate": 5.913035941805467e-06,
+      "loss": 0.3434,
+      "step": 5933
+    },
+    {
+      "epoch": 1.4924547283702214,
+      "grad_norm": 0.32990938425064087,
+      "learning_rate": 5.911597258218733e-06,
+      "loss": 0.3676,
+      "step": 5934
+    },
+    {
+      "epoch": 1.4927062374245472,
+      "grad_norm": 0.33850571513175964,
+      "learning_rate": 5.9101584965593975e-06,
+      "loss": 0.342,
+      "step": 5935
+    },
+    {
+      "epoch": 1.4929577464788732,
+      "grad_norm": 0.3495611846446991,
+      "learning_rate": 5.90871965695068e-06,
+      "loss": 0.3366,
+      "step": 5936
+    },
+    {
+      "epoch": 1.4932092555331993,
+      "grad_norm": 0.32349708676338196,
+      "learning_rate": 5.907280739515809e-06,
+      "loss": 0.3408,
+      "step": 5937
+    },
+    {
+      "epoch": 1.493460764587525,
+      "grad_norm": 0.3027772307395935,
+      "learning_rate": 5.905841744378019e-06,
+      "loss": 0.3582,
+      "step": 5938
+    },
+    {
+      "epoch": 1.493712273641851,
+      "grad_norm": 0.3265613913536072,
+      "learning_rate": 5.904402671660551e-06,
+      "loss": 0.3438,
+      "step": 5939
+    },
+    {
+      "epoch": 1.4939637826961771,
+      "grad_norm": 0.3158648908138275,
+      "learning_rate": 5.902963521486651e-06,
+      "loss": 0.3569,
+      "step": 5940
+    },
+    {
+      "epoch": 1.494215291750503,
+      "grad_norm": 0.3225020468235016,
+      "learning_rate": 5.901524293979575e-06,
+      "loss": 0.3403,
+      "step": 5941
+    },
+    {
+      "epoch": 1.494466800804829,
+      "grad_norm": 0.33680132031440735,
+      "learning_rate": 5.900084989262581e-06,
+      "loss": 0.3514,
+      "step": 5942
+    },
+    {
+      "epoch": 1.494718309859155,
+      "grad_norm": 0.30138635635375977,
+      "learning_rate": 5.898645607458941e-06,
+      "loss": 0.3465,
+      "step": 5943
+    },
+    {
+      "epoch": 1.4949698189134808,
+      "grad_norm": 0.3087744116783142,
+      "learning_rate": 5.897206148691925e-06,
+      "loss": 0.3362,
+      "step": 5944
+    },
+    {
+      "epoch": 1.4952213279678068,
+      "grad_norm": 0.3069364130496979,
+      "learning_rate": 5.895766613084817e-06,
+      "loss": 0.3122,
+      "step": 5945
+    },
+    {
+      "epoch": 1.4954728370221329,
+      "grad_norm": 0.289499968290329,
+      "learning_rate": 5.8943270007609026e-06,
+      "loss": 0.3384,
+      "step": 5946
+    },
+    {
+      "epoch": 1.4957243460764587,
+      "grad_norm": 0.3094853162765503,
+      "learning_rate": 5.892887311843474e-06,
+      "loss": 0.333,
+      "step": 5947
+    },
+    {
+      "epoch": 1.4959758551307847,
+      "grad_norm": 0.28878265619277954,
+      "learning_rate": 5.891447546455833e-06,
+      "loss": 0.32,
+      "step": 5948
+    },
+    {
+      "epoch": 1.4962273641851107,
+      "grad_norm": 0.32799869775772095,
+      "learning_rate": 5.890007704721288e-06,
+      "loss": 0.3462,
+      "step": 5949
+    },
+    {
+      "epoch": 1.4964788732394365,
+      "grad_norm": 0.30716651678085327,
+      "learning_rate": 5.88856778676315e-06,
+      "loss": 0.3626,
+      "step": 5950
+    },
+    {
+      "epoch": 1.4967303822937625,
+      "grad_norm": 0.31167250871658325,
+      "learning_rate": 5.88712779270474e-06,
+      "loss": 0.3524,
+      "step": 5951
+    },
+    {
+      "epoch": 1.4969818913480886,
+      "grad_norm": 0.31440672278404236,
+      "learning_rate": 5.885687722669384e-06,
+      "loss": 0.3277,
+      "step": 5952
+    },
+    {
+      "epoch": 1.4972334004024144,
+      "grad_norm": 0.28327929973602295,
+      "learning_rate": 5.884247576780416e-06,
+      "loss": 0.347,
+      "step": 5953
+    },
+    {
+      "epoch": 1.4974849094567404,
+      "grad_norm": 0.33180156350135803,
+      "learning_rate": 5.882807355161174e-06,
+      "loss": 0.3556,
+      "step": 5954
+    },
+    {
+      "epoch": 1.4977364185110664,
+      "grad_norm": 0.30268731713294983,
+      "learning_rate": 5.881367057935005e-06,
+      "loss": 0.337,
+      "step": 5955
+    },
+    {
+      "epoch": 1.4979879275653922,
+      "grad_norm": 0.32002800703048706,
+      "learning_rate": 5.879926685225264e-06,
+      "loss": 0.3529,
+      "step": 5956
+    },
+    {
+      "epoch": 1.4982394366197183,
+      "grad_norm": 0.31586262583732605,
+      "learning_rate": 5.878486237155304e-06,
+      "loss": 0.3336,
+      "step": 5957
+    },
+    {
+      "epoch": 1.4984909456740443,
+      "grad_norm": 0.3224563002586365,
+      "learning_rate": 5.877045713848495e-06,
+      "loss": 0.3587,
+      "step": 5958
+    },
+    {
+      "epoch": 1.49874245472837,
+      "grad_norm": 0.32255980372428894,
+      "learning_rate": 5.875605115428207e-06,
+      "loss": 0.3308,
+      "step": 5959
+    },
+    {
+      "epoch": 1.4989939637826961,
+      "grad_norm": 0.3053814172744751,
+      "learning_rate": 5.874164442017819e-06,
+      "loss": 0.3301,
+      "step": 5960
+    },
+    {
+      "epoch": 1.4992454728370221,
+      "grad_norm": 0.32673344016075134,
+      "learning_rate": 5.872723693740715e-06,
+      "loss": 0.3345,
+      "step": 5961
+    },
+    {
+      "epoch": 1.4994969818913482,
+      "grad_norm": 0.3168867230415344,
+      "learning_rate": 5.871282870720286e-06,
+      "loss": 0.3376,
+      "step": 5962
+    },
+    {
+      "epoch": 1.499748490945674,
+      "grad_norm": 0.3028486371040344,
+      "learning_rate": 5.869841973079931e-06,
+      "loss": 0.3365,
+      "step": 5963
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.34289392828941345,
+      "learning_rate": 5.868401000943051e-06,
+      "loss": 0.348,
+      "step": 5964
+    },
+    {
+      "epoch": 1.5002515090543258,
+      "grad_norm": 0.3447049856185913,
+      "learning_rate": 5.866959954433058e-06,
+      "loss": 0.3576,
+      "step": 5965
+    },
+    {
+      "epoch": 1.500503018108652,
+      "grad_norm": 0.3279689848423004,
+      "learning_rate": 5.86551883367337e-06,
+      "loss": 0.3803,
+      "step": 5966
+    },
+    {
+      "epoch": 1.5007545271629779,
+      "grad_norm": 0.3094026744365692,
+      "learning_rate": 5.8640776387874085e-06,
+      "loss": 0.351,
+      "step": 5967
+    },
+    {
+      "epoch": 1.5010060362173037,
+      "grad_norm": 0.3381515443325043,
+      "learning_rate": 5.8626363698986025e-06,
+      "loss": 0.3711,
+      "step": 5968
+    },
+    {
+      "epoch": 1.50125754527163,
+      "grad_norm": 0.3412519693374634,
+      "learning_rate": 5.861195027130388e-06,
+      "loss": 0.3376,
+      "step": 5969
+    },
+    {
+      "epoch": 1.5015090543259557,
+      "grad_norm": 0.33772408962249756,
+      "learning_rate": 5.859753610606207e-06,
+      "loss": 0.3629,
+      "step": 5970
+    },
+    {
+      "epoch": 1.5017605633802817,
+      "grad_norm": 0.3265193700790405,
+      "learning_rate": 5.858312120449507e-06,
+      "loss": 0.347,
+      "step": 5971
+    },
+    {
+      "epoch": 1.5020120724346078,
+      "grad_norm": 0.3511604368686676,
+      "learning_rate": 5.856870556783746e-06,
+      "loss": 0.3296,
+      "step": 5972
+    },
+    {
+      "epoch": 1.5022635814889336,
+      "grad_norm": 0.3105302155017853,
+      "learning_rate": 5.85542891973238e-06,
+      "loss": 0.3374,
+      "step": 5973
+    },
+    {
+      "epoch": 1.5025150905432596,
+      "grad_norm": 0.34747567772865295,
+      "learning_rate": 5.85398720941888e-06,
+      "loss": 0.3545,
+      "step": 5974
+    },
+    {
+      "epoch": 1.5027665995975856,
+      "grad_norm": 0.3080263137817383,
+      "learning_rate": 5.852545425966717e-06,
+      "loss": 0.3281,
+      "step": 5975
+    },
+    {
+      "epoch": 1.5030181086519114,
+      "grad_norm": 0.34550175070762634,
+      "learning_rate": 5.851103569499372e-06,
+      "loss": 0.3526,
+      "step": 5976
+    },
+    {
+      "epoch": 1.5032696177062375,
+      "grad_norm": 0.3092076778411865,
+      "learning_rate": 5.849661640140332e-06,
+      "loss": 0.3381,
+      "step": 5977
+    },
+    {
+      "epoch": 1.5035211267605635,
+      "grad_norm": 0.31573188304901123,
+      "learning_rate": 5.848219638013086e-06,
+      "loss": 0.3599,
+      "step": 5978
+    },
+    {
+      "epoch": 1.5037726358148893,
+      "grad_norm": 0.2985522449016571,
+      "learning_rate": 5.846777563241136e-06,
+      "loss": 0.3504,
+      "step": 5979
+    },
+    {
+      "epoch": 1.5040241448692153,
+      "grad_norm": 0.2984470725059509,
+      "learning_rate": 5.845335415947985e-06,
+      "loss": 0.334,
+      "step": 5980
+    },
+    {
+      "epoch": 1.5042756539235413,
+      "grad_norm": 0.3081437349319458,
+      "learning_rate": 5.8438931962571435e-06,
+      "loss": 0.3274,
+      "step": 5981
+    },
+    {
+      "epoch": 1.5045271629778671,
+      "grad_norm": 0.31499478220939636,
+      "learning_rate": 5.842450904292128e-06,
+      "loss": 0.3468,
+      "step": 5982
+    },
+    {
+      "epoch": 1.5047786720321932,
+      "grad_norm": 0.30446550250053406,
+      "learning_rate": 5.841008540176465e-06,
+      "loss": 0.3299,
+      "step": 5983
+    },
+    {
+      "epoch": 1.5050301810865192,
+      "grad_norm": 0.2896031141281128,
+      "learning_rate": 5.83956610403368e-06,
+      "loss": 0.3371,
+      "step": 5984
+    },
+    {
+      "epoch": 1.505281690140845,
+      "grad_norm": 0.3181334435939789,
+      "learning_rate": 5.838123595987312e-06,
+      "loss": 0.3294,
+      "step": 5985
+    },
+    {
+      "epoch": 1.505533199195171,
+      "grad_norm": 0.32969263195991516,
+      "learning_rate": 5.8366810161609e-06,
+      "loss": 0.3314,
+      "step": 5986
+    },
+    {
+      "epoch": 1.505784708249497,
+      "grad_norm": 0.3354569375514984,
+      "learning_rate": 5.835238364677994e-06,
+      "loss": 0.3647,
+      "step": 5987
+    },
+    {
+      "epoch": 1.5060362173038229,
+      "grad_norm": 0.30361291766166687,
+      "learning_rate": 5.8337956416621465e-06,
+      "loss": 0.3302,
+      "step": 5988
+    },
+    {
+      "epoch": 1.506287726358149,
+      "grad_norm": 0.3423859775066376,
+      "learning_rate": 5.832352847236919e-06,
+      "loss": 0.3275,
+      "step": 5989
+    },
+    {
+      "epoch": 1.506539235412475,
+      "grad_norm": 0.3418254554271698,
+      "learning_rate": 5.830909981525879e-06,
+      "loss": 0.3161,
+      "step": 5990
+    },
+    {
+      "epoch": 1.5067907444668007,
+      "grad_norm": 0.33632370829582214,
+      "learning_rate": 5.829467044652595e-06,
+      "loss": 0.3405,
+      "step": 5991
+    },
+    {
+      "epoch": 1.5070422535211268,
+      "grad_norm": 0.31483152508735657,
+      "learning_rate": 5.828024036740649e-06,
+      "loss": 0.3588,
+      "step": 5992
+    },
+    {
+      "epoch": 1.5072937625754528,
+      "grad_norm": 0.3332071900367737,
+      "learning_rate": 5.826580957913624e-06,
+      "loss": 0.3365,
+      "step": 5993
+    },
+    {
+      "epoch": 1.5075452716297786,
+      "grad_norm": 0.3333156704902649,
+      "learning_rate": 5.825137808295111e-06,
+      "loss": 0.3691,
+      "step": 5994
+    },
+    {
+      "epoch": 1.5077967806841046,
+      "grad_norm": 0.32617276906967163,
+      "learning_rate": 5.823694588008707e-06,
+      "loss": 0.3303,
+      "step": 5995
+    },
+    {
+      "epoch": 1.5080482897384306,
+      "grad_norm": 0.3059340715408325,
+      "learning_rate": 5.822251297178014e-06,
+      "loss": 0.3357,
+      "step": 5996
+    },
+    {
+      "epoch": 1.5082997987927564,
+      "grad_norm": 0.3015100359916687,
+      "learning_rate": 5.82080793592664e-06,
+      "loss": 0.3366,
+      "step": 5997
+    },
+    {
+      "epoch": 1.5085513078470825,
+      "grad_norm": 0.3392370939254761,
+      "learning_rate": 5.819364504378203e-06,
+      "loss": 0.3774,
+      "step": 5998
+    },
+    {
+      "epoch": 1.5088028169014085,
+      "grad_norm": 0.34105855226516724,
+      "learning_rate": 5.817921002656323e-06,
+      "loss": 0.3507,
+      "step": 5999
+    },
+    {
+      "epoch": 1.5090543259557343,
+      "grad_norm": 0.3226035237312317,
+      "learning_rate": 5.816477430884625e-06,
+      "loss": 0.3672,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5093058350100603,
+      "grad_norm": 0.32897132635116577,
+      "learning_rate": 5.815033789186743e-06,
+      "loss": 0.354,
+      "step": 6001
+    },
+    {
+      "epoch": 1.5095573440643864,
+      "grad_norm": 0.30205658078193665,
+      "learning_rate": 5.813590077686315e-06,
+      "loss": 0.3416,
+      "step": 6002
+    },
+    {
+      "epoch": 1.5098088531187122,
+      "grad_norm": 0.33135828375816345,
+      "learning_rate": 5.812146296506987e-06,
+      "loss": 0.3654,
+      "step": 6003
+    },
+    {
+      "epoch": 1.5100603621730382,
+      "grad_norm": 0.3252605199813843,
+      "learning_rate": 5.810702445772411e-06,
+      "loss": 0.3617,
+      "step": 6004
+    },
+    {
+      "epoch": 1.5103118712273642,
+      "grad_norm": 0.3258083760738373,
+      "learning_rate": 5.809258525606241e-06,
+      "loss": 0.3513,
+      "step": 6005
+    },
+    {
+      "epoch": 1.51056338028169,
+      "grad_norm": 0.3420291244983673,
+      "learning_rate": 5.807814536132141e-06,
+      "loss": 0.3803,
+      "step": 6006
+    },
+    {
+      "epoch": 1.510814889336016,
+      "grad_norm": 0.317022442817688,
+      "learning_rate": 5.806370477473778e-06,
+      "loss": 0.35,
+      "step": 6007
+    },
+    {
+      "epoch": 1.511066398390342,
+      "grad_norm": 0.30522531270980835,
+      "learning_rate": 5.80492634975483e-06,
+      "loss": 0.3376,
+      "step": 6008
+    },
+    {
+      "epoch": 1.5113179074446679,
+      "grad_norm": 0.3109252452850342,
+      "learning_rate": 5.803482153098976e-06,
+      "loss": 0.3218,
+      "step": 6009
+    },
+    {
+      "epoch": 1.5115694164989941,
+      "grad_norm": 0.3221626281738281,
+      "learning_rate": 5.802037887629902e-06,
+      "loss": 0.3244,
+      "step": 6010
+    },
+    {
+      "epoch": 1.51182092555332,
+      "grad_norm": 0.31988272070884705,
+      "learning_rate": 5.800593553471301e-06,
+      "loss": 0.3714,
+      "step": 6011
+    },
+    {
+      "epoch": 1.5120724346076457,
+      "grad_norm": 0.298286497592926,
+      "learning_rate": 5.79914915074687e-06,
+      "loss": 0.3457,
+      "step": 6012
+    },
+    {
+      "epoch": 1.512323943661972,
+      "grad_norm": 0.3138798773288727,
+      "learning_rate": 5.797704679580313e-06,
+      "loss": 0.3422,
+      "step": 6013
+    },
+    {
+      "epoch": 1.5125754527162978,
+      "grad_norm": 0.307769238948822,
+      "learning_rate": 5.796260140095343e-06,
+      "loss": 0.3435,
+      "step": 6014
+    },
+    {
+      "epoch": 1.5128269617706236,
+      "grad_norm": 0.30271801352500916,
+      "learning_rate": 5.79481553241567e-06,
+      "loss": 0.3565,
+      "step": 6015
+    },
+    {
+      "epoch": 1.5130784708249498,
+      "grad_norm": 0.31750625371932983,
+      "learning_rate": 5.793370856665021e-06,
+      "loss": 0.3732,
+      "step": 6016
+    },
+    {
+      "epoch": 1.5133299798792756,
+      "grad_norm": 0.3040974736213684,
+      "learning_rate": 5.791926112967122e-06,
+      "loss": 0.3527,
+      "step": 6017
+    },
+    {
+      "epoch": 1.5135814889336014,
+      "grad_norm": 0.2947345972061157,
+      "learning_rate": 5.790481301445704e-06,
+      "loss": 0.3503,
+      "step": 6018
+    },
+    {
+      "epoch": 1.5138329979879277,
+      "grad_norm": 0.3347560167312622,
+      "learning_rate": 5.789036422224508e-06,
+      "loss": 0.3437,
+      "step": 6019
+    },
+    {
+      "epoch": 1.5140845070422535,
+      "grad_norm": 0.2965729236602783,
+      "learning_rate": 5.78759147542728e-06,
+      "loss": 0.3459,
+      "step": 6020
+    },
+    {
+      "epoch": 1.5143360160965795,
+      "grad_norm": 0.2949177622795105,
+      "learning_rate": 5.786146461177769e-06,
+      "loss": 0.3359,
+      "step": 6021
+    },
+    {
+      "epoch": 1.5145875251509056,
+      "grad_norm": 0.3233572542667389,
+      "learning_rate": 5.7847013795997306e-06,
+      "loss": 0.376,
+      "step": 6022
+    },
+    {
+      "epoch": 1.5148390342052314,
+      "grad_norm": 0.3046778440475464,
+      "learning_rate": 5.7832562308169286e-06,
+      "loss": 0.339,
+      "step": 6023
+    },
+    {
+      "epoch": 1.5150905432595574,
+      "grad_norm": 0.2932363450527191,
+      "learning_rate": 5.78181101495313e-06,
+      "loss": 0.3297,
+      "step": 6024
+    },
+    {
+      "epoch": 1.5153420523138834,
+      "grad_norm": 0.29733753204345703,
+      "learning_rate": 5.78036573213211e-06,
+      "loss": 0.3531,
+      "step": 6025
+    },
+    {
+      "epoch": 1.5155935613682092,
+      "grad_norm": 0.2947341501712799,
+      "learning_rate": 5.778920382477647e-06,
+      "loss": 0.3486,
+      "step": 6026
+    },
+    {
+      "epoch": 1.5158450704225352,
+      "grad_norm": 0.3160892724990845,
+      "learning_rate": 5.7774749661135255e-06,
+      "loss": 0.3706,
+      "step": 6027
+    },
+    {
+      "epoch": 1.5160965794768613,
+      "grad_norm": 0.3119431734085083,
+      "learning_rate": 5.776029483163538e-06,
+      "loss": 0.3675,
+      "step": 6028
+    },
+    {
+      "epoch": 1.516348088531187,
+      "grad_norm": 0.30674445629119873,
+      "learning_rate": 5.774583933751477e-06,
+      "loss": 0.3458,
+      "step": 6029
+    },
+    {
+      "epoch": 1.516599597585513,
+      "grad_norm": 0.3203643560409546,
+      "learning_rate": 5.773138318001151e-06,
+      "loss": 0.3581,
+      "step": 6030
+    },
+    {
+      "epoch": 1.5168511066398391,
+      "grad_norm": 0.3252681493759155,
+      "learning_rate": 5.771692636036365e-06,
+      "loss": 0.3449,
+      "step": 6031
+    },
+    {
+      "epoch": 1.517102615694165,
+      "grad_norm": 0.30788522958755493,
+      "learning_rate": 5.77024688798093e-06,
+      "loss": 0.3391,
+      "step": 6032
+    },
+    {
+      "epoch": 1.517354124748491,
+      "grad_norm": 0.3264254331588745,
+      "learning_rate": 5.76880107395867e-06,
+      "loss": 0.3492,
+      "step": 6033
+    },
+    {
+      "epoch": 1.517605633802817,
+      "grad_norm": 0.31012457609176636,
+      "learning_rate": 5.767355194093407e-06,
+      "loss": 0.3388,
+      "step": 6034
+    },
+    {
+      "epoch": 1.5178571428571428,
+      "grad_norm": 0.2838296592235565,
+      "learning_rate": 5.765909248508972e-06,
+      "loss": 0.344,
+      "step": 6035
+    },
+    {
+      "epoch": 1.5181086519114688,
+      "grad_norm": 0.29018688201904297,
+      "learning_rate": 5.764463237329201e-06,
+      "loss": 0.3587,
+      "step": 6036
+    },
+    {
+      "epoch": 1.5183601609657948,
+      "grad_norm": 0.3031754195690155,
+      "learning_rate": 5.7630171606779365e-06,
+      "loss": 0.3667,
+      "step": 6037
+    },
+    {
+      "epoch": 1.5186116700201207,
+      "grad_norm": 0.3265339732170105,
+      "learning_rate": 5.761571018679025e-06,
+      "loss": 0.3765,
+      "step": 6038
+    },
+    {
+      "epoch": 1.5188631790744467,
+      "grad_norm": 0.3091321289539337,
+      "learning_rate": 5.760124811456322e-06,
+      "loss": 0.3388,
+      "step": 6039
+    },
+    {
+      "epoch": 1.5191146881287727,
+      "grad_norm": 0.32367005944252014,
+      "learning_rate": 5.758678539133682e-06,
+      "loss": 0.342,
+      "step": 6040
+    },
+    {
+      "epoch": 1.5193661971830985,
+      "grad_norm": 0.3425564765930176,
+      "learning_rate": 5.757232201834973e-06,
+      "loss": 0.3465,
+      "step": 6041
+    },
+    {
+      "epoch": 1.5196177062374245,
+      "grad_norm": 0.28837791085243225,
+      "learning_rate": 5.755785799684063e-06,
+      "loss": 0.3377,
+      "step": 6042
+    },
+    {
+      "epoch": 1.5198692152917506,
+      "grad_norm": 0.31125667691230774,
+      "learning_rate": 5.754339332804826e-06,
+      "loss": 0.3478,
+      "step": 6043
+    },
+    {
+      "epoch": 1.5201207243460764,
+      "grad_norm": 0.342917263507843,
+      "learning_rate": 5.752892801321146e-06,
+      "loss": 0.3409,
+      "step": 6044
+    },
+    {
+      "epoch": 1.5203722334004024,
+      "grad_norm": 0.3406009376049042,
+      "learning_rate": 5.751446205356906e-06,
+      "loss": 0.3429,
+      "step": 6045
+    },
+    {
+      "epoch": 1.5206237424547284,
+      "grad_norm": 0.32107996940612793,
+      "learning_rate": 5.749999545036001e-06,
+      "loss": 0.332,
+      "step": 6046
+    },
+    {
+      "epoch": 1.5208752515090542,
+      "grad_norm": 0.31515464186668396,
+      "learning_rate": 5.7485528204823275e-06,
+      "loss": 0.3561,
+      "step": 6047
+    },
+    {
+      "epoch": 1.5211267605633803,
+      "grad_norm": 0.2875670790672302,
+      "learning_rate": 5.7471060318197856e-06,
+      "loss": 0.3564,
+      "step": 6048
+    },
+    {
+      "epoch": 1.5213782696177063,
+      "grad_norm": 0.3263548016548157,
+      "learning_rate": 5.7456591791722875e-06,
+      "loss": 0.3738,
+      "step": 6049
+    },
+    {
+      "epoch": 1.521629778672032,
+      "grad_norm": 0.30848369002342224,
+      "learning_rate": 5.744212262663745e-06,
+      "loss": 0.351,
+      "step": 6050
+    },
+    {
+      "epoch": 1.5218812877263581,
+      "grad_norm": 0.29957401752471924,
+      "learning_rate": 5.742765282418077e-06,
+      "loss": 0.3695,
+      "step": 6051
+    },
+    {
+      "epoch": 1.5221327967806841,
+      "grad_norm": 0.30603325366973877,
+      "learning_rate": 5.74131823855921e-06,
+      "loss": 0.3635,
+      "step": 6052
+    },
+    {
+      "epoch": 1.52238430583501,
+      "grad_norm": 0.31580379605293274,
+      "learning_rate": 5.739871131211074e-06,
+      "loss": 0.3334,
+      "step": 6053
+    },
+    {
+      "epoch": 1.522635814889336,
+      "grad_norm": 0.31090569496154785,
+      "learning_rate": 5.738423960497604e-06,
+      "loss": 0.3362,
+      "step": 6054
+    },
+    {
+      "epoch": 1.522887323943662,
+      "grad_norm": 0.33442890644073486,
+      "learning_rate": 5.736976726542742e-06,
+      "loss": 0.363,
+      "step": 6055
+    },
+    {
+      "epoch": 1.5231388329979878,
+      "grad_norm": 0.3184659481048584,
+      "learning_rate": 5.735529429470433e-06,
+      "loss": 0.3529,
+      "step": 6056
+    },
+    {
+      "epoch": 1.5233903420523138,
+      "grad_norm": 0.3236134946346283,
+      "learning_rate": 5.734082069404631e-06,
+      "loss": 0.3506,
+      "step": 6057
+    },
+    {
+      "epoch": 1.5236418511066399,
+      "grad_norm": 0.29839229583740234,
+      "learning_rate": 5.732634646469291e-06,
+      "loss": 0.3242,
+      "step": 6058
+    },
+    {
+      "epoch": 1.5238933601609657,
+      "grad_norm": 0.30354323983192444,
+      "learning_rate": 5.731187160788377e-06,
+      "loss": 0.3657,
+      "step": 6059
+    },
+    {
+      "epoch": 1.524144869215292,
+      "grad_norm": 0.3145615756511688,
+      "learning_rate": 5.729739612485857e-06,
+      "loss": 0.3592,
+      "step": 6060
+    },
+    {
+      "epoch": 1.5243963782696177,
+      "grad_norm": 0.322421133518219,
+      "learning_rate": 5.728292001685706e-06,
+      "loss": 0.3392,
+      "step": 6061
+    },
+    {
+      "epoch": 1.5246478873239435,
+      "grad_norm": 0.3052886128425598,
+      "learning_rate": 5.726844328511898e-06,
+      "loss": 0.3416,
+      "step": 6062
+    },
+    {
+      "epoch": 1.5248993963782698,
+      "grad_norm": 0.2889304459095001,
+      "learning_rate": 5.725396593088423e-06,
+      "loss": 0.3424,
+      "step": 6063
+    },
+    {
+      "epoch": 1.5251509054325956,
+      "grad_norm": 0.3219023048877716,
+      "learning_rate": 5.723948795539267e-06,
+      "loss": 0.3305,
+      "step": 6064
+    },
+    {
+      "epoch": 1.5254024144869214,
+      "grad_norm": 0.3205178678035736,
+      "learning_rate": 5.722500935988425e-06,
+      "loss": 0.3336,
+      "step": 6065
+    },
+    {
+      "epoch": 1.5256539235412476,
+      "grad_norm": 0.3172166347503662,
+      "learning_rate": 5.721053014559898e-06,
+      "loss": 0.3496,
+      "step": 6066
+    },
+    {
+      "epoch": 1.5259054325955734,
+      "grad_norm": 0.30717623233795166,
+      "learning_rate": 5.719605031377693e-06,
+      "loss": 0.3197,
+      "step": 6067
+    },
+    {
+      "epoch": 1.5261569416498992,
+      "grad_norm": 0.33775895833969116,
+      "learning_rate": 5.718156986565817e-06,
+      "loss": 0.3463,
+      "step": 6068
+    },
+    {
+      "epoch": 1.5264084507042255,
+      "grad_norm": 0.31737133860588074,
+      "learning_rate": 5.71670888024829e-06,
+      "loss": 0.3468,
+      "step": 6069
+    },
+    {
+      "epoch": 1.5266599597585513,
+      "grad_norm": 0.33389195799827576,
+      "learning_rate": 5.715260712549129e-06,
+      "loss": 0.364,
+      "step": 6070
+    },
+    {
+      "epoch": 1.5269114688128773,
+      "grad_norm": 0.32125452160835266,
+      "learning_rate": 5.713812483592364e-06,
+      "loss": 0.331,
+      "step": 6071
+    },
+    {
+      "epoch": 1.5271629778672033,
+      "grad_norm": 0.319833904504776,
+      "learning_rate": 5.712364193502024e-06,
+      "loss": 0.3332,
+      "step": 6072
+    },
+    {
+      "epoch": 1.5274144869215291,
+      "grad_norm": 0.3113959729671478,
+      "learning_rate": 5.710915842402147e-06,
+      "loss": 0.3545,
+      "step": 6073
+    },
+    {
+      "epoch": 1.5276659959758552,
+      "grad_norm": 0.31851494312286377,
+      "learning_rate": 5.7094674304167766e-06,
+      "loss": 0.3371,
+      "step": 6074
+    },
+    {
+      "epoch": 1.5279175050301812,
+      "grad_norm": 0.3372229337692261,
+      "learning_rate": 5.708018957669959e-06,
+      "loss": 0.3374,
+      "step": 6075
+    },
+    {
+      "epoch": 1.528169014084507,
+      "grad_norm": 0.3625347316265106,
+      "learning_rate": 5.706570424285747e-06,
+      "loss": 0.3519,
+      "step": 6076
+    },
+    {
+      "epoch": 1.528420523138833,
+      "grad_norm": 0.31328415870666504,
+      "learning_rate": 5.705121830388199e-06,
+      "loss": 0.3275,
+      "step": 6077
+    },
+    {
+      "epoch": 1.528672032193159,
+      "grad_norm": 0.36509522795677185,
+      "learning_rate": 5.703673176101377e-06,
+      "loss": 0.3426,
+      "step": 6078
+    },
+    {
+      "epoch": 1.5289235412474849,
+      "grad_norm": 0.3331649601459503,
+      "learning_rate": 5.702224461549351e-06,
+      "loss": 0.348,
+      "step": 6079
+    },
+    {
+      "epoch": 1.529175050301811,
+      "grad_norm": 0.33841848373413086,
+      "learning_rate": 5.700775686856192e-06,
+      "loss": 0.3439,
+      "step": 6080
+    },
+    {
+      "epoch": 1.529426559356137,
+      "grad_norm": 0.35388943552970886,
+      "learning_rate": 5.699326852145979e-06,
+      "loss": 0.3504,
+      "step": 6081
+    },
+    {
+      "epoch": 1.5296780684104627,
+      "grad_norm": 0.31031563878059387,
+      "learning_rate": 5.697877957542795e-06,
+      "loss": 0.3449,
+      "step": 6082
+    },
+    {
+      "epoch": 1.5299295774647887,
+      "grad_norm": 0.34058165550231934,
+      "learning_rate": 5.696429003170732e-06,
+      "loss": 0.347,
+      "step": 6083
+    },
+    {
+      "epoch": 1.5301810865191148,
+      "grad_norm": 0.30972498655319214,
+      "learning_rate": 5.69497998915388e-06,
+      "loss": 0.3401,
+      "step": 6084
+    },
+    {
+      "epoch": 1.5304325955734406,
+      "grad_norm": 0.32734933495521545,
+      "learning_rate": 5.693530915616341e-06,
+      "loss": 0.3593,
+      "step": 6085
+    },
+    {
+      "epoch": 1.5306841046277666,
+      "grad_norm": 0.34089869260787964,
+      "learning_rate": 5.6920817826822176e-06,
+      "loss": 0.3545,
+      "step": 6086
+    },
+    {
+      "epoch": 1.5309356136820926,
+      "grad_norm": 0.3333452641963959,
+      "learning_rate": 5.690632590475618e-06,
+      "loss": 0.3751,
+      "step": 6087
+    },
+    {
+      "epoch": 1.5311871227364184,
+      "grad_norm": 0.3344314992427826,
+      "learning_rate": 5.6891833391206585e-06,
+      "loss": 0.3573,
+      "step": 6088
+    },
+    {
+      "epoch": 1.5314386317907445,
+      "grad_norm": 0.2930935323238373,
+      "learning_rate": 5.687734028741459e-06,
+      "loss": 0.3553,
+      "step": 6089
+    },
+    {
+      "epoch": 1.5316901408450705,
+      "grad_norm": 0.32319191098213196,
+      "learning_rate": 5.68628465946214e-06,
+      "loss": 0.3395,
+      "step": 6090
+    },
+    {
+      "epoch": 1.5319416498993963,
+      "grad_norm": 0.30757129192352295,
+      "learning_rate": 5.684835231406835e-06,
+      "loss": 0.3526,
+      "step": 6091
+    },
+    {
+      "epoch": 1.5321931589537223,
+      "grad_norm": 0.31132009625434875,
+      "learning_rate": 5.683385744699675e-06,
+      "loss": 0.3385,
+      "step": 6092
+    },
+    {
+      "epoch": 1.5324446680080483,
+      "grad_norm": 0.31980809569358826,
+      "learning_rate": 5.681936199464803e-06,
+      "loss": 0.3532,
+      "step": 6093
+    },
+    {
+      "epoch": 1.5326961770623742,
+      "grad_norm": 0.29808422923088074,
+      "learning_rate": 5.680486595826361e-06,
+      "loss": 0.3306,
+      "step": 6094
+    },
+    {
+      "epoch": 1.5329476861167002,
+      "grad_norm": 0.3331838548183441,
+      "learning_rate": 5.6790369339085e-06,
+      "loss": 0.3695,
+      "step": 6095
+    },
+    {
+      "epoch": 1.5331991951710262,
+      "grad_norm": 0.3027997612953186,
+      "learning_rate": 5.677587213835372e-06,
+      "loss": 0.3521,
+      "step": 6096
+    },
+    {
+      "epoch": 1.533450704225352,
+      "grad_norm": 0.3264061212539673,
+      "learning_rate": 5.676137435731139e-06,
+      "loss": 0.3425,
+      "step": 6097
+    },
+    {
+      "epoch": 1.533702213279678,
+      "grad_norm": 0.32032856345176697,
+      "learning_rate": 5.674687599719963e-06,
+      "loss": 0.343,
+      "step": 6098
+    },
+    {
+      "epoch": 1.533953722334004,
+      "grad_norm": 0.33043432235717773,
+      "learning_rate": 5.673237705926018e-06,
+      "loss": 0.3522,
+      "step": 6099
+    },
+    {
+      "epoch": 1.5342052313883299,
+      "grad_norm": 0.30190953612327576,
+      "learning_rate": 5.6717877544734735e-06,
+      "loss": 0.3511,
+      "step": 6100
+    },
+    {
+      "epoch": 1.534456740442656,
+      "grad_norm": 0.2926798462867737,
+      "learning_rate": 5.670337745486511e-06,
+      "loss": 0.3662,
+      "step": 6101
+    },
+    {
+      "epoch": 1.534708249496982,
+      "grad_norm": 0.289917528629303,
+      "learning_rate": 5.668887679089314e-06,
+      "loss": 0.3499,
+      "step": 6102
+    },
+    {
+      "epoch": 1.5349597585513077,
+      "grad_norm": 0.3212122619152069,
+      "learning_rate": 5.66743755540607e-06,
+      "loss": 0.3655,
+      "step": 6103
+    },
+    {
+      "epoch": 1.5352112676056338,
+      "grad_norm": 0.31146296858787537,
+      "learning_rate": 5.665987374560977e-06,
+      "loss": 0.3524,
+      "step": 6104
+    },
+    {
+      "epoch": 1.5354627766599598,
+      "grad_norm": 0.3085387647151947,
+      "learning_rate": 5.66453713667823e-06,
+      "loss": 0.3444,
+      "step": 6105
+    },
+    {
+      "epoch": 1.5357142857142856,
+      "grad_norm": 0.3076164126396179,
+      "learning_rate": 5.663086841882036e-06,
+      "loss": 0.3589,
+      "step": 6106
+    },
+    {
+      "epoch": 1.5359657947686118,
+      "grad_norm": 0.36159980297088623,
+      "learning_rate": 5.661636490296602e-06,
+      "loss": 0.3798,
+      "step": 6107
+    },
+    {
+      "epoch": 1.5362173038229376,
+      "grad_norm": 0.33638373017311096,
+      "learning_rate": 5.660186082046142e-06,
+      "loss": 0.3668,
+      "step": 6108
+    },
+    {
+      "epoch": 1.5364688128772634,
+      "grad_norm": 0.2957940697669983,
+      "learning_rate": 5.658735617254874e-06,
+      "loss": 0.3312,
+      "step": 6109
+    },
+    {
+      "epoch": 1.5367203219315897,
+      "grad_norm": 0.3306633532047272,
+      "learning_rate": 5.6572850960470215e-06,
+      "loss": 0.3716,
+      "step": 6110
+    },
+    {
+      "epoch": 1.5369718309859155,
+      "grad_norm": 0.2976609468460083,
+      "learning_rate": 5.655834518546813e-06,
+      "loss": 0.3471,
+      "step": 6111
+    },
+    {
+      "epoch": 1.5372233400402413,
+      "grad_norm": 0.34043240547180176,
+      "learning_rate": 5.654383884878481e-06,
+      "loss": 0.332,
+      "step": 6112
+    },
+    {
+      "epoch": 1.5374748490945676,
+      "grad_norm": 0.3192569315433502,
+      "learning_rate": 5.6529331951662615e-06,
+      "loss": 0.3513,
+      "step": 6113
+    },
+    {
+      "epoch": 1.5377263581488934,
+      "grad_norm": 0.2940548062324524,
+      "learning_rate": 5.6514824495344e-06,
+      "loss": 0.3441,
+      "step": 6114
+    },
+    {
+      "epoch": 1.5379778672032192,
+      "grad_norm": 0.3519013524055481,
+      "learning_rate": 5.650031648107142e-06,
+      "loss": 0.3306,
+      "step": 6115
+    },
+    {
+      "epoch": 1.5382293762575454,
+      "grad_norm": 0.3195810616016388,
+      "learning_rate": 5.648580791008739e-06,
+      "loss": 0.35,
+      "step": 6116
+    },
+    {
+      "epoch": 1.5384808853118712,
+      "grad_norm": 0.324769526720047,
+      "learning_rate": 5.647129878363449e-06,
+      "loss": 0.3416,
+      "step": 6117
+    },
+    {
+      "epoch": 1.538732394366197,
+      "grad_norm": 0.310882568359375,
+      "learning_rate": 5.645678910295533e-06,
+      "loss": 0.3592,
+      "step": 6118
+    },
+    {
+      "epoch": 1.5389839034205233,
+      "grad_norm": 0.3428153097629547,
+      "learning_rate": 5.64422788692926e-06,
+      "loss": 0.3694,
+      "step": 6119
+    },
+    {
+      "epoch": 1.539235412474849,
+      "grad_norm": 0.39914482831954956,
+      "learning_rate": 5.642776808388897e-06,
+      "loss": 0.3532,
+      "step": 6120
+    },
+    {
+      "epoch": 1.539486921529175,
+      "grad_norm": 0.3099226653575897,
+      "learning_rate": 5.641325674798722e-06,
+      "loss": 0.3303,
+      "step": 6121
+    },
+    {
+      "epoch": 1.5397384305835011,
+      "grad_norm": 0.32525354623794556,
+      "learning_rate": 5.639874486283015e-06,
+      "loss": 0.3348,
+      "step": 6122
+    },
+    {
+      "epoch": 1.539989939637827,
+      "grad_norm": 0.3201609253883362,
+      "learning_rate": 5.638423242966061e-06,
+      "loss": 0.3441,
+      "step": 6123
+    },
+    {
+      "epoch": 1.540241448692153,
+      "grad_norm": 0.31889769434928894,
+      "learning_rate": 5.63697194497215e-06,
+      "loss": 0.3829,
+      "step": 6124
+    },
+    {
+      "epoch": 1.540492957746479,
+      "grad_norm": 0.31827834248542786,
+      "learning_rate": 5.635520592425579e-06,
+      "loss": 0.3438,
+      "step": 6125
+    },
+    {
+      "epoch": 1.5407444668008048,
+      "grad_norm": 0.30125829577445984,
+      "learning_rate": 5.634069185450642e-06,
+      "loss": 0.3563,
+      "step": 6126
+    },
+    {
+      "epoch": 1.5409959758551308,
+      "grad_norm": 0.2951527237892151,
+      "learning_rate": 5.6326177241716466e-06,
+      "loss": 0.3313,
+      "step": 6127
+    },
+    {
+      "epoch": 1.5412474849094568,
+      "grad_norm": 0.32351893186569214,
+      "learning_rate": 5.631166208712902e-06,
+      "loss": 0.354,
+      "step": 6128
+    },
+    {
+      "epoch": 1.5414989939637826,
+      "grad_norm": 0.3129163384437561,
+      "learning_rate": 5.629714639198719e-06,
+      "loss": 0.3546,
+      "step": 6129
+    },
+    {
+      "epoch": 1.5417505030181087,
+      "grad_norm": 0.29270297288894653,
+      "learning_rate": 5.628263015753418e-06,
+      "loss": 0.3406,
+      "step": 6130
+    },
+    {
+      "epoch": 1.5420020120724347,
+      "grad_norm": 0.29806989431381226,
+      "learning_rate": 5.626811338501319e-06,
+      "loss": 0.3735,
+      "step": 6131
+    },
+    {
+      "epoch": 1.5422535211267605,
+      "grad_norm": 0.31030187010765076,
+      "learning_rate": 5.625359607566751e-06,
+      "loss": 0.3505,
+      "step": 6132
+    },
+    {
+      "epoch": 1.5425050301810865,
+      "grad_norm": 0.30579960346221924,
+      "learning_rate": 5.623907823074044e-06,
+      "loss": 0.3578,
+      "step": 6133
+    },
+    {
+      "epoch": 1.5427565392354126,
+      "grad_norm": 0.3121272325515747,
+      "learning_rate": 5.622455985147536e-06,
+      "loss": 0.3878,
+      "step": 6134
+    },
+    {
+      "epoch": 1.5430080482897384,
+      "grad_norm": 0.3233649730682373,
+      "learning_rate": 5.621004093911566e-06,
+      "loss": 0.3413,
+      "step": 6135
+    },
+    {
+      "epoch": 1.5432595573440644,
+      "grad_norm": 0.32058221101760864,
+      "learning_rate": 5.6195521494904815e-06,
+      "loss": 0.3535,
+      "step": 6136
+    },
+    {
+      "epoch": 1.5435110663983904,
+      "grad_norm": 0.3023648262023926,
+      "learning_rate": 5.61810015200863e-06,
+      "loss": 0.3657,
+      "step": 6137
+    },
+    {
+      "epoch": 1.5437625754527162,
+      "grad_norm": 0.3155832290649414,
+      "learning_rate": 5.616648101590367e-06,
+      "loss": 0.3481,
+      "step": 6138
+    },
+    {
+      "epoch": 1.5440140845070423,
+      "grad_norm": 0.33503812551498413,
+      "learning_rate": 5.615195998360053e-06,
+      "loss": 0.3623,
+      "step": 6139
+    },
+    {
+      "epoch": 1.5442655935613683,
+      "grad_norm": 0.31416836380958557,
+      "learning_rate": 5.61374384244205e-06,
+      "loss": 0.3691,
+      "step": 6140
+    },
+    {
+      "epoch": 1.544517102615694,
+      "grad_norm": 0.3015648424625397,
+      "learning_rate": 5.612291633960727e-06,
+      "loss": 0.3305,
+      "step": 6141
+    },
+    {
+      "epoch": 1.54476861167002,
+      "grad_norm": 0.33664670586586,
+      "learning_rate": 5.610839373040455e-06,
+      "loss": 0.3321,
+      "step": 6142
+    },
+    {
+      "epoch": 1.5450201207243461,
+      "grad_norm": 0.31471818685531616,
+      "learning_rate": 5.609387059805614e-06,
+      "loss": 0.3553,
+      "step": 6143
+    },
+    {
+      "epoch": 1.545271629778672,
+      "grad_norm": 0.3041958212852478,
+      "learning_rate": 5.607934694380581e-06,
+      "loss": 0.3485,
+      "step": 6144
+    },
+    {
+      "epoch": 1.545523138832998,
+      "grad_norm": 0.3049089312553406,
+      "learning_rate": 5.606482276889746e-06,
+      "loss": 0.3366,
+      "step": 6145
+    },
+    {
+      "epoch": 1.545774647887324,
+      "grad_norm": 0.2998283803462982,
+      "learning_rate": 5.605029807457499e-06,
+      "loss": 0.3243,
+      "step": 6146
+    },
+    {
+      "epoch": 1.5460261569416498,
+      "grad_norm": 0.32313594222068787,
+      "learning_rate": 5.603577286208234e-06,
+      "loss": 0.3821,
+      "step": 6147
+    },
+    {
+      "epoch": 1.5462776659959758,
+      "grad_norm": 0.3039160668849945,
+      "learning_rate": 5.602124713266349e-06,
+      "loss": 0.3721,
+      "step": 6148
+    },
+    {
+      "epoch": 1.5465291750503019,
+      "grad_norm": 0.31081444025039673,
+      "learning_rate": 5.60067208875625e-06,
+      "loss": 0.3554,
+      "step": 6149
+    },
+    {
+      "epoch": 1.5467806841046277,
+      "grad_norm": 0.3036777675151825,
+      "learning_rate": 5.599219412802344e-06,
+      "loss": 0.3459,
+      "step": 6150
+    },
+    {
+      "epoch": 1.5470321931589537,
+      "grad_norm": 0.2954856753349304,
+      "learning_rate": 5.597766685529043e-06,
+      "loss": 0.3439,
+      "step": 6151
+    },
+    {
+      "epoch": 1.5472837022132797,
+      "grad_norm": 0.3063904941082001,
+      "learning_rate": 5.596313907060766e-06,
+      "loss": 0.3452,
+      "step": 6152
+    },
+    {
+      "epoch": 1.5475352112676055,
+      "grad_norm": 0.33368048071861267,
+      "learning_rate": 5.594861077521935e-06,
+      "loss": 0.3523,
+      "step": 6153
+    },
+    {
+      "epoch": 1.5477867203219315,
+      "grad_norm": 0.31460052728652954,
+      "learning_rate": 5.593408197036973e-06,
+      "loss": 0.3389,
+      "step": 6154
+    },
+    {
+      "epoch": 1.5480382293762576,
+      "grad_norm": 0.32115140557289124,
+      "learning_rate": 5.59195526573031e-06,
+      "loss": 0.3714,
+      "step": 6155
+    },
+    {
+      "epoch": 1.5482897384305834,
+      "grad_norm": 0.3152860105037689,
+      "learning_rate": 5.590502283726383e-06,
+      "loss": 0.3539,
+      "step": 6156
+    },
+    {
+      "epoch": 1.5485412474849096,
+      "grad_norm": 0.33016350865364075,
+      "learning_rate": 5.5890492511496294e-06,
+      "loss": 0.366,
+      "step": 6157
+    },
+    {
+      "epoch": 1.5487927565392354,
+      "grad_norm": 0.3130359351634979,
+      "learning_rate": 5.587596168124493e-06,
+      "loss": 0.3789,
+      "step": 6158
+    },
+    {
+      "epoch": 1.5490442655935612,
+      "grad_norm": 0.3576628565788269,
+      "learning_rate": 5.5861430347754195e-06,
+      "loss": 0.3626,
+      "step": 6159
+    },
+    {
+      "epoch": 1.5492957746478875,
+      "grad_norm": 0.28835195302963257,
+      "learning_rate": 5.584689851226863e-06,
+      "loss": 0.3559,
+      "step": 6160
+    },
+    {
+      "epoch": 1.5495472837022133,
+      "grad_norm": 0.32597804069519043,
+      "learning_rate": 5.583236617603278e-06,
+      "loss": 0.3582,
+      "step": 6161
+    },
+    {
+      "epoch": 1.549798792756539,
+      "grad_norm": 0.29985523223876953,
+      "learning_rate": 5.5817833340291265e-06,
+      "loss": 0.3467,
+      "step": 6162
+    },
+    {
+      "epoch": 1.5500503018108653,
+      "grad_norm": 0.3118489384651184,
+      "learning_rate": 5.5803300006288704e-06,
+      "loss": 0.3401,
+      "step": 6163
+    },
+    {
+      "epoch": 1.5503018108651911,
+      "grad_norm": 0.302613228559494,
+      "learning_rate": 5.578876617526982e-06,
+      "loss": 0.3266,
+      "step": 6164
+    },
+    {
+      "epoch": 1.550553319919517,
+      "grad_norm": 0.3214384615421295,
+      "learning_rate": 5.577423184847932e-06,
+      "loss": 0.3369,
+      "step": 6165
+    },
+    {
+      "epoch": 1.5508048289738432,
+      "grad_norm": 0.2934117019176483,
+      "learning_rate": 5.575969702716199e-06,
+      "loss": 0.3559,
+      "step": 6166
+    },
+    {
+      "epoch": 1.551056338028169,
+      "grad_norm": 0.3137935698032379,
+      "learning_rate": 5.574516171256263e-06,
+      "loss": 0.3543,
+      "step": 6167
+    },
+    {
+      "epoch": 1.5513078470824948,
+      "grad_norm": 0.29603636264801025,
+      "learning_rate": 5.5730625905926114e-06,
+      "loss": 0.3514,
+      "step": 6168
+    },
+    {
+      "epoch": 1.551559356136821,
+      "grad_norm": 0.3102024793624878,
+      "learning_rate": 5.571608960849735e-06,
+      "loss": 0.3519,
+      "step": 6169
+    },
+    {
+      "epoch": 1.5518108651911469,
+      "grad_norm": 0.2983706593513489,
+      "learning_rate": 5.570155282152125e-06,
+      "loss": 0.3428,
+      "step": 6170
+    },
+    {
+      "epoch": 1.5520623742454729,
+      "grad_norm": 0.31096917390823364,
+      "learning_rate": 5.568701554624284e-06,
+      "loss": 0.3497,
+      "step": 6171
+    },
+    {
+      "epoch": 1.552313883299799,
+      "grad_norm": 0.28655806183815,
+      "learning_rate": 5.567247778390712e-06,
+      "loss": 0.3459,
+      "step": 6172
+    },
+    {
+      "epoch": 1.5525653923541247,
+      "grad_norm": 0.30332696437835693,
+      "learning_rate": 5.565793953575916e-06,
+      "loss": 0.3527,
+      "step": 6173
+    },
+    {
+      "epoch": 1.5528169014084507,
+      "grad_norm": 0.3258723020553589,
+      "learning_rate": 5.5643400803044075e-06,
+      "loss": 0.3454,
+      "step": 6174
+    },
+    {
+      "epoch": 1.5530684104627768,
+      "grad_norm": 0.34094491600990295,
+      "learning_rate": 5.5628861587007035e-06,
+      "loss": 0.3808,
+      "step": 6175
+    },
+    {
+      "epoch": 1.5533199195171026,
+      "grad_norm": 0.2914903461933136,
+      "learning_rate": 5.5614321888893195e-06,
+      "loss": 0.346,
+      "step": 6176
+    },
+    {
+      "epoch": 1.5535714285714286,
+      "grad_norm": 0.338826060295105,
+      "learning_rate": 5.559978170994781e-06,
+      "loss": 0.3567,
+      "step": 6177
+    },
+    {
+      "epoch": 1.5538229376257546,
+      "grad_norm": 0.3329686224460602,
+      "learning_rate": 5.558524105141616e-06,
+      "loss": 0.3439,
+      "step": 6178
+    },
+    {
+      "epoch": 1.5540744466800804,
+      "grad_norm": 0.3257056474685669,
+      "learning_rate": 5.557069991454356e-06,
+      "loss": 0.3378,
+      "step": 6179
+    },
+    {
+      "epoch": 1.5543259557344065,
+      "grad_norm": 0.3230164349079132,
+      "learning_rate": 5.5556158300575345e-06,
+      "loss": 0.3419,
+      "step": 6180
+    },
+    {
+      "epoch": 1.5545774647887325,
+      "grad_norm": 0.3321927487850189,
+      "learning_rate": 5.554161621075693e-06,
+      "loss": 0.3163,
+      "step": 6181
+    },
+    {
+      "epoch": 1.5548289738430583,
+      "grad_norm": 0.36797481775283813,
+      "learning_rate": 5.552707364633376e-06,
+      "loss": 0.3797,
+      "step": 6182
+    },
+    {
+      "epoch": 1.5550804828973843,
+      "grad_norm": 0.3352421820163727,
+      "learning_rate": 5.5512530608551315e-06,
+      "loss": 0.3408,
+      "step": 6183
+    },
+    {
+      "epoch": 1.5553319919517103,
+      "grad_norm": 0.2971833050251007,
+      "learning_rate": 5.549798709865512e-06,
+      "loss": 0.3406,
+      "step": 6184
+    },
+    {
+      "epoch": 1.5555835010060362,
+      "grad_norm": 0.30301129817962646,
+      "learning_rate": 5.5483443117890715e-06,
+      "loss": 0.3458,
+      "step": 6185
+    },
+    {
+      "epoch": 1.5558350100603622,
+      "grad_norm": 0.34085342288017273,
+      "learning_rate": 5.546889866750371e-06,
+      "loss": 0.3523,
+      "step": 6186
+    },
+    {
+      "epoch": 1.5560865191146882,
+      "grad_norm": 0.3280344307422638,
+      "learning_rate": 5.5454353748739755e-06,
+      "loss": 0.3563,
+      "step": 6187
+    },
+    {
+      "epoch": 1.556338028169014,
+      "grad_norm": 0.32044655084609985,
+      "learning_rate": 5.543980836284451e-06,
+      "loss": 0.3555,
+      "step": 6188
+    },
+    {
+      "epoch": 1.55658953722334,
+      "grad_norm": 0.30782240629196167,
+      "learning_rate": 5.542526251106372e-06,
+      "loss": 0.3402,
+      "step": 6189
+    },
+    {
+      "epoch": 1.556841046277666,
+      "grad_norm": 0.3036595284938812,
+      "learning_rate": 5.541071619464314e-06,
+      "loss": 0.3148,
+      "step": 6190
+    },
+    {
+      "epoch": 1.5570925553319919,
+      "grad_norm": 0.30741608142852783,
+      "learning_rate": 5.539616941482855e-06,
+      "loss": 0.3522,
+      "step": 6191
+    },
+    {
+      "epoch": 1.557344064386318,
+      "grad_norm": 0.32885679602622986,
+      "learning_rate": 5.538162217286581e-06,
+      "loss": 0.3384,
+      "step": 6192
+    },
+    {
+      "epoch": 1.557595573440644,
+      "grad_norm": 0.29733267426490784,
+      "learning_rate": 5.53670744700008e-06,
+      "loss": 0.3465,
+      "step": 6193
+    },
+    {
+      "epoch": 1.5578470824949697,
+      "grad_norm": 0.33164188265800476,
+      "learning_rate": 5.535252630747945e-06,
+      "loss": 0.3429,
+      "step": 6194
+    },
+    {
+      "epoch": 1.5580985915492958,
+      "grad_norm": 0.3217810392379761,
+      "learning_rate": 5.533797768654771e-06,
+      "loss": 0.3594,
+      "step": 6195
+    },
+    {
+      "epoch": 1.5583501006036218,
+      "grad_norm": 0.29132455587387085,
+      "learning_rate": 5.532342860845157e-06,
+      "loss": 0.3428,
+      "step": 6196
+    },
+    {
+      "epoch": 1.5586016096579476,
+      "grad_norm": 0.30485719442367554,
+      "learning_rate": 5.5308879074437065e-06,
+      "loss": 0.3193,
+      "step": 6197
+    },
+    {
+      "epoch": 1.5588531187122736,
+      "grad_norm": 0.310776025056839,
+      "learning_rate": 5.529432908575029e-06,
+      "loss": 0.3504,
+      "step": 6198
+    },
+    {
+      "epoch": 1.5591046277665996,
+      "grad_norm": 0.32021844387054443,
+      "learning_rate": 5.527977864363734e-06,
+      "loss": 0.3495,
+      "step": 6199
+    },
+    {
+      "epoch": 1.5593561368209254,
+      "grad_norm": 0.2971351146697998,
+      "learning_rate": 5.5265227749344385e-06,
+      "loss": 0.3765,
+      "step": 6200
+    },
+    {
+      "epoch": 1.5596076458752515,
+      "grad_norm": 0.30885204672813416,
+      "learning_rate": 5.525067640411761e-06,
+      "loss": 0.3426,
+      "step": 6201
+    },
+    {
+      "epoch": 1.5598591549295775,
+      "grad_norm": 0.31158149242401123,
+      "learning_rate": 5.523612460920326e-06,
+      "loss": 0.3468,
+      "step": 6202
+    },
+    {
+      "epoch": 1.5601106639839033,
+      "grad_norm": 0.3428870439529419,
+      "learning_rate": 5.5221572365847565e-06,
+      "loss": 0.3758,
+      "step": 6203
+    },
+    {
+      "epoch": 1.5603621730382293,
+      "grad_norm": 0.35247480869293213,
+      "learning_rate": 5.520701967529689e-06,
+      "loss": 0.3744,
+      "step": 6204
+    },
+    {
+      "epoch": 1.5606136820925554,
+      "grad_norm": 0.3332814872264862,
+      "learning_rate": 5.519246653879754e-06,
+      "loss": 0.3641,
+      "step": 6205
+    },
+    {
+      "epoch": 1.5608651911468812,
+      "grad_norm": 0.3046468198299408,
+      "learning_rate": 5.517791295759592e-06,
+      "loss": 0.3338,
+      "step": 6206
+    },
+    {
+      "epoch": 1.5611167002012074,
+      "grad_norm": 0.32283511757850647,
+      "learning_rate": 5.516335893293846e-06,
+      "loss": 0.3549,
+      "step": 6207
+    },
+    {
+      "epoch": 1.5613682092555332,
+      "grad_norm": 0.2913826107978821,
+      "learning_rate": 5.514880446607161e-06,
+      "loss": 0.319,
+      "step": 6208
+    },
+    {
+      "epoch": 1.561619718309859,
+      "grad_norm": 0.33693376183509827,
+      "learning_rate": 5.513424955824185e-06,
+      "loss": 0.3555,
+      "step": 6209
+    },
+    {
+      "epoch": 1.5618712273641853,
+      "grad_norm": 0.348818838596344,
+      "learning_rate": 5.511969421069574e-06,
+      "loss": 0.3493,
+      "step": 6210
+    },
+    {
+      "epoch": 1.562122736418511,
+      "grad_norm": 0.3071136474609375,
+      "learning_rate": 5.510513842467986e-06,
+      "loss": 0.3714,
+      "step": 6211
+    },
+    {
+      "epoch": 1.5623742454728369,
+      "grad_norm": 0.34389618039131165,
+      "learning_rate": 5.50905822014408e-06,
+      "loss": 0.3539,
+      "step": 6212
+    },
+    {
+      "epoch": 1.5626257545271631,
+      "grad_norm": 0.3115088641643524,
+      "learning_rate": 5.507602554222523e-06,
+      "loss": 0.3671,
+      "step": 6213
+    },
+    {
+      "epoch": 1.562877263581489,
+      "grad_norm": 0.3112759292125702,
+      "learning_rate": 5.506146844827981e-06,
+      "loss": 0.3616,
+      "step": 6214
+    },
+    {
+      "epoch": 1.5631287726358147,
+      "grad_norm": 0.32954081892967224,
+      "learning_rate": 5.50469109208513e-06,
+      "loss": 0.333,
+      "step": 6215
+    },
+    {
+      "epoch": 1.563380281690141,
+      "grad_norm": 0.3064148426055908,
+      "learning_rate": 5.503235296118643e-06,
+      "loss": 0.3339,
+      "step": 6216
+    },
+    {
+      "epoch": 1.5636317907444668,
+      "grad_norm": 0.29228824377059937,
+      "learning_rate": 5.501779457053202e-06,
+      "loss": 0.3435,
+      "step": 6217
+    },
+    {
+      "epoch": 1.5638832997987926,
+      "grad_norm": 0.32584089040756226,
+      "learning_rate": 5.50032357501349e-06,
+      "loss": 0.3494,
+      "step": 6218
+    },
+    {
+      "epoch": 1.5641348088531188,
+      "grad_norm": 0.29713818430900574,
+      "learning_rate": 5.498867650124193e-06,
+      "loss": 0.3463,
+      "step": 6219
+    },
+    {
+      "epoch": 1.5643863179074446,
+      "grad_norm": 0.31487134099006653,
+      "learning_rate": 5.4974116825100035e-06,
+      "loss": 0.3632,
+      "step": 6220
+    },
+    {
+      "epoch": 1.5646378269617707,
+      "grad_norm": 0.30101948976516724,
+      "learning_rate": 5.495955672295615e-06,
+      "loss": 0.3582,
+      "step": 6221
+    },
+    {
+      "epoch": 1.5648893360160967,
+      "grad_norm": 0.30046340823173523,
+      "learning_rate": 5.494499619605725e-06,
+      "loss": 0.3532,
+      "step": 6222
+    },
+    {
+      "epoch": 1.5651408450704225,
+      "grad_norm": 0.3313204050064087,
+      "learning_rate": 5.493043524565037e-06,
+      "loss": 0.3435,
+      "step": 6223
+    },
+    {
+      "epoch": 1.5653923541247485,
+      "grad_norm": 0.30352333188056946,
+      "learning_rate": 5.491587387298256e-06,
+      "loss": 0.3368,
+      "step": 6224
+    },
+    {
+      "epoch": 1.5656438631790746,
+      "grad_norm": 0.30736225843429565,
+      "learning_rate": 5.490131207930089e-06,
+      "loss": 0.3715,
+      "step": 6225
+    },
+    {
+      "epoch": 1.5658953722334004,
+      "grad_norm": 0.3192325234413147,
+      "learning_rate": 5.488674986585252e-06,
+      "loss": 0.3587,
+      "step": 6226
+    },
+    {
+      "epoch": 1.5661468812877264,
+      "grad_norm": 0.32336869835853577,
+      "learning_rate": 5.487218723388459e-06,
+      "loss": 0.3587,
+      "step": 6227
+    },
+    {
+      "epoch": 1.5663983903420524,
+      "grad_norm": 0.33127787709236145,
+      "learning_rate": 5.48576241846443e-06,
+      "loss": 0.3737,
+      "step": 6228
+    },
+    {
+      "epoch": 1.5666498993963782,
+      "grad_norm": 0.30386435985565186,
+      "learning_rate": 5.484306071937889e-06,
+      "loss": 0.3378,
+      "step": 6229
+    },
+    {
+      "epoch": 1.5669014084507042,
+      "grad_norm": 0.30401545763015747,
+      "learning_rate": 5.4828496839335635e-06,
+      "loss": 0.3511,
+      "step": 6230
+    },
+    {
+      "epoch": 1.5671529175050303,
+      "grad_norm": 0.327298641204834,
+      "learning_rate": 5.4813932545761815e-06,
+      "loss": 0.3497,
+      "step": 6231
+    },
+    {
+      "epoch": 1.567404426559356,
+      "grad_norm": 0.3015185296535492,
+      "learning_rate": 5.4799367839904805e-06,
+      "loss": 0.3462,
+      "step": 6232
+    },
+    {
+      "epoch": 1.567655935613682,
+      "grad_norm": 0.3208427429199219,
+      "learning_rate": 5.478480272301195e-06,
+      "loss": 0.3293,
+      "step": 6233
+    },
+    {
+      "epoch": 1.5679074446680081,
+      "grad_norm": 0.34195858240127563,
+      "learning_rate": 5.477023719633069e-06,
+      "loss": 0.3793,
+      "step": 6234
+    },
+    {
+      "epoch": 1.568158953722334,
+      "grad_norm": 0.31113073229789734,
+      "learning_rate": 5.4755671261108445e-06,
+      "loss": 0.3565,
+      "step": 6235
+    },
+    {
+      "epoch": 1.56841046277666,
+      "grad_norm": 0.3247259557247162,
+      "learning_rate": 5.474110491859272e-06,
+      "loss": 0.3352,
+      "step": 6236
+    },
+    {
+      "epoch": 1.568661971830986,
+      "grad_norm": 0.3409235179424286,
+      "learning_rate": 5.4726538170031e-06,
+      "loss": 0.3498,
+      "step": 6237
+    },
+    {
+      "epoch": 1.5689134808853118,
+      "grad_norm": 0.32643982768058777,
+      "learning_rate": 5.471197101667087e-06,
+      "loss": 0.3437,
+      "step": 6238
+    },
+    {
+      "epoch": 1.5691649899396378,
+      "grad_norm": 0.3240383267402649,
+      "learning_rate": 5.469740345975989e-06,
+      "loss": 0.3527,
+      "step": 6239
+    },
+    {
+      "epoch": 1.5694164989939638,
+      "grad_norm": 0.34534913301467896,
+      "learning_rate": 5.468283550054571e-06,
+      "loss": 0.3658,
+      "step": 6240
+    },
+    {
+      "epoch": 1.5696680080482897,
+      "grad_norm": 0.2972480356693268,
+      "learning_rate": 5.466826714027595e-06,
+      "loss": 0.3357,
+      "step": 6241
+    },
+    {
+      "epoch": 1.5699195171026157,
+      "grad_norm": 0.3179325461387634,
+      "learning_rate": 5.465369838019832e-06,
+      "loss": 0.3268,
+      "step": 6242
+    },
+    {
+      "epoch": 1.5701710261569417,
+      "grad_norm": 0.3161191940307617,
+      "learning_rate": 5.463912922156053e-06,
+      "loss": 0.3631,
+      "step": 6243
+    },
+    {
+      "epoch": 1.5704225352112675,
+      "grad_norm": 0.32680991291999817,
+      "learning_rate": 5.462455966561034e-06,
+      "loss": 0.3689,
+      "step": 6244
+    },
+    {
+      "epoch": 1.5706740442655935,
+      "grad_norm": 0.31402167677879333,
+      "learning_rate": 5.460998971359556e-06,
+      "loss": 0.3468,
+      "step": 6245
+    },
+    {
+      "epoch": 1.5709255533199196,
+      "grad_norm": 0.3343561589717865,
+      "learning_rate": 5.459541936676398e-06,
+      "loss": 0.3655,
+      "step": 6246
+    },
+    {
+      "epoch": 1.5711770623742454,
+      "grad_norm": 0.30885884165763855,
+      "learning_rate": 5.45808486263635e-06,
+      "loss": 0.3284,
+      "step": 6247
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.3167196214199066,
+      "learning_rate": 5.456627749364198e-06,
+      "loss": 0.3454,
+      "step": 6248
+    },
+    {
+      "epoch": 1.5716800804828974,
+      "grad_norm": 0.33008265495300293,
+      "learning_rate": 5.455170596984738e-06,
+      "loss": 0.3528,
+      "step": 6249
+    },
+    {
+      "epoch": 1.5719315895372232,
+      "grad_norm": 0.30799832940101624,
+      "learning_rate": 5.4537134056227626e-06,
+      "loss": 0.3536,
+      "step": 6250
+    },
+    {
+      "epoch": 1.5721830985915493,
+      "grad_norm": 0.29477211833000183,
+      "learning_rate": 5.452256175403072e-06,
+      "loss": 0.3448,
+      "step": 6251
+    },
+    {
+      "epoch": 1.5724346076458753,
+      "grad_norm": 0.3047988712787628,
+      "learning_rate": 5.4507989064504695e-06,
+      "loss": 0.3506,
+      "step": 6252
+    },
+    {
+      "epoch": 1.572686116700201,
+      "grad_norm": 0.32256054878234863,
+      "learning_rate": 5.4493415988897615e-06,
+      "loss": 0.3653,
+      "step": 6253
+    },
+    {
+      "epoch": 1.5729376257545271,
+      "grad_norm": 0.3178042471408844,
+      "learning_rate": 5.4478842528457565e-06,
+      "loss": 0.3709,
+      "step": 6254
+    },
+    {
+      "epoch": 1.5731891348088531,
+      "grad_norm": 0.329878568649292,
+      "learning_rate": 5.4464268684432664e-06,
+      "loss": 0.3329,
+      "step": 6255
+    },
+    {
+      "epoch": 1.573440643863179,
+      "grad_norm": 0.30290189385414124,
+      "learning_rate": 5.444969445807109e-06,
+      "loss": 0.3458,
+      "step": 6256
+    },
+    {
+      "epoch": 1.5736921529175052,
+      "grad_norm": 0.3331491947174072,
+      "learning_rate": 5.4435119850621e-06,
+      "loss": 0.3516,
+      "step": 6257
+    },
+    {
+      "epoch": 1.573943661971831,
+      "grad_norm": 0.2870064675807953,
+      "learning_rate": 5.442054486333066e-06,
+      "loss": 0.3636,
+      "step": 6258
+    },
+    {
+      "epoch": 1.5741951710261568,
+      "grad_norm": 0.3208286762237549,
+      "learning_rate": 5.440596949744831e-06,
+      "loss": 0.3486,
+      "step": 6259
+    },
+    {
+      "epoch": 1.574446680080483,
+      "grad_norm": 0.30544915795326233,
+      "learning_rate": 5.439139375422223e-06,
+      "loss": 0.3647,
+      "step": 6260
+    },
+    {
+      "epoch": 1.5746981891348089,
+      "grad_norm": 0.3159911036491394,
+      "learning_rate": 5.437681763490075e-06,
+      "loss": 0.3524,
+      "step": 6261
+    },
+    {
+      "epoch": 1.5749496981891347,
+      "grad_norm": 0.315897136926651,
+      "learning_rate": 5.4362241140732215e-06,
+      "loss": 0.3794,
+      "step": 6262
+    },
+    {
+      "epoch": 1.575201207243461,
+      "grad_norm": 0.3100701570510864,
+      "learning_rate": 5.434766427296502e-06,
+      "loss": 0.3537,
+      "step": 6263
+    },
+    {
+      "epoch": 1.5754527162977867,
+      "grad_norm": 0.327070951461792,
+      "learning_rate": 5.433308703284759e-06,
+      "loss": 0.3974,
+      "step": 6264
+    },
+    {
+      "epoch": 1.5757042253521125,
+      "grad_norm": 0.31196388602256775,
+      "learning_rate": 5.431850942162834e-06,
+      "loss": 0.3373,
+      "step": 6265
+    },
+    {
+      "epoch": 1.5759557344064388,
+      "grad_norm": 0.28920623660087585,
+      "learning_rate": 5.430393144055579e-06,
+      "loss": 0.3308,
+      "step": 6266
+    },
+    {
+      "epoch": 1.5762072434607646,
+      "grad_norm": 0.30474618077278137,
+      "learning_rate": 5.428935309087844e-06,
+      "loss": 0.3323,
+      "step": 6267
+    },
+    {
+      "epoch": 1.5764587525150904,
+      "grad_norm": 0.3304254412651062,
+      "learning_rate": 5.427477437384482e-06,
+      "loss": 0.3647,
+      "step": 6268
+    },
+    {
+      "epoch": 1.5767102615694166,
+      "grad_norm": 0.28863289952278137,
+      "learning_rate": 5.426019529070352e-06,
+      "loss": 0.362,
+      "step": 6269
+    },
+    {
+      "epoch": 1.5769617706237424,
+      "grad_norm": 0.31353408098220825,
+      "learning_rate": 5.4245615842703146e-06,
+      "loss": 0.3637,
+      "step": 6270
+    },
+    {
+      "epoch": 1.5772132796780685,
+      "grad_norm": 0.3224716782569885,
+      "learning_rate": 5.4231036031092345e-06,
+      "loss": 0.3273,
+      "step": 6271
+    },
+    {
+      "epoch": 1.5774647887323945,
+      "grad_norm": 0.31004685163497925,
+      "learning_rate": 5.4216455857119765e-06,
+      "loss": 0.354,
+      "step": 6272
+    },
+    {
+      "epoch": 1.5777162977867203,
+      "grad_norm": 0.32508277893066406,
+      "learning_rate": 5.420187532203413e-06,
+      "loss": 0.3552,
+      "step": 6273
+    },
+    {
+      "epoch": 1.5779678068410463,
+      "grad_norm": 0.33112356066703796,
+      "learning_rate": 5.418729442708416e-06,
+      "loss": 0.3846,
+      "step": 6274
+    },
+    {
+      "epoch": 1.5782193158953723,
+      "grad_norm": 0.32484641671180725,
+      "learning_rate": 5.417271317351861e-06,
+      "loss": 0.3589,
+      "step": 6275
+    },
+    {
+      "epoch": 1.5784708249496981,
+      "grad_norm": 0.3176564574241638,
+      "learning_rate": 5.415813156258628e-06,
+      "loss": 0.3525,
+      "step": 6276
+    },
+    {
+      "epoch": 1.5787223340040242,
+      "grad_norm": 0.31098705530166626,
+      "learning_rate": 5.4143549595536e-06,
+      "loss": 0.3355,
+      "step": 6277
+    },
+    {
+      "epoch": 1.5789738430583502,
+      "grad_norm": 0.34657394886016846,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.3566,
+      "step": 6278
+    },
+    {
+      "epoch": 1.579225352112676,
+      "grad_norm": 0.31053581833839417,
+      "learning_rate": 5.411438459807703e-06,
+      "loss": 0.3741,
+      "step": 6279
+    },
+    {
+      "epoch": 1.579476861167002,
+      "grad_norm": 0.28460362553596497,
+      "learning_rate": 5.4099801570166135e-06,
+      "loss": 0.3299,
+      "step": 6280
+    },
+    {
+      "epoch": 1.579728370221328,
+      "grad_norm": 0.3039803206920624,
+      "learning_rate": 5.408521819113287e-06,
+      "loss": 0.3297,
+      "step": 6281
+    },
+    {
+      "epoch": 1.5799798792756539,
+      "grad_norm": 0.31220072507858276,
+      "learning_rate": 5.407063446222623e-06,
+      "loss": 0.3368,
+      "step": 6282
+    },
+    {
+      "epoch": 1.58023138832998,
+      "grad_norm": 0.31574392318725586,
+      "learning_rate": 5.4056050384695225e-06,
+      "loss": 0.3582,
+      "step": 6283
+    },
+    {
+      "epoch": 1.580482897384306,
+      "grad_norm": 0.34987935423851013,
+      "learning_rate": 5.404146595978887e-06,
+      "loss": 0.3484,
+      "step": 6284
+    },
+    {
+      "epoch": 1.5807344064386317,
+      "grad_norm": 0.3038938641548157,
+      "learning_rate": 5.402688118875624e-06,
+      "loss": 0.3401,
+      "step": 6285
+    },
+    {
+      "epoch": 1.5809859154929577,
+      "grad_norm": 0.29709485173225403,
+      "learning_rate": 5.401229607284644e-06,
+      "loss": 0.3588,
+      "step": 6286
+    },
+    {
+      "epoch": 1.5812374245472838,
+      "grad_norm": 0.3312198221683502,
+      "learning_rate": 5.3997710613308565e-06,
+      "loss": 0.3632,
+      "step": 6287
+    },
+    {
+      "epoch": 1.5814889336016096,
+      "grad_norm": 0.30179715156555176,
+      "learning_rate": 5.398312481139179e-06,
+      "loss": 0.3404,
+      "step": 6288
+    },
+    {
+      "epoch": 1.5817404426559356,
+      "grad_norm": 0.3145090639591217,
+      "learning_rate": 5.396853866834529e-06,
+      "loss": 0.3781,
+      "step": 6289
+    },
+    {
+      "epoch": 1.5819919517102616,
+      "grad_norm": 0.33599305152893066,
+      "learning_rate": 5.395395218541829e-06,
+      "loss": 0.3496,
+      "step": 6290
+    },
+    {
+      "epoch": 1.5822434607645874,
+      "grad_norm": 0.30032995343208313,
+      "learning_rate": 5.393936536386001e-06,
+      "loss": 0.3498,
+      "step": 6291
+    },
+    {
+      "epoch": 1.5824949698189135,
+      "grad_norm": 0.30552947521209717,
+      "learning_rate": 5.392477820491974e-06,
+      "loss": 0.3472,
+      "step": 6292
+    },
+    {
+      "epoch": 1.5827464788732395,
+      "grad_norm": 0.29774901270866394,
+      "learning_rate": 5.391019070984676e-06,
+      "loss": 0.3306,
+      "step": 6293
+    },
+    {
+      "epoch": 1.5829979879275653,
+      "grad_norm": 0.3048769235610962,
+      "learning_rate": 5.389560287989043e-06,
+      "loss": 0.3284,
+      "step": 6294
+    },
+    {
+      "epoch": 1.5832494969818913,
+      "grad_norm": 0.33023905754089355,
+      "learning_rate": 5.388101471630006e-06,
+      "loss": 0.3384,
+      "step": 6295
+    },
+    {
+      "epoch": 1.5835010060362174,
+      "grad_norm": 0.30348482728004456,
+      "learning_rate": 5.3866426220325075e-06,
+      "loss": 0.3713,
+      "step": 6296
+    },
+    {
+      "epoch": 1.5837525150905432,
+      "grad_norm": 0.3093292713165283,
+      "learning_rate": 5.385183739321486e-06,
+      "loss": 0.3527,
+      "step": 6297
+    },
+    {
+      "epoch": 1.5840040241448692,
+      "grad_norm": 0.30583062767982483,
+      "learning_rate": 5.383724823621889e-06,
+      "loss": 0.3345,
+      "step": 6298
+    },
+    {
+      "epoch": 1.5842555331991952,
+      "grad_norm": 0.33037832379341125,
+      "learning_rate": 5.38226587505866e-06,
+      "loss": 0.3787,
+      "step": 6299
+    },
+    {
+      "epoch": 1.584507042253521,
+      "grad_norm": 0.312497615814209,
+      "learning_rate": 5.380806893756748e-06,
+      "loss": 0.3446,
+      "step": 6300
+    },
+    {
+      "epoch": 1.584758551307847,
+      "grad_norm": 0.3048454225063324,
+      "learning_rate": 5.3793478798411105e-06,
+      "loss": 0.3678,
+      "step": 6301
+    },
+    {
+      "epoch": 1.585010060362173,
+      "grad_norm": 0.290546178817749,
+      "learning_rate": 5.3778888334367e-06,
+      "loss": 0.3401,
+      "step": 6302
+    },
+    {
+      "epoch": 1.5852615694164989,
+      "grad_norm": 0.3122500479221344,
+      "learning_rate": 5.376429754668475e-06,
+      "loss": 0.3374,
+      "step": 6303
+    },
+    {
+      "epoch": 1.585513078470825,
+      "grad_norm": 0.3183901011943817,
+      "learning_rate": 5.374970643661397e-06,
+      "loss": 0.3668,
+      "step": 6304
+    },
+    {
+      "epoch": 1.585764587525151,
+      "grad_norm": 0.3228647708892822,
+      "learning_rate": 5.373511500540428e-06,
+      "loss": 0.3438,
+      "step": 6305
+    },
+    {
+      "epoch": 1.5860160965794767,
+      "grad_norm": 0.3075144290924072,
+      "learning_rate": 5.372052325430537e-06,
+      "loss": 0.3626,
+      "step": 6306
+    },
+    {
+      "epoch": 1.586267605633803,
+      "grad_norm": 0.3228583335876465,
+      "learning_rate": 5.37059311845669e-06,
+      "loss": 0.3396,
+      "step": 6307
+    },
+    {
+      "epoch": 1.5865191146881288,
+      "grad_norm": 0.33655041456222534,
+      "learning_rate": 5.3691338797438615e-06,
+      "loss": 0.3342,
+      "step": 6308
+    },
+    {
+      "epoch": 1.5867706237424546,
+      "grad_norm": 0.3635300397872925,
+      "learning_rate": 5.3676746094170265e-06,
+      "loss": 0.355,
+      "step": 6309
+    },
+    {
+      "epoch": 1.5870221327967808,
+      "grad_norm": 0.32811570167541504,
+      "learning_rate": 5.3662153076011614e-06,
+      "loss": 0.3313,
+      "step": 6310
+    },
+    {
+      "epoch": 1.5872736418511066,
+      "grad_norm": 0.3333826959133148,
+      "learning_rate": 5.364755974421244e-06,
+      "loss": 0.342,
+      "step": 6311
+    },
+    {
+      "epoch": 1.5875251509054324,
+      "grad_norm": 0.34751078486442566,
+      "learning_rate": 5.363296610002261e-06,
+      "loss": 0.348,
+      "step": 6312
+    },
+    {
+      "epoch": 1.5877766599597587,
+      "grad_norm": 0.3453575372695923,
+      "learning_rate": 5.361837214469197e-06,
+      "loss": 0.349,
+      "step": 6313
+    },
+    {
+      "epoch": 1.5880281690140845,
+      "grad_norm": 0.3269634544849396,
+      "learning_rate": 5.360377787947037e-06,
+      "loss": 0.3284,
+      "step": 6314
+    },
+    {
+      "epoch": 1.5882796780684103,
+      "grad_norm": 0.3597438335418701,
+      "learning_rate": 5.358918330560776e-06,
+      "loss": 0.3663,
+      "step": 6315
+    },
+    {
+      "epoch": 1.5885311871227366,
+      "grad_norm": 0.3321945369243622,
+      "learning_rate": 5.357458842435405e-06,
+      "loss": 0.3514,
+      "step": 6316
+    },
+    {
+      "epoch": 1.5887826961770624,
+      "grad_norm": 0.33890068531036377,
+      "learning_rate": 5.35599932369592e-06,
+      "loss": 0.3409,
+      "step": 6317
+    },
+    {
+      "epoch": 1.5890342052313882,
+      "grad_norm": 0.33739712834358215,
+      "learning_rate": 5.35453977446732e-06,
+      "loss": 0.3655,
+      "step": 6318
+    },
+    {
+      "epoch": 1.5892857142857144,
+      "grad_norm": 0.3216194808483124,
+      "learning_rate": 5.353080194874606e-06,
+      "loss": 0.325,
+      "step": 6319
+    },
+    {
+      "epoch": 1.5895372233400402,
+      "grad_norm": 0.2983427941799164,
+      "learning_rate": 5.351620585042783e-06,
+      "loss": 0.354,
+      "step": 6320
+    },
+    {
+      "epoch": 1.5897887323943662,
+      "grad_norm": 0.3386406898498535,
+      "learning_rate": 5.350160945096856e-06,
+      "loss": 0.3736,
+      "step": 6321
+    },
+    {
+      "epoch": 1.5900402414486923,
+      "grad_norm": 0.3367750346660614,
+      "learning_rate": 5.348701275161834e-06,
+      "loss": 0.3222,
+      "step": 6322
+    },
+    {
+      "epoch": 1.590291750503018,
+      "grad_norm": 0.3235786259174347,
+      "learning_rate": 5.347241575362729e-06,
+      "loss": 0.3701,
+      "step": 6323
+    },
+    {
+      "epoch": 1.590543259557344,
+      "grad_norm": 0.29877564311027527,
+      "learning_rate": 5.345781845824557e-06,
+      "loss": 0.3367,
+      "step": 6324
+    },
+    {
+      "epoch": 1.5907947686116701,
+      "grad_norm": 0.30587801337242126,
+      "learning_rate": 5.344322086672332e-06,
+      "loss": 0.3363,
+      "step": 6325
+    },
+    {
+      "epoch": 1.591046277665996,
+      "grad_norm": 0.31401360034942627,
+      "learning_rate": 5.3428622980310755e-06,
+      "loss": 0.3309,
+      "step": 6326
+    },
+    {
+      "epoch": 1.591297786720322,
+      "grad_norm": 0.2830161452293396,
+      "learning_rate": 5.341402480025808e-06,
+      "loss": 0.3416,
+      "step": 6327
+    },
+    {
+      "epoch": 1.591549295774648,
+      "grad_norm": 0.3165720999240875,
+      "learning_rate": 5.339942632781553e-06,
+      "loss": 0.3532,
+      "step": 6328
+    },
+    {
+      "epoch": 1.5918008048289738,
+      "grad_norm": 0.3214383125305176,
+      "learning_rate": 5.338482756423339e-06,
+      "loss": 0.3412,
+      "step": 6329
+    },
+    {
+      "epoch": 1.5920523138832998,
+      "grad_norm": 0.2742040455341339,
+      "learning_rate": 5.337022851076193e-06,
+      "loss": 0.3516,
+      "step": 6330
+    },
+    {
+      "epoch": 1.5923038229376258,
+      "grad_norm": 0.2959214150905609,
+      "learning_rate": 5.33556291686515e-06,
+      "loss": 0.3405,
+      "step": 6331
+    },
+    {
+      "epoch": 1.5925553319919517,
+      "grad_norm": 0.28972509503364563,
+      "learning_rate": 5.334102953915242e-06,
+      "loss": 0.3324,
+      "step": 6332
+    },
+    {
+      "epoch": 1.5928068410462777,
+      "grad_norm": 0.34085187315940857,
+      "learning_rate": 5.332642962351505e-06,
+      "loss": 0.3419,
+      "step": 6333
+    },
+    {
+      "epoch": 1.5930583501006037,
+      "grad_norm": 0.32344910502433777,
+      "learning_rate": 5.331182942298981e-06,
+      "loss": 0.3387,
+      "step": 6334
+    },
+    {
+      "epoch": 1.5933098591549295,
+      "grad_norm": 0.3067835867404938,
+      "learning_rate": 5.329722893882708e-06,
+      "loss": 0.3325,
+      "step": 6335
+    },
+    {
+      "epoch": 1.5935613682092555,
+      "grad_norm": 0.31385210156440735,
+      "learning_rate": 5.328262817227733e-06,
+      "loss": 0.3588,
+      "step": 6336
+    },
+    {
+      "epoch": 1.5938128772635816,
+      "grad_norm": 0.30561771988868713,
+      "learning_rate": 5.326802712459101e-06,
+      "loss": 0.3438,
+      "step": 6337
+    },
+    {
+      "epoch": 1.5940643863179074,
+      "grad_norm": 0.30326759815216064,
+      "learning_rate": 5.325342579701862e-06,
+      "loss": 0.3669,
+      "step": 6338
+    },
+    {
+      "epoch": 1.5943158953722334,
+      "grad_norm": 0.31566211581230164,
+      "learning_rate": 5.323882419081066e-06,
+      "loss": 0.3275,
+      "step": 6339
+    },
+    {
+      "epoch": 1.5945674044265594,
+      "grad_norm": 0.32427820563316345,
+      "learning_rate": 5.3224222307217665e-06,
+      "loss": 0.3351,
+      "step": 6340
+    },
+    {
+      "epoch": 1.5948189134808852,
+      "grad_norm": 0.33437636494636536,
+      "learning_rate": 5.32096201474902e-06,
+      "loss": 0.3573,
+      "step": 6341
+    },
+    {
+      "epoch": 1.5950704225352113,
+      "grad_norm": 0.3199196457862854,
+      "learning_rate": 5.319501771287885e-06,
+      "loss": 0.3591,
+      "step": 6342
+    },
+    {
+      "epoch": 1.5953219315895373,
+      "grad_norm": 0.332581102848053,
+      "learning_rate": 5.318041500463423e-06,
+      "loss": 0.3374,
+      "step": 6343
+    },
+    {
+      "epoch": 1.595573440643863,
+      "grad_norm": 0.3242298662662506,
+      "learning_rate": 5.316581202400694e-06,
+      "loss": 0.3266,
+      "step": 6344
+    },
+    {
+      "epoch": 1.595824949698189,
+      "grad_norm": 0.3219207525253296,
+      "learning_rate": 5.315120877224767e-06,
+      "loss": 0.3477,
+      "step": 6345
+    },
+    {
+      "epoch": 1.5960764587525151,
+      "grad_norm": 0.3285122215747833,
+      "learning_rate": 5.313660525060709e-06,
+      "loss": 0.3691,
+      "step": 6346
+    },
+    {
+      "epoch": 1.596327967806841,
+      "grad_norm": 0.32978108525276184,
+      "learning_rate": 5.312200146033588e-06,
+      "loss": 0.3434,
+      "step": 6347
+    },
+    {
+      "epoch": 1.596579476861167,
+      "grad_norm": 0.3488015830516815,
+      "learning_rate": 5.310739740268478e-06,
+      "loss": 0.3484,
+      "step": 6348
+    },
+    {
+      "epoch": 1.596830985915493,
+      "grad_norm": 0.31981801986694336,
+      "learning_rate": 5.309279307890453e-06,
+      "loss": 0.3259,
+      "step": 6349
+    },
+    {
+      "epoch": 1.5970824949698188,
+      "grad_norm": 0.31969207525253296,
+      "learning_rate": 5.3078188490245905e-06,
+      "loss": 0.3416,
+      "step": 6350
+    },
+    {
+      "epoch": 1.5973340040241448,
+      "grad_norm": 0.3058636784553528,
+      "learning_rate": 5.30635836379597e-06,
+      "loss": 0.3111,
+      "step": 6351
+    },
+    {
+      "epoch": 1.5975855130784709,
+      "grad_norm": 0.3224903345108032,
+      "learning_rate": 5.304897852329671e-06,
+      "loss": 0.3454,
+      "step": 6352
+    },
+    {
+      "epoch": 1.5978370221327967,
+      "grad_norm": 0.33611172437667847,
+      "learning_rate": 5.303437314750779e-06,
+      "loss": 0.3447,
+      "step": 6353
+    },
+    {
+      "epoch": 1.5980885311871227,
+      "grad_norm": 0.3095466196537018,
+      "learning_rate": 5.301976751184379e-06,
+      "loss": 0.3523,
+      "step": 6354
+    },
+    {
+      "epoch": 1.5983400402414487,
+      "grad_norm": 0.3098112642765045,
+      "learning_rate": 5.300516161755559e-06,
+      "loss": 0.3419,
+      "step": 6355
+    },
+    {
+      "epoch": 1.5985915492957745,
+      "grad_norm": 0.3331829309463501,
+      "learning_rate": 5.29905554658941e-06,
+      "loss": 0.3652,
+      "step": 6356
+    },
+    {
+      "epoch": 1.5988430583501008,
+      "grad_norm": 0.30637791752815247,
+      "learning_rate": 5.297594905811024e-06,
+      "loss": 0.3591,
+      "step": 6357
+    },
+    {
+      "epoch": 1.5990945674044266,
+      "grad_norm": 0.3300563395023346,
+      "learning_rate": 5.296134239545497e-06,
+      "loss": 0.3505,
+      "step": 6358
+    },
+    {
+      "epoch": 1.5993460764587524,
+      "grad_norm": 0.34214162826538086,
+      "learning_rate": 5.294673547917925e-06,
+      "loss": 0.3663,
+      "step": 6359
+    },
+    {
+      "epoch": 1.5995975855130786,
+      "grad_norm": 0.3289168179035187,
+      "learning_rate": 5.293212831053407e-06,
+      "loss": 0.3356,
+      "step": 6360
+    },
+    {
+      "epoch": 1.5998490945674044,
+      "grad_norm": 0.3313751816749573,
+      "learning_rate": 5.291752089077044e-06,
+      "loss": 0.3586,
+      "step": 6361
+    },
+    {
+      "epoch": 1.6001006036217302,
+      "grad_norm": 0.32884520292282104,
+      "learning_rate": 5.29029132211394e-06,
+      "loss": 0.3481,
+      "step": 6362
+    },
+    {
+      "epoch": 1.6003521126760565,
+      "grad_norm": 0.3014932870864868,
+      "learning_rate": 5.2888305302891996e-06,
+      "loss": 0.3677,
+      "step": 6363
+    },
+    {
+      "epoch": 1.6006036217303823,
+      "grad_norm": 0.3143293857574463,
+      "learning_rate": 5.287369713727933e-06,
+      "loss": 0.3352,
+      "step": 6364
+    },
+    {
+      "epoch": 1.600855130784708,
+      "grad_norm": 0.30233773589134216,
+      "learning_rate": 5.285908872555247e-06,
+      "loss": 0.3466,
+      "step": 6365
+    },
+    {
+      "epoch": 1.6011066398390343,
+      "grad_norm": 0.3170585036277771,
+      "learning_rate": 5.284448006896252e-06,
+      "loss": 0.3253,
+      "step": 6366
+    },
+    {
+      "epoch": 1.6013581488933601,
+      "grad_norm": 0.3214609920978546,
+      "learning_rate": 5.282987116876068e-06,
+      "loss": 0.3341,
+      "step": 6367
+    },
+    {
+      "epoch": 1.6016096579476862,
+      "grad_norm": 0.3339797854423523,
+      "learning_rate": 5.281526202619808e-06,
+      "loss": 0.3594,
+      "step": 6368
+    },
+    {
+      "epoch": 1.6018611670020122,
+      "grad_norm": 0.30358412861824036,
+      "learning_rate": 5.2800652642525885e-06,
+      "loss": 0.3265,
+      "step": 6369
+    },
+    {
+      "epoch": 1.602112676056338,
+      "grad_norm": 0.3258301615715027,
+      "learning_rate": 5.278604301899531e-06,
+      "loss": 0.335,
+      "step": 6370
+    },
+    {
+      "epoch": 1.602364185110664,
+      "grad_norm": 0.32706478238105774,
+      "learning_rate": 5.27714331568576e-06,
+      "loss": 0.3561,
+      "step": 6371
+    },
+    {
+      "epoch": 1.60261569416499,
+      "grad_norm": 0.3139190077781677,
+      "learning_rate": 5.275682305736396e-06,
+      "loss": 0.3447,
+      "step": 6372
+    },
+    {
+      "epoch": 1.6028672032193159,
+      "grad_norm": 0.30157265067100525,
+      "learning_rate": 5.274221272176569e-06,
+      "loss": 0.3914,
+      "step": 6373
+    },
+    {
+      "epoch": 1.6031187122736419,
+      "grad_norm": 0.311866819858551,
+      "learning_rate": 5.2727602151314035e-06,
+      "loss": 0.3594,
+      "step": 6374
+    },
+    {
+      "epoch": 1.603370221327968,
+      "grad_norm": 0.29979294538497925,
+      "learning_rate": 5.271299134726034e-06,
+      "loss": 0.3382,
+      "step": 6375
+    },
+    {
+      "epoch": 1.6036217303822937,
+      "grad_norm": 0.3057425320148468,
+      "learning_rate": 5.269838031085588e-06,
+      "loss": 0.3378,
+      "step": 6376
+    },
+    {
+      "epoch": 1.6038732394366197,
+      "grad_norm": 0.33399319648742676,
+      "learning_rate": 5.268376904335204e-06,
+      "loss": 0.3625,
+      "step": 6377
+    },
+    {
+      "epoch": 1.6041247484909458,
+      "grad_norm": 0.31966760754585266,
+      "learning_rate": 5.266915754600018e-06,
+      "loss": 0.3683,
+      "step": 6378
+    },
+    {
+      "epoch": 1.6043762575452716,
+      "grad_norm": 0.3125464916229248,
+      "learning_rate": 5.265454582005167e-06,
+      "loss": 0.3498,
+      "step": 6379
+    },
+    {
+      "epoch": 1.6046277665995976,
+      "grad_norm": 0.30603522062301636,
+      "learning_rate": 5.263993386675792e-06,
+      "loss": 0.3682,
+      "step": 6380
+    },
+    {
+      "epoch": 1.6048792756539236,
+      "grad_norm": 0.3024091422557831,
+      "learning_rate": 5.2625321687370345e-06,
+      "loss": 0.3384,
+      "step": 6381
+    },
+    {
+      "epoch": 1.6051307847082494,
+      "grad_norm": 0.2936726212501526,
+      "learning_rate": 5.261070928314039e-06,
+      "loss": 0.373,
+      "step": 6382
+    },
+    {
+      "epoch": 1.6053822937625755,
+      "grad_norm": 0.31139662861824036,
+      "learning_rate": 5.259609665531951e-06,
+      "loss": 0.3445,
+      "step": 6383
+    },
+    {
+      "epoch": 1.6056338028169015,
+      "grad_norm": 0.30623286962509155,
+      "learning_rate": 5.258148380515922e-06,
+      "loss": 0.3562,
+      "step": 6384
+    },
+    {
+      "epoch": 1.6058853118712273,
+      "grad_norm": 0.3077852129936218,
+      "learning_rate": 5.256687073391097e-06,
+      "loss": 0.3313,
+      "step": 6385
+    },
+    {
+      "epoch": 1.6061368209255533,
+      "grad_norm": 0.28741592168807983,
+      "learning_rate": 5.255225744282631e-06,
+      "loss": 0.354,
+      "step": 6386
+    },
+    {
+      "epoch": 1.6063883299798793,
+      "grad_norm": 0.32665884494781494,
+      "learning_rate": 5.253764393315674e-06,
+      "loss": 0.3458,
+      "step": 6387
+    },
+    {
+      "epoch": 1.6066398390342052,
+      "grad_norm": 0.31474220752716064,
+      "learning_rate": 5.252303020615387e-06,
+      "loss": 0.3425,
+      "step": 6388
+    },
+    {
+      "epoch": 1.6068913480885312,
+      "grad_norm": 0.3298763930797577,
+      "learning_rate": 5.250841626306924e-06,
+      "loss": 0.3634,
+      "step": 6389
+    },
+    {
+      "epoch": 1.6071428571428572,
+      "grad_norm": 0.2954833209514618,
+      "learning_rate": 5.249380210515446e-06,
+      "loss": 0.3702,
+      "step": 6390
+    },
+    {
+      "epoch": 1.607394366197183,
+      "grad_norm": 0.3127298057079315,
+      "learning_rate": 5.247918773366112e-06,
+      "loss": 0.348,
+      "step": 6391
+    },
+    {
+      "epoch": 1.607645875251509,
+      "grad_norm": 0.2956436574459076,
+      "learning_rate": 5.246457314984086e-06,
+      "loss": 0.3504,
+      "step": 6392
+    },
+    {
+      "epoch": 1.607897384305835,
+      "grad_norm": 0.35721251368522644,
+      "learning_rate": 5.2449958354945326e-06,
+      "loss": 0.3357,
+      "step": 6393
+    },
+    {
+      "epoch": 1.6081488933601609,
+      "grad_norm": 0.31743982434272766,
+      "learning_rate": 5.24353433502262e-06,
+      "loss": 0.3412,
+      "step": 6394
+    },
+    {
+      "epoch": 1.608400402414487,
+      "grad_norm": 0.31196045875549316,
+      "learning_rate": 5.242072813693514e-06,
+      "loss": 0.3253,
+      "step": 6395
+    },
+    {
+      "epoch": 1.608651911468813,
+      "grad_norm": 0.29665273427963257,
+      "learning_rate": 5.240611271632386e-06,
+      "loss": 0.3622,
+      "step": 6396
+    },
+    {
+      "epoch": 1.6089034205231387,
+      "grad_norm": 0.36455845832824707,
+      "learning_rate": 5.239149708964409e-06,
+      "loss": 0.3435,
+      "step": 6397
+    },
+    {
+      "epoch": 1.6091549295774648,
+      "grad_norm": 0.3024010956287384,
+      "learning_rate": 5.237688125814752e-06,
+      "loss": 0.3549,
+      "step": 6398
+    },
+    {
+      "epoch": 1.6094064386317908,
+      "grad_norm": 0.3012137711048126,
+      "learning_rate": 5.236226522308596e-06,
+      "loss": 0.3332,
+      "step": 6399
+    },
+    {
+      "epoch": 1.6096579476861166,
+      "grad_norm": 0.31425225734710693,
+      "learning_rate": 5.234764898571118e-06,
+      "loss": 0.3436,
+      "step": 6400
+    },
+    {
+      "epoch": 1.6099094567404426,
+      "grad_norm": 0.3358578383922577,
+      "learning_rate": 5.233303254727493e-06,
+      "loss": 0.3441,
+      "step": 6401
+    },
+    {
+      "epoch": 1.6101609657947686,
+      "grad_norm": 0.3274340033531189,
+      "learning_rate": 5.231841590902905e-06,
+      "loss": 0.3244,
+      "step": 6402
+    },
+    {
+      "epoch": 1.6104124748490944,
+      "grad_norm": 0.3106974959373474,
+      "learning_rate": 5.230379907222535e-06,
+      "loss": 0.3764,
+      "step": 6403
+    },
+    {
+      "epoch": 1.6106639839034205,
+      "grad_norm": 0.2865849435329437,
+      "learning_rate": 5.228918203811566e-06,
+      "loss": 0.3393,
+      "step": 6404
+    },
+    {
+      "epoch": 1.6109154929577465,
+      "grad_norm": 0.31629478931427,
+      "learning_rate": 5.227456480795187e-06,
+      "loss": 0.3581,
+      "step": 6405
+    },
+    {
+      "epoch": 1.6111670020120723,
+      "grad_norm": 0.3469734787940979,
+      "learning_rate": 5.225994738298582e-06,
+      "loss": 0.3523,
+      "step": 6406
+    },
+    {
+      "epoch": 1.6114185110663986,
+      "grad_norm": 0.31348875164985657,
+      "learning_rate": 5.224532976446941e-06,
+      "loss": 0.3381,
+      "step": 6407
+    },
+    {
+      "epoch": 1.6116700201207244,
+      "grad_norm": 0.29473093152046204,
+      "learning_rate": 5.223071195365456e-06,
+      "loss": 0.3483,
+      "step": 6408
+    },
+    {
+      "epoch": 1.6119215291750502,
+      "grad_norm": 0.3042793869972229,
+      "learning_rate": 5.221609395179319e-06,
+      "loss": 0.3646,
+      "step": 6409
+    },
+    {
+      "epoch": 1.6121730382293764,
+      "grad_norm": 0.3180801272392273,
+      "learning_rate": 5.220147576013724e-06,
+      "loss": 0.3661,
+      "step": 6410
+    },
+    {
+      "epoch": 1.6124245472837022,
+      "grad_norm": 0.318566232919693,
+      "learning_rate": 5.218685737993865e-06,
+      "loss": 0.3538,
+      "step": 6411
+    },
+    {
+      "epoch": 1.612676056338028,
+      "grad_norm": 0.3094564974308014,
+      "learning_rate": 5.217223881244942e-06,
+      "loss": 0.3289,
+      "step": 6412
+    },
+    {
+      "epoch": 1.6129275653923543,
+      "grad_norm": 0.3494105041027069,
+      "learning_rate": 5.215762005892151e-06,
+      "loss": 0.3474,
+      "step": 6413
+    },
+    {
+      "epoch": 1.61317907444668,
+      "grad_norm": 0.312987357378006,
+      "learning_rate": 5.214300112060695e-06,
+      "loss": 0.3601,
+      "step": 6414
+    },
+    {
+      "epoch": 1.6134305835010059,
+      "grad_norm": 0.30890733003616333,
+      "learning_rate": 5.212838199875775e-06,
+      "loss": 0.3478,
+      "step": 6415
+    },
+    {
+      "epoch": 1.6136820925553321,
+      "grad_norm": 0.3125660717487335,
+      "learning_rate": 5.211376269462594e-06,
+      "loss": 0.3657,
+      "step": 6416
+    },
+    {
+      "epoch": 1.613933601609658,
+      "grad_norm": 0.32715532183647156,
+      "learning_rate": 5.209914320946359e-06,
+      "loss": 0.3662,
+      "step": 6417
+    },
+    {
+      "epoch": 1.614185110663984,
+      "grad_norm": 0.3133423626422882,
+      "learning_rate": 5.208452354452275e-06,
+      "loss": 0.3472,
+      "step": 6418
+    },
+    {
+      "epoch": 1.61443661971831,
+      "grad_norm": 0.3222804069519043,
+      "learning_rate": 5.20699037010555e-06,
+      "loss": 0.3355,
+      "step": 6419
+    },
+    {
+      "epoch": 1.6146881287726358,
+      "grad_norm": 0.3086826205253601,
+      "learning_rate": 5.205528368031395e-06,
+      "loss": 0.3552,
+      "step": 6420
+    },
+    {
+      "epoch": 1.6149396378269618,
+      "grad_norm": 0.3171122968196869,
+      "learning_rate": 5.204066348355022e-06,
+      "loss": 0.3557,
+      "step": 6421
+    },
+    {
+      "epoch": 1.6151911468812878,
+      "grad_norm": 0.31034934520721436,
+      "learning_rate": 5.202604311201642e-06,
+      "loss": 0.3478,
+      "step": 6422
+    },
+    {
+      "epoch": 1.6154426559356136,
+      "grad_norm": 0.29711443185806274,
+      "learning_rate": 5.201142256696472e-06,
+      "loss": 0.349,
+      "step": 6423
+    },
+    {
+      "epoch": 1.6156941649899397,
+      "grad_norm": 0.3413085639476776,
+      "learning_rate": 5.199680184964725e-06,
+      "loss": 0.3472,
+      "step": 6424
+    },
+    {
+      "epoch": 1.6159456740442657,
+      "grad_norm": 0.2911038100719452,
+      "learning_rate": 5.198218096131619e-06,
+      "loss": 0.342,
+      "step": 6425
+    },
+    {
+      "epoch": 1.6161971830985915,
+      "grad_norm": 0.291364461183548,
+      "learning_rate": 5.196755990322373e-06,
+      "loss": 0.3439,
+      "step": 6426
+    },
+    {
+      "epoch": 1.6164486921529175,
+      "grad_norm": 0.3112189769744873,
+      "learning_rate": 5.195293867662208e-06,
+      "loss": 0.3266,
+      "step": 6427
+    },
+    {
+      "epoch": 1.6167002012072436,
+      "grad_norm": 0.33983948826789856,
+      "learning_rate": 5.193831728276345e-06,
+      "loss": 0.3386,
+      "step": 6428
+    },
+    {
+      "epoch": 1.6169517102615694,
+      "grad_norm": 0.3202919363975525,
+      "learning_rate": 5.192369572290007e-06,
+      "loss": 0.3472,
+      "step": 6429
+    },
+    {
+      "epoch": 1.6172032193158954,
+      "grad_norm": 0.31747010350227356,
+      "learning_rate": 5.190907399828418e-06,
+      "loss": 0.3429,
+      "step": 6430
+    },
+    {
+      "epoch": 1.6174547283702214,
+      "grad_norm": 0.3446648120880127,
+      "learning_rate": 5.189445211016804e-06,
+      "loss": 0.3562,
+      "step": 6431
+    },
+    {
+      "epoch": 1.6177062374245472,
+      "grad_norm": 0.3192874491214752,
+      "learning_rate": 5.187983005980393e-06,
+      "loss": 0.3591,
+      "step": 6432
+    },
+    {
+      "epoch": 1.6179577464788732,
+      "grad_norm": 0.3468160331249237,
+      "learning_rate": 5.186520784844416e-06,
+      "loss": 0.3513,
+      "step": 6433
+    },
+    {
+      "epoch": 1.6182092555331993,
+      "grad_norm": 0.32596200704574585,
+      "learning_rate": 5.185058547734098e-06,
+      "loss": 0.3382,
+      "step": 6434
+    },
+    {
+      "epoch": 1.618460764587525,
+      "grad_norm": 0.2944319546222687,
+      "learning_rate": 5.1835962947746744e-06,
+      "loss": 0.3301,
+      "step": 6435
+    },
+    {
+      "epoch": 1.618712273641851,
+      "grad_norm": 0.32907602190971375,
+      "learning_rate": 5.1821340260913765e-06,
+      "loss": 0.3355,
+      "step": 6436
+    },
+    {
+      "epoch": 1.6189637826961771,
+      "grad_norm": 0.3250274956226349,
+      "learning_rate": 5.180671741809439e-06,
+      "loss": 0.3566,
+      "step": 6437
+    },
+    {
+      "epoch": 1.619215291750503,
+      "grad_norm": 0.3051334321498871,
+      "learning_rate": 5.179209442054096e-06,
+      "loss": 0.3507,
+      "step": 6438
+    },
+    {
+      "epoch": 1.619466800804829,
+      "grad_norm": 0.3153745234012604,
+      "learning_rate": 5.177747126950587e-06,
+      "loss": 0.3558,
+      "step": 6439
+    },
+    {
+      "epoch": 1.619718309859155,
+      "grad_norm": 0.31299081444740295,
+      "learning_rate": 5.176284796624147e-06,
+      "loss": 0.3314,
+      "step": 6440
+    },
+    {
+      "epoch": 1.6199698189134808,
+      "grad_norm": 0.29618367552757263,
+      "learning_rate": 5.174822451200018e-06,
+      "loss": 0.3563,
+      "step": 6441
+    },
+    {
+      "epoch": 1.6202213279678068,
+      "grad_norm": 0.3016599118709564,
+      "learning_rate": 5.173360090803437e-06,
+      "loss": 0.3545,
+      "step": 6442
+    },
+    {
+      "epoch": 1.6204728370221329,
+      "grad_norm": 0.30213463306427,
+      "learning_rate": 5.1718977155596515e-06,
+      "loss": 0.348,
+      "step": 6443
+    },
+    {
+      "epoch": 1.6207243460764587,
+      "grad_norm": 0.3281251788139343,
+      "learning_rate": 5.170435325593902e-06,
+      "loss": 0.3423,
+      "step": 6444
+    },
+    {
+      "epoch": 1.6209758551307847,
+      "grad_norm": 0.33480846881866455,
+      "learning_rate": 5.168972921031433e-06,
+      "loss": 0.3396,
+      "step": 6445
+    },
+    {
+      "epoch": 1.6212273641851107,
+      "grad_norm": 0.32940176129341125,
+      "learning_rate": 5.1675105019974905e-06,
+      "loss": 0.3567,
+      "step": 6446
+    },
+    {
+      "epoch": 1.6214788732394365,
+      "grad_norm": 0.3426181674003601,
+      "learning_rate": 5.166048068617321e-06,
+      "loss": 0.3681,
+      "step": 6447
+    },
+    {
+      "epoch": 1.6217303822937625,
+      "grad_norm": 0.3707010746002197,
+      "learning_rate": 5.164585621016174e-06,
+      "loss": 0.3475,
+      "step": 6448
+    },
+    {
+      "epoch": 1.6219818913480886,
+      "grad_norm": 0.32672780752182007,
+      "learning_rate": 5.163123159319298e-06,
+      "loss": 0.3372,
+      "step": 6449
+    },
+    {
+      "epoch": 1.6222334004024144,
+      "grad_norm": 0.32823729515075684,
+      "learning_rate": 5.161660683651943e-06,
+      "loss": 0.3486,
+      "step": 6450
+    },
+    {
+      "epoch": 1.6224849094567404,
+      "grad_norm": 0.3079800009727478,
+      "learning_rate": 5.160198194139362e-06,
+      "loss": 0.3535,
+      "step": 6451
+    },
+    {
+      "epoch": 1.6227364185110664,
+      "grad_norm": 0.34681087732315063,
+      "learning_rate": 5.158735690906808e-06,
+      "loss": 0.3498,
+      "step": 6452
+    },
+    {
+      "epoch": 1.6229879275653922,
+      "grad_norm": 0.33672547340393066,
+      "learning_rate": 5.157273174079535e-06,
+      "loss": 0.3343,
+      "step": 6453
+    },
+    {
+      "epoch": 1.6232394366197183,
+      "grad_norm": 0.34650522470474243,
+      "learning_rate": 5.155810643782798e-06,
+      "loss": 0.3391,
+      "step": 6454
+    },
+    {
+      "epoch": 1.6234909456740443,
+      "grad_norm": 0.3248770236968994,
+      "learning_rate": 5.154348100141855e-06,
+      "loss": 0.3862,
+      "step": 6455
+    },
+    {
+      "epoch": 1.62374245472837,
+      "grad_norm": 0.3029992878437042,
+      "learning_rate": 5.152885543281964e-06,
+      "loss": 0.3286,
+      "step": 6456
+    },
+    {
+      "epoch": 1.6239939637826963,
+      "grad_norm": 0.32535600662231445,
+      "learning_rate": 5.151422973328381e-06,
+      "loss": 0.3105,
+      "step": 6457
+    },
+    {
+      "epoch": 1.6242454728370221,
+      "grad_norm": 0.3738095164299011,
+      "learning_rate": 5.149960390406368e-06,
+      "loss": 0.3534,
+      "step": 6458
+    },
+    {
+      "epoch": 1.624496981891348,
+      "grad_norm": 0.31076717376708984,
+      "learning_rate": 5.1484977946411855e-06,
+      "loss": 0.3666,
+      "step": 6459
+    },
+    {
+      "epoch": 1.6247484909456742,
+      "grad_norm": 0.32721608877182007,
+      "learning_rate": 5.147035186158096e-06,
+      "loss": 0.3678,
+      "step": 6460
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.3344939947128296,
+      "learning_rate": 5.145572565082363e-06,
+      "loss": 0.3726,
+      "step": 6461
+    },
+    {
+      "epoch": 1.6252515090543258,
+      "grad_norm": 0.29688549041748047,
+      "learning_rate": 5.144109931539251e-06,
+      "loss": 0.3751,
+      "step": 6462
+    },
+    {
+      "epoch": 1.625503018108652,
+      "grad_norm": 0.29007625579833984,
+      "learning_rate": 5.142647285654023e-06,
+      "loss": 0.3259,
+      "step": 6463
+    },
+    {
+      "epoch": 1.6257545271629779,
+      "grad_norm": 0.3192233443260193,
+      "learning_rate": 5.14118462755195e-06,
+      "loss": 0.378,
+      "step": 6464
+    },
+    {
+      "epoch": 1.6260060362173037,
+      "grad_norm": 0.3485788106918335,
+      "learning_rate": 5.139721957358295e-06,
+      "loss": 0.3603,
+      "step": 6465
+    },
+    {
+      "epoch": 1.62625754527163,
+      "grad_norm": 0.30432939529418945,
+      "learning_rate": 5.138259275198329e-06,
+      "loss": 0.3284,
+      "step": 6466
+    },
+    {
+      "epoch": 1.6265090543259557,
+      "grad_norm": 0.3238847851753235,
+      "learning_rate": 5.1367965811973204e-06,
+      "loss": 0.3459,
+      "step": 6467
+    },
+    {
+      "epoch": 1.6267605633802817,
+      "grad_norm": 0.2998338043689728,
+      "learning_rate": 5.135333875480541e-06,
+      "loss": 0.3437,
+      "step": 6468
+    },
+    {
+      "epoch": 1.6270120724346078,
+      "grad_norm": 0.3004499673843384,
+      "learning_rate": 5.133871158173262e-06,
+      "loss": 0.3473,
+      "step": 6469
+    },
+    {
+      "epoch": 1.6272635814889336,
+      "grad_norm": 0.3121063709259033,
+      "learning_rate": 5.132408429400755e-06,
+      "loss": 0.3421,
+      "step": 6470
+    },
+    {
+      "epoch": 1.6275150905432596,
+      "grad_norm": 0.3265574276447296,
+      "learning_rate": 5.130945689288295e-06,
+      "loss": 0.3465,
+      "step": 6471
+    },
+    {
+      "epoch": 1.6277665995975856,
+      "grad_norm": 0.32837405800819397,
+      "learning_rate": 5.129482937961155e-06,
+      "loss": 0.3612,
+      "step": 6472
+    },
+    {
+      "epoch": 1.6280181086519114,
+      "grad_norm": 0.3113304078578949,
+      "learning_rate": 5.128020175544612e-06,
+      "loss": 0.3388,
+      "step": 6473
+    },
+    {
+      "epoch": 1.6282696177062375,
+      "grad_norm": 0.32058820128440857,
+      "learning_rate": 5.126557402163943e-06,
+      "loss": 0.3485,
+      "step": 6474
+    },
+    {
+      "epoch": 1.6285211267605635,
+      "grad_norm": 0.34431353211402893,
+      "learning_rate": 5.125094617944424e-06,
+      "loss": 0.3268,
+      "step": 6475
+    },
+    {
+      "epoch": 1.6287726358148893,
+      "grad_norm": 0.3116149604320526,
+      "learning_rate": 5.123631823011333e-06,
+      "loss": 0.3588,
+      "step": 6476
+    },
+    {
+      "epoch": 1.6290241448692153,
+      "grad_norm": 0.2909471392631531,
+      "learning_rate": 5.122169017489949e-06,
+      "loss": 0.3541,
+      "step": 6477
+    },
+    {
+      "epoch": 1.6292756539235413,
+      "grad_norm": 0.35828930139541626,
+      "learning_rate": 5.120706201505554e-06,
+      "loss": 0.3597,
+      "step": 6478
+    },
+    {
+      "epoch": 1.6295271629778671,
+      "grad_norm": 0.28314337134361267,
+      "learning_rate": 5.119243375183427e-06,
+      "loss": 0.3599,
+      "step": 6479
+    },
+    {
+      "epoch": 1.6297786720321932,
+      "grad_norm": 0.2970763146877289,
+      "learning_rate": 5.1177805386488525e-06,
+      "loss": 0.3418,
+      "step": 6480
+    },
+    {
+      "epoch": 1.6300301810865192,
+      "grad_norm": 0.3072241246700287,
+      "learning_rate": 5.116317692027111e-06,
+      "loss": 0.3745,
+      "step": 6481
+    },
+    {
+      "epoch": 1.630281690140845,
+      "grad_norm": 0.32199907302856445,
+      "learning_rate": 5.114854835443486e-06,
+      "loss": 0.355,
+      "step": 6482
+    },
+    {
+      "epoch": 1.630533199195171,
+      "grad_norm": 0.328183650970459,
+      "learning_rate": 5.113391969023264e-06,
+      "loss": 0.358,
+      "step": 6483
+    },
+    {
+      "epoch": 1.630784708249497,
+      "grad_norm": 0.3165997266769409,
+      "learning_rate": 5.1119290928917285e-06,
+      "loss": 0.3576,
+      "step": 6484
+    },
+    {
+      "epoch": 1.6310362173038229,
+      "grad_norm": 0.3254222273826599,
+      "learning_rate": 5.110466207174165e-06,
+      "loss": 0.3737,
+      "step": 6485
+    },
+    {
+      "epoch": 1.631287726358149,
+      "grad_norm": 0.3073231279850006,
+      "learning_rate": 5.109003311995864e-06,
+      "loss": 0.3435,
+      "step": 6486
+    },
+    {
+      "epoch": 1.631539235412475,
+      "grad_norm": 0.3206403851509094,
+      "learning_rate": 5.107540407482111e-06,
+      "loss": 0.3568,
+      "step": 6487
+    },
+    {
+      "epoch": 1.6317907444668007,
+      "grad_norm": 0.29873090982437134,
+      "learning_rate": 5.106077493758195e-06,
+      "loss": 0.3167,
+      "step": 6488
+    },
+    {
+      "epoch": 1.6320422535211268,
+      "grad_norm": 0.2889578342437744,
+      "learning_rate": 5.104614570949404e-06,
+      "loss": 0.3574,
+      "step": 6489
+    },
+    {
+      "epoch": 1.6322937625754528,
+      "grad_norm": 0.3086026608943939,
+      "learning_rate": 5.1031516391810306e-06,
+      "loss": 0.3079,
+      "step": 6490
+    },
+    {
+      "epoch": 1.6325452716297786,
+      "grad_norm": 0.2884649634361267,
+      "learning_rate": 5.101688698578364e-06,
+      "loss": 0.3466,
+      "step": 6491
+    },
+    {
+      "epoch": 1.6327967806841046,
+      "grad_norm": 0.31144869327545166,
+      "learning_rate": 5.100225749266698e-06,
+      "loss": 0.3572,
+      "step": 6492
+    },
+    {
+      "epoch": 1.6330482897384306,
+      "grad_norm": 0.3050963282585144,
+      "learning_rate": 5.098762791371322e-06,
+      "loss": 0.3376,
+      "step": 6493
+    },
+    {
+      "epoch": 1.6332997987927564,
+      "grad_norm": 0.2924579083919525,
+      "learning_rate": 5.097299825017532e-06,
+      "loss": 0.3409,
+      "step": 6494
+    },
+    {
+      "epoch": 1.6335513078470825,
+      "grad_norm": 0.31079205870628357,
+      "learning_rate": 5.09583685033062e-06,
+      "loss": 0.3827,
+      "step": 6495
+    },
+    {
+      "epoch": 1.6338028169014085,
+      "grad_norm": 0.30637675523757935,
+      "learning_rate": 5.09437386743588e-06,
+      "loss": 0.3354,
+      "step": 6496
+    },
+    {
+      "epoch": 1.6340543259557343,
+      "grad_norm": 0.29085665941238403,
+      "learning_rate": 5.09291087645861e-06,
+      "loss": 0.349,
+      "step": 6497
+    },
+    {
+      "epoch": 1.6343058350100603,
+      "grad_norm": 0.2851710915565491,
+      "learning_rate": 5.091447877524105e-06,
+      "loss": 0.3623,
+      "step": 6498
+    },
+    {
+      "epoch": 1.6345573440643864,
+      "grad_norm": 0.29473739862442017,
+      "learning_rate": 5.089984870757661e-06,
+      "loss": 0.3461,
+      "step": 6499
+    },
+    {
+      "epoch": 1.6348088531187122,
+      "grad_norm": 0.3163786828517914,
+      "learning_rate": 5.088521856284576e-06,
+      "loss": 0.3542,
+      "step": 6500
+    },
+    {
+      "epoch": 1.6350603621730382,
+      "grad_norm": 0.3166099786758423,
+      "learning_rate": 5.087058834230148e-06,
+      "loss": 0.3335,
+      "step": 6501
+    },
+    {
+      "epoch": 1.6353118712273642,
+      "grad_norm": 0.3093219995498657,
+      "learning_rate": 5.0855958047196744e-06,
+      "loss": 0.3258,
+      "step": 6502
+    },
+    {
+      "epoch": 1.63556338028169,
+      "grad_norm": 0.3152346611022949,
+      "learning_rate": 5.084132767878457e-06,
+      "loss": 0.3328,
+      "step": 6503
+    },
+    {
+      "epoch": 1.635814889336016,
+      "grad_norm": 0.2982290983200073,
+      "learning_rate": 5.082669723831793e-06,
+      "loss": 0.35,
+      "step": 6504
+    },
+    {
+      "epoch": 1.636066398390342,
+      "grad_norm": 0.3094666004180908,
+      "learning_rate": 5.081206672704986e-06,
+      "loss": 0.3647,
+      "step": 6505
+    },
+    {
+      "epoch": 1.6363179074446679,
+      "grad_norm": 0.31432268023490906,
+      "learning_rate": 5.079743614623334e-06,
+      "loss": 0.3449,
+      "step": 6506
+    },
+    {
+      "epoch": 1.6365694164989941,
+      "grad_norm": 0.32027164101600647,
+      "learning_rate": 5.07828054971214e-06,
+      "loss": 0.3704,
+      "step": 6507
+    },
+    {
+      "epoch": 1.63682092555332,
+      "grad_norm": 0.30292704701423645,
+      "learning_rate": 5.076817478096707e-06,
+      "loss": 0.3522,
+      "step": 6508
+    },
+    {
+      "epoch": 1.6370724346076457,
+      "grad_norm": 0.30981820821762085,
+      "learning_rate": 5.075354399902338e-06,
+      "loss": 0.3399,
+      "step": 6509
+    },
+    {
+      "epoch": 1.637323943661972,
+      "grad_norm": 0.3107950687408447,
+      "learning_rate": 5.073891315254337e-06,
+      "loss": 0.3439,
+      "step": 6510
+    },
+    {
+      "epoch": 1.6375754527162978,
+      "grad_norm": 0.29735639691352844,
+      "learning_rate": 5.072428224278005e-06,
+      "loss": 0.3326,
+      "step": 6511
+    },
+    {
+      "epoch": 1.6378269617706236,
+      "grad_norm": 0.34328925609588623,
+      "learning_rate": 5.07096512709865e-06,
+      "loss": 0.3414,
+      "step": 6512
+    },
+    {
+      "epoch": 1.6380784708249498,
+      "grad_norm": 0.306029736995697,
+      "learning_rate": 5.069502023841576e-06,
+      "loss": 0.3733,
+      "step": 6513
+    },
+    {
+      "epoch": 1.6383299798792756,
+      "grad_norm": 0.33722612261772156,
+      "learning_rate": 5.068038914632088e-06,
+      "loss": 0.3623,
+      "step": 6514
+    },
+    {
+      "epoch": 1.6385814889336014,
+      "grad_norm": 0.3304372727870941,
+      "learning_rate": 5.066575799595494e-06,
+      "loss": 0.3489,
+      "step": 6515
+    },
+    {
+      "epoch": 1.6388329979879277,
+      "grad_norm": 0.32422342896461487,
+      "learning_rate": 5.065112678857097e-06,
+      "loss": 0.3753,
+      "step": 6516
+    },
+    {
+      "epoch": 1.6390845070422535,
+      "grad_norm": 0.341206431388855,
+      "learning_rate": 5.063649552542208e-06,
+      "loss": 0.3352,
+      "step": 6517
+    },
+    {
+      "epoch": 1.6393360160965795,
+      "grad_norm": 0.30149486660957336,
+      "learning_rate": 5.062186420776132e-06,
+      "loss": 0.3636,
+      "step": 6518
+    },
+    {
+      "epoch": 1.6395875251509056,
+      "grad_norm": 0.31636372208595276,
+      "learning_rate": 5.060723283684178e-06,
+      "loss": 0.3196,
+      "step": 6519
+    },
+    {
+      "epoch": 1.6398390342052314,
+      "grad_norm": 0.3157700300216675,
+      "learning_rate": 5.0592601413916555e-06,
+      "loss": 0.3368,
+      "step": 6520
+    },
+    {
+      "epoch": 1.6400905432595574,
+      "grad_norm": 0.3472077548503876,
+      "learning_rate": 5.057796994023873e-06,
+      "loss": 0.3561,
+      "step": 6521
+    },
+    {
+      "epoch": 1.6403420523138834,
+      "grad_norm": 0.3037931025028229,
+      "learning_rate": 5.056333841706138e-06,
+      "loss": 0.3218,
+      "step": 6522
+    },
+    {
+      "epoch": 1.6405935613682092,
+      "grad_norm": 0.3278881907463074,
+      "learning_rate": 5.054870684563763e-06,
+      "loss": 0.3403,
+      "step": 6523
+    },
+    {
+      "epoch": 1.6408450704225352,
+      "grad_norm": 0.326613187789917,
+      "learning_rate": 5.053407522722057e-06,
+      "loss": 0.3386,
+      "step": 6524
+    },
+    {
+      "epoch": 1.6410965794768613,
+      "grad_norm": 0.31864145398139954,
+      "learning_rate": 5.0519443563063306e-06,
+      "loss": 0.3662,
+      "step": 6525
+    },
+    {
+      "epoch": 1.641348088531187,
+      "grad_norm": 0.29436439275741577,
+      "learning_rate": 5.0504811854418946e-06,
+      "loss": 0.3386,
+      "step": 6526
+    },
+    {
+      "epoch": 1.641599597585513,
+      "grad_norm": 0.3037160634994507,
+      "learning_rate": 5.049018010254062e-06,
+      "loss": 0.3429,
+      "step": 6527
+    },
+    {
+      "epoch": 1.6418511066398391,
+      "grad_norm": 0.29753682017326355,
+      "learning_rate": 5.047554830868142e-06,
+      "loss": 0.3608,
+      "step": 6528
+    },
+    {
+      "epoch": 1.642102615694165,
+      "grad_norm": 0.3260503113269806,
+      "learning_rate": 5.04609164740945e-06,
+      "loss": 0.3306,
+      "step": 6529
+    },
+    {
+      "epoch": 1.642354124748491,
+      "grad_norm": 0.34160879254341125,
+      "learning_rate": 5.044628460003296e-06,
+      "loss": 0.3614,
+      "step": 6530
+    },
+    {
+      "epoch": 1.642605633802817,
+      "grad_norm": 0.31074759364128113,
+      "learning_rate": 5.043165268774993e-06,
+      "loss": 0.3435,
+      "step": 6531
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.29596835374832153,
+      "learning_rate": 5.041702073849856e-06,
+      "loss": 0.3331,
+      "step": 6532
+    },
+    {
+      "epoch": 1.6431086519114688,
+      "grad_norm": 0.3086645305156708,
+      "learning_rate": 5.040238875353196e-06,
+      "loss": 0.3534,
+      "step": 6533
+    },
+    {
+      "epoch": 1.6433601609657948,
+      "grad_norm": 0.3223525285720825,
+      "learning_rate": 5.038775673410329e-06,
+      "loss": 0.3513,
+      "step": 6534
+    },
+    {
+      "epoch": 1.6436116700201207,
+      "grad_norm": 0.3153035044670105,
+      "learning_rate": 5.037312468146567e-06,
+      "loss": 0.3537,
+      "step": 6535
+    },
+    {
+      "epoch": 1.6438631790744467,
+      "grad_norm": 0.3239792287349701,
+      "learning_rate": 5.035849259687227e-06,
+      "loss": 0.3684,
+      "step": 6536
+    },
+    {
+      "epoch": 1.6441146881287727,
+      "grad_norm": 0.31517869234085083,
+      "learning_rate": 5.034386048157622e-06,
+      "loss": 0.3477,
+      "step": 6537
+    },
+    {
+      "epoch": 1.6443661971830985,
+      "grad_norm": 0.334823876619339,
+      "learning_rate": 5.032922833683066e-06,
+      "loss": 0.3537,
+      "step": 6538
+    },
+    {
+      "epoch": 1.6446177062374245,
+      "grad_norm": 0.2976732552051544,
+      "learning_rate": 5.031459616388874e-06,
+      "loss": 0.3599,
+      "step": 6539
+    },
+    {
+      "epoch": 1.6448692152917506,
+      "grad_norm": 0.3268083333969116,
+      "learning_rate": 5.029996396400365e-06,
+      "loss": 0.3449,
+      "step": 6540
+    },
+    {
+      "epoch": 1.6451207243460764,
+      "grad_norm": 0.3218528628349304,
+      "learning_rate": 5.028533173842851e-06,
+      "loss": 0.3338,
+      "step": 6541
+    },
+    {
+      "epoch": 1.6453722334004024,
+      "grad_norm": 0.28955259919166565,
+      "learning_rate": 5.02706994884165e-06,
+      "loss": 0.3325,
+      "step": 6542
+    },
+    {
+      "epoch": 1.6456237424547284,
+      "grad_norm": 0.2912788987159729,
+      "learning_rate": 5.025606721522077e-06,
+      "loss": 0.3465,
+      "step": 6543
+    },
+    {
+      "epoch": 1.6458752515090542,
+      "grad_norm": 0.33705586194992065,
+      "learning_rate": 5.024143492009449e-06,
+      "loss": 0.3647,
+      "step": 6544
+    },
+    {
+      "epoch": 1.6461267605633803,
+      "grad_norm": 0.32726725935935974,
+      "learning_rate": 5.022680260429082e-06,
+      "loss": 0.3435,
+      "step": 6545
+    },
+    {
+      "epoch": 1.6463782696177063,
+      "grad_norm": 0.32821711897850037,
+      "learning_rate": 5.021217026906292e-06,
+      "loss": 0.3583,
+      "step": 6546
+    },
+    {
+      "epoch": 1.646629778672032,
+      "grad_norm": 0.3105396330356598,
+      "learning_rate": 5.019753791566396e-06,
+      "loss": 0.34,
+      "step": 6547
+    },
+    {
+      "epoch": 1.6468812877263581,
+      "grad_norm": 0.3292068839073181,
+      "learning_rate": 5.0182905545347125e-06,
+      "loss": 0.3265,
+      "step": 6548
+    },
+    {
+      "epoch": 1.6471327967806841,
+      "grad_norm": 0.3053254783153534,
+      "learning_rate": 5.016827315936557e-06,
+      "loss": 0.3391,
+      "step": 6549
+    },
+    {
+      "epoch": 1.64738430583501,
+      "grad_norm": 0.3218199908733368,
+      "learning_rate": 5.015364075897246e-06,
+      "loss": 0.3596,
+      "step": 6550
+    },
+    {
+      "epoch": 1.647635814889336,
+      "grad_norm": 0.28854915499687195,
+      "learning_rate": 5.013900834542099e-06,
+      "loss": 0.3419,
+      "step": 6551
+    },
+    {
+      "epoch": 1.647887323943662,
+      "grad_norm": 0.33648034930229187,
+      "learning_rate": 5.012437591996432e-06,
+      "loss": 0.3496,
+      "step": 6552
+    },
+    {
+      "epoch": 1.6481388329979878,
+      "grad_norm": 0.2896369993686676,
+      "learning_rate": 5.010974348385565e-06,
+      "loss": 0.3423,
+      "step": 6553
+    },
+    {
+      "epoch": 1.6483903420523138,
+      "grad_norm": 0.31751370429992676,
+      "learning_rate": 5.009511103834811e-06,
+      "loss": 0.3562,
+      "step": 6554
+    },
+    {
+      "epoch": 1.6486418511066399,
+      "grad_norm": 0.2887558341026306,
+      "learning_rate": 5.008047858469492e-06,
+      "loss": 0.3316,
+      "step": 6555
+    },
+    {
+      "epoch": 1.6488933601609657,
+      "grad_norm": 0.3214130103588104,
+      "learning_rate": 5.006584612414924e-06,
+      "loss": 0.3598,
+      "step": 6556
+    },
+    {
+      "epoch": 1.649144869215292,
+      "grad_norm": 0.32030197978019714,
+      "learning_rate": 5.0051213657964245e-06,
+      "loss": 0.3675,
+      "step": 6557
+    },
+    {
+      "epoch": 1.6493963782696177,
+      "grad_norm": 0.3157503008842468,
+      "learning_rate": 5.003658118739313e-06,
+      "loss": 0.3495,
+      "step": 6558
+    },
+    {
+      "epoch": 1.6496478873239435,
+      "grad_norm": 0.31887757778167725,
+      "learning_rate": 5.0021948713689064e-06,
+      "loss": 0.3527,
+      "step": 6559
+    },
+    {
+      "epoch": 1.6498993963782698,
+      "grad_norm": 0.3290422558784485,
+      "learning_rate": 5.000731623810523e-06,
+      "loss": 0.3633,
+      "step": 6560
+    },
+    {
+      "epoch": 1.6501509054325956,
+      "grad_norm": 0.31094884872436523,
+      "learning_rate": 4.99926837618948e-06,
+      "loss": 0.3623,
+      "step": 6561
+    },
+    {
+      "epoch": 1.6504024144869214,
+      "grad_norm": 0.2901002764701843,
+      "learning_rate": 4.997805128631095e-06,
+      "loss": 0.355,
+      "step": 6562
+    },
+    {
+      "epoch": 1.6506539235412476,
+      "grad_norm": 0.28659021854400635,
+      "learning_rate": 4.996341881260689e-06,
+      "loss": 0.3392,
+      "step": 6563
+    },
+    {
+      "epoch": 1.6509054325955734,
+      "grad_norm": 0.324534147977829,
+      "learning_rate": 4.994878634203576e-06,
+      "loss": 0.3179,
+      "step": 6564
+    },
+    {
+      "epoch": 1.6511569416498992,
+      "grad_norm": 0.3127976357936859,
+      "learning_rate": 4.993415387585079e-06,
+      "loss": 0.3493,
+      "step": 6565
+    },
+    {
+      "epoch": 1.6514084507042255,
+      "grad_norm": 0.3226813077926636,
+      "learning_rate": 4.991952141530509e-06,
+      "loss": 0.3516,
+      "step": 6566
+    },
+    {
+      "epoch": 1.6516599597585513,
+      "grad_norm": 0.30921557545661926,
+      "learning_rate": 4.9904888961651895e-06,
+      "loss": 0.3334,
+      "step": 6567
+    },
+    {
+      "epoch": 1.6519114688128773,
+      "grad_norm": 0.31748196482658386,
+      "learning_rate": 4.989025651614438e-06,
+      "loss": 0.3315,
+      "step": 6568
+    },
+    {
+      "epoch": 1.6521629778672033,
+      "grad_norm": 0.3072637915611267,
+      "learning_rate": 4.987562408003568e-06,
+      "loss": 0.3572,
+      "step": 6569
+    },
+    {
+      "epoch": 1.6524144869215291,
+      "grad_norm": 0.3248019218444824,
+      "learning_rate": 4.9860991654579025e-06,
+      "loss": 0.3596,
+      "step": 6570
+    },
+    {
+      "epoch": 1.6526659959758552,
+      "grad_norm": 0.3171079754829407,
+      "learning_rate": 4.984635924102754e-06,
+      "loss": 0.3551,
+      "step": 6571
+    },
+    {
+      "epoch": 1.6529175050301812,
+      "grad_norm": 0.3320086598396301,
+      "learning_rate": 4.983172684063446e-06,
+      "loss": 0.3329,
+      "step": 6572
+    },
+    {
+      "epoch": 1.653169014084507,
+      "grad_norm": 0.3475695848464966,
+      "learning_rate": 4.981709445465288e-06,
+      "loss": 0.3538,
+      "step": 6573
+    },
+    {
+      "epoch": 1.653420523138833,
+      "grad_norm": 0.29529550671577454,
+      "learning_rate": 4.980246208433606e-06,
+      "loss": 0.315,
+      "step": 6574
+    },
+    {
+      "epoch": 1.653672032193159,
+      "grad_norm": 0.31380680203437805,
+      "learning_rate": 4.978782973093709e-06,
+      "loss": 0.3334,
+      "step": 6575
+    },
+    {
+      "epoch": 1.6539235412474849,
+      "grad_norm": 0.3308258354663849,
+      "learning_rate": 4.977319739570921e-06,
+      "loss": 0.3196,
+      "step": 6576
+    },
+    {
+      "epoch": 1.654175050301811,
+      "grad_norm": 0.3524826169013977,
+      "learning_rate": 4.975856507990552e-06,
+      "loss": 0.3344,
+      "step": 6577
+    },
+    {
+      "epoch": 1.654426559356137,
+      "grad_norm": 0.3204286992549896,
+      "learning_rate": 4.974393278477923e-06,
+      "loss": 0.3661,
+      "step": 6578
+    },
+    {
+      "epoch": 1.6546780684104627,
+      "grad_norm": 0.3220480978488922,
+      "learning_rate": 4.972930051158351e-06,
+      "loss": 0.3719,
+      "step": 6579
+    },
+    {
+      "epoch": 1.6549295774647887,
+      "grad_norm": 0.3321162760257721,
+      "learning_rate": 4.971466826157149e-06,
+      "loss": 0.3661,
+      "step": 6580
+    },
+    {
+      "epoch": 1.6551810865191148,
+      "grad_norm": 0.3444232642650604,
+      "learning_rate": 4.970003603599637e-06,
+      "loss": 0.3676,
+      "step": 6581
+    },
+    {
+      "epoch": 1.6554325955734406,
+      "grad_norm": 0.33695197105407715,
+      "learning_rate": 4.968540383611126e-06,
+      "loss": 0.3604,
+      "step": 6582
+    },
+    {
+      "epoch": 1.6556841046277666,
+      "grad_norm": 0.35238873958587646,
+      "learning_rate": 4.967077166316937e-06,
+      "loss": 0.3389,
+      "step": 6583
+    },
+    {
+      "epoch": 1.6559356136820926,
+      "grad_norm": 0.30956074595451355,
+      "learning_rate": 4.96561395184238e-06,
+      "loss": 0.3846,
+      "step": 6584
+    },
+    {
+      "epoch": 1.6561871227364184,
+      "grad_norm": 0.29510757327079773,
+      "learning_rate": 4.964150740312776e-06,
+      "loss": 0.3391,
+      "step": 6585
+    },
+    {
+      "epoch": 1.6564386317907445,
+      "grad_norm": 0.309314489364624,
+      "learning_rate": 4.962687531853434e-06,
+      "loss": 0.3519,
+      "step": 6586
+    },
+    {
+      "epoch": 1.6566901408450705,
+      "grad_norm": 0.3318527638912201,
+      "learning_rate": 4.961224326589674e-06,
+      "loss": 0.3607,
+      "step": 6587
+    },
+    {
+      "epoch": 1.6569416498993963,
+      "grad_norm": 0.3277136981487274,
+      "learning_rate": 4.959761124646805e-06,
+      "loss": 0.351,
+      "step": 6588
+    },
+    {
+      "epoch": 1.6571931589537223,
+      "grad_norm": 0.29636046290397644,
+      "learning_rate": 4.958297926150146e-06,
+      "loss": 0.3345,
+      "step": 6589
+    },
+    {
+      "epoch": 1.6574446680080483,
+      "grad_norm": 0.2861737310886383,
+      "learning_rate": 4.956834731225008e-06,
+      "loss": 0.3306,
+      "step": 6590
+    },
+    {
+      "epoch": 1.6576961770623742,
+      "grad_norm": 0.31142330169677734,
+      "learning_rate": 4.955371539996706e-06,
+      "loss": 0.3601,
+      "step": 6591
+    },
+    {
+      "epoch": 1.6579476861167002,
+      "grad_norm": 0.319341778755188,
+      "learning_rate": 4.953908352590552e-06,
+      "loss": 0.3525,
+      "step": 6592
+    },
+    {
+      "epoch": 1.6581991951710262,
+      "grad_norm": 0.295574426651001,
+      "learning_rate": 4.9524451691318585e-06,
+      "loss": 0.338,
+      "step": 6593
+    },
+    {
+      "epoch": 1.658450704225352,
+      "grad_norm": 0.3026335537433624,
+      "learning_rate": 4.95098198974594e-06,
+      "loss": 0.3506,
+      "step": 6594
+    },
+    {
+      "epoch": 1.658702213279678,
+      "grad_norm": 0.3181147873401642,
+      "learning_rate": 4.949518814558106e-06,
+      "loss": 0.3806,
+      "step": 6595
+    },
+    {
+      "epoch": 1.658953722334004,
+      "grad_norm": 0.30840447545051575,
+      "learning_rate": 4.948055643693671e-06,
+      "loss": 0.3635,
+      "step": 6596
+    },
+    {
+      "epoch": 1.6592052313883299,
+      "grad_norm": 0.3271403908729553,
+      "learning_rate": 4.946592477277945e-06,
+      "loss": 0.3833,
+      "step": 6597
+    },
+    {
+      "epoch": 1.659456740442656,
+      "grad_norm": 0.3057454228401184,
+      "learning_rate": 4.945129315436239e-06,
+      "loss": 0.3459,
+      "step": 6598
+    },
+    {
+      "epoch": 1.659708249496982,
+      "grad_norm": 0.30508214235305786,
+      "learning_rate": 4.943666158293864e-06,
+      "loss": 0.3659,
+      "step": 6599
+    },
+    {
+      "epoch": 1.6599597585513077,
+      "grad_norm": 0.3431748151779175,
+      "learning_rate": 4.942203005976128e-06,
+      "loss": 0.3873,
+      "step": 6600
+    },
+    {
+      "epoch": 1.6602112676056338,
+      "grad_norm": 0.3456651270389557,
+      "learning_rate": 4.940739858608346e-06,
+      "loss": 0.3472,
+      "step": 6601
+    },
+    {
+      "epoch": 1.6604627766599598,
+      "grad_norm": 0.3188634514808655,
+      "learning_rate": 4.939276716315822e-06,
+      "loss": 0.3422,
+      "step": 6602
+    },
+    {
+      "epoch": 1.6607142857142856,
+      "grad_norm": 0.31093767285346985,
+      "learning_rate": 4.937813579223871e-06,
+      "loss": 0.359,
+      "step": 6603
+    },
+    {
+      "epoch": 1.6609657947686118,
+      "grad_norm": 0.30429354310035706,
+      "learning_rate": 4.9363504474577936e-06,
+      "loss": 0.3307,
+      "step": 6604
+    },
+    {
+      "epoch": 1.6612173038229376,
+      "grad_norm": 0.31555190682411194,
+      "learning_rate": 4.934887321142905e-06,
+      "loss": 0.3468,
+      "step": 6605
+    },
+    {
+      "epoch": 1.6614688128772634,
+      "grad_norm": 0.30331486463546753,
+      "learning_rate": 4.933424200404508e-06,
+      "loss": 0.3412,
+      "step": 6606
+    },
+    {
+      "epoch": 1.6617203219315897,
+      "grad_norm": 0.3019322454929352,
+      "learning_rate": 4.9319610853679136e-06,
+      "loss": 0.3432,
+      "step": 6607
+    },
+    {
+      "epoch": 1.6619718309859155,
+      "grad_norm": 0.3192391097545624,
+      "learning_rate": 4.9304979761584256e-06,
+      "loss": 0.3571,
+      "step": 6608
+    },
+    {
+      "epoch": 1.6622233400402413,
+      "grad_norm": 0.31620514392852783,
+      "learning_rate": 4.929034872901352e-06,
+      "loss": 0.3415,
+      "step": 6609
+    },
+    {
+      "epoch": 1.6624748490945676,
+      "grad_norm": 0.32536083459854126,
+      "learning_rate": 4.927571775721996e-06,
+      "loss": 0.3433,
+      "step": 6610
+    },
+    {
+      "epoch": 1.6627263581488934,
+      "grad_norm": 0.2909981608390808,
+      "learning_rate": 4.926108684745664e-06,
+      "loss": 0.3614,
+      "step": 6611
+    },
+    {
+      "epoch": 1.6629778672032192,
+      "grad_norm": 0.30204445123672485,
+      "learning_rate": 4.924645600097663e-06,
+      "loss": 0.3639,
+      "step": 6612
+    },
+    {
+      "epoch": 1.6632293762575454,
+      "grad_norm": 0.3101191818714142,
+      "learning_rate": 4.923182521903293e-06,
+      "loss": 0.3451,
+      "step": 6613
+    },
+    {
+      "epoch": 1.6634808853118712,
+      "grad_norm": 0.3223671615123749,
+      "learning_rate": 4.9217194502878615e-06,
+      "loss": 0.353,
+      "step": 6614
+    },
+    {
+      "epoch": 1.663732394366197,
+      "grad_norm": 0.3006145656108856,
+      "learning_rate": 4.920256385376668e-06,
+      "loss": 0.3356,
+      "step": 6615
+    },
+    {
+      "epoch": 1.6639839034205233,
+      "grad_norm": 0.2919999957084656,
+      "learning_rate": 4.918793327295018e-06,
+      "loss": 0.3371,
+      "step": 6616
+    },
+    {
+      "epoch": 1.664235412474849,
+      "grad_norm": 0.3268287777900696,
+      "learning_rate": 4.917330276168208e-06,
+      "loss": 0.3501,
+      "step": 6617
+    },
+    {
+      "epoch": 1.664486921529175,
+      "grad_norm": 0.29942867159843445,
+      "learning_rate": 4.915867232121546e-06,
+      "loss": 0.3629,
+      "step": 6618
+    },
+    {
+      "epoch": 1.6647384305835011,
+      "grad_norm": 0.30160030722618103,
+      "learning_rate": 4.914404195280326e-06,
+      "loss": 0.3347,
+      "step": 6619
+    },
+    {
+      "epoch": 1.664989939637827,
+      "grad_norm": 0.30396658182144165,
+      "learning_rate": 4.912941165769855e-06,
+      "loss": 0.367,
+      "step": 6620
+    },
+    {
+      "epoch": 1.665241448692153,
+      "grad_norm": 0.34666627645492554,
+      "learning_rate": 4.9114781437154255e-06,
+      "loss": 0.3528,
+      "step": 6621
+    },
+    {
+      "epoch": 1.665492957746479,
+      "grad_norm": 0.324301153421402,
+      "learning_rate": 4.910015129242339e-06,
+      "loss": 0.3496,
+      "step": 6622
+    },
+    {
+      "epoch": 1.6657444668008048,
+      "grad_norm": 0.3196171224117279,
+      "learning_rate": 4.908552122475897e-06,
+      "loss": 0.338,
+      "step": 6623
+    },
+    {
+      "epoch": 1.6659959758551308,
+      "grad_norm": 0.2906225323677063,
+      "learning_rate": 4.90708912354139e-06,
+      "loss": 0.3447,
+      "step": 6624
+    },
+    {
+      "epoch": 1.6662474849094568,
+      "grad_norm": 0.3057789206504822,
+      "learning_rate": 4.905626132564121e-06,
+      "loss": 0.3315,
+      "step": 6625
+    },
+    {
+      "epoch": 1.6664989939637826,
+      "grad_norm": 0.34539157152175903,
+      "learning_rate": 4.904163149669382e-06,
+      "loss": 0.3487,
+      "step": 6626
+    },
+    {
+      "epoch": 1.6667505030181087,
+      "grad_norm": 0.3429698944091797,
+      "learning_rate": 4.902700174982471e-06,
+      "loss": 0.3373,
+      "step": 6627
+    },
+    {
+      "epoch": 1.6670020120724347,
+      "grad_norm": 0.29753679037094116,
+      "learning_rate": 4.901237208628679e-06,
+      "loss": 0.3371,
+      "step": 6628
+    },
+    {
+      "epoch": 1.6672535211267605,
+      "grad_norm": 0.3603936433792114,
+      "learning_rate": 4.899774250733305e-06,
+      "loss": 0.3302,
+      "step": 6629
+    },
+    {
+      "epoch": 1.6675050301810865,
+      "grad_norm": 0.315326452255249,
+      "learning_rate": 4.8983113014216365e-06,
+      "loss": 0.3192,
+      "step": 6630
+    },
+    {
+      "epoch": 1.6677565392354126,
+      "grad_norm": 0.28168460726737976,
+      "learning_rate": 4.896848360818971e-06,
+      "loss": 0.3242,
+      "step": 6631
+    },
+    {
+      "epoch": 1.6680080482897384,
+      "grad_norm": 0.3165770173072815,
+      "learning_rate": 4.895385429050597e-06,
+      "loss": 0.3665,
+      "step": 6632
+    },
+    {
+      "epoch": 1.6682595573440644,
+      "grad_norm": 0.3269800543785095,
+      "learning_rate": 4.893922506241806e-06,
+      "loss": 0.3463,
+      "step": 6633
+    },
+    {
+      "epoch": 1.6685110663983904,
+      "grad_norm": 0.3024964928627014,
+      "learning_rate": 4.8924595925178905e-06,
+      "loss": 0.3233,
+      "step": 6634
+    },
+    {
+      "epoch": 1.6687625754527162,
+      "grad_norm": 0.3181130290031433,
+      "learning_rate": 4.890996688004136e-06,
+      "loss": 0.3512,
+      "step": 6635
+    },
+    {
+      "epoch": 1.6690140845070423,
+      "grad_norm": 0.3129771053791046,
+      "learning_rate": 4.889533792825836e-06,
+      "loss": 0.3609,
+      "step": 6636
+    },
+    {
+      "epoch": 1.6692655935613683,
+      "grad_norm": 0.297387957572937,
+      "learning_rate": 4.888070907108273e-06,
+      "loss": 0.3696,
+      "step": 6637
+    },
+    {
+      "epoch": 1.669517102615694,
+      "grad_norm": 0.3127029836177826,
+      "learning_rate": 4.886608030976739e-06,
+      "loss": 0.3338,
+      "step": 6638
+    },
+    {
+      "epoch": 1.66976861167002,
+      "grad_norm": 0.326416939496994,
+      "learning_rate": 4.885145164556516e-06,
+      "loss": 0.3449,
+      "step": 6639
+    },
+    {
+      "epoch": 1.6700201207243461,
+      "grad_norm": 0.3048277795314789,
+      "learning_rate": 4.8836823079728925e-06,
+      "loss": 0.3643,
+      "step": 6640
+    },
+    {
+      "epoch": 1.670271629778672,
+      "grad_norm": 0.3344593644142151,
+      "learning_rate": 4.882219461351149e-06,
+      "loss": 0.3471,
+      "step": 6641
+    },
+    {
+      "epoch": 1.670523138832998,
+      "grad_norm": 0.3117375075817108,
+      "learning_rate": 4.880756624816574e-06,
+      "loss": 0.3416,
+      "step": 6642
+    },
+    {
+      "epoch": 1.670774647887324,
+      "grad_norm": 0.31037411093711853,
+      "learning_rate": 4.879293798494448e-06,
+      "loss": 0.3518,
+      "step": 6643
+    },
+    {
+      "epoch": 1.6710261569416498,
+      "grad_norm": 0.3362254500389099,
+      "learning_rate": 4.877830982510052e-06,
+      "loss": 0.3156,
+      "step": 6644
+    },
+    {
+      "epoch": 1.6712776659959758,
+      "grad_norm": 0.3232327699661255,
+      "learning_rate": 4.876368176988669e-06,
+      "loss": 0.3462,
+      "step": 6645
+    },
+    {
+      "epoch": 1.6715291750503019,
+      "grad_norm": 0.29828834533691406,
+      "learning_rate": 4.874905382055578e-06,
+      "loss": 0.3654,
+      "step": 6646
+    },
+    {
+      "epoch": 1.6717806841046277,
+      "grad_norm": 0.3095129430294037,
+      "learning_rate": 4.873442597836058e-06,
+      "loss": 0.3525,
+      "step": 6647
+    },
+    {
+      "epoch": 1.6720321931589537,
+      "grad_norm": 0.3279082775115967,
+      "learning_rate": 4.871979824455388e-06,
+      "loss": 0.3591,
+      "step": 6648
+    },
+    {
+      "epoch": 1.6722837022132797,
+      "grad_norm": 0.33682674169540405,
+      "learning_rate": 4.870517062038846e-06,
+      "loss": 0.3456,
+      "step": 6649
+    },
+    {
+      "epoch": 1.6725352112676055,
+      "grad_norm": 0.3893029987812042,
+      "learning_rate": 4.869054310711707e-06,
+      "loss": 0.3861,
+      "step": 6650
+    },
+    {
+      "epoch": 1.6727867203219315,
+      "grad_norm": 0.346723735332489,
+      "learning_rate": 4.867591570599247e-06,
+      "loss": 0.3562,
+      "step": 6651
+    },
+    {
+      "epoch": 1.6730382293762576,
+      "grad_norm": 0.3105736970901489,
+      "learning_rate": 4.8661288418267395e-06,
+      "loss": 0.3493,
+      "step": 6652
+    },
+    {
+      "epoch": 1.6732897384305834,
+      "grad_norm": 0.33905136585235596,
+      "learning_rate": 4.8646661245194605e-06,
+      "loss": 0.336,
+      "step": 6653
+    },
+    {
+      "epoch": 1.6735412474849096,
+      "grad_norm": 0.3350204527378082,
+      "learning_rate": 4.863203418802681e-06,
+      "loss": 0.3214,
+      "step": 6654
+    },
+    {
+      "epoch": 1.6737927565392354,
+      "grad_norm": 0.3009431064128876,
+      "learning_rate": 4.861740724801673e-06,
+      "loss": 0.336,
+      "step": 6655
+    },
+    {
+      "epoch": 1.6740442655935612,
+      "grad_norm": 0.3083348572254181,
+      "learning_rate": 4.860278042641707e-06,
+      "loss": 0.3652,
+      "step": 6656
+    },
+    {
+      "epoch": 1.6742957746478875,
+      "grad_norm": 0.30141690373420715,
+      "learning_rate": 4.858815372448053e-06,
+      "loss": 0.3533,
+      "step": 6657
+    },
+    {
+      "epoch": 1.6745472837022133,
+      "grad_norm": 0.2887999415397644,
+      "learning_rate": 4.857352714345978e-06,
+      "loss": 0.349,
+      "step": 6658
+    },
+    {
+      "epoch": 1.674798792756539,
+      "grad_norm": 0.28986456990242004,
+      "learning_rate": 4.8558900684607515e-06,
+      "loss": 0.3252,
+      "step": 6659
+    },
+    {
+      "epoch": 1.6750503018108653,
+      "grad_norm": 0.3002329468727112,
+      "learning_rate": 4.854427434917638e-06,
+      "loss": 0.3624,
+      "step": 6660
+    },
+    {
+      "epoch": 1.6753018108651911,
+      "grad_norm": 0.3201340138912201,
+      "learning_rate": 4.852964813841906e-06,
+      "loss": 0.3457,
+      "step": 6661
+    },
+    {
+      "epoch": 1.675553319919517,
+      "grad_norm": 0.30298128724098206,
+      "learning_rate": 4.851502205358816e-06,
+      "loss": 0.365,
+      "step": 6662
+    },
+    {
+      "epoch": 1.6758048289738432,
+      "grad_norm": 0.33912432193756104,
+      "learning_rate": 4.850039609593634e-06,
+      "loss": 0.3453,
+      "step": 6663
+    },
+    {
+      "epoch": 1.676056338028169,
+      "grad_norm": 0.3226577341556549,
+      "learning_rate": 4.84857702667162e-06,
+      "loss": 0.3462,
+      "step": 6664
+    },
+    {
+      "epoch": 1.6763078470824948,
+      "grad_norm": 0.2974978983402252,
+      "learning_rate": 4.847114456718039e-06,
+      "loss": 0.3518,
+      "step": 6665
+    },
+    {
+      "epoch": 1.676559356136821,
+      "grad_norm": 0.3077523708343506,
+      "learning_rate": 4.8456518998581445e-06,
+      "loss": 0.3547,
+      "step": 6666
+    },
+    {
+      "epoch": 1.6768108651911469,
+      "grad_norm": 0.32342010736465454,
+      "learning_rate": 4.844189356217203e-06,
+      "loss": 0.3691,
+      "step": 6667
+    },
+    {
+      "epoch": 1.6770623742454729,
+      "grad_norm": 0.33299824595451355,
+      "learning_rate": 4.842726825920466e-06,
+      "loss": 0.368,
+      "step": 6668
+    },
+    {
+      "epoch": 1.677313883299799,
+      "grad_norm": 0.2916051149368286,
+      "learning_rate": 4.8412643090931945e-06,
+      "loss": 0.3292,
+      "step": 6669
+    },
+    {
+      "epoch": 1.6775653923541247,
+      "grad_norm": 0.3211728632450104,
+      "learning_rate": 4.839801805860639e-06,
+      "loss": 0.3667,
+      "step": 6670
+    },
+    {
+      "epoch": 1.6778169014084507,
+      "grad_norm": 0.3185436725616455,
+      "learning_rate": 4.83833931634806e-06,
+      "loss": 0.3416,
+      "step": 6671
+    },
+    {
+      "epoch": 1.6780684104627768,
+      "grad_norm": 0.3072052001953125,
+      "learning_rate": 4.8368768406807045e-06,
+      "loss": 0.3526,
+      "step": 6672
+    },
+    {
+      "epoch": 1.6783199195171026,
+      "grad_norm": 0.29810407757759094,
+      "learning_rate": 4.8354143789838285e-06,
+      "loss": 0.339,
+      "step": 6673
+    },
+    {
+      "epoch": 1.6785714285714286,
+      "grad_norm": 0.3369450271129608,
+      "learning_rate": 4.83395193138268e-06,
+      "loss": 0.3568,
+      "step": 6674
+    },
+    {
+      "epoch": 1.6788229376257546,
+      "grad_norm": 0.33037352561950684,
+      "learning_rate": 4.83248949800251e-06,
+      "loss": 0.3522,
+      "step": 6675
+    },
+    {
+      "epoch": 1.6790744466800804,
+      "grad_norm": 0.36880165338516235,
+      "learning_rate": 4.831027078968568e-06,
+      "loss": 0.3664,
+      "step": 6676
+    },
+    {
+      "epoch": 1.6793259557344065,
+      "grad_norm": 0.3295780420303345,
+      "learning_rate": 4.829564674406098e-06,
+      "loss": 0.3236,
+      "step": 6677
+    },
+    {
+      "epoch": 1.6795774647887325,
+      "grad_norm": 0.327239066362381,
+      "learning_rate": 4.828102284440349e-06,
+      "loss": 0.3442,
+      "step": 6678
+    },
+    {
+      "epoch": 1.6798289738430583,
+      "grad_norm": 0.3397291898727417,
+      "learning_rate": 4.826639909196562e-06,
+      "loss": 0.3433,
+      "step": 6679
+    },
+    {
+      "epoch": 1.6800804828973843,
+      "grad_norm": 0.3069293797016144,
+      "learning_rate": 4.825177548799985e-06,
+      "loss": 0.3455,
+      "step": 6680
+    },
+    {
+      "epoch": 1.6803319919517103,
+      "grad_norm": 0.35793232917785645,
+      "learning_rate": 4.823715203375854e-06,
+      "loss": 0.3675,
+      "step": 6681
+    },
+    {
+      "epoch": 1.6805835010060362,
+      "grad_norm": 0.291510671377182,
+      "learning_rate": 4.822252873049416e-06,
+      "loss": 0.3461,
+      "step": 6682
+    },
+    {
+      "epoch": 1.6808350100603622,
+      "grad_norm": 0.32121041417121887,
+      "learning_rate": 4.8207905579459054e-06,
+      "loss": 0.3352,
+      "step": 6683
+    },
+    {
+      "epoch": 1.6810865191146882,
+      "grad_norm": 0.337119460105896,
+      "learning_rate": 4.819328258190564e-06,
+      "loss": 0.3171,
+      "step": 6684
+    },
+    {
+      "epoch": 1.681338028169014,
+      "grad_norm": 0.32898882031440735,
+      "learning_rate": 4.817865973908625e-06,
+      "loss": 0.35,
+      "step": 6685
+    },
+    {
+      "epoch": 1.68158953722334,
+      "grad_norm": 0.3288995623588562,
+      "learning_rate": 4.816403705225326e-06,
+      "loss": 0.3474,
+      "step": 6686
+    },
+    {
+      "epoch": 1.681841046277666,
+      "grad_norm": 0.3149915039539337,
+      "learning_rate": 4.814941452265903e-06,
+      "loss": 0.3283,
+      "step": 6687
+    },
+    {
+      "epoch": 1.6820925553319919,
+      "grad_norm": 0.3553139865398407,
+      "learning_rate": 4.813479215155585e-06,
+      "loss": 0.378,
+      "step": 6688
+    },
+    {
+      "epoch": 1.682344064386318,
+      "grad_norm": 0.3398416042327881,
+      "learning_rate": 4.812016994019607e-06,
+      "loss": 0.3614,
+      "step": 6689
+    },
+    {
+      "epoch": 1.682595573440644,
+      "grad_norm": 0.33960628509521484,
+      "learning_rate": 4.810554788983196e-06,
+      "loss": 0.348,
+      "step": 6690
+    },
+    {
+      "epoch": 1.6828470824949697,
+      "grad_norm": 0.3453803062438965,
+      "learning_rate": 4.809092600171584e-06,
+      "loss": 0.3615,
+      "step": 6691
+    },
+    {
+      "epoch": 1.6830985915492958,
+      "grad_norm": 0.3408578634262085,
+      "learning_rate": 4.807630427709995e-06,
+      "loss": 0.3595,
+      "step": 6692
+    },
+    {
+      "epoch": 1.6833501006036218,
+      "grad_norm": 0.3130299746990204,
+      "learning_rate": 4.806168271723657e-06,
+      "loss": 0.3441,
+      "step": 6693
+    },
+    {
+      "epoch": 1.6836016096579476,
+      "grad_norm": 0.3115828037261963,
+      "learning_rate": 4.804706132337793e-06,
+      "loss": 0.3505,
+      "step": 6694
+    },
+    {
+      "epoch": 1.6838531187122736,
+      "grad_norm": 0.34020087122917175,
+      "learning_rate": 4.803244009677629e-06,
+      "loss": 0.3569,
+      "step": 6695
+    },
+    {
+      "epoch": 1.6841046277665996,
+      "grad_norm": 0.3176514804363251,
+      "learning_rate": 4.801781903868383e-06,
+      "loss": 0.3452,
+      "step": 6696
+    },
+    {
+      "epoch": 1.6843561368209254,
+      "grad_norm": 0.33271244168281555,
+      "learning_rate": 4.8003198150352755e-06,
+      "loss": 0.3536,
+      "step": 6697
+    },
+    {
+      "epoch": 1.6846076458752515,
+      "grad_norm": 0.3202335834503174,
+      "learning_rate": 4.79885774330353e-06,
+      "loss": 0.3692,
+      "step": 6698
+    },
+    {
+      "epoch": 1.6848591549295775,
+      "grad_norm": 0.314801424741745,
+      "learning_rate": 4.797395688798358e-06,
+      "loss": 0.3468,
+      "step": 6699
+    },
+    {
+      "epoch": 1.6851106639839033,
+      "grad_norm": 0.35798442363739014,
+      "learning_rate": 4.7959336516449795e-06,
+      "loss": 0.3449,
+      "step": 6700
+    },
+    {
+      "epoch": 1.6853621730382293,
+      "grad_norm": 0.35215941071510315,
+      "learning_rate": 4.794471631968606e-06,
+      "loss": 0.3699,
+      "step": 6701
+    },
+    {
+      "epoch": 1.6856136820925554,
+      "grad_norm": 0.30209723114967346,
+      "learning_rate": 4.793009629894451e-06,
+      "loss": 0.3386,
+      "step": 6702
+    },
+    {
+      "epoch": 1.6858651911468812,
+      "grad_norm": 0.3090112507343292,
+      "learning_rate": 4.791547645547727e-06,
+      "loss": 0.3756,
+      "step": 6703
+    },
+    {
+      "epoch": 1.6861167002012074,
+      "grad_norm": 0.32377490401268005,
+      "learning_rate": 4.790085679053644e-06,
+      "loss": 0.3814,
+      "step": 6704
+    },
+    {
+      "epoch": 1.6863682092555332,
+      "grad_norm": 0.30719804763793945,
+      "learning_rate": 4.788623730537407e-06,
+      "loss": 0.3386,
+      "step": 6705
+    },
+    {
+      "epoch": 1.686619718309859,
+      "grad_norm": 0.3107820451259613,
+      "learning_rate": 4.787161800124228e-06,
+      "loss": 0.3603,
+      "step": 6706
+    },
+    {
+      "epoch": 1.6868712273641853,
+      "grad_norm": 0.30711448192596436,
+      "learning_rate": 4.785699887939307e-06,
+      "loss": 0.344,
+      "step": 6707
+    },
+    {
+      "epoch": 1.687122736418511,
+      "grad_norm": 0.30240827798843384,
+      "learning_rate": 4.78423799410785e-06,
+      "loss": 0.3494,
+      "step": 6708
+    },
+    {
+      "epoch": 1.6873742454728369,
+      "grad_norm": 0.30051419138908386,
+      "learning_rate": 4.782776118755061e-06,
+      "loss": 0.3477,
+      "step": 6709
+    },
+    {
+      "epoch": 1.6876257545271631,
+      "grad_norm": 0.3172437846660614,
+      "learning_rate": 4.7813142620061365e-06,
+      "loss": 0.3885,
+      "step": 6710
+    },
+    {
+      "epoch": 1.687877263581489,
+      "grad_norm": 0.30793458223342896,
+      "learning_rate": 4.779852423986278e-06,
+      "loss": 0.3312,
+      "step": 6711
+    },
+    {
+      "epoch": 1.6881287726358147,
+      "grad_norm": 0.31784626841545105,
+      "learning_rate": 4.778390604820683e-06,
+      "loss": 0.3646,
+      "step": 6712
+    },
+    {
+      "epoch": 1.688380281690141,
+      "grad_norm": 0.3140884041786194,
+      "learning_rate": 4.776928804634545e-06,
+      "loss": 0.3276,
+      "step": 6713
+    },
+    {
+      "epoch": 1.6886317907444668,
+      "grad_norm": 0.3013300895690918,
+      "learning_rate": 4.775467023553061e-06,
+      "loss": 0.3607,
+      "step": 6714
+    },
+    {
+      "epoch": 1.6888832997987926,
+      "grad_norm": 0.30073946714401245,
+      "learning_rate": 4.77400526170142e-06,
+      "loss": 0.348,
+      "step": 6715
+    },
+    {
+      "epoch": 1.6891348088531188,
+      "grad_norm": 0.3390004336833954,
+      "learning_rate": 4.7725435192048156e-06,
+      "loss": 0.324,
+      "step": 6716
+    },
+    {
+      "epoch": 1.6893863179074446,
+      "grad_norm": 0.31539881229400635,
+      "learning_rate": 4.771081796188435e-06,
+      "loss": 0.3331,
+      "step": 6717
+    },
+    {
+      "epoch": 1.6896378269617707,
+      "grad_norm": 0.34204819798469543,
+      "learning_rate": 4.7696200927774675e-06,
+      "loss": 0.3243,
+      "step": 6718
+    },
+    {
+      "epoch": 1.6898893360160967,
+      "grad_norm": 0.312477171421051,
+      "learning_rate": 4.768158409097096e-06,
+      "loss": 0.3378,
+      "step": 6719
+    },
+    {
+      "epoch": 1.6901408450704225,
+      "grad_norm": 0.32249870896339417,
+      "learning_rate": 4.766696745272508e-06,
+      "loss": 0.3498,
+      "step": 6720
+    },
+    {
+      "epoch": 1.6903923541247485,
+      "grad_norm": 0.31593847274780273,
+      "learning_rate": 4.765235101428883e-06,
+      "loss": 0.3244,
+      "step": 6721
+    },
+    {
+      "epoch": 1.6906438631790746,
+      "grad_norm": 0.31690171360969543,
+      "learning_rate": 4.7637734776914045e-06,
+      "loss": 0.3346,
+      "step": 6722
+    },
+    {
+      "epoch": 1.6908953722334004,
+      "grad_norm": 0.3169795274734497,
+      "learning_rate": 4.7623118741852484e-06,
+      "loss": 0.3551,
+      "step": 6723
+    },
+    {
+      "epoch": 1.6911468812877264,
+      "grad_norm": 0.2886812686920166,
+      "learning_rate": 4.760850291035595e-06,
+      "loss": 0.3425,
+      "step": 6724
+    },
+    {
+      "epoch": 1.6913983903420524,
+      "grad_norm": 0.3049076199531555,
+      "learning_rate": 4.759388728367615e-06,
+      "loss": 0.3521,
+      "step": 6725
+    },
+    {
+      "epoch": 1.6916498993963782,
+      "grad_norm": 0.2920475900173187,
+      "learning_rate": 4.757927186306489e-06,
+      "loss": 0.3332,
+      "step": 6726
+    },
+    {
+      "epoch": 1.6919014084507042,
+      "grad_norm": 0.3055409789085388,
+      "learning_rate": 4.756465664977381e-06,
+      "loss": 0.3586,
+      "step": 6727
+    },
+    {
+      "epoch": 1.6921529175050303,
+      "grad_norm": 0.3127298057079315,
+      "learning_rate": 4.75500416450547e-06,
+      "loss": 0.3309,
+      "step": 6728
+    },
+    {
+      "epoch": 1.692404426559356,
+      "grad_norm": 0.3161258101463318,
+      "learning_rate": 4.753542685015916e-06,
+      "loss": 0.3436,
+      "step": 6729
+    },
+    {
+      "epoch": 1.692655935613682,
+      "grad_norm": 0.3124731481075287,
+      "learning_rate": 4.752081226633888e-06,
+      "loss": 0.3495,
+      "step": 6730
+    },
+    {
+      "epoch": 1.6929074446680081,
+      "grad_norm": 0.30646365880966187,
+      "learning_rate": 4.750619789484556e-06,
+      "loss": 0.328,
+      "step": 6731
+    },
+    {
+      "epoch": 1.693158953722334,
+      "grad_norm": 0.30611762404441833,
+      "learning_rate": 4.749158373693076e-06,
+      "loss": 0.363,
+      "step": 6732
+    },
+    {
+      "epoch": 1.69341046277666,
+      "grad_norm": 0.3001714050769806,
+      "learning_rate": 4.7476969793846136e-06,
+      "loss": 0.3489,
+      "step": 6733
+    },
+    {
+      "epoch": 1.693661971830986,
+      "grad_norm": 0.30453163385391235,
+      "learning_rate": 4.746235606684326e-06,
+      "loss": 0.3424,
+      "step": 6734
+    },
+    {
+      "epoch": 1.6939134808853118,
+      "grad_norm": 0.31667834520339966,
+      "learning_rate": 4.744774255717372e-06,
+      "loss": 0.3638,
+      "step": 6735
+    },
+    {
+      "epoch": 1.6941649899396378,
+      "grad_norm": 0.30841580033302307,
+      "learning_rate": 4.7433129266089045e-06,
+      "loss": 0.3691,
+      "step": 6736
+    },
+    {
+      "epoch": 1.6944164989939638,
+      "grad_norm": 0.3112299144268036,
+      "learning_rate": 4.74185161948408e-06,
+      "loss": 0.3445,
+      "step": 6737
+    },
+    {
+      "epoch": 1.6946680080482897,
+      "grad_norm": 0.32629552483558655,
+      "learning_rate": 4.7403903344680495e-06,
+      "loss": 0.344,
+      "step": 6738
+    },
+    {
+      "epoch": 1.6949195171026157,
+      "grad_norm": 0.3291604220867157,
+      "learning_rate": 4.7389290716859634e-06,
+      "loss": 0.3614,
+      "step": 6739
+    },
+    {
+      "epoch": 1.6951710261569417,
+      "grad_norm": 0.3310836851596832,
+      "learning_rate": 4.737467831262967e-06,
+      "loss": 0.3459,
+      "step": 6740
+    },
+    {
+      "epoch": 1.6954225352112675,
+      "grad_norm": 0.3385966718196869,
+      "learning_rate": 4.736006613324209e-06,
+      "loss": 0.3583,
+      "step": 6741
+    },
+    {
+      "epoch": 1.6956740442655935,
+      "grad_norm": 0.29955390095710754,
+      "learning_rate": 4.734545417994834e-06,
+      "loss": 0.3815,
+      "step": 6742
+    },
+    {
+      "epoch": 1.6959255533199196,
+      "grad_norm": 0.30086326599121094,
+      "learning_rate": 4.7330842453999825e-06,
+      "loss": 0.3576,
+      "step": 6743
+    },
+    {
+      "epoch": 1.6961770623742454,
+      "grad_norm": 0.29754674434661865,
+      "learning_rate": 4.731623095664797e-06,
+      "loss": 0.3435,
+      "step": 6744
+    },
+    {
+      "epoch": 1.6964285714285714,
+      "grad_norm": 0.34076476097106934,
+      "learning_rate": 4.730161968914412e-06,
+      "loss": 0.3372,
+      "step": 6745
+    },
+    {
+      "epoch": 1.6966800804828974,
+      "grad_norm": 0.3299402594566345,
+      "learning_rate": 4.72870086527397e-06,
+      "loss": 0.34,
+      "step": 6746
+    },
+    {
+      "epoch": 1.6969315895372232,
+      "grad_norm": 0.3097115755081177,
+      "learning_rate": 4.727239784868597e-06,
+      "loss": 0.3463,
+      "step": 6747
+    },
+    {
+      "epoch": 1.6971830985915493,
+      "grad_norm": 0.30106401443481445,
+      "learning_rate": 4.725778727823434e-06,
+      "loss": 0.3376,
+      "step": 6748
+    },
+    {
+      "epoch": 1.6974346076458753,
+      "grad_norm": 0.3247431814670563,
+      "learning_rate": 4.724317694263605e-06,
+      "loss": 0.3387,
+      "step": 6749
+    },
+    {
+      "epoch": 1.697686116700201,
+      "grad_norm": 0.30322572588920593,
+      "learning_rate": 4.7228566843142426e-06,
+      "loss": 0.3589,
+      "step": 6750
+    },
+    {
+      "epoch": 1.6979376257545271,
+      "grad_norm": 0.3284371793270111,
+      "learning_rate": 4.72139569810047e-06,
+      "loss": 0.3592,
+      "step": 6751
+    },
+    {
+      "epoch": 1.6981891348088531,
+      "grad_norm": 0.3332383632659912,
+      "learning_rate": 4.7199347357474115e-06,
+      "loss": 0.357,
+      "step": 6752
+    },
+    {
+      "epoch": 1.698440643863179,
+      "grad_norm": 0.32514792680740356,
+      "learning_rate": 4.7184737973801945e-06,
+      "loss": 0.3224,
+      "step": 6753
+    },
+    {
+      "epoch": 1.6986921529175052,
+      "grad_norm": 0.2966492176055908,
+      "learning_rate": 4.717012883123932e-06,
+      "loss": 0.3646,
+      "step": 6754
+    },
+    {
+      "epoch": 1.698943661971831,
+      "grad_norm": 0.32919812202453613,
+      "learning_rate": 4.715551993103749e-06,
+      "loss": 0.3455,
+      "step": 6755
+    },
+    {
+      "epoch": 1.6991951710261568,
+      "grad_norm": 0.29699820280075073,
+      "learning_rate": 4.714091127444755e-06,
+      "loss": 0.3478,
+      "step": 6756
+    },
+    {
+      "epoch": 1.699446680080483,
+      "grad_norm": 0.33212509751319885,
+      "learning_rate": 4.712630286272071e-06,
+      "loss": 0.351,
+      "step": 6757
+    },
+    {
+      "epoch": 1.6996981891348089,
+      "grad_norm": 0.323261022567749,
+      "learning_rate": 4.711169469710802e-06,
+      "loss": 0.353,
+      "step": 6758
+    },
+    {
+      "epoch": 1.6999496981891347,
+      "grad_norm": 0.3002278804779053,
+      "learning_rate": 4.7097086778860625e-06,
+      "loss": 0.3433,
+      "step": 6759
+    },
+    {
+      "epoch": 1.700201207243461,
+      "grad_norm": 0.31812816858291626,
+      "learning_rate": 4.708247910922958e-06,
+      "loss": 0.3633,
+      "step": 6760
+    },
+    {
+      "epoch": 1.7004527162977867,
+      "grad_norm": 0.3050650656223297,
+      "learning_rate": 4.706787168946596e-06,
+      "loss": 0.3376,
+      "step": 6761
+    },
+    {
+      "epoch": 1.7007042253521125,
+      "grad_norm": 0.3200231194496155,
+      "learning_rate": 4.705326452082076e-06,
+      "loss": 0.3421,
+      "step": 6762
+    },
+    {
+      "epoch": 1.7009557344064388,
+      "grad_norm": 0.32024839520454407,
+      "learning_rate": 4.703865760454503e-06,
+      "loss": 0.3384,
+      "step": 6763
+    },
+    {
+      "epoch": 1.7012072434607646,
+      "grad_norm": 0.3096054494380951,
+      "learning_rate": 4.702405094188977e-06,
+      "loss": 0.3466,
+      "step": 6764
+    },
+    {
+      "epoch": 1.7014587525150904,
+      "grad_norm": 0.30760541558265686,
+      "learning_rate": 4.70094445341059e-06,
+      "loss": 0.3337,
+      "step": 6765
+    },
+    {
+      "epoch": 1.7017102615694166,
+      "grad_norm": 0.2908000349998474,
+      "learning_rate": 4.699483838244443e-06,
+      "loss": 0.3563,
+      "step": 6766
+    },
+    {
+      "epoch": 1.7019617706237424,
+      "grad_norm": 0.3102479875087738,
+      "learning_rate": 4.698023248815623e-06,
+      "loss": 0.34,
+      "step": 6767
+    },
+    {
+      "epoch": 1.7022132796780685,
+      "grad_norm": 0.317518413066864,
+      "learning_rate": 4.6965626852492235e-06,
+      "loss": 0.3706,
+      "step": 6768
+    },
+    {
+      "epoch": 1.7024647887323945,
+      "grad_norm": 0.33676648139953613,
+      "learning_rate": 4.6951021476703304e-06,
+      "loss": 0.348,
+      "step": 6769
+    },
+    {
+      "epoch": 1.7027162977867203,
+      "grad_norm": 0.30047839879989624,
+      "learning_rate": 4.6936416362040325e-06,
+      "loss": 0.3411,
+      "step": 6770
+    },
+    {
+      "epoch": 1.7029678068410463,
+      "grad_norm": 0.325817734003067,
+      "learning_rate": 4.69218115097541e-06,
+      "loss": 0.3657,
+      "step": 6771
+    },
+    {
+      "epoch": 1.7032193158953723,
+      "grad_norm": 0.28770050406455994,
+      "learning_rate": 4.690720692109549e-06,
+      "loss": 0.3375,
+      "step": 6772
+    },
+    {
+      "epoch": 1.7034708249496981,
+      "grad_norm": 0.30655109882354736,
+      "learning_rate": 4.689260259731523e-06,
+      "loss": 0.3397,
+      "step": 6773
+    },
+    {
+      "epoch": 1.7037223340040242,
+      "grad_norm": 0.3254563510417938,
+      "learning_rate": 4.687799853966413e-06,
+      "loss": 0.3299,
+      "step": 6774
+    },
+    {
+      "epoch": 1.7039738430583502,
+      "grad_norm": 0.3062305450439453,
+      "learning_rate": 4.686339474939293e-06,
+      "loss": 0.3617,
+      "step": 6775
+    },
+    {
+      "epoch": 1.704225352112676,
+      "grad_norm": 0.3447815179824829,
+      "learning_rate": 4.6848791227752335e-06,
+      "loss": 0.3491,
+      "step": 6776
+    },
+    {
+      "epoch": 1.704476861167002,
+      "grad_norm": 0.3192574083805084,
+      "learning_rate": 4.6834187975993065e-06,
+      "loss": 0.3619,
+      "step": 6777
+    },
+    {
+      "epoch": 1.704728370221328,
+      "grad_norm": 0.3120723068714142,
+      "learning_rate": 4.681958499536579e-06,
+      "loss": 0.3453,
+      "step": 6778
+    },
+    {
+      "epoch": 1.7049798792756539,
+      "grad_norm": 0.31706029176712036,
+      "learning_rate": 4.680498228712116e-06,
+      "loss": 0.3423,
+      "step": 6779
+    },
+    {
+      "epoch": 1.70523138832998,
+      "grad_norm": 0.3090144693851471,
+      "learning_rate": 4.679037985250981e-06,
+      "loss": 0.3368,
+      "step": 6780
+    },
+    {
+      "epoch": 1.705482897384306,
+      "grad_norm": 0.3182971179485321,
+      "learning_rate": 4.677577769278235e-06,
+      "loss": 0.3429,
+      "step": 6781
+    },
+    {
+      "epoch": 1.7057344064386317,
+      "grad_norm": 0.2918050289154053,
+      "learning_rate": 4.6761175809189366e-06,
+      "loss": 0.3477,
+      "step": 6782
+    },
+    {
+      "epoch": 1.7059859154929577,
+      "grad_norm": 0.3405851721763611,
+      "learning_rate": 4.67465742029814e-06,
+      "loss": 0.3507,
+      "step": 6783
+    },
+    {
+      "epoch": 1.7062374245472838,
+      "grad_norm": 0.2979619801044464,
+      "learning_rate": 4.6731972875409e-06,
+      "loss": 0.3183,
+      "step": 6784
+    },
+    {
+      "epoch": 1.7064889336016096,
+      "grad_norm": 0.3200087249279022,
+      "learning_rate": 4.671737182772267e-06,
+      "loss": 0.3588,
+      "step": 6785
+    },
+    {
+      "epoch": 1.7067404426559356,
+      "grad_norm": 0.3113258183002472,
+      "learning_rate": 4.6702771061172935e-06,
+      "loss": 0.3237,
+      "step": 6786
+    },
+    {
+      "epoch": 1.7069919517102616,
+      "grad_norm": 0.3095763325691223,
+      "learning_rate": 4.66881705770102e-06,
+      "loss": 0.3481,
+      "step": 6787
+    },
+    {
+      "epoch": 1.7072434607645874,
+      "grad_norm": 0.32275089621543884,
+      "learning_rate": 4.667357037648496e-06,
+      "loss": 0.3434,
+      "step": 6788
+    },
+    {
+      "epoch": 1.7074949698189135,
+      "grad_norm": 0.3030339777469635,
+      "learning_rate": 4.665897046084759e-06,
+      "loss": 0.3369,
+      "step": 6789
+    },
+    {
+      "epoch": 1.7077464788732395,
+      "grad_norm": 0.32319900393486023,
+      "learning_rate": 4.6644370831348524e-06,
+      "loss": 0.3355,
+      "step": 6790
+    },
+    {
+      "epoch": 1.7079979879275653,
+      "grad_norm": 0.2972315847873688,
+      "learning_rate": 4.662977148923808e-06,
+      "loss": 0.337,
+      "step": 6791
+    },
+    {
+      "epoch": 1.7082494969818913,
+      "grad_norm": 0.3062729239463806,
+      "learning_rate": 4.6615172435766636e-06,
+      "loss": 0.3504,
+      "step": 6792
+    },
+    {
+      "epoch": 1.7085010060362174,
+      "grad_norm": 0.3187355697154999,
+      "learning_rate": 4.660057367218448e-06,
+      "loss": 0.3579,
+      "step": 6793
+    },
+    {
+      "epoch": 1.7087525150905432,
+      "grad_norm": 0.33169421553611755,
+      "learning_rate": 4.658597519974193e-06,
+      "loss": 0.3504,
+      "step": 6794
+    },
+    {
+      "epoch": 1.7090040241448692,
+      "grad_norm": 0.3317626118659973,
+      "learning_rate": 4.657137701968925e-06,
+      "loss": 0.3571,
+      "step": 6795
+    },
+    {
+      "epoch": 1.7092555331991952,
+      "grad_norm": 0.31924623250961304,
+      "learning_rate": 4.655677913327668e-06,
+      "loss": 0.3457,
+      "step": 6796
+    },
+    {
+      "epoch": 1.709507042253521,
+      "grad_norm": 0.3141118288040161,
+      "learning_rate": 4.654218154175444e-06,
+      "loss": 0.3178,
+      "step": 6797
+    },
+    {
+      "epoch": 1.709758551307847,
+      "grad_norm": 0.3203374445438385,
+      "learning_rate": 4.652758424637271e-06,
+      "loss": 0.3472,
+      "step": 6798
+    },
+    {
+      "epoch": 1.710010060362173,
+      "grad_norm": 0.33453568816185,
+      "learning_rate": 4.651298724838168e-06,
+      "loss": 0.3485,
+      "step": 6799
+    },
+    {
+      "epoch": 1.7102615694164989,
+      "grad_norm": 0.3020557761192322,
+      "learning_rate": 4.649839054903146e-06,
+      "loss": 0.3516,
+      "step": 6800
+    },
+    {
+      "epoch": 1.710513078470825,
+      "grad_norm": 0.32627707719802856,
+      "learning_rate": 4.6483794149572196e-06,
+      "loss": 0.3663,
+      "step": 6801
+    },
+    {
+      "epoch": 1.710764587525151,
+      "grad_norm": 0.29445070028305054,
+      "learning_rate": 4.646919805125396e-06,
+      "loss": 0.3552,
+      "step": 6802
+    },
+    {
+      "epoch": 1.7110160965794767,
+      "grad_norm": 0.30278632044792175,
+      "learning_rate": 4.645460225532683e-06,
+      "loss": 0.3602,
+      "step": 6803
+    },
+    {
+      "epoch": 1.711267605633803,
+      "grad_norm": 0.3449458181858063,
+      "learning_rate": 4.644000676304082e-06,
+      "loss": 0.3329,
+      "step": 6804
+    },
+    {
+      "epoch": 1.7115191146881288,
+      "grad_norm": 0.32875609397888184,
+      "learning_rate": 4.642541157564596e-06,
+      "loss": 0.3313,
+      "step": 6805
+    },
+    {
+      "epoch": 1.7117706237424546,
+      "grad_norm": 0.3220992386341095,
+      "learning_rate": 4.641081669439226e-06,
+      "loss": 0.3739,
+      "step": 6806
+    },
+    {
+      "epoch": 1.7120221327967808,
+      "grad_norm": 0.30020537972450256,
+      "learning_rate": 4.6396222120529625e-06,
+      "loss": 0.33,
+      "step": 6807
+    },
+    {
+      "epoch": 1.7122736418511066,
+      "grad_norm": 0.3377690315246582,
+      "learning_rate": 4.638162785530805e-06,
+      "loss": 0.3478,
+      "step": 6808
+    },
+    {
+      "epoch": 1.7125251509054324,
+      "grad_norm": 0.30975767970085144,
+      "learning_rate": 4.636703389997739e-06,
+      "loss": 0.3479,
+      "step": 6809
+    },
+    {
+      "epoch": 1.7127766599597587,
+      "grad_norm": 0.31768518686294556,
+      "learning_rate": 4.635244025578757e-06,
+      "loss": 0.3301,
+      "step": 6810
+    },
+    {
+      "epoch": 1.7130281690140845,
+      "grad_norm": 0.28654173016548157,
+      "learning_rate": 4.63378469239884e-06,
+      "loss": 0.3402,
+      "step": 6811
+    },
+    {
+      "epoch": 1.7132796780684103,
+      "grad_norm": 0.33102425932884216,
+      "learning_rate": 4.632325390582976e-06,
+      "loss": 0.3487,
+      "step": 6812
+    },
+    {
+      "epoch": 1.7135311871227366,
+      "grad_norm": 0.30774351954460144,
+      "learning_rate": 4.630866120256139e-06,
+      "loss": 0.3409,
+      "step": 6813
+    },
+    {
+      "epoch": 1.7137826961770624,
+      "grad_norm": 0.3273797035217285,
+      "learning_rate": 4.629406881543312e-06,
+      "loss": 0.3329,
+      "step": 6814
+    },
+    {
+      "epoch": 1.7140342052313882,
+      "grad_norm": 0.31635209918022156,
+      "learning_rate": 4.6279476745694655e-06,
+      "loss": 0.3387,
+      "step": 6815
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.3162465989589691,
+      "learning_rate": 4.6264884994595725e-06,
+      "loss": 0.3405,
+      "step": 6816
+    },
+    {
+      "epoch": 1.7145372233400402,
+      "grad_norm": 0.35132935643196106,
+      "learning_rate": 4.625029356338605e-06,
+      "loss": 0.3522,
+      "step": 6817
+    },
+    {
+      "epoch": 1.7147887323943662,
+      "grad_norm": 0.31276100873947144,
+      "learning_rate": 4.623570245331525e-06,
+      "loss": 0.3402,
+      "step": 6818
+    },
+    {
+      "epoch": 1.7150402414486923,
+      "grad_norm": 0.2836771309375763,
+      "learning_rate": 4.622111166563301e-06,
+      "loss": 0.3389,
+      "step": 6819
+    },
+    {
+      "epoch": 1.715291750503018,
+      "grad_norm": 0.33196336030960083,
+      "learning_rate": 4.6206521201588894e-06,
+      "loss": 0.3398,
+      "step": 6820
+    },
+    {
+      "epoch": 1.715543259557344,
+      "grad_norm": 0.28593510389328003,
+      "learning_rate": 4.6191931062432526e-06,
+      "loss": 0.366,
+      "step": 6821
+    },
+    {
+      "epoch": 1.7157947686116701,
+      "grad_norm": 0.3059460520744324,
+      "learning_rate": 4.617734124941342e-06,
+      "loss": 0.3467,
+      "step": 6822
+    },
+    {
+      "epoch": 1.716046277665996,
+      "grad_norm": 0.3157053589820862,
+      "learning_rate": 4.6162751763781146e-06,
+      "loss": 0.309,
+      "step": 6823
+    },
+    {
+      "epoch": 1.716297786720322,
+      "grad_norm": 0.29486018419265747,
+      "learning_rate": 4.6148162606785144e-06,
+      "loss": 0.3461,
+      "step": 6824
+    },
+    {
+      "epoch": 1.716549295774648,
+      "grad_norm": 0.32476910948753357,
+      "learning_rate": 4.613357377967495e-06,
+      "loss": 0.3405,
+      "step": 6825
+    },
+    {
+      "epoch": 1.7168008048289738,
+      "grad_norm": 0.3237256705760956,
+      "learning_rate": 4.611898528369995e-06,
+      "loss": 0.3262,
+      "step": 6826
+    },
+    {
+      "epoch": 1.7170523138832998,
+      "grad_norm": 0.3211307227611542,
+      "learning_rate": 4.610439712010959e-06,
+      "loss": 0.3476,
+      "step": 6827
+    },
+    {
+      "epoch": 1.7173038229376258,
+      "grad_norm": 0.29768112301826477,
+      "learning_rate": 4.6089809290153245e-06,
+      "loss": 0.3629,
+      "step": 6828
+    },
+    {
+      "epoch": 1.7175553319919517,
+      "grad_norm": 0.2961174547672272,
+      "learning_rate": 4.607522179508027e-06,
+      "loss": 0.343,
+      "step": 6829
+    },
+    {
+      "epoch": 1.7178068410462777,
+      "grad_norm": 0.2991441488265991,
+      "learning_rate": 4.606063463614e-06,
+      "loss": 0.3571,
+      "step": 6830
+    },
+    {
+      "epoch": 1.7180583501006037,
+      "grad_norm": 0.2887977063655853,
+      "learning_rate": 4.604604781458173e-06,
+      "loss": 0.3516,
+      "step": 6831
+    },
+    {
+      "epoch": 1.7183098591549295,
+      "grad_norm": 0.31125563383102417,
+      "learning_rate": 4.6031461331654725e-06,
+      "loss": 0.3294,
+      "step": 6832
+    },
+    {
+      "epoch": 1.7185613682092555,
+      "grad_norm": 0.3178209960460663,
+      "learning_rate": 4.601687518860823e-06,
+      "loss": 0.3519,
+      "step": 6833
+    },
+    {
+      "epoch": 1.7188128772635816,
+      "grad_norm": 0.29017218947410583,
+      "learning_rate": 4.600228938669146e-06,
+      "loss": 0.3451,
+      "step": 6834
+    },
+    {
+      "epoch": 1.7190643863179074,
+      "grad_norm": 0.3055189549922943,
+      "learning_rate": 4.5987703927153575e-06,
+      "loss": 0.3346,
+      "step": 6835
+    },
+    {
+      "epoch": 1.7193158953722334,
+      "grad_norm": 0.3083680272102356,
+      "learning_rate": 4.597311881124378e-06,
+      "loss": 0.3386,
+      "step": 6836
+    },
+    {
+      "epoch": 1.7195674044265594,
+      "grad_norm": 0.32472705841064453,
+      "learning_rate": 4.595853404021114e-06,
+      "loss": 0.3384,
+      "step": 6837
+    },
+    {
+      "epoch": 1.7198189134808852,
+      "grad_norm": 0.3365086317062378,
+      "learning_rate": 4.594394961530479e-06,
+      "loss": 0.3496,
+      "step": 6838
+    },
+    {
+      "epoch": 1.7200704225352113,
+      "grad_norm": 0.33195602893829346,
+      "learning_rate": 4.592936553777378e-06,
+      "loss": 0.3715,
+      "step": 6839
+    },
+    {
+      "epoch": 1.7203219315895373,
+      "grad_norm": 0.3010450005531311,
+      "learning_rate": 4.591478180886714e-06,
+      "loss": 0.3476,
+      "step": 6840
+    },
+    {
+      "epoch": 1.720573440643863,
+      "grad_norm": 0.3121878206729889,
+      "learning_rate": 4.590019842983389e-06,
+      "loss": 0.3477,
+      "step": 6841
+    },
+    {
+      "epoch": 1.720824949698189,
+      "grad_norm": 0.3093717396259308,
+      "learning_rate": 4.588561540192299e-06,
+      "loss": 0.3335,
+      "step": 6842
+    },
+    {
+      "epoch": 1.7210764587525151,
+      "grad_norm": 0.3174624741077423,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.355,
+      "step": 6843
+    },
+    {
+      "epoch": 1.721327967806841,
+      "grad_norm": 0.3113724887371063,
+      "learning_rate": 4.585645040446401e-06,
+      "loss": 0.3357,
+      "step": 6844
+    },
+    {
+      "epoch": 1.721579476861167,
+      "grad_norm": 0.33507058024406433,
+      "learning_rate": 4.584186843741373e-06,
+      "loss": 0.3498,
+      "step": 6845
+    },
+    {
+      "epoch": 1.721830985915493,
+      "grad_norm": 0.34897923469543457,
+      "learning_rate": 4.58272868264814e-06,
+      "loss": 0.3539,
+      "step": 6846
+    },
+    {
+      "epoch": 1.7220824949698188,
+      "grad_norm": 0.3107760548591614,
+      "learning_rate": 4.581270557291586e-06,
+      "loss": 0.3618,
+      "step": 6847
+    },
+    {
+      "epoch": 1.7223340040241448,
+      "grad_norm": 0.3432888984680176,
+      "learning_rate": 4.579812467796588e-06,
+      "loss": 0.3521,
+      "step": 6848
+    },
+    {
+      "epoch": 1.7225855130784709,
+      "grad_norm": 0.35445329546928406,
+      "learning_rate": 4.5783544142880235e-06,
+      "loss": 0.3892,
+      "step": 6849
+    },
+    {
+      "epoch": 1.7228370221327967,
+      "grad_norm": 0.35715755820274353,
+      "learning_rate": 4.576896396890767e-06,
+      "loss": 0.3621,
+      "step": 6850
+    },
+    {
+      "epoch": 1.7230885311871227,
+      "grad_norm": 0.30892953276634216,
+      "learning_rate": 4.575438415729685e-06,
+      "loss": 0.3596,
+      "step": 6851
+    },
+    {
+      "epoch": 1.7233400402414487,
+      "grad_norm": 0.339541494846344,
+      "learning_rate": 4.573980470929649e-06,
+      "loss": 0.3262,
+      "step": 6852
+    },
+    {
+      "epoch": 1.7235915492957745,
+      "grad_norm": 0.33129116892814636,
+      "learning_rate": 4.572522562615519e-06,
+      "loss": 0.3431,
+      "step": 6853
+    },
+    {
+      "epoch": 1.7238430583501008,
+      "grad_norm": 0.3051982522010803,
+      "learning_rate": 4.5710646909121585e-06,
+      "loss": 0.3451,
+      "step": 6854
+    },
+    {
+      "epoch": 1.7240945674044266,
+      "grad_norm": 0.3270474076271057,
+      "learning_rate": 4.5696068559444225e-06,
+      "loss": 0.3626,
+      "step": 6855
+    },
+    {
+      "epoch": 1.7243460764587524,
+      "grad_norm": 0.3316960334777832,
+      "learning_rate": 4.568149057837168e-06,
+      "loss": 0.3443,
+      "step": 6856
+    },
+    {
+      "epoch": 1.7245975855130786,
+      "grad_norm": 0.32199814915657043,
+      "learning_rate": 4.5666912967152435e-06,
+      "loss": 0.3544,
+      "step": 6857
+    },
+    {
+      "epoch": 1.7248490945674044,
+      "grad_norm": 0.32749658823013306,
+      "learning_rate": 4.5652335727035e-06,
+      "loss": 0.3256,
+      "step": 6858
+    },
+    {
+      "epoch": 1.7251006036217302,
+      "grad_norm": 0.3357948362827301,
+      "learning_rate": 4.56377588592678e-06,
+      "loss": 0.3441,
+      "step": 6859
+    },
+    {
+      "epoch": 1.7253521126760565,
+      "grad_norm": 0.3046301603317261,
+      "learning_rate": 4.562318236509926e-06,
+      "loss": 0.3464,
+      "step": 6860
+    },
+    {
+      "epoch": 1.7256036217303823,
+      "grad_norm": 0.33477479219436646,
+      "learning_rate": 4.560860624577779e-06,
+      "loss": 0.3365,
+      "step": 6861
+    },
+    {
+      "epoch": 1.725855130784708,
+      "grad_norm": 0.3378863036632538,
+      "learning_rate": 4.559403050255169e-06,
+      "loss": 0.3526,
+      "step": 6862
+    },
+    {
+      "epoch": 1.7261066398390343,
+      "grad_norm": 0.3373473584651947,
+      "learning_rate": 4.557945513666935e-06,
+      "loss": 0.3376,
+      "step": 6863
+    },
+    {
+      "epoch": 1.7263581488933601,
+      "grad_norm": 0.33986005187034607,
+      "learning_rate": 4.5564880149378995e-06,
+      "loss": 0.3471,
+      "step": 6864
+    },
+    {
+      "epoch": 1.7266096579476862,
+      "grad_norm": 0.2948022782802582,
+      "learning_rate": 4.555030554192894e-06,
+      "loss": 0.3384,
+      "step": 6865
+    },
+    {
+      "epoch": 1.7268611670020122,
+      "grad_norm": 0.3061770796775818,
+      "learning_rate": 4.553573131556734e-06,
+      "loss": 0.3311,
+      "step": 6866
+    },
+    {
+      "epoch": 1.727112676056338,
+      "grad_norm": 0.32700714468955994,
+      "learning_rate": 4.552115747154247e-06,
+      "loss": 0.3456,
+      "step": 6867
+    },
+    {
+      "epoch": 1.727364185110664,
+      "grad_norm": 0.3140343725681305,
+      "learning_rate": 4.55065840111024e-06,
+      "loss": 0.3903,
+      "step": 6868
+    },
+    {
+      "epoch": 1.72761569416499,
+      "grad_norm": 0.32476603984832764,
+      "learning_rate": 4.549201093549533e-06,
+      "loss": 0.3226,
+      "step": 6869
+    },
+    {
+      "epoch": 1.7278672032193159,
+      "grad_norm": 0.32490959763526917,
+      "learning_rate": 4.547743824596929e-06,
+      "loss": 0.3408,
+      "step": 6870
+    },
+    {
+      "epoch": 1.7281187122736419,
+      "grad_norm": 0.3557870388031006,
+      "learning_rate": 4.546286594377238e-06,
+      "loss": 0.3438,
+      "step": 6871
+    },
+    {
+      "epoch": 1.728370221327968,
+      "grad_norm": 0.3313372731208801,
+      "learning_rate": 4.544829403015264e-06,
+      "loss": 0.3448,
+      "step": 6872
+    },
+    {
+      "epoch": 1.7286217303822937,
+      "grad_norm": 0.3483891189098358,
+      "learning_rate": 4.543372250635801e-06,
+      "loss": 0.3346,
+      "step": 6873
+    },
+    {
+      "epoch": 1.7288732394366197,
+      "grad_norm": 0.3317503035068512,
+      "learning_rate": 4.541915137363651e-06,
+      "loss": 0.3537,
+      "step": 6874
+    },
+    {
+      "epoch": 1.7291247484909458,
+      "grad_norm": 0.33802077174186707,
+      "learning_rate": 4.540458063323601e-06,
+      "loss": 0.3488,
+      "step": 6875
+    },
+    {
+      "epoch": 1.7293762575452716,
+      "grad_norm": 0.30749836564064026,
+      "learning_rate": 4.539001028640447e-06,
+      "loss": 0.35,
+      "step": 6876
+    },
+    {
+      "epoch": 1.7296277665995976,
+      "grad_norm": 0.3114013671875,
+      "learning_rate": 4.537544033438967e-06,
+      "loss": 0.3305,
+      "step": 6877
+    },
+    {
+      "epoch": 1.7298792756539236,
+      "grad_norm": 0.3545249104499817,
+      "learning_rate": 4.53608707784395e-06,
+      "loss": 0.347,
+      "step": 6878
+    },
+    {
+      "epoch": 1.7301307847082494,
+      "grad_norm": 0.3405088484287262,
+      "learning_rate": 4.534630161980171e-06,
+      "loss": 0.3498,
+      "step": 6879
+    },
+    {
+      "epoch": 1.7303822937625755,
+      "grad_norm": 0.3269940912723541,
+      "learning_rate": 4.533173285972408e-06,
+      "loss": 0.3648,
+      "step": 6880
+    },
+    {
+      "epoch": 1.7306338028169015,
+      "grad_norm": 0.30962666869163513,
+      "learning_rate": 4.531716449945431e-06,
+      "loss": 0.3474,
+      "step": 6881
+    },
+    {
+      "epoch": 1.7308853118712273,
+      "grad_norm": 0.3638971149921417,
+      "learning_rate": 4.530259654024011e-06,
+      "loss": 0.3424,
+      "step": 6882
+    },
+    {
+      "epoch": 1.7311368209255533,
+      "grad_norm": 0.38066476583480835,
+      "learning_rate": 4.528802898332914e-06,
+      "loss": 0.3332,
+      "step": 6883
+    },
+    {
+      "epoch": 1.7313883299798793,
+      "grad_norm": 0.30477452278137207,
+      "learning_rate": 4.5273461829969e-06,
+      "loss": 0.3625,
+      "step": 6884
+    },
+    {
+      "epoch": 1.7316398390342052,
+      "grad_norm": 0.3417944014072418,
+      "learning_rate": 4.525889508140731e-06,
+      "loss": 0.3661,
+      "step": 6885
+    },
+    {
+      "epoch": 1.7318913480885312,
+      "grad_norm": 0.340406209230423,
+      "learning_rate": 4.524432873889156e-06,
+      "loss": 0.3576,
+      "step": 6886
+    },
+    {
+      "epoch": 1.7321428571428572,
+      "grad_norm": 0.40321967005729675,
+      "learning_rate": 4.522976280366934e-06,
+      "loss": 0.346,
+      "step": 6887
+    },
+    {
+      "epoch": 1.732394366197183,
+      "grad_norm": 0.3876849412918091,
+      "learning_rate": 4.5215197276988055e-06,
+      "loss": 0.3598,
+      "step": 6888
+    },
+    {
+      "epoch": 1.732645875251509,
+      "grad_norm": 0.3189951479434967,
+      "learning_rate": 4.520063216009522e-06,
+      "loss": 0.3289,
+      "step": 6889
+    },
+    {
+      "epoch": 1.732897384305835,
+      "grad_norm": 0.3129585087299347,
+      "learning_rate": 4.518606745423819e-06,
+      "loss": 0.325,
+      "step": 6890
+    },
+    {
+      "epoch": 1.7331488933601609,
+      "grad_norm": 0.2990803122520447,
+      "learning_rate": 4.517150316066439e-06,
+      "loss": 0.3333,
+      "step": 6891
+    },
+    {
+      "epoch": 1.733400402414487,
+      "grad_norm": 0.31391507387161255,
+      "learning_rate": 4.515693928062112e-06,
+      "loss": 0.3527,
+      "step": 6892
+    },
+    {
+      "epoch": 1.733651911468813,
+      "grad_norm": 0.3259546756744385,
+      "learning_rate": 4.514237581535571e-06,
+      "loss": 0.3729,
+      "step": 6893
+    },
+    {
+      "epoch": 1.7339034205231387,
+      "grad_norm": 0.3225707411766052,
+      "learning_rate": 4.512781276611542e-06,
+      "loss": 0.3587,
+      "step": 6894
+    },
+    {
+      "epoch": 1.7341549295774648,
+      "grad_norm": 0.311168909072876,
+      "learning_rate": 4.511325013414749e-06,
+      "loss": 0.3533,
+      "step": 6895
+    },
+    {
+      "epoch": 1.7344064386317908,
+      "grad_norm": 0.30714693665504456,
+      "learning_rate": 4.509868792069912e-06,
+      "loss": 0.3373,
+      "step": 6896
+    },
+    {
+      "epoch": 1.7346579476861166,
+      "grad_norm": 0.31528759002685547,
+      "learning_rate": 4.508412612701746e-06,
+      "loss": 0.3632,
+      "step": 6897
+    },
+    {
+      "epoch": 1.7349094567404426,
+      "grad_norm": 0.29422181844711304,
+      "learning_rate": 4.506956475434964e-06,
+      "loss": 0.3536,
+      "step": 6898
+    },
+    {
+      "epoch": 1.7351609657947686,
+      "grad_norm": 0.2980017364025116,
+      "learning_rate": 4.505500380394276e-06,
+      "loss": 0.3205,
+      "step": 6899
+    },
+    {
+      "epoch": 1.7354124748490944,
+      "grad_norm": 0.32730206847190857,
+      "learning_rate": 4.504044327704387e-06,
+      "loss": 0.3645,
+      "step": 6900
+    },
+    {
+      "epoch": 1.7356639839034205,
+      "grad_norm": 0.3445815145969391,
+      "learning_rate": 4.502588317489997e-06,
+      "loss": 0.382,
+      "step": 6901
+    },
+    {
+      "epoch": 1.7359154929577465,
+      "grad_norm": 0.3304295241832733,
+      "learning_rate": 4.501132349875808e-06,
+      "loss": 0.3683,
+      "step": 6902
+    },
+    {
+      "epoch": 1.7361670020120723,
+      "grad_norm": 0.30265411734580994,
+      "learning_rate": 4.499676424986512e-06,
+      "loss": 0.3542,
+      "step": 6903
+    },
+    {
+      "epoch": 1.7364185110663986,
+      "grad_norm": 0.3029594421386719,
+      "learning_rate": 4.498220542946798e-06,
+      "loss": 0.3597,
+      "step": 6904
+    },
+    {
+      "epoch": 1.7366700201207244,
+      "grad_norm": 0.327571302652359,
+      "learning_rate": 4.4967647038813575e-06,
+      "loss": 0.353,
+      "step": 6905
+    },
+    {
+      "epoch": 1.7369215291750502,
+      "grad_norm": 0.2830412685871124,
+      "learning_rate": 4.495308907914871e-06,
+      "loss": 0.3255,
+      "step": 6906
+    },
+    {
+      "epoch": 1.7371730382293764,
+      "grad_norm": 0.32870203256607056,
+      "learning_rate": 4.49385315517202e-06,
+      "loss": 0.3445,
+      "step": 6907
+    },
+    {
+      "epoch": 1.7374245472837022,
+      "grad_norm": 0.34851449728012085,
+      "learning_rate": 4.492397445777479e-06,
+      "loss": 0.3283,
+      "step": 6908
+    },
+    {
+      "epoch": 1.737676056338028,
+      "grad_norm": 0.3187304735183716,
+      "learning_rate": 4.490941779855922e-06,
+      "loss": 0.3495,
+      "step": 6909
+    },
+    {
+      "epoch": 1.7379275653923543,
+      "grad_norm": 0.31370389461517334,
+      "learning_rate": 4.489486157532016e-06,
+      "loss": 0.3392,
+      "step": 6910
+    },
+    {
+      "epoch": 1.73817907444668,
+      "grad_norm": 0.3236584961414337,
+      "learning_rate": 4.488030578930428e-06,
+      "loss": 0.3764,
+      "step": 6911
+    },
+    {
+      "epoch": 1.7384305835010059,
+      "grad_norm": 0.3015926778316498,
+      "learning_rate": 4.486575044175817e-06,
+      "loss": 0.348,
+      "step": 6912
+    },
+    {
+      "epoch": 1.7386820925553321,
+      "grad_norm": 0.33164599537849426,
+      "learning_rate": 4.485119553392843e-06,
+      "loss": 0.3407,
+      "step": 6913
+    },
+    {
+      "epoch": 1.738933601609658,
+      "grad_norm": 0.3237990736961365,
+      "learning_rate": 4.483664106706155e-06,
+      "loss": 0.3614,
+      "step": 6914
+    },
+    {
+      "epoch": 1.739185110663984,
+      "grad_norm": 0.30540236830711365,
+      "learning_rate": 4.482208704240408e-06,
+      "loss": 0.3377,
+      "step": 6915
+    },
+    {
+      "epoch": 1.73943661971831,
+      "grad_norm": 0.32697343826293945,
+      "learning_rate": 4.480753346120247e-06,
+      "loss": 0.336,
+      "step": 6916
+    },
+    {
+      "epoch": 1.7396881287726358,
+      "grad_norm": 0.32882174849510193,
+      "learning_rate": 4.479298032470312e-06,
+      "loss": 0.3385,
+      "step": 6917
+    },
+    {
+      "epoch": 1.7399396378269618,
+      "grad_norm": 0.315578430891037,
+      "learning_rate": 4.477842763415244e-06,
+      "loss": 0.3481,
+      "step": 6918
+    },
+    {
+      "epoch": 1.7401911468812878,
+      "grad_norm": 0.2991701066493988,
+      "learning_rate": 4.476387539079676e-06,
+      "loss": 0.321,
+      "step": 6919
+    },
+    {
+      "epoch": 1.7404426559356136,
+      "grad_norm": 0.3147648274898529,
+      "learning_rate": 4.474932359588241e-06,
+      "loss": 0.3375,
+      "step": 6920
+    },
+    {
+      "epoch": 1.7406941649899397,
+      "grad_norm": 0.34548136591911316,
+      "learning_rate": 4.473477225065563e-06,
+      "loss": 0.3568,
+      "step": 6921
+    },
+    {
+      "epoch": 1.7409456740442657,
+      "grad_norm": 0.3195883631706238,
+      "learning_rate": 4.472022135636268e-06,
+      "loss": 0.3509,
+      "step": 6922
+    },
+    {
+      "epoch": 1.7411971830985915,
+      "grad_norm": 0.2943090498447418,
+      "learning_rate": 4.470567091424973e-06,
+      "loss": 0.3495,
+      "step": 6923
+    },
+    {
+      "epoch": 1.7414486921529175,
+      "grad_norm": 0.32778680324554443,
+      "learning_rate": 4.469112092556296e-06,
+      "loss": 0.3478,
+      "step": 6924
+    },
+    {
+      "epoch": 1.7417002012072436,
+      "grad_norm": 0.2987135052680969,
+      "learning_rate": 4.467657139154845e-06,
+      "loss": 0.3446,
+      "step": 6925
+    },
+    {
+      "epoch": 1.7419517102615694,
+      "grad_norm": 0.3135947287082672,
+      "learning_rate": 4.466202231345229e-06,
+      "loss": 0.3638,
+      "step": 6926
+    },
+    {
+      "epoch": 1.7422032193158954,
+      "grad_norm": 0.29023781418800354,
+      "learning_rate": 4.464747369252056e-06,
+      "loss": 0.3696,
+      "step": 6927
+    },
+    {
+      "epoch": 1.7424547283702214,
+      "grad_norm": 0.31927579641342163,
+      "learning_rate": 4.463292552999919e-06,
+      "loss": 0.3555,
+      "step": 6928
+    },
+    {
+      "epoch": 1.7427062374245472,
+      "grad_norm": 0.36215946078300476,
+      "learning_rate": 4.4618377827134205e-06,
+      "loss": 0.3435,
+      "step": 6929
+    },
+    {
+      "epoch": 1.7429577464788732,
+      "grad_norm": 0.30876463651657104,
+      "learning_rate": 4.460383058517146e-06,
+      "loss": 0.3416,
+      "step": 6930
+    },
+    {
+      "epoch": 1.7432092555331993,
+      "grad_norm": 0.2847077250480652,
+      "learning_rate": 4.458928380535689e-06,
+      "loss": 0.3375,
+      "step": 6931
+    },
+    {
+      "epoch": 1.743460764587525,
+      "grad_norm": 0.30569058656692505,
+      "learning_rate": 4.45747374889363e-06,
+      "loss": 0.3246,
+      "step": 6932
+    },
+    {
+      "epoch": 1.743712273641851,
+      "grad_norm": 0.2917396128177643,
+      "learning_rate": 4.456019163715552e-06,
+      "loss": 0.3466,
+      "step": 6933
+    },
+    {
+      "epoch": 1.7439637826961771,
+      "grad_norm": 0.29794490337371826,
+      "learning_rate": 4.454564625126026e-06,
+      "loss": 0.3593,
+      "step": 6934
+    },
+    {
+      "epoch": 1.744215291750503,
+      "grad_norm": 0.31993839144706726,
+      "learning_rate": 4.45311013324963e-06,
+      "loss": 0.3431,
+      "step": 6935
+    },
+    {
+      "epoch": 1.744466800804829,
+      "grad_norm": 0.2840554714202881,
+      "learning_rate": 4.45165568821093e-06,
+      "loss": 0.3485,
+      "step": 6936
+    },
+    {
+      "epoch": 1.744718309859155,
+      "grad_norm": 0.32249221205711365,
+      "learning_rate": 4.450201290134489e-06,
+      "loss": 0.3301,
+      "step": 6937
+    },
+    {
+      "epoch": 1.7449698189134808,
+      "grad_norm": 0.290263831615448,
+      "learning_rate": 4.448746939144869e-06,
+      "loss": 0.3413,
+      "step": 6938
+    },
+    {
+      "epoch": 1.7452213279678068,
+      "grad_norm": 0.30761563777923584,
+      "learning_rate": 4.447292635366623e-06,
+      "loss": 0.3572,
+      "step": 6939
+    },
+    {
+      "epoch": 1.7454728370221329,
+      "grad_norm": 0.3331648111343384,
+      "learning_rate": 4.4458383789243086e-06,
+      "loss": 0.3372,
+      "step": 6940
+    },
+    {
+      "epoch": 1.7457243460764587,
+      "grad_norm": 0.3228667080402374,
+      "learning_rate": 4.444384169942466e-06,
+      "loss": 0.3245,
+      "step": 6941
+    },
+    {
+      "epoch": 1.7459758551307847,
+      "grad_norm": 0.2900792360305786,
+      "learning_rate": 4.4429300085456475e-06,
+      "loss": 0.3437,
+      "step": 6942
+    },
+    {
+      "epoch": 1.7462273641851107,
+      "grad_norm": 0.31676149368286133,
+      "learning_rate": 4.4414758948583855e-06,
+      "loss": 0.3397,
+      "step": 6943
+    },
+    {
+      "epoch": 1.7464788732394365,
+      "grad_norm": 0.32281917333602905,
+      "learning_rate": 4.440021829005221e-06,
+      "loss": 0.3333,
+      "step": 6944
+    },
+    {
+      "epoch": 1.7467303822937625,
+      "grad_norm": 0.3074769377708435,
+      "learning_rate": 4.438567811110682e-06,
+      "loss": 0.3588,
+      "step": 6945
+    },
+    {
+      "epoch": 1.7469818913480886,
+      "grad_norm": 0.31130295991897583,
+      "learning_rate": 4.437113841299297e-06,
+      "loss": 0.3617,
+      "step": 6946
+    },
+    {
+      "epoch": 1.7472334004024144,
+      "grad_norm": 0.3043542802333832,
+      "learning_rate": 4.435659919695593e-06,
+      "loss": 0.3479,
+      "step": 6947
+    },
+    {
+      "epoch": 1.7474849094567404,
+      "grad_norm": 0.37658563256263733,
+      "learning_rate": 4.434206046424085e-06,
+      "loss": 0.3292,
+      "step": 6948
+    },
+    {
+      "epoch": 1.7477364185110664,
+      "grad_norm": 0.3038911521434784,
+      "learning_rate": 4.43275222160929e-06,
+      "loss": 0.3645,
+      "step": 6949
+    },
+    {
+      "epoch": 1.7479879275653922,
+      "grad_norm": 0.3258115351200104,
+      "learning_rate": 4.431298445375717e-06,
+      "loss": 0.3271,
+      "step": 6950
+    },
+    {
+      "epoch": 1.7482394366197183,
+      "grad_norm": 0.3135131001472473,
+      "learning_rate": 4.429844717847876e-06,
+      "loss": 0.349,
+      "step": 6951
+    },
+    {
+      "epoch": 1.7484909456740443,
+      "grad_norm": 0.32563337683677673,
+      "learning_rate": 4.428391039150266e-06,
+      "loss": 0.3503,
+      "step": 6952
+    },
+    {
+      "epoch": 1.74874245472837,
+      "grad_norm": 0.3328086733818054,
+      "learning_rate": 4.426937409407391e-06,
+      "loss": 0.3368,
+      "step": 6953
+    },
+    {
+      "epoch": 1.7489939637826963,
+      "grad_norm": 0.3279817998409271,
+      "learning_rate": 4.4254838287437386e-06,
+      "loss": 0.3378,
+      "step": 6954
+    },
+    {
+      "epoch": 1.7492454728370221,
+      "grad_norm": 0.30334845185279846,
+      "learning_rate": 4.424030297283805e-06,
+      "loss": 0.3261,
+      "step": 6955
+    },
+    {
+      "epoch": 1.749496981891348,
+      "grad_norm": 0.32851114869117737,
+      "learning_rate": 4.42257681515207e-06,
+      "loss": 0.3525,
+      "step": 6956
+    },
+    {
+      "epoch": 1.7497484909456742,
+      "grad_norm": 0.32159623503685,
+      "learning_rate": 4.42112338247302e-06,
+      "loss": 0.3465,
+      "step": 6957
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3375226855278015,
+      "learning_rate": 4.41966999937113e-06,
+      "loss": 0.3525,
+      "step": 6958
+    },
+    {
+      "epoch": 1.7502515090543258,
+      "grad_norm": 0.34559088945388794,
+      "learning_rate": 4.418216665970875e-06,
+      "loss": 0.3692,
+      "step": 6959
+    },
+    {
+      "epoch": 1.750503018108652,
+      "grad_norm": 0.2969364523887634,
+      "learning_rate": 4.416763382396723e-06,
+      "loss": 0.3436,
+      "step": 6960
+    },
+    {
+      "epoch": 1.7507545271629779,
+      "grad_norm": 0.3050820529460907,
+      "learning_rate": 4.4153101487731385e-06,
+      "loss": 0.3235,
+      "step": 6961
+    },
+    {
+      "epoch": 1.7510060362173037,
+      "grad_norm": 0.32389354705810547,
+      "learning_rate": 4.413856965224581e-06,
+      "loss": 0.3538,
+      "step": 6962
+    },
+    {
+      "epoch": 1.75125754527163,
+      "grad_norm": 0.36416080594062805,
+      "learning_rate": 4.412403831875509e-06,
+      "loss": 0.3653,
+      "step": 6963
+    },
+    {
+      "epoch": 1.7515090543259557,
+      "grad_norm": 0.31853196024894714,
+      "learning_rate": 4.410950748850372e-06,
+      "loss": 0.3283,
+      "step": 6964
+    },
+    {
+      "epoch": 1.7517605633802817,
+      "grad_norm": 0.29971179366111755,
+      "learning_rate": 4.409497716273618e-06,
+      "loss": 0.3397,
+      "step": 6965
+    },
+    {
+      "epoch": 1.7520120724346078,
+      "grad_norm": 0.3168172836303711,
+      "learning_rate": 4.408044734269692e-06,
+      "loss": 0.3361,
+      "step": 6966
+    },
+    {
+      "epoch": 1.7522635814889336,
+      "grad_norm": 0.3141171336174011,
+      "learning_rate": 4.40659180296303e-06,
+      "loss": 0.3341,
+      "step": 6967
+    },
+    {
+      "epoch": 1.7525150905432596,
+      "grad_norm": 0.3332131803035736,
+      "learning_rate": 4.405138922478066e-06,
+      "loss": 0.327,
+      "step": 6968
+    },
+    {
+      "epoch": 1.7527665995975856,
+      "grad_norm": 0.3187667727470398,
+      "learning_rate": 4.403686092939235e-06,
+      "loss": 0.3512,
+      "step": 6969
+    },
+    {
+      "epoch": 1.7530181086519114,
+      "grad_norm": 0.3000306487083435,
+      "learning_rate": 4.4022333144709566e-06,
+      "loss": 0.363,
+      "step": 6970
+    },
+    {
+      "epoch": 1.7532696177062375,
+      "grad_norm": 0.32261931896209717,
+      "learning_rate": 4.400780587197658e-06,
+      "loss": 0.3832,
+      "step": 6971
+    },
+    {
+      "epoch": 1.7535211267605635,
+      "grad_norm": 0.31589117646217346,
+      "learning_rate": 4.399327911243751e-06,
+      "loss": 0.3646,
+      "step": 6972
+    },
+    {
+      "epoch": 1.7537726358148893,
+      "grad_norm": 0.3187357187271118,
+      "learning_rate": 4.3978752867336536e-06,
+      "loss": 0.3653,
+      "step": 6973
+    },
+    {
+      "epoch": 1.7540241448692153,
+      "grad_norm": 0.3000258505344391,
+      "learning_rate": 4.396422713791768e-06,
+      "loss": 0.3428,
+      "step": 6974
+    },
+    {
+      "epoch": 1.7542756539235413,
+      "grad_norm": 0.30427271127700806,
+      "learning_rate": 4.394970192542504e-06,
+      "loss": 0.3423,
+      "step": 6975
+    },
+    {
+      "epoch": 1.7545271629778671,
+      "grad_norm": 0.3039720952510834,
+      "learning_rate": 4.3935177231102544e-06,
+      "loss": 0.3351,
+      "step": 6976
+    },
+    {
+      "epoch": 1.7547786720321932,
+      "grad_norm": 0.29543939232826233,
+      "learning_rate": 4.3920653056194205e-06,
+      "loss": 0.3216,
+      "step": 6977
+    },
+    {
+      "epoch": 1.7550301810865192,
+      "grad_norm": 0.3125627934932709,
+      "learning_rate": 4.390612940194388e-06,
+      "loss": 0.3434,
+      "step": 6978
+    },
+    {
+      "epoch": 1.755281690140845,
+      "grad_norm": 0.2896956205368042,
+      "learning_rate": 4.389160626959545e-06,
+      "loss": 0.3435,
+      "step": 6979
+    },
+    {
+      "epoch": 1.755533199195171,
+      "grad_norm": 0.3043994903564453,
+      "learning_rate": 4.387708366039275e-06,
+      "loss": 0.3524,
+      "step": 6980
+    },
+    {
+      "epoch": 1.755784708249497,
+      "grad_norm": 0.3003641664981842,
+      "learning_rate": 4.38625615755795e-06,
+      "loss": 0.321,
+      "step": 6981
+    },
+    {
+      "epoch": 1.7560362173038229,
+      "grad_norm": 0.33770516514778137,
+      "learning_rate": 4.384804001639948e-06,
+      "loss": 0.3518,
+      "step": 6982
+    },
+    {
+      "epoch": 1.756287726358149,
+      "grad_norm": 0.3379949927330017,
+      "learning_rate": 4.383351898409634e-06,
+      "loss": 0.34,
+      "step": 6983
+    },
+    {
+      "epoch": 1.756539235412475,
+      "grad_norm": 0.33452707529067993,
+      "learning_rate": 4.381899847991372e-06,
+      "loss": 0.385,
+      "step": 6984
+    },
+    {
+      "epoch": 1.7567907444668007,
+      "grad_norm": 0.30171477794647217,
+      "learning_rate": 4.38044785050952e-06,
+      "loss": 0.3481,
+      "step": 6985
+    },
+    {
+      "epoch": 1.7570422535211268,
+      "grad_norm": 0.2875131666660309,
+      "learning_rate": 4.378995906088436e-06,
+      "loss": 0.3327,
+      "step": 6986
+    },
+    {
+      "epoch": 1.7572937625754528,
+      "grad_norm": 0.2994171977043152,
+      "learning_rate": 4.377544014852466e-06,
+      "loss": 0.3355,
+      "step": 6987
+    },
+    {
+      "epoch": 1.7575452716297786,
+      "grad_norm": 0.3051261007785797,
+      "learning_rate": 4.3760921769259585e-06,
+      "loss": 0.3314,
+      "step": 6988
+    },
+    {
+      "epoch": 1.7577967806841046,
+      "grad_norm": 0.29593223333358765,
+      "learning_rate": 4.374640392433251e-06,
+      "loss": 0.3445,
+      "step": 6989
+    },
+    {
+      "epoch": 1.7580482897384306,
+      "grad_norm": 0.3013860285282135,
+      "learning_rate": 4.3731886614986815e-06,
+      "loss": 0.3661,
+      "step": 6990
+    },
+    {
+      "epoch": 1.7582997987927564,
+      "grad_norm": 0.2953547537326813,
+      "learning_rate": 4.371736984246584e-06,
+      "loss": 0.3256,
+      "step": 6991
+    },
+    {
+      "epoch": 1.7585513078470825,
+      "grad_norm": 0.3246706426143646,
+      "learning_rate": 4.370285360801281e-06,
+      "loss": 0.3298,
+      "step": 6992
+    },
+    {
+      "epoch": 1.7588028169014085,
+      "grad_norm": 0.3016349673271179,
+      "learning_rate": 4.3688337912871e-06,
+      "loss": 0.3447,
+      "step": 6993
+    },
+    {
+      "epoch": 1.7590543259557343,
+      "grad_norm": 0.295219361782074,
+      "learning_rate": 4.367382275828353e-06,
+      "loss": 0.3534,
+      "step": 6994
+    },
+    {
+      "epoch": 1.7593058350100603,
+      "grad_norm": 0.3270989954471588,
+      "learning_rate": 4.36593081454936e-06,
+      "loss": 0.3554,
+      "step": 6995
+    },
+    {
+      "epoch": 1.7595573440643864,
+      "grad_norm": 0.31494227051734924,
+      "learning_rate": 4.364479407574424e-06,
+      "loss": 0.3628,
+      "step": 6996
+    },
+    {
+      "epoch": 1.7598088531187122,
+      "grad_norm": 0.3002457022666931,
+      "learning_rate": 4.363028055027852e-06,
+      "loss": 0.3426,
+      "step": 6997
+    },
+    {
+      "epoch": 1.7600603621730382,
+      "grad_norm": 0.3095671236515045,
+      "learning_rate": 4.36157675703394e-06,
+      "loss": 0.3341,
+      "step": 6998
+    },
+    {
+      "epoch": 1.7603118712273642,
+      "grad_norm": 0.2987361252307892,
+      "learning_rate": 4.360125513716988e-06,
+      "loss": 0.3406,
+      "step": 6999
+    },
+    {
+      "epoch": 1.76056338028169,
+      "grad_norm": 0.30728045105934143,
+      "learning_rate": 4.35867432520128e-06,
+      "loss": 0.3648,
+      "step": 7000
+    },
+    {
+      "epoch": 1.760814889336016,
+      "grad_norm": 0.279835969209671,
+      "learning_rate": 4.357223191611103e-06,
+      "loss": 0.3338,
+      "step": 7001
+    },
+    {
+      "epoch": 1.761066398390342,
+      "grad_norm": 0.32082340121269226,
+      "learning_rate": 4.355772113070742e-06,
+      "loss": 0.364,
+      "step": 7002
+    },
+    {
+      "epoch": 1.7613179074446679,
+      "grad_norm": 0.2994687259197235,
+      "learning_rate": 4.354321089704466e-06,
+      "loss": 0.3723,
+      "step": 7003
+    },
+    {
+      "epoch": 1.7615694164989941,
+      "grad_norm": 0.3012014329433441,
+      "learning_rate": 4.352870121636553e-06,
+      "loss": 0.3351,
+      "step": 7004
+    },
+    {
+      "epoch": 1.76182092555332,
+      "grad_norm": 0.288789838552475,
+      "learning_rate": 4.351419208991262e-06,
+      "loss": 0.3324,
+      "step": 7005
+    },
+    {
+      "epoch": 1.7620724346076457,
+      "grad_norm": 0.2942191958427429,
+      "learning_rate": 4.349968351892861e-06,
+      "loss": 0.3425,
+      "step": 7006
+    },
+    {
+      "epoch": 1.762323943661972,
+      "grad_norm": 0.29075539112091064,
+      "learning_rate": 4.348517550465602e-06,
+      "loss": 0.3457,
+      "step": 7007
+    },
+    {
+      "epoch": 1.7625754527162978,
+      "grad_norm": 0.31270796060562134,
+      "learning_rate": 4.34706680483374e-06,
+      "loss": 0.3489,
+      "step": 7008
+    },
+    {
+      "epoch": 1.7628269617706236,
+      "grad_norm": 0.2807566821575165,
+      "learning_rate": 4.345616115121521e-06,
+      "loss": 0.3408,
+      "step": 7009
+    },
+    {
+      "epoch": 1.7630784708249498,
+      "grad_norm": 0.3200899064540863,
+      "learning_rate": 4.34416548145319e-06,
+      "loss": 0.3233,
+      "step": 7010
+    },
+    {
+      "epoch": 1.7633299798792756,
+      "grad_norm": 0.2903282642364502,
+      "learning_rate": 4.342714903952979e-06,
+      "loss": 0.3452,
+      "step": 7011
+    },
+    {
+      "epoch": 1.7635814889336014,
+      "grad_norm": 0.3277839720249176,
+      "learning_rate": 4.341264382745127e-06,
+      "loss": 0.3368,
+      "step": 7012
+    },
+    {
+      "epoch": 1.7638329979879277,
+      "grad_norm": 0.28098973631858826,
+      "learning_rate": 4.339813917953859e-06,
+      "loss": 0.3404,
+      "step": 7013
+    },
+    {
+      "epoch": 1.7640845070422535,
+      "grad_norm": 0.3015226423740387,
+      "learning_rate": 4.338363509703399e-06,
+      "loss": 0.3395,
+      "step": 7014
+    },
+    {
+      "epoch": 1.7643360160965795,
+      "grad_norm": 0.33274605870246887,
+      "learning_rate": 4.336913158117965e-06,
+      "loss": 0.3596,
+      "step": 7015
+    },
+    {
+      "epoch": 1.7645875251509056,
+      "grad_norm": 0.31674689054489136,
+      "learning_rate": 4.33546286332177e-06,
+      "loss": 0.3396,
+      "step": 7016
+    },
+    {
+      "epoch": 1.7648390342052314,
+      "grad_norm": 0.31130534410476685,
+      "learning_rate": 4.3340126254390255e-06,
+      "loss": 0.3906,
+      "step": 7017
+    },
+    {
+      "epoch": 1.7650905432595574,
+      "grad_norm": 0.30620113015174866,
+      "learning_rate": 4.3325624445939306e-06,
+      "loss": 0.3371,
+      "step": 7018
+    },
+    {
+      "epoch": 1.7653420523138834,
+      "grad_norm": 0.33001548051834106,
+      "learning_rate": 4.33111232091069e-06,
+      "loss": 0.3542,
+      "step": 7019
+    },
+    {
+      "epoch": 1.7655935613682092,
+      "grad_norm": 0.3208904266357422,
+      "learning_rate": 4.329662254513492e-06,
+      "loss": 0.3387,
+      "step": 7020
+    },
+    {
+      "epoch": 1.7658450704225352,
+      "grad_norm": 0.32378435134887695,
+      "learning_rate": 4.32821224552653e-06,
+      "loss": 0.3505,
+      "step": 7021
+    },
+    {
+      "epoch": 1.7660965794768613,
+      "grad_norm": 0.3116562068462372,
+      "learning_rate": 4.326762294073984e-06,
+      "loss": 0.3457,
+      "step": 7022
+    },
+    {
+      "epoch": 1.766348088531187,
+      "grad_norm": 0.3385581374168396,
+      "learning_rate": 4.3253124002800376e-06,
+      "loss": 0.3637,
+      "step": 7023
+    },
+    {
+      "epoch": 1.766599597585513,
+      "grad_norm": 0.29773086309432983,
+      "learning_rate": 4.323862564268862e-06,
+      "loss": 0.334,
+      "step": 7024
+    },
+    {
+      "epoch": 1.7668511066398391,
+      "grad_norm": 0.3357481062412262,
+      "learning_rate": 4.32241278616463e-06,
+      "loss": 0.3617,
+      "step": 7025
+    },
+    {
+      "epoch": 1.767102615694165,
+      "grad_norm": 0.32160642743110657,
+      "learning_rate": 4.320963066091503e-06,
+      "loss": 0.3617,
+      "step": 7026
+    },
+    {
+      "epoch": 1.767354124748491,
+      "grad_norm": 0.29699718952178955,
+      "learning_rate": 4.319513404173641e-06,
+      "loss": 0.3469,
+      "step": 7027
+    },
+    {
+      "epoch": 1.767605633802817,
+      "grad_norm": 0.31124669313430786,
+      "learning_rate": 4.318063800535199e-06,
+      "loss": 0.3523,
+      "step": 7028
+    },
+    {
+      "epoch": 1.7678571428571428,
+      "grad_norm": 0.3128409683704376,
+      "learning_rate": 4.316614255300326e-06,
+      "loss": 0.3102,
+      "step": 7029
+    },
+    {
+      "epoch": 1.7681086519114688,
+      "grad_norm": 0.2870144844055176,
+      "learning_rate": 4.315164768593167e-06,
+      "loss": 0.3445,
+      "step": 7030
+    },
+    {
+      "epoch": 1.7683601609657948,
+      "grad_norm": 0.3071323037147522,
+      "learning_rate": 4.313715340537861e-06,
+      "loss": 0.3552,
+      "step": 7031
+    },
+    {
+      "epoch": 1.7686116700201207,
+      "grad_norm": 0.33453187346458435,
+      "learning_rate": 4.312265971258544e-06,
+      "loss": 0.3461,
+      "step": 7032
+    },
+    {
+      "epoch": 1.7688631790744467,
+      "grad_norm": 0.3303498923778534,
+      "learning_rate": 4.310816660879342e-06,
+      "loss": 0.3508,
+      "step": 7033
+    },
+    {
+      "epoch": 1.7691146881287727,
+      "grad_norm": 0.30803173780441284,
+      "learning_rate": 4.3093674095243825e-06,
+      "loss": 0.3221,
+      "step": 7034
+    },
+    {
+      "epoch": 1.7693661971830985,
+      "grad_norm": 0.3237571716308594,
+      "learning_rate": 4.307918217317785e-06,
+      "loss": 0.3409,
+      "step": 7035
+    },
+    {
+      "epoch": 1.7696177062374245,
+      "grad_norm": 0.3842838704586029,
+      "learning_rate": 4.30646908438366e-06,
+      "loss": 0.3455,
+      "step": 7036
+    },
+    {
+      "epoch": 1.7698692152917506,
+      "grad_norm": 0.3077596426010132,
+      "learning_rate": 4.305020010846121e-06,
+      "loss": 0.3569,
+      "step": 7037
+    },
+    {
+      "epoch": 1.7701207243460764,
+      "grad_norm": 0.32035428285598755,
+      "learning_rate": 4.303570996829269e-06,
+      "loss": 0.3468,
+      "step": 7038
+    },
+    {
+      "epoch": 1.7703722334004024,
+      "grad_norm": 0.3082960546016693,
+      "learning_rate": 4.302122042457206e-06,
+      "loss": 0.3595,
+      "step": 7039
+    },
+    {
+      "epoch": 1.7706237424547284,
+      "grad_norm": 0.32630711793899536,
+      "learning_rate": 4.300673147854023e-06,
+      "loss": 0.323,
+      "step": 7040
+    },
+    {
+      "epoch": 1.7708752515090542,
+      "grad_norm": 0.3460298180580139,
+      "learning_rate": 4.299224313143811e-06,
+      "loss": 0.3253,
+      "step": 7041
+    },
+    {
+      "epoch": 1.7711267605633803,
+      "grad_norm": 0.30822697281837463,
+      "learning_rate": 4.297775538450651e-06,
+      "loss": 0.3232,
+      "step": 7042
+    },
+    {
+      "epoch": 1.7713782696177063,
+      "grad_norm": 0.3184284567832947,
+      "learning_rate": 4.296326823898625e-06,
+      "loss": 0.3431,
+      "step": 7043
+    },
+    {
+      "epoch": 1.771629778672032,
+      "grad_norm": 0.3198661804199219,
+      "learning_rate": 4.294878169611802e-06,
+      "loss": 0.3349,
+      "step": 7044
+    },
+    {
+      "epoch": 1.7718812877263581,
+      "grad_norm": 0.3085322976112366,
+      "learning_rate": 4.2934295757142526e-06,
+      "loss": 0.3351,
+      "step": 7045
+    },
+    {
+      "epoch": 1.7721327967806841,
+      "grad_norm": 0.31835395097732544,
+      "learning_rate": 4.291981042330042e-06,
+      "loss": 0.3702,
+      "step": 7046
+    },
+    {
+      "epoch": 1.77238430583501,
+      "grad_norm": 0.3139986991882324,
+      "learning_rate": 4.290532569583223e-06,
+      "loss": 0.33,
+      "step": 7047
+    },
+    {
+      "epoch": 1.772635814889336,
+      "grad_norm": 0.30537134408950806,
+      "learning_rate": 4.289084157597854e-06,
+      "loss": 0.3513,
+      "step": 7048
+    },
+    {
+      "epoch": 1.772887323943662,
+      "grad_norm": 0.2940613627433777,
+      "learning_rate": 4.287635806497977e-06,
+      "loss": 0.3523,
+      "step": 7049
+    },
+    {
+      "epoch": 1.7731388329979878,
+      "grad_norm": 0.3057369887828827,
+      "learning_rate": 4.2861875164076394e-06,
+      "loss": 0.3706,
+      "step": 7050
+    },
+    {
+      "epoch": 1.7733903420523138,
+      "grad_norm": 0.328139066696167,
+      "learning_rate": 4.284739287450873e-06,
+      "loss": 0.3603,
+      "step": 7051
+    },
+    {
+      "epoch": 1.7736418511066399,
+      "grad_norm": 0.32239678502082825,
+      "learning_rate": 4.283291119751714e-06,
+      "loss": 0.3415,
+      "step": 7052
+    },
+    {
+      "epoch": 1.7738933601609657,
+      "grad_norm": 0.3281613290309906,
+      "learning_rate": 4.2818430134341835e-06,
+      "loss": 0.3683,
+      "step": 7053
+    },
+    {
+      "epoch": 1.774144869215292,
+      "grad_norm": 0.29835787415504456,
+      "learning_rate": 4.280394968622309e-06,
+      "loss": 0.3449,
+      "step": 7054
+    },
+    {
+      "epoch": 1.7743963782696177,
+      "grad_norm": 0.3036617040634155,
+      "learning_rate": 4.2789469854401025e-06,
+      "loss": 0.3389,
+      "step": 7055
+    },
+    {
+      "epoch": 1.7746478873239435,
+      "grad_norm": 0.2952130138874054,
+      "learning_rate": 4.277499064011575e-06,
+      "loss": 0.3251,
+      "step": 7056
+    },
+    {
+      "epoch": 1.7748993963782698,
+      "grad_norm": 0.3045006990432739,
+      "learning_rate": 4.276051204460735e-06,
+      "loss": 0.3491,
+      "step": 7057
+    },
+    {
+      "epoch": 1.7751509054325956,
+      "grad_norm": 0.304092675447464,
+      "learning_rate": 4.274603406911578e-06,
+      "loss": 0.3373,
+      "step": 7058
+    },
+    {
+      "epoch": 1.7754024144869214,
+      "grad_norm": 0.3368243873119354,
+      "learning_rate": 4.273155671488103e-06,
+      "loss": 0.3633,
+      "step": 7059
+    },
+    {
+      "epoch": 1.7756539235412476,
+      "grad_norm": 0.3125411868095398,
+      "learning_rate": 4.271707998314296e-06,
+      "loss": 0.3279,
+      "step": 7060
+    },
+    {
+      "epoch": 1.7759054325955734,
+      "grad_norm": 0.31142720580101013,
+      "learning_rate": 4.270260387514145e-06,
+      "loss": 0.3408,
+      "step": 7061
+    },
+    {
+      "epoch": 1.7761569416498992,
+      "grad_norm": 0.3212254047393799,
+      "learning_rate": 4.268812839211624e-06,
+      "loss": 0.3604,
+      "step": 7062
+    },
+    {
+      "epoch": 1.7764084507042255,
+      "grad_norm": 0.3029959797859192,
+      "learning_rate": 4.267365353530711e-06,
+      "loss": 0.3625,
+      "step": 7063
+    },
+    {
+      "epoch": 1.7766599597585513,
+      "grad_norm": 0.30155423283576965,
+      "learning_rate": 4.265917930595371e-06,
+      "loss": 0.3484,
+      "step": 7064
+    },
+    {
+      "epoch": 1.7769114688128773,
+      "grad_norm": 0.3170860707759857,
+      "learning_rate": 4.264470570529569e-06,
+      "loss": 0.3534,
+      "step": 7065
+    },
+    {
+      "epoch": 1.7771629778672033,
+      "grad_norm": 0.3197275698184967,
+      "learning_rate": 4.2630232734572594e-06,
+      "loss": 0.3507,
+      "step": 7066
+    },
+    {
+      "epoch": 1.7774144869215291,
+      "grad_norm": 0.3387235999107361,
+      "learning_rate": 4.2615760395023956e-06,
+      "loss": 0.3531,
+      "step": 7067
+    },
+    {
+      "epoch": 1.7776659959758552,
+      "grad_norm": 0.30116546154022217,
+      "learning_rate": 4.260128868788927e-06,
+      "loss": 0.3455,
+      "step": 7068
+    },
+    {
+      "epoch": 1.7779175050301812,
+      "grad_norm": 0.32368162274360657,
+      "learning_rate": 4.25868176144079e-06,
+      "loss": 0.3489,
+      "step": 7069
+    },
+    {
+      "epoch": 1.778169014084507,
+      "grad_norm": 0.29791736602783203,
+      "learning_rate": 4.2572347175819245e-06,
+      "loss": 0.3504,
+      "step": 7070
+    },
+    {
+      "epoch": 1.778420523138833,
+      "grad_norm": 0.33331525325775146,
+      "learning_rate": 4.255787737336257e-06,
+      "loss": 0.3276,
+      "step": 7071
+    },
+    {
+      "epoch": 1.778672032193159,
+      "grad_norm": 0.28021112084388733,
+      "learning_rate": 4.254340820827715e-06,
+      "loss": 0.3205,
+      "step": 7072
+    },
+    {
+      "epoch": 1.7789235412474849,
+      "grad_norm": 0.29119399189949036,
+      "learning_rate": 4.252893968180215e-06,
+      "loss": 0.3536,
+      "step": 7073
+    },
+    {
+      "epoch": 1.779175050301811,
+      "grad_norm": 0.29139408469200134,
+      "learning_rate": 4.251447179517676e-06,
+      "loss": 0.3535,
+      "step": 7074
+    },
+    {
+      "epoch": 1.779426559356137,
+      "grad_norm": 0.3197605311870575,
+      "learning_rate": 4.250000454964001e-06,
+      "loss": 0.336,
+      "step": 7075
+    },
+    {
+      "epoch": 1.7796780684104627,
+      "grad_norm": 0.3005029261112213,
+      "learning_rate": 4.248553794643096e-06,
+      "loss": 0.3385,
+      "step": 7076
+    },
+    {
+      "epoch": 1.7799295774647887,
+      "grad_norm": 0.31606951355934143,
+      "learning_rate": 4.247107198678856e-06,
+      "loss": 0.3525,
+      "step": 7077
+    },
+    {
+      "epoch": 1.7801810865191148,
+      "grad_norm": 0.3125983476638794,
+      "learning_rate": 4.245660667195175e-06,
+      "loss": 0.3413,
+      "step": 7078
+    },
+    {
+      "epoch": 1.7804325955734406,
+      "grad_norm": 0.3108173906803131,
+      "learning_rate": 4.244214200315939e-06,
+      "loss": 0.3516,
+      "step": 7079
+    },
+    {
+      "epoch": 1.7806841046277666,
+      "grad_norm": 0.31591638922691345,
+      "learning_rate": 4.242767798165028e-06,
+      "loss": 0.3521,
+      "step": 7080
+    },
+    {
+      "epoch": 1.7809356136820926,
+      "grad_norm": 0.30157795548439026,
+      "learning_rate": 4.241321460866319e-06,
+      "loss": 0.3294,
+      "step": 7081
+    },
+    {
+      "epoch": 1.7811871227364184,
+      "grad_norm": 0.3349739611148834,
+      "learning_rate": 4.23987518854368e-06,
+      "loss": 0.3625,
+      "step": 7082
+    },
+    {
+      "epoch": 1.7814386317907445,
+      "grad_norm": 0.3226284086704254,
+      "learning_rate": 4.2384289813209754e-06,
+      "loss": 0.3247,
+      "step": 7083
+    },
+    {
+      "epoch": 1.7816901408450705,
+      "grad_norm": 0.3141878545284271,
+      "learning_rate": 4.236982839322064e-06,
+      "loss": 0.3171,
+      "step": 7084
+    },
+    {
+      "epoch": 1.7819416498993963,
+      "grad_norm": 0.3055501878261566,
+      "learning_rate": 4.235536762670801e-06,
+      "loss": 0.3376,
+      "step": 7085
+    },
+    {
+      "epoch": 1.7821931589537223,
+      "grad_norm": 0.31992486119270325,
+      "learning_rate": 4.23409075149103e-06,
+      "loss": 0.3444,
+      "step": 7086
+    },
+    {
+      "epoch": 1.7824446680080483,
+      "grad_norm": 0.3101184070110321,
+      "learning_rate": 4.2326448059065935e-06,
+      "loss": 0.3499,
+      "step": 7087
+    },
+    {
+      "epoch": 1.7826961770623742,
+      "grad_norm": 0.33392560482025146,
+      "learning_rate": 4.231198926041332e-06,
+      "loss": 0.3843,
+      "step": 7088
+    },
+    {
+      "epoch": 1.7829476861167002,
+      "grad_norm": 0.2991205155849457,
+      "learning_rate": 4.229753112019069e-06,
+      "loss": 0.3313,
+      "step": 7089
+    },
+    {
+      "epoch": 1.7831991951710262,
+      "grad_norm": 0.314127653837204,
+      "learning_rate": 4.2283073639636376e-06,
+      "loss": 0.3379,
+      "step": 7090
+    },
+    {
+      "epoch": 1.783450704225352,
+      "grad_norm": 0.31345435976982117,
+      "learning_rate": 4.226861681998849e-06,
+      "loss": 0.3603,
+      "step": 7091
+    },
+    {
+      "epoch": 1.783702213279678,
+      "grad_norm": 0.3256407082080841,
+      "learning_rate": 4.2254160662485236e-06,
+      "loss": 0.3277,
+      "step": 7092
+    },
+    {
+      "epoch": 1.783953722334004,
+      "grad_norm": 0.3122624158859253,
+      "learning_rate": 4.2239705168364636e-06,
+      "loss": 0.3499,
+      "step": 7093
+    },
+    {
+      "epoch": 1.7842052313883299,
+      "grad_norm": 0.2897671163082123,
+      "learning_rate": 4.222525033886476e-06,
+      "loss": 0.3465,
+      "step": 7094
+    },
+    {
+      "epoch": 1.784456740442656,
+      "grad_norm": 0.31568416953086853,
+      "learning_rate": 4.221079617522354e-06,
+      "loss": 0.3393,
+      "step": 7095
+    },
+    {
+      "epoch": 1.784708249496982,
+      "grad_norm": 0.31326398253440857,
+      "learning_rate": 4.219634267867892e-06,
+      "loss": 0.3521,
+      "step": 7096
+    },
+    {
+      "epoch": 1.7849597585513077,
+      "grad_norm": 0.346648246049881,
+      "learning_rate": 4.2181889850468704e-06,
+      "loss": 0.3414,
+      "step": 7097
+    },
+    {
+      "epoch": 1.7852112676056338,
+      "grad_norm": 0.29954788088798523,
+      "learning_rate": 4.216743769183071e-06,
+      "loss": 0.3569,
+      "step": 7098
+    },
+    {
+      "epoch": 1.7854627766599598,
+      "grad_norm": 0.3176994323730469,
+      "learning_rate": 4.215298620400271e-06,
+      "loss": 0.3513,
+      "step": 7099
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.3110980987548828,
+      "learning_rate": 4.213853538822232e-06,
+      "loss": 0.3463,
+      "step": 7100
+    },
+    {
+      "epoch": 1.7859657947686118,
+      "grad_norm": 0.3204716145992279,
+      "learning_rate": 4.212408524572722e-06,
+      "loss": 0.3722,
+      "step": 7101
+    },
+    {
+      "epoch": 1.7862173038229376,
+      "grad_norm": 0.3436601161956787,
+      "learning_rate": 4.210963577775492e-06,
+      "loss": 0.3489,
+      "step": 7102
+    },
+    {
+      "epoch": 1.7864688128772634,
+      "grad_norm": 0.31376877427101135,
+      "learning_rate": 4.209518698554298e-06,
+      "loss": 0.3449,
+      "step": 7103
+    },
+    {
+      "epoch": 1.7867203219315897,
+      "grad_norm": 0.31126469373703003,
+      "learning_rate": 4.2080738870328795e-06,
+      "loss": 0.3685,
+      "step": 7104
+    },
+    {
+      "epoch": 1.7869718309859155,
+      "grad_norm": 0.29963016510009766,
+      "learning_rate": 4.206629143334981e-06,
+      "loss": 0.3533,
+      "step": 7105
+    },
+    {
+      "epoch": 1.7872233400402413,
+      "grad_norm": 0.3086322546005249,
+      "learning_rate": 4.20518446758433e-06,
+      "loss": 0.3215,
+      "step": 7106
+    },
+    {
+      "epoch": 1.7874748490945676,
+      "grad_norm": 0.35310494899749756,
+      "learning_rate": 4.20373985990466e-06,
+      "loss": 0.3563,
+      "step": 7107
+    },
+    {
+      "epoch": 1.7877263581488934,
+      "grad_norm": 0.2881607413291931,
+      "learning_rate": 4.202295320419687e-06,
+      "loss": 0.3393,
+      "step": 7108
+    },
+    {
+      "epoch": 1.7879778672032192,
+      "grad_norm": 0.3135796785354614,
+      "learning_rate": 4.2008508492531305e-06,
+      "loss": 0.3605,
+      "step": 7109
+    },
+    {
+      "epoch": 1.7882293762575454,
+      "grad_norm": 0.3010595440864563,
+      "learning_rate": 4.1994064465287e-06,
+      "loss": 0.356,
+      "step": 7110
+    },
+    {
+      "epoch": 1.7884808853118712,
+      "grad_norm": 0.3232562839984894,
+      "learning_rate": 4.1979621123700976e-06,
+      "loss": 0.335,
+      "step": 7111
+    },
+    {
+      "epoch": 1.788732394366197,
+      "grad_norm": 0.3245859742164612,
+      "learning_rate": 4.196517846901025e-06,
+      "loss": 0.3427,
+      "step": 7112
+    },
+    {
+      "epoch": 1.7889839034205233,
+      "grad_norm": 0.323621928691864,
+      "learning_rate": 4.195073650245169e-06,
+      "loss": 0.3566,
+      "step": 7113
+    },
+    {
+      "epoch": 1.789235412474849,
+      "grad_norm": 0.3078625500202179,
+      "learning_rate": 4.193629522526223e-06,
+      "loss": 0.3304,
+      "step": 7114
+    },
+    {
+      "epoch": 1.789486921529175,
+      "grad_norm": 0.3693898320198059,
+      "learning_rate": 4.19218546386786e-06,
+      "loss": 0.3466,
+      "step": 7115
+    },
+    {
+      "epoch": 1.7897384305835011,
+      "grad_norm": 0.2945307493209839,
+      "learning_rate": 4.190741474393762e-06,
+      "loss": 0.3452,
+      "step": 7116
+    },
+    {
+      "epoch": 1.789989939637827,
+      "grad_norm": 0.3134191930294037,
+      "learning_rate": 4.18929755422759e-06,
+      "loss": 0.3609,
+      "step": 7117
+    },
+    {
+      "epoch": 1.790241448692153,
+      "grad_norm": 0.3266321122646332,
+      "learning_rate": 4.187853703493014e-06,
+      "loss": 0.3513,
+      "step": 7118
+    },
+    {
+      "epoch": 1.790492957746479,
+      "grad_norm": 0.3312418758869171,
+      "learning_rate": 4.186409922313686e-06,
+      "loss": 0.3385,
+      "step": 7119
+    },
+    {
+      "epoch": 1.7907444668008048,
+      "grad_norm": 0.30360767245292664,
+      "learning_rate": 4.184966210813258e-06,
+      "loss": 0.3557,
+      "step": 7120
+    },
+    {
+      "epoch": 1.7909959758551308,
+      "grad_norm": 0.3027835786342621,
+      "learning_rate": 4.183522569115377e-06,
+      "loss": 0.3553,
+      "step": 7121
+    },
+    {
+      "epoch": 1.7912474849094568,
+      "grad_norm": 0.317096084356308,
+      "learning_rate": 4.182078997343678e-06,
+      "loss": 0.3794,
+      "step": 7122
+    },
+    {
+      "epoch": 1.7914989939637826,
+      "grad_norm": 0.31175005435943604,
+      "learning_rate": 4.180635495621798e-06,
+      "loss": 0.3567,
+      "step": 7123
+    },
+    {
+      "epoch": 1.7917505030181087,
+      "grad_norm": 0.30733904242515564,
+      "learning_rate": 4.1791920640733596e-06,
+      "loss": 0.3446,
+      "step": 7124
+    },
+    {
+      "epoch": 1.7920020120724347,
+      "grad_norm": 0.3084380030632019,
+      "learning_rate": 4.177748702821988e-06,
+      "loss": 0.3611,
+      "step": 7125
+    },
+    {
+      "epoch": 1.7922535211267605,
+      "grad_norm": 0.3090236783027649,
+      "learning_rate": 4.176305411991295e-06,
+      "loss": 0.3313,
+      "step": 7126
+    },
+    {
+      "epoch": 1.7925050301810865,
+      "grad_norm": 0.2927131950855255,
+      "learning_rate": 4.174862191704892e-06,
+      "loss": 0.3476,
+      "step": 7127
+    },
+    {
+      "epoch": 1.7927565392354126,
+      "grad_norm": 0.3258267343044281,
+      "learning_rate": 4.173419042086377e-06,
+      "loss": 0.3354,
+      "step": 7128
+    },
+    {
+      "epoch": 1.7930080482897384,
+      "grad_norm": 0.2800391614437103,
+      "learning_rate": 4.171975963259354e-06,
+      "loss": 0.352,
+      "step": 7129
+    },
+    {
+      "epoch": 1.7932595573440644,
+      "grad_norm": 0.32517415285110474,
+      "learning_rate": 4.170532955347406e-06,
+      "loss": 0.3151,
+      "step": 7130
+    },
+    {
+      "epoch": 1.7935110663983904,
+      "grad_norm": 0.30167156457901,
+      "learning_rate": 4.169090018474122e-06,
+      "loss": 0.3481,
+      "step": 7131
+    },
+    {
+      "epoch": 1.7937625754527162,
+      "grad_norm": 0.31728750467300415,
+      "learning_rate": 4.1676471527630815e-06,
+      "loss": 0.3554,
+      "step": 7132
+    },
+    {
+      "epoch": 1.7940140845070423,
+      "grad_norm": 0.306044340133667,
+      "learning_rate": 4.1662043583378534e-06,
+      "loss": 0.3496,
+      "step": 7133
+    },
+    {
+      "epoch": 1.7942655935613683,
+      "grad_norm": 0.3083249628543854,
+      "learning_rate": 4.164761635322007e-06,
+      "loss": 0.3371,
+      "step": 7134
+    },
+    {
+      "epoch": 1.794517102615694,
+      "grad_norm": 0.2971269190311432,
+      "learning_rate": 4.163318983839101e-06,
+      "loss": 0.3446,
+      "step": 7135
+    },
+    {
+      "epoch": 1.79476861167002,
+      "grad_norm": 0.29189154505729675,
+      "learning_rate": 4.1618764040126905e-06,
+      "loss": 0.3526,
+      "step": 7136
+    },
+    {
+      "epoch": 1.7950201207243461,
+      "grad_norm": 0.31266725063323975,
+      "learning_rate": 4.1604338959663204e-06,
+      "loss": 0.3506,
+      "step": 7137
+    },
+    {
+      "epoch": 1.795271629778672,
+      "grad_norm": 0.294013112783432,
+      "learning_rate": 4.158991459823538e-06,
+      "loss": 0.3593,
+      "step": 7138
+    },
+    {
+      "epoch": 1.795523138832998,
+      "grad_norm": 0.31676429510116577,
+      "learning_rate": 4.1575490957078725e-06,
+      "loss": 0.3357,
+      "step": 7139
+    },
+    {
+      "epoch": 1.795774647887324,
+      "grad_norm": 0.3218402862548828,
+      "learning_rate": 4.15610680374286e-06,
+      "loss": 0.3462,
+      "step": 7140
+    },
+    {
+      "epoch": 1.7960261569416498,
+      "grad_norm": 0.2809396982192993,
+      "learning_rate": 4.154664584052018e-06,
+      "loss": 0.3368,
+      "step": 7141
+    },
+    {
+      "epoch": 1.7962776659959758,
+      "grad_norm": 0.30327579379081726,
+      "learning_rate": 4.153222436758866e-06,
+      "loss": 0.3773,
+      "step": 7142
+    },
+    {
+      "epoch": 1.7965291750503019,
+      "grad_norm": 0.2996901869773865,
+      "learning_rate": 4.151780361986915e-06,
+      "loss": 0.3433,
+      "step": 7143
+    },
+    {
+      "epoch": 1.7967806841046277,
+      "grad_norm": 0.30090850591659546,
+      "learning_rate": 4.1503383598596705e-06,
+      "loss": 0.3518,
+      "step": 7144
+    },
+    {
+      "epoch": 1.7970321931589537,
+      "grad_norm": 0.3128022849559784,
+      "learning_rate": 4.148896430500629e-06,
+      "loss": 0.3521,
+      "step": 7145
+    },
+    {
+      "epoch": 1.7972837022132797,
+      "grad_norm": 0.26742905378341675,
+      "learning_rate": 4.147454574033284e-06,
+      "loss": 0.3435,
+      "step": 7146
+    },
+    {
+      "epoch": 1.7975352112676055,
+      "grad_norm": 0.31213289499282837,
+      "learning_rate": 4.146012790581121e-06,
+      "loss": 0.3451,
+      "step": 7147
+    },
+    {
+      "epoch": 1.7977867203219315,
+      "grad_norm": 0.2901371419429779,
+      "learning_rate": 4.144571080267621e-06,
+      "loss": 0.3496,
+      "step": 7148
+    },
+    {
+      "epoch": 1.7980382293762576,
+      "grad_norm": 0.3160885274410248,
+      "learning_rate": 4.143129443216256e-06,
+      "loss": 0.3258,
+      "step": 7149
+    },
+    {
+      "epoch": 1.7982897384305834,
+      "grad_norm": 0.3092404007911682,
+      "learning_rate": 4.141687879550494e-06,
+      "loss": 0.3679,
+      "step": 7150
+    },
+    {
+      "epoch": 1.7985412474849096,
+      "grad_norm": 0.3174445629119873,
+      "learning_rate": 4.140246389393794e-06,
+      "loss": 0.3338,
+      "step": 7151
+    },
+    {
+      "epoch": 1.7987927565392354,
+      "grad_norm": 0.28808581829071045,
+      "learning_rate": 4.138804972869613e-06,
+      "loss": 0.3477,
+      "step": 7152
+    },
+    {
+      "epoch": 1.7990442655935612,
+      "grad_norm": 0.3383561670780182,
+      "learning_rate": 4.137363630101398e-06,
+      "loss": 0.3548,
+      "step": 7153
+    },
+    {
+      "epoch": 1.7992957746478875,
+      "grad_norm": 0.30686813592910767,
+      "learning_rate": 4.135922361212593e-06,
+      "loss": 0.3549,
+      "step": 7154
+    },
+    {
+      "epoch": 1.7995472837022133,
+      "grad_norm": 0.2914334833621979,
+      "learning_rate": 4.134481166326631e-06,
+      "loss": 0.3306,
+      "step": 7155
+    },
+    {
+      "epoch": 1.799798792756539,
+      "grad_norm": 0.3026004135608673,
+      "learning_rate": 4.133040045566942e-06,
+      "loss": 0.3397,
+      "step": 7156
+    },
+    {
+      "epoch": 1.8000503018108653,
+      "grad_norm": 0.3466000258922577,
+      "learning_rate": 4.13159899905695e-06,
+      "loss": 0.3348,
+      "step": 7157
+    },
+    {
+      "epoch": 1.8003018108651911,
+      "grad_norm": 0.30805760622024536,
+      "learning_rate": 4.130158026920072e-06,
+      "loss": 0.3512,
+      "step": 7158
+    },
+    {
+      "epoch": 1.800553319919517,
+      "grad_norm": 0.32429513335227966,
+      "learning_rate": 4.128717129279715e-06,
+      "loss": 0.3828,
+      "step": 7159
+    },
+    {
+      "epoch": 1.8008048289738432,
+      "grad_norm": 0.3071700930595398,
+      "learning_rate": 4.127276306259288e-06,
+      "loss": 0.3582,
+      "step": 7160
+    },
+    {
+      "epoch": 1.801056338028169,
+      "grad_norm": 0.31725555658340454,
+      "learning_rate": 4.125835557982183e-06,
+      "loss": 0.361,
+      "step": 7161
+    },
+    {
+      "epoch": 1.8013078470824948,
+      "grad_norm": 0.31528329849243164,
+      "learning_rate": 4.124394884571796e-06,
+      "loss": 0.3451,
+      "step": 7162
+    },
+    {
+      "epoch": 1.801559356136821,
+      "grad_norm": 0.3092787563800812,
+      "learning_rate": 4.122954286151507e-06,
+      "loss": 0.3431,
+      "step": 7163
+    },
+    {
+      "epoch": 1.8018108651911469,
+      "grad_norm": 0.3064475953578949,
+      "learning_rate": 4.121513762844696e-06,
+      "loss": 0.3313,
+      "step": 7164
+    },
+    {
+      "epoch": 1.8020623742454729,
+      "grad_norm": 0.30549487471580505,
+      "learning_rate": 4.120073314774739e-06,
+      "loss": 0.3576,
+      "step": 7165
+    },
+    {
+      "epoch": 1.802313883299799,
+      "grad_norm": 0.30158209800720215,
+      "learning_rate": 4.118632942064995e-06,
+      "loss": 0.3607,
+      "step": 7166
+    },
+    {
+      "epoch": 1.8025653923541247,
+      "grad_norm": 0.2950282394886017,
+      "learning_rate": 4.117192644838827e-06,
+      "loss": 0.3414,
+      "step": 7167
+    },
+    {
+      "epoch": 1.8028169014084507,
+      "grad_norm": 0.3308418393135071,
+      "learning_rate": 4.115752423219585e-06,
+      "loss": 0.3561,
+      "step": 7168
+    },
+    {
+      "epoch": 1.8030684104627768,
+      "grad_norm": 0.30721715092658997,
+      "learning_rate": 4.114312277330617e-06,
+      "loss": 0.3459,
+      "step": 7169
+    },
+    {
+      "epoch": 1.8033199195171026,
+      "grad_norm": 0.300238698720932,
+      "learning_rate": 4.112872207295262e-06,
+      "loss": 0.3266,
+      "step": 7170
+    },
+    {
+      "epoch": 1.8035714285714286,
+      "grad_norm": 0.3531527519226074,
+      "learning_rate": 4.1114322132368524e-06,
+      "loss": 0.3471,
+      "step": 7171
+    },
+    {
+      "epoch": 1.8038229376257546,
+      "grad_norm": 0.3371135890483856,
+      "learning_rate": 4.109992295278714e-06,
+      "loss": 0.3284,
+      "step": 7172
+    },
+    {
+      "epoch": 1.8040744466800804,
+      "grad_norm": 0.30015435814857483,
+      "learning_rate": 4.108552453544169e-06,
+      "loss": 0.3471,
+      "step": 7173
+    },
+    {
+      "epoch": 1.8043259557344065,
+      "grad_norm": 0.29010263085365295,
+      "learning_rate": 4.107112688156528e-06,
+      "loss": 0.3366,
+      "step": 7174
+    },
+    {
+      "epoch": 1.8045774647887325,
+      "grad_norm": 0.3293246626853943,
+      "learning_rate": 4.105672999239098e-06,
+      "loss": 0.3481,
+      "step": 7175
+    },
+    {
+      "epoch": 1.8048289738430583,
+      "grad_norm": 0.34103676676750183,
+      "learning_rate": 4.104233386915185e-06,
+      "loss": 0.341,
+      "step": 7176
+    },
+    {
+      "epoch": 1.8050804828973843,
+      "grad_norm": 0.2750874161720276,
+      "learning_rate": 4.102793851308074e-06,
+      "loss": 0.3327,
+      "step": 7177
+    },
+    {
+      "epoch": 1.8053319919517103,
+      "grad_norm": 0.30653923749923706,
+      "learning_rate": 4.101354392541061e-06,
+      "loss": 0.335,
+      "step": 7178
+    },
+    {
+      "epoch": 1.8055835010060362,
+      "grad_norm": 0.32137489318847656,
+      "learning_rate": 4.099915010737419e-06,
+      "loss": 0.3409,
+      "step": 7179
+    },
+    {
+      "epoch": 1.8058350100603622,
+      "grad_norm": 0.3056570291519165,
+      "learning_rate": 4.098475706020428e-06,
+      "loss": 0.3229,
+      "step": 7180
+    },
+    {
+      "epoch": 1.8060865191146882,
+      "grad_norm": 0.30447155237197876,
+      "learning_rate": 4.0970364785133506e-06,
+      "loss": 0.3495,
+      "step": 7181
+    },
+    {
+      "epoch": 1.806338028169014,
+      "grad_norm": 0.2888796925544739,
+      "learning_rate": 4.0955973283394525e-06,
+      "loss": 0.3445,
+      "step": 7182
+    },
+    {
+      "epoch": 1.80658953722334,
+      "grad_norm": 0.29918986558914185,
+      "learning_rate": 4.094158255621983e-06,
+      "loss": 0.3205,
+      "step": 7183
+    },
+    {
+      "epoch": 1.806841046277666,
+      "grad_norm": 0.31096193194389343,
+      "learning_rate": 4.0927192604841935e-06,
+      "loss": 0.3415,
+      "step": 7184
+    },
+    {
+      "epoch": 1.8070925553319919,
+      "grad_norm": 0.3002293109893799,
+      "learning_rate": 4.0912803430493215e-06,
+      "loss": 0.33,
+      "step": 7185
+    },
+    {
+      "epoch": 1.807344064386318,
+      "grad_norm": 0.2879883348941803,
+      "learning_rate": 4.089841503440603e-06,
+      "loss": 0.3347,
+      "step": 7186
+    },
+    {
+      "epoch": 1.807595573440644,
+      "grad_norm": 0.2978774607181549,
+      "learning_rate": 4.088402741781269e-06,
+      "loss": 0.3424,
+      "step": 7187
+    },
+    {
+      "epoch": 1.8078470824949697,
+      "grad_norm": 0.3050556480884552,
+      "learning_rate": 4.086964058194534e-06,
+      "loss": 0.3397,
+      "step": 7188
+    },
+    {
+      "epoch": 1.8080985915492958,
+      "grad_norm": 0.30784785747528076,
+      "learning_rate": 4.085525452803618e-06,
+      "loss": 0.3586,
+      "step": 7189
+    },
+    {
+      "epoch": 1.8083501006036218,
+      "grad_norm": 0.3016720712184906,
+      "learning_rate": 4.084086925731723e-06,
+      "loss": 0.3537,
+      "step": 7190
+    },
+    {
+      "epoch": 1.8086016096579476,
+      "grad_norm": 0.3045290410518646,
+      "learning_rate": 4.0826484771020565e-06,
+      "loss": 0.3393,
+      "step": 7191
+    },
+    {
+      "epoch": 1.8088531187122736,
+      "grad_norm": 0.2894311249256134,
+      "learning_rate": 4.081210107037806e-06,
+      "loss": 0.351,
+      "step": 7192
+    },
+    {
+      "epoch": 1.8091046277665996,
+      "grad_norm": 0.30208760499954224,
+      "learning_rate": 4.079771815662164e-06,
+      "loss": 0.3349,
+      "step": 7193
+    },
+    {
+      "epoch": 1.8093561368209254,
+      "grad_norm": 0.3042428195476532,
+      "learning_rate": 4.078333603098307e-06,
+      "loss": 0.3831,
+      "step": 7194
+    },
+    {
+      "epoch": 1.8096076458752515,
+      "grad_norm": 0.3321954905986786,
+      "learning_rate": 4.076895469469413e-06,
+      "loss": 0.3566,
+      "step": 7195
+    },
+    {
+      "epoch": 1.8098591549295775,
+      "grad_norm": 0.3012005686759949,
+      "learning_rate": 4.075457414898646e-06,
+      "loss": 0.3242,
+      "step": 7196
+    },
+    {
+      "epoch": 1.8101106639839033,
+      "grad_norm": 0.29138419032096863,
+      "learning_rate": 4.074019439509168e-06,
+      "loss": 0.3078,
+      "step": 7197
+    },
+    {
+      "epoch": 1.8103621730382293,
+      "grad_norm": 0.29968154430389404,
+      "learning_rate": 4.072581543424132e-06,
+      "loss": 0.3189,
+      "step": 7198
+    },
+    {
+      "epoch": 1.8106136820925554,
+      "grad_norm": 0.3333386778831482,
+      "learning_rate": 4.071143726766683e-06,
+      "loss": 0.3542,
+      "step": 7199
+    },
+    {
+      "epoch": 1.8108651911468812,
+      "grad_norm": 0.31147098541259766,
+      "learning_rate": 4.069705989659966e-06,
+      "loss": 0.3586,
+      "step": 7200
+    },
+    {
+      "epoch": 1.8111167002012074,
+      "grad_norm": 0.31270739436149597,
+      "learning_rate": 4.0682683322271086e-06,
+      "loss": 0.3408,
+      "step": 7201
+    },
+    {
+      "epoch": 1.8113682092555332,
+      "grad_norm": 0.30901169776916504,
+      "learning_rate": 4.066830754591242e-06,
+      "loss": 0.346,
+      "step": 7202
+    },
+    {
+      "epoch": 1.811619718309859,
+      "grad_norm": 0.30854368209838867,
+      "learning_rate": 4.065393256875481e-06,
+      "loss": 0.3383,
+      "step": 7203
+    },
+    {
+      "epoch": 1.8118712273641853,
+      "grad_norm": 0.304512083530426,
+      "learning_rate": 4.063955839202943e-06,
+      "loss": 0.312,
+      "step": 7204
+    },
+    {
+      "epoch": 1.812122736418511,
+      "grad_norm": 0.29708781838417053,
+      "learning_rate": 4.062518501696729e-06,
+      "loss": 0.3454,
+      "step": 7205
+    },
+    {
+      "epoch": 1.8123742454728369,
+      "grad_norm": 0.304839551448822,
+      "learning_rate": 4.061081244479943e-06,
+      "loss": 0.3573,
+      "step": 7206
+    },
+    {
+      "epoch": 1.8126257545271631,
+      "grad_norm": 0.3408084213733673,
+      "learning_rate": 4.059644067675673e-06,
+      "loss": 0.3337,
+      "step": 7207
+    },
+    {
+      "epoch": 1.812877263581489,
+      "grad_norm": 0.2825142443180084,
+      "learning_rate": 4.058206971407006e-06,
+      "loss": 0.378,
+      "step": 7208
+    },
+    {
+      "epoch": 1.8131287726358147,
+      "grad_norm": 0.285431444644928,
+      "learning_rate": 4.05676995579702e-06,
+      "loss": 0.3331,
+      "step": 7209
+    },
+    {
+      "epoch": 1.813380281690141,
+      "grad_norm": 0.32098081707954407,
+      "learning_rate": 4.055333020968787e-06,
+      "loss": 0.342,
+      "step": 7210
+    },
+    {
+      "epoch": 1.8136317907444668,
+      "grad_norm": 0.2935009002685547,
+      "learning_rate": 4.05389616704537e-06,
+      "loss": 0.3568,
+      "step": 7211
+    },
+    {
+      "epoch": 1.8138832997987926,
+      "grad_norm": 0.29505932331085205,
+      "learning_rate": 4.052459394149829e-06,
+      "loss": 0.351,
+      "step": 7212
+    },
+    {
+      "epoch": 1.8141348088531188,
+      "grad_norm": 0.30334311723709106,
+      "learning_rate": 4.0510227024052115e-06,
+      "loss": 0.3446,
+      "step": 7213
+    },
+    {
+      "epoch": 1.8143863179074446,
+      "grad_norm": 0.3207496404647827,
+      "learning_rate": 4.049586091934563e-06,
+      "loss": 0.3443,
+      "step": 7214
+    },
+    {
+      "epoch": 1.8146378269617707,
+      "grad_norm": 0.31141701340675354,
+      "learning_rate": 4.048149562860921e-06,
+      "loss": 0.3418,
+      "step": 7215
+    },
+    {
+      "epoch": 1.8148893360160967,
+      "grad_norm": 0.3144972622394562,
+      "learning_rate": 4.046713115307314e-06,
+      "loss": 0.3316,
+      "step": 7216
+    },
+    {
+      "epoch": 1.8151408450704225,
+      "grad_norm": 0.30265048146247864,
+      "learning_rate": 4.045276749396764e-06,
+      "loss": 0.3368,
+      "step": 7217
+    },
+    {
+      "epoch": 1.8153923541247485,
+      "grad_norm": 0.2881247401237488,
+      "learning_rate": 4.043840465252289e-06,
+      "loss": 0.3493,
+      "step": 7218
+    },
+    {
+      "epoch": 1.8156438631790746,
+      "grad_norm": 0.30661842226982117,
+      "learning_rate": 4.042404262996894e-06,
+      "loss": 0.3615,
+      "step": 7219
+    },
+    {
+      "epoch": 1.8158953722334004,
+      "grad_norm": 0.2946421205997467,
+      "learning_rate": 4.0409681427535855e-06,
+      "loss": 0.331,
+      "step": 7220
+    },
+    {
+      "epoch": 1.8161468812877264,
+      "grad_norm": 0.34275466203689575,
+      "learning_rate": 4.039532104645354e-06,
+      "loss": 0.3441,
+      "step": 7221
+    },
+    {
+      "epoch": 1.8163983903420524,
+      "grad_norm": 0.3059110641479492,
+      "learning_rate": 4.0380961487951915e-06,
+      "loss": 0.3422,
+      "step": 7222
+    },
+    {
+      "epoch": 1.8166498993963782,
+      "grad_norm": 0.29724571108818054,
+      "learning_rate": 4.0366602753260745e-06,
+      "loss": 0.3245,
+      "step": 7223
+    },
+    {
+      "epoch": 1.8169014084507042,
+      "grad_norm": 0.31005269289016724,
+      "learning_rate": 4.035224484360979e-06,
+      "loss": 0.335,
+      "step": 7224
+    },
+    {
+      "epoch": 1.8171529175050303,
+      "grad_norm": 0.31909066438674927,
+      "learning_rate": 4.03378877602287e-06,
+      "loss": 0.359,
+      "step": 7225
+    },
+    {
+      "epoch": 1.817404426559356,
+      "grad_norm": 0.2967131435871124,
+      "learning_rate": 4.032353150434709e-06,
+      "loss": 0.3227,
+      "step": 7226
+    },
+    {
+      "epoch": 1.817655935613682,
+      "grad_norm": 0.3029683828353882,
+      "learning_rate": 4.030917607719446e-06,
+      "loss": 0.3581,
+      "step": 7227
+    },
+    {
+      "epoch": 1.8179074446680081,
+      "grad_norm": 0.3194609582424164,
+      "learning_rate": 4.029482148000028e-06,
+      "loss": 0.3354,
+      "step": 7228
+    },
+    {
+      "epoch": 1.818158953722334,
+      "grad_norm": 0.3186091482639313,
+      "learning_rate": 4.028046771399391e-06,
+      "loss": 0.3477,
+      "step": 7229
+    },
+    {
+      "epoch": 1.81841046277666,
+      "grad_norm": 0.28328919410705566,
+      "learning_rate": 4.026611478040468e-06,
+      "loss": 0.3498,
+      "step": 7230
+    },
+    {
+      "epoch": 1.818661971830986,
+      "grad_norm": 0.28632745146751404,
+      "learning_rate": 4.025176268046184e-06,
+      "loss": 0.3525,
+      "step": 7231
+    },
+    {
+      "epoch": 1.8189134808853118,
+      "grad_norm": 0.310872346162796,
+      "learning_rate": 4.023741141539453e-06,
+      "loss": 0.3418,
+      "step": 7232
+    },
+    {
+      "epoch": 1.8191649899396378,
+      "grad_norm": 0.3138779401779175,
+      "learning_rate": 4.022306098643186e-06,
+      "loss": 0.3636,
+      "step": 7233
+    },
+    {
+      "epoch": 1.8194164989939638,
+      "grad_norm": 0.31211575865745544,
+      "learning_rate": 4.020871139480285e-06,
+      "loss": 0.3466,
+      "step": 7234
+    },
+    {
+      "epoch": 1.8196680080482897,
+      "grad_norm": 0.2843218445777893,
+      "learning_rate": 4.019436264173646e-06,
+      "loss": 0.3238,
+      "step": 7235
+    },
+    {
+      "epoch": 1.8199195171026157,
+      "grad_norm": 0.3115207552909851,
+      "learning_rate": 4.018001472846156e-06,
+      "loss": 0.342,
+      "step": 7236
+    },
+    {
+      "epoch": 1.8201710261569417,
+      "grad_norm": 0.3359984755516052,
+      "learning_rate": 4.0165667656206975e-06,
+      "loss": 0.3548,
+      "step": 7237
+    },
+    {
+      "epoch": 1.8204225352112675,
+      "grad_norm": 0.31638795137405396,
+      "learning_rate": 4.0151321426201414e-06,
+      "loss": 0.363,
+      "step": 7238
+    },
+    {
+      "epoch": 1.8206740442655935,
+      "grad_norm": 0.3151678442955017,
+      "learning_rate": 4.013697603967356e-06,
+      "loss": 0.3564,
+      "step": 7239
+    },
+    {
+      "epoch": 1.8209255533199196,
+      "grad_norm": 0.2856763005256653,
+      "learning_rate": 4.012263149785203e-06,
+      "loss": 0.3497,
+      "step": 7240
+    },
+    {
+      "epoch": 1.8211770623742454,
+      "grad_norm": 0.32092490792274475,
+      "learning_rate": 4.010828780196529e-06,
+      "loss": 0.3394,
+      "step": 7241
+    },
+    {
+      "epoch": 1.8214285714285714,
+      "grad_norm": 0.3086574971675873,
+      "learning_rate": 4.009394495324185e-06,
+      "loss": 0.3707,
+      "step": 7242
+    },
+    {
+      "epoch": 1.8216800804828974,
+      "grad_norm": 0.3423568606376648,
+      "learning_rate": 4.007960295291002e-06,
+      "loss": 0.3474,
+      "step": 7243
+    },
+    {
+      "epoch": 1.8219315895372232,
+      "grad_norm": 0.28429266810417175,
+      "learning_rate": 4.006526180219816e-06,
+      "loss": 0.3341,
+      "step": 7244
+    },
+    {
+      "epoch": 1.8221830985915493,
+      "grad_norm": 0.2965366244316101,
+      "learning_rate": 4.005092150233445e-06,
+      "loss": 0.3395,
+      "step": 7245
+    },
+    {
+      "epoch": 1.8224346076458753,
+      "grad_norm": 0.34402769804000854,
+      "learning_rate": 4.00365820545471e-06,
+      "loss": 0.3643,
+      "step": 7246
+    },
+    {
+      "epoch": 1.822686116700201,
+      "grad_norm": 0.30077889561653137,
+      "learning_rate": 4.002224346006415e-06,
+      "loss": 0.3334,
+      "step": 7247
+    },
+    {
+      "epoch": 1.8229376257545271,
+      "grad_norm": 0.31439444422721863,
+      "learning_rate": 4.000790572011365e-06,
+      "loss": 0.3382,
+      "step": 7248
+    },
+    {
+      "epoch": 1.8231891348088531,
+      "grad_norm": 0.34265372157096863,
+      "learning_rate": 3.999356883592348e-06,
+      "loss": 0.3576,
+      "step": 7249
+    },
+    {
+      "epoch": 1.823440643863179,
+      "grad_norm": 0.3146701753139496,
+      "learning_rate": 3.997923280872154e-06,
+      "loss": 0.3728,
+      "step": 7250
+    },
+    {
+      "epoch": 1.8236921529175052,
+      "grad_norm": 0.3109131455421448,
+      "learning_rate": 3.9964897639735644e-06,
+      "loss": 0.3297,
+      "step": 7251
+    },
+    {
+      "epoch": 1.823943661971831,
+      "grad_norm": 0.28482386469841003,
+      "learning_rate": 3.995056333019347e-06,
+      "loss": 0.33,
+      "step": 7252
+    },
+    {
+      "epoch": 1.8241951710261568,
+      "grad_norm": 0.3102104067802429,
+      "learning_rate": 3.993622988132269e-06,
+      "loss": 0.3738,
+      "step": 7253
+    },
+    {
+      "epoch": 1.824446680080483,
+      "grad_norm": 0.34564486145973206,
+      "learning_rate": 3.992189729435085e-06,
+      "loss": 0.3482,
+      "step": 7254
+    },
+    {
+      "epoch": 1.8246981891348089,
+      "grad_norm": 0.3167095482349396,
+      "learning_rate": 3.990756557050548e-06,
+      "loss": 0.3487,
+      "step": 7255
+    },
+    {
+      "epoch": 1.8249496981891347,
+      "grad_norm": 0.28097379207611084,
+      "learning_rate": 3.989323471101395e-06,
+      "loss": 0.3564,
+      "step": 7256
+    },
+    {
+      "epoch": 1.825201207243461,
+      "grad_norm": 0.29964274168014526,
+      "learning_rate": 3.987890471710367e-06,
+      "loss": 0.3434,
+      "step": 7257
+    },
+    {
+      "epoch": 1.8254527162977867,
+      "grad_norm": 0.32765746116638184,
+      "learning_rate": 3.986457559000185e-06,
+      "loss": 0.3327,
+      "step": 7258
+    },
+    {
+      "epoch": 1.8257042253521125,
+      "grad_norm": 0.3484762907028198,
+      "learning_rate": 3.985024733093576e-06,
+      "loss": 0.3504,
+      "step": 7259
+    },
+    {
+      "epoch": 1.8259557344064388,
+      "grad_norm": 0.3308793306350708,
+      "learning_rate": 3.9835919941132464e-06,
+      "loss": 0.3428,
+      "step": 7260
+    },
+    {
+      "epoch": 1.8262072434607646,
+      "grad_norm": 0.3149087429046631,
+      "learning_rate": 3.982159342181904e-06,
+      "loss": 0.3478,
+      "step": 7261
+    },
+    {
+      "epoch": 1.8264587525150904,
+      "grad_norm": 0.3215365707874298,
+      "learning_rate": 3.9807267774222475e-06,
+      "loss": 0.3417,
+      "step": 7262
+    },
+    {
+      "epoch": 1.8267102615694166,
+      "grad_norm": 0.3201581835746765,
+      "learning_rate": 3.979294299956965e-06,
+      "loss": 0.3392,
+      "step": 7263
+    },
+    {
+      "epoch": 1.8269617706237424,
+      "grad_norm": 0.3202977478504181,
+      "learning_rate": 3.977861909908741e-06,
+      "loss": 0.3425,
+      "step": 7264
+    },
+    {
+      "epoch": 1.8272132796780685,
+      "grad_norm": 0.31886807084083557,
+      "learning_rate": 3.976429607400249e-06,
+      "loss": 0.371,
+      "step": 7265
+    },
+    {
+      "epoch": 1.8274647887323945,
+      "grad_norm": 0.3250950872898102,
+      "learning_rate": 3.9749973925541585e-06,
+      "loss": 0.3473,
+      "step": 7266
+    },
+    {
+      "epoch": 1.8277162977867203,
+      "grad_norm": 0.30726975202560425,
+      "learning_rate": 3.973565265493129e-06,
+      "loss": 0.3644,
+      "step": 7267
+    },
+    {
+      "epoch": 1.8279678068410463,
+      "grad_norm": 0.3059992492198944,
+      "learning_rate": 3.972133226339812e-06,
+      "loss": 0.3607,
+      "step": 7268
+    },
+    {
+      "epoch": 1.8282193158953723,
+      "grad_norm": 0.32286500930786133,
+      "learning_rate": 3.970701275216855e-06,
+      "loss": 0.339,
+      "step": 7269
+    },
+    {
+      "epoch": 1.8284708249496981,
+      "grad_norm": 0.319062203168869,
+      "learning_rate": 3.969269412246895e-06,
+      "loss": 0.3551,
+      "step": 7270
+    },
+    {
+      "epoch": 1.8287223340040242,
+      "grad_norm": 0.3119877576828003,
+      "learning_rate": 3.967837637552561e-06,
+      "loss": 0.3506,
+      "step": 7271
+    },
+    {
+      "epoch": 1.8289738430583502,
+      "grad_norm": 0.32227200269699097,
+      "learning_rate": 3.966405951256475e-06,
+      "loss": 0.3581,
+      "step": 7272
+    },
+    {
+      "epoch": 1.829225352112676,
+      "grad_norm": 0.3127002716064453,
+      "learning_rate": 3.964974353481254e-06,
+      "loss": 0.3448,
+      "step": 7273
+    },
+    {
+      "epoch": 1.829476861167002,
+      "grad_norm": 0.3057059049606323,
+      "learning_rate": 3.963542844349505e-06,
+      "loss": 0.374,
+      "step": 7274
+    },
+    {
+      "epoch": 1.829728370221328,
+      "grad_norm": 0.29945117235183716,
+      "learning_rate": 3.962111423983827e-06,
+      "loss": 0.3443,
+      "step": 7275
+    },
+    {
+      "epoch": 1.8299798792756539,
+      "grad_norm": 0.30039167404174805,
+      "learning_rate": 3.960680092506812e-06,
+      "loss": 0.3194,
+      "step": 7276
+    },
+    {
+      "epoch": 1.83023138832998,
+      "grad_norm": 0.32396718859672546,
+      "learning_rate": 3.9592488500410465e-06,
+      "loss": 0.3649,
+      "step": 7277
+    },
+    {
+      "epoch": 1.830482897384306,
+      "grad_norm": 0.33545982837677,
+      "learning_rate": 3.957817696709104e-06,
+      "loss": 0.3568,
+      "step": 7278
+    },
+    {
+      "epoch": 1.8307344064386317,
+      "grad_norm": 0.2859671413898468,
+      "learning_rate": 3.9563866326335575e-06,
+      "loss": 0.3704,
+      "step": 7279
+    },
+    {
+      "epoch": 1.8309859154929577,
+      "grad_norm": 0.2837298512458801,
+      "learning_rate": 3.9549556579369665e-06,
+      "loss": 0.3228,
+      "step": 7280
+    },
+    {
+      "epoch": 1.8312374245472838,
+      "grad_norm": 0.2976028025150299,
+      "learning_rate": 3.953524772741886e-06,
+      "loss": 0.3342,
+      "step": 7281
+    },
+    {
+      "epoch": 1.8314889336016096,
+      "grad_norm": 0.30670884251594543,
+      "learning_rate": 3.952093977170861e-06,
+      "loss": 0.3578,
+      "step": 7282
+    },
+    {
+      "epoch": 1.8317404426559356,
+      "grad_norm": 0.2988281846046448,
+      "learning_rate": 3.950663271346432e-06,
+      "loss": 0.3489,
+      "step": 7283
+    },
+    {
+      "epoch": 1.8319919517102616,
+      "grad_norm": 0.276311993598938,
+      "learning_rate": 3.94923265539113e-06,
+      "loss": 0.3466,
+      "step": 7284
+    },
+    {
+      "epoch": 1.8322434607645874,
+      "grad_norm": 0.3069443106651306,
+      "learning_rate": 3.947802129427476e-06,
+      "loss": 0.3689,
+      "step": 7285
+    },
+    {
+      "epoch": 1.8324949698189135,
+      "grad_norm": 0.3089459240436554,
+      "learning_rate": 3.946371693577988e-06,
+      "loss": 0.3199,
+      "step": 7286
+    },
+    {
+      "epoch": 1.8327464788732395,
+      "grad_norm": 0.3473307192325592,
+      "learning_rate": 3.9449413479651715e-06,
+      "loss": 0.3832,
+      "step": 7287
+    },
+    {
+      "epoch": 1.8329979879275653,
+      "grad_norm": 0.2900719940662384,
+      "learning_rate": 3.94351109271153e-06,
+      "loss": 0.3357,
+      "step": 7288
+    },
+    {
+      "epoch": 1.8332494969818913,
+      "grad_norm": 0.29746437072753906,
+      "learning_rate": 3.9420809279395525e-06,
+      "loss": 0.3304,
+      "step": 7289
+    },
+    {
+      "epoch": 1.8335010060362174,
+      "grad_norm": 0.3016701340675354,
+      "learning_rate": 3.940650853771727e-06,
+      "loss": 0.371,
+      "step": 7290
+    },
+    {
+      "epoch": 1.8337525150905432,
+      "grad_norm": 0.2984379231929779,
+      "learning_rate": 3.939220870330527e-06,
+      "loss": 0.3555,
+      "step": 7291
+    },
+    {
+      "epoch": 1.8340040241448692,
+      "grad_norm": 0.3294828534126282,
+      "learning_rate": 3.937790977738425e-06,
+      "loss": 0.3452,
+      "step": 7292
+    },
+    {
+      "epoch": 1.8342555331991952,
+      "grad_norm": 0.3317073881626129,
+      "learning_rate": 3.936361176117879e-06,
+      "loss": 0.3459,
+      "step": 7293
+    },
+    {
+      "epoch": 1.834507042253521,
+      "grad_norm": 0.2974916994571686,
+      "learning_rate": 3.934931465591343e-06,
+      "loss": 0.3601,
+      "step": 7294
+    },
+    {
+      "epoch": 1.834758551307847,
+      "grad_norm": 0.30047857761383057,
+      "learning_rate": 3.9335018462812664e-06,
+      "loss": 0.3436,
+      "step": 7295
+    },
+    {
+      "epoch": 1.835010060362173,
+      "grad_norm": 0.3131597936153412,
+      "learning_rate": 3.9320723183100824e-06,
+      "loss": 0.3716,
+      "step": 7296
+    },
+    {
+      "epoch": 1.8352615694164989,
+      "grad_norm": 0.3238302171230316,
+      "learning_rate": 3.930642881800227e-06,
+      "loss": 0.3654,
+      "step": 7297
+    },
+    {
+      "epoch": 1.835513078470825,
+      "grad_norm": 0.31614577770233154,
+      "learning_rate": 3.9292135368741155e-06,
+      "loss": 0.3463,
+      "step": 7298
+    },
+    {
+      "epoch": 1.835764587525151,
+      "grad_norm": 0.31243595480918884,
+      "learning_rate": 3.927784283654168e-06,
+      "loss": 0.3459,
+      "step": 7299
+    },
+    {
+      "epoch": 1.8360160965794767,
+      "grad_norm": 0.2986874580383301,
+      "learning_rate": 3.926355122262787e-06,
+      "loss": 0.3263,
+      "step": 7300
+    },
+    {
+      "epoch": 1.836267605633803,
+      "grad_norm": 0.33278465270996094,
+      "learning_rate": 3.9249260528223745e-06,
+      "loss": 0.3431,
+      "step": 7301
+    },
+    {
+      "epoch": 1.8365191146881288,
+      "grad_norm": 0.3187868893146515,
+      "learning_rate": 3.923497075455319e-06,
+      "loss": 0.3582,
+      "step": 7302
+    },
+    {
+      "epoch": 1.8367706237424546,
+      "grad_norm": 0.33840519189834595,
+      "learning_rate": 3.922068190284005e-06,
+      "loss": 0.363,
+      "step": 7303
+    },
+    {
+      "epoch": 1.8370221327967808,
+      "grad_norm": 0.32831573486328125,
+      "learning_rate": 3.920639397430806e-06,
+      "loss": 0.3694,
+      "step": 7304
+    },
+    {
+      "epoch": 1.8372736418511066,
+      "grad_norm": 0.3055534362792969,
+      "learning_rate": 3.91921069701809e-06,
+      "loss": 0.3389,
+      "step": 7305
+    },
+    {
+      "epoch": 1.8375251509054324,
+      "grad_norm": 0.31282302737236023,
+      "learning_rate": 3.917782089168218e-06,
+      "loss": 0.3161,
+      "step": 7306
+    },
+    {
+      "epoch": 1.8377766599597587,
+      "grad_norm": 0.350875586271286,
+      "learning_rate": 3.916353574003538e-06,
+      "loss": 0.3565,
+      "step": 7307
+    },
+    {
+      "epoch": 1.8380281690140845,
+      "grad_norm": 0.3590453267097473,
+      "learning_rate": 3.914925151646397e-06,
+      "loss": 0.3605,
+      "step": 7308
+    },
+    {
+      "epoch": 1.8382796780684103,
+      "grad_norm": 0.3325250446796417,
+      "learning_rate": 3.913496822219127e-06,
+      "loss": 0.3876,
+      "step": 7309
+    },
+    {
+      "epoch": 1.8385311871227366,
+      "grad_norm": 0.2975671589374542,
+      "learning_rate": 3.912068585844059e-06,
+      "loss": 0.3617,
+      "step": 7310
+    },
+    {
+      "epoch": 1.8387826961770624,
+      "grad_norm": 0.30971285700798035,
+      "learning_rate": 3.910640442643508e-06,
+      "loss": 0.3263,
+      "step": 7311
+    },
+    {
+      "epoch": 1.8390342052313882,
+      "grad_norm": 0.34714826941490173,
+      "learning_rate": 3.909212392739791e-06,
+      "loss": 0.333,
+      "step": 7312
+    },
+    {
+      "epoch": 1.8392857142857144,
+      "grad_norm": 0.34102630615234375,
+      "learning_rate": 3.907784436255205e-06,
+      "loss": 0.3454,
+      "step": 7313
+    },
+    {
+      "epoch": 1.8395372233400402,
+      "grad_norm": 0.31585681438446045,
+      "learning_rate": 3.906356573312052e-06,
+      "loss": 0.3302,
+      "step": 7314
+    },
+    {
+      "epoch": 1.8397887323943662,
+      "grad_norm": 0.32975485920906067,
+      "learning_rate": 3.904928804032615e-06,
+      "loss": 0.3249,
+      "step": 7315
+    },
+    {
+      "epoch": 1.8400402414486923,
+      "grad_norm": 0.3263443112373352,
+      "learning_rate": 3.903501128539175e-06,
+      "loss": 0.331,
+      "step": 7316
+    },
+    {
+      "epoch": 1.840291750503018,
+      "grad_norm": 0.33371224999427795,
+      "learning_rate": 3.902073546954006e-06,
+      "loss": 0.3425,
+      "step": 7317
+    },
+    {
+      "epoch": 1.840543259557344,
+      "grad_norm": 0.35445427894592285,
+      "learning_rate": 3.900646059399367e-06,
+      "loss": 0.3474,
+      "step": 7318
+    },
+    {
+      "epoch": 1.8407947686116701,
+      "grad_norm": 0.3176020085811615,
+      "learning_rate": 3.899218665997517e-06,
+      "loss": 0.3538,
+      "step": 7319
+    },
+    {
+      "epoch": 1.841046277665996,
+      "grad_norm": 0.34081193804740906,
+      "learning_rate": 3.8977913668707e-06,
+      "loss": 0.3394,
+      "step": 7320
+    },
+    {
+      "epoch": 1.841297786720322,
+      "grad_norm": 0.35305309295654297,
+      "learning_rate": 3.896364162141159e-06,
+      "loss": 0.3447,
+      "step": 7321
+    },
+    {
+      "epoch": 1.841549295774648,
+      "grad_norm": 0.2991912364959717,
+      "learning_rate": 3.894937051931122e-06,
+      "loss": 0.3299,
+      "step": 7322
+    },
+    {
+      "epoch": 1.8418008048289738,
+      "grad_norm": 0.32608288526535034,
+      "learning_rate": 3.8935100363628135e-06,
+      "loss": 0.3316,
+      "step": 7323
+    },
+    {
+      "epoch": 1.8420523138832998,
+      "grad_norm": 0.3336857855319977,
+      "learning_rate": 3.892083115558447e-06,
+      "loss": 0.3469,
+      "step": 7324
+    },
+    {
+      "epoch": 1.8423038229376258,
+      "grad_norm": 0.3458016514778137,
+      "learning_rate": 3.890656289640233e-06,
+      "loss": 0.3666,
+      "step": 7325
+    },
+    {
+      "epoch": 1.8425553319919517,
+      "grad_norm": 0.3222261667251587,
+      "learning_rate": 3.889229558730365e-06,
+      "loss": 0.3593,
+      "step": 7326
+    },
+    {
+      "epoch": 1.8428068410462777,
+      "grad_norm": 0.31532028317451477,
+      "learning_rate": 3.887802922951038e-06,
+      "loss": 0.3456,
+      "step": 7327
+    },
+    {
+      "epoch": 1.8430583501006037,
+      "grad_norm": 0.306510865688324,
+      "learning_rate": 3.886376382424433e-06,
+      "loss": 0.3506,
+      "step": 7328
+    },
+    {
+      "epoch": 1.8433098591549295,
+      "grad_norm": 0.3103088438510895,
+      "learning_rate": 3.884949937272724e-06,
+      "loss": 0.3657,
+      "step": 7329
+    },
+    {
+      "epoch": 1.8435613682092555,
+      "grad_norm": 0.3240601718425751,
+      "learning_rate": 3.883523587618077e-06,
+      "loss": 0.34,
+      "step": 7330
+    },
+    {
+      "epoch": 1.8438128772635816,
+      "grad_norm": 0.30873456597328186,
+      "learning_rate": 3.8820973335826494e-06,
+      "loss": 0.3478,
+      "step": 7331
+    },
+    {
+      "epoch": 1.8440643863179074,
+      "grad_norm": 0.34828251600265503,
+      "learning_rate": 3.880671175288592e-06,
+      "loss": 0.3613,
+      "step": 7332
+    },
+    {
+      "epoch": 1.8443158953722334,
+      "grad_norm": 0.30683434009552,
+      "learning_rate": 3.879245112858046e-06,
+      "loss": 0.3606,
+      "step": 7333
+    },
+    {
+      "epoch": 1.8445674044265594,
+      "grad_norm": 0.2935948073863983,
+      "learning_rate": 3.877819146413144e-06,
+      "loss": 0.3395,
+      "step": 7334
+    },
+    {
+      "epoch": 1.8448189134808852,
+      "grad_norm": 0.2853759527206421,
+      "learning_rate": 3.876393276076013e-06,
+      "loss": 0.3525,
+      "step": 7335
+    },
+    {
+      "epoch": 1.8450704225352113,
+      "grad_norm": 0.31282341480255127,
+      "learning_rate": 3.8749675019687684e-06,
+      "loss": 0.3469,
+      "step": 7336
+    },
+    {
+      "epoch": 1.8453219315895373,
+      "grad_norm": 0.2779003083705902,
+      "learning_rate": 3.873541824213518e-06,
+      "loss": 0.3664,
+      "step": 7337
+    },
+    {
+      "epoch": 1.845573440643863,
+      "grad_norm": 0.3223724961280823,
+      "learning_rate": 3.872116242932363e-06,
+      "loss": 0.3689,
+      "step": 7338
+    },
+    {
+      "epoch": 1.845824949698189,
+      "grad_norm": 0.3017749786376953,
+      "learning_rate": 3.870690758247399e-06,
+      "loss": 0.3387,
+      "step": 7339
+    },
+    {
+      "epoch": 1.8460764587525151,
+      "grad_norm": 0.3117123544216156,
+      "learning_rate": 3.869265370280702e-06,
+      "loss": 0.318,
+      "step": 7340
+    },
+    {
+      "epoch": 1.846327967806841,
+      "grad_norm": 0.29091158509254456,
+      "learning_rate": 3.867840079154356e-06,
+      "loss": 0.3381,
+      "step": 7341
+    },
+    {
+      "epoch": 1.846579476861167,
+      "grad_norm": 0.3203681707382202,
+      "learning_rate": 3.866414884990422e-06,
+      "loss": 0.3435,
+      "step": 7342
+    },
+    {
+      "epoch": 1.846830985915493,
+      "grad_norm": 0.3184633255004883,
+      "learning_rate": 3.864989787910964e-06,
+      "loss": 0.3623,
+      "step": 7343
+    },
+    {
+      "epoch": 1.8470824949698188,
+      "grad_norm": 0.2918001413345337,
+      "learning_rate": 3.863564788038027e-06,
+      "loss": 0.37,
+      "step": 7344
+    },
+    {
+      "epoch": 1.8473340040241448,
+      "grad_norm": 0.3107215166091919,
+      "learning_rate": 3.862139885493659e-06,
+      "loss": 0.33,
+      "step": 7345
+    },
+    {
+      "epoch": 1.8475855130784709,
+      "grad_norm": 0.3049178719520569,
+      "learning_rate": 3.860715080399889e-06,
+      "loss": 0.3381,
+      "step": 7346
+    },
+    {
+      "epoch": 1.8478370221327967,
+      "grad_norm": 0.30147770047187805,
+      "learning_rate": 3.859290372878748e-06,
+      "loss": 0.3285,
+      "step": 7347
+    },
+    {
+      "epoch": 1.8480885311871227,
+      "grad_norm": 0.2963356375694275,
+      "learning_rate": 3.857865763052247e-06,
+      "loss": 0.3157,
+      "step": 7348
+    },
+    {
+      "epoch": 1.8483400402414487,
+      "grad_norm": 0.29773834347724915,
+      "learning_rate": 3.856441251042399e-06,
+      "loss": 0.3511,
+      "step": 7349
+    },
+    {
+      "epoch": 1.8485915492957745,
+      "grad_norm": 0.3260302245616913,
+      "learning_rate": 3.8550168369712055e-06,
+      "loss": 0.3778,
+      "step": 7350
+    },
+    {
+      "epoch": 1.8488430583501008,
+      "grad_norm": 0.30447903275489807,
+      "learning_rate": 3.8535925209606554e-06,
+      "loss": 0.3527,
+      "step": 7351
+    },
+    {
+      "epoch": 1.8490945674044266,
+      "grad_norm": 0.3211756944656372,
+      "learning_rate": 3.852168303132735e-06,
+      "loss": 0.3288,
+      "step": 7352
+    },
+    {
+      "epoch": 1.8493460764587524,
+      "grad_norm": 0.3038700819015503,
+      "learning_rate": 3.8507441836094175e-06,
+      "loss": 0.3574,
+      "step": 7353
+    },
+    {
+      "epoch": 1.8495975855130786,
+      "grad_norm": 0.3078562915325165,
+      "learning_rate": 3.849320162512672e-06,
+      "loss": 0.3248,
+      "step": 7354
+    },
+    {
+      "epoch": 1.8498490945674044,
+      "grad_norm": 0.31693726778030396,
+      "learning_rate": 3.847896239964455e-06,
+      "loss": 0.3387,
+      "step": 7355
+    },
+    {
+      "epoch": 1.8501006036217302,
+      "grad_norm": 0.32334980368614197,
+      "learning_rate": 3.8464724160867195e-06,
+      "loss": 0.3423,
+      "step": 7356
+    },
+    {
+      "epoch": 1.8503521126760565,
+      "grad_norm": 0.3063230812549591,
+      "learning_rate": 3.845048691001402e-06,
+      "loss": 0.3663,
+      "step": 7357
+    },
+    {
+      "epoch": 1.8506036217303823,
+      "grad_norm": 0.30338233709335327,
+      "learning_rate": 3.8436250648304415e-06,
+      "loss": 0.3548,
+      "step": 7358
+    },
+    {
+      "epoch": 1.850855130784708,
+      "grad_norm": 0.31649091839790344,
+      "learning_rate": 3.842201537695758e-06,
+      "loss": 0.3481,
+      "step": 7359
+    },
+    {
+      "epoch": 1.8511066398390343,
+      "grad_norm": 0.31132972240448,
+      "learning_rate": 3.84077810971927e-06,
+      "loss": 0.3467,
+      "step": 7360
+    },
+    {
+      "epoch": 1.8513581488933601,
+      "grad_norm": 0.3053782284259796,
+      "learning_rate": 3.839354781022886e-06,
+      "loss": 0.3211,
+      "step": 7361
+    },
+    {
+      "epoch": 1.8516096579476862,
+      "grad_norm": 0.314434677362442,
+      "learning_rate": 3.8379315517285025e-06,
+      "loss": 0.3406,
+      "step": 7362
+    },
+    {
+      "epoch": 1.8518611670020122,
+      "grad_norm": 0.29468590021133423,
+      "learning_rate": 3.836508421958014e-06,
+      "loss": 0.3442,
+      "step": 7363
+    },
+    {
+      "epoch": 1.852112676056338,
+      "grad_norm": 0.3091891407966614,
+      "learning_rate": 3.8350853918332974e-06,
+      "loss": 0.324,
+      "step": 7364
+    },
+    {
+      "epoch": 1.852364185110664,
+      "grad_norm": 0.30145037174224854,
+      "learning_rate": 3.833662461476233e-06,
+      "loss": 0.3225,
+      "step": 7365
+    },
+    {
+      "epoch": 1.85261569416499,
+      "grad_norm": 0.31025686860084534,
+      "learning_rate": 3.8322396310086785e-06,
+      "loss": 0.3593,
+      "step": 7366
+    },
+    {
+      "epoch": 1.8528672032193159,
+      "grad_norm": 0.32085564732551575,
+      "learning_rate": 3.8308169005524964e-06,
+      "loss": 0.3509,
+      "step": 7367
+    },
+    {
+      "epoch": 1.8531187122736419,
+      "grad_norm": 0.3120383322238922,
+      "learning_rate": 3.829394270229531e-06,
+      "loss": 0.3462,
+      "step": 7368
+    },
+    {
+      "epoch": 1.853370221327968,
+      "grad_norm": 0.3061147928237915,
+      "learning_rate": 3.827971740161625e-06,
+      "loss": 0.365,
+      "step": 7369
+    },
+    {
+      "epoch": 1.8536217303822937,
+      "grad_norm": 0.2813125550746918,
+      "learning_rate": 3.826549310470605e-06,
+      "loss": 0.3114,
+      "step": 7370
+    },
+    {
+      "epoch": 1.8538732394366197,
+      "grad_norm": 0.30677247047424316,
+      "learning_rate": 3.825126981278296e-06,
+      "loss": 0.368,
+      "step": 7371
+    },
+    {
+      "epoch": 1.8541247484909458,
+      "grad_norm": 0.3084120452404022,
+      "learning_rate": 3.823704752706512e-06,
+      "loss": 0.3603,
+      "step": 7372
+    },
+    {
+      "epoch": 1.8543762575452716,
+      "grad_norm": 0.29872360825538635,
+      "learning_rate": 3.8222826248770555e-06,
+      "loss": 0.3452,
+      "step": 7373
+    },
+    {
+      "epoch": 1.8546277665995976,
+      "grad_norm": 0.32304060459136963,
+      "learning_rate": 3.820860597911726e-06,
+      "loss": 0.3371,
+      "step": 7374
+    },
+    {
+      "epoch": 1.8548792756539236,
+      "grad_norm": 0.3058139681816101,
+      "learning_rate": 3.819438671932308e-06,
+      "loss": 0.3539,
+      "step": 7375
+    },
+    {
+      "epoch": 1.8551307847082494,
+      "grad_norm": 0.3089655041694641,
+      "learning_rate": 3.818016847060585e-06,
+      "loss": 0.3519,
+      "step": 7376
+    },
+    {
+      "epoch": 1.8553822937625755,
+      "grad_norm": 0.2986527681350708,
+      "learning_rate": 3.816595123418322e-06,
+      "loss": 0.3401,
+      "step": 7377
+    },
+    {
+      "epoch": 1.8556338028169015,
+      "grad_norm": 0.33130231499671936,
+      "learning_rate": 3.815173501127285e-06,
+      "loss": 0.3656,
+      "step": 7378
+    },
+    {
+      "epoch": 1.8558853118712273,
+      "grad_norm": 0.328773558139801,
+      "learning_rate": 3.813751980309224e-06,
+      "loss": 0.3466,
+      "step": 7379
+    },
+    {
+      "epoch": 1.8561368209255533,
+      "grad_norm": 0.32407474517822266,
+      "learning_rate": 3.8123305610858863e-06,
+      "loss": 0.35,
+      "step": 7380
+    },
+    {
+      "epoch": 1.8563883299798793,
+      "grad_norm": 0.3152269721031189,
+      "learning_rate": 3.810909243579004e-06,
+      "loss": 0.3328,
+      "step": 7381
+    },
+    {
+      "epoch": 1.8566398390342052,
+      "grad_norm": 0.32206395268440247,
+      "learning_rate": 3.8094880279103063e-06,
+      "loss": 0.353,
+      "step": 7382
+    },
+    {
+      "epoch": 1.8568913480885312,
+      "grad_norm": 0.3166285455226898,
+      "learning_rate": 3.808066914201513e-06,
+      "loss": 0.3548,
+      "step": 7383
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.3051060438156128,
+      "learning_rate": 3.80664590257433e-06,
+      "loss": 0.3394,
+      "step": 7384
+    },
+    {
+      "epoch": 1.857394366197183,
+      "grad_norm": 0.33410540223121643,
+      "learning_rate": 3.8052249931504614e-06,
+      "loss": 0.3551,
+      "step": 7385
+    },
+    {
+      "epoch": 1.857645875251509,
+      "grad_norm": 0.2926395535469055,
+      "learning_rate": 3.8038041860515962e-06,
+      "loss": 0.3339,
+      "step": 7386
+    },
+    {
+      "epoch": 1.857897384305835,
+      "grad_norm": 0.3076823949813843,
+      "learning_rate": 3.802383481399421e-06,
+      "loss": 0.3417,
+      "step": 7387
+    },
+    {
+      "epoch": 1.8581488933601609,
+      "grad_norm": 0.3127098083496094,
+      "learning_rate": 3.8009628793156056e-06,
+      "loss": 0.3316,
+      "step": 7388
+    },
+    {
+      "epoch": 1.858400402414487,
+      "grad_norm": 0.3178151845932007,
+      "learning_rate": 3.799542379921821e-06,
+      "loss": 0.3539,
+      "step": 7389
+    },
+    {
+      "epoch": 1.858651911468813,
+      "grad_norm": 0.3181746006011963,
+      "learning_rate": 3.79812198333972e-06,
+      "loss": 0.3584,
+      "step": 7390
+    },
+    {
+      "epoch": 1.8589034205231387,
+      "grad_norm": 0.33825597167015076,
+      "learning_rate": 3.7967016896909524e-06,
+      "loss": 0.3672,
+      "step": 7391
+    },
+    {
+      "epoch": 1.8591549295774648,
+      "grad_norm": 0.2978416085243225,
+      "learning_rate": 3.7952814990971575e-06,
+      "loss": 0.3398,
+      "step": 7392
+    },
+    {
+      "epoch": 1.8594064386317908,
+      "grad_norm": 0.3217383623123169,
+      "learning_rate": 3.7938614116799655e-06,
+      "loss": 0.3486,
+      "step": 7393
+    },
+    {
+      "epoch": 1.8596579476861166,
+      "grad_norm": 0.34177547693252563,
+      "learning_rate": 3.792441427560998e-06,
+      "loss": 0.3472,
+      "step": 7394
+    },
+    {
+      "epoch": 1.8599094567404426,
+      "grad_norm": 0.30881890654563904,
+      "learning_rate": 3.791021546861868e-06,
+      "loss": 0.3487,
+      "step": 7395
+    },
+    {
+      "epoch": 1.8601609657947686,
+      "grad_norm": 0.28196999430656433,
+      "learning_rate": 3.7896017697041788e-06,
+      "loss": 0.3386,
+      "step": 7396
+    },
+    {
+      "epoch": 1.8604124748490944,
+      "grad_norm": 0.2953253984451294,
+      "learning_rate": 3.788182096209526e-06,
+      "loss": 0.353,
+      "step": 7397
+    },
+    {
+      "epoch": 1.8606639839034205,
+      "grad_norm": 0.31603649258613586,
+      "learning_rate": 3.7867625264994954e-06,
+      "loss": 0.357,
+      "step": 7398
+    },
+    {
+      "epoch": 1.8609154929577465,
+      "grad_norm": 0.2854761481285095,
+      "learning_rate": 3.7853430606956647e-06,
+      "loss": 0.3246,
+      "step": 7399
+    },
+    {
+      "epoch": 1.8611670020120723,
+      "grad_norm": 0.3070826530456543,
+      "learning_rate": 3.783923698919602e-06,
+      "loss": 0.3417,
+      "step": 7400
+    },
+    {
+      "epoch": 1.8614185110663986,
+      "grad_norm": 0.3086259365081787,
+      "learning_rate": 3.782504441292867e-06,
+      "loss": 0.3459,
+      "step": 7401
+    },
+    {
+      "epoch": 1.8616700201207244,
+      "grad_norm": 0.30796676874160767,
+      "learning_rate": 3.7810852879370084e-06,
+      "loss": 0.3557,
+      "step": 7402
+    },
+    {
+      "epoch": 1.8619215291750502,
+      "grad_norm": 0.3138282001018524,
+      "learning_rate": 3.7796662389735718e-06,
+      "loss": 0.3609,
+      "step": 7403
+    },
+    {
+      "epoch": 1.8621730382293764,
+      "grad_norm": 0.29625728726387024,
+      "learning_rate": 3.778247294524085e-06,
+      "loss": 0.339,
+      "step": 7404
+    },
+    {
+      "epoch": 1.8624245472837022,
+      "grad_norm": 0.29120931029319763,
+      "learning_rate": 3.7768284547100763e-06,
+      "loss": 0.3245,
+      "step": 7405
+    },
+    {
+      "epoch": 1.862676056338028,
+      "grad_norm": 0.30197152495384216,
+      "learning_rate": 3.7754097196530566e-06,
+      "loss": 0.3421,
+      "step": 7406
+    },
+    {
+      "epoch": 1.8629275653923543,
+      "grad_norm": 0.2998182773590088,
+      "learning_rate": 3.7739910894745345e-06,
+      "loss": 0.3357,
+      "step": 7407
+    },
+    {
+      "epoch": 1.86317907444668,
+      "grad_norm": 0.33410176634788513,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.3829,
+      "step": 7408
+    },
+    {
+      "epoch": 1.8634305835010059,
+      "grad_norm": 0.31223413348197937,
+      "learning_rate": 3.771154144238958e-06,
+      "loss": 0.3493,
+      "step": 7409
+    },
+    {
+      "epoch": 1.8636820925553321,
+      "grad_norm": 0.32594525814056396,
+      "learning_rate": 3.7697358294248687e-06,
+      "loss": 0.3278,
+      "step": 7410
+    },
+    {
+      "epoch": 1.863933601609658,
+      "grad_norm": 0.3015677034854889,
+      "learning_rate": 3.7683176199752115e-06,
+      "loss": 0.3588,
+      "step": 7411
+    },
+    {
+      "epoch": 1.864185110663984,
+      "grad_norm": 0.33221563696861267,
+      "learning_rate": 3.7668995160114424e-06,
+      "loss": 0.368,
+      "step": 7412
+    },
+    {
+      "epoch": 1.86443661971831,
+      "grad_norm": 0.3092586100101471,
+      "learning_rate": 3.765481517655015e-06,
+      "loss": 0.3353,
+      "step": 7413
+    },
+    {
+      "epoch": 1.8646881287726358,
+      "grad_norm": 0.3521566092967987,
+      "learning_rate": 3.7640636250273754e-06,
+      "loss": 0.3502,
+      "step": 7414
+    },
+    {
+      "epoch": 1.8649396378269618,
+      "grad_norm": 0.33217549324035645,
+      "learning_rate": 3.7626458382499525e-06,
+      "loss": 0.3509,
+      "step": 7415
+    },
+    {
+      "epoch": 1.8651911468812878,
+      "grad_norm": 0.30001208186149597,
+      "learning_rate": 3.7612281574441744e-06,
+      "loss": 0.3576,
+      "step": 7416
+    },
+    {
+      "epoch": 1.8654426559356136,
+      "grad_norm": 0.32171934843063354,
+      "learning_rate": 3.7598105827314524e-06,
+      "loss": 0.3404,
+      "step": 7417
+    },
+    {
+      "epoch": 1.8656941649899397,
+      "grad_norm": 0.31113913655281067,
+      "learning_rate": 3.758393114233198e-06,
+      "loss": 0.3632,
+      "step": 7418
+    },
+    {
+      "epoch": 1.8659456740442657,
+      "grad_norm": 0.31468674540519714,
+      "learning_rate": 3.7569757520708034e-06,
+      "loss": 0.3457,
+      "step": 7419
+    },
+    {
+      "epoch": 1.8661971830985915,
+      "grad_norm": 0.29202166199684143,
+      "learning_rate": 3.7555584963656615e-06,
+      "loss": 0.3184,
+      "step": 7420
+    },
+    {
+      "epoch": 1.8664486921529175,
+      "grad_norm": 0.3300865888595581,
+      "learning_rate": 3.7541413472391474e-06,
+      "loss": 0.3406,
+      "step": 7421
+    },
+    {
+      "epoch": 1.8667002012072436,
+      "grad_norm": 0.32692527770996094,
+      "learning_rate": 3.752724304812635e-06,
+      "loss": 0.3436,
+      "step": 7422
+    },
+    {
+      "epoch": 1.8669517102615694,
+      "grad_norm": 0.3171963393688202,
+      "learning_rate": 3.7513073692074802e-06,
+      "loss": 0.3179,
+      "step": 7423
+    },
+    {
+      "epoch": 1.8672032193158954,
+      "grad_norm": 0.31962621212005615,
+      "learning_rate": 3.749890540545038e-06,
+      "loss": 0.3683,
+      "step": 7424
+    },
+    {
+      "epoch": 1.8674547283702214,
+      "grad_norm": 0.3360038995742798,
+      "learning_rate": 3.748473818946652e-06,
+      "loss": 0.3318,
+      "step": 7425
+    },
+    {
+      "epoch": 1.8677062374245472,
+      "grad_norm": 0.3103548586368561,
+      "learning_rate": 3.7470572045336518e-06,
+      "loss": 0.3393,
+      "step": 7426
+    },
+    {
+      "epoch": 1.8679577464788732,
+      "grad_norm": 0.3288777470588684,
+      "learning_rate": 3.745640697427366e-06,
+      "loss": 0.3435,
+      "step": 7427
+    },
+    {
+      "epoch": 1.8682092555331993,
+      "grad_norm": 0.3266128897666931,
+      "learning_rate": 3.744224297749105e-06,
+      "loss": 0.328,
+      "step": 7428
+    },
+    {
+      "epoch": 1.868460764587525,
+      "grad_norm": 0.2997075617313385,
+      "learning_rate": 3.7428080056201794e-06,
+      "loss": 0.3406,
+      "step": 7429
+    },
+    {
+      "epoch": 1.868712273641851,
+      "grad_norm": 0.3005223870277405,
+      "learning_rate": 3.7413918211618804e-06,
+      "loss": 0.3256,
+      "step": 7430
+    },
+    {
+      "epoch": 1.8689637826961771,
+      "grad_norm": 0.3092925548553467,
+      "learning_rate": 3.739975744495501e-06,
+      "loss": 0.3591,
+      "step": 7431
+    },
+    {
+      "epoch": 1.869215291750503,
+      "grad_norm": 0.31816989183425903,
+      "learning_rate": 3.738559775742313e-06,
+      "loss": 0.3276,
+      "step": 7432
+    },
+    {
+      "epoch": 1.869466800804829,
+      "grad_norm": 0.29660165309906006,
+      "learning_rate": 3.7371439150235923e-06,
+      "loss": 0.3396,
+      "step": 7433
+    },
+    {
+      "epoch": 1.869718309859155,
+      "grad_norm": 0.3243357241153717,
+      "learning_rate": 3.735728162460591e-06,
+      "loss": 0.3516,
+      "step": 7434
+    },
+    {
+      "epoch": 1.8699698189134808,
+      "grad_norm": 0.29959172010421753,
+      "learning_rate": 3.734312518174565e-06,
+      "loss": 0.3762,
+      "step": 7435
+    },
+    {
+      "epoch": 1.8702213279678068,
+      "grad_norm": 0.29366421699523926,
+      "learning_rate": 3.732896982286755e-06,
+      "loss": 0.3441,
+      "step": 7436
+    },
+    {
+      "epoch": 1.8704728370221329,
+      "grad_norm": 0.29330679774284363,
+      "learning_rate": 3.731481554918389e-06,
+      "loss": 0.3299,
+      "step": 7437
+    },
+    {
+      "epoch": 1.8707243460764587,
+      "grad_norm": 0.3113592565059662,
+      "learning_rate": 3.7300662361906946e-06,
+      "loss": 0.3302,
+      "step": 7438
+    },
+    {
+      "epoch": 1.8709758551307847,
+      "grad_norm": 0.31629472970962524,
+      "learning_rate": 3.728651026224881e-06,
+      "loss": 0.3492,
+      "step": 7439
+    },
+    {
+      "epoch": 1.8712273641851107,
+      "grad_norm": 0.3160915672779083,
+      "learning_rate": 3.727235925142154e-06,
+      "loss": 0.3628,
+      "step": 7440
+    },
+    {
+      "epoch": 1.8714788732394365,
+      "grad_norm": 0.2999648153781891,
+      "learning_rate": 3.725820933063707e-06,
+      "loss": 0.342,
+      "step": 7441
+    },
+    {
+      "epoch": 1.8717303822937625,
+      "grad_norm": 0.2990424931049347,
+      "learning_rate": 3.724406050110727e-06,
+      "loss": 0.3564,
+      "step": 7442
+    },
+    {
+      "epoch": 1.8719818913480886,
+      "grad_norm": 0.29632776975631714,
+      "learning_rate": 3.7229912764043874e-06,
+      "loss": 0.3448,
+      "step": 7443
+    },
+    {
+      "epoch": 1.8722334004024144,
+      "grad_norm": 0.33509159088134766,
+      "learning_rate": 3.7215766120658568e-06,
+      "loss": 0.3618,
+      "step": 7444
+    },
+    {
+      "epoch": 1.8724849094567404,
+      "grad_norm": 0.328966349363327,
+      "learning_rate": 3.720162057216291e-06,
+      "loss": 0.3296,
+      "step": 7445
+    },
+    {
+      "epoch": 1.8727364185110664,
+      "grad_norm": 0.293011337518692,
+      "learning_rate": 3.7187476119768383e-06,
+      "loss": 0.3452,
+      "step": 7446
+    },
+    {
+      "epoch": 1.8729879275653922,
+      "grad_norm": 0.3068729043006897,
+      "learning_rate": 3.7173332764686375e-06,
+      "loss": 0.3468,
+      "step": 7447
+    },
+    {
+      "epoch": 1.8732394366197183,
+      "grad_norm": 0.3060571849346161,
+      "learning_rate": 3.715919050812817e-06,
+      "loss": 0.3509,
+      "step": 7448
+    },
+    {
+      "epoch": 1.8734909456740443,
+      "grad_norm": 0.31499195098876953,
+      "learning_rate": 3.7145049351304973e-06,
+      "loss": 0.3422,
+      "step": 7449
+    },
+    {
+      "epoch": 1.87374245472837,
+      "grad_norm": 0.2951659858226776,
+      "learning_rate": 3.7130909295427873e-06,
+      "loss": 0.331,
+      "step": 7450
+    },
+    {
+      "epoch": 1.8739939637826963,
+      "grad_norm": 0.30359193682670593,
+      "learning_rate": 3.7116770341707893e-06,
+      "loss": 0.3479,
+      "step": 7451
+    },
+    {
+      "epoch": 1.8742454728370221,
+      "grad_norm": 0.28634950518608093,
+      "learning_rate": 3.710263249135593e-06,
+      "loss": 0.3286,
+      "step": 7452
+    },
+    {
+      "epoch": 1.874496981891348,
+      "grad_norm": 0.30430909991264343,
+      "learning_rate": 3.7088495745582803e-06,
+      "loss": 0.3244,
+      "step": 7453
+    },
+    {
+      "epoch": 1.8747484909456742,
+      "grad_norm": 0.3412688970565796,
+      "learning_rate": 3.7074360105599246e-06,
+      "loss": 0.3657,
+      "step": 7454
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.3176121413707733,
+      "learning_rate": 3.706022557261588e-06,
+      "loss": 0.3633,
+      "step": 7455
+    },
+    {
+      "epoch": 1.8752515090543258,
+      "grad_norm": 0.2989017069339752,
+      "learning_rate": 3.704609214784325e-06,
+      "loss": 0.3417,
+      "step": 7456
+    },
+    {
+      "epoch": 1.875503018108652,
+      "grad_norm": 0.31169548630714417,
+      "learning_rate": 3.703195983249177e-06,
+      "loss": 0.3268,
+      "step": 7457
+    },
+    {
+      "epoch": 1.8757545271629779,
+      "grad_norm": 0.306718647480011,
+      "learning_rate": 3.7017828627771825e-06,
+      "loss": 0.3479,
+      "step": 7458
+    },
+    {
+      "epoch": 1.8760060362173037,
+      "grad_norm": 0.29229748249053955,
+      "learning_rate": 3.7003698534893623e-06,
+      "loss": 0.3337,
+      "step": 7459
+    },
+    {
+      "epoch": 1.87625754527163,
+      "grad_norm": 0.3112187683582306,
+      "learning_rate": 3.6989569555067357e-06,
+      "loss": 0.3334,
+      "step": 7460
+    },
+    {
+      "epoch": 1.8765090543259557,
+      "grad_norm": 0.2908962368965149,
+      "learning_rate": 3.6975441689503044e-06,
+      "loss": 0.3319,
+      "step": 7461
+    },
+    {
+      "epoch": 1.8767605633802817,
+      "grad_norm": 0.2986694872379303,
+      "learning_rate": 3.6961314939410674e-06,
+      "loss": 0.3649,
+      "step": 7462
+    },
+    {
+      "epoch": 1.8770120724346078,
+      "grad_norm": 0.2823163568973541,
+      "learning_rate": 3.694718930600012e-06,
+      "loss": 0.3357,
+      "step": 7463
+    },
+    {
+      "epoch": 1.8772635814889336,
+      "grad_norm": 0.28905975818634033,
+      "learning_rate": 3.693306479048114e-06,
+      "loss": 0.3284,
+      "step": 7464
+    },
+    {
+      "epoch": 1.8775150905432596,
+      "grad_norm": 0.3094346821308136,
+      "learning_rate": 3.6918941394063414e-06,
+      "loss": 0.3834,
+      "step": 7465
+    },
+    {
+      "epoch": 1.8777665995975856,
+      "grad_norm": 0.30385082960128784,
+      "learning_rate": 3.6904819117956526e-06,
+      "loss": 0.3556,
+      "step": 7466
+    },
+    {
+      "epoch": 1.8780181086519114,
+      "grad_norm": 0.31416597962379456,
+      "learning_rate": 3.6890697963369947e-06,
+      "loss": 0.3567,
+      "step": 7467
+    },
+    {
+      "epoch": 1.8782696177062375,
+      "grad_norm": 0.2962150275707245,
+      "learning_rate": 3.6876577931513076e-06,
+      "loss": 0.3439,
+      "step": 7468
+    },
+    {
+      "epoch": 1.8785211267605635,
+      "grad_norm": 0.30000972747802734,
+      "learning_rate": 3.686245902359522e-06,
+      "loss": 0.3308,
+      "step": 7469
+    },
+    {
+      "epoch": 1.8787726358148893,
+      "grad_norm": 0.31348156929016113,
+      "learning_rate": 3.6848341240825548e-06,
+      "loss": 0.3402,
+      "step": 7470
+    },
+    {
+      "epoch": 1.8790241448692153,
+      "grad_norm": 0.34603050351142883,
+      "learning_rate": 3.6834224584413183e-06,
+      "loss": 0.3684,
+      "step": 7471
+    },
+    {
+      "epoch": 1.8792756539235413,
+      "grad_norm": 0.3139626681804657,
+      "learning_rate": 3.68201090555671e-06,
+      "loss": 0.3524,
+      "step": 7472
+    },
+    {
+      "epoch": 1.8795271629778671,
+      "grad_norm": 0.3043857216835022,
+      "learning_rate": 3.6805994655496245e-06,
+      "loss": 0.3485,
+      "step": 7473
+    },
+    {
+      "epoch": 1.8797786720321932,
+      "grad_norm": 0.33150768280029297,
+      "learning_rate": 3.6791881385409383e-06,
+      "loss": 0.3492,
+      "step": 7474
+    },
+    {
+      "epoch": 1.8800301810865192,
+      "grad_norm": 0.30486994981765747,
+      "learning_rate": 3.6777769246515275e-06,
+      "loss": 0.3585,
+      "step": 7475
+    },
+    {
+      "epoch": 1.880281690140845,
+      "grad_norm": 0.2925189733505249,
+      "learning_rate": 3.6763658240022495e-06,
+      "loss": 0.31,
+      "step": 7476
+    },
+    {
+      "epoch": 1.880533199195171,
+      "grad_norm": 0.2866535484790802,
+      "learning_rate": 3.67495483671396e-06,
+      "loss": 0.3426,
+      "step": 7477
+    },
+    {
+      "epoch": 1.880784708249497,
+      "grad_norm": 0.32085758447647095,
+      "learning_rate": 3.6735439629074964e-06,
+      "loss": 0.3441,
+      "step": 7478
+    },
+    {
+      "epoch": 1.8810362173038229,
+      "grad_norm": 0.3063550293445587,
+      "learning_rate": 3.672133202703694e-06,
+      "loss": 0.3211,
+      "step": 7479
+    },
+    {
+      "epoch": 1.881287726358149,
+      "grad_norm": 0.32170259952545166,
+      "learning_rate": 3.670722556223379e-06,
+      "loss": 0.3347,
+      "step": 7480
+    },
+    {
+      "epoch": 1.881539235412475,
+      "grad_norm": 0.30713140964508057,
+      "learning_rate": 3.669312023587358e-06,
+      "loss": 0.3515,
+      "step": 7481
+    },
+    {
+      "epoch": 1.8817907444668007,
+      "grad_norm": 0.3304310142993927,
+      "learning_rate": 3.66790160491644e-06,
+      "loss": 0.3393,
+      "step": 7482
+    },
+    {
+      "epoch": 1.8820422535211268,
+      "grad_norm": 0.3182964026927948,
+      "learning_rate": 3.666491300331414e-06,
+      "loss": 0.3511,
+      "step": 7483
+    },
+    {
+      "epoch": 1.8822937625754528,
+      "grad_norm": 0.2859238088130951,
+      "learning_rate": 3.6650811099530673e-06,
+      "loss": 0.3363,
+      "step": 7484
+    },
+    {
+      "epoch": 1.8825452716297786,
+      "grad_norm": 0.3283844292163849,
+      "learning_rate": 3.663671033902171e-06,
+      "loss": 0.3292,
+      "step": 7485
+    },
+    {
+      "epoch": 1.8827967806841046,
+      "grad_norm": 0.31571048498153687,
+      "learning_rate": 3.662261072299492e-06,
+      "loss": 0.359,
+      "step": 7486
+    },
+    {
+      "epoch": 1.8830482897384306,
+      "grad_norm": 0.32221272587776184,
+      "learning_rate": 3.660851225265781e-06,
+      "loss": 0.3513,
+      "step": 7487
+    },
+    {
+      "epoch": 1.8832997987927564,
+      "grad_norm": 0.29501035809516907,
+      "learning_rate": 3.659441492921788e-06,
+      "loss": 0.3411,
+      "step": 7488
+    },
+    {
+      "epoch": 1.8835513078470825,
+      "grad_norm": 0.31255054473876953,
+      "learning_rate": 3.6580318753882414e-06,
+      "loss": 0.3369,
+      "step": 7489
+    },
+    {
+      "epoch": 1.8838028169014085,
+      "grad_norm": 0.3012523055076599,
+      "learning_rate": 3.6566223727858697e-06,
+      "loss": 0.333,
+      "step": 7490
+    },
+    {
+      "epoch": 1.8840543259557343,
+      "grad_norm": 0.36300086975097656,
+      "learning_rate": 3.65521298523539e-06,
+      "loss": 0.3814,
+      "step": 7491
+    },
+    {
+      "epoch": 1.8843058350100603,
+      "grad_norm": 0.3129633069038391,
+      "learning_rate": 3.653803712857503e-06,
+      "loss": 0.3338,
+      "step": 7492
+    },
+    {
+      "epoch": 1.8845573440643864,
+      "grad_norm": 0.29744789004325867,
+      "learning_rate": 3.652394555772908e-06,
+      "loss": 0.3582,
+      "step": 7493
+    },
+    {
+      "epoch": 1.8848088531187122,
+      "grad_norm": 0.30003800988197327,
+      "learning_rate": 3.650985514102287e-06,
+      "loss": 0.3399,
+      "step": 7494
+    },
+    {
+      "epoch": 1.8850603621730382,
+      "grad_norm": 0.2842726707458496,
+      "learning_rate": 3.6495765879663194e-06,
+      "loss": 0.3373,
+      "step": 7495
+    },
+    {
+      "epoch": 1.8853118712273642,
+      "grad_norm": 0.2723131775856018,
+      "learning_rate": 3.6481677774856666e-06,
+      "loss": 0.3369,
+      "step": 7496
+    },
+    {
+      "epoch": 1.88556338028169,
+      "grad_norm": 0.3011725842952728,
+      "learning_rate": 3.6467590827809885e-06,
+      "loss": 0.3421,
+      "step": 7497
+    },
+    {
+      "epoch": 1.885814889336016,
+      "grad_norm": 0.31559064984321594,
+      "learning_rate": 3.6453505039729274e-06,
+      "loss": 0.3782,
+      "step": 7498
+    },
+    {
+      "epoch": 1.886066398390342,
+      "grad_norm": 0.29440516233444214,
+      "learning_rate": 3.6439420411821226e-06,
+      "loss": 0.3317,
+      "step": 7499
+    },
+    {
+      "epoch": 1.8863179074446679,
+      "grad_norm": 0.2825719714164734,
+      "learning_rate": 3.642533694529197e-06,
+      "loss": 0.3552,
+      "step": 7500
+    },
+    {
+      "epoch": 1.8865694164989941,
+      "grad_norm": 0.3134790360927582,
+      "learning_rate": 3.641125464134768e-06,
+      "loss": 0.3633,
+      "step": 7501
+    },
+    {
+      "epoch": 1.88682092555332,
+      "grad_norm": 0.33232244849205017,
+      "learning_rate": 3.6397173501194436e-06,
+      "loss": 0.3448,
+      "step": 7502
+    },
+    {
+      "epoch": 1.8870724346076457,
+      "grad_norm": 0.31418442726135254,
+      "learning_rate": 3.638309352603816e-06,
+      "loss": 0.3756,
+      "step": 7503
+    },
+    {
+      "epoch": 1.887323943661972,
+      "grad_norm": 0.31419721245765686,
+      "learning_rate": 3.6369014717084768e-06,
+      "loss": 0.3438,
+      "step": 7504
+    },
+    {
+      "epoch": 1.8875754527162978,
+      "grad_norm": 0.2999892234802246,
+      "learning_rate": 3.635493707553996e-06,
+      "loss": 0.3494,
+      "step": 7505
+    },
+    {
+      "epoch": 1.8878269617706236,
+      "grad_norm": 0.2898130714893341,
+      "learning_rate": 3.634086060260945e-06,
+      "loss": 0.3625,
+      "step": 7506
+    },
+    {
+      "epoch": 1.8880784708249498,
+      "grad_norm": 0.30754363536834717,
+      "learning_rate": 3.6326785299498758e-06,
+      "loss": 0.3224,
+      "step": 7507
+    },
+    {
+      "epoch": 1.8883299798792756,
+      "grad_norm": 0.3098406493663788,
+      "learning_rate": 3.6312711167413394e-06,
+      "loss": 0.3585,
+      "step": 7508
+    },
+    {
+      "epoch": 1.8885814889336014,
+      "grad_norm": 0.29480767250061035,
+      "learning_rate": 3.629863820755866e-06,
+      "loss": 0.3193,
+      "step": 7509
+    },
+    {
+      "epoch": 1.8888329979879277,
+      "grad_norm": 0.3022323548793793,
+      "learning_rate": 3.628456642113988e-06,
+      "loss": 0.3516,
+      "step": 7510
+    },
+    {
+      "epoch": 1.8890845070422535,
+      "grad_norm": 0.2908898591995239,
+      "learning_rate": 3.627049580936215e-06,
+      "loss": 0.3337,
+      "step": 7511
+    },
+    {
+      "epoch": 1.8893360160965795,
+      "grad_norm": 0.3488399088382721,
+      "learning_rate": 3.6256426373430577e-06,
+      "loss": 0.3307,
+      "step": 7512
+    },
+    {
+      "epoch": 1.8895875251509056,
+      "grad_norm": 0.33839771151542664,
+      "learning_rate": 3.6242358114550104e-06,
+      "loss": 0.3452,
+      "step": 7513
+    },
+    {
+      "epoch": 1.8898390342052314,
+      "grad_norm": 0.31028321385383606,
+      "learning_rate": 3.6228291033925596e-06,
+      "loss": 0.3599,
+      "step": 7514
+    },
+    {
+      "epoch": 1.8900905432595574,
+      "grad_norm": 0.3041622042655945,
+      "learning_rate": 3.6214225132761806e-06,
+      "loss": 0.3322,
+      "step": 7515
+    },
+    {
+      "epoch": 1.8903420523138834,
+      "grad_norm": 0.2997733950614929,
+      "learning_rate": 3.620016041226338e-06,
+      "loss": 0.3191,
+      "step": 7516
+    },
+    {
+      "epoch": 1.8905935613682092,
+      "grad_norm": 0.3369154632091522,
+      "learning_rate": 3.618609687363489e-06,
+      "loss": 0.35,
+      "step": 7517
+    },
+    {
+      "epoch": 1.8908450704225352,
+      "grad_norm": 0.3129628896713257,
+      "learning_rate": 3.6172034518080785e-06,
+      "loss": 0.3106,
+      "step": 7518
+    },
+    {
+      "epoch": 1.8910965794768613,
+      "grad_norm": 0.2884010672569275,
+      "learning_rate": 3.615797334680541e-06,
+      "loss": 0.3325,
+      "step": 7519
+    },
+    {
+      "epoch": 1.891348088531187,
+      "grad_norm": 0.2941190004348755,
+      "learning_rate": 3.6143913361013026e-06,
+      "loss": 0.3346,
+      "step": 7520
+    },
+    {
+      "epoch": 1.891599597585513,
+      "grad_norm": 0.314097136259079,
+      "learning_rate": 3.6129854561907786e-06,
+      "loss": 0.3595,
+      "step": 7521
+    },
+    {
+      "epoch": 1.8918511066398391,
+      "grad_norm": 0.30389559268951416,
+      "learning_rate": 3.611579695069372e-06,
+      "loss": 0.3565,
+      "step": 7522
+    },
+    {
+      "epoch": 1.892102615694165,
+      "grad_norm": 0.3276069760322571,
+      "learning_rate": 3.610174052857478e-06,
+      "loss": 0.3509,
+      "step": 7523
+    },
+    {
+      "epoch": 1.892354124748491,
+      "grad_norm": 0.3453889787197113,
+      "learning_rate": 3.608768529675484e-06,
+      "loss": 0.3276,
+      "step": 7524
+    },
+    {
+      "epoch": 1.892605633802817,
+      "grad_norm": 0.28621137142181396,
+      "learning_rate": 3.607363125643759e-06,
+      "loss": 0.3435,
+      "step": 7525
+    },
+    {
+      "epoch": 1.8928571428571428,
+      "grad_norm": 0.33448565006256104,
+      "learning_rate": 3.6059578408826734e-06,
+      "loss": 0.346,
+      "step": 7526
+    },
+    {
+      "epoch": 1.8931086519114688,
+      "grad_norm": 0.3352438509464264,
+      "learning_rate": 3.604552675512574e-06,
+      "loss": 0.3487,
+      "step": 7527
+    },
+    {
+      "epoch": 1.8933601609657948,
+      "grad_norm": 0.3266422152519226,
+      "learning_rate": 3.6031476296538113e-06,
+      "loss": 0.341,
+      "step": 7528
+    },
+    {
+      "epoch": 1.8936116700201207,
+      "grad_norm": 0.2940256595611572,
+      "learning_rate": 3.601742703426713e-06,
+      "loss": 0.3504,
+      "step": 7529
+    },
+    {
+      "epoch": 1.8938631790744467,
+      "grad_norm": 0.34617501497268677,
+      "learning_rate": 3.6003378969516067e-06,
+      "loss": 0.3384,
+      "step": 7530
+    },
+    {
+      "epoch": 1.8941146881287727,
+      "grad_norm": 0.33195579051971436,
+      "learning_rate": 3.5989332103488013e-06,
+      "loss": 0.3491,
+      "step": 7531
+    },
+    {
+      "epoch": 1.8943661971830985,
+      "grad_norm": 0.3452162742614746,
+      "learning_rate": 3.5975286437386014e-06,
+      "loss": 0.3862,
+      "step": 7532
+    },
+    {
+      "epoch": 1.8946177062374245,
+      "grad_norm": 0.2940990626811981,
+      "learning_rate": 3.5961241972413012e-06,
+      "loss": 0.3175,
+      "step": 7533
+    },
+    {
+      "epoch": 1.8948692152917506,
+      "grad_norm": 0.30977126955986023,
+      "learning_rate": 3.594719870977179e-06,
+      "loss": 0.3298,
+      "step": 7534
+    },
+    {
+      "epoch": 1.8951207243460764,
+      "grad_norm": 0.2974280118942261,
+      "learning_rate": 3.5933156650665102e-06,
+      "loss": 0.3208,
+      "step": 7535
+    },
+    {
+      "epoch": 1.8953722334004024,
+      "grad_norm": 0.323830246925354,
+      "learning_rate": 3.591911579629553e-06,
+      "loss": 0.363,
+      "step": 7536
+    },
+    {
+      "epoch": 1.8956237424547284,
+      "grad_norm": 0.308124840259552,
+      "learning_rate": 3.590507614786561e-06,
+      "loss": 0.3466,
+      "step": 7537
+    },
+    {
+      "epoch": 1.8958752515090542,
+      "grad_norm": 0.2933880090713501,
+      "learning_rate": 3.5891037706577736e-06,
+      "loss": 0.3256,
+      "step": 7538
+    },
+    {
+      "epoch": 1.8961267605633803,
+      "grad_norm": 0.30009788274765015,
+      "learning_rate": 3.5877000473634227e-06,
+      "loss": 0.3625,
+      "step": 7539
+    },
+    {
+      "epoch": 1.8963782696177063,
+      "grad_norm": 0.3329128324985504,
+      "learning_rate": 3.586296445023726e-06,
+      "loss": 0.3725,
+      "step": 7540
+    },
+    {
+      "epoch": 1.896629778672032,
+      "grad_norm": 0.2923327088356018,
+      "learning_rate": 3.584892963758896e-06,
+      "loss": 0.3407,
+      "step": 7541
+    },
+    {
+      "epoch": 1.8968812877263581,
+      "grad_norm": 0.336452454328537,
+      "learning_rate": 3.583489603689129e-06,
+      "loss": 0.3387,
+      "step": 7542
+    },
+    {
+      "epoch": 1.8971327967806841,
+      "grad_norm": 0.3008471131324768,
+      "learning_rate": 3.5820863649346162e-06,
+      "loss": 0.3415,
+      "step": 7543
+    },
+    {
+      "epoch": 1.89738430583501,
+      "grad_norm": 0.3129260241985321,
+      "learning_rate": 3.5806832476155373e-06,
+      "loss": 0.365,
+      "step": 7544
+    },
+    {
+      "epoch": 1.897635814889336,
+      "grad_norm": 0.3290889263153076,
+      "learning_rate": 3.579280251852057e-06,
+      "loss": 0.3611,
+      "step": 7545
+    },
+    {
+      "epoch": 1.897887323943662,
+      "grad_norm": 0.2925894856452942,
+      "learning_rate": 3.577877377764337e-06,
+      "loss": 0.3765,
+      "step": 7546
+    },
+    {
+      "epoch": 1.8981388329979878,
+      "grad_norm": 0.3327069878578186,
+      "learning_rate": 3.5764746254725213e-06,
+      "loss": 0.3524,
+      "step": 7547
+    },
+    {
+      "epoch": 1.8983903420523138,
+      "grad_norm": 0.3372514545917511,
+      "learning_rate": 3.5750719950967507e-06,
+      "loss": 0.365,
+      "step": 7548
+    },
+    {
+      "epoch": 1.8986418511066399,
+      "grad_norm": 0.2978571057319641,
+      "learning_rate": 3.5736694867571465e-06,
+      "loss": 0.3321,
+      "step": 7549
+    },
+    {
+      "epoch": 1.8988933601609657,
+      "grad_norm": 0.33006635308265686,
+      "learning_rate": 3.5722671005738303e-06,
+      "loss": 0.3283,
+      "step": 7550
+    },
+    {
+      "epoch": 1.899144869215292,
+      "grad_norm": 0.32429298758506775,
+      "learning_rate": 3.570864836666903e-06,
+      "loss": 0.3622,
+      "step": 7551
+    },
+    {
+      "epoch": 1.8993963782696177,
+      "grad_norm": 0.3145343065261841,
+      "learning_rate": 3.5694626951564637e-06,
+      "loss": 0.3564,
+      "step": 7552
+    },
+    {
+      "epoch": 1.8996478873239435,
+      "grad_norm": 0.30636870861053467,
+      "learning_rate": 3.5680606761625925e-06,
+      "loss": 0.3545,
+      "step": 7553
+    },
+    {
+      "epoch": 1.8998993963782698,
+      "grad_norm": 0.27128949761390686,
+      "learning_rate": 3.566658779805367e-06,
+      "loss": 0.351,
+      "step": 7554
+    },
+    {
+      "epoch": 1.9001509054325956,
+      "grad_norm": 0.318464070558548,
+      "learning_rate": 3.565257006204852e-06,
+      "loss": 0.3615,
+      "step": 7555
+    },
+    {
+      "epoch": 1.9004024144869214,
+      "grad_norm": 0.3404373824596405,
+      "learning_rate": 3.5638553554810963e-06,
+      "loss": 0.366,
+      "step": 7556
+    },
+    {
+      "epoch": 1.9006539235412476,
+      "grad_norm": 0.29662925004959106,
+      "learning_rate": 3.5624538277541474e-06,
+      "loss": 0.3436,
+      "step": 7557
+    },
+    {
+      "epoch": 1.9009054325955734,
+      "grad_norm": 0.2770460546016693,
+      "learning_rate": 3.5610524231440324e-06,
+      "loss": 0.3663,
+      "step": 7558
+    },
+    {
+      "epoch": 1.9011569416498992,
+      "grad_norm": 0.3167244493961334,
+      "learning_rate": 3.559651141770778e-06,
+      "loss": 0.3376,
+      "step": 7559
+    },
+    {
+      "epoch": 1.9014084507042255,
+      "grad_norm": 0.30453893542289734,
+      "learning_rate": 3.5582499837543894e-06,
+      "loss": 0.3545,
+      "step": 7560
+    },
+    {
+      "epoch": 1.9016599597585513,
+      "grad_norm": 0.3127380907535553,
+      "learning_rate": 3.5568489492148728e-06,
+      "loss": 0.3502,
+      "step": 7561
+    },
+    {
+      "epoch": 1.9019114688128773,
+      "grad_norm": 0.3051506280899048,
+      "learning_rate": 3.5554480382722134e-06,
+      "loss": 0.3508,
+      "step": 7562
+    },
+    {
+      "epoch": 1.9021629778672033,
+      "grad_norm": 0.28955167531967163,
+      "learning_rate": 3.5540472510463947e-06,
+      "loss": 0.3445,
+      "step": 7563
+    },
+    {
+      "epoch": 1.9024144869215291,
+      "grad_norm": 0.30066999793052673,
+      "learning_rate": 3.552646587657381e-06,
+      "loss": 0.3658,
+      "step": 7564
+    },
+    {
+      "epoch": 1.9026659959758552,
+      "grad_norm": 0.2882605791091919,
+      "learning_rate": 3.551246048225132e-06,
+      "loss": 0.362,
+      "step": 7565
+    },
+    {
+      "epoch": 1.9029175050301812,
+      "grad_norm": 0.29483652114868164,
+      "learning_rate": 3.549845632869598e-06,
+      "loss": 0.3343,
+      "step": 7566
+    },
+    {
+      "epoch": 1.903169014084507,
+      "grad_norm": 0.28981560468673706,
+      "learning_rate": 3.5484453417107113e-06,
+      "loss": 0.3553,
+      "step": 7567
+    },
+    {
+      "epoch": 1.903420523138833,
+      "grad_norm": 0.321000337600708,
+      "learning_rate": 3.547045174868402e-06,
+      "loss": 0.3451,
+      "step": 7568
+    },
+    {
+      "epoch": 1.903672032193159,
+      "grad_norm": 0.29945504665374756,
+      "learning_rate": 3.545645132462582e-06,
+      "loss": 0.3361,
+      "step": 7569
+    },
+    {
+      "epoch": 1.9039235412474849,
+      "grad_norm": 0.31492796540260315,
+      "learning_rate": 3.54424521461316e-06,
+      "loss": 0.3388,
+      "step": 7570
+    },
+    {
+      "epoch": 1.904175050301811,
+      "grad_norm": 0.28294259309768677,
+      "learning_rate": 3.5428454214400265e-06,
+      "loss": 0.3443,
+      "step": 7571
+    },
+    {
+      "epoch": 1.904426559356137,
+      "grad_norm": 0.3141169846057892,
+      "learning_rate": 3.541445753063068e-06,
+      "loss": 0.3318,
+      "step": 7572
+    },
+    {
+      "epoch": 1.9046780684104627,
+      "grad_norm": 0.2912863492965698,
+      "learning_rate": 3.5400462096021547e-06,
+      "loss": 0.3491,
+      "step": 7573
+    },
+    {
+      "epoch": 1.9049295774647887,
+      "grad_norm": 0.32144850492477417,
+      "learning_rate": 3.5386467911771518e-06,
+      "loss": 0.3553,
+      "step": 7574
+    },
+    {
+      "epoch": 1.9051810865191148,
+      "grad_norm": 0.33859172463417053,
+      "learning_rate": 3.5372474979079067e-06,
+      "loss": 0.3514,
+      "step": 7575
+    },
+    {
+      "epoch": 1.9054325955734406,
+      "grad_norm": 0.3066120743751526,
+      "learning_rate": 3.5358483299142645e-06,
+      "loss": 0.3434,
+      "step": 7576
+    },
+    {
+      "epoch": 1.9056841046277666,
+      "grad_norm": 0.35578417778015137,
+      "learning_rate": 3.534449287316052e-06,
+      "loss": 0.3699,
+      "step": 7577
+    },
+    {
+      "epoch": 1.9059356136820926,
+      "grad_norm": 0.30244138836860657,
+      "learning_rate": 3.5330503702330898e-06,
+      "loss": 0.3566,
+      "step": 7578
+    },
+    {
+      "epoch": 1.9061871227364184,
+      "grad_norm": 0.2991444170475006,
+      "learning_rate": 3.5316515787851867e-06,
+      "loss": 0.3412,
+      "step": 7579
+    },
+    {
+      "epoch": 1.9064386317907445,
+      "grad_norm": 0.29319190979003906,
+      "learning_rate": 3.53025291309214e-06,
+      "loss": 0.343,
+      "step": 7580
+    },
+    {
+      "epoch": 1.9066901408450705,
+      "grad_norm": 0.3237687945365906,
+      "learning_rate": 3.528854373273736e-06,
+      "loss": 0.339,
+      "step": 7581
+    },
+    {
+      "epoch": 1.9069416498993963,
+      "grad_norm": 0.31021758913993835,
+      "learning_rate": 3.5274559594497513e-06,
+      "loss": 0.3483,
+      "step": 7582
+    },
+    {
+      "epoch": 1.9071931589537223,
+      "grad_norm": 0.3291212320327759,
+      "learning_rate": 3.5260576717399518e-06,
+      "loss": 0.3483,
+      "step": 7583
+    },
+    {
+      "epoch": 1.9074446680080483,
+      "grad_norm": 0.2789864242076874,
+      "learning_rate": 3.5246595102640924e-06,
+      "loss": 0.3251,
+      "step": 7584
+    },
+    {
+      "epoch": 1.9076961770623742,
+      "grad_norm": 0.29820239543914795,
+      "learning_rate": 3.523261475141916e-06,
+      "loss": 0.3354,
+      "step": 7585
+    },
+    {
+      "epoch": 1.9079476861167002,
+      "grad_norm": 0.33340615034103394,
+      "learning_rate": 3.5218635664931556e-06,
+      "loss": 0.352,
+      "step": 7586
+    },
+    {
+      "epoch": 1.9081991951710262,
+      "grad_norm": 0.3217755854129791,
+      "learning_rate": 3.5204657844375323e-06,
+      "loss": 0.3632,
+      "step": 7587
+    },
+    {
+      "epoch": 1.908450704225352,
+      "grad_norm": 0.28968545794487,
+      "learning_rate": 3.5190681290947603e-06,
+      "loss": 0.3159,
+      "step": 7588
+    },
+    {
+      "epoch": 1.908702213279678,
+      "grad_norm": 0.2866925895214081,
+      "learning_rate": 3.517670600584537e-06,
+      "loss": 0.3443,
+      "step": 7589
+    },
+    {
+      "epoch": 1.908953722334004,
+      "grad_norm": 0.30135875940322876,
+      "learning_rate": 3.5162731990265553e-06,
+      "loss": 0.3629,
+      "step": 7590
+    },
+    {
+      "epoch": 1.9092052313883299,
+      "grad_norm": 0.28496021032333374,
+      "learning_rate": 3.5148759245404895e-06,
+      "loss": 0.3375,
+      "step": 7591
+    },
+    {
+      "epoch": 1.909456740442656,
+      "grad_norm": 0.29872947931289673,
+      "learning_rate": 3.513478777246012e-06,
+      "loss": 0.32,
+      "step": 7592
+    },
+    {
+      "epoch": 1.909708249496982,
+      "grad_norm": 0.3028433620929718,
+      "learning_rate": 3.5120817572627763e-06,
+      "loss": 0.3223,
+      "step": 7593
+    },
+    {
+      "epoch": 1.9099597585513077,
+      "grad_norm": 0.32273510098457336,
+      "learning_rate": 3.510684864710431e-06,
+      "loss": 0.3305,
+      "step": 7594
+    },
+    {
+      "epoch": 1.9102112676056338,
+      "grad_norm": 0.30309391021728516,
+      "learning_rate": 3.5092880997086076e-06,
+      "loss": 0.3564,
+      "step": 7595
+    },
+    {
+      "epoch": 1.9104627766599598,
+      "grad_norm": 0.3080253601074219,
+      "learning_rate": 3.5078914623769357e-06,
+      "loss": 0.3443,
+      "step": 7596
+    },
+    {
+      "epoch": 1.9107142857142856,
+      "grad_norm": 0.299723356962204,
+      "learning_rate": 3.506494952835022e-06,
+      "loss": 0.3494,
+      "step": 7597
+    },
+    {
+      "epoch": 1.9109657947686118,
+      "grad_norm": 0.3054339587688446,
+      "learning_rate": 3.505098571202473e-06,
+      "loss": 0.3533,
+      "step": 7598
+    },
+    {
+      "epoch": 1.9112173038229376,
+      "grad_norm": 0.2950621545314789,
+      "learning_rate": 3.5037023175988818e-06,
+      "loss": 0.3482,
+      "step": 7599
+    },
+    {
+      "epoch": 1.9114688128772634,
+      "grad_norm": 0.279410183429718,
+      "learning_rate": 3.502306192143824e-06,
+      "loss": 0.346,
+      "step": 7600
+    },
+    {
+      "epoch": 1.9117203219315897,
+      "grad_norm": 0.3337315320968628,
+      "learning_rate": 3.500910194956873e-06,
+      "loss": 0.3456,
+      "step": 7601
+    },
+    {
+      "epoch": 1.9119718309859155,
+      "grad_norm": 0.3044759929180145,
+      "learning_rate": 3.4995143261575835e-06,
+      "loss": 0.3345,
+      "step": 7602
+    },
+    {
+      "epoch": 1.9122233400402413,
+      "grad_norm": 0.33510109782218933,
+      "learning_rate": 3.4981185858655076e-06,
+      "loss": 0.3427,
+      "step": 7603
+    },
+    {
+      "epoch": 1.9124748490945676,
+      "grad_norm": 0.30306974053382874,
+      "learning_rate": 3.4967229742001764e-06,
+      "loss": 0.3364,
+      "step": 7604
+    },
+    {
+      "epoch": 1.9127263581488934,
+      "grad_norm": 0.31048643589019775,
+      "learning_rate": 3.4953274912811198e-06,
+      "loss": 0.3343,
+      "step": 7605
+    },
+    {
+      "epoch": 1.9129778672032192,
+      "grad_norm": 0.30860239267349243,
+      "learning_rate": 3.493932137227849e-06,
+      "loss": 0.3387,
+      "step": 7606
+    },
+    {
+      "epoch": 1.9132293762575454,
+      "grad_norm": 0.33586663007736206,
+      "learning_rate": 3.4925369121598708e-06,
+      "loss": 0.3428,
+      "step": 7607
+    },
+    {
+      "epoch": 1.9134808853118712,
+      "grad_norm": 0.2864849269390106,
+      "learning_rate": 3.4911418161966726e-06,
+      "loss": 0.3449,
+      "step": 7608
+    },
+    {
+      "epoch": 1.913732394366197,
+      "grad_norm": 0.3048544228076935,
+      "learning_rate": 3.489746849457739e-06,
+      "loss": 0.349,
+      "step": 7609
+    },
+    {
+      "epoch": 1.9139839034205233,
+      "grad_norm": 0.3124396502971649,
+      "learning_rate": 3.4883520120625414e-06,
+      "loss": 0.3296,
+      "step": 7610
+    },
+    {
+      "epoch": 1.914235412474849,
+      "grad_norm": 0.32257482409477234,
+      "learning_rate": 3.486957304130535e-06,
+      "loss": 0.3414,
+      "step": 7611
+    },
+    {
+      "epoch": 1.914486921529175,
+      "grad_norm": 0.30354657769203186,
+      "learning_rate": 3.4855627257811727e-06,
+      "loss": 0.3571,
+      "step": 7612
+    },
+    {
+      "epoch": 1.9147384305835011,
+      "grad_norm": 0.2927219867706299,
+      "learning_rate": 3.484168277133886e-06,
+      "loss": 0.3492,
+      "step": 7613
+    },
+    {
+      "epoch": 1.914989939637827,
+      "grad_norm": 0.3123704195022583,
+      "learning_rate": 3.4827739583081054e-06,
+      "loss": 0.3536,
+      "step": 7614
+    },
+    {
+      "epoch": 1.915241448692153,
+      "grad_norm": 0.3322194814682007,
+      "learning_rate": 3.481379769423242e-06,
+      "loss": 0.366,
+      "step": 7615
+    },
+    {
+      "epoch": 1.915492957746479,
+      "grad_norm": 0.3077843487262726,
+      "learning_rate": 3.479985710598702e-06,
+      "loss": 0.3643,
+      "step": 7616
+    },
+    {
+      "epoch": 1.9157444668008048,
+      "grad_norm": 0.33019596338272095,
+      "learning_rate": 3.4785917819538757e-06,
+      "loss": 0.3169,
+      "step": 7617
+    },
+    {
+      "epoch": 1.9159959758551308,
+      "grad_norm": 0.29199883341789246,
+      "learning_rate": 3.477197983608147e-06,
+      "loss": 0.3315,
+      "step": 7618
+    },
+    {
+      "epoch": 1.9162474849094568,
+      "grad_norm": 0.29495760798454285,
+      "learning_rate": 3.475804315680882e-06,
+      "loss": 0.3453,
+      "step": 7619
+    },
+    {
+      "epoch": 1.9164989939637826,
+      "grad_norm": 0.2928338646888733,
+      "learning_rate": 3.4744107782914425e-06,
+      "loss": 0.3661,
+      "step": 7620
+    },
+    {
+      "epoch": 1.9167505030181087,
+      "grad_norm": 0.3239118754863739,
+      "learning_rate": 3.4730173715591773e-06,
+      "loss": 0.3537,
+      "step": 7621
+    },
+    {
+      "epoch": 1.9170020120724347,
+      "grad_norm": 0.29765039682388306,
+      "learning_rate": 3.4716240956034197e-06,
+      "loss": 0.3481,
+      "step": 7622
+    },
+    {
+      "epoch": 1.9172535211267605,
+      "grad_norm": 0.2840639054775238,
+      "learning_rate": 3.4702309505434996e-06,
+      "loss": 0.3504,
+      "step": 7623
+    },
+    {
+      "epoch": 1.9175050301810865,
+      "grad_norm": 0.3031284511089325,
+      "learning_rate": 3.468837936498725e-06,
+      "loss": 0.3553,
+      "step": 7624
+    },
+    {
+      "epoch": 1.9177565392354126,
+      "grad_norm": 0.31842097640037537,
+      "learning_rate": 3.4674450535884053e-06,
+      "loss": 0.3413,
+      "step": 7625
+    },
+    {
+      "epoch": 1.9180080482897384,
+      "grad_norm": 0.3358910083770752,
+      "learning_rate": 3.4660523019318267e-06,
+      "loss": 0.3566,
+      "step": 7626
+    },
+    {
+      "epoch": 1.9182595573440644,
+      "grad_norm": 0.33151108026504517,
+      "learning_rate": 3.4646596816482743e-06,
+      "loss": 0.3433,
+      "step": 7627
+    },
+    {
+      "epoch": 1.9185110663983904,
+      "grad_norm": 0.31569045782089233,
+      "learning_rate": 3.4632671928570126e-06,
+      "loss": 0.3264,
+      "step": 7628
+    },
+    {
+      "epoch": 1.9187625754527162,
+      "grad_norm": 0.30424806475639343,
+      "learning_rate": 3.4618748356773046e-06,
+      "loss": 0.3341,
+      "step": 7629
+    },
+    {
+      "epoch": 1.9190140845070423,
+      "grad_norm": 0.29679644107818604,
+      "learning_rate": 3.460482610228392e-06,
+      "loss": 0.3414,
+      "step": 7630
+    },
+    {
+      "epoch": 1.9192655935613683,
+      "grad_norm": 0.3199297785758972,
+      "learning_rate": 3.459090516629514e-06,
+      "loss": 0.3571,
+      "step": 7631
+    },
+    {
+      "epoch": 1.919517102615694,
+      "grad_norm": 0.3305947482585907,
+      "learning_rate": 3.457698554999893e-06,
+      "loss": 0.353,
+      "step": 7632
+    },
+    {
+      "epoch": 1.91976861167002,
+      "grad_norm": 0.31420543789863586,
+      "learning_rate": 3.4563067254587424e-06,
+      "loss": 0.3556,
+      "step": 7633
+    },
+    {
+      "epoch": 1.9200201207243461,
+      "grad_norm": 0.2856319844722748,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.3611,
+      "step": 7634
+    },
+    {
+      "epoch": 1.920271629778672,
+      "grad_norm": 0.30721721053123474,
+      "learning_rate": 3.4535234631186466e-06,
+      "loss": 0.324,
+      "step": 7635
+    },
+    {
+      "epoch": 1.920523138832998,
+      "grad_norm": 0.3075389266014099,
+      "learning_rate": 3.4521320305580697e-06,
+      "loss": 0.3438,
+      "step": 7636
+    },
+    {
+      "epoch": 1.920774647887324,
+      "grad_norm": 0.32247766852378845,
+      "learning_rate": 3.4507407305627018e-06,
+      "loss": 0.3569,
+      "step": 7637
+    },
+    {
+      "epoch": 1.9210261569416498,
+      "grad_norm": 0.32027557492256165,
+      "learning_rate": 3.449349563251697e-06,
+      "loss": 0.3591,
+      "step": 7638
+    },
+    {
+      "epoch": 1.9212776659959758,
+      "grad_norm": 0.292683482170105,
+      "learning_rate": 3.4479585287442025e-06,
+      "loss": 0.3264,
+      "step": 7639
+    },
+    {
+      "epoch": 1.9215291750503019,
+      "grad_norm": 0.3263826072216034,
+      "learning_rate": 3.4465676271593495e-06,
+      "loss": 0.3507,
+      "step": 7640
+    },
+    {
+      "epoch": 1.9217806841046277,
+      "grad_norm": 0.3174888491630554,
+      "learning_rate": 3.445176858616262e-06,
+      "loss": 0.346,
+      "step": 7641
+    },
+    {
+      "epoch": 1.9220321931589537,
+      "grad_norm": 0.301901638507843,
+      "learning_rate": 3.443786223234048e-06,
+      "loss": 0.357,
+      "step": 7642
+    },
+    {
+      "epoch": 1.9222837022132797,
+      "grad_norm": 0.3539835810661316,
+      "learning_rate": 3.4423957211318092e-06,
+      "loss": 0.3506,
+      "step": 7643
+    },
+    {
+      "epoch": 1.9225352112676055,
+      "grad_norm": 0.3143352270126343,
+      "learning_rate": 3.441005352428633e-06,
+      "loss": 0.3439,
+      "step": 7644
+    },
+    {
+      "epoch": 1.9227867203219315,
+      "grad_norm": 0.30168989300727844,
+      "learning_rate": 3.4396151172435954e-06,
+      "loss": 0.3677,
+      "step": 7645
+    },
+    {
+      "epoch": 1.9230382293762576,
+      "grad_norm": 0.3380453884601593,
+      "learning_rate": 3.4382250156957607e-06,
+      "loss": 0.3338,
+      "step": 7646
+    },
+    {
+      "epoch": 1.9232897384305834,
+      "grad_norm": 0.2964382767677307,
+      "learning_rate": 3.4368350479041836e-06,
+      "loss": 0.3358,
+      "step": 7647
+    },
+    {
+      "epoch": 1.9235412474849096,
+      "grad_norm": 0.2817126512527466,
+      "learning_rate": 3.4354452139879044e-06,
+      "loss": 0.3568,
+      "step": 7648
+    },
+    {
+      "epoch": 1.9237927565392354,
+      "grad_norm": 0.32317665219306946,
+      "learning_rate": 3.434055514065956e-06,
+      "loss": 0.358,
+      "step": 7649
+    },
+    {
+      "epoch": 1.9240442655935612,
+      "grad_norm": 0.3008497953414917,
+      "learning_rate": 3.4326659482573556e-06,
+      "loss": 0.3482,
+      "step": 7650
+    },
+    {
+      "epoch": 1.9242957746478875,
+      "grad_norm": 0.29205819964408875,
+      "learning_rate": 3.431276516681112e-06,
+      "loss": 0.3532,
+      "step": 7651
+    },
+    {
+      "epoch": 1.9245472837022133,
+      "grad_norm": 0.3241797983646393,
+      "learning_rate": 3.4298872194562203e-06,
+      "loss": 0.3549,
+      "step": 7652
+    },
+    {
+      "epoch": 1.924798792756539,
+      "grad_norm": 0.31223297119140625,
+      "learning_rate": 3.428498056701665e-06,
+      "loss": 0.3619,
+      "step": 7653
+    },
+    {
+      "epoch": 1.9250503018108653,
+      "grad_norm": 0.3212898075580597,
+      "learning_rate": 3.4271090285364216e-06,
+      "loss": 0.367,
+      "step": 7654
+    },
+    {
+      "epoch": 1.9253018108651911,
+      "grad_norm": 0.29246649146080017,
+      "learning_rate": 3.4257201350794487e-06,
+      "loss": 0.3404,
+      "step": 7655
+    },
+    {
+      "epoch": 1.925553319919517,
+      "grad_norm": 0.33508625626564026,
+      "learning_rate": 3.424331376449699e-06,
+      "loss": 0.3407,
+      "step": 7656
+    },
+    {
+      "epoch": 1.9258048289738432,
+      "grad_norm": 0.3068259358406067,
+      "learning_rate": 3.4229427527661074e-06,
+      "loss": 0.3378,
+      "step": 7657
+    },
+    {
+      "epoch": 1.926056338028169,
+      "grad_norm": 0.2951129972934723,
+      "learning_rate": 3.4215542641476053e-06,
+      "loss": 0.3404,
+      "step": 7658
+    },
+    {
+      "epoch": 1.9263078470824948,
+      "grad_norm": 0.28135162591934204,
+      "learning_rate": 3.4201659107131036e-06,
+      "loss": 0.3448,
+      "step": 7659
+    },
+    {
+      "epoch": 1.926559356136821,
+      "grad_norm": 0.2802700698375702,
+      "learning_rate": 3.4187776925815103e-06,
+      "loss": 0.3163,
+      "step": 7660
+    },
+    {
+      "epoch": 1.9268108651911469,
+      "grad_norm": 0.29353809356689453,
+      "learning_rate": 3.4173896098717134e-06,
+      "loss": 0.3673,
+      "step": 7661
+    },
+    {
+      "epoch": 1.9270623742454729,
+      "grad_norm": 0.3133329153060913,
+      "learning_rate": 3.4160016627025976e-06,
+      "loss": 0.3452,
+      "step": 7662
+    },
+    {
+      "epoch": 1.927313883299799,
+      "grad_norm": 0.29814791679382324,
+      "learning_rate": 3.414613851193028e-06,
+      "loss": 0.3653,
+      "step": 7663
+    },
+    {
+      "epoch": 1.9275653923541247,
+      "grad_norm": 0.28661277890205383,
+      "learning_rate": 3.4132261754618646e-06,
+      "loss": 0.3484,
+      "step": 7664
+    },
+    {
+      "epoch": 1.9278169014084507,
+      "grad_norm": 0.3048580586910248,
+      "learning_rate": 3.411838635627953e-06,
+      "loss": 0.3434,
+      "step": 7665
+    },
+    {
+      "epoch": 1.9280684104627768,
+      "grad_norm": 0.3117848336696625,
+      "learning_rate": 3.4104512318101256e-06,
+      "loss": 0.3471,
+      "step": 7666
+    },
+    {
+      "epoch": 1.9283199195171026,
+      "grad_norm": 0.31257185339927673,
+      "learning_rate": 3.4090639641272085e-06,
+      "loss": 0.3702,
+      "step": 7667
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.30880677700042725,
+      "learning_rate": 3.407676832698007e-06,
+      "loss": 0.337,
+      "step": 7668
+    },
+    {
+      "epoch": 1.9288229376257546,
+      "grad_norm": 0.294489324092865,
+      "learning_rate": 3.4062898376413257e-06,
+      "loss": 0.3519,
+      "step": 7669
+    },
+    {
+      "epoch": 1.9290744466800804,
+      "grad_norm": 0.2875889539718628,
+      "learning_rate": 3.404902979075948e-06,
+      "loss": 0.3557,
+      "step": 7670
+    },
+    {
+      "epoch": 1.9293259557344065,
+      "grad_norm": 0.33200985193252563,
+      "learning_rate": 3.4035162571206528e-06,
+      "loss": 0.364,
+      "step": 7671
+    },
+    {
+      "epoch": 1.9295774647887325,
+      "grad_norm": 0.31349629163742065,
+      "learning_rate": 3.4021296718942006e-06,
+      "loss": 0.3433,
+      "step": 7672
+    },
+    {
+      "epoch": 1.9298289738430583,
+      "grad_norm": 0.32347166538238525,
+      "learning_rate": 3.400743223515348e-06,
+      "loss": 0.3675,
+      "step": 7673
+    },
+    {
+      "epoch": 1.9300804828973843,
+      "grad_norm": 0.30907049775123596,
+      "learning_rate": 3.3993569121028306e-06,
+      "loss": 0.3577,
+      "step": 7674
+    },
+    {
+      "epoch": 1.9303319919517103,
+      "grad_norm": 0.28248360753059387,
+      "learning_rate": 3.397970737775381e-06,
+      "loss": 0.3327,
+      "step": 7675
+    },
+    {
+      "epoch": 1.9305835010060362,
+      "grad_norm": 0.3216158449649811,
+      "learning_rate": 3.396584700651717e-06,
+      "loss": 0.3264,
+      "step": 7676
+    },
+    {
+      "epoch": 1.9308350100603622,
+      "grad_norm": 0.3292233943939209,
+      "learning_rate": 3.395198800850541e-06,
+      "loss": 0.368,
+      "step": 7677
+    },
+    {
+      "epoch": 1.9310865191146882,
+      "grad_norm": 0.30121901631355286,
+      "learning_rate": 3.3938130384905495e-06,
+      "loss": 0.3396,
+      "step": 7678
+    },
+    {
+      "epoch": 1.931338028169014,
+      "grad_norm": 0.32539746165275574,
+      "learning_rate": 3.3924274136904214e-06,
+      "loss": 0.3689,
+      "step": 7679
+    },
+    {
+      "epoch": 1.93158953722334,
+      "grad_norm": 0.3132033944129944,
+      "learning_rate": 3.39104192656883e-06,
+      "loss": 0.3617,
+      "step": 7680
+    },
+    {
+      "epoch": 1.931841046277666,
+      "grad_norm": 0.30390465259552,
+      "learning_rate": 3.3896565772444303e-06,
+      "loss": 0.3432,
+      "step": 7681
+    },
+    {
+      "epoch": 1.9320925553319919,
+      "grad_norm": 0.3210819959640503,
+      "learning_rate": 3.3882713658358716e-06,
+      "loss": 0.3315,
+      "step": 7682
+    },
+    {
+      "epoch": 1.932344064386318,
+      "grad_norm": 0.2830940783023834,
+      "learning_rate": 3.3868862924617862e-06,
+      "loss": 0.319,
+      "step": 7683
+    },
+    {
+      "epoch": 1.932595573440644,
+      "grad_norm": 0.3306945264339447,
+      "learning_rate": 3.385501357240798e-06,
+      "loss": 0.3682,
+      "step": 7684
+    },
+    {
+      "epoch": 1.9328470824949697,
+      "grad_norm": 0.3089801073074341,
+      "learning_rate": 3.3841165602915206e-06,
+      "loss": 0.3496,
+      "step": 7685
+    },
+    {
+      "epoch": 1.9330985915492958,
+      "grad_norm": 0.31166577339172363,
+      "learning_rate": 3.3827319017325486e-06,
+      "loss": 0.3641,
+      "step": 7686
+    },
+    {
+      "epoch": 1.9333501006036218,
+      "grad_norm": 0.3063109517097473,
+      "learning_rate": 3.3813473816824743e-06,
+      "loss": 0.3454,
+      "step": 7687
+    },
+    {
+      "epoch": 1.9336016096579476,
+      "grad_norm": 0.29401347041130066,
+      "learning_rate": 3.3799630002598683e-06,
+      "loss": 0.3354,
+      "step": 7688
+    },
+    {
+      "epoch": 1.9338531187122736,
+      "grad_norm": 0.2971067726612091,
+      "learning_rate": 3.3785787575832974e-06,
+      "loss": 0.3436,
+      "step": 7689
+    },
+    {
+      "epoch": 1.9341046277665996,
+      "grad_norm": 0.28379392623901367,
+      "learning_rate": 3.377194653771311e-06,
+      "loss": 0.3655,
+      "step": 7690
+    },
+    {
+      "epoch": 1.9343561368209254,
+      "grad_norm": 0.3166361451148987,
+      "learning_rate": 3.3758106889424526e-06,
+      "loss": 0.3555,
+      "step": 7691
+    },
+    {
+      "epoch": 1.9346076458752515,
+      "grad_norm": 0.2777943015098572,
+      "learning_rate": 3.3744268632152454e-06,
+      "loss": 0.3448,
+      "step": 7692
+    },
+    {
+      "epoch": 1.9348591549295775,
+      "grad_norm": 0.3278443217277527,
+      "learning_rate": 3.37304317670821e-06,
+      "loss": 0.373,
+      "step": 7693
+    },
+    {
+      "epoch": 1.9351106639839033,
+      "grad_norm": 0.30939531326293945,
+      "learning_rate": 3.371659629539846e-06,
+      "loss": 0.3366,
+      "step": 7694
+    },
+    {
+      "epoch": 1.9353621730382293,
+      "grad_norm": 0.3135591447353363,
+      "learning_rate": 3.3702762218286487e-06,
+      "loss": 0.337,
+      "step": 7695
+    },
+    {
+      "epoch": 1.9356136820925554,
+      "grad_norm": 0.3024998903274536,
+      "learning_rate": 3.368892953693098e-06,
+      "loss": 0.3488,
+      "step": 7696
+    },
+    {
+      "epoch": 1.9358651911468812,
+      "grad_norm": 0.2925957143306732,
+      "learning_rate": 3.367509825251662e-06,
+      "loss": 0.3415,
+      "step": 7697
+    },
+    {
+      "epoch": 1.9361167002012074,
+      "grad_norm": 0.26913854479789734,
+      "learning_rate": 3.366126836622796e-06,
+      "loss": 0.3147,
+      "step": 7698
+    },
+    {
+      "epoch": 1.9363682092555332,
+      "grad_norm": 0.331188827753067,
+      "learning_rate": 3.3647439879249453e-06,
+      "loss": 0.3634,
+      "step": 7699
+    },
+    {
+      "epoch": 1.936619718309859,
+      "grad_norm": 0.3402562439441681,
+      "learning_rate": 3.363361279276541e-06,
+      "loss": 0.356,
+      "step": 7700
+    },
+    {
+      "epoch": 1.9368712273641853,
+      "grad_norm": 0.3227146863937378,
+      "learning_rate": 3.3619787107960054e-06,
+      "loss": 0.3238,
+      "step": 7701
+    },
+    {
+      "epoch": 1.937122736418511,
+      "grad_norm": 0.31124117970466614,
+      "learning_rate": 3.3605962826017457e-06,
+      "loss": 0.3313,
+      "step": 7702
+    },
+    {
+      "epoch": 1.9373742454728369,
+      "grad_norm": 0.303874671459198,
+      "learning_rate": 3.359213994812158e-06,
+      "loss": 0.338,
+      "step": 7703
+    },
+    {
+      "epoch": 1.9376257545271631,
+      "grad_norm": 0.3281771242618561,
+      "learning_rate": 3.357831847545627e-06,
+      "loss": 0.3628,
+      "step": 7704
+    },
+    {
+      "epoch": 1.937877263581489,
+      "grad_norm": 0.3109123408794403,
+      "learning_rate": 3.356449840920525e-06,
+      "loss": 0.3361,
+      "step": 7705
+    },
+    {
+      "epoch": 1.9381287726358147,
+      "grad_norm": 0.30376991629600525,
+      "learning_rate": 3.3550679750552107e-06,
+      "loss": 0.3625,
+      "step": 7706
+    },
+    {
+      "epoch": 1.938380281690141,
+      "grad_norm": 0.29432567954063416,
+      "learning_rate": 3.3536862500680354e-06,
+      "loss": 0.3326,
+      "step": 7707
+    },
+    {
+      "epoch": 1.9386317907444668,
+      "grad_norm": 0.3185214102268219,
+      "learning_rate": 3.3523046660773327e-06,
+      "loss": 0.3599,
+      "step": 7708
+    },
+    {
+      "epoch": 1.9388832997987926,
+      "grad_norm": 0.35394561290740967,
+      "learning_rate": 3.3509232232014287e-06,
+      "loss": 0.376,
+      "step": 7709
+    },
+    {
+      "epoch": 1.9391348088531188,
+      "grad_norm": 0.3000744581222534,
+      "learning_rate": 3.3495419215586324e-06,
+      "loss": 0.3304,
+      "step": 7710
+    },
+    {
+      "epoch": 1.9393863179074446,
+      "grad_norm": 0.29905176162719727,
+      "learning_rate": 3.3481607612672464e-06,
+      "loss": 0.3643,
+      "step": 7711
+    },
+    {
+      "epoch": 1.9396378269617707,
+      "grad_norm": 0.3088197410106659,
+      "learning_rate": 3.346779742445556e-06,
+      "loss": 0.3307,
+      "step": 7712
+    },
+    {
+      "epoch": 1.9398893360160967,
+      "grad_norm": 0.2897097170352936,
+      "learning_rate": 3.3453988652118398e-06,
+      "loss": 0.3296,
+      "step": 7713
+    },
+    {
+      "epoch": 1.9401408450704225,
+      "grad_norm": 0.31739160418510437,
+      "learning_rate": 3.344018129684358e-06,
+      "loss": 0.3387,
+      "step": 7714
+    },
+    {
+      "epoch": 1.9403923541247485,
+      "grad_norm": 0.315039724111557,
+      "learning_rate": 3.3426375359813655e-06,
+      "loss": 0.3463,
+      "step": 7715
+    },
+    {
+      "epoch": 1.9406438631790746,
+      "grad_norm": 0.28398001194000244,
+      "learning_rate": 3.341257084221098e-06,
+      "loss": 0.378,
+      "step": 7716
+    },
+    {
+      "epoch": 1.9408953722334004,
+      "grad_norm": 0.3012690246105194,
+      "learning_rate": 3.339876774521783e-06,
+      "loss": 0.3328,
+      "step": 7717
+    },
+    {
+      "epoch": 1.9411468812877264,
+      "grad_norm": 0.3159896731376648,
+      "learning_rate": 3.33849660700164e-06,
+      "loss": 0.3328,
+      "step": 7718
+    },
+    {
+      "epoch": 1.9413983903420524,
+      "grad_norm": 0.3031710088253021,
+      "learning_rate": 3.3371165817788655e-06,
+      "loss": 0.336,
+      "step": 7719
+    },
+    {
+      "epoch": 1.9416498993963782,
+      "grad_norm": 0.3224204182624817,
+      "learning_rate": 3.3357366989716544e-06,
+      "loss": 0.3553,
+      "step": 7720
+    },
+    {
+      "epoch": 1.9419014084507042,
+      "grad_norm": 0.31610408425331116,
+      "learning_rate": 3.3343569586981823e-06,
+      "loss": 0.307,
+      "step": 7721
+    },
+    {
+      "epoch": 1.9421529175050303,
+      "grad_norm": 0.32398226857185364,
+      "learning_rate": 3.332977361076618e-06,
+      "loss": 0.375,
+      "step": 7722
+    },
+    {
+      "epoch": 1.942404426559356,
+      "grad_norm": 0.3373188078403473,
+      "learning_rate": 3.331597906225112e-06,
+      "loss": 0.3535,
+      "step": 7723
+    },
+    {
+      "epoch": 1.942655935613682,
+      "grad_norm": 0.3140803575515747,
+      "learning_rate": 3.330218594261809e-06,
+      "loss": 0.3512,
+      "step": 7724
+    },
+    {
+      "epoch": 1.9429074446680081,
+      "grad_norm": 0.29968270659446716,
+      "learning_rate": 3.3288394253048365e-06,
+      "loss": 0.3476,
+      "step": 7725
+    },
+    {
+      "epoch": 1.943158953722334,
+      "grad_norm": 0.29010340571403503,
+      "learning_rate": 3.3274603994723144e-06,
+      "loss": 0.3188,
+      "step": 7726
+    },
+    {
+      "epoch": 1.94341046277666,
+      "grad_norm": 0.3180834949016571,
+      "learning_rate": 3.3260815168823433e-06,
+      "loss": 0.3317,
+      "step": 7727
+    },
+    {
+      "epoch": 1.943661971830986,
+      "grad_norm": 0.3009601831436157,
+      "learning_rate": 3.3247027776530183e-06,
+      "loss": 0.3546,
+      "step": 7728
+    },
+    {
+      "epoch": 1.9439134808853118,
+      "grad_norm": 0.29234665632247925,
+      "learning_rate": 3.323324181902422e-06,
+      "loss": 0.3357,
+      "step": 7729
+    },
+    {
+      "epoch": 1.9441649899396378,
+      "grad_norm": 0.2999419867992401,
+      "learning_rate": 3.321945729748618e-06,
+      "loss": 0.3627,
+      "step": 7730
+    },
+    {
+      "epoch": 1.9444164989939638,
+      "grad_norm": 0.2987615466117859,
+      "learning_rate": 3.3205674213096662e-06,
+      "loss": 0.3133,
+      "step": 7731
+    },
+    {
+      "epoch": 1.9446680080482897,
+      "grad_norm": 0.30129966139793396,
+      "learning_rate": 3.3191892567036065e-06,
+      "loss": 0.3278,
+      "step": 7732
+    },
+    {
+      "epoch": 1.9449195171026157,
+      "grad_norm": 0.2908097505569458,
+      "learning_rate": 3.317811236048474e-06,
+      "loss": 0.3454,
+      "step": 7733
+    },
+    {
+      "epoch": 1.9451710261569417,
+      "grad_norm": 0.30645039677619934,
+      "learning_rate": 3.316433359462283e-06,
+      "loss": 0.336,
+      "step": 7734
+    },
+    {
+      "epoch": 1.9454225352112675,
+      "grad_norm": 0.30733421444892883,
+      "learning_rate": 3.315055627063045e-06,
+      "loss": 0.3457,
+      "step": 7735
+    },
+    {
+      "epoch": 1.9456740442655935,
+      "grad_norm": 0.3086591362953186,
+      "learning_rate": 3.313678038968749e-06,
+      "loss": 0.3393,
+      "step": 7736
+    },
+    {
+      "epoch": 1.9459255533199196,
+      "grad_norm": 0.3118712902069092,
+      "learning_rate": 3.312300595297382e-06,
+      "loss": 0.3467,
+      "step": 7737
+    },
+    {
+      "epoch": 1.9461770623742454,
+      "grad_norm": 0.29625868797302246,
+      "learning_rate": 3.310923296166908e-06,
+      "loss": 0.3222,
+      "step": 7738
+    },
+    {
+      "epoch": 1.9464285714285714,
+      "grad_norm": 0.3383052349090576,
+      "learning_rate": 3.309546141695287e-06,
+      "loss": 0.3333,
+      "step": 7739
+    },
+    {
+      "epoch": 1.9466800804828974,
+      "grad_norm": 0.29828548431396484,
+      "learning_rate": 3.308169132000466e-06,
+      "loss": 0.3462,
+      "step": 7740
+    },
+    {
+      "epoch": 1.9469315895372232,
+      "grad_norm": 0.31392189860343933,
+      "learning_rate": 3.3067922672003727e-06,
+      "loss": 0.3538,
+      "step": 7741
+    },
+    {
+      "epoch": 1.9471830985915493,
+      "grad_norm": 0.28369051218032837,
+      "learning_rate": 3.3054155474129306e-06,
+      "loss": 0.3093,
+      "step": 7742
+    },
+    {
+      "epoch": 1.9474346076458753,
+      "grad_norm": 0.3025501072406769,
+      "learning_rate": 3.304038972756044e-06,
+      "loss": 0.3583,
+      "step": 7743
+    },
+    {
+      "epoch": 1.947686116700201,
+      "grad_norm": 0.3110698461532593,
+      "learning_rate": 3.3026625433476112e-06,
+      "loss": 0.3538,
+      "step": 7744
+    },
+    {
+      "epoch": 1.9479376257545271,
+      "grad_norm": 0.3204960823059082,
+      "learning_rate": 3.301286259305511e-06,
+      "loss": 0.3606,
+      "step": 7745
+    },
+    {
+      "epoch": 1.9481891348088531,
+      "grad_norm": 0.28646546602249146,
+      "learning_rate": 3.299910120747618e-06,
+      "loss": 0.3416,
+      "step": 7746
+    },
+    {
+      "epoch": 1.948440643863179,
+      "grad_norm": 0.31837978959083557,
+      "learning_rate": 3.298534127791785e-06,
+      "loss": 0.3218,
+      "step": 7747
+    },
+    {
+      "epoch": 1.9486921529175052,
+      "grad_norm": 0.2954898476600647,
+      "learning_rate": 3.2971582805558622e-06,
+      "loss": 0.37,
+      "step": 7748
+    },
+    {
+      "epoch": 1.948943661971831,
+      "grad_norm": 0.3241606652736664,
+      "learning_rate": 3.295782579157677e-06,
+      "loss": 0.3637,
+      "step": 7749
+    },
+    {
+      "epoch": 1.9491951710261568,
+      "grad_norm": 0.3172752559185028,
+      "learning_rate": 3.294407023715053e-06,
+      "loss": 0.3458,
+      "step": 7750
+    },
+    {
+      "epoch": 1.949446680080483,
+      "grad_norm": 0.3054000735282898,
+      "learning_rate": 3.2930316143457984e-06,
+      "loss": 0.3666,
+      "step": 7751
+    },
+    {
+      "epoch": 1.9496981891348089,
+      "grad_norm": 0.27285468578338623,
+      "learning_rate": 3.2916563511677057e-06,
+      "loss": 0.3397,
+      "step": 7752
+    },
+    {
+      "epoch": 1.9499496981891347,
+      "grad_norm": 0.32400456070899963,
+      "learning_rate": 3.2902812342985613e-06,
+      "loss": 0.3159,
+      "step": 7753
+    },
+    {
+      "epoch": 1.950201207243461,
+      "grad_norm": 0.30845338106155396,
+      "learning_rate": 3.2889062638561313e-06,
+      "loss": 0.3652,
+      "step": 7754
+    },
+    {
+      "epoch": 1.9504527162977867,
+      "grad_norm": 0.2901265621185303,
+      "learning_rate": 3.287531439958177e-06,
+      "loss": 0.3283,
+      "step": 7755
+    },
+    {
+      "epoch": 1.9507042253521125,
+      "grad_norm": 0.32865509390830994,
+      "learning_rate": 3.28615676272244e-06,
+      "loss": 0.3355,
+      "step": 7756
+    },
+    {
+      "epoch": 1.9509557344064388,
+      "grad_norm": 0.28665393590927124,
+      "learning_rate": 3.2847822322666564e-06,
+      "loss": 0.3413,
+      "step": 7757
+    },
+    {
+      "epoch": 1.9512072434607646,
+      "grad_norm": 0.3147144019603729,
+      "learning_rate": 3.283407848708542e-06,
+      "loss": 0.3519,
+      "step": 7758
+    },
+    {
+      "epoch": 1.9514587525150904,
+      "grad_norm": 0.32015910744667053,
+      "learning_rate": 3.2820336121658084e-06,
+      "loss": 0.3478,
+      "step": 7759
+    },
+    {
+      "epoch": 1.9517102615694166,
+      "grad_norm": 0.29439136385917664,
+      "learning_rate": 3.2806595227561464e-06,
+      "loss": 0.3274,
+      "step": 7760
+    },
+    {
+      "epoch": 1.9519617706237424,
+      "grad_norm": 0.303396075963974,
+      "learning_rate": 3.279285580597241e-06,
+      "loss": 0.3506,
+      "step": 7761
+    },
+    {
+      "epoch": 1.9522132796780685,
+      "grad_norm": 0.2776741087436676,
+      "learning_rate": 3.277911785806761e-06,
+      "loss": 0.344,
+      "step": 7762
+    },
+    {
+      "epoch": 1.9524647887323945,
+      "grad_norm": 0.3228535056114197,
+      "learning_rate": 3.2765381385023638e-06,
+      "loss": 0.3522,
+      "step": 7763
+    },
+    {
+      "epoch": 1.9527162977867203,
+      "grad_norm": 0.3100894093513489,
+      "learning_rate": 3.2751646388016924e-06,
+      "loss": 0.3582,
+      "step": 7764
+    },
+    {
+      "epoch": 1.9529678068410463,
+      "grad_norm": 0.3215596079826355,
+      "learning_rate": 3.27379128682238e-06,
+      "loss": 0.3395,
+      "step": 7765
+    },
+    {
+      "epoch": 1.9532193158953723,
+      "grad_norm": 0.30445247888565063,
+      "learning_rate": 3.2724180826820436e-06,
+      "loss": 0.3348,
+      "step": 7766
+    },
+    {
+      "epoch": 1.9534708249496981,
+      "grad_norm": 0.29653722047805786,
+      "learning_rate": 3.2710450264982906e-06,
+      "loss": 0.3616,
+      "step": 7767
+    },
+    {
+      "epoch": 1.9537223340040242,
+      "grad_norm": 0.31945815682411194,
+      "learning_rate": 3.269672118388716e-06,
+      "loss": 0.3714,
+      "step": 7768
+    },
+    {
+      "epoch": 1.9539738430583502,
+      "grad_norm": 0.3002147972583771,
+      "learning_rate": 3.2682993584708988e-06,
+      "loss": 0.3323,
+      "step": 7769
+    },
+    {
+      "epoch": 1.954225352112676,
+      "grad_norm": 0.3391788601875305,
+      "learning_rate": 3.2669267468624077e-06,
+      "loss": 0.3334,
+      "step": 7770
+    },
+    {
+      "epoch": 1.954476861167002,
+      "grad_norm": 0.28778892755508423,
+      "learning_rate": 3.2655542836807998e-06,
+      "loss": 0.3374,
+      "step": 7771
+    },
+    {
+      "epoch": 1.954728370221328,
+      "grad_norm": 0.3348093032836914,
+      "learning_rate": 3.264181969043615e-06,
+      "loss": 0.3448,
+      "step": 7772
+    },
+    {
+      "epoch": 1.9549798792756539,
+      "grad_norm": 0.3041500449180603,
+      "learning_rate": 3.2628098030683873e-06,
+      "loss": 0.3534,
+      "step": 7773
+    },
+    {
+      "epoch": 1.95523138832998,
+      "grad_norm": 0.32908380031585693,
+      "learning_rate": 3.26143778587263e-06,
+      "loss": 0.3498,
+      "step": 7774
+    },
+    {
+      "epoch": 1.955482897384306,
+      "grad_norm": 0.33852681517601013,
+      "learning_rate": 3.2600659175738524e-06,
+      "loss": 0.3642,
+      "step": 7775
+    },
+    {
+      "epoch": 1.9557344064386317,
+      "grad_norm": 0.3201028108596802,
+      "learning_rate": 3.2586941982895414e-06,
+      "loss": 0.3265,
+      "step": 7776
+    },
+    {
+      "epoch": 1.9559859154929577,
+      "grad_norm": 0.2781035602092743,
+      "learning_rate": 3.2573226281371817e-06,
+      "loss": 0.336,
+      "step": 7777
+    },
+    {
+      "epoch": 1.9562374245472838,
+      "grad_norm": 0.30619314312934875,
+      "learning_rate": 3.2559512072342342e-06,
+      "loss": 0.3426,
+      "step": 7778
+    },
+    {
+      "epoch": 1.9564889336016096,
+      "grad_norm": 0.31019851565361023,
+      "learning_rate": 3.2545799356981566e-06,
+      "loss": 0.3383,
+      "step": 7779
+    },
+    {
+      "epoch": 1.9567404426559356,
+      "grad_norm": 0.32283902168273926,
+      "learning_rate": 3.2532088136463867e-06,
+      "loss": 0.3363,
+      "step": 7780
+    },
+    {
+      "epoch": 1.9569919517102616,
+      "grad_norm": 0.2992129325866699,
+      "learning_rate": 3.2518378411963565e-06,
+      "loss": 0.328,
+      "step": 7781
+    },
+    {
+      "epoch": 1.9572434607645874,
+      "grad_norm": 0.2824527323246002,
+      "learning_rate": 3.2504670184654764e-06,
+      "loss": 0.3299,
+      "step": 7782
+    },
+    {
+      "epoch": 1.9574949698189135,
+      "grad_norm": 0.3210359811782837,
+      "learning_rate": 3.2490963455711506e-06,
+      "loss": 0.3574,
+      "step": 7783
+    },
+    {
+      "epoch": 1.9577464788732395,
+      "grad_norm": 0.3022248446941376,
+      "learning_rate": 3.2477258226307716e-06,
+      "loss": 0.3502,
+      "step": 7784
+    },
+    {
+      "epoch": 1.9579979879275653,
+      "grad_norm": 0.292578786611557,
+      "learning_rate": 3.2463554497617113e-06,
+      "loss": 0.3535,
+      "step": 7785
+    },
+    {
+      "epoch": 1.9582494969818913,
+      "grad_norm": 0.2895413935184479,
+      "learning_rate": 3.2449852270813386e-06,
+      "loss": 0.3292,
+      "step": 7786
+    },
+    {
+      "epoch": 1.9585010060362174,
+      "grad_norm": 0.3182920515537262,
+      "learning_rate": 3.243615154706999e-06,
+      "loss": 0.336,
+      "step": 7787
+    },
+    {
+      "epoch": 1.9587525150905432,
+      "grad_norm": 0.3026027977466583,
+      "learning_rate": 3.242245232756036e-06,
+      "loss": 0.345,
+      "step": 7788
+    },
+    {
+      "epoch": 1.9590040241448692,
+      "grad_norm": 0.28621864318847656,
+      "learning_rate": 3.2408754613457703e-06,
+      "loss": 0.3604,
+      "step": 7789
+    },
+    {
+      "epoch": 1.9592555331991952,
+      "grad_norm": 0.2878339886665344,
+      "learning_rate": 3.2395058405935186e-06,
+      "loss": 0.3392,
+      "step": 7790
+    },
+    {
+      "epoch": 1.959507042253521,
+      "grad_norm": 0.3039291203022003,
+      "learning_rate": 3.238136370616576e-06,
+      "loss": 0.3317,
+      "step": 7791
+    },
+    {
+      "epoch": 1.959758551307847,
+      "grad_norm": 0.30178239941596985,
+      "learning_rate": 3.2367670515322324e-06,
+      "loss": 0.3338,
+      "step": 7792
+    },
+    {
+      "epoch": 1.960010060362173,
+      "grad_norm": 0.2928571403026581,
+      "learning_rate": 3.2353978834577587e-06,
+      "loss": 0.3344,
+      "step": 7793
+    },
+    {
+      "epoch": 1.9602615694164989,
+      "grad_norm": 0.3058040142059326,
+      "learning_rate": 3.2340288665104167e-06,
+      "loss": 0.3786,
+      "step": 7794
+    },
+    {
+      "epoch": 1.960513078470825,
+      "grad_norm": 0.3003310263156891,
+      "learning_rate": 3.232660000807457e-06,
+      "loss": 0.351,
+      "step": 7795
+    },
+    {
+      "epoch": 1.960764587525151,
+      "grad_norm": 0.30328527092933655,
+      "learning_rate": 3.231291286466109e-06,
+      "loss": 0.3547,
+      "step": 7796
+    },
+    {
+      "epoch": 1.9610160965794767,
+      "grad_norm": 0.2882510721683502,
+      "learning_rate": 3.2299227236035996e-06,
+      "loss": 0.3301,
+      "step": 7797
+    },
+    {
+      "epoch": 1.961267605633803,
+      "grad_norm": 0.3310767114162445,
+      "learning_rate": 3.2285543123371333e-06,
+      "loss": 0.3515,
+      "step": 7798
+    },
+    {
+      "epoch": 1.9615191146881288,
+      "grad_norm": 0.31823989748954773,
+      "learning_rate": 3.2271860527839104e-06,
+      "loss": 0.3701,
+      "step": 7799
+    },
+    {
+      "epoch": 1.9617706237424546,
+      "grad_norm": 0.28764277696609497,
+      "learning_rate": 3.2258179450611086e-06,
+      "loss": 0.3319,
+      "step": 7800
+    },
+    {
+      "epoch": 1.9620221327967808,
+      "grad_norm": 0.26832863688468933,
+      "learning_rate": 3.2244499892859032e-06,
+      "loss": 0.3329,
+      "step": 7801
+    },
+    {
+      "epoch": 1.9622736418511066,
+      "grad_norm": 0.29785555601119995,
+      "learning_rate": 3.2230821855754464e-06,
+      "loss": 0.3667,
+      "step": 7802
+    },
+    {
+      "epoch": 1.9625251509054324,
+      "grad_norm": 0.3114742934703827,
+      "learning_rate": 3.221714534046886e-06,
+      "loss": 0.3327,
+      "step": 7803
+    },
+    {
+      "epoch": 1.9627766599597587,
+      "grad_norm": 0.32278159260749817,
+      "learning_rate": 3.2203470348173483e-06,
+      "loss": 0.3631,
+      "step": 7804
+    },
+    {
+      "epoch": 1.9630281690140845,
+      "grad_norm": 0.283717542886734,
+      "learning_rate": 3.2189796880039535e-06,
+      "loss": 0.3244,
+      "step": 7805
+    },
+    {
+      "epoch": 1.9632796780684103,
+      "grad_norm": 0.3093319833278656,
+      "learning_rate": 3.2176124937238094e-06,
+      "loss": 0.3288,
+      "step": 7806
+    },
+    {
+      "epoch": 1.9635311871227366,
+      "grad_norm": 0.295280396938324,
+      "learning_rate": 3.2162454520940024e-06,
+      "loss": 0.3391,
+      "step": 7807
+    },
+    {
+      "epoch": 1.9637826961770624,
+      "grad_norm": 0.2879166007041931,
+      "learning_rate": 3.214878563231615e-06,
+      "loss": 0.3515,
+      "step": 7808
+    },
+    {
+      "epoch": 1.9640342052313882,
+      "grad_norm": 0.32011473178863525,
+      "learning_rate": 3.2135118272537093e-06,
+      "loss": 0.3262,
+      "step": 7809
+    },
+    {
+      "epoch": 1.9642857142857144,
+      "grad_norm": 0.305389940738678,
+      "learning_rate": 3.2121452442773405e-06,
+      "loss": 0.3509,
+      "step": 7810
+    },
+    {
+      "epoch": 1.9645372233400402,
+      "grad_norm": 0.3013128936290741,
+      "learning_rate": 3.210778814419545e-06,
+      "loss": 0.3232,
+      "step": 7811
+    },
+    {
+      "epoch": 1.9647887323943662,
+      "grad_norm": 0.28125154972076416,
+      "learning_rate": 3.2094125377973534e-06,
+      "loss": 0.3342,
+      "step": 7812
+    },
+    {
+      "epoch": 1.9650402414486923,
+      "grad_norm": 0.30323272943496704,
+      "learning_rate": 3.2080464145277736e-06,
+      "loss": 0.3375,
+      "step": 7813
+    },
+    {
+      "epoch": 1.965291750503018,
+      "grad_norm": 0.29133641719818115,
+      "learning_rate": 3.20668044472781e-06,
+      "loss": 0.3335,
+      "step": 7814
+    },
+    {
+      "epoch": 1.965543259557344,
+      "grad_norm": 0.30253052711486816,
+      "learning_rate": 3.2053146285144456e-06,
+      "loss": 0.3318,
+      "step": 7815
+    },
+    {
+      "epoch": 1.9657947686116701,
+      "grad_norm": 0.2923412024974823,
+      "learning_rate": 3.2039489660046565e-06,
+      "loss": 0.33,
+      "step": 7816
+    },
+    {
+      "epoch": 1.966046277665996,
+      "grad_norm": 0.30035942792892456,
+      "learning_rate": 3.2025834573154025e-06,
+      "loss": 0.3195,
+      "step": 7817
+    },
+    {
+      "epoch": 1.966297786720322,
+      "grad_norm": 0.30417102575302124,
+      "learning_rate": 3.2012181025636303e-06,
+      "loss": 0.3386,
+      "step": 7818
+    },
+    {
+      "epoch": 1.966549295774648,
+      "grad_norm": 0.288310170173645,
+      "learning_rate": 3.1998529018662748e-06,
+      "loss": 0.3736,
+      "step": 7819
+    },
+    {
+      "epoch": 1.9668008048289738,
+      "grad_norm": 0.2810823321342468,
+      "learning_rate": 3.1984878553402566e-06,
+      "loss": 0.322,
+      "step": 7820
+    },
+    {
+      "epoch": 1.9670523138832998,
+      "grad_norm": 0.28224313259124756,
+      "learning_rate": 3.1971229631024836e-06,
+      "loss": 0.3359,
+      "step": 7821
+    },
+    {
+      "epoch": 1.9673038229376258,
+      "grad_norm": 0.3008633553981781,
+      "learning_rate": 3.19575822526985e-06,
+      "loss": 0.3528,
+      "step": 7822
+    },
+    {
+      "epoch": 1.9675553319919517,
+      "grad_norm": 0.3109699487686157,
+      "learning_rate": 3.194393641959237e-06,
+      "loss": 0.335,
+      "step": 7823
+    },
+    {
+      "epoch": 1.9678068410462777,
+      "grad_norm": 0.2929900288581848,
+      "learning_rate": 3.193029213287513e-06,
+      "loss": 0.3471,
+      "step": 7824
+    },
+    {
+      "epoch": 1.9680583501006037,
+      "grad_norm": 0.3091834485530853,
+      "learning_rate": 3.1916649393715314e-06,
+      "loss": 0.3395,
+      "step": 7825
+    },
+    {
+      "epoch": 1.9683098591549295,
+      "grad_norm": 0.2836105525493622,
+      "learning_rate": 3.190300820328135e-06,
+      "loss": 0.3434,
+      "step": 7826
+    },
+    {
+      "epoch": 1.9685613682092555,
+      "grad_norm": 0.331920325756073,
+      "learning_rate": 3.1889368562741527e-06,
+      "loss": 0.3478,
+      "step": 7827
+    },
+    {
+      "epoch": 1.9688128772635816,
+      "grad_norm": 0.34226587414741516,
+      "learning_rate": 3.187573047326398e-06,
+      "loss": 0.3698,
+      "step": 7828
+    },
+    {
+      "epoch": 1.9690643863179074,
+      "grad_norm": 0.31301939487457275,
+      "learning_rate": 3.186209393601674e-06,
+      "loss": 0.3542,
+      "step": 7829
+    },
+    {
+      "epoch": 1.9693158953722334,
+      "grad_norm": 0.31900009512901306,
+      "learning_rate": 3.184845895216768e-06,
+      "loss": 0.3532,
+      "step": 7830
+    },
+    {
+      "epoch": 1.9695674044265594,
+      "grad_norm": 0.29443228244781494,
+      "learning_rate": 3.183482552288456e-06,
+      "loss": 0.3383,
+      "step": 7831
+    },
+    {
+      "epoch": 1.9698189134808852,
+      "grad_norm": 0.31553661823272705,
+      "learning_rate": 3.1821193649334993e-06,
+      "loss": 0.3484,
+      "step": 7832
+    },
+    {
+      "epoch": 1.9700704225352113,
+      "grad_norm": 0.2916860580444336,
+      "learning_rate": 3.180756333268646e-06,
+      "loss": 0.3332,
+      "step": 7833
+    },
+    {
+      "epoch": 1.9703219315895373,
+      "grad_norm": 0.3077201247215271,
+      "learning_rate": 3.1793934574106317e-06,
+      "loss": 0.3442,
+      "step": 7834
+    },
+    {
+      "epoch": 1.970573440643863,
+      "grad_norm": 0.32148656249046326,
+      "learning_rate": 3.1780307374761777e-06,
+      "loss": 0.3529,
+      "step": 7835
+    },
+    {
+      "epoch": 1.970824949698189,
+      "grad_norm": 0.30224621295928955,
+      "learning_rate": 3.1766681735819926e-06,
+      "loss": 0.3389,
+      "step": 7836
+    },
+    {
+      "epoch": 1.9710764587525151,
+      "grad_norm": 0.3319030702114105,
+      "learning_rate": 3.1753057658447726e-06,
+      "loss": 0.3552,
+      "step": 7837
+    },
+    {
+      "epoch": 1.971327967806841,
+      "grad_norm": 0.28931450843811035,
+      "learning_rate": 3.173943514381198e-06,
+      "loss": 0.326,
+      "step": 7838
+    },
+    {
+      "epoch": 1.971579476861167,
+      "grad_norm": 0.3092653453350067,
+      "learning_rate": 3.1725814193079384e-06,
+      "loss": 0.3511,
+      "step": 7839
+    },
+    {
+      "epoch": 1.971830985915493,
+      "grad_norm": 0.2999701499938965,
+      "learning_rate": 3.171219480741646e-06,
+      "loss": 0.3566,
+      "step": 7840
+    },
+    {
+      "epoch": 1.9720824949698188,
+      "grad_norm": 0.29582321643829346,
+      "learning_rate": 3.1698576987989672e-06,
+      "loss": 0.3297,
+      "step": 7841
+    },
+    {
+      "epoch": 1.9723340040241448,
+      "grad_norm": 0.288314551115036,
+      "learning_rate": 3.168496073596524e-06,
+      "loss": 0.3499,
+      "step": 7842
+    },
+    {
+      "epoch": 1.9725855130784709,
+      "grad_norm": 0.292122483253479,
+      "learning_rate": 3.167134605250938e-06,
+      "loss": 0.3154,
+      "step": 7843
+    },
+    {
+      "epoch": 1.9728370221327967,
+      "grad_norm": 0.3173863887786865,
+      "learning_rate": 3.1657732938788033e-06,
+      "loss": 0.3427,
+      "step": 7844
+    },
+    {
+      "epoch": 1.9730885311871227,
+      "grad_norm": 0.3070237934589386,
+      "learning_rate": 3.164412139596713e-06,
+      "loss": 0.3399,
+      "step": 7845
+    },
+    {
+      "epoch": 1.9733400402414487,
+      "grad_norm": 0.30909422039985657,
+      "learning_rate": 3.163051142521238e-06,
+      "loss": 0.3436,
+      "step": 7846
+    },
+    {
+      "epoch": 1.9735915492957745,
+      "grad_norm": 0.3138062059879303,
+      "learning_rate": 3.1616903027689407e-06,
+      "loss": 0.3271,
+      "step": 7847
+    },
+    {
+      "epoch": 1.9738430583501008,
+      "grad_norm": 0.30243563652038574,
+      "learning_rate": 3.1603296204563707e-06,
+      "loss": 0.3389,
+      "step": 7848
+    },
+    {
+      "epoch": 1.9740945674044266,
+      "grad_norm": 0.31402406096458435,
+      "learning_rate": 3.158969095700057e-06,
+      "loss": 0.3529,
+      "step": 7849
+    },
+    {
+      "epoch": 1.9743460764587524,
+      "grad_norm": 0.32227790355682373,
+      "learning_rate": 3.157608728616525e-06,
+      "loss": 0.3545,
+      "step": 7850
+    },
+    {
+      "epoch": 1.9745975855130786,
+      "grad_norm": 0.301570862531662,
+      "learning_rate": 3.156248519322278e-06,
+      "loss": 0.337,
+      "step": 7851
+    },
+    {
+      "epoch": 1.9748490945674044,
+      "grad_norm": 0.27347323298454285,
+      "learning_rate": 3.154888467933812e-06,
+      "loss": 0.328,
+      "step": 7852
+    },
+    {
+      "epoch": 1.9751006036217302,
+      "grad_norm": 0.2885780930519104,
+      "learning_rate": 3.153528574567605e-06,
+      "loss": 0.3225,
+      "step": 7853
+    },
+    {
+      "epoch": 1.9753521126760565,
+      "grad_norm": 0.30733510851860046,
+      "learning_rate": 3.152168839340125e-06,
+      "loss": 0.348,
+      "step": 7854
+    },
+    {
+      "epoch": 1.9756036217303823,
+      "grad_norm": 0.30902665853500366,
+      "learning_rate": 3.1508092623678223e-06,
+      "loss": 0.3374,
+      "step": 7855
+    },
+    {
+      "epoch": 1.975855130784708,
+      "grad_norm": 0.28816184401512146,
+      "learning_rate": 3.14944984376714e-06,
+      "loss": 0.3434,
+      "step": 7856
+    },
+    {
+      "epoch": 1.9761066398390343,
+      "grad_norm": 0.315157026052475,
+      "learning_rate": 3.1480905836544996e-06,
+      "loss": 0.3565,
+      "step": 7857
+    },
+    {
+      "epoch": 1.9763581488933601,
+      "grad_norm": 0.3467448651790619,
+      "learning_rate": 3.1467314821463147e-06,
+      "loss": 0.3529,
+      "step": 7858
+    },
+    {
+      "epoch": 1.9766096579476862,
+      "grad_norm": 0.3209885358810425,
+      "learning_rate": 3.145372539358987e-06,
+      "loss": 0.34,
+      "step": 7859
+    },
+    {
+      "epoch": 1.9768611670020122,
+      "grad_norm": 0.3004145920276642,
+      "learning_rate": 3.1440137554088957e-06,
+      "loss": 0.3567,
+      "step": 7860
+    },
+    {
+      "epoch": 1.977112676056338,
+      "grad_norm": 0.31923210620880127,
+      "learning_rate": 3.1426551304124187e-06,
+      "loss": 0.3249,
+      "step": 7861
+    },
+    {
+      "epoch": 1.977364185110664,
+      "grad_norm": 0.2933943569660187,
+      "learning_rate": 3.1412966644859073e-06,
+      "loss": 0.3495,
+      "step": 7862
+    },
+    {
+      "epoch": 1.97761569416499,
+      "grad_norm": 0.31692373752593994,
+      "learning_rate": 3.139938357745711e-06,
+      "loss": 0.3705,
+      "step": 7863
+    },
+    {
+      "epoch": 1.9778672032193159,
+      "grad_norm": 0.30593782663345337,
+      "learning_rate": 3.138580210308155e-06,
+      "loss": 0.3254,
+      "step": 7864
+    },
+    {
+      "epoch": 1.9781187122736419,
+      "grad_norm": 0.34505024552345276,
+      "learning_rate": 3.137222222289562e-06,
+      "loss": 0.3704,
+      "step": 7865
+    },
+    {
+      "epoch": 1.978370221327968,
+      "grad_norm": 0.322317898273468,
+      "learning_rate": 3.1358643938062295e-06,
+      "loss": 0.3434,
+      "step": 7866
+    },
+    {
+      "epoch": 1.9786217303822937,
+      "grad_norm": 0.2983636260032654,
+      "learning_rate": 3.134506724974452e-06,
+      "loss": 0.3634,
+      "step": 7867
+    },
+    {
+      "epoch": 1.9788732394366197,
+      "grad_norm": 0.2886245846748352,
+      "learning_rate": 3.1331492159105007e-06,
+      "loss": 0.3471,
+      "step": 7868
+    },
+    {
+      "epoch": 1.9791247484909458,
+      "grad_norm": 0.2910291254520416,
+      "learning_rate": 3.1317918667306406e-06,
+      "loss": 0.3388,
+      "step": 7869
+    },
+    {
+      "epoch": 1.9793762575452716,
+      "grad_norm": 0.32111358642578125,
+      "learning_rate": 3.130434677551122e-06,
+      "loss": 0.3532,
+      "step": 7870
+    },
+    {
+      "epoch": 1.9796277665995976,
+      "grad_norm": 0.3142852783203125,
+      "learning_rate": 3.129077648488174e-06,
+      "loss": 0.3354,
+      "step": 7871
+    },
+    {
+      "epoch": 1.9798792756539236,
+      "grad_norm": 0.29093170166015625,
+      "learning_rate": 3.1277207796580237e-06,
+      "loss": 0.3475,
+      "step": 7872
+    },
+    {
+      "epoch": 1.9801307847082494,
+      "grad_norm": 0.3069877624511719,
+      "learning_rate": 3.126364071176874e-06,
+      "loss": 0.3158,
+      "step": 7873
+    },
+    {
+      "epoch": 1.9803822937625755,
+      "grad_norm": 0.3028368353843689,
+      "learning_rate": 3.125007523160921e-06,
+      "loss": 0.3464,
+      "step": 7874
+    },
+    {
+      "epoch": 1.9806338028169015,
+      "grad_norm": 0.31962573528289795,
+      "learning_rate": 3.123651135726343e-06,
+      "loss": 0.3508,
+      "step": 7875
+    },
+    {
+      "epoch": 1.9808853118712273,
+      "grad_norm": 0.28765198588371277,
+      "learning_rate": 3.1222949089893085e-06,
+      "loss": 0.3446,
+      "step": 7876
+    },
+    {
+      "epoch": 1.9811368209255533,
+      "grad_norm": 0.30417191982269287,
+      "learning_rate": 3.120938843065966e-06,
+      "loss": 0.3452,
+      "step": 7877
+    },
+    {
+      "epoch": 1.9813883299798793,
+      "grad_norm": 0.2888725996017456,
+      "learning_rate": 3.1195829380724585e-06,
+      "loss": 0.348,
+      "step": 7878
+    },
+    {
+      "epoch": 1.9816398390342052,
+      "grad_norm": 0.3087455630302429,
+      "learning_rate": 3.1182271941249054e-06,
+      "loss": 0.361,
+      "step": 7879
+    },
+    {
+      "epoch": 1.9818913480885312,
+      "grad_norm": 0.2925972044467926,
+      "learning_rate": 3.1168716113394224e-06,
+      "loss": 0.3458,
+      "step": 7880
+    },
+    {
+      "epoch": 1.9821428571428572,
+      "grad_norm": 0.3094989061355591,
+      "learning_rate": 3.1155161898321064e-06,
+      "loss": 0.3298,
+      "step": 7881
+    },
+    {
+      "epoch": 1.982394366197183,
+      "grad_norm": 0.29126372933387756,
+      "learning_rate": 3.114160929719038e-06,
+      "loss": 0.3506,
+      "step": 7882
+    },
+    {
+      "epoch": 1.982645875251509,
+      "grad_norm": 0.29920557141304016,
+      "learning_rate": 3.1128058311162885e-06,
+      "loss": 0.3524,
+      "step": 7883
+    },
+    {
+      "epoch": 1.982897384305835,
+      "grad_norm": 0.3097069263458252,
+      "learning_rate": 3.1114508941399135e-06,
+      "loss": 0.3685,
+      "step": 7884
+    },
+    {
+      "epoch": 1.9831488933601609,
+      "grad_norm": 0.28832173347473145,
+      "learning_rate": 3.110096118905954e-06,
+      "loss": 0.3692,
+      "step": 7885
+    },
+    {
+      "epoch": 1.983400402414487,
+      "grad_norm": 0.30503520369529724,
+      "learning_rate": 3.1087415055304392e-06,
+      "loss": 0.3686,
+      "step": 7886
+    },
+    {
+      "epoch": 1.983651911468813,
+      "grad_norm": 0.2966339886188507,
+      "learning_rate": 3.1073870541293834e-06,
+      "loss": 0.3295,
+      "step": 7887
+    },
+    {
+      "epoch": 1.9839034205231387,
+      "grad_norm": 0.30514541268348694,
+      "learning_rate": 3.1060327648187855e-06,
+      "loss": 0.3639,
+      "step": 7888
+    },
+    {
+      "epoch": 1.9841549295774648,
+      "grad_norm": 0.3116115629673004,
+      "learning_rate": 3.1046786377146332e-06,
+      "loss": 0.3457,
+      "step": 7889
+    },
+    {
+      "epoch": 1.9844064386317908,
+      "grad_norm": 0.31909826397895813,
+      "learning_rate": 3.103324672932898e-06,
+      "loss": 0.3306,
+      "step": 7890
+    },
+    {
+      "epoch": 1.9846579476861166,
+      "grad_norm": 0.3044300973415375,
+      "learning_rate": 3.101970870589538e-06,
+      "loss": 0.3527,
+      "step": 7891
+    },
+    {
+      "epoch": 1.9849094567404426,
+      "grad_norm": 0.29158785939216614,
+      "learning_rate": 3.1006172308005012e-06,
+      "loss": 0.3387,
+      "step": 7892
+    },
+    {
+      "epoch": 1.9851609657947686,
+      "grad_norm": 0.308321475982666,
+      "learning_rate": 3.099263753681714e-06,
+      "loss": 0.3492,
+      "step": 7893
+    },
+    {
+      "epoch": 1.9854124748490944,
+      "grad_norm": 0.2938082218170166,
+      "learning_rate": 3.0979104393490965e-06,
+      "loss": 0.3443,
+      "step": 7894
+    },
+    {
+      "epoch": 1.9856639839034205,
+      "grad_norm": 0.30831268429756165,
+      "learning_rate": 3.0965572879185495e-06,
+      "loss": 0.354,
+      "step": 7895
+    },
+    {
+      "epoch": 1.9859154929577465,
+      "grad_norm": 0.32580068707466125,
+      "learning_rate": 3.095204299505965e-06,
+      "loss": 0.3326,
+      "step": 7896
+    },
+    {
+      "epoch": 1.9861670020120723,
+      "grad_norm": 0.2983664572238922,
+      "learning_rate": 3.093851474227213e-06,
+      "loss": 0.3367,
+      "step": 7897
+    },
+    {
+      "epoch": 1.9864185110663986,
+      "grad_norm": 0.2757643759250641,
+      "learning_rate": 3.0924988121981604e-06,
+      "loss": 0.3575,
+      "step": 7898
+    },
+    {
+      "epoch": 1.9866700201207244,
+      "grad_norm": 0.2955786883831024,
+      "learning_rate": 3.0911463135346486e-06,
+      "loss": 0.3523,
+      "step": 7899
+    },
+    {
+      "epoch": 1.9869215291750502,
+      "grad_norm": 0.285170316696167,
+      "learning_rate": 3.0897939783525156e-06,
+      "loss": 0.3152,
+      "step": 7900
+    },
+    {
+      "epoch": 1.9871730382293764,
+      "grad_norm": 0.28338173031806946,
+      "learning_rate": 3.0884418067675755e-06,
+      "loss": 0.3367,
+      "step": 7901
+    },
+    {
+      "epoch": 1.9874245472837022,
+      "grad_norm": 0.3066754937171936,
+      "learning_rate": 3.0870897988956362e-06,
+      "loss": 0.3347,
+      "step": 7902
+    },
+    {
+      "epoch": 1.987676056338028,
+      "grad_norm": 0.30550864338874817,
+      "learning_rate": 3.0857379548524914e-06,
+      "loss": 0.3358,
+      "step": 7903
+    },
+    {
+      "epoch": 1.9879275653923543,
+      "grad_norm": 0.32733213901519775,
+      "learning_rate": 3.0843862747539123e-06,
+      "loss": 0.3602,
+      "step": 7904
+    },
+    {
+      "epoch": 1.98817907444668,
+      "grad_norm": 0.2825915813446045,
+      "learning_rate": 3.0830347587156667e-06,
+      "loss": 0.3433,
+      "step": 7905
+    },
+    {
+      "epoch": 1.9884305835010059,
+      "grad_norm": 0.3268314003944397,
+      "learning_rate": 3.0816834068534994e-06,
+      "loss": 0.3502,
+      "step": 7906
+    },
+    {
+      "epoch": 1.9886820925553321,
+      "grad_norm": 0.3060625195503235,
+      "learning_rate": 3.0803322192831496e-06,
+      "loss": 0.3397,
+      "step": 7907
+    },
+    {
+      "epoch": 1.988933601609658,
+      "grad_norm": 0.28180480003356934,
+      "learning_rate": 3.0789811961203342e-06,
+      "loss": 0.3592,
+      "step": 7908
+    },
+    {
+      "epoch": 1.989185110663984,
+      "grad_norm": 0.28377094864845276,
+      "learning_rate": 3.077630337480764e-06,
+      "loss": 0.3577,
+      "step": 7909
+    },
+    {
+      "epoch": 1.98943661971831,
+      "grad_norm": 0.28484049439430237,
+      "learning_rate": 3.076279643480126e-06,
+      "loss": 0.3419,
+      "step": 7910
+    },
+    {
+      "epoch": 1.9896881287726358,
+      "grad_norm": 0.3092370927333832,
+      "learning_rate": 3.0749291142341037e-06,
+      "loss": 0.3364,
+      "step": 7911
+    },
+    {
+      "epoch": 1.9899396378269618,
+      "grad_norm": 0.29064294695854187,
+      "learning_rate": 3.073578749858358e-06,
+      "loss": 0.3469,
+      "step": 7912
+    },
+    {
+      "epoch": 1.9901911468812878,
+      "grad_norm": 0.2736586630344391,
+      "learning_rate": 3.0722285504685405e-06,
+      "loss": 0.3434,
+      "step": 7913
+    },
+    {
+      "epoch": 1.9904426559356136,
+      "grad_norm": 0.3202298581600189,
+      "learning_rate": 3.0708785161802902e-06,
+      "loss": 0.3541,
+      "step": 7914
+    },
+    {
+      "epoch": 1.9906941649899397,
+      "grad_norm": 0.2941378355026245,
+      "learning_rate": 3.0695286471092235e-06,
+      "loss": 0.3528,
+      "step": 7915
+    },
+    {
+      "epoch": 1.9909456740442657,
+      "grad_norm": 0.28272390365600586,
+      "learning_rate": 3.0681789433709535e-06,
+      "loss": 0.3324,
+      "step": 7916
+    },
+    {
+      "epoch": 1.9911971830985915,
+      "grad_norm": 0.3015698492527008,
+      "learning_rate": 3.06682940508107e-06,
+      "loss": 0.3348,
+      "step": 7917
+    },
+    {
+      "epoch": 1.9914486921529175,
+      "grad_norm": 0.3570209741592407,
+      "learning_rate": 3.065480032355156e-06,
+      "loss": 0.3712,
+      "step": 7918
+    },
+    {
+      "epoch": 1.9917002012072436,
+      "grad_norm": 0.29426610469818115,
+      "learning_rate": 3.0641308253087722e-06,
+      "loss": 0.3465,
+      "step": 7919
+    },
+    {
+      "epoch": 1.9919517102615694,
+      "grad_norm": 0.3420696258544922,
+      "learning_rate": 3.0627817840574747e-06,
+      "loss": 0.3412,
+      "step": 7920
+    },
+    {
+      "epoch": 1.9922032193158954,
+      "grad_norm": 0.3334300220012665,
+      "learning_rate": 3.061432908716797e-06,
+      "loss": 0.3811,
+      "step": 7921
+    },
+    {
+      "epoch": 1.9924547283702214,
+      "grad_norm": 0.3178221881389618,
+      "learning_rate": 3.0600841994022645e-06,
+      "loss": 0.3462,
+      "step": 7922
+    },
+    {
+      "epoch": 1.9927062374245472,
+      "grad_norm": 0.2942996919155121,
+      "learning_rate": 3.058735656229382e-06,
+      "loss": 0.3455,
+      "step": 7923
+    },
+    {
+      "epoch": 1.9929577464788732,
+      "grad_norm": 0.3083072006702423,
+      "learning_rate": 3.057387279313646e-06,
+      "loss": 0.353,
+      "step": 7924
+    },
+    {
+      "epoch": 1.9932092555331993,
+      "grad_norm": 0.30187562108039856,
+      "learning_rate": 3.056039068770539e-06,
+      "loss": 0.3473,
+      "step": 7925
+    },
+    {
+      "epoch": 1.993460764587525,
+      "grad_norm": 0.3180212080478668,
+      "learning_rate": 3.0546910247155224e-06,
+      "loss": 0.34,
+      "step": 7926
+    },
+    {
+      "epoch": 1.993712273641851,
+      "grad_norm": 0.3074866235256195,
+      "learning_rate": 3.053343147264052e-06,
+      "loss": 0.3273,
+      "step": 7927
+    },
+    {
+      "epoch": 1.9939637826961771,
+      "grad_norm": 0.3176601231098175,
+      "learning_rate": 3.0519954365315595e-06,
+      "loss": 0.3411,
+      "step": 7928
+    },
+    {
+      "epoch": 1.994215291750503,
+      "grad_norm": 0.31003740429878235,
+      "learning_rate": 3.050647892633474e-06,
+      "loss": 0.3533,
+      "step": 7929
+    },
+    {
+      "epoch": 1.994466800804829,
+      "grad_norm": 0.2907339632511139,
+      "learning_rate": 3.0493005156851997e-06,
+      "loss": 0.3365,
+      "step": 7930
+    },
+    {
+      "epoch": 1.994718309859155,
+      "grad_norm": 0.31127139925956726,
+      "learning_rate": 3.0479533058021345e-06,
+      "loss": 0.3386,
+      "step": 7931
+    },
+    {
+      "epoch": 1.9949698189134808,
+      "grad_norm": 0.30205103754997253,
+      "learning_rate": 3.046606263099654e-06,
+      "loss": 0.3306,
+      "step": 7932
+    },
+    {
+      "epoch": 1.9952213279678068,
+      "grad_norm": 0.30589067935943604,
+      "learning_rate": 3.0452593876931296e-06,
+      "loss": 0.378,
+      "step": 7933
+    },
+    {
+      "epoch": 1.9954728370221329,
+      "grad_norm": 0.3329513669013977,
+      "learning_rate": 3.0439126796979074e-06,
+      "loss": 0.3734,
+      "step": 7934
+    },
+    {
+      "epoch": 1.9957243460764587,
+      "grad_norm": 0.2932153046131134,
+      "learning_rate": 3.042566139229327e-06,
+      "loss": 0.3653,
+      "step": 7935
+    },
+    {
+      "epoch": 1.9959758551307847,
+      "grad_norm": 0.2880392074584961,
+      "learning_rate": 3.041219766402713e-06,
+      "loss": 0.3714,
+      "step": 7936
+    },
+    {
+      "epoch": 1.9962273641851107,
+      "grad_norm": 0.31852367520332336,
+      "learning_rate": 3.03987356133337e-06,
+      "loss": 0.3305,
+      "step": 7937
+    },
+    {
+      "epoch": 1.9964788732394365,
+      "grad_norm": 0.3372204899787903,
+      "learning_rate": 3.0385275241365965e-06,
+      "loss": 0.334,
+      "step": 7938
+    },
+    {
+      "epoch": 1.9967303822937625,
+      "grad_norm": 0.3119423985481262,
+      "learning_rate": 3.0371816549276667e-06,
+      "loss": 0.3488,
+      "step": 7939
+    },
+    {
+      "epoch": 1.9969818913480886,
+      "grad_norm": 0.2920767366886139,
+      "learning_rate": 3.035835953821851e-06,
+      "loss": 0.3238,
+      "step": 7940
+    },
+    {
+      "epoch": 1.9972334004024144,
+      "grad_norm": 0.3219856321811676,
+      "learning_rate": 3.0344904209343962e-06,
+      "loss": 0.3492,
+      "step": 7941
+    },
+    {
+      "epoch": 1.9974849094567404,
+      "grad_norm": 0.2855657935142517,
+      "learning_rate": 3.0331450563805433e-06,
+      "loss": 0.3,
+      "step": 7942
+    },
+    {
+      "epoch": 1.9977364185110664,
+      "grad_norm": 0.3001307547092438,
+      "learning_rate": 3.0317998602755087e-06,
+      "loss": 0.3469,
+      "step": 7943
+    },
+    {
+      "epoch": 1.9979879275653922,
+      "grad_norm": 0.29662996530532837,
+      "learning_rate": 3.0304548327345056e-06,
+      "loss": 0.3425,
+      "step": 7944
+    },
+    {
+      "epoch": 1.9982394366197183,
+      "grad_norm": 0.28861871361732483,
+      "learning_rate": 3.0291099738727226e-06,
+      "loss": 0.3553,
+      "step": 7945
+    },
+    {
+      "epoch": 1.9984909456740443,
+      "grad_norm": 0.30145004391670227,
+      "learning_rate": 3.0277652838053416e-06,
+      "loss": 0.3693,
+      "step": 7946
+    },
+    {
+      "epoch": 1.99874245472837,
+      "grad_norm": 0.32663676142692566,
+      "learning_rate": 3.0264207626475254e-06,
+      "loss": 0.343,
+      "step": 7947
+    },
+    {
+      "epoch": 1.9989939637826963,
+      "grad_norm": 0.3272598385810852,
+      "learning_rate": 3.025076410514425e-06,
+      "loss": 0.3422,
+      "step": 7948
+    },
+    {
+      "epoch": 1.9992454728370221,
+      "grad_norm": 0.3103751540184021,
+      "learning_rate": 3.023732227521174e-06,
+      "loss": 0.3515,
+      "step": 7949
+    },
+    {
+      "epoch": 1.999496981891348,
+      "grad_norm": 0.297633558511734,
+      "learning_rate": 3.0223882137828947e-06,
+      "loss": 0.3398,
+      "step": 7950
+    },
+    {
+      "epoch": 1.9997484909456742,
+      "grad_norm": 0.3038352131843567,
+      "learning_rate": 3.021044369414693e-06,
+      "loss": 0.3405,
+      "step": 7951
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3005836606025696,
+      "learning_rate": 3.0197006945316604e-06,
+      "loss": 0.3303,
+      "step": 7952
+    },
+    {
+      "epoch": 2.000251509054326,
+      "grad_norm": 0.3221258819103241,
+      "learning_rate": 3.018357189248875e-06,
+      "loss": 0.3005,
+      "step": 7953
+    },
+    {
+      "epoch": 2.000503018108652,
+      "grad_norm": 0.31759047508239746,
+      "learning_rate": 3.0170138536813984e-06,
+      "loss": 0.3198,
+      "step": 7954
+    },
+    {
+      "epoch": 2.000754527162978,
+      "grad_norm": 0.327646940946579,
+      "learning_rate": 3.015670687944281e-06,
+      "loss": 0.3294,
+      "step": 7955
+    },
+    {
+      "epoch": 2.0010060362173037,
+      "grad_norm": 0.2987602949142456,
+      "learning_rate": 3.014327692152554e-06,
+      "loss": 0.3277,
+      "step": 7956
+    },
+    {
+      "epoch": 2.00125754527163,
+      "grad_norm": 0.31388530135154724,
+      "learning_rate": 3.012984866421238e-06,
+      "loss": 0.333,
+      "step": 7957
+    },
+    {
+      "epoch": 2.0015090543259557,
+      "grad_norm": 0.29405900835990906,
+      "learning_rate": 3.0116422108653387e-06,
+      "loss": 0.3272,
+      "step": 7958
+    },
+    {
+      "epoch": 2.0017605633802815,
+      "grad_norm": 0.3055470287799835,
+      "learning_rate": 3.0102997255998433e-06,
+      "loss": 0.3186,
+      "step": 7959
+    },
+    {
+      "epoch": 2.0020120724346078,
+      "grad_norm": 0.308438777923584,
+      "learning_rate": 3.0089574107397306e-06,
+      "loss": 0.3175,
+      "step": 7960
+    },
+    {
+      "epoch": 2.0022635814889336,
+      "grad_norm": 0.3100303113460541,
+      "learning_rate": 3.007615266399958e-06,
+      "loss": 0.2978,
+      "step": 7961
+    },
+    {
+      "epoch": 2.0025150905432594,
+      "grad_norm": 0.28166472911834717,
+      "learning_rate": 3.006273292695475e-06,
+      "loss": 0.3129,
+      "step": 7962
+    },
+    {
+      "epoch": 2.0027665995975856,
+      "grad_norm": 0.2950986921787262,
+      "learning_rate": 3.0049314897412106e-06,
+      "loss": 0.3231,
+      "step": 7963
+    },
+    {
+      "epoch": 2.0030181086519114,
+      "grad_norm": 0.3130662441253662,
+      "learning_rate": 3.0035898576520844e-06,
+      "loss": 0.3189,
+      "step": 7964
+    },
+    {
+      "epoch": 2.0032696177062372,
+      "grad_norm": 0.3303288221359253,
+      "learning_rate": 3.002248396542996e-06,
+      "loss": 0.3069,
+      "step": 7965
+    },
+    {
+      "epoch": 2.0035211267605635,
+      "grad_norm": 0.30254948139190674,
+      "learning_rate": 3.000907106528836e-06,
+      "loss": 0.3085,
+      "step": 7966
+    },
+    {
+      "epoch": 2.0037726358148893,
+      "grad_norm": 0.3039596676826477,
+      "learning_rate": 2.9995659877244736e-06,
+      "loss": 0.3117,
+      "step": 7967
+    },
+    {
+      "epoch": 2.004024144869215,
+      "grad_norm": 0.30226194858551025,
+      "learning_rate": 2.9982250402447706e-06,
+      "loss": 0.3352,
+      "step": 7968
+    },
+    {
+      "epoch": 2.0042756539235413,
+      "grad_norm": 0.29361942410469055,
+      "learning_rate": 2.9968842642045713e-06,
+      "loss": 0.332,
+      "step": 7969
+    },
+    {
+      "epoch": 2.004527162977867,
+      "grad_norm": 0.2824215888977051,
+      "learning_rate": 2.9955436597187016e-06,
+      "loss": 0.3018,
+      "step": 7970
+    },
+    {
+      "epoch": 2.004778672032193,
+      "grad_norm": 0.3032889664173126,
+      "learning_rate": 2.9942032269019792e-06,
+      "loss": 0.3291,
+      "step": 7971
+    },
+    {
+      "epoch": 2.005030181086519,
+      "grad_norm": 0.2901354432106018,
+      "learning_rate": 2.9928629658692006e-06,
+      "loss": 0.2918,
+      "step": 7972
+    },
+    {
+      "epoch": 2.005281690140845,
+      "grad_norm": 0.3217187225818634,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.3387,
+      "step": 7973
+    },
+    {
+      "epoch": 2.005533199195171,
+      "grad_norm": 0.326134592294693,
+      "learning_rate": 2.9901829596146057e-06,
+      "loss": 0.3327,
+      "step": 7974
+    },
+    {
+      "epoch": 2.005784708249497,
+      "grad_norm": 0.30647000670433044,
+      "learning_rate": 2.9888432146223167e-06,
+      "loss": 0.3048,
+      "step": 7975
+    },
+    {
+      "epoch": 2.006036217303823,
+      "grad_norm": 0.2938840389251709,
+      "learning_rate": 2.9875036418730218e-06,
+      "loss": 0.3367,
+      "step": 7976
+    },
+    {
+      "epoch": 2.006287726358149,
+      "grad_norm": 0.3004414737224579,
+      "learning_rate": 2.9861642414814502e-06,
+      "loss": 0.3116,
+      "step": 7977
+    },
+    {
+      "epoch": 2.006539235412475,
+      "grad_norm": 0.3037022650241852,
+      "learning_rate": 2.984825013562315e-06,
+      "loss": 0.3275,
+      "step": 7978
+    },
+    {
+      "epoch": 2.0067907444668007,
+      "grad_norm": 0.3147777318954468,
+      "learning_rate": 2.983485958230308e-06,
+      "loss": 0.3348,
+      "step": 7979
+    },
+    {
+      "epoch": 2.007042253521127,
+      "grad_norm": 0.30514058470726013,
+      "learning_rate": 2.9821470756001148e-06,
+      "loss": 0.3299,
+      "step": 7980
+    },
+    {
+      "epoch": 2.0072937625754528,
+      "grad_norm": 0.33660414814949036,
+      "learning_rate": 2.9808083657863994e-06,
+      "loss": 0.3422,
+      "step": 7981
+    },
+    {
+      "epoch": 2.0075452716297786,
+      "grad_norm": 0.29854559898376465,
+      "learning_rate": 2.9794698289038183e-06,
+      "loss": 0.319,
+      "step": 7982
+    },
+    {
+      "epoch": 2.007796780684105,
+      "grad_norm": 0.31165847182273865,
+      "learning_rate": 2.9781314650670033e-06,
+      "loss": 0.3134,
+      "step": 7983
+    },
+    {
+      "epoch": 2.0080482897384306,
+      "grad_norm": 0.3158718943595886,
+      "learning_rate": 2.9767932743905813e-06,
+      "loss": 0.3041,
+      "step": 7984
+    },
+    {
+      "epoch": 2.0082997987927564,
+      "grad_norm": 0.3166303336620331,
+      "learning_rate": 2.9754552569891566e-06,
+      "loss": 0.3155,
+      "step": 7985
+    },
+    {
+      "epoch": 2.0085513078470827,
+      "grad_norm": 0.3297767639160156,
+      "learning_rate": 2.9741174129773253e-06,
+      "loss": 0.3683,
+      "step": 7986
+    },
+    {
+      "epoch": 2.0088028169014085,
+      "grad_norm": 0.2979738414287567,
+      "learning_rate": 2.972779742469662e-06,
+      "loss": 0.2971,
+      "step": 7987
+    },
+    {
+      "epoch": 2.0090543259557343,
+      "grad_norm": 0.2994316518306732,
+      "learning_rate": 2.971442245580731e-06,
+      "loss": 0.3391,
+      "step": 7988
+    },
+    {
+      "epoch": 2.0093058350100605,
+      "grad_norm": 0.31496062874794006,
+      "learning_rate": 2.970104922425084e-06,
+      "loss": 0.3329,
+      "step": 7989
+    },
+    {
+      "epoch": 2.0095573440643864,
+      "grad_norm": 0.3337770700454712,
+      "learning_rate": 2.9687677731172486e-06,
+      "loss": 0.3214,
+      "step": 7990
+    },
+    {
+      "epoch": 2.009808853118712,
+      "grad_norm": 0.31632599234580994,
+      "learning_rate": 2.9674307977717486e-06,
+      "loss": 0.3217,
+      "step": 7991
+    },
+    {
+      "epoch": 2.0100603621730384,
+      "grad_norm": 0.33396008610725403,
+      "learning_rate": 2.9660939965030826e-06,
+      "loss": 0.3301,
+      "step": 7992
+    },
+    {
+      "epoch": 2.010311871227364,
+      "grad_norm": 0.3123500347137451,
+      "learning_rate": 2.9647573694257436e-06,
+      "loss": 0.3126,
+      "step": 7993
+    },
+    {
+      "epoch": 2.01056338028169,
+      "grad_norm": 0.31851232051849365,
+      "learning_rate": 2.963420916654202e-06,
+      "loss": 0.351,
+      "step": 7994
+    },
+    {
+      "epoch": 2.0108148893360163,
+      "grad_norm": 0.31659290194511414,
+      "learning_rate": 2.96208463830292e-06,
+      "loss": 0.3098,
+      "step": 7995
+    },
+    {
+      "epoch": 2.011066398390342,
+      "grad_norm": 0.31261491775512695,
+      "learning_rate": 2.9607485344863375e-06,
+      "loss": 0.3452,
+      "step": 7996
+    },
+    {
+      "epoch": 2.011317907444668,
+      "grad_norm": 0.3377598822116852,
+      "learning_rate": 2.9594126053188874e-06,
+      "loss": 0.3072,
+      "step": 7997
+    },
+    {
+      "epoch": 2.011569416498994,
+      "grad_norm": 0.3177874684333801,
+      "learning_rate": 2.95807685091498e-06,
+      "loss": 0.3158,
+      "step": 7998
+    },
+    {
+      "epoch": 2.01182092555332,
+      "grad_norm": 0.3485296368598938,
+      "learning_rate": 2.9567412713890163e-06,
+      "loss": 0.3123,
+      "step": 7999
+    },
+    {
+      "epoch": 2.0120724346076457,
+      "grad_norm": 0.3071430027484894,
+      "learning_rate": 2.955405866855381e-06,
+      "loss": 0.3191,
+      "step": 8000
+    },
+    {
+      "epoch": 2.012323943661972,
+      "grad_norm": 0.30622658133506775,
+      "learning_rate": 2.9540706374284423e-06,
+      "loss": 0.3353,
+      "step": 8001
+    },
+    {
+      "epoch": 2.012575452716298,
+      "grad_norm": 0.32716530561447144,
+      "learning_rate": 2.9527355832225542e-06,
+      "loss": 0.3129,
+      "step": 8002
+    },
+    {
+      "epoch": 2.0128269617706236,
+      "grad_norm": 0.31558316946029663,
+      "learning_rate": 2.9514007043520555e-06,
+      "loss": 0.3257,
+      "step": 8003
+    },
+    {
+      "epoch": 2.01307847082495,
+      "grad_norm": 0.3272193968296051,
+      "learning_rate": 2.9500660009312698e-06,
+      "loss": 0.3263,
+      "step": 8004
+    },
+    {
+      "epoch": 2.0133299798792756,
+      "grad_norm": 0.30284756422042847,
+      "learning_rate": 2.9487314730745075e-06,
+      "loss": 0.3047,
+      "step": 8005
+    },
+    {
+      "epoch": 2.0135814889336014,
+      "grad_norm": 0.3058840036392212,
+      "learning_rate": 2.947397120896062e-06,
+      "loss": 0.3387,
+      "step": 8006
+    },
+    {
+      "epoch": 2.0138329979879277,
+      "grad_norm": 0.31466108560562134,
+      "learning_rate": 2.9460629445102106e-06,
+      "loss": 0.3362,
+      "step": 8007
+    },
+    {
+      "epoch": 2.0140845070422535,
+      "grad_norm": 0.3358847200870514,
+      "learning_rate": 2.944728944031221e-06,
+      "loss": 0.3371,
+      "step": 8008
+    },
+    {
+      "epoch": 2.0143360160965793,
+      "grad_norm": 0.28841134905815125,
+      "learning_rate": 2.9433951195733374e-06,
+      "loss": 0.302,
+      "step": 8009
+    },
+    {
+      "epoch": 2.0145875251509056,
+      "grad_norm": 0.32336318492889404,
+      "learning_rate": 2.9420614712507966e-06,
+      "loss": 0.3229,
+      "step": 8010
+    },
+    {
+      "epoch": 2.0148390342052314,
+      "grad_norm": 0.3032439053058624,
+      "learning_rate": 2.940727999177817e-06,
+      "loss": 0.3102,
+      "step": 8011
+    },
+    {
+      "epoch": 2.015090543259557,
+      "grad_norm": 0.29863134026527405,
+      "learning_rate": 2.939394703468601e-06,
+      "loss": 0.3196,
+      "step": 8012
+    },
+    {
+      "epoch": 2.0153420523138834,
+      "grad_norm": 0.30820432305336,
+      "learning_rate": 2.9380615842373372e-06,
+      "loss": 0.3429,
+      "step": 8013
+    },
+    {
+      "epoch": 2.015593561368209,
+      "grad_norm": 0.32752588391304016,
+      "learning_rate": 2.9367286415982e-06,
+      "loss": 0.3339,
+      "step": 8014
+    },
+    {
+      "epoch": 2.015845070422535,
+      "grad_norm": 0.3171319365501404,
+      "learning_rate": 2.935395875665346e-06,
+      "loss": 0.3269,
+      "step": 8015
+    },
+    {
+      "epoch": 2.0160965794768613,
+      "grad_norm": 0.3042650520801544,
+      "learning_rate": 2.9340632865529194e-06,
+      "loss": 0.3174,
+      "step": 8016
+    },
+    {
+      "epoch": 2.016348088531187,
+      "grad_norm": 0.3020657002925873,
+      "learning_rate": 2.9327308743750483e-06,
+      "loss": 0.3195,
+      "step": 8017
+    },
+    {
+      "epoch": 2.016599597585513,
+      "grad_norm": 0.298667311668396,
+      "learning_rate": 2.931398639245845e-06,
+      "loss": 0.3408,
+      "step": 8018
+    },
+    {
+      "epoch": 2.016851106639839,
+      "grad_norm": 0.30726972222328186,
+      "learning_rate": 2.9300665812794073e-06,
+      "loss": 0.2885,
+      "step": 8019
+    },
+    {
+      "epoch": 2.017102615694165,
+      "grad_norm": 0.2989729642868042,
+      "learning_rate": 2.9287347005898162e-06,
+      "loss": 0.3166,
+      "step": 8020
+    },
+    {
+      "epoch": 2.0173541247484907,
+      "grad_norm": 0.305683970451355,
+      "learning_rate": 2.9274029972911404e-06,
+      "loss": 0.3279,
+      "step": 8021
+    },
+    {
+      "epoch": 2.017605633802817,
+      "grad_norm": 0.30576571822166443,
+      "learning_rate": 2.926071471497434e-06,
+      "loss": 0.3285,
+      "step": 8022
+    },
+    {
+      "epoch": 2.017857142857143,
+      "grad_norm": 0.3315543830394745,
+      "learning_rate": 2.9247401233227285e-06,
+      "loss": 0.336,
+      "step": 8023
+    },
+    {
+      "epoch": 2.0181086519114686,
+      "grad_norm": 0.3203072249889374,
+      "learning_rate": 2.923408952881051e-06,
+      "loss": 0.3441,
+      "step": 8024
+    },
+    {
+      "epoch": 2.018360160965795,
+      "grad_norm": 0.3002256751060486,
+      "learning_rate": 2.9220779602864035e-06,
+      "loss": 0.3027,
+      "step": 8025
+    },
+    {
+      "epoch": 2.0186116700201207,
+      "grad_norm": 0.3109540641307831,
+      "learning_rate": 2.920747145652782e-06,
+      "loss": 0.3432,
+      "step": 8026
+    },
+    {
+      "epoch": 2.0188631790744465,
+      "grad_norm": 0.3139438033103943,
+      "learning_rate": 2.9194165090941575e-06,
+      "loss": 0.321,
+      "step": 8027
+    },
+    {
+      "epoch": 2.0191146881287727,
+      "grad_norm": 0.32630038261413574,
+      "learning_rate": 2.9180860507244936e-06,
+      "loss": 0.3284,
+      "step": 8028
+    },
+    {
+      "epoch": 2.0193661971830985,
+      "grad_norm": 0.31174954771995544,
+      "learning_rate": 2.916755770657733e-06,
+      "loss": 0.2986,
+      "step": 8029
+    },
+    {
+      "epoch": 2.0196177062374248,
+      "grad_norm": 0.3617261052131653,
+      "learning_rate": 2.915425669007812e-06,
+      "loss": 0.323,
+      "step": 8030
+    },
+    {
+      "epoch": 2.0198692152917506,
+      "grad_norm": 0.28990018367767334,
+      "learning_rate": 2.914095745888638e-06,
+      "loss": 0.3193,
+      "step": 8031
+    },
+    {
+      "epoch": 2.0201207243460764,
+      "grad_norm": 0.2929520905017853,
+      "learning_rate": 2.912766001414116e-06,
+      "loss": 0.3319,
+      "step": 8032
+    },
+    {
+      "epoch": 2.0203722334004026,
+      "grad_norm": 0.3127540647983551,
+      "learning_rate": 2.9114364356981274e-06,
+      "loss": 0.3179,
+      "step": 8033
+    },
+    {
+      "epoch": 2.0206237424547284,
+      "grad_norm": 0.3118465840816498,
+      "learning_rate": 2.9101070488545424e-06,
+      "loss": 0.3215,
+      "step": 8034
+    },
+    {
+      "epoch": 2.0208752515090542,
+      "grad_norm": 0.31171420216560364,
+      "learning_rate": 2.9087778409972132e-06,
+      "loss": 0.3351,
+      "step": 8035
+    },
+    {
+      "epoch": 2.0211267605633805,
+      "grad_norm": 0.3322293758392334,
+      "learning_rate": 2.9074488122399813e-06,
+      "loss": 0.3329,
+      "step": 8036
+    },
+    {
+      "epoch": 2.0213782696177063,
+      "grad_norm": 0.28532707691192627,
+      "learning_rate": 2.906119962696666e-06,
+      "loss": 0.3257,
+      "step": 8037
+    },
+    {
+      "epoch": 2.021629778672032,
+      "grad_norm": 0.2843002676963806,
+      "learning_rate": 2.9047912924810786e-06,
+      "loss": 0.3388,
+      "step": 8038
+    },
+    {
+      "epoch": 2.0218812877263583,
+      "grad_norm": 0.3153465688228607,
+      "learning_rate": 2.9034628017070064e-06,
+      "loss": 0.3385,
+      "step": 8039
+    },
+    {
+      "epoch": 2.022132796780684,
+      "grad_norm": 0.29112154245376587,
+      "learning_rate": 2.9021344904882324e-06,
+      "loss": 0.3294,
+      "step": 8040
+    },
+    {
+      "epoch": 2.02238430583501,
+      "grad_norm": 0.307882696390152,
+      "learning_rate": 2.9008063589385127e-06,
+      "loss": 0.2989,
+      "step": 8041
+    },
+    {
+      "epoch": 2.022635814889336,
+      "grad_norm": 0.3012144863605499,
+      "learning_rate": 2.899478407171598e-06,
+      "loss": 0.3385,
+      "step": 8042
+    },
+    {
+      "epoch": 2.022887323943662,
+      "grad_norm": 0.26836854219436646,
+      "learning_rate": 2.8981506353012145e-06,
+      "loss": 0.3189,
+      "step": 8043
+    },
+    {
+      "epoch": 2.023138832997988,
+      "grad_norm": 0.29367396235466003,
+      "learning_rate": 2.896823043441083e-06,
+      "loss": 0.3315,
+      "step": 8044
+    },
+    {
+      "epoch": 2.023390342052314,
+      "grad_norm": 0.29954785108566284,
+      "learning_rate": 2.895495631704898e-06,
+      "loss": 0.3305,
+      "step": 8045
+    },
+    {
+      "epoch": 2.02364185110664,
+      "grad_norm": 0.28079143166542053,
+      "learning_rate": 2.8941684002063473e-06,
+      "loss": 0.3255,
+      "step": 8046
+    },
+    {
+      "epoch": 2.0238933601609657,
+      "grad_norm": 0.28768080472946167,
+      "learning_rate": 2.892841349059098e-06,
+      "loss": 0.3205,
+      "step": 8047
+    },
+    {
+      "epoch": 2.024144869215292,
+      "grad_norm": 0.29650044441223145,
+      "learning_rate": 2.8915144783768047e-06,
+      "loss": 0.3387,
+      "step": 8048
+    },
+    {
+      "epoch": 2.0243963782696177,
+      "grad_norm": 0.2993822991847992,
+      "learning_rate": 2.8901877882731076e-06,
+      "loss": 0.293,
+      "step": 8049
+    },
+    {
+      "epoch": 2.0246478873239435,
+      "grad_norm": 0.29370445013046265,
+      "learning_rate": 2.8888612788616256e-06,
+      "loss": 0.3055,
+      "step": 8050
+    },
+    {
+      "epoch": 2.0248993963782698,
+      "grad_norm": 0.29673752188682556,
+      "learning_rate": 2.887534950255969e-06,
+      "loss": 0.2896,
+      "step": 8051
+    },
+    {
+      "epoch": 2.0251509054325956,
+      "grad_norm": 0.2987542748451233,
+      "learning_rate": 2.886208802569728e-06,
+      "loss": 0.31,
+      "step": 8052
+    },
+    {
+      "epoch": 2.0254024144869214,
+      "grad_norm": 0.30512362718582153,
+      "learning_rate": 2.8848828359164797e-06,
+      "loss": 0.3223,
+      "step": 8053
+    },
+    {
+      "epoch": 2.0256539235412476,
+      "grad_norm": 0.3173840343952179,
+      "learning_rate": 2.883557050409783e-06,
+      "loss": 0.3265,
+      "step": 8054
+    },
+    {
+      "epoch": 2.0259054325955734,
+      "grad_norm": 0.3171699047088623,
+      "learning_rate": 2.882231446163187e-06,
+      "loss": 0.3339,
+      "step": 8055
+    },
+    {
+      "epoch": 2.0261569416498992,
+      "grad_norm": 0.3157178461551666,
+      "learning_rate": 2.8809060232902165e-06,
+      "loss": 0.3051,
+      "step": 8056
+    },
+    {
+      "epoch": 2.0264084507042255,
+      "grad_norm": 0.3298434615135193,
+      "learning_rate": 2.8795807819043898e-06,
+      "loss": 0.3176,
+      "step": 8057
+    },
+    {
+      "epoch": 2.0266599597585513,
+      "grad_norm": 0.290964812040329,
+      "learning_rate": 2.878255722119202e-06,
+      "loss": 0.3387,
+      "step": 8058
+    },
+    {
+      "epoch": 2.026911468812877,
+      "grad_norm": 0.3040768504142761,
+      "learning_rate": 2.87693084404814e-06,
+      "loss": 0.3093,
+      "step": 8059
+    },
+    {
+      "epoch": 2.0271629778672033,
+      "grad_norm": 0.3241004943847656,
+      "learning_rate": 2.875606147804667e-06,
+      "loss": 0.313,
+      "step": 8060
+    },
+    {
+      "epoch": 2.027414486921529,
+      "grad_norm": 0.296888142824173,
+      "learning_rate": 2.87428163350224e-06,
+      "loss": 0.3126,
+      "step": 8061
+    },
+    {
+      "epoch": 2.027665995975855,
+      "grad_norm": 0.29444047808647156,
+      "learning_rate": 2.87295730125429e-06,
+      "loss": 0.3275,
+      "step": 8062
+    },
+    {
+      "epoch": 2.027917505030181,
+      "grad_norm": 0.2897621989250183,
+      "learning_rate": 2.871633151174243e-06,
+      "loss": 0.3072,
+      "step": 8063
+    },
+    {
+      "epoch": 2.028169014084507,
+      "grad_norm": 0.31098416447639465,
+      "learning_rate": 2.8703091833754993e-06,
+      "loss": 0.3159,
+      "step": 8064
+    },
+    {
+      "epoch": 2.028420523138833,
+      "grad_norm": 0.2953088581562042,
+      "learning_rate": 2.8689853979714505e-06,
+      "loss": 0.314,
+      "step": 8065
+    },
+    {
+      "epoch": 2.028672032193159,
+      "grad_norm": 0.2950742840766907,
+      "learning_rate": 2.8676617950754733e-06,
+      "loss": 0.3236,
+      "step": 8066
+    },
+    {
+      "epoch": 2.028923541247485,
+      "grad_norm": 0.29954662919044495,
+      "learning_rate": 2.866338374800921e-06,
+      "loss": 0.3318,
+      "step": 8067
+    },
+    {
+      "epoch": 2.0291750503018107,
+      "grad_norm": 0.28929510712623596,
+      "learning_rate": 2.8650151372611414e-06,
+      "loss": 0.341,
+      "step": 8068
+    },
+    {
+      "epoch": 2.029426559356137,
+      "grad_norm": 0.32224392890930176,
+      "learning_rate": 2.8636920825694557e-06,
+      "loss": 0.3236,
+      "step": 8069
+    },
+    {
+      "epoch": 2.0296780684104627,
+      "grad_norm": 0.2966301739215851,
+      "learning_rate": 2.8623692108391808e-06,
+      "loss": 0.3327,
+      "step": 8070
+    },
+    {
+      "epoch": 2.0299295774647885,
+      "grad_norm": 0.2904188930988312,
+      "learning_rate": 2.8610465221836094e-06,
+      "loss": 0.3303,
+      "step": 8071
+    },
+    {
+      "epoch": 2.0301810865191148,
+      "grad_norm": 0.3125072419643402,
+      "learning_rate": 2.859724016716022e-06,
+      "loss": 0.3256,
+      "step": 8072
+    },
+    {
+      "epoch": 2.0304325955734406,
+      "grad_norm": 0.3028983771800995,
+      "learning_rate": 2.858401694549683e-06,
+      "loss": 0.3228,
+      "step": 8073
+    },
+    {
+      "epoch": 2.0306841046277664,
+      "grad_norm": 0.30204835534095764,
+      "learning_rate": 2.8570795557978413e-06,
+      "loss": 0.3266,
+      "step": 8074
+    },
+    {
+      "epoch": 2.0309356136820926,
+      "grad_norm": 0.29210224747657776,
+      "learning_rate": 2.8557576005737286e-06,
+      "loss": 0.3212,
+      "step": 8075
+    },
+    {
+      "epoch": 2.0311871227364184,
+      "grad_norm": 0.32311296463012695,
+      "learning_rate": 2.854435828990563e-06,
+      "loss": 0.3015,
+      "step": 8076
+    },
+    {
+      "epoch": 2.0314386317907447,
+      "grad_norm": 0.3028305470943451,
+      "learning_rate": 2.853114241161549e-06,
+      "loss": 0.3349,
+      "step": 8077
+    },
+    {
+      "epoch": 2.0316901408450705,
+      "grad_norm": 0.3027438223361969,
+      "learning_rate": 2.851792837199866e-06,
+      "loss": 0.3319,
+      "step": 8078
+    },
+    {
+      "epoch": 2.0319416498993963,
+      "grad_norm": 0.30897650122642517,
+      "learning_rate": 2.85047161721869e-06,
+      "loss": 0.3249,
+      "step": 8079
+    },
+    {
+      "epoch": 2.0321931589537225,
+      "grad_norm": 0.30179494619369507,
+      "learning_rate": 2.849150581331169e-06,
+      "loss": 0.3486,
+      "step": 8080
+    },
+    {
+      "epoch": 2.0324446680080483,
+      "grad_norm": 0.3074062466621399,
+      "learning_rate": 2.8478297296504487e-06,
+      "loss": 0.3201,
+      "step": 8081
+    },
+    {
+      "epoch": 2.032696177062374,
+      "grad_norm": 0.29912763833999634,
+      "learning_rate": 2.846509062289646e-06,
+      "loss": 0.3105,
+      "step": 8082
+    },
+    {
+      "epoch": 2.0329476861167004,
+      "grad_norm": 0.32092058658599854,
+      "learning_rate": 2.8451885793618716e-06,
+      "loss": 0.3287,
+      "step": 8083
+    },
+    {
+      "epoch": 2.033199195171026,
+      "grad_norm": 0.3172852694988251,
+      "learning_rate": 2.8438682809802133e-06,
+      "loss": 0.3266,
+      "step": 8084
+    },
+    {
+      "epoch": 2.033450704225352,
+      "grad_norm": 0.2866196036338806,
+      "learning_rate": 2.8425481672577494e-06,
+      "loss": 0.3152,
+      "step": 8085
+    },
+    {
+      "epoch": 2.0337022132796783,
+      "grad_norm": 0.3131830394268036,
+      "learning_rate": 2.8412282383075362e-06,
+      "loss": 0.3061,
+      "step": 8086
+    },
+    {
+      "epoch": 2.033953722334004,
+      "grad_norm": 0.308741956949234,
+      "learning_rate": 2.8399084942426193e-06,
+      "loss": 0.3321,
+      "step": 8087
+    },
+    {
+      "epoch": 2.03420523138833,
+      "grad_norm": 0.29598724842071533,
+      "learning_rate": 2.8385889351760283e-06,
+      "loss": 0.3332,
+      "step": 8088
+    },
+    {
+      "epoch": 2.034456740442656,
+      "grad_norm": 0.2867431640625,
+      "learning_rate": 2.8372695612207715e-06,
+      "loss": 0.3383,
+      "step": 8089
+    },
+    {
+      "epoch": 2.034708249496982,
+      "grad_norm": 0.30722376704216003,
+      "learning_rate": 2.8359503724898485e-06,
+      "loss": 0.3236,
+      "step": 8090
+    },
+    {
+      "epoch": 2.0349597585513077,
+      "grad_norm": 0.305992990732193,
+      "learning_rate": 2.8346313690962358e-06,
+      "loss": 0.3416,
+      "step": 8091
+    },
+    {
+      "epoch": 2.035211267605634,
+      "grad_norm": 0.30428627133369446,
+      "learning_rate": 2.8333125511529012e-06,
+      "loss": 0.3309,
+      "step": 8092
+    },
+    {
+      "epoch": 2.03546277665996,
+      "grad_norm": 0.3013574779033661,
+      "learning_rate": 2.8319939187727913e-06,
+      "loss": 0.3321,
+      "step": 8093
+    },
+    {
+      "epoch": 2.0357142857142856,
+      "grad_norm": 0.2905578315258026,
+      "learning_rate": 2.83067547206884e-06,
+      "loss": 0.2902,
+      "step": 8094
+    },
+    {
+      "epoch": 2.035965794768612,
+      "grad_norm": 0.3262609839439392,
+      "learning_rate": 2.8293572111539625e-06,
+      "loss": 0.3395,
+      "step": 8095
+    },
+    {
+      "epoch": 2.0362173038229376,
+      "grad_norm": 0.31515759229660034,
+      "learning_rate": 2.8280391361410614e-06,
+      "loss": 0.347,
+      "step": 8096
+    },
+    {
+      "epoch": 2.0364688128772634,
+      "grad_norm": 0.3180714547634125,
+      "learning_rate": 2.826721247143018e-06,
+      "loss": 0.3489,
+      "step": 8097
+    },
+    {
+      "epoch": 2.0367203219315897,
+      "grad_norm": 0.30598127841949463,
+      "learning_rate": 2.825403544272706e-06,
+      "loss": 0.3448,
+      "step": 8098
+    },
+    {
+      "epoch": 2.0369718309859155,
+      "grad_norm": 0.3095637261867523,
+      "learning_rate": 2.824086027642976e-06,
+      "loss": 0.3007,
+      "step": 8099
+    },
+    {
+      "epoch": 2.0372233400402413,
+      "grad_norm": 0.32037293910980225,
+      "learning_rate": 2.822768697366664e-06,
+      "loss": 0.3288,
+      "step": 8100
+    },
+    {
+      "epoch": 2.0374748490945676,
+      "grad_norm": 0.32354459166526794,
+      "learning_rate": 2.8214515535565946e-06,
+      "loss": 0.3091,
+      "step": 8101
+    },
+    {
+      "epoch": 2.0377263581488934,
+      "grad_norm": 0.28764376044273376,
+      "learning_rate": 2.820134596325568e-06,
+      "loss": 0.3336,
+      "step": 8102
+    },
+    {
+      "epoch": 2.037977867203219,
+      "grad_norm": 0.31268617510795593,
+      "learning_rate": 2.8188178257863784e-06,
+      "loss": 0.3147,
+      "step": 8103
+    },
+    {
+      "epoch": 2.0382293762575454,
+      "grad_norm": 0.3202211856842041,
+      "learning_rate": 2.8175012420517954e-06,
+      "loss": 0.3116,
+      "step": 8104
+    },
+    {
+      "epoch": 2.038480885311871,
+      "grad_norm": 0.32938629388809204,
+      "learning_rate": 2.8161848452345784e-06,
+      "loss": 0.3222,
+      "step": 8105
+    },
+    {
+      "epoch": 2.038732394366197,
+      "grad_norm": 0.31832754611968994,
+      "learning_rate": 2.8148686354474657e-06,
+      "loss": 0.3221,
+      "step": 8106
+    },
+    {
+      "epoch": 2.0389839034205233,
+      "grad_norm": 0.29280537366867065,
+      "learning_rate": 2.8135526128031864e-06,
+      "loss": 0.3317,
+      "step": 8107
+    },
+    {
+      "epoch": 2.039235412474849,
+      "grad_norm": 0.2781398594379425,
+      "learning_rate": 2.8122367774144454e-06,
+      "loss": 0.3139,
+      "step": 8108
+    },
+    {
+      "epoch": 2.039486921529175,
+      "grad_norm": 0.33072832226753235,
+      "learning_rate": 2.8109211293939376e-06,
+      "loss": 0.3226,
+      "step": 8109
+    },
+    {
+      "epoch": 2.039738430583501,
+      "grad_norm": 0.32349687814712524,
+      "learning_rate": 2.809605668854343e-06,
+      "loss": 0.3172,
+      "step": 8110
+    },
+    {
+      "epoch": 2.039989939637827,
+      "grad_norm": 0.2853679358959198,
+      "learning_rate": 2.8082903959083165e-06,
+      "loss": 0.3007,
+      "step": 8111
+    },
+    {
+      "epoch": 2.0402414486921527,
+      "grad_norm": 0.2898256778717041,
+      "learning_rate": 2.8069753106685093e-06,
+      "loss": 0.3371,
+      "step": 8112
+    },
+    {
+      "epoch": 2.040492957746479,
+      "grad_norm": 0.31160539388656616,
+      "learning_rate": 2.8056604132475445e-06,
+      "loss": 0.3333,
+      "step": 8113
+    },
+    {
+      "epoch": 2.040744466800805,
+      "grad_norm": 0.2927866578102112,
+      "learning_rate": 2.80434570375804e-06,
+      "loss": 0.307,
+      "step": 8114
+    },
+    {
+      "epoch": 2.0409959758551306,
+      "grad_norm": 0.34478503465652466,
+      "learning_rate": 2.8030311823125877e-06,
+      "loss": 0.3505,
+      "step": 8115
+    },
+    {
+      "epoch": 2.041247484909457,
+      "grad_norm": 0.29528501629829407,
+      "learning_rate": 2.8017168490237735e-06,
+      "loss": 0.328,
+      "step": 8116
+    },
+    {
+      "epoch": 2.0414989939637826,
+      "grad_norm": 0.30806538462638855,
+      "learning_rate": 2.8004027040041555e-06,
+      "loss": 0.3352,
+      "step": 8117
+    },
+    {
+      "epoch": 2.0417505030181085,
+      "grad_norm": 0.34178322553634644,
+      "learning_rate": 2.7990887473662875e-06,
+      "loss": 0.3195,
+      "step": 8118
+    },
+    {
+      "epoch": 2.0420020120724347,
+      "grad_norm": 0.3116021454334259,
+      "learning_rate": 2.7977749792226978e-06,
+      "loss": 0.3202,
+      "step": 8119
+    },
+    {
+      "epoch": 2.0422535211267605,
+      "grad_norm": 0.33535781502723694,
+      "learning_rate": 2.7964613996859037e-06,
+      "loss": 0.3406,
+      "step": 8120
+    },
+    {
+      "epoch": 2.0425050301810863,
+      "grad_norm": 0.3147551417350769,
+      "learning_rate": 2.795148008868408e-06,
+      "loss": 0.312,
+      "step": 8121
+    },
+    {
+      "epoch": 2.0427565392354126,
+      "grad_norm": 0.30535510182380676,
+      "learning_rate": 2.7938348068826893e-06,
+      "loss": 0.3357,
+      "step": 8122
+    },
+    {
+      "epoch": 2.0430080482897384,
+      "grad_norm": 0.27879422903060913,
+      "learning_rate": 2.79252179384122e-06,
+      "loss": 0.2919,
+      "step": 8123
+    },
+    {
+      "epoch": 2.043259557344064,
+      "grad_norm": 0.29287123680114746,
+      "learning_rate": 2.791208969856447e-06,
+      "loss": 0.3204,
+      "step": 8124
+    },
+    {
+      "epoch": 2.0435110663983904,
+      "grad_norm": 0.30926448106765747,
+      "learning_rate": 2.7898963350408093e-06,
+      "loss": 0.3161,
+      "step": 8125
+    },
+    {
+      "epoch": 2.0437625754527162,
+      "grad_norm": 0.29730355739593506,
+      "learning_rate": 2.788583889506722e-06,
+      "loss": 0.3248,
+      "step": 8126
+    },
+    {
+      "epoch": 2.044014084507042,
+      "grad_norm": 0.3108483850955963,
+      "learning_rate": 2.7872716333665928e-06,
+      "loss": 0.3357,
+      "step": 8127
+    },
+    {
+      "epoch": 2.0442655935613683,
+      "grad_norm": 0.3125215470790863,
+      "learning_rate": 2.7859595667328027e-06,
+      "loss": 0.3266,
+      "step": 8128
+    },
+    {
+      "epoch": 2.044517102615694,
+      "grad_norm": 0.31513240933418274,
+      "learning_rate": 2.784647689717725e-06,
+      "loss": 0.3317,
+      "step": 8129
+    },
+    {
+      "epoch": 2.0447686116700203,
+      "grad_norm": 0.344375342130661,
+      "learning_rate": 2.7833360024337152e-06,
+      "loss": 0.3385,
+      "step": 8130
+    },
+    {
+      "epoch": 2.045020120724346,
+      "grad_norm": 0.309125691652298,
+      "learning_rate": 2.782024504993108e-06,
+      "loss": 0.337,
+      "step": 8131
+    },
+    {
+      "epoch": 2.045271629778672,
+      "grad_norm": 0.3255023956298828,
+      "learning_rate": 2.780713197508228e-06,
+      "loss": 0.3483,
+      "step": 8132
+    },
+    {
+      "epoch": 2.045523138832998,
+      "grad_norm": 0.3013308644294739,
+      "learning_rate": 2.779402080091377e-06,
+      "loss": 0.3441,
+      "step": 8133
+    },
+    {
+      "epoch": 2.045774647887324,
+      "grad_norm": 0.34748101234436035,
+      "learning_rate": 2.778091152854847e-06,
+      "loss": 0.3485,
+      "step": 8134
+    },
+    {
+      "epoch": 2.04602615694165,
+      "grad_norm": 0.3031235337257385,
+      "learning_rate": 2.776780415910908e-06,
+      "loss": 0.3231,
+      "step": 8135
+    },
+    {
+      "epoch": 2.046277665995976,
+      "grad_norm": 0.3195328116416931,
+      "learning_rate": 2.7754698693718206e-06,
+      "loss": 0.308,
+      "step": 8136
+    },
+    {
+      "epoch": 2.046529175050302,
+      "grad_norm": 0.2759738266468048,
+      "learning_rate": 2.77415951334982e-06,
+      "loss": 0.3014,
+      "step": 8137
+    },
+    {
+      "epoch": 2.0467806841046277,
+      "grad_norm": 0.3147771656513214,
+      "learning_rate": 2.772849347957134e-06,
+      "loss": 0.3274,
+      "step": 8138
+    },
+    {
+      "epoch": 2.047032193158954,
+      "grad_norm": 0.2973617911338806,
+      "learning_rate": 2.7715393733059657e-06,
+      "loss": 0.3163,
+      "step": 8139
+    },
+    {
+      "epoch": 2.0472837022132797,
+      "grad_norm": 0.3099863529205322,
+      "learning_rate": 2.7702295895085097e-06,
+      "loss": 0.3241,
+      "step": 8140
+    },
+    {
+      "epoch": 2.0475352112676055,
+      "grad_norm": 0.2844668924808502,
+      "learning_rate": 2.768919996676942e-06,
+      "loss": 0.3176,
+      "step": 8141
+    },
+    {
+      "epoch": 2.0477867203219318,
+      "grad_norm": 0.34302541613578796,
+      "learning_rate": 2.7676105949234168e-06,
+      "loss": 0.3257,
+      "step": 8142
+    },
+    {
+      "epoch": 2.0480382293762576,
+      "grad_norm": 0.32527220249176025,
+      "learning_rate": 2.7663013843600805e-06,
+      "loss": 0.3046,
+      "step": 8143
+    },
+    {
+      "epoch": 2.0482897384305834,
+      "grad_norm": 0.31658607721328735,
+      "learning_rate": 2.764992365099054e-06,
+      "loss": 0.3138,
+      "step": 8144
+    },
+    {
+      "epoch": 2.0485412474849096,
+      "grad_norm": 0.31021907925605774,
+      "learning_rate": 2.7636835372524516e-06,
+      "loss": 0.3358,
+      "step": 8145
+    },
+    {
+      "epoch": 2.0487927565392354,
+      "grad_norm": 0.3204159736633301,
+      "learning_rate": 2.7623749009323626e-06,
+      "loss": 0.3216,
+      "step": 8146
+    },
+    {
+      "epoch": 2.0490442655935612,
+      "grad_norm": 0.2946203947067261,
+      "learning_rate": 2.761066456250866e-06,
+      "loss": 0.309,
+      "step": 8147
+    },
+    {
+      "epoch": 2.0492957746478875,
+      "grad_norm": 0.29937252402305603,
+      "learning_rate": 2.759758203320019e-06,
+      "loss": 0.326,
+      "step": 8148
+    },
+    {
+      "epoch": 2.0495472837022133,
+      "grad_norm": 0.30016854405403137,
+      "learning_rate": 2.7584501422518696e-06,
+      "loss": 0.3309,
+      "step": 8149
+    },
+    {
+      "epoch": 2.049798792756539,
+      "grad_norm": 0.3086334466934204,
+      "learning_rate": 2.75714227315844e-06,
+      "loss": 0.3039,
+      "step": 8150
+    },
+    {
+      "epoch": 2.0500503018108653,
+      "grad_norm": 0.30553144216537476,
+      "learning_rate": 2.7558345961517422e-06,
+      "loss": 0.3332,
+      "step": 8151
+    },
+    {
+      "epoch": 2.050301810865191,
+      "grad_norm": 0.32114413380622864,
+      "learning_rate": 2.754527111343775e-06,
+      "loss": 0.3279,
+      "step": 8152
+    },
+    {
+      "epoch": 2.050553319919517,
+      "grad_norm": 0.3155249357223511,
+      "learning_rate": 2.753219818846511e-06,
+      "loss": 0.3356,
+      "step": 8153
+    },
+    {
+      "epoch": 2.050804828973843,
+      "grad_norm": 0.30092206597328186,
+      "learning_rate": 2.751912718771915e-06,
+      "loss": 0.3133,
+      "step": 8154
+    },
+    {
+      "epoch": 2.051056338028169,
+      "grad_norm": 0.2950682044029236,
+      "learning_rate": 2.75060581123193e-06,
+      "loss": 0.3122,
+      "step": 8155
+    },
+    {
+      "epoch": 2.051307847082495,
+      "grad_norm": 0.3297138214111328,
+      "learning_rate": 2.749299096338486e-06,
+      "loss": 0.3259,
+      "step": 8156
+    },
+    {
+      "epoch": 2.051559356136821,
+      "grad_norm": 0.29814019799232483,
+      "learning_rate": 2.7479925742034926e-06,
+      "loss": 0.3146,
+      "step": 8157
+    },
+    {
+      "epoch": 2.051810865191147,
+      "grad_norm": 0.29181772470474243,
+      "learning_rate": 2.7466862449388483e-06,
+      "loss": 0.3268,
+      "step": 8158
+    },
+    {
+      "epoch": 2.0520623742454727,
+      "grad_norm": 0.31455108523368835,
+      "learning_rate": 2.7453801086564284e-06,
+      "loss": 0.3424,
+      "step": 8159
+    },
+    {
+      "epoch": 2.052313883299799,
+      "grad_norm": 0.2971799969673157,
+      "learning_rate": 2.7440741654680995e-06,
+      "loss": 0.3083,
+      "step": 8160
+    },
+    {
+      "epoch": 2.0525653923541247,
+      "grad_norm": 0.3010028898715973,
+      "learning_rate": 2.7427684154857036e-06,
+      "loss": 0.3197,
+      "step": 8161
+    },
+    {
+      "epoch": 2.0528169014084505,
+      "grad_norm": 0.30157679319381714,
+      "learning_rate": 2.7414628588210736e-06,
+      "loss": 0.3374,
+      "step": 8162
+    },
+    {
+      "epoch": 2.0530684104627768,
+      "grad_norm": 0.3262830078601837,
+      "learning_rate": 2.7401574955860177e-06,
+      "loss": 0.332,
+      "step": 8163
+    },
+    {
+      "epoch": 2.0533199195171026,
+      "grad_norm": 0.3156241178512573,
+      "learning_rate": 2.7388523258923373e-06,
+      "loss": 0.351,
+      "step": 8164
+    },
+    {
+      "epoch": 2.0535714285714284,
+      "grad_norm": 0.28493210673332214,
+      "learning_rate": 2.737547349851808e-06,
+      "loss": 0.3083,
+      "step": 8165
+    },
+    {
+      "epoch": 2.0538229376257546,
+      "grad_norm": 0.3349147140979767,
+      "learning_rate": 2.7362425675761955e-06,
+      "loss": 0.3097,
+      "step": 8166
+    },
+    {
+      "epoch": 2.0540744466800804,
+      "grad_norm": 0.32758259773254395,
+      "learning_rate": 2.7349379791772434e-06,
+      "loss": 0.3391,
+      "step": 8167
+    },
+    {
+      "epoch": 2.0543259557344062,
+      "grad_norm": 0.3020482361316681,
+      "learning_rate": 2.733633584766685e-06,
+      "loss": 0.3545,
+      "step": 8168
+    },
+    {
+      "epoch": 2.0545774647887325,
+      "grad_norm": 0.31516534090042114,
+      "learning_rate": 2.7323293844562305e-06,
+      "loss": 0.3087,
+      "step": 8169
+    },
+    {
+      "epoch": 2.0548289738430583,
+      "grad_norm": 0.2857208549976349,
+      "learning_rate": 2.731025378357579e-06,
+      "loss": 0.316,
+      "step": 8170
+    },
+    {
+      "epoch": 2.055080482897384,
+      "grad_norm": 0.3077753186225891,
+      "learning_rate": 2.729721566582407e-06,
+      "loss": 0.3417,
+      "step": 8171
+    },
+    {
+      "epoch": 2.0553319919517103,
+      "grad_norm": 0.31536969542503357,
+      "learning_rate": 2.7284179492423825e-06,
+      "loss": 0.3375,
+      "step": 8172
+    },
+    {
+      "epoch": 2.055583501006036,
+      "grad_norm": 0.3076031804084778,
+      "learning_rate": 2.7271145264491473e-06,
+      "loss": 0.3405,
+      "step": 8173
+    },
+    {
+      "epoch": 2.055835010060362,
+      "grad_norm": 0.30033233761787415,
+      "learning_rate": 2.725811298314336e-06,
+      "loss": 0.3091,
+      "step": 8174
+    },
+    {
+      "epoch": 2.056086519114688,
+      "grad_norm": 0.28484421968460083,
+      "learning_rate": 2.724508264949558e-06,
+      "loss": 0.3294,
+      "step": 8175
+    },
+    {
+      "epoch": 2.056338028169014,
+      "grad_norm": 0.29345014691352844,
+      "learning_rate": 2.723205426466413e-06,
+      "loss": 0.3031,
+      "step": 8176
+    },
+    {
+      "epoch": 2.0565895372233403,
+      "grad_norm": 0.3001652657985687,
+      "learning_rate": 2.7219027829764777e-06,
+      "loss": 0.3259,
+      "step": 8177
+    },
+    {
+      "epoch": 2.056841046277666,
+      "grad_norm": 0.29563117027282715,
+      "learning_rate": 2.72060033459132e-06,
+      "loss": 0.3285,
+      "step": 8178
+    },
+    {
+      "epoch": 2.057092555331992,
+      "grad_norm": 0.28922346234321594,
+      "learning_rate": 2.719298081422481e-06,
+      "loss": 0.323,
+      "step": 8179
+    },
+    {
+      "epoch": 2.057344064386318,
+      "grad_norm": 0.3013193905353546,
+      "learning_rate": 2.7179960235814963e-06,
+      "loss": 0.307,
+      "step": 8180
+    },
+    {
+      "epoch": 2.057595573440644,
+      "grad_norm": 0.301833838224411,
+      "learning_rate": 2.716694161179873e-06,
+      "loss": 0.3321,
+      "step": 8181
+    },
+    {
+      "epoch": 2.0578470824949697,
+      "grad_norm": 0.3148650825023651,
+      "learning_rate": 2.7153924943291125e-06,
+      "loss": 0.2919,
+      "step": 8182
+    },
+    {
+      "epoch": 2.058098591549296,
+      "grad_norm": 0.3130474090576172,
+      "learning_rate": 2.7140910231406915e-06,
+      "loss": 0.3154,
+      "step": 8183
+    },
+    {
+      "epoch": 2.058350100603622,
+      "grad_norm": 0.30896395444869995,
+      "learning_rate": 2.7127897477260723e-06,
+      "loss": 0.2989,
+      "step": 8184
+    },
+    {
+      "epoch": 2.0586016096579476,
+      "grad_norm": 0.3000273108482361,
+      "learning_rate": 2.711488668196706e-06,
+      "loss": 0.3157,
+      "step": 8185
+    },
+    {
+      "epoch": 2.058853118712274,
+      "grad_norm": 0.32448527216911316,
+      "learning_rate": 2.710187784664015e-06,
+      "loss": 0.3304,
+      "step": 8186
+    },
+    {
+      "epoch": 2.0591046277665996,
+      "grad_norm": 0.327384889125824,
+      "learning_rate": 2.708887097239418e-06,
+      "loss": 0.3002,
+      "step": 8187
+    },
+    {
+      "epoch": 2.0593561368209254,
+      "grad_norm": 0.340179443359375,
+      "learning_rate": 2.7075866060343057e-06,
+      "loss": 0.3246,
+      "step": 8188
+    },
+    {
+      "epoch": 2.0596076458752517,
+      "grad_norm": 0.35086414217948914,
+      "learning_rate": 2.706286311160061e-06,
+      "loss": 0.3246,
+      "step": 8189
+    },
+    {
+      "epoch": 2.0598591549295775,
+      "grad_norm": 0.2781917154788971,
+      "learning_rate": 2.704986212728043e-06,
+      "loss": 0.3393,
+      "step": 8190
+    },
+    {
+      "epoch": 2.0601106639839033,
+      "grad_norm": 0.3145841360092163,
+      "learning_rate": 2.7036863108495996e-06,
+      "loss": 0.2988,
+      "step": 8191
+    },
+    {
+      "epoch": 2.0603621730382295,
+      "grad_norm": 0.32019394636154175,
+      "learning_rate": 2.702386605636057e-06,
+      "loss": 0.3337,
+      "step": 8192
+    },
+    {
+      "epoch": 2.0606136820925554,
+      "grad_norm": 0.2975797951221466,
+      "learning_rate": 2.701087097198729e-06,
+      "loss": 0.296,
+      "step": 8193
+    },
+    {
+      "epoch": 2.060865191146881,
+      "grad_norm": 0.2773379385471344,
+      "learning_rate": 2.6997877856489073e-06,
+      "loss": 0.32,
+      "step": 8194
+    },
+    {
+      "epoch": 2.0611167002012074,
+      "grad_norm": 0.31435322761535645,
+      "learning_rate": 2.698488671097872e-06,
+      "loss": 0.3106,
+      "step": 8195
+    },
+    {
+      "epoch": 2.061368209255533,
+      "grad_norm": 0.32029151916503906,
+      "learning_rate": 2.6971897536568853e-06,
+      "loss": 0.3177,
+      "step": 8196
+    },
+    {
+      "epoch": 2.061619718309859,
+      "grad_norm": 0.3515484035015106,
+      "learning_rate": 2.695891033437188e-06,
+      "loss": 0.3305,
+      "step": 8197
+    },
+    {
+      "epoch": 2.0618712273641853,
+      "grad_norm": 0.29757028818130493,
+      "learning_rate": 2.6945925105500117e-06,
+      "loss": 0.3383,
+      "step": 8198
+    },
+    {
+      "epoch": 2.062122736418511,
+      "grad_norm": 0.32565537095069885,
+      "learning_rate": 2.693294185106562e-06,
+      "loss": 0.3371,
+      "step": 8199
+    },
+    {
+      "epoch": 2.062374245472837,
+      "grad_norm": 0.3038772642612457,
+      "learning_rate": 2.691996057218036e-06,
+      "loss": 0.3313,
+      "step": 8200
+    },
+    {
+      "epoch": 2.062625754527163,
+      "grad_norm": 0.3129657506942749,
+      "learning_rate": 2.6906981269956077e-06,
+      "loss": 0.3235,
+      "step": 8201
+    },
+    {
+      "epoch": 2.062877263581489,
+      "grad_norm": 0.3186679780483246,
+      "learning_rate": 2.6894003945504393e-06,
+      "loss": 0.3371,
+      "step": 8202
+    },
+    {
+      "epoch": 2.0631287726358147,
+      "grad_norm": 0.29854926466941833,
+      "learning_rate": 2.6881028599936705e-06,
+      "loss": 0.3208,
+      "step": 8203
+    },
+    {
+      "epoch": 2.063380281690141,
+      "grad_norm": 0.28703761100769043,
+      "learning_rate": 2.6868055234364304e-06,
+      "loss": 0.3137,
+      "step": 8204
+    },
+    {
+      "epoch": 2.063631790744467,
+      "grad_norm": 0.3135847747325897,
+      "learning_rate": 2.685508384989824e-06,
+      "loss": 0.3263,
+      "step": 8205
+    },
+    {
+      "epoch": 2.0638832997987926,
+      "grad_norm": 0.32722026109695435,
+      "learning_rate": 2.684211444764945e-06,
+      "loss": 0.3446,
+      "step": 8206
+    },
+    {
+      "epoch": 2.064134808853119,
+      "grad_norm": 0.2794181704521179,
+      "learning_rate": 2.6829147028728695e-06,
+      "loss": 0.3239,
+      "step": 8207
+    },
+    {
+      "epoch": 2.0643863179074446,
+      "grad_norm": 0.29036521911621094,
+      "learning_rate": 2.6816181594246534e-06,
+      "loss": 0.3363,
+      "step": 8208
+    },
+    {
+      "epoch": 2.0646378269617705,
+      "grad_norm": 0.29604461789131165,
+      "learning_rate": 2.6803218145313392e-06,
+      "loss": 0.2908,
+      "step": 8209
+    },
+    {
+      "epoch": 2.0648893360160967,
+      "grad_norm": 0.3000899851322174,
+      "learning_rate": 2.6790256683039485e-06,
+      "loss": 0.314,
+      "step": 8210
+    },
+    {
+      "epoch": 2.0651408450704225,
+      "grad_norm": 0.3037608563899994,
+      "learning_rate": 2.6777297208534903e-06,
+      "loss": 0.3712,
+      "step": 8211
+    },
+    {
+      "epoch": 2.0653923541247483,
+      "grad_norm": 0.30880311131477356,
+      "learning_rate": 2.6764339722909523e-06,
+      "loss": 0.3349,
+      "step": 8212
+    },
+    {
+      "epoch": 2.0656438631790746,
+      "grad_norm": 0.3168122470378876,
+      "learning_rate": 2.67513842272731e-06,
+      "loss": 0.307,
+      "step": 8213
+    },
+    {
+      "epoch": 2.0658953722334004,
+      "grad_norm": 0.29152563214302063,
+      "learning_rate": 2.6738430722735155e-06,
+      "loss": 0.3272,
+      "step": 8214
+    },
+    {
+      "epoch": 2.066146881287726,
+      "grad_norm": 0.29437023401260376,
+      "learning_rate": 2.6725479210405114e-06,
+      "loss": 0.3171,
+      "step": 8215
+    },
+    {
+      "epoch": 2.0663983903420524,
+      "grad_norm": 0.2880735397338867,
+      "learning_rate": 2.671252969139216e-06,
+      "loss": 0.3026,
+      "step": 8216
+    },
+    {
+      "epoch": 2.066649899396378,
+      "grad_norm": 0.3284808099269867,
+      "learning_rate": 2.669958216680535e-06,
+      "loss": 0.3173,
+      "step": 8217
+    },
+    {
+      "epoch": 2.066901408450704,
+      "grad_norm": 0.30216220021247864,
+      "learning_rate": 2.668663663775357e-06,
+      "loss": 0.3389,
+      "step": 8218
+    },
+    {
+      "epoch": 2.0671529175050303,
+      "grad_norm": 0.3077835738658905,
+      "learning_rate": 2.6673693105345506e-06,
+      "loss": 0.318,
+      "step": 8219
+    },
+    {
+      "epoch": 2.067404426559356,
+      "grad_norm": 0.3287010192871094,
+      "learning_rate": 2.6660751570689715e-06,
+      "loss": 0.3402,
+      "step": 8220
+    },
+    {
+      "epoch": 2.067655935613682,
+      "grad_norm": 0.28192028403282166,
+      "learning_rate": 2.6647812034894516e-06,
+      "loss": 0.301,
+      "step": 8221
+    },
+    {
+      "epoch": 2.067907444668008,
+      "grad_norm": 0.3122316598892212,
+      "learning_rate": 2.6634874499068154e-06,
+      "loss": 0.3272,
+      "step": 8222
+    },
+    {
+      "epoch": 2.068158953722334,
+      "grad_norm": 0.2841907739639282,
+      "learning_rate": 2.6621938964318593e-06,
+      "loss": 0.3069,
+      "step": 8223
+    },
+    {
+      "epoch": 2.0684104627766597,
+      "grad_norm": 0.2972943186759949,
+      "learning_rate": 2.6609005431753733e-06,
+      "loss": 0.3083,
+      "step": 8224
+    },
+    {
+      "epoch": 2.068661971830986,
+      "grad_norm": 0.29851382970809937,
+      "learning_rate": 2.65960739024812e-06,
+      "loss": 0.3166,
+      "step": 8225
+    },
+    {
+      "epoch": 2.068913480885312,
+      "grad_norm": 0.32675349712371826,
+      "learning_rate": 2.658314437760855e-06,
+      "loss": 0.3267,
+      "step": 8226
+    },
+    {
+      "epoch": 2.0691649899396376,
+      "grad_norm": 0.2977888584136963,
+      "learning_rate": 2.6570216858243057e-06,
+      "loss": 0.3502,
+      "step": 8227
+    },
+    {
+      "epoch": 2.069416498993964,
+      "grad_norm": 0.30917125940322876,
+      "learning_rate": 2.655729134549192e-06,
+      "loss": 0.3134,
+      "step": 8228
+    },
+    {
+      "epoch": 2.0696680080482897,
+      "grad_norm": 0.3009907603263855,
+      "learning_rate": 2.654436784046214e-06,
+      "loss": 0.3215,
+      "step": 8229
+    },
+    {
+      "epoch": 2.069919517102616,
+      "grad_norm": 0.29445791244506836,
+      "learning_rate": 2.6531446344260503e-06,
+      "loss": 0.3027,
+      "step": 8230
+    },
+    {
+      "epoch": 2.0701710261569417,
+      "grad_norm": 0.2821619212627411,
+      "learning_rate": 2.651852685799368e-06,
+      "loss": 0.2985,
+      "step": 8231
+    },
+    {
+      "epoch": 2.0704225352112675,
+      "grad_norm": 0.28280389308929443,
+      "learning_rate": 2.6505609382768117e-06,
+      "loss": 0.3142,
+      "step": 8232
+    },
+    {
+      "epoch": 2.0706740442655938,
+      "grad_norm": 0.3183631896972656,
+      "learning_rate": 2.649269391969015e-06,
+      "loss": 0.3409,
+      "step": 8233
+    },
+    {
+      "epoch": 2.0709255533199196,
+      "grad_norm": 0.30372148752212524,
+      "learning_rate": 2.6479780469865864e-06,
+      "loss": 0.3102,
+      "step": 8234
+    },
+    {
+      "epoch": 2.0711770623742454,
+      "grad_norm": 0.29154670238494873,
+      "learning_rate": 2.646686903440126e-06,
+      "loss": 0.3123,
+      "step": 8235
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.30079391598701477,
+      "learning_rate": 2.645395961440208e-06,
+      "loss": 0.3023,
+      "step": 8236
+    },
+    {
+      "epoch": 2.0716800804828974,
+      "grad_norm": 0.304799348115921,
+      "learning_rate": 2.6441052210973974e-06,
+      "loss": 0.3398,
+      "step": 8237
+    },
+    {
+      "epoch": 2.0719315895372232,
+      "grad_norm": 0.2910962998867035,
+      "learning_rate": 2.6428146825222344e-06,
+      "loss": 0.3074,
+      "step": 8238
+    },
+    {
+      "epoch": 2.0721830985915495,
+      "grad_norm": 0.28885748982429504,
+      "learning_rate": 2.641524345825248e-06,
+      "loss": 0.3215,
+      "step": 8239
+    },
+    {
+      "epoch": 2.0724346076458753,
+      "grad_norm": 0.28982120752334595,
+      "learning_rate": 2.6402342111169476e-06,
+      "loss": 0.324,
+      "step": 8240
+    },
+    {
+      "epoch": 2.072686116700201,
+      "grad_norm": 0.30088311433792114,
+      "learning_rate": 2.6389442785078227e-06,
+      "loss": 0.325,
+      "step": 8241
+    },
+    {
+      "epoch": 2.0729376257545273,
+      "grad_norm": 0.30647531151771545,
+      "learning_rate": 2.637654548108352e-06,
+      "loss": 0.3286,
+      "step": 8242
+    },
+    {
+      "epoch": 2.073189134808853,
+      "grad_norm": 0.3003426492214203,
+      "learning_rate": 2.636365020028988e-06,
+      "loss": 0.3363,
+      "step": 8243
+    },
+    {
+      "epoch": 2.073440643863179,
+      "grad_norm": 0.30914726853370667,
+      "learning_rate": 2.635075694380176e-06,
+      "loss": 0.3139,
+      "step": 8244
+    },
+    {
+      "epoch": 2.073692152917505,
+      "grad_norm": 0.3180095851421356,
+      "learning_rate": 2.633786571272333e-06,
+      "loss": 0.3075,
+      "step": 8245
+    },
+    {
+      "epoch": 2.073943661971831,
+      "grad_norm": 0.31122735142707825,
+      "learning_rate": 2.6324976508158697e-06,
+      "loss": 0.3222,
+      "step": 8246
+    },
+    {
+      "epoch": 2.074195171026157,
+      "grad_norm": 0.30536743998527527,
+      "learning_rate": 2.6312089331211693e-06,
+      "loss": 0.3163,
+      "step": 8247
+    },
+    {
+      "epoch": 2.074446680080483,
+      "grad_norm": 0.32808443903923035,
+      "learning_rate": 2.6299204182986072e-06,
+      "loss": 0.3442,
+      "step": 8248
+    },
+    {
+      "epoch": 2.074698189134809,
+      "grad_norm": 0.2878994941711426,
+      "learning_rate": 2.6286321064585315e-06,
+      "loss": 0.3282,
+      "step": 8249
+    },
+    {
+      "epoch": 2.0749496981891347,
+      "grad_norm": 0.303153395652771,
+      "learning_rate": 2.6273439977112803e-06,
+      "loss": 0.3128,
+      "step": 8250
+    },
+    {
+      "epoch": 2.075201207243461,
+      "grad_norm": 0.29553067684173584,
+      "learning_rate": 2.626056092167175e-06,
+      "loss": 0.3316,
+      "step": 8251
+    },
+    {
+      "epoch": 2.0754527162977867,
+      "grad_norm": 0.3366231918334961,
+      "learning_rate": 2.6247683899365117e-06,
+      "loss": 0.3172,
+      "step": 8252
+    },
+    {
+      "epoch": 2.0757042253521125,
+      "grad_norm": 0.33030417561531067,
+      "learning_rate": 2.623480891129579e-06,
+      "loss": 0.3245,
+      "step": 8253
+    },
+    {
+      "epoch": 2.0759557344064388,
+      "grad_norm": 0.30303242802619934,
+      "learning_rate": 2.622193595856638e-06,
+      "loss": 0.3321,
+      "step": 8254
+    },
+    {
+      "epoch": 2.0762072434607646,
+      "grad_norm": 0.29723554849624634,
+      "learning_rate": 2.6209065042279426e-06,
+      "loss": 0.3069,
+      "step": 8255
+    },
+    {
+      "epoch": 2.0764587525150904,
+      "grad_norm": 0.294047087430954,
+      "learning_rate": 2.619619616353719e-06,
+      "loss": 0.3519,
+      "step": 8256
+    },
+    {
+      "epoch": 2.0767102615694166,
+      "grad_norm": 0.31295087933540344,
+      "learning_rate": 2.618332932344185e-06,
+      "loss": 0.3249,
+      "step": 8257
+    },
+    {
+      "epoch": 2.0769617706237424,
+      "grad_norm": 0.3490563929080963,
+      "learning_rate": 2.617046452309535e-06,
+      "loss": 0.326,
+      "step": 8258
+    },
+    {
+      "epoch": 2.0772132796780682,
+      "grad_norm": 0.30121055245399475,
+      "learning_rate": 2.6157601763599504e-06,
+      "loss": 0.3128,
+      "step": 8259
+    },
+    {
+      "epoch": 2.0774647887323945,
+      "grad_norm": 0.29791775345802307,
+      "learning_rate": 2.614474104605589e-06,
+      "loss": 0.3224,
+      "step": 8260
+    },
+    {
+      "epoch": 2.0777162977867203,
+      "grad_norm": 0.3081105947494507,
+      "learning_rate": 2.613188237156596e-06,
+      "loss": 0.3093,
+      "step": 8261
+    },
+    {
+      "epoch": 2.077967806841046,
+      "grad_norm": 0.30400124192237854,
+      "learning_rate": 2.6119025741231007e-06,
+      "loss": 0.3172,
+      "step": 8262
+    },
+    {
+      "epoch": 2.0782193158953723,
+      "grad_norm": 0.3303559422492981,
+      "learning_rate": 2.610617115615208e-06,
+      "loss": 0.3163,
+      "step": 8263
+    },
+    {
+      "epoch": 2.078470824949698,
+      "grad_norm": 0.2974235415458679,
+      "learning_rate": 2.609331861743014e-06,
+      "loss": 0.3333,
+      "step": 8264
+    },
+    {
+      "epoch": 2.078722334004024,
+      "grad_norm": 0.32158637046813965,
+      "learning_rate": 2.608046812616588e-06,
+      "loss": 0.328,
+      "step": 8265
+    },
+    {
+      "epoch": 2.07897384305835,
+      "grad_norm": 0.3166216313838959,
+      "learning_rate": 2.6067619683459904e-06,
+      "loss": 0.317,
+      "step": 8266
+    },
+    {
+      "epoch": 2.079225352112676,
+      "grad_norm": 0.2718227803707123,
+      "learning_rate": 2.605477329041256e-06,
+      "loss": 0.3305,
+      "step": 8267
+    },
+    {
+      "epoch": 2.079476861167002,
+      "grad_norm": 0.33164918422698975,
+      "learning_rate": 2.6041928948124107e-06,
+      "loss": 0.322,
+      "step": 8268
+    },
+    {
+      "epoch": 2.079728370221328,
+      "grad_norm": 0.3085431158542633,
+      "learning_rate": 2.6029086657694537e-06,
+      "loss": 0.3276,
+      "step": 8269
+    },
+    {
+      "epoch": 2.079979879275654,
+      "grad_norm": 0.30075356364250183,
+      "learning_rate": 2.6016246420223744e-06,
+      "loss": 0.325,
+      "step": 8270
+    },
+    {
+      "epoch": 2.0802313883299797,
+      "grad_norm": 0.3089926540851593,
+      "learning_rate": 2.600340823681139e-06,
+      "loss": 0.3003,
+      "step": 8271
+    },
+    {
+      "epoch": 2.080482897384306,
+      "grad_norm": 0.3290434777736664,
+      "learning_rate": 2.5990572108557e-06,
+      "loss": 0.3306,
+      "step": 8272
+    },
+    {
+      "epoch": 2.0807344064386317,
+      "grad_norm": 0.3045661747455597,
+      "learning_rate": 2.597773803655993e-06,
+      "loss": 0.3234,
+      "step": 8273
+    },
+    {
+      "epoch": 2.080985915492958,
+      "grad_norm": 0.2929016053676605,
+      "learning_rate": 2.596490602191929e-06,
+      "loss": 0.312,
+      "step": 8274
+    },
+    {
+      "epoch": 2.0812374245472838,
+      "grad_norm": 0.2901984453201294,
+      "learning_rate": 2.59520760657341e-06,
+      "loss": 0.3147,
+      "step": 8275
+    },
+    {
+      "epoch": 2.0814889336016096,
+      "grad_norm": 0.33070576190948486,
+      "learning_rate": 2.5939248169103136e-06,
+      "loss": 0.3309,
+      "step": 8276
+    },
+    {
+      "epoch": 2.081740442655936,
+      "grad_norm": 0.3062666654586792,
+      "learning_rate": 2.5926422333125066e-06,
+      "loss": 0.3098,
+      "step": 8277
+    },
+    {
+      "epoch": 2.0819919517102616,
+      "grad_norm": 0.31025469303131104,
+      "learning_rate": 2.59135985588983e-06,
+      "loss": 0.3265,
+      "step": 8278
+    },
+    {
+      "epoch": 2.0822434607645874,
+      "grad_norm": 0.29651591181755066,
+      "learning_rate": 2.5900776847521148e-06,
+      "loss": 0.3186,
+      "step": 8279
+    },
+    {
+      "epoch": 2.0824949698189137,
+      "grad_norm": 0.31347307562828064,
+      "learning_rate": 2.588795720009168e-06,
+      "loss": 0.3182,
+      "step": 8280
+    },
+    {
+      "epoch": 2.0827464788732395,
+      "grad_norm": 0.3125113546848297,
+      "learning_rate": 2.587513961770785e-06,
+      "loss": 0.339,
+      "step": 8281
+    },
+    {
+      "epoch": 2.0829979879275653,
+      "grad_norm": 0.3104122281074524,
+      "learning_rate": 2.586232410146737e-06,
+      "loss": 0.3019,
+      "step": 8282
+    },
+    {
+      "epoch": 2.0832494969818915,
+      "grad_norm": 0.7815317511558533,
+      "learning_rate": 2.584951065246784e-06,
+      "loss": 0.3071,
+      "step": 8283
+    },
+    {
+      "epoch": 2.0835010060362174,
+      "grad_norm": 0.32731983065605164,
+      "learning_rate": 2.583669927180662e-06,
+      "loss": 0.3127,
+      "step": 8284
+    },
+    {
+      "epoch": 2.083752515090543,
+      "grad_norm": 0.3034411668777466,
+      "learning_rate": 2.5823889960580967e-06,
+      "loss": 0.3335,
+      "step": 8285
+    },
+    {
+      "epoch": 2.0840040241448694,
+      "grad_norm": 0.30236032605171204,
+      "learning_rate": 2.581108271988787e-06,
+      "loss": 0.3237,
+      "step": 8286
+    },
+    {
+      "epoch": 2.084255533199195,
+      "grad_norm": 0.3101537525653839,
+      "learning_rate": 2.5798277550824238e-06,
+      "loss": 0.2826,
+      "step": 8287
+    },
+    {
+      "epoch": 2.084507042253521,
+      "grad_norm": 0.2881515324115753,
+      "learning_rate": 2.5785474454486696e-06,
+      "loss": 0.3407,
+      "step": 8288
+    },
+    {
+      "epoch": 2.0847585513078473,
+      "grad_norm": 0.31731492280960083,
+      "learning_rate": 2.5772673431971805e-06,
+      "loss": 0.3017,
+      "step": 8289
+    },
+    {
+      "epoch": 2.085010060362173,
+      "grad_norm": 0.29667213559150696,
+      "learning_rate": 2.575987448437586e-06,
+      "loss": 0.3199,
+      "step": 8290
+    },
+    {
+      "epoch": 2.085261569416499,
+      "grad_norm": 0.29330283403396606,
+      "learning_rate": 2.574707761279503e-06,
+      "loss": 0.3388,
+      "step": 8291
+    },
+    {
+      "epoch": 2.085513078470825,
+      "grad_norm": 0.2838901877403259,
+      "learning_rate": 2.5734282818325256e-06,
+      "loss": 0.3097,
+      "step": 8292
+    },
+    {
+      "epoch": 2.085764587525151,
+      "grad_norm": 0.3035966753959656,
+      "learning_rate": 2.5721490102062373e-06,
+      "loss": 0.3277,
+      "step": 8293
+    },
+    {
+      "epoch": 2.0860160965794767,
+      "grad_norm": 0.32544568181037903,
+      "learning_rate": 2.570869946510196e-06,
+      "loss": 0.347,
+      "step": 8294
+    },
+    {
+      "epoch": 2.086267605633803,
+      "grad_norm": 0.30072399973869324,
+      "learning_rate": 2.5695910908539494e-06,
+      "loss": 0.3299,
+      "step": 8295
+    },
+    {
+      "epoch": 2.086519114688129,
+      "grad_norm": 0.3020775318145752,
+      "learning_rate": 2.568312443347019e-06,
+      "loss": 0.3178,
+      "step": 8296
+    },
+    {
+      "epoch": 2.0867706237424546,
+      "grad_norm": 0.3153866231441498,
+      "learning_rate": 2.567034004098917e-06,
+      "loss": 0.3386,
+      "step": 8297
+    },
+    {
+      "epoch": 2.087022132796781,
+      "grad_norm": 0.29892367124557495,
+      "learning_rate": 2.565755773219131e-06,
+      "loss": 0.3268,
+      "step": 8298
+    },
+    {
+      "epoch": 2.0872736418511066,
+      "grad_norm": 0.30764761567115784,
+      "learning_rate": 2.564477750817135e-06,
+      "loss": 0.3481,
+      "step": 8299
+    },
+    {
+      "epoch": 2.0875251509054324,
+      "grad_norm": 0.2927669882774353,
+      "learning_rate": 2.563199937002382e-06,
+      "loss": 0.3064,
+      "step": 8300
+    },
+    {
+      "epoch": 2.0877766599597587,
+      "grad_norm": 0.30873027443885803,
+      "learning_rate": 2.561922331884311e-06,
+      "loss": 0.3375,
+      "step": 8301
+    },
+    {
+      "epoch": 2.0880281690140845,
+      "grad_norm": 0.30623045563697815,
+      "learning_rate": 2.560644935572338e-06,
+      "loss": 0.3257,
+      "step": 8302
+    },
+    {
+      "epoch": 2.0882796780684103,
+      "grad_norm": 0.29799333214759827,
+      "learning_rate": 2.559367748175867e-06,
+      "loss": 0.3113,
+      "step": 8303
+    },
+    {
+      "epoch": 2.0885311871227366,
+      "grad_norm": 0.31221136450767517,
+      "learning_rate": 2.5580907698042802e-06,
+      "loss": 0.3235,
+      "step": 8304
+    },
+    {
+      "epoch": 2.0887826961770624,
+      "grad_norm": 0.30643758177757263,
+      "learning_rate": 2.5568140005669414e-06,
+      "loss": 0.3348,
+      "step": 8305
+    },
+    {
+      "epoch": 2.089034205231388,
+      "grad_norm": 0.3108401894569397,
+      "learning_rate": 2.5555374405732e-06,
+      "loss": 0.3332,
+      "step": 8306
+    },
+    {
+      "epoch": 2.0892857142857144,
+      "grad_norm": 0.31820324063301086,
+      "learning_rate": 2.5542610899323826e-06,
+      "loss": 0.3098,
+      "step": 8307
+    },
+    {
+      "epoch": 2.08953722334004,
+      "grad_norm": 0.30521127581596375,
+      "learning_rate": 2.552984948753805e-06,
+      "loss": 0.3333,
+      "step": 8308
+    },
+    {
+      "epoch": 2.089788732394366,
+      "grad_norm": 0.304718941450119,
+      "learning_rate": 2.5517090171467557e-06,
+      "loss": 0.3161,
+      "step": 8309
+    },
+    {
+      "epoch": 2.0900402414486923,
+      "grad_norm": 0.33663034439086914,
+      "learning_rate": 2.550433295220515e-06,
+      "loss": 0.3282,
+      "step": 8310
+    },
+    {
+      "epoch": 2.090291750503018,
+      "grad_norm": 0.3158341646194458,
+      "learning_rate": 2.549157783084335e-06,
+      "loss": 0.3157,
+      "step": 8311
+    },
+    {
+      "epoch": 2.090543259557344,
+      "grad_norm": 0.3130324184894562,
+      "learning_rate": 2.5478824808474613e-06,
+      "loss": 0.3192,
+      "step": 8312
+    },
+    {
+      "epoch": 2.09079476861167,
+      "grad_norm": 0.3025255799293518,
+      "learning_rate": 2.546607388619111e-06,
+      "loss": 0.3188,
+      "step": 8313
+    },
+    {
+      "epoch": 2.091046277665996,
+      "grad_norm": 0.3400377333164215,
+      "learning_rate": 2.5453325065084887e-06,
+      "loss": 0.3063,
+      "step": 8314
+    },
+    {
+      "epoch": 2.0912977867203217,
+      "grad_norm": 0.300741046667099,
+      "learning_rate": 2.5440578346247834e-06,
+      "loss": 0.3058,
+      "step": 8315
+    },
+    {
+      "epoch": 2.091549295774648,
+      "grad_norm": 0.31331273913383484,
+      "learning_rate": 2.5427833730771577e-06,
+      "loss": 0.3098,
+      "step": 8316
+    },
+    {
+      "epoch": 2.091800804828974,
+      "grad_norm": 0.2970178425312042,
+      "learning_rate": 2.541509121974766e-06,
+      "loss": 0.3035,
+      "step": 8317
+    },
+    {
+      "epoch": 2.0920523138832996,
+      "grad_norm": 0.31109151244163513,
+      "learning_rate": 2.5402350814267364e-06,
+      "loss": 0.3272,
+      "step": 8318
+    },
+    {
+      "epoch": 2.092303822937626,
+      "grad_norm": 0.29535481333732605,
+      "learning_rate": 2.538961251542185e-06,
+      "loss": 0.3169,
+      "step": 8319
+    },
+    {
+      "epoch": 2.0925553319919517,
+      "grad_norm": 0.30848875641822815,
+      "learning_rate": 2.5376876324302045e-06,
+      "loss": 0.3327,
+      "step": 8320
+    },
+    {
+      "epoch": 2.0928068410462775,
+      "grad_norm": 0.28316324949264526,
+      "learning_rate": 2.5364142241998755e-06,
+      "loss": 0.3399,
+      "step": 8321
+    },
+    {
+      "epoch": 2.0930583501006037,
+      "grad_norm": 0.3024502992630005,
+      "learning_rate": 2.535141026960255e-06,
+      "loss": 0.3063,
+      "step": 8322
+    },
+    {
+      "epoch": 2.0933098591549295,
+      "grad_norm": 0.33002567291259766,
+      "learning_rate": 2.5338680408203875e-06,
+      "loss": 0.3255,
+      "step": 8323
+    },
+    {
+      "epoch": 2.0935613682092553,
+      "grad_norm": 0.2928857207298279,
+      "learning_rate": 2.5325952658892916e-06,
+      "loss": 0.3213,
+      "step": 8324
+    },
+    {
+      "epoch": 2.0938128772635816,
+      "grad_norm": 0.30271077156066895,
+      "learning_rate": 2.531322702275976e-06,
+      "loss": 0.3324,
+      "step": 8325
+    },
+    {
+      "epoch": 2.0940643863179074,
+      "grad_norm": 0.2813225984573364,
+      "learning_rate": 2.530050350089428e-06,
+      "loss": 0.3231,
+      "step": 8326
+    },
+    {
+      "epoch": 2.094315895372233,
+      "grad_norm": 0.3095519542694092,
+      "learning_rate": 2.528778209438614e-06,
+      "loss": 0.3165,
+      "step": 8327
+    },
+    {
+      "epoch": 2.0945674044265594,
+      "grad_norm": 0.30373334884643555,
+      "learning_rate": 2.527506280432488e-06,
+      "loss": 0.3349,
+      "step": 8328
+    },
+    {
+      "epoch": 2.0948189134808852,
+      "grad_norm": 0.30294981598854065,
+      "learning_rate": 2.5262345631799794e-06,
+      "loss": 0.3178,
+      "step": 8329
+    },
+    {
+      "epoch": 2.0950704225352115,
+      "grad_norm": 0.3043777644634247,
+      "learning_rate": 2.524963057790007e-06,
+      "loss": 0.3115,
+      "step": 8330
+    },
+    {
+      "epoch": 2.0953219315895373,
+      "grad_norm": 0.288493275642395,
+      "learning_rate": 2.5236917643714628e-06,
+      "loss": 0.3383,
+      "step": 8331
+    },
+    {
+      "epoch": 2.095573440643863,
+      "grad_norm": 0.29598721861839294,
+      "learning_rate": 2.5224206830332286e-06,
+      "loss": 0.3202,
+      "step": 8332
+    },
+    {
+      "epoch": 2.0958249496981893,
+      "grad_norm": 0.299293577671051,
+      "learning_rate": 2.521149813884162e-06,
+      "loss": 0.3185,
+      "step": 8333
+    },
+    {
+      "epoch": 2.096076458752515,
+      "grad_norm": 0.2833903729915619,
+      "learning_rate": 2.5198791570331083e-06,
+      "loss": 0.3333,
+      "step": 8334
+    },
+    {
+      "epoch": 2.096327967806841,
+      "grad_norm": 0.33051058650016785,
+      "learning_rate": 2.5186087125888863e-06,
+      "loss": 0.3338,
+      "step": 8335
+    },
+    {
+      "epoch": 2.096579476861167,
+      "grad_norm": 0.293345183134079,
+      "learning_rate": 2.5173384806603052e-06,
+      "loss": 0.3244,
+      "step": 8336
+    },
+    {
+      "epoch": 2.096830985915493,
+      "grad_norm": 0.2857302725315094,
+      "learning_rate": 2.516068461356154e-06,
+      "loss": 0.2933,
+      "step": 8337
+    },
+    {
+      "epoch": 2.097082494969819,
+      "grad_norm": 0.3116956949234009,
+      "learning_rate": 2.514798654785197e-06,
+      "loss": 0.315,
+      "step": 8338
+    },
+    {
+      "epoch": 2.097334004024145,
+      "grad_norm": 0.278083473443985,
+      "learning_rate": 2.51352906105619e-06,
+      "loss": 0.3115,
+      "step": 8339
+    },
+    {
+      "epoch": 2.097585513078471,
+      "grad_norm": 0.30503949522972107,
+      "learning_rate": 2.512259680277862e-06,
+      "loss": 0.3144,
+      "step": 8340
+    },
+    {
+      "epoch": 2.0978370221327967,
+      "grad_norm": 0.3300677537918091,
+      "learning_rate": 2.510990512558931e-06,
+      "loss": 0.3183,
+      "step": 8341
+    },
+    {
+      "epoch": 2.098088531187123,
+      "grad_norm": 0.3177797496318817,
+      "learning_rate": 2.509721558008089e-06,
+      "loss": 0.3577,
+      "step": 8342
+    },
+    {
+      "epoch": 2.0983400402414487,
+      "grad_norm": 0.3060501515865326,
+      "learning_rate": 2.508452816734019e-06,
+      "loss": 0.3328,
+      "step": 8343
+    },
+    {
+      "epoch": 2.0985915492957745,
+      "grad_norm": 0.3117055594921112,
+      "learning_rate": 2.507184288845376e-06,
+      "loss": 0.314,
+      "step": 8344
+    },
+    {
+      "epoch": 2.0988430583501008,
+      "grad_norm": 0.31024450063705444,
+      "learning_rate": 2.5059159744508055e-06,
+      "loss": 0.2982,
+      "step": 8345
+    },
+    {
+      "epoch": 2.0990945674044266,
+      "grad_norm": 0.33131489157676697,
+      "learning_rate": 2.5046478736589264e-06,
+      "loss": 0.3093,
+      "step": 8346
+    },
+    {
+      "epoch": 2.0993460764587524,
+      "grad_norm": 0.3035459816455841,
+      "learning_rate": 2.503379986578347e-06,
+      "loss": 0.321,
+      "step": 8347
+    },
+    {
+      "epoch": 2.0995975855130786,
+      "grad_norm": 0.3351083993911743,
+      "learning_rate": 2.502112313317654e-06,
+      "loss": 0.3098,
+      "step": 8348
+    },
+    {
+      "epoch": 2.0998490945674044,
+      "grad_norm": 0.3084772825241089,
+      "learning_rate": 2.5008448539854134e-06,
+      "loss": 0.3327,
+      "step": 8349
+    },
+    {
+      "epoch": 2.1001006036217302,
+      "grad_norm": 0.3130403459072113,
+      "learning_rate": 2.499577608690178e-06,
+      "loss": 0.3342,
+      "step": 8350
+    },
+    {
+      "epoch": 2.1003521126760565,
+      "grad_norm": 0.32867324352264404,
+      "learning_rate": 2.498310577540476e-06,
+      "loss": 0.3216,
+      "step": 8351
+    },
+    {
+      "epoch": 2.1006036217303823,
+      "grad_norm": 0.3281696140766144,
+      "learning_rate": 2.4970437606448245e-06,
+      "loss": 0.3087,
+      "step": 8352
+    },
+    {
+      "epoch": 2.100855130784708,
+      "grad_norm": 0.3127487897872925,
+      "learning_rate": 2.495777158111714e-06,
+      "loss": 0.3194,
+      "step": 8353
+    },
+    {
+      "epoch": 2.1011066398390343,
+      "grad_norm": 0.29324328899383545,
+      "learning_rate": 2.4945107700496263e-06,
+      "loss": 0.3323,
+      "step": 8354
+    },
+    {
+      "epoch": 2.10135814889336,
+      "grad_norm": 0.27764442563056946,
+      "learning_rate": 2.4932445965670145e-06,
+      "loss": 0.2898,
+      "step": 8355
+    },
+    {
+      "epoch": 2.101609657947686,
+      "grad_norm": 0.3088850677013397,
+      "learning_rate": 2.4919786377723225e-06,
+      "loss": 0.3096,
+      "step": 8356
+    },
+    {
+      "epoch": 2.101861167002012,
+      "grad_norm": 0.29290542006492615,
+      "learning_rate": 2.490712893773968e-06,
+      "loss": 0.3038,
+      "step": 8357
+    },
+    {
+      "epoch": 2.102112676056338,
+      "grad_norm": 0.30638036131858826,
+      "learning_rate": 2.489447364680357e-06,
+      "loss": 0.3554,
+      "step": 8358
+    },
+    {
+      "epoch": 2.102364185110664,
+      "grad_norm": 0.3109561800956726,
+      "learning_rate": 2.4881820505998743e-06,
+      "loss": 0.319,
+      "step": 8359
+    },
+    {
+      "epoch": 2.10261569416499,
+      "grad_norm": 0.2938198745250702,
+      "learning_rate": 2.486916951640884e-06,
+      "loss": 0.319,
+      "step": 8360
+    },
+    {
+      "epoch": 2.102867203219316,
+      "grad_norm": 0.32792484760284424,
+      "learning_rate": 2.4856520679117357e-06,
+      "loss": 0.3347,
+      "step": 8361
+    },
+    {
+      "epoch": 2.1031187122736417,
+      "grad_norm": 0.3298247158527374,
+      "learning_rate": 2.4843873995207567e-06,
+      "loss": 0.3105,
+      "step": 8362
+    },
+    {
+      "epoch": 2.103370221327968,
+      "grad_norm": 0.32708612084388733,
+      "learning_rate": 2.483122946576262e-06,
+      "loss": 0.3043,
+      "step": 8363
+    },
+    {
+      "epoch": 2.1036217303822937,
+      "grad_norm": 0.31453192234039307,
+      "learning_rate": 2.4818587091865386e-06,
+      "loss": 0.3036,
+      "step": 8364
+    },
+    {
+      "epoch": 2.1038732394366195,
+      "grad_norm": 0.2795354425907135,
+      "learning_rate": 2.480594687459865e-06,
+      "loss": 0.3167,
+      "step": 8365
+    },
+    {
+      "epoch": 2.1041247484909458,
+      "grad_norm": 0.29401612281799316,
+      "learning_rate": 2.4793308815044943e-06,
+      "loss": 0.3266,
+      "step": 8366
+    },
+    {
+      "epoch": 2.1043762575452716,
+      "grad_norm": 0.3023243844509125,
+      "learning_rate": 2.4780672914286652e-06,
+      "loss": 0.322,
+      "step": 8367
+    },
+    {
+      "epoch": 2.1046277665995974,
+      "grad_norm": 0.3229201138019562,
+      "learning_rate": 2.476803917340594e-06,
+      "loss": 0.3315,
+      "step": 8368
+    },
+    {
+      "epoch": 2.1048792756539236,
+      "grad_norm": 0.3104352355003357,
+      "learning_rate": 2.475540759348482e-06,
+      "loss": 0.3132,
+      "step": 8369
+    },
+    {
+      "epoch": 2.1051307847082494,
+      "grad_norm": 0.30430683493614197,
+      "learning_rate": 2.474277817560513e-06,
+      "loss": 0.3303,
+      "step": 8370
+    },
+    {
+      "epoch": 2.1053822937625752,
+      "grad_norm": 0.296283483505249,
+      "learning_rate": 2.473015092084846e-06,
+      "loss": 0.3168,
+      "step": 8371
+    },
+    {
+      "epoch": 2.1056338028169015,
+      "grad_norm": 0.31060439348220825,
+      "learning_rate": 2.4717525830296295e-06,
+      "loss": 0.3165,
+      "step": 8372
+    },
+    {
+      "epoch": 2.1058853118712273,
+      "grad_norm": 0.3210805654525757,
+      "learning_rate": 2.470490290502986e-06,
+      "loss": 0.3222,
+      "step": 8373
+    },
+    {
+      "epoch": 2.1061368209255535,
+      "grad_norm": 0.3033636212348938,
+      "learning_rate": 2.4692282146130266e-06,
+      "loss": 0.3395,
+      "step": 8374
+    },
+    {
+      "epoch": 2.1063883299798793,
+      "grad_norm": 0.3115089535713196,
+      "learning_rate": 2.4679663554678357e-06,
+      "loss": 0.3245,
+      "step": 8375
+    },
+    {
+      "epoch": 2.106639839034205,
+      "grad_norm": 0.319614976644516,
+      "learning_rate": 2.4667047131754884e-06,
+      "loss": 0.3062,
+      "step": 8376
+    },
+    {
+      "epoch": 2.1068913480885314,
+      "grad_norm": 0.30483153462409973,
+      "learning_rate": 2.465443287844032e-06,
+      "loss": 0.3157,
+      "step": 8377
+    },
+    {
+      "epoch": 2.107142857142857,
+      "grad_norm": 0.30842599272727966,
+      "learning_rate": 2.464182079581504e-06,
+      "loss": 0.3497,
+      "step": 8378
+    },
+    {
+      "epoch": 2.107394366197183,
+      "grad_norm": 0.30280083417892456,
+      "learning_rate": 2.462921088495915e-06,
+      "loss": 0.3356,
+      "step": 8379
+    },
+    {
+      "epoch": 2.1076458752515093,
+      "grad_norm": 0.2851159870624542,
+      "learning_rate": 2.4616603146952628e-06,
+      "loss": 0.3035,
+      "step": 8380
+    },
+    {
+      "epoch": 2.107897384305835,
+      "grad_norm": 0.3186687231063843,
+      "learning_rate": 2.4603997582875266e-06,
+      "loss": 0.322,
+      "step": 8381
+    },
+    {
+      "epoch": 2.108148893360161,
+      "grad_norm": 0.2962452471256256,
+      "learning_rate": 2.4591394193806615e-06,
+      "loss": 0.3278,
+      "step": 8382
+    },
+    {
+      "epoch": 2.108400402414487,
+      "grad_norm": 0.29877567291259766,
+      "learning_rate": 2.4578792980826114e-06,
+      "loss": 0.3511,
+      "step": 8383
+    },
+    {
+      "epoch": 2.108651911468813,
+      "grad_norm": 0.2987919747829437,
+      "learning_rate": 2.4566193945012946e-06,
+      "loss": 0.3165,
+      "step": 8384
+    },
+    {
+      "epoch": 2.1089034205231387,
+      "grad_norm": 0.28243282437324524,
+      "learning_rate": 2.4553597087446163e-06,
+      "loss": 0.3505,
+      "step": 8385
+    },
+    {
+      "epoch": 2.109154929577465,
+      "grad_norm": 0.31294065713882446,
+      "learning_rate": 2.4541002409204584e-06,
+      "loss": 0.3179,
+      "step": 8386
+    },
+    {
+      "epoch": 2.109406438631791,
+      "grad_norm": 0.2762846052646637,
+      "learning_rate": 2.4528409911366897e-06,
+      "loss": 0.3258,
+      "step": 8387
+    },
+    {
+      "epoch": 2.1096579476861166,
+      "grad_norm": 0.28698083758354187,
+      "learning_rate": 2.4515819595011532e-06,
+      "loss": 0.3219,
+      "step": 8388
+    },
+    {
+      "epoch": 2.109909456740443,
+      "grad_norm": 0.32928720116615295,
+      "learning_rate": 2.450323146121681e-06,
+      "loss": 0.3351,
+      "step": 8389
+    },
+    {
+      "epoch": 2.1101609657947686,
+      "grad_norm": 0.30943191051483154,
+      "learning_rate": 2.4490645511060784e-06,
+      "loss": 0.3292,
+      "step": 8390
+    },
+    {
+      "epoch": 2.1104124748490944,
+      "grad_norm": 0.2983676493167877,
+      "learning_rate": 2.4478061745621383e-06,
+      "loss": 0.3354,
+      "step": 8391
+    },
+    {
+      "epoch": 2.1106639839034207,
+      "grad_norm": 0.29229018092155457,
+      "learning_rate": 2.446548016597635e-06,
+      "loss": 0.3172,
+      "step": 8392
+    },
+    {
+      "epoch": 2.1109154929577465,
+      "grad_norm": 0.30713313817977905,
+      "learning_rate": 2.4452900773203182e-06,
+      "loss": 0.3452,
+      "step": 8393
+    },
+    {
+      "epoch": 2.1111670020120723,
+      "grad_norm": 0.3006666600704193,
+      "learning_rate": 2.4440323568379255e-06,
+      "loss": 0.2936,
+      "step": 8394
+    },
+    {
+      "epoch": 2.1114185110663986,
+      "grad_norm": 0.30126234889030457,
+      "learning_rate": 2.4427748552581694e-06,
+      "loss": 0.3048,
+      "step": 8395
+    },
+    {
+      "epoch": 2.1116700201207244,
+      "grad_norm": 0.2818053960800171,
+      "learning_rate": 2.4415175726887513e-06,
+      "loss": 0.3103,
+      "step": 8396
+    },
+    {
+      "epoch": 2.11192152917505,
+      "grad_norm": 0.2906033992767334,
+      "learning_rate": 2.440260509237345e-06,
+      "loss": 0.3261,
+      "step": 8397
+    },
+    {
+      "epoch": 2.1121730382293764,
+      "grad_norm": 0.28882506489753723,
+      "learning_rate": 2.4390036650116144e-06,
+      "loss": 0.3137,
+      "step": 8398
+    },
+    {
+      "epoch": 2.112424547283702,
+      "grad_norm": 0.31607574224472046,
+      "learning_rate": 2.4377470401191965e-06,
+      "loss": 0.3077,
+      "step": 8399
+    },
+    {
+      "epoch": 2.112676056338028,
+      "grad_norm": 0.3000749945640564,
+      "learning_rate": 2.436490634667717e-06,
+      "loss": 0.3225,
+      "step": 8400
+    },
+    {
+      "epoch": 2.1129275653923543,
+      "grad_norm": 0.29554128646850586,
+      "learning_rate": 2.4352344487647755e-06,
+      "loss": 0.3399,
+      "step": 8401
+    },
+    {
+      "epoch": 2.11317907444668,
+      "grad_norm": 0.308362752199173,
+      "learning_rate": 2.4339784825179606e-06,
+      "loss": 0.3279,
+      "step": 8402
+    },
+    {
+      "epoch": 2.113430583501006,
+      "grad_norm": 0.28427183628082275,
+      "learning_rate": 2.4327227360348333e-06,
+      "loss": 0.3256,
+      "step": 8403
+    },
+    {
+      "epoch": 2.113682092555332,
+      "grad_norm": 0.3028574287891388,
+      "learning_rate": 2.431467209422945e-06,
+      "loss": 0.3003,
+      "step": 8404
+    },
+    {
+      "epoch": 2.113933601609658,
+      "grad_norm": 0.32093802094459534,
+      "learning_rate": 2.4302119027898195e-06,
+      "loss": 0.3018,
+      "step": 8405
+    },
+    {
+      "epoch": 2.1141851106639837,
+      "grad_norm": 0.3018190562725067,
+      "learning_rate": 2.42895681624297e-06,
+      "loss": 0.3397,
+      "step": 8406
+    },
+    {
+      "epoch": 2.11443661971831,
+      "grad_norm": 0.30792275071144104,
+      "learning_rate": 2.427701949889883e-06,
+      "loss": 0.3254,
+      "step": 8407
+    },
+    {
+      "epoch": 2.114688128772636,
+      "grad_norm": 0.2988547384738922,
+      "learning_rate": 2.426447303838033e-06,
+      "loss": 0.3281,
+      "step": 8408
+    },
+    {
+      "epoch": 2.1149396378269616,
+      "grad_norm": 0.2943371832370758,
+      "learning_rate": 2.4251928781948704e-06,
+      "loss": 0.3151,
+      "step": 8409
+    },
+    {
+      "epoch": 2.115191146881288,
+      "grad_norm": 0.31441688537597656,
+      "learning_rate": 2.423938673067831e-06,
+      "loss": 0.3286,
+      "step": 8410
+    },
+    {
+      "epoch": 2.1154426559356136,
+      "grad_norm": 0.31938257813453674,
+      "learning_rate": 2.422684688564327e-06,
+      "loss": 0.3401,
+      "step": 8411
+    },
+    {
+      "epoch": 2.1156941649899395,
+      "grad_norm": 0.29643604159355164,
+      "learning_rate": 2.4214309247917558e-06,
+      "loss": 0.3549,
+      "step": 8412
+    },
+    {
+      "epoch": 2.1159456740442657,
+      "grad_norm": 0.30069106817245483,
+      "learning_rate": 2.4201773818574956e-06,
+      "loss": 0.3058,
+      "step": 8413
+    },
+    {
+      "epoch": 2.1161971830985915,
+      "grad_norm": 0.2893010079860687,
+      "learning_rate": 2.4189240598689025e-06,
+      "loss": 0.3402,
+      "step": 8414
+    },
+    {
+      "epoch": 2.1164486921529173,
+      "grad_norm": 0.30323195457458496,
+      "learning_rate": 2.4176709589333173e-06,
+      "loss": 0.3333,
+      "step": 8415
+    },
+    {
+      "epoch": 2.1167002012072436,
+      "grad_norm": 0.29671111702919006,
+      "learning_rate": 2.4164180791580584e-06,
+      "loss": 0.3183,
+      "step": 8416
+    },
+    {
+      "epoch": 2.1169517102615694,
+      "grad_norm": 0.2824574112892151,
+      "learning_rate": 2.41516542065043e-06,
+      "loss": 0.3322,
+      "step": 8417
+    },
+    {
+      "epoch": 2.117203219315895,
+      "grad_norm": 0.30182933807373047,
+      "learning_rate": 2.4139129835177104e-06,
+      "loss": 0.313,
+      "step": 8418
+    },
+    {
+      "epoch": 2.1174547283702214,
+      "grad_norm": 0.325040727853775,
+      "learning_rate": 2.4126607678671672e-06,
+      "loss": 0.3391,
+      "step": 8419
+    },
+    {
+      "epoch": 2.1177062374245472,
+      "grad_norm": 0.3018413484096527,
+      "learning_rate": 2.411408773806041e-06,
+      "loss": 0.316,
+      "step": 8420
+    },
+    {
+      "epoch": 2.117957746478873,
+      "grad_norm": 0.29537421464920044,
+      "learning_rate": 2.410157001441561e-06,
+      "loss": 0.3042,
+      "step": 8421
+    },
+    {
+      "epoch": 2.1182092555331993,
+      "grad_norm": 0.2987903952598572,
+      "learning_rate": 2.40890545088093e-06,
+      "loss": 0.3171,
+      "step": 8422
+    },
+    {
+      "epoch": 2.118460764587525,
+      "grad_norm": 0.30431345105171204,
+      "learning_rate": 2.407654122231339e-06,
+      "loss": 0.3215,
+      "step": 8423
+    },
+    {
+      "epoch": 2.118712273641851,
+      "grad_norm": 0.277266263961792,
+      "learning_rate": 2.4064030155999534e-06,
+      "loss": 0.3201,
+      "step": 8424
+    },
+    {
+      "epoch": 2.118963782696177,
+      "grad_norm": 0.3193032443523407,
+      "learning_rate": 2.4051521310939258e-06,
+      "loss": 0.3517,
+      "step": 8425
+    },
+    {
+      "epoch": 2.119215291750503,
+      "grad_norm": 0.31553915143013,
+      "learning_rate": 2.4039014688203825e-06,
+      "loss": 0.3249,
+      "step": 8426
+    },
+    {
+      "epoch": 2.119466800804829,
+      "grad_norm": 0.2827959656715393,
+      "learning_rate": 2.4026510288864396e-06,
+      "loss": 0.3281,
+      "step": 8427
+    },
+    {
+      "epoch": 2.119718309859155,
+      "grad_norm": 0.28876793384552,
+      "learning_rate": 2.4014008113991855e-06,
+      "loss": 0.3331,
+      "step": 8428
+    },
+    {
+      "epoch": 2.119969818913481,
+      "grad_norm": 0.2849387526512146,
+      "learning_rate": 2.400150816465696e-06,
+      "loss": 0.3234,
+      "step": 8429
+    },
+    {
+      "epoch": 2.120221327967807,
+      "grad_norm": 0.3101980984210968,
+      "learning_rate": 2.398901044193023e-06,
+      "loss": 0.34,
+      "step": 8430
+    },
+    {
+      "epoch": 2.120472837022133,
+      "grad_norm": 0.2996865212917328,
+      "learning_rate": 2.3976514946882057e-06,
+      "loss": 0.3353,
+      "step": 8431
+    },
+    {
+      "epoch": 2.1207243460764587,
+      "grad_norm": 0.2950674593448639,
+      "learning_rate": 2.396402168058255e-06,
+      "loss": 0.3319,
+      "step": 8432
+    },
+    {
+      "epoch": 2.120975855130785,
+      "grad_norm": 0.28375566005706787,
+      "learning_rate": 2.395153064410171e-06,
+      "loss": 0.34,
+      "step": 8433
+    },
+    {
+      "epoch": 2.1212273641851107,
+      "grad_norm": 0.33272629976272583,
+      "learning_rate": 2.3939041838509324e-06,
+      "loss": 0.3399,
+      "step": 8434
+    },
+    {
+      "epoch": 2.1214788732394365,
+      "grad_norm": 0.28813275694847107,
+      "learning_rate": 2.3926555264874956e-06,
+      "loss": 0.3091,
+      "step": 8435
+    },
+    {
+      "epoch": 2.1217303822937628,
+      "grad_norm": 0.3202812075614929,
+      "learning_rate": 2.391407092426803e-06,
+      "loss": 0.3143,
+      "step": 8436
+    },
+    {
+      "epoch": 2.1219818913480886,
+      "grad_norm": 0.2992876470088959,
+      "learning_rate": 2.390158881775772e-06,
+      "loss": 0.3201,
+      "step": 8437
+    },
+    {
+      "epoch": 2.1222334004024144,
+      "grad_norm": 0.3007931113243103,
+      "learning_rate": 2.388910894641307e-06,
+      "loss": 0.322,
+      "step": 8438
+    },
+    {
+      "epoch": 2.1224849094567406,
+      "grad_norm": 0.2985740303993225,
+      "learning_rate": 2.387663131130288e-06,
+      "loss": 0.324,
+      "step": 8439
+    },
+    {
+      "epoch": 2.1227364185110664,
+      "grad_norm": 0.29932522773742676,
+      "learning_rate": 2.3864155913495803e-06,
+      "loss": 0.3135,
+      "step": 8440
+    },
+    {
+      "epoch": 2.1229879275653922,
+      "grad_norm": 0.3008473813533783,
+      "learning_rate": 2.3851682754060247e-06,
+      "loss": 0.3353,
+      "step": 8441
+    },
+    {
+      "epoch": 2.1232394366197185,
+      "grad_norm": 0.3180418610572815,
+      "learning_rate": 2.3839211834064496e-06,
+      "loss": 0.2981,
+      "step": 8442
+    },
+    {
+      "epoch": 2.1234909456740443,
+      "grad_norm": 0.29798197746276855,
+      "learning_rate": 2.3826743154576576e-06,
+      "loss": 0.333,
+      "step": 8443
+    },
+    {
+      "epoch": 2.12374245472837,
+      "grad_norm": 0.29840409755706787,
+      "learning_rate": 2.3814276716664365e-06,
+      "loss": 0.3147,
+      "step": 8444
+    },
+    {
+      "epoch": 2.1239939637826963,
+      "grad_norm": 0.30087023973464966,
+      "learning_rate": 2.3801812521395557e-06,
+      "loss": 0.3251,
+      "step": 8445
+    },
+    {
+      "epoch": 2.124245472837022,
+      "grad_norm": 0.27059686183929443,
+      "learning_rate": 2.3789350569837588e-06,
+      "loss": 0.3134,
+      "step": 8446
+    },
+    {
+      "epoch": 2.124496981891348,
+      "grad_norm": 0.2948879599571228,
+      "learning_rate": 2.377689086305779e-06,
+      "loss": 0.3309,
+      "step": 8447
+    },
+    {
+      "epoch": 2.124748490945674,
+      "grad_norm": 0.30460086464881897,
+      "learning_rate": 2.3764433402123223e-06,
+      "loss": 0.3201,
+      "step": 8448
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 0.2858458161354065,
+      "learning_rate": 2.3751978188100816e-06,
+      "loss": 0.3085,
+      "step": 8449
+    },
+    {
+      "epoch": 2.125251509054326,
+      "grad_norm": 0.2818026542663574,
+      "learning_rate": 2.3739525222057257e-06,
+      "loss": 0.3384,
+      "step": 8450
+    },
+    {
+      "epoch": 2.125503018108652,
+      "grad_norm": 0.2949872612953186,
+      "learning_rate": 2.37270745050591e-06,
+      "loss": 0.3338,
+      "step": 8451
+    },
+    {
+      "epoch": 2.125754527162978,
+      "grad_norm": 0.33432409167289734,
+      "learning_rate": 2.3714626038172623e-06,
+      "loss": 0.3343,
+      "step": 8452
+    },
+    {
+      "epoch": 2.1260060362173037,
+      "grad_norm": 0.32601264119148254,
+      "learning_rate": 2.3702179822464006e-06,
+      "loss": 0.3158,
+      "step": 8453
+    },
+    {
+      "epoch": 2.12625754527163,
+      "grad_norm": 0.31106406450271606,
+      "learning_rate": 2.3689735858999152e-06,
+      "loss": 0.3377,
+      "step": 8454
+    },
+    {
+      "epoch": 2.1265090543259557,
+      "grad_norm": 0.3100818991661072,
+      "learning_rate": 2.367729414884383e-06,
+      "loss": 0.31,
+      "step": 8455
+    },
+    {
+      "epoch": 2.1267605633802815,
+      "grad_norm": 0.29893726110458374,
+      "learning_rate": 2.3664854693063598e-06,
+      "loss": 0.3122,
+      "step": 8456
+    },
+    {
+      "epoch": 2.1270120724346078,
+      "grad_norm": 0.2957155108451843,
+      "learning_rate": 2.3652417492723795e-06,
+      "loss": 0.3108,
+      "step": 8457
+    },
+    {
+      "epoch": 2.1272635814889336,
+      "grad_norm": 0.3101329803466797,
+      "learning_rate": 2.3639982548889623e-06,
+      "loss": 0.3182,
+      "step": 8458
+    },
+    {
+      "epoch": 2.1275150905432594,
+      "grad_norm": 0.29328593611717224,
+      "learning_rate": 2.3627549862626014e-06,
+      "loss": 0.324,
+      "step": 8459
+    },
+    {
+      "epoch": 2.1277665995975856,
+      "grad_norm": 0.30886930227279663,
+      "learning_rate": 2.3615119434997803e-06,
+      "loss": 0.312,
+      "step": 8460
+    },
+    {
+      "epoch": 2.1280181086519114,
+      "grad_norm": 0.339282751083374,
+      "learning_rate": 2.360269126706952e-06,
+      "loss": 0.3421,
+      "step": 8461
+    },
+    {
+      "epoch": 2.1282696177062372,
+      "grad_norm": 0.3135293126106262,
+      "learning_rate": 2.359026535990561e-06,
+      "loss": 0.2935,
+      "step": 8462
+    },
+    {
+      "epoch": 2.1285211267605635,
+      "grad_norm": 0.3130805492401123,
+      "learning_rate": 2.357784171457024e-06,
+      "loss": 0.3256,
+      "step": 8463
+    },
+    {
+      "epoch": 2.1287726358148893,
+      "grad_norm": 0.3069337010383606,
+      "learning_rate": 2.3565420332127447e-06,
+      "loss": 0.3241,
+      "step": 8464
+    },
+    {
+      "epoch": 2.129024144869215,
+      "grad_norm": 0.3093840181827545,
+      "learning_rate": 2.355300121364101e-06,
+      "loss": 0.31,
+      "step": 8465
+    },
+    {
+      "epoch": 2.1292756539235413,
+      "grad_norm": 0.2900824546813965,
+      "learning_rate": 2.354058436017456e-06,
+      "loss": 0.355,
+      "step": 8466
+    },
+    {
+      "epoch": 2.129527162977867,
+      "grad_norm": 0.3184683322906494,
+      "learning_rate": 2.352816977279156e-06,
+      "loss": 0.3043,
+      "step": 8467
+    },
+    {
+      "epoch": 2.129778672032193,
+      "grad_norm": 0.29197773337364197,
+      "learning_rate": 2.351575745255519e-06,
+      "loss": 0.3052,
+      "step": 8468
+    },
+    {
+      "epoch": 2.130030181086519,
+      "grad_norm": 0.2952864468097687,
+      "learning_rate": 2.350334740052852e-06,
+      "loss": 0.319,
+      "step": 8469
+    },
+    {
+      "epoch": 2.130281690140845,
+      "grad_norm": 0.29490870237350464,
+      "learning_rate": 2.349093961777437e-06,
+      "loss": 0.3397,
+      "step": 8470
+    },
+    {
+      "epoch": 2.1305331991951713,
+      "grad_norm": 0.33022499084472656,
+      "learning_rate": 2.3478534105355423e-06,
+      "loss": 0.3102,
+      "step": 8471
+    },
+    {
+      "epoch": 2.130784708249497,
+      "grad_norm": 0.3145847022533417,
+      "learning_rate": 2.3466130864334085e-06,
+      "loss": 0.3431,
+      "step": 8472
+    },
+    {
+      "epoch": 2.131036217303823,
+      "grad_norm": 0.2934006154537201,
+      "learning_rate": 2.345372989577267e-06,
+      "loss": 0.3187,
+      "step": 8473
+    },
+    {
+      "epoch": 2.131287726358149,
+      "grad_norm": 0.29067620635032654,
+      "learning_rate": 2.344133120073319e-06,
+      "loss": 0.2961,
+      "step": 8474
+    },
+    {
+      "epoch": 2.131539235412475,
+      "grad_norm": 0.28378283977508545,
+      "learning_rate": 2.3428934780277567e-06,
+      "loss": 0.3215,
+      "step": 8475
+    },
+    {
+      "epoch": 2.1317907444668007,
+      "grad_norm": 0.2955572009086609,
+      "learning_rate": 2.341654063546743e-06,
+      "loss": 0.3309,
+      "step": 8476
+    },
+    {
+      "epoch": 2.132042253521127,
+      "grad_norm": 0.31690630316734314,
+      "learning_rate": 2.340414876736429e-06,
+      "loss": 0.3254,
+      "step": 8477
+    },
+    {
+      "epoch": 2.1322937625754528,
+      "grad_norm": 0.297673761844635,
+      "learning_rate": 2.339175917702943e-06,
+      "loss": 0.3179,
+      "step": 8478
+    },
+    {
+      "epoch": 2.1325452716297786,
+      "grad_norm": 0.31934690475463867,
+      "learning_rate": 2.3379371865523926e-06,
+      "loss": 0.3356,
+      "step": 8479
+    },
+    {
+      "epoch": 2.132796780684105,
+      "grad_norm": 0.2896483540534973,
+      "learning_rate": 2.33669868339087e-06,
+      "loss": 0.3258,
+      "step": 8480
+    },
+    {
+      "epoch": 2.1330482897384306,
+      "grad_norm": 0.2923082113265991,
+      "learning_rate": 2.335460408324442e-06,
+      "loss": 0.326,
+      "step": 8481
+    },
+    {
+      "epoch": 2.1332997987927564,
+      "grad_norm": 0.30978333950042725,
+      "learning_rate": 2.3342223614591623e-06,
+      "loss": 0.3199,
+      "step": 8482
+    },
+    {
+      "epoch": 2.1335513078470827,
+      "grad_norm": 0.2992372512817383,
+      "learning_rate": 2.332984542901059e-06,
+      "loss": 0.3611,
+      "step": 8483
+    },
+    {
+      "epoch": 2.1338028169014085,
+      "grad_norm": 0.3130771517753601,
+      "learning_rate": 2.331746952756146e-06,
+      "loss": 0.3089,
+      "step": 8484
+    },
+    {
+      "epoch": 2.1340543259557343,
+      "grad_norm": 0.31406500935554504,
+      "learning_rate": 2.3305095911304123e-06,
+      "loss": 0.3132,
+      "step": 8485
+    },
+    {
+      "epoch": 2.1343058350100605,
+      "grad_norm": 0.30472251772880554,
+      "learning_rate": 2.3292724581298338e-06,
+      "loss": 0.3216,
+      "step": 8486
+    },
+    {
+      "epoch": 2.1345573440643864,
+      "grad_norm": 0.31913644075393677,
+      "learning_rate": 2.328035553860359e-06,
+      "loss": 0.3051,
+      "step": 8487
+    },
+    {
+      "epoch": 2.134808853118712,
+      "grad_norm": 0.3231859803199768,
+      "learning_rate": 2.326798878427924e-06,
+      "loss": 0.3293,
+      "step": 8488
+    },
+    {
+      "epoch": 2.1350603621730384,
+      "grad_norm": 0.2944040298461914,
+      "learning_rate": 2.325562431938442e-06,
+      "loss": 0.3373,
+      "step": 8489
+    },
+    {
+      "epoch": 2.135311871227364,
+      "grad_norm": 0.2909225523471832,
+      "learning_rate": 2.3243262144978063e-06,
+      "loss": 0.3285,
+      "step": 8490
+    },
+    {
+      "epoch": 2.13556338028169,
+      "grad_norm": 0.283944696187973,
+      "learning_rate": 2.3230902262118922e-06,
+      "loss": 0.3177,
+      "step": 8491
+    },
+    {
+      "epoch": 2.1358148893360163,
+      "grad_norm": 0.31086745858192444,
+      "learning_rate": 2.321854467186552e-06,
+      "loss": 0.3492,
+      "step": 8492
+    },
+    {
+      "epoch": 2.136066398390342,
+      "grad_norm": 0.3098077178001404,
+      "learning_rate": 2.320618937527624e-06,
+      "loss": 0.3144,
+      "step": 8493
+    },
+    {
+      "epoch": 2.136317907444668,
+      "grad_norm": 0.31440508365631104,
+      "learning_rate": 2.31938363734092e-06,
+      "loss": 0.3136,
+      "step": 8494
+    },
+    {
+      "epoch": 2.136569416498994,
+      "grad_norm": 0.3078277111053467,
+      "learning_rate": 2.3181485667322397e-06,
+      "loss": 0.3126,
+      "step": 8495
+    },
+    {
+      "epoch": 2.13682092555332,
+      "grad_norm": 0.3191978633403778,
+      "learning_rate": 2.316913725807355e-06,
+      "loss": 0.2969,
+      "step": 8496
+    },
+    {
+      "epoch": 2.1370724346076457,
+      "grad_norm": 0.2999849319458008,
+      "learning_rate": 2.3156791146720266e-06,
+      "loss": 0.3327,
+      "step": 8497
+    },
+    {
+      "epoch": 2.137323943661972,
+      "grad_norm": 0.2959345281124115,
+      "learning_rate": 2.3144447334319866e-06,
+      "loss": 0.3355,
+      "step": 8498
+    },
+    {
+      "epoch": 2.137575452716298,
+      "grad_norm": 0.29375961422920227,
+      "learning_rate": 2.313210582192954e-06,
+      "loss": 0.3151,
+      "step": 8499
+    },
+    {
+      "epoch": 2.1378269617706236,
+      "grad_norm": 0.2988927960395813,
+      "learning_rate": 2.3119766610606293e-06,
+      "loss": 0.3479,
+      "step": 8500
+    },
+    {
+      "epoch": 2.13807847082495,
+      "grad_norm": 0.32058557868003845,
+      "learning_rate": 2.3107429701406845e-06,
+      "loss": 0.3118,
+      "step": 8501
+    },
+    {
+      "epoch": 2.1383299798792756,
+      "grad_norm": 0.31546249985694885,
+      "learning_rate": 2.309509509538783e-06,
+      "loss": 0.3289,
+      "step": 8502
+    },
+    {
+      "epoch": 2.1385814889336014,
+      "grad_norm": 0.3331628739833832,
+      "learning_rate": 2.3082762793605582e-06,
+      "loss": 0.3276,
+      "step": 8503
+    },
+    {
+      "epoch": 2.1388329979879277,
+      "grad_norm": 0.3143128752708435,
+      "learning_rate": 2.307043279711633e-06,
+      "loss": 0.3154,
+      "step": 8504
+    },
+    {
+      "epoch": 2.1390845070422535,
+      "grad_norm": 0.33894532918930054,
+      "learning_rate": 2.3058105106976013e-06,
+      "loss": 0.3275,
+      "step": 8505
+    },
+    {
+      "epoch": 2.1393360160965793,
+      "grad_norm": 0.30815622210502625,
+      "learning_rate": 2.3045779724240468e-06,
+      "loss": 0.3163,
+      "step": 8506
+    },
+    {
+      "epoch": 2.1395875251509056,
+      "grad_norm": 0.29994967579841614,
+      "learning_rate": 2.3033456649965246e-06,
+      "loss": 0.3283,
+      "step": 8507
+    },
+    {
+      "epoch": 2.1398390342052314,
+      "grad_norm": 0.3090510666370392,
+      "learning_rate": 2.302113588520578e-06,
+      "loss": 0.3087,
+      "step": 8508
+    },
+    {
+      "epoch": 2.140090543259557,
+      "grad_norm": 0.3417157530784607,
+      "learning_rate": 2.3008817431017225e-06,
+      "loss": 0.3144,
+      "step": 8509
+    },
+    {
+      "epoch": 2.1403420523138834,
+      "grad_norm": 0.3093573749065399,
+      "learning_rate": 2.2996501288454606e-06,
+      "loss": 0.3099,
+      "step": 8510
+    },
+    {
+      "epoch": 2.140593561368209,
+      "grad_norm": 0.2987154722213745,
+      "learning_rate": 2.2984187458572727e-06,
+      "loss": 0.3293,
+      "step": 8511
+    },
+    {
+      "epoch": 2.140845070422535,
+      "grad_norm": 0.31684228777885437,
+      "learning_rate": 2.297187594242617e-06,
+      "loss": 0.3171,
+      "step": 8512
+    },
+    {
+      "epoch": 2.1410965794768613,
+      "grad_norm": 0.2938196063041687,
+      "learning_rate": 2.2959566741069365e-06,
+      "loss": 0.3196,
+      "step": 8513
+    },
+    {
+      "epoch": 2.141348088531187,
+      "grad_norm": 0.30638089776039124,
+      "learning_rate": 2.2947259855556493e-06,
+      "loss": 0.3256,
+      "step": 8514
+    },
+    {
+      "epoch": 2.141599597585513,
+      "grad_norm": 0.2842661738395691,
+      "learning_rate": 2.2934955286941583e-06,
+      "loss": 0.3199,
+      "step": 8515
+    },
+    {
+      "epoch": 2.141851106639839,
+      "grad_norm": 0.30369311571121216,
+      "learning_rate": 2.292265303627842e-06,
+      "loss": 0.3241,
+      "step": 8516
+    },
+    {
+      "epoch": 2.142102615694165,
+      "grad_norm": 0.3095688223838806,
+      "learning_rate": 2.2910353104620647e-06,
+      "loss": 0.3437,
+      "step": 8517
+    },
+    {
+      "epoch": 2.1423541247484907,
+      "grad_norm": 0.29922717809677124,
+      "learning_rate": 2.2898055493021644e-06,
+      "loss": 0.3326,
+      "step": 8518
+    },
+    {
+      "epoch": 2.142605633802817,
+      "grad_norm": 0.30769720673561096,
+      "learning_rate": 2.288576020253465e-06,
+      "loss": 0.3249,
+      "step": 8519
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.2776482403278351,
+      "learning_rate": 2.2873467234212654e-06,
+      "loss": 0.327,
+      "step": 8520
+    },
+    {
+      "epoch": 2.1431086519114686,
+      "grad_norm": 0.30429205298423767,
+      "learning_rate": 2.2861176589108487e-06,
+      "loss": 0.3109,
+      "step": 8521
+    },
+    {
+      "epoch": 2.143360160965795,
+      "grad_norm": 0.2737998366355896,
+      "learning_rate": 2.284888826827478e-06,
+      "loss": 0.3143,
+      "step": 8522
+    },
+    {
+      "epoch": 2.1436116700201207,
+      "grad_norm": 0.29733067750930786,
+      "learning_rate": 2.2836602272763924e-06,
+      "loss": 0.3157,
+      "step": 8523
+    },
+    {
+      "epoch": 2.1438631790744465,
+      "grad_norm": 0.32914045453071594,
+      "learning_rate": 2.2824318603628163e-06,
+      "loss": 0.3138,
+      "step": 8524
+    },
+    {
+      "epoch": 2.1441146881287727,
+      "grad_norm": 0.3098900020122528,
+      "learning_rate": 2.2812037261919483e-06,
+      "loss": 0.3287,
+      "step": 8525
+    },
+    {
+      "epoch": 2.1443661971830985,
+      "grad_norm": 0.3037329316139221,
+      "learning_rate": 2.2799758248689747e-06,
+      "loss": 0.3236,
+      "step": 8526
+    },
+    {
+      "epoch": 2.1446177062374243,
+      "grad_norm": 0.334514856338501,
+      "learning_rate": 2.2787481564990533e-06,
+      "loss": 0.3309,
+      "step": 8527
+    },
+    {
+      "epoch": 2.1448692152917506,
+      "grad_norm": 0.28961655497550964,
+      "learning_rate": 2.27752072118733e-06,
+      "loss": 0.3469,
+      "step": 8528
+    },
+    {
+      "epoch": 2.1451207243460764,
+      "grad_norm": 0.31919920444488525,
+      "learning_rate": 2.2762935190389233e-06,
+      "loss": 0.3104,
+      "step": 8529
+    },
+    {
+      "epoch": 2.1453722334004026,
+      "grad_norm": 0.29657045006752014,
+      "learning_rate": 2.275066550158939e-06,
+      "loss": 0.3139,
+      "step": 8530
+    },
+    {
+      "epoch": 2.1456237424547284,
+      "grad_norm": 0.29360705614089966,
+      "learning_rate": 2.273839814652456e-06,
+      "loss": 0.3169,
+      "step": 8531
+    },
+    {
+      "epoch": 2.1458752515090542,
+      "grad_norm": 0.331571102142334,
+      "learning_rate": 2.27261331262454e-06,
+      "loss": 0.3003,
+      "step": 8532
+    },
+    {
+      "epoch": 2.1461267605633805,
+      "grad_norm": 0.28920167684555054,
+      "learning_rate": 2.2713870441802287e-06,
+      "loss": 0.3175,
+      "step": 8533
+    },
+    {
+      "epoch": 2.1463782696177063,
+      "grad_norm": 0.27948108315467834,
+      "learning_rate": 2.2701610094245496e-06,
+      "loss": 0.2903,
+      "step": 8534
+    },
+    {
+      "epoch": 2.146629778672032,
+      "grad_norm": 0.2875995934009552,
+      "learning_rate": 2.2689352084625e-06,
+      "loss": 0.3286,
+      "step": 8535
+    },
+    {
+      "epoch": 2.1468812877263583,
+      "grad_norm": 0.2751031219959259,
+      "learning_rate": 2.267709641399066e-06,
+      "loss": 0.2945,
+      "step": 8536
+    },
+    {
+      "epoch": 2.147132796780684,
+      "grad_norm": 0.3058602213859558,
+      "learning_rate": 2.2664843083392063e-06,
+      "loss": 0.3593,
+      "step": 8537
+    },
+    {
+      "epoch": 2.14738430583501,
+      "grad_norm": 0.31076836585998535,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.332,
+      "step": 8538
+    },
+    {
+      "epoch": 2.147635814889336,
+      "grad_norm": 0.2831321954727173,
+      "learning_rate": 2.2640343446499656e-06,
+      "loss": 0.3137,
+      "step": 8539
+    },
+    {
+      "epoch": 2.147887323943662,
+      "grad_norm": 0.2820286452770233,
+      "learning_rate": 2.2628097142304083e-06,
+      "loss": 0.3312,
+      "step": 8540
+    },
+    {
+      "epoch": 2.148138832997988,
+      "grad_norm": 0.29765576124191284,
+      "learning_rate": 2.2615853182340737e-06,
+      "loss": 0.3077,
+      "step": 8541
+    },
+    {
+      "epoch": 2.148390342052314,
+      "grad_norm": 0.3157854676246643,
+      "learning_rate": 2.2603611567658267e-06,
+      "loss": 0.3455,
+      "step": 8542
+    },
+    {
+      "epoch": 2.14864185110664,
+      "grad_norm": 0.30632486939430237,
+      "learning_rate": 2.259137229930506e-06,
+      "loss": 0.3374,
+      "step": 8543
+    },
+    {
+      "epoch": 2.1488933601609657,
+      "grad_norm": 0.312173068523407,
+      "learning_rate": 2.2579135378329357e-06,
+      "loss": 0.3233,
+      "step": 8544
+    },
+    {
+      "epoch": 2.149144869215292,
+      "grad_norm": 0.2874966263771057,
+      "learning_rate": 2.2566900805779157e-06,
+      "loss": 0.3079,
+      "step": 8545
+    },
+    {
+      "epoch": 2.1493963782696177,
+      "grad_norm": 0.30250847339630127,
+      "learning_rate": 2.2554668582702294e-06,
+      "loss": 0.3519,
+      "step": 8546
+    },
+    {
+      "epoch": 2.1496478873239435,
+      "grad_norm": 0.2963038384914398,
+      "learning_rate": 2.2542438710146354e-06,
+      "loss": 0.3195,
+      "step": 8547
+    },
+    {
+      "epoch": 2.1498993963782698,
+      "grad_norm": 0.299123615026474,
+      "learning_rate": 2.253021118915878e-06,
+      "loss": 0.3204,
+      "step": 8548
+    },
+    {
+      "epoch": 2.1501509054325956,
+      "grad_norm": 0.28678908944129944,
+      "learning_rate": 2.2517986020786745e-06,
+      "loss": 0.308,
+      "step": 8549
+    },
+    {
+      "epoch": 2.1504024144869214,
+      "grad_norm": 0.2993088662624359,
+      "learning_rate": 2.2505763206077306e-06,
+      "loss": 0.306,
+      "step": 8550
+    },
+    {
+      "epoch": 2.1506539235412476,
+      "grad_norm": 0.30504879355430603,
+      "learning_rate": 2.249354274607723e-06,
+      "loss": 0.3226,
+      "step": 8551
+    },
+    {
+      "epoch": 2.1509054325955734,
+      "grad_norm": 0.33471235632896423,
+      "learning_rate": 2.2481324641833146e-06,
+      "loss": 0.3101,
+      "step": 8552
+    },
+    {
+      "epoch": 2.1511569416498992,
+      "grad_norm": 0.30294346809387207,
+      "learning_rate": 2.246910889439144e-06,
+      "loss": 0.3211,
+      "step": 8553
+    },
+    {
+      "epoch": 2.1514084507042255,
+      "grad_norm": 0.2953152656555176,
+      "learning_rate": 2.2456895504798322e-06,
+      "loss": 0.3226,
+      "step": 8554
+    },
+    {
+      "epoch": 2.1516599597585513,
+      "grad_norm": 0.2956666052341461,
+      "learning_rate": 2.2444684474099807e-06,
+      "loss": 0.3312,
+      "step": 8555
+    },
+    {
+      "epoch": 2.151911468812877,
+      "grad_norm": 0.3139225244522095,
+      "learning_rate": 2.2432475803341674e-06,
+      "loss": 0.3122,
+      "step": 8556
+    },
+    {
+      "epoch": 2.1521629778672033,
+      "grad_norm": 0.3028757572174072,
+      "learning_rate": 2.242026949356954e-06,
+      "loss": 0.3361,
+      "step": 8557
+    },
+    {
+      "epoch": 2.152414486921529,
+      "grad_norm": 0.31237322092056274,
+      "learning_rate": 2.2408065545828765e-06,
+      "loss": 0.338,
+      "step": 8558
+    },
+    {
+      "epoch": 2.152665995975855,
+      "grad_norm": 0.3021441400051117,
+      "learning_rate": 2.239586396116458e-06,
+      "loss": 0.2959,
+      "step": 8559
+    },
+    {
+      "epoch": 2.152917505030181,
+      "grad_norm": 0.29930517077445984,
+      "learning_rate": 2.238366474062194e-06,
+      "loss": 0.3408,
+      "step": 8560
+    },
+    {
+      "epoch": 2.153169014084507,
+      "grad_norm": 0.2866148352622986,
+      "learning_rate": 2.2371467885245667e-06,
+      "loss": 0.3145,
+      "step": 8561
+    },
+    {
+      "epoch": 2.153420523138833,
+      "grad_norm": 0.3019120693206787,
+      "learning_rate": 2.2359273396080306e-06,
+      "loss": 0.3423,
+      "step": 8562
+    },
+    {
+      "epoch": 2.153672032193159,
+      "grad_norm": 0.31220588088035583,
+      "learning_rate": 2.2347081274170273e-06,
+      "loss": 0.3306,
+      "step": 8563
+    },
+    {
+      "epoch": 2.153923541247485,
+      "grad_norm": 0.31017324328422546,
+      "learning_rate": 2.233489152055972e-06,
+      "loss": 0.3089,
+      "step": 8564
+    },
+    {
+      "epoch": 2.1541750503018107,
+      "grad_norm": 0.2911800146102905,
+      "learning_rate": 2.2322704136292632e-06,
+      "loss": 0.3092,
+      "step": 8565
+    },
+    {
+      "epoch": 2.154426559356137,
+      "grad_norm": 0.2987365424633026,
+      "learning_rate": 2.231051912241281e-06,
+      "loss": 0.3316,
+      "step": 8566
+    },
+    {
+      "epoch": 2.1546780684104627,
+      "grad_norm": 0.30714505910873413,
+      "learning_rate": 2.229833647996378e-06,
+      "loss": 0.2903,
+      "step": 8567
+    },
+    {
+      "epoch": 2.1549295774647885,
+      "grad_norm": 0.31266728043556213,
+      "learning_rate": 2.228615620998894e-06,
+      "loss": 0.3185,
+      "step": 8568
+    },
+    {
+      "epoch": 2.1551810865191148,
+      "grad_norm": 0.27466437220573425,
+      "learning_rate": 2.2273978313531436e-06,
+      "loss": 0.294,
+      "step": 8569
+    },
+    {
+      "epoch": 2.1554325955734406,
+      "grad_norm": 0.3066222071647644,
+      "learning_rate": 2.2261802791634245e-06,
+      "loss": 0.3285,
+      "step": 8570
+    },
+    {
+      "epoch": 2.155684104627767,
+      "grad_norm": 0.31608250737190247,
+      "learning_rate": 2.2249629645340104e-06,
+      "loss": 0.3345,
+      "step": 8571
+    },
+    {
+      "epoch": 2.1559356136820926,
+      "grad_norm": 0.3242324888706207,
+      "learning_rate": 2.2237458875691592e-06,
+      "loss": 0.2912,
+      "step": 8572
+    },
+    {
+      "epoch": 2.1561871227364184,
+      "grad_norm": 0.2953372895717621,
+      "learning_rate": 2.2225290483731034e-06,
+      "loss": 0.3208,
+      "step": 8573
+    },
+    {
+      "epoch": 2.1564386317907447,
+      "grad_norm": 0.3065500259399414,
+      "learning_rate": 2.221312447050058e-06,
+      "loss": 0.333,
+      "step": 8574
+    },
+    {
+      "epoch": 2.1566901408450705,
+      "grad_norm": 0.2919135093688965,
+      "learning_rate": 2.2200960837042202e-06,
+      "loss": 0.3145,
+      "step": 8575
+    },
+    {
+      "epoch": 2.1569416498993963,
+      "grad_norm": 0.2694363594055176,
+      "learning_rate": 2.2188799584397604e-06,
+      "loss": 0.3248,
+      "step": 8576
+    },
+    {
+      "epoch": 2.1571931589537225,
+      "grad_norm": 0.29707813262939453,
+      "learning_rate": 2.2176640713608345e-06,
+      "loss": 0.2982,
+      "step": 8577
+    },
+    {
+      "epoch": 2.1574446680080483,
+      "grad_norm": 0.3023638129234314,
+      "learning_rate": 2.2164484225715734e-06,
+      "loss": 0.3199,
+      "step": 8578
+    },
+    {
+      "epoch": 2.157696177062374,
+      "grad_norm": 0.29433873295783997,
+      "learning_rate": 2.215233012176093e-06,
+      "loss": 0.3304,
+      "step": 8579
+    },
+    {
+      "epoch": 2.1579476861167004,
+      "grad_norm": 0.2980920672416687,
+      "learning_rate": 2.2140178402784814e-06,
+      "loss": 0.3326,
+      "step": 8580
+    },
+    {
+      "epoch": 2.158199195171026,
+      "grad_norm": 0.3132590651512146,
+      "learning_rate": 2.212802906982815e-06,
+      "loss": 0.3246,
+      "step": 8581
+    },
+    {
+      "epoch": 2.158450704225352,
+      "grad_norm": 0.29961881041526794,
+      "learning_rate": 2.2115882123931403e-06,
+      "loss": 0.3313,
+      "step": 8582
+    },
+    {
+      "epoch": 2.1587022132796783,
+      "grad_norm": 0.3205260634422302,
+      "learning_rate": 2.210373756613494e-06,
+      "loss": 0.3257,
+      "step": 8583
+    },
+    {
+      "epoch": 2.158953722334004,
+      "grad_norm": 0.2848275601863861,
+      "learning_rate": 2.209159539747881e-06,
+      "loss": 0.3306,
+      "step": 8584
+    },
+    {
+      "epoch": 2.15920523138833,
+      "grad_norm": 0.3031473457813263,
+      "learning_rate": 2.2079455619002936e-06,
+      "loss": 0.3369,
+      "step": 8585
+    },
+    {
+      "epoch": 2.159456740442656,
+      "grad_norm": 0.3208923935890198,
+      "learning_rate": 2.2067318231747047e-06,
+      "loss": 0.3092,
+      "step": 8586
+    },
+    {
+      "epoch": 2.159708249496982,
+      "grad_norm": 0.31306546926498413,
+      "learning_rate": 2.2055183236750577e-06,
+      "loss": 0.3051,
+      "step": 8587
+    },
+    {
+      "epoch": 2.1599597585513077,
+      "grad_norm": 0.3087550103664398,
+      "learning_rate": 2.2043050635052866e-06,
+      "loss": 0.3105,
+      "step": 8588
+    },
+    {
+      "epoch": 2.160211267605634,
+      "grad_norm": 0.2876415252685547,
+      "learning_rate": 2.2030920427692947e-06,
+      "loss": 0.3006,
+      "step": 8589
+    },
+    {
+      "epoch": 2.16046277665996,
+      "grad_norm": 0.3287197947502136,
+      "learning_rate": 2.201879261570974e-06,
+      "loss": 0.3298,
+      "step": 8590
+    },
+    {
+      "epoch": 2.1607142857142856,
+      "grad_norm": 0.30581316351890564,
+      "learning_rate": 2.2006667200141877e-06,
+      "loss": 0.3205,
+      "step": 8591
+    },
+    {
+      "epoch": 2.160965794768612,
+      "grad_norm": 0.30748024582862854,
+      "learning_rate": 2.199454418202786e-06,
+      "loss": 0.3159,
+      "step": 8592
+    },
+    {
+      "epoch": 2.1612173038229376,
+      "grad_norm": 0.32226288318634033,
+      "learning_rate": 2.1982423562405915e-06,
+      "loss": 0.3129,
+      "step": 8593
+    },
+    {
+      "epoch": 2.1614688128772634,
+      "grad_norm": 0.2897011935710907,
+      "learning_rate": 2.1970305342314135e-06,
+      "loss": 0.336,
+      "step": 8594
+    },
+    {
+      "epoch": 2.1617203219315897,
+      "grad_norm": 0.3219929039478302,
+      "learning_rate": 2.1958189522790325e-06,
+      "loss": 0.3294,
+      "step": 8595
+    },
+    {
+      "epoch": 2.1619718309859155,
+      "grad_norm": 0.3067570626735687,
+      "learning_rate": 2.194607610487216e-06,
+      "loss": 0.3422,
+      "step": 8596
+    },
+    {
+      "epoch": 2.1622233400402413,
+      "grad_norm": 0.2897886633872986,
+      "learning_rate": 2.1933965089597087e-06,
+      "loss": 0.3096,
+      "step": 8597
+    },
+    {
+      "epoch": 2.1624748490945676,
+      "grad_norm": 0.2907717525959015,
+      "learning_rate": 2.1921856478002302e-06,
+      "loss": 0.3176,
+      "step": 8598
+    },
+    {
+      "epoch": 2.1627263581488934,
+      "grad_norm": 0.31856968998908997,
+      "learning_rate": 2.1909750271124873e-06,
+      "loss": 0.324,
+      "step": 8599
+    },
+    {
+      "epoch": 2.162977867203219,
+      "grad_norm": 0.30686262249946594,
+      "learning_rate": 2.1897646470001588e-06,
+      "loss": 0.3286,
+      "step": 8600
+    },
+    {
+      "epoch": 2.1632293762575454,
+      "grad_norm": 0.30653515458106995,
+      "learning_rate": 2.188554507566909e-06,
+      "loss": 0.3303,
+      "step": 8601
+    },
+    {
+      "epoch": 2.163480885311871,
+      "grad_norm": 0.3120279908180237,
+      "learning_rate": 2.187344608916375e-06,
+      "loss": 0.3467,
+      "step": 8602
+    },
+    {
+      "epoch": 2.163732394366197,
+      "grad_norm": 0.32078132033348083,
+      "learning_rate": 2.1861349511521817e-06,
+      "loss": 0.3224,
+      "step": 8603
+    },
+    {
+      "epoch": 2.1639839034205233,
+      "grad_norm": 0.2890268862247467,
+      "learning_rate": 2.1849255343779246e-06,
+      "loss": 0.3249,
+      "step": 8604
+    },
+    {
+      "epoch": 2.164235412474849,
+      "grad_norm": 0.3207026720046997,
+      "learning_rate": 2.183716358697186e-06,
+      "loss": 0.3349,
+      "step": 8605
+    },
+    {
+      "epoch": 2.164486921529175,
+      "grad_norm": 0.30875587463378906,
+      "learning_rate": 2.1825074242135206e-06,
+      "loss": 0.2937,
+      "step": 8606
+    },
+    {
+      "epoch": 2.164738430583501,
+      "grad_norm": 0.27507445216178894,
+      "learning_rate": 2.181298731030469e-06,
+      "loss": 0.3357,
+      "step": 8607
+    },
+    {
+      "epoch": 2.164989939637827,
+      "grad_norm": 0.2858785092830658,
+      "learning_rate": 2.180090279251548e-06,
+      "loss": 0.3203,
+      "step": 8608
+    },
+    {
+      "epoch": 2.1652414486921527,
+      "grad_norm": 0.3202469050884247,
+      "learning_rate": 2.1788820689802524e-06,
+      "loss": 0.329,
+      "step": 8609
+    },
+    {
+      "epoch": 2.165492957746479,
+      "grad_norm": 0.30295151472091675,
+      "learning_rate": 2.1776741003200603e-06,
+      "loss": 0.3369,
+      "step": 8610
+    },
+    {
+      "epoch": 2.165744466800805,
+      "grad_norm": 0.3162943720817566,
+      "learning_rate": 2.1764663733744234e-06,
+      "loss": 0.3167,
+      "step": 8611
+    },
+    {
+      "epoch": 2.1659959758551306,
+      "grad_norm": 0.2848318815231323,
+      "learning_rate": 2.175258888246779e-06,
+      "loss": 0.311,
+      "step": 8612
+    },
+    {
+      "epoch": 2.166247484909457,
+      "grad_norm": 0.31551775336265564,
+      "learning_rate": 2.174051645040538e-06,
+      "loss": 0.3374,
+      "step": 8613
+    },
+    {
+      "epoch": 2.1664989939637826,
+      "grad_norm": 0.2893933057785034,
+      "learning_rate": 2.172844643859096e-06,
+      "loss": 0.3212,
+      "step": 8614
+    },
+    {
+      "epoch": 2.1667505030181085,
+      "grad_norm": 0.295808881521225,
+      "learning_rate": 2.1716378848058217e-06,
+      "loss": 0.3155,
+      "step": 8615
+    },
+    {
+      "epoch": 2.1670020120724347,
+      "grad_norm": 0.3089081346988678,
+      "learning_rate": 2.1704313679840706e-06,
+      "loss": 0.3483,
+      "step": 8616
+    },
+    {
+      "epoch": 2.1672535211267605,
+      "grad_norm": 0.3219684064388275,
+      "learning_rate": 2.169225093497169e-06,
+      "loss": 0.3358,
+      "step": 8617
+    },
+    {
+      "epoch": 2.1675050301810863,
+      "grad_norm": 0.3217262029647827,
+      "learning_rate": 2.1680190614484292e-06,
+      "loss": 0.3385,
+      "step": 8618
+    },
+    {
+      "epoch": 2.1677565392354126,
+      "grad_norm": 0.28920111060142517,
+      "learning_rate": 2.166813271941141e-06,
+      "loss": 0.3109,
+      "step": 8619
+    },
+    {
+      "epoch": 2.1680080482897384,
+      "grad_norm": 0.32903382182121277,
+      "learning_rate": 2.165607725078571e-06,
+      "loss": 0.3133,
+      "step": 8620
+    },
+    {
+      "epoch": 2.168259557344064,
+      "grad_norm": 0.3116111159324646,
+      "learning_rate": 2.1644024209639687e-06,
+      "loss": 0.3415,
+      "step": 8621
+    },
+    {
+      "epoch": 2.1685110663983904,
+      "grad_norm": 0.30731502175331116,
+      "learning_rate": 2.1631973597005574e-06,
+      "loss": 0.3426,
+      "step": 8622
+    },
+    {
+      "epoch": 2.1687625754527162,
+      "grad_norm": 0.29497724771499634,
+      "learning_rate": 2.1619925413915475e-06,
+      "loss": 0.3216,
+      "step": 8623
+    },
+    {
+      "epoch": 2.169014084507042,
+      "grad_norm": 0.30117177963256836,
+      "learning_rate": 2.16078796614012e-06,
+      "loss": 0.3228,
+      "step": 8624
+    },
+    {
+      "epoch": 2.1692655935613683,
+      "grad_norm": 0.29439517855644226,
+      "learning_rate": 2.159583634049443e-06,
+      "loss": 0.3107,
+      "step": 8625
+    },
+    {
+      "epoch": 2.169517102615694,
+      "grad_norm": 0.2719261944293976,
+      "learning_rate": 2.1583795452226563e-06,
+      "loss": 0.2919,
+      "step": 8626
+    },
+    {
+      "epoch": 2.16976861167002,
+      "grad_norm": 0.32442960143089294,
+      "learning_rate": 2.157175699762886e-06,
+      "loss": 0.3018,
+      "step": 8627
+    },
+    {
+      "epoch": 2.170020120724346,
+      "grad_norm": 0.2717970013618469,
+      "learning_rate": 2.155972097773231e-06,
+      "loss": 0.3407,
+      "step": 8628
+    },
+    {
+      "epoch": 2.170271629778672,
+      "grad_norm": 0.2874637842178345,
+      "learning_rate": 2.1547687393567736e-06,
+      "loss": 0.2937,
+      "step": 8629
+    },
+    {
+      "epoch": 2.170523138832998,
+      "grad_norm": 0.28952640295028687,
+      "learning_rate": 2.153565624616576e-06,
+      "loss": 0.3032,
+      "step": 8630
+    },
+    {
+      "epoch": 2.170774647887324,
+      "grad_norm": 0.30068325996398926,
+      "learning_rate": 2.1523627536556728e-06,
+      "loss": 0.344,
+      "step": 8631
+    },
+    {
+      "epoch": 2.17102615694165,
+      "grad_norm": 0.2989504933357239,
+      "learning_rate": 2.1511601265770876e-06,
+      "loss": 0.3452,
+      "step": 8632
+    },
+    {
+      "epoch": 2.171277665995976,
+      "grad_norm": 0.30707207322120667,
+      "learning_rate": 2.1499577434838132e-06,
+      "loss": 0.2972,
+      "step": 8633
+    },
+    {
+      "epoch": 2.171529175050302,
+      "grad_norm": 0.30241820216178894,
+      "learning_rate": 2.14875560447883e-06,
+      "loss": 0.3093,
+      "step": 8634
+    },
+    {
+      "epoch": 2.1717806841046277,
+      "grad_norm": 0.28262701630592346,
+      "learning_rate": 2.14755370966509e-06,
+      "loss": 0.3204,
+      "step": 8635
+    },
+    {
+      "epoch": 2.172032193158954,
+      "grad_norm": 0.3182533085346222,
+      "learning_rate": 2.146352059145532e-06,
+      "loss": 0.3132,
+      "step": 8636
+    },
+    {
+      "epoch": 2.1722837022132797,
+      "grad_norm": 0.30099421739578247,
+      "learning_rate": 2.1451506530230654e-06,
+      "loss": 0.3211,
+      "step": 8637
+    },
+    {
+      "epoch": 2.1725352112676055,
+      "grad_norm": 0.28418776392936707,
+      "learning_rate": 2.1439494914005877e-06,
+      "loss": 0.3228,
+      "step": 8638
+    },
+    {
+      "epoch": 2.1727867203219318,
+      "grad_norm": 0.31524962186813354,
+      "learning_rate": 2.1427485743809667e-06,
+      "loss": 0.3276,
+      "step": 8639
+    },
+    {
+      "epoch": 2.1730382293762576,
+      "grad_norm": 0.28617793321609497,
+      "learning_rate": 2.141547902067056e-06,
+      "loss": 0.3266,
+      "step": 8640
+    },
+    {
+      "epoch": 2.1732897384305834,
+      "grad_norm": 0.2999754548072815,
+      "learning_rate": 2.1403474745616863e-06,
+      "loss": 0.3383,
+      "step": 8641
+    },
+    {
+      "epoch": 2.1735412474849096,
+      "grad_norm": 0.3075141906738281,
+      "learning_rate": 2.1391472919676637e-06,
+      "loss": 0.313,
+      "step": 8642
+    },
+    {
+      "epoch": 2.1737927565392354,
+      "grad_norm": 0.3021905720233917,
+      "learning_rate": 2.13794735438778e-06,
+      "loss": 0.3151,
+      "step": 8643
+    },
+    {
+      "epoch": 2.1740442655935612,
+      "grad_norm": 0.3037421703338623,
+      "learning_rate": 2.136747661924799e-06,
+      "loss": 0.3213,
+      "step": 8644
+    },
+    {
+      "epoch": 2.1742957746478875,
+      "grad_norm": 0.3154190480709076,
+      "learning_rate": 2.1355482146814693e-06,
+      "loss": 0.3147,
+      "step": 8645
+    },
+    {
+      "epoch": 2.1745472837022133,
+      "grad_norm": 0.3106088638305664,
+      "learning_rate": 2.1343490127605136e-06,
+      "loss": 0.3102,
+      "step": 8646
+    },
+    {
+      "epoch": 2.174798792756539,
+      "grad_norm": 0.3053724765777588,
+      "learning_rate": 2.1331500562646396e-06,
+      "loss": 0.3177,
+      "step": 8647
+    },
+    {
+      "epoch": 2.1750503018108653,
+      "grad_norm": 0.30620327591896057,
+      "learning_rate": 2.1319513452965264e-06,
+      "loss": 0.314,
+      "step": 8648
+    },
+    {
+      "epoch": 2.175301810865191,
+      "grad_norm": 0.291100412607193,
+      "learning_rate": 2.13075287995884e-06,
+      "loss": 0.3018,
+      "step": 8649
+    },
+    {
+      "epoch": 2.175553319919517,
+      "grad_norm": 0.3090902864933014,
+      "learning_rate": 2.129554660354217e-06,
+      "loss": 0.3476,
+      "step": 8650
+    },
+    {
+      "epoch": 2.175804828973843,
+      "grad_norm": 0.27171772718429565,
+      "learning_rate": 2.1283566865852824e-06,
+      "loss": 0.3119,
+      "step": 8651
+    },
+    {
+      "epoch": 2.176056338028169,
+      "grad_norm": 0.31004947423934937,
+      "learning_rate": 2.1271589587546303e-06,
+      "loss": 0.3367,
+      "step": 8652
+    },
+    {
+      "epoch": 2.176307847082495,
+      "grad_norm": 0.2942144274711609,
+      "learning_rate": 2.1259614769648434e-06,
+      "loss": 0.3417,
+      "step": 8653
+    },
+    {
+      "epoch": 2.176559356136821,
+      "grad_norm": 0.29157233238220215,
+      "learning_rate": 2.124764241318474e-06,
+      "loss": 0.3177,
+      "step": 8654
+    },
+    {
+      "epoch": 2.176810865191147,
+      "grad_norm": 0.34676676988601685,
+      "learning_rate": 2.1235672519180615e-06,
+      "loss": 0.3086,
+      "step": 8655
+    },
+    {
+      "epoch": 2.1770623742454727,
+      "grad_norm": 0.30454134941101074,
+      "learning_rate": 2.1223705088661174e-06,
+      "loss": 0.3145,
+      "step": 8656
+    },
+    {
+      "epoch": 2.177313883299799,
+      "grad_norm": 0.2989632487297058,
+      "learning_rate": 2.121174012265138e-06,
+      "loss": 0.3293,
+      "step": 8657
+    },
+    {
+      "epoch": 2.1775653923541247,
+      "grad_norm": 0.3130037784576416,
+      "learning_rate": 2.119977762217594e-06,
+      "loss": 0.3147,
+      "step": 8658
+    },
+    {
+      "epoch": 2.1778169014084505,
+      "grad_norm": 0.28079167008399963,
+      "learning_rate": 2.118781758825938e-06,
+      "loss": 0.3225,
+      "step": 8659
+    },
+    {
+      "epoch": 2.1780684104627768,
+      "grad_norm": 0.30156850814819336,
+      "learning_rate": 2.117586002192598e-06,
+      "loss": 0.3157,
+      "step": 8660
+    },
+    {
+      "epoch": 2.1783199195171026,
+      "grad_norm": 0.3073710799217224,
+      "learning_rate": 2.1163904924199865e-06,
+      "loss": 0.3166,
+      "step": 8661
+    },
+    {
+      "epoch": 2.1785714285714284,
+      "grad_norm": 0.3458421528339386,
+      "learning_rate": 2.1151952296104876e-06,
+      "loss": 0.3273,
+      "step": 8662
+    },
+    {
+      "epoch": 2.1788229376257546,
+      "grad_norm": 0.3005492687225342,
+      "learning_rate": 2.1140002138664718e-06,
+      "loss": 0.3192,
+      "step": 8663
+    },
+    {
+      "epoch": 2.1790744466800804,
+      "grad_norm": 0.301626056432724,
+      "learning_rate": 2.1128054452902812e-06,
+      "loss": 0.2996,
+      "step": 8664
+    },
+    {
+      "epoch": 2.1793259557344062,
+      "grad_norm": 0.28979891538619995,
+      "learning_rate": 2.111610923984244e-06,
+      "loss": 0.3323,
+      "step": 8665
+    },
+    {
+      "epoch": 2.1795774647887325,
+      "grad_norm": 0.32507914304733276,
+      "learning_rate": 2.1104166500506596e-06,
+      "loss": 0.3134,
+      "step": 8666
+    },
+    {
+      "epoch": 2.1798289738430583,
+      "grad_norm": 0.30215296149253845,
+      "learning_rate": 2.1092226235918135e-06,
+      "loss": 0.343,
+      "step": 8667
+    },
+    {
+      "epoch": 2.1800804828973845,
+      "grad_norm": 0.3291065990924835,
+      "learning_rate": 2.1080288447099635e-06,
+      "loss": 0.3444,
+      "step": 8668
+    },
+    {
+      "epoch": 2.1803319919517103,
+      "grad_norm": 0.29154881834983826,
+      "learning_rate": 2.106835313507352e-06,
+      "loss": 0.3347,
+      "step": 8669
+    },
+    {
+      "epoch": 2.180583501006036,
+      "grad_norm": 0.3173246383666992,
+      "learning_rate": 2.1056420300861953e-06,
+      "loss": 0.3234,
+      "step": 8670
+    },
+    {
+      "epoch": 2.1808350100603624,
+      "grad_norm": 0.29975295066833496,
+      "learning_rate": 2.104448994548693e-06,
+      "loss": 0.3172,
+      "step": 8671
+    },
+    {
+      "epoch": 2.181086519114688,
+      "grad_norm": 0.29570671916007996,
+      "learning_rate": 2.103256206997018e-06,
+      "loss": 0.3228,
+      "step": 8672
+    },
+    {
+      "epoch": 2.181338028169014,
+      "grad_norm": 0.3010045289993286,
+      "learning_rate": 2.1020636675333273e-06,
+      "loss": 0.3299,
+      "step": 8673
+    },
+    {
+      "epoch": 2.1815895372233403,
+      "grad_norm": 0.32294243574142456,
+      "learning_rate": 2.1008713762597554e-06,
+      "loss": 0.3119,
+      "step": 8674
+    },
+    {
+      "epoch": 2.181841046277666,
+      "grad_norm": 0.3337547779083252,
+      "learning_rate": 2.0996793332784116e-06,
+      "loss": 0.3317,
+      "step": 8675
+    },
+    {
+      "epoch": 2.182092555331992,
+      "grad_norm": 0.3114114999771118,
+      "learning_rate": 2.0984875386913904e-06,
+      "loss": 0.3526,
+      "step": 8676
+    },
+    {
+      "epoch": 2.182344064386318,
+      "grad_norm": 0.31616508960723877,
+      "learning_rate": 2.097295992600758e-06,
+      "loss": 0.2981,
+      "step": 8677
+    },
+    {
+      "epoch": 2.182595573440644,
+      "grad_norm": 0.2805394232273102,
+      "learning_rate": 2.0961046951085662e-06,
+      "loss": 0.3403,
+      "step": 8678
+    },
+    {
+      "epoch": 2.1828470824949697,
+      "grad_norm": 0.28356558084487915,
+      "learning_rate": 2.094913646316839e-06,
+      "loss": 0.3279,
+      "step": 8679
+    },
+    {
+      "epoch": 2.183098591549296,
+      "grad_norm": 0.2977490723133087,
+      "learning_rate": 2.0937228463275854e-06,
+      "loss": 0.3316,
+      "step": 8680
+    },
+    {
+      "epoch": 2.183350100603622,
+      "grad_norm": 0.27247682213783264,
+      "learning_rate": 2.092532295242787e-06,
+      "loss": 0.3144,
+      "step": 8681
+    },
+    {
+      "epoch": 2.1836016096579476,
+      "grad_norm": 0.307102233171463,
+      "learning_rate": 2.0913419931644095e-06,
+      "loss": 0.3271,
+      "step": 8682
+    },
+    {
+      "epoch": 2.183853118712274,
+      "grad_norm": 0.2990267276763916,
+      "learning_rate": 2.0901519401943924e-06,
+      "loss": 0.2964,
+      "step": 8683
+    },
+    {
+      "epoch": 2.1841046277665996,
+      "grad_norm": 0.30906108021736145,
+      "learning_rate": 2.088962136434658e-06,
+      "loss": 0.3318,
+      "step": 8684
+    },
+    {
+      "epoch": 2.1843561368209254,
+      "grad_norm": 0.2764362394809723,
+      "learning_rate": 2.0877725819871065e-06,
+      "loss": 0.3023,
+      "step": 8685
+    },
+    {
+      "epoch": 2.1846076458752517,
+      "grad_norm": 0.3090154826641083,
+      "learning_rate": 2.0865832769536125e-06,
+      "loss": 0.346,
+      "step": 8686
+    },
+    {
+      "epoch": 2.1848591549295775,
+      "grad_norm": 0.3105575442314148,
+      "learning_rate": 2.0853942214360365e-06,
+      "loss": 0.3426,
+      "step": 8687
+    },
+    {
+      "epoch": 2.1851106639839033,
+      "grad_norm": 0.2777365744113922,
+      "learning_rate": 2.0842054155362105e-06,
+      "loss": 0.3196,
+      "step": 8688
+    },
+    {
+      "epoch": 2.1853621730382295,
+      "grad_norm": 0.29215675592422485,
+      "learning_rate": 2.0830168593559513e-06,
+      "loss": 0.3114,
+      "step": 8689
+    },
+    {
+      "epoch": 2.1856136820925554,
+      "grad_norm": 0.29614973068237305,
+      "learning_rate": 2.081828552997047e-06,
+      "loss": 0.3229,
+      "step": 8690
+    },
+    {
+      "epoch": 2.185865191146881,
+      "grad_norm": 0.30876439809799194,
+      "learning_rate": 2.0806404965612737e-06,
+      "loss": 0.3022,
+      "step": 8691
+    },
+    {
+      "epoch": 2.1861167002012074,
+      "grad_norm": 0.3041656017303467,
+      "learning_rate": 2.0794526901503757e-06,
+      "loss": 0.3057,
+      "step": 8692
+    },
+    {
+      "epoch": 2.186368209255533,
+      "grad_norm": 0.29282814264297485,
+      "learning_rate": 2.0782651338660862e-06,
+      "loss": 0.3257,
+      "step": 8693
+    },
+    {
+      "epoch": 2.186619718309859,
+      "grad_norm": 0.3103329837322235,
+      "learning_rate": 2.077077827810108e-06,
+      "loss": 0.3163,
+      "step": 8694
+    },
+    {
+      "epoch": 2.1868712273641853,
+      "grad_norm": 0.28378820419311523,
+      "learning_rate": 2.075890772084128e-06,
+      "loss": 0.3226,
+      "step": 8695
+    },
+    {
+      "epoch": 2.187122736418511,
+      "grad_norm": 0.30652692914009094,
+      "learning_rate": 2.074703966789812e-06,
+      "loss": 0.3087,
+      "step": 8696
+    },
+    {
+      "epoch": 2.187374245472837,
+      "grad_norm": 0.3030233681201935,
+      "learning_rate": 2.073517412028799e-06,
+      "loss": 0.3096,
+      "step": 8697
+    },
+    {
+      "epoch": 2.187625754527163,
+      "grad_norm": 0.28076255321502686,
+      "learning_rate": 2.072331107902713e-06,
+      "loss": 0.3153,
+      "step": 8698
+    },
+    {
+      "epoch": 2.187877263581489,
+      "grad_norm": 0.2828161418437958,
+      "learning_rate": 2.0711450545131505e-06,
+      "loss": 0.3112,
+      "step": 8699
+    },
+    {
+      "epoch": 2.1881287726358147,
+      "grad_norm": 0.29777970910072327,
+      "learning_rate": 2.0699592519616934e-06,
+      "loss": 0.3253,
+      "step": 8700
+    },
+    {
+      "epoch": 2.188380281690141,
+      "grad_norm": 0.32333430647850037,
+      "learning_rate": 2.0687737003498944e-06,
+      "loss": 0.3359,
+      "step": 8701
+    },
+    {
+      "epoch": 2.188631790744467,
+      "grad_norm": 0.30716678500175476,
+      "learning_rate": 2.0675883997792913e-06,
+      "loss": 0.3126,
+      "step": 8702
+    },
+    {
+      "epoch": 2.1888832997987926,
+      "grad_norm": 0.28720593452453613,
+      "learning_rate": 2.0664033503513953e-06,
+      "loss": 0.3196,
+      "step": 8703
+    },
+    {
+      "epoch": 2.189134808853119,
+      "grad_norm": 0.29374513030052185,
+      "learning_rate": 2.0652185521677016e-06,
+      "loss": 0.3285,
+      "step": 8704
+    },
+    {
+      "epoch": 2.1893863179074446,
+      "grad_norm": 0.30496105551719666,
+      "learning_rate": 2.064034005329677e-06,
+      "loss": 0.3008,
+      "step": 8705
+    },
+    {
+      "epoch": 2.1896378269617705,
+      "grad_norm": 0.29388949275016785,
+      "learning_rate": 2.0628497099387727e-06,
+      "loss": 0.3009,
+      "step": 8706
+    },
+    {
+      "epoch": 2.1898893360160967,
+      "grad_norm": 0.3176174461841583,
+      "learning_rate": 2.061665666096418e-06,
+      "loss": 0.3277,
+      "step": 8707
+    },
+    {
+      "epoch": 2.1901408450704225,
+      "grad_norm": 0.3095850646495819,
+      "learning_rate": 2.0604818739040143e-06,
+      "loss": 0.3092,
+      "step": 8708
+    },
+    {
+      "epoch": 2.1903923541247483,
+      "grad_norm": 0.28293004631996155,
+      "learning_rate": 2.0592983334629506e-06,
+      "loss": 0.3213,
+      "step": 8709
+    },
+    {
+      "epoch": 2.1906438631790746,
+      "grad_norm": 0.2949610650539398,
+      "learning_rate": 2.0581150448745863e-06,
+      "loss": 0.3194,
+      "step": 8710
+    },
+    {
+      "epoch": 2.1908953722334004,
+      "grad_norm": 0.299869179725647,
+      "learning_rate": 2.0569320082402654e-06,
+      "loss": 0.3376,
+      "step": 8711
+    },
+    {
+      "epoch": 2.191146881287726,
+      "grad_norm": 0.28059983253479004,
+      "learning_rate": 2.055749223661305e-06,
+      "loss": 0.3159,
+      "step": 8712
+    },
+    {
+      "epoch": 2.1913983903420524,
+      "grad_norm": 0.3038329780101776,
+      "learning_rate": 2.0545666912390053e-06,
+      "loss": 0.3033,
+      "step": 8713
+    },
+    {
+      "epoch": 2.191649899396378,
+      "grad_norm": 0.3102431893348694,
+      "learning_rate": 2.053384411074641e-06,
+      "loss": 0.3305,
+      "step": 8714
+    },
+    {
+      "epoch": 2.191901408450704,
+      "grad_norm": 0.30974650382995605,
+      "learning_rate": 2.0522023832694694e-06,
+      "loss": 0.3425,
+      "step": 8715
+    },
+    {
+      "epoch": 2.1921529175050303,
+      "grad_norm": 0.31163522601127625,
+      "learning_rate": 2.051020607924721e-06,
+      "loss": 0.3385,
+      "step": 8716
+    },
+    {
+      "epoch": 2.192404426559356,
+      "grad_norm": 0.29782193899154663,
+      "learning_rate": 2.049839085141608e-06,
+      "loss": 0.3237,
+      "step": 8717
+    },
+    {
+      "epoch": 2.192655935613682,
+      "grad_norm": 0.3062836229801178,
+      "learning_rate": 2.048657815021323e-06,
+      "loss": 0.3187,
+      "step": 8718
+    },
+    {
+      "epoch": 2.192907444668008,
+      "grad_norm": 0.32062938809394836,
+      "learning_rate": 2.0474767976650313e-06,
+      "loss": 0.2985,
+      "step": 8719
+    },
+    {
+      "epoch": 2.193158953722334,
+      "grad_norm": 0.3006410300731659,
+      "learning_rate": 2.0462960331738824e-06,
+      "loss": 0.3242,
+      "step": 8720
+    },
+    {
+      "epoch": 2.1934104627766597,
+      "grad_norm": 0.28875410556793213,
+      "learning_rate": 2.0451155216489983e-06,
+      "loss": 0.3338,
+      "step": 8721
+    },
+    {
+      "epoch": 2.193661971830986,
+      "grad_norm": 0.3076934218406677,
+      "learning_rate": 2.043935263191486e-06,
+      "loss": 0.3354,
+      "step": 8722
+    },
+    {
+      "epoch": 2.193913480885312,
+      "grad_norm": 0.301831990480423,
+      "learning_rate": 2.0427552579024234e-06,
+      "loss": 0.3407,
+      "step": 8723
+    },
+    {
+      "epoch": 2.1941649899396376,
+      "grad_norm": 0.31272050738334656,
+      "learning_rate": 2.041575505882874e-06,
+      "loss": 0.3296,
+      "step": 8724
+    },
+    {
+      "epoch": 2.194416498993964,
+      "grad_norm": 0.29415902495384216,
+      "learning_rate": 2.040396007233873e-06,
+      "loss": 0.315,
+      "step": 8725
+    },
+    {
+      "epoch": 2.1946680080482897,
+      "grad_norm": 0.32999396324157715,
+      "learning_rate": 2.039216762056439e-06,
+      "loss": 0.3289,
+      "step": 8726
+    },
+    {
+      "epoch": 2.1949195171026155,
+      "grad_norm": 0.3093070089817047,
+      "learning_rate": 2.0380377704515687e-06,
+      "loss": 0.3259,
+      "step": 8727
+    },
+    {
+      "epoch": 2.1951710261569417,
+      "grad_norm": 0.31014779210090637,
+      "learning_rate": 2.0368590325202315e-06,
+      "loss": 0.318,
+      "step": 8728
+    },
+    {
+      "epoch": 2.1954225352112675,
+      "grad_norm": 0.29741162061691284,
+      "learning_rate": 2.035680548363382e-06,
+      "loss": 0.3241,
+      "step": 8729
+    },
+    {
+      "epoch": 2.1956740442655938,
+      "grad_norm": 0.27723386883735657,
+      "learning_rate": 2.0345023180819474e-06,
+      "loss": 0.3286,
+      "step": 8730
+    },
+    {
+      "epoch": 2.1959255533199196,
+      "grad_norm": 0.31631845235824585,
+      "learning_rate": 2.033324341776839e-06,
+      "loss": 0.2967,
+      "step": 8731
+    },
+    {
+      "epoch": 2.1961770623742454,
+      "grad_norm": 0.3349016606807709,
+      "learning_rate": 2.03214661954894e-06,
+      "loss": 0.3171,
+      "step": 8732
+    },
+    {
+      "epoch": 2.1964285714285716,
+      "grad_norm": 0.3033004701137543,
+      "learning_rate": 2.030969151499117e-06,
+      "loss": 0.3221,
+      "step": 8733
+    },
+    {
+      "epoch": 2.1966800804828974,
+      "grad_norm": 0.29788362979888916,
+      "learning_rate": 2.0297919377282106e-06,
+      "loss": 0.3317,
+      "step": 8734
+    },
+    {
+      "epoch": 2.1969315895372232,
+      "grad_norm": 0.2985301911830902,
+      "learning_rate": 2.0286149783370453e-06,
+      "loss": 0.3315,
+      "step": 8735
+    },
+    {
+      "epoch": 2.1971830985915495,
+      "grad_norm": 0.30903440713882446,
+      "learning_rate": 2.027438273426416e-06,
+      "loss": 0.3339,
+      "step": 8736
+    },
+    {
+      "epoch": 2.1974346076458753,
+      "grad_norm": 0.33624449372291565,
+      "learning_rate": 2.0262618230971023e-06,
+      "loss": 0.3379,
+      "step": 8737
+    },
+    {
+      "epoch": 2.197686116700201,
+      "grad_norm": 0.31500041484832764,
+      "learning_rate": 2.0250856274498617e-06,
+      "loss": 0.3312,
+      "step": 8738
+    },
+    {
+      "epoch": 2.1979376257545273,
+      "grad_norm": 0.3018837571144104,
+      "learning_rate": 2.0239096865854243e-06,
+      "loss": 0.3333,
+      "step": 8739
+    },
+    {
+      "epoch": 2.198189134808853,
+      "grad_norm": 0.305803120136261,
+      "learning_rate": 2.0227340006045056e-06,
+      "loss": 0.3403,
+      "step": 8740
+    },
+    {
+      "epoch": 2.198440643863179,
+      "grad_norm": 0.2978799045085907,
+      "learning_rate": 2.021558569607792e-06,
+      "loss": 0.3093,
+      "step": 8741
+    },
+    {
+      "epoch": 2.198692152917505,
+      "grad_norm": 0.2844066023826599,
+      "learning_rate": 2.020383393695956e-06,
+      "loss": 0.3106,
+      "step": 8742
+    },
+    {
+      "epoch": 2.198943661971831,
+      "grad_norm": 0.30605459213256836,
+      "learning_rate": 2.01920847296964e-06,
+      "loss": 0.3195,
+      "step": 8743
+    },
+    {
+      "epoch": 2.199195171026157,
+      "grad_norm": 0.30778858065605164,
+      "learning_rate": 2.0180338075294726e-06,
+      "loss": 0.3221,
+      "step": 8744
+    },
+    {
+      "epoch": 2.199446680080483,
+      "grad_norm": 0.29119858145713806,
+      "learning_rate": 2.016859397476052e-06,
+      "loss": 0.3172,
+      "step": 8745
+    },
+    {
+      "epoch": 2.199698189134809,
+      "grad_norm": 0.3088303804397583,
+      "learning_rate": 2.0156852429099638e-06,
+      "loss": 0.3166,
+      "step": 8746
+    },
+    {
+      "epoch": 2.1999496981891347,
+      "grad_norm": 0.281408429145813,
+      "learning_rate": 2.014511343931763e-06,
+      "loss": 0.3352,
+      "step": 8747
+    },
+    {
+      "epoch": 2.200201207243461,
+      "grad_norm": 0.30251288414001465,
+      "learning_rate": 2.0133377006419885e-06,
+      "loss": 0.3042,
+      "step": 8748
+    },
+    {
+      "epoch": 2.2004527162977867,
+      "grad_norm": 0.3207908868789673,
+      "learning_rate": 2.0121643131411568e-06,
+      "loss": 0.3338,
+      "step": 8749
+    },
+    {
+      "epoch": 2.2007042253521125,
+      "grad_norm": 0.3297126889228821,
+      "learning_rate": 2.0109911815297585e-06,
+      "loss": 0.3362,
+      "step": 8750
+    },
+    {
+      "epoch": 2.2009557344064388,
+      "grad_norm": 0.29778042435646057,
+      "learning_rate": 2.0098183059082675e-06,
+      "loss": 0.3456,
+      "step": 8751
+    },
+    {
+      "epoch": 2.2012072434607646,
+      "grad_norm": 0.30129680037498474,
+      "learning_rate": 2.008645686377131e-06,
+      "loss": 0.3625,
+      "step": 8752
+    },
+    {
+      "epoch": 2.2014587525150904,
+      "grad_norm": 0.2970339059829712,
+      "learning_rate": 2.007473323036779e-06,
+      "loss": 0.3177,
+      "step": 8753
+    },
+    {
+      "epoch": 2.2017102615694166,
+      "grad_norm": 0.3131774961948395,
+      "learning_rate": 2.0063012159876138e-06,
+      "loss": 0.3363,
+      "step": 8754
+    },
+    {
+      "epoch": 2.2019617706237424,
+      "grad_norm": 0.31431764364242554,
+      "learning_rate": 2.005129365330023e-06,
+      "loss": 0.3446,
+      "step": 8755
+    },
+    {
+      "epoch": 2.2022132796780682,
+      "grad_norm": 0.2815970182418823,
+      "learning_rate": 2.0039577711643642e-06,
+      "loss": 0.3325,
+      "step": 8756
+    },
+    {
+      "epoch": 2.2024647887323945,
+      "grad_norm": 0.2978397309780121,
+      "learning_rate": 2.002786433590981e-06,
+      "loss": 0.3338,
+      "step": 8757
+    },
+    {
+      "epoch": 2.2027162977867203,
+      "grad_norm": 0.3014795780181885,
+      "learning_rate": 2.001615352710188e-06,
+      "loss": 0.3291,
+      "step": 8758
+    },
+    {
+      "epoch": 2.202967806841046,
+      "grad_norm": 0.3280890882015228,
+      "learning_rate": 2.0004445286222818e-06,
+      "loss": 0.3253,
+      "step": 8759
+    },
+    {
+      "epoch": 2.2032193158953723,
+      "grad_norm": 0.3170720636844635,
+      "learning_rate": 1.999273961427538e-06,
+      "loss": 0.3304,
+      "step": 8760
+    },
+    {
+      "epoch": 2.203470824949698,
+      "grad_norm": 0.31641075015068054,
+      "learning_rate": 1.9981036512262054e-06,
+      "loss": 0.3233,
+      "step": 8761
+    },
+    {
+      "epoch": 2.203722334004024,
+      "grad_norm": 0.30692046880722046,
+      "learning_rate": 1.9969335981185173e-06,
+      "loss": 0.3543,
+      "step": 8762
+    },
+    {
+      "epoch": 2.20397384305835,
+      "grad_norm": 0.31193432211875916,
+      "learning_rate": 1.9957638022046773e-06,
+      "loss": 0.3305,
+      "step": 8763
+    },
+    {
+      "epoch": 2.204225352112676,
+      "grad_norm": 0.3167397677898407,
+      "learning_rate": 1.9945942635848745e-06,
+      "loss": 0.3031,
+      "step": 8764
+    },
+    {
+      "epoch": 2.204476861167002,
+      "grad_norm": 0.29315993189811707,
+      "learning_rate": 1.9934249823592703e-06,
+      "loss": 0.3366,
+      "step": 8765
+    },
+    {
+      "epoch": 2.204728370221328,
+      "grad_norm": 0.2992209196090698,
+      "learning_rate": 1.992255958628009e-06,
+      "loss": 0.3327,
+      "step": 8766
+    },
+    {
+      "epoch": 2.204979879275654,
+      "grad_norm": 0.29840826988220215,
+      "learning_rate": 1.9910871924912063e-06,
+      "loss": 0.3068,
+      "step": 8767
+    },
+    {
+      "epoch": 2.20523138832998,
+      "grad_norm": 0.318357914686203,
+      "learning_rate": 1.989918684048964e-06,
+      "loss": 0.3294,
+      "step": 8768
+    },
+    {
+      "epoch": 2.205482897384306,
+      "grad_norm": 0.28239601850509644,
+      "learning_rate": 1.9887504334013534e-06,
+      "loss": 0.3296,
+      "step": 8769
+    },
+    {
+      "epoch": 2.2057344064386317,
+      "grad_norm": 0.32247522473335266,
+      "learning_rate": 1.9875824406484318e-06,
+      "loss": 0.332,
+      "step": 8770
+    },
+    {
+      "epoch": 2.205985915492958,
+      "grad_norm": 0.2952177822589874,
+      "learning_rate": 1.986414705890226e-06,
+      "loss": 0.3287,
+      "step": 8771
+    },
+    {
+      "epoch": 2.2062374245472838,
+      "grad_norm": 0.3215632438659668,
+      "learning_rate": 1.9852472292267505e-06,
+      "loss": 0.3458,
+      "step": 8772
+    },
+    {
+      "epoch": 2.2064889336016096,
+      "grad_norm": 0.30382493138313293,
+      "learning_rate": 1.9840800107579872e-06,
+      "loss": 0.3165,
+      "step": 8773
+    },
+    {
+      "epoch": 2.206740442655936,
+      "grad_norm": 0.31115108728408813,
+      "learning_rate": 1.9829130505839058e-06,
+      "loss": 0.3306,
+      "step": 8774
+    },
+    {
+      "epoch": 2.2069919517102616,
+      "grad_norm": 0.29950037598609924,
+      "learning_rate": 1.9817463488044446e-06,
+      "loss": 0.3125,
+      "step": 8775
+    },
+    {
+      "epoch": 2.2072434607645874,
+      "grad_norm": 0.2986210882663727,
+      "learning_rate": 1.9805799055195264e-06,
+      "loss": 0.3062,
+      "step": 8776
+    },
+    {
+      "epoch": 2.2074949698189137,
+      "grad_norm": 0.3211692273616791,
+      "learning_rate": 1.9794137208290516e-06,
+      "loss": 0.3129,
+      "step": 8777
+    },
+    {
+      "epoch": 2.2077464788732395,
+      "grad_norm": 0.29951202869415283,
+      "learning_rate": 1.9782477948328933e-06,
+      "loss": 0.3337,
+      "step": 8778
+    },
+    {
+      "epoch": 2.2079979879275653,
+      "grad_norm": 0.3311370611190796,
+      "learning_rate": 1.9770821276309093e-06,
+      "loss": 0.3185,
+      "step": 8779
+    },
+    {
+      "epoch": 2.2082494969818915,
+      "grad_norm": 0.31037113070487976,
+      "learning_rate": 1.9759167193229277e-06,
+      "loss": 0.3226,
+      "step": 8780
+    },
+    {
+      "epoch": 2.2085010060362174,
+      "grad_norm": 0.315602570772171,
+      "learning_rate": 1.974751570008762e-06,
+      "loss": 0.334,
+      "step": 8781
+    },
+    {
+      "epoch": 2.208752515090543,
+      "grad_norm": 0.2930223345756531,
+      "learning_rate": 1.9735866797881977e-06,
+      "loss": 0.2912,
+      "step": 8782
+    },
+    {
+      "epoch": 2.2090040241448694,
+      "grad_norm": 0.2838136553764343,
+      "learning_rate": 1.972422048761002e-06,
+      "loss": 0.3148,
+      "step": 8783
+    },
+    {
+      "epoch": 2.209255533199195,
+      "grad_norm": 0.3084590435028076,
+      "learning_rate": 1.9712576770269155e-06,
+      "loss": 0.3255,
+      "step": 8784
+    },
+    {
+      "epoch": 2.209507042253521,
+      "grad_norm": 0.3039499819278717,
+      "learning_rate": 1.9700935646856634e-06,
+      "loss": 0.341,
+      "step": 8785
+    },
+    {
+      "epoch": 2.2097585513078473,
+      "grad_norm": 0.30746522545814514,
+      "learning_rate": 1.9689297118369403e-06,
+      "loss": 0.3122,
+      "step": 8786
+    },
+    {
+      "epoch": 2.210010060362173,
+      "grad_norm": 0.31201913952827454,
+      "learning_rate": 1.967766118580427e-06,
+      "loss": 0.2904,
+      "step": 8787
+    },
+    {
+      "epoch": 2.210261569416499,
+      "grad_norm": 0.3182002007961273,
+      "learning_rate": 1.9666027850157745e-06,
+      "loss": 0.3133,
+      "step": 8788
+    },
+    {
+      "epoch": 2.210513078470825,
+      "grad_norm": 0.3027884066104889,
+      "learning_rate": 1.965439711242618e-06,
+      "loss": 0.307,
+      "step": 8789
+    },
+    {
+      "epoch": 2.210764587525151,
+      "grad_norm": 0.3081648051738739,
+      "learning_rate": 1.964276897360565e-06,
+      "loss": 0.3231,
+      "step": 8790
+    },
+    {
+      "epoch": 2.2110160965794767,
+      "grad_norm": 0.30565327405929565,
+      "learning_rate": 1.9631143434692054e-06,
+      "loss": 0.3237,
+      "step": 8791
+    },
+    {
+      "epoch": 2.211267605633803,
+      "grad_norm": 0.2992105484008789,
+      "learning_rate": 1.9619520496681015e-06,
+      "loss": 0.3361,
+      "step": 8792
+    },
+    {
+      "epoch": 2.211519114688129,
+      "grad_norm": 0.28107115626335144,
+      "learning_rate": 1.960790016056801e-06,
+      "loss": 0.3208,
+      "step": 8793
+    },
+    {
+      "epoch": 2.2117706237424546,
+      "grad_norm": 0.31731727719306946,
+      "learning_rate": 1.9596282427348206e-06,
+      "loss": 0.3032,
+      "step": 8794
+    },
+    {
+      "epoch": 2.212022132796781,
+      "grad_norm": 0.30339574813842773,
+      "learning_rate": 1.958466729801662e-06,
+      "loss": 0.3053,
+      "step": 8795
+    },
+    {
+      "epoch": 2.2122736418511066,
+      "grad_norm": 0.29197800159454346,
+      "learning_rate": 1.957305477356799e-06,
+      "loss": 0.3513,
+      "step": 8796
+    },
+    {
+      "epoch": 2.2125251509054324,
+      "grad_norm": 0.2851543426513672,
+      "learning_rate": 1.956144485499688e-06,
+      "loss": 0.3059,
+      "step": 8797
+    },
+    {
+      "epoch": 2.2127766599597587,
+      "grad_norm": 0.2786988615989685,
+      "learning_rate": 1.9549837543297585e-06,
+      "loss": 0.3259,
+      "step": 8798
+    },
+    {
+      "epoch": 2.2130281690140845,
+      "grad_norm": 0.2844609022140503,
+      "learning_rate": 1.953823283946422e-06,
+      "loss": 0.3198,
+      "step": 8799
+    },
+    {
+      "epoch": 2.2132796780684103,
+      "grad_norm": 0.32058948278427124,
+      "learning_rate": 1.952663074449063e-06,
+      "loss": 0.3098,
+      "step": 8800
+    },
+    {
+      "epoch": 2.2135311871227366,
+      "grad_norm": 0.2962084412574768,
+      "learning_rate": 1.9515031259370493e-06,
+      "loss": 0.3343,
+      "step": 8801
+    },
+    {
+      "epoch": 2.2137826961770624,
+      "grad_norm": 0.2991352081298828,
+      "learning_rate": 1.95034343850972e-06,
+      "loss": 0.3374,
+      "step": 8802
+    },
+    {
+      "epoch": 2.214034205231388,
+      "grad_norm": 0.3038289248943329,
+      "learning_rate": 1.9491840122663965e-06,
+      "loss": 0.3438,
+      "step": 8803
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.3067372143268585,
+      "learning_rate": 1.948024847306378e-06,
+      "loss": 0.302,
+      "step": 8804
+    },
+    {
+      "epoch": 2.21453722334004,
+      "grad_norm": 0.27954307198524475,
+      "learning_rate": 1.946865943728936e-06,
+      "loss": 0.3406,
+      "step": 8805
+    },
+    {
+      "epoch": 2.214788732394366,
+      "grad_norm": 0.2837284207344055,
+      "learning_rate": 1.945707301633328e-06,
+      "loss": 0.3044,
+      "step": 8806
+    },
+    {
+      "epoch": 2.2150402414486923,
+      "grad_norm": 0.29118677973747253,
+      "learning_rate": 1.94454892111878e-06,
+      "loss": 0.3169,
+      "step": 8807
+    },
+    {
+      "epoch": 2.215291750503018,
+      "grad_norm": 0.301509827375412,
+      "learning_rate": 1.9433908022845046e-06,
+      "loss": 0.3166,
+      "step": 8808
+    },
+    {
+      "epoch": 2.215543259557344,
+      "grad_norm": 0.3074178099632263,
+      "learning_rate": 1.9422329452296825e-06,
+      "loss": 0.3188,
+      "step": 8809
+    },
+    {
+      "epoch": 2.21579476861167,
+      "grad_norm": 0.2955106496810913,
+      "learning_rate": 1.941075350053481e-06,
+      "loss": 0.3116,
+      "step": 8810
+    },
+    {
+      "epoch": 2.216046277665996,
+      "grad_norm": 0.2759988009929657,
+      "learning_rate": 1.9399180168550374e-06,
+      "loss": 0.311,
+      "step": 8811
+    },
+    {
+      "epoch": 2.2162977867203217,
+      "grad_norm": 0.3274937570095062,
+      "learning_rate": 1.9387609457334734e-06,
+      "loss": 0.3431,
+      "step": 8812
+    },
+    {
+      "epoch": 2.216549295774648,
+      "grad_norm": 0.2869795858860016,
+      "learning_rate": 1.937604136787882e-06,
+      "loss": 0.3158,
+      "step": 8813
+    },
+    {
+      "epoch": 2.216800804828974,
+      "grad_norm": 0.33137309551239014,
+      "learning_rate": 1.936447590117338e-06,
+      "loss": 0.3242,
+      "step": 8814
+    },
+    {
+      "epoch": 2.2170523138832996,
+      "grad_norm": 0.2800137996673584,
+      "learning_rate": 1.935291305820894e-06,
+      "loss": 0.3108,
+      "step": 8815
+    },
+    {
+      "epoch": 2.217303822937626,
+      "grad_norm": 0.29524585604667664,
+      "learning_rate": 1.9341352839975753e-06,
+      "loss": 0.3126,
+      "step": 8816
+    },
+    {
+      "epoch": 2.2175553319919517,
+      "grad_norm": 0.30382615327835083,
+      "learning_rate": 1.9329795247463913e-06,
+      "loss": 0.3184,
+      "step": 8817
+    },
+    {
+      "epoch": 2.2178068410462775,
+      "grad_norm": 0.3121460974216461,
+      "learning_rate": 1.9318240281663215e-06,
+      "loss": 0.3162,
+      "step": 8818
+    },
+    {
+      "epoch": 2.2180583501006037,
+      "grad_norm": 0.28468015789985657,
+      "learning_rate": 1.930668794356331e-06,
+      "loss": 0.3268,
+      "step": 8819
+    },
+    {
+      "epoch": 2.2183098591549295,
+      "grad_norm": 0.3230193257331848,
+      "learning_rate": 1.929513823415356e-06,
+      "loss": 0.309,
+      "step": 8820
+    },
+    {
+      "epoch": 2.2185613682092553,
+      "grad_norm": 0.30588406324386597,
+      "learning_rate": 1.928359115442314e-06,
+      "loss": 0.3162,
+      "step": 8821
+    },
+    {
+      "epoch": 2.2188128772635816,
+      "grad_norm": 0.2859831750392914,
+      "learning_rate": 1.9272046705360958e-06,
+      "loss": 0.3161,
+      "step": 8822
+    },
+    {
+      "epoch": 2.2190643863179074,
+      "grad_norm": 0.2836097478866577,
+      "learning_rate": 1.926050488795576e-06,
+      "loss": 0.3085,
+      "step": 8823
+    },
+    {
+      "epoch": 2.219315895372233,
+      "grad_norm": 0.29898855090141296,
+      "learning_rate": 1.9248965703196e-06,
+      "loss": 0.3115,
+      "step": 8824
+    },
+    {
+      "epoch": 2.2195674044265594,
+      "grad_norm": 0.3057454228401184,
+      "learning_rate": 1.9237429152069948e-06,
+      "loss": 0.2977,
+      "step": 8825
+    },
+    {
+      "epoch": 2.2198189134808852,
+      "grad_norm": 0.30841681361198425,
+      "learning_rate": 1.922589523556565e-06,
+      "loss": 0.3491,
+      "step": 8826
+    },
+    {
+      "epoch": 2.2200704225352115,
+      "grad_norm": 0.3082001507282257,
+      "learning_rate": 1.9214363954670895e-06,
+      "loss": 0.3165,
+      "step": 8827
+    },
+    {
+      "epoch": 2.2203219315895373,
+      "grad_norm": 0.3125540614128113,
+      "learning_rate": 1.9202835310373285e-06,
+      "loss": 0.3291,
+      "step": 8828
+    },
+    {
+      "epoch": 2.220573440643863,
+      "grad_norm": 0.3130517303943634,
+      "learning_rate": 1.9191309303660145e-06,
+      "loss": 0.3229,
+      "step": 8829
+    },
+    {
+      "epoch": 2.2208249496981893,
+      "grad_norm": 0.3235696852207184,
+      "learning_rate": 1.9179785935518647e-06,
+      "loss": 0.328,
+      "step": 8830
+    },
+    {
+      "epoch": 2.221076458752515,
+      "grad_norm": 0.30910971760749817,
+      "learning_rate": 1.9168265206935655e-06,
+      "loss": 0.3185,
+      "step": 8831
+    },
+    {
+      "epoch": 2.221327967806841,
+      "grad_norm": 0.282240092754364,
+      "learning_rate": 1.9156747118897878e-06,
+      "loss": 0.32,
+      "step": 8832
+    },
+    {
+      "epoch": 2.221579476861167,
+      "grad_norm": 0.30179286003112793,
+      "learning_rate": 1.914523167239174e-06,
+      "loss": 0.3176,
+      "step": 8833
+    },
+    {
+      "epoch": 2.221830985915493,
+      "grad_norm": 0.3033350110054016,
+      "learning_rate": 1.91337188684035e-06,
+      "loss": 0.3161,
+      "step": 8834
+    },
+    {
+      "epoch": 2.222082494969819,
+      "grad_norm": 0.30857256054878235,
+      "learning_rate": 1.9122208707919125e-06,
+      "loss": 0.3243,
+      "step": 8835
+    },
+    {
+      "epoch": 2.222334004024145,
+      "grad_norm": 0.3050358295440674,
+      "learning_rate": 1.9110701191924403e-06,
+      "loss": 0.3213,
+      "step": 8836
+    },
+    {
+      "epoch": 2.222585513078471,
+      "grad_norm": 0.3051794171333313,
+      "learning_rate": 1.9099196321404895e-06,
+      "loss": 0.3034,
+      "step": 8837
+    },
+    {
+      "epoch": 2.2228370221327967,
+      "grad_norm": 0.30538561940193176,
+      "learning_rate": 1.9087694097345895e-06,
+      "loss": 0.3213,
+      "step": 8838
+    },
+    {
+      "epoch": 2.223088531187123,
+      "grad_norm": 0.3018161952495575,
+      "learning_rate": 1.9076194520732523e-06,
+      "loss": 0.3263,
+      "step": 8839
+    },
+    {
+      "epoch": 2.2233400402414487,
+      "grad_norm": 0.3152303397655487,
+      "learning_rate": 1.9064697592549613e-06,
+      "loss": 0.323,
+      "step": 8840
+    },
+    {
+      "epoch": 2.2235915492957745,
+      "grad_norm": 0.2975950539112091,
+      "learning_rate": 1.9053203313781843e-06,
+      "loss": 0.3556,
+      "step": 8841
+    },
+    {
+      "epoch": 2.2238430583501008,
+      "grad_norm": 0.3171286880970001,
+      "learning_rate": 1.9041711685413588e-06,
+      "loss": 0.3144,
+      "step": 8842
+    },
+    {
+      "epoch": 2.2240945674044266,
+      "grad_norm": 0.2995269298553467,
+      "learning_rate": 1.903022270842907e-06,
+      "loss": 0.3153,
+      "step": 8843
+    },
+    {
+      "epoch": 2.2243460764587524,
+      "grad_norm": 0.3209341764450073,
+      "learning_rate": 1.9018736383812214e-06,
+      "loss": 0.3278,
+      "step": 8844
+    },
+    {
+      "epoch": 2.2245975855130786,
+      "grad_norm": 0.3014548718929291,
+      "learning_rate": 1.9007252712546786e-06,
+      "loss": 0.3342,
+      "step": 8845
+    },
+    {
+      "epoch": 2.2248490945674044,
+      "grad_norm": 0.3181338310241699,
+      "learning_rate": 1.8995771695616255e-06,
+      "loss": 0.3352,
+      "step": 8846
+    },
+    {
+      "epoch": 2.2251006036217302,
+      "grad_norm": 0.28974583745002747,
+      "learning_rate": 1.8984293334003917e-06,
+      "loss": 0.3346,
+      "step": 8847
+    },
+    {
+      "epoch": 2.2253521126760565,
+      "grad_norm": 0.30787721276283264,
+      "learning_rate": 1.897281762869284e-06,
+      "loss": 0.3173,
+      "step": 8848
+    },
+    {
+      "epoch": 2.2256036217303823,
+      "grad_norm": 0.29022449254989624,
+      "learning_rate": 1.8961344580665808e-06,
+      "loss": 0.3134,
+      "step": 8849
+    },
+    {
+      "epoch": 2.225855130784708,
+      "grad_norm": 0.2811683416366577,
+      "learning_rate": 1.8949874190905453e-06,
+      "loss": 0.2889,
+      "step": 8850
+    },
+    {
+      "epoch": 2.2261066398390343,
+      "grad_norm": 0.2896963059902191,
+      "learning_rate": 1.893840646039411e-06,
+      "loss": 0.3004,
+      "step": 8851
+    },
+    {
+      "epoch": 2.22635814889336,
+      "grad_norm": 0.2802124619483948,
+      "learning_rate": 1.8926941390113946e-06,
+      "loss": 0.3045,
+      "step": 8852
+    },
+    {
+      "epoch": 2.226609657947686,
+      "grad_norm": 0.293868750333786,
+      "learning_rate": 1.8915478981046847e-06,
+      "loss": 0.3134,
+      "step": 8853
+    },
+    {
+      "epoch": 2.226861167002012,
+      "grad_norm": 0.28831586241722107,
+      "learning_rate": 1.8904019234174526e-06,
+      "loss": 0.3105,
+      "step": 8854
+    },
+    {
+      "epoch": 2.227112676056338,
+      "grad_norm": 0.2980342507362366,
+      "learning_rate": 1.88925621504784e-06,
+      "loss": 0.3228,
+      "step": 8855
+    },
+    {
+      "epoch": 2.227364185110664,
+      "grad_norm": 0.30006951093673706,
+      "learning_rate": 1.8881107730939734e-06,
+      "loss": 0.3245,
+      "step": 8856
+    },
+    {
+      "epoch": 2.22761569416499,
+      "grad_norm": 0.2884455919265747,
+      "learning_rate": 1.8869655976539502e-06,
+      "loss": 0.3142,
+      "step": 8857
+    },
+    {
+      "epoch": 2.227867203219316,
+      "grad_norm": 0.32939469814300537,
+      "learning_rate": 1.885820688825848e-06,
+      "loss": 0.3118,
+      "step": 8858
+    },
+    {
+      "epoch": 2.2281187122736417,
+      "grad_norm": 0.30000847578048706,
+      "learning_rate": 1.8846760467077236e-06,
+      "loss": 0.3248,
+      "step": 8859
+    },
+    {
+      "epoch": 2.228370221327968,
+      "grad_norm": 0.29850441217422485,
+      "learning_rate": 1.8835316713976043e-06,
+      "loss": 0.3101,
+      "step": 8860
+    },
+    {
+      "epoch": 2.2286217303822937,
+      "grad_norm": 0.2936405837535858,
+      "learning_rate": 1.882387562993503e-06,
+      "loss": 0.3186,
+      "step": 8861
+    },
+    {
+      "epoch": 2.2288732394366195,
+      "grad_norm": 0.2933601438999176,
+      "learning_rate": 1.881243721593401e-06,
+      "loss": 0.3265,
+      "step": 8862
+    },
+    {
+      "epoch": 2.2291247484909458,
+      "grad_norm": 0.2859182059764862,
+      "learning_rate": 1.8801001472952651e-06,
+      "loss": 0.3317,
+      "step": 8863
+    },
+    {
+      "epoch": 2.2293762575452716,
+      "grad_norm": 0.2935737371444702,
+      "learning_rate": 1.878956840197032e-06,
+      "loss": 0.3221,
+      "step": 8864
+    },
+    {
+      "epoch": 2.2296277665995974,
+      "grad_norm": 0.31956246495246887,
+      "learning_rate": 1.8778138003966218e-06,
+      "loss": 0.3337,
+      "step": 8865
+    },
+    {
+      "epoch": 2.2298792756539236,
+      "grad_norm": 0.3022806644439697,
+      "learning_rate": 1.876671027991926e-06,
+      "loss": 0.3045,
+      "step": 8866
+    },
+    {
+      "epoch": 2.2301307847082494,
+      "grad_norm": 0.29735061526298523,
+      "learning_rate": 1.8755285230808185e-06,
+      "loss": 0.3275,
+      "step": 8867
+    },
+    {
+      "epoch": 2.2303822937625757,
+      "grad_norm": 0.3103269934654236,
+      "learning_rate": 1.874386285761145e-06,
+      "loss": 0.3247,
+      "step": 8868
+    },
+    {
+      "epoch": 2.2306338028169015,
+      "grad_norm": 0.3051782250404358,
+      "learning_rate": 1.8732443161307323e-06,
+      "loss": 0.3151,
+      "step": 8869
+    },
+    {
+      "epoch": 2.2308853118712273,
+      "grad_norm": 0.2836689352989197,
+      "learning_rate": 1.8721026142873843e-06,
+      "loss": 0.2943,
+      "step": 8870
+    },
+    {
+      "epoch": 2.2311368209255535,
+      "grad_norm": 0.31249019503593445,
+      "learning_rate": 1.8709611803288779e-06,
+      "loss": 0.3151,
+      "step": 8871
+    },
+    {
+      "epoch": 2.2313883299798793,
+      "grad_norm": 0.289578378200531,
+      "learning_rate": 1.8698200143529733e-06,
+      "loss": 0.321,
+      "step": 8872
+    },
+    {
+      "epoch": 2.231639839034205,
+      "grad_norm": 0.3105667531490326,
+      "learning_rate": 1.8686791164573997e-06,
+      "loss": 0.3235,
+      "step": 8873
+    },
+    {
+      "epoch": 2.2318913480885314,
+      "grad_norm": 0.31519603729248047,
+      "learning_rate": 1.8675384867398722e-06,
+      "loss": 0.3308,
+      "step": 8874
+    },
+    {
+      "epoch": 2.232142857142857,
+      "grad_norm": 0.30213382840156555,
+      "learning_rate": 1.866398125298075e-06,
+      "loss": 0.2995,
+      "step": 8875
+    },
+    {
+      "epoch": 2.232394366197183,
+      "grad_norm": 0.30496177077293396,
+      "learning_rate": 1.8652580322296766e-06,
+      "loss": 0.3019,
+      "step": 8876
+    },
+    {
+      "epoch": 2.2326458752515093,
+      "grad_norm": 0.30280569195747375,
+      "learning_rate": 1.864118207632315e-06,
+      "loss": 0.355,
+      "step": 8877
+    },
+    {
+      "epoch": 2.232897384305835,
+      "grad_norm": 0.3069732189178467,
+      "learning_rate": 1.8629786516036109e-06,
+      "loss": 0.3714,
+      "step": 8878
+    },
+    {
+      "epoch": 2.233148893360161,
+      "grad_norm": 0.3117799162864685,
+      "learning_rate": 1.861839364241162e-06,
+      "loss": 0.3448,
+      "step": 8879
+    },
+    {
+      "epoch": 2.233400402414487,
+      "grad_norm": 0.29441627860069275,
+      "learning_rate": 1.860700345642537e-06,
+      "loss": 0.324,
+      "step": 8880
+    },
+    {
+      "epoch": 2.233651911468813,
+      "grad_norm": 0.29074686765670776,
+      "learning_rate": 1.8595615959052905e-06,
+      "loss": 0.323,
+      "step": 8881
+    },
+    {
+      "epoch": 2.2339034205231387,
+      "grad_norm": 0.2976393699645996,
+      "learning_rate": 1.8584231151269444e-06,
+      "loss": 0.3263,
+      "step": 8882
+    },
+    {
+      "epoch": 2.234154929577465,
+      "grad_norm": 0.3090425133705139,
+      "learning_rate": 1.8572849034050066e-06,
+      "loss": 0.347,
+      "step": 8883
+    },
+    {
+      "epoch": 2.234406438631791,
+      "grad_norm": 0.2783370316028595,
+      "learning_rate": 1.8561469608369547e-06,
+      "loss": 0.3137,
+      "step": 8884
+    },
+    {
+      "epoch": 2.2346579476861166,
+      "grad_norm": 0.2869153320789337,
+      "learning_rate": 1.8550092875202497e-06,
+      "loss": 0.3286,
+      "step": 8885
+    },
+    {
+      "epoch": 2.234909456740443,
+      "grad_norm": 0.3169291317462921,
+      "learning_rate": 1.8538718835523217e-06,
+      "loss": 0.3386,
+      "step": 8886
+    },
+    {
+      "epoch": 2.2351609657947686,
+      "grad_norm": 0.2993890643119812,
+      "learning_rate": 1.852734749030587e-06,
+      "loss": 0.3145,
+      "step": 8887
+    },
+    {
+      "epoch": 2.2354124748490944,
+      "grad_norm": 0.3018781244754791,
+      "learning_rate": 1.8515978840524302e-06,
+      "loss": 0.3272,
+      "step": 8888
+    },
+    {
+      "epoch": 2.2356639839034207,
+      "grad_norm": 0.29869964718818665,
+      "learning_rate": 1.850461288715218e-06,
+      "loss": 0.3214,
+      "step": 8889
+    },
+    {
+      "epoch": 2.2359154929577465,
+      "grad_norm": 0.2824086844921112,
+      "learning_rate": 1.8493249631162947e-06,
+      "loss": 0.3178,
+      "step": 8890
+    },
+    {
+      "epoch": 2.2361670020120723,
+      "grad_norm": 0.2874877452850342,
+      "learning_rate": 1.8481889073529762e-06,
+      "loss": 0.3227,
+      "step": 8891
+    },
+    {
+      "epoch": 2.2364185110663986,
+      "grad_norm": 0.29150328040122986,
+      "learning_rate": 1.8470531215225617e-06,
+      "loss": 0.3245,
+      "step": 8892
+    },
+    {
+      "epoch": 2.2366700201207244,
+      "grad_norm": 0.28248754143714905,
+      "learning_rate": 1.845917605722321e-06,
+      "loss": 0.325,
+      "step": 8893
+    },
+    {
+      "epoch": 2.23692152917505,
+      "grad_norm": 0.28792303800582886,
+      "learning_rate": 1.8447823600495068e-06,
+      "loss": 0.3175,
+      "step": 8894
+    },
+    {
+      "epoch": 2.2371730382293764,
+      "grad_norm": 0.28080111742019653,
+      "learning_rate": 1.8436473846013432e-06,
+      "loss": 0.3382,
+      "step": 8895
+    },
+    {
+      "epoch": 2.237424547283702,
+      "grad_norm": 0.28693875670433044,
+      "learning_rate": 1.842512679475037e-06,
+      "loss": 0.3263,
+      "step": 8896
+    },
+    {
+      "epoch": 2.237676056338028,
+      "grad_norm": 0.2953447103500366,
+      "learning_rate": 1.8413782447677641e-06,
+      "loss": 0.3322,
+      "step": 8897
+    },
+    {
+      "epoch": 2.2379275653923543,
+      "grad_norm": 0.3262675404548645,
+      "learning_rate": 1.8402440805766863e-06,
+      "loss": 0.317,
+      "step": 8898
+    },
+    {
+      "epoch": 2.23817907444668,
+      "grad_norm": 0.28162655234336853,
+      "learning_rate": 1.8391101869989341e-06,
+      "loss": 0.3345,
+      "step": 8899
+    },
+    {
+      "epoch": 2.238430583501006,
+      "grad_norm": 0.27264997363090515,
+      "learning_rate": 1.8379765641316216e-06,
+      "loss": 0.2932,
+      "step": 8900
+    },
+    {
+      "epoch": 2.238682092555332,
+      "grad_norm": 0.30955466628074646,
+      "learning_rate": 1.8368432120718332e-06,
+      "loss": 0.3388,
+      "step": 8901
+    },
+    {
+      "epoch": 2.238933601609658,
+      "grad_norm": 0.29133230447769165,
+      "learning_rate": 1.8357101309166364e-06,
+      "loss": 0.3317,
+      "step": 8902
+    },
+    {
+      "epoch": 2.2391851106639837,
+      "grad_norm": 0.3130471110343933,
+      "learning_rate": 1.8345773207630696e-06,
+      "loss": 0.31,
+      "step": 8903
+    },
+    {
+      "epoch": 2.23943661971831,
+      "grad_norm": 0.30992329120635986,
+      "learning_rate": 1.833444781708154e-06,
+      "loss": 0.3204,
+      "step": 8904
+    },
+    {
+      "epoch": 2.239688128772636,
+      "grad_norm": 0.2960168719291687,
+      "learning_rate": 1.8323125138488818e-06,
+      "loss": 0.3244,
+      "step": 8905
+    },
+    {
+      "epoch": 2.2399396378269616,
+      "grad_norm": 0.31481972336769104,
+      "learning_rate": 1.8311805172822272e-06,
+      "loss": 0.3318,
+      "step": 8906
+    },
+    {
+      "epoch": 2.240191146881288,
+      "grad_norm": 0.3201163113117218,
+      "learning_rate": 1.8300487921051352e-06,
+      "loss": 0.3213,
+      "step": 8907
+    },
+    {
+      "epoch": 2.2404426559356136,
+      "grad_norm": 0.31222665309906006,
+      "learning_rate": 1.8289173384145354e-06,
+      "loss": 0.2939,
+      "step": 8908
+    },
+    {
+      "epoch": 2.2406941649899395,
+      "grad_norm": 0.3136330246925354,
+      "learning_rate": 1.8277861563073252e-06,
+      "loss": 0.3286,
+      "step": 8909
+    },
+    {
+      "epoch": 2.2409456740442657,
+      "grad_norm": 0.3031679391860962,
+      "learning_rate": 1.8266552458803872e-06,
+      "loss": 0.3048,
+      "step": 8910
+    },
+    {
+      "epoch": 2.2411971830985915,
+      "grad_norm": 0.29609373211860657,
+      "learning_rate": 1.8255246072305727e-06,
+      "loss": 0.2845,
+      "step": 8911
+    },
+    {
+      "epoch": 2.2414486921529173,
+      "grad_norm": 0.32585108280181885,
+      "learning_rate": 1.8243942404547183e-06,
+      "loss": 0.3267,
+      "step": 8912
+    },
+    {
+      "epoch": 2.2417002012072436,
+      "grad_norm": 0.30860912799835205,
+      "learning_rate": 1.8232641456496292e-06,
+      "loss": 0.3092,
+      "step": 8913
+    },
+    {
+      "epoch": 2.2419517102615694,
+      "grad_norm": 0.30172842741012573,
+      "learning_rate": 1.822134322912094e-06,
+      "loss": 0.2914,
+      "step": 8914
+    },
+    {
+      "epoch": 2.242203219315895,
+      "grad_norm": 0.30393847823143005,
+      "learning_rate": 1.8210047723388718e-06,
+      "loss": 0.3236,
+      "step": 8915
+    },
+    {
+      "epoch": 2.2424547283702214,
+      "grad_norm": 0.2871536314487457,
+      "learning_rate": 1.8198754940267044e-06,
+      "loss": 0.3247,
+      "step": 8916
+    },
+    {
+      "epoch": 2.2427062374245472,
+      "grad_norm": 0.31319135427474976,
+      "learning_rate": 1.818746488072305e-06,
+      "loss": 0.3169,
+      "step": 8917
+    },
+    {
+      "epoch": 2.242957746478873,
+      "grad_norm": 0.3419165015220642,
+      "learning_rate": 1.8176177545723683e-06,
+      "loss": 0.3388,
+      "step": 8918
+    },
+    {
+      "epoch": 2.2432092555331993,
+      "grad_norm": 0.308427631855011,
+      "learning_rate": 1.8164892936235602e-06,
+      "loss": 0.3216,
+      "step": 8919
+    },
+    {
+      "epoch": 2.243460764587525,
+      "grad_norm": 0.33541321754455566,
+      "learning_rate": 1.81536110532253e-06,
+      "loss": 0.3406,
+      "step": 8920
+    },
+    {
+      "epoch": 2.243712273641851,
+      "grad_norm": 0.3217279314994812,
+      "learning_rate": 1.8142331897658967e-06,
+      "loss": 0.3293,
+      "step": 8921
+    },
+    {
+      "epoch": 2.243963782696177,
+      "grad_norm": 0.30454570055007935,
+      "learning_rate": 1.8131055470502601e-06,
+      "loss": 0.3415,
+      "step": 8922
+    },
+    {
+      "epoch": 2.244215291750503,
+      "grad_norm": 0.30971288681030273,
+      "learning_rate": 1.8119781772721984e-06,
+      "loss": 0.3416,
+      "step": 8923
+    },
+    {
+      "epoch": 2.2444668008048287,
+      "grad_norm": 0.3124922513961792,
+      "learning_rate": 1.8108510805282598e-06,
+      "loss": 0.3392,
+      "step": 8924
+    },
+    {
+      "epoch": 2.244718309859155,
+      "grad_norm": 0.29240474104881287,
+      "learning_rate": 1.8097242569149765e-06,
+      "loss": 0.3084,
+      "step": 8925
+    },
+    {
+      "epoch": 2.244969818913481,
+      "grad_norm": 0.2906413674354553,
+      "learning_rate": 1.8085977065288502e-06,
+      "loss": 0.3244,
+      "step": 8926
+    },
+    {
+      "epoch": 2.245221327967807,
+      "grad_norm": 0.27822086215019226,
+      "learning_rate": 1.807471429466367e-06,
+      "loss": 0.3249,
+      "step": 8927
+    },
+    {
+      "epoch": 2.245472837022133,
+      "grad_norm": 0.3114204704761505,
+      "learning_rate": 1.8063454258239821e-06,
+      "loss": 0.3361,
+      "step": 8928
+    },
+    {
+      "epoch": 2.2457243460764587,
+      "grad_norm": 0.3120749890804291,
+      "learning_rate": 1.8052196956981333e-06,
+      "loss": 0.3021,
+      "step": 8929
+    },
+    {
+      "epoch": 2.245975855130785,
+      "grad_norm": 0.31990909576416016,
+      "learning_rate": 1.8040942391852296e-06,
+      "loss": 0.3188,
+      "step": 8930
+    },
+    {
+      "epoch": 2.2462273641851107,
+      "grad_norm": 0.3102615475654602,
+      "learning_rate": 1.8029690563816626e-06,
+      "loss": 0.3298,
+      "step": 8931
+    },
+    {
+      "epoch": 2.2464788732394365,
+      "grad_norm": 0.29477155208587646,
+      "learning_rate": 1.8018441473837934e-06,
+      "loss": 0.3149,
+      "step": 8932
+    },
+    {
+      "epoch": 2.2467303822937628,
+      "grad_norm": 0.2773665189743042,
+      "learning_rate": 1.8007195122879656e-06,
+      "loss": 0.3093,
+      "step": 8933
+    },
+    {
+      "epoch": 2.2469818913480886,
+      "grad_norm": 0.31135937571525574,
+      "learning_rate": 1.7995951511904985e-06,
+      "loss": 0.334,
+      "step": 8934
+    },
+    {
+      "epoch": 2.2472334004024144,
+      "grad_norm": 0.28784048557281494,
+      "learning_rate": 1.7984710641876829e-06,
+      "loss": 0.3057,
+      "step": 8935
+    },
+    {
+      "epoch": 2.2474849094567406,
+      "grad_norm": 0.3066675364971161,
+      "learning_rate": 1.7973472513757945e-06,
+      "loss": 0.3211,
+      "step": 8936
+    },
+    {
+      "epoch": 2.2477364185110664,
+      "grad_norm": 0.30136793851852417,
+      "learning_rate": 1.7962237128510761e-06,
+      "loss": 0.3269,
+      "step": 8937
+    },
+    {
+      "epoch": 2.2479879275653922,
+      "grad_norm": 0.30192577838897705,
+      "learning_rate": 1.7951004487097557e-06,
+      "loss": 0.3124,
+      "step": 8938
+    },
+    {
+      "epoch": 2.2482394366197185,
+      "grad_norm": 0.29201701283454895,
+      "learning_rate": 1.7939774590480301e-06,
+      "loss": 0.3281,
+      "step": 8939
+    },
+    {
+      "epoch": 2.2484909456740443,
+      "grad_norm": 0.29174402356147766,
+      "learning_rate": 1.7928547439620808e-06,
+      "loss": 0.3287,
+      "step": 8940
+    },
+    {
+      "epoch": 2.24874245472837,
+      "grad_norm": 0.3136996924877167,
+      "learning_rate": 1.7917323035480567e-06,
+      "loss": 0.3081,
+      "step": 8941
+    },
+    {
+      "epoch": 2.2489939637826963,
+      "grad_norm": 0.2897532880306244,
+      "learning_rate": 1.7906101379020912e-06,
+      "loss": 0.3015,
+      "step": 8942
+    },
+    {
+      "epoch": 2.249245472837022,
+      "grad_norm": 0.30037471652030945,
+      "learning_rate": 1.7894882471202884e-06,
+      "loss": 0.344,
+      "step": 8943
+    },
+    {
+      "epoch": 2.249496981891348,
+      "grad_norm": 0.32366058230400085,
+      "learning_rate": 1.7883666312987319e-06,
+      "loss": 0.328,
+      "step": 8944
+    },
+    {
+      "epoch": 2.249748490945674,
+      "grad_norm": 0.295149564743042,
+      "learning_rate": 1.7872452905334836e-06,
+      "loss": 0.3338,
+      "step": 8945
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.31807345151901245,
+      "learning_rate": 1.7861242249205752e-06,
+      "loss": 0.3177,
+      "step": 8946
+    },
+    {
+      "epoch": 2.250251509054326,
+      "grad_norm": 0.31685301661491394,
+      "learning_rate": 1.785003434556023e-06,
+      "loss": 0.3012,
+      "step": 8947
+    },
+    {
+      "epoch": 2.250503018108652,
+      "grad_norm": 0.3292388617992401,
+      "learning_rate": 1.783882919535812e-06,
+      "loss": 0.3129,
+      "step": 8948
+    },
+    {
+      "epoch": 2.250754527162978,
+      "grad_norm": 0.3227696716785431,
+      "learning_rate": 1.7827626799559105e-06,
+      "loss": 0.3408,
+      "step": 8949
+    },
+    {
+      "epoch": 2.2510060362173037,
+      "grad_norm": 0.2962954044342041,
+      "learning_rate": 1.7816427159122569e-06,
+      "loss": 0.339,
+      "step": 8950
+    },
+    {
+      "epoch": 2.25125754527163,
+      "grad_norm": 0.3421285152435303,
+      "learning_rate": 1.7805230275007724e-06,
+      "loss": 0.3196,
+      "step": 8951
+    },
+    {
+      "epoch": 2.2515090543259557,
+      "grad_norm": 0.3049623668193817,
+      "learning_rate": 1.7794036148173477e-06,
+      "loss": 0.344,
+      "step": 8952
+    },
+    {
+      "epoch": 2.2517605633802815,
+      "grad_norm": 0.29739803075790405,
+      "learning_rate": 1.7782844779578574e-06,
+      "loss": 0.3286,
+      "step": 8953
+    },
+    {
+      "epoch": 2.2520120724346078,
+      "grad_norm": 0.31580233573913574,
+      "learning_rate": 1.7771656170181445e-06,
+      "loss": 0.3177,
+      "step": 8954
+    },
+    {
+      "epoch": 2.2522635814889336,
+      "grad_norm": 0.321826308965683,
+      "learning_rate": 1.7760470320940348e-06,
+      "loss": 0.3254,
+      "step": 8955
+    },
+    {
+      "epoch": 2.2525150905432594,
+      "grad_norm": 0.2998609244823456,
+      "learning_rate": 1.7749287232813296e-06,
+      "loss": 0.3193,
+      "step": 8956
+    },
+    {
+      "epoch": 2.2527665995975856,
+      "grad_norm": 0.29203975200653076,
+      "learning_rate": 1.7738106906758013e-06,
+      "loss": 0.3187,
+      "step": 8957
+    },
+    {
+      "epoch": 2.2530181086519114,
+      "grad_norm": 0.30509132146835327,
+      "learning_rate": 1.7726929343732059e-06,
+      "loss": 0.3282,
+      "step": 8958
+    },
+    {
+      "epoch": 2.2532696177062372,
+      "grad_norm": 0.29113608598709106,
+      "learning_rate": 1.7715754544692692e-06,
+      "loss": 0.3167,
+      "step": 8959
+    },
+    {
+      "epoch": 2.2535211267605635,
+      "grad_norm": 0.322614848613739,
+      "learning_rate": 1.7704582510596996e-06,
+      "loss": 0.3343,
+      "step": 8960
+    },
+    {
+      "epoch": 2.2537726358148893,
+      "grad_norm": 0.31486186385154724,
+      "learning_rate": 1.7693413242401753e-06,
+      "loss": 0.3303,
+      "step": 8961
+    },
+    {
+      "epoch": 2.2540241448692155,
+      "grad_norm": 0.30907559394836426,
+      "learning_rate": 1.7682246741063568e-06,
+      "loss": 0.3308,
+      "step": 8962
+    },
+    {
+      "epoch": 2.2542756539235413,
+      "grad_norm": 0.3086276054382324,
+      "learning_rate": 1.7671083007538765e-06,
+      "loss": 0.3373,
+      "step": 8963
+    },
+    {
+      "epoch": 2.254527162977867,
+      "grad_norm": 0.31900346279144287,
+      "learning_rate": 1.7659922042783463e-06,
+      "loss": 0.3326,
+      "step": 8964
+    },
+    {
+      "epoch": 2.2547786720321934,
+      "grad_norm": 0.29411938786506653,
+      "learning_rate": 1.7648763847753497e-06,
+      "loss": 0.3219,
+      "step": 8965
+    },
+    {
+      "epoch": 2.255030181086519,
+      "grad_norm": 0.2990444004535675,
+      "learning_rate": 1.7637608423404524e-06,
+      "loss": 0.3265,
+      "step": 8966
+    },
+    {
+      "epoch": 2.255281690140845,
+      "grad_norm": 0.3038404881954193,
+      "learning_rate": 1.7626455770691947e-06,
+      "loss": 0.2958,
+      "step": 8967
+    },
+    {
+      "epoch": 2.2555331991951713,
+      "grad_norm": 0.2940180003643036,
+      "learning_rate": 1.7615305890570888e-06,
+      "loss": 0.3148,
+      "step": 8968
+    },
+    {
+      "epoch": 2.255784708249497,
+      "grad_norm": 0.31063640117645264,
+      "learning_rate": 1.76041587839963e-06,
+      "loss": 0.3319,
+      "step": 8969
+    },
+    {
+      "epoch": 2.256036217303823,
+      "grad_norm": 0.2945641875267029,
+      "learning_rate": 1.759301445192283e-06,
+      "loss": 0.3332,
+      "step": 8970
+    },
+    {
+      "epoch": 2.256287726358149,
+      "grad_norm": 0.2893473505973816,
+      "learning_rate": 1.7581872895304947e-06,
+      "loss": 0.3137,
+      "step": 8971
+    },
+    {
+      "epoch": 2.256539235412475,
+      "grad_norm": 0.2820425033569336,
+      "learning_rate": 1.7570734115096827e-06,
+      "loss": 0.3094,
+      "step": 8972
+    },
+    {
+      "epoch": 2.2567907444668007,
+      "grad_norm": 0.27644819021224976,
+      "learning_rate": 1.7559598112252475e-06,
+      "loss": 0.326,
+      "step": 8973
+    },
+    {
+      "epoch": 2.257042253521127,
+      "grad_norm": 0.276979923248291,
+      "learning_rate": 1.7548464887725576e-06,
+      "loss": 0.3127,
+      "step": 8974
+    },
+    {
+      "epoch": 2.2572937625754528,
+      "grad_norm": 0.3153163492679596,
+      "learning_rate": 1.753733444246966e-06,
+      "loss": 0.3206,
+      "step": 8975
+    },
+    {
+      "epoch": 2.2575452716297786,
+      "grad_norm": 0.286739706993103,
+      "learning_rate": 1.7526206777437948e-06,
+      "loss": 0.3228,
+      "step": 8976
+    },
+    {
+      "epoch": 2.257796780684105,
+      "grad_norm": 0.3067798912525177,
+      "learning_rate": 1.7515081893583469e-06,
+      "loss": 0.3415,
+      "step": 8977
+    },
+    {
+      "epoch": 2.2580482897384306,
+      "grad_norm": 0.28727710247039795,
+      "learning_rate": 1.7503959791859016e-06,
+      "loss": 0.3038,
+      "step": 8978
+    },
+    {
+      "epoch": 2.2582997987927564,
+      "grad_norm": 0.29552116990089417,
+      "learning_rate": 1.7492840473217099e-06,
+      "loss": 0.3148,
+      "step": 8979
+    },
+    {
+      "epoch": 2.2585513078470827,
+      "grad_norm": 0.2867220342159271,
+      "learning_rate": 1.7481723938610045e-06,
+      "loss": 0.3154,
+      "step": 8980
+    },
+    {
+      "epoch": 2.2588028169014085,
+      "grad_norm": 0.2924220860004425,
+      "learning_rate": 1.747061018898989e-06,
+      "loss": 0.288,
+      "step": 8981
+    },
+    {
+      "epoch": 2.2590543259557343,
+      "grad_norm": 0.29958319664001465,
+      "learning_rate": 1.745949922530848e-06,
+      "loss": 0.3132,
+      "step": 8982
+    },
+    {
+      "epoch": 2.2593058350100605,
+      "grad_norm": 0.3001472055912018,
+      "learning_rate": 1.7448391048517378e-06,
+      "loss": 0.3353,
+      "step": 8983
+    },
+    {
+      "epoch": 2.2595573440643864,
+      "grad_norm": 0.32707712054252625,
+      "learning_rate": 1.7437285659567954e-06,
+      "loss": 0.3129,
+      "step": 8984
+    },
+    {
+      "epoch": 2.259808853118712,
+      "grad_norm": 0.2877408266067505,
+      "learning_rate": 1.7426183059411284e-06,
+      "loss": 0.334,
+      "step": 8985
+    },
+    {
+      "epoch": 2.2600603621730384,
+      "grad_norm": 0.3208245635032654,
+      "learning_rate": 1.7415083248998271e-06,
+      "loss": 0.3472,
+      "step": 8986
+    },
+    {
+      "epoch": 2.260311871227364,
+      "grad_norm": 0.31004029512405396,
+      "learning_rate": 1.7403986229279506e-06,
+      "loss": 0.3382,
+      "step": 8987
+    },
+    {
+      "epoch": 2.26056338028169,
+      "grad_norm": 0.3009289503097534,
+      "learning_rate": 1.7392892001205409e-06,
+      "loss": 0.3078,
+      "step": 8988
+    },
+    {
+      "epoch": 2.2608148893360163,
+      "grad_norm": 0.29705020785331726,
+      "learning_rate": 1.7381800565726138e-06,
+      "loss": 0.3133,
+      "step": 8989
+    },
+    {
+      "epoch": 2.261066398390342,
+      "grad_norm": 0.30671584606170654,
+      "learning_rate": 1.7370711923791567e-06,
+      "loss": 0.3117,
+      "step": 8990
+    },
+    {
+      "epoch": 2.261317907444668,
+      "grad_norm": 0.2912043333053589,
+      "learning_rate": 1.735962607635141e-06,
+      "loss": 0.3099,
+      "step": 8991
+    },
+    {
+      "epoch": 2.261569416498994,
+      "grad_norm": 0.3244048058986664,
+      "learning_rate": 1.7348543024355068e-06,
+      "loss": 0.329,
+      "step": 8992
+    },
+    {
+      "epoch": 2.26182092555332,
+      "grad_norm": 0.28510650992393494,
+      "learning_rate": 1.7337462768751766e-06,
+      "loss": 0.3192,
+      "step": 8993
+    },
+    {
+      "epoch": 2.2620724346076457,
+      "grad_norm": 0.3037286698818207,
+      "learning_rate": 1.7326385310490424e-06,
+      "loss": 0.3109,
+      "step": 8994
+    },
+    {
+      "epoch": 2.262323943661972,
+      "grad_norm": 0.3154160678386688,
+      "learning_rate": 1.731531065051979e-06,
+      "loss": 0.3267,
+      "step": 8995
+    },
+    {
+      "epoch": 2.262575452716298,
+      "grad_norm": 0.2968575358390808,
+      "learning_rate": 1.7304238789788308e-06,
+      "loss": 0.3137,
+      "step": 8996
+    },
+    {
+      "epoch": 2.2628269617706236,
+      "grad_norm": 0.2917513847351074,
+      "learning_rate": 1.7293169729244247e-06,
+      "loss": 0.2899,
+      "step": 8997
+    },
+    {
+      "epoch": 2.26307847082495,
+      "grad_norm": 0.29885080456733704,
+      "learning_rate": 1.728210346983557e-06,
+      "loss": 0.3232,
+      "step": 8998
+    },
+    {
+      "epoch": 2.2633299798792756,
+      "grad_norm": 0.2979438006877899,
+      "learning_rate": 1.7271040012510044e-06,
+      "loss": 0.3017,
+      "step": 8999
+    },
+    {
+      "epoch": 2.2635814889336014,
+      "grad_norm": 0.2976253628730774,
+      "learning_rate": 1.7259979358215213e-06,
+      "loss": 0.3289,
+      "step": 9000
+    },
+    {
+      "epoch": 2.2638329979879277,
+      "grad_norm": 0.2976281940937042,
+      "learning_rate": 1.7248921507898304e-06,
+      "loss": 0.3013,
+      "step": 9001
+    },
+    {
+      "epoch": 2.2640845070422535,
+      "grad_norm": 0.31368160247802734,
+      "learning_rate": 1.7237866462506398e-06,
+      "loss": 0.2938,
+      "step": 9002
+    },
+    {
+      "epoch": 2.2643360160965793,
+      "grad_norm": 0.30961135029792786,
+      "learning_rate": 1.7226814222986255e-06,
+      "loss": 0.3168,
+      "step": 9003
+    },
+    {
+      "epoch": 2.2645875251509056,
+      "grad_norm": 0.2974013686180115,
+      "learning_rate": 1.7215764790284462e-06,
+      "loss": 0.3062,
+      "step": 9004
+    },
+    {
+      "epoch": 2.2648390342052314,
+      "grad_norm": 0.2913673520088196,
+      "learning_rate": 1.7204718165347301e-06,
+      "loss": 0.3068,
+      "step": 9005
+    },
+    {
+      "epoch": 2.265090543259557,
+      "grad_norm": 0.3057686686515808,
+      "learning_rate": 1.7193674349120877e-06,
+      "loss": 0.3078,
+      "step": 9006
+    },
+    {
+      "epoch": 2.2653420523138834,
+      "grad_norm": 0.29526326060295105,
+      "learning_rate": 1.7182633342550991e-06,
+      "loss": 0.2993,
+      "step": 9007
+    },
+    {
+      "epoch": 2.265593561368209,
+      "grad_norm": 0.30530017614364624,
+      "learning_rate": 1.7171595146583275e-06,
+      "loss": 0.3163,
+      "step": 9008
+    },
+    {
+      "epoch": 2.265845070422535,
+      "grad_norm": 0.31532004475593567,
+      "learning_rate": 1.7160559762163042e-06,
+      "loss": 0.3276,
+      "step": 9009
+    },
+    {
+      "epoch": 2.2660965794768613,
+      "grad_norm": 0.30409690737724304,
+      "learning_rate": 1.714952719023542e-06,
+      "loss": 0.3166,
+      "step": 9010
+    },
+    {
+      "epoch": 2.266348088531187,
+      "grad_norm": 0.3003282845020294,
+      "learning_rate": 1.7138497431745304e-06,
+      "loss": 0.3452,
+      "step": 9011
+    },
+    {
+      "epoch": 2.266599597585513,
+      "grad_norm": 0.29305264353752136,
+      "learning_rate": 1.7127470487637282e-06,
+      "loss": 0.3278,
+      "step": 9012
+    },
+    {
+      "epoch": 2.266851106639839,
+      "grad_norm": 0.27770164608955383,
+      "learning_rate": 1.7116446358855781e-06,
+      "loss": 0.3196,
+      "step": 9013
+    },
+    {
+      "epoch": 2.267102615694165,
+      "grad_norm": 0.2922765016555786,
+      "learning_rate": 1.7105425046344914e-06,
+      "loss": 0.3444,
+      "step": 9014
+    },
+    {
+      "epoch": 2.2673541247484907,
+      "grad_norm": 0.32455912232398987,
+      "learning_rate": 1.709440655104862e-06,
+      "loss": 0.2895,
+      "step": 9015
+    },
+    {
+      "epoch": 2.267605633802817,
+      "grad_norm": 0.2932599186897278,
+      "learning_rate": 1.7083390873910533e-06,
+      "loss": 0.3285,
+      "step": 9016
+    },
+    {
+      "epoch": 2.267857142857143,
+      "grad_norm": 0.29908427596092224,
+      "learning_rate": 1.7072378015874107e-06,
+      "loss": 0.3236,
+      "step": 9017
+    },
+    {
+      "epoch": 2.2681086519114686,
+      "grad_norm": 0.31389617919921875,
+      "learning_rate": 1.7061367977882493e-06,
+      "loss": 0.3431,
+      "step": 9018
+    },
+    {
+      "epoch": 2.268360160965795,
+      "grad_norm": 0.2915651798248291,
+      "learning_rate": 1.7050360760878665e-06,
+      "loss": 0.3442,
+      "step": 9019
+    },
+    {
+      "epoch": 2.2686116700201207,
+      "grad_norm": 0.3061988949775696,
+      "learning_rate": 1.7039356365805286e-06,
+      "loss": 0.3118,
+      "step": 9020
+    },
+    {
+      "epoch": 2.2688631790744465,
+      "grad_norm": 0.3040686249732971,
+      "learning_rate": 1.7028354793604857e-06,
+      "loss": 0.3257,
+      "step": 9021
+    },
+    {
+      "epoch": 2.2691146881287727,
+      "grad_norm": 0.3051724135875702,
+      "learning_rate": 1.7017356045219545e-06,
+      "loss": 0.3085,
+      "step": 9022
+    },
+    {
+      "epoch": 2.2693661971830985,
+      "grad_norm": 0.32119080424308777,
+      "learning_rate": 1.700636012159137e-06,
+      "loss": 0.3363,
+      "step": 9023
+    },
+    {
+      "epoch": 2.2696177062374243,
+      "grad_norm": 0.28925296664237976,
+      "learning_rate": 1.699536702366203e-06,
+      "loss": 0.3113,
+      "step": 9024
+    },
+    {
+      "epoch": 2.2698692152917506,
+      "grad_norm": 0.30859729647636414,
+      "learning_rate": 1.6984376752373038e-06,
+      "loss": 0.2988,
+      "step": 9025
+    },
+    {
+      "epoch": 2.2701207243460764,
+      "grad_norm": 0.3168920576572418,
+      "learning_rate": 1.6973389308665617e-06,
+      "loss": 0.3182,
+      "step": 9026
+    },
+    {
+      "epoch": 2.270372233400402,
+      "grad_norm": 0.3211313784122467,
+      "learning_rate": 1.6962404693480805e-06,
+      "loss": 0.3277,
+      "step": 9027
+    },
+    {
+      "epoch": 2.2706237424547284,
+      "grad_norm": 0.30997171998023987,
+      "learning_rate": 1.6951422907759336e-06,
+      "loss": 0.3244,
+      "step": 9028
+    },
+    {
+      "epoch": 2.2708752515090542,
+      "grad_norm": 0.3075638711452484,
+      "learning_rate": 1.6940443952441755e-06,
+      "loss": 0.3263,
+      "step": 9029
+    },
+    {
+      "epoch": 2.2711267605633805,
+      "grad_norm": 0.31429892778396606,
+      "learning_rate": 1.6929467828468316e-06,
+      "loss": 0.3171,
+      "step": 9030
+    },
+    {
+      "epoch": 2.2713782696177063,
+      "grad_norm": 0.2928819954395294,
+      "learning_rate": 1.6918494536779084e-06,
+      "loss": 0.2967,
+      "step": 9031
+    },
+    {
+      "epoch": 2.271629778672032,
+      "grad_norm": 0.2912546694278717,
+      "learning_rate": 1.690752407831382e-06,
+      "loss": 0.3185,
+      "step": 9032
+    },
+    {
+      "epoch": 2.2718812877263583,
+      "grad_norm": 0.31666284799575806,
+      "learning_rate": 1.689655645401211e-06,
+      "loss": 0.351,
+      "step": 9033
+    },
+    {
+      "epoch": 2.272132796780684,
+      "grad_norm": 0.30313748121261597,
+      "learning_rate": 1.6885591664813228e-06,
+      "loss": 0.3183,
+      "step": 9034
+    },
+    {
+      "epoch": 2.27238430583501,
+      "grad_norm": 0.2988080084323883,
+      "learning_rate": 1.6874629711656275e-06,
+      "loss": 0.3295,
+      "step": 9035
+    },
+    {
+      "epoch": 2.272635814889336,
+      "grad_norm": 0.2955930531024933,
+      "learning_rate": 1.6863670595480042e-06,
+      "loss": 0.3343,
+      "step": 9036
+    },
+    {
+      "epoch": 2.272887323943662,
+      "grad_norm": 0.33139654994010925,
+      "learning_rate": 1.6852714317223118e-06,
+      "loss": 0.3049,
+      "step": 9037
+    },
+    {
+      "epoch": 2.273138832997988,
+      "grad_norm": 0.29346272349357605,
+      "learning_rate": 1.684176087782386e-06,
+      "loss": 0.3158,
+      "step": 9038
+    },
+    {
+      "epoch": 2.273390342052314,
+      "grad_norm": 0.3030097186565399,
+      "learning_rate": 1.6830810278220327e-06,
+      "loss": 0.3183,
+      "step": 9039
+    },
+    {
+      "epoch": 2.27364185110664,
+      "grad_norm": 0.28629571199417114,
+      "learning_rate": 1.681986251935041e-06,
+      "loss": 0.3128,
+      "step": 9040
+    },
+    {
+      "epoch": 2.2738933601609657,
+      "grad_norm": 0.31391236186027527,
+      "learning_rate": 1.6808917602151676e-06,
+      "loss": 0.3154,
+      "step": 9041
+    },
+    {
+      "epoch": 2.274144869215292,
+      "grad_norm": 0.27841538190841675,
+      "learning_rate": 1.6797975527561522e-06,
+      "loss": 0.3176,
+      "step": 9042
+    },
+    {
+      "epoch": 2.2743963782696177,
+      "grad_norm": 0.2939876914024353,
+      "learning_rate": 1.6787036296517034e-06,
+      "loss": 0.3298,
+      "step": 9043
+    },
+    {
+      "epoch": 2.2746478873239435,
+      "grad_norm": 0.3155568540096283,
+      "learning_rate": 1.677609990995513e-06,
+      "loss": 0.3457,
+      "step": 9044
+    },
+    {
+      "epoch": 2.2748993963782698,
+      "grad_norm": 0.30136945843696594,
+      "learning_rate": 1.6765166368812403e-06,
+      "loss": 0.3086,
+      "step": 9045
+    },
+    {
+      "epoch": 2.2751509054325956,
+      "grad_norm": 0.2959335744380951,
+      "learning_rate": 1.6754235674025271e-06,
+      "loss": 0.3092,
+      "step": 9046
+    },
+    {
+      "epoch": 2.2754024144869214,
+      "grad_norm": 0.2871185541152954,
+      "learning_rate": 1.6743307826529858e-06,
+      "loss": 0.3144,
+      "step": 9047
+    },
+    {
+      "epoch": 2.2756539235412476,
+      "grad_norm": 0.30033811926841736,
+      "learning_rate": 1.6732382827262089e-06,
+      "loss": 0.31,
+      "step": 9048
+    },
+    {
+      "epoch": 2.2759054325955734,
+      "grad_norm": 0.3050321936607361,
+      "learning_rate": 1.6721460677157591e-06,
+      "loss": 0.3331,
+      "step": 9049
+    },
+    {
+      "epoch": 2.2761569416498992,
+      "grad_norm": 0.33643415570259094,
+      "learning_rate": 1.6710541377151818e-06,
+      "loss": 0.3273,
+      "step": 9050
+    },
+    {
+      "epoch": 2.2764084507042255,
+      "grad_norm": 0.32750841975212097,
+      "learning_rate": 1.6699624928179897e-06,
+      "loss": 0.3269,
+      "step": 9051
+    },
+    {
+      "epoch": 2.2766599597585513,
+      "grad_norm": 0.29840290546417236,
+      "learning_rate": 1.6688711331176777e-06,
+      "loss": 0.313,
+      "step": 9052
+    },
+    {
+      "epoch": 2.276911468812877,
+      "grad_norm": 0.3018534481525421,
+      "learning_rate": 1.6677800587077153e-06,
+      "loss": 0.3041,
+      "step": 9053
+    },
+    {
+      "epoch": 2.2771629778672033,
+      "grad_norm": 0.297966867685318,
+      "learning_rate": 1.6666892696815428e-06,
+      "loss": 0.3085,
+      "step": 9054
+    },
+    {
+      "epoch": 2.277414486921529,
+      "grad_norm": 0.29202499985694885,
+      "learning_rate": 1.6655987661325835e-06,
+      "loss": 0.313,
+      "step": 9055
+    },
+    {
+      "epoch": 2.277665995975855,
+      "grad_norm": 0.27722373604774475,
+      "learning_rate": 1.6645085481542273e-06,
+      "loss": 0.3335,
+      "step": 9056
+    },
+    {
+      "epoch": 2.277917505030181,
+      "grad_norm": 0.3057959973812103,
+      "learning_rate": 1.6634186158398496e-06,
+      "loss": 0.3158,
+      "step": 9057
+    },
+    {
+      "epoch": 2.278169014084507,
+      "grad_norm": 0.29990729689598083,
+      "learning_rate": 1.6623289692827916e-06,
+      "loss": 0.3162,
+      "step": 9058
+    },
+    {
+      "epoch": 2.278420523138833,
+      "grad_norm": 0.3051862418651581,
+      "learning_rate": 1.6612396085763794e-06,
+      "loss": 0.2969,
+      "step": 9059
+    },
+    {
+      "epoch": 2.278672032193159,
+      "grad_norm": 0.30120375752449036,
+      "learning_rate": 1.660150533813905e-06,
+      "loss": 0.3275,
+      "step": 9060
+    },
+    {
+      "epoch": 2.278923541247485,
+      "grad_norm": 0.3087018132209778,
+      "learning_rate": 1.6590617450886453e-06,
+      "loss": 0.3146,
+      "step": 9061
+    },
+    {
+      "epoch": 2.279175050301811,
+      "grad_norm": 0.3511127531528473,
+      "learning_rate": 1.657973242493845e-06,
+      "loss": 0.3205,
+      "step": 9062
+    },
+    {
+      "epoch": 2.279426559356137,
+      "grad_norm": 0.29214683175086975,
+      "learning_rate": 1.6568850261227282e-06,
+      "loss": 0.3271,
+      "step": 9063
+    },
+    {
+      "epoch": 2.2796780684104627,
+      "grad_norm": 0.3279742896556854,
+      "learning_rate": 1.6557970960684965e-06,
+      "loss": 0.3187,
+      "step": 9064
+    },
+    {
+      "epoch": 2.279929577464789,
+      "grad_norm": 0.2921777367591858,
+      "learning_rate": 1.6547094524243207e-06,
+      "loss": 0.3241,
+      "step": 9065
+    },
+    {
+      "epoch": 2.2801810865191148,
+      "grad_norm": 0.2994241714477539,
+      "learning_rate": 1.6536220952833536e-06,
+      "loss": 0.3162,
+      "step": 9066
+    },
+    {
+      "epoch": 2.2804325955734406,
+      "grad_norm": 0.29721885919570923,
+      "learning_rate": 1.6525350247387178e-06,
+      "loss": 0.3329,
+      "step": 9067
+    },
+    {
+      "epoch": 2.280684104627767,
+      "grad_norm": 0.3295932412147522,
+      "learning_rate": 1.651448240883517e-06,
+      "loss": 0.312,
+      "step": 9068
+    },
+    {
+      "epoch": 2.2809356136820926,
+      "grad_norm": 0.3002535402774811,
+      "learning_rate": 1.6503617438108243e-06,
+      "loss": 0.3275,
+      "step": 9069
+    },
+    {
+      "epoch": 2.2811871227364184,
+      "grad_norm": 0.2872998118400574,
+      "learning_rate": 1.6492755336136945e-06,
+      "loss": 0.306,
+      "step": 9070
+    },
+    {
+      "epoch": 2.2814386317907447,
+      "grad_norm": 0.29634666442871094,
+      "learning_rate": 1.6481896103851513e-06,
+      "loss": 0.3245,
+      "step": 9071
+    },
+    {
+      "epoch": 2.2816901408450705,
+      "grad_norm": 0.28420212864875793,
+      "learning_rate": 1.647103974218201e-06,
+      "loss": 0.3308,
+      "step": 9072
+    },
+    {
+      "epoch": 2.2819416498993963,
+      "grad_norm": 0.3027746081352234,
+      "learning_rate": 1.6460186252058181e-06,
+      "loss": 0.3234,
+      "step": 9073
+    },
+    {
+      "epoch": 2.2821931589537225,
+      "grad_norm": 0.3052777945995331,
+      "learning_rate": 1.6449335634409575e-06,
+      "loss": 0.3114,
+      "step": 9074
+    },
+    {
+      "epoch": 2.2824446680080483,
+      "grad_norm": 0.299187570810318,
+      "learning_rate": 1.6438487890165494e-06,
+      "loss": 0.3206,
+      "step": 9075
+    },
+    {
+      "epoch": 2.282696177062374,
+      "grad_norm": 0.2914719879627228,
+      "learning_rate": 1.6427643020254951e-06,
+      "loss": 0.3431,
+      "step": 9076
+    },
+    {
+      "epoch": 2.2829476861167004,
+      "grad_norm": 0.29015323519706726,
+      "learning_rate": 1.641680102560677e-06,
+      "loss": 0.3297,
+      "step": 9077
+    },
+    {
+      "epoch": 2.283199195171026,
+      "grad_norm": 0.28477156162261963,
+      "learning_rate": 1.640596190714947e-06,
+      "loss": 0.3376,
+      "step": 9078
+    },
+    {
+      "epoch": 2.283450704225352,
+      "grad_norm": 0.2982317805290222,
+      "learning_rate": 1.6395125665811385e-06,
+      "loss": 0.3274,
+      "step": 9079
+    },
+    {
+      "epoch": 2.2837022132796783,
+      "grad_norm": 0.3212500512599945,
+      "learning_rate": 1.6384292302520537e-06,
+      "loss": 0.3396,
+      "step": 9080
+    },
+    {
+      "epoch": 2.283953722334004,
+      "grad_norm": 0.2954564392566681,
+      "learning_rate": 1.6373461818204773e-06,
+      "loss": 0.3362,
+      "step": 9081
+    },
+    {
+      "epoch": 2.28420523138833,
+      "grad_norm": 0.306144118309021,
+      "learning_rate": 1.6362634213791617e-06,
+      "loss": 0.34,
+      "step": 9082
+    },
+    {
+      "epoch": 2.284456740442656,
+      "grad_norm": 0.28161266446113586,
+      "learning_rate": 1.6351809490208426e-06,
+      "loss": 0.3189,
+      "step": 9083
+    },
+    {
+      "epoch": 2.284708249496982,
+      "grad_norm": 0.3151404857635498,
+      "learning_rate": 1.6340987648382233e-06,
+      "loss": 0.3294,
+      "step": 9084
+    },
+    {
+      "epoch": 2.2849597585513077,
+      "grad_norm": 0.28344619274139404,
+      "learning_rate": 1.6330168689239879e-06,
+      "loss": 0.317,
+      "step": 9085
+    },
+    {
+      "epoch": 2.285211267605634,
+      "grad_norm": 0.2821776568889618,
+      "learning_rate": 1.6319352613707956e-06,
+      "loss": 0.3065,
+      "step": 9086
+    },
+    {
+      "epoch": 2.28546277665996,
+      "grad_norm": 0.29392242431640625,
+      "learning_rate": 1.6308539422712756e-06,
+      "loss": 0.3496,
+      "step": 9087
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.2900041937828064,
+      "learning_rate": 1.62977291171804e-06,
+      "loss": 0.303,
+      "step": 9088
+    },
+    {
+      "epoch": 2.285965794768612,
+      "grad_norm": 0.30908921360969543,
+      "learning_rate": 1.6286921698036685e-06,
+      "loss": 0.3404,
+      "step": 9089
+    },
+    {
+      "epoch": 2.2862173038229376,
+      "grad_norm": 0.28658002614974976,
+      "learning_rate": 1.627611716620724e-06,
+      "loss": 0.3158,
+      "step": 9090
+    },
+    {
+      "epoch": 2.2864688128772634,
+      "grad_norm": 0.2914651930332184,
+      "learning_rate": 1.6265315522617365e-06,
+      "loss": 0.3237,
+      "step": 9091
+    },
+    {
+      "epoch": 2.2867203219315897,
+      "grad_norm": 0.3053009808063507,
+      "learning_rate": 1.6254516768192185e-06,
+      "loss": 0.3383,
+      "step": 9092
+    },
+    {
+      "epoch": 2.2869718309859155,
+      "grad_norm": 0.2992967367172241,
+      "learning_rate": 1.6243720903856518e-06,
+      "loss": 0.3272,
+      "step": 9093
+    },
+    {
+      "epoch": 2.2872233400402413,
+      "grad_norm": 0.2941541075706482,
+      "learning_rate": 1.6232927930534997e-06,
+      "loss": 0.3164,
+      "step": 9094
+    },
+    {
+      "epoch": 2.2874748490945676,
+      "grad_norm": 0.28971970081329346,
+      "learning_rate": 1.6222137849151932e-06,
+      "loss": 0.3291,
+      "step": 9095
+    },
+    {
+      "epoch": 2.2877263581488934,
+      "grad_norm": 0.27451595664024353,
+      "learning_rate": 1.6211350660631448e-06,
+      "loss": 0.2967,
+      "step": 9096
+    },
+    {
+      "epoch": 2.287977867203219,
+      "grad_norm": 0.2858002483844757,
+      "learning_rate": 1.6200566365897414e-06,
+      "loss": 0.3311,
+      "step": 9097
+    },
+    {
+      "epoch": 2.2882293762575454,
+      "grad_norm": 0.28279751539230347,
+      "learning_rate": 1.6189784965873407e-06,
+      "loss": 0.3334,
+      "step": 9098
+    },
+    {
+      "epoch": 2.288480885311871,
+      "grad_norm": 0.29283228516578674,
+      "learning_rate": 1.617900646148282e-06,
+      "loss": 0.3158,
+      "step": 9099
+    },
+    {
+      "epoch": 2.288732394366197,
+      "grad_norm": 0.29870137572288513,
+      "learning_rate": 1.6168230853648725e-06,
+      "loss": 0.3458,
+      "step": 9100
+    },
+    {
+      "epoch": 2.2889839034205233,
+      "grad_norm": 0.2867933213710785,
+      "learning_rate": 1.615745814329403e-06,
+      "loss": 0.3067,
+      "step": 9101
+    },
+    {
+      "epoch": 2.289235412474849,
+      "grad_norm": 0.27168020606040955,
+      "learning_rate": 1.6146688331341303e-06,
+      "loss": 0.3111,
+      "step": 9102
+    },
+    {
+      "epoch": 2.289486921529175,
+      "grad_norm": 0.2864697277545929,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.3208,
+      "step": 9103
+    },
+    {
+      "epoch": 2.289738430583501,
+      "grad_norm": 0.2777710556983948,
+      "learning_rate": 1.6125157406331065e-06,
+      "loss": 0.3309,
+      "step": 9104
+    },
+    {
+      "epoch": 2.289989939637827,
+      "grad_norm": 0.27852359414100647,
+      "learning_rate": 1.6114396295117547e-06,
+      "loss": 0.3057,
+      "step": 9105
+    },
+    {
+      "epoch": 2.2902414486921527,
+      "grad_norm": 0.30460140109062195,
+      "learning_rate": 1.6103638085993972e-06,
+      "loss": 0.3214,
+      "step": 9106
+    },
+    {
+      "epoch": 2.290492957746479,
+      "grad_norm": 0.3020101487636566,
+      "learning_rate": 1.6092882779881746e-06,
+      "loss": 0.3219,
+      "step": 9107
+    },
+    {
+      "epoch": 2.290744466800805,
+      "grad_norm": 0.29373854398727417,
+      "learning_rate": 1.6082130377702004e-06,
+      "loss": 0.3454,
+      "step": 9108
+    },
+    {
+      "epoch": 2.2909959758551306,
+      "grad_norm": 0.27670902013778687,
+      "learning_rate": 1.6071380880375586e-06,
+      "loss": 0.3154,
+      "step": 9109
+    },
+    {
+      "epoch": 2.291247484909457,
+      "grad_norm": 0.3227442502975464,
+      "learning_rate": 1.6060634288823158e-06,
+      "loss": 0.3082,
+      "step": 9110
+    },
+    {
+      "epoch": 2.2914989939637826,
+      "grad_norm": 0.3054077923297882,
+      "learning_rate": 1.6049890603965063e-06,
+      "loss": 0.3121,
+      "step": 9111
+    },
+    {
+      "epoch": 2.2917505030181085,
+      "grad_norm": 0.2847161293029785,
+      "learning_rate": 1.6039149826721462e-06,
+      "loss": 0.2923,
+      "step": 9112
+    },
+    {
+      "epoch": 2.2920020120724347,
+      "grad_norm": 0.3265150487422943,
+      "learning_rate": 1.6028411958012203e-06,
+      "loss": 0.3319,
+      "step": 9113
+    },
+    {
+      "epoch": 2.2922535211267605,
+      "grad_norm": 0.2997869551181793,
+      "learning_rate": 1.6017676998756947e-06,
+      "loss": 0.3079,
+      "step": 9114
+    },
+    {
+      "epoch": 2.2925050301810863,
+      "grad_norm": 0.3061138391494751,
+      "learning_rate": 1.6006944949875052e-06,
+      "loss": 0.3217,
+      "step": 9115
+    },
+    {
+      "epoch": 2.2927565392354126,
+      "grad_norm": 0.2861214876174927,
+      "learning_rate": 1.5996215812285682e-06,
+      "loss": 0.3015,
+      "step": 9116
+    },
+    {
+      "epoch": 2.2930080482897384,
+      "grad_norm": 0.31791844964027405,
+      "learning_rate": 1.5985489586907676e-06,
+      "loss": 0.3303,
+      "step": 9117
+    },
+    {
+      "epoch": 2.293259557344064,
+      "grad_norm": 0.2969111502170563,
+      "learning_rate": 1.5974766274659697e-06,
+      "loss": 0.3213,
+      "step": 9118
+    },
+    {
+      "epoch": 2.2935110663983904,
+      "grad_norm": 0.2913528084754944,
+      "learning_rate": 1.5964045876460143e-06,
+      "loss": 0.3531,
+      "step": 9119
+    },
+    {
+      "epoch": 2.2937625754527162,
+      "grad_norm": 0.30313563346862793,
+      "learning_rate": 1.5953328393227113e-06,
+      "loss": 0.3048,
+      "step": 9120
+    },
+    {
+      "epoch": 2.294014084507042,
+      "grad_norm": 0.3050040304660797,
+      "learning_rate": 1.5942613825878527e-06,
+      "loss": 0.3299,
+      "step": 9121
+    },
+    {
+      "epoch": 2.2942655935613683,
+      "grad_norm": 0.2819101810455322,
+      "learning_rate": 1.5931902175331986e-06,
+      "loss": 0.3041,
+      "step": 9122
+    },
+    {
+      "epoch": 2.294517102615694,
+      "grad_norm": 0.28545475006103516,
+      "learning_rate": 1.5921193442504918e-06,
+      "loss": 0.3063,
+      "step": 9123
+    },
+    {
+      "epoch": 2.29476861167002,
+      "grad_norm": 0.33305785059928894,
+      "learning_rate": 1.5910487628314414e-06,
+      "loss": 0.3053,
+      "step": 9124
+    },
+    {
+      "epoch": 2.295020120724346,
+      "grad_norm": 0.30768680572509766,
+      "learning_rate": 1.5899784733677397e-06,
+      "loss": 0.3215,
+      "step": 9125
+    },
+    {
+      "epoch": 2.295271629778672,
+      "grad_norm": 0.28684544563293457,
+      "learning_rate": 1.588908475951047e-06,
+      "loss": 0.3256,
+      "step": 9126
+    },
+    {
+      "epoch": 2.2955231388329977,
+      "grad_norm": 0.2967512011528015,
+      "learning_rate": 1.5878387706730053e-06,
+      "loss": 0.3377,
+      "step": 9127
+    },
+    {
+      "epoch": 2.295774647887324,
+      "grad_norm": 0.29876625537872314,
+      "learning_rate": 1.5867693576252252e-06,
+      "loss": 0.3176,
+      "step": 9128
+    },
+    {
+      "epoch": 2.29602615694165,
+      "grad_norm": 0.3200327754020691,
+      "learning_rate": 1.5857002368992963e-06,
+      "loss": 0.3297,
+      "step": 9129
+    },
+    {
+      "epoch": 2.296277665995976,
+      "grad_norm": 0.31204643845558167,
+      "learning_rate": 1.5846314085867836e-06,
+      "loss": 0.3412,
+      "step": 9130
+    },
+    {
+      "epoch": 2.296529175050302,
+      "grad_norm": 0.29177501797676086,
+      "learning_rate": 1.5835628727792228e-06,
+      "loss": 0.3191,
+      "step": 9131
+    },
+    {
+      "epoch": 2.2967806841046277,
+      "grad_norm": 0.2729922831058502,
+      "learning_rate": 1.582494629568131e-06,
+      "loss": 0.3133,
+      "step": 9132
+    },
+    {
+      "epoch": 2.297032193158954,
+      "grad_norm": 0.3033842146396637,
+      "learning_rate": 1.5814266790449922e-06,
+      "loss": 0.313,
+      "step": 9133
+    },
+    {
+      "epoch": 2.2972837022132797,
+      "grad_norm": 0.29062163829803467,
+      "learning_rate": 1.5803590213012738e-06,
+      "loss": 0.2842,
+      "step": 9134
+    },
+    {
+      "epoch": 2.2975352112676055,
+      "grad_norm": 0.2890336215496063,
+      "learning_rate": 1.57929165642841e-06,
+      "loss": 0.3044,
+      "step": 9135
+    },
+    {
+      "epoch": 2.2977867203219318,
+      "grad_norm": 0.2879352569580078,
+      "learning_rate": 1.578224584517818e-06,
+      "loss": 0.3019,
+      "step": 9136
+    },
+    {
+      "epoch": 2.2980382293762576,
+      "grad_norm": 0.3032044768333435,
+      "learning_rate": 1.5771578056608816e-06,
+      "loss": 0.3173,
+      "step": 9137
+    },
+    {
+      "epoch": 2.2982897384305834,
+      "grad_norm": 0.3006070852279663,
+      "learning_rate": 1.5760913199489674e-06,
+      "loss": 0.3308,
+      "step": 9138
+    },
+    {
+      "epoch": 2.2985412474849096,
+      "grad_norm": 0.3287603259086609,
+      "learning_rate": 1.5750251274734107e-06,
+      "loss": 0.3182,
+      "step": 9139
+    },
+    {
+      "epoch": 2.2987927565392354,
+      "grad_norm": 0.2758319675922394,
+      "learning_rate": 1.5739592283255251e-06,
+      "loss": 0.3166,
+      "step": 9140
+    },
+    {
+      "epoch": 2.2990442655935612,
+      "grad_norm": 0.29488644003868103,
+      "learning_rate": 1.5728936225966002e-06,
+      "loss": 0.293,
+      "step": 9141
+    },
+    {
+      "epoch": 2.2992957746478875,
+      "grad_norm": 0.31970569491386414,
+      "learning_rate": 1.5718283103778941e-06,
+      "loss": 0.313,
+      "step": 9142
+    },
+    {
+      "epoch": 2.2995472837022133,
+      "grad_norm": 0.29525116086006165,
+      "learning_rate": 1.5707632917606491e-06,
+      "loss": 0.3212,
+      "step": 9143
+    },
+    {
+      "epoch": 2.299798792756539,
+      "grad_norm": 0.29070043563842773,
+      "learning_rate": 1.569698566836073e-06,
+      "loss": 0.3151,
+      "step": 9144
+    },
+    {
+      "epoch": 2.3000503018108653,
+      "grad_norm": 0.3086007237434387,
+      "learning_rate": 1.5686341356953566e-06,
+      "loss": 0.3343,
+      "step": 9145
+    },
+    {
+      "epoch": 2.300301810865191,
+      "grad_norm": 0.3106892704963684,
+      "learning_rate": 1.5675699984296584e-06,
+      "loss": 0.3108,
+      "step": 9146
+    },
+    {
+      "epoch": 2.300553319919517,
+      "grad_norm": 0.2823115289211273,
+      "learning_rate": 1.5665061551301175e-06,
+      "loss": 0.3255,
+      "step": 9147
+    },
+    {
+      "epoch": 2.300804828973843,
+      "grad_norm": 0.3261948227882385,
+      "learning_rate": 1.5654426058878436e-06,
+      "loss": 0.3265,
+      "step": 9148
+    },
+    {
+      "epoch": 2.301056338028169,
+      "grad_norm": 0.30333390831947327,
+      "learning_rate": 1.5643793507939253e-06,
+      "loss": 0.3367,
+      "step": 9149
+    },
+    {
+      "epoch": 2.301307847082495,
+      "grad_norm": 0.28751394152641296,
+      "learning_rate": 1.5633163899394211e-06,
+      "loss": 0.3166,
+      "step": 9150
+    },
+    {
+      "epoch": 2.301559356136821,
+      "grad_norm": 0.2967854142189026,
+      "learning_rate": 1.5622537234153695e-06,
+      "loss": 0.3093,
+      "step": 9151
+    },
+    {
+      "epoch": 2.301810865191147,
+      "grad_norm": 0.2856552004814148,
+      "learning_rate": 1.561191351312779e-06,
+      "loss": 0.3194,
+      "step": 9152
+    },
+    {
+      "epoch": 2.3020623742454727,
+      "grad_norm": 0.30799177289009094,
+      "learning_rate": 1.560129273722637e-06,
+      "loss": 0.313,
+      "step": 9153
+    },
+    {
+      "epoch": 2.302313883299799,
+      "grad_norm": 0.2922762334346771,
+      "learning_rate": 1.559067490735902e-06,
+      "loss": 0.3232,
+      "step": 9154
+    },
+    {
+      "epoch": 2.3025653923541247,
+      "grad_norm": 0.28383079171180725,
+      "learning_rate": 1.5580060024435112e-06,
+      "loss": 0.3039,
+      "step": 9155
+    },
+    {
+      "epoch": 2.3028169014084505,
+      "grad_norm": 0.2966978847980499,
+      "learning_rate": 1.556944808936372e-06,
+      "loss": 0.3438,
+      "step": 9156
+    },
+    {
+      "epoch": 2.3030684104627768,
+      "grad_norm": 0.3017922639846802,
+      "learning_rate": 1.5558839103053713e-06,
+      "loss": 0.3177,
+      "step": 9157
+    },
+    {
+      "epoch": 2.3033199195171026,
+      "grad_norm": 0.2967662513256073,
+      "learning_rate": 1.554823306641366e-06,
+      "loss": 0.2977,
+      "step": 9158
+    },
+    {
+      "epoch": 2.3035714285714284,
+      "grad_norm": 0.2964570224285126,
+      "learning_rate": 1.5537629980351932e-06,
+      "loss": 0.3134,
+      "step": 9159
+    },
+    {
+      "epoch": 2.3038229376257546,
+      "grad_norm": 0.307412713766098,
+      "learning_rate": 1.5527029845776587e-06,
+      "loss": 0.3429,
+      "step": 9160
+    },
+    {
+      "epoch": 2.3040744466800804,
+      "grad_norm": 0.32412174344062805,
+      "learning_rate": 1.5516432663595483e-06,
+      "loss": 0.3201,
+      "step": 9161
+    },
+    {
+      "epoch": 2.3043259557344067,
+      "grad_norm": 0.3185696601867676,
+      "learning_rate": 1.550583843471618e-06,
+      "loss": 0.3388,
+      "step": 9162
+    },
+    {
+      "epoch": 2.3045774647887325,
+      "grad_norm": 0.2888423502445221,
+      "learning_rate": 1.5495247160046039e-06,
+      "loss": 0.3304,
+      "step": 9163
+    },
+    {
+      "epoch": 2.3048289738430583,
+      "grad_norm": 0.2889467477798462,
+      "learning_rate": 1.54846588404921e-06,
+      "loss": 0.3088,
+      "step": 9164
+    },
+    {
+      "epoch": 2.3050804828973845,
+      "grad_norm": 0.3054656982421875,
+      "learning_rate": 1.5474073476961216e-06,
+      "loss": 0.3256,
+      "step": 9165
+    },
+    {
+      "epoch": 2.3053319919517103,
+      "grad_norm": 0.30939677357673645,
+      "learning_rate": 1.5463491070359937e-06,
+      "loss": 0.3316,
+      "step": 9166
+    },
+    {
+      "epoch": 2.305583501006036,
+      "grad_norm": 0.31722491979599,
+      "learning_rate": 1.5452911621594596e-06,
+      "loss": 0.314,
+      "step": 9167
+    },
+    {
+      "epoch": 2.3058350100603624,
+      "grad_norm": 0.31227657198905945,
+      "learning_rate": 1.544233513157124e-06,
+      "loss": 0.3222,
+      "step": 9168
+    },
+    {
+      "epoch": 2.306086519114688,
+      "grad_norm": 0.2999171018600464,
+      "learning_rate": 1.54317616011957e-06,
+      "loss": 0.3119,
+      "step": 9169
+    },
+    {
+      "epoch": 2.306338028169014,
+      "grad_norm": 0.30300676822662354,
+      "learning_rate": 1.54211910313735e-06,
+      "loss": 0.2988,
+      "step": 9170
+    },
+    {
+      "epoch": 2.3065895372233403,
+      "grad_norm": 0.31904926896095276,
+      "learning_rate": 1.541062342300997e-06,
+      "loss": 0.3285,
+      "step": 9171
+    },
+    {
+      "epoch": 2.306841046277666,
+      "grad_norm": 0.29842498898506165,
+      "learning_rate": 1.5400058777010169e-06,
+      "loss": 0.3163,
+      "step": 9172
+    },
+    {
+      "epoch": 2.307092555331992,
+      "grad_norm": 0.3134821951389313,
+      "learning_rate": 1.5389497094278861e-06,
+      "loss": 0.3168,
+      "step": 9173
+    },
+    {
+      "epoch": 2.307344064386318,
+      "grad_norm": 0.3033539950847626,
+      "learning_rate": 1.537893837572062e-06,
+      "loss": 0.2992,
+      "step": 9174
+    },
+    {
+      "epoch": 2.307595573440644,
+      "grad_norm": 0.2625449001789093,
+      "learning_rate": 1.5368382622239703e-06,
+      "loss": 0.3167,
+      "step": 9175
+    },
+    {
+      "epoch": 2.3078470824949697,
+      "grad_norm": 0.3125070035457611,
+      "learning_rate": 1.5357829834740174e-06,
+      "loss": 0.3273,
+      "step": 9176
+    },
+    {
+      "epoch": 2.308098591549296,
+      "grad_norm": 0.3032437264919281,
+      "learning_rate": 1.5347280014125782e-06,
+      "loss": 0.3069,
+      "step": 9177
+    },
+    {
+      "epoch": 2.308350100603622,
+      "grad_norm": 0.3202145993709564,
+      "learning_rate": 1.5336733161300088e-06,
+      "loss": 0.3211,
+      "step": 9178
+    },
+    {
+      "epoch": 2.3086016096579476,
+      "grad_norm": 0.30158209800720215,
+      "learning_rate": 1.5326189277166325e-06,
+      "loss": 0.3295,
+      "step": 9179
+    },
+    {
+      "epoch": 2.308853118712274,
+      "grad_norm": 0.3015068769454956,
+      "learning_rate": 1.5315648362627556e-06,
+      "loss": 0.3257,
+      "step": 9180
+    },
+    {
+      "epoch": 2.3091046277665996,
+      "grad_norm": 0.30036771297454834,
+      "learning_rate": 1.53051104185865e-06,
+      "loss": 0.3132,
+      "step": 9181
+    },
+    {
+      "epoch": 2.3093561368209254,
+      "grad_norm": 0.2994939386844635,
+      "learning_rate": 1.5294575445945687e-06,
+      "loss": 0.3275,
+      "step": 9182
+    },
+    {
+      "epoch": 2.3096076458752517,
+      "grad_norm": 0.3159419298171997,
+      "learning_rate": 1.5284043445607383e-06,
+      "loss": 0.2978,
+      "step": 9183
+    },
+    {
+      "epoch": 2.3098591549295775,
+      "grad_norm": 0.3039325475692749,
+      "learning_rate": 1.5273514418473566e-06,
+      "loss": 0.3127,
+      "step": 9184
+    },
+    {
+      "epoch": 2.3101106639839033,
+      "grad_norm": 0.32907170057296753,
+      "learning_rate": 1.5262988365446002e-06,
+      "loss": 0.3283,
+      "step": 9185
+    },
+    {
+      "epoch": 2.3103621730382295,
+      "grad_norm": 0.30235663056373596,
+      "learning_rate": 1.5252465287426154e-06,
+      "loss": 0.3055,
+      "step": 9186
+    },
+    {
+      "epoch": 2.3106136820925554,
+      "grad_norm": 0.30011188983917236,
+      "learning_rate": 1.5241945185315292e-06,
+      "loss": 0.3435,
+      "step": 9187
+    },
+    {
+      "epoch": 2.310865191146881,
+      "grad_norm": 0.3140828311443329,
+      "learning_rate": 1.5231428060014363e-06,
+      "loss": 0.3229,
+      "step": 9188
+    },
+    {
+      "epoch": 2.3111167002012074,
+      "grad_norm": 0.30959945917129517,
+      "learning_rate": 1.5220913912424128e-06,
+      "loss": 0.3224,
+      "step": 9189
+    },
+    {
+      "epoch": 2.311368209255533,
+      "grad_norm": 0.2961542010307312,
+      "learning_rate": 1.5210402743445018e-06,
+      "loss": 0.343,
+      "step": 9190
+    },
+    {
+      "epoch": 2.311619718309859,
+      "grad_norm": 0.31126269698143005,
+      "learning_rate": 1.519989455397729e-06,
+      "loss": 0.2833,
+      "step": 9191
+    },
+    {
+      "epoch": 2.3118712273641853,
+      "grad_norm": 0.2786300480365753,
+      "learning_rate": 1.518938934492087e-06,
+      "loss": 0.3245,
+      "step": 9192
+    },
+    {
+      "epoch": 2.312122736418511,
+      "grad_norm": 0.30616483092308044,
+      "learning_rate": 1.5178887117175472e-06,
+      "loss": 0.3253,
+      "step": 9193
+    },
+    {
+      "epoch": 2.312374245472837,
+      "grad_norm": 0.27918025851249695,
+      "learning_rate": 1.5168387871640572e-06,
+      "loss": 0.325,
+      "step": 9194
+    },
+    {
+      "epoch": 2.312625754527163,
+      "grad_norm": 0.2911374866962433,
+      "learning_rate": 1.5157891609215331e-06,
+      "loss": 0.3209,
+      "step": 9195
+    },
+    {
+      "epoch": 2.312877263581489,
+      "grad_norm": 0.28203797340393066,
+      "learning_rate": 1.5147398330798712e-06,
+      "loss": 0.3063,
+      "step": 9196
+    },
+    {
+      "epoch": 2.3131287726358147,
+      "grad_norm": 0.29836559295654297,
+      "learning_rate": 1.5136908037289377e-06,
+      "loss": 0.3154,
+      "step": 9197
+    },
+    {
+      "epoch": 2.313380281690141,
+      "grad_norm": 0.3110038936138153,
+      "learning_rate": 1.5126420729585784e-06,
+      "loss": 0.3066,
+      "step": 9198
+    },
+    {
+      "epoch": 2.313631790744467,
+      "grad_norm": 0.3166778087615967,
+      "learning_rate": 1.5115936408586069e-06,
+      "loss": 0.3624,
+      "step": 9199
+    },
+    {
+      "epoch": 2.3138832997987926,
+      "grad_norm": 0.2844400107860565,
+      "learning_rate": 1.510545507518818e-06,
+      "loss": 0.3125,
+      "step": 9200
+    },
+    {
+      "epoch": 2.314134808853119,
+      "grad_norm": 0.3151662051677704,
+      "learning_rate": 1.5094976730289751e-06,
+      "loss": 0.3391,
+      "step": 9201
+    },
+    {
+      "epoch": 2.3143863179074446,
+      "grad_norm": 0.30355745553970337,
+      "learning_rate": 1.5084501374788213e-06,
+      "loss": 0.3132,
+      "step": 9202
+    },
+    {
+      "epoch": 2.3146378269617705,
+      "grad_norm": 0.3106285631656647,
+      "learning_rate": 1.5074029009580687e-06,
+      "loss": 0.3235,
+      "step": 9203
+    },
+    {
+      "epoch": 2.3148893360160967,
+      "grad_norm": 0.31202206015586853,
+      "learning_rate": 1.5063559635564078e-06,
+      "loss": 0.362,
+      "step": 9204
+    },
+    {
+      "epoch": 2.3151408450704225,
+      "grad_norm": 0.2972472012042999,
+      "learning_rate": 1.5053093253635043e-06,
+      "loss": 0.3213,
+      "step": 9205
+    },
+    {
+      "epoch": 2.3153923541247483,
+      "grad_norm": 0.29215607047080994,
+      "learning_rate": 1.5042629864689927e-06,
+      "loss": 0.3171,
+      "step": 9206
+    },
+    {
+      "epoch": 2.3156438631790746,
+      "grad_norm": 0.3004598617553711,
+      "learning_rate": 1.5032169469624892e-06,
+      "loss": 0.3167,
+      "step": 9207
+    },
+    {
+      "epoch": 2.3158953722334004,
+      "grad_norm": 0.2856040894985199,
+      "learning_rate": 1.502171206933576e-06,
+      "loss": 0.3203,
+      "step": 9208
+    },
+    {
+      "epoch": 2.316146881287726,
+      "grad_norm": 0.2829803228378296,
+      "learning_rate": 1.5011257664718187e-06,
+      "loss": 0.315,
+      "step": 9209
+    },
+    {
+      "epoch": 2.3163983903420524,
+      "grad_norm": 0.28959813714027405,
+      "learning_rate": 1.500080625666749e-06,
+      "loss": 0.3013,
+      "step": 9210
+    },
+    {
+      "epoch": 2.316649899396378,
+      "grad_norm": 0.3246660530567169,
+      "learning_rate": 1.4990357846078795e-06,
+      "loss": 0.3109,
+      "step": 9211
+    },
+    {
+      "epoch": 2.316901408450704,
+      "grad_norm": 0.30734169483184814,
+      "learning_rate": 1.4979912433846917e-06,
+      "loss": 0.3233,
+      "step": 9212
+    },
+    {
+      "epoch": 2.3171529175050303,
+      "grad_norm": 0.3003253638744354,
+      "learning_rate": 1.4969470020866467e-06,
+      "loss": 0.3286,
+      "step": 9213
+    },
+    {
+      "epoch": 2.317404426559356,
+      "grad_norm": 0.31494301557540894,
+      "learning_rate": 1.4959030608031749e-06,
+      "loss": 0.3244,
+      "step": 9214
+    },
+    {
+      "epoch": 2.317655935613682,
+      "grad_norm": 0.3177368640899658,
+      "learning_rate": 1.4948594196236838e-06,
+      "loss": 0.2982,
+      "step": 9215
+    },
+    {
+      "epoch": 2.317907444668008,
+      "grad_norm": 0.2995319664478302,
+      "learning_rate": 1.4938160786375571e-06,
+      "loss": 0.3289,
+      "step": 9216
+    },
+    {
+      "epoch": 2.318158953722334,
+      "grad_norm": 0.32530516386032104,
+      "learning_rate": 1.4927730379341476e-06,
+      "loss": 0.3391,
+      "step": 9217
+    },
+    {
+      "epoch": 2.3184104627766597,
+      "grad_norm": 0.2980848252773285,
+      "learning_rate": 1.491730297602787e-06,
+      "loss": 0.3546,
+      "step": 9218
+    },
+    {
+      "epoch": 2.318661971830986,
+      "grad_norm": 0.2969314455986023,
+      "learning_rate": 1.4906878577327776e-06,
+      "loss": 0.3178,
+      "step": 9219
+    },
+    {
+      "epoch": 2.318913480885312,
+      "grad_norm": 0.2775447368621826,
+      "learning_rate": 1.4896457184134005e-06,
+      "loss": 0.3162,
+      "step": 9220
+    },
+    {
+      "epoch": 2.3191649899396376,
+      "grad_norm": 0.2855249047279358,
+      "learning_rate": 1.4886038797339058e-06,
+      "loss": 0.3313,
+      "step": 9221
+    },
+    {
+      "epoch": 2.319416498993964,
+      "grad_norm": 0.3028770089149475,
+      "learning_rate": 1.4875623417835227e-06,
+      "loss": 0.3522,
+      "step": 9222
+    },
+    {
+      "epoch": 2.3196680080482897,
+      "grad_norm": 0.2914048731327057,
+      "learning_rate": 1.4865211046514494e-06,
+      "loss": 0.2839,
+      "step": 9223
+    },
+    {
+      "epoch": 2.3199195171026155,
+      "grad_norm": 0.3016049563884735,
+      "learning_rate": 1.4854801684268655e-06,
+      "loss": 0.3268,
+      "step": 9224
+    },
+    {
+      "epoch": 2.3201710261569417,
+      "grad_norm": 0.3333890438079834,
+      "learning_rate": 1.4844395331989164e-06,
+      "loss": 0.324,
+      "step": 9225
+    },
+    {
+      "epoch": 2.3204225352112675,
+      "grad_norm": 0.2845839858055115,
+      "learning_rate": 1.4833991990567281e-06,
+      "loss": 0.3258,
+      "step": 9226
+    },
+    {
+      "epoch": 2.3206740442655933,
+      "grad_norm": 0.30603617429733276,
+      "learning_rate": 1.4823591660894e-06,
+      "loss": 0.3091,
+      "step": 9227
+    },
+    {
+      "epoch": 2.3209255533199196,
+      "grad_norm": 0.2999797761440277,
+      "learning_rate": 1.4813194343860015e-06,
+      "loss": 0.3111,
+      "step": 9228
+    },
+    {
+      "epoch": 2.3211770623742454,
+      "grad_norm": 0.29560771584510803,
+      "learning_rate": 1.4802800040355825e-06,
+      "loss": 0.3128,
+      "step": 9229
+    },
+    {
+      "epoch": 2.3214285714285716,
+      "grad_norm": 0.28773796558380127,
+      "learning_rate": 1.4792408751271603e-06,
+      "loss": 0.3219,
+      "step": 9230
+    },
+    {
+      "epoch": 2.3216800804828974,
+      "grad_norm": 0.3093893826007843,
+      "learning_rate": 1.4782020477497328e-06,
+      "loss": 0.3332,
+      "step": 9231
+    },
+    {
+      "epoch": 2.3219315895372232,
+      "grad_norm": 0.3124389350414276,
+      "learning_rate": 1.4771635219922658e-06,
+      "loss": 0.3106,
+      "step": 9232
+    },
+    {
+      "epoch": 2.3221830985915495,
+      "grad_norm": 0.29394763708114624,
+      "learning_rate": 1.4761252979437062e-06,
+      "loss": 0.3145,
+      "step": 9233
+    },
+    {
+      "epoch": 2.3224346076458753,
+      "grad_norm": 0.29244881868362427,
+      "learning_rate": 1.475087375692968e-06,
+      "loss": 0.3295,
+      "step": 9234
+    },
+    {
+      "epoch": 2.322686116700201,
+      "grad_norm": 0.3140466511249542,
+      "learning_rate": 1.4740497553289456e-06,
+      "loss": 0.3344,
+      "step": 9235
+    },
+    {
+      "epoch": 2.3229376257545273,
+      "grad_norm": 0.2974216341972351,
+      "learning_rate": 1.473012436940502e-06,
+      "loss": 0.3238,
+      "step": 9236
+    },
+    {
+      "epoch": 2.323189134808853,
+      "grad_norm": 0.29725712537765503,
+      "learning_rate": 1.4719754206164782e-06,
+      "loss": 0.3307,
+      "step": 9237
+    },
+    {
+      "epoch": 2.323440643863179,
+      "grad_norm": 0.3167237937450409,
+      "learning_rate": 1.4709387064456899e-06,
+      "loss": 0.3255,
+      "step": 9238
+    },
+    {
+      "epoch": 2.323692152917505,
+      "grad_norm": 0.28957289457321167,
+      "learning_rate": 1.4699022945169221e-06,
+      "loss": 0.3095,
+      "step": 9239
+    },
+    {
+      "epoch": 2.323943661971831,
+      "grad_norm": 0.2918396294116974,
+      "learning_rate": 1.4688661849189407e-06,
+      "loss": 0.34,
+      "step": 9240
+    },
+    {
+      "epoch": 2.324195171026157,
+      "grad_norm": 0.29980701208114624,
+      "learning_rate": 1.4678303777404778e-06,
+      "loss": 0.3276,
+      "step": 9241
+    },
+    {
+      "epoch": 2.324446680080483,
+      "grad_norm": 0.2971269488334656,
+      "learning_rate": 1.4667948730702474e-06,
+      "loss": 0.3113,
+      "step": 9242
+    },
+    {
+      "epoch": 2.324698189134809,
+      "grad_norm": 0.29910194873809814,
+      "learning_rate": 1.4657596709969313e-06,
+      "loss": 0.3286,
+      "step": 9243
+    },
+    {
+      "epoch": 2.3249496981891347,
+      "grad_norm": 0.2877163887023926,
+      "learning_rate": 1.46472477160919e-06,
+      "loss": 0.3216,
+      "step": 9244
+    },
+    {
+      "epoch": 2.325201207243461,
+      "grad_norm": 0.308214396238327,
+      "learning_rate": 1.4636901749956544e-06,
+      "loss": 0.3337,
+      "step": 9245
+    },
+    {
+      "epoch": 2.3254527162977867,
+      "grad_norm": 0.3051459491252899,
+      "learning_rate": 1.4626558812449337e-06,
+      "loss": 0.3487,
+      "step": 9246
+    },
+    {
+      "epoch": 2.3257042253521125,
+      "grad_norm": 0.29465314745903015,
+      "learning_rate": 1.461621890445606e-06,
+      "loss": 0.3044,
+      "step": 9247
+    },
+    {
+      "epoch": 2.3259557344064388,
+      "grad_norm": 0.3071502447128296,
+      "learning_rate": 1.4605882026862267e-06,
+      "loss": 0.3422,
+      "step": 9248
+    },
+    {
+      "epoch": 2.3262072434607646,
+      "grad_norm": 0.3024512529373169,
+      "learning_rate": 1.4595548180553275e-06,
+      "loss": 0.3562,
+      "step": 9249
+    },
+    {
+      "epoch": 2.3264587525150904,
+      "grad_norm": 0.2800012528896332,
+      "learning_rate": 1.4585217366414072e-06,
+      "loss": 0.3417,
+      "step": 9250
+    },
+    {
+      "epoch": 2.3267102615694166,
+      "grad_norm": 0.30198147892951965,
+      "learning_rate": 1.4574889585329466e-06,
+      "loss": 0.3352,
+      "step": 9251
+    },
+    {
+      "epoch": 2.3269617706237424,
+      "grad_norm": 0.2976208031177521,
+      "learning_rate": 1.456456483818393e-06,
+      "loss": 0.3333,
+      "step": 9252
+    },
+    {
+      "epoch": 2.3272132796780682,
+      "grad_norm": 0.30566921830177307,
+      "learning_rate": 1.455424312586175e-06,
+      "loss": 0.3108,
+      "step": 9253
+    },
+    {
+      "epoch": 2.3274647887323945,
+      "grad_norm": 0.2987402379512787,
+      "learning_rate": 1.4543924449246882e-06,
+      "loss": 0.3167,
+      "step": 9254
+    },
+    {
+      "epoch": 2.3277162977867203,
+      "grad_norm": 0.3012560307979584,
+      "learning_rate": 1.4533608809223087e-06,
+      "loss": 0.3464,
+      "step": 9255
+    },
+    {
+      "epoch": 2.327967806841046,
+      "grad_norm": 0.3095744848251343,
+      "learning_rate": 1.452329620667381e-06,
+      "loss": 0.3325,
+      "step": 9256
+    },
+    {
+      "epoch": 2.3282193158953723,
+      "grad_norm": 0.29700353741645813,
+      "learning_rate": 1.4512986642482279e-06,
+      "loss": 0.3131,
+      "step": 9257
+    },
+    {
+      "epoch": 2.328470824949698,
+      "grad_norm": 0.2864930033683777,
+      "learning_rate": 1.4502680117531425e-06,
+      "loss": 0.334,
+      "step": 9258
+    },
+    {
+      "epoch": 2.328722334004024,
+      "grad_norm": 0.3204570710659027,
+      "learning_rate": 1.4492376632703947e-06,
+      "loss": 0.345,
+      "step": 9259
+    },
+    {
+      "epoch": 2.32897384305835,
+      "grad_norm": 0.2854207158088684,
+      "learning_rate": 1.4482076188882293e-06,
+      "loss": 0.3135,
+      "step": 9260
+    },
+    {
+      "epoch": 2.329225352112676,
+      "grad_norm": 0.2826644480228424,
+      "learning_rate": 1.4471778786948598e-06,
+      "loss": 0.3473,
+      "step": 9261
+    },
+    {
+      "epoch": 2.3294768611670023,
+      "grad_norm": 0.29891934990882874,
+      "learning_rate": 1.44614844277848e-06,
+      "loss": 0.3143,
+      "step": 9262
+    },
+    {
+      "epoch": 2.329728370221328,
+      "grad_norm": 0.3028249144554138,
+      "learning_rate": 1.4451193112272515e-06,
+      "loss": 0.3143,
+      "step": 9263
+    },
+    {
+      "epoch": 2.329979879275654,
+      "grad_norm": 0.30594244599342346,
+      "learning_rate": 1.4440904841293168e-06,
+      "loss": 0.3291,
+      "step": 9264
+    },
+    {
+      "epoch": 2.33023138832998,
+      "grad_norm": 0.3240032196044922,
+      "learning_rate": 1.4430619615727842e-06,
+      "loss": 0.3143,
+      "step": 9265
+    },
+    {
+      "epoch": 2.330482897384306,
+      "grad_norm": 0.2798107862472534,
+      "learning_rate": 1.442033743645745e-06,
+      "loss": 0.3249,
+      "step": 9266
+    },
+    {
+      "epoch": 2.3307344064386317,
+      "grad_norm": 0.3072926104068756,
+      "learning_rate": 1.4410058304362546e-06,
+      "loss": 0.347,
+      "step": 9267
+    },
+    {
+      "epoch": 2.330985915492958,
+      "grad_norm": 0.3058062493801117,
+      "learning_rate": 1.4399782220323515e-06,
+      "loss": 0.3252,
+      "step": 9268
+    },
+    {
+      "epoch": 2.3312374245472838,
+      "grad_norm": 0.2971736490726471,
+      "learning_rate": 1.4389509185220412e-06,
+      "loss": 0.3435,
+      "step": 9269
+    },
+    {
+      "epoch": 2.3314889336016096,
+      "grad_norm": 0.2837081253528595,
+      "learning_rate": 1.4379239199933082e-06,
+      "loss": 0.3381,
+      "step": 9270
+    },
+    {
+      "epoch": 2.331740442655936,
+      "grad_norm": 0.3149973452091217,
+      "learning_rate": 1.4368972265341052e-06,
+      "loss": 0.3382,
+      "step": 9271
+    },
+    {
+      "epoch": 2.3319919517102616,
+      "grad_norm": 0.3013942837715149,
+      "learning_rate": 1.435870838232366e-06,
+      "loss": 0.3117,
+      "step": 9272
+    },
+    {
+      "epoch": 2.3322434607645874,
+      "grad_norm": 0.3119725286960602,
+      "learning_rate": 1.4348447551759908e-06,
+      "loss": 0.3132,
+      "step": 9273
+    },
+    {
+      "epoch": 2.3324949698189137,
+      "grad_norm": 0.3200342059135437,
+      "learning_rate": 1.4338189774528605e-06,
+      "loss": 0.3183,
+      "step": 9274
+    },
+    {
+      "epoch": 2.3327464788732395,
+      "grad_norm": 0.2716279923915863,
+      "learning_rate": 1.432793505150823e-06,
+      "loss": 0.3251,
+      "step": 9275
+    },
+    {
+      "epoch": 2.3329979879275653,
+      "grad_norm": 0.31028926372528076,
+      "learning_rate": 1.4317683383577074e-06,
+      "loss": 0.3335,
+      "step": 9276
+    },
+    {
+      "epoch": 2.3332494969818915,
+      "grad_norm": 0.2938306927680969,
+      "learning_rate": 1.4307434771613087e-06,
+      "loss": 0.3327,
+      "step": 9277
+    },
+    {
+      "epoch": 2.3335010060362174,
+      "grad_norm": 0.30416056513786316,
+      "learning_rate": 1.429718921649404e-06,
+      "loss": 0.3397,
+      "step": 9278
+    },
+    {
+      "epoch": 2.333752515090543,
+      "grad_norm": 0.3303104341030121,
+      "learning_rate": 1.428694671909736e-06,
+      "loss": 0.3197,
+      "step": 9279
+    },
+    {
+      "epoch": 2.3340040241448694,
+      "grad_norm": 0.29095789790153503,
+      "learning_rate": 1.4276707280300295e-06,
+      "loss": 0.3101,
+      "step": 9280
+    },
+    {
+      "epoch": 2.334255533199195,
+      "grad_norm": 0.308417946100235,
+      "learning_rate": 1.4266470900979746e-06,
+      "loss": 0.2987,
+      "step": 9281
+    },
+    {
+      "epoch": 2.334507042253521,
+      "grad_norm": 0.33862265944480896,
+      "learning_rate": 1.4256237582012433e-06,
+      "loss": 0.3286,
+      "step": 9282
+    },
+    {
+      "epoch": 2.3347585513078473,
+      "grad_norm": 0.28483128547668457,
+      "learning_rate": 1.4246007324274747e-06,
+      "loss": 0.3082,
+      "step": 9283
+    },
+    {
+      "epoch": 2.335010060362173,
+      "grad_norm": 0.3114943504333496,
+      "learning_rate": 1.4235780128642867e-06,
+      "loss": 0.3211,
+      "step": 9284
+    },
+    {
+      "epoch": 2.335261569416499,
+      "grad_norm": 0.30490976572036743,
+      "learning_rate": 1.4225555995992668e-06,
+      "loss": 0.3289,
+      "step": 9285
+    },
+    {
+      "epoch": 2.335513078470825,
+      "grad_norm": 0.29261600971221924,
+      "learning_rate": 1.4215334927199808e-06,
+      "loss": 0.3234,
+      "step": 9286
+    },
+    {
+      "epoch": 2.335764587525151,
+      "grad_norm": 0.30190035700798035,
+      "learning_rate": 1.4205116923139628e-06,
+      "loss": 0.3004,
+      "step": 9287
+    },
+    {
+      "epoch": 2.3360160965794767,
+      "grad_norm": 0.30952534079551697,
+      "learning_rate": 1.4194901984687266e-06,
+      "loss": 0.3236,
+      "step": 9288
+    },
+    {
+      "epoch": 2.336267605633803,
+      "grad_norm": 0.3120158910751343,
+      "learning_rate": 1.4184690112717536e-06,
+      "loss": 0.3316,
+      "step": 9289
+    },
+    {
+      "epoch": 2.336519114688129,
+      "grad_norm": 0.2920396029949188,
+      "learning_rate": 1.417448130810506e-06,
+      "loss": 0.3116,
+      "step": 9290
+    },
+    {
+      "epoch": 2.3367706237424546,
+      "grad_norm": 0.31028714776039124,
+      "learning_rate": 1.4164275571724112e-06,
+      "loss": 0.3198,
+      "step": 9291
+    },
+    {
+      "epoch": 2.337022132796781,
+      "grad_norm": 0.33299851417541504,
+      "learning_rate": 1.4154072904448778e-06,
+      "loss": 0.3254,
+      "step": 9292
+    },
+    {
+      "epoch": 2.3372736418511066,
+      "grad_norm": 0.2730899155139923,
+      "learning_rate": 1.4143873307152867e-06,
+      "loss": 0.3093,
+      "step": 9293
+    },
+    {
+      "epoch": 2.3375251509054324,
+      "grad_norm": 0.30700913071632385,
+      "learning_rate": 1.413367678070987e-06,
+      "loss": 0.322,
+      "step": 9294
+    },
+    {
+      "epoch": 2.3377766599597587,
+      "grad_norm": 0.31583860516548157,
+      "learning_rate": 1.41234833259931e-06,
+      "loss": 0.3232,
+      "step": 9295
+    },
+    {
+      "epoch": 2.3380281690140845,
+      "grad_norm": 0.31791380047798157,
+      "learning_rate": 1.4113292943875518e-06,
+      "loss": 0.3248,
+      "step": 9296
+    },
+    {
+      "epoch": 2.3382796780684103,
+      "grad_norm": 0.2840316593647003,
+      "learning_rate": 1.4103105635229907e-06,
+      "loss": 0.311,
+      "step": 9297
+    },
+    {
+      "epoch": 2.3385311871227366,
+      "grad_norm": 0.31344839930534363,
+      "learning_rate": 1.4092921400928717e-06,
+      "loss": 0.3031,
+      "step": 9298
+    },
+    {
+      "epoch": 2.3387826961770624,
+      "grad_norm": 0.29697540402412415,
+      "learning_rate": 1.4082740241844185e-06,
+      "loss": 0.3319,
+      "step": 9299
+    },
+    {
+      "epoch": 2.339034205231388,
+      "grad_norm": 0.30100035667419434,
+      "learning_rate": 1.4072562158848241e-06,
+      "loss": 0.2954,
+      "step": 9300
+    },
+    {
+      "epoch": 2.3392857142857144,
+      "grad_norm": 0.31279945373535156,
+      "learning_rate": 1.4062387152812595e-06,
+      "loss": 0.3215,
+      "step": 9301
+    },
+    {
+      "epoch": 2.33953722334004,
+      "grad_norm": 0.29303503036499023,
+      "learning_rate": 1.4052215224608656e-06,
+      "loss": 0.3127,
+      "step": 9302
+    },
+    {
+      "epoch": 2.339788732394366,
+      "grad_norm": 0.3038380444049835,
+      "learning_rate": 1.4042046375107592e-06,
+      "loss": 0.3176,
+      "step": 9303
+    },
+    {
+      "epoch": 2.3400402414486923,
+      "grad_norm": 0.3016948103904724,
+      "learning_rate": 1.4031880605180325e-06,
+      "loss": 0.3134,
+      "step": 9304
+    },
+    {
+      "epoch": 2.340291750503018,
+      "grad_norm": 0.3166118264198303,
+      "learning_rate": 1.4021717915697448e-06,
+      "loss": 0.3081,
+      "step": 9305
+    },
+    {
+      "epoch": 2.340543259557344,
+      "grad_norm": 0.31366658210754395,
+      "learning_rate": 1.4011558307529366e-06,
+      "loss": 0.3166,
+      "step": 9306
+    },
+    {
+      "epoch": 2.34079476861167,
+      "grad_norm": 0.3000653088092804,
+      "learning_rate": 1.400140178154616e-06,
+      "loss": 0.3153,
+      "step": 9307
+    },
+    {
+      "epoch": 2.341046277665996,
+      "grad_norm": 0.2916873097419739,
+      "learning_rate": 1.3991248338617697e-06,
+      "loss": 0.3262,
+      "step": 9308
+    },
+    {
+      "epoch": 2.3412977867203217,
+      "grad_norm": 0.29192009568214417,
+      "learning_rate": 1.3981097979613528e-06,
+      "loss": 0.323,
+      "step": 9309
+    },
+    {
+      "epoch": 2.341549295774648,
+      "grad_norm": 0.31591734290122986,
+      "learning_rate": 1.3970950705403003e-06,
+      "loss": 0.3382,
+      "step": 9310
+    },
+    {
+      "epoch": 2.341800804828974,
+      "grad_norm": 0.32917144894599915,
+      "learning_rate": 1.3960806516855136e-06,
+      "loss": 0.3127,
+      "step": 9311
+    },
+    {
+      "epoch": 2.3420523138832996,
+      "grad_norm": 0.29949262738227844,
+      "learning_rate": 1.3950665414838744e-06,
+      "loss": 0.3415,
+      "step": 9312
+    },
+    {
+      "epoch": 2.342303822937626,
+      "grad_norm": 0.31109297275543213,
+      "learning_rate": 1.394052740022232e-06,
+      "loss": 0.3125,
+      "step": 9313
+    },
+    {
+      "epoch": 2.3425553319919517,
+      "grad_norm": 0.31198766827583313,
+      "learning_rate": 1.3930392473874138e-06,
+      "loss": 0.3121,
+      "step": 9314
+    },
+    {
+      "epoch": 2.3428068410462775,
+      "grad_norm": 0.30766913294792175,
+      "learning_rate": 1.3920260636662208e-06,
+      "loss": 0.3147,
+      "step": 9315
+    },
+    {
+      "epoch": 2.3430583501006037,
+      "grad_norm": 0.3334919512271881,
+      "learning_rate": 1.391013188945422e-06,
+      "loss": 0.3141,
+      "step": 9316
+    },
+    {
+      "epoch": 2.3433098591549295,
+      "grad_norm": 0.3124428391456604,
+      "learning_rate": 1.3900006233117675e-06,
+      "loss": 0.341,
+      "step": 9317
+    },
+    {
+      "epoch": 2.3435613682092553,
+      "grad_norm": 0.29679644107818604,
+      "learning_rate": 1.3889883668519744e-06,
+      "loss": 0.304,
+      "step": 9318
+    },
+    {
+      "epoch": 2.3438128772635816,
+      "grad_norm": 0.2995522916316986,
+      "learning_rate": 1.3879764196527384e-06,
+      "loss": 0.321,
+      "step": 9319
+    },
+    {
+      "epoch": 2.3440643863179074,
+      "grad_norm": 0.31305447220802307,
+      "learning_rate": 1.3869647818007236e-06,
+      "loss": 0.2999,
+      "step": 9320
+    },
+    {
+      "epoch": 2.344315895372233,
+      "grad_norm": 0.2792012095451355,
+      "learning_rate": 1.385953453382574e-06,
+      "loss": 0.3074,
+      "step": 9321
+    },
+    {
+      "epoch": 2.3445674044265594,
+      "grad_norm": 0.29924044013023376,
+      "learning_rate": 1.3849424344849e-06,
+      "loss": 0.2991,
+      "step": 9322
+    },
+    {
+      "epoch": 2.3448189134808852,
+      "grad_norm": 0.30037921667099,
+      "learning_rate": 1.3839317251942907e-06,
+      "loss": 0.3035,
+      "step": 9323
+    },
+    {
+      "epoch": 2.345070422535211,
+      "grad_norm": 0.3079024851322174,
+      "learning_rate": 1.3829213255973089e-06,
+      "loss": 0.321,
+      "step": 9324
+    },
+    {
+      "epoch": 2.3453219315895373,
+      "grad_norm": 0.2921197712421417,
+      "learning_rate": 1.3819112357804859e-06,
+      "loss": 0.3223,
+      "step": 9325
+    },
+    {
+      "epoch": 2.345573440643863,
+      "grad_norm": 0.28539982438087463,
+      "learning_rate": 1.3809014558303319e-06,
+      "loss": 0.321,
+      "step": 9326
+    },
+    {
+      "epoch": 2.345824949698189,
+      "grad_norm": 0.2913080155849457,
+      "learning_rate": 1.3798919858333254e-06,
+      "loss": 0.3126,
+      "step": 9327
+    },
+    {
+      "epoch": 2.346076458752515,
+      "grad_norm": 0.31199610233306885,
+      "learning_rate": 1.3788828258759251e-06,
+      "loss": 0.3611,
+      "step": 9328
+    },
+    {
+      "epoch": 2.346327967806841,
+      "grad_norm": 0.301520973443985,
+      "learning_rate": 1.3778739760445552e-06,
+      "loss": 0.3339,
+      "step": 9329
+    },
+    {
+      "epoch": 2.346579476861167,
+      "grad_norm": 0.29440373182296753,
+      "learning_rate": 1.3768654364256212e-06,
+      "loss": 0.3115,
+      "step": 9330
+    },
+    {
+      "epoch": 2.346830985915493,
+      "grad_norm": 0.29134684801101685,
+      "learning_rate": 1.375857207105495e-06,
+      "loss": 0.3267,
+      "step": 9331
+    },
+    {
+      "epoch": 2.347082494969819,
+      "grad_norm": 0.2973969280719757,
+      "learning_rate": 1.3748492881705272e-06,
+      "loss": 0.3272,
+      "step": 9332
+    },
+    {
+      "epoch": 2.347334004024145,
+      "grad_norm": 0.30077022314071655,
+      "learning_rate": 1.373841679707038e-06,
+      "loss": 0.3198,
+      "step": 9333
+    },
+    {
+      "epoch": 2.347585513078471,
+      "grad_norm": 0.27227941155433655,
+      "learning_rate": 1.3728343818013233e-06,
+      "loss": 0.282,
+      "step": 9334
+    },
+    {
+      "epoch": 2.3478370221327967,
+      "grad_norm": 0.30526816844940186,
+      "learning_rate": 1.3718273945396542e-06,
+      "loss": 0.3189,
+      "step": 9335
+    },
+    {
+      "epoch": 2.348088531187123,
+      "grad_norm": 0.27349480986595154,
+      "learning_rate": 1.3708207180082694e-06,
+      "loss": 0.3208,
+      "step": 9336
+    },
+    {
+      "epoch": 2.3483400402414487,
+      "grad_norm": 0.30724623799324036,
+      "learning_rate": 1.3698143522933876e-06,
+      "loss": 0.3169,
+      "step": 9337
+    },
+    {
+      "epoch": 2.3485915492957745,
+      "grad_norm": 0.2816196084022522,
+      "learning_rate": 1.368808297481195e-06,
+      "loss": 0.3152,
+      "step": 9338
+    },
+    {
+      "epoch": 2.3488430583501008,
+      "grad_norm": 0.2888517379760742,
+      "learning_rate": 1.3678025536578559e-06,
+      "loss": 0.3093,
+      "step": 9339
+    },
+    {
+      "epoch": 2.3490945674044266,
+      "grad_norm": 0.2849167287349701,
+      "learning_rate": 1.3667971209095039e-06,
+      "loss": 0.3124,
+      "step": 9340
+    },
+    {
+      "epoch": 2.3493460764587524,
+      "grad_norm": 0.29156047105789185,
+      "learning_rate": 1.3657919993222507e-06,
+      "loss": 0.3193,
+      "step": 9341
+    },
+    {
+      "epoch": 2.3495975855130786,
+      "grad_norm": 0.30401670932769775,
+      "learning_rate": 1.3647871889821762e-06,
+      "loss": 0.3127,
+      "step": 9342
+    },
+    {
+      "epoch": 2.3498490945674044,
+      "grad_norm": 0.2958643436431885,
+      "learning_rate": 1.363782689975338e-06,
+      "loss": 0.3034,
+      "step": 9343
+    },
+    {
+      "epoch": 2.3501006036217302,
+      "grad_norm": 0.2841004729270935,
+      "learning_rate": 1.3627785023877633e-06,
+      "loss": 0.3383,
+      "step": 9344
+    },
+    {
+      "epoch": 2.3503521126760565,
+      "grad_norm": 0.2793041169643402,
+      "learning_rate": 1.3617746263054548e-06,
+      "loss": 0.3506,
+      "step": 9345
+    },
+    {
+      "epoch": 2.3506036217303823,
+      "grad_norm": 0.29398274421691895,
+      "learning_rate": 1.360771061814391e-06,
+      "loss": 0.3331,
+      "step": 9346
+    },
+    {
+      "epoch": 2.350855130784708,
+      "grad_norm": 0.3107350468635559,
+      "learning_rate": 1.3597678090005168e-06,
+      "loss": 0.3286,
+      "step": 9347
+    },
+    {
+      "epoch": 2.3511066398390343,
+      "grad_norm": 0.2840394973754883,
+      "learning_rate": 1.3587648679497583e-06,
+      "loss": 0.3182,
+      "step": 9348
+    },
+    {
+      "epoch": 2.35135814889336,
+      "grad_norm": 0.3232376277446747,
+      "learning_rate": 1.3577622387480082e-06,
+      "loss": 0.3361,
+      "step": 9349
+    },
+    {
+      "epoch": 2.351609657947686,
+      "grad_norm": 0.30280598998069763,
+      "learning_rate": 1.3567599214811379e-06,
+      "loss": 0.3451,
+      "step": 9350
+    },
+    {
+      "epoch": 2.351861167002012,
+      "grad_norm": 0.2855769693851471,
+      "learning_rate": 1.3557579162349864e-06,
+      "loss": 0.2765,
+      "step": 9351
+    },
+    {
+      "epoch": 2.352112676056338,
+      "grad_norm": 0.30225083231925964,
+      "learning_rate": 1.3547562230953726e-06,
+      "loss": 0.3393,
+      "step": 9352
+    },
+    {
+      "epoch": 2.352364185110664,
+      "grad_norm": 0.29712530970573425,
+      "learning_rate": 1.353754842148083e-06,
+      "loss": 0.316,
+      "step": 9353
+    },
+    {
+      "epoch": 2.35261569416499,
+      "grad_norm": 0.3048205077648163,
+      "learning_rate": 1.352753773478881e-06,
+      "loss": 0.3233,
+      "step": 9354
+    },
+    {
+      "epoch": 2.352867203219316,
+      "grad_norm": 0.3305099308490753,
+      "learning_rate": 1.3517530171735e-06,
+      "loss": 0.3363,
+      "step": 9355
+    },
+    {
+      "epoch": 2.3531187122736417,
+      "grad_norm": 0.27165278792381287,
+      "learning_rate": 1.3507525733176497e-06,
+      "loss": 0.331,
+      "step": 9356
+    },
+    {
+      "epoch": 2.353370221327968,
+      "grad_norm": 0.3076144754886627,
+      "learning_rate": 1.3497524419970132e-06,
+      "loss": 0.3293,
+      "step": 9357
+    },
+    {
+      "epoch": 2.3536217303822937,
+      "grad_norm": 0.31551241874694824,
+      "learning_rate": 1.348752623297243e-06,
+      "loss": 0.3077,
+      "step": 9358
+    },
+    {
+      "epoch": 2.3538732394366195,
+      "grad_norm": 0.3101344108581543,
+      "learning_rate": 1.3477531173039697e-06,
+      "loss": 0.3015,
+      "step": 9359
+    },
+    {
+      "epoch": 2.3541247484909458,
+      "grad_norm": 0.3111810088157654,
+      "learning_rate": 1.3467539241027922e-06,
+      "loss": 0.3292,
+      "step": 9360
+    },
+    {
+      "epoch": 2.3543762575452716,
+      "grad_norm": 0.32074761390686035,
+      "learning_rate": 1.3457550437792876e-06,
+      "loss": 0.3281,
+      "step": 9361
+    },
+    {
+      "epoch": 2.354627766599598,
+      "grad_norm": 0.2921229600906372,
+      "learning_rate": 1.344756476419002e-06,
+      "loss": 0.3177,
+      "step": 9362
+    },
+    {
+      "epoch": 2.3548792756539236,
+      "grad_norm": 0.28211191296577454,
+      "learning_rate": 1.3437582221074574e-06,
+      "loss": 0.3583,
+      "step": 9363
+    },
+    {
+      "epoch": 2.3551307847082494,
+      "grad_norm": 0.31738191843032837,
+      "learning_rate": 1.342760280930147e-06,
+      "loss": 0.3441,
+      "step": 9364
+    },
+    {
+      "epoch": 2.3553822937625757,
+      "grad_norm": 0.28729522228240967,
+      "learning_rate": 1.3417626529725402e-06,
+      "loss": 0.3326,
+      "step": 9365
+    },
+    {
+      "epoch": 2.3556338028169015,
+      "grad_norm": 0.3042784333229065,
+      "learning_rate": 1.3407653383200747e-06,
+      "loss": 0.3347,
+      "step": 9366
+    },
+    {
+      "epoch": 2.3558853118712273,
+      "grad_norm": 0.2783034145832062,
+      "learning_rate": 1.339768337058166e-06,
+      "loss": 0.2803,
+      "step": 9367
+    },
+    {
+      "epoch": 2.3561368209255535,
+      "grad_norm": 0.2911806106567383,
+      "learning_rate": 1.3387716492722025e-06,
+      "loss": 0.303,
+      "step": 9368
+    },
+    {
+      "epoch": 2.3563883299798793,
+      "grad_norm": 0.2815414071083069,
+      "learning_rate": 1.337775275047541e-06,
+      "loss": 0.3222,
+      "step": 9369
+    },
+    {
+      "epoch": 2.356639839034205,
+      "grad_norm": 0.2898373007774353,
+      "learning_rate": 1.336779214469518e-06,
+      "loss": 0.3224,
+      "step": 9370
+    },
+    {
+      "epoch": 2.3568913480885314,
+      "grad_norm": 0.3051788806915283,
+      "learning_rate": 1.3357834676234366e-06,
+      "loss": 0.3103,
+      "step": 9371
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.31733277440071106,
+      "learning_rate": 1.3347880345945796e-06,
+      "loss": 0.3446,
+      "step": 9372
+    },
+    {
+      "epoch": 2.357394366197183,
+      "grad_norm": 0.30566954612731934,
+      "learning_rate": 1.333792915468196e-06,
+      "loss": 0.3109,
+      "step": 9373
+    },
+    {
+      "epoch": 2.3576458752515093,
+      "grad_norm": 0.2991909682750702,
+      "learning_rate": 1.3327981103295156e-06,
+      "loss": 0.3123,
+      "step": 9374
+    },
+    {
+      "epoch": 2.357897384305835,
+      "grad_norm": 0.2768964469432831,
+      "learning_rate": 1.3318036192637334e-06,
+      "loss": 0.3224,
+      "step": 9375
+    },
+    {
+      "epoch": 2.358148893360161,
+      "grad_norm": 0.2982604205608368,
+      "learning_rate": 1.3308094423560242e-06,
+      "loss": 0.3151,
+      "step": 9376
+    },
+    {
+      "epoch": 2.358400402414487,
+      "grad_norm": 0.31598472595214844,
+      "learning_rate": 1.3298155796915307e-06,
+      "loss": 0.3188,
+      "step": 9377
+    },
+    {
+      "epoch": 2.358651911468813,
+      "grad_norm": 0.29461905360221863,
+      "learning_rate": 1.3288220313553723e-06,
+      "loss": 0.3171,
+      "step": 9378
+    },
+    {
+      "epoch": 2.3589034205231387,
+      "grad_norm": 0.2900139391422272,
+      "learning_rate": 1.3278287974326415e-06,
+      "loss": 0.3277,
+      "step": 9379
+    },
+    {
+      "epoch": 2.359154929577465,
+      "grad_norm": 0.2897052764892578,
+      "learning_rate": 1.3268358780083995e-06,
+      "loss": 0.3078,
+      "step": 9380
+    },
+    {
+      "epoch": 2.359406438631791,
+      "grad_norm": 0.2946465015411377,
+      "learning_rate": 1.3258432731676867e-06,
+      "loss": 0.3077,
+      "step": 9381
+    },
+    {
+      "epoch": 2.3596579476861166,
+      "grad_norm": 0.29683101177215576,
+      "learning_rate": 1.324850982995511e-06,
+      "loss": 0.3171,
+      "step": 9382
+    },
+    {
+      "epoch": 2.359909456740443,
+      "grad_norm": 0.32809895277023315,
+      "learning_rate": 1.3238590075768582e-06,
+      "loss": 0.3467,
+      "step": 9383
+    },
+    {
+      "epoch": 2.3601609657947686,
+      "grad_norm": 0.28296294808387756,
+      "learning_rate": 1.3228673469966819e-06,
+      "loss": 0.3348,
+      "step": 9384
+    },
+    {
+      "epoch": 2.3604124748490944,
+      "grad_norm": 0.31060582399368286,
+      "learning_rate": 1.321876001339915e-06,
+      "loss": 0.317,
+      "step": 9385
+    },
+    {
+      "epoch": 2.3606639839034207,
+      "grad_norm": 0.2970585525035858,
+      "learning_rate": 1.3208849706914567e-06,
+      "loss": 0.3378,
+      "step": 9386
+    },
+    {
+      "epoch": 2.3609154929577465,
+      "grad_norm": 0.310907781124115,
+      "learning_rate": 1.319894255136186e-06,
+      "loss": 0.3151,
+      "step": 9387
+    },
+    {
+      "epoch": 2.3611670020120723,
+      "grad_norm": 0.3074939250946045,
+      "learning_rate": 1.3189038547589479e-06,
+      "loss": 0.3049,
+      "step": 9388
+    },
+    {
+      "epoch": 2.3614185110663986,
+      "grad_norm": 0.2910417318344116,
+      "learning_rate": 1.317913769644567e-06,
+      "loss": 0.325,
+      "step": 9389
+    },
+    {
+      "epoch": 2.3616700201207244,
+      "grad_norm": 0.29934588074684143,
+      "learning_rate": 1.3169239998778361e-06,
+      "loss": 0.3217,
+      "step": 9390
+    },
+    {
+      "epoch": 2.36192152917505,
+      "grad_norm": 0.3142791986465454,
+      "learning_rate": 1.3159345455435241e-06,
+      "loss": 0.3297,
+      "step": 9391
+    },
+    {
+      "epoch": 2.3621730382293764,
+      "grad_norm": 0.30525627732276917,
+      "learning_rate": 1.3149454067263696e-06,
+      "loss": 0.3088,
+      "step": 9392
+    },
+    {
+      "epoch": 2.362424547283702,
+      "grad_norm": 0.2881489396095276,
+      "learning_rate": 1.3139565835110884e-06,
+      "loss": 0.308,
+      "step": 9393
+    },
+    {
+      "epoch": 2.362676056338028,
+      "grad_norm": 0.32259997725486755,
+      "learning_rate": 1.3129680759823648e-06,
+      "loss": 0.3287,
+      "step": 9394
+    },
+    {
+      "epoch": 2.3629275653923543,
+      "grad_norm": 0.30433225631713867,
+      "learning_rate": 1.311979884224861e-06,
+      "loss": 0.3147,
+      "step": 9395
+    },
+    {
+      "epoch": 2.36317907444668,
+      "grad_norm": 0.3119969964027405,
+      "learning_rate": 1.3109920083232064e-06,
+      "loss": 0.3184,
+      "step": 9396
+    },
+    {
+      "epoch": 2.363430583501006,
+      "grad_norm": 0.2946832776069641,
+      "learning_rate": 1.3100044483620094e-06,
+      "loss": 0.318,
+      "step": 9397
+    },
+    {
+      "epoch": 2.363682092555332,
+      "grad_norm": 0.282425194978714,
+      "learning_rate": 1.3090172044258458e-06,
+      "loss": 0.3321,
+      "step": 9398
+    },
+    {
+      "epoch": 2.363933601609658,
+      "grad_norm": 0.3065638244152069,
+      "learning_rate": 1.3080302765992692e-06,
+      "loss": 0.3006,
+      "step": 9399
+    },
+    {
+      "epoch": 2.3641851106639837,
+      "grad_norm": 0.2910194993019104,
+      "learning_rate": 1.3070436649668006e-06,
+      "loss": 0.3033,
+      "step": 9400
+    },
+    {
+      "epoch": 2.36443661971831,
+      "grad_norm": 0.306165874004364,
+      "learning_rate": 1.3060573696129396e-06,
+      "loss": 0.3083,
+      "step": 9401
+    },
+    {
+      "epoch": 2.364688128772636,
+      "grad_norm": 0.3192187249660492,
+      "learning_rate": 1.305071390622157e-06,
+      "loss": 0.302,
+      "step": 9402
+    },
+    {
+      "epoch": 2.3649396378269616,
+      "grad_norm": 0.3046712577342987,
+      "learning_rate": 1.3040857280788927e-06,
+      "loss": 0.302,
+      "step": 9403
+    },
+    {
+      "epoch": 2.365191146881288,
+      "grad_norm": 0.2873310446739197,
+      "learning_rate": 1.3031003820675659e-06,
+      "loss": 0.2988,
+      "step": 9404
+    },
+    {
+      "epoch": 2.3654426559356136,
+      "grad_norm": 0.31300392746925354,
+      "learning_rate": 1.3021153526725615e-06,
+      "loss": 0.317,
+      "step": 9405
+    },
+    {
+      "epoch": 2.3656941649899395,
+      "grad_norm": 0.29510292410850525,
+      "learning_rate": 1.3011306399782458e-06,
+      "loss": 0.2967,
+      "step": 9406
+    },
+    {
+      "epoch": 2.3659456740442657,
+      "grad_norm": 0.3187848627567291,
+      "learning_rate": 1.3001462440689488e-06,
+      "loss": 0.3349,
+      "step": 9407
+    },
+    {
+      "epoch": 2.3661971830985915,
+      "grad_norm": 0.2933845818042755,
+      "learning_rate": 1.2991621650289809e-06,
+      "loss": 0.2987,
+      "step": 9408
+    },
+    {
+      "epoch": 2.3664486921529173,
+      "grad_norm": 0.2807075083255768,
+      "learning_rate": 1.2981784029426203e-06,
+      "loss": 0.2975,
+      "step": 9409
+    },
+    {
+      "epoch": 2.3667002012072436,
+      "grad_norm": 0.2867673635482788,
+      "learning_rate": 1.2971949578941217e-06,
+      "loss": 0.3161,
+      "step": 9410
+    },
+    {
+      "epoch": 2.3669517102615694,
+      "grad_norm": 0.2991779148578644,
+      "learning_rate": 1.2962118299677095e-06,
+      "loss": 0.3163,
+      "step": 9411
+    },
+    {
+      "epoch": 2.367203219315895,
+      "grad_norm": 0.3035879135131836,
+      "learning_rate": 1.2952290192475848e-06,
+      "loss": 0.29,
+      "step": 9412
+    },
+    {
+      "epoch": 2.3674547283702214,
+      "grad_norm": 0.3021244704723358,
+      "learning_rate": 1.2942465258179155e-06,
+      "loss": 0.3111,
+      "step": 9413
+    },
+    {
+      "epoch": 2.3677062374245472,
+      "grad_norm": 0.2943519949913025,
+      "learning_rate": 1.29326434976285e-06,
+      "loss": 0.3085,
+      "step": 9414
+    },
+    {
+      "epoch": 2.367957746478873,
+      "grad_norm": 0.29150792956352234,
+      "learning_rate": 1.2922824911665021e-06,
+      "loss": 0.3083,
+      "step": 9415
+    },
+    {
+      "epoch": 2.3682092555331993,
+      "grad_norm": 0.3039129972457886,
+      "learning_rate": 1.2913009501129653e-06,
+      "loss": 0.3093,
+      "step": 9416
+    },
+    {
+      "epoch": 2.368460764587525,
+      "grad_norm": 0.28910478949546814,
+      "learning_rate": 1.2903197266862989e-06,
+      "loss": 0.3038,
+      "step": 9417
+    },
+    {
+      "epoch": 2.368712273641851,
+      "grad_norm": 0.29740893840789795,
+      "learning_rate": 1.289338820970541e-06,
+      "loss": 0.3274,
+      "step": 9418
+    },
+    {
+      "epoch": 2.368963782696177,
+      "grad_norm": 0.2945566773414612,
+      "learning_rate": 1.2883582330496986e-06,
+      "loss": 0.3546,
+      "step": 9419
+    },
+    {
+      "epoch": 2.369215291750503,
+      "grad_norm": 0.2992318272590637,
+      "learning_rate": 1.287377963007755e-06,
+      "loss": 0.3222,
+      "step": 9420
+    },
+    {
+      "epoch": 2.3694668008048287,
+      "grad_norm": 0.2896125614643097,
+      "learning_rate": 1.2863980109286605e-06,
+      "loss": 0.3324,
+      "step": 9421
+    },
+    {
+      "epoch": 2.369718309859155,
+      "grad_norm": 0.31392091512680054,
+      "learning_rate": 1.2854183768963453e-06,
+      "loss": 0.318,
+      "step": 9422
+    },
+    {
+      "epoch": 2.369969818913481,
+      "grad_norm": 0.2963343560695648,
+      "learning_rate": 1.2844390609947082e-06,
+      "loss": 0.3077,
+      "step": 9423
+    },
+    {
+      "epoch": 2.3702213279678066,
+      "grad_norm": 0.32114240527153015,
+      "learning_rate": 1.2834600633076205e-06,
+      "loss": 0.332,
+      "step": 9424
+    },
+    {
+      "epoch": 2.370472837022133,
+      "grad_norm": 0.28790682554244995,
+      "learning_rate": 1.2824813839189288e-06,
+      "loss": 0.3242,
+      "step": 9425
+    },
+    {
+      "epoch": 2.3707243460764587,
+      "grad_norm": 0.31781139969825745,
+      "learning_rate": 1.2815030229124481e-06,
+      "loss": 0.3148,
+      "step": 9426
+    },
+    {
+      "epoch": 2.3709758551307845,
+      "grad_norm": 0.3021380305290222,
+      "learning_rate": 1.2805249803719722e-06,
+      "loss": 0.3122,
+      "step": 9427
+    },
+    {
+      "epoch": 2.3712273641851107,
+      "grad_norm": 0.3446398973464966,
+      "learning_rate": 1.2795472563812617e-06,
+      "loss": 0.3437,
+      "step": 9428
+    },
+    {
+      "epoch": 2.3714788732394365,
+      "grad_norm": 0.2881867587566376,
+      "learning_rate": 1.2785698510240546e-06,
+      "loss": 0.3348,
+      "step": 9429
+    },
+    {
+      "epoch": 2.3717303822937628,
+      "grad_norm": 0.28368863463401794,
+      "learning_rate": 1.2775927643840575e-06,
+      "loss": 0.3034,
+      "step": 9430
+    },
+    {
+      "epoch": 2.3719818913480886,
+      "grad_norm": 0.281934529542923,
+      "learning_rate": 1.2766159965449543e-06,
+      "loss": 0.3107,
+      "step": 9431
+    },
+    {
+      "epoch": 2.3722334004024144,
+      "grad_norm": 0.2791815996170044,
+      "learning_rate": 1.275639547590396e-06,
+      "loss": 0.3286,
+      "step": 9432
+    },
+    {
+      "epoch": 2.3724849094567406,
+      "grad_norm": 0.29985862970352173,
+      "learning_rate": 1.2746634176040107e-06,
+      "loss": 0.3342,
+      "step": 9433
+    },
+    {
+      "epoch": 2.3727364185110664,
+      "grad_norm": 0.31482386589050293,
+      "learning_rate": 1.2736876066693999e-06,
+      "loss": 0.3313,
+      "step": 9434
+    },
+    {
+      "epoch": 2.3729879275653922,
+      "grad_norm": 0.2999832332134247,
+      "learning_rate": 1.2727121148701322e-06,
+      "loss": 0.3334,
+      "step": 9435
+    },
+    {
+      "epoch": 2.3732394366197185,
+      "grad_norm": 0.3035413920879364,
+      "learning_rate": 1.2717369422897552e-06,
+      "loss": 0.3196,
+      "step": 9436
+    },
+    {
+      "epoch": 2.3734909456740443,
+      "grad_norm": 0.28583768010139465,
+      "learning_rate": 1.270762089011784e-06,
+      "loss": 0.3175,
+      "step": 9437
+    },
+    {
+      "epoch": 2.37374245472837,
+      "grad_norm": 0.2991746962070465,
+      "learning_rate": 1.2697875551197113e-06,
+      "loss": 0.3391,
+      "step": 9438
+    },
+    {
+      "epoch": 2.3739939637826963,
+      "grad_norm": 0.29370561242103577,
+      "learning_rate": 1.268813340696997e-06,
+      "loss": 0.3131,
+      "step": 9439
+    },
+    {
+      "epoch": 2.374245472837022,
+      "grad_norm": 0.31026825308799744,
+      "learning_rate": 1.2678394458270794e-06,
+      "loss": 0.3207,
+      "step": 9440
+    },
+    {
+      "epoch": 2.374496981891348,
+      "grad_norm": 0.31360548734664917,
+      "learning_rate": 1.2668658705933628e-06,
+      "loss": 0.314,
+      "step": 9441
+    },
+    {
+      "epoch": 2.374748490945674,
+      "grad_norm": 0.31127217411994934,
+      "learning_rate": 1.2658926150792321e-06,
+      "loss": 0.3276,
+      "step": 9442
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 0.31766191124916077,
+      "learning_rate": 1.2649196793680367e-06,
+      "loss": 0.328,
+      "step": 9443
+    },
+    {
+      "epoch": 2.375251509054326,
+      "grad_norm": 0.32087278366088867,
+      "learning_rate": 1.2639470635431044e-06,
+      "loss": 0.3228,
+      "step": 9444
+    },
+    {
+      "epoch": 2.375503018108652,
+      "grad_norm": 0.29224544763565063,
+      "learning_rate": 1.2629747676877347e-06,
+      "loss": 0.3348,
+      "step": 9445
+    },
+    {
+      "epoch": 2.375754527162978,
+      "grad_norm": 0.3009779155254364,
+      "learning_rate": 1.2620027918851956e-06,
+      "loss": 0.3244,
+      "step": 9446
+    },
+    {
+      "epoch": 2.3760060362173037,
+      "grad_norm": 0.304565966129303,
+      "learning_rate": 1.2610311362187343e-06,
+      "loss": 0.3357,
+      "step": 9447
+    },
+    {
+      "epoch": 2.37625754527163,
+      "grad_norm": 0.3026162087917328,
+      "learning_rate": 1.260059800771564e-06,
+      "loss": 0.3098,
+      "step": 9448
+    },
+    {
+      "epoch": 2.3765090543259557,
+      "grad_norm": 0.3093135952949524,
+      "learning_rate": 1.2590887856268764e-06,
+      "loss": 0.3289,
+      "step": 9449
+    },
+    {
+      "epoch": 2.3767605633802815,
+      "grad_norm": 0.2858860194683075,
+      "learning_rate": 1.2581180908678291e-06,
+      "loss": 0.3005,
+      "step": 9450
+    },
+    {
+      "epoch": 2.3770120724346078,
+      "grad_norm": 0.28006115555763245,
+      "learning_rate": 1.2571477165775597e-06,
+      "loss": 0.3127,
+      "step": 9451
+    },
+    {
+      "epoch": 2.3772635814889336,
+      "grad_norm": 0.2762000560760498,
+      "learning_rate": 1.2561776628391725e-06,
+      "loss": 0.3349,
+      "step": 9452
+    },
+    {
+      "epoch": 2.3775150905432594,
+      "grad_norm": 0.3034118413925171,
+      "learning_rate": 1.2552079297357478e-06,
+      "loss": 0.318,
+      "step": 9453
+    },
+    {
+      "epoch": 2.3777665995975856,
+      "grad_norm": 0.29778799414634705,
+      "learning_rate": 1.2542385173503359e-06,
+      "loss": 0.3035,
+      "step": 9454
+    },
+    {
+      "epoch": 2.3780181086519114,
+      "grad_norm": 0.30300748348236084,
+      "learning_rate": 1.2532694257659605e-06,
+      "loss": 0.3182,
+      "step": 9455
+    },
+    {
+      "epoch": 2.3782696177062372,
+      "grad_norm": 0.27301889657974243,
+      "learning_rate": 1.252300655065622e-06,
+      "loss": 0.3322,
+      "step": 9456
+    },
+    {
+      "epoch": 2.3785211267605635,
+      "grad_norm": 0.3212936222553253,
+      "learning_rate": 1.2513322053322847e-06,
+      "loss": 0.3413,
+      "step": 9457
+    },
+    {
+      "epoch": 2.3787726358148893,
+      "grad_norm": 0.3001827597618103,
+      "learning_rate": 1.250364076648894e-06,
+      "loss": 0.3137,
+      "step": 9458
+    },
+    {
+      "epoch": 2.3790241448692155,
+      "grad_norm": 0.29002684354782104,
+      "learning_rate": 1.2493962690983608e-06,
+      "loss": 0.296,
+      "step": 9459
+    },
+    {
+      "epoch": 2.3792756539235413,
+      "grad_norm": 0.30628740787506104,
+      "learning_rate": 1.248428782763575e-06,
+      "loss": 0.3288,
+      "step": 9460
+    },
+    {
+      "epoch": 2.379527162977867,
+      "grad_norm": 0.30561593174934387,
+      "learning_rate": 1.2474616177273928e-06,
+      "loss": 0.3324,
+      "step": 9461
+    },
+    {
+      "epoch": 2.3797786720321934,
+      "grad_norm": 0.3026500642299652,
+      "learning_rate": 1.2464947740726491e-06,
+      "loss": 0.3171,
+      "step": 9462
+    },
+    {
+      "epoch": 2.380030181086519,
+      "grad_norm": 0.2937304377555847,
+      "learning_rate": 1.2455282518821442e-06,
+      "loss": 0.3395,
+      "step": 9463
+    },
+    {
+      "epoch": 2.380281690140845,
+      "grad_norm": 0.30851301550865173,
+      "learning_rate": 1.244562051238658e-06,
+      "loss": 0.3576,
+      "step": 9464
+    },
+    {
+      "epoch": 2.3805331991951713,
+      "grad_norm": 0.3095259666442871,
+      "learning_rate": 1.2435961722249374e-06,
+      "loss": 0.332,
+      "step": 9465
+    },
+    {
+      "epoch": 2.380784708249497,
+      "grad_norm": 0.2995477616786957,
+      "learning_rate": 1.2426306149237039e-06,
+      "loss": 0.3151,
+      "step": 9466
+    },
+    {
+      "epoch": 2.381036217303823,
+      "grad_norm": 0.2909550964832306,
+      "learning_rate": 1.2416653794176542e-06,
+      "loss": 0.3098,
+      "step": 9467
+    },
+    {
+      "epoch": 2.381287726358149,
+      "grad_norm": 0.28925126791000366,
+      "learning_rate": 1.2407004657894505e-06,
+      "loss": 0.3065,
+      "step": 9468
+    },
+    {
+      "epoch": 2.381539235412475,
+      "grad_norm": 0.28470999002456665,
+      "learning_rate": 1.2397358741217359e-06,
+      "loss": 0.3151,
+      "step": 9469
+    },
+    {
+      "epoch": 2.3817907444668007,
+      "grad_norm": 0.3109789192676544,
+      "learning_rate": 1.2387716044971181e-06,
+      "loss": 0.3137,
+      "step": 9470
+    },
+    {
+      "epoch": 2.382042253521127,
+      "grad_norm": 0.3258463740348816,
+      "learning_rate": 1.2378076569981833e-06,
+      "loss": 0.3246,
+      "step": 9471
+    },
+    {
+      "epoch": 2.3822937625754528,
+      "grad_norm": 0.3138425052165985,
+      "learning_rate": 1.2368440317074854e-06,
+      "loss": 0.3248,
+      "step": 9472
+    },
+    {
+      "epoch": 2.3825452716297786,
+      "grad_norm": 0.31474605202674866,
+      "learning_rate": 1.2358807287075553e-06,
+      "loss": 0.3273,
+      "step": 9473
+    },
+    {
+      "epoch": 2.382796780684105,
+      "grad_norm": 0.30715155601501465,
+      "learning_rate": 1.234917748080891e-06,
+      "loss": 0.3247,
+      "step": 9474
+    },
+    {
+      "epoch": 2.3830482897384306,
+      "grad_norm": 0.29947811365127563,
+      "learning_rate": 1.2339550899099673e-06,
+      "loss": 0.319,
+      "step": 9475
+    },
+    {
+      "epoch": 2.3832997987927564,
+      "grad_norm": 0.27288398146629333,
+      "learning_rate": 1.2329927542772314e-06,
+      "loss": 0.338,
+      "step": 9476
+    },
+    {
+      "epoch": 2.3835513078470827,
+      "grad_norm": 0.30378079414367676,
+      "learning_rate": 1.2320307412650978e-06,
+      "loss": 0.3123,
+      "step": 9477
+    },
+    {
+      "epoch": 2.3838028169014085,
+      "grad_norm": 0.2979941666126251,
+      "learning_rate": 1.2310690509559609e-06,
+      "loss": 0.3077,
+      "step": 9478
+    },
+    {
+      "epoch": 2.3840543259557343,
+      "grad_norm": 0.3131090998649597,
+      "learning_rate": 1.2301076834321796e-06,
+      "loss": 0.3274,
+      "step": 9479
+    },
+    {
+      "epoch": 2.3843058350100605,
+      "grad_norm": 0.2883933186531067,
+      "learning_rate": 1.2291466387760925e-06,
+      "loss": 0.3344,
+      "step": 9480
+    },
+    {
+      "epoch": 2.3845573440643864,
+      "grad_norm": 0.3163345158100128,
+      "learning_rate": 1.2281859170700039e-06,
+      "loss": 0.3107,
+      "step": 9481
+    },
+    {
+      "epoch": 2.384808853118712,
+      "grad_norm": 0.2901724874973297,
+      "learning_rate": 1.2272255183961968e-06,
+      "loss": 0.3282,
+      "step": 9482
+    },
+    {
+      "epoch": 2.3850603621730384,
+      "grad_norm": 0.31056809425354004,
+      "learning_rate": 1.2262654428369198e-06,
+      "loss": 0.3323,
+      "step": 9483
+    },
+    {
+      "epoch": 2.385311871227364,
+      "grad_norm": 0.299684077501297,
+      "learning_rate": 1.225305690474401e-06,
+      "loss": 0.3099,
+      "step": 9484
+    },
+    {
+      "epoch": 2.38556338028169,
+      "grad_norm": 0.31283077597618103,
+      "learning_rate": 1.2243462613908336e-06,
+      "loss": 0.3079,
+      "step": 9485
+    },
+    {
+      "epoch": 2.3858148893360163,
+      "grad_norm": 0.28932955861091614,
+      "learning_rate": 1.2233871556683891e-06,
+      "loss": 0.2907,
+      "step": 9486
+    },
+    {
+      "epoch": 2.386066398390342,
+      "grad_norm": 0.30000898241996765,
+      "learning_rate": 1.222428373389209e-06,
+      "loss": 0.3257,
+      "step": 9487
+    },
+    {
+      "epoch": 2.386317907444668,
+      "grad_norm": 0.3037915825843811,
+      "learning_rate": 1.2214699146354054e-06,
+      "loss": 0.3359,
+      "step": 9488
+    },
+    {
+      "epoch": 2.386569416498994,
+      "grad_norm": 0.30737778544425964,
+      "learning_rate": 1.2205117794890665e-06,
+      "loss": 0.3159,
+      "step": 9489
+    },
+    {
+      "epoch": 2.38682092555332,
+      "grad_norm": 0.318010538816452,
+      "learning_rate": 1.2195539680322476e-06,
+      "loss": 0.3197,
+      "step": 9490
+    },
+    {
+      "epoch": 2.3870724346076457,
+      "grad_norm": 0.32174649834632874,
+      "learning_rate": 1.2185964803469824e-06,
+      "loss": 0.3168,
+      "step": 9491
+    },
+    {
+      "epoch": 2.387323943661972,
+      "grad_norm": 0.28798073530197144,
+      "learning_rate": 1.2176393165152712e-06,
+      "loss": 0.3204,
+      "step": 9492
+    },
+    {
+      "epoch": 2.387575452716298,
+      "grad_norm": 0.3134564757347107,
+      "learning_rate": 1.2166824766190916e-06,
+      "loss": 0.3338,
+      "step": 9493
+    },
+    {
+      "epoch": 2.3878269617706236,
+      "grad_norm": 0.2944769859313965,
+      "learning_rate": 1.2157259607403877e-06,
+      "loss": 0.3366,
+      "step": 9494
+    },
+    {
+      "epoch": 2.38807847082495,
+      "grad_norm": 0.28981396555900574,
+      "learning_rate": 1.2147697689610826e-06,
+      "loss": 0.3254,
+      "step": 9495
+    },
+    {
+      "epoch": 2.3883299798792756,
+      "grad_norm": 0.28675341606140137,
+      "learning_rate": 1.213813901363065e-06,
+      "loss": 0.3123,
+      "step": 9496
+    },
+    {
+      "epoch": 2.3885814889336014,
+      "grad_norm": 0.29373446106910706,
+      "learning_rate": 1.2128583580282005e-06,
+      "loss": 0.3495,
+      "step": 9497
+    },
+    {
+      "epoch": 2.3888329979879277,
+      "grad_norm": 0.286771684885025,
+      "learning_rate": 1.2119031390383268e-06,
+      "loss": 0.3153,
+      "step": 9498
+    },
+    {
+      "epoch": 2.3890845070422535,
+      "grad_norm": 0.29528728127479553,
+      "learning_rate": 1.210948244475249e-06,
+      "loss": 0.3318,
+      "step": 9499
+    },
+    {
+      "epoch": 2.3893360160965793,
+      "grad_norm": 0.2970670759677887,
+      "learning_rate": 1.209993674420752e-06,
+      "loss": 0.3444,
+      "step": 9500
+    },
+    {
+      "epoch": 2.3895875251509056,
+      "grad_norm": 0.2939586639404297,
+      "learning_rate": 1.2090394289565849e-06,
+      "loss": 0.3132,
+      "step": 9501
+    },
+    {
+      "epoch": 2.3898390342052314,
+      "grad_norm": 0.2839318513870239,
+      "learning_rate": 1.208085508164476e-06,
+      "loss": 0.3193,
+      "step": 9502
+    },
+    {
+      "epoch": 2.390090543259557,
+      "grad_norm": 0.31266748905181885,
+      "learning_rate": 1.2071319121261194e-06,
+      "loss": 0.3341,
+      "step": 9503
+    },
+    {
+      "epoch": 2.3903420523138834,
+      "grad_norm": 0.29940277338027954,
+      "learning_rate": 1.2061786409231884e-06,
+      "loss": 0.3222,
+      "step": 9504
+    },
+    {
+      "epoch": 2.390593561368209,
+      "grad_norm": 0.2878960967063904,
+      "learning_rate": 1.2052256946373209e-06,
+      "loss": 0.3252,
+      "step": 9505
+    },
+    {
+      "epoch": 2.390845070422535,
+      "grad_norm": 0.3191094398498535,
+      "learning_rate": 1.204273073350134e-06,
+      "loss": 0.3082,
+      "step": 9506
+    },
+    {
+      "epoch": 2.3910965794768613,
+      "grad_norm": 0.3221496343612671,
+      "learning_rate": 1.203320777143211e-06,
+      "loss": 0.3168,
+      "step": 9507
+    },
+    {
+      "epoch": 2.391348088531187,
+      "grad_norm": 0.30812880396842957,
+      "learning_rate": 1.202368806098112e-06,
+      "loss": 0.3156,
+      "step": 9508
+    },
+    {
+      "epoch": 2.391599597585513,
+      "grad_norm": 0.28987714648246765,
+      "learning_rate": 1.2014171602963676e-06,
+      "loss": 0.3236,
+      "step": 9509
+    },
+    {
+      "epoch": 2.391851106639839,
+      "grad_norm": 0.2956389784812927,
+      "learning_rate": 1.2004658398194786e-06,
+      "loss": 0.3101,
+      "step": 9510
+    },
+    {
+      "epoch": 2.392102615694165,
+      "grad_norm": 0.2824747860431671,
+      "learning_rate": 1.199514844748922e-06,
+      "loss": 0.3145,
+      "step": 9511
+    },
+    {
+      "epoch": 2.3923541247484907,
+      "grad_norm": 0.3140904903411865,
+      "learning_rate": 1.1985641751661415e-06,
+      "loss": 0.3101,
+      "step": 9512
+    },
+    {
+      "epoch": 2.392605633802817,
+      "grad_norm": 0.288084477186203,
+      "learning_rate": 1.1976138311525592e-06,
+      "loss": 0.3343,
+      "step": 9513
+    },
+    {
+      "epoch": 2.392857142857143,
+      "grad_norm": 0.27887189388275146,
+      "learning_rate": 1.196663812789563e-06,
+      "loss": 0.3242,
+      "step": 9514
+    },
+    {
+      "epoch": 2.3931086519114686,
+      "grad_norm": 0.30040696263313293,
+      "learning_rate": 1.1957141201585193e-06,
+      "loss": 0.3303,
+      "step": 9515
+    },
+    {
+      "epoch": 2.393360160965795,
+      "grad_norm": 0.2744455635547638,
+      "learning_rate": 1.1947647533407602e-06,
+      "loss": 0.3199,
+      "step": 9516
+    },
+    {
+      "epoch": 2.3936116700201207,
+      "grad_norm": 0.3067541718482971,
+      "learning_rate": 1.1938157124175959e-06,
+      "loss": 0.3125,
+      "step": 9517
+    },
+    {
+      "epoch": 2.3938631790744465,
+      "grad_norm": 0.29640766978263855,
+      "learning_rate": 1.1928669974703033e-06,
+      "loss": 0.3321,
+      "step": 9518
+    },
+    {
+      "epoch": 2.3941146881287727,
+      "grad_norm": 0.30596280097961426,
+      "learning_rate": 1.191918608580136e-06,
+      "loss": 0.3078,
+      "step": 9519
+    },
+    {
+      "epoch": 2.3943661971830985,
+      "grad_norm": 0.27518826723098755,
+      "learning_rate": 1.1909705458283155e-06,
+      "loss": 0.3197,
+      "step": 9520
+    },
+    {
+      "epoch": 2.3946177062374243,
+      "grad_norm": 0.27755483984947205,
+      "learning_rate": 1.1900228092960398e-06,
+      "loss": 0.3256,
+      "step": 9521
+    },
+    {
+      "epoch": 2.3948692152917506,
+      "grad_norm": 0.30145710706710815,
+      "learning_rate": 1.1890753990644738e-06,
+      "loss": 0.3207,
+      "step": 9522
+    },
+    {
+      "epoch": 2.3951207243460764,
+      "grad_norm": 0.3164133131504059,
+      "learning_rate": 1.1881283152147606e-06,
+      "loss": 0.3314,
+      "step": 9523
+    },
+    {
+      "epoch": 2.395372233400402,
+      "grad_norm": 0.3069988787174225,
+      "learning_rate": 1.1871815578280083e-06,
+      "loss": 0.3108,
+      "step": 9524
+    },
+    {
+      "epoch": 2.3956237424547284,
+      "grad_norm": 0.3104003965854645,
+      "learning_rate": 1.186235126985304e-06,
+      "loss": 0.3224,
+      "step": 9525
+    },
+    {
+      "epoch": 2.3958752515090542,
+      "grad_norm": 0.2976137399673462,
+      "learning_rate": 1.185289022767701e-06,
+      "loss": 0.3175,
+      "step": 9526
+    },
+    {
+      "epoch": 2.3961267605633805,
+      "grad_norm": 0.3276063799858093,
+      "learning_rate": 1.1843432452562303e-06,
+      "loss": 0.3342,
+      "step": 9527
+    },
+    {
+      "epoch": 2.3963782696177063,
+      "grad_norm": 0.3153345584869385,
+      "learning_rate": 1.183397794531888e-06,
+      "loss": 0.3397,
+      "step": 9528
+    },
+    {
+      "epoch": 2.396629778672032,
+      "grad_norm": 0.3022550940513611,
+      "learning_rate": 1.18245267067565e-06,
+      "loss": 0.3021,
+      "step": 9529
+    },
+    {
+      "epoch": 2.3968812877263583,
+      "grad_norm": 0.3086456060409546,
+      "learning_rate": 1.1815078737684566e-06,
+      "loss": 0.3211,
+      "step": 9530
+    },
+    {
+      "epoch": 2.397132796780684,
+      "grad_norm": 0.3157150447368622,
+      "learning_rate": 1.1805634038912268e-06,
+      "loss": 0.3374,
+      "step": 9531
+    },
+    {
+      "epoch": 2.39738430583501,
+      "grad_norm": 0.2981681823730469,
+      "learning_rate": 1.1796192611248452e-06,
+      "loss": 0.3354,
+      "step": 9532
+    },
+    {
+      "epoch": 2.397635814889336,
+      "grad_norm": 0.2994699478149414,
+      "learning_rate": 1.1786754455501759e-06,
+      "loss": 0.3359,
+      "step": 9533
+    },
+    {
+      "epoch": 2.397887323943662,
+      "grad_norm": 0.3035818338394165,
+      "learning_rate": 1.1777319572480468e-06,
+      "loss": 0.3408,
+      "step": 9534
+    },
+    {
+      "epoch": 2.398138832997988,
+      "grad_norm": 0.2822018265724182,
+      "learning_rate": 1.1767887962992647e-06,
+      "loss": 0.3129,
+      "step": 9535
+    },
+    {
+      "epoch": 2.398390342052314,
+      "grad_norm": 0.2857167720794678,
+      "learning_rate": 1.1758459627846031e-06,
+      "loss": 0.3171,
+      "step": 9536
+    },
+    {
+      "epoch": 2.39864185110664,
+      "grad_norm": 0.2896609306335449,
+      "learning_rate": 1.1749034567848122e-06,
+      "loss": 0.3178,
+      "step": 9537
+    },
+    {
+      "epoch": 2.3988933601609657,
+      "grad_norm": 0.3051433563232422,
+      "learning_rate": 1.1739612783806092e-06,
+      "loss": 0.3435,
+      "step": 9538
+    },
+    {
+      "epoch": 2.399144869215292,
+      "grad_norm": 0.3050413727760315,
+      "learning_rate": 1.1730194276526885e-06,
+      "loss": 0.3223,
+      "step": 9539
+    },
+    {
+      "epoch": 2.3993963782696177,
+      "grad_norm": 0.30371037125587463,
+      "learning_rate": 1.1720779046817104e-06,
+      "loss": 0.3187,
+      "step": 9540
+    },
+    {
+      "epoch": 2.3996478873239435,
+      "grad_norm": 0.28884759545326233,
+      "learning_rate": 1.1711367095483134e-06,
+      "loss": 0.3246,
+      "step": 9541
+    },
+    {
+      "epoch": 2.3998993963782698,
+      "grad_norm": 0.3120008409023285,
+      "learning_rate": 1.1701958423331044e-06,
+      "loss": 0.3344,
+      "step": 9542
+    },
+    {
+      "epoch": 2.4001509054325956,
+      "grad_norm": 0.28870880603790283,
+      "learning_rate": 1.1692553031166616e-06,
+      "loss": 0.3225,
+      "step": 9543
+    },
+    {
+      "epoch": 2.4004024144869214,
+      "grad_norm": 0.28078019618988037,
+      "learning_rate": 1.1683150919795378e-06,
+      "loss": 0.2904,
+      "step": 9544
+    },
+    {
+      "epoch": 2.4006539235412476,
+      "grad_norm": 0.28186681866645813,
+      "learning_rate": 1.1673752090022544e-06,
+      "loss": 0.3233,
+      "step": 9545
+    },
+    {
+      "epoch": 2.4009054325955734,
+      "grad_norm": 0.2807130813598633,
+      "learning_rate": 1.1664356542653088e-06,
+      "loss": 0.3163,
+      "step": 9546
+    },
+    {
+      "epoch": 2.4011569416498992,
+      "grad_norm": 0.28840017318725586,
+      "learning_rate": 1.1654964278491653e-06,
+      "loss": 0.3297,
+      "step": 9547
+    },
+    {
+      "epoch": 2.4014084507042255,
+      "grad_norm": 0.2861834466457367,
+      "learning_rate": 1.1645575298342659e-06,
+      "loss": 0.3246,
+      "step": 9548
+    },
+    {
+      "epoch": 2.4016599597585513,
+      "grad_norm": 0.2845926582813263,
+      "learning_rate": 1.1636189603010179e-06,
+      "loss": 0.3242,
+      "step": 9549
+    },
+    {
+      "epoch": 2.401911468812877,
+      "grad_norm": 0.28834056854248047,
+      "learning_rate": 1.1626807193298073e-06,
+      "loss": 0.3234,
+      "step": 9550
+    },
+    {
+      "epoch": 2.4021629778672033,
+      "grad_norm": 0.31181180477142334,
+      "learning_rate": 1.161742807000985e-06,
+      "loss": 0.3505,
+      "step": 9551
+    },
+    {
+      "epoch": 2.402414486921529,
+      "grad_norm": 0.30017805099487305,
+      "learning_rate": 1.1608052233948797e-06,
+      "loss": 0.332,
+      "step": 9552
+    },
+    {
+      "epoch": 2.402665995975855,
+      "grad_norm": 0.2945021986961365,
+      "learning_rate": 1.1598679685917901e-06,
+      "loss": 0.3185,
+      "step": 9553
+    },
+    {
+      "epoch": 2.402917505030181,
+      "grad_norm": 0.29077422618865967,
+      "learning_rate": 1.158931042671984e-06,
+      "loss": 0.3215,
+      "step": 9554
+    },
+    {
+      "epoch": 2.403169014084507,
+      "grad_norm": 0.28947049379348755,
+      "learning_rate": 1.157994445715706e-06,
+      "loss": 0.3264,
+      "step": 9555
+    },
+    {
+      "epoch": 2.403420523138833,
+      "grad_norm": 0.2847462296485901,
+      "learning_rate": 1.1570581778031665e-06,
+      "loss": 0.3298,
+      "step": 9556
+    },
+    {
+      "epoch": 2.403672032193159,
+      "grad_norm": 0.29877012968063354,
+      "learning_rate": 1.1561222390145543e-06,
+      "loss": 0.324,
+      "step": 9557
+    },
+    {
+      "epoch": 2.403923541247485,
+      "grad_norm": 0.2849181294441223,
+      "learning_rate": 1.1551866294300234e-06,
+      "loss": 0.321,
+      "step": 9558
+    },
+    {
+      "epoch": 2.404175050301811,
+      "grad_norm": 0.3224382698535919,
+      "learning_rate": 1.1542513491297063e-06,
+      "loss": 0.3194,
+      "step": 9559
+    },
+    {
+      "epoch": 2.404426559356137,
+      "grad_norm": 0.2912129759788513,
+      "learning_rate": 1.1533163981937012e-06,
+      "loss": 0.3128,
+      "step": 9560
+    },
+    {
+      "epoch": 2.4046780684104627,
+      "grad_norm": 0.2958779036998749,
+      "learning_rate": 1.1523817767020829e-06,
+      "loss": 0.3136,
+      "step": 9561
+    },
+    {
+      "epoch": 2.404929577464789,
+      "grad_norm": 0.300225168466568,
+      "learning_rate": 1.1514474847348934e-06,
+      "loss": 0.338,
+      "step": 9562
+    },
+    {
+      "epoch": 2.4051810865191148,
+      "grad_norm": 0.31059449911117554,
+      "learning_rate": 1.1505135223721498e-06,
+      "loss": 0.3326,
+      "step": 9563
+    },
+    {
+      "epoch": 2.4054325955734406,
+      "grad_norm": 0.293338805437088,
+      "learning_rate": 1.1495798896938426e-06,
+      "loss": 0.3294,
+      "step": 9564
+    },
+    {
+      "epoch": 2.405684104627767,
+      "grad_norm": 0.3122572898864746,
+      "learning_rate": 1.1486465867799284e-06,
+      "loss": 0.3319,
+      "step": 9565
+    },
+    {
+      "epoch": 2.4059356136820926,
+      "grad_norm": 0.2884367108345032,
+      "learning_rate": 1.147713613710341e-06,
+      "loss": 0.2962,
+      "step": 9566
+    },
+    {
+      "epoch": 2.4061871227364184,
+      "grad_norm": 0.3003442585468292,
+      "learning_rate": 1.1467809705649817e-06,
+      "loss": 0.3139,
+      "step": 9567
+    },
+    {
+      "epoch": 2.4064386317907447,
+      "grad_norm": 0.2922108471393585,
+      "learning_rate": 1.1458486574237281e-06,
+      "loss": 0.3144,
+      "step": 9568
+    },
+    {
+      "epoch": 2.4066901408450705,
+      "grad_norm": 0.2977858781814575,
+      "learning_rate": 1.144916674366424e-06,
+      "loss": 0.3182,
+      "step": 9569
+    },
+    {
+      "epoch": 2.4069416498993963,
+      "grad_norm": 0.2883601486682892,
+      "learning_rate": 1.1439850214728908e-06,
+      "loss": 0.3377,
+      "step": 9570
+    },
+    {
+      "epoch": 2.4071931589537225,
+      "grad_norm": 0.28815150260925293,
+      "learning_rate": 1.1430536988229157e-06,
+      "loss": 0.3322,
+      "step": 9571
+    },
+    {
+      "epoch": 2.4074446680080483,
+      "grad_norm": 0.3212835192680359,
+      "learning_rate": 1.1421227064962641e-06,
+      "loss": 0.303,
+      "step": 9572
+    },
+    {
+      "epoch": 2.407696177062374,
+      "grad_norm": 0.2907494008541107,
+      "learning_rate": 1.1411920445726666e-06,
+      "loss": 0.3007,
+      "step": 9573
+    },
+    {
+      "epoch": 2.4079476861167004,
+      "grad_norm": 0.3153952360153198,
+      "learning_rate": 1.1402617131318295e-06,
+      "loss": 0.3223,
+      "step": 9574
+    },
+    {
+      "epoch": 2.408199195171026,
+      "grad_norm": 0.2907307744026184,
+      "learning_rate": 1.1393317122534315e-06,
+      "loss": 0.3325,
+      "step": 9575
+    },
+    {
+      "epoch": 2.408450704225352,
+      "grad_norm": 0.30540433526039124,
+      "learning_rate": 1.1384020420171194e-06,
+      "loss": 0.3305,
+      "step": 9576
+    },
+    {
+      "epoch": 2.4087022132796783,
+      "grad_norm": 0.2847049832344055,
+      "learning_rate": 1.137472702502515e-06,
+      "loss": 0.3229,
+      "step": 9577
+    },
+    {
+      "epoch": 2.408953722334004,
+      "grad_norm": 0.29284361004829407,
+      "learning_rate": 1.1365436937892082e-06,
+      "loss": 0.3176,
+      "step": 9578
+    },
+    {
+      "epoch": 2.40920523138833,
+      "grad_norm": 0.28661638498306274,
+      "learning_rate": 1.1356150159567664e-06,
+      "loss": 0.338,
+      "step": 9579
+    },
+    {
+      "epoch": 2.409456740442656,
+      "grad_norm": 0.2953725755214691,
+      "learning_rate": 1.1346866690847214e-06,
+      "loss": 0.3218,
+      "step": 9580
+    },
+    {
+      "epoch": 2.409708249496982,
+      "grad_norm": 0.30523139238357544,
+      "learning_rate": 1.133758653252583e-06,
+      "loss": 0.3014,
+      "step": 9581
+    },
+    {
+      "epoch": 2.4099597585513077,
+      "grad_norm": 0.2769249379634857,
+      "learning_rate": 1.1328309685398275e-06,
+      "loss": 0.3114,
+      "step": 9582
+    },
+    {
+      "epoch": 2.410211267605634,
+      "grad_norm": 0.2869117856025696,
+      "learning_rate": 1.1319036150259078e-06,
+      "loss": 0.339,
+      "step": 9583
+    },
+    {
+      "epoch": 2.41046277665996,
+      "grad_norm": 0.2989134192466736,
+      "learning_rate": 1.1309765927902439e-06,
+      "loss": 0.2997,
+      "step": 9584
+    },
+    {
+      "epoch": 2.4107142857142856,
+      "grad_norm": 0.29306572675704956,
+      "learning_rate": 1.1300499019122295e-06,
+      "loss": 0.321,
+      "step": 9585
+    },
+    {
+      "epoch": 2.410965794768612,
+      "grad_norm": 0.27259159088134766,
+      "learning_rate": 1.1291235424712328e-06,
+      "loss": 0.3093,
+      "step": 9586
+    },
+    {
+      "epoch": 2.4112173038229376,
+      "grad_norm": 0.3309045732021332,
+      "learning_rate": 1.1281975145465867e-06,
+      "loss": 0.316,
+      "step": 9587
+    },
+    {
+      "epoch": 2.4114688128772634,
+      "grad_norm": 0.3107389211654663,
+      "learning_rate": 1.1272718182176034e-06,
+      "loss": 0.3225,
+      "step": 9588
+    },
+    {
+      "epoch": 2.4117203219315897,
+      "grad_norm": 0.29425039887428284,
+      "learning_rate": 1.1263464535635594e-06,
+      "loss": 0.336,
+      "step": 9589
+    },
+    {
+      "epoch": 2.4119718309859155,
+      "grad_norm": 0.29913732409477234,
+      "learning_rate": 1.1254214206637099e-06,
+      "loss": 0.3135,
+      "step": 9590
+    },
+    {
+      "epoch": 2.4122233400402413,
+      "grad_norm": 0.2878449261188507,
+      "learning_rate": 1.1244967195972745e-06,
+      "loss": 0.3308,
+      "step": 9591
+    },
+    {
+      "epoch": 2.4124748490945676,
+      "grad_norm": 0.2862037122249603,
+      "learning_rate": 1.123572350443452e-06,
+      "loss": 0.3186,
+      "step": 9592
+    },
+    {
+      "epoch": 2.4127263581488934,
+      "grad_norm": 0.30815479159355164,
+      "learning_rate": 1.1226483132814048e-06,
+      "loss": 0.3305,
+      "step": 9593
+    },
+    {
+      "epoch": 2.412977867203219,
+      "grad_norm": 0.29880401492118835,
+      "learning_rate": 1.121724608190275e-06,
+      "loss": 0.3288,
+      "step": 9594
+    },
+    {
+      "epoch": 2.4132293762575454,
+      "grad_norm": 0.3055490553379059,
+      "learning_rate": 1.1208012352491681e-06,
+      "loss": 0.3273,
+      "step": 9595
+    },
+    {
+      "epoch": 2.413480885311871,
+      "grad_norm": 0.31553953886032104,
+      "learning_rate": 1.1198781945371673e-06,
+      "loss": 0.2943,
+      "step": 9596
+    },
+    {
+      "epoch": 2.413732394366197,
+      "grad_norm": 0.293268084526062,
+      "learning_rate": 1.118955486133327e-06,
+      "loss": 0.323,
+      "step": 9597
+    },
+    {
+      "epoch": 2.4139839034205233,
+      "grad_norm": 0.304035484790802,
+      "learning_rate": 1.1180331101166675e-06,
+      "loss": 0.3229,
+      "step": 9598
+    },
+    {
+      "epoch": 2.414235412474849,
+      "grad_norm": 0.2874673008918762,
+      "learning_rate": 1.1171110665661888e-06,
+      "loss": 0.3199,
+      "step": 9599
+    },
+    {
+      "epoch": 2.414486921529175,
+      "grad_norm": 0.2967674434185028,
+      "learning_rate": 1.1161893555608538e-06,
+      "loss": 0.3396,
+      "step": 9600
+    },
+    {
+      "epoch": 2.414738430583501,
+      "grad_norm": 0.3168094754219055,
+      "learning_rate": 1.1152679771796054e-06,
+      "loss": 0.3291,
+      "step": 9601
+    },
+    {
+      "epoch": 2.414989939637827,
+      "grad_norm": 0.3081880509853363,
+      "learning_rate": 1.1143469315013505e-06,
+      "loss": 0.3123,
+      "step": 9602
+    },
+    {
+      "epoch": 2.4152414486921527,
+      "grad_norm": 0.3019590377807617,
+      "learning_rate": 1.1134262186049732e-06,
+      "loss": 0.3103,
+      "step": 9603
+    },
+    {
+      "epoch": 2.415492957746479,
+      "grad_norm": 0.3133490979671478,
+      "learning_rate": 1.1125058385693255e-06,
+      "loss": 0.3362,
+      "step": 9604
+    },
+    {
+      "epoch": 2.415744466800805,
+      "grad_norm": 0.2901817262172699,
+      "learning_rate": 1.1115857914732332e-06,
+      "loss": 0.3293,
+      "step": 9605
+    },
+    {
+      "epoch": 2.4159959758551306,
+      "grad_norm": 0.2860255241394043,
+      "learning_rate": 1.1106660773954908e-06,
+      "loss": 0.3074,
+      "step": 9606
+    },
+    {
+      "epoch": 2.416247484909457,
+      "grad_norm": 0.29888466000556946,
+      "learning_rate": 1.109746696414868e-06,
+      "loss": 0.3458,
+      "step": 9607
+    },
+    {
+      "epoch": 2.4164989939637826,
+      "grad_norm": 0.30433139204978943,
+      "learning_rate": 1.1088276486101034e-06,
+      "loss": 0.3235,
+      "step": 9608
+    },
+    {
+      "epoch": 2.4167505030181085,
+      "grad_norm": 0.30524951219558716,
+      "learning_rate": 1.107908934059907e-06,
+      "loss": 0.3208,
+      "step": 9609
+    },
+    {
+      "epoch": 2.4170020120724347,
+      "grad_norm": 0.29082682728767395,
+      "learning_rate": 1.1069905528429631e-06,
+      "loss": 0.3234,
+      "step": 9610
+    },
+    {
+      "epoch": 2.4172535211267605,
+      "grad_norm": 0.2806370258331299,
+      "learning_rate": 1.1060725050379223e-06,
+      "loss": 0.3113,
+      "step": 9611
+    },
+    {
+      "epoch": 2.4175050301810863,
+      "grad_norm": 0.2872363328933716,
+      "learning_rate": 1.1051547907234122e-06,
+      "loss": 0.3013,
+      "step": 9612
+    },
+    {
+      "epoch": 2.4177565392354126,
+      "grad_norm": 0.2908620834350586,
+      "learning_rate": 1.1042374099780268e-06,
+      "loss": 0.3258,
+      "step": 9613
+    },
+    {
+      "epoch": 2.4180080482897384,
+      "grad_norm": 0.2825472354888916,
+      "learning_rate": 1.1033203628803369e-06,
+      "loss": 0.3147,
+      "step": 9614
+    },
+    {
+      "epoch": 2.418259557344064,
+      "grad_norm": 0.29791221022605896,
+      "learning_rate": 1.1024036495088792e-06,
+      "loss": 0.3144,
+      "step": 9615
+    },
+    {
+      "epoch": 2.4185110663983904,
+      "grad_norm": 0.3021547198295593,
+      "learning_rate": 1.1014872699421669e-06,
+      "loss": 0.3039,
+      "step": 9616
+    },
+    {
+      "epoch": 2.4187625754527162,
+      "grad_norm": 0.2768807113170624,
+      "learning_rate": 1.100571224258679e-06,
+      "loss": 0.3238,
+      "step": 9617
+    },
+    {
+      "epoch": 2.419014084507042,
+      "grad_norm": 0.29537469148635864,
+      "learning_rate": 1.099655512536872e-06,
+      "loss": 0.3087,
+      "step": 9618
+    },
+    {
+      "epoch": 2.4192655935613683,
+      "grad_norm": 0.28753364086151123,
+      "learning_rate": 1.0987401348551702e-06,
+      "loss": 0.3266,
+      "step": 9619
+    },
+    {
+      "epoch": 2.419517102615694,
+      "grad_norm": 0.30412012338638306,
+      "learning_rate": 1.097825091291969e-06,
+      "loss": 0.3196,
+      "step": 9620
+    },
+    {
+      "epoch": 2.41976861167002,
+      "grad_norm": 0.29906362295150757,
+      "learning_rate": 1.0969103819256377e-06,
+      "loss": 0.3135,
+      "step": 9621
+    },
+    {
+      "epoch": 2.420020120724346,
+      "grad_norm": 0.3171037435531616,
+      "learning_rate": 1.0959960068345139e-06,
+      "loss": 0.3056,
+      "step": 9622
+    },
+    {
+      "epoch": 2.420271629778672,
+      "grad_norm": 0.2972031831741333,
+      "learning_rate": 1.0950819660969092e-06,
+      "loss": 0.3344,
+      "step": 9623
+    },
+    {
+      "epoch": 2.4205231388329977,
+      "grad_norm": 0.29488030076026917,
+      "learning_rate": 1.094168259791104e-06,
+      "loss": 0.3244,
+      "step": 9624
+    },
+    {
+      "epoch": 2.420774647887324,
+      "grad_norm": 0.290930837392807,
+      "learning_rate": 1.0932548879953543e-06,
+      "loss": 0.3018,
+      "step": 9625
+    },
+    {
+      "epoch": 2.42102615694165,
+      "grad_norm": 0.29524073004722595,
+      "learning_rate": 1.0923418507878807e-06,
+      "loss": 0.3232,
+      "step": 9626
+    },
+    {
+      "epoch": 2.421277665995976,
+      "grad_norm": 0.30457374453544617,
+      "learning_rate": 1.091429148246882e-06,
+      "loss": 0.3264,
+      "step": 9627
+    },
+    {
+      "epoch": 2.421529175050302,
+      "grad_norm": 0.27748963236808777,
+      "learning_rate": 1.0905167804505263e-06,
+      "loss": 0.3311,
+      "step": 9628
+    },
+    {
+      "epoch": 2.4217806841046277,
+      "grad_norm": 0.29854872822761536,
+      "learning_rate": 1.0896047474769489e-06,
+      "loss": 0.3514,
+      "step": 9629
+    },
+    {
+      "epoch": 2.422032193158954,
+      "grad_norm": 0.3030698597431183,
+      "learning_rate": 1.088693049404263e-06,
+      "loss": 0.3006,
+      "step": 9630
+    },
+    {
+      "epoch": 2.4222837022132797,
+      "grad_norm": 0.31721991300582886,
+      "learning_rate": 1.0877816863105473e-06,
+      "loss": 0.3031,
+      "step": 9631
+    },
+    {
+      "epoch": 2.4225352112676055,
+      "grad_norm": 0.2863737642765045,
+      "learning_rate": 1.0868706582738563e-06,
+      "loss": 0.3057,
+      "step": 9632
+    },
+    {
+      "epoch": 2.4227867203219318,
+      "grad_norm": 0.303460031747818,
+      "learning_rate": 1.085959965372212e-06,
+      "loss": 0.3319,
+      "step": 9633
+    },
+    {
+      "epoch": 2.4230382293762576,
+      "grad_norm": 0.29836776852607727,
+      "learning_rate": 1.085049607683612e-06,
+      "loss": 0.3197,
+      "step": 9634
+    },
+    {
+      "epoch": 2.4232897384305834,
+      "grad_norm": 0.2951622009277344,
+      "learning_rate": 1.0841395852860194e-06,
+      "loss": 0.2886,
+      "step": 9635
+    },
+    {
+      "epoch": 2.4235412474849096,
+      "grad_norm": 0.29916274547576904,
+      "learning_rate": 1.0832298982573753e-06,
+      "loss": 0.3294,
+      "step": 9636
+    },
+    {
+      "epoch": 2.4237927565392354,
+      "grad_norm": 0.29172125458717346,
+      "learning_rate": 1.0823205466755858e-06,
+      "loss": 0.3011,
+      "step": 9637
+    },
+    {
+      "epoch": 2.4240442655935612,
+      "grad_norm": 0.28251975774765015,
+      "learning_rate": 1.0814115306185342e-06,
+      "loss": 0.3193,
+      "step": 9638
+    },
+    {
+      "epoch": 2.4242957746478875,
+      "grad_norm": 0.31814032793045044,
+      "learning_rate": 1.0805028501640686e-06,
+      "loss": 0.3236,
+      "step": 9639
+    },
+    {
+      "epoch": 2.4245472837022133,
+      "grad_norm": 0.313716858625412,
+      "learning_rate": 1.0795945053900153e-06,
+      "loss": 0.2972,
+      "step": 9640
+    },
+    {
+      "epoch": 2.424798792756539,
+      "grad_norm": 0.29684916138648987,
+      "learning_rate": 1.078686496374165e-06,
+      "loss": 0.3449,
+      "step": 9641
+    },
+    {
+      "epoch": 2.4250503018108653,
+      "grad_norm": 0.2841853201389313,
+      "learning_rate": 1.0777788231942859e-06,
+      "loss": 0.3192,
+      "step": 9642
+    },
+    {
+      "epoch": 2.425301810865191,
+      "grad_norm": 0.29728415608406067,
+      "learning_rate": 1.076871485928112e-06,
+      "loss": 0.3058,
+      "step": 9643
+    },
+    {
+      "epoch": 2.425553319919517,
+      "grad_norm": 0.29951390624046326,
+      "learning_rate": 1.075964484653354e-06,
+      "loss": 0.3084,
+      "step": 9644
+    },
+    {
+      "epoch": 2.425804828973843,
+      "grad_norm": 0.3010987937450409,
+      "learning_rate": 1.0750578194476875e-06,
+      "loss": 0.3049,
+      "step": 9645
+    },
+    {
+      "epoch": 2.426056338028169,
+      "grad_norm": 0.29705509543418884,
+      "learning_rate": 1.0741514903887657e-06,
+      "loss": 0.3206,
+      "step": 9646
+    },
+    {
+      "epoch": 2.426307847082495,
+      "grad_norm": 0.28296321630477905,
+      "learning_rate": 1.0732454975542079e-06,
+      "loss": 0.3216,
+      "step": 9647
+    },
+    {
+      "epoch": 2.426559356136821,
+      "grad_norm": 0.2731419503688812,
+      "learning_rate": 1.0723398410216085e-06,
+      "loss": 0.3335,
+      "step": 9648
+    },
+    {
+      "epoch": 2.426810865191147,
+      "grad_norm": 0.29805314540863037,
+      "learning_rate": 1.071434520868529e-06,
+      "loss": 0.3268,
+      "step": 9649
+    },
+    {
+      "epoch": 2.4270623742454727,
+      "grad_norm": 0.27815183997154236,
+      "learning_rate": 1.0705295371725066e-06,
+      "loss": 0.3374,
+      "step": 9650
+    },
+    {
+      "epoch": 2.427313883299799,
+      "grad_norm": 0.31619152426719666,
+      "learning_rate": 1.0696248900110461e-06,
+      "loss": 0.332,
+      "step": 9651
+    },
+    {
+      "epoch": 2.4275653923541247,
+      "grad_norm": 0.27347874641418457,
+      "learning_rate": 1.0687205794616262e-06,
+      "loss": 0.2954,
+      "step": 9652
+    },
+    {
+      "epoch": 2.4278169014084505,
+      "grad_norm": 0.323209285736084,
+      "learning_rate": 1.0678166056016936e-06,
+      "loss": 0.3199,
+      "step": 9653
+    },
+    {
+      "epoch": 2.4280684104627768,
+      "grad_norm": 0.2880942225456238,
+      "learning_rate": 1.0669129685086705e-06,
+      "loss": 0.3555,
+      "step": 9654
+    },
+    {
+      "epoch": 2.4283199195171026,
+      "grad_norm": 0.30394867062568665,
+      "learning_rate": 1.0660096682599453e-06,
+      "loss": 0.3058,
+      "step": 9655
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.29747632145881653,
+      "learning_rate": 1.0651067049328818e-06,
+      "loss": 0.324,
+      "step": 9656
+    },
+    {
+      "epoch": 2.4288229376257546,
+      "grad_norm": 0.305722177028656,
+      "learning_rate": 1.0642040786048113e-06,
+      "loss": 0.3026,
+      "step": 9657
+    },
+    {
+      "epoch": 2.4290744466800804,
+      "grad_norm": 0.30282852053642273,
+      "learning_rate": 1.0633017893530407e-06,
+      "loss": 0.3272,
+      "step": 9658
+    },
+    {
+      "epoch": 2.4293259557344067,
+      "grad_norm": 0.29805997014045715,
+      "learning_rate": 1.0623998372548423e-06,
+      "loss": 0.3306,
+      "step": 9659
+    },
+    {
+      "epoch": 2.4295774647887325,
+      "grad_norm": 0.2963772118091583,
+      "learning_rate": 1.0614982223874642e-06,
+      "loss": 0.3272,
+      "step": 9660
+    },
+    {
+      "epoch": 2.4298289738430583,
+      "grad_norm": 0.3043665587902069,
+      "learning_rate": 1.0605969448281257e-06,
+      "loss": 0.3119,
+      "step": 9661
+    },
+    {
+      "epoch": 2.4300804828973845,
+      "grad_norm": 0.2912227511405945,
+      "learning_rate": 1.0596960046540129e-06,
+      "loss": 0.3362,
+      "step": 9662
+    },
+    {
+      "epoch": 2.4303319919517103,
+      "grad_norm": 0.3193990886211395,
+      "learning_rate": 1.0587954019422874e-06,
+      "loss": 0.3238,
+      "step": 9663
+    },
+    {
+      "epoch": 2.430583501006036,
+      "grad_norm": 0.2979458272457123,
+      "learning_rate": 1.057895136770079e-06,
+      "loss": 0.3141,
+      "step": 9664
+    },
+    {
+      "epoch": 2.4308350100603624,
+      "grad_norm": 0.2856670916080475,
+      "learning_rate": 1.056995209214492e-06,
+      "loss": 0.3083,
+      "step": 9665
+    },
+    {
+      "epoch": 2.431086519114688,
+      "grad_norm": 0.3085869550704956,
+      "learning_rate": 1.0560956193525961e-06,
+      "loss": 0.3193,
+      "step": 9666
+    },
+    {
+      "epoch": 2.431338028169014,
+      "grad_norm": 0.30326932668685913,
+      "learning_rate": 1.0551963672614385e-06,
+      "loss": 0.3145,
+      "step": 9667
+    },
+    {
+      "epoch": 2.4315895372233403,
+      "grad_norm": 0.30177420377731323,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.3422,
+      "step": 9668
+    },
+    {
+      "epoch": 2.431841046277666,
+      "grad_norm": 0.28768646717071533,
+      "learning_rate": 1.0533988766993668e-06,
+      "loss": 0.3213,
+      "step": 9669
+    },
+    {
+      "epoch": 2.432092555331992,
+      "grad_norm": 0.3230489194393158,
+      "learning_rate": 1.052500638382396e-06,
+      "loss": 0.3326,
+      "step": 9670
+    },
+    {
+      "epoch": 2.432344064386318,
+      "grad_norm": 0.3124891519546509,
+      "learning_rate": 1.0516027381440502e-06,
+      "loss": 0.3219,
+      "step": 9671
+    },
+    {
+      "epoch": 2.432595573440644,
+      "grad_norm": 0.2979585826396942,
+      "learning_rate": 1.0507051760612302e-06,
+      "loss": 0.304,
+      "step": 9672
+    },
+    {
+      "epoch": 2.4328470824949697,
+      "grad_norm": 0.2868764400482178,
+      "learning_rate": 1.0498079522108034e-06,
+      "loss": 0.349,
+      "step": 9673
+    },
+    {
+      "epoch": 2.433098591549296,
+      "grad_norm": 0.32363319396972656,
+      "learning_rate": 1.0489110666696144e-06,
+      "loss": 0.3378,
+      "step": 9674
+    },
+    {
+      "epoch": 2.433350100603622,
+      "grad_norm": 0.28327086567878723,
+      "learning_rate": 1.0480145195144736e-06,
+      "loss": 0.3041,
+      "step": 9675
+    },
+    {
+      "epoch": 2.4336016096579476,
+      "grad_norm": 0.30430731177330017,
+      "learning_rate": 1.0471183108221673e-06,
+      "loss": 0.3099,
+      "step": 9676
+    },
+    {
+      "epoch": 2.433853118712274,
+      "grad_norm": 0.3252510130405426,
+      "learning_rate": 1.0462224406694471e-06,
+      "loss": 0.3302,
+      "step": 9677
+    },
+    {
+      "epoch": 2.4341046277665996,
+      "grad_norm": 0.29767587780952454,
+      "learning_rate": 1.045326909133041e-06,
+      "loss": 0.2966,
+      "step": 9678
+    },
+    {
+      "epoch": 2.4343561368209254,
+      "grad_norm": 0.3064614534378052,
+      "learning_rate": 1.0444317162896433e-06,
+      "loss": 0.3123,
+      "step": 9679
+    },
+    {
+      "epoch": 2.4346076458752517,
+      "grad_norm": 0.30733683705329895,
+      "learning_rate": 1.0435368622159254e-06,
+      "loss": 0.2995,
+      "step": 9680
+    },
+    {
+      "epoch": 2.4348591549295775,
+      "grad_norm": 0.31988590955734253,
+      "learning_rate": 1.0426423469885216e-06,
+      "loss": 0.3176,
+      "step": 9681
+    },
+    {
+      "epoch": 2.4351106639839033,
+      "grad_norm": 0.29145556688308716,
+      "learning_rate": 1.0417481706840439e-06,
+      "loss": 0.3056,
+      "step": 9682
+    },
+    {
+      "epoch": 2.4353621730382295,
+      "grad_norm": 0.2926798462867737,
+      "learning_rate": 1.0408543333790738e-06,
+      "loss": 0.3149,
+      "step": 9683
+    },
+    {
+      "epoch": 2.4356136820925554,
+      "grad_norm": 0.2882416844367981,
+      "learning_rate": 1.0399608351501606e-06,
+      "loss": 0.3247,
+      "step": 9684
+    },
+    {
+      "epoch": 2.435865191146881,
+      "grad_norm": 0.3121548593044281,
+      "learning_rate": 1.0390676760738289e-06,
+      "loss": 0.336,
+      "step": 9685
+    },
+    {
+      "epoch": 2.4361167002012074,
+      "grad_norm": 0.2849773168563843,
+      "learning_rate": 1.0381748562265704e-06,
+      "loss": 0.3229,
+      "step": 9686
+    },
+    {
+      "epoch": 2.436368209255533,
+      "grad_norm": 0.30011287331581116,
+      "learning_rate": 1.037282375684851e-06,
+      "loss": 0.3302,
+      "step": 9687
+    },
+    {
+      "epoch": 2.436619718309859,
+      "grad_norm": 0.298657089471817,
+      "learning_rate": 1.0363902345251048e-06,
+      "loss": 0.3117,
+      "step": 9688
+    },
+    {
+      "epoch": 2.4368712273641853,
+      "grad_norm": 0.3116760551929474,
+      "learning_rate": 1.0354984328237399e-06,
+      "loss": 0.3293,
+      "step": 9689
+    },
+    {
+      "epoch": 2.437122736418511,
+      "grad_norm": 0.30939438939094543,
+      "learning_rate": 1.034606970657131e-06,
+      "loss": 0.3208,
+      "step": 9690
+    },
+    {
+      "epoch": 2.437374245472837,
+      "grad_norm": 0.3038461208343506,
+      "learning_rate": 1.0337158481016285e-06,
+      "loss": 0.3762,
+      "step": 9691
+    },
+    {
+      "epoch": 2.437625754527163,
+      "grad_norm": 0.2885994613170624,
+      "learning_rate": 1.0328250652335497e-06,
+      "loss": 0.3434,
+      "step": 9692
+    },
+    {
+      "epoch": 2.437877263581489,
+      "grad_norm": 0.28599095344543457,
+      "learning_rate": 1.031934622129186e-06,
+      "loss": 0.3283,
+      "step": 9693
+    },
+    {
+      "epoch": 2.4381287726358147,
+      "grad_norm": 0.2979143559932709,
+      "learning_rate": 1.0310445188647983e-06,
+      "loss": 0.3298,
+      "step": 9694
+    },
+    {
+      "epoch": 2.438380281690141,
+      "grad_norm": 0.300065815448761,
+      "learning_rate": 1.030154755516617e-06,
+      "loss": 0.3153,
+      "step": 9695
+    },
+    {
+      "epoch": 2.438631790744467,
+      "grad_norm": 0.29073888063430786,
+      "learning_rate": 1.029265332160847e-06,
+      "loss": 0.3157,
+      "step": 9696
+    },
+    {
+      "epoch": 2.4388832997987926,
+      "grad_norm": 0.28464123606681824,
+      "learning_rate": 1.0283762488736588e-06,
+      "loss": 0.2994,
+      "step": 9697
+    },
+    {
+      "epoch": 2.439134808853119,
+      "grad_norm": 0.299254447221756,
+      "learning_rate": 1.0274875057312001e-06,
+      "loss": 0.3058,
+      "step": 9698
+    },
+    {
+      "epoch": 2.4393863179074446,
+      "grad_norm": 0.29035767912864685,
+      "learning_rate": 1.0265991028095828e-06,
+      "loss": 0.3407,
+      "step": 9699
+    },
+    {
+      "epoch": 2.4396378269617705,
+      "grad_norm": 0.30013376474380493,
+      "learning_rate": 1.0257110401848963e-06,
+      "loss": 0.3296,
+      "step": 9700
+    },
+    {
+      "epoch": 2.4398893360160967,
+      "grad_norm": 0.30098024010658264,
+      "learning_rate": 1.0248233179331952e-06,
+      "loss": 0.3096,
+      "step": 9701
+    },
+    {
+      "epoch": 2.4401408450704225,
+      "grad_norm": 0.3151046633720398,
+      "learning_rate": 1.0239359361305091e-06,
+      "loss": 0.2879,
+      "step": 9702
+    },
+    {
+      "epoch": 2.4403923541247483,
+      "grad_norm": 0.273389607667923,
+      "learning_rate": 1.023048894852835e-06,
+      "loss": 0.3119,
+      "step": 9703
+    },
+    {
+      "epoch": 2.4406438631790746,
+      "grad_norm": 0.31045857071876526,
+      "learning_rate": 1.0221621941761428e-06,
+      "loss": 0.3154,
+      "step": 9704
+    },
+    {
+      "epoch": 2.4408953722334004,
+      "grad_norm": 0.27846378087997437,
+      "learning_rate": 1.0212758341763752e-06,
+      "loss": 0.3276,
+      "step": 9705
+    },
+    {
+      "epoch": 2.441146881287726,
+      "grad_norm": 0.3135693371295929,
+      "learning_rate": 1.02038981492944e-06,
+      "loss": 0.3199,
+      "step": 9706
+    },
+    {
+      "epoch": 2.4413983903420524,
+      "grad_norm": 0.30014142394065857,
+      "learning_rate": 1.0195041365112224e-06,
+      "loss": 0.3388,
+      "step": 9707
+    },
+    {
+      "epoch": 2.441649899396378,
+      "grad_norm": 0.30678755044937134,
+      "learning_rate": 1.0186187989975722e-06,
+      "loss": 0.3294,
+      "step": 9708
+    },
+    {
+      "epoch": 2.441901408450704,
+      "grad_norm": 0.2942301332950592,
+      "learning_rate": 1.0177338024643157e-06,
+      "loss": 0.3288,
+      "step": 9709
+    },
+    {
+      "epoch": 2.4421529175050303,
+      "grad_norm": 0.2993881404399872,
+      "learning_rate": 1.0168491469872444e-06,
+      "loss": 0.3219,
+      "step": 9710
+    },
+    {
+      "epoch": 2.442404426559356,
+      "grad_norm": 0.322645366191864,
+      "learning_rate": 1.0159648326421268e-06,
+      "loss": 0.2956,
+      "step": 9711
+    },
+    {
+      "epoch": 2.442655935613682,
+      "grad_norm": 0.2843986749649048,
+      "learning_rate": 1.0150808595046963e-06,
+      "loss": 0.3088,
+      "step": 9712
+    },
+    {
+      "epoch": 2.442907444668008,
+      "grad_norm": 0.2960312068462372,
+      "learning_rate": 1.0141972276506612e-06,
+      "loss": 0.3394,
+      "step": 9713
+    },
+    {
+      "epoch": 2.443158953722334,
+      "grad_norm": 0.27722686529159546,
+      "learning_rate": 1.013313937155697e-06,
+      "loss": 0.3286,
+      "step": 9714
+    },
+    {
+      "epoch": 2.4434104627766597,
+      "grad_norm": 0.31136950850486755,
+      "learning_rate": 1.012430988095454e-06,
+      "loss": 0.2942,
+      "step": 9715
+    },
+    {
+      "epoch": 2.443661971830986,
+      "grad_norm": 0.28949692845344543,
+      "learning_rate": 1.0115483805455522e-06,
+      "loss": 0.3309,
+      "step": 9716
+    },
+    {
+      "epoch": 2.443913480885312,
+      "grad_norm": 0.29630956053733826,
+      "learning_rate": 1.0106661145815783e-06,
+      "loss": 0.3099,
+      "step": 9717
+    },
+    {
+      "epoch": 2.4441649899396376,
+      "grad_norm": 0.3057538866996765,
+      "learning_rate": 1.0097841902790955e-06,
+      "loss": 0.3265,
+      "step": 9718
+    },
+    {
+      "epoch": 2.444416498993964,
+      "grad_norm": 0.2971940338611603,
+      "learning_rate": 1.0089026077136333e-06,
+      "loss": 0.3257,
+      "step": 9719
+    },
+    {
+      "epoch": 2.4446680080482897,
+      "grad_norm": 0.30409669876098633,
+      "learning_rate": 1.0080213669606958e-06,
+      "loss": 0.3457,
+      "step": 9720
+    },
+    {
+      "epoch": 2.4449195171026155,
+      "grad_norm": 0.30436453223228455,
+      "learning_rate": 1.007140468095753e-06,
+      "loss": 0.3052,
+      "step": 9721
+    },
+    {
+      "epoch": 2.4451710261569417,
+      "grad_norm": 0.28796908259391785,
+      "learning_rate": 1.006259911194251e-06,
+      "loss": 0.3238,
+      "step": 9722
+    },
+    {
+      "epoch": 2.4454225352112675,
+      "grad_norm": 0.3076968491077423,
+      "learning_rate": 1.005379696331602e-06,
+      "loss": 0.3161,
+      "step": 9723
+    },
+    {
+      "epoch": 2.4456740442655933,
+      "grad_norm": 0.2974853813648224,
+      "learning_rate": 1.0044998235831927e-06,
+      "loss": 0.3177,
+      "step": 9724
+    },
+    {
+      "epoch": 2.4459255533199196,
+      "grad_norm": 0.2881965637207031,
+      "learning_rate": 1.0036202930243766e-06,
+      "loss": 0.3106,
+      "step": 9725
+    },
+    {
+      "epoch": 2.4461770623742454,
+      "grad_norm": 0.2931132912635803,
+      "learning_rate": 1.0027411047304808e-06,
+      "loss": 0.319,
+      "step": 9726
+    },
+    {
+      "epoch": 2.4464285714285716,
+      "grad_norm": 0.2949577569961548,
+      "learning_rate": 1.001862258776804e-06,
+      "loss": 0.3085,
+      "step": 9727
+    },
+    {
+      "epoch": 2.4466800804828974,
+      "grad_norm": 0.2667643129825592,
+      "learning_rate": 1.0009837552386114e-06,
+      "loss": 0.3312,
+      "step": 9728
+    },
+    {
+      "epoch": 2.4469315895372232,
+      "grad_norm": 0.30988672375679016,
+      "learning_rate": 1.0001055941911437e-06,
+      "loss": 0.3215,
+      "step": 9729
+    },
+    {
+      "epoch": 2.4471830985915495,
+      "grad_norm": 0.3098249137401581,
+      "learning_rate": 9.992277757096069e-07,
+      "loss": 0.303,
+      "step": 9730
+    },
+    {
+      "epoch": 2.4474346076458753,
+      "grad_norm": 0.27878841757774353,
+      "learning_rate": 9.98350299869184e-07,
+      "loss": 0.322,
+      "step": 9731
+    },
+    {
+      "epoch": 2.447686116700201,
+      "grad_norm": 0.306535005569458,
+      "learning_rate": 9.974731667450227e-07,
+      "loss": 0.3044,
+      "step": 9732
+    },
+    {
+      "epoch": 2.4479376257545273,
+      "grad_norm": 0.3038245737552643,
+      "learning_rate": 9.965963764122455e-07,
+      "loss": 0.342,
+      "step": 9733
+    },
+    {
+      "epoch": 2.448189134808853,
+      "grad_norm": 0.2671661972999573,
+      "learning_rate": 9.957199289459424e-07,
+      "loss": 0.3162,
+      "step": 9734
+    },
+    {
+      "epoch": 2.448440643863179,
+      "grad_norm": 0.3104800879955292,
+      "learning_rate": 9.948438244211784e-07,
+      "loss": 0.3096,
+      "step": 9735
+    },
+    {
+      "epoch": 2.448692152917505,
+      "grad_norm": 0.3143140971660614,
+      "learning_rate": 9.939680629129828e-07,
+      "loss": 0.318,
+      "step": 9736
+    },
+    {
+      "epoch": 2.448943661971831,
+      "grad_norm": 0.2797510325908661,
+      "learning_rate": 9.930926444963612e-07,
+      "loss": 0.311,
+      "step": 9737
+    },
+    {
+      "epoch": 2.449195171026157,
+      "grad_norm": 0.2918142080307007,
+      "learning_rate": 9.922175692462887e-07,
+      "loss": 0.3394,
+      "step": 9738
+    },
+    {
+      "epoch": 2.449446680080483,
+      "grad_norm": 0.3032245337963104,
+      "learning_rate": 9.913428372377077e-07,
+      "loss": 0.3326,
+      "step": 9739
+    },
+    {
+      "epoch": 2.449698189134809,
+      "grad_norm": 0.2985784411430359,
+      "learning_rate": 9.904684485455358e-07,
+      "loss": 0.3224,
+      "step": 9740
+    },
+    {
+      "epoch": 2.4499496981891347,
+      "grad_norm": 0.30142176151275635,
+      "learning_rate": 9.895944032446563e-07,
+      "loss": 0.335,
+      "step": 9741
+    },
+    {
+      "epoch": 2.450201207243461,
+      "grad_norm": 0.3030637204647064,
+      "learning_rate": 9.887207014099288e-07,
+      "loss": 0.3076,
+      "step": 9742
+    },
+    {
+      "epoch": 2.4504527162977867,
+      "grad_norm": 0.2667844593524933,
+      "learning_rate": 9.878473431161767e-07,
+      "loss": 0.3327,
+      "step": 9743
+    },
+    {
+      "epoch": 2.4507042253521125,
+      "grad_norm": 0.3067658245563507,
+      "learning_rate": 9.869743284382016e-07,
+      "loss": 0.3115,
+      "step": 9744
+    },
+    {
+      "epoch": 2.4509557344064388,
+      "grad_norm": 0.2902940511703491,
+      "learning_rate": 9.861016574507686e-07,
+      "loss": 0.3084,
+      "step": 9745
+    },
+    {
+      "epoch": 2.4512072434607646,
+      "grad_norm": 0.28992483019828796,
+      "learning_rate": 9.852293302286186e-07,
+      "loss": 0.3156,
+      "step": 9746
+    },
+    {
+      "epoch": 2.4514587525150904,
+      "grad_norm": 0.3165602385997772,
+      "learning_rate": 9.843573468464596e-07,
+      "loss": 0.3362,
+      "step": 9747
+    },
+    {
+      "epoch": 2.4517102615694166,
+      "grad_norm": 0.2829355299472809,
+      "learning_rate": 9.834857073789716e-07,
+      "loss": 0.3233,
+      "step": 9748
+    },
+    {
+      "epoch": 2.4519617706237424,
+      "grad_norm": 0.3075374960899353,
+      "learning_rate": 9.826144119008068e-07,
+      "loss": 0.3281,
+      "step": 9749
+    },
+    {
+      "epoch": 2.4522132796780682,
+      "grad_norm": 0.32207030057907104,
+      "learning_rate": 9.817434604865833e-07,
+      "loss": 0.332,
+      "step": 9750
+    },
+    {
+      "epoch": 2.4524647887323945,
+      "grad_norm": 0.29507938027381897,
+      "learning_rate": 9.808728532108963e-07,
+      "loss": 0.3202,
+      "step": 9751
+    },
+    {
+      "epoch": 2.4527162977867203,
+      "grad_norm": 0.29577136039733887,
+      "learning_rate": 9.80002590148304e-07,
+      "loss": 0.349,
+      "step": 9752
+    },
+    {
+      "epoch": 2.452967806841046,
+      "grad_norm": 0.2732732594013214,
+      "learning_rate": 9.791326713733424e-07,
+      "loss": 0.3364,
+      "step": 9753
+    },
+    {
+      "epoch": 2.4532193158953723,
+      "grad_norm": 0.2873142957687378,
+      "learning_rate": 9.782630969605116e-07,
+      "loss": 0.3449,
+      "step": 9754
+    },
+    {
+      "epoch": 2.453470824949698,
+      "grad_norm": 0.28171807527542114,
+      "learning_rate": 9.77393866984288e-07,
+      "loss": 0.3179,
+      "step": 9755
+    },
+    {
+      "epoch": 2.453722334004024,
+      "grad_norm": 0.28104138374328613,
+      "learning_rate": 9.76524981519113e-07,
+      "loss": 0.3137,
+      "step": 9756
+    },
+    {
+      "epoch": 2.45397384305835,
+      "grad_norm": 0.3065677881240845,
+      "learning_rate": 9.756564406394042e-07,
+      "loss": 0.3079,
+      "step": 9757
+    },
+    {
+      "epoch": 2.454225352112676,
+      "grad_norm": 0.2809961438179016,
+      "learning_rate": 9.747882444195434e-07,
+      "loss": 0.3275,
+      "step": 9758
+    },
+    {
+      "epoch": 2.4544768611670023,
+      "grad_norm": 0.2860454022884369,
+      "learning_rate": 9.739203929338892e-07,
+      "loss": 0.3225,
+      "step": 9759
+    },
+    {
+      "epoch": 2.454728370221328,
+      "grad_norm": 0.29092496633529663,
+      "learning_rate": 9.730528862567645e-07,
+      "loss": 0.3164,
+      "step": 9760
+    },
+    {
+      "epoch": 2.454979879275654,
+      "grad_norm": 0.31237539649009705,
+      "learning_rate": 9.721857244624695e-07,
+      "loss": 0.3227,
+      "step": 9761
+    },
+    {
+      "epoch": 2.45523138832998,
+      "grad_norm": 0.28859943151474,
+      "learning_rate": 9.713189076252676e-07,
+      "loss": 0.3339,
+      "step": 9762
+    },
+    {
+      "epoch": 2.455482897384306,
+      "grad_norm": 0.2893824875354767,
+      "learning_rate": 9.704524358193996e-07,
+      "loss": 0.3166,
+      "step": 9763
+    },
+    {
+      "epoch": 2.4557344064386317,
+      "grad_norm": 0.2801450788974762,
+      "learning_rate": 9.6958630911907e-07,
+      "loss": 0.3077,
+      "step": 9764
+    },
+    {
+      "epoch": 2.455985915492958,
+      "grad_norm": 0.2814953029155731,
+      "learning_rate": 9.68720527598459e-07,
+      "loss": 0.3253,
+      "step": 9765
+    },
+    {
+      "epoch": 2.4562374245472838,
+      "grad_norm": 0.2909027338027954,
+      "learning_rate": 9.678550913317169e-07,
+      "loss": 0.3334,
+      "step": 9766
+    },
+    {
+      "epoch": 2.4564889336016096,
+      "grad_norm": 0.3029521703720093,
+      "learning_rate": 9.669900003929595e-07,
+      "loss": 0.321,
+      "step": 9767
+    },
+    {
+      "epoch": 2.456740442655936,
+      "grad_norm": 0.2950665056705475,
+      "learning_rate": 9.661252548562794e-07,
+      "loss": 0.3358,
+      "step": 9768
+    },
+    {
+      "epoch": 2.4569919517102616,
+      "grad_norm": 0.2943587005138397,
+      "learning_rate": 9.652608547957343e-07,
+      "loss": 0.3022,
+      "step": 9769
+    },
+    {
+      "epoch": 2.4572434607645874,
+      "grad_norm": 0.2907446622848511,
+      "learning_rate": 9.643968002853566e-07,
+      "loss": 0.3282,
+      "step": 9770
+    },
+    {
+      "epoch": 2.4574949698189137,
+      "grad_norm": 0.2860673666000366,
+      "learning_rate": 9.635330913991453e-07,
+      "loss": 0.3258,
+      "step": 9771
+    },
+    {
+      "epoch": 2.4577464788732395,
+      "grad_norm": 0.2849615514278412,
+      "learning_rate": 9.626697282110743e-07,
+      "loss": 0.3418,
+      "step": 9772
+    },
+    {
+      "epoch": 2.4579979879275653,
+      "grad_norm": 0.2882292866706848,
+      "learning_rate": 9.61806710795082e-07,
+      "loss": 0.3055,
+      "step": 9773
+    },
+    {
+      "epoch": 2.4582494969818915,
+      "grad_norm": 0.28438448905944824,
+      "learning_rate": 9.609440392250829e-07,
+      "loss": 0.3548,
+      "step": 9774
+    },
+    {
+      "epoch": 2.4585010060362174,
+      "grad_norm": 0.28415539860725403,
+      "learning_rate": 9.600817135749579e-07,
+      "loss": 0.3328,
+      "step": 9775
+    },
+    {
+      "epoch": 2.458752515090543,
+      "grad_norm": 0.2894395887851715,
+      "learning_rate": 9.592197339185617e-07,
+      "loss": 0.3477,
+      "step": 9776
+    },
+    {
+      "epoch": 2.4590040241448694,
+      "grad_norm": 0.2790740728378296,
+      "learning_rate": 9.583581003297148e-07,
+      "loss": 0.326,
+      "step": 9777
+    },
+    {
+      "epoch": 2.459255533199195,
+      "grad_norm": 0.29626601934432983,
+      "learning_rate": 9.574968128822138e-07,
+      "loss": 0.3037,
+      "step": 9778
+    },
+    {
+      "epoch": 2.459507042253521,
+      "grad_norm": 0.29274749755859375,
+      "learning_rate": 9.566358716498192e-07,
+      "loss": 0.3609,
+      "step": 9779
+    },
+    {
+      "epoch": 2.4597585513078473,
+      "grad_norm": 0.2905212640762329,
+      "learning_rate": 9.557752767062683e-07,
+      "loss": 0.3246,
+      "step": 9780
+    },
+    {
+      "epoch": 2.460010060362173,
+      "grad_norm": 0.2953580915927887,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.3493,
+      "step": 9781
+    },
+    {
+      "epoch": 2.460261569416499,
+      "grad_norm": 0.2971348464488983,
+      "learning_rate": 9.540551259804814e-07,
+      "loss": 0.3454,
+      "step": 9782
+    },
+    {
+      "epoch": 2.460513078470825,
+      "grad_norm": 0.2992364168167114,
+      "learning_rate": 9.531955703455654e-07,
+      "loss": 0.3294,
+      "step": 9783
+    },
+    {
+      "epoch": 2.460764587525151,
+      "grad_norm": 0.3017844259738922,
+      "learning_rate": 9.523363612941333e-07,
+      "loss": 0.3127,
+      "step": 9784
+    },
+    {
+      "epoch": 2.4610160965794767,
+      "grad_norm": 0.2864784300327301,
+      "learning_rate": 9.514774988997683e-07,
+      "loss": 0.3399,
+      "step": 9785
+    },
+    {
+      "epoch": 2.461267605633803,
+      "grad_norm": 0.29324662685394287,
+      "learning_rate": 9.506189832360296e-07,
+      "loss": 0.3104,
+      "step": 9786
+    },
+    {
+      "epoch": 2.461519114688129,
+      "grad_norm": 0.30676597356796265,
+      "learning_rate": 9.497608143764403e-07,
+      "loss": 0.298,
+      "step": 9787
+    },
+    {
+      "epoch": 2.4617706237424546,
+      "grad_norm": 0.2943531274795532,
+      "learning_rate": 9.489029923945009e-07,
+      "loss": 0.3265,
+      "step": 9788
+    },
+    {
+      "epoch": 2.462022132796781,
+      "grad_norm": 0.30594533681869507,
+      "learning_rate": 9.480455173636754e-07,
+      "loss": 0.3198,
+      "step": 9789
+    },
+    {
+      "epoch": 2.4622736418511066,
+      "grad_norm": 0.2777751684188843,
+      "learning_rate": 9.471883893574019e-07,
+      "loss": 0.3289,
+      "step": 9790
+    },
+    {
+      "epoch": 2.4625251509054324,
+      "grad_norm": 0.3079993426799774,
+      "learning_rate": 9.463316084490903e-07,
+      "loss": 0.3278,
+      "step": 9791
+    },
+    {
+      "epoch": 2.4627766599597587,
+      "grad_norm": 0.29290157556533813,
+      "learning_rate": 9.454751747121149e-07,
+      "loss": 0.3278,
+      "step": 9792
+    },
+    {
+      "epoch": 2.4630281690140845,
+      "grad_norm": 0.2852899432182312,
+      "learning_rate": 9.446190882198275e-07,
+      "loss": 0.3066,
+      "step": 9793
+    },
+    {
+      "epoch": 2.4632796780684103,
+      "grad_norm": 0.2716057300567627,
+      "learning_rate": 9.437633490455434e-07,
+      "loss": 0.3427,
+      "step": 9794
+    },
+    {
+      "epoch": 2.4635311871227366,
+      "grad_norm": 0.2801603674888611,
+      "learning_rate": 9.429079572625543e-07,
+      "loss": 0.331,
+      "step": 9795
+    },
+    {
+      "epoch": 2.4637826961770624,
+      "grad_norm": 0.2798592150211334,
+      "learning_rate": 9.420529129441159e-07,
+      "loss": 0.3137,
+      "step": 9796
+    },
+    {
+      "epoch": 2.464034205231388,
+      "grad_norm": 0.2886095345020294,
+      "learning_rate": 9.411982161634603e-07,
+      "loss": 0.2978,
+      "step": 9797
+    },
+    {
+      "epoch": 2.4642857142857144,
+      "grad_norm": 0.31059303879737854,
+      "learning_rate": 9.403438669937848e-07,
+      "loss": 0.3069,
+      "step": 9798
+    },
+    {
+      "epoch": 2.46453722334004,
+      "grad_norm": 0.2892918884754181,
+      "learning_rate": 9.394898655082607e-07,
+      "loss": 0.3412,
+      "step": 9799
+    },
+    {
+      "epoch": 2.464788732394366,
+      "grad_norm": 0.29814252257347107,
+      "learning_rate": 9.386362117800262e-07,
+      "loss": 0.313,
+      "step": 9800
+    },
+    {
+      "epoch": 2.4650402414486923,
+      "grad_norm": 0.30424758791923523,
+      "learning_rate": 9.377829058821924e-07,
+      "loss": 0.3106,
+      "step": 9801
+    },
+    {
+      "epoch": 2.465291750503018,
+      "grad_norm": 0.3049749732017517,
+      "learning_rate": 9.369299478878408e-07,
+      "loss": 0.3122,
+      "step": 9802
+    },
+    {
+      "epoch": 2.465543259557344,
+      "grad_norm": 0.2787761092185974,
+      "learning_rate": 9.360773378700194e-07,
+      "loss": 0.319,
+      "step": 9803
+    },
+    {
+      "epoch": 2.46579476861167,
+      "grad_norm": 0.3029969334602356,
+      "learning_rate": 9.352250759017517e-07,
+      "loss": 0.3046,
+      "step": 9804
+    },
+    {
+      "epoch": 2.466046277665996,
+      "grad_norm": 0.2986536920070648,
+      "learning_rate": 9.343731620560254e-07,
+      "loss": 0.3045,
+      "step": 9805
+    },
+    {
+      "epoch": 2.4662977867203217,
+      "grad_norm": 0.3007206618785858,
+      "learning_rate": 9.335215964058047e-07,
+      "loss": 0.3246,
+      "step": 9806
+    },
+    {
+      "epoch": 2.466549295774648,
+      "grad_norm": 0.31535202264785767,
+      "learning_rate": 9.326703790240183e-07,
+      "loss": 0.3106,
+      "step": 9807
+    },
+    {
+      "epoch": 2.466800804828974,
+      "grad_norm": 0.29556044936180115,
+      "learning_rate": 9.318195099835697e-07,
+      "loss": 0.3332,
+      "step": 9808
+    },
+    {
+      "epoch": 2.4670523138832996,
+      "grad_norm": 0.2860323190689087,
+      "learning_rate": 9.309689893573287e-07,
+      "loss": 0.3225,
+      "step": 9809
+    },
+    {
+      "epoch": 2.467303822937626,
+      "grad_norm": 0.3032556176185608,
+      "learning_rate": 9.301188172181391e-07,
+      "loss": 0.2963,
+      "step": 9810
+    },
+    {
+      "epoch": 2.4675553319919517,
+      "grad_norm": 0.29713839292526245,
+      "learning_rate": 9.292689936388106e-07,
+      "loss": 0.327,
+      "step": 9811
+    },
+    {
+      "epoch": 2.4678068410462775,
+      "grad_norm": 0.2727225124835968,
+      "learning_rate": 9.284195186921263e-07,
+      "loss": 0.3128,
+      "step": 9812
+    },
+    {
+      "epoch": 2.4680583501006037,
+      "grad_norm": 0.2951027452945709,
+      "learning_rate": 9.2757039245084e-07,
+      "loss": 0.3165,
+      "step": 9813
+    },
+    {
+      "epoch": 2.4683098591549295,
+      "grad_norm": 0.31366926431655884,
+      "learning_rate": 9.267216149876712e-07,
+      "loss": 0.3337,
+      "step": 9814
+    },
+    {
+      "epoch": 2.4685613682092553,
+      "grad_norm": 0.30813443660736084,
+      "learning_rate": 9.258731863753145e-07,
+      "loss": 0.3278,
+      "step": 9815
+    },
+    {
+      "epoch": 2.4688128772635816,
+      "grad_norm": 0.30502599477767944,
+      "learning_rate": 9.250251066864308e-07,
+      "loss": 0.3154,
+      "step": 9816
+    },
+    {
+      "epoch": 2.4690643863179074,
+      "grad_norm": 0.2972979247570038,
+      "learning_rate": 9.241773759936551e-07,
+      "loss": 0.3481,
+      "step": 9817
+    },
+    {
+      "epoch": 2.469315895372233,
+      "grad_norm": 0.32507577538490295,
+      "learning_rate": 9.233299943695878e-07,
+      "loss": 0.3105,
+      "step": 9818
+    },
+    {
+      "epoch": 2.4695674044265594,
+      "grad_norm": 0.30700263381004333,
+      "learning_rate": 9.224829618868037e-07,
+      "loss": 0.3226,
+      "step": 9819
+    },
+    {
+      "epoch": 2.4698189134808852,
+      "grad_norm": 0.3043920397758484,
+      "learning_rate": 9.21636278617844e-07,
+      "loss": 0.3128,
+      "step": 9820
+    },
+    {
+      "epoch": 2.470070422535211,
+      "grad_norm": 0.2863234579563141,
+      "learning_rate": 9.207899446352242e-07,
+      "loss": 0.3091,
+      "step": 9821
+    },
+    {
+      "epoch": 2.4703219315895373,
+      "grad_norm": 0.3094150722026825,
+      "learning_rate": 9.199439600114247e-07,
+      "loss": 0.3213,
+      "step": 9822
+    },
+    {
+      "epoch": 2.470573440643863,
+      "grad_norm": 0.2926952838897705,
+      "learning_rate": 9.190983248189006e-07,
+      "loss": 0.3025,
+      "step": 9823
+    },
+    {
+      "epoch": 2.470824949698189,
+      "grad_norm": 0.3109200596809387,
+      "learning_rate": 9.18253039130076e-07,
+      "loss": 0.3124,
+      "step": 9824
+    },
+    {
+      "epoch": 2.471076458752515,
+      "grad_norm": 0.32803231477737427,
+      "learning_rate": 9.174081030173421e-07,
+      "loss": 0.2898,
+      "step": 9825
+    },
+    {
+      "epoch": 2.471327967806841,
+      "grad_norm": 0.30386537313461304,
+      "learning_rate": 9.165635165530645e-07,
+      "loss": 0.3086,
+      "step": 9826
+    },
+    {
+      "epoch": 2.471579476861167,
+      "grad_norm": 0.27869728207588196,
+      "learning_rate": 9.157192798095748e-07,
+      "loss": 0.341,
+      "step": 9827
+    },
+    {
+      "epoch": 2.471830985915493,
+      "grad_norm": 0.2737056612968445,
+      "learning_rate": 9.148753928591791e-07,
+      "loss": 0.318,
+      "step": 9828
+    },
+    {
+      "epoch": 2.472082494969819,
+      "grad_norm": 0.3147308826446533,
+      "learning_rate": 9.140318557741479e-07,
+      "loss": 0.309,
+      "step": 9829
+    },
+    {
+      "epoch": 2.472334004024145,
+      "grad_norm": 0.324812650680542,
+      "learning_rate": 9.131886686267277e-07,
+      "loss": 0.3185,
+      "step": 9830
+    },
+    {
+      "epoch": 2.472585513078471,
+      "grad_norm": 0.29894959926605225,
+      "learning_rate": 9.123458314891304e-07,
+      "loss": 0.3313,
+      "step": 9831
+    },
+    {
+      "epoch": 2.4728370221327967,
+      "grad_norm": 0.278921514749527,
+      "learning_rate": 9.115033444335408e-07,
+      "loss": 0.3172,
+      "step": 9832
+    },
+    {
+      "epoch": 2.473088531187123,
+      "grad_norm": 0.27208349108695984,
+      "learning_rate": 9.106612075321114e-07,
+      "loss": 0.3138,
+      "step": 9833
+    },
+    {
+      "epoch": 2.4733400402414487,
+      "grad_norm": 0.27421247959136963,
+      "learning_rate": 9.098194208569666e-07,
+      "loss": 0.3112,
+      "step": 9834
+    },
+    {
+      "epoch": 2.4735915492957745,
+      "grad_norm": 0.3077579736709595,
+      "learning_rate": 9.089779844802016e-07,
+      "loss": 0.3335,
+      "step": 9835
+    },
+    {
+      "epoch": 2.4738430583501008,
+      "grad_norm": 0.2903458774089813,
+      "learning_rate": 9.081368984738781e-07,
+      "loss": 0.3324,
+      "step": 9836
+    },
+    {
+      "epoch": 2.4740945674044266,
+      "grad_norm": 0.2951434254646301,
+      "learning_rate": 9.072961629100313e-07,
+      "loss": 0.3341,
+      "step": 9837
+    },
+    {
+      "epoch": 2.4743460764587524,
+      "grad_norm": 0.28142881393432617,
+      "learning_rate": 9.064557778606631e-07,
+      "loss": 0.3054,
+      "step": 9838
+    },
+    {
+      "epoch": 2.4745975855130786,
+      "grad_norm": 0.2758212387561798,
+      "learning_rate": 9.056157433977497e-07,
+      "loss": 0.3227,
+      "step": 9839
+    },
+    {
+      "epoch": 2.4748490945674044,
+      "grad_norm": 0.2981395125389099,
+      "learning_rate": 9.04776059593232e-07,
+      "loss": 0.3408,
+      "step": 9840
+    },
+    {
+      "epoch": 2.4751006036217302,
+      "grad_norm": 0.2696145176887512,
+      "learning_rate": 9.039367265190268e-07,
+      "loss": 0.3244,
+      "step": 9841
+    },
+    {
+      "epoch": 2.4753521126760565,
+      "grad_norm": 0.2982428967952728,
+      "learning_rate": 9.030977442470146e-07,
+      "loss": 0.3094,
+      "step": 9842
+    },
+    {
+      "epoch": 2.4756036217303823,
+      "grad_norm": 0.2985045611858368,
+      "learning_rate": 9.022591128490521e-07,
+      "loss": 0.3323,
+      "step": 9843
+    },
+    {
+      "epoch": 2.475855130784708,
+      "grad_norm": 0.2910267114639282,
+      "learning_rate": 9.014208323969598e-07,
+      "loss": 0.3386,
+      "step": 9844
+    },
+    {
+      "epoch": 2.4761066398390343,
+      "grad_norm": 0.290507972240448,
+      "learning_rate": 9.005829029625324e-07,
+      "loss": 0.3003,
+      "step": 9845
+    },
+    {
+      "epoch": 2.47635814889336,
+      "grad_norm": 0.2682448923587799,
+      "learning_rate": 8.997453246175347e-07,
+      "loss": 0.3272,
+      "step": 9846
+    },
+    {
+      "epoch": 2.476609657947686,
+      "grad_norm": 0.2868577241897583,
+      "learning_rate": 8.989080974336972e-07,
+      "loss": 0.3217,
+      "step": 9847
+    },
+    {
+      "epoch": 2.476861167002012,
+      "grad_norm": 0.28790926933288574,
+      "learning_rate": 8.98071221482727e-07,
+      "loss": 0.3042,
+      "step": 9848
+    },
+    {
+      "epoch": 2.477112676056338,
+      "grad_norm": 0.29729732871055603,
+      "learning_rate": 8.972346968362932e-07,
+      "loss": 0.304,
+      "step": 9849
+    },
+    {
+      "epoch": 2.477364185110664,
+      "grad_norm": 0.28235840797424316,
+      "learning_rate": 8.963985235660422e-07,
+      "loss": 0.325,
+      "step": 9850
+    },
+    {
+      "epoch": 2.47761569416499,
+      "grad_norm": 0.2943761646747589,
+      "learning_rate": 8.955627017435841e-07,
+      "loss": 0.3345,
+      "step": 9851
+    },
+    {
+      "epoch": 2.477867203219316,
+      "grad_norm": 0.29644346237182617,
+      "learning_rate": 8.947272314405048e-07,
+      "loss": 0.326,
+      "step": 9852
+    },
+    {
+      "epoch": 2.4781187122736417,
+      "grad_norm": 0.29010266065597534,
+      "learning_rate": 8.938921127283545e-07,
+      "loss": 0.337,
+      "step": 9853
+    },
+    {
+      "epoch": 2.478370221327968,
+      "grad_norm": 0.31179898977279663,
+      "learning_rate": 8.930573456786584e-07,
+      "loss": 0.3329,
+      "step": 9854
+    },
+    {
+      "epoch": 2.4786217303822937,
+      "grad_norm": 0.30490022897720337,
+      "learning_rate": 8.922229303629059e-07,
+      "loss": 0.3509,
+      "step": 9855
+    },
+    {
+      "epoch": 2.4788732394366195,
+      "grad_norm": 0.2934214472770691,
+      "learning_rate": 8.913888668525616e-07,
+      "loss": 0.3173,
+      "step": 9856
+    },
+    {
+      "epoch": 2.4791247484909458,
+      "grad_norm": 0.29519298672676086,
+      "learning_rate": 8.905551552190589e-07,
+      "loss": 0.3348,
+      "step": 9857
+    },
+    {
+      "epoch": 2.4793762575452716,
+      "grad_norm": 0.2930413484573364,
+      "learning_rate": 8.897217955337967e-07,
+      "loss": 0.3564,
+      "step": 9858
+    },
+    {
+      "epoch": 2.479627766599598,
+      "grad_norm": 0.2853700816631317,
+      "learning_rate": 8.888887878681507e-07,
+      "loss": 0.3438,
+      "step": 9859
+    },
+    {
+      "epoch": 2.4798792756539236,
+      "grad_norm": 0.30569860339164734,
+      "learning_rate": 8.880561322934595e-07,
+      "loss": 0.3153,
+      "step": 9860
+    },
+    {
+      "epoch": 2.4801307847082494,
+      "grad_norm": 0.30227360129356384,
+      "learning_rate": 8.872238288810375e-07,
+      "loss": 0.3207,
+      "step": 9861
+    },
+    {
+      "epoch": 2.4803822937625757,
+      "grad_norm": 0.2963835895061493,
+      "learning_rate": 8.86391877702164e-07,
+      "loss": 0.3277,
+      "step": 9862
+    },
+    {
+      "epoch": 2.4806338028169015,
+      "grad_norm": 0.29063740372657776,
+      "learning_rate": 8.855602788280926e-07,
+      "loss": 0.3126,
+      "step": 9863
+    },
+    {
+      "epoch": 2.4808853118712273,
+      "grad_norm": 0.2818237245082855,
+      "learning_rate": 8.847290323300423e-07,
+      "loss": 0.3289,
+      "step": 9864
+    },
+    {
+      "epoch": 2.4811368209255535,
+      "grad_norm": 0.2938326299190521,
+      "learning_rate": 8.838981382792067e-07,
+      "loss": 0.3429,
+      "step": 9865
+    },
+    {
+      "epoch": 2.4813883299798793,
+      "grad_norm": 0.30507832765579224,
+      "learning_rate": 8.830675967467439e-07,
+      "loss": 0.3219,
+      "step": 9866
+    },
+    {
+      "epoch": 2.481639839034205,
+      "grad_norm": 0.2869739234447479,
+      "learning_rate": 8.822374078037859e-07,
+      "loss": 0.305,
+      "step": 9867
+    },
+    {
+      "epoch": 2.4818913480885314,
+      "grad_norm": 0.3017198145389557,
+      "learning_rate": 8.814075715214348e-07,
+      "loss": 0.3455,
+      "step": 9868
+    },
+    {
+      "epoch": 2.482142857142857,
+      "grad_norm": 0.2792239189147949,
+      "learning_rate": 8.805780879707582e-07,
+      "loss": 0.3077,
+      "step": 9869
+    },
+    {
+      "epoch": 2.482394366197183,
+      "grad_norm": 0.2923182249069214,
+      "learning_rate": 8.797489572227985e-07,
+      "loss": 0.3207,
+      "step": 9870
+    },
+    {
+      "epoch": 2.4826458752515093,
+      "grad_norm": 0.3037014901638031,
+      "learning_rate": 8.78920179348563e-07,
+      "loss": 0.3091,
+      "step": 9871
+    },
+    {
+      "epoch": 2.482897384305835,
+      "grad_norm": 0.312212735414505,
+      "learning_rate": 8.780917544190337e-07,
+      "loss": 0.3114,
+      "step": 9872
+    },
+    {
+      "epoch": 2.483148893360161,
+      "grad_norm": 0.29248368740081787,
+      "learning_rate": 8.772636825051584e-07,
+      "loss": 0.3165,
+      "step": 9873
+    },
+    {
+      "epoch": 2.483400402414487,
+      "grad_norm": 0.291723370552063,
+      "learning_rate": 8.76435963677858e-07,
+      "loss": 0.3269,
+      "step": 9874
+    },
+    {
+      "epoch": 2.483651911468813,
+      "grad_norm": 0.26432013511657715,
+      "learning_rate": 8.756085980080193e-07,
+      "loss": 0.3193,
+      "step": 9875
+    },
+    {
+      "epoch": 2.4839034205231387,
+      "grad_norm": 0.2941466271877289,
+      "learning_rate": 8.747815855665026e-07,
+      "loss": 0.3385,
+      "step": 9876
+    },
+    {
+      "epoch": 2.484154929577465,
+      "grad_norm": 0.2962441146373749,
+      "learning_rate": 8.739549264241353e-07,
+      "loss": 0.3287,
+      "step": 9877
+    },
+    {
+      "epoch": 2.484406438631791,
+      "grad_norm": 0.29611530900001526,
+      "learning_rate": 8.731286206517158e-07,
+      "loss": 0.3197,
+      "step": 9878
+    },
+    {
+      "epoch": 2.4846579476861166,
+      "grad_norm": 0.2853570878505707,
+      "learning_rate": 8.723026683200136e-07,
+      "loss": 0.3227,
+      "step": 9879
+    },
+    {
+      "epoch": 2.484909456740443,
+      "grad_norm": 0.3189356029033661,
+      "learning_rate": 8.714770694997637e-07,
+      "loss": 0.351,
+      "step": 9880
+    },
+    {
+      "epoch": 2.4851609657947686,
+      "grad_norm": 0.2843788266181946,
+      "learning_rate": 8.706518242616762e-07,
+      "loss": 0.357,
+      "step": 9881
+    },
+    {
+      "epoch": 2.4854124748490944,
+      "grad_norm": 0.2974276542663574,
+      "learning_rate": 8.698269326764258e-07,
+      "loss": 0.3195,
+      "step": 9882
+    },
+    {
+      "epoch": 2.4856639839034207,
+      "grad_norm": 0.28935566544532776,
+      "learning_rate": 8.690023948146614e-07,
+      "loss": 0.3518,
+      "step": 9883
+    },
+    {
+      "epoch": 2.4859154929577465,
+      "grad_norm": 0.29735007882118225,
+      "learning_rate": 8.681782107469971e-07,
+      "loss": 0.3126,
+      "step": 9884
+    },
+    {
+      "epoch": 2.4861670020120723,
+      "grad_norm": 0.29163575172424316,
+      "learning_rate": 8.673543805440216e-07,
+      "loss": 0.3077,
+      "step": 9885
+    },
+    {
+      "epoch": 2.4864185110663986,
+      "grad_norm": 0.29763737320899963,
+      "learning_rate": 8.665309042762888e-07,
+      "loss": 0.3097,
+      "step": 9886
+    },
+    {
+      "epoch": 2.4866700201207244,
+      "grad_norm": 0.3058236539363861,
+      "learning_rate": 8.657077820143262e-07,
+      "loss": 0.296,
+      "step": 9887
+    },
+    {
+      "epoch": 2.48692152917505,
+      "grad_norm": 0.2766462564468384,
+      "learning_rate": 8.648850138286263e-07,
+      "loss": 0.3218,
+      "step": 9888
+    },
+    {
+      "epoch": 2.4871730382293764,
+      "grad_norm": 0.3000984489917755,
+      "learning_rate": 8.640625997896579e-07,
+      "loss": 0.3407,
+      "step": 9889
+    },
+    {
+      "epoch": 2.487424547283702,
+      "grad_norm": 0.3003200888633728,
+      "learning_rate": 8.632405399678518e-07,
+      "loss": 0.3151,
+      "step": 9890
+    },
+    {
+      "epoch": 2.487676056338028,
+      "grad_norm": 0.288412481546402,
+      "learning_rate": 8.624188344336148e-07,
+      "loss": 0.3449,
+      "step": 9891
+    },
+    {
+      "epoch": 2.4879275653923543,
+      "grad_norm": 0.3167831599712372,
+      "learning_rate": 8.615974832573193e-07,
+      "loss": 0.3157,
+      "step": 9892
+    },
+    {
+      "epoch": 2.48817907444668,
+      "grad_norm": 0.27706781029701233,
+      "learning_rate": 8.607764865093104e-07,
+      "loss": 0.3028,
+      "step": 9893
+    },
+    {
+      "epoch": 2.488430583501006,
+      "grad_norm": 0.2621653079986572,
+      "learning_rate": 8.599558442598998e-07,
+      "loss": 0.3303,
+      "step": 9894
+    },
+    {
+      "epoch": 2.488682092555332,
+      "grad_norm": 0.29237544536590576,
+      "learning_rate": 8.591355565793724e-07,
+      "loss": 0.3163,
+      "step": 9895
+    },
+    {
+      "epoch": 2.488933601609658,
+      "grad_norm": 0.31671497225761414,
+      "learning_rate": 8.583156235379774e-07,
+      "loss": 0.3258,
+      "step": 9896
+    },
+    {
+      "epoch": 2.4891851106639837,
+      "grad_norm": 0.3017057478427887,
+      "learning_rate": 8.574960452059411e-07,
+      "loss": 0.3216,
+      "step": 9897
+    },
+    {
+      "epoch": 2.48943661971831,
+      "grad_norm": 0.2894316017627716,
+      "learning_rate": 8.566768216534516e-07,
+      "loss": 0.3205,
+      "step": 9898
+    },
+    {
+      "epoch": 2.489688128772636,
+      "grad_norm": 0.28005433082580566,
+      "learning_rate": 8.558579529506728e-07,
+      "loss": 0.3178,
+      "step": 9899
+    },
+    {
+      "epoch": 2.4899396378269616,
+      "grad_norm": 0.3050450086593628,
+      "learning_rate": 8.55039439167733e-07,
+      "loss": 0.3008,
+      "step": 9900
+    },
+    {
+      "epoch": 2.490191146881288,
+      "grad_norm": 0.2795218527317047,
+      "learning_rate": 8.542212803747363e-07,
+      "loss": 0.3197,
+      "step": 9901
+    },
+    {
+      "epoch": 2.4904426559356136,
+      "grad_norm": 0.2750023901462555,
+      "learning_rate": 8.534034766417498e-07,
+      "loss": 0.3155,
+      "step": 9902
+    },
+    {
+      "epoch": 2.4906941649899395,
+      "grad_norm": 0.30676746368408203,
+      "learning_rate": 8.525860280388154e-07,
+      "loss": 0.31,
+      "step": 9903
+    },
+    {
+      "epoch": 2.4909456740442657,
+      "grad_norm": 0.28755173087120056,
+      "learning_rate": 8.517689346359409e-07,
+      "loss": 0.3182,
+      "step": 9904
+    },
+    {
+      "epoch": 2.4911971830985915,
+      "grad_norm": 0.30206596851348877,
+      "learning_rate": 8.509521965031064e-07,
+      "loss": 0.299,
+      "step": 9905
+    },
+    {
+      "epoch": 2.4914486921529173,
+      "grad_norm": 0.28528329730033875,
+      "learning_rate": 8.501358137102589e-07,
+      "loss": 0.3261,
+      "step": 9906
+    },
+    {
+      "epoch": 2.4917002012072436,
+      "grad_norm": 0.27517518401145935,
+      "learning_rate": 8.493197863273189e-07,
+      "loss": 0.3225,
+      "step": 9907
+    },
+    {
+      "epoch": 2.4919517102615694,
+      "grad_norm": 0.29692593216896057,
+      "learning_rate": 8.485041144241712e-07,
+      "loss": 0.3461,
+      "step": 9908
+    },
+    {
+      "epoch": 2.492203219315895,
+      "grad_norm": 0.288084477186203,
+      "learning_rate": 8.476887980706761e-07,
+      "loss": 0.3184,
+      "step": 9909
+    },
+    {
+      "epoch": 2.4924547283702214,
+      "grad_norm": 0.2816907465457916,
+      "learning_rate": 8.468738373366569e-07,
+      "loss": 0.3151,
+      "step": 9910
+    },
+    {
+      "epoch": 2.4927062374245472,
+      "grad_norm": 0.3093920946121216,
+      "learning_rate": 8.46059232291912e-07,
+      "loss": 0.3172,
+      "step": 9911
+    },
+    {
+      "epoch": 2.492957746478873,
+      "grad_norm": 0.30304044485092163,
+      "learning_rate": 8.452449830062082e-07,
+      "loss": 0.3061,
+      "step": 9912
+    },
+    {
+      "epoch": 2.4932092555331993,
+      "grad_norm": 0.2883797287940979,
+      "learning_rate": 8.44431089549278e-07,
+      "loss": 0.3292,
+      "step": 9913
+    },
+    {
+      "epoch": 2.493460764587525,
+      "grad_norm": 0.2927064597606659,
+      "learning_rate": 8.436175519908291e-07,
+      "loss": 0.3416,
+      "step": 9914
+    },
+    {
+      "epoch": 2.493712273641851,
+      "grad_norm": 0.29678410291671753,
+      "learning_rate": 8.428043704005334e-07,
+      "loss": 0.3239,
+      "step": 9915
+    },
+    {
+      "epoch": 2.493963782696177,
+      "grad_norm": 0.299121230840683,
+      "learning_rate": 8.419915448480376e-07,
+      "loss": 0.3275,
+      "step": 9916
+    },
+    {
+      "epoch": 2.494215291750503,
+      "grad_norm": 0.2699803411960602,
+      "learning_rate": 8.41179075402952e-07,
+      "loss": 0.3286,
+      "step": 9917
+    },
+    {
+      "epoch": 2.4944668008048287,
+      "grad_norm": 0.29386457800865173,
+      "learning_rate": 8.40366962134862e-07,
+      "loss": 0.3343,
+      "step": 9918
+    },
+    {
+      "epoch": 2.494718309859155,
+      "grad_norm": 0.29827818274497986,
+      "learning_rate": 8.395552051133182e-07,
+      "loss": 0.3185,
+      "step": 9919
+    },
+    {
+      "epoch": 2.494969818913481,
+      "grad_norm": 0.3134204149246216,
+      "learning_rate": 8.387438044078439e-07,
+      "loss": 0.3283,
+      "step": 9920
+    },
+    {
+      "epoch": 2.4952213279678066,
+      "grad_norm": 0.2914014458656311,
+      "learning_rate": 8.379327600879306e-07,
+      "loss": 0.3361,
+      "step": 9921
+    },
+    {
+      "epoch": 2.495472837022133,
+      "grad_norm": 0.29312896728515625,
+      "learning_rate": 8.371220722230378e-07,
+      "loss": 0.3204,
+      "step": 9922
+    },
+    {
+      "epoch": 2.4957243460764587,
+      "grad_norm": 0.29510068893432617,
+      "learning_rate": 8.363117408825972e-07,
+      "loss": 0.3093,
+      "step": 9923
+    },
+    {
+      "epoch": 2.4959758551307845,
+      "grad_norm": 0.29542407393455505,
+      "learning_rate": 8.355017661360077e-07,
+      "loss": 0.3169,
+      "step": 9924
+    },
+    {
+      "epoch": 2.4962273641851107,
+      "grad_norm": 0.3483355939388275,
+      "learning_rate": 8.346921480526393e-07,
+      "loss": 0.3323,
+      "step": 9925
+    },
+    {
+      "epoch": 2.4964788732394365,
+      "grad_norm": 0.302891343832016,
+      "learning_rate": 8.338828867018295e-07,
+      "loss": 0.3261,
+      "step": 9926
+    },
+    {
+      "epoch": 2.4967303822937628,
+      "grad_norm": 0.28868502378463745,
+      "learning_rate": 8.330739821528888e-07,
+      "loss": 0.3307,
+      "step": 9927
+    },
+    {
+      "epoch": 2.4969818913480886,
+      "grad_norm": 0.2844146192073822,
+      "learning_rate": 8.322654344750914e-07,
+      "loss": 0.3226,
+      "step": 9928
+    },
+    {
+      "epoch": 2.4972334004024144,
+      "grad_norm": 0.30544909834861755,
+      "learning_rate": 8.314572437376883e-07,
+      "loss": 0.3061,
+      "step": 9929
+    },
+    {
+      "epoch": 2.4974849094567406,
+      "grad_norm": 0.2997104525566101,
+      "learning_rate": 8.306494100098922e-07,
+      "loss": 0.3298,
+      "step": 9930
+    },
+    {
+      "epoch": 2.4977364185110664,
+      "grad_norm": 0.311212956905365,
+      "learning_rate": 8.298419333608909e-07,
+      "loss": 0.3136,
+      "step": 9931
+    },
+    {
+      "epoch": 2.4979879275653922,
+      "grad_norm": 0.3107351064682007,
+      "learning_rate": 8.290348138598408e-07,
+      "loss": 0.3482,
+      "step": 9932
+    },
+    {
+      "epoch": 2.4982394366197185,
+      "grad_norm": 0.3184583783149719,
+      "learning_rate": 8.282280515758639e-07,
+      "loss": 0.321,
+      "step": 9933
+    },
+    {
+      "epoch": 2.4984909456740443,
+      "grad_norm": 0.3004460036754608,
+      "learning_rate": 8.274216465780577e-07,
+      "loss": 0.3046,
+      "step": 9934
+    },
+    {
+      "epoch": 2.49874245472837,
+      "grad_norm": 0.26762136816978455,
+      "learning_rate": 8.266155989354823e-07,
+      "loss": 0.3174,
+      "step": 9935
+    },
+    {
+      "epoch": 2.4989939637826963,
+      "grad_norm": 0.31421777606010437,
+      "learning_rate": 8.258099087171734e-07,
+      "loss": 0.2931,
+      "step": 9936
+    },
+    {
+      "epoch": 2.499245472837022,
+      "grad_norm": 0.311367005109787,
+      "learning_rate": 8.25004575992131e-07,
+      "loss": 0.3506,
+      "step": 9937
+    },
+    {
+      "epoch": 2.499496981891348,
+      "grad_norm": 0.28332027792930603,
+      "learning_rate": 8.241996008293295e-07,
+      "loss": 0.307,
+      "step": 9938
+    },
+    {
+      "epoch": 2.499748490945674,
+      "grad_norm": 0.2985660433769226,
+      "learning_rate": 8.233949832977067e-07,
+      "loss": 0.3112,
+      "step": 9939
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.30060288310050964,
+      "learning_rate": 8.225907234661767e-07,
+      "loss": 0.344,
+      "step": 9940
+    },
+    {
+      "epoch": 2.500251509054326,
+      "grad_norm": 0.2916584610939026,
+      "learning_rate": 8.217868214036156e-07,
+      "loss": 0.3283,
+      "step": 9941
+    },
+    {
+      "epoch": 2.500503018108652,
+      "grad_norm": 0.3100978434085846,
+      "learning_rate": 8.20983277178875e-07,
+      "loss": 0.3318,
+      "step": 9942
+    },
+    {
+      "epoch": 2.500754527162978,
+      "grad_norm": 0.27890169620513916,
+      "learning_rate": 8.201800908607738e-07,
+      "loss": 0.3292,
+      "step": 9943
+    },
+    {
+      "epoch": 2.5010060362173037,
+      "grad_norm": 0.29491671919822693,
+      "learning_rate": 8.193772625180974e-07,
+      "loss": 0.3101,
+      "step": 9944
+    },
+    {
+      "epoch": 2.50125754527163,
+      "grad_norm": 0.2769851088523865,
+      "learning_rate": 8.185747922196063e-07,
+      "loss": 0.3225,
+      "step": 9945
+    },
+    {
+      "epoch": 2.5015090543259557,
+      "grad_norm": 0.2963421046733856,
+      "learning_rate": 8.177726800340235e-07,
+      "loss": 0.318,
+      "step": 9946
+    },
+    {
+      "epoch": 2.501760563380282,
+      "grad_norm": 0.2658748924732208,
+      "learning_rate": 8.169709260300485e-07,
+      "loss": 0.319,
+      "step": 9947
+    },
+    {
+      "epoch": 2.5020120724346078,
+      "grad_norm": 0.2909032106399536,
+      "learning_rate": 8.161695302763434e-07,
+      "loss": 0.3245,
+      "step": 9948
+    },
+    {
+      "epoch": 2.5022635814889336,
+      "grad_norm": 0.28727254271507263,
+      "learning_rate": 8.15368492841545e-07,
+      "loss": 0.3029,
+      "step": 9949
+    },
+    {
+      "epoch": 2.50251509054326,
+      "grad_norm": 0.29924967885017395,
+      "learning_rate": 8.145678137942553e-07,
+      "loss": 0.3283,
+      "step": 9950
+    },
+    {
+      "epoch": 2.5027665995975856,
+      "grad_norm": 0.28090178966522217,
+      "learning_rate": 8.137674932030499e-07,
+      "loss": 0.3244,
+      "step": 9951
+    },
+    {
+      "epoch": 2.5030181086519114,
+      "grad_norm": 0.29318609833717346,
+      "learning_rate": 8.129675311364682e-07,
+      "loss": 0.334,
+      "step": 9952
+    },
+    {
+      "epoch": 2.5032696177062377,
+      "grad_norm": 0.28759893774986267,
+      "learning_rate": 8.121679276630235e-07,
+      "loss": 0.3267,
+      "step": 9953
+    },
+    {
+      "epoch": 2.5035211267605635,
+      "grad_norm": 0.3086696267127991,
+      "learning_rate": 8.113686828511974e-07,
+      "loss": 0.3411,
+      "step": 9954
+    },
+    {
+      "epoch": 2.5037726358148893,
+      "grad_norm": 0.298218697309494,
+      "learning_rate": 8.105697967694393e-07,
+      "loss": 0.3243,
+      "step": 9955
+    },
+    {
+      "epoch": 2.5040241448692155,
+      "grad_norm": 0.31472504138946533,
+      "learning_rate": 8.097712694861698e-07,
+      "loss": 0.3349,
+      "step": 9956
+    },
+    {
+      "epoch": 2.5042756539235413,
+      "grad_norm": 0.3033403754234314,
+      "learning_rate": 8.08973101069776e-07,
+      "loss": 0.3157,
+      "step": 9957
+    },
+    {
+      "epoch": 2.504527162977867,
+      "grad_norm": 0.25785255432128906,
+      "learning_rate": 8.081752915886182e-07,
+      "loss": 0.3112,
+      "step": 9958
+    },
+    {
+      "epoch": 2.5047786720321934,
+      "grad_norm": 0.29093462228775024,
+      "learning_rate": 8.073778411110216e-07,
+      "loss": 0.33,
+      "step": 9959
+    },
+    {
+      "epoch": 2.505030181086519,
+      "grad_norm": 0.28467297554016113,
+      "learning_rate": 8.065807497052852e-07,
+      "loss": 0.3359,
+      "step": 9960
+    },
+    {
+      "epoch": 2.505281690140845,
+      "grad_norm": 0.28814148902893066,
+      "learning_rate": 8.057840174396725e-07,
+      "loss": 0.3229,
+      "step": 9961
+    },
+    {
+      "epoch": 2.5055331991951713,
+      "grad_norm": 0.301241934299469,
+      "learning_rate": 8.049876443824212e-07,
+      "loss": 0.3187,
+      "step": 9962
+    },
+    {
+      "epoch": 2.505784708249497,
+      "grad_norm": 0.2975336015224457,
+      "learning_rate": 8.041916306017322e-07,
+      "loss": 0.3204,
+      "step": 9963
+    },
+    {
+      "epoch": 2.506036217303823,
+      "grad_norm": 0.30471301078796387,
+      "learning_rate": 8.033959761657817e-07,
+      "loss": 0.3381,
+      "step": 9964
+    },
+    {
+      "epoch": 2.506287726358149,
+      "grad_norm": 0.2744012176990509,
+      "learning_rate": 8.026006811427134e-07,
+      "loss": 0.3128,
+      "step": 9965
+    },
+    {
+      "epoch": 2.506539235412475,
+      "grad_norm": 0.2855711877346039,
+      "learning_rate": 8.018057456006362e-07,
+      "loss": 0.32,
+      "step": 9966
+    },
+    {
+      "epoch": 2.5067907444668007,
+      "grad_norm": 0.28882476687431335,
+      "learning_rate": 8.010111696076344e-07,
+      "loss": 0.3253,
+      "step": 9967
+    },
+    {
+      "epoch": 2.507042253521127,
+      "grad_norm": 0.29218101501464844,
+      "learning_rate": 8.002169532317566e-07,
+      "loss": 0.3323,
+      "step": 9968
+    },
+    {
+      "epoch": 2.5072937625754528,
+      "grad_norm": 0.3086097836494446,
+      "learning_rate": 7.994230965410232e-07,
+      "loss": 0.3173,
+      "step": 9969
+    },
+    {
+      "epoch": 2.5075452716297786,
+      "grad_norm": 0.2888603210449219,
+      "learning_rate": 7.986295996034221e-07,
+      "loss": 0.3351,
+      "step": 9970
+    },
+    {
+      "epoch": 2.507796780684105,
+      "grad_norm": 0.27783140540122986,
+      "learning_rate": 7.978364624869134e-07,
+      "loss": 0.3177,
+      "step": 9971
+    },
+    {
+      "epoch": 2.5080482897384306,
+      "grad_norm": 0.2933598756790161,
+      "learning_rate": 7.970436852594221e-07,
+      "loss": 0.3139,
+      "step": 9972
+    },
+    {
+      "epoch": 2.5082997987927564,
+      "grad_norm": 0.2918246388435364,
+      "learning_rate": 7.962512679888462e-07,
+      "loss": 0.3066,
+      "step": 9973
+    },
+    {
+      "epoch": 2.5085513078470827,
+      "grad_norm": 0.28106623888015747,
+      "learning_rate": 7.954592107430498e-07,
+      "loss": 0.3278,
+      "step": 9974
+    },
+    {
+      "epoch": 2.5088028169014085,
+      "grad_norm": 0.3085310459136963,
+      "learning_rate": 7.946675135898679e-07,
+      "loss": 0.2984,
+      "step": 9975
+    },
+    {
+      "epoch": 2.5090543259557343,
+      "grad_norm": 0.2928948402404785,
+      "learning_rate": 7.938761765971065e-07,
+      "loss": 0.3113,
+      "step": 9976
+    },
+    {
+      "epoch": 2.5093058350100605,
+      "grad_norm": 0.29120776057243347,
+      "learning_rate": 7.930851998325362e-07,
+      "loss": 0.3081,
+      "step": 9977
+    },
+    {
+      "epoch": 2.5095573440643864,
+      "grad_norm": 0.289899080991745,
+      "learning_rate": 7.922945833639012e-07,
+      "loss": 0.3236,
+      "step": 9978
+    },
+    {
+      "epoch": 2.509808853118712,
+      "grad_norm": 0.27034732699394226,
+      "learning_rate": 7.915043272589106e-07,
+      "loss": 0.3194,
+      "step": 9979
+    },
+    {
+      "epoch": 2.5100603621730384,
+      "grad_norm": 0.2954181730747223,
+      "learning_rate": 7.907144315852472e-07,
+      "loss": 0.3179,
+      "step": 9980
+    },
+    {
+      "epoch": 2.510311871227364,
+      "grad_norm": 0.2821013033390045,
+      "learning_rate": 7.899248964105583e-07,
+      "loss": 0.3298,
+      "step": 9981
+    },
+    {
+      "epoch": 2.51056338028169,
+      "grad_norm": 0.2877882122993469,
+      "learning_rate": 7.891357218024653e-07,
+      "loss": 0.3265,
+      "step": 9982
+    },
+    {
+      "epoch": 2.5108148893360163,
+      "grad_norm": 0.32784876227378845,
+      "learning_rate": 7.883469078285533e-07,
+      "loss": 0.3191,
+      "step": 9983
+    },
+    {
+      "epoch": 2.511066398390342,
+      "grad_norm": 0.28723639249801636,
+      "learning_rate": 7.875584545563819e-07,
+      "loss": 0.346,
+      "step": 9984
+    },
+    {
+      "epoch": 2.511317907444668,
+      "grad_norm": 0.2926671504974365,
+      "learning_rate": 7.867703620534744e-07,
+      "loss": 0.2905,
+      "step": 9985
+    },
+    {
+      "epoch": 2.511569416498994,
+      "grad_norm": 0.2825871706008911,
+      "learning_rate": 7.85982630387328e-07,
+      "loss": 0.3087,
+      "step": 9986
+    },
+    {
+      "epoch": 2.51182092555332,
+      "grad_norm": 0.28117263317108154,
+      "learning_rate": 7.851952596254076e-07,
+      "loss": 0.3221,
+      "step": 9987
+    },
+    {
+      "epoch": 2.5120724346076457,
+      "grad_norm": 0.2999492883682251,
+      "learning_rate": 7.844082498351441e-07,
+      "loss": 0.3078,
+      "step": 9988
+    },
+    {
+      "epoch": 2.512323943661972,
+      "grad_norm": 0.2990661859512329,
+      "learning_rate": 7.836216010839426e-07,
+      "loss": 0.3162,
+      "step": 9989
+    },
+    {
+      "epoch": 2.512575452716298,
+      "grad_norm": 0.30873674154281616,
+      "learning_rate": 7.82835313439172e-07,
+      "loss": 0.312,
+      "step": 9990
+    },
+    {
+      "epoch": 2.5128269617706236,
+      "grad_norm": 0.300627201795578,
+      "learning_rate": 7.820493869681761e-07,
+      "loss": 0.318,
+      "step": 9991
+    },
+    {
+      "epoch": 2.51307847082495,
+      "grad_norm": 0.30525755882263184,
+      "learning_rate": 7.812638217382612e-07,
+      "loss": 0.3613,
+      "step": 9992
+    },
+    {
+      "epoch": 2.5133299798792756,
+      "grad_norm": 0.30198240280151367,
+      "learning_rate": 7.804786178167085e-07,
+      "loss": 0.3029,
+      "step": 9993
+    },
+    {
+      "epoch": 2.5135814889336014,
+      "grad_norm": 0.3166106343269348,
+      "learning_rate": 7.796937752707639e-07,
+      "loss": 0.3309,
+      "step": 9994
+    },
+    {
+      "epoch": 2.5138329979879277,
+      "grad_norm": 0.3065146207809448,
+      "learning_rate": 7.789092941676468e-07,
+      "loss": 0.345,
+      "step": 9995
+    },
+    {
+      "epoch": 2.5140845070422535,
+      "grad_norm": 0.3031245470046997,
+      "learning_rate": 7.781251745745405e-07,
+      "loss": 0.3172,
+      "step": 9996
+    },
+    {
+      "epoch": 2.5143360160965793,
+      "grad_norm": 0.29413849115371704,
+      "learning_rate": 7.773414165586007e-07,
+      "loss": 0.322,
+      "step": 9997
+    },
+    {
+      "epoch": 2.5145875251509056,
+      "grad_norm": 0.29748326539993286,
+      "learning_rate": 7.765580201869527e-07,
+      "loss": 0.2931,
+      "step": 9998
+    },
+    {
+      "epoch": 2.5148390342052314,
+      "grad_norm": 0.3018585443496704,
+      "learning_rate": 7.757749855266878e-07,
+      "loss": 0.3339,
+      "step": 9999
+    },
+    {
+      "epoch": 2.515090543259557,
+      "grad_norm": 0.30707988142967224,
+      "learning_rate": 7.749923126448694e-07,
+      "loss": 0.3281,
+      "step": 10000
+    },
+    {
+      "epoch": 2.5153420523138834,
+      "grad_norm": 0.3032652735710144,
+      "learning_rate": 7.742100016085269e-07,
+      "loss": 0.3183,
+      "step": 10001
+    },
+    {
+      "epoch": 2.515593561368209,
+      "grad_norm": 0.30448129773139954,
+      "learning_rate": 7.734280524846627e-07,
+      "loss": 0.298,
+      "step": 10002
+    },
+    {
+      "epoch": 2.515845070422535,
+      "grad_norm": 0.3087362051010132,
+      "learning_rate": 7.726464653402432e-07,
+      "loss": 0.3155,
+      "step": 10003
+    },
+    {
+      "epoch": 2.5160965794768613,
+      "grad_norm": 0.2881805896759033,
+      "learning_rate": 7.718652402422088e-07,
+      "loss": 0.3456,
+      "step": 10004
+    },
+    {
+      "epoch": 2.516348088531187,
+      "grad_norm": 0.301011323928833,
+      "learning_rate": 7.710843772574644e-07,
+      "loss": 0.3288,
+      "step": 10005
+    },
+    {
+      "epoch": 2.516599597585513,
+      "grad_norm": 0.31224364042282104,
+      "learning_rate": 7.70303876452888e-07,
+      "loss": 0.3128,
+      "step": 10006
+    },
+    {
+      "epoch": 2.516851106639839,
+      "grad_norm": 0.2763179838657379,
+      "learning_rate": 7.695237378953224e-07,
+      "loss": 0.3003,
+      "step": 10007
+    },
+    {
+      "epoch": 2.517102615694165,
+      "grad_norm": 0.3021027445793152,
+      "learning_rate": 7.687439616515846e-07,
+      "loss": 0.3198,
+      "step": 10008
+    },
+    {
+      "epoch": 2.5173541247484907,
+      "grad_norm": 0.2829072177410126,
+      "learning_rate": 7.67964547788454e-07,
+      "loss": 0.2887,
+      "step": 10009
+    },
+    {
+      "epoch": 2.517605633802817,
+      "grad_norm": 0.31430134177207947,
+      "learning_rate": 7.67185496372686e-07,
+      "loss": 0.3293,
+      "step": 10010
+    },
+    {
+      "epoch": 2.517857142857143,
+      "grad_norm": 0.30119824409484863,
+      "learning_rate": 7.664068074709985e-07,
+      "loss": 0.3175,
+      "step": 10011
+    },
+    {
+      "epoch": 2.5181086519114686,
+      "grad_norm": 0.2803148329257965,
+      "learning_rate": 7.656284811500842e-07,
+      "loss": 0.3335,
+      "step": 10012
+    },
+    {
+      "epoch": 2.518360160965795,
+      "grad_norm": 0.28627124428749084,
+      "learning_rate": 7.648505174765986e-07,
+      "loss": 0.3127,
+      "step": 10013
+    },
+    {
+      "epoch": 2.5186116700201207,
+      "grad_norm": 0.2865979075431824,
+      "learning_rate": 7.640729165171723e-07,
+      "loss": 0.2974,
+      "step": 10014
+    },
+    {
+      "epoch": 2.5188631790744465,
+      "grad_norm": 0.2760888636112213,
+      "learning_rate": 7.632956783383999e-07,
+      "loss": 0.3136,
+      "step": 10015
+    },
+    {
+      "epoch": 2.5191146881287727,
+      "grad_norm": 0.2837379276752472,
+      "learning_rate": 7.625188030068492e-07,
+      "loss": 0.324,
+      "step": 10016
+    },
+    {
+      "epoch": 2.5193661971830985,
+      "grad_norm": 0.3180212676525116,
+      "learning_rate": 7.617422905890521e-07,
+      "loss": 0.3085,
+      "step": 10017
+    },
+    {
+      "epoch": 2.5196177062374243,
+      "grad_norm": 0.29850107431411743,
+      "learning_rate": 7.609661411515146e-07,
+      "loss": 0.3149,
+      "step": 10018
+    },
+    {
+      "epoch": 2.5198692152917506,
+      "grad_norm": 0.32031044363975525,
+      "learning_rate": 7.601903547607064e-07,
+      "loss": 0.3302,
+      "step": 10019
+    },
+    {
+      "epoch": 2.5201207243460764,
+      "grad_norm": 0.30878177285194397,
+      "learning_rate": 7.594149314830717e-07,
+      "loss": 0.3003,
+      "step": 10020
+    },
+    {
+      "epoch": 2.520372233400402,
+      "grad_norm": 0.2950962781906128,
+      "learning_rate": 7.586398713850179e-07,
+      "loss": 0.327,
+      "step": 10021
+    },
+    {
+      "epoch": 2.5206237424547284,
+      "grad_norm": 0.29020336270332336,
+      "learning_rate": 7.578651745329263e-07,
+      "loss": 0.3268,
+      "step": 10022
+    },
+    {
+      "epoch": 2.5208752515090542,
+      "grad_norm": 0.3015232980251312,
+      "learning_rate": 7.570908409931427e-07,
+      "loss": 0.3087,
+      "step": 10023
+    },
+    {
+      "epoch": 2.52112676056338,
+      "grad_norm": 0.29202380776405334,
+      "learning_rate": 7.563168708319857e-07,
+      "loss": 0.3301,
+      "step": 10024
+    },
+    {
+      "epoch": 2.5213782696177063,
+      "grad_norm": 0.2775900065898895,
+      "learning_rate": 7.555432641157396e-07,
+      "loss": 0.3149,
+      "step": 10025
+    },
+    {
+      "epoch": 2.521629778672032,
+      "grad_norm": 0.2708669900894165,
+      "learning_rate": 7.547700209106606e-07,
+      "loss": 0.3238,
+      "step": 10026
+    },
+    {
+      "epoch": 2.521881287726358,
+      "grad_norm": 0.30167317390441895,
+      "learning_rate": 7.539971412829705e-07,
+      "loss": 0.3001,
+      "step": 10027
+    },
+    {
+      "epoch": 2.522132796780684,
+      "grad_norm": 0.3165789246559143,
+      "learning_rate": 7.532246252988617e-07,
+      "loss": 0.3382,
+      "step": 10028
+    },
+    {
+      "epoch": 2.52238430583501,
+      "grad_norm": 0.29492056369781494,
+      "learning_rate": 7.524524730244975e-07,
+      "loss": 0.3383,
+      "step": 10029
+    },
+    {
+      "epoch": 2.5226358148893357,
+      "grad_norm": 0.2822633981704712,
+      "learning_rate": 7.516806845260055e-07,
+      "loss": 0.3003,
+      "step": 10030
+    },
+    {
+      "epoch": 2.522887323943662,
+      "grad_norm": 0.3063894212245941,
+      "learning_rate": 7.509092598694861e-07,
+      "loss": 0.3138,
+      "step": 10031
+    },
+    {
+      "epoch": 2.523138832997988,
+      "grad_norm": 0.32523077726364136,
+      "learning_rate": 7.501381991210061e-07,
+      "loss": 0.3164,
+      "step": 10032
+    },
+    {
+      "epoch": 2.5233903420523136,
+      "grad_norm": 0.28941982984542847,
+      "learning_rate": 7.493675023466024e-07,
+      "loss": 0.3347,
+      "step": 10033
+    },
+    {
+      "epoch": 2.52364185110664,
+      "grad_norm": 0.29666292667388916,
+      "learning_rate": 7.485971696122796e-07,
+      "loss": 0.324,
+      "step": 10034
+    },
+    {
+      "epoch": 2.5238933601609657,
+      "grad_norm": 0.3100225031375885,
+      "learning_rate": 7.478272009840137e-07,
+      "loss": 0.3254,
+      "step": 10035
+    },
+    {
+      "epoch": 2.524144869215292,
+      "grad_norm": 0.28848564624786377,
+      "learning_rate": 7.470575965277455e-07,
+      "loss": 0.3452,
+      "step": 10036
+    },
+    {
+      "epoch": 2.5243963782696177,
+      "grad_norm": 0.2774829864501953,
+      "learning_rate": 7.462883563093887e-07,
+      "loss": 0.3218,
+      "step": 10037
+    },
+    {
+      "epoch": 2.5246478873239435,
+      "grad_norm": 0.3099637031555176,
+      "learning_rate": 7.455194803948218e-07,
+      "loss": 0.2875,
+      "step": 10038
+    },
+    {
+      "epoch": 2.5248993963782698,
+      "grad_norm": 0.28145939111709595,
+      "learning_rate": 7.447509688498971e-07,
+      "loss": 0.3332,
+      "step": 10039
+    },
+    {
+      "epoch": 2.5251509054325956,
+      "grad_norm": 0.296622633934021,
+      "learning_rate": 7.439828217404293e-07,
+      "loss": 0.3478,
+      "step": 10040
+    },
+    {
+      "epoch": 2.5254024144869214,
+      "grad_norm": 0.27812138199806213,
+      "learning_rate": 7.432150391322079e-07,
+      "loss": 0.3302,
+      "step": 10041
+    },
+    {
+      "epoch": 2.5256539235412476,
+      "grad_norm": 0.26120641827583313,
+      "learning_rate": 7.424476210909893e-07,
+      "loss": 0.2886,
+      "step": 10042
+    },
+    {
+      "epoch": 2.5259054325955734,
+      "grad_norm": 0.29901203513145447,
+      "learning_rate": 7.41680567682495e-07,
+      "loss": 0.3021,
+      "step": 10043
+    },
+    {
+      "epoch": 2.5261569416498992,
+      "grad_norm": 0.28617119789123535,
+      "learning_rate": 7.409138789724213e-07,
+      "loss": 0.321,
+      "step": 10044
+    },
+    {
+      "epoch": 2.5264084507042255,
+      "grad_norm": 0.28953617811203003,
+      "learning_rate": 7.401475550264286e-07,
+      "loss": 0.346,
+      "step": 10045
+    },
+    {
+      "epoch": 2.5266599597585513,
+      "grad_norm": 0.28805407881736755,
+      "learning_rate": 7.393815959101491e-07,
+      "loss": 0.3206,
+      "step": 10046
+    },
+    {
+      "epoch": 2.5269114688128775,
+      "grad_norm": 0.29841816425323486,
+      "learning_rate": 7.386160016891802e-07,
+      "loss": 0.3242,
+      "step": 10047
+    },
+    {
+      "epoch": 2.5271629778672033,
+      "grad_norm": 0.2983440160751343,
+      "learning_rate": 7.378507724290929e-07,
+      "loss": 0.3347,
+      "step": 10048
+    },
+    {
+      "epoch": 2.527414486921529,
+      "grad_norm": 0.2765296697616577,
+      "learning_rate": 7.370859081954219e-07,
+      "loss": 0.3093,
+      "step": 10049
+    },
+    {
+      "epoch": 2.5276659959758554,
+      "grad_norm": 0.2982845902442932,
+      "learning_rate": 7.363214090536752e-07,
+      "loss": 0.3454,
+      "step": 10050
+    },
+    {
+      "epoch": 2.527917505030181,
+      "grad_norm": 0.2902871370315552,
+      "learning_rate": 7.355572750693252e-07,
+      "loss": 0.3436,
+      "step": 10051
+    },
+    {
+      "epoch": 2.528169014084507,
+      "grad_norm": 0.319418340921402,
+      "learning_rate": 7.347935063078165e-07,
+      "loss": 0.3195,
+      "step": 10052
+    },
+    {
+      "epoch": 2.5284205231388333,
+      "grad_norm": 0.2917003631591797,
+      "learning_rate": 7.34030102834562e-07,
+      "loss": 0.3125,
+      "step": 10053
+    },
+    {
+      "epoch": 2.528672032193159,
+      "grad_norm": 0.30017781257629395,
+      "learning_rate": 7.332670647149398e-07,
+      "loss": 0.3478,
+      "step": 10054
+    },
+    {
+      "epoch": 2.528923541247485,
+      "grad_norm": 0.3094007968902588,
+      "learning_rate": 7.325043920143027e-07,
+      "loss": 0.3284,
+      "step": 10055
+    },
+    {
+      "epoch": 2.529175050301811,
+      "grad_norm": 0.3049699366092682,
+      "learning_rate": 7.317420847979656e-07,
+      "loss": 0.3051,
+      "step": 10056
+    },
+    {
+      "epoch": 2.529426559356137,
+      "grad_norm": 0.29030466079711914,
+      "learning_rate": 7.30980143131218e-07,
+      "loss": 0.3203,
+      "step": 10057
+    },
+    {
+      "epoch": 2.5296780684104627,
+      "grad_norm": 0.2822028398513794,
+      "learning_rate": 7.302185670793132e-07,
+      "loss": 0.3209,
+      "step": 10058
+    },
+    {
+      "epoch": 2.529929577464789,
+      "grad_norm": 0.28473353385925293,
+      "learning_rate": 7.294573567074776e-07,
+      "loss": 0.3205,
+      "step": 10059
+    },
+    {
+      "epoch": 2.5301810865191148,
+      "grad_norm": 0.2997729778289795,
+      "learning_rate": 7.286965120809014e-07,
+      "loss": 0.3134,
+      "step": 10060
+    },
+    {
+      "epoch": 2.5304325955734406,
+      "grad_norm": 0.2938189208507538,
+      "learning_rate": 7.279360332647495e-07,
+      "loss": 0.313,
+      "step": 10061
+    },
+    {
+      "epoch": 2.530684104627767,
+      "grad_norm": 0.291318416595459,
+      "learning_rate": 7.271759203241485e-07,
+      "loss": 0.2945,
+      "step": 10062
+    },
+    {
+      "epoch": 2.5309356136820926,
+      "grad_norm": 0.29242825508117676,
+      "learning_rate": 7.264161733241998e-07,
+      "loss": 0.2912,
+      "step": 10063
+    },
+    {
+      "epoch": 2.5311871227364184,
+      "grad_norm": 0.2778705358505249,
+      "learning_rate": 7.256567923299712e-07,
+      "loss": 0.3547,
+      "step": 10064
+    },
+    {
+      "epoch": 2.5314386317907447,
+      "grad_norm": 0.29071563482284546,
+      "learning_rate": 7.248977774064975e-07,
+      "loss": 0.3357,
+      "step": 10065
+    },
+    {
+      "epoch": 2.5316901408450705,
+      "grad_norm": 0.2635411322116852,
+      "learning_rate": 7.241391286187849e-07,
+      "loss": 0.3257,
+      "step": 10066
+    },
+    {
+      "epoch": 2.5319416498993963,
+      "grad_norm": 0.2990947961807251,
+      "learning_rate": 7.233808460318054e-07,
+      "loss": 0.3175,
+      "step": 10067
+    },
+    {
+      "epoch": 2.5321931589537225,
+      "grad_norm": 0.2650429904460907,
+      "learning_rate": 7.226229297105031e-07,
+      "loss": 0.3322,
+      "step": 10068
+    },
+    {
+      "epoch": 2.5324446680080483,
+      "grad_norm": 0.2894965410232544,
+      "learning_rate": 7.218653797197861e-07,
+      "loss": 0.307,
+      "step": 10069
+    },
+    {
+      "epoch": 2.532696177062374,
+      "grad_norm": 0.26556655764579773,
+      "learning_rate": 7.21108196124537e-07,
+      "loss": 0.3372,
+      "step": 10070
+    },
+    {
+      "epoch": 2.5329476861167004,
+      "grad_norm": 0.3001581132411957,
+      "learning_rate": 7.20351378989601e-07,
+      "loss": 0.3338,
+      "step": 10071
+    },
+    {
+      "epoch": 2.533199195171026,
+      "grad_norm": 0.3141966164112091,
+      "learning_rate": 7.195949283797959e-07,
+      "loss": 0.3287,
+      "step": 10072
+    },
+    {
+      "epoch": 2.533450704225352,
+      "grad_norm": 0.2945497930049896,
+      "learning_rate": 7.188388443599081e-07,
+      "loss": 0.3309,
+      "step": 10073
+    },
+    {
+      "epoch": 2.5337022132796783,
+      "grad_norm": 0.3076508641242981,
+      "learning_rate": 7.180831269946898e-07,
+      "loss": 0.335,
+      "step": 10074
+    },
+    {
+      "epoch": 2.533953722334004,
+      "grad_norm": 0.2986814081668854,
+      "learning_rate": 7.173277763488646e-07,
+      "loss": 0.3138,
+      "step": 10075
+    },
+    {
+      "epoch": 2.53420523138833,
+      "grad_norm": 0.2792844772338867,
+      "learning_rate": 7.165727924871224e-07,
+      "loss": 0.3147,
+      "step": 10076
+    },
+    {
+      "epoch": 2.534456740442656,
+      "grad_norm": 0.3163931667804718,
+      "learning_rate": 7.158181754741239e-07,
+      "loss": 0.3245,
+      "step": 10077
+    },
+    {
+      "epoch": 2.534708249496982,
+      "grad_norm": 0.2911807894706726,
+      "learning_rate": 7.150639253744967e-07,
+      "loss": 0.3191,
+      "step": 10078
+    },
+    {
+      "epoch": 2.5349597585513077,
+      "grad_norm": 0.284397691488266,
+      "learning_rate": 7.143100422528382e-07,
+      "loss": 0.3186,
+      "step": 10079
+    },
+    {
+      "epoch": 2.535211267605634,
+      "grad_norm": 0.2803077697753906,
+      "learning_rate": 7.135565261737126e-07,
+      "loss": 0.3276,
+      "step": 10080
+    },
+    {
+      "epoch": 2.53546277665996,
+      "grad_norm": 0.27358850836753845,
+      "learning_rate": 7.128033772016557e-07,
+      "loss": 0.3073,
+      "step": 10081
+    },
+    {
+      "epoch": 2.5357142857142856,
+      "grad_norm": 0.30038154125213623,
+      "learning_rate": 7.120505954011681e-07,
+      "loss": 0.3076,
+      "step": 10082
+    },
+    {
+      "epoch": 2.535965794768612,
+      "grad_norm": 0.3066074848175049,
+      "learning_rate": 7.112981808367214e-07,
+      "loss": 0.3354,
+      "step": 10083
+    },
+    {
+      "epoch": 2.5362173038229376,
+      "grad_norm": 0.30170679092407227,
+      "learning_rate": 7.105461335727564e-07,
+      "loss": 0.3102,
+      "step": 10084
+    },
+    {
+      "epoch": 2.5364688128772634,
+      "grad_norm": 0.3071909546852112,
+      "learning_rate": 7.097944536736795e-07,
+      "loss": 0.3112,
+      "step": 10085
+    },
+    {
+      "epoch": 2.5367203219315897,
+      "grad_norm": 0.2948451638221741,
+      "learning_rate": 7.09043141203869e-07,
+      "loss": 0.3097,
+      "step": 10086
+    },
+    {
+      "epoch": 2.5369718309859155,
+      "grad_norm": 0.31133005023002625,
+      "learning_rate": 7.082921962276684e-07,
+      "loss": 0.3196,
+      "step": 10087
+    },
+    {
+      "epoch": 2.5372233400402413,
+      "grad_norm": 0.29081660509109497,
+      "learning_rate": 7.075416188093936e-07,
+      "loss": 0.2977,
+      "step": 10088
+    },
+    {
+      "epoch": 2.5374748490945676,
+      "grad_norm": 0.2829488515853882,
+      "learning_rate": 7.067914090133244e-07,
+      "loss": 0.3158,
+      "step": 10089
+    },
+    {
+      "epoch": 2.5377263581488934,
+      "grad_norm": 0.2930876612663269,
+      "learning_rate": 7.060415669037135e-07,
+      "loss": 0.3419,
+      "step": 10090
+    },
+    {
+      "epoch": 2.537977867203219,
+      "grad_norm": 0.32287168502807617,
+      "learning_rate": 7.052920925447792e-07,
+      "loss": 0.3263,
+      "step": 10091
+    },
+    {
+      "epoch": 2.5382293762575454,
+      "grad_norm": 0.31697362661361694,
+      "learning_rate": 7.045429860007102e-07,
+      "loss": 0.2973,
+      "step": 10092
+    },
+    {
+      "epoch": 2.538480885311871,
+      "grad_norm": 0.28679776191711426,
+      "learning_rate": 7.037942473356607e-07,
+      "loss": 0.3167,
+      "step": 10093
+    },
+    {
+      "epoch": 2.538732394366197,
+      "grad_norm": 0.3053102195262909,
+      "learning_rate": 7.030458766137577e-07,
+      "loss": 0.3345,
+      "step": 10094
+    },
+    {
+      "epoch": 2.5389839034205233,
+      "grad_norm": 0.2991425395011902,
+      "learning_rate": 7.022978738990943e-07,
+      "loss": 0.3406,
+      "step": 10095
+    },
+    {
+      "epoch": 2.539235412474849,
+      "grad_norm": 0.3642655909061432,
+      "learning_rate": 7.015502392557305e-07,
+      "loss": 0.3412,
+      "step": 10096
+    },
+    {
+      "epoch": 2.539486921529175,
+      "grad_norm": 0.31416499614715576,
+      "learning_rate": 7.00802972747699e-07,
+      "loss": 0.3052,
+      "step": 10097
+    },
+    {
+      "epoch": 2.539738430583501,
+      "grad_norm": 0.296009361743927,
+      "learning_rate": 7.000560744389962e-07,
+      "loss": 0.3015,
+      "step": 10098
+    },
+    {
+      "epoch": 2.539989939637827,
+      "grad_norm": 0.31365057826042175,
+      "learning_rate": 6.99309544393591e-07,
+      "loss": 0.3134,
+      "step": 10099
+    },
+    {
+      "epoch": 2.5402414486921527,
+      "grad_norm": 0.2986956536769867,
+      "learning_rate": 6.985633826754173e-07,
+      "loss": 0.3184,
+      "step": 10100
+    },
+    {
+      "epoch": 2.540492957746479,
+      "grad_norm": 0.2894130349159241,
+      "learning_rate": 6.978175893483812e-07,
+      "loss": 0.312,
+      "step": 10101
+    },
+    {
+      "epoch": 2.540744466800805,
+      "grad_norm": 0.29604774713516235,
+      "learning_rate": 6.970721644763534e-07,
+      "loss": 0.3198,
+      "step": 10102
+    },
+    {
+      "epoch": 2.5409959758551306,
+      "grad_norm": 0.28080493211746216,
+      "learning_rate": 6.963271081231765e-07,
+      "loss": 0.3098,
+      "step": 10103
+    },
+    {
+      "epoch": 2.541247484909457,
+      "grad_norm": 0.2949555218219757,
+      "learning_rate": 6.955824203526585e-07,
+      "loss": 0.3102,
+      "step": 10104
+    },
+    {
+      "epoch": 2.5414989939637826,
+      "grad_norm": 0.304453581571579,
+      "learning_rate": 6.948381012285771e-07,
+      "loss": 0.3332,
+      "step": 10105
+    },
+    {
+      "epoch": 2.5417505030181085,
+      "grad_norm": 0.3298734128475189,
+      "learning_rate": 6.940941508146809e-07,
+      "loss": 0.3199,
+      "step": 10106
+    },
+    {
+      "epoch": 2.5420020120724347,
+      "grad_norm": 0.2873101830482483,
+      "learning_rate": 6.933505691746816e-07,
+      "loss": 0.3262,
+      "step": 10107
+    },
+    {
+      "epoch": 2.5422535211267605,
+      "grad_norm": 0.2953130900859833,
+      "learning_rate": 6.926073563722652e-07,
+      "loss": 0.3264,
+      "step": 10108
+    },
+    {
+      "epoch": 2.5425050301810863,
+      "grad_norm": 0.3112882971763611,
+      "learning_rate": 6.918645124710805e-07,
+      "loss": 0.3349,
+      "step": 10109
+    },
+    {
+      "epoch": 2.5427565392354126,
+      "grad_norm": 0.28245386481285095,
+      "learning_rate": 6.911220375347499e-07,
+      "loss": 0.2928,
+      "step": 10110
+    },
+    {
+      "epoch": 2.5430080482897384,
+      "grad_norm": 0.30176132917404175,
+      "learning_rate": 6.903799316268595e-07,
+      "loss": 0.3107,
+      "step": 10111
+    },
+    {
+      "epoch": 2.543259557344064,
+      "grad_norm": 0.29093387722969055,
+      "learning_rate": 6.896381948109682e-07,
+      "loss": 0.2997,
+      "step": 10112
+    },
+    {
+      "epoch": 2.5435110663983904,
+      "grad_norm": 0.2985793650150299,
+      "learning_rate": 6.888968271505986e-07,
+      "loss": 0.317,
+      "step": 10113
+    },
+    {
+      "epoch": 2.5437625754527162,
+      "grad_norm": 0.3139978349208832,
+      "learning_rate": 6.881558287092466e-07,
+      "loss": 0.3097,
+      "step": 10114
+    },
+    {
+      "epoch": 2.544014084507042,
+      "grad_norm": 0.30869972705841064,
+      "learning_rate": 6.87415199550372e-07,
+      "loss": 0.3236,
+      "step": 10115
+    },
+    {
+      "epoch": 2.5442655935613683,
+      "grad_norm": 0.2910330593585968,
+      "learning_rate": 6.866749397374062e-07,
+      "loss": 0.303,
+      "step": 10116
+    },
+    {
+      "epoch": 2.544517102615694,
+      "grad_norm": 0.31802523136138916,
+      "learning_rate": 6.859350493337491e-07,
+      "loss": 0.3032,
+      "step": 10117
+    },
+    {
+      "epoch": 2.54476861167002,
+      "grad_norm": 0.2831653356552124,
+      "learning_rate": 6.85195528402765e-07,
+      "loss": 0.3226,
+      "step": 10118
+    },
+    {
+      "epoch": 2.545020120724346,
+      "grad_norm": 0.29345449805259705,
+      "learning_rate": 6.844563770077917e-07,
+      "loss": 0.3364,
+      "step": 10119
+    },
+    {
+      "epoch": 2.545271629778672,
+      "grad_norm": 0.30557841062545776,
+      "learning_rate": 6.837175952121305e-07,
+      "loss": 0.3013,
+      "step": 10120
+    },
+    {
+      "epoch": 2.5455231388329977,
+      "grad_norm": 0.2902613580226898,
+      "learning_rate": 6.829791830790555e-07,
+      "loss": 0.3354,
+      "step": 10121
+    },
+    {
+      "epoch": 2.545774647887324,
+      "grad_norm": 0.3239326775074005,
+      "learning_rate": 6.82241140671806e-07,
+      "loss": 0.3009,
+      "step": 10122
+    },
+    {
+      "epoch": 2.54602615694165,
+      "grad_norm": 0.30113399028778076,
+      "learning_rate": 6.815034680535915e-07,
+      "loss": 0.3348,
+      "step": 10123
+    },
+    {
+      "epoch": 2.5462776659959756,
+      "grad_norm": 0.30587083101272583,
+      "learning_rate": 6.807661652875875e-07,
+      "loss": 0.3163,
+      "step": 10124
+    },
+    {
+      "epoch": 2.546529175050302,
+      "grad_norm": 0.2883608639240265,
+      "learning_rate": 6.800292324369417e-07,
+      "loss": 0.3324,
+      "step": 10125
+    },
+    {
+      "epoch": 2.5467806841046277,
+      "grad_norm": 0.30463874340057373,
+      "learning_rate": 6.79292669564765e-07,
+      "loss": 0.3423,
+      "step": 10126
+    },
+    {
+      "epoch": 2.5470321931589535,
+      "grad_norm": 0.28268906474113464,
+      "learning_rate": 6.785564767341423e-07,
+      "loss": 0.301,
+      "step": 10127
+    },
+    {
+      "epoch": 2.5472837022132797,
+      "grad_norm": 0.29041969776153564,
+      "learning_rate": 6.778206540081211e-07,
+      "loss": 0.3189,
+      "step": 10128
+    },
+    {
+      "epoch": 2.5475352112676055,
+      "grad_norm": 0.2949490547180176,
+      "learning_rate": 6.77085201449722e-07,
+      "loss": 0.3236,
+      "step": 10129
+    },
+    {
+      "epoch": 2.5477867203219313,
+      "grad_norm": 0.31995368003845215,
+      "learning_rate": 6.763501191219319e-07,
+      "loss": 0.3262,
+      "step": 10130
+    },
+    {
+      "epoch": 2.5480382293762576,
+      "grad_norm": 0.29409512877464294,
+      "learning_rate": 6.756154070877047e-07,
+      "loss": 0.3152,
+      "step": 10131
+    },
+    {
+      "epoch": 2.5482897384305834,
+      "grad_norm": 0.3013671040534973,
+      "learning_rate": 6.748810654099652e-07,
+      "loss": 0.3187,
+      "step": 10132
+    },
+    {
+      "epoch": 2.5485412474849096,
+      "grad_norm": 0.3068297207355499,
+      "learning_rate": 6.741470941516043e-07,
+      "loss": 0.302,
+      "step": 10133
+    },
+    {
+      "epoch": 2.5487927565392354,
+      "grad_norm": 0.2889004945755005,
+      "learning_rate": 6.734134933754826e-07,
+      "loss": 0.3019,
+      "step": 10134
+    },
+    {
+      "epoch": 2.5490442655935612,
+      "grad_norm": 0.3107074499130249,
+      "learning_rate": 6.726802631444274e-07,
+      "loss": 0.3373,
+      "step": 10135
+    },
+    {
+      "epoch": 2.5492957746478875,
+      "grad_norm": 0.30381691455841064,
+      "learning_rate": 6.719474035212376e-07,
+      "loss": 0.3,
+      "step": 10136
+    },
+    {
+      "epoch": 2.5495472837022133,
+      "grad_norm": 0.29392245411872864,
+      "learning_rate": 6.712149145686748e-07,
+      "loss": 0.3328,
+      "step": 10137
+    },
+    {
+      "epoch": 2.549798792756539,
+      "grad_norm": 0.2984176576137543,
+      "learning_rate": 6.704827963494748e-07,
+      "loss": 0.2996,
+      "step": 10138
+    },
+    {
+      "epoch": 2.5500503018108653,
+      "grad_norm": 0.31401655077934265,
+      "learning_rate": 6.697510489263371e-07,
+      "loss": 0.3237,
+      "step": 10139
+    },
+    {
+      "epoch": 2.550301810865191,
+      "grad_norm": 0.2747150659561157,
+      "learning_rate": 6.69019672361933e-07,
+      "loss": 0.323,
+      "step": 10140
+    },
+    {
+      "epoch": 2.550553319919517,
+      "grad_norm": 0.29566800594329834,
+      "learning_rate": 6.682886667188987e-07,
+      "loss": 0.3063,
+      "step": 10141
+    },
+    {
+      "epoch": 2.550804828973843,
+      "grad_norm": 0.3180108964443207,
+      "learning_rate": 6.675580320598418e-07,
+      "loss": 0.3439,
+      "step": 10142
+    },
+    {
+      "epoch": 2.551056338028169,
+      "grad_norm": 0.2787119150161743,
+      "learning_rate": 6.668277684473346e-07,
+      "loss": 0.3225,
+      "step": 10143
+    },
+    {
+      "epoch": 2.551307847082495,
+      "grad_norm": 0.2893412709236145,
+      "learning_rate": 6.660978759439219e-07,
+      "loss": 0.3334,
+      "step": 10144
+    },
+    {
+      "epoch": 2.551559356136821,
+      "grad_norm": 0.30765849351882935,
+      "learning_rate": 6.653683546121126e-07,
+      "loss": 0.3181,
+      "step": 10145
+    },
+    {
+      "epoch": 2.551810865191147,
+      "grad_norm": 0.29015520215034485,
+      "learning_rate": 6.646392045143868e-07,
+      "loss": 0.327,
+      "step": 10146
+    },
+    {
+      "epoch": 2.552062374245473,
+      "grad_norm": 0.29602178931236267,
+      "learning_rate": 6.639104257131907e-07,
+      "loss": 0.3178,
+      "step": 10147
+    },
+    {
+      "epoch": 2.552313883299799,
+      "grad_norm": 0.3179301917552948,
+      "learning_rate": 6.631820182709409e-07,
+      "loss": 0.334,
+      "step": 10148
+    },
+    {
+      "epoch": 2.5525653923541247,
+      "grad_norm": 0.315472811460495,
+      "learning_rate": 6.624539822500192e-07,
+      "loss": 0.3132,
+      "step": 10149
+    },
+    {
+      "epoch": 2.552816901408451,
+      "grad_norm": 0.3242339491844177,
+      "learning_rate": 6.617263177127797e-07,
+      "loss": 0.3154,
+      "step": 10150
+    },
+    {
+      "epoch": 2.5530684104627768,
+      "grad_norm": 0.26861920952796936,
+      "learning_rate": 6.609990247215393e-07,
+      "loss": 0.3406,
+      "step": 10151
+    },
+    {
+      "epoch": 2.5533199195171026,
+      "grad_norm": 0.29673323035240173,
+      "learning_rate": 6.602721033385889e-07,
+      "loss": 0.322,
+      "step": 10152
+    },
+    {
+      "epoch": 2.553571428571429,
+      "grad_norm": 0.32738587260246277,
+      "learning_rate": 6.59545553626183e-07,
+      "loss": 0.3157,
+      "step": 10153
+    },
+    {
+      "epoch": 2.5538229376257546,
+      "grad_norm": 0.3015281558036804,
+      "learning_rate": 6.588193756465472e-07,
+      "loss": 0.3277,
+      "step": 10154
+    },
+    {
+      "epoch": 2.5540744466800804,
+      "grad_norm": 0.2911074161529541,
+      "learning_rate": 6.580935694618728e-07,
+      "loss": 0.3128,
+      "step": 10155
+    },
+    {
+      "epoch": 2.5543259557344067,
+      "grad_norm": 0.29950693249702454,
+      "learning_rate": 6.573681351343226e-07,
+      "loss": 0.3063,
+      "step": 10156
+    },
+    {
+      "epoch": 2.5545774647887325,
+      "grad_norm": 0.28311774134635925,
+      "learning_rate": 6.566430727260226e-07,
+      "loss": 0.2949,
+      "step": 10157
+    },
+    {
+      "epoch": 2.5548289738430583,
+      "grad_norm": 0.27676114439964294,
+      "learning_rate": 6.559183822990727e-07,
+      "loss": 0.3299,
+      "step": 10158
+    },
+    {
+      "epoch": 2.5550804828973845,
+      "grad_norm": 0.2759498059749603,
+      "learning_rate": 6.551940639155357e-07,
+      "loss": 0.3193,
+      "step": 10159
+    },
+    {
+      "epoch": 2.5553319919517103,
+      "grad_norm": 0.29512062668800354,
+      "learning_rate": 6.544701176374462e-07,
+      "loss": 0.3159,
+      "step": 10160
+    },
+    {
+      "epoch": 2.555583501006036,
+      "grad_norm": 0.2883027195930481,
+      "learning_rate": 6.537465435268065e-07,
+      "loss": 0.3145,
+      "step": 10161
+    },
+    {
+      "epoch": 2.5558350100603624,
+      "grad_norm": 0.2832835614681244,
+      "learning_rate": 6.530233416455845e-07,
+      "loss": 0.3234,
+      "step": 10162
+    },
+    {
+      "epoch": 2.556086519114688,
+      "grad_norm": 0.28715407848358154,
+      "learning_rate": 6.523005120557197e-07,
+      "loss": 0.3323,
+      "step": 10163
+    },
+    {
+      "epoch": 2.556338028169014,
+      "grad_norm": 0.3091478645801544,
+      "learning_rate": 6.515780548191159e-07,
+      "loss": 0.3248,
+      "step": 10164
+    },
+    {
+      "epoch": 2.5565895372233403,
+      "grad_norm": 0.2894414961338043,
+      "learning_rate": 6.508559699976486e-07,
+      "loss": 0.3113,
+      "step": 10165
+    },
+    {
+      "epoch": 2.556841046277666,
+      "grad_norm": 0.2907220721244812,
+      "learning_rate": 6.501342576531589e-07,
+      "loss": 0.3341,
+      "step": 10166
+    },
+    {
+      "epoch": 2.557092555331992,
+      "grad_norm": 0.31087636947631836,
+      "learning_rate": 6.494129178474579e-07,
+      "loss": 0.338,
+      "step": 10167
+    },
+    {
+      "epoch": 2.557344064386318,
+      "grad_norm": 0.2967066168785095,
+      "learning_rate": 6.486919506423228e-07,
+      "loss": 0.3225,
+      "step": 10168
+    },
+    {
+      "epoch": 2.557595573440644,
+      "grad_norm": 0.31147968769073486,
+      "learning_rate": 6.47971356099501e-07,
+      "loss": 0.3556,
+      "step": 10169
+    },
+    {
+      "epoch": 2.5578470824949697,
+      "grad_norm": 0.3024539053440094,
+      "learning_rate": 6.472511342807052e-07,
+      "loss": 0.3199,
+      "step": 10170
+    },
+    {
+      "epoch": 2.558098591549296,
+      "grad_norm": 0.28230634331703186,
+      "learning_rate": 6.465312852476197e-07,
+      "loss": 0.3253,
+      "step": 10171
+    },
+    {
+      "epoch": 2.558350100603622,
+      "grad_norm": 0.30267390608787537,
+      "learning_rate": 6.458118090618948e-07,
+      "loss": 0.3296,
+      "step": 10172
+    },
+    {
+      "epoch": 2.5586016096579476,
+      "grad_norm": 0.27803835272789,
+      "learning_rate": 6.450927057851481e-07,
+      "loss": 0.296,
+      "step": 10173
+    },
+    {
+      "epoch": 2.558853118712274,
+      "grad_norm": 0.2754371166229248,
+      "learning_rate": 6.44373975478968e-07,
+      "loss": 0.3001,
+      "step": 10174
+    },
+    {
+      "epoch": 2.5591046277665996,
+      "grad_norm": 0.2934453785419464,
+      "learning_rate": 6.436556182049069e-07,
+      "loss": 0.3301,
+      "step": 10175
+    },
+    {
+      "epoch": 2.5593561368209254,
+      "grad_norm": 0.29431700706481934,
+      "learning_rate": 6.429376340244897e-07,
+      "loss": 0.3223,
+      "step": 10176
+    },
+    {
+      "epoch": 2.5596076458752517,
+      "grad_norm": 0.26826512813568115,
+      "learning_rate": 6.42220022999206e-07,
+      "loss": 0.3048,
+      "step": 10177
+    },
+    {
+      "epoch": 2.5598591549295775,
+      "grad_norm": 0.2905735671520233,
+      "learning_rate": 6.415027851905159e-07,
+      "loss": 0.3369,
+      "step": 10178
+    },
+    {
+      "epoch": 2.5601106639839033,
+      "grad_norm": 0.2780333161354065,
+      "learning_rate": 6.407859206598443e-07,
+      "loss": 0.3325,
+      "step": 10179
+    },
+    {
+      "epoch": 2.5603621730382295,
+      "grad_norm": 0.3008924126625061,
+      "learning_rate": 6.400694294685889e-07,
+      "loss": 0.3048,
+      "step": 10180
+    },
+    {
+      "epoch": 2.5606136820925554,
+      "grad_norm": 0.2865042984485626,
+      "learning_rate": 6.393533116781098e-07,
+      "loss": 0.3104,
+      "step": 10181
+    },
+    {
+      "epoch": 2.560865191146881,
+      "grad_norm": 0.28145965933799744,
+      "learning_rate": 6.386375673497397e-07,
+      "loss": 0.3036,
+      "step": 10182
+    },
+    {
+      "epoch": 2.5611167002012074,
+      "grad_norm": 0.2906494438648224,
+      "learning_rate": 6.379221965447785e-07,
+      "loss": 0.312,
+      "step": 10183
+    },
+    {
+      "epoch": 2.561368209255533,
+      "grad_norm": 0.2934057116508484,
+      "learning_rate": 6.37207199324491e-07,
+      "loss": 0.3203,
+      "step": 10184
+    },
+    {
+      "epoch": 2.561619718309859,
+      "grad_norm": 0.3017788827419281,
+      "learning_rate": 6.364925757501139e-07,
+      "loss": 0.3339,
+      "step": 10185
+    },
+    {
+      "epoch": 2.5618712273641853,
+      "grad_norm": 0.2982634902000427,
+      "learning_rate": 6.357783258828493e-07,
+      "loss": 0.3241,
+      "step": 10186
+    },
+    {
+      "epoch": 2.562122736418511,
+      "grad_norm": 0.28276070952415466,
+      "learning_rate": 6.350644497838692e-07,
+      "loss": 0.3201,
+      "step": 10187
+    },
+    {
+      "epoch": 2.562374245472837,
+      "grad_norm": 0.27246373891830444,
+      "learning_rate": 6.343509475143112e-07,
+      "loss": 0.3196,
+      "step": 10188
+    },
+    {
+      "epoch": 2.562625754527163,
+      "grad_norm": 0.2898280918598175,
+      "learning_rate": 6.336378191352838e-07,
+      "loss": 0.3311,
+      "step": 10189
+    },
+    {
+      "epoch": 2.562877263581489,
+      "grad_norm": 0.29259827733039856,
+      "learning_rate": 6.329250647078605e-07,
+      "loss": 0.3205,
+      "step": 10190
+    },
+    {
+      "epoch": 2.5631287726358147,
+      "grad_norm": 0.3051588237285614,
+      "learning_rate": 6.322126842930864e-07,
+      "loss": 0.3302,
+      "step": 10191
+    },
+    {
+      "epoch": 2.563380281690141,
+      "grad_norm": 0.28429844975471497,
+      "learning_rate": 6.3150067795197e-07,
+      "loss": 0.3169,
+      "step": 10192
+    },
+    {
+      "epoch": 2.563631790744467,
+      "grad_norm": 0.2904362082481384,
+      "learning_rate": 6.307890457454907e-07,
+      "loss": 0.3111,
+      "step": 10193
+    },
+    {
+      "epoch": 2.5638832997987926,
+      "grad_norm": 0.28321024775505066,
+      "learning_rate": 6.300777877345976e-07,
+      "loss": 0.3277,
+      "step": 10194
+    },
+    {
+      "epoch": 2.564134808853119,
+      "grad_norm": 0.2912009656429291,
+      "learning_rate": 6.293669039802025e-07,
+      "loss": 0.3177,
+      "step": 10195
+    },
+    {
+      "epoch": 2.5643863179074446,
+      "grad_norm": 0.29499733448028564,
+      "learning_rate": 6.286563945431906e-07,
+      "loss": 0.3204,
+      "step": 10196
+    },
+    {
+      "epoch": 2.5646378269617705,
+      "grad_norm": 0.2899915874004364,
+      "learning_rate": 6.279462594844105e-07,
+      "loss": 0.3003,
+      "step": 10197
+    },
+    {
+      "epoch": 2.5648893360160967,
+      "grad_norm": 0.29533299803733826,
+      "learning_rate": 6.272364988646828e-07,
+      "loss": 0.3252,
+      "step": 10198
+    },
+    {
+      "epoch": 2.5651408450704225,
+      "grad_norm": 0.2786716818809509,
+      "learning_rate": 6.26527112744792e-07,
+      "loss": 0.3268,
+      "step": 10199
+    },
+    {
+      "epoch": 2.5653923541247483,
+      "grad_norm": 0.28114524483680725,
+      "learning_rate": 6.258181011854947e-07,
+      "loss": 0.3183,
+      "step": 10200
+    },
+    {
+      "epoch": 2.5656438631790746,
+      "grad_norm": 0.30359089374542236,
+      "learning_rate": 6.251094642475108e-07,
+      "loss": 0.3293,
+      "step": 10201
+    },
+    {
+      "epoch": 2.5658953722334004,
+      "grad_norm": 0.295454204082489,
+      "learning_rate": 6.244012019915335e-07,
+      "loss": 0.3227,
+      "step": 10202
+    },
+    {
+      "epoch": 2.566146881287726,
+      "grad_norm": 0.29742130637168884,
+      "learning_rate": 6.236933144782187e-07,
+      "loss": 0.3293,
+      "step": 10203
+    },
+    {
+      "epoch": 2.5663983903420524,
+      "grad_norm": 0.29131677746772766,
+      "learning_rate": 6.229858017681933e-07,
+      "loss": 0.3234,
+      "step": 10204
+    },
+    {
+      "epoch": 2.566649899396378,
+      "grad_norm": 0.280320942401886,
+      "learning_rate": 6.222786639220524e-07,
+      "loss": 0.3104,
+      "step": 10205
+    },
+    {
+      "epoch": 2.566901408450704,
+      "grad_norm": 0.33856505155563354,
+      "learning_rate": 6.215719010003557e-07,
+      "loss": 0.3169,
+      "step": 10206
+    },
+    {
+      "epoch": 2.5671529175050303,
+      "grad_norm": 0.2823481559753418,
+      "learning_rate": 6.208655130636354e-07,
+      "loss": 0.3126,
+      "step": 10207
+    },
+    {
+      "epoch": 2.567404426559356,
+      "grad_norm": 0.28553295135498047,
+      "learning_rate": 6.201595001723876e-07,
+      "loss": 0.3226,
+      "step": 10208
+    },
+    {
+      "epoch": 2.567655935613682,
+      "grad_norm": 0.28577616810798645,
+      "learning_rate": 6.194538623870794e-07,
+      "loss": 0.3168,
+      "step": 10209
+    },
+    {
+      "epoch": 2.567907444668008,
+      "grad_norm": 0.29407092928886414,
+      "learning_rate": 6.187485997681419e-07,
+      "loss": 0.31,
+      "step": 10210
+    },
+    {
+      "epoch": 2.568158953722334,
+      "grad_norm": 0.29761141538619995,
+      "learning_rate": 6.180437123759786e-07,
+      "loss": 0.2955,
+      "step": 10211
+    },
+    {
+      "epoch": 2.5684104627766597,
+      "grad_norm": 0.27446889877319336,
+      "learning_rate": 6.173392002709572e-07,
+      "loss": 0.3325,
+      "step": 10212
+    },
+    {
+      "epoch": 2.568661971830986,
+      "grad_norm": 0.2877654731273651,
+      "learning_rate": 6.166350635134166e-07,
+      "loss": 0.3209,
+      "step": 10213
+    },
+    {
+      "epoch": 2.568913480885312,
+      "grad_norm": 0.30045285820961,
+      "learning_rate": 6.159313021636593e-07,
+      "loss": 0.3168,
+      "step": 10214
+    },
+    {
+      "epoch": 2.5691649899396376,
+      "grad_norm": 0.28202614188194275,
+      "learning_rate": 6.152279162819597e-07,
+      "loss": 0.3216,
+      "step": 10215
+    },
+    {
+      "epoch": 2.569416498993964,
+      "grad_norm": 0.33453133702278137,
+      "learning_rate": 6.145249059285585e-07,
+      "loss": 0.3074,
+      "step": 10216
+    },
+    {
+      "epoch": 2.5696680080482897,
+      "grad_norm": 0.28775471448898315,
+      "learning_rate": 6.138222711636632e-07,
+      "loss": 0.3357,
+      "step": 10217
+    },
+    {
+      "epoch": 2.5699195171026155,
+      "grad_norm": 0.29132527112960815,
+      "learning_rate": 6.131200120474512e-07,
+      "loss": 0.3223,
+      "step": 10218
+    },
+    {
+      "epoch": 2.5701710261569417,
+      "grad_norm": 0.29309695959091187,
+      "learning_rate": 6.124181286400649e-07,
+      "loss": 0.3157,
+      "step": 10219
+    },
+    {
+      "epoch": 2.5704225352112675,
+      "grad_norm": 0.2960768938064575,
+      "learning_rate": 6.117166210016184e-07,
+      "loss": 0.3223,
+      "step": 10220
+    },
+    {
+      "epoch": 2.5706740442655933,
+      "grad_norm": 0.2980702817440033,
+      "learning_rate": 6.110154891921894e-07,
+      "loss": 0.3371,
+      "step": 10221
+    },
+    {
+      "epoch": 2.5709255533199196,
+      "grad_norm": 0.2795586884021759,
+      "learning_rate": 6.103147332718274e-07,
+      "loss": 0.3141,
+      "step": 10222
+    },
+    {
+      "epoch": 2.5711770623742454,
+      "grad_norm": 0.2727452516555786,
+      "learning_rate": 6.096143533005455e-07,
+      "loss": 0.3137,
+      "step": 10223
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.28229624032974243,
+      "learning_rate": 6.089143493383281e-07,
+      "loss": 0.3211,
+      "step": 10224
+    },
+    {
+      "epoch": 2.5716800804828974,
+      "grad_norm": 0.3388809263706207,
+      "learning_rate": 6.082147214451272e-07,
+      "loss": 0.3339,
+      "step": 10225
+    },
+    {
+      "epoch": 2.5719315895372232,
+      "grad_norm": 0.2950616776943207,
+      "learning_rate": 6.075154696808594e-07,
+      "loss": 0.3158,
+      "step": 10226
+    },
+    {
+      "epoch": 2.572183098591549,
+      "grad_norm": 0.2900426983833313,
+      "learning_rate": 6.068165941054133e-07,
+      "loss": 0.3371,
+      "step": 10227
+    },
+    {
+      "epoch": 2.5724346076458753,
+      "grad_norm": 0.29088470339775085,
+      "learning_rate": 6.061180947786411e-07,
+      "loss": 0.3193,
+      "step": 10228
+    },
+    {
+      "epoch": 2.572686116700201,
+      "grad_norm": 0.30321744084358215,
+      "learning_rate": 6.054199717603671e-07,
+      "loss": 0.3315,
+      "step": 10229
+    },
+    {
+      "epoch": 2.572937625754527,
+      "grad_norm": 0.3024834990501404,
+      "learning_rate": 6.047222251103796e-07,
+      "loss": 0.3233,
+      "step": 10230
+    },
+    {
+      "epoch": 2.573189134808853,
+      "grad_norm": 0.2948141396045685,
+      "learning_rate": 6.04024854888437e-07,
+      "loss": 0.3318,
+      "step": 10231
+    },
+    {
+      "epoch": 2.573440643863179,
+      "grad_norm": 0.323097825050354,
+      "learning_rate": 6.033278611542642e-07,
+      "loss": 0.3179,
+      "step": 10232
+    },
+    {
+      "epoch": 2.573692152917505,
+      "grad_norm": 0.2814940810203552,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.3333,
+      "step": 10233
+    },
+    {
+      "epoch": 2.573943661971831,
+      "grad_norm": 0.2813478708267212,
+      "learning_rate": 6.01935003387969e-07,
+      "loss": 0.3158,
+      "step": 10234
+    },
+    {
+      "epoch": 2.574195171026157,
+      "grad_norm": 0.27595117688179016,
+      "learning_rate": 6.01239139475136e-07,
+      "loss": 0.2883,
+      "step": 10235
+    },
+    {
+      "epoch": 2.574446680080483,
+      "grad_norm": 0.278484970331192,
+      "learning_rate": 6.005436522886532e-07,
+      "loss": 0.308,
+      "step": 10236
+    },
+    {
+      "epoch": 2.574698189134809,
+      "grad_norm": 0.285978227853775,
+      "learning_rate": 5.998485418880822e-07,
+      "loss": 0.3014,
+      "step": 10237
+    },
+    {
+      "epoch": 2.5749496981891347,
+      "grad_norm": 0.3101520538330078,
+      "learning_rate": 5.991538083329579e-07,
+      "loss": 0.3149,
+      "step": 10238
+    },
+    {
+      "epoch": 2.575201207243461,
+      "grad_norm": 0.2651280462741852,
+      "learning_rate": 5.984594516827769e-07,
+      "loss": 0.3223,
+      "step": 10239
+    },
+    {
+      "epoch": 2.5754527162977867,
+      "grad_norm": 0.307229608297348,
+      "learning_rate": 5.977654719970088e-07,
+      "loss": 0.3224,
+      "step": 10240
+    },
+    {
+      "epoch": 2.5757042253521125,
+      "grad_norm": 0.28054678440093994,
+      "learning_rate": 5.970718693350874e-07,
+      "loss": 0.3274,
+      "step": 10241
+    },
+    {
+      "epoch": 2.5759557344064388,
+      "grad_norm": 0.2816450297832489,
+      "learning_rate": 5.963786437564161e-07,
+      "loss": 0.3536,
+      "step": 10242
+    },
+    {
+      "epoch": 2.5762072434607646,
+      "grad_norm": 0.281474769115448,
+      "learning_rate": 5.956857953203643e-07,
+      "loss": 0.324,
+      "step": 10243
+    },
+    {
+      "epoch": 2.5764587525150904,
+      "grad_norm": 0.3121716380119324,
+      "learning_rate": 5.949933240862715e-07,
+      "loss": 0.3248,
+      "step": 10244
+    },
+    {
+      "epoch": 2.5767102615694166,
+      "grad_norm": 0.28747543692588806,
+      "learning_rate": 5.94301230113442e-07,
+      "loss": 0.3225,
+      "step": 10245
+    },
+    {
+      "epoch": 2.5769617706237424,
+      "grad_norm": 0.29833775758743286,
+      "learning_rate": 5.936095134611508e-07,
+      "loss": 0.2994,
+      "step": 10246
+    },
+    {
+      "epoch": 2.5772132796780687,
+      "grad_norm": 0.2817133069038391,
+      "learning_rate": 5.929181741886386e-07,
+      "loss": 0.3039,
+      "step": 10247
+    },
+    {
+      "epoch": 2.5774647887323945,
+      "grad_norm": 0.28056350350379944,
+      "learning_rate": 5.922272123551137e-07,
+      "loss": 0.3196,
+      "step": 10248
+    },
+    {
+      "epoch": 2.5777162977867203,
+      "grad_norm": 0.28009307384490967,
+      "learning_rate": 5.915366280197537e-07,
+      "loss": 0.3259,
+      "step": 10249
+    },
+    {
+      "epoch": 2.5779678068410465,
+      "grad_norm": 0.303177148103714,
+      "learning_rate": 5.908464212417014e-07,
+      "loss": 0.3336,
+      "step": 10250
+    },
+    {
+      "epoch": 2.5782193158953723,
+      "grad_norm": 0.2899198532104492,
+      "learning_rate": 5.901565920800711e-07,
+      "loss": 0.2883,
+      "step": 10251
+    },
+    {
+      "epoch": 2.578470824949698,
+      "grad_norm": 0.28011229634284973,
+      "learning_rate": 5.894671405939389e-07,
+      "loss": 0.331,
+      "step": 10252
+    },
+    {
+      "epoch": 2.5787223340040244,
+      "grad_norm": 0.2783331274986267,
+      "learning_rate": 5.887780668423553e-07,
+      "loss": 0.3173,
+      "step": 10253
+    },
+    {
+      "epoch": 2.57897384305835,
+      "grad_norm": 0.31792569160461426,
+      "learning_rate": 5.880893708843332e-07,
+      "loss": 0.3361,
+      "step": 10254
+    },
+    {
+      "epoch": 2.579225352112676,
+      "grad_norm": 0.2781875729560852,
+      "learning_rate": 5.87401052778856e-07,
+      "loss": 0.3333,
+      "step": 10255
+    },
+    {
+      "epoch": 2.5794768611670023,
+      "grad_norm": 0.29065370559692383,
+      "learning_rate": 5.867131125848729e-07,
+      "loss": 0.3436,
+      "step": 10256
+    },
+    {
+      "epoch": 2.579728370221328,
+      "grad_norm": 0.2745533585548401,
+      "learning_rate": 5.860255503613033e-07,
+      "loss": 0.3263,
+      "step": 10257
+    },
+    {
+      "epoch": 2.579979879275654,
+      "grad_norm": 0.2924485206604004,
+      "learning_rate": 5.853383661670303e-07,
+      "loss": 0.3291,
+      "step": 10258
+    },
+    {
+      "epoch": 2.58023138832998,
+      "grad_norm": 0.288885235786438,
+      "learning_rate": 5.846515600609093e-07,
+      "loss": 0.3057,
+      "step": 10259
+    },
+    {
+      "epoch": 2.580482897384306,
+      "grad_norm": 0.2955508530139923,
+      "learning_rate": 5.839651321017586e-07,
+      "loss": 0.3173,
+      "step": 10260
+    },
+    {
+      "epoch": 2.5807344064386317,
+      "grad_norm": 0.289850115776062,
+      "learning_rate": 5.832790823483691e-07,
+      "loss": 0.3112,
+      "step": 10261
+    },
+    {
+      "epoch": 2.580985915492958,
+      "grad_norm": 0.30250605940818787,
+      "learning_rate": 5.825934108594938e-07,
+      "loss": 0.3184,
+      "step": 10262
+    },
+    {
+      "epoch": 2.5812374245472838,
+      "grad_norm": 0.28620943427085876,
+      "learning_rate": 5.81908117693859e-07,
+      "loss": 0.3132,
+      "step": 10263
+    },
+    {
+      "epoch": 2.5814889336016096,
+      "grad_norm": 0.2834034264087677,
+      "learning_rate": 5.812232029101533e-07,
+      "loss": 0.3057,
+      "step": 10264
+    },
+    {
+      "epoch": 2.581740442655936,
+      "grad_norm": 0.2939947545528412,
+      "learning_rate": 5.805386665670376e-07,
+      "loss": 0.3077,
+      "step": 10265
+    },
+    {
+      "epoch": 2.5819919517102616,
+      "grad_norm": 0.3035759925842285,
+      "learning_rate": 5.79854508723136e-07,
+      "loss": 0.3171,
+      "step": 10266
+    },
+    {
+      "epoch": 2.5822434607645874,
+      "grad_norm": 0.28236502408981323,
+      "learning_rate": 5.791707294370447e-07,
+      "loss": 0.3288,
+      "step": 10267
+    },
+    {
+      "epoch": 2.5824949698189137,
+      "grad_norm": 0.2845684587955475,
+      "learning_rate": 5.784873287673226e-07,
+      "loss": 0.3225,
+      "step": 10268
+    },
+    {
+      "epoch": 2.5827464788732395,
+      "grad_norm": 0.31281614303588867,
+      "learning_rate": 5.778043067725009e-07,
+      "loss": 0.3177,
+      "step": 10269
+    },
+    {
+      "epoch": 2.5829979879275653,
+      "grad_norm": 0.33208024501800537,
+      "learning_rate": 5.771216635110738e-07,
+      "loss": 0.2962,
+      "step": 10270
+    },
+    {
+      "epoch": 2.5832494969818915,
+      "grad_norm": 0.28506436944007874,
+      "learning_rate": 5.764393990415079e-07,
+      "loss": 0.3374,
+      "step": 10271
+    },
+    {
+      "epoch": 2.5835010060362174,
+      "grad_norm": 0.29046157002449036,
+      "learning_rate": 5.757575134222332e-07,
+      "loss": 0.3216,
+      "step": 10272
+    },
+    {
+      "epoch": 2.583752515090543,
+      "grad_norm": 0.2891055941581726,
+      "learning_rate": 5.750760067116501e-07,
+      "loss": 0.3129,
+      "step": 10273
+    },
+    {
+      "epoch": 2.5840040241448694,
+      "grad_norm": 0.2932874858379364,
+      "learning_rate": 5.743948789681236e-07,
+      "loss": 0.2921,
+      "step": 10274
+    },
+    {
+      "epoch": 2.584255533199195,
+      "grad_norm": 0.3003051280975342,
+      "learning_rate": 5.73714130249991e-07,
+      "loss": 0.3158,
+      "step": 10275
+    },
+    {
+      "epoch": 2.584507042253521,
+      "grad_norm": 0.278804749250412,
+      "learning_rate": 5.730337606155506e-07,
+      "loss": 0.3074,
+      "step": 10276
+    },
+    {
+      "epoch": 2.5847585513078473,
+      "grad_norm": 0.297124445438385,
+      "learning_rate": 5.723537701230747e-07,
+      "loss": 0.3316,
+      "step": 10277
+    },
+    {
+      "epoch": 2.585010060362173,
+      "grad_norm": 0.28858664631843567,
+      "learning_rate": 5.716741588307983e-07,
+      "loss": 0.3064,
+      "step": 10278
+    },
+    {
+      "epoch": 2.585261569416499,
+      "grad_norm": 0.30511143803596497,
+      "learning_rate": 5.709949267969267e-07,
+      "loss": 0.3197,
+      "step": 10279
+    },
+    {
+      "epoch": 2.585513078470825,
+      "grad_norm": 0.2882242500782013,
+      "learning_rate": 5.703160740796332e-07,
+      "loss": 0.3344,
+      "step": 10280
+    },
+    {
+      "epoch": 2.585764587525151,
+      "grad_norm": 0.32658329606056213,
+      "learning_rate": 5.696376007370541e-07,
+      "loss": 0.3257,
+      "step": 10281
+    },
+    {
+      "epoch": 2.5860160965794767,
+      "grad_norm": 0.3152960538864136,
+      "learning_rate": 5.689595068273002e-07,
+      "loss": 0.3103,
+      "step": 10282
+    },
+    {
+      "epoch": 2.586267605633803,
+      "grad_norm": 0.2786172330379486,
+      "learning_rate": 5.682817924084422e-07,
+      "loss": 0.3174,
+      "step": 10283
+    },
+    {
+      "epoch": 2.586519114688129,
+      "grad_norm": 0.29683375358581543,
+      "learning_rate": 5.676044575385254e-07,
+      "loss": 0.3426,
+      "step": 10284
+    },
+    {
+      "epoch": 2.5867706237424546,
+      "grad_norm": 0.28705403208732605,
+      "learning_rate": 5.669275022755566e-07,
+      "loss": 0.3055,
+      "step": 10285
+    },
+    {
+      "epoch": 2.587022132796781,
+      "grad_norm": 0.29597151279449463,
+      "learning_rate": 5.662509266775151e-07,
+      "loss": 0.3117,
+      "step": 10286
+    },
+    {
+      "epoch": 2.5872736418511066,
+      "grad_norm": 0.29061004519462585,
+      "learning_rate": 5.655747308023434e-07,
+      "loss": 0.3414,
+      "step": 10287
+    },
+    {
+      "epoch": 2.5875251509054324,
+      "grad_norm": 0.3225204646587372,
+      "learning_rate": 5.648989147079553e-07,
+      "loss": 0.3319,
+      "step": 10288
+    },
+    {
+      "epoch": 2.5877766599597587,
+      "grad_norm": 0.27866053581237793,
+      "learning_rate": 5.642234784522282e-07,
+      "loss": 0.3036,
+      "step": 10289
+    },
+    {
+      "epoch": 2.5880281690140845,
+      "grad_norm": 0.3126038610935211,
+      "learning_rate": 5.635484220930098e-07,
+      "loss": 0.3094,
+      "step": 10290
+    },
+    {
+      "epoch": 2.5882796780684103,
+      "grad_norm": 0.2966510057449341,
+      "learning_rate": 5.628737456881161e-07,
+      "loss": 0.3066,
+      "step": 10291
+    },
+    {
+      "epoch": 2.5885311871227366,
+      "grad_norm": 0.2831152677536011,
+      "learning_rate": 5.621994492953264e-07,
+      "loss": 0.3296,
+      "step": 10292
+    },
+    {
+      "epoch": 2.5887826961770624,
+      "grad_norm": 0.2895983159542084,
+      "learning_rate": 5.615255329723917e-07,
+      "loss": 0.3145,
+      "step": 10293
+    },
+    {
+      "epoch": 2.589034205231388,
+      "grad_norm": 0.29133400321006775,
+      "learning_rate": 5.608519967770276e-07,
+      "loss": 0.3194,
+      "step": 10294
+    },
+    {
+      "epoch": 2.5892857142857144,
+      "grad_norm": 0.2729475200176239,
+      "learning_rate": 5.601788407669196e-07,
+      "loss": 0.3218,
+      "step": 10295
+    },
+    {
+      "epoch": 2.58953722334004,
+      "grad_norm": 0.2915710508823395,
+      "learning_rate": 5.595060649997175e-07,
+      "loss": 0.3297,
+      "step": 10296
+    },
+    {
+      "epoch": 2.589788732394366,
+      "grad_norm": 0.31030383706092834,
+      "learning_rate": 5.588336695330421e-07,
+      "loss": 0.3357,
+      "step": 10297
+    },
+    {
+      "epoch": 2.5900402414486923,
+      "grad_norm": 0.3029539883136749,
+      "learning_rate": 5.581616544244778e-07,
+      "loss": 0.2825,
+      "step": 10298
+    },
+    {
+      "epoch": 2.590291750503018,
+      "grad_norm": 0.30874502658843994,
+      "learning_rate": 5.574900197315814e-07,
+      "loss": 0.3133,
+      "step": 10299
+    },
+    {
+      "epoch": 2.590543259557344,
+      "grad_norm": 0.3056495785713196,
+      "learning_rate": 5.568187655118712e-07,
+      "loss": 0.3328,
+      "step": 10300
+    },
+    {
+      "epoch": 2.59079476861167,
+      "grad_norm": 0.29649117588996887,
+      "learning_rate": 5.561478918228369e-07,
+      "loss": 0.31,
+      "step": 10301
+    },
+    {
+      "epoch": 2.591046277665996,
+      "grad_norm": 0.2896314263343811,
+      "learning_rate": 5.55477398721937e-07,
+      "loss": 0.3237,
+      "step": 10302
+    },
+    {
+      "epoch": 2.5912977867203217,
+      "grad_norm": 0.2920903265476227,
+      "learning_rate": 5.548072862665909e-07,
+      "loss": 0.2982,
+      "step": 10303
+    },
+    {
+      "epoch": 2.591549295774648,
+      "grad_norm": 0.309803307056427,
+      "learning_rate": 5.541375545141936e-07,
+      "loss": 0.3153,
+      "step": 10304
+    },
+    {
+      "epoch": 2.591800804828974,
+      "grad_norm": 0.3066963255405426,
+      "learning_rate": 5.534682035221e-07,
+      "loss": 0.3182,
+      "step": 10305
+    },
+    {
+      "epoch": 2.5920523138832996,
+      "grad_norm": 0.27324411273002625,
+      "learning_rate": 5.527992333476389e-07,
+      "loss": 0.3304,
+      "step": 10306
+    },
+    {
+      "epoch": 2.592303822937626,
+      "grad_norm": 0.3093414604663849,
+      "learning_rate": 5.521306440481005e-07,
+      "loss": 0.331,
+      "step": 10307
+    },
+    {
+      "epoch": 2.5925553319919517,
+      "grad_norm": 0.29584071040153503,
+      "learning_rate": 5.514624356807474e-07,
+      "loss": 0.3306,
+      "step": 10308
+    },
+    {
+      "epoch": 2.5928068410462775,
+      "grad_norm": 0.27764448523521423,
+      "learning_rate": 5.507946083028059e-07,
+      "loss": 0.3037,
+      "step": 10309
+    },
+    {
+      "epoch": 2.5930583501006037,
+      "grad_norm": 0.30795976519584656,
+      "learning_rate": 5.501271619714732e-07,
+      "loss": 0.311,
+      "step": 10310
+    },
+    {
+      "epoch": 2.5933098591549295,
+      "grad_norm": 0.29350659251213074,
+      "learning_rate": 5.494600967439095e-07,
+      "loss": 0.3163,
+      "step": 10311
+    },
+    {
+      "epoch": 2.5935613682092553,
+      "grad_norm": 0.3054122030735016,
+      "learning_rate": 5.487934126772465e-07,
+      "loss": 0.327,
+      "step": 10312
+    },
+    {
+      "epoch": 2.5938128772635816,
+      "grad_norm": 0.29255709052085876,
+      "learning_rate": 5.481271098285818e-07,
+      "loss": 0.3198,
+      "step": 10313
+    },
+    {
+      "epoch": 2.5940643863179074,
+      "grad_norm": 0.26492074131965637,
+      "learning_rate": 5.474611882549785e-07,
+      "loss": 0.3185,
+      "step": 10314
+    },
+    {
+      "epoch": 2.594315895372233,
+      "grad_norm": 0.306562215089798,
+      "learning_rate": 5.4679564801347e-07,
+      "loss": 0.3028,
+      "step": 10315
+    },
+    {
+      "epoch": 2.5945674044265594,
+      "grad_norm": 0.3125389814376831,
+      "learning_rate": 5.461304891610541e-07,
+      "loss": 0.3101,
+      "step": 10316
+    },
+    {
+      "epoch": 2.5948189134808852,
+      "grad_norm": 0.31623390316963196,
+      "learning_rate": 5.454657117546996e-07,
+      "loss": 0.3141,
+      "step": 10317
+    },
+    {
+      "epoch": 2.595070422535211,
+      "grad_norm": 0.2654491066932678,
+      "learning_rate": 5.448013158513388e-07,
+      "loss": 0.3211,
+      "step": 10318
+    },
+    {
+      "epoch": 2.5953219315895373,
+      "grad_norm": 0.30861228704452515,
+      "learning_rate": 5.441373015078744e-07,
+      "loss": 0.3259,
+      "step": 10319
+    },
+    {
+      "epoch": 2.595573440643863,
+      "grad_norm": 0.26960867643356323,
+      "learning_rate": 5.434736687811731e-07,
+      "loss": 0.3155,
+      "step": 10320
+    },
+    {
+      "epoch": 2.595824949698189,
+      "grad_norm": 0.26485928893089294,
+      "learning_rate": 5.428104177280735e-07,
+      "loss": 0.3251,
+      "step": 10321
+    },
+    {
+      "epoch": 2.596076458752515,
+      "grad_norm": 0.29853057861328125,
+      "learning_rate": 5.421475484053762e-07,
+      "loss": 0.3669,
+      "step": 10322
+    },
+    {
+      "epoch": 2.596327967806841,
+      "grad_norm": 0.29327720403671265,
+      "learning_rate": 5.414850608698535e-07,
+      "loss": 0.3355,
+      "step": 10323
+    },
+    {
+      "epoch": 2.5965794768611667,
+      "grad_norm": 0.2685185372829437,
+      "learning_rate": 5.408229551782435e-07,
+      "loss": 0.3319,
+      "step": 10324
+    },
+    {
+      "epoch": 2.596830985915493,
+      "grad_norm": 0.3062201142311096,
+      "learning_rate": 5.4016123138725e-07,
+      "loss": 0.3081,
+      "step": 10325
+    },
+    {
+      "epoch": 2.597082494969819,
+      "grad_norm": 0.3145555555820465,
+      "learning_rate": 5.394998895535475e-07,
+      "loss": 0.312,
+      "step": 10326
+    },
+    {
+      "epoch": 2.5973340040241446,
+      "grad_norm": 0.30714958906173706,
+      "learning_rate": 5.388389297337737e-07,
+      "loss": 0.3083,
+      "step": 10327
+    },
+    {
+      "epoch": 2.597585513078471,
+      "grad_norm": 0.28293806314468384,
+      "learning_rate": 5.381783519845374e-07,
+      "loss": 0.3066,
+      "step": 10328
+    },
+    {
+      "epoch": 2.5978370221327967,
+      "grad_norm": 0.2907370626926422,
+      "learning_rate": 5.375181563624116e-07,
+      "loss": 0.3577,
+      "step": 10329
+    },
+    {
+      "epoch": 2.5980885311871225,
+      "grad_norm": 0.3050544559955597,
+      "learning_rate": 5.368583429239394e-07,
+      "loss": 0.3285,
+      "step": 10330
+    },
+    {
+      "epoch": 2.5983400402414487,
+      "grad_norm": 0.3092591166496277,
+      "learning_rate": 5.361989117256277e-07,
+      "loss": 0.3384,
+      "step": 10331
+    },
+    {
+      "epoch": 2.5985915492957745,
+      "grad_norm": 0.27648666501045227,
+      "learning_rate": 5.355398628239544e-07,
+      "loss": 0.3422,
+      "step": 10332
+    },
+    {
+      "epoch": 2.5988430583501008,
+      "grad_norm": 0.3023586869239807,
+      "learning_rate": 5.34881196275362e-07,
+      "loss": 0.3052,
+      "step": 10333
+    },
+    {
+      "epoch": 2.5990945674044266,
+      "grad_norm": 0.2839665710926056,
+      "learning_rate": 5.342229121362607e-07,
+      "loss": 0.2988,
+      "step": 10334
+    },
+    {
+      "epoch": 2.5993460764587524,
+      "grad_norm": 0.3002241849899292,
+      "learning_rate": 5.335650104630308e-07,
+      "loss": 0.3427,
+      "step": 10335
+    },
+    {
+      "epoch": 2.5995975855130786,
+      "grad_norm": 0.27492570877075195,
+      "learning_rate": 5.329074913120141e-07,
+      "loss": 0.3233,
+      "step": 10336
+    },
+    {
+      "epoch": 2.5998490945674044,
+      "grad_norm": 0.29533281922340393,
+      "learning_rate": 5.322503547395263e-07,
+      "loss": 0.3129,
+      "step": 10337
+    },
+    {
+      "epoch": 2.6001006036217302,
+      "grad_norm": 0.285576194524765,
+      "learning_rate": 5.31593600801844e-07,
+      "loss": 0.314,
+      "step": 10338
+    },
+    {
+      "epoch": 2.6003521126760565,
+      "grad_norm": 0.2748103439807892,
+      "learning_rate": 5.309372295552173e-07,
+      "loss": 0.3299,
+      "step": 10339
+    },
+    {
+      "epoch": 2.6006036217303823,
+      "grad_norm": 0.2951182425022125,
+      "learning_rate": 5.302812410558567e-07,
+      "loss": 0.3202,
+      "step": 10340
+    },
+    {
+      "epoch": 2.600855130784708,
+      "grad_norm": 0.30300065875053406,
+      "learning_rate": 5.296256353599466e-07,
+      "loss": 0.3153,
+      "step": 10341
+    },
+    {
+      "epoch": 2.6011066398390343,
+      "grad_norm": 0.3175194561481476,
+      "learning_rate": 5.289704125236333e-07,
+      "loss": 0.3119,
+      "step": 10342
+    },
+    {
+      "epoch": 2.60135814889336,
+      "grad_norm": 0.2759941518306732,
+      "learning_rate": 5.283155726030348e-07,
+      "loss": 0.3194,
+      "step": 10343
+    },
+    {
+      "epoch": 2.6016096579476864,
+      "grad_norm": 0.30253252387046814,
+      "learning_rate": 5.276611156542316e-07,
+      "loss": 0.2918,
+      "step": 10344
+    },
+    {
+      "epoch": 2.601861167002012,
+      "grad_norm": 0.2967860698699951,
+      "learning_rate": 5.270070417332745e-07,
+      "loss": 0.3106,
+      "step": 10345
+    },
+    {
+      "epoch": 2.602112676056338,
+      "grad_norm": 0.2763741612434387,
+      "learning_rate": 5.263533508961827e-07,
+      "loss": 0.3111,
+      "step": 10346
+    },
+    {
+      "epoch": 2.6023641851106643,
+      "grad_norm": 0.28950735926628113,
+      "learning_rate": 5.257000431989384e-07,
+      "loss": 0.3368,
+      "step": 10347
+    },
+    {
+      "epoch": 2.60261569416499,
+      "grad_norm": 0.2865469753742218,
+      "learning_rate": 5.250471186974954e-07,
+      "loss": 0.3163,
+      "step": 10348
+    },
+    {
+      "epoch": 2.602867203219316,
+      "grad_norm": 0.27980104088783264,
+      "learning_rate": 5.243945774477699e-07,
+      "loss": 0.333,
+      "step": 10349
+    },
+    {
+      "epoch": 2.603118712273642,
+      "grad_norm": 0.2943830192089081,
+      "learning_rate": 5.237424195056512e-07,
+      "loss": 0.3181,
+      "step": 10350
+    },
+    {
+      "epoch": 2.603370221327968,
+      "grad_norm": 0.2930108606815338,
+      "learning_rate": 5.230906449269895e-07,
+      "loss": 0.3005,
+      "step": 10351
+    },
+    {
+      "epoch": 2.6036217303822937,
+      "grad_norm": 0.291398823261261,
+      "learning_rate": 5.224392537676077e-07,
+      "loss": 0.3232,
+      "step": 10352
+    },
+    {
+      "epoch": 2.60387323943662,
+      "grad_norm": 0.30126404762268066,
+      "learning_rate": 5.217882460832912e-07,
+      "loss": 0.3264,
+      "step": 10353
+    },
+    {
+      "epoch": 2.6041247484909458,
+      "grad_norm": 0.28334009647369385,
+      "learning_rate": 5.21137621929797e-07,
+      "loss": 0.3262,
+      "step": 10354
+    },
+    {
+      "epoch": 2.6043762575452716,
+      "grad_norm": 0.2912385165691376,
+      "learning_rate": 5.204873813628447e-07,
+      "loss": 0.3388,
+      "step": 10355
+    },
+    {
+      "epoch": 2.604627766599598,
+      "grad_norm": 0.29965826869010925,
+      "learning_rate": 5.198375244381243e-07,
+      "loss": 0.324,
+      "step": 10356
+    },
+    {
+      "epoch": 2.6048792756539236,
+      "grad_norm": 0.2777501046657562,
+      "learning_rate": 5.191880512112934e-07,
+      "loss": 0.3292,
+      "step": 10357
+    },
+    {
+      "epoch": 2.6051307847082494,
+      "grad_norm": 0.2996978163719177,
+      "learning_rate": 5.185389617379727e-07,
+      "loss": 0.337,
+      "step": 10358
+    },
+    {
+      "epoch": 2.6053822937625757,
+      "grad_norm": 0.2818640470504761,
+      "learning_rate": 5.178902560737554e-07,
+      "loss": 0.3316,
+      "step": 10359
+    },
+    {
+      "epoch": 2.6056338028169015,
+      "grad_norm": 0.2935522794723511,
+      "learning_rate": 5.172419342741963e-07,
+      "loss": 0.3255,
+      "step": 10360
+    },
+    {
+      "epoch": 2.6058853118712273,
+      "grad_norm": 0.27610835433006287,
+      "learning_rate": 5.165939963948225e-07,
+      "loss": 0.3366,
+      "step": 10361
+    },
+    {
+      "epoch": 2.6061368209255535,
+      "grad_norm": 0.3033030331134796,
+      "learning_rate": 5.159464424911242e-07,
+      "loss": 0.3233,
+      "step": 10362
+    },
+    {
+      "epoch": 2.6063883299798793,
+      "grad_norm": 0.29915034770965576,
+      "learning_rate": 5.152992726185619e-07,
+      "loss": 0.3066,
+      "step": 10363
+    },
+    {
+      "epoch": 2.606639839034205,
+      "grad_norm": 0.3014090955257416,
+      "learning_rate": 5.146524868325592e-07,
+      "loss": 0.3178,
+      "step": 10364
+    },
+    {
+      "epoch": 2.6068913480885314,
+      "grad_norm": 0.31432896852493286,
+      "learning_rate": 5.140060851885109e-07,
+      "loss": 0.3133,
+      "step": 10365
+    },
+    {
+      "epoch": 2.607142857142857,
+      "grad_norm": 0.28758054971694946,
+      "learning_rate": 5.133600677417782e-07,
+      "loss": 0.3389,
+      "step": 10366
+    },
+    {
+      "epoch": 2.607394366197183,
+      "grad_norm": 0.2802176773548126,
+      "learning_rate": 5.127144345476865e-07,
+      "loss": 0.2957,
+      "step": 10367
+    },
+    {
+      "epoch": 2.6076458752515093,
+      "grad_norm": 0.30097344517707825,
+      "learning_rate": 5.120691856615323e-07,
+      "loss": 0.3412,
+      "step": 10368
+    },
+    {
+      "epoch": 2.607897384305835,
+      "grad_norm": 0.28366386890411377,
+      "learning_rate": 5.114243211385744e-07,
+      "loss": 0.3119,
+      "step": 10369
+    },
+    {
+      "epoch": 2.608148893360161,
+      "grad_norm": 0.2719464898109436,
+      "learning_rate": 5.107798410340442e-07,
+      "loss": 0.3054,
+      "step": 10370
+    },
+    {
+      "epoch": 2.608400402414487,
+      "grad_norm": 0.2885161340236664,
+      "learning_rate": 5.101357454031352e-07,
+      "loss": 0.3178,
+      "step": 10371
+    },
+    {
+      "epoch": 2.608651911468813,
+      "grad_norm": 0.29005712270736694,
+      "learning_rate": 5.094920343010124e-07,
+      "loss": 0.3128,
+      "step": 10372
+    },
+    {
+      "epoch": 2.6089034205231387,
+      "grad_norm": 0.27925872802734375,
+      "learning_rate": 5.08848707782803e-07,
+      "loss": 0.3333,
+      "step": 10373
+    },
+    {
+      "epoch": 2.609154929577465,
+      "grad_norm": 0.272777795791626,
+      "learning_rate": 5.082057659036061e-07,
+      "loss": 0.3407,
+      "step": 10374
+    },
+    {
+      "epoch": 2.609406438631791,
+      "grad_norm": 0.2825217843055725,
+      "learning_rate": 5.07563208718484e-07,
+      "loss": 0.3213,
+      "step": 10375
+    },
+    {
+      "epoch": 2.6096579476861166,
+      "grad_norm": 0.2770220637321472,
+      "learning_rate": 5.069210362824694e-07,
+      "loss": 0.3095,
+      "step": 10376
+    },
+    {
+      "epoch": 2.609909456740443,
+      "grad_norm": 0.296953946352005,
+      "learning_rate": 5.062792486505586e-07,
+      "loss": 0.3356,
+      "step": 10377
+    },
+    {
+      "epoch": 2.6101609657947686,
+      "grad_norm": 0.305620402097702,
+      "learning_rate": 5.056378458777183e-07,
+      "loss": 0.3265,
+      "step": 10378
+    },
+    {
+      "epoch": 2.6104124748490944,
+      "grad_norm": 0.2921883463859558,
+      "learning_rate": 5.049968280188788e-07,
+      "loss": 0.3275,
+      "step": 10379
+    },
+    {
+      "epoch": 2.6106639839034207,
+      "grad_norm": 0.31497353315353394,
+      "learning_rate": 5.043561951289411e-07,
+      "loss": 0.3168,
+      "step": 10380
+    },
+    {
+      "epoch": 2.6109154929577465,
+      "grad_norm": 0.2880299687385559,
+      "learning_rate": 5.0371594726277e-07,
+      "loss": 0.3245,
+      "step": 10381
+    },
+    {
+      "epoch": 2.6111670020120723,
+      "grad_norm": 0.2965189814567566,
+      "learning_rate": 5.030760844752003e-07,
+      "loss": 0.3289,
+      "step": 10382
+    },
+    {
+      "epoch": 2.6114185110663986,
+      "grad_norm": 0.29246169328689575,
+      "learning_rate": 5.024366068210307e-07,
+      "loss": 0.3146,
+      "step": 10383
+    },
+    {
+      "epoch": 2.6116700201207244,
+      "grad_norm": 0.2918647229671478,
+      "learning_rate": 5.017975143550296e-07,
+      "loss": 0.3293,
+      "step": 10384
+    },
+    {
+      "epoch": 2.61192152917505,
+      "grad_norm": 0.30444636940956116,
+      "learning_rate": 5.011588071319295e-07,
+      "loss": 0.3029,
+      "step": 10385
+    },
+    {
+      "epoch": 2.6121730382293764,
+      "grad_norm": 0.3097932040691376,
+      "learning_rate": 5.005204852064344e-07,
+      "loss": 0.3443,
+      "step": 10386
+    },
+    {
+      "epoch": 2.612424547283702,
+      "grad_norm": 0.2824104428291321,
+      "learning_rate": 4.998825486332098e-07,
+      "loss": 0.3114,
+      "step": 10387
+    },
+    {
+      "epoch": 2.612676056338028,
+      "grad_norm": 0.27352190017700195,
+      "learning_rate": 4.992449974668933e-07,
+      "loss": 0.2861,
+      "step": 10388
+    },
+    {
+      "epoch": 2.6129275653923543,
+      "grad_norm": 0.31323835253715515,
+      "learning_rate": 4.986078317620852e-07,
+      "loss": 0.3132,
+      "step": 10389
+    },
+    {
+      "epoch": 2.61317907444668,
+      "grad_norm": 0.3092406690120697,
+      "learning_rate": 4.979710515733566e-07,
+      "loss": 0.2916,
+      "step": 10390
+    },
+    {
+      "epoch": 2.613430583501006,
+      "grad_norm": 0.28336817026138306,
+      "learning_rate": 4.973346569552417e-07,
+      "loss": 0.314,
+      "step": 10391
+    },
+    {
+      "epoch": 2.613682092555332,
+      "grad_norm": 0.27843061089515686,
+      "learning_rate": 4.966986479622454e-07,
+      "loss": 0.3101,
+      "step": 10392
+    },
+    {
+      "epoch": 2.613933601609658,
+      "grad_norm": 0.2896105945110321,
+      "learning_rate": 4.96063024648838e-07,
+      "loss": 0.3245,
+      "step": 10393
+    },
+    {
+      "epoch": 2.6141851106639837,
+      "grad_norm": 0.2984098792076111,
+      "learning_rate": 4.954277870694552e-07,
+      "loss": 0.3126,
+      "step": 10394
+    },
+    {
+      "epoch": 2.61443661971831,
+      "grad_norm": 0.3082030117511749,
+      "learning_rate": 4.947929352785024e-07,
+      "loss": 0.298,
+      "step": 10395
+    },
+    {
+      "epoch": 2.614688128772636,
+      "grad_norm": 0.3350622355937958,
+      "learning_rate": 4.941584693303497e-07,
+      "loss": 0.3467,
+      "step": 10396
+    },
+    {
+      "epoch": 2.6149396378269616,
+      "grad_norm": 0.2979256510734558,
+      "learning_rate": 4.935243892793362e-07,
+      "loss": 0.3165,
+      "step": 10397
+    },
+    {
+      "epoch": 2.615191146881288,
+      "grad_norm": 0.3071826100349426,
+      "learning_rate": 4.928906951797657e-07,
+      "loss": 0.3366,
+      "step": 10398
+    },
+    {
+      "epoch": 2.6154426559356136,
+      "grad_norm": 0.2692726254463196,
+      "learning_rate": 4.922573870859115e-07,
+      "loss": 0.2949,
+      "step": 10399
+    },
+    {
+      "epoch": 2.6156941649899395,
+      "grad_norm": 0.2713705003261566,
+      "learning_rate": 4.916244650520108e-07,
+      "loss": 0.3231,
+      "step": 10400
+    },
+    {
+      "epoch": 2.6159456740442657,
+      "grad_norm": 0.282787561416626,
+      "learning_rate": 4.909919291322718e-07,
+      "loss": 0.3244,
+      "step": 10401
+    },
+    {
+      "epoch": 2.6161971830985915,
+      "grad_norm": 0.2967943549156189,
+      "learning_rate": 4.90359779380864e-07,
+      "loss": 0.3158,
+      "step": 10402
+    },
+    {
+      "epoch": 2.6164486921529173,
+      "grad_norm": 0.30457451939582825,
+      "learning_rate": 4.897280158519307e-07,
+      "loss": 0.3176,
+      "step": 10403
+    },
+    {
+      "epoch": 2.6167002012072436,
+      "grad_norm": 0.3144105076789856,
+      "learning_rate": 4.890966385995754e-07,
+      "loss": 0.3444,
+      "step": 10404
+    },
+    {
+      "epoch": 2.6169517102615694,
+      "grad_norm": 0.28179723024368286,
+      "learning_rate": 4.884656476778738e-07,
+      "loss": 0.3197,
+      "step": 10405
+    },
+    {
+      "epoch": 2.617203219315895,
+      "grad_norm": 0.2976565361022949,
+      "learning_rate": 4.878350431408641e-07,
+      "loss": 0.32,
+      "step": 10406
+    },
+    {
+      "epoch": 2.6174547283702214,
+      "grad_norm": 0.2884789705276489,
+      "learning_rate": 4.872048250425565e-07,
+      "loss": 0.3244,
+      "step": 10407
+    },
+    {
+      "epoch": 2.6177062374245472,
+      "grad_norm": 0.30798858404159546,
+      "learning_rate": 4.865749934369224e-07,
+      "loss": 0.3102,
+      "step": 10408
+    },
+    {
+      "epoch": 2.617957746478873,
+      "grad_norm": 0.2883943021297455,
+      "learning_rate": 4.859455483779041e-07,
+      "loss": 0.3245,
+      "step": 10409
+    },
+    {
+      "epoch": 2.6182092555331993,
+      "grad_norm": 0.28204745054244995,
+      "learning_rate": 4.853164899194107e-07,
+      "loss": 0.3222,
+      "step": 10410
+    },
+    {
+      "epoch": 2.618460764587525,
+      "grad_norm": 0.28429892659187317,
+      "learning_rate": 4.846878181153153e-07,
+      "loss": 0.3085,
+      "step": 10411
+    },
+    {
+      "epoch": 2.618712273641851,
+      "grad_norm": 0.2938331663608551,
+      "learning_rate": 4.840595330194614e-07,
+      "loss": 0.3001,
+      "step": 10412
+    },
+    {
+      "epoch": 2.618963782696177,
+      "grad_norm": 0.28804659843444824,
+      "learning_rate": 4.834316346856565e-07,
+      "loss": 0.352,
+      "step": 10413
+    },
+    {
+      "epoch": 2.619215291750503,
+      "grad_norm": 0.2770189642906189,
+      "learning_rate": 4.828041231676766e-07,
+      "loss": 0.3067,
+      "step": 10414
+    },
+    {
+      "epoch": 2.6194668008048287,
+      "grad_norm": 0.2969238758087158,
+      "learning_rate": 4.821769985192637e-07,
+      "loss": 0.3361,
+      "step": 10415
+    },
+    {
+      "epoch": 2.619718309859155,
+      "grad_norm": 0.2775512933731079,
+      "learning_rate": 4.815502607941286e-07,
+      "loss": 0.3031,
+      "step": 10416
+    },
+    {
+      "epoch": 2.619969818913481,
+      "grad_norm": 0.287936806678772,
+      "learning_rate": 4.809239100459451e-07,
+      "loss": 0.3292,
+      "step": 10417
+    },
+    {
+      "epoch": 2.6202213279678066,
+      "grad_norm": 0.2935434877872467,
+      "learning_rate": 4.80297946328358e-07,
+      "loss": 0.2995,
+      "step": 10418
+    },
+    {
+      "epoch": 2.620472837022133,
+      "grad_norm": 0.3020229637622833,
+      "learning_rate": 4.796723696949762e-07,
+      "loss": 0.3263,
+      "step": 10419
+    },
+    {
+      "epoch": 2.6207243460764587,
+      "grad_norm": 0.29107797145843506,
+      "learning_rate": 4.790471801993768e-07,
+      "loss": 0.2952,
+      "step": 10420
+    },
+    {
+      "epoch": 2.6209758551307845,
+      "grad_norm": 0.29532748460769653,
+      "learning_rate": 4.784223778951042e-07,
+      "loss": 0.3259,
+      "step": 10421
+    },
+    {
+      "epoch": 2.6212273641851107,
+      "grad_norm": 0.29053303599357605,
+      "learning_rate": 4.777979628356672e-07,
+      "loss": 0.3274,
+      "step": 10422
+    },
+    {
+      "epoch": 2.6214788732394365,
+      "grad_norm": 0.3015473783016205,
+      "learning_rate": 4.771739350745447e-07,
+      "loss": 0.3069,
+      "step": 10423
+    },
+    {
+      "epoch": 2.6217303822937623,
+      "grad_norm": 0.29153281450271606,
+      "learning_rate": 4.7655029466517897e-07,
+      "loss": 0.3218,
+      "step": 10424
+    },
+    {
+      "epoch": 2.6219818913480886,
+      "grad_norm": 0.30175068974494934,
+      "learning_rate": 4.759270416609829e-07,
+      "loss": 0.3358,
+      "step": 10425
+    },
+    {
+      "epoch": 2.6222334004024144,
+      "grad_norm": 0.2931424081325531,
+      "learning_rate": 4.753041761153326e-07,
+      "loss": 0.3102,
+      "step": 10426
+    },
+    {
+      "epoch": 2.62248490945674,
+      "grad_norm": 0.2768230438232422,
+      "learning_rate": 4.746816980815738e-07,
+      "loss": 0.3365,
+      "step": 10427
+    },
+    {
+      "epoch": 2.6227364185110664,
+      "grad_norm": 0.2788086235523224,
+      "learning_rate": 4.7405960761301606e-07,
+      "loss": 0.3253,
+      "step": 10428
+    },
+    {
+      "epoch": 2.6229879275653922,
+      "grad_norm": 0.27975523471832275,
+      "learning_rate": 4.7343790476294005e-07,
+      "loss": 0.3132,
+      "step": 10429
+    },
+    {
+      "epoch": 2.623239436619718,
+      "grad_norm": 0.28822508454322815,
+      "learning_rate": 4.7281658958458877e-07,
+      "loss": 0.3114,
+      "step": 10430
+    },
+    {
+      "epoch": 2.6234909456740443,
+      "grad_norm": 0.29682502150535583,
+      "learning_rate": 4.7219566213117406e-07,
+      "loss": 0.326,
+      "step": 10431
+    },
+    {
+      "epoch": 2.62374245472837,
+      "grad_norm": 0.290158748626709,
+      "learning_rate": 4.7157512245587623e-07,
+      "loss": 0.3218,
+      "step": 10432
+    },
+    {
+      "epoch": 2.6239939637826963,
+      "grad_norm": 0.3001720607280731,
+      "learning_rate": 4.7095497061183826e-07,
+      "loss": 0.3339,
+      "step": 10433
+    },
+    {
+      "epoch": 2.624245472837022,
+      "grad_norm": 0.2894241511821747,
+      "learning_rate": 4.703352066521749e-07,
+      "loss": 0.3424,
+      "step": 10434
+    },
+    {
+      "epoch": 2.624496981891348,
+      "grad_norm": 0.2992819547653198,
+      "learning_rate": 4.697158306299621e-07,
+      "loss": 0.2913,
+      "step": 10435
+    },
+    {
+      "epoch": 2.624748490945674,
+      "grad_norm": 0.2854137420654297,
+      "learning_rate": 4.6909684259824785e-07,
+      "loss": 0.3385,
+      "step": 10436
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.28658536076545715,
+      "learning_rate": 4.6847824261004313e-07,
+      "loss": 0.3146,
+      "step": 10437
+    },
+    {
+      "epoch": 2.625251509054326,
+      "grad_norm": 0.2668541371822357,
+      "learning_rate": 4.6786003071832895e-07,
+      "loss": 0.3246,
+      "step": 10438
+    },
+    {
+      "epoch": 2.625503018108652,
+      "grad_norm": 0.3052661120891571,
+      "learning_rate": 4.6724220697604904e-07,
+      "loss": 0.3195,
+      "step": 10439
+    },
+    {
+      "epoch": 2.625754527162978,
+      "grad_norm": 0.2899504005908966,
+      "learning_rate": 4.666247714361183e-07,
+      "loss": 0.3266,
+      "step": 10440
+    },
+    {
+      "epoch": 2.6260060362173037,
+      "grad_norm": 0.3089952766895294,
+      "learning_rate": 4.6600772415141437e-07,
+      "loss": 0.3129,
+      "step": 10441
+    },
+    {
+      "epoch": 2.62625754527163,
+      "grad_norm": 0.281039834022522,
+      "learning_rate": 4.6539106517478394e-07,
+      "loss": 0.2899,
+      "step": 10442
+    },
+    {
+      "epoch": 2.6265090543259557,
+      "grad_norm": 0.2719120383262634,
+      "learning_rate": 4.647747945590414e-07,
+      "loss": 0.3245,
+      "step": 10443
+    },
+    {
+      "epoch": 2.626760563380282,
+      "grad_norm": 0.2826182544231415,
+      "learning_rate": 4.6415891235696453e-07,
+      "loss": 0.3268,
+      "step": 10444
+    },
+    {
+      "epoch": 2.6270120724346078,
+      "grad_norm": 0.30722248554229736,
+      "learning_rate": 4.635434186213017e-07,
+      "loss": 0.3347,
+      "step": 10445
+    },
+    {
+      "epoch": 2.6272635814889336,
+      "grad_norm": 0.2918623685836792,
+      "learning_rate": 4.6292831340476406e-07,
+      "loss": 0.3351,
+      "step": 10446
+    },
+    {
+      "epoch": 2.62751509054326,
+      "grad_norm": 0.29145464301109314,
+      "learning_rate": 4.623135967600334e-07,
+      "loss": 0.3048,
+      "step": 10447
+    },
+    {
+      "epoch": 2.6277665995975856,
+      "grad_norm": 0.298068106174469,
+      "learning_rate": 4.616992687397548e-07,
+      "loss": 0.2789,
+      "step": 10448
+    },
+    {
+      "epoch": 2.6280181086519114,
+      "grad_norm": 0.3069632649421692,
+      "learning_rate": 4.610853293965434e-07,
+      "loss": 0.3214,
+      "step": 10449
+    },
+    {
+      "epoch": 2.6282696177062377,
+      "grad_norm": 0.28947851061820984,
+      "learning_rate": 4.6047177878297654e-07,
+      "loss": 0.3121,
+      "step": 10450
+    },
+    {
+      "epoch": 2.6285211267605635,
+      "grad_norm": 0.2965468168258667,
+      "learning_rate": 4.5985861695160393e-07,
+      "loss": 0.3294,
+      "step": 10451
+    },
+    {
+      "epoch": 2.6287726358148893,
+      "grad_norm": 0.2832191288471222,
+      "learning_rate": 4.592458439549369e-07,
+      "loss": 0.3213,
+      "step": 10452
+    },
+    {
+      "epoch": 2.6290241448692155,
+      "grad_norm": 0.3064347803592682,
+      "learning_rate": 4.586334598454567e-07,
+      "loss": 0.3124,
+      "step": 10453
+    },
+    {
+      "epoch": 2.6292756539235413,
+      "grad_norm": 0.3119833469390869,
+      "learning_rate": 4.5802146467561035e-07,
+      "loss": 0.3171,
+      "step": 10454
+    },
+    {
+      "epoch": 2.629527162977867,
+      "grad_norm": 0.2910575568675995,
+      "learning_rate": 4.574098584978104e-07,
+      "loss": 0.2897,
+      "step": 10455
+    },
+    {
+      "epoch": 2.6297786720321934,
+      "grad_norm": 0.30219748616218567,
+      "learning_rate": 4.5679864136443874e-07,
+      "loss": 0.3135,
+      "step": 10456
+    },
+    {
+      "epoch": 2.630030181086519,
+      "grad_norm": 0.30393391847610474,
+      "learning_rate": 4.5618781332784026e-07,
+      "loss": 0.343,
+      "step": 10457
+    },
+    {
+      "epoch": 2.630281690140845,
+      "grad_norm": 0.3024074137210846,
+      "learning_rate": 4.5557737444033025e-07,
+      "loss": 0.3359,
+      "step": 10458
+    },
+    {
+      "epoch": 2.6305331991951713,
+      "grad_norm": 0.2811712324619293,
+      "learning_rate": 4.549673247541875e-07,
+      "loss": 0.3059,
+      "step": 10459
+    },
+    {
+      "epoch": 2.630784708249497,
+      "grad_norm": 0.2942906320095062,
+      "learning_rate": 4.543576643216607e-07,
+      "loss": 0.3207,
+      "step": 10460
+    },
+    {
+      "epoch": 2.631036217303823,
+      "grad_norm": 0.2751673758029938,
+      "learning_rate": 4.5374839319496156e-07,
+      "loss": 0.3229,
+      "step": 10461
+    },
+    {
+      "epoch": 2.631287726358149,
+      "grad_norm": 0.2935747504234314,
+      "learning_rate": 4.531395114262721e-07,
+      "loss": 0.3112,
+      "step": 10462
+    },
+    {
+      "epoch": 2.631539235412475,
+      "grad_norm": 0.27842676639556885,
+      "learning_rate": 4.525310190677379e-07,
+      "loss": 0.3132,
+      "step": 10463
+    },
+    {
+      "epoch": 2.6317907444668007,
+      "grad_norm": 0.27436089515686035,
+      "learning_rate": 4.5192291617147274e-07,
+      "loss": 0.3149,
+      "step": 10464
+    },
+    {
+      "epoch": 2.632042253521127,
+      "grad_norm": 0.27890241146087646,
+      "learning_rate": 4.5131520278955785e-07,
+      "loss": 0.2893,
+      "step": 10465
+    },
+    {
+      "epoch": 2.6322937625754528,
+      "grad_norm": 0.27693304419517517,
+      "learning_rate": 4.507078789740388e-07,
+      "loss": 0.315,
+      "step": 10466
+    },
+    {
+      "epoch": 2.6325452716297786,
+      "grad_norm": 0.309488445520401,
+      "learning_rate": 4.5010094477693057e-07,
+      "loss": 0.3331,
+      "step": 10467
+    },
+    {
+      "epoch": 2.632796780684105,
+      "grad_norm": 0.29617559909820557,
+      "learning_rate": 4.4949440025021105e-07,
+      "loss": 0.3234,
+      "step": 10468
+    },
+    {
+      "epoch": 2.6330482897384306,
+      "grad_norm": 0.2881315350532532,
+      "learning_rate": 4.488882454458299e-07,
+      "loss": 0.3431,
+      "step": 10469
+    },
+    {
+      "epoch": 2.6332997987927564,
+      "grad_norm": 0.3251035213470459,
+      "learning_rate": 4.482824804156971e-07,
+      "loss": 0.3014,
+      "step": 10470
+    },
+    {
+      "epoch": 2.6335513078470827,
+      "grad_norm": 0.3029896020889282,
+      "learning_rate": 4.476771052116957e-07,
+      "loss": 0.2974,
+      "step": 10471
+    },
+    {
+      "epoch": 2.6338028169014085,
+      "grad_norm": 0.2852858901023865,
+      "learning_rate": 4.4707211988567036e-07,
+      "loss": 0.2965,
+      "step": 10472
+    },
+    {
+      "epoch": 2.6340543259557343,
+      "grad_norm": 0.2950040400028229,
+      "learning_rate": 4.464675244894351e-07,
+      "loss": 0.3049,
+      "step": 10473
+    },
+    {
+      "epoch": 2.6343058350100605,
+      "grad_norm": 0.3013119399547577,
+      "learning_rate": 4.458633190747691e-07,
+      "loss": 0.3099,
+      "step": 10474
+    },
+    {
+      "epoch": 2.6345573440643864,
+      "grad_norm": 0.30011510848999023,
+      "learning_rate": 4.452595036934193e-07,
+      "loss": 0.3137,
+      "step": 10475
+    },
+    {
+      "epoch": 2.634808853118712,
+      "grad_norm": 0.2907141149044037,
+      "learning_rate": 4.4465607839709934e-07,
+      "loss": 0.3165,
+      "step": 10476
+    },
+    {
+      "epoch": 2.6350603621730384,
+      "grad_norm": 0.28316307067871094,
+      "learning_rate": 4.440530432374873e-07,
+      "loss": 0.3043,
+      "step": 10477
+    },
+    {
+      "epoch": 2.635311871227364,
+      "grad_norm": 0.315193772315979,
+      "learning_rate": 4.434503982662314e-07,
+      "loss": 0.3224,
+      "step": 10478
+    },
+    {
+      "epoch": 2.63556338028169,
+      "grad_norm": 0.27484557032585144,
+      "learning_rate": 4.4284814353494187e-07,
+      "loss": 0.2935,
+      "step": 10479
+    },
+    {
+      "epoch": 2.6358148893360163,
+      "grad_norm": 0.30151981115341187,
+      "learning_rate": 4.4224627909520034e-07,
+      "loss": 0.3119,
+      "step": 10480
+    },
+    {
+      "epoch": 2.636066398390342,
+      "grad_norm": 0.2999078631401062,
+      "learning_rate": 4.4164480499855114e-07,
+      "loss": 0.3054,
+      "step": 10481
+    },
+    {
+      "epoch": 2.636317907444668,
+      "grad_norm": 0.30388715863227844,
+      "learning_rate": 4.410437212965085e-07,
+      "loss": 0.3041,
+      "step": 10482
+    },
+    {
+      "epoch": 2.636569416498994,
+      "grad_norm": 0.2887566387653351,
+      "learning_rate": 4.404430280405492e-07,
+      "loss": 0.2967,
+      "step": 10483
+    },
+    {
+      "epoch": 2.63682092555332,
+      "grad_norm": 0.27546751499176025,
+      "learning_rate": 4.3984272528212077e-07,
+      "loss": 0.3144,
+      "step": 10484
+    },
+    {
+      "epoch": 2.6370724346076457,
+      "grad_norm": 0.2938852310180664,
+      "learning_rate": 4.3924281307263397e-07,
+      "loss": 0.3284,
+      "step": 10485
+    },
+    {
+      "epoch": 2.637323943661972,
+      "grad_norm": 0.30135706067085266,
+      "learning_rate": 4.3864329146346804e-07,
+      "loss": 0.3305,
+      "step": 10486
+    },
+    {
+      "epoch": 2.637575452716298,
+      "grad_norm": 0.2923177182674408,
+      "learning_rate": 4.3804416050596933e-07,
+      "loss": 0.3229,
+      "step": 10487
+    },
+    {
+      "epoch": 2.6378269617706236,
+      "grad_norm": 0.28316444158554077,
+      "learning_rate": 4.374454202514483e-07,
+      "loss": 0.3427,
+      "step": 10488
+    },
+    {
+      "epoch": 2.63807847082495,
+      "grad_norm": 0.28909602761268616,
+      "learning_rate": 4.3684707075118403e-07,
+      "loss": 0.3228,
+      "step": 10489
+    },
+    {
+      "epoch": 2.6383299798792756,
+      "grad_norm": 0.28920796513557434,
+      "learning_rate": 4.362491120564205e-07,
+      "loss": 0.3346,
+      "step": 10490
+    },
+    {
+      "epoch": 2.6385814889336014,
+      "grad_norm": 0.2695193290710449,
+      "learning_rate": 4.356515442183712e-07,
+      "loss": 0.3424,
+      "step": 10491
+    },
+    {
+      "epoch": 2.6388329979879277,
+      "grad_norm": 0.28552255034446716,
+      "learning_rate": 4.3505436728821125e-07,
+      "loss": 0.3054,
+      "step": 10492
+    },
+    {
+      "epoch": 2.6390845070422535,
+      "grad_norm": 0.27092212438583374,
+      "learning_rate": 4.344575813170876e-07,
+      "loss": 0.3152,
+      "step": 10493
+    },
+    {
+      "epoch": 2.6393360160965793,
+      "grad_norm": 0.29936766624450684,
+      "learning_rate": 4.3386118635610875e-07,
+      "loss": 0.3311,
+      "step": 10494
+    },
+    {
+      "epoch": 2.6395875251509056,
+      "grad_norm": 0.2581084966659546,
+      "learning_rate": 4.3326518245635494e-07,
+      "loss": 0.3198,
+      "step": 10495
+    },
+    {
+      "epoch": 2.6398390342052314,
+      "grad_norm": 0.27823901176452637,
+      "learning_rate": 4.326695696688682e-07,
+      "loss": 0.3174,
+      "step": 10496
+    },
+    {
+      "epoch": 2.640090543259557,
+      "grad_norm": 0.2890242636203766,
+      "learning_rate": 4.320743480446593e-07,
+      "loss": 0.3019,
+      "step": 10497
+    },
+    {
+      "epoch": 2.6403420523138834,
+      "grad_norm": 0.2868576645851135,
+      "learning_rate": 4.31479517634707e-07,
+      "loss": 0.3138,
+      "step": 10498
+    },
+    {
+      "epoch": 2.640593561368209,
+      "grad_norm": 0.3102249205112457,
+      "learning_rate": 4.308850784899521e-07,
+      "loss": 0.3264,
+      "step": 10499
+    },
+    {
+      "epoch": 2.640845070422535,
+      "grad_norm": 0.27739331126213074,
+      "learning_rate": 4.3029103066130673e-07,
+      "loss": 0.3361,
+      "step": 10500
+    },
+    {
+      "epoch": 2.6410965794768613,
+      "grad_norm": 0.28932255506515503,
+      "learning_rate": 4.296973741996463e-07,
+      "loss": 0.3229,
+      "step": 10501
+    },
+    {
+      "epoch": 2.641348088531187,
+      "grad_norm": 0.2557661235332489,
+      "learning_rate": 4.291041091558146e-07,
+      "loss": 0.3158,
+      "step": 10502
+    },
+    {
+      "epoch": 2.641599597585513,
+      "grad_norm": 0.282735675573349,
+      "learning_rate": 4.2851123558061927e-07,
+      "loss": 0.3341,
+      "step": 10503
+    },
+    {
+      "epoch": 2.641851106639839,
+      "grad_norm": 0.2887149155139923,
+      "learning_rate": 4.2791875352483857e-07,
+      "loss": 0.3283,
+      "step": 10504
+    },
+    {
+      "epoch": 2.642102615694165,
+      "grad_norm": 0.29206418991088867,
+      "learning_rate": 4.273266630392131e-07,
+      "loss": 0.3085,
+      "step": 10505
+    },
+    {
+      "epoch": 2.6423541247484907,
+      "grad_norm": 0.26142215728759766,
+      "learning_rate": 4.267349641744534e-07,
+      "loss": 0.3334,
+      "step": 10506
+    },
+    {
+      "epoch": 2.642605633802817,
+      "grad_norm": 0.2744672894477844,
+      "learning_rate": 4.261436569812322e-07,
+      "loss": 0.3323,
+      "step": 10507
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.2745734453201294,
+      "learning_rate": 4.2555274151019456e-07,
+      "loss": 0.335,
+      "step": 10508
+    },
+    {
+      "epoch": 2.6431086519114686,
+      "grad_norm": 0.2822088897228241,
+      "learning_rate": 4.2496221781194557e-07,
+      "loss": 0.3042,
+      "step": 10509
+    },
+    {
+      "epoch": 2.643360160965795,
+      "grad_norm": 0.2912130653858185,
+      "learning_rate": 4.2437208593706204e-07,
+      "loss": 0.3195,
+      "step": 10510
+    },
+    {
+      "epoch": 2.6436116700201207,
+      "grad_norm": 0.284824401140213,
+      "learning_rate": 4.2378234593608346e-07,
+      "loss": 0.3012,
+      "step": 10511
+    },
+    {
+      "epoch": 2.6438631790744465,
+      "grad_norm": 0.28413233160972595,
+      "learning_rate": 4.231929978595195e-07,
+      "loss": 0.3298,
+      "step": 10512
+    },
+    {
+      "epoch": 2.6441146881287727,
+      "grad_norm": 0.3148362636566162,
+      "learning_rate": 4.226040417578414e-07,
+      "loss": 0.3313,
+      "step": 10513
+    },
+    {
+      "epoch": 2.6443661971830985,
+      "grad_norm": 0.3029150366783142,
+      "learning_rate": 4.2201547768149277e-07,
+      "loss": 0.319,
+      "step": 10514
+    },
+    {
+      "epoch": 2.6446177062374243,
+      "grad_norm": 0.3129327595233917,
+      "learning_rate": 4.214273056808771e-07,
+      "loss": 0.2861,
+      "step": 10515
+    },
+    {
+      "epoch": 2.6448692152917506,
+      "grad_norm": 0.2981908917427063,
+      "learning_rate": 4.208395258063702e-07,
+      "loss": 0.3343,
+      "step": 10516
+    },
+    {
+      "epoch": 2.6451207243460764,
+      "grad_norm": 0.2968314588069916,
+      "learning_rate": 4.202521381083102e-07,
+      "loss": 0.3344,
+      "step": 10517
+    },
+    {
+      "epoch": 2.645372233400402,
+      "grad_norm": 0.30850720405578613,
+      "learning_rate": 4.196651426370041e-07,
+      "loss": 0.3308,
+      "step": 10518
+    },
+    {
+      "epoch": 2.6456237424547284,
+      "grad_norm": 0.2996951937675476,
+      "learning_rate": 4.190785394427238e-07,
+      "loss": 0.3206,
+      "step": 10519
+    },
+    {
+      "epoch": 2.6458752515090542,
+      "grad_norm": 0.2981989085674286,
+      "learning_rate": 4.184923285757092e-07,
+      "loss": 0.334,
+      "step": 10520
+    },
+    {
+      "epoch": 2.64612676056338,
+      "grad_norm": 0.26357463002204895,
+      "learning_rate": 4.1790651008616343e-07,
+      "loss": 0.335,
+      "step": 10521
+    },
+    {
+      "epoch": 2.6463782696177063,
+      "grad_norm": 0.2741795480251312,
+      "learning_rate": 4.1732108402426087e-07,
+      "loss": 0.3304,
+      "step": 10522
+    },
+    {
+      "epoch": 2.646629778672032,
+      "grad_norm": 0.292894572019577,
+      "learning_rate": 4.167360504401374e-07,
+      "loss": 0.3099,
+      "step": 10523
+    },
+    {
+      "epoch": 2.646881287726358,
+      "grad_norm": 0.30272337794303894,
+      "learning_rate": 4.161514093838992e-07,
+      "loss": 0.3106,
+      "step": 10524
+    },
+    {
+      "epoch": 2.647132796780684,
+      "grad_norm": 0.31396594643592834,
+      "learning_rate": 4.155671609056156e-07,
+      "loss": 0.314,
+      "step": 10525
+    },
+    {
+      "epoch": 2.64738430583501,
+      "grad_norm": 0.29578897356987,
+      "learning_rate": 4.1498330505532533e-07,
+      "loss": 0.3251,
+      "step": 10526
+    },
+    {
+      "epoch": 2.6476358148893357,
+      "grad_norm": 0.3138672709465027,
+      "learning_rate": 4.1439984188303027e-07,
+      "loss": 0.3185,
+      "step": 10527
+    },
+    {
+      "epoch": 2.647887323943662,
+      "grad_norm": 0.28872033953666687,
+      "learning_rate": 4.1381677143870147e-07,
+      "loss": 0.3277,
+      "step": 10528
+    },
+    {
+      "epoch": 2.648138832997988,
+      "grad_norm": 0.28270837664604187,
+      "learning_rate": 4.1323409377227565e-07,
+      "loss": 0.3312,
+      "step": 10529
+    },
+    {
+      "epoch": 2.6483903420523136,
+      "grad_norm": 0.3057926595211029,
+      "learning_rate": 4.1265180893365453e-07,
+      "loss": 0.3222,
+      "step": 10530
+    },
+    {
+      "epoch": 2.64864185110664,
+      "grad_norm": 0.30965495109558105,
+      "learning_rate": 4.1206991697270825e-07,
+      "loss": 0.3254,
+      "step": 10531
+    },
+    {
+      "epoch": 2.6488933601609657,
+      "grad_norm": 0.2973712086677551,
+      "learning_rate": 4.114884179392709e-07,
+      "loss": 0.332,
+      "step": 10532
+    },
+    {
+      "epoch": 2.649144869215292,
+      "grad_norm": 0.26278215646743774,
+      "learning_rate": 4.1090731188314583e-07,
+      "loss": 0.32,
+      "step": 10533
+    },
+    {
+      "epoch": 2.6493963782696177,
+      "grad_norm": 0.28730037808418274,
+      "learning_rate": 4.103265988540989e-07,
+      "loss": 0.3395,
+      "step": 10534
+    },
+    {
+      "epoch": 2.6496478873239435,
+      "grad_norm": 0.2832060754299164,
+      "learning_rate": 4.0974627890186745e-07,
+      "loss": 0.2925,
+      "step": 10535
+    },
+    {
+      "epoch": 2.6498993963782698,
+      "grad_norm": 0.28094425797462463,
+      "learning_rate": 4.0916635207614906e-07,
+      "loss": 0.3131,
+      "step": 10536
+    },
+    {
+      "epoch": 2.6501509054325956,
+      "grad_norm": 0.2945941984653473,
+      "learning_rate": 4.0858681842661395e-07,
+      "loss": 0.3151,
+      "step": 10537
+    },
+    {
+      "epoch": 2.6504024144869214,
+      "grad_norm": 0.28159478306770325,
+      "learning_rate": 4.080076780028924e-07,
+      "loss": 0.3046,
+      "step": 10538
+    },
+    {
+      "epoch": 2.6506539235412476,
+      "grad_norm": 0.3074738085269928,
+      "learning_rate": 4.0742893085458644e-07,
+      "loss": 0.3226,
+      "step": 10539
+    },
+    {
+      "epoch": 2.6509054325955734,
+      "grad_norm": 0.28362908959388733,
+      "learning_rate": 4.068505770312625e-07,
+      "loss": 0.3322,
+      "step": 10540
+    },
+    {
+      "epoch": 2.6511569416498992,
+      "grad_norm": 0.3007882833480835,
+      "learning_rate": 4.062726165824504e-07,
+      "loss": 0.3386,
+      "step": 10541
+    },
+    {
+      "epoch": 2.6514084507042255,
+      "grad_norm": 0.3021930158138275,
+      "learning_rate": 4.0569504955765227e-07,
+      "loss": 0.3079,
+      "step": 10542
+    },
+    {
+      "epoch": 2.6516599597585513,
+      "grad_norm": 0.2768537700176239,
+      "learning_rate": 4.0511787600632955e-07,
+      "loss": 0.3495,
+      "step": 10543
+    },
+    {
+      "epoch": 2.6519114688128775,
+      "grad_norm": 0.2638303339481354,
+      "learning_rate": 4.045410959779167e-07,
+      "loss": 0.3116,
+      "step": 10544
+    },
+    {
+      "epoch": 2.6521629778672033,
+      "grad_norm": 0.27994871139526367,
+      "learning_rate": 4.0396470952180857e-07,
+      "loss": 0.3143,
+      "step": 10545
+    },
+    {
+      "epoch": 2.652414486921529,
+      "grad_norm": 0.286160409450531,
+      "learning_rate": 4.033887166873712e-07,
+      "loss": 0.3084,
+      "step": 10546
+    },
+    {
+      "epoch": 2.6526659959758554,
+      "grad_norm": 0.28356102108955383,
+      "learning_rate": 4.028131175239336e-07,
+      "loss": 0.3367,
+      "step": 10547
+    },
+    {
+      "epoch": 2.652917505030181,
+      "grad_norm": 0.31378769874572754,
+      "learning_rate": 4.022379120807929e-07,
+      "loss": 0.3136,
+      "step": 10548
+    },
+    {
+      "epoch": 2.653169014084507,
+      "grad_norm": 0.29306289553642273,
+      "learning_rate": 4.016631004072108e-07,
+      "loss": 0.3319,
+      "step": 10549
+    },
+    {
+      "epoch": 2.6534205231388333,
+      "grad_norm": 0.2860982418060303,
+      "learning_rate": 4.010886825524174e-07,
+      "loss": 0.3403,
+      "step": 10550
+    },
+    {
+      "epoch": 2.653672032193159,
+      "grad_norm": 0.29843586683273315,
+      "learning_rate": 4.0051465856560836e-07,
+      "loss": 0.2994,
+      "step": 10551
+    },
+    {
+      "epoch": 2.653923541247485,
+      "grad_norm": 0.29828014969825745,
+      "learning_rate": 3.999410284959432e-07,
+      "loss": 0.3126,
+      "step": 10552
+    },
+    {
+      "epoch": 2.654175050301811,
+      "grad_norm": 0.2804122567176819,
+      "learning_rate": 3.9936779239255207e-07,
+      "loss": 0.3178,
+      "step": 10553
+    },
+    {
+      "epoch": 2.654426559356137,
+      "grad_norm": 0.29577210545539856,
+      "learning_rate": 3.987949503045274e-07,
+      "loss": 0.326,
+      "step": 10554
+    },
+    {
+      "epoch": 2.6546780684104627,
+      "grad_norm": 0.2835969030857086,
+      "learning_rate": 3.982225022809311e-07,
+      "loss": 0.3026,
+      "step": 10555
+    },
+    {
+      "epoch": 2.654929577464789,
+      "grad_norm": 0.27092668414115906,
+      "learning_rate": 3.9765044837078825e-07,
+      "loss": 0.2987,
+      "step": 10556
+    },
+    {
+      "epoch": 2.6551810865191148,
+      "grad_norm": 0.28689372539520264,
+      "learning_rate": 3.970787886230926e-07,
+      "loss": 0.3423,
+      "step": 10557
+    },
+    {
+      "epoch": 2.6554325955734406,
+      "grad_norm": 0.294495165348053,
+      "learning_rate": 3.965075230868026e-07,
+      "loss": 0.3245,
+      "step": 10558
+    },
+    {
+      "epoch": 2.655684104627767,
+      "grad_norm": 0.2807731628417969,
+      "learning_rate": 3.9593665181084427e-07,
+      "loss": 0.3214,
+      "step": 10559
+    },
+    {
+      "epoch": 2.6559356136820926,
+      "grad_norm": 0.29218292236328125,
+      "learning_rate": 3.953661748441079e-07,
+      "loss": 0.2978,
+      "step": 10560
+    },
+    {
+      "epoch": 2.6561871227364184,
+      "grad_norm": 0.3011685013771057,
+      "learning_rate": 3.947960922354527e-07,
+      "loss": 0.3007,
+      "step": 10561
+    },
+    {
+      "epoch": 2.6564386317907447,
+      "grad_norm": 0.2898184359073639,
+      "learning_rate": 3.9422640403370236e-07,
+      "loss": 0.3316,
+      "step": 10562
+    },
+    {
+      "epoch": 2.6566901408450705,
+      "grad_norm": 0.28175485134124756,
+      "learning_rate": 3.936571102876463e-07,
+      "loss": 0.319,
+      "step": 10563
+    },
+    {
+      "epoch": 2.6569416498993963,
+      "grad_norm": 0.2649247944355011,
+      "learning_rate": 3.930882110460421e-07,
+      "loss": 0.3151,
+      "step": 10564
+    },
+    {
+      "epoch": 2.6571931589537225,
+      "grad_norm": 0.2886830270290375,
+      "learning_rate": 3.925197063576114e-07,
+      "loss": 0.3035,
+      "step": 10565
+    },
+    {
+      "epoch": 2.6574446680080483,
+      "grad_norm": 0.28880345821380615,
+      "learning_rate": 3.9195159627104465e-07,
+      "loss": 0.3268,
+      "step": 10566
+    },
+    {
+      "epoch": 2.657696177062374,
+      "grad_norm": 0.2972525954246521,
+      "learning_rate": 3.913838808349946e-07,
+      "loss": 0.3078,
+      "step": 10567
+    },
+    {
+      "epoch": 2.6579476861167004,
+      "grad_norm": 0.298945814371109,
+      "learning_rate": 3.908165600980845e-07,
+      "loss": 0.3142,
+      "step": 10568
+    },
+    {
+      "epoch": 2.658199195171026,
+      "grad_norm": 0.3075450658798218,
+      "learning_rate": 3.9024963410890015e-07,
+      "loss": 0.3152,
+      "step": 10569
+    },
+    {
+      "epoch": 2.658450704225352,
+      "grad_norm": 0.31689146161079407,
+      "learning_rate": 3.8968310291599753e-07,
+      "loss": 0.3172,
+      "step": 10570
+    },
+    {
+      "epoch": 2.6587022132796783,
+      "grad_norm": 0.2927457094192505,
+      "learning_rate": 3.89116966567894e-07,
+      "loss": 0.3291,
+      "step": 10571
+    },
+    {
+      "epoch": 2.658953722334004,
+      "grad_norm": 0.2956581115722656,
+      "learning_rate": 3.885512251130763e-07,
+      "loss": 0.3041,
+      "step": 10572
+    },
+    {
+      "epoch": 2.65920523138833,
+      "grad_norm": 0.29633039236068726,
+      "learning_rate": 3.879858785999979e-07,
+      "loss": 0.3256,
+      "step": 10573
+    },
+    {
+      "epoch": 2.659456740442656,
+      "grad_norm": 0.2878909707069397,
+      "learning_rate": 3.874209270770762e-07,
+      "loss": 0.3356,
+      "step": 10574
+    },
+    {
+      "epoch": 2.659708249496982,
+      "grad_norm": 0.3077471852302551,
+      "learning_rate": 3.8685637059269587e-07,
+      "loss": 0.3131,
+      "step": 10575
+    },
+    {
+      "epoch": 2.6599597585513077,
+      "grad_norm": 0.28095880150794983,
+      "learning_rate": 3.862922091952076e-07,
+      "loss": 0.3165,
+      "step": 10576
+    },
+    {
+      "epoch": 2.660211267605634,
+      "grad_norm": 0.3052351772785187,
+      "learning_rate": 3.857284429329289e-07,
+      "loss": 0.3114,
+      "step": 10577
+    },
+    {
+      "epoch": 2.66046277665996,
+      "grad_norm": 0.3149491250514984,
+      "learning_rate": 3.851650718541411e-07,
+      "loss": 0.3186,
+      "step": 10578
+    },
+    {
+      "epoch": 2.6607142857142856,
+      "grad_norm": 0.30292919278144836,
+      "learning_rate": 3.846020960070956e-07,
+      "loss": 0.3441,
+      "step": 10579
+    },
+    {
+      "epoch": 2.660965794768612,
+      "grad_norm": 0.3036801218986511,
+      "learning_rate": 3.8403951544000617e-07,
+      "loss": 0.3152,
+      "step": 10580
+    },
+    {
+      "epoch": 2.6612173038229376,
+      "grad_norm": 0.2941226661205292,
+      "learning_rate": 3.834773302010553e-07,
+      "loss": 0.3112,
+      "step": 10581
+    },
+    {
+      "epoch": 2.6614688128772634,
+      "grad_norm": 0.30887505412101746,
+      "learning_rate": 3.829155403383894e-07,
+      "loss": 0.3293,
+      "step": 10582
+    },
+    {
+      "epoch": 2.6617203219315897,
+      "grad_norm": 0.28087976574897766,
+      "learning_rate": 3.823541459001234e-07,
+      "loss": 0.3246,
+      "step": 10583
+    },
+    {
+      "epoch": 2.6619718309859155,
+      "grad_norm": 0.28773432970046997,
+      "learning_rate": 3.8179314693433775e-07,
+      "loss": 0.3033,
+      "step": 10584
+    },
+    {
+      "epoch": 2.6622233400402413,
+      "grad_norm": 0.27361011505126953,
+      "learning_rate": 3.8123254348907676e-07,
+      "loss": 0.3338,
+      "step": 10585
+    },
+    {
+      "epoch": 2.6624748490945676,
+      "grad_norm": 0.28717923164367676,
+      "learning_rate": 3.806723356123543e-07,
+      "loss": 0.3316,
+      "step": 10586
+    },
+    {
+      "epoch": 2.6627263581488934,
+      "grad_norm": 0.2967682182788849,
+      "learning_rate": 3.801125233521469e-07,
+      "loss": 0.3631,
+      "step": 10587
+    },
+    {
+      "epoch": 2.662977867203219,
+      "grad_norm": 0.25790077447891235,
+      "learning_rate": 3.7955310675640066e-07,
+      "loss": 0.3155,
+      "step": 10588
+    },
+    {
+      "epoch": 2.6632293762575454,
+      "grad_norm": 0.25773194432258606,
+      "learning_rate": 3.789940858730251e-07,
+      "loss": 0.2978,
+      "step": 10589
+    },
+    {
+      "epoch": 2.663480885311871,
+      "grad_norm": 0.295448899269104,
+      "learning_rate": 3.7843546074989747e-07,
+      "loss": 0.3467,
+      "step": 10590
+    },
+    {
+      "epoch": 2.663732394366197,
+      "grad_norm": 0.27285072207450867,
+      "learning_rate": 3.778772314348594e-07,
+      "loss": 0.3165,
+      "step": 10591
+    },
+    {
+      "epoch": 2.6639839034205233,
+      "grad_norm": 0.2875916361808777,
+      "learning_rate": 3.773193979757217e-07,
+      "loss": 0.2996,
+      "step": 10592
+    },
+    {
+      "epoch": 2.664235412474849,
+      "grad_norm": 0.3071140646934509,
+      "learning_rate": 3.7676196042025715e-07,
+      "loss": 0.3157,
+      "step": 10593
+    },
+    {
+      "epoch": 2.664486921529175,
+      "grad_norm": 0.3103017807006836,
+      "learning_rate": 3.7620491881620814e-07,
+      "loss": 0.3411,
+      "step": 10594
+    },
+    {
+      "epoch": 2.664738430583501,
+      "grad_norm": 0.3201421797275543,
+      "learning_rate": 3.7564827321128203e-07,
+      "loss": 0.3409,
+      "step": 10595
+    },
+    {
+      "epoch": 2.664989939637827,
+      "grad_norm": 0.2804863154888153,
+      "learning_rate": 3.750920236531502e-07,
+      "loss": 0.3188,
+      "step": 10596
+    },
+    {
+      "epoch": 2.6652414486921527,
+      "grad_norm": 0.30754902958869934,
+      "learning_rate": 3.7453617018945453e-07,
+      "loss": 0.3183,
+      "step": 10597
+    },
+    {
+      "epoch": 2.665492957746479,
+      "grad_norm": 0.2903851866722107,
+      "learning_rate": 3.739807128677986e-07,
+      "loss": 0.3096,
+      "step": 10598
+    },
+    {
+      "epoch": 2.665744466800805,
+      "grad_norm": 0.2837523818016052,
+      "learning_rate": 3.734256517357543e-07,
+      "loss": 0.3035,
+      "step": 10599
+    },
+    {
+      "epoch": 2.6659959758551306,
+      "grad_norm": 0.3047906160354614,
+      "learning_rate": 3.7287098684085867e-07,
+      "loss": 0.3157,
+      "step": 10600
+    },
+    {
+      "epoch": 2.666247484909457,
+      "grad_norm": 0.2783104181289673,
+      "learning_rate": 3.72316718230617e-07,
+      "loss": 0.3306,
+      "step": 10601
+    },
+    {
+      "epoch": 2.6664989939637826,
+      "grad_norm": 0.27863505482673645,
+      "learning_rate": 3.717628459524963e-07,
+      "loss": 0.3356,
+      "step": 10602
+    },
+    {
+      "epoch": 2.6667505030181085,
+      "grad_norm": 0.2805593013763428,
+      "learning_rate": 3.7120937005393487e-07,
+      "loss": 0.3201,
+      "step": 10603
+    },
+    {
+      "epoch": 2.6670020120724347,
+      "grad_norm": 0.30254265666007996,
+      "learning_rate": 3.7065629058233245e-07,
+      "loss": 0.2921,
+      "step": 10604
+    },
+    {
+      "epoch": 2.6672535211267605,
+      "grad_norm": 0.30693769454956055,
+      "learning_rate": 3.701036075850578e-07,
+      "loss": 0.3277,
+      "step": 10605
+    },
+    {
+      "epoch": 2.6675050301810863,
+      "grad_norm": 0.29164475202560425,
+      "learning_rate": 3.695513211094448e-07,
+      "loss": 0.3082,
+      "step": 10606
+    },
+    {
+      "epoch": 2.6677565392354126,
+      "grad_norm": 0.2975282371044159,
+      "learning_rate": 3.689994312027928e-07,
+      "loss": 0.3147,
+      "step": 10607
+    },
+    {
+      "epoch": 2.6680080482897384,
+      "grad_norm": 0.2964397072792053,
+      "learning_rate": 3.6844793791236897e-07,
+      "loss": 0.3245,
+      "step": 10608
+    },
+    {
+      "epoch": 2.668259557344064,
+      "grad_norm": 0.29004722833633423,
+      "learning_rate": 3.678968412854034e-07,
+      "loss": 0.3261,
+      "step": 10609
+    },
+    {
+      "epoch": 2.6685110663983904,
+      "grad_norm": 0.27896592020988464,
+      "learning_rate": 3.67346141369096e-07,
+      "loss": 0.3172,
+      "step": 10610
+    },
+    {
+      "epoch": 2.6687625754527162,
+      "grad_norm": 0.2943844497203827,
+      "learning_rate": 3.6679583821060904e-07,
+      "loss": 0.3059,
+      "step": 10611
+    },
+    {
+      "epoch": 2.669014084507042,
+      "grad_norm": 0.3017599880695343,
+      "learning_rate": 3.662459318570738e-07,
+      "loss": 0.3441,
+      "step": 10612
+    },
+    {
+      "epoch": 2.6692655935613683,
+      "grad_norm": 0.29490116238594055,
+      "learning_rate": 3.656964223555848e-07,
+      "loss": 0.3213,
+      "step": 10613
+    },
+    {
+      "epoch": 2.669517102615694,
+      "grad_norm": 0.30418360233306885,
+      "learning_rate": 3.6514730975320657e-07,
+      "loss": 0.3226,
+      "step": 10614
+    },
+    {
+      "epoch": 2.66976861167002,
+      "grad_norm": 0.2704751789569855,
+      "learning_rate": 3.645985940969643e-07,
+      "loss": 0.3592,
+      "step": 10615
+    },
+    {
+      "epoch": 2.670020120724346,
+      "grad_norm": 0.27038201689720154,
+      "learning_rate": 3.6405027543385374e-07,
+      "loss": 0.3128,
+      "step": 10616
+    },
+    {
+      "epoch": 2.670271629778672,
+      "grad_norm": 0.3021993637084961,
+      "learning_rate": 3.6350235381083563e-07,
+      "loss": 0.3064,
+      "step": 10617
+    },
+    {
+      "epoch": 2.6705231388329977,
+      "grad_norm": 0.2776448428630829,
+      "learning_rate": 3.629548292748342e-07,
+      "loss": 0.3311,
+      "step": 10618
+    },
+    {
+      "epoch": 2.670774647887324,
+      "grad_norm": 0.2895011901855469,
+      "learning_rate": 3.624077018727429e-07,
+      "loss": 0.2909,
+      "step": 10619
+    },
+    {
+      "epoch": 2.67102615694165,
+      "grad_norm": 0.27862903475761414,
+      "learning_rate": 3.6186097165141875e-07,
+      "loss": 0.3213,
+      "step": 10620
+    },
+    {
+      "epoch": 2.6712776659959756,
+      "grad_norm": 0.3124517798423767,
+      "learning_rate": 3.6131463865768714e-07,
+      "loss": 0.3084,
+      "step": 10621
+    },
+    {
+      "epoch": 2.671529175050302,
+      "grad_norm": 0.28285762667655945,
+      "learning_rate": 3.607687029383361e-07,
+      "loss": 0.3213,
+      "step": 10622
+    },
+    {
+      "epoch": 2.6717806841046277,
+      "grad_norm": 0.30333417654037476,
+      "learning_rate": 3.6022316454012387e-07,
+      "loss": 0.3243,
+      "step": 10623
+    },
+    {
+      "epoch": 2.6720321931589535,
+      "grad_norm": 0.2860763370990753,
+      "learning_rate": 3.5967802350977024e-07,
+      "loss": 0.3055,
+      "step": 10624
+    },
+    {
+      "epoch": 2.6722837022132797,
+      "grad_norm": 0.3174950182437897,
+      "learning_rate": 3.591332798939651e-07,
+      "loss": 0.3369,
+      "step": 10625
+    },
+    {
+      "epoch": 2.6725352112676055,
+      "grad_norm": 0.3028983175754547,
+      "learning_rate": 3.5858893373936e-07,
+      "loss": 0.3352,
+      "step": 10626
+    },
+    {
+      "epoch": 2.6727867203219313,
+      "grad_norm": 0.27620789408683777,
+      "learning_rate": 3.5804498509257766e-07,
+      "loss": 0.3241,
+      "step": 10627
+    },
+    {
+      "epoch": 2.6730382293762576,
+      "grad_norm": 0.2942667007446289,
+      "learning_rate": 3.575014340002009e-07,
+      "loss": 0.3004,
+      "step": 10628
+    },
+    {
+      "epoch": 2.6732897384305834,
+      "grad_norm": 0.300465852022171,
+      "learning_rate": 3.56958280508784e-07,
+      "loss": 0.3163,
+      "step": 10629
+    },
+    {
+      "epoch": 2.6735412474849096,
+      "grad_norm": 0.30546244978904724,
+      "learning_rate": 3.564155246648432e-07,
+      "loss": 0.3497,
+      "step": 10630
+    },
+    {
+      "epoch": 2.6737927565392354,
+      "grad_norm": 0.27925771474838257,
+      "learning_rate": 3.558731665148629e-07,
+      "loss": 0.3144,
+      "step": 10631
+    },
+    {
+      "epoch": 2.6740442655935612,
+      "grad_norm": 0.2924439013004303,
+      "learning_rate": 3.5533120610529157e-07,
+      "loss": 0.3182,
+      "step": 10632
+    },
+    {
+      "epoch": 2.6742957746478875,
+      "grad_norm": 0.27821362018585205,
+      "learning_rate": 3.547896434825465e-07,
+      "loss": 0.3016,
+      "step": 10633
+    },
+    {
+      "epoch": 2.6745472837022133,
+      "grad_norm": 0.3052501976490021,
+      "learning_rate": 3.542484786930073e-07,
+      "loss": 0.3394,
+      "step": 10634
+    },
+    {
+      "epoch": 2.674798792756539,
+      "grad_norm": 0.3052050769329071,
+      "learning_rate": 3.537077117830229e-07,
+      "loss": 0.2972,
+      "step": 10635
+    },
+    {
+      "epoch": 2.6750503018108653,
+      "grad_norm": 0.2863207757472992,
+      "learning_rate": 3.531673427989046e-07,
+      "loss": 0.3262,
+      "step": 10636
+    },
+    {
+      "epoch": 2.675301810865191,
+      "grad_norm": 0.26985007524490356,
+      "learning_rate": 3.526273717869344e-07,
+      "loss": 0.3168,
+      "step": 10637
+    },
+    {
+      "epoch": 2.675553319919517,
+      "grad_norm": 0.31778794527053833,
+      "learning_rate": 3.5208779879335465e-07,
+      "loss": 0.3115,
+      "step": 10638
+    },
+    {
+      "epoch": 2.675804828973843,
+      "grad_norm": 0.2770874500274658,
+      "learning_rate": 3.5154862386437894e-07,
+      "loss": 0.3377,
+      "step": 10639
+    },
+    {
+      "epoch": 2.676056338028169,
+      "grad_norm": 0.2914242148399353,
+      "learning_rate": 3.5100984704618145e-07,
+      "loss": 0.3152,
+      "step": 10640
+    },
+    {
+      "epoch": 2.676307847082495,
+      "grad_norm": 0.313717246055603,
+      "learning_rate": 3.504714683849081e-07,
+      "loss": 0.3311,
+      "step": 10641
+    },
+    {
+      "epoch": 2.676559356136821,
+      "grad_norm": 0.31820422410964966,
+      "learning_rate": 3.499334879266653e-07,
+      "loss": 0.3342,
+      "step": 10642
+    },
+    {
+      "epoch": 2.676810865191147,
+      "grad_norm": 0.2878066897392273,
+      "learning_rate": 3.4939590571752893e-07,
+      "loss": 0.3092,
+      "step": 10643
+    },
+    {
+      "epoch": 2.677062374245473,
+      "grad_norm": 0.32320746779441833,
+      "learning_rate": 3.488587218035383e-07,
+      "loss": 0.3276,
+      "step": 10644
+    },
+    {
+      "epoch": 2.677313883299799,
+      "grad_norm": 0.26483622193336487,
+      "learning_rate": 3.4832193623070167e-07,
+      "loss": 0.2976,
+      "step": 10645
+    },
+    {
+      "epoch": 2.6775653923541247,
+      "grad_norm": 0.29713913798332214,
+      "learning_rate": 3.4778554904498996e-07,
+      "loss": 0.3227,
+      "step": 10646
+    },
+    {
+      "epoch": 2.677816901408451,
+      "grad_norm": 0.2784908711910248,
+      "learning_rate": 3.47249560292342e-07,
+      "loss": 0.3209,
+      "step": 10647
+    },
+    {
+      "epoch": 2.6780684104627768,
+      "grad_norm": 0.3123219609260559,
+      "learning_rate": 3.4671397001866116e-07,
+      "loss": 0.3209,
+      "step": 10648
+    },
+    {
+      "epoch": 2.6783199195171026,
+      "grad_norm": 0.27715906500816345,
+      "learning_rate": 3.4617877826981785e-07,
+      "loss": 0.3067,
+      "step": 10649
+    },
+    {
+      "epoch": 2.678571428571429,
+      "grad_norm": 0.2937963902950287,
+      "learning_rate": 3.4564398509164877e-07,
+      "loss": 0.3111,
+      "step": 10650
+    },
+    {
+      "epoch": 2.6788229376257546,
+      "grad_norm": 0.2874857783317566,
+      "learning_rate": 3.451095905299545e-07,
+      "loss": 0.3135,
+      "step": 10651
+    },
+    {
+      "epoch": 2.6790744466800804,
+      "grad_norm": 0.2816859781742096,
+      "learning_rate": 3.4457559463050293e-07,
+      "loss": 0.3099,
+      "step": 10652
+    },
+    {
+      "epoch": 2.6793259557344067,
+      "grad_norm": 0.28699028491973877,
+      "learning_rate": 3.440419974390269e-07,
+      "loss": 0.3174,
+      "step": 10653
+    },
+    {
+      "epoch": 2.6795774647887325,
+      "grad_norm": 0.30524781346321106,
+      "learning_rate": 3.43508799001227e-07,
+      "loss": 0.3135,
+      "step": 10654
+    },
+    {
+      "epoch": 2.6798289738430583,
+      "grad_norm": 0.2902679741382599,
+      "learning_rate": 3.429759993627674e-07,
+      "loss": 0.3229,
+      "step": 10655
+    },
+    {
+      "epoch": 2.6800804828973845,
+      "grad_norm": 0.331782728433609,
+      "learning_rate": 3.424435985692792e-07,
+      "loss": 0.3208,
+      "step": 10656
+    },
+    {
+      "epoch": 2.6803319919517103,
+      "grad_norm": 0.2915107309818268,
+      "learning_rate": 3.419115966663583e-07,
+      "loss": 0.3151,
+      "step": 10657
+    },
+    {
+      "epoch": 2.680583501006036,
+      "grad_norm": 0.3128476142883301,
+      "learning_rate": 3.413799936995699e-07,
+      "loss": 0.3463,
+      "step": 10658
+    },
+    {
+      "epoch": 2.6808350100603624,
+      "grad_norm": 0.2803356349468231,
+      "learning_rate": 3.4084878971443914e-07,
+      "loss": 0.3348,
+      "step": 10659
+    },
+    {
+      "epoch": 2.681086519114688,
+      "grad_norm": 0.2853963375091553,
+      "learning_rate": 3.403179847564625e-07,
+      "loss": 0.3282,
+      "step": 10660
+    },
+    {
+      "epoch": 2.681338028169014,
+      "grad_norm": 0.28760576248168945,
+      "learning_rate": 3.397875788711003e-07,
+      "loss": 0.3046,
+      "step": 10661
+    },
+    {
+      "epoch": 2.6815895372233403,
+      "grad_norm": 0.298556923866272,
+      "learning_rate": 3.3925757210377664e-07,
+      "loss": 0.2909,
+      "step": 10662
+    },
+    {
+      "epoch": 2.681841046277666,
+      "grad_norm": 0.2910586893558502,
+      "learning_rate": 3.387279644998853e-07,
+      "loss": 0.3091,
+      "step": 10663
+    },
+    {
+      "epoch": 2.682092555331992,
+      "grad_norm": 0.3114005923271179,
+      "learning_rate": 3.3819875610478225e-07,
+      "loss": 0.3163,
+      "step": 10664
+    },
+    {
+      "epoch": 2.682344064386318,
+      "grad_norm": 0.29191139340400696,
+      "learning_rate": 3.3766994696379283e-07,
+      "loss": 0.3394,
+      "step": 10665
+    },
+    {
+      "epoch": 2.682595573440644,
+      "grad_norm": 0.2941957116127014,
+      "learning_rate": 3.3714153712220364e-07,
+      "loss": 0.3197,
+      "step": 10666
+    },
+    {
+      "epoch": 2.6828470824949697,
+      "grad_norm": 0.269555002450943,
+      "learning_rate": 3.3661352662527234e-07,
+      "loss": 0.2953,
+      "step": 10667
+    },
+    {
+      "epoch": 2.683098591549296,
+      "grad_norm": 0.30141013860702515,
+      "learning_rate": 3.3608591551821724e-07,
+      "loss": 0.3079,
+      "step": 10668
+    },
+    {
+      "epoch": 2.683350100603622,
+      "grad_norm": 0.3198448121547699,
+      "learning_rate": 3.35558703846226e-07,
+      "loss": 0.3246,
+      "step": 10669
+    },
+    {
+      "epoch": 2.6836016096579476,
+      "grad_norm": 0.3114519715309143,
+      "learning_rate": 3.3503189165445205e-07,
+      "loss": 0.3223,
+      "step": 10670
+    },
+    {
+      "epoch": 2.683853118712274,
+      "grad_norm": 0.2955555319786072,
+      "learning_rate": 3.34505478988012e-07,
+      "loss": 0.3222,
+      "step": 10671
+    },
+    {
+      "epoch": 2.6841046277665996,
+      "grad_norm": 0.31283295154571533,
+      "learning_rate": 3.339794658919915e-07,
+      "loss": 0.3082,
+      "step": 10672
+    },
+    {
+      "epoch": 2.6843561368209254,
+      "grad_norm": 0.2927229702472687,
+      "learning_rate": 3.334538524114378e-07,
+      "loss": 0.3144,
+      "step": 10673
+    },
+    {
+      "epoch": 2.6846076458752517,
+      "grad_norm": 0.2944906949996948,
+      "learning_rate": 3.329286385913688e-07,
+      "loss": 0.3193,
+      "step": 10674
+    },
+    {
+      "epoch": 2.6848591549295775,
+      "grad_norm": 0.2984817326068878,
+      "learning_rate": 3.3240382447676413e-07,
+      "loss": 0.3145,
+      "step": 10675
+    },
+    {
+      "epoch": 2.6851106639839033,
+      "grad_norm": 0.28597378730773926,
+      "learning_rate": 3.3187941011257217e-07,
+      "loss": 0.3065,
+      "step": 10676
+    },
+    {
+      "epoch": 2.6853621730382295,
+      "grad_norm": 0.2896621525287628,
+      "learning_rate": 3.313553955437049e-07,
+      "loss": 0.3322,
+      "step": 10677
+    },
+    {
+      "epoch": 2.6856136820925554,
+      "grad_norm": 0.29271525144577026,
+      "learning_rate": 3.308317808150413e-07,
+      "loss": 0.3155,
+      "step": 10678
+    },
+    {
+      "epoch": 2.685865191146881,
+      "grad_norm": 0.2880185544490814,
+      "learning_rate": 3.303085659714245e-07,
+      "loss": 0.3204,
+      "step": 10679
+    },
+    {
+      "epoch": 2.6861167002012074,
+      "grad_norm": 0.29037216305732727,
+      "learning_rate": 3.297857510576658e-07,
+      "loss": 0.3253,
+      "step": 10680
+    },
+    {
+      "epoch": 2.686368209255533,
+      "grad_norm": 0.2900163531303406,
+      "learning_rate": 3.29263336118541e-07,
+      "loss": 0.3198,
+      "step": 10681
+    },
+    {
+      "epoch": 2.686619718309859,
+      "grad_norm": 0.28541356325149536,
+      "learning_rate": 3.287413211987911e-07,
+      "loss": 0.3171,
+      "step": 10682
+    },
+    {
+      "epoch": 2.6868712273641853,
+      "grad_norm": 0.3070562183856964,
+      "learning_rate": 3.2821970634312474e-07,
+      "loss": 0.3107,
+      "step": 10683
+    },
+    {
+      "epoch": 2.687122736418511,
+      "grad_norm": 0.296577513217926,
+      "learning_rate": 3.276984915962128e-07,
+      "loss": 0.3507,
+      "step": 10684
+    },
+    {
+      "epoch": 2.687374245472837,
+      "grad_norm": 0.286769300699234,
+      "learning_rate": 3.271776770026963e-07,
+      "loss": 0.3122,
+      "step": 10685
+    },
+    {
+      "epoch": 2.687625754527163,
+      "grad_norm": 0.31535083055496216,
+      "learning_rate": 3.266572626071773e-07,
+      "loss": 0.3166,
+      "step": 10686
+    },
+    {
+      "epoch": 2.687877263581489,
+      "grad_norm": 0.2796669602394104,
+      "learning_rate": 3.261372484542291e-07,
+      "loss": 0.3233,
+      "step": 10687
+    },
+    {
+      "epoch": 2.6881287726358147,
+      "grad_norm": 0.29072704911231995,
+      "learning_rate": 3.2561763458838436e-07,
+      "loss": 0.3206,
+      "step": 10688
+    },
+    {
+      "epoch": 2.688380281690141,
+      "grad_norm": 0.291552871465683,
+      "learning_rate": 3.25098421054148e-07,
+      "loss": 0.3319,
+      "step": 10689
+    },
+    {
+      "epoch": 2.688631790744467,
+      "grad_norm": 0.2940013110637665,
+      "learning_rate": 3.2457960789598453e-07,
+      "loss": 0.3319,
+      "step": 10690
+    },
+    {
+      "epoch": 2.6888832997987926,
+      "grad_norm": 0.2845515310764313,
+      "learning_rate": 3.2406119515832777e-07,
+      "loss": 0.3211,
+      "step": 10691
+    },
+    {
+      "epoch": 2.689134808853119,
+      "grad_norm": 0.30485305190086365,
+      "learning_rate": 3.235431828855784e-07,
+      "loss": 0.3121,
+      "step": 10692
+    },
+    {
+      "epoch": 2.6893863179074446,
+      "grad_norm": 0.2907279431819916,
+      "learning_rate": 3.230255711220992e-07,
+      "loss": 0.3204,
+      "step": 10693
+    },
+    {
+      "epoch": 2.6896378269617705,
+      "grad_norm": 0.2880786061286926,
+      "learning_rate": 3.2250835991222084e-07,
+      "loss": 0.3015,
+      "step": 10694
+    },
+    {
+      "epoch": 2.6898893360160967,
+      "grad_norm": 0.28251487016677856,
+      "learning_rate": 3.2199154930023903e-07,
+      "loss": 0.3232,
+      "step": 10695
+    },
+    {
+      "epoch": 2.6901408450704225,
+      "grad_norm": 0.27719196677207947,
+      "learning_rate": 3.214751393304155e-07,
+      "loss": 0.3025,
+      "step": 10696
+    },
+    {
+      "epoch": 2.6903923541247483,
+      "grad_norm": 0.2866508364677429,
+      "learning_rate": 3.209591300469772e-07,
+      "loss": 0.3137,
+      "step": 10697
+    },
+    {
+      "epoch": 2.6906438631790746,
+      "grad_norm": 0.269786536693573,
+      "learning_rate": 3.204435214941182e-07,
+      "loss": 0.3266,
+      "step": 10698
+    },
+    {
+      "epoch": 2.6908953722334004,
+      "grad_norm": 0.27028945088386536,
+      "learning_rate": 3.1992831371599584e-07,
+      "loss": 0.3135,
+      "step": 10699
+    },
+    {
+      "epoch": 2.691146881287726,
+      "grad_norm": 0.2978735864162445,
+      "learning_rate": 3.1941350675673557e-07,
+      "loss": 0.3323,
+      "step": 10700
+    },
+    {
+      "epoch": 2.6913983903420524,
+      "grad_norm": 0.3069797456264496,
+      "learning_rate": 3.1889910066042587e-07,
+      "loss": 0.3128,
+      "step": 10701
+    },
+    {
+      "epoch": 2.691649899396378,
+      "grad_norm": 0.2964140474796295,
+      "learning_rate": 3.183850954711232e-07,
+      "loss": 0.3123,
+      "step": 10702
+    },
+    {
+      "epoch": 2.691901408450704,
+      "grad_norm": 0.3021068572998047,
+      "learning_rate": 3.178714912328501e-07,
+      "loss": 0.287,
+      "step": 10703
+    },
+    {
+      "epoch": 2.6921529175050303,
+      "grad_norm": 0.2956176698207855,
+      "learning_rate": 3.1735828798959146e-07,
+      "loss": 0.3147,
+      "step": 10704
+    },
+    {
+      "epoch": 2.692404426559356,
+      "grad_norm": 0.27685198187828064,
+      "learning_rate": 3.168454857853015e-07,
+      "loss": 0.3162,
+      "step": 10705
+    },
+    {
+      "epoch": 2.692655935613682,
+      "grad_norm": 0.285156786441803,
+      "learning_rate": 3.163330846638979e-07,
+      "loss": 0.324,
+      "step": 10706
+    },
+    {
+      "epoch": 2.692907444668008,
+      "grad_norm": 0.31111207604408264,
+      "learning_rate": 3.1582108466926497e-07,
+      "loss": 0.3137,
+      "step": 10707
+    },
+    {
+      "epoch": 2.693158953722334,
+      "grad_norm": 0.32243606448173523,
+      "learning_rate": 3.15309485845251e-07,
+      "loss": 0.3186,
+      "step": 10708
+    },
+    {
+      "epoch": 2.6934104627766597,
+      "grad_norm": 0.2985250651836395,
+      "learning_rate": 3.147982882356732e-07,
+      "loss": 0.3265,
+      "step": 10709
+    },
+    {
+      "epoch": 2.693661971830986,
+      "grad_norm": 0.2669808566570282,
+      "learning_rate": 3.1428749188431086e-07,
+      "loss": 0.3053,
+      "step": 10710
+    },
+    {
+      "epoch": 2.693913480885312,
+      "grad_norm": 0.3001328706741333,
+      "learning_rate": 3.1377709683491185e-07,
+      "loss": 0.3213,
+      "step": 10711
+    },
+    {
+      "epoch": 2.6941649899396376,
+      "grad_norm": 0.2827167809009552,
+      "learning_rate": 3.132671031311868e-07,
+      "loss": 0.3212,
+      "step": 10712
+    },
+    {
+      "epoch": 2.694416498993964,
+      "grad_norm": 0.27102166414260864,
+      "learning_rate": 3.127575108168146e-07,
+      "loss": 0.3306,
+      "step": 10713
+    },
+    {
+      "epoch": 2.6946680080482897,
+      "grad_norm": 0.2781563997268677,
+      "learning_rate": 3.1224831993543867e-07,
+      "loss": 0.3274,
+      "step": 10714
+    },
+    {
+      "epoch": 2.6949195171026155,
+      "grad_norm": 0.3081703782081604,
+      "learning_rate": 3.11739530530667e-07,
+      "loss": 0.3186,
+      "step": 10715
+    },
+    {
+      "epoch": 2.6951710261569417,
+      "grad_norm": 0.31926673650741577,
+      "learning_rate": 3.1123114264607515e-07,
+      "loss": 0.3128,
+      "step": 10716
+    },
+    {
+      "epoch": 2.6954225352112675,
+      "grad_norm": 0.28970199823379517,
+      "learning_rate": 3.1072315632520277e-07,
+      "loss": 0.2909,
+      "step": 10717
+    },
+    {
+      "epoch": 2.6956740442655933,
+      "grad_norm": 0.28923845291137695,
+      "learning_rate": 3.102155716115568e-07,
+      "loss": 0.3278,
+      "step": 10718
+    },
+    {
+      "epoch": 2.6959255533199196,
+      "grad_norm": 0.3150719106197357,
+      "learning_rate": 3.097083885486074e-07,
+      "loss": 0.3314,
+      "step": 10719
+    },
+    {
+      "epoch": 2.6961770623742454,
+      "grad_norm": 0.2829613983631134,
+      "learning_rate": 3.0920160717979264e-07,
+      "loss": 0.3328,
+      "step": 10720
+    },
+    {
+      "epoch": 2.696428571428571,
+      "grad_norm": 0.2828682065010071,
+      "learning_rate": 3.0869522754851387e-07,
+      "loss": 0.3164,
+      "step": 10721
+    },
+    {
+      "epoch": 2.6966800804828974,
+      "grad_norm": 0.2709435522556305,
+      "learning_rate": 3.081892496981409e-07,
+      "loss": 0.3154,
+      "step": 10722
+    },
+    {
+      "epoch": 2.6969315895372232,
+      "grad_norm": 0.3398783504962921,
+      "learning_rate": 3.0768367367200624e-07,
+      "loss": 0.3052,
+      "step": 10723
+    },
+    {
+      "epoch": 2.697183098591549,
+      "grad_norm": 0.30523836612701416,
+      "learning_rate": 3.0717849951340973e-07,
+      "loss": 0.3241,
+      "step": 10724
+    },
+    {
+      "epoch": 2.6974346076458753,
+      "grad_norm": 0.2881937623023987,
+      "learning_rate": 3.066737272656173e-07,
+      "loss": 0.3551,
+      "step": 10725
+    },
+    {
+      "epoch": 2.697686116700201,
+      "grad_norm": 0.28372272849082947,
+      "learning_rate": 3.061693569718577e-07,
+      "loss": 0.3136,
+      "step": 10726
+    },
+    {
+      "epoch": 2.697937625754527,
+      "grad_norm": 0.2901321053504944,
+      "learning_rate": 3.056653886753297e-07,
+      "loss": 0.3264,
+      "step": 10727
+    },
+    {
+      "epoch": 2.698189134808853,
+      "grad_norm": 0.2705398499965668,
+      "learning_rate": 3.051618224191921e-07,
+      "loss": 0.3243,
+      "step": 10728
+    },
+    {
+      "epoch": 2.698440643863179,
+      "grad_norm": 0.2847079336643219,
+      "learning_rate": 3.046586582465744e-07,
+      "loss": 0.3083,
+      "step": 10729
+    },
+    {
+      "epoch": 2.698692152917505,
+      "grad_norm": 0.29345259070396423,
+      "learning_rate": 3.0415589620056807e-07,
+      "loss": 0.3141,
+      "step": 10730
+    },
+    {
+      "epoch": 2.698943661971831,
+      "grad_norm": 0.2997452914714813,
+      "learning_rate": 3.036535363242327e-07,
+      "loss": 0.2915,
+      "step": 10731
+    },
+    {
+      "epoch": 2.699195171026157,
+      "grad_norm": 0.2906875014305115,
+      "learning_rate": 3.0315157866059154e-07,
+      "loss": 0.3336,
+      "step": 10732
+    },
+    {
+      "epoch": 2.699446680080483,
+      "grad_norm": 0.2780359387397766,
+      "learning_rate": 3.0265002325263415e-07,
+      "loss": 0.3313,
+      "step": 10733
+    },
+    {
+      "epoch": 2.699698189134809,
+      "grad_norm": 0.2860005497932434,
+      "learning_rate": 3.021488701433156e-07,
+      "loss": 0.3464,
+      "step": 10734
+    },
+    {
+      "epoch": 2.6999496981891347,
+      "grad_norm": 0.30452266335487366,
+      "learning_rate": 3.016481193755566e-07,
+      "loss": 0.3243,
+      "step": 10735
+    },
+    {
+      "epoch": 2.700201207243461,
+      "grad_norm": 0.28733357787132263,
+      "learning_rate": 3.011477709922439e-07,
+      "loss": 0.3178,
+      "step": 10736
+    },
+    {
+      "epoch": 2.7004527162977867,
+      "grad_norm": 0.26170480251312256,
+      "learning_rate": 3.0064782503622827e-07,
+      "loss": 0.3241,
+      "step": 10737
+    },
+    {
+      "epoch": 2.7007042253521125,
+      "grad_norm": 0.27204710245132446,
+      "learning_rate": 3.001482815503276e-07,
+      "loss": 0.3139,
+      "step": 10738
+    },
+    {
+      "epoch": 2.7009557344064388,
+      "grad_norm": 0.3037412464618683,
+      "learning_rate": 2.9964914057732386e-07,
+      "loss": 0.3181,
+      "step": 10739
+    },
+    {
+      "epoch": 2.7012072434607646,
+      "grad_norm": 0.27614688873291016,
+      "learning_rate": 2.9915040215996726e-07,
+      "loss": 0.3135,
+      "step": 10740
+    },
+    {
+      "epoch": 2.7014587525150904,
+      "grad_norm": 0.2942797839641571,
+      "learning_rate": 2.986520663409687e-07,
+      "loss": 0.3279,
+      "step": 10741
+    },
+    {
+      "epoch": 2.7017102615694166,
+      "grad_norm": 0.2742573320865631,
+      "learning_rate": 2.9815413316301055e-07,
+      "loss": 0.3289,
+      "step": 10742
+    },
+    {
+      "epoch": 2.7019617706237424,
+      "grad_norm": 0.27266111969947815,
+      "learning_rate": 2.97656602668735e-07,
+      "loss": 0.3331,
+      "step": 10743
+    },
+    {
+      "epoch": 2.7022132796780687,
+      "grad_norm": 0.29972144961357117,
+      "learning_rate": 2.9715947490075447e-07,
+      "loss": 0.3313,
+      "step": 10744
+    },
+    {
+      "epoch": 2.7024647887323945,
+      "grad_norm": 0.30087292194366455,
+      "learning_rate": 2.9666274990164336e-07,
+      "loss": 0.3195,
+      "step": 10745
+    },
+    {
+      "epoch": 2.7027162977867203,
+      "grad_norm": 0.28101980686187744,
+      "learning_rate": 2.961664277139448e-07,
+      "loss": 0.3034,
+      "step": 10746
+    },
+    {
+      "epoch": 2.7029678068410465,
+      "grad_norm": 0.2926780879497528,
+      "learning_rate": 2.956705083801636e-07,
+      "loss": 0.3363,
+      "step": 10747
+    },
+    {
+      "epoch": 2.7032193158953723,
+      "grad_norm": 0.30072760581970215,
+      "learning_rate": 2.951749919427738e-07,
+      "loss": 0.3311,
+      "step": 10748
+    },
+    {
+      "epoch": 2.703470824949698,
+      "grad_norm": 0.3085022568702698,
+      "learning_rate": 2.9467987844421173e-07,
+      "loss": 0.3272,
+      "step": 10749
+    },
+    {
+      "epoch": 2.7037223340040244,
+      "grad_norm": 0.28719770908355713,
+      "learning_rate": 2.941851679268831e-07,
+      "loss": 0.3186,
+      "step": 10750
+    },
+    {
+      "epoch": 2.70397384305835,
+      "grad_norm": 0.30047592520713806,
+      "learning_rate": 2.93690860433154e-07,
+      "loss": 0.308,
+      "step": 10751
+    },
+    {
+      "epoch": 2.704225352112676,
+      "grad_norm": 0.29841721057891846,
+      "learning_rate": 2.93196956005361e-07,
+      "loss": 0.315,
+      "step": 10752
+    },
+    {
+      "epoch": 2.7044768611670023,
+      "grad_norm": 0.3019961416721344,
+      "learning_rate": 2.927034546858026e-07,
+      "loss": 0.3273,
+      "step": 10753
+    },
+    {
+      "epoch": 2.704728370221328,
+      "grad_norm": 0.2860671579837799,
+      "learning_rate": 2.92210356516745e-07,
+      "loss": 0.3405,
+      "step": 10754
+    },
+    {
+      "epoch": 2.704979879275654,
+      "grad_norm": 0.2890463173389435,
+      "learning_rate": 2.917176615404177e-07,
+      "loss": 0.3264,
+      "step": 10755
+    },
+    {
+      "epoch": 2.70523138832998,
+      "grad_norm": 0.27873101830482483,
+      "learning_rate": 2.9122536979901807e-07,
+      "loss": 0.3134,
+      "step": 10756
+    },
+    {
+      "epoch": 2.705482897384306,
+      "grad_norm": 0.28098541498184204,
+      "learning_rate": 2.9073348133470844e-07,
+      "loss": 0.3087,
+      "step": 10757
+    },
+    {
+      "epoch": 2.7057344064386317,
+      "grad_norm": 0.27469319105148315,
+      "learning_rate": 2.902419961896136e-07,
+      "loss": 0.3253,
+      "step": 10758
+    },
+    {
+      "epoch": 2.705985915492958,
+      "grad_norm": 0.295611172914505,
+      "learning_rate": 2.8975091440582915e-07,
+      "loss": 0.3224,
+      "step": 10759
+    },
+    {
+      "epoch": 2.7062374245472838,
+      "grad_norm": 0.2876059114933014,
+      "learning_rate": 2.892602360254104e-07,
+      "loss": 0.3465,
+      "step": 10760
+    },
+    {
+      "epoch": 2.7064889336016096,
+      "grad_norm": 0.285182923078537,
+      "learning_rate": 2.887699610903827e-07,
+      "loss": 0.3092,
+      "step": 10761
+    },
+    {
+      "epoch": 2.706740442655936,
+      "grad_norm": 0.2798616290092468,
+      "learning_rate": 2.8828008964273404e-07,
+      "loss": 0.3096,
+      "step": 10762
+    },
+    {
+      "epoch": 2.7069919517102616,
+      "grad_norm": 0.2900584042072296,
+      "learning_rate": 2.877906217244203e-07,
+      "loss": 0.3103,
+      "step": 10763
+    },
+    {
+      "epoch": 2.7072434607645874,
+      "grad_norm": 0.2769961357116699,
+      "learning_rate": 2.873015573773591e-07,
+      "loss": 0.3217,
+      "step": 10764
+    },
+    {
+      "epoch": 2.7074949698189137,
+      "grad_norm": 0.2770218849182129,
+      "learning_rate": 2.86812896643438e-07,
+      "loss": 0.3292,
+      "step": 10765
+    },
+    {
+      "epoch": 2.7077464788732395,
+      "grad_norm": 0.29273056983947754,
+      "learning_rate": 2.8632463956450583e-07,
+      "loss": 0.3286,
+      "step": 10766
+    },
+    {
+      "epoch": 2.7079979879275653,
+      "grad_norm": 0.2939911186695099,
+      "learning_rate": 2.858367861823802e-07,
+      "loss": 0.3022,
+      "step": 10767
+    },
+    {
+      "epoch": 2.7082494969818915,
+      "grad_norm": 0.3015715777873993,
+      "learning_rate": 2.8534933653884157e-07,
+      "loss": 0.3051,
+      "step": 10768
+    },
+    {
+      "epoch": 2.7085010060362174,
+      "grad_norm": 0.28284311294555664,
+      "learning_rate": 2.848622906756382e-07,
+      "loss": 0.3136,
+      "step": 10769
+    },
+    {
+      "epoch": 2.708752515090543,
+      "grad_norm": 0.2819089889526367,
+      "learning_rate": 2.843756486344812e-07,
+      "loss": 0.3146,
+      "step": 10770
+    },
+    {
+      "epoch": 2.7090040241448694,
+      "grad_norm": 0.3178308606147766,
+      "learning_rate": 2.8388941045705e-07,
+      "loss": 0.3401,
+      "step": 10771
+    },
+    {
+      "epoch": 2.709255533199195,
+      "grad_norm": 0.2823493182659149,
+      "learning_rate": 2.834035761849857e-07,
+      "loss": 0.3334,
+      "step": 10772
+    },
+    {
+      "epoch": 2.709507042253521,
+      "grad_norm": 0.2783769369125366,
+      "learning_rate": 2.8291814585989894e-07,
+      "loss": 0.3232,
+      "step": 10773
+    },
+    {
+      "epoch": 2.7097585513078473,
+      "grad_norm": 0.2950974404811859,
+      "learning_rate": 2.824331195233626e-07,
+      "loss": 0.3237,
+      "step": 10774
+    },
+    {
+      "epoch": 2.710010060362173,
+      "grad_norm": 0.2713472247123718,
+      "learning_rate": 2.8194849721691673e-07,
+      "loss": 0.2991,
+      "step": 10775
+    },
+    {
+      "epoch": 2.710261569416499,
+      "grad_norm": 0.2863728702068329,
+      "learning_rate": 2.81464278982066e-07,
+      "loss": 0.3167,
+      "step": 10776
+    },
+    {
+      "epoch": 2.710513078470825,
+      "grad_norm": 0.285849392414093,
+      "learning_rate": 2.80980464860281e-07,
+      "loss": 0.3086,
+      "step": 10777
+    },
+    {
+      "epoch": 2.710764587525151,
+      "grad_norm": 0.31550291180610657,
+      "learning_rate": 2.804970548929964e-07,
+      "loss": 0.3127,
+      "step": 10778
+    },
+    {
+      "epoch": 2.7110160965794767,
+      "grad_norm": 0.29624804854393005,
+      "learning_rate": 2.8001404912161413e-07,
+      "loss": 0.3238,
+      "step": 10779
+    },
+    {
+      "epoch": 2.711267605633803,
+      "grad_norm": 0.2810332179069519,
+      "learning_rate": 2.7953144758750107e-07,
+      "loss": 0.3298,
+      "step": 10780
+    },
+    {
+      "epoch": 2.711519114688129,
+      "grad_norm": 0.29689785838127136,
+      "learning_rate": 2.790492503319875e-07,
+      "loss": 0.3364,
+      "step": 10781
+    },
+    {
+      "epoch": 2.7117706237424546,
+      "grad_norm": 0.30866745114326477,
+      "learning_rate": 2.7856745739637206e-07,
+      "loss": 0.3216,
+      "step": 10782
+    },
+    {
+      "epoch": 2.712022132796781,
+      "grad_norm": 0.26995787024497986,
+      "learning_rate": 2.7808606882191615e-07,
+      "loss": 0.328,
+      "step": 10783
+    },
+    {
+      "epoch": 2.7122736418511066,
+      "grad_norm": 0.2834140956401825,
+      "learning_rate": 2.7760508464984904e-07,
+      "loss": 0.3015,
+      "step": 10784
+    },
+    {
+      "epoch": 2.7125251509054324,
+      "grad_norm": 0.28610095381736755,
+      "learning_rate": 2.771245049213628e-07,
+      "loss": 0.3174,
+      "step": 10785
+    },
+    {
+      "epoch": 2.7127766599597587,
+      "grad_norm": 0.2732725441455841,
+      "learning_rate": 2.7664432967761667e-07,
+      "loss": 0.3077,
+      "step": 10786
+    },
+    {
+      "epoch": 2.7130281690140845,
+      "grad_norm": 0.2708452641963959,
+      "learning_rate": 2.7616455895973393e-07,
+      "loss": 0.3036,
+      "step": 10787
+    },
+    {
+      "epoch": 2.7132796780684103,
+      "grad_norm": 0.28488287329673767,
+      "learning_rate": 2.756851928088056e-07,
+      "loss": 0.3226,
+      "step": 10788
+    },
+    {
+      "epoch": 2.7135311871227366,
+      "grad_norm": 0.3026772737503052,
+      "learning_rate": 2.752062312658838e-07,
+      "loss": 0.3404,
+      "step": 10789
+    },
+    {
+      "epoch": 2.7137826961770624,
+      "grad_norm": 0.28403767943382263,
+      "learning_rate": 2.7472767437199067e-07,
+      "loss": 0.325,
+      "step": 10790
+    },
+    {
+      "epoch": 2.714034205231388,
+      "grad_norm": 0.2924075424671173,
+      "learning_rate": 2.742495221681113e-07,
+      "loss": 0.322,
+      "step": 10791
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.28228023648262024,
+      "learning_rate": 2.7377177469519565e-07,
+      "loss": 0.3184,
+      "step": 10792
+    },
+    {
+      "epoch": 2.71453722334004,
+      "grad_norm": 0.29336363077163696,
+      "learning_rate": 2.7329443199416105e-07,
+      "loss": 0.3155,
+      "step": 10793
+    },
+    {
+      "epoch": 2.714788732394366,
+      "grad_norm": 0.31586167216300964,
+      "learning_rate": 2.7281749410588753e-07,
+      "loss": 0.3242,
+      "step": 10794
+    },
+    {
+      "epoch": 2.7150402414486923,
+      "grad_norm": 0.30801641941070557,
+      "learning_rate": 2.7234096107122357e-07,
+      "loss": 0.3315,
+      "step": 10795
+    },
+    {
+      "epoch": 2.715291750503018,
+      "grad_norm": 0.29662755131721497,
+      "learning_rate": 2.7186483293097863e-07,
+      "loss": 0.3071,
+      "step": 10796
+    },
+    {
+      "epoch": 2.715543259557344,
+      "grad_norm": 0.28053396940231323,
+      "learning_rate": 2.71389109725933e-07,
+      "loss": 0.3257,
+      "step": 10797
+    },
+    {
+      "epoch": 2.71579476861167,
+      "grad_norm": 0.29082953929901123,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.2968,
+      "step": 10798
+    },
+    {
+      "epoch": 2.716046277665996,
+      "grad_norm": 0.30169904232025146,
+      "learning_rate": 2.7043887828437033e-07,
+      "loss": 0.3146,
+      "step": 10799
+    },
+    {
+      "epoch": 2.7162977867203217,
+      "grad_norm": 0.3031311333179474,
+      "learning_rate": 2.699643701292348e-07,
+      "loss": 0.3383,
+      "step": 10800
+    },
+    {
+      "epoch": 2.716549295774648,
+      "grad_norm": 0.28714466094970703,
+      "learning_rate": 2.694902670720606e-07,
+      "loss": 0.3333,
+      "step": 10801
+    },
+    {
+      "epoch": 2.716800804828974,
+      "grad_norm": 0.2786526679992676,
+      "learning_rate": 2.690165691534513e-07,
+      "loss": 0.3273,
+      "step": 10802
+    },
+    {
+      "epoch": 2.7170523138832996,
+      "grad_norm": 0.2807994484901428,
+      "learning_rate": 2.6854327641397504e-07,
+      "loss": 0.2979,
+      "step": 10803
+    },
+    {
+      "epoch": 2.717303822937626,
+      "grad_norm": 0.3144852817058563,
+      "learning_rate": 2.6807038889416824e-07,
+      "loss": 0.3153,
+      "step": 10804
+    },
+    {
+      "epoch": 2.7175553319919517,
+      "grad_norm": 0.2780417501926422,
+      "learning_rate": 2.675979066345291e-07,
+      "loss": 0.3198,
+      "step": 10805
+    },
+    {
+      "epoch": 2.7178068410462775,
+      "grad_norm": 0.27267181873321533,
+      "learning_rate": 2.671258296755241e-07,
+      "loss": 0.2898,
+      "step": 10806
+    },
+    {
+      "epoch": 2.7180583501006037,
+      "grad_norm": 0.29780054092407227,
+      "learning_rate": 2.6665415805758264e-07,
+      "loss": 0.3345,
+      "step": 10807
+    },
+    {
+      "epoch": 2.7183098591549295,
+      "grad_norm": 0.286492258310318,
+      "learning_rate": 2.661828918211012e-07,
+      "loss": 0.3259,
+      "step": 10808
+    },
+    {
+      "epoch": 2.7185613682092553,
+      "grad_norm": 0.2818504571914673,
+      "learning_rate": 2.657120310064393e-07,
+      "loss": 0.32,
+      "step": 10809
+    },
+    {
+      "epoch": 2.7188128772635816,
+      "grad_norm": 0.28963735699653625,
+      "learning_rate": 2.6524157565392506e-07,
+      "loss": 0.3339,
+      "step": 10810
+    },
+    {
+      "epoch": 2.7190643863179074,
+      "grad_norm": 0.268822580575943,
+      "learning_rate": 2.647715258038497e-07,
+      "loss": 0.3148,
+      "step": 10811
+    },
+    {
+      "epoch": 2.719315895372233,
+      "grad_norm": 0.304812490940094,
+      "learning_rate": 2.643018814964687e-07,
+      "loss": 0.3164,
+      "step": 10812
+    },
+    {
+      "epoch": 2.7195674044265594,
+      "grad_norm": 0.28156810998916626,
+      "learning_rate": 2.6383264277200616e-07,
+      "loss": 0.3312,
+      "step": 10813
+    },
+    {
+      "epoch": 2.7198189134808852,
+      "grad_norm": 0.2927444875240326,
+      "learning_rate": 2.6336380967064754e-07,
+      "loss": 0.3059,
+      "step": 10814
+    },
+    {
+      "epoch": 2.720070422535211,
+      "grad_norm": 0.2759106159210205,
+      "learning_rate": 2.6289538223254695e-07,
+      "loss": 0.3408,
+      "step": 10815
+    },
+    {
+      "epoch": 2.7203219315895373,
+      "grad_norm": 0.2832070589065552,
+      "learning_rate": 2.624273604978211e-07,
+      "loss": 0.3276,
+      "step": 10816
+    },
+    {
+      "epoch": 2.720573440643863,
+      "grad_norm": 0.27949801087379456,
+      "learning_rate": 2.6195974450655415e-07,
+      "loss": 0.3654,
+      "step": 10817
+    },
+    {
+      "epoch": 2.720824949698189,
+      "grad_norm": 0.29671213030815125,
+      "learning_rate": 2.6149253429879397e-07,
+      "loss": 0.3069,
+      "step": 10818
+    },
+    {
+      "epoch": 2.721076458752515,
+      "grad_norm": 0.31909680366516113,
+      "learning_rate": 2.610257299145541e-07,
+      "loss": 0.3114,
+      "step": 10819
+    },
+    {
+      "epoch": 2.721327967806841,
+      "grad_norm": 0.3081822395324707,
+      "learning_rate": 2.6055933139381315e-07,
+      "loss": 0.3447,
+      "step": 10820
+    },
+    {
+      "epoch": 2.7215794768611667,
+      "grad_norm": 0.30881619453430176,
+      "learning_rate": 2.600933387765159e-07,
+      "loss": 0.2926,
+      "step": 10821
+    },
+    {
+      "epoch": 2.721830985915493,
+      "grad_norm": 0.28827062249183655,
+      "learning_rate": 2.596277521025714e-07,
+      "loss": 0.326,
+      "step": 10822
+    },
+    {
+      "epoch": 2.722082494969819,
+      "grad_norm": 0.30842483043670654,
+      "learning_rate": 2.5916257141185395e-07,
+      "loss": 0.3216,
+      "step": 10823
+    },
+    {
+      "epoch": 2.7223340040241446,
+      "grad_norm": 0.3006271421909332,
+      "learning_rate": 2.586977967442045e-07,
+      "loss": 0.3144,
+      "step": 10824
+    },
+    {
+      "epoch": 2.722585513078471,
+      "grad_norm": 0.2951270341873169,
+      "learning_rate": 2.5823342813942665e-07,
+      "loss": 0.329,
+      "step": 10825
+    },
+    {
+      "epoch": 2.7228370221327967,
+      "grad_norm": 0.2893744111061096,
+      "learning_rate": 2.57769465637292e-07,
+      "loss": 0.3137,
+      "step": 10826
+    },
+    {
+      "epoch": 2.7230885311871225,
+      "grad_norm": 0.2886696457862854,
+      "learning_rate": 2.573059092775343e-07,
+      "loss": 0.3488,
+      "step": 10827
+    },
+    {
+      "epoch": 2.7233400402414487,
+      "grad_norm": 0.3092022240161896,
+      "learning_rate": 2.568427590998557e-07,
+      "loss": 0.3293,
+      "step": 10828
+    },
+    {
+      "epoch": 2.7235915492957745,
+      "grad_norm": 0.2914911210536957,
+      "learning_rate": 2.563800151439216e-07,
+      "loss": 0.3183,
+      "step": 10829
+    },
+    {
+      "epoch": 2.7238430583501008,
+      "grad_norm": 0.29686030745506287,
+      "learning_rate": 2.559176774493638e-07,
+      "loss": 0.3196,
+      "step": 10830
+    },
+    {
+      "epoch": 2.7240945674044266,
+      "grad_norm": 0.303915411233902,
+      "learning_rate": 2.554557460557772e-07,
+      "loss": 0.3234,
+      "step": 10831
+    },
+    {
+      "epoch": 2.7243460764587524,
+      "grad_norm": 0.3081934452056885,
+      "learning_rate": 2.54994221002724e-07,
+      "loss": 0.3263,
+      "step": 10832
+    },
+    {
+      "epoch": 2.7245975855130786,
+      "grad_norm": 0.290477454662323,
+      "learning_rate": 2.5453310232973205e-07,
+      "loss": 0.3441,
+      "step": 10833
+    },
+    {
+      "epoch": 2.7248490945674044,
+      "grad_norm": 0.2770725190639496,
+      "learning_rate": 2.5407239007629145e-07,
+      "loss": 0.3166,
+      "step": 10834
+    },
+    {
+      "epoch": 2.7251006036217302,
+      "grad_norm": 0.28399229049682617,
+      "learning_rate": 2.536120842818612e-07,
+      "loss": 0.2891,
+      "step": 10835
+    },
+    {
+      "epoch": 2.7253521126760565,
+      "grad_norm": 0.2806536853313446,
+      "learning_rate": 2.53152184985862e-07,
+      "loss": 0.3178,
+      "step": 10836
+    },
+    {
+      "epoch": 2.7256036217303823,
+      "grad_norm": 0.27946868538856506,
+      "learning_rate": 2.5269269222768234e-07,
+      "loss": 0.3259,
+      "step": 10837
+    },
+    {
+      "epoch": 2.725855130784708,
+      "grad_norm": 0.2891778349876404,
+      "learning_rate": 2.5223360604667404e-07,
+      "loss": 0.3271,
+      "step": 10838
+    },
+    {
+      "epoch": 2.7261066398390343,
+      "grad_norm": 0.2895258069038391,
+      "learning_rate": 2.517749264821556e-07,
+      "loss": 0.3193,
+      "step": 10839
+    },
+    {
+      "epoch": 2.72635814889336,
+      "grad_norm": 0.2947310507297516,
+      "learning_rate": 2.5131665357340963e-07,
+      "loss": 0.3362,
+      "step": 10840
+    },
+    {
+      "epoch": 2.7266096579476864,
+      "grad_norm": 0.28745022416114807,
+      "learning_rate": 2.508587873596857e-07,
+      "loss": 0.2958,
+      "step": 10841
+    },
+    {
+      "epoch": 2.726861167002012,
+      "grad_norm": 0.2851249873638153,
+      "learning_rate": 2.504013278801948e-07,
+      "loss": 0.327,
+      "step": 10842
+    },
+    {
+      "epoch": 2.727112676056338,
+      "grad_norm": 0.28858688473701477,
+      "learning_rate": 2.499442751741171e-07,
+      "loss": 0.3361,
+      "step": 10843
+    },
+    {
+      "epoch": 2.7273641851106643,
+      "grad_norm": 0.29522156715393066,
+      "learning_rate": 2.4948762928059647e-07,
+      "loss": 0.3238,
+      "step": 10844
+    },
+    {
+      "epoch": 2.72761569416499,
+      "grad_norm": 0.2840255796909332,
+      "learning_rate": 2.490313902387409e-07,
+      "loss": 0.3464,
+      "step": 10845
+    },
+    {
+      "epoch": 2.727867203219316,
+      "grad_norm": 0.29043489694595337,
+      "learning_rate": 2.485755580876248e-07,
+      "loss": 0.3275,
+      "step": 10846
+    },
+    {
+      "epoch": 2.728118712273642,
+      "grad_norm": 0.29880261421203613,
+      "learning_rate": 2.4812013286628747e-07,
+      "loss": 0.3568,
+      "step": 10847
+    },
+    {
+      "epoch": 2.728370221327968,
+      "grad_norm": 0.31489261984825134,
+      "learning_rate": 2.4766511461373324e-07,
+      "loss": 0.3404,
+      "step": 10848
+    },
+    {
+      "epoch": 2.7286217303822937,
+      "grad_norm": 0.28683874011039734,
+      "learning_rate": 2.4721050336893094e-07,
+      "loss": 0.3227,
+      "step": 10849
+    },
+    {
+      "epoch": 2.72887323943662,
+      "grad_norm": 0.27683448791503906,
+      "learning_rate": 2.467562991708161e-07,
+      "loss": 0.3275,
+      "step": 10850
+    },
+    {
+      "epoch": 2.7291247484909458,
+      "grad_norm": 0.28187745809555054,
+      "learning_rate": 2.4630250205828767e-07,
+      "loss": 0.331,
+      "step": 10851
+    },
+    {
+      "epoch": 2.7293762575452716,
+      "grad_norm": 0.30056044459342957,
+      "learning_rate": 2.458491120702117e-07,
+      "loss": 0.2948,
+      "step": 10852
+    },
+    {
+      "epoch": 2.729627766599598,
+      "grad_norm": 0.287810742855072,
+      "learning_rate": 2.453961292454166e-07,
+      "loss": 0.327,
+      "step": 10853
+    },
+    {
+      "epoch": 2.7298792756539236,
+      "grad_norm": 0.2935195565223694,
+      "learning_rate": 2.4494355362269796e-07,
+      "loss": 0.3268,
+      "step": 10854
+    },
+    {
+      "epoch": 2.7301307847082494,
+      "grad_norm": 0.2871311902999878,
+      "learning_rate": 2.444913852408176e-07,
+      "loss": 0.3373,
+      "step": 10855
+    },
+    {
+      "epoch": 2.7303822937625757,
+      "grad_norm": 0.2980871796607971,
+      "learning_rate": 2.440396241384985e-07,
+      "loss": 0.3115,
+      "step": 10856
+    },
+    {
+      "epoch": 2.7306338028169015,
+      "grad_norm": 0.29173511266708374,
+      "learning_rate": 2.435882703544334e-07,
+      "loss": 0.3228,
+      "step": 10857
+    },
+    {
+      "epoch": 2.7308853118712273,
+      "grad_norm": 0.3014184534549713,
+      "learning_rate": 2.4313732392727664e-07,
+      "loss": 0.3397,
+      "step": 10858
+    },
+    {
+      "epoch": 2.7311368209255535,
+      "grad_norm": 0.29081061482429504,
+      "learning_rate": 2.4268678489564935e-07,
+      "loss": 0.3183,
+      "step": 10859
+    },
+    {
+      "epoch": 2.7313883299798793,
+      "grad_norm": 0.3027068078517914,
+      "learning_rate": 2.422366532981368e-07,
+      "loss": 0.3241,
+      "step": 10860
+    },
+    {
+      "epoch": 2.731639839034205,
+      "grad_norm": 0.29450279474258423,
+      "learning_rate": 2.4178692917329106e-07,
+      "loss": 0.3406,
+      "step": 10861
+    },
+    {
+      "epoch": 2.7318913480885314,
+      "grad_norm": 0.26403874158859253,
+      "learning_rate": 2.413376125596267e-07,
+      "loss": 0.2853,
+      "step": 10862
+    },
+    {
+      "epoch": 2.732142857142857,
+      "grad_norm": 0.3012741804122925,
+      "learning_rate": 2.4088870349562644e-07,
+      "loss": 0.3511,
+      "step": 10863
+    },
+    {
+      "epoch": 2.732394366197183,
+      "grad_norm": 0.2905932068824768,
+      "learning_rate": 2.404402020197355e-07,
+      "loss": 0.303,
+      "step": 10864
+    },
+    {
+      "epoch": 2.7326458752515093,
+      "grad_norm": 0.2958056926727295,
+      "learning_rate": 2.399921081703654e-07,
+      "loss": 0.3264,
+      "step": 10865
+    },
+    {
+      "epoch": 2.732897384305835,
+      "grad_norm": 0.2873068153858185,
+      "learning_rate": 2.3954442198589334e-07,
+      "loss": 0.2966,
+      "step": 10866
+    },
+    {
+      "epoch": 2.733148893360161,
+      "grad_norm": 0.3035353720188141,
+      "learning_rate": 2.390971435046596e-07,
+      "loss": 0.3086,
+      "step": 10867
+    },
+    {
+      "epoch": 2.733400402414487,
+      "grad_norm": 0.3155568838119507,
+      "learning_rate": 2.38650272764972e-07,
+      "loss": 0.3157,
+      "step": 10868
+    },
+    {
+      "epoch": 2.733651911468813,
+      "grad_norm": 0.2832005023956299,
+      "learning_rate": 2.3820380980510093e-07,
+      "loss": 0.3133,
+      "step": 10869
+    },
+    {
+      "epoch": 2.7339034205231387,
+      "grad_norm": 0.28932392597198486,
+      "learning_rate": 2.3775775466328422e-07,
+      "loss": 0.3163,
+      "step": 10870
+    },
+    {
+      "epoch": 2.734154929577465,
+      "grad_norm": 0.30421337485313416,
+      "learning_rate": 2.3731210737772293e-07,
+      "loss": 0.3355,
+      "step": 10871
+    },
+    {
+      "epoch": 2.734406438631791,
+      "grad_norm": 0.2934373915195465,
+      "learning_rate": 2.3686686798658543e-07,
+      "loss": 0.3051,
+      "step": 10872
+    },
+    {
+      "epoch": 2.7346579476861166,
+      "grad_norm": 0.28928595781326294,
+      "learning_rate": 2.364220365280012e-07,
+      "loss": 0.3199,
+      "step": 10873
+    },
+    {
+      "epoch": 2.734909456740443,
+      "grad_norm": 0.30236467719078064,
+      "learning_rate": 2.3597761304006984e-07,
+      "loss": 0.3511,
+      "step": 10874
+    },
+    {
+      "epoch": 2.7351609657947686,
+      "grad_norm": 0.29965662956237793,
+      "learning_rate": 2.3553359756085192e-07,
+      "loss": 0.2945,
+      "step": 10875
+    },
+    {
+      "epoch": 2.7354124748490944,
+      "grad_norm": 0.31335750222206116,
+      "learning_rate": 2.3508999012837484e-07,
+      "loss": 0.3408,
+      "step": 10876
+    },
+    {
+      "epoch": 2.7356639839034207,
+      "grad_norm": 0.28357815742492676,
+      "learning_rate": 2.3464679078063102e-07,
+      "loss": 0.3215,
+      "step": 10877
+    },
+    {
+      "epoch": 2.7359154929577465,
+      "grad_norm": 0.30339914560317993,
+      "learning_rate": 2.3420399955557782e-07,
+      "loss": 0.3115,
+      "step": 10878
+    },
+    {
+      "epoch": 2.7361670020120723,
+      "grad_norm": 0.31013423204421997,
+      "learning_rate": 2.337616164911366e-07,
+      "loss": 0.3253,
+      "step": 10879
+    },
+    {
+      "epoch": 2.7364185110663986,
+      "grad_norm": 0.30664506554603577,
+      "learning_rate": 2.333196416251965e-07,
+      "loss": 0.306,
+      "step": 10880
+    },
+    {
+      "epoch": 2.7366700201207244,
+      "grad_norm": 0.27958711981773376,
+      "learning_rate": 2.3287807499560777e-07,
+      "loss": 0.3085,
+      "step": 10881
+    },
+    {
+      "epoch": 2.73692152917505,
+      "grad_norm": 0.27122312784194946,
+      "learning_rate": 2.3243691664018964e-07,
+      "loss": 0.3313,
+      "step": 10882
+    },
+    {
+      "epoch": 2.7371730382293764,
+      "grad_norm": 0.2907143533229828,
+      "learning_rate": 2.3199616659672352e-07,
+      "loss": 0.3293,
+      "step": 10883
+    },
+    {
+      "epoch": 2.737424547283702,
+      "grad_norm": 0.29304689168930054,
+      "learning_rate": 2.315558249029576e-07,
+      "loss": 0.3059,
+      "step": 10884
+    },
+    {
+      "epoch": 2.737676056338028,
+      "grad_norm": 0.2985115945339203,
+      "learning_rate": 2.311158915966033e-07,
+      "loss": 0.3035,
+      "step": 10885
+    },
+    {
+      "epoch": 2.7379275653923543,
+      "grad_norm": 0.3043895661830902,
+      "learning_rate": 2.3067636671533944e-07,
+      "loss": 0.3192,
+      "step": 10886
+    },
+    {
+      "epoch": 2.73817907444668,
+      "grad_norm": 0.2754703462123871,
+      "learning_rate": 2.30237250296807e-07,
+      "loss": 0.3498,
+      "step": 10887
+    },
+    {
+      "epoch": 2.738430583501006,
+      "grad_norm": 0.27213260531425476,
+      "learning_rate": 2.2979854237861588e-07,
+      "loss": 0.3159,
+      "step": 10888
+    },
+    {
+      "epoch": 2.738682092555332,
+      "grad_norm": 0.29739394783973694,
+      "learning_rate": 2.2936024299833605e-07,
+      "loss": 0.3277,
+      "step": 10889
+    },
+    {
+      "epoch": 2.738933601609658,
+      "grad_norm": 0.2812947928905487,
+      "learning_rate": 2.2892235219350745e-07,
+      "loss": 0.3192,
+      "step": 10890
+    },
+    {
+      "epoch": 2.7391851106639837,
+      "grad_norm": 0.29224130511283875,
+      "learning_rate": 2.2848487000163067e-07,
+      "loss": 0.3248,
+      "step": 10891
+    },
+    {
+      "epoch": 2.73943661971831,
+      "grad_norm": 0.2976722717285156,
+      "learning_rate": 2.2804779646017517e-07,
+      "loss": 0.3271,
+      "step": 10892
+    },
+    {
+      "epoch": 2.739688128772636,
+      "grad_norm": 0.28829827904701233,
+      "learning_rate": 2.276111316065721e-07,
+      "loss": 0.3171,
+      "step": 10893
+    },
+    {
+      "epoch": 2.7399396378269616,
+      "grad_norm": 0.29329490661621094,
+      "learning_rate": 2.2717487547821992e-07,
+      "loss": 0.3202,
+      "step": 10894
+    },
+    {
+      "epoch": 2.740191146881288,
+      "grad_norm": 0.29280152916908264,
+      "learning_rate": 2.267390281124804e-07,
+      "loss": 0.305,
+      "step": 10895
+    },
+    {
+      "epoch": 2.7404426559356136,
+      "grad_norm": 0.2917034924030304,
+      "learning_rate": 2.2630358954668253e-07,
+      "loss": 0.3526,
+      "step": 10896
+    },
+    {
+      "epoch": 2.7406941649899395,
+      "grad_norm": 0.2969276010990143,
+      "learning_rate": 2.258685598181176e-07,
+      "loss": 0.333,
+      "step": 10897
+    },
+    {
+      "epoch": 2.7409456740442657,
+      "grad_norm": 0.2692883610725403,
+      "learning_rate": 2.2543393896404308e-07,
+      "loss": 0.329,
+      "step": 10898
+    },
+    {
+      "epoch": 2.7411971830985915,
+      "grad_norm": 0.29979148507118225,
+      "learning_rate": 2.2499972702168304e-07,
+      "loss": 0.3182,
+      "step": 10899
+    },
+    {
+      "epoch": 2.7414486921529173,
+      "grad_norm": 0.2951495349407196,
+      "learning_rate": 2.245659240282233e-07,
+      "loss": 0.3015,
+      "step": 10900
+    },
+    {
+      "epoch": 2.7417002012072436,
+      "grad_norm": 0.2918504774570465,
+      "learning_rate": 2.2413253002081803e-07,
+      "loss": 0.3195,
+      "step": 10901
+    },
+    {
+      "epoch": 2.7419517102615694,
+      "grad_norm": 0.28916507959365845,
+      "learning_rate": 2.2369954503658308e-07,
+      "loss": 0.3268,
+      "step": 10902
+    },
+    {
+      "epoch": 2.742203219315895,
+      "grad_norm": 0.29391786456108093,
+      "learning_rate": 2.2326696911260215e-07,
+      "loss": 0.3068,
+      "step": 10903
+    },
+    {
+      "epoch": 2.7424547283702214,
+      "grad_norm": 0.28322654962539673,
+      "learning_rate": 2.2283480228592168e-07,
+      "loss": 0.3066,
+      "step": 10904
+    },
+    {
+      "epoch": 2.7427062374245472,
+      "grad_norm": 0.3049914538860321,
+      "learning_rate": 2.2240304459355544e-07,
+      "loss": 0.3235,
+      "step": 10905
+    },
+    {
+      "epoch": 2.742957746478873,
+      "grad_norm": 0.26875460147857666,
+      "learning_rate": 2.2197169607247882e-07,
+      "loss": 0.3386,
+      "step": 10906
+    },
+    {
+      "epoch": 2.7432092555331993,
+      "grad_norm": 0.29541999101638794,
+      "learning_rate": 2.2154075675963617e-07,
+      "loss": 0.3232,
+      "step": 10907
+    },
+    {
+      "epoch": 2.743460764587525,
+      "grad_norm": 0.26899954676628113,
+      "learning_rate": 2.2111022669193293e-07,
+      "loss": 0.3177,
+      "step": 10908
+    },
+    {
+      "epoch": 2.743712273641851,
+      "grad_norm": 0.27981895208358765,
+      "learning_rate": 2.206801059062419e-07,
+      "loss": 0.3132,
+      "step": 10909
+    },
+    {
+      "epoch": 2.743963782696177,
+      "grad_norm": 0.28188595175743103,
+      "learning_rate": 2.2025039443940134e-07,
+      "loss": 0.3502,
+      "step": 10910
+    },
+    {
+      "epoch": 2.744215291750503,
+      "grad_norm": 0.27629154920578003,
+      "learning_rate": 2.198210923282118e-07,
+      "loss": 0.324,
+      "step": 10911
+    },
+    {
+      "epoch": 2.7444668008048287,
+      "grad_norm": 0.2914685904979706,
+      "learning_rate": 2.1939219960944168e-07,
+      "loss": 0.3242,
+      "step": 10912
+    },
+    {
+      "epoch": 2.744718309859155,
+      "grad_norm": 0.2767096757888794,
+      "learning_rate": 2.1896371631982162e-07,
+      "loss": 0.3063,
+      "step": 10913
+    },
+    {
+      "epoch": 2.744969818913481,
+      "grad_norm": 0.28286007046699524,
+      "learning_rate": 2.1853564249604996e-07,
+      "loss": 0.3,
+      "step": 10914
+    },
+    {
+      "epoch": 2.7452213279678066,
+      "grad_norm": 0.28479135036468506,
+      "learning_rate": 2.181079781747869e-07,
+      "loss": 0.3035,
+      "step": 10915
+    },
+    {
+      "epoch": 2.745472837022133,
+      "grad_norm": 0.27827152609825134,
+      "learning_rate": 2.176807233926609e-07,
+      "loss": 0.3327,
+      "step": 10916
+    },
+    {
+      "epoch": 2.7457243460764587,
+      "grad_norm": 0.29502740502357483,
+      "learning_rate": 2.172538781862621e-07,
+      "loss": 0.3055,
+      "step": 10917
+    },
+    {
+      "epoch": 2.7459758551307845,
+      "grad_norm": 0.2725423276424408,
+      "learning_rate": 2.168274425921485e-07,
+      "loss": 0.3063,
+      "step": 10918
+    },
+    {
+      "epoch": 2.7462273641851107,
+      "grad_norm": 0.29838913679122925,
+      "learning_rate": 2.1640141664684034e-07,
+      "loss": 0.2981,
+      "step": 10919
+    },
+    {
+      "epoch": 2.7464788732394365,
+      "grad_norm": 0.29055625200271606,
+      "learning_rate": 2.1597580038682453e-07,
+      "loss": 0.32,
+      "step": 10920
+    },
+    {
+      "epoch": 2.7467303822937623,
+      "grad_norm": 0.27832192182540894,
+      "learning_rate": 2.1555059384855358e-07,
+      "loss": 0.3281,
+      "step": 10921
+    },
+    {
+      "epoch": 2.7469818913480886,
+      "grad_norm": 0.27420729398727417,
+      "learning_rate": 2.1512579706844227e-07,
+      "loss": 0.3345,
+      "step": 10922
+    },
+    {
+      "epoch": 2.7472334004024144,
+      "grad_norm": 0.3057540953159332,
+      "learning_rate": 2.1470141008287316e-07,
+      "loss": 0.3079,
+      "step": 10923
+    },
+    {
+      "epoch": 2.74748490945674,
+      "grad_norm": 0.27829062938690186,
+      "learning_rate": 2.1427743292819047e-07,
+      "loss": 0.3347,
+      "step": 10924
+    },
+    {
+      "epoch": 2.7477364185110664,
+      "grad_norm": 0.3032315969467163,
+      "learning_rate": 2.1385386564070688e-07,
+      "loss": 0.3091,
+      "step": 10925
+    },
+    {
+      "epoch": 2.7479879275653922,
+      "grad_norm": 0.30215707421302795,
+      "learning_rate": 2.1343070825669776e-07,
+      "loss": 0.3233,
+      "step": 10926
+    },
+    {
+      "epoch": 2.748239436619718,
+      "grad_norm": 0.3079783320426941,
+      "learning_rate": 2.130079608124036e-07,
+      "loss": 0.3209,
+      "step": 10927
+    },
+    {
+      "epoch": 2.7484909456740443,
+      "grad_norm": 0.3027101457118988,
+      "learning_rate": 2.1258562334402987e-07,
+      "loss": 0.3379,
+      "step": 10928
+    },
+    {
+      "epoch": 2.74874245472837,
+      "grad_norm": 0.3107944130897522,
+      "learning_rate": 2.121636958877482e-07,
+      "loss": 0.3205,
+      "step": 10929
+    },
+    {
+      "epoch": 2.7489939637826963,
+      "grad_norm": 0.2983773946762085,
+      "learning_rate": 2.1174217847969302e-07,
+      "loss": 0.3062,
+      "step": 10930
+    },
+    {
+      "epoch": 2.749245472837022,
+      "grad_norm": 0.29391881823539734,
+      "learning_rate": 2.1132107115596434e-07,
+      "loss": 0.3116,
+      "step": 10931
+    },
+    {
+      "epoch": 2.749496981891348,
+      "grad_norm": 0.3001260459423065,
+      "learning_rate": 2.1090037395262941e-07,
+      "loss": 0.3488,
+      "step": 10932
+    },
+    {
+      "epoch": 2.749748490945674,
+      "grad_norm": 0.3142930269241333,
+      "learning_rate": 2.104800869057161e-07,
+      "loss": 0.3339,
+      "step": 10933
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.28879308700561523,
+      "learning_rate": 2.1006021005122057e-07,
+      "loss": 0.291,
+      "step": 10934
+    },
+    {
+      "epoch": 2.750251509054326,
+      "grad_norm": 0.3108893930912018,
+      "learning_rate": 2.0964074342510187e-07,
+      "loss": 0.3433,
+      "step": 10935
+    },
+    {
+      "epoch": 2.750503018108652,
+      "grad_norm": 0.30376648902893066,
+      "learning_rate": 2.0922168706328572e-07,
+      "loss": 0.3144,
+      "step": 10936
+    },
+    {
+      "epoch": 2.750754527162978,
+      "grad_norm": 0.2809644639492035,
+      "learning_rate": 2.0880304100166004e-07,
+      "loss": 0.3439,
+      "step": 10937
+    },
+    {
+      "epoch": 2.7510060362173037,
+      "grad_norm": 0.29834315180778503,
+      "learning_rate": 2.0838480527608118e-07,
+      "loss": 0.3044,
+      "step": 10938
+    },
+    {
+      "epoch": 2.75125754527163,
+      "grad_norm": 0.28084221482276917,
+      "learning_rate": 2.0796697992236713e-07,
+      "loss": 0.3288,
+      "step": 10939
+    },
+    {
+      "epoch": 2.7515090543259557,
+      "grad_norm": 0.2842169404029846,
+      "learning_rate": 2.0754956497630262e-07,
+      "loss": 0.3257,
+      "step": 10940
+    },
+    {
+      "epoch": 2.751760563380282,
+      "grad_norm": 0.2855300009250641,
+      "learning_rate": 2.0713256047363573e-07,
+      "loss": 0.3319,
+      "step": 10941
+    },
+    {
+      "epoch": 2.7520120724346078,
+      "grad_norm": 0.2749643921852112,
+      "learning_rate": 2.067159664500812e-07,
+      "loss": 0.3204,
+      "step": 10942
+    },
+    {
+      "epoch": 2.7522635814889336,
+      "grad_norm": 0.2852446436882019,
+      "learning_rate": 2.0629978294131824e-07,
+      "loss": 0.3053,
+      "step": 10943
+    },
+    {
+      "epoch": 2.75251509054326,
+      "grad_norm": 0.2823444604873657,
+      "learning_rate": 2.058840099829884e-07,
+      "loss": 0.3115,
+      "step": 10944
+    },
+    {
+      "epoch": 2.7527665995975856,
+      "grad_norm": 0.28557664155960083,
+      "learning_rate": 2.0546864761070262e-07,
+      "loss": 0.3398,
+      "step": 10945
+    },
+    {
+      "epoch": 2.7530181086519114,
+      "grad_norm": 0.28932204842567444,
+      "learning_rate": 2.050536958600313e-07,
+      "loss": 0.3197,
+      "step": 10946
+    },
+    {
+      "epoch": 2.7532696177062377,
+      "grad_norm": 0.27391183376312256,
+      "learning_rate": 2.0463915476651496e-07,
+      "loss": 0.3173,
+      "step": 10947
+    },
+    {
+      "epoch": 2.7535211267605635,
+      "grad_norm": 0.28741297125816345,
+      "learning_rate": 2.0422502436565462e-07,
+      "loss": 0.3155,
+      "step": 10948
+    },
+    {
+      "epoch": 2.7537726358148893,
+      "grad_norm": 0.2882639169692993,
+      "learning_rate": 2.038113046929191e-07,
+      "loss": 0.3166,
+      "step": 10949
+    },
+    {
+      "epoch": 2.7540241448692155,
+      "grad_norm": 0.27754950523376465,
+      "learning_rate": 2.0339799578373954e-07,
+      "loss": 0.3292,
+      "step": 10950
+    },
+    {
+      "epoch": 2.7542756539235413,
+      "grad_norm": 0.2686426043510437,
+      "learning_rate": 2.0298509767351538e-07,
+      "loss": 0.3227,
+      "step": 10951
+    },
+    {
+      "epoch": 2.754527162977867,
+      "grad_norm": 0.3171316981315613,
+      "learning_rate": 2.0257261039760667e-07,
+      "loss": 0.351,
+      "step": 10952
+    },
+    {
+      "epoch": 2.7547786720321934,
+      "grad_norm": 0.3160313367843628,
+      "learning_rate": 2.0216053399134127e-07,
+      "loss": 0.3276,
+      "step": 10953
+    },
+    {
+      "epoch": 2.755030181086519,
+      "grad_norm": 0.3004869520664215,
+      "learning_rate": 2.0174886849001207e-07,
+      "loss": 0.2926,
+      "step": 10954
+    },
+    {
+      "epoch": 2.755281690140845,
+      "grad_norm": 0.2918458580970764,
+      "learning_rate": 2.0133761392887308e-07,
+      "loss": 0.3214,
+      "step": 10955
+    },
+    {
+      "epoch": 2.7555331991951713,
+      "grad_norm": 0.279135137796402,
+      "learning_rate": 2.0092677034314834e-07,
+      "loss": 0.3276,
+      "step": 10956
+    },
+    {
+      "epoch": 2.755784708249497,
+      "grad_norm": 0.28750714659690857,
+      "learning_rate": 2.0051633776802192e-07,
+      "loss": 0.3329,
+      "step": 10957
+    },
+    {
+      "epoch": 2.756036217303823,
+      "grad_norm": 0.29169023036956787,
+      "learning_rate": 2.001063162386463e-07,
+      "loss": 0.3052,
+      "step": 10958
+    },
+    {
+      "epoch": 2.756287726358149,
+      "grad_norm": 0.3238070011138916,
+      "learning_rate": 1.996967057901361e-07,
+      "loss": 0.339,
+      "step": 10959
+    },
+    {
+      "epoch": 2.756539235412475,
+      "grad_norm": 0.31088271737098694,
+      "learning_rate": 1.9928750645757332e-07,
+      "loss": 0.3617,
+      "step": 10960
+    },
+    {
+      "epoch": 2.7567907444668007,
+      "grad_norm": 0.28064385056495667,
+      "learning_rate": 1.9887871827600158e-07,
+      "loss": 0.3263,
+      "step": 10961
+    },
+    {
+      "epoch": 2.757042253521127,
+      "grad_norm": 0.2962087094783783,
+      "learning_rate": 1.9847034128043175e-07,
+      "loss": 0.3306,
+      "step": 10962
+    },
+    {
+      "epoch": 2.7572937625754528,
+      "grad_norm": 0.2970735728740692,
+      "learning_rate": 1.9806237550583974e-07,
+      "loss": 0.3042,
+      "step": 10963
+    },
+    {
+      "epoch": 2.7575452716297786,
+      "grad_norm": 0.30512532591819763,
+      "learning_rate": 1.976548209871637e-07,
+      "loss": 0.2982,
+      "step": 10964
+    },
+    {
+      "epoch": 2.757796780684105,
+      "grad_norm": 0.3158949315547943,
+      "learning_rate": 1.9724767775930965e-07,
+      "loss": 0.3253,
+      "step": 10965
+    },
+    {
+      "epoch": 2.7580482897384306,
+      "grad_norm": 0.30506396293640137,
+      "learning_rate": 1.9684094585714575e-07,
+      "loss": 0.3229,
+      "step": 10966
+    },
+    {
+      "epoch": 2.7582997987927564,
+      "grad_norm": 0.2968839704990387,
+      "learning_rate": 1.964346253155064e-07,
+      "loss": 0.3332,
+      "step": 10967
+    },
+    {
+      "epoch": 2.7585513078470827,
+      "grad_norm": 0.28158625960350037,
+      "learning_rate": 1.9602871616918985e-07,
+      "loss": 0.3014,
+      "step": 10968
+    },
+    {
+      "epoch": 2.7588028169014085,
+      "grad_norm": 0.30511099100112915,
+      "learning_rate": 1.9562321845296106e-07,
+      "loss": 0.3046,
+      "step": 10969
+    },
+    {
+      "epoch": 2.7590543259557343,
+      "grad_norm": 0.3017299175262451,
+      "learning_rate": 1.9521813220154672e-07,
+      "loss": 0.2923,
+      "step": 10970
+    },
+    {
+      "epoch": 2.7593058350100605,
+      "grad_norm": 0.3004690706729889,
+      "learning_rate": 1.948134574496413e-07,
+      "loss": 0.3107,
+      "step": 10971
+    },
+    {
+      "epoch": 2.7595573440643864,
+      "grad_norm": 0.28788694739341736,
+      "learning_rate": 1.9440919423190208e-07,
+      "loss": 0.3172,
+      "step": 10972
+    },
+    {
+      "epoch": 2.759808853118712,
+      "grad_norm": 0.265638530254364,
+      "learning_rate": 1.9400534258295078e-07,
+      "loss": 0.3227,
+      "step": 10973
+    },
+    {
+      "epoch": 2.7600603621730384,
+      "grad_norm": 0.32271093130111694,
+      "learning_rate": 1.9360190253737698e-07,
+      "loss": 0.297,
+      "step": 10974
+    },
+    {
+      "epoch": 2.760311871227364,
+      "grad_norm": 0.30242830514907837,
+      "learning_rate": 1.9319887412973083e-07,
+      "loss": 0.3182,
+      "step": 10975
+    },
+    {
+      "epoch": 2.76056338028169,
+      "grad_norm": 0.2825239300727844,
+      "learning_rate": 1.9279625739453022e-07,
+      "loss": 0.3349,
+      "step": 10976
+    },
+    {
+      "epoch": 2.7608148893360163,
+      "grad_norm": 0.2780284285545349,
+      "learning_rate": 1.923940523662554e-07,
+      "loss": 0.3108,
+      "step": 10977
+    },
+    {
+      "epoch": 2.761066398390342,
+      "grad_norm": 0.2897617816925049,
+      "learning_rate": 1.9199225907935492e-07,
+      "loss": 0.3242,
+      "step": 10978
+    },
+    {
+      "epoch": 2.761317907444668,
+      "grad_norm": 0.287583589553833,
+      "learning_rate": 1.9159087756823792e-07,
+      "loss": 0.3197,
+      "step": 10979
+    },
+    {
+      "epoch": 2.761569416498994,
+      "grad_norm": 0.27631574869155884,
+      "learning_rate": 1.911899078672813e-07,
+      "loss": 0.3144,
+      "step": 10980
+    },
+    {
+      "epoch": 2.76182092555332,
+      "grad_norm": 0.2759281396865845,
+      "learning_rate": 1.9078935001082487e-07,
+      "loss": 0.3438,
+      "step": 10981
+    },
+    {
+      "epoch": 2.7620724346076457,
+      "grad_norm": 0.30593448877334595,
+      "learning_rate": 1.9038920403317507e-07,
+      "loss": 0.3062,
+      "step": 10982
+    },
+    {
+      "epoch": 2.762323943661972,
+      "grad_norm": 0.30097198486328125,
+      "learning_rate": 1.8998946996860002e-07,
+      "loss": 0.316,
+      "step": 10983
+    },
+    {
+      "epoch": 2.762575452716298,
+      "grad_norm": 0.28567543625831604,
+      "learning_rate": 1.8959014785133621e-07,
+      "loss": 0.3111,
+      "step": 10984
+    },
+    {
+      "epoch": 2.7628269617706236,
+      "grad_norm": 0.2774016857147217,
+      "learning_rate": 1.8919123771558246e-07,
+      "loss": 0.3204,
+      "step": 10985
+    },
+    {
+      "epoch": 2.76307847082495,
+      "grad_norm": 0.317721962928772,
+      "learning_rate": 1.8879273959550248e-07,
+      "loss": 0.3089,
+      "step": 10986
+    },
+    {
+      "epoch": 2.7633299798792756,
+      "grad_norm": 0.28425851464271545,
+      "learning_rate": 1.8839465352522623e-07,
+      "loss": 0.3328,
+      "step": 10987
+    },
+    {
+      "epoch": 2.7635814889336014,
+      "grad_norm": 0.2964561879634857,
+      "learning_rate": 1.8799697953884587e-07,
+      "loss": 0.3114,
+      "step": 10988
+    },
+    {
+      "epoch": 2.7638329979879277,
+      "grad_norm": 0.2809069752693176,
+      "learning_rate": 1.8759971767042085e-07,
+      "loss": 0.3146,
+      "step": 10989
+    },
+    {
+      "epoch": 2.7640845070422535,
+      "grad_norm": 0.305742084980011,
+      "learning_rate": 1.8720286795397335e-07,
+      "loss": 0.3204,
+      "step": 10990
+    },
+    {
+      "epoch": 2.7643360160965793,
+      "grad_norm": 0.298260360956192,
+      "learning_rate": 1.868064304234918e-07,
+      "loss": 0.3394,
+      "step": 10991
+    },
+    {
+      "epoch": 2.7645875251509056,
+      "grad_norm": 0.2834104299545288,
+      "learning_rate": 1.8641040511292786e-07,
+      "loss": 0.3074,
+      "step": 10992
+    },
+    {
+      "epoch": 2.7648390342052314,
+      "grad_norm": 0.2777165472507477,
+      "learning_rate": 1.8601479205619945e-07,
+      "loss": 0.3359,
+      "step": 10993
+    },
+    {
+      "epoch": 2.765090543259557,
+      "grad_norm": 0.29171082377433777,
+      "learning_rate": 1.856195912871872e-07,
+      "loss": 0.3155,
+      "step": 10994
+    },
+    {
+      "epoch": 2.7653420523138834,
+      "grad_norm": 0.29711124300956726,
+      "learning_rate": 1.8522480283973908e-07,
+      "loss": 0.3459,
+      "step": 10995
+    },
+    {
+      "epoch": 2.765593561368209,
+      "grad_norm": 0.32558995485305786,
+      "learning_rate": 1.8483042674766527e-07,
+      "loss": 0.3161,
+      "step": 10996
+    },
+    {
+      "epoch": 2.765845070422535,
+      "grad_norm": 0.29762402176856995,
+      "learning_rate": 1.8443646304474206e-07,
+      "loss": 0.3293,
+      "step": 10997
+    },
+    {
+      "epoch": 2.7660965794768613,
+      "grad_norm": 0.2941772937774658,
+      "learning_rate": 1.8404291176470857e-07,
+      "loss": 0.3248,
+      "step": 10998
+    },
+    {
+      "epoch": 2.766348088531187,
+      "grad_norm": 0.2893567681312561,
+      "learning_rate": 1.836497729412723e-07,
+      "loss": 0.3181,
+      "step": 10999
+    },
+    {
+      "epoch": 2.766599597585513,
+      "grad_norm": 0.28409847617149353,
+      "learning_rate": 1.8325704660810128e-07,
+      "loss": 0.3346,
+      "step": 11000
+    },
+    {
+      "epoch": 2.766851106639839,
+      "grad_norm": 0.2963768541812897,
+      "learning_rate": 1.8286473279883142e-07,
+      "loss": 0.3117,
+      "step": 11001
+    },
+    {
+      "epoch": 2.767102615694165,
+      "grad_norm": 0.30031341314315796,
+      "learning_rate": 1.8247283154706085e-07,
+      "loss": 0.3261,
+      "step": 11002
+    },
+    {
+      "epoch": 2.7673541247484907,
+      "grad_norm": 0.292661190032959,
+      "learning_rate": 1.8208134288635438e-07,
+      "loss": 0.3137,
+      "step": 11003
+    },
+    {
+      "epoch": 2.767605633802817,
+      "grad_norm": 0.2892690598964691,
+      "learning_rate": 1.816902668502396e-07,
+      "loss": 0.3256,
+      "step": 11004
+    },
+    {
+      "epoch": 2.767857142857143,
+      "grad_norm": 0.2964717745780945,
+      "learning_rate": 1.812996034722103e-07,
+      "loss": 0.3335,
+      "step": 11005
+    },
+    {
+      "epoch": 2.7681086519114686,
+      "grad_norm": 0.26202383637428284,
+      "learning_rate": 1.809093527857242e-07,
+      "loss": 0.3231,
+      "step": 11006
+    },
+    {
+      "epoch": 2.768360160965795,
+      "grad_norm": 0.26758965849876404,
+      "learning_rate": 1.805195148242045e-07,
+      "loss": 0.3121,
+      "step": 11007
+    },
+    {
+      "epoch": 2.7686116700201207,
+      "grad_norm": 0.2875462472438812,
+      "learning_rate": 1.8013008962103674e-07,
+      "loss": 0.3221,
+      "step": 11008
+    },
+    {
+      "epoch": 2.7688631790744465,
+      "grad_norm": 0.2862987220287323,
+      "learning_rate": 1.7974107720957478e-07,
+      "loss": 0.33,
+      "step": 11009
+    },
+    {
+      "epoch": 2.7691146881287727,
+      "grad_norm": 0.289815753698349,
+      "learning_rate": 1.7935247762313312e-07,
+      "loss": 0.318,
+      "step": 11010
+    },
+    {
+      "epoch": 2.7693661971830985,
+      "grad_norm": 0.2913775146007538,
+      "learning_rate": 1.7896429089499455e-07,
+      "loss": 0.3035,
+      "step": 11011
+    },
+    {
+      "epoch": 2.7696177062374243,
+      "grad_norm": 0.2775018513202667,
+      "learning_rate": 1.7857651705840419e-07,
+      "loss": 0.318,
+      "step": 11012
+    },
+    {
+      "epoch": 2.7698692152917506,
+      "grad_norm": 0.27448347210884094,
+      "learning_rate": 1.781891561465726e-07,
+      "loss": 0.339,
+      "step": 11013
+    },
+    {
+      "epoch": 2.7701207243460764,
+      "grad_norm": 0.31805574893951416,
+      "learning_rate": 1.778022081926739e-07,
+      "loss": 0.3193,
+      "step": 11014
+    },
+    {
+      "epoch": 2.770372233400402,
+      "grad_norm": 0.292459100484848,
+      "learning_rate": 1.774156732298493e-07,
+      "loss": 0.3134,
+      "step": 11015
+    },
+    {
+      "epoch": 2.7706237424547284,
+      "grad_norm": 0.28790146112442017,
+      "learning_rate": 1.7702955129120125e-07,
+      "loss": 0.3145,
+      "step": 11016
+    },
+    {
+      "epoch": 2.7708752515090542,
+      "grad_norm": 0.26574012637138367,
+      "learning_rate": 1.7664384240979993e-07,
+      "loss": 0.3131,
+      "step": 11017
+    },
+    {
+      "epoch": 2.77112676056338,
+      "grad_norm": 0.2886962890625,
+      "learning_rate": 1.7625854661867947e-07,
+      "loss": 0.2903,
+      "step": 11018
+    },
+    {
+      "epoch": 2.7713782696177063,
+      "grad_norm": 0.2845294773578644,
+      "learning_rate": 1.7587366395083683e-07,
+      "loss": 0.3289,
+      "step": 11019
+    },
+    {
+      "epoch": 2.771629778672032,
+      "grad_norm": 0.27203091979026794,
+      "learning_rate": 1.754891944392356e-07,
+      "loss": 0.3224,
+      "step": 11020
+    },
+    {
+      "epoch": 2.771881287726358,
+      "grad_norm": 0.27143409848213196,
+      "learning_rate": 1.751051381168023e-07,
+      "loss": 0.3132,
+      "step": 11021
+    },
+    {
+      "epoch": 2.772132796780684,
+      "grad_norm": 0.2910093665122986,
+      "learning_rate": 1.7472149501643e-07,
+      "loss": 0.3269,
+      "step": 11022
+    },
+    {
+      "epoch": 2.77238430583501,
+      "grad_norm": 0.29057279229164124,
+      "learning_rate": 1.7433826517097407e-07,
+      "loss": 0.3368,
+      "step": 11023
+    },
+    {
+      "epoch": 2.7726358148893357,
+      "grad_norm": 0.2875846326351166,
+      "learning_rate": 1.7395544861325718e-07,
+      "loss": 0.3195,
+      "step": 11024
+    },
+    {
+      "epoch": 2.772887323943662,
+      "grad_norm": 0.2830538749694824,
+      "learning_rate": 1.7357304537606367e-07,
+      "loss": 0.2901,
+      "step": 11025
+    },
+    {
+      "epoch": 2.773138832997988,
+      "grad_norm": 0.29084324836730957,
+      "learning_rate": 1.7319105549214564e-07,
+      "loss": 0.3114,
+      "step": 11026
+    },
+    {
+      "epoch": 2.7733903420523136,
+      "grad_norm": 0.29110777378082275,
+      "learning_rate": 1.7280947899421695e-07,
+      "loss": 0.3129,
+      "step": 11027
+    },
+    {
+      "epoch": 2.77364185110664,
+      "grad_norm": 0.2726239562034607,
+      "learning_rate": 1.7242831591495701e-07,
+      "loss": 0.3164,
+      "step": 11028
+    },
+    {
+      "epoch": 2.7738933601609657,
+      "grad_norm": 0.28240305185317993,
+      "learning_rate": 1.7204756628701192e-07,
+      "loss": 0.3127,
+      "step": 11029
+    },
+    {
+      "epoch": 2.774144869215292,
+      "grad_norm": 0.29125747084617615,
+      "learning_rate": 1.716672301429878e-07,
+      "loss": 0.3089,
+      "step": 11030
+    },
+    {
+      "epoch": 2.7743963782696177,
+      "grad_norm": 0.2842182219028473,
+      "learning_rate": 1.7128730751546086e-07,
+      "loss": 0.3137,
+      "step": 11031
+    },
+    {
+      "epoch": 2.7746478873239435,
+      "grad_norm": 0.28499892354011536,
+      "learning_rate": 1.709077984369667e-07,
+      "loss": 0.3305,
+      "step": 11032
+    },
+    {
+      "epoch": 2.7748993963782698,
+      "grad_norm": 0.29583901166915894,
+      "learning_rate": 1.7052870294000933e-07,
+      "loss": 0.3467,
+      "step": 11033
+    },
+    {
+      "epoch": 2.7751509054325956,
+      "grad_norm": 0.2947319746017456,
+      "learning_rate": 1.7015002105705502e-07,
+      "loss": 0.3388,
+      "step": 11034
+    },
+    {
+      "epoch": 2.7754024144869214,
+      "grad_norm": 0.28705930709838867,
+      "learning_rate": 1.6977175282053672e-07,
+      "loss": 0.3503,
+      "step": 11035
+    },
+    {
+      "epoch": 2.7756539235412476,
+      "grad_norm": 0.27521228790283203,
+      "learning_rate": 1.6939389826284903e-07,
+      "loss": 0.3195,
+      "step": 11036
+    },
+    {
+      "epoch": 2.7759054325955734,
+      "grad_norm": 0.2797527313232422,
+      "learning_rate": 1.690164574163544e-07,
+      "loss": 0.3026,
+      "step": 11037
+    },
+    {
+      "epoch": 2.7761569416498992,
+      "grad_norm": 0.29513728618621826,
+      "learning_rate": 1.686394303133776e-07,
+      "loss": 0.3057,
+      "step": 11038
+    },
+    {
+      "epoch": 2.7764084507042255,
+      "grad_norm": 0.27043822407722473,
+      "learning_rate": 1.6826281698620827e-07,
+      "loss": 0.3386,
+      "step": 11039
+    },
+    {
+      "epoch": 2.7766599597585513,
+      "grad_norm": 0.2766996920108795,
+      "learning_rate": 1.6788661746710178e-07,
+      "loss": 0.3178,
+      "step": 11040
+    },
+    {
+      "epoch": 2.7769114688128775,
+      "grad_norm": 0.2824966609477997,
+      "learning_rate": 1.6751083178827675e-07,
+      "loss": 0.3275,
+      "step": 11041
+    },
+    {
+      "epoch": 2.7771629778672033,
+      "grad_norm": 0.27287888526916504,
+      "learning_rate": 1.6713545998191748e-07,
+      "loss": 0.3203,
+      "step": 11042
+    },
+    {
+      "epoch": 2.777414486921529,
+      "grad_norm": 0.28277119994163513,
+      "learning_rate": 1.6676050208017102e-07,
+      "loss": 0.3313,
+      "step": 11043
+    },
+    {
+      "epoch": 2.7776659959758554,
+      "grad_norm": 0.2877527177333832,
+      "learning_rate": 1.663859581151517e-07,
+      "loss": 0.3023,
+      "step": 11044
+    },
+    {
+      "epoch": 2.777917505030181,
+      "grad_norm": 0.2927461266517639,
+      "learning_rate": 1.6601182811893545e-07,
+      "loss": 0.334,
+      "step": 11045
+    },
+    {
+      "epoch": 2.778169014084507,
+      "grad_norm": 0.28662925958633423,
+      "learning_rate": 1.6563811212356506e-07,
+      "loss": 0.32,
+      "step": 11046
+    },
+    {
+      "epoch": 2.7784205231388333,
+      "grad_norm": 0.300090491771698,
+      "learning_rate": 1.6526481016104655e-07,
+      "loss": 0.3101,
+      "step": 11047
+    },
+    {
+      "epoch": 2.778672032193159,
+      "grad_norm": 0.28218507766723633,
+      "learning_rate": 1.6489192226335104e-07,
+      "loss": 0.3205,
+      "step": 11048
+    },
+    {
+      "epoch": 2.778923541247485,
+      "grad_norm": 0.2887854278087616,
+      "learning_rate": 1.6451944846241408e-07,
+      "loss": 0.3097,
+      "step": 11049
+    },
+    {
+      "epoch": 2.779175050301811,
+      "grad_norm": 0.28003057837486267,
+      "learning_rate": 1.641473887901357e-07,
+      "loss": 0.3098,
+      "step": 11050
+    },
+    {
+      "epoch": 2.779426559356137,
+      "grad_norm": 0.2808135449886322,
+      "learning_rate": 1.6377574327838041e-07,
+      "loss": 0.319,
+      "step": 11051
+    },
+    {
+      "epoch": 2.7796780684104627,
+      "grad_norm": 0.2711102366447449,
+      "learning_rate": 1.6340451195897722e-07,
+      "loss": 0.3376,
+      "step": 11052
+    },
+    {
+      "epoch": 2.779929577464789,
+      "grad_norm": 0.30259403586387634,
+      "learning_rate": 1.6303369486372067e-07,
+      "loss": 0.3333,
+      "step": 11053
+    },
+    {
+      "epoch": 2.7801810865191148,
+      "grad_norm": 0.27301356196403503,
+      "learning_rate": 1.6266329202436758e-07,
+      "loss": 0.318,
+      "step": 11054
+    },
+    {
+      "epoch": 2.7804325955734406,
+      "grad_norm": 0.28594970703125,
+      "learning_rate": 1.622933034726415e-07,
+      "loss": 0.3366,
+      "step": 11055
+    },
+    {
+      "epoch": 2.780684104627767,
+      "grad_norm": 0.2891000509262085,
+      "learning_rate": 1.6192372924022925e-07,
+      "loss": 0.3187,
+      "step": 11056
+    },
+    {
+      "epoch": 2.7809356136820926,
+      "grad_norm": 0.299022376537323,
+      "learning_rate": 1.6155456935878277e-07,
+      "loss": 0.3196,
+      "step": 11057
+    },
+    {
+      "epoch": 2.7811871227364184,
+      "grad_norm": 0.27598682045936584,
+      "learning_rate": 1.6118582385991787e-07,
+      "loss": 0.3121,
+      "step": 11058
+    },
+    {
+      "epoch": 2.7814386317907447,
+      "grad_norm": 0.29095298051834106,
+      "learning_rate": 1.6081749277521598e-07,
+      "loss": 0.2999,
+      "step": 11059
+    },
+    {
+      "epoch": 2.7816901408450705,
+      "grad_norm": 0.28317791223526,
+      "learning_rate": 1.604495761362218e-07,
+      "loss": 0.3121,
+      "step": 11060
+    },
+    {
+      "epoch": 2.7819416498993963,
+      "grad_norm": 0.2773258686065674,
+      "learning_rate": 1.600820739744452e-07,
+      "loss": 0.3309,
+      "step": 11061
+    },
+    {
+      "epoch": 2.7821931589537225,
+      "grad_norm": 0.28628113865852356,
+      "learning_rate": 1.5971498632136096e-07,
+      "loss": 0.3194,
+      "step": 11062
+    },
+    {
+      "epoch": 2.7824446680080483,
+      "grad_norm": 0.28664273023605347,
+      "learning_rate": 1.5934831320840672e-07,
+      "loss": 0.3022,
+      "step": 11063
+    },
+    {
+      "epoch": 2.782696177062374,
+      "grad_norm": 0.2975166141986847,
+      "learning_rate": 1.5898205466698736e-07,
+      "loss": 0.3185,
+      "step": 11064
+    },
+    {
+      "epoch": 2.7829476861167004,
+      "grad_norm": 0.27436116337776184,
+      "learning_rate": 1.586162107284689e-07,
+      "loss": 0.3214,
+      "step": 11065
+    },
+    {
+      "epoch": 2.783199195171026,
+      "grad_norm": 0.2933892607688904,
+      "learning_rate": 1.5825078142418516e-07,
+      "loss": 0.3318,
+      "step": 11066
+    },
+    {
+      "epoch": 2.783450704225352,
+      "grad_norm": 0.2979046702384949,
+      "learning_rate": 1.578857667854311e-07,
+      "loss": 0.3252,
+      "step": 11067
+    },
+    {
+      "epoch": 2.7837022132796783,
+      "grad_norm": 0.2885100841522217,
+      "learning_rate": 1.5752116684347008e-07,
+      "loss": 0.3235,
+      "step": 11068
+    },
+    {
+      "epoch": 2.783953722334004,
+      "grad_norm": 0.2883824110031128,
+      "learning_rate": 1.5715698162952597e-07,
+      "loss": 0.3425,
+      "step": 11069
+    },
+    {
+      "epoch": 2.78420523138833,
+      "grad_norm": 0.2968808114528656,
+      "learning_rate": 1.5679321117478995e-07,
+      "loss": 0.3058,
+      "step": 11070
+    },
+    {
+      "epoch": 2.784456740442656,
+      "grad_norm": 0.27541083097457886,
+      "learning_rate": 1.56429855510416e-07,
+      "loss": 0.3413,
+      "step": 11071
+    },
+    {
+      "epoch": 2.784708249496982,
+      "grad_norm": 0.26949572563171387,
+      "learning_rate": 1.5606691466752366e-07,
+      "loss": 0.3173,
+      "step": 11072
+    },
+    {
+      "epoch": 2.7849597585513077,
+      "grad_norm": 0.28457048535346985,
+      "learning_rate": 1.5570438867719695e-07,
+      "loss": 0.3225,
+      "step": 11073
+    },
+    {
+      "epoch": 2.785211267605634,
+      "grad_norm": 0.27355846762657166,
+      "learning_rate": 1.553422775704827e-07,
+      "loss": 0.322,
+      "step": 11074
+    },
+    {
+      "epoch": 2.78546277665996,
+      "grad_norm": 0.2736510634422302,
+      "learning_rate": 1.5498058137839555e-07,
+      "loss": 0.3038,
+      "step": 11075
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.2820262908935547,
+      "learning_rate": 1.5461930013191018e-07,
+      "loss": 0.3123,
+      "step": 11076
+    },
+    {
+      "epoch": 2.785965794768612,
+      "grad_norm": 0.2771058976650238,
+      "learning_rate": 1.5425843386197015e-07,
+      "loss": 0.3065,
+      "step": 11077
+    },
+    {
+      "epoch": 2.7862173038229376,
+      "grad_norm": 0.2988390326499939,
+      "learning_rate": 1.5389798259947908e-07,
+      "loss": 0.3116,
+      "step": 11078
+    },
+    {
+      "epoch": 2.7864688128772634,
+      "grad_norm": 0.27741125226020813,
+      "learning_rate": 1.5353794637531005e-07,
+      "loss": 0.305,
+      "step": 11079
+    },
+    {
+      "epoch": 2.7867203219315897,
+      "grad_norm": 0.3002147078514099,
+      "learning_rate": 1.5317832522029563e-07,
+      "loss": 0.3438,
+      "step": 11080
+    },
+    {
+      "epoch": 2.7869718309859155,
+      "grad_norm": 0.2924443483352661,
+      "learning_rate": 1.5281911916523672e-07,
+      "loss": 0.3429,
+      "step": 11081
+    },
+    {
+      "epoch": 2.7872233400402413,
+      "grad_norm": 0.2645252048969269,
+      "learning_rate": 1.524603282408954e-07,
+      "loss": 0.3502,
+      "step": 11082
+    },
+    {
+      "epoch": 2.7874748490945676,
+      "grad_norm": 0.26844504475593567,
+      "learning_rate": 1.5210195247800153e-07,
+      "loss": 0.3112,
+      "step": 11083
+    },
+    {
+      "epoch": 2.7877263581488934,
+      "grad_norm": 0.2783125638961792,
+      "learning_rate": 1.517439919072472e-07,
+      "loss": 0.3137,
+      "step": 11084
+    },
+    {
+      "epoch": 2.787977867203219,
+      "grad_norm": 0.29496461153030396,
+      "learning_rate": 1.5138644655928848e-07,
+      "loss": 0.3234,
+      "step": 11085
+    },
+    {
+      "epoch": 2.7882293762575454,
+      "grad_norm": 0.3056720793247223,
+      "learning_rate": 1.5102931646474917e-07,
+      "loss": 0.3139,
+      "step": 11086
+    },
+    {
+      "epoch": 2.788480885311871,
+      "grad_norm": 0.2965206801891327,
+      "learning_rate": 1.506726016542126e-07,
+      "loss": 0.3233,
+      "step": 11087
+    },
+    {
+      "epoch": 2.788732394366197,
+      "grad_norm": 0.27822843194007874,
+      "learning_rate": 1.50316302158231e-07,
+      "loss": 0.3161,
+      "step": 11088
+    },
+    {
+      "epoch": 2.7889839034205233,
+      "grad_norm": 0.2970999479293823,
+      "learning_rate": 1.4996041800731832e-07,
+      "loss": 0.3139,
+      "step": 11089
+    },
+    {
+      "epoch": 2.789235412474849,
+      "grad_norm": 0.2687007188796997,
+      "learning_rate": 1.4960494923195457e-07,
+      "loss": 0.3199,
+      "step": 11090
+    },
+    {
+      "epoch": 2.789486921529175,
+      "grad_norm": 0.31234511733055115,
+      "learning_rate": 1.4924989586258265e-07,
+      "loss": 0.3306,
+      "step": 11091
+    },
+    {
+      "epoch": 2.789738430583501,
+      "grad_norm": 0.312133252620697,
+      "learning_rate": 1.4889525792961103e-07,
+      "loss": 0.3264,
+      "step": 11092
+    },
+    {
+      "epoch": 2.789989939637827,
+      "grad_norm": 0.28549107909202576,
+      "learning_rate": 1.4854103546341204e-07,
+      "loss": 0.3121,
+      "step": 11093
+    },
+    {
+      "epoch": 2.7902414486921527,
+      "grad_norm": 0.28895583748817444,
+      "learning_rate": 1.4818722849432253e-07,
+      "loss": 0.3081,
+      "step": 11094
+    },
+    {
+      "epoch": 2.790492957746479,
+      "grad_norm": 0.2902131974697113,
+      "learning_rate": 1.4783383705264444e-07,
+      "loss": 0.3254,
+      "step": 11095
+    },
+    {
+      "epoch": 2.790744466800805,
+      "grad_norm": 0.2827642858028412,
+      "learning_rate": 1.474808611686429e-07,
+      "loss": 0.3218,
+      "step": 11096
+    },
+    {
+      "epoch": 2.7909959758551306,
+      "grad_norm": 0.3049061894416809,
+      "learning_rate": 1.4712830087254825e-07,
+      "loss": 0.3173,
+      "step": 11097
+    },
+    {
+      "epoch": 2.791247484909457,
+      "grad_norm": 0.2864426076412201,
+      "learning_rate": 1.467761561945552e-07,
+      "loss": 0.326,
+      "step": 11098
+    },
+    {
+      "epoch": 2.7914989939637826,
+      "grad_norm": 0.28067395091056824,
+      "learning_rate": 1.4642442716482298e-07,
+      "loss": 0.3218,
+      "step": 11099
+    },
+    {
+      "epoch": 2.7917505030181085,
+      "grad_norm": 0.28718701004981995,
+      "learning_rate": 1.4607311381347467e-07,
+      "loss": 0.3006,
+      "step": 11100
+    },
+    {
+      "epoch": 2.7920020120724347,
+      "grad_norm": 0.30900463461875916,
+      "learning_rate": 1.457222161705979e-07,
+      "loss": 0.3336,
+      "step": 11101
+    },
+    {
+      "epoch": 2.7922535211267605,
+      "grad_norm": 0.31647589802742004,
+      "learning_rate": 1.453717342662453e-07,
+      "loss": 0.3224,
+      "step": 11102
+    },
+    {
+      "epoch": 2.7925050301810863,
+      "grad_norm": 0.27756398916244507,
+      "learning_rate": 1.4502166813043283e-07,
+      "loss": 0.3268,
+      "step": 11103
+    },
+    {
+      "epoch": 2.7927565392354126,
+      "grad_norm": 0.2882534861564636,
+      "learning_rate": 1.446720177931421e-07,
+      "loss": 0.3331,
+      "step": 11104
+    },
+    {
+      "epoch": 2.7930080482897384,
+      "grad_norm": 0.2910817861557007,
+      "learning_rate": 1.4432278328431748e-07,
+      "loss": 0.3222,
+      "step": 11105
+    },
+    {
+      "epoch": 2.793259557344064,
+      "grad_norm": 0.26894426345825195,
+      "learning_rate": 1.4397396463387059e-07,
+      "loss": 0.3128,
+      "step": 11106
+    },
+    {
+      "epoch": 2.7935110663983904,
+      "grad_norm": 0.28713172674179077,
+      "learning_rate": 1.4362556187167365e-07,
+      "loss": 0.3278,
+      "step": 11107
+    },
+    {
+      "epoch": 2.7937625754527162,
+      "grad_norm": 0.28895851969718933,
+      "learning_rate": 1.4327757502756668e-07,
+      "loss": 0.3292,
+      "step": 11108
+    },
+    {
+      "epoch": 2.794014084507042,
+      "grad_norm": 0.285756379365921,
+      "learning_rate": 1.4293000413135084e-07,
+      "loss": 0.3139,
+      "step": 11109
+    },
+    {
+      "epoch": 2.7942655935613683,
+      "grad_norm": 0.3005828857421875,
+      "learning_rate": 1.4258284921279565e-07,
+      "loss": 0.3371,
+      "step": 11110
+    },
+    {
+      "epoch": 2.794517102615694,
+      "grad_norm": 0.282633900642395,
+      "learning_rate": 1.4223611030163064e-07,
+      "loss": 0.3343,
+      "step": 11111
+    },
+    {
+      "epoch": 2.79476861167002,
+      "grad_norm": 0.29902184009552,
+      "learning_rate": 1.4188978742755322e-07,
+      "loss": 0.2928,
+      "step": 11112
+    },
+    {
+      "epoch": 2.795020120724346,
+      "grad_norm": 0.3080330491065979,
+      "learning_rate": 1.415438806202224e-07,
+      "loss": 0.3043,
+      "step": 11113
+    },
+    {
+      "epoch": 2.795271629778672,
+      "grad_norm": 0.2750524878501892,
+      "learning_rate": 1.4119838990926448e-07,
+      "loss": 0.3241,
+      "step": 11114
+    },
+    {
+      "epoch": 2.7955231388329977,
+      "grad_norm": 0.27586209774017334,
+      "learning_rate": 1.4085331532426748e-07,
+      "loss": 0.326,
+      "step": 11115
+    },
+    {
+      "epoch": 2.795774647887324,
+      "grad_norm": 0.29278305172920227,
+      "learning_rate": 1.405086568947861e-07,
+      "loss": 0.3371,
+      "step": 11116
+    },
+    {
+      "epoch": 2.79602615694165,
+      "grad_norm": 0.30196496844291687,
+      "learning_rate": 1.401644146503367e-07,
+      "loss": 0.3354,
+      "step": 11117
+    },
+    {
+      "epoch": 2.7962776659959756,
+      "grad_norm": 0.2952038645744324,
+      "learning_rate": 1.3982058862040238e-07,
+      "loss": 0.2915,
+      "step": 11118
+    },
+    {
+      "epoch": 2.796529175050302,
+      "grad_norm": 0.29630571603775024,
+      "learning_rate": 1.3947717883442903e-07,
+      "loss": 0.3196,
+      "step": 11119
+    },
+    {
+      "epoch": 2.7967806841046277,
+      "grad_norm": 0.27728769183158875,
+      "learning_rate": 1.391341853218281e-07,
+      "loss": 0.3405,
+      "step": 11120
+    },
+    {
+      "epoch": 2.7970321931589535,
+      "grad_norm": 0.272607684135437,
+      "learning_rate": 1.3879160811197556e-07,
+      "loss": 0.3274,
+      "step": 11121
+    },
+    {
+      "epoch": 2.7972837022132797,
+      "grad_norm": 0.29455098509788513,
+      "learning_rate": 1.3844944723420906e-07,
+      "loss": 0.3242,
+      "step": 11122
+    },
+    {
+      "epoch": 2.7975352112676055,
+      "grad_norm": 0.2698933184146881,
+      "learning_rate": 1.3810770271783457e-07,
+      "loss": 0.3153,
+      "step": 11123
+    },
+    {
+      "epoch": 2.7977867203219313,
+      "grad_norm": 0.2703859806060791,
+      "learning_rate": 1.3776637459211872e-07,
+      "loss": 0.3044,
+      "step": 11124
+    },
+    {
+      "epoch": 2.7980382293762576,
+      "grad_norm": 0.29078856110572815,
+      "learning_rate": 1.3742546288629532e-07,
+      "loss": 0.3257,
+      "step": 11125
+    },
+    {
+      "epoch": 2.7982897384305834,
+      "grad_norm": 0.2826457917690277,
+      "learning_rate": 1.3708496762956047e-07,
+      "loss": 0.3087,
+      "step": 11126
+    },
+    {
+      "epoch": 2.7985412474849096,
+      "grad_norm": 0.2665698826313019,
+      "learning_rate": 1.367448888510764e-07,
+      "loss": 0.3194,
+      "step": 11127
+    },
+    {
+      "epoch": 2.7987927565392354,
+      "grad_norm": 0.27445247769355774,
+      "learning_rate": 1.3640522657996757e-07,
+      "loss": 0.3108,
+      "step": 11128
+    },
+    {
+      "epoch": 2.7990442655935612,
+      "grad_norm": 0.2883208096027374,
+      "learning_rate": 1.3606598084532517e-07,
+      "loss": 0.3259,
+      "step": 11129
+    },
+    {
+      "epoch": 2.7992957746478875,
+      "grad_norm": 0.2990553677082062,
+      "learning_rate": 1.357271516762021e-07,
+      "loss": 0.3369,
+      "step": 11130
+    },
+    {
+      "epoch": 2.7995472837022133,
+      "grad_norm": 0.2815016806125641,
+      "learning_rate": 1.3538873910161788e-07,
+      "loss": 0.3235,
+      "step": 11131
+    },
+    {
+      "epoch": 2.799798792756539,
+      "grad_norm": 0.2700446844100952,
+      "learning_rate": 1.3505074315055545e-07,
+      "loss": 0.3025,
+      "step": 11132
+    },
+    {
+      "epoch": 2.8000503018108653,
+      "grad_norm": 0.2885153889656067,
+      "learning_rate": 1.347131638519622e-07,
+      "loss": 0.3216,
+      "step": 11133
+    },
+    {
+      "epoch": 2.800301810865191,
+      "grad_norm": 0.2758094370365143,
+      "learning_rate": 1.3437600123474837e-07,
+      "loss": 0.3108,
+      "step": 11134
+    },
+    {
+      "epoch": 2.800553319919517,
+      "grad_norm": 0.2886246144771576,
+      "learning_rate": 1.3403925532779137e-07,
+      "loss": 0.3076,
+      "step": 11135
+    },
+    {
+      "epoch": 2.800804828973843,
+      "grad_norm": 0.28727173805236816,
+      "learning_rate": 1.3370292615993098e-07,
+      "loss": 0.3063,
+      "step": 11136
+    },
+    {
+      "epoch": 2.801056338028169,
+      "grad_norm": 0.2814168632030487,
+      "learning_rate": 1.333670137599713e-07,
+      "loss": 0.3179,
+      "step": 11137
+    },
+    {
+      "epoch": 2.801307847082495,
+      "grad_norm": 0.28780585527420044,
+      "learning_rate": 1.3303151815668103e-07,
+      "loss": 0.3431,
+      "step": 11138
+    },
+    {
+      "epoch": 2.801559356136821,
+      "grad_norm": 0.296987384557724,
+      "learning_rate": 1.3269643937879384e-07,
+      "loss": 0.3119,
+      "step": 11139
+    },
+    {
+      "epoch": 2.801810865191147,
+      "grad_norm": 0.2830405831336975,
+      "learning_rate": 1.323617774550068e-07,
+      "loss": 0.3119,
+      "step": 11140
+    },
+    {
+      "epoch": 2.802062374245473,
+      "grad_norm": 0.29808276891708374,
+      "learning_rate": 1.3202753241398192e-07,
+      "loss": 0.3218,
+      "step": 11141
+    },
+    {
+      "epoch": 2.802313883299799,
+      "grad_norm": 0.27036961913108826,
+      "learning_rate": 1.316937042843447e-07,
+      "loss": 0.3352,
+      "step": 11142
+    },
+    {
+      "epoch": 2.8025653923541247,
+      "grad_norm": 0.2818096876144409,
+      "learning_rate": 1.3136029309468612e-07,
+      "loss": 0.3312,
+      "step": 11143
+    },
+    {
+      "epoch": 2.802816901408451,
+      "grad_norm": 0.3205699026584625,
+      "learning_rate": 1.3102729887355947e-07,
+      "loss": 0.3306,
+      "step": 11144
+    },
+    {
+      "epoch": 2.8030684104627768,
+      "grad_norm": 0.28567755222320557,
+      "learning_rate": 1.3069472164948526e-07,
+      "loss": 0.3176,
+      "step": 11145
+    },
+    {
+      "epoch": 2.8033199195171026,
+      "grad_norm": 0.30168774724006653,
+      "learning_rate": 1.3036256145094516e-07,
+      "loss": 0.3242,
+      "step": 11146
+    },
+    {
+      "epoch": 2.803571428571429,
+      "grad_norm": 0.304684579372406,
+      "learning_rate": 1.3003081830638752e-07,
+      "loss": 0.3341,
+      "step": 11147
+    },
+    {
+      "epoch": 2.8038229376257546,
+      "grad_norm": 0.29443493485450745,
+      "learning_rate": 1.2969949224422407e-07,
+      "loss": 0.3268,
+      "step": 11148
+    },
+    {
+      "epoch": 2.8040744466800804,
+      "grad_norm": 0.2812149226665497,
+      "learning_rate": 1.2936858329283043e-07,
+      "loss": 0.3193,
+      "step": 11149
+    },
+    {
+      "epoch": 2.8043259557344067,
+      "grad_norm": 0.3072992265224457,
+      "learning_rate": 1.290380914805478e-07,
+      "loss": 0.3232,
+      "step": 11150
+    },
+    {
+      "epoch": 2.8045774647887325,
+      "grad_norm": 0.2899894416332245,
+      "learning_rate": 1.2870801683567913e-07,
+      "loss": 0.3289,
+      "step": 11151
+    },
+    {
+      "epoch": 2.8048289738430583,
+      "grad_norm": 0.29002153873443604,
+      "learning_rate": 1.2837835938649456e-07,
+      "loss": 0.3479,
+      "step": 11152
+    },
+    {
+      "epoch": 2.8050804828973845,
+      "grad_norm": 0.30149370431900024,
+      "learning_rate": 1.2804911916122596e-07,
+      "loss": 0.305,
+      "step": 11153
+    },
+    {
+      "epoch": 2.8053319919517103,
+      "grad_norm": 0.2866424322128296,
+      "learning_rate": 1.2772029618807247e-07,
+      "loss": 0.3309,
+      "step": 11154
+    },
+    {
+      "epoch": 2.805583501006036,
+      "grad_norm": 0.3057340383529663,
+      "learning_rate": 1.2739189049519429e-07,
+      "loss": 0.2923,
+      "step": 11155
+    },
+    {
+      "epoch": 2.8058350100603624,
+      "grad_norm": 0.27844423055648804,
+      "learning_rate": 1.2706390211071784e-07,
+      "loss": 0.3382,
+      "step": 11156
+    },
+    {
+      "epoch": 2.806086519114688,
+      "grad_norm": 0.28191810846328735,
+      "learning_rate": 1.2673633106273286e-07,
+      "loss": 0.3181,
+      "step": 11157
+    },
+    {
+      "epoch": 2.806338028169014,
+      "grad_norm": 0.2744201719760895,
+      "learning_rate": 1.2640917737929414e-07,
+      "loss": 0.3168,
+      "step": 11158
+    },
+    {
+      "epoch": 2.8065895372233403,
+      "grad_norm": 0.28745532035827637,
+      "learning_rate": 1.2608244108842094e-07,
+      "loss": 0.3212,
+      "step": 11159
+    },
+    {
+      "epoch": 2.806841046277666,
+      "grad_norm": 0.2886742055416107,
+      "learning_rate": 1.2575612221809476e-07,
+      "loss": 0.3253,
+      "step": 11160
+    },
+    {
+      "epoch": 2.807092555331992,
+      "grad_norm": 0.2701511085033417,
+      "learning_rate": 1.2543022079626376e-07,
+      "loss": 0.3295,
+      "step": 11161
+    },
+    {
+      "epoch": 2.807344064386318,
+      "grad_norm": 0.2751189172267914,
+      "learning_rate": 1.25104736850839e-07,
+      "loss": 0.3468,
+      "step": 11162
+    },
+    {
+      "epoch": 2.807595573440644,
+      "grad_norm": 0.2670914828777313,
+      "learning_rate": 1.2477967040969708e-07,
+      "loss": 0.3246,
+      "step": 11163
+    },
+    {
+      "epoch": 2.8078470824949697,
+      "grad_norm": 0.27149543166160583,
+      "learning_rate": 1.2445502150067623e-07,
+      "loss": 0.3093,
+      "step": 11164
+    },
+    {
+      "epoch": 2.808098591549296,
+      "grad_norm": 0.3040313124656677,
+      "learning_rate": 1.2413079015158202e-07,
+      "loss": 0.3027,
+      "step": 11165
+    },
+    {
+      "epoch": 2.808350100603622,
+      "grad_norm": 0.2989967465400696,
+      "learning_rate": 1.238069763901817e-07,
+      "loss": 0.3109,
+      "step": 11166
+    },
+    {
+      "epoch": 2.8086016096579476,
+      "grad_norm": 0.2808802127838135,
+      "learning_rate": 1.2348358024420914e-07,
+      "loss": 0.3019,
+      "step": 11167
+    },
+    {
+      "epoch": 2.808853118712274,
+      "grad_norm": 0.28532177209854126,
+      "learning_rate": 1.2316060174136e-07,
+      "loss": 0.3205,
+      "step": 11168
+    },
+    {
+      "epoch": 2.8091046277665996,
+      "grad_norm": 0.29730409383773804,
+      "learning_rate": 1.2283804090929608e-07,
+      "loss": 0.3281,
+      "step": 11169
+    },
+    {
+      "epoch": 2.8093561368209254,
+      "grad_norm": 0.28388866782188416,
+      "learning_rate": 1.2251589777564242e-07,
+      "loss": 0.3285,
+      "step": 11170
+    },
+    {
+      "epoch": 2.8096076458752517,
+      "grad_norm": 0.2799801230430603,
+      "learning_rate": 1.221941723679887e-07,
+      "loss": 0.3403,
+      "step": 11171
+    },
+    {
+      "epoch": 2.8098591549295775,
+      "grad_norm": 0.2923330068588257,
+      "learning_rate": 1.2187286471388893e-07,
+      "loss": 0.36,
+      "step": 11172
+    },
+    {
+      "epoch": 2.8101106639839033,
+      "grad_norm": 0.31289908289909363,
+      "learning_rate": 1.2155197484086055e-07,
+      "loss": 0.3146,
+      "step": 11173
+    },
+    {
+      "epoch": 2.8103621730382295,
+      "grad_norm": 0.30059537291526794,
+      "learning_rate": 1.2123150277638662e-07,
+      "loss": 0.3224,
+      "step": 11174
+    },
+    {
+      "epoch": 2.8106136820925554,
+      "grad_norm": 0.27516844868659973,
+      "learning_rate": 1.2091144854791237e-07,
+      "loss": 0.3201,
+      "step": 11175
+    },
+    {
+      "epoch": 2.810865191146881,
+      "grad_norm": 0.2999837398529053,
+      "learning_rate": 1.2059181218284922e-07,
+      "loss": 0.3061,
+      "step": 11176
+    },
+    {
+      "epoch": 2.8111167002012074,
+      "grad_norm": 0.2801435887813568,
+      "learning_rate": 1.2027259370857193e-07,
+      "loss": 0.3218,
+      "step": 11177
+    },
+    {
+      "epoch": 2.811368209255533,
+      "grad_norm": 0.28390711545944214,
+      "learning_rate": 1.1995379315241972e-07,
+      "loss": 0.3263,
+      "step": 11178
+    },
+    {
+      "epoch": 2.811619718309859,
+      "grad_norm": 0.28144189715385437,
+      "learning_rate": 1.1963541054169526e-07,
+      "loss": 0.3042,
+      "step": 11179
+    },
+    {
+      "epoch": 2.8118712273641853,
+      "grad_norm": 0.3061021566390991,
+      "learning_rate": 1.193174459036661e-07,
+      "loss": 0.3118,
+      "step": 11180
+    },
+    {
+      "epoch": 2.812122736418511,
+      "grad_norm": 0.267780601978302,
+      "learning_rate": 1.1899989926556498e-07,
+      "loss": 0.334,
+      "step": 11181
+    },
+    {
+      "epoch": 2.812374245472837,
+      "grad_norm": 0.301758736371994,
+      "learning_rate": 1.1868277065458677e-07,
+      "loss": 0.3052,
+      "step": 11182
+    },
+    {
+      "epoch": 2.812625754527163,
+      "grad_norm": 0.2828764319419861,
+      "learning_rate": 1.1836606009789198e-07,
+      "loss": 0.3152,
+      "step": 11183
+    },
+    {
+      "epoch": 2.812877263581489,
+      "grad_norm": 0.27735185623168945,
+      "learning_rate": 1.1804976762260445e-07,
+      "loss": 0.3104,
+      "step": 11184
+    },
+    {
+      "epoch": 2.8131287726358147,
+      "grad_norm": 0.2653750777244568,
+      "learning_rate": 1.1773389325581363e-07,
+      "loss": 0.3207,
+      "step": 11185
+    },
+    {
+      "epoch": 2.813380281690141,
+      "grad_norm": 0.2903718948364258,
+      "learning_rate": 1.1741843702457068e-07,
+      "loss": 0.3296,
+      "step": 11186
+    },
+    {
+      "epoch": 2.813631790744467,
+      "grad_norm": 0.2752559185028076,
+      "learning_rate": 1.171033989558934e-07,
+      "loss": 0.3011,
+      "step": 11187
+    },
+    {
+      "epoch": 2.8138832997987926,
+      "grad_norm": 0.3056381642818451,
+      "learning_rate": 1.16788779076763e-07,
+      "loss": 0.3116,
+      "step": 11188
+    },
+    {
+      "epoch": 2.814134808853119,
+      "grad_norm": 0.28672274947166443,
+      "learning_rate": 1.164745774141246e-07,
+      "loss": 0.3503,
+      "step": 11189
+    },
+    {
+      "epoch": 2.8143863179074446,
+      "grad_norm": 0.2924738824367523,
+      "learning_rate": 1.1616079399488667e-07,
+      "loss": 0.3123,
+      "step": 11190
+    },
+    {
+      "epoch": 2.8146378269617705,
+      "grad_norm": 0.2763361930847168,
+      "learning_rate": 1.1584742884592382e-07,
+      "loss": 0.2954,
+      "step": 11191
+    },
+    {
+      "epoch": 2.8148893360160967,
+      "grad_norm": 0.28140226006507874,
+      "learning_rate": 1.1553448199407346e-07,
+      "loss": 0.3254,
+      "step": 11192
+    },
+    {
+      "epoch": 2.8151408450704225,
+      "grad_norm": 0.281065434217453,
+      "learning_rate": 1.1522195346613752e-07,
+      "loss": 0.3213,
+      "step": 11193
+    },
+    {
+      "epoch": 2.8153923541247483,
+      "grad_norm": 0.2887207269668579,
+      "learning_rate": 1.1490984328888288e-07,
+      "loss": 0.3029,
+      "step": 11194
+    },
+    {
+      "epoch": 2.8156438631790746,
+      "grad_norm": 0.26352542638778687,
+      "learning_rate": 1.1459815148903819e-07,
+      "loss": 0.3126,
+      "step": 11195
+    },
+    {
+      "epoch": 2.8158953722334004,
+      "grad_norm": 0.2894093692302704,
+      "learning_rate": 1.1428687809329986e-07,
+      "loss": 0.3023,
+      "step": 11196
+    },
+    {
+      "epoch": 2.816146881287726,
+      "grad_norm": 0.285803884267807,
+      "learning_rate": 1.139760231283249e-07,
+      "loss": 0.3372,
+      "step": 11197
+    },
+    {
+      "epoch": 2.8163983903420524,
+      "grad_norm": 0.2947225868701935,
+      "learning_rate": 1.1366558662073646e-07,
+      "loss": 0.3078,
+      "step": 11198
+    },
+    {
+      "epoch": 2.816649899396378,
+      "grad_norm": 0.3021707236766815,
+      "learning_rate": 1.1335556859712216e-07,
+      "loss": 0.3293,
+      "step": 11199
+    },
+    {
+      "epoch": 2.816901408450704,
+      "grad_norm": 0.27112025022506714,
+      "learning_rate": 1.1304596908403243e-07,
+      "loss": 0.3064,
+      "step": 11200
+    },
+    {
+      "epoch": 2.8171529175050303,
+      "grad_norm": 0.2846040427684784,
+      "learning_rate": 1.1273678810798272e-07,
+      "loss": 0.3141,
+      "step": 11201
+    },
+    {
+      "epoch": 2.817404426559356,
+      "grad_norm": 0.28327932953834534,
+      "learning_rate": 1.1242802569545241e-07,
+      "loss": 0.334,
+      "step": 11202
+    },
+    {
+      "epoch": 2.817655935613682,
+      "grad_norm": 0.2747574746608734,
+      "learning_rate": 1.1211968187288535e-07,
+      "loss": 0.3375,
+      "step": 11203
+    },
+    {
+      "epoch": 2.817907444668008,
+      "grad_norm": 0.2933041751384735,
+      "learning_rate": 1.1181175666668931e-07,
+      "loss": 0.3076,
+      "step": 11204
+    },
+    {
+      "epoch": 2.818158953722334,
+      "grad_norm": 0.288150817155838,
+      "learning_rate": 1.1150425010323541e-07,
+      "loss": 0.3362,
+      "step": 11205
+    },
+    {
+      "epoch": 2.8184104627766597,
+      "grad_norm": 0.27248165011405945,
+      "learning_rate": 1.1119716220886034e-07,
+      "loss": 0.3292,
+      "step": 11206
+    },
+    {
+      "epoch": 2.818661971830986,
+      "grad_norm": 0.2919750511646271,
+      "learning_rate": 1.1089049300986421e-07,
+      "loss": 0.3198,
+      "step": 11207
+    },
+    {
+      "epoch": 2.818913480885312,
+      "grad_norm": 0.28943032026290894,
+      "learning_rate": 1.1058424253251099e-07,
+      "loss": 0.3201,
+      "step": 11208
+    },
+    {
+      "epoch": 2.8191649899396376,
+      "grad_norm": 0.28583675622940063,
+      "learning_rate": 1.102784108030297e-07,
+      "loss": 0.3293,
+      "step": 11209
+    },
+    {
+      "epoch": 2.819416498993964,
+      "grad_norm": 0.2858569025993347,
+      "learning_rate": 1.0997299784761218e-07,
+      "loss": 0.3149,
+      "step": 11210
+    },
+    {
+      "epoch": 2.8196680080482897,
+      "grad_norm": 0.3017948269844055,
+      "learning_rate": 1.0966800369241526e-07,
+      "loss": 0.3415,
+      "step": 11211
+    },
+    {
+      "epoch": 2.8199195171026155,
+      "grad_norm": 0.2693357765674591,
+      "learning_rate": 1.0936342836356028e-07,
+      "loss": 0.3366,
+      "step": 11212
+    },
+    {
+      "epoch": 2.8201710261569417,
+      "grad_norm": 0.29551607370376587,
+      "learning_rate": 1.0905927188713195e-07,
+      "loss": 0.3274,
+      "step": 11213
+    },
+    {
+      "epoch": 2.8204225352112675,
+      "grad_norm": 0.3079835772514343,
+      "learning_rate": 1.0875553428917939e-07,
+      "loss": 0.3176,
+      "step": 11214
+    },
+    {
+      "epoch": 2.8206740442655933,
+      "grad_norm": 0.28076887130737305,
+      "learning_rate": 1.0845221559571517e-07,
+      "loss": 0.3163,
+      "step": 11215
+    },
+    {
+      "epoch": 2.8209255533199196,
+      "grad_norm": 0.2994937598705292,
+      "learning_rate": 1.0814931583271848e-07,
+      "loss": 0.3265,
+      "step": 11216
+    },
+    {
+      "epoch": 2.8211770623742454,
+      "grad_norm": 0.293682724237442,
+      "learning_rate": 1.0784683502612858e-07,
+      "loss": 0.3358,
+      "step": 11217
+    },
+    {
+      "epoch": 2.821428571428571,
+      "grad_norm": 0.3048734664916992,
+      "learning_rate": 1.0754477320185253e-07,
+      "loss": 0.3298,
+      "step": 11218
+    },
+    {
+      "epoch": 2.8216800804828974,
+      "grad_norm": 0.27152693271636963,
+      "learning_rate": 1.0724313038575906e-07,
+      "loss": 0.3252,
+      "step": 11219
+    },
+    {
+      "epoch": 2.8219315895372232,
+      "grad_norm": 0.3005521893501282,
+      "learning_rate": 1.0694190660368309e-07,
+      "loss": 0.3093,
+      "step": 11220
+    },
+    {
+      "epoch": 2.822183098591549,
+      "grad_norm": 0.281059205532074,
+      "learning_rate": 1.0664110188142118e-07,
+      "loss": 0.3335,
+      "step": 11221
+    },
+    {
+      "epoch": 2.8224346076458753,
+      "grad_norm": 0.2841849625110626,
+      "learning_rate": 1.0634071624473719e-07,
+      "loss": 0.3297,
+      "step": 11222
+    },
+    {
+      "epoch": 2.822686116700201,
+      "grad_norm": 0.29039302468299866,
+      "learning_rate": 1.0604074971935497e-07,
+      "loss": 0.323,
+      "step": 11223
+    },
+    {
+      "epoch": 2.822937625754527,
+      "grad_norm": 0.2728997766971588,
+      "learning_rate": 1.0574120233096674e-07,
+      "loss": 0.3357,
+      "step": 11224
+    },
+    {
+      "epoch": 2.823189134808853,
+      "grad_norm": 0.2747051417827606,
+      "learning_rate": 1.0544207410522644e-07,
+      "loss": 0.324,
+      "step": 11225
+    },
+    {
+      "epoch": 2.823440643863179,
+      "grad_norm": 0.286855548620224,
+      "learning_rate": 1.0514336506775135e-07,
+      "loss": 0.2887,
+      "step": 11226
+    },
+    {
+      "epoch": 2.823692152917505,
+      "grad_norm": 0.28169578313827515,
+      "learning_rate": 1.0484507524412602e-07,
+      "loss": 0.3123,
+      "step": 11227
+    },
+    {
+      "epoch": 2.823943661971831,
+      "grad_norm": 0.2875922918319702,
+      "learning_rate": 1.0454720465989498e-07,
+      "loss": 0.3339,
+      "step": 11228
+    },
+    {
+      "epoch": 2.824195171026157,
+      "grad_norm": 0.31413838267326355,
+      "learning_rate": 1.0424975334057064e-07,
+      "loss": 0.3335,
+      "step": 11229
+    },
+    {
+      "epoch": 2.824446680080483,
+      "grad_norm": 0.2952876091003418,
+      "learning_rate": 1.0395272131162703e-07,
+      "loss": 0.3128,
+      "step": 11230
+    },
+    {
+      "epoch": 2.824698189134809,
+      "grad_norm": 0.2815072238445282,
+      "learning_rate": 1.0365610859850328e-07,
+      "loss": 0.324,
+      "step": 11231
+    },
+    {
+      "epoch": 2.8249496981891347,
+      "grad_norm": 0.2810538709163666,
+      "learning_rate": 1.0335991522660239e-07,
+      "loss": 0.3117,
+      "step": 11232
+    },
+    {
+      "epoch": 2.825201207243461,
+      "grad_norm": 0.2817157804965973,
+      "learning_rate": 1.0306414122129127e-07,
+      "loss": 0.3148,
+      "step": 11233
+    },
+    {
+      "epoch": 2.8254527162977867,
+      "grad_norm": 0.2859732210636139,
+      "learning_rate": 1.0276878660790135e-07,
+      "loss": 0.3146,
+      "step": 11234
+    },
+    {
+      "epoch": 2.8257042253521125,
+      "grad_norm": 0.2922956347465515,
+      "learning_rate": 1.0247385141172794e-07,
+      "loss": 0.3036,
+      "step": 11235
+    },
+    {
+      "epoch": 2.8259557344064388,
+      "grad_norm": 0.2939150631427765,
+      "learning_rate": 1.0217933565803085e-07,
+      "loss": 0.3114,
+      "step": 11236
+    },
+    {
+      "epoch": 2.8262072434607646,
+      "grad_norm": 0.2930050492286682,
+      "learning_rate": 1.018852393720321e-07,
+      "loss": 0.3278,
+      "step": 11237
+    },
+    {
+      "epoch": 2.8264587525150904,
+      "grad_norm": 0.2771678566932678,
+      "learning_rate": 1.01591562578921e-07,
+      "loss": 0.3314,
+      "step": 11238
+    },
+    {
+      "epoch": 2.8267102615694166,
+      "grad_norm": 0.2838783264160156,
+      "learning_rate": 1.0129830530384743e-07,
+      "loss": 0.302,
+      "step": 11239
+    },
+    {
+      "epoch": 2.8269617706237424,
+      "grad_norm": 0.28282177448272705,
+      "learning_rate": 1.0100546757192853e-07,
+      "loss": 0.3397,
+      "step": 11240
+    },
+    {
+      "epoch": 2.8272132796780687,
+      "grad_norm": 0.28225719928741455,
+      "learning_rate": 1.0071304940824255e-07,
+      "loss": 0.3246,
+      "step": 11241
+    },
+    {
+      "epoch": 2.8274647887323945,
+      "grad_norm": 0.29289302229881287,
+      "learning_rate": 1.0042105083783449e-07,
+      "loss": 0.312,
+      "step": 11242
+    },
+    {
+      "epoch": 2.8277162977867203,
+      "grad_norm": 0.2851783335208893,
+      "learning_rate": 1.0012947188571154e-07,
+      "loss": 0.3026,
+      "step": 11243
+    },
+    {
+      "epoch": 2.8279678068410465,
+      "grad_norm": 0.3011089265346527,
+      "learning_rate": 9.983831257684651e-08,
+      "loss": 0.3134,
+      "step": 11244
+    },
+    {
+      "epoch": 2.8282193158953723,
+      "grad_norm": 0.28961464762687683,
+      "learning_rate": 9.95475729361739e-08,
+      "loss": 0.3119,
+      "step": 11245
+    },
+    {
+      "epoch": 2.828470824949698,
+      "grad_norm": 0.30916205048561096,
+      "learning_rate": 9.92572529885949e-08,
+      "loss": 0.3098,
+      "step": 11246
+    },
+    {
+      "epoch": 2.8287223340040244,
+      "grad_norm": 0.29237234592437744,
+      "learning_rate": 9.896735275897296e-08,
+      "loss": 0.3082,
+      "step": 11247
+    },
+    {
+      "epoch": 2.82897384305835,
+      "grad_norm": 0.3051128387451172,
+      "learning_rate": 9.867787227213655e-08,
+      "loss": 0.3121,
+      "step": 11248
+    },
+    {
+      "epoch": 2.829225352112676,
+      "grad_norm": 0.2796797454357147,
+      "learning_rate": 9.838881155287806e-08,
+      "loss": 0.3058,
+      "step": 11249
+    },
+    {
+      "epoch": 2.8294768611670023,
+      "grad_norm": 0.28661051392555237,
+      "learning_rate": 9.810017062595322e-08,
+      "loss": 0.3158,
+      "step": 11250
+    },
+    {
+      "epoch": 2.829728370221328,
+      "grad_norm": 0.28525805473327637,
+      "learning_rate": 9.781194951608286e-08,
+      "loss": 0.3155,
+      "step": 11251
+    },
+    {
+      "epoch": 2.829979879275654,
+      "grad_norm": 0.2928963601589203,
+      "learning_rate": 9.75241482479511e-08,
+      "loss": 0.3174,
+      "step": 11252
+    },
+    {
+      "epoch": 2.83023138832998,
+      "grad_norm": 0.28580477833747864,
+      "learning_rate": 9.723676684620542e-08,
+      "loss": 0.3443,
+      "step": 11253
+    },
+    {
+      "epoch": 2.830482897384306,
+      "grad_norm": 0.29559531807899475,
+      "learning_rate": 9.694980533546005e-08,
+      "loss": 0.3303,
+      "step": 11254
+    },
+    {
+      "epoch": 2.8307344064386317,
+      "grad_norm": 0.27034255862236023,
+      "learning_rate": 9.666326374028979e-08,
+      "loss": 0.3093,
+      "step": 11255
+    },
+    {
+      "epoch": 2.830985915492958,
+      "grad_norm": 0.29185721278190613,
+      "learning_rate": 9.637714208523664e-08,
+      "loss": 0.331,
+      "step": 11256
+    },
+    {
+      "epoch": 2.8312374245472838,
+      "grad_norm": 0.2903611660003662,
+      "learning_rate": 9.609144039480323e-08,
+      "loss": 0.3167,
+      "step": 11257
+    },
+    {
+      "epoch": 2.8314889336016096,
+      "grad_norm": 0.28841114044189453,
+      "learning_rate": 9.580615869345999e-08,
+      "loss": 0.3126,
+      "step": 11258
+    },
+    {
+      "epoch": 2.831740442655936,
+      "grad_norm": 0.2874075174331665,
+      "learning_rate": 9.55212970056385e-08,
+      "loss": 0.3075,
+      "step": 11259
+    },
+    {
+      "epoch": 2.8319919517102616,
+      "grad_norm": 0.2626476287841797,
+      "learning_rate": 9.52368553557359e-08,
+      "loss": 0.3212,
+      "step": 11260
+    },
+    {
+      "epoch": 2.8322434607645874,
+      "grad_norm": 0.28571265935897827,
+      "learning_rate": 9.49528337681116e-08,
+      "loss": 0.3141,
+      "step": 11261
+    },
+    {
+      "epoch": 2.8324949698189137,
+      "grad_norm": 0.27694886922836304,
+      "learning_rate": 9.466923226709223e-08,
+      "loss": 0.3197,
+      "step": 11262
+    },
+    {
+      "epoch": 2.8327464788732395,
+      "grad_norm": 0.30933791399002075,
+      "learning_rate": 9.43860508769645e-08,
+      "loss": 0.333,
+      "step": 11263
+    },
+    {
+      "epoch": 2.8329979879275653,
+      "grad_norm": 0.28589534759521484,
+      "learning_rate": 9.410328962198289e-08,
+      "loss": 0.3427,
+      "step": 11264
+    },
+    {
+      "epoch": 2.8332494969818915,
+      "grad_norm": 0.2972225248813629,
+      "learning_rate": 9.382094852636303e-08,
+      "loss": 0.3231,
+      "step": 11265
+    },
+    {
+      "epoch": 2.8335010060362174,
+      "grad_norm": 0.2830668091773987,
+      "learning_rate": 9.353902761428557e-08,
+      "loss": 0.3156,
+      "step": 11266
+    },
+    {
+      "epoch": 2.833752515090543,
+      "grad_norm": 0.30200186371803284,
+      "learning_rate": 9.325752690989676e-08,
+      "loss": 0.3334,
+      "step": 11267
+    },
+    {
+      "epoch": 2.8340040241448694,
+      "grad_norm": 0.27993977069854736,
+      "learning_rate": 9.297644643730342e-08,
+      "loss": 0.3043,
+      "step": 11268
+    },
+    {
+      "epoch": 2.834255533199195,
+      "grad_norm": 0.27513226866722107,
+      "learning_rate": 9.269578622057962e-08,
+      "loss": 0.3329,
+      "step": 11269
+    },
+    {
+      "epoch": 2.834507042253521,
+      "grad_norm": 0.2780003249645233,
+      "learning_rate": 9.241554628376115e-08,
+      "loss": 0.3006,
+      "step": 11270
+    },
+    {
+      "epoch": 2.8347585513078473,
+      "grad_norm": 0.2845858633518219,
+      "learning_rate": 9.213572665084991e-08,
+      "loss": 0.3041,
+      "step": 11271
+    },
+    {
+      "epoch": 2.835010060362173,
+      "grad_norm": 0.27587416768074036,
+      "learning_rate": 9.185632734581008e-08,
+      "loss": 0.3032,
+      "step": 11272
+    },
+    {
+      "epoch": 2.835261569416499,
+      "grad_norm": 0.28815388679504395,
+      "learning_rate": 9.157734839257026e-08,
+      "loss": 0.3105,
+      "step": 11273
+    },
+    {
+      "epoch": 2.835513078470825,
+      "grad_norm": 0.279449999332428,
+      "learning_rate": 9.12987898150236e-08,
+      "loss": 0.3063,
+      "step": 11274
+    },
+    {
+      "epoch": 2.835764587525151,
+      "grad_norm": 0.2929026484489441,
+      "learning_rate": 9.102065163702767e-08,
+      "loss": 0.3263,
+      "step": 11275
+    },
+    {
+      "epoch": 2.8360160965794767,
+      "grad_norm": 0.28392016887664795,
+      "learning_rate": 9.074293388240118e-08,
+      "loss": 0.3048,
+      "step": 11276
+    },
+    {
+      "epoch": 2.836267605633803,
+      "grad_norm": 0.27088797092437744,
+      "learning_rate": 9.046563657493068e-08,
+      "loss": 0.3295,
+      "step": 11277
+    },
+    {
+      "epoch": 2.836519114688129,
+      "grad_norm": 0.28667935729026794,
+      "learning_rate": 9.018875973836493e-08,
+      "loss": 0.3453,
+      "step": 11278
+    },
+    {
+      "epoch": 2.8367706237424546,
+      "grad_norm": 0.29696646332740784,
+      "learning_rate": 8.991230339641554e-08,
+      "loss": 0.286,
+      "step": 11279
+    },
+    {
+      "epoch": 2.837022132796781,
+      "grad_norm": 0.2868374288082123,
+      "learning_rate": 8.963626757276078e-08,
+      "loss": 0.3175,
+      "step": 11280
+    },
+    {
+      "epoch": 2.8372736418511066,
+      "grad_norm": 0.2921113967895508,
+      "learning_rate": 8.936065229104007e-08,
+      "loss": 0.3318,
+      "step": 11281
+    },
+    {
+      "epoch": 2.8375251509054324,
+      "grad_norm": 0.2965297996997833,
+      "learning_rate": 8.908545757485843e-08,
+      "loss": 0.3135,
+      "step": 11282
+    },
+    {
+      "epoch": 2.8377766599597587,
+      "grad_norm": 0.2960835099220276,
+      "learning_rate": 8.881068344778476e-08,
+      "loss": 0.3107,
+      "step": 11283
+    },
+    {
+      "epoch": 2.8380281690140845,
+      "grad_norm": 0.27558842301368713,
+      "learning_rate": 8.853632993335248e-08,
+      "loss": 0.3091,
+      "step": 11284
+    },
+    {
+      "epoch": 2.8382796780684103,
+      "grad_norm": 0.28575974702835083,
+      "learning_rate": 8.826239705505668e-08,
+      "loss": 0.3177,
+      "step": 11285
+    },
+    {
+      "epoch": 2.8385311871227366,
+      "grad_norm": 0.2759111225605011,
+      "learning_rate": 8.79888848363597e-08,
+      "loss": 0.3333,
+      "step": 11286
+    },
+    {
+      "epoch": 2.8387826961770624,
+      "grad_norm": 0.2808525264263153,
+      "learning_rate": 8.771579330068447e-08,
+      "loss": 0.3325,
+      "step": 11287
+    },
+    {
+      "epoch": 2.839034205231388,
+      "grad_norm": 0.2843991816043854,
+      "learning_rate": 8.744312247142005e-08,
+      "loss": 0.297,
+      "step": 11288
+    },
+    {
+      "epoch": 2.8392857142857144,
+      "grad_norm": 0.28934431076049805,
+      "learning_rate": 8.717087237192057e-08,
+      "loss": 0.3257,
+      "step": 11289
+    },
+    {
+      "epoch": 2.83953722334004,
+      "grad_norm": 0.3172331750392914,
+      "learning_rate": 8.689904302550011e-08,
+      "loss": 0.3271,
+      "step": 11290
+    },
+    {
+      "epoch": 2.839788732394366,
+      "grad_norm": 0.27618086338043213,
+      "learning_rate": 8.662763445544121e-08,
+      "loss": 0.34,
+      "step": 11291
+    },
+    {
+      "epoch": 2.8400402414486923,
+      "grad_norm": 0.29746612906455994,
+      "learning_rate": 8.635664668498744e-08,
+      "loss": 0.3051,
+      "step": 11292
+    },
+    {
+      "epoch": 2.840291750503018,
+      "grad_norm": 0.3068857789039612,
+      "learning_rate": 8.608607973734695e-08,
+      "loss": 0.3271,
+      "step": 11293
+    },
+    {
+      "epoch": 2.840543259557344,
+      "grad_norm": 0.28964129090309143,
+      "learning_rate": 8.58159336356923e-08,
+      "loss": 0.3228,
+      "step": 11294
+    },
+    {
+      "epoch": 2.84079476861167,
+      "grad_norm": 0.28352829813957214,
+      "learning_rate": 8.554620840315997e-08,
+      "loss": 0.3148,
+      "step": 11295
+    },
+    {
+      "epoch": 2.841046277665996,
+      "grad_norm": 0.3184901475906372,
+      "learning_rate": 8.527690406285039e-08,
+      "loss": 0.3446,
+      "step": 11296
+    },
+    {
+      "epoch": 2.8412977867203217,
+      "grad_norm": 0.2846779525279999,
+      "learning_rate": 8.500802063782732e-08,
+      "loss": 0.3211,
+      "step": 11297
+    },
+    {
+      "epoch": 2.841549295774648,
+      "grad_norm": 0.2861534655094147,
+      "learning_rate": 8.4739558151119e-08,
+      "loss": 0.3138,
+      "step": 11298
+    },
+    {
+      "epoch": 2.841800804828974,
+      "grad_norm": 0.3019251227378845,
+      "learning_rate": 8.447151662571762e-08,
+      "loss": 0.3251,
+      "step": 11299
+    },
+    {
+      "epoch": 2.8420523138832996,
+      "grad_norm": 0.27594611048698425,
+      "learning_rate": 8.420389608458035e-08,
+      "loss": 0.3178,
+      "step": 11300
+    },
+    {
+      "epoch": 2.842303822937626,
+      "grad_norm": 0.28519585728645325,
+      "learning_rate": 8.393669655062553e-08,
+      "loss": 0.3354,
+      "step": 11301
+    },
+    {
+      "epoch": 2.8425553319919517,
+      "grad_norm": 0.31750738620758057,
+      "learning_rate": 8.366991804673818e-08,
+      "loss": 0.3143,
+      "step": 11302
+    },
+    {
+      "epoch": 2.8428068410462775,
+      "grad_norm": 0.3031100332736969,
+      "learning_rate": 8.340356059576505e-08,
+      "loss": 0.3148,
+      "step": 11303
+    },
+    {
+      "epoch": 2.8430583501006037,
+      "grad_norm": 0.2872277498245239,
+      "learning_rate": 8.313762422052008e-08,
+      "loss": 0.321,
+      "step": 11304
+    },
+    {
+      "epoch": 2.8433098591549295,
+      "grad_norm": 0.2903568744659424,
+      "learning_rate": 8.287210894377673e-08,
+      "loss": 0.312,
+      "step": 11305
+    },
+    {
+      "epoch": 2.8435613682092553,
+      "grad_norm": 0.2857918441295624,
+      "learning_rate": 8.260701478827626e-08,
+      "loss": 0.3355,
+      "step": 11306
+    },
+    {
+      "epoch": 2.8438128772635816,
+      "grad_norm": 0.2868918180465698,
+      "learning_rate": 8.234234177672217e-08,
+      "loss": 0.3188,
+      "step": 11307
+    },
+    {
+      "epoch": 2.8440643863179074,
+      "grad_norm": 0.2801552414894104,
+      "learning_rate": 8.207808993178134e-08,
+      "loss": 0.3317,
+      "step": 11308
+    },
+    {
+      "epoch": 2.844315895372233,
+      "grad_norm": 0.28546252846717834,
+      "learning_rate": 8.181425927608566e-08,
+      "loss": 0.3225,
+      "step": 11309
+    },
+    {
+      "epoch": 2.8445674044265594,
+      "grad_norm": 0.3166808485984802,
+      "learning_rate": 8.155084983223038e-08,
+      "loss": 0.3293,
+      "step": 11310
+    },
+    {
+      "epoch": 2.8448189134808852,
+      "grad_norm": 0.287341445684433,
+      "learning_rate": 8.12878616227758e-08,
+      "loss": 0.3176,
+      "step": 11311
+    },
+    {
+      "epoch": 2.845070422535211,
+      "grad_norm": 0.27534937858581543,
+      "learning_rate": 8.102529467024389e-08,
+      "loss": 0.3313,
+      "step": 11312
+    },
+    {
+      "epoch": 2.8453219315895373,
+      "grad_norm": 0.28979775309562683,
+      "learning_rate": 8.076314899712279e-08,
+      "loss": 0.3389,
+      "step": 11313
+    },
+    {
+      "epoch": 2.845573440643863,
+      "grad_norm": 0.2847757339477539,
+      "learning_rate": 8.050142462586285e-08,
+      "loss": 0.3058,
+      "step": 11314
+    },
+    {
+      "epoch": 2.845824949698189,
+      "grad_norm": 0.2654733657836914,
+      "learning_rate": 8.024012157888062e-08,
+      "loss": 0.3408,
+      "step": 11315
+    },
+    {
+      "epoch": 2.846076458752515,
+      "grad_norm": 0.2785297632217407,
+      "learning_rate": 7.997923987855316e-08,
+      "loss": 0.3114,
+      "step": 11316
+    },
+    {
+      "epoch": 2.846327967806841,
+      "grad_norm": 0.27078497409820557,
+      "learning_rate": 7.971877954722429e-08,
+      "loss": 0.305,
+      "step": 11317
+    },
+    {
+      "epoch": 2.8465794768611667,
+      "grad_norm": 0.29301923513412476,
+      "learning_rate": 7.945874060720116e-08,
+      "loss": 0.3411,
+      "step": 11318
+    },
+    {
+      "epoch": 2.846830985915493,
+      "grad_norm": 0.2895153760910034,
+      "learning_rate": 7.919912308075428e-08,
+      "loss": 0.2994,
+      "step": 11319
+    },
+    {
+      "epoch": 2.847082494969819,
+      "grad_norm": 0.3017180263996124,
+      "learning_rate": 7.893992699011754e-08,
+      "loss": 0.3418,
+      "step": 11320
+    },
+    {
+      "epoch": 2.8473340040241446,
+      "grad_norm": 0.3073883652687073,
+      "learning_rate": 7.868115235748986e-08,
+      "loss": 0.3158,
+      "step": 11321
+    },
+    {
+      "epoch": 2.847585513078471,
+      "grad_norm": 0.28526899218559265,
+      "learning_rate": 7.842279920503404e-08,
+      "loss": 0.3253,
+      "step": 11322
+    },
+    {
+      "epoch": 2.8478370221327967,
+      "grad_norm": 0.2872920036315918,
+      "learning_rate": 7.81648675548763e-08,
+      "loss": 0.3285,
+      "step": 11323
+    },
+    {
+      "epoch": 2.8480885311871225,
+      "grad_norm": 0.2851753532886505,
+      "learning_rate": 7.790735742910671e-08,
+      "loss": 0.3204,
+      "step": 11324
+    },
+    {
+      "epoch": 2.8483400402414487,
+      "grad_norm": 0.28143948316574097,
+      "learning_rate": 7.765026884977934e-08,
+      "loss": 0.3232,
+      "step": 11325
+    },
+    {
+      "epoch": 2.8485915492957745,
+      "grad_norm": 0.28405773639678955,
+      "learning_rate": 7.739360183891265e-08,
+      "loss": 0.3175,
+      "step": 11326
+    },
+    {
+      "epoch": 2.8488430583501008,
+      "grad_norm": 0.3014451265335083,
+      "learning_rate": 7.713735641848796e-08,
+      "loss": 0.3247,
+      "step": 11327
+    },
+    {
+      "epoch": 2.8490945674044266,
+      "grad_norm": 0.2807544469833374,
+      "learning_rate": 7.688153261045161e-08,
+      "loss": 0.3086,
+      "step": 11328
+    },
+    {
+      "epoch": 2.8493460764587524,
+      "grad_norm": 0.2870287001132965,
+      "learning_rate": 7.662613043671274e-08,
+      "loss": 0.3077,
+      "step": 11329
+    },
+    {
+      "epoch": 2.8495975855130786,
+      "grad_norm": 0.2920602560043335,
+      "learning_rate": 7.637114991914552e-08,
+      "loss": 0.3255,
+      "step": 11330
+    },
+    {
+      "epoch": 2.8498490945674044,
+      "grad_norm": 0.29726293683052063,
+      "learning_rate": 7.611659107958692e-08,
+      "loss": 0.3174,
+      "step": 11331
+    },
+    {
+      "epoch": 2.8501006036217302,
+      "grad_norm": 0.2930297553539276,
+      "learning_rate": 7.586245393983837e-08,
+      "loss": 0.3149,
+      "step": 11332
+    },
+    {
+      "epoch": 2.8503521126760565,
+      "grad_norm": 0.29084110260009766,
+      "learning_rate": 7.560873852166584e-08,
+      "loss": 0.3169,
+      "step": 11333
+    },
+    {
+      "epoch": 2.8506036217303823,
+      "grad_norm": 0.29585593938827515,
+      "learning_rate": 7.535544484679747e-08,
+      "loss": 0.3074,
+      "step": 11334
+    },
+    {
+      "epoch": 2.850855130784708,
+      "grad_norm": 0.29283884167671204,
+      "learning_rate": 7.510257293692702e-08,
+      "loss": 0.3137,
+      "step": 11335
+    },
+    {
+      "epoch": 2.8511066398390343,
+      "grad_norm": 0.31102699041366577,
+      "learning_rate": 7.485012281371107e-08,
+      "loss": 0.3049,
+      "step": 11336
+    },
+    {
+      "epoch": 2.85135814889336,
+      "grad_norm": 0.3003372848033905,
+      "learning_rate": 7.45980944987712e-08,
+      "loss": 0.3204,
+      "step": 11337
+    },
+    {
+      "epoch": 2.8516096579476864,
+      "grad_norm": 0.2767249643802643,
+      "learning_rate": 7.434648801369015e-08,
+      "loss": 0.3144,
+      "step": 11338
+    },
+    {
+      "epoch": 2.851861167002012,
+      "grad_norm": 0.2756442725658417,
+      "learning_rate": 7.409530338001846e-08,
+      "loss": 0.3237,
+      "step": 11339
+    },
+    {
+      "epoch": 2.852112676056338,
+      "grad_norm": 0.29652678966522217,
+      "learning_rate": 7.384454061926727e-08,
+      "loss": 0.3038,
+      "step": 11340
+    },
+    {
+      "epoch": 2.8523641851106643,
+      "grad_norm": 0.266886830329895,
+      "learning_rate": 7.359419975291326e-08,
+      "loss": 0.3405,
+      "step": 11341
+    },
+    {
+      "epoch": 2.85261569416499,
+      "grad_norm": 0.284145325422287,
+      "learning_rate": 7.334428080239653e-08,
+      "loss": 0.3099,
+      "step": 11342
+    },
+    {
+      "epoch": 2.852867203219316,
+      "grad_norm": 0.2697547674179077,
+      "learning_rate": 7.309478378912105e-08,
+      "loss": 0.309,
+      "step": 11343
+    },
+    {
+      "epoch": 2.853118712273642,
+      "grad_norm": 0.29870113730430603,
+      "learning_rate": 7.28457087344553e-08,
+      "loss": 0.3209,
+      "step": 11344
+    },
+    {
+      "epoch": 2.853370221327968,
+      "grad_norm": 0.2862483561038971,
+      "learning_rate": 7.259705565972941e-08,
+      "loss": 0.3286,
+      "step": 11345
+    },
+    {
+      "epoch": 2.8536217303822937,
+      "grad_norm": 0.29490405321121216,
+      "learning_rate": 7.234882458624081e-08,
+      "loss": 0.3179,
+      "step": 11346
+    },
+    {
+      "epoch": 2.85387323943662,
+      "grad_norm": 0.2868093252182007,
+      "learning_rate": 7.210101553524751e-08,
+      "loss": 0.3436,
+      "step": 11347
+    },
+    {
+      "epoch": 2.8541247484909458,
+      "grad_norm": 0.2712860703468323,
+      "learning_rate": 7.185362852797417e-08,
+      "loss": 0.331,
+      "step": 11348
+    },
+    {
+      "epoch": 2.8543762575452716,
+      "grad_norm": 0.2829849421977997,
+      "learning_rate": 7.160666358560664e-08,
+      "loss": 0.32,
+      "step": 11349
+    },
+    {
+      "epoch": 2.854627766599598,
+      "grad_norm": 0.28862541913986206,
+      "learning_rate": 7.136012072929632e-08,
+      "loss": 0.3101,
+      "step": 11350
+    },
+    {
+      "epoch": 2.8548792756539236,
+      "grad_norm": 0.27747079730033875,
+      "learning_rate": 7.111399998015856e-08,
+      "loss": 0.3098,
+      "step": 11351
+    },
+    {
+      "epoch": 2.8551307847082494,
+      "grad_norm": 0.2933838963508606,
+      "learning_rate": 7.086830135927148e-08,
+      "loss": 0.3195,
+      "step": 11352
+    },
+    {
+      "epoch": 2.8553822937625757,
+      "grad_norm": 0.2843649685382843,
+      "learning_rate": 7.06230248876777e-08,
+      "loss": 0.3176,
+      "step": 11353
+    },
+    {
+      "epoch": 2.8556338028169015,
+      "grad_norm": 0.2857533395290375,
+      "learning_rate": 7.037817058638375e-08,
+      "loss": 0.2968,
+      "step": 11354
+    },
+    {
+      "epoch": 2.8558853118712273,
+      "grad_norm": 0.27660998702049255,
+      "learning_rate": 7.013373847636007e-08,
+      "loss": 0.3313,
+      "step": 11355
+    },
+    {
+      "epoch": 2.8561368209255535,
+      "grad_norm": 0.28614291548728943,
+      "learning_rate": 6.988972857853993e-08,
+      "loss": 0.3525,
+      "step": 11356
+    },
+    {
+      "epoch": 2.8563883299798793,
+      "grad_norm": 0.2815568447113037,
+      "learning_rate": 6.964614091382272e-08,
+      "loss": 0.318,
+      "step": 11357
+    },
+    {
+      "epoch": 2.856639839034205,
+      "grad_norm": 0.2796972393989563,
+      "learning_rate": 6.940297550306895e-08,
+      "loss": 0.3181,
+      "step": 11358
+    },
+    {
+      "epoch": 2.8568913480885314,
+      "grad_norm": 0.2727015018463135,
+      "learning_rate": 6.916023236710478e-08,
+      "loss": 0.339,
+      "step": 11359
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.2959345281124115,
+      "learning_rate": 6.891791152671912e-08,
+      "loss": 0.3253,
+      "step": 11360
+    },
+    {
+      "epoch": 2.857394366197183,
+      "grad_norm": 0.29117295145988464,
+      "learning_rate": 6.867601300266647e-08,
+      "loss": 0.3153,
+      "step": 11361
+    },
+    {
+      "epoch": 2.8576458752515093,
+      "grad_norm": 0.31557777523994446,
+      "learning_rate": 6.843453681566192e-08,
+      "loss": 0.3248,
+      "step": 11362
+    },
+    {
+      "epoch": 2.857897384305835,
+      "grad_norm": 0.28794246912002563,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.3368,
+      "step": 11363
+    },
+    {
+      "epoch": 2.858148893360161,
+      "grad_norm": 0.2903740108013153,
+      "learning_rate": 6.79528515354888e-08,
+      "loss": 0.3016,
+      "step": 11364
+    },
+    {
+      "epoch": 2.858400402414487,
+      "grad_norm": 0.2773711681365967,
+      "learning_rate": 6.771264248357389e-08,
+      "loss": 0.3324,
+      "step": 11365
+    },
+    {
+      "epoch": 2.858651911468813,
+      "grad_norm": 0.3121846318244934,
+      "learning_rate": 6.747285585121388e-08,
+      "loss": 0.3131,
+      "step": 11366
+    },
+    {
+      "epoch": 2.8589034205231387,
+      "grad_norm": 0.27057939767837524,
+      "learning_rate": 6.723349165894621e-08,
+      "loss": 0.3408,
+      "step": 11367
+    },
+    {
+      "epoch": 2.859154929577465,
+      "grad_norm": 0.3090129494667053,
+      "learning_rate": 6.699454992727061e-08,
+      "loss": 0.3223,
+      "step": 11368
+    },
+    {
+      "epoch": 2.859406438631791,
+      "grad_norm": 0.2833699882030487,
+      "learning_rate": 6.675603067665182e-08,
+      "loss": 0.3158,
+      "step": 11369
+    },
+    {
+      "epoch": 2.8596579476861166,
+      "grad_norm": 0.29948365688323975,
+      "learning_rate": 6.651793392751571e-08,
+      "loss": 0.325,
+      "step": 11370
+    },
+    {
+      "epoch": 2.859909456740443,
+      "grad_norm": 0.27591371536254883,
+      "learning_rate": 6.628025970025542e-08,
+      "loss": 0.3463,
+      "step": 11371
+    },
+    {
+      "epoch": 2.8601609657947686,
+      "grad_norm": 0.3021274209022522,
+      "learning_rate": 6.604300801522523e-08,
+      "loss": 0.2923,
+      "step": 11372
+    },
+    {
+      "epoch": 2.8604124748490944,
+      "grad_norm": 0.26336097717285156,
+      "learning_rate": 6.580617889274498e-08,
+      "loss": 0.3211,
+      "step": 11373
+    },
+    {
+      "epoch": 2.8606639839034207,
+      "grad_norm": 0.26855531334877014,
+      "learning_rate": 6.55697723530968e-08,
+      "loss": 0.3167,
+      "step": 11374
+    },
+    {
+      "epoch": 2.8609154929577465,
+      "grad_norm": 0.28841209411621094,
+      "learning_rate": 6.533378841652893e-08,
+      "loss": 0.3344,
+      "step": 11375
+    },
+    {
+      "epoch": 2.8611670020120723,
+      "grad_norm": 0.28485679626464844,
+      "learning_rate": 6.509822710324964e-08,
+      "loss": 0.3074,
+      "step": 11376
+    },
+    {
+      "epoch": 2.8614185110663986,
+      "grad_norm": 0.3013753294944763,
+      "learning_rate": 6.486308843343558e-08,
+      "loss": 0.3302,
+      "step": 11377
+    },
+    {
+      "epoch": 2.8616700201207244,
+      "grad_norm": 0.29902198910713196,
+      "learning_rate": 6.462837242722342e-08,
+      "loss": 0.341,
+      "step": 11378
+    },
+    {
+      "epoch": 2.86192152917505,
+      "grad_norm": 0.29432061314582825,
+      "learning_rate": 6.43940791047154e-08,
+      "loss": 0.3172,
+      "step": 11379
+    },
+    {
+      "epoch": 2.8621730382293764,
+      "grad_norm": 0.28921979665756226,
+      "learning_rate": 6.41602084859777e-08,
+      "loss": 0.3307,
+      "step": 11380
+    },
+    {
+      "epoch": 2.862424547283702,
+      "grad_norm": 0.3158515989780426,
+      "learning_rate": 6.392676059103986e-08,
+      "loss": 0.3208,
+      "step": 11381
+    },
+    {
+      "epoch": 2.862676056338028,
+      "grad_norm": 0.27476269006729126,
+      "learning_rate": 6.36937354398942e-08,
+      "loss": 0.3132,
+      "step": 11382
+    },
+    {
+      "epoch": 2.8629275653923543,
+      "grad_norm": 0.2913151681423187,
+      "learning_rate": 6.346113305249923e-08,
+      "loss": 0.2988,
+      "step": 11383
+    },
+    {
+      "epoch": 2.86317907444668,
+      "grad_norm": 0.3120485544204712,
+      "learning_rate": 6.322895344877566e-08,
+      "loss": 0.3129,
+      "step": 11384
+    },
+    {
+      "epoch": 2.863430583501006,
+      "grad_norm": 0.2824515402317047,
+      "learning_rate": 6.299719664860704e-08,
+      "loss": 0.301,
+      "step": 11385
+    },
+    {
+      "epoch": 2.863682092555332,
+      "grad_norm": 0.28883475065231323,
+      "learning_rate": 6.27658626718436e-08,
+      "loss": 0.3207,
+      "step": 11386
+    },
+    {
+      "epoch": 2.863933601609658,
+      "grad_norm": 0.3073241710662842,
+      "learning_rate": 6.253495153829614e-08,
+      "loss": 0.3431,
+      "step": 11387
+    },
+    {
+      "epoch": 2.8641851106639837,
+      "grad_norm": 0.2900136709213257,
+      "learning_rate": 6.230446326774165e-08,
+      "loss": 0.3236,
+      "step": 11388
+    },
+    {
+      "epoch": 2.86443661971831,
+      "grad_norm": 0.3047274351119995,
+      "learning_rate": 6.207439787991986e-08,
+      "loss": 0.3098,
+      "step": 11389
+    },
+    {
+      "epoch": 2.864688128772636,
+      "grad_norm": 0.2926124930381775,
+      "learning_rate": 6.184475539453394e-08,
+      "loss": 0.3156,
+      "step": 11390
+    },
+    {
+      "epoch": 2.8649396378269616,
+      "grad_norm": 0.290855348110199,
+      "learning_rate": 6.161553583125202e-08,
+      "loss": 0.3187,
+      "step": 11391
+    },
+    {
+      "epoch": 2.865191146881288,
+      "grad_norm": 0.30072471499443054,
+      "learning_rate": 6.138673920970505e-08,
+      "loss": 0.3089,
+      "step": 11392
+    },
+    {
+      "epoch": 2.8654426559356136,
+      "grad_norm": 0.2740122675895691,
+      "learning_rate": 6.115836554948795e-08,
+      "loss": 0.3369,
+      "step": 11393
+    },
+    {
+      "epoch": 2.8656941649899395,
+      "grad_norm": 0.2824988067150116,
+      "learning_rate": 6.09304148701595e-08,
+      "loss": 0.3327,
+      "step": 11394
+    },
+    {
+      "epoch": 2.8659456740442657,
+      "grad_norm": 0.28075578808784485,
+      "learning_rate": 6.07028871912424e-08,
+      "loss": 0.3295,
+      "step": 11395
+    },
+    {
+      "epoch": 2.8661971830985915,
+      "grad_norm": 0.32692331075668335,
+      "learning_rate": 6.047578253222274e-08,
+      "loss": 0.3303,
+      "step": 11396
+    },
+    {
+      "epoch": 2.8664486921529173,
+      "grad_norm": 0.27784934639930725,
+      "learning_rate": 6.024910091255054e-08,
+      "loss": 0.3183,
+      "step": 11397
+    },
+    {
+      "epoch": 2.8667002012072436,
+      "grad_norm": 0.30311983823776245,
+      "learning_rate": 6.002284235164024e-08,
+      "loss": 0.3179,
+      "step": 11398
+    },
+    {
+      "epoch": 2.8669517102615694,
+      "grad_norm": 0.2793339490890503,
+      "learning_rate": 5.979700686886914e-08,
+      "loss": 0.3189,
+      "step": 11399
+    },
+    {
+      "epoch": 2.867203219315895,
+      "grad_norm": 0.3019621968269348,
+      "learning_rate": 5.9571594483577855e-08,
+      "loss": 0.3127,
+      "step": 11400
+    },
+    {
+      "epoch": 2.8674547283702214,
+      "grad_norm": 0.28011688590049744,
+      "learning_rate": 5.9346605215073185e-08,
+      "loss": 0.3086,
+      "step": 11401
+    },
+    {
+      "epoch": 2.8677062374245472,
+      "grad_norm": 0.31461960077285767,
+      "learning_rate": 5.9122039082622486e-08,
+      "loss": 0.3178,
+      "step": 11402
+    },
+    {
+      "epoch": 2.867957746478873,
+      "grad_norm": 0.27287060022354126,
+      "learning_rate": 5.889789610545982e-08,
+      "loss": 0.3329,
+      "step": 11403
+    },
+    {
+      "epoch": 2.8682092555331993,
+      "grad_norm": 0.28381481766700745,
+      "learning_rate": 5.867417630277983e-08,
+      "loss": 0.3345,
+      "step": 11404
+    },
+    {
+      "epoch": 2.868460764587525,
+      "grad_norm": 0.29278185963630676,
+      "learning_rate": 5.845087969374497e-08,
+      "loss": 0.3284,
+      "step": 11405
+    },
+    {
+      "epoch": 2.868712273641851,
+      "grad_norm": 0.27680209279060364,
+      "learning_rate": 5.8228006297477156e-08,
+      "loss": 0.3217,
+      "step": 11406
+    },
+    {
+      "epoch": 2.868963782696177,
+      "grad_norm": 0.2894262373447418,
+      "learning_rate": 5.8005556133065575e-08,
+      "loss": 0.2931,
+      "step": 11407
+    },
+    {
+      "epoch": 2.869215291750503,
+      "grad_norm": 0.26620855927467346,
+      "learning_rate": 5.7783529219560544e-08,
+      "loss": 0.3274,
+      "step": 11408
+    },
+    {
+      "epoch": 2.8694668008048287,
+      "grad_norm": 0.2928333878517151,
+      "learning_rate": 5.756192557597795e-08,
+      "loss": 0.314,
+      "step": 11409
+    },
+    {
+      "epoch": 2.869718309859155,
+      "grad_norm": 0.2882349193096161,
+      "learning_rate": 5.734074522129707e-08,
+      "loss": 0.34,
+      "step": 11410
+    },
+    {
+      "epoch": 2.869969818913481,
+      "grad_norm": 0.2803259789943695,
+      "learning_rate": 5.711998817445996e-08,
+      "loss": 0.3263,
+      "step": 11411
+    },
+    {
+      "epoch": 2.8702213279678066,
+      "grad_norm": 0.3050937354564667,
+      "learning_rate": 5.689965445437318e-08,
+      "loss": 0.3102,
+      "step": 11412
+    },
+    {
+      "epoch": 2.870472837022133,
+      "grad_norm": 0.2841440439224243,
+      "learning_rate": 5.6679744079907176e-08,
+      "loss": 0.3356,
+      "step": 11413
+    },
+    {
+      "epoch": 2.8707243460764587,
+      "grad_norm": 0.28401076793670654,
+      "learning_rate": 5.646025706989577e-08,
+      "loss": 0.3222,
+      "step": 11414
+    },
+    {
+      "epoch": 2.8709758551307845,
+      "grad_norm": 0.3013293147087097,
+      "learning_rate": 5.624119344313672e-08,
+      "loss": 0.3261,
+      "step": 11415
+    },
+    {
+      "epoch": 2.8712273641851107,
+      "grad_norm": 0.2837771773338318,
+      "learning_rate": 5.6022553218391674e-08,
+      "loss": 0.313,
+      "step": 11416
+    },
+    {
+      "epoch": 2.8714788732394365,
+      "grad_norm": 0.28893938660621643,
+      "learning_rate": 5.580433641438454e-08,
+      "loss": 0.3093,
+      "step": 11417
+    },
+    {
+      "epoch": 2.8717303822937623,
+      "grad_norm": 0.2678500711917877,
+      "learning_rate": 5.558654304980593e-08,
+      "loss": 0.3037,
+      "step": 11418
+    },
+    {
+      "epoch": 2.8719818913480886,
+      "grad_norm": 0.29921528697013855,
+      "learning_rate": 5.536917314330759e-08,
+      "loss": 0.3199,
+      "step": 11419
+    },
+    {
+      "epoch": 2.8722334004024144,
+      "grad_norm": 0.2949788570404053,
+      "learning_rate": 5.5152226713506285e-08,
+      "loss": 0.311,
+      "step": 11420
+    },
+    {
+      "epoch": 2.87248490945674,
+      "grad_norm": 0.31055039167404175,
+      "learning_rate": 5.493570377898161e-08,
+      "loss": 0.3221,
+      "step": 11421
+    },
+    {
+      "epoch": 2.8727364185110664,
+      "grad_norm": 0.28630343079566956,
+      "learning_rate": 5.4719604358277615e-08,
+      "loss": 0.33,
+      "step": 11422
+    },
+    {
+      "epoch": 2.8729879275653922,
+      "grad_norm": 0.30379021167755127,
+      "learning_rate": 5.450392846990227e-08,
+      "loss": 0.3115,
+      "step": 11423
+    },
+    {
+      "epoch": 2.873239436619718,
+      "grad_norm": 0.2785795032978058,
+      "learning_rate": 5.428867613232636e-08,
+      "loss": 0.2873,
+      "step": 11424
+    },
+    {
+      "epoch": 2.8734909456740443,
+      "grad_norm": 0.271603524684906,
+      "learning_rate": 5.407384736398513e-08,
+      "loss": 0.3243,
+      "step": 11425
+    },
+    {
+      "epoch": 2.87374245472837,
+      "grad_norm": 0.2870146930217743,
+      "learning_rate": 5.385944218327721e-08,
+      "loss": 0.3185,
+      "step": 11426
+    },
+    {
+      "epoch": 2.8739939637826963,
+      "grad_norm": 0.2926969528198242,
+      "learning_rate": 5.3645460608565124e-08,
+      "loss": 0.301,
+      "step": 11427
+    },
+    {
+      "epoch": 2.874245472837022,
+      "grad_norm": 0.2959466278553009,
+      "learning_rate": 5.3431902658174776e-08,
+      "loss": 0.3264,
+      "step": 11428
+    },
+    {
+      "epoch": 2.874496981891348,
+      "grad_norm": 0.3116055130958557,
+      "learning_rate": 5.321876835039652e-08,
+      "loss": 0.2992,
+      "step": 11429
+    },
+    {
+      "epoch": 2.874748490945674,
+      "grad_norm": 0.28724607825279236,
+      "learning_rate": 5.30060577034841e-08,
+      "loss": 0.3454,
+      "step": 11430
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 0.29540908336639404,
+      "learning_rate": 5.279377073565406e-08,
+      "loss": 0.2933,
+      "step": 11431
+    },
+    {
+      "epoch": 2.875251509054326,
+      "grad_norm": 0.3000968098640442,
+      "learning_rate": 5.258190746508796e-08,
+      "loss": 0.3071,
+      "step": 11432
+    },
+    {
+      "epoch": 2.875503018108652,
+      "grad_norm": 0.29470768570899963,
+      "learning_rate": 5.237046790993072e-08,
+      "loss": 0.324,
+      "step": 11433
+    },
+    {
+      "epoch": 2.875754527162978,
+      "grad_norm": 0.29645785689353943,
+      "learning_rate": 5.2159452088290654e-08,
+      "loss": 0.3147,
+      "step": 11434
+    },
+    {
+      "epoch": 2.8760060362173037,
+      "grad_norm": 0.27087098360061646,
+      "learning_rate": 5.1948860018239954e-08,
+      "loss": 0.3253,
+      "step": 11435
+    },
+    {
+      "epoch": 2.87625754527163,
+      "grad_norm": 0.30059030652046204,
+      "learning_rate": 5.17386917178142e-08,
+      "loss": 0.3133,
+      "step": 11436
+    },
+    {
+      "epoch": 2.8765090543259557,
+      "grad_norm": 0.27557048201560974,
+      "learning_rate": 5.152894720501345e-08,
+      "loss": 0.312,
+      "step": 11437
+    },
+    {
+      "epoch": 2.876760563380282,
+      "grad_norm": 0.2941986322402954,
+      "learning_rate": 5.131962649780109e-08,
+      "loss": 0.3058,
+      "step": 11438
+    },
+    {
+      "epoch": 2.8770120724346078,
+      "grad_norm": 0.2895359694957733,
+      "learning_rate": 5.111072961410335e-08,
+      "loss": 0.3238,
+      "step": 11439
+    },
+    {
+      "epoch": 2.8772635814889336,
+      "grad_norm": 0.3026026785373688,
+      "learning_rate": 5.090225657181203e-08,
+      "loss": 0.3418,
+      "step": 11440
+    },
+    {
+      "epoch": 2.87751509054326,
+      "grad_norm": 0.29785895347595215,
+      "learning_rate": 5.069420738878061e-08,
+      "loss": 0.3185,
+      "step": 11441
+    },
+    {
+      "epoch": 2.8777665995975856,
+      "grad_norm": 0.3039131164550781,
+      "learning_rate": 5.048658208282764e-08,
+      "loss": 0.3096,
+      "step": 11442
+    },
+    {
+      "epoch": 2.8780181086519114,
+      "grad_norm": 0.29077792167663574,
+      "learning_rate": 5.027938067173499e-08,
+      "loss": 0.3351,
+      "step": 11443
+    },
+    {
+      "epoch": 2.8782696177062377,
+      "grad_norm": 0.2855316698551178,
+      "learning_rate": 5.0072603173247914e-08,
+      "loss": 0.3141,
+      "step": 11444
+    },
+    {
+      "epoch": 2.8785211267605635,
+      "grad_norm": 0.2709047496318817,
+      "learning_rate": 4.986624960507558e-08,
+      "loss": 0.3076,
+      "step": 11445
+    },
+    {
+      "epoch": 2.8787726358148893,
+      "grad_norm": 0.27553707361221313,
+      "learning_rate": 4.966031998489107e-08,
+      "loss": 0.3167,
+      "step": 11446
+    },
+    {
+      "epoch": 2.8790241448692155,
+      "grad_norm": 0.29868531227111816,
+      "learning_rate": 4.945481433033139e-08,
+      "loss": 0.3118,
+      "step": 11447
+    },
+    {
+      "epoch": 2.8792756539235413,
+      "grad_norm": 0.28663668036460876,
+      "learning_rate": 4.92497326589958e-08,
+      "loss": 0.3408,
+      "step": 11448
+    },
+    {
+      "epoch": 2.879527162977867,
+      "grad_norm": 0.2877984046936035,
+      "learning_rate": 4.9045074988449125e-08,
+      "loss": 0.2728,
+      "step": 11449
+    },
+    {
+      "epoch": 2.8797786720321934,
+      "grad_norm": 0.2943393290042877,
+      "learning_rate": 4.884084133621847e-08,
+      "loss": 0.3304,
+      "step": 11450
+    },
+    {
+      "epoch": 2.880030181086519,
+      "grad_norm": 0.28473910689353943,
+      "learning_rate": 4.863703171979539e-08,
+      "loss": 0.3062,
+      "step": 11451
+    },
+    {
+      "epoch": 2.880281690140845,
+      "grad_norm": 0.2870289981365204,
+      "learning_rate": 4.843364615663537e-08,
+      "loss": 0.3047,
+      "step": 11452
+    },
+    {
+      "epoch": 2.8805331991951713,
+      "grad_norm": 0.2811976671218872,
+      "learning_rate": 4.823068466415615e-08,
+      "loss": 0.3187,
+      "step": 11453
+    },
+    {
+      "epoch": 2.880784708249497,
+      "grad_norm": 0.2888941466808319,
+      "learning_rate": 4.802814725974048e-08,
+      "loss": 0.3171,
+      "step": 11454
+    },
+    {
+      "epoch": 2.881036217303823,
+      "grad_norm": 0.2838338613510132,
+      "learning_rate": 4.782603396073504e-08,
+      "loss": 0.3013,
+      "step": 11455
+    },
+    {
+      "epoch": 2.881287726358149,
+      "grad_norm": 0.29402175545692444,
+      "learning_rate": 4.7624344784448774e-08,
+      "loss": 0.3276,
+      "step": 11456
+    },
+    {
+      "epoch": 2.881539235412475,
+      "grad_norm": 0.28359174728393555,
+      "learning_rate": 4.742307974815563e-08,
+      "loss": 0.3127,
+      "step": 11457
+    },
+    {
+      "epoch": 2.8817907444668007,
+      "grad_norm": 0.3000732660293579,
+      "learning_rate": 4.7222238869092386e-08,
+      "loss": 0.3278,
+      "step": 11458
+    },
+    {
+      "epoch": 2.882042253521127,
+      "grad_norm": 0.2840462327003479,
+      "learning_rate": 4.702182216445972e-08,
+      "loss": 0.3335,
+      "step": 11459
+    },
+    {
+      "epoch": 2.8822937625754528,
+      "grad_norm": 0.3007800281047821,
+      "learning_rate": 4.682182965142279e-08,
+      "loss": 0.3129,
+      "step": 11460
+    },
+    {
+      "epoch": 2.8825452716297786,
+      "grad_norm": 0.291432648897171,
+      "learning_rate": 4.6622261347108456e-08,
+      "loss": 0.3206,
+      "step": 11461
+    },
+    {
+      "epoch": 2.882796780684105,
+      "grad_norm": 0.2831004858016968,
+      "learning_rate": 4.6423117268609704e-08,
+      "loss": 0.3142,
+      "step": 11462
+    },
+    {
+      "epoch": 2.8830482897384306,
+      "grad_norm": 0.29433655738830566,
+      "learning_rate": 4.622439743298124e-08,
+      "loss": 0.3303,
+      "step": 11463
+    },
+    {
+      "epoch": 2.8832997987927564,
+      "grad_norm": 0.28008735179901123,
+      "learning_rate": 4.6026101857242765e-08,
+      "loss": 0.3134,
+      "step": 11464
+    },
+    {
+      "epoch": 2.8835513078470827,
+      "grad_norm": 0.26906129717826843,
+      "learning_rate": 4.582823055837626e-08,
+      "loss": 0.3519,
+      "step": 11465
+    },
+    {
+      "epoch": 2.8838028169014085,
+      "grad_norm": 0.27299466729164124,
+      "learning_rate": 4.563078355332873e-08,
+      "loss": 0.3303,
+      "step": 11466
+    },
+    {
+      "epoch": 2.8840543259557343,
+      "grad_norm": 0.29059502482414246,
+      "learning_rate": 4.543376085901052e-08,
+      "loss": 0.3287,
+      "step": 11467
+    },
+    {
+      "epoch": 2.8843058350100605,
+      "grad_norm": 0.28941747546195984,
+      "learning_rate": 4.523716249229426e-08,
+      "loss": 0.3179,
+      "step": 11468
+    },
+    {
+      "epoch": 2.8845573440643864,
+      "grad_norm": 0.2766706645488739,
+      "learning_rate": 4.504098847001925e-08,
+      "loss": 0.307,
+      "step": 11469
+    },
+    {
+      "epoch": 2.884808853118712,
+      "grad_norm": 0.28143224120140076,
+      "learning_rate": 4.484523880898428e-08,
+      "loss": 0.3123,
+      "step": 11470
+    },
+    {
+      "epoch": 2.8850603621730384,
+      "grad_norm": 0.30837175250053406,
+      "learning_rate": 4.464991352595593e-08,
+      "loss": 0.2922,
+      "step": 11471
+    },
+    {
+      "epoch": 2.885311871227364,
+      "grad_norm": 0.26661986112594604,
+      "learning_rate": 4.445501263766194e-08,
+      "loss": 0.3325,
+      "step": 11472
+    },
+    {
+      "epoch": 2.88556338028169,
+      "grad_norm": 0.2966688275337219,
+      "learning_rate": 4.426053616079395e-08,
+      "loss": 0.3129,
+      "step": 11473
+    },
+    {
+      "epoch": 2.8858148893360163,
+      "grad_norm": 0.2732764184474945,
+      "learning_rate": 4.406648411200809e-08,
+      "loss": 0.2998,
+      "step": 11474
+    },
+    {
+      "epoch": 2.886066398390342,
+      "grad_norm": 0.3051884174346924,
+      "learning_rate": 4.3872856507923835e-08,
+      "loss": 0.3071,
+      "step": 11475
+    },
+    {
+      "epoch": 2.886317907444668,
+      "grad_norm": 0.2977435290813446,
+      "learning_rate": 4.367965336512403e-08,
+      "loss": 0.3247,
+      "step": 11476
+    },
+    {
+      "epoch": 2.886569416498994,
+      "grad_norm": 0.27795737981796265,
+      "learning_rate": 4.348687470015489e-08,
+      "loss": 0.3144,
+      "step": 11477
+    },
+    {
+      "epoch": 2.88682092555332,
+      "grad_norm": 0.29412925243377686,
+      "learning_rate": 4.329452052952765e-08,
+      "loss": 0.3354,
+      "step": 11478
+    },
+    {
+      "epoch": 2.8870724346076457,
+      "grad_norm": 0.30797427892684937,
+      "learning_rate": 4.3102590869715246e-08,
+      "loss": 0.3209,
+      "step": 11479
+    },
+    {
+      "epoch": 2.887323943661972,
+      "grad_norm": 0.3010835349559784,
+      "learning_rate": 4.291108573715563e-08,
+      "loss": 0.3181,
+      "step": 11480
+    },
+    {
+      "epoch": 2.887575452716298,
+      "grad_norm": 0.27650344371795654,
+      "learning_rate": 4.272000514825014e-08,
+      "loss": 0.3232,
+      "step": 11481
+    },
+    {
+      "epoch": 2.8878269617706236,
+      "grad_norm": 0.284249871969223,
+      "learning_rate": 4.2529349119364014e-08,
+      "loss": 0.3218,
+      "step": 11482
+    },
+    {
+      "epoch": 2.88807847082495,
+      "grad_norm": 0.27574625611305237,
+      "learning_rate": 4.233911766682475e-08,
+      "loss": 0.3078,
+      "step": 11483
+    },
+    {
+      "epoch": 2.8883299798792756,
+      "grad_norm": 0.31248676776885986,
+      "learning_rate": 4.214931080692486e-08,
+      "loss": 0.3008,
+      "step": 11484
+    },
+    {
+      "epoch": 2.8885814889336014,
+      "grad_norm": 0.2914111316204071,
+      "learning_rate": 4.195992855592079e-08,
+      "loss": 0.3229,
+      "step": 11485
+    },
+    {
+      "epoch": 2.8888329979879277,
+      "grad_norm": 0.31353798508644104,
+      "learning_rate": 4.177097093003124e-08,
+      "loss": 0.3246,
+      "step": 11486
+    },
+    {
+      "epoch": 2.8890845070422535,
+      "grad_norm": 0.2765747308731079,
+      "learning_rate": 4.158243794543992e-08,
+      "loss": 0.3021,
+      "step": 11487
+    },
+    {
+      "epoch": 2.8893360160965793,
+      "grad_norm": 0.2932724058628082,
+      "learning_rate": 4.1394329618292265e-08,
+      "loss": 0.3156,
+      "step": 11488
+    },
+    {
+      "epoch": 2.8895875251509056,
+      "grad_norm": 0.2794465720653534,
+      "learning_rate": 4.120664596469981e-08,
+      "loss": 0.3164,
+      "step": 11489
+    },
+    {
+      "epoch": 2.8898390342052314,
+      "grad_norm": 0.2837032973766327,
+      "learning_rate": 4.101938700073582e-08,
+      "loss": 0.3202,
+      "step": 11490
+    },
+    {
+      "epoch": 2.890090543259557,
+      "grad_norm": 0.27074161171913147,
+      "learning_rate": 4.083255274243858e-08,
+      "loss": 0.3186,
+      "step": 11491
+    },
+    {
+      "epoch": 2.8903420523138834,
+      "grad_norm": 0.290571391582489,
+      "learning_rate": 4.0646143205808063e-08,
+      "loss": 0.3222,
+      "step": 11492
+    },
+    {
+      "epoch": 2.890593561368209,
+      "grad_norm": 0.2796350419521332,
+      "learning_rate": 4.046015840680984e-08,
+      "loss": 0.3166,
+      "step": 11493
+    },
+    {
+      "epoch": 2.890845070422535,
+      "grad_norm": 0.30049675703048706,
+      "learning_rate": 4.0274598361372266e-08,
+      "loss": 0.3055,
+      "step": 11494
+    },
+    {
+      "epoch": 2.8910965794768613,
+      "grad_norm": 0.3056521415710449,
+      "learning_rate": 4.008946308538764e-08,
+      "loss": 0.3097,
+      "step": 11495
+    },
+    {
+      "epoch": 2.891348088531187,
+      "grad_norm": 0.2760073244571686,
+      "learning_rate": 3.990475259471105e-08,
+      "loss": 0.3126,
+      "step": 11496
+    },
+    {
+      "epoch": 2.891599597585513,
+      "grad_norm": 0.2556244134902954,
+      "learning_rate": 3.9720466905162625e-08,
+      "loss": 0.3084,
+      "step": 11497
+    },
+    {
+      "epoch": 2.891851106639839,
+      "grad_norm": 0.2833138704299927,
+      "learning_rate": 3.953660603252474e-08,
+      "loss": 0.3218,
+      "step": 11498
+    },
+    {
+      "epoch": 2.892102615694165,
+      "grad_norm": 0.27805668115615845,
+      "learning_rate": 3.9353169992543684e-08,
+      "loss": 0.3332,
+      "step": 11499
+    },
+    {
+      "epoch": 2.8923541247484907,
+      "grad_norm": 0.30059728026390076,
+      "learning_rate": 3.917015880092967e-08,
+      "loss": 0.3212,
+      "step": 11500
+    },
+    {
+      "epoch": 2.892605633802817,
+      "grad_norm": 0.27901148796081543,
+      "learning_rate": 3.898757247335738e-08,
+      "loss": 0.3197,
+      "step": 11501
+    },
+    {
+      "epoch": 2.892857142857143,
+      "grad_norm": 0.29673993587493896,
+      "learning_rate": 3.8805411025463204e-08,
+      "loss": 0.3502,
+      "step": 11502
+    },
+    {
+      "epoch": 2.8931086519114686,
+      "grad_norm": 0.2697283923625946,
+      "learning_rate": 3.862367447284854e-08,
+      "loss": 0.2868,
+      "step": 11503
+    },
+    {
+      "epoch": 2.893360160965795,
+      "grad_norm": 0.27969419956207275,
+      "learning_rate": 3.8442362831077603e-08,
+      "loss": 0.3168,
+      "step": 11504
+    },
+    {
+      "epoch": 2.8936116700201207,
+      "grad_norm": 0.29848575592041016,
+      "learning_rate": 3.82614761156791e-08,
+      "loss": 0.3243,
+      "step": 11505
+    },
+    {
+      "epoch": 2.8938631790744465,
+      "grad_norm": 0.2888948619365692,
+      "learning_rate": 3.8081014342144506e-08,
+      "loss": 0.3113,
+      "step": 11506
+    },
+    {
+      "epoch": 2.8941146881287727,
+      "grad_norm": 0.2911056876182556,
+      "learning_rate": 3.790097752592925e-08,
+      "loss": 0.3319,
+      "step": 11507
+    },
+    {
+      "epoch": 2.8943661971830985,
+      "grad_norm": 0.2891317903995514,
+      "learning_rate": 3.772136568245266e-08,
+      "loss": 0.3201,
+      "step": 11508
+    },
+    {
+      "epoch": 2.8946177062374243,
+      "grad_norm": 0.3018838167190552,
+      "learning_rate": 3.754217882709743e-08,
+      "loss": 0.2983,
+      "step": 11509
+    },
+    {
+      "epoch": 2.8948692152917506,
+      "grad_norm": 0.30178946256637573,
+      "learning_rate": 3.7363416975209065e-08,
+      "loss": 0.3322,
+      "step": 11510
+    },
+    {
+      "epoch": 2.8951207243460764,
+      "grad_norm": 0.28581055998802185,
+      "learning_rate": 3.718508014209809e-08,
+      "loss": 0.3176,
+      "step": 11511
+    },
+    {
+      "epoch": 2.895372233400402,
+      "grad_norm": 0.3123716711997986,
+      "learning_rate": 3.700716834303786e-08,
+      "loss": 0.328,
+      "step": 11512
+    },
+    {
+      "epoch": 2.8956237424547284,
+      "grad_norm": 0.2836924195289612,
+      "learning_rate": 3.682968159326505e-08,
+      "loss": 0.3073,
+      "step": 11513
+    },
+    {
+      "epoch": 2.8958752515090542,
+      "grad_norm": 0.2993314266204834,
+      "learning_rate": 3.665261990798086e-08,
+      "loss": 0.327,
+      "step": 11514
+    },
+    {
+      "epoch": 2.89612676056338,
+      "grad_norm": 0.2848893105983734,
+      "learning_rate": 3.6475983302348695e-08,
+      "loss": 0.3163,
+      "step": 11515
+    },
+    {
+      "epoch": 2.8963782696177063,
+      "grad_norm": 0.30268558859825134,
+      "learning_rate": 3.629977179149702e-08,
+      "loss": 0.3199,
+      "step": 11516
+    },
+    {
+      "epoch": 2.896629778672032,
+      "grad_norm": 0.28077784180641174,
+      "learning_rate": 3.6123985390517094e-08,
+      "loss": 0.3212,
+      "step": 11517
+    },
+    {
+      "epoch": 2.896881287726358,
+      "grad_norm": 0.31762558221817017,
+      "learning_rate": 3.5948624114464094e-08,
+      "loss": 0.3415,
+      "step": 11518
+    },
+    {
+      "epoch": 2.897132796780684,
+      "grad_norm": 0.267600953578949,
+      "learning_rate": 3.577368797835601e-08,
+      "loss": 0.3174,
+      "step": 11519
+    },
+    {
+      "epoch": 2.89738430583501,
+      "grad_norm": 0.29275646805763245,
+      "learning_rate": 3.5599176997175853e-08,
+      "loss": 0.308,
+      "step": 11520
+    },
+    {
+      "epoch": 2.8976358148893357,
+      "grad_norm": 0.29874688386917114,
+      "learning_rate": 3.54250911858689e-08,
+      "loss": 0.2987,
+      "step": 11521
+    },
+    {
+      "epoch": 2.897887323943662,
+      "grad_norm": 0.29817038774490356,
+      "learning_rate": 3.525143055934488e-08,
+      "loss": 0.3041,
+      "step": 11522
+    },
+    {
+      "epoch": 2.898138832997988,
+      "grad_norm": 0.28878530859947205,
+      "learning_rate": 3.507819513247579e-08,
+      "loss": 0.3076,
+      "step": 11523
+    },
+    {
+      "epoch": 2.8983903420523136,
+      "grad_norm": 0.2917855679988861,
+      "learning_rate": 3.4905384920099204e-08,
+      "loss": 0.3071,
+      "step": 11524
+    },
+    {
+      "epoch": 2.89864185110664,
+      "grad_norm": 0.28223884105682373,
+      "learning_rate": 3.473299993701496e-08,
+      "loss": 0.3096,
+      "step": 11525
+    },
+    {
+      "epoch": 2.8988933601609657,
+      "grad_norm": 0.2846827805042267,
+      "learning_rate": 3.4561040197986785e-08,
+      "loss": 0.3293,
+      "step": 11526
+    },
+    {
+      "epoch": 2.899144869215292,
+      "grad_norm": 0.31608250737190247,
+      "learning_rate": 3.4389505717741246e-08,
+      "loss": 0.3281,
+      "step": 11527
+    },
+    {
+      "epoch": 2.8993963782696177,
+      "grad_norm": 0.29066160321235657,
+      "learning_rate": 3.421839651096992e-08,
+      "loss": 0.3134,
+      "step": 11528
+    },
+    {
+      "epoch": 2.8996478873239435,
+      "grad_norm": 0.28401198983192444,
+      "learning_rate": 3.4047712592327753e-08,
+      "loss": 0.3228,
+      "step": 11529
+    },
+    {
+      "epoch": 2.8998993963782698,
+      "grad_norm": 0.27312031388282776,
+      "learning_rate": 3.3877453976431386e-08,
+      "loss": 0.3142,
+      "step": 11530
+    },
+    {
+      "epoch": 2.9001509054325956,
+      "grad_norm": 0.27587342262268066,
+      "learning_rate": 3.37076206778636e-08,
+      "loss": 0.3202,
+      "step": 11531
+    },
+    {
+      "epoch": 2.9004024144869214,
+      "grad_norm": 0.2901318073272705,
+      "learning_rate": 3.353821271116886e-08,
+      "loss": 0.3086,
+      "step": 11532
+    },
+    {
+      "epoch": 2.9006539235412476,
+      "grad_norm": 0.28231751918792725,
+      "learning_rate": 3.336923009085613e-08,
+      "loss": 0.3257,
+      "step": 11533
+    },
+    {
+      "epoch": 2.9009054325955734,
+      "grad_norm": 0.27121472358703613,
+      "learning_rate": 3.320067283139772e-08,
+      "loss": 0.3256,
+      "step": 11534
+    },
+    {
+      "epoch": 2.9011569416498992,
+      "grad_norm": 0.2887016832828522,
+      "learning_rate": 3.3032540947229296e-08,
+      "loss": 0.3162,
+      "step": 11535
+    },
+    {
+      "epoch": 2.9014084507042255,
+      "grad_norm": 0.2676198184490204,
+      "learning_rate": 3.286483445275046e-08,
+      "loss": 0.3181,
+      "step": 11536
+    },
+    {
+      "epoch": 2.9016599597585513,
+      "grad_norm": 0.2852703630924225,
+      "learning_rate": 3.269755336232472e-08,
+      "loss": 0.2963,
+      "step": 11537
+    },
+    {
+      "epoch": 2.9019114688128775,
+      "grad_norm": 0.2976604402065277,
+      "learning_rate": 3.2530697690277835e-08,
+      "loss": 0.348,
+      "step": 11538
+    },
+    {
+      "epoch": 2.9021629778672033,
+      "grad_norm": 0.27679896354675293,
+      "learning_rate": 3.236426745090004e-08,
+      "loss": 0.3239,
+      "step": 11539
+    },
+    {
+      "epoch": 2.902414486921529,
+      "grad_norm": 0.2862226068973541,
+      "learning_rate": 3.219826265844606e-08,
+      "loss": 0.3037,
+      "step": 11540
+    },
+    {
+      "epoch": 2.9026659959758554,
+      "grad_norm": 0.27300092577934265,
+      "learning_rate": 3.203268332713172e-08,
+      "loss": 0.328,
+      "step": 11541
+    },
+    {
+      "epoch": 2.902917505030181,
+      "grad_norm": 0.2864827513694763,
+      "learning_rate": 3.1867529471139025e-08,
+      "loss": 0.3227,
+      "step": 11542
+    },
+    {
+      "epoch": 2.903169014084507,
+      "grad_norm": 0.2821025848388672,
+      "learning_rate": 3.1702801104611655e-08,
+      "loss": 0.312,
+      "step": 11543
+    },
+    {
+      "epoch": 2.9034205231388333,
+      "grad_norm": 0.27062085270881653,
+      "learning_rate": 3.153849824165778e-08,
+      "loss": 0.2998,
+      "step": 11544
+    },
+    {
+      "epoch": 2.903672032193159,
+      "grad_norm": 0.2897660434246063,
+      "learning_rate": 3.1374620896348905e-08,
+      "loss": 0.3393,
+      "step": 11545
+    },
+    {
+      "epoch": 2.903923541247485,
+      "grad_norm": 0.3015936017036438,
+      "learning_rate": 3.121116908272048e-08,
+      "loss": 0.315,
+      "step": 11546
+    },
+    {
+      "epoch": 2.904175050301811,
+      "grad_norm": 0.28271710872650146,
+      "learning_rate": 3.10481428147702e-08,
+      "loss": 0.3177,
+      "step": 11547
+    },
+    {
+      "epoch": 2.904426559356137,
+      "grad_norm": 0.2824450135231018,
+      "learning_rate": 3.088554210646133e-08,
+      "loss": 0.3422,
+      "step": 11548
+    },
+    {
+      "epoch": 2.9046780684104627,
+      "grad_norm": 0.289212703704834,
+      "learning_rate": 3.072336697171885e-08,
+      "loss": 0.3493,
+      "step": 11549
+    },
+    {
+      "epoch": 2.904929577464789,
+      "grad_norm": 0.29587748646736145,
+      "learning_rate": 3.05616174244322e-08,
+      "loss": 0.2915,
+      "step": 11550
+    },
+    {
+      "epoch": 2.9051810865191148,
+      "grad_norm": 0.26594677567481995,
+      "learning_rate": 3.040029347845419e-08,
+      "loss": 0.3406,
+      "step": 11551
+    },
+    {
+      "epoch": 2.9054325955734406,
+      "grad_norm": 0.28409308195114136,
+      "learning_rate": 3.0239395147601547e-08,
+      "loss": 0.3331,
+      "step": 11552
+    },
+    {
+      "epoch": 2.905684104627767,
+      "grad_norm": 0.29760369658470154,
+      "learning_rate": 3.007892244565436e-08,
+      "loss": 0.3406,
+      "step": 11553
+    },
+    {
+      "epoch": 2.9059356136820926,
+      "grad_norm": 0.276475191116333,
+      "learning_rate": 2.991887538635496e-08,
+      "loss": 0.3105,
+      "step": 11554
+    },
+    {
+      "epoch": 2.9061871227364184,
+      "grad_norm": 0.29626041650772095,
+      "learning_rate": 2.975925398341184e-08,
+      "loss": 0.3304,
+      "step": 11555
+    },
+    {
+      "epoch": 2.9064386317907447,
+      "grad_norm": 0.29270699620246887,
+      "learning_rate": 2.960005825049461e-08,
+      "loss": 0.342,
+      "step": 11556
+    },
+    {
+      "epoch": 2.9066901408450705,
+      "grad_norm": 0.2765859365463257,
+      "learning_rate": 2.944128820123737e-08,
+      "loss": 0.326,
+      "step": 11557
+    },
+    {
+      "epoch": 2.9069416498993963,
+      "grad_norm": 0.2837342917919159,
+      "learning_rate": 2.9282943849238687e-08,
+      "loss": 0.3028,
+      "step": 11558
+    },
+    {
+      "epoch": 2.9071931589537225,
+      "grad_norm": 0.3145996630191803,
+      "learning_rate": 2.9125025208058823e-08,
+      "loss": 0.3317,
+      "step": 11559
+    },
+    {
+      "epoch": 2.9074446680080483,
+      "grad_norm": 0.28416410088539124,
+      "learning_rate": 2.8967532291222512e-08,
+      "loss": 0.3253,
+      "step": 11560
+    },
+    {
+      "epoch": 2.907696177062374,
+      "grad_norm": 0.2724005877971649,
+      "learning_rate": 2.8810465112218965e-08,
+      "loss": 0.3177,
+      "step": 11561
+    },
+    {
+      "epoch": 2.9079476861167004,
+      "grad_norm": 0.2829074561595917,
+      "learning_rate": 2.8653823684499093e-08,
+      "loss": 0.3286,
+      "step": 11562
+    },
+    {
+      "epoch": 2.908199195171026,
+      "grad_norm": 0.2702285349369049,
+      "learning_rate": 2.849760802147883e-08,
+      "loss": 0.3161,
+      "step": 11563
+    },
+    {
+      "epoch": 2.908450704225352,
+      "grad_norm": 0.2694869637489319,
+      "learning_rate": 2.8341818136536915e-08,
+      "loss": 0.3177,
+      "step": 11564
+    },
+    {
+      "epoch": 2.9087022132796783,
+      "grad_norm": 0.302658349275589,
+      "learning_rate": 2.8186454043014898e-08,
+      "loss": 0.3182,
+      "step": 11565
+    },
+    {
+      "epoch": 2.908953722334004,
+      "grad_norm": 0.2830633521080017,
+      "learning_rate": 2.8031515754220473e-08,
+      "loss": 0.3262,
+      "step": 11566
+    },
+    {
+      "epoch": 2.90920523138833,
+      "grad_norm": 0.25605013966560364,
+      "learning_rate": 2.7877003283421356e-08,
+      "loss": 0.3138,
+      "step": 11567
+    },
+    {
+      "epoch": 2.909456740442656,
+      "grad_norm": 0.30849385261535645,
+      "learning_rate": 2.772291664385196e-08,
+      "loss": 0.3408,
+      "step": 11568
+    },
+    {
+      "epoch": 2.909708249496982,
+      "grad_norm": 0.2905694246292114,
+      "learning_rate": 2.7569255848708397e-08,
+      "loss": 0.332,
+      "step": 11569
+    },
+    {
+      "epoch": 2.9099597585513077,
+      "grad_norm": 0.31673821806907654,
+      "learning_rate": 2.7416020911150144e-08,
+      "loss": 0.3283,
+      "step": 11570
+    },
+    {
+      "epoch": 2.910211267605634,
+      "grad_norm": 0.28510385751724243,
+      "learning_rate": 2.7263211844301695e-08,
+      "loss": 0.3045,
+      "step": 11571
+    },
+    {
+      "epoch": 2.91046277665996,
+      "grad_norm": 0.2822716534137726,
+      "learning_rate": 2.7110828661249255e-08,
+      "loss": 0.3522,
+      "step": 11572
+    },
+    {
+      "epoch": 2.9107142857142856,
+      "grad_norm": 0.3001321256160736,
+      "learning_rate": 2.6958871375044605e-08,
+      "loss": 0.3117,
+      "step": 11573
+    },
+    {
+      "epoch": 2.910965794768612,
+      "grad_norm": 0.2811189889907837,
+      "learning_rate": 2.680733999870122e-08,
+      "loss": 0.3131,
+      "step": 11574
+    },
+    {
+      "epoch": 2.9112173038229376,
+      "grad_norm": 0.3301267921924591,
+      "learning_rate": 2.6656234545197057e-08,
+      "loss": 0.3362,
+      "step": 11575
+    },
+    {
+      "epoch": 2.9114688128772634,
+      "grad_norm": 0.2918173372745514,
+      "learning_rate": 2.6505555027472875e-08,
+      "loss": 0.3034,
+      "step": 11576
+    },
+    {
+      "epoch": 2.9117203219315897,
+      "grad_norm": 0.29059723019599915,
+      "learning_rate": 2.6355301458434457e-08,
+      "loss": 0.3262,
+      "step": 11577
+    },
+    {
+      "epoch": 2.9119718309859155,
+      "grad_norm": 0.29665839672088623,
+      "learning_rate": 2.620547385094929e-08,
+      "loss": 0.3288,
+      "step": 11578
+    },
+    {
+      "epoch": 2.9122233400402413,
+      "grad_norm": 0.275056391954422,
+      "learning_rate": 2.6056072217848783e-08,
+      "loss": 0.3151,
+      "step": 11579
+    },
+    {
+      "epoch": 2.9124748490945676,
+      "grad_norm": 0.2742392122745514,
+      "learning_rate": 2.5907096571929357e-08,
+      "loss": 0.3128,
+      "step": 11580
+    },
+    {
+      "epoch": 2.9127263581488934,
+      "grad_norm": 0.2982200086116791,
+      "learning_rate": 2.5758546925949148e-08,
+      "loss": 0.3303,
+      "step": 11581
+    },
+    {
+      "epoch": 2.912977867203219,
+      "grad_norm": 0.284332275390625,
+      "learning_rate": 2.5610423292630195e-08,
+      "loss": 0.335,
+      "step": 11582
+    },
+    {
+      "epoch": 2.9132293762575454,
+      "grad_norm": 0.289644718170166,
+      "learning_rate": 2.5462725684659573e-08,
+      "loss": 0.3368,
+      "step": 11583
+    },
+    {
+      "epoch": 2.913480885311871,
+      "grad_norm": 0.27438437938690186,
+      "learning_rate": 2.53154541146855e-08,
+      "loss": 0.3279,
+      "step": 11584
+    },
+    {
+      "epoch": 2.913732394366197,
+      "grad_norm": 0.2902214527130127,
+      "learning_rate": 2.516860859532122e-08,
+      "loss": 0.3175,
+      "step": 11585
+    },
+    {
+      "epoch": 2.9139839034205233,
+      "grad_norm": 0.28719666600227356,
+      "learning_rate": 2.5022189139143338e-08,
+      "loss": 0.3358,
+      "step": 11586
+    },
+    {
+      "epoch": 2.914235412474849,
+      "grad_norm": 0.298453688621521,
+      "learning_rate": 2.487619575869127e-08,
+      "loss": 0.3301,
+      "step": 11587
+    },
+    {
+      "epoch": 2.914486921529175,
+      "grad_norm": 0.2861069142818451,
+      "learning_rate": 2.4730628466468898e-08,
+      "loss": 0.288,
+      "step": 11588
+    },
+    {
+      "epoch": 2.914738430583501,
+      "grad_norm": 0.3006594479084015,
+      "learning_rate": 2.4585487274942922e-08,
+      "loss": 0.3185,
+      "step": 11589
+    },
+    {
+      "epoch": 2.914989939637827,
+      "grad_norm": 0.2652274966239929,
+      "learning_rate": 2.444077219654395e-08,
+      "loss": 0.3174,
+      "step": 11590
+    },
+    {
+      "epoch": 2.9152414486921527,
+      "grad_norm": 0.29787126183509827,
+      "learning_rate": 2.4296483243665958e-08,
+      "loss": 0.3119,
+      "step": 11591
+    },
+    {
+      "epoch": 2.915492957746479,
+      "grad_norm": 0.26815494894981384,
+      "learning_rate": 2.4152620428666284e-08,
+      "loss": 0.3128,
+      "step": 11592
+    },
+    {
+      "epoch": 2.915744466800805,
+      "grad_norm": 0.28812792897224426,
+      "learning_rate": 2.400918376386563e-08,
+      "loss": 0.3161,
+      "step": 11593
+    },
+    {
+      "epoch": 2.9159959758551306,
+      "grad_norm": 0.27825263142585754,
+      "learning_rate": 2.386617326154861e-08,
+      "loss": 0.2885,
+      "step": 11594
+    },
+    {
+      "epoch": 2.916247484909457,
+      "grad_norm": 0.30951130390167236,
+      "learning_rate": 2.372358893396376e-08,
+      "loss": 0.3131,
+      "step": 11595
+    },
+    {
+      "epoch": 2.9164989939637826,
+      "grad_norm": 0.3053039610385895,
+      "learning_rate": 2.358143079332187e-08,
+      "loss": 0.3242,
+      "step": 11596
+    },
+    {
+      "epoch": 2.9167505030181085,
+      "grad_norm": 0.2770836651325226,
+      "learning_rate": 2.3439698851797643e-08,
+      "loss": 0.33,
+      "step": 11597
+    },
+    {
+      "epoch": 2.9170020120724347,
+      "grad_norm": 0.29645803570747375,
+      "learning_rate": 2.329839312153026e-08,
+      "loss": 0.3375,
+      "step": 11598
+    },
+    {
+      "epoch": 2.9172535211267605,
+      "grad_norm": 0.26674768328666687,
+      "learning_rate": 2.3157513614621706e-08,
+      "loss": 0.3247,
+      "step": 11599
+    },
+    {
+      "epoch": 2.9175050301810863,
+      "grad_norm": 0.2916862666606903,
+      "learning_rate": 2.3017060343136223e-08,
+      "loss": 0.322,
+      "step": 11600
+    },
+    {
+      "epoch": 2.9177565392354126,
+      "grad_norm": 0.2905628979206085,
+      "learning_rate": 2.287703331910418e-08,
+      "loss": 0.2903,
+      "step": 11601
+    },
+    {
+      "epoch": 2.9180080482897384,
+      "grad_norm": 0.28413960337638855,
+      "learning_rate": 2.273743255451766e-08,
+      "loss": 0.3085,
+      "step": 11602
+    },
+    {
+      "epoch": 2.918259557344064,
+      "grad_norm": 0.290414035320282,
+      "learning_rate": 2.2598258061331536e-08,
+      "loss": 0.3394,
+      "step": 11603
+    },
+    {
+      "epoch": 2.9185110663983904,
+      "grad_norm": 0.2848317623138428,
+      "learning_rate": 2.2459509851466833e-08,
+      "loss": 0.3058,
+      "step": 11604
+    },
+    {
+      "epoch": 2.9187625754527162,
+      "grad_norm": 0.2922884225845337,
+      "learning_rate": 2.232118793680571e-08,
+      "loss": 0.3092,
+      "step": 11605
+    },
+    {
+      "epoch": 2.919014084507042,
+      "grad_norm": 0.2839473485946655,
+      "learning_rate": 2.218329232919425e-08,
+      "loss": 0.3318,
+      "step": 11606
+    },
+    {
+      "epoch": 2.9192655935613683,
+      "grad_norm": 0.28309670090675354,
+      "learning_rate": 2.2045823040443005e-08,
+      "loss": 0.31,
+      "step": 11607
+    },
+    {
+      "epoch": 2.919517102615694,
+      "grad_norm": 0.2878527045249939,
+      "learning_rate": 2.1908780082324777e-08,
+      "loss": 0.3089,
+      "step": 11608
+    },
+    {
+      "epoch": 2.91976861167002,
+      "grad_norm": 0.3030252158641815,
+      "learning_rate": 2.177216346657629e-08,
+      "loss": 0.2988,
+      "step": 11609
+    },
+    {
+      "epoch": 2.920020120724346,
+      "grad_norm": 0.2812774181365967,
+      "learning_rate": 2.1635973204899296e-08,
+      "loss": 0.3232,
+      "step": 11610
+    },
+    {
+      "epoch": 2.920271629778672,
+      "grad_norm": 0.2975054383277893,
+      "learning_rate": 2.1500209308956132e-08,
+      "loss": 0.3178,
+      "step": 11611
+    },
+    {
+      "epoch": 2.9205231388329977,
+      "grad_norm": 0.262926310300827,
+      "learning_rate": 2.136487179037472e-08,
+      "loss": 0.3276,
+      "step": 11612
+    },
+    {
+      "epoch": 2.920774647887324,
+      "grad_norm": 0.30193185806274414,
+      "learning_rate": 2.122996066074523e-08,
+      "loss": 0.3339,
+      "step": 11613
+    },
+    {
+      "epoch": 2.92102615694165,
+      "grad_norm": 0.28565213084220886,
+      "learning_rate": 2.1095475931623422e-08,
+      "loss": 0.3043,
+      "step": 11614
+    },
+    {
+      "epoch": 2.9212776659959756,
+      "grad_norm": 0.32027721405029297,
+      "learning_rate": 2.0961417614525638e-08,
+      "loss": 0.2899,
+      "step": 11615
+    },
+    {
+      "epoch": 2.921529175050302,
+      "grad_norm": 0.29877468943595886,
+      "learning_rate": 2.0827785720933803e-08,
+      "loss": 0.3295,
+      "step": 11616
+    },
+    {
+      "epoch": 2.9217806841046277,
+      "grad_norm": 0.291421502828598,
+      "learning_rate": 2.0694580262292096e-08,
+      "loss": 0.3109,
+      "step": 11617
+    },
+    {
+      "epoch": 2.9220321931589535,
+      "grad_norm": 0.2898477613925934,
+      "learning_rate": 2.0561801250009727e-08,
+      "loss": 0.3352,
+      "step": 11618
+    },
+    {
+      "epoch": 2.9222837022132797,
+      "grad_norm": 0.27962225675582886,
+      "learning_rate": 2.04294486954576e-08,
+      "loss": 0.329,
+      "step": 11619
+    },
+    {
+      "epoch": 2.9225352112676055,
+      "grad_norm": 0.31196466088294983,
+      "learning_rate": 2.0297522609971087e-08,
+      "loss": 0.3335,
+      "step": 11620
+    },
+    {
+      "epoch": 2.9227867203219313,
+      "grad_norm": 0.2565595507621765,
+      "learning_rate": 2.0166023004848934e-08,
+      "loss": 0.2932,
+      "step": 11621
+    },
+    {
+      "epoch": 2.9230382293762576,
+      "grad_norm": 0.2930534780025482,
+      "learning_rate": 2.003494989135324e-08,
+      "loss": 0.329,
+      "step": 11622
+    },
+    {
+      "epoch": 2.9232897384305834,
+      "grad_norm": 0.2888554632663727,
+      "learning_rate": 1.990430328070947e-08,
+      "loss": 0.3193,
+      "step": 11623
+    },
+    {
+      "epoch": 2.9235412474849096,
+      "grad_norm": 0.27373868227005005,
+      "learning_rate": 1.977408318410645e-08,
+      "loss": 0.3213,
+      "step": 11624
+    },
+    {
+      "epoch": 2.9237927565392354,
+      "grad_norm": 0.2846567630767822,
+      "learning_rate": 1.9644289612697487e-08,
+      "loss": 0.3173,
+      "step": 11625
+    },
+    {
+      "epoch": 2.9240442655935612,
+      "grad_norm": 0.25684884190559387,
+      "learning_rate": 1.951492257759757e-08,
+      "loss": 0.3325,
+      "step": 11626
+    },
+    {
+      "epoch": 2.9242957746478875,
+      "grad_norm": 0.2954108417034149,
+      "learning_rate": 1.938598208988729e-08,
+      "loss": 0.3258,
+      "step": 11627
+    },
+    {
+      "epoch": 2.9245472837022133,
+      "grad_norm": 0.28941529989242554,
+      "learning_rate": 1.9257468160608917e-08,
+      "loss": 0.3436,
+      "step": 11628
+    },
+    {
+      "epoch": 2.924798792756539,
+      "grad_norm": 0.310966432094574,
+      "learning_rate": 1.912938080076865e-08,
+      "loss": 0.3186,
+      "step": 11629
+    },
+    {
+      "epoch": 2.9250503018108653,
+      "grad_norm": 0.29559624195098877,
+      "learning_rate": 1.90017200213366e-08,
+      "loss": 0.3081,
+      "step": 11630
+    },
+    {
+      "epoch": 2.925301810865191,
+      "grad_norm": 0.2723540663719177,
+      "learning_rate": 1.887448583324625e-08,
+      "loss": 0.3291,
+      "step": 11631
+    },
+    {
+      "epoch": 2.925553319919517,
+      "grad_norm": 0.290840208530426,
+      "learning_rate": 1.8747678247394984e-08,
+      "loss": 0.3344,
+      "step": 11632
+    },
+    {
+      "epoch": 2.925804828973843,
+      "grad_norm": 0.31380322575569153,
+      "learning_rate": 1.8621297274641904e-08,
+      "loss": 0.2994,
+      "step": 11633
+    },
+    {
+      "epoch": 2.926056338028169,
+      "grad_norm": 0.279624342918396,
+      "learning_rate": 1.8495342925811122e-08,
+      "loss": 0.3212,
+      "step": 11634
+    },
+    {
+      "epoch": 2.926307847082495,
+      "grad_norm": 0.27431485056877136,
+      "learning_rate": 1.8369815211690678e-08,
+      "loss": 0.2976,
+      "step": 11635
+    },
+    {
+      "epoch": 2.926559356136821,
+      "grad_norm": 0.27514395117759705,
+      "learning_rate": 1.8244714143029752e-08,
+      "loss": 0.305,
+      "step": 11636
+    },
+    {
+      "epoch": 2.926810865191147,
+      "grad_norm": 0.27842938899993896,
+      "learning_rate": 1.8120039730544214e-08,
+      "loss": 0.3311,
+      "step": 11637
+    },
+    {
+      "epoch": 2.927062374245473,
+      "grad_norm": 0.26175856590270996,
+      "learning_rate": 1.799579198490997e-08,
+      "loss": 0.323,
+      "step": 11638
+    },
+    {
+      "epoch": 2.927313883299799,
+      "grad_norm": 0.3041093647480011,
+      "learning_rate": 1.7871970916769067e-08,
+      "loss": 0.3209,
+      "step": 11639
+    },
+    {
+      "epoch": 2.9275653923541247,
+      "grad_norm": 0.28131407499313354,
+      "learning_rate": 1.774857653672579e-08,
+      "loss": 0.3049,
+      "step": 11640
+    },
+    {
+      "epoch": 2.927816901408451,
+      "grad_norm": 0.3082786500453949,
+      "learning_rate": 1.7625608855348365e-08,
+      "loss": 0.302,
+      "step": 11641
+    },
+    {
+      "epoch": 2.9280684104627768,
+      "grad_norm": 0.28668394684791565,
+      "learning_rate": 1.7503067883167247e-08,
+      "loss": 0.3034,
+      "step": 11642
+    },
+    {
+      "epoch": 2.9283199195171026,
+      "grad_norm": 0.2656381130218506,
+      "learning_rate": 1.7380953630678488e-08,
+      "loss": 0.3191,
+      "step": 11643
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.30544763803482056,
+      "learning_rate": 1.7259266108339833e-08,
+      "loss": 0.3299,
+      "step": 11644
+    },
+    {
+      "epoch": 2.9288229376257546,
+      "grad_norm": 0.2998693287372589,
+      "learning_rate": 1.713800532657295e-08,
+      "loss": 0.3154,
+      "step": 11645
+    },
+    {
+      "epoch": 2.9290744466800804,
+      "grad_norm": 0.2941758334636688,
+      "learning_rate": 1.7017171295763412e-08,
+      "loss": 0.3054,
+      "step": 11646
+    },
+    {
+      "epoch": 2.9293259557344067,
+      "grad_norm": 0.29245221614837646,
+      "learning_rate": 1.6896764026259616e-08,
+      "loss": 0.326,
+      "step": 11647
+    },
+    {
+      "epoch": 2.9295774647887325,
+      "grad_norm": 0.2588028907775879,
+      "learning_rate": 1.6776783528373864e-08,
+      "loss": 0.2949,
+      "step": 11648
+    },
+    {
+      "epoch": 2.9298289738430583,
+      "grad_norm": 0.27226272225379944,
+      "learning_rate": 1.665722981238127e-08,
+      "loss": 0.3068,
+      "step": 11649
+    },
+    {
+      "epoch": 2.9300804828973845,
+      "grad_norm": 0.27903351187705994,
+      "learning_rate": 1.6538102888521423e-08,
+      "loss": 0.3074,
+      "step": 11650
+    },
+    {
+      "epoch": 2.9303319919517103,
+      "grad_norm": 0.2830352187156677,
+      "learning_rate": 1.6419402766996717e-08,
+      "loss": 0.3179,
+      "step": 11651
+    },
+    {
+      "epoch": 2.930583501006036,
+      "grad_norm": 0.29141873121261597,
+      "learning_rate": 1.630112945797291e-08,
+      "loss": 0.3243,
+      "step": 11652
+    },
+    {
+      "epoch": 2.9308350100603624,
+      "grad_norm": 0.28866273164749146,
+      "learning_rate": 1.6183282971579673e-08,
+      "loss": 0.3409,
+      "step": 11653
+    },
+    {
+      "epoch": 2.931086519114688,
+      "grad_norm": 0.27328038215637207,
+      "learning_rate": 1.606586331790949e-08,
+      "loss": 0.3436,
+      "step": 11654
+    },
+    {
+      "epoch": 2.931338028169014,
+      "grad_norm": 0.2887606620788574,
+      "learning_rate": 1.5948870507018766e-08,
+      "loss": 0.3276,
+      "step": 11655
+    },
+    {
+      "epoch": 2.9315895372233403,
+      "grad_norm": 0.2820006012916565,
+      "learning_rate": 1.5832304548926703e-08,
+      "loss": 0.2983,
+      "step": 11656
+    },
+    {
+      "epoch": 2.931841046277666,
+      "grad_norm": 0.2761184573173523,
+      "learning_rate": 1.571616545361754e-08,
+      "loss": 0.3168,
+      "step": 11657
+    },
+    {
+      "epoch": 2.932092555331992,
+      "grad_norm": 0.3011208474636078,
+      "learning_rate": 1.560045323103665e-08,
+      "loss": 0.3198,
+      "step": 11658
+    },
+    {
+      "epoch": 2.932344064386318,
+      "grad_norm": 0.27800092101097107,
+      "learning_rate": 1.5485167891095e-08,
+      "loss": 0.321,
+      "step": 11659
+    },
+    {
+      "epoch": 2.932595573440644,
+      "grad_norm": 0.2734206020832062,
+      "learning_rate": 1.53703094436658e-08,
+      "loss": 0.3157,
+      "step": 11660
+    },
+    {
+      "epoch": 2.9328470824949697,
+      "grad_norm": 0.26894569396972656,
+      "learning_rate": 1.5255877898585624e-08,
+      "loss": 0.3178,
+      "step": 11661
+    },
+    {
+      "epoch": 2.933098591549296,
+      "grad_norm": 0.2924859821796417,
+      "learning_rate": 1.5141873265654973e-08,
+      "loss": 0.3364,
+      "step": 11662
+    },
+    {
+      "epoch": 2.933350100603622,
+      "grad_norm": 0.2777864634990692,
+      "learning_rate": 1.5028295554637695e-08,
+      "loss": 0.3277,
+      "step": 11663
+    },
+    {
+      "epoch": 2.9336016096579476,
+      "grad_norm": 0.3020499348640442,
+      "learning_rate": 1.4915144775261014e-08,
+      "loss": 0.3211,
+      "step": 11664
+    },
+    {
+      "epoch": 2.933853118712274,
+      "grad_norm": 0.2888851761817932,
+      "learning_rate": 1.4802420937216066e-08,
+      "loss": 0.3298,
+      "step": 11665
+    },
+    {
+      "epoch": 2.9341046277665996,
+      "grad_norm": 0.27491968870162964,
+      "learning_rate": 1.4690124050155686e-08,
+      "loss": 0.3193,
+      "step": 11666
+    },
+    {
+      "epoch": 2.9343561368209254,
+      "grad_norm": 0.29931673407554626,
+      "learning_rate": 1.4578254123698844e-08,
+      "loss": 0.299,
+      "step": 11667
+    },
+    {
+      "epoch": 2.9346076458752517,
+      "grad_norm": 0.2893234193325043,
+      "learning_rate": 1.4466811167425655e-08,
+      "loss": 0.3203,
+      "step": 11668
+    },
+    {
+      "epoch": 2.9348591549295775,
+      "grad_norm": 0.2912534475326538,
+      "learning_rate": 1.4355795190880707e-08,
+      "loss": 0.3138,
+      "step": 11669
+    },
+    {
+      "epoch": 2.9351106639839033,
+      "grad_norm": 0.28769242763519287,
+      "learning_rate": 1.4245206203571393e-08,
+      "loss": 0.3178,
+      "step": 11670
+    },
+    {
+      "epoch": 2.9353621730382295,
+      "grad_norm": 0.31816333532333374,
+      "learning_rate": 1.4135044214969585e-08,
+      "loss": 0.3455,
+      "step": 11671
+    },
+    {
+      "epoch": 2.9356136820925554,
+      "grad_norm": 0.3172784149646759,
+      "learning_rate": 1.4025309234510509e-08,
+      "loss": 0.3227,
+      "step": 11672
+    },
+    {
+      "epoch": 2.935865191146881,
+      "grad_norm": 0.31336894631385803,
+      "learning_rate": 1.391600127159054e-08,
+      "loss": 0.3134,
+      "step": 11673
+    },
+    {
+      "epoch": 2.9361167002012074,
+      "grad_norm": 0.2920610010623932,
+      "learning_rate": 1.3807120335572743e-08,
+      "loss": 0.3017,
+      "step": 11674
+    },
+    {
+      "epoch": 2.936368209255533,
+      "grad_norm": 0.27723726630210876,
+      "learning_rate": 1.3698666435781327e-08,
+      "loss": 0.3341,
+      "step": 11675
+    },
+    {
+      "epoch": 2.936619718309859,
+      "grad_norm": 0.28226420283317566,
+      "learning_rate": 1.3590639581504971e-08,
+      "loss": 0.3079,
+      "step": 11676
+    },
+    {
+      "epoch": 2.9368712273641853,
+      "grad_norm": 0.3186614215373993,
+      "learning_rate": 1.3483039781995721e-08,
+      "loss": 0.3145,
+      "step": 11677
+    },
+    {
+      "epoch": 2.937122736418511,
+      "grad_norm": 0.27414506673812866,
+      "learning_rate": 1.3375867046468427e-08,
+      "loss": 0.2952,
+      "step": 11678
+    },
+    {
+      "epoch": 2.937374245472837,
+      "grad_norm": 0.28623631596565247,
+      "learning_rate": 1.3269121384101857e-08,
+      "loss": 0.34,
+      "step": 11679
+    },
+    {
+      "epoch": 2.937625754527163,
+      "grad_norm": 0.29475000500679016,
+      "learning_rate": 1.316280280403759e-08,
+      "loss": 0.3194,
+      "step": 11680
+    },
+    {
+      "epoch": 2.937877263581489,
+      "grad_norm": 0.3122510015964508,
+      "learning_rate": 1.3056911315382226e-08,
+      "loss": 0.3142,
+      "step": 11681
+    },
+    {
+      "epoch": 2.9381287726358147,
+      "grad_norm": 0.29271435737609863,
+      "learning_rate": 1.2951446927204625e-08,
+      "loss": 0.3127,
+      "step": 11682
+    },
+    {
+      "epoch": 2.938380281690141,
+      "grad_norm": 0.3412143886089325,
+      "learning_rate": 1.2846409648535896e-08,
+      "loss": 0.3269,
+      "step": 11683
+    },
+    {
+      "epoch": 2.938631790744467,
+      "grad_norm": 0.2909950613975525,
+      "learning_rate": 1.2741799488373285e-08,
+      "loss": 0.3061,
+      "step": 11684
+    },
+    {
+      "epoch": 2.9388832997987926,
+      "grad_norm": 0.276928573846817,
+      "learning_rate": 1.2637616455675183e-08,
+      "loss": 0.3038,
+      "step": 11685
+    },
+    {
+      "epoch": 2.939134808853119,
+      "grad_norm": 0.2644422948360443,
+      "learning_rate": 1.2533860559363897e-08,
+      "loss": 0.332,
+      "step": 11686
+    },
+    {
+      "epoch": 2.9393863179074446,
+      "grad_norm": 0.28948256373405457,
+      "learning_rate": 1.2430531808326763e-08,
+      "loss": 0.325,
+      "step": 11687
+    },
+    {
+      "epoch": 2.9396378269617705,
+      "grad_norm": 0.2958967387676239,
+      "learning_rate": 1.23276302114117e-08,
+      "loss": 0.3028,
+      "step": 11688
+    },
+    {
+      "epoch": 2.9398893360160967,
+      "grad_norm": 0.2886587977409363,
+      "learning_rate": 1.2225155777432773e-08,
+      "loss": 0.3312,
+      "step": 11689
+    },
+    {
+      "epoch": 2.9401408450704225,
+      "grad_norm": 0.3025323748588562,
+      "learning_rate": 1.2123108515165738e-08,
+      "loss": 0.3017,
+      "step": 11690
+    },
+    {
+      "epoch": 2.9403923541247483,
+      "grad_norm": 0.28794747591018677,
+      "learning_rate": 1.202148843335027e-08,
+      "loss": 0.3094,
+      "step": 11691
+    },
+    {
+      "epoch": 2.9406438631790746,
+      "grad_norm": 0.287518709897995,
+      "learning_rate": 1.1920295540689964e-08,
+      "loss": 0.3325,
+      "step": 11692
+    },
+    {
+      "epoch": 2.9408953722334004,
+      "grad_norm": 0.3010087013244629,
+      "learning_rate": 1.1819529845850664e-08,
+      "loss": 0.3359,
+      "step": 11693
+    },
+    {
+      "epoch": 2.941146881287726,
+      "grad_norm": 0.28999802470207214,
+      "learning_rate": 1.1719191357463245e-08,
+      "loss": 0.3375,
+      "step": 11694
+    },
+    {
+      "epoch": 2.9413983903420524,
+      "grad_norm": 0.2955481708049774,
+      "learning_rate": 1.1619280084119722e-08,
+      "loss": 0.3295,
+      "step": 11695
+    },
+    {
+      "epoch": 2.941649899396378,
+      "grad_norm": 0.28115126490592957,
+      "learning_rate": 1.151979603437825e-08,
+      "loss": 0.3115,
+      "step": 11696
+    },
+    {
+      "epoch": 2.941901408450704,
+      "grad_norm": 0.2838100492954254,
+      "learning_rate": 1.1420739216758125e-08,
+      "loss": 0.3389,
+      "step": 11697
+    },
+    {
+      "epoch": 2.9421529175050303,
+      "grad_norm": 0.291797935962677,
+      "learning_rate": 1.1322109639743117e-08,
+      "loss": 0.3357,
+      "step": 11698
+    },
+    {
+      "epoch": 2.942404426559356,
+      "grad_norm": 0.2834652066230774,
+      "learning_rate": 1.1223907311780358e-08,
+      "loss": 0.303,
+      "step": 11699
+    },
+    {
+      "epoch": 2.942655935613682,
+      "grad_norm": 0.2929726243019104,
+      "learning_rate": 1.1126132241280342e-08,
+      "loss": 0.3303,
+      "step": 11700
+    },
+    {
+      "epoch": 2.942907444668008,
+      "grad_norm": 0.290096253156662,
+      "learning_rate": 1.1028784436616923e-08,
+      "loss": 0.323,
+      "step": 11701
+    },
+    {
+      "epoch": 2.943158953722334,
+      "grad_norm": 0.2806238830089569,
+      "learning_rate": 1.0931863906127327e-08,
+      "loss": 0.3357,
+      "step": 11702
+    },
+    {
+      "epoch": 2.9434104627766597,
+      "grad_norm": 0.2727678716182709,
+      "learning_rate": 1.0835370658111577e-08,
+      "loss": 0.3191,
+      "step": 11703
+    },
+    {
+      "epoch": 2.943661971830986,
+      "grad_norm": 0.3112581670284271,
+      "learning_rate": 1.073930470083473e-08,
+      "loss": 0.314,
+      "step": 11704
+    },
+    {
+      "epoch": 2.943913480885312,
+      "grad_norm": 0.29780250787734985,
+      "learning_rate": 1.0643666042523537e-08,
+      "loss": 0.3292,
+      "step": 11705
+    },
+    {
+      "epoch": 2.9441649899396376,
+      "grad_norm": 0.283771276473999,
+      "learning_rate": 1.054845469136867e-08,
+      "loss": 0.3173,
+      "step": 11706
+    },
+    {
+      "epoch": 2.944416498993964,
+      "grad_norm": 0.2684139311313629,
+      "learning_rate": 1.0453670655525273e-08,
+      "loss": 0.3129,
+      "step": 11707
+    },
+    {
+      "epoch": 2.9446680080482897,
+      "grad_norm": 0.29727786779403687,
+      "learning_rate": 1.0359313943110183e-08,
+      "loss": 0.3189,
+      "step": 11708
+    },
+    {
+      "epoch": 2.9449195171026155,
+      "grad_norm": 0.28367939591407776,
+      "learning_rate": 1.0265384562205272e-08,
+      "loss": 0.3273,
+      "step": 11709
+    },
+    {
+      "epoch": 2.9451710261569417,
+      "grad_norm": 0.28062888979911804,
+      "learning_rate": 1.0171882520853548e-08,
+      "loss": 0.3059,
+      "step": 11710
+    },
+    {
+      "epoch": 2.9454225352112675,
+      "grad_norm": 0.28939807415008545,
+      "learning_rate": 1.0078807827064718e-08,
+      "loss": 0.3318,
+      "step": 11711
+    },
+    {
+      "epoch": 2.9456740442655933,
+      "grad_norm": 0.27794498205184937,
+      "learning_rate": 9.986160488808517e-09,
+      "loss": 0.3177,
+      "step": 11712
+    },
+    {
+      "epoch": 2.9459255533199196,
+      "grad_norm": 0.3011113107204437,
+      "learning_rate": 9.893940514020817e-09,
+      "loss": 0.3293,
+      "step": 11713
+    },
+    {
+      "epoch": 2.9461770623742454,
+      "grad_norm": 0.2947550117969513,
+      "learning_rate": 9.802147910598637e-09,
+      "loss": 0.3097,
+      "step": 11714
+    },
+    {
+      "epoch": 2.946428571428571,
+      "grad_norm": 0.2753145694732666,
+      "learning_rate": 9.71078268640402e-09,
+      "loss": 0.3261,
+      "step": 11715
+    },
+    {
+      "epoch": 2.9466800804828974,
+      "grad_norm": 0.29693037271499634,
+      "learning_rate": 9.619844849261816e-09,
+      "loss": 0.3104,
+      "step": 11716
+    },
+    {
+      "epoch": 2.9469315895372232,
+      "grad_norm": 0.2914050221443176,
+      "learning_rate": 9.529334406960245e-09,
+      "loss": 0.3201,
+      "step": 11717
+    },
+    {
+      "epoch": 2.947183098591549,
+      "grad_norm": 0.3129412829875946,
+      "learning_rate": 9.439251367250879e-09,
+      "loss": 0.3052,
+      "step": 11718
+    },
+    {
+      "epoch": 2.9474346076458753,
+      "grad_norm": 0.26742586493492126,
+      "learning_rate": 9.349595737848105e-09,
+      "loss": 0.3316,
+      "step": 11719
+    },
+    {
+      "epoch": 2.947686116700201,
+      "grad_norm": 0.2630806565284729,
+      "learning_rate": 9.260367526431891e-09,
+      "loss": 0.3085,
+      "step": 11720
+    },
+    {
+      "epoch": 2.947937625754527,
+      "grad_norm": 0.28567931056022644,
+      "learning_rate": 9.17156674064279e-09,
+      "loss": 0.3279,
+      "step": 11721
+    },
+    {
+      "epoch": 2.948189134808853,
+      "grad_norm": 0.2721531391143799,
+      "learning_rate": 9.083193388086941e-09,
+      "loss": 0.3058,
+      "step": 11722
+    },
+    {
+      "epoch": 2.948440643863179,
+      "grad_norm": 0.31495407223701477,
+      "learning_rate": 8.99524747633218e-09,
+      "loss": 0.3336,
+      "step": 11723
+    },
+    {
+      "epoch": 2.948692152917505,
+      "grad_norm": 0.26969239115715027,
+      "learning_rate": 8.907729012910814e-09,
+      "loss": 0.3234,
+      "step": 11724
+    },
+    {
+      "epoch": 2.948943661971831,
+      "grad_norm": 0.3003351092338562,
+      "learning_rate": 8.820638005317961e-09,
+      "loss": 0.3474,
+      "step": 11725
+    },
+    {
+      "epoch": 2.949195171026157,
+      "grad_norm": 0.2833467125892639,
+      "learning_rate": 8.733974461013207e-09,
+      "loss": 0.3048,
+      "step": 11726
+    },
+    {
+      "epoch": 2.949446680080483,
+      "grad_norm": 0.2887994050979614,
+      "learning_rate": 8.647738387418391e-09,
+      "loss": 0.3077,
+      "step": 11727
+    },
+    {
+      "epoch": 2.949698189134809,
+      "grad_norm": 0.2787559926509857,
+      "learning_rate": 8.561929791918722e-09,
+      "loss": 0.3414,
+      "step": 11728
+    },
+    {
+      "epoch": 2.9499496981891347,
+      "grad_norm": 0.2879605293273926,
+      "learning_rate": 8.476548681863316e-09,
+      "loss": 0.3144,
+      "step": 11729
+    },
+    {
+      "epoch": 2.950201207243461,
+      "grad_norm": 0.2889518141746521,
+      "learning_rate": 8.391595064564661e-09,
+      "loss": 0.3077,
+      "step": 11730
+    },
+    {
+      "epoch": 2.9504527162977867,
+      "grad_norm": 0.2559886872768402,
+      "learning_rate": 8.307068947299157e-09,
+      "loss": 0.3284,
+      "step": 11731
+    },
+    {
+      "epoch": 2.9507042253521125,
+      "grad_norm": 0.3104651868343353,
+      "learning_rate": 8.222970337304347e-09,
+      "loss": 0.3091,
+      "step": 11732
+    },
+    {
+      "epoch": 2.9509557344064388,
+      "grad_norm": 0.2882402241230011,
+      "learning_rate": 8.139299241783916e-09,
+      "loss": 0.3232,
+      "step": 11733
+    },
+    {
+      "epoch": 2.9512072434607646,
+      "grad_norm": 0.2959415018558502,
+      "learning_rate": 8.056055667903795e-09,
+      "loss": 0.3191,
+      "step": 11734
+    },
+    {
+      "epoch": 2.9514587525150904,
+      "grad_norm": 0.2943558394908905,
+      "learning_rate": 7.973239622792728e-09,
+      "loss": 0.3241,
+      "step": 11735
+    },
+    {
+      "epoch": 2.9517102615694166,
+      "grad_norm": 0.271892786026001,
+      "learning_rate": 7.89085111354393e-09,
+      "loss": 0.3026,
+      "step": 11736
+    },
+    {
+      "epoch": 2.9519617706237424,
+      "grad_norm": 0.2949487268924713,
+      "learning_rate": 7.808890147213422e-09,
+      "loss": 0.3119,
+      "step": 11737
+    },
+    {
+      "epoch": 2.9522132796780687,
+      "grad_norm": 0.2683142125606537,
+      "learning_rate": 7.727356730820035e-09,
+      "loss": 0.3042,
+      "step": 11738
+    },
+    {
+      "epoch": 2.9524647887323945,
+      "grad_norm": 0.3025915026664734,
+      "learning_rate": 7.646250871347072e-09,
+      "loss": 0.3253,
+      "step": 11739
+    },
+    {
+      "epoch": 2.9527162977867203,
+      "grad_norm": 0.2929423451423645,
+      "learning_rate": 7.565572575740087e-09,
+      "loss": 0.3055,
+      "step": 11740
+    },
+    {
+      "epoch": 2.9529678068410465,
+      "grad_norm": 0.27529823780059814,
+      "learning_rate": 7.485321850910221e-09,
+      "loss": 0.3363,
+      "step": 11741
+    },
+    {
+      "epoch": 2.9532193158953723,
+      "grad_norm": 0.28709059953689575,
+      "learning_rate": 7.405498703728642e-09,
+      "loss": 0.3161,
+      "step": 11742
+    },
+    {
+      "epoch": 2.953470824949698,
+      "grad_norm": 0.28413447737693787,
+      "learning_rate": 7.326103141033214e-09,
+      "loss": 0.2844,
+      "step": 11743
+    },
+    {
+      "epoch": 2.9537223340040244,
+      "grad_norm": 0.2665967643260956,
+      "learning_rate": 7.247135169622388e-09,
+      "loss": 0.2992,
+      "step": 11744
+    },
+    {
+      "epoch": 2.95397384305835,
+      "grad_norm": 0.29413825273513794,
+      "learning_rate": 7.1685947962601976e-09,
+      "loss": 0.3074,
+      "step": 11745
+    },
+    {
+      "epoch": 2.954225352112676,
+      "grad_norm": 0.29388710856437683,
+      "learning_rate": 7.0904820276729294e-09,
+      "loss": 0.311,
+      "step": 11746
+    },
+    {
+      "epoch": 2.9544768611670023,
+      "grad_norm": 0.30368688702583313,
+      "learning_rate": 7.012796870549676e-09,
+      "loss": 0.3278,
+      "step": 11747
+    },
+    {
+      "epoch": 2.954728370221328,
+      "grad_norm": 0.30217698216438293,
+      "learning_rate": 6.935539331545116e-09,
+      "loss": 0.3486,
+      "step": 11748
+    },
+    {
+      "epoch": 2.954979879275654,
+      "grad_norm": 0.30664876103401184,
+      "learning_rate": 6.858709417274512e-09,
+      "loss": 0.3202,
+      "step": 11749
+    },
+    {
+      "epoch": 2.95523138832998,
+      "grad_norm": 0.2738303244113922,
+      "learning_rate": 6.7823071343187106e-09,
+      "loss": 0.3249,
+      "step": 11750
+    },
+    {
+      "epoch": 2.955482897384306,
+      "grad_norm": 0.30232658982276917,
+      "learning_rate": 6.7063324892208125e-09,
+      "loss": 0.3419,
+      "step": 11751
+    },
+    {
+      "epoch": 2.9557344064386317,
+      "grad_norm": 0.2941600978374481,
+      "learning_rate": 6.630785488487834e-09,
+      "loss": 0.3004,
+      "step": 11752
+    },
+    {
+      "epoch": 2.955985915492958,
+      "grad_norm": 0.27397286891937256,
+      "learning_rate": 6.5556661385896005e-09,
+      "loss": 0.3072,
+      "step": 11753
+    },
+    {
+      "epoch": 2.9562374245472838,
+      "grad_norm": 0.2965390980243683,
+      "learning_rate": 6.480974445959298e-09,
+      "loss": 0.345,
+      "step": 11754
+    },
+    {
+      "epoch": 2.9564889336016096,
+      "grad_norm": 0.27063122391700745,
+      "learning_rate": 6.406710416994588e-09,
+      "loss": 0.2997,
+      "step": 11755
+    },
+    {
+      "epoch": 2.956740442655936,
+      "grad_norm": 0.2771882712841034,
+      "learning_rate": 6.3328740580548275e-09,
+      "loss": 0.3317,
+      "step": 11756
+    },
+    {
+      "epoch": 2.9569919517102616,
+      "grad_norm": 0.2760043442249298,
+      "learning_rate": 6.259465375464402e-09,
+      "loss": 0.3197,
+      "step": 11757
+    },
+    {
+      "epoch": 2.9572434607645874,
+      "grad_norm": 0.30299848318099976,
+      "learning_rate": 6.1864843755099495e-09,
+      "loss": 0.3115,
+      "step": 11758
+    },
+    {
+      "epoch": 2.9574949698189137,
+      "grad_norm": 0.2945834994316101,
+      "learning_rate": 6.113931064442025e-09,
+      "loss": 0.3084,
+      "step": 11759
+    },
+    {
+      "epoch": 2.9577464788732395,
+      "grad_norm": 0.2729983329772949,
+      "learning_rate": 6.041805448474547e-09,
+      "loss": 0.3012,
+      "step": 11760
+    },
+    {
+      "epoch": 2.9579979879275653,
+      "grad_norm": 0.29064232110977173,
+      "learning_rate": 5.970107533783687e-09,
+      "loss": 0.3246,
+      "step": 11761
+    },
+    {
+      "epoch": 2.9582494969818915,
+      "grad_norm": 0.29307058453559875,
+      "learning_rate": 5.898837326511197e-09,
+      "loss": 0.3051,
+      "step": 11762
+    },
+    {
+      "epoch": 2.9585010060362174,
+      "grad_norm": 0.2878889739513397,
+      "learning_rate": 5.8279948327599754e-09,
+      "loss": 0.3432,
+      "step": 11763
+    },
+    {
+      "epoch": 2.958752515090543,
+      "grad_norm": 0.272480309009552,
+      "learning_rate": 5.75758005859739e-09,
+      "loss": 0.3215,
+      "step": 11764
+    },
+    {
+      "epoch": 2.9590040241448694,
+      "grad_norm": 0.280304878950119,
+      "learning_rate": 5.6875930100541706e-09,
+      "loss": 0.3251,
+      "step": 11765
+    },
+    {
+      "epoch": 2.959255533199195,
+      "grad_norm": 0.2817912697792053,
+      "learning_rate": 5.618033693124414e-09,
+      "loss": 0.3462,
+      "step": 11766
+    },
+    {
+      "epoch": 2.959507042253521,
+      "grad_norm": 0.2780134677886963,
+      "learning_rate": 5.548902113765575e-09,
+      "loss": 0.299,
+      "step": 11767
+    },
+    {
+      "epoch": 2.9597585513078473,
+      "grad_norm": 0.28865960240364075,
+      "learning_rate": 5.480198277897919e-09,
+      "loss": 0.3421,
+      "step": 11768
+    },
+    {
+      "epoch": 2.960010060362173,
+      "grad_norm": 0.2794874608516693,
+      "learning_rate": 5.411922191405627e-09,
+      "loss": 0.3387,
+      "step": 11769
+    },
+    {
+      "epoch": 2.960261569416499,
+      "grad_norm": 0.27201390266418457,
+      "learning_rate": 5.34407386013569e-09,
+      "loss": 0.3346,
+      "step": 11770
+    },
+    {
+      "epoch": 2.960513078470825,
+      "grad_norm": 0.28588712215423584,
+      "learning_rate": 5.276653289900124e-09,
+      "loss": 0.3482,
+      "step": 11771
+    },
+    {
+      "epoch": 2.960764587525151,
+      "grad_norm": 0.2717914283275604,
+      "learning_rate": 5.20966048647209e-09,
+      "loss": 0.3096,
+      "step": 11772
+    },
+    {
+      "epoch": 2.9610160965794767,
+      "grad_norm": 0.2592178285121918,
+      "learning_rate": 5.14309545558922e-09,
+      "loss": 0.2934,
+      "step": 11773
+    },
+    {
+      "epoch": 2.961267605633803,
+      "grad_norm": 0.2764432728290558,
+      "learning_rate": 5.076958202952509e-09,
+      "loss": 0.3238,
+      "step": 11774
+    },
+    {
+      "epoch": 2.961519114688129,
+      "grad_norm": 0.3098163902759552,
+      "learning_rate": 5.01124873422576e-09,
+      "loss": 0.3151,
+      "step": 11775
+    },
+    {
+      "epoch": 2.9617706237424546,
+      "grad_norm": 0.2991572916507721,
+      "learning_rate": 4.945967055037803e-09,
+      "loss": 0.3173,
+      "step": 11776
+    },
+    {
+      "epoch": 2.962022132796781,
+      "grad_norm": 0.26591500639915466,
+      "learning_rate": 4.881113170978058e-09,
+      "loss": 0.3266,
+      "step": 11777
+    },
+    {
+      "epoch": 2.9622736418511066,
+      "grad_norm": 0.29749444127082825,
+      "learning_rate": 4.8166870876020786e-09,
+      "loss": 0.3228,
+      "step": 11778
+    },
+    {
+      "epoch": 2.9625251509054324,
+      "grad_norm": 0.2841416001319885,
+      "learning_rate": 4.75268881042712e-09,
+      "loss": 0.3343,
+      "step": 11779
+    },
+    {
+      "epoch": 2.9627766599597587,
+      "grad_norm": 0.26449331641197205,
+      "learning_rate": 4.689118344933796e-09,
+      "loss": 0.3179,
+      "step": 11780
+    },
+    {
+      "epoch": 2.9630281690140845,
+      "grad_norm": 0.3073911666870117,
+      "learning_rate": 4.625975696567197e-09,
+      "loss": 0.3033,
+      "step": 11781
+    },
+    {
+      "epoch": 2.9632796780684103,
+      "grad_norm": 0.28680184483528137,
+      "learning_rate": 4.56326087073522e-09,
+      "loss": 0.2853,
+      "step": 11782
+    },
+    {
+      "epoch": 2.9635311871227366,
+      "grad_norm": 0.2971070408821106,
+      "learning_rate": 4.500973872808012e-09,
+      "loss": 0.3466,
+      "step": 11783
+    },
+    {
+      "epoch": 2.9637826961770624,
+      "grad_norm": 0.2824273705482483,
+      "learning_rate": 4.439114708120751e-09,
+      "loss": 0.3272,
+      "step": 11784
+    },
+    {
+      "epoch": 2.964034205231388,
+      "grad_norm": 0.30176204442977905,
+      "learning_rate": 4.37768338197142e-09,
+      "loss": 0.3077,
+      "step": 11785
+    },
+    {
+      "epoch": 2.9642857142857144,
+      "grad_norm": 0.27803835272789,
+      "learning_rate": 4.3166798996208125e-09,
+      "loss": 0.3356,
+      "step": 11786
+    },
+    {
+      "epoch": 2.96453722334004,
+      "grad_norm": 0.30758580565452576,
+      "learning_rate": 4.256104266293637e-09,
+      "loss": 0.3263,
+      "step": 11787
+    },
+    {
+      "epoch": 2.964788732394366,
+      "grad_norm": 0.27492862939834595,
+      "learning_rate": 4.19595648717741e-09,
+      "loss": 0.3243,
+      "step": 11788
+    },
+    {
+      "epoch": 2.9650402414486923,
+      "grad_norm": 0.2788017988204956,
+      "learning_rate": 4.136236567424679e-09,
+      "loss": 0.3405,
+      "step": 11789
+    },
+    {
+      "epoch": 2.965291750503018,
+      "grad_norm": 0.2948797047138214,
+      "learning_rate": 4.076944512148573e-09,
+      "loss": 0.3143,
+      "step": 11790
+    },
+    {
+      "epoch": 2.965543259557344,
+      "grad_norm": 0.2663567364215851,
+      "learning_rate": 4.018080326428364e-09,
+      "loss": 0.2989,
+      "step": 11791
+    },
+    {
+      "epoch": 2.96579476861167,
+      "grad_norm": 0.2669561803340912,
+      "learning_rate": 3.9596440153044645e-09,
+      "loss": 0.3026,
+      "step": 11792
+    },
+    {
+      "epoch": 2.966046277665996,
+      "grad_norm": 0.28775742650032043,
+      "learning_rate": 3.901635583782315e-09,
+      "loss": 0.333,
+      "step": 11793
+    },
+    {
+      "epoch": 2.9662977867203217,
+      "grad_norm": 0.293270468711853,
+      "learning_rate": 3.844055036829053e-09,
+      "loss": 0.3608,
+      "step": 11794
+    },
+    {
+      "epoch": 2.966549295774648,
+      "grad_norm": 0.2894499897956848,
+      "learning_rate": 3.786902379376844e-09,
+      "loss": 0.3251,
+      "step": 11795
+    },
+    {
+      "epoch": 2.966800804828974,
+      "grad_norm": 0.2831835448741913,
+      "learning_rate": 3.730177616320108e-09,
+      "loss": 0.3216,
+      "step": 11796
+    },
+    {
+      "epoch": 2.9670523138832996,
+      "grad_norm": 0.2966955900192261,
+      "learning_rate": 3.6738807525171784e-09,
+      "loss": 0.3252,
+      "step": 11797
+    },
+    {
+      "epoch": 2.967303822937626,
+      "grad_norm": 0.2994356155395508,
+      "learning_rate": 3.618011792789755e-09,
+      "loss": 0.3243,
+      "step": 11798
+    },
+    {
+      "epoch": 2.9675553319919517,
+      "grad_norm": 0.2818380892276764,
+      "learning_rate": 3.562570741921789e-09,
+      "loss": 0.3147,
+      "step": 11799
+    },
+    {
+      "epoch": 2.9678068410462775,
+      "grad_norm": 0.29389482736587524,
+      "learning_rate": 3.5075576046628145e-09,
+      "loss": 0.3545,
+      "step": 11800
+    },
+    {
+      "epoch": 2.9680583501006037,
+      "grad_norm": 0.2830759584903717,
+      "learning_rate": 3.4529723857229526e-09,
+      "loss": 0.3221,
+      "step": 11801
+    },
+    {
+      "epoch": 2.9683098591549295,
+      "grad_norm": 0.28202909231185913,
+      "learning_rate": 3.3988150897779073e-09,
+      "loss": 0.3163,
+      "step": 11802
+    },
+    {
+      "epoch": 2.9685613682092553,
+      "grad_norm": 0.3006686866283417,
+      "learning_rate": 3.345085721465635e-09,
+      "loss": 0.3173,
+      "step": 11803
+    },
+    {
+      "epoch": 2.9688128772635816,
+      "grad_norm": 0.28550851345062256,
+      "learning_rate": 3.291784285387456e-09,
+      "loss": 0.3222,
+      "step": 11804
+    },
+    {
+      "epoch": 2.9690643863179074,
+      "grad_norm": 0.2890929877758026,
+      "learning_rate": 3.238910786109162e-09,
+      "loss": 0.305,
+      "step": 11805
+    },
+    {
+      "epoch": 2.969315895372233,
+      "grad_norm": 0.27377450466156006,
+      "learning_rate": 3.186465228158242e-09,
+      "loss": 0.3526,
+      "step": 11806
+    },
+    {
+      "epoch": 2.9695674044265594,
+      "grad_norm": 0.285478413105011,
+      "learning_rate": 3.1344476160266592e-09,
+      "loss": 0.3079,
+      "step": 11807
+    },
+    {
+      "epoch": 2.9698189134808852,
+      "grad_norm": 0.29703089594841003,
+      "learning_rate": 3.082857954169738e-09,
+      "loss": 0.2883,
+      "step": 11808
+    },
+    {
+      "epoch": 2.970070422535211,
+      "grad_norm": 0.2617098391056061,
+      "learning_rate": 3.031696247005056e-09,
+      "loss": 0.3055,
+      "step": 11809
+    },
+    {
+      "epoch": 2.9703219315895373,
+      "grad_norm": 0.2684277594089508,
+      "learning_rate": 2.9809624989146633e-09,
+      "loss": 0.3213,
+      "step": 11810
+    },
+    {
+      "epoch": 2.970573440643863,
+      "grad_norm": 0.28313401341438293,
+      "learning_rate": 2.9306567142434183e-09,
+      "loss": 0.3155,
+      "step": 11811
+    },
+    {
+      "epoch": 2.970824949698189,
+      "grad_norm": 0.2776716351509094,
+      "learning_rate": 2.8807788973000962e-09,
+      "loss": 0.3116,
+      "step": 11812
+    },
+    {
+      "epoch": 2.971076458752515,
+      "grad_norm": 0.29418906569480896,
+      "learning_rate": 2.83132905235628e-09,
+      "loss": 0.2839,
+      "step": 11813
+    },
+    {
+      "epoch": 2.971327967806841,
+      "grad_norm": 0.2912454903125763,
+      "learning_rate": 2.78230718364636e-09,
+      "loss": 0.3275,
+      "step": 11814
+    },
+    {
+      "epoch": 2.9715794768611667,
+      "grad_norm": 0.2744319438934326,
+      "learning_rate": 2.7337132953697555e-09,
+      "loss": 0.3291,
+      "step": 11815
+    },
+    {
+      "epoch": 2.971830985915493,
+      "grad_norm": 0.2988802194595337,
+      "learning_rate": 2.685547391688137e-09,
+      "loss": 0.3212,
+      "step": 11816
+    },
+    {
+      "epoch": 2.972082494969819,
+      "grad_norm": 0.28176239132881165,
+      "learning_rate": 2.6378094767259833e-09,
+      "loss": 0.3297,
+      "step": 11817
+    },
+    {
+      "epoch": 2.9723340040241446,
+      "grad_norm": 0.30423882603645325,
+      "learning_rate": 2.5904995545716903e-09,
+      "loss": 0.3004,
+      "step": 11818
+    },
+    {
+      "epoch": 2.972585513078471,
+      "grad_norm": 0.27662205696105957,
+      "learning_rate": 2.5436176292781277e-09,
+      "loss": 0.3055,
+      "step": 11819
+    },
+    {
+      "epoch": 2.9728370221327967,
+      "grad_norm": 0.28385183215141296,
+      "learning_rate": 2.497163704859307e-09,
+      "loss": 0.2986,
+      "step": 11820
+    },
+    {
+      "epoch": 2.9730885311871225,
+      "grad_norm": 0.2943600118160248,
+      "learning_rate": 2.4511377852937114e-09,
+      "loss": 0.3284,
+      "step": 11821
+    },
+    {
+      "epoch": 2.9733400402414487,
+      "grad_norm": 0.27290356159210205,
+      "learning_rate": 2.4055398745242987e-09,
+      "loss": 0.3114,
+      "step": 11822
+    },
+    {
+      "epoch": 2.9735915492957745,
+      "grad_norm": 0.2819903790950775,
+      "learning_rate": 2.360369976455168e-09,
+      "loss": 0.3206,
+      "step": 11823
+    },
+    {
+      "epoch": 2.9738430583501008,
+      "grad_norm": 0.28330057859420776,
+      "learning_rate": 2.3156280949554465e-09,
+      "loss": 0.3173,
+      "step": 11824
+    },
+    {
+      "epoch": 2.9740945674044266,
+      "grad_norm": 0.3153158724308014,
+      "learning_rate": 2.2713142338565142e-09,
+      "loss": 0.3447,
+      "step": 11825
+    },
+    {
+      "epoch": 2.9743460764587524,
+      "grad_norm": 0.278213232755661,
+      "learning_rate": 2.2274283969542233e-09,
+      "loss": 0.3002,
+      "step": 11826
+    },
+    {
+      "epoch": 2.9745975855130786,
+      "grad_norm": 0.29102084040641785,
+      "learning_rate": 2.1839705880061235e-09,
+      "loss": 0.3253,
+      "step": 11827
+    },
+    {
+      "epoch": 2.9748490945674044,
+      "grad_norm": 0.28229936957359314,
+      "learning_rate": 2.1409408107353478e-09,
+      "loss": 0.3337,
+      "step": 11828
+    },
+    {
+      "epoch": 2.9751006036217302,
+      "grad_norm": 0.28082865476608276,
+      "learning_rate": 2.0983390688261718e-09,
+      "loss": 0.3255,
+      "step": 11829
+    },
+    {
+      "epoch": 2.9753521126760565,
+      "grad_norm": 0.27319207787513733,
+      "learning_rate": 2.056165365927343e-09,
+      "loss": 0.3132,
+      "step": 11830
+    },
+    {
+      "epoch": 2.9756036217303823,
+      "grad_norm": 0.2930520474910736,
+      "learning_rate": 2.0144197056509717e-09,
+      "loss": 0.3183,
+      "step": 11831
+    },
+    {
+      "epoch": 2.975855130784708,
+      "grad_norm": 0.28141433000564575,
+      "learning_rate": 1.9731020915725317e-09,
+      "loss": 0.3456,
+      "step": 11832
+    },
+    {
+      "epoch": 2.9761066398390343,
+      "grad_norm": 0.28628331422805786,
+      "learning_rate": 1.9322125272297488e-09,
+      "loss": 0.3227,
+      "step": 11833
+    },
+    {
+      "epoch": 2.97635814889336,
+      "grad_norm": 0.30301791429519653,
+      "learning_rate": 1.8917510161259312e-09,
+      "loss": 0.315,
+      "step": 11834
+    },
+    {
+      "epoch": 2.9766096579476864,
+      "grad_norm": 0.28687921166419983,
+      "learning_rate": 1.851717561724975e-09,
+      "loss": 0.3036,
+      "step": 11835
+    },
+    {
+      "epoch": 2.976861167002012,
+      "grad_norm": 0.26738491654396057,
+      "learning_rate": 1.812112167456359e-09,
+      "loss": 0.3211,
+      "step": 11836
+    },
+    {
+      "epoch": 2.977112676056338,
+      "grad_norm": 0.3000773787498474,
+      "learning_rate": 1.7729348367118148e-09,
+      "loss": 0.3202,
+      "step": 11837
+    },
+    {
+      "epoch": 2.9773641851106643,
+      "grad_norm": 0.28814077377319336,
+      "learning_rate": 1.7341855728464363e-09,
+      "loss": 0.2894,
+      "step": 11838
+    },
+    {
+      "epoch": 2.97761569416499,
+      "grad_norm": 0.26308709383010864,
+      "learning_rate": 1.695864379179235e-09,
+      "loss": 0.3039,
+      "step": 11839
+    },
+    {
+      "epoch": 2.977867203219316,
+      "grad_norm": 0.28434452414512634,
+      "learning_rate": 1.6579712589914754e-09,
+      "loss": 0.322,
+      "step": 11840
+    },
+    {
+      "epoch": 2.978118712273642,
+      "grad_norm": 0.2894355058670044,
+      "learning_rate": 1.6205062155294494e-09,
+      "loss": 0.3197,
+      "step": 11841
+    },
+    {
+      "epoch": 2.978370221327968,
+      "grad_norm": 0.31105583906173706,
+      "learning_rate": 1.5834692520011462e-09,
+      "loss": 0.3124,
+      "step": 11842
+    },
+    {
+      "epoch": 2.9786217303822937,
+      "grad_norm": 0.2865217924118042,
+      "learning_rate": 1.5468603715784736e-09,
+      "loss": 0.3336,
+      "step": 11843
+    },
+    {
+      "epoch": 2.97887323943662,
+      "grad_norm": 0.3048990070819855,
+      "learning_rate": 1.510679577397256e-09,
+      "loss": 0.3265,
+      "step": 11844
+    },
+    {
+      "epoch": 2.9791247484909458,
+      "grad_norm": 0.27794545888900757,
+      "learning_rate": 1.4749268725555709e-09,
+      "loss": 0.3072,
+      "step": 11845
+    },
+    {
+      "epoch": 2.9793762575452716,
+      "grad_norm": 0.31118738651275635,
+      "learning_rate": 1.4396022601159687e-09,
+      "loss": 0.3398,
+      "step": 11846
+    },
+    {
+      "epoch": 2.979627766599598,
+      "grad_norm": 0.28356632590293884,
+      "learning_rate": 1.404705743103807e-09,
+      "loss": 0.3246,
+      "step": 11847
+    },
+    {
+      "epoch": 2.9798792756539236,
+      "grad_norm": 0.28750303387641907,
+      "learning_rate": 1.370237324507251e-09,
+      "loss": 0.3162,
+      "step": 11848
+    },
+    {
+      "epoch": 2.9801307847082494,
+      "grad_norm": 0.346623957157135,
+      "learning_rate": 1.3361970072783836e-09,
+      "loss": 0.2966,
+      "step": 11849
+    },
+    {
+      "epoch": 2.9803822937625757,
+      "grad_norm": 0.309396892786026,
+      "learning_rate": 1.3025847943326508e-09,
+      "loss": 0.3277,
+      "step": 11850
+    },
+    {
+      "epoch": 2.9806338028169015,
+      "grad_norm": 0.2721833884716034,
+      "learning_rate": 1.2694006885488609e-09,
+      "loss": 0.3412,
+      "step": 11851
+    },
+    {
+      "epoch": 2.9808853118712273,
+      "grad_norm": 0.3036994934082031,
+      "learning_rate": 1.2366446927691844e-09,
+      "loss": 0.2995,
+      "step": 11852
+    },
+    {
+      "epoch": 2.9811368209255535,
+      "grad_norm": 0.2673596441745758,
+      "learning_rate": 1.2043168097986002e-09,
+      "loss": 0.3089,
+      "step": 11853
+    },
+    {
+      "epoch": 2.9813883299798793,
+      "grad_norm": 0.3089800179004669,
+      "learning_rate": 1.1724170424054493e-09,
+      "loss": 0.3185,
+      "step": 11854
+    },
+    {
+      "epoch": 2.981639839034205,
+      "grad_norm": 0.26984110474586487,
+      "learning_rate": 1.1409453933225457e-09,
+      "loss": 0.3277,
+      "step": 11855
+    },
+    {
+      "epoch": 2.9818913480885314,
+      "grad_norm": 0.2922016382217407,
+      "learning_rate": 1.1099018652449557e-09,
+      "loss": 0.3058,
+      "step": 11856
+    },
+    {
+      "epoch": 2.982142857142857,
+      "grad_norm": 0.2652096152305603,
+      "learning_rate": 1.0792864608316634e-09,
+      "loss": 0.3217,
+      "step": 11857
+    },
+    {
+      "epoch": 2.982394366197183,
+      "grad_norm": 0.3048350512981415,
+      "learning_rate": 1.0490991827039055e-09,
+      "loss": 0.2961,
+      "step": 11858
+    },
+    {
+      "epoch": 2.9826458752515093,
+      "grad_norm": 0.26128822565078735,
+      "learning_rate": 1.0193400334473913e-09,
+      "loss": 0.3407,
+      "step": 11859
+    },
+    {
+      "epoch": 2.982897384305835,
+      "grad_norm": 0.28069064021110535,
+      "learning_rate": 9.90009015611193e-10,
+      "loss": 0.3106,
+      "step": 11860
+    },
+    {
+      "epoch": 2.983148893360161,
+      "grad_norm": 0.2878744602203369,
+      "learning_rate": 9.61106131706635e-10,
+      "loss": 0.3249,
+      "step": 11861
+    },
+    {
+      "epoch": 2.983400402414487,
+      "grad_norm": 0.2882971167564392,
+      "learning_rate": 9.326313842100698e-10,
+      "loss": 0.3099,
+      "step": 11862
+    },
+    {
+      "epoch": 2.983651911468813,
+      "grad_norm": 0.2869325578212738,
+      "learning_rate": 9.045847755589921e-10,
+      "loss": 0.3298,
+      "step": 11863
+    },
+    {
+      "epoch": 2.9839034205231387,
+      "grad_norm": 0.29187971353530884,
+      "learning_rate": 8.769663081559243e-10,
+      "loss": 0.3112,
+      "step": 11864
+    },
+    {
+      "epoch": 2.984154929577465,
+      "grad_norm": 0.2968631684780121,
+      "learning_rate": 8.497759843667519e-10,
+      "loss": 0.3326,
+      "step": 11865
+    },
+    {
+      "epoch": 2.984406438631791,
+      "grad_norm": 0.3074916899204254,
+      "learning_rate": 8.230138065196125e-10,
+      "loss": 0.3237,
+      "step": 11866
+    },
+    {
+      "epoch": 2.9846579476861166,
+      "grad_norm": 0.2778856158256531,
+      "learning_rate": 7.966797769065615e-10,
+      "loss": 0.3291,
+      "step": 11867
+    },
+    {
+      "epoch": 2.984909456740443,
+      "grad_norm": 0.2621375620365143,
+      "learning_rate": 7.707738977824619e-10,
+      "loss": 0.3146,
+      "step": 11868
+    },
+    {
+      "epoch": 2.9851609657947686,
+      "grad_norm": 0.25857844948768616,
+      "learning_rate": 7.452961713672046e-10,
+      "loss": 0.3171,
+      "step": 11869
+    },
+    {
+      "epoch": 2.9854124748490944,
+      "grad_norm": 0.2810332775115967,
+      "learning_rate": 7.202465998412678e-10,
+      "loss": 0.3306,
+      "step": 11870
+    },
+    {
+      "epoch": 2.9856639839034207,
+      "grad_norm": 0.2880176603794098,
+      "learning_rate": 6.956251853512675e-10,
+      "loss": 0.3065,
+      "step": 11871
+    },
+    {
+      "epoch": 2.9859154929577465,
+      "grad_norm": 0.2959819436073303,
+      "learning_rate": 6.714319300055172e-10,
+      "loss": 0.332,
+      "step": 11872
+    },
+    {
+      "epoch": 2.9861670020120723,
+      "grad_norm": 0.2593652606010437,
+      "learning_rate": 6.476668358762483e-10,
+      "loss": 0.3388,
+      "step": 11873
+    },
+    {
+      "epoch": 2.9864185110663986,
+      "grad_norm": 0.28062987327575684,
+      "learning_rate": 6.243299049979445e-10,
+      "loss": 0.3382,
+      "step": 11874
+    },
+    {
+      "epoch": 2.9866700201207244,
+      "grad_norm": 0.27674224972724915,
+      "learning_rate": 6.014211393695623e-10,
+      "loss": 0.3121,
+      "step": 11875
+    },
+    {
+      "epoch": 2.98692152917505,
+      "grad_norm": 0.2833153009414673,
+      "learning_rate": 5.789405409539761e-10,
+      "loss": 0.3147,
+      "step": 11876
+    },
+    {
+      "epoch": 2.9871730382293764,
+      "grad_norm": 0.2868058383464813,
+      "learning_rate": 5.568881116752023e-10,
+      "loss": 0.3182,
+      "step": 11877
+    },
+    {
+      "epoch": 2.987424547283702,
+      "grad_norm": 0.2811480760574341,
+      "learning_rate": 5.352638534228405e-10,
+      "loss": 0.2941,
+      "step": 11878
+    },
+    {
+      "epoch": 2.987676056338028,
+      "grad_norm": 0.29555627703666687,
+      "learning_rate": 5.140677680487427e-10,
+      "loss": 0.3058,
+      "step": 11879
+    },
+    {
+      "epoch": 2.9879275653923543,
+      "grad_norm": 0.2923000752925873,
+      "learning_rate": 4.932998573681236e-10,
+      "loss": 0.3494,
+      "step": 11880
+    },
+    {
+      "epoch": 2.98817907444668,
+      "grad_norm": 0.27389127016067505,
+      "learning_rate": 4.729601231590053e-10,
+      "loss": 0.3071,
+      "step": 11881
+    },
+    {
+      "epoch": 2.988430583501006,
+      "grad_norm": 0.27962055802345276,
+      "learning_rate": 4.5304856716443803e-10,
+      "loss": 0.3109,
+      "step": 11882
+    },
+    {
+      "epoch": 2.988682092555332,
+      "grad_norm": 0.30079811811447144,
+      "learning_rate": 4.3356519108916914e-10,
+      "loss": 0.3221,
+      "step": 11883
+    },
+    {
+      "epoch": 2.988933601609658,
+      "grad_norm": 0.2800407409667969,
+      "learning_rate": 4.14509996601864e-10,
+      "loss": 0.3115,
+      "step": 11884
+    },
+    {
+      "epoch": 2.9891851106639837,
+      "grad_norm": 0.27782750129699707,
+      "learning_rate": 3.9588298533399515e-10,
+      "loss": 0.3386,
+      "step": 11885
+    },
+    {
+      "epoch": 2.98943661971831,
+      "grad_norm": 0.2804793119430542,
+      "learning_rate": 3.7768415888150835e-10,
+      "loss": 0.3244,
+      "step": 11886
+    },
+    {
+      "epoch": 2.989688128772636,
+      "grad_norm": 0.2901947796344757,
+      "learning_rate": 3.5991351880315663e-10,
+      "loss": 0.3037,
+      "step": 11887
+    },
+    {
+      "epoch": 2.9899396378269616,
+      "grad_norm": 0.2956394553184509,
+      "learning_rate": 3.4257106662050066e-10,
+      "loss": 0.3442,
+      "step": 11888
+    },
+    {
+      "epoch": 2.990191146881288,
+      "grad_norm": 0.3086378574371338,
+      "learning_rate": 3.2565680381846377e-10,
+      "loss": 0.3066,
+      "step": 11889
+    },
+    {
+      "epoch": 2.9904426559356136,
+      "grad_norm": 0.30673331022262573,
+      "learning_rate": 3.091707318464421e-10,
+      "loss": 0.3219,
+      "step": 11890
+    },
+    {
+      "epoch": 2.9906941649899395,
+      "grad_norm": 0.28686726093292236,
+      "learning_rate": 2.931128521160842e-10,
+      "loss": 0.3265,
+      "step": 11891
+    },
+    {
+      "epoch": 2.9909456740442657,
+      "grad_norm": 0.2747303247451782,
+      "learning_rate": 2.774831660018462e-10,
+      "loss": 0.3198,
+      "step": 11892
+    },
+    {
+      "epoch": 2.9911971830985915,
+      "grad_norm": 0.27951404452323914,
+      "learning_rate": 2.622816748437673e-10,
+      "loss": 0.3252,
+      "step": 11893
+    },
+    {
+      "epoch": 2.9914486921529173,
+      "grad_norm": 0.27234378457069397,
+      "learning_rate": 2.475083799424738e-10,
+      "loss": 0.2957,
+      "step": 11894
+    },
+    {
+      "epoch": 2.9917002012072436,
+      "grad_norm": 0.28382253646850586,
+      "learning_rate": 2.33163282564175e-10,
+      "loss": 0.3286,
+      "step": 11895
+    },
+    {
+      "epoch": 2.9919517102615694,
+      "grad_norm": 0.28595641255378723,
+      "learning_rate": 2.1924638393677755e-10,
+      "loss": 0.3089,
+      "step": 11896
+    },
+    {
+      "epoch": 2.992203219315895,
+      "grad_norm": 0.28392264246940613,
+      "learning_rate": 2.0575768525266105e-10,
+      "loss": 0.3039,
+      "step": 11897
+    },
+    {
+      "epoch": 2.9924547283702214,
+      "grad_norm": 0.307187020778656,
+      "learning_rate": 1.926971876664574e-10,
+      "loss": 0.3291,
+      "step": 11898
+    },
+    {
+      "epoch": 2.9927062374245472,
+      "grad_norm": 0.29181286692619324,
+      "learning_rate": 1.800648922967163e-10,
+      "loss": 0.3418,
+      "step": 11899
+    },
+    {
+      "epoch": 2.992957746478873,
+      "grad_norm": 0.2649078965187073,
+      "learning_rate": 1.6786080022646034e-10,
+      "loss": 0.3342,
+      "step": 11900
+    },
+    {
+      "epoch": 2.9932092555331993,
+      "grad_norm": 0.2709779739379883,
+      "learning_rate": 1.5608491249929913e-10,
+      "loss": 0.3176,
+      "step": 11901
+    },
+    {
+      "epoch": 2.993460764587525,
+      "grad_norm": 0.2838386595249176,
+      "learning_rate": 1.4473723012498053e-10,
+      "loss": 0.3059,
+      "step": 11902
+    },
+    {
+      "epoch": 2.993712273641851,
+      "grad_norm": 0.2875151038169861,
+      "learning_rate": 1.338177540749497e-10,
+      "loss": 0.3346,
+      "step": 11903
+    },
+    {
+      "epoch": 2.993963782696177,
+      "grad_norm": 0.29026007652282715,
+      "learning_rate": 1.2332648528401436e-10,
+      "loss": 0.3377,
+      "step": 11904
+    },
+    {
+      "epoch": 2.994215291750503,
+      "grad_norm": 0.29916590452194214,
+      "learning_rate": 1.1326342465145523e-10,
+      "loss": 0.3209,
+      "step": 11905
+    },
+    {
+      "epoch": 2.9944668008048287,
+      "grad_norm": 0.28423720598220825,
+      "learning_rate": 1.0362857303825025e-10,
+      "loss": 0.3078,
+      "step": 11906
+    },
+    {
+      "epoch": 2.994718309859155,
+      "grad_norm": 0.2835725247859955,
+      "learning_rate": 9.442193127040533e-11,
+      "loss": 0.318,
+      "step": 11907
+    },
+    {
+      "epoch": 2.994969818913481,
+      "grad_norm": 0.27757924795150757,
+      "learning_rate": 8.564350013617884e-11,
+      "loss": 0.3137,
+      "step": 11908
+    },
+    {
+      "epoch": 2.9952213279678066,
+      "grad_norm": 0.28157174587249756,
+      "learning_rate": 7.729328038663663e-11,
+      "loss": 0.3522,
+      "step": 11909
+    },
+    {
+      "epoch": 2.995472837022133,
+      "grad_norm": 0.3048500716686249,
+      "learning_rate": 6.937127273787258e-11,
+      "loss": 0.3175,
+      "step": 11910
+    },
+    {
+      "epoch": 2.9957243460764587,
+      "grad_norm": 0.28835025429725647,
+      "learning_rate": 6.187747786767783e-11,
+      "loss": 0.3322,
+      "step": 11911
+    },
+    {
+      "epoch": 2.9959758551307845,
+      "grad_norm": 0.2930144965648651,
+      "learning_rate": 5.4811896418316414e-11,
+      "loss": 0.3168,
+      "step": 11912
+    },
+    {
+      "epoch": 2.9962273641851107,
+      "grad_norm": 0.2962915599346161,
+      "learning_rate": 4.817452899485986e-11,
+      "loss": 0.3037,
+      "step": 11913
+    },
+    {
+      "epoch": 2.9964788732394365,
+      "grad_norm": 0.2874841094017029,
+      "learning_rate": 4.1965376165742365e-11,
+      "loss": 0.3398,
+      "step": 11914
+    },
+    {
+      "epoch": 2.9967303822937623,
+      "grad_norm": 0.2807866632938385,
+      "learning_rate": 3.618443846276076e-11,
+      "loss": 0.3102,
+      "step": 11915
+    },
+    {
+      "epoch": 2.9969818913480886,
+      "grad_norm": 0.27110809087753296,
+      "learning_rate": 3.0831716380519406e-11,
+      "loss": 0.3245,
+      "step": 11916
+    },
+    {
+      "epoch": 2.9972334004024144,
+      "grad_norm": 0.30204376578330994,
+      "learning_rate": 2.5907210378095514e-11,
+      "loss": 0.3111,
+      "step": 11917
+    },
+    {
+      "epoch": 2.99748490945674,
+      "grad_norm": 0.29356303811073303,
+      "learning_rate": 2.141092087681873e-11,
+      "loss": 0.3641,
+      "step": 11918
+    },
+    {
+      "epoch": 2.9977364185110664,
+      "grad_norm": 0.3049434721469879,
+      "learning_rate": 1.734284826193644e-11,
+      "loss": 0.3259,
+      "step": 11919
+    },
+    {
+      "epoch": 2.9979879275653922,
+      "grad_norm": 0.28138816356658936,
+      "learning_rate": 1.3702992882058675e-11,
+      "loss": 0.3219,
+      "step": 11920
+    },
+    {
+      "epoch": 2.998239436619718,
+      "grad_norm": 0.29172345995903015,
+      "learning_rate": 1.049135504804788e-11,
+      "loss": 0.3215,
+      "step": 11921
+    },
+    {
+      "epoch": 2.9984909456740443,
+      "grad_norm": 0.28759902715682983,
+      "learning_rate": 7.70793503634959e-12,
+      "loss": 0.3142,
+      "step": 11922
+    },
+    {
+      "epoch": 2.99874245472837,
+      "grad_norm": 0.3014575242996216,
+      "learning_rate": 5.352733084551531e-12,
+      "loss": 0.3054,
+      "step": 11923
+    },
+    {
+      "epoch": 2.9989939637826963,
+      "grad_norm": 0.2920631766319275,
+      "learning_rate": 3.4257493941591837e-12,
+      "loss": 0.3298,
+      "step": 11924
+    },
+    {
+      "epoch": 2.999245472837022,
+      "grad_norm": 0.2961787283420563,
+      "learning_rate": 1.926984130595777e-12,
+      "loss": 0.3011,
+      "step": 11925
+    },
+    {
+      "epoch": 2.999496981891348,
+      "grad_norm": 0.28382429480552673,
+      "learning_rate": 8.564374220920713e-13,
+      "loss": 0.3406,
+      "step": 11926
+    },
+    {
+      "epoch": 2.999748490945674,
+      "grad_norm": 0.2791459858417511,
+      "learning_rate": 2.1410936024146568e-13,
+      "loss": 0.3243,
+      "step": 11927
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2965925633907318,
+      "learning_rate": 0.0,
+      "loss": 0.2831,
+      "step": 11928
+    },
+    {
+      "epoch": 3.0,
+      "step": 11928,
+      "total_flos": 1.1291892680425472e+16,
+      "train_loss": 0.36148936700982826,
+      "train_runtime": 214787.5482,
+      "train_samples_per_second": 5.331,
+      "train_steps_per_second": 0.056
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11928,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1291892680425472e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}