diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,70873 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.964806124969128,
+  "eval_steps": 500,
+  "global_step": 10120,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000395159298592245,
+      "grad_norm": 45.00149749159182,
+      "learning_rate": 1.0638297872340427e-08,
+      "loss": 5.3324,
+      "step": 1
+    },
+    {
+      "epoch": 0.00079031859718449,
+      "grad_norm": 47.65956613810862,
+      "learning_rate": 2.1276595744680853e-08,
+      "loss": 5.4979,
+      "step": 2
+    },
+    {
+      "epoch": 0.001185477895776735,
+      "grad_norm": 44.18618860656718,
+      "learning_rate": 3.191489361702128e-08,
+      "loss": 5.1904,
+      "step": 3
+    },
+    {
+      "epoch": 0.00158063719436898,
+      "grad_norm": 46.48562838932194,
+      "learning_rate": 4.2553191489361707e-08,
+      "loss": 5.3019,
+      "step": 4
+    },
+    {
+      "epoch": 0.001975796492961225,
+      "grad_norm": 46.390295089186544,
+      "learning_rate": 5.319148936170213e-08,
+      "loss": 5.3445,
+      "step": 5
+    },
+    {
+      "epoch": 0.00237095579155347,
+      "grad_norm": 46.76761442399584,
+      "learning_rate": 6.382978723404255e-08,
+      "loss": 5.3375,
+      "step": 6
+    },
+    {
+      "epoch": 0.002766115090145715,
+      "grad_norm": 48.279218165562085,
+      "learning_rate": 7.446808510638299e-08,
+      "loss": 5.3938,
+      "step": 7
+    },
+    {
+      "epoch": 0.00316127438873796,
+      "grad_norm": 49.205312835168,
+      "learning_rate": 8.510638297872341e-08,
+      "loss": 5.4594,
+      "step": 8
+    },
+    {
+      "epoch": 0.003556433687330205,
+      "grad_norm": 46.56040077093359,
+      "learning_rate": 9.574468085106384e-08,
+      "loss": 5.3369,
+      "step": 9
+    },
+    {
+      "epoch": 0.00395159298592245,
+      "grad_norm": 45.73670653221075,
+      "learning_rate": 1.0638297872340426e-07,
+      "loss": 5.28,
+      "step": 10
+    },
+    {
+      "epoch": 0.004346752284514695,
+      "grad_norm": 45.65613647161846,
+      "learning_rate": 1.1702127659574468e-07,
+      "loss": 5.4326,
+      "step": 11
+    },
+    {
+      "epoch": 0.00474191158310694,
+      "grad_norm": 48.50884423593061,
+      "learning_rate": 1.276595744680851e-07,
+      "loss": 5.4455,
+      "step": 12
+    },
+    {
+      "epoch": 0.005137070881699185,
+      "grad_norm": 48.08180573773191,
+      "learning_rate": 1.3829787234042553e-07,
+      "loss": 5.5208,
+      "step": 13
+    },
+    {
+      "epoch": 0.00553223018029143,
+      "grad_norm": 45.43412625685486,
+      "learning_rate": 1.4893617021276598e-07,
+      "loss": 5.3535,
+      "step": 14
+    },
+    {
+      "epoch": 0.005927389478883675,
+      "grad_norm": 46.86389365193829,
+      "learning_rate": 1.5957446808510638e-07,
+      "loss": 5.3197,
+      "step": 15
+    },
+    {
+      "epoch": 0.00632254877747592,
+      "grad_norm": 44.6253221018194,
+      "learning_rate": 1.7021276595744683e-07,
+      "loss": 5.4598,
+      "step": 16
+    },
+    {
+      "epoch": 0.006717708076068165,
+      "grad_norm": 45.02802440676611,
+      "learning_rate": 1.8085106382978722e-07,
+      "loss": 5.1813,
+      "step": 17
+    },
+    {
+      "epoch": 0.00711286737466041,
+      "grad_norm": 44.458003741695514,
+      "learning_rate": 1.9148936170212767e-07,
+      "loss": 5.299,
+      "step": 18
+    },
+    {
+      "epoch": 0.007508026673252655,
+      "grad_norm": 42.134533089902625,
+      "learning_rate": 2.0212765957446812e-07,
+      "loss": 5.0162,
+      "step": 19
+    },
+    {
+      "epoch": 0.0079031859718449,
+      "grad_norm": 45.3211377446349,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 5.3317,
+      "step": 20
+    },
+    {
+      "epoch": 0.008298345270437145,
+      "grad_norm": 39.99349721590117,
+      "learning_rate": 2.2340425531914897e-07,
+      "loss": 5.1135,
+      "step": 21
+    },
+    {
+      "epoch": 0.00869350456902939,
+      "grad_norm": 38.01803171664136,
+      "learning_rate": 2.3404255319148937e-07,
+      "loss": 5.1482,
+      "step": 22
+    },
+    {
+      "epoch": 0.009088663867621635,
+      "grad_norm": 39.354061184796315,
+      "learning_rate": 2.446808510638298e-07,
+      "loss": 5.2229,
+      "step": 23
+    },
+    {
+      "epoch": 0.00948382316621388,
+      "grad_norm": 40.80324954040139,
+      "learning_rate": 2.553191489361702e-07,
+      "loss": 5.2,
+      "step": 24
+    },
+    {
+      "epoch": 0.009878982464806126,
+      "grad_norm": 38.10588841551195,
+      "learning_rate": 2.6595744680851066e-07,
+      "loss": 5.0376,
+      "step": 25
+    },
+    {
+      "epoch": 0.01027414176339837,
+      "grad_norm": 38.040044463686115,
+      "learning_rate": 2.7659574468085106e-07,
+      "loss": 5.234,
+      "step": 26
+    },
+    {
+      "epoch": 0.010669301061990615,
+      "grad_norm": 38.93658315055219,
+      "learning_rate": 2.872340425531915e-07,
+      "loss": 5.2858,
+      "step": 27
+    },
+    {
+      "epoch": 0.01106446036058286,
+      "grad_norm": 27.7819822121004,
+      "learning_rate": 2.9787234042553196e-07,
+      "loss": 4.8375,
+      "step": 28
+    },
+    {
+      "epoch": 0.011459619659175104,
+      "grad_norm": 27.368335853965625,
+      "learning_rate": 3.0851063829787236e-07,
+      "loss": 4.9516,
+      "step": 29
+    },
+    {
+      "epoch": 0.01185477895776735,
+      "grad_norm": 27.250102842852968,
+      "learning_rate": 3.1914893617021275e-07,
+      "loss": 4.926,
+      "step": 30
+    },
+    {
+      "epoch": 0.012249938256359595,
+      "grad_norm": 26.50459159867008,
+      "learning_rate": 3.297872340425532e-07,
+      "loss": 4.8223,
+      "step": 31
+    },
+    {
+      "epoch": 0.01264509755495184,
+      "grad_norm": 27.513101530119382,
+      "learning_rate": 3.4042553191489365e-07,
+      "loss": 5.0109,
+      "step": 32
+    },
+    {
+      "epoch": 0.013040256853544084,
+      "grad_norm": 26.6602907635092,
+      "learning_rate": 3.510638297872341e-07,
+      "loss": 4.8947,
+      "step": 33
+    },
+    {
+      "epoch": 0.01343541615213633,
+      "grad_norm": 23.204176582249318,
+      "learning_rate": 3.6170212765957445e-07,
+      "loss": 4.6691,
+      "step": 34
+    },
+    {
+      "epoch": 0.013830575450728575,
+      "grad_norm": 24.66170575750984,
+      "learning_rate": 3.723404255319149e-07,
+      "loss": 4.9299,
+      "step": 35
+    },
+    {
+      "epoch": 0.01422573474932082,
+      "grad_norm": 24.263834716366237,
+      "learning_rate": 3.8297872340425535e-07,
+      "loss": 4.9347,
+      "step": 36
+    },
+    {
+      "epoch": 0.014620894047913065,
+      "grad_norm": 23.965936128051734,
+      "learning_rate": 3.936170212765958e-07,
+      "loss": 4.8474,
+      "step": 37
+    },
+    {
+      "epoch": 0.01501605334650531,
+      "grad_norm": 18.157336850120906,
+      "learning_rate": 4.0425531914893625e-07,
+      "loss": 4.6366,
+      "step": 38
+    },
+    {
+      "epoch": 0.015411212645097556,
+      "grad_norm": 17.163104051613317,
+      "learning_rate": 4.148936170212766e-07,
+      "loss": 4.5056,
+      "step": 39
+    },
+    {
+      "epoch": 0.0158063719436898,
+      "grad_norm": 18.14853573769075,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 4.5394,
+      "step": 40
+    },
+    {
+      "epoch": 0.016201531242282043,
+      "grad_norm": 17.031326941596244,
+      "learning_rate": 4.361702127659575e-07,
+      "loss": 4.5494,
+      "step": 41
+    },
+    {
+      "epoch": 0.01659669054087429,
+      "grad_norm": 16.530898546137706,
+      "learning_rate": 4.4680851063829794e-07,
+      "loss": 4.3963,
+      "step": 42
+    },
+    {
+      "epoch": 0.016991849839466534,
+      "grad_norm": 15.10305267365516,
+      "learning_rate": 4.574468085106383e-07,
+      "loss": 4.3983,
+      "step": 43
+    },
+    {
+      "epoch": 0.01738700913805878,
+      "grad_norm": 15.265493803692115,
+      "learning_rate": 4.6808510638297873e-07,
+      "loss": 4.3438,
+      "step": 44
+    },
+    {
+      "epoch": 0.017782168436651025,
+      "grad_norm": 14.891772118131469,
+      "learning_rate": 4.787234042553192e-07,
+      "loss": 4.356,
+      "step": 45
+    },
+    {
+      "epoch": 0.01817732773524327,
+      "grad_norm": 14.448417003857594,
+      "learning_rate": 4.893617021276596e-07,
+      "loss": 4.3505,
+      "step": 46
+    },
+    {
+      "epoch": 0.018572487033835516,
+      "grad_norm": 14.013807338161678,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 4.3902,
+      "step": 47
+    },
+    {
+      "epoch": 0.01896764633242776,
+      "grad_norm": 12.903706951798439,
+      "learning_rate": 5.106382978723404e-07,
+      "loss": 4.2129,
+      "step": 48
+    },
+    {
+      "epoch": 0.019362805631020004,
+      "grad_norm": 12.74392330179218,
+      "learning_rate": 5.212765957446809e-07,
+      "loss": 4.2245,
+      "step": 49
+    },
+    {
+      "epoch": 0.01975796492961225,
+      "grad_norm": 11.409859418409809,
+      "learning_rate": 5.319148936170213e-07,
+      "loss": 4.0668,
+      "step": 50
+    },
+    {
+      "epoch": 0.020153124228204495,
+      "grad_norm": 12.382991509917876,
+      "learning_rate": 5.425531914893618e-07,
+      "loss": 4.0922,
+      "step": 51
+    },
+    {
+      "epoch": 0.02054828352679674,
+      "grad_norm": 11.858955825560276,
+      "learning_rate": 5.531914893617021e-07,
+      "loss": 4.0067,
+      "step": 52
+    },
+    {
+      "epoch": 0.020943442825388986,
+      "grad_norm": 12.68072624432379,
+      "learning_rate": 5.638297872340426e-07,
+      "loss": 4.0186,
+      "step": 53
+    },
+    {
+      "epoch": 0.02133860212398123,
+      "grad_norm": 10.958508064800348,
+      "learning_rate": 5.74468085106383e-07,
+      "loss": 3.873,
+      "step": 54
+    },
+    {
+      "epoch": 0.021733761422573473,
+      "grad_norm": 10.842426212742,
+      "learning_rate": 5.851063829787235e-07,
+      "loss": 3.9817,
+      "step": 55
+    },
+    {
+      "epoch": 0.02212892072116572,
+      "grad_norm": 10.471356929993854,
+      "learning_rate": 5.957446808510639e-07,
+      "loss": 3.9189,
+      "step": 56
+    },
+    {
+      "epoch": 0.022524080019757965,
+      "grad_norm": 9.522689670857774,
+      "learning_rate": 6.063829787234043e-07,
+      "loss": 3.7485,
+      "step": 57
+    },
+    {
+      "epoch": 0.02291923931835021,
+      "grad_norm": 9.89336870076738,
+      "learning_rate": 6.170212765957447e-07,
+      "loss": 3.7502,
+      "step": 58
+    },
+    {
+      "epoch": 0.023314398616942456,
+      "grad_norm": 10.12284131423492,
+      "learning_rate": 6.276595744680851e-07,
+      "loss": 3.784,
+      "step": 59
+    },
+    {
+      "epoch": 0.0237095579155347,
+      "grad_norm": 9.549611583907577,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 3.7651,
+      "step": 60
+    },
+    {
+      "epoch": 0.024104717214126947,
+      "grad_norm": 9.025126475730096,
+      "learning_rate": 6.48936170212766e-07,
+      "loss": 3.6512,
+      "step": 61
+    },
+    {
+      "epoch": 0.02449987651271919,
+      "grad_norm": 8.597629129461852,
+      "learning_rate": 6.595744680851064e-07,
+      "loss": 3.7376,
+      "step": 62
+    },
+    {
+      "epoch": 0.024895035811311434,
+      "grad_norm": 8.345445781598853,
+      "learning_rate": 6.702127659574469e-07,
+      "loss": 3.4983,
+      "step": 63
+    },
+    {
+      "epoch": 0.02529019510990368,
+      "grad_norm": 7.990649600345105,
+      "learning_rate": 6.808510638297873e-07,
+      "loss": 3.5068,
+      "step": 64
+    },
+    {
+      "epoch": 0.025685354408495925,
+      "grad_norm": 7.513578806398596,
+      "learning_rate": 6.914893617021278e-07,
+      "loss": 3.5573,
+      "step": 65
+    },
+    {
+      "epoch": 0.02608051370708817,
+      "grad_norm": 7.9623181239477105,
+      "learning_rate": 7.021276595744682e-07,
+      "loss": 3.5052,
+      "step": 66
+    },
+    {
+      "epoch": 0.026475673005680416,
+      "grad_norm": 7.637547791371253,
+      "learning_rate": 7.127659574468087e-07,
+      "loss": 3.4709,
+      "step": 67
+    },
+    {
+      "epoch": 0.02687083230427266,
+      "grad_norm": 8.558533087051865,
+      "learning_rate": 7.234042553191489e-07,
+      "loss": 3.3483,
+      "step": 68
+    },
+    {
+      "epoch": 0.027265991602864904,
+      "grad_norm": 8.199113740225883,
+      "learning_rate": 7.340425531914893e-07,
+      "loss": 3.2996,
+      "step": 69
+    },
+    {
+      "epoch": 0.02766115090145715,
+      "grad_norm": 8.450241754459654,
+      "learning_rate": 7.446808510638298e-07,
+      "loss": 3.1916,
+      "step": 70
+    },
+    {
+      "epoch": 0.028056310200049395,
+      "grad_norm": 7.103052632657081,
+      "learning_rate": 7.553191489361702e-07,
+      "loss": 3.165,
+      "step": 71
+    },
+    {
+      "epoch": 0.02845146949864164,
+      "grad_norm": 7.397343059451777,
+      "learning_rate": 7.659574468085107e-07,
+      "loss": 3.1549,
+      "step": 72
+    },
+    {
+      "epoch": 0.028846628797233886,
+      "grad_norm": 7.532992807551275,
+      "learning_rate": 7.765957446808511e-07,
+      "loss": 3.0944,
+      "step": 73
+    },
+    {
+      "epoch": 0.02924178809582613,
+      "grad_norm": 6.784491281435092,
+      "learning_rate": 7.872340425531916e-07,
+      "loss": 3.0294,
+      "step": 74
+    },
+    {
+      "epoch": 0.029636947394418373,
+      "grad_norm": 7.083810212959109,
+      "learning_rate": 7.97872340425532e-07,
+      "loss": 3.0021,
+      "step": 75
+    },
+    {
+      "epoch": 0.03003210669301062,
+      "grad_norm": 6.443669804960258,
+      "learning_rate": 8.085106382978725e-07,
+      "loss": 2.9301,
+      "step": 76
+    },
+    {
+      "epoch": 0.030427265991602864,
+      "grad_norm": 6.650880145536403,
+      "learning_rate": 8.191489361702127e-07,
+      "loss": 2.9397,
+      "step": 77
+    },
+    {
+      "epoch": 0.03082242529019511,
+      "grad_norm": 6.36715615120205,
+      "learning_rate": 8.297872340425532e-07,
+      "loss": 2.8843,
+      "step": 78
+    },
+    {
+      "epoch": 0.031217584588787355,
+      "grad_norm": 6.238954181171991,
+      "learning_rate": 8.404255319148936e-07,
+      "loss": 2.7895,
+      "step": 79
+    },
+    {
+      "epoch": 0.0316127438873796,
+      "grad_norm": 6.232900546667891,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 2.8601,
+      "step": 80
+    },
+    {
+      "epoch": 0.032007903185971846,
+      "grad_norm": 5.726543181847973,
+      "learning_rate": 8.617021276595745e-07,
+      "loss": 2.6892,
+      "step": 81
+    },
+    {
+      "epoch": 0.03240306248456409,
+      "grad_norm": 5.502730599732169,
+      "learning_rate": 8.72340425531915e-07,
+      "loss": 2.6352,
+      "step": 82
+    },
+    {
+      "epoch": 0.032798221783156334,
+      "grad_norm": 5.739593573701677,
+      "learning_rate": 8.829787234042554e-07,
+      "loss": 2.6015,
+      "step": 83
+    },
+    {
+      "epoch": 0.03319338108174858,
+      "grad_norm": 5.6507901978379955,
+      "learning_rate": 8.936170212765959e-07,
+      "loss": 2.5505,
+      "step": 84
+    },
+    {
+      "epoch": 0.03358854038034083,
+      "grad_norm": 5.183242056909587,
+      "learning_rate": 9.042553191489363e-07,
+      "loss": 2.5001,
+      "step": 85
+    },
+    {
+      "epoch": 0.03398369967893307,
+      "grad_norm": 5.362205769595081,
+      "learning_rate": 9.148936170212766e-07,
+      "loss": 2.5267,
+      "step": 86
+    },
+    {
+      "epoch": 0.034378858977525316,
+      "grad_norm": 4.811942843640751,
+      "learning_rate": 9.25531914893617e-07,
+      "loss": 2.3826,
+      "step": 87
+    },
+    {
+      "epoch": 0.03477401827611756,
+      "grad_norm": 4.904542737672584,
+      "learning_rate": 9.361702127659575e-07,
+      "loss": 2.4234,
+      "step": 88
+    },
+    {
+      "epoch": 0.0351691775747098,
+      "grad_norm": 4.870107420594597,
+      "learning_rate": 9.468085106382979e-07,
+      "loss": 2.321,
+      "step": 89
+    },
+    {
+      "epoch": 0.03556433687330205,
+      "grad_norm": 4.581596683493078,
+      "learning_rate": 9.574468085106384e-07,
+      "loss": 2.2948,
+      "step": 90
+    },
+    {
+      "epoch": 0.0359594961718943,
+      "grad_norm": 4.661567248806833,
+      "learning_rate": 9.680851063829788e-07,
+      "loss": 2.2851,
+      "step": 91
+    },
+    {
+      "epoch": 0.03635465547048654,
+      "grad_norm": 4.1499647842616385,
+      "learning_rate": 9.787234042553193e-07,
+      "loss": 2.2358,
+      "step": 92
+    },
+    {
+      "epoch": 0.036749814769078785,
+      "grad_norm": 4.064054364491757,
+      "learning_rate": 9.893617021276597e-07,
+      "loss": 2.2021,
+      "step": 93
+    },
+    {
+      "epoch": 0.03714497406767103,
+      "grad_norm": 4.019497224170159,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.1558,
+      "step": 94
+    },
+    {
+      "epoch": 0.03754013336626327,
+      "grad_norm": 3.7739064116455254,
+      "learning_rate": 1.0106382978723404e-06,
+      "loss": 2.145,
+      "step": 95
+    },
+    {
+      "epoch": 0.03793529266485552,
+      "grad_norm": 3.587043554533132,
+      "learning_rate": 1.0212765957446809e-06,
+      "loss": 2.0649,
+      "step": 96
+    },
+    {
+      "epoch": 0.03833045196344777,
+      "grad_norm": 3.5425961325072732,
+      "learning_rate": 1.0319148936170213e-06,
+      "loss": 2.0494,
+      "step": 97
+    },
+    {
+      "epoch": 0.03872561126204001,
+      "grad_norm": 3.4793118680889914,
+      "learning_rate": 1.0425531914893618e-06,
+      "loss": 1.9564,
+      "step": 98
+    },
+    {
+      "epoch": 0.039120770560632255,
+      "grad_norm": 3.480928663874862,
+      "learning_rate": 1.0531914893617022e-06,
+      "loss": 1.9516,
+      "step": 99
+    },
+    {
+      "epoch": 0.0395159298592245,
+      "grad_norm": 3.4270001249424427,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 1.943,
+      "step": 100
+    },
+    {
+      "epoch": 0.03991108915781674,
+      "grad_norm": 3.142501786484995,
+      "learning_rate": 1.074468085106383e-06,
+      "loss": 1.858,
+      "step": 101
+    },
+    {
+      "epoch": 0.04030624845640899,
+      "grad_norm": 3.2413282091965376,
+      "learning_rate": 1.0851063829787236e-06,
+      "loss": 1.8773,
+      "step": 102
+    },
+    {
+      "epoch": 0.04070140775500124,
+      "grad_norm": 3.2180251637134223,
+      "learning_rate": 1.095744680851064e-06,
+      "loss": 1.7858,
+      "step": 103
+    },
+    {
+      "epoch": 0.04109656705359348,
+      "grad_norm": 3.0827158001171715,
+      "learning_rate": 1.1063829787234042e-06,
+      "loss": 1.8103,
+      "step": 104
+    },
+    {
+      "epoch": 0.041491726352185725,
+      "grad_norm": 3.0156067280676284,
+      "learning_rate": 1.1170212765957447e-06,
+      "loss": 1.7703,
+      "step": 105
+    },
+    {
+      "epoch": 0.04188688565077797,
+      "grad_norm": 3.1564655772648536,
+      "learning_rate": 1.1276595744680851e-06,
+      "loss": 1.7584,
+      "step": 106
+    },
+    {
+      "epoch": 0.04228204494937021,
+      "grad_norm": 2.8464417493034384,
+      "learning_rate": 1.1382978723404256e-06,
+      "loss": 1.7098,
+      "step": 107
+    },
+    {
+      "epoch": 0.04267720424796246,
+      "grad_norm": 2.979579920610495,
+      "learning_rate": 1.148936170212766e-06,
+      "loss": 1.6593,
+      "step": 108
+    },
+    {
+      "epoch": 0.04307236354655471,
+      "grad_norm": 2.7758712165539996,
+      "learning_rate": 1.1595744680851065e-06,
+      "loss": 1.6563,
+      "step": 109
+    },
+    {
+      "epoch": 0.04346752284514695,
+      "grad_norm": 2.5891620159779016,
+      "learning_rate": 1.170212765957447e-06,
+      "loss": 1.6248,
+      "step": 110
+    },
+    {
+      "epoch": 0.043862682143739194,
+      "grad_norm": 2.4867507899845473,
+      "learning_rate": 1.1808510638297874e-06,
+      "loss": 1.5999,
+      "step": 111
+    },
+    {
+      "epoch": 0.04425784144233144,
+      "grad_norm": 2.411241253335211,
+      "learning_rate": 1.1914893617021278e-06,
+      "loss": 1.6233,
+      "step": 112
+    },
+    {
+      "epoch": 0.04465300074092368,
+      "grad_norm": 2.675802892730842,
+      "learning_rate": 1.202127659574468e-06,
+      "loss": 1.5911,
+      "step": 113
+    },
+    {
+      "epoch": 0.04504816003951593,
+      "grad_norm": 2.4484713846419783,
+      "learning_rate": 1.2127659574468085e-06,
+      "loss": 1.552,
+      "step": 114
+    },
+    {
+      "epoch": 0.045443319338108176,
+      "grad_norm": 2.126895225481791,
+      "learning_rate": 1.223404255319149e-06,
+      "loss": 1.5116,
+      "step": 115
+    },
+    {
+      "epoch": 0.04583847863670042,
+      "grad_norm": 2.108035496774095,
+      "learning_rate": 1.2340425531914894e-06,
+      "loss": 1.5332,
+      "step": 116
+    },
+    {
+      "epoch": 0.046233637935292664,
+      "grad_norm": 2.170204737476494,
+      "learning_rate": 1.2446808510638299e-06,
+      "loss": 1.5099,
+      "step": 117
+    },
+    {
+      "epoch": 0.04662879723388491,
+      "grad_norm": 1.9897269791458945,
+      "learning_rate": 1.2553191489361701e-06,
+      "loss": 1.4747,
+      "step": 118
+    },
+    {
+      "epoch": 0.04702395653247716,
+      "grad_norm": 1.924882132114436,
+      "learning_rate": 1.2659574468085106e-06,
+      "loss": 1.4877,
+      "step": 119
+    },
+    {
+      "epoch": 0.0474191158310694,
+      "grad_norm": 2.1236023483825934,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 1.4593,
+      "step": 120
+    },
+    {
+      "epoch": 0.047814275129661646,
+      "grad_norm": 1.8555875106493023,
+      "learning_rate": 1.2872340425531915e-06,
+      "loss": 1.3795,
+      "step": 121
+    },
+    {
+      "epoch": 0.04820943442825389,
+      "grad_norm": 1.9400614964548064,
+      "learning_rate": 1.297872340425532e-06,
+      "loss": 1.404,
+      "step": 122
+    },
+    {
+      "epoch": 0.04860459372684613,
+      "grad_norm": 1.6699536554778585,
+      "learning_rate": 1.3085106382978724e-06,
+      "loss": 1.3696,
+      "step": 123
+    },
+    {
+      "epoch": 0.04899975302543838,
+      "grad_norm": 1.689620342600094,
+      "learning_rate": 1.3191489361702128e-06,
+      "loss": 1.344,
+      "step": 124
+    },
+    {
+      "epoch": 0.04939491232403063,
+      "grad_norm": 1.7509004836307809,
+      "learning_rate": 1.3297872340425533e-06,
+      "loss": 1.3462,
+      "step": 125
+    },
+    {
+      "epoch": 0.04979007162262287,
+      "grad_norm": 1.7664945294015308,
+      "learning_rate": 1.3404255319148937e-06,
+      "loss": 1.3477,
+      "step": 126
+    },
+    {
+      "epoch": 0.050185230921215115,
+      "grad_norm": 2.084264973245835,
+      "learning_rate": 1.3510638297872342e-06,
+      "loss": 1.3218,
+      "step": 127
+    },
+    {
+      "epoch": 0.05058039021980736,
+      "grad_norm": 1.5965337608217485,
+      "learning_rate": 1.3617021276595746e-06,
+      "loss": 1.3084,
+      "step": 128
+    },
+    {
+      "epoch": 0.0509755495183996,
+      "grad_norm": 1.619614456507732,
+      "learning_rate": 1.372340425531915e-06,
+      "loss": 1.2938,
+      "step": 129
+    },
+    {
+      "epoch": 0.05137070881699185,
+      "grad_norm": 1.8247931156223327,
+      "learning_rate": 1.3829787234042555e-06,
+      "loss": 1.2944,
+      "step": 130
+    },
+    {
+      "epoch": 0.0517658681155841,
+      "grad_norm": 1.580039976604606,
+      "learning_rate": 1.393617021276596e-06,
+      "loss": 1.2777,
+      "step": 131
+    },
+    {
+      "epoch": 0.05216102741417634,
+      "grad_norm": 1.509843143895781,
+      "learning_rate": 1.4042553191489364e-06,
+      "loss": 1.2841,
+      "step": 132
+    },
+    {
+      "epoch": 0.052556186712768585,
+      "grad_norm": 1.6756920394271753,
+      "learning_rate": 1.4148936170212769e-06,
+      "loss": 1.2419,
+      "step": 133
+    },
+    {
+      "epoch": 0.05295134601136083,
+      "grad_norm": 1.4623629043957789,
+      "learning_rate": 1.4255319148936173e-06,
+      "loss": 1.2077,
+      "step": 134
+    },
+    {
+      "epoch": 0.05334650530995307,
+      "grad_norm": 1.383626312375799,
+      "learning_rate": 1.4361702127659578e-06,
+      "loss": 1.2265,
+      "step": 135
+    },
+    {
+      "epoch": 0.05374166460854532,
+      "grad_norm": 1.7925919805177952,
+      "learning_rate": 1.4468085106382978e-06,
+      "loss": 1.2186,
+      "step": 136
+    },
+    {
+      "epoch": 0.05413682390713757,
+      "grad_norm": 1.3738854790506727,
+      "learning_rate": 1.4574468085106382e-06,
+      "loss": 1.2159,
+      "step": 137
+    },
+    {
+      "epoch": 0.05453198320572981,
+      "grad_norm": 1.4567831484040439,
+      "learning_rate": 1.4680851063829787e-06,
+      "loss": 1.1814,
+      "step": 138
+    },
+    {
+      "epoch": 0.054927142504322055,
+      "grad_norm": 1.3546122882454086,
+      "learning_rate": 1.4787234042553191e-06,
+      "loss": 1.1519,
+      "step": 139
+    },
+    {
+      "epoch": 0.0553223018029143,
+      "grad_norm": 1.3541293229496851,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 1.1625,
+      "step": 140
+    },
+    {
+      "epoch": 0.05571746110150654,
+      "grad_norm": 1.5328490619353512,
+      "learning_rate": 1.5e-06,
+      "loss": 1.1564,
+      "step": 141
+    },
+    {
+      "epoch": 0.05611262040009879,
+      "grad_norm": 1.2335978296031311,
+      "learning_rate": 1.5106382978723405e-06,
+      "loss": 1.1294,
+      "step": 142
+    },
+    {
+      "epoch": 0.05650777969869104,
+      "grad_norm": 1.281543201298622,
+      "learning_rate": 1.521276595744681e-06,
+      "loss": 1.1539,
+      "step": 143
+    },
+    {
+      "epoch": 0.05690293899728328,
+      "grad_norm": 1.1347832342130402,
+      "learning_rate": 1.5319148936170214e-06,
+      "loss": 1.1475,
+      "step": 144
+    },
+    {
+      "epoch": 0.057298098295875524,
+      "grad_norm": 1.272401514207382,
+      "learning_rate": 1.5425531914893618e-06,
+      "loss": 1.1062,
+      "step": 145
+    },
+    {
+      "epoch": 0.05769325759446777,
+      "grad_norm": 1.1113360919213993,
+      "learning_rate": 1.5531914893617023e-06,
+      "loss": 1.0904,
+      "step": 146
+    },
+    {
+      "epoch": 0.05808841689306001,
+      "grad_norm": 1.2143209476879704,
+      "learning_rate": 1.5638297872340427e-06,
+      "loss": 1.1221,
+      "step": 147
+    },
+    {
+      "epoch": 0.05848357619165226,
+      "grad_norm": 1.1975123411456277,
+      "learning_rate": 1.5744680851063832e-06,
+      "loss": 1.1134,
+      "step": 148
+    },
+    {
+      "epoch": 0.058878735490244506,
+      "grad_norm": 1.1752282768767688,
+      "learning_rate": 1.5851063829787236e-06,
+      "loss": 1.1073,
+      "step": 149
+    },
+    {
+      "epoch": 0.059273894788836747,
+      "grad_norm": 1.1221206874706884,
+      "learning_rate": 1.595744680851064e-06,
+      "loss": 1.0863,
+      "step": 150
+    },
+    {
+      "epoch": 0.059669054087428994,
+      "grad_norm": 1.0719072351198553,
+      "learning_rate": 1.6063829787234045e-06,
+      "loss": 1.087,
+      "step": 151
+    },
+    {
+      "epoch": 0.06006421338602124,
+      "grad_norm": 1.0618749015123967,
+      "learning_rate": 1.617021276595745e-06,
+      "loss": 1.0724,
+      "step": 152
+    },
+    {
+      "epoch": 0.06045937268461349,
+      "grad_norm": 1.289175259046802,
+      "learning_rate": 1.6276595744680854e-06,
+      "loss": 1.0913,
+      "step": 153
+    },
+    {
+      "epoch": 0.06085453198320573,
+      "grad_norm": 1.1077238693214941,
+      "learning_rate": 1.6382978723404255e-06,
+      "loss": 1.0408,
+      "step": 154
+    },
+    {
+      "epoch": 0.061249691281797976,
+      "grad_norm": 1.3121363094248382,
+      "learning_rate": 1.648936170212766e-06,
+      "loss": 1.0649,
+      "step": 155
+    },
+    {
+      "epoch": 0.06164485058039022,
+      "grad_norm": 1.0047858271903711,
+      "learning_rate": 1.6595744680851064e-06,
+      "loss": 1.0339,
+      "step": 156
+    },
+    {
+      "epoch": 0.06204000987898246,
+      "grad_norm": 1.222322248701329,
+      "learning_rate": 1.6702127659574468e-06,
+      "loss": 1.0367,
+      "step": 157
+    },
+    {
+      "epoch": 0.06243516917757471,
+      "grad_norm": 0.9790824508239406,
+      "learning_rate": 1.6808510638297873e-06,
+      "loss": 1.0221,
+      "step": 158
+    },
+    {
+      "epoch": 0.06283032847616696,
+      "grad_norm": 1.067740944312297,
+      "learning_rate": 1.6914893617021277e-06,
+      "loss": 1.0334,
+      "step": 159
+    },
+    {
+      "epoch": 0.0632254877747592,
+      "grad_norm": 0.896625566421198,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 1.0092,
+      "step": 160
+    },
+    {
+      "epoch": 0.06362064707335144,
+      "grad_norm": 0.9627618743848442,
+      "learning_rate": 1.7127659574468086e-06,
+      "loss": 1.0463,
+      "step": 161
+    },
+    {
+      "epoch": 0.06401580637194369,
+      "grad_norm": 1.0734207189542146,
+      "learning_rate": 1.723404255319149e-06,
+      "loss": 1.0392,
+      "step": 162
+    },
+    {
+      "epoch": 0.06441096567053593,
+      "grad_norm": 0.9349753594761937,
+      "learning_rate": 1.7340425531914895e-06,
+      "loss": 0.9892,
+      "step": 163
+    },
+    {
+      "epoch": 0.06480612496912817,
+      "grad_norm": 0.9980254538111385,
+      "learning_rate": 1.74468085106383e-06,
+      "loss": 0.9982,
+      "step": 164
+    },
+    {
+      "epoch": 0.06520128426772043,
+      "grad_norm": 1.0811827248370265,
+      "learning_rate": 1.7553191489361704e-06,
+      "loss": 1.0152,
+      "step": 165
+    },
+    {
+      "epoch": 0.06559644356631267,
+      "grad_norm": 0.9331981369264469,
+      "learning_rate": 1.7659574468085109e-06,
+      "loss": 0.98,
+      "step": 166
+    },
+    {
+      "epoch": 0.06599160286490491,
+      "grad_norm": 0.9206315543562278,
+      "learning_rate": 1.7765957446808513e-06,
+      "loss": 1.0212,
+      "step": 167
+    },
+    {
+      "epoch": 0.06638676216349716,
+      "grad_norm": 1.1010365241697404,
+      "learning_rate": 1.7872340425531918e-06,
+      "loss": 0.9817,
+      "step": 168
+    },
+    {
+      "epoch": 0.0667819214620894,
+      "grad_norm": 0.9070688042861361,
+      "learning_rate": 1.7978723404255322e-06,
+      "loss": 0.9974,
+      "step": 169
+    },
+    {
+      "epoch": 0.06717708076068166,
+      "grad_norm": 0.9259977243442546,
+      "learning_rate": 1.8085106382978727e-06,
+      "loss": 0.9979,
+      "step": 170
+    },
+    {
+      "epoch": 0.0675722400592739,
+      "grad_norm": 0.9016076582389854,
+      "learning_rate": 1.8191489361702131e-06,
+      "loss": 0.9735,
+      "step": 171
+    },
+    {
+      "epoch": 0.06796739935786614,
+      "grad_norm": 0.9114028424001606,
+      "learning_rate": 1.8297872340425531e-06,
+      "loss": 0.9806,
+      "step": 172
+    },
+    {
+      "epoch": 0.06836255865645839,
+      "grad_norm": 0.9002141700632048,
+      "learning_rate": 1.8404255319148936e-06,
+      "loss": 0.9922,
+      "step": 173
+    },
+    {
+      "epoch": 0.06875771795505063,
+      "grad_norm": 0.9028305681256402,
+      "learning_rate": 1.851063829787234e-06,
+      "loss": 0.9698,
+      "step": 174
+    },
+    {
+      "epoch": 0.06915287725364287,
+      "grad_norm": 0.8797232554167207,
+      "learning_rate": 1.8617021276595745e-06,
+      "loss": 0.9627,
+      "step": 175
+    },
+    {
+      "epoch": 0.06954803655223513,
+      "grad_norm": 0.8502350286865529,
+      "learning_rate": 1.872340425531915e-06,
+      "loss": 0.9788,
+      "step": 176
+    },
+    {
+      "epoch": 0.06994319585082737,
+      "grad_norm": 0.9168330903333247,
+      "learning_rate": 1.8829787234042554e-06,
+      "loss": 0.9566,
+      "step": 177
+    },
+    {
+      "epoch": 0.0703383551494196,
+      "grad_norm": 0.8633709197803816,
+      "learning_rate": 1.8936170212765958e-06,
+      "loss": 0.9403,
+      "step": 178
+    },
+    {
+      "epoch": 0.07073351444801186,
+      "grad_norm": 0.9428773416078605,
+      "learning_rate": 1.9042553191489363e-06,
+      "loss": 0.973,
+      "step": 179
+    },
+    {
+      "epoch": 0.0711286737466041,
+      "grad_norm": 0.8934209230197848,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 0.9531,
+      "step": 180
+    },
+    {
+      "epoch": 0.07152383304519634,
+      "grad_norm": 0.891758559309114,
+      "learning_rate": 1.925531914893617e-06,
+      "loss": 0.9379,
+      "step": 181
+    },
+    {
+      "epoch": 0.0719189923437886,
+      "grad_norm": 0.8289117111620914,
+      "learning_rate": 1.9361702127659576e-06,
+      "loss": 0.9594,
+      "step": 182
+    },
+    {
+      "epoch": 0.07231415164238084,
+      "grad_norm": 1.0538786469036632,
+      "learning_rate": 1.946808510638298e-06,
+      "loss": 0.9784,
+      "step": 183
+    },
+    {
+      "epoch": 0.07270931094097308,
+      "grad_norm": 0.8224300832472103,
+      "learning_rate": 1.9574468085106385e-06,
+      "loss": 0.9571,
+      "step": 184
+    },
+    {
+      "epoch": 0.07310447023956533,
+      "grad_norm": 0.8493893795427978,
+      "learning_rate": 1.968085106382979e-06,
+      "loss": 0.9436,
+      "step": 185
+    },
+    {
+      "epoch": 0.07349962953815757,
+      "grad_norm": 0.7674702264899148,
+      "learning_rate": 1.9787234042553194e-06,
+      "loss": 0.9391,
+      "step": 186
+    },
+    {
+      "epoch": 0.07389478883674981,
+      "grad_norm": 0.7756609871194469,
+      "learning_rate": 1.98936170212766e-06,
+      "loss": 0.9097,
+      "step": 187
+    },
+    {
+      "epoch": 0.07428994813534207,
+      "grad_norm": 0.8401464399948625,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.9084,
+      "step": 188
+    },
+    {
+      "epoch": 0.0746851074339343,
+      "grad_norm": 0.800464792845907,
+      "learning_rate": 2.0106382978723408e-06,
+      "loss": 0.955,
+      "step": 189
+    },
+    {
+      "epoch": 0.07508026673252655,
+      "grad_norm": 0.7963237467038586,
+      "learning_rate": 2.021276595744681e-06,
+      "loss": 0.9077,
+      "step": 190
+    },
+    {
+      "epoch": 0.0754754260311188,
+      "grad_norm": 0.8597621463693865,
+      "learning_rate": 2.0319148936170213e-06,
+      "loss": 0.9145,
+      "step": 191
+    },
+    {
+      "epoch": 0.07587058532971104,
+      "grad_norm": 0.8089654571622219,
+      "learning_rate": 2.0425531914893617e-06,
+      "loss": 0.896,
+      "step": 192
+    },
+    {
+      "epoch": 0.07626574462830328,
+      "grad_norm": 0.8373533262314947,
+      "learning_rate": 2.053191489361702e-06,
+      "loss": 0.9412,
+      "step": 193
+    },
+    {
+      "epoch": 0.07666090392689554,
+      "grad_norm": 0.7388812923284674,
+      "learning_rate": 2.0638297872340426e-06,
+      "loss": 0.9099,
+      "step": 194
+    },
+    {
+      "epoch": 0.07705606322548778,
+      "grad_norm": 0.7430489194577857,
+      "learning_rate": 2.074468085106383e-06,
+      "loss": 0.8875,
+      "step": 195
+    },
+    {
+      "epoch": 0.07745122252408002,
+      "grad_norm": 0.9205612468817582,
+      "learning_rate": 2.0851063829787235e-06,
+      "loss": 0.8999,
+      "step": 196
+    },
+    {
+      "epoch": 0.07784638182267227,
+      "grad_norm": 0.7981256350175032,
+      "learning_rate": 2.095744680851064e-06,
+      "loss": 0.906,
+      "step": 197
+    },
+    {
+      "epoch": 0.07824154112126451,
+      "grad_norm": 0.934795287215565,
+      "learning_rate": 2.1063829787234044e-06,
+      "loss": 0.8908,
+      "step": 198
+    },
+    {
+      "epoch": 0.07863670041985675,
+      "grad_norm": 0.8225561510286254,
+      "learning_rate": 2.117021276595745e-06,
+      "loss": 0.901,
+      "step": 199
+    },
+    {
+      "epoch": 0.079031859718449,
+      "grad_norm": 0.814892859258292,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.897,
+      "step": 200
+    },
+    {
+      "epoch": 0.07942701901704124,
+      "grad_norm": 0.9225501505784203,
+      "learning_rate": 2.1382978723404258e-06,
+      "loss": 0.9174,
+      "step": 201
+    },
+    {
+      "epoch": 0.07982217831563349,
+      "grad_norm": 0.7007925087166256,
+      "learning_rate": 2.148936170212766e-06,
+      "loss": 0.9093,
+      "step": 202
+    },
+    {
+      "epoch": 0.08021733761422574,
+      "grad_norm": 0.7613541888010653,
+      "learning_rate": 2.1595744680851067e-06,
+      "loss": 0.8826,
+      "step": 203
+    },
+    {
+      "epoch": 0.08061249691281798,
+      "grad_norm": 0.726817783841255,
+      "learning_rate": 2.170212765957447e-06,
+      "loss": 0.8738,
+      "step": 204
+    },
+    {
+      "epoch": 0.08100765621141022,
+      "grad_norm": 0.7037687083978766,
+      "learning_rate": 2.1808510638297876e-06,
+      "loss": 0.8989,
+      "step": 205
+    },
+    {
+      "epoch": 0.08140281551000247,
+      "grad_norm": 0.7363960547281249,
+      "learning_rate": 2.191489361702128e-06,
+      "loss": 0.8902,
+      "step": 206
+    },
+    {
+      "epoch": 0.08179797480859471,
+      "grad_norm": 0.7580280652504182,
+      "learning_rate": 2.2021276595744685e-06,
+      "loss": 0.9102,
+      "step": 207
+    },
+    {
+      "epoch": 0.08219313410718695,
+      "grad_norm": 0.7757175348061099,
+      "learning_rate": 2.2127659574468085e-06,
+      "loss": 0.8779,
+      "step": 208
+    },
+    {
+      "epoch": 0.08258829340577921,
+      "grad_norm": 0.7482271345832553,
+      "learning_rate": 2.223404255319149e-06,
+      "loss": 0.8865,
+      "step": 209
+    },
+    {
+      "epoch": 0.08298345270437145,
+      "grad_norm": 0.7398985474265629,
+      "learning_rate": 2.2340425531914894e-06,
+      "loss": 0.8801,
+      "step": 210
+    },
+    {
+      "epoch": 0.08337861200296369,
+      "grad_norm": 0.7323608656730606,
+      "learning_rate": 2.24468085106383e-06,
+      "loss": 0.9002,
+      "step": 211
+    },
+    {
+      "epoch": 0.08377377130155594,
+      "grad_norm": 0.8148667910929218,
+      "learning_rate": 2.2553191489361703e-06,
+      "loss": 0.8647,
+      "step": 212
+    },
+    {
+      "epoch": 0.08416893060014818,
+      "grad_norm": 0.7262927203241404,
+      "learning_rate": 2.2659574468085107e-06,
+      "loss": 0.8741,
+      "step": 213
+    },
+    {
+      "epoch": 0.08456408989874042,
+      "grad_norm": 0.8201240188887061,
+      "learning_rate": 2.276595744680851e-06,
+      "loss": 0.9041,
+      "step": 214
+    },
+    {
+      "epoch": 0.08495924919733268,
+      "grad_norm": 0.7492499292809485,
+      "learning_rate": 2.2872340425531916e-06,
+      "loss": 0.853,
+      "step": 215
+    },
+    {
+      "epoch": 0.08535440849592492,
+      "grad_norm": 0.7916341013954181,
+      "learning_rate": 2.297872340425532e-06,
+      "loss": 0.8871,
+      "step": 216
+    },
+    {
+      "epoch": 0.08574956779451716,
+      "grad_norm": 0.7520394996291526,
+      "learning_rate": 2.3085106382978725e-06,
+      "loss": 0.8562,
+      "step": 217
+    },
+    {
+      "epoch": 0.08614472709310941,
+      "grad_norm": 1.1281408516742564,
+      "learning_rate": 2.319148936170213e-06,
+      "loss": 0.8645,
+      "step": 218
+    },
+    {
+      "epoch": 0.08653988639170165,
+      "grad_norm": 0.8788573403694457,
+      "learning_rate": 2.3297872340425534e-06,
+      "loss": 0.8623,
+      "step": 219
+    },
+    {
+      "epoch": 0.0869350456902939,
+      "grad_norm": 0.867944142979568,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 0.8572,
+      "step": 220
+    },
+    {
+      "epoch": 0.08733020498888615,
+      "grad_norm": 0.7926936676598336,
+      "learning_rate": 2.3510638297872343e-06,
+      "loss": 0.8713,
+      "step": 221
+    },
+    {
+      "epoch": 0.08772536428747839,
+      "grad_norm": 0.76255556176971,
+      "learning_rate": 2.3617021276595748e-06,
+      "loss": 0.8494,
+      "step": 222
+    },
+    {
+      "epoch": 0.08812052358607063,
+      "grad_norm": 0.7600673269310736,
+      "learning_rate": 2.3723404255319152e-06,
+      "loss": 0.8482,
+      "step": 223
+    },
+    {
+      "epoch": 0.08851568288466288,
+      "grad_norm": 0.7112814516322407,
+      "learning_rate": 2.3829787234042557e-06,
+      "loss": 0.8533,
+      "step": 224
+    },
+    {
+      "epoch": 0.08891084218325512,
+      "grad_norm": 0.7036773729937497,
+      "learning_rate": 2.393617021276596e-06,
+      "loss": 0.847,
+      "step": 225
+    },
+    {
+      "epoch": 0.08930600148184736,
+      "grad_norm": 0.6590954821985295,
+      "learning_rate": 2.404255319148936e-06,
+      "loss": 0.8561,
+      "step": 226
+    },
+    {
+      "epoch": 0.08970116078043962,
+      "grad_norm": 0.7847232653490934,
+      "learning_rate": 2.4148936170212766e-06,
+      "loss": 0.8536,
+      "step": 227
+    },
+    {
+      "epoch": 0.09009632007903186,
+      "grad_norm": 0.6509851787949374,
+      "learning_rate": 2.425531914893617e-06,
+      "loss": 0.85,
+      "step": 228
+    },
+    {
+      "epoch": 0.0904914793776241,
+      "grad_norm": 0.648445158504499,
+      "learning_rate": 2.4361702127659575e-06,
+      "loss": 0.8012,
+      "step": 229
+    },
+    {
+      "epoch": 0.09088663867621635,
+      "grad_norm": 0.736497309836382,
+      "learning_rate": 2.446808510638298e-06,
+      "loss": 0.8639,
+      "step": 230
+    },
+    {
+      "epoch": 0.09128179797480859,
+      "grad_norm": 0.645898512592661,
+      "learning_rate": 2.4574468085106384e-06,
+      "loss": 0.8243,
+      "step": 231
+    },
+    {
+      "epoch": 0.09167695727340083,
+      "grad_norm": 0.67481419189577,
+      "learning_rate": 2.468085106382979e-06,
+      "loss": 0.8682,
+      "step": 232
+    },
+    {
+      "epoch": 0.09207211657199309,
+      "grad_norm": 0.7664263353306855,
+      "learning_rate": 2.4787234042553193e-06,
+      "loss": 0.8623,
+      "step": 233
+    },
+    {
+      "epoch": 0.09246727587058533,
+      "grad_norm": 0.6372201971565544,
+      "learning_rate": 2.4893617021276598e-06,
+      "loss": 0.8153,
+      "step": 234
+    },
+    {
+      "epoch": 0.09286243516917757,
+      "grad_norm": 0.7305559386754572,
+      "learning_rate": 2.5e-06,
+      "loss": 0.85,
+      "step": 235
+    },
+    {
+      "epoch": 0.09325759446776982,
+      "grad_norm": 0.8258720994595878,
+      "learning_rate": 2.5106382978723402e-06,
+      "loss": 0.8588,
+      "step": 236
+    },
+    {
+      "epoch": 0.09365275376636206,
+      "grad_norm": 0.762878968342536,
+      "learning_rate": 2.521276595744681e-06,
+      "loss": 0.8606,
+      "step": 237
+    },
+    {
+      "epoch": 0.09404791306495432,
+      "grad_norm": 0.7142096057042978,
+      "learning_rate": 2.531914893617021e-06,
+      "loss": 0.8337,
+      "step": 238
+    },
+    {
+      "epoch": 0.09444307236354656,
+      "grad_norm": 0.6622420389390998,
+      "learning_rate": 2.542553191489362e-06,
+      "loss": 0.8475,
+      "step": 239
+    },
+    {
+      "epoch": 0.0948382316621388,
+      "grad_norm": 0.7610481119780552,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.8637,
+      "step": 240
+    },
+    {
+      "epoch": 0.09523339096073105,
+      "grad_norm": 0.8093851995765269,
+      "learning_rate": 2.563829787234043e-06,
+      "loss": 0.8521,
+      "step": 241
+    },
+    {
+      "epoch": 0.09562855025932329,
+      "grad_norm": 0.6922332669648779,
+      "learning_rate": 2.574468085106383e-06,
+      "loss": 0.8318,
+      "step": 242
+    },
+    {
+      "epoch": 0.09602370955791553,
+      "grad_norm": 0.7693846197783168,
+      "learning_rate": 2.585106382978724e-06,
+      "loss": 0.8344,
+      "step": 243
+    },
+    {
+      "epoch": 0.09641886885650779,
+      "grad_norm": 0.827949068079929,
+      "learning_rate": 2.595744680851064e-06,
+      "loss": 0.8486,
+      "step": 244
+    },
+    {
+      "epoch": 0.09681402815510003,
+      "grad_norm": 0.6650613101391282,
+      "learning_rate": 2.6063829787234047e-06,
+      "loss": 0.8562,
+      "step": 245
+    },
+    {
+      "epoch": 0.09720918745369227,
+      "grad_norm": 0.8278434657863171,
+      "learning_rate": 2.6170212765957447e-06,
+      "loss": 0.8416,
+      "step": 246
+    },
+    {
+      "epoch": 0.09760434675228452,
+      "grad_norm": 0.7938328023685507,
+      "learning_rate": 2.6276595744680856e-06,
+      "loss": 0.8311,
+      "step": 247
+    },
+    {
+      "epoch": 0.09799950605087676,
+      "grad_norm": 0.6143610103260296,
+      "learning_rate": 2.6382978723404256e-06,
+      "loss": 0.8254,
+      "step": 248
+    },
+    {
+      "epoch": 0.098394665349469,
+      "grad_norm": 0.7619119256406987,
+      "learning_rate": 2.6489361702127665e-06,
+      "loss": 0.839,
+      "step": 249
+    },
+    {
+      "epoch": 0.09878982464806126,
+      "grad_norm": 0.7986196919326185,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 0.8455,
+      "step": 250
+    },
+    {
+      "epoch": 0.0991849839466535,
+      "grad_norm": 0.6717154185196351,
+      "learning_rate": 2.6702127659574474e-06,
+      "loss": 0.8336,
+      "step": 251
+    },
+    {
+      "epoch": 0.09958014324524574,
+      "grad_norm": 0.7767129776930844,
+      "learning_rate": 2.6808510638297874e-06,
+      "loss": 0.8145,
+      "step": 252
+    },
+    {
+      "epoch": 0.09997530254383799,
+      "grad_norm": 0.6439894121807275,
+      "learning_rate": 2.6914893617021283e-06,
+      "loss": 0.8111,
+      "step": 253
+    },
+    {
+      "epoch": 0.10037046184243023,
+      "grad_norm": 0.607679779354355,
+      "learning_rate": 2.7021276595744683e-06,
+      "loss": 0.8291,
+      "step": 254
+    },
+    {
+      "epoch": 0.10076562114102247,
+      "grad_norm": 0.7464563131206438,
+      "learning_rate": 2.7127659574468084e-06,
+      "loss": 0.8017,
+      "step": 255
+    },
+    {
+      "epoch": 0.10116078043961473,
+      "grad_norm": 0.7095071407490194,
+      "learning_rate": 2.7234042553191492e-06,
+      "loss": 0.8497,
+      "step": 256
+    },
+    {
+      "epoch": 0.10155593973820697,
+      "grad_norm": 0.6459106720980804,
+      "learning_rate": 2.7340425531914893e-06,
+      "loss": 0.8405,
+      "step": 257
+    },
+    {
+      "epoch": 0.1019510990367992,
+      "grad_norm": 0.7520801305481896,
+      "learning_rate": 2.74468085106383e-06,
+      "loss": 0.8074,
+      "step": 258
+    },
+    {
+      "epoch": 0.10234625833539146,
+      "grad_norm": 0.7352594748784674,
+      "learning_rate": 2.75531914893617e-06,
+      "loss": 0.8185,
+      "step": 259
+    },
+    {
+      "epoch": 0.1027414176339837,
+      "grad_norm": 0.7017112677861721,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 0.8266,
+      "step": 260
+    },
+    {
+      "epoch": 0.10313657693257594,
+      "grad_norm": 0.761095345544042,
+      "learning_rate": 2.776595744680851e-06,
+      "loss": 0.7901,
+      "step": 261
+    },
+    {
+      "epoch": 0.1035317362311682,
+      "grad_norm": 0.7630856502083389,
+      "learning_rate": 2.787234042553192e-06,
+      "loss": 0.8038,
+      "step": 262
+    },
+    {
+      "epoch": 0.10392689552976044,
+      "grad_norm": 0.7132952976277227,
+      "learning_rate": 2.797872340425532e-06,
+      "loss": 0.8157,
+      "step": 263
+    },
+    {
+      "epoch": 0.10432205482835268,
+      "grad_norm": 0.900249895843058,
+      "learning_rate": 2.808510638297873e-06,
+      "loss": 0.8025,
+      "step": 264
+    },
+    {
+      "epoch": 0.10471721412694493,
+      "grad_norm": 0.6280451084365367,
+      "learning_rate": 2.819148936170213e-06,
+      "loss": 0.7995,
+      "step": 265
+    },
+    {
+      "epoch": 0.10511237342553717,
+      "grad_norm": 0.7227719924319213,
+      "learning_rate": 2.8297872340425537e-06,
+      "loss": 0.8428,
+      "step": 266
+    },
+    {
+      "epoch": 0.10550753272412941,
+      "grad_norm": 0.7918390149442235,
+      "learning_rate": 2.8404255319148938e-06,
+      "loss": 0.7932,
+      "step": 267
+    },
+    {
+      "epoch": 0.10590269202272166,
+      "grad_norm": 0.6180037290493834,
+      "learning_rate": 2.8510638297872346e-06,
+      "loss": 0.8234,
+      "step": 268
+    },
+    {
+      "epoch": 0.1062978513213139,
+      "grad_norm": 0.6527196921600206,
+      "learning_rate": 2.8617021276595747e-06,
+      "loss": 0.8047,
+      "step": 269
+    },
+    {
+      "epoch": 0.10669301061990615,
+      "grad_norm": 1.0085142638197158,
+      "learning_rate": 2.8723404255319155e-06,
+      "loss": 0.8164,
+      "step": 270
+    },
+    {
+      "epoch": 0.1070881699184984,
+      "grad_norm": 0.9812147195881408,
+      "learning_rate": 2.8829787234042556e-06,
+      "loss": 0.7965,
+      "step": 271
+    },
+    {
+      "epoch": 0.10748332921709064,
+      "grad_norm": 1.0448400314614967,
+      "learning_rate": 2.8936170212765956e-06,
+      "loss": 0.8107,
+      "step": 272
+    },
+    {
+      "epoch": 0.10787848851568288,
+      "grad_norm": 0.6984652148632842,
+      "learning_rate": 2.9042553191489365e-06,
+      "loss": 0.7865,
+      "step": 273
+    },
+    {
+      "epoch": 0.10827364781427513,
+      "grad_norm": 0.615065445787656,
+      "learning_rate": 2.9148936170212765e-06,
+      "loss": 0.7847,
+      "step": 274
+    },
+    {
+      "epoch": 0.10866880711286737,
+      "grad_norm": 0.8697531900111486,
+      "learning_rate": 2.9255319148936174e-06,
+      "loss": 0.8037,
+      "step": 275
+    },
+    {
+      "epoch": 0.10906396641145961,
+      "grad_norm": 0.6197520935871196,
+      "learning_rate": 2.9361702127659574e-06,
+      "loss": 0.8151,
+      "step": 276
+    },
+    {
+      "epoch": 0.10945912571005187,
+      "grad_norm": 0.8060011131830561,
+      "learning_rate": 2.9468085106382983e-06,
+      "loss": 0.7873,
+      "step": 277
+    },
+    {
+      "epoch": 0.10985428500864411,
+      "grad_norm": 0.6756833173956306,
+      "learning_rate": 2.9574468085106383e-06,
+      "loss": 0.8293,
+      "step": 278
+    },
+    {
+      "epoch": 0.11024944430723635,
+      "grad_norm": 0.6099851620784188,
+      "learning_rate": 2.968085106382979e-06,
+      "loss": 0.8241,
+      "step": 279
+    },
+    {
+      "epoch": 0.1106446036058286,
+      "grad_norm": 0.7283434437348605,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 0.7936,
+      "step": 280
+    },
+    {
+      "epoch": 0.11103976290442084,
+      "grad_norm": 0.6267015873151124,
+      "learning_rate": 2.98936170212766e-06,
+      "loss": 0.7962,
+      "step": 281
+    },
+    {
+      "epoch": 0.11143492220301308,
+      "grad_norm": 0.6116326742433275,
+      "learning_rate": 3e-06,
+      "loss": 0.8187,
+      "step": 282
+    },
+    {
+      "epoch": 0.11183008150160534,
+      "grad_norm": 0.7049851223926784,
+      "learning_rate": 3.010638297872341e-06,
+      "loss": 0.7831,
+      "step": 283
+    },
+    {
+      "epoch": 0.11222524080019758,
+      "grad_norm": 0.6257674496318532,
+      "learning_rate": 3.021276595744681e-06,
+      "loss": 0.7781,
+      "step": 284
+    },
+    {
+      "epoch": 0.11262040009878982,
+      "grad_norm": 0.6190650007298025,
+      "learning_rate": 3.031914893617022e-06,
+      "loss": 0.7847,
+      "step": 285
+    },
+    {
+      "epoch": 0.11301555939738207,
+      "grad_norm": 0.6089050965059565,
+      "learning_rate": 3.042553191489362e-06,
+      "loss": 0.7922,
+      "step": 286
+    },
+    {
+      "epoch": 0.11341071869597431,
+      "grad_norm": 0.6557347874370106,
+      "learning_rate": 3.0531914893617027e-06,
+      "loss": 0.7963,
+      "step": 287
+    },
+    {
+      "epoch": 0.11380587799456655,
+      "grad_norm": 0.605070002885156,
+      "learning_rate": 3.0638297872340428e-06,
+      "loss": 0.8013,
+      "step": 288
+    },
+    {
+      "epoch": 0.11420103729315881,
+      "grad_norm": 0.6311761000607505,
+      "learning_rate": 3.0744680851063836e-06,
+      "loss": 0.8005,
+      "step": 289
+    },
+    {
+      "epoch": 0.11459619659175105,
+      "grad_norm": 0.6053509981952528,
+      "learning_rate": 3.0851063829787237e-06,
+      "loss": 0.7998,
+      "step": 290
+    },
+    {
+      "epoch": 0.11499135589034329,
+      "grad_norm": 0.6235992193861427,
+      "learning_rate": 3.0957446808510637e-06,
+      "loss": 0.7804,
+      "step": 291
+    },
+    {
+      "epoch": 0.11538651518893554,
+      "grad_norm": 0.7433766960086654,
+      "learning_rate": 3.1063829787234046e-06,
+      "loss": 0.7867,
+      "step": 292
+    },
+    {
+      "epoch": 0.11578167448752778,
+      "grad_norm": 1.272372503947218,
+      "learning_rate": 3.1170212765957446e-06,
+      "loss": 0.7929,
+      "step": 293
+    },
+    {
+      "epoch": 0.11617683378612002,
+      "grad_norm": 0.8186585325087611,
+      "learning_rate": 3.1276595744680855e-06,
+      "loss": 0.7854,
+      "step": 294
+    },
+    {
+      "epoch": 0.11657199308471228,
+      "grad_norm": 0.6739199686171928,
+      "learning_rate": 3.1382978723404255e-06,
+      "loss": 0.805,
+      "step": 295
+    },
+    {
+      "epoch": 0.11696715238330452,
+      "grad_norm": 0.6306549835549222,
+      "learning_rate": 3.1489361702127664e-06,
+      "loss": 0.786,
+      "step": 296
+    },
+    {
+      "epoch": 0.11736231168189676,
+      "grad_norm": 0.7237829920169022,
+      "learning_rate": 3.1595744680851064e-06,
+      "loss": 0.7921,
+      "step": 297
+    },
+    {
+      "epoch": 0.11775747098048901,
+      "grad_norm": 0.6263525455488399,
+      "learning_rate": 3.1702127659574473e-06,
+      "loss": 0.7794,
+      "step": 298
+    },
+    {
+      "epoch": 0.11815263027908125,
+      "grad_norm": 0.6260917117859388,
+      "learning_rate": 3.1808510638297873e-06,
+      "loss": 0.7821,
+      "step": 299
+    },
+    {
+      "epoch": 0.11854778957767349,
+      "grad_norm": 0.6670782296472709,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 0.7889,
+      "step": 300
+    },
+    {
+      "epoch": 0.11894294887626575,
+      "grad_norm": 0.7021296376843668,
+      "learning_rate": 3.202127659574468e-06,
+      "loss": 0.7879,
+      "step": 301
+    },
+    {
+      "epoch": 0.11933810817485799,
+      "grad_norm": 0.6436206115473436,
+      "learning_rate": 3.212765957446809e-06,
+      "loss": 0.7463,
+      "step": 302
+    },
+    {
+      "epoch": 0.11973326747345023,
+      "grad_norm": 0.6581559963727542,
+      "learning_rate": 3.223404255319149e-06,
+      "loss": 0.7833,
+      "step": 303
+    },
+    {
+      "epoch": 0.12012842677204248,
+      "grad_norm": 0.6085378327556474,
+      "learning_rate": 3.23404255319149e-06,
+      "loss": 0.7761,
+      "step": 304
+    },
+    {
+      "epoch": 0.12052358607063472,
+      "grad_norm": 0.6332080516803438,
+      "learning_rate": 3.24468085106383e-06,
+      "loss": 0.7687,
+      "step": 305
+    },
+    {
+      "epoch": 0.12091874536922698,
+      "grad_norm": 0.5853426452014024,
+      "learning_rate": 3.255319148936171e-06,
+      "loss": 0.7878,
+      "step": 306
+    },
+    {
+      "epoch": 0.12131390466781922,
+      "grad_norm": 0.6223748575158505,
+      "learning_rate": 3.265957446808511e-06,
+      "loss": 0.7739,
+      "step": 307
+    },
+    {
+      "epoch": 0.12170906396641146,
+      "grad_norm": 0.6981138823006418,
+      "learning_rate": 3.276595744680851e-06,
+      "loss": 0.7695,
+      "step": 308
+    },
+    {
+      "epoch": 0.12210422326500371,
+      "grad_norm": 0.6663161701946723,
+      "learning_rate": 3.287234042553192e-06,
+      "loss": 0.8053,
+      "step": 309
+    },
+    {
+      "epoch": 0.12249938256359595,
+      "grad_norm": 0.5681680114078884,
+      "learning_rate": 3.297872340425532e-06,
+      "loss": 0.7752,
+      "step": 310
+    },
+    {
+      "epoch": 0.12289454186218819,
+      "grad_norm": 0.5858906412963043,
+      "learning_rate": 3.3085106382978727e-06,
+      "loss": 0.7916,
+      "step": 311
+    },
+    {
+      "epoch": 0.12328970116078045,
+      "grad_norm": 0.6534395144260531,
+      "learning_rate": 3.3191489361702127e-06,
+      "loss": 0.7912,
+      "step": 312
+    },
+    {
+      "epoch": 0.12368486045937269,
+      "grad_norm": 0.586198108795242,
+      "learning_rate": 3.3297872340425536e-06,
+      "loss": 0.7813,
+      "step": 313
+    },
+    {
+      "epoch": 0.12408001975796493,
+      "grad_norm": 0.647904927155932,
+      "learning_rate": 3.3404255319148936e-06,
+      "loss": 0.7936,
+      "step": 314
+    },
+    {
+      "epoch": 0.12447517905655718,
+      "grad_norm": 0.58805476345561,
+      "learning_rate": 3.3510638297872345e-06,
+      "loss": 0.7973,
+      "step": 315
+    },
+    {
+      "epoch": 0.12487033835514942,
+      "grad_norm": 0.5947946152867843,
+      "learning_rate": 3.3617021276595745e-06,
+      "loss": 0.7632,
+      "step": 316
+    },
+    {
+      "epoch": 0.12526549765374168,
+      "grad_norm": 0.6193828087421168,
+      "learning_rate": 3.3723404255319154e-06,
+      "loss": 0.7848,
+      "step": 317
+    },
+    {
+      "epoch": 0.12566065695233392,
+      "grad_norm": 0.6258831363958954,
+      "learning_rate": 3.3829787234042554e-06,
+      "loss": 0.7678,
+      "step": 318
+    },
+    {
+      "epoch": 0.12605581625092616,
+      "grad_norm": 0.5747152594546233,
+      "learning_rate": 3.3936170212765963e-06,
+      "loss": 0.7817,
+      "step": 319
+    },
+    {
+      "epoch": 0.1264509755495184,
+      "grad_norm": 0.731674333424822,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 0.7643,
+      "step": 320
+    },
+    {
+      "epoch": 0.12684613484811064,
+      "grad_norm": 0.6529010150169795,
+      "learning_rate": 3.414893617021277e-06,
+      "loss": 0.7667,
+      "step": 321
+    },
+    {
+      "epoch": 0.12724129414670288,
+      "grad_norm": 0.7014365604795955,
+      "learning_rate": 3.4255319148936172e-06,
+      "loss": 0.7787,
+      "step": 322
+    },
+    {
+      "epoch": 0.12763645344529514,
+      "grad_norm": 0.5834784417943027,
+      "learning_rate": 3.436170212765958e-06,
+      "loss": 0.779,
+      "step": 323
+    },
+    {
+      "epoch": 0.12803161274388739,
+      "grad_norm": 0.7742370051056543,
+      "learning_rate": 3.446808510638298e-06,
+      "loss": 0.7733,
+      "step": 324
+    },
+    {
+      "epoch": 0.12842677204247963,
+      "grad_norm": 0.5865654907094333,
+      "learning_rate": 3.457446808510639e-06,
+      "loss": 0.7456,
+      "step": 325
+    },
+    {
+      "epoch": 0.12882193134107187,
+      "grad_norm": 0.682149646741674,
+      "learning_rate": 3.468085106382979e-06,
+      "loss": 0.7994,
+      "step": 326
+    },
+    {
+      "epoch": 0.1292170906396641,
+      "grad_norm": 0.6701640838897713,
+      "learning_rate": 3.478723404255319e-06,
+      "loss": 0.7757,
+      "step": 327
+    },
+    {
+      "epoch": 0.12961224993825635,
+      "grad_norm": 0.6614941446237965,
+      "learning_rate": 3.48936170212766e-06,
+      "loss": 0.7725,
+      "step": 328
+    },
+    {
+      "epoch": 0.13000740923684861,
+      "grad_norm": 0.6095415461958933,
+      "learning_rate": 3.5e-06,
+      "loss": 0.7653,
+      "step": 329
+    },
+    {
+      "epoch": 0.13040256853544085,
+      "grad_norm": 0.6190745963316048,
+      "learning_rate": 3.510638297872341e-06,
+      "loss": 0.7805,
+      "step": 330
+    },
+    {
+      "epoch": 0.1307977278340331,
+      "grad_norm": 0.6147443464940033,
+      "learning_rate": 3.521276595744681e-06,
+      "loss": 0.7763,
+      "step": 331
+    },
+    {
+      "epoch": 0.13119288713262534,
+      "grad_norm": 0.5866635120263384,
+      "learning_rate": 3.5319148936170217e-06,
+      "loss": 0.7698,
+      "step": 332
+    },
+    {
+      "epoch": 0.13158804643121758,
+      "grad_norm": 0.6186148281750584,
+      "learning_rate": 3.5425531914893617e-06,
+      "loss": 0.7771,
+      "step": 333
+    },
+    {
+      "epoch": 0.13198320572980982,
+      "grad_norm": 0.6620495658950574,
+      "learning_rate": 3.5531914893617026e-06,
+      "loss": 0.7622,
+      "step": 334
+    },
+    {
+      "epoch": 0.13237836502840208,
+      "grad_norm": 0.6402051904915239,
+      "learning_rate": 3.5638297872340426e-06,
+      "loss": 0.7734,
+      "step": 335
+    },
+    {
+      "epoch": 0.13277352432699432,
+      "grad_norm": 0.6309802907790648,
+      "learning_rate": 3.5744680851063835e-06,
+      "loss": 0.7665,
+      "step": 336
+    },
+    {
+      "epoch": 0.13316868362558656,
+      "grad_norm": 0.6408680224306361,
+      "learning_rate": 3.5851063829787235e-06,
+      "loss": 0.7686,
+      "step": 337
+    },
+    {
+      "epoch": 0.1335638429241788,
+      "grad_norm": 0.6160995866186305,
+      "learning_rate": 3.5957446808510644e-06,
+      "loss": 0.7373,
+      "step": 338
+    },
+    {
+      "epoch": 0.13395900222277105,
+      "grad_norm": 0.6840190156980204,
+      "learning_rate": 3.6063829787234044e-06,
+      "loss": 0.7648,
+      "step": 339
+    },
+    {
+      "epoch": 0.1343541615213633,
+      "grad_norm": 0.6581837356659743,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 0.7531,
+      "step": 340
+    },
+    {
+      "epoch": 0.13474932081995555,
+      "grad_norm": 0.7148764476179186,
+      "learning_rate": 3.6276595744680853e-06,
+      "loss": 0.7819,
+      "step": 341
+    },
+    {
+      "epoch": 0.1351444801185478,
+      "grad_norm": 0.6600259458370601,
+      "learning_rate": 3.6382978723404262e-06,
+      "loss": 0.7842,
+      "step": 342
+    },
+    {
+      "epoch": 0.13553963941714003,
+      "grad_norm": 0.6126556516156064,
+      "learning_rate": 3.6489361702127662e-06,
+      "loss": 0.7639,
+      "step": 343
+    },
+    {
+      "epoch": 0.13593479871573227,
+      "grad_norm": 0.6672986551005967,
+      "learning_rate": 3.6595744680851063e-06,
+      "loss": 0.7592,
+      "step": 344
+    },
+    {
+      "epoch": 0.13632995801432451,
+      "grad_norm": 0.6223043791589471,
+      "learning_rate": 3.670212765957447e-06,
+      "loss": 0.7564,
+      "step": 345
+    },
+    {
+      "epoch": 0.13672511731291678,
+      "grad_norm": 0.8110409579655677,
+      "learning_rate": 3.680851063829787e-06,
+      "loss": 0.7651,
+      "step": 346
+    },
+    {
+      "epoch": 0.13712027661150902,
+      "grad_norm": 0.6581211301407619,
+      "learning_rate": 3.691489361702128e-06,
+      "loss": 0.7611,
+      "step": 347
+    },
+    {
+      "epoch": 0.13751543591010126,
+      "grad_norm": 0.6744339312700035,
+      "learning_rate": 3.702127659574468e-06,
+      "loss": 0.7613,
+      "step": 348
+    },
+    {
+      "epoch": 0.1379105952086935,
+      "grad_norm": 0.605969858747699,
+      "learning_rate": 3.712765957446809e-06,
+      "loss": 0.7847,
+      "step": 349
+    },
+    {
+      "epoch": 0.13830575450728574,
+      "grad_norm": 0.6796481089227777,
+      "learning_rate": 3.723404255319149e-06,
+      "loss": 0.7698,
+      "step": 350
+    },
+    {
+      "epoch": 0.13870091380587798,
+      "grad_norm": 0.6445216487073453,
+      "learning_rate": 3.73404255319149e-06,
+      "loss": 0.7452,
+      "step": 351
+    },
+    {
+      "epoch": 0.13909607310447025,
+      "grad_norm": 0.5848343681445489,
+      "learning_rate": 3.74468085106383e-06,
+      "loss": 0.7532,
+      "step": 352
+    },
+    {
+      "epoch": 0.1394912324030625,
+      "grad_norm": 0.686482496408356,
+      "learning_rate": 3.7553191489361707e-06,
+      "loss": 0.745,
+      "step": 353
+    },
+    {
+      "epoch": 0.13988639170165473,
+      "grad_norm": 0.6484269721297492,
+      "learning_rate": 3.7659574468085108e-06,
+      "loss": 0.7408,
+      "step": 354
+    },
+    {
+      "epoch": 0.14028155100024697,
+      "grad_norm": 0.5809696958364097,
+      "learning_rate": 3.7765957446808516e-06,
+      "loss": 0.7471,
+      "step": 355
+    },
+    {
+      "epoch": 0.1406767102988392,
+      "grad_norm": 0.6182389029108604,
+      "learning_rate": 3.7872340425531917e-06,
+      "loss": 0.7654,
+      "step": 356
+    },
+    {
+      "epoch": 0.14107186959743145,
+      "grad_norm": 0.6745002737038325,
+      "learning_rate": 3.7978723404255325e-06,
+      "loss": 0.7604,
+      "step": 357
+    },
+    {
+      "epoch": 0.14146702889602372,
+      "grad_norm": 0.7457299483196483,
+      "learning_rate": 3.8085106382978726e-06,
+      "loss": 0.771,
+      "step": 358
+    },
+    {
+      "epoch": 0.14186218819461596,
+      "grad_norm": 0.626223558343114,
+      "learning_rate": 3.819148936170213e-06,
+      "loss": 0.7856,
+      "step": 359
+    },
+    {
+      "epoch": 0.1422573474932082,
+      "grad_norm": 0.6297264134042032,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 0.748,
+      "step": 360
+    },
+    {
+      "epoch": 0.14265250679180044,
+      "grad_norm": 0.5951480073799157,
+      "learning_rate": 3.840425531914894e-06,
+      "loss": 0.7782,
+      "step": 361
+    },
+    {
+      "epoch": 0.14304766609039268,
+      "grad_norm": 0.5886768213260241,
+      "learning_rate": 3.851063829787234e-06,
+      "loss": 0.7455,
+      "step": 362
+    },
+    {
+      "epoch": 0.14344282538898492,
+      "grad_norm": 0.673440091158271,
+      "learning_rate": 3.861702127659575e-06,
+      "loss": 0.7419,
+      "step": 363
+    },
+    {
+      "epoch": 0.1438379846875772,
+      "grad_norm": 0.6400671142527239,
+      "learning_rate": 3.872340425531915e-06,
+      "loss": 0.7648,
+      "step": 364
+    },
+    {
+      "epoch": 0.14423314398616943,
+      "grad_norm": 0.6378848588398233,
+      "learning_rate": 3.882978723404256e-06,
+      "loss": 0.764,
+      "step": 365
+    },
+    {
+      "epoch": 0.14462830328476167,
+      "grad_norm": 0.5855533735447821,
+      "learning_rate": 3.893617021276596e-06,
+      "loss": 0.755,
+      "step": 366
+    },
+    {
+      "epoch": 0.1450234625833539,
+      "grad_norm": 0.5988399094182794,
+      "learning_rate": 3.904255319148937e-06,
+      "loss": 0.7717,
+      "step": 367
+    },
+    {
+      "epoch": 0.14541862188194615,
+      "grad_norm": 0.5909689175761067,
+      "learning_rate": 3.914893617021277e-06,
+      "loss": 0.7305,
+      "step": 368
+    },
+    {
+      "epoch": 0.1458137811805384,
+      "grad_norm": 0.6091829357239786,
+      "learning_rate": 3.9255319148936175e-06,
+      "loss": 0.7582,
+      "step": 369
+    },
+    {
+      "epoch": 0.14620894047913066,
+      "grad_norm": 0.663665703959751,
+      "learning_rate": 3.936170212765958e-06,
+      "loss": 0.7634,
+      "step": 370
+    },
+    {
+      "epoch": 0.1466040997777229,
+      "grad_norm": 0.6131032136468707,
+      "learning_rate": 3.946808510638298e-06,
+      "loss": 0.7352,
+      "step": 371
+    },
+    {
+      "epoch": 0.14699925907631514,
+      "grad_norm": 0.6438458775790014,
+      "learning_rate": 3.957446808510639e-06,
+      "loss": 0.7558,
+      "step": 372
+    },
+    {
+      "epoch": 0.14739441837490738,
+      "grad_norm": 0.6221026846089203,
+      "learning_rate": 3.968085106382979e-06,
+      "loss": 0.7483,
+      "step": 373
+    },
+    {
+      "epoch": 0.14778957767349962,
+      "grad_norm": 0.718970949273493,
+      "learning_rate": 3.97872340425532e-06,
+      "loss": 0.7582,
+      "step": 374
+    },
+    {
+      "epoch": 0.14818473697209186,
+      "grad_norm": 0.7632958625164173,
+      "learning_rate": 3.98936170212766e-06,
+      "loss": 0.7487,
+      "step": 375
+    },
+    {
+      "epoch": 0.14857989627068413,
+      "grad_norm": 0.6975538226598186,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7548,
+      "step": 376
+    },
+    {
+      "epoch": 0.14897505556927637,
+      "grad_norm": 1.011413729830943,
+      "learning_rate": 4.010638297872341e-06,
+      "loss": 0.7355,
+      "step": 377
+    },
+    {
+      "epoch": 0.1493702148678686,
+      "grad_norm": 0.6337564143926464,
+      "learning_rate": 4.0212765957446816e-06,
+      "loss": 0.7519,
+      "step": 378
+    },
+    {
+      "epoch": 0.14976537416646085,
+      "grad_norm": 0.668831153796517,
+      "learning_rate": 4.031914893617022e-06,
+      "loss": 0.7624,
+      "step": 379
+    },
+    {
+      "epoch": 0.1501605334650531,
+      "grad_norm": 0.6436926511990773,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 0.7474,
+      "step": 380
+    },
+    {
+      "epoch": 0.15055569276364533,
+      "grad_norm": 0.6272202096101458,
+      "learning_rate": 4.053191489361702e-06,
+      "loss": 0.7602,
+      "step": 381
+    },
+    {
+      "epoch": 0.1509508520622376,
+      "grad_norm": 0.6212797540013587,
+      "learning_rate": 4.0638297872340425e-06,
+      "loss": 0.7772,
+      "step": 382
+    },
+    {
+      "epoch": 0.15134601136082984,
+      "grad_norm": 0.6289946916321798,
+      "learning_rate": 4.074468085106383e-06,
+      "loss": 0.7689,
+      "step": 383
+    },
+    {
+      "epoch": 0.15174117065942208,
+      "grad_norm": 0.6380191041059411,
+      "learning_rate": 4.085106382978723e-06,
+      "loss": 0.7572,
+      "step": 384
+    },
+    {
+      "epoch": 0.15213632995801432,
+      "grad_norm": 0.5779795688082269,
+      "learning_rate": 4.095744680851064e-06,
+      "loss": 0.7546,
+      "step": 385
+    },
+    {
+      "epoch": 0.15253148925660656,
+      "grad_norm": 0.7842430533418746,
+      "learning_rate": 4.106382978723404e-06,
+      "loss": 0.7751,
+      "step": 386
+    },
+    {
+      "epoch": 0.1529266485551988,
+      "grad_norm": 0.6729893760136558,
+      "learning_rate": 4.117021276595745e-06,
+      "loss": 0.7455,
+      "step": 387
+    },
+    {
+      "epoch": 0.15332180785379107,
+      "grad_norm": 0.5882912759438818,
+      "learning_rate": 4.127659574468085e-06,
+      "loss": 0.7696,
+      "step": 388
+    },
+    {
+      "epoch": 0.1537169671523833,
+      "grad_norm": 0.6227877790472663,
+      "learning_rate": 4.138297872340426e-06,
+      "loss": 0.7608,
+      "step": 389
+    },
+    {
+      "epoch": 0.15411212645097555,
+      "grad_norm": 0.6619432056903835,
+      "learning_rate": 4.148936170212766e-06,
+      "loss": 0.7473,
+      "step": 390
+    },
+    {
+      "epoch": 0.1545072857495678,
+      "grad_norm": 0.628215165860546,
+      "learning_rate": 4.1595744680851066e-06,
+      "loss": 0.7489,
+      "step": 391
+    },
+    {
+      "epoch": 0.15490244504816003,
+      "grad_norm": 0.5841047919949067,
+      "learning_rate": 4.170212765957447e-06,
+      "loss": 0.7447,
+      "step": 392
+    },
+    {
+      "epoch": 0.15529760434675227,
+      "grad_norm": 0.5825058321729848,
+      "learning_rate": 4.1808510638297875e-06,
+      "loss": 0.7518,
+      "step": 393
+    },
+    {
+      "epoch": 0.15569276364534454,
+      "grad_norm": 0.5989180700052267,
+      "learning_rate": 4.191489361702128e-06,
+      "loss": 0.7463,
+      "step": 394
+    },
+    {
+      "epoch": 0.15608792294393678,
+      "grad_norm": 0.5485054555998863,
+      "learning_rate": 4.202127659574468e-06,
+      "loss": 0.7254,
+      "step": 395
+    },
+    {
+      "epoch": 0.15648308224252902,
+      "grad_norm": 0.646716189095031,
+      "learning_rate": 4.212765957446809e-06,
+      "loss": 0.7583,
+      "step": 396
+    },
+    {
+      "epoch": 0.15687824154112126,
+      "grad_norm": 0.6866055568311499,
+      "learning_rate": 4.223404255319149e-06,
+      "loss": 0.7372,
+      "step": 397
+    },
+    {
+      "epoch": 0.1572734008397135,
+      "grad_norm": 0.7140385676935366,
+      "learning_rate": 4.23404255319149e-06,
+      "loss": 0.739,
+      "step": 398
+    },
+    {
+      "epoch": 0.15766856013830574,
+      "grad_norm": 0.643023278201698,
+      "learning_rate": 4.24468085106383e-06,
+      "loss": 0.7522,
+      "step": 399
+    },
+    {
+      "epoch": 0.158063719436898,
+      "grad_norm": 0.6312967460011237,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 0.7676,
+      "step": 400
+    },
+    {
+      "epoch": 0.15845887873549025,
+      "grad_norm": 0.6382138110089172,
+      "learning_rate": 4.265957446808511e-06,
+      "loss": 0.7629,
+      "step": 401
+    },
+    {
+      "epoch": 0.1588540380340825,
+      "grad_norm": 0.6332459333263118,
+      "learning_rate": 4.2765957446808515e-06,
+      "loss": 0.7311,
+      "step": 402
+    },
+    {
+      "epoch": 0.15924919733267473,
+      "grad_norm": 0.6412487468227186,
+      "learning_rate": 4.287234042553192e-06,
+      "loss": 0.7567,
+      "step": 403
+    },
+    {
+      "epoch": 0.15964435663126697,
+      "grad_norm": 0.7012222995798256,
+      "learning_rate": 4.297872340425532e-06,
+      "loss": 0.7436,
+      "step": 404
+    },
+    {
+      "epoch": 0.1600395159298592,
+      "grad_norm": 0.6181464501586335,
+      "learning_rate": 4.308510638297873e-06,
+      "loss": 0.7298,
+      "step": 405
+    },
+    {
+      "epoch": 0.16043467522845148,
+      "grad_norm": 0.6207970587414706,
+      "learning_rate": 4.319148936170213e-06,
+      "loss": 0.7376,
+      "step": 406
+    },
+    {
+      "epoch": 0.16082983452704372,
+      "grad_norm": 0.6106494887539161,
+      "learning_rate": 4.329787234042554e-06,
+      "loss": 0.7613,
+      "step": 407
+    },
+    {
+      "epoch": 0.16122499382563596,
+      "grad_norm": 0.6219536291754221,
+      "learning_rate": 4.340425531914894e-06,
+      "loss": 0.7397,
+      "step": 408
+    },
+    {
+      "epoch": 0.1616201531242282,
+      "grad_norm": 0.6668879597496847,
+      "learning_rate": 4.351063829787235e-06,
+      "loss": 0.7469,
+      "step": 409
+    },
+    {
+      "epoch": 0.16201531242282044,
+      "grad_norm": 0.6048928125977155,
+      "learning_rate": 4.361702127659575e-06,
+      "loss": 0.7184,
+      "step": 410
+    },
+    {
+      "epoch": 0.1624104717214127,
+      "grad_norm": 0.6204217080705178,
+      "learning_rate": 4.3723404255319156e-06,
+      "loss": 0.7433,
+      "step": 411
+    },
+    {
+      "epoch": 0.16280563102000495,
+      "grad_norm": 0.6071870434011859,
+      "learning_rate": 4.382978723404256e-06,
+      "loss": 0.7522,
+      "step": 412
+    },
+    {
+      "epoch": 0.1632007903185972,
+      "grad_norm": 0.6179635375795429,
+      "learning_rate": 4.3936170212765965e-06,
+      "loss": 0.7521,
+      "step": 413
+    },
+    {
+      "epoch": 0.16359594961718943,
+      "grad_norm": 0.8065353676281373,
+      "learning_rate": 4.404255319148937e-06,
+      "loss": 0.7371,
+      "step": 414
+    },
+    {
+      "epoch": 0.16399110891578167,
+      "grad_norm": 0.7223783073823268,
+      "learning_rate": 4.414893617021277e-06,
+      "loss": 0.749,
+      "step": 415
+    },
+    {
+      "epoch": 0.1643862682143739,
+      "grad_norm": 0.6418436008798059,
+      "learning_rate": 4.425531914893617e-06,
+      "loss": 0.7074,
+      "step": 416
+    },
+    {
+      "epoch": 0.16478142751296618,
+      "grad_norm": 0.649001786754692,
+      "learning_rate": 4.436170212765957e-06,
+      "loss": 0.7409,
+      "step": 417
+    },
+    {
+      "epoch": 0.16517658681155842,
+      "grad_norm": 0.6488082324591959,
+      "learning_rate": 4.446808510638298e-06,
+      "loss": 0.7294,
+      "step": 418
+    },
+    {
+      "epoch": 0.16557174611015066,
+      "grad_norm": 0.6815680138497775,
+      "learning_rate": 4.457446808510638e-06,
+      "loss": 0.7516,
+      "step": 419
+    },
+    {
+      "epoch": 0.1659669054087429,
+      "grad_norm": 0.627178650684025,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 0.7462,
+      "step": 420
+    },
+    {
+      "epoch": 0.16636206470733514,
+      "grad_norm": 0.6847077790393925,
+      "learning_rate": 4.478723404255319e-06,
+      "loss": 0.7642,
+      "step": 421
+    },
+    {
+      "epoch": 0.16675722400592738,
+      "grad_norm": 0.6677432032398735,
+      "learning_rate": 4.48936170212766e-06,
+      "loss": 0.7376,
+      "step": 422
+    },
+    {
+      "epoch": 0.16715238330451965,
+      "grad_norm": 0.6622301887078457,
+      "learning_rate": 4.5e-06,
+      "loss": 0.7281,
+      "step": 423
+    },
+    {
+      "epoch": 0.1675475426031119,
+      "grad_norm": 0.7032852662429785,
+      "learning_rate": 4.5106382978723406e-06,
+      "loss": 0.7295,
+      "step": 424
+    },
+    {
+      "epoch": 0.16794270190170413,
+      "grad_norm": 0.9182307524223992,
+      "learning_rate": 4.521276595744681e-06,
+      "loss": 0.7251,
+      "step": 425
+    },
+    {
+      "epoch": 0.16833786120029637,
+      "grad_norm": 0.6549474458937882,
+      "learning_rate": 4.5319148936170215e-06,
+      "loss": 0.7393,
+      "step": 426
+    },
+    {
+      "epoch": 0.1687330204988886,
+      "grad_norm": 0.6260937965574087,
+      "learning_rate": 4.542553191489362e-06,
+      "loss": 0.7466,
+      "step": 427
+    },
+    {
+      "epoch": 0.16912817979748085,
+      "grad_norm": 0.6098749910337613,
+      "learning_rate": 4.553191489361702e-06,
+      "loss": 0.7429,
+      "step": 428
+    },
+    {
+      "epoch": 0.16952333909607312,
+      "grad_norm": 0.6258409309068056,
+      "learning_rate": 4.563829787234043e-06,
+      "loss": 0.7437,
+      "step": 429
+    },
+    {
+      "epoch": 0.16991849839466536,
+      "grad_norm": 0.6998727053444407,
+      "learning_rate": 4.574468085106383e-06,
+      "loss": 0.7349,
+      "step": 430
+    },
+    {
+      "epoch": 0.1703136576932576,
+      "grad_norm": 0.635839700174999,
+      "learning_rate": 4.585106382978724e-06,
+      "loss": 0.742,
+      "step": 431
+    },
+    {
+      "epoch": 0.17070881699184984,
+      "grad_norm": 0.6881295298901814,
+      "learning_rate": 4.595744680851064e-06,
+      "loss": 0.7635,
+      "step": 432
+    },
+    {
+      "epoch": 0.17110397629044208,
+      "grad_norm": 0.6138576739147359,
+      "learning_rate": 4.606382978723405e-06,
+      "loss": 0.7215,
+      "step": 433
+    },
+    {
+      "epoch": 0.17149913558903432,
+      "grad_norm": 0.6786451341607185,
+      "learning_rate": 4.617021276595745e-06,
+      "loss": 0.746,
+      "step": 434
+    },
+    {
+      "epoch": 0.1718942948876266,
+      "grad_norm": 0.6085640309923145,
+      "learning_rate": 4.6276595744680855e-06,
+      "loss": 0.7201,
+      "step": 435
+    },
+    {
+      "epoch": 0.17228945418621883,
+      "grad_norm": 0.6543509049521442,
+      "learning_rate": 4.638297872340426e-06,
+      "loss": 0.7376,
+      "step": 436
+    },
+    {
+      "epoch": 0.17268461348481107,
+      "grad_norm": 0.6281733392631065,
+      "learning_rate": 4.648936170212766e-06,
+      "loss": 0.7386,
+      "step": 437
+    },
+    {
+      "epoch": 0.1730797727834033,
+      "grad_norm": 0.5766921427888596,
+      "learning_rate": 4.659574468085107e-06,
+      "loss": 0.7344,
+      "step": 438
+    },
+    {
+      "epoch": 0.17347493208199555,
+      "grad_norm": 0.6238350960095946,
+      "learning_rate": 4.670212765957447e-06,
+      "loss": 0.7401,
+      "step": 439
+    },
+    {
+      "epoch": 0.1738700913805878,
+      "grad_norm": 0.6308341725603104,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 0.7311,
+      "step": 440
+    },
+    {
+      "epoch": 0.17426525067918006,
+      "grad_norm": 0.6418577607256782,
+      "learning_rate": 4.691489361702128e-06,
+      "loss": 0.7234,
+      "step": 441
+    },
+    {
+      "epoch": 0.1746604099777723,
+      "grad_norm": 0.5854790033107359,
+      "learning_rate": 4.702127659574469e-06,
+      "loss": 0.7323,
+      "step": 442
+    },
+    {
+      "epoch": 0.17505556927636454,
+      "grad_norm": 0.6823496981208785,
+      "learning_rate": 4.712765957446809e-06,
+      "loss": 0.7323,
+      "step": 443
+    },
+    {
+      "epoch": 0.17545072857495678,
+      "grad_norm": 0.6175278963352572,
+      "learning_rate": 4.7234042553191496e-06,
+      "loss": 0.7147,
+      "step": 444
+    },
+    {
+      "epoch": 0.17584588787354902,
+      "grad_norm": 0.7360888497887278,
+      "learning_rate": 4.73404255319149e-06,
+      "loss": 0.7348,
+      "step": 445
+    },
+    {
+      "epoch": 0.17624104717214126,
+      "grad_norm": 0.6230101684392804,
+      "learning_rate": 4.7446808510638305e-06,
+      "loss": 0.7119,
+      "step": 446
+    },
+    {
+      "epoch": 0.17663620647073353,
+      "grad_norm": 0.6554804476316878,
+      "learning_rate": 4.755319148936171e-06,
+      "loss": 0.7294,
+      "step": 447
+    },
+    {
+      "epoch": 0.17703136576932577,
+      "grad_norm": 0.8564206801667128,
+      "learning_rate": 4.765957446808511e-06,
+      "loss": 0.7423,
+      "step": 448
+    },
+    {
+      "epoch": 0.177426525067918,
+      "grad_norm": 1.3629336168086166,
+      "learning_rate": 4.776595744680852e-06,
+      "loss": 0.7484,
+      "step": 449
+    },
+    {
+      "epoch": 0.17782168436651025,
+      "grad_norm": 0.7691195276316329,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 0.7097,
+      "step": 450
+    },
+    {
+      "epoch": 0.1782168436651025,
+      "grad_norm": 0.6194923429303832,
+      "learning_rate": 4.797872340425533e-06,
+      "loss": 0.7285,
+      "step": 451
+    },
+    {
+      "epoch": 0.17861200296369473,
+      "grad_norm": 0.673233699921991,
+      "learning_rate": 4.808510638297872e-06,
+      "loss": 0.7278,
+      "step": 452
+    },
+    {
+      "epoch": 0.179007162262287,
+      "grad_norm": 0.6766336054718974,
+      "learning_rate": 4.819148936170213e-06,
+      "loss": 0.7356,
+      "step": 453
+    },
+    {
+      "epoch": 0.17940232156087924,
+      "grad_norm": 0.57599098500566,
+      "learning_rate": 4.829787234042553e-06,
+      "loss": 0.7115,
+      "step": 454
+    },
+    {
+      "epoch": 0.17979748085947148,
+      "grad_norm": 0.8080267118445753,
+      "learning_rate": 4.840425531914894e-06,
+      "loss": 0.7534,
+      "step": 455
+    },
+    {
+      "epoch": 0.18019264015806372,
+      "grad_norm": 0.603929876457928,
+      "learning_rate": 4.851063829787234e-06,
+      "loss": 0.7105,
+      "step": 456
+    },
+    {
+      "epoch": 0.18058779945665596,
+      "grad_norm": 0.643703372655632,
+      "learning_rate": 4.8617021276595746e-06,
+      "loss": 0.7279,
+      "step": 457
+    },
+    {
+      "epoch": 0.1809829587552482,
+      "grad_norm": 0.8198396590432799,
+      "learning_rate": 4.872340425531915e-06,
+      "loss": 0.7198,
+      "step": 458
+    },
+    {
+      "epoch": 0.18137811805384046,
+      "grad_norm": 0.65771945253315,
+      "learning_rate": 4.8829787234042555e-06,
+      "loss": 0.7113,
+      "step": 459
+    },
+    {
+      "epoch": 0.1817732773524327,
+      "grad_norm": 0.6963361768452864,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 0.74,
+      "step": 460
+    },
+    {
+      "epoch": 0.18216843665102495,
+      "grad_norm": 0.6261965671131124,
+      "learning_rate": 4.904255319148936e-06,
+      "loss": 0.7126,
+      "step": 461
+    },
+    {
+      "epoch": 0.18256359594961719,
+      "grad_norm": 0.6148512179719411,
+      "learning_rate": 4.914893617021277e-06,
+      "loss": 0.726,
+      "step": 462
+    },
+    {
+      "epoch": 0.18295875524820943,
+      "grad_norm": 0.676252763119563,
+      "learning_rate": 4.925531914893617e-06,
+      "loss": 0.73,
+      "step": 463
+    },
+    {
+      "epoch": 0.18335391454680167,
+      "grad_norm": 0.7917693042214963,
+      "learning_rate": 4.936170212765958e-06,
+      "loss": 0.745,
+      "step": 464
+    },
+    {
+      "epoch": 0.18374907384539393,
+      "grad_norm": 0.623944439707609,
+      "learning_rate": 4.946808510638298e-06,
+      "loss": 0.7178,
+      "step": 465
+    },
+    {
+      "epoch": 0.18414423314398617,
+      "grad_norm": 0.7595206103625429,
+      "learning_rate": 4.957446808510639e-06,
+      "loss": 0.7511,
+      "step": 466
+    },
+    {
+      "epoch": 0.18453939244257841,
+      "grad_norm": 0.6236195124431673,
+      "learning_rate": 4.968085106382979e-06,
+      "loss": 0.7443,
+      "step": 467
+    },
+    {
+      "epoch": 0.18493455174117066,
+      "grad_norm": 0.6344774337895079,
+      "learning_rate": 4.9787234042553195e-06,
+      "loss": 0.7502,
+      "step": 468
+    },
+    {
+      "epoch": 0.1853297110397629,
+      "grad_norm": 0.6180463187517735,
+      "learning_rate": 4.98936170212766e-06,
+      "loss": 0.7089,
+      "step": 469
+    },
+    {
+      "epoch": 0.18572487033835514,
+      "grad_norm": 0.6437262535137148,
+      "learning_rate": 5e-06,
+      "loss": 0.7118,
+      "step": 470
+    },
+    {
+      "epoch": 0.1861200296369474,
+      "grad_norm": 0.8247696667741479,
+      "learning_rate": 4.999999995090759e-06,
+      "loss": 0.7448,
+      "step": 471
+    },
+    {
+      "epoch": 0.18651518893553964,
+      "grad_norm": 0.6821659327496776,
+      "learning_rate": 4.9999999803630365e-06,
+      "loss": 0.7096,
+      "step": 472
+    },
+    {
+      "epoch": 0.18691034823413188,
+      "grad_norm": 0.6937300518785938,
+      "learning_rate": 4.999999955816832e-06,
+      "loss": 0.7505,
+      "step": 473
+    },
+    {
+      "epoch": 0.18730550753272412,
+      "grad_norm": 0.64493393853509,
+      "learning_rate": 4.999999921452146e-06,
+      "loss": 0.7246,
+      "step": 474
+    },
+    {
+      "epoch": 0.18770066683131637,
+      "grad_norm": 0.699335103430049,
+      "learning_rate": 4.999999877268977e-06,
+      "loss": 0.7362,
+      "step": 475
+    },
+    {
+      "epoch": 0.18809582612990863,
+      "grad_norm": 0.6958675091974922,
+      "learning_rate": 4.999999823267328e-06,
+      "loss": 0.7425,
+      "step": 476
+    },
+    {
+      "epoch": 0.18849098542850087,
+      "grad_norm": 0.6561148703214672,
+      "learning_rate": 4.999999759447197e-06,
+      "loss": 0.7193,
+      "step": 477
+    },
+    {
+      "epoch": 0.1888861447270931,
+      "grad_norm": 0.7332659964684728,
+      "learning_rate": 4.999999685808585e-06,
+      "loss": 0.7367,
+      "step": 478
+    },
+    {
+      "epoch": 0.18928130402568535,
+      "grad_norm": 0.6434935616449193,
+      "learning_rate": 4.999999602351493e-06,
+      "loss": 0.7241,
+      "step": 479
+    },
+    {
+      "epoch": 0.1896764633242776,
+      "grad_norm": 0.7085133055543303,
+      "learning_rate": 4.99999950907592e-06,
+      "loss": 0.7294,
+      "step": 480
+    },
+    {
+      "epoch": 0.19007162262286983,
+      "grad_norm": 0.6037858004944545,
+      "learning_rate": 4.999999405981868e-06,
+      "loss": 0.7169,
+      "step": 481
+    },
+    {
+      "epoch": 0.1904667819214621,
+      "grad_norm": 0.6353596283356293,
+      "learning_rate": 4.999999293069335e-06,
+      "loss": 0.7255,
+      "step": 482
+    },
+    {
+      "epoch": 0.19086194122005434,
+      "grad_norm": 0.6011323842788836,
+      "learning_rate": 4.999999170338324e-06,
+      "loss": 0.7098,
+      "step": 483
+    },
+    {
+      "epoch": 0.19125710051864658,
+      "grad_norm": 0.685601090106948,
+      "learning_rate": 4.999999037788834e-06,
+      "loss": 0.7333,
+      "step": 484
+    },
+    {
+      "epoch": 0.19165225981723882,
+      "grad_norm": 0.6377038979338291,
+      "learning_rate": 4.999998895420866e-06,
+      "loss": 0.7257,
+      "step": 485
+    },
+    {
+      "epoch": 0.19204741911583106,
+      "grad_norm": 0.6192918261508922,
+      "learning_rate": 4.9999987432344195e-06,
+      "loss": 0.7275,
+      "step": 486
+    },
+    {
+      "epoch": 0.1924425784144233,
+      "grad_norm": 0.6720726722815289,
+      "learning_rate": 4.999998581229497e-06,
+      "loss": 0.7209,
+      "step": 487
+    },
+    {
+      "epoch": 0.19283773771301557,
+      "grad_norm": 0.5958748564691253,
+      "learning_rate": 4.999998409406098e-06,
+      "loss": 0.7182,
+      "step": 488
+    },
+    {
+      "epoch": 0.1932328970116078,
+      "grad_norm": 0.6523018913824231,
+      "learning_rate": 4.999998227764223e-06,
+      "loss": 0.7223,
+      "step": 489
+    },
+    {
+      "epoch": 0.19362805631020005,
+      "grad_norm": 0.6969026955713506,
+      "learning_rate": 4.999998036303873e-06,
+      "loss": 0.7017,
+      "step": 490
+    },
+    {
+      "epoch": 0.1940232156087923,
+      "grad_norm": 0.6199690598086954,
+      "learning_rate": 4.999997835025049e-06,
+      "loss": 0.724,
+      "step": 491
+    },
+    {
+      "epoch": 0.19441837490738453,
+      "grad_norm": 0.6899050040288978,
+      "learning_rate": 4.999997623927752e-06,
+      "loss": 0.702,
+      "step": 492
+    },
+    {
+      "epoch": 0.19481353420597677,
+      "grad_norm": 0.6023461037422734,
+      "learning_rate": 4.999997403011982e-06,
+      "loss": 0.7192,
+      "step": 493
+    },
+    {
+      "epoch": 0.19520869350456904,
+      "grad_norm": 0.7949013705151751,
+      "learning_rate": 4.9999971722777395e-06,
+      "loss": 0.7214,
+      "step": 494
+    },
+    {
+      "epoch": 0.19560385280316128,
+      "grad_norm": 0.8406697228012115,
+      "learning_rate": 4.9999969317250276e-06,
+      "loss": 0.7511,
+      "step": 495
+    },
+    {
+      "epoch": 0.19599901210175352,
+      "grad_norm": 0.6165479711982665,
+      "learning_rate": 4.999996681353845e-06,
+      "loss": 0.7129,
+      "step": 496
+    },
+    {
+      "epoch": 0.19639417140034576,
+      "grad_norm": 0.6365598818327172,
+      "learning_rate": 4.999996421164194e-06,
+      "loss": 0.724,
+      "step": 497
+    },
+    {
+      "epoch": 0.196789330698938,
+      "grad_norm": 0.5804862939924182,
+      "learning_rate": 4.999996151156075e-06,
+      "loss": 0.7208,
+      "step": 498
+    },
+    {
+      "epoch": 0.19718448999753024,
+      "grad_norm": 0.8043069867513684,
+      "learning_rate": 4.9999958713294886e-06,
+      "loss": 0.7316,
+      "step": 499
+    },
+    {
+      "epoch": 0.1975796492961225,
+      "grad_norm": 0.7039727180630397,
+      "learning_rate": 4.999995581684437e-06,
+      "loss": 0.7273,
+      "step": 500
+    },
+    {
+      "epoch": 0.19797480859471475,
+      "grad_norm": 0.5796311427949719,
+      "learning_rate": 4.999995282220921e-06,
+      "loss": 0.7353,
+      "step": 501
+    },
+    {
+      "epoch": 0.198369967893307,
+      "grad_norm": 0.640816004878865,
+      "learning_rate": 4.999994972938941e-06,
+      "loss": 0.718,
+      "step": 502
+    },
+    {
+      "epoch": 0.19876512719189923,
+      "grad_norm": 0.6625901203049049,
+      "learning_rate": 4.9999946538384995e-06,
+      "loss": 0.717,
+      "step": 503
+    },
+    {
+      "epoch": 0.19916028649049147,
+      "grad_norm": 0.731900514664347,
+      "learning_rate": 4.999994324919596e-06,
+      "loss": 0.7154,
+      "step": 504
+    },
+    {
+      "epoch": 0.1995554457890837,
+      "grad_norm": 0.6358220991451766,
+      "learning_rate": 4.9999939861822345e-06,
+      "loss": 0.7135,
+      "step": 505
+    },
+    {
+      "epoch": 0.19995060508767598,
+      "grad_norm": 0.6772742599590151,
+      "learning_rate": 4.999993637626413e-06,
+      "loss": 0.7142,
+      "step": 506
+    },
+    {
+      "epoch": 0.20034576438626822,
+      "grad_norm": 0.6203815751394517,
+      "learning_rate": 4.999993279252136e-06,
+      "loss": 0.7094,
+      "step": 507
+    },
+    {
+      "epoch": 0.20074092368486046,
+      "grad_norm": 0.6159319745437677,
+      "learning_rate": 4.999992911059402e-06,
+      "loss": 0.7097,
+      "step": 508
+    },
+    {
+      "epoch": 0.2011360829834527,
+      "grad_norm": 0.6394572431227349,
+      "learning_rate": 4.999992533048216e-06,
+      "loss": 0.7389,
+      "step": 509
+    },
+    {
+      "epoch": 0.20153124228204494,
+      "grad_norm": 0.5730980921754206,
+      "learning_rate": 4.999992145218576e-06,
+      "loss": 0.7148,
+      "step": 510
+    },
+    {
+      "epoch": 0.20192640158063718,
+      "grad_norm": 0.642558910052563,
+      "learning_rate": 4.999991747570485e-06,
+      "loss": 0.6986,
+      "step": 511
+    },
+    {
+      "epoch": 0.20232156087922945,
+      "grad_norm": 0.7198501180731144,
+      "learning_rate": 4.999991340103944e-06,
+      "loss": 0.7169,
+      "step": 512
+    },
+    {
+      "epoch": 0.2027167201778217,
+      "grad_norm": 0.5448656307534819,
+      "learning_rate": 4.999990922818957e-06,
+      "loss": 0.6977,
+      "step": 513
+    },
+    {
+      "epoch": 0.20311187947641393,
+      "grad_norm": 0.6205152271493748,
+      "learning_rate": 4.999990495715522e-06,
+      "loss": 0.7289,
+      "step": 514
+    },
+    {
+      "epoch": 0.20350703877500617,
+      "grad_norm": 0.5899827817607862,
+      "learning_rate": 4.9999900587936425e-06,
+      "loss": 0.7269,
+      "step": 515
+    },
+    {
+      "epoch": 0.2039021980735984,
+      "grad_norm": 0.6379914033363828,
+      "learning_rate": 4.999989612053321e-06,
+      "loss": 0.7031,
+      "step": 516
+    },
+    {
+      "epoch": 0.20429735737219065,
+      "grad_norm": 0.63396345510458,
+      "learning_rate": 4.999989155494558e-06,
+      "loss": 0.7113,
+      "step": 517
+    },
+    {
+      "epoch": 0.20469251667078292,
+      "grad_norm": 0.5769043304193019,
+      "learning_rate": 4.999988689117355e-06,
+      "loss": 0.7371,
+      "step": 518
+    },
+    {
+      "epoch": 0.20508767596937516,
+      "grad_norm": 0.6508775630628256,
+      "learning_rate": 4.999988212921715e-06,
+      "loss": 0.6937,
+      "step": 519
+    },
+    {
+      "epoch": 0.2054828352679674,
+      "grad_norm": 0.6868129416605413,
+      "learning_rate": 4.99998772690764e-06,
+      "loss": 0.7383,
+      "step": 520
+    },
+    {
+      "epoch": 0.20587799456655964,
+      "grad_norm": 0.5389127646966064,
+      "learning_rate": 4.9999872310751305e-06,
+      "loss": 0.6993,
+      "step": 521
+    },
+    {
+      "epoch": 0.20627315386515188,
+      "grad_norm": 0.6060273410792214,
+      "learning_rate": 4.9999867254241894e-06,
+      "loss": 0.7118,
+      "step": 522
+    },
+    {
+      "epoch": 0.20666831316374412,
+      "grad_norm": 0.5800631862306052,
+      "learning_rate": 4.9999862099548186e-06,
+      "loss": 0.7323,
+      "step": 523
+    },
+    {
+      "epoch": 0.2070634724623364,
+      "grad_norm": 0.7534794475732837,
+      "learning_rate": 4.9999856846670195e-06,
+      "loss": 0.7228,
+      "step": 524
+    },
+    {
+      "epoch": 0.20745863176092863,
+      "grad_norm": 0.5812327339287102,
+      "learning_rate": 4.999985149560796e-06,
+      "loss": 0.7072,
+      "step": 525
+    },
+    {
+      "epoch": 0.20785379105952087,
+      "grad_norm": 0.5768618184120087,
+      "learning_rate": 4.999984604636148e-06,
+      "loss": 0.7088,
+      "step": 526
+    },
+    {
+      "epoch": 0.2082489503581131,
+      "grad_norm": 0.6399784431138891,
+      "learning_rate": 4.999984049893078e-06,
+      "loss": 0.7149,
+      "step": 527
+    },
+    {
+      "epoch": 0.20864410965670535,
+      "grad_norm": 0.5623651744286807,
+      "learning_rate": 4.99998348533159e-06,
+      "loss": 0.7226,
+      "step": 528
+    },
+    {
+      "epoch": 0.2090392689552976,
+      "grad_norm": 0.6963865673449156,
+      "learning_rate": 4.999982910951684e-06,
+      "loss": 0.7233,
+      "step": 529
+    },
+    {
+      "epoch": 0.20943442825388986,
+      "grad_norm": 0.6478334973673012,
+      "learning_rate": 4.999982326753363e-06,
+      "loss": 0.7262,
+      "step": 530
+    },
+    {
+      "epoch": 0.2098295875524821,
+      "grad_norm": 0.5449147628310747,
+      "learning_rate": 4.9999817327366305e-06,
+      "loss": 0.7072,
+      "step": 531
+    },
+    {
+      "epoch": 0.21022474685107434,
+      "grad_norm": 0.5817196907571065,
+      "learning_rate": 4.9999811289014876e-06,
+      "loss": 0.7127,
+      "step": 532
+    },
+    {
+      "epoch": 0.21061990614966658,
+      "grad_norm": 0.5908688502826027,
+      "learning_rate": 4.999980515247936e-06,
+      "loss": 0.7232,
+      "step": 533
+    },
+    {
+      "epoch": 0.21101506544825882,
+      "grad_norm": 0.5636398728531682,
+      "learning_rate": 4.99997989177598e-06,
+      "loss": 0.7351,
+      "step": 534
+    },
+    {
+      "epoch": 0.21141022474685106,
+      "grad_norm": 0.626447595665311,
+      "learning_rate": 4.9999792584856204e-06,
+      "loss": 0.7069,
+      "step": 535
+    },
+    {
+      "epoch": 0.21180538404544333,
+      "grad_norm": 0.5779788065467234,
+      "learning_rate": 4.999978615376861e-06,
+      "loss": 0.7109,
+      "step": 536
+    },
+    {
+      "epoch": 0.21220054334403557,
+      "grad_norm": 0.5720137928958787,
+      "learning_rate": 4.999977962449703e-06,
+      "loss": 0.7117,
+      "step": 537
+    },
+    {
+      "epoch": 0.2125957026426278,
+      "grad_norm": 0.5741484180144606,
+      "learning_rate": 4.99997729970415e-06,
+      "loss": 0.7004,
+      "step": 538
+    },
+    {
+      "epoch": 0.21299086194122005,
+      "grad_norm": 0.6385491107900715,
+      "learning_rate": 4.999976627140204e-06,
+      "loss": 0.7269,
+      "step": 539
+    },
+    {
+      "epoch": 0.2133860212398123,
+      "grad_norm": 0.5816500717084315,
+      "learning_rate": 4.999975944757868e-06,
+      "loss": 0.7036,
+      "step": 540
+    },
+    {
+      "epoch": 0.21378118053840453,
+      "grad_norm": 0.5659777534165991,
+      "learning_rate": 4.999975252557145e-06,
+      "loss": 0.7105,
+      "step": 541
+    },
+    {
+      "epoch": 0.2141763398369968,
+      "grad_norm": 0.5749663433605801,
+      "learning_rate": 4.999974550538037e-06,
+      "loss": 0.7264,
+      "step": 542
+    },
+    {
+      "epoch": 0.21457149913558904,
+      "grad_norm": 0.6112261810413965,
+      "learning_rate": 4.999973838700547e-06,
+      "loss": 0.7153,
+      "step": 543
+    },
+    {
+      "epoch": 0.21496665843418128,
+      "grad_norm": 0.5490028470815621,
+      "learning_rate": 4.999973117044678e-06,
+      "loss": 0.7083,
+      "step": 544
+    },
+    {
+      "epoch": 0.21536181773277352,
+      "grad_norm": 0.7622632899364211,
+      "learning_rate": 4.9999723855704315e-06,
+      "loss": 0.736,
+      "step": 545
+    },
+    {
+      "epoch": 0.21575697703136576,
+      "grad_norm": 0.596735586828792,
+      "learning_rate": 4.999971644277812e-06,
+      "loss": 0.6858,
+      "step": 546
+    },
+    {
+      "epoch": 0.21615213632995803,
+      "grad_norm": 0.6527535984449617,
+      "learning_rate": 4.999970893166823e-06,
+      "loss": 0.722,
+      "step": 547
+    },
+    {
+      "epoch": 0.21654729562855027,
+      "grad_norm": 0.652324953771554,
+      "learning_rate": 4.999970132237466e-06,
+      "loss": 0.7424,
+      "step": 548
+    },
+    {
+      "epoch": 0.2169424549271425,
+      "grad_norm": 0.5908339949824033,
+      "learning_rate": 4.999969361489744e-06,
+      "loss": 0.7227,
+      "step": 549
+    },
+    {
+      "epoch": 0.21733761422573475,
+      "grad_norm": 0.6629809077478138,
+      "learning_rate": 4.99996858092366e-06,
+      "loss": 0.6958,
+      "step": 550
+    },
+    {
+      "epoch": 0.217732773524327,
+      "grad_norm": 0.5704721132671529,
+      "learning_rate": 4.9999677905392185e-06,
+      "loss": 0.7056,
+      "step": 551
+    },
+    {
+      "epoch": 0.21812793282291923,
+      "grad_norm": 0.5883161489370882,
+      "learning_rate": 4.9999669903364215e-06,
+      "loss": 0.7131,
+      "step": 552
+    },
+    {
+      "epoch": 0.2185230921215115,
+      "grad_norm": 0.6166470675766057,
+      "learning_rate": 4.999966180315271e-06,
+      "loss": 0.7088,
+      "step": 553
+    },
+    {
+      "epoch": 0.21891825142010374,
+      "grad_norm": 0.5742237649639247,
+      "learning_rate": 4.999965360475773e-06,
+      "loss": 0.6977,
+      "step": 554
+    },
+    {
+      "epoch": 0.21931341071869598,
+      "grad_norm": 0.6751101103583117,
+      "learning_rate": 4.9999645308179275e-06,
+      "loss": 0.7072,
+      "step": 555
+    },
+    {
+      "epoch": 0.21970857001728822,
+      "grad_norm": 0.5984737347193803,
+      "learning_rate": 4.999963691341741e-06,
+      "loss": 0.7147,
+      "step": 556
+    },
+    {
+      "epoch": 0.22010372931588046,
+      "grad_norm": 0.6110992520398403,
+      "learning_rate": 4.999962842047214e-06,
+      "loss": 0.7013,
+      "step": 557
+    },
+    {
+      "epoch": 0.2204988886144727,
+      "grad_norm": 0.768651010687167,
+      "learning_rate": 4.999961982934352e-06,
+      "loss": 0.708,
+      "step": 558
+    },
+    {
+      "epoch": 0.22089404791306497,
+      "grad_norm": 0.5750080834238912,
+      "learning_rate": 4.999961114003156e-06,
+      "loss": 0.7163,
+      "step": 559
+    },
+    {
+      "epoch": 0.2212892072116572,
+      "grad_norm": 0.6908491657265623,
+      "learning_rate": 4.9999602352536316e-06,
+      "loss": 0.703,
+      "step": 560
+    },
+    {
+      "epoch": 0.22168436651024945,
+      "grad_norm": 0.6747348114487975,
+      "learning_rate": 4.999959346685782e-06,
+      "loss": 0.7156,
+      "step": 561
+    },
+    {
+      "epoch": 0.2220795258088417,
+      "grad_norm": 0.6777010126648408,
+      "learning_rate": 4.999958448299609e-06,
+      "loss": 0.7221,
+      "step": 562
+    },
+    {
+      "epoch": 0.22247468510743393,
+      "grad_norm": 0.7149124473295017,
+      "learning_rate": 4.9999575400951185e-06,
+      "loss": 0.7433,
+      "step": 563
+    },
+    {
+      "epoch": 0.22286984440602617,
+      "grad_norm": 0.5509899586140755,
+      "learning_rate": 4.999956622072312e-06,
+      "loss": 0.7023,
+      "step": 564
+    },
+    {
+      "epoch": 0.22326500370461844,
+      "grad_norm": 0.7820264763430183,
+      "learning_rate": 4.999955694231194e-06,
+      "loss": 0.7288,
+      "step": 565
+    },
+    {
+      "epoch": 0.22366016300321068,
+      "grad_norm": 0.5919945717677796,
+      "learning_rate": 4.999954756571769e-06,
+      "loss": 0.7152,
+      "step": 566
+    },
+    {
+      "epoch": 0.22405532230180292,
+      "grad_norm": 0.6535677585414817,
+      "learning_rate": 4.999953809094038e-06,
+      "loss": 0.7083,
+      "step": 567
+    },
+    {
+      "epoch": 0.22445048160039516,
+      "grad_norm": 0.7785050994873816,
+      "learning_rate": 4.999952851798008e-06,
+      "loss": 0.7062,
+      "step": 568
+    },
+    {
+      "epoch": 0.2248456408989874,
+      "grad_norm": 0.5482528691036233,
+      "learning_rate": 4.99995188468368e-06,
+      "loss": 0.6968,
+      "step": 569
+    },
+    {
+      "epoch": 0.22524080019757964,
+      "grad_norm": 0.7313401830631631,
+      "learning_rate": 4.99995090775106e-06,
+      "loss": 0.7196,
+      "step": 570
+    },
+    {
+      "epoch": 0.2256359594961719,
+      "grad_norm": 0.7311084419182904,
+      "learning_rate": 4.999949921000151e-06,
+      "loss": 0.6959,
+      "step": 571
+    },
+    {
+      "epoch": 0.22603111879476415,
+      "grad_norm": 0.640055542754839,
+      "learning_rate": 4.999948924430956e-06,
+      "loss": 0.7233,
+      "step": 572
+    },
+    {
+      "epoch": 0.2264262780933564,
+      "grad_norm": 0.7564396262457012,
+      "learning_rate": 4.9999479180434805e-06,
+      "loss": 0.7259,
+      "step": 573
+    },
+    {
+      "epoch": 0.22682143739194863,
+      "grad_norm": 0.5567836517749442,
+      "learning_rate": 4.999946901837727e-06,
+      "loss": 0.732,
+      "step": 574
+    },
+    {
+      "epoch": 0.22721659669054087,
+      "grad_norm": 0.6565826035434957,
+      "learning_rate": 4.999945875813701e-06,
+      "loss": 0.7107,
+      "step": 575
+    },
+    {
+      "epoch": 0.2276117559891331,
+      "grad_norm": 0.6555275013117167,
+      "learning_rate": 4.999944839971404e-06,
+      "loss": 0.7065,
+      "step": 576
+    },
+    {
+      "epoch": 0.22800691528772538,
+      "grad_norm": 0.5750121597094802,
+      "learning_rate": 4.999943794310844e-06,
+      "loss": 0.7168,
+      "step": 577
+    },
+    {
+      "epoch": 0.22840207458631762,
+      "grad_norm": 0.7172808104044253,
+      "learning_rate": 4.999942738832021e-06,
+      "loss": 0.7035,
+      "step": 578
+    },
+    {
+      "epoch": 0.22879723388490986,
+      "grad_norm": 0.681813584446898,
+      "learning_rate": 4.999941673534941e-06,
+      "loss": 0.7278,
+      "step": 579
+    },
+    {
+      "epoch": 0.2291923931835021,
+      "grad_norm": 0.6129736074024745,
+      "learning_rate": 4.999940598419609e-06,
+      "loss": 0.7207,
+      "step": 580
+    },
+    {
+      "epoch": 0.22958755248209434,
+      "grad_norm": 0.6921872225622865,
+      "learning_rate": 4.999939513486028e-06,
+      "loss": 0.7079,
+      "step": 581
+    },
+    {
+      "epoch": 0.22998271178068658,
+      "grad_norm": 0.6070410601571897,
+      "learning_rate": 4.9999384187342035e-06,
+      "loss": 0.7149,
+      "step": 582
+    },
+    {
+      "epoch": 0.23037787107927885,
+      "grad_norm": 0.6700679528371282,
+      "learning_rate": 4.999937314164138e-06,
+      "loss": 0.7091,
+      "step": 583
+    },
+    {
+      "epoch": 0.23077303037787109,
+      "grad_norm": 0.7175503981303464,
+      "learning_rate": 4.999936199775836e-06,
+      "loss": 0.7044,
+      "step": 584
+    },
+    {
+      "epoch": 0.23116818967646333,
+      "grad_norm": 0.5372712731465968,
+      "learning_rate": 4.999935075569304e-06,
+      "loss": 0.7185,
+      "step": 585
+    },
+    {
+      "epoch": 0.23156334897505557,
+      "grad_norm": 0.6626410452980375,
+      "learning_rate": 4.999933941544544e-06,
+      "loss": 0.7061,
+      "step": 586
+    },
+    {
+      "epoch": 0.2319585082736478,
+      "grad_norm": 0.6219349768601438,
+      "learning_rate": 4.999932797701563e-06,
+      "loss": 0.7027,
+      "step": 587
+    },
+    {
+      "epoch": 0.23235366757224005,
+      "grad_norm": 0.622944974108998,
+      "learning_rate": 4.999931644040363e-06,
+      "loss": 0.7041,
+      "step": 588
+    },
+    {
+      "epoch": 0.23274882687083231,
+      "grad_norm": 0.6403160681420452,
+      "learning_rate": 4.999930480560949e-06,
+      "loss": 0.713,
+      "step": 589
+    },
+    {
+      "epoch": 0.23314398616942456,
+      "grad_norm": 0.5810884528951928,
+      "learning_rate": 4.999929307263327e-06,
+      "loss": 0.6959,
+      "step": 590
+    },
+    {
+      "epoch": 0.2335391454680168,
+      "grad_norm": 0.5617194856604512,
+      "learning_rate": 4.999928124147501e-06,
+      "loss": 0.7134,
+      "step": 591
+    },
+    {
+      "epoch": 0.23393430476660904,
+      "grad_norm": 0.5794486892578056,
+      "learning_rate": 4.999926931213475e-06,
+      "loss": 0.6581,
+      "step": 592
+    },
+    {
+      "epoch": 0.23432946406520128,
+      "grad_norm": 0.6010638355759388,
+      "learning_rate": 4.999925728461254e-06,
+      "loss": 0.6905,
+      "step": 593
+    },
+    {
+      "epoch": 0.23472462336379352,
+      "grad_norm": 0.5698458903870885,
+      "learning_rate": 4.999924515890843e-06,
+      "loss": 0.6929,
+      "step": 594
+    },
+    {
+      "epoch": 0.23511978266238578,
+      "grad_norm": 0.6038481326467747,
+      "learning_rate": 4.9999232935022455e-06,
+      "loss": 0.7134,
+      "step": 595
+    },
+    {
+      "epoch": 0.23551494196097802,
+      "grad_norm": 0.5575437681930879,
+      "learning_rate": 4.9999220612954685e-06,
+      "loss": 0.686,
+      "step": 596
+    },
+    {
+      "epoch": 0.23591010125957027,
+      "grad_norm": 0.5582827483407896,
+      "learning_rate": 4.999920819270515e-06,
+      "loss": 0.677,
+      "step": 597
+    },
+    {
+      "epoch": 0.2363052605581625,
+      "grad_norm": 0.5581039778387297,
+      "learning_rate": 4.999919567427391e-06,
+      "loss": 0.7041,
+      "step": 598
+    },
+    {
+      "epoch": 0.23670041985675475,
+      "grad_norm": 0.5926274884146744,
+      "learning_rate": 4.999918305766101e-06,
+      "loss": 0.7131,
+      "step": 599
+    },
+    {
+      "epoch": 0.23709557915534699,
+      "grad_norm": 0.5683616438789352,
+      "learning_rate": 4.99991703428665e-06,
+      "loss": 0.7004,
+      "step": 600
+    },
+    {
+      "epoch": 0.23749073845393925,
+      "grad_norm": 0.5730277632443862,
+      "learning_rate": 4.999915752989042e-06,
+      "loss": 0.7056,
+      "step": 601
+    },
+    {
+      "epoch": 0.2378858977525315,
+      "grad_norm": 0.582494416565038,
+      "learning_rate": 4.999914461873283e-06,
+      "loss": 0.7116,
+      "step": 602
+    },
+    {
+      "epoch": 0.23828105705112373,
+      "grad_norm": 0.5966610530038993,
+      "learning_rate": 4.9999131609393795e-06,
+      "loss": 0.7284,
+      "step": 603
+    },
+    {
+      "epoch": 0.23867621634971597,
+      "grad_norm": 0.5558683547691745,
+      "learning_rate": 4.999911850187335e-06,
+      "loss": 0.6998,
+      "step": 604
+    },
+    {
+      "epoch": 0.23907137564830822,
+      "grad_norm": 0.6258177810251072,
+      "learning_rate": 4.999910529617154e-06,
+      "loss": 0.6855,
+      "step": 605
+    },
+    {
+      "epoch": 0.23946653494690046,
+      "grad_norm": 0.5318478473731086,
+      "learning_rate": 4.999909199228841e-06,
+      "loss": 0.6795,
+      "step": 606
+    },
+    {
+      "epoch": 0.23986169424549272,
+      "grad_norm": 0.5547878659686439,
+      "learning_rate": 4.999907859022405e-06,
+      "loss": 0.6953,
+      "step": 607
+    },
+    {
+      "epoch": 0.24025685354408496,
+      "grad_norm": 0.5931673987268137,
+      "learning_rate": 4.9999065089978485e-06,
+      "loss": 0.7148,
+      "step": 608
+    },
+    {
+      "epoch": 0.2406520128426772,
+      "grad_norm": 0.568231220260507,
+      "learning_rate": 4.9999051491551766e-06,
+      "loss": 0.695,
+      "step": 609
+    },
+    {
+      "epoch": 0.24104717214126944,
+      "grad_norm": 0.5808912372610399,
+      "learning_rate": 4.999903779494395e-06,
+      "loss": 0.7074,
+      "step": 610
+    },
+    {
+      "epoch": 0.24144233143986168,
+      "grad_norm": 0.5824501699853736,
+      "learning_rate": 4.9999024000155104e-06,
+      "loss": 0.7017,
+      "step": 611
+    },
+    {
+      "epoch": 0.24183749073845395,
+      "grad_norm": 0.5393723920629134,
+      "learning_rate": 4.9999010107185264e-06,
+      "loss": 0.6646,
+      "step": 612
+    },
+    {
+      "epoch": 0.2422326500370462,
+      "grad_norm": 0.5980491294226428,
+      "learning_rate": 4.99989961160345e-06,
+      "loss": 0.6794,
+      "step": 613
+    },
+    {
+      "epoch": 0.24262780933563843,
+      "grad_norm": 0.6188242993162584,
+      "learning_rate": 4.999898202670286e-06,
+      "loss": 0.705,
+      "step": 614
+    },
+    {
+      "epoch": 0.24302296863423067,
+      "grad_norm": 0.5890575366759165,
+      "learning_rate": 4.9998967839190384e-06,
+      "loss": 0.7118,
+      "step": 615
+    },
+    {
+      "epoch": 0.24341812793282291,
+      "grad_norm": 0.6031363408649272,
+      "learning_rate": 4.999895355349716e-06,
+      "loss": 0.7248,
+      "step": 616
+    },
+    {
+      "epoch": 0.24381328723141515,
+      "grad_norm": 0.5917614651923631,
+      "learning_rate": 4.9998939169623215e-06,
+      "loss": 0.696,
+      "step": 617
+    },
+    {
+      "epoch": 0.24420844653000742,
+      "grad_norm": 0.594908607458186,
+      "learning_rate": 4.9998924687568625e-06,
+      "loss": 0.7036,
+      "step": 618
+    },
+    {
+      "epoch": 0.24460360582859966,
+      "grad_norm": 0.6727052500603081,
+      "learning_rate": 4.999891010733343e-06,
+      "loss": 0.6824,
+      "step": 619
+    },
+    {
+      "epoch": 0.2449987651271919,
+      "grad_norm": 0.598235520036059,
+      "learning_rate": 4.99988954289177e-06,
+      "loss": 0.7238,
+      "step": 620
+    },
+    {
+      "epoch": 0.24539392442578414,
+      "grad_norm": 0.7176084933336286,
+      "learning_rate": 4.999888065232149e-06,
+      "loss": 0.7076,
+      "step": 621
+    },
+    {
+      "epoch": 0.24578908372437638,
+      "grad_norm": 0.5872860940803798,
+      "learning_rate": 4.999886577754486e-06,
+      "loss": 0.702,
+      "step": 622
+    },
+    {
+      "epoch": 0.24618424302296862,
+      "grad_norm": 0.6143484421432265,
+      "learning_rate": 4.999885080458786e-06,
+      "loss": 0.6923,
+      "step": 623
+    },
+    {
+      "epoch": 0.2465794023215609,
+      "grad_norm": 0.6537296010257171,
+      "learning_rate": 4.9998835733450564e-06,
+      "loss": 0.6996,
+      "step": 624
+    },
+    {
+      "epoch": 0.24697456162015313,
+      "grad_norm": 0.5652281166809588,
+      "learning_rate": 4.9998820564133016e-06,
+      "loss": 0.6926,
+      "step": 625
+    },
+    {
+      "epoch": 0.24736972091874537,
+      "grad_norm": 0.6841779465916256,
+      "learning_rate": 4.999880529663528e-06,
+      "loss": 0.708,
+      "step": 626
+    },
+    {
+      "epoch": 0.2477648802173376,
+      "grad_norm": 0.618999997435258,
+      "learning_rate": 4.9998789930957415e-06,
+      "loss": 0.7009,
+      "step": 627
+    },
+    {
+      "epoch": 0.24816003951592985,
+      "grad_norm": 0.5814213814123862,
+      "learning_rate": 4.999877446709949e-06,
+      "loss": 0.7166,
+      "step": 628
+    },
+    {
+      "epoch": 0.2485551988145221,
+      "grad_norm": 0.6410458092239222,
+      "learning_rate": 4.999875890506155e-06,
+      "loss": 0.701,
+      "step": 629
+    },
+    {
+      "epoch": 0.24895035811311436,
+      "grad_norm": 0.5643287481712287,
+      "learning_rate": 4.999874324484367e-06,
+      "loss": 0.6781,
+      "step": 630
+    },
+    {
+      "epoch": 0.2493455174117066,
+      "grad_norm": 0.719129519371788,
+      "learning_rate": 4.9998727486445895e-06,
+      "loss": 0.71,
+      "step": 631
+    },
+    {
+      "epoch": 0.24974067671029884,
+      "grad_norm": 0.643749171305181,
+      "learning_rate": 4.999871162986831e-06,
+      "loss": 0.6954,
+      "step": 632
+    },
+    {
+      "epoch": 0.2501358360088911,
+      "grad_norm": 0.5312855320703463,
+      "learning_rate": 4.999869567511097e-06,
+      "loss": 0.6619,
+      "step": 633
+    },
+    {
+      "epoch": 0.25053099530748335,
+      "grad_norm": 0.5908234255298699,
+      "learning_rate": 4.9998679622173925e-06,
+      "loss": 0.7128,
+      "step": 634
+    },
+    {
+      "epoch": 0.2509261546060756,
+      "grad_norm": 0.6073203092125553,
+      "learning_rate": 4.999866347105725e-06,
+      "loss": 0.6908,
+      "step": 635
+    },
+    {
+      "epoch": 0.25132131390466783,
+      "grad_norm": 0.5785049800687088,
+      "learning_rate": 4.999864722176101e-06,
+      "loss": 0.6863,
+      "step": 636
+    },
+    {
+      "epoch": 0.25171647320326007,
+      "grad_norm": 0.6312245040925751,
+      "learning_rate": 4.999863087428526e-06,
+      "loss": 0.6905,
+      "step": 637
+    },
+    {
+      "epoch": 0.2521116325018523,
+      "grad_norm": 0.596727469474082,
+      "learning_rate": 4.999861442863005e-06,
+      "loss": 0.6822,
+      "step": 638
+    },
+    {
+      "epoch": 0.25250679180044455,
+      "grad_norm": 0.5678026535695422,
+      "learning_rate": 4.999859788479548e-06,
+      "loss": 0.6999,
+      "step": 639
+    },
+    {
+      "epoch": 0.2529019510990368,
+      "grad_norm": 0.548613923585835,
+      "learning_rate": 4.99985812427816e-06,
+      "loss": 0.7132,
+      "step": 640
+    },
+    {
+      "epoch": 0.25329711039762903,
+      "grad_norm": 0.5873264929153031,
+      "learning_rate": 4.999856450258847e-06,
+      "loss": 0.7207,
+      "step": 641
+    },
+    {
+      "epoch": 0.2536922696962213,
+      "grad_norm": 0.5678140101223359,
+      "learning_rate": 4.9998547664216155e-06,
+      "loss": 0.7106,
+      "step": 642
+    },
+    {
+      "epoch": 0.2540874289948135,
+      "grad_norm": 0.5578970132548404,
+      "learning_rate": 4.999853072766473e-06,
+      "loss": 0.7125,
+      "step": 643
+    },
+    {
+      "epoch": 0.25448258829340575,
+      "grad_norm": 0.6041251604589954,
+      "learning_rate": 4.999851369293425e-06,
+      "loss": 0.6901,
+      "step": 644
+    },
+    {
+      "epoch": 0.25487774759199805,
+      "grad_norm": 0.5736028260838277,
+      "learning_rate": 4.99984965600248e-06,
+      "loss": 0.7116,
+      "step": 645
+    },
+    {
+      "epoch": 0.2552729068905903,
+      "grad_norm": 0.532105428102255,
+      "learning_rate": 4.999847932893642e-06,
+      "loss": 0.6927,
+      "step": 646
+    },
+    {
+      "epoch": 0.25566806618918253,
+      "grad_norm": 0.6413099649877119,
+      "learning_rate": 4.99984619996692e-06,
+      "loss": 0.6916,
+      "step": 647
+    },
+    {
+      "epoch": 0.25606322548777477,
+      "grad_norm": 0.5690991015722213,
+      "learning_rate": 4.9998444572223205e-06,
+      "loss": 0.7192,
+      "step": 648
+    },
+    {
+      "epoch": 0.256458384786367,
+      "grad_norm": 0.5557660152572236,
+      "learning_rate": 4.999842704659849e-06,
+      "loss": 0.6832,
+      "step": 649
+    },
+    {
+      "epoch": 0.25685354408495925,
+      "grad_norm": 0.5617270616167598,
+      "learning_rate": 4.999840942279514e-06,
+      "loss": 0.6878,
+      "step": 650
+    },
+    {
+      "epoch": 0.2572487033835515,
+      "grad_norm": 0.5666842673944367,
+      "learning_rate": 4.999839170081322e-06,
+      "loss": 0.6672,
+      "step": 651
+    },
+    {
+      "epoch": 0.25764386268214373,
+      "grad_norm": 0.5623407827389078,
+      "learning_rate": 4.999837388065279e-06,
+      "loss": 0.689,
+      "step": 652
+    },
+    {
+      "epoch": 0.25803902198073597,
+      "grad_norm": 0.5610766033323613,
+      "learning_rate": 4.999835596231392e-06,
+      "loss": 0.7217,
+      "step": 653
+    },
+    {
+      "epoch": 0.2584341812793282,
+      "grad_norm": 0.5941036719520371,
+      "learning_rate": 4.999833794579671e-06,
+      "loss": 0.704,
+      "step": 654
+    },
+    {
+      "epoch": 0.25882934057792045,
+      "grad_norm": 0.5527677491351838,
+      "learning_rate": 4.999831983110119e-06,
+      "loss": 0.7063,
+      "step": 655
+    },
+    {
+      "epoch": 0.2592244998765127,
+      "grad_norm": 0.5896815383284895,
+      "learning_rate": 4.999830161822746e-06,
+      "loss": 0.6823,
+      "step": 656
+    },
+    {
+      "epoch": 0.259619659175105,
+      "grad_norm": 0.5617257867146465,
+      "learning_rate": 4.999828330717558e-06,
+      "loss": 0.6952,
+      "step": 657
+    },
+    {
+      "epoch": 0.26001481847369723,
+      "grad_norm": 0.5841626881089839,
+      "learning_rate": 4.999826489794562e-06,
+      "loss": 0.6801,
+      "step": 658
+    },
+    {
+      "epoch": 0.26040997777228947,
+      "grad_norm": 0.5646406469988224,
+      "learning_rate": 4.9998246390537655e-06,
+      "loss": 0.6812,
+      "step": 659
+    },
+    {
+      "epoch": 0.2608051370708817,
+      "grad_norm": 0.5750406721528946,
+      "learning_rate": 4.999822778495176e-06,
+      "loss": 0.6943,
+      "step": 660
+    },
+    {
+      "epoch": 0.26120029636947395,
+      "grad_norm": 0.5559705390104022,
+      "learning_rate": 4.999820908118801e-06,
+      "loss": 0.6763,
+      "step": 661
+    },
+    {
+      "epoch": 0.2615954556680662,
+      "grad_norm": 0.5562879316008681,
+      "learning_rate": 4.999819027924646e-06,
+      "loss": 0.6832,
+      "step": 662
+    },
+    {
+      "epoch": 0.26199061496665843,
+      "grad_norm": 0.5921377833036137,
+      "learning_rate": 4.999817137912721e-06,
+      "loss": 0.6858,
+      "step": 663
+    },
+    {
+      "epoch": 0.26238577426525067,
+      "grad_norm": 0.567426550861614,
+      "learning_rate": 4.9998152380830325e-06,
+      "loss": 0.7036,
+      "step": 664
+    },
+    {
+      "epoch": 0.2627809335638429,
+      "grad_norm": 0.6061252730535107,
+      "learning_rate": 4.999813328435587e-06,
+      "loss": 0.6858,
+      "step": 665
+    },
+    {
+      "epoch": 0.26317609286243515,
+      "grad_norm": 0.5678734043512008,
+      "learning_rate": 4.999811408970392e-06,
+      "loss": 0.7152,
+      "step": 666
+    },
+    {
+      "epoch": 0.2635712521610274,
+      "grad_norm": 0.6033593786362805,
+      "learning_rate": 4.999809479687457e-06,
+      "loss": 0.6913,
+      "step": 667
+    },
+    {
+      "epoch": 0.26396641145961963,
+      "grad_norm": 0.5744590883863819,
+      "learning_rate": 4.999807540586788e-06,
+      "loss": 0.7029,
+      "step": 668
+    },
+    {
+      "epoch": 0.26436157075821193,
+      "grad_norm": 0.5281778655339753,
+      "learning_rate": 4.999805591668393e-06,
+      "loss": 0.7049,
+      "step": 669
+    },
+    {
+      "epoch": 0.26475673005680417,
+      "grad_norm": 0.5585259313584738,
+      "learning_rate": 4.99980363293228e-06,
+      "loss": 0.673,
+      "step": 670
+    },
+    {
+      "epoch": 0.2651518893553964,
+      "grad_norm": 0.5594694999229931,
+      "learning_rate": 4.999801664378455e-06,
+      "loss": 0.6827,
+      "step": 671
+    },
+    {
+      "epoch": 0.26554704865398865,
+      "grad_norm": 0.5996606992971393,
+      "learning_rate": 4.999799686006928e-06,
+      "loss": 0.7072,
+      "step": 672
+    },
+    {
+      "epoch": 0.2659422079525809,
+      "grad_norm": 0.5580082863684926,
+      "learning_rate": 4.999797697817706e-06,
+      "loss": 0.7039,
+      "step": 673
+    },
+    {
+      "epoch": 0.26633736725117313,
+      "grad_norm": 0.521406203903939,
+      "learning_rate": 4.999795699810796e-06,
+      "loss": 0.6886,
+      "step": 674
+    },
+    {
+      "epoch": 0.26673252654976537,
+      "grad_norm": 0.5672071403136022,
+      "learning_rate": 4.9997936919862065e-06,
+      "loss": 0.6799,
+      "step": 675
+    },
+    {
+      "epoch": 0.2671276858483576,
+      "grad_norm": 0.5568748196172835,
+      "learning_rate": 4.999791674343945e-06,
+      "loss": 0.7012,
+      "step": 676
+    },
+    {
+      "epoch": 0.26752284514694985,
+      "grad_norm": 0.5593131719229575,
+      "learning_rate": 4.999789646884021e-06,
+      "loss": 0.6841,
+      "step": 677
+    },
+    {
+      "epoch": 0.2679180044455421,
+      "grad_norm": 0.5731216777632498,
+      "learning_rate": 4.99978760960644e-06,
+      "loss": 0.7111,
+      "step": 678
+    },
+    {
+      "epoch": 0.26831316374413433,
+      "grad_norm": 0.6026088473088514,
+      "learning_rate": 4.999785562511212e-06,
+      "loss": 0.7065,
+      "step": 679
+    },
+    {
+      "epoch": 0.2687083230427266,
+      "grad_norm": 0.5516806935780276,
+      "learning_rate": 4.999783505598344e-06,
+      "loss": 0.6909,
+      "step": 680
+    },
+    {
+      "epoch": 0.26910348234131887,
+      "grad_norm": 0.5702038400823601,
+      "learning_rate": 4.999781438867844e-06,
+      "loss": 0.6734,
+      "step": 681
+    },
+    {
+      "epoch": 0.2694986416399111,
+      "grad_norm": 0.5746782762066555,
+      "learning_rate": 4.999779362319721e-06,
+      "loss": 0.7042,
+      "step": 682
+    },
+    {
+      "epoch": 0.26989380093850335,
+      "grad_norm": 0.6007328759815763,
+      "learning_rate": 4.9997772759539825e-06,
+      "loss": 0.6971,
+      "step": 683
+    },
+    {
+      "epoch": 0.2702889602370956,
+      "grad_norm": 0.5658032828435371,
+      "learning_rate": 4.999775179770637e-06,
+      "loss": 0.6741,
+      "step": 684
+    },
+    {
+      "epoch": 0.27068411953568783,
+      "grad_norm": 0.5445910362376495,
+      "learning_rate": 4.999773073769692e-06,
+      "loss": 0.6914,
+      "step": 685
+    },
+    {
+      "epoch": 0.27107927883428007,
+      "grad_norm": 0.5625355418392146,
+      "learning_rate": 4.9997709579511566e-06,
+      "loss": 0.6778,
+      "step": 686
+    },
+    {
+      "epoch": 0.2714744381328723,
+      "grad_norm": 0.5558770779852484,
+      "learning_rate": 4.999768832315038e-06,
+      "loss": 0.675,
+      "step": 687
+    },
+    {
+      "epoch": 0.27186959743146455,
+      "grad_norm": 0.5384961627107955,
+      "learning_rate": 4.999766696861346e-06,
+      "loss": 0.6881,
+      "step": 688
+    },
+    {
+      "epoch": 0.2722647567300568,
+      "grad_norm": 0.5646517085086373,
+      "learning_rate": 4.9997645515900895e-06,
+      "loss": 0.6737,
+      "step": 689
+    },
+    {
+      "epoch": 0.27265991602864903,
+      "grad_norm": 0.5769838182468013,
+      "learning_rate": 4.9997623965012745e-06,
+      "loss": 0.6718,
+      "step": 690
+    },
+    {
+      "epoch": 0.27305507532724127,
+      "grad_norm": 0.7282357435191615,
+      "learning_rate": 4.999760231594911e-06,
+      "loss": 0.7049,
+      "step": 691
+    },
+    {
+      "epoch": 0.27345023462583357,
+      "grad_norm": 0.5279126762131758,
+      "learning_rate": 4.999758056871007e-06,
+      "loss": 0.6621,
+      "step": 692
+    },
+    {
+      "epoch": 0.2738453939244258,
+      "grad_norm": 0.5706899106672201,
+      "learning_rate": 4.999755872329571e-06,
+      "loss": 0.6883,
+      "step": 693
+    },
+    {
+      "epoch": 0.27424055322301805,
+      "grad_norm": 0.5563908836582794,
+      "learning_rate": 4.999753677970614e-06,
+      "loss": 0.7143,
+      "step": 694
+    },
+    {
+      "epoch": 0.2746357125216103,
+      "grad_norm": 0.5362870289511571,
+      "learning_rate": 4.9997514737941396e-06,
+      "loss": 0.684,
+      "step": 695
+    },
+    {
+      "epoch": 0.2750308718202025,
+      "grad_norm": 0.6153860851792133,
+      "learning_rate": 4.9997492598001605e-06,
+      "loss": 0.6883,
+      "step": 696
+    },
+    {
+      "epoch": 0.27542603111879477,
+      "grad_norm": 0.546338501558778,
+      "learning_rate": 4.999747035988684e-06,
+      "loss": 0.6853,
+      "step": 697
+    },
+    {
+      "epoch": 0.275821190417387,
+      "grad_norm": 0.5411393398480797,
+      "learning_rate": 4.999744802359721e-06,
+      "loss": 0.6759,
+      "step": 698
+    },
+    {
+      "epoch": 0.27621634971597925,
+      "grad_norm": 0.5820635145009669,
+      "learning_rate": 4.999742558913276e-06,
+      "loss": 0.69,
+      "step": 699
+    },
+    {
+      "epoch": 0.2766115090145715,
+      "grad_norm": 0.519802903820158,
+      "learning_rate": 4.999740305649361e-06,
+      "loss": 0.689,
+      "step": 700
+    },
+    {
+      "epoch": 0.27700666831316373,
+      "grad_norm": 0.5335002406037318,
+      "learning_rate": 4.999738042567984e-06,
+      "loss": 0.6733,
+      "step": 701
+    },
+    {
+      "epoch": 0.27740182761175597,
+      "grad_norm": 0.6251340494736801,
+      "learning_rate": 4.999735769669154e-06,
+      "loss": 0.6947,
+      "step": 702
+    },
+    {
+      "epoch": 0.2777969869103482,
+      "grad_norm": 0.5568465012685252,
+      "learning_rate": 4.999733486952879e-06,
+      "loss": 0.6791,
+      "step": 703
+    },
+    {
+      "epoch": 0.2781921462089405,
+      "grad_norm": 0.5677313063564176,
+      "learning_rate": 4.9997311944191695e-06,
+      "loss": 0.6653,
+      "step": 704
+    },
+    {
+      "epoch": 0.27858730550753275,
+      "grad_norm": 0.5398469806124272,
+      "learning_rate": 4.9997288920680345e-06,
+      "loss": 0.6708,
+      "step": 705
+    },
+    {
+      "epoch": 0.278982464806125,
+      "grad_norm": 0.5249202618448952,
+      "learning_rate": 4.999726579899481e-06,
+      "loss": 0.6714,
+      "step": 706
+    },
+    {
+      "epoch": 0.2793776241047172,
+      "grad_norm": 0.5479464511682306,
+      "learning_rate": 4.99972425791352e-06,
+      "loss": 0.6884,
+      "step": 707
+    },
+    {
+      "epoch": 0.27977278340330947,
+      "grad_norm": 0.612951514131732,
+      "learning_rate": 4.99972192611016e-06,
+      "loss": 0.6934,
+      "step": 708
+    },
+    {
+      "epoch": 0.2801679427019017,
+      "grad_norm": 0.5472856818707729,
+      "learning_rate": 4.999719584489409e-06,
+      "loss": 0.6927,
+      "step": 709
+    },
+    {
+      "epoch": 0.28056310200049395,
+      "grad_norm": 0.5711839364829995,
+      "learning_rate": 4.99971723305128e-06,
+      "loss": 0.6993,
+      "step": 710
+    },
+    {
+      "epoch": 0.2809582612990862,
+      "grad_norm": 0.5344843617908501,
+      "learning_rate": 4.999714871795778e-06,
+      "loss": 0.6713,
+      "step": 711
+    },
+    {
+      "epoch": 0.2813534205976784,
+      "grad_norm": 0.5725098490316971,
+      "learning_rate": 4.9997125007229144e-06,
+      "loss": 0.7028,
+      "step": 712
+    },
+    {
+      "epoch": 0.28174857989627067,
+      "grad_norm": 0.5706604408107072,
+      "learning_rate": 4.999710119832697e-06,
+      "loss": 0.7134,
+      "step": 713
+    },
+    {
+      "epoch": 0.2821437391948629,
+      "grad_norm": 0.5261160242380305,
+      "learning_rate": 4.9997077291251366e-06,
+      "loss": 0.6767,
+      "step": 714
+    },
+    {
+      "epoch": 0.28253889849345515,
+      "grad_norm": 0.5469719762517126,
+      "learning_rate": 4.999705328600243e-06,
+      "loss": 0.7025,
+      "step": 715
+    },
+    {
+      "epoch": 0.28293405779204744,
+      "grad_norm": 0.5324472833124879,
+      "learning_rate": 4.9997029182580236e-06,
+      "loss": 0.6743,
+      "step": 716
+    },
+    {
+      "epoch": 0.2833292170906397,
+      "grad_norm": 0.5401067135045965,
+      "learning_rate": 4.99970049809849e-06,
+      "loss": 0.6931,
+      "step": 717
+    },
+    {
+      "epoch": 0.2837243763892319,
+      "grad_norm": 0.5544406474161367,
+      "learning_rate": 4.99969806812165e-06,
+      "loss": 0.6968,
+      "step": 718
+    },
+    {
+      "epoch": 0.28411953568782417,
+      "grad_norm": 0.5506360149437247,
+      "learning_rate": 4.999695628327514e-06,
+      "loss": 0.6981,
+      "step": 719
+    },
+    {
+      "epoch": 0.2845146949864164,
+      "grad_norm": 0.5478511076951501,
+      "learning_rate": 4.999693178716092e-06,
+      "loss": 0.6782,
+      "step": 720
+    },
+    {
+      "epoch": 0.28490985428500865,
+      "grad_norm": 0.5487601492507348,
+      "learning_rate": 4.999690719287392e-06,
+      "loss": 0.682,
+      "step": 721
+    },
+    {
+      "epoch": 0.2853050135836009,
+      "grad_norm": 0.5568074889646241,
+      "learning_rate": 4.999688250041426e-06,
+      "loss": 0.7005,
+      "step": 722
+    },
+    {
+      "epoch": 0.2857001728821931,
+      "grad_norm": 0.5211276732961236,
+      "learning_rate": 4.999685770978202e-06,
+      "loss": 0.685,
+      "step": 723
+    },
+    {
+      "epoch": 0.28609533218078537,
+      "grad_norm": 0.5238564370246029,
+      "learning_rate": 4.9996832820977305e-06,
+      "loss": 0.6733,
+      "step": 724
+    },
+    {
+      "epoch": 0.2864904914793776,
+      "grad_norm": 0.5488717039839817,
+      "learning_rate": 4.999680783400021e-06,
+      "loss": 0.6824,
+      "step": 725
+    },
+    {
+      "epoch": 0.28688565077796985,
+      "grad_norm": 0.5590783909424277,
+      "learning_rate": 4.999678274885082e-06,
+      "loss": 0.6852,
+      "step": 726
+    },
+    {
+      "epoch": 0.2872808100765621,
+      "grad_norm": 0.5718775481303447,
+      "learning_rate": 4.999675756552926e-06,
+      "loss": 0.6684,
+      "step": 727
+    },
+    {
+      "epoch": 0.2876759693751544,
+      "grad_norm": 0.5537561512785655,
+      "learning_rate": 4.99967322840356e-06,
+      "loss": 0.7021,
+      "step": 728
+    },
+    {
+      "epoch": 0.2880711286737466,
+      "grad_norm": 0.595756815576963,
+      "learning_rate": 4.999670690436997e-06,
+      "loss": 0.7033,
+      "step": 729
+    },
+    {
+      "epoch": 0.28846628797233886,
+      "grad_norm": 0.5734638717038859,
+      "learning_rate": 4.999668142653244e-06,
+      "loss": 0.6825,
+      "step": 730
+    },
+    {
+      "epoch": 0.2888614472709311,
+      "grad_norm": 0.5649785931782922,
+      "learning_rate": 4.999665585052314e-06,
+      "loss": 0.7142,
+      "step": 731
+    },
+    {
+      "epoch": 0.28925660656952334,
+      "grad_norm": 0.6530890960010839,
+      "learning_rate": 4.999663017634214e-06,
+      "loss": 0.6788,
+      "step": 732
+    },
+    {
+      "epoch": 0.2896517658681156,
+      "grad_norm": 0.6271597238579767,
+      "learning_rate": 4.999660440398957e-06,
+      "loss": 0.6949,
+      "step": 733
+    },
+    {
+      "epoch": 0.2900469251667078,
+      "grad_norm": 0.5695950102858286,
+      "learning_rate": 4.999657853346551e-06,
+      "loss": 0.6833,
+      "step": 734
+    },
+    {
+      "epoch": 0.29044208446530007,
+      "grad_norm": 0.6916391987946482,
+      "learning_rate": 4.999655256477006e-06,
+      "loss": 0.6961,
+      "step": 735
+    },
+    {
+      "epoch": 0.2908372437638923,
+      "grad_norm": 0.6074484081013837,
+      "learning_rate": 4.9996526497903345e-06,
+      "loss": 0.7025,
+      "step": 736
+    },
+    {
+      "epoch": 0.29123240306248455,
+      "grad_norm": 0.5356095341699845,
+      "learning_rate": 4.9996500332865445e-06,
+      "loss": 0.6576,
+      "step": 737
+    },
+    {
+      "epoch": 0.2916275623610768,
+      "grad_norm": 0.5270122623233058,
+      "learning_rate": 4.999647406965647e-06,
+      "loss": 0.6493,
+      "step": 738
+    },
+    {
+      "epoch": 0.292022721659669,
+      "grad_norm": 0.5844224564006655,
+      "learning_rate": 4.999644770827652e-06,
+      "loss": 0.7082,
+      "step": 739
+    },
+    {
+      "epoch": 0.2924178809582613,
+      "grad_norm": 0.5731795031934965,
+      "learning_rate": 4.999642124872571e-06,
+      "loss": 0.6796,
+      "step": 740
+    },
+    {
+      "epoch": 0.29281304025685356,
+      "grad_norm": 0.5464280422780536,
+      "learning_rate": 4.999639469100413e-06,
+      "loss": 0.6919,
+      "step": 741
+    },
+    {
+      "epoch": 0.2932081995554458,
+      "grad_norm": 0.5488021923081339,
+      "learning_rate": 4.99963680351119e-06,
+      "loss": 0.6991,
+      "step": 742
+    },
+    {
+      "epoch": 0.29360335885403804,
+      "grad_norm": 0.5999425296366285,
+      "learning_rate": 4.99963412810491e-06,
+      "loss": 0.6892,
+      "step": 743
+    },
+    {
+      "epoch": 0.2939985181526303,
+      "grad_norm": 0.5301599791894791,
+      "learning_rate": 4.999631442881586e-06,
+      "loss": 0.7089,
+      "step": 744
+    },
+    {
+      "epoch": 0.2943936774512225,
+      "grad_norm": 0.5563197028629331,
+      "learning_rate": 4.999628747841227e-06,
+      "loss": 0.6775,
+      "step": 745
+    },
+    {
+      "epoch": 0.29478883674981476,
+      "grad_norm": 0.5446511718424091,
+      "learning_rate": 4.999626042983845e-06,
+      "loss": 0.6963,
+      "step": 746
+    },
+    {
+      "epoch": 0.295183996048407,
+      "grad_norm": 0.5507487342171807,
+      "learning_rate": 4.9996233283094485e-06,
+      "loss": 0.6755,
+      "step": 747
+    },
+    {
+      "epoch": 0.29557915534699924,
+      "grad_norm": 0.5266055784439764,
+      "learning_rate": 4.999620603818051e-06,
+      "loss": 0.6929,
+      "step": 748
+    },
+    {
+      "epoch": 0.2959743146455915,
+      "grad_norm": 0.5626530509305415,
+      "learning_rate": 4.99961786950966e-06,
+      "loss": 0.6898,
+      "step": 749
+    },
+    {
+      "epoch": 0.2963694739441837,
+      "grad_norm": 0.5399861793258945,
+      "learning_rate": 4.999615125384289e-06,
+      "loss": 0.6801,
+      "step": 750
+    },
+    {
+      "epoch": 0.296764633242776,
+      "grad_norm": 0.5610645829088754,
+      "learning_rate": 4.999612371441947e-06,
+      "loss": 0.6709,
+      "step": 751
+    },
+    {
+      "epoch": 0.29715979254136826,
+      "grad_norm": 0.5394716972805861,
+      "learning_rate": 4.999609607682645e-06,
+      "loss": 0.6514,
+      "step": 752
+    },
+    {
+      "epoch": 0.2975549518399605,
+      "grad_norm": 0.5430055925792929,
+      "learning_rate": 4.999606834106395e-06,
+      "loss": 0.6849,
+      "step": 753
+    },
+    {
+      "epoch": 0.29795011113855274,
+      "grad_norm": 0.544962481828332,
+      "learning_rate": 4.999604050713207e-06,
+      "loss": 0.7015,
+      "step": 754
+    },
+    {
+      "epoch": 0.298345270437145,
+      "grad_norm": 0.5344852748367742,
+      "learning_rate": 4.999601257503093e-06,
+      "loss": 0.6949,
+      "step": 755
+    },
+    {
+      "epoch": 0.2987404297357372,
+      "grad_norm": 0.5719742664992333,
+      "learning_rate": 4.999598454476062e-06,
+      "loss": 0.6901,
+      "step": 756
+    },
+    {
+      "epoch": 0.29913558903432946,
+      "grad_norm": 0.584662964003699,
+      "learning_rate": 4.999595641632126e-06,
+      "loss": 0.7033,
+      "step": 757
+    },
+    {
+      "epoch": 0.2995307483329217,
+      "grad_norm": 0.552326717376869,
+      "learning_rate": 4.999592818971297e-06,
+      "loss": 0.6863,
+      "step": 758
+    },
+    {
+      "epoch": 0.29992590763151394,
+      "grad_norm": 0.5475655007217732,
+      "learning_rate": 4.999589986493585e-06,
+      "loss": 0.66,
+      "step": 759
+    },
+    {
+      "epoch": 0.3003210669301062,
+      "grad_norm": 0.5882240228280694,
+      "learning_rate": 4.999587144199001e-06,
+      "loss": 0.6962,
+      "step": 760
+    },
+    {
+      "epoch": 0.3007162262286984,
+      "grad_norm": 0.5699153356709309,
+      "learning_rate": 4.9995842920875575e-06,
+      "loss": 0.6803,
+      "step": 761
+    },
+    {
+      "epoch": 0.30111138552729066,
+      "grad_norm": 0.6280558268501941,
+      "learning_rate": 4.9995814301592635e-06,
+      "loss": 0.6793,
+      "step": 762
+    },
+    {
+      "epoch": 0.30150654482588296,
+      "grad_norm": 0.6650094210932691,
+      "learning_rate": 4.999578558414132e-06,
+      "loss": 0.6975,
+      "step": 763
+    },
+    {
+      "epoch": 0.3019017041244752,
+      "grad_norm": 0.5542990218134582,
+      "learning_rate": 4.999575676852174e-06,
+      "loss": 0.6897,
+      "step": 764
+    },
+    {
+      "epoch": 0.30229686342306744,
+      "grad_norm": 0.585799531977493,
+      "learning_rate": 4.9995727854734e-06,
+      "loss": 0.6721,
+      "step": 765
+    },
+    {
+      "epoch": 0.3026920227216597,
+      "grad_norm": 0.7770451245382635,
+      "learning_rate": 4.9995698842778225e-06,
+      "loss": 0.6981,
+      "step": 766
+    },
+    {
+      "epoch": 0.3030871820202519,
+      "grad_norm": 0.5471559718041984,
+      "learning_rate": 4.999566973265452e-06,
+      "loss": 0.6524,
+      "step": 767
+    },
+    {
+      "epoch": 0.30348234131884416,
+      "grad_norm": 0.5680704664062982,
+      "learning_rate": 4.999564052436301e-06,
+      "loss": 0.6481,
+      "step": 768
+    },
+    {
+      "epoch": 0.3038775006174364,
+      "grad_norm": 0.5983428486623749,
+      "learning_rate": 4.999561121790379e-06,
+      "loss": 0.6848,
+      "step": 769
+    },
+    {
+      "epoch": 0.30427265991602864,
+      "grad_norm": 0.5641224737953592,
+      "learning_rate": 4.9995581813276995e-06,
+      "loss": 0.6703,
+      "step": 770
+    },
+    {
+      "epoch": 0.3046678192146209,
+      "grad_norm": 0.5647226711303925,
+      "learning_rate": 4.999555231048273e-06,
+      "loss": 0.7158,
+      "step": 771
+    },
+    {
+      "epoch": 0.3050629785132131,
+      "grad_norm": 0.7023254822593299,
+      "learning_rate": 4.999552270952112e-06,
+      "loss": 0.684,
+      "step": 772
+    },
+    {
+      "epoch": 0.30545813781180536,
+      "grad_norm": 0.6146604133584318,
+      "learning_rate": 4.999549301039226e-06,
+      "loss": 0.7085,
+      "step": 773
+    },
+    {
+      "epoch": 0.3058532971103976,
+      "grad_norm": 0.562728967949529,
+      "learning_rate": 4.999546321309629e-06,
+      "loss": 0.6681,
+      "step": 774
+    },
+    {
+      "epoch": 0.3062484564089899,
+      "grad_norm": 1.4500324454053806,
+      "learning_rate": 4.999543331763332e-06,
+      "loss": 0.6917,
+      "step": 775
+    },
+    {
+      "epoch": 0.30664361570758214,
+      "grad_norm": 0.6576023968965049,
+      "learning_rate": 4.999540332400346e-06,
+      "loss": 0.6953,
+      "step": 776
+    },
+    {
+      "epoch": 0.3070387750061744,
+      "grad_norm": 0.5706530250137147,
+      "learning_rate": 4.999537323220684e-06,
+      "loss": 0.6794,
+      "step": 777
+    },
+    {
+      "epoch": 0.3074339343047666,
+      "grad_norm": 0.5428676827482191,
+      "learning_rate": 4.999534304224357e-06,
+      "loss": 0.6851,
+      "step": 778
+    },
+    {
+      "epoch": 0.30782909360335886,
+      "grad_norm": 0.5981739162493036,
+      "learning_rate": 4.999531275411377e-06,
+      "loss": 0.671,
+      "step": 779
+    },
+    {
+      "epoch": 0.3082242529019511,
+      "grad_norm": 0.6170758554603312,
+      "learning_rate": 4.999528236781757e-06,
+      "loss": 0.6957,
+      "step": 780
+    },
+    {
+      "epoch": 0.30861941220054334,
+      "grad_norm": 0.5823666001801212,
+      "learning_rate": 4.999525188335507e-06,
+      "loss": 0.6719,
+      "step": 781
+    },
+    {
+      "epoch": 0.3090145714991356,
+      "grad_norm": 0.5997874539667468,
+      "learning_rate": 4.99952213007264e-06,
+      "loss": 0.6917,
+      "step": 782
+    },
+    {
+      "epoch": 0.3094097307977278,
+      "grad_norm": 0.5821609633692683,
+      "learning_rate": 4.999519061993166e-06,
+      "loss": 0.6838,
+      "step": 783
+    },
+    {
+      "epoch": 0.30980489009632006,
+      "grad_norm": 0.5479337929157194,
+      "learning_rate": 4.999515984097101e-06,
+      "loss": 0.6837,
+      "step": 784
+    },
+    {
+      "epoch": 0.3102000493949123,
+      "grad_norm": 0.5937117790026203,
+      "learning_rate": 4.999512896384455e-06,
+      "loss": 0.6899,
+      "step": 785
+    },
+    {
+      "epoch": 0.31059520869350454,
+      "grad_norm": 0.5782838146122935,
+      "learning_rate": 4.99950979885524e-06,
+      "loss": 0.6802,
+      "step": 786
+    },
+    {
+      "epoch": 0.31099036799209684,
+      "grad_norm": 0.546006734321019,
+      "learning_rate": 4.999506691509467e-06,
+      "loss": 0.6776,
+      "step": 787
+    },
+    {
+      "epoch": 0.3113855272906891,
+      "grad_norm": 0.5621083736104908,
+      "learning_rate": 4.999503574347151e-06,
+      "loss": 0.679,
+      "step": 788
+    },
+    {
+      "epoch": 0.3117806865892813,
+      "grad_norm": 0.6018666658407275,
+      "learning_rate": 4.9995004473683016e-06,
+      "loss": 0.7095,
+      "step": 789
+    },
+    {
+      "epoch": 0.31217584588787356,
+      "grad_norm": 0.5665313992465836,
+      "learning_rate": 4.999497310572933e-06,
+      "loss": 0.6906,
+      "step": 790
+    },
+    {
+      "epoch": 0.3125710051864658,
+      "grad_norm": 0.5795816586087552,
+      "learning_rate": 4.999494163961057e-06,
+      "loss": 0.6792,
+      "step": 791
+    },
+    {
+      "epoch": 0.31296616448505804,
+      "grad_norm": 0.5868236665755461,
+      "learning_rate": 4.999491007532685e-06,
+      "loss": 0.6851,
+      "step": 792
+    },
+    {
+      "epoch": 0.3133613237836503,
+      "grad_norm": 0.5722398773037819,
+      "learning_rate": 4.999487841287831e-06,
+      "loss": 0.6667,
+      "step": 793
+    },
+    {
+      "epoch": 0.3137564830822425,
+      "grad_norm": 0.5946352610623089,
+      "learning_rate": 4.999484665226506e-06,
+      "loss": 0.6909,
+      "step": 794
+    },
+    {
+      "epoch": 0.31415164238083476,
+      "grad_norm": 1.2564866117315103,
+      "learning_rate": 4.999481479348722e-06,
+      "loss": 0.6852,
+      "step": 795
+    },
+    {
+      "epoch": 0.314546801679427,
+      "grad_norm": 0.5575268193016416,
+      "learning_rate": 4.999478283654494e-06,
+      "loss": 0.6742,
+      "step": 796
+    },
+    {
+      "epoch": 0.31494196097801924,
+      "grad_norm": 0.5545559884431172,
+      "learning_rate": 4.999475078143832e-06,
+      "loss": 0.6743,
+      "step": 797
+    },
+    {
+      "epoch": 0.3153371202766115,
+      "grad_norm": 0.5460260402768534,
+      "learning_rate": 4.999471862816751e-06,
+      "loss": 0.6682,
+      "step": 798
+    },
+    {
+      "epoch": 0.3157322795752038,
+      "grad_norm": 0.545727768192459,
+      "learning_rate": 4.9994686376732605e-06,
+      "loss": 0.675,
+      "step": 799
+    },
+    {
+      "epoch": 0.316127438873796,
+      "grad_norm": 0.5493756349564834,
+      "learning_rate": 4.999465402713376e-06,
+      "loss": 0.6785,
+      "step": 800
+    },
+    {
+      "epoch": 0.31652259817238826,
+      "grad_norm": 0.5521530157011729,
+      "learning_rate": 4.99946215793711e-06,
+      "loss": 0.6723,
+      "step": 801
+    },
+    {
+      "epoch": 0.3169177574709805,
+      "grad_norm": 0.6919194319875192,
+      "learning_rate": 4.999458903344473e-06,
+      "loss": 0.6801,
+      "step": 802
+    },
+    {
+      "epoch": 0.31731291676957274,
+      "grad_norm": 0.5635837786929486,
+      "learning_rate": 4.99945563893548e-06,
+      "loss": 0.6983,
+      "step": 803
+    },
+    {
+      "epoch": 0.317708076068165,
+      "grad_norm": 0.5577079529859765,
+      "learning_rate": 4.999452364710142e-06,
+      "loss": 0.6869,
+      "step": 804
+    },
+    {
+      "epoch": 0.3181032353667572,
+      "grad_norm": 0.522245723084247,
+      "learning_rate": 4.999449080668473e-06,
+      "loss": 0.6711,
+      "step": 805
+    },
+    {
+      "epoch": 0.31849839466534946,
+      "grad_norm": 0.5543341617543934,
+      "learning_rate": 4.999445786810486e-06,
+      "loss": 0.6528,
+      "step": 806
+    },
+    {
+      "epoch": 0.3188935539639417,
+      "grad_norm": 0.5404102043718793,
+      "learning_rate": 4.999442483136194e-06,
+      "loss": 0.6706,
+      "step": 807
+    },
+    {
+      "epoch": 0.31928871326253394,
+      "grad_norm": 0.5463105642491327,
+      "learning_rate": 4.99943916964561e-06,
+      "loss": 0.6675,
+      "step": 808
+    },
+    {
+      "epoch": 0.3196838725611262,
+      "grad_norm": 0.6177688829664617,
+      "learning_rate": 4.999435846338746e-06,
+      "loss": 0.6753,
+      "step": 809
+    },
+    {
+      "epoch": 0.3200790318597184,
+      "grad_norm": 0.5490308724780782,
+      "learning_rate": 4.999432513215616e-06,
+      "loss": 0.6751,
+      "step": 810
+    },
+    {
+      "epoch": 0.3204741911583107,
+      "grad_norm": 0.6090150943921979,
+      "learning_rate": 4.999429170276233e-06,
+      "loss": 0.6852,
+      "step": 811
+    },
+    {
+      "epoch": 0.32086935045690296,
+      "grad_norm": 0.5611702933211862,
+      "learning_rate": 4.99942581752061e-06,
+      "loss": 0.6857,
+      "step": 812
+    },
+    {
+      "epoch": 0.3212645097554952,
+      "grad_norm": 0.5471158991492043,
+      "learning_rate": 4.999422454948759e-06,
+      "loss": 0.668,
+      "step": 813
+    },
+    {
+      "epoch": 0.32165966905408744,
+      "grad_norm": 0.6040235246117969,
+      "learning_rate": 4.9994190825606955e-06,
+      "loss": 0.6738,
+      "step": 814
+    },
+    {
+      "epoch": 0.3220548283526797,
+      "grad_norm": 0.5461114975582116,
+      "learning_rate": 4.999415700356431e-06,
+      "loss": 0.6935,
+      "step": 815
+    },
+    {
+      "epoch": 0.3224499876512719,
+      "grad_norm": 0.5531585597793667,
+      "learning_rate": 4.99941230833598e-06,
+      "loss": 0.6964,
+      "step": 816
+    },
+    {
+      "epoch": 0.32284514694986416,
+      "grad_norm": 0.555832857990718,
+      "learning_rate": 4.999408906499355e-06,
+      "loss": 0.6878,
+      "step": 817
+    },
+    {
+      "epoch": 0.3232403062484564,
+      "grad_norm": 0.5159924465583967,
+      "learning_rate": 4.999405494846569e-06,
+      "loss": 0.6666,
+      "step": 818
+    },
+    {
+      "epoch": 0.32363546554704864,
+      "grad_norm": 0.5242278141632275,
+      "learning_rate": 4.999402073377637e-06,
+      "loss": 0.6556,
+      "step": 819
+    },
+    {
+      "epoch": 0.3240306248456409,
+      "grad_norm": 0.5485685565466117,
+      "learning_rate": 4.99939864209257e-06,
+      "loss": 0.6867,
+      "step": 820
+    },
+    {
+      "epoch": 0.3244257841442331,
+      "grad_norm": 0.5599138503166448,
+      "learning_rate": 4.999395200991384e-06,
+      "loss": 0.6741,
+      "step": 821
+    },
+    {
+      "epoch": 0.3248209434428254,
+      "grad_norm": 0.5314314737207193,
+      "learning_rate": 4.999391750074091e-06,
+      "loss": 0.6597,
+      "step": 822
+    },
+    {
+      "epoch": 0.32521610274141766,
+      "grad_norm": 0.5505001683410817,
+      "learning_rate": 4.999388289340705e-06,
+      "loss": 0.6734,
+      "step": 823
+    },
+    {
+      "epoch": 0.3256112620400099,
+      "grad_norm": 0.5483028368326526,
+      "learning_rate": 4.999384818791239e-06,
+      "loss": 0.6871,
+      "step": 824
+    },
+    {
+      "epoch": 0.32600642133860214,
+      "grad_norm": 0.5553659611188733,
+      "learning_rate": 4.9993813384257075e-06,
+      "loss": 0.6879,
+      "step": 825
+    },
+    {
+      "epoch": 0.3264015806371944,
+      "grad_norm": 0.5508245461247865,
+      "learning_rate": 4.999377848244123e-06,
+      "loss": 0.6608,
+      "step": 826
+    },
+    {
+      "epoch": 0.3267967399357866,
+      "grad_norm": 0.5468555275628424,
+      "learning_rate": 4.9993743482465015e-06,
+      "loss": 0.6885,
+      "step": 827
+    },
+    {
+      "epoch": 0.32719189923437886,
+      "grad_norm": 0.5518450089645637,
+      "learning_rate": 4.999370838432854e-06,
+      "loss": 0.6784,
+      "step": 828
+    },
+    {
+      "epoch": 0.3275870585329711,
+      "grad_norm": 0.5381677755962232,
+      "learning_rate": 4.999367318803197e-06,
+      "loss": 0.6813,
+      "step": 829
+    },
+    {
+      "epoch": 0.32798221783156334,
+      "grad_norm": 0.5400069304709476,
+      "learning_rate": 4.999363789357542e-06,
+      "loss": 0.6664,
+      "step": 830
+    },
+    {
+      "epoch": 0.3283773771301556,
+      "grad_norm": 0.6129541548516995,
+      "learning_rate": 4.999360250095903e-06,
+      "loss": 0.6813,
+      "step": 831
+    },
+    {
+      "epoch": 0.3287725364287478,
+      "grad_norm": 0.551379733965531,
+      "learning_rate": 4.999356701018295e-06,
+      "loss": 0.6698,
+      "step": 832
+    },
+    {
+      "epoch": 0.32916769572734006,
+      "grad_norm": 0.530243445703882,
+      "learning_rate": 4.999353142124732e-06,
+      "loss": 0.682,
+      "step": 833
+    },
+    {
+      "epoch": 0.32956285502593236,
+      "grad_norm": 0.5348358317715631,
+      "learning_rate": 4.999349573415227e-06,
+      "loss": 0.6489,
+      "step": 834
+    },
+    {
+      "epoch": 0.3299580143245246,
+      "grad_norm": 0.5547085181902088,
+      "learning_rate": 4.999345994889796e-06,
+      "loss": 0.6938,
+      "step": 835
+    },
+    {
+      "epoch": 0.33035317362311684,
+      "grad_norm": 0.5243281588616451,
+      "learning_rate": 4.99934240654845e-06,
+      "loss": 0.6671,
+      "step": 836
+    },
+    {
+      "epoch": 0.3307483329217091,
+      "grad_norm": 0.5493088801401717,
+      "learning_rate": 4.9993388083912054e-06,
+      "loss": 0.6706,
+      "step": 837
+    },
+    {
+      "epoch": 0.3311434922203013,
+      "grad_norm": 0.545645002523389,
+      "learning_rate": 4.999335200418076e-06,
+      "loss": 0.6901,
+      "step": 838
+    },
+    {
+      "epoch": 0.33153865151889356,
+      "grad_norm": 0.5613907384738281,
+      "learning_rate": 4.999331582629075e-06,
+      "loss": 0.6908,
+      "step": 839
+    },
+    {
+      "epoch": 0.3319338108174858,
+      "grad_norm": 0.6000984842384804,
+      "learning_rate": 4.999327955024217e-06,
+      "loss": 0.6828,
+      "step": 840
+    },
+    {
+      "epoch": 0.33232897011607804,
+      "grad_norm": 0.51885138747256,
+      "learning_rate": 4.9993243176035175e-06,
+      "loss": 0.6597,
+      "step": 841
+    },
+    {
+      "epoch": 0.3327241294146703,
+      "grad_norm": 0.6231011537404787,
+      "learning_rate": 4.999320670366989e-06,
+      "loss": 0.6571,
+      "step": 842
+    },
+    {
+      "epoch": 0.3331192887132625,
+      "grad_norm": 0.5214424709505693,
+      "learning_rate": 4.999317013314646e-06,
+      "loss": 0.673,
+      "step": 843
+    },
+    {
+      "epoch": 0.33351444801185476,
+      "grad_norm": 0.5410014959338921,
+      "learning_rate": 4.999313346446505e-06,
+      "loss": 0.6747,
+      "step": 844
+    },
+    {
+      "epoch": 0.333909607310447,
+      "grad_norm": 0.6300322367055629,
+      "learning_rate": 4.999309669762578e-06,
+      "loss": 0.6944,
+      "step": 845
+    },
+    {
+      "epoch": 0.3343047666090393,
+      "grad_norm": 0.6141171186975185,
+      "learning_rate": 4.99930598326288e-06,
+      "loss": 0.6746,
+      "step": 846
+    },
+    {
+      "epoch": 0.33469992590763153,
+      "grad_norm": 0.5592903856619154,
+      "learning_rate": 4.9993022869474264e-06,
+      "loss": 0.6846,
+      "step": 847
+    },
+    {
+      "epoch": 0.3350950852062238,
+      "grad_norm": 0.5914019521534136,
+      "learning_rate": 4.99929858081623e-06,
+      "loss": 0.6562,
+      "step": 848
+    },
+    {
+      "epoch": 0.335490244504816,
+      "grad_norm": 0.5556986261868628,
+      "learning_rate": 4.999294864869307e-06,
+      "loss": 0.695,
+      "step": 849
+    },
+    {
+      "epoch": 0.33588540380340826,
+      "grad_norm": 0.5361036550501209,
+      "learning_rate": 4.999291139106672e-06,
+      "loss": 0.6733,
+      "step": 850
+    },
+    {
+      "epoch": 0.3362805631020005,
+      "grad_norm": 0.5663681947389955,
+      "learning_rate": 4.9992874035283375e-06,
+      "loss": 0.6727,
+      "step": 851
+    },
+    {
+      "epoch": 0.33667572240059274,
+      "grad_norm": 0.5385125766471879,
+      "learning_rate": 4.999283658134322e-06,
+      "loss": 0.6717,
+      "step": 852
+    },
+    {
+      "epoch": 0.337070881699185,
+      "grad_norm": 0.58237023128874,
+      "learning_rate": 4.999279902924636e-06,
+      "loss": 0.6751,
+      "step": 853
+    },
+    {
+      "epoch": 0.3374660409977772,
+      "grad_norm": 0.6219007403124622,
+      "learning_rate": 4.999276137899297e-06,
+      "loss": 0.6515,
+      "step": 854
+    },
+    {
+      "epoch": 0.33786120029636946,
+      "grad_norm": 0.5581232382993533,
+      "learning_rate": 4.999272363058319e-06,
+      "loss": 0.6876,
+      "step": 855
+    },
+    {
+      "epoch": 0.3382563595949617,
+      "grad_norm": 0.5570097230753517,
+      "learning_rate": 4.999268578401717e-06,
+      "loss": 0.6696,
+      "step": 856
+    },
+    {
+      "epoch": 0.33865151889355394,
+      "grad_norm": 0.5502292291243254,
+      "learning_rate": 4.999264783929505e-06,
+      "loss": 0.6651,
+      "step": 857
+    },
+    {
+      "epoch": 0.33904667819214623,
+      "grad_norm": 0.5289187848980834,
+      "learning_rate": 4.999260979641699e-06,
+      "loss": 0.6586,
+      "step": 858
+    },
+    {
+      "epoch": 0.3394418374907385,
+      "grad_norm": 0.5677219337773182,
+      "learning_rate": 4.999257165538314e-06,
+      "loss": 0.6764,
+      "step": 859
+    },
+    {
+      "epoch": 0.3398369967893307,
+      "grad_norm": 0.6248356948027501,
+      "learning_rate": 4.999253341619363e-06,
+      "loss": 0.6685,
+      "step": 860
+    },
+    {
+      "epoch": 0.34023215608792295,
+      "grad_norm": 0.5245870998609313,
+      "learning_rate": 4.999249507884864e-06,
+      "loss": 0.6637,
+      "step": 861
+    },
+    {
+      "epoch": 0.3406273153865152,
+      "grad_norm": 0.5633998533768565,
+      "learning_rate": 4.9992456643348296e-06,
+      "loss": 0.6626,
+      "step": 862
+    },
+    {
+      "epoch": 0.34102247468510744,
+      "grad_norm": 0.6163731317461442,
+      "learning_rate": 4.999241810969276e-06,
+      "loss": 0.7031,
+      "step": 863
+    },
+    {
+      "epoch": 0.3414176339836997,
+      "grad_norm": 0.5319060256897468,
+      "learning_rate": 4.999237947788218e-06,
+      "loss": 0.6661,
+      "step": 864
+    },
+    {
+      "epoch": 0.3418127932822919,
+      "grad_norm": 0.5747234912624848,
+      "learning_rate": 4.999234074791673e-06,
+      "loss": 0.6637,
+      "step": 865
+    },
+    {
+      "epoch": 0.34220795258088416,
+      "grad_norm": 0.6217626123594077,
+      "learning_rate": 4.9992301919796515e-06,
+      "loss": 0.6956,
+      "step": 866
+    },
+    {
+      "epoch": 0.3426031118794764,
+      "grad_norm": 0.5291571135659683,
+      "learning_rate": 4.999226299352172e-06,
+      "loss": 0.6592,
+      "step": 867
+    },
+    {
+      "epoch": 0.34299827117806864,
+      "grad_norm": 0.5929800922433651,
+      "learning_rate": 4.99922239690925e-06,
+      "loss": 0.6758,
+      "step": 868
+    },
+    {
+      "epoch": 0.3433934304766609,
+      "grad_norm": 0.8263016280842782,
+      "learning_rate": 4.999218484650899e-06,
+      "loss": 0.6924,
+      "step": 869
+    },
+    {
+      "epoch": 0.3437885897752532,
+      "grad_norm": 0.5373424767279488,
+      "learning_rate": 4.999214562577137e-06,
+      "loss": 0.6778,
+      "step": 870
+    },
+    {
+      "epoch": 0.3441837490738454,
+      "grad_norm": 0.6161942738080999,
+      "learning_rate": 4.999210630687976e-06,
+      "loss": 0.6868,
+      "step": 871
+    },
+    {
+      "epoch": 0.34457890837243765,
+      "grad_norm": 0.5544247743850119,
+      "learning_rate": 4.999206688983435e-06,
+      "loss": 0.6472,
+      "step": 872
+    },
+    {
+      "epoch": 0.3449740676710299,
+      "grad_norm": 0.5436735454536356,
+      "learning_rate": 4.9992027374635265e-06,
+      "loss": 0.6681,
+      "step": 873
+    },
+    {
+      "epoch": 0.34536922696962213,
+      "grad_norm": 0.5870467900357642,
+      "learning_rate": 4.999198776128268e-06,
+      "loss": 0.6488,
+      "step": 874
+    },
+    {
+      "epoch": 0.3457643862682144,
+      "grad_norm": 0.5226483330016569,
+      "learning_rate": 4.999194804977674e-06,
+      "loss": 0.6474,
+      "step": 875
+    },
+    {
+      "epoch": 0.3461595455668066,
+      "grad_norm": 0.5480980063898397,
+      "learning_rate": 4.99919082401176e-06,
+      "loss": 0.6809,
+      "step": 876
+    },
+    {
+      "epoch": 0.34655470486539885,
+      "grad_norm": 0.5845327636648194,
+      "learning_rate": 4.999186833230542e-06,
+      "loss": 0.6461,
+      "step": 877
+    },
+    {
+      "epoch": 0.3469498641639911,
+      "grad_norm": 0.700267095825688,
+      "learning_rate": 4.999182832634036e-06,
+      "loss": 0.6864,
+      "step": 878
+    },
+    {
+      "epoch": 0.34734502346258334,
+      "grad_norm": 0.5704163766747798,
+      "learning_rate": 4.999178822222258e-06,
+      "loss": 0.6624,
+      "step": 879
+    },
+    {
+      "epoch": 0.3477401827611756,
+      "grad_norm": 0.5688194389353117,
+      "learning_rate": 4.999174801995222e-06,
+      "loss": 0.6597,
+      "step": 880
+    },
+    {
+      "epoch": 0.34813534205976787,
+      "grad_norm": 0.5210130312006719,
+      "learning_rate": 4.999170771952946e-06,
+      "loss": 0.6618,
+      "step": 881
+    },
+    {
+      "epoch": 0.3485305013583601,
+      "grad_norm": 0.5429969490828442,
+      "learning_rate": 4.999166732095445e-06,
+      "loss": 0.6677,
+      "step": 882
+    },
+    {
+      "epoch": 0.34892566065695235,
+      "grad_norm": 0.5679202548589564,
+      "learning_rate": 4.999162682422733e-06,
+      "loss": 0.6668,
+      "step": 883
+    },
+    {
+      "epoch": 0.3493208199555446,
+      "grad_norm": 0.5391777536262067,
+      "learning_rate": 4.999158622934829e-06,
+      "loss": 0.6704,
+      "step": 884
+    },
+    {
+      "epoch": 0.34971597925413683,
+      "grad_norm": 0.5622016243906001,
+      "learning_rate": 4.999154553631748e-06,
+      "loss": 0.6435,
+      "step": 885
+    },
+    {
+      "epoch": 0.3501111385527291,
+      "grad_norm": 0.5628428119759383,
+      "learning_rate": 4.999150474513504e-06,
+      "loss": 0.6618,
+      "step": 886
+    },
+    {
+      "epoch": 0.3505062978513213,
+      "grad_norm": 0.5303772562973009,
+      "learning_rate": 4.999146385580114e-06,
+      "loss": 0.6628,
+      "step": 887
+    },
+    {
+      "epoch": 0.35090145714991355,
+      "grad_norm": 0.5813782031762684,
+      "learning_rate": 4.999142286831596e-06,
+      "loss": 0.6876,
+      "step": 888
+    },
+    {
+      "epoch": 0.3512966164485058,
+      "grad_norm": 0.563823831771851,
+      "learning_rate": 4.999138178267965e-06,
+      "loss": 0.6604,
+      "step": 889
+    },
+    {
+      "epoch": 0.35169177574709803,
+      "grad_norm": 0.5253757499381441,
+      "learning_rate": 4.999134059889236e-06,
+      "loss": 0.6639,
+      "step": 890
+    },
+    {
+      "epoch": 0.3520869350456903,
+      "grad_norm": 0.5593726055000605,
+      "learning_rate": 4.9991299316954255e-06,
+      "loss": 0.6476,
+      "step": 891
+    },
+    {
+      "epoch": 0.3524820943442825,
+      "grad_norm": 0.5367006366591701,
+      "learning_rate": 4.9991257936865515e-06,
+      "loss": 0.6937,
+      "step": 892
+    },
+    {
+      "epoch": 0.3528772536428748,
+      "grad_norm": 0.5567610106868094,
+      "learning_rate": 4.999121645862628e-06,
+      "loss": 0.6731,
+      "step": 893
+    },
+    {
+      "epoch": 0.35327241294146705,
+      "grad_norm": 0.5324611058908338,
+      "learning_rate": 4.999117488223672e-06,
+      "loss": 0.6735,
+      "step": 894
+    },
+    {
+      "epoch": 0.3536675722400593,
+      "grad_norm": 0.845928580999717,
+      "learning_rate": 4.999113320769701e-06,
+      "loss": 0.6675,
+      "step": 895
+    },
+    {
+      "epoch": 0.35406273153865153,
+      "grad_norm": 0.5597594694408724,
+      "learning_rate": 4.999109143500729e-06,
+      "loss": 0.6982,
+      "step": 896
+    },
+    {
+      "epoch": 0.35445789083724377,
+      "grad_norm": 0.5742575981877034,
+      "learning_rate": 4.999104956416775e-06,
+      "loss": 0.6431,
+      "step": 897
+    },
+    {
+      "epoch": 0.354853050135836,
+      "grad_norm": 0.5446258761551195,
+      "learning_rate": 4.999100759517854e-06,
+      "loss": 0.6749,
+      "step": 898
+    },
+    {
+      "epoch": 0.35524820943442825,
+      "grad_norm": 0.5292659599253334,
+      "learning_rate": 4.999096552803983e-06,
+      "loss": 0.6775,
+      "step": 899
+    },
+    {
+      "epoch": 0.3556433687330205,
+      "grad_norm": 0.5292261917552322,
+      "learning_rate": 4.9990923362751776e-06,
+      "loss": 0.6716,
+      "step": 900
+    },
+    {
+      "epoch": 0.35603852803161273,
+      "grad_norm": 0.5470660498231109,
+      "learning_rate": 4.999088109931456e-06,
+      "loss": 0.6599,
+      "step": 901
+    },
+    {
+      "epoch": 0.356433687330205,
+      "grad_norm": 0.5338026521581444,
+      "learning_rate": 4.999083873772833e-06,
+      "loss": 0.6691,
+      "step": 902
+    },
+    {
+      "epoch": 0.3568288466287972,
+      "grad_norm": 0.5613881734474877,
+      "learning_rate": 4.9990796277993255e-06,
+      "loss": 0.6798,
+      "step": 903
+    },
+    {
+      "epoch": 0.35722400592738945,
+      "grad_norm": 0.5478029825120281,
+      "learning_rate": 4.999075372010952e-06,
+      "loss": 0.6623,
+      "step": 904
+    },
+    {
+      "epoch": 0.35761916522598175,
+      "grad_norm": 0.5165063149329737,
+      "learning_rate": 4.999071106407728e-06,
+      "loss": 0.661,
+      "step": 905
+    },
+    {
+      "epoch": 0.358014324524574,
+      "grad_norm": 0.5992216629583358,
+      "learning_rate": 4.999066830989669e-06,
+      "loss": 0.653,
+      "step": 906
+    },
+    {
+      "epoch": 0.35840948382316623,
+      "grad_norm": 0.5526267591056174,
+      "learning_rate": 4.999062545756794e-06,
+      "loss": 0.6657,
+      "step": 907
+    },
+    {
+      "epoch": 0.35880464312175847,
+      "grad_norm": 0.5231560128614834,
+      "learning_rate": 4.999058250709119e-06,
+      "loss": 0.6427,
+      "step": 908
+    },
+    {
+      "epoch": 0.3591998024203507,
+      "grad_norm": 0.5496671975565968,
+      "learning_rate": 4.99905394584666e-06,
+      "loss": 0.6728,
+      "step": 909
+    },
+    {
+      "epoch": 0.35959496171894295,
+      "grad_norm": 0.5814232320207219,
+      "learning_rate": 4.999049631169435e-06,
+      "loss": 0.6716,
+      "step": 910
+    },
+    {
+      "epoch": 0.3599901210175352,
+      "grad_norm": 0.5284531514261371,
+      "learning_rate": 4.99904530667746e-06,
+      "loss": 0.6784,
+      "step": 911
+    },
+    {
+      "epoch": 0.36038528031612743,
+      "grad_norm": 0.5830847990400584,
+      "learning_rate": 4.999040972370753e-06,
+      "loss": 0.6728,
+      "step": 912
+    },
+    {
+      "epoch": 0.3607804396147197,
+      "grad_norm": 0.5265753919458841,
+      "learning_rate": 4.999036628249331e-06,
+      "loss": 0.6958,
+      "step": 913
+    },
+    {
+      "epoch": 0.3611755989133119,
+      "grad_norm": 0.528588620481184,
+      "learning_rate": 4.99903227431321e-06,
+      "loss": 0.6548,
+      "step": 914
+    },
+    {
+      "epoch": 0.36157075821190415,
+      "grad_norm": 0.5182035242903082,
+      "learning_rate": 4.9990279105624076e-06,
+      "loss": 0.6598,
+      "step": 915
+    },
+    {
+      "epoch": 0.3619659175104964,
+      "grad_norm": 0.5286689369003267,
+      "learning_rate": 4.9990235369969406e-06,
+      "loss": 0.6872,
+      "step": 916
+    },
+    {
+      "epoch": 0.3623610768090887,
+      "grad_norm": 0.5386957143651905,
+      "learning_rate": 4.9990191536168274e-06,
+      "loss": 0.6491,
+      "step": 917
+    },
+    {
+      "epoch": 0.36275623610768093,
+      "grad_norm": 0.5514686370106001,
+      "learning_rate": 4.999014760422085e-06,
+      "loss": 0.6651,
+      "step": 918
+    },
+    {
+      "epoch": 0.36315139540627317,
+      "grad_norm": 0.5648373707016645,
+      "learning_rate": 4.999010357412729e-06,
+      "loss": 0.6731,
+      "step": 919
+    },
+    {
+      "epoch": 0.3635465547048654,
+      "grad_norm": 0.5152295186463809,
+      "learning_rate": 4.999005944588779e-06,
+      "loss": 0.6428,
+      "step": 920
+    },
+    {
+      "epoch": 0.36394171400345765,
+      "grad_norm": 0.5066022878135411,
+      "learning_rate": 4.9990015219502505e-06,
+      "loss": 0.658,
+      "step": 921
+    },
+    {
+      "epoch": 0.3643368733020499,
+      "grad_norm": 0.5428977376504636,
+      "learning_rate": 4.998997089497161e-06,
+      "loss": 0.6613,
+      "step": 922
+    },
+    {
+      "epoch": 0.36473203260064213,
+      "grad_norm": 0.5678413901240047,
+      "learning_rate": 4.998992647229529e-06,
+      "loss": 0.6737,
+      "step": 923
+    },
+    {
+      "epoch": 0.36512719189923437,
+      "grad_norm": 0.536401567098352,
+      "learning_rate": 4.9989881951473706e-06,
+      "loss": 0.652,
+      "step": 924
+    },
+    {
+      "epoch": 0.3655223511978266,
+      "grad_norm": 0.5768953101789457,
+      "learning_rate": 4.998983733250705e-06,
+      "loss": 0.6314,
+      "step": 925
+    },
+    {
+      "epoch": 0.36591751049641885,
+      "grad_norm": 0.5352996178968072,
+      "learning_rate": 4.998979261539548e-06,
+      "loss": 0.6683,
+      "step": 926
+    },
+    {
+      "epoch": 0.3663126697950111,
+      "grad_norm": 0.5174696386468567,
+      "learning_rate": 4.998974780013919e-06,
+      "loss": 0.6607,
+      "step": 927
+    },
+    {
+      "epoch": 0.36670782909360333,
+      "grad_norm": 0.6030535589781871,
+      "learning_rate": 4.998970288673833e-06,
+      "loss": 0.6578,
+      "step": 928
+    },
+    {
+      "epoch": 0.36710298839219563,
+      "grad_norm": 0.5500821850702614,
+      "learning_rate": 4.99896578751931e-06,
+      "loss": 0.6622,
+      "step": 929
+    },
+    {
+      "epoch": 0.36749814769078787,
+      "grad_norm": 0.5283486083712549,
+      "learning_rate": 4.998961276550367e-06,
+      "loss": 0.6518,
+      "step": 930
+    },
+    {
+      "epoch": 0.3678933069893801,
+      "grad_norm": 0.5663497776849602,
+      "learning_rate": 4.998956755767021e-06,
+      "loss": 0.6697,
+      "step": 931
+    },
+    {
+      "epoch": 0.36828846628797235,
+      "grad_norm": 0.5211239249646628,
+      "learning_rate": 4.998952225169291e-06,
+      "loss": 0.6658,
+      "step": 932
+    },
+    {
+      "epoch": 0.3686836255865646,
+      "grad_norm": 0.5375278433902325,
+      "learning_rate": 4.9989476847571935e-06,
+      "loss": 0.6672,
+      "step": 933
+    },
+    {
+      "epoch": 0.36907878488515683,
+      "grad_norm": 0.5125111796062831,
+      "learning_rate": 4.998943134530748e-06,
+      "loss": 0.6664,
+      "step": 934
+    },
+    {
+      "epoch": 0.36947394418374907,
+      "grad_norm": 0.7149723112166187,
+      "learning_rate": 4.9989385744899705e-06,
+      "loss": 0.6792,
+      "step": 935
+    },
+    {
+      "epoch": 0.3698691034823413,
+      "grad_norm": 0.5195495635576596,
+      "learning_rate": 4.99893400463488e-06,
+      "loss": 0.6424,
+      "step": 936
+    },
+    {
+      "epoch": 0.37026426278093355,
+      "grad_norm": 0.5261355467385814,
+      "learning_rate": 4.998929424965494e-06,
+      "loss": 0.6794,
+      "step": 937
+    },
+    {
+      "epoch": 0.3706594220795258,
+      "grad_norm": 0.5371603404887052,
+      "learning_rate": 4.99892483548183e-06,
+      "loss": 0.6556,
+      "step": 938
+    },
+    {
+      "epoch": 0.37105458137811803,
+      "grad_norm": 0.5288627513817326,
+      "learning_rate": 4.998920236183908e-06,
+      "loss": 0.6408,
+      "step": 939
+    },
+    {
+      "epoch": 0.37144974067671027,
+      "grad_norm": 0.5366598263301546,
+      "learning_rate": 4.998915627071743e-06,
+      "loss": 0.6745,
+      "step": 940
+    },
+    {
+      "epoch": 0.37184489997530257,
+      "grad_norm": 0.5209858322992073,
+      "learning_rate": 4.998911008145357e-06,
+      "loss": 0.6716,
+      "step": 941
+    },
+    {
+      "epoch": 0.3722400592738948,
+      "grad_norm": 0.5092473497788766,
+      "learning_rate": 4.998906379404764e-06,
+      "loss": 0.6588,
+      "step": 942
+    },
+    {
+      "epoch": 0.37263521857248705,
+      "grad_norm": 0.5301240404384194,
+      "learning_rate": 4.998901740849985e-06,
+      "loss": 0.6475,
+      "step": 943
+    },
+    {
+      "epoch": 0.3730303778710793,
+      "grad_norm": 0.5369766245496886,
+      "learning_rate": 4.998897092481037e-06,
+      "loss": 0.6704,
+      "step": 944
+    },
+    {
+      "epoch": 0.37342553716967153,
+      "grad_norm": 0.5677722223194425,
+      "learning_rate": 4.998892434297939e-06,
+      "loss": 0.6489,
+      "step": 945
+    },
+    {
+      "epoch": 0.37382069646826377,
+      "grad_norm": 0.5231100890683938,
+      "learning_rate": 4.998887766300708e-06,
+      "loss": 0.6746,
+      "step": 946
+    },
+    {
+      "epoch": 0.374215855766856,
+      "grad_norm": 0.5295898234845133,
+      "learning_rate": 4.998883088489365e-06,
+      "loss": 0.6814,
+      "step": 947
+    },
+    {
+      "epoch": 0.37461101506544825,
+      "grad_norm": 0.5785367347699937,
+      "learning_rate": 4.9988784008639254e-06,
+      "loss": 0.6677,
+      "step": 948
+    },
+    {
+      "epoch": 0.3750061743640405,
+      "grad_norm": 0.5052116039346995,
+      "learning_rate": 4.99887370342441e-06,
+      "loss": 0.6693,
+      "step": 949
+    },
+    {
+      "epoch": 0.37540133366263273,
+      "grad_norm": 0.5090885753544528,
+      "learning_rate": 4.998868996170835e-06,
+      "loss": 0.6308,
+      "step": 950
+    },
+    {
+      "epoch": 0.37579649296122497,
+      "grad_norm": 0.6095866408280952,
+      "learning_rate": 4.9988642791032205e-06,
+      "loss": 0.6728,
+      "step": 951
+    },
+    {
+      "epoch": 0.37619165225981727,
+      "grad_norm": 0.5542782397017351,
+      "learning_rate": 4.998859552221584e-06,
+      "loss": 0.6645,
+      "step": 952
+    },
+    {
+      "epoch": 0.3765868115584095,
+      "grad_norm": 0.6102273000867942,
+      "learning_rate": 4.9988548155259446e-06,
+      "loss": 0.6714,
+      "step": 953
+    },
+    {
+      "epoch": 0.37698197085700175,
+      "grad_norm": 0.5788528638282513,
+      "learning_rate": 4.998850069016321e-06,
+      "loss": 0.6676,
+      "step": 954
+    },
+    {
+      "epoch": 0.377377130155594,
+      "grad_norm": 0.5259169323825259,
+      "learning_rate": 4.998845312692732e-06,
+      "loss": 0.6705,
+      "step": 955
+    },
+    {
+      "epoch": 0.3777722894541862,
+      "grad_norm": 0.5396876703479768,
+      "learning_rate": 4.998840546555196e-06,
+      "loss": 0.6619,
+      "step": 956
+    },
+    {
+      "epoch": 0.37816744875277847,
+      "grad_norm": 0.5417300333578844,
+      "learning_rate": 4.9988357706037315e-06,
+      "loss": 0.675,
+      "step": 957
+    },
+    {
+      "epoch": 0.3785626080513707,
+      "grad_norm": 0.5212701698383503,
+      "learning_rate": 4.998830984838358e-06,
+      "loss": 0.6446,
+      "step": 958
+    },
+    {
+      "epoch": 0.37895776734996295,
+      "grad_norm": 0.5585604547421766,
+      "learning_rate": 4.9988261892590925e-06,
+      "loss": 0.6649,
+      "step": 959
+    },
+    {
+      "epoch": 0.3793529266485552,
+      "grad_norm": 0.5356740144146942,
+      "learning_rate": 4.998821383865956e-06,
+      "loss": 0.6612,
+      "step": 960
+    },
+    {
+      "epoch": 0.37974808594714743,
+      "grad_norm": 0.5389707866617842,
+      "learning_rate": 4.9988165686589665e-06,
+      "loss": 0.6612,
+      "step": 961
+    },
+    {
+      "epoch": 0.38014324524573967,
+      "grad_norm": 0.5470563785584632,
+      "learning_rate": 4.998811743638142e-06,
+      "loss": 0.6502,
+      "step": 962
+    },
+    {
+      "epoch": 0.3805384045443319,
+      "grad_norm": 0.5497560302772653,
+      "learning_rate": 4.998806908803504e-06,
+      "loss": 0.6656,
+      "step": 963
+    },
+    {
+      "epoch": 0.3809335638429242,
+      "grad_norm": 0.5504911983499313,
+      "learning_rate": 4.998802064155068e-06,
+      "loss": 0.6502,
+      "step": 964
+    },
+    {
+      "epoch": 0.38132872314151645,
+      "grad_norm": 0.5526354169398469,
+      "learning_rate": 4.998797209692856e-06,
+      "loss": 0.6807,
+      "step": 965
+    },
+    {
+      "epoch": 0.3817238824401087,
+      "grad_norm": 0.5608969389308869,
+      "learning_rate": 4.998792345416886e-06,
+      "loss": 0.6562,
+      "step": 966
+    },
+    {
+      "epoch": 0.3821190417387009,
+      "grad_norm": 0.5361562466884632,
+      "learning_rate": 4.998787471327177e-06,
+      "loss": 0.6781,
+      "step": 967
+    },
+    {
+      "epoch": 0.38251420103729317,
+      "grad_norm": 0.6165770215935255,
+      "learning_rate": 4.998782587423747e-06,
+      "loss": 0.668,
+      "step": 968
+    },
+    {
+      "epoch": 0.3829093603358854,
+      "grad_norm": 0.6202442629402494,
+      "learning_rate": 4.9987776937066175e-06,
+      "loss": 0.6746,
+      "step": 969
+    },
+    {
+      "epoch": 0.38330451963447765,
+      "grad_norm": 0.5192514425229897,
+      "learning_rate": 4.998772790175806e-06,
+      "loss": 0.6578,
+      "step": 970
+    },
+    {
+      "epoch": 0.3836996789330699,
+      "grad_norm": 0.621919978812942,
+      "learning_rate": 4.998767876831333e-06,
+      "loss": 0.6718,
+      "step": 971
+    },
+    {
+      "epoch": 0.38409483823166213,
+      "grad_norm": 0.5374983489076336,
+      "learning_rate": 4.998762953673216e-06,
+      "loss": 0.6818,
+      "step": 972
+    },
+    {
+      "epoch": 0.38448999753025437,
+      "grad_norm": 0.5336474258984356,
+      "learning_rate": 4.998758020701476e-06,
+      "loss": 0.6751,
+      "step": 973
+    },
+    {
+      "epoch": 0.3848851568288466,
+      "grad_norm": 0.5621510282866113,
+      "learning_rate": 4.998753077916132e-06,
+      "loss": 0.6844,
+      "step": 974
+    },
+    {
+      "epoch": 0.38528031612743885,
+      "grad_norm": 0.5441555376095901,
+      "learning_rate": 4.998748125317203e-06,
+      "loss": 0.6607,
+      "step": 975
+    },
+    {
+      "epoch": 0.38567547542603114,
+      "grad_norm": 0.5419600341782435,
+      "learning_rate": 4.998743162904709e-06,
+      "loss": 0.6722,
+      "step": 976
+    },
+    {
+      "epoch": 0.3860706347246234,
+      "grad_norm": 0.5739218284794436,
+      "learning_rate": 4.998738190678669e-06,
+      "loss": 0.6466,
+      "step": 977
+    },
+    {
+      "epoch": 0.3864657940232156,
+      "grad_norm": 0.5369050064214791,
+      "learning_rate": 4.998733208639103e-06,
+      "loss": 0.6476,
+      "step": 978
+    },
+    {
+      "epoch": 0.38686095332180787,
+      "grad_norm": 0.5377019972492644,
+      "learning_rate": 4.99872821678603e-06,
+      "loss": 0.6768,
+      "step": 979
+    },
+    {
+      "epoch": 0.3872561126204001,
+      "grad_norm": 0.6081827614560933,
+      "learning_rate": 4.99872321511947e-06,
+      "loss": 0.6694,
+      "step": 980
+    },
+    {
+      "epoch": 0.38765127191899235,
+      "grad_norm": 0.5640672539104963,
+      "learning_rate": 4.998718203639442e-06,
+      "loss": 0.6632,
+      "step": 981
+    },
+    {
+      "epoch": 0.3880464312175846,
+      "grad_norm": 0.5004089536671517,
+      "learning_rate": 4.998713182345967e-06,
+      "loss": 0.6379,
+      "step": 982
+    },
+    {
+      "epoch": 0.3884415905161768,
+      "grad_norm": 0.5836270463868607,
+      "learning_rate": 4.998708151239063e-06,
+      "loss": 0.658,
+      "step": 983
+    },
+    {
+      "epoch": 0.38883674981476907,
+      "grad_norm": 0.5251488931681385,
+      "learning_rate": 4.998703110318751e-06,
+      "loss": 0.696,
+      "step": 984
+    },
+    {
+      "epoch": 0.3892319091133613,
+      "grad_norm": 0.5242683309546348,
+      "learning_rate": 4.998698059585051e-06,
+      "loss": 0.6634,
+      "step": 985
+    },
+    {
+      "epoch": 0.38962706841195355,
+      "grad_norm": 0.5380044203359831,
+      "learning_rate": 4.998692999037982e-06,
+      "loss": 0.639,
+      "step": 986
+    },
+    {
+      "epoch": 0.3900222277105458,
+      "grad_norm": 0.5462572647473274,
+      "learning_rate": 4.998687928677563e-06,
+      "loss": 0.6645,
+      "step": 987
+    },
+    {
+      "epoch": 0.3904173870091381,
+      "grad_norm": 0.5464260530979489,
+      "learning_rate": 4.998682848503817e-06,
+      "loss": 0.6602,
+      "step": 988
+    },
+    {
+      "epoch": 0.3908125463077303,
+      "grad_norm": 0.535112702170672,
+      "learning_rate": 4.998677758516761e-06,
+      "loss": 0.6743,
+      "step": 989
+    },
+    {
+      "epoch": 0.39120770560632256,
+      "grad_norm": 0.5295319813464742,
+      "learning_rate": 4.998672658716416e-06,
+      "loss": 0.6655,
+      "step": 990
+    },
+    {
+      "epoch": 0.3916028649049148,
+      "grad_norm": 0.541811550848016,
+      "learning_rate": 4.998667549102803e-06,
+      "loss": 0.6758,
+      "step": 991
+    },
+    {
+      "epoch": 0.39199802420350705,
+      "grad_norm": 0.5451452325701546,
+      "learning_rate": 4.99866242967594e-06,
+      "loss": 0.6884,
+      "step": 992
+    },
+    {
+      "epoch": 0.3923931835020993,
+      "grad_norm": 0.5341028641904682,
+      "learning_rate": 4.998657300435849e-06,
+      "loss": 0.6585,
+      "step": 993
+    },
+    {
+      "epoch": 0.3927883428006915,
+      "grad_norm": 0.5871334739980254,
+      "learning_rate": 4.9986521613825486e-06,
+      "loss": 0.6512,
+      "step": 994
+    },
+    {
+      "epoch": 0.39318350209928377,
+      "grad_norm": 0.5236524401205415,
+      "learning_rate": 4.9986470125160605e-06,
+      "loss": 0.668,
+      "step": 995
+    },
+    {
+      "epoch": 0.393578661397876,
+      "grad_norm": 0.5213997123202212,
+      "learning_rate": 4.998641853836404e-06,
+      "loss": 0.6478,
+      "step": 996
+    },
+    {
+      "epoch": 0.39397382069646825,
+      "grad_norm": 0.5407197664544282,
+      "learning_rate": 4.9986366853436e-06,
+      "loss": 0.6936,
+      "step": 997
+    },
+    {
+      "epoch": 0.3943689799950605,
+      "grad_norm": 0.5280672011633306,
+      "learning_rate": 4.9986315070376675e-06,
+      "loss": 0.6498,
+      "step": 998
+    },
+    {
+      "epoch": 0.3947641392936527,
+      "grad_norm": 0.5204342792742052,
+      "learning_rate": 4.998626318918628e-06,
+      "loss": 0.6775,
+      "step": 999
+    },
+    {
+      "epoch": 0.395159298592245,
+      "grad_norm": 0.5029429193286309,
+      "learning_rate": 4.998621120986502e-06,
+      "loss": 0.6573,
+      "step": 1000
+    },
+    {
+      "epoch": 0.39555445789083726,
+      "grad_norm": 0.5297718137035743,
+      "learning_rate": 4.998615913241309e-06,
+      "loss": 0.6645,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3959496171894295,
+      "grad_norm": 0.5082079772490311,
+      "learning_rate": 4.9986106956830705e-06,
+      "loss": 0.6609,
+      "step": 1002
+    },
+    {
+      "epoch": 0.39634477648802174,
+      "grad_norm": 0.5349674831031332,
+      "learning_rate": 4.998605468311805e-06,
+      "loss": 0.6563,
+      "step": 1003
+    },
+    {
+      "epoch": 0.396739935786614,
+      "grad_norm": 0.5733947545706467,
+      "learning_rate": 4.9986002311275365e-06,
+      "loss": 0.6363,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3971350950852062,
+      "grad_norm": 0.5182307185611646,
+      "learning_rate": 4.9985949841302825e-06,
+      "loss": 0.6516,
+      "step": 1005
+    },
+    {
+      "epoch": 0.39753025438379846,
+      "grad_norm": 0.5062700575538346,
+      "learning_rate": 4.9985897273200645e-06,
+      "loss": 0.6529,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3979254136823907,
+      "grad_norm": 0.5384297745518919,
+      "learning_rate": 4.998584460696904e-06,
+      "loss": 0.6737,
+      "step": 1007
+    },
+    {
+      "epoch": 0.39832057298098295,
+      "grad_norm": 0.5816595934297174,
+      "learning_rate": 4.99857918426082e-06,
+      "loss": 0.6739,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3987157322795752,
+      "grad_norm": 0.5179842596066364,
+      "learning_rate": 4.998573898011835e-06,
+      "loss": 0.6551,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3991108915781674,
+      "grad_norm": 0.5782261510920916,
+      "learning_rate": 4.998568601949968e-06,
+      "loss": 0.6653,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39950605087675967,
+      "grad_norm": 0.5070377777043106,
+      "learning_rate": 4.998563296075241e-06,
+      "loss": 0.6637,
+      "step": 1011
+    },
+    {
+      "epoch": 0.39990121017535196,
+      "grad_norm": 0.5348184367082748,
+      "learning_rate": 4.998557980387675e-06,
+      "loss": 0.6573,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4002963694739442,
+      "grad_norm": 0.5533134020078849,
+      "learning_rate": 4.99855265488729e-06,
+      "loss": 0.6513,
+      "step": 1013
+    },
+    {
+      "epoch": 0.40069152877253644,
+      "grad_norm": 0.5214559414090275,
+      "learning_rate": 4.998547319574108e-06,
+      "loss": 0.6607,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4010866880711287,
+      "grad_norm": 0.5454794781961575,
+      "learning_rate": 4.998541974448149e-06,
+      "loss": 0.657,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4014818473697209,
+      "grad_norm": 0.5323970053630921,
+      "learning_rate": 4.998536619509434e-06,
+      "loss": 0.6727,
+      "step": 1016
+    },
+    {
+      "epoch": 0.40187700666831316,
+      "grad_norm": 0.5677904629043107,
+      "learning_rate": 4.998531254757984e-06,
+      "loss": 0.6439,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4022721659669054,
+      "grad_norm": 0.5203197455098383,
+      "learning_rate": 4.998525880193822e-06,
+      "loss": 0.6694,
+      "step": 1018
+    },
+    {
+      "epoch": 0.40266732526549764,
+      "grad_norm": 0.5251699628155582,
+      "learning_rate": 4.998520495816967e-06,
+      "loss": 0.6546,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4030624845640899,
+      "grad_norm": 0.5539885023636764,
+      "learning_rate": 4.99851510162744e-06,
+      "loss": 0.6802,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4034576438626821,
+      "grad_norm": 0.5328047319862138,
+      "learning_rate": 4.998509697625262e-06,
+      "loss": 0.65,
+      "step": 1021
+    },
+    {
+      "epoch": 0.40385280316127437,
+      "grad_norm": 0.5505034070242261,
+      "learning_rate": 4.998504283810457e-06,
+      "loss": 0.6507,
+      "step": 1022
+    },
+    {
+      "epoch": 0.40424796245986666,
+      "grad_norm": 0.571221870438257,
+      "learning_rate": 4.998498860183043e-06,
+      "loss": 0.6449,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4046431217584589,
+      "grad_norm": 0.534352298409385,
+      "learning_rate": 4.998493426743044e-06,
+      "loss": 0.6806,
+      "step": 1024
+    },
+    {
+      "epoch": 0.40503828105705114,
+      "grad_norm": 0.5487946183182154,
+      "learning_rate": 4.9984879834904785e-06,
+      "loss": 0.6471,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4054334403556434,
+      "grad_norm": 0.5244300620347428,
+      "learning_rate": 4.99848253042537e-06,
+      "loss": 0.6555,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4058285996542356,
+      "grad_norm": 0.5408629309328332,
+      "learning_rate": 4.99847706754774e-06,
+      "loss": 0.6642,
+      "step": 1027
+    },
+    {
+      "epoch": 0.40622375895282786,
+      "grad_norm": 0.5333823797340622,
+      "learning_rate": 4.998471594857608e-06,
+      "loss": 0.6543,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4066189182514201,
+      "grad_norm": 0.512514585835694,
+      "learning_rate": 4.998466112354998e-06,
+      "loss": 0.663,
+      "step": 1029
+    },
+    {
+      "epoch": 0.40701407755001234,
+      "grad_norm": 0.5066646847695025,
+      "learning_rate": 4.998460620039929e-06,
+      "loss": 0.6483,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4074092368486046,
+      "grad_norm": 0.5408686381957716,
+      "learning_rate": 4.998455117912425e-06,
+      "loss": 0.6556,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4078043961471968,
+      "grad_norm": 0.5118425141052046,
+      "learning_rate": 4.998449605972505e-06,
+      "loss": 0.6727,
+      "step": 1032
+    },
+    {
+      "epoch": 0.40819955544578906,
+      "grad_norm": 0.5596018207482388,
+      "learning_rate": 4.9984440842201935e-06,
+      "loss": 0.662,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4085947147443813,
+      "grad_norm": 0.5081268236908408,
+      "learning_rate": 4.99843855265551e-06,
+      "loss": 0.6472,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4089898740429736,
+      "grad_norm": 0.5208429105965913,
+      "learning_rate": 4.998433011278477e-06,
+      "loss": 0.6542,
+      "step": 1035
+    },
+    {
+      "epoch": 0.40938503334156584,
+      "grad_norm": 0.5476716325972726,
+      "learning_rate": 4.998427460089117e-06,
+      "loss": 0.6636,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4097801926401581,
+      "grad_norm": 0.5581628717427036,
+      "learning_rate": 4.9984218990874504e-06,
+      "loss": 0.6688,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4101753519387503,
+      "grad_norm": 0.5313422288525836,
+      "learning_rate": 4.9984163282734995e-06,
+      "loss": 0.6614,
+      "step": 1038
+    },
+    {
+      "epoch": 0.41057051123734256,
+      "grad_norm": 0.5152033372832628,
+      "learning_rate": 4.998410747647287e-06,
+      "loss": 0.6793,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4109656705359348,
+      "grad_norm": 0.5352102398766864,
+      "learning_rate": 4.998405157208833e-06,
+      "loss": 0.6502,
+      "step": 1040
+    },
+    {
+      "epoch": 0.41136082983452704,
+      "grad_norm": 0.6112119061779702,
+      "learning_rate": 4.998399556958162e-06,
+      "loss": 0.6765,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4117559891331193,
+      "grad_norm": 0.5012000160841078,
+      "learning_rate": 4.9983939468952945e-06,
+      "loss": 0.6602,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4121511484317115,
+      "grad_norm": 0.5623209902661767,
+      "learning_rate": 4.9983883270202525e-06,
+      "loss": 0.6458,
+      "step": 1043
+    },
+    {
+      "epoch": 0.41254630773030376,
+      "grad_norm": 0.52453782255717,
+      "learning_rate": 4.998382697333058e-06,
+      "loss": 0.6542,
+      "step": 1044
+    },
+    {
+      "epoch": 0.412941467028896,
+      "grad_norm": 0.5822299190642851,
+      "learning_rate": 4.998377057833733e-06,
+      "loss": 0.679,
+      "step": 1045
+    },
+    {
+      "epoch": 0.41333662632748824,
+      "grad_norm": 0.5421354227434607,
+      "learning_rate": 4.998371408522302e-06,
+      "loss": 0.6717,
+      "step": 1046
+    },
+    {
+      "epoch": 0.41373178562608054,
+      "grad_norm": 0.5286757169293261,
+      "learning_rate": 4.998365749398783e-06,
+      "loss": 0.6711,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4141269449246728,
+      "grad_norm": 0.5177607288874907,
+      "learning_rate": 4.9983600804632e-06,
+      "loss": 0.6686,
+      "step": 1048
+    },
+    {
+      "epoch": 0.414522104223265,
+      "grad_norm": 0.5543980251697767,
+      "learning_rate": 4.998354401715577e-06,
+      "loss": 0.6663,
+      "step": 1049
+    },
+    {
+      "epoch": 0.41491726352185726,
+      "grad_norm": 0.5459993812213396,
+      "learning_rate": 4.9983487131559354e-06,
+      "loss": 0.6407,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4153124228204495,
+      "grad_norm": 0.5359069513585107,
+      "learning_rate": 4.998343014784296e-06,
+      "loss": 0.6575,
+      "step": 1051
+    },
+    {
+      "epoch": 0.41570758211904174,
+      "grad_norm": 0.5300909673600657,
+      "learning_rate": 4.998337306600683e-06,
+      "loss": 0.6667,
+      "step": 1052
+    },
+    {
+      "epoch": 0.416102741417634,
+      "grad_norm": 0.5392983958526001,
+      "learning_rate": 4.9983315886051185e-06,
+      "loss": 0.6721,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4164979007162262,
+      "grad_norm": 0.5070297189799001,
+      "learning_rate": 4.998325860797624e-06,
+      "loss": 0.6561,
+      "step": 1054
+    },
+    {
+      "epoch": 0.41689306001481846,
+      "grad_norm": 0.5224307231811263,
+      "learning_rate": 4.998320123178223e-06,
+      "loss": 0.6618,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4172882193134107,
+      "grad_norm": 0.604920198130287,
+      "learning_rate": 4.998314375746937e-06,
+      "loss": 0.6532,
+      "step": 1056
+    },
+    {
+      "epoch": 0.41768337861200294,
+      "grad_norm": 0.5168038782011498,
+      "learning_rate": 4.9983086185037896e-06,
+      "loss": 0.6651,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4180785379105952,
+      "grad_norm": 0.5297744100346249,
+      "learning_rate": 4.998302851448803e-06,
+      "loss": 0.6608,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4184736972091875,
+      "grad_norm": 0.5574740674977203,
+      "learning_rate": 4.998297074581999e-06,
+      "loss": 0.6881,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4188688565077797,
+      "grad_norm": 0.521702687451242,
+      "learning_rate": 4.9982912879034025e-06,
+      "loss": 0.6839,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41926401580637196,
+      "grad_norm": 0.5188090920413589,
+      "learning_rate": 4.9982854914130345e-06,
+      "loss": 0.6602,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4196591751049642,
+      "grad_norm": 0.5270636307682602,
+      "learning_rate": 4.998279685110917e-06,
+      "loss": 0.6606,
+      "step": 1062
+    },
+    {
+      "epoch": 0.42005433440355644,
+      "grad_norm": 0.5366763635966227,
+      "learning_rate": 4.998273868997075e-06,
+      "loss": 0.6704,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4204494937021487,
+      "grad_norm": 0.5031111535411266,
+      "learning_rate": 4.9982680430715305e-06,
+      "loss": 0.6747,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4208446530007409,
+      "grad_norm": 0.5151445274947559,
+      "learning_rate": 4.998262207334306e-06,
+      "loss": 0.6645,
+      "step": 1065
+    },
+    {
+      "epoch": 0.42123981229933316,
+      "grad_norm": 0.5210208888874347,
+      "learning_rate": 4.998256361785424e-06,
+      "loss": 0.65,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4216349715979254,
+      "grad_norm": 0.5169090685583989,
+      "learning_rate": 4.998250506424908e-06,
+      "loss": 0.6517,
+      "step": 1067
+    },
+    {
+      "epoch": 0.42203013089651764,
+      "grad_norm": 0.5413286895905355,
+      "learning_rate": 4.998244641252781e-06,
+      "loss": 0.6795,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4224252901951099,
+      "grad_norm": 0.6433702861587849,
+      "learning_rate": 4.998238766269067e-06,
+      "loss": 0.6951,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4228204494937021,
+      "grad_norm": 0.5023044231984531,
+      "learning_rate": 4.998232881473787e-06,
+      "loss": 0.6694,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4232156087922944,
+      "grad_norm": 0.5994864572308269,
+      "learning_rate": 4.998226986866966e-06,
+      "loss": 0.6532,
+      "step": 1071
+    },
+    {
+      "epoch": 0.42361076809088666,
+      "grad_norm": 0.5126827044965871,
+      "learning_rate": 4.998221082448627e-06,
+      "loss": 0.6496,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4240059273894789,
+      "grad_norm": 0.6035162881851404,
+      "learning_rate": 4.998215168218791e-06,
+      "loss": 0.6633,
+      "step": 1073
+    },
+    {
+      "epoch": 0.42440108668807114,
+      "grad_norm": 0.5296590133461134,
+      "learning_rate": 4.998209244177484e-06,
+      "loss": 0.6691,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4247962459866634,
+      "grad_norm": 0.5435187950208628,
+      "learning_rate": 4.998203310324727e-06,
+      "loss": 0.6636,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4251914052852556,
+      "grad_norm": 0.5931608594717878,
+      "learning_rate": 4.998197366660546e-06,
+      "loss": 0.6772,
+      "step": 1076
+    },
+    {
+      "epoch": 0.42558656458384786,
+      "grad_norm": 0.524768630466991,
+      "learning_rate": 4.9981914131849614e-06,
+      "loss": 0.657,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4259817238824401,
+      "grad_norm": 0.5518814321211608,
+      "learning_rate": 4.998185449897999e-06,
+      "loss": 0.6483,
+      "step": 1078
+    },
+    {
+      "epoch": 0.42637688318103234,
+      "grad_norm": 0.536584834748668,
+      "learning_rate": 4.998179476799679e-06,
+      "loss": 0.6678,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4267720424796246,
+      "grad_norm": 0.5621190263657815,
+      "learning_rate": 4.998173493890029e-06,
+      "loss": 0.6416,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4271672017782168,
+      "grad_norm": 0.5755170038945299,
+      "learning_rate": 4.99816750116907e-06,
+      "loss": 0.6638,
+      "step": 1081
+    },
+    {
+      "epoch": 0.42756236107680906,
+      "grad_norm": 0.6441072902907758,
+      "learning_rate": 4.998161498636826e-06,
+      "loss": 0.6661,
+      "step": 1082
+    },
+    {
+      "epoch": 0.42795752037540136,
+      "grad_norm": 0.565628041779372,
+      "learning_rate": 4.998155486293321e-06,
+      "loss": 0.6417,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4283526796739936,
+      "grad_norm": 1.8488838880474752,
+      "learning_rate": 4.9981494641385775e-06,
+      "loss": 0.6564,
+      "step": 1084
+    },
+    {
+      "epoch": 0.42874783897258584,
+      "grad_norm": 0.5160251022562632,
+      "learning_rate": 4.99814343217262e-06,
+      "loss": 0.6748,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4291429982711781,
+      "grad_norm": 0.5216784182870835,
+      "learning_rate": 4.998137390395472e-06,
+      "loss": 0.6496,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4295381575697703,
+      "grad_norm": 0.569161680789975,
+      "learning_rate": 4.998131338807158e-06,
+      "loss": 0.6629,
+      "step": 1087
+    },
+    {
+      "epoch": 0.42993331686836256,
+      "grad_norm": 0.5479677358649104,
+      "learning_rate": 4.9981252774077e-06,
+      "loss": 0.658,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4303284761669548,
+      "grad_norm": 0.5267588064158593,
+      "learning_rate": 4.998119206197124e-06,
+      "loss": 0.646,
+      "step": 1089
+    },
+    {
+      "epoch": 0.43072363546554704,
+      "grad_norm": 0.6215218453962831,
+      "learning_rate": 4.9981131251754516e-06,
+      "loss": 0.6473,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4311187947641393,
+      "grad_norm": 0.5829418685005663,
+      "learning_rate": 4.998107034342708e-06,
+      "loss": 0.6489,
+      "step": 1091
+    },
+    {
+      "epoch": 0.4315139540627315,
+      "grad_norm": 0.5445716063606678,
+      "learning_rate": 4.998100933698917e-06,
+      "loss": 0.6563,
+      "step": 1092
+    },
+    {
+      "epoch": 0.43190911336132376,
+      "grad_norm": 0.5717920439154024,
+      "learning_rate": 4.998094823244103e-06,
+      "loss": 0.6455,
+      "step": 1093
+    },
+    {
+      "epoch": 0.43230427265991606,
+      "grad_norm": 0.5194203985812613,
+      "learning_rate": 4.9980887029782895e-06,
+      "loss": 0.6441,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4326994319585083,
+      "grad_norm": 0.5380215301586405,
+      "learning_rate": 4.9980825729015e-06,
+      "loss": 0.6592,
+      "step": 1095
+    },
+    {
+      "epoch": 0.43309459125710054,
+      "grad_norm": 0.5656627049153228,
+      "learning_rate": 4.998076433013758e-06,
+      "loss": 0.6538,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4334897505556928,
+      "grad_norm": 0.5266370469773144,
+      "learning_rate": 4.998070283315091e-06,
+      "loss": 0.6631,
+      "step": 1097
+    },
+    {
+      "epoch": 0.433884909854285,
+      "grad_norm": 0.6484381198976229,
+      "learning_rate": 4.998064123805519e-06,
+      "loss": 0.6539,
+      "step": 1098
+    },
+    {
+      "epoch": 0.43428006915287726,
+      "grad_norm": 0.5474765769152128,
+      "learning_rate": 4.99805795448507e-06,
+      "loss": 0.6589,
+      "step": 1099
+    },
+    {
+      "epoch": 0.4346752284514695,
+      "grad_norm": 0.532857291701254,
+      "learning_rate": 4.998051775353764e-06,
+      "loss": 0.6592,
+      "step": 1100
+    },
+    {
+      "epoch": 0.43507038775006174,
+      "grad_norm": 0.58383601467703,
+      "learning_rate": 4.998045586411629e-06,
+      "loss": 0.6857,
+      "step": 1101
+    },
+    {
+      "epoch": 0.435465547048654,
+      "grad_norm": 0.5473181451591314,
+      "learning_rate": 4.998039387658686e-06,
+      "loss": 0.6637,
+      "step": 1102
+    },
+    {
+      "epoch": 0.4358607063472462,
+      "grad_norm": 0.5385435795744777,
+      "learning_rate": 4.998033179094963e-06,
+      "loss": 0.6259,
+      "step": 1103
+    },
+    {
+      "epoch": 0.43625586564583846,
+      "grad_norm": 0.5531552648985082,
+      "learning_rate": 4.998026960720483e-06,
+      "loss": 0.6586,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4366510249444307,
+      "grad_norm": 0.528398105412856,
+      "learning_rate": 4.998020732535268e-06,
+      "loss": 0.6423,
+      "step": 1105
+    },
+    {
+      "epoch": 0.437046184243023,
+      "grad_norm": 0.5211440156999154,
+      "learning_rate": 4.998014494539345e-06,
+      "loss": 0.6316,
+      "step": 1106
+    },
+    {
+      "epoch": 0.43744134354161524,
+      "grad_norm": 0.547888131353762,
+      "learning_rate": 4.998008246732739e-06,
+      "loss": 0.6342,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4378365028402075,
+      "grad_norm": 0.5557989460207667,
+      "learning_rate": 4.998001989115473e-06,
+      "loss": 0.6574,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4382316621387997,
+      "grad_norm": 0.5500706961475933,
+      "learning_rate": 4.997995721687572e-06,
+      "loss": 0.6471,
+      "step": 1109
+    },
+    {
+      "epoch": 0.43862682143739196,
+      "grad_norm": 0.5504155370955836,
+      "learning_rate": 4.997989444449061e-06,
+      "loss": 0.6576,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4390219807359842,
+      "grad_norm": 0.5547627447818888,
+      "learning_rate": 4.997983157399963e-06,
+      "loss": 0.6764,
+      "step": 1111
+    },
+    {
+      "epoch": 0.43941714003457644,
+      "grad_norm": 0.5804481448072446,
+      "learning_rate": 4.997976860540305e-06,
+      "loss": 0.6682,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4398122993331687,
+      "grad_norm": 0.5397528058437261,
+      "learning_rate": 4.997970553870111e-06,
+      "loss": 0.6572,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4402074586317609,
+      "grad_norm": 0.6196448584341162,
+      "learning_rate": 4.997964237389405e-06,
+      "loss": 0.658,
+      "step": 1114
+    },
+    {
+      "epoch": 0.44060261793035316,
+      "grad_norm": 0.6066819778157864,
+      "learning_rate": 4.997957911098212e-06,
+      "loss": 0.6799,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4409977772289454,
+      "grad_norm": 0.5478230864583113,
+      "learning_rate": 4.997951574996558e-06,
+      "loss": 0.6656,
+      "step": 1116
+    },
+    {
+      "epoch": 0.44139293652753764,
+      "grad_norm": 0.6388534140119478,
+      "learning_rate": 4.997945229084467e-06,
+      "loss": 0.6733,
+      "step": 1117
+    },
+    {
+      "epoch": 0.44178809582612993,
+      "grad_norm": 0.5602979323078272,
+      "learning_rate": 4.997938873361964e-06,
+      "loss": 0.6499,
+      "step": 1118
+    },
+    {
+      "epoch": 0.4421832551247222,
+      "grad_norm": 0.6110344898706578,
+      "learning_rate": 4.997932507829073e-06,
+      "loss": 0.6486,
+      "step": 1119
+    },
+    {
+      "epoch": 0.4425784144233144,
+      "grad_norm": 0.5439975635095607,
+      "learning_rate": 4.997926132485821e-06,
+      "loss": 0.6615,
+      "step": 1120
+    },
+    {
+      "epoch": 0.44297357372190665,
+      "grad_norm": 0.5670773229357888,
+      "learning_rate": 4.9979197473322315e-06,
+      "loss": 0.6644,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4433687330204989,
+      "grad_norm": 0.6135668184552048,
+      "learning_rate": 4.99791335236833e-06,
+      "loss": 0.6586,
+      "step": 1122
+    },
+    {
+      "epoch": 0.44376389231909114,
+      "grad_norm": 0.6934295929969644,
+      "learning_rate": 4.997906947594142e-06,
+      "loss": 0.6446,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4441590516176834,
+      "grad_norm": 0.5521875793962009,
+      "learning_rate": 4.997900533009692e-06,
+      "loss": 0.6845,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4445542109162756,
+      "grad_norm": 0.5802844349362782,
+      "learning_rate": 4.9978941086150055e-06,
+      "loss": 0.6313,
+      "step": 1125
+    },
+    {
+      "epoch": 0.44494937021486786,
+      "grad_norm": 0.5443010267154513,
+      "learning_rate": 4.997887674410108e-06,
+      "loss": 0.6602,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4453445295134601,
+      "grad_norm": 0.5391556371847563,
+      "learning_rate": 4.997881230395024e-06,
+      "loss": 0.6605,
+      "step": 1127
+    },
+    {
+      "epoch": 0.44573968881205234,
+      "grad_norm": 0.6625220158972982,
+      "learning_rate": 4.99787477656978e-06,
+      "loss": 0.6627,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4461348481106446,
+      "grad_norm": 0.5147363265798429,
+      "learning_rate": 4.9978683129344e-06,
+      "loss": 0.6382,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4465300074092369,
+      "grad_norm": 0.5482662281477236,
+      "learning_rate": 4.99786183948891e-06,
+      "loss": 0.6558,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4469251667078291,
+      "grad_norm": 0.5604183997920332,
+      "learning_rate": 4.997855356233337e-06,
+      "loss": 0.6517,
+      "step": 1131
+    },
+    {
+      "epoch": 0.44732032600642135,
+      "grad_norm": 0.5353748475279002,
+      "learning_rate": 4.997848863167703e-06,
+      "loss": 0.6341,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4477154853050136,
+      "grad_norm": 0.5488398069149468,
+      "learning_rate": 4.997842360292036e-06,
+      "loss": 0.6676,
+      "step": 1133
+    },
+    {
+      "epoch": 0.44811064460360583,
+      "grad_norm": 0.5663070948772903,
+      "learning_rate": 4.997835847606361e-06,
+      "loss": 0.64,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4485058039021981,
+      "grad_norm": 0.5403481201660154,
+      "learning_rate": 4.997829325110705e-06,
+      "loss": 0.6599,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4489009632007903,
+      "grad_norm": 0.5222149042010814,
+      "learning_rate": 4.997822792805091e-06,
+      "loss": 0.6262,
+      "step": 1136
+    },
+    {
+      "epoch": 0.44929612249938256,
+      "grad_norm": 0.6013913868061602,
+      "learning_rate": 4.997816250689545e-06,
+      "loss": 0.6661,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4496912817979748,
+      "grad_norm": 0.6787491243452577,
+      "learning_rate": 4.997809698764094e-06,
+      "loss": 0.653,
+      "step": 1138
+    },
+    {
+      "epoch": 0.45008644109656704,
+      "grad_norm": 0.6230119957241225,
+      "learning_rate": 4.997803137028764e-06,
+      "loss": 0.6444,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4504816003951593,
+      "grad_norm": 0.5712264595241187,
+      "learning_rate": 4.9977965654835795e-06,
+      "loss": 0.6115,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4508767596937515,
+      "grad_norm": 0.5074125803443317,
+      "learning_rate": 4.997789984128567e-06,
+      "loss": 0.6439,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4512719189923438,
+      "grad_norm": 0.5525141812309801,
+      "learning_rate": 4.997783392963752e-06,
+      "loss": 0.6223,
+      "step": 1142
+    },
+    {
+      "epoch": 0.45166707829093605,
+      "grad_norm": 0.5882221184228417,
+      "learning_rate": 4.997776791989161e-06,
+      "loss": 0.6506,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4520622375895283,
+      "grad_norm": 0.5063672919950463,
+      "learning_rate": 4.9977701812048185e-06,
+      "loss": 0.6345,
+      "step": 1144
+    },
+    {
+      "epoch": 0.45245739688812053,
+      "grad_norm": 0.5980686687629019,
+      "learning_rate": 4.997763560610752e-06,
+      "loss": 0.6694,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4528525561867128,
+      "grad_norm": 0.5737769969500374,
+      "learning_rate": 4.997756930206987e-06,
+      "loss": 0.6576,
+      "step": 1146
+    },
+    {
+      "epoch": 0.453247715485305,
+      "grad_norm": 0.5254065661728536,
+      "learning_rate": 4.99775028999355e-06,
+      "loss": 0.6406,
+      "step": 1147
+    },
+    {
+      "epoch": 0.45364287478389725,
+      "grad_norm": 0.5556443229887936,
+      "learning_rate": 4.997743639970466e-06,
+      "loss": 0.6353,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4540380340824895,
+      "grad_norm": 0.6029582429283775,
+      "learning_rate": 4.997736980137762e-06,
+      "loss": 0.6525,
+      "step": 1149
+    },
+    {
+      "epoch": 0.45443319338108173,
+      "grad_norm": 0.5498673723447964,
+      "learning_rate": 4.997730310495464e-06,
+      "loss": 0.6499,
+      "step": 1150
+    },
+    {
+      "epoch": 0.454828352679674,
+      "grad_norm": 0.5478846740663137,
+      "learning_rate": 4.997723631043597e-06,
+      "loss": 0.6598,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4552235119782662,
+      "grad_norm": 0.5257136286019838,
+      "learning_rate": 4.997716941782189e-06,
+      "loss": 0.642,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4556186712768585,
+      "grad_norm": 0.5545888280208816,
+      "learning_rate": 4.997710242711266e-06,
+      "loss": 0.6351,
+      "step": 1153
+    },
+    {
+      "epoch": 0.45601383057545075,
+      "grad_norm": 0.537711633021132,
+      "learning_rate": 4.997703533830853e-06,
+      "loss": 0.6504,
+      "step": 1154
+    },
+    {
+      "epoch": 0.456408989874043,
+      "grad_norm": 0.8090471632091599,
+      "learning_rate": 4.997696815140978e-06,
+      "loss": 0.6535,
+      "step": 1155
+    },
+    {
+      "epoch": 0.45680414917263523,
+      "grad_norm": 0.5180994899415312,
+      "learning_rate": 4.997690086641666e-06,
+      "loss": 0.6526,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4571993084712275,
+      "grad_norm": 0.5504445838265242,
+      "learning_rate": 4.997683348332945e-06,
+      "loss": 0.6491,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4575944677698197,
+      "grad_norm": 0.5189150345527932,
+      "learning_rate": 4.997676600214839e-06,
+      "loss": 0.6358,
+      "step": 1158
+    },
+    {
+      "epoch": 0.45798962706841195,
+      "grad_norm": 0.5120657160011092,
+      "learning_rate": 4.997669842287377e-06,
+      "loss": 0.6447,
+      "step": 1159
+    },
+    {
+      "epoch": 0.4583847863670042,
+      "grad_norm": 0.5051586925759178,
+      "learning_rate": 4.997663074550584e-06,
+      "loss": 0.6385,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45877994566559643,
+      "grad_norm": 0.5168398913045842,
+      "learning_rate": 4.997656297004487e-06,
+      "loss": 0.6743,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4591751049641887,
+      "grad_norm": 0.5043203559885473,
+      "learning_rate": 4.997649509649114e-06,
+      "loss": 0.6449,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4595702642627809,
+      "grad_norm": 0.5459507585407142,
+      "learning_rate": 4.997642712484489e-06,
+      "loss": 0.6476,
+      "step": 1163
+    },
+    {
+      "epoch": 0.45996542356137315,
+      "grad_norm": 0.5245847133766435,
+      "learning_rate": 4.99763590551064e-06,
+      "loss": 0.6596,
+      "step": 1164
+    },
+    {
+      "epoch": 0.46036058285996545,
+      "grad_norm": 0.4986446564658458,
+      "learning_rate": 4.997629088727594e-06,
+      "loss": 0.6525,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4607557421585577,
+      "grad_norm": 0.507187864739981,
+      "learning_rate": 4.997622262135379e-06,
+      "loss": 0.6522,
+      "step": 1166
+    },
+    {
+      "epoch": 0.46115090145714993,
+      "grad_norm": 0.5093539527497493,
+      "learning_rate": 4.997615425734019e-06,
+      "loss": 0.6483,
+      "step": 1167
+    },
+    {
+      "epoch": 0.46154606075574217,
+      "grad_norm": 0.5152474618083938,
+      "learning_rate": 4.997608579523543e-06,
+      "loss": 0.6435,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4619412200543344,
+      "grad_norm": 0.5340544828015156,
+      "learning_rate": 4.997601723503977e-06,
+      "loss": 0.661,
+      "step": 1169
+    },
+    {
+      "epoch": 0.46233637935292665,
+      "grad_norm": 0.5035751988607956,
+      "learning_rate": 4.997594857675347e-06,
+      "loss": 0.6365,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4627315386515189,
+      "grad_norm": 0.4881941716047279,
+      "learning_rate": 4.997587982037682e-06,
+      "loss": 0.6539,
+      "step": 1171
+    },
+    {
+      "epoch": 0.46312669795011113,
+      "grad_norm": 0.5200636790139908,
+      "learning_rate": 4.997581096591007e-06,
+      "loss": 0.66,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4635218572487034,
+      "grad_norm": 0.575490383484796,
+      "learning_rate": 4.9975742013353515e-06,
+      "loss": 0.6833,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4639170165472956,
+      "grad_norm": 0.5118745281561939,
+      "learning_rate": 4.99756729627074e-06,
+      "loss": 0.6538,
+      "step": 1174
+    },
+    {
+      "epoch": 0.46431217584588785,
+      "grad_norm": 0.5207712676596821,
+      "learning_rate": 4.997560381397201e-06,
+      "loss": 0.6575,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4647073351444801,
+      "grad_norm": 0.5196060135878352,
+      "learning_rate": 4.997553456714762e-06,
+      "loss": 0.6606,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4651024944430724,
+      "grad_norm": 0.4993473638807208,
+      "learning_rate": 4.997546522223449e-06,
+      "loss": 0.6473,
+      "step": 1177
+    },
+    {
+      "epoch": 0.46549765374166463,
+      "grad_norm": 0.6632587575022358,
+      "learning_rate": 4.99753957792329e-06,
+      "loss": 0.6479,
+      "step": 1178
+    },
+    {
+      "epoch": 0.46589281304025687,
+      "grad_norm": 0.5118246850008393,
+      "learning_rate": 4.997532623814312e-06,
+      "loss": 0.6508,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4662879723388491,
+      "grad_norm": 0.5216848433144801,
+      "learning_rate": 4.997525659896543e-06,
+      "loss": 0.6507,
+      "step": 1180
+    },
+    {
+      "epoch": 0.46668313163744135,
+      "grad_norm": 0.5116873956882975,
+      "learning_rate": 4.99751868617001e-06,
+      "loss": 0.6321,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4670782909360336,
+      "grad_norm": 0.5465288573676182,
+      "learning_rate": 4.997511702634739e-06,
+      "loss": 0.6723,
+      "step": 1182
+    },
+    {
+      "epoch": 0.46747345023462583,
+      "grad_norm": 0.5229836616121926,
+      "learning_rate": 4.997504709290759e-06,
+      "loss": 0.6572,
+      "step": 1183
+    },
+    {
+      "epoch": 0.46786860953321807,
+      "grad_norm": 0.5055148283070077,
+      "learning_rate": 4.997497706138098e-06,
+      "loss": 0.651,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4682637688318103,
+      "grad_norm": 0.5073671397211292,
+      "learning_rate": 4.997490693176782e-06,
+      "loss": 0.6663,
+      "step": 1185
+    },
+    {
+      "epoch": 0.46865892813040255,
+      "grad_norm": 0.5187411381755869,
+      "learning_rate": 4.997483670406839e-06,
+      "loss": 0.6536,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4690540874289948,
+      "grad_norm": 0.4896382189093635,
+      "learning_rate": 4.9974766378282964e-06,
+      "loss": 0.6376,
+      "step": 1187
+    },
+    {
+      "epoch": 0.46944924672758703,
+      "grad_norm": 0.5403722513951374,
+      "learning_rate": 4.997469595441182e-06,
+      "loss": 0.6384,
+      "step": 1188
+    },
+    {
+      "epoch": 0.46984440602617933,
+      "grad_norm": 0.5283649386398254,
+      "learning_rate": 4.9974625432455245e-06,
+      "loss": 0.6619,
+      "step": 1189
+    },
+    {
+      "epoch": 0.47023956532477157,
+      "grad_norm": 0.5145771307045552,
+      "learning_rate": 4.99745548124135e-06,
+      "loss": 0.6608,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4706347246233638,
+      "grad_norm": 0.5088786133567975,
+      "learning_rate": 4.997448409428687e-06,
+      "loss": 0.652,
+      "step": 1191
+    },
+    {
+      "epoch": 0.47102988392195605,
+      "grad_norm": 0.5298568145169689,
+      "learning_rate": 4.997441327807563e-06,
+      "loss": 0.626,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4714250432205483,
+      "grad_norm": 0.525542401872513,
+      "learning_rate": 4.997434236378006e-06,
+      "loss": 0.6535,
+      "step": 1193
+    },
+    {
+      "epoch": 0.47182020251914053,
+      "grad_norm": 0.5112712732618009,
+      "learning_rate": 4.997427135140045e-06,
+      "loss": 0.6463,
+      "step": 1194
+    },
+    {
+      "epoch": 0.47221536181773277,
+      "grad_norm": 0.5333131112367618,
+      "learning_rate": 4.997420024093705e-06,
+      "loss": 0.6479,
+      "step": 1195
+    },
+    {
+      "epoch": 0.472610521116325,
+      "grad_norm": 0.5314791312504851,
+      "learning_rate": 4.997412903239017e-06,
+      "loss": 0.6412,
+      "step": 1196
+    },
+    {
+      "epoch": 0.47300568041491725,
+      "grad_norm": 0.49879520819742884,
+      "learning_rate": 4.997405772576007e-06,
+      "loss": 0.6429,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4734008397135095,
+      "grad_norm": 0.5301052895426869,
+      "learning_rate": 4.997398632104703e-06,
+      "loss": 0.662,
+      "step": 1198
+    },
+    {
+      "epoch": 0.47379599901210173,
+      "grad_norm": 0.5114526203823233,
+      "learning_rate": 4.997391481825135e-06,
+      "loss": 0.6488,
+      "step": 1199
+    },
+    {
+      "epoch": 0.47419115831069397,
+      "grad_norm": 0.7049934317889363,
+      "learning_rate": 4.9973843217373295e-06,
+      "loss": 0.6367,
+      "step": 1200
+    },
+    {
+      "epoch": 0.47458631760928627,
+      "grad_norm": 0.5070332705862394,
+      "learning_rate": 4.997377151841314e-06,
+      "loss": 0.6449,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4749814769078785,
+      "grad_norm": 0.5056739665660747,
+      "learning_rate": 4.997369972137118e-06,
+      "loss": 0.6098,
+      "step": 1202
+    },
+    {
+      "epoch": 0.47537663620647075,
+      "grad_norm": 0.5196505706221262,
+      "learning_rate": 4.9973627826247695e-06,
+      "loss": 0.6412,
+      "step": 1203
+    },
+    {
+      "epoch": 0.475771795505063,
+      "grad_norm": 0.542890120550589,
+      "learning_rate": 4.997355583304297e-06,
+      "loss": 0.6798,
+      "step": 1204
+    },
+    {
+      "epoch": 0.47616695480365523,
+      "grad_norm": 0.5360546233685582,
+      "learning_rate": 4.997348374175727e-06,
+      "loss": 0.6472,
+      "step": 1205
+    },
+    {
+      "epoch": 0.47656211410224747,
+      "grad_norm": 0.5130309786882297,
+      "learning_rate": 4.997341155239089e-06,
+      "loss": 0.6628,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4769572734008397,
+      "grad_norm": 0.5155911732166051,
+      "learning_rate": 4.997333926494412e-06,
+      "loss": 0.6589,
+      "step": 1207
+    },
+    {
+      "epoch": 0.47735243269943195,
+      "grad_norm": 0.5151851351284802,
+      "learning_rate": 4.997326687941724e-06,
+      "loss": 0.6426,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4777475919980242,
+      "grad_norm": 0.5421565514451022,
+      "learning_rate": 4.997319439581053e-06,
+      "loss": 0.6286,
+      "step": 1209
+    },
+    {
+      "epoch": 0.47814275129661643,
+      "grad_norm": 0.5426134587047164,
+      "learning_rate": 4.997312181412428e-06,
+      "loss": 0.657,
+      "step": 1210
+    },
+    {
+      "epoch": 0.47853791059520867,
+      "grad_norm": 0.5155176144597221,
+      "learning_rate": 4.997304913435876e-06,
+      "loss": 0.6545,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4789330698938009,
+      "grad_norm": 0.510072359951836,
+      "learning_rate": 4.997297635651428e-06,
+      "loss": 0.669,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4793282291923932,
+      "grad_norm": 0.5446367757282072,
+      "learning_rate": 4.997290348059111e-06,
+      "loss": 0.6428,
+      "step": 1213
+    },
+    {
+      "epoch": 0.47972338849098545,
+      "grad_norm": 0.4942516591572794,
+      "learning_rate": 4.997283050658954e-06,
+      "loss": 0.6276,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4801185477895777,
+      "grad_norm": 0.5495732208117189,
+      "learning_rate": 4.997275743450986e-06,
+      "loss": 0.6542,
+      "step": 1215
+    },
+    {
+      "epoch": 0.48051370708816993,
+      "grad_norm": 0.5299388692393306,
+      "learning_rate": 4.997268426435234e-06,
+      "loss": 0.6659,
+      "step": 1216
+    },
+    {
+      "epoch": 0.48090886638676217,
+      "grad_norm": 0.5053176342784106,
+      "learning_rate": 4.99726109961173e-06,
+      "loss": 0.6309,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4813040256853544,
+      "grad_norm": 0.6098039985268289,
+      "learning_rate": 4.997253762980499e-06,
+      "loss": 0.667,
+      "step": 1218
+    },
+    {
+      "epoch": 0.48169918498394665,
+      "grad_norm": 0.5376800287875747,
+      "learning_rate": 4.9972464165415726e-06,
+      "loss": 0.6506,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4820943442825389,
+      "grad_norm": 0.5696775883802918,
+      "learning_rate": 4.997239060294978e-06,
+      "loss": 0.6567,
+      "step": 1220
+    },
+    {
+      "epoch": 0.48248950358113113,
+      "grad_norm": 0.5370925937539444,
+      "learning_rate": 4.997231694240745e-06,
+      "loss": 0.6395,
+      "step": 1221
+    },
+    {
+      "epoch": 0.48288466287972337,
+      "grad_norm": 0.5518524860262873,
+      "learning_rate": 4.997224318378903e-06,
+      "loss": 0.6545,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4832798221783156,
+      "grad_norm": 0.6934350958942014,
+      "learning_rate": 4.99721693270948e-06,
+      "loss": 0.6425,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4836749814769079,
+      "grad_norm": 0.5319811886445309,
+      "learning_rate": 4.997209537232505e-06,
+      "loss": 0.6848,
+      "step": 1224
+    },
+    {
+      "epoch": 0.48407014077550015,
+      "grad_norm": 0.5133175351958729,
+      "learning_rate": 4.9972021319480065e-06,
+      "loss": 0.6422,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4844653000740924,
+      "grad_norm": 0.515963610700411,
+      "learning_rate": 4.997194716856016e-06,
+      "loss": 0.6544,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4848604593726846,
+      "grad_norm": 0.5152566753981194,
+      "learning_rate": 4.99718729195656e-06,
+      "loss": 0.6651,
+      "step": 1227
+    },
+    {
+      "epoch": 0.48525561867127687,
+      "grad_norm": 0.51916191994593,
+      "learning_rate": 4.997179857249669e-06,
+      "loss": 0.6539,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4856507779698691,
+      "grad_norm": 0.550013608585077,
+      "learning_rate": 4.9971724127353725e-06,
+      "loss": 0.6719,
+      "step": 1229
+    },
+    {
+      "epoch": 0.48604593726846135,
+      "grad_norm": 0.5152452965696306,
+      "learning_rate": 4.997164958413698e-06,
+      "loss": 0.6558,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4864410965670536,
+      "grad_norm": 0.5149055111216204,
+      "learning_rate": 4.997157494284677e-06,
+      "loss": 0.6652,
+      "step": 1231
+    },
+    {
+      "epoch": 0.48683625586564583,
+      "grad_norm": 0.5332858943526043,
+      "learning_rate": 4.997150020348337e-06,
+      "loss": 0.6394,
+      "step": 1232
+    },
+    {
+      "epoch": 0.48723141516423807,
+      "grad_norm": 0.56714602215336,
+      "learning_rate": 4.997142536604708e-06,
+      "loss": 0.6531,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4876265744628303,
+      "grad_norm": 0.5545157033195995,
+      "learning_rate": 4.99713504305382e-06,
+      "loss": 0.6412,
+      "step": 1234
+    },
+    {
+      "epoch": 0.48802173376142255,
+      "grad_norm": 0.5587339548936653,
+      "learning_rate": 4.997127539695701e-06,
+      "loss": 0.6749,
+      "step": 1235
+    },
+    {
+      "epoch": 0.48841689306001485,
+      "grad_norm": 0.5115432181953685,
+      "learning_rate": 4.997120026530382e-06,
+      "loss": 0.6409,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4888120523586071,
+      "grad_norm": 0.5464551003108359,
+      "learning_rate": 4.997112503557892e-06,
+      "loss": 0.6289,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4892072116571993,
+      "grad_norm": 0.5486875395190994,
+      "learning_rate": 4.99710497077826e-06,
+      "loss": 0.6764,
+      "step": 1238
+    },
+    {
+      "epoch": 0.48960237095579157,
+      "grad_norm": 0.5032189382025016,
+      "learning_rate": 4.997097428191516e-06,
+      "loss": 0.658,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4899975302543838,
+      "grad_norm": 0.6362327602346085,
+      "learning_rate": 4.99708987579769e-06,
+      "loss": 0.6591,
+      "step": 1240
+    },
+    {
+      "epoch": 0.49039268955297605,
+      "grad_norm": 0.5560505192991964,
+      "learning_rate": 4.9970823135968115e-06,
+      "loss": 0.6808,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4907878488515683,
+      "grad_norm": 0.5180049644630822,
+      "learning_rate": 4.997074741588909e-06,
+      "loss": 0.6491,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4911830081501605,
+      "grad_norm": 0.5090961137400876,
+      "learning_rate": 4.997067159774014e-06,
+      "loss": 0.6519,
+      "step": 1243
+    },
+    {
+      "epoch": 0.49157816744875277,
+      "grad_norm": 0.5358018751041574,
+      "learning_rate": 4.997059568152155e-06,
+      "loss": 0.6689,
+      "step": 1244
+    },
+    {
+      "epoch": 0.491973326747345,
+      "grad_norm": 0.5417084482265679,
+      "learning_rate": 4.997051966723363e-06,
+      "loss": 0.6412,
+      "step": 1245
+    },
+    {
+      "epoch": 0.49236848604593725,
+      "grad_norm": 0.524183120106596,
+      "learning_rate": 4.997044355487667e-06,
+      "loss": 0.6574,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4927636453445295,
+      "grad_norm": 0.5104588248452702,
+      "learning_rate": 4.9970367344450966e-06,
+      "loss": 0.6336,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4931588046431218,
+      "grad_norm": 0.5390658040510667,
+      "learning_rate": 4.997029103595682e-06,
+      "loss": 0.6575,
+      "step": 1248
+    },
+    {
+      "epoch": 0.493553963941714,
+      "grad_norm": 0.5085283720065422,
+      "learning_rate": 4.997021462939454e-06,
+      "loss": 0.644,
+      "step": 1249
+    },
+    {
+      "epoch": 0.49394912324030626,
+      "grad_norm": 0.5039511304830326,
+      "learning_rate": 4.997013812476442e-06,
+      "loss": 0.6587,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4943442825388985,
+      "grad_norm": 0.530739785584054,
+      "learning_rate": 4.997006152206675e-06,
+      "loss": 0.6646,
+      "step": 1251
+    },
+    {
+      "epoch": 0.49473944183749075,
+      "grad_norm": 0.5419758659886152,
+      "learning_rate": 4.9969984821301855e-06,
+      "loss": 0.6505,
+      "step": 1252
+    },
+    {
+      "epoch": 0.495134601136083,
+      "grad_norm": 0.49355655865907666,
+      "learning_rate": 4.996990802247002e-06,
+      "loss": 0.6272,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4955297604346752,
+      "grad_norm": 0.5263427398496581,
+      "learning_rate": 4.996983112557154e-06,
+      "loss": 0.6216,
+      "step": 1254
+    },
+    {
+      "epoch": 0.49592491973326747,
+      "grad_norm": 0.5361314380353814,
+      "learning_rate": 4.996975413060673e-06,
+      "loss": 0.6352,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4963200790318597,
+      "grad_norm": 0.5697308230357576,
+      "learning_rate": 4.996967703757589e-06,
+      "loss": 0.6623,
+      "step": 1256
+    },
+    {
+      "epoch": 0.49671523833045195,
+      "grad_norm": 0.5281054703403194,
+      "learning_rate": 4.996959984647931e-06,
+      "loss": 0.6538,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4971103976290442,
+      "grad_norm": 0.5251657598014784,
+      "learning_rate": 4.996952255731732e-06,
+      "loss": 0.6471,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4975055569276364,
+      "grad_norm": 0.5279909788842021,
+      "learning_rate": 4.99694451700902e-06,
+      "loss": 0.6367,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4979007162262287,
+      "grad_norm": 0.5181820037607305,
+      "learning_rate": 4.996936768479826e-06,
+      "loss": 0.661,
+      "step": 1260
+    },
+    {
+      "epoch": 0.49829587552482096,
+      "grad_norm": 0.526657102607614,
+      "learning_rate": 4.9969290101441815e-06,
+      "loss": 0.6481,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4986910348234132,
+      "grad_norm": 0.5547575204328595,
+      "learning_rate": 4.996921242002115e-06,
+      "loss": 0.6392,
+      "step": 1262
+    },
+    {
+      "epoch": 0.49908619412200544,
+      "grad_norm": 0.5292550322998774,
+      "learning_rate": 4.996913464053659e-06,
+      "loss": 0.6462,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4994813534205977,
+      "grad_norm": 0.5051213982225501,
+      "learning_rate": 4.996905676298843e-06,
+      "loss": 0.6243,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4998765127191899,
+      "grad_norm": 0.5404121143898882,
+      "learning_rate": 4.996897878737697e-06,
+      "loss": 0.6508,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5002716720177822,
+      "grad_norm": 0.5853244704579353,
+      "learning_rate": 4.996890071370253e-06,
+      "loss": 0.6692,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5006668313163745,
+      "grad_norm": 0.49722504428159636,
+      "learning_rate": 4.99688225419654e-06,
+      "loss": 0.6241,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5010619906149667,
+      "grad_norm": 0.5391014672758346,
+      "learning_rate": 4.996874427216591e-06,
+      "loss": 0.6517,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5014571499135589,
+      "grad_norm": 0.5284804206913505,
+      "learning_rate": 4.996866590430435e-06,
+      "loss": 0.6511,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5018523092121512,
+      "grad_norm": 0.5270981555373312,
+      "learning_rate": 4.996858743838103e-06,
+      "loss": 0.6426,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5022474685107434,
+      "grad_norm": 0.5246251269006239,
+      "learning_rate": 4.996850887439626e-06,
+      "loss": 0.6558,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5026426278093357,
+      "grad_norm": 0.5432048241827167,
+      "learning_rate": 4.996843021235035e-06,
+      "loss": 0.6398,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5030377871079279,
+      "grad_norm": 0.5293220925783506,
+      "learning_rate": 4.9968351452243605e-06,
+      "loss": 0.6422,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5034329464065201,
+      "grad_norm": 0.5133980649059295,
+      "learning_rate": 4.996827259407634e-06,
+      "loss": 0.6459,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5038281057051124,
+      "grad_norm": 0.5382972682689461,
+      "learning_rate": 4.996819363784886e-06,
+      "loss": 0.6625,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5042232650037046,
+      "grad_norm": 0.9692398606608912,
+      "learning_rate": 4.996811458356148e-06,
+      "loss": 0.6399,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5046184243022969,
+      "grad_norm": 0.5168785866731923,
+      "learning_rate": 4.99680354312145e-06,
+      "loss": 0.6608,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5050135836008891,
+      "grad_norm": 0.5770550825296323,
+      "learning_rate": 4.996795618080824e-06,
+      "loss": 0.626,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5054087428994813,
+      "grad_norm": 0.5677109698383079,
+      "learning_rate": 4.996787683234302e-06,
+      "loss": 0.6436,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5058039021980736,
+      "grad_norm": 0.5046529589482918,
+      "learning_rate": 4.9967797385819135e-06,
+      "loss": 0.632,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5061990614966658,
+      "grad_norm": 0.5511822639718909,
+      "learning_rate": 4.99677178412369e-06,
+      "loss": 0.6504,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5065942207952581,
+      "grad_norm": 0.5162230611570326,
+      "learning_rate": 4.996763819859663e-06,
+      "loss": 0.6545,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5069893800938503,
+      "grad_norm": 0.5229128926155301,
+      "learning_rate": 4.996755845789865e-06,
+      "loss": 0.6812,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5073845393924425,
+      "grad_norm": 0.5205567049268548,
+      "learning_rate": 4.9967478619143244e-06,
+      "loss": 0.6311,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5077796986910348,
+      "grad_norm": 0.5104439487974587,
+      "learning_rate": 4.996739868233076e-06,
+      "loss": 0.6355,
+      "step": 1285
+    },
+    {
+      "epoch": 0.508174857989627,
+      "grad_norm": 0.5216314503599059,
+      "learning_rate": 4.996731864746148e-06,
+      "loss": 0.6466,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5085700172882193,
+      "grad_norm": 0.6055026988137685,
+      "learning_rate": 4.9967238514535745e-06,
+      "loss": 0.6442,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5089651765868115,
+      "grad_norm": 0.5106442297323203,
+      "learning_rate": 4.9967158283553856e-06,
+      "loss": 0.6419,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5093603358854039,
+      "grad_norm": 0.5231952656265989,
+      "learning_rate": 4.996707795451612e-06,
+      "loss": 0.6562,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5097554951839961,
+      "grad_norm": 0.5138942047359026,
+      "learning_rate": 4.996699752742287e-06,
+      "loss": 0.6576,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5101506544825883,
+      "grad_norm": 0.5320229519854904,
+      "learning_rate": 4.996691700227441e-06,
+      "loss": 0.6552,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5105458137811806,
+      "grad_norm": 0.523798873093951,
+      "learning_rate": 4.996683637907107e-06,
+      "loss": 0.657,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5109409730797728,
+      "grad_norm": 0.4798254266039342,
+      "learning_rate": 4.996675565781315e-06,
+      "loss": 0.6393,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5113361323783651,
+      "grad_norm": 0.493096648178815,
+      "learning_rate": 4.996667483850097e-06,
+      "loss": 0.6446,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5117312916769573,
+      "grad_norm": 0.49595320560319645,
+      "learning_rate": 4.996659392113486e-06,
+      "loss": 0.6405,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5121264509755495,
+      "grad_norm": 0.5364956991750699,
+      "learning_rate": 4.9966512905715135e-06,
+      "loss": 0.6527,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5125216102741418,
+      "grad_norm": 0.5294308228967807,
+      "learning_rate": 4.99664317922421e-06,
+      "loss": 0.6481,
+      "step": 1297
+    },
+    {
+      "epoch": 0.512916769572734,
+      "grad_norm": 0.5414401474478188,
+      "learning_rate": 4.996635058071609e-06,
+      "loss": 0.6669,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5133119288713263,
+      "grad_norm": 0.5036105503103926,
+      "learning_rate": 4.99662692711374e-06,
+      "loss": 0.6214,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5137070881699185,
+      "grad_norm": 0.4980125974467828,
+      "learning_rate": 4.996618786350637e-06,
+      "loss": 0.6462,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5141022474685107,
+      "grad_norm": 0.5214248630142702,
+      "learning_rate": 4.996610635782332e-06,
+      "loss": 0.6398,
+      "step": 1301
+    },
+    {
+      "epoch": 0.514497406767103,
+      "grad_norm": 0.500280326327944,
+      "learning_rate": 4.996602475408856e-06,
+      "loss": 0.6231,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5148925660656952,
+      "grad_norm": 0.5289763112497039,
+      "learning_rate": 4.996594305230241e-06,
+      "loss": 0.6456,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5152877253642875,
+      "grad_norm": 0.4963706186676594,
+      "learning_rate": 4.996586125246521e-06,
+      "loss": 0.6507,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5156828846628797,
+      "grad_norm": 0.5107270935037411,
+      "learning_rate": 4.9965779354577254e-06,
+      "loss": 0.6505,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5160780439614719,
+      "grad_norm": 0.5263648487154836,
+      "learning_rate": 4.996569735863888e-06,
+      "loss": 0.6441,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5164732032600642,
+      "grad_norm": 0.5235885042012381,
+      "learning_rate": 4.9965615264650416e-06,
+      "loss": 0.6108,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5168683625586564,
+      "grad_norm": 0.5141835232080435,
+      "learning_rate": 4.996553307261216e-06,
+      "loss": 0.6605,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5172635218572487,
+      "grad_norm": 0.5388043396561775,
+      "learning_rate": 4.996545078252446e-06,
+      "loss": 0.6464,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5176586811558409,
+      "grad_norm": 0.5082281116024677,
+      "learning_rate": 4.996536839438763e-06,
+      "loss": 0.6323,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5180538404544331,
+      "grad_norm": 0.5635524885120492,
+      "learning_rate": 4.996528590820199e-06,
+      "loss": 0.6691,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5184489997530254,
+      "grad_norm": 0.5229240130623062,
+      "learning_rate": 4.996520332396786e-06,
+      "loss": 0.6452,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5188441590516177,
+      "grad_norm": 0.5812217140701571,
+      "learning_rate": 4.996512064168558e-06,
+      "loss": 0.6566,
+      "step": 1313
+    },
+    {
+      "epoch": 0.51923931835021,
+      "grad_norm": 0.5219118851284731,
+      "learning_rate": 4.996503786135546e-06,
+      "loss": 0.6663,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5196344776488022,
+      "grad_norm": 0.507821654017062,
+      "learning_rate": 4.996495498297783e-06,
+      "loss": 0.6512,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5200296369473945,
+      "grad_norm": 0.5307845396993616,
+      "learning_rate": 4.9964872006553025e-06,
+      "loss": 0.6269,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5204247962459867,
+      "grad_norm": 0.5240077658361868,
+      "learning_rate": 4.996478893208135e-06,
+      "loss": 0.6816,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5208199555445789,
+      "grad_norm": 0.5167433367959745,
+      "learning_rate": 4.996470575956316e-06,
+      "loss": 0.6555,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5212151148431712,
+      "grad_norm": 0.5185457506960603,
+      "learning_rate": 4.996462248899876e-06,
+      "loss": 0.6492,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5216102741417634,
+      "grad_norm": 0.5259749401931311,
+      "learning_rate": 4.9964539120388475e-06,
+      "loss": 0.6385,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5220054334403557,
+      "grad_norm": 0.5363760379563364,
+      "learning_rate": 4.996445565373264e-06,
+      "loss": 0.6563,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5224005927389479,
+      "grad_norm": 0.5118945344997703,
+      "learning_rate": 4.996437208903159e-06,
+      "loss": 0.6388,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5227957520375401,
+      "grad_norm": 0.5105505561913978,
+      "learning_rate": 4.996428842628563e-06,
+      "loss": 0.6374,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5231909113361324,
+      "grad_norm": 0.545025639173082,
+      "learning_rate": 4.996420466549512e-06,
+      "loss": 0.6304,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5235860706347246,
+      "grad_norm": 0.5181486179482868,
+      "learning_rate": 4.996412080666036e-06,
+      "loss": 0.6607,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5239812299333169,
+      "grad_norm": 0.5918376928214296,
+      "learning_rate": 4.99640368497817e-06,
+      "loss": 0.65,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5243763892319091,
+      "grad_norm": 0.6440341218450398,
+      "learning_rate": 4.9963952794859475e-06,
+      "loss": 0.6314,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5247715485305013,
+      "grad_norm": 0.506859989727924,
+      "learning_rate": 4.996386864189399e-06,
+      "loss": 0.6405,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5251667078290936,
+      "grad_norm": 0.604693581937193,
+      "learning_rate": 4.9963784390885585e-06,
+      "loss": 0.6301,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5255618671276858,
+      "grad_norm": 0.5388769958954528,
+      "learning_rate": 4.99637000418346e-06,
+      "loss": 0.6563,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5259570264262781,
+      "grad_norm": 0.5525987392165671,
+      "learning_rate": 4.996361559474135e-06,
+      "loss": 0.6408,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5263521857248703,
+      "grad_norm": 0.5438403662996905,
+      "learning_rate": 4.996353104960619e-06,
+      "loss": 0.6463,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5267473450234625,
+      "grad_norm": 0.5205416563435727,
+      "learning_rate": 4.996344640642943e-06,
+      "loss": 0.6281,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5271425043220548,
+      "grad_norm": 0.49717641720078637,
+      "learning_rate": 4.9963361665211404e-06,
+      "loss": 0.6438,
+      "step": 1334
+    },
+    {
+      "epoch": 0.527537663620647,
+      "grad_norm": 0.5720531214872092,
+      "learning_rate": 4.996327682595247e-06,
+      "loss": 0.6549,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5279328229192393,
+      "grad_norm": 0.5083062127313356,
+      "learning_rate": 4.996319188865293e-06,
+      "loss": 0.6343,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5283279822178316,
+      "grad_norm": 0.5326331393036251,
+      "learning_rate": 4.996310685331314e-06,
+      "loss": 0.6654,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5287231415164239,
+      "grad_norm": 0.5263704388039211,
+      "learning_rate": 4.996302171993341e-06,
+      "loss": 0.6319,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5291183008150161,
+      "grad_norm": 0.5004961835420784,
+      "learning_rate": 4.99629364885141e-06,
+      "loss": 0.6383,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5295134601136083,
+      "grad_norm": 0.5474008702250441,
+      "learning_rate": 4.996285115905554e-06,
+      "loss": 0.644,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5299086194122006,
+      "grad_norm": 0.5214317652571568,
+      "learning_rate": 4.996276573155805e-06,
+      "loss": 0.6278,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5303037787107928,
+      "grad_norm": 0.5178793183599104,
+      "learning_rate": 4.996268020602198e-06,
+      "loss": 0.6527,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5306989380093851,
+      "grad_norm": 0.5106695795609324,
+      "learning_rate": 4.9962594582447654e-06,
+      "loss": 0.6194,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5310940973079773,
+      "grad_norm": 0.5553450350073675,
+      "learning_rate": 4.996250886083541e-06,
+      "loss": 0.6651,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5314892566065695,
+      "grad_norm": 0.5020511547000661,
+      "learning_rate": 4.996242304118561e-06,
+      "loss": 0.6579,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5318844159051618,
+      "grad_norm": 0.5056151563934147,
+      "learning_rate": 4.996233712349855e-06,
+      "loss": 0.6522,
+      "step": 1346
+    },
+    {
+      "epoch": 0.532279575203754,
+      "grad_norm": 0.5141576495869976,
+      "learning_rate": 4.996225110777459e-06,
+      "loss": 0.6233,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5326747345023463,
+      "grad_norm": 0.5060024011484795,
+      "learning_rate": 4.996216499401408e-06,
+      "loss": 0.6496,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5330698938009385,
+      "grad_norm": 0.513393743699311,
+      "learning_rate": 4.996207878221732e-06,
+      "loss": 0.6363,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5334650530995307,
+      "grad_norm": 0.516451601423634,
+      "learning_rate": 4.9961992472384695e-06,
+      "loss": 0.6144,
+      "step": 1350
+    },
+    {
+      "epoch": 0.533860212398123,
+      "grad_norm": 0.543940796081213,
+      "learning_rate": 4.99619060645165e-06,
+      "loss": 0.6636,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5342553716967152,
+      "grad_norm": 0.523839395723129,
+      "learning_rate": 4.996181955861311e-06,
+      "loss": 0.6522,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5346505309953075,
+      "grad_norm": 0.5346665570634227,
+      "learning_rate": 4.996173295467485e-06,
+      "loss": 0.6247,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5350456902938997,
+      "grad_norm": 0.5053515545286174,
+      "learning_rate": 4.996164625270206e-06,
+      "loss": 0.628,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5354408495924919,
+      "grad_norm": 0.5385469662689469,
+      "learning_rate": 4.9961559452695075e-06,
+      "loss": 0.6403,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5358360088910842,
+      "grad_norm": 0.530821893645594,
+      "learning_rate": 4.996147255465425e-06,
+      "loss": 0.6493,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5362311681896764,
+      "grad_norm": 0.49520095478056697,
+      "learning_rate": 4.9961385558579915e-06,
+      "loss": 0.6391,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5366263274882687,
+      "grad_norm": 0.5012636197893197,
+      "learning_rate": 4.996129846447241e-06,
+      "loss": 0.6265,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5370214867868609,
+      "grad_norm": 0.5412008937924457,
+      "learning_rate": 4.996121127233209e-06,
+      "loss": 0.6583,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5374166460854533,
+      "grad_norm": 0.51146837063799,
+      "learning_rate": 4.996112398215929e-06,
+      "loss": 0.6552,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5378118053840455,
+      "grad_norm": 0.5164037939572933,
+      "learning_rate": 4.996103659395434e-06,
+      "loss": 0.6441,
+      "step": 1361
+    },
+    {
+      "epoch": 0.5382069646826377,
+      "grad_norm": 0.507221251387215,
+      "learning_rate": 4.99609491077176e-06,
+      "loss": 0.6183,
+      "step": 1362
+    },
+    {
+      "epoch": 0.53860212398123,
+      "grad_norm": 0.5110021327827016,
+      "learning_rate": 4.996086152344942e-06,
+      "loss": 0.6445,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5389972832798222,
+      "grad_norm": 0.4950203261367757,
+      "learning_rate": 4.996077384115012e-06,
+      "loss": 0.6317,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5393924425784145,
+      "grad_norm": 0.513880385000124,
+      "learning_rate": 4.9960686060820065e-06,
+      "loss": 0.6537,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5397876018770067,
+      "grad_norm": 0.5047955828045061,
+      "learning_rate": 4.99605981824596e-06,
+      "loss": 0.6385,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5401827611755989,
+      "grad_norm": 0.5444433922369591,
+      "learning_rate": 4.996051020606904e-06,
+      "loss": 0.6333,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5405779204741912,
+      "grad_norm": 0.5391605933331952,
+      "learning_rate": 4.9960422131648765e-06,
+      "loss": 0.6553,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5409730797727834,
+      "grad_norm": 0.5490109139703936,
+      "learning_rate": 4.996033395919911e-06,
+      "loss": 0.6352,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5413682390713757,
+      "grad_norm": 0.5044155900096984,
+      "learning_rate": 4.996024568872042e-06,
+      "loss": 0.6328,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5417633983699679,
+      "grad_norm": 0.5426896659207923,
+      "learning_rate": 4.9960157320213046e-06,
+      "loss": 0.6419,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5421585576685601,
+      "grad_norm": 0.5475008282374658,
+      "learning_rate": 4.996006885367732e-06,
+      "loss": 0.661,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5425537169671524,
+      "grad_norm": 0.50456290195716,
+      "learning_rate": 4.995998028911361e-06,
+      "loss": 0.6411,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5429488762657446,
+      "grad_norm": 0.517404197086199,
+      "learning_rate": 4.995989162652224e-06,
+      "loss": 0.6244,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5433440355643369,
+      "grad_norm": 0.52119175194665,
+      "learning_rate": 4.995980286590358e-06,
+      "loss": 0.6285,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5437391948629291,
+      "grad_norm": 0.5535606068992925,
+      "learning_rate": 4.995971400725798e-06,
+      "loss": 0.6512,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5441343541615213,
+      "grad_norm": 0.5096581348064602,
+      "learning_rate": 4.995962505058577e-06,
+      "loss": 0.6209,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5445295134601136,
+      "grad_norm": 0.507954709115174,
+      "learning_rate": 4.995953599588731e-06,
+      "loss": 0.6347,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5449246727587058,
+      "grad_norm": 0.5455172501120015,
+      "learning_rate": 4.995944684316295e-06,
+      "loss": 0.6544,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5453198320572981,
+      "grad_norm": 0.5195045840736692,
+      "learning_rate": 4.995935759241304e-06,
+      "loss": 0.6308,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5457149913558903,
+      "grad_norm": 0.5128980270231307,
+      "learning_rate": 4.995926824363793e-06,
+      "loss": 0.6451,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5461101506544825,
+      "grad_norm": 0.5694454666753539,
+      "learning_rate": 4.995917879683796e-06,
+      "loss": 0.6544,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5465053099530748,
+      "grad_norm": 0.5222383396969622,
+      "learning_rate": 4.995908925201351e-06,
+      "loss": 0.6215,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5469004692516671,
+      "grad_norm": 0.5033825276170869,
+      "learning_rate": 4.995899960916489e-06,
+      "loss": 0.6485,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5472956285502594,
+      "grad_norm": 0.5217250381657061,
+      "learning_rate": 4.9958909868292495e-06,
+      "loss": 0.6392,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5476907878488516,
+      "grad_norm": 0.5275739257953245,
+      "learning_rate": 4.995882002939665e-06,
+      "loss": 0.6309,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5480859471474439,
+      "grad_norm": 0.4888893706726107,
+      "learning_rate": 4.995873009247771e-06,
+      "loss": 0.6266,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5484811064460361,
+      "grad_norm": 0.6091283696485257,
+      "learning_rate": 4.995864005753605e-06,
+      "loss": 0.6451,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5488762657446283,
+      "grad_norm": 0.5292867693675212,
+      "learning_rate": 4.9958549924572e-06,
+      "loss": 0.6369,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5492714250432206,
+      "grad_norm": 0.516530358022225,
+      "learning_rate": 4.995845969358591e-06,
+      "loss": 0.6427,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5496665843418128,
+      "grad_norm": 0.5302839204130153,
+      "learning_rate": 4.995836936457816e-06,
+      "loss": 0.6581,
+      "step": 1391
+    },
+    {
+      "epoch": 0.550061743640405,
+      "grad_norm": 0.5392670526192246,
+      "learning_rate": 4.995827893754909e-06,
+      "loss": 0.6354,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5504569029389973,
+      "grad_norm": 0.533934176126299,
+      "learning_rate": 4.995818841249905e-06,
+      "loss": 0.6514,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5508520622375895,
+      "grad_norm": 0.5423235074943602,
+      "learning_rate": 4.99580977894284e-06,
+      "loss": 0.652,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5512472215361818,
+      "grad_norm": 0.5548664924819643,
+      "learning_rate": 4.995800706833751e-06,
+      "loss": 0.6514,
+      "step": 1395
+    },
+    {
+      "epoch": 0.551642380834774,
+      "grad_norm": 0.5230139409187969,
+      "learning_rate": 4.995791624922671e-06,
+      "loss": 0.6449,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5520375401333663,
+      "grad_norm": 0.4882535391793784,
+      "learning_rate": 4.995782533209638e-06,
+      "loss": 0.6171,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5524326994319585,
+      "grad_norm": 0.5602125565323922,
+      "learning_rate": 4.995773431694686e-06,
+      "loss": 0.6535,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5528278587305507,
+      "grad_norm": 0.5396171131081425,
+      "learning_rate": 4.995764320377852e-06,
+      "loss": 0.6597,
+      "step": 1399
+    },
+    {
+      "epoch": 0.553223018029143,
+      "grad_norm": 0.5233921034680511,
+      "learning_rate": 4.99575519925917e-06,
+      "loss": 0.6498,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5536181773277352,
+      "grad_norm": 0.49235649050585056,
+      "learning_rate": 4.995746068338679e-06,
+      "loss": 0.6381,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5540133366263275,
+      "grad_norm": 0.5307247825728258,
+      "learning_rate": 4.995736927616412e-06,
+      "loss": 0.6624,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5544084959249197,
+      "grad_norm": 0.5423744236754792,
+      "learning_rate": 4.995727777092406e-06,
+      "loss": 0.6447,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5548036552235119,
+      "grad_norm": 0.5784643067877503,
+      "learning_rate": 4.995718616766696e-06,
+      "loss": 0.661,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5551988145221042,
+      "grad_norm": 0.5122057994466735,
+      "learning_rate": 4.995709446639319e-06,
+      "loss": 0.656,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5555939738206964,
+      "grad_norm": 0.5174768644606775,
+      "learning_rate": 4.995700266710312e-06,
+      "loss": 0.6455,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5559891331192887,
+      "grad_norm": 0.5496336917317313,
+      "learning_rate": 4.9956910769797085e-06,
+      "loss": 0.6422,
+      "step": 1407
+    },
+    {
+      "epoch": 0.556384292417881,
+      "grad_norm": 0.502618362332198,
+      "learning_rate": 4.9956818774475465e-06,
+      "loss": 0.6368,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5567794517164733,
+      "grad_norm": 0.5691680887325605,
+      "learning_rate": 4.995672668113861e-06,
+      "loss": 0.6551,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5571746110150655,
+      "grad_norm": 0.4970626341833687,
+      "learning_rate": 4.99566344897869e-06,
+      "loss": 0.6235,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5575697703136577,
+      "grad_norm": 0.5189759440475544,
+      "learning_rate": 4.995654220042067e-06,
+      "loss": 0.6591,
+      "step": 1411
+    },
+    {
+      "epoch": 0.55796492961225,
+      "grad_norm": 0.5200276974925322,
+      "learning_rate": 4.99564498130403e-06,
+      "loss": 0.627,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5583600889108422,
+      "grad_norm": 0.4968048227090689,
+      "learning_rate": 4.9956357327646155e-06,
+      "loss": 0.6456,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5587552482094345,
+      "grad_norm": 0.5006447246560516,
+      "learning_rate": 4.995626474423859e-06,
+      "loss": 0.624,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5591504075080267,
+      "grad_norm": 0.5408193413913738,
+      "learning_rate": 4.995617206281797e-06,
+      "loss": 0.6521,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5595455668066189,
+      "grad_norm": 0.5102142633356043,
+      "learning_rate": 4.995607928338466e-06,
+      "loss": 0.6271,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5599407261052112,
+      "grad_norm": 0.7566967338630805,
+      "learning_rate": 4.9955986405939025e-06,
+      "loss": 0.667,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5603358854038034,
+      "grad_norm": 0.5002937738203951,
+      "learning_rate": 4.995589343048144e-06,
+      "loss": 0.5957,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5607310447023957,
+      "grad_norm": 0.5307076787963949,
+      "learning_rate": 4.9955800357012245e-06,
+      "loss": 0.6303,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5611262040009879,
+      "grad_norm": 0.5707224007581209,
+      "learning_rate": 4.995570718553182e-06,
+      "loss": 0.6358,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5615213632995801,
+      "grad_norm": 0.5184302142265548,
+      "learning_rate": 4.995561391604054e-06,
+      "loss": 0.6429,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5619165225981724,
+      "grad_norm": 0.5242472513502078,
+      "learning_rate": 4.995552054853876e-06,
+      "loss": 0.6352,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5623116818967646,
+      "grad_norm": 0.5984667209493643,
+      "learning_rate": 4.995542708302684e-06,
+      "loss": 0.6449,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5627068411953569,
+      "grad_norm": 0.5211234764982716,
+      "learning_rate": 4.995533351950517e-06,
+      "loss": 0.6336,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5631020004939491,
+      "grad_norm": 0.5005552076190557,
+      "learning_rate": 4.9955239857974095e-06,
+      "loss": 0.6315,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5634971597925413,
+      "grad_norm": 0.5156815083287939,
+      "learning_rate": 4.995514609843399e-06,
+      "loss": 0.6128,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5638923190911336,
+      "grad_norm": 0.8025016430790596,
+      "learning_rate": 4.995505224088524e-06,
+      "loss": 0.6324,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5642874783897258,
+      "grad_norm": 0.5143928596776913,
+      "learning_rate": 4.995495828532818e-06,
+      "loss": 0.6298,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5646826376883181,
+      "grad_norm": 0.5420804076927369,
+      "learning_rate": 4.99548642317632e-06,
+      "loss": 0.6263,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5650777969869103,
+      "grad_norm": 0.5455732492128239,
+      "learning_rate": 4.995477008019067e-06,
+      "loss": 0.6599,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5654729562855026,
+      "grad_norm": 0.4976835769324547,
+      "learning_rate": 4.995467583061096e-06,
+      "loss": 0.6287,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5658681155840949,
+      "grad_norm": 0.4891898546027407,
+      "learning_rate": 4.995458148302444e-06,
+      "loss": 0.6389,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5662632748826871,
+      "grad_norm": 0.5330421535897775,
+      "learning_rate": 4.995448703743147e-06,
+      "loss": 0.6551,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5666584341812794,
+      "grad_norm": 0.4928122884897478,
+      "learning_rate": 4.995439249383243e-06,
+      "loss": 0.6453,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5670535934798716,
+      "grad_norm": 0.4982212210514846,
+      "learning_rate": 4.995429785222768e-06,
+      "loss": 0.6364,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5674487527784638,
+      "grad_norm": 0.5090391879426845,
+      "learning_rate": 4.995420311261761e-06,
+      "loss": 0.6527,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5678439120770561,
+      "grad_norm": 0.49179123014556764,
+      "learning_rate": 4.9954108275002586e-06,
+      "loss": 0.6198,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5682390713756483,
+      "grad_norm": 0.49426838823272856,
+      "learning_rate": 4.9954013339382975e-06,
+      "loss": 0.6373,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5686342306742406,
+      "grad_norm": 0.51127919787096,
+      "learning_rate": 4.995391830575915e-06,
+      "loss": 0.6447,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5690293899728328,
+      "grad_norm": 0.5080627757706936,
+      "learning_rate": 4.995382317413149e-06,
+      "loss": 0.6239,
+      "step": 1440
+    },
+    {
+      "epoch": 0.569424549271425,
+      "grad_norm": 0.49636389856600227,
+      "learning_rate": 4.995372794450037e-06,
+      "loss": 0.6433,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5698197085700173,
+      "grad_norm": 0.5337319531991955,
+      "learning_rate": 4.995363261686615e-06,
+      "loss": 0.6335,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5702148678686095,
+      "grad_norm": 0.5283909447365106,
+      "learning_rate": 4.995353719122921e-06,
+      "loss": 0.6246,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5706100271672018,
+      "grad_norm": 0.5050664542458486,
+      "learning_rate": 4.995344166758994e-06,
+      "loss": 0.6547,
+      "step": 1444
+    },
+    {
+      "epoch": 0.571005186465794,
+      "grad_norm": 0.47674622000290146,
+      "learning_rate": 4.99533460459487e-06,
+      "loss": 0.637,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5714003457643863,
+      "grad_norm": 0.522395740013937,
+      "learning_rate": 4.995325032630588e-06,
+      "loss": 0.6294,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5717955050629785,
+      "grad_norm": 0.5076861343641647,
+      "learning_rate": 4.995315450866183e-06,
+      "loss": 0.6478,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5721906643615707,
+      "grad_norm": 0.5123340353219106,
+      "learning_rate": 4.995305859301695e-06,
+      "loss": 0.6324,
+      "step": 1448
+    },
+    {
+      "epoch": 0.572585823660163,
+      "grad_norm": 0.5116830386496904,
+      "learning_rate": 4.9952962579371595e-06,
+      "loss": 0.6541,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5729809829587552,
+      "grad_norm": 0.5029055259684767,
+      "learning_rate": 4.995286646772616e-06,
+      "loss": 0.6411,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5733761422573475,
+      "grad_norm": 0.54340913147798,
+      "learning_rate": 4.995277025808103e-06,
+      "loss": 0.6318,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5737713015559397,
+      "grad_norm": 0.6008588857149761,
+      "learning_rate": 4.995267395043656e-06,
+      "loss": 0.6313,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5741664608545319,
+      "grad_norm": 0.5046550392300445,
+      "learning_rate": 4.995257754479313e-06,
+      "loss": 0.6093,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5745616201531242,
+      "grad_norm": 0.5075601586098186,
+      "learning_rate": 4.995248104115114e-06,
+      "loss": 0.6715,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5749567794517165,
+      "grad_norm": 0.5175079376115269,
+      "learning_rate": 4.995238443951096e-06,
+      "loss": 0.6332,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5753519387503088,
+      "grad_norm": 0.5251599308274112,
+      "learning_rate": 4.995228773987296e-06,
+      "loss": 0.6471,
+      "step": 1456
+    },
+    {
+      "epoch": 0.575747098048901,
+      "grad_norm": 0.5112664273121302,
+      "learning_rate": 4.995219094223753e-06,
+      "loss": 0.6487,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5761422573474932,
+      "grad_norm": 0.5161124213557929,
+      "learning_rate": 4.9952094046605034e-06,
+      "loss": 0.6371,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5765374166460855,
+      "grad_norm": 0.5122993630545928,
+      "learning_rate": 4.995199705297587e-06,
+      "loss": 0.6402,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5769325759446777,
+      "grad_norm": 0.48934561452372555,
+      "learning_rate": 4.995189996135042e-06,
+      "loss": 0.6453,
+      "step": 1460
+    },
+    {
+      "epoch": 0.57732773524327,
+      "grad_norm": 0.5015789106369151,
+      "learning_rate": 4.995180277172905e-06,
+      "loss": 0.6405,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5777228945418622,
+      "grad_norm": 0.4896116736893514,
+      "learning_rate": 4.995170548411215e-06,
+      "loss": 0.604,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5781180538404544,
+      "grad_norm": 0.5056206160242199,
+      "learning_rate": 4.995160809850012e-06,
+      "loss": 0.6501,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5785132131390467,
+      "grad_norm": 0.5271567630707523,
+      "learning_rate": 4.99515106148933e-06,
+      "loss": 0.6637,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5789083724376389,
+      "grad_norm": 0.5161640907927971,
+      "learning_rate": 4.9951413033292115e-06,
+      "loss": 0.6575,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5793035317362312,
+      "grad_norm": 0.5366526893281445,
+      "learning_rate": 4.995131535369693e-06,
+      "loss": 0.6442,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5796986910348234,
+      "grad_norm": 0.478707795838876,
+      "learning_rate": 4.995121757610812e-06,
+      "loss": 0.6264,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5800938503334157,
+      "grad_norm": 0.5109398474629543,
+      "learning_rate": 4.995111970052608e-06,
+      "loss": 0.6432,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5804890096320079,
+      "grad_norm": 0.5029075819574539,
+      "learning_rate": 4.99510217269512e-06,
+      "loss": 0.641,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5808841689306001,
+      "grad_norm": 0.5123704050877957,
+      "learning_rate": 4.995092365538385e-06,
+      "loss": 0.6311,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5812793282291924,
+      "grad_norm": 0.5232609805684696,
+      "learning_rate": 4.995082548582443e-06,
+      "loss": 0.6722,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5816744875277846,
+      "grad_norm": 0.5140040836372506,
+      "learning_rate": 4.995072721827331e-06,
+      "loss": 0.6263,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5820696468263769,
+      "grad_norm": 0.5133765736650968,
+      "learning_rate": 4.995062885273089e-06,
+      "loss": 0.6297,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5824648061249691,
+      "grad_norm": 0.4941943476821983,
+      "learning_rate": 4.995053038919755e-06,
+      "loss": 0.6439,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5828599654235613,
+      "grad_norm": 0.49939101156839866,
+      "learning_rate": 4.9950431827673676e-06,
+      "loss": 0.6221,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5832551247221536,
+      "grad_norm": 0.5284234536221161,
+      "learning_rate": 4.995033316815966e-06,
+      "loss": 0.6466,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5836502840207458,
+      "grad_norm": 0.5232984307977716,
+      "learning_rate": 4.9950234410655886e-06,
+      "loss": 0.6657,
+      "step": 1477
+    },
+    {
+      "epoch": 0.584045443319338,
+      "grad_norm": 0.49111166592428446,
+      "learning_rate": 4.995013555516274e-06,
+      "loss": 0.6372,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5844406026179304,
+      "grad_norm": 0.49817946819904,
+      "learning_rate": 4.99500366016806e-06,
+      "loss": 0.6564,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5848357619165226,
+      "grad_norm": 0.5045229504599369,
+      "learning_rate": 4.994993755020989e-06,
+      "loss": 0.6236,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5852309212151149,
+      "grad_norm": 0.5137046231454804,
+      "learning_rate": 4.994983840075096e-06,
+      "loss": 0.638,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5856260805137071,
+      "grad_norm": 0.5090356242899768,
+      "learning_rate": 4.9949739153304224e-06,
+      "loss": 0.607,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5860212398122994,
+      "grad_norm": 0.5086542750896299,
+      "learning_rate": 4.994963980787005e-06,
+      "loss": 0.6137,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5864163991108916,
+      "grad_norm": 0.5359161859801612,
+      "learning_rate": 4.994954036444886e-06,
+      "loss": 0.6126,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5868115584094838,
+      "grad_norm": 0.507773612302792,
+      "learning_rate": 4.994944082304102e-06,
+      "loss": 0.6161,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5872067177080761,
+      "grad_norm": 0.4931028875320953,
+      "learning_rate": 4.9949341183646914e-06,
+      "loss": 0.6326,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5876018770066683,
+      "grad_norm": 0.5370606115690266,
+      "learning_rate": 4.994924144626695e-06,
+      "loss": 0.6421,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5879970363052606,
+      "grad_norm": 0.5028578672986176,
+      "learning_rate": 4.994914161090152e-06,
+      "loss": 0.6482,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5883921956038528,
+      "grad_norm": 0.49503149805703456,
+      "learning_rate": 4.994904167755102e-06,
+      "loss": 0.6424,
+      "step": 1489
+    },
+    {
+      "epoch": 0.588787354902445,
+      "grad_norm": 0.52378156821829,
+      "learning_rate": 4.994894164621581e-06,
+      "loss": 0.6372,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5891825142010373,
+      "grad_norm": 0.5049018438428512,
+      "learning_rate": 4.994884151689633e-06,
+      "loss": 0.6026,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5895776734996295,
+      "grad_norm": 0.5056297740574257,
+      "learning_rate": 4.994874128959294e-06,
+      "loss": 0.6439,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5899728327982218,
+      "grad_norm": 0.5059061924424931,
+      "learning_rate": 4.994864096430604e-06,
+      "loss": 0.6496,
+      "step": 1493
+    },
+    {
+      "epoch": 0.590367992096814,
+      "grad_norm": 0.5252330272280762,
+      "learning_rate": 4.994854054103604e-06,
+      "loss": 0.6431,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5907631513954062,
+      "grad_norm": 0.5082734525840502,
+      "learning_rate": 4.994844001978331e-06,
+      "loss": 0.665,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5911583106939985,
+      "grad_norm": 0.5783213762110898,
+      "learning_rate": 4.994833940054827e-06,
+      "loss": 0.6641,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5915534699925907,
+      "grad_norm": 0.5579102147905942,
+      "learning_rate": 4.994823868333129e-06,
+      "loss": 0.6442,
+      "step": 1497
+    },
+    {
+      "epoch": 0.591948629291183,
+      "grad_norm": 0.49660206731752937,
+      "learning_rate": 4.9948137868132785e-06,
+      "loss": 0.6256,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5923437885897752,
+      "grad_norm": 0.5338722501575794,
+      "learning_rate": 4.994803695495315e-06,
+      "loss": 0.6324,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5927389478883675,
+      "grad_norm": 0.5435806297907863,
+      "learning_rate": 4.994793594379275e-06,
+      "loss": 0.642,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5931341071869597,
+      "grad_norm": 0.5244058039956626,
+      "learning_rate": 4.9947834834652035e-06,
+      "loss": 0.6351,
+      "step": 1501
+    },
+    {
+      "epoch": 0.593529266485552,
+      "grad_norm": 0.5397763750210784,
+      "learning_rate": 4.9947733627531365e-06,
+      "loss": 0.5993,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5939244257841443,
+      "grad_norm": 0.5288352373402283,
+      "learning_rate": 4.994763232243114e-06,
+      "loss": 0.6477,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5943195850827365,
+      "grad_norm": 0.5090715743441777,
+      "learning_rate": 4.994753091935177e-06,
+      "loss": 0.645,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5947147443813288,
+      "grad_norm": 0.6719556320052742,
+      "learning_rate": 4.994742941829364e-06,
+      "loss": 0.6573,
+      "step": 1505
+    },
+    {
+      "epoch": 0.595109903679921,
+      "grad_norm": 0.5147973826184885,
+      "learning_rate": 4.994732781925717e-06,
+      "loss": 0.6464,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5955050629785132,
+      "grad_norm": 0.522878872773451,
+      "learning_rate": 4.994722612224274e-06,
+      "loss": 0.6592,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5959002222771055,
+      "grad_norm": 0.5037339588735743,
+      "learning_rate": 4.9947124327250755e-06,
+      "loss": 0.6281,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5962953815756977,
+      "grad_norm": 0.5142576687522799,
+      "learning_rate": 4.99470224342816e-06,
+      "loss": 0.6397,
+      "step": 1509
+    },
+    {
+      "epoch": 0.59669054087429,
+      "grad_norm": 0.5442114812958629,
+      "learning_rate": 4.99469204433357e-06,
+      "loss": 0.6681,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5970857001728822,
+      "grad_norm": 0.49806465072349676,
+      "learning_rate": 4.994681835441345e-06,
+      "loss": 0.6411,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5974808594714744,
+      "grad_norm": 0.5113177774148318,
+      "learning_rate": 4.994671616751524e-06,
+      "loss": 0.6365,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5978760187700667,
+      "grad_norm": 0.5044746925510942,
+      "learning_rate": 4.994661388264148e-06,
+      "loss": 0.6245,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5982711780686589,
+      "grad_norm": 0.5077014321420371,
+      "learning_rate": 4.994651149979257e-06,
+      "loss": 0.6296,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5986663373672512,
+      "grad_norm": 0.5577617502488906,
+      "learning_rate": 4.9946409018968915e-06,
+      "loss": 0.6166,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5990614966658434,
+      "grad_norm": 0.5454684436241605,
+      "learning_rate": 4.99463064401709e-06,
+      "loss": 0.6386,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5994566559644356,
+      "grad_norm": 0.5330025000783452,
+      "learning_rate": 4.994620376339895e-06,
+      "loss": 0.6405,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5998518152630279,
+      "grad_norm": 0.501164395733037,
+      "learning_rate": 4.994610098865346e-06,
+      "loss": 0.6225,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6002469745616201,
+      "grad_norm": 0.49869645020020736,
+      "learning_rate": 4.994599811593484e-06,
+      "loss": 0.6316,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6006421338602124,
+      "grad_norm": 0.5203023850194525,
+      "learning_rate": 4.9945895145243476e-06,
+      "loss": 0.6428,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6010372931588046,
+      "grad_norm": 0.5288551722444358,
+      "learning_rate": 4.994579207657979e-06,
+      "loss": 0.6164,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6014324524573968,
+      "grad_norm": 0.4919113743752538,
+      "learning_rate": 4.9945688909944175e-06,
+      "loss": 0.6213,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6018276117559891,
+      "grad_norm": 0.5595834215690251,
+      "learning_rate": 4.994558564533705e-06,
+      "loss": 0.6436,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6022227710545813,
+      "grad_norm": 0.5247902498260631,
+      "learning_rate": 4.9945482282758806e-06,
+      "loss": 0.638,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6026179303531736,
+      "grad_norm": 0.5291480908300407,
+      "learning_rate": 4.994537882220985e-06,
+      "loss": 0.6253,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6030130896517659,
+      "grad_norm": 0.501147812137641,
+      "learning_rate": 4.994527526369061e-06,
+      "loss": 0.664,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6034082489503582,
+      "grad_norm": 0.505734785729107,
+      "learning_rate": 4.994517160720146e-06,
+      "loss": 0.6385,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6038034082489504,
+      "grad_norm": 0.5275768661720794,
+      "learning_rate": 4.994506785274283e-06,
+      "loss": 0.6387,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6041985675475426,
+      "grad_norm": 0.5056505262627228,
+      "learning_rate": 4.994496400031512e-06,
+      "loss": 0.64,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6045937268461349,
+      "grad_norm": 0.4908463182821926,
+      "learning_rate": 4.9944860049918746e-06,
+      "loss": 0.621,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6049888861447271,
+      "grad_norm": 0.5234617249072642,
+      "learning_rate": 4.99447560015541e-06,
+      "loss": 0.6418,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6053840454433194,
+      "grad_norm": 0.501149633946628,
+      "learning_rate": 4.994465185522161e-06,
+      "loss": 0.6636,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6057792047419116,
+      "grad_norm": 0.4720640412577634,
+      "learning_rate": 4.994454761092166e-06,
+      "loss": 0.6411,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6061743640405038,
+      "grad_norm": 0.524342790903981,
+      "learning_rate": 4.994444326865469e-06,
+      "loss": 0.6452,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6065695233390961,
+      "grad_norm": 0.5013477729936378,
+      "learning_rate": 4.994433882842108e-06,
+      "loss": 0.6304,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6069646826376883,
+      "grad_norm": 0.5240831322483116,
+      "learning_rate": 4.994423429022126e-06,
+      "loss": 0.6339,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6073598419362806,
+      "grad_norm": 0.4969641999990712,
+      "learning_rate": 4.994412965405563e-06,
+      "loss": 0.6443,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6077550012348728,
+      "grad_norm": 0.5272037913210349,
+      "learning_rate": 4.9944024919924615e-06,
+      "loss": 0.6468,
+      "step": 1538
+    },
+    {
+      "epoch": 0.608150160533465,
+      "grad_norm": 0.529008774394857,
+      "learning_rate": 4.9943920087828615e-06,
+      "loss": 0.6362,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6085453198320573,
+      "grad_norm": 0.4915768675876793,
+      "learning_rate": 4.994381515776804e-06,
+      "loss": 0.6435,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6089404791306495,
+      "grad_norm": 0.5822531861044621,
+      "learning_rate": 4.9943710129743304e-06,
+      "loss": 0.6608,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6093356384292418,
+      "grad_norm": 0.5146971943259008,
+      "learning_rate": 4.994360500375482e-06,
+      "loss": 0.6351,
+      "step": 1542
+    },
+    {
+      "epoch": 0.609730797727834,
+      "grad_norm": 0.5047870000075502,
+      "learning_rate": 4.994349977980301e-06,
+      "loss": 0.6387,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6101259570264262,
+      "grad_norm": 0.4924385862974,
+      "learning_rate": 4.994339445788827e-06,
+      "loss": 0.6378,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6105211163250185,
+      "grad_norm": 0.5195506832924177,
+      "learning_rate": 4.9943289038011035e-06,
+      "loss": 0.6422,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6109162756236107,
+      "grad_norm": 0.504416370915793,
+      "learning_rate": 4.99431835201717e-06,
+      "loss": 0.6521,
+      "step": 1546
+    },
+    {
+      "epoch": 0.611311434922203,
+      "grad_norm": 0.5134028045383864,
+      "learning_rate": 4.9943077904370684e-06,
+      "loss": 0.6486,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6117065942207952,
+      "grad_norm": 0.5438989106962773,
+      "learning_rate": 4.994297219060841e-06,
+      "loss": 0.6402,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6121017535193874,
+      "grad_norm": 0.502459375072004,
+      "learning_rate": 4.994286637888528e-06,
+      "loss": 0.6421,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6124969128179798,
+      "grad_norm": 0.5162824244852758,
+      "learning_rate": 4.994276046920172e-06,
+      "loss": 0.6403,
+      "step": 1550
+    },
+    {
+      "epoch": 0.612892072116572,
+      "grad_norm": 0.5178670603935037,
+      "learning_rate": 4.994265446155814e-06,
+      "loss": 0.6309,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6132872314151643,
+      "grad_norm": 0.5117410334804755,
+      "learning_rate": 4.994254835595497e-06,
+      "loss": 0.6681,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6136823907137565,
+      "grad_norm": 0.4965002156326863,
+      "learning_rate": 4.994244215239261e-06,
+      "loss": 0.6358,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6140775500123488,
+      "grad_norm": 0.5151456093040961,
+      "learning_rate": 4.994233585087148e-06,
+      "loss": 0.6345,
+      "step": 1554
+    },
+    {
+      "epoch": 0.614472709310941,
+      "grad_norm": 0.5107017962950764,
+      "learning_rate": 4.9942229451392e-06,
+      "loss": 0.6224,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6148678686095332,
+      "grad_norm": 0.5362359326835328,
+      "learning_rate": 4.99421229539546e-06,
+      "loss": 0.6402,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6152630279081255,
+      "grad_norm": 0.5128666635843286,
+      "learning_rate": 4.994201635855967e-06,
+      "loss": 0.6359,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6156581872067177,
+      "grad_norm": 0.5273421131964295,
+      "learning_rate": 4.994190966520765e-06,
+      "loss": 0.6503,
+      "step": 1558
+    },
+    {
+      "epoch": 0.61605334650531,
+      "grad_norm": 0.49230499474339184,
+      "learning_rate": 4.994180287389896e-06,
+      "loss": 0.6261,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6164485058039022,
+      "grad_norm": 0.5217099067284172,
+      "learning_rate": 4.994169598463401e-06,
+      "loss": 0.6586,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6168436651024944,
+      "grad_norm": 0.5102545311764888,
+      "learning_rate": 4.994158899741323e-06,
+      "loss": 0.6147,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6172388244010867,
+      "grad_norm": 0.538265791682114,
+      "learning_rate": 4.9941481912237024e-06,
+      "loss": 0.6526,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6176339836996789,
+      "grad_norm": 0.5087707653077916,
+      "learning_rate": 4.994137472910583e-06,
+      "loss": 0.6103,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6180291429982712,
+      "grad_norm": 0.4985542462986526,
+      "learning_rate": 4.994126744802006e-06,
+      "loss": 0.6245,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6184243022968634,
+      "grad_norm": 0.48942059314144304,
+      "learning_rate": 4.994116006898013e-06,
+      "loss": 0.6214,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6188194615954556,
+      "grad_norm": 0.5116181294437978,
+      "learning_rate": 4.994105259198649e-06,
+      "loss": 0.622,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6192146208940479,
+      "grad_norm": 0.5138679125630047,
+      "learning_rate": 4.994094501703951e-06,
+      "loss": 0.6434,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6196097801926401,
+      "grad_norm": 0.5680217102402336,
+      "learning_rate": 4.994083734413966e-06,
+      "loss": 0.6421,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6200049394912324,
+      "grad_norm": 0.5120382178875652,
+      "learning_rate": 4.9940729573287346e-06,
+      "loss": 0.6318,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6204000987898246,
+      "grad_norm": 0.5470369708755527,
+      "learning_rate": 4.994062170448298e-06,
+      "loss": 0.6225,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6207952580884168,
+      "grad_norm": 0.5586073977661113,
+      "learning_rate": 4.994051373772701e-06,
+      "loss": 0.6429,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6211904173870091,
+      "grad_norm": 0.5780818912171855,
+      "learning_rate": 4.9940405673019844e-06,
+      "loss": 0.6378,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6215855766856014,
+      "grad_norm": 0.5199314186964572,
+      "learning_rate": 4.99402975103619e-06,
+      "loss": 0.642,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6219807359841937,
+      "grad_norm": 0.5292899229264992,
+      "learning_rate": 4.994018924975362e-06,
+      "loss": 0.638,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6223758952827859,
+      "grad_norm": 0.5190979503571678,
+      "learning_rate": 4.994008089119542e-06,
+      "loss": 0.6242,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6227710545813782,
+      "grad_norm": 0.575963095804118,
+      "learning_rate": 4.993997243468772e-06,
+      "loss": 0.6266,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6231662138799704,
+      "grad_norm": 0.5340579778193899,
+      "learning_rate": 4.993986388023096e-06,
+      "loss": 0.6663,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6235613731785626,
+      "grad_norm": 0.4959581900916024,
+      "learning_rate": 4.993975522782556e-06,
+      "loss": 0.6311,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6239565324771549,
+      "grad_norm": 0.4981648627664517,
+      "learning_rate": 4.993964647747195e-06,
+      "loss": 0.6364,
+      "step": 1579
+    },
+    {
+      "epoch": 0.6243516917757471,
+      "grad_norm": 0.5042116842190024,
+      "learning_rate": 4.993953762917054e-06,
+      "loss": 0.6367,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6247468510743394,
+      "grad_norm": 0.5027206712870845,
+      "learning_rate": 4.993942868292178e-06,
+      "loss": 0.6408,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6251420103729316,
+      "grad_norm": 0.494384710791813,
+      "learning_rate": 4.993931963872608e-06,
+      "loss": 0.6388,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6255371696715238,
+      "grad_norm": 0.48929240453544737,
+      "learning_rate": 4.993921049658389e-06,
+      "loss": 0.6359,
+      "step": 1583
+    },
+    {
+      "epoch": 0.6259323289701161,
+      "grad_norm": 0.5275911914444981,
+      "learning_rate": 4.993910125649561e-06,
+      "loss": 0.6435,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6263274882687083,
+      "grad_norm": 0.5354103744385167,
+      "learning_rate": 4.993899191846169e-06,
+      "loss": 0.6214,
+      "step": 1585
+    },
+    {
+      "epoch": 0.6267226475673006,
+      "grad_norm": 0.5116129949120737,
+      "learning_rate": 4.9938882482482555e-06,
+      "loss": 0.636,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6271178068658928,
+      "grad_norm": 0.5185673630960299,
+      "learning_rate": 4.993877294855863e-06,
+      "loss": 0.6366,
+      "step": 1587
+    },
+    {
+      "epoch": 0.627512966164485,
+      "grad_norm": 0.5025695303228517,
+      "learning_rate": 4.993866331669035e-06,
+      "loss": 0.6266,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6279081254630773,
+      "grad_norm": 0.4939243554026485,
+      "learning_rate": 4.993855358687814e-06,
+      "loss": 0.6247,
+      "step": 1589
+    },
+    {
+      "epoch": 0.6283032847616695,
+      "grad_norm": 0.5013958381045799,
+      "learning_rate": 4.993844375912244e-06,
+      "loss": 0.6374,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6286984440602618,
+      "grad_norm": 0.5165249055138371,
+      "learning_rate": 4.993833383342368e-06,
+      "loss": 0.6276,
+      "step": 1591
+    },
+    {
+      "epoch": 0.629093603358854,
+      "grad_norm": 0.5010885019204411,
+      "learning_rate": 4.993822380978228e-06,
+      "loss": 0.6273,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6294887626574462,
+      "grad_norm": 0.47308571192974885,
+      "learning_rate": 4.993811368819869e-06,
+      "loss": 0.6406,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6298839219560385,
+      "grad_norm": 0.5181491819308544,
+      "learning_rate": 4.993800346867333e-06,
+      "loss": 0.6577,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6302790812546307,
+      "grad_norm": 0.7592473307425937,
+      "learning_rate": 4.993789315120663e-06,
+      "loss": 0.6136,
+      "step": 1595
+    },
+    {
+      "epoch": 0.630674240553223,
+      "grad_norm": 0.48322179308535596,
+      "learning_rate": 4.993778273579903e-06,
+      "loss": 0.6233,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6310693998518153,
+      "grad_norm": 0.4989991809584226,
+      "learning_rate": 4.993767222245096e-06,
+      "loss": 0.615,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6314645591504076,
+      "grad_norm": 0.5102524351860097,
+      "learning_rate": 4.993756161116287e-06,
+      "loss": 0.646,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6318597184489998,
+      "grad_norm": 0.48881189426787136,
+      "learning_rate": 4.9937450901935166e-06,
+      "loss": 0.6309,
+      "step": 1599
+    },
+    {
+      "epoch": 0.632254877747592,
+      "grad_norm": 0.5162222090652335,
+      "learning_rate": 4.993734009476831e-06,
+      "loss": 0.6234,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6326500370461843,
+      "grad_norm": 0.5219467840999582,
+      "learning_rate": 4.99372291896627e-06,
+      "loss": 0.6623,
+      "step": 1601
+    },
+    {
+      "epoch": 0.6330451963447765,
+      "grad_norm": 0.4812462236064405,
+      "learning_rate": 4.993711818661882e-06,
+      "loss": 0.6357,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6334403556433688,
+      "grad_norm": 0.5121012568296751,
+      "learning_rate": 4.993700708563708e-06,
+      "loss": 0.6475,
+      "step": 1603
+    },
+    {
+      "epoch": 0.633835514941961,
+      "grad_norm": 0.5243902599376221,
+      "learning_rate": 4.993689588671792e-06,
+      "loss": 0.6204,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6342306742405532,
+      "grad_norm": 0.5038240130102247,
+      "learning_rate": 4.9936784589861765e-06,
+      "loss": 0.6381,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6346258335391455,
+      "grad_norm": 0.5056284357196895,
+      "learning_rate": 4.993667319506907e-06,
+      "loss": 0.6124,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6350209928377377,
+      "grad_norm": 0.5210968104913557,
+      "learning_rate": 4.993656170234026e-06,
+      "loss": 0.6254,
+      "step": 1607
+    },
+    {
+      "epoch": 0.63541615213633,
+      "grad_norm": 0.5529274377061739,
+      "learning_rate": 4.9936450111675785e-06,
+      "loss": 0.6312,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6358113114349222,
+      "grad_norm": 0.5464579530723207,
+      "learning_rate": 4.993633842307607e-06,
+      "loss": 0.616,
+      "step": 1609
+    },
+    {
+      "epoch": 0.6362064707335144,
+      "grad_norm": 0.5262401377076071,
+      "learning_rate": 4.9936226636541564e-06,
+      "loss": 0.6426,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6366016300321067,
+      "grad_norm": 0.543994117033289,
+      "learning_rate": 4.993611475207269e-06,
+      "loss": 0.6192,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6369967893306989,
+      "grad_norm": 0.5341939948990526,
+      "learning_rate": 4.993600276966992e-06,
+      "loss": 0.6271,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6373919486292912,
+      "grad_norm": 0.5093829321261388,
+      "learning_rate": 4.993589068933366e-06,
+      "loss": 0.6177,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6377871079278834,
+      "grad_norm": 0.6103509660198887,
+      "learning_rate": 4.993577851106437e-06,
+      "loss": 0.6292,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6381822672264756,
+      "grad_norm": 0.5815804502591855,
+      "learning_rate": 4.993566623486247e-06,
+      "loss": 0.6644,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6385774265250679,
+      "grad_norm": 0.5192475918715985,
+      "learning_rate": 4.993555386072843e-06,
+      "loss": 0.6439,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6389725858236601,
+      "grad_norm": 0.5474735931150393,
+      "learning_rate": 4.993544138866266e-06,
+      "loss": 0.6393,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6393677451222524,
+      "grad_norm": 0.5395600119026902,
+      "learning_rate": 4.993532881866564e-06,
+      "loss": 0.6386,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6397629044208446,
+      "grad_norm": 0.5461307453791069,
+      "learning_rate": 4.993521615073777e-06,
+      "loss": 0.6423,
+      "step": 1619
+    },
+    {
+      "epoch": 0.6401580637194368,
+      "grad_norm": 0.5129146758889324,
+      "learning_rate": 4.9935103384879525e-06,
+      "loss": 0.6263,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6405532230180292,
+      "grad_norm": 0.5338093770005661,
+      "learning_rate": 4.9934990521091335e-06,
+      "loss": 0.6464,
+      "step": 1621
+    },
+    {
+      "epoch": 0.6409483823166214,
+      "grad_norm": 0.5019435816787823,
+      "learning_rate": 4.993487755937363e-06,
+      "loss": 0.5966,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6413435416152137,
+      "grad_norm": 0.520040205195016,
+      "learning_rate": 4.993476449972689e-06,
+      "loss": 0.6175,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6417387009138059,
+      "grad_norm": 0.6219021349863824,
+      "learning_rate": 4.993465134215151e-06,
+      "loss": 0.635,
+      "step": 1624
+    },
+    {
+      "epoch": 0.6421338602123982,
+      "grad_norm": 0.4961034586172033,
+      "learning_rate": 4.993453808664798e-06,
+      "loss": 0.6482,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6425290195109904,
+      "grad_norm": 0.5096787165995785,
+      "learning_rate": 4.9934424733216715e-06,
+      "loss": 0.6633,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6429241788095826,
+      "grad_norm": 0.5237229903942505,
+      "learning_rate": 4.993431128185818e-06,
+      "loss": 0.6385,
+      "step": 1627
+    },
+    {
+      "epoch": 0.6433193381081749,
+      "grad_norm": 0.4967553534767619,
+      "learning_rate": 4.9934197732572794e-06,
+      "loss": 0.6326,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6437144974067671,
+      "grad_norm": 0.49402878207474876,
+      "learning_rate": 4.993408408536104e-06,
+      "loss": 0.6437,
+      "step": 1629
+    },
+    {
+      "epoch": 0.6441096567053594,
+      "grad_norm": 0.5955393513897179,
+      "learning_rate": 4.993397034022333e-06,
+      "loss": 0.6238,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6445048160039516,
+      "grad_norm": 0.5038622383402742,
+      "learning_rate": 4.993385649716014e-06,
+      "loss": 0.6286,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6448999753025438,
+      "grad_norm": 0.47945185187430733,
+      "learning_rate": 4.9933742556171895e-06,
+      "loss": 0.6099,
+      "step": 1632
+    },
+    {
+      "epoch": 0.6452951346011361,
+      "grad_norm": 0.489392501913111,
+      "learning_rate": 4.993362851725905e-06,
+      "loss": 0.6319,
+      "step": 1633
+    },
+    {
+      "epoch": 0.6456902938997283,
+      "grad_norm": 0.5010165910463097,
+      "learning_rate": 4.993351438042204e-06,
+      "loss": 0.6394,
+      "step": 1634
+    },
+    {
+      "epoch": 0.6460854531983206,
+      "grad_norm": 0.4934200799303604,
+      "learning_rate": 4.993340014566135e-06,
+      "loss": 0.6357,
+      "step": 1635
+    },
+    {
+      "epoch": 0.6464806124969128,
+      "grad_norm": 0.4912427926805577,
+      "learning_rate": 4.993328581297738e-06,
+      "loss": 0.623,
+      "step": 1636
+    },
+    {
+      "epoch": 0.646875771795505,
+      "grad_norm": 0.4873510707511348,
+      "learning_rate": 4.993317138237062e-06,
+      "loss": 0.6391,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6472709310940973,
+      "grad_norm": 0.4948293050023501,
+      "learning_rate": 4.99330568538415e-06,
+      "loss": 0.6412,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6476660903926895,
+      "grad_norm": 0.4936219175231247,
+      "learning_rate": 4.993294222739047e-06,
+      "loss": 0.6268,
+      "step": 1639
+    },
+    {
+      "epoch": 0.6480612496912818,
+      "grad_norm": 0.49996197505527556,
+      "learning_rate": 4.993282750301799e-06,
+      "loss": 0.6429,
+      "step": 1640
+    },
+    {
+      "epoch": 0.648456408989874,
+      "grad_norm": 0.4941502207114622,
+      "learning_rate": 4.993271268072449e-06,
+      "loss": 0.6079,
+      "step": 1641
+    },
+    {
+      "epoch": 0.6488515682884662,
+      "grad_norm": 0.49029556645390376,
+      "learning_rate": 4.993259776051045e-06,
+      "loss": 0.6307,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6492467275870585,
+      "grad_norm": 0.5033458347363889,
+      "learning_rate": 4.9932482742376295e-06,
+      "loss": 0.6331,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6496418868856508,
+      "grad_norm": 0.4996027449996174,
+      "learning_rate": 4.993236762632248e-06,
+      "loss": 0.6174,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6500370461842431,
+      "grad_norm": 0.4954371092717256,
+      "learning_rate": 4.993225241234949e-06,
+      "loss": 0.6355,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6504322054828353,
+      "grad_norm": 0.525740581307838,
+      "learning_rate": 4.9932137100457735e-06,
+      "loss": 0.6318,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6508273647814276,
+      "grad_norm": 0.5096572884835409,
+      "learning_rate": 4.993202169064769e-06,
+      "loss": 0.6369,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6512225240800198,
+      "grad_norm": 0.8344702188550295,
+      "learning_rate": 4.993190618291979e-06,
+      "loss": 0.6473,
+      "step": 1648
+    },
+    {
+      "epoch": 0.651617683378612,
+      "grad_norm": 0.4943527469749559,
+      "learning_rate": 4.993179057727452e-06,
+      "loss": 0.6305,
+      "step": 1649
+    },
+    {
+      "epoch": 0.6520128426772043,
+      "grad_norm": 0.5123537897586123,
+      "learning_rate": 4.993167487371231e-06,
+      "loss": 0.6363,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6524080019757965,
+      "grad_norm": 0.48337632188932905,
+      "learning_rate": 4.993155907223362e-06,
+      "loss": 0.6462,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6528031612743888,
+      "grad_norm": 0.4941193733493147,
+      "learning_rate": 4.993144317283891e-06,
+      "loss": 0.6395,
+      "step": 1652
+    },
+    {
+      "epoch": 0.653198320572981,
+      "grad_norm": 0.5171725409360232,
+      "learning_rate": 4.993132717552862e-06,
+      "loss": 0.615,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6535934798715732,
+      "grad_norm": 0.5028964520752601,
+      "learning_rate": 4.9931211080303225e-06,
+      "loss": 0.6264,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6539886391701655,
+      "grad_norm": 0.5170261798292486,
+      "learning_rate": 4.9931094887163165e-06,
+      "loss": 0.6232,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6543837984687577,
+      "grad_norm": 0.5393500555133617,
+      "learning_rate": 4.993097859610891e-06,
+      "loss": 0.6295,
+      "step": 1656
+    },
+    {
+      "epoch": 0.65477895776735,
+      "grad_norm": 0.5170632791663832,
+      "learning_rate": 4.99308622071409e-06,
+      "loss": 0.6457,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6551741170659422,
+      "grad_norm": 0.5088037328336552,
+      "learning_rate": 4.993074572025961e-06,
+      "loss": 0.6329,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6555692763645344,
+      "grad_norm": 0.5517685712418717,
+      "learning_rate": 4.993062913546549e-06,
+      "loss": 0.6151,
+      "step": 1659
+    },
+    {
+      "epoch": 0.6559644356631267,
+      "grad_norm": 0.6574193345875109,
+      "learning_rate": 4.9930512452758996e-06,
+      "loss": 0.6541,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6563595949617189,
+      "grad_norm": 0.50641713145241,
+      "learning_rate": 4.993039567214058e-06,
+      "loss": 0.6439,
+      "step": 1661
+    },
+    {
+      "epoch": 0.6567547542603112,
+      "grad_norm": 0.5183226080631096,
+      "learning_rate": 4.993027879361072e-06,
+      "loss": 0.6441,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6571499135589034,
+      "grad_norm": 0.5026167336939322,
+      "learning_rate": 4.993016181716987e-06,
+      "loss": 0.6335,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6575450728574956,
+      "grad_norm": 0.5060260690365562,
+      "learning_rate": 4.993004474281846e-06,
+      "loss": 0.6447,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6579402321560879,
+      "grad_norm": 0.5057941400086423,
+      "learning_rate": 4.992992757055699e-06,
+      "loss": 0.6291,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6583353914546801,
+      "grad_norm": 0.506979357825653,
+      "learning_rate": 4.9929810300385894e-06,
+      "loss": 0.635,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6587305507532724,
+      "grad_norm": 0.4818917872233048,
+      "learning_rate": 4.992969293230565e-06,
+      "loss": 0.6382,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6591257100518647,
+      "grad_norm": 0.4925360002260035,
+      "learning_rate": 4.992957546631671e-06,
+      "loss": 0.6142,
+      "step": 1668
+    },
+    {
+      "epoch": 0.659520869350457,
+      "grad_norm": 0.5046203653064728,
+      "learning_rate": 4.992945790241952e-06,
+      "loss": 0.6304,
+      "step": 1669
+    },
+    {
+      "epoch": 0.6599160286490492,
+      "grad_norm": 0.49179177381723277,
+      "learning_rate": 4.992934024061456e-06,
+      "loss": 0.6293,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6603111879476414,
+      "grad_norm": 0.4818457491170986,
+      "learning_rate": 4.9929222480902305e-06,
+      "loss": 0.612,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6607063472462337,
+      "grad_norm": 0.4853439003519475,
+      "learning_rate": 4.992910462328319e-06,
+      "loss": 0.6331,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6611015065448259,
+      "grad_norm": 0.4736906753344473,
+      "learning_rate": 4.99289866677577e-06,
+      "loss": 0.6233,
+      "step": 1673
+    },
+    {
+      "epoch": 0.6614966658434182,
+      "grad_norm": 0.4875165225064638,
+      "learning_rate": 4.992886861432628e-06,
+      "loss": 0.6267,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6618918251420104,
+      "grad_norm": 0.4808588973957292,
+      "learning_rate": 4.99287504629894e-06,
+      "loss": 0.6172,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6622869844406026,
+      "grad_norm": 0.4907208193211784,
+      "learning_rate": 4.992863221374753e-06,
+      "loss": 0.617,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6626821437391949,
+      "grad_norm": 0.48958473331720903,
+      "learning_rate": 4.992851386660114e-06,
+      "loss": 0.6482,
+      "step": 1677
+    },
+    {
+      "epoch": 0.6630773030377871,
+      "grad_norm": 0.4746449843260302,
+      "learning_rate": 4.992839542155067e-06,
+      "loss": 0.6363,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6634724623363794,
+      "grad_norm": 0.5081053936197776,
+      "learning_rate": 4.9928276878596605e-06,
+      "loss": 0.6349,
+      "step": 1679
+    },
+    {
+      "epoch": 0.6638676216349716,
+      "grad_norm": 0.5059779639168953,
+      "learning_rate": 4.99281582377394e-06,
+      "loss": 0.6479,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6642627809335638,
+      "grad_norm": 0.49460659054649947,
+      "learning_rate": 4.992803949897954e-06,
+      "loss": 0.6161,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6646579402321561,
+      "grad_norm": 0.48860656949577014,
+      "learning_rate": 4.992792066231746e-06,
+      "loss": 0.6147,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6650530995307483,
+      "grad_norm": 0.4990949380069287,
+      "learning_rate": 4.992780172775366e-06,
+      "loss": 0.6687,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6654482588293406,
+      "grad_norm": 0.50449571493179,
+      "learning_rate": 4.9927682695288584e-06,
+      "loss": 0.6234,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6658434181279328,
+      "grad_norm": 0.5097265264943369,
+      "learning_rate": 4.992756356492271e-06,
+      "loss": 0.6345,
+      "step": 1685
+    },
+    {
+      "epoch": 0.666238577426525,
+      "grad_norm": 0.4883277462528845,
+      "learning_rate": 4.99274443366565e-06,
+      "loss": 0.6419,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6666337367251173,
+      "grad_norm": 0.47970413042564153,
+      "learning_rate": 4.992732501049044e-06,
+      "loss": 0.6036,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6670288960237095,
+      "grad_norm": 0.4935176772839213,
+      "learning_rate": 4.992720558642496e-06,
+      "loss": 0.6463,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6674240553223018,
+      "grad_norm": 0.4759709026940183,
+      "learning_rate": 4.9927086064460575e-06,
+      "loss": 0.6537,
+      "step": 1689
+    },
+    {
+      "epoch": 0.667819214620894,
+      "grad_norm": 0.47430486816189027,
+      "learning_rate": 4.992696644459771e-06,
+      "loss": 0.6317,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6682143739194862,
+      "grad_norm": 0.5034833730971029,
+      "learning_rate": 4.992684672683688e-06,
+      "loss": 0.627,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6686095332180786,
+      "grad_norm": 0.519836264221229,
+      "learning_rate": 4.992672691117852e-06,
+      "loss": 0.6538,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6690046925166708,
+      "grad_norm": 0.5050122558075558,
+      "learning_rate": 4.992660699762311e-06,
+      "loss": 0.6473,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6693998518152631,
+      "grad_norm": 0.4707732129541408,
+      "learning_rate": 4.992648698617113e-06,
+      "loss": 0.6076,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6697950111138553,
+      "grad_norm": 0.49511795117802215,
+      "learning_rate": 4.9926366876823054e-06,
+      "loss": 0.6271,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6701901704124476,
+      "grad_norm": 0.49908332035521424,
+      "learning_rate": 4.992624666957932e-06,
+      "loss": 0.6431,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6705853297110398,
+      "grad_norm": 0.48240971327251186,
+      "learning_rate": 4.992612636444045e-06,
+      "loss": 0.6218,
+      "step": 1697
+    },
+    {
+      "epoch": 0.670980489009632,
+      "grad_norm": 0.5028260213765751,
+      "learning_rate": 4.992600596140688e-06,
+      "loss": 0.6142,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6713756483082243,
+      "grad_norm": 0.49372992278625305,
+      "learning_rate": 4.99258854604791e-06,
+      "loss": 0.6289,
+      "step": 1699
+    },
+    {
+      "epoch": 0.6717708076068165,
+      "grad_norm": 0.48990901770488715,
+      "learning_rate": 4.9925764861657575e-06,
+      "loss": 0.6139,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6721659669054088,
+      "grad_norm": 0.5062662633187875,
+      "learning_rate": 4.9925644164942776e-06,
+      "loss": 0.6363,
+      "step": 1701
+    },
+    {
+      "epoch": 0.672561126204001,
+      "grad_norm": 0.5000686626853466,
+      "learning_rate": 4.992552337033519e-06,
+      "loss": 0.6062,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6729562855025932,
+      "grad_norm": 0.49142806262561756,
+      "learning_rate": 4.992540247783528e-06,
+      "loss": 0.639,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6733514448011855,
+      "grad_norm": 0.49049346875912636,
+      "learning_rate": 4.992528148744353e-06,
+      "loss": 0.6244,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6737466040997777,
+      "grad_norm": 0.47745945762367137,
+      "learning_rate": 4.99251603991604e-06,
+      "loss": 0.6541,
+      "step": 1705
+    },
+    {
+      "epoch": 0.67414176339837,
+      "grad_norm": 0.4842192684853816,
+      "learning_rate": 4.992503921298638e-06,
+      "loss": 0.621,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6745369226969622,
+      "grad_norm": 0.4870080088135584,
+      "learning_rate": 4.992491792892194e-06,
+      "loss": 0.6359,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6749320819955544,
+      "grad_norm": 0.48850423253720654,
+      "learning_rate": 4.992479654696757e-06,
+      "loss": 0.6375,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6753272412941467,
+      "grad_norm": 0.49822947926968186,
+      "learning_rate": 4.992467506712372e-06,
+      "loss": 0.6399,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6757224005927389,
+      "grad_norm": 0.48888282315789844,
+      "learning_rate": 4.992455348939088e-06,
+      "loss": 0.6275,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6761175598913312,
+      "grad_norm": 0.4782716794651655,
+      "learning_rate": 4.992443181376954e-06,
+      "loss": 0.6353,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6765127191899234,
+      "grad_norm": 0.49789653142772966,
+      "learning_rate": 4.992431004026016e-06,
+      "loss": 0.6202,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6769078784885156,
+      "grad_norm": 0.49660191988375124,
+      "learning_rate": 4.992418816886322e-06,
+      "loss": 0.6237,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6773030377871079,
+      "grad_norm": 0.5012417152818854,
+      "learning_rate": 4.992406619957922e-06,
+      "loss": 0.6216,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6776981970857002,
+      "grad_norm": 0.4839082971693346,
+      "learning_rate": 4.992394413240861e-06,
+      "loss": 0.6118,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6780933563842925,
+      "grad_norm": 0.4965208023800787,
+      "learning_rate": 4.992382196735188e-06,
+      "loss": 0.6224,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6784885156828847,
+      "grad_norm": 0.5042338428138545,
+      "learning_rate": 4.992369970440952e-06,
+      "loss": 0.643,
+      "step": 1717
+    },
+    {
+      "epoch": 0.678883674981477,
+      "grad_norm": 0.4975143700279896,
+      "learning_rate": 4.9923577343582e-06,
+      "loss": 0.6136,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6792788342800692,
+      "grad_norm": 0.5281177934717292,
+      "learning_rate": 4.992345488486979e-06,
+      "loss": 0.6397,
+      "step": 1719
+    },
+    {
+      "epoch": 0.6796739935786614,
+      "grad_norm": 0.4881426918498292,
+      "learning_rate": 4.99233323282734e-06,
+      "loss": 0.6249,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6800691528772537,
+      "grad_norm": 0.49409156351025163,
+      "learning_rate": 4.992320967379329e-06,
+      "loss": 0.6397,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6804643121758459,
+      "grad_norm": 0.48346096985864084,
+      "learning_rate": 4.992308692142995e-06,
+      "loss": 0.633,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6808594714744381,
+      "grad_norm": 0.4926531840900509,
+      "learning_rate": 4.992296407118385e-06,
+      "loss": 0.6346,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6812546307730304,
+      "grad_norm": 0.4894311266855414,
+      "learning_rate": 4.992284112305549e-06,
+      "loss": 0.6086,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6816497900716226,
+      "grad_norm": 0.49490153693793343,
+      "learning_rate": 4.992271807704534e-06,
+      "loss": 0.6318,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6820449493702149,
+      "grad_norm": 0.4828850751455796,
+      "learning_rate": 4.9922594933153884e-06,
+      "loss": 0.632,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6824401086688071,
+      "grad_norm": 0.5330475769017348,
+      "learning_rate": 4.992247169138161e-06,
+      "loss": 0.6269,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6828352679673994,
+      "grad_norm": 0.48050985014818776,
+      "learning_rate": 4.9922348351729e-06,
+      "loss": 0.6349,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6832304272659916,
+      "grad_norm": 0.48483280529770134,
+      "learning_rate": 4.992222491419655e-06,
+      "loss": 0.6555,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6836255865645838,
+      "grad_norm": 0.5080406635977245,
+      "learning_rate": 4.992210137878472e-06,
+      "loss": 0.6359,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6840207458631761,
+      "grad_norm": 0.5089264759702186,
+      "learning_rate": 4.9921977745494025e-06,
+      "loss": 0.6406,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6844159051617683,
+      "grad_norm": 0.48659054045056965,
+      "learning_rate": 4.992185401432493e-06,
+      "loss": 0.602,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6848110644603606,
+      "grad_norm": 0.49059051256711417,
+      "learning_rate": 4.992173018527791e-06,
+      "loss": 0.6035,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6852062237589528,
+      "grad_norm": 0.5182172495516812,
+      "learning_rate": 4.992160625835348e-06,
+      "loss": 0.643,
+      "step": 1734
+    },
+    {
+      "epoch": 0.685601383057545,
+      "grad_norm": 0.49425214892911673,
+      "learning_rate": 4.992148223355211e-06,
+      "loss": 0.6326,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6859965423561373,
+      "grad_norm": 0.5023535053495725,
+      "learning_rate": 4.9921358110874295e-06,
+      "loss": 0.6416,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6863917016547295,
+      "grad_norm": 0.6023947929077457,
+      "learning_rate": 4.992123389032052e-06,
+      "loss": 0.6429,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6867868609533218,
+      "grad_norm": 0.5037834252439649,
+      "learning_rate": 4.992110957189126e-06,
+      "loss": 0.6277,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6871820202519141,
+      "grad_norm": 0.49549521269729524,
+      "learning_rate": 4.992098515558702e-06,
+      "loss": 0.6305,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6875771795505063,
+      "grad_norm": 0.49551911885336786,
+      "learning_rate": 4.992086064140829e-06,
+      "loss": 0.6248,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6879723388490986,
+      "grad_norm": 0.5246656926007554,
+      "learning_rate": 4.9920736029355544e-06,
+      "loss": 0.6281,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6883674981476908,
+      "grad_norm": 0.5001944530177468,
+      "learning_rate": 4.992061131942929e-06,
+      "loss": 0.6261,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6887626574462831,
+      "grad_norm": 0.4826806905887991,
+      "learning_rate": 4.992048651163e-06,
+      "loss": 0.6205,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6891578167448753,
+      "grad_norm": 0.5077638034008282,
+      "learning_rate": 4.992036160595817e-06,
+      "loss": 0.6366,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6895529760434675,
+      "grad_norm": 0.4957736986969226,
+      "learning_rate": 4.9920236602414295e-06,
+      "loss": 0.6197,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6899481353420598,
+      "grad_norm": 0.4934343040818385,
+      "learning_rate": 4.992011150099886e-06,
+      "loss": 0.6381,
+      "step": 1746
+    },
+    {
+      "epoch": 0.690343294640652,
+      "grad_norm": 0.492375853768171,
+      "learning_rate": 4.991998630171236e-06,
+      "loss": 0.6313,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6907384539392443,
+      "grad_norm": 0.47870636501116376,
+      "learning_rate": 4.991986100455529e-06,
+      "loss": 0.6038,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6911336132378365,
+      "grad_norm": 0.5048304913807063,
+      "learning_rate": 4.991973560952813e-06,
+      "loss": 0.6293,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6915287725364287,
+      "grad_norm": 0.5249637722841536,
+      "learning_rate": 4.991961011663139e-06,
+      "loss": 0.6307,
+      "step": 1750
+    },
+    {
+      "epoch": 0.691923931835021,
+      "grad_norm": 0.5208870005299038,
+      "learning_rate": 4.991948452586555e-06,
+      "loss": 0.6332,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6923190911336132,
+      "grad_norm": 0.5015944555669404,
+      "learning_rate": 4.991935883723111e-06,
+      "loss": 0.5839,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6927142504322055,
+      "grad_norm": 0.5163525589308585,
+      "learning_rate": 4.991923305072856e-06,
+      "loss": 0.6235,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6931094097307977,
+      "grad_norm": 0.5032310673814737,
+      "learning_rate": 4.991910716635838e-06,
+      "loss": 0.6263,
+      "step": 1754
+    },
+    {
+      "epoch": 0.69350456902939,
+      "grad_norm": 0.47440600362048463,
+      "learning_rate": 4.991898118412109e-06,
+      "loss": 0.6376,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6938997283279822,
+      "grad_norm": 0.47177908043565303,
+      "learning_rate": 4.991885510401717e-06,
+      "loss": 0.5965,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6942948876265744,
+      "grad_norm": 0.5050010124065104,
+      "learning_rate": 4.991872892604713e-06,
+      "loss": 0.6443,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6946900469251667,
+      "grad_norm": 0.4703390805427689,
+      "learning_rate": 4.991860265021144e-06,
+      "loss": 0.6371,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6950852062237589,
+      "grad_norm": 0.5965162740676478,
+      "learning_rate": 4.991847627651062e-06,
+      "loss": 0.6325,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6954803655223512,
+      "grad_norm": 0.4995953085902974,
+      "learning_rate": 4.991834980494515e-06,
+      "loss": 0.6124,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6958755248209434,
+      "grad_norm": 0.48410450082082324,
+      "learning_rate": 4.991822323551554e-06,
+      "loss": 0.6543,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6962706841195357,
+      "grad_norm": 0.5097060752514024,
+      "learning_rate": 4.991809656822227e-06,
+      "loss": 0.6341,
+      "step": 1762
+    },
+    {
+      "epoch": 0.696665843418128,
+      "grad_norm": 0.5028866911312128,
+      "learning_rate": 4.991796980306586e-06,
+      "loss": 0.6196,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6970610027167202,
+      "grad_norm": 0.48181973244537724,
+      "learning_rate": 4.991784294004679e-06,
+      "loss": 0.6142,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6974561620153125,
+      "grad_norm": 0.481960114220102,
+      "learning_rate": 4.991771597916556e-06,
+      "loss": 0.6064,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6978513213139047,
+      "grad_norm": 0.4962930971445703,
+      "learning_rate": 4.9917588920422675e-06,
+      "loss": 0.6415,
+      "step": 1766
+    },
+    {
+      "epoch": 0.698246480612497,
+      "grad_norm": 0.49440862308297445,
+      "learning_rate": 4.991746176381863e-06,
+      "loss": 0.6169,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6986416399110892,
+      "grad_norm": 0.5047352213306215,
+      "learning_rate": 4.991733450935393e-06,
+      "loss": 0.6304,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6990367992096814,
+      "grad_norm": 0.4733163582966052,
+      "learning_rate": 4.991720715702907e-06,
+      "loss": 0.6272,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6994319585082737,
+      "grad_norm": 0.5098227044422302,
+      "learning_rate": 4.991707970684455e-06,
+      "loss": 0.6364,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6998271178068659,
+      "grad_norm": 0.49090533225232885,
+      "learning_rate": 4.991695215880087e-06,
+      "loss": 0.6415,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7002222771054581,
+      "grad_norm": 0.49614105702660316,
+      "learning_rate": 4.991682451289853e-06,
+      "loss": 0.6588,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7006174364040504,
+      "grad_norm": 0.47774693460503354,
+      "learning_rate": 4.991669676913804e-06,
+      "loss": 0.6316,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7010125957026426,
+      "grad_norm": 0.4822380940669454,
+      "learning_rate": 4.991656892751989e-06,
+      "loss": 0.6086,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7014077550012349,
+      "grad_norm": 0.472072856981578,
+      "learning_rate": 4.99164409880446e-06,
+      "loss": 0.6126,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7018029142998271,
+      "grad_norm": 0.5033099153233312,
+      "learning_rate": 4.991631295071265e-06,
+      "loss": 0.6334,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7021980735984193,
+      "grad_norm": 0.48177193738949203,
+      "learning_rate": 4.991618481552455e-06,
+      "loss": 0.6191,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7025932328970116,
+      "grad_norm": 0.5635001881112269,
+      "learning_rate": 4.9916056582480805e-06,
+      "loss": 0.6255,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7029883921956038,
+      "grad_norm": 0.497837050692373,
+      "learning_rate": 4.991592825158192e-06,
+      "loss": 0.6277,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7033835514941961,
+      "grad_norm": 0.5046834438070316,
+      "learning_rate": 4.991579982282841e-06,
+      "loss": 0.6419,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7037787107927883,
+      "grad_norm": 0.4902373827129985,
+      "learning_rate": 4.991567129622076e-06,
+      "loss": 0.6432,
+      "step": 1781
+    },
+    {
+      "epoch": 0.7041738700913805,
+      "grad_norm": 0.5067434826741889,
+      "learning_rate": 4.991554267175947e-06,
+      "loss": 0.6453,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7045690293899728,
+      "grad_norm": 0.49398025290400144,
+      "learning_rate": 4.991541394944508e-06,
+      "loss": 0.6082,
+      "step": 1783
+    },
+    {
+      "epoch": 0.704964188688565,
+      "grad_norm": 0.500932111828932,
+      "learning_rate": 4.991528512927806e-06,
+      "loss": 0.6222,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7053593479871573,
+      "grad_norm": 0.4855703145150172,
+      "learning_rate": 4.991515621125893e-06,
+      "loss": 0.6395,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7057545072857496,
+      "grad_norm": 0.48895788982354876,
+      "learning_rate": 4.99150271953882e-06,
+      "loss": 0.6254,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7061496665843419,
+      "grad_norm": 0.49706293010128083,
+      "learning_rate": 4.9914898081666375e-06,
+      "loss": 0.6225,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7065448258829341,
+      "grad_norm": 0.5177841757480415,
+      "learning_rate": 4.991476887009395e-06,
+      "loss": 0.6477,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7069399851815263,
+      "grad_norm": 0.4768336877804729,
+      "learning_rate": 4.991463956067145e-06,
+      "loss": 0.6427,
+      "step": 1789
+    },
+    {
+      "epoch": 0.7073351444801186,
+      "grad_norm": 0.48872337485295864,
+      "learning_rate": 4.9914510153399375e-06,
+      "loss": 0.6283,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7077303037787108,
+      "grad_norm": 0.5233174541443238,
+      "learning_rate": 4.9914380648278224e-06,
+      "loss": 0.6483,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7081254630773031,
+      "grad_norm": 0.48167766487074626,
+      "learning_rate": 4.991425104530852e-06,
+      "loss": 0.6253,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7085206223758953,
+      "grad_norm": 0.49635669925532444,
+      "learning_rate": 4.991412134449078e-06,
+      "loss": 0.6354,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7089157816744875,
+      "grad_norm": 0.494107036823578,
+      "learning_rate": 4.991399154582548e-06,
+      "loss": 0.618,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7093109409730798,
+      "grad_norm": 0.48336185208770716,
+      "learning_rate": 4.991386164931316e-06,
+      "loss": 0.6236,
+      "step": 1795
+    },
+    {
+      "epoch": 0.709706100271672,
+      "grad_norm": 0.49749633745530736,
+      "learning_rate": 4.991373165495431e-06,
+      "loss": 0.6336,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7101012595702643,
+      "grad_norm": 0.5034095057979783,
+      "learning_rate": 4.991360156274946e-06,
+      "loss": 0.647,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7104964188688565,
+      "grad_norm": 0.5322413221010839,
+      "learning_rate": 4.9913471372699115e-06,
+      "loss": 0.6011,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7108915781674487,
+      "grad_norm": 0.49545526697939474,
+      "learning_rate": 4.991334108480377e-06,
+      "loss": 0.6364,
+      "step": 1799
+    },
+    {
+      "epoch": 0.711286737466041,
+      "grad_norm": 0.49309817729132516,
+      "learning_rate": 4.9913210699063965e-06,
+      "loss": 0.614,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7116818967646332,
+      "grad_norm": 0.5691944811153536,
+      "learning_rate": 4.991308021548018e-06,
+      "loss": 0.6186,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7120770560632255,
+      "grad_norm": 0.5449350122228545,
+      "learning_rate": 4.9912949634052955e-06,
+      "loss": 0.6247,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7124722153618177,
+      "grad_norm": 0.5049416194812504,
+      "learning_rate": 4.991281895478279e-06,
+      "loss": 0.5903,
+      "step": 1803
+    },
+    {
+      "epoch": 0.71286737466041,
+      "grad_norm": 0.47208342204692466,
+      "learning_rate": 4.9912688177670195e-06,
+      "loss": 0.6231,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7132625339590022,
+      "grad_norm": 0.5243565371621691,
+      "learning_rate": 4.991255730271569e-06,
+      "loss": 0.6277,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7136576932575944,
+      "grad_norm": 0.5670677049532014,
+      "learning_rate": 4.991242632991979e-06,
+      "loss": 0.6372,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7140528525561867,
+      "grad_norm": 0.5158307483356673,
+      "learning_rate": 4.9912295259283015e-06,
+      "loss": 0.6125,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7144480118547789,
+      "grad_norm": 0.508525557153672,
+      "learning_rate": 4.991216409080586e-06,
+      "loss": 0.6101,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7148431711533711,
+      "grad_norm": 0.5076780173186427,
+      "learning_rate": 4.9912032824488855e-06,
+      "loss": 0.6345,
+      "step": 1809
+    },
+    {
+      "epoch": 0.7152383304519635,
+      "grad_norm": 0.5069119882804742,
+      "learning_rate": 4.991190146033251e-06,
+      "loss": 0.5988,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7156334897505557,
+      "grad_norm": 0.5111101135704393,
+      "learning_rate": 4.991176999833734e-06,
+      "loss": 0.6285,
+      "step": 1811
+    },
+    {
+      "epoch": 0.716028649049148,
+      "grad_norm": 0.494023778451646,
+      "learning_rate": 4.991163843850388e-06,
+      "loss": 0.6271,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7164238083477402,
+      "grad_norm": 0.49401028308719136,
+      "learning_rate": 4.991150678083262e-06,
+      "loss": 0.6162,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7168189676463325,
+      "grad_norm": 0.5455894933511418,
+      "learning_rate": 4.99113750253241e-06,
+      "loss": 0.6207,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7172141269449247,
+      "grad_norm": 0.5038767858357743,
+      "learning_rate": 4.991124317197881e-06,
+      "loss": 0.6508,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7176092862435169,
+      "grad_norm": 0.49228914244907324,
+      "learning_rate": 4.991111122079729e-06,
+      "loss": 0.6566,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7180044455421092,
+      "grad_norm": 0.4939671059206662,
+      "learning_rate": 4.991097917178005e-06,
+      "loss": 0.6153,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7183996048407014,
+      "grad_norm": 0.48627356885183,
+      "learning_rate": 4.991084702492761e-06,
+      "loss": 0.6016,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7187947641392937,
+      "grad_norm": 0.4940616525370284,
+      "learning_rate": 4.99107147802405e-06,
+      "loss": 0.6139,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7191899234378859,
+      "grad_norm": 0.498285217304123,
+      "learning_rate": 4.991058243771922e-06,
+      "loss": 0.6108,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7195850827364781,
+      "grad_norm": 0.5053900244808598,
+      "learning_rate": 4.9910449997364295e-06,
+      "loss": 0.632,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7199802420350704,
+      "grad_norm": 0.5073136517973986,
+      "learning_rate": 4.991031745917626e-06,
+      "loss": 0.6369,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7203754013336626,
+      "grad_norm": 0.5195708029162378,
+      "learning_rate": 4.991018482315561e-06,
+      "loss": 0.6273,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7207705606322549,
+      "grad_norm": 0.48766761832670025,
+      "learning_rate": 4.99100520893029e-06,
+      "loss": 0.6362,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7211657199308471,
+      "grad_norm": 0.5320190857436362,
+      "learning_rate": 4.990991925761862e-06,
+      "loss": 0.6209,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7215608792294393,
+      "grad_norm": 0.5362331175084071,
+      "learning_rate": 4.99097863281033e-06,
+      "loss": 0.6234,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7219560385280316,
+      "grad_norm": 0.46797239650415506,
+      "learning_rate": 4.990965330075746e-06,
+      "loss": 0.6149,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7223511978266238,
+      "grad_norm": 0.49165256567192805,
+      "learning_rate": 4.990952017558164e-06,
+      "loss": 0.6197,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7227463571252161,
+      "grad_norm": 0.5734478891642341,
+      "learning_rate": 4.9909386952576355e-06,
+      "loss": 0.6099,
+      "step": 1829
+    },
+    {
+      "epoch": 0.7231415164238083,
+      "grad_norm": 0.5067294740219436,
+      "learning_rate": 4.9909253631742115e-06,
+      "loss": 0.6278,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7235366757224005,
+      "grad_norm": 0.5289952166160319,
+      "learning_rate": 4.990912021307945e-06,
+      "loss": 0.6258,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7239318350209928,
+      "grad_norm": 0.5043539597608305,
+      "learning_rate": 4.990898669658889e-06,
+      "loss": 0.6166,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7243269943195851,
+      "grad_norm": 0.5137481323998266,
+      "learning_rate": 4.990885308227096e-06,
+      "loss": 0.6349,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7247221536181774,
+      "grad_norm": 0.5223800914526201,
+      "learning_rate": 4.9908719370126175e-06,
+      "loss": 0.6335,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7251173129167696,
+      "grad_norm": 0.49483421508060577,
+      "learning_rate": 4.990858556015507e-06,
+      "loss": 0.6243,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7255124722153619,
+      "grad_norm": 0.507533775647517,
+      "learning_rate": 4.990845165235816e-06,
+      "loss": 0.6288,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7259076315139541,
+      "grad_norm": 0.5337126385462883,
+      "learning_rate": 4.990831764673598e-06,
+      "loss": 0.6422,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7263027908125463,
+      "grad_norm": 0.8373622357864616,
+      "learning_rate": 4.9908183543289055e-06,
+      "loss": 0.6366,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7266979501111386,
+      "grad_norm": 0.4867027048473409,
+      "learning_rate": 4.99080493420179e-06,
+      "loss": 0.6196,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7270931094097308,
+      "grad_norm": 0.5143251024891161,
+      "learning_rate": 4.990791504292307e-06,
+      "loss": 0.6378,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7274882687083231,
+      "grad_norm": 0.4925899162999197,
+      "learning_rate": 4.990778064600506e-06,
+      "loss": 0.6077,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7278834280069153,
+      "grad_norm": 0.46958077754810756,
+      "learning_rate": 4.990764615126442e-06,
+      "loss": 0.6249,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7282785873055075,
+      "grad_norm": 0.502560018511133,
+      "learning_rate": 4.990751155870167e-06,
+      "loss": 0.6106,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7286737466040998,
+      "grad_norm": 0.5082154553642297,
+      "learning_rate": 4.990737686831734e-06,
+      "loss": 0.6111,
+      "step": 1844
+    },
+    {
+      "epoch": 0.729068905902692,
+      "grad_norm": 0.48176602480847014,
+      "learning_rate": 4.990724208011195e-06,
+      "loss": 0.6225,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7294640652012843,
+      "grad_norm": 0.47402002408096394,
+      "learning_rate": 4.990710719408604e-06,
+      "loss": 0.6258,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7298592244998765,
+      "grad_norm": 0.4718364904587041,
+      "learning_rate": 4.9906972210240146e-06,
+      "loss": 0.6206,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7302543837984687,
+      "grad_norm": 0.48775817606336896,
+      "learning_rate": 4.990683712857479e-06,
+      "loss": 0.6176,
+      "step": 1848
+    },
+    {
+      "epoch": 0.730649543097061,
+      "grad_norm": 0.48862016729905866,
+      "learning_rate": 4.99067019490905e-06,
+      "loss": 0.6477,
+      "step": 1849
+    },
+    {
+      "epoch": 0.7310447023956532,
+      "grad_norm": 0.4851256824616345,
+      "learning_rate": 4.990656667178781e-06,
+      "loss": 0.6297,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7314398616942455,
+      "grad_norm": 0.4900449934686768,
+      "learning_rate": 4.9906431296667235e-06,
+      "loss": 0.6225,
+      "step": 1851
+    },
+    {
+      "epoch": 0.7318350209928377,
+      "grad_norm": 0.4803057428812473,
+      "learning_rate": 4.9906295823729334e-06,
+      "loss": 0.6096,
+      "step": 1852
+    },
+    {
+      "epoch": 0.7322301802914299,
+      "grad_norm": 0.5084616847011403,
+      "learning_rate": 4.990616025297462e-06,
+      "loss": 0.6402,
+      "step": 1853
+    },
+    {
+      "epoch": 0.7326253395900222,
+      "grad_norm": 0.48222218139870393,
+      "learning_rate": 4.990602458440364e-06,
+      "loss": 0.618,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7330204988886144,
+      "grad_norm": 0.4768756457104444,
+      "learning_rate": 4.990588881801692e-06,
+      "loss": 0.6277,
+      "step": 1855
+    },
+    {
+      "epoch": 0.7334156581872067,
+      "grad_norm": 0.47054304737078606,
+      "learning_rate": 4.990575295381499e-06,
+      "loss": 0.639,
+      "step": 1856
+    },
+    {
+      "epoch": 0.733810817485799,
+      "grad_norm": 0.4954513777862497,
+      "learning_rate": 4.990561699179838e-06,
+      "loss": 0.6336,
+      "step": 1857
+    },
+    {
+      "epoch": 0.7342059767843913,
+      "grad_norm": 0.4977664948935915,
+      "learning_rate": 4.990548093196765e-06,
+      "loss": 0.6408,
+      "step": 1858
+    },
+    {
+      "epoch": 0.7346011360829835,
+      "grad_norm": 0.4821488918104782,
+      "learning_rate": 4.9905344774323285e-06,
+      "loss": 0.6477,
+      "step": 1859
+    },
+    {
+      "epoch": 0.7349962953815757,
+      "grad_norm": 0.47164432904373776,
+      "learning_rate": 4.990520851886586e-06,
+      "loss": 0.6116,
+      "step": 1860
+    },
+    {
+      "epoch": 0.735391454680168,
+      "grad_norm": 0.5130522518926274,
+      "learning_rate": 4.990507216559591e-06,
+      "loss": 0.6285,
+      "step": 1861
+    },
+    {
+      "epoch": 0.7357866139787602,
+      "grad_norm": 0.5085662976075583,
+      "learning_rate": 4.990493571451396e-06,
+      "loss": 0.6263,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7361817732773525,
+      "grad_norm": 0.48426222103509337,
+      "learning_rate": 4.990479916562053e-06,
+      "loss": 0.6231,
+      "step": 1863
+    },
+    {
+      "epoch": 0.7365769325759447,
+      "grad_norm": 0.5176828196274378,
+      "learning_rate": 4.990466251891618e-06,
+      "loss": 0.6608,
+      "step": 1864
+    },
+    {
+      "epoch": 0.7369720918745369,
+      "grad_norm": 0.5221468288688574,
+      "learning_rate": 4.990452577440144e-06,
+      "loss": 0.6274,
+      "step": 1865
+    },
+    {
+      "epoch": 0.7373672511731292,
+      "grad_norm": 0.47850681281032353,
+      "learning_rate": 4.990438893207684e-06,
+      "loss": 0.6229,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7377624104717214,
+      "grad_norm": 0.543170513792422,
+      "learning_rate": 4.990425199194293e-06,
+      "loss": 0.6256,
+      "step": 1867
+    },
+    {
+      "epoch": 0.7381575697703137,
+      "grad_norm": 0.5436786053164533,
+      "learning_rate": 4.990411495400024e-06,
+      "loss": 0.6326,
+      "step": 1868
+    },
+    {
+      "epoch": 0.7385527290689059,
+      "grad_norm": 0.4779013731244117,
+      "learning_rate": 4.9903977818249305e-06,
+      "loss": 0.624,
+      "step": 1869
+    },
+    {
+      "epoch": 0.7389478883674981,
+      "grad_norm": 0.5107248359446434,
+      "learning_rate": 4.9903840584690675e-06,
+      "loss": 0.6465,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7393430476660904,
+      "grad_norm": 0.5191444772993868,
+      "learning_rate": 4.990370325332488e-06,
+      "loss": 0.6215,
+      "step": 1871
+    },
+    {
+      "epoch": 0.7397382069646826,
+      "grad_norm": 0.5148010616996448,
+      "learning_rate": 4.990356582415245e-06,
+      "loss": 0.6254,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7401333662632749,
+      "grad_norm": 0.4822825362300129,
+      "learning_rate": 4.990342829717394e-06,
+      "loss": 0.6275,
+      "step": 1873
+    },
+    {
+      "epoch": 0.7405285255618671,
+      "grad_norm": 0.5111622322537774,
+      "learning_rate": 4.9903290672389895e-06,
+      "loss": 0.6222,
+      "step": 1874
+    },
+    {
+      "epoch": 0.7409236848604593,
+      "grad_norm": 0.524971372504675,
+      "learning_rate": 4.990315294980083e-06,
+      "loss": 0.6329,
+      "step": 1875
+    },
+    {
+      "epoch": 0.7413188441590516,
+      "grad_norm": 0.46981819397533325,
+      "learning_rate": 4.990301512940732e-06,
+      "loss": 0.6255,
+      "step": 1876
+    },
+    {
+      "epoch": 0.7417140034576438,
+      "grad_norm": 0.5061645450254407,
+      "learning_rate": 4.990287721120988e-06,
+      "loss": 0.6332,
+      "step": 1877
+    },
+    {
+      "epoch": 0.7421091627562361,
+      "grad_norm": 0.5081802490588557,
+      "learning_rate": 4.990273919520906e-06,
+      "loss": 0.6223,
+      "step": 1878
+    },
+    {
+      "epoch": 0.7425043220548283,
+      "grad_norm": 0.4904066696859732,
+      "learning_rate": 4.990260108140541e-06,
+      "loss": 0.6403,
+      "step": 1879
+    },
+    {
+      "epoch": 0.7428994813534205,
+      "grad_norm": 0.4991704339399477,
+      "learning_rate": 4.990246286979945e-06,
+      "loss": 0.6378,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7432946406520129,
+      "grad_norm": 0.4834570028068084,
+      "learning_rate": 4.9902324560391745e-06,
+      "loss": 0.6003,
+      "step": 1881
+    },
+    {
+      "epoch": 0.7436897999506051,
+      "grad_norm": 0.4946859173466393,
+      "learning_rate": 4.990218615318283e-06,
+      "loss": 0.609,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7440849592491974,
+      "grad_norm": 0.4988829525100141,
+      "learning_rate": 4.990204764817326e-06,
+      "loss": 0.6338,
+      "step": 1883
+    },
+    {
+      "epoch": 0.7444801185477896,
+      "grad_norm": 0.5882873111602047,
+      "learning_rate": 4.990190904536355e-06,
+      "loss": 0.652,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7448752778463819,
+      "grad_norm": 0.4988958396253975,
+      "learning_rate": 4.990177034475427e-06,
+      "loss": 0.626,
+      "step": 1885
+    },
+    {
+      "epoch": 0.7452704371449741,
+      "grad_norm": 0.495275464799001,
+      "learning_rate": 4.990163154634596e-06,
+      "loss": 0.6095,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7456655964435663,
+      "grad_norm": 0.4919380989234139,
+      "learning_rate": 4.990149265013916e-06,
+      "loss": 0.6211,
+      "step": 1887
+    },
+    {
+      "epoch": 0.7460607557421586,
+      "grad_norm": 0.5019734847912949,
+      "learning_rate": 4.990135365613442e-06,
+      "loss": 0.6299,
+      "step": 1888
+    },
+    {
+      "epoch": 0.7464559150407508,
+      "grad_norm": 0.4828661219472996,
+      "learning_rate": 4.9901214564332275e-06,
+      "loss": 0.6202,
+      "step": 1889
+    },
+    {
+      "epoch": 0.7468510743393431,
+      "grad_norm": 0.48107070668497365,
+      "learning_rate": 4.990107537473329e-06,
+      "loss": 0.6092,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7472462336379353,
+      "grad_norm": 0.4791747545680655,
+      "learning_rate": 4.9900936087338e-06,
+      "loss": 0.6447,
+      "step": 1891
+    },
+    {
+      "epoch": 0.7476413929365275,
+      "grad_norm": 0.49731269148004137,
+      "learning_rate": 4.990079670214696e-06,
+      "loss": 0.6332,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7480365522351198,
+      "grad_norm": 0.49528011354972196,
+      "learning_rate": 4.99006572191607e-06,
+      "loss": 0.6231,
+      "step": 1893
+    },
+    {
+      "epoch": 0.748431711533712,
+      "grad_norm": 0.4967662330741853,
+      "learning_rate": 4.990051763837978e-06,
+      "loss": 0.635,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7488268708323043,
+      "grad_norm": 0.5265304272477109,
+      "learning_rate": 4.990037795980475e-06,
+      "loss": 0.622,
+      "step": 1895
+    },
+    {
+      "epoch": 0.7492220301308965,
+      "grad_norm": 0.49557723563808614,
+      "learning_rate": 4.990023818343615e-06,
+      "loss": 0.6204,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7496171894294887,
+      "grad_norm": 0.4721310887893004,
+      "learning_rate": 4.9900098309274544e-06,
+      "loss": 0.6307,
+      "step": 1897
+    },
+    {
+      "epoch": 0.750012348728081,
+      "grad_norm": 0.5090822375536443,
+      "learning_rate": 4.989995833732047e-06,
+      "loss": 0.6415,
+      "step": 1898
+    },
+    {
+      "epoch": 0.7504075080266732,
+      "grad_norm": 0.48847487647832843,
+      "learning_rate": 4.989981826757447e-06,
+      "loss": 0.6407,
+      "step": 1899
+    },
+    {
+      "epoch": 0.7508026673252655,
+      "grad_norm": 0.4756812146928004,
+      "learning_rate": 4.989967810003712e-06,
+      "loss": 0.6259,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7511978266238577,
+      "grad_norm": 0.4824209290886366,
+      "learning_rate": 4.989953783470895e-06,
+      "loss": 0.6306,
+      "step": 1901
+    },
+    {
+      "epoch": 0.7515929859224499,
+      "grad_norm": 0.48714913834725626,
+      "learning_rate": 4.9899397471590505e-06,
+      "loss": 0.6251,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7519881452210422,
+      "grad_norm": 0.6529166302870585,
+      "learning_rate": 4.9899257010682355e-06,
+      "loss": 0.628,
+      "step": 1903
+    },
+    {
+      "epoch": 0.7523833045196345,
+      "grad_norm": 0.4958403702141253,
+      "learning_rate": 4.989911645198504e-06,
+      "loss": 0.6198,
+      "step": 1904
+    },
+    {
+      "epoch": 0.7527784638182268,
+      "grad_norm": 0.4834392149932408,
+      "learning_rate": 4.989897579549912e-06,
+      "loss": 0.6155,
+      "step": 1905
+    },
+    {
+      "epoch": 0.753173623116819,
+      "grad_norm": 0.48216115631971956,
+      "learning_rate": 4.989883504122514e-06,
+      "loss": 0.6083,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7535687824154113,
+      "grad_norm": 0.5044002777393403,
+      "learning_rate": 4.989869418916364e-06,
+      "loss": 0.6608,
+      "step": 1907
+    },
+    {
+      "epoch": 0.7539639417140035,
+      "grad_norm": 0.4915044040208007,
+      "learning_rate": 4.98985532393152e-06,
+      "loss": 0.6306,
+      "step": 1908
+    },
+    {
+      "epoch": 0.7543591010125957,
+      "grad_norm": 0.4808889993199748,
+      "learning_rate": 4.989841219168037e-06,
+      "loss": 0.6045,
+      "step": 1909
+    },
+    {
+      "epoch": 0.754754260311188,
+      "grad_norm": 0.48355433928580754,
+      "learning_rate": 4.989827104625969e-06,
+      "loss": 0.6225,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7551494196097802,
+      "grad_norm": 0.4946507696546552,
+      "learning_rate": 4.989812980305372e-06,
+      "loss": 0.622,
+      "step": 1911
+    },
+    {
+      "epoch": 0.7555445789083725,
+      "grad_norm": 0.4821780404795558,
+      "learning_rate": 4.989798846206302e-06,
+      "loss": 0.6334,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7559397382069647,
+      "grad_norm": 0.4729609905608424,
+      "learning_rate": 4.989784702328814e-06,
+      "loss": 0.6138,
+      "step": 1913
+    },
+    {
+      "epoch": 0.7563348975055569,
+      "grad_norm": 0.47732682551126054,
+      "learning_rate": 4.989770548672962e-06,
+      "loss": 0.6308,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7567300568041492,
+      "grad_norm": 0.49741773014790336,
+      "learning_rate": 4.9897563852388046e-06,
+      "loss": 0.6096,
+      "step": 1915
+    },
+    {
+      "epoch": 0.7571252161027414,
+      "grad_norm": 0.491814907631123,
+      "learning_rate": 4.989742212026396e-06,
+      "loss": 0.6148,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7575203754013337,
+      "grad_norm": 0.4953592826702089,
+      "learning_rate": 4.989728029035791e-06,
+      "loss": 0.6176,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7579155346999259,
+      "grad_norm": 0.5041003634543201,
+      "learning_rate": 4.989713836267047e-06,
+      "loss": 0.6147,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7583106939985181,
+      "grad_norm": 0.6158285253583131,
+      "learning_rate": 4.989699633720218e-06,
+      "loss": 0.6389,
+      "step": 1919
+    },
+    {
+      "epoch": 0.7587058532971104,
+      "grad_norm": 0.5193736067753019,
+      "learning_rate": 4.989685421395361e-06,
+      "loss": 0.6441,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7591010125957026,
+      "grad_norm": 0.5216730101645742,
+      "learning_rate": 4.989671199292533e-06,
+      "loss": 0.6293,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7594961718942949,
+      "grad_norm": 0.4902177383288198,
+      "learning_rate": 4.989656967411787e-06,
+      "loss": 0.6349,
+      "step": 1922
+    },
+    {
+      "epoch": 0.7598913311928871,
+      "grad_norm": 0.48667537873584354,
+      "learning_rate": 4.9896427257531795e-06,
+      "loss": 0.5971,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7602864904914793,
+      "grad_norm": 0.5011754592981034,
+      "learning_rate": 4.9896284743167685e-06,
+      "loss": 0.6287,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7606816497900716,
+      "grad_norm": 0.49307058655487895,
+      "learning_rate": 4.989614213102608e-06,
+      "loss": 0.6161,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7610768090886638,
+      "grad_norm": 0.48115110544146544,
+      "learning_rate": 4.989599942110754e-06,
+      "loss": 0.6203,
+      "step": 1926
+    },
+    {
+      "epoch": 0.7614719683872561,
+      "grad_norm": 0.5197946460675674,
+      "learning_rate": 4.9895856613412645e-06,
+      "loss": 0.6194,
+      "step": 1927
+    },
+    {
+      "epoch": 0.7618671276858484,
+      "grad_norm": 0.4970888947426336,
+      "learning_rate": 4.989571370794194e-06,
+      "loss": 0.6471,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7622622869844407,
+      "grad_norm": 0.5938997863900486,
+      "learning_rate": 4.989557070469598e-06,
+      "loss": 0.627,
+      "step": 1929
+    },
+    {
+      "epoch": 0.7626574462830329,
+      "grad_norm": 0.4912343551544669,
+      "learning_rate": 4.989542760367535e-06,
+      "loss": 0.6073,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7630526055816251,
+      "grad_norm": 0.5035910915834263,
+      "learning_rate": 4.989528440488059e-06,
+      "loss": 0.6061,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7634477648802174,
+      "grad_norm": 0.47771768086688043,
+      "learning_rate": 4.9895141108312264e-06,
+      "loss": 0.6155,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7638429241788096,
+      "grad_norm": 0.49893837750925285,
+      "learning_rate": 4.9894997713970945e-06,
+      "loss": 0.6194,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7642380834774019,
+      "grad_norm": 0.48183105529501513,
+      "learning_rate": 4.989485422185719e-06,
+      "loss": 0.6072,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7646332427759941,
+      "grad_norm": 0.49437340564091187,
+      "learning_rate": 4.989471063197157e-06,
+      "loss": 0.6235,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7650284020745863,
+      "grad_norm": 0.5086950926898877,
+      "learning_rate": 4.989456694431464e-06,
+      "loss": 0.6204,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7654235613731786,
+      "grad_norm": 0.4769480906144647,
+      "learning_rate": 4.989442315888697e-06,
+      "loss": 0.5961,
+      "step": 1937
+    },
+    {
+      "epoch": 0.7658187206717708,
+      "grad_norm": 0.49314554262535976,
+      "learning_rate": 4.9894279275689124e-06,
+      "loss": 0.6249,
+      "step": 1938
+    },
+    {
+      "epoch": 0.766213879970363,
+      "grad_norm": 0.5154666684010718,
+      "learning_rate": 4.989413529472166e-06,
+      "loss": 0.6252,
+      "step": 1939
+    },
+    {
+      "epoch": 0.7666090392689553,
+      "grad_norm": 0.4870804225575094,
+      "learning_rate": 4.989399121598516e-06,
+      "loss": 0.6125,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7670041985675475,
+      "grad_norm": 0.5588292063163853,
+      "learning_rate": 4.989384703948017e-06,
+      "loss": 0.6152,
+      "step": 1941
+    },
+    {
+      "epoch": 0.7673993578661398,
+      "grad_norm": 0.49714045565400894,
+      "learning_rate": 4.989370276520726e-06,
+      "loss": 0.6175,
+      "step": 1942
+    },
+    {
+      "epoch": 0.767794517164732,
+      "grad_norm": 0.501785140415218,
+      "learning_rate": 4.989355839316701e-06,
+      "loss": 0.6385,
+      "step": 1943
+    },
+    {
+      "epoch": 0.7681896764633243,
+      "grad_norm": 0.4727502626196803,
+      "learning_rate": 4.989341392335998e-06,
+      "loss": 0.619,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7685848357619165,
+      "grad_norm": 0.5109703415941235,
+      "learning_rate": 4.989326935578673e-06,
+      "loss": 0.5984,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7689799950605087,
+      "grad_norm": 0.512283569257176,
+      "learning_rate": 4.9893124690447835e-06,
+      "loss": 0.6236,
+      "step": 1946
+    },
+    {
+      "epoch": 0.769375154359101,
+      "grad_norm": 0.4941171640474088,
+      "learning_rate": 4.989297992734386e-06,
+      "loss": 0.6163,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7697703136576932,
+      "grad_norm": 0.5134713647285842,
+      "learning_rate": 4.989283506647539e-06,
+      "loss": 0.6033,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7701654729562855,
+      "grad_norm": 0.5065873835627485,
+      "learning_rate": 4.9892690107842964e-06,
+      "loss": 0.6306,
+      "step": 1949
+    },
+    {
+      "epoch": 0.7705606322548777,
+      "grad_norm": 0.488059732764682,
+      "learning_rate": 4.9892545051447175e-06,
+      "loss": 0.6095,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7709557915534699,
+      "grad_norm": 0.5039415788029225,
+      "learning_rate": 4.989239989728859e-06,
+      "loss": 0.614,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7713509508520623,
+      "grad_norm": 0.5032937006287896,
+      "learning_rate": 4.989225464536776e-06,
+      "loss": 0.6511,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7717461101506545,
+      "grad_norm": 0.49467260366888127,
+      "learning_rate": 4.989210929568527e-06,
+      "loss": 0.6123,
+      "step": 1953
+    },
+    {
+      "epoch": 0.7721412694492468,
+      "grad_norm": 0.5221851652881345,
+      "learning_rate": 4.98919638482417e-06,
+      "loss": 0.6117,
+      "step": 1954
+    },
+    {
+      "epoch": 0.772536428747839,
+      "grad_norm": 0.5123539208719938,
+      "learning_rate": 4.989181830303761e-06,
+      "loss": 0.6081,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7729315880464313,
+      "grad_norm": 0.49070829163767216,
+      "learning_rate": 4.9891672660073566e-06,
+      "loss": 0.6169,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7733267473450235,
+      "grad_norm": 0.504303254265576,
+      "learning_rate": 4.989152691935015e-06,
+      "loss": 0.6099,
+      "step": 1957
+    },
+    {
+      "epoch": 0.7737219066436157,
+      "grad_norm": 0.4823674667046708,
+      "learning_rate": 4.989138108086793e-06,
+      "loss": 0.6057,
+      "step": 1958
+    },
+    {
+      "epoch": 0.774117065942208,
+      "grad_norm": 0.488893702740108,
+      "learning_rate": 4.989123514462748e-06,
+      "loss": 0.6215,
+      "step": 1959
+    },
+    {
+      "epoch": 0.7745122252408002,
+      "grad_norm": 0.5065060131368263,
+      "learning_rate": 4.989108911062938e-06,
+      "loss": 0.6268,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7749073845393925,
+      "grad_norm": 0.46321552663150034,
+      "learning_rate": 4.989094297887419e-06,
+      "loss": 0.622,
+      "step": 1961
+    },
+    {
+      "epoch": 0.7753025438379847,
+      "grad_norm": 0.5020834149871061,
+      "learning_rate": 4.989079674936249e-06,
+      "loss": 0.6185,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7756977031365769,
+      "grad_norm": 0.4981082643061992,
+      "learning_rate": 4.989065042209486e-06,
+      "loss": 0.6292,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7760928624351692,
+      "grad_norm": 0.47491786674487524,
+      "learning_rate": 4.989050399707186e-06,
+      "loss": 0.6186,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7764880217337614,
+      "grad_norm": 0.47555277744404306,
+      "learning_rate": 4.989035747429409e-06,
+      "loss": 0.6205,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7768831810323537,
+      "grad_norm": 0.5232964357542301,
+      "learning_rate": 4.989021085376209e-06,
+      "loss": 0.6325,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7772783403309459,
+      "grad_norm": 0.48296952218572786,
+      "learning_rate": 4.989006413547647e-06,
+      "loss": 0.6232,
+      "step": 1967
+    },
+    {
+      "epoch": 0.7776734996295381,
+      "grad_norm": 0.49104169294101535,
+      "learning_rate": 4.988991731943778e-06,
+      "loss": 0.6095,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7780686589281304,
+      "grad_norm": 0.4706225330715028,
+      "learning_rate": 4.988977040564662e-06,
+      "loss": 0.6001,
+      "step": 1969
+    },
+    {
+      "epoch": 0.7784638182267226,
+      "grad_norm": 0.4916599710014126,
+      "learning_rate": 4.988962339410356e-06,
+      "loss": 0.6018,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7788589775253149,
+      "grad_norm": 0.4930270017350692,
+      "learning_rate": 4.988947628480917e-06,
+      "loss": 0.6232,
+      "step": 1971
+    },
+    {
+      "epoch": 0.7792541368239071,
+      "grad_norm": 0.49545246948058913,
+      "learning_rate": 4.988932907776402e-06,
+      "loss": 0.6423,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7796492961224993,
+      "grad_norm": 0.48258013902876534,
+      "learning_rate": 4.988918177296871e-06,
+      "loss": 0.6234,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7800444554210916,
+      "grad_norm": 0.4654851797716931,
+      "learning_rate": 4.988903437042379e-06,
+      "loss": 0.6061,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7804396147196839,
+      "grad_norm": 0.4709686787629398,
+      "learning_rate": 4.988888687012988e-06,
+      "loss": 0.6047,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7808347740182762,
+      "grad_norm": 0.480518188036582,
+      "learning_rate": 4.988873927208753e-06,
+      "loss": 0.6272,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7812299333168684,
+      "grad_norm": 0.4830740477285781,
+      "learning_rate": 4.9888591576297315e-06,
+      "loss": 0.6199,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7816250926154606,
+      "grad_norm": 0.49273772037327457,
+      "learning_rate": 4.988844378275983e-06,
+      "loss": 0.6278,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7820202519140529,
+      "grad_norm": 0.5246304983064629,
+      "learning_rate": 4.988829589147566e-06,
+      "loss": 0.642,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7824154112126451,
+      "grad_norm": 0.5572740103595891,
+      "learning_rate": 4.988814790244536e-06,
+      "loss": 0.6414,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7828105705112374,
+      "grad_norm": 0.5029929788394535,
+      "learning_rate": 4.988799981566954e-06,
+      "loss": 0.6119,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7832057298098296,
+      "grad_norm": 0.5256809229609704,
+      "learning_rate": 4.988785163114876e-06,
+      "loss": 0.6411,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7836008891084218,
+      "grad_norm": 0.46688383401448613,
+      "learning_rate": 4.988770334888362e-06,
+      "loss": 0.6118,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7839960484070141,
+      "grad_norm": 0.4953447635940505,
+      "learning_rate": 4.988755496887469e-06,
+      "loss": 0.616,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7843912077056063,
+      "grad_norm": 0.5398841768039435,
+      "learning_rate": 4.988740649112256e-06,
+      "loss": 0.6177,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7847863670041986,
+      "grad_norm": 0.5019947174695433,
+      "learning_rate": 4.988725791562782e-06,
+      "loss": 0.6296,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7851815263027908,
+      "grad_norm": 0.49373501939581677,
+      "learning_rate": 4.988710924239103e-06,
+      "loss": 0.6365,
+      "step": 1987
+    },
+    {
+      "epoch": 0.785576685601383,
+      "grad_norm": 0.5151169309400688,
+      "learning_rate": 4.988696047141278e-06,
+      "loss": 0.5958,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7859718448999753,
+      "grad_norm": 0.4864586261342781,
+      "learning_rate": 4.988681160269367e-06,
+      "loss": 0.6389,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7863670041985675,
+      "grad_norm": 0.47700282379110287,
+      "learning_rate": 4.988666263623428e-06,
+      "loss": 0.6333,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7867621634971598,
+      "grad_norm": 0.5030597816770237,
+      "learning_rate": 4.988651357203519e-06,
+      "loss": 0.6201,
+      "step": 1991
+    },
+    {
+      "epoch": 0.787157322795752,
+      "grad_norm": 0.4781975328115511,
+      "learning_rate": 4.988636441009698e-06,
+      "loss": 0.6175,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7875524820943443,
+      "grad_norm": 0.5174944437050455,
+      "learning_rate": 4.988621515042025e-06,
+      "loss": 0.6462,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7879476413929365,
+      "grad_norm": 0.5374787592098473,
+      "learning_rate": 4.988606579300557e-06,
+      "loss": 0.634,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7883428006915287,
+      "grad_norm": 0.47432495552537524,
+      "learning_rate": 4.988591633785354e-06,
+      "loss": 0.6332,
+      "step": 1995
+    },
+    {
+      "epoch": 0.788737959990121,
+      "grad_norm": 0.48163792175353165,
+      "learning_rate": 4.988576678496474e-06,
+      "loss": 0.6038,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7891331192887132,
+      "grad_norm": 0.46928084110206164,
+      "learning_rate": 4.988561713433977e-06,
+      "loss": 0.6053,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7895282785873055,
+      "grad_norm": 0.5029560676742602,
+      "learning_rate": 4.988546738597919e-06,
+      "loss": 0.6366,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7899234378858978,
+      "grad_norm": 0.4850619096705073,
+      "learning_rate": 4.988531753988361e-06,
+      "loss": 0.6231,
+      "step": 1999
+    },
+    {
+      "epoch": 0.79031859718449,
+      "grad_norm": 0.4946994316101753,
+      "learning_rate": 4.988516759605363e-06,
+      "loss": 0.6331,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7907137564830823,
+      "grad_norm": 0.5367867666042327,
+      "learning_rate": 4.988501755448981e-06,
+      "loss": 0.639,
+      "step": 2001
+    },
+    {
+      "epoch": 0.7911089157816745,
+      "grad_norm": 0.5010435528360841,
+      "learning_rate": 4.988486741519275e-06,
+      "loss": 0.62,
+      "step": 2002
+    },
+    {
+      "epoch": 0.7915040750802668,
+      "grad_norm": 0.4847944569719835,
+      "learning_rate": 4.988471717816305e-06,
+      "loss": 0.6629,
+      "step": 2003
+    },
+    {
+      "epoch": 0.791899234378859,
+      "grad_norm": 0.7692429223559273,
+      "learning_rate": 4.988456684340128e-06,
+      "loss": 0.6407,
+      "step": 2004
+    },
+    {
+      "epoch": 0.7922943936774512,
+      "grad_norm": 0.5018665724568167,
+      "learning_rate": 4.9884416410908055e-06,
+      "loss": 0.6318,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7926895529760435,
+      "grad_norm": 0.47570736353057425,
+      "learning_rate": 4.988426588068394e-06,
+      "loss": 0.6203,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7930847122746357,
+      "grad_norm": 0.49443151633573906,
+      "learning_rate": 4.988411525272954e-06,
+      "loss": 0.6314,
+      "step": 2007
+    },
+    {
+      "epoch": 0.793479871573228,
+      "grad_norm": 0.47649470527071924,
+      "learning_rate": 4.988396452704546e-06,
+      "loss": 0.6122,
+      "step": 2008
+    },
+    {
+      "epoch": 0.7938750308718202,
+      "grad_norm": 0.4959125816855865,
+      "learning_rate": 4.988381370363227e-06,
+      "loss": 0.6264,
+      "step": 2009
+    },
+    {
+      "epoch": 0.7942701901704124,
+      "grad_norm": 0.4670880342830564,
+      "learning_rate": 4.9883662782490576e-06,
+      "loss": 0.5956,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7946653494690047,
+      "grad_norm": 0.5050700439632813,
+      "learning_rate": 4.988351176362095e-06,
+      "loss": 0.6234,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7950605087675969,
+      "grad_norm": 0.48271559918802115,
+      "learning_rate": 4.9883360647024e-06,
+      "loss": 0.6295,
+      "step": 2012
+    },
+    {
+      "epoch": 0.7954556680661892,
+      "grad_norm": 0.4636363520372164,
+      "learning_rate": 4.988320943270034e-06,
+      "loss": 0.6177,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7958508273647814,
+      "grad_norm": 0.6019915166951819,
+      "learning_rate": 4.988305812065053e-06,
+      "loss": 0.6307,
+      "step": 2014
+    },
+    {
+      "epoch": 0.7962459866633737,
+      "grad_norm": 0.5005076532485925,
+      "learning_rate": 4.988290671087517e-06,
+      "loss": 0.6331,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7966411459619659,
+      "grad_norm": 0.5057151799248627,
+      "learning_rate": 4.988275520337488e-06,
+      "loss": 0.6397,
+      "step": 2016
+    },
+    {
+      "epoch": 0.7970363052605581,
+      "grad_norm": 0.4620738413199476,
+      "learning_rate": 4.988260359815022e-06,
+      "loss": 0.6243,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7974314645591504,
+      "grad_norm": 0.48041825341679295,
+      "learning_rate": 4.988245189520181e-06,
+      "loss": 0.6347,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7978266238577426,
+      "grad_norm": 0.4854380738784906,
+      "learning_rate": 4.9882300094530236e-06,
+      "loss": 0.6321,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7982217831563349,
+      "grad_norm": 0.4732758526185407,
+      "learning_rate": 4.988214819613611e-06,
+      "loss": 0.613,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7986169424549271,
+      "grad_norm": 0.4902286255287152,
+      "learning_rate": 4.988199620002e-06,
+      "loss": 0.642,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7990121017535193,
+      "grad_norm": 0.48018089371628836,
+      "learning_rate": 4.988184410618252e-06,
+      "loss": 0.6281,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7994072610521117,
+      "grad_norm": 0.48276881530797655,
+      "learning_rate": 4.988169191462426e-06,
+      "loss": 0.624,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7998024203507039,
+      "grad_norm": 0.4805829544314743,
+      "learning_rate": 4.988153962534583e-06,
+      "loss": 0.6355,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8001975796492962,
+      "grad_norm": 0.4650087756290116,
+      "learning_rate": 4.988138723834783e-06,
+      "loss": 0.6159,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8005927389478884,
+      "grad_norm": 0.500302525183259,
+      "learning_rate": 4.9881234753630835e-06,
+      "loss": 0.6207,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8009878982464806,
+      "grad_norm": 0.4633513547631363,
+      "learning_rate": 4.988108217119547e-06,
+      "loss": 0.603,
+      "step": 2027
+    },
+    {
+      "epoch": 0.8013830575450729,
+      "grad_norm": 0.4566681685008273,
+      "learning_rate": 4.988092949104232e-06,
+      "loss": 0.6074,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8017782168436651,
+      "grad_norm": 0.4988424175671541,
+      "learning_rate": 4.988077671317198e-06,
+      "loss": 0.6013,
+      "step": 2029
+    },
+    {
+      "epoch": 0.8021733761422574,
+      "grad_norm": 0.4929308612546276,
+      "learning_rate": 4.988062383758506e-06,
+      "loss": 0.6244,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8025685354408496,
+      "grad_norm": 0.4781963632718603,
+      "learning_rate": 4.988047086428217e-06,
+      "loss": 0.6197,
+      "step": 2031
+    },
+    {
+      "epoch": 0.8029636947394418,
+      "grad_norm": 0.6162145939572514,
+      "learning_rate": 4.988031779326389e-06,
+      "loss": 0.6274,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8033588540380341,
+      "grad_norm": 0.48495525673379924,
+      "learning_rate": 4.988016462453082e-06,
+      "loss": 0.6293,
+      "step": 2033
+    },
+    {
+      "epoch": 0.8037540133366263,
+      "grad_norm": 0.5043930111987597,
+      "learning_rate": 4.988001135808358e-06,
+      "loss": 0.6138,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8041491726352186,
+      "grad_norm": 0.46666576704300994,
+      "learning_rate": 4.987985799392277e-06,
+      "loss": 0.6072,
+      "step": 2035
+    },
+    {
+      "epoch": 0.8045443319338108,
+      "grad_norm": 0.48805370345241733,
+      "learning_rate": 4.987970453204898e-06,
+      "loss": 0.6249,
+      "step": 2036
+    },
+    {
+      "epoch": 0.804939491232403,
+      "grad_norm": 0.48377604545911973,
+      "learning_rate": 4.987955097246282e-06,
+      "loss": 0.617,
+      "step": 2037
+    },
+    {
+      "epoch": 0.8053346505309953,
+      "grad_norm": 0.4720053686199778,
+      "learning_rate": 4.987939731516489e-06,
+      "loss": 0.6046,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8057298098295875,
+      "grad_norm": 0.5828932827171168,
+      "learning_rate": 4.987924356015579e-06,
+      "loss": 0.6315,
+      "step": 2039
+    },
+    {
+      "epoch": 0.8061249691281798,
+      "grad_norm": 0.4891162177690198,
+      "learning_rate": 4.987908970743614e-06,
+      "loss": 0.6332,
+      "step": 2040
+    },
+    {
+      "epoch": 0.806520128426772,
+      "grad_norm": 0.4827837446920491,
+      "learning_rate": 4.987893575700652e-06,
+      "loss": 0.6258,
+      "step": 2041
+    },
+    {
+      "epoch": 0.8069152877253642,
+      "grad_norm": 0.46077989778395895,
+      "learning_rate": 4.987878170886755e-06,
+      "loss": 0.5977,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8073104470239565,
+      "grad_norm": 0.48314109012199535,
+      "learning_rate": 4.987862756301984e-06,
+      "loss": 0.6335,
+      "step": 2043
+    },
+    {
+      "epoch": 0.8077056063225487,
+      "grad_norm": 0.47542653192334977,
+      "learning_rate": 4.987847331946398e-06,
+      "loss": 0.6009,
+      "step": 2044
+    },
+    {
+      "epoch": 0.808100765621141,
+      "grad_norm": 0.46969812561265184,
+      "learning_rate": 4.987831897820059e-06,
+      "loss": 0.6262,
+      "step": 2045
+    },
+    {
+      "epoch": 0.8084959249197333,
+      "grad_norm": 0.4697880436971334,
+      "learning_rate": 4.987816453923027e-06,
+      "loss": 0.6212,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8088910842183256,
+      "grad_norm": 0.5290819232360799,
+      "learning_rate": 4.987801000255362e-06,
+      "loss": 0.6359,
+      "step": 2047
+    },
+    {
+      "epoch": 0.8092862435169178,
+      "grad_norm": 0.4774269454534689,
+      "learning_rate": 4.987785536817127e-06,
+      "loss": 0.6069,
+      "step": 2048
+    },
+    {
+      "epoch": 0.80968140281551,
+      "grad_norm": 0.4925043786131469,
+      "learning_rate": 4.987770063608379e-06,
+      "loss": 0.6237,
+      "step": 2049
+    },
+    {
+      "epoch": 0.8100765621141023,
+      "grad_norm": 0.48913030453903555,
+      "learning_rate": 4.987754580629182e-06,
+      "loss": 0.617,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8104717214126945,
+      "grad_norm": 0.4870701601969778,
+      "learning_rate": 4.987739087879596e-06,
+      "loss": 0.6152,
+      "step": 2051
+    },
+    {
+      "epoch": 0.8108668807112868,
+      "grad_norm": 0.4732126768516727,
+      "learning_rate": 4.987723585359681e-06,
+      "loss": 0.619,
+      "step": 2052
+    },
+    {
+      "epoch": 0.811262040009879,
+      "grad_norm": 0.4889602356350729,
+      "learning_rate": 4.987708073069498e-06,
+      "loss": 0.6352,
+      "step": 2053
+    },
+    {
+      "epoch": 0.8116571993084712,
+      "grad_norm": 0.48795631546959883,
+      "learning_rate": 4.9876925510091085e-06,
+      "loss": 0.6143,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8120523586070635,
+      "grad_norm": 0.46539271557990114,
+      "learning_rate": 4.987677019178573e-06,
+      "loss": 0.6046,
+      "step": 2055
+    },
+    {
+      "epoch": 0.8124475179056557,
+      "grad_norm": 0.49405822426322776,
+      "learning_rate": 4.987661477577953e-06,
+      "loss": 0.6014,
+      "step": 2056
+    },
+    {
+      "epoch": 0.812842677204248,
+      "grad_norm": 0.48386934068049814,
+      "learning_rate": 4.98764592620731e-06,
+      "loss": 0.6001,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8132378365028402,
+      "grad_norm": 0.5187696026651084,
+      "learning_rate": 4.987630365066703e-06,
+      "loss": 0.6317,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8136329958014324,
+      "grad_norm": 0.5042378749814072,
+      "learning_rate": 4.987614794156196e-06,
+      "loss": 0.6258,
+      "step": 2059
+    },
+    {
+      "epoch": 0.8140281551000247,
+      "grad_norm": 0.4644958030378567,
+      "learning_rate": 4.987599213475848e-06,
+      "loss": 0.6111,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8144233143986169,
+      "grad_norm": 0.45523155834489054,
+      "learning_rate": 4.98758362302572e-06,
+      "loss": 0.614,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8148184736972092,
+      "grad_norm": 0.4798484565189588,
+      "learning_rate": 4.987568022805875e-06,
+      "loss": 0.6285,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8152136329958014,
+      "grad_norm": 0.48279313141569785,
+      "learning_rate": 4.987552412816373e-06,
+      "loss": 0.6223,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8156087922943936,
+      "grad_norm": 0.452316831060907,
+      "learning_rate": 4.9875367930572764e-06,
+      "loss": 0.5993,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8160039515929859,
+      "grad_norm": 0.4964123620781851,
+      "learning_rate": 4.987521163528645e-06,
+      "loss": 0.6159,
+      "step": 2065
+    },
+    {
+      "epoch": 0.8163991108915781,
+      "grad_norm": 0.487979292436224,
+      "learning_rate": 4.9875055242305414e-06,
+      "loss": 0.6027,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8167942701901704,
+      "grad_norm": 0.46013675854663044,
+      "learning_rate": 4.987489875163027e-06,
+      "loss": 0.6076,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8171894294887626,
+      "grad_norm": 0.4929449077318516,
+      "learning_rate": 4.987474216326162e-06,
+      "loss": 0.6387,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8175845887873548,
+      "grad_norm": 0.4845064197629273,
+      "learning_rate": 4.987458547720009e-06,
+      "loss": 0.624,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8179797480859472,
+      "grad_norm": 0.5780674060390795,
+      "learning_rate": 4.987442869344629e-06,
+      "loss": 0.6473,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8183749073845394,
+      "grad_norm": 0.5093404649783988,
+      "learning_rate": 4.987427181200084e-06,
+      "loss": 0.6399,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8187700666831317,
+      "grad_norm": 0.5027970250091759,
+      "learning_rate": 4.987411483286436e-06,
+      "loss": 0.6364,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8191652259817239,
+      "grad_norm": 0.4681596123762207,
+      "learning_rate": 4.987395775603746e-06,
+      "loss": 0.5992,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8195603852803162,
+      "grad_norm": 0.4774118024924558,
+      "learning_rate": 4.987380058152076e-06,
+      "loss": 0.6076,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8199555445789084,
+      "grad_norm": 0.4758447061485989,
+      "learning_rate": 4.987364330931487e-06,
+      "loss": 0.6348,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8203507038775006,
+      "grad_norm": 0.4820431900570343,
+      "learning_rate": 4.9873485939420405e-06,
+      "loss": 0.6126,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8207458631760929,
+      "grad_norm": 0.46836863605198487,
+      "learning_rate": 4.987332847183801e-06,
+      "loss": 0.6143,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8211410224746851,
+      "grad_norm": 0.47555062710148005,
+      "learning_rate": 4.987317090656827e-06,
+      "loss": 0.6142,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8215361817732774,
+      "grad_norm": 0.4870036833640785,
+      "learning_rate": 4.987301324361182e-06,
+      "loss": 0.6083,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8219313410718696,
+      "grad_norm": 0.4844543461263999,
+      "learning_rate": 4.9872855482969284e-06,
+      "loss": 0.6168,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8223265003704618,
+      "grad_norm": 0.4746621256551825,
+      "learning_rate": 4.987269762464127e-06,
+      "loss": 0.62,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8227216596690541,
+      "grad_norm": 0.4718461178840169,
+      "learning_rate": 4.987253966862841e-06,
+      "loss": 0.626,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8231168189676463,
+      "grad_norm": 0.47940330337486675,
+      "learning_rate": 4.987238161493132e-06,
+      "loss": 0.6384,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8235119782662386,
+      "grad_norm": 0.4831189596082736,
+      "learning_rate": 4.987222346355061e-06,
+      "loss": 0.6006,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8239071375648308,
+      "grad_norm": 0.5111828704378958,
+      "learning_rate": 4.987206521448691e-06,
+      "loss": 0.6275,
+      "step": 2085
+    },
+    {
+      "epoch": 0.824302296863423,
+      "grad_norm": 0.4761799453175209,
+      "learning_rate": 4.987190686774084e-06,
+      "loss": 0.6166,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8246974561620153,
+      "grad_norm": 0.479237633298224,
+      "learning_rate": 4.987174842331303e-06,
+      "loss": 0.6157,
+      "step": 2087
+    },
+    {
+      "epoch": 0.8250926154606075,
+      "grad_norm": 0.4877816511875846,
+      "learning_rate": 4.9871589881204085e-06,
+      "loss": 0.633,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8254877747591998,
+      "grad_norm": 0.5597118422036577,
+      "learning_rate": 4.987143124141465e-06,
+      "loss": 0.6281,
+      "step": 2089
+    },
+    {
+      "epoch": 0.825882934057792,
+      "grad_norm": 0.4808185647219305,
+      "learning_rate": 4.987127250394532e-06,
+      "loss": 0.6284,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8262780933563842,
+      "grad_norm": 0.49350572331519976,
+      "learning_rate": 4.987111366879674e-06,
+      "loss": 0.6327,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8266732526549765,
+      "grad_norm": 0.4950625588459782,
+      "learning_rate": 4.987095473596954e-06,
+      "loss": 0.5991,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8270684119535687,
+      "grad_norm": 0.48717004314629225,
+      "learning_rate": 4.987079570546432e-06,
+      "loss": 0.6125,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8274635712521611,
+      "grad_norm": 0.477632541157664,
+      "learning_rate": 4.987063657728172e-06,
+      "loss": 0.6184,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8278587305507533,
+      "grad_norm": 0.490920143190879,
+      "learning_rate": 4.987047735142236e-06,
+      "loss": 0.6112,
+      "step": 2095
+    },
+    {
+      "epoch": 0.8282538898493456,
+      "grad_norm": 0.4963440592377515,
+      "learning_rate": 4.9870318027886874e-06,
+      "loss": 0.6122,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8286490491479378,
+      "grad_norm": 0.4867870903292657,
+      "learning_rate": 4.9870158606675875e-06,
+      "loss": 0.5993,
+      "step": 2097
+    },
+    {
+      "epoch": 0.82904420844653,
+      "grad_norm": 0.4976198098833027,
+      "learning_rate": 4.986999908779e-06,
+      "loss": 0.6465,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8294393677451223,
+      "grad_norm": 0.4870232342621309,
+      "learning_rate": 4.986983947122986e-06,
+      "loss": 0.6201,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8298345270437145,
+      "grad_norm": 0.46830678342177057,
+      "learning_rate": 4.9869679756996105e-06,
+      "loss": 0.6107,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8302296863423068,
+      "grad_norm": 0.6549407593688914,
+      "learning_rate": 4.986951994508934e-06,
+      "loss": 0.6266,
+      "step": 2101
+    },
+    {
+      "epoch": 0.830624845640899,
+      "grad_norm": 0.47125597170756056,
+      "learning_rate": 4.98693600355102e-06,
+      "loss": 0.6293,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8310200049394912,
+      "grad_norm": 0.491814138267252,
+      "learning_rate": 4.9869200028259325e-06,
+      "loss": 0.6273,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8314151642380835,
+      "grad_norm": 0.4992441103556944,
+      "learning_rate": 4.986903992333734e-06,
+      "loss": 0.6485,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8318103235366757,
+      "grad_norm": 0.4838463362223243,
+      "learning_rate": 4.986887972074485e-06,
+      "loss": 0.6233,
+      "step": 2105
+    },
+    {
+      "epoch": 0.832205482835268,
+      "grad_norm": 0.5162232947613259,
+      "learning_rate": 4.986871942048252e-06,
+      "loss": 0.5937,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8326006421338602,
+      "grad_norm": 0.491642550409257,
+      "learning_rate": 4.986855902255094e-06,
+      "loss": 0.6275,
+      "step": 2107
+    },
+    {
+      "epoch": 0.8329958014324524,
+      "grad_norm": 0.4697163915777976,
+      "learning_rate": 4.9868398526950765e-06,
+      "loss": 0.6012,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8333909607310447,
+      "grad_norm": 0.5029059751532935,
+      "learning_rate": 4.986823793368263e-06,
+      "loss": 0.6184,
+      "step": 2109
+    },
+    {
+      "epoch": 0.8337861200296369,
+      "grad_norm": 0.49890701973097873,
+      "learning_rate": 4.9868077242747156e-06,
+      "loss": 0.6489,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8341812793282292,
+      "grad_norm": 0.47086406769377864,
+      "learning_rate": 4.986791645414498e-06,
+      "loss": 0.6158,
+      "step": 2111
+    },
+    {
+      "epoch": 0.8345764386268214,
+      "grad_norm": 0.4711411719674825,
+      "learning_rate": 4.986775556787672e-06,
+      "loss": 0.6108,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8349715979254136,
+      "grad_norm": 0.5170133429016198,
+      "learning_rate": 4.986759458394302e-06,
+      "loss": 0.6205,
+      "step": 2113
+    },
+    {
+      "epoch": 0.8353667572240059,
+      "grad_norm": 0.48113051440505156,
+      "learning_rate": 4.986743350234451e-06,
+      "loss": 0.5975,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8357619165225981,
+      "grad_norm": 0.4867487446570137,
+      "learning_rate": 4.986727232308182e-06,
+      "loss": 0.6193,
+      "step": 2115
+    },
+    {
+      "epoch": 0.8361570758211904,
+      "grad_norm": 0.5120874526734966,
+      "learning_rate": 4.986711104615558e-06,
+      "loss": 0.6233,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8365522351197827,
+      "grad_norm": 0.5108538780702159,
+      "learning_rate": 4.986694967156644e-06,
+      "loss": 0.6112,
+      "step": 2117
+    },
+    {
+      "epoch": 0.836947394418375,
+      "grad_norm": 0.5009432477935026,
+      "learning_rate": 4.986678819931501e-06,
+      "loss": 0.6129,
+      "step": 2118
+    },
+    {
+      "epoch": 0.8373425537169672,
+      "grad_norm": 0.5394279204944199,
+      "learning_rate": 4.986662662940193e-06,
+      "loss": 0.6328,
+      "step": 2119
+    },
+    {
+      "epoch": 0.8377377130155594,
+      "grad_norm": 0.5074352200988693,
+      "learning_rate": 4.986646496182786e-06,
+      "loss": 0.6297,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8381328723141517,
+      "grad_norm": 0.4623558465170015,
+      "learning_rate": 4.98663031965934e-06,
+      "loss": 0.5988,
+      "step": 2121
+    },
+    {
+      "epoch": 0.8385280316127439,
+      "grad_norm": 0.5048081966048025,
+      "learning_rate": 4.9866141333699215e-06,
+      "loss": 0.6172,
+      "step": 2122
+    },
+    {
+      "epoch": 0.8389231909113362,
+      "grad_norm": 0.4969733235844862,
+      "learning_rate": 4.986597937314591e-06,
+      "loss": 0.6126,
+      "step": 2123
+    },
+    {
+      "epoch": 0.8393183502099284,
+      "grad_norm": 0.4644808410529684,
+      "learning_rate": 4.986581731493415e-06,
+      "loss": 0.6122,
+      "step": 2124
+    },
+    {
+      "epoch": 0.8397135095085206,
+      "grad_norm": 0.49621141240128863,
+      "learning_rate": 4.986565515906455e-06,
+      "loss": 0.627,
+      "step": 2125
+    },
+    {
+      "epoch": 0.8401086688071129,
+      "grad_norm": 0.5140052050923385,
+      "learning_rate": 4.986549290553777e-06,
+      "loss": 0.6076,
+      "step": 2126
+    },
+    {
+      "epoch": 0.8405038281057051,
+      "grad_norm": 0.4774110621495217,
+      "learning_rate": 4.986533055435442e-06,
+      "loss": 0.638,
+      "step": 2127
+    },
+    {
+      "epoch": 0.8408989874042974,
+      "grad_norm": 0.4831024360099805,
+      "learning_rate": 4.986516810551515e-06,
+      "loss": 0.5982,
+      "step": 2128
+    },
+    {
+      "epoch": 0.8412941467028896,
+      "grad_norm": 0.489641326798273,
+      "learning_rate": 4.9865005559020605e-06,
+      "loss": 0.6139,
+      "step": 2129
+    },
+    {
+      "epoch": 0.8416893060014818,
+      "grad_norm": 0.4700688822738762,
+      "learning_rate": 4.986484291487142e-06,
+      "loss": 0.6094,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8420844653000741,
+      "grad_norm": 0.5498327067539832,
+      "learning_rate": 4.9864680173068215e-06,
+      "loss": 0.6159,
+      "step": 2131
+    },
+    {
+      "epoch": 0.8424796245986663,
+      "grad_norm": 0.5209813131829413,
+      "learning_rate": 4.986451733361165e-06,
+      "loss": 0.6058,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8428747838972586,
+      "grad_norm": 0.4755760046484086,
+      "learning_rate": 4.986435439650236e-06,
+      "loss": 0.6139,
+      "step": 2133
+    },
+    {
+      "epoch": 0.8432699431958508,
+      "grad_norm": 0.4980942846797616,
+      "learning_rate": 4.9864191361741e-06,
+      "loss": 0.6253,
+      "step": 2134
+    },
+    {
+      "epoch": 0.843665102494443,
+      "grad_norm": 0.5023114342986108,
+      "learning_rate": 4.986402822932818e-06,
+      "loss": 0.6174,
+      "step": 2135
+    },
+    {
+      "epoch": 0.8440602617930353,
+      "grad_norm": 0.487624480191535,
+      "learning_rate": 4.986386499926456e-06,
+      "loss": 0.6156,
+      "step": 2136
+    },
+    {
+      "epoch": 0.8444554210916275,
+      "grad_norm": 0.47126111959332073,
+      "learning_rate": 4.986370167155078e-06,
+      "loss": 0.604,
+      "step": 2137
+    },
+    {
+      "epoch": 0.8448505803902198,
+      "grad_norm": 0.4859939062367633,
+      "learning_rate": 4.986353824618747e-06,
+      "loss": 0.6085,
+      "step": 2138
+    },
+    {
+      "epoch": 0.845245739688812,
+      "grad_norm": 0.4865870625029928,
+      "learning_rate": 4.9863374723175285e-06,
+      "loss": 0.6155,
+      "step": 2139
+    },
+    {
+      "epoch": 0.8456408989874042,
+      "grad_norm": 0.5154805544761918,
+      "learning_rate": 4.9863211102514855e-06,
+      "loss": 0.6092,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8460360582859966,
+      "grad_norm": 0.479204794222624,
+      "learning_rate": 4.986304738420684e-06,
+      "loss": 0.6202,
+      "step": 2141
+    },
+    {
+      "epoch": 0.8464312175845888,
+      "grad_norm": 0.4854511339075718,
+      "learning_rate": 4.986288356825186e-06,
+      "loss": 0.5973,
+      "step": 2142
+    },
+    {
+      "epoch": 0.8468263768831811,
+      "grad_norm": 0.4947439475538887,
+      "learning_rate": 4.986271965465058e-06,
+      "loss": 0.606,
+      "step": 2143
+    },
+    {
+      "epoch": 0.8472215361817733,
+      "grad_norm": 0.4772366979723403,
+      "learning_rate": 4.9862555643403634e-06,
+      "loss": 0.6412,
+      "step": 2144
+    },
+    {
+      "epoch": 0.8476166954803656,
+      "grad_norm": 0.45202495018365624,
+      "learning_rate": 4.986239153451167e-06,
+      "loss": 0.6009,
+      "step": 2145
+    },
+    {
+      "epoch": 0.8480118547789578,
+      "grad_norm": 0.4674899554356609,
+      "learning_rate": 4.986222732797532e-06,
+      "loss": 0.6083,
+      "step": 2146
+    },
+    {
+      "epoch": 0.84840701407755,
+      "grad_norm": 0.4720229690067215,
+      "learning_rate": 4.986206302379524e-06,
+      "loss": 0.6193,
+      "step": 2147
+    },
+    {
+      "epoch": 0.8488021733761423,
+      "grad_norm": 0.5068197579171579,
+      "learning_rate": 4.986189862197208e-06,
+      "loss": 0.6188,
+      "step": 2148
+    },
+    {
+      "epoch": 0.8491973326747345,
+      "grad_norm": 0.475281381106489,
+      "learning_rate": 4.9861734122506475e-06,
+      "loss": 0.6115,
+      "step": 2149
+    },
+    {
+      "epoch": 0.8495924919733268,
+      "grad_norm": 0.46986989066929974,
+      "learning_rate": 4.986156952539908e-06,
+      "loss": 0.6269,
+      "step": 2150
+    },
+    {
+      "epoch": 0.849987651271919,
+      "grad_norm": 0.46811282834432916,
+      "learning_rate": 4.986140483065053e-06,
+      "loss": 0.6215,
+      "step": 2151
+    },
+    {
+      "epoch": 0.8503828105705112,
+      "grad_norm": 0.4910081036842562,
+      "learning_rate": 4.986124003826148e-06,
+      "loss": 0.6181,
+      "step": 2152
+    },
+    {
+      "epoch": 0.8507779698691035,
+      "grad_norm": 0.4872472385487101,
+      "learning_rate": 4.986107514823257e-06,
+      "loss": 0.6192,
+      "step": 2153
+    },
+    {
+      "epoch": 0.8511731291676957,
+      "grad_norm": 0.47883592093053695,
+      "learning_rate": 4.986091016056446e-06,
+      "loss": 0.6201,
+      "step": 2154
+    },
+    {
+      "epoch": 0.851568288466288,
+      "grad_norm": 0.485056443392643,
+      "learning_rate": 4.986074507525779e-06,
+      "loss": 0.6156,
+      "step": 2155
+    },
+    {
+      "epoch": 0.8519634477648802,
+      "grad_norm": 0.4833961803335379,
+      "learning_rate": 4.986057989231321e-06,
+      "loss": 0.5979,
+      "step": 2156
+    },
+    {
+      "epoch": 0.8523586070634724,
+      "grad_norm": 0.48491015695364686,
+      "learning_rate": 4.9860414611731375e-06,
+      "loss": 0.6379,
+      "step": 2157
+    },
+    {
+      "epoch": 0.8527537663620647,
+      "grad_norm": 0.47943852412179266,
+      "learning_rate": 4.986024923351292e-06,
+      "loss": 0.6266,
+      "step": 2158
+    },
+    {
+      "epoch": 0.8531489256606569,
+      "grad_norm": 0.4899269774275039,
+      "learning_rate": 4.9860083757658505e-06,
+      "loss": 0.6266,
+      "step": 2159
+    },
+    {
+      "epoch": 0.8535440849592492,
+      "grad_norm": 0.47032529957003527,
+      "learning_rate": 4.985991818416877e-06,
+      "loss": 0.6491,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8539392442578414,
+      "grad_norm": 0.4902687851540872,
+      "learning_rate": 4.9859752513044375e-06,
+      "loss": 0.6301,
+      "step": 2161
+    },
+    {
+      "epoch": 0.8543344035564336,
+      "grad_norm": 0.48553682325045194,
+      "learning_rate": 4.985958674428597e-06,
+      "loss": 0.6061,
+      "step": 2162
+    },
+    {
+      "epoch": 0.8547295628550259,
+      "grad_norm": 0.48411182285405713,
+      "learning_rate": 4.98594208778942e-06,
+      "loss": 0.6247,
+      "step": 2163
+    },
+    {
+      "epoch": 0.8551247221536181,
+      "grad_norm": 0.47951482859948863,
+      "learning_rate": 4.985925491386973e-06,
+      "loss": 0.6312,
+      "step": 2164
+    },
+    {
+      "epoch": 0.8555198814522105,
+      "grad_norm": 0.4746537184476964,
+      "learning_rate": 4.98590888522132e-06,
+      "loss": 0.6109,
+      "step": 2165
+    },
+    {
+      "epoch": 0.8559150407508027,
+      "grad_norm": 0.4730418292508103,
+      "learning_rate": 4.985892269292526e-06,
+      "loss": 0.6161,
+      "step": 2166
+    },
+    {
+      "epoch": 0.856310200049395,
+      "grad_norm": 0.4904685249853375,
+      "learning_rate": 4.985875643600656e-06,
+      "loss": 0.6025,
+      "step": 2167
+    },
+    {
+      "epoch": 0.8567053593479872,
+      "grad_norm": 0.48706770829729257,
+      "learning_rate": 4.985859008145777e-06,
+      "loss": 0.6083,
+      "step": 2168
+    },
+    {
+      "epoch": 0.8571005186465794,
+      "grad_norm": 0.4789542146071817,
+      "learning_rate": 4.9858423629279525e-06,
+      "loss": 0.6022,
+      "step": 2169
+    },
+    {
+      "epoch": 0.8574956779451717,
+      "grad_norm": 0.4764312440150958,
+      "learning_rate": 4.98582570794725e-06,
+      "loss": 0.6269,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8578908372437639,
+      "grad_norm": 0.4905768390433796,
+      "learning_rate": 4.985809043203732e-06,
+      "loss": 0.6202,
+      "step": 2171
+    },
+    {
+      "epoch": 0.8582859965423562,
+      "grad_norm": 0.48600066838590206,
+      "learning_rate": 4.9857923686974664e-06,
+      "loss": 0.6293,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8586811558409484,
+      "grad_norm": 0.5278954101657808,
+      "learning_rate": 4.985775684428518e-06,
+      "loss": 0.626,
+      "step": 2173
+    },
+    {
+      "epoch": 0.8590763151395406,
+      "grad_norm": 0.4772206937537348,
+      "learning_rate": 4.985758990396952e-06,
+      "loss": 0.6127,
+      "step": 2174
+    },
+    {
+      "epoch": 0.8594714744381329,
+      "grad_norm": 0.5032599137233393,
+      "learning_rate": 4.985742286602834e-06,
+      "loss": 0.6247,
+      "step": 2175
+    },
+    {
+      "epoch": 0.8598666337367251,
+      "grad_norm": 0.4890130823058908,
+      "learning_rate": 4.985725573046229e-06,
+      "loss": 0.6138,
+      "step": 2176
+    },
+    {
+      "epoch": 0.8602617930353174,
+      "grad_norm": 0.5090384794958555,
+      "learning_rate": 4.985708849727205e-06,
+      "loss": 0.6221,
+      "step": 2177
+    },
+    {
+      "epoch": 0.8606569523339096,
+      "grad_norm": 0.5064691369911446,
+      "learning_rate": 4.985692116645825e-06,
+      "loss": 0.6365,
+      "step": 2178
+    },
+    {
+      "epoch": 0.8610521116325018,
+      "grad_norm": 0.4884256917672367,
+      "learning_rate": 4.985675373802155e-06,
+      "loss": 0.6245,
+      "step": 2179
+    },
+    {
+      "epoch": 0.8614472709310941,
+      "grad_norm": 0.48277194568900933,
+      "learning_rate": 4.9856586211962636e-06,
+      "loss": 0.6303,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8618424302296863,
+      "grad_norm": 0.4796258764505248,
+      "learning_rate": 4.985641858828213e-06,
+      "loss": 0.5917,
+      "step": 2181
+    },
+    {
+      "epoch": 0.8622375895282786,
+      "grad_norm": 0.4742026777974947,
+      "learning_rate": 4.985625086698071e-06,
+      "loss": 0.6014,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8626327488268708,
+      "grad_norm": 0.4798722649367731,
+      "learning_rate": 4.9856083048059025e-06,
+      "loss": 0.6058,
+      "step": 2183
+    },
+    {
+      "epoch": 0.863027908125463,
+      "grad_norm": 0.484402428507696,
+      "learning_rate": 4.985591513151775e-06,
+      "loss": 0.6244,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8634230674240553,
+      "grad_norm": 0.4859733762622107,
+      "learning_rate": 4.985574711735752e-06,
+      "loss": 0.6098,
+      "step": 2185
+    },
+    {
+      "epoch": 0.8638182267226475,
+      "grad_norm": 0.4810113395505376,
+      "learning_rate": 4.985557900557902e-06,
+      "loss": 0.6142,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8642133860212398,
+      "grad_norm": 0.504453340452416,
+      "learning_rate": 4.985541079618289e-06,
+      "loss": 0.6289,
+      "step": 2187
+    },
+    {
+      "epoch": 0.8646085453198321,
+      "grad_norm": 0.4855393706146718,
+      "learning_rate": 4.985524248916981e-06,
+      "loss": 0.6215,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8650037046184244,
+      "grad_norm": 0.4816372880869928,
+      "learning_rate": 4.985507408454042e-06,
+      "loss": 0.6214,
+      "step": 2189
+    },
+    {
+      "epoch": 0.8653988639170166,
+      "grad_norm": 0.4971312058231278,
+      "learning_rate": 4.98549055822954e-06,
+      "loss": 0.6184,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8657940232156088,
+      "grad_norm": 0.4737419973037183,
+      "learning_rate": 4.985473698243539e-06,
+      "loss": 0.6362,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8661891825142011,
+      "grad_norm": 0.4819594165338205,
+      "learning_rate": 4.985456828496108e-06,
+      "loss": 0.6265,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8665843418127933,
+      "grad_norm": 0.4713033614828359,
+      "learning_rate": 4.985439948987311e-06,
+      "loss": 0.6062,
+      "step": 2193
+    },
+    {
+      "epoch": 0.8669795011113856,
+      "grad_norm": 0.4731870858016755,
+      "learning_rate": 4.985423059717216e-06,
+      "loss": 0.6222,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8673746604099778,
+      "grad_norm": 0.4835316201063381,
+      "learning_rate": 4.9854061606858875e-06,
+      "loss": 0.636,
+      "step": 2195
+    },
+    {
+      "epoch": 0.86776981970857,
+      "grad_norm": 0.48541889377733993,
+      "learning_rate": 4.985389251893393e-06,
+      "loss": 0.6183,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8681649790071623,
+      "grad_norm": 0.49391047126615967,
+      "learning_rate": 4.985372333339799e-06,
+      "loss": 0.6044,
+      "step": 2197
+    },
+    {
+      "epoch": 0.8685601383057545,
+      "grad_norm": 0.5015005291120314,
+      "learning_rate": 4.985355405025172e-06,
+      "loss": 0.6142,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8689552976043468,
+      "grad_norm": 0.4898017203240075,
+      "learning_rate": 4.985338466949577e-06,
+      "loss": 0.6299,
+      "step": 2199
+    },
+    {
+      "epoch": 0.869350456902939,
+      "grad_norm": 0.49139107555550443,
+      "learning_rate": 4.985321519113083e-06,
+      "loss": 0.5891,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8697456162015312,
+      "grad_norm": 0.4653030042469201,
+      "learning_rate": 4.985304561515754e-06,
+      "loss": 0.6056,
+      "step": 2201
+    },
+    {
+      "epoch": 0.8701407755001235,
+      "grad_norm": 0.48204406635629615,
+      "learning_rate": 4.985287594157659e-06,
+      "loss": 0.6149,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8705359347987157,
+      "grad_norm": 0.4944551279372508,
+      "learning_rate": 4.9852706170388635e-06,
+      "loss": 0.6122,
+      "step": 2203
+    },
+    {
+      "epoch": 0.870931094097308,
+      "grad_norm": 0.4632483556840172,
+      "learning_rate": 4.985253630159434e-06,
+      "loss": 0.6117,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8713262533959002,
+      "grad_norm": 0.4758369257065419,
+      "learning_rate": 4.9852366335194365e-06,
+      "loss": 0.631,
+      "step": 2205
+    },
+    {
+      "epoch": 0.8717214126944924,
+      "grad_norm": 0.4784261261826601,
+      "learning_rate": 4.985219627118939e-06,
+      "loss": 0.6254,
+      "step": 2206
+    },
+    {
+      "epoch": 0.8721165719930847,
+      "grad_norm": 0.46820459867095077,
+      "learning_rate": 4.985202610958008e-06,
+      "loss": 0.6041,
+      "step": 2207
+    },
+    {
+      "epoch": 0.8725117312916769,
+      "grad_norm": 0.6774502372821436,
+      "learning_rate": 4.98518558503671e-06,
+      "loss": 0.6349,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8729068905902692,
+      "grad_norm": 0.477847341760241,
+      "learning_rate": 4.985168549355113e-06,
+      "loss": 0.5999,
+      "step": 2209
+    },
+    {
+      "epoch": 0.8733020498888614,
+      "grad_norm": 0.4709206605453225,
+      "learning_rate": 4.985151503913283e-06,
+      "loss": 0.6219,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8736972091874536,
+      "grad_norm": 0.5073393890746453,
+      "learning_rate": 4.985134448711285e-06,
+      "loss": 0.6131,
+      "step": 2211
+    },
+    {
+      "epoch": 0.874092368486046,
+      "grad_norm": 0.47895189242680536,
+      "learning_rate": 4.98511738374919e-06,
+      "loss": 0.6276,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8744875277846382,
+      "grad_norm": 0.4732689476332835,
+      "learning_rate": 4.985100309027062e-06,
+      "loss": 0.6256,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8748826870832305,
+      "grad_norm": 0.4682497810387466,
+      "learning_rate": 4.985083224544969e-06,
+      "loss": 0.6024,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8752778463818227,
+      "grad_norm": 0.5044101641071467,
+      "learning_rate": 4.985066130302979e-06,
+      "loss": 0.6292,
+      "step": 2215
+    },
+    {
+      "epoch": 0.875673005680415,
+      "grad_norm": 0.5079557931579949,
+      "learning_rate": 4.985049026301158e-06,
+      "loss": 0.6228,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8760681649790072,
+      "grad_norm": 0.4867583650129786,
+      "learning_rate": 4.985031912539572e-06,
+      "loss": 0.6211,
+      "step": 2217
+    },
+    {
+      "epoch": 0.8764633242775994,
+      "grad_norm": 0.48476520801738077,
+      "learning_rate": 4.985014789018291e-06,
+      "loss": 0.5925,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8768584835761917,
+      "grad_norm": 0.49568692882589194,
+      "learning_rate": 4.9849976557373805e-06,
+      "loss": 0.6223,
+      "step": 2219
+    },
+    {
+      "epoch": 0.8772536428747839,
+      "grad_norm": 0.5244025646487809,
+      "learning_rate": 4.984980512696908e-06,
+      "loss": 0.6113,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8776488021733762,
+      "grad_norm": 0.4647233758348524,
+      "learning_rate": 4.984963359896941e-06,
+      "loss": 0.6245,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8780439614719684,
+      "grad_norm": 0.493296948580835,
+      "learning_rate": 4.984946197337548e-06,
+      "loss": 0.6289,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8784391207705606,
+      "grad_norm": 0.48487371283817793,
+      "learning_rate": 4.984929025018794e-06,
+      "loss": 0.618,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8788342800691529,
+      "grad_norm": 0.4628323988563299,
+      "learning_rate": 4.9849118429407486e-06,
+      "loss": 0.6116,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8792294393677451,
+      "grad_norm": 0.4708565386034292,
+      "learning_rate": 4.984894651103478e-06,
+      "loss": 0.6073,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8796245986663374,
+      "grad_norm": 0.47596980477482004,
+      "learning_rate": 4.98487744950705e-06,
+      "loss": 0.6134,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8800197579649296,
+      "grad_norm": 0.48436684059520585,
+      "learning_rate": 4.984860238151533e-06,
+      "loss": 0.6306,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8804149172635218,
+      "grad_norm": 2.901406553646194,
+      "learning_rate": 4.984843017036993e-06,
+      "loss": 0.6343,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8808100765621141,
+      "grad_norm": 0.5052255791290899,
+      "learning_rate": 4.984825786163499e-06,
+      "loss": 0.6131,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8812052358607063,
+      "grad_norm": 0.4790543312612506,
+      "learning_rate": 4.984808545531118e-06,
+      "loss": 0.6215,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8816003951592986,
+      "grad_norm": 0.46921948088144955,
+      "learning_rate": 4.984791295139917e-06,
+      "loss": 0.6004,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8819955544578908,
+      "grad_norm": 0.4774028382073232,
+      "learning_rate": 4.984774034989965e-06,
+      "loss": 0.6351,
+      "step": 2232
+    },
+    {
+      "epoch": 0.882390713756483,
+      "grad_norm": 0.48613593924598036,
+      "learning_rate": 4.98475676508133e-06,
+      "loss": 0.6223,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8827858730550753,
+      "grad_norm": 0.47488141594904604,
+      "learning_rate": 4.9847394854140796e-06,
+      "loss": 0.6361,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8831810323536676,
+      "grad_norm": 0.469545225131135,
+      "learning_rate": 4.984722195988281e-06,
+      "loss": 0.6227,
+      "step": 2235
+    },
+    {
+      "epoch": 0.8835761916522599,
+      "grad_norm": 0.5460234062255062,
+      "learning_rate": 4.984704896804003e-06,
+      "loss": 0.5994,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8839713509508521,
+      "grad_norm": 0.5087695798593977,
+      "learning_rate": 4.984687587861311e-06,
+      "loss": 0.6432,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8843665102494443,
+      "grad_norm": 0.49848654741613146,
+      "learning_rate": 4.984670269160277e-06,
+      "loss": 0.6081,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8847616695480366,
+      "grad_norm": 0.47222113192747045,
+      "learning_rate": 4.984652940700966e-06,
+      "loss": 0.5955,
+      "step": 2239
+    },
+    {
+      "epoch": 0.8851568288466288,
+      "grad_norm": 0.47776469283687745,
+      "learning_rate": 4.984635602483447e-06,
+      "loss": 0.6107,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8855519881452211,
+      "grad_norm": 0.4761875116964049,
+      "learning_rate": 4.984618254507788e-06,
+      "loss": 0.6104,
+      "step": 2241
+    },
+    {
+      "epoch": 0.8859471474438133,
+      "grad_norm": 0.4772606111305406,
+      "learning_rate": 4.984600896774058e-06,
+      "loss": 0.612,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8863423067424056,
+      "grad_norm": 0.4875078890714507,
+      "learning_rate": 4.984583529282323e-06,
+      "loss": 0.6046,
+      "step": 2243
+    },
+    {
+      "epoch": 0.8867374660409978,
+      "grad_norm": 0.5351729625323042,
+      "learning_rate": 4.984566152032654e-06,
+      "loss": 0.6302,
+      "step": 2244
+    },
+    {
+      "epoch": 0.88713262533959,
+      "grad_norm": 0.5134865903542387,
+      "learning_rate": 4.984548765025117e-06,
+      "loss": 0.6398,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8875277846381823,
+      "grad_norm": 0.47943572967975917,
+      "learning_rate": 4.984531368259782e-06,
+      "loss": 0.583,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8879229439367745,
+      "grad_norm": 0.48674628922099783,
+      "learning_rate": 4.984513961736716e-06,
+      "loss": 0.6189,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8883181032353668,
+      "grad_norm": 0.5054075145507174,
+      "learning_rate": 4.984496545455988e-06,
+      "loss": 0.6236,
+      "step": 2248
+    },
+    {
+      "epoch": 0.888713262533959,
+      "grad_norm": 0.4717072903007971,
+      "learning_rate": 4.984479119417666e-06,
+      "loss": 0.6108,
+      "step": 2249
+    },
+    {
+      "epoch": 0.8891084218325512,
+      "grad_norm": 0.4973332505120243,
+      "learning_rate": 4.984461683621818e-06,
+      "loss": 0.6141,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8895035811311435,
+      "grad_norm": 0.5202503935334306,
+      "learning_rate": 4.984444238068515e-06,
+      "loss": 0.611,
+      "step": 2251
+    },
+    {
+      "epoch": 0.8898987404297357,
+      "grad_norm": 0.4868713771823043,
+      "learning_rate": 4.984426782757822e-06,
+      "loss": 0.623,
+      "step": 2252
+    },
+    {
+      "epoch": 0.890293899728328,
+      "grad_norm": 0.4797581634559235,
+      "learning_rate": 4.984409317689809e-06,
+      "loss": 0.6192,
+      "step": 2253
+    },
+    {
+      "epoch": 0.8906890590269202,
+      "grad_norm": 0.49427284931935045,
+      "learning_rate": 4.984391842864546e-06,
+      "loss": 0.6149,
+      "step": 2254
+    },
+    {
+      "epoch": 0.8910842183255124,
+      "grad_norm": 0.4869169065393114,
+      "learning_rate": 4.9843743582821005e-06,
+      "loss": 0.621,
+      "step": 2255
+    },
+    {
+      "epoch": 0.8914793776241047,
+      "grad_norm": 0.4876328379453221,
+      "learning_rate": 4.98435686394254e-06,
+      "loss": 0.5969,
+      "step": 2256
+    },
+    {
+      "epoch": 0.8918745369226969,
+      "grad_norm": 0.4832121246843812,
+      "learning_rate": 4.984339359845935e-06,
+      "loss": 0.6265,
+      "step": 2257
+    },
+    {
+      "epoch": 0.8922696962212892,
+      "grad_norm": 0.566324624003132,
+      "learning_rate": 4.9843218459923535e-06,
+      "loss": 0.6097,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8926648555198815,
+      "grad_norm": 0.49414708906211646,
+      "learning_rate": 4.9843043223818646e-06,
+      "loss": 0.6336,
+      "step": 2259
+    },
+    {
+      "epoch": 0.8930600148184737,
+      "grad_norm": 0.4670689454184053,
+      "learning_rate": 4.984286789014536e-06,
+      "loss": 0.6231,
+      "step": 2260
+    },
+    {
+      "epoch": 0.893455174117066,
+      "grad_norm": 0.47425632523334776,
+      "learning_rate": 4.984269245890438e-06,
+      "loss": 0.6116,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8938503334156582,
+      "grad_norm": 0.481098132153482,
+      "learning_rate": 4.98425169300964e-06,
+      "loss": 0.6339,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8942454927142505,
+      "grad_norm": 0.4881412900652154,
+      "learning_rate": 4.984234130372209e-06,
+      "loss": 0.617,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8946406520128427,
+      "grad_norm": 0.49122021239183733,
+      "learning_rate": 4.984216557978214e-06,
+      "loss": 0.6199,
+      "step": 2264
+    },
+    {
+      "epoch": 0.895035811311435,
+      "grad_norm": 0.4890052599536881,
+      "learning_rate": 4.9841989758277255e-06,
+      "loss": 0.6359,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8954309706100272,
+      "grad_norm": 0.5014418130563002,
+      "learning_rate": 4.984181383920812e-06,
+      "loss": 0.6162,
+      "step": 2266
+    },
+    {
+      "epoch": 0.8958261299086194,
+      "grad_norm": 0.48774048149914556,
+      "learning_rate": 4.984163782257543e-06,
+      "loss": 0.6175,
+      "step": 2267
+    },
+    {
+      "epoch": 0.8962212892072117,
+      "grad_norm": 0.4619145815324718,
+      "learning_rate": 4.9841461708379865e-06,
+      "loss": 0.6129,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8966164485058039,
+      "grad_norm": 0.4706359033079191,
+      "learning_rate": 4.9841285496622124e-06,
+      "loss": 0.6217,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8970116078043961,
+      "grad_norm": 0.5330737136007412,
+      "learning_rate": 4.984110918730289e-06,
+      "loss": 0.6302,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8974067671029884,
+      "grad_norm": 0.4929056846260133,
+      "learning_rate": 4.984093278042288e-06,
+      "loss": 0.6106,
+      "step": 2271
+    },
+    {
+      "epoch": 0.8978019264015806,
+      "grad_norm": 0.4780478045107924,
+      "learning_rate": 4.984075627598276e-06,
+      "loss": 0.6254,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8981970857001729,
+      "grad_norm": 0.48039178372010166,
+      "learning_rate": 4.984057967398324e-06,
+      "loss": 0.613,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8985922449987651,
+      "grad_norm": 0.4902534031318573,
+      "learning_rate": 4.984040297442499e-06,
+      "loss": 0.6196,
+      "step": 2274
+    },
+    {
+      "epoch": 0.8989874042973574,
+      "grad_norm": 0.48251041315749466,
+      "learning_rate": 4.9840226177308745e-06,
+      "loss": 0.635,
+      "step": 2275
+    },
+    {
+      "epoch": 0.8993825635959496,
+      "grad_norm": 0.5107948423795806,
+      "learning_rate": 4.984004928263516e-06,
+      "loss": 0.6371,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8997777228945418,
+      "grad_norm": 0.4849099292899462,
+      "learning_rate": 4.983987229040495e-06,
+      "loss": 0.5988,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9001728821931341,
+      "grad_norm": 0.45033030391921003,
+      "learning_rate": 4.9839695200618804e-06,
+      "loss": 0.5957,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9005680414917263,
+      "grad_norm": 0.48511884291874885,
+      "learning_rate": 4.9839518013277425e-06,
+      "loss": 0.5963,
+      "step": 2279
+    },
+    {
+      "epoch": 0.9009632007903186,
+      "grad_norm": 0.4818554448134295,
+      "learning_rate": 4.983934072838149e-06,
+      "loss": 0.6199,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9013583600889108,
+      "grad_norm": 0.4864050546230534,
+      "learning_rate": 4.983916334593171e-06,
+      "loss": 0.6268,
+      "step": 2281
+    },
+    {
+      "epoch": 0.901753519387503,
+      "grad_norm": 0.48126531375723003,
+      "learning_rate": 4.9838985865928794e-06,
+      "loss": 0.645,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9021486786860954,
+      "grad_norm": 0.47513264097168434,
+      "learning_rate": 4.9838808288373405e-06,
+      "loss": 0.628,
+      "step": 2283
+    },
+    {
+      "epoch": 0.9025438379846876,
+      "grad_norm": 0.4858642717988499,
+      "learning_rate": 4.983863061326627e-06,
+      "loss": 0.6167,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9029389972832799,
+      "grad_norm": 0.5024695807134403,
+      "learning_rate": 4.983845284060808e-06,
+      "loss": 0.6108,
+      "step": 2285
+    },
+    {
+      "epoch": 0.9033341565818721,
+      "grad_norm": 0.47725860181210494,
+      "learning_rate": 4.983827497039953e-06,
+      "loss": 0.5905,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9037293158804643,
+      "grad_norm": 0.4725442142623689,
+      "learning_rate": 4.983809700264131e-06,
+      "loss": 0.6251,
+      "step": 2287
+    },
+    {
+      "epoch": 0.9041244751790566,
+      "grad_norm": 0.4917289285832563,
+      "learning_rate": 4.9837918937334125e-06,
+      "loss": 0.642,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9045196344776488,
+      "grad_norm": 0.49254790176708696,
+      "learning_rate": 4.983774077447869e-06,
+      "loss": 0.6035,
+      "step": 2289
+    },
+    {
+      "epoch": 0.9049147937762411,
+      "grad_norm": 0.47663891087530386,
+      "learning_rate": 4.983756251407569e-06,
+      "loss": 0.6165,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9053099530748333,
+      "grad_norm": 0.46771621207039893,
+      "learning_rate": 4.983738415612581e-06,
+      "loss": 0.5982,
+      "step": 2291
+    },
+    {
+      "epoch": 0.9057051123734255,
+      "grad_norm": 0.48007080844590533,
+      "learning_rate": 4.983720570062979e-06,
+      "loss": 0.6118,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9061002716720178,
+      "grad_norm": 0.47976244296713044,
+      "learning_rate": 4.9837027147588294e-06,
+      "loss": 0.6042,
+      "step": 2293
+    },
+    {
+      "epoch": 0.90649543097061,
+      "grad_norm": 0.5070646462309771,
+      "learning_rate": 4.983684849700204e-06,
+      "loss": 0.6336,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9068905902692023,
+      "grad_norm": 0.4746142937787273,
+      "learning_rate": 4.983666974887172e-06,
+      "loss": 0.6275,
+      "step": 2295
+    },
+    {
+      "epoch": 0.9072857495677945,
+      "grad_norm": 0.4773646681690295,
+      "learning_rate": 4.983649090319806e-06,
+      "loss": 0.6222,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9076809088663867,
+      "grad_norm": 0.4870955712187313,
+      "learning_rate": 4.983631195998173e-06,
+      "loss": 0.6336,
+      "step": 2297
+    },
+    {
+      "epoch": 0.908076068164979,
+      "grad_norm": 0.5385238468634763,
+      "learning_rate": 4.983613291922345e-06,
+      "loss": 0.6218,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9084712274635712,
+      "grad_norm": 0.4704713115373773,
+      "learning_rate": 4.983595378092393e-06,
+      "loss": 0.6295,
+      "step": 2299
+    },
+    {
+      "epoch": 0.9088663867621635,
+      "grad_norm": 0.5046914027772894,
+      "learning_rate": 4.9835774545083856e-06,
+      "loss": 0.6145,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9092615460607557,
+      "grad_norm": 0.4839085038708857,
+      "learning_rate": 4.983559521170394e-06,
+      "loss": 0.6254,
+      "step": 2301
+    },
+    {
+      "epoch": 0.909656705359348,
+      "grad_norm": 0.5087717682587891,
+      "learning_rate": 4.98354157807849e-06,
+      "loss": 0.6161,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9100518646579402,
+      "grad_norm": 0.46452927922636916,
+      "learning_rate": 4.983523625232741e-06,
+      "loss": 0.621,
+      "step": 2303
+    },
+    {
+      "epoch": 0.9104470239565324,
+      "grad_norm": 0.46904690117204156,
+      "learning_rate": 4.983505662633221e-06,
+      "loss": 0.6,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9108421832551247,
+      "grad_norm": 0.5301523549946244,
+      "learning_rate": 4.983487690279998e-06,
+      "loss": 0.6086,
+      "step": 2305
+    },
+    {
+      "epoch": 0.911237342553717,
+      "grad_norm": 0.5060704545892506,
+      "learning_rate": 4.983469708173143e-06,
+      "loss": 0.6229,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9116325018523093,
+      "grad_norm": 0.4813420559082125,
+      "learning_rate": 4.9834517163127275e-06,
+      "loss": 0.6211,
+      "step": 2307
+    },
+    {
+      "epoch": 0.9120276611509015,
+      "grad_norm": 0.5220914006598277,
+      "learning_rate": 4.983433714698821e-06,
+      "loss": 0.6152,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9124228204494937,
+      "grad_norm": 0.49458336480253934,
+      "learning_rate": 4.983415703331496e-06,
+      "loss": 0.5969,
+      "step": 2309
+    },
+    {
+      "epoch": 0.912817979748086,
+      "grad_norm": 0.48129884758049346,
+      "learning_rate": 4.983397682210821e-06,
+      "loss": 0.621,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9132131390466782,
+      "grad_norm": 0.4727468640393369,
+      "learning_rate": 4.983379651336869e-06,
+      "loss": 0.6197,
+      "step": 2311
+    },
+    {
+      "epoch": 0.9136082983452705,
+      "grad_norm": 0.49587024829033005,
+      "learning_rate": 4.983361610709709e-06,
+      "loss": 0.6351,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9140034576438627,
+      "grad_norm": 0.4887985778014849,
+      "learning_rate": 4.983343560329413e-06,
+      "loss": 0.6036,
+      "step": 2313
+    },
+    {
+      "epoch": 0.914398616942455,
+      "grad_norm": 0.5094208232865344,
+      "learning_rate": 4.983325500196051e-06,
+      "loss": 0.6281,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9147937762410472,
+      "grad_norm": 0.520779337465162,
+      "learning_rate": 4.983307430309695e-06,
+      "loss": 0.5966,
+      "step": 2315
+    },
+    {
+      "epoch": 0.9151889355396394,
+      "grad_norm": 0.4983724492164491,
+      "learning_rate": 4.983289350670415e-06,
+      "loss": 0.6002,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9155840948382317,
+      "grad_norm": 0.4977585587251415,
+      "learning_rate": 4.983271261278282e-06,
+      "loss": 0.6048,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9159792541368239,
+      "grad_norm": 0.5074344584060536,
+      "learning_rate": 4.983253162133368e-06,
+      "loss": 0.6061,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9163744134354161,
+      "grad_norm": 0.47393303007323573,
+      "learning_rate": 4.983235053235743e-06,
+      "loss": 0.6107,
+      "step": 2319
+    },
+    {
+      "epoch": 0.9167695727340084,
+      "grad_norm": 0.48092444595983924,
+      "learning_rate": 4.983216934585478e-06,
+      "loss": 0.6252,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9171647320326006,
+      "grad_norm": 0.5173009133356737,
+      "learning_rate": 4.983198806182647e-06,
+      "loss": 0.6342,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9175598913311929,
+      "grad_norm": 0.4823899650053501,
+      "learning_rate": 4.983180668027317e-06,
+      "loss": 0.6254,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9179550506297851,
+      "grad_norm": 0.48307570444067927,
+      "learning_rate": 4.9831625201195625e-06,
+      "loss": 0.6074,
+      "step": 2323
+    },
+    {
+      "epoch": 0.9183502099283773,
+      "grad_norm": 0.5114385837411004,
+      "learning_rate": 4.983144362459452e-06,
+      "loss": 0.5995,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9187453692269696,
+      "grad_norm": 0.49319776581630537,
+      "learning_rate": 4.9831261950470595e-06,
+      "loss": 0.6033,
+      "step": 2325
+    },
+    {
+      "epoch": 0.9191405285255618,
+      "grad_norm": 0.48172601408960075,
+      "learning_rate": 4.9831080178824545e-06,
+      "loss": 0.6144,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9195356878241541,
+      "grad_norm": 0.6998545259966715,
+      "learning_rate": 4.983089830965709e-06,
+      "loss": 0.5958,
+      "step": 2327
+    },
+    {
+      "epoch": 0.9199308471227463,
+      "grad_norm": 0.47578948749489164,
+      "learning_rate": 4.983071634296895e-06,
+      "loss": 0.6066,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9203260064213385,
+      "grad_norm": 0.4924183719099034,
+      "learning_rate": 4.983053427876083e-06,
+      "loss": 0.6328,
+      "step": 2329
+    },
+    {
+      "epoch": 0.9207211657199309,
+      "grad_norm": 0.46828818805254213,
+      "learning_rate": 4.983035211703345e-06,
+      "loss": 0.6272,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9211163250185231,
+      "grad_norm": 0.5014119318637013,
+      "learning_rate": 4.9830169857787524e-06,
+      "loss": 0.6223,
+      "step": 2331
+    },
+    {
+      "epoch": 0.9215114843171154,
+      "grad_norm": 0.509639756966488,
+      "learning_rate": 4.982998750102378e-06,
+      "loss": 0.6451,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9219066436157076,
+      "grad_norm": 0.4874868797487242,
+      "learning_rate": 4.982980504674291e-06,
+      "loss": 0.6198,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9223018029142999,
+      "grad_norm": 0.4910768108400356,
+      "learning_rate": 4.982962249494564e-06,
+      "loss": 0.6133,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9226969622128921,
+      "grad_norm": 0.600038361094374,
+      "learning_rate": 4.98294398456327e-06,
+      "loss": 0.5906,
+      "step": 2335
+    },
+    {
+      "epoch": 0.9230921215114843,
+      "grad_norm": 0.4882683625038398,
+      "learning_rate": 4.982925709880479e-06,
+      "loss": 0.6216,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9234872808100766,
+      "grad_norm": 0.4877944511418589,
+      "learning_rate": 4.982907425446264e-06,
+      "loss": 0.6142,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9238824401086688,
+      "grad_norm": 0.47849837901974185,
+      "learning_rate": 4.982889131260696e-06,
+      "loss": 0.629,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9242775994072611,
+      "grad_norm": 0.49657514695675,
+      "learning_rate": 4.982870827323847e-06,
+      "loss": 0.6344,
+      "step": 2339
+    },
+    {
+      "epoch": 0.9246727587058533,
+      "grad_norm": 0.4756867812643004,
+      "learning_rate": 4.98285251363579e-06,
+      "loss": 0.5892,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9250679180044455,
+      "grad_norm": 0.4792345020023528,
+      "learning_rate": 4.982834190196595e-06,
+      "loss": 0.6405,
+      "step": 2341
+    },
+    {
+      "epoch": 0.9254630773030378,
+      "grad_norm": 0.47568126691061463,
+      "learning_rate": 4.982815857006336e-06,
+      "loss": 0.6218,
+      "step": 2342
+    },
+    {
+      "epoch": 0.92585823660163,
+      "grad_norm": 0.4622150942965156,
+      "learning_rate": 4.982797514065083e-06,
+      "loss": 0.5966,
+      "step": 2343
+    },
+    {
+      "epoch": 0.9262533959002223,
+      "grad_norm": 0.48062685913560826,
+      "learning_rate": 4.982779161372909e-06,
+      "loss": 0.6067,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9266485551988145,
+      "grad_norm": 0.49832908386195485,
+      "learning_rate": 4.9827607989298874e-06,
+      "loss": 0.6213,
+      "step": 2345
+    },
+    {
+      "epoch": 0.9270437144974067,
+      "grad_norm": 0.4651264905943558,
+      "learning_rate": 4.982742426736088e-06,
+      "loss": 0.6104,
+      "step": 2346
+    },
+    {
+      "epoch": 0.927438873795999,
+      "grad_norm": 0.4627917129880916,
+      "learning_rate": 4.982724044791584e-06,
+      "loss": 0.5947,
+      "step": 2347
+    },
+    {
+      "epoch": 0.9278340330945912,
+      "grad_norm": 0.47459773400411004,
+      "learning_rate": 4.982705653096447e-06,
+      "loss": 0.6171,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9282291923931835,
+      "grad_norm": 0.4754459015910213,
+      "learning_rate": 4.982687251650751e-06,
+      "loss": 0.5867,
+      "step": 2349
+    },
+    {
+      "epoch": 0.9286243516917757,
+      "grad_norm": 0.4693623306765774,
+      "learning_rate": 4.9826688404545655e-06,
+      "loss": 0.6016,
+      "step": 2350
+    },
+    {
+      "epoch": 0.929019510990368,
+      "grad_norm": 0.4595592610393575,
+      "learning_rate": 4.982650419507965e-06,
+      "loss": 0.6083,
+      "step": 2351
+    },
+    {
+      "epoch": 0.9294146702889602,
+      "grad_norm": 1.0329197407008157,
+      "learning_rate": 4.982631988811022e-06,
+      "loss": 0.6151,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9298098295875524,
+      "grad_norm": 0.46434445302211896,
+      "learning_rate": 4.982613548363807e-06,
+      "loss": 0.6109,
+      "step": 2353
+    },
+    {
+      "epoch": 0.9302049888861448,
+      "grad_norm": 0.4720496810881737,
+      "learning_rate": 4.982595098166394e-06,
+      "loss": 0.6135,
+      "step": 2354
+    },
+    {
+      "epoch": 0.930600148184737,
+      "grad_norm": 0.46567296934174945,
+      "learning_rate": 4.982576638218855e-06,
+      "loss": 0.6092,
+      "step": 2355
+    },
+    {
+      "epoch": 0.9309953074833293,
+      "grad_norm": 0.4783627968108407,
+      "learning_rate": 4.982558168521263e-06,
+      "loss": 0.6226,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9313904667819215,
+      "grad_norm": 0.46771411865600776,
+      "learning_rate": 4.982539689073689e-06,
+      "loss": 0.6216,
+      "step": 2357
+    },
+    {
+      "epoch": 0.9317856260805137,
+      "grad_norm": 0.4824264375712755,
+      "learning_rate": 4.982521199876207e-06,
+      "loss": 0.6331,
+      "step": 2358
+    },
+    {
+      "epoch": 0.932180785379106,
+      "grad_norm": 0.4870691371321396,
+      "learning_rate": 4.9825027009288896e-06,
+      "loss": 0.6275,
+      "step": 2359
+    },
+    {
+      "epoch": 0.9325759446776982,
+      "grad_norm": 0.48392613415399993,
+      "learning_rate": 4.982484192231808e-06,
+      "loss": 0.6143,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9329711039762905,
+      "grad_norm": 0.4622678197643799,
+      "learning_rate": 4.9824656737850365e-06,
+      "loss": 0.6188,
+      "step": 2361
+    },
+    {
+      "epoch": 0.9333662632748827,
+      "grad_norm": 0.49520864864087916,
+      "learning_rate": 4.982447145588648e-06,
+      "loss": 0.5983,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9337614225734749,
+      "grad_norm": 0.49238331015657344,
+      "learning_rate": 4.982428607642713e-06,
+      "loss": 0.6304,
+      "step": 2363
+    },
+    {
+      "epoch": 0.9341565818720672,
+      "grad_norm": 0.46810832966740035,
+      "learning_rate": 4.9824100599473065e-06,
+      "loss": 0.6057,
+      "step": 2364
+    },
+    {
+      "epoch": 0.9345517411706594,
+      "grad_norm": 0.4789733237570732,
+      "learning_rate": 4.982391502502501e-06,
+      "loss": 0.5931,
+      "step": 2365
+    },
+    {
+      "epoch": 0.9349469004692517,
+      "grad_norm": 0.48015565027094287,
+      "learning_rate": 4.98237293530837e-06,
+      "loss": 0.6304,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9353420597678439,
+      "grad_norm": 0.4851134573753921,
+      "learning_rate": 4.9823543583649846e-06,
+      "loss": 0.6296,
+      "step": 2367
+    },
+    {
+      "epoch": 0.9357372190664361,
+      "grad_norm": 0.48585992460960437,
+      "learning_rate": 4.982335771672418e-06,
+      "loss": 0.5901,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9361323783650284,
+      "grad_norm": 0.47063745016814984,
+      "learning_rate": 4.982317175230744e-06,
+      "loss": 0.6024,
+      "step": 2369
+    },
+    {
+      "epoch": 0.9365275376636206,
+      "grad_norm": 0.4799775780201179,
+      "learning_rate": 4.982298569040036e-06,
+      "loss": 0.6158,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9369226969622129,
+      "grad_norm": 0.487398135567962,
+      "learning_rate": 4.982279953100366e-06,
+      "loss": 0.6207,
+      "step": 2371
+    },
+    {
+      "epoch": 0.9373178562608051,
+      "grad_norm": 0.4707569265394489,
+      "learning_rate": 4.9822613274118085e-06,
+      "loss": 0.6129,
+      "step": 2372
+    },
+    {
+      "epoch": 0.9377130155593973,
+      "grad_norm": 0.48484895295531333,
+      "learning_rate": 4.9822426919744355e-06,
+      "loss": 0.5927,
+      "step": 2373
+    },
+    {
+      "epoch": 0.9381081748579896,
+      "grad_norm": 0.5086631550313602,
+      "learning_rate": 4.9822240467883205e-06,
+      "loss": 0.6513,
+      "step": 2374
+    },
+    {
+      "epoch": 0.9385033341565818,
+      "grad_norm": 0.4854307750752647,
+      "learning_rate": 4.9822053918535366e-06,
+      "loss": 0.639,
+      "step": 2375
+    },
+    {
+      "epoch": 0.9388984934551741,
+      "grad_norm": 0.49908919962486925,
+      "learning_rate": 4.982186727170157e-06,
+      "loss": 0.6171,
+      "step": 2376
+    },
+    {
+      "epoch": 0.9392936527537664,
+      "grad_norm": 0.47151786983504285,
+      "learning_rate": 4.982168052738255e-06,
+      "loss": 0.6127,
+      "step": 2377
+    },
+    {
+      "epoch": 0.9396888120523587,
+      "grad_norm": 0.48250797464606165,
+      "learning_rate": 4.982149368557905e-06,
+      "loss": 0.636,
+      "step": 2378
+    },
+    {
+      "epoch": 0.9400839713509509,
+      "grad_norm": 0.4839100853335766,
+      "learning_rate": 4.982130674629179e-06,
+      "loss": 0.6175,
+      "step": 2379
+    },
+    {
+      "epoch": 0.9404791306495431,
+      "grad_norm": 0.46593074774551896,
+      "learning_rate": 4.982111970952151e-06,
+      "loss": 0.6,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9408742899481354,
+      "grad_norm": 0.4726472043264624,
+      "learning_rate": 4.982093257526894e-06,
+      "loss": 0.6306,
+      "step": 2381
+    },
+    {
+      "epoch": 0.9412694492467276,
+      "grad_norm": 0.4707519207258955,
+      "learning_rate": 4.982074534353482e-06,
+      "loss": 0.6089,
+      "step": 2382
+    },
+    {
+      "epoch": 0.9416646085453199,
+      "grad_norm": 0.47810887989613376,
+      "learning_rate": 4.9820558014319895e-06,
+      "loss": 0.6068,
+      "step": 2383
+    },
+    {
+      "epoch": 0.9420597678439121,
+      "grad_norm": 0.46370942325328973,
+      "learning_rate": 4.9820370587624875e-06,
+      "loss": 0.5938,
+      "step": 2384
+    },
+    {
+      "epoch": 0.9424549271425043,
+      "grad_norm": 0.5035785783227827,
+      "learning_rate": 4.9820183063450525e-06,
+      "loss": 0.6265,
+      "step": 2385
+    },
+    {
+      "epoch": 0.9428500864410966,
+      "grad_norm": 0.5024383370795096,
+      "learning_rate": 4.981999544179756e-06,
+      "loss": 0.6236,
+      "step": 2386
+    },
+    {
+      "epoch": 0.9432452457396888,
+      "grad_norm": 0.45990102437488745,
+      "learning_rate": 4.981980772266672e-06,
+      "loss": 0.5894,
+      "step": 2387
+    },
+    {
+      "epoch": 0.9436404050382811,
+      "grad_norm": 0.4654023117324646,
+      "learning_rate": 4.981961990605876e-06,
+      "loss": 0.5898,
+      "step": 2388
+    },
+    {
+      "epoch": 0.9440355643368733,
+      "grad_norm": 0.4762087512767081,
+      "learning_rate": 4.981943199197439e-06,
+      "loss": 0.6056,
+      "step": 2389
+    },
+    {
+      "epoch": 0.9444307236354655,
+      "grad_norm": 0.4749986650314227,
+      "learning_rate": 4.981924398041436e-06,
+      "loss": 0.6135,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9448258829340578,
+      "grad_norm": 0.4719526074802648,
+      "learning_rate": 4.981905587137943e-06,
+      "loss": 0.5893,
+      "step": 2391
+    },
+    {
+      "epoch": 0.94522104223265,
+      "grad_norm": 0.4755247323470361,
+      "learning_rate": 4.9818867664870306e-06,
+      "loss": 0.6073,
+      "step": 2392
+    },
+    {
+      "epoch": 0.9456162015312423,
+      "grad_norm": 0.48675290749713584,
+      "learning_rate": 4.981867936088774e-06,
+      "loss": 0.5924,
+      "step": 2393
+    },
+    {
+      "epoch": 0.9460113608298345,
+      "grad_norm": 0.4743633378571163,
+      "learning_rate": 4.9818490959432485e-06,
+      "loss": 0.62,
+      "step": 2394
+    },
+    {
+      "epoch": 0.9464065201284267,
+      "grad_norm": 0.49288950026353556,
+      "learning_rate": 4.981830246050526e-06,
+      "loss": 0.6045,
+      "step": 2395
+    },
+    {
+      "epoch": 0.946801679427019,
+      "grad_norm": 0.4884082163788517,
+      "learning_rate": 4.981811386410681e-06,
+      "loss": 0.6118,
+      "step": 2396
+    },
+    {
+      "epoch": 0.9471968387256112,
+      "grad_norm": 0.4818633127585905,
+      "learning_rate": 4.981792517023788e-06,
+      "loss": 0.6154,
+      "step": 2397
+    },
+    {
+      "epoch": 0.9475919980242035,
+      "grad_norm": 0.5212061592296549,
+      "learning_rate": 4.981773637889921e-06,
+      "loss": 0.5999,
+      "step": 2398
+    },
+    {
+      "epoch": 0.9479871573227957,
+      "grad_norm": 0.4921362608346308,
+      "learning_rate": 4.981754749009154e-06,
+      "loss": 0.6248,
+      "step": 2399
+    },
+    {
+      "epoch": 0.9483823166213879,
+      "grad_norm": 0.6375277486862304,
+      "learning_rate": 4.9817358503815616e-06,
+      "loss": 0.6205,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9487774759199803,
+      "grad_norm": 0.4822713530796236,
+      "learning_rate": 4.9817169420072175e-06,
+      "loss": 0.6068,
+      "step": 2401
+    },
+    {
+      "epoch": 0.9491726352185725,
+      "grad_norm": 0.45735313205333067,
+      "learning_rate": 4.981698023886197e-06,
+      "loss": 0.6045,
+      "step": 2402
+    },
+    {
+      "epoch": 0.9495677945171648,
+      "grad_norm": 0.4870688284599884,
+      "learning_rate": 4.9816790960185725e-06,
+      "loss": 0.6169,
+      "step": 2403
+    },
+    {
+      "epoch": 0.949962953815757,
+      "grad_norm": 0.4508977591163079,
+      "learning_rate": 4.981660158404421e-06,
+      "loss": 0.6081,
+      "step": 2404
+    },
+    {
+      "epoch": 0.9503581131143493,
+      "grad_norm": 0.48177734751539525,
+      "learning_rate": 4.981641211043814e-06,
+      "loss": 0.6075,
+      "step": 2405
+    },
+    {
+      "epoch": 0.9507532724129415,
+      "grad_norm": 0.45900423930549045,
+      "learning_rate": 4.981622253936828e-06,
+      "loss": 0.5804,
+      "step": 2406
+    },
+    {
+      "epoch": 0.9511484317115337,
+      "grad_norm": 0.46636251306928633,
+      "learning_rate": 4.9816032870835366e-06,
+      "loss": 0.6093,
+      "step": 2407
+    },
+    {
+      "epoch": 0.951543591010126,
+      "grad_norm": 0.49173914273079045,
+      "learning_rate": 4.981584310484014e-06,
+      "loss": 0.6423,
+      "step": 2408
+    },
+    {
+      "epoch": 0.9519387503087182,
+      "grad_norm": 0.4871067267842084,
+      "learning_rate": 4.981565324138336e-06,
+      "loss": 0.6404,
+      "step": 2409
+    },
+    {
+      "epoch": 0.9523339096073105,
+      "grad_norm": 0.48801903916719275,
+      "learning_rate": 4.981546328046575e-06,
+      "loss": 0.6091,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9527290689059027,
+      "grad_norm": 0.47631638584783426,
+      "learning_rate": 4.9815273222088075e-06,
+      "loss": 0.6086,
+      "step": 2411
+    },
+    {
+      "epoch": 0.9531242282044949,
+      "grad_norm": 0.45326422779082315,
+      "learning_rate": 4.981508306625108e-06,
+      "loss": 0.612,
+      "step": 2412
+    },
+    {
+      "epoch": 0.9535193875030872,
+      "grad_norm": 0.5469615293144247,
+      "learning_rate": 4.981489281295549e-06,
+      "loss": 0.5988,
+      "step": 2413
+    },
+    {
+      "epoch": 0.9539145468016794,
+      "grad_norm": 0.49738406140305297,
+      "learning_rate": 4.9814702462202094e-06,
+      "loss": 0.6225,
+      "step": 2414
+    },
+    {
+      "epoch": 0.9543097061002717,
+      "grad_norm": 0.4621955995020316,
+      "learning_rate": 4.98145120139916e-06,
+      "loss": 0.6016,
+      "step": 2415
+    },
+    {
+      "epoch": 0.9547048653988639,
+      "grad_norm": 0.46293528627278396,
+      "learning_rate": 4.981432146832478e-06,
+      "loss": 0.6214,
+      "step": 2416
+    },
+    {
+      "epoch": 0.9551000246974561,
+      "grad_norm": 0.49663994492008307,
+      "learning_rate": 4.981413082520236e-06,
+      "loss": 0.6133,
+      "step": 2417
+    },
+    {
+      "epoch": 0.9554951839960484,
+      "grad_norm": 0.4817582642896156,
+      "learning_rate": 4.981394008462511e-06,
+      "loss": 0.6238,
+      "step": 2418
+    },
+    {
+      "epoch": 0.9558903432946406,
+      "grad_norm": 0.5586123030609345,
+      "learning_rate": 4.981374924659377e-06,
+      "loss": 0.6243,
+      "step": 2419
+    },
+    {
+      "epoch": 0.9562855025932329,
+      "grad_norm": 0.4850500636019948,
+      "learning_rate": 4.9813558311109095e-06,
+      "loss": 0.622,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9566806618918251,
+      "grad_norm": 0.4689441752707809,
+      "learning_rate": 4.9813367278171835e-06,
+      "loss": 0.5765,
+      "step": 2421
+    },
+    {
+      "epoch": 0.9570758211904173,
+      "grad_norm": 0.47158958076654856,
+      "learning_rate": 4.981317614778272e-06,
+      "loss": 0.6296,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9574709804890096,
+      "grad_norm": 0.514101025815252,
+      "learning_rate": 4.981298491994252e-06,
+      "loss": 0.6234,
+      "step": 2423
+    },
+    {
+      "epoch": 0.9578661397876018,
+      "grad_norm": 0.48399944464009775,
+      "learning_rate": 4.981279359465199e-06,
+      "loss": 0.6107,
+      "step": 2424
+    },
+    {
+      "epoch": 0.9582612990861942,
+      "grad_norm": 0.48729938368312087,
+      "learning_rate": 4.981260217191187e-06,
+      "loss": 0.5893,
+      "step": 2425
+    },
+    {
+      "epoch": 0.9586564583847864,
+      "grad_norm": 0.49004477541134556,
+      "learning_rate": 4.981241065172292e-06,
+      "loss": 0.6008,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9590516176833787,
+      "grad_norm": 0.48347415116566556,
+      "learning_rate": 4.9812219034085886e-06,
+      "loss": 0.618,
+      "step": 2427
+    },
+    {
+      "epoch": 0.9594467769819709,
+      "grad_norm": 0.4793590490738295,
+      "learning_rate": 4.981202731900152e-06,
+      "loss": 0.6251,
+      "step": 2428
+    },
+    {
+      "epoch": 0.9598419362805631,
+      "grad_norm": 0.48347957011954834,
+      "learning_rate": 4.9811835506470575e-06,
+      "loss": 0.6144,
+      "step": 2429
+    },
+    {
+      "epoch": 0.9602370955791554,
+      "grad_norm": 0.5000951926275022,
+      "learning_rate": 4.981164359649381e-06,
+      "loss": 0.6217,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9606322548777476,
+      "grad_norm": 0.4853268721052706,
+      "learning_rate": 4.981145158907198e-06,
+      "loss": 0.6107,
+      "step": 2431
+    },
+    {
+      "epoch": 0.9610274141763399,
+      "grad_norm": 0.4775051457662627,
+      "learning_rate": 4.981125948420583e-06,
+      "loss": 0.619,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9614225734749321,
+      "grad_norm": 0.4610581366616115,
+      "learning_rate": 4.981106728189612e-06,
+      "loss": 0.6011,
+      "step": 2433
+    },
+    {
+      "epoch": 0.9618177327735243,
+      "grad_norm": 0.46908821425073965,
+      "learning_rate": 4.98108749821436e-06,
+      "loss": 0.5991,
+      "step": 2434
+    },
+    {
+      "epoch": 0.9622128920721166,
+      "grad_norm": 0.4643140368630478,
+      "learning_rate": 4.981068258494903e-06,
+      "loss": 0.5848,
+      "step": 2435
+    },
+    {
+      "epoch": 0.9626080513707088,
+      "grad_norm": 0.48640039378791994,
+      "learning_rate": 4.981049009031317e-06,
+      "loss": 0.6175,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9630032106693011,
+      "grad_norm": 0.4707625851268911,
+      "learning_rate": 4.9810297498236765e-06,
+      "loss": 0.5903,
+      "step": 2437
+    },
+    {
+      "epoch": 0.9633983699678933,
+      "grad_norm": 0.47186949155654456,
+      "learning_rate": 4.981010480872058e-06,
+      "loss": 0.6051,
+      "step": 2438
+    },
+    {
+      "epoch": 0.9637935292664855,
+      "grad_norm": 0.47253458553373584,
+      "learning_rate": 4.980991202176536e-06,
+      "loss": 0.6135,
+      "step": 2439
+    },
+    {
+      "epoch": 0.9641886885650778,
+      "grad_norm": 0.4739469724306039,
+      "learning_rate": 4.980971913737188e-06,
+      "loss": 0.5884,
+      "step": 2440
+    },
+    {
+      "epoch": 0.96458384786367,
+      "grad_norm": 0.47756433353936395,
+      "learning_rate": 4.980952615554089e-06,
+      "loss": 0.6054,
+      "step": 2441
+    },
+    {
+      "epoch": 0.9649790071622623,
+      "grad_norm": 0.46338471724667013,
+      "learning_rate": 4.980933307627315e-06,
+      "loss": 0.611,
+      "step": 2442
+    },
+    {
+      "epoch": 0.9653741664608545,
+      "grad_norm": 0.487680669856884,
+      "learning_rate": 4.98091398995694e-06,
+      "loss": 0.6192,
+      "step": 2443
+    },
+    {
+      "epoch": 0.9657693257594467,
+      "grad_norm": 0.5227924200190431,
+      "learning_rate": 4.9808946625430425e-06,
+      "loss": 0.6034,
+      "step": 2444
+    },
+    {
+      "epoch": 0.966164485058039,
+      "grad_norm": 0.49032597318513343,
+      "learning_rate": 4.980875325385697e-06,
+      "loss": 0.588,
+      "step": 2445
+    },
+    {
+      "epoch": 0.9665596443566312,
+      "grad_norm": 0.46454318608059514,
+      "learning_rate": 4.98085597848498e-06,
+      "loss": 0.5968,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9669548036552235,
+      "grad_norm": 0.5078562623229934,
+      "learning_rate": 4.980836621840967e-06,
+      "loss": 0.6195,
+      "step": 2447
+    },
+    {
+      "epoch": 0.9673499629538158,
+      "grad_norm": 0.4752546102671055,
+      "learning_rate": 4.980817255453734e-06,
+      "loss": 0.6118,
+      "step": 2448
+    },
+    {
+      "epoch": 0.967745122252408,
+      "grad_norm": 0.4833480233978842,
+      "learning_rate": 4.980797879323357e-06,
+      "loss": 0.6261,
+      "step": 2449
+    },
+    {
+      "epoch": 0.9681402815510003,
+      "grad_norm": 0.4771570217763964,
+      "learning_rate": 4.980778493449912e-06,
+      "loss": 0.6318,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9685354408495925,
+      "grad_norm": 0.47469969618301255,
+      "learning_rate": 4.980759097833476e-06,
+      "loss": 0.6027,
+      "step": 2451
+    },
+    {
+      "epoch": 0.9689306001481848,
+      "grad_norm": 0.4781745301282861,
+      "learning_rate": 4.980739692474125e-06,
+      "loss": 0.6172,
+      "step": 2452
+    },
+    {
+      "epoch": 0.969325759446777,
+      "grad_norm": 0.48953552287496055,
+      "learning_rate": 4.980720277371934e-06,
+      "loss": 0.6059,
+      "step": 2453
+    },
+    {
+      "epoch": 0.9697209187453693,
+      "grad_norm": 0.4697257378860957,
+      "learning_rate": 4.980700852526981e-06,
+      "loss": 0.6108,
+      "step": 2454
+    },
+    {
+      "epoch": 0.9701160780439615,
+      "grad_norm": 0.5170911041975621,
+      "learning_rate": 4.980681417939341e-06,
+      "loss": 0.6353,
+      "step": 2455
+    },
+    {
+      "epoch": 0.9705112373425537,
+      "grad_norm": 0.4765311737964586,
+      "learning_rate": 4.980661973609091e-06,
+      "loss": 0.5911,
+      "step": 2456
+    },
+    {
+      "epoch": 0.970906396641146,
+      "grad_norm": 0.4571918246118576,
+      "learning_rate": 4.980642519536307e-06,
+      "loss": 0.6204,
+      "step": 2457
+    },
+    {
+      "epoch": 0.9713015559397382,
+      "grad_norm": 0.47402704361826625,
+      "learning_rate": 4.980623055721065e-06,
+      "loss": 0.6075,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9716967152383305,
+      "grad_norm": 0.4823370913476268,
+      "learning_rate": 4.980603582163443e-06,
+      "loss": 0.612,
+      "step": 2459
+    },
+    {
+      "epoch": 0.9720918745369227,
+      "grad_norm": 0.4665199872651599,
+      "learning_rate": 4.980584098863516e-06,
+      "loss": 0.6255,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9724870338355149,
+      "grad_norm": 0.45850568990455864,
+      "learning_rate": 4.980564605821361e-06,
+      "loss": 0.6003,
+      "step": 2461
+    },
+    {
+      "epoch": 0.9728821931341072,
+      "grad_norm": 0.4660323061963318,
+      "learning_rate": 4.980545103037054e-06,
+      "loss": 0.6153,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9732773524326994,
+      "grad_norm": 0.47426339473199003,
+      "learning_rate": 4.9805255905106735e-06,
+      "loss": 0.6203,
+      "step": 2463
+    },
+    {
+      "epoch": 0.9736725117312917,
+      "grad_norm": 0.4648762466922138,
+      "learning_rate": 4.9805060682422925e-06,
+      "loss": 0.602,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9740676710298839,
+      "grad_norm": 0.4774681753525283,
+      "learning_rate": 4.980486536231992e-06,
+      "loss": 0.6199,
+      "step": 2465
+    },
+    {
+      "epoch": 0.9744628303284761,
+      "grad_norm": 0.5025939174795709,
+      "learning_rate": 4.980466994479845e-06,
+      "loss": 0.6342,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9748579896270684,
+      "grad_norm": 0.46520163938175313,
+      "learning_rate": 4.980447442985931e-06,
+      "loss": 0.6082,
+      "step": 2467
+    },
+    {
+      "epoch": 0.9752531489256606,
+      "grad_norm": 0.5093555297811326,
+      "learning_rate": 4.980427881750325e-06,
+      "loss": 0.6406,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9756483082242529,
+      "grad_norm": 0.4675282128894731,
+      "learning_rate": 4.980408310773105e-06,
+      "loss": 0.6013,
+      "step": 2469
+    },
+    {
+      "epoch": 0.9760434675228451,
+      "grad_norm": 0.4600464903665783,
+      "learning_rate": 4.980388730054347e-06,
+      "loss": 0.6021,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9764386268214373,
+      "grad_norm": 0.4781471662113073,
+      "learning_rate": 4.980369139594128e-06,
+      "loss": 0.6302,
+      "step": 2471
+    },
+    {
+      "epoch": 0.9768337861200297,
+      "grad_norm": 0.569664962295512,
+      "learning_rate": 4.980349539392526e-06,
+      "loss": 0.6076,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9772289454186219,
+      "grad_norm": 0.474728795793001,
+      "learning_rate": 4.980329929449616e-06,
+      "loss": 0.5973,
+      "step": 2473
+    },
+    {
+      "epoch": 0.9776241047172142,
+      "grad_norm": 0.466971617619074,
+      "learning_rate": 4.980310309765477e-06,
+      "loss": 0.5881,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9780192640158064,
+      "grad_norm": 0.49734334610344294,
+      "learning_rate": 4.980290680340185e-06,
+      "loss": 0.6377,
+      "step": 2475
+    },
+    {
+      "epoch": 0.9784144233143987,
+      "grad_norm": 0.4660369065720144,
+      "learning_rate": 4.980271041173818e-06,
+      "loss": 0.5939,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9788095826129909,
+      "grad_norm": 0.4541019720752234,
+      "learning_rate": 4.9802513922664506e-06,
+      "loss": 0.6013,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9792047419115831,
+      "grad_norm": 0.46864260551647896,
+      "learning_rate": 4.980231733618164e-06,
+      "loss": 0.5934,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9795999012101754,
+      "grad_norm": 0.49812016619203664,
+      "learning_rate": 4.980212065229032e-06,
+      "loss": 0.6074,
+      "step": 2479
+    },
+    {
+      "epoch": 0.9799950605087676,
+      "grad_norm": 0.47269225015752886,
+      "learning_rate": 4.980192387099133e-06,
+      "loss": 0.6075,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9803902198073599,
+      "grad_norm": 0.46117462938553433,
+      "learning_rate": 4.980172699228545e-06,
+      "loss": 0.5987,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9807853791059521,
+      "grad_norm": 0.47234556720051585,
+      "learning_rate": 4.980153001617344e-06,
+      "loss": 0.6008,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9811805384045443,
+      "grad_norm": 0.46430799343376644,
+      "learning_rate": 4.980133294265608e-06,
+      "loss": 0.6126,
+      "step": 2483
+    },
+    {
+      "epoch": 0.9815756977031366,
+      "grad_norm": 0.48231158605504654,
+      "learning_rate": 4.980113577173415e-06,
+      "loss": 0.6166,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9819708570017288,
+      "grad_norm": 0.4544084646577223,
+      "learning_rate": 4.980093850340842e-06,
+      "loss": 0.5911,
+      "step": 2485
+    },
+    {
+      "epoch": 0.982366016300321,
+      "grad_norm": 0.49807318396062805,
+      "learning_rate": 4.980074113767966e-06,
+      "loss": 0.596,
+      "step": 2486
+    },
+    {
+      "epoch": 0.9827611755989133,
+      "grad_norm": 0.47544603743432695,
+      "learning_rate": 4.980054367454865e-06,
+      "loss": 0.6095,
+      "step": 2487
+    },
+    {
+      "epoch": 0.9831563348975055,
+      "grad_norm": 0.4659987694152846,
+      "learning_rate": 4.9800346114016165e-06,
+      "loss": 0.606,
+      "step": 2488
+    },
+    {
+      "epoch": 0.9835514941960978,
+      "grad_norm": 0.47287386440876494,
+      "learning_rate": 4.980014845608298e-06,
+      "loss": 0.6352,
+      "step": 2489
+    },
+    {
+      "epoch": 0.98394665349469,
+      "grad_norm": 0.4864128975388848,
+      "learning_rate": 4.979995070074987e-06,
+      "loss": 0.6278,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9843418127932823,
+      "grad_norm": 0.4544891691408379,
+      "learning_rate": 4.979975284801761e-06,
+      "loss": 0.608,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9847369720918745,
+      "grad_norm": 0.4795256509587486,
+      "learning_rate": 4.979955489788698e-06,
+      "loss": 0.6273,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9851321313904667,
+      "grad_norm": 0.4766741019260381,
+      "learning_rate": 4.979935685035876e-06,
+      "loss": 0.608,
+      "step": 2493
+    },
+    {
+      "epoch": 0.985527290689059,
+      "grad_norm": 0.4737813917288424,
+      "learning_rate": 4.979915870543372e-06,
+      "loss": 0.6259,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9859224499876512,
+      "grad_norm": 0.4923664866297512,
+      "learning_rate": 4.979896046311266e-06,
+      "loss": 0.5989,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9863176092862436,
+      "grad_norm": 0.49326185549267143,
+      "learning_rate": 4.979876212339632e-06,
+      "loss": 0.6077,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9867127685848358,
+      "grad_norm": 0.4660497012513692,
+      "learning_rate": 4.9798563686285515e-06,
+      "loss": 0.6168,
+      "step": 2497
+    },
+    {
+      "epoch": 0.987107927883428,
+      "grad_norm": 0.4800478363437438,
+      "learning_rate": 4.979836515178101e-06,
+      "loss": 0.6211,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9875030871820203,
+      "grad_norm": 0.47487319014526613,
+      "learning_rate": 4.979816651988358e-06,
+      "loss": 0.6136,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9878982464806125,
+      "grad_norm": 0.47991505886589036,
+      "learning_rate": 4.979796779059401e-06,
+      "loss": 0.6081,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9882934057792048,
+      "grad_norm": 0.472571865692289,
+      "learning_rate": 4.979776896391308e-06,
+      "loss": 0.6208,
+      "step": 2501
+    },
+    {
+      "epoch": 0.988688565077797,
+      "grad_norm": 0.5058758260925268,
+      "learning_rate": 4.979757003984158e-06,
+      "loss": 0.6318,
+      "step": 2502
+    },
+    {
+      "epoch": 0.9890837243763893,
+      "grad_norm": 0.45916735017432975,
+      "learning_rate": 4.979737101838028e-06,
+      "loss": 0.5949,
+      "step": 2503
+    },
+    {
+      "epoch": 0.9894788836749815,
+      "grad_norm": 0.46426645842628067,
+      "learning_rate": 4.979717189952996e-06,
+      "loss": 0.6169,
+      "step": 2504
+    },
+    {
+      "epoch": 0.9898740429735737,
+      "grad_norm": 0.46140290638450737,
+      "learning_rate": 4.9796972683291415e-06,
+      "loss": 0.6021,
+      "step": 2505
+    },
+    {
+      "epoch": 0.990269202272166,
+      "grad_norm": 0.4503137341547014,
+      "learning_rate": 4.979677336966541e-06,
+      "loss": 0.5937,
+      "step": 2506
+    },
+    {
+      "epoch": 0.9906643615707582,
+      "grad_norm": 0.4734628943802484,
+      "learning_rate": 4.9796573958652735e-06,
+      "loss": 0.6208,
+      "step": 2507
+    },
+    {
+      "epoch": 0.9910595208693505,
+      "grad_norm": 0.4665327413909791,
+      "learning_rate": 4.979637445025418e-06,
+      "loss": 0.5911,
+      "step": 2508
+    },
+    {
+      "epoch": 0.9914546801679427,
+      "grad_norm": 0.5283523686632157,
+      "learning_rate": 4.979617484447052e-06,
+      "loss": 0.6054,
+      "step": 2509
+    },
+    {
+      "epoch": 0.9918498394665349,
+      "grad_norm": 0.4671594121597241,
+      "learning_rate": 4.979597514130254e-06,
+      "loss": 0.6187,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9922449987651272,
+      "grad_norm": 0.4893067227078538,
+      "learning_rate": 4.979577534075103e-06,
+      "loss": 0.6339,
+      "step": 2511
+    },
+    {
+      "epoch": 0.9926401580637194,
+      "grad_norm": 0.47208844535109873,
+      "learning_rate": 4.979557544281677e-06,
+      "loss": 0.6087,
+      "step": 2512
+    },
+    {
+      "epoch": 0.9930353173623117,
+      "grad_norm": 0.48119092703509836,
+      "learning_rate": 4.979537544750055e-06,
+      "loss": 0.6201,
+      "step": 2513
+    },
+    {
+      "epoch": 0.9934304766609039,
+      "grad_norm": 0.477557865345246,
+      "learning_rate": 4.979517535480315e-06,
+      "loss": 0.6143,
+      "step": 2514
+    },
+    {
+      "epoch": 0.9938256359594961,
+      "grad_norm": 0.4585124296524369,
+      "learning_rate": 4.979497516472535e-06,
+      "loss": 0.5979,
+      "step": 2515
+    },
+    {
+      "epoch": 0.9942207952580884,
+      "grad_norm": 0.47684697114164715,
+      "learning_rate": 4.979477487726796e-06,
+      "loss": 0.6031,
+      "step": 2516
+    },
+    {
+      "epoch": 0.9946159545566806,
+      "grad_norm": 0.46484256613871744,
+      "learning_rate": 4.979457449243174e-06,
+      "loss": 0.5965,
+      "step": 2517
+    },
+    {
+      "epoch": 0.9950111138552729,
+      "grad_norm": 0.46418868296419014,
+      "learning_rate": 4.97943740102175e-06,
+      "loss": 0.6187,
+      "step": 2518
+    },
+    {
+      "epoch": 0.9954062731538652,
+      "grad_norm": 0.4712799089967701,
+      "learning_rate": 4.9794173430626e-06,
+      "loss": 0.6179,
+      "step": 2519
+    },
+    {
+      "epoch": 0.9958014324524574,
+      "grad_norm": 0.4898574169755568,
+      "learning_rate": 4.979397275365804e-06,
+      "loss": 0.6182,
+      "step": 2520
+    },
+    {
+      "epoch": 0.9961965917510497,
+      "grad_norm": 0.4721726580593012,
+      "learning_rate": 4.979377197931442e-06,
+      "loss": 0.6141,
+      "step": 2521
+    },
+    {
+      "epoch": 0.9965917510496419,
+      "grad_norm": 0.4696253306480775,
+      "learning_rate": 4.979357110759592e-06,
+      "loss": 0.6024,
+      "step": 2522
+    },
+    {
+      "epoch": 0.9969869103482342,
+      "grad_norm": 0.4695198005530173,
+      "learning_rate": 4.979337013850332e-06,
+      "loss": 0.6183,
+      "step": 2523
+    },
+    {
+      "epoch": 0.9973820696468264,
+      "grad_norm": 0.46544585435480235,
+      "learning_rate": 4.979316907203743e-06,
+      "loss": 0.5947,
+      "step": 2524
+    },
+    {
+      "epoch": 0.9977772289454186,
+      "grad_norm": 0.4673922524133896,
+      "learning_rate": 4.979296790819901e-06,
+      "loss": 0.6276,
+      "step": 2525
+    },
+    {
+      "epoch": 0.9981723882440109,
+      "grad_norm": 0.4784936628964307,
+      "learning_rate": 4.979276664698888e-06,
+      "loss": 0.6163,
+      "step": 2526
+    },
+    {
+      "epoch": 0.9985675475426031,
+      "grad_norm": 0.46022244379627936,
+      "learning_rate": 4.979256528840782e-06,
+      "loss": 0.6152,
+      "step": 2527
+    },
+    {
+      "epoch": 0.9989627068411954,
+      "grad_norm": 0.46842270777733813,
+      "learning_rate": 4.979236383245661e-06,
+      "loss": 0.611,
+      "step": 2528
+    },
+    {
+      "epoch": 0.9993578661397876,
+      "grad_norm": 0.45419251266565824,
+      "learning_rate": 4.979216227913605e-06,
+      "loss": 0.601,
+      "step": 2529
+    },
+    {
+      "epoch": 0.9997530254383798,
+      "grad_norm": 0.4731512690482939,
+      "learning_rate": 4.9791960628446935e-06,
+      "loss": 0.6229,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0003951592985922,
+      "grad_norm": 0.9039832839745028,
+      "learning_rate": 4.979175888039005e-06,
+      "loss": 0.5558,
+      "step": 2531
+    },
+    {
+      "epoch": 1.0007903185971845,
+      "grad_norm": 2.641950999929186,
+      "learning_rate": 4.979155703496619e-06,
+      "loss": 0.5752,
+      "step": 2532
+    },
+    {
+      "epoch": 1.0011854778957767,
+      "grad_norm": 1.076378753297078,
+      "learning_rate": 4.979135509217615e-06,
+      "loss": 0.5912,
+      "step": 2533
+    },
+    {
+      "epoch": 1.001580637194369,
+      "grad_norm": 0.8858294098499361,
+      "learning_rate": 4.979115305202073e-06,
+      "loss": 0.5723,
+      "step": 2534
+    },
+    {
+      "epoch": 1.0019757964929612,
+      "grad_norm": 0.9699126038857411,
+      "learning_rate": 4.9790950914500705e-06,
+      "loss": 0.5786,
+      "step": 2535
+    },
+    {
+      "epoch": 1.0023709557915534,
+      "grad_norm": 0.6877025575827704,
+      "learning_rate": 4.979074867961687e-06,
+      "loss": 0.5676,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0027661150901457,
+      "grad_norm": 0.6507721272783759,
+      "learning_rate": 4.979054634737004e-06,
+      "loss": 0.5813,
+      "step": 2537
+    },
+    {
+      "epoch": 1.003161274388738,
+      "grad_norm": 0.7752079766225948,
+      "learning_rate": 4.9790343917761e-06,
+      "loss": 0.5741,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0035564336873302,
+      "grad_norm": 0.728864402953281,
+      "learning_rate": 4.979014139079053e-06,
+      "loss": 0.5594,
+      "step": 2539
+    },
+    {
+      "epoch": 1.0039515929859224,
+      "grad_norm": 0.5774709293399436,
+      "learning_rate": 4.978993876645945e-06,
+      "loss": 0.5635,
+      "step": 2540
+    },
+    {
+      "epoch": 1.0043467522845146,
+      "grad_norm": 0.5686297060429989,
+      "learning_rate": 4.978973604476855e-06,
+      "loss": 0.5701,
+      "step": 2541
+    },
+    {
+      "epoch": 1.0047419115831069,
+      "grad_norm": 0.5887544543582653,
+      "learning_rate": 4.97895332257186e-06,
+      "loss": 0.5632,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0051370708816991,
+      "grad_norm": 0.6127637272155544,
+      "learning_rate": 4.978933030931042e-06,
+      "loss": 0.5594,
+      "step": 2543
+    },
+    {
+      "epoch": 1.0055322301802914,
+      "grad_norm": 0.5483995865865325,
+      "learning_rate": 4.978912729554481e-06,
+      "loss": 0.5557,
+      "step": 2544
+    },
+    {
+      "epoch": 1.0059273894788836,
+      "grad_norm": 0.5341010434760727,
+      "learning_rate": 4.978892418442256e-06,
+      "loss": 0.5553,
+      "step": 2545
+    },
+    {
+      "epoch": 1.0063225487774758,
+      "grad_norm": 0.5396189198412058,
+      "learning_rate": 4.9788720975944475e-06,
+      "loss": 0.572,
+      "step": 2546
+    },
+    {
+      "epoch": 1.006717708076068,
+      "grad_norm": 0.5603122829844251,
+      "learning_rate": 4.978851767011134e-06,
+      "loss": 0.554,
+      "step": 2547
+    },
+    {
+      "epoch": 1.0071128673746603,
+      "grad_norm": 0.5973006671495175,
+      "learning_rate": 4.978831426692397e-06,
+      "loss": 0.573,
+      "step": 2548
+    },
+    {
+      "epoch": 1.0075080266732526,
+      "grad_norm": 0.5537555552956865,
+      "learning_rate": 4.9788110766383135e-06,
+      "loss": 0.5725,
+      "step": 2549
+    },
+    {
+      "epoch": 1.0079031859718448,
+      "grad_norm": 0.506831024700252,
+      "learning_rate": 4.978790716848966e-06,
+      "loss": 0.5605,
+      "step": 2550
+    },
+    {
+      "epoch": 1.008298345270437,
+      "grad_norm": 0.5146187226178434,
+      "learning_rate": 4.9787703473244346e-06,
+      "loss": 0.5725,
+      "step": 2551
+    },
+    {
+      "epoch": 1.0086935045690293,
+      "grad_norm": 0.584671063878508,
+      "learning_rate": 4.9787499680647975e-06,
+      "loss": 0.5535,
+      "step": 2552
+    },
+    {
+      "epoch": 1.0090886638676215,
+      "grad_norm": 0.517272250939791,
+      "learning_rate": 4.978729579070136e-06,
+      "loss": 0.5694,
+      "step": 2553
+    },
+    {
+      "epoch": 1.009483823166214,
+      "grad_norm": 0.5197158410846429,
+      "learning_rate": 4.978709180340529e-06,
+      "loss": 0.5703,
+      "step": 2554
+    },
+    {
+      "epoch": 1.0098789824648062,
+      "grad_norm": 0.5316647592420108,
+      "learning_rate": 4.978688771876059e-06,
+      "loss": 0.5614,
+      "step": 2555
+    },
+    {
+      "epoch": 1.0102741417633985,
+      "grad_norm": 0.47946328549941336,
+      "learning_rate": 4.978668353676804e-06,
+      "loss": 0.574,
+      "step": 2556
+    },
+    {
+      "epoch": 1.0106693010619907,
+      "grad_norm": 0.4799139075965923,
+      "learning_rate": 4.9786479257428455e-06,
+      "loss": 0.5638,
+      "step": 2557
+    },
+    {
+      "epoch": 1.011064460360583,
+      "grad_norm": 0.5034447217830896,
+      "learning_rate": 4.978627488074263e-06,
+      "loss": 0.576,
+      "step": 2558
+    },
+    {
+      "epoch": 1.0114596196591752,
+      "grad_norm": 0.5176884626827072,
+      "learning_rate": 4.978607040671136e-06,
+      "loss": 0.5636,
+      "step": 2559
+    },
+    {
+      "epoch": 1.0118547789577674,
+      "grad_norm": 0.5148490669814971,
+      "learning_rate": 4.978586583533545e-06,
+      "loss": 0.5838,
+      "step": 2560
+    },
+    {
+      "epoch": 1.0122499382563597,
+      "grad_norm": 0.4542757059816822,
+      "learning_rate": 4.978566116661573e-06,
+      "loss": 0.5532,
+      "step": 2561
+    },
+    {
+      "epoch": 1.012645097554952,
+      "grad_norm": 0.4811513221663893,
+      "learning_rate": 4.978545640055297e-06,
+      "loss": 0.5658,
+      "step": 2562
+    },
+    {
+      "epoch": 1.0130402568535442,
+      "grad_norm": 0.500097716390493,
+      "learning_rate": 4.978525153714799e-06,
+      "loss": 0.5744,
+      "step": 2563
+    },
+    {
+      "epoch": 1.0134354161521364,
+      "grad_norm": 0.5142140734871467,
+      "learning_rate": 4.978504657640159e-06,
+      "loss": 0.5538,
+      "step": 2564
+    },
+    {
+      "epoch": 1.0138305754507286,
+      "grad_norm": 0.47546571996143905,
+      "learning_rate": 4.978484151831458e-06,
+      "loss": 0.5587,
+      "step": 2565
+    },
+    {
+      "epoch": 1.0142257347493209,
+      "grad_norm": 0.472451374021239,
+      "learning_rate": 4.978463636288776e-06,
+      "loss": 0.5881,
+      "step": 2566
+    },
+    {
+      "epoch": 1.0146208940479131,
+      "grad_norm": 0.4541293970758113,
+      "learning_rate": 4.978443111012195e-06,
+      "loss": 0.5572,
+      "step": 2567
+    },
+    {
+      "epoch": 1.0150160533465054,
+      "grad_norm": 0.47790081006874097,
+      "learning_rate": 4.978422576001793e-06,
+      "loss": 0.552,
+      "step": 2568
+    },
+    {
+      "epoch": 1.0154112126450976,
+      "grad_norm": 0.4983815727839767,
+      "learning_rate": 4.978402031257653e-06,
+      "loss": 0.5655,
+      "step": 2569
+    },
+    {
+      "epoch": 1.0158063719436898,
+      "grad_norm": 0.484727652450049,
+      "learning_rate": 4.9783814767798545e-06,
+      "loss": 0.5726,
+      "step": 2570
+    },
+    {
+      "epoch": 1.016201531242282,
+      "grad_norm": 0.4852767066955579,
+      "learning_rate": 4.978360912568479e-06,
+      "loss": 0.5537,
+      "step": 2571
+    },
+    {
+      "epoch": 1.0165966905408743,
+      "grad_norm": 0.4485626151233385,
+      "learning_rate": 4.978340338623606e-06,
+      "loss": 0.5555,
+      "step": 2572
+    },
+    {
+      "epoch": 1.0169918498394666,
+      "grad_norm": 0.5060512578748023,
+      "learning_rate": 4.9783197549453164e-06,
+      "loss": 0.5898,
+      "step": 2573
+    },
+    {
+      "epoch": 1.0173870091380588,
+      "grad_norm": 0.4646178927626715,
+      "learning_rate": 4.978299161533693e-06,
+      "loss": 0.5744,
+      "step": 2574
+    },
+    {
+      "epoch": 1.017782168436651,
+      "grad_norm": 0.461846845018728,
+      "learning_rate": 4.978278558388815e-06,
+      "loss": 0.5654,
+      "step": 2575
+    },
+    {
+      "epoch": 1.0181773277352433,
+      "grad_norm": 0.6011786743216133,
+      "learning_rate": 4.978257945510764e-06,
+      "loss": 0.5593,
+      "step": 2576
+    },
+    {
+      "epoch": 1.0185724870338355,
+      "grad_norm": 0.523836605126601,
+      "learning_rate": 4.978237322899621e-06,
+      "loss": 0.572,
+      "step": 2577
+    },
+    {
+      "epoch": 1.0189676463324278,
+      "grad_norm": 0.4752809951230506,
+      "learning_rate": 4.978216690555465e-06,
+      "loss": 0.583,
+      "step": 2578
+    },
+    {
+      "epoch": 1.01936280563102,
+      "grad_norm": 0.45433798164574274,
+      "learning_rate": 4.978196048478381e-06,
+      "loss": 0.568,
+      "step": 2579
+    },
+    {
+      "epoch": 1.0197579649296122,
+      "grad_norm": 0.46744153312914954,
+      "learning_rate": 4.9781753966684455e-06,
+      "loss": 0.5526,
+      "step": 2580
+    },
+    {
+      "epoch": 1.0201531242282045,
+      "grad_norm": 0.47136324904331967,
+      "learning_rate": 4.978154735125743e-06,
+      "loss": 0.5734,
+      "step": 2581
+    },
+    {
+      "epoch": 1.0205482835267967,
+      "grad_norm": 0.476581814748425,
+      "learning_rate": 4.9781340638503536e-06,
+      "loss": 0.596,
+      "step": 2582
+    },
+    {
+      "epoch": 1.020943442825389,
+      "grad_norm": 0.4541300104338508,
+      "learning_rate": 4.9781133828423585e-06,
+      "loss": 0.5553,
+      "step": 2583
+    },
+    {
+      "epoch": 1.0213386021239812,
+      "grad_norm": 0.45621750533610406,
+      "learning_rate": 4.978092692101838e-06,
+      "loss": 0.5561,
+      "step": 2584
+    },
+    {
+      "epoch": 1.0217337614225734,
+      "grad_norm": 0.47098693586154,
+      "learning_rate": 4.978071991628875e-06,
+      "loss": 0.5524,
+      "step": 2585
+    },
+    {
+      "epoch": 1.0221289207211657,
+      "grad_norm": 0.4259588250799944,
+      "learning_rate": 4.97805128142355e-06,
+      "loss": 0.5235,
+      "step": 2586
+    },
+    {
+      "epoch": 1.022524080019758,
+      "grad_norm": 0.4686581079113878,
+      "learning_rate": 4.9780305614859435e-06,
+      "loss": 0.5574,
+      "step": 2587
+    },
+    {
+      "epoch": 1.0229192393183502,
+      "grad_norm": 0.46911512358386387,
+      "learning_rate": 4.9780098318161385e-06,
+      "loss": 0.5777,
+      "step": 2588
+    },
+    {
+      "epoch": 1.0233143986169424,
+      "grad_norm": 0.4642643584963139,
+      "learning_rate": 4.9779890924142155e-06,
+      "loss": 0.5726,
+      "step": 2589
+    },
+    {
+      "epoch": 1.0237095579155346,
+      "grad_norm": 0.4712873655824191,
+      "learning_rate": 4.977968343280256e-06,
+      "loss": 0.5823,
+      "step": 2590
+    },
+    {
+      "epoch": 1.0241047172141269,
+      "grad_norm": 0.4482243472573749,
+      "learning_rate": 4.977947584414341e-06,
+      "loss": 0.564,
+      "step": 2591
+    },
+    {
+      "epoch": 1.0244998765127191,
+      "grad_norm": 0.4530647693811047,
+      "learning_rate": 4.977926815816553e-06,
+      "loss": 0.5809,
+      "step": 2592
+    },
+    {
+      "epoch": 1.0248950358113114,
+      "grad_norm": 0.4479882772757105,
+      "learning_rate": 4.977906037486974e-06,
+      "loss": 0.5616,
+      "step": 2593
+    },
+    {
+      "epoch": 1.0252901951099036,
+      "grad_norm": 0.4610471046625982,
+      "learning_rate": 4.977885249425684e-06,
+      "loss": 0.5472,
+      "step": 2594
+    },
+    {
+      "epoch": 1.0256853544084958,
+      "grad_norm": 0.47862490507477046,
+      "learning_rate": 4.977864451632764e-06,
+      "loss": 0.5902,
+      "step": 2595
+    },
+    {
+      "epoch": 1.026080513707088,
+      "grad_norm": 0.5047546461386095,
+      "learning_rate": 4.977843644108299e-06,
+      "loss": 0.5884,
+      "step": 2596
+    },
+    {
+      "epoch": 1.0264756730056803,
+      "grad_norm": 0.4451423123413334,
+      "learning_rate": 4.977822826852369e-06,
+      "loss": 0.5642,
+      "step": 2597
+    },
+    {
+      "epoch": 1.0268708323042726,
+      "grad_norm": 0.4560284371925842,
+      "learning_rate": 4.977801999865054e-06,
+      "loss": 0.5606,
+      "step": 2598
+    },
+    {
+      "epoch": 1.0272659916028648,
+      "grad_norm": 0.4461067403640707,
+      "learning_rate": 4.977781163146438e-06,
+      "loss": 0.5649,
+      "step": 2599
+    },
+    {
+      "epoch": 1.027661150901457,
+      "grad_norm": 0.451597159891192,
+      "learning_rate": 4.977760316696603e-06,
+      "loss": 0.5634,
+      "step": 2600
+    },
+    {
+      "epoch": 1.0280563102000495,
+      "grad_norm": 0.4519191659487248,
+      "learning_rate": 4.977739460515629e-06,
+      "loss": 0.5796,
+      "step": 2601
+    },
+    {
+      "epoch": 1.0284514694986417,
+      "grad_norm": 0.4596070611618375,
+      "learning_rate": 4.977718594603599e-06,
+      "loss": 0.5764,
+      "step": 2602
+    },
+    {
+      "epoch": 1.028846628797234,
+      "grad_norm": 0.440619261667536,
+      "learning_rate": 4.977697718960595e-06,
+      "loss": 0.557,
+      "step": 2603
+    },
+    {
+      "epoch": 1.0292417880958262,
+      "grad_norm": 0.4585810235456444,
+      "learning_rate": 4.977676833586699e-06,
+      "loss": 0.5618,
+      "step": 2604
+    },
+    {
+      "epoch": 1.0296369473944185,
+      "grad_norm": 0.489223887277486,
+      "learning_rate": 4.977655938481994e-06,
+      "loss": 0.6054,
+      "step": 2605
+    },
+    {
+      "epoch": 1.0300321066930107,
+      "grad_norm": 0.4590644227459909,
+      "learning_rate": 4.97763503364656e-06,
+      "loss": 0.5575,
+      "step": 2606
+    },
+    {
+      "epoch": 1.030427265991603,
+      "grad_norm": 0.4676586979127894,
+      "learning_rate": 4.97761411908048e-06,
+      "loss": 0.5715,
+      "step": 2607
+    },
+    {
+      "epoch": 1.0308224252901952,
+      "grad_norm": 0.44950227149153915,
+      "learning_rate": 4.977593194783836e-06,
+      "loss": 0.5473,
+      "step": 2608
+    },
+    {
+      "epoch": 1.0312175845887874,
+      "grad_norm": 0.44607152531853783,
+      "learning_rate": 4.977572260756711e-06,
+      "loss": 0.5318,
+      "step": 2609
+    },
+    {
+      "epoch": 1.0316127438873797,
+      "grad_norm": 0.45941977696779834,
+      "learning_rate": 4.977551316999186e-06,
+      "loss": 0.5453,
+      "step": 2610
+    },
+    {
+      "epoch": 1.032007903185972,
+      "grad_norm": 0.4624807401390588,
+      "learning_rate": 4.977530363511344e-06,
+      "loss": 0.5721,
+      "step": 2611
+    },
+    {
+      "epoch": 1.0324030624845641,
+      "grad_norm": 0.4486414918582425,
+      "learning_rate": 4.977509400293268e-06,
+      "loss": 0.5634,
+      "step": 2612
+    },
+    {
+      "epoch": 1.0327982217831564,
+      "grad_norm": 0.460795352195087,
+      "learning_rate": 4.977488427345039e-06,
+      "loss": 0.5663,
+      "step": 2613
+    },
+    {
+      "epoch": 1.0331933810817486,
+      "grad_norm": 0.47567928790908864,
+      "learning_rate": 4.977467444666739e-06,
+      "loss": 0.586,
+      "step": 2614
+    },
+    {
+      "epoch": 1.0335885403803409,
+      "grad_norm": 0.4486438855377463,
+      "learning_rate": 4.977446452258452e-06,
+      "loss": 0.5904,
+      "step": 2615
+    },
+    {
+      "epoch": 1.033983699678933,
+      "grad_norm": 0.4415373719504563,
+      "learning_rate": 4.97742545012026e-06,
+      "loss": 0.5523,
+      "step": 2616
+    },
+    {
+      "epoch": 1.0343788589775254,
+      "grad_norm": 0.46461856034519633,
+      "learning_rate": 4.977404438252245e-06,
+      "loss": 0.556,
+      "step": 2617
+    },
+    {
+      "epoch": 1.0347740182761176,
+      "grad_norm": 0.45898412914751924,
+      "learning_rate": 4.97738341665449e-06,
+      "loss": 0.5868,
+      "step": 2618
+    },
+    {
+      "epoch": 1.0351691775747098,
+      "grad_norm": 0.44026432413454175,
+      "learning_rate": 4.977362385327077e-06,
+      "loss": 0.5399,
+      "step": 2619
+    },
+    {
+      "epoch": 1.035564336873302,
+      "grad_norm": 0.4536936643831469,
+      "learning_rate": 4.977341344270088e-06,
+      "loss": 0.5572,
+      "step": 2620
+    },
+    {
+      "epoch": 1.0359594961718943,
+      "grad_norm": 0.45300210621077247,
+      "learning_rate": 4.977320293483608e-06,
+      "loss": 0.5581,
+      "step": 2621
+    },
+    {
+      "epoch": 1.0363546554704866,
+      "grad_norm": 0.4768261944797516,
+      "learning_rate": 4.977299232967717e-06,
+      "loss": 0.5567,
+      "step": 2622
+    },
+    {
+      "epoch": 1.0367498147690788,
+      "grad_norm": 0.4490373651630668,
+      "learning_rate": 4.9772781627225e-06,
+      "loss": 0.5613,
+      "step": 2623
+    },
+    {
+      "epoch": 1.037144974067671,
+      "grad_norm": 0.4542757319001758,
+      "learning_rate": 4.977257082748038e-06,
+      "loss": 0.5598,
+      "step": 2624
+    },
+    {
+      "epoch": 1.0375401333662633,
+      "grad_norm": 0.4696319591373662,
+      "learning_rate": 4.977235993044415e-06,
+      "loss": 0.574,
+      "step": 2625
+    },
+    {
+      "epoch": 1.0379352926648555,
+      "grad_norm": 0.4544516411565146,
+      "learning_rate": 4.977214893611713e-06,
+      "loss": 0.5647,
+      "step": 2626
+    },
+    {
+      "epoch": 1.0383304519634478,
+      "grad_norm": 0.4636968131984274,
+      "learning_rate": 4.977193784450015e-06,
+      "loss": 0.5776,
+      "step": 2627
+    },
+    {
+      "epoch": 1.03872561126204,
+      "grad_norm": 0.4549508464045885,
+      "learning_rate": 4.977172665559403e-06,
+      "loss": 0.5721,
+      "step": 2628
+    },
+    {
+      "epoch": 1.0391207705606322,
+      "grad_norm": 0.4432377804345881,
+      "learning_rate": 4.9771515369399625e-06,
+      "loss": 0.5624,
+      "step": 2629
+    },
+    {
+      "epoch": 1.0395159298592245,
+      "grad_norm": 0.4512832309113786,
+      "learning_rate": 4.977130398591775e-06,
+      "loss": 0.5793,
+      "step": 2630
+    },
+    {
+      "epoch": 1.0399110891578167,
+      "grad_norm": 0.44957500296377784,
+      "learning_rate": 4.977109250514923e-06,
+      "loss": 0.5649,
+      "step": 2631
+    },
+    {
+      "epoch": 1.040306248456409,
+      "grad_norm": 0.45057963341150464,
+      "learning_rate": 4.97708809270949e-06,
+      "loss": 0.5607,
+      "step": 2632
+    },
+    {
+      "epoch": 1.0407014077550012,
+      "grad_norm": 0.4542174831876239,
+      "learning_rate": 4.977066925175559e-06,
+      "loss": 0.5539,
+      "step": 2633
+    },
+    {
+      "epoch": 1.0410965670535934,
+      "grad_norm": 0.46807162702065624,
+      "learning_rate": 4.977045747913213e-06,
+      "loss": 0.563,
+      "step": 2634
+    },
+    {
+      "epoch": 1.0414917263521857,
+      "grad_norm": 0.457076849040818,
+      "learning_rate": 4.977024560922537e-06,
+      "loss": 0.561,
+      "step": 2635
+    },
+    {
+      "epoch": 1.041886885650778,
+      "grad_norm": 0.45562089282056256,
+      "learning_rate": 4.9770033642036105e-06,
+      "loss": 0.5779,
+      "step": 2636
+    },
+    {
+      "epoch": 1.0422820449493702,
+      "grad_norm": 0.4525510725985377,
+      "learning_rate": 4.97698215775652e-06,
+      "loss": 0.6033,
+      "step": 2637
+    },
+    {
+      "epoch": 1.0426772042479624,
+      "grad_norm": 0.5400191406643862,
+      "learning_rate": 4.976960941581348e-06,
+      "loss": 0.5704,
+      "step": 2638
+    },
+    {
+      "epoch": 1.0430723635465546,
+      "grad_norm": 0.44608237054831834,
+      "learning_rate": 4.976939715678178e-06,
+      "loss": 0.5611,
+      "step": 2639
+    },
+    {
+      "epoch": 1.0434675228451469,
+      "grad_norm": 0.5052264389295701,
+      "learning_rate": 4.9769184800470915e-06,
+      "loss": 0.5755,
+      "step": 2640
+    },
+    {
+      "epoch": 1.0438626821437391,
+      "grad_norm": 0.44348147340006067,
+      "learning_rate": 4.976897234688174e-06,
+      "loss": 0.5598,
+      "step": 2641
+    },
+    {
+      "epoch": 1.0442578414423314,
+      "grad_norm": 0.451998380744831,
+      "learning_rate": 4.976875979601508e-06,
+      "loss": 0.5866,
+      "step": 2642
+    },
+    {
+      "epoch": 1.0446530007409236,
+      "grad_norm": 0.4450349262521982,
+      "learning_rate": 4.976854714787177e-06,
+      "loss": 0.5612,
+      "step": 2643
+    },
+    {
+      "epoch": 1.0450481600395158,
+      "grad_norm": 0.4554004140171642,
+      "learning_rate": 4.976833440245265e-06,
+      "loss": 0.5752,
+      "step": 2644
+    },
+    {
+      "epoch": 1.045443319338108,
+      "grad_norm": 0.4717355641847888,
+      "learning_rate": 4.976812155975855e-06,
+      "loss": 0.5696,
+      "step": 2645
+    },
+    {
+      "epoch": 1.0458384786367003,
+      "grad_norm": 0.451389235554898,
+      "learning_rate": 4.976790861979031e-06,
+      "loss": 0.5896,
+      "step": 2646
+    },
+    {
+      "epoch": 1.0462336379352926,
+      "grad_norm": 0.4587582612615821,
+      "learning_rate": 4.976769558254877e-06,
+      "loss": 0.5905,
+      "step": 2647
+    },
+    {
+      "epoch": 1.0466287972338848,
+      "grad_norm": 0.4587087573396345,
+      "learning_rate": 4.976748244803475e-06,
+      "loss": 0.5823,
+      "step": 2648
+    },
+    {
+      "epoch": 1.0470239565324773,
+      "grad_norm": 0.4477959820273805,
+      "learning_rate": 4.976726921624911e-06,
+      "loss": 0.5657,
+      "step": 2649
+    },
+    {
+      "epoch": 1.0474191158310695,
+      "grad_norm": 0.43767555375829137,
+      "learning_rate": 4.976705588719267e-06,
+      "loss": 0.5649,
+      "step": 2650
+    },
+    {
+      "epoch": 1.0478142751296617,
+      "grad_norm": 0.453863897792215,
+      "learning_rate": 4.976684246086627e-06,
+      "loss": 0.563,
+      "step": 2651
+    },
+    {
+      "epoch": 1.048209434428254,
+      "grad_norm": 0.44586538451216967,
+      "learning_rate": 4.976662893727076e-06,
+      "loss": 0.5541,
+      "step": 2652
+    },
+    {
+      "epoch": 1.0486045937268462,
+      "grad_norm": 0.46349135729431057,
+      "learning_rate": 4.9766415316406965e-06,
+      "loss": 0.5621,
+      "step": 2653
+    },
+    {
+      "epoch": 1.0489997530254385,
+      "grad_norm": 0.4485724204935941,
+      "learning_rate": 4.976620159827574e-06,
+      "loss": 0.5465,
+      "step": 2654
+    },
+    {
+      "epoch": 1.0493949123240307,
+      "grad_norm": 0.4388668643967193,
+      "learning_rate": 4.97659877828779e-06,
+      "loss": 0.5592,
+      "step": 2655
+    },
+    {
+      "epoch": 1.049790071622623,
+      "grad_norm": 0.46226835952230305,
+      "learning_rate": 4.976577387021431e-06,
+      "loss": 0.5814,
+      "step": 2656
+    },
+    {
+      "epoch": 1.0501852309212152,
+      "grad_norm": 0.4551414886420647,
+      "learning_rate": 4.97655598602858e-06,
+      "loss": 0.5489,
+      "step": 2657
+    },
+    {
+      "epoch": 1.0505803902198074,
+      "grad_norm": 0.4503034545836757,
+      "learning_rate": 4.97653457530932e-06,
+      "loss": 0.5847,
+      "step": 2658
+    },
+    {
+      "epoch": 1.0509755495183997,
+      "grad_norm": 0.4510190880823167,
+      "learning_rate": 4.976513154863735e-06,
+      "loss": 0.5638,
+      "step": 2659
+    },
+    {
+      "epoch": 1.051370708816992,
+      "grad_norm": 0.45601750337175206,
+      "learning_rate": 4.976491724691912e-06,
+      "loss": 0.5629,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0517658681155841,
+      "grad_norm": 0.44161293110782124,
+      "learning_rate": 4.976470284793933e-06,
+      "loss": 0.5638,
+      "step": 2661
+    },
+    {
+      "epoch": 1.0521610274141764,
+      "grad_norm": 0.4971279919030968,
+      "learning_rate": 4.976448835169882e-06,
+      "loss": 0.5602,
+      "step": 2662
+    },
+    {
+      "epoch": 1.0525561867127686,
+      "grad_norm": 0.46201452125116643,
+      "learning_rate": 4.976427375819844e-06,
+      "loss": 0.5645,
+      "step": 2663
+    },
+    {
+      "epoch": 1.0529513460113609,
+      "grad_norm": 0.45839768743558873,
+      "learning_rate": 4.9764059067439045e-06,
+      "loss": 0.5729,
+      "step": 2664
+    },
+    {
+      "epoch": 1.053346505309953,
+      "grad_norm": 0.45571009890945263,
+      "learning_rate": 4.9763844279421444e-06,
+      "loss": 0.5669,
+      "step": 2665
+    },
+    {
+      "epoch": 1.0537416646085453,
+      "grad_norm": 0.4404670408454727,
+      "learning_rate": 4.97636293941465e-06,
+      "loss": 0.5537,
+      "step": 2666
+    },
+    {
+      "epoch": 1.0541368239071376,
+      "grad_norm": 0.4618688908161834,
+      "learning_rate": 4.976341441161507e-06,
+      "loss": 0.5666,
+      "step": 2667
+    },
+    {
+      "epoch": 1.0545319832057298,
+      "grad_norm": 0.4678051099127983,
+      "learning_rate": 4.976319933182797e-06,
+      "loss": 0.5852,
+      "step": 2668
+    },
+    {
+      "epoch": 1.054927142504322,
+      "grad_norm": 0.45030359679512383,
+      "learning_rate": 4.9762984154786075e-06,
+      "loss": 0.5749,
+      "step": 2669
+    },
+    {
+      "epoch": 1.0553223018029143,
+      "grad_norm": 0.4553020196001742,
+      "learning_rate": 4.9762768880490205e-06,
+      "loss": 0.5595,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0557174611015065,
+      "grad_norm": 0.45256331743406897,
+      "learning_rate": 4.976255350894122e-06,
+      "loss": 0.5635,
+      "step": 2671
+    },
+    {
+      "epoch": 1.0561126204000988,
+      "grad_norm": 0.44922321136235305,
+      "learning_rate": 4.9762338040139965e-06,
+      "loss": 0.5488,
+      "step": 2672
+    },
+    {
+      "epoch": 1.056507779698691,
+      "grad_norm": 0.45045666513262506,
+      "learning_rate": 4.976212247408727e-06,
+      "loss": 0.5595,
+      "step": 2673
+    },
+    {
+      "epoch": 1.0569029389972833,
+      "grad_norm": 0.45788018345436676,
+      "learning_rate": 4.976190681078401e-06,
+      "loss": 0.5693,
+      "step": 2674
+    },
+    {
+      "epoch": 1.0572980982958755,
+      "grad_norm": 0.48385791595148814,
+      "learning_rate": 4.976169105023101e-06,
+      "loss": 0.5967,
+      "step": 2675
+    },
+    {
+      "epoch": 1.0576932575944678,
+      "grad_norm": 0.45899699704778635,
+      "learning_rate": 4.976147519242912e-06,
+      "loss": 0.5755,
+      "step": 2676
+    },
+    {
+      "epoch": 1.05808841689306,
+      "grad_norm": 0.44447734769605085,
+      "learning_rate": 4.976125923737919e-06,
+      "loss": 0.5678,
+      "step": 2677
+    },
+    {
+      "epoch": 1.0584835761916522,
+      "grad_norm": 0.4578877530549744,
+      "learning_rate": 4.976104318508207e-06,
+      "loss": 0.5781,
+      "step": 2678
+    },
+    {
+      "epoch": 1.0588787354902445,
+      "grad_norm": 0.4522345638325663,
+      "learning_rate": 4.9760827035538614e-06,
+      "loss": 0.5733,
+      "step": 2679
+    },
+    {
+      "epoch": 1.0592738947888367,
+      "grad_norm": 0.45242325394735017,
+      "learning_rate": 4.9760610788749665e-06,
+      "loss": 0.5781,
+      "step": 2680
+    },
+    {
+      "epoch": 1.059669054087429,
+      "grad_norm": 0.45775529697399175,
+      "learning_rate": 4.976039444471607e-06,
+      "loss": 0.5627,
+      "step": 2681
+    },
+    {
+      "epoch": 1.0600642133860212,
+      "grad_norm": 0.44249752472475534,
+      "learning_rate": 4.976017800343868e-06,
+      "loss": 0.5471,
+      "step": 2682
+    },
+    {
+      "epoch": 1.0604593726846134,
+      "grad_norm": 0.47201137623979633,
+      "learning_rate": 4.975996146491834e-06,
+      "loss": 0.5786,
+      "step": 2683
+    },
+    {
+      "epoch": 1.0608545319832057,
+      "grad_norm": 0.4775255993907024,
+      "learning_rate": 4.975974482915592e-06,
+      "loss": 0.5936,
+      "step": 2684
+    },
+    {
+      "epoch": 1.061249691281798,
+      "grad_norm": 0.44639736931494717,
+      "learning_rate": 4.975952809615225e-06,
+      "loss": 0.5579,
+      "step": 2685
+    },
+    {
+      "epoch": 1.0616448505803902,
+      "grad_norm": 0.44958051847488756,
+      "learning_rate": 4.975931126590819e-06,
+      "loss": 0.5706,
+      "step": 2686
+    },
+    {
+      "epoch": 1.0620400098789824,
+      "grad_norm": 0.4507031529753823,
+      "learning_rate": 4.975909433842459e-06,
+      "loss": 0.557,
+      "step": 2687
+    },
+    {
+      "epoch": 1.0624351691775746,
+      "grad_norm": 0.45147386789174,
+      "learning_rate": 4.975887731370229e-06,
+      "loss": 0.5561,
+      "step": 2688
+    },
+    {
+      "epoch": 1.0628303284761669,
+      "grad_norm": 0.45782580821366853,
+      "learning_rate": 4.975866019174217e-06,
+      "loss": 0.5511,
+      "step": 2689
+    },
+    {
+      "epoch": 1.0632254877747591,
+      "grad_norm": 0.4707517108424949,
+      "learning_rate": 4.975844297254506e-06,
+      "loss": 0.5485,
+      "step": 2690
+    },
+    {
+      "epoch": 1.0636206470733514,
+      "grad_norm": 0.451561493525649,
+      "learning_rate": 4.975822565611183e-06,
+      "loss": 0.5615,
+      "step": 2691
+    },
+    {
+      "epoch": 1.0640158063719436,
+      "grad_norm": 0.4669716121243911,
+      "learning_rate": 4.975800824244331e-06,
+      "loss": 0.5676,
+      "step": 2692
+    },
+    {
+      "epoch": 1.0644109656705358,
+      "grad_norm": 0.4657398109153327,
+      "learning_rate": 4.975779073154038e-06,
+      "loss": 0.5609,
+      "step": 2693
+    },
+    {
+      "epoch": 1.064806124969128,
+      "grad_norm": 0.4577124290606098,
+      "learning_rate": 4.975757312340387e-06,
+      "loss": 0.5649,
+      "step": 2694
+    },
+    {
+      "epoch": 1.0652012842677205,
+      "grad_norm": 0.459041880221058,
+      "learning_rate": 4.975735541803465e-06,
+      "loss": 0.5649,
+      "step": 2695
+    },
+    {
+      "epoch": 1.0655964435663128,
+      "grad_norm": 0.4665780766862844,
+      "learning_rate": 4.975713761543357e-06,
+      "loss": 0.5837,
+      "step": 2696
+    },
+    {
+      "epoch": 1.065991602864905,
+      "grad_norm": 0.45328946916325785,
+      "learning_rate": 4.975691971560149e-06,
+      "loss": 0.5687,
+      "step": 2697
+    },
+    {
+      "epoch": 1.0663867621634973,
+      "grad_norm": 0.46683872005405097,
+      "learning_rate": 4.975670171853926e-06,
+      "loss": 0.564,
+      "step": 2698
+    },
+    {
+      "epoch": 1.0667819214620895,
+      "grad_norm": 0.4490782215754304,
+      "learning_rate": 4.9756483624247745e-06,
+      "loss": 0.5694,
+      "step": 2699
+    },
+    {
+      "epoch": 1.0671770807606817,
+      "grad_norm": 0.4602344049001532,
+      "learning_rate": 4.975626543272779e-06,
+      "loss": 0.5829,
+      "step": 2700
+    },
+    {
+      "epoch": 1.067572240059274,
+      "grad_norm": 0.47778949312409547,
+      "learning_rate": 4.975604714398026e-06,
+      "loss": 0.5849,
+      "step": 2701
+    },
+    {
+      "epoch": 1.0679673993578662,
+      "grad_norm": 0.4766565902077033,
+      "learning_rate": 4.9755828758006e-06,
+      "loss": 0.5803,
+      "step": 2702
+    },
+    {
+      "epoch": 1.0683625586564585,
+      "grad_norm": 0.46556612189369995,
+      "learning_rate": 4.975561027480589e-06,
+      "loss": 0.565,
+      "step": 2703
+    },
+    {
+      "epoch": 1.0687577179550507,
+      "grad_norm": 0.4681074034544366,
+      "learning_rate": 4.975539169438077e-06,
+      "loss": 0.5826,
+      "step": 2704
+    },
+    {
+      "epoch": 1.069152877253643,
+      "grad_norm": 0.4638743898250165,
+      "learning_rate": 4.975517301673151e-06,
+      "loss": 0.5734,
+      "step": 2705
+    },
+    {
+      "epoch": 1.0695480365522352,
+      "grad_norm": 0.4791989380557464,
+      "learning_rate": 4.975495424185895e-06,
+      "loss": 0.5706,
+      "step": 2706
+    },
+    {
+      "epoch": 1.0699431958508274,
+      "grad_norm": 0.4648061747202406,
+      "learning_rate": 4.975473536976397e-06,
+      "loss": 0.5613,
+      "step": 2707
+    },
+    {
+      "epoch": 1.0703383551494197,
+      "grad_norm": 0.4758277574553136,
+      "learning_rate": 4.975451640044742e-06,
+      "loss": 0.5758,
+      "step": 2708
+    },
+    {
+      "epoch": 1.070733514448012,
+      "grad_norm": 0.46226325782703626,
+      "learning_rate": 4.975429733391016e-06,
+      "loss": 0.5685,
+      "step": 2709
+    },
+    {
+      "epoch": 1.0711286737466041,
+      "grad_norm": 0.45170576388419686,
+      "learning_rate": 4.975407817015306e-06,
+      "loss": 0.5501,
+      "step": 2710
+    },
+    {
+      "epoch": 1.0715238330451964,
+      "grad_norm": 0.4587950204803136,
+      "learning_rate": 4.975385890917696e-06,
+      "loss": 0.5627,
+      "step": 2711
+    },
+    {
+      "epoch": 1.0719189923437886,
+      "grad_norm": 0.5105258657179963,
+      "learning_rate": 4.975363955098273e-06,
+      "loss": 0.5664,
+      "step": 2712
+    },
+    {
+      "epoch": 1.0723141516423809,
+      "grad_norm": 0.49413851191960967,
+      "learning_rate": 4.975342009557125e-06,
+      "loss": 0.562,
+      "step": 2713
+    },
+    {
+      "epoch": 1.072709310940973,
+      "grad_norm": 0.45320917178937237,
+      "learning_rate": 4.975320054294336e-06,
+      "loss": 0.5385,
+      "step": 2714
+    },
+    {
+      "epoch": 1.0731044702395653,
+      "grad_norm": 0.47467856876046166,
+      "learning_rate": 4.975298089309993e-06,
+      "loss": 0.5721,
+      "step": 2715
+    },
+    {
+      "epoch": 1.0734996295381576,
+      "grad_norm": 0.4600018582339907,
+      "learning_rate": 4.975276114604182e-06,
+      "loss": 0.5653,
+      "step": 2716
+    },
+    {
+      "epoch": 1.0738947888367498,
+      "grad_norm": 0.4708300385350632,
+      "learning_rate": 4.97525413017699e-06,
+      "loss": 0.5603,
+      "step": 2717
+    },
+    {
+      "epoch": 1.074289948135342,
+      "grad_norm": 0.4506434681952806,
+      "learning_rate": 4.975232136028502e-06,
+      "loss": 0.5381,
+      "step": 2718
+    },
+    {
+      "epoch": 1.0746851074339343,
+      "grad_norm": 0.46055573294638946,
+      "learning_rate": 4.975210132158805e-06,
+      "loss": 0.5503,
+      "step": 2719
+    },
+    {
+      "epoch": 1.0750802667325265,
+      "grad_norm": 0.5855856780535553,
+      "learning_rate": 4.975188118567987e-06,
+      "loss": 0.582,
+      "step": 2720
+    },
+    {
+      "epoch": 1.0754754260311188,
+      "grad_norm": 0.48475880634450325,
+      "learning_rate": 4.975166095256132e-06,
+      "loss": 0.5735,
+      "step": 2721
+    },
+    {
+      "epoch": 1.075870585329711,
+      "grad_norm": 0.48720473179501395,
+      "learning_rate": 4.975144062223328e-06,
+      "loss": 0.5719,
+      "step": 2722
+    },
+    {
+      "epoch": 1.0762657446283033,
+      "grad_norm": 0.4703550004447043,
+      "learning_rate": 4.9751220194696615e-06,
+      "loss": 0.6014,
+      "step": 2723
+    },
+    {
+      "epoch": 1.0766609039268955,
+      "grad_norm": 0.47236077192721254,
+      "learning_rate": 4.975099966995218e-06,
+      "loss": 0.5712,
+      "step": 2724
+    },
+    {
+      "epoch": 1.0770560632254877,
+      "grad_norm": 0.46826552794346366,
+      "learning_rate": 4.975077904800086e-06,
+      "loss": 0.5742,
+      "step": 2725
+    },
+    {
+      "epoch": 1.07745122252408,
+      "grad_norm": 0.4976523466784567,
+      "learning_rate": 4.975055832884349e-06,
+      "loss": 0.5837,
+      "step": 2726
+    },
+    {
+      "epoch": 1.0778463818226722,
+      "grad_norm": 0.5218434843940184,
+      "learning_rate": 4.975033751248096e-06,
+      "loss": 0.5422,
+      "step": 2727
+    },
+    {
+      "epoch": 1.0782415411212645,
+      "grad_norm": 0.4931023411702474,
+      "learning_rate": 4.975011659891415e-06,
+      "loss": 0.5729,
+      "step": 2728
+    },
+    {
+      "epoch": 1.0786367004198567,
+      "grad_norm": 0.5049728108245582,
+      "learning_rate": 4.974989558814389e-06,
+      "loss": 0.5665,
+      "step": 2729
+    },
+    {
+      "epoch": 1.079031859718449,
+      "grad_norm": 0.49433774891751775,
+      "learning_rate": 4.974967448017109e-06,
+      "loss": 0.5706,
+      "step": 2730
+    },
+    {
+      "epoch": 1.0794270190170412,
+      "grad_norm": 0.4620589073782666,
+      "learning_rate": 4.974945327499658e-06,
+      "loss": 0.5621,
+      "step": 2731
+    },
+    {
+      "epoch": 1.0798221783156334,
+      "grad_norm": 0.48242430406086245,
+      "learning_rate": 4.974923197262126e-06,
+      "loss": 0.5584,
+      "step": 2732
+    },
+    {
+      "epoch": 1.0802173376142257,
+      "grad_norm": 0.46908301271787106,
+      "learning_rate": 4.974901057304598e-06,
+      "loss": 0.575,
+      "step": 2733
+    },
+    {
+      "epoch": 1.080612496912818,
+      "grad_norm": 0.4635792423887607,
+      "learning_rate": 4.974878907627161e-06,
+      "loss": 0.5666,
+      "step": 2734
+    },
+    {
+      "epoch": 1.0810076562114102,
+      "grad_norm": 0.45678202788616845,
+      "learning_rate": 4.974856748229902e-06,
+      "loss": 0.5513,
+      "step": 2735
+    },
+    {
+      "epoch": 1.0814028155100024,
+      "grad_norm": 0.48311485248397,
+      "learning_rate": 4.97483457911291e-06,
+      "loss": 0.5517,
+      "step": 2736
+    },
+    {
+      "epoch": 1.0817979748085946,
+      "grad_norm": 0.4636165852005326,
+      "learning_rate": 4.97481240027627e-06,
+      "loss": 0.5597,
+      "step": 2737
+    },
+    {
+      "epoch": 1.0821931341071869,
+      "grad_norm": 0.4417447450007961,
+      "learning_rate": 4.97479021172007e-06,
+      "loss": 0.5705,
+      "step": 2738
+    },
+    {
+      "epoch": 1.0825882934057791,
+      "grad_norm": 0.4518840030411349,
+      "learning_rate": 4.974768013444395e-06,
+      "loss": 0.5653,
+      "step": 2739
+    },
+    {
+      "epoch": 1.0829834527043714,
+      "grad_norm": 1.1302182473095272,
+      "learning_rate": 4.974745805449336e-06,
+      "loss": 0.5999,
+      "step": 2740
+    },
+    {
+      "epoch": 1.0833786120029636,
+      "grad_norm": 0.4646198565639412,
+      "learning_rate": 4.974723587734977e-06,
+      "loss": 0.5814,
+      "step": 2741
+    },
+    {
+      "epoch": 1.0837737713015558,
+      "grad_norm": 0.4561905295641323,
+      "learning_rate": 4.974701360301408e-06,
+      "loss": 0.5699,
+      "step": 2742
+    },
+    {
+      "epoch": 1.084168930600148,
+      "grad_norm": 0.48160057985917565,
+      "learning_rate": 4.974679123148713e-06,
+      "loss": 0.556,
+      "step": 2743
+    },
+    {
+      "epoch": 1.0845640898987403,
+      "grad_norm": 0.4611451474715039,
+      "learning_rate": 4.974656876276983e-06,
+      "loss": 0.577,
+      "step": 2744
+    },
+    {
+      "epoch": 1.0849592491973328,
+      "grad_norm": 0.4954859328073638,
+      "learning_rate": 4.974634619686303e-06,
+      "loss": 0.5785,
+      "step": 2745
+    },
+    {
+      "epoch": 1.085354408495925,
+      "grad_norm": 0.46526474266981815,
+      "learning_rate": 4.97461235337676e-06,
+      "loss": 0.5627,
+      "step": 2746
+    },
+    {
+      "epoch": 1.0857495677945173,
+      "grad_norm": 0.47088544330789683,
+      "learning_rate": 4.974590077348442e-06,
+      "loss": 0.5735,
+      "step": 2747
+    },
+    {
+      "epoch": 1.0861447270931095,
+      "grad_norm": 0.44275077749114805,
+      "learning_rate": 4.974567791601438e-06,
+      "loss": 0.56,
+      "step": 2748
+    },
+    {
+      "epoch": 1.0865398863917017,
+      "grad_norm": 0.44909868576930495,
+      "learning_rate": 4.974545496135834e-06,
+      "loss": 0.5626,
+      "step": 2749
+    },
+    {
+      "epoch": 1.086935045690294,
+      "grad_norm": 0.4435390432235627,
+      "learning_rate": 4.974523190951718e-06,
+      "loss": 0.5489,
+      "step": 2750
+    },
+    {
+      "epoch": 1.0873302049888862,
+      "grad_norm": 0.46104078067735216,
+      "learning_rate": 4.974500876049177e-06,
+      "loss": 0.5575,
+      "step": 2751
+    },
+    {
+      "epoch": 1.0877253642874785,
+      "grad_norm": 0.47232606154854884,
+      "learning_rate": 4.9744785514283e-06,
+      "loss": 0.5636,
+      "step": 2752
+    },
+    {
+      "epoch": 1.0881205235860707,
+      "grad_norm": 0.45996006809985085,
+      "learning_rate": 4.974456217089173e-06,
+      "loss": 0.5707,
+      "step": 2753
+    },
+    {
+      "epoch": 1.088515682884663,
+      "grad_norm": 0.45991550984476076,
+      "learning_rate": 4.9744338730318846e-06,
+      "loss": 0.565,
+      "step": 2754
+    },
+    {
+      "epoch": 1.0889108421832552,
+      "grad_norm": 0.45264942810273506,
+      "learning_rate": 4.974411519256523e-06,
+      "loss": 0.5611,
+      "step": 2755
+    },
+    {
+      "epoch": 1.0893060014818474,
+      "grad_norm": 0.4603092385989538,
+      "learning_rate": 4.974389155763175e-06,
+      "loss": 0.5729,
+      "step": 2756
+    },
+    {
+      "epoch": 1.0897011607804397,
+      "grad_norm": 0.4622551865908486,
+      "learning_rate": 4.974366782551929e-06,
+      "loss": 0.5706,
+      "step": 2757
+    },
+    {
+      "epoch": 1.090096320079032,
+      "grad_norm": 0.45746013561890175,
+      "learning_rate": 4.974344399622874e-06,
+      "loss": 0.5527,
+      "step": 2758
+    },
+    {
+      "epoch": 1.0904914793776241,
+      "grad_norm": 0.4503667034204893,
+      "learning_rate": 4.974322006976095e-06,
+      "loss": 0.56,
+      "step": 2759
+    },
+    {
+      "epoch": 1.0908866386762164,
+      "grad_norm": 0.45807333150826124,
+      "learning_rate": 4.974299604611682e-06,
+      "loss": 0.5475,
+      "step": 2760
+    },
+    {
+      "epoch": 1.0912817979748086,
+      "grad_norm": 0.46938442181289014,
+      "learning_rate": 4.974277192529723e-06,
+      "loss": 0.5607,
+      "step": 2761
+    },
+    {
+      "epoch": 1.0916769572734009,
+      "grad_norm": 0.4504533242644673,
+      "learning_rate": 4.974254770730306e-06,
+      "loss": 0.5817,
+      "step": 2762
+    },
+    {
+      "epoch": 1.092072116571993,
+      "grad_norm": 0.44892490581416017,
+      "learning_rate": 4.974232339213519e-06,
+      "loss": 0.5572,
+      "step": 2763
+    },
+    {
+      "epoch": 1.0924672758705853,
+      "grad_norm": 0.5598674028961942,
+      "learning_rate": 4.974209897979448e-06,
+      "loss": 0.5688,
+      "step": 2764
+    },
+    {
+      "epoch": 1.0928624351691776,
+      "grad_norm": 0.4463594082886901,
+      "learning_rate": 4.974187447028185e-06,
+      "loss": 0.5657,
+      "step": 2765
+    },
+    {
+      "epoch": 1.0932575944677698,
+      "grad_norm": 0.45947250553137825,
+      "learning_rate": 4.974164986359814e-06,
+      "loss": 0.5719,
+      "step": 2766
+    },
+    {
+      "epoch": 1.093652753766362,
+      "grad_norm": 0.4621085715770585,
+      "learning_rate": 4.974142515974427e-06,
+      "loss": 0.5801,
+      "step": 2767
+    },
+    {
+      "epoch": 1.0940479130649543,
+      "grad_norm": 0.44428474945478147,
+      "learning_rate": 4.97412003587211e-06,
+      "loss": 0.5583,
+      "step": 2768
+    },
+    {
+      "epoch": 1.0944430723635465,
+      "grad_norm": 0.4573957820670806,
+      "learning_rate": 4.974097546052952e-06,
+      "loss": 0.5731,
+      "step": 2769
+    },
+    {
+      "epoch": 1.0948382316621388,
+      "grad_norm": 0.45565450583105144,
+      "learning_rate": 4.9740750465170415e-06,
+      "loss": 0.5809,
+      "step": 2770
+    },
+    {
+      "epoch": 1.095233390960731,
+      "grad_norm": 0.4482788589021872,
+      "learning_rate": 4.974052537264465e-06,
+      "loss": 0.5792,
+      "step": 2771
+    },
+    {
+      "epoch": 1.0956285502593233,
+      "grad_norm": 0.4461820839459798,
+      "learning_rate": 4.974030018295314e-06,
+      "loss": 0.5786,
+      "step": 2772
+    },
+    {
+      "epoch": 1.0960237095579155,
+      "grad_norm": 0.44439297274093054,
+      "learning_rate": 4.974007489609675e-06,
+      "loss": 0.5887,
+      "step": 2773
+    },
+    {
+      "epoch": 1.0964188688565077,
+      "grad_norm": 0.44582844653174675,
+      "learning_rate": 4.973984951207638e-06,
+      "loss": 0.5897,
+      "step": 2774
+    },
+    {
+      "epoch": 1.0968140281551,
+      "grad_norm": 0.4553585727658302,
+      "learning_rate": 4.9739624030892885e-06,
+      "loss": 0.5661,
+      "step": 2775
+    },
+    {
+      "epoch": 1.0972091874536922,
+      "grad_norm": 0.4479235148347027,
+      "learning_rate": 4.973939845254718e-06,
+      "loss": 0.5931,
+      "step": 2776
+    },
+    {
+      "epoch": 1.0976043467522845,
+      "grad_norm": 0.4428344941159122,
+      "learning_rate": 4.973917277704014e-06,
+      "loss": 0.5598,
+      "step": 2777
+    },
+    {
+      "epoch": 1.0979995060508767,
+      "grad_norm": 0.476027880314366,
+      "learning_rate": 4.973894700437265e-06,
+      "loss": 0.5873,
+      "step": 2778
+    },
+    {
+      "epoch": 1.098394665349469,
+      "grad_norm": 0.4404118923653951,
+      "learning_rate": 4.97387211345456e-06,
+      "loss": 0.5493,
+      "step": 2779
+    },
+    {
+      "epoch": 1.0987898246480612,
+      "grad_norm": 0.4574987015710223,
+      "learning_rate": 4.973849516755987e-06,
+      "loss": 0.5498,
+      "step": 2780
+    },
+    {
+      "epoch": 1.0991849839466534,
+      "grad_norm": 0.45123369032502647,
+      "learning_rate": 4.973826910341636e-06,
+      "loss": 0.5636,
+      "step": 2781
+    },
+    {
+      "epoch": 1.0995801432452457,
+      "grad_norm": 0.4617618303547862,
+      "learning_rate": 4.973804294211595e-06,
+      "loss": 0.5835,
+      "step": 2782
+    },
+    {
+      "epoch": 1.099975302543838,
+      "grad_norm": 0.4579139636190123,
+      "learning_rate": 4.973781668365953e-06,
+      "loss": 0.5737,
+      "step": 2783
+    },
+    {
+      "epoch": 1.1003704618424301,
+      "grad_norm": 0.4771120050290431,
+      "learning_rate": 4.973759032804798e-06,
+      "loss": 0.581,
+      "step": 2784
+    },
+    {
+      "epoch": 1.1007656211410224,
+      "grad_norm": 0.4573835551232583,
+      "learning_rate": 4.973736387528219e-06,
+      "loss": 0.5536,
+      "step": 2785
+    },
+    {
+      "epoch": 1.1011607804396146,
+      "grad_norm": 0.43751748320789247,
+      "learning_rate": 4.973713732536307e-06,
+      "loss": 0.5533,
+      "step": 2786
+    },
+    {
+      "epoch": 1.1015559397382069,
+      "grad_norm": 0.4551112523172073,
+      "learning_rate": 4.97369106782915e-06,
+      "loss": 0.5727,
+      "step": 2787
+    },
+    {
+      "epoch": 1.101951099036799,
+      "grad_norm": 0.4728095976072211,
+      "learning_rate": 4.973668393406835e-06,
+      "loss": 0.5828,
+      "step": 2788
+    },
+    {
+      "epoch": 1.1023462583353916,
+      "grad_norm": 0.46554008898501564,
+      "learning_rate": 4.9736457092694545e-06,
+      "loss": 0.5754,
+      "step": 2789
+    },
+    {
+      "epoch": 1.1027414176339838,
+      "grad_norm": 0.44230286912396166,
+      "learning_rate": 4.9736230154170945e-06,
+      "loss": 0.5659,
+      "step": 2790
+    },
+    {
+      "epoch": 1.103136576932576,
+      "grad_norm": 0.44947861957297847,
+      "learning_rate": 4.973600311849845e-06,
+      "loss": 0.5709,
+      "step": 2791
+    },
+    {
+      "epoch": 1.1035317362311683,
+      "grad_norm": 0.44653133824766644,
+      "learning_rate": 4.973577598567797e-06,
+      "loss": 0.5887,
+      "step": 2792
+    },
+    {
+      "epoch": 1.1039268955297605,
+      "grad_norm": 0.46671707574431837,
+      "learning_rate": 4.9735548755710374e-06,
+      "loss": 0.5871,
+      "step": 2793
+    },
+    {
+      "epoch": 1.1043220548283528,
+      "grad_norm": 0.45380131690571734,
+      "learning_rate": 4.973532142859656e-06,
+      "loss": 0.5517,
+      "step": 2794
+    },
+    {
+      "epoch": 1.104717214126945,
+      "grad_norm": 0.47357589275123324,
+      "learning_rate": 4.973509400433743e-06,
+      "loss": 0.5503,
+      "step": 2795
+    },
+    {
+      "epoch": 1.1051123734255373,
+      "grad_norm": 0.45961644968240184,
+      "learning_rate": 4.9734866482933865e-06,
+      "loss": 0.5652,
+      "step": 2796
+    },
+    {
+      "epoch": 1.1055075327241295,
+      "grad_norm": 0.4625546092783896,
+      "learning_rate": 4.973463886438676e-06,
+      "loss": 0.5787,
+      "step": 2797
+    },
+    {
+      "epoch": 1.1059026920227217,
+      "grad_norm": 0.46020081411653113,
+      "learning_rate": 4.9734411148697025e-06,
+      "loss": 0.5609,
+      "step": 2798
+    },
+    {
+      "epoch": 1.106297851321314,
+      "grad_norm": 0.46162652087403155,
+      "learning_rate": 4.973418333586553e-06,
+      "loss": 0.5872,
+      "step": 2799
+    },
+    {
+      "epoch": 1.1066930106199062,
+      "grad_norm": 0.44195379442185856,
+      "learning_rate": 4.97339554258932e-06,
+      "loss": 0.5497,
+      "step": 2800
+    },
+    {
+      "epoch": 1.1070881699184985,
+      "grad_norm": 0.449613283566657,
+      "learning_rate": 4.97337274187809e-06,
+      "loss": 0.5603,
+      "step": 2801
+    },
+    {
+      "epoch": 1.1074833292170907,
+      "grad_norm": 0.45072734141619586,
+      "learning_rate": 4.973349931452953e-06,
+      "loss": 0.5577,
+      "step": 2802
+    },
+    {
+      "epoch": 1.107878488515683,
+      "grad_norm": 0.4674608199843895,
+      "learning_rate": 4.973327111314e-06,
+      "loss": 0.5786,
+      "step": 2803
+    },
+    {
+      "epoch": 1.1082736478142752,
+      "grad_norm": 0.4630627894242256,
+      "learning_rate": 4.9733042814613205e-06,
+      "loss": 0.5735,
+      "step": 2804
+    },
+    {
+      "epoch": 1.1086688071128674,
+      "grad_norm": 0.4583380745367057,
+      "learning_rate": 4.973281441895004e-06,
+      "loss": 0.56,
+      "step": 2805
+    },
+    {
+      "epoch": 1.1090639664114597,
+      "grad_norm": 0.4733069612546895,
+      "learning_rate": 4.973258592615139e-06,
+      "loss": 0.5684,
+      "step": 2806
+    },
+    {
+      "epoch": 1.109459125710052,
+      "grad_norm": 0.454537320554346,
+      "learning_rate": 4.973235733621816e-06,
+      "loss": 0.5611,
+      "step": 2807
+    },
+    {
+      "epoch": 1.1098542850086441,
+      "grad_norm": 0.46762331052618233,
+      "learning_rate": 4.973212864915126e-06,
+      "loss": 0.5712,
+      "step": 2808
+    },
+    {
+      "epoch": 1.1102494443072364,
+      "grad_norm": 0.46039286243703786,
+      "learning_rate": 4.973189986495157e-06,
+      "loss": 0.5706,
+      "step": 2809
+    },
+    {
+      "epoch": 1.1106446036058286,
+      "grad_norm": 0.45645844421382126,
+      "learning_rate": 4.973167098362e-06,
+      "loss": 0.5969,
+      "step": 2810
+    },
+    {
+      "epoch": 1.1110397629044209,
+      "grad_norm": 0.44815006924446793,
+      "learning_rate": 4.973144200515742e-06,
+      "loss": 0.5622,
+      "step": 2811
+    },
+    {
+      "epoch": 1.111434922203013,
+      "grad_norm": 0.45673663581682494,
+      "learning_rate": 4.9731212929564785e-06,
+      "loss": 0.5713,
+      "step": 2812
+    },
+    {
+      "epoch": 1.1118300815016053,
+      "grad_norm": 0.4410699289674135,
+      "learning_rate": 4.973098375684295e-06,
+      "loss": 0.5572,
+      "step": 2813
+    },
+    {
+      "epoch": 1.1122252408001976,
+      "grad_norm": 0.45953779864395006,
+      "learning_rate": 4.973075448699283e-06,
+      "loss": 0.5721,
+      "step": 2814
+    },
+    {
+      "epoch": 1.1126204000987898,
+      "grad_norm": 0.48969886056398254,
+      "learning_rate": 4.973052512001532e-06,
+      "loss": 0.5961,
+      "step": 2815
+    },
+    {
+      "epoch": 1.113015559397382,
+      "grad_norm": 0.4582083675335303,
+      "learning_rate": 4.9730295655911325e-06,
+      "loss": 0.5664,
+      "step": 2816
+    },
+    {
+      "epoch": 1.1134107186959743,
+      "grad_norm": 0.45873792204035635,
+      "learning_rate": 4.973006609468175e-06,
+      "loss": 0.5798,
+      "step": 2817
+    },
+    {
+      "epoch": 1.1138058779945665,
+      "grad_norm": 0.4595469818479189,
+      "learning_rate": 4.9729836436327486e-06,
+      "loss": 0.5583,
+      "step": 2818
+    },
+    {
+      "epoch": 1.1142010372931588,
+      "grad_norm": 0.4660954028217541,
+      "learning_rate": 4.972960668084945e-06,
+      "loss": 0.586,
+      "step": 2819
+    },
+    {
+      "epoch": 1.114596196591751,
+      "grad_norm": 0.45385089149228297,
+      "learning_rate": 4.972937682824853e-06,
+      "loss": 0.5617,
+      "step": 2820
+    },
+    {
+      "epoch": 1.1149913558903433,
+      "grad_norm": 0.4541868015985183,
+      "learning_rate": 4.972914687852564e-06,
+      "loss": 0.5709,
+      "step": 2821
+    },
+    {
+      "epoch": 1.1153865151889355,
+      "grad_norm": 0.4712498210514924,
+      "learning_rate": 4.972891683168166e-06,
+      "loss": 0.5643,
+      "step": 2822
+    },
+    {
+      "epoch": 1.1157816744875277,
+      "grad_norm": 0.4528912615721515,
+      "learning_rate": 4.9728686687717534e-06,
+      "loss": 0.5677,
+      "step": 2823
+    },
+    {
+      "epoch": 1.11617683378612,
+      "grad_norm": 0.5454379549098186,
+      "learning_rate": 4.972845644663412e-06,
+      "loss": 0.5848,
+      "step": 2824
+    },
+    {
+      "epoch": 1.1165719930847122,
+      "grad_norm": 0.46204902182734076,
+      "learning_rate": 4.972822610843236e-06,
+      "loss": 0.5509,
+      "step": 2825
+    },
+    {
+      "epoch": 1.1169671523833045,
+      "grad_norm": 0.45981650702917953,
+      "learning_rate": 4.972799567311314e-06,
+      "loss": 0.5589,
+      "step": 2826
+    },
+    {
+      "epoch": 1.1173623116818967,
+      "grad_norm": 0.4748451628776366,
+      "learning_rate": 4.9727765140677374e-06,
+      "loss": 0.5735,
+      "step": 2827
+    },
+    {
+      "epoch": 1.117757470980489,
+      "grad_norm": 0.43824092167576956,
+      "learning_rate": 4.972753451112596e-06,
+      "loss": 0.5595,
+      "step": 2828
+    },
+    {
+      "epoch": 1.1181526302790812,
+      "grad_norm": 0.4532662779925716,
+      "learning_rate": 4.97273037844598e-06,
+      "loss": 0.5644,
+      "step": 2829
+    },
+    {
+      "epoch": 1.1185477895776734,
+      "grad_norm": 0.45748152598351954,
+      "learning_rate": 4.972707296067981e-06,
+      "loss": 0.5677,
+      "step": 2830
+    },
+    {
+      "epoch": 1.1189429488762657,
+      "grad_norm": 0.4588021846664917,
+      "learning_rate": 4.972684203978689e-06,
+      "loss": 0.5741,
+      "step": 2831
+    },
+    {
+      "epoch": 1.119338108174858,
+      "grad_norm": 0.4571577453972498,
+      "learning_rate": 4.972661102178196e-06,
+      "loss": 0.5521,
+      "step": 2832
+    },
+    {
+      "epoch": 1.1197332674734501,
+      "grad_norm": 0.4584124940006824,
+      "learning_rate": 4.97263799066659e-06,
+      "loss": 0.555,
+      "step": 2833
+    },
+    {
+      "epoch": 1.1201284267720424,
+      "grad_norm": 0.4768334321839287,
+      "learning_rate": 4.972614869443965e-06,
+      "loss": 0.5554,
+      "step": 2834
+    },
+    {
+      "epoch": 1.1205235860706346,
+      "grad_norm": 0.45878358249431184,
+      "learning_rate": 4.972591738510409e-06,
+      "loss": 0.5652,
+      "step": 2835
+    },
+    {
+      "epoch": 1.1209187453692269,
+      "grad_norm": 0.4603002756472343,
+      "learning_rate": 4.972568597866014e-06,
+      "loss": 0.5706,
+      "step": 2836
+    },
+    {
+      "epoch": 1.121313904667819,
+      "grad_norm": 0.4331713760224906,
+      "learning_rate": 4.9725454475108714e-06,
+      "loss": 0.5544,
+      "step": 2837
+    },
+    {
+      "epoch": 1.1217090639664113,
+      "grad_norm": 0.4527006951431627,
+      "learning_rate": 4.972522287445072e-06,
+      "loss": 0.5435,
+      "step": 2838
+    },
+    {
+      "epoch": 1.1221042232650038,
+      "grad_norm": 0.4757703333155185,
+      "learning_rate": 4.972499117668707e-06,
+      "loss": 0.5851,
+      "step": 2839
+    },
+    {
+      "epoch": 1.122499382563596,
+      "grad_norm": 0.45557041972406903,
+      "learning_rate": 4.972475938181866e-06,
+      "loss": 0.5624,
+      "step": 2840
+    },
+    {
+      "epoch": 1.1228945418621883,
+      "grad_norm": 0.4512928810062557,
+      "learning_rate": 4.972452748984641e-06,
+      "loss": 0.5485,
+      "step": 2841
+    },
+    {
+      "epoch": 1.1232897011607805,
+      "grad_norm": 0.47660422203668557,
+      "learning_rate": 4.972429550077122e-06,
+      "loss": 0.5785,
+      "step": 2842
+    },
+    {
+      "epoch": 1.1236848604593728,
+      "grad_norm": 0.45195553481987427,
+      "learning_rate": 4.972406341459403e-06,
+      "loss": 0.5754,
+      "step": 2843
+    },
+    {
+      "epoch": 1.124080019757965,
+      "grad_norm": 0.4692003584527425,
+      "learning_rate": 4.972383123131572e-06,
+      "loss": 0.5713,
+      "step": 2844
+    },
+    {
+      "epoch": 1.1244751790565573,
+      "grad_norm": 0.46001788143359484,
+      "learning_rate": 4.9723598950937216e-06,
+      "loss": 0.5527,
+      "step": 2845
+    },
+    {
+      "epoch": 1.1248703383551495,
+      "grad_norm": 0.4530700630207127,
+      "learning_rate": 4.972336657345943e-06,
+      "loss": 0.5762,
+      "step": 2846
+    },
+    {
+      "epoch": 1.1252654976537417,
+      "grad_norm": 0.4814747648947358,
+      "learning_rate": 4.972313409888327e-06,
+      "loss": 0.5757,
+      "step": 2847
+    },
+    {
+      "epoch": 1.125660656952334,
+      "grad_norm": 0.4644316998225435,
+      "learning_rate": 4.972290152720965e-06,
+      "loss": 0.5932,
+      "step": 2848
+    },
+    {
+      "epoch": 1.1260558162509262,
+      "grad_norm": 0.43522985761057875,
+      "learning_rate": 4.97226688584395e-06,
+      "loss": 0.548,
+      "step": 2849
+    },
+    {
+      "epoch": 1.1264509755495185,
+      "grad_norm": 0.44885828698030505,
+      "learning_rate": 4.97224360925737e-06,
+      "loss": 0.5673,
+      "step": 2850
+    },
+    {
+      "epoch": 1.1268461348481107,
+      "grad_norm": 0.4546348191990088,
+      "learning_rate": 4.972220322961318e-06,
+      "loss": 0.5717,
+      "step": 2851
+    },
+    {
+      "epoch": 1.127241294146703,
+      "grad_norm": 0.45766425039161684,
+      "learning_rate": 4.972197026955888e-06,
+      "loss": 0.5738,
+      "step": 2852
+    },
+    {
+      "epoch": 1.1276364534452952,
+      "grad_norm": 0.48259567077817345,
+      "learning_rate": 4.972173721241168e-06,
+      "loss": 0.5711,
+      "step": 2853
+    },
+    {
+      "epoch": 1.1280316127438874,
+      "grad_norm": 0.4465075582537828,
+      "learning_rate": 4.972150405817251e-06,
+      "loss": 0.5608,
+      "step": 2854
+    },
+    {
+      "epoch": 1.1284267720424797,
+      "grad_norm": 0.4494060907226681,
+      "learning_rate": 4.972127080684228e-06,
+      "loss": 0.5597,
+      "step": 2855
+    },
+    {
+      "epoch": 1.128821931341072,
+      "grad_norm": 0.45776592151810525,
+      "learning_rate": 4.972103745842192e-06,
+      "loss": 0.5696,
+      "step": 2856
+    },
+    {
+      "epoch": 1.1292170906396641,
+      "grad_norm": 0.4517897855267896,
+      "learning_rate": 4.9720804012912325e-06,
+      "loss": 0.5597,
+      "step": 2857
+    },
+    {
+      "epoch": 1.1296122499382564,
+      "grad_norm": 0.45557106783434753,
+      "learning_rate": 4.9720570470314435e-06,
+      "loss": 0.585,
+      "step": 2858
+    },
+    {
+      "epoch": 1.1300074092368486,
+      "grad_norm": 0.44670409565234737,
+      "learning_rate": 4.9720336830629145e-06,
+      "loss": 0.5609,
+      "step": 2859
+    },
+    {
+      "epoch": 1.1304025685354409,
+      "grad_norm": 0.4616073635588755,
+      "learning_rate": 4.972010309385739e-06,
+      "loss": 0.5669,
+      "step": 2860
+    },
+    {
+      "epoch": 1.130797727834033,
+      "grad_norm": 0.48898771950037234,
+      "learning_rate": 4.971986926000008e-06,
+      "loss": 0.5781,
+      "step": 2861
+    },
+    {
+      "epoch": 1.1311928871326253,
+      "grad_norm": 0.47129122376722427,
+      "learning_rate": 4.971963532905812e-06,
+      "loss": 0.5856,
+      "step": 2862
+    },
+    {
+      "epoch": 1.1315880464312176,
+      "grad_norm": 0.4367114279377543,
+      "learning_rate": 4.971940130103245e-06,
+      "loss": 0.5498,
+      "step": 2863
+    },
+    {
+      "epoch": 1.1319832057298098,
+      "grad_norm": 0.45445984797614775,
+      "learning_rate": 4.9719167175924e-06,
+      "loss": 0.5585,
+      "step": 2864
+    },
+    {
+      "epoch": 1.132378365028402,
+      "grad_norm": 0.47288776531263177,
+      "learning_rate": 4.971893295373366e-06,
+      "loss": 0.5699,
+      "step": 2865
+    },
+    {
+      "epoch": 1.1327735243269943,
+      "grad_norm": 0.46037768185257905,
+      "learning_rate": 4.971869863446235e-06,
+      "loss": 0.555,
+      "step": 2866
+    },
+    {
+      "epoch": 1.1331686836255865,
+      "grad_norm": 0.4409672704352513,
+      "learning_rate": 4.971846421811101e-06,
+      "loss": 0.5741,
+      "step": 2867
+    },
+    {
+      "epoch": 1.1335638429241788,
+      "grad_norm": 0.4735947826477942,
+      "learning_rate": 4.971822970468056e-06,
+      "loss": 0.5751,
+      "step": 2868
+    },
+    {
+      "epoch": 1.133959002222771,
+      "grad_norm": 0.4985409552375511,
+      "learning_rate": 4.97179950941719e-06,
+      "loss": 0.5831,
+      "step": 2869
+    },
+    {
+      "epoch": 1.1343541615213633,
+      "grad_norm": 0.44396729751506975,
+      "learning_rate": 4.971776038658598e-06,
+      "loss": 0.5666,
+      "step": 2870
+    },
+    {
+      "epoch": 1.1347493208199555,
+      "grad_norm": 0.46267223426001386,
+      "learning_rate": 4.97175255819237e-06,
+      "loss": 0.5596,
+      "step": 2871
+    },
+    {
+      "epoch": 1.1351444801185477,
+      "grad_norm": 0.46568413072052744,
+      "learning_rate": 4.9717290680185985e-06,
+      "loss": 0.5686,
+      "step": 2872
+    },
+    {
+      "epoch": 1.13553963941714,
+      "grad_norm": 0.4673886583205963,
+      "learning_rate": 4.971705568137376e-06,
+      "loss": 0.5756,
+      "step": 2873
+    },
+    {
+      "epoch": 1.1359347987157322,
+      "grad_norm": 0.45942651811103113,
+      "learning_rate": 4.971682058548795e-06,
+      "loss": 0.5676,
+      "step": 2874
+    },
+    {
+      "epoch": 1.1363299580143245,
+      "grad_norm": 0.45192092503324155,
+      "learning_rate": 4.971658539252948e-06,
+      "loss": 0.5729,
+      "step": 2875
+    },
+    {
+      "epoch": 1.1367251173129167,
+      "grad_norm": 0.4428557809778483,
+      "learning_rate": 4.971635010249928e-06,
+      "loss": 0.5751,
+      "step": 2876
+    },
+    {
+      "epoch": 1.137120276611509,
+      "grad_norm": 0.4506504247352054,
+      "learning_rate": 4.971611471539826e-06,
+      "loss": 0.5628,
+      "step": 2877
+    },
+    {
+      "epoch": 1.1375154359101012,
+      "grad_norm": 0.45545331444170223,
+      "learning_rate": 4.971587923122734e-06,
+      "loss": 0.5642,
+      "step": 2878
+    },
+    {
+      "epoch": 1.1379105952086934,
+      "grad_norm": 0.479413221980225,
+      "learning_rate": 4.971564364998747e-06,
+      "loss": 0.5749,
+      "step": 2879
+    },
+    {
+      "epoch": 1.1383057545072857,
+      "grad_norm": 0.4439935558628392,
+      "learning_rate": 4.971540797167954e-06,
+      "loss": 0.5599,
+      "step": 2880
+    },
+    {
+      "epoch": 1.138700913805878,
+      "grad_norm": 0.4444069130578539,
+      "learning_rate": 4.971517219630451e-06,
+      "loss": 0.5862,
+      "step": 2881
+    },
+    {
+      "epoch": 1.1390960731044704,
+      "grad_norm": 0.4552609885225358,
+      "learning_rate": 4.971493632386329e-06,
+      "loss": 0.5575,
+      "step": 2882
+    },
+    {
+      "epoch": 1.1394912324030626,
+      "grad_norm": 0.4382492830615158,
+      "learning_rate": 4.97147003543568e-06,
+      "loss": 0.5589,
+      "step": 2883
+    },
+    {
+      "epoch": 1.1398863917016548,
+      "grad_norm": 0.4437390367384225,
+      "learning_rate": 4.971446428778599e-06,
+      "loss": 0.5714,
+      "step": 2884
+    },
+    {
+      "epoch": 1.140281551000247,
+      "grad_norm": 0.4456567659509046,
+      "learning_rate": 4.971422812415176e-06,
+      "loss": 0.5526,
+      "step": 2885
+    },
+    {
+      "epoch": 1.1406767102988393,
+      "grad_norm": 0.44730543435927383,
+      "learning_rate": 4.971399186345505e-06,
+      "loss": 0.5579,
+      "step": 2886
+    },
+    {
+      "epoch": 1.1410718695974316,
+      "grad_norm": 0.45573671820566486,
+      "learning_rate": 4.97137555056968e-06,
+      "loss": 0.56,
+      "step": 2887
+    },
+    {
+      "epoch": 1.1414670288960238,
+      "grad_norm": 0.4505940700048066,
+      "learning_rate": 4.971351905087791e-06,
+      "loss": 0.5677,
+      "step": 2888
+    },
+    {
+      "epoch": 1.141862188194616,
+      "grad_norm": 0.44421960402517957,
+      "learning_rate": 4.971328249899932e-06,
+      "loss": 0.5494,
+      "step": 2889
+    },
+    {
+      "epoch": 1.1422573474932083,
+      "grad_norm": 0.4494388053299841,
+      "learning_rate": 4.971304585006198e-06,
+      "loss": 0.5808,
+      "step": 2890
+    },
+    {
+      "epoch": 1.1426525067918005,
+      "grad_norm": 0.454206814685918,
+      "learning_rate": 4.971280910406679e-06,
+      "loss": 0.5647,
+      "step": 2891
+    },
+    {
+      "epoch": 1.1430476660903928,
+      "grad_norm": 0.4570937269415416,
+      "learning_rate": 4.971257226101469e-06,
+      "loss": 0.5645,
+      "step": 2892
+    },
+    {
+      "epoch": 1.143442825388985,
+      "grad_norm": 0.4424306427628629,
+      "learning_rate": 4.971233532090662e-06,
+      "loss": 0.5617,
+      "step": 2893
+    },
+    {
+      "epoch": 1.1438379846875772,
+      "grad_norm": 0.43897777777159125,
+      "learning_rate": 4.97120982837435e-06,
+      "loss": 0.5649,
+      "step": 2894
+    },
+    {
+      "epoch": 1.1442331439861695,
+      "grad_norm": 0.44861248994251357,
+      "learning_rate": 4.971186114952628e-06,
+      "loss": 0.565,
+      "step": 2895
+    },
+    {
+      "epoch": 1.1446283032847617,
+      "grad_norm": 0.47001095027879825,
+      "learning_rate": 4.971162391825586e-06,
+      "loss": 0.5803,
+      "step": 2896
+    },
+    {
+      "epoch": 1.145023462583354,
+      "grad_norm": 0.44454980270407224,
+      "learning_rate": 4.971138658993318e-06,
+      "loss": 0.561,
+      "step": 2897
+    },
+    {
+      "epoch": 1.1454186218819462,
+      "grad_norm": 0.4467127010573676,
+      "learning_rate": 4.971114916455919e-06,
+      "loss": 0.5531,
+      "step": 2898
+    },
+    {
+      "epoch": 1.1458137811805384,
+      "grad_norm": 0.46879881901374293,
+      "learning_rate": 4.9710911642134805e-06,
+      "loss": 0.5943,
+      "step": 2899
+    },
+    {
+      "epoch": 1.1462089404791307,
+      "grad_norm": 0.46048699268271154,
+      "learning_rate": 4.971067402266097e-06,
+      "loss": 0.5785,
+      "step": 2900
+    },
+    {
+      "epoch": 1.146604099777723,
+      "grad_norm": 0.4578268867042688,
+      "learning_rate": 4.971043630613861e-06,
+      "loss": 0.5627,
+      "step": 2901
+    },
+    {
+      "epoch": 1.1469992590763152,
+      "grad_norm": 0.44834695982191136,
+      "learning_rate": 4.971019849256866e-06,
+      "loss": 0.5596,
+      "step": 2902
+    },
+    {
+      "epoch": 1.1473944183749074,
+      "grad_norm": 0.4447032230484451,
+      "learning_rate": 4.970996058195206e-06,
+      "loss": 0.555,
+      "step": 2903
+    },
+    {
+      "epoch": 1.1477895776734997,
+      "grad_norm": 0.4655101994383645,
+      "learning_rate": 4.970972257428973e-06,
+      "loss": 0.5904,
+      "step": 2904
+    },
+    {
+      "epoch": 1.148184736972092,
+      "grad_norm": 0.48666828141831653,
+      "learning_rate": 4.970948446958262e-06,
+      "loss": 0.5872,
+      "step": 2905
+    },
+    {
+      "epoch": 1.1485798962706841,
+      "grad_norm": 0.45560221158301,
+      "learning_rate": 4.970924626783165e-06,
+      "loss": 0.572,
+      "step": 2906
+    },
+    {
+      "epoch": 1.1489750555692764,
+      "grad_norm": 0.4542327270574395,
+      "learning_rate": 4.970900796903778e-06,
+      "loss": 0.5807,
+      "step": 2907
+    },
+    {
+      "epoch": 1.1493702148678686,
+      "grad_norm": 0.45302601284820004,
+      "learning_rate": 4.970876957320193e-06,
+      "loss": 0.5556,
+      "step": 2908
+    },
+    {
+      "epoch": 1.1497653741664609,
+      "grad_norm": 0.45085641190037123,
+      "learning_rate": 4.970853108032503e-06,
+      "loss": 0.553,
+      "step": 2909
+    },
+    {
+      "epoch": 1.150160533465053,
+      "grad_norm": 0.44311699037774255,
+      "learning_rate": 4.970829249040803e-06,
+      "loss": 0.5783,
+      "step": 2910
+    },
+    {
+      "epoch": 1.1505556927636453,
+      "grad_norm": 0.4565247742626593,
+      "learning_rate": 4.970805380345186e-06,
+      "loss": 0.5668,
+      "step": 2911
+    },
+    {
+      "epoch": 1.1509508520622376,
+      "grad_norm": 0.442670058390068,
+      "learning_rate": 4.970781501945745e-06,
+      "loss": 0.5804,
+      "step": 2912
+    },
+    {
+      "epoch": 1.1513460113608298,
+      "grad_norm": 0.4587903036700737,
+      "learning_rate": 4.970757613842575e-06,
+      "loss": 0.5436,
+      "step": 2913
+    },
+    {
+      "epoch": 1.151741170659422,
+      "grad_norm": 0.4605719839378815,
+      "learning_rate": 4.970733716035769e-06,
+      "loss": 0.5708,
+      "step": 2914
+    },
+    {
+      "epoch": 1.1521363299580143,
+      "grad_norm": 0.4811516114353964,
+      "learning_rate": 4.970709808525423e-06,
+      "loss": 0.5695,
+      "step": 2915
+    },
+    {
+      "epoch": 1.1525314892566065,
+      "grad_norm": 0.4465598040548844,
+      "learning_rate": 4.970685891311627e-06,
+      "loss": 0.5585,
+      "step": 2916
+    },
+    {
+      "epoch": 1.1529266485551988,
+      "grad_norm": 0.4482126904622925,
+      "learning_rate": 4.970661964394479e-06,
+      "loss": 0.5531,
+      "step": 2917
+    },
+    {
+      "epoch": 1.153321807853791,
+      "grad_norm": 0.45465037670607433,
+      "learning_rate": 4.97063802777407e-06,
+      "loss": 0.5552,
+      "step": 2918
+    },
+    {
+      "epoch": 1.1537169671523833,
+      "grad_norm": 0.44832182812879506,
+      "learning_rate": 4.970614081450495e-06,
+      "loss": 0.5564,
+      "step": 2919
+    },
+    {
+      "epoch": 1.1541121264509755,
+      "grad_norm": 0.45584014338048706,
+      "learning_rate": 4.9705901254238485e-06,
+      "loss": 0.5569,
+      "step": 2920
+    },
+    {
+      "epoch": 1.1545072857495677,
+      "grad_norm": 0.44834971508394056,
+      "learning_rate": 4.970566159694224e-06,
+      "loss": 0.5461,
+      "step": 2921
+    },
+    {
+      "epoch": 1.15490244504816,
+      "grad_norm": 0.4578941775430072,
+      "learning_rate": 4.970542184261716e-06,
+      "loss": 0.5668,
+      "step": 2922
+    },
+    {
+      "epoch": 1.1552976043467522,
+      "grad_norm": 0.4817857268031156,
+      "learning_rate": 4.9705181991264185e-06,
+      "loss": 0.5776,
+      "step": 2923
+    },
+    {
+      "epoch": 1.1556927636453445,
+      "grad_norm": 0.4608535112462551,
+      "learning_rate": 4.9704942042884256e-06,
+      "loss": 0.5665,
+      "step": 2924
+    },
+    {
+      "epoch": 1.1560879229439367,
+      "grad_norm": 0.45851740053649265,
+      "learning_rate": 4.970470199747831e-06,
+      "loss": 0.5699,
+      "step": 2925
+    },
+    {
+      "epoch": 1.156483082242529,
+      "grad_norm": 0.47180228641422883,
+      "learning_rate": 4.97044618550473e-06,
+      "loss": 0.572,
+      "step": 2926
+    },
+    {
+      "epoch": 1.1568782415411212,
+      "grad_norm": 0.47156814207993075,
+      "learning_rate": 4.970422161559217e-06,
+      "loss": 0.5437,
+      "step": 2927
+    },
+    {
+      "epoch": 1.1572734008397134,
+      "grad_norm": 0.4543173767246186,
+      "learning_rate": 4.970398127911386e-06,
+      "loss": 0.5796,
+      "step": 2928
+    },
+    {
+      "epoch": 1.1576685601383057,
+      "grad_norm": 0.4467556140355343,
+      "learning_rate": 4.97037408456133e-06,
+      "loss": 0.5739,
+      "step": 2929
+    },
+    {
+      "epoch": 1.158063719436898,
+      "grad_norm": 0.4572687150991262,
+      "learning_rate": 4.970350031509146e-06,
+      "loss": 0.5789,
+      "step": 2930
+    },
+    {
+      "epoch": 1.1584588787354901,
+      "grad_norm": 0.46752603817680133,
+      "learning_rate": 4.970325968754926e-06,
+      "loss": 0.5902,
+      "step": 2931
+    },
+    {
+      "epoch": 1.1588540380340824,
+      "grad_norm": 0.44878914889504196,
+      "learning_rate": 4.970301896298767e-06,
+      "loss": 0.5721,
+      "step": 2932
+    },
+    {
+      "epoch": 1.1592491973326746,
+      "grad_norm": 0.4485973988983018,
+      "learning_rate": 4.9702778141407615e-06,
+      "loss": 0.5553,
+      "step": 2933
+    },
+    {
+      "epoch": 1.1596443566312669,
+      "grad_norm": 0.46028448862594296,
+      "learning_rate": 4.970253722281006e-06,
+      "loss": 0.5714,
+      "step": 2934
+    },
+    {
+      "epoch": 1.160039515929859,
+      "grad_norm": 0.43839159073806405,
+      "learning_rate": 4.970229620719592e-06,
+      "loss": 0.5682,
+      "step": 2935
+    },
+    {
+      "epoch": 1.1604346752284516,
+      "grad_norm": 0.4502203330745019,
+      "learning_rate": 4.970205509456617e-06,
+      "loss": 0.5633,
+      "step": 2936
+    },
+    {
+      "epoch": 1.1608298345270438,
+      "grad_norm": 0.4577454214170737,
+      "learning_rate": 4.970181388492174e-06,
+      "loss": 0.56,
+      "step": 2937
+    },
+    {
+      "epoch": 1.161224993825636,
+      "grad_norm": 0.5137258014666423,
+      "learning_rate": 4.970157257826359e-06,
+      "loss": 0.5807,
+      "step": 2938
+    },
+    {
+      "epoch": 1.1616201531242283,
+      "grad_norm": 0.456240539508233,
+      "learning_rate": 4.970133117459266e-06,
+      "loss": 0.5698,
+      "step": 2939
+    },
+    {
+      "epoch": 1.1620153124228205,
+      "grad_norm": 0.4571385214359614,
+      "learning_rate": 4.9701089673909905e-06,
+      "loss": 0.5712,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1624104717214128,
+      "grad_norm": 0.43698412892461164,
+      "learning_rate": 4.970084807621627e-06,
+      "loss": 0.5518,
+      "step": 2941
+    },
+    {
+      "epoch": 1.162805631020005,
+      "grad_norm": 0.4508512464344563,
+      "learning_rate": 4.97006063815127e-06,
+      "loss": 0.5543,
+      "step": 2942
+    },
+    {
+      "epoch": 1.1632007903185972,
+      "grad_norm": 0.444206994150076,
+      "learning_rate": 4.970036458980014e-06,
+      "loss": 0.5581,
+      "step": 2943
+    },
+    {
+      "epoch": 1.1635959496171895,
+      "grad_norm": 0.4524722456542353,
+      "learning_rate": 4.9700122701079566e-06,
+      "loss": 0.5542,
+      "step": 2944
+    },
+    {
+      "epoch": 1.1639911089157817,
+      "grad_norm": 0.44578046719303255,
+      "learning_rate": 4.969988071535189e-06,
+      "loss": 0.5725,
+      "step": 2945
+    },
+    {
+      "epoch": 1.164386268214374,
+      "grad_norm": 0.4549968477995691,
+      "learning_rate": 4.969963863261808e-06,
+      "loss": 0.5848,
+      "step": 2946
+    },
+    {
+      "epoch": 1.1647814275129662,
+      "grad_norm": 0.4552555332588111,
+      "learning_rate": 4.969939645287911e-06,
+      "loss": 0.557,
+      "step": 2947
+    },
+    {
+      "epoch": 1.1651765868115584,
+      "grad_norm": 0.4750098537013762,
+      "learning_rate": 4.969915417613589e-06,
+      "loss": 0.5864,
+      "step": 2948
+    },
+    {
+      "epoch": 1.1655717461101507,
+      "grad_norm": 0.45082337345793494,
+      "learning_rate": 4.96989118023894e-06,
+      "loss": 0.5866,
+      "step": 2949
+    },
+    {
+      "epoch": 1.165966905408743,
+      "grad_norm": 0.4455097203171396,
+      "learning_rate": 4.969866933164057e-06,
+      "loss": 0.5607,
+      "step": 2950
+    },
+    {
+      "epoch": 1.1663620647073352,
+      "grad_norm": 0.4489253199947083,
+      "learning_rate": 4.969842676389038e-06,
+      "loss": 0.5662,
+      "step": 2951
+    },
+    {
+      "epoch": 1.1667572240059274,
+      "grad_norm": 0.46293264922861316,
+      "learning_rate": 4.969818409913976e-06,
+      "loss": 0.5551,
+      "step": 2952
+    },
+    {
+      "epoch": 1.1671523833045196,
+      "grad_norm": 0.45718754430330005,
+      "learning_rate": 4.969794133738967e-06,
+      "loss": 0.5541,
+      "step": 2953
+    },
+    {
+      "epoch": 1.1675475426031119,
+      "grad_norm": 0.45394398829100463,
+      "learning_rate": 4.9697698478641056e-06,
+      "loss": 0.5543,
+      "step": 2954
+    },
+    {
+      "epoch": 1.1679427019017041,
+      "grad_norm": 0.4580575806358461,
+      "learning_rate": 4.969745552289489e-06,
+      "loss": 0.5648,
+      "step": 2955
+    },
+    {
+      "epoch": 1.1683378612002964,
+      "grad_norm": 0.45929415970435655,
+      "learning_rate": 4.969721247015212e-06,
+      "loss": 0.5802,
+      "step": 2956
+    },
+    {
+      "epoch": 1.1687330204988886,
+      "grad_norm": 0.43706127656723726,
+      "learning_rate": 4.969696932041369e-06,
+      "loss": 0.5658,
+      "step": 2957
+    },
+    {
+      "epoch": 1.1691281797974808,
+      "grad_norm": 0.45589757018205523,
+      "learning_rate": 4.969672607368056e-06,
+      "loss": 0.5451,
+      "step": 2958
+    },
+    {
+      "epoch": 1.169523339096073,
+      "grad_norm": 0.4453274113893662,
+      "learning_rate": 4.969648272995368e-06,
+      "loss": 0.5651,
+      "step": 2959
+    },
+    {
+      "epoch": 1.1699184983946653,
+      "grad_norm": 0.4422585180450584,
+      "learning_rate": 4.9696239289234025e-06,
+      "loss": 0.5513,
+      "step": 2960
+    },
+    {
+      "epoch": 1.1703136576932576,
+      "grad_norm": 0.4609612248644218,
+      "learning_rate": 4.969599575152253e-06,
+      "loss": 0.579,
+      "step": 2961
+    },
+    {
+      "epoch": 1.1707088169918498,
+      "grad_norm": 0.46053331483814797,
+      "learning_rate": 4.969575211682016e-06,
+      "loss": 0.5456,
+      "step": 2962
+    },
+    {
+      "epoch": 1.171103976290442,
+      "grad_norm": 0.46253809484884606,
+      "learning_rate": 4.969550838512787e-06,
+      "loss": 0.5738,
+      "step": 2963
+    },
+    {
+      "epoch": 1.1714991355890343,
+      "grad_norm": 0.45757469040052795,
+      "learning_rate": 4.969526455644664e-06,
+      "loss": 0.5835,
+      "step": 2964
+    },
+    {
+      "epoch": 1.1718942948876265,
+      "grad_norm": 0.4649235194612874,
+      "learning_rate": 4.969502063077738e-06,
+      "loss": 0.5656,
+      "step": 2965
+    },
+    {
+      "epoch": 1.1722894541862188,
+      "grad_norm": 0.4529442270814776,
+      "learning_rate": 4.969477660812108e-06,
+      "loss": 0.5475,
+      "step": 2966
+    },
+    {
+      "epoch": 1.172684613484811,
+      "grad_norm": 0.46009271161789567,
+      "learning_rate": 4.969453248847871e-06,
+      "loss": 0.5638,
+      "step": 2967
+    },
+    {
+      "epoch": 1.1730797727834033,
+      "grad_norm": 0.4543848817490557,
+      "learning_rate": 4.96942882718512e-06,
+      "loss": 0.556,
+      "step": 2968
+    },
+    {
+      "epoch": 1.1734749320819955,
+      "grad_norm": 0.4486250704854233,
+      "learning_rate": 4.9694043958239515e-06,
+      "loss": 0.5475,
+      "step": 2969
+    },
+    {
+      "epoch": 1.1738700913805877,
+      "grad_norm": 0.475212385037125,
+      "learning_rate": 4.969379954764463e-06,
+      "loss": 0.5694,
+      "step": 2970
+    },
+    {
+      "epoch": 1.17426525067918,
+      "grad_norm": 0.46354530802940175,
+      "learning_rate": 4.969355504006749e-06,
+      "loss": 0.5693,
+      "step": 2971
+    },
+    {
+      "epoch": 1.1746604099777722,
+      "grad_norm": 0.4654693704390786,
+      "learning_rate": 4.969331043550907e-06,
+      "loss": 0.5821,
+      "step": 2972
+    },
+    {
+      "epoch": 1.1750555692763645,
+      "grad_norm": 0.45975428248741407,
+      "learning_rate": 4.969306573397031e-06,
+      "loss": 0.5763,
+      "step": 2973
+    },
+    {
+      "epoch": 1.1754507285749567,
+      "grad_norm": 0.4567317408231308,
+      "learning_rate": 4.969282093545218e-06,
+      "loss": 0.5785,
+      "step": 2974
+    },
+    {
+      "epoch": 1.175845887873549,
+      "grad_norm": 0.44978056274129696,
+      "learning_rate": 4.969257603995566e-06,
+      "loss": 0.5624,
+      "step": 2975
+    },
+    {
+      "epoch": 1.1762410471721412,
+      "grad_norm": 0.4535109810307646,
+      "learning_rate": 4.969233104748168e-06,
+      "loss": 0.5568,
+      "step": 2976
+    },
+    {
+      "epoch": 1.1766362064707336,
+      "grad_norm": 0.512694154610389,
+      "learning_rate": 4.9692085958031225e-06,
+      "loss": 0.5537,
+      "step": 2977
+    },
+    {
+      "epoch": 1.1770313657693259,
+      "grad_norm": 0.4657562385746197,
+      "learning_rate": 4.969184077160524e-06,
+      "loss": 0.5754,
+      "step": 2978
+    },
+    {
+      "epoch": 1.1774265250679181,
+      "grad_norm": 0.4676379011730835,
+      "learning_rate": 4.96915954882047e-06,
+      "loss": 0.5763,
+      "step": 2979
+    },
+    {
+      "epoch": 1.1778216843665104,
+      "grad_norm": 0.44433392586627835,
+      "learning_rate": 4.9691350107830575e-06,
+      "loss": 0.5544,
+      "step": 2980
+    },
+    {
+      "epoch": 1.1782168436651026,
+      "grad_norm": 0.4655458760369557,
+      "learning_rate": 4.9691104630483825e-06,
+      "loss": 0.5511,
+      "step": 2981
+    },
+    {
+      "epoch": 1.1786120029636948,
+      "grad_norm": 0.4685666784171711,
+      "learning_rate": 4.96908590561654e-06,
+      "loss": 0.5601,
+      "step": 2982
+    },
+    {
+      "epoch": 1.179007162262287,
+      "grad_norm": 0.46307465210699217,
+      "learning_rate": 4.969061338487627e-06,
+      "loss": 0.5804,
+      "step": 2983
+    },
+    {
+      "epoch": 1.1794023215608793,
+      "grad_norm": 0.4751945587922622,
+      "learning_rate": 4.969036761661741e-06,
+      "loss": 0.5761,
+      "step": 2984
+    },
+    {
+      "epoch": 1.1797974808594716,
+      "grad_norm": 0.4354064786650696,
+      "learning_rate": 4.969012175138978e-06,
+      "loss": 0.5496,
+      "step": 2985
+    },
+    {
+      "epoch": 1.1801926401580638,
+      "grad_norm": 0.4410526844854051,
+      "learning_rate": 4.968987578919434e-06,
+      "loss": 0.574,
+      "step": 2986
+    },
+    {
+      "epoch": 1.180587799456656,
+      "grad_norm": 0.5073421248174037,
+      "learning_rate": 4.9689629730032065e-06,
+      "loss": 0.5713,
+      "step": 2987
+    },
+    {
+      "epoch": 1.1809829587552483,
+      "grad_norm": 0.45094193815320693,
+      "learning_rate": 4.968938357390391e-06,
+      "loss": 0.5618,
+      "step": 2988
+    },
+    {
+      "epoch": 1.1813781180538405,
+      "grad_norm": 0.45984337128911706,
+      "learning_rate": 4.968913732081085e-06,
+      "loss": 0.5518,
+      "step": 2989
+    },
+    {
+      "epoch": 1.1817732773524328,
+      "grad_norm": 0.44962428315636743,
+      "learning_rate": 4.968889097075386e-06,
+      "loss": 0.5603,
+      "step": 2990
+    },
+    {
+      "epoch": 1.182168436651025,
+      "grad_norm": 0.4791783985253456,
+      "learning_rate": 4.968864452373388e-06,
+      "loss": 0.5695,
+      "step": 2991
+    },
+    {
+      "epoch": 1.1825635959496172,
+      "grad_norm": 0.46504234897652474,
+      "learning_rate": 4.968839797975192e-06,
+      "loss": 0.57,
+      "step": 2992
+    },
+    {
+      "epoch": 1.1829587552482095,
+      "grad_norm": 0.46532115508012195,
+      "learning_rate": 4.96881513388089e-06,
+      "loss": 0.5848,
+      "step": 2993
+    },
+    {
+      "epoch": 1.1833539145468017,
+      "grad_norm": 0.4736857697301262,
+      "learning_rate": 4.968790460090584e-06,
+      "loss": 0.5536,
+      "step": 2994
+    },
+    {
+      "epoch": 1.183749073845394,
+      "grad_norm": 0.47327160279398156,
+      "learning_rate": 4.968765776604366e-06,
+      "loss": 0.59,
+      "step": 2995
+    },
+    {
+      "epoch": 1.1841442331439862,
+      "grad_norm": 0.44968273792335417,
+      "learning_rate": 4.968741083422335e-06,
+      "loss": 0.5602,
+      "step": 2996
+    },
+    {
+      "epoch": 1.1845393924425784,
+      "grad_norm": 0.47160397918194813,
+      "learning_rate": 4.968716380544589e-06,
+      "loss": 0.578,
+      "step": 2997
+    },
+    {
+      "epoch": 1.1849345517411707,
+      "grad_norm": 0.4674565580152445,
+      "learning_rate": 4.968691667971224e-06,
+      "loss": 0.565,
+      "step": 2998
+    },
+    {
+      "epoch": 1.185329711039763,
+      "grad_norm": 0.45935108094182536,
+      "learning_rate": 4.968666945702338e-06,
+      "loss": 0.5511,
+      "step": 2999
+    },
+    {
+      "epoch": 1.1857248703383552,
+      "grad_norm": 0.4551811442826113,
+      "learning_rate": 4.9686422137380265e-06,
+      "loss": 0.5646,
+      "step": 3000
+    },
+    {
+      "epoch": 1.1861200296369474,
+      "grad_norm": 0.45211507022800734,
+      "learning_rate": 4.968617472078388e-06,
+      "loss": 0.5573,
+      "step": 3001
+    },
+    {
+      "epoch": 1.1865151889355396,
+      "grad_norm": 0.451017086480957,
+      "learning_rate": 4.968592720723518e-06,
+      "loss": 0.5851,
+      "step": 3002
+    },
+    {
+      "epoch": 1.1869103482341319,
+      "grad_norm": 0.45071015068421694,
+      "learning_rate": 4.968567959673515e-06,
+      "loss": 0.5484,
+      "step": 3003
+    },
+    {
+      "epoch": 1.1873055075327241,
+      "grad_norm": 0.47688770951933446,
+      "learning_rate": 4.968543188928476e-06,
+      "loss": 0.577,
+      "step": 3004
+    },
+    {
+      "epoch": 1.1877006668313164,
+      "grad_norm": 0.43827997699657956,
+      "learning_rate": 4.9685184084885e-06,
+      "loss": 0.5566,
+      "step": 3005
+    },
+    {
+      "epoch": 1.1880958261299086,
+      "grad_norm": 0.4566633468170921,
+      "learning_rate": 4.968493618353681e-06,
+      "loss": 0.5539,
+      "step": 3006
+    },
+    {
+      "epoch": 1.1884909854285008,
+      "grad_norm": 0.46120055132005444,
+      "learning_rate": 4.968468818524118e-06,
+      "loss": 0.5704,
+      "step": 3007
+    },
+    {
+      "epoch": 1.188886144727093,
+      "grad_norm": 0.44778171241028825,
+      "learning_rate": 4.968444008999909e-06,
+      "loss": 0.5855,
+      "step": 3008
+    },
+    {
+      "epoch": 1.1892813040256853,
+      "grad_norm": 0.4750224432409252,
+      "learning_rate": 4.9684191897811505e-06,
+      "loss": 0.5565,
+      "step": 3009
+    },
+    {
+      "epoch": 1.1896764633242776,
+      "grad_norm": 0.4458948426804506,
+      "learning_rate": 4.968394360867941e-06,
+      "loss": 0.5585,
+      "step": 3010
+    },
+    {
+      "epoch": 1.1900716226228698,
+      "grad_norm": 0.46227629169335344,
+      "learning_rate": 4.968369522260377e-06,
+      "loss": 0.5754,
+      "step": 3011
+    },
+    {
+      "epoch": 1.190466781921462,
+      "grad_norm": 0.46386108301772255,
+      "learning_rate": 4.968344673958556e-06,
+      "loss": 0.5743,
+      "step": 3012
+    },
+    {
+      "epoch": 1.1908619412200543,
+      "grad_norm": 0.45605726481124903,
+      "learning_rate": 4.968319815962577e-06,
+      "loss": 0.5698,
+      "step": 3013
+    },
+    {
+      "epoch": 1.1912571005186465,
+      "grad_norm": 0.4430878703628007,
+      "learning_rate": 4.968294948272535e-06,
+      "loss": 0.5432,
+      "step": 3014
+    },
+    {
+      "epoch": 1.1916522598172388,
+      "grad_norm": 0.45206251975568795,
+      "learning_rate": 4.96827007088853e-06,
+      "loss": 0.571,
+      "step": 3015
+    },
+    {
+      "epoch": 1.192047419115831,
+      "grad_norm": 0.4552471461255095,
+      "learning_rate": 4.968245183810659e-06,
+      "loss": 0.5691,
+      "step": 3016
+    },
+    {
+      "epoch": 1.1924425784144232,
+      "grad_norm": 0.4432473462755384,
+      "learning_rate": 4.968220287039021e-06,
+      "loss": 0.5401,
+      "step": 3017
+    },
+    {
+      "epoch": 1.1928377377130155,
+      "grad_norm": 0.44587530242670237,
+      "learning_rate": 4.9681953805737106e-06,
+      "loss": 0.5461,
+      "step": 3018
+    },
+    {
+      "epoch": 1.1932328970116077,
+      "grad_norm": 0.47050677473839214,
+      "learning_rate": 4.968170464414828e-06,
+      "loss": 0.563,
+      "step": 3019
+    },
+    {
+      "epoch": 1.1936280563102,
+      "grad_norm": 0.44130303947137217,
+      "learning_rate": 4.968145538562471e-06,
+      "loss": 0.5693,
+      "step": 3020
+    },
+    {
+      "epoch": 1.1940232156087922,
+      "grad_norm": 0.4544886703368673,
+      "learning_rate": 4.968120603016737e-06,
+      "loss": 0.5653,
+      "step": 3021
+    },
+    {
+      "epoch": 1.1944183749073845,
+      "grad_norm": 0.4896624605231788,
+      "learning_rate": 4.968095657777724e-06,
+      "loss": 0.5822,
+      "step": 3022
+    },
+    {
+      "epoch": 1.1948135342059767,
+      "grad_norm": 0.4703893538726537,
+      "learning_rate": 4.968070702845529e-06,
+      "loss": 0.5384,
+      "step": 3023
+    },
+    {
+      "epoch": 1.195208693504569,
+      "grad_norm": 0.5189597430778528,
+      "learning_rate": 4.968045738220252e-06,
+      "loss": 0.578,
+      "step": 3024
+    },
+    {
+      "epoch": 1.1956038528031612,
+      "grad_norm": 0.44433555063115776,
+      "learning_rate": 4.96802076390199e-06,
+      "loss": 0.5672,
+      "step": 3025
+    },
+    {
+      "epoch": 1.1959990121017534,
+      "grad_norm": 0.44690653875268826,
+      "learning_rate": 4.96799577989084e-06,
+      "loss": 0.5483,
+      "step": 3026
+    },
+    {
+      "epoch": 1.1963941714003457,
+      "grad_norm": 0.4456693441690437,
+      "learning_rate": 4.967970786186903e-06,
+      "loss": 0.5917,
+      "step": 3027
+    },
+    {
+      "epoch": 1.196789330698938,
+      "grad_norm": 0.4656615654209235,
+      "learning_rate": 4.967945782790275e-06,
+      "loss": 0.6074,
+      "step": 3028
+    },
+    {
+      "epoch": 1.1971844899975301,
+      "grad_norm": 0.4679283421472829,
+      "learning_rate": 4.967920769701053e-06,
+      "loss": 0.5562,
+      "step": 3029
+    },
+    {
+      "epoch": 1.1975796492961226,
+      "grad_norm": 0.4599878221931696,
+      "learning_rate": 4.967895746919339e-06,
+      "loss": 0.5898,
+      "step": 3030
+    },
+    {
+      "epoch": 1.1979748085947148,
+      "grad_norm": 0.44364183466084445,
+      "learning_rate": 4.967870714445227e-06,
+      "loss": 0.5518,
+      "step": 3031
+    },
+    {
+      "epoch": 1.198369967893307,
+      "grad_norm": 0.45964094956857143,
+      "learning_rate": 4.967845672278819e-06,
+      "loss": 0.5554,
+      "step": 3032
+    },
+    {
+      "epoch": 1.1987651271918993,
+      "grad_norm": 0.45133637183308356,
+      "learning_rate": 4.967820620420211e-06,
+      "loss": 0.5651,
+      "step": 3033
+    },
+    {
+      "epoch": 1.1991602864904916,
+      "grad_norm": 0.4427547623086265,
+      "learning_rate": 4.9677955588695025e-06,
+      "loss": 0.5748,
+      "step": 3034
+    },
+    {
+      "epoch": 1.1995554457890838,
+      "grad_norm": 0.6682331551539357,
+      "learning_rate": 4.967770487626791e-06,
+      "loss": 0.5567,
+      "step": 3035
+    },
+    {
+      "epoch": 1.199950605087676,
+      "grad_norm": 0.47306363628338366,
+      "learning_rate": 4.967745406692176e-06,
+      "loss": 0.5675,
+      "step": 3036
+    },
+    {
+      "epoch": 1.2003457643862683,
+      "grad_norm": 0.48270419923550345,
+      "learning_rate": 4.967720316065756e-06,
+      "loss": 0.5661,
+      "step": 3037
+    },
+    {
+      "epoch": 1.2007409236848605,
+      "grad_norm": 0.4423873238428253,
+      "learning_rate": 4.9676952157476285e-06,
+      "loss": 0.5681,
+      "step": 3038
+    },
+    {
+      "epoch": 1.2011360829834528,
+      "grad_norm": 0.4487466875408833,
+      "learning_rate": 4.967670105737892e-06,
+      "loss": 0.5489,
+      "step": 3039
+    },
+    {
+      "epoch": 1.201531242282045,
+      "grad_norm": 0.5042106639999401,
+      "learning_rate": 4.967644986036647e-06,
+      "loss": 0.5611,
+      "step": 3040
+    },
+    {
+      "epoch": 1.2019264015806372,
+      "grad_norm": 0.4759632066357272,
+      "learning_rate": 4.96761985664399e-06,
+      "loss": 0.5622,
+      "step": 3041
+    },
+    {
+      "epoch": 1.2023215608792295,
+      "grad_norm": 0.47451217869994206,
+      "learning_rate": 4.967594717560022e-06,
+      "loss": 0.5804,
+      "step": 3042
+    },
+    {
+      "epoch": 1.2027167201778217,
+      "grad_norm": 0.4593740341418387,
+      "learning_rate": 4.967569568784839e-06,
+      "loss": 0.5563,
+      "step": 3043
+    },
+    {
+      "epoch": 1.203111879476414,
+      "grad_norm": 0.4790716838554254,
+      "learning_rate": 4.967544410318541e-06,
+      "loss": 0.5562,
+      "step": 3044
+    },
+    {
+      "epoch": 1.2035070387750062,
+      "grad_norm": 0.44309498656578405,
+      "learning_rate": 4.967519242161227e-06,
+      "loss": 0.5516,
+      "step": 3045
+    },
+    {
+      "epoch": 1.2039021980735984,
+      "grad_norm": 0.4895189026135667,
+      "learning_rate": 4.967494064312996e-06,
+      "loss": 0.5667,
+      "step": 3046
+    },
+    {
+      "epoch": 1.2042973573721907,
+      "grad_norm": 0.4561602202621074,
+      "learning_rate": 4.967468876773948e-06,
+      "loss": 0.5629,
+      "step": 3047
+    },
+    {
+      "epoch": 1.204692516670783,
+      "grad_norm": 0.4697256795259107,
+      "learning_rate": 4.9674436795441795e-06,
+      "loss": 0.5786,
+      "step": 3048
+    },
+    {
+      "epoch": 1.2050876759693752,
+      "grad_norm": 0.5259340703235517,
+      "learning_rate": 4.96741847262379e-06,
+      "loss": 0.5583,
+      "step": 3049
+    },
+    {
+      "epoch": 1.2054828352679674,
+      "grad_norm": 0.4894487391041935,
+      "learning_rate": 4.967393256012879e-06,
+      "loss": 0.5904,
+      "step": 3050
+    },
+    {
+      "epoch": 1.2058779945665596,
+      "grad_norm": 0.46711138204567687,
+      "learning_rate": 4.967368029711547e-06,
+      "loss": 0.5702,
+      "step": 3051
+    },
+    {
+      "epoch": 1.2062731538651519,
+      "grad_norm": 0.4888902860202162,
+      "learning_rate": 4.96734279371989e-06,
+      "loss": 0.5771,
+      "step": 3052
+    },
+    {
+      "epoch": 1.2066683131637441,
+      "grad_norm": 0.5333072923347787,
+      "learning_rate": 4.96731754803801e-06,
+      "loss": 0.561,
+      "step": 3053
+    },
+    {
+      "epoch": 1.2070634724623364,
+      "grad_norm": 0.46654604201036526,
+      "learning_rate": 4.967292292666004e-06,
+      "loss": 0.5799,
+      "step": 3054
+    },
+    {
+      "epoch": 1.2074586317609286,
+      "grad_norm": 0.45887016609605846,
+      "learning_rate": 4.967267027603972e-06,
+      "loss": 0.5672,
+      "step": 3055
+    },
+    {
+      "epoch": 1.2078537910595208,
+      "grad_norm": 0.477396686518973,
+      "learning_rate": 4.967241752852015e-06,
+      "loss": 0.5846,
+      "step": 3056
+    },
+    {
+      "epoch": 1.208248950358113,
+      "grad_norm": 0.4613218129856247,
+      "learning_rate": 4.967216468410229e-06,
+      "loss": 0.5661,
+      "step": 3057
+    },
+    {
+      "epoch": 1.2086441096567053,
+      "grad_norm": 0.5240029995870954,
+      "learning_rate": 4.9671911742787145e-06,
+      "loss": 0.5709,
+      "step": 3058
+    },
+    {
+      "epoch": 1.2090392689552976,
+      "grad_norm": 0.4363976879978063,
+      "learning_rate": 4.967165870457573e-06,
+      "loss": 0.5458,
+      "step": 3059
+    },
+    {
+      "epoch": 1.2094344282538898,
+      "grad_norm": 0.43183191736087323,
+      "learning_rate": 4.9671405569469e-06,
+      "loss": 0.5615,
+      "step": 3060
+    },
+    {
+      "epoch": 1.209829587552482,
+      "grad_norm": 0.48194750159483546,
+      "learning_rate": 4.967115233746798e-06,
+      "loss": 0.5692,
+      "step": 3061
+    },
+    {
+      "epoch": 1.2102247468510743,
+      "grad_norm": 0.4403230569615395,
+      "learning_rate": 4.967089900857366e-06,
+      "loss": 0.5541,
+      "step": 3062
+    },
+    {
+      "epoch": 1.2106199061496665,
+      "grad_norm": 0.4521141643968015,
+      "learning_rate": 4.9670645582787025e-06,
+      "loss": 0.5683,
+      "step": 3063
+    },
+    {
+      "epoch": 1.2110150654482588,
+      "grad_norm": 0.4436203962884348,
+      "learning_rate": 4.967039206010908e-06,
+      "loss": 0.5293,
+      "step": 3064
+    },
+    {
+      "epoch": 1.211410224746851,
+      "grad_norm": 0.4412702021651393,
+      "learning_rate": 4.967013844054081e-06,
+      "loss": 0.5575,
+      "step": 3065
+    },
+    {
+      "epoch": 1.2118053840454432,
+      "grad_norm": 0.44313069086594686,
+      "learning_rate": 4.966988472408322e-06,
+      "loss": 0.5701,
+      "step": 3066
+    },
+    {
+      "epoch": 1.2122005433440355,
+      "grad_norm": 0.45845037915893444,
+      "learning_rate": 4.96696309107373e-06,
+      "loss": 0.5817,
+      "step": 3067
+    },
+    {
+      "epoch": 1.2125957026426277,
+      "grad_norm": 0.4574367758060889,
+      "learning_rate": 4.966937700050405e-06,
+      "loss": 0.5705,
+      "step": 3068
+    },
+    {
+      "epoch": 1.21299086194122,
+      "grad_norm": 0.4471000171164327,
+      "learning_rate": 4.966912299338447e-06,
+      "loss": 0.5413,
+      "step": 3069
+    },
+    {
+      "epoch": 1.2133860212398122,
+      "grad_norm": 0.45685994308151356,
+      "learning_rate": 4.966886888937955e-06,
+      "loss": 0.5688,
+      "step": 3070
+    },
+    {
+      "epoch": 1.2137811805384044,
+      "grad_norm": 0.4397052448052631,
+      "learning_rate": 4.96686146884903e-06,
+      "loss": 0.5555,
+      "step": 3071
+    },
+    {
+      "epoch": 1.214176339836997,
+      "grad_norm": 0.4522674726127247,
+      "learning_rate": 4.96683603907177e-06,
+      "loss": 0.5723,
+      "step": 3072
+    },
+    {
+      "epoch": 1.2145714991355891,
+      "grad_norm": 0.46257064891670174,
+      "learning_rate": 4.966810599606277e-06,
+      "loss": 0.5795,
+      "step": 3073
+    },
+    {
+      "epoch": 1.2149666584341814,
+      "grad_norm": 0.4638428914792885,
+      "learning_rate": 4.9667851504526495e-06,
+      "loss": 0.5765,
+      "step": 3074
+    },
+    {
+      "epoch": 1.2153618177327736,
+      "grad_norm": 0.4633324633840189,
+      "learning_rate": 4.966759691610989e-06,
+      "loss": 0.5901,
+      "step": 3075
+    },
+    {
+      "epoch": 1.2157569770313659,
+      "grad_norm": 0.4694538016404579,
+      "learning_rate": 4.966734223081392e-06,
+      "loss": 0.5755,
+      "step": 3076
+    },
+    {
+      "epoch": 1.2161521363299581,
+      "grad_norm": 0.44870540089281646,
+      "learning_rate": 4.966708744863962e-06,
+      "loss": 0.5664,
+      "step": 3077
+    },
+    {
+      "epoch": 1.2165472956285504,
+      "grad_norm": 0.4633771864468751,
+      "learning_rate": 4.966683256958799e-06,
+      "loss": 0.5683,
+      "step": 3078
+    },
+    {
+      "epoch": 1.2169424549271426,
+      "grad_norm": 0.4746476377257439,
+      "learning_rate": 4.966657759366e-06,
+      "loss": 0.5647,
+      "step": 3079
+    },
+    {
+      "epoch": 1.2173376142257348,
+      "grad_norm": 0.45147878211938924,
+      "learning_rate": 4.966632252085669e-06,
+      "loss": 0.5752,
+      "step": 3080
+    },
+    {
+      "epoch": 1.217732773524327,
+      "grad_norm": 0.44156118894699065,
+      "learning_rate": 4.966606735117902e-06,
+      "loss": 0.5563,
+      "step": 3081
+    },
+    {
+      "epoch": 1.2181279328229193,
+      "grad_norm": 0.45196789781220703,
+      "learning_rate": 4.966581208462804e-06,
+      "loss": 0.5475,
+      "step": 3082
+    },
+    {
+      "epoch": 1.2185230921215116,
+      "grad_norm": 0.4442256161242522,
+      "learning_rate": 4.966555672120472e-06,
+      "loss": 0.5695,
+      "step": 3083
+    },
+    {
+      "epoch": 1.2189182514201038,
+      "grad_norm": 0.44730065552547726,
+      "learning_rate": 4.966530126091007e-06,
+      "loss": 0.5599,
+      "step": 3084
+    },
+    {
+      "epoch": 1.219313410718696,
+      "grad_norm": 0.4544579855898216,
+      "learning_rate": 4.966504570374509e-06,
+      "loss": 0.5451,
+      "step": 3085
+    },
+    {
+      "epoch": 1.2197085700172883,
+      "grad_norm": 0.45542641771326065,
+      "learning_rate": 4.9664790049710795e-06,
+      "loss": 0.5691,
+      "step": 3086
+    },
+    {
+      "epoch": 1.2201037293158805,
+      "grad_norm": 0.46855478341533124,
+      "learning_rate": 4.966453429880818e-06,
+      "loss": 0.5757,
+      "step": 3087
+    },
+    {
+      "epoch": 1.2204988886144728,
+      "grad_norm": 0.43240914194877017,
+      "learning_rate": 4.966427845103825e-06,
+      "loss": 0.5719,
+      "step": 3088
+    },
+    {
+      "epoch": 1.220894047913065,
+      "grad_norm": 0.4182707219417738,
+      "learning_rate": 4.966402250640201e-06,
+      "loss": 0.5522,
+      "step": 3089
+    },
+    {
+      "epoch": 1.2212892072116572,
+      "grad_norm": 0.44846470069043676,
+      "learning_rate": 4.9663766464900465e-06,
+      "loss": 0.5701,
+      "step": 3090
+    },
+    {
+      "epoch": 1.2216843665102495,
+      "grad_norm": 0.48379385795172264,
+      "learning_rate": 4.966351032653463e-06,
+      "loss": 0.5632,
+      "step": 3091
+    },
+    {
+      "epoch": 1.2220795258088417,
+      "grad_norm": 0.4608900432943671,
+      "learning_rate": 4.966325409130549e-06,
+      "loss": 0.5672,
+      "step": 3092
+    },
+    {
+      "epoch": 1.222474685107434,
+      "grad_norm": 0.465833480879931,
+      "learning_rate": 4.9662997759214074e-06,
+      "loss": 0.5651,
+      "step": 3093
+    },
+    {
+      "epoch": 1.2228698444060262,
+      "grad_norm": 0.45118044139514435,
+      "learning_rate": 4.966274133026138e-06,
+      "loss": 0.573,
+      "step": 3094
+    },
+    {
+      "epoch": 1.2232650037046184,
+      "grad_norm": 0.4421531493404346,
+      "learning_rate": 4.966248480444841e-06,
+      "loss": 0.5657,
+      "step": 3095
+    },
+    {
+      "epoch": 1.2236601630032107,
+      "grad_norm": 0.44492466262445596,
+      "learning_rate": 4.966222818177617e-06,
+      "loss": 0.5542,
+      "step": 3096
+    },
+    {
+      "epoch": 1.224055322301803,
+      "grad_norm": 0.4389512642580722,
+      "learning_rate": 4.966197146224568e-06,
+      "loss": 0.5646,
+      "step": 3097
+    },
+    {
+      "epoch": 1.2244504816003952,
+      "grad_norm": 0.44609105957073,
+      "learning_rate": 4.966171464585794e-06,
+      "loss": 0.5512,
+      "step": 3098
+    },
+    {
+      "epoch": 1.2248456408989874,
+      "grad_norm": 0.45028716091193877,
+      "learning_rate": 4.966145773261396e-06,
+      "loss": 0.5528,
+      "step": 3099
+    },
+    {
+      "epoch": 1.2252408001975796,
+      "grad_norm": 0.4430745962507206,
+      "learning_rate": 4.966120072251475e-06,
+      "loss": 0.5486,
+      "step": 3100
+    },
+    {
+      "epoch": 1.2256359594961719,
+      "grad_norm": 0.4565725474643139,
+      "learning_rate": 4.966094361556132e-06,
+      "loss": 0.5637,
+      "step": 3101
+    },
+    {
+      "epoch": 1.2260311187947641,
+      "grad_norm": 0.4603270298388148,
+      "learning_rate": 4.966068641175469e-06,
+      "loss": 0.5742,
+      "step": 3102
+    },
+    {
+      "epoch": 1.2264262780933564,
+      "grad_norm": 0.4413064560746586,
+      "learning_rate": 4.966042911109584e-06,
+      "loss": 0.5513,
+      "step": 3103
+    },
+    {
+      "epoch": 1.2268214373919486,
+      "grad_norm": 0.45702241651020653,
+      "learning_rate": 4.9660171713585805e-06,
+      "loss": 0.5678,
+      "step": 3104
+    },
+    {
+      "epoch": 1.2272165966905408,
+      "grad_norm": 0.47067105579936214,
+      "learning_rate": 4.965991421922559e-06,
+      "loss": 0.5632,
+      "step": 3105
+    },
+    {
+      "epoch": 1.227611755989133,
+      "grad_norm": 0.5372601558221561,
+      "learning_rate": 4.965965662801621e-06,
+      "loss": 0.5575,
+      "step": 3106
+    },
+    {
+      "epoch": 1.2280069152877253,
+      "grad_norm": 0.48016720590726975,
+      "learning_rate": 4.965939893995867e-06,
+      "loss": 0.574,
+      "step": 3107
+    },
+    {
+      "epoch": 1.2284020745863176,
+      "grad_norm": 0.43658120984806364,
+      "learning_rate": 4.965914115505398e-06,
+      "loss": 0.5547,
+      "step": 3108
+    },
+    {
+      "epoch": 1.2287972338849098,
+      "grad_norm": 0.4464718194752851,
+      "learning_rate": 4.965888327330316e-06,
+      "loss": 0.5738,
+      "step": 3109
+    },
+    {
+      "epoch": 1.229192393183502,
+      "grad_norm": 0.44348094873374344,
+      "learning_rate": 4.9658625294707226e-06,
+      "loss": 0.5558,
+      "step": 3110
+    },
+    {
+      "epoch": 1.2295875524820943,
+      "grad_norm": 0.45414022219810424,
+      "learning_rate": 4.965836721926718e-06,
+      "loss": 0.5591,
+      "step": 3111
+    },
+    {
+      "epoch": 1.2299827117806865,
+      "grad_norm": 0.45465803398220034,
+      "learning_rate": 4.965810904698404e-06,
+      "loss": 0.5784,
+      "step": 3112
+    },
+    {
+      "epoch": 1.2303778710792788,
+      "grad_norm": 0.42507296664604205,
+      "learning_rate": 4.965785077785882e-06,
+      "loss": 0.5589,
+      "step": 3113
+    },
+    {
+      "epoch": 1.230773030377871,
+      "grad_norm": 0.4451171220895154,
+      "learning_rate": 4.965759241189254e-06,
+      "loss": 0.5653,
+      "step": 3114
+    },
+    {
+      "epoch": 1.2311681896764632,
+      "grad_norm": 0.44320611707236784,
+      "learning_rate": 4.965733394908621e-06,
+      "loss": 0.5764,
+      "step": 3115
+    },
+    {
+      "epoch": 1.2315633489750555,
+      "grad_norm": 0.4467513021648378,
+      "learning_rate": 4.965707538944085e-06,
+      "loss": 0.5681,
+      "step": 3116
+    },
+    {
+      "epoch": 1.2319585082736477,
+      "grad_norm": 0.4962560415068218,
+      "learning_rate": 4.9656816732957454e-06,
+      "loss": 0.5772,
+      "step": 3117
+    },
+    {
+      "epoch": 1.23235366757224,
+      "grad_norm": 0.4393887859931017,
+      "learning_rate": 4.965655797963707e-06,
+      "loss": 0.5647,
+      "step": 3118
+    },
+    {
+      "epoch": 1.2327488268708322,
+      "grad_norm": 0.464778893222507,
+      "learning_rate": 4.965629912948069e-06,
+      "loss": 0.5851,
+      "step": 3119
+    },
+    {
+      "epoch": 1.2331439861694244,
+      "grad_norm": 0.4525476192977244,
+      "learning_rate": 4.965604018248934e-06,
+      "loss": 0.578,
+      "step": 3120
+    },
+    {
+      "epoch": 1.2335391454680167,
+      "grad_norm": 0.461394081784411,
+      "learning_rate": 4.965578113866404e-06,
+      "loss": 0.561,
+      "step": 3121
+    },
+    {
+      "epoch": 1.233934304766609,
+      "grad_norm": 0.43936664461165786,
+      "learning_rate": 4.96555219980058e-06,
+      "loss": 0.5476,
+      "step": 3122
+    },
+    {
+      "epoch": 1.2343294640652012,
+      "grad_norm": 0.4509074107094845,
+      "learning_rate": 4.965526276051564e-06,
+      "loss": 0.567,
+      "step": 3123
+    },
+    {
+      "epoch": 1.2347246233637934,
+      "grad_norm": 0.45459696038357283,
+      "learning_rate": 4.965500342619458e-06,
+      "loss": 0.5763,
+      "step": 3124
+    },
+    {
+      "epoch": 1.2351197826623859,
+      "grad_norm": 0.44821748870137185,
+      "learning_rate": 4.965474399504364e-06,
+      "loss": 0.5668,
+      "step": 3125
+    },
+    {
+      "epoch": 1.235514941960978,
+      "grad_norm": 0.46197320609039444,
+      "learning_rate": 4.965448446706384e-06,
+      "loss": 0.5682,
+      "step": 3126
+    },
+    {
+      "epoch": 1.2359101012595703,
+      "grad_norm": 0.4396093700838116,
+      "learning_rate": 4.96542248422562e-06,
+      "loss": 0.5632,
+      "step": 3127
+    },
+    {
+      "epoch": 1.2363052605581626,
+      "grad_norm": 0.4308392401888905,
+      "learning_rate": 4.965396512062171e-06,
+      "loss": 0.5433,
+      "step": 3128
+    },
+    {
+      "epoch": 1.2367004198567548,
+      "grad_norm": 0.462787889383493,
+      "learning_rate": 4.9653705302161446e-06,
+      "loss": 0.5841,
+      "step": 3129
+    },
+    {
+      "epoch": 1.237095579155347,
+      "grad_norm": 0.43840660693999445,
+      "learning_rate": 4.965344538687638e-06,
+      "loss": 0.5452,
+      "step": 3130
+    },
+    {
+      "epoch": 1.2374907384539393,
+      "grad_norm": 0.4602319352251958,
+      "learning_rate": 4.965318537476756e-06,
+      "loss": 0.5811,
+      "step": 3131
+    },
+    {
+      "epoch": 1.2378858977525315,
+      "grad_norm": 0.45813906029186197,
+      "learning_rate": 4.9652925265836e-06,
+      "loss": 0.5694,
+      "step": 3132
+    },
+    {
+      "epoch": 1.2382810570511238,
+      "grad_norm": 0.4492594870763079,
+      "learning_rate": 4.965266506008271e-06,
+      "loss": 0.5884,
+      "step": 3133
+    },
+    {
+      "epoch": 1.238676216349716,
+      "grad_norm": 0.4627556622877327,
+      "learning_rate": 4.9652404757508726e-06,
+      "loss": 0.571,
+      "step": 3134
+    },
+    {
+      "epoch": 1.2390713756483083,
+      "grad_norm": 0.4629993337469515,
+      "learning_rate": 4.965214435811506e-06,
+      "loss": 0.5669,
+      "step": 3135
+    },
+    {
+      "epoch": 1.2394665349469005,
+      "grad_norm": 0.46398715244786704,
+      "learning_rate": 4.965188386190275e-06,
+      "loss": 0.5735,
+      "step": 3136
+    },
+    {
+      "epoch": 1.2398616942454928,
+      "grad_norm": 0.44440876198440876,
+      "learning_rate": 4.965162326887281e-06,
+      "loss": 0.5669,
+      "step": 3137
+    },
+    {
+      "epoch": 1.240256853544085,
+      "grad_norm": 0.49597914064913107,
+      "learning_rate": 4.965136257902626e-06,
+      "loss": 0.5807,
+      "step": 3138
+    },
+    {
+      "epoch": 1.2406520128426772,
+      "grad_norm": 0.44555439287267584,
+      "learning_rate": 4.965110179236412e-06,
+      "loss": 0.5439,
+      "step": 3139
+    },
+    {
+      "epoch": 1.2410471721412695,
+      "grad_norm": 0.45626121734971314,
+      "learning_rate": 4.965084090888743e-06,
+      "loss": 0.5689,
+      "step": 3140
+    },
+    {
+      "epoch": 1.2414423314398617,
+      "grad_norm": 0.4363299527581059,
+      "learning_rate": 4.96505799285972e-06,
+      "loss": 0.5476,
+      "step": 3141
+    },
+    {
+      "epoch": 1.241837490738454,
+      "grad_norm": 0.4671748329287142,
+      "learning_rate": 4.9650318851494465e-06,
+      "loss": 0.5817,
+      "step": 3142
+    },
+    {
+      "epoch": 1.2422326500370462,
+      "grad_norm": 0.472754493128498,
+      "learning_rate": 4.965005767758024e-06,
+      "loss": 0.5701,
+      "step": 3143
+    },
+    {
+      "epoch": 1.2426278093356384,
+      "grad_norm": 0.4760983983168848,
+      "learning_rate": 4.964979640685557e-06,
+      "loss": 0.5804,
+      "step": 3144
+    },
+    {
+      "epoch": 1.2430229686342307,
+      "grad_norm": 0.4717372783720608,
+      "learning_rate": 4.964953503932146e-06,
+      "loss": 0.5867,
+      "step": 3145
+    },
+    {
+      "epoch": 1.243418127932823,
+      "grad_norm": 0.45391804628270654,
+      "learning_rate": 4.964927357497894e-06,
+      "loss": 0.5214,
+      "step": 3146
+    },
+    {
+      "epoch": 1.2438132872314152,
+      "grad_norm": 0.4407103558661319,
+      "learning_rate": 4.964901201382905e-06,
+      "loss": 0.572,
+      "step": 3147
+    },
+    {
+      "epoch": 1.2442084465300074,
+      "grad_norm": 0.45045690316166226,
+      "learning_rate": 4.96487503558728e-06,
+      "loss": 0.5663,
+      "step": 3148
+    },
+    {
+      "epoch": 1.2446036058285996,
+      "grad_norm": 0.4434691882497208,
+      "learning_rate": 4.964848860111122e-06,
+      "loss": 0.5473,
+      "step": 3149
+    },
+    {
+      "epoch": 1.2449987651271919,
+      "grad_norm": 0.4460639321688961,
+      "learning_rate": 4.964822674954536e-06,
+      "loss": 0.5719,
+      "step": 3150
+    },
+    {
+      "epoch": 1.2453939244257841,
+      "grad_norm": 0.4709745072765176,
+      "learning_rate": 4.964796480117623e-06,
+      "loss": 0.5756,
+      "step": 3151
+    },
+    {
+      "epoch": 1.2457890837243764,
+      "grad_norm": 0.47049229358105993,
+      "learning_rate": 4.9647702756004855e-06,
+      "loss": 0.5714,
+      "step": 3152
+    },
+    {
+      "epoch": 1.2461842430229686,
+      "grad_norm": 0.4649418473083078,
+      "learning_rate": 4.964744061403227e-06,
+      "loss": 0.5749,
+      "step": 3153
+    },
+    {
+      "epoch": 1.2465794023215608,
+      "grad_norm": 0.47164232942914297,
+      "learning_rate": 4.964717837525951e-06,
+      "loss": 0.5735,
+      "step": 3154
+    },
+    {
+      "epoch": 1.246974561620153,
+      "grad_norm": 0.4529213434049233,
+      "learning_rate": 4.9646916039687594e-06,
+      "loss": 0.5836,
+      "step": 3155
+    },
+    {
+      "epoch": 1.2473697209187453,
+      "grad_norm": 0.45289976023026995,
+      "learning_rate": 4.964665360731757e-06,
+      "loss": 0.5695,
+      "step": 3156
+    },
+    {
+      "epoch": 1.2477648802173376,
+      "grad_norm": 0.4780170631528699,
+      "learning_rate": 4.964639107815044e-06,
+      "loss": 0.5602,
+      "step": 3157
+    },
+    {
+      "epoch": 1.2481600395159298,
+      "grad_norm": 0.44393148384313114,
+      "learning_rate": 4.964612845218726e-06,
+      "loss": 0.5618,
+      "step": 3158
+    },
+    {
+      "epoch": 1.248555198814522,
+      "grad_norm": 0.4486878703274451,
+      "learning_rate": 4.964586572942905e-06,
+      "loss": 0.5715,
+      "step": 3159
+    },
+    {
+      "epoch": 1.2489503581131143,
+      "grad_norm": 0.45262089196133376,
+      "learning_rate": 4.964560290987686e-06,
+      "loss": 0.5538,
+      "step": 3160
+    },
+    {
+      "epoch": 1.2493455174117065,
+      "grad_norm": 0.46646019654372667,
+      "learning_rate": 4.964533999353169e-06,
+      "loss": 0.5561,
+      "step": 3161
+    },
+    {
+      "epoch": 1.2497406767102988,
+      "grad_norm": 0.46395577187095943,
+      "learning_rate": 4.96450769803946e-06,
+      "loss": 0.5703,
+      "step": 3162
+    },
+    {
+      "epoch": 1.2501358360088912,
+      "grad_norm": 0.43225782469002116,
+      "learning_rate": 4.9644813870466605e-06,
+      "loss": 0.5475,
+      "step": 3163
+    },
+    {
+      "epoch": 1.2505309953074835,
+      "grad_norm": 0.4477412714097664,
+      "learning_rate": 4.9644550663748755e-06,
+      "loss": 0.588,
+      "step": 3164
+    },
+    {
+      "epoch": 1.2509261546060757,
+      "grad_norm": 0.4488220491257586,
+      "learning_rate": 4.964428736024207e-06,
+      "loss": 0.5456,
+      "step": 3165
+    },
+    {
+      "epoch": 1.251321313904668,
+      "grad_norm": 0.4591216638115739,
+      "learning_rate": 4.964402395994759e-06,
+      "loss": 0.5962,
+      "step": 3166
+    },
+    {
+      "epoch": 1.2517164732032602,
+      "grad_norm": 0.4496651980462876,
+      "learning_rate": 4.964376046286635e-06,
+      "loss": 0.5773,
+      "step": 3167
+    },
+    {
+      "epoch": 1.2521116325018524,
+      "grad_norm": 0.43068834285244056,
+      "learning_rate": 4.964349686899938e-06,
+      "loss": 0.5649,
+      "step": 3168
+    },
+    {
+      "epoch": 1.2525067918004447,
+      "grad_norm": 0.43656858666444814,
+      "learning_rate": 4.964323317834772e-06,
+      "loss": 0.5606,
+      "step": 3169
+    },
+    {
+      "epoch": 1.252901951099037,
+      "grad_norm": 0.4465521206805361,
+      "learning_rate": 4.96429693909124e-06,
+      "loss": 0.5581,
+      "step": 3170
+    },
+    {
+      "epoch": 1.2532971103976291,
+      "grad_norm": 0.4356583772476069,
+      "learning_rate": 4.964270550669447e-06,
+      "loss": 0.554,
+      "step": 3171
+    },
+    {
+      "epoch": 1.2536922696962214,
+      "grad_norm": 0.4538626705056506,
+      "learning_rate": 4.964244152569495e-06,
+      "loss": 0.5724,
+      "step": 3172
+    },
+    {
+      "epoch": 1.2540874289948136,
+      "grad_norm": 0.44499801851560644,
+      "learning_rate": 4.964217744791489e-06,
+      "loss": 0.5675,
+      "step": 3173
+    },
+    {
+      "epoch": 1.2544825882934059,
+      "grad_norm": 0.44634942356882007,
+      "learning_rate": 4.964191327335531e-06,
+      "loss": 0.5507,
+      "step": 3174
+    },
+    {
+      "epoch": 1.254877747591998,
+      "grad_norm": 0.47867361712053463,
+      "learning_rate": 4.964164900201726e-06,
+      "loss": 0.5587,
+      "step": 3175
+    },
+    {
+      "epoch": 1.2552729068905903,
+      "grad_norm": 0.4555734443969186,
+      "learning_rate": 4.964138463390178e-06,
+      "loss": 0.5646,
+      "step": 3176
+    },
+    {
+      "epoch": 1.2556680661891826,
+      "grad_norm": 0.4584704800532935,
+      "learning_rate": 4.964112016900991e-06,
+      "loss": 0.5517,
+      "step": 3177
+    },
+    {
+      "epoch": 1.2560632254877748,
+      "grad_norm": 0.45549434221447166,
+      "learning_rate": 4.964085560734267e-06,
+      "loss": 0.5771,
+      "step": 3178
+    },
+    {
+      "epoch": 1.256458384786367,
+      "grad_norm": 0.44914572871190456,
+      "learning_rate": 4.964059094890112e-06,
+      "loss": 0.5617,
+      "step": 3179
+    },
+    {
+      "epoch": 1.2568535440849593,
+      "grad_norm": 0.4512882813462357,
+      "learning_rate": 4.964032619368629e-06,
+      "loss": 0.5723,
+      "step": 3180
+    },
+    {
+      "epoch": 1.2572487033835515,
+      "grad_norm": 0.4509607204943829,
+      "learning_rate": 4.964006134169922e-06,
+      "loss": 0.5643,
+      "step": 3181
+    },
+    {
+      "epoch": 1.2576438626821438,
+      "grad_norm": 0.45921297434875,
+      "learning_rate": 4.9639796392940955e-06,
+      "loss": 0.5844,
+      "step": 3182
+    },
+    {
+      "epoch": 1.258039021980736,
+      "grad_norm": 0.48942904209316174,
+      "learning_rate": 4.963953134741253e-06,
+      "loss": 0.5802,
+      "step": 3183
+    },
+    {
+      "epoch": 1.2584341812793283,
+      "grad_norm": 0.45854271793023527,
+      "learning_rate": 4.963926620511497e-06,
+      "loss": 0.5815,
+      "step": 3184
+    },
+    {
+      "epoch": 1.2588293405779205,
+      "grad_norm": 0.4560649261852496,
+      "learning_rate": 4.963900096604936e-06,
+      "loss": 0.5534,
+      "step": 3185
+    },
+    {
+      "epoch": 1.2592244998765127,
+      "grad_norm": 0.4518649540196938,
+      "learning_rate": 4.9638735630216704e-06,
+      "loss": 0.5754,
+      "step": 3186
+    },
+    {
+      "epoch": 1.259619659175105,
+      "grad_norm": 0.6754753158823785,
+      "learning_rate": 4.963847019761806e-06,
+      "loss": 0.5704,
+      "step": 3187
+    },
+    {
+      "epoch": 1.2600148184736972,
+      "grad_norm": 0.452944137497009,
+      "learning_rate": 4.9638204668254465e-06,
+      "loss": 0.5644,
+      "step": 3188
+    },
+    {
+      "epoch": 1.2604099777722895,
+      "grad_norm": 0.44138818835118054,
+      "learning_rate": 4.9637939042126965e-06,
+      "loss": 0.5624,
+      "step": 3189
+    },
+    {
+      "epoch": 1.2608051370708817,
+      "grad_norm": 0.467795887447353,
+      "learning_rate": 4.96376733192366e-06,
+      "loss": 0.5909,
+      "step": 3190
+    },
+    {
+      "epoch": 1.261200296369474,
+      "grad_norm": 0.4567278109168832,
+      "learning_rate": 4.963740749958441e-06,
+      "loss": 0.5645,
+      "step": 3191
+    },
+    {
+      "epoch": 1.2615954556680662,
+      "grad_norm": 0.603665506544785,
+      "learning_rate": 4.9637141583171456e-06,
+      "loss": 0.5607,
+      "step": 3192
+    },
+    {
+      "epoch": 1.2619906149666584,
+      "grad_norm": 0.44076800996835624,
+      "learning_rate": 4.9636875569998756e-06,
+      "loss": 0.5479,
+      "step": 3193
+    },
+    {
+      "epoch": 1.2623857742652507,
+      "grad_norm": 0.4597963136042489,
+      "learning_rate": 4.963660946006737e-06,
+      "loss": 0.5677,
+      "step": 3194
+    },
+    {
+      "epoch": 1.262780933563843,
+      "grad_norm": 0.45342024851568213,
+      "learning_rate": 4.963634325337836e-06,
+      "loss": 0.5452,
+      "step": 3195
+    },
+    {
+      "epoch": 1.2631760928624352,
+      "grad_norm": 0.44313171394307665,
+      "learning_rate": 4.9636076949932736e-06,
+      "loss": 0.5699,
+      "step": 3196
+    },
+    {
+      "epoch": 1.2635712521610274,
+      "grad_norm": 0.45404430511750704,
+      "learning_rate": 4.9635810549731565e-06,
+      "loss": 0.579,
+      "step": 3197
+    },
+    {
+      "epoch": 1.2639664114596196,
+      "grad_norm": 0.4739171320560697,
+      "learning_rate": 4.9635544052775895e-06,
+      "loss": 0.5475,
+      "step": 3198
+    },
+    {
+      "epoch": 1.2643615707582119,
+      "grad_norm": 0.4528216117045259,
+      "learning_rate": 4.963527745906677e-06,
+      "loss": 0.5431,
+      "step": 3199
+    },
+    {
+      "epoch": 1.2647567300568041,
+      "grad_norm": 0.44237602152913297,
+      "learning_rate": 4.963501076860522e-06,
+      "loss": 0.5514,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2651518893553964,
+      "grad_norm": 0.4472333931200183,
+      "learning_rate": 4.9634743981392316e-06,
+      "loss": 0.5597,
+      "step": 3201
+    },
+    {
+      "epoch": 1.2655470486539886,
+      "grad_norm": 0.4951585973064483,
+      "learning_rate": 4.9634477097429105e-06,
+      "loss": 0.5993,
+      "step": 3202
+    },
+    {
+      "epoch": 1.2659422079525808,
+      "grad_norm": 0.4588298365010358,
+      "learning_rate": 4.9634210116716606e-06,
+      "loss": 0.5612,
+      "step": 3203
+    },
+    {
+      "epoch": 1.266337367251173,
+      "grad_norm": 0.43949401153597856,
+      "learning_rate": 4.96339430392559e-06,
+      "loss": 0.5758,
+      "step": 3204
+    },
+    {
+      "epoch": 1.2667325265497653,
+      "grad_norm": 0.4429353608767374,
+      "learning_rate": 4.963367586504803e-06,
+      "loss": 0.5581,
+      "step": 3205
+    },
+    {
+      "epoch": 1.2671276858483576,
+      "grad_norm": 0.44990618359944873,
+      "learning_rate": 4.963340859409404e-06,
+      "loss": 0.554,
+      "step": 3206
+    },
+    {
+      "epoch": 1.2675228451469498,
+      "grad_norm": 0.46874642934363553,
+      "learning_rate": 4.963314122639497e-06,
+      "loss": 0.5549,
+      "step": 3207
+    },
+    {
+      "epoch": 1.267918004445542,
+      "grad_norm": 0.4366561437308232,
+      "learning_rate": 4.963287376195188e-06,
+      "loss": 0.568,
+      "step": 3208
+    },
+    {
+      "epoch": 1.2683131637441343,
+      "grad_norm": 0.43922781969850494,
+      "learning_rate": 4.963260620076582e-06,
+      "loss": 0.5612,
+      "step": 3209
+    },
+    {
+      "epoch": 1.2687083230427265,
+      "grad_norm": 0.46732316449861777,
+      "learning_rate": 4.963233854283785e-06,
+      "loss": 0.5675,
+      "step": 3210
+    },
+    {
+      "epoch": 1.2691034823413188,
+      "grad_norm": 0.4499324082172506,
+      "learning_rate": 4.9632070788169e-06,
+      "loss": 0.572,
+      "step": 3211
+    },
+    {
+      "epoch": 1.269498641639911,
+      "grad_norm": 0.4448079688254288,
+      "learning_rate": 4.9631802936760345e-06,
+      "loss": 0.5777,
+      "step": 3212
+    },
+    {
+      "epoch": 1.2698938009385032,
+      "grad_norm": 0.4681934528119531,
+      "learning_rate": 4.963153498861292e-06,
+      "loss": 0.5642,
+      "step": 3213
+    },
+    {
+      "epoch": 1.2702889602370955,
+      "grad_norm": 0.47693235289926394,
+      "learning_rate": 4.963126694372777e-06,
+      "loss": 0.5895,
+      "step": 3214
+    },
+    {
+      "epoch": 1.2706841195356877,
+      "grad_norm": 0.448800942766577,
+      "learning_rate": 4.963099880210598e-06,
+      "loss": 0.5754,
+      "step": 3215
+    },
+    {
+      "epoch": 1.27107927883428,
+      "grad_norm": 0.44749924388535084,
+      "learning_rate": 4.9630730563748575e-06,
+      "loss": 0.5618,
+      "step": 3216
+    },
+    {
+      "epoch": 1.2714744381328722,
+      "grad_norm": 0.45578491730563103,
+      "learning_rate": 4.963046222865662e-06,
+      "loss": 0.5689,
+      "step": 3217
+    },
+    {
+      "epoch": 1.2718695974314644,
+      "grad_norm": 0.5195796262214436,
+      "learning_rate": 4.963019379683116e-06,
+      "loss": 0.5939,
+      "step": 3218
+    },
+    {
+      "epoch": 1.2722647567300567,
+      "grad_norm": 0.47649433003286457,
+      "learning_rate": 4.962992526827326e-06,
+      "loss": 0.5771,
+      "step": 3219
+    },
+    {
+      "epoch": 1.272659916028649,
+      "grad_norm": 0.4410004496915249,
+      "learning_rate": 4.962965664298396e-06,
+      "loss": 0.5511,
+      "step": 3220
+    },
+    {
+      "epoch": 1.2730550753272412,
+      "grad_norm": 0.4636086332760471,
+      "learning_rate": 4.9629387920964335e-06,
+      "loss": 0.5684,
+      "step": 3221
+    },
+    {
+      "epoch": 1.2734502346258336,
+      "grad_norm": 0.46246815925142704,
+      "learning_rate": 4.962911910221543e-06,
+      "loss": 0.559,
+      "step": 3222
+    },
+    {
+      "epoch": 1.2738453939244259,
+      "grad_norm": 0.46339823849118905,
+      "learning_rate": 4.96288501867383e-06,
+      "loss": 0.5571,
+      "step": 3223
+    },
+    {
+      "epoch": 1.274240553223018,
+      "grad_norm": 0.4677333897938672,
+      "learning_rate": 4.9628581174534e-06,
+      "loss": 0.5793,
+      "step": 3224
+    },
+    {
+      "epoch": 1.2746357125216103,
+      "grad_norm": 0.4710260149695448,
+      "learning_rate": 4.962831206560358e-06,
+      "loss": 0.5927,
+      "step": 3225
+    },
+    {
+      "epoch": 1.2750308718202026,
+      "grad_norm": 0.45906156067530524,
+      "learning_rate": 4.962804285994811e-06,
+      "loss": 0.5508,
+      "step": 3226
+    },
+    {
+      "epoch": 1.2754260311187948,
+      "grad_norm": 0.4389981450527275,
+      "learning_rate": 4.962777355756865e-06,
+      "loss": 0.5544,
+      "step": 3227
+    },
+    {
+      "epoch": 1.275821190417387,
+      "grad_norm": 0.4566293309775638,
+      "learning_rate": 4.962750415846624e-06,
+      "loss": 0.5698,
+      "step": 3228
+    },
+    {
+      "epoch": 1.2762163497159793,
+      "grad_norm": 0.4565239233936326,
+      "learning_rate": 4.9627234662641965e-06,
+      "loss": 0.5843,
+      "step": 3229
+    },
+    {
+      "epoch": 1.2766115090145715,
+      "grad_norm": 0.4402634375950171,
+      "learning_rate": 4.962696507009686e-06,
+      "loss": 0.5716,
+      "step": 3230
+    },
+    {
+      "epoch": 1.2770066683131638,
+      "grad_norm": 0.43986945562125607,
+      "learning_rate": 4.962669538083198e-06,
+      "loss": 0.5645,
+      "step": 3231
+    },
+    {
+      "epoch": 1.277401827611756,
+      "grad_norm": 0.44151108436527936,
+      "learning_rate": 4.9626425594848404e-06,
+      "loss": 0.5351,
+      "step": 3232
+    },
+    {
+      "epoch": 1.2777969869103483,
+      "grad_norm": 0.46054451936494795,
+      "learning_rate": 4.962615571214718e-06,
+      "loss": 0.575,
+      "step": 3233
+    },
+    {
+      "epoch": 1.2781921462089405,
+      "grad_norm": 0.4832601682517206,
+      "learning_rate": 4.9625885732729365e-06,
+      "loss": 0.5724,
+      "step": 3234
+    },
+    {
+      "epoch": 1.2785873055075327,
+      "grad_norm": 0.4496650178160313,
+      "learning_rate": 4.962561565659603e-06,
+      "loss": 0.571,
+      "step": 3235
+    },
+    {
+      "epoch": 1.278982464806125,
+      "grad_norm": 0.4459537839833863,
+      "learning_rate": 4.962534548374823e-06,
+      "loss": 0.5496,
+      "step": 3236
+    },
+    {
+      "epoch": 1.2793776241047172,
+      "grad_norm": 0.4820361417897107,
+      "learning_rate": 4.962507521418703e-06,
+      "loss": 0.5488,
+      "step": 3237
+    },
+    {
+      "epoch": 1.2797727834033095,
+      "grad_norm": 0.4471059519027043,
+      "learning_rate": 4.962480484791348e-06,
+      "loss": 0.5557,
+      "step": 3238
+    },
+    {
+      "epoch": 1.2801679427019017,
+      "grad_norm": 0.44540448662756593,
+      "learning_rate": 4.962453438492865e-06,
+      "loss": 0.5403,
+      "step": 3239
+    },
+    {
+      "epoch": 1.280563102000494,
+      "grad_norm": 0.4706472390621206,
+      "learning_rate": 4.962426382523361e-06,
+      "loss": 0.575,
+      "step": 3240
+    },
+    {
+      "epoch": 1.2809582612990862,
+      "grad_norm": 0.531102576848962,
+      "learning_rate": 4.962399316882941e-06,
+      "loss": 0.5533,
+      "step": 3241
+    },
+    {
+      "epoch": 1.2813534205976784,
+      "grad_norm": 0.461239203184141,
+      "learning_rate": 4.962372241571711e-06,
+      "loss": 0.5704,
+      "step": 3242
+    },
+    {
+      "epoch": 1.2817485798962707,
+      "grad_norm": 0.43407314953672027,
+      "learning_rate": 4.962345156589779e-06,
+      "loss": 0.5503,
+      "step": 3243
+    },
+    {
+      "epoch": 1.282143739194863,
+      "grad_norm": 0.4493863750894706,
+      "learning_rate": 4.9623180619372505e-06,
+      "loss": 0.5661,
+      "step": 3244
+    },
+    {
+      "epoch": 1.2825388984934551,
+      "grad_norm": 0.4832832516667832,
+      "learning_rate": 4.962290957614231e-06,
+      "loss": 0.572,
+      "step": 3245
+    },
+    {
+      "epoch": 1.2829340577920474,
+      "grad_norm": 0.4503676002104791,
+      "learning_rate": 4.962263843620828e-06,
+      "loss": 0.5575,
+      "step": 3246
+    },
+    {
+      "epoch": 1.2833292170906396,
+      "grad_norm": 0.44527990189860395,
+      "learning_rate": 4.9622367199571485e-06,
+      "loss": 0.5598,
+      "step": 3247
+    },
+    {
+      "epoch": 1.2837243763892319,
+      "grad_norm": 0.4360326535872034,
+      "learning_rate": 4.962209586623298e-06,
+      "loss": 0.5625,
+      "step": 3248
+    },
+    {
+      "epoch": 1.284119535687824,
+      "grad_norm": 0.46369867577006174,
+      "learning_rate": 4.962182443619383e-06,
+      "loss": 0.5547,
+      "step": 3249
+    },
+    {
+      "epoch": 1.2845146949864163,
+      "grad_norm": 0.47697119598465904,
+      "learning_rate": 4.962155290945511e-06,
+      "loss": 0.5644,
+      "step": 3250
+    },
+    {
+      "epoch": 1.2849098542850086,
+      "grad_norm": 0.4572682032044699,
+      "learning_rate": 4.962128128601787e-06,
+      "loss": 0.6004,
+      "step": 3251
+    },
+    {
+      "epoch": 1.2853050135836008,
+      "grad_norm": 0.46293963615887146,
+      "learning_rate": 4.96210095658832e-06,
+      "loss": 0.5735,
+      "step": 3252
+    },
+    {
+      "epoch": 1.285700172882193,
+      "grad_norm": 0.4353007660494593,
+      "learning_rate": 4.962073774905216e-06,
+      "loss": 0.5515,
+      "step": 3253
+    },
+    {
+      "epoch": 1.2860953321807853,
+      "grad_norm": 0.4336239776816601,
+      "learning_rate": 4.96204658355258e-06,
+      "loss": 0.5479,
+      "step": 3254
+    },
+    {
+      "epoch": 1.2864904914793776,
+      "grad_norm": 0.4385631160161934,
+      "learning_rate": 4.962019382530521e-06,
+      "loss": 0.5581,
+      "step": 3255
+    },
+    {
+      "epoch": 1.2868856507779698,
+      "grad_norm": 0.47634454262632553,
+      "learning_rate": 4.961992171839144e-06,
+      "loss": 0.5794,
+      "step": 3256
+    },
+    {
+      "epoch": 1.287280810076562,
+      "grad_norm": 0.4672479339967555,
+      "learning_rate": 4.961964951478557e-06,
+      "loss": 0.5606,
+      "step": 3257
+    },
+    {
+      "epoch": 1.2876759693751545,
+      "grad_norm": 0.46957567960558677,
+      "learning_rate": 4.961937721448867e-06,
+      "loss": 0.5643,
+      "step": 3258
+    },
+    {
+      "epoch": 1.2880711286737467,
+      "grad_norm": 0.4773737016736467,
+      "learning_rate": 4.961910481750181e-06,
+      "loss": 0.5854,
+      "step": 3259
+    },
+    {
+      "epoch": 1.288466287972339,
+      "grad_norm": 0.4587288767028595,
+      "learning_rate": 4.961883232382604e-06,
+      "loss": 0.5533,
+      "step": 3260
+    },
+    {
+      "epoch": 1.2888614472709312,
+      "grad_norm": 0.4752043572517741,
+      "learning_rate": 4.961855973346246e-06,
+      "loss": 0.5837,
+      "step": 3261
+    },
+    {
+      "epoch": 1.2892566065695235,
+      "grad_norm": 0.4587769372346428,
+      "learning_rate": 4.961828704641212e-06,
+      "loss": 0.5547,
+      "step": 3262
+    },
+    {
+      "epoch": 1.2896517658681157,
+      "grad_norm": 0.4660024170483272,
+      "learning_rate": 4.96180142626761e-06,
+      "loss": 0.5511,
+      "step": 3263
+    },
+    {
+      "epoch": 1.290046925166708,
+      "grad_norm": 0.483664785494738,
+      "learning_rate": 4.961774138225547e-06,
+      "loss": 0.5766,
+      "step": 3264
+    },
+    {
+      "epoch": 1.2904420844653002,
+      "grad_norm": 0.4518446012693695,
+      "learning_rate": 4.96174684051513e-06,
+      "loss": 0.5557,
+      "step": 3265
+    },
+    {
+      "epoch": 1.2908372437638924,
+      "grad_norm": 0.44730158472491605,
+      "learning_rate": 4.961719533136466e-06,
+      "loss": 0.5657,
+      "step": 3266
+    },
+    {
+      "epoch": 1.2912324030624847,
+      "grad_norm": 0.45919328333593834,
+      "learning_rate": 4.961692216089663e-06,
+      "loss": 0.5659,
+      "step": 3267
+    },
+    {
+      "epoch": 1.291627562361077,
+      "grad_norm": 0.4670554649733668,
+      "learning_rate": 4.961664889374827e-06,
+      "loss": 0.5705,
+      "step": 3268
+    },
+    {
+      "epoch": 1.2920227216596691,
+      "grad_norm": 0.4470155299015426,
+      "learning_rate": 4.961637552992067e-06,
+      "loss": 0.5422,
+      "step": 3269
+    },
+    {
+      "epoch": 1.2924178809582614,
+      "grad_norm": 0.44944938927627665,
+      "learning_rate": 4.961610206941488e-06,
+      "loss": 0.5603,
+      "step": 3270
+    },
+    {
+      "epoch": 1.2928130402568536,
+      "grad_norm": 0.45882995864498305,
+      "learning_rate": 4.961582851223201e-06,
+      "loss": 0.5623,
+      "step": 3271
+    },
+    {
+      "epoch": 1.2932081995554459,
+      "grad_norm": 0.44864834182329033,
+      "learning_rate": 4.96155548583731e-06,
+      "loss": 0.5536,
+      "step": 3272
+    },
+    {
+      "epoch": 1.293603358854038,
+      "grad_norm": 0.4365082513173151,
+      "learning_rate": 4.961528110783924e-06,
+      "loss": 0.5492,
+      "step": 3273
+    },
+    {
+      "epoch": 1.2939985181526303,
+      "grad_norm": 0.4637763227170593,
+      "learning_rate": 4.961500726063151e-06,
+      "loss": 0.5704,
+      "step": 3274
+    },
+    {
+      "epoch": 1.2943936774512226,
+      "grad_norm": 0.4389304914923759,
+      "learning_rate": 4.961473331675096e-06,
+      "loss": 0.561,
+      "step": 3275
+    },
+    {
+      "epoch": 1.2947888367498148,
+      "grad_norm": 0.44926731822340105,
+      "learning_rate": 4.9614459276198705e-06,
+      "loss": 0.582,
+      "step": 3276
+    },
+    {
+      "epoch": 1.295183996048407,
+      "grad_norm": 0.4486660860991371,
+      "learning_rate": 4.961418513897579e-06,
+      "loss": 0.5635,
+      "step": 3277
+    },
+    {
+      "epoch": 1.2955791553469993,
+      "grad_norm": 0.4527867437707233,
+      "learning_rate": 4.96139109050833e-06,
+      "loss": 0.5733,
+      "step": 3278
+    },
+    {
+      "epoch": 1.2959743146455915,
+      "grad_norm": 0.45347608716724236,
+      "learning_rate": 4.961363657452232e-06,
+      "loss": 0.5518,
+      "step": 3279
+    },
+    {
+      "epoch": 1.2963694739441838,
+      "grad_norm": 0.4508697870043093,
+      "learning_rate": 4.961336214729392e-06,
+      "loss": 0.5548,
+      "step": 3280
+    },
+    {
+      "epoch": 1.296764633242776,
+      "grad_norm": 0.4637754516249654,
+      "learning_rate": 4.961308762339918e-06,
+      "loss": 0.5638,
+      "step": 3281
+    },
+    {
+      "epoch": 1.2971597925413683,
+      "grad_norm": 0.4574438583788397,
+      "learning_rate": 4.961281300283918e-06,
+      "loss": 0.5516,
+      "step": 3282
+    },
+    {
+      "epoch": 1.2975549518399605,
+      "grad_norm": 0.4476663326378099,
+      "learning_rate": 4.961253828561499e-06,
+      "loss": 0.5494,
+      "step": 3283
+    },
+    {
+      "epoch": 1.2979501111385527,
+      "grad_norm": 0.45758420853662424,
+      "learning_rate": 4.96122634717277e-06,
+      "loss": 0.5669,
+      "step": 3284
+    },
+    {
+      "epoch": 1.298345270437145,
+      "grad_norm": 0.4730047069561211,
+      "learning_rate": 4.9611988561178385e-06,
+      "loss": 0.5748,
+      "step": 3285
+    },
+    {
+      "epoch": 1.2987404297357372,
+      "grad_norm": 0.44003229484144685,
+      "learning_rate": 4.9611713553968125e-06,
+      "loss": 0.5465,
+      "step": 3286
+    },
+    {
+      "epoch": 1.2991355890343295,
+      "grad_norm": 0.4411529968194715,
+      "learning_rate": 4.9611438450098e-06,
+      "loss": 0.5559,
+      "step": 3287
+    },
+    {
+      "epoch": 1.2995307483329217,
+      "grad_norm": 0.4778237900402621,
+      "learning_rate": 4.9611163249569085e-06,
+      "loss": 0.5639,
+      "step": 3288
+    },
+    {
+      "epoch": 1.299925907631514,
+      "grad_norm": 0.44974909893119686,
+      "learning_rate": 4.961088795238247e-06,
+      "loss": 0.5758,
+      "step": 3289
+    },
+    {
+      "epoch": 1.3003210669301062,
+      "grad_norm": 0.4378290859300243,
+      "learning_rate": 4.9610612558539214e-06,
+      "loss": 0.5689,
+      "step": 3290
+    },
+    {
+      "epoch": 1.3007162262286984,
+      "grad_norm": 0.4429720821425789,
+      "learning_rate": 4.961033706804044e-06,
+      "loss": 0.5606,
+      "step": 3291
+    },
+    {
+      "epoch": 1.3011113855272907,
+      "grad_norm": 0.4633926009741077,
+      "learning_rate": 4.961006148088719e-06,
+      "loss": 0.5631,
+      "step": 3292
+    },
+    {
+      "epoch": 1.301506544825883,
+      "grad_norm": 0.46706115854462443,
+      "learning_rate": 4.960978579708058e-06,
+      "loss": 0.5813,
+      "step": 3293
+    },
+    {
+      "epoch": 1.3019017041244751,
+      "grad_norm": 0.47368413285383676,
+      "learning_rate": 4.9609510016621655e-06,
+      "loss": 0.5883,
+      "step": 3294
+    },
+    {
+      "epoch": 1.3022968634230674,
+      "grad_norm": 0.4737507951875103,
+      "learning_rate": 4.960923413951153e-06,
+      "loss": 0.5748,
+      "step": 3295
+    },
+    {
+      "epoch": 1.3026920227216596,
+      "grad_norm": 0.4570347719095943,
+      "learning_rate": 4.960895816575127e-06,
+      "loss": 0.5692,
+      "step": 3296
+    },
+    {
+      "epoch": 1.3030871820202519,
+      "grad_norm": 0.4577457839859938,
+      "learning_rate": 4.960868209534198e-06,
+      "loss": 0.5562,
+      "step": 3297
+    },
+    {
+      "epoch": 1.303482341318844,
+      "grad_norm": 0.47003885252402855,
+      "learning_rate": 4.960840592828472e-06,
+      "loss": 0.5734,
+      "step": 3298
+    },
+    {
+      "epoch": 1.3038775006174363,
+      "grad_norm": 0.4624743876597724,
+      "learning_rate": 4.960812966458058e-06,
+      "loss": 0.5603,
+      "step": 3299
+    },
+    {
+      "epoch": 1.3042726599160286,
+      "grad_norm": 0.48901567850108524,
+      "learning_rate": 4.960785330423066e-06,
+      "loss": 0.5893,
+      "step": 3300
+    },
+    {
+      "epoch": 1.3046678192146208,
+      "grad_norm": 0.4522014166198177,
+      "learning_rate": 4.960757684723603e-06,
+      "loss": 0.5538,
+      "step": 3301
+    },
+    {
+      "epoch": 1.305062978513213,
+      "grad_norm": 0.45770794284942456,
+      "learning_rate": 4.9607300293597774e-06,
+      "loss": 0.5656,
+      "step": 3302
+    },
+    {
+      "epoch": 1.3054581378118053,
+      "grad_norm": 0.4322173217917462,
+      "learning_rate": 4.960702364331699e-06,
+      "loss": 0.5497,
+      "step": 3303
+    },
+    {
+      "epoch": 1.3058532971103975,
+      "grad_norm": 0.4381343619517814,
+      "learning_rate": 4.960674689639477e-06,
+      "loss": 0.5591,
+      "step": 3304
+    },
+    {
+      "epoch": 1.3062484564089898,
+      "grad_norm": 0.44151218550571697,
+      "learning_rate": 4.960647005283217e-06,
+      "loss": 0.5694,
+      "step": 3305
+    },
+    {
+      "epoch": 1.306643615707582,
+      "grad_norm": 0.43450554135557107,
+      "learning_rate": 4.960619311263031e-06,
+      "loss": 0.5803,
+      "step": 3306
+    },
+    {
+      "epoch": 1.3070387750061743,
+      "grad_norm": 0.42935555212955734,
+      "learning_rate": 4.960591607579026e-06,
+      "loss": 0.5367,
+      "step": 3307
+    },
+    {
+      "epoch": 1.3074339343047665,
+      "grad_norm": 0.44216935598010304,
+      "learning_rate": 4.960563894231312e-06,
+      "loss": 0.5526,
+      "step": 3308
+    },
+    {
+      "epoch": 1.3078290936033587,
+      "grad_norm": 0.4436461836567033,
+      "learning_rate": 4.960536171219997e-06,
+      "loss": 0.5658,
+      "step": 3309
+    },
+    {
+      "epoch": 1.308224252901951,
+      "grad_norm": 0.46145987300596963,
+      "learning_rate": 4.96050843854519e-06,
+      "loss": 0.5712,
+      "step": 3310
+    },
+    {
+      "epoch": 1.3086194122005432,
+      "grad_norm": 0.4540758807723103,
+      "learning_rate": 4.960480696206999e-06,
+      "loss": 0.5551,
+      "step": 3311
+    },
+    {
+      "epoch": 1.3090145714991355,
+      "grad_norm": 0.4371619617022349,
+      "learning_rate": 4.960452944205535e-06,
+      "loss": 0.5526,
+      "step": 3312
+    },
+    {
+      "epoch": 1.3094097307977277,
+      "grad_norm": 0.44906152086035417,
+      "learning_rate": 4.960425182540905e-06,
+      "loss": 0.556,
+      "step": 3313
+    },
+    {
+      "epoch": 1.30980489009632,
+      "grad_norm": 0.6283881072933288,
+      "learning_rate": 4.96039741121322e-06,
+      "loss": 0.6207,
+      "step": 3314
+    },
+    {
+      "epoch": 1.3102000493949122,
+      "grad_norm": 0.43354415802766594,
+      "learning_rate": 4.960369630222588e-06,
+      "loss": 0.5782,
+      "step": 3315
+    },
+    {
+      "epoch": 1.3105952086935044,
+      "grad_norm": 0.4573902248334659,
+      "learning_rate": 4.960341839569117e-06,
+      "loss": 0.5657,
+      "step": 3316
+    },
+    {
+      "epoch": 1.310990367992097,
+      "grad_norm": 0.4618875741095821,
+      "learning_rate": 4.9603140392529185e-06,
+      "loss": 0.5504,
+      "step": 3317
+    },
+    {
+      "epoch": 1.3113855272906891,
+      "grad_norm": 0.4335794680087329,
+      "learning_rate": 4.9602862292740995e-06,
+      "loss": 0.547,
+      "step": 3318
+    },
+    {
+      "epoch": 1.3117806865892814,
+      "grad_norm": 0.46798359661776007,
+      "learning_rate": 4.960258409632771e-06,
+      "loss": 0.5565,
+      "step": 3319
+    },
+    {
+      "epoch": 1.3121758458878736,
+      "grad_norm": 0.44285226370838654,
+      "learning_rate": 4.960230580329041e-06,
+      "loss": 0.5641,
+      "step": 3320
+    },
+    {
+      "epoch": 1.3125710051864659,
+      "grad_norm": 0.4656857860504189,
+      "learning_rate": 4.960202741363018e-06,
+      "loss": 0.5633,
+      "step": 3321
+    },
+    {
+      "epoch": 1.312966164485058,
+      "grad_norm": 0.4524731635101246,
+      "learning_rate": 4.960174892734813e-06,
+      "loss": 0.5607,
+      "step": 3322
+    },
+    {
+      "epoch": 1.3133613237836503,
+      "grad_norm": 0.456230055946175,
+      "learning_rate": 4.960147034444537e-06,
+      "loss": 0.5584,
+      "step": 3323
+    },
+    {
+      "epoch": 1.3137564830822426,
+      "grad_norm": 0.4772423713788212,
+      "learning_rate": 4.960119166492295e-06,
+      "loss": 0.57,
+      "step": 3324
+    },
+    {
+      "epoch": 1.3141516423808348,
+      "grad_norm": 0.4586733788184475,
+      "learning_rate": 4.9600912888782e-06,
+      "loss": 0.5758,
+      "step": 3325
+    },
+    {
+      "epoch": 1.314546801679427,
+      "grad_norm": 0.5480041169557188,
+      "learning_rate": 4.9600634016023606e-06,
+      "loss": 0.5772,
+      "step": 3326
+    },
+    {
+      "epoch": 1.3149419609780193,
+      "grad_norm": 0.4415783584140013,
+      "learning_rate": 4.960035504664885e-06,
+      "loss": 0.5404,
+      "step": 3327
+    },
+    {
+      "epoch": 1.3153371202766115,
+      "grad_norm": 0.4291556202863314,
+      "learning_rate": 4.960007598065884e-06,
+      "loss": 0.5612,
+      "step": 3328
+    },
+    {
+      "epoch": 1.3157322795752038,
+      "grad_norm": 0.44774827375999054,
+      "learning_rate": 4.959979681805467e-06,
+      "loss": 0.5776,
+      "step": 3329
+    },
+    {
+      "epoch": 1.316127438873796,
+      "grad_norm": 0.44973594796897337,
+      "learning_rate": 4.959951755883744e-06,
+      "loss": 0.5768,
+      "step": 3330
+    },
+    {
+      "epoch": 1.3165225981723883,
+      "grad_norm": 0.44813772825578874,
+      "learning_rate": 4.959923820300824e-06,
+      "loss": 0.5623,
+      "step": 3331
+    },
+    {
+      "epoch": 1.3169177574709805,
+      "grad_norm": 0.44654411240174396,
+      "learning_rate": 4.959895875056816e-06,
+      "loss": 0.5655,
+      "step": 3332
+    },
+    {
+      "epoch": 1.3173129167695727,
+      "grad_norm": 0.4373133409477842,
+      "learning_rate": 4.959867920151832e-06,
+      "loss": 0.5592,
+      "step": 3333
+    },
+    {
+      "epoch": 1.317708076068165,
+      "grad_norm": 0.4363324235460165,
+      "learning_rate": 4.95983995558598e-06,
+      "loss": 0.5492,
+      "step": 3334
+    },
+    {
+      "epoch": 1.3181032353667572,
+      "grad_norm": 0.4346900282961237,
+      "learning_rate": 4.95981198135937e-06,
+      "loss": 0.5444,
+      "step": 3335
+    },
+    {
+      "epoch": 1.3184983946653495,
+      "grad_norm": 0.4523110093393848,
+      "learning_rate": 4.959783997472113e-06,
+      "loss": 0.5536,
+      "step": 3336
+    },
+    {
+      "epoch": 1.3188935539639417,
+      "grad_norm": 0.46544382830839454,
+      "learning_rate": 4.959756003924317e-06,
+      "loss": 0.5702,
+      "step": 3337
+    },
+    {
+      "epoch": 1.319288713262534,
+      "grad_norm": 0.44772145946647923,
+      "learning_rate": 4.959728000716094e-06,
+      "loss": 0.5616,
+      "step": 3338
+    },
+    {
+      "epoch": 1.3196838725611262,
+      "grad_norm": 0.4384987036176576,
+      "learning_rate": 4.959699987847554e-06,
+      "loss": 0.5685,
+      "step": 3339
+    },
+    {
+      "epoch": 1.3200790318597184,
+      "grad_norm": 0.43847040276558563,
+      "learning_rate": 4.9596719653188045e-06,
+      "loss": 0.5518,
+      "step": 3340
+    },
+    {
+      "epoch": 1.3204741911583107,
+      "grad_norm": 0.45117712104049984,
+      "learning_rate": 4.959643933129958e-06,
+      "loss": 0.5734,
+      "step": 3341
+    },
+    {
+      "epoch": 1.320869350456903,
+      "grad_norm": 0.45533362827402557,
+      "learning_rate": 4.9596158912811235e-06,
+      "loss": 0.5649,
+      "step": 3342
+    },
+    {
+      "epoch": 1.3212645097554951,
+      "grad_norm": 0.44551751419590213,
+      "learning_rate": 4.9595878397724106e-06,
+      "loss": 0.5778,
+      "step": 3343
+    },
+    {
+      "epoch": 1.3216596690540874,
+      "grad_norm": 0.4370316849726543,
+      "learning_rate": 4.959559778603931e-06,
+      "loss": 0.5602,
+      "step": 3344
+    },
+    {
+      "epoch": 1.3220548283526796,
+      "grad_norm": 0.4599955825604526,
+      "learning_rate": 4.959531707775793e-06,
+      "loss": 0.571,
+      "step": 3345
+    },
+    {
+      "epoch": 1.3224499876512719,
+      "grad_norm": 0.4424416975622541,
+      "learning_rate": 4.959503627288109e-06,
+      "loss": 0.561,
+      "step": 3346
+    },
+    {
+      "epoch": 1.322845146949864,
+      "grad_norm": 0.43661943086497584,
+      "learning_rate": 4.959475537140989e-06,
+      "loss": 0.5523,
+      "step": 3347
+    },
+    {
+      "epoch": 1.3232403062484563,
+      "grad_norm": 0.4373998487954772,
+      "learning_rate": 4.959447437334541e-06,
+      "loss": 0.5613,
+      "step": 3348
+    },
+    {
+      "epoch": 1.3236354655470486,
+      "grad_norm": 0.4494818246912831,
+      "learning_rate": 4.959419327868878e-06,
+      "loss": 0.5775,
+      "step": 3349
+    },
+    {
+      "epoch": 1.3240306248456408,
+      "grad_norm": 0.43842069386255866,
+      "learning_rate": 4.959391208744108e-06,
+      "loss": 0.5677,
+      "step": 3350
+    },
+    {
+      "epoch": 1.324425784144233,
+      "grad_norm": 0.4494362228050278,
+      "learning_rate": 4.959363079960344e-06,
+      "loss": 0.5517,
+      "step": 3351
+    },
+    {
+      "epoch": 1.3248209434428255,
+      "grad_norm": 0.45526558767555503,
+      "learning_rate": 4.959334941517695e-06,
+      "loss": 0.572,
+      "step": 3352
+    },
+    {
+      "epoch": 1.3252161027414178,
+      "grad_norm": 0.44027079643934225,
+      "learning_rate": 4.959306793416271e-06,
+      "loss": 0.5698,
+      "step": 3353
+    },
+    {
+      "epoch": 1.32561126204001,
+      "grad_norm": 0.4405471076038371,
+      "learning_rate": 4.959278635656184e-06,
+      "loss": 0.5653,
+      "step": 3354
+    },
+    {
+      "epoch": 1.3260064213386022,
+      "grad_norm": 0.45167721473900385,
+      "learning_rate": 4.959250468237544e-06,
+      "loss": 0.5757,
+      "step": 3355
+    },
+    {
+      "epoch": 1.3264015806371945,
+      "grad_norm": 0.4421108044099741,
+      "learning_rate": 4.959222291160461e-06,
+      "loss": 0.5642,
+      "step": 3356
+    },
+    {
+      "epoch": 1.3267967399357867,
+      "grad_norm": 0.4431212984296979,
+      "learning_rate": 4.959194104425047e-06,
+      "loss": 0.5505,
+      "step": 3357
+    },
+    {
+      "epoch": 1.327191899234379,
+      "grad_norm": 0.4521254913136757,
+      "learning_rate": 4.95916590803141e-06,
+      "loss": 0.5674,
+      "step": 3358
+    },
+    {
+      "epoch": 1.3275870585329712,
+      "grad_norm": 0.4738728564671993,
+      "learning_rate": 4.959137701979665e-06,
+      "loss": 0.5737,
+      "step": 3359
+    },
+    {
+      "epoch": 1.3279822178315634,
+      "grad_norm": 0.4503578411569048,
+      "learning_rate": 4.9591094862699185e-06,
+      "loss": 0.5767,
+      "step": 3360
+    },
+    {
+      "epoch": 1.3283773771301557,
+      "grad_norm": 0.4593996821030863,
+      "learning_rate": 4.959081260902284e-06,
+      "loss": 0.5819,
+      "step": 3361
+    },
+    {
+      "epoch": 1.328772536428748,
+      "grad_norm": 0.47458910560702716,
+      "learning_rate": 4.959053025876871e-06,
+      "loss": 0.5751,
+      "step": 3362
+    },
+    {
+      "epoch": 1.3291676957273402,
+      "grad_norm": 0.4601597399895584,
+      "learning_rate": 4.959024781193792e-06,
+      "loss": 0.5502,
+      "step": 3363
+    },
+    {
+      "epoch": 1.3295628550259324,
+      "grad_norm": 0.43863004190964605,
+      "learning_rate": 4.958996526853156e-06,
+      "loss": 0.552,
+      "step": 3364
+    },
+    {
+      "epoch": 1.3299580143245247,
+      "grad_norm": 0.449223468368665,
+      "learning_rate": 4.958968262855075e-06,
+      "loss": 0.5726,
+      "step": 3365
+    },
+    {
+      "epoch": 1.330353173623117,
+      "grad_norm": 0.45615386443341344,
+      "learning_rate": 4.958939989199659e-06,
+      "loss": 0.5764,
+      "step": 3366
+    },
+    {
+      "epoch": 1.3307483329217091,
+      "grad_norm": 0.44070542771087373,
+      "learning_rate": 4.958911705887022e-06,
+      "loss": 0.5687,
+      "step": 3367
+    },
+    {
+      "epoch": 1.3311434922203014,
+      "grad_norm": 0.44414182367934024,
+      "learning_rate": 4.958883412917271e-06,
+      "loss": 0.5658,
+      "step": 3368
+    },
+    {
+      "epoch": 1.3315386515188936,
+      "grad_norm": 0.44838583407126426,
+      "learning_rate": 4.9588551102905205e-06,
+      "loss": 0.5716,
+      "step": 3369
+    },
+    {
+      "epoch": 1.3319338108174859,
+      "grad_norm": 0.889969275288809,
+      "learning_rate": 4.9588267980068795e-06,
+      "loss": 0.556,
+      "step": 3370
+    },
+    {
+      "epoch": 1.332328970116078,
+      "grad_norm": 0.46358754540200775,
+      "learning_rate": 4.95879847606646e-06,
+      "loss": 0.5861,
+      "step": 3371
+    },
+    {
+      "epoch": 1.3327241294146703,
+      "grad_norm": 0.4282860491355915,
+      "learning_rate": 4.958770144469372e-06,
+      "loss": 0.5428,
+      "step": 3372
+    },
+    {
+      "epoch": 1.3331192887132626,
+      "grad_norm": 0.45729755865769983,
+      "learning_rate": 4.95874180321573e-06,
+      "loss": 0.5856,
+      "step": 3373
+    },
+    {
+      "epoch": 1.3335144480118548,
+      "grad_norm": 0.45588398035202177,
+      "learning_rate": 4.958713452305642e-06,
+      "loss": 0.5713,
+      "step": 3374
+    },
+    {
+      "epoch": 1.333909607310447,
+      "grad_norm": 0.44858104138575744,
+      "learning_rate": 4.958685091739221e-06,
+      "loss": 0.5657,
+      "step": 3375
+    },
+    {
+      "epoch": 1.3343047666090393,
+      "grad_norm": 0.44036953218064384,
+      "learning_rate": 4.958656721516577e-06,
+      "loss": 0.5587,
+      "step": 3376
+    },
+    {
+      "epoch": 1.3346999259076315,
+      "grad_norm": 0.4540748920166802,
+      "learning_rate": 4.958628341637823e-06,
+      "loss": 0.5646,
+      "step": 3377
+    },
+    {
+      "epoch": 1.3350950852062238,
+      "grad_norm": 0.4396313054654993,
+      "learning_rate": 4.9585999521030704e-06,
+      "loss": 0.5464,
+      "step": 3378
+    },
+    {
+      "epoch": 1.335490244504816,
+      "grad_norm": 0.4454412703380863,
+      "learning_rate": 4.958571552912429e-06,
+      "loss": 0.5545,
+      "step": 3379
+    },
+    {
+      "epoch": 1.3358854038034083,
+      "grad_norm": 0.45082119473828774,
+      "learning_rate": 4.9585431440660125e-06,
+      "loss": 0.561,
+      "step": 3380
+    },
+    {
+      "epoch": 1.3362805631020005,
+      "grad_norm": 0.4374755472502395,
+      "learning_rate": 4.958514725563931e-06,
+      "loss": 0.5571,
+      "step": 3381
+    },
+    {
+      "epoch": 1.3366757224005927,
+      "grad_norm": 0.46478260478991495,
+      "learning_rate": 4.958486297406296e-06,
+      "loss": 0.5891,
+      "step": 3382
+    },
+    {
+      "epoch": 1.337070881699185,
+      "grad_norm": 0.43814804662705964,
+      "learning_rate": 4.958457859593221e-06,
+      "loss": 0.565,
+      "step": 3383
+    },
+    {
+      "epoch": 1.3374660409977772,
+      "grad_norm": 0.44507180809627767,
+      "learning_rate": 4.958429412124816e-06,
+      "loss": 0.5599,
+      "step": 3384
+    },
+    {
+      "epoch": 1.3378612002963695,
+      "grad_norm": 0.441421838525371,
+      "learning_rate": 4.958400955001193e-06,
+      "loss": 0.5428,
+      "step": 3385
+    },
+    {
+      "epoch": 1.3382563595949617,
+      "grad_norm": 0.44374387122522124,
+      "learning_rate": 4.958372488222463e-06,
+      "loss": 0.5403,
+      "step": 3386
+    },
+    {
+      "epoch": 1.338651518893554,
+      "grad_norm": 0.4467964188449642,
+      "learning_rate": 4.958344011788739e-06,
+      "loss": 0.5503,
+      "step": 3387
+    },
+    {
+      "epoch": 1.3390466781921462,
+      "grad_norm": 0.4479117153175756,
+      "learning_rate": 4.958315525700134e-06,
+      "loss": 0.5775,
+      "step": 3388
+    },
+    {
+      "epoch": 1.3394418374907384,
+      "grad_norm": 0.5332367747737651,
+      "learning_rate": 4.958287029956757e-06,
+      "loss": 0.5715,
+      "step": 3389
+    },
+    {
+      "epoch": 1.3398369967893307,
+      "grad_norm": 0.45663522892567066,
+      "learning_rate": 4.958258524558721e-06,
+      "loss": 0.5881,
+      "step": 3390
+    },
+    {
+      "epoch": 1.340232156087923,
+      "grad_norm": 0.4521841960229859,
+      "learning_rate": 4.958230009506139e-06,
+      "loss": 0.5603,
+      "step": 3391
+    },
+    {
+      "epoch": 1.3406273153865151,
+      "grad_norm": 0.45825530510638524,
+      "learning_rate": 4.958201484799122e-06,
+      "loss": 0.5664,
+      "step": 3392
+    },
+    {
+      "epoch": 1.3410224746851074,
+      "grad_norm": 0.45965721161303025,
+      "learning_rate": 4.958172950437782e-06,
+      "loss": 0.5715,
+      "step": 3393
+    },
+    {
+      "epoch": 1.3414176339836996,
+      "grad_norm": 0.4412893524397775,
+      "learning_rate": 4.958144406422232e-06,
+      "loss": 0.5709,
+      "step": 3394
+    },
+    {
+      "epoch": 1.3418127932822919,
+      "grad_norm": 0.45819164319217937,
+      "learning_rate": 4.958115852752583e-06,
+      "loss": 0.5666,
+      "step": 3395
+    },
+    {
+      "epoch": 1.342207952580884,
+      "grad_norm": 0.45080426168635684,
+      "learning_rate": 4.958087289428948e-06,
+      "loss": 0.5452,
+      "step": 3396
+    },
+    {
+      "epoch": 1.3426031118794763,
+      "grad_norm": 0.46299421201178154,
+      "learning_rate": 4.958058716451438e-06,
+      "loss": 0.5788,
+      "step": 3397
+    },
+    {
+      "epoch": 1.3429982711780686,
+      "grad_norm": 0.44631538749161104,
+      "learning_rate": 4.9580301338201665e-06,
+      "loss": 0.5605,
+      "step": 3398
+    },
+    {
+      "epoch": 1.3433934304766608,
+      "grad_norm": 0.4753236947479198,
+      "learning_rate": 4.9580015415352446e-06,
+      "loss": 0.5833,
+      "step": 3399
+    },
+    {
+      "epoch": 1.343788589775253,
+      "grad_norm": 0.4543478319902544,
+      "learning_rate": 4.9579729395967855e-06,
+      "loss": 0.5459,
+      "step": 3400
+    },
+    {
+      "epoch": 1.3441837490738453,
+      "grad_norm": 0.42397174780842645,
+      "learning_rate": 4.957944328004902e-06,
+      "loss": 0.5408,
+      "step": 3401
+    },
+    {
+      "epoch": 1.3445789083724375,
+      "grad_norm": 0.44366683749260666,
+      "learning_rate": 4.957915706759705e-06,
+      "loss": 0.5672,
+      "step": 3402
+    },
+    {
+      "epoch": 1.3449740676710298,
+      "grad_norm": 0.4620326978290853,
+      "learning_rate": 4.957887075861308e-06,
+      "loss": 0.5912,
+      "step": 3403
+    },
+    {
+      "epoch": 1.345369226969622,
+      "grad_norm": 0.437254776572285,
+      "learning_rate": 4.957858435309822e-06,
+      "loss": 0.5584,
+      "step": 3404
+    },
+    {
+      "epoch": 1.3457643862682143,
+      "grad_norm": 0.44334059234747064,
+      "learning_rate": 4.9578297851053626e-06,
+      "loss": 0.5552,
+      "step": 3405
+    },
+    {
+      "epoch": 1.3461595455668065,
+      "grad_norm": 0.45277563926558906,
+      "learning_rate": 4.957801125248038e-06,
+      "loss": 0.5796,
+      "step": 3406
+    },
+    {
+      "epoch": 1.3465547048653987,
+      "grad_norm": 0.4529953395968256,
+      "learning_rate": 4.957772455737965e-06,
+      "loss": 0.5645,
+      "step": 3407
+    },
+    {
+      "epoch": 1.346949864163991,
+      "grad_norm": 0.45953659145635833,
+      "learning_rate": 4.9577437765752535e-06,
+      "loss": 0.5551,
+      "step": 3408
+    },
+    {
+      "epoch": 1.3473450234625832,
+      "grad_norm": 0.45243007027630033,
+      "learning_rate": 4.957715087760017e-06,
+      "loss": 0.562,
+      "step": 3409
+    },
+    {
+      "epoch": 1.3477401827611755,
+      "grad_norm": 0.44813612053487434,
+      "learning_rate": 4.9576863892923675e-06,
+      "loss": 0.5669,
+      "step": 3410
+    },
+    {
+      "epoch": 1.348135342059768,
+      "grad_norm": 0.4414843229708112,
+      "learning_rate": 4.957657681172419e-06,
+      "loss": 0.5564,
+      "step": 3411
+    },
+    {
+      "epoch": 1.3485305013583602,
+      "grad_norm": 0.4762734336021678,
+      "learning_rate": 4.957628963400283e-06,
+      "loss": 0.5576,
+      "step": 3412
+    },
+    {
+      "epoch": 1.3489256606569524,
+      "grad_norm": 0.4651845435138158,
+      "learning_rate": 4.957600235976072e-06,
+      "loss": 0.5471,
+      "step": 3413
+    },
+    {
+      "epoch": 1.3493208199555446,
+      "grad_norm": 0.4394461136350005,
+      "learning_rate": 4.957571498899901e-06,
+      "loss": 0.5542,
+      "step": 3414
+    },
+    {
+      "epoch": 1.3497159792541369,
+      "grad_norm": 0.44448501672351937,
+      "learning_rate": 4.95754275217188e-06,
+      "loss": 0.5635,
+      "step": 3415
+    },
+    {
+      "epoch": 1.3501111385527291,
+      "grad_norm": 0.4620591465811548,
+      "learning_rate": 4.9575139957921245e-06,
+      "loss": 0.5587,
+      "step": 3416
+    },
+    {
+      "epoch": 1.3505062978513214,
+      "grad_norm": 0.45815520871269744,
+      "learning_rate": 4.957485229760747e-06,
+      "loss": 0.5652,
+      "step": 3417
+    },
+    {
+      "epoch": 1.3509014571499136,
+      "grad_norm": 0.4493406020218002,
+      "learning_rate": 4.957456454077858e-06,
+      "loss": 0.5441,
+      "step": 3418
+    },
+    {
+      "epoch": 1.3512966164485058,
+      "grad_norm": 0.454603776283902,
+      "learning_rate": 4.957427668743573e-06,
+      "loss": 0.5598,
+      "step": 3419
+    },
+    {
+      "epoch": 1.351691775747098,
+      "grad_norm": 0.47114620064149954,
+      "learning_rate": 4.9573988737580045e-06,
+      "loss": 0.5714,
+      "step": 3420
+    },
+    {
+      "epoch": 1.3520869350456903,
+      "grad_norm": 0.4559096213569245,
+      "learning_rate": 4.957370069121265e-06,
+      "loss": 0.5616,
+      "step": 3421
+    },
+    {
+      "epoch": 1.3524820943442826,
+      "grad_norm": 0.42729193122760173,
+      "learning_rate": 4.95734125483347e-06,
+      "loss": 0.5561,
+      "step": 3422
+    },
+    {
+      "epoch": 1.3528772536428748,
+      "grad_norm": 0.46360094062631574,
+      "learning_rate": 4.957312430894729e-06,
+      "loss": 0.5475,
+      "step": 3423
+    },
+    {
+      "epoch": 1.353272412941467,
+      "grad_norm": 0.44865908216487266,
+      "learning_rate": 4.957283597305157e-06,
+      "loss": 0.5744,
+      "step": 3424
+    },
+    {
+      "epoch": 1.3536675722400593,
+      "grad_norm": 0.4527404050788081,
+      "learning_rate": 4.957254754064867e-06,
+      "loss": 0.5691,
+      "step": 3425
+    },
+    {
+      "epoch": 1.3540627315386515,
+      "grad_norm": 0.4380566310678888,
+      "learning_rate": 4.957225901173973e-06,
+      "loss": 0.5675,
+      "step": 3426
+    },
+    {
+      "epoch": 1.3544578908372438,
+      "grad_norm": 0.4480119306626472,
+      "learning_rate": 4.957197038632587e-06,
+      "loss": 0.5794,
+      "step": 3427
+    },
+    {
+      "epoch": 1.354853050135836,
+      "grad_norm": 0.44898242236087227,
+      "learning_rate": 4.957168166440824e-06,
+      "loss": 0.5749,
+      "step": 3428
+    },
+    {
+      "epoch": 1.3552482094344283,
+      "grad_norm": 0.4395620002210985,
+      "learning_rate": 4.9571392845987965e-06,
+      "loss": 0.5539,
+      "step": 3429
+    },
+    {
+      "epoch": 1.3556433687330205,
+      "grad_norm": 0.4828976972353129,
+      "learning_rate": 4.957110393106618e-06,
+      "loss": 0.5486,
+      "step": 3430
+    },
+    {
+      "epoch": 1.3560385280316127,
+      "grad_norm": 0.4340264093323101,
+      "learning_rate": 4.9570814919644015e-06,
+      "loss": 0.5528,
+      "step": 3431
+    },
+    {
+      "epoch": 1.356433687330205,
+      "grad_norm": 0.43633467432037276,
+      "learning_rate": 4.9570525811722604e-06,
+      "loss": 0.5823,
+      "step": 3432
+    },
+    {
+      "epoch": 1.3568288466287972,
+      "grad_norm": 0.44896667799179996,
+      "learning_rate": 4.957023660730309e-06,
+      "loss": 0.5674,
+      "step": 3433
+    },
+    {
+      "epoch": 1.3572240059273895,
+      "grad_norm": 0.4424965400029668,
+      "learning_rate": 4.9569947306386614e-06,
+      "loss": 0.5597,
+      "step": 3434
+    },
+    {
+      "epoch": 1.3576191652259817,
+      "grad_norm": 0.4505551734794091,
+      "learning_rate": 4.95696579089743e-06,
+      "loss": 0.5732,
+      "step": 3435
+    },
+    {
+      "epoch": 1.358014324524574,
+      "grad_norm": 0.42939268083383797,
+      "learning_rate": 4.956936841506729e-06,
+      "loss": 0.5455,
+      "step": 3436
+    },
+    {
+      "epoch": 1.3584094838231662,
+      "grad_norm": 0.4283591881731488,
+      "learning_rate": 4.956907882466672e-06,
+      "loss": 0.5599,
+      "step": 3437
+    },
+    {
+      "epoch": 1.3588046431217584,
+      "grad_norm": 0.45404223789984394,
+      "learning_rate": 4.956878913777373e-06,
+      "loss": 0.5664,
+      "step": 3438
+    },
+    {
+      "epoch": 1.3591998024203507,
+      "grad_norm": 0.4403257215242389,
+      "learning_rate": 4.956849935438945e-06,
+      "loss": 0.5615,
+      "step": 3439
+    },
+    {
+      "epoch": 1.359594961718943,
+      "grad_norm": 0.4557250216642013,
+      "learning_rate": 4.956820947451503e-06,
+      "loss": 0.569,
+      "step": 3440
+    },
+    {
+      "epoch": 1.3599901210175351,
+      "grad_norm": 0.45833724590035085,
+      "learning_rate": 4.956791949815159e-06,
+      "loss": 0.5823,
+      "step": 3441
+    },
+    {
+      "epoch": 1.3603852803161274,
+      "grad_norm": 0.4530682419897322,
+      "learning_rate": 4.956762942530029e-06,
+      "loss": 0.5707,
+      "step": 3442
+    },
+    {
+      "epoch": 1.3607804396147196,
+      "grad_norm": 0.4376158015116781,
+      "learning_rate": 4.956733925596227e-06,
+      "loss": 0.5708,
+      "step": 3443
+    },
+    {
+      "epoch": 1.3611755989133119,
+      "grad_norm": 0.4307700914051484,
+      "learning_rate": 4.956704899013864e-06,
+      "loss": 0.5522,
+      "step": 3444
+    },
+    {
+      "epoch": 1.361570758211904,
+      "grad_norm": 0.46554414982116393,
+      "learning_rate": 4.956675862783057e-06,
+      "loss": 0.5655,
+      "step": 3445
+    },
+    {
+      "epoch": 1.3619659175104963,
+      "grad_norm": 0.4671181738319043,
+      "learning_rate": 4.9566468169039185e-06,
+      "loss": 0.566,
+      "step": 3446
+    },
+    {
+      "epoch": 1.3623610768090888,
+      "grad_norm": 0.4297413408882112,
+      "learning_rate": 4.956617761376563e-06,
+      "loss": 0.5542,
+      "step": 3447
+    },
+    {
+      "epoch": 1.362756236107681,
+      "grad_norm": 0.44684012819132296,
+      "learning_rate": 4.956588696201105e-06,
+      "loss": 0.5513,
+      "step": 3448
+    },
+    {
+      "epoch": 1.3631513954062733,
+      "grad_norm": 0.4693033252542671,
+      "learning_rate": 4.956559621377658e-06,
+      "loss": 0.5557,
+      "step": 3449
+    },
+    {
+      "epoch": 1.3635465547048655,
+      "grad_norm": 0.4486147777129349,
+      "learning_rate": 4.956530536906337e-06,
+      "loss": 0.5835,
+      "step": 3450
+    },
+    {
+      "epoch": 1.3639417140034578,
+      "grad_norm": 0.4520688255329486,
+      "learning_rate": 4.956501442787256e-06,
+      "loss": 0.5975,
+      "step": 3451
+    },
+    {
+      "epoch": 1.36433687330205,
+      "grad_norm": 0.4478220251661161,
+      "learning_rate": 4.956472339020528e-06,
+      "loss": 0.5653,
+      "step": 3452
+    },
+    {
+      "epoch": 1.3647320326006422,
+      "grad_norm": 0.49319113298532397,
+      "learning_rate": 4.956443225606269e-06,
+      "loss": 0.5619,
+      "step": 3453
+    },
+    {
+      "epoch": 1.3651271918992345,
+      "grad_norm": 0.4472089262359022,
+      "learning_rate": 4.9564141025445924e-06,
+      "loss": 0.5851,
+      "step": 3454
+    },
+    {
+      "epoch": 1.3655223511978267,
+      "grad_norm": 0.45279260771258567,
+      "learning_rate": 4.956384969835613e-06,
+      "loss": 0.5825,
+      "step": 3455
+    },
+    {
+      "epoch": 1.365917510496419,
+      "grad_norm": 0.44893832422728336,
+      "learning_rate": 4.956355827479445e-06,
+      "loss": 0.54,
+      "step": 3456
+    },
+    {
+      "epoch": 1.3663126697950112,
+      "grad_norm": 0.44861167684626757,
+      "learning_rate": 4.9563266754762025e-06,
+      "loss": 0.5798,
+      "step": 3457
+    },
+    {
+      "epoch": 1.3667078290936034,
+      "grad_norm": 0.457323231515835,
+      "learning_rate": 4.956297513826e-06,
+      "loss": 0.5578,
+      "step": 3458
+    },
+    {
+      "epoch": 1.3671029883921957,
+      "grad_norm": 0.4497319765240072,
+      "learning_rate": 4.9562683425289535e-06,
+      "loss": 0.5738,
+      "step": 3459
+    },
+    {
+      "epoch": 1.367498147690788,
+      "grad_norm": 0.44725620193345855,
+      "learning_rate": 4.956239161585176e-06,
+      "loss": 0.5727,
+      "step": 3460
+    },
+    {
+      "epoch": 1.3678933069893802,
+      "grad_norm": 0.4464598621025446,
+      "learning_rate": 4.956209970994783e-06,
+      "loss": 0.5661,
+      "step": 3461
+    },
+    {
+      "epoch": 1.3682884662879724,
+      "grad_norm": 0.4446893412743254,
+      "learning_rate": 4.956180770757888e-06,
+      "loss": 0.5732,
+      "step": 3462
+    },
+    {
+      "epoch": 1.3686836255865646,
+      "grad_norm": 0.43091284408122377,
+      "learning_rate": 4.956151560874607e-06,
+      "loss": 0.5573,
+      "step": 3463
+    },
+    {
+      "epoch": 1.3690787848851569,
+      "grad_norm": 0.4503486991715494,
+      "learning_rate": 4.9561223413450544e-06,
+      "loss": 0.5629,
+      "step": 3464
+    },
+    {
+      "epoch": 1.3694739441837491,
+      "grad_norm": 0.4661947809210117,
+      "learning_rate": 4.956093112169343e-06,
+      "loss": 0.5912,
+      "step": 3465
+    },
+    {
+      "epoch": 1.3698691034823414,
+      "grad_norm": 0.4420942969976414,
+      "learning_rate": 4.956063873347591e-06,
+      "loss": 0.5605,
+      "step": 3466
+    },
+    {
+      "epoch": 1.3702642627809336,
+      "grad_norm": 0.4299631865847097,
+      "learning_rate": 4.956034624879911e-06,
+      "loss": 0.5311,
+      "step": 3467
+    },
+    {
+      "epoch": 1.3706594220795258,
+      "grad_norm": 0.444327734857436,
+      "learning_rate": 4.956005366766419e-06,
+      "loss": 0.5744,
+      "step": 3468
+    },
+    {
+      "epoch": 1.371054581378118,
+      "grad_norm": 0.42540391618117634,
+      "learning_rate": 4.955976099007228e-06,
+      "loss": 0.572,
+      "step": 3469
+    },
+    {
+      "epoch": 1.3714497406767103,
+      "grad_norm": 0.4460903620872563,
+      "learning_rate": 4.955946821602455e-06,
+      "loss": 0.5715,
+      "step": 3470
+    },
+    {
+      "epoch": 1.3718448999753026,
+      "grad_norm": 0.4464339414945937,
+      "learning_rate": 4.9559175345522135e-06,
+      "loss": 0.5477,
+      "step": 3471
+    },
+    {
+      "epoch": 1.3722400592738948,
+      "grad_norm": 0.4496737708805918,
+      "learning_rate": 4.95588823785662e-06,
+      "loss": 0.566,
+      "step": 3472
+    },
+    {
+      "epoch": 1.372635218572487,
+      "grad_norm": 0.46530073162081453,
+      "learning_rate": 4.955858931515789e-06,
+      "loss": 0.5863,
+      "step": 3473
+    },
+    {
+      "epoch": 1.3730303778710793,
+      "grad_norm": 0.4304612284801493,
+      "learning_rate": 4.955829615529835e-06,
+      "loss": 0.5608,
+      "step": 3474
+    },
+    {
+      "epoch": 1.3734255371696715,
+      "grad_norm": 0.43919382720802014,
+      "learning_rate": 4.955800289898874e-06,
+      "loss": 0.5455,
+      "step": 3475
+    },
+    {
+      "epoch": 1.3738206964682638,
+      "grad_norm": 0.4570983105615984,
+      "learning_rate": 4.95577095462302e-06,
+      "loss": 0.5623,
+      "step": 3476
+    },
+    {
+      "epoch": 1.374215855766856,
+      "grad_norm": 0.4377489629257734,
+      "learning_rate": 4.955741609702389e-06,
+      "loss": 0.5599,
+      "step": 3477
+    },
+    {
+      "epoch": 1.3746110150654482,
+      "grad_norm": 0.457732827339079,
+      "learning_rate": 4.9557122551370964e-06,
+      "loss": 0.5607,
+      "step": 3478
+    },
+    {
+      "epoch": 1.3750061743640405,
+      "grad_norm": 0.44043665757163497,
+      "learning_rate": 4.955682890927257e-06,
+      "loss": 0.5563,
+      "step": 3479
+    },
+    {
+      "epoch": 1.3754013336626327,
+      "grad_norm": 0.4583584687679824,
+      "learning_rate": 4.955653517072986e-06,
+      "loss": 0.5743,
+      "step": 3480
+    },
+    {
+      "epoch": 1.375796492961225,
+      "grad_norm": 0.45153202964049305,
+      "learning_rate": 4.955624133574401e-06,
+      "loss": 0.5677,
+      "step": 3481
+    },
+    {
+      "epoch": 1.3761916522598172,
+      "grad_norm": 0.4374756911240904,
+      "learning_rate": 4.955594740431613e-06,
+      "loss": 0.5654,
+      "step": 3482
+    },
+    {
+      "epoch": 1.3765868115584095,
+      "grad_norm": 0.4556124603164875,
+      "learning_rate": 4.9555653376447416e-06,
+      "loss": 0.5649,
+      "step": 3483
+    },
+    {
+      "epoch": 1.3769819708570017,
+      "grad_norm": 0.4757752596484646,
+      "learning_rate": 4.9555359252139e-06,
+      "loss": 0.5888,
+      "step": 3484
+    },
+    {
+      "epoch": 1.377377130155594,
+      "grad_norm": 0.4465560956225096,
+      "learning_rate": 4.955506503139205e-06,
+      "loss": 0.5591,
+      "step": 3485
+    },
+    {
+      "epoch": 1.3777722894541862,
+      "grad_norm": 0.46143614783960507,
+      "learning_rate": 4.955477071420771e-06,
+      "loss": 0.5767,
+      "step": 3486
+    },
+    {
+      "epoch": 1.3781674487527784,
+      "grad_norm": 0.4625557795823603,
+      "learning_rate": 4.955447630058714e-06,
+      "loss": 0.5471,
+      "step": 3487
+    },
+    {
+      "epoch": 1.3785626080513707,
+      "grad_norm": 0.45122028220004684,
+      "learning_rate": 4.95541817905315e-06,
+      "loss": 0.5763,
+      "step": 3488
+    },
+    {
+      "epoch": 1.378957767349963,
+      "grad_norm": 0.4442976727237174,
+      "learning_rate": 4.955388718404194e-06,
+      "loss": 0.5665,
+      "step": 3489
+    },
+    {
+      "epoch": 1.3793529266485551,
+      "grad_norm": 0.46811037604603395,
+      "learning_rate": 4.955359248111963e-06,
+      "loss": 0.5561,
+      "step": 3490
+    },
+    {
+      "epoch": 1.3797480859471474,
+      "grad_norm": 0.45401020607684417,
+      "learning_rate": 4.955329768176571e-06,
+      "loss": 0.5673,
+      "step": 3491
+    },
+    {
+      "epoch": 1.3801432452457396,
+      "grad_norm": 0.44325397259906013,
+      "learning_rate": 4.955300278598135e-06,
+      "loss": 0.5571,
+      "step": 3492
+    },
+    {
+      "epoch": 1.3805384045443319,
+      "grad_norm": 0.4477474281264573,
+      "learning_rate": 4.955270779376771e-06,
+      "loss": 0.5621,
+      "step": 3493
+    },
+    {
+      "epoch": 1.380933563842924,
+      "grad_norm": 0.4768670398130126,
+      "learning_rate": 4.955241270512593e-06,
+      "loss": 0.5839,
+      "step": 3494
+    },
+    {
+      "epoch": 1.3813287231415163,
+      "grad_norm": 0.4460401555017405,
+      "learning_rate": 4.955211752005719e-06,
+      "loss": 0.5539,
+      "step": 3495
+    },
+    {
+      "epoch": 1.3817238824401086,
+      "grad_norm": 0.4366552836817992,
+      "learning_rate": 4.955182223856264e-06,
+      "loss": 0.5556,
+      "step": 3496
+    },
+    {
+      "epoch": 1.3821190417387008,
+      "grad_norm": 0.4493868061509665,
+      "learning_rate": 4.955152686064344e-06,
+      "loss": 0.5756,
+      "step": 3497
+    },
+    {
+      "epoch": 1.382514201037293,
+      "grad_norm": 0.42939645657453296,
+      "learning_rate": 4.955123138630075e-06,
+      "loss": 0.5705,
+      "step": 3498
+    },
+    {
+      "epoch": 1.3829093603358853,
+      "grad_norm": 0.4338686438980797,
+      "learning_rate": 4.955093581553574e-06,
+      "loss": 0.5505,
+      "step": 3499
+    },
+    {
+      "epoch": 1.3833045196344775,
+      "grad_norm": 0.44361106507257514,
+      "learning_rate": 4.955064014834955e-06,
+      "loss": 0.5488,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3836996789330698,
+      "grad_norm": 0.4420112874538524,
+      "learning_rate": 4.9550344384743365e-06,
+      "loss": 0.5624,
+      "step": 3501
+    },
+    {
+      "epoch": 1.384094838231662,
+      "grad_norm": 0.44997332812869023,
+      "learning_rate": 4.955004852471832e-06,
+      "loss": 0.5697,
+      "step": 3502
+    },
+    {
+      "epoch": 1.3844899975302543,
+      "grad_norm": 0.4591341499895522,
+      "learning_rate": 4.9549752568275605e-06,
+      "loss": 0.5587,
+      "step": 3503
+    },
+    {
+      "epoch": 1.3848851568288465,
+      "grad_norm": 0.460676095154433,
+      "learning_rate": 4.954945651541636e-06,
+      "loss": 0.565,
+      "step": 3504
+    },
+    {
+      "epoch": 1.3852803161274387,
+      "grad_norm": 0.4299069716032761,
+      "learning_rate": 4.954916036614177e-06,
+      "loss": 0.5545,
+      "step": 3505
+    },
+    {
+      "epoch": 1.3856754754260312,
+      "grad_norm": 0.43887083866057874,
+      "learning_rate": 4.954886412045298e-06,
+      "loss": 0.5921,
+      "step": 3506
+    },
+    {
+      "epoch": 1.3860706347246234,
+      "grad_norm": 0.42648077434777515,
+      "learning_rate": 4.954856777835115e-06,
+      "loss": 0.5456,
+      "step": 3507
+    },
+    {
+      "epoch": 1.3864657940232157,
+      "grad_norm": 0.4290139512004522,
+      "learning_rate": 4.954827133983746e-06,
+      "loss": 0.5867,
+      "step": 3508
+    },
+    {
+      "epoch": 1.386860953321808,
+      "grad_norm": 0.44830270070683675,
+      "learning_rate": 4.954797480491307e-06,
+      "loss": 0.5784,
+      "step": 3509
+    },
+    {
+      "epoch": 1.3872561126204002,
+      "grad_norm": 0.4477655067908414,
+      "learning_rate": 4.954767817357913e-06,
+      "loss": 0.5638,
+      "step": 3510
+    },
+    {
+      "epoch": 1.3876512719189924,
+      "grad_norm": 0.43658218747964356,
+      "learning_rate": 4.954738144583683e-06,
+      "loss": 0.5844,
+      "step": 3511
+    },
+    {
+      "epoch": 1.3880464312175846,
+      "grad_norm": 0.4393922211507673,
+      "learning_rate": 4.954708462168731e-06,
+      "loss": 0.5658,
+      "step": 3512
+    },
+    {
+      "epoch": 1.3884415905161769,
+      "grad_norm": 0.44104621414355816,
+      "learning_rate": 4.954678770113175e-06,
+      "loss": 0.5566,
+      "step": 3513
+    },
+    {
+      "epoch": 1.3888367498147691,
+      "grad_norm": 0.4578661431463306,
+      "learning_rate": 4.954649068417132e-06,
+      "loss": 0.581,
+      "step": 3514
+    },
+    {
+      "epoch": 1.3892319091133614,
+      "grad_norm": 0.44817762228122565,
+      "learning_rate": 4.954619357080717e-06,
+      "loss": 0.5633,
+      "step": 3515
+    },
+    {
+      "epoch": 1.3896270684119536,
+      "grad_norm": 0.4817436014812288,
+      "learning_rate": 4.954589636104049e-06,
+      "loss": 0.5447,
+      "step": 3516
+    },
+    {
+      "epoch": 1.3900222277105458,
+      "grad_norm": 0.46133823688460374,
+      "learning_rate": 4.954559905487242e-06,
+      "loss": 0.5746,
+      "step": 3517
+    },
+    {
+      "epoch": 1.390417387009138,
+      "grad_norm": 0.44329286540723245,
+      "learning_rate": 4.954530165230415e-06,
+      "loss": 0.5564,
+      "step": 3518
+    },
+    {
+      "epoch": 1.3908125463077303,
+      "grad_norm": 0.43383834817026123,
+      "learning_rate": 4.954500415333684e-06,
+      "loss": 0.5645,
+      "step": 3519
+    },
+    {
+      "epoch": 1.3912077056063226,
+      "grad_norm": 0.4378169397587514,
+      "learning_rate": 4.954470655797165e-06,
+      "loss": 0.5463,
+      "step": 3520
+    },
+    {
+      "epoch": 1.3916028649049148,
+      "grad_norm": 0.4550593061487435,
+      "learning_rate": 4.954440886620977e-06,
+      "loss": 0.5516,
+      "step": 3521
+    },
+    {
+      "epoch": 1.391998024203507,
+      "grad_norm": 0.4400286853367591,
+      "learning_rate": 4.9544111078052345e-06,
+      "loss": 0.5699,
+      "step": 3522
+    },
+    {
+      "epoch": 1.3923931835020993,
+      "grad_norm": 0.45981608819949793,
+      "learning_rate": 4.954381319350056e-06,
+      "loss": 0.569,
+      "step": 3523
+    },
+    {
+      "epoch": 1.3927883428006915,
+      "grad_norm": 0.44060684429684627,
+      "learning_rate": 4.9543515212555585e-06,
+      "loss": 0.5732,
+      "step": 3524
+    },
+    {
+      "epoch": 1.3931835020992838,
+      "grad_norm": 0.4516861113642003,
+      "learning_rate": 4.954321713521858e-06,
+      "loss": 0.5519,
+      "step": 3525
+    },
+    {
+      "epoch": 1.393578661397876,
+      "grad_norm": 0.4616718149420385,
+      "learning_rate": 4.954291896149072e-06,
+      "loss": 0.5516,
+      "step": 3526
+    },
+    {
+      "epoch": 1.3939738206964682,
+      "grad_norm": 0.4452993405905123,
+      "learning_rate": 4.954262069137318e-06,
+      "loss": 0.5572,
+      "step": 3527
+    },
+    {
+      "epoch": 1.3943689799950605,
+      "grad_norm": 0.434383788130944,
+      "learning_rate": 4.9542322324867136e-06,
+      "loss": 0.5472,
+      "step": 3528
+    },
+    {
+      "epoch": 1.3947641392936527,
+      "grad_norm": 0.44451550141787594,
+      "learning_rate": 4.954202386197375e-06,
+      "loss": 0.5587,
+      "step": 3529
+    },
+    {
+      "epoch": 1.395159298592245,
+      "grad_norm": 0.4481545332604534,
+      "learning_rate": 4.954172530269419e-06,
+      "loss": 0.5513,
+      "step": 3530
+    },
+    {
+      "epoch": 1.3955544578908372,
+      "grad_norm": 0.4396843141846436,
+      "learning_rate": 4.954142664702963e-06,
+      "loss": 0.5759,
+      "step": 3531
+    },
+    {
+      "epoch": 1.3959496171894294,
+      "grad_norm": 0.44315777199892625,
+      "learning_rate": 4.954112789498126e-06,
+      "loss": 0.5745,
+      "step": 3532
+    },
+    {
+      "epoch": 1.3963447764880217,
+      "grad_norm": 0.463836596007923,
+      "learning_rate": 4.9540829046550245e-06,
+      "loss": 0.5844,
+      "step": 3533
+    },
+    {
+      "epoch": 1.396739935786614,
+      "grad_norm": 0.527126180105154,
+      "learning_rate": 4.954053010173774e-06,
+      "loss": 0.5632,
+      "step": 3534
+    },
+    {
+      "epoch": 1.3971350950852062,
+      "grad_norm": 0.471207695771745,
+      "learning_rate": 4.954023106054495e-06,
+      "loss": 0.5727,
+      "step": 3535
+    },
+    {
+      "epoch": 1.3975302543837984,
+      "grad_norm": 0.4547374119776966,
+      "learning_rate": 4.953993192297303e-06,
+      "loss": 0.5639,
+      "step": 3536
+    },
+    {
+      "epoch": 1.3979254136823906,
+      "grad_norm": 0.4567625911659694,
+      "learning_rate": 4.953963268902315e-06,
+      "loss": 0.5772,
+      "step": 3537
+    },
+    {
+      "epoch": 1.398320572980983,
+      "grad_norm": 0.4241364653534491,
+      "learning_rate": 4.953933335869651e-06,
+      "loss": 0.5501,
+      "step": 3538
+    },
+    {
+      "epoch": 1.3987157322795751,
+      "grad_norm": 0.45962400820648397,
+      "learning_rate": 4.9539033931994255e-06,
+      "loss": 0.5597,
+      "step": 3539
+    },
+    {
+      "epoch": 1.3991108915781674,
+      "grad_norm": 0.44479747715613505,
+      "learning_rate": 4.953873440891758e-06,
+      "loss": 0.567,
+      "step": 3540
+    },
+    {
+      "epoch": 1.3995060508767596,
+      "grad_norm": 0.44139630414780884,
+      "learning_rate": 4.953843478946766e-06,
+      "loss": 0.5611,
+      "step": 3541
+    },
+    {
+      "epoch": 1.399901210175352,
+      "grad_norm": 0.44038065639815877,
+      "learning_rate": 4.953813507364566e-06,
+      "loss": 0.5572,
+      "step": 3542
+    },
+    {
+      "epoch": 1.4002963694739443,
+      "grad_norm": 0.44652707139959935,
+      "learning_rate": 4.9537835261452785e-06,
+      "loss": 0.552,
+      "step": 3543
+    },
+    {
+      "epoch": 1.4006915287725366,
+      "grad_norm": 0.447649590870608,
+      "learning_rate": 4.953753535289017e-06,
+      "loss": 0.55,
+      "step": 3544
+    },
+    {
+      "epoch": 1.4010866880711288,
+      "grad_norm": 0.4544049896666161,
+      "learning_rate": 4.953723534795903e-06,
+      "loss": 0.5504,
+      "step": 3545
+    },
+    {
+      "epoch": 1.401481847369721,
+      "grad_norm": 0.466964841617532,
+      "learning_rate": 4.953693524666054e-06,
+      "loss": 0.5722,
+      "step": 3546
+    },
+    {
+      "epoch": 1.4018770066683133,
+      "grad_norm": 0.527230905472802,
+      "learning_rate": 4.953663504899585e-06,
+      "loss": 0.5524,
+      "step": 3547
+    },
+    {
+      "epoch": 1.4022721659669055,
+      "grad_norm": 0.440070467466554,
+      "learning_rate": 4.953633475496615e-06,
+      "loss": 0.5412,
+      "step": 3548
+    },
+    {
+      "epoch": 1.4026673252654978,
+      "grad_norm": 0.4364642690771747,
+      "learning_rate": 4.9536034364572645e-06,
+      "loss": 0.5551,
+      "step": 3549
+    },
+    {
+      "epoch": 1.40306248456409,
+      "grad_norm": 0.45824115876308913,
+      "learning_rate": 4.953573387781649e-06,
+      "loss": 0.5731,
+      "step": 3550
+    },
+    {
+      "epoch": 1.4034576438626822,
+      "grad_norm": 0.4559809654130625,
+      "learning_rate": 4.9535433294698865e-06,
+      "loss": 0.5735,
+      "step": 3551
+    },
+    {
+      "epoch": 1.4038528031612745,
+      "grad_norm": 0.4437519963912747,
+      "learning_rate": 4.9535132615220965e-06,
+      "loss": 0.5715,
+      "step": 3552
+    },
+    {
+      "epoch": 1.4042479624598667,
+      "grad_norm": 0.45338834415060414,
+      "learning_rate": 4.953483183938395e-06,
+      "loss": 0.5724,
+      "step": 3553
+    },
+    {
+      "epoch": 1.404643121758459,
+      "grad_norm": 0.4568209368624167,
+      "learning_rate": 4.953453096718903e-06,
+      "loss": 0.5581,
+      "step": 3554
+    },
+    {
+      "epoch": 1.4050382810570512,
+      "grad_norm": 0.43314731805578655,
+      "learning_rate": 4.953422999863736e-06,
+      "loss": 0.5379,
+      "step": 3555
+    },
+    {
+      "epoch": 1.4054334403556434,
+      "grad_norm": 0.45449692655904694,
+      "learning_rate": 4.953392893373015e-06,
+      "loss": 0.5739,
+      "step": 3556
+    },
+    {
+      "epoch": 1.4058285996542357,
+      "grad_norm": 0.4562200435387982,
+      "learning_rate": 4.953362777246855e-06,
+      "loss": 0.5643,
+      "step": 3557
+    },
+    {
+      "epoch": 1.406223758952828,
+      "grad_norm": 0.4440850256076502,
+      "learning_rate": 4.953332651485375e-06,
+      "loss": 0.5689,
+      "step": 3558
+    },
+    {
+      "epoch": 1.4066189182514202,
+      "grad_norm": 0.43280961705826465,
+      "learning_rate": 4.953302516088695e-06,
+      "loss": 0.5452,
+      "step": 3559
+    },
+    {
+      "epoch": 1.4070140775500124,
+      "grad_norm": 0.43383955128353996,
+      "learning_rate": 4.953272371056933e-06,
+      "loss": 0.5437,
+      "step": 3560
+    },
+    {
+      "epoch": 1.4074092368486046,
+      "grad_norm": 0.46587228167336536,
+      "learning_rate": 4.953242216390206e-06,
+      "loss": 0.5626,
+      "step": 3561
+    },
+    {
+      "epoch": 1.4078043961471969,
+      "grad_norm": 0.4627748388388923,
+      "learning_rate": 4.953212052088634e-06,
+      "loss": 0.5578,
+      "step": 3562
+    },
+    {
+      "epoch": 1.4081995554457891,
+      "grad_norm": 0.4494606350210151,
+      "learning_rate": 4.953181878152334e-06,
+      "loss": 0.559,
+      "step": 3563
+    },
+    {
+      "epoch": 1.4085947147443814,
+      "grad_norm": 0.4689503889787178,
+      "learning_rate": 4.953151694581425e-06,
+      "loss": 0.5853,
+      "step": 3564
+    },
+    {
+      "epoch": 1.4089898740429736,
+      "grad_norm": 0.425074031613252,
+      "learning_rate": 4.953121501376027e-06,
+      "loss": 0.5469,
+      "step": 3565
+    },
+    {
+      "epoch": 1.4093850333415658,
+      "grad_norm": 0.43908518140110436,
+      "learning_rate": 4.953091298536256e-06,
+      "loss": 0.5573,
+      "step": 3566
+    },
+    {
+      "epoch": 1.409780192640158,
+      "grad_norm": 0.456588536080327,
+      "learning_rate": 4.953061086062233e-06,
+      "loss": 0.5738,
+      "step": 3567
+    },
+    {
+      "epoch": 1.4101753519387503,
+      "grad_norm": 0.4407899689038523,
+      "learning_rate": 4.953030863954075e-06,
+      "loss": 0.5485,
+      "step": 3568
+    },
+    {
+      "epoch": 1.4105705112373426,
+      "grad_norm": 0.44983477733657745,
+      "learning_rate": 4.953000632211902e-06,
+      "loss": 0.5587,
+      "step": 3569
+    },
+    {
+      "epoch": 1.4109656705359348,
+      "grad_norm": 0.4444516608223304,
+      "learning_rate": 4.952970390835831e-06,
+      "loss": 0.56,
+      "step": 3570
+    },
+    {
+      "epoch": 1.411360829834527,
+      "grad_norm": 0.44234362530280924,
+      "learning_rate": 4.952940139825982e-06,
+      "loss": 0.58,
+      "step": 3571
+    },
+    {
+      "epoch": 1.4117559891331193,
+      "grad_norm": 0.4303479209786135,
+      "learning_rate": 4.952909879182475e-06,
+      "loss": 0.5487,
+      "step": 3572
+    },
+    {
+      "epoch": 1.4121511484317115,
+      "grad_norm": 0.4395260223716139,
+      "learning_rate": 4.952879608905427e-06,
+      "loss": 0.5697,
+      "step": 3573
+    },
+    {
+      "epoch": 1.4125463077303038,
+      "grad_norm": 0.43025433618329334,
+      "learning_rate": 4.952849328994957e-06,
+      "loss": 0.5622,
+      "step": 3574
+    },
+    {
+      "epoch": 1.412941467028896,
+      "grad_norm": 0.4425864856013525,
+      "learning_rate": 4.9528190394511835e-06,
+      "loss": 0.5472,
+      "step": 3575
+    },
+    {
+      "epoch": 1.4133366263274882,
+      "grad_norm": 0.4515220954726901,
+      "learning_rate": 4.9527887402742266e-06,
+      "loss": 0.5593,
+      "step": 3576
+    },
+    {
+      "epoch": 1.4137317856260805,
+      "grad_norm": 0.4435809550125933,
+      "learning_rate": 4.952758431464206e-06,
+      "loss": 0.5604,
+      "step": 3577
+    },
+    {
+      "epoch": 1.4141269449246727,
+      "grad_norm": 0.45887561975872704,
+      "learning_rate": 4.952728113021239e-06,
+      "loss": 0.5573,
+      "step": 3578
+    },
+    {
+      "epoch": 1.414522104223265,
+      "grad_norm": 0.4382166236259494,
+      "learning_rate": 4.952697784945445e-06,
+      "loss": 0.5767,
+      "step": 3579
+    },
+    {
+      "epoch": 1.4149172635218572,
+      "grad_norm": 0.43576829414699486,
+      "learning_rate": 4.952667447236944e-06,
+      "loss": 0.5485,
+      "step": 3580
+    },
+    {
+      "epoch": 1.4153124228204494,
+      "grad_norm": 0.6295771726774199,
+      "learning_rate": 4.952637099895854e-06,
+      "loss": 0.5594,
+      "step": 3581
+    },
+    {
+      "epoch": 1.4157075821190417,
+      "grad_norm": 0.44289940633979036,
+      "learning_rate": 4.952606742922296e-06,
+      "loss": 0.5492,
+      "step": 3582
+    },
+    {
+      "epoch": 1.416102741417634,
+      "grad_norm": 0.44005534309853256,
+      "learning_rate": 4.952576376316387e-06,
+      "loss": 0.5488,
+      "step": 3583
+    },
+    {
+      "epoch": 1.4164979007162262,
+      "grad_norm": 0.5156052999824663,
+      "learning_rate": 4.952546000078247e-06,
+      "loss": 0.5623,
+      "step": 3584
+    },
+    {
+      "epoch": 1.4168930600148184,
+      "grad_norm": 0.4720203506988176,
+      "learning_rate": 4.952515614207996e-06,
+      "loss": 0.5709,
+      "step": 3585
+    },
+    {
+      "epoch": 1.4172882193134106,
+      "grad_norm": 0.4403789400135003,
+      "learning_rate": 4.952485218705753e-06,
+      "loss": 0.5612,
+      "step": 3586
+    },
+    {
+      "epoch": 1.4176833786120029,
+      "grad_norm": 0.4346034283155917,
+      "learning_rate": 4.952454813571638e-06,
+      "loss": 0.5532,
+      "step": 3587
+    },
+    {
+      "epoch": 1.4180785379105951,
+      "grad_norm": 0.4914958621897474,
+      "learning_rate": 4.952424398805769e-06,
+      "loss": 0.5874,
+      "step": 3588
+    },
+    {
+      "epoch": 1.4184736972091874,
+      "grad_norm": 0.44158849976736936,
+      "learning_rate": 4.952393974408265e-06,
+      "loss": 0.558,
+      "step": 3589
+    },
+    {
+      "epoch": 1.4188688565077796,
+      "grad_norm": 0.47059542127549936,
+      "learning_rate": 4.952363540379248e-06,
+      "loss": 0.56,
+      "step": 3590
+    },
+    {
+      "epoch": 1.4192640158063718,
+      "grad_norm": 0.4428834340575035,
+      "learning_rate": 4.952333096718837e-06,
+      "loss": 0.5779,
+      "step": 3591
+    },
+    {
+      "epoch": 1.419659175104964,
+      "grad_norm": 0.4318718915548664,
+      "learning_rate": 4.952302643427149e-06,
+      "loss": 0.5545,
+      "step": 3592
+    },
+    {
+      "epoch": 1.4200543344035563,
+      "grad_norm": 0.4399415267342125,
+      "learning_rate": 4.952272180504306e-06,
+      "loss": 0.5605,
+      "step": 3593
+    },
+    {
+      "epoch": 1.4204494937021486,
+      "grad_norm": 0.44720706398972543,
+      "learning_rate": 4.952241707950427e-06,
+      "loss": 0.5722,
+      "step": 3594
+    },
+    {
+      "epoch": 1.4208446530007408,
+      "grad_norm": 0.43958222609485287,
+      "learning_rate": 4.9522112257656315e-06,
+      "loss": 0.5474,
+      "step": 3595
+    },
+    {
+      "epoch": 1.421239812299333,
+      "grad_norm": 0.4335981477530248,
+      "learning_rate": 4.952180733950039e-06,
+      "loss": 0.5406,
+      "step": 3596
+    },
+    {
+      "epoch": 1.4216349715979253,
+      "grad_norm": 0.4452745792193868,
+      "learning_rate": 4.952150232503771e-06,
+      "loss": 0.5774,
+      "step": 3597
+    },
+    {
+      "epoch": 1.4220301308965175,
+      "grad_norm": 0.450111748575881,
+      "learning_rate": 4.952119721426945e-06,
+      "loss": 0.564,
+      "step": 3598
+    },
+    {
+      "epoch": 1.4224252901951098,
+      "grad_norm": 0.4514178883673883,
+      "learning_rate": 4.952089200719682e-06,
+      "loss": 0.5702,
+      "step": 3599
+    },
+    {
+      "epoch": 1.422820449493702,
+      "grad_norm": 0.44869890712620886,
+      "learning_rate": 4.9520586703821006e-06,
+      "loss": 0.5695,
+      "step": 3600
+    },
+    {
+      "epoch": 1.4232156087922945,
+      "grad_norm": 0.47920918782737604,
+      "learning_rate": 4.952028130414322e-06,
+      "loss": 0.5538,
+      "step": 3601
+    },
+    {
+      "epoch": 1.4236107680908867,
+      "grad_norm": 0.4359999152882843,
+      "learning_rate": 4.951997580816466e-06,
+      "loss": 0.5561,
+      "step": 3602
+    },
+    {
+      "epoch": 1.424005927389479,
+      "grad_norm": 0.4603603266157431,
+      "learning_rate": 4.951967021588654e-06,
+      "loss": 0.5539,
+      "step": 3603
+    },
+    {
+      "epoch": 1.4244010866880712,
+      "grad_norm": 0.4470887413164905,
+      "learning_rate": 4.9519364527310035e-06,
+      "loss": 0.557,
+      "step": 3604
+    },
+    {
+      "epoch": 1.4247962459866634,
+      "grad_norm": 0.45851610322617087,
+      "learning_rate": 4.9519058742436345e-06,
+      "loss": 0.5501,
+      "step": 3605
+    },
+    {
+      "epoch": 1.4251914052852557,
+      "grad_norm": 0.47419186727157,
+      "learning_rate": 4.951875286126669e-06,
+      "loss": 0.577,
+      "step": 3606
+    },
+    {
+      "epoch": 1.425586564583848,
+      "grad_norm": 0.4495038740005354,
+      "learning_rate": 4.951844688380226e-06,
+      "loss": 0.5761,
+      "step": 3607
+    },
+    {
+      "epoch": 1.4259817238824402,
+      "grad_norm": 0.42780708612897866,
+      "learning_rate": 4.951814081004426e-06,
+      "loss": 0.5507,
+      "step": 3608
+    },
+    {
+      "epoch": 1.4263768831810324,
+      "grad_norm": 0.4420018249267614,
+      "learning_rate": 4.951783463999389e-06,
+      "loss": 0.5659,
+      "step": 3609
+    },
+    {
+      "epoch": 1.4267720424796246,
+      "grad_norm": 0.4321949197983691,
+      "learning_rate": 4.951752837365236e-06,
+      "loss": 0.5578,
+      "step": 3610
+    },
+    {
+      "epoch": 1.4271672017782169,
+      "grad_norm": 0.45648209967593256,
+      "learning_rate": 4.951722201102085e-06,
+      "loss": 0.5829,
+      "step": 3611
+    },
+    {
+      "epoch": 1.4275623610768091,
+      "grad_norm": 0.4448948600493821,
+      "learning_rate": 4.9516915552100594e-06,
+      "loss": 0.5672,
+      "step": 3612
+    },
+    {
+      "epoch": 1.4279575203754014,
+      "grad_norm": 0.4357660265947906,
+      "learning_rate": 4.951660899689278e-06,
+      "loss": 0.5473,
+      "step": 3613
+    },
+    {
+      "epoch": 1.4283526796739936,
+      "grad_norm": 0.4697567901251499,
+      "learning_rate": 4.951630234539861e-06,
+      "loss": 0.5793,
+      "step": 3614
+    },
+    {
+      "epoch": 1.4287478389725858,
+      "grad_norm": 0.4474239868294111,
+      "learning_rate": 4.951599559761929e-06,
+      "loss": 0.5559,
+      "step": 3615
+    },
+    {
+      "epoch": 1.429142998271178,
+      "grad_norm": 0.46638942204948125,
+      "learning_rate": 4.951568875355603e-06,
+      "loss": 0.5674,
+      "step": 3616
+    },
+    {
+      "epoch": 1.4295381575697703,
+      "grad_norm": 0.46201733877960605,
+      "learning_rate": 4.951538181321003e-06,
+      "loss": 0.565,
+      "step": 3617
+    },
+    {
+      "epoch": 1.4299333168683626,
+      "grad_norm": 0.4468591448162063,
+      "learning_rate": 4.9515074776582495e-06,
+      "loss": 0.5644,
+      "step": 3618
+    },
+    {
+      "epoch": 1.4303284761669548,
+      "grad_norm": 0.4563343700830998,
+      "learning_rate": 4.951476764367463e-06,
+      "loss": 0.5749,
+      "step": 3619
+    },
+    {
+      "epoch": 1.430723635465547,
+      "grad_norm": 0.43658851767260776,
+      "learning_rate": 4.951446041448765e-06,
+      "loss": 0.5736,
+      "step": 3620
+    },
+    {
+      "epoch": 1.4311187947641393,
+      "grad_norm": 0.4349415610360004,
+      "learning_rate": 4.951415308902275e-06,
+      "loss": 0.5596,
+      "step": 3621
+    },
+    {
+      "epoch": 1.4315139540627315,
+      "grad_norm": 0.462135397048311,
+      "learning_rate": 4.951384566728115e-06,
+      "loss": 0.5595,
+      "step": 3622
+    },
+    {
+      "epoch": 1.4319091133613238,
+      "grad_norm": 0.4543875910238101,
+      "learning_rate": 4.951353814926405e-06,
+      "loss": 0.5489,
+      "step": 3623
+    },
+    {
+      "epoch": 1.432304272659916,
+      "grad_norm": 0.4297719147230641,
+      "learning_rate": 4.951323053497265e-06,
+      "loss": 0.565,
+      "step": 3624
+    },
+    {
+      "epoch": 1.4326994319585082,
+      "grad_norm": 0.45442855434813045,
+      "learning_rate": 4.951292282440817e-06,
+      "loss": 0.5698,
+      "step": 3625
+    },
+    {
+      "epoch": 1.4330945912571005,
+      "grad_norm": 0.5225354098317048,
+      "learning_rate": 4.951261501757182e-06,
+      "loss": 0.5531,
+      "step": 3626
+    },
+    {
+      "epoch": 1.4334897505556927,
+      "grad_norm": 0.4420068105595844,
+      "learning_rate": 4.951230711446479e-06,
+      "loss": 0.5766,
+      "step": 3627
+    },
+    {
+      "epoch": 1.433884909854285,
+      "grad_norm": 0.44501142719182346,
+      "learning_rate": 4.951199911508831e-06,
+      "loss": 0.5568,
+      "step": 3628
+    },
+    {
+      "epoch": 1.4342800691528772,
+      "grad_norm": 0.5349036907902572,
+      "learning_rate": 4.951169101944358e-06,
+      "loss": 0.5508,
+      "step": 3629
+    },
+    {
+      "epoch": 1.4346752284514694,
+      "grad_norm": 0.46471714809625664,
+      "learning_rate": 4.951138282753181e-06,
+      "loss": 0.579,
+      "step": 3630
+    },
+    {
+      "epoch": 1.4350703877500617,
+      "grad_norm": 0.4425488778078077,
+      "learning_rate": 4.951107453935421e-06,
+      "loss": 0.5469,
+      "step": 3631
+    },
+    {
+      "epoch": 1.435465547048654,
+      "grad_norm": 0.4395384121942282,
+      "learning_rate": 4.951076615491201e-06,
+      "loss": 0.5469,
+      "step": 3632
+    },
+    {
+      "epoch": 1.4358607063472462,
+      "grad_norm": 0.45264206492150943,
+      "learning_rate": 4.9510457674206385e-06,
+      "loss": 0.5669,
+      "step": 3633
+    },
+    {
+      "epoch": 1.4362558656458384,
+      "grad_norm": 0.4359952505534556,
+      "learning_rate": 4.951014909723858e-06,
+      "loss": 0.5581,
+      "step": 3634
+    },
+    {
+      "epoch": 1.4366510249444306,
+      "grad_norm": 0.43031272641598334,
+      "learning_rate": 4.950984042400978e-06,
+      "loss": 0.5614,
+      "step": 3635
+    },
+    {
+      "epoch": 1.437046184243023,
+      "grad_norm": 0.44952139574949684,
+      "learning_rate": 4.9509531654521216e-06,
+      "loss": 0.564,
+      "step": 3636
+    },
+    {
+      "epoch": 1.4374413435416153,
+      "grad_norm": 0.4894458991545237,
+      "learning_rate": 4.950922278877409e-06,
+      "loss": 0.5373,
+      "step": 3637
+    },
+    {
+      "epoch": 1.4378365028402076,
+      "grad_norm": 0.4435614177631397,
+      "learning_rate": 4.950891382676963e-06,
+      "loss": 0.561,
+      "step": 3638
+    },
+    {
+      "epoch": 1.4382316621387998,
+      "grad_norm": 0.4556594196014496,
+      "learning_rate": 4.950860476850903e-06,
+      "loss": 0.5612,
+      "step": 3639
+    },
+    {
+      "epoch": 1.438626821437392,
+      "grad_norm": 0.43556778440010696,
+      "learning_rate": 4.9508295613993515e-06,
+      "loss": 0.5494,
+      "step": 3640
+    },
+    {
+      "epoch": 1.4390219807359843,
+      "grad_norm": 0.4521754341667638,
+      "learning_rate": 4.9507986363224305e-06,
+      "loss": 0.5526,
+      "step": 3641
+    },
+    {
+      "epoch": 1.4394171400345765,
+      "grad_norm": 0.439981945538117,
+      "learning_rate": 4.950767701620259e-06,
+      "loss": 0.548,
+      "step": 3642
+    },
+    {
+      "epoch": 1.4398122993331688,
+      "grad_norm": 0.4351939027647017,
+      "learning_rate": 4.950736757292962e-06,
+      "loss": 0.5698,
+      "step": 3643
+    },
+    {
+      "epoch": 1.440207458631761,
+      "grad_norm": 0.43466458525745466,
+      "learning_rate": 4.950705803340657e-06,
+      "loss": 0.5595,
+      "step": 3644
+    },
+    {
+      "epoch": 1.4406026179303533,
+      "grad_norm": 0.4482718750704879,
+      "learning_rate": 4.9506748397634695e-06,
+      "loss": 0.5714,
+      "step": 3645
+    },
+    {
+      "epoch": 1.4409977772289455,
+      "grad_norm": 0.4424752556047753,
+      "learning_rate": 4.9506438665615195e-06,
+      "loss": 0.5542,
+      "step": 3646
+    },
+    {
+      "epoch": 1.4413929365275377,
+      "grad_norm": 0.4455607912735244,
+      "learning_rate": 4.950612883734928e-06,
+      "loss": 0.5591,
+      "step": 3647
+    },
+    {
+      "epoch": 1.44178809582613,
+      "grad_norm": 0.44170634526996,
+      "learning_rate": 4.950581891283816e-06,
+      "loss": 0.5546,
+      "step": 3648
+    },
+    {
+      "epoch": 1.4421832551247222,
+      "grad_norm": 0.45300602248436533,
+      "learning_rate": 4.950550889208308e-06,
+      "loss": 0.5494,
+      "step": 3649
+    },
+    {
+      "epoch": 1.4425784144233145,
+      "grad_norm": 0.45683270516581864,
+      "learning_rate": 4.950519877508524e-06,
+      "loss": 0.5805,
+      "step": 3650
+    },
+    {
+      "epoch": 1.4429735737219067,
+      "grad_norm": 0.4527318497547736,
+      "learning_rate": 4.950488856184585e-06,
+      "loss": 0.5625,
+      "step": 3651
+    },
+    {
+      "epoch": 1.443368733020499,
+      "grad_norm": 0.4384333802721848,
+      "learning_rate": 4.950457825236615e-06,
+      "loss": 0.5624,
+      "step": 3652
+    },
+    {
+      "epoch": 1.4437638923190912,
+      "grad_norm": 0.48213960406408873,
+      "learning_rate": 4.950426784664734e-06,
+      "loss": 0.5796,
+      "step": 3653
+    },
+    {
+      "epoch": 1.4441590516176834,
+      "grad_norm": 0.4656284274871252,
+      "learning_rate": 4.950395734469065e-06,
+      "loss": 0.5488,
+      "step": 3654
+    },
+    {
+      "epoch": 1.4445542109162757,
+      "grad_norm": 0.4342682580148919,
+      "learning_rate": 4.950364674649729e-06,
+      "loss": 0.5465,
+      "step": 3655
+    },
+    {
+      "epoch": 1.444949370214868,
+      "grad_norm": 0.4371079121432642,
+      "learning_rate": 4.9503336052068485e-06,
+      "loss": 0.5622,
+      "step": 3656
+    },
+    {
+      "epoch": 1.4453445295134602,
+      "grad_norm": 0.43992242202525345,
+      "learning_rate": 4.9503025261405455e-06,
+      "loss": 0.5514,
+      "step": 3657
+    },
+    {
+      "epoch": 1.4457396888120524,
+      "grad_norm": 0.4414816507834958,
+      "learning_rate": 4.950271437450943e-06,
+      "loss": 0.5563,
+      "step": 3658
+    },
+    {
+      "epoch": 1.4461348481106446,
+      "grad_norm": 0.44005325021141295,
+      "learning_rate": 4.950240339138161e-06,
+      "loss": 0.5747,
+      "step": 3659
+    },
+    {
+      "epoch": 1.4465300074092369,
+      "grad_norm": 0.4421664700998617,
+      "learning_rate": 4.950209231202323e-06,
+      "loss": 0.5498,
+      "step": 3660
+    },
+    {
+      "epoch": 1.4469251667078291,
+      "grad_norm": 0.43113979255459345,
+      "learning_rate": 4.950178113643551e-06,
+      "loss": 0.5798,
+      "step": 3661
+    },
+    {
+      "epoch": 1.4473203260064214,
+      "grad_norm": 0.4708087923129291,
+      "learning_rate": 4.950146986461968e-06,
+      "loss": 0.5421,
+      "step": 3662
+    },
+    {
+      "epoch": 1.4477154853050136,
+      "grad_norm": 0.45141956435396696,
+      "learning_rate": 4.9501158496576945e-06,
+      "loss": 0.5613,
+      "step": 3663
+    },
+    {
+      "epoch": 1.4481106446036058,
+      "grad_norm": 0.45661303582282825,
+      "learning_rate": 4.950084703230854e-06,
+      "loss": 0.5821,
+      "step": 3664
+    },
+    {
+      "epoch": 1.448505803902198,
+      "grad_norm": 0.4467554956414753,
+      "learning_rate": 4.9500535471815696e-06,
+      "loss": 0.5614,
+      "step": 3665
+    },
+    {
+      "epoch": 1.4489009632007903,
+      "grad_norm": 0.4375691215444136,
+      "learning_rate": 4.950022381509961e-06,
+      "loss": 0.5546,
+      "step": 3666
+    },
+    {
+      "epoch": 1.4492961224993826,
+      "grad_norm": 0.44417059941621506,
+      "learning_rate": 4.949991206216152e-06,
+      "loss": 0.5474,
+      "step": 3667
+    },
+    {
+      "epoch": 1.4496912817979748,
+      "grad_norm": 0.5239384393901027,
+      "learning_rate": 4.949960021300267e-06,
+      "loss": 0.5724,
+      "step": 3668
+    },
+    {
+      "epoch": 1.450086441096567,
+      "grad_norm": 0.4600101751343761,
+      "learning_rate": 4.949928826762425e-06,
+      "loss": 0.5643,
+      "step": 3669
+    },
+    {
+      "epoch": 1.4504816003951593,
+      "grad_norm": 0.4398883196726322,
+      "learning_rate": 4.949897622602752e-06,
+      "loss": 0.5645,
+      "step": 3670
+    },
+    {
+      "epoch": 1.4508767596937515,
+      "grad_norm": 0.4243916986024868,
+      "learning_rate": 4.949866408821368e-06,
+      "loss": 0.5414,
+      "step": 3671
+    },
+    {
+      "epoch": 1.4512719189923438,
+      "grad_norm": 0.46429052461409404,
+      "learning_rate": 4.949835185418397e-06,
+      "loss": 0.5675,
+      "step": 3672
+    },
+    {
+      "epoch": 1.451667078290936,
+      "grad_norm": 0.43727146868324845,
+      "learning_rate": 4.94980395239396e-06,
+      "loss": 0.5411,
+      "step": 3673
+    },
+    {
+      "epoch": 1.4520622375895282,
+      "grad_norm": 0.44296006305796626,
+      "learning_rate": 4.94977270974818e-06,
+      "loss": 0.5706,
+      "step": 3674
+    },
+    {
+      "epoch": 1.4524573968881205,
+      "grad_norm": 0.43581736362129647,
+      "learning_rate": 4.949741457481182e-06,
+      "loss": 0.5579,
+      "step": 3675
+    },
+    {
+      "epoch": 1.4528525561867127,
+      "grad_norm": 0.43935390623902243,
+      "learning_rate": 4.949710195593087e-06,
+      "loss": 0.5522,
+      "step": 3676
+    },
+    {
+      "epoch": 1.453247715485305,
+      "grad_norm": 0.4604206853638542,
+      "learning_rate": 4.949678924084017e-06,
+      "loss": 0.5861,
+      "step": 3677
+    },
+    {
+      "epoch": 1.4536428747838972,
+      "grad_norm": 0.4666269167795469,
+      "learning_rate": 4.949647642954096e-06,
+      "loss": 0.5733,
+      "step": 3678
+    },
+    {
+      "epoch": 1.4540380340824894,
+      "grad_norm": 0.4425711605975332,
+      "learning_rate": 4.949616352203447e-06,
+      "loss": 0.5599,
+      "step": 3679
+    },
+    {
+      "epoch": 1.4544331933810817,
+      "grad_norm": 0.438259767331312,
+      "learning_rate": 4.949585051832192e-06,
+      "loss": 0.5512,
+      "step": 3680
+    },
+    {
+      "epoch": 1.454828352679674,
+      "grad_norm": 0.47194140077374513,
+      "learning_rate": 4.949553741840455e-06,
+      "loss": 0.5684,
+      "step": 3681
+    },
+    {
+      "epoch": 1.4552235119782662,
+      "grad_norm": 0.4581698913385663,
+      "learning_rate": 4.9495224222283576e-06,
+      "loss": 0.5731,
+      "step": 3682
+    },
+    {
+      "epoch": 1.4556186712768584,
+      "grad_norm": 0.4729218162827143,
+      "learning_rate": 4.949491092996024e-06,
+      "loss": 0.5581,
+      "step": 3683
+    },
+    {
+      "epoch": 1.4560138305754506,
+      "grad_norm": 0.4631393154498066,
+      "learning_rate": 4.9494597541435764e-06,
+      "loss": 0.5891,
+      "step": 3684
+    },
+    {
+      "epoch": 1.4564089898740429,
+      "grad_norm": 0.44412962722285376,
+      "learning_rate": 4.949428405671138e-06,
+      "loss": 0.5669,
+      "step": 3685
+    },
+    {
+      "epoch": 1.4568041491726351,
+      "grad_norm": 0.44251521812090155,
+      "learning_rate": 4.949397047578833e-06,
+      "loss": 0.5767,
+      "step": 3686
+    },
+    {
+      "epoch": 1.4571993084712274,
+      "grad_norm": 0.5568404678574649,
+      "learning_rate": 4.949365679866783e-06,
+      "loss": 0.5695,
+      "step": 3687
+    },
+    {
+      "epoch": 1.4575944677698196,
+      "grad_norm": 0.4597237274269838,
+      "learning_rate": 4.9493343025351125e-06,
+      "loss": 0.5841,
+      "step": 3688
+    },
+    {
+      "epoch": 1.4579896270684118,
+      "grad_norm": 0.433320088836893,
+      "learning_rate": 4.9493029155839435e-06,
+      "loss": 0.5549,
+      "step": 3689
+    },
+    {
+      "epoch": 1.458384786367004,
+      "grad_norm": 0.4408848451488535,
+      "learning_rate": 4.949271519013401e-06,
+      "loss": 0.5628,
+      "step": 3690
+    },
+    {
+      "epoch": 1.4587799456655963,
+      "grad_norm": 0.45052941581991246,
+      "learning_rate": 4.949240112823606e-06,
+      "loss": 0.5777,
+      "step": 3691
+    },
+    {
+      "epoch": 1.4591751049641886,
+      "grad_norm": 0.5775749617611324,
+      "learning_rate": 4.949208697014685e-06,
+      "loss": 0.5599,
+      "step": 3692
+    },
+    {
+      "epoch": 1.4595702642627808,
+      "grad_norm": 0.44576099623969545,
+      "learning_rate": 4.949177271586758e-06,
+      "loss": 0.5525,
+      "step": 3693
+    },
+    {
+      "epoch": 1.459965423561373,
+      "grad_norm": 0.44043843064525395,
+      "learning_rate": 4.94914583653995e-06,
+      "loss": 0.5702,
+      "step": 3694
+    },
+    {
+      "epoch": 1.4603605828599655,
+      "grad_norm": 0.47947522923316277,
+      "learning_rate": 4.9491143918743845e-06,
+      "loss": 0.5558,
+      "step": 3695
+    },
+    {
+      "epoch": 1.4607557421585577,
+      "grad_norm": 0.4592839139494569,
+      "learning_rate": 4.949082937590185e-06,
+      "loss": 0.5702,
+      "step": 3696
+    },
+    {
+      "epoch": 1.46115090145715,
+      "grad_norm": 0.45188516970792914,
+      "learning_rate": 4.949051473687475e-06,
+      "loss": 0.5528,
+      "step": 3697
+    },
+    {
+      "epoch": 1.4615460607557422,
+      "grad_norm": 2.2730337039449893,
+      "learning_rate": 4.949020000166378e-06,
+      "loss": 0.5541,
+      "step": 3698
+    },
+    {
+      "epoch": 1.4619412200543345,
+      "grad_norm": 0.4339995210953989,
+      "learning_rate": 4.948988517027017e-06,
+      "loss": 0.5729,
+      "step": 3699
+    },
+    {
+      "epoch": 1.4623363793529267,
+      "grad_norm": 0.4500818289037516,
+      "learning_rate": 4.948957024269516e-06,
+      "loss": 0.5613,
+      "step": 3700
+    },
+    {
+      "epoch": 1.462731538651519,
+      "grad_norm": 0.44742467459872376,
+      "learning_rate": 4.948925521894e-06,
+      "loss": 0.5351,
+      "step": 3701
+    },
+    {
+      "epoch": 1.4631266979501112,
+      "grad_norm": 0.4499910805435301,
+      "learning_rate": 4.948894009900591e-06,
+      "loss": 0.5574,
+      "step": 3702
+    },
+    {
+      "epoch": 1.4635218572487034,
+      "grad_norm": 0.4357059717410755,
+      "learning_rate": 4.948862488289413e-06,
+      "loss": 0.552,
+      "step": 3703
+    },
+    {
+      "epoch": 1.4639170165472957,
+      "grad_norm": 0.6496169252590506,
+      "learning_rate": 4.948830957060591e-06,
+      "loss": 0.5701,
+      "step": 3704
+    },
+    {
+      "epoch": 1.464312175845888,
+      "grad_norm": 0.6005901899539847,
+      "learning_rate": 4.948799416214247e-06,
+      "loss": 0.5604,
+      "step": 3705
+    },
+    {
+      "epoch": 1.4647073351444801,
+      "grad_norm": 0.45141656213054204,
+      "learning_rate": 4.9487678657505065e-06,
+      "loss": 0.573,
+      "step": 3706
+    },
+    {
+      "epoch": 1.4651024944430724,
+      "grad_norm": 0.46166246230338953,
+      "learning_rate": 4.948736305669494e-06,
+      "loss": 0.5826,
+      "step": 3707
+    },
+    {
+      "epoch": 1.4654976537416646,
+      "grad_norm": 0.4485874822386171,
+      "learning_rate": 4.9487047359713304e-06,
+      "loss": 0.5683,
+      "step": 3708
+    },
+    {
+      "epoch": 1.4658928130402569,
+      "grad_norm": 0.4504494694062807,
+      "learning_rate": 4.9486731566561416e-06,
+      "loss": 0.5597,
+      "step": 3709
+    },
+    {
+      "epoch": 1.466287972338849,
+      "grad_norm": 0.46237624966282265,
+      "learning_rate": 4.948641567724053e-06,
+      "loss": 0.5587,
+      "step": 3710
+    },
+    {
+      "epoch": 1.4666831316374414,
+      "grad_norm": 0.4412338330627608,
+      "learning_rate": 4.948609969175186e-06,
+      "loss": 0.58,
+      "step": 3711
+    },
+    {
+      "epoch": 1.4670782909360336,
+      "grad_norm": 0.46261210643628975,
+      "learning_rate": 4.9485783610096664e-06,
+      "loss": 0.5919,
+      "step": 3712
+    },
+    {
+      "epoch": 1.4674734502346258,
+      "grad_norm": 0.4570667260557455,
+      "learning_rate": 4.948546743227617e-06,
+      "loss": 0.5623,
+      "step": 3713
+    },
+    {
+      "epoch": 1.467868609533218,
+      "grad_norm": 0.43799780214279865,
+      "learning_rate": 4.948515115829164e-06,
+      "loss": 0.5526,
+      "step": 3714
+    },
+    {
+      "epoch": 1.4682637688318103,
+      "grad_norm": 0.4867222574376945,
+      "learning_rate": 4.9484834788144295e-06,
+      "loss": 0.5734,
+      "step": 3715
+    },
+    {
+      "epoch": 1.4686589281304026,
+      "grad_norm": 0.46745491220808105,
+      "learning_rate": 4.948451832183539e-06,
+      "loss": 0.5864,
+      "step": 3716
+    },
+    {
+      "epoch": 1.4690540874289948,
+      "grad_norm": 0.45356535226498157,
+      "learning_rate": 4.948420175936618e-06,
+      "loss": 0.563,
+      "step": 3717
+    },
+    {
+      "epoch": 1.469449246727587,
+      "grad_norm": 0.46464161309085605,
+      "learning_rate": 4.9483885100737875e-06,
+      "loss": 0.5494,
+      "step": 3718
+    },
+    {
+      "epoch": 1.4698444060261793,
+      "grad_norm": 0.4296514326477871,
+      "learning_rate": 4.9483568345951735e-06,
+      "loss": 0.561,
+      "step": 3719
+    },
+    {
+      "epoch": 1.4702395653247715,
+      "grad_norm": 0.4449958968332145,
+      "learning_rate": 4.948325149500902e-06,
+      "loss": 0.5767,
+      "step": 3720
+    },
+    {
+      "epoch": 1.4706347246233638,
+      "grad_norm": 0.44304606813408115,
+      "learning_rate": 4.948293454791095e-06,
+      "loss": 0.5768,
+      "step": 3721
+    },
+    {
+      "epoch": 1.471029883921956,
+      "grad_norm": 0.4438521427352953,
+      "learning_rate": 4.948261750465878e-06,
+      "loss": 0.5467,
+      "step": 3722
+    },
+    {
+      "epoch": 1.4714250432205482,
+      "grad_norm": 0.4377012072556724,
+      "learning_rate": 4.948230036525375e-06,
+      "loss": 0.5613,
+      "step": 3723
+    },
+    {
+      "epoch": 1.4718202025191405,
+      "grad_norm": 0.43727189711397746,
+      "learning_rate": 4.948198312969712e-06,
+      "loss": 0.5428,
+      "step": 3724
+    },
+    {
+      "epoch": 1.4722153618177327,
+      "grad_norm": 0.48964762021273595,
+      "learning_rate": 4.948166579799013e-06,
+      "loss": 0.5568,
+      "step": 3725
+    },
+    {
+      "epoch": 1.472610521116325,
+      "grad_norm": 0.5323132325510923,
+      "learning_rate": 4.948134837013402e-06,
+      "loss": 0.5578,
+      "step": 3726
+    },
+    {
+      "epoch": 1.4730056804149172,
+      "grad_norm": 0.4398364066065741,
+      "learning_rate": 4.948103084613003e-06,
+      "loss": 0.5609,
+      "step": 3727
+    },
+    {
+      "epoch": 1.4734008397135094,
+      "grad_norm": 0.4310398347472198,
+      "learning_rate": 4.948071322597943e-06,
+      "loss": 0.5651,
+      "step": 3728
+    },
+    {
+      "epoch": 1.4737959990121017,
+      "grad_norm": 0.539562325199801,
+      "learning_rate": 4.948039550968345e-06,
+      "loss": 0.5635,
+      "step": 3729
+    },
+    {
+      "epoch": 1.474191158310694,
+      "grad_norm": 0.4463656077676281,
+      "learning_rate": 4.948007769724333e-06,
+      "loss": 0.55,
+      "step": 3730
+    },
+    {
+      "epoch": 1.4745863176092864,
+      "grad_norm": 0.44864795312044087,
+      "learning_rate": 4.947975978866034e-06,
+      "loss": 0.5385,
+      "step": 3731
+    },
+    {
+      "epoch": 1.4749814769078786,
+      "grad_norm": 3.1516964419751026,
+      "learning_rate": 4.947944178393572e-06,
+      "loss": 0.582,
+      "step": 3732
+    },
+    {
+      "epoch": 1.4753766362064709,
+      "grad_norm": 0.46934564073210333,
+      "learning_rate": 4.947912368307071e-06,
+      "loss": 0.555,
+      "step": 3733
+    },
+    {
+      "epoch": 1.475771795505063,
+      "grad_norm": 0.46254224035417973,
+      "learning_rate": 4.9478805486066575e-06,
+      "loss": 0.5879,
+      "step": 3734
+    },
+    {
+      "epoch": 1.4761669548036553,
+      "grad_norm": 0.4469674187179735,
+      "learning_rate": 4.947848719292455e-06,
+      "loss": 0.554,
+      "step": 3735
+    },
+    {
+      "epoch": 1.4765621141022476,
+      "grad_norm": 0.45010349935865934,
+      "learning_rate": 4.947816880364589e-06,
+      "loss": 0.5465,
+      "step": 3736
+    },
+    {
+      "epoch": 1.4769572734008398,
+      "grad_norm": 0.4671526430772705,
+      "learning_rate": 4.9477850318231855e-06,
+      "loss": 0.5768,
+      "step": 3737
+    },
+    {
+      "epoch": 1.477352432699432,
+      "grad_norm": 0.44222712097266215,
+      "learning_rate": 4.947753173668368e-06,
+      "loss": 0.5647,
+      "step": 3738
+    },
+    {
+      "epoch": 1.4777475919980243,
+      "grad_norm": 0.4346426270801945,
+      "learning_rate": 4.947721305900263e-06,
+      "loss": 0.5883,
+      "step": 3739
+    },
+    {
+      "epoch": 1.4781427512966165,
+      "grad_norm": 0.4428949633974912,
+      "learning_rate": 4.947689428518994e-06,
+      "loss": 0.5585,
+      "step": 3740
+    },
+    {
+      "epoch": 1.4785379105952088,
+      "grad_norm": 0.45290050689849504,
+      "learning_rate": 4.947657541524689e-06,
+      "loss": 0.5698,
+      "step": 3741
+    },
+    {
+      "epoch": 1.478933069893801,
+      "grad_norm": 0.46047272522893107,
+      "learning_rate": 4.947625644917471e-06,
+      "loss": 0.5853,
+      "step": 3742
+    },
+    {
+      "epoch": 1.4793282291923933,
+      "grad_norm": 0.436358562052887,
+      "learning_rate": 4.9475937386974645e-06,
+      "loss": 0.5556,
+      "step": 3743
+    },
+    {
+      "epoch": 1.4797233884909855,
+      "grad_norm": 0.44605536186917205,
+      "learning_rate": 4.947561822864797e-06,
+      "loss": 0.5691,
+      "step": 3744
+    },
+    {
+      "epoch": 1.4801185477895777,
+      "grad_norm": 0.45340872193909043,
+      "learning_rate": 4.947529897419593e-06,
+      "loss": 0.5595,
+      "step": 3745
+    },
+    {
+      "epoch": 1.48051370708817,
+      "grad_norm": 0.44732431526653443,
+      "learning_rate": 4.947497962361977e-06,
+      "loss": 0.5469,
+      "step": 3746
+    },
+    {
+      "epoch": 1.4809088663867622,
+      "grad_norm": 0.43950554191375485,
+      "learning_rate": 4.947466017692075e-06,
+      "loss": 0.5493,
+      "step": 3747
+    },
+    {
+      "epoch": 1.4813040256853545,
+      "grad_norm": 0.451698847161343,
+      "learning_rate": 4.947434063410014e-06,
+      "loss": 0.5763,
+      "step": 3748
+    },
+    {
+      "epoch": 1.4816991849839467,
+      "grad_norm": 0.4383967173667518,
+      "learning_rate": 4.947402099515918e-06,
+      "loss": 0.5574,
+      "step": 3749
+    },
+    {
+      "epoch": 1.482094344282539,
+      "grad_norm": 0.46921309981868,
+      "learning_rate": 4.947370126009912e-06,
+      "loss": 0.5698,
+      "step": 3750
+    },
+    {
+      "epoch": 1.4824895035811312,
+      "grad_norm": 0.4576340583848198,
+      "learning_rate": 4.947338142892123e-06,
+      "loss": 0.5555,
+      "step": 3751
+    },
+    {
+      "epoch": 1.4828846628797234,
+      "grad_norm": 0.45431167387415017,
+      "learning_rate": 4.947306150162675e-06,
+      "loss": 0.5584,
+      "step": 3752
+    },
+    {
+      "epoch": 1.4832798221783157,
+      "grad_norm": 0.44151975322514186,
+      "learning_rate": 4.947274147821694e-06,
+      "loss": 0.5395,
+      "step": 3753
+    },
+    {
+      "epoch": 1.483674981476908,
+      "grad_norm": 0.4508179501579237,
+      "learning_rate": 4.947242135869308e-06,
+      "loss": 0.5629,
+      "step": 3754
+    },
+    {
+      "epoch": 1.4840701407755001,
+      "grad_norm": 0.44449630571041165,
+      "learning_rate": 4.94721011430564e-06,
+      "loss": 0.5543,
+      "step": 3755
+    },
+    {
+      "epoch": 1.4844653000740924,
+      "grad_norm": 0.44088055023127704,
+      "learning_rate": 4.947178083130817e-06,
+      "loss": 0.5525,
+      "step": 3756
+    },
+    {
+      "epoch": 1.4848604593726846,
+      "grad_norm": 0.4349780866471166,
+      "learning_rate": 4.947146042344964e-06,
+      "loss": 0.5431,
+      "step": 3757
+    },
+    {
+      "epoch": 1.4852556186712769,
+      "grad_norm": 0.4690057952385603,
+      "learning_rate": 4.947113991948207e-06,
+      "loss": 0.609,
+      "step": 3758
+    },
+    {
+      "epoch": 1.485650777969869,
+      "grad_norm": 0.4570722839955001,
+      "learning_rate": 4.947081931940673e-06,
+      "loss": 0.5589,
+      "step": 3759
+    },
+    {
+      "epoch": 1.4860459372684613,
+      "grad_norm": 0.44584022143653507,
+      "learning_rate": 4.9470498623224875e-06,
+      "loss": 0.5509,
+      "step": 3760
+    },
+    {
+      "epoch": 1.4864410965670536,
+      "grad_norm": 0.44483253459798316,
+      "learning_rate": 4.947017783093775e-06,
+      "loss": 0.5671,
+      "step": 3761
+    },
+    {
+      "epoch": 1.4868362558656458,
+      "grad_norm": 0.4637492181059857,
+      "learning_rate": 4.946985694254662e-06,
+      "loss": 0.5649,
+      "step": 3762
+    },
+    {
+      "epoch": 1.487231415164238,
+      "grad_norm": 0.4443774539885124,
+      "learning_rate": 4.946953595805277e-06,
+      "loss": 0.5664,
+      "step": 3763
+    },
+    {
+      "epoch": 1.4876265744628303,
+      "grad_norm": 0.4371331634851273,
+      "learning_rate": 4.946921487745743e-06,
+      "loss": 0.5585,
+      "step": 3764
+    },
+    {
+      "epoch": 1.4880217337614225,
+      "grad_norm": 0.44138169702671043,
+      "learning_rate": 4.9468893700761874e-06,
+      "loss": 0.5576,
+      "step": 3765
+    },
+    {
+      "epoch": 1.4884168930600148,
+      "grad_norm": 0.4525359288431999,
+      "learning_rate": 4.946857242796737e-06,
+      "loss": 0.5688,
+      "step": 3766
+    },
+    {
+      "epoch": 1.488812052358607,
+      "grad_norm": 0.4488955169238718,
+      "learning_rate": 4.946825105907516e-06,
+      "loss": 0.5708,
+      "step": 3767
+    },
+    {
+      "epoch": 1.4892072116571993,
+      "grad_norm": 0.453722892192494,
+      "learning_rate": 4.946792959408652e-06,
+      "loss": 0.574,
+      "step": 3768
+    },
+    {
+      "epoch": 1.4896023709557915,
+      "grad_norm": 0.4438539782138186,
+      "learning_rate": 4.9467608033002715e-06,
+      "loss": 0.5736,
+      "step": 3769
+    },
+    {
+      "epoch": 1.4899975302543838,
+      "grad_norm": 0.4734446349651656,
+      "learning_rate": 4.9467286375824995e-06,
+      "loss": 0.5587,
+      "step": 3770
+    },
+    {
+      "epoch": 1.490392689552976,
+      "grad_norm": 0.45483078234475843,
+      "learning_rate": 4.946696462255464e-06,
+      "loss": 0.5843,
+      "step": 3771
+    },
+    {
+      "epoch": 1.4907878488515682,
+      "grad_norm": 0.43491380893057063,
+      "learning_rate": 4.94666427731929e-06,
+      "loss": 0.5568,
+      "step": 3772
+    },
+    {
+      "epoch": 1.4911830081501605,
+      "grad_norm": 0.4593799609237185,
+      "learning_rate": 4.946632082774105e-06,
+      "loss": 0.568,
+      "step": 3773
+    },
+    {
+      "epoch": 1.4915781674487527,
+      "grad_norm": 0.4756685082546809,
+      "learning_rate": 4.946599878620034e-06,
+      "loss": 0.5525,
+      "step": 3774
+    },
+    {
+      "epoch": 1.491973326747345,
+      "grad_norm": 0.4375243311510868,
+      "learning_rate": 4.946567664857205e-06,
+      "loss": 0.5688,
+      "step": 3775
+    },
+    {
+      "epoch": 1.4923684860459372,
+      "grad_norm": 0.44023853517164957,
+      "learning_rate": 4.946535441485744e-06,
+      "loss": 0.5461,
+      "step": 3776
+    },
+    {
+      "epoch": 1.4927636453445294,
+      "grad_norm": 0.4417756181030543,
+      "learning_rate": 4.946503208505776e-06,
+      "loss": 0.5483,
+      "step": 3777
+    },
+    {
+      "epoch": 1.4931588046431217,
+      "grad_norm": 0.4459801582543887,
+      "learning_rate": 4.94647096591743e-06,
+      "loss": 0.5494,
+      "step": 3778
+    },
+    {
+      "epoch": 1.493553963941714,
+      "grad_norm": 0.44654308211555016,
+      "learning_rate": 4.9464387137208326e-06,
+      "loss": 0.5711,
+      "step": 3779
+    },
+    {
+      "epoch": 1.4939491232403062,
+      "grad_norm": 0.44171711692198834,
+      "learning_rate": 4.946406451916108e-06,
+      "loss": 0.5607,
+      "step": 3780
+    },
+    {
+      "epoch": 1.4943442825388984,
+      "grad_norm": 0.4390434383203796,
+      "learning_rate": 4.946374180503385e-06,
+      "loss": 0.5462,
+      "step": 3781
+    },
+    {
+      "epoch": 1.4947394418374906,
+      "grad_norm": 0.4548707333166944,
+      "learning_rate": 4.94634189948279e-06,
+      "loss": 0.554,
+      "step": 3782
+    },
+    {
+      "epoch": 1.4951346011360829,
+      "grad_norm": 0.4499388348643594,
+      "learning_rate": 4.946309608854449e-06,
+      "loss": 0.5705,
+      "step": 3783
+    },
+    {
+      "epoch": 1.4955297604346751,
+      "grad_norm": 0.4451819872634511,
+      "learning_rate": 4.94627730861849e-06,
+      "loss": 0.5653,
+      "step": 3784
+    },
+    {
+      "epoch": 1.4959249197332674,
+      "grad_norm": 0.4429157766074593,
+      "learning_rate": 4.946244998775039e-06,
+      "loss": 0.5591,
+      "step": 3785
+    },
+    {
+      "epoch": 1.4963200790318596,
+      "grad_norm": 0.45149504764423787,
+      "learning_rate": 4.946212679324222e-06,
+      "loss": 0.5531,
+      "step": 3786
+    },
+    {
+      "epoch": 1.4967152383304518,
+      "grad_norm": 0.45311480641045254,
+      "learning_rate": 4.946180350266168e-06,
+      "loss": 0.5738,
+      "step": 3787
+    },
+    {
+      "epoch": 1.497110397629044,
+      "grad_norm": 0.5577932771269037,
+      "learning_rate": 4.946148011601003e-06,
+      "loss": 0.5918,
+      "step": 3788
+    },
+    {
+      "epoch": 1.4975055569276363,
+      "grad_norm": 0.4442006706613438,
+      "learning_rate": 4.9461156633288535e-06,
+      "loss": 0.5573,
+      "step": 3789
+    },
+    {
+      "epoch": 1.4979007162262288,
+      "grad_norm": 0.448842435666703,
+      "learning_rate": 4.946083305449847e-06,
+      "loss": 0.5694,
+      "step": 3790
+    },
+    {
+      "epoch": 1.498295875524821,
+      "grad_norm": 0.47063189404977285,
+      "learning_rate": 4.946050937964112e-06,
+      "loss": 0.5663,
+      "step": 3791
+    },
+    {
+      "epoch": 1.4986910348234133,
+      "grad_norm": 0.4222521912580929,
+      "learning_rate": 4.946018560871772e-06,
+      "loss": 0.5333,
+      "step": 3792
+    },
+    {
+      "epoch": 1.4990861941220055,
+      "grad_norm": 0.708119310733299,
+      "learning_rate": 4.945986174172958e-06,
+      "loss": 0.5816,
+      "step": 3793
+    },
+    {
+      "epoch": 1.4994813534205977,
+      "grad_norm": 0.44863172461117107,
+      "learning_rate": 4.9459537778677955e-06,
+      "loss": 0.5848,
+      "step": 3794
+    },
+    {
+      "epoch": 1.49987651271919,
+      "grad_norm": 0.4411602360525524,
+      "learning_rate": 4.945921371956411e-06,
+      "loss": 0.5733,
+      "step": 3795
+    },
+    {
+      "epoch": 1.5002716720177822,
+      "grad_norm": 0.433024812052623,
+      "learning_rate": 4.945888956438933e-06,
+      "loss": 0.5644,
+      "step": 3796
+    },
+    {
+      "epoch": 1.5006668313163745,
+      "grad_norm": 0.5014185664609002,
+      "learning_rate": 4.945856531315489e-06,
+      "loss": 0.564,
+      "step": 3797
+    },
+    {
+      "epoch": 1.5010619906149667,
+      "grad_norm": 0.46228942218632185,
+      "learning_rate": 4.945824096586205e-06,
+      "loss": 0.569,
+      "step": 3798
+    },
+    {
+      "epoch": 1.501457149913559,
+      "grad_norm": 0.4623049423084713,
+      "learning_rate": 4.94579165225121e-06,
+      "loss": 0.5614,
+      "step": 3799
+    },
+    {
+      "epoch": 1.5018523092121512,
+      "grad_norm": 0.4294050033616448,
+      "learning_rate": 4.945759198310629e-06,
+      "loss": 0.5661,
+      "step": 3800
+    },
+    {
+      "epoch": 1.5022474685107434,
+      "grad_norm": 0.4383701048448765,
+      "learning_rate": 4.945726734764592e-06,
+      "loss": 0.5541,
+      "step": 3801
+    },
+    {
+      "epoch": 1.5026426278093357,
+      "grad_norm": 0.43965998205136625,
+      "learning_rate": 4.945694261613225e-06,
+      "loss": 0.5503,
+      "step": 3802
+    },
+    {
+      "epoch": 1.503037787107928,
+      "grad_norm": 0.43862817298801543,
+      "learning_rate": 4.945661778856658e-06,
+      "loss": 0.5793,
+      "step": 3803
+    },
+    {
+      "epoch": 1.5034329464065201,
+      "grad_norm": 0.45968905394437315,
+      "learning_rate": 4.945629286495014e-06,
+      "loss": 0.5836,
+      "step": 3804
+    },
+    {
+      "epoch": 1.5038281057051124,
+      "grad_norm": 0.43146289605158555,
+      "learning_rate": 4.945596784528425e-06,
+      "loss": 0.5553,
+      "step": 3805
+    },
+    {
+      "epoch": 1.5042232650037046,
+      "grad_norm": 0.4227325494043722,
+      "learning_rate": 4.945564272957016e-06,
+      "loss": 0.5467,
+      "step": 3806
+    },
+    {
+      "epoch": 1.5046184243022969,
+      "grad_norm": 0.45665440959944387,
+      "learning_rate": 4.945531751780915e-06,
+      "loss": 0.5825,
+      "step": 3807
+    },
+    {
+      "epoch": 1.505013583600889,
+      "grad_norm": 0.4370355896617275,
+      "learning_rate": 4.9454992210002515e-06,
+      "loss": 0.5757,
+      "step": 3808
+    },
+    {
+      "epoch": 1.5054087428994813,
+      "grad_norm": 0.44183906572444237,
+      "learning_rate": 4.9454666806151515e-06,
+      "loss": 0.5619,
+      "step": 3809
+    },
+    {
+      "epoch": 1.5058039021980736,
+      "grad_norm": 0.4214971459805865,
+      "learning_rate": 4.945434130625744e-06,
+      "loss": 0.5638,
+      "step": 3810
+    },
+    {
+      "epoch": 1.5061990614966658,
+      "grad_norm": 0.43677389141426,
+      "learning_rate": 4.945401571032156e-06,
+      "loss": 0.5626,
+      "step": 3811
+    },
+    {
+      "epoch": 1.506594220795258,
+      "grad_norm": 0.43865734507834114,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.5353,
+      "step": 3812
+    },
+    {
+      "epoch": 1.5069893800938503,
+      "grad_norm": 0.5342126887725891,
+      "learning_rate": 4.945336423032949e-06,
+      "loss": 0.5695,
+      "step": 3813
+    },
+    {
+      "epoch": 1.5073845393924425,
+      "grad_norm": 0.4349734124143589,
+      "learning_rate": 4.945303834627587e-06,
+      "loss": 0.5418,
+      "step": 3814
+    },
+    {
+      "epoch": 1.5077796986910348,
+      "grad_norm": 0.4385647900898778,
+      "learning_rate": 4.945271236618557e-06,
+      "loss": 0.5658,
+      "step": 3815
+    },
+    {
+      "epoch": 1.508174857989627,
+      "grad_norm": 0.42971941105827866,
+      "learning_rate": 4.945238629005986e-06,
+      "loss": 0.5727,
+      "step": 3816
+    },
+    {
+      "epoch": 1.5085700172882193,
+      "grad_norm": 0.4454445834803687,
+      "learning_rate": 4.945206011790002e-06,
+      "loss": 0.582,
+      "step": 3817
+    },
+    {
+      "epoch": 1.5089651765868115,
+      "grad_norm": 0.4534020504734814,
+      "learning_rate": 4.945173384970734e-06,
+      "loss": 0.5659,
+      "step": 3818
+    },
+    {
+      "epoch": 1.509360335885404,
+      "grad_norm": 0.4615252795337526,
+      "learning_rate": 4.945140748548309e-06,
+      "loss": 0.5762,
+      "step": 3819
+    },
+    {
+      "epoch": 1.5097554951839962,
+      "grad_norm": 0.43867764806826415,
+      "learning_rate": 4.945108102522858e-06,
+      "loss": 0.5667,
+      "step": 3820
+    },
+    {
+      "epoch": 1.5101506544825885,
+      "grad_norm": 0.43752103633853423,
+      "learning_rate": 4.945075446894505e-06,
+      "loss": 0.5648,
+      "step": 3821
+    },
+    {
+      "epoch": 1.5105458137811807,
+      "grad_norm": 0.48724137180843774,
+      "learning_rate": 4.945042781663381e-06,
+      "loss": 0.5496,
+      "step": 3822
+    },
+    {
+      "epoch": 1.510940973079773,
+      "grad_norm": 0.4386933082638948,
+      "learning_rate": 4.945010106829614e-06,
+      "loss": 0.5613,
+      "step": 3823
+    },
+    {
+      "epoch": 1.5113361323783652,
+      "grad_norm": 0.5117191506598214,
+      "learning_rate": 4.944977422393332e-06,
+      "loss": 0.5537,
+      "step": 3824
+    },
+    {
+      "epoch": 1.5117312916769574,
+      "grad_norm": 0.43364355513059377,
+      "learning_rate": 4.944944728354663e-06,
+      "loss": 0.5655,
+      "step": 3825
+    },
+    {
+      "epoch": 1.5121264509755497,
+      "grad_norm": 0.4467784125266709,
+      "learning_rate": 4.9449120247137365e-06,
+      "loss": 0.5664,
+      "step": 3826
+    },
+    {
+      "epoch": 1.512521610274142,
+      "grad_norm": 0.5300397203257904,
+      "learning_rate": 4.944879311470679e-06,
+      "loss": 0.5769,
+      "step": 3827
+    },
+    {
+      "epoch": 1.5129167695727341,
+      "grad_norm": 0.4257934452475102,
+      "learning_rate": 4.944846588625621e-06,
+      "loss": 0.5436,
+      "step": 3828
+    },
+    {
+      "epoch": 1.5133119288713264,
+      "grad_norm": 0.4455795318956715,
+      "learning_rate": 4.94481385617869e-06,
+      "loss": 0.5385,
+      "step": 3829
+    },
+    {
+      "epoch": 1.5137070881699186,
+      "grad_norm": 0.45355241250172695,
+      "learning_rate": 4.944781114130015e-06,
+      "loss": 0.5721,
+      "step": 3830
+    },
+    {
+      "epoch": 1.5141022474685109,
+      "grad_norm": 0.44014963072857977,
+      "learning_rate": 4.944748362479723e-06,
+      "loss": 0.562,
+      "step": 3831
+    },
+    {
+      "epoch": 1.514497406767103,
+      "grad_norm": 0.42555153523264766,
+      "learning_rate": 4.9447156012279455e-06,
+      "loss": 0.5624,
+      "step": 3832
+    },
+    {
+      "epoch": 1.5148925660656953,
+      "grad_norm": 0.48013397168835886,
+      "learning_rate": 4.944682830374809e-06,
+      "loss": 0.5513,
+      "step": 3833
+    },
+    {
+      "epoch": 1.5152877253642876,
+      "grad_norm": 0.45697256603832853,
+      "learning_rate": 4.944650049920443e-06,
+      "loss": 0.5629,
+      "step": 3834
+    },
+    {
+      "epoch": 1.5156828846628798,
+      "grad_norm": 0.44411668965768414,
+      "learning_rate": 4.944617259864976e-06,
+      "loss": 0.5506,
+      "step": 3835
+    },
+    {
+      "epoch": 1.516078043961472,
+      "grad_norm": 0.4416796017067383,
+      "learning_rate": 4.944584460208537e-06,
+      "loss": 0.5536,
+      "step": 3836
+    },
+    {
+      "epoch": 1.5164732032600643,
+      "grad_norm": 0.4349355230140983,
+      "learning_rate": 4.944551650951255e-06,
+      "loss": 0.5599,
+      "step": 3837
+    },
+    {
+      "epoch": 1.5168683625586565,
+      "grad_norm": 0.42051333332637836,
+      "learning_rate": 4.944518832093258e-06,
+      "loss": 0.549,
+      "step": 3838
+    },
+    {
+      "epoch": 1.5172635218572488,
+      "grad_norm": 0.43470424455730705,
+      "learning_rate": 4.944486003634675e-06,
+      "loss": 0.5705,
+      "step": 3839
+    },
+    {
+      "epoch": 1.517658681155841,
+      "grad_norm": 0.4484568492506127,
+      "learning_rate": 4.944453165575635e-06,
+      "loss": 0.5527,
+      "step": 3840
+    },
+    {
+      "epoch": 1.5180538404544333,
+      "grad_norm": 0.43871369779576824,
+      "learning_rate": 4.944420317916269e-06,
+      "loss": 0.5572,
+      "step": 3841
+    },
+    {
+      "epoch": 1.5184489997530255,
+      "grad_norm": 0.4267428186392819,
+      "learning_rate": 4.944387460656703e-06,
+      "loss": 0.5602,
+      "step": 3842
+    },
+    {
+      "epoch": 1.5188441590516177,
+      "grad_norm": 0.43416355081762653,
+      "learning_rate": 4.9443545937970686e-06,
+      "loss": 0.567,
+      "step": 3843
+    },
+    {
+      "epoch": 1.51923931835021,
+      "grad_norm": 0.49590515295094667,
+      "learning_rate": 4.944321717337493e-06,
+      "loss": 0.5598,
+      "step": 3844
+    },
+    {
+      "epoch": 1.5196344776488022,
+      "grad_norm": 0.4626680777777951,
+      "learning_rate": 4.9442888312781056e-06,
+      "loss": 0.5583,
+      "step": 3845
+    },
+    {
+      "epoch": 1.5200296369473945,
+      "grad_norm": 0.4449293851147631,
+      "learning_rate": 4.944255935619036e-06,
+      "loss": 0.562,
+      "step": 3846
+    },
+    {
+      "epoch": 1.5204247962459867,
+      "grad_norm": 0.43950123990367407,
+      "learning_rate": 4.944223030360414e-06,
+      "loss": 0.5642,
+      "step": 3847
+    },
+    {
+      "epoch": 1.520819955544579,
+      "grad_norm": 0.44149298105890766,
+      "learning_rate": 4.9441901155023675e-06,
+      "loss": 0.5796,
+      "step": 3848
+    },
+    {
+      "epoch": 1.5212151148431712,
+      "grad_norm": 0.4482301026578816,
+      "learning_rate": 4.944157191045027e-06,
+      "loss": 0.5688,
+      "step": 3849
+    },
+    {
+      "epoch": 1.5216102741417634,
+      "grad_norm": 0.5488343962637722,
+      "learning_rate": 4.94412425698852e-06,
+      "loss": 0.5546,
+      "step": 3850
+    },
+    {
+      "epoch": 1.5220054334403557,
+      "grad_norm": 0.44100420516713945,
+      "learning_rate": 4.944091313332978e-06,
+      "loss": 0.5624,
+      "step": 3851
+    },
+    {
+      "epoch": 1.522400592738948,
+      "grad_norm": 0.42706076298730056,
+      "learning_rate": 4.94405836007853e-06,
+      "loss": 0.5365,
+      "step": 3852
+    },
+    {
+      "epoch": 1.5227957520375401,
+      "grad_norm": 0.4359259767312831,
+      "learning_rate": 4.944025397225304e-06,
+      "loss": 0.5411,
+      "step": 3853
+    },
+    {
+      "epoch": 1.5231909113361324,
+      "grad_norm": 0.4388004674269903,
+      "learning_rate": 4.943992424773431e-06,
+      "loss": 0.5655,
+      "step": 3854
+    },
+    {
+      "epoch": 1.5235860706347246,
+      "grad_norm": 0.4747569771286888,
+      "learning_rate": 4.943959442723039e-06,
+      "loss": 0.5583,
+      "step": 3855
+    },
+    {
+      "epoch": 1.5239812299333169,
+      "grad_norm": 0.4673031379896023,
+      "learning_rate": 4.943926451074258e-06,
+      "loss": 0.5663,
+      "step": 3856
+    },
+    {
+      "epoch": 1.524376389231909,
+      "grad_norm": 0.45989463226172245,
+      "learning_rate": 4.943893449827219e-06,
+      "loss": 0.554,
+      "step": 3857
+    },
+    {
+      "epoch": 1.5247715485305013,
+      "grad_norm": 0.4488593649799981,
+      "learning_rate": 4.94386043898205e-06,
+      "loss": 0.5733,
+      "step": 3858
+    },
+    {
+      "epoch": 1.5251667078290936,
+      "grad_norm": 0.4485345293642178,
+      "learning_rate": 4.943827418538882e-06,
+      "loss": 0.5603,
+      "step": 3859
+    },
+    {
+      "epoch": 1.5255618671276858,
+      "grad_norm": 0.4535620010441292,
+      "learning_rate": 4.943794388497842e-06,
+      "loss": 0.5548,
+      "step": 3860
+    },
+    {
+      "epoch": 1.525957026426278,
+      "grad_norm": 0.4735064699244355,
+      "learning_rate": 4.943761348859063e-06,
+      "loss": 0.5524,
+      "step": 3861
+    },
+    {
+      "epoch": 1.5263521857248703,
+      "grad_norm": 0.4513665077962669,
+      "learning_rate": 4.9437282996226734e-06,
+      "loss": 0.5677,
+      "step": 3862
+    },
+    {
+      "epoch": 1.5267473450234625,
+      "grad_norm": 0.4556372954566632,
+      "learning_rate": 4.943695240788803e-06,
+      "loss": 0.559,
+      "step": 3863
+    },
+    {
+      "epoch": 1.5271425043220548,
+      "grad_norm": 0.4597655079944917,
+      "learning_rate": 4.943662172357582e-06,
+      "loss": 0.5529,
+      "step": 3864
+    },
+    {
+      "epoch": 1.527537663620647,
+      "grad_norm": 0.45852506910423174,
+      "learning_rate": 4.943629094329139e-06,
+      "loss": 0.556,
+      "step": 3865
+    },
+    {
+      "epoch": 1.5279328229192393,
+      "grad_norm": 0.4475865214143378,
+      "learning_rate": 4.9435960067036045e-06,
+      "loss": 0.5638,
+      "step": 3866
+    },
+    {
+      "epoch": 1.5283279822178315,
+      "grad_norm": 0.44719416313388277,
+      "learning_rate": 4.943562909481109e-06,
+      "loss": 0.5652,
+      "step": 3867
+    },
+    {
+      "epoch": 1.5287231415164237,
+      "grad_norm": 0.4288620192017668,
+      "learning_rate": 4.943529802661783e-06,
+      "loss": 0.5436,
+      "step": 3868
+    },
+    {
+      "epoch": 1.529118300815016,
+      "grad_norm": 0.45692475861207277,
+      "learning_rate": 4.943496686245754e-06,
+      "loss": 0.5723,
+      "step": 3869
+    },
+    {
+      "epoch": 1.5295134601136082,
+      "grad_norm": 0.43936779754731264,
+      "learning_rate": 4.943463560233155e-06,
+      "loss": 0.5631,
+      "step": 3870
+    },
+    {
+      "epoch": 1.5299086194122005,
+      "grad_norm": 0.45634250616291433,
+      "learning_rate": 4.943430424624115e-06,
+      "loss": 0.5876,
+      "step": 3871
+    },
+    {
+      "epoch": 1.5303037787107927,
+      "grad_norm": 0.46250472654805214,
+      "learning_rate": 4.943397279418764e-06,
+      "loss": 0.5704,
+      "step": 3872
+    },
+    {
+      "epoch": 1.530698938009385,
+      "grad_norm": 0.4445910904786816,
+      "learning_rate": 4.943364124617232e-06,
+      "loss": 0.5762,
+      "step": 3873
+    },
+    {
+      "epoch": 1.5310940973079772,
+      "grad_norm": 0.4492442500667238,
+      "learning_rate": 4.9433309602196494e-06,
+      "loss": 0.5812,
+      "step": 3874
+    },
+    {
+      "epoch": 1.5314892566065694,
+      "grad_norm": 0.4375970310778139,
+      "learning_rate": 4.943297786226147e-06,
+      "loss": 0.5673,
+      "step": 3875
+    },
+    {
+      "epoch": 1.5318844159051617,
+      "grad_norm": 0.43961087035282753,
+      "learning_rate": 4.9432646026368535e-06,
+      "loss": 0.5671,
+      "step": 3876
+    },
+    {
+      "epoch": 1.532279575203754,
+      "grad_norm": 0.4415868685793641,
+      "learning_rate": 4.943231409451901e-06,
+      "loss": 0.5559,
+      "step": 3877
+    },
+    {
+      "epoch": 1.5326747345023461,
+      "grad_norm": 0.43726516852584973,
+      "learning_rate": 4.943198206671419e-06,
+      "loss": 0.5684,
+      "step": 3878
+    },
+    {
+      "epoch": 1.5330698938009384,
+      "grad_norm": 0.5804283001744497,
+      "learning_rate": 4.943164994295538e-06,
+      "loss": 0.5731,
+      "step": 3879
+    },
+    {
+      "epoch": 1.5334650530995306,
+      "grad_norm": 0.45583485277725394,
+      "learning_rate": 4.943131772324388e-06,
+      "loss": 0.5572,
+      "step": 3880
+    },
+    {
+      "epoch": 1.5338602123981229,
+      "grad_norm": 0.4399328351020496,
+      "learning_rate": 4.9430985407581e-06,
+      "loss": 0.5427,
+      "step": 3881
+    },
+    {
+      "epoch": 1.534255371696715,
+      "grad_norm": 0.42834809808516533,
+      "learning_rate": 4.943065299596806e-06,
+      "loss": 0.5681,
+      "step": 3882
+    },
+    {
+      "epoch": 1.5346505309953073,
+      "grad_norm": 0.5826505399905085,
+      "learning_rate": 4.943032048840633e-06,
+      "loss": 0.5551,
+      "step": 3883
+    },
+    {
+      "epoch": 1.5350456902938996,
+      "grad_norm": 0.4471917315958813,
+      "learning_rate": 4.942998788489715e-06,
+      "loss": 0.5762,
+      "step": 3884
+    },
+    {
+      "epoch": 1.5354408495924918,
+      "grad_norm": 0.44668514989336283,
+      "learning_rate": 4.94296551854418e-06,
+      "loss": 0.5721,
+      "step": 3885
+    },
+    {
+      "epoch": 1.535836008891084,
+      "grad_norm": 0.4631760060913264,
+      "learning_rate": 4.942932239004161e-06,
+      "loss": 0.5684,
+      "step": 3886
+    },
+    {
+      "epoch": 1.5362311681896763,
+      "grad_norm": 0.45324498503225535,
+      "learning_rate": 4.942898949869787e-06,
+      "loss": 0.5749,
+      "step": 3887
+    },
+    {
+      "epoch": 1.5366263274882686,
+      "grad_norm": 0.43838680281308623,
+      "learning_rate": 4.942865651141189e-06,
+      "loss": 0.5723,
+      "step": 3888
+    },
+    {
+      "epoch": 1.5370214867868608,
+      "grad_norm": 0.44695717165338794,
+      "learning_rate": 4.942832342818499e-06,
+      "loss": 0.58,
+      "step": 3889
+    },
+    {
+      "epoch": 1.5374166460854533,
+      "grad_norm": 0.42767213191242315,
+      "learning_rate": 4.942799024901846e-06,
+      "loss": 0.5508,
+      "step": 3890
+    },
+    {
+      "epoch": 1.5378118053840455,
+      "grad_norm": 0.440495140617439,
+      "learning_rate": 4.942765697391363e-06,
+      "loss": 0.5735,
+      "step": 3891
+    },
+    {
+      "epoch": 1.5382069646826377,
+      "grad_norm": 0.4607800617717655,
+      "learning_rate": 4.942732360287179e-06,
+      "loss": 0.5874,
+      "step": 3892
+    },
+    {
+      "epoch": 1.53860212398123,
+      "grad_norm": 0.44999899303277263,
+      "learning_rate": 4.942699013589425e-06,
+      "loss": 0.5793,
+      "step": 3893
+    },
+    {
+      "epoch": 1.5389972832798222,
+      "grad_norm": 0.4558760646566193,
+      "learning_rate": 4.942665657298233e-06,
+      "loss": 0.5714,
+      "step": 3894
+    },
+    {
+      "epoch": 1.5393924425784145,
+      "grad_norm": 0.43776240719293746,
+      "learning_rate": 4.9426322914137335e-06,
+      "loss": 0.5676,
+      "step": 3895
+    },
+    {
+      "epoch": 1.5397876018770067,
+      "grad_norm": 0.4437159939338884,
+      "learning_rate": 4.942598915936058e-06,
+      "loss": 0.5893,
+      "step": 3896
+    },
+    {
+      "epoch": 1.540182761175599,
+      "grad_norm": 0.4374941926171507,
+      "learning_rate": 4.942565530865337e-06,
+      "loss": 0.555,
+      "step": 3897
+    },
+    {
+      "epoch": 1.5405779204741912,
+      "grad_norm": 0.4208573300497547,
+      "learning_rate": 4.942532136201702e-06,
+      "loss": 0.5565,
+      "step": 3898
+    },
+    {
+      "epoch": 1.5409730797727834,
+      "grad_norm": 0.4337013789375931,
+      "learning_rate": 4.942498731945283e-06,
+      "loss": 0.5616,
+      "step": 3899
+    },
+    {
+      "epoch": 1.5413682390713757,
+      "grad_norm": 0.4527938862232506,
+      "learning_rate": 4.942465318096213e-06,
+      "loss": 0.556,
+      "step": 3900
+    },
+    {
+      "epoch": 1.541763398369968,
+      "grad_norm": 0.4427938210330589,
+      "learning_rate": 4.942431894654622e-06,
+      "loss": 0.5716,
+      "step": 3901
+    },
+    {
+      "epoch": 1.5421585576685601,
+      "grad_norm": 0.439043234576677,
+      "learning_rate": 4.942398461620642e-06,
+      "loss": 0.5627,
+      "step": 3902
+    },
+    {
+      "epoch": 1.5425537169671524,
+      "grad_norm": 0.4556850317164177,
+      "learning_rate": 4.942365018994404e-06,
+      "loss": 0.564,
+      "step": 3903
+    },
+    {
+      "epoch": 1.5429488762657446,
+      "grad_norm": 0.5109127133225726,
+      "learning_rate": 4.942331566776039e-06,
+      "loss": 0.5574,
+      "step": 3904
+    },
+    {
+      "epoch": 1.5433440355643369,
+      "grad_norm": 0.4363940727408752,
+      "learning_rate": 4.942298104965679e-06,
+      "loss": 0.575,
+      "step": 3905
+    },
+    {
+      "epoch": 1.543739194862929,
+      "grad_norm": 0.42938184413018377,
+      "learning_rate": 4.942264633563455e-06,
+      "loss": 0.5562,
+      "step": 3906
+    },
+    {
+      "epoch": 1.5441343541615213,
+      "grad_norm": 0.44370460437623127,
+      "learning_rate": 4.942231152569499e-06,
+      "loss": 0.5753,
+      "step": 3907
+    },
+    {
+      "epoch": 1.5445295134601136,
+      "grad_norm": 0.4367799811992645,
+      "learning_rate": 4.94219766198394e-06,
+      "loss": 0.5531,
+      "step": 3908
+    },
+    {
+      "epoch": 1.5449246727587058,
+      "grad_norm": 0.436026557215798,
+      "learning_rate": 4.942164161806914e-06,
+      "loss": 0.5641,
+      "step": 3909
+    },
+    {
+      "epoch": 1.545319832057298,
+      "grad_norm": 0.43686930660689016,
+      "learning_rate": 4.94213065203855e-06,
+      "loss": 0.5514,
+      "step": 3910
+    },
+    {
+      "epoch": 1.5457149913558903,
+      "grad_norm": 0.459912181031538,
+      "learning_rate": 4.942097132678978e-06,
+      "loss": 0.5717,
+      "step": 3911
+    },
+    {
+      "epoch": 1.5461101506544825,
+      "grad_norm": 0.432296005037305,
+      "learning_rate": 4.942063603728332e-06,
+      "loss": 0.5598,
+      "step": 3912
+    },
+    {
+      "epoch": 1.5465053099530748,
+      "grad_norm": 0.44580566374151687,
+      "learning_rate": 4.942030065186744e-06,
+      "loss": 0.5649,
+      "step": 3913
+    },
+    {
+      "epoch": 1.5469004692516672,
+      "grad_norm": 0.4362823236216351,
+      "learning_rate": 4.941996517054344e-06,
+      "loss": 0.5658,
+      "step": 3914
+    },
+    {
+      "epoch": 1.5472956285502595,
+      "grad_norm": 0.42956539131805577,
+      "learning_rate": 4.941962959331265e-06,
+      "loss": 0.564,
+      "step": 3915
+    },
+    {
+      "epoch": 1.5476907878488517,
+      "grad_norm": 0.445691619032847,
+      "learning_rate": 4.941929392017637e-06,
+      "loss": 0.5615,
+      "step": 3916
+    },
+    {
+      "epoch": 1.548085947147444,
+      "grad_norm": 0.43714883646638897,
+      "learning_rate": 4.9418958151135946e-06,
+      "loss": 0.5746,
+      "step": 3917
+    },
+    {
+      "epoch": 1.5484811064460362,
+      "grad_norm": 0.4574932956702348,
+      "learning_rate": 4.941862228619267e-06,
+      "loss": 0.5527,
+      "step": 3918
+    },
+    {
+      "epoch": 1.5488762657446284,
+      "grad_norm": 0.4530888175518143,
+      "learning_rate": 4.941828632534789e-06,
+      "loss": 0.5569,
+      "step": 3919
+    },
+    {
+      "epoch": 1.5492714250432207,
+      "grad_norm": 0.44308964244366394,
+      "learning_rate": 4.94179502686029e-06,
+      "loss": 0.5578,
+      "step": 3920
+    },
+    {
+      "epoch": 1.549666584341813,
+      "grad_norm": 0.4395901315241771,
+      "learning_rate": 4.941761411595903e-06,
+      "loss": 0.5559,
+      "step": 3921
+    },
+    {
+      "epoch": 1.5500617436404052,
+      "grad_norm": 0.4506494793506838,
+      "learning_rate": 4.94172778674176e-06,
+      "loss": 0.5751,
+      "step": 3922
+    },
+    {
+      "epoch": 1.5504569029389974,
+      "grad_norm": 0.43273605402011445,
+      "learning_rate": 4.9416941522979926e-06,
+      "loss": 0.5476,
+      "step": 3923
+    },
+    {
+      "epoch": 1.5508520622375896,
+      "grad_norm": 0.43415805767715054,
+      "learning_rate": 4.9416605082647325e-06,
+      "loss": 0.5693,
+      "step": 3924
+    },
+    {
+      "epoch": 1.5512472215361819,
+      "grad_norm": 0.4434283494921018,
+      "learning_rate": 4.941626854642114e-06,
+      "loss": 0.546,
+      "step": 3925
+    },
+    {
+      "epoch": 1.5516423808347741,
+      "grad_norm": 0.443351882260677,
+      "learning_rate": 4.941593191430267e-06,
+      "loss": 0.5511,
+      "step": 3926
+    },
+    {
+      "epoch": 1.5520375401333664,
+      "grad_norm": 0.4371115950925369,
+      "learning_rate": 4.941559518629325e-06,
+      "loss": 0.5655,
+      "step": 3927
+    },
+    {
+      "epoch": 1.5524326994319586,
+      "grad_norm": 0.4500093849359018,
+      "learning_rate": 4.94152583623942e-06,
+      "loss": 0.5556,
+      "step": 3928
+    },
+    {
+      "epoch": 1.5528278587305508,
+      "grad_norm": 0.4353898954337321,
+      "learning_rate": 4.941492144260683e-06,
+      "loss": 0.5571,
+      "step": 3929
+    },
+    {
+      "epoch": 1.553223018029143,
+      "grad_norm": 0.45540296825899435,
+      "learning_rate": 4.941458442693249e-06,
+      "loss": 0.5649,
+      "step": 3930
+    },
+    {
+      "epoch": 1.5536181773277353,
+      "grad_norm": 0.43590100396307097,
+      "learning_rate": 4.9414247315372474e-06,
+      "loss": 0.5592,
+      "step": 3931
+    },
+    {
+      "epoch": 1.5540133366263276,
+      "grad_norm": 0.4273362760763734,
+      "learning_rate": 4.9413910107928136e-06,
+      "loss": 0.5559,
+      "step": 3932
+    },
+    {
+      "epoch": 1.5544084959249198,
+      "grad_norm": 0.43873328194495376,
+      "learning_rate": 4.941357280460076e-06,
+      "loss": 0.5707,
+      "step": 3933
+    },
+    {
+      "epoch": 1.554803655223512,
+      "grad_norm": 0.44105964043710244,
+      "learning_rate": 4.941323540539171e-06,
+      "loss": 0.5683,
+      "step": 3934
+    },
+    {
+      "epoch": 1.5551988145221043,
+      "grad_norm": 0.436934179465898,
+      "learning_rate": 4.941289791030229e-06,
+      "loss": 0.5757,
+      "step": 3935
+    },
+    {
+      "epoch": 1.5555939738206965,
+      "grad_norm": 0.45737730913945995,
+      "learning_rate": 4.9412560319333844e-06,
+      "loss": 0.5711,
+      "step": 3936
+    },
+    {
+      "epoch": 1.5559891331192888,
+      "grad_norm": 0.42468924058156193,
+      "learning_rate": 4.941222263248767e-06,
+      "loss": 0.5478,
+      "step": 3937
+    },
+    {
+      "epoch": 1.556384292417881,
+      "grad_norm": 0.43195983842147717,
+      "learning_rate": 4.941188484976512e-06,
+      "loss": 0.5659,
+      "step": 3938
+    },
+    {
+      "epoch": 1.5567794517164733,
+      "grad_norm": 0.44375257055854056,
+      "learning_rate": 4.9411546971167505e-06,
+      "loss": 0.5867,
+      "step": 3939
+    },
+    {
+      "epoch": 1.5571746110150655,
+      "grad_norm": 0.43710025950059056,
+      "learning_rate": 4.941120899669616e-06,
+      "loss": 0.5735,
+      "step": 3940
+    },
+    {
+      "epoch": 1.5575697703136577,
+      "grad_norm": 0.4474954937975031,
+      "learning_rate": 4.94108709263524e-06,
+      "loss": 0.602,
+      "step": 3941
+    },
+    {
+      "epoch": 1.55796492961225,
+      "grad_norm": 0.4305607613140482,
+      "learning_rate": 4.941053276013758e-06,
+      "loss": 0.5548,
+      "step": 3942
+    },
+    {
+      "epoch": 1.5583600889108422,
+      "grad_norm": 0.4398028265526488,
+      "learning_rate": 4.941019449805299e-06,
+      "loss": 0.5495,
+      "step": 3943
+    },
+    {
+      "epoch": 1.5587552482094345,
+      "grad_norm": 0.427010931690335,
+      "learning_rate": 4.940985614009999e-06,
+      "loss": 0.561,
+      "step": 3944
+    },
+    {
+      "epoch": 1.5591504075080267,
+      "grad_norm": 0.43014008416807015,
+      "learning_rate": 4.9409517686279895e-06,
+      "loss": 0.5796,
+      "step": 3945
+    },
+    {
+      "epoch": 1.559545566806619,
+      "grad_norm": 0.4473068429218742,
+      "learning_rate": 4.940917913659404e-06,
+      "loss": 0.5594,
+      "step": 3946
+    },
+    {
+      "epoch": 1.5599407261052112,
+      "grad_norm": 0.44348060278334916,
+      "learning_rate": 4.940884049104374e-06,
+      "loss": 0.5603,
+      "step": 3947
+    },
+    {
+      "epoch": 1.5603358854038034,
+      "grad_norm": 0.4254990006538986,
+      "learning_rate": 4.940850174963035e-06,
+      "loss": 0.5496,
+      "step": 3948
+    },
+    {
+      "epoch": 1.5607310447023957,
+      "grad_norm": 0.4599229925834786,
+      "learning_rate": 4.9408162912355185e-06,
+      "loss": 0.569,
+      "step": 3949
+    },
+    {
+      "epoch": 1.561126204000988,
+      "grad_norm": 0.4608183190992476,
+      "learning_rate": 4.940782397921957e-06,
+      "loss": 0.5475,
+      "step": 3950
+    },
+    {
+      "epoch": 1.5615213632995801,
+      "grad_norm": 0.4334203860785032,
+      "learning_rate": 4.940748495022485e-06,
+      "loss": 0.5617,
+      "step": 3951
+    },
+    {
+      "epoch": 1.5619165225981724,
+      "grad_norm": 0.4286064511326075,
+      "learning_rate": 4.940714582537235e-06,
+      "loss": 0.5508,
+      "step": 3952
+    },
+    {
+      "epoch": 1.5623116818967646,
+      "grad_norm": 0.4451771504215982,
+      "learning_rate": 4.94068066046634e-06,
+      "loss": 0.5679,
+      "step": 3953
+    },
+    {
+      "epoch": 1.5627068411953569,
+      "grad_norm": 0.44950544874820486,
+      "learning_rate": 4.940646728809933e-06,
+      "loss": 0.5571,
+      "step": 3954
+    },
+    {
+      "epoch": 1.563102000493949,
+      "grad_norm": 0.43332984605668107,
+      "learning_rate": 4.940612787568148e-06,
+      "loss": 0.5412,
+      "step": 3955
+    },
+    {
+      "epoch": 1.5634971597925413,
+      "grad_norm": 0.4537080751951872,
+      "learning_rate": 4.940578836741119e-06,
+      "loss": 0.5922,
+      "step": 3956
+    },
+    {
+      "epoch": 1.5638923190911336,
+      "grad_norm": 0.45065137229775004,
+      "learning_rate": 4.940544876328977e-06,
+      "loss": 0.5761,
+      "step": 3957
+    },
+    {
+      "epoch": 1.5642874783897258,
+      "grad_norm": 0.468459520127438,
+      "learning_rate": 4.940510906331856e-06,
+      "loss": 0.5766,
+      "step": 3958
+    },
+    {
+      "epoch": 1.564682637688318,
+      "grad_norm": 0.4841953450723878,
+      "learning_rate": 4.940476926749892e-06,
+      "loss": 0.5509,
+      "step": 3959
+    },
+    {
+      "epoch": 1.5650777969869103,
+      "grad_norm": 0.43421269689782305,
+      "learning_rate": 4.940442937583216e-06,
+      "loss": 0.5428,
+      "step": 3960
+    },
+    {
+      "epoch": 1.5654729562855025,
+      "grad_norm": 0.44740726473549824,
+      "learning_rate": 4.940408938831962e-06,
+      "loss": 0.5543,
+      "step": 3961
+    },
+    {
+      "epoch": 1.5658681155840948,
+      "grad_norm": 0.43972490377414003,
+      "learning_rate": 4.9403749304962635e-06,
+      "loss": 0.5711,
+      "step": 3962
+    },
+    {
+      "epoch": 1.566263274882687,
+      "grad_norm": 0.42877178820545214,
+      "learning_rate": 4.940340912576254e-06,
+      "loss": 0.5674,
+      "step": 3963
+    },
+    {
+      "epoch": 1.5666584341812793,
+      "grad_norm": 0.4180015219456009,
+      "learning_rate": 4.940306885072067e-06,
+      "loss": 0.5465,
+      "step": 3964
+    },
+    {
+      "epoch": 1.5670535934798715,
+      "grad_norm": 0.44880071394905674,
+      "learning_rate": 4.940272847983837e-06,
+      "loss": 0.5638,
+      "step": 3965
+    },
+    {
+      "epoch": 1.5674487527784637,
+      "grad_norm": 0.4470580270494271,
+      "learning_rate": 4.9402388013116965e-06,
+      "loss": 0.5577,
+      "step": 3966
+    },
+    {
+      "epoch": 1.567843912077056,
+      "grad_norm": 0.46986742626918304,
+      "learning_rate": 4.94020474505578e-06,
+      "loss": 0.5754,
+      "step": 3967
+    },
+    {
+      "epoch": 1.5682390713756482,
+      "grad_norm": 0.4719511276473468,
+      "learning_rate": 4.940170679216222e-06,
+      "loss": 0.5465,
+      "step": 3968
+    },
+    {
+      "epoch": 1.5686342306742405,
+      "grad_norm": 0.4819185669692573,
+      "learning_rate": 4.940136603793154e-06,
+      "loss": 0.5903,
+      "step": 3969
+    },
+    {
+      "epoch": 1.5690293899728327,
+      "grad_norm": 0.4596392737276118,
+      "learning_rate": 4.940102518786711e-06,
+      "loss": 0.5827,
+      "step": 3970
+    },
+    {
+      "epoch": 1.569424549271425,
+      "grad_norm": 0.4350274926677015,
+      "learning_rate": 4.9400684241970285e-06,
+      "loss": 0.5732,
+      "step": 3971
+    },
+    {
+      "epoch": 1.5698197085700172,
+      "grad_norm": 0.43547065854272937,
+      "learning_rate": 4.940034320024237e-06,
+      "loss": 0.5587,
+      "step": 3972
+    },
+    {
+      "epoch": 1.5702148678686094,
+      "grad_norm": 0.44965772836569196,
+      "learning_rate": 4.940000206268474e-06,
+      "loss": 0.5757,
+      "step": 3973
+    },
+    {
+      "epoch": 1.5706100271672017,
+      "grad_norm": 0.42960226035038346,
+      "learning_rate": 4.939966082929872e-06,
+      "loss": 0.5411,
+      "step": 3974
+    },
+    {
+      "epoch": 1.571005186465794,
+      "grad_norm": 0.43654243807808146,
+      "learning_rate": 4.939931950008563e-06,
+      "loss": 0.5681,
+      "step": 3975
+    },
+    {
+      "epoch": 1.5714003457643861,
+      "grad_norm": 0.46158348779701347,
+      "learning_rate": 4.939897807504684e-06,
+      "loss": 0.5637,
+      "step": 3976
+    },
+    {
+      "epoch": 1.5717955050629784,
+      "grad_norm": 0.4506664040924429,
+      "learning_rate": 4.939863655418368e-06,
+      "loss": 0.5716,
+      "step": 3977
+    },
+    {
+      "epoch": 1.5721906643615706,
+      "grad_norm": 0.43736830707484653,
+      "learning_rate": 4.939829493749749e-06,
+      "loss": 0.5634,
+      "step": 3978
+    },
+    {
+      "epoch": 1.5725858236601629,
+      "grad_norm": 0.47956231415468115,
+      "learning_rate": 4.939795322498961e-06,
+      "loss": 0.5768,
+      "step": 3979
+    },
+    {
+      "epoch": 1.572980982958755,
+      "grad_norm": 0.47009053670758705,
+      "learning_rate": 4.93976114166614e-06,
+      "loss": 0.5739,
+      "step": 3980
+    },
+    {
+      "epoch": 1.5733761422573473,
+      "grad_norm": 0.4519666405286993,
+      "learning_rate": 4.9397269512514175e-06,
+      "loss": 0.5548,
+      "step": 3981
+    },
+    {
+      "epoch": 1.5737713015559396,
+      "grad_norm": 0.46600888709363636,
+      "learning_rate": 4.939692751254929e-06,
+      "loss": 0.5409,
+      "step": 3982
+    },
+    {
+      "epoch": 1.5741664608545318,
+      "grad_norm": 0.4362128507908923,
+      "learning_rate": 4.939658541676809e-06,
+      "loss": 0.557,
+      "step": 3983
+    },
+    {
+      "epoch": 1.574561620153124,
+      "grad_norm": 0.45311061929248453,
+      "learning_rate": 4.9396243225171916e-06,
+      "loss": 0.5534,
+      "step": 3984
+    },
+    {
+      "epoch": 1.5749567794517165,
+      "grad_norm": 0.4344985151811654,
+      "learning_rate": 4.939590093776211e-06,
+      "loss": 0.5646,
+      "step": 3985
+    },
+    {
+      "epoch": 1.5753519387503088,
+      "grad_norm": 0.4480209913301821,
+      "learning_rate": 4.939555855454003e-06,
+      "loss": 0.5561,
+      "step": 3986
+    },
+    {
+      "epoch": 1.575747098048901,
+      "grad_norm": 0.4531075976767362,
+      "learning_rate": 4.9395216075507e-06,
+      "loss": 0.565,
+      "step": 3987
+    },
+    {
+      "epoch": 1.5761422573474932,
+      "grad_norm": 0.4291612629531522,
+      "learning_rate": 4.939487350066438e-06,
+      "loss": 0.5561,
+      "step": 3988
+    },
+    {
+      "epoch": 1.5765374166460855,
+      "grad_norm": 0.4411400070808928,
+      "learning_rate": 4.93945308300135e-06,
+      "loss": 0.5722,
+      "step": 3989
+    },
+    {
+      "epoch": 1.5769325759446777,
+      "grad_norm": 0.4411690987818465,
+      "learning_rate": 4.939418806355573e-06,
+      "loss": 0.5744,
+      "step": 3990
+    },
+    {
+      "epoch": 1.57732773524327,
+      "grad_norm": 0.4234846841720286,
+      "learning_rate": 4.939384520129239e-06,
+      "loss": 0.5242,
+      "step": 3991
+    },
+    {
+      "epoch": 1.5777228945418622,
+      "grad_norm": 0.43794297369469665,
+      "learning_rate": 4.9393502243224844e-06,
+      "loss": 0.5702,
+      "step": 3992
+    },
+    {
+      "epoch": 1.5781180538404544,
+      "grad_norm": 0.4619795012466095,
+      "learning_rate": 4.9393159189354435e-06,
+      "loss": 0.5602,
+      "step": 3993
+    },
+    {
+      "epoch": 1.5785132131390467,
+      "grad_norm": 0.4248706656017066,
+      "learning_rate": 4.9392816039682516e-06,
+      "loss": 0.5593,
+      "step": 3994
+    },
+    {
+      "epoch": 1.578908372437639,
+      "grad_norm": 0.4581079886861394,
+      "learning_rate": 4.939247279421041e-06,
+      "loss": 0.5602,
+      "step": 3995
+    },
+    {
+      "epoch": 1.5793035317362312,
+      "grad_norm": 0.4305195033029715,
+      "learning_rate": 4.93921294529395e-06,
+      "loss": 0.5596,
+      "step": 3996
+    },
+    {
+      "epoch": 1.5796986910348234,
+      "grad_norm": 0.4366623407663496,
+      "learning_rate": 4.9391786015871106e-06,
+      "loss": 0.5597,
+      "step": 3997
+    },
+    {
+      "epoch": 1.5800938503334157,
+      "grad_norm": 0.44069624259394685,
+      "learning_rate": 4.939144248300659e-06,
+      "loss": 0.5897,
+      "step": 3998
+    },
+    {
+      "epoch": 1.580489009632008,
+      "grad_norm": 0.43890619357978694,
+      "learning_rate": 4.939109885434731e-06,
+      "loss": 0.569,
+      "step": 3999
+    },
+    {
+      "epoch": 1.5808841689306001,
+      "grad_norm": 0.4364576916674103,
+      "learning_rate": 4.939075512989459e-06,
+      "loss": 0.5787,
+      "step": 4000
+    },
+    {
+      "epoch": 1.5812793282291924,
+      "grad_norm": 0.45319416604702467,
+      "learning_rate": 4.93904113096498e-06,
+      "loss": 0.5832,
+      "step": 4001
+    },
+    {
+      "epoch": 1.5816744875277846,
+      "grad_norm": 0.4408144136335299,
+      "learning_rate": 4.939006739361429e-06,
+      "loss": 0.5757,
+      "step": 4002
+    },
+    {
+      "epoch": 1.5820696468263769,
+      "grad_norm": 0.43847351893891434,
+      "learning_rate": 4.93897233817894e-06,
+      "loss": 0.57,
+      "step": 4003
+    },
+    {
+      "epoch": 1.582464806124969,
+      "grad_norm": 0.4272786242930415,
+      "learning_rate": 4.9389379274176485e-06,
+      "loss": 0.561,
+      "step": 4004
+    },
+    {
+      "epoch": 1.5828599654235613,
+      "grad_norm": 0.4273445360462422,
+      "learning_rate": 4.93890350707769e-06,
+      "loss": 0.5543,
+      "step": 4005
+    },
+    {
+      "epoch": 1.5832551247221536,
+      "grad_norm": 0.43334855462016986,
+      "learning_rate": 4.9388690771592e-06,
+      "loss": 0.5632,
+      "step": 4006
+    },
+    {
+      "epoch": 1.5836502840207458,
+      "grad_norm": 0.4412911003838902,
+      "learning_rate": 4.938834637662313e-06,
+      "loss": 0.5717,
+      "step": 4007
+    },
+    {
+      "epoch": 1.584045443319338,
+      "grad_norm": 0.4384873516991556,
+      "learning_rate": 4.9388001885871635e-06,
+      "loss": 0.5569,
+      "step": 4008
+    },
+    {
+      "epoch": 1.5844406026179305,
+      "grad_norm": 0.4357385688585848,
+      "learning_rate": 4.938765729933889e-06,
+      "loss": 0.5656,
+      "step": 4009
+    },
+    {
+      "epoch": 1.5848357619165228,
+      "grad_norm": 0.43820335981947794,
+      "learning_rate": 4.938731261702624e-06,
+      "loss": 0.5726,
+      "step": 4010
+    },
+    {
+      "epoch": 1.585230921215115,
+      "grad_norm": 0.4369840237848306,
+      "learning_rate": 4.938696783893502e-06,
+      "loss": 0.5543,
+      "step": 4011
+    },
+    {
+      "epoch": 1.5856260805137072,
+      "grad_norm": 0.4472393204243838,
+      "learning_rate": 4.938662296506661e-06,
+      "loss": 0.5565,
+      "step": 4012
+    },
+    {
+      "epoch": 1.5860212398122995,
+      "grad_norm": 0.4615011233799181,
+      "learning_rate": 4.938627799542235e-06,
+      "loss": 0.5587,
+      "step": 4013
+    },
+    {
+      "epoch": 1.5864163991108917,
+      "grad_norm": 0.42914755332494525,
+      "learning_rate": 4.93859329300036e-06,
+      "loss": 0.5551,
+      "step": 4014
+    },
+    {
+      "epoch": 1.586811558409484,
+      "grad_norm": 0.6994500397802149,
+      "learning_rate": 4.93855877688117e-06,
+      "loss": 0.5641,
+      "step": 4015
+    },
+    {
+      "epoch": 1.5872067177080762,
+      "grad_norm": 0.4539471102905313,
+      "learning_rate": 4.938524251184803e-06,
+      "loss": 0.5772,
+      "step": 4016
+    },
+    {
+      "epoch": 1.5876018770066684,
+      "grad_norm": 0.4737130066010192,
+      "learning_rate": 4.938489715911394e-06,
+      "loss": 0.5753,
+      "step": 4017
+    },
+    {
+      "epoch": 1.5879970363052607,
+      "grad_norm": 0.4459029139930523,
+      "learning_rate": 4.938455171061077e-06,
+      "loss": 0.5774,
+      "step": 4018
+    },
+    {
+      "epoch": 1.588392195603853,
+      "grad_norm": 0.4434353397507251,
+      "learning_rate": 4.93842061663399e-06,
+      "loss": 0.5581,
+      "step": 4019
+    },
+    {
+      "epoch": 1.5887873549024452,
+      "grad_norm": 0.43023249449874645,
+      "learning_rate": 4.938386052630267e-06,
+      "loss": 0.5429,
+      "step": 4020
+    },
+    {
+      "epoch": 1.5891825142010374,
+      "grad_norm": 0.45310635393738163,
+      "learning_rate": 4.938351479050044e-06,
+      "loss": 0.5674,
+      "step": 4021
+    },
+    {
+      "epoch": 1.5895776734996296,
+      "grad_norm": 0.45352464255321795,
+      "learning_rate": 4.938316895893458e-06,
+      "loss": 0.5661,
+      "step": 4022
+    },
+    {
+      "epoch": 1.5899728327982219,
+      "grad_norm": 0.4532822555467358,
+      "learning_rate": 4.938282303160643e-06,
+      "loss": 0.5515,
+      "step": 4023
+    },
+    {
+      "epoch": 1.5903679920968141,
+      "grad_norm": 0.44533687524094095,
+      "learning_rate": 4.938247700851735e-06,
+      "loss": 0.5511,
+      "step": 4024
+    },
+    {
+      "epoch": 1.5907631513954064,
+      "grad_norm": 0.45327422167542913,
+      "learning_rate": 4.938213088966872e-06,
+      "loss": 0.5691,
+      "step": 4025
+    },
+    {
+      "epoch": 1.5911583106939986,
+      "grad_norm": 0.43728559584246246,
+      "learning_rate": 4.938178467506187e-06,
+      "loss": 0.5565,
+      "step": 4026
+    },
+    {
+      "epoch": 1.5915534699925908,
+      "grad_norm": 0.4393345012809165,
+      "learning_rate": 4.938143836469818e-06,
+      "loss": 0.5432,
+      "step": 4027
+    },
+    {
+      "epoch": 1.591948629291183,
+      "grad_norm": 0.4446012831175294,
+      "learning_rate": 4.938109195857902e-06,
+      "loss": 0.5782,
+      "step": 4028
+    },
+    {
+      "epoch": 1.5923437885897753,
+      "grad_norm": 0.4476104890005792,
+      "learning_rate": 4.938074545670573e-06,
+      "loss": 0.58,
+      "step": 4029
+    },
+    {
+      "epoch": 1.5927389478883676,
+      "grad_norm": 0.4469302927676119,
+      "learning_rate": 4.938039885907967e-06,
+      "loss": 0.582,
+      "step": 4030
+    },
+    {
+      "epoch": 1.5931341071869598,
+      "grad_norm": 0.4326506856607228,
+      "learning_rate": 4.938005216570221e-06,
+      "loss": 0.5587,
+      "step": 4031
+    },
+    {
+      "epoch": 1.593529266485552,
+      "grad_norm": 0.45215823190175686,
+      "learning_rate": 4.9379705376574705e-06,
+      "loss": 0.5416,
+      "step": 4032
+    },
+    {
+      "epoch": 1.5939244257841443,
+      "grad_norm": 0.4604923926371955,
+      "learning_rate": 4.937935849169853e-06,
+      "loss": 0.5814,
+      "step": 4033
+    },
+    {
+      "epoch": 1.5943195850827365,
+      "grad_norm": 0.4483793169947381,
+      "learning_rate": 4.937901151107504e-06,
+      "loss": 0.5606,
+      "step": 4034
+    },
+    {
+      "epoch": 1.5947147443813288,
+      "grad_norm": 0.4731607560798562,
+      "learning_rate": 4.937866443470558e-06,
+      "loss": 0.5616,
+      "step": 4035
+    },
+    {
+      "epoch": 1.595109903679921,
+      "grad_norm": 0.44139056004958066,
+      "learning_rate": 4.9378317262591545e-06,
+      "loss": 0.5704,
+      "step": 4036
+    },
+    {
+      "epoch": 1.5955050629785132,
+      "grad_norm": 0.43225652789170976,
+      "learning_rate": 4.937796999473427e-06,
+      "loss": 0.545,
+      "step": 4037
+    },
+    {
+      "epoch": 1.5959002222771055,
+      "grad_norm": 0.4376950142264724,
+      "learning_rate": 4.937762263113515e-06,
+      "loss": 0.5607,
+      "step": 4038
+    },
+    {
+      "epoch": 1.5962953815756977,
+      "grad_norm": 0.43266642151648355,
+      "learning_rate": 4.937727517179552e-06,
+      "loss": 0.5734,
+      "step": 4039
+    },
+    {
+      "epoch": 1.59669054087429,
+      "grad_norm": 0.4491269605040788,
+      "learning_rate": 4.9376927616716764e-06,
+      "loss": 0.5593,
+      "step": 4040
+    },
+    {
+      "epoch": 1.5970857001728822,
+      "grad_norm": 0.4339310588220914,
+      "learning_rate": 4.937657996590023e-06,
+      "loss": 0.5706,
+      "step": 4041
+    },
+    {
+      "epoch": 1.5974808594714744,
+      "grad_norm": 0.4300915782723815,
+      "learning_rate": 4.93762322193473e-06,
+      "loss": 0.549,
+      "step": 4042
+    },
+    {
+      "epoch": 1.5978760187700667,
+      "grad_norm": 0.4349429575023993,
+      "learning_rate": 4.937588437705933e-06,
+      "loss": 0.5673,
+      "step": 4043
+    },
+    {
+      "epoch": 1.598271178068659,
+      "grad_norm": 0.43087812609014015,
+      "learning_rate": 4.937553643903768e-06,
+      "loss": 0.5514,
+      "step": 4044
+    },
+    {
+      "epoch": 1.5986663373672512,
+      "grad_norm": 0.4497997607400426,
+      "learning_rate": 4.937518840528373e-06,
+      "loss": 0.5593,
+      "step": 4045
+    },
+    {
+      "epoch": 1.5990614966658434,
+      "grad_norm": 0.44562506368324917,
+      "learning_rate": 4.937484027579883e-06,
+      "loss": 0.5587,
+      "step": 4046
+    },
+    {
+      "epoch": 1.5994566559644356,
+      "grad_norm": 0.43652547554219967,
+      "learning_rate": 4.937449205058438e-06,
+      "loss": 0.58,
+      "step": 4047
+    },
+    {
+      "epoch": 1.5998518152630279,
+      "grad_norm": 0.4240519663611341,
+      "learning_rate": 4.937414372964171e-06,
+      "loss": 0.5486,
+      "step": 4048
+    },
+    {
+      "epoch": 1.6002469745616201,
+      "grad_norm": 0.45799730349541395,
+      "learning_rate": 4.937379531297221e-06,
+      "loss": 0.5567,
+      "step": 4049
+    },
+    {
+      "epoch": 1.6006421338602124,
+      "grad_norm": 0.4292261145352039,
+      "learning_rate": 4.937344680057724e-06,
+      "loss": 0.5505,
+      "step": 4050
+    },
+    {
+      "epoch": 1.6010372931588046,
+      "grad_norm": 0.42937598054316073,
+      "learning_rate": 4.937309819245818e-06,
+      "loss": 0.5601,
+      "step": 4051
+    },
+    {
+      "epoch": 1.6014324524573968,
+      "grad_norm": 0.4380122077896125,
+      "learning_rate": 4.937274948861638e-06,
+      "loss": 0.5608,
+      "step": 4052
+    },
+    {
+      "epoch": 1.601827611755989,
+      "grad_norm": 0.4548432558319798,
+      "learning_rate": 4.937240068905322e-06,
+      "loss": 0.5685,
+      "step": 4053
+    },
+    {
+      "epoch": 1.6022227710545813,
+      "grad_norm": 0.43181199483513577,
+      "learning_rate": 4.937205179377008e-06,
+      "loss": 0.5646,
+      "step": 4054
+    },
+    {
+      "epoch": 1.6026179303531736,
+      "grad_norm": 0.43979884745381725,
+      "learning_rate": 4.937170280276831e-06,
+      "loss": 0.582,
+      "step": 4055
+    },
+    {
+      "epoch": 1.6030130896517658,
+      "grad_norm": 0.428767044050838,
+      "learning_rate": 4.93713537160493e-06,
+      "loss": 0.545,
+      "step": 4056
+    },
+    {
+      "epoch": 1.603408248950358,
+      "grad_norm": 0.4444074071525543,
+      "learning_rate": 4.9371004533614395e-06,
+      "loss": 0.5357,
+      "step": 4057
+    },
+    {
+      "epoch": 1.6038034082489503,
+      "grad_norm": 0.4305444155233475,
+      "learning_rate": 4.9370655255465e-06,
+      "loss": 0.5483,
+      "step": 4058
+    },
+    {
+      "epoch": 1.6041985675475425,
+      "grad_norm": 0.44066258955092896,
+      "learning_rate": 4.937030588160247e-06,
+      "loss": 0.5631,
+      "step": 4059
+    },
+    {
+      "epoch": 1.6045937268461348,
+      "grad_norm": 0.4487310628582567,
+      "learning_rate": 4.936995641202816e-06,
+      "loss": 0.5761,
+      "step": 4060
+    },
+    {
+      "epoch": 1.604988886144727,
+      "grad_norm": 0.44499130482437343,
+      "learning_rate": 4.936960684674348e-06,
+      "loss": 0.5692,
+      "step": 4061
+    },
+    {
+      "epoch": 1.6053840454433193,
+      "grad_norm": 0.4412506035853422,
+      "learning_rate": 4.9369257185749766e-06,
+      "loss": 0.5496,
+      "step": 4062
+    },
+    {
+      "epoch": 1.6057792047419115,
+      "grad_norm": 0.4359722621598534,
+      "learning_rate": 4.936890742904842e-06,
+      "loss": 0.5493,
+      "step": 4063
+    },
+    {
+      "epoch": 1.6061743640405037,
+      "grad_norm": 0.4340637645593551,
+      "learning_rate": 4.936855757664079e-06,
+      "loss": 0.5566,
+      "step": 4064
+    },
+    {
+      "epoch": 1.606569523339096,
+      "grad_norm": 0.4289469251048877,
+      "learning_rate": 4.936820762852827e-06,
+      "loss": 0.5362,
+      "step": 4065
+    },
+    {
+      "epoch": 1.6069646826376882,
+      "grad_norm": 0.4319109469211228,
+      "learning_rate": 4.936785758471223e-06,
+      "loss": 0.5601,
+      "step": 4066
+    },
+    {
+      "epoch": 1.6073598419362805,
+      "grad_norm": 0.4476806397110603,
+      "learning_rate": 4.936750744519404e-06,
+      "loss": 0.5868,
+      "step": 4067
+    },
+    {
+      "epoch": 1.6077550012348727,
+      "grad_norm": 0.4386152539571968,
+      "learning_rate": 4.936715720997508e-06,
+      "loss": 0.5643,
+      "step": 4068
+    },
+    {
+      "epoch": 1.608150160533465,
+      "grad_norm": 0.4279438753027654,
+      "learning_rate": 4.936680687905673e-06,
+      "loss": 0.5575,
+      "step": 4069
+    },
+    {
+      "epoch": 1.6085453198320572,
+      "grad_norm": 0.4326188532708726,
+      "learning_rate": 4.936645645244034e-06,
+      "loss": 0.5598,
+      "step": 4070
+    },
+    {
+      "epoch": 1.6089404791306494,
+      "grad_norm": 0.44654932545961695,
+      "learning_rate": 4.936610593012732e-06,
+      "loss": 0.5678,
+      "step": 4071
+    },
+    {
+      "epoch": 1.6093356384292417,
+      "grad_norm": 0.4350127525034427,
+      "learning_rate": 4.936575531211902e-06,
+      "loss": 0.552,
+      "step": 4072
+    },
+    {
+      "epoch": 1.609730797727834,
+      "grad_norm": 0.4334061095799106,
+      "learning_rate": 4.936540459841684e-06,
+      "loss": 0.5631,
+      "step": 4073
+    },
+    {
+      "epoch": 1.6101259570264261,
+      "grad_norm": 0.4380393010701646,
+      "learning_rate": 4.9365053789022145e-06,
+      "loss": 0.5505,
+      "step": 4074
+    },
+    {
+      "epoch": 1.6105211163250184,
+      "grad_norm": 0.44462390814852865,
+      "learning_rate": 4.936470288393631e-06,
+      "loss": 0.548,
+      "step": 4075
+    },
+    {
+      "epoch": 1.6109162756236106,
+      "grad_norm": 0.4391814847764043,
+      "learning_rate": 4.936435188316071e-06,
+      "loss": 0.5508,
+      "step": 4076
+    },
+    {
+      "epoch": 1.6113114349222029,
+      "grad_norm": 0.45314650675562296,
+      "learning_rate": 4.936400078669674e-06,
+      "loss": 0.5763,
+      "step": 4077
+    },
+    {
+      "epoch": 1.611706594220795,
+      "grad_norm": 0.42859513707939456,
+      "learning_rate": 4.936364959454577e-06,
+      "loss": 0.5534,
+      "step": 4078
+    },
+    {
+      "epoch": 1.6121017535193873,
+      "grad_norm": 0.43651873348895415,
+      "learning_rate": 4.936329830670918e-06,
+      "loss": 0.5539,
+      "step": 4079
+    },
+    {
+      "epoch": 1.6124969128179798,
+      "grad_norm": 0.4464140156441266,
+      "learning_rate": 4.9362946923188345e-06,
+      "loss": 0.5639,
+      "step": 4080
+    },
+    {
+      "epoch": 1.612892072116572,
+      "grad_norm": 0.43474468854326337,
+      "learning_rate": 4.936259544398465e-06,
+      "loss": 0.567,
+      "step": 4081
+    },
+    {
+      "epoch": 1.6132872314151643,
+      "grad_norm": 0.446752616186375,
+      "learning_rate": 4.936224386909947e-06,
+      "loss": 0.5701,
+      "step": 4082
+    },
+    {
+      "epoch": 1.6136823907137565,
+      "grad_norm": 0.4281198845056971,
+      "learning_rate": 4.93618921985342e-06,
+      "loss": 0.5582,
+      "step": 4083
+    },
+    {
+      "epoch": 1.6140775500123488,
+      "grad_norm": 0.4460784589294146,
+      "learning_rate": 4.93615404322902e-06,
+      "loss": 0.5565,
+      "step": 4084
+    },
+    {
+      "epoch": 1.614472709310941,
+      "grad_norm": 0.4488735337458113,
+      "learning_rate": 4.936118857036887e-06,
+      "loss": 0.5498,
+      "step": 4085
+    },
+    {
+      "epoch": 1.6148678686095332,
+      "grad_norm": 0.4364780045470906,
+      "learning_rate": 4.936083661277158e-06,
+      "loss": 0.5603,
+      "step": 4086
+    },
+    {
+      "epoch": 1.6152630279081255,
+      "grad_norm": 0.4263849821332742,
+      "learning_rate": 4.936048455949971e-06,
+      "loss": 0.5407,
+      "step": 4087
+    },
+    {
+      "epoch": 1.6156581872067177,
+      "grad_norm": 0.4400718397147201,
+      "learning_rate": 4.936013241055465e-06,
+      "loss": 0.5762,
+      "step": 4088
+    },
+    {
+      "epoch": 1.61605334650531,
+      "grad_norm": 0.4308631720264388,
+      "learning_rate": 4.935978016593779e-06,
+      "loss": 0.566,
+      "step": 4089
+    },
+    {
+      "epoch": 1.6164485058039022,
+      "grad_norm": 0.4299942853443749,
+      "learning_rate": 4.935942782565051e-06,
+      "loss": 0.5337,
+      "step": 4090
+    },
+    {
+      "epoch": 1.6168436651024944,
+      "grad_norm": 0.4366671539606536,
+      "learning_rate": 4.935907538969418e-06,
+      "loss": 0.555,
+      "step": 4091
+    },
+    {
+      "epoch": 1.6172388244010867,
+      "grad_norm": 0.48254923126585936,
+      "learning_rate": 4.93587228580702e-06,
+      "loss": 0.5595,
+      "step": 4092
+    },
+    {
+      "epoch": 1.617633983699679,
+      "grad_norm": 0.4239126920096438,
+      "learning_rate": 4.935837023077994e-06,
+      "loss": 0.5517,
+      "step": 4093
+    },
+    {
+      "epoch": 1.6180291429982712,
+      "grad_norm": 0.436879102852562,
+      "learning_rate": 4.93580175078248e-06,
+      "loss": 0.545,
+      "step": 4094
+    },
+    {
+      "epoch": 1.6184243022968634,
+      "grad_norm": 0.44498901921393713,
+      "learning_rate": 4.935766468920615e-06,
+      "loss": 0.5639,
+      "step": 4095
+    },
+    {
+      "epoch": 1.6188194615954556,
+      "grad_norm": 0.4208783410765207,
+      "learning_rate": 4.935731177492539e-06,
+      "loss": 0.5513,
+      "step": 4096
+    },
+    {
+      "epoch": 1.6192146208940479,
+      "grad_norm": 0.4403934391341918,
+      "learning_rate": 4.93569587649839e-06,
+      "loss": 0.5597,
+      "step": 4097
+    },
+    {
+      "epoch": 1.6196097801926401,
+      "grad_norm": 0.4445680674509993,
+      "learning_rate": 4.935660565938306e-06,
+      "loss": 0.5747,
+      "step": 4098
+    },
+    {
+      "epoch": 1.6200049394912324,
+      "grad_norm": 0.4542236992519073,
+      "learning_rate": 4.935625245812426e-06,
+      "loss": 0.5557,
+      "step": 4099
+    },
+    {
+      "epoch": 1.6204000987898246,
+      "grad_norm": 0.4448277099086342,
+      "learning_rate": 4.935589916120891e-06,
+      "loss": 0.5652,
+      "step": 4100
+    },
+    {
+      "epoch": 1.6207952580884168,
+      "grad_norm": 0.4383946322249754,
+      "learning_rate": 4.935554576863837e-06,
+      "loss": 0.5433,
+      "step": 4101
+    },
+    {
+      "epoch": 1.621190417387009,
+      "grad_norm": 0.43502119286992535,
+      "learning_rate": 4.9355192280414024e-06,
+      "loss": 0.5693,
+      "step": 4102
+    },
+    {
+      "epoch": 1.6215855766856015,
+      "grad_norm": 0.4316383774413981,
+      "learning_rate": 4.935483869653728e-06,
+      "loss": 0.5677,
+      "step": 4103
+    },
+    {
+      "epoch": 1.6219807359841938,
+      "grad_norm": 0.4341260683812164,
+      "learning_rate": 4.935448501700953e-06,
+      "loss": 0.5391,
+      "step": 4104
+    },
+    {
+      "epoch": 1.622375895282786,
+      "grad_norm": 0.43195315752672536,
+      "learning_rate": 4.935413124183212e-06,
+      "loss": 0.5639,
+      "step": 4105
+    },
+    {
+      "epoch": 1.6227710545813783,
+      "grad_norm": 0.43507822462287415,
+      "learning_rate": 4.93537773710065e-06,
+      "loss": 0.5576,
+      "step": 4106
+    },
+    {
+      "epoch": 1.6231662138799705,
+      "grad_norm": 0.46638962293882175,
+      "learning_rate": 4.935342340453402e-06,
+      "loss": 0.5631,
+      "step": 4107
+    },
+    {
+      "epoch": 1.6235613731785628,
+      "grad_norm": 0.4558364456192893,
+      "learning_rate": 4.9353069342416085e-06,
+      "loss": 0.5566,
+      "step": 4108
+    },
+    {
+      "epoch": 1.623956532477155,
+      "grad_norm": 0.43312180311952486,
+      "learning_rate": 4.935271518465408e-06,
+      "loss": 0.56,
+      "step": 4109
+    },
+    {
+      "epoch": 1.6243516917757472,
+      "grad_norm": 0.4365748990534713,
+      "learning_rate": 4.93523609312494e-06,
+      "loss": 0.5629,
+      "step": 4110
+    },
+    {
+      "epoch": 1.6247468510743395,
+      "grad_norm": 0.4587870833960197,
+      "learning_rate": 4.935200658220342e-06,
+      "loss": 0.5741,
+      "step": 4111
+    },
+    {
+      "epoch": 1.6251420103729317,
+      "grad_norm": 0.43339139678044175,
+      "learning_rate": 4.935165213751757e-06,
+      "loss": 0.5569,
+      "step": 4112
+    },
+    {
+      "epoch": 1.625537169671524,
+      "grad_norm": 0.4428057531936081,
+      "learning_rate": 4.93512975971932e-06,
+      "loss": 0.5683,
+      "step": 4113
+    },
+    {
+      "epoch": 1.6259323289701162,
+      "grad_norm": 0.4578729689342045,
+      "learning_rate": 4.935094296123172e-06,
+      "loss": 0.5501,
+      "step": 4114
+    },
+    {
+      "epoch": 1.6263274882687084,
+      "grad_norm": 0.45476951551439365,
+      "learning_rate": 4.935058822963454e-06,
+      "loss": 0.5709,
+      "step": 4115
+    },
+    {
+      "epoch": 1.6267226475673007,
+      "grad_norm": 0.45641536816268186,
+      "learning_rate": 4.935023340240301e-06,
+      "loss": 0.5564,
+      "step": 4116
+    },
+    {
+      "epoch": 1.627117806865893,
+      "grad_norm": 0.4488224757255472,
+      "learning_rate": 4.934987847953856e-06,
+      "loss": 0.5619,
+      "step": 4117
+    },
+    {
+      "epoch": 1.6275129661644852,
+      "grad_norm": 0.46880383951054816,
+      "learning_rate": 4.934952346104258e-06,
+      "loss": 0.5604,
+      "step": 4118
+    },
+    {
+      "epoch": 1.6279081254630774,
+      "grad_norm": 0.545037272078235,
+      "learning_rate": 4.9349168346916454e-06,
+      "loss": 0.543,
+      "step": 4119
+    },
+    {
+      "epoch": 1.6283032847616696,
+      "grad_norm": 0.4366418096115771,
+      "learning_rate": 4.934881313716158e-06,
+      "loss": 0.5722,
+      "step": 4120
+    },
+    {
+      "epoch": 1.6286984440602619,
+      "grad_norm": 0.43777846292338235,
+      "learning_rate": 4.934845783177935e-06,
+      "loss": 0.5524,
+      "step": 4121
+    },
+    {
+      "epoch": 1.6290936033588541,
+      "grad_norm": 0.422011213186665,
+      "learning_rate": 4.934810243077117e-06,
+      "loss": 0.5733,
+      "step": 4122
+    },
+    {
+      "epoch": 1.6294887626574464,
+      "grad_norm": 0.44138201150444306,
+      "learning_rate": 4.934774693413843e-06,
+      "loss": 0.5686,
+      "step": 4123
+    },
+    {
+      "epoch": 1.6298839219560386,
+      "grad_norm": 0.44921636818156546,
+      "learning_rate": 4.934739134188251e-06,
+      "loss": 0.5724,
+      "step": 4124
+    },
+    {
+      "epoch": 1.6302790812546308,
+      "grad_norm": 0.44859391331022486,
+      "learning_rate": 4.934703565400484e-06,
+      "loss": 0.5821,
+      "step": 4125
+    },
+    {
+      "epoch": 1.630674240553223,
+      "grad_norm": 0.4233228039840541,
+      "learning_rate": 4.934667987050678e-06,
+      "loss": 0.5571,
+      "step": 4126
+    },
+    {
+      "epoch": 1.6310693998518153,
+      "grad_norm": 0.4338865601440852,
+      "learning_rate": 4.934632399138976e-06,
+      "loss": 0.5675,
+      "step": 4127
+    },
+    {
+      "epoch": 1.6314645591504076,
+      "grad_norm": 0.4394241216685595,
+      "learning_rate": 4.934596801665515e-06,
+      "loss": 0.5571,
+      "step": 4128
+    },
+    {
+      "epoch": 1.6318597184489998,
+      "grad_norm": 0.43864069069871264,
+      "learning_rate": 4.934561194630437e-06,
+      "loss": 0.5381,
+      "step": 4129
+    },
+    {
+      "epoch": 1.632254877747592,
+      "grad_norm": 0.42429399086993547,
+      "learning_rate": 4.934525578033881e-06,
+      "loss": 0.5587,
+      "step": 4130
+    },
+    {
+      "epoch": 1.6326500370461843,
+      "grad_norm": 0.4404107423141753,
+      "learning_rate": 4.934489951875987e-06,
+      "loss": 0.5721,
+      "step": 4131
+    },
+    {
+      "epoch": 1.6330451963447765,
+      "grad_norm": 0.4332961401456829,
+      "learning_rate": 4.934454316156894e-06,
+      "loss": 0.5638,
+      "step": 4132
+    },
+    {
+      "epoch": 1.6334403556433688,
+      "grad_norm": 0.4695705815950691,
+      "learning_rate": 4.934418670876743e-06,
+      "loss": 0.561,
+      "step": 4133
+    },
+    {
+      "epoch": 1.633835514941961,
+      "grad_norm": 0.45507102791172976,
+      "learning_rate": 4.9343830160356744e-06,
+      "loss": 0.5713,
+      "step": 4134
+    },
+    {
+      "epoch": 1.6342306742405532,
+      "grad_norm": 0.43768991836081356,
+      "learning_rate": 4.934347351633827e-06,
+      "loss": 0.556,
+      "step": 4135
+    },
+    {
+      "epoch": 1.6346258335391455,
+      "grad_norm": 0.43283166401271617,
+      "learning_rate": 4.934311677671342e-06,
+      "loss": 0.5596,
+      "step": 4136
+    },
+    {
+      "epoch": 1.6350209928377377,
+      "grad_norm": 0.42555330658880147,
+      "learning_rate": 4.934275994148357e-06,
+      "loss": 0.5593,
+      "step": 4137
+    },
+    {
+      "epoch": 1.63541615213633,
+      "grad_norm": 0.44485035790650657,
+      "learning_rate": 4.934240301065016e-06,
+      "loss": 0.5739,
+      "step": 4138
+    },
+    {
+      "epoch": 1.6358113114349222,
+      "grad_norm": 0.4409925294439782,
+      "learning_rate": 4.934204598421457e-06,
+      "loss": 0.5643,
+      "step": 4139
+    },
+    {
+      "epoch": 1.6362064707335144,
+      "grad_norm": 0.4310591741238124,
+      "learning_rate": 4.934168886217821e-06,
+      "loss": 0.571,
+      "step": 4140
+    },
+    {
+      "epoch": 1.6366016300321067,
+      "grad_norm": 0.43546524799796604,
+      "learning_rate": 4.934133164454246e-06,
+      "loss": 0.5682,
+      "step": 4141
+    },
+    {
+      "epoch": 1.636996789330699,
+      "grad_norm": 0.42766476649034174,
+      "learning_rate": 4.934097433130875e-06,
+      "loss": 0.5587,
+      "step": 4142
+    },
+    {
+      "epoch": 1.6373919486292912,
+      "grad_norm": 0.4381721056390369,
+      "learning_rate": 4.934061692247847e-06,
+      "loss": 0.5627,
+      "step": 4143
+    },
+    {
+      "epoch": 1.6377871079278834,
+      "grad_norm": 0.4289920275995806,
+      "learning_rate": 4.9340259418053035e-06,
+      "loss": 0.547,
+      "step": 4144
+    },
+    {
+      "epoch": 1.6381822672264756,
+      "grad_norm": 0.4264307971370297,
+      "learning_rate": 4.933990181803383e-06,
+      "loss": 0.5569,
+      "step": 4145
+    },
+    {
+      "epoch": 1.6385774265250679,
+      "grad_norm": 0.4296221573612008,
+      "learning_rate": 4.933954412242228e-06,
+      "loss": 0.5523,
+      "step": 4146
+    },
+    {
+      "epoch": 1.6389725858236601,
+      "grad_norm": 0.43507977648083157,
+      "learning_rate": 4.933918633121978e-06,
+      "loss": 0.5561,
+      "step": 4147
+    },
+    {
+      "epoch": 1.6393677451222524,
+      "grad_norm": 0.42056933159661486,
+      "learning_rate": 4.933882844442773e-06,
+      "loss": 0.5636,
+      "step": 4148
+    },
+    {
+      "epoch": 1.6397629044208446,
+      "grad_norm": 0.4433116034912783,
+      "learning_rate": 4.933847046204754e-06,
+      "loss": 0.5553,
+      "step": 4149
+    },
+    {
+      "epoch": 1.6401580637194368,
+      "grad_norm": 0.4459630268656738,
+      "learning_rate": 4.933811238408063e-06,
+      "loss": 0.5577,
+      "step": 4150
+    },
+    {
+      "epoch": 1.640553223018029,
+      "grad_norm": 0.4403056379106962,
+      "learning_rate": 4.933775421052838e-06,
+      "loss": 0.5833,
+      "step": 4151
+    },
+    {
+      "epoch": 1.6409483823166213,
+      "grad_norm": 0.42886531696805363,
+      "learning_rate": 4.933739594139221e-06,
+      "loss": 0.541,
+      "step": 4152
+    },
+    {
+      "epoch": 1.6413435416152136,
+      "grad_norm": 0.4315557640640133,
+      "learning_rate": 4.933703757667353e-06,
+      "loss": 0.5575,
+      "step": 4153
+    },
+    {
+      "epoch": 1.6417387009138058,
+      "grad_norm": 0.48140901833929156,
+      "learning_rate": 4.933667911637375e-06,
+      "loss": 0.5783,
+      "step": 4154
+    },
+    {
+      "epoch": 1.642133860212398,
+      "grad_norm": 0.42997667165062453,
+      "learning_rate": 4.933632056049427e-06,
+      "loss": 0.5482,
+      "step": 4155
+    },
+    {
+      "epoch": 1.6425290195109903,
+      "grad_norm": 0.429706934348882,
+      "learning_rate": 4.93359619090365e-06,
+      "loss": 0.5721,
+      "step": 4156
+    },
+    {
+      "epoch": 1.6429241788095825,
+      "grad_norm": 0.4445206300843515,
+      "learning_rate": 4.933560316200185e-06,
+      "loss": 0.5742,
+      "step": 4157
+    },
+    {
+      "epoch": 1.6433193381081748,
+      "grad_norm": 0.433653943288636,
+      "learning_rate": 4.933524431939173e-06,
+      "loss": 0.5682,
+      "step": 4158
+    },
+    {
+      "epoch": 1.643714497406767,
+      "grad_norm": 0.43516906274908146,
+      "learning_rate": 4.933488538120754e-06,
+      "loss": 0.5658,
+      "step": 4159
+    },
+    {
+      "epoch": 1.6441096567053592,
+      "grad_norm": 0.43315071912764175,
+      "learning_rate": 4.933452634745071e-06,
+      "loss": 0.5605,
+      "step": 4160
+    },
+    {
+      "epoch": 1.6445048160039515,
+      "grad_norm": 0.4623128049190355,
+      "learning_rate": 4.933416721812262e-06,
+      "loss": 0.5451,
+      "step": 4161
+    },
+    {
+      "epoch": 1.6448999753025437,
+      "grad_norm": 0.4322419596499138,
+      "learning_rate": 4.933380799322471e-06,
+      "loss": 0.5405,
+      "step": 4162
+    },
+    {
+      "epoch": 1.645295134601136,
+      "grad_norm": 0.4353017445216181,
+      "learning_rate": 4.933344867275837e-06,
+      "loss": 0.5465,
+      "step": 4163
+    },
+    {
+      "epoch": 1.6456902938997282,
+      "grad_norm": 0.42774307582614995,
+      "learning_rate": 4.9333089256725034e-06,
+      "loss": 0.5272,
+      "step": 4164
+    },
+    {
+      "epoch": 1.6460854531983204,
+      "grad_norm": 0.43158324134092596,
+      "learning_rate": 4.9332729745126085e-06,
+      "loss": 0.5593,
+      "step": 4165
+    },
+    {
+      "epoch": 1.6464806124969127,
+      "grad_norm": 0.43107023737645744,
+      "learning_rate": 4.933237013796295e-06,
+      "loss": 0.5674,
+      "step": 4166
+    },
+    {
+      "epoch": 1.646875771795505,
+      "grad_norm": 0.470689528328689,
+      "learning_rate": 4.9332010435237045e-06,
+      "loss": 0.5735,
+      "step": 4167
+    },
+    {
+      "epoch": 1.6472709310940972,
+      "grad_norm": 0.4337591992128468,
+      "learning_rate": 4.933165063694978e-06,
+      "loss": 0.5651,
+      "step": 4168
+    },
+    {
+      "epoch": 1.6476660903926894,
+      "grad_norm": 0.433598238285656,
+      "learning_rate": 4.933129074310257e-06,
+      "loss": 0.5609,
+      "step": 4169
+    },
+    {
+      "epoch": 1.6480612496912816,
+      "grad_norm": 0.4383209332082589,
+      "learning_rate": 4.933093075369681e-06,
+      "loss": 0.5675,
+      "step": 4170
+    },
+    {
+      "epoch": 1.648456408989874,
+      "grad_norm": 0.47207585803002206,
+      "learning_rate": 4.933057066873394e-06,
+      "loss": 0.5662,
+      "step": 4171
+    },
+    {
+      "epoch": 1.6488515682884661,
+      "grad_norm": 0.472525910555258,
+      "learning_rate": 4.933021048821536e-06,
+      "loss": 0.5402,
+      "step": 4172
+    },
+    {
+      "epoch": 1.6492467275870584,
+      "grad_norm": 0.4760221302508032,
+      "learning_rate": 4.932985021214248e-06,
+      "loss": 0.5783,
+      "step": 4173
+    },
+    {
+      "epoch": 1.6496418868856508,
+      "grad_norm": 0.433597879420229,
+      "learning_rate": 4.932948984051673e-06,
+      "loss": 0.5731,
+      "step": 4174
+    },
+    {
+      "epoch": 1.650037046184243,
+      "grad_norm": 0.4402059167564898,
+      "learning_rate": 4.9329129373339525e-06,
+      "loss": 0.547,
+      "step": 4175
+    },
+    {
+      "epoch": 1.6504322054828353,
+      "grad_norm": 0.4546433582700222,
+      "learning_rate": 4.932876881061226e-06,
+      "loss": 0.5469,
+      "step": 4176
+    },
+    {
+      "epoch": 1.6508273647814276,
+      "grad_norm": 0.4316382194174931,
+      "learning_rate": 4.932840815233637e-06,
+      "loss": 0.5392,
+      "step": 4177
+    },
+    {
+      "epoch": 1.6512225240800198,
+      "grad_norm": 0.4334651261656393,
+      "learning_rate": 4.9328047398513265e-06,
+      "loss": 0.5556,
+      "step": 4178
+    },
+    {
+      "epoch": 1.651617683378612,
+      "grad_norm": 0.5167325693733595,
+      "learning_rate": 4.932768654914437e-06,
+      "loss": 0.5574,
+      "step": 4179
+    },
+    {
+      "epoch": 1.6520128426772043,
+      "grad_norm": 0.4537887100737346,
+      "learning_rate": 4.932732560423108e-06,
+      "loss": 0.5519,
+      "step": 4180
+    },
+    {
+      "epoch": 1.6524080019757965,
+      "grad_norm": 0.4407088531206237,
+      "learning_rate": 4.932696456377484e-06,
+      "loss": 0.5602,
+      "step": 4181
+    },
+    {
+      "epoch": 1.6528031612743888,
+      "grad_norm": 0.4239271803777524,
+      "learning_rate": 4.932660342777705e-06,
+      "loss": 0.5344,
+      "step": 4182
+    },
+    {
+      "epoch": 1.653198320572981,
+      "grad_norm": 0.4221587419527885,
+      "learning_rate": 4.932624219623913e-06,
+      "loss": 0.5419,
+      "step": 4183
+    },
+    {
+      "epoch": 1.6535934798715732,
+      "grad_norm": 0.45091328428360944,
+      "learning_rate": 4.932588086916251e-06,
+      "loss": 0.5569,
+      "step": 4184
+    },
+    {
+      "epoch": 1.6539886391701655,
+      "grad_norm": 0.4535113787923153,
+      "learning_rate": 4.93255194465486e-06,
+      "loss": 0.5311,
+      "step": 4185
+    },
+    {
+      "epoch": 1.6543837984687577,
+      "grad_norm": 0.4495246513741634,
+      "learning_rate": 4.932515792839882e-06,
+      "loss": 0.5459,
+      "step": 4186
+    },
+    {
+      "epoch": 1.65477895776735,
+      "grad_norm": 0.45103027154827796,
+      "learning_rate": 4.932479631471459e-06,
+      "loss": 0.5617,
+      "step": 4187
+    },
+    {
+      "epoch": 1.6551741170659422,
+      "grad_norm": 0.46161108604777257,
+      "learning_rate": 4.932443460549733e-06,
+      "loss": 0.5744,
+      "step": 4188
+    },
+    {
+      "epoch": 1.6555692763645344,
+      "grad_norm": 0.4585641137181729,
+      "learning_rate": 4.932407280074846e-06,
+      "loss": 0.5571,
+      "step": 4189
+    },
+    {
+      "epoch": 1.6559644356631267,
+      "grad_norm": 0.47002181393717396,
+      "learning_rate": 4.93237109004694e-06,
+      "loss": 0.5591,
+      "step": 4190
+    },
+    {
+      "epoch": 1.656359594961719,
+      "grad_norm": 0.45813305720090625,
+      "learning_rate": 4.932334890466158e-06,
+      "loss": 0.5509,
+      "step": 4191
+    },
+    {
+      "epoch": 1.6567547542603112,
+      "grad_norm": 0.4442101744900535,
+      "learning_rate": 4.932298681332641e-06,
+      "loss": 0.5834,
+      "step": 4192
+    },
+    {
+      "epoch": 1.6571499135589034,
+      "grad_norm": 0.43904506024703893,
+      "learning_rate": 4.932262462646532e-06,
+      "loss": 0.5628,
+      "step": 4193
+    },
+    {
+      "epoch": 1.6575450728574956,
+      "grad_norm": 0.45081395145304054,
+      "learning_rate": 4.932226234407973e-06,
+      "loss": 0.5597,
+      "step": 4194
+    },
+    {
+      "epoch": 1.6579402321560879,
+      "grad_norm": 0.45308321586467587,
+      "learning_rate": 4.932189996617106e-06,
+      "loss": 0.5424,
+      "step": 4195
+    },
+    {
+      "epoch": 1.6583353914546801,
+      "grad_norm": 0.43748276416781745,
+      "learning_rate": 4.932153749274074e-06,
+      "loss": 0.5478,
+      "step": 4196
+    },
+    {
+      "epoch": 1.6587305507532724,
+      "grad_norm": 0.4390093028892108,
+      "learning_rate": 4.932117492379019e-06,
+      "loss": 0.5576,
+      "step": 4197
+    },
+    {
+      "epoch": 1.6591257100518648,
+      "grad_norm": 0.43009631591048475,
+      "learning_rate": 4.932081225932084e-06,
+      "loss": 0.5461,
+      "step": 4198
+    },
+    {
+      "epoch": 1.659520869350457,
+      "grad_norm": 0.4421965633593312,
+      "learning_rate": 4.93204494993341e-06,
+      "loss": 0.5644,
+      "step": 4199
+    },
+    {
+      "epoch": 1.6599160286490493,
+      "grad_norm": 0.44435453925130375,
+      "learning_rate": 4.93200866438314e-06,
+      "loss": 0.5522,
+      "step": 4200
+    },
+    {
+      "epoch": 1.6603111879476415,
+      "grad_norm": 0.47458082371469706,
+      "learning_rate": 4.931972369281417e-06,
+      "loss": 0.5471,
+      "step": 4201
+    },
+    {
+      "epoch": 1.6607063472462338,
+      "grad_norm": 0.4301353638936425,
+      "learning_rate": 4.931936064628383e-06,
+      "loss": 0.5703,
+      "step": 4202
+    },
+    {
+      "epoch": 1.661101506544826,
+      "grad_norm": 0.4372380561131763,
+      "learning_rate": 4.931899750424182e-06,
+      "loss": 0.5683,
+      "step": 4203
+    },
+    {
+      "epoch": 1.6614966658434183,
+      "grad_norm": 0.46217752541691776,
+      "learning_rate": 4.931863426668955e-06,
+      "loss": 0.5559,
+      "step": 4204
+    },
+    {
+      "epoch": 1.6618918251420105,
+      "grad_norm": 0.4414697866376089,
+      "learning_rate": 4.931827093362844e-06,
+      "loss": 0.5533,
+      "step": 4205
+    },
+    {
+      "epoch": 1.6622869844406027,
+      "grad_norm": 0.44026690012982383,
+      "learning_rate": 4.931790750505994e-06,
+      "loss": 0.5473,
+      "step": 4206
+    },
+    {
+      "epoch": 1.662682143739195,
+      "grad_norm": 0.44908034456035184,
+      "learning_rate": 4.931754398098546e-06,
+      "loss": 0.5676,
+      "step": 4207
+    },
+    {
+      "epoch": 1.6630773030377872,
+      "grad_norm": 0.43506228540161684,
+      "learning_rate": 4.931718036140645e-06,
+      "loss": 0.56,
+      "step": 4208
+    },
+    {
+      "epoch": 1.6634724623363795,
+      "grad_norm": 0.4414844514234546,
+      "learning_rate": 4.9316816646324305e-06,
+      "loss": 0.5395,
+      "step": 4209
+    },
+    {
+      "epoch": 1.6638676216349717,
+      "grad_norm": 0.4353535377517495,
+      "learning_rate": 4.931645283574047e-06,
+      "loss": 0.5706,
+      "step": 4210
+    },
+    {
+      "epoch": 1.664262780933564,
+      "grad_norm": 0.4352610887294283,
+      "learning_rate": 4.931608892965638e-06,
+      "loss": 0.5529,
+      "step": 4211
+    },
+    {
+      "epoch": 1.6646579402321562,
+      "grad_norm": 0.48475567073134784,
+      "learning_rate": 4.931572492807346e-06,
+      "loss": 0.5691,
+      "step": 4212
+    },
+    {
+      "epoch": 1.6650530995307484,
+      "grad_norm": 0.43914859193682443,
+      "learning_rate": 4.931536083099313e-06,
+      "loss": 0.5447,
+      "step": 4213
+    },
+    {
+      "epoch": 1.6654482588293407,
+      "grad_norm": 0.42574696114957944,
+      "learning_rate": 4.931499663841683e-06,
+      "loss": 0.5506,
+      "step": 4214
+    },
+    {
+      "epoch": 1.665843418127933,
+      "grad_norm": 0.44003634839285843,
+      "learning_rate": 4.9314632350345995e-06,
+      "loss": 0.5714,
+      "step": 4215
+    },
+    {
+      "epoch": 1.6662385774265251,
+      "grad_norm": 0.44345794553567824,
+      "learning_rate": 4.931426796678204e-06,
+      "loss": 0.5525,
+      "step": 4216
+    },
+    {
+      "epoch": 1.6666337367251174,
+      "grad_norm": 0.44184543039510044,
+      "learning_rate": 4.9313903487726415e-06,
+      "loss": 0.5533,
+      "step": 4217
+    },
+    {
+      "epoch": 1.6670288960237096,
+      "grad_norm": 0.4482583566446144,
+      "learning_rate": 4.931353891318053e-06,
+      "loss": 0.5567,
+      "step": 4218
+    },
+    {
+      "epoch": 1.6674240553223019,
+      "grad_norm": 0.44359624620051485,
+      "learning_rate": 4.931317424314583e-06,
+      "loss": 0.5417,
+      "step": 4219
+    },
+    {
+      "epoch": 1.667819214620894,
+      "grad_norm": 0.45039139024204555,
+      "learning_rate": 4.9312809477623755e-06,
+      "loss": 0.5756,
+      "step": 4220
+    },
+    {
+      "epoch": 1.6682143739194863,
+      "grad_norm": 0.46242291351750414,
+      "learning_rate": 4.931244461661571e-06,
+      "loss": 0.5548,
+      "step": 4221
+    },
+    {
+      "epoch": 1.6686095332180786,
+      "grad_norm": 0.44761433940772233,
+      "learning_rate": 4.9312079660123165e-06,
+      "loss": 0.5545,
+      "step": 4222
+    },
+    {
+      "epoch": 1.6690046925166708,
+      "grad_norm": 0.42745307952513,
+      "learning_rate": 4.931171460814752e-06,
+      "loss": 0.5612,
+      "step": 4223
+    },
+    {
+      "epoch": 1.669399851815263,
+      "grad_norm": 0.4409759305327771,
+      "learning_rate": 4.9311349460690235e-06,
+      "loss": 0.5295,
+      "step": 4224
+    },
+    {
+      "epoch": 1.6697950111138553,
+      "grad_norm": 0.44041060145128974,
+      "learning_rate": 4.9310984217752725e-06,
+      "loss": 0.5807,
+      "step": 4225
+    },
+    {
+      "epoch": 1.6701901704124476,
+      "grad_norm": 0.45082681364452676,
+      "learning_rate": 4.931061887933643e-06,
+      "loss": 0.5525,
+      "step": 4226
+    },
+    {
+      "epoch": 1.6705853297110398,
+      "grad_norm": 0.4325179559967354,
+      "learning_rate": 4.931025344544279e-06,
+      "loss": 0.5614,
+      "step": 4227
+    },
+    {
+      "epoch": 1.670980489009632,
+      "grad_norm": 0.4491688766617088,
+      "learning_rate": 4.930988791607324e-06,
+      "loss": 0.5547,
+      "step": 4228
+    },
+    {
+      "epoch": 1.6713756483082243,
+      "grad_norm": 0.43251674883633673,
+      "learning_rate": 4.930952229122921e-06,
+      "loss": 0.5525,
+      "step": 4229
+    },
+    {
+      "epoch": 1.6717708076068165,
+      "grad_norm": 0.450173373250462,
+      "learning_rate": 4.930915657091213e-06,
+      "loss": 0.5539,
+      "step": 4230
+    },
+    {
+      "epoch": 1.6721659669054088,
+      "grad_norm": 0.452547822179787,
+      "learning_rate": 4.930879075512345e-06,
+      "loss": 0.5783,
+      "step": 4231
+    },
+    {
+      "epoch": 1.672561126204001,
+      "grad_norm": 0.456787048326293,
+      "learning_rate": 4.93084248438646e-06,
+      "loss": 0.5851,
+      "step": 4232
+    },
+    {
+      "epoch": 1.6729562855025932,
+      "grad_norm": 0.4398970129307844,
+      "learning_rate": 4.930805883713702e-06,
+      "loss": 0.5588,
+      "step": 4233
+    },
+    {
+      "epoch": 1.6733514448011855,
+      "grad_norm": 0.4344964137010482,
+      "learning_rate": 4.930769273494215e-06,
+      "loss": 0.5422,
+      "step": 4234
+    },
+    {
+      "epoch": 1.6737466040997777,
+      "grad_norm": 0.4480357286208341,
+      "learning_rate": 4.930732653728141e-06,
+      "loss": 0.5719,
+      "step": 4235
+    },
+    {
+      "epoch": 1.67414176339837,
+      "grad_norm": 0.4468533257133995,
+      "learning_rate": 4.930696024415626e-06,
+      "loss": 0.5706,
+      "step": 4236
+    },
+    {
+      "epoch": 1.6745369226969622,
+      "grad_norm": 0.43589038296389493,
+      "learning_rate": 4.930659385556813e-06,
+      "loss": 0.5718,
+      "step": 4237
+    },
+    {
+      "epoch": 1.6749320819955544,
+      "grad_norm": 0.46213277997599467,
+      "learning_rate": 4.9306227371518455e-06,
+      "loss": 0.543,
+      "step": 4238
+    },
+    {
+      "epoch": 1.6753272412941467,
+      "grad_norm": 0.4418860283426867,
+      "learning_rate": 4.930586079200869e-06,
+      "loss": 0.555,
+      "step": 4239
+    },
+    {
+      "epoch": 1.675722400592739,
+      "grad_norm": 0.41996717679739026,
+      "learning_rate": 4.930549411704025e-06,
+      "loss": 0.5357,
+      "step": 4240
+    },
+    {
+      "epoch": 1.6761175598913312,
+      "grad_norm": 0.4575919872171913,
+      "learning_rate": 4.930512734661459e-06,
+      "loss": 0.5545,
+      "step": 4241
+    },
+    {
+      "epoch": 1.6765127191899234,
+      "grad_norm": 0.44970217578452903,
+      "learning_rate": 4.930476048073316e-06,
+      "loss": 0.5634,
+      "step": 4242
+    },
+    {
+      "epoch": 1.6769078784885156,
+      "grad_norm": 0.4436734001641958,
+      "learning_rate": 4.930439351939738e-06,
+      "loss": 0.5848,
+      "step": 4243
+    },
+    {
+      "epoch": 1.6773030377871079,
+      "grad_norm": 0.43230787717115343,
+      "learning_rate": 4.93040264626087e-06,
+      "loss": 0.5758,
+      "step": 4244
+    },
+    {
+      "epoch": 1.6776981970857001,
+      "grad_norm": 0.45704860376232215,
+      "learning_rate": 4.9303659310368565e-06,
+      "loss": 0.5644,
+      "step": 4245
+    },
+    {
+      "epoch": 1.6780933563842924,
+      "grad_norm": 0.458499629507116,
+      "learning_rate": 4.930329206267841e-06,
+      "loss": 0.5714,
+      "step": 4246
+    },
+    {
+      "epoch": 1.6784885156828846,
+      "grad_norm": 0.4436057764210825,
+      "learning_rate": 4.930292471953969e-06,
+      "loss": 0.5512,
+      "step": 4247
+    },
+    {
+      "epoch": 1.6788836749814768,
+      "grad_norm": 0.4322151502298006,
+      "learning_rate": 4.930255728095383e-06,
+      "loss": 0.5475,
+      "step": 4248
+    },
+    {
+      "epoch": 1.679278834280069,
+      "grad_norm": 0.4549937231291737,
+      "learning_rate": 4.930218974692229e-06,
+      "loss": 0.5568,
+      "step": 4249
+    },
+    {
+      "epoch": 1.6796739935786613,
+      "grad_norm": 0.44063968344614446,
+      "learning_rate": 4.930182211744649e-06,
+      "loss": 0.562,
+      "step": 4250
+    },
+    {
+      "epoch": 1.6800691528772536,
+      "grad_norm": 0.4635898234904489,
+      "learning_rate": 4.930145439252791e-06,
+      "loss": 0.566,
+      "step": 4251
+    },
+    {
+      "epoch": 1.6804643121758458,
+      "grad_norm": 0.4347065121603174,
+      "learning_rate": 4.930108657216796e-06,
+      "loss": 0.5579,
+      "step": 4252
+    },
+    {
+      "epoch": 1.680859471474438,
+      "grad_norm": 0.45191051205485905,
+      "learning_rate": 4.930071865636811e-06,
+      "loss": 0.5616,
+      "step": 4253
+    },
+    {
+      "epoch": 1.6812546307730303,
+      "grad_norm": 0.5003942515624423,
+      "learning_rate": 4.930035064512979e-06,
+      "loss": 0.5544,
+      "step": 4254
+    },
+    {
+      "epoch": 1.6816497900716225,
+      "grad_norm": 0.43280130558001995,
+      "learning_rate": 4.929998253845444e-06,
+      "loss": 0.5839,
+      "step": 4255
+    },
+    {
+      "epoch": 1.6820449493702148,
+      "grad_norm": 0.4347329458274132,
+      "learning_rate": 4.929961433634352e-06,
+      "loss": 0.5512,
+      "step": 4256
+    },
+    {
+      "epoch": 1.682440108668807,
+      "grad_norm": 0.4595377167928187,
+      "learning_rate": 4.9299246038798474e-06,
+      "loss": 0.5797,
+      "step": 4257
+    },
+    {
+      "epoch": 1.6828352679673992,
+      "grad_norm": 0.4542171341695038,
+      "learning_rate": 4.9298877645820735e-06,
+      "loss": 0.5421,
+      "step": 4258
+    },
+    {
+      "epoch": 1.6832304272659915,
+      "grad_norm": 0.43845775532831804,
+      "learning_rate": 4.929850915741177e-06,
+      "loss": 0.565,
+      "step": 4259
+    },
+    {
+      "epoch": 1.6836255865645837,
+      "grad_norm": 0.44402420830903166,
+      "learning_rate": 4.929814057357301e-06,
+      "loss": 0.5605,
+      "step": 4260
+    },
+    {
+      "epoch": 1.684020745863176,
+      "grad_norm": 0.44396077168291875,
+      "learning_rate": 4.929777189430591e-06,
+      "loss": 0.5694,
+      "step": 4261
+    },
+    {
+      "epoch": 1.6844159051617682,
+      "grad_norm": 0.4413223103845851,
+      "learning_rate": 4.929740311961192e-06,
+      "loss": 0.5681,
+      "step": 4262
+    },
+    {
+      "epoch": 1.6848110644603604,
+      "grad_norm": 0.4432712724622185,
+      "learning_rate": 4.929703424949248e-06,
+      "loss": 0.5791,
+      "step": 4263
+    },
+    {
+      "epoch": 1.6852062237589527,
+      "grad_norm": 0.4291806701241977,
+      "learning_rate": 4.929666528394904e-06,
+      "loss": 0.5608,
+      "step": 4264
+    },
+    {
+      "epoch": 1.685601383057545,
+      "grad_norm": 0.44175554248806165,
+      "learning_rate": 4.929629622298307e-06,
+      "loss": 0.5367,
+      "step": 4265
+    },
+    {
+      "epoch": 1.6859965423561372,
+      "grad_norm": 0.43417237826237876,
+      "learning_rate": 4.929592706659599e-06,
+      "loss": 0.5621,
+      "step": 4266
+    },
+    {
+      "epoch": 1.6863917016547294,
+      "grad_norm": 0.436619782617058,
+      "learning_rate": 4.929555781478925e-06,
+      "loss": 0.556,
+      "step": 4267
+    },
+    {
+      "epoch": 1.6867868609533216,
+      "grad_norm": 0.4412555558446154,
+      "learning_rate": 4.9295188467564326e-06,
+      "loss": 0.5549,
+      "step": 4268
+    },
+    {
+      "epoch": 1.687182020251914,
+      "grad_norm": 0.4429817647687938,
+      "learning_rate": 4.929481902492265e-06,
+      "loss": 0.5451,
+      "step": 4269
+    },
+    {
+      "epoch": 1.6875771795505063,
+      "grad_norm": 0.42710204000708135,
+      "learning_rate": 4.929444948686568e-06,
+      "loss": 0.5574,
+      "step": 4270
+    },
+    {
+      "epoch": 1.6879723388490986,
+      "grad_norm": 0.4345670015513589,
+      "learning_rate": 4.929407985339486e-06,
+      "loss": 0.573,
+      "step": 4271
+    },
+    {
+      "epoch": 1.6883674981476908,
+      "grad_norm": 0.42754387278545575,
+      "learning_rate": 4.929371012451165e-06,
+      "loss": 0.5504,
+      "step": 4272
+    },
+    {
+      "epoch": 1.688762657446283,
+      "grad_norm": 0.44346425482357044,
+      "learning_rate": 4.9293340300217505e-06,
+      "loss": 0.566,
+      "step": 4273
+    },
+    {
+      "epoch": 1.6891578167448753,
+      "grad_norm": 0.43395132844534345,
+      "learning_rate": 4.929297038051386e-06,
+      "loss": 0.5751,
+      "step": 4274
+    },
+    {
+      "epoch": 1.6895529760434675,
+      "grad_norm": 0.4289911868889898,
+      "learning_rate": 4.929260036540218e-06,
+      "loss": 0.5545,
+      "step": 4275
+    },
+    {
+      "epoch": 1.6899481353420598,
+      "grad_norm": 0.4299481945107919,
+      "learning_rate": 4.929223025488393e-06,
+      "loss": 0.5655,
+      "step": 4276
+    },
+    {
+      "epoch": 1.690343294640652,
+      "grad_norm": 0.4315808805682797,
+      "learning_rate": 4.929186004896054e-06,
+      "loss": 0.5586,
+      "step": 4277
+    },
+    {
+      "epoch": 1.6907384539392443,
+      "grad_norm": 0.43737402235911826,
+      "learning_rate": 4.929148974763347e-06,
+      "loss": 0.5677,
+      "step": 4278
+    },
+    {
+      "epoch": 1.6911336132378365,
+      "grad_norm": 0.44163862598713893,
+      "learning_rate": 4.929111935090418e-06,
+      "loss": 0.5484,
+      "step": 4279
+    },
+    {
+      "epoch": 1.6915287725364287,
+      "grad_norm": 0.44050203475372246,
+      "learning_rate": 4.929074885877414e-06,
+      "loss": 0.5639,
+      "step": 4280
+    },
+    {
+      "epoch": 1.691923931835021,
+      "grad_norm": 0.44036371463743945,
+      "learning_rate": 4.929037827124477e-06,
+      "loss": 0.5532,
+      "step": 4281
+    },
+    {
+      "epoch": 1.6923190911336132,
+      "grad_norm": 0.42541252276330227,
+      "learning_rate": 4.929000758831755e-06,
+      "loss": 0.5541,
+      "step": 4282
+    },
+    {
+      "epoch": 1.6927142504322055,
+      "grad_norm": 0.45077617747089965,
+      "learning_rate": 4.928963680999393e-06,
+      "loss": 0.5728,
+      "step": 4283
+    },
+    {
+      "epoch": 1.6931094097307977,
+      "grad_norm": 0.43862795202138594,
+      "learning_rate": 4.928926593627537e-06,
+      "loss": 0.5609,
+      "step": 4284
+    },
+    {
+      "epoch": 1.69350456902939,
+      "grad_norm": 0.4355391950216865,
+      "learning_rate": 4.928889496716331e-06,
+      "loss": 0.5577,
+      "step": 4285
+    },
+    {
+      "epoch": 1.6938997283279822,
+      "grad_norm": 0.449424153003478,
+      "learning_rate": 4.928852390265923e-06,
+      "loss": 0.5589,
+      "step": 4286
+    },
+    {
+      "epoch": 1.6942948876265744,
+      "grad_norm": 0.42444603371479195,
+      "learning_rate": 4.928815274276458e-06,
+      "loss": 0.5542,
+      "step": 4287
+    },
+    {
+      "epoch": 1.6946900469251667,
+      "grad_norm": 0.4374300606831901,
+      "learning_rate": 4.928778148748081e-06,
+      "loss": 0.5768,
+      "step": 4288
+    },
+    {
+      "epoch": 1.695085206223759,
+      "grad_norm": 0.4267602747759363,
+      "learning_rate": 4.928741013680939e-06,
+      "loss": 0.5606,
+      "step": 4289
+    },
+    {
+      "epoch": 1.6954803655223512,
+      "grad_norm": 0.4366846729132836,
+      "learning_rate": 4.928703869075176e-06,
+      "loss": 0.5498,
+      "step": 4290
+    },
+    {
+      "epoch": 1.6958755248209434,
+      "grad_norm": 0.4435134184494798,
+      "learning_rate": 4.92866671493094e-06,
+      "loss": 0.5485,
+      "step": 4291
+    },
+    {
+      "epoch": 1.6962706841195359,
+      "grad_norm": 0.42497901408157157,
+      "learning_rate": 4.928629551248375e-06,
+      "loss": 0.5383,
+      "step": 4292
+    },
+    {
+      "epoch": 1.696665843418128,
+      "grad_norm": 0.4443410896472178,
+      "learning_rate": 4.928592378027628e-06,
+      "loss": 0.543,
+      "step": 4293
+    },
+    {
+      "epoch": 1.6970610027167203,
+      "grad_norm": 0.4546919285525924,
+      "learning_rate": 4.928555195268845e-06,
+      "loss": 0.5598,
+      "step": 4294
+    },
+    {
+      "epoch": 1.6974561620153126,
+      "grad_norm": 0.4327018959048797,
+      "learning_rate": 4.928518002972172e-06,
+      "loss": 0.5607,
+      "step": 4295
+    },
+    {
+      "epoch": 1.6978513213139048,
+      "grad_norm": 0.43177140667053937,
+      "learning_rate": 4.928480801137755e-06,
+      "loss": 0.5468,
+      "step": 4296
+    },
+    {
+      "epoch": 1.698246480612497,
+      "grad_norm": 0.43141416580935815,
+      "learning_rate": 4.92844358976574e-06,
+      "loss": 0.5663,
+      "step": 4297
+    },
+    {
+      "epoch": 1.6986416399110893,
+      "grad_norm": 0.4787754721973709,
+      "learning_rate": 4.928406368856273e-06,
+      "loss": 0.5636,
+      "step": 4298
+    },
+    {
+      "epoch": 1.6990367992096815,
+      "grad_norm": 0.442334169803762,
+      "learning_rate": 4.9283691384095e-06,
+      "loss": 0.5823,
+      "step": 4299
+    },
+    {
+      "epoch": 1.6994319585082738,
+      "grad_norm": 0.432453303132636,
+      "learning_rate": 4.928331898425568e-06,
+      "loss": 0.5589,
+      "step": 4300
+    },
+    {
+      "epoch": 1.699827117806866,
+      "grad_norm": 0.4490638732729113,
+      "learning_rate": 4.9282946489046235e-06,
+      "loss": 0.5754,
+      "step": 4301
+    },
+    {
+      "epoch": 1.7002222771054583,
+      "grad_norm": 0.4853856757096015,
+      "learning_rate": 4.928257389846812e-06,
+      "loss": 0.5503,
+      "step": 4302
+    },
+    {
+      "epoch": 1.7006174364040505,
+      "grad_norm": 0.42812505775715426,
+      "learning_rate": 4.92822012125228e-06,
+      "loss": 0.5477,
+      "step": 4303
+    },
+    {
+      "epoch": 1.7010125957026427,
+      "grad_norm": 0.43886286312335837,
+      "learning_rate": 4.928182843121173e-06,
+      "loss": 0.5716,
+      "step": 4304
+    },
+    {
+      "epoch": 1.701407755001235,
+      "grad_norm": 0.4523032752577509,
+      "learning_rate": 4.928145555453638e-06,
+      "loss": 0.5509,
+      "step": 4305
+    },
+    {
+      "epoch": 1.7018029142998272,
+      "grad_norm": 0.4517070813549554,
+      "learning_rate": 4.928108258249823e-06,
+      "loss": 0.566,
+      "step": 4306
+    },
+    {
+      "epoch": 1.7021980735984195,
+      "grad_norm": 0.5646470946712949,
+      "learning_rate": 4.928070951509873e-06,
+      "loss": 0.575,
+      "step": 4307
+    },
+    {
+      "epoch": 1.7025932328970117,
+      "grad_norm": 0.42398116604836683,
+      "learning_rate": 4.928033635233934e-06,
+      "loss": 0.5449,
+      "step": 4308
+    },
+    {
+      "epoch": 1.702988392195604,
+      "grad_norm": 0.45125784631407473,
+      "learning_rate": 4.927996309422154e-06,
+      "loss": 0.555,
+      "step": 4309
+    },
+    {
+      "epoch": 1.7033835514941962,
+      "grad_norm": 0.4516857652827014,
+      "learning_rate": 4.927958974074678e-06,
+      "loss": 0.5358,
+      "step": 4310
+    },
+    {
+      "epoch": 1.7037787107927884,
+      "grad_norm": 0.4486541477736022,
+      "learning_rate": 4.927921629191654e-06,
+      "loss": 0.5706,
+      "step": 4311
+    },
+    {
+      "epoch": 1.7041738700913807,
+      "grad_norm": 0.5504549124395234,
+      "learning_rate": 4.927884274773229e-06,
+      "loss": 0.5499,
+      "step": 4312
+    },
+    {
+      "epoch": 1.704569029389973,
+      "grad_norm": 0.4377036298330416,
+      "learning_rate": 4.927846910819547e-06,
+      "loss": 0.5624,
+      "step": 4313
+    },
+    {
+      "epoch": 1.7049641886885651,
+      "grad_norm": 0.44142723756552443,
+      "learning_rate": 4.9278095373307586e-06,
+      "loss": 0.5641,
+      "step": 4314
+    },
+    {
+      "epoch": 1.7053593479871574,
+      "grad_norm": 0.4305246703883127,
+      "learning_rate": 4.927772154307007e-06,
+      "loss": 0.5584,
+      "step": 4315
+    },
+    {
+      "epoch": 1.7057545072857496,
+      "grad_norm": 0.42333610423419793,
+      "learning_rate": 4.927734761748441e-06,
+      "loss": 0.5662,
+      "step": 4316
+    },
+    {
+      "epoch": 1.7061496665843419,
+      "grad_norm": 0.43713955439883345,
+      "learning_rate": 4.927697359655208e-06,
+      "loss": 0.5574,
+      "step": 4317
+    },
+    {
+      "epoch": 1.706544825882934,
+      "grad_norm": 0.4350887115565443,
+      "learning_rate": 4.927659948027453e-06,
+      "loss": 0.5484,
+      "step": 4318
+    },
+    {
+      "epoch": 1.7069399851815263,
+      "grad_norm": 0.4305661898067747,
+      "learning_rate": 4.927622526865324e-06,
+      "loss": 0.5721,
+      "step": 4319
+    },
+    {
+      "epoch": 1.7073351444801186,
+      "grad_norm": 0.4237425713774883,
+      "learning_rate": 4.927585096168967e-06,
+      "loss": 0.5385,
+      "step": 4320
+    },
+    {
+      "epoch": 1.7077303037787108,
+      "grad_norm": 0.42768772713800246,
+      "learning_rate": 4.9275476559385316e-06,
+      "loss": 0.5615,
+      "step": 4321
+    },
+    {
+      "epoch": 1.708125463077303,
+      "grad_norm": 0.45101823860302,
+      "learning_rate": 4.927510206174162e-06,
+      "loss": 0.5777,
+      "step": 4322
+    },
+    {
+      "epoch": 1.7085206223758953,
+      "grad_norm": 0.45186205819480785,
+      "learning_rate": 4.927472746876007e-06,
+      "loss": 0.55,
+      "step": 4323
+    },
+    {
+      "epoch": 1.7089157816744875,
+      "grad_norm": 0.4632522948748882,
+      "learning_rate": 4.9274352780442125e-06,
+      "loss": 0.5719,
+      "step": 4324
+    },
+    {
+      "epoch": 1.7093109409730798,
+      "grad_norm": 0.43100101608453223,
+      "learning_rate": 4.927397799678927e-06,
+      "loss": 0.5549,
+      "step": 4325
+    },
+    {
+      "epoch": 1.709706100271672,
+      "grad_norm": 0.434763481025286,
+      "learning_rate": 4.927360311780296e-06,
+      "loss": 0.566,
+      "step": 4326
+    },
+    {
+      "epoch": 1.7101012595702643,
+      "grad_norm": 0.43309030464048226,
+      "learning_rate": 4.927322814348468e-06,
+      "loss": 0.5345,
+      "step": 4327
+    },
+    {
+      "epoch": 1.7104964188688565,
+      "grad_norm": 0.48013463816215374,
+      "learning_rate": 4.92728530738359e-06,
+      "loss": 0.5912,
+      "step": 4328
+    },
+    {
+      "epoch": 1.7108915781674487,
+      "grad_norm": 0.47034903875978573,
+      "learning_rate": 4.927247790885809e-06,
+      "loss": 0.5778,
+      "step": 4329
+    },
+    {
+      "epoch": 1.711286737466041,
+      "grad_norm": 0.4278937802281345,
+      "learning_rate": 4.927210264855274e-06,
+      "loss": 0.5597,
+      "step": 4330
+    },
+    {
+      "epoch": 1.7116818967646332,
+      "grad_norm": 0.43626768378812447,
+      "learning_rate": 4.927172729292129e-06,
+      "loss": 0.5651,
+      "step": 4331
+    },
+    {
+      "epoch": 1.7120770560632255,
+      "grad_norm": 0.4465422521558346,
+      "learning_rate": 4.927135184196524e-06,
+      "loss": 0.5756,
+      "step": 4332
+    },
+    {
+      "epoch": 1.7124722153618177,
+      "grad_norm": 0.44974916569628687,
+      "learning_rate": 4.927097629568606e-06,
+      "loss": 0.568,
+      "step": 4333
+    },
+    {
+      "epoch": 1.71286737466041,
+      "grad_norm": 0.4472334023755386,
+      "learning_rate": 4.927060065408522e-06,
+      "loss": 0.5797,
+      "step": 4334
+    },
+    {
+      "epoch": 1.7132625339590022,
+      "grad_norm": 0.4481801045998597,
+      "learning_rate": 4.9270224917164204e-06,
+      "loss": 0.5675,
+      "step": 4335
+    },
+    {
+      "epoch": 1.7136576932575944,
+      "grad_norm": 0.4549341973795611,
+      "learning_rate": 4.926984908492448e-06,
+      "loss": 0.5803,
+      "step": 4336
+    },
+    {
+      "epoch": 1.7140528525561867,
+      "grad_norm": 0.4359842869982814,
+      "learning_rate": 4.9269473157367535e-06,
+      "loss": 0.5499,
+      "step": 4337
+    },
+    {
+      "epoch": 1.714448011854779,
+      "grad_norm": 0.5200059350206986,
+      "learning_rate": 4.926909713449482e-06,
+      "loss": 0.5577,
+      "step": 4338
+    },
+    {
+      "epoch": 1.7148431711533711,
+      "grad_norm": 0.4689154411180137,
+      "learning_rate": 4.926872101630784e-06,
+      "loss": 0.5545,
+      "step": 4339
+    },
+    {
+      "epoch": 1.7152383304519634,
+      "grad_norm": 0.45440101006285427,
+      "learning_rate": 4.9268344802808055e-06,
+      "loss": 0.5633,
+      "step": 4340
+    },
+    {
+      "epoch": 1.7156334897505556,
+      "grad_norm": 0.45895219511156227,
+      "learning_rate": 4.926796849399694e-06,
+      "loss": 0.5515,
+      "step": 4341
+    },
+    {
+      "epoch": 1.7160286490491479,
+      "grad_norm": 0.43433747556834423,
+      "learning_rate": 4.9267592089876e-06,
+      "loss": 0.55,
+      "step": 4342
+    },
+    {
+      "epoch": 1.71642380834774,
+      "grad_norm": 0.4338247101488299,
+      "learning_rate": 4.926721559044668e-06,
+      "loss": 0.5563,
+      "step": 4343
+    },
+    {
+      "epoch": 1.7168189676463324,
+      "grad_norm": 0.45944141139935163,
+      "learning_rate": 4.926683899571048e-06,
+      "loss": 0.5357,
+      "step": 4344
+    },
+    {
+      "epoch": 1.7172141269449246,
+      "grad_norm": 0.4264619359745203,
+      "learning_rate": 4.9266462305668876e-06,
+      "loss": 0.5385,
+      "step": 4345
+    },
+    {
+      "epoch": 1.7176092862435168,
+      "grad_norm": 0.43636381830305015,
+      "learning_rate": 4.926608552032334e-06,
+      "loss": 0.5781,
+      "step": 4346
+    },
+    {
+      "epoch": 1.718004445542109,
+      "grad_norm": 0.4488377673647985,
+      "learning_rate": 4.926570863967535e-06,
+      "loss": 0.559,
+      "step": 4347
+    },
+    {
+      "epoch": 1.7183996048407013,
+      "grad_norm": 0.44153153641448994,
+      "learning_rate": 4.926533166372639e-06,
+      "loss": 0.5704,
+      "step": 4348
+    },
+    {
+      "epoch": 1.7187947641392936,
+      "grad_norm": 0.43552915303217893,
+      "learning_rate": 4.926495459247795e-06,
+      "loss": 0.5564,
+      "step": 4349
+    },
+    {
+      "epoch": 1.7191899234378858,
+      "grad_norm": 0.4306421081394027,
+      "learning_rate": 4.9264577425931505e-06,
+      "loss": 0.5668,
+      "step": 4350
+    },
+    {
+      "epoch": 1.719585082736478,
+      "grad_norm": 0.43170183294012887,
+      "learning_rate": 4.926420016408852e-06,
+      "loss": 0.56,
+      "step": 4351
+    },
+    {
+      "epoch": 1.7199802420350703,
+      "grad_norm": 0.44310661756844827,
+      "learning_rate": 4.92638228069505e-06,
+      "loss": 0.5444,
+      "step": 4352
+    },
+    {
+      "epoch": 1.7203754013336625,
+      "grad_norm": 0.5473701951430664,
+      "learning_rate": 4.926344535451892e-06,
+      "loss": 0.5518,
+      "step": 4353
+    },
+    {
+      "epoch": 1.7207705606322548,
+      "grad_norm": 0.45773670023142704,
+      "learning_rate": 4.926306780679526e-06,
+      "loss": 0.5879,
+      "step": 4354
+    },
+    {
+      "epoch": 1.721165719930847,
+      "grad_norm": 0.4470158656152022,
+      "learning_rate": 4.926269016378099e-06,
+      "loss": 0.5574,
+      "step": 4355
+    },
+    {
+      "epoch": 1.7215608792294392,
+      "grad_norm": 0.4326444239390642,
+      "learning_rate": 4.926231242547762e-06,
+      "loss": 0.5575,
+      "step": 4356
+    },
+    {
+      "epoch": 1.7219560385280315,
+      "grad_norm": 0.43096022881772944,
+      "learning_rate": 4.926193459188662e-06,
+      "loss": 0.5237,
+      "step": 4357
+    },
+    {
+      "epoch": 1.7223511978266237,
+      "grad_norm": 0.4335631980043337,
+      "learning_rate": 4.926155666300947e-06,
+      "loss": 0.5508,
+      "step": 4358
+    },
+    {
+      "epoch": 1.722746357125216,
+      "grad_norm": 0.43344519231600975,
+      "learning_rate": 4.926117863884765e-06,
+      "loss": 0.5578,
+      "step": 4359
+    },
+    {
+      "epoch": 1.7231415164238082,
+      "grad_norm": 0.439357665476649,
+      "learning_rate": 4.926080051940267e-06,
+      "loss": 0.5677,
+      "step": 4360
+    },
+    {
+      "epoch": 1.7235366757224004,
+      "grad_norm": 0.45079399416037436,
+      "learning_rate": 4.926042230467598e-06,
+      "loss": 0.5653,
+      "step": 4361
+    },
+    {
+      "epoch": 1.7239318350209927,
+      "grad_norm": 0.41470626875209715,
+      "learning_rate": 4.9260043994669094e-06,
+      "loss": 0.5375,
+      "step": 4362
+    },
+    {
+      "epoch": 1.7243269943195851,
+      "grad_norm": 0.4313994825808373,
+      "learning_rate": 4.925966558938348e-06,
+      "loss": 0.5692,
+      "step": 4363
+    },
+    {
+      "epoch": 1.7247221536181774,
+      "grad_norm": 0.4321560662738567,
+      "learning_rate": 4.925928708882064e-06,
+      "loss": 0.5467,
+      "step": 4364
+    },
+    {
+      "epoch": 1.7251173129167696,
+      "grad_norm": 0.43615051233389873,
+      "learning_rate": 4.925890849298204e-06,
+      "loss": 0.561,
+      "step": 4365
+    },
+    {
+      "epoch": 1.7255124722153619,
+      "grad_norm": 0.4386066095911232,
+      "learning_rate": 4.925852980186918e-06,
+      "loss": 0.5702,
+      "step": 4366
+    },
+    {
+      "epoch": 1.725907631513954,
+      "grad_norm": 0.4294008469837109,
+      "learning_rate": 4.9258151015483555e-06,
+      "loss": 0.5611,
+      "step": 4367
+    },
+    {
+      "epoch": 1.7263027908125463,
+      "grad_norm": 0.43462199012351793,
+      "learning_rate": 4.925777213382663e-06,
+      "loss": 0.571,
+      "step": 4368
+    },
+    {
+      "epoch": 1.7266979501111386,
+      "grad_norm": 0.45940417820864915,
+      "learning_rate": 4.925739315689991e-06,
+      "loss": 0.576,
+      "step": 4369
+    },
+    {
+      "epoch": 1.7270931094097308,
+      "grad_norm": 0.43409099796596096,
+      "learning_rate": 4.925701408470489e-06,
+      "loss": 0.5583,
+      "step": 4370
+    },
+    {
+      "epoch": 1.727488268708323,
+      "grad_norm": 0.4364436014175337,
+      "learning_rate": 4.925663491724304e-06,
+      "loss": 0.5465,
+      "step": 4371
+    },
+    {
+      "epoch": 1.7278834280069153,
+      "grad_norm": 0.4242367032389472,
+      "learning_rate": 4.9256255654515865e-06,
+      "loss": 0.5566,
+      "step": 4372
+    },
+    {
+      "epoch": 1.7282785873055075,
+      "grad_norm": 0.4348950743141649,
+      "learning_rate": 4.925587629652483e-06,
+      "loss": 0.5524,
+      "step": 4373
+    },
+    {
+      "epoch": 1.7286737466040998,
+      "grad_norm": 0.440868038863271,
+      "learning_rate": 4.925549684327145e-06,
+      "loss": 0.5432,
+      "step": 4374
+    },
+    {
+      "epoch": 1.729068905902692,
+      "grad_norm": 0.4594372909416891,
+      "learning_rate": 4.925511729475722e-06,
+      "loss": 0.573,
+      "step": 4375
+    },
+    {
+      "epoch": 1.7294640652012843,
+      "grad_norm": 0.442038328983773,
+      "learning_rate": 4.92547376509836e-06,
+      "loss": 0.5436,
+      "step": 4376
+    },
+    {
+      "epoch": 1.7298592244998765,
+      "grad_norm": 0.4488082560018164,
+      "learning_rate": 4.925435791195211e-06,
+      "loss": 0.5382,
+      "step": 4377
+    },
+    {
+      "epoch": 1.7302543837984687,
+      "grad_norm": 0.4451552451369125,
+      "learning_rate": 4.925397807766422e-06,
+      "loss": 0.5708,
+      "step": 4378
+    },
+    {
+      "epoch": 1.730649543097061,
+      "grad_norm": 0.44829170627702375,
+      "learning_rate": 4.925359814812144e-06,
+      "loss": 0.5566,
+      "step": 4379
+    },
+    {
+      "epoch": 1.7310447023956532,
+      "grad_norm": 0.45319114553972123,
+      "learning_rate": 4.925321812332526e-06,
+      "loss": 0.5812,
+      "step": 4380
+    },
+    {
+      "epoch": 1.7314398616942455,
+      "grad_norm": 0.4730466509935319,
+      "learning_rate": 4.925283800327715e-06,
+      "loss": 0.5595,
+      "step": 4381
+    },
+    {
+      "epoch": 1.7318350209928377,
+      "grad_norm": 0.44863564694406577,
+      "learning_rate": 4.925245778797863e-06,
+      "loss": 0.5537,
+      "step": 4382
+    },
+    {
+      "epoch": 1.73223018029143,
+      "grad_norm": 0.449712427848464,
+      "learning_rate": 4.925207747743118e-06,
+      "loss": 0.5548,
+      "step": 4383
+    },
+    {
+      "epoch": 1.7326253395900222,
+      "grad_norm": 0.4353187498862786,
+      "learning_rate": 4.925169707163629e-06,
+      "loss": 0.5699,
+      "step": 4384
+    },
+    {
+      "epoch": 1.7330204988886144,
+      "grad_norm": 0.4552613973646662,
+      "learning_rate": 4.925131657059548e-06,
+      "loss": 0.5589,
+      "step": 4385
+    },
+    {
+      "epoch": 1.7334156581872067,
+      "grad_norm": 0.43804313299916375,
+      "learning_rate": 4.925093597431021e-06,
+      "loss": 0.5665,
+      "step": 4386
+    },
+    {
+      "epoch": 1.7338108174857991,
+      "grad_norm": 0.44336856585398643,
+      "learning_rate": 4.925055528278199e-06,
+      "loss": 0.5691,
+      "step": 4387
+    },
+    {
+      "epoch": 1.7342059767843914,
+      "grad_norm": 0.4453678495451275,
+      "learning_rate": 4.925017449601231e-06,
+      "loss": 0.5799,
+      "step": 4388
+    },
+    {
+      "epoch": 1.7346011360829836,
+      "grad_norm": 0.43883849979746115,
+      "learning_rate": 4.924979361400268e-06,
+      "loss": 0.5714,
+      "step": 4389
+    },
+    {
+      "epoch": 1.7349962953815758,
+      "grad_norm": 0.4301004660209356,
+      "learning_rate": 4.924941263675458e-06,
+      "loss": 0.5542,
+      "step": 4390
+    },
+    {
+      "epoch": 1.735391454680168,
+      "grad_norm": 0.4371671225872987,
+      "learning_rate": 4.924903156426952e-06,
+      "loss": 0.5634,
+      "step": 4391
+    },
+    {
+      "epoch": 1.7357866139787603,
+      "grad_norm": 0.4467984256797525,
+      "learning_rate": 4.924865039654898e-06,
+      "loss": 0.5741,
+      "step": 4392
+    },
+    {
+      "epoch": 1.7361817732773526,
+      "grad_norm": 0.4373830450072716,
+      "learning_rate": 4.9248269133594464e-06,
+      "loss": 0.5546,
+      "step": 4393
+    },
+    {
+      "epoch": 1.7365769325759448,
+      "grad_norm": 0.44157114771440137,
+      "learning_rate": 4.924788777540748e-06,
+      "loss": 0.5758,
+      "step": 4394
+    },
+    {
+      "epoch": 1.736972091874537,
+      "grad_norm": 0.42686715929933167,
+      "learning_rate": 4.9247506321989514e-06,
+      "loss": 0.5612,
+      "step": 4395
+    },
+    {
+      "epoch": 1.7373672511731293,
+      "grad_norm": 0.429547791413384,
+      "learning_rate": 4.924712477334206e-06,
+      "loss": 0.5474,
+      "step": 4396
+    },
+    {
+      "epoch": 1.7377624104717215,
+      "grad_norm": 0.4456343541323995,
+      "learning_rate": 4.924674312946663e-06,
+      "loss": 0.5648,
+      "step": 4397
+    },
+    {
+      "epoch": 1.7381575697703138,
+      "grad_norm": 0.4557318656179386,
+      "learning_rate": 4.924636139036472e-06,
+      "loss": 0.5811,
+      "step": 4398
+    },
+    {
+      "epoch": 1.738552729068906,
+      "grad_norm": 0.42877428492133096,
+      "learning_rate": 4.924597955603782e-06,
+      "loss": 0.5719,
+      "step": 4399
+    },
+    {
+      "epoch": 1.7389478883674983,
+      "grad_norm": 0.44201823103262,
+      "learning_rate": 4.924559762648744e-06,
+      "loss": 0.5527,
+      "step": 4400
+    },
+    {
+      "epoch": 1.7393430476660905,
+      "grad_norm": 0.4406878460734057,
+      "learning_rate": 4.924521560171507e-06,
+      "loss": 0.5506,
+      "step": 4401
+    },
+    {
+      "epoch": 1.7397382069646827,
+      "grad_norm": 0.4238997986809783,
+      "learning_rate": 4.924483348172222e-06,
+      "loss": 0.555,
+      "step": 4402
+    },
+    {
+      "epoch": 1.740133366263275,
+      "grad_norm": 0.444786923090035,
+      "learning_rate": 4.924445126651038e-06,
+      "loss": 0.5585,
+      "step": 4403
+    },
+    {
+      "epoch": 1.7405285255618672,
+      "grad_norm": 0.42484925220727704,
+      "learning_rate": 4.924406895608106e-06,
+      "loss": 0.5482,
+      "step": 4404
+    },
+    {
+      "epoch": 1.7409236848604595,
+      "grad_norm": 0.44349133709535676,
+      "learning_rate": 4.924368655043577e-06,
+      "loss": 0.5573,
+      "step": 4405
+    },
+    {
+      "epoch": 1.7413188441590517,
+      "grad_norm": 0.4227263535085141,
+      "learning_rate": 4.924330404957599e-06,
+      "loss": 0.5465,
+      "step": 4406
+    },
+    {
+      "epoch": 1.741714003457644,
+      "grad_norm": 0.44828080019324945,
+      "learning_rate": 4.924292145350323e-06,
+      "loss": 0.5723,
+      "step": 4407
+    },
+    {
+      "epoch": 1.7421091627562362,
+      "grad_norm": 0.426266339908026,
+      "learning_rate": 4.924253876221899e-06,
+      "loss": 0.5772,
+      "step": 4408
+    },
+    {
+      "epoch": 1.7425043220548284,
+      "grad_norm": 0.43419093036176604,
+      "learning_rate": 4.92421559757248e-06,
+      "loss": 0.5596,
+      "step": 4409
+    },
+    {
+      "epoch": 1.7428994813534207,
+      "grad_norm": 0.43554329163688116,
+      "learning_rate": 4.924177309402213e-06,
+      "loss": 0.5697,
+      "step": 4410
+    },
+    {
+      "epoch": 1.743294640652013,
+      "grad_norm": 0.4473385143087953,
+      "learning_rate": 4.9241390117112495e-06,
+      "loss": 0.5642,
+      "step": 4411
+    },
+    {
+      "epoch": 1.7436897999506051,
+      "grad_norm": 0.4387919180383067,
+      "learning_rate": 4.92410070449974e-06,
+      "loss": 0.5712,
+      "step": 4412
+    },
+    {
+      "epoch": 1.7440849592491974,
+      "grad_norm": 0.4440101663736749,
+      "learning_rate": 4.924062387767835e-06,
+      "loss": 0.5767,
+      "step": 4413
+    },
+    {
+      "epoch": 1.7444801185477896,
+      "grad_norm": 0.43157217680041976,
+      "learning_rate": 4.924024061515684e-06,
+      "loss": 0.5589,
+      "step": 4414
+    },
+    {
+      "epoch": 1.7448752778463819,
+      "grad_norm": 0.4207925328836993,
+      "learning_rate": 4.9239857257434395e-06,
+      "loss": 0.5494,
+      "step": 4415
+    },
+    {
+      "epoch": 1.745270437144974,
+      "grad_norm": 0.4343427301041821,
+      "learning_rate": 4.923947380451252e-06,
+      "loss": 0.5418,
+      "step": 4416
+    },
+    {
+      "epoch": 1.7456655964435663,
+      "grad_norm": 0.43915247356661835,
+      "learning_rate": 4.92390902563927e-06,
+      "loss": 0.5457,
+      "step": 4417
+    },
+    {
+      "epoch": 1.7460607557421586,
+      "grad_norm": 0.412423518922869,
+      "learning_rate": 4.923870661307645e-06,
+      "loss": 0.543,
+      "step": 4418
+    },
+    {
+      "epoch": 1.7464559150407508,
+      "grad_norm": 0.4659475964234698,
+      "learning_rate": 4.923832287456527e-06,
+      "loss": 0.5727,
+      "step": 4419
+    },
+    {
+      "epoch": 1.746851074339343,
+      "grad_norm": 0.4386733240756286,
+      "learning_rate": 4.92379390408607e-06,
+      "loss": 0.546,
+      "step": 4420
+    },
+    {
+      "epoch": 1.7472462336379353,
+      "grad_norm": 0.4566302235906808,
+      "learning_rate": 4.9237555111964204e-06,
+      "loss": 0.577,
+      "step": 4421
+    },
+    {
+      "epoch": 1.7476413929365275,
+      "grad_norm": 0.4544004401768208,
+      "learning_rate": 4.923717108787731e-06,
+      "loss": 0.5863,
+      "step": 4422
+    },
+    {
+      "epoch": 1.7480365522351198,
+      "grad_norm": 0.4364008166512911,
+      "learning_rate": 4.923678696860153e-06,
+      "loss": 0.5588,
+      "step": 4423
+    },
+    {
+      "epoch": 1.748431711533712,
+      "grad_norm": 0.44646539196245927,
+      "learning_rate": 4.923640275413838e-06,
+      "loss": 0.5797,
+      "step": 4424
+    },
+    {
+      "epoch": 1.7488268708323043,
+      "grad_norm": 0.42595898093134776,
+      "learning_rate": 4.923601844448934e-06,
+      "loss": 0.5326,
+      "step": 4425
+    },
+    {
+      "epoch": 1.7492220301308965,
+      "grad_norm": 0.43632414312178214,
+      "learning_rate": 4.923563403965595e-06,
+      "loss": 0.5477,
+      "step": 4426
+    },
+    {
+      "epoch": 1.7496171894294887,
+      "grad_norm": 0.43161605146182297,
+      "learning_rate": 4.923524953963969e-06,
+      "loss": 0.5652,
+      "step": 4427
+    },
+    {
+      "epoch": 1.750012348728081,
+      "grad_norm": 0.42112039961812653,
+      "learning_rate": 4.923486494444209e-06,
+      "loss": 0.5403,
+      "step": 4428
+    },
+    {
+      "epoch": 1.7504075080266732,
+      "grad_norm": 0.4387403266260913,
+      "learning_rate": 4.923448025406467e-06,
+      "loss": 0.5646,
+      "step": 4429
+    },
+    {
+      "epoch": 1.7508026673252655,
+      "grad_norm": 0.44195630371865374,
+      "learning_rate": 4.923409546850891e-06,
+      "loss": 0.5514,
+      "step": 4430
+    },
+    {
+      "epoch": 1.7511978266238577,
+      "grad_norm": 0.44395804443010556,
+      "learning_rate": 4.923371058777635e-06,
+      "loss": 0.5645,
+      "step": 4431
+    },
+    {
+      "epoch": 1.75159298592245,
+      "grad_norm": 0.4326703585981714,
+      "learning_rate": 4.923332561186849e-06,
+      "loss": 0.5616,
+      "step": 4432
+    },
+    {
+      "epoch": 1.7519881452210422,
+      "grad_norm": 0.41917281291116726,
+      "learning_rate": 4.923294054078684e-06,
+      "loss": 0.5365,
+      "step": 4433
+    },
+    {
+      "epoch": 1.7523833045196344,
+      "grad_norm": 0.43629197498616906,
+      "learning_rate": 4.923255537453292e-06,
+      "loss": 0.5715,
+      "step": 4434
+    },
+    {
+      "epoch": 1.7527784638182267,
+      "grad_norm": 0.43248199576929525,
+      "learning_rate": 4.923217011310823e-06,
+      "loss": 0.5339,
+      "step": 4435
+    },
+    {
+      "epoch": 1.753173623116819,
+      "grad_norm": 0.4796555742650671,
+      "learning_rate": 4.923178475651429e-06,
+      "loss": 0.5628,
+      "step": 4436
+    },
+    {
+      "epoch": 1.7535687824154111,
+      "grad_norm": 0.44016438303089994,
+      "learning_rate": 4.923139930475262e-06,
+      "loss": 0.5796,
+      "step": 4437
+    },
+    {
+      "epoch": 1.7539639417140034,
+      "grad_norm": 0.44061698759839063,
+      "learning_rate": 4.923101375782472e-06,
+      "loss": 0.5676,
+      "step": 4438
+    },
+    {
+      "epoch": 1.7543591010125956,
+      "grad_norm": 0.43587278759955905,
+      "learning_rate": 4.923062811573211e-06,
+      "loss": 0.5731,
+      "step": 4439
+    },
+    {
+      "epoch": 1.7547542603111879,
+      "grad_norm": 0.4324193780709019,
+      "learning_rate": 4.9230242378476325e-06,
+      "loss": 0.5662,
+      "step": 4440
+    },
+    {
+      "epoch": 1.75514941960978,
+      "grad_norm": 0.4330363556293551,
+      "learning_rate": 4.922985654605884e-06,
+      "loss": 0.5526,
+      "step": 4441
+    },
+    {
+      "epoch": 1.7555445789083723,
+      "grad_norm": 0.4314258458445518,
+      "learning_rate": 4.922947061848121e-06,
+      "loss": 0.5468,
+      "step": 4442
+    },
+    {
+      "epoch": 1.7559397382069646,
+      "grad_norm": 0.4367345050157589,
+      "learning_rate": 4.922908459574492e-06,
+      "loss": 0.5573,
+      "step": 4443
+    },
+    {
+      "epoch": 1.7563348975055568,
+      "grad_norm": 0.5054619145164584,
+      "learning_rate": 4.92286984778515e-06,
+      "loss": 0.5611,
+      "step": 4444
+    },
+    {
+      "epoch": 1.756730056804149,
+      "grad_norm": 0.4244999054734717,
+      "learning_rate": 4.922831226480247e-06,
+      "loss": 0.5533,
+      "step": 4445
+    },
+    {
+      "epoch": 1.7571252161027413,
+      "grad_norm": 0.4172564528587188,
+      "learning_rate": 4.9227925956599336e-06,
+      "loss": 0.5394,
+      "step": 4446
+    },
+    {
+      "epoch": 1.7575203754013335,
+      "grad_norm": 0.4487356151355405,
+      "learning_rate": 4.922753955324362e-06,
+      "loss": 0.5593,
+      "step": 4447
+    },
+    {
+      "epoch": 1.7579155346999258,
+      "grad_norm": 0.4471793604107619,
+      "learning_rate": 4.922715305473684e-06,
+      "loss": 0.545,
+      "step": 4448
+    },
+    {
+      "epoch": 1.758310693998518,
+      "grad_norm": 0.42116771593933916,
+      "learning_rate": 4.922676646108052e-06,
+      "loss": 0.5565,
+      "step": 4449
+    },
+    {
+      "epoch": 1.7587058532971103,
+      "grad_norm": 0.42986579453156515,
+      "learning_rate": 4.9226379772276165e-06,
+      "loss": 0.5542,
+      "step": 4450
+    },
+    {
+      "epoch": 1.7591010125957025,
+      "grad_norm": 0.4247741467500481,
+      "learning_rate": 4.922599298832531e-06,
+      "loss": 0.5277,
+      "step": 4451
+    },
+    {
+      "epoch": 1.7594961718942947,
+      "grad_norm": 0.43883975962836197,
+      "learning_rate": 4.922560610922946e-06,
+      "loss": 0.5554,
+      "step": 4452
+    },
+    {
+      "epoch": 1.759891331192887,
+      "grad_norm": 0.4269542206063633,
+      "learning_rate": 4.922521913499014e-06,
+      "loss": 0.55,
+      "step": 4453
+    },
+    {
+      "epoch": 1.7602864904914792,
+      "grad_norm": 0.44358761831837534,
+      "learning_rate": 4.922483206560888e-06,
+      "loss": 0.5681,
+      "step": 4454
+    },
+    {
+      "epoch": 1.7606816497900715,
+      "grad_norm": 0.41524870259004626,
+      "learning_rate": 4.9224444901087174e-06,
+      "loss": 0.5287,
+      "step": 4455
+    },
+    {
+      "epoch": 1.7610768090886637,
+      "grad_norm": 0.47449363146584195,
+      "learning_rate": 4.922405764142656e-06,
+      "loss": 0.5534,
+      "step": 4456
+    },
+    {
+      "epoch": 1.761471968387256,
+      "grad_norm": 0.4274899599226474,
+      "learning_rate": 4.9223670286628566e-06,
+      "loss": 0.5513,
+      "step": 4457
+    },
+    {
+      "epoch": 1.7618671276858484,
+      "grad_norm": 0.4229395462347126,
+      "learning_rate": 4.92232828366947e-06,
+      "loss": 0.5643,
+      "step": 4458
+    },
+    {
+      "epoch": 1.7622622869844407,
+      "grad_norm": 0.43795229928660484,
+      "learning_rate": 4.922289529162649e-06,
+      "loss": 0.5577,
+      "step": 4459
+    },
+    {
+      "epoch": 1.762657446283033,
+      "grad_norm": 0.4509336716051313,
+      "learning_rate": 4.922250765142546e-06,
+      "loss": 0.5369,
+      "step": 4460
+    },
+    {
+      "epoch": 1.7630526055816251,
+      "grad_norm": 0.4963694155527903,
+      "learning_rate": 4.9222119916093115e-06,
+      "loss": 0.5729,
+      "step": 4461
+    },
+    {
+      "epoch": 1.7634477648802174,
+      "grad_norm": 0.4311986192626347,
+      "learning_rate": 4.9221732085631e-06,
+      "loss": 0.5636,
+      "step": 4462
+    },
+    {
+      "epoch": 1.7638429241788096,
+      "grad_norm": 0.4319021698000714,
+      "learning_rate": 4.9221344160040626e-06,
+      "loss": 0.554,
+      "step": 4463
+    },
+    {
+      "epoch": 1.7642380834774019,
+      "grad_norm": 0.42976334810476374,
+      "learning_rate": 4.922095613932353e-06,
+      "loss": 0.5721,
+      "step": 4464
+    },
+    {
+      "epoch": 1.764633242775994,
+      "grad_norm": 0.4348605034262508,
+      "learning_rate": 4.922056802348122e-06,
+      "loss": 0.5481,
+      "step": 4465
+    },
+    {
+      "epoch": 1.7650284020745863,
+      "grad_norm": 0.4402318664595867,
+      "learning_rate": 4.9220179812515226e-06,
+      "loss": 0.5659,
+      "step": 4466
+    },
+    {
+      "epoch": 1.7654235613731786,
+      "grad_norm": 0.43124398530041586,
+      "learning_rate": 4.921979150642707e-06,
+      "loss": 0.5737,
+      "step": 4467
+    },
+    {
+      "epoch": 1.7658187206717708,
+      "grad_norm": 0.43936196155478946,
+      "learning_rate": 4.921940310521828e-06,
+      "loss": 0.5595,
+      "step": 4468
+    },
+    {
+      "epoch": 1.766213879970363,
+      "grad_norm": 0.43223733140815723,
+      "learning_rate": 4.921901460889039e-06,
+      "loss": 0.5488,
+      "step": 4469
+    },
+    {
+      "epoch": 1.7666090392689553,
+      "grad_norm": 0.43010489859753376,
+      "learning_rate": 4.921862601744491e-06,
+      "loss": 0.5597,
+      "step": 4470
+    },
+    {
+      "epoch": 1.7670041985675475,
+      "grad_norm": 0.4258097789495025,
+      "learning_rate": 4.9218237330883375e-06,
+      "loss": 0.5548,
+      "step": 4471
+    },
+    {
+      "epoch": 1.7673993578661398,
+      "grad_norm": 0.4441676656075857,
+      "learning_rate": 4.921784854920731e-06,
+      "loss": 0.5439,
+      "step": 4472
+    },
+    {
+      "epoch": 1.767794517164732,
+      "grad_norm": 0.429127198914778,
+      "learning_rate": 4.921745967241825e-06,
+      "loss": 0.5624,
+      "step": 4473
+    },
+    {
+      "epoch": 1.7681896764633243,
+      "grad_norm": 0.42973439941503405,
+      "learning_rate": 4.921707070051769e-06,
+      "loss": 0.5405,
+      "step": 4474
+    },
+    {
+      "epoch": 1.7685848357619165,
+      "grad_norm": 0.4368917818588069,
+      "learning_rate": 4.92166816335072e-06,
+      "loss": 0.571,
+      "step": 4475
+    },
+    {
+      "epoch": 1.7689799950605087,
+      "grad_norm": 0.4329313735901494,
+      "learning_rate": 4.921629247138829e-06,
+      "loss": 0.5425,
+      "step": 4476
+    },
+    {
+      "epoch": 1.769375154359101,
+      "grad_norm": 0.4522903482231501,
+      "learning_rate": 4.9215903214162485e-06,
+      "loss": 0.5566,
+      "step": 4477
+    },
+    {
+      "epoch": 1.7697703136576932,
+      "grad_norm": 0.4460656956099865,
+      "learning_rate": 4.921551386183131e-06,
+      "loss": 0.5558,
+      "step": 4478
+    },
+    {
+      "epoch": 1.7701654729562855,
+      "grad_norm": 0.41968856516312,
+      "learning_rate": 4.921512441439631e-06,
+      "loss": 0.558,
+      "step": 4479
+    },
+    {
+      "epoch": 1.7705606322548777,
+      "grad_norm": 0.4512166443513065,
+      "learning_rate": 4.9214734871859e-06,
+      "loss": 0.543,
+      "step": 4480
+    },
+    {
+      "epoch": 1.77095579155347,
+      "grad_norm": 0.44040585068512283,
+      "learning_rate": 4.921434523422093e-06,
+      "loss": 0.5606,
+      "step": 4481
+    },
+    {
+      "epoch": 1.7713509508520624,
+      "grad_norm": 0.41848653488225906,
+      "learning_rate": 4.9213955501483605e-06,
+      "loss": 0.5361,
+      "step": 4482
+    },
+    {
+      "epoch": 1.7717461101506546,
+      "grad_norm": 0.43807590314563183,
+      "learning_rate": 4.921356567364856e-06,
+      "loss": 0.5314,
+      "step": 4483
+    },
+    {
+      "epoch": 1.7721412694492469,
+      "grad_norm": 0.4700972659356998,
+      "learning_rate": 4.921317575071733e-06,
+      "loss": 0.5636,
+      "step": 4484
+    },
+    {
+      "epoch": 1.7725364287478391,
+      "grad_norm": 0.45042036525685325,
+      "learning_rate": 4.921278573269146e-06,
+      "loss": 0.5544,
+      "step": 4485
+    },
+    {
+      "epoch": 1.7729315880464314,
+      "grad_norm": 0.44614873313304904,
+      "learning_rate": 4.9212395619572474e-06,
+      "loss": 0.5693,
+      "step": 4486
+    },
+    {
+      "epoch": 1.7733267473450236,
+      "grad_norm": 0.44860751788633424,
+      "learning_rate": 4.92120054113619e-06,
+      "loss": 0.558,
+      "step": 4487
+    },
+    {
+      "epoch": 1.7737219066436158,
+      "grad_norm": 0.43151386401272274,
+      "learning_rate": 4.921161510806125e-06,
+      "loss": 0.5542,
+      "step": 4488
+    },
+    {
+      "epoch": 1.774117065942208,
+      "grad_norm": 0.4356644343958426,
+      "learning_rate": 4.92112247096721e-06,
+      "loss": 0.5368,
+      "step": 4489
+    },
+    {
+      "epoch": 1.7745122252408003,
+      "grad_norm": 0.4548142559509372,
+      "learning_rate": 4.921083421619595e-06,
+      "loss": 0.5736,
+      "step": 4490
+    },
+    {
+      "epoch": 1.7749073845393926,
+      "grad_norm": 0.4459993625684455,
+      "learning_rate": 4.921044362763436e-06,
+      "loss": 0.557,
+      "step": 4491
+    },
+    {
+      "epoch": 1.7753025438379848,
+      "grad_norm": 0.44551019969621913,
+      "learning_rate": 4.921005294398883e-06,
+      "loss": 0.5612,
+      "step": 4492
+    },
+    {
+      "epoch": 1.775697703136577,
+      "grad_norm": 0.4534954772992248,
+      "learning_rate": 4.9209662165260916e-06,
+      "loss": 0.5764,
+      "step": 4493
+    },
+    {
+      "epoch": 1.7760928624351693,
+      "grad_norm": 0.45806011580752654,
+      "learning_rate": 4.9209271291452156e-06,
+      "loss": 0.5565,
+      "step": 4494
+    },
+    {
+      "epoch": 1.7764880217337615,
+      "grad_norm": 0.42599165598927907,
+      "learning_rate": 4.920888032256408e-06,
+      "loss": 0.5434,
+      "step": 4495
+    },
+    {
+      "epoch": 1.7768831810323538,
+      "grad_norm": 0.427438042814662,
+      "learning_rate": 4.920848925859822e-06,
+      "loss": 0.5496,
+      "step": 4496
+    },
+    {
+      "epoch": 1.777278340330946,
+      "grad_norm": 0.4584387507023162,
+      "learning_rate": 4.9208098099556114e-06,
+      "loss": 0.5712,
+      "step": 4497
+    },
+    {
+      "epoch": 1.7776734996295382,
+      "grad_norm": 0.4355189310861454,
+      "learning_rate": 4.920770684543929e-06,
+      "loss": 0.5534,
+      "step": 4498
+    },
+    {
+      "epoch": 1.7780686589281305,
+      "grad_norm": 0.43597371943223134,
+      "learning_rate": 4.920731549624931e-06,
+      "loss": 0.565,
+      "step": 4499
+    },
+    {
+      "epoch": 1.7784638182267227,
+      "grad_norm": 0.4268466727675861,
+      "learning_rate": 4.920692405198769e-06,
+      "loss": 0.5426,
+      "step": 4500
+    },
+    {
+      "epoch": 1.778858977525315,
+      "grad_norm": 0.4224690255424684,
+      "learning_rate": 4.920653251265597e-06,
+      "loss": 0.555,
+      "step": 4501
+    },
+    {
+      "epoch": 1.7792541368239072,
+      "grad_norm": 0.42762282174827215,
+      "learning_rate": 4.920614087825568e-06,
+      "loss": 0.5525,
+      "step": 4502
+    },
+    {
+      "epoch": 1.7796492961224994,
+      "grad_norm": 0.4290253808006352,
+      "learning_rate": 4.9205749148788376e-06,
+      "loss": 0.5723,
+      "step": 4503
+    },
+    {
+      "epoch": 1.7800444554210917,
+      "grad_norm": 0.4316979082254998,
+      "learning_rate": 4.920535732425559e-06,
+      "loss": 0.555,
+      "step": 4504
+    },
+    {
+      "epoch": 1.780439614719684,
+      "grad_norm": 0.438272777807001,
+      "learning_rate": 4.920496540465885e-06,
+      "loss": 0.5421,
+      "step": 4505
+    },
+    {
+      "epoch": 1.7808347740182762,
+      "grad_norm": 0.455364594727949,
+      "learning_rate": 4.920457338999971e-06,
+      "loss": 0.5472,
+      "step": 4506
+    },
+    {
+      "epoch": 1.7812299333168684,
+      "grad_norm": 0.44459923067197527,
+      "learning_rate": 4.920418128027971e-06,
+      "loss": 0.5707,
+      "step": 4507
+    },
+    {
+      "epoch": 1.7816250926154606,
+      "grad_norm": 0.4417890387860095,
+      "learning_rate": 4.920378907550037e-06,
+      "loss": 0.5691,
+      "step": 4508
+    },
+    {
+      "epoch": 1.7820202519140529,
+      "grad_norm": 0.44085319317841215,
+      "learning_rate": 4.9203396775663245e-06,
+      "loss": 0.5439,
+      "step": 4509
+    },
+    {
+      "epoch": 1.7824154112126451,
+      "grad_norm": 0.45781235310847124,
+      "learning_rate": 4.920300438076989e-06,
+      "loss": 0.5535,
+      "step": 4510
+    },
+    {
+      "epoch": 1.7828105705112374,
+      "grad_norm": 0.42751841581481953,
+      "learning_rate": 4.9202611890821815e-06,
+      "loss": 0.5404,
+      "step": 4511
+    },
+    {
+      "epoch": 1.7832057298098296,
+      "grad_norm": 0.4341840586645018,
+      "learning_rate": 4.920221930582059e-06,
+      "loss": 0.5357,
+      "step": 4512
+    },
+    {
+      "epoch": 1.7836008891084218,
+      "grad_norm": 0.44062665450398303,
+      "learning_rate": 4.920182662576773e-06,
+      "loss": 0.5533,
+      "step": 4513
+    },
+    {
+      "epoch": 1.783996048407014,
+      "grad_norm": 0.4569945381667453,
+      "learning_rate": 4.920143385066479e-06,
+      "loss": 0.5605,
+      "step": 4514
+    },
+    {
+      "epoch": 1.7843912077056063,
+      "grad_norm": 0.5311247214308134,
+      "learning_rate": 4.920104098051333e-06,
+      "loss": 0.5786,
+      "step": 4515
+    },
+    {
+      "epoch": 1.7847863670041986,
+      "grad_norm": 0.46139739197081714,
+      "learning_rate": 4.920064801531486e-06,
+      "loss": 0.5773,
+      "step": 4516
+    },
+    {
+      "epoch": 1.7851815263027908,
+      "grad_norm": 0.43510101422224146,
+      "learning_rate": 4.920025495507095e-06,
+      "loss": 0.5375,
+      "step": 4517
+    },
+    {
+      "epoch": 1.785576685601383,
+      "grad_norm": 0.4299696722046226,
+      "learning_rate": 4.919986179978313e-06,
+      "loss": 0.5596,
+      "step": 4518
+    },
+    {
+      "epoch": 1.7859718448999753,
+      "grad_norm": 0.4245007247792506,
+      "learning_rate": 4.9199468549452956e-06,
+      "loss": 0.5575,
+      "step": 4519
+    },
+    {
+      "epoch": 1.7863670041985675,
+      "grad_norm": 0.427078611580864,
+      "learning_rate": 4.919907520408196e-06,
+      "loss": 0.5594,
+      "step": 4520
+    },
+    {
+      "epoch": 1.7867621634971598,
+      "grad_norm": 0.43840491567979,
+      "learning_rate": 4.919868176367168e-06,
+      "loss": 0.5404,
+      "step": 4521
+    },
+    {
+      "epoch": 1.787157322795752,
+      "grad_norm": 0.4569134416853893,
+      "learning_rate": 4.919828822822369e-06,
+      "loss": 0.5508,
+      "step": 4522
+    },
+    {
+      "epoch": 1.7875524820943443,
+      "grad_norm": 0.43107835382000415,
+      "learning_rate": 4.91978945977395e-06,
+      "loss": 0.5594,
+      "step": 4523
+    },
+    {
+      "epoch": 1.7879476413929365,
+      "grad_norm": 0.4361231870002236,
+      "learning_rate": 4.919750087222068e-06,
+      "loss": 0.5726,
+      "step": 4524
+    },
+    {
+      "epoch": 1.7883428006915287,
+      "grad_norm": 0.4353192973361471,
+      "learning_rate": 4.919710705166878e-06,
+      "loss": 0.5641,
+      "step": 4525
+    },
+    {
+      "epoch": 1.788737959990121,
+      "grad_norm": 0.4364332444547788,
+      "learning_rate": 4.919671313608533e-06,
+      "loss": 0.541,
+      "step": 4526
+    },
+    {
+      "epoch": 1.7891331192887132,
+      "grad_norm": 0.43409123496491997,
+      "learning_rate": 4.919631912547188e-06,
+      "loss": 0.5687,
+      "step": 4527
+    },
+    {
+      "epoch": 1.7895282785873055,
+      "grad_norm": 0.43446982982003957,
+      "learning_rate": 4.919592501982998e-06,
+      "loss": 0.5556,
+      "step": 4528
+    },
+    {
+      "epoch": 1.7899234378858977,
+      "grad_norm": 0.4345438030521792,
+      "learning_rate": 4.9195530819161185e-06,
+      "loss": 0.5474,
+      "step": 4529
+    },
+    {
+      "epoch": 1.79031859718449,
+      "grad_norm": 0.43436414603534473,
+      "learning_rate": 4.919513652346704e-06,
+      "loss": 0.5548,
+      "step": 4530
+    },
+    {
+      "epoch": 1.7907137564830822,
+      "grad_norm": 0.4320792577761697,
+      "learning_rate": 4.919474213274908e-06,
+      "loss": 0.5293,
+      "step": 4531
+    },
+    {
+      "epoch": 1.7911089157816744,
+      "grad_norm": 0.4417297782885469,
+      "learning_rate": 4.919434764700888e-06,
+      "loss": 0.5532,
+      "step": 4532
+    },
+    {
+      "epoch": 1.7915040750802667,
+      "grad_norm": 0.4491898130529519,
+      "learning_rate": 4.9193953066247965e-06,
+      "loss": 0.5691,
+      "step": 4533
+    },
+    {
+      "epoch": 1.791899234378859,
+      "grad_norm": 0.44010561319843694,
+      "learning_rate": 4.919355839046789e-06,
+      "loss": 0.5601,
+      "step": 4534
+    },
+    {
+      "epoch": 1.7922943936774511,
+      "grad_norm": 0.4326683763258569,
+      "learning_rate": 4.919316361967021e-06,
+      "loss": 0.5601,
+      "step": 4535
+    },
+    {
+      "epoch": 1.7926895529760434,
+      "grad_norm": 0.49880321858382415,
+      "learning_rate": 4.919276875385648e-06,
+      "loss": 0.5802,
+      "step": 4536
+    },
+    {
+      "epoch": 1.7930847122746356,
+      "grad_norm": 0.4402605348668624,
+      "learning_rate": 4.919237379302824e-06,
+      "loss": 0.5442,
+      "step": 4537
+    },
+    {
+      "epoch": 1.7934798715732279,
+      "grad_norm": 0.4487929856641043,
+      "learning_rate": 4.919197873718705e-06,
+      "loss": 0.5683,
+      "step": 4538
+    },
+    {
+      "epoch": 1.79387503087182,
+      "grad_norm": 0.42534492592015277,
+      "learning_rate": 4.919158358633445e-06,
+      "loss": 0.5652,
+      "step": 4539
+    },
+    {
+      "epoch": 1.7942701901704123,
+      "grad_norm": 0.4457639959321754,
+      "learning_rate": 4.919118834047201e-06,
+      "loss": 0.5658,
+      "step": 4540
+    },
+    {
+      "epoch": 1.7946653494690046,
+      "grad_norm": 0.4541593885926761,
+      "learning_rate": 4.919079299960127e-06,
+      "loss": 0.5688,
+      "step": 4541
+    },
+    {
+      "epoch": 1.7950605087675968,
+      "grad_norm": 0.42333628277611585,
+      "learning_rate": 4.919039756372378e-06,
+      "loss": 0.5474,
+      "step": 4542
+    },
+    {
+      "epoch": 1.795455668066189,
+      "grad_norm": 0.42037667148139624,
+      "learning_rate": 4.91900020328411e-06,
+      "loss": 0.5578,
+      "step": 4543
+    },
+    {
+      "epoch": 1.7958508273647813,
+      "grad_norm": 0.41437397951305344,
+      "learning_rate": 4.918960640695478e-06,
+      "loss": 0.5507,
+      "step": 4544
+    },
+    {
+      "epoch": 1.7962459866633735,
+      "grad_norm": 0.43166169225760287,
+      "learning_rate": 4.918921068606638e-06,
+      "loss": 0.5518,
+      "step": 4545
+    },
+    {
+      "epoch": 1.7966411459619658,
+      "grad_norm": 0.4327031422658885,
+      "learning_rate": 4.9188814870177435e-06,
+      "loss": 0.5559,
+      "step": 4546
+    },
+    {
+      "epoch": 1.797036305260558,
+      "grad_norm": 0.43810822912426284,
+      "learning_rate": 4.918841895928953e-06,
+      "loss": 0.5538,
+      "step": 4547
+    },
+    {
+      "epoch": 1.7974314645591503,
+      "grad_norm": 0.43509704083525685,
+      "learning_rate": 4.918802295340419e-06,
+      "loss": 0.5534,
+      "step": 4548
+    },
+    {
+      "epoch": 1.7978266238577425,
+      "grad_norm": 0.4204249306753487,
+      "learning_rate": 4.918762685252299e-06,
+      "loss": 0.555,
+      "step": 4549
+    },
+    {
+      "epoch": 1.7982217831563347,
+      "grad_norm": 0.429252379340873,
+      "learning_rate": 4.918723065664747e-06,
+      "loss": 0.5569,
+      "step": 4550
+    },
+    {
+      "epoch": 1.798616942454927,
+      "grad_norm": 0.4222980081248242,
+      "learning_rate": 4.918683436577921e-06,
+      "loss": 0.5271,
+      "step": 4551
+    },
+    {
+      "epoch": 1.7990121017535192,
+      "grad_norm": 0.4313877563692727,
+      "learning_rate": 4.918643797991975e-06,
+      "loss": 0.5702,
+      "step": 4552
+    },
+    {
+      "epoch": 1.7994072610521117,
+      "grad_norm": 0.43938081010488067,
+      "learning_rate": 4.918604149907064e-06,
+      "loss": 0.5914,
+      "step": 4553
+    },
+    {
+      "epoch": 1.799802420350704,
+      "grad_norm": 0.42842791710741546,
+      "learning_rate": 4.918564492323346e-06,
+      "loss": 0.5464,
+      "step": 4554
+    },
+    {
+      "epoch": 1.8001975796492962,
+      "grad_norm": 0.4346701478393733,
+      "learning_rate": 4.918524825240973e-06,
+      "loss": 0.5426,
+      "step": 4555
+    },
+    {
+      "epoch": 1.8005927389478884,
+      "grad_norm": 0.4300228304840291,
+      "learning_rate": 4.918485148660105e-06,
+      "loss": 0.5718,
+      "step": 4556
+    },
+    {
+      "epoch": 1.8009878982464806,
+      "grad_norm": 0.4328620116394069,
+      "learning_rate": 4.918445462580895e-06,
+      "loss": 0.567,
+      "step": 4557
+    },
+    {
+      "epoch": 1.8013830575450729,
+      "grad_norm": 0.43079626932416043,
+      "learning_rate": 4.9184057670035e-06,
+      "loss": 0.5518,
+      "step": 4558
+    },
+    {
+      "epoch": 1.8017782168436651,
+      "grad_norm": 0.43706587631711175,
+      "learning_rate": 4.918366061928076e-06,
+      "loss": 0.5492,
+      "step": 4559
+    },
+    {
+      "epoch": 1.8021733761422574,
+      "grad_norm": 0.4403528370210141,
+      "learning_rate": 4.918326347354778e-06,
+      "loss": 0.564,
+      "step": 4560
+    },
+    {
+      "epoch": 1.8025685354408496,
+      "grad_norm": 0.4265423519618484,
+      "learning_rate": 4.918286623283763e-06,
+      "loss": 0.5441,
+      "step": 4561
+    },
+    {
+      "epoch": 1.8029636947394418,
+      "grad_norm": 0.4497893586884997,
+      "learning_rate": 4.918246889715186e-06,
+      "loss": 0.5729,
+      "step": 4562
+    },
+    {
+      "epoch": 1.803358854038034,
+      "grad_norm": 0.4394008615994604,
+      "learning_rate": 4.918207146649204e-06,
+      "loss": 0.5577,
+      "step": 4563
+    },
+    {
+      "epoch": 1.8037540133366263,
+      "grad_norm": 0.4541079886766186,
+      "learning_rate": 4.918167394085974e-06,
+      "loss": 0.5629,
+      "step": 4564
+    },
+    {
+      "epoch": 1.8041491726352186,
+      "grad_norm": 0.45506948243874085,
+      "learning_rate": 4.91812763202565e-06,
+      "loss": 0.5561,
+      "step": 4565
+    },
+    {
+      "epoch": 1.8045443319338108,
+      "grad_norm": 0.4342638593346457,
+      "learning_rate": 4.918087860468388e-06,
+      "loss": 0.576,
+      "step": 4566
+    },
+    {
+      "epoch": 1.804939491232403,
+      "grad_norm": 0.4340982482700696,
+      "learning_rate": 4.918048079414346e-06,
+      "loss": 0.5554,
+      "step": 4567
+    },
+    {
+      "epoch": 1.8053346505309953,
+      "grad_norm": 0.4409861288626772,
+      "learning_rate": 4.91800828886368e-06,
+      "loss": 0.5726,
+      "step": 4568
+    },
+    {
+      "epoch": 1.8057298098295875,
+      "grad_norm": 0.4322705316149436,
+      "learning_rate": 4.917968488816545e-06,
+      "loss": 0.5329,
+      "step": 4569
+    },
+    {
+      "epoch": 1.8061249691281798,
+      "grad_norm": 0.4192848854350575,
+      "learning_rate": 4.917928679273098e-06,
+      "loss": 0.5427,
+      "step": 4570
+    },
+    {
+      "epoch": 1.806520128426772,
+      "grad_norm": 0.44939852167500016,
+      "learning_rate": 4.917888860233496e-06,
+      "loss": 0.556,
+      "step": 4571
+    },
+    {
+      "epoch": 1.8069152877253642,
+      "grad_norm": 0.5167355570733759,
+      "learning_rate": 4.917849031697894e-06,
+      "loss": 0.6005,
+      "step": 4572
+    },
+    {
+      "epoch": 1.8073104470239565,
+      "grad_norm": 0.4403454466703087,
+      "learning_rate": 4.91780919366645e-06,
+      "loss": 0.5508,
+      "step": 4573
+    },
+    {
+      "epoch": 1.8077056063225487,
+      "grad_norm": 0.4293181191267813,
+      "learning_rate": 4.917769346139319e-06,
+      "loss": 0.5607,
+      "step": 4574
+    },
+    {
+      "epoch": 1.808100765621141,
+      "grad_norm": 0.4338122274539976,
+      "learning_rate": 4.9177294891166585e-06,
+      "loss": 0.5423,
+      "step": 4575
+    },
+    {
+      "epoch": 1.8084959249197334,
+      "grad_norm": 0.43148032764038485,
+      "learning_rate": 4.917689622598625e-06,
+      "loss": 0.5718,
+      "step": 4576
+    },
+    {
+      "epoch": 1.8088910842183257,
+      "grad_norm": 0.43602229802024295,
+      "learning_rate": 4.917649746585374e-06,
+      "loss": 0.5404,
+      "step": 4577
+    },
+    {
+      "epoch": 1.809286243516918,
+      "grad_norm": 0.44355545237967436,
+      "learning_rate": 4.917609861077064e-06,
+      "loss": 0.5673,
+      "step": 4578
+    },
+    {
+      "epoch": 1.8096814028155102,
+      "grad_norm": 0.45356140574084836,
+      "learning_rate": 4.917569966073849e-06,
+      "loss": 0.5466,
+      "step": 4579
+    },
+    {
+      "epoch": 1.8100765621141024,
+      "grad_norm": 0.4133512710636827,
+      "learning_rate": 4.917530061575888e-06,
+      "loss": 0.5353,
+      "step": 4580
+    },
+    {
+      "epoch": 1.8104717214126946,
+      "grad_norm": 0.4368743815868589,
+      "learning_rate": 4.917490147583337e-06,
+      "loss": 0.5711,
+      "step": 4581
+    },
+    {
+      "epoch": 1.8108668807112869,
+      "grad_norm": 0.429475779733978,
+      "learning_rate": 4.917450224096353e-06,
+      "loss": 0.5663,
+      "step": 4582
+    },
+    {
+      "epoch": 1.8112620400098791,
+      "grad_norm": 0.4489675045116446,
+      "learning_rate": 4.917410291115092e-06,
+      "loss": 0.569,
+      "step": 4583
+    },
+    {
+      "epoch": 1.8116571993084714,
+      "grad_norm": 0.4253227142982292,
+      "learning_rate": 4.917370348639712e-06,
+      "loss": 0.5518,
+      "step": 4584
+    },
+    {
+      "epoch": 1.8120523586070636,
+      "grad_norm": 0.43792903853943244,
+      "learning_rate": 4.917330396670368e-06,
+      "loss": 0.5547,
+      "step": 4585
+    },
+    {
+      "epoch": 1.8124475179056558,
+      "grad_norm": 0.431983453096147,
+      "learning_rate": 4.917290435207219e-06,
+      "loss": 0.5622,
+      "step": 4586
+    },
+    {
+      "epoch": 1.812842677204248,
+      "grad_norm": 0.4421445444609276,
+      "learning_rate": 4.9172504642504204e-06,
+      "loss": 0.5734,
+      "step": 4587
+    },
+    {
+      "epoch": 1.8132378365028403,
+      "grad_norm": 0.4410353497824584,
+      "learning_rate": 4.91721048380013e-06,
+      "loss": 0.5692,
+      "step": 4588
+    },
+    {
+      "epoch": 1.8136329958014326,
+      "grad_norm": 0.41201401996131587,
+      "learning_rate": 4.917170493856504e-06,
+      "loss": 0.5427,
+      "step": 4589
+    },
+    {
+      "epoch": 1.8140281551000248,
+      "grad_norm": 0.4334126547698007,
+      "learning_rate": 4.917130494419702e-06,
+      "loss": 0.553,
+      "step": 4590
+    },
+    {
+      "epoch": 1.814423314398617,
+      "grad_norm": 0.4336096435165992,
+      "learning_rate": 4.917090485489877e-06,
+      "loss": 0.5694,
+      "step": 4591
+    },
+    {
+      "epoch": 1.8148184736972093,
+      "grad_norm": 0.43402803896434733,
+      "learning_rate": 4.91705046706719e-06,
+      "loss": 0.5671,
+      "step": 4592
+    },
+    {
+      "epoch": 1.8152136329958015,
+      "grad_norm": 0.43863493460328706,
+      "learning_rate": 4.917010439151796e-06,
+      "loss": 0.5469,
+      "step": 4593
+    },
+    {
+      "epoch": 1.8156087922943938,
+      "grad_norm": 0.4232411523985182,
+      "learning_rate": 4.916970401743852e-06,
+      "loss": 0.5397,
+      "step": 4594
+    },
+    {
+      "epoch": 1.816003951592986,
+      "grad_norm": 0.41773834042402014,
+      "learning_rate": 4.916930354843516e-06,
+      "loss": 0.5595,
+      "step": 4595
+    },
+    {
+      "epoch": 1.8163991108915782,
+      "grad_norm": 0.42847635907614967,
+      "learning_rate": 4.9168902984509456e-06,
+      "loss": 0.5675,
+      "step": 4596
+    },
+    {
+      "epoch": 1.8167942701901705,
+      "grad_norm": 0.43702921400183176,
+      "learning_rate": 4.9168502325662985e-06,
+      "loss": 0.544,
+      "step": 4597
+    },
+    {
+      "epoch": 1.8171894294887627,
+      "grad_norm": 0.4359050509194601,
+      "learning_rate": 4.91681015718973e-06,
+      "loss": 0.5531,
+      "step": 4598
+    },
+    {
+      "epoch": 1.817584588787355,
+      "grad_norm": 0.4238189029244474,
+      "learning_rate": 4.9167700723214e-06,
+      "loss": 0.5222,
+      "step": 4599
+    },
+    {
+      "epoch": 1.8179797480859472,
+      "grad_norm": 0.42884238541026204,
+      "learning_rate": 4.916729977961463e-06,
+      "loss": 0.5486,
+      "step": 4600
+    },
+    {
+      "epoch": 1.8183749073845394,
+      "grad_norm": 0.42474360636234726,
+      "learning_rate": 4.91668987411008e-06,
+      "loss": 0.5553,
+      "step": 4601
+    },
+    {
+      "epoch": 1.8187700666831317,
+      "grad_norm": 0.44168858012742196,
+      "learning_rate": 4.916649760767405e-06,
+      "loss": 0.5518,
+      "step": 4602
+    },
+    {
+      "epoch": 1.819165225981724,
+      "grad_norm": 0.4556359965693222,
+      "learning_rate": 4.916609637933598e-06,
+      "loss": 0.5668,
+      "step": 4603
+    },
+    {
+      "epoch": 1.8195603852803162,
+      "grad_norm": 0.43307847780187875,
+      "learning_rate": 4.916569505608816e-06,
+      "loss": 0.5498,
+      "step": 4604
+    },
+    {
+      "epoch": 1.8199555445789084,
+      "grad_norm": 0.4311936492838269,
+      "learning_rate": 4.916529363793216e-06,
+      "loss": 0.554,
+      "step": 4605
+    },
+    {
+      "epoch": 1.8203507038775006,
+      "grad_norm": 0.42973054647548137,
+      "learning_rate": 4.916489212486956e-06,
+      "loss": 0.565,
+      "step": 4606
+    },
+    {
+      "epoch": 1.8207458631760929,
+      "grad_norm": 0.42836235280682294,
+      "learning_rate": 4.916449051690194e-06,
+      "loss": 0.5599,
+      "step": 4607
+    },
+    {
+      "epoch": 1.8211410224746851,
+      "grad_norm": 0.41648067518333354,
+      "learning_rate": 4.916408881403087e-06,
+      "loss": 0.5499,
+      "step": 4608
+    },
+    {
+      "epoch": 1.8215361817732774,
+      "grad_norm": 0.4389273257682643,
+      "learning_rate": 4.916368701625795e-06,
+      "loss": 0.5582,
+      "step": 4609
+    },
+    {
+      "epoch": 1.8219313410718696,
+      "grad_norm": 0.4996066817560874,
+      "learning_rate": 4.916328512358472e-06,
+      "loss": 0.544,
+      "step": 4610
+    },
+    {
+      "epoch": 1.8223265003704618,
+      "grad_norm": 0.4320671880478789,
+      "learning_rate": 4.916288313601278e-06,
+      "loss": 0.5684,
+      "step": 4611
+    },
+    {
+      "epoch": 1.822721659669054,
+      "grad_norm": 0.47994501476908125,
+      "learning_rate": 4.916248105354372e-06,
+      "loss": 0.5627,
+      "step": 4612
+    },
+    {
+      "epoch": 1.8231168189676463,
+      "grad_norm": 0.44725918814624716,
+      "learning_rate": 4.91620788761791e-06,
+      "loss": 0.5801,
+      "step": 4613
+    },
+    {
+      "epoch": 1.8235119782662386,
+      "grad_norm": 0.41780605996678455,
+      "learning_rate": 4.9161676603920505e-06,
+      "loss": 0.5368,
+      "step": 4614
+    },
+    {
+      "epoch": 1.8239071375648308,
+      "grad_norm": 0.42149016682401924,
+      "learning_rate": 4.9161274236769516e-06,
+      "loss": 0.5417,
+      "step": 4615
+    },
+    {
+      "epoch": 1.824302296863423,
+      "grad_norm": 0.4159145131021071,
+      "learning_rate": 4.916087177472771e-06,
+      "loss": 0.5634,
+      "step": 4616
+    },
+    {
+      "epoch": 1.8246974561620153,
+      "grad_norm": 0.427772247807,
+      "learning_rate": 4.916046921779668e-06,
+      "loss": 0.5597,
+      "step": 4617
+    },
+    {
+      "epoch": 1.8250926154606075,
+      "grad_norm": 0.504175016285204,
+      "learning_rate": 4.916006656597799e-06,
+      "loss": 0.5534,
+      "step": 4618
+    },
+    {
+      "epoch": 1.8254877747591998,
+      "grad_norm": 0.43930170796960316,
+      "learning_rate": 4.915966381927324e-06,
+      "loss": 0.5625,
+      "step": 4619
+    },
+    {
+      "epoch": 1.825882934057792,
+      "grad_norm": 0.4317015969701484,
+      "learning_rate": 4.9159260977683986e-06,
+      "loss": 0.5535,
+      "step": 4620
+    },
+    {
+      "epoch": 1.8262780933563842,
+      "grad_norm": 0.4319583379497269,
+      "learning_rate": 4.915885804121184e-06,
+      "loss": 0.5443,
+      "step": 4621
+    },
+    {
+      "epoch": 1.8266732526549765,
+      "grad_norm": 0.46283454804644775,
+      "learning_rate": 4.915845500985836e-06,
+      "loss": 0.5798,
+      "step": 4622
+    },
+    {
+      "epoch": 1.8270684119535687,
+      "grad_norm": 0.4578316028623526,
+      "learning_rate": 4.915805188362514e-06,
+      "loss": 0.5873,
+      "step": 4623
+    },
+    {
+      "epoch": 1.827463571252161,
+      "grad_norm": 0.4380807378788219,
+      "learning_rate": 4.915764866251376e-06,
+      "loss": 0.5702,
+      "step": 4624
+    },
+    {
+      "epoch": 1.8278587305507532,
+      "grad_norm": 0.4635611902517724,
+      "learning_rate": 4.915724534652581e-06,
+      "loss": 0.5676,
+      "step": 4625
+    },
+    {
+      "epoch": 1.8282538898493454,
+      "grad_norm": 0.44336086993981816,
+      "learning_rate": 4.915684193566287e-06,
+      "loss": 0.571,
+      "step": 4626
+    },
+    {
+      "epoch": 1.8286490491479377,
+      "grad_norm": 0.49804959145623445,
+      "learning_rate": 4.915643842992652e-06,
+      "loss": 0.5704,
+      "step": 4627
+    },
+    {
+      "epoch": 1.82904420844653,
+      "grad_norm": 0.4484804351198767,
+      "learning_rate": 4.915603482931835e-06,
+      "loss": 0.5601,
+      "step": 4628
+    },
+    {
+      "epoch": 1.8294393677451222,
+      "grad_norm": 0.4259382481340384,
+      "learning_rate": 4.915563113383994e-06,
+      "loss": 0.5414,
+      "step": 4629
+    },
+    {
+      "epoch": 1.8298345270437144,
+      "grad_norm": 0.46410055417082385,
+      "learning_rate": 4.915522734349289e-06,
+      "loss": 0.5668,
+      "step": 4630
+    },
+    {
+      "epoch": 1.8302296863423066,
+      "grad_norm": 0.4450378060474863,
+      "learning_rate": 4.915482345827876e-06,
+      "loss": 0.5625,
+      "step": 4631
+    },
+    {
+      "epoch": 1.830624845640899,
+      "grad_norm": 0.4469682638808664,
+      "learning_rate": 4.915441947819916e-06,
+      "loss": 0.56,
+      "step": 4632
+    },
+    {
+      "epoch": 1.8310200049394911,
+      "grad_norm": 0.4386874685983757,
+      "learning_rate": 4.915401540325566e-06,
+      "loss": 0.5763,
+      "step": 4633
+    },
+    {
+      "epoch": 1.8314151642380834,
+      "grad_norm": 0.4386404133941846,
+      "learning_rate": 4.9153611233449864e-06,
+      "loss": 0.5513,
+      "step": 4634
+    },
+    {
+      "epoch": 1.8318103235366756,
+      "grad_norm": 0.45589832428091503,
+      "learning_rate": 4.915320696878335e-06,
+      "loss": 0.5545,
+      "step": 4635
+    },
+    {
+      "epoch": 1.8322054828352679,
+      "grad_norm": 0.4388397614681877,
+      "learning_rate": 4.91528026092577e-06,
+      "loss": 0.5369,
+      "step": 4636
+    },
+    {
+      "epoch": 1.83260064213386,
+      "grad_norm": 0.4220869515007803,
+      "learning_rate": 4.915239815487451e-06,
+      "loss": 0.5536,
+      "step": 4637
+    },
+    {
+      "epoch": 1.8329958014324523,
+      "grad_norm": 0.4516307562377459,
+      "learning_rate": 4.915199360563536e-06,
+      "loss": 0.5663,
+      "step": 4638
+    },
+    {
+      "epoch": 1.8333909607310446,
+      "grad_norm": 0.43082783021013155,
+      "learning_rate": 4.915158896154185e-06,
+      "loss": 0.5711,
+      "step": 4639
+    },
+    {
+      "epoch": 1.8337861200296368,
+      "grad_norm": 0.4280112586626305,
+      "learning_rate": 4.915118422259557e-06,
+      "loss": 0.5586,
+      "step": 4640
+    },
+    {
+      "epoch": 1.834181279328229,
+      "grad_norm": 0.455862847226725,
+      "learning_rate": 4.91507793887981e-06,
+      "loss": 0.5567,
+      "step": 4641
+    },
+    {
+      "epoch": 1.8345764386268213,
+      "grad_norm": 0.43432786099005233,
+      "learning_rate": 4.915037446015103e-06,
+      "loss": 0.5722,
+      "step": 4642
+    },
+    {
+      "epoch": 1.8349715979254135,
+      "grad_norm": 0.45288339991619825,
+      "learning_rate": 4.914996943665596e-06,
+      "loss": 0.5548,
+      "step": 4643
+    },
+    {
+      "epoch": 1.8353667572240058,
+      "grad_norm": 0.430304761849555,
+      "learning_rate": 4.914956431831447e-06,
+      "loss": 0.5569,
+      "step": 4644
+    },
+    {
+      "epoch": 1.835761916522598,
+      "grad_norm": 0.4516766791005426,
+      "learning_rate": 4.914915910512815e-06,
+      "loss": 0.5439,
+      "step": 4645
+    },
+    {
+      "epoch": 1.8361570758211903,
+      "grad_norm": 0.4262026911256271,
+      "learning_rate": 4.914875379709861e-06,
+      "loss": 0.5502,
+      "step": 4646
+    },
+    {
+      "epoch": 1.8365522351197827,
+      "grad_norm": 0.44982296126025034,
+      "learning_rate": 4.914834839422742e-06,
+      "loss": 0.5753,
+      "step": 4647
+    },
+    {
+      "epoch": 1.836947394418375,
+      "grad_norm": 0.4305160106289168,
+      "learning_rate": 4.914794289651619e-06,
+      "loss": 0.5476,
+      "step": 4648
+    },
+    {
+      "epoch": 1.8373425537169672,
+      "grad_norm": 0.43698124283033885,
+      "learning_rate": 4.91475373039665e-06,
+      "loss": 0.5626,
+      "step": 4649
+    },
+    {
+      "epoch": 1.8377377130155594,
+      "grad_norm": 0.4324559418236934,
+      "learning_rate": 4.914713161657993e-06,
+      "loss": 0.5465,
+      "step": 4650
+    },
+    {
+      "epoch": 1.8381328723141517,
+      "grad_norm": 0.46505067847109016,
+      "learning_rate": 4.914672583435811e-06,
+      "loss": 0.5773,
+      "step": 4651
+    },
+    {
+      "epoch": 1.838528031612744,
+      "grad_norm": 0.44054104719934584,
+      "learning_rate": 4.9146319957302615e-06,
+      "loss": 0.5462,
+      "step": 4652
+    },
+    {
+      "epoch": 1.8389231909113362,
+      "grad_norm": 0.4293700652645379,
+      "learning_rate": 4.914591398541503e-06,
+      "loss": 0.5514,
+      "step": 4653
+    },
+    {
+      "epoch": 1.8393183502099284,
+      "grad_norm": 0.45753087472514964,
+      "learning_rate": 4.9145507918696956e-06,
+      "loss": 0.5478,
+      "step": 4654
+    },
+    {
+      "epoch": 1.8397135095085206,
+      "grad_norm": 0.4461916227350306,
+      "learning_rate": 4.9145101757149994e-06,
+      "loss": 0.5548,
+      "step": 4655
+    },
+    {
+      "epoch": 1.8401086688071129,
+      "grad_norm": 0.4471190461133453,
+      "learning_rate": 4.914469550077573e-06,
+      "loss": 0.5522,
+      "step": 4656
+    },
+    {
+      "epoch": 1.8405038281057051,
+      "grad_norm": 0.44384611797235896,
+      "learning_rate": 4.914428914957576e-06,
+      "loss": 0.572,
+      "step": 4657
+    },
+    {
+      "epoch": 1.8408989874042974,
+      "grad_norm": 0.4574610337875189,
+      "learning_rate": 4.9143882703551685e-06,
+      "loss": 0.5625,
+      "step": 4658
+    },
+    {
+      "epoch": 1.8412941467028896,
+      "grad_norm": 0.5048055930740011,
+      "learning_rate": 4.914347616270511e-06,
+      "loss": 0.5431,
+      "step": 4659
+    },
+    {
+      "epoch": 1.8416893060014818,
+      "grad_norm": 0.4421724478683854,
+      "learning_rate": 4.914306952703761e-06,
+      "loss": 0.5472,
+      "step": 4660
+    },
+    {
+      "epoch": 1.842084465300074,
+      "grad_norm": 0.43918097376339305,
+      "learning_rate": 4.914266279655079e-06,
+      "loss": 0.5677,
+      "step": 4661
+    },
+    {
+      "epoch": 1.8424796245986663,
+      "grad_norm": 0.4518794942674701,
+      "learning_rate": 4.914225597124626e-06,
+      "loss": 0.5492,
+      "step": 4662
+    },
+    {
+      "epoch": 1.8428747838972586,
+      "grad_norm": 0.47641842318354,
+      "learning_rate": 4.9141849051125614e-06,
+      "loss": 0.5769,
+      "step": 4663
+    },
+    {
+      "epoch": 1.8432699431958508,
+      "grad_norm": 0.43242823970216165,
+      "learning_rate": 4.9141442036190435e-06,
+      "loss": 0.5474,
+      "step": 4664
+    },
+    {
+      "epoch": 1.843665102494443,
+      "grad_norm": 0.43975219203201016,
+      "learning_rate": 4.914103492644233e-06,
+      "loss": 0.5428,
+      "step": 4665
+    },
+    {
+      "epoch": 1.8440602617930353,
+      "grad_norm": 0.44273932980174646,
+      "learning_rate": 4.91406277218829e-06,
+      "loss": 0.5429,
+      "step": 4666
+    },
+    {
+      "epoch": 1.8444554210916275,
+      "grad_norm": 0.4371689468098226,
+      "learning_rate": 4.914022042251375e-06,
+      "loss": 0.5443,
+      "step": 4667
+    },
+    {
+      "epoch": 1.8448505803902198,
+      "grad_norm": 0.46091840391838906,
+      "learning_rate": 4.9139813028336465e-06,
+      "loss": 0.5623,
+      "step": 4668
+    },
+    {
+      "epoch": 1.845245739688812,
+      "grad_norm": 0.4515022813696254,
+      "learning_rate": 4.9139405539352655e-06,
+      "loss": 0.5615,
+      "step": 4669
+    },
+    {
+      "epoch": 1.8456408989874042,
+      "grad_norm": 0.44395923841481416,
+      "learning_rate": 4.913899795556391e-06,
+      "loss": 0.5688,
+      "step": 4670
+    },
+    {
+      "epoch": 1.8460360582859967,
+      "grad_norm": 0.44117565523570207,
+      "learning_rate": 4.913859027697185e-06,
+      "loss": 0.5654,
+      "step": 4671
+    },
+    {
+      "epoch": 1.846431217584589,
+      "grad_norm": 0.42429096303861064,
+      "learning_rate": 4.913818250357807e-06,
+      "loss": 0.5462,
+      "step": 4672
+    },
+    {
+      "epoch": 1.8468263768831812,
+      "grad_norm": 0.44958055380119594,
+      "learning_rate": 4.913777463538416e-06,
+      "loss": 0.5686,
+      "step": 4673
+    },
+    {
+      "epoch": 1.8472215361817734,
+      "grad_norm": 0.4443331190570723,
+      "learning_rate": 4.913736667239173e-06,
+      "loss": 0.5567,
+      "step": 4674
+    },
+    {
+      "epoch": 1.8476166954803657,
+      "grad_norm": 0.42627047085078384,
+      "learning_rate": 4.913695861460238e-06,
+      "loss": 0.5426,
+      "step": 4675
+    },
+    {
+      "epoch": 1.848011854778958,
+      "grad_norm": 0.4500558947718227,
+      "learning_rate": 4.9136550462017716e-06,
+      "loss": 0.5542,
+      "step": 4676
+    },
+    {
+      "epoch": 1.8484070140775501,
+      "grad_norm": 0.4414303454674209,
+      "learning_rate": 4.913614221463932e-06,
+      "loss": 0.555,
+      "step": 4677
+    },
+    {
+      "epoch": 1.8488021733761424,
+      "grad_norm": 0.43962204738708255,
+      "learning_rate": 4.913573387246884e-06,
+      "loss": 0.5704,
+      "step": 4678
+    },
+    {
+      "epoch": 1.8491973326747346,
+      "grad_norm": 0.44664322555736324,
+      "learning_rate": 4.9135325435507845e-06,
+      "loss": 0.57,
+      "step": 4679
+    },
+    {
+      "epoch": 1.8495924919733269,
+      "grad_norm": 0.4539113882483917,
+      "learning_rate": 4.913491690375794e-06,
+      "loss": 0.5679,
+      "step": 4680
+    },
+    {
+      "epoch": 1.849987651271919,
+      "grad_norm": 0.4392523191394003,
+      "learning_rate": 4.913450827722074e-06,
+      "loss": 0.5388,
+      "step": 4681
+    },
+    {
+      "epoch": 1.8503828105705113,
+      "grad_norm": 0.430956543329709,
+      "learning_rate": 4.913409955589785e-06,
+      "loss": 0.5727,
+      "step": 4682
+    },
+    {
+      "epoch": 1.8507779698691036,
+      "grad_norm": 0.4509366979993098,
+      "learning_rate": 4.9133690739790864e-06,
+      "loss": 0.5503,
+      "step": 4683
+    },
+    {
+      "epoch": 1.8511731291676958,
+      "grad_norm": 0.4412502685023789,
+      "learning_rate": 4.91332818289014e-06,
+      "loss": 0.542,
+      "step": 4684
+    },
+    {
+      "epoch": 1.851568288466288,
+      "grad_norm": 0.4421900270354689,
+      "learning_rate": 4.913287282323107e-06,
+      "loss": 0.5562,
+      "step": 4685
+    },
+    {
+      "epoch": 1.8519634477648803,
+      "grad_norm": 0.5017741843721322,
+      "learning_rate": 4.913246372278145e-06,
+      "loss": 0.5546,
+      "step": 4686
+    },
+    {
+      "epoch": 1.8523586070634726,
+      "grad_norm": 0.42598558349235277,
+      "learning_rate": 4.913205452755418e-06,
+      "loss": 0.5565,
+      "step": 4687
+    },
+    {
+      "epoch": 1.8527537663620648,
+      "grad_norm": 0.44801891691212664,
+      "learning_rate": 4.913164523755085e-06,
+      "loss": 0.5622,
+      "step": 4688
+    },
+    {
+      "epoch": 1.853148925660657,
+      "grad_norm": 0.46504021620652847,
+      "learning_rate": 4.9131235852773075e-06,
+      "loss": 0.5486,
+      "step": 4689
+    },
+    {
+      "epoch": 1.8535440849592493,
+      "grad_norm": 0.42392415227635827,
+      "learning_rate": 4.913082637322245e-06,
+      "loss": 0.5462,
+      "step": 4690
+    },
+    {
+      "epoch": 1.8539392442578415,
+      "grad_norm": 0.43104217850617627,
+      "learning_rate": 4.91304167989006e-06,
+      "loss": 0.5406,
+      "step": 4691
+    },
+    {
+      "epoch": 1.8543344035564338,
+      "grad_norm": 0.47167572808016,
+      "learning_rate": 4.9130007129809135e-06,
+      "loss": 0.5469,
+      "step": 4692
+    },
+    {
+      "epoch": 1.854729562855026,
+      "grad_norm": 0.4747344328170053,
+      "learning_rate": 4.912959736594963e-06,
+      "loss": 0.5653,
+      "step": 4693
+    },
+    {
+      "epoch": 1.8551247221536182,
+      "grad_norm": 0.4343101893171199,
+      "learning_rate": 4.912918750732374e-06,
+      "loss": 0.5373,
+      "step": 4694
+    },
+    {
+      "epoch": 1.8555198814522105,
+      "grad_norm": 0.42499504251024883,
+      "learning_rate": 4.9128777553933035e-06,
+      "loss": 0.5499,
+      "step": 4695
+    },
+    {
+      "epoch": 1.8559150407508027,
+      "grad_norm": 0.45857770081696003,
+      "learning_rate": 4.9128367505779165e-06,
+      "loss": 0.5665,
+      "step": 4696
+    },
+    {
+      "epoch": 1.856310200049395,
+      "grad_norm": 0.47188633170732824,
+      "learning_rate": 4.91279573628637e-06,
+      "loss": 0.5745,
+      "step": 4697
+    },
+    {
+      "epoch": 1.8567053593479872,
+      "grad_norm": 0.43004014361775683,
+      "learning_rate": 4.912754712518828e-06,
+      "loss": 0.5387,
+      "step": 4698
+    },
+    {
+      "epoch": 1.8571005186465794,
+      "grad_norm": 0.450368047124033,
+      "learning_rate": 4.912713679275451e-06,
+      "loss": 0.551,
+      "step": 4699
+    },
+    {
+      "epoch": 1.8574956779451717,
+      "grad_norm": 0.4561974358066812,
+      "learning_rate": 4.912672636556398e-06,
+      "loss": 0.5645,
+      "step": 4700
+    },
+    {
+      "epoch": 1.857890837243764,
+      "grad_norm": 0.4327321080207485,
+      "learning_rate": 4.912631584361833e-06,
+      "loss": 0.5449,
+      "step": 4701
+    },
+    {
+      "epoch": 1.8582859965423562,
+      "grad_norm": 0.42737763737038065,
+      "learning_rate": 4.912590522691917e-06,
+      "loss": 0.5381,
+      "step": 4702
+    },
+    {
+      "epoch": 1.8586811558409484,
+      "grad_norm": 0.43781196017871354,
+      "learning_rate": 4.912549451546809e-06,
+      "loss": 0.5615,
+      "step": 4703
+    },
+    {
+      "epoch": 1.8590763151395406,
+      "grad_norm": 0.43337841503621866,
+      "learning_rate": 4.912508370926672e-06,
+      "loss": 0.5569,
+      "step": 4704
+    },
+    {
+      "epoch": 1.8594714744381329,
+      "grad_norm": 0.451977525541082,
+      "learning_rate": 4.912467280831668e-06,
+      "loss": 0.5494,
+      "step": 4705
+    },
+    {
+      "epoch": 1.8598666337367251,
+      "grad_norm": 0.4728561044232118,
+      "learning_rate": 4.9124261812619566e-06,
+      "loss": 0.5483,
+      "step": 4706
+    },
+    {
+      "epoch": 1.8602617930353174,
+      "grad_norm": 0.4415020307643117,
+      "learning_rate": 4.9123850722177e-06,
+      "loss": 0.548,
+      "step": 4707
+    },
+    {
+      "epoch": 1.8606569523339096,
+      "grad_norm": 0.45173525627796496,
+      "learning_rate": 4.912343953699061e-06,
+      "loss": 0.5574,
+      "step": 4708
+    },
+    {
+      "epoch": 1.8610521116325018,
+      "grad_norm": 0.4267764458634889,
+      "learning_rate": 4.912302825706198e-06,
+      "loss": 0.553,
+      "step": 4709
+    },
+    {
+      "epoch": 1.861447270931094,
+      "grad_norm": 0.4357451577511683,
+      "learning_rate": 4.912261688239275e-06,
+      "loss": 0.5541,
+      "step": 4710
+    },
+    {
+      "epoch": 1.8618424302296863,
+      "grad_norm": 0.46999902555719913,
+      "learning_rate": 4.912220541298454e-06,
+      "loss": 0.5777,
+      "step": 4711
+    },
+    {
+      "epoch": 1.8622375895282786,
+      "grad_norm": 0.4574693196717579,
+      "learning_rate": 4.912179384883894e-06,
+      "loss": 0.5587,
+      "step": 4712
+    },
+    {
+      "epoch": 1.8626327488268708,
+      "grad_norm": 0.44445542433033436,
+      "learning_rate": 4.912138218995759e-06,
+      "loss": 0.5649,
+      "step": 4713
+    },
+    {
+      "epoch": 1.863027908125463,
+      "grad_norm": 0.4342553915239652,
+      "learning_rate": 4.9120970436342095e-06,
+      "loss": 0.562,
+      "step": 4714
+    },
+    {
+      "epoch": 1.8634230674240553,
+      "grad_norm": 0.4543224773995171,
+      "learning_rate": 4.912055858799407e-06,
+      "loss": 0.5718,
+      "step": 4715
+    },
+    {
+      "epoch": 1.8638182267226475,
+      "grad_norm": 0.42547741598029465,
+      "learning_rate": 4.912014664491514e-06,
+      "loss": 0.5439,
+      "step": 4716
+    },
+    {
+      "epoch": 1.8642133860212398,
+      "grad_norm": 0.45619787933952977,
+      "learning_rate": 4.911973460710692e-06,
+      "loss": 0.5515,
+      "step": 4717
+    },
+    {
+      "epoch": 1.864608545319832,
+      "grad_norm": 0.46059272985729816,
+      "learning_rate": 4.911932247457104e-06,
+      "loss": 0.5648,
+      "step": 4718
+    },
+    {
+      "epoch": 1.8650037046184242,
+      "grad_norm": 0.44116740865007975,
+      "learning_rate": 4.911891024730911e-06,
+      "loss": 0.5574,
+      "step": 4719
+    },
+    {
+      "epoch": 1.8653988639170165,
+      "grad_norm": 0.41870208648406154,
+      "learning_rate": 4.9118497925322725e-06,
+      "loss": 0.537,
+      "step": 4720
+    },
+    {
+      "epoch": 1.8657940232156087,
+      "grad_norm": 0.5142452344989176,
+      "learning_rate": 4.911808550861353e-06,
+      "loss": 0.5673,
+      "step": 4721
+    },
+    {
+      "epoch": 1.866189182514201,
+      "grad_norm": 0.4401669678300714,
+      "learning_rate": 4.9117672997183155e-06,
+      "loss": 0.5578,
+      "step": 4722
+    },
+    {
+      "epoch": 1.8665843418127932,
+      "grad_norm": 0.4455903304729346,
+      "learning_rate": 4.911726039103319e-06,
+      "loss": 0.5608,
+      "step": 4723
+    },
+    {
+      "epoch": 1.8669795011113854,
+      "grad_norm": 0.44233877100194063,
+      "learning_rate": 4.911684769016528e-06,
+      "loss": 0.5764,
+      "step": 4724
+    },
+    {
+      "epoch": 1.8673746604099777,
+      "grad_norm": 0.43688108562586403,
+      "learning_rate": 4.911643489458104e-06,
+      "loss": 0.5526,
+      "step": 4725
+    },
+    {
+      "epoch": 1.86776981970857,
+      "grad_norm": 0.4534632460594352,
+      "learning_rate": 4.911602200428208e-06,
+      "loss": 0.5542,
+      "step": 4726
+    },
+    {
+      "epoch": 1.8681649790071622,
+      "grad_norm": 0.4803823461812875,
+      "learning_rate": 4.911560901927003e-06,
+      "loss": 0.5458,
+      "step": 4727
+    },
+    {
+      "epoch": 1.8685601383057544,
+      "grad_norm": 0.4390720418788202,
+      "learning_rate": 4.911519593954652e-06,
+      "loss": 0.5585,
+      "step": 4728
+    },
+    {
+      "epoch": 1.8689552976043466,
+      "grad_norm": 0.43151939301169534,
+      "learning_rate": 4.9114782765113155e-06,
+      "loss": 0.5717,
+      "step": 4729
+    },
+    {
+      "epoch": 1.8693504569029389,
+      "grad_norm": 0.43397827245657605,
+      "learning_rate": 4.911436949597157e-06,
+      "loss": 0.5294,
+      "step": 4730
+    },
+    {
+      "epoch": 1.8697456162015311,
+      "grad_norm": 0.4348332309375543,
+      "learning_rate": 4.911395613212339e-06,
+      "loss": 0.5472,
+      "step": 4731
+    },
+    {
+      "epoch": 1.8701407755001234,
+      "grad_norm": 0.4244300368938872,
+      "learning_rate": 4.911354267357022e-06,
+      "loss": 0.565,
+      "step": 4732
+    },
+    {
+      "epoch": 1.8705359347987156,
+      "grad_norm": 0.4353867070970591,
+      "learning_rate": 4.911312912031371e-06,
+      "loss": 0.5642,
+      "step": 4733
+    },
+    {
+      "epoch": 1.8709310940973078,
+      "grad_norm": 0.42839969001511663,
+      "learning_rate": 4.9112715472355464e-06,
+      "loss": 0.5475,
+      "step": 4734
+    },
+    {
+      "epoch": 1.8713262533959,
+      "grad_norm": 0.44110960690425427,
+      "learning_rate": 4.911230172969711e-06,
+      "loss": 0.5614,
+      "step": 4735
+    },
+    {
+      "epoch": 1.8717214126944923,
+      "grad_norm": 0.4995723449596207,
+      "learning_rate": 4.911188789234028e-06,
+      "loss": 0.5662,
+      "step": 4736
+    },
+    {
+      "epoch": 1.8721165719930846,
+      "grad_norm": 0.423171830947674,
+      "learning_rate": 4.91114739602866e-06,
+      "loss": 0.5564,
+      "step": 4737
+    },
+    {
+      "epoch": 1.8725117312916768,
+      "grad_norm": 0.44108599595757897,
+      "learning_rate": 4.911105993353769e-06,
+      "loss": 0.5664,
+      "step": 4738
+    },
+    {
+      "epoch": 1.872906890590269,
+      "grad_norm": 0.4502796994319581,
+      "learning_rate": 4.9110645812095174e-06,
+      "loss": 0.5572,
+      "step": 4739
+    },
+    {
+      "epoch": 1.8733020498888613,
+      "grad_norm": 0.4245422544885071,
+      "learning_rate": 4.911023159596069e-06,
+      "loss": 0.538,
+      "step": 4740
+    },
+    {
+      "epoch": 1.8736972091874535,
+      "grad_norm": 0.41908462248992445,
+      "learning_rate": 4.910981728513586e-06,
+      "loss": 0.5383,
+      "step": 4741
+    },
+    {
+      "epoch": 1.874092368486046,
+      "grad_norm": 0.4232818477813306,
+      "learning_rate": 4.910940287962229e-06,
+      "loss": 0.5348,
+      "step": 4742
+    },
+    {
+      "epoch": 1.8744875277846382,
+      "grad_norm": 0.43217070766658694,
+      "learning_rate": 4.910898837942163e-06,
+      "loss": 0.5527,
+      "step": 4743
+    },
+    {
+      "epoch": 1.8748826870832305,
+      "grad_norm": 0.4312801007800965,
+      "learning_rate": 4.9108573784535515e-06,
+      "loss": 0.5627,
+      "step": 4744
+    },
+    {
+      "epoch": 1.8752778463818227,
+      "grad_norm": 0.44318569935703905,
+      "learning_rate": 4.910815909496555e-06,
+      "loss": 0.5666,
+      "step": 4745
+    },
+    {
+      "epoch": 1.875673005680415,
+      "grad_norm": 0.4264547180929095,
+      "learning_rate": 4.910774431071338e-06,
+      "loss": 0.5566,
+      "step": 4746
+    },
+    {
+      "epoch": 1.8760681649790072,
+      "grad_norm": 0.43447024061939904,
+      "learning_rate": 4.910732943178063e-06,
+      "loss": 0.5388,
+      "step": 4747
+    },
+    {
+      "epoch": 1.8764633242775994,
+      "grad_norm": 0.44201190438269433,
+      "learning_rate": 4.9106914458168934e-06,
+      "loss": 0.537,
+      "step": 4748
+    },
+    {
+      "epoch": 1.8768584835761917,
+      "grad_norm": 0.43667832935964324,
+      "learning_rate": 4.91064993898799e-06,
+      "loss": 0.554,
+      "step": 4749
+    },
+    {
+      "epoch": 1.877253642874784,
+      "grad_norm": 0.4332594072357512,
+      "learning_rate": 4.910608422691519e-06,
+      "loss": 0.5458,
+      "step": 4750
+    },
+    {
+      "epoch": 1.8776488021733762,
+      "grad_norm": 0.4281199350474413,
+      "learning_rate": 4.910566896927642e-06,
+      "loss": 0.5343,
+      "step": 4751
+    },
+    {
+      "epoch": 1.8780439614719684,
+      "grad_norm": 0.4300151216135207,
+      "learning_rate": 4.910525361696521e-06,
+      "loss": 0.5442,
+      "step": 4752
+    },
+    {
+      "epoch": 1.8784391207705606,
+      "grad_norm": 0.4448357946018024,
+      "learning_rate": 4.91048381699832e-06,
+      "loss": 0.5608,
+      "step": 4753
+    },
+    {
+      "epoch": 1.8788342800691529,
+      "grad_norm": 0.43367677911395175,
+      "learning_rate": 4.910442262833204e-06,
+      "loss": 0.5497,
+      "step": 4754
+    },
+    {
+      "epoch": 1.8792294393677451,
+      "grad_norm": 0.4307967775333535,
+      "learning_rate": 4.9104006992013335e-06,
+      "loss": 0.5522,
+      "step": 4755
+    },
+    {
+      "epoch": 1.8796245986663374,
+      "grad_norm": 0.44915265676312033,
+      "learning_rate": 4.910359126102872e-06,
+      "loss": 0.5541,
+      "step": 4756
+    },
+    {
+      "epoch": 1.8800197579649296,
+      "grad_norm": 0.4513270587407442,
+      "learning_rate": 4.910317543537984e-06,
+      "loss": 0.5704,
+      "step": 4757
+    },
+    {
+      "epoch": 1.8804149172635218,
+      "grad_norm": 0.42836930222487574,
+      "learning_rate": 4.910275951506832e-06,
+      "loss": 0.5455,
+      "step": 4758
+    },
+    {
+      "epoch": 1.880810076562114,
+      "grad_norm": 0.4200332636052868,
+      "learning_rate": 4.91023435000958e-06,
+      "loss": 0.5493,
+      "step": 4759
+    },
+    {
+      "epoch": 1.8812052358607063,
+      "grad_norm": 0.4322816077163206,
+      "learning_rate": 4.910192739046392e-06,
+      "loss": 0.5828,
+      "step": 4760
+    },
+    {
+      "epoch": 1.8816003951592986,
+      "grad_norm": 0.42914581453180367,
+      "learning_rate": 4.910151118617429e-06,
+      "loss": 0.5444,
+      "step": 4761
+    },
+    {
+      "epoch": 1.8819955544578908,
+      "grad_norm": 0.4164751887870663,
+      "learning_rate": 4.910109488722857e-06,
+      "loss": 0.5345,
+      "step": 4762
+    },
+    {
+      "epoch": 1.882390713756483,
+      "grad_norm": 0.4231358763864964,
+      "learning_rate": 4.910067849362838e-06,
+      "loss": 0.5464,
+      "step": 4763
+    },
+    {
+      "epoch": 1.8827858730550753,
+      "grad_norm": 0.4289620642079432,
+      "learning_rate": 4.910026200537535e-06,
+      "loss": 0.5629,
+      "step": 4764
+    },
+    {
+      "epoch": 1.8831810323536677,
+      "grad_norm": 0.4296872173261681,
+      "learning_rate": 4.909984542247115e-06,
+      "loss": 0.5323,
+      "step": 4765
+    },
+    {
+      "epoch": 1.88357619165226,
+      "grad_norm": 0.4405891662125567,
+      "learning_rate": 4.909942874491736e-06,
+      "loss": 0.5581,
+      "step": 4766
+    },
+    {
+      "epoch": 1.8839713509508522,
+      "grad_norm": 0.49421695499827306,
+      "learning_rate": 4.9099011972715674e-06,
+      "loss": 0.561,
+      "step": 4767
+    },
+    {
+      "epoch": 1.8843665102494445,
+      "grad_norm": 0.44152259881922284,
+      "learning_rate": 4.909859510586769e-06,
+      "loss": 0.55,
+      "step": 4768
+    },
+    {
+      "epoch": 1.8847616695480367,
+      "grad_norm": 0.5509522024405635,
+      "learning_rate": 4.909817814437506e-06,
+      "loss": 0.5424,
+      "step": 4769
+    },
+    {
+      "epoch": 1.885156828846629,
+      "grad_norm": 0.42019717906356363,
+      "learning_rate": 4.909776108823941e-06,
+      "loss": 0.5529,
+      "step": 4770
+    },
+    {
+      "epoch": 1.8855519881452212,
+      "grad_norm": 0.42580576090703337,
+      "learning_rate": 4.909734393746241e-06,
+      "loss": 0.5465,
+      "step": 4771
+    },
+    {
+      "epoch": 1.8859471474438134,
+      "grad_norm": 0.429426688268032,
+      "learning_rate": 4.909692669204565e-06,
+      "loss": 0.5518,
+      "step": 4772
+    },
+    {
+      "epoch": 1.8863423067424057,
+      "grad_norm": 0.4597075112764053,
+      "learning_rate": 4.909650935199082e-06,
+      "loss": 0.593,
+      "step": 4773
+    },
+    {
+      "epoch": 1.886737466040998,
+      "grad_norm": 0.4231583750958899,
+      "learning_rate": 4.909609191729951e-06,
+      "loss": 0.5623,
+      "step": 4774
+    },
+    {
+      "epoch": 1.8871326253395901,
+      "grad_norm": 0.42211619397536165,
+      "learning_rate": 4.90956743879734e-06,
+      "loss": 0.5507,
+      "step": 4775
+    },
+    {
+      "epoch": 1.8875277846381824,
+      "grad_norm": 0.42252973826533957,
+      "learning_rate": 4.90952567640141e-06,
+      "loss": 0.5614,
+      "step": 4776
+    },
+    {
+      "epoch": 1.8879229439367746,
+      "grad_norm": 0.4531797772700644,
+      "learning_rate": 4.909483904542327e-06,
+      "loss": 0.5533,
+      "step": 4777
+    },
+    {
+      "epoch": 1.8883181032353669,
+      "grad_norm": 0.4376214503057026,
+      "learning_rate": 4.909442123220255e-06,
+      "loss": 0.5564,
+      "step": 4778
+    },
+    {
+      "epoch": 1.888713262533959,
+      "grad_norm": 0.43083585059658475,
+      "learning_rate": 4.909400332435357e-06,
+      "loss": 0.5523,
+      "step": 4779
+    },
+    {
+      "epoch": 1.8891084218325513,
+      "grad_norm": 0.41258509304143803,
+      "learning_rate": 4.909358532187796e-06,
+      "loss": 0.5389,
+      "step": 4780
+    },
+    {
+      "epoch": 1.8895035811311436,
+      "grad_norm": 0.4343631923478858,
+      "learning_rate": 4.909316722477739e-06,
+      "loss": 0.5707,
+      "step": 4781
+    },
+    {
+      "epoch": 1.8898987404297358,
+      "grad_norm": 0.4432077662590313,
+      "learning_rate": 4.909274903305349e-06,
+      "loss": 0.5461,
+      "step": 4782
+    },
+    {
+      "epoch": 1.890293899728328,
+      "grad_norm": 0.43233155649566796,
+      "learning_rate": 4.909233074670791e-06,
+      "loss": 0.568,
+      "step": 4783
+    },
+    {
+      "epoch": 1.8906890590269203,
+      "grad_norm": 0.42852851485106386,
+      "learning_rate": 4.909191236574227e-06,
+      "loss": 0.5621,
+      "step": 4784
+    },
+    {
+      "epoch": 1.8910842183255125,
+      "grad_norm": 0.4322477381113751,
+      "learning_rate": 4.909149389015823e-06,
+      "loss": 0.5531,
+      "step": 4785
+    },
+    {
+      "epoch": 1.8914793776241048,
+      "grad_norm": 0.43499368654441956,
+      "learning_rate": 4.909107531995744e-06,
+      "loss": 0.5732,
+      "step": 4786
+    },
+    {
+      "epoch": 1.891874536922697,
+      "grad_norm": 0.43733806705560363,
+      "learning_rate": 4.909065665514152e-06,
+      "loss": 0.5763,
+      "step": 4787
+    },
+    {
+      "epoch": 1.8922696962212893,
+      "grad_norm": 0.45421207033142263,
+      "learning_rate": 4.909023789571214e-06,
+      "loss": 0.5762,
+      "step": 4788
+    },
+    {
+      "epoch": 1.8926648555198815,
+      "grad_norm": 0.47864113189844826,
+      "learning_rate": 4.908981904167094e-06,
+      "loss": 0.5378,
+      "step": 4789
+    },
+    {
+      "epoch": 1.8930600148184737,
+      "grad_norm": 0.4498106966078856,
+      "learning_rate": 4.908940009301955e-06,
+      "loss": 0.554,
+      "step": 4790
+    },
+    {
+      "epoch": 1.893455174117066,
+      "grad_norm": 0.47655100568930403,
+      "learning_rate": 4.908898104975962e-06,
+      "loss": 0.5498,
+      "step": 4791
+    },
+    {
+      "epoch": 1.8938503334156582,
+      "grad_norm": 0.43699221579300107,
+      "learning_rate": 4.908856191189281e-06,
+      "loss": 0.5803,
+      "step": 4792
+    },
+    {
+      "epoch": 1.8942454927142505,
+      "grad_norm": 0.44031567020576257,
+      "learning_rate": 4.908814267942075e-06,
+      "loss": 0.5603,
+      "step": 4793
+    },
+    {
+      "epoch": 1.8946406520128427,
+      "grad_norm": 0.4381512496766988,
+      "learning_rate": 4.908772335234509e-06,
+      "loss": 0.56,
+      "step": 4794
+    },
+    {
+      "epoch": 1.895035811311435,
+      "grad_norm": 0.4387758165017901,
+      "learning_rate": 4.9087303930667485e-06,
+      "loss": 0.5642,
+      "step": 4795
+    },
+    {
+      "epoch": 1.8954309706100272,
+      "grad_norm": 0.4239025821881657,
+      "learning_rate": 4.908688441438957e-06,
+      "loss": 0.5522,
+      "step": 4796
+    },
+    {
+      "epoch": 1.8958261299086194,
+      "grad_norm": 0.4273577795828686,
+      "learning_rate": 4.908646480351301e-06,
+      "loss": 0.5502,
+      "step": 4797
+    },
+    {
+      "epoch": 1.8962212892072117,
+      "grad_norm": 0.44421556427677866,
+      "learning_rate": 4.908604509803944e-06,
+      "loss": 0.5495,
+      "step": 4798
+    },
+    {
+      "epoch": 1.896616448505804,
+      "grad_norm": 0.44492416211462005,
+      "learning_rate": 4.908562529797051e-06,
+      "loss": 0.5577,
+      "step": 4799
+    },
+    {
+      "epoch": 1.8970116078043961,
+      "grad_norm": 0.4388278234952724,
+      "learning_rate": 4.908520540330786e-06,
+      "loss": 0.5615,
+      "step": 4800
+    },
+    {
+      "epoch": 1.8974067671029884,
+      "grad_norm": 0.43228612343838674,
+      "learning_rate": 4.908478541405316e-06,
+      "loss": 0.5467,
+      "step": 4801
+    },
+    {
+      "epoch": 1.8978019264015806,
+      "grad_norm": 0.4309700517194293,
+      "learning_rate": 4.908436533020804e-06,
+      "loss": 0.5321,
+      "step": 4802
+    },
+    {
+      "epoch": 1.8981970857001729,
+      "grad_norm": 0.4281007407461444,
+      "learning_rate": 4.908394515177416e-06,
+      "loss": 0.5356,
+      "step": 4803
+    },
+    {
+      "epoch": 1.898592244998765,
+      "grad_norm": 0.4323019441412228,
+      "learning_rate": 4.908352487875317e-06,
+      "loss": 0.5679,
+      "step": 4804
+    },
+    {
+      "epoch": 1.8989874042973574,
+      "grad_norm": 0.4403836139433748,
+      "learning_rate": 4.908310451114672e-06,
+      "loss": 0.5662,
+      "step": 4805
+    },
+    {
+      "epoch": 1.8993825635959496,
+      "grad_norm": 0.44165797245282984,
+      "learning_rate": 4.908268404895645e-06,
+      "loss": 0.543,
+      "step": 4806
+    },
+    {
+      "epoch": 1.8997777228945418,
+      "grad_norm": 0.4437790299847047,
+      "learning_rate": 4.908226349218404e-06,
+      "loss": 0.5511,
+      "step": 4807
+    },
+    {
+      "epoch": 1.900172882193134,
+      "grad_norm": 0.44724367559807077,
+      "learning_rate": 4.908184284083111e-06,
+      "loss": 0.5635,
+      "step": 4808
+    },
+    {
+      "epoch": 1.9005680414917263,
+      "grad_norm": 0.4365212081127553,
+      "learning_rate": 4.908142209489932e-06,
+      "loss": 0.5542,
+      "step": 4809
+    },
+    {
+      "epoch": 1.9009632007903186,
+      "grad_norm": 0.45163950732319247,
+      "learning_rate": 4.908100125439033e-06,
+      "loss": 0.5627,
+      "step": 4810
+    },
+    {
+      "epoch": 1.9013583600889108,
+      "grad_norm": 0.4316244913107137,
+      "learning_rate": 4.90805803193058e-06,
+      "loss": 0.5498,
+      "step": 4811
+    },
+    {
+      "epoch": 1.901753519387503,
+      "grad_norm": 0.445880369848173,
+      "learning_rate": 4.908015928964735e-06,
+      "loss": 0.5565,
+      "step": 4812
+    },
+    {
+      "epoch": 1.9021486786860953,
+      "grad_norm": 0.4344296261072089,
+      "learning_rate": 4.9079738165416676e-06,
+      "loss": 0.5358,
+      "step": 4813
+    },
+    {
+      "epoch": 1.9025438379846875,
+      "grad_norm": 0.4352424451727997,
+      "learning_rate": 4.907931694661541e-06,
+      "loss": 0.5495,
+      "step": 4814
+    },
+    {
+      "epoch": 1.9029389972832798,
+      "grad_norm": 0.4200141642812343,
+      "learning_rate": 4.907889563324521e-06,
+      "loss": 0.5529,
+      "step": 4815
+    },
+    {
+      "epoch": 1.903334156581872,
+      "grad_norm": 0.42328744956338943,
+      "learning_rate": 4.907847422530773e-06,
+      "loss": 0.5408,
+      "step": 4816
+    },
+    {
+      "epoch": 1.9037293158804642,
+      "grad_norm": 0.43808904720068553,
+      "learning_rate": 4.907805272280461e-06,
+      "loss": 0.5654,
+      "step": 4817
+    },
+    {
+      "epoch": 1.9041244751790565,
+      "grad_norm": 0.4345122275007242,
+      "learning_rate": 4.907763112573754e-06,
+      "loss": 0.5558,
+      "step": 4818
+    },
+    {
+      "epoch": 1.9045196344776487,
+      "grad_norm": 0.41639788021854496,
+      "learning_rate": 4.907720943410814e-06,
+      "loss": 0.5506,
+      "step": 4819
+    },
+    {
+      "epoch": 1.904914793776241,
+      "grad_norm": 0.42802462554370857,
+      "learning_rate": 4.90767876479181e-06,
+      "loss": 0.5433,
+      "step": 4820
+    },
+    {
+      "epoch": 1.9053099530748332,
+      "grad_norm": 0.43712860812027704,
+      "learning_rate": 4.907636576716904e-06,
+      "loss": 0.5482,
+      "step": 4821
+    },
+    {
+      "epoch": 1.9057051123734254,
+      "grad_norm": 0.4442981118293308,
+      "learning_rate": 4.9075943791862645e-06,
+      "loss": 0.5516,
+      "step": 4822
+    },
+    {
+      "epoch": 1.9061002716720177,
+      "grad_norm": 0.42923155588906153,
+      "learning_rate": 4.907552172200056e-06,
+      "loss": 0.5646,
+      "step": 4823
+    },
+    {
+      "epoch": 1.90649543097061,
+      "grad_norm": 0.4175409742504318,
+      "learning_rate": 4.907509955758444e-06,
+      "loss": 0.5476,
+      "step": 4824
+    },
+    {
+      "epoch": 1.9068905902692022,
+      "grad_norm": 0.43260553320884426,
+      "learning_rate": 4.907467729861595e-06,
+      "loss": 0.5576,
+      "step": 4825
+    },
+    {
+      "epoch": 1.9072857495677944,
+      "grad_norm": 0.4789723482405854,
+      "learning_rate": 4.907425494509675e-06,
+      "loss": 0.5863,
+      "step": 4826
+    },
+    {
+      "epoch": 1.9076809088663866,
+      "grad_norm": 0.43256077905417156,
+      "learning_rate": 4.90738324970285e-06,
+      "loss": 0.5648,
+      "step": 4827
+    },
+    {
+      "epoch": 1.9080760681649789,
+      "grad_norm": 0.4301753052479235,
+      "learning_rate": 4.907340995441284e-06,
+      "loss": 0.5765,
+      "step": 4828
+    },
+    {
+      "epoch": 1.9084712274635711,
+      "grad_norm": 0.4262264166103991,
+      "learning_rate": 4.907298731725146e-06,
+      "loss": 0.549,
+      "step": 4829
+    },
+    {
+      "epoch": 1.9088663867621634,
+      "grad_norm": 0.41997042490994213,
+      "learning_rate": 4.9072564585546e-06,
+      "loss": 0.558,
+      "step": 4830
+    },
+    {
+      "epoch": 1.9092615460607556,
+      "grad_norm": 0.4364550869207615,
+      "learning_rate": 4.9072141759298114e-06,
+      "loss": 0.5643,
+      "step": 4831
+    },
+    {
+      "epoch": 1.9096567053593478,
+      "grad_norm": 0.42201288173328566,
+      "learning_rate": 4.907171883850948e-06,
+      "loss": 0.534,
+      "step": 4832
+    },
+    {
+      "epoch": 1.91005186465794,
+      "grad_norm": 0.4417197054389254,
+      "learning_rate": 4.907129582318175e-06,
+      "loss": 0.5823,
+      "step": 4833
+    },
+    {
+      "epoch": 1.9104470239565323,
+      "grad_norm": 0.43489230519142597,
+      "learning_rate": 4.907087271331658e-06,
+      "loss": 0.5634,
+      "step": 4834
+    },
+    {
+      "epoch": 1.9108421832551246,
+      "grad_norm": 0.4273968780202027,
+      "learning_rate": 4.907044950891565e-06,
+      "loss": 0.5668,
+      "step": 4835
+    },
+    {
+      "epoch": 1.911237342553717,
+      "grad_norm": 0.42553680975953273,
+      "learning_rate": 4.907002620998061e-06,
+      "loss": 0.5562,
+      "step": 4836
+    },
+    {
+      "epoch": 1.9116325018523093,
+      "grad_norm": 0.45057537156888056,
+      "learning_rate": 4.906960281651312e-06,
+      "loss": 0.5906,
+      "step": 4837
+    },
+    {
+      "epoch": 1.9120276611509015,
+      "grad_norm": 0.43561521398581576,
+      "learning_rate": 4.906917932851484e-06,
+      "loss": 0.5673,
+      "step": 4838
+    },
+    {
+      "epoch": 1.9124228204494937,
+      "grad_norm": 0.41696070212722824,
+      "learning_rate": 4.906875574598745e-06,
+      "loss": 0.545,
+      "step": 4839
+    },
+    {
+      "epoch": 1.912817979748086,
+      "grad_norm": 0.4542669385242667,
+      "learning_rate": 4.90683320689326e-06,
+      "loss": 0.5732,
+      "step": 4840
+    },
+    {
+      "epoch": 1.9132131390466782,
+      "grad_norm": 0.42925333709930363,
+      "learning_rate": 4.906790829735195e-06,
+      "loss": 0.5553,
+      "step": 4841
+    },
+    {
+      "epoch": 1.9136082983452705,
+      "grad_norm": 0.4189953420798717,
+      "learning_rate": 4.906748443124718e-06,
+      "loss": 0.5412,
+      "step": 4842
+    },
+    {
+      "epoch": 1.9140034576438627,
+      "grad_norm": 0.43842151872555235,
+      "learning_rate": 4.906706047061994e-06,
+      "loss": 0.5842,
+      "step": 4843
+    },
+    {
+      "epoch": 1.914398616942455,
+      "grad_norm": 0.43303434492630205,
+      "learning_rate": 4.906663641547191e-06,
+      "loss": 0.5446,
+      "step": 4844
+    },
+    {
+      "epoch": 1.9147937762410472,
+      "grad_norm": 0.44750651700901584,
+      "learning_rate": 4.906621226580473e-06,
+      "loss": 0.5565,
+      "step": 4845
+    },
+    {
+      "epoch": 1.9151889355396394,
+      "grad_norm": 0.43669244850265104,
+      "learning_rate": 4.906578802162008e-06,
+      "loss": 0.5539,
+      "step": 4846
+    },
+    {
+      "epoch": 1.9155840948382317,
+      "grad_norm": 0.42310364393737065,
+      "learning_rate": 4.906536368291964e-06,
+      "loss": 0.5537,
+      "step": 4847
+    },
+    {
+      "epoch": 1.915979254136824,
+      "grad_norm": 0.41308731234212065,
+      "learning_rate": 4.9064939249705066e-06,
+      "loss": 0.5426,
+      "step": 4848
+    },
+    {
+      "epoch": 1.9163744134354161,
+      "grad_norm": 0.4284823800137259,
+      "learning_rate": 4.906451472197802e-06,
+      "loss": 0.5337,
+      "step": 4849
+    },
+    {
+      "epoch": 1.9167695727340084,
+      "grad_norm": 0.44109819338095785,
+      "learning_rate": 4.906409009974018e-06,
+      "loss": 0.5352,
+      "step": 4850
+    },
+    {
+      "epoch": 1.9171647320326006,
+      "grad_norm": 0.43268349819650925,
+      "learning_rate": 4.90636653829932e-06,
+      "loss": 0.5475,
+      "step": 4851
+    },
+    {
+      "epoch": 1.9175598913311929,
+      "grad_norm": 0.430803043302547,
+      "learning_rate": 4.906324057173875e-06,
+      "loss": 0.5426,
+      "step": 4852
+    },
+    {
+      "epoch": 1.917955050629785,
+      "grad_norm": 0.4266825335162365,
+      "learning_rate": 4.9062815665978504e-06,
+      "loss": 0.5547,
+      "step": 4853
+    },
+    {
+      "epoch": 1.9183502099283773,
+      "grad_norm": 0.4498735174649945,
+      "learning_rate": 4.906239066571413e-06,
+      "loss": 0.5621,
+      "step": 4854
+    },
+    {
+      "epoch": 1.9187453692269696,
+      "grad_norm": 0.4363381817986835,
+      "learning_rate": 4.90619655709473e-06,
+      "loss": 0.5578,
+      "step": 4855
+    },
+    {
+      "epoch": 1.9191405285255618,
+      "grad_norm": 0.4320421517760889,
+      "learning_rate": 4.906154038167968e-06,
+      "loss": 0.5516,
+      "step": 4856
+    },
+    {
+      "epoch": 1.919535687824154,
+      "grad_norm": 0.4377025215430826,
+      "learning_rate": 4.9061115097912944e-06,
+      "loss": 0.5528,
+      "step": 4857
+    },
+    {
+      "epoch": 1.9199308471227463,
+      "grad_norm": 0.45155283623556963,
+      "learning_rate": 4.906068971964876e-06,
+      "loss": 0.5677,
+      "step": 4858
+    },
+    {
+      "epoch": 1.9203260064213385,
+      "grad_norm": 0.4400211424480233,
+      "learning_rate": 4.906026424688879e-06,
+      "loss": 0.562,
+      "step": 4859
+    },
+    {
+      "epoch": 1.920721165719931,
+      "grad_norm": 0.4343248105860274,
+      "learning_rate": 4.905983867963472e-06,
+      "loss": 0.5668,
+      "step": 4860
+    },
+    {
+      "epoch": 1.9211163250185233,
+      "grad_norm": 0.4353078976406567,
+      "learning_rate": 4.905941301788821e-06,
+      "loss": 0.5536,
+      "step": 4861
+    },
+    {
+      "epoch": 1.9215114843171155,
+      "grad_norm": 0.44191754785521875,
+      "learning_rate": 4.905898726165093e-06,
+      "loss": 0.5397,
+      "step": 4862
+    },
+    {
+      "epoch": 1.9219066436157077,
+      "grad_norm": 0.4319364852789675,
+      "learning_rate": 4.905856141092457e-06,
+      "loss": 0.5535,
+      "step": 4863
+    },
+    {
+      "epoch": 1.9223018029143,
+      "grad_norm": 0.4303876762850279,
+      "learning_rate": 4.9058135465710776e-06,
+      "loss": 0.5388,
+      "step": 4864
+    },
+    {
+      "epoch": 1.9226969622128922,
+      "grad_norm": 0.4264199761833403,
+      "learning_rate": 4.9057709426011236e-06,
+      "loss": 0.5561,
+      "step": 4865
+    },
+    {
+      "epoch": 1.9230921215114845,
+      "grad_norm": 0.44314277804736335,
+      "learning_rate": 4.905728329182763e-06,
+      "loss": 0.5609,
+      "step": 4866
+    },
+    {
+      "epoch": 1.9234872808100767,
+      "grad_norm": 0.4404975868482517,
+      "learning_rate": 4.905685706316162e-06,
+      "loss": 0.5404,
+      "step": 4867
+    },
+    {
+      "epoch": 1.923882440108669,
+      "grad_norm": 0.45402270420659185,
+      "learning_rate": 4.9056430740014885e-06,
+      "loss": 0.5515,
+      "step": 4868
+    },
+    {
+      "epoch": 1.9242775994072612,
+      "grad_norm": 0.43115642120987685,
+      "learning_rate": 4.90560043223891e-06,
+      "loss": 0.5809,
+      "step": 4869
+    },
+    {
+      "epoch": 1.9246727587058534,
+      "grad_norm": 0.4508455280612912,
+      "learning_rate": 4.905557781028593e-06,
+      "loss": 0.5532,
+      "step": 4870
+    },
+    {
+      "epoch": 1.9250679180044457,
+      "grad_norm": 0.42373614920434,
+      "learning_rate": 4.905515120370706e-06,
+      "loss": 0.5552,
+      "step": 4871
+    },
+    {
+      "epoch": 1.925463077303038,
+      "grad_norm": 0.5279871782928506,
+      "learning_rate": 4.905472450265416e-06,
+      "loss": 0.5716,
+      "step": 4872
+    },
+    {
+      "epoch": 1.9258582366016301,
+      "grad_norm": 0.42805135252978943,
+      "learning_rate": 4.905429770712892e-06,
+      "loss": 0.5521,
+      "step": 4873
+    },
+    {
+      "epoch": 1.9262533959002224,
+      "grad_norm": 0.4368318688771723,
+      "learning_rate": 4.9053870817133e-06,
+      "loss": 0.5486,
+      "step": 4874
+    },
+    {
+      "epoch": 1.9266485551988146,
+      "grad_norm": 0.44355571797903176,
+      "learning_rate": 4.905344383266808e-06,
+      "loss": 0.5667,
+      "step": 4875
+    },
+    {
+      "epoch": 1.9270437144974069,
+      "grad_norm": 0.42560534528870747,
+      "learning_rate": 4.9053016753735836e-06,
+      "loss": 0.5561,
+      "step": 4876
+    },
+    {
+      "epoch": 1.927438873795999,
+      "grad_norm": 0.43288235759183613,
+      "learning_rate": 4.905258958033795e-06,
+      "loss": 0.5374,
+      "step": 4877
+    },
+    {
+      "epoch": 1.9278340330945913,
+      "grad_norm": 0.42388914415273304,
+      "learning_rate": 4.90521623124761e-06,
+      "loss": 0.5719,
+      "step": 4878
+    },
+    {
+      "epoch": 1.9282291923931836,
+      "grad_norm": 0.4261428683430966,
+      "learning_rate": 4.905173495015196e-06,
+      "loss": 0.5525,
+      "step": 4879
+    },
+    {
+      "epoch": 1.9286243516917758,
+      "grad_norm": 0.4238364660888561,
+      "learning_rate": 4.9051307493367205e-06,
+      "loss": 0.5456,
+      "step": 4880
+    },
+    {
+      "epoch": 1.929019510990368,
+      "grad_norm": 0.43787334764811864,
+      "learning_rate": 4.905087994212353e-06,
+      "loss": 0.5666,
+      "step": 4881
+    },
+    {
+      "epoch": 1.9294146702889603,
+      "grad_norm": 0.41851611764846186,
+      "learning_rate": 4.9050452296422595e-06,
+      "loss": 0.5355,
+      "step": 4882
+    },
+    {
+      "epoch": 1.9298098295875525,
+      "grad_norm": 0.4221693340127139,
+      "learning_rate": 4.905002455626609e-06,
+      "loss": 0.5524,
+      "step": 4883
+    },
+    {
+      "epoch": 1.9302049888861448,
+      "grad_norm": 0.4334189440111412,
+      "learning_rate": 4.904959672165569e-06,
+      "loss": 0.5448,
+      "step": 4884
+    },
+    {
+      "epoch": 1.930600148184737,
+      "grad_norm": 0.4304554597293204,
+      "learning_rate": 4.904916879259308e-06,
+      "loss": 0.5651,
+      "step": 4885
+    },
+    {
+      "epoch": 1.9309953074833293,
+      "grad_norm": 0.4325837099787629,
+      "learning_rate": 4.904874076907994e-06,
+      "loss": 0.571,
+      "step": 4886
+    },
+    {
+      "epoch": 1.9313904667819215,
+      "grad_norm": 0.43207355005966136,
+      "learning_rate": 4.904831265111795e-06,
+      "loss": 0.5691,
+      "step": 4887
+    },
+    {
+      "epoch": 1.9317856260805137,
+      "grad_norm": 0.432638152421682,
+      "learning_rate": 4.904788443870879e-06,
+      "loss": 0.5524,
+      "step": 4888
+    },
+    {
+      "epoch": 1.932180785379106,
+      "grad_norm": 0.4512618098367983,
+      "learning_rate": 4.904745613185415e-06,
+      "loss": 0.5496,
+      "step": 4889
+    },
+    {
+      "epoch": 1.9325759446776982,
+      "grad_norm": 0.4374527551224479,
+      "learning_rate": 4.904702773055568e-06,
+      "loss": 0.5732,
+      "step": 4890
+    },
+    {
+      "epoch": 1.9329711039762905,
+      "grad_norm": 0.4468449203480439,
+      "learning_rate": 4.9046599234815105e-06,
+      "loss": 0.5587,
+      "step": 4891
+    },
+    {
+      "epoch": 1.9333662632748827,
+      "grad_norm": 0.4316133274910195,
+      "learning_rate": 4.90461706446341e-06,
+      "loss": 0.5355,
+      "step": 4892
+    },
+    {
+      "epoch": 1.933761422573475,
+      "grad_norm": 0.4246061221952136,
+      "learning_rate": 4.904574196001432e-06,
+      "loss": 0.5326,
+      "step": 4893
+    },
+    {
+      "epoch": 1.9341565818720672,
+      "grad_norm": 0.45695254659882967,
+      "learning_rate": 4.9045313180957474e-06,
+      "loss": 0.5655,
+      "step": 4894
+    },
+    {
+      "epoch": 1.9345517411706594,
+      "grad_norm": 0.44841848061126316,
+      "learning_rate": 4.904488430746524e-06,
+      "loss": 0.566,
+      "step": 4895
+    },
+    {
+      "epoch": 1.9349469004692517,
+      "grad_norm": 0.43653829446312165,
+      "learning_rate": 4.90444553395393e-06,
+      "loss": 0.5223,
+      "step": 4896
+    },
+    {
+      "epoch": 1.935342059767844,
+      "grad_norm": 0.44040644249133537,
+      "learning_rate": 4.904402627718134e-06,
+      "loss": 0.5481,
+      "step": 4897
+    },
+    {
+      "epoch": 1.9357372190664361,
+      "grad_norm": 0.4403357250362934,
+      "learning_rate": 4.904359712039304e-06,
+      "loss": 0.5652,
+      "step": 4898
+    },
+    {
+      "epoch": 1.9361323783650284,
+      "grad_norm": 0.4196447153062507,
+      "learning_rate": 4.90431678691761e-06,
+      "loss": 0.5494,
+      "step": 4899
+    },
+    {
+      "epoch": 1.9365275376636206,
+      "grad_norm": 0.4507476048268143,
+      "learning_rate": 4.904273852353219e-06,
+      "loss": 0.5388,
+      "step": 4900
+    },
+    {
+      "epoch": 1.9369226969622129,
+      "grad_norm": 0.45818820763144874,
+      "learning_rate": 4.9042309083463e-06,
+      "loss": 0.5647,
+      "step": 4901
+    },
+    {
+      "epoch": 1.937317856260805,
+      "grad_norm": 0.4169132869917713,
+      "learning_rate": 4.904187954897023e-06,
+      "loss": 0.5328,
+      "step": 4902
+    },
+    {
+      "epoch": 1.9377130155593973,
+      "grad_norm": 0.42776873057488746,
+      "learning_rate": 4.904144992005555e-06,
+      "loss": 0.5401,
+      "step": 4903
+    },
+    {
+      "epoch": 1.9381081748579896,
+      "grad_norm": 0.4452737398761506,
+      "learning_rate": 4.904102019672066e-06,
+      "loss": 0.5849,
+      "step": 4904
+    },
+    {
+      "epoch": 1.9385033341565818,
+      "grad_norm": 0.4268122814283881,
+      "learning_rate": 4.904059037896723e-06,
+      "loss": 0.5577,
+      "step": 4905
+    },
+    {
+      "epoch": 1.938898493455174,
+      "grad_norm": 0.43474540031672093,
+      "learning_rate": 4.904016046679696e-06,
+      "loss": 0.5651,
+      "step": 4906
+    },
+    {
+      "epoch": 1.9392936527537663,
+      "grad_norm": 0.4439080883548797,
+      "learning_rate": 4.9039730460211545e-06,
+      "loss": 0.5525,
+      "step": 4907
+    },
+    {
+      "epoch": 1.9396888120523585,
+      "grad_norm": 0.4380749282591096,
+      "learning_rate": 4.9039300359212665e-06,
+      "loss": 0.5647,
+      "step": 4908
+    },
+    {
+      "epoch": 1.9400839713509508,
+      "grad_norm": 0.4409470401445144,
+      "learning_rate": 4.9038870163802e-06,
+      "loss": 0.5645,
+      "step": 4909
+    },
+    {
+      "epoch": 1.940479130649543,
+      "grad_norm": 0.4453293955228133,
+      "learning_rate": 4.903843987398127e-06,
+      "loss": 0.5473,
+      "step": 4910
+    },
+    {
+      "epoch": 1.9408742899481353,
+      "grad_norm": 0.4327605271211098,
+      "learning_rate": 4.903800948975213e-06,
+      "loss": 0.5391,
+      "step": 4911
+    },
+    {
+      "epoch": 1.9412694492467275,
+      "grad_norm": 0.4436439359315798,
+      "learning_rate": 4.903757901111629e-06,
+      "loss": 0.5592,
+      "step": 4912
+    },
+    {
+      "epoch": 1.9416646085453197,
+      "grad_norm": 0.4224199241527376,
+      "learning_rate": 4.903714843807543e-06,
+      "loss": 0.5408,
+      "step": 4913
+    },
+    {
+      "epoch": 1.942059767843912,
+      "grad_norm": 0.44551596874456995,
+      "learning_rate": 4.903671777063126e-06,
+      "loss": 0.5426,
+      "step": 4914
+    },
+    {
+      "epoch": 1.9424549271425042,
+      "grad_norm": 0.4296694239782312,
+      "learning_rate": 4.9036287008785446e-06,
+      "loss": 0.5388,
+      "step": 4915
+    },
+    {
+      "epoch": 1.9428500864410965,
+      "grad_norm": 0.44248224422217924,
+      "learning_rate": 4.903585615253969e-06,
+      "loss": 0.5597,
+      "step": 4916
+    },
+    {
+      "epoch": 1.9432452457396887,
+      "grad_norm": 0.4494129495563011,
+      "learning_rate": 4.90354252018957e-06,
+      "loss": 0.5744,
+      "step": 4917
+    },
+    {
+      "epoch": 1.943640405038281,
+      "grad_norm": 0.4668938690509174,
+      "learning_rate": 4.903499415685515e-06,
+      "loss": 0.5719,
+      "step": 4918
+    },
+    {
+      "epoch": 1.9440355643368732,
+      "grad_norm": 0.4397994559767837,
+      "learning_rate": 4.903456301741973e-06,
+      "loss": 0.5566,
+      "step": 4919
+    },
+    {
+      "epoch": 1.9444307236354654,
+      "grad_norm": 0.44596460788657327,
+      "learning_rate": 4.903413178359115e-06,
+      "loss": 0.5804,
+      "step": 4920
+    },
+    {
+      "epoch": 1.9448258829340577,
+      "grad_norm": 0.43722931620391176,
+      "learning_rate": 4.9033700455371095e-06,
+      "loss": 0.5356,
+      "step": 4921
+    },
+    {
+      "epoch": 1.94522104223265,
+      "grad_norm": 0.42315317154912496,
+      "learning_rate": 4.903326903276125e-06,
+      "loss": 0.5544,
+      "step": 4922
+    },
+    {
+      "epoch": 1.9456162015312422,
+      "grad_norm": 0.43461478276395443,
+      "learning_rate": 4.903283751576333e-06,
+      "loss": 0.5385,
+      "step": 4923
+    },
+    {
+      "epoch": 1.9460113608298344,
+      "grad_norm": 0.43858622832546995,
+      "learning_rate": 4.903240590437901e-06,
+      "loss": 0.5365,
+      "step": 4924
+    },
+    {
+      "epoch": 1.9464065201284266,
+      "grad_norm": 0.4381635670286083,
+      "learning_rate": 4.903197419861e-06,
+      "loss": 0.5522,
+      "step": 4925
+    },
+    {
+      "epoch": 1.9468016794270189,
+      "grad_norm": 0.4211868434644149,
+      "learning_rate": 4.903154239845798e-06,
+      "loss": 0.5383,
+      "step": 4926
+    },
+    {
+      "epoch": 1.9471968387256111,
+      "grad_norm": 0.4367363445108265,
+      "learning_rate": 4.903111050392465e-06,
+      "loss": 0.5585,
+      "step": 4927
+    },
+    {
+      "epoch": 1.9475919980242034,
+      "grad_norm": 0.43254406845062815,
+      "learning_rate": 4.903067851501172e-06,
+      "loss": 0.5458,
+      "step": 4928
+    },
+    {
+      "epoch": 1.9479871573227956,
+      "grad_norm": 0.42229361659520137,
+      "learning_rate": 4.9030246431720875e-06,
+      "loss": 0.5438,
+      "step": 4929
+    },
+    {
+      "epoch": 1.9483823166213878,
+      "grad_norm": 0.44413774805607853,
+      "learning_rate": 4.902981425405381e-06,
+      "loss": 0.5668,
+      "step": 4930
+    },
+    {
+      "epoch": 1.9487774759199803,
+      "grad_norm": 0.4380542711297867,
+      "learning_rate": 4.902938198201223e-06,
+      "loss": 0.5631,
+      "step": 4931
+    },
+    {
+      "epoch": 1.9491726352185725,
+      "grad_norm": 0.42049828655472166,
+      "learning_rate": 4.902894961559783e-06,
+      "loss": 0.5418,
+      "step": 4932
+    },
+    {
+      "epoch": 1.9495677945171648,
+      "grad_norm": 0.42075791692900755,
+      "learning_rate": 4.90285171548123e-06,
+      "loss": 0.543,
+      "step": 4933
+    },
+    {
+      "epoch": 1.949962953815757,
+      "grad_norm": 0.43396399038359434,
+      "learning_rate": 4.9028084599657355e-06,
+      "loss": 0.5334,
+      "step": 4934
+    },
+    {
+      "epoch": 1.9503581131143493,
+      "grad_norm": 0.4238423271020982,
+      "learning_rate": 4.902765195013468e-06,
+      "loss": 0.5629,
+      "step": 4935
+    },
+    {
+      "epoch": 1.9507532724129415,
+      "grad_norm": 0.4390345402600252,
+      "learning_rate": 4.902721920624598e-06,
+      "loss": 0.5535,
+      "step": 4936
+    },
+    {
+      "epoch": 1.9511484317115337,
+      "grad_norm": 0.433652416350482,
+      "learning_rate": 4.9026786367992955e-06,
+      "loss": 0.569,
+      "step": 4937
+    },
+    {
+      "epoch": 1.951543591010126,
+      "grad_norm": 0.43072162266053016,
+      "learning_rate": 4.90263534353773e-06,
+      "loss": 0.5486,
+      "step": 4938
+    },
+    {
+      "epoch": 1.9519387503087182,
+      "grad_norm": 0.4219105439697877,
+      "learning_rate": 4.902592040840071e-06,
+      "loss": 0.5656,
+      "step": 4939
+    },
+    {
+      "epoch": 1.9523339096073105,
+      "grad_norm": 0.4397667987358108,
+      "learning_rate": 4.9025487287064905e-06,
+      "loss": 0.5836,
+      "step": 4940
+    },
+    {
+      "epoch": 1.9527290689059027,
+      "grad_norm": 0.4444122529654959,
+      "learning_rate": 4.9025054071371565e-06,
+      "loss": 0.5395,
+      "step": 4941
+    },
+    {
+      "epoch": 1.953124228204495,
+      "grad_norm": 0.446787855529709,
+      "learning_rate": 4.9024620761322415e-06,
+      "loss": 0.5595,
+      "step": 4942
+    },
+    {
+      "epoch": 1.9535193875030872,
+      "grad_norm": 0.42700864336105265,
+      "learning_rate": 4.902418735691914e-06,
+      "loss": 0.5347,
+      "step": 4943
+    },
+    {
+      "epoch": 1.9539145468016794,
+      "grad_norm": 0.44112913377219515,
+      "learning_rate": 4.902375385816344e-06,
+      "loss": 0.55,
+      "step": 4944
+    },
+    {
+      "epoch": 1.9543097061002717,
+      "grad_norm": 0.452174058015207,
+      "learning_rate": 4.902332026505703e-06,
+      "loss": 0.5553,
+      "step": 4945
+    },
+    {
+      "epoch": 1.954704865398864,
+      "grad_norm": 0.5493673198052214,
+      "learning_rate": 4.902288657760159e-06,
+      "loss": 0.5628,
+      "step": 4946
+    },
+    {
+      "epoch": 1.9551000246974561,
+      "grad_norm": 0.4388936013126185,
+      "learning_rate": 4.902245279579886e-06,
+      "loss": 0.5442,
+      "step": 4947
+    },
+    {
+      "epoch": 1.9554951839960484,
+      "grad_norm": 0.4342127172483669,
+      "learning_rate": 4.9022018919650505e-06,
+      "loss": 0.5716,
+      "step": 4948
+    },
+    {
+      "epoch": 1.9558903432946406,
+      "grad_norm": 0.4409460564547779,
+      "learning_rate": 4.902158494915825e-06,
+      "loss": 0.555,
+      "step": 4949
+    },
+    {
+      "epoch": 1.9562855025932329,
+      "grad_norm": 0.43477091609114554,
+      "learning_rate": 4.90211508843238e-06,
+      "loss": 0.5425,
+      "step": 4950
+    },
+    {
+      "epoch": 1.956680661891825,
+      "grad_norm": 0.44011026483098825,
+      "learning_rate": 4.902071672514886e-06,
+      "loss": 0.5586,
+      "step": 4951
+    },
+    {
+      "epoch": 1.9570758211904173,
+      "grad_norm": 0.4279433040735619,
+      "learning_rate": 4.902028247163512e-06,
+      "loss": 0.5603,
+      "step": 4952
+    },
+    {
+      "epoch": 1.9574709804890096,
+      "grad_norm": 0.43493371661347063,
+      "learning_rate": 4.901984812378431e-06,
+      "loss": 0.5496,
+      "step": 4953
+    },
+    {
+      "epoch": 1.9578661397876018,
+      "grad_norm": 0.429552671257518,
+      "learning_rate": 4.901941368159812e-06,
+      "loss": 0.5667,
+      "step": 4954
+    },
+    {
+      "epoch": 1.9582612990861943,
+      "grad_norm": 0.45149418076182646,
+      "learning_rate": 4.901897914507825e-06,
+      "loss": 0.5721,
+      "step": 4955
+    },
+    {
+      "epoch": 1.9586564583847865,
+      "grad_norm": 0.4440050617964623,
+      "learning_rate": 4.901854451422642e-06,
+      "loss": 0.5447,
+      "step": 4956
+    },
+    {
+      "epoch": 1.9590516176833788,
+      "grad_norm": 0.4271232711419705,
+      "learning_rate": 4.901810978904433e-06,
+      "loss": 0.5303,
+      "step": 4957
+    },
+    {
+      "epoch": 1.959446776981971,
+      "grad_norm": 0.4178784895693934,
+      "learning_rate": 4.901767496953369e-06,
+      "loss": 0.5398,
+      "step": 4958
+    },
+    {
+      "epoch": 1.9598419362805632,
+      "grad_norm": 0.45219556101457836,
+      "learning_rate": 4.90172400556962e-06,
+      "loss": 0.5464,
+      "step": 4959
+    },
+    {
+      "epoch": 1.9602370955791555,
+      "grad_norm": 0.43549913764734244,
+      "learning_rate": 4.901680504753358e-06,
+      "loss": 0.5469,
+      "step": 4960
+    },
+    {
+      "epoch": 1.9606322548777477,
+      "grad_norm": 0.4288069140164951,
+      "learning_rate": 4.901636994504754e-06,
+      "loss": 0.5438,
+      "step": 4961
+    },
+    {
+      "epoch": 1.96102741417634,
+      "grad_norm": 0.558867701364055,
+      "learning_rate": 4.901593474823978e-06,
+      "loss": 0.5632,
+      "step": 4962
+    },
+    {
+      "epoch": 1.9614225734749322,
+      "grad_norm": 0.4402971314310295,
+      "learning_rate": 4.9015499457112e-06,
+      "loss": 0.5496,
+      "step": 4963
+    },
+    {
+      "epoch": 1.9618177327735244,
+      "grad_norm": 0.4280995907810528,
+      "learning_rate": 4.901506407166594e-06,
+      "loss": 0.5323,
+      "step": 4964
+    },
+    {
+      "epoch": 1.9622128920721167,
+      "grad_norm": 0.43660734305856075,
+      "learning_rate": 4.901462859190328e-06,
+      "loss": 0.5439,
+      "step": 4965
+    },
+    {
+      "epoch": 1.962608051370709,
+      "grad_norm": 0.4364007886164183,
+      "learning_rate": 4.9014193017825735e-06,
+      "loss": 0.5532,
+      "step": 4966
+    },
+    {
+      "epoch": 1.9630032106693012,
+      "grad_norm": 0.4597936604604986,
+      "learning_rate": 4.901375734943504e-06,
+      "loss": 0.5571,
+      "step": 4967
+    },
+    {
+      "epoch": 1.9633983699678934,
+      "grad_norm": 0.45327235813160105,
+      "learning_rate": 4.901332158673288e-06,
+      "loss": 0.5574,
+      "step": 4968
+    },
+    {
+      "epoch": 1.9637935292664856,
+      "grad_norm": 0.44715737005782,
+      "learning_rate": 4.901288572972097e-06,
+      "loss": 0.579,
+      "step": 4969
+    },
+    {
+      "epoch": 1.964188688565078,
+      "grad_norm": 0.42524834539399503,
+      "learning_rate": 4.901244977840103e-06,
+      "loss": 0.5186,
+      "step": 4970
+    },
+    {
+      "epoch": 1.9645838478636701,
+      "grad_norm": 0.45527798672472874,
+      "learning_rate": 4.9012013732774765e-06,
+      "loss": 0.5373,
+      "step": 4971
+    },
+    {
+      "epoch": 1.9649790071622624,
+      "grad_norm": 0.4343605181054587,
+      "learning_rate": 4.901157759284389e-06,
+      "loss": 0.5686,
+      "step": 4972
+    },
+    {
+      "epoch": 1.9653741664608546,
+      "grad_norm": 0.44258891979240134,
+      "learning_rate": 4.901114135861013e-06,
+      "loss": 0.5697,
+      "step": 4973
+    },
+    {
+      "epoch": 1.9657693257594469,
+      "grad_norm": 0.42190562697963224,
+      "learning_rate": 4.901070503007516e-06,
+      "loss": 0.5309,
+      "step": 4974
+    },
+    {
+      "epoch": 1.966164485058039,
+      "grad_norm": 0.45195265147050434,
+      "learning_rate": 4.901026860724075e-06,
+      "loss": 0.5568,
+      "step": 4975
+    },
+    {
+      "epoch": 1.9665596443566313,
+      "grad_norm": 0.4352261875981932,
+      "learning_rate": 4.900983209010858e-06,
+      "loss": 0.5507,
+      "step": 4976
+    },
+    {
+      "epoch": 1.9669548036552236,
+      "grad_norm": 0.43978785905912626,
+      "learning_rate": 4.9009395478680355e-06,
+      "loss": 0.5599,
+      "step": 4977
+    },
+    {
+      "epoch": 1.9673499629538158,
+      "grad_norm": 0.4475750883161317,
+      "learning_rate": 4.9008958772957815e-06,
+      "loss": 0.5493,
+      "step": 4978
+    },
+    {
+      "epoch": 1.967745122252408,
+      "grad_norm": 0.4328811619048258,
+      "learning_rate": 4.900852197294266e-06,
+      "loss": 0.5361,
+      "step": 4979
+    },
+    {
+      "epoch": 1.9681402815510003,
+      "grad_norm": 0.43773523890770105,
+      "learning_rate": 4.900808507863661e-06,
+      "loss": 0.5548,
+      "step": 4980
+    },
+    {
+      "epoch": 1.9685354408495925,
+      "grad_norm": 0.43452862600059244,
+      "learning_rate": 4.900764809004138e-06,
+      "loss": 0.5336,
+      "step": 4981
+    },
+    {
+      "epoch": 1.9689306001481848,
+      "grad_norm": 0.419201512372097,
+      "learning_rate": 4.900721100715869e-06,
+      "loss": 0.551,
+      "step": 4982
+    },
+    {
+      "epoch": 1.969325759446777,
+      "grad_norm": 0.45373661373694074,
+      "learning_rate": 4.900677382999025e-06,
+      "loss": 0.5657,
+      "step": 4983
+    },
+    {
+      "epoch": 1.9697209187453693,
+      "grad_norm": 0.4370236493158466,
+      "learning_rate": 4.900633655853778e-06,
+      "loss": 0.528,
+      "step": 4984
+    },
+    {
+      "epoch": 1.9701160780439615,
+      "grad_norm": 0.4528150258511628,
+      "learning_rate": 4.9005899192803e-06,
+      "loss": 0.5427,
+      "step": 4985
+    },
+    {
+      "epoch": 1.9705112373425537,
+      "grad_norm": 0.4506026367726158,
+      "learning_rate": 4.900546173278762e-06,
+      "loss": 0.5555,
+      "step": 4986
+    },
+    {
+      "epoch": 1.970906396641146,
+      "grad_norm": 0.4207709811591268,
+      "learning_rate": 4.900502417849337e-06,
+      "loss": 0.5502,
+      "step": 4987
+    },
+    {
+      "epoch": 1.9713015559397382,
+      "grad_norm": 0.4266100877818548,
+      "learning_rate": 4.900458652992196e-06,
+      "loss": 0.5469,
+      "step": 4988
+    },
+    {
+      "epoch": 1.9716967152383305,
+      "grad_norm": 0.4387635930216414,
+      "learning_rate": 4.900414878707511e-06,
+      "loss": 0.5587,
+      "step": 4989
+    },
+    {
+      "epoch": 1.9720918745369227,
+      "grad_norm": 0.42751985993998615,
+      "learning_rate": 4.9003710949954535e-06,
+      "loss": 0.5488,
+      "step": 4990
+    },
+    {
+      "epoch": 1.972487033835515,
+      "grad_norm": 0.429513289497151,
+      "learning_rate": 4.900327301856196e-06,
+      "loss": 0.5538,
+      "step": 4991
+    },
+    {
+      "epoch": 1.9728821931341072,
+      "grad_norm": 0.42988406007517393,
+      "learning_rate": 4.9002834992899104e-06,
+      "loss": 0.5279,
+      "step": 4992
+    },
+    {
+      "epoch": 1.9732773524326994,
+      "grad_norm": 0.4380519285585141,
+      "learning_rate": 4.90023968729677e-06,
+      "loss": 0.5651,
+      "step": 4993
+    },
+    {
+      "epoch": 1.9736725117312917,
+      "grad_norm": 0.4489032418464982,
+      "learning_rate": 4.900195865876944e-06,
+      "loss": 0.5577,
+      "step": 4994
+    },
+    {
+      "epoch": 1.974067671029884,
+      "grad_norm": 0.45508873672141004,
+      "learning_rate": 4.900152035030607e-06,
+      "loss": 0.5582,
+      "step": 4995
+    },
+    {
+      "epoch": 1.9744628303284761,
+      "grad_norm": 0.43159558253407226,
+      "learning_rate": 4.90010819475793e-06,
+      "loss": 0.5623,
+      "step": 4996
+    },
+    {
+      "epoch": 1.9748579896270684,
+      "grad_norm": 0.4372402377219953,
+      "learning_rate": 4.900064345059086e-06,
+      "loss": 0.5519,
+      "step": 4997
+    },
+    {
+      "epoch": 1.9752531489256606,
+      "grad_norm": 0.4627096781522281,
+      "learning_rate": 4.900020485934245e-06,
+      "loss": 0.5693,
+      "step": 4998
+    },
+    {
+      "epoch": 1.9756483082242529,
+      "grad_norm": 0.4518788906698329,
+      "learning_rate": 4.899976617383583e-06,
+      "loss": 0.5577,
+      "step": 4999
+    },
+    {
+      "epoch": 1.976043467522845,
+      "grad_norm": 0.44234025817296496,
+      "learning_rate": 4.899932739407268e-06,
+      "loss": 0.552,
+      "step": 5000
+    },
+    {
+      "epoch": 1.9764386268214373,
+      "grad_norm": 0.42902794698726177,
+      "learning_rate": 4.899888852005477e-06,
+      "loss": 0.535,
+      "step": 5001
+    },
+    {
+      "epoch": 1.9768337861200296,
+      "grad_norm": 0.4337101022811792,
+      "learning_rate": 4.899844955178378e-06,
+      "loss": 0.5477,
+      "step": 5002
+    },
+    {
+      "epoch": 1.9772289454186218,
+      "grad_norm": 0.4334561135618806,
+      "learning_rate": 4.899801048926146e-06,
+      "loss": 0.5536,
+      "step": 5003
+    },
+    {
+      "epoch": 1.977624104717214,
+      "grad_norm": 0.4539433104177194,
+      "learning_rate": 4.899757133248953e-06,
+      "loss": 0.5634,
+      "step": 5004
+    },
+    {
+      "epoch": 1.9780192640158063,
+      "grad_norm": 0.46398410451302946,
+      "learning_rate": 4.89971320814697e-06,
+      "loss": 0.5579,
+      "step": 5005
+    },
+    {
+      "epoch": 1.9784144233143985,
+      "grad_norm": 0.4224716161481286,
+      "learning_rate": 4.899669273620372e-06,
+      "loss": 0.5213,
+      "step": 5006
+    },
+    {
+      "epoch": 1.9788095826129908,
+      "grad_norm": 0.43890124660317376,
+      "learning_rate": 4.899625329669329e-06,
+      "loss": 0.5585,
+      "step": 5007
+    },
+    {
+      "epoch": 1.979204741911583,
+      "grad_norm": 0.45718913393571137,
+      "learning_rate": 4.899581376294016e-06,
+      "loss": 0.5739,
+      "step": 5008
+    },
+    {
+      "epoch": 1.9795999012101753,
+      "grad_norm": 0.4274912725996382,
+      "learning_rate": 4.899537413494604e-06,
+      "loss": 0.5284,
+      "step": 5009
+    },
+    {
+      "epoch": 1.9799950605087675,
+      "grad_norm": 0.4263688372409198,
+      "learning_rate": 4.899493441271266e-06,
+      "loss": 0.5371,
+      "step": 5010
+    },
+    {
+      "epoch": 1.9803902198073597,
+      "grad_norm": 0.42390616403436115,
+      "learning_rate": 4.899449459624175e-06,
+      "loss": 0.5525,
+      "step": 5011
+    },
+    {
+      "epoch": 1.980785379105952,
+      "grad_norm": 0.439575208819799,
+      "learning_rate": 4.899405468553503e-06,
+      "loss": 0.5596,
+      "step": 5012
+    },
+    {
+      "epoch": 1.9811805384045442,
+      "grad_norm": 0.438514692310827,
+      "learning_rate": 4.899361468059424e-06,
+      "loss": 0.5656,
+      "step": 5013
+    },
+    {
+      "epoch": 1.9815756977031365,
+      "grad_norm": 0.4286665241622738,
+      "learning_rate": 4.8993174581421095e-06,
+      "loss": 0.5462,
+      "step": 5014
+    },
+    {
+      "epoch": 1.9819708570017287,
+      "grad_norm": 0.4310129047836364,
+      "learning_rate": 4.899273438801734e-06,
+      "loss": 0.5717,
+      "step": 5015
+    },
+    {
+      "epoch": 1.982366016300321,
+      "grad_norm": 0.4410419994536329,
+      "learning_rate": 4.899229410038468e-06,
+      "loss": 0.5432,
+      "step": 5016
+    },
+    {
+      "epoch": 1.9827611755989132,
+      "grad_norm": 0.4421044202734814,
+      "learning_rate": 4.899185371852487e-06,
+      "loss": 0.552,
+      "step": 5017
+    },
+    {
+      "epoch": 1.9831563348975054,
+      "grad_norm": 0.43915212841424395,
+      "learning_rate": 4.899141324243962e-06,
+      "loss": 0.5384,
+      "step": 5018
+    },
+    {
+      "epoch": 1.9835514941960977,
+      "grad_norm": 0.42929780686144275,
+      "learning_rate": 4.8990972672130675e-06,
+      "loss": 0.5467,
+      "step": 5019
+    },
+    {
+      "epoch": 1.98394665349469,
+      "grad_norm": 0.44257765457839277,
+      "learning_rate": 4.899053200759975e-06,
+      "loss": 0.5723,
+      "step": 5020
+    },
+    {
+      "epoch": 1.9843418127932821,
+      "grad_norm": 0.4491299367521981,
+      "learning_rate": 4.8990091248848586e-06,
+      "loss": 0.56,
+      "step": 5021
+    },
+    {
+      "epoch": 1.9847369720918744,
+      "grad_norm": 0.42594398347130197,
+      "learning_rate": 4.898965039587891e-06,
+      "loss": 0.544,
+      "step": 5022
+    },
+    {
+      "epoch": 1.9851321313904666,
+      "grad_norm": 0.430300760036231,
+      "learning_rate": 4.898920944869245e-06,
+      "loss": 0.542,
+      "step": 5023
+    },
+    {
+      "epoch": 1.9855272906890589,
+      "grad_norm": 0.43273660717146456,
+      "learning_rate": 4.898876840729095e-06,
+      "loss": 0.5361,
+      "step": 5024
+    },
+    {
+      "epoch": 1.985922449987651,
+      "grad_norm": 0.44664865716579133,
+      "learning_rate": 4.898832727167613e-06,
+      "loss": 0.5703,
+      "step": 5025
+    },
+    {
+      "epoch": 1.9863176092862436,
+      "grad_norm": 0.4268197117087961,
+      "learning_rate": 4.898788604184973e-06,
+      "loss": 0.5773,
+      "step": 5026
+    },
+    {
+      "epoch": 1.9867127685848358,
+      "grad_norm": 0.4453737100715003,
+      "learning_rate": 4.8987444717813475e-06,
+      "loss": 0.5687,
+      "step": 5027
+    },
+    {
+      "epoch": 1.987107927883428,
+      "grad_norm": 0.43272023462491616,
+      "learning_rate": 4.898700329956911e-06,
+      "loss": 0.558,
+      "step": 5028
+    },
+    {
+      "epoch": 1.9875030871820203,
+      "grad_norm": 0.4264184633199343,
+      "learning_rate": 4.898656178711836e-06,
+      "loss": 0.5548,
+      "step": 5029
+    },
+    {
+      "epoch": 1.9878982464806125,
+      "grad_norm": 0.4221266341561921,
+      "learning_rate": 4.898612018046296e-06,
+      "loss": 0.5594,
+      "step": 5030
+    },
+    {
+      "epoch": 1.9882934057792048,
+      "grad_norm": 0.4347414491375779,
+      "learning_rate": 4.898567847960463e-06,
+      "loss": 0.5357,
+      "step": 5031
+    },
+    {
+      "epoch": 1.988688565077797,
+      "grad_norm": 0.43586397162379814,
+      "learning_rate": 4.898523668454514e-06,
+      "loss": 0.5446,
+      "step": 5032
+    },
+    {
+      "epoch": 1.9890837243763893,
+      "grad_norm": 0.5110052342258178,
+      "learning_rate": 4.89847947952862e-06,
+      "loss": 0.5537,
+      "step": 5033
+    },
+    {
+      "epoch": 1.9894788836749815,
+      "grad_norm": 0.4451068478930137,
+      "learning_rate": 4.898435281182955e-06,
+      "loss": 0.5476,
+      "step": 5034
+    },
+    {
+      "epoch": 1.9898740429735737,
+      "grad_norm": 0.4416455338525789,
+      "learning_rate": 4.898391073417692e-06,
+      "loss": 0.5623,
+      "step": 5035
+    },
+    {
+      "epoch": 1.990269202272166,
+      "grad_norm": 0.4258928324683576,
+      "learning_rate": 4.898346856233006e-06,
+      "loss": 0.5516,
+      "step": 5036
+    },
+    {
+      "epoch": 1.9906643615707582,
+      "grad_norm": 0.43772124082226155,
+      "learning_rate": 4.89830262962907e-06,
+      "loss": 0.5732,
+      "step": 5037
+    },
+    {
+      "epoch": 1.9910595208693505,
+      "grad_norm": 0.44967948499329063,
+      "learning_rate": 4.898258393606057e-06,
+      "loss": 0.5642,
+      "step": 5038
+    },
+    {
+      "epoch": 1.9914546801679427,
+      "grad_norm": 0.43814703911183817,
+      "learning_rate": 4.898214148164142e-06,
+      "loss": 0.5469,
+      "step": 5039
+    },
+    {
+      "epoch": 1.991849839466535,
+      "grad_norm": 0.4230077406376979,
+      "learning_rate": 4.898169893303497e-06,
+      "loss": 0.5466,
+      "step": 5040
+    },
+    {
+      "epoch": 1.9922449987651272,
+      "grad_norm": 0.43628697564315577,
+      "learning_rate": 4.898125629024298e-06,
+      "loss": 0.5445,
+      "step": 5041
+    },
+    {
+      "epoch": 1.9926401580637194,
+      "grad_norm": 0.45113306979240975,
+      "learning_rate": 4.898081355326717e-06,
+      "loss": 0.5658,
+      "step": 5042
+    },
+    {
+      "epoch": 1.9930353173623117,
+      "grad_norm": 0.4334951853126296,
+      "learning_rate": 4.898037072210929e-06,
+      "loss": 0.557,
+      "step": 5043
+    },
+    {
+      "epoch": 1.993430476660904,
+      "grad_norm": 0.42752413492781205,
+      "learning_rate": 4.897992779677108e-06,
+      "loss": 0.5546,
+      "step": 5044
+    },
+    {
+      "epoch": 1.9938256359594961,
+      "grad_norm": 0.4435208315687634,
+      "learning_rate": 4.8979484777254275e-06,
+      "loss": 0.5583,
+      "step": 5045
+    },
+    {
+      "epoch": 1.9942207952580884,
+      "grad_norm": 0.4487454343438756,
+      "learning_rate": 4.89790416635606e-06,
+      "loss": 0.55,
+      "step": 5046
+    },
+    {
+      "epoch": 1.9946159545566806,
+      "grad_norm": 0.4504942707382518,
+      "learning_rate": 4.8978598455691825e-06,
+      "loss": 0.5575,
+      "step": 5047
+    },
+    {
+      "epoch": 1.9950111138552729,
+      "grad_norm": 0.45640354880823075,
+      "learning_rate": 4.8978155153649674e-06,
+      "loss": 0.5693,
+      "step": 5048
+    },
+    {
+      "epoch": 1.9954062731538653,
+      "grad_norm": 0.442089413525396,
+      "learning_rate": 4.897771175743588e-06,
+      "loss": 0.559,
+      "step": 5049
+    },
+    {
+      "epoch": 1.9958014324524576,
+      "grad_norm": 0.45021406883084675,
+      "learning_rate": 4.89772682670522e-06,
+      "loss": 0.566,
+      "step": 5050
+    },
+    {
+      "epoch": 1.9961965917510498,
+      "grad_norm": 0.45325350234496126,
+      "learning_rate": 4.897682468250038e-06,
+      "loss": 0.5639,
+      "step": 5051
+    },
+    {
+      "epoch": 1.996591751049642,
+      "grad_norm": 0.4320207290666673,
+      "learning_rate": 4.897638100378214e-06,
+      "loss": 0.559,
+      "step": 5052
+    },
+    {
+      "epoch": 1.9969869103482343,
+      "grad_norm": 0.4471743726766185,
+      "learning_rate": 4.897593723089924e-06,
+      "loss": 0.5859,
+      "step": 5053
+    },
+    {
+      "epoch": 1.9973820696468265,
+      "grad_norm": 0.4526408463021309,
+      "learning_rate": 4.897549336385341e-06,
+      "loss": 0.562,
+      "step": 5054
+    },
+    {
+      "epoch": 1.9977772289454188,
+      "grad_norm": 0.4204803007502452,
+      "learning_rate": 4.897504940264641e-06,
+      "loss": 0.544,
+      "step": 5055
+    },
+    {
+      "epoch": 1.998172388244011,
+      "grad_norm": 0.4293273991156161,
+      "learning_rate": 4.897460534727997e-06,
+      "loss": 0.5356,
+      "step": 5056
+    },
+    {
+      "epoch": 1.9985675475426032,
+      "grad_norm": 0.43667313213086545,
+      "learning_rate": 4.897416119775584e-06,
+      "loss": 0.5415,
+      "step": 5057
+    },
+    {
+      "epoch": 1.9989627068411955,
+      "grad_norm": 0.4463243145959753,
+      "learning_rate": 4.897371695407576e-06,
+      "loss": 0.5554,
+      "step": 5058
+    },
+    {
+      "epoch": 1.9993578661397877,
+      "grad_norm": 0.4449190372166922,
+      "learning_rate": 4.897327261624148e-06,
+      "loss": 0.5541,
+      "step": 5059
+    },
+    {
+      "epoch": 1.99975302543838,
+      "grad_norm": 0.4548254873381931,
+      "learning_rate": 4.897282818425474e-06,
+      "loss": 0.5736,
+      "step": 5060
+    },
+    {
+      "epoch": 2.000148184736972,
+      "grad_norm": 0.45300066146970297,
+      "learning_rate": 4.89723836581173e-06,
+      "loss": 0.5866,
+      "step": 5061
+    },
+    {
+      "epoch": 2.0005433440355644,
+      "grad_norm": 0.44240572381302623,
+      "learning_rate": 4.897193903783087e-06,
+      "loss": 0.5543,
+      "step": 5062
+    },
+    {
+      "epoch": 2.0009385033341567,
+      "grad_norm": 0.4507961288693878,
+      "learning_rate": 4.8971494323397236e-06,
+      "loss": 0.5534,
+      "step": 5063
+    },
+    {
+      "epoch": 2.001333662632749,
+      "grad_norm": 0.42461245435649125,
+      "learning_rate": 4.897104951481813e-06,
+      "loss": 0.546,
+      "step": 5064
+    },
+    {
+      "epoch": 2.001728821931341,
+      "grad_norm": 0.43117780751367724,
+      "learning_rate": 4.897060461209529e-06,
+      "loss": 0.5413,
+      "step": 5065
+    },
+    {
+      "epoch": 2.0021239812299334,
+      "grad_norm": 0.4590693569061441,
+      "learning_rate": 4.8970159615230476e-06,
+      "loss": 0.5553,
+      "step": 5066
+    },
+    {
+      "epoch": 2.0025191405285256,
+      "grad_norm": 0.43910645784012625,
+      "learning_rate": 4.896971452422543e-06,
+      "loss": 0.5453,
+      "step": 5067
+    },
+    {
+      "epoch": 2.002914299827118,
+      "grad_norm": 0.4491473519263672,
+      "learning_rate": 4.89692693390819e-06,
+      "loss": 0.5344,
+      "step": 5068
+    },
+    {
+      "epoch": 2.00330945912571,
+      "grad_norm": 0.4558312418893952,
+      "learning_rate": 4.896882405980164e-06,
+      "loss": 0.5424,
+      "step": 5069
+    },
+    {
+      "epoch": 2.0037046184243024,
+      "grad_norm": 0.4342578385234224,
+      "learning_rate": 4.896837868638638e-06,
+      "loss": 0.5609,
+      "step": 5070
+    },
+    {
+      "epoch": 2.0040997777228946,
+      "grad_norm": 0.472485898038852,
+      "learning_rate": 4.896793321883789e-06,
+      "loss": 0.5507,
+      "step": 5071
+    },
+    {
+      "epoch": 2.004494937021487,
+      "grad_norm": 0.43961186761669674,
+      "learning_rate": 4.896748765715792e-06,
+      "loss": 0.5627,
+      "step": 5072
+    },
+    {
+      "epoch": 2.004890096320079,
+      "grad_norm": 0.5682972949137233,
+      "learning_rate": 4.89670420013482e-06,
+      "loss": 0.5572,
+      "step": 5073
+    },
+    {
+      "epoch": 2.0052852556186713,
+      "grad_norm": 0.4414444744926705,
+      "learning_rate": 4.89665962514105e-06,
+      "loss": 0.5509,
+      "step": 5074
+    },
+    {
+      "epoch": 2.0056804149172636,
+      "grad_norm": 0.41346400399643507,
+      "learning_rate": 4.896615040734656e-06,
+      "loss": 0.5201,
+      "step": 5075
+    },
+    {
+      "epoch": 2.006075574215856,
+      "grad_norm": 0.4444163402767339,
+      "learning_rate": 4.896570446915814e-06,
+      "loss": 0.5373,
+      "step": 5076
+    },
+    {
+      "epoch": 2.006470733514448,
+      "grad_norm": 0.4296320259295841,
+      "learning_rate": 4.896525843684698e-06,
+      "loss": 0.5574,
+      "step": 5077
+    },
+    {
+      "epoch": 2.0068658928130403,
+      "grad_norm": 0.4324921273129445,
+      "learning_rate": 4.896481231041483e-06,
+      "loss": 0.5597,
+      "step": 5078
+    },
+    {
+      "epoch": 2.0072610521116325,
+      "grad_norm": 0.5733673612744304,
+      "learning_rate": 4.896436608986347e-06,
+      "loss": 0.5648,
+      "step": 5079
+    },
+    {
+      "epoch": 2.0076562114102248,
+      "grad_norm": 0.42676323466611693,
+      "learning_rate": 4.896391977519461e-06,
+      "loss": 0.5537,
+      "step": 5080
+    },
+    {
+      "epoch": 2.008051370708817,
+      "grad_norm": 0.4430572425107361,
+      "learning_rate": 4.896347336641004e-06,
+      "loss": 0.5509,
+      "step": 5081
+    },
+    {
+      "epoch": 2.0084465300074092,
+      "grad_norm": 0.45815046775746454,
+      "learning_rate": 4.896302686351149e-06,
+      "loss": 0.5684,
+      "step": 5082
+    },
+    {
+      "epoch": 2.0088416893060015,
+      "grad_norm": 0.4498740575972979,
+      "learning_rate": 4.896258026650072e-06,
+      "loss": 0.5895,
+      "step": 5083
+    },
+    {
+      "epoch": 2.0092368486045937,
+      "grad_norm": 0.4208347192689959,
+      "learning_rate": 4.89621335753795e-06,
+      "loss": 0.546,
+      "step": 5084
+    },
+    {
+      "epoch": 2.009632007903186,
+      "grad_norm": 0.4859059756580774,
+      "learning_rate": 4.8961686790149554e-06,
+      "loss": 0.5585,
+      "step": 5085
+    },
+    {
+      "epoch": 2.010027167201778,
+      "grad_norm": 0.43391858881607787,
+      "learning_rate": 4.896123991081266e-06,
+      "loss": 0.563,
+      "step": 5086
+    },
+    {
+      "epoch": 2.0104223265003704,
+      "grad_norm": 0.4314428130560567,
+      "learning_rate": 4.8960792937370565e-06,
+      "loss": 0.544,
+      "step": 5087
+    },
+    {
+      "epoch": 2.0108174857989627,
+      "grad_norm": 0.42222335118568166,
+      "learning_rate": 4.896034586982502e-06,
+      "loss": 0.5453,
+      "step": 5088
+    },
+    {
+      "epoch": 2.011212645097555,
+      "grad_norm": 0.4360640333190303,
+      "learning_rate": 4.89598987081778e-06,
+      "loss": 0.5461,
+      "step": 5089
+    },
+    {
+      "epoch": 2.011607804396147,
+      "grad_norm": 0.44026271873516537,
+      "learning_rate": 4.8959451452430635e-06,
+      "loss": 0.5541,
+      "step": 5090
+    },
+    {
+      "epoch": 2.0120029636947394,
+      "grad_norm": 0.42715031631521877,
+      "learning_rate": 4.895900410258529e-06,
+      "loss": 0.537,
+      "step": 5091
+    },
+    {
+      "epoch": 2.0123981229933317,
+      "grad_norm": 0.41412343827373277,
+      "learning_rate": 4.8958556658643535e-06,
+      "loss": 0.5494,
+      "step": 5092
+    },
+    {
+      "epoch": 2.012793282291924,
+      "grad_norm": 0.4358377535786107,
+      "learning_rate": 4.8958109120607115e-06,
+      "loss": 0.5513,
+      "step": 5093
+    },
+    {
+      "epoch": 2.013188441590516,
+      "grad_norm": 0.42454225257888584,
+      "learning_rate": 4.895766148847779e-06,
+      "loss": 0.542,
+      "step": 5094
+    },
+    {
+      "epoch": 2.0135836008891084,
+      "grad_norm": 0.42998865577635037,
+      "learning_rate": 4.895721376225732e-06,
+      "loss": 0.5523,
+      "step": 5095
+    },
+    {
+      "epoch": 2.0139787601877006,
+      "grad_norm": 0.4344501069089372,
+      "learning_rate": 4.8956765941947456e-06,
+      "loss": 0.5486,
+      "step": 5096
+    },
+    {
+      "epoch": 2.014373919486293,
+      "grad_norm": 0.43353826879791363,
+      "learning_rate": 4.895631802754997e-06,
+      "loss": 0.543,
+      "step": 5097
+    },
+    {
+      "epoch": 2.014769078784885,
+      "grad_norm": 0.4575411830513003,
+      "learning_rate": 4.895587001906661e-06,
+      "loss": 0.5689,
+      "step": 5098
+    },
+    {
+      "epoch": 2.0151642380834773,
+      "grad_norm": 0.42786800682913717,
+      "learning_rate": 4.895542191649914e-06,
+      "loss": 0.5578,
+      "step": 5099
+    },
+    {
+      "epoch": 2.0155593973820696,
+      "grad_norm": 0.4964495652383933,
+      "learning_rate": 4.895497371984932e-06,
+      "loss": 0.564,
+      "step": 5100
+    },
+    {
+      "epoch": 2.015954556680662,
+      "grad_norm": 0.42115240187280495,
+      "learning_rate": 4.895452542911891e-06,
+      "loss": 0.5498,
+      "step": 5101
+    },
+    {
+      "epoch": 2.016349715979254,
+      "grad_norm": 0.42588276441255346,
+      "learning_rate": 4.895407704430967e-06,
+      "loss": 0.5556,
+      "step": 5102
+    },
+    {
+      "epoch": 2.0167448752778463,
+      "grad_norm": 0.4302156099061691,
+      "learning_rate": 4.895362856542336e-06,
+      "loss": 0.5442,
+      "step": 5103
+    },
+    {
+      "epoch": 2.0171400345764385,
+      "grad_norm": 0.43604945564161973,
+      "learning_rate": 4.895317999246174e-06,
+      "loss": 0.5656,
+      "step": 5104
+    },
+    {
+      "epoch": 2.0000987898246483,
+      "grad_norm": 0.44789874867544277,
+      "learning_rate": 4.895273132542658e-06,
+      "loss": 0.5109,
+      "step": 5105
+    },
+    {
+      "epoch": 2.0004939491232405,
+      "grad_norm": 0.7893264350562275,
+      "learning_rate": 4.895228256431963e-06,
+      "loss": 0.4954,
+      "step": 5106
+    },
+    {
+      "epoch": 2.0008891084218328,
+      "grad_norm": 0.5938590122165337,
+      "learning_rate": 4.895183370914267e-06,
+      "loss": 0.504,
+      "step": 5107
+    },
+    {
+      "epoch": 2.001284267720425,
+      "grad_norm": 0.5235383132112481,
+      "learning_rate": 4.8951384759897435e-06,
+      "loss": 0.5104,
+      "step": 5108
+    },
+    {
+      "epoch": 2.0016794270190172,
+      "grad_norm": 0.6882530599009438,
+      "learning_rate": 4.895093571658571e-06,
+      "loss": 0.4983,
+      "step": 5109
+    },
+    {
+      "epoch": 2.0020745863176095,
+      "grad_norm": 0.7510915200577702,
+      "learning_rate": 4.895048657920926e-06,
+      "loss": 0.4983,
+      "step": 5110
+    },
+    {
+      "epoch": 2.0024697456162017,
+      "grad_norm": 0.6065619194319212,
+      "learning_rate": 4.895003734776984e-06,
+      "loss": 0.49,
+      "step": 5111
+    },
+    {
+      "epoch": 2.002864904914794,
+      "grad_norm": 0.5356219470458906,
+      "learning_rate": 4.894958802226921e-06,
+      "loss": 0.4856,
+      "step": 5112
+    },
+    {
+      "epoch": 2.003260064213386,
+      "grad_norm": 0.5780607077723401,
+      "learning_rate": 4.894913860270915e-06,
+      "loss": 0.4935,
+      "step": 5113
+    },
+    {
+      "epoch": 2.0036552235119784,
+      "grad_norm": 0.5595862883969074,
+      "learning_rate": 4.8948689089091414e-06,
+      "loss": 0.4943,
+      "step": 5114
+    },
+    {
+      "epoch": 2.0040503828105707,
+      "grad_norm": 0.5217610935709599,
+      "learning_rate": 4.8948239481417766e-06,
+      "loss": 0.5112,
+      "step": 5115
+    },
+    {
+      "epoch": 2.004445542109163,
+      "grad_norm": 0.49305801115666537,
+      "learning_rate": 4.894778977968998e-06,
+      "loss": 0.4907,
+      "step": 5116
+    },
+    {
+      "epoch": 2.004840701407755,
+      "grad_norm": 0.530247044656353,
+      "learning_rate": 4.894733998390982e-06,
+      "loss": 0.4904,
+      "step": 5117
+    },
+    {
+      "epoch": 2.0052358607063474,
+      "grad_norm": 0.552426427747116,
+      "learning_rate": 4.894689009407903e-06,
+      "loss": 0.49,
+      "step": 5118
+    },
+    {
+      "epoch": 2.0056310200049396,
+      "grad_norm": 0.5685491982079328,
+      "learning_rate": 4.894644011019942e-06,
+      "loss": 0.4948,
+      "step": 5119
+    },
+    {
+      "epoch": 2.006026179303532,
+      "grad_norm": 0.502595184515819,
+      "learning_rate": 4.894599003227273e-06,
+      "loss": 0.4962,
+      "step": 5120
+    },
+    {
+      "epoch": 2.006421338602124,
+      "grad_norm": 0.5130021858315302,
+      "learning_rate": 4.8945539860300725e-06,
+      "loss": 0.5107,
+      "step": 5121
+    },
+    {
+      "epoch": 2.0068164979007164,
+      "grad_norm": 0.5350797674291279,
+      "learning_rate": 4.8945089594285185e-06,
+      "loss": 0.5076,
+      "step": 5122
+    },
+    {
+      "epoch": 2.0072116571993086,
+      "grad_norm": 0.49090717650738347,
+      "learning_rate": 4.894463923422787e-06,
+      "loss": 0.4939,
+      "step": 5123
+    },
+    {
+      "epoch": 2.007606816497901,
+      "grad_norm": 0.48813921337582705,
+      "learning_rate": 4.8944188780130555e-06,
+      "loss": 0.5004,
+      "step": 5124
+    },
+    {
+      "epoch": 2.008001975796493,
+      "grad_norm": 0.4912844935858703,
+      "learning_rate": 4.8943738231995005e-06,
+      "loss": 0.4914,
+      "step": 5125
+    },
+    {
+      "epoch": 2.0083971350950853,
+      "grad_norm": 0.5117780002863588,
+      "learning_rate": 4.894328758982299e-06,
+      "loss": 0.4892,
+      "step": 5126
+    },
+    {
+      "epoch": 2.0087922943936776,
+      "grad_norm": 0.4929822170606211,
+      "learning_rate": 4.894283685361628e-06,
+      "loss": 0.5052,
+      "step": 5127
+    },
+    {
+      "epoch": 2.00918745369227,
+      "grad_norm": 0.47940057487781756,
+      "learning_rate": 4.894238602337665e-06,
+      "loss": 0.4986,
+      "step": 5128
+    },
+    {
+      "epoch": 2.009582612990862,
+      "grad_norm": 0.4738863333886923,
+      "learning_rate": 4.894193509910586e-06,
+      "loss": 0.4841,
+      "step": 5129
+    },
+    {
+      "epoch": 2.0099777722894543,
+      "grad_norm": 0.470276754348612,
+      "learning_rate": 4.8941484080805695e-06,
+      "loss": 0.5098,
+      "step": 5130
+    },
+    {
+      "epoch": 2.0103729315880465,
+      "grad_norm": 0.4767147238302886,
+      "learning_rate": 4.8941032968477914e-06,
+      "loss": 0.4909,
+      "step": 5131
+    },
+    {
+      "epoch": 2.0107680908866388,
+      "grad_norm": 0.4721168271949362,
+      "learning_rate": 4.894058176212429e-06,
+      "loss": 0.484,
+      "step": 5132
+    },
+    {
+      "epoch": 2.011163250185231,
+      "grad_norm": 0.4694032432392925,
+      "learning_rate": 4.89401304617466e-06,
+      "loss": 0.4952,
+      "step": 5133
+    },
+    {
+      "epoch": 2.0115584094838233,
+      "grad_norm": 0.5016813592368355,
+      "learning_rate": 4.8939679067346625e-06,
+      "loss": 0.4858,
+      "step": 5134
+    },
+    {
+      "epoch": 2.0119535687824155,
+      "grad_norm": 0.4764925899996865,
+      "learning_rate": 4.893922757892612e-06,
+      "loss": 0.4997,
+      "step": 5135
+    },
+    {
+      "epoch": 2.0123487280810077,
+      "grad_norm": 0.48398112841687313,
+      "learning_rate": 4.893877599648686e-06,
+      "loss": 0.5019,
+      "step": 5136
+    },
+    {
+      "epoch": 2.0127438873796,
+      "grad_norm": 0.47136229426376053,
+      "learning_rate": 4.893832432003062e-06,
+      "loss": 0.4867,
+      "step": 5137
+    },
+    {
+      "epoch": 2.013139046678192,
+      "grad_norm": 0.48191425839984575,
+      "learning_rate": 4.893787254955919e-06,
+      "loss": 0.4825,
+      "step": 5138
+    },
+    {
+      "epoch": 2.0135342059767845,
+      "grad_norm": 0.4670311308730599,
+      "learning_rate": 4.893742068507434e-06,
+      "loss": 0.4951,
+      "step": 5139
+    },
+    {
+      "epoch": 2.0139293652753767,
+      "grad_norm": 0.4890864968827146,
+      "learning_rate": 4.893696872657782e-06,
+      "loss": 0.5113,
+      "step": 5140
+    },
+    {
+      "epoch": 2.014324524573969,
+      "grad_norm": 0.4584922074591445,
+      "learning_rate": 4.893651667407143e-06,
+      "loss": 0.4888,
+      "step": 5141
+    },
+    {
+      "epoch": 2.014719683872561,
+      "grad_norm": 0.48036994938581246,
+      "learning_rate": 4.893606452755693e-06,
+      "loss": 0.4916,
+      "step": 5142
+    },
+    {
+      "epoch": 2.0151148431711534,
+      "grad_norm": 0.6649796099167191,
+      "learning_rate": 4.893561228703611e-06,
+      "loss": 0.5061,
+      "step": 5143
+    },
+    {
+      "epoch": 2.0155100024697457,
+      "grad_norm": 0.46584443556518795,
+      "learning_rate": 4.8935159952510745e-06,
+      "loss": 0.4786,
+      "step": 5144
+    },
+    {
+      "epoch": 2.015905161768338,
+      "grad_norm": 0.46787656590183857,
+      "learning_rate": 4.893470752398261e-06,
+      "loss": 0.4822,
+      "step": 5145
+    },
+    {
+      "epoch": 2.01630032106693,
+      "grad_norm": 0.46824424287806576,
+      "learning_rate": 4.893425500145346e-06,
+      "loss": 0.5001,
+      "step": 5146
+    },
+    {
+      "epoch": 2.0166954803655224,
+      "grad_norm": 0.49194078179454165,
+      "learning_rate": 4.89338023849251e-06,
+      "loss": 0.4799,
+      "step": 5147
+    },
+    {
+      "epoch": 2.0170906396641146,
+      "grad_norm": 0.476320179653451,
+      "learning_rate": 4.893334967439929e-06,
+      "loss": 0.5194,
+      "step": 5148
+    },
+    {
+      "epoch": 2.017485798962707,
+      "grad_norm": 0.4780909181467547,
+      "learning_rate": 4.893289686987782e-06,
+      "loss": 0.5032,
+      "step": 5149
+    },
+    {
+      "epoch": 2.017880958261299,
+      "grad_norm": 0.4548175332915629,
+      "learning_rate": 4.893244397136247e-06,
+      "loss": 0.4723,
+      "step": 5150
+    },
+    {
+      "epoch": 2.0182761175598913,
+      "grad_norm": 0.44903119971257,
+      "learning_rate": 4.8931990978855005e-06,
+      "loss": 0.4873,
+      "step": 5151
+    },
+    {
+      "epoch": 2.0186712768584836,
+      "grad_norm": 0.46396916970746166,
+      "learning_rate": 4.893153789235722e-06,
+      "loss": 0.5,
+      "step": 5152
+    },
+    {
+      "epoch": 2.019066436157076,
+      "grad_norm": 0.4569485462711414,
+      "learning_rate": 4.893108471187088e-06,
+      "loss": 0.5015,
+      "step": 5153
+    },
+    {
+      "epoch": 2.019461595455668,
+      "grad_norm": 0.45922275497265813,
+      "learning_rate": 4.893063143739777e-06,
+      "loss": 0.502,
+      "step": 5154
+    },
+    {
+      "epoch": 2.0198567547542603,
+      "grad_norm": 0.4695075729482603,
+      "learning_rate": 4.893017806893967e-06,
+      "loss": 0.5134,
+      "step": 5155
+    },
+    {
+      "epoch": 2.0202519140528525,
+      "grad_norm": 0.4530917617622478,
+      "learning_rate": 4.892972460649836e-06,
+      "loss": 0.5076,
+      "step": 5156
+    },
+    {
+      "epoch": 2.020647073351445,
+      "grad_norm": 0.4504342750468262,
+      "learning_rate": 4.892927105007563e-06,
+      "loss": 0.505,
+      "step": 5157
+    },
+    {
+      "epoch": 2.021042232650037,
+      "grad_norm": 0.4667012843063457,
+      "learning_rate": 4.892881739967325e-06,
+      "loss": 0.5167,
+      "step": 5158
+    },
+    {
+      "epoch": 2.0214373919486293,
+      "grad_norm": 0.4494930932302058,
+      "learning_rate": 4.892836365529301e-06,
+      "loss": 0.4803,
+      "step": 5159
+    },
+    {
+      "epoch": 2.0218325512472215,
+      "grad_norm": 0.4588896369546433,
+      "learning_rate": 4.892790981693668e-06,
+      "loss": 0.5003,
+      "step": 5160
+    },
+    {
+      "epoch": 2.0222277105458137,
+      "grad_norm": 0.4497347107376531,
+      "learning_rate": 4.892745588460606e-06,
+      "loss": 0.4777,
+      "step": 5161
+    },
+    {
+      "epoch": 2.022622869844406,
+      "grad_norm": 0.4547275239393893,
+      "learning_rate": 4.892700185830291e-06,
+      "loss": 0.4908,
+      "step": 5162
+    },
+    {
+      "epoch": 2.023018029142998,
+      "grad_norm": 0.4776327997808457,
+      "learning_rate": 4.892654773802904e-06,
+      "loss": 0.5075,
+      "step": 5163
+    },
+    {
+      "epoch": 2.0234131884415905,
+      "grad_norm": 0.45100034823235746,
+      "learning_rate": 4.892609352378621e-06,
+      "loss": 0.4954,
+      "step": 5164
+    },
+    {
+      "epoch": 2.0238083477401827,
+      "grad_norm": 0.453949322209561,
+      "learning_rate": 4.8925639215576215e-06,
+      "loss": 0.4988,
+      "step": 5165
+    },
+    {
+      "epoch": 2.024203507038775,
+      "grad_norm": 0.4485757581479012,
+      "learning_rate": 4.8925184813400835e-06,
+      "loss": 0.4954,
+      "step": 5166
+    },
+    {
+      "epoch": 2.024598666337367,
+      "grad_norm": 0.4523702947118703,
+      "learning_rate": 4.892473031726187e-06,
+      "loss": 0.5083,
+      "step": 5167
+    },
+    {
+      "epoch": 2.0249938256359594,
+      "grad_norm": 0.4510797356282732,
+      "learning_rate": 4.8924275727161075e-06,
+      "loss": 0.4964,
+      "step": 5168
+    },
+    {
+      "epoch": 2.0253889849345517,
+      "grad_norm": 0.464165953716701,
+      "learning_rate": 4.892382104310026e-06,
+      "loss": 0.506,
+      "step": 5169
+    },
+    {
+      "epoch": 2.025784144233144,
+      "grad_norm": 0.4598347669774403,
+      "learning_rate": 4.892336626508121e-06,
+      "loss": 0.5139,
+      "step": 5170
+    },
+    {
+      "epoch": 2.026179303531736,
+      "grad_norm": 0.44063169455818113,
+      "learning_rate": 4.89229113931057e-06,
+      "loss": 0.5005,
+      "step": 5171
+    },
+    {
+      "epoch": 2.0265744628303284,
+      "grad_norm": 0.4520710240225769,
+      "learning_rate": 4.892245642717551e-06,
+      "loss": 0.4917,
+      "step": 5172
+    },
+    {
+      "epoch": 2.0269696221289206,
+      "grad_norm": 0.4558674076863556,
+      "learning_rate": 4.8922001367292445e-06,
+      "loss": 0.487,
+      "step": 5173
+    },
+    {
+      "epoch": 2.027364781427513,
+      "grad_norm": 0.4637585590513638,
+      "learning_rate": 4.892154621345829e-06,
+      "loss": 0.5094,
+      "step": 5174
+    },
+    {
+      "epoch": 2.027759940726105,
+      "grad_norm": 0.45275964455192436,
+      "learning_rate": 4.8921090965674825e-06,
+      "loss": 0.4866,
+      "step": 5175
+    },
+    {
+      "epoch": 2.0281551000246973,
+      "grad_norm": 0.4544576816718463,
+      "learning_rate": 4.892063562394384e-06,
+      "loss": 0.5135,
+      "step": 5176
+    },
+    {
+      "epoch": 2.0285502593232896,
+      "grad_norm": 0.4713873273127142,
+      "learning_rate": 4.892018018826712e-06,
+      "loss": 0.4952,
+      "step": 5177
+    },
+    {
+      "epoch": 2.028945418621882,
+      "grad_norm": 0.44993919688256234,
+      "learning_rate": 4.8919724658646465e-06,
+      "loss": 0.4896,
+      "step": 5178
+    },
+    {
+      "epoch": 2.029340577920474,
+      "grad_norm": 0.4689922683938695,
+      "learning_rate": 4.891926903508365e-06,
+      "loss": 0.4828,
+      "step": 5179
+    },
+    {
+      "epoch": 2.0297357372190663,
+      "grad_norm": 0.4838250197458008,
+      "learning_rate": 4.891881331758047e-06,
+      "loss": 0.5096,
+      "step": 5180
+    },
+    {
+      "epoch": 2.0301308965176585,
+      "grad_norm": 0.4675284522861168,
+      "learning_rate": 4.891835750613872e-06,
+      "loss": 0.4982,
+      "step": 5181
+    },
+    {
+      "epoch": 2.030526055816251,
+      "grad_norm": 0.45385770764223,
+      "learning_rate": 4.891790160076018e-06,
+      "loss": 0.5001,
+      "step": 5182
+    },
+    {
+      "epoch": 2.030921215114843,
+      "grad_norm": 0.4600669620147579,
+      "learning_rate": 4.8917445601446656e-06,
+      "loss": 0.4839,
+      "step": 5183
+    },
+    {
+      "epoch": 2.0313163744134353,
+      "grad_norm": 0.4530041700317838,
+      "learning_rate": 4.891698950819992e-06,
+      "loss": 0.501,
+      "step": 5184
+    },
+    {
+      "epoch": 2.0317115337120275,
+      "grad_norm": 0.47142109619231753,
+      "learning_rate": 4.891653332102177e-06,
+      "loss": 0.5274,
+      "step": 5185
+    },
+    {
+      "epoch": 2.0321066930106197,
+      "grad_norm": 0.46742031828739516,
+      "learning_rate": 4.891607703991401e-06,
+      "loss": 0.5046,
+      "step": 5186
+    },
+    {
+      "epoch": 2.032501852309212,
+      "grad_norm": 0.7198025805905561,
+      "learning_rate": 4.891562066487842e-06,
+      "loss": 0.4781,
+      "step": 5187
+    },
+    {
+      "epoch": 2.0328970116078042,
+      "grad_norm": 0.47277610272117676,
+      "learning_rate": 4.891516419591679e-06,
+      "loss": 0.4994,
+      "step": 5188
+    },
+    {
+      "epoch": 2.0332921709063965,
+      "grad_norm": 0.49484180211119877,
+      "learning_rate": 4.891470763303092e-06,
+      "loss": 0.4852,
+      "step": 5189
+    },
+    {
+      "epoch": 2.0336873302049887,
+      "grad_norm": 0.4859621511332183,
+      "learning_rate": 4.89142509762226e-06,
+      "loss": 0.5134,
+      "step": 5190
+    },
+    {
+      "epoch": 2.034082489503581,
+      "grad_norm": 0.4577229468236649,
+      "learning_rate": 4.891379422549361e-06,
+      "loss": 0.4986,
+      "step": 5191
+    },
+    {
+      "epoch": 2.034477648802173,
+      "grad_norm": 0.47417760848062124,
+      "learning_rate": 4.891333738084578e-06,
+      "loss": 0.511,
+      "step": 5192
+    },
+    {
+      "epoch": 2.0348728081007654,
+      "grad_norm": 0.48539047796098145,
+      "learning_rate": 4.891288044228088e-06,
+      "loss": 0.496,
+      "step": 5193
+    },
+    {
+      "epoch": 2.0352679673993577,
+      "grad_norm": 0.46849711925970705,
+      "learning_rate": 4.891242340980069e-06,
+      "loss": 0.5081,
+      "step": 5194
+    },
+    {
+      "epoch": 2.03566312669795,
+      "grad_norm": 0.4690072112312264,
+      "learning_rate": 4.891196628340703e-06,
+      "loss": 0.5097,
+      "step": 5195
+    },
+    {
+      "epoch": 2.036058285996542,
+      "grad_norm": 0.46808264537794086,
+      "learning_rate": 4.8911509063101685e-06,
+      "loss": 0.488,
+      "step": 5196
+    },
+    {
+      "epoch": 2.0364534452951344,
+      "grad_norm": 0.5519283415442977,
+      "learning_rate": 4.891105174888645e-06,
+      "loss": 0.5065,
+      "step": 5197
+    },
+    {
+      "epoch": 2.0368486045937266,
+      "grad_norm": 0.46362954842671233,
+      "learning_rate": 4.891059434076313e-06,
+      "loss": 0.4878,
+      "step": 5198
+    },
+    {
+      "epoch": 2.0372437638923193,
+      "grad_norm": 0.5173109880832085,
+      "learning_rate": 4.891013683873351e-06,
+      "loss": 0.4982,
+      "step": 5199
+    },
+    {
+      "epoch": 2.0376389231909116,
+      "grad_norm": 0.46175244060924875,
+      "learning_rate": 4.890967924279939e-06,
+      "loss": 0.4923,
+      "step": 5200
+    },
+    {
+      "epoch": 2.038034082489504,
+      "grad_norm": 0.44259574184550005,
+      "learning_rate": 4.8909221552962574e-06,
+      "loss": 0.4824,
+      "step": 5201
+    },
+    {
+      "epoch": 2.038429241788096,
+      "grad_norm": 0.4586966341918503,
+      "learning_rate": 4.890876376922486e-06,
+      "loss": 0.4779,
+      "step": 5202
+    },
+    {
+      "epoch": 2.0388244010866883,
+      "grad_norm": 0.46102142747514496,
+      "learning_rate": 4.890830589158802e-06,
+      "loss": 0.49,
+      "step": 5203
+    },
+    {
+      "epoch": 2.0392195603852805,
+      "grad_norm": 0.471145434062063,
+      "learning_rate": 4.8907847920053885e-06,
+      "loss": 0.5156,
+      "step": 5204
+    },
+    {
+      "epoch": 2.0396147196838728,
+      "grad_norm": 0.4549769045467522,
+      "learning_rate": 4.890738985462424e-06,
+      "loss": 0.4871,
+      "step": 5205
+    },
+    {
+      "epoch": 2.040009878982465,
+      "grad_norm": 0.44669007469835803,
+      "learning_rate": 4.890693169530088e-06,
+      "loss": 0.48,
+      "step": 5206
+    },
+    {
+      "epoch": 2.0404050382810572,
+      "grad_norm": 0.4617411167739797,
+      "learning_rate": 4.890647344208562e-06,
+      "loss": 0.4818,
+      "step": 5207
+    },
+    {
+      "epoch": 2.0408001975796495,
+      "grad_norm": 0.4689178727542056,
+      "learning_rate": 4.8906015094980246e-06,
+      "loss": 0.4983,
+      "step": 5208
+    },
+    {
+      "epoch": 2.0411953568782417,
+      "grad_norm": 0.45516172538535027,
+      "learning_rate": 4.890555665398656e-06,
+      "loss": 0.4964,
+      "step": 5209
+    },
+    {
+      "epoch": 2.041590516176834,
+      "grad_norm": 0.4391222692490777,
+      "learning_rate": 4.890509811910637e-06,
+      "loss": 0.472,
+      "step": 5210
+    },
+    {
+      "epoch": 2.041985675475426,
+      "grad_norm": 0.4739508445179121,
+      "learning_rate": 4.890463949034145e-06,
+      "loss": 0.5124,
+      "step": 5211
+    },
+    {
+      "epoch": 2.0423808347740184,
+      "grad_norm": 0.4597047754530406,
+      "learning_rate": 4.890418076769364e-06,
+      "loss": 0.4901,
+      "step": 5212
+    },
+    {
+      "epoch": 2.0427759940726107,
+      "grad_norm": 0.4400780410171843,
+      "learning_rate": 4.890372195116472e-06,
+      "loss": 0.4955,
+      "step": 5213
+    },
+    {
+      "epoch": 2.043171153371203,
+      "grad_norm": 0.4525419825437649,
+      "learning_rate": 4.890326304075649e-06,
+      "loss": 0.4837,
+      "step": 5214
+    },
+    {
+      "epoch": 2.043566312669795,
+      "grad_norm": 0.45871098059276433,
+      "learning_rate": 4.890280403647076e-06,
+      "loss": 0.4951,
+      "step": 5215
+    },
+    {
+      "epoch": 2.0439614719683874,
+      "grad_norm": 0.4643059677550961,
+      "learning_rate": 4.890234493830933e-06,
+      "loss": 0.503,
+      "step": 5216
+    },
+    {
+      "epoch": 2.0443566312669796,
+      "grad_norm": 0.4499775130020552,
+      "learning_rate": 4.8901885746274e-06,
+      "loss": 0.484,
+      "step": 5217
+    },
+    {
+      "epoch": 2.044751790565572,
+      "grad_norm": 0.45820087196407705,
+      "learning_rate": 4.890142646036659e-06,
+      "loss": 0.504,
+      "step": 5218
+    },
+    {
+      "epoch": 2.045146949864164,
+      "grad_norm": 0.4532918197234839,
+      "learning_rate": 4.890096708058888e-06,
+      "loss": 0.4833,
+      "step": 5219
+    },
+    {
+      "epoch": 2.0455421091627564,
+      "grad_norm": 0.4596033870554187,
+      "learning_rate": 4.890050760694268e-06,
+      "loss": 0.4818,
+      "step": 5220
+    },
+    {
+      "epoch": 2.0459372684613486,
+      "grad_norm": 0.5275495863257932,
+      "learning_rate": 4.890004803942982e-06,
+      "loss": 0.5084,
+      "step": 5221
+    },
+    {
+      "epoch": 2.046332427759941,
+      "grad_norm": 0.44983675009756247,
+      "learning_rate": 4.889958837805207e-06,
+      "loss": 0.4839,
+      "step": 5222
+    },
+    {
+      "epoch": 2.046727587058533,
+      "grad_norm": 0.46932242931178536,
+      "learning_rate": 4.889912862281124e-06,
+      "loss": 0.4891,
+      "step": 5223
+    },
+    {
+      "epoch": 2.0471227463571253,
+      "grad_norm": 0.4580616059487756,
+      "learning_rate": 4.889866877370915e-06,
+      "loss": 0.4697,
+      "step": 5224
+    },
+    {
+      "epoch": 2.0475179056557176,
+      "grad_norm": 0.4543027878441082,
+      "learning_rate": 4.8898208830747615e-06,
+      "loss": 0.4971,
+      "step": 5225
+    },
+    {
+      "epoch": 2.04791306495431,
+      "grad_norm": 0.45234992583478045,
+      "learning_rate": 4.889774879392841e-06,
+      "loss": 0.5034,
+      "step": 5226
+    },
+    {
+      "epoch": 2.048308224252902,
+      "grad_norm": 0.45898065072529526,
+      "learning_rate": 4.889728866325337e-06,
+      "loss": 0.5033,
+      "step": 5227
+    },
+    {
+      "epoch": 2.0487033835514943,
+      "grad_norm": 0.4601022979706201,
+      "learning_rate": 4.889682843872429e-06,
+      "loss": 0.4865,
+      "step": 5228
+    },
+    {
+      "epoch": 2.0490985428500865,
+      "grad_norm": 0.44681491251197947,
+      "learning_rate": 4.889636812034298e-06,
+      "loss": 0.5066,
+      "step": 5229
+    },
+    {
+      "epoch": 2.0494937021486788,
+      "grad_norm": 0.4506415649381756,
+      "learning_rate": 4.889590770811125e-06,
+      "loss": 0.4907,
+      "step": 5230
+    },
+    {
+      "epoch": 2.049888861447271,
+      "grad_norm": 0.45120214151761606,
+      "learning_rate": 4.88954472020309e-06,
+      "loss": 0.5062,
+      "step": 5231
+    },
+    {
+      "epoch": 2.0502840207458632,
+      "grad_norm": 0.45154401506062164,
+      "learning_rate": 4.8894986602103735e-06,
+      "loss": 0.4961,
+      "step": 5232
+    },
+    {
+      "epoch": 2.0506791800444555,
+      "grad_norm": 0.4511811539872512,
+      "learning_rate": 4.889452590833158e-06,
+      "loss": 0.4858,
+      "step": 5233
+    },
+    {
+      "epoch": 2.0510743393430477,
+      "grad_norm": 0.44882241957643904,
+      "learning_rate": 4.8894065120716235e-06,
+      "loss": 0.5002,
+      "step": 5234
+    },
+    {
+      "epoch": 2.05146949864164,
+      "grad_norm": 0.44681664863949433,
+      "learning_rate": 4.889360423925952e-06,
+      "loss": 0.4889,
+      "step": 5235
+    },
+    {
+      "epoch": 2.051864657940232,
+      "grad_norm": 0.45495838986619974,
+      "learning_rate": 4.889314326396323e-06,
+      "loss": 0.4913,
+      "step": 5236
+    },
+    {
+      "epoch": 2.0522598172388244,
+      "grad_norm": 0.45058039719748044,
+      "learning_rate": 4.889268219482918e-06,
+      "loss": 0.4974,
+      "step": 5237
+    },
+    {
+      "epoch": 2.0526549765374167,
+      "grad_norm": 0.4673937972973294,
+      "learning_rate": 4.889222103185919e-06,
+      "loss": 0.4963,
+      "step": 5238
+    },
+    {
+      "epoch": 2.053050135836009,
+      "grad_norm": 0.4650737282011602,
+      "learning_rate": 4.889175977505505e-06,
+      "loss": 0.4903,
+      "step": 5239
+    },
+    {
+      "epoch": 2.053445295134601,
+      "grad_norm": 0.45116470306677475,
+      "learning_rate": 4.88912984244186e-06,
+      "loss": 0.4992,
+      "step": 5240
+    },
+    {
+      "epoch": 2.0538404544331934,
+      "grad_norm": 0.5298176727549352,
+      "learning_rate": 4.889083697995163e-06,
+      "loss": 0.4919,
+      "step": 5241
+    },
+    {
+      "epoch": 2.0542356137317856,
+      "grad_norm": 0.45177097015636897,
+      "learning_rate": 4.889037544165596e-06,
+      "loss": 0.4904,
+      "step": 5242
+    },
+    {
+      "epoch": 2.054630773030378,
+      "grad_norm": 0.5119412810802527,
+      "learning_rate": 4.8889913809533404e-06,
+      "loss": 0.4968,
+      "step": 5243
+    },
+    {
+      "epoch": 2.05502593232897,
+      "grad_norm": 0.4616855361620579,
+      "learning_rate": 4.888945208358577e-06,
+      "loss": 0.5005,
+      "step": 5244
+    },
+    {
+      "epoch": 2.0554210916275624,
+      "grad_norm": 0.4672209154136625,
+      "learning_rate": 4.888899026381487e-06,
+      "loss": 0.5107,
+      "step": 5245
+    },
+    {
+      "epoch": 2.0558162509261546,
+      "grad_norm": 0.4528232853036076,
+      "learning_rate": 4.888852835022253e-06,
+      "loss": 0.4762,
+      "step": 5246
+    },
+    {
+      "epoch": 2.056211410224747,
+      "grad_norm": 0.4558490462001704,
+      "learning_rate": 4.8888066342810555e-06,
+      "loss": 0.491,
+      "step": 5247
+    },
+    {
+      "epoch": 2.056606569523339,
+      "grad_norm": 0.4527416296938898,
+      "learning_rate": 4.888760424158077e-06,
+      "loss": 0.4922,
+      "step": 5248
+    },
+    {
+      "epoch": 2.0570017288219313,
+      "grad_norm": 0.46237981652455606,
+      "learning_rate": 4.8887142046534975e-06,
+      "loss": 0.5024,
+      "step": 5249
+    },
+    {
+      "epoch": 2.0573968881205236,
+      "grad_norm": 0.4637558739077874,
+      "learning_rate": 4.888667975767499e-06,
+      "loss": 0.4894,
+      "step": 5250
+    },
+    {
+      "epoch": 2.057792047419116,
+      "grad_norm": 0.4744951570899205,
+      "learning_rate": 4.888621737500262e-06,
+      "loss": 0.4908,
+      "step": 5251
+    },
+    {
+      "epoch": 2.058187206717708,
+      "grad_norm": 0.4584214020537685,
+      "learning_rate": 4.888575489851971e-06,
+      "loss": 0.5037,
+      "step": 5252
+    },
+    {
+      "epoch": 2.0585823660163003,
+      "grad_norm": 0.4727044592007818,
+      "learning_rate": 4.888529232822805e-06,
+      "loss": 0.4776,
+      "step": 5253
+    },
+    {
+      "epoch": 2.0589775253148925,
+      "grad_norm": 0.4632461314737739,
+      "learning_rate": 4.888482966412947e-06,
+      "loss": 0.4921,
+      "step": 5254
+    },
+    {
+      "epoch": 2.0593726846134848,
+      "grad_norm": 0.46365343578610424,
+      "learning_rate": 4.888436690622578e-06,
+      "loss": 0.4982,
+      "step": 5255
+    },
+    {
+      "epoch": 2.059767843912077,
+      "grad_norm": 0.4927502347234346,
+      "learning_rate": 4.8883904054518805e-06,
+      "loss": 0.5025,
+      "step": 5256
+    },
+    {
+      "epoch": 2.0601630032106693,
+      "grad_norm": 0.45431480591589724,
+      "learning_rate": 4.888344110901035e-06,
+      "loss": 0.4862,
+      "step": 5257
+    },
+    {
+      "epoch": 2.0605581625092615,
+      "grad_norm": 0.44639395998441367,
+      "learning_rate": 4.888297806970225e-06,
+      "loss": 0.5002,
+      "step": 5258
+    },
+    {
+      "epoch": 2.0609533218078537,
+      "grad_norm": 0.4581725919289095,
+      "learning_rate": 4.888251493659631e-06,
+      "loss": 0.4763,
+      "step": 5259
+    },
+    {
+      "epoch": 2.061348481106446,
+      "grad_norm": 0.455160272080671,
+      "learning_rate": 4.888205170969435e-06,
+      "loss": 0.499,
+      "step": 5260
+    },
+    {
+      "epoch": 2.061743640405038,
+      "grad_norm": 0.4643189428275059,
+      "learning_rate": 4.888158838899819e-06,
+      "loss": 0.4932,
+      "step": 5261
+    },
+    {
+      "epoch": 2.0621387997036305,
+      "grad_norm": 0.4463638667832112,
+      "learning_rate": 4.888112497450966e-06,
+      "loss": 0.5011,
+      "step": 5262
+    },
+    {
+      "epoch": 2.0625339590022227,
+      "grad_norm": 0.44491650271349903,
+      "learning_rate": 4.888066146623058e-06,
+      "loss": 0.4968,
+      "step": 5263
+    },
+    {
+      "epoch": 2.062929118300815,
+      "grad_norm": 0.4741916119379606,
+      "learning_rate": 4.888019786416275e-06,
+      "loss": 0.5109,
+      "step": 5264
+    },
+    {
+      "epoch": 2.063324277599407,
+      "grad_norm": 0.46077105210136354,
+      "learning_rate": 4.887973416830801e-06,
+      "loss": 0.4847,
+      "step": 5265
+    },
+    {
+      "epoch": 2.0637194368979994,
+      "grad_norm": 0.5849360991852471,
+      "learning_rate": 4.887927037866817e-06,
+      "loss": 0.4926,
+      "step": 5266
+    },
+    {
+      "epoch": 2.0641145961965917,
+      "grad_norm": 0.4633200832882633,
+      "learning_rate": 4.8878806495245055e-06,
+      "loss": 0.4919,
+      "step": 5267
+    },
+    {
+      "epoch": 2.064509755495184,
+      "grad_norm": 0.46236301178780204,
+      "learning_rate": 4.887834251804049e-06,
+      "loss": 0.4954,
+      "step": 5268
+    },
+    {
+      "epoch": 2.064904914793776,
+      "grad_norm": 0.457629583658407,
+      "learning_rate": 4.8877878447056305e-06,
+      "loss": 0.4923,
+      "step": 5269
+    },
+    {
+      "epoch": 2.0653000740923684,
+      "grad_norm": 0.4673120651485547,
+      "learning_rate": 4.88774142822943e-06,
+      "loss": 0.4808,
+      "step": 5270
+    },
+    {
+      "epoch": 2.0656952333909606,
+      "grad_norm": 0.4568906136006192,
+      "learning_rate": 4.887695002375631e-06,
+      "loss": 0.485,
+      "step": 5271
+    },
+    {
+      "epoch": 2.066090392689553,
+      "grad_norm": 0.4734779549670619,
+      "learning_rate": 4.8876485671444175e-06,
+      "loss": 0.4925,
+      "step": 5272
+    },
+    {
+      "epoch": 2.066485551988145,
+      "grad_norm": 0.47832808820634337,
+      "learning_rate": 4.887602122535969e-06,
+      "loss": 0.4957,
+      "step": 5273
+    },
+    {
+      "epoch": 2.0668807112867373,
+      "grad_norm": 0.4749938197663773,
+      "learning_rate": 4.887555668550469e-06,
+      "loss": 0.5011,
+      "step": 5274
+    },
+    {
+      "epoch": 2.0672758705853296,
+      "grad_norm": 0.46181910473113197,
+      "learning_rate": 4.887509205188101e-06,
+      "loss": 0.4954,
+      "step": 5275
+    },
+    {
+      "epoch": 2.067671029883922,
+      "grad_norm": 0.4476900675547803,
+      "learning_rate": 4.887462732449046e-06,
+      "loss": 0.5033,
+      "step": 5276
+    },
+    {
+      "epoch": 2.068066189182514,
+      "grad_norm": 0.4540259110444361,
+      "learning_rate": 4.887416250333487e-06,
+      "loss": 0.4902,
+      "step": 5277
+    },
+    {
+      "epoch": 2.0684613484811063,
+      "grad_norm": 0.448149981706839,
+      "learning_rate": 4.8873697588416075e-06,
+      "loss": 0.4998,
+      "step": 5278
+    },
+    {
+      "epoch": 2.0688565077796985,
+      "grad_norm": 0.44190241176056827,
+      "learning_rate": 4.887323257973589e-06,
+      "loss": 0.4946,
+      "step": 5279
+    },
+    {
+      "epoch": 2.069251667078291,
+      "grad_norm": 0.4462613061115609,
+      "learning_rate": 4.887276747729614e-06,
+      "loss": 0.4817,
+      "step": 5280
+    },
+    {
+      "epoch": 2.069646826376883,
+      "grad_norm": 0.45013660019262763,
+      "learning_rate": 4.887230228109866e-06,
+      "loss": 0.4897,
+      "step": 5281
+    },
+    {
+      "epoch": 2.0700419856754753,
+      "grad_norm": 0.4482753274443419,
+      "learning_rate": 4.887183699114526e-06,
+      "loss": 0.5006,
+      "step": 5282
+    },
+    {
+      "epoch": 2.0704371449740675,
+      "grad_norm": 0.46170470070203723,
+      "learning_rate": 4.88713716074378e-06,
+      "loss": 0.5052,
+      "step": 5283
+    },
+    {
+      "epoch": 2.0708323042726597,
+      "grad_norm": 0.4669933900589111,
+      "learning_rate": 4.887090612997808e-06,
+      "loss": 0.4917,
+      "step": 5284
+    },
+    {
+      "epoch": 2.071227463571252,
+      "grad_norm": 0.483843226668854,
+      "learning_rate": 4.887044055876793e-06,
+      "loss": 0.5046,
+      "step": 5285
+    },
+    {
+      "epoch": 2.071622622869844,
+      "grad_norm": 0.46714368758553043,
+      "learning_rate": 4.886997489380919e-06,
+      "loss": 0.489,
+      "step": 5286
+    },
+    {
+      "epoch": 2.0720177821684365,
+      "grad_norm": 0.44469009126903564,
+      "learning_rate": 4.886950913510368e-06,
+      "loss": 0.4925,
+      "step": 5287
+    },
+    {
+      "epoch": 2.0724129414670287,
+      "grad_norm": 0.46045303664787235,
+      "learning_rate": 4.886904328265323e-06,
+      "loss": 0.4839,
+      "step": 5288
+    },
+    {
+      "epoch": 2.072808100765621,
+      "grad_norm": 0.4601616045035097,
+      "learning_rate": 4.886857733645968e-06,
+      "loss": 0.4876,
+      "step": 5289
+    },
+    {
+      "epoch": 2.073203260064213,
+      "grad_norm": 0.48030791644212956,
+      "learning_rate": 4.886811129652484e-06,
+      "loss": 0.5098,
+      "step": 5290
+    },
+    {
+      "epoch": 2.0735984193628054,
+      "grad_norm": 0.4499859205882746,
+      "learning_rate": 4.886764516285057e-06,
+      "loss": 0.4892,
+      "step": 5291
+    },
+    {
+      "epoch": 2.0739935786613977,
+      "grad_norm": 0.46426479744305693,
+      "learning_rate": 4.886717893543868e-06,
+      "loss": 0.4841,
+      "step": 5292
+    },
+    {
+      "epoch": 2.07438873795999,
+      "grad_norm": 0.4559454042901413,
+      "learning_rate": 4.886671261429099e-06,
+      "loss": 0.488,
+      "step": 5293
+    },
+    {
+      "epoch": 2.074783897258582,
+      "grad_norm": 0.4563383108320409,
+      "learning_rate": 4.8866246199409354e-06,
+      "loss": 0.4792,
+      "step": 5294
+    },
+    {
+      "epoch": 2.0751790565571744,
+      "grad_norm": 0.4641041054840762,
+      "learning_rate": 4.886577969079559e-06,
+      "loss": 0.4883,
+      "step": 5295
+    },
+    {
+      "epoch": 2.075574215855767,
+      "grad_norm": 0.47969805202822735,
+      "learning_rate": 4.8865313088451544e-06,
+      "loss": 0.5032,
+      "step": 5296
+    },
+    {
+      "epoch": 2.0759693751543593,
+      "grad_norm": 0.46863300818021786,
+      "learning_rate": 4.886484639237903e-06,
+      "loss": 0.498,
+      "step": 5297
+    },
+    {
+      "epoch": 2.0763645344529515,
+      "grad_norm": 0.4657095895273564,
+      "learning_rate": 4.88643796025799e-06,
+      "loss": 0.4925,
+      "step": 5298
+    },
+    {
+      "epoch": 2.076759693751544,
+      "grad_norm": 0.6267052883363237,
+      "learning_rate": 4.886391271905597e-06,
+      "loss": 0.5167,
+      "step": 5299
+    },
+    {
+      "epoch": 2.077154853050136,
+      "grad_norm": 0.4773221517600891,
+      "learning_rate": 4.886344574180909e-06,
+      "loss": 0.4994,
+      "step": 5300
+    },
+    {
+      "epoch": 2.0775500123487283,
+      "grad_norm": 0.48647171806501055,
+      "learning_rate": 4.886297867084109e-06,
+      "loss": 0.4895,
+      "step": 5301
+    },
+    {
+      "epoch": 2.0779451716473205,
+      "grad_norm": 0.47734592420468774,
+      "learning_rate": 4.886251150615379e-06,
+      "loss": 0.4964,
+      "step": 5302
+    },
+    {
+      "epoch": 2.0783403309459128,
+      "grad_norm": 0.4822474079732969,
+      "learning_rate": 4.886204424774904e-06,
+      "loss": 0.5145,
+      "step": 5303
+    },
+    {
+      "epoch": 2.078735490244505,
+      "grad_norm": 0.4648912162923145,
+      "learning_rate": 4.886157689562866e-06,
+      "loss": 0.5063,
+      "step": 5304
+    },
+    {
+      "epoch": 2.0791306495430972,
+      "grad_norm": 0.45932677809466166,
+      "learning_rate": 4.886110944979451e-06,
+      "loss": 0.495,
+      "step": 5305
+    },
+    {
+      "epoch": 2.0795258088416895,
+      "grad_norm": 0.47706737183878567,
+      "learning_rate": 4.88606419102484e-06,
+      "loss": 0.5103,
+      "step": 5306
+    },
+    {
+      "epoch": 2.0799209681402817,
+      "grad_norm": 0.4532623596914375,
+      "learning_rate": 4.886017427699218e-06,
+      "loss": 0.4873,
+      "step": 5307
+    },
+    {
+      "epoch": 2.080316127438874,
+      "grad_norm": 0.4483128492884054,
+      "learning_rate": 4.885970655002768e-06,
+      "loss": 0.5029,
+      "step": 5308
+    },
+    {
+      "epoch": 2.080711286737466,
+      "grad_norm": 0.46648648387318337,
+      "learning_rate": 4.885923872935675e-06,
+      "loss": 0.4932,
+      "step": 5309
+    },
+    {
+      "epoch": 2.0811064460360584,
+      "grad_norm": 0.470504146656533,
+      "learning_rate": 4.885877081498122e-06,
+      "loss": 0.4986,
+      "step": 5310
+    },
+    {
+      "epoch": 2.0815016053346507,
+      "grad_norm": 0.46815656364735725,
+      "learning_rate": 4.8858302806902925e-06,
+      "loss": 0.5053,
+      "step": 5311
+    },
+    {
+      "epoch": 2.081896764633243,
+      "grad_norm": 0.47676947419158733,
+      "learning_rate": 4.88578347051237e-06,
+      "loss": 0.5048,
+      "step": 5312
+    },
+    {
+      "epoch": 2.082291923931835,
+      "grad_norm": 0.4551948922256932,
+      "learning_rate": 4.885736650964539e-06,
+      "loss": 0.4846,
+      "step": 5313
+    },
+    {
+      "epoch": 2.0826870832304274,
+      "grad_norm": 0.4689297776862242,
+      "learning_rate": 4.885689822046983e-06,
+      "loss": 0.5138,
+      "step": 5314
+    },
+    {
+      "epoch": 2.0830822425290196,
+      "grad_norm": 0.5266371128832054,
+      "learning_rate": 4.885642983759885e-06,
+      "loss": 0.502,
+      "step": 5315
+    },
+    {
+      "epoch": 2.083477401827612,
+      "grad_norm": 0.45232336122965455,
+      "learning_rate": 4.885596136103432e-06,
+      "loss": 0.4926,
+      "step": 5316
+    },
+    {
+      "epoch": 2.083872561126204,
+      "grad_norm": 0.45815345336313995,
+      "learning_rate": 4.885549279077805e-06,
+      "loss": 0.4706,
+      "step": 5317
+    },
+    {
+      "epoch": 2.0842677204247964,
+      "grad_norm": 0.4508115268547949,
+      "learning_rate": 4.885502412683189e-06,
+      "loss": 0.5158,
+      "step": 5318
+    },
+    {
+      "epoch": 2.0846628797233886,
+      "grad_norm": 0.4506873596094119,
+      "learning_rate": 4.885455536919767e-06,
+      "loss": 0.4826,
+      "step": 5319
+    },
+    {
+      "epoch": 2.085058039021981,
+      "grad_norm": 0.47613431415557844,
+      "learning_rate": 4.885408651787725e-06,
+      "loss": 0.5037,
+      "step": 5320
+    },
+    {
+      "epoch": 2.085453198320573,
+      "grad_norm": 0.4603502263078611,
+      "learning_rate": 4.885361757287247e-06,
+      "loss": 0.4886,
+      "step": 5321
+    },
+    {
+      "epoch": 2.0858483576191653,
+      "grad_norm": 0.46036610887217155,
+      "learning_rate": 4.8853148534185165e-06,
+      "loss": 0.4953,
+      "step": 5322
+    },
+    {
+      "epoch": 2.0862435169177576,
+      "grad_norm": 0.4529241365840035,
+      "learning_rate": 4.885267940181717e-06,
+      "loss": 0.4953,
+      "step": 5323
+    },
+    {
+      "epoch": 2.08663867621635,
+      "grad_norm": 0.48675482273393644,
+      "learning_rate": 4.885221017577033e-06,
+      "loss": 0.5134,
+      "step": 5324
+    },
+    {
+      "epoch": 2.087033835514942,
+      "grad_norm": 0.45569874226039914,
+      "learning_rate": 4.88517408560465e-06,
+      "loss": 0.5097,
+      "step": 5325
+    },
+    {
+      "epoch": 2.0874289948135343,
+      "grad_norm": 0.4673253599053655,
+      "learning_rate": 4.885127144264752e-06,
+      "loss": 0.5054,
+      "step": 5326
+    },
+    {
+      "epoch": 2.0878241541121265,
+      "grad_norm": 0.4531561763675708,
+      "learning_rate": 4.885080193557522e-06,
+      "loss": 0.5016,
+      "step": 5327
+    },
+    {
+      "epoch": 2.0882193134107188,
+      "grad_norm": 0.4872090735746867,
+      "learning_rate": 4.885033233483146e-06,
+      "loss": 0.5176,
+      "step": 5328
+    },
+    {
+      "epoch": 2.088614472709311,
+      "grad_norm": 0.4513837913446904,
+      "learning_rate": 4.884986264041808e-06,
+      "loss": 0.469,
+      "step": 5329
+    },
+    {
+      "epoch": 2.0890096320079032,
+      "grad_norm": 0.4590797104168329,
+      "learning_rate": 4.884939285233691e-06,
+      "loss": 0.504,
+      "step": 5330
+    },
+    {
+      "epoch": 2.0894047913064955,
+      "grad_norm": 0.4627972424027732,
+      "learning_rate": 4.884892297058981e-06,
+      "loss": 0.4901,
+      "step": 5331
+    },
+    {
+      "epoch": 2.0897999506050877,
+      "grad_norm": 0.44688340629975803,
+      "learning_rate": 4.884845299517863e-06,
+      "loss": 0.4877,
+      "step": 5332
+    },
+    {
+      "epoch": 2.09019510990368,
+      "grad_norm": 0.4611503172302672,
+      "learning_rate": 4.88479829261052e-06,
+      "loss": 0.4921,
+      "step": 5333
+    },
+    {
+      "epoch": 2.090590269202272,
+      "grad_norm": 0.46829455437768785,
+      "learning_rate": 4.884751276337138e-06,
+      "loss": 0.5113,
+      "step": 5334
+    },
+    {
+      "epoch": 2.0909854285008644,
+      "grad_norm": 0.4473009959944776,
+      "learning_rate": 4.8847042506979e-06,
+      "loss": 0.4904,
+      "step": 5335
+    },
+    {
+      "epoch": 2.0913805877994567,
+      "grad_norm": 0.4604208483059509,
+      "learning_rate": 4.8846572156929936e-06,
+      "loss": 0.5003,
+      "step": 5336
+    },
+    {
+      "epoch": 2.091775747098049,
+      "grad_norm": 0.45767779762723915,
+      "learning_rate": 4.8846101713226005e-06,
+      "loss": 0.4935,
+      "step": 5337
+    },
+    {
+      "epoch": 2.092170906396641,
+      "grad_norm": 0.4615944313852972,
+      "learning_rate": 4.884563117586907e-06,
+      "loss": 0.4979,
+      "step": 5338
+    },
+    {
+      "epoch": 2.0925660656952334,
+      "grad_norm": 0.4810789677245922,
+      "learning_rate": 4.884516054486097e-06,
+      "loss": 0.5252,
+      "step": 5339
+    },
+    {
+      "epoch": 2.0929612249938256,
+      "grad_norm": 0.4573861349101158,
+      "learning_rate": 4.884468982020357e-06,
+      "loss": 0.5129,
+      "step": 5340
+    },
+    {
+      "epoch": 2.093356384292418,
+      "grad_norm": 0.4528184175451582,
+      "learning_rate": 4.88442190018987e-06,
+      "loss": 0.5138,
+      "step": 5341
+    },
+    {
+      "epoch": 2.09375154359101,
+      "grad_norm": 0.4566897018760855,
+      "learning_rate": 4.884374808994822e-06,
+      "loss": 0.5044,
+      "step": 5342
+    },
+    {
+      "epoch": 2.0941467028896024,
+      "grad_norm": 0.4516180831443502,
+      "learning_rate": 4.884327708435397e-06,
+      "loss": 0.4991,
+      "step": 5343
+    },
+    {
+      "epoch": 2.0945418621881946,
+      "grad_norm": 0.44852554152023116,
+      "learning_rate": 4.884280598511781e-06,
+      "loss": 0.5032,
+      "step": 5344
+    },
+    {
+      "epoch": 2.094937021486787,
+      "grad_norm": 0.45516969248231826,
+      "learning_rate": 4.8842334792241586e-06,
+      "loss": 0.4983,
+      "step": 5345
+    },
+    {
+      "epoch": 2.095332180785379,
+      "grad_norm": 0.46631444223982815,
+      "learning_rate": 4.884186350572715e-06,
+      "loss": 0.502,
+      "step": 5346
+    },
+    {
+      "epoch": 2.0957273400839713,
+      "grad_norm": 0.47398673780754186,
+      "learning_rate": 4.884139212557635e-06,
+      "loss": 0.4953,
+      "step": 5347
+    },
+    {
+      "epoch": 2.0961224993825636,
+      "grad_norm": 0.47579743471894775,
+      "learning_rate": 4.884092065179103e-06,
+      "loss": 0.5056,
+      "step": 5348
+    },
+    {
+      "epoch": 2.096517658681156,
+      "grad_norm": 0.48515241623044925,
+      "learning_rate": 4.884044908437306e-06,
+      "loss": 0.5059,
+      "step": 5349
+    },
+    {
+      "epoch": 2.096912817979748,
+      "grad_norm": 0.4652147279266291,
+      "learning_rate": 4.883997742332429e-06,
+      "loss": 0.4976,
+      "step": 5350
+    },
+    {
+      "epoch": 2.0973079772783403,
+      "grad_norm": 0.4453104642842236,
+      "learning_rate": 4.883950566864656e-06,
+      "loss": 0.5074,
+      "step": 5351
+    },
+    {
+      "epoch": 2.0977031365769325,
+      "grad_norm": 0.45143539446002273,
+      "learning_rate": 4.883903382034172e-06,
+      "loss": 0.4835,
+      "step": 5352
+    },
+    {
+      "epoch": 2.0980982958755248,
+      "grad_norm": 0.4628598192673955,
+      "learning_rate": 4.883856187841164e-06,
+      "loss": 0.5055,
+      "step": 5353
+    },
+    {
+      "epoch": 2.098493455174117,
+      "grad_norm": 0.4564188127449865,
+      "learning_rate": 4.883808984285816e-06,
+      "loss": 0.4958,
+      "step": 5354
+    },
+    {
+      "epoch": 2.0988886144727092,
+      "grad_norm": 0.45860123136973613,
+      "learning_rate": 4.8837617713683146e-06,
+      "loss": 0.4906,
+      "step": 5355
+    },
+    {
+      "epoch": 2.0992837737713015,
+      "grad_norm": 0.4732141389245651,
+      "learning_rate": 4.883714549088844e-06,
+      "loss": 0.493,
+      "step": 5356
+    },
+    {
+      "epoch": 2.0996789330698937,
+      "grad_norm": 0.4611681662392404,
+      "learning_rate": 4.8836673174475894e-06,
+      "loss": 0.5004,
+      "step": 5357
+    },
+    {
+      "epoch": 2.100074092368486,
+      "grad_norm": 0.4779009985374861,
+      "learning_rate": 4.883620076444738e-06,
+      "loss": 0.4939,
+      "step": 5358
+    },
+    {
+      "epoch": 2.100469251667078,
+      "grad_norm": 0.4644243183886433,
+      "learning_rate": 4.883572826080474e-06,
+      "loss": 0.4901,
+      "step": 5359
+    },
+    {
+      "epoch": 2.1008644109656704,
+      "grad_norm": 0.4730815340398359,
+      "learning_rate": 4.883525566354983e-06,
+      "loss": 0.4929,
+      "step": 5360
+    },
+    {
+      "epoch": 2.1012595702642627,
+      "grad_norm": 0.4645212217941286,
+      "learning_rate": 4.883478297268451e-06,
+      "loss": 0.485,
+      "step": 5361
+    },
+    {
+      "epoch": 2.101654729562855,
+      "grad_norm": 0.5883162574204609,
+      "learning_rate": 4.883431018821064e-06,
+      "loss": 0.4936,
+      "step": 5362
+    },
+    {
+      "epoch": 2.102049888861447,
+      "grad_norm": 0.44813227041664244,
+      "learning_rate": 4.883383731013007e-06,
+      "loss": 0.4999,
+      "step": 5363
+    },
+    {
+      "epoch": 2.1024450481600394,
+      "grad_norm": 0.4693538528184338,
+      "learning_rate": 4.883336433844465e-06,
+      "loss": 0.5014,
+      "step": 5364
+    },
+    {
+      "epoch": 2.1028402074586316,
+      "grad_norm": 0.4613356297326133,
+      "learning_rate": 4.883289127315627e-06,
+      "loss": 0.4986,
+      "step": 5365
+    },
+    {
+      "epoch": 2.103235366757224,
+      "grad_norm": 0.47462410453708387,
+      "learning_rate": 4.883241811426675e-06,
+      "loss": 0.4964,
+      "step": 5366
+    },
+    {
+      "epoch": 2.103630526055816,
+      "grad_norm": 0.4585830157369119,
+      "learning_rate": 4.883194486177796e-06,
+      "loss": 0.5305,
+      "step": 5367
+    },
+    {
+      "epoch": 2.1040256853544084,
+      "grad_norm": 0.5374049651711528,
+      "learning_rate": 4.883147151569178e-06,
+      "loss": 0.5013,
+      "step": 5368
+    },
+    {
+      "epoch": 2.1044208446530006,
+      "grad_norm": 0.4505696100137808,
+      "learning_rate": 4.883099807601003e-06,
+      "loss": 0.4905,
+      "step": 5369
+    },
+    {
+      "epoch": 2.104816003951593,
+      "grad_norm": 0.472748675203154,
+      "learning_rate": 4.88305245427346e-06,
+      "loss": 0.5043,
+      "step": 5370
+    },
+    {
+      "epoch": 2.105211163250185,
+      "grad_norm": 0.4703878721715555,
+      "learning_rate": 4.883005091586734e-06,
+      "loss": 0.484,
+      "step": 5371
+    },
+    {
+      "epoch": 2.1056063225487773,
+      "grad_norm": 0.4556588761245561,
+      "learning_rate": 4.882957719541011e-06,
+      "loss": 0.5031,
+      "step": 5372
+    },
+    {
+      "epoch": 2.1060014818473696,
+      "grad_norm": 0.44797746752549056,
+      "learning_rate": 4.882910338136478e-06,
+      "loss": 0.4944,
+      "step": 5373
+    },
+    {
+      "epoch": 2.106396641145962,
+      "grad_norm": 0.4886233397662977,
+      "learning_rate": 4.882862947373318e-06,
+      "loss": 0.5049,
+      "step": 5374
+    },
+    {
+      "epoch": 2.106791800444554,
+      "grad_norm": 0.4669275499791975,
+      "learning_rate": 4.882815547251721e-06,
+      "loss": 0.5095,
+      "step": 5375
+    },
+    {
+      "epoch": 2.1071869597431463,
+      "grad_norm": 0.4797695708025504,
+      "learning_rate": 4.8827681377718715e-06,
+      "loss": 0.5128,
+      "step": 5376
+    },
+    {
+      "epoch": 2.1075821190417385,
+      "grad_norm": 0.45320448299389626,
+      "learning_rate": 4.8827207189339545e-06,
+      "loss": 0.4986,
+      "step": 5377
+    },
+    {
+      "epoch": 2.1079772783403308,
+      "grad_norm": 0.44937616960762794,
+      "learning_rate": 4.882673290738158e-06,
+      "loss": 0.499,
+      "step": 5378
+    },
+    {
+      "epoch": 2.108372437638923,
+      "grad_norm": 0.4609581797871518,
+      "learning_rate": 4.8826258531846686e-06,
+      "loss": 0.5022,
+      "step": 5379
+    },
+    {
+      "epoch": 2.1087675969375153,
+      "grad_norm": 0.4582795163905733,
+      "learning_rate": 4.882578406273671e-06,
+      "loss": 0.4887,
+      "step": 5380
+    },
+    {
+      "epoch": 2.1091627562361075,
+      "grad_norm": 0.45000378134977503,
+      "learning_rate": 4.882530950005351e-06,
+      "loss": 0.5033,
+      "step": 5381
+    },
+    {
+      "epoch": 2.1095579155346997,
+      "grad_norm": 0.4535208499466116,
+      "learning_rate": 4.882483484379898e-06,
+      "loss": 0.4893,
+      "step": 5382
+    },
+    {
+      "epoch": 2.109953074833292,
+      "grad_norm": 0.4976229552655962,
+      "learning_rate": 4.8824360093974945e-06,
+      "loss": 0.4984,
+      "step": 5383
+    },
+    {
+      "epoch": 2.110348234131884,
+      "grad_norm": 0.4747654674353325,
+      "learning_rate": 4.88238852505833e-06,
+      "loss": 0.5173,
+      "step": 5384
+    },
+    {
+      "epoch": 2.1107433934304765,
+      "grad_norm": 0.45542034736390435,
+      "learning_rate": 4.88234103136259e-06,
+      "loss": 0.4833,
+      "step": 5385
+    },
+    {
+      "epoch": 2.1111385527290687,
+      "grad_norm": 0.46024000204519655,
+      "learning_rate": 4.882293528310462e-06,
+      "loss": 0.5103,
+      "step": 5386
+    },
+    {
+      "epoch": 2.111533712027661,
+      "grad_norm": 0.4874077584347664,
+      "learning_rate": 4.882246015902131e-06,
+      "loss": 0.5052,
+      "step": 5387
+    },
+    {
+      "epoch": 2.1119288713262536,
+      "grad_norm": 0.44865374080295306,
+      "learning_rate": 4.882198494137785e-06,
+      "loss": 0.4952,
+      "step": 5388
+    },
+    {
+      "epoch": 2.112324030624846,
+      "grad_norm": 0.46109394452242536,
+      "learning_rate": 4.882150963017609e-06,
+      "loss": 0.5005,
+      "step": 5389
+    },
+    {
+      "epoch": 2.112719189923438,
+      "grad_norm": 0.45930292764238206,
+      "learning_rate": 4.88210342254179e-06,
+      "loss": 0.4739,
+      "step": 5390
+    },
+    {
+      "epoch": 2.1131143492220303,
+      "grad_norm": 0.5451627115880612,
+      "learning_rate": 4.882055872710516e-06,
+      "loss": 0.5029,
+      "step": 5391
+    },
+    {
+      "epoch": 2.1135095085206226,
+      "grad_norm": 0.45461534307695234,
+      "learning_rate": 4.882008313523973e-06,
+      "loss": 0.4902,
+      "step": 5392
+    },
+    {
+      "epoch": 2.113904667819215,
+      "grad_norm": 0.4700402964260748,
+      "learning_rate": 4.881960744982348e-06,
+      "loss": 0.5006,
+      "step": 5393
+    },
+    {
+      "epoch": 2.114299827117807,
+      "grad_norm": 0.4558104260192114,
+      "learning_rate": 4.881913167085826e-06,
+      "loss": 0.4926,
+      "step": 5394
+    },
+    {
+      "epoch": 2.1146949864163993,
+      "grad_norm": 0.46103896695677804,
+      "learning_rate": 4.881865579834598e-06,
+      "loss": 0.5009,
+      "step": 5395
+    },
+    {
+      "epoch": 2.1150901457149915,
+      "grad_norm": 0.46008105522327936,
+      "learning_rate": 4.881817983228847e-06,
+      "loss": 0.5072,
+      "step": 5396
+    },
+    {
+      "epoch": 2.115485305013584,
+      "grad_norm": 0.459804457470388,
+      "learning_rate": 4.881770377268761e-06,
+      "loss": 0.4966,
+      "step": 5397
+    },
+    {
+      "epoch": 2.115880464312176,
+      "grad_norm": 0.4564873017201415,
+      "learning_rate": 4.8817227619545274e-06,
+      "loss": 0.5065,
+      "step": 5398
+    },
+    {
+      "epoch": 2.1162756236107683,
+      "grad_norm": 0.4572790123320898,
+      "learning_rate": 4.881675137286334e-06,
+      "loss": 0.4917,
+      "step": 5399
+    },
+    {
+      "epoch": 2.1166707829093605,
+      "grad_norm": 0.4583018544454832,
+      "learning_rate": 4.881627503264365e-06,
+      "loss": 0.5122,
+      "step": 5400
+    },
+    {
+      "epoch": 2.1170659422079527,
+      "grad_norm": 0.45766918038131965,
+      "learning_rate": 4.881579859888811e-06,
+      "loss": 0.5008,
+      "step": 5401
+    },
+    {
+      "epoch": 2.117461101506545,
+      "grad_norm": 0.47089609794076337,
+      "learning_rate": 4.881532207159857e-06,
+      "loss": 0.5111,
+      "step": 5402
+    },
+    {
+      "epoch": 2.1178562608051372,
+      "grad_norm": 0.48698638605698286,
+      "learning_rate": 4.881484545077691e-06,
+      "loss": 0.4955,
+      "step": 5403
+    },
+    {
+      "epoch": 2.1182514201037295,
+      "grad_norm": 0.47367199549589856,
+      "learning_rate": 4.881436873642499e-06,
+      "loss": 0.4843,
+      "step": 5404
+    },
+    {
+      "epoch": 2.1186465794023217,
+      "grad_norm": 0.4548370429625071,
+      "learning_rate": 4.881389192854469e-06,
+      "loss": 0.5109,
+      "step": 5405
+    },
+    {
+      "epoch": 2.119041738700914,
+      "grad_norm": 0.46207813913976026,
+      "learning_rate": 4.881341502713789e-06,
+      "loss": 0.5102,
+      "step": 5406
+    },
+    {
+      "epoch": 2.119436897999506,
+      "grad_norm": 0.4665523026201837,
+      "learning_rate": 4.881293803220646e-06,
+      "loss": 0.4937,
+      "step": 5407
+    },
+    {
+      "epoch": 2.1198320572980984,
+      "grad_norm": 0.44584389144997866,
+      "learning_rate": 4.881246094375226e-06,
+      "loss": 0.4929,
+      "step": 5408
+    },
+    {
+      "epoch": 2.1202272165966907,
+      "grad_norm": 0.45411441910971817,
+      "learning_rate": 4.881198376177717e-06,
+      "loss": 0.49,
+      "step": 5409
+    },
+    {
+      "epoch": 2.120622375895283,
+      "grad_norm": 0.45916208922818735,
+      "learning_rate": 4.8811506486283075e-06,
+      "loss": 0.4891,
+      "step": 5410
+    },
+    {
+      "epoch": 2.121017535193875,
+      "grad_norm": 0.45426066117787084,
+      "learning_rate": 4.881102911727184e-06,
+      "loss": 0.4934,
+      "step": 5411
+    },
+    {
+      "epoch": 2.1214126944924674,
+      "grad_norm": 0.4676617535350159,
+      "learning_rate": 4.881055165474535e-06,
+      "loss": 0.5035,
+      "step": 5412
+    },
+    {
+      "epoch": 2.1218078537910596,
+      "grad_norm": 0.47051348318804287,
+      "learning_rate": 4.881007409870546e-06,
+      "loss": 0.5064,
+      "step": 5413
+    },
+    {
+      "epoch": 2.122203013089652,
+      "grad_norm": 0.4496266026428253,
+      "learning_rate": 4.880959644915406e-06,
+      "loss": 0.5047,
+      "step": 5414
+    },
+    {
+      "epoch": 2.122598172388244,
+      "grad_norm": 0.4595779522195174,
+      "learning_rate": 4.880911870609302e-06,
+      "loss": 0.4893,
+      "step": 5415
+    },
+    {
+      "epoch": 2.1229933316868363,
+      "grad_norm": 0.4569941314020582,
+      "learning_rate": 4.880864086952423e-06,
+      "loss": 0.4922,
+      "step": 5416
+    },
+    {
+      "epoch": 2.1233884909854286,
+      "grad_norm": 0.469415079482959,
+      "learning_rate": 4.880816293944955e-06,
+      "loss": 0.4938,
+      "step": 5417
+    },
+    {
+      "epoch": 2.123783650284021,
+      "grad_norm": 0.4447409912431963,
+      "learning_rate": 4.880768491587085e-06,
+      "loss": 0.5183,
+      "step": 5418
+    },
+    {
+      "epoch": 2.124178809582613,
+      "grad_norm": 0.45255535919526285,
+      "learning_rate": 4.880720679879004e-06,
+      "loss": 0.4958,
+      "step": 5419
+    },
+    {
+      "epoch": 2.1245739688812053,
+      "grad_norm": 0.4614305972818154,
+      "learning_rate": 4.880672858820897e-06,
+      "loss": 0.4974,
+      "step": 5420
+    },
+    {
+      "epoch": 2.1249691281797976,
+      "grad_norm": 0.47210760737931706,
+      "learning_rate": 4.880625028412952e-06,
+      "loss": 0.4918,
+      "step": 5421
+    },
+    {
+      "epoch": 2.12536428747839,
+      "grad_norm": 0.46822573176234256,
+      "learning_rate": 4.880577188655359e-06,
+      "loss": 0.5063,
+      "step": 5422
+    },
+    {
+      "epoch": 2.125759446776982,
+      "grad_norm": 0.480592867802166,
+      "learning_rate": 4.880529339548303e-06,
+      "loss": 0.4923,
+      "step": 5423
+    },
+    {
+      "epoch": 2.1261546060755743,
+      "grad_norm": 0.45661426326462,
+      "learning_rate": 4.880481481091974e-06,
+      "loss": 0.5092,
+      "step": 5424
+    },
+    {
+      "epoch": 2.1265497653741665,
+      "grad_norm": 0.4704280977766564,
+      "learning_rate": 4.8804336132865595e-06,
+      "loss": 0.5105,
+      "step": 5425
+    },
+    {
+      "epoch": 2.1269449246727588,
+      "grad_norm": 0.4618387396924572,
+      "learning_rate": 4.880385736132246e-06,
+      "loss": 0.4752,
+      "step": 5426
+    },
+    {
+      "epoch": 2.127340083971351,
+      "grad_norm": 0.45751503112924735,
+      "learning_rate": 4.8803378496292244e-06,
+      "loss": 0.4965,
+      "step": 5427
+    },
+    {
+      "epoch": 2.1277352432699432,
+      "grad_norm": 0.4641374268101188,
+      "learning_rate": 4.88028995377768e-06,
+      "loss": 0.5184,
+      "step": 5428
+    },
+    {
+      "epoch": 2.1281304025685355,
+      "grad_norm": 0.46565955389095914,
+      "learning_rate": 4.880242048577802e-06,
+      "loss": 0.5042,
+      "step": 5429
+    },
+    {
+      "epoch": 2.1285255618671277,
+      "grad_norm": 0.45584303689164063,
+      "learning_rate": 4.8801941340297795e-06,
+      "loss": 0.5004,
+      "step": 5430
+    },
+    {
+      "epoch": 2.12892072116572,
+      "grad_norm": 0.4710850087326468,
+      "learning_rate": 4.8801462101338e-06,
+      "loss": 0.5192,
+      "step": 5431
+    },
+    {
+      "epoch": 2.129315880464312,
+      "grad_norm": 0.44061458101376594,
+      "learning_rate": 4.88009827689005e-06,
+      "loss": 0.4933,
+      "step": 5432
+    },
+    {
+      "epoch": 2.1297110397629044,
+      "grad_norm": 0.4519911211493473,
+      "learning_rate": 4.88005033429872e-06,
+      "loss": 0.4949,
+      "step": 5433
+    },
+    {
+      "epoch": 2.1301061990614967,
+      "grad_norm": 0.4504066589256489,
+      "learning_rate": 4.880002382359998e-06,
+      "loss": 0.4791,
+      "step": 5434
+    },
+    {
+      "epoch": 2.130501358360089,
+      "grad_norm": 0.45764033049204544,
+      "learning_rate": 4.879954421074071e-06,
+      "loss": 0.4903,
+      "step": 5435
+    },
+    {
+      "epoch": 2.130896517658681,
+      "grad_norm": 0.4706731789649233,
+      "learning_rate": 4.879906450441129e-06,
+      "loss": 0.5127,
+      "step": 5436
+    },
+    {
+      "epoch": 2.1312916769572734,
+      "grad_norm": 0.44961446232384666,
+      "learning_rate": 4.8798584704613585e-06,
+      "loss": 0.4987,
+      "step": 5437
+    },
+    {
+      "epoch": 2.1316868362558656,
+      "grad_norm": 0.47280770131697647,
+      "learning_rate": 4.87981048113495e-06,
+      "loss": 0.5161,
+      "step": 5438
+    },
+    {
+      "epoch": 2.132081995554458,
+      "grad_norm": 0.46339098157068087,
+      "learning_rate": 4.879762482462091e-06,
+      "loss": 0.4894,
+      "step": 5439
+    },
+    {
+      "epoch": 2.13247715485305,
+      "grad_norm": 0.4487372634389615,
+      "learning_rate": 4.87971447444297e-06,
+      "loss": 0.5221,
+      "step": 5440
+    },
+    {
+      "epoch": 2.1328723141516424,
+      "grad_norm": 0.4562917626396976,
+      "learning_rate": 4.879666457077775e-06,
+      "loss": 0.524,
+      "step": 5441
+    },
+    {
+      "epoch": 2.1332674734502346,
+      "grad_norm": 0.4821307353976656,
+      "learning_rate": 4.879618430366696e-06,
+      "loss": 0.5079,
+      "step": 5442
+    },
+    {
+      "epoch": 2.133662632748827,
+      "grad_norm": 0.5353325642318681,
+      "learning_rate": 4.879570394309921e-06,
+      "loss": 0.5145,
+      "step": 5443
+    },
+    {
+      "epoch": 2.134057792047419,
+      "grad_norm": 0.47763771900604635,
+      "learning_rate": 4.879522348907637e-06,
+      "loss": 0.4835,
+      "step": 5444
+    },
+    {
+      "epoch": 2.1344529513460113,
+      "grad_norm": 0.44913009790356634,
+      "learning_rate": 4.879474294160035e-06,
+      "loss": 0.4757,
+      "step": 5445
+    },
+    {
+      "epoch": 2.1348481106446036,
+      "grad_norm": 0.4597607997906684,
+      "learning_rate": 4.879426230067303e-06,
+      "loss": 0.5005,
+      "step": 5446
+    },
+    {
+      "epoch": 2.135243269943196,
+      "grad_norm": 0.4623134190589581,
+      "learning_rate": 4.8793781566296294e-06,
+      "loss": 0.5043,
+      "step": 5447
+    },
+    {
+      "epoch": 2.135638429241788,
+      "grad_norm": 0.4506998220430185,
+      "learning_rate": 4.8793300738472025e-06,
+      "loss": 0.475,
+      "step": 5448
+    },
+    {
+      "epoch": 2.1360335885403803,
+      "grad_norm": 0.45507178878371285,
+      "learning_rate": 4.879281981720213e-06,
+      "loss": 0.498,
+      "step": 5449
+    },
+    {
+      "epoch": 2.1364287478389725,
+      "grad_norm": 0.46575574764050215,
+      "learning_rate": 4.879233880248848e-06,
+      "loss": 0.5143,
+      "step": 5450
+    },
+    {
+      "epoch": 2.1368239071375648,
+      "grad_norm": 0.4468946988002143,
+      "learning_rate": 4.879185769433298e-06,
+      "loss": 0.5058,
+      "step": 5451
+    },
+    {
+      "epoch": 2.137219066436157,
+      "grad_norm": 0.45705479305728886,
+      "learning_rate": 4.87913764927375e-06,
+      "loss": 0.4844,
+      "step": 5452
+    },
+    {
+      "epoch": 2.1376142257347492,
+      "grad_norm": 0.4535460076744396,
+      "learning_rate": 4.8790895197703945e-06,
+      "loss": 0.5187,
+      "step": 5453
+    },
+    {
+      "epoch": 2.1380093850333415,
+      "grad_norm": 0.4508238974301958,
+      "learning_rate": 4.879041380923421e-06,
+      "loss": 0.498,
+      "step": 5454
+    },
+    {
+      "epoch": 2.1384045443319337,
+      "grad_norm": 0.47611124879857625,
+      "learning_rate": 4.878993232733016e-06,
+      "loss": 0.5015,
+      "step": 5455
+    },
+    {
+      "epoch": 2.138799703630526,
+      "grad_norm": 0.47623598181038684,
+      "learning_rate": 4.8789450751993705e-06,
+      "loss": 0.4918,
+      "step": 5456
+    },
+    {
+      "epoch": 2.139194862929118,
+      "grad_norm": 0.4463328063702259,
+      "learning_rate": 4.878896908322673e-06,
+      "loss": 0.4949,
+      "step": 5457
+    },
+    {
+      "epoch": 2.1395900222277104,
+      "grad_norm": 0.4688848998545471,
+      "learning_rate": 4.878848732103114e-06,
+      "loss": 0.4784,
+      "step": 5458
+    },
+    {
+      "epoch": 2.1399851815263027,
+      "grad_norm": 0.44121522041937494,
+      "learning_rate": 4.878800546540881e-06,
+      "loss": 0.4849,
+      "step": 5459
+    },
+    {
+      "epoch": 2.140380340824895,
+      "grad_norm": 0.46258456174979196,
+      "learning_rate": 4.878752351636164e-06,
+      "loss": 0.498,
+      "step": 5460
+    },
+    {
+      "epoch": 2.140775500123487,
+      "grad_norm": 0.4515779539663671,
+      "learning_rate": 4.878704147389153e-06,
+      "loss": 0.5277,
+      "step": 5461
+    },
+    {
+      "epoch": 2.1411706594220794,
+      "grad_norm": 0.44675655713159357,
+      "learning_rate": 4.878655933800036e-06,
+      "loss": 0.4781,
+      "step": 5462
+    },
+    {
+      "epoch": 2.1415658187206716,
+      "grad_norm": 0.4787595797738257,
+      "learning_rate": 4.878607710869002e-06,
+      "loss": 0.5159,
+      "step": 5463
+    },
+    {
+      "epoch": 2.141960978019264,
+      "grad_norm": 0.46495746367564106,
+      "learning_rate": 4.878559478596242e-06,
+      "loss": 0.5156,
+      "step": 5464
+    },
+    {
+      "epoch": 2.142356137317856,
+      "grad_norm": 0.4502923794920052,
+      "learning_rate": 4.8785112369819455e-06,
+      "loss": 0.4891,
+      "step": 5465
+    },
+    {
+      "epoch": 2.1427512966164484,
+      "grad_norm": 0.451977669133556,
+      "learning_rate": 4.8784629860263e-06,
+      "loss": 0.5063,
+      "step": 5466
+    },
+    {
+      "epoch": 2.1431464559150406,
+      "grad_norm": 0.45073347613474773,
+      "learning_rate": 4.878414725729497e-06,
+      "loss": 0.5037,
+      "step": 5467
+    },
+    {
+      "epoch": 2.143541615213633,
+      "grad_norm": 0.4739115032234395,
+      "learning_rate": 4.878366456091724e-06,
+      "loss": 0.4945,
+      "step": 5468
+    },
+    {
+      "epoch": 2.143936774512225,
+      "grad_norm": 0.46171831924852413,
+      "learning_rate": 4.8783181771131735e-06,
+      "loss": 0.5097,
+      "step": 5469
+    },
+    {
+      "epoch": 2.1443319338108173,
+      "grad_norm": 0.4657774448437735,
+      "learning_rate": 4.878269888794032e-06,
+      "loss": 0.4919,
+      "step": 5470
+    },
+    {
+      "epoch": 2.1447270931094096,
+      "grad_norm": 0.47080691090966,
+      "learning_rate": 4.878221591134491e-06,
+      "loss": 0.5119,
+      "step": 5471
+    },
+    {
+      "epoch": 2.145122252408002,
+      "grad_norm": 0.4513544487027582,
+      "learning_rate": 4.8781732841347395e-06,
+      "loss": 0.4851,
+      "step": 5472
+    },
+    {
+      "epoch": 2.145517411706594,
+      "grad_norm": 0.47195481464231376,
+      "learning_rate": 4.878124967794968e-06,
+      "loss": 0.5161,
+      "step": 5473
+    },
+    {
+      "epoch": 2.1459125710051863,
+      "grad_norm": 0.4609154414666813,
+      "learning_rate": 4.878076642115366e-06,
+      "loss": 0.4985,
+      "step": 5474
+    },
+    {
+      "epoch": 2.1463077303037785,
+      "grad_norm": 0.46676548499669845,
+      "learning_rate": 4.878028307096122e-06,
+      "loss": 0.4724,
+      "step": 5475
+    },
+    {
+      "epoch": 2.1467028896023708,
+      "grad_norm": 0.4611191108433873,
+      "learning_rate": 4.8779799627374265e-06,
+      "loss": 0.4943,
+      "step": 5476
+    },
+    {
+      "epoch": 2.147098048900963,
+      "grad_norm": 0.4521869519269894,
+      "learning_rate": 4.877931609039471e-06,
+      "loss": 0.4997,
+      "step": 5477
+    },
+    {
+      "epoch": 2.1474932081995552,
+      "grad_norm": 0.505573034174829,
+      "learning_rate": 4.877883246002444e-06,
+      "loss": 0.5042,
+      "step": 5478
+    },
+    {
+      "epoch": 2.1478883674981475,
+      "grad_norm": 0.47546280424045617,
+      "learning_rate": 4.877834873626535e-06,
+      "loss": 0.5167,
+      "step": 5479
+    },
+    {
+      "epoch": 2.1482835267967397,
+      "grad_norm": 0.4634538760180319,
+      "learning_rate": 4.877786491911935e-06,
+      "loss": 0.4926,
+      "step": 5480
+    },
+    {
+      "epoch": 2.148678686095332,
+      "grad_norm": 0.4677531886753567,
+      "learning_rate": 4.877738100858832e-06,
+      "loss": 0.4891,
+      "step": 5481
+    },
+    {
+      "epoch": 2.149073845393924,
+      "grad_norm": 0.4690745882384834,
+      "learning_rate": 4.877689700467419e-06,
+      "loss": 0.5207,
+      "step": 5482
+    },
+    {
+      "epoch": 2.1494690046925164,
+      "grad_norm": 0.4692422873547366,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.4979,
+      "step": 5483
+    },
+    {
+      "epoch": 2.1498641639911087,
+      "grad_norm": 0.45669923066085627,
+      "learning_rate": 4.877592871670419e-06,
+      "loss": 0.5017,
+      "step": 5484
+    },
+    {
+      "epoch": 2.150259323289701,
+      "grad_norm": 0.4647143592549349,
+      "learning_rate": 4.877544443265212e-06,
+      "loss": 0.5213,
+      "step": 5485
+    },
+    {
+      "epoch": 2.150654482588293,
+      "grad_norm": 0.4487735037385304,
+      "learning_rate": 4.877496005522454e-06,
+      "loss": 0.4956,
+      "step": 5486
+    },
+    {
+      "epoch": 2.151049641886886,
+      "grad_norm": 0.4732217043292161,
+      "learning_rate": 4.877447558442335e-06,
+      "loss": 0.5098,
+      "step": 5487
+    },
+    {
+      "epoch": 2.151444801185478,
+      "grad_norm": 0.4868586481311084,
+      "learning_rate": 4.877399102025046e-06,
+      "loss": 0.4968,
+      "step": 5488
+    },
+    {
+      "epoch": 2.1518399604840703,
+      "grad_norm": 0.4679716331191469,
+      "learning_rate": 4.877350636270778e-06,
+      "loss": 0.5025,
+      "step": 5489
+    },
+    {
+      "epoch": 2.1522351197826626,
+      "grad_norm": 0.455600959345513,
+      "learning_rate": 4.87730216117972e-06,
+      "loss": 0.5006,
+      "step": 5490
+    },
+    {
+      "epoch": 2.152630279081255,
+      "grad_norm": 0.4729311871353748,
+      "learning_rate": 4.877253676752062e-06,
+      "loss": 0.4905,
+      "step": 5491
+    },
+    {
+      "epoch": 2.153025438379847,
+      "grad_norm": 0.47191264067689354,
+      "learning_rate": 4.877205182987995e-06,
+      "loss": 0.5033,
+      "step": 5492
+    },
+    {
+      "epoch": 2.1534205976784393,
+      "grad_norm": 0.4734254855624861,
+      "learning_rate": 4.87715667988771e-06,
+      "loss": 0.5113,
+      "step": 5493
+    },
+    {
+      "epoch": 2.1538157569770315,
+      "grad_norm": 0.46092003008011734,
+      "learning_rate": 4.8771081674513965e-06,
+      "loss": 0.4878,
+      "step": 5494
+    },
+    {
+      "epoch": 2.1542109162756238,
+      "grad_norm": 0.45908545472439594,
+      "learning_rate": 4.877059645679246e-06,
+      "loss": 0.4934,
+      "step": 5495
+    },
+    {
+      "epoch": 2.154606075574216,
+      "grad_norm": 0.4468868294393995,
+      "learning_rate": 4.877011114571449e-06,
+      "loss": 0.4947,
+      "step": 5496
+    },
+    {
+      "epoch": 2.1550012348728083,
+      "grad_norm": 0.444489465663762,
+      "learning_rate": 4.876962574128196e-06,
+      "loss": 0.4967,
+      "step": 5497
+    },
+    {
+      "epoch": 2.1553963941714005,
+      "grad_norm": 0.4393378102668866,
+      "learning_rate": 4.876914024349676e-06,
+      "loss": 0.4844,
+      "step": 5498
+    },
+    {
+      "epoch": 2.1557915534699927,
+      "grad_norm": 0.4640805678204216,
+      "learning_rate": 4.876865465236082e-06,
+      "loss": 0.5285,
+      "step": 5499
+    },
+    {
+      "epoch": 2.156186712768585,
+      "grad_norm": 0.5289230739266989,
+      "learning_rate": 4.876816896787603e-06,
+      "loss": 0.4989,
+      "step": 5500
+    },
+    {
+      "epoch": 2.156581872067177,
+      "grad_norm": 0.5419539291904234,
+      "learning_rate": 4.876768319004431e-06,
+      "loss": 0.5113,
+      "step": 5501
+    },
+    {
+      "epoch": 2.1569770313657695,
+      "grad_norm": 0.4517913697063531,
+      "learning_rate": 4.876719731886757e-06,
+      "loss": 0.4963,
+      "step": 5502
+    },
+    {
+      "epoch": 2.1573721906643617,
+      "grad_norm": 0.4504437690654093,
+      "learning_rate": 4.87667113543477e-06,
+      "loss": 0.4958,
+      "step": 5503
+    },
+    {
+      "epoch": 2.157767349962954,
+      "grad_norm": 0.4619724381275875,
+      "learning_rate": 4.876622529648663e-06,
+      "loss": 0.4998,
+      "step": 5504
+    },
+    {
+      "epoch": 2.158162509261546,
+      "grad_norm": 0.46378800765710737,
+      "learning_rate": 4.876573914528625e-06,
+      "loss": 0.5158,
+      "step": 5505
+    },
+    {
+      "epoch": 2.1585576685601384,
+      "grad_norm": 0.45305071404161723,
+      "learning_rate": 4.876525290074848e-06,
+      "loss": 0.4932,
+      "step": 5506
+    },
+    {
+      "epoch": 2.1589528278587307,
+      "grad_norm": 0.46744932250978694,
+      "learning_rate": 4.8764766562875235e-06,
+      "loss": 0.5075,
+      "step": 5507
+    },
+    {
+      "epoch": 2.159347987157323,
+      "grad_norm": 0.44330667766097176,
+      "learning_rate": 4.87642801316684e-06,
+      "loss": 0.4809,
+      "step": 5508
+    },
+    {
+      "epoch": 2.159743146455915,
+      "grad_norm": 0.46354232388964045,
+      "learning_rate": 4.876379360712993e-06,
+      "loss": 0.5029,
+      "step": 5509
+    },
+    {
+      "epoch": 2.1601383057545074,
+      "grad_norm": 0.46977655177705924,
+      "learning_rate": 4.876330698926169e-06,
+      "loss": 0.5001,
+      "step": 5510
+    },
+    {
+      "epoch": 2.1605334650530996,
+      "grad_norm": 0.45685117233330014,
+      "learning_rate": 4.876282027806561e-06,
+      "loss": 0.5151,
+      "step": 5511
+    },
+    {
+      "epoch": 2.160928624351692,
+      "grad_norm": 0.47320943013293004,
+      "learning_rate": 4.87623334735436e-06,
+      "loss": 0.5153,
+      "step": 5512
+    },
+    {
+      "epoch": 2.161323783650284,
+      "grad_norm": 0.45887136394693434,
+      "learning_rate": 4.876184657569759e-06,
+      "loss": 0.4955,
+      "step": 5513
+    },
+    {
+      "epoch": 2.1617189429488763,
+      "grad_norm": 0.45614045191471664,
+      "learning_rate": 4.876135958452946e-06,
+      "loss": 0.5055,
+      "step": 5514
+    },
+    {
+      "epoch": 2.1621141022474686,
+      "grad_norm": 0.4545945601618243,
+      "learning_rate": 4.876087250004114e-06,
+      "loss": 0.5069,
+      "step": 5515
+    },
+    {
+      "epoch": 2.162509261546061,
+      "grad_norm": 0.447681420448537,
+      "learning_rate": 4.876038532223454e-06,
+      "loss": 0.4999,
+      "step": 5516
+    },
+    {
+      "epoch": 2.162904420844653,
+      "grad_norm": 0.47058053249246107,
+      "learning_rate": 4.875989805111158e-06,
+      "loss": 0.4918,
+      "step": 5517
+    },
+    {
+      "epoch": 2.1632995801432453,
+      "grad_norm": 0.7930113043114789,
+      "learning_rate": 4.875941068667417e-06,
+      "loss": 0.4778,
+      "step": 5518
+    },
+    {
+      "epoch": 2.1636947394418375,
+      "grad_norm": 0.445244286794454,
+      "learning_rate": 4.875892322892421e-06,
+      "loss": 0.4965,
+      "step": 5519
+    },
+    {
+      "epoch": 2.16408989874043,
+      "grad_norm": 0.4577281165835478,
+      "learning_rate": 4.875843567786364e-06,
+      "loss": 0.4979,
+      "step": 5520
+    },
+    {
+      "epoch": 2.164485058039022,
+      "grad_norm": 0.464223962296654,
+      "learning_rate": 4.8757948033494365e-06,
+      "loss": 0.4995,
+      "step": 5521
+    },
+    {
+      "epoch": 2.1648802173376143,
+      "grad_norm": 0.44936416883238395,
+      "learning_rate": 4.875746029581828e-06,
+      "loss": 0.5189,
+      "step": 5522
+    },
+    {
+      "epoch": 2.1652753766362065,
+      "grad_norm": 0.45656574932816285,
+      "learning_rate": 4.875697246483733e-06,
+      "loss": 0.4947,
+      "step": 5523
+    },
+    {
+      "epoch": 2.1656705359347987,
+      "grad_norm": 0.45384101923535536,
+      "learning_rate": 4.875648454055341e-06,
+      "loss": 0.5006,
+      "step": 5524
+    },
+    {
+      "epoch": 2.166065695233391,
+      "grad_norm": 0.5008396943055093,
+      "learning_rate": 4.875599652296845e-06,
+      "loss": 0.4967,
+      "step": 5525
+    },
+    {
+      "epoch": 2.1664608545319832,
+      "grad_norm": 0.45982000765099335,
+      "learning_rate": 4.8755508412084364e-06,
+      "loss": 0.4969,
+      "step": 5526
+    },
+    {
+      "epoch": 2.1668560138305755,
+      "grad_norm": 0.45510607157877675,
+      "learning_rate": 4.875502020790306e-06,
+      "loss": 0.5089,
+      "step": 5527
+    },
+    {
+      "epoch": 2.1672511731291677,
+      "grad_norm": 0.45718988116180564,
+      "learning_rate": 4.875453191042646e-06,
+      "loss": 0.5141,
+      "step": 5528
+    },
+    {
+      "epoch": 2.16764633242776,
+      "grad_norm": 0.46235453264969417,
+      "learning_rate": 4.875404351965648e-06,
+      "loss": 0.4939,
+      "step": 5529
+    },
+    {
+      "epoch": 2.168041491726352,
+      "grad_norm": 0.45617899670326717,
+      "learning_rate": 4.875355503559506e-06,
+      "loss": 0.4993,
+      "step": 5530
+    },
+    {
+      "epoch": 2.1684366510249444,
+      "grad_norm": 0.47662452130332766,
+      "learning_rate": 4.875306645824408e-06,
+      "loss": 0.5141,
+      "step": 5531
+    },
+    {
+      "epoch": 2.1688318103235367,
+      "grad_norm": 0.4622948877225635,
+      "learning_rate": 4.875257778760549e-06,
+      "loss": 0.515,
+      "step": 5532
+    },
+    {
+      "epoch": 2.169226969622129,
+      "grad_norm": 0.4842991640779124,
+      "learning_rate": 4.8752089023681195e-06,
+      "loss": 0.5125,
+      "step": 5533
+    },
+    {
+      "epoch": 2.169622128920721,
+      "grad_norm": 0.48514678815733153,
+      "learning_rate": 4.875160016647311e-06,
+      "loss": 0.5048,
+      "step": 5534
+    },
+    {
+      "epoch": 2.1700172882193134,
+      "grad_norm": 0.45733286495961495,
+      "learning_rate": 4.875111121598317e-06,
+      "loss": 0.5015,
+      "step": 5535
+    },
+    {
+      "epoch": 2.1704124475179056,
+      "grad_norm": 0.45364214188982693,
+      "learning_rate": 4.875062217221329e-06,
+      "loss": 0.5115,
+      "step": 5536
+    },
+    {
+      "epoch": 2.170807606816498,
+      "grad_norm": 0.4421312637696224,
+      "learning_rate": 4.875013303516538e-06,
+      "loss": 0.4895,
+      "step": 5537
+    },
+    {
+      "epoch": 2.17120276611509,
+      "grad_norm": 0.4908166354246267,
+      "learning_rate": 4.874964380484138e-06,
+      "loss": 0.4812,
+      "step": 5538
+    },
+    {
+      "epoch": 2.1715979254136824,
+      "grad_norm": 0.4616856713589806,
+      "learning_rate": 4.874915448124319e-06,
+      "loss": 0.4974,
+      "step": 5539
+    },
+    {
+      "epoch": 2.1719930847122746,
+      "grad_norm": 0.4646457406849082,
+      "learning_rate": 4.874866506437275e-06,
+      "loss": 0.5034,
+      "step": 5540
+    },
+    {
+      "epoch": 2.172388244010867,
+      "grad_norm": 0.45047160391577984,
+      "learning_rate": 4.874817555423196e-06,
+      "loss": 0.4877,
+      "step": 5541
+    },
+    {
+      "epoch": 2.172783403309459,
+      "grad_norm": 0.4500708056755459,
+      "learning_rate": 4.874768595082277e-06,
+      "loss": 0.4918,
+      "step": 5542
+    },
+    {
+      "epoch": 2.1731785626080513,
+      "grad_norm": 0.47456977571950365,
+      "learning_rate": 4.874719625414709e-06,
+      "loss": 0.4933,
+      "step": 5543
+    },
+    {
+      "epoch": 2.1735737219066436,
+      "grad_norm": 0.46200205295141555,
+      "learning_rate": 4.874670646420684e-06,
+      "loss": 0.5142,
+      "step": 5544
+    },
+    {
+      "epoch": 2.173968881205236,
+      "grad_norm": 0.4734243176175764,
+      "learning_rate": 4.874621658100395e-06,
+      "loss": 0.5208,
+      "step": 5545
+    },
+    {
+      "epoch": 2.174364040503828,
+      "grad_norm": 0.44815049758957626,
+      "learning_rate": 4.874572660454034e-06,
+      "loss": 0.4968,
+      "step": 5546
+    },
+    {
+      "epoch": 2.1747591998024203,
+      "grad_norm": 0.4614058105293498,
+      "learning_rate": 4.874523653481793e-06,
+      "loss": 0.5072,
+      "step": 5547
+    },
+    {
+      "epoch": 2.1751543591010125,
+      "grad_norm": 0.44451885664415813,
+      "learning_rate": 4.874474637183866e-06,
+      "loss": 0.5043,
+      "step": 5548
+    },
+    {
+      "epoch": 2.1755495183996048,
+      "grad_norm": 0.4591881542386408,
+      "learning_rate": 4.874425611560444e-06,
+      "loss": 0.4854,
+      "step": 5549
+    },
+    {
+      "epoch": 2.175944677698197,
+      "grad_norm": 0.47533595495611164,
+      "learning_rate": 4.874376576611719e-06,
+      "loss": 0.5194,
+      "step": 5550
+    },
+    {
+      "epoch": 2.1763398369967892,
+      "grad_norm": 0.46610351057678207,
+      "learning_rate": 4.874327532337886e-06,
+      "loss": 0.5002,
+      "step": 5551
+    },
+    {
+      "epoch": 2.1767349962953815,
+      "grad_norm": 0.4633556164186955,
+      "learning_rate": 4.8742784787391355e-06,
+      "loss": 0.5062,
+      "step": 5552
+    },
+    {
+      "epoch": 2.1771301555939737,
+      "grad_norm": 0.4676833644787226,
+      "learning_rate": 4.874229415815661e-06,
+      "loss": 0.5077,
+      "step": 5553
+    },
+    {
+      "epoch": 2.177525314892566,
+      "grad_norm": 0.4861231482509832,
+      "learning_rate": 4.874180343567655e-06,
+      "loss": 0.5149,
+      "step": 5554
+    },
+    {
+      "epoch": 2.177920474191158,
+      "grad_norm": 0.47393974173121844,
+      "learning_rate": 4.8741312619953106e-06,
+      "loss": 0.4931,
+      "step": 5555
+    },
+    {
+      "epoch": 2.1783156334897504,
+      "grad_norm": 0.46438049892681393,
+      "learning_rate": 4.87408217109882e-06,
+      "loss": 0.4922,
+      "step": 5556
+    },
+    {
+      "epoch": 2.1787107927883427,
+      "grad_norm": 0.480552674800388,
+      "learning_rate": 4.874033070878377e-06,
+      "loss": 0.5093,
+      "step": 5557
+    },
+    {
+      "epoch": 2.179105952086935,
+      "grad_norm": 0.49032805311209543,
+      "learning_rate": 4.873983961334172e-06,
+      "loss": 0.5129,
+      "step": 5558
+    },
+    {
+      "epoch": 2.179501111385527,
+      "grad_norm": 0.46500235243264837,
+      "learning_rate": 4.873934842466401e-06,
+      "loss": 0.4974,
+      "step": 5559
+    },
+    {
+      "epoch": 2.1798962706841194,
+      "grad_norm": 0.4601347793294418,
+      "learning_rate": 4.873885714275255e-06,
+      "loss": 0.5178,
+      "step": 5560
+    },
+    {
+      "epoch": 2.1802914299827116,
+      "grad_norm": 0.4601391528297951,
+      "learning_rate": 4.873836576760927e-06,
+      "loss": 0.501,
+      "step": 5561
+    },
+    {
+      "epoch": 2.180686589281304,
+      "grad_norm": 0.5067694421292206,
+      "learning_rate": 4.873787429923611e-06,
+      "loss": 0.484,
+      "step": 5562
+    },
+    {
+      "epoch": 2.181081748579896,
+      "grad_norm": 0.4663242902766598,
+      "learning_rate": 4.8737382737635e-06,
+      "loss": 0.5084,
+      "step": 5563
+    },
+    {
+      "epoch": 2.1814769078784884,
+      "grad_norm": 0.4553398223962099,
+      "learning_rate": 4.873689108280786e-06,
+      "loss": 0.4965,
+      "step": 5564
+    },
+    {
+      "epoch": 2.1818720671770806,
+      "grad_norm": 0.529541497892225,
+      "learning_rate": 4.873639933475662e-06,
+      "loss": 0.4955,
+      "step": 5565
+    },
+    {
+      "epoch": 2.182267226475673,
+      "grad_norm": 0.4713945957848912,
+      "learning_rate": 4.8735907493483216e-06,
+      "loss": 0.5047,
+      "step": 5566
+    },
+    {
+      "epoch": 2.182662385774265,
+      "grad_norm": 0.4572708785437191,
+      "learning_rate": 4.873541555898959e-06,
+      "loss": 0.4803,
+      "step": 5567
+    },
+    {
+      "epoch": 2.1830575450728573,
+      "grad_norm": 0.45611106467714163,
+      "learning_rate": 4.873492353127765e-06,
+      "loss": 0.4936,
+      "step": 5568
+    },
+    {
+      "epoch": 2.1834527043714496,
+      "grad_norm": 0.4595792613339848,
+      "learning_rate": 4.873443141034936e-06,
+      "loss": 0.4874,
+      "step": 5569
+    },
+    {
+      "epoch": 2.183847863670042,
+      "grad_norm": 0.4640866973139749,
+      "learning_rate": 4.873393919620663e-06,
+      "loss": 0.4924,
+      "step": 5570
+    },
+    {
+      "epoch": 2.184243022968634,
+      "grad_norm": 0.46111098813030116,
+      "learning_rate": 4.873344688885139e-06,
+      "loss": 0.5057,
+      "step": 5571
+    },
+    {
+      "epoch": 2.1846381822672263,
+      "grad_norm": 0.46999556887546373,
+      "learning_rate": 4.873295448828559e-06,
+      "loss": 0.5093,
+      "step": 5572
+    },
+    {
+      "epoch": 2.1850333415658185,
+      "grad_norm": 0.4718054070690503,
+      "learning_rate": 4.873246199451116e-06,
+      "loss": 0.4879,
+      "step": 5573
+    },
+    {
+      "epoch": 2.1854285008644108,
+      "grad_norm": 0.4710239940304868,
+      "learning_rate": 4.873196940753002e-06,
+      "loss": 0.4993,
+      "step": 5574
+    },
+    {
+      "epoch": 2.1858236601630034,
+      "grad_norm": 0.49030453082387426,
+      "learning_rate": 4.873147672734412e-06,
+      "loss": 0.5107,
+      "step": 5575
+    },
+    {
+      "epoch": 2.1862188194615957,
+      "grad_norm": 0.46406412622436066,
+      "learning_rate": 4.873098395395539e-06,
+      "loss": 0.5081,
+      "step": 5576
+    },
+    {
+      "epoch": 2.186613978760188,
+      "grad_norm": 0.4628191661905533,
+      "learning_rate": 4.873049108736577e-06,
+      "loss": 0.5031,
+      "step": 5577
+    },
+    {
+      "epoch": 2.18700913805878,
+      "grad_norm": 0.44331929710296053,
+      "learning_rate": 4.872999812757718e-06,
+      "loss": 0.5046,
+      "step": 5578
+    },
+    {
+      "epoch": 2.1874042973573724,
+      "grad_norm": 0.446907560016826,
+      "learning_rate": 4.872950507459158e-06,
+      "loss": 0.5187,
+      "step": 5579
+    },
+    {
+      "epoch": 2.1877994566559646,
+      "grad_norm": 0.45847831566326597,
+      "learning_rate": 4.872901192841089e-06,
+      "loss": 0.4828,
+      "step": 5580
+    },
+    {
+      "epoch": 2.188194615954557,
+      "grad_norm": 0.4487958830148146,
+      "learning_rate": 4.872851868903704e-06,
+      "loss": 0.5003,
+      "step": 5581
+    },
+    {
+      "epoch": 2.188589775253149,
+      "grad_norm": 0.4535135827385076,
+      "learning_rate": 4.872802535647199e-06,
+      "loss": 0.5003,
+      "step": 5582
+    },
+    {
+      "epoch": 2.1889849345517414,
+      "grad_norm": 0.4375345693693656,
+      "learning_rate": 4.872753193071766e-06,
+      "loss": 0.4881,
+      "step": 5583
+    },
+    {
+      "epoch": 2.1893800938503336,
+      "grad_norm": 0.4641343320330823,
+      "learning_rate": 4.872703841177599e-06,
+      "loss": 0.4941,
+      "step": 5584
+    },
+    {
+      "epoch": 2.189775253148926,
+      "grad_norm": 0.45644605766947677,
+      "learning_rate": 4.872654479964892e-06,
+      "loss": 0.5066,
+      "step": 5585
+    },
+    {
+      "epoch": 2.190170412447518,
+      "grad_norm": 0.46839020316567503,
+      "learning_rate": 4.87260510943384e-06,
+      "loss": 0.5116,
+      "step": 5586
+    },
+    {
+      "epoch": 2.1905655717461103,
+      "grad_norm": 0.4731473500021345,
+      "learning_rate": 4.872555729584635e-06,
+      "loss": 0.5096,
+      "step": 5587
+    },
+    {
+      "epoch": 2.1909607310447026,
+      "grad_norm": 0.4433081776671587,
+      "learning_rate": 4.872506340417471e-06,
+      "loss": 0.4933,
+      "step": 5588
+    },
+    {
+      "epoch": 2.191355890343295,
+      "grad_norm": 0.48768292944297426,
+      "learning_rate": 4.872456941932544e-06,
+      "loss": 0.507,
+      "step": 5589
+    },
+    {
+      "epoch": 2.191751049641887,
+      "grad_norm": 0.4436677280755508,
+      "learning_rate": 4.872407534130047e-06,
+      "loss": 0.5024,
+      "step": 5590
+    },
+    {
+      "epoch": 2.1921462089404793,
+      "grad_norm": 0.4536188851299345,
+      "learning_rate": 4.8723581170101734e-06,
+      "loss": 0.4982,
+      "step": 5591
+    },
+    {
+      "epoch": 2.1925413682390715,
+      "grad_norm": 0.44412822568870913,
+      "learning_rate": 4.872308690573118e-06,
+      "loss": 0.4957,
+      "step": 5592
+    },
+    {
+      "epoch": 2.1929365275376638,
+      "grad_norm": 0.6642864573262021,
+      "learning_rate": 4.872259254819073e-06,
+      "loss": 0.5231,
+      "step": 5593
+    },
+    {
+      "epoch": 2.193331686836256,
+      "grad_norm": 0.46456261247267744,
+      "learning_rate": 4.872209809748236e-06,
+      "loss": 0.5085,
+      "step": 5594
+    },
+    {
+      "epoch": 2.1937268461348483,
+      "grad_norm": 0.45989758175778567,
+      "learning_rate": 4.872160355360798e-06,
+      "loss": 0.4912,
+      "step": 5595
+    },
+    {
+      "epoch": 2.1941220054334405,
+      "grad_norm": 0.47722595433382775,
+      "learning_rate": 4.8721108916569555e-06,
+      "loss": 0.5026,
+      "step": 5596
+    },
+    {
+      "epoch": 2.1945171647320327,
+      "grad_norm": 0.4588929552955127,
+      "learning_rate": 4.872061418636902e-06,
+      "loss": 0.4941,
+      "step": 5597
+    },
+    {
+      "epoch": 2.194912324030625,
+      "grad_norm": 0.43968483295091,
+      "learning_rate": 4.872011936300831e-06,
+      "loss": 0.5046,
+      "step": 5598
+    },
+    {
+      "epoch": 2.195307483329217,
+      "grad_norm": 0.44462587632009504,
+      "learning_rate": 4.871962444648938e-06,
+      "loss": 0.4819,
+      "step": 5599
+    },
+    {
+      "epoch": 2.1957026426278095,
+      "grad_norm": 0.45683508501671893,
+      "learning_rate": 4.871912943681416e-06,
+      "loss": 0.504,
+      "step": 5600
+    },
+    {
+      "epoch": 2.1960978019264017,
+      "grad_norm": 0.4446678291894044,
+      "learning_rate": 4.87186343339846e-06,
+      "loss": 0.4973,
+      "step": 5601
+    },
+    {
+      "epoch": 2.196492961224994,
+      "grad_norm": 0.4736599203705845,
+      "learning_rate": 4.871813913800266e-06,
+      "loss": 0.5027,
+      "step": 5602
+    },
+    {
+      "epoch": 2.196888120523586,
+      "grad_norm": 0.45846116866212805,
+      "learning_rate": 4.8717643848870265e-06,
+      "loss": 0.5032,
+      "step": 5603
+    },
+    {
+      "epoch": 2.1972832798221784,
+      "grad_norm": 0.5481111316900571,
+      "learning_rate": 4.871714846658937e-06,
+      "loss": 0.5066,
+      "step": 5604
+    },
+    {
+      "epoch": 2.1976784391207707,
+      "grad_norm": 0.45324845161744187,
+      "learning_rate": 4.871665299116192e-06,
+      "loss": 0.496,
+      "step": 5605
+    },
+    {
+      "epoch": 2.198073598419363,
+      "grad_norm": 0.46469652330820316,
+      "learning_rate": 4.871615742258985e-06,
+      "loss": 0.5095,
+      "step": 5606
+    },
+    {
+      "epoch": 2.198468757717955,
+      "grad_norm": 0.5196428348937057,
+      "learning_rate": 4.871566176087512e-06,
+      "loss": 0.4955,
+      "step": 5607
+    },
+    {
+      "epoch": 2.1988639170165474,
+      "grad_norm": 0.47622508477020703,
+      "learning_rate": 4.871516600601968e-06,
+      "loss": 0.5222,
+      "step": 5608
+    },
+    {
+      "epoch": 2.1992590763151396,
+      "grad_norm": 0.4644820525166222,
+      "learning_rate": 4.871467015802545e-06,
+      "loss": 0.5046,
+      "step": 5609
+    },
+    {
+      "epoch": 2.199654235613732,
+      "grad_norm": 0.4563914565316861,
+      "learning_rate": 4.871417421689442e-06,
+      "loss": 0.502,
+      "step": 5610
+    },
+    {
+      "epoch": 2.200049394912324,
+      "grad_norm": 0.4521533603619674,
+      "learning_rate": 4.871367818262849e-06,
+      "loss": 0.4989,
+      "step": 5611
+    },
+    {
+      "epoch": 2.2004445542109163,
+      "grad_norm": 0.4679852547890778,
+      "learning_rate": 4.871318205522965e-06,
+      "loss": 0.5271,
+      "step": 5612
+    },
+    {
+      "epoch": 2.2008397135095086,
+      "grad_norm": 0.44239899933907284,
+      "learning_rate": 4.871268583469982e-06,
+      "loss": 0.4922,
+      "step": 5613
+    },
+    {
+      "epoch": 2.201234872808101,
+      "grad_norm": 0.46044033765149545,
+      "learning_rate": 4.8712189521040955e-06,
+      "loss": 0.505,
+      "step": 5614
+    },
+    {
+      "epoch": 2.201630032106693,
+      "grad_norm": 0.4509076270729074,
+      "learning_rate": 4.871169311425501e-06,
+      "loss": 0.4841,
+      "step": 5615
+    },
+    {
+      "epoch": 2.2020251914052853,
+      "grad_norm": 0.46744577639760204,
+      "learning_rate": 4.871119661434395e-06,
+      "loss": 0.4923,
+      "step": 5616
+    },
+    {
+      "epoch": 2.2024203507038775,
+      "grad_norm": 0.46269441156741004,
+      "learning_rate": 4.871070002130968e-06,
+      "loss": 0.5057,
+      "step": 5617
+    },
+    {
+      "epoch": 2.20281551000247,
+      "grad_norm": 0.4615270330282239,
+      "learning_rate": 4.871020333515421e-06,
+      "loss": 0.4931,
+      "step": 5618
+    },
+    {
+      "epoch": 2.203210669301062,
+      "grad_norm": 0.45781453501465114,
+      "learning_rate": 4.870970655587943e-06,
+      "loss": 0.4993,
+      "step": 5619
+    },
+    {
+      "epoch": 2.2036058285996543,
+      "grad_norm": 0.46448710878947935,
+      "learning_rate": 4.870920968348734e-06,
+      "loss": 0.513,
+      "step": 5620
+    },
+    {
+      "epoch": 2.2040009878982465,
+      "grad_norm": 0.45923364426870794,
+      "learning_rate": 4.870871271797986e-06,
+      "loss": 0.4827,
+      "step": 5621
+    },
+    {
+      "epoch": 2.2043961471968387,
+      "grad_norm": 0.4663678964949085,
+      "learning_rate": 4.870821565935896e-06,
+      "loss": 0.487,
+      "step": 5622
+    },
+    {
+      "epoch": 2.204791306495431,
+      "grad_norm": 0.4652080789551877,
+      "learning_rate": 4.870771850762658e-06,
+      "loss": 0.5101,
+      "step": 5623
+    },
+    {
+      "epoch": 2.205186465794023,
+      "grad_norm": 0.7697658307753772,
+      "learning_rate": 4.870722126278468e-06,
+      "loss": 0.4702,
+      "step": 5624
+    },
+    {
+      "epoch": 2.2055816250926155,
+      "grad_norm": 0.46255357306164907,
+      "learning_rate": 4.870672392483521e-06,
+      "loss": 0.5035,
+      "step": 5625
+    },
+    {
+      "epoch": 2.2059767843912077,
+      "grad_norm": 0.49091601504343524,
+      "learning_rate": 4.870622649378012e-06,
+      "loss": 0.5232,
+      "step": 5626
+    },
+    {
+      "epoch": 2.2063719436898,
+      "grad_norm": 0.46764422378827775,
+      "learning_rate": 4.870572896962138e-06,
+      "loss": 0.4822,
+      "step": 5627
+    },
+    {
+      "epoch": 2.206767102988392,
+      "grad_norm": 0.4580646300847785,
+      "learning_rate": 4.870523135236092e-06,
+      "loss": 0.516,
+      "step": 5628
+    },
+    {
+      "epoch": 2.2071622622869844,
+      "grad_norm": 0.4716946592361151,
+      "learning_rate": 4.8704733642000714e-06,
+      "loss": 0.5058,
+      "step": 5629
+    },
+    {
+      "epoch": 2.2075574215855767,
+      "grad_norm": 0.4610497435658121,
+      "learning_rate": 4.8704235838542705e-06,
+      "loss": 0.5128,
+      "step": 5630
+    },
+    {
+      "epoch": 2.207952580884169,
+      "grad_norm": 0.46580653784035464,
+      "learning_rate": 4.870373794198885e-06,
+      "loss": 0.5168,
+      "step": 5631
+    },
+    {
+      "epoch": 2.208347740182761,
+      "grad_norm": 0.45033186180479673,
+      "learning_rate": 4.870323995234109e-06,
+      "loss": 0.4908,
+      "step": 5632
+    },
+    {
+      "epoch": 2.2087428994813534,
+      "grad_norm": 0.4623037931279241,
+      "learning_rate": 4.870274186960142e-06,
+      "loss": 0.5007,
+      "step": 5633
+    },
+    {
+      "epoch": 2.2091380587799456,
+      "grad_norm": 0.4543745068546596,
+      "learning_rate": 4.870224369377176e-06,
+      "loss": 0.4904,
+      "step": 5634
+    },
+    {
+      "epoch": 2.209533218078538,
+      "grad_norm": 0.4525710119538143,
+      "learning_rate": 4.87017454248541e-06,
+      "loss": 0.4789,
+      "step": 5635
+    },
+    {
+      "epoch": 2.20992837737713,
+      "grad_norm": 0.475347791482502,
+      "learning_rate": 4.870124706285036e-06,
+      "loss": 0.5116,
+      "step": 5636
+    },
+    {
+      "epoch": 2.2103235366757223,
+      "grad_norm": 0.4668581607053355,
+      "learning_rate": 4.8700748607762515e-06,
+      "loss": 0.4948,
+      "step": 5637
+    },
+    {
+      "epoch": 2.2107186959743146,
+      "grad_norm": 0.4562173473097316,
+      "learning_rate": 4.870025005959252e-06,
+      "loss": 0.5139,
+      "step": 5638
+    },
+    {
+      "epoch": 2.211113855272907,
+      "grad_norm": 0.4638745581699392,
+      "learning_rate": 4.869975141834234e-06,
+      "loss": 0.5042,
+      "step": 5639
+    },
+    {
+      "epoch": 2.211509014571499,
+      "grad_norm": 0.45690466092589344,
+      "learning_rate": 4.869925268401392e-06,
+      "loss": 0.5122,
+      "step": 5640
+    },
+    {
+      "epoch": 2.2119041738700913,
+      "grad_norm": 0.46129161407797814,
+      "learning_rate": 4.869875385660923e-06,
+      "loss": 0.5034,
+      "step": 5641
+    },
+    {
+      "epoch": 2.2122993331686835,
+      "grad_norm": 0.45249931427852574,
+      "learning_rate": 4.869825493613023e-06,
+      "loss": 0.4944,
+      "step": 5642
+    },
+    {
+      "epoch": 2.212694492467276,
+      "grad_norm": 0.4717240333098036,
+      "learning_rate": 4.869775592257887e-06,
+      "loss": 0.5079,
+      "step": 5643
+    },
+    {
+      "epoch": 2.213089651765868,
+      "grad_norm": 0.45646269978521453,
+      "learning_rate": 4.869725681595712e-06,
+      "loss": 0.5086,
+      "step": 5644
+    },
+    {
+      "epoch": 2.2134848110644603,
+      "grad_norm": 0.4660158700586118,
+      "learning_rate": 4.869675761626693e-06,
+      "loss": 0.5024,
+      "step": 5645
+    },
+    {
+      "epoch": 2.2138799703630525,
+      "grad_norm": 0.46107902427465175,
+      "learning_rate": 4.869625832351026e-06,
+      "loss": 0.4842,
+      "step": 5646
+    },
+    {
+      "epoch": 2.2142751296616447,
+      "grad_norm": 0.4610113532446741,
+      "learning_rate": 4.869575893768909e-06,
+      "loss": 0.5025,
+      "step": 5647
+    },
+    {
+      "epoch": 2.214670288960237,
+      "grad_norm": 0.46330124836282083,
+      "learning_rate": 4.869525945880536e-06,
+      "loss": 0.5011,
+      "step": 5648
+    },
+    {
+      "epoch": 2.2150654482588292,
+      "grad_norm": 0.44825650922902094,
+      "learning_rate": 4.869475988686105e-06,
+      "loss": 0.5132,
+      "step": 5649
+    },
+    {
+      "epoch": 2.2154606075574215,
+      "grad_norm": 0.438523700300251,
+      "learning_rate": 4.8694260221858095e-06,
+      "loss": 0.4888,
+      "step": 5650
+    },
+    {
+      "epoch": 2.2158557668560137,
+      "grad_norm": 0.446477475948747,
+      "learning_rate": 4.869376046379848e-06,
+      "loss": 0.5027,
+      "step": 5651
+    },
+    {
+      "epoch": 2.216250926154606,
+      "grad_norm": 0.47359530183611437,
+      "learning_rate": 4.869326061268416e-06,
+      "loss": 0.5057,
+      "step": 5652
+    },
+    {
+      "epoch": 2.216646085453198,
+      "grad_norm": 0.45346141292171366,
+      "learning_rate": 4.869276066851711e-06,
+      "loss": 0.5238,
+      "step": 5653
+    },
+    {
+      "epoch": 2.2170412447517904,
+      "grad_norm": 0.4697574903847061,
+      "learning_rate": 4.869226063129926e-06,
+      "loss": 0.5157,
+      "step": 5654
+    },
+    {
+      "epoch": 2.2174364040503827,
+      "grad_norm": 0.4992277490563748,
+      "learning_rate": 4.869176050103262e-06,
+      "loss": 0.5312,
+      "step": 5655
+    },
+    {
+      "epoch": 2.217831563348975,
+      "grad_norm": 0.4463421488924404,
+      "learning_rate": 4.869126027771912e-06,
+      "loss": 0.4857,
+      "step": 5656
+    },
+    {
+      "epoch": 2.218226722647567,
+      "grad_norm": 0.445536527726352,
+      "learning_rate": 4.8690759961360736e-06,
+      "loss": 0.511,
+      "step": 5657
+    },
+    {
+      "epoch": 2.2186218819461594,
+      "grad_norm": 0.4608081468830687,
+      "learning_rate": 4.869025955195944e-06,
+      "loss": 0.4942,
+      "step": 5658
+    },
+    {
+      "epoch": 2.2190170412447516,
+      "grad_norm": 0.471512152814528,
+      "learning_rate": 4.868975904951718e-06,
+      "loss": 0.5061,
+      "step": 5659
+    },
+    {
+      "epoch": 2.219412200543344,
+      "grad_norm": 0.4550754501913592,
+      "learning_rate": 4.868925845403594e-06,
+      "loss": 0.4992,
+      "step": 5660
+    },
+    {
+      "epoch": 2.219807359841936,
+      "grad_norm": 0.456877328887177,
+      "learning_rate": 4.868875776551767e-06,
+      "loss": 0.5084,
+      "step": 5661
+    },
+    {
+      "epoch": 2.2202025191405284,
+      "grad_norm": 0.44849181279718897,
+      "learning_rate": 4.868825698396435e-06,
+      "loss": 0.4798,
+      "step": 5662
+    },
+    {
+      "epoch": 2.2205976784391206,
+      "grad_norm": 0.4478184080732161,
+      "learning_rate": 4.8687756109377935e-06,
+      "loss": 0.4878,
+      "step": 5663
+    },
+    {
+      "epoch": 2.220992837737713,
+      "grad_norm": 0.4558088044427646,
+      "learning_rate": 4.86872551417604e-06,
+      "loss": 0.5042,
+      "step": 5664
+    },
+    {
+      "epoch": 2.221387997036305,
+      "grad_norm": 0.45237458187762586,
+      "learning_rate": 4.8686754081113715e-06,
+      "loss": 0.4871,
+      "step": 5665
+    },
+    {
+      "epoch": 2.2217831563348973,
+      "grad_norm": 0.4344840693507297,
+      "learning_rate": 4.868625292743985e-06,
+      "loss": 0.4937,
+      "step": 5666
+    },
+    {
+      "epoch": 2.2221783156334896,
+      "grad_norm": 0.45680335357644025,
+      "learning_rate": 4.868575168074075e-06,
+      "loss": 0.5051,
+      "step": 5667
+    },
+    {
+      "epoch": 2.222573474932082,
+      "grad_norm": 0.4786605604468794,
+      "learning_rate": 4.8685250341018405e-06,
+      "loss": 0.5145,
+      "step": 5668
+    },
+    {
+      "epoch": 2.222968634230674,
+      "grad_norm": 0.45133256657617904,
+      "learning_rate": 4.868474890827479e-06,
+      "loss": 0.4852,
+      "step": 5669
+    },
+    {
+      "epoch": 2.2233637935292663,
+      "grad_norm": 0.45874816511999167,
+      "learning_rate": 4.8684247382511855e-06,
+      "loss": 0.506,
+      "step": 5670
+    },
+    {
+      "epoch": 2.2237589528278585,
+      "grad_norm": 0.47105938523376145,
+      "learning_rate": 4.868374576373157e-06,
+      "loss": 0.5193,
+      "step": 5671
+    },
+    {
+      "epoch": 2.2241541121264508,
+      "grad_norm": 0.4499761946748633,
+      "learning_rate": 4.868324405193593e-06,
+      "loss": 0.4957,
+      "step": 5672
+    },
+    {
+      "epoch": 2.224549271425043,
+      "grad_norm": 0.44396274883575054,
+      "learning_rate": 4.868274224712688e-06,
+      "loss": 0.4968,
+      "step": 5673
+    },
+    {
+      "epoch": 2.2249444307236352,
+      "grad_norm": 0.45650883917884044,
+      "learning_rate": 4.86822403493064e-06,
+      "loss": 0.4963,
+      "step": 5674
+    },
+    {
+      "epoch": 2.2253395900222275,
+      "grad_norm": 0.45717604881834667,
+      "learning_rate": 4.868173835847646e-06,
+      "loss": 0.5214,
+      "step": 5675
+    },
+    {
+      "epoch": 2.22573474932082,
+      "grad_norm": 0.4587246120001337,
+      "learning_rate": 4.8681236274639024e-06,
+      "loss": 0.4964,
+      "step": 5676
+    },
+    {
+      "epoch": 2.2261299086194124,
+      "grad_norm": 0.46577038706127155,
+      "learning_rate": 4.868073409779609e-06,
+      "loss": 0.501,
+      "step": 5677
+    },
+    {
+      "epoch": 2.2265250679180046,
+      "grad_norm": 0.45292880378261396,
+      "learning_rate": 4.86802318279496e-06,
+      "loss": 0.5076,
+      "step": 5678
+    },
+    {
+      "epoch": 2.226920227216597,
+      "grad_norm": 0.46823804273718916,
+      "learning_rate": 4.867972946510154e-06,
+      "loss": 0.4996,
+      "step": 5679
+    },
+    {
+      "epoch": 2.227315386515189,
+      "grad_norm": 0.4651990124613078,
+      "learning_rate": 4.867922700925388e-06,
+      "loss": 0.4958,
+      "step": 5680
+    },
+    {
+      "epoch": 2.2277105458137814,
+      "grad_norm": 0.48024213114162967,
+      "learning_rate": 4.86787244604086e-06,
+      "loss": 0.5189,
+      "step": 5681
+    },
+    {
+      "epoch": 2.2281057051123736,
+      "grad_norm": 0.5081055107145902,
+      "learning_rate": 4.867822181856766e-06,
+      "loss": 0.4902,
+      "step": 5682
+    },
+    {
+      "epoch": 2.228500864410966,
+      "grad_norm": 0.4430423080870507,
+      "learning_rate": 4.867771908373306e-06,
+      "loss": 0.5037,
+      "step": 5683
+    },
+    {
+      "epoch": 2.228896023709558,
+      "grad_norm": 0.46192941764536816,
+      "learning_rate": 4.867721625590674e-06,
+      "loss": 0.5109,
+      "step": 5684
+    },
+    {
+      "epoch": 2.2292911830081503,
+      "grad_norm": 0.47441393038250607,
+      "learning_rate": 4.8676713335090694e-06,
+      "loss": 0.4973,
+      "step": 5685
+    },
+    {
+      "epoch": 2.2296863423067426,
+      "grad_norm": 0.4466027577701095,
+      "learning_rate": 4.867621032128691e-06,
+      "loss": 0.4853,
+      "step": 5686
+    },
+    {
+      "epoch": 2.230081501605335,
+      "grad_norm": 0.46435968422699964,
+      "learning_rate": 4.867570721449734e-06,
+      "loss": 0.5097,
+      "step": 5687
+    },
+    {
+      "epoch": 2.230476660903927,
+      "grad_norm": 0.45475109267398633,
+      "learning_rate": 4.867520401472396e-06,
+      "loss": 0.4937,
+      "step": 5688
+    },
+    {
+      "epoch": 2.2308718202025193,
+      "grad_norm": 0.46084767398038173,
+      "learning_rate": 4.867470072196876e-06,
+      "loss": 0.4913,
+      "step": 5689
+    },
+    {
+      "epoch": 2.2312669795011115,
+      "grad_norm": 0.45201456818615127,
+      "learning_rate": 4.867419733623372e-06,
+      "loss": 0.494,
+      "step": 5690
+    },
+    {
+      "epoch": 2.2316621387997038,
+      "grad_norm": 0.4669002434804088,
+      "learning_rate": 4.86736938575208e-06,
+      "loss": 0.5061,
+      "step": 5691
+    },
+    {
+      "epoch": 2.232057298098296,
+      "grad_norm": 0.5274175443355472,
+      "learning_rate": 4.867319028583199e-06,
+      "loss": 0.4944,
+      "step": 5692
+    },
+    {
+      "epoch": 2.2324524573968882,
+      "grad_norm": 0.45829216977076004,
+      "learning_rate": 4.867268662116926e-06,
+      "loss": 0.5064,
+      "step": 5693
+    },
+    {
+      "epoch": 2.2328476166954805,
+      "grad_norm": 0.45758070298208187,
+      "learning_rate": 4.86721828635346e-06,
+      "loss": 0.49,
+      "step": 5694
+    },
+    {
+      "epoch": 2.2332427759940727,
+      "grad_norm": 0.47928993400834125,
+      "learning_rate": 4.867167901292997e-06,
+      "loss": 0.5075,
+      "step": 5695
+    },
+    {
+      "epoch": 2.233637935292665,
+      "grad_norm": 0.46840348523008235,
+      "learning_rate": 4.867117506935737e-06,
+      "loss": 0.517,
+      "step": 5696
+    },
+    {
+      "epoch": 2.234033094591257,
+      "grad_norm": 0.48509443499209903,
+      "learning_rate": 4.867067103281876e-06,
+      "loss": 0.5159,
+      "step": 5697
+    },
+    {
+      "epoch": 2.2344282538898494,
+      "grad_norm": 0.5618447420828913,
+      "learning_rate": 4.867016690331613e-06,
+      "loss": 0.498,
+      "step": 5698
+    },
+    {
+      "epoch": 2.2348234131884417,
+      "grad_norm": 0.46216383707180914,
+      "learning_rate": 4.866966268085146e-06,
+      "loss": 0.5139,
+      "step": 5699
+    },
+    {
+      "epoch": 2.235218572487034,
+      "grad_norm": 0.45344618660738734,
+      "learning_rate": 4.866915836542672e-06,
+      "loss": 0.4922,
+      "step": 5700
+    },
+    {
+      "epoch": 2.235613731785626,
+      "grad_norm": 0.4679880087576633,
+      "learning_rate": 4.866865395704391e-06,
+      "loss": 0.4877,
+      "step": 5701
+    },
+    {
+      "epoch": 2.2360088910842184,
+      "grad_norm": 0.4683019810711741,
+      "learning_rate": 4.8668149455705e-06,
+      "loss": 0.4969,
+      "step": 5702
+    },
+    {
+      "epoch": 2.2364040503828106,
+      "grad_norm": 0.44030516087498117,
+      "learning_rate": 4.866764486141195e-06,
+      "loss": 0.5104,
+      "step": 5703
+    },
+    {
+      "epoch": 2.236799209681403,
+      "grad_norm": 0.4568263053340173,
+      "learning_rate": 4.866714017416678e-06,
+      "loss": 0.4835,
+      "step": 5704
+    },
+    {
+      "epoch": 2.237194368979995,
+      "grad_norm": 0.4855558978372768,
+      "learning_rate": 4.866663539397145e-06,
+      "loss": 0.5128,
+      "step": 5705
+    },
+    {
+      "epoch": 2.2375895282785874,
+      "grad_norm": 0.4698552204107376,
+      "learning_rate": 4.866613052082795e-06,
+      "loss": 0.4989,
+      "step": 5706
+    },
+    {
+      "epoch": 2.2379846875771796,
+      "grad_norm": 0.4573675129961296,
+      "learning_rate": 4.866562555473826e-06,
+      "loss": 0.5163,
+      "step": 5707
+    },
+    {
+      "epoch": 2.238379846875772,
+      "grad_norm": 0.4951264964599398,
+      "learning_rate": 4.866512049570437e-06,
+      "loss": 0.5161,
+      "step": 5708
+    },
+    {
+      "epoch": 2.238775006174364,
+      "grad_norm": 0.4687182481039132,
+      "learning_rate": 4.866461534372825e-06,
+      "loss": 0.5193,
+      "step": 5709
+    },
+    {
+      "epoch": 2.2391701654729563,
+      "grad_norm": 0.4524062924624281,
+      "learning_rate": 4.866411009881189e-06,
+      "loss": 0.5038,
+      "step": 5710
+    },
+    {
+      "epoch": 2.2395653247715486,
+      "grad_norm": 0.44991629928323695,
+      "learning_rate": 4.866360476095727e-06,
+      "loss": 0.4756,
+      "step": 5711
+    },
+    {
+      "epoch": 2.239960484070141,
+      "grad_norm": 0.45306535987909613,
+      "learning_rate": 4.866309933016639e-06,
+      "loss": 0.5075,
+      "step": 5712
+    },
+    {
+      "epoch": 2.240355643368733,
+      "grad_norm": 0.4582923597497876,
+      "learning_rate": 4.866259380644122e-06,
+      "loss": 0.4876,
+      "step": 5713
+    },
+    {
+      "epoch": 2.2407508026673253,
+      "grad_norm": 0.4756588236793085,
+      "learning_rate": 4.866208818978375e-06,
+      "loss": 0.5107,
+      "step": 5714
+    },
+    {
+      "epoch": 2.2411459619659175,
+      "grad_norm": 0.4463905269080376,
+      "learning_rate": 4.866158248019597e-06,
+      "loss": 0.4959,
+      "step": 5715
+    },
+    {
+      "epoch": 2.2415411212645098,
+      "grad_norm": 0.46283025201138467,
+      "learning_rate": 4.866107667767986e-06,
+      "loss": 0.5118,
+      "step": 5716
+    },
+    {
+      "epoch": 2.241936280563102,
+      "grad_norm": 0.45145005862470505,
+      "learning_rate": 4.866057078223741e-06,
+      "loss": 0.5036,
+      "step": 5717
+    },
+    {
+      "epoch": 2.2423314398616943,
+      "grad_norm": 0.4496155378592595,
+      "learning_rate": 4.86600647938706e-06,
+      "loss": 0.4873,
+      "step": 5718
+    },
+    {
+      "epoch": 2.2427265991602865,
+      "grad_norm": 0.46541226899418187,
+      "learning_rate": 4.865955871258142e-06,
+      "loss": 0.4932,
+      "step": 5719
+    },
+    {
+      "epoch": 2.2431217584588787,
+      "grad_norm": 0.46041587527729605,
+      "learning_rate": 4.865905253837187e-06,
+      "loss": 0.4907,
+      "step": 5720
+    },
+    {
+      "epoch": 2.243516917757471,
+      "grad_norm": 0.45128087864299693,
+      "learning_rate": 4.865854627124392e-06,
+      "loss": 0.5083,
+      "step": 5721
+    },
+    {
+      "epoch": 2.243912077056063,
+      "grad_norm": 0.4597446068090033,
+      "learning_rate": 4.8658039911199575e-06,
+      "loss": 0.4876,
+      "step": 5722
+    },
+    {
+      "epoch": 2.2443072363546555,
+      "grad_norm": 0.4531642673031476,
+      "learning_rate": 4.8657533458240814e-06,
+      "loss": 0.5167,
+      "step": 5723
+    },
+    {
+      "epoch": 2.2447023956532477,
+      "grad_norm": 0.4436869230047204,
+      "learning_rate": 4.865702691236962e-06,
+      "loss": 0.4931,
+      "step": 5724
+    },
+    {
+      "epoch": 2.24509755495184,
+      "grad_norm": 0.4651798857195074,
+      "learning_rate": 4.865652027358799e-06,
+      "loss": 0.5097,
+      "step": 5725
+    },
+    {
+      "epoch": 2.245492714250432,
+      "grad_norm": 0.45772796873736676,
+      "learning_rate": 4.865601354189791e-06,
+      "loss": 0.4916,
+      "step": 5726
+    },
+    {
+      "epoch": 2.2458878735490244,
+      "grad_norm": 0.46051001258089186,
+      "learning_rate": 4.865550671730139e-06,
+      "loss": 0.505,
+      "step": 5727
+    },
+    {
+      "epoch": 2.2462830328476167,
+      "grad_norm": 0.46185516258558523,
+      "learning_rate": 4.8654999799800394e-06,
+      "loss": 0.489,
+      "step": 5728
+    },
+    {
+      "epoch": 2.246678192146209,
+      "grad_norm": 0.47864882828994293,
+      "learning_rate": 4.865449278939693e-06,
+      "loss": 0.528,
+      "step": 5729
+    },
+    {
+      "epoch": 2.247073351444801,
+      "grad_norm": 0.49285960618658853,
+      "learning_rate": 4.865398568609297e-06,
+      "loss": 0.5121,
+      "step": 5730
+    },
+    {
+      "epoch": 2.2474685107433934,
+      "grad_norm": 0.470443068321539,
+      "learning_rate": 4.865347848989052e-06,
+      "loss": 0.5095,
+      "step": 5731
+    },
+    {
+      "epoch": 2.2478636700419856,
+      "grad_norm": 0.4560396366626315,
+      "learning_rate": 4.865297120079157e-06,
+      "loss": 0.4911,
+      "step": 5732
+    },
+    {
+      "epoch": 2.248258829340578,
+      "grad_norm": 0.4510249709001742,
+      "learning_rate": 4.8652463818798115e-06,
+      "loss": 0.491,
+      "step": 5733
+    },
+    {
+      "epoch": 2.24865398863917,
+      "grad_norm": 0.4646681611937035,
+      "learning_rate": 4.8651956343912145e-06,
+      "loss": 0.4974,
+      "step": 5734
+    },
+    {
+      "epoch": 2.2490491479377623,
+      "grad_norm": 0.45710106589102445,
+      "learning_rate": 4.8651448776135655e-06,
+      "loss": 0.4957,
+      "step": 5735
+    },
+    {
+      "epoch": 2.2494443072363546,
+      "grad_norm": 0.4432218553818469,
+      "learning_rate": 4.8650941115470636e-06,
+      "loss": 0.5001,
+      "step": 5736
+    },
+    {
+      "epoch": 2.249839466534947,
+      "grad_norm": 0.4461130216348829,
+      "learning_rate": 4.865043336191908e-06,
+      "loss": 0.4993,
+      "step": 5737
+    },
+    {
+      "epoch": 2.250234625833539,
+      "grad_norm": 0.4765715644527521,
+      "learning_rate": 4.864992551548298e-06,
+      "loss": 0.4984,
+      "step": 5738
+    },
+    {
+      "epoch": 2.2506297851321313,
+      "grad_norm": 0.44134482235233946,
+      "learning_rate": 4.864941757616434e-06,
+      "loss": 0.4966,
+      "step": 5739
+    },
+    {
+      "epoch": 2.2510249444307235,
+      "grad_norm": 0.4565907937298236,
+      "learning_rate": 4.864890954396514e-06,
+      "loss": 0.5124,
+      "step": 5740
+    },
+    {
+      "epoch": 2.251420103729316,
+      "grad_norm": 0.4445014385916519,
+      "learning_rate": 4.8648401418887385e-06,
+      "loss": 0.5036,
+      "step": 5741
+    },
+    {
+      "epoch": 2.251815263027908,
+      "grad_norm": 0.5046116665567862,
+      "learning_rate": 4.864789320093307e-06,
+      "loss": 0.4885,
+      "step": 5742
+    },
+    {
+      "epoch": 2.2522104223265003,
+      "grad_norm": 0.45743791687221036,
+      "learning_rate": 4.86473848901042e-06,
+      "loss": 0.5157,
+      "step": 5743
+    },
+    {
+      "epoch": 2.2526055816250925,
+      "grad_norm": 0.45713971041275187,
+      "learning_rate": 4.864687648640275e-06,
+      "loss": 0.5209,
+      "step": 5744
+    },
+    {
+      "epoch": 2.2530007409236847,
+      "grad_norm": 0.45916966205747045,
+      "learning_rate": 4.864636798983073e-06,
+      "loss": 0.5218,
+      "step": 5745
+    },
+    {
+      "epoch": 2.253395900222277,
+      "grad_norm": 0.4645094613852396,
+      "learning_rate": 4.864585940039014e-06,
+      "loss": 0.4919,
+      "step": 5746
+    },
+    {
+      "epoch": 2.2537910595208692,
+      "grad_norm": 0.4611192692789078,
+      "learning_rate": 4.864535071808298e-06,
+      "loss": 0.5181,
+      "step": 5747
+    },
+    {
+      "epoch": 2.2541862188194615,
+      "grad_norm": 0.462528547985611,
+      "learning_rate": 4.8644841942911225e-06,
+      "loss": 0.5048,
+      "step": 5748
+    },
+    {
+      "epoch": 2.2545813781180537,
+      "grad_norm": 0.4622175816401159,
+      "learning_rate": 4.8644333074876896e-06,
+      "loss": 0.5138,
+      "step": 5749
+    },
+    {
+      "epoch": 2.254976537416646,
+      "grad_norm": 0.4802227886602715,
+      "learning_rate": 4.864382411398198e-06,
+      "loss": 0.4925,
+      "step": 5750
+    },
+    {
+      "epoch": 2.255371696715238,
+      "grad_norm": 0.4542300433876316,
+      "learning_rate": 4.864331506022848e-06,
+      "loss": 0.5204,
+      "step": 5751
+    },
+    {
+      "epoch": 2.2557668560138304,
+      "grad_norm": 0.46121165508412626,
+      "learning_rate": 4.86428059136184e-06,
+      "loss": 0.489,
+      "step": 5752
+    },
+    {
+      "epoch": 2.2561620153124227,
+      "grad_norm": 0.45065089171005923,
+      "learning_rate": 4.864229667415373e-06,
+      "loss": 0.5082,
+      "step": 5753
+    },
+    {
+      "epoch": 2.256557174611015,
+      "grad_norm": 0.4664108459131774,
+      "learning_rate": 4.864178734183649e-06,
+      "loss": 0.5203,
+      "step": 5754
+    },
+    {
+      "epoch": 2.256952333909607,
+      "grad_norm": 0.46599726286903415,
+      "learning_rate": 4.864127791666865e-06,
+      "loss": 0.5044,
+      "step": 5755
+    },
+    {
+      "epoch": 2.2573474932081994,
+      "grad_norm": 0.4579348177710053,
+      "learning_rate": 4.864076839865223e-06,
+      "loss": 0.4966,
+      "step": 5756
+    },
+    {
+      "epoch": 2.2577426525067916,
+      "grad_norm": 0.4710610355784568,
+      "learning_rate": 4.864025878778923e-06,
+      "loss": 0.5039,
+      "step": 5757
+    },
+    {
+      "epoch": 2.258137811805384,
+      "grad_norm": 0.5098836289434676,
+      "learning_rate": 4.863974908408164e-06,
+      "loss": 0.4864,
+      "step": 5758
+    },
+    {
+      "epoch": 2.258532971103976,
+      "grad_norm": 0.444685462826527,
+      "learning_rate": 4.863923928753148e-06,
+      "loss": 0.4929,
+      "step": 5759
+    },
+    {
+      "epoch": 2.2589281304025683,
+      "grad_norm": 0.45233517234419085,
+      "learning_rate": 4.8638729398140735e-06,
+      "loss": 0.5061,
+      "step": 5760
+    },
+    {
+      "epoch": 2.2593232897011606,
+      "grad_norm": 0.465371345730281,
+      "learning_rate": 4.863821941591142e-06,
+      "loss": 0.4998,
+      "step": 5761
+    },
+    {
+      "epoch": 2.2597184489997533,
+      "grad_norm": 0.4562128947163187,
+      "learning_rate": 4.863770934084553e-06,
+      "loss": 0.5238,
+      "step": 5762
+    },
+    {
+      "epoch": 2.2601136082983455,
+      "grad_norm": 0.45638820275354225,
+      "learning_rate": 4.863719917294507e-06,
+      "loss": 0.5035,
+      "step": 5763
+    },
+    {
+      "epoch": 2.2605087675969378,
+      "grad_norm": 0.47757782000563964,
+      "learning_rate": 4.863668891221206e-06,
+      "loss": 0.5079,
+      "step": 5764
+    },
+    {
+      "epoch": 2.26090392689553,
+      "grad_norm": 0.4662080915894657,
+      "learning_rate": 4.863617855864847e-06,
+      "loss": 0.5124,
+      "step": 5765
+    },
+    {
+      "epoch": 2.2612990861941222,
+      "grad_norm": 0.47892291257771435,
+      "learning_rate": 4.863566811225634e-06,
+      "loss": 0.5184,
+      "step": 5766
+    },
+    {
+      "epoch": 2.2616942454927145,
+      "grad_norm": 0.45977277292864505,
+      "learning_rate": 4.863515757303764e-06,
+      "loss": 0.4941,
+      "step": 5767
+    },
+    {
+      "epoch": 2.2620894047913067,
+      "grad_norm": 0.46916509655903266,
+      "learning_rate": 4.863464694099441e-06,
+      "loss": 0.5056,
+      "step": 5768
+    },
+    {
+      "epoch": 2.262484564089899,
+      "grad_norm": 0.4504962671579038,
+      "learning_rate": 4.863413621612862e-06,
+      "loss": 0.5062,
+      "step": 5769
+    },
+    {
+      "epoch": 2.262879723388491,
+      "grad_norm": 0.44867824093263253,
+      "learning_rate": 4.863362539844231e-06,
+      "loss": 0.4993,
+      "step": 5770
+    },
+    {
+      "epoch": 2.2632748826870834,
+      "grad_norm": 0.4670694822124429,
+      "learning_rate": 4.863311448793747e-06,
+      "loss": 0.5106,
+      "step": 5771
+    },
+    {
+      "epoch": 2.2636700419856757,
+      "grad_norm": 0.4632432969652908,
+      "learning_rate": 4.8632603484616095e-06,
+      "loss": 0.5138,
+      "step": 5772
+    },
+    {
+      "epoch": 2.264065201284268,
+      "grad_norm": 0.5486722092316492,
+      "learning_rate": 4.8632092388480216e-06,
+      "loss": 0.4994,
+      "step": 5773
+    },
+    {
+      "epoch": 2.26446036058286,
+      "grad_norm": 0.46134894866499143,
+      "learning_rate": 4.863158119953182e-06,
+      "loss": 0.5041,
+      "step": 5774
+    },
+    {
+      "epoch": 2.2648555198814524,
+      "grad_norm": 0.4645211586085868,
+      "learning_rate": 4.863106991777293e-06,
+      "loss": 0.5062,
+      "step": 5775
+    },
+    {
+      "epoch": 2.2652506791800446,
+      "grad_norm": 0.46537662412865477,
+      "learning_rate": 4.863055854320554e-06,
+      "loss": 0.4913,
+      "step": 5776
+    },
+    {
+      "epoch": 2.265645838478637,
+      "grad_norm": 0.4463910826863991,
+      "learning_rate": 4.863004707583167e-06,
+      "loss": 0.5051,
+      "step": 5777
+    },
+    {
+      "epoch": 2.266040997777229,
+      "grad_norm": 0.4461844055881152,
+      "learning_rate": 4.862953551565332e-06,
+      "loss": 0.4879,
+      "step": 5778
+    },
+    {
+      "epoch": 2.2664361570758214,
+      "grad_norm": 0.4671496378324098,
+      "learning_rate": 4.862902386267251e-06,
+      "loss": 0.4962,
+      "step": 5779
+    },
+    {
+      "epoch": 2.2668313163744136,
+      "grad_norm": 0.47268040765077685,
+      "learning_rate": 4.862851211689124e-06,
+      "loss": 0.5075,
+      "step": 5780
+    },
+    {
+      "epoch": 2.267226475673006,
+      "grad_norm": 0.4482777628793822,
+      "learning_rate": 4.8628000278311515e-06,
+      "loss": 0.5135,
+      "step": 5781
+    },
+    {
+      "epoch": 2.267621634971598,
+      "grad_norm": 0.4589821173101706,
+      "learning_rate": 4.862748834693536e-06,
+      "loss": 0.4909,
+      "step": 5782
+    },
+    {
+      "epoch": 2.2680167942701903,
+      "grad_norm": 0.4670920261314897,
+      "learning_rate": 4.862697632276477e-06,
+      "loss": 0.5232,
+      "step": 5783
+    },
+    {
+      "epoch": 2.2684119535687826,
+      "grad_norm": 0.4421128362683695,
+      "learning_rate": 4.862646420580178e-06,
+      "loss": 0.4887,
+      "step": 5784
+    },
+    {
+      "epoch": 2.268807112867375,
+      "grad_norm": 0.44505858305719176,
+      "learning_rate": 4.862595199604837e-06,
+      "loss": 0.5003,
+      "step": 5785
+    },
+    {
+      "epoch": 2.269202272165967,
+      "grad_norm": 0.49135116256536704,
+      "learning_rate": 4.862543969350657e-06,
+      "loss": 0.4949,
+      "step": 5786
+    },
+    {
+      "epoch": 2.2695974314645593,
+      "grad_norm": 0.46918700677450953,
+      "learning_rate": 4.86249272981784e-06,
+      "loss": 0.4885,
+      "step": 5787
+    },
+    {
+      "epoch": 2.2699925907631515,
+      "grad_norm": 0.4485662070030722,
+      "learning_rate": 4.862441481006586e-06,
+      "loss": 0.5049,
+      "step": 5788
+    },
+    {
+      "epoch": 2.2703877500617438,
+      "grad_norm": 0.43803428307843834,
+      "learning_rate": 4.862390222917095e-06,
+      "loss": 0.4808,
+      "step": 5789
+    },
+    {
+      "epoch": 2.270782909360336,
+      "grad_norm": 0.46807207385887406,
+      "learning_rate": 4.86233895554957e-06,
+      "loss": 0.5187,
+      "step": 5790
+    },
+    {
+      "epoch": 2.2711780686589282,
+      "grad_norm": 0.44793935494961473,
+      "learning_rate": 4.862287678904213e-06,
+      "loss": 0.4933,
+      "step": 5791
+    },
+    {
+      "epoch": 2.2715732279575205,
+      "grad_norm": 0.46535343446328026,
+      "learning_rate": 4.862236392981225e-06,
+      "loss": 0.5276,
+      "step": 5792
+    },
+    {
+      "epoch": 2.2719683872561127,
+      "grad_norm": 0.4698550219155634,
+      "learning_rate": 4.8621850977808046e-06,
+      "loss": 0.4872,
+      "step": 5793
+    },
+    {
+      "epoch": 2.272363546554705,
+      "grad_norm": 0.4682879416283325,
+      "learning_rate": 4.862133793303157e-06,
+      "loss": 0.4903,
+      "step": 5794
+    },
+    {
+      "epoch": 2.272758705853297,
+      "grad_norm": 0.48270114542131143,
+      "learning_rate": 4.862082479548482e-06,
+      "loss": 0.4993,
+      "step": 5795
+    },
+    {
+      "epoch": 2.2731538651518894,
+      "grad_norm": 0.5493323187094451,
+      "learning_rate": 4.862031156516982e-06,
+      "loss": 0.4965,
+      "step": 5796
+    },
+    {
+      "epoch": 2.2735490244504817,
+      "grad_norm": 0.4516215367368103,
+      "learning_rate": 4.861979824208857e-06,
+      "loss": 0.5034,
+      "step": 5797
+    },
+    {
+      "epoch": 2.273944183749074,
+      "grad_norm": 0.4538027339697062,
+      "learning_rate": 4.86192848262431e-06,
+      "loss": 0.5006,
+      "step": 5798
+    },
+    {
+      "epoch": 2.274339343047666,
+      "grad_norm": 0.4616229995385204,
+      "learning_rate": 4.861877131763542e-06,
+      "loss": 0.5013,
+      "step": 5799
+    },
+    {
+      "epoch": 2.2747345023462584,
+      "grad_norm": 0.46227058646164154,
+      "learning_rate": 4.861825771626755e-06,
+      "loss": 0.4987,
+      "step": 5800
+    },
+    {
+      "epoch": 2.2751296616448506,
+      "grad_norm": 0.45156002524949446,
+      "learning_rate": 4.86177440221415e-06,
+      "loss": 0.4905,
+      "step": 5801
+    },
+    {
+      "epoch": 2.275524820943443,
+      "grad_norm": 0.45398486859479054,
+      "learning_rate": 4.861723023525929e-06,
+      "loss": 0.5127,
+      "step": 5802
+    },
+    {
+      "epoch": 2.275919980242035,
+      "grad_norm": 0.4546600141282232,
+      "learning_rate": 4.861671635562295e-06,
+      "loss": 0.512,
+      "step": 5803
+    },
+    {
+      "epoch": 2.2763151395406274,
+      "grad_norm": 0.4420173118969808,
+      "learning_rate": 4.861620238323449e-06,
+      "loss": 0.508,
+      "step": 5804
+    },
+    {
+      "epoch": 2.2767102988392196,
+      "grad_norm": 0.4481000176106849,
+      "learning_rate": 4.861568831809592e-06,
+      "loss": 0.505,
+      "step": 5805
+    },
+    {
+      "epoch": 2.277105458137812,
+      "grad_norm": 0.48239910462261965,
+      "learning_rate": 4.861517416020928e-06,
+      "loss": 0.5103,
+      "step": 5806
+    },
+    {
+      "epoch": 2.277500617436404,
+      "grad_norm": 0.47340057062183605,
+      "learning_rate": 4.861465990957656e-06,
+      "loss": 0.4927,
+      "step": 5807
+    },
+    {
+      "epoch": 2.2778957767349963,
+      "grad_norm": 0.4552226094608049,
+      "learning_rate": 4.86141455661998e-06,
+      "loss": 0.4982,
+      "step": 5808
+    },
+    {
+      "epoch": 2.2782909360335886,
+      "grad_norm": 0.580540080884017,
+      "learning_rate": 4.861363113008102e-06,
+      "loss": 0.4945,
+      "step": 5809
+    },
+    {
+      "epoch": 2.278686095332181,
+      "grad_norm": 0.4563722531620255,
+      "learning_rate": 4.861311660122223e-06,
+      "loss": 0.491,
+      "step": 5810
+    },
+    {
+      "epoch": 2.279081254630773,
+      "grad_norm": 0.44479481415363753,
+      "learning_rate": 4.861260197962546e-06,
+      "loss": 0.4733,
+      "step": 5811
+    },
+    {
+      "epoch": 2.2794764139293653,
+      "grad_norm": 0.46120540937626275,
+      "learning_rate": 4.861208726529273e-06,
+      "loss": 0.4971,
+      "step": 5812
+    },
+    {
+      "epoch": 2.2798715732279575,
+      "grad_norm": 0.46393191062150557,
+      "learning_rate": 4.861157245822605e-06,
+      "loss": 0.5004,
+      "step": 5813
+    },
+    {
+      "epoch": 2.2802667325265498,
+      "grad_norm": 0.4515647504745636,
+      "learning_rate": 4.861105755842747e-06,
+      "loss": 0.4777,
+      "step": 5814
+    },
+    {
+      "epoch": 2.280661891825142,
+      "grad_norm": 0.438205927734934,
+      "learning_rate": 4.8610542565898975e-06,
+      "loss": 0.4916,
+      "step": 5815
+    },
+    {
+      "epoch": 2.2810570511237342,
+      "grad_norm": 0.457098274611986,
+      "learning_rate": 4.861002748064261e-06,
+      "loss": 0.5009,
+      "step": 5816
+    },
+    {
+      "epoch": 2.2814522104223265,
+      "grad_norm": 0.4615257787626016,
+      "learning_rate": 4.86095123026604e-06,
+      "loss": 0.4986,
+      "step": 5817
+    },
+    {
+      "epoch": 2.2818473697209187,
+      "grad_norm": 0.45965059976307776,
+      "learning_rate": 4.860899703195435e-06,
+      "loss": 0.4986,
+      "step": 5818
+    },
+    {
+      "epoch": 2.282242529019511,
+      "grad_norm": 0.46417580461089847,
+      "learning_rate": 4.860848166852651e-06,
+      "loss": 0.4937,
+      "step": 5819
+    },
+    {
+      "epoch": 2.282637688318103,
+      "grad_norm": 0.47161773166956317,
+      "learning_rate": 4.860796621237888e-06,
+      "loss": 0.4979,
+      "step": 5820
+    },
+    {
+      "epoch": 2.2830328476166954,
+      "grad_norm": 0.4660616983606871,
+      "learning_rate": 4.86074506635135e-06,
+      "loss": 0.509,
+      "step": 5821
+    },
+    {
+      "epoch": 2.2834280069152877,
+      "grad_norm": 0.4832205275826543,
+      "learning_rate": 4.860693502193239e-06,
+      "loss": 0.516,
+      "step": 5822
+    },
+    {
+      "epoch": 2.28382316621388,
+      "grad_norm": 0.49641441825377863,
+      "learning_rate": 4.860641928763757e-06,
+      "loss": 0.5138,
+      "step": 5823
+    },
+    {
+      "epoch": 2.284218325512472,
+      "grad_norm": 0.4591396621913851,
+      "learning_rate": 4.860590346063107e-06,
+      "loss": 0.4956,
+      "step": 5824
+    },
+    {
+      "epoch": 2.2846134848110644,
+      "grad_norm": 0.46808791796018534,
+      "learning_rate": 4.8605387540914915e-06,
+      "loss": 0.4979,
+      "step": 5825
+    },
+    {
+      "epoch": 2.2850086441096567,
+      "grad_norm": 0.44616669574023354,
+      "learning_rate": 4.8604871528491135e-06,
+      "loss": 0.4809,
+      "step": 5826
+    },
+    {
+      "epoch": 2.285403803408249,
+      "grad_norm": 0.46293514962772786,
+      "learning_rate": 4.860435542336175e-06,
+      "loss": 0.5058,
+      "step": 5827
+    },
+    {
+      "epoch": 2.285798962706841,
+      "grad_norm": 0.47952265408731376,
+      "learning_rate": 4.86038392255288e-06,
+      "loss": 0.4998,
+      "step": 5828
+    },
+    {
+      "epoch": 2.2861941220054334,
+      "grad_norm": 0.4585566892611743,
+      "learning_rate": 4.8603322934994284e-06,
+      "loss": 0.5039,
+      "step": 5829
+    },
+    {
+      "epoch": 2.2865892813040256,
+      "grad_norm": 0.45394838026461104,
+      "learning_rate": 4.860280655176026e-06,
+      "loss": 0.5084,
+      "step": 5830
+    },
+    {
+      "epoch": 2.286984440602618,
+      "grad_norm": 0.4556930344196025,
+      "learning_rate": 4.860229007582874e-06,
+      "loss": 0.512,
+      "step": 5831
+    },
+    {
+      "epoch": 2.28737959990121,
+      "grad_norm": 0.5160699721279627,
+      "learning_rate": 4.860177350720176e-06,
+      "loss": 0.4992,
+      "step": 5832
+    },
+    {
+      "epoch": 2.2877747591998023,
+      "grad_norm": 0.4683352173737334,
+      "learning_rate": 4.860125684588135e-06,
+      "loss": 0.4984,
+      "step": 5833
+    },
+    {
+      "epoch": 2.2881699184983946,
+      "grad_norm": 0.44948200378048164,
+      "learning_rate": 4.860074009186952e-06,
+      "loss": 0.5026,
+      "step": 5834
+    },
+    {
+      "epoch": 2.288565077796987,
+      "grad_norm": 0.4462063995104546,
+      "learning_rate": 4.8600223245168325e-06,
+      "loss": 0.4989,
+      "step": 5835
+    },
+    {
+      "epoch": 2.288960237095579,
+      "grad_norm": 0.4769414323903259,
+      "learning_rate": 4.8599706305779785e-06,
+      "loss": 0.5109,
+      "step": 5836
+    },
+    {
+      "epoch": 2.2893553963941713,
+      "grad_norm": 0.4838732772951883,
+      "learning_rate": 4.8599189273705926e-06,
+      "loss": 0.4982,
+      "step": 5837
+    },
+    {
+      "epoch": 2.2897505556927635,
+      "grad_norm": 0.47336977014120124,
+      "learning_rate": 4.859867214894878e-06,
+      "loss": 0.501,
+      "step": 5838
+    },
+    {
+      "epoch": 2.2901457149913558,
+      "grad_norm": 0.4708351136561112,
+      "learning_rate": 4.8598154931510385e-06,
+      "loss": 0.5142,
+      "step": 5839
+    },
+    {
+      "epoch": 2.290540874289948,
+      "grad_norm": 0.4335493511834357,
+      "learning_rate": 4.859763762139276e-06,
+      "loss": 0.4897,
+      "step": 5840
+    },
+    {
+      "epoch": 2.2909360335885403,
+      "grad_norm": 0.4551238002159216,
+      "learning_rate": 4.859712021859795e-06,
+      "loss": 0.506,
+      "step": 5841
+    },
+    {
+      "epoch": 2.2913311928871325,
+      "grad_norm": 0.46419964327160607,
+      "learning_rate": 4.8596602723127975e-06,
+      "loss": 0.4924,
+      "step": 5842
+    },
+    {
+      "epoch": 2.2917263521857247,
+      "grad_norm": 0.4504273710533772,
+      "learning_rate": 4.859608513498488e-06,
+      "loss": 0.5141,
+      "step": 5843
+    },
+    {
+      "epoch": 2.292121511484317,
+      "grad_norm": 0.4592991376285661,
+      "learning_rate": 4.859556745417068e-06,
+      "loss": 0.5224,
+      "step": 5844
+    },
+    {
+      "epoch": 2.292516670782909,
+      "grad_norm": 0.45863529913627665,
+      "learning_rate": 4.859504968068743e-06,
+      "loss": 0.5019,
+      "step": 5845
+    },
+    {
+      "epoch": 2.2929118300815015,
+      "grad_norm": 0.46302406671745044,
+      "learning_rate": 4.859453181453715e-06,
+      "loss": 0.5123,
+      "step": 5846
+    },
+    {
+      "epoch": 2.2933069893800937,
+      "grad_norm": 0.5210997355300256,
+      "learning_rate": 4.8594013855721875e-06,
+      "loss": 0.5157,
+      "step": 5847
+    },
+    {
+      "epoch": 2.293702148678686,
+      "grad_norm": 0.4698909559181017,
+      "learning_rate": 4.859349580424364e-06,
+      "loss": 0.5115,
+      "step": 5848
+    },
+    {
+      "epoch": 2.294097307977278,
+      "grad_norm": 0.4753355992191938,
+      "learning_rate": 4.859297766010448e-06,
+      "loss": 0.508,
+      "step": 5849
+    },
+    {
+      "epoch": 2.2944924672758704,
+      "grad_norm": 0.44406133299641976,
+      "learning_rate": 4.859245942330643e-06,
+      "loss": 0.5031,
+      "step": 5850
+    },
+    {
+      "epoch": 2.2948876265744627,
+      "grad_norm": 0.46793872675388737,
+      "learning_rate": 4.859194109385152e-06,
+      "loss": 0.5028,
+      "step": 5851
+    },
+    {
+      "epoch": 2.295282785873055,
+      "grad_norm": 0.456450080256325,
+      "learning_rate": 4.85914226717418e-06,
+      "loss": 0.4996,
+      "step": 5852
+    },
+    {
+      "epoch": 2.295677945171647,
+      "grad_norm": 0.4678454186698638,
+      "learning_rate": 4.85909041569793e-06,
+      "loss": 0.4944,
+      "step": 5853
+    },
+    {
+      "epoch": 2.2960731044702394,
+      "grad_norm": 0.4558794558467581,
+      "learning_rate": 4.8590385549566046e-06,
+      "loss": 0.5009,
+      "step": 5854
+    },
+    {
+      "epoch": 2.2964682637688316,
+      "grad_norm": 0.45402991768060114,
+      "learning_rate": 4.858986684950408e-06,
+      "loss": 0.5311,
+      "step": 5855
+    },
+    {
+      "epoch": 2.296863423067424,
+      "grad_norm": 0.46642376112323874,
+      "learning_rate": 4.858934805679545e-06,
+      "loss": 0.4982,
+      "step": 5856
+    },
+    {
+      "epoch": 2.297258582366016,
+      "grad_norm": 0.4582563144966381,
+      "learning_rate": 4.858882917144218e-06,
+      "loss": 0.4966,
+      "step": 5857
+    },
+    {
+      "epoch": 2.2976537416646083,
+      "grad_norm": 0.44640576325625786,
+      "learning_rate": 4.858831019344632e-06,
+      "loss": 0.5014,
+      "step": 5858
+    },
+    {
+      "epoch": 2.2980489009632006,
+      "grad_norm": 0.45330919927536356,
+      "learning_rate": 4.858779112280989e-06,
+      "loss": 0.4981,
+      "step": 5859
+    },
+    {
+      "epoch": 2.298444060261793,
+      "grad_norm": 0.4575939467608079,
+      "learning_rate": 4.858727195953495e-06,
+      "loss": 0.5004,
+      "step": 5860
+    },
+    {
+      "epoch": 2.298839219560385,
+      "grad_norm": 0.45894810163579486,
+      "learning_rate": 4.858675270362352e-06,
+      "loss": 0.5005,
+      "step": 5861
+    },
+    {
+      "epoch": 2.2992343788589773,
+      "grad_norm": 0.4667448223237067,
+      "learning_rate": 4.858623335507765e-06,
+      "loss": 0.5131,
+      "step": 5862
+    },
+    {
+      "epoch": 2.2996295381575695,
+      "grad_norm": 0.4466389707322938,
+      "learning_rate": 4.858571391389938e-06,
+      "loss": 0.5065,
+      "step": 5863
+    },
+    {
+      "epoch": 2.300024697456162,
+      "grad_norm": 0.46052845698144745,
+      "learning_rate": 4.858519438009075e-06,
+      "loss": 0.4981,
+      "step": 5864
+    },
+    {
+      "epoch": 2.300419856754754,
+      "grad_norm": 0.4521693219071108,
+      "learning_rate": 4.8584674753653795e-06,
+      "loss": 0.4932,
+      "step": 5865
+    },
+    {
+      "epoch": 2.3008150160533463,
+      "grad_norm": 0.4747140612853695,
+      "learning_rate": 4.858415503459056e-06,
+      "loss": 0.5212,
+      "step": 5866
+    },
+    {
+      "epoch": 2.3012101753519385,
+      "grad_norm": 0.45520487283506444,
+      "learning_rate": 4.858363522290308e-06,
+      "loss": 0.5063,
+      "step": 5867
+    },
+    {
+      "epoch": 2.301605334650531,
+      "grad_norm": 0.4551319537165788,
+      "learning_rate": 4.858311531859341e-06,
+      "loss": 0.488,
+      "step": 5868
+    },
+    {
+      "epoch": 2.3020004939491234,
+      "grad_norm": 0.4794988475592662,
+      "learning_rate": 4.858259532166358e-06,
+      "loss": 0.5088,
+      "step": 5869
+    },
+    {
+      "epoch": 2.3023956532477157,
+      "grad_norm": 0.4581559846294469,
+      "learning_rate": 4.858207523211563e-06,
+      "loss": 0.5049,
+      "step": 5870
+    },
+    {
+      "epoch": 2.302790812546308,
+      "grad_norm": 0.4616610380469891,
+      "learning_rate": 4.858155504995162e-06,
+      "loss": 0.4891,
+      "step": 5871
+    },
+    {
+      "epoch": 2.3031859718449,
+      "grad_norm": 0.46422859211150336,
+      "learning_rate": 4.8581034775173575e-06,
+      "loss": 0.512,
+      "step": 5872
+    },
+    {
+      "epoch": 2.3035811311434924,
+      "grad_norm": 0.4693415161744398,
+      "learning_rate": 4.858051440778354e-06,
+      "loss": 0.5236,
+      "step": 5873
+    },
+    {
+      "epoch": 2.3039762904420846,
+      "grad_norm": 0.45548329510092883,
+      "learning_rate": 4.857999394778357e-06,
+      "loss": 0.5045,
+      "step": 5874
+    },
+    {
+      "epoch": 2.304371449740677,
+      "grad_norm": 0.4568687640660631,
+      "learning_rate": 4.857947339517571e-06,
+      "loss": 0.5055,
+      "step": 5875
+    },
+    {
+      "epoch": 2.304766609039269,
+      "grad_norm": 0.46035875823989625,
+      "learning_rate": 4.857895274996198e-06,
+      "loss": 0.4936,
+      "step": 5876
+    },
+    {
+      "epoch": 2.3051617683378613,
+      "grad_norm": 0.4852045992439535,
+      "learning_rate": 4.857843201214445e-06,
+      "loss": 0.515,
+      "step": 5877
+    },
+    {
+      "epoch": 2.3055569276364536,
+      "grad_norm": 0.4595559022314449,
+      "learning_rate": 4.857791118172515e-06,
+      "loss": 0.4981,
+      "step": 5878
+    },
+    {
+      "epoch": 2.305952086935046,
+      "grad_norm": 0.4647597601816075,
+      "learning_rate": 4.857739025870614e-06,
+      "loss": 0.4914,
+      "step": 5879
+    },
+    {
+      "epoch": 2.306347246233638,
+      "grad_norm": 0.4745172978982465,
+      "learning_rate": 4.857686924308946e-06,
+      "loss": 0.4986,
+      "step": 5880
+    },
+    {
+      "epoch": 2.3067424055322303,
+      "grad_norm": 0.4738840639632119,
+      "learning_rate": 4.857634813487715e-06,
+      "loss": 0.5071,
+      "step": 5881
+    },
+    {
+      "epoch": 2.3071375648308226,
+      "grad_norm": 0.4528727368292258,
+      "learning_rate": 4.857582693407126e-06,
+      "loss": 0.4974,
+      "step": 5882
+    },
+    {
+      "epoch": 2.307532724129415,
+      "grad_norm": 0.46653799947338975,
+      "learning_rate": 4.857530564067383e-06,
+      "loss": 0.4883,
+      "step": 5883
+    },
+    {
+      "epoch": 2.307927883428007,
+      "grad_norm": 0.46309044360177304,
+      "learning_rate": 4.857478425468693e-06,
+      "loss": 0.4971,
+      "step": 5884
+    },
+    {
+      "epoch": 2.3083230427265993,
+      "grad_norm": 0.46207981907616164,
+      "learning_rate": 4.857426277611258e-06,
+      "loss": 0.512,
+      "step": 5885
+    },
+    {
+      "epoch": 2.3087182020251915,
+      "grad_norm": 0.46314752976399587,
+      "learning_rate": 4.857374120495285e-06,
+      "loss": 0.5063,
+      "step": 5886
+    },
+    {
+      "epoch": 2.3091133613237838,
+      "grad_norm": 0.4649842929739362,
+      "learning_rate": 4.857321954120977e-06,
+      "loss": 0.5114,
+      "step": 5887
+    },
+    {
+      "epoch": 2.309508520622376,
+      "grad_norm": 0.5535102504683219,
+      "learning_rate": 4.857269778488541e-06,
+      "loss": 0.5134,
+      "step": 5888
+    },
+    {
+      "epoch": 2.3099036799209682,
+      "grad_norm": 0.4423259127513932,
+      "learning_rate": 4.85721759359818e-06,
+      "loss": 0.4975,
+      "step": 5889
+    },
+    {
+      "epoch": 2.3102988392195605,
+      "grad_norm": 0.44102253287067894,
+      "learning_rate": 4.8571653994501e-06,
+      "loss": 0.487,
+      "step": 5890
+    },
+    {
+      "epoch": 2.3106939985181527,
+      "grad_norm": 0.4528723880487849,
+      "learning_rate": 4.857113196044505e-06,
+      "loss": 0.5174,
+      "step": 5891
+    },
+    {
+      "epoch": 2.311089157816745,
+      "grad_norm": 0.4594030254108174,
+      "learning_rate": 4.857060983381601e-06,
+      "loss": 0.5041,
+      "step": 5892
+    },
+    {
+      "epoch": 2.311484317115337,
+      "grad_norm": 0.4688774998319715,
+      "learning_rate": 4.857008761461593e-06,
+      "loss": 0.5058,
+      "step": 5893
+    },
+    {
+      "epoch": 2.3118794764139294,
+      "grad_norm": 0.4649598927489743,
+      "learning_rate": 4.856956530284686e-06,
+      "loss": 0.5022,
+      "step": 5894
+    },
+    {
+      "epoch": 2.3122746357125217,
+      "grad_norm": 0.4498229566988718,
+      "learning_rate": 4.856904289851084e-06,
+      "loss": 0.5122,
+      "step": 5895
+    },
+    {
+      "epoch": 2.312669795011114,
+      "grad_norm": 0.4683470822663338,
+      "learning_rate": 4.856852040160994e-06,
+      "loss": 0.5005,
+      "step": 5896
+    },
+    {
+      "epoch": 2.313064954309706,
+      "grad_norm": 0.4911558695679091,
+      "learning_rate": 4.856799781214621e-06,
+      "loss": 0.5253,
+      "step": 5897
+    },
+    {
+      "epoch": 2.3134601136082984,
+      "grad_norm": 0.464409082612125,
+      "learning_rate": 4.856747513012168e-06,
+      "loss": 0.5019,
+      "step": 5898
+    },
+    {
+      "epoch": 2.3138552729068906,
+      "grad_norm": 0.4579973840895186,
+      "learning_rate": 4.856695235553843e-06,
+      "loss": 0.4942,
+      "step": 5899
+    },
+    {
+      "epoch": 2.314250432205483,
+      "grad_norm": 0.4478579726762588,
+      "learning_rate": 4.85664294883985e-06,
+      "loss": 0.4892,
+      "step": 5900
+    },
+    {
+      "epoch": 2.314645591504075,
+      "grad_norm": 0.4693973591026146,
+      "learning_rate": 4.856590652870395e-06,
+      "loss": 0.5006,
+      "step": 5901
+    },
+    {
+      "epoch": 2.3150407508026674,
+      "grad_norm": 0.5045358380289714,
+      "learning_rate": 4.856538347645681e-06,
+      "loss": 0.5154,
+      "step": 5902
+    },
+    {
+      "epoch": 2.3154359101012596,
+      "grad_norm": 0.4649328917632037,
+      "learning_rate": 4.856486033165917e-06,
+      "loss": 0.4955,
+      "step": 5903
+    },
+    {
+      "epoch": 2.315831069399852,
+      "grad_norm": 0.46867742048507893,
+      "learning_rate": 4.856433709431307e-06,
+      "loss": 0.5016,
+      "step": 5904
+    },
+    {
+      "epoch": 2.316226228698444,
+      "grad_norm": 0.4520581706582322,
+      "learning_rate": 4.8563813764420555e-06,
+      "loss": 0.5125,
+      "step": 5905
+    },
+    {
+      "epoch": 2.3166213879970363,
+      "grad_norm": 0.45026206385228246,
+      "learning_rate": 4.856329034198368e-06,
+      "loss": 0.4939,
+      "step": 5906
+    },
+    {
+      "epoch": 2.3170165472956286,
+      "grad_norm": 0.45030736202082045,
+      "learning_rate": 4.8562766827004525e-06,
+      "loss": 0.5242,
+      "step": 5907
+    },
+    {
+      "epoch": 2.317411706594221,
+      "grad_norm": 0.4469905948107838,
+      "learning_rate": 4.856224321948512e-06,
+      "loss": 0.5073,
+      "step": 5908
+    },
+    {
+      "epoch": 2.317806865892813,
+      "grad_norm": 0.45894383299198227,
+      "learning_rate": 4.856171951942754e-06,
+      "loss": 0.5101,
+      "step": 5909
+    },
+    {
+      "epoch": 2.3182020251914053,
+      "grad_norm": 0.4395899572240971,
+      "learning_rate": 4.856119572683383e-06,
+      "loss": 0.4837,
+      "step": 5910
+    },
+    {
+      "epoch": 2.3185971844899975,
+      "grad_norm": 0.44751988426963857,
+      "learning_rate": 4.856067184170604e-06,
+      "loss": 0.5126,
+      "step": 5911
+    },
+    {
+      "epoch": 2.3189923437885898,
+      "grad_norm": 0.45656584660916594,
+      "learning_rate": 4.856014786404625e-06,
+      "loss": 0.4979,
+      "step": 5912
+    },
+    {
+      "epoch": 2.319387503087182,
+      "grad_norm": 0.4506209307224017,
+      "learning_rate": 4.8559623793856505e-06,
+      "loss": 0.4942,
+      "step": 5913
+    },
+    {
+      "epoch": 2.3197826623857742,
+      "grad_norm": 0.4461798611971729,
+      "learning_rate": 4.855909963113886e-06,
+      "loss": 0.4802,
+      "step": 5914
+    },
+    {
+      "epoch": 2.3201778216843665,
+      "grad_norm": 0.46130265251689806,
+      "learning_rate": 4.8558575375895375e-06,
+      "loss": 0.5252,
+      "step": 5915
+    },
+    {
+      "epoch": 2.3205729809829587,
+      "grad_norm": 0.45320094317982595,
+      "learning_rate": 4.855805102812811e-06,
+      "loss": 0.5089,
+      "step": 5916
+    },
+    {
+      "epoch": 2.320968140281551,
+      "grad_norm": 0.4508029872376069,
+      "learning_rate": 4.855752658783914e-06,
+      "loss": 0.5247,
+      "step": 5917
+    },
+    {
+      "epoch": 2.321363299580143,
+      "grad_norm": 0.4448389675565296,
+      "learning_rate": 4.85570020550305e-06,
+      "loss": 0.4821,
+      "step": 5918
+    },
+    {
+      "epoch": 2.3217584588787354,
+      "grad_norm": 0.46839616082852714,
+      "learning_rate": 4.8556477429704265e-06,
+      "loss": 0.4997,
+      "step": 5919
+    },
+    {
+      "epoch": 2.3221536181773277,
+      "grad_norm": 0.45777076533689726,
+      "learning_rate": 4.855595271186249e-06,
+      "loss": 0.4868,
+      "step": 5920
+    },
+    {
+      "epoch": 2.32254877747592,
+      "grad_norm": 0.4606460723597928,
+      "learning_rate": 4.855542790150723e-06,
+      "loss": 0.4996,
+      "step": 5921
+    },
+    {
+      "epoch": 2.322943936774512,
+      "grad_norm": 0.474039852550981,
+      "learning_rate": 4.855490299864055e-06,
+      "loss": 0.5024,
+      "step": 5922
+    },
+    {
+      "epoch": 2.3233390960731044,
+      "grad_norm": 0.45026962258097275,
+      "learning_rate": 4.8554378003264525e-06,
+      "loss": 0.4911,
+      "step": 5923
+    },
+    {
+      "epoch": 2.3237342553716966,
+      "grad_norm": 0.47088446275779033,
+      "learning_rate": 4.85538529153812e-06,
+      "loss": 0.5054,
+      "step": 5924
+    },
+    {
+      "epoch": 2.324129414670289,
+      "grad_norm": 0.4520409677257949,
+      "learning_rate": 4.855332773499265e-06,
+      "loss": 0.5,
+      "step": 5925
+    },
+    {
+      "epoch": 2.324524573968881,
+      "grad_norm": 0.5002895000010738,
+      "learning_rate": 4.855280246210093e-06,
+      "loss": 0.5013,
+      "step": 5926
+    },
+    {
+      "epoch": 2.3249197332674734,
+      "grad_norm": 0.43977050907935966,
+      "learning_rate": 4.8552277096708104e-06,
+      "loss": 0.4955,
+      "step": 5927
+    },
+    {
+      "epoch": 2.3253148925660656,
+      "grad_norm": 0.44665214038970635,
+      "learning_rate": 4.855175163881623e-06,
+      "loss": 0.491,
+      "step": 5928
+    },
+    {
+      "epoch": 2.325710051864658,
+      "grad_norm": 0.4543680288607131,
+      "learning_rate": 4.855122608842738e-06,
+      "loss": 0.5021,
+      "step": 5929
+    },
+    {
+      "epoch": 2.32610521116325,
+      "grad_norm": 0.47105886533075386,
+      "learning_rate": 4.855070044554361e-06,
+      "loss": 0.4931,
+      "step": 5930
+    },
+    {
+      "epoch": 2.3265003704618423,
+      "grad_norm": 0.4542183871966276,
+      "learning_rate": 4.8550174710167e-06,
+      "loss": 0.4898,
+      "step": 5931
+    },
+    {
+      "epoch": 2.3268955297604346,
+      "grad_norm": 0.44138796987873863,
+      "learning_rate": 4.854964888229959e-06,
+      "loss": 0.4778,
+      "step": 5932
+    },
+    {
+      "epoch": 2.327290689059027,
+      "grad_norm": 0.4453768384169529,
+      "learning_rate": 4.854912296194347e-06,
+      "loss": 0.4952,
+      "step": 5933
+    },
+    {
+      "epoch": 2.327685848357619,
+      "grad_norm": 0.46292907268573785,
+      "learning_rate": 4.854859694910069e-06,
+      "loss": 0.5194,
+      "step": 5934
+    },
+    {
+      "epoch": 2.3280810076562113,
+      "grad_norm": 0.5289798354022537,
+      "learning_rate": 4.854807084377332e-06,
+      "loss": 0.4982,
+      "step": 5935
+    },
+    {
+      "epoch": 2.3284761669548035,
+      "grad_norm": 0.4844679893735134,
+      "learning_rate": 4.854754464596344e-06,
+      "loss": 0.5083,
+      "step": 5936
+    },
+    {
+      "epoch": 2.3288713262533958,
+      "grad_norm": 0.46477016057992765,
+      "learning_rate": 4.854701835567309e-06,
+      "loss": 0.5174,
+      "step": 5937
+    },
+    {
+      "epoch": 2.329266485551988,
+      "grad_norm": 0.4434767608118626,
+      "learning_rate": 4.8546491972904354e-06,
+      "loss": 0.4904,
+      "step": 5938
+    },
+    {
+      "epoch": 2.3296616448505802,
+      "grad_norm": 0.4481318939794802,
+      "learning_rate": 4.854596549765929e-06,
+      "loss": 0.5045,
+      "step": 5939
+    },
+    {
+      "epoch": 2.3300568041491725,
+      "grad_norm": 0.4534393794450779,
+      "learning_rate": 4.8545438929939985e-06,
+      "loss": 0.4957,
+      "step": 5940
+    },
+    {
+      "epoch": 2.3304519634477647,
+      "grad_norm": 0.4687124392558075,
+      "learning_rate": 4.854491226974848e-06,
+      "loss": 0.5176,
+      "step": 5941
+    },
+    {
+      "epoch": 2.330847122746357,
+      "grad_norm": 0.6355213440329489,
+      "learning_rate": 4.854438551708686e-06,
+      "loss": 0.5251,
+      "step": 5942
+    },
+    {
+      "epoch": 2.331242282044949,
+      "grad_norm": 0.45957482273724554,
+      "learning_rate": 4.854385867195719e-06,
+      "loss": 0.4918,
+      "step": 5943
+    },
+    {
+      "epoch": 2.3316374413435415,
+      "grad_norm": 0.46063015196640106,
+      "learning_rate": 4.854333173436154e-06,
+      "loss": 0.5072,
+      "step": 5944
+    },
+    {
+      "epoch": 2.3320326006421337,
+      "grad_norm": 0.46027939590902706,
+      "learning_rate": 4.854280470430199e-06,
+      "loss": 0.5212,
+      "step": 5945
+    },
+    {
+      "epoch": 2.332427759940726,
+      "grad_norm": 0.4697655644185264,
+      "learning_rate": 4.854227758178058e-06,
+      "loss": 0.4959,
+      "step": 5946
+    },
+    {
+      "epoch": 2.332822919239318,
+      "grad_norm": 0.45342411422226236,
+      "learning_rate": 4.854175036679941e-06,
+      "loss": 0.5064,
+      "step": 5947
+    },
+    {
+      "epoch": 2.3332180785379104,
+      "grad_norm": 0.4534823543771811,
+      "learning_rate": 4.854122305936054e-06,
+      "loss": 0.5074,
+      "step": 5948
+    },
+    {
+      "epoch": 2.3336132378365027,
+      "grad_norm": 0.46415220608251223,
+      "learning_rate": 4.8540695659466045e-06,
+      "loss": 0.4908,
+      "step": 5949
+    },
+    {
+      "epoch": 2.334008397135095,
+      "grad_norm": 0.4555359616647736,
+      "learning_rate": 4.854016816711799e-06,
+      "loss": 0.5075,
+      "step": 5950
+    },
+    {
+      "epoch": 2.3344035564336876,
+      "grad_norm": 0.4602201174143169,
+      "learning_rate": 4.853964058231844e-06,
+      "loss": 0.5022,
+      "step": 5951
+    },
+    {
+      "epoch": 2.33479871573228,
+      "grad_norm": 0.445668129634111,
+      "learning_rate": 4.853911290506949e-06,
+      "loss": 0.4977,
+      "step": 5952
+    },
+    {
+      "epoch": 2.335193875030872,
+      "grad_norm": 0.4677802262684022,
+      "learning_rate": 4.853858513537319e-06,
+      "loss": 0.5155,
+      "step": 5953
+    },
+    {
+      "epoch": 2.3355890343294643,
+      "grad_norm": 0.46379592618135573,
+      "learning_rate": 4.853805727323162e-06,
+      "loss": 0.5065,
+      "step": 5954
+    },
+    {
+      "epoch": 2.3359841936280565,
+      "grad_norm": 0.4438035631679444,
+      "learning_rate": 4.853752931864685e-06,
+      "loss": 0.4997,
+      "step": 5955
+    },
+    {
+      "epoch": 2.336379352926649,
+      "grad_norm": 0.46352707418602923,
+      "learning_rate": 4.853700127162097e-06,
+      "loss": 0.499,
+      "step": 5956
+    },
+    {
+      "epoch": 2.336774512225241,
+      "grad_norm": 0.4428618122246438,
+      "learning_rate": 4.8536473132156025e-06,
+      "loss": 0.4866,
+      "step": 5957
+    },
+    {
+      "epoch": 2.3371696715238333,
+      "grad_norm": 0.4521576676325371,
+      "learning_rate": 4.8535944900254115e-06,
+      "loss": 0.5002,
+      "step": 5958
+    },
+    {
+      "epoch": 2.3375648308224255,
+      "grad_norm": 0.4626467646021148,
+      "learning_rate": 4.853541657591731e-06,
+      "loss": 0.5019,
+      "step": 5959
+    },
+    {
+      "epoch": 2.3379599901210177,
+      "grad_norm": 0.46120148479836487,
+      "learning_rate": 4.853488815914768e-06,
+      "loss": 0.5007,
+      "step": 5960
+    },
+    {
+      "epoch": 2.33835514941961,
+      "grad_norm": 0.4506914517703311,
+      "learning_rate": 4.85343596499473e-06,
+      "loss": 0.4884,
+      "step": 5961
+    },
+    {
+      "epoch": 2.338750308718202,
+      "grad_norm": 0.43940905649704776,
+      "learning_rate": 4.853383104831823e-06,
+      "loss": 0.4956,
+      "step": 5962
+    },
+    {
+      "epoch": 2.3391454680167945,
+      "grad_norm": 0.46981121876573256,
+      "learning_rate": 4.853330235426258e-06,
+      "loss": 0.5144,
+      "step": 5963
+    },
+    {
+      "epoch": 2.3395406273153867,
+      "grad_norm": 0.4910013553279312,
+      "learning_rate": 4.85327735677824e-06,
+      "loss": 0.5202,
+      "step": 5964
+    },
+    {
+      "epoch": 2.339935786613979,
+      "grad_norm": 0.45060719865992904,
+      "learning_rate": 4.853224468887978e-06,
+      "loss": 0.502,
+      "step": 5965
+    },
+    {
+      "epoch": 2.340330945912571,
+      "grad_norm": 0.4553193621696011,
+      "learning_rate": 4.853171571755679e-06,
+      "loss": 0.4781,
+      "step": 5966
+    },
+    {
+      "epoch": 2.3407261052111634,
+      "grad_norm": 0.4515357883492096,
+      "learning_rate": 4.853118665381551e-06,
+      "loss": 0.4948,
+      "step": 5967
+    },
+    {
+      "epoch": 2.3411212645097557,
+      "grad_norm": 0.45077739345730894,
+      "learning_rate": 4.853065749765802e-06,
+      "loss": 0.494,
+      "step": 5968
+    },
+    {
+      "epoch": 2.341516423808348,
+      "grad_norm": 0.6697241027454435,
+      "learning_rate": 4.853012824908639e-06,
+      "loss": 0.5153,
+      "step": 5969
+    },
+    {
+      "epoch": 2.34191158310694,
+      "grad_norm": 0.46052939599652937,
+      "learning_rate": 4.852959890810271e-06,
+      "loss": 0.4899,
+      "step": 5970
+    },
+    {
+      "epoch": 2.3423067424055324,
+      "grad_norm": 0.4629470441078412,
+      "learning_rate": 4.852906947470905e-06,
+      "loss": 0.5122,
+      "step": 5971
+    },
+    {
+      "epoch": 2.3427019017041246,
+      "grad_norm": 0.45554485923089544,
+      "learning_rate": 4.8528539948907495e-06,
+      "loss": 0.5118,
+      "step": 5972
+    },
+    {
+      "epoch": 2.343097061002717,
+      "grad_norm": 0.46703177457997613,
+      "learning_rate": 4.8528010330700125e-06,
+      "loss": 0.4909,
+      "step": 5973
+    },
+    {
+      "epoch": 2.343492220301309,
+      "grad_norm": 0.4592646549588022,
+      "learning_rate": 4.852748062008901e-06,
+      "loss": 0.4904,
+      "step": 5974
+    },
+    {
+      "epoch": 2.3438873795999013,
+      "grad_norm": 0.474115303348079,
+      "learning_rate": 4.8526950817076244e-06,
+      "loss": 0.5038,
+      "step": 5975
+    },
+    {
+      "epoch": 2.3442825388984936,
+      "grad_norm": 0.48149869168465126,
+      "learning_rate": 4.85264209216639e-06,
+      "loss": 0.5108,
+      "step": 5976
+    },
+    {
+      "epoch": 2.344677698197086,
+      "grad_norm": 0.45837357450934973,
+      "learning_rate": 4.852589093385406e-06,
+      "loss": 0.5158,
+      "step": 5977
+    },
+    {
+      "epoch": 2.345072857495678,
+      "grad_norm": 0.44593351395167496,
+      "learning_rate": 4.852536085364881e-06,
+      "loss": 0.4967,
+      "step": 5978
+    },
+    {
+      "epoch": 2.3454680167942703,
+      "grad_norm": 0.4531644610848413,
+      "learning_rate": 4.852483068105022e-06,
+      "loss": 0.4981,
+      "step": 5979
+    },
+    {
+      "epoch": 2.3458631760928625,
+      "grad_norm": 0.4595627732654401,
+      "learning_rate": 4.852430041606039e-06,
+      "loss": 0.4888,
+      "step": 5980
+    },
+    {
+      "epoch": 2.346258335391455,
+      "grad_norm": 0.4581903361601766,
+      "learning_rate": 4.852377005868138e-06,
+      "loss": 0.4822,
+      "step": 5981
+    },
+    {
+      "epoch": 2.346653494690047,
+      "grad_norm": 0.4483989827690783,
+      "learning_rate": 4.85232396089153e-06,
+      "loss": 0.5124,
+      "step": 5982
+    },
+    {
+      "epoch": 2.3470486539886393,
+      "grad_norm": 0.45908454030513607,
+      "learning_rate": 4.8522709066764204e-06,
+      "loss": 0.5086,
+      "step": 5983
+    },
+    {
+      "epoch": 2.3474438132872315,
+      "grad_norm": 0.4583400146391861,
+      "learning_rate": 4.85221784322302e-06,
+      "loss": 0.5112,
+      "step": 5984
+    },
+    {
+      "epoch": 2.3478389725858237,
+      "grad_norm": 0.45763389698982093,
+      "learning_rate": 4.852164770531536e-06,
+      "loss": 0.5045,
+      "step": 5985
+    },
+    {
+      "epoch": 2.348234131884416,
+      "grad_norm": 0.460537227961999,
+      "learning_rate": 4.852111688602177e-06,
+      "loss": 0.5215,
+      "step": 5986
+    },
+    {
+      "epoch": 2.3486292911830082,
+      "grad_norm": 0.46679943581021927,
+      "learning_rate": 4.852058597435152e-06,
+      "loss": 0.5116,
+      "step": 5987
+    },
+    {
+      "epoch": 2.3490244504816005,
+      "grad_norm": 0.44817207334568054,
+      "learning_rate": 4.852005497030669e-06,
+      "loss": 0.5015,
+      "step": 5988
+    },
+    {
+      "epoch": 2.3494196097801927,
+      "grad_norm": 0.4489419200289153,
+      "learning_rate": 4.851952387388936e-06,
+      "loss": 0.5115,
+      "step": 5989
+    },
+    {
+      "epoch": 2.349814769078785,
+      "grad_norm": 0.45473940760337633,
+      "learning_rate": 4.851899268510163e-06,
+      "loss": 0.4971,
+      "step": 5990
+    },
+    {
+      "epoch": 2.350209928377377,
+      "grad_norm": 0.4595345160766586,
+      "learning_rate": 4.851846140394557e-06,
+      "loss": 0.5165,
+      "step": 5991
+    },
+    {
+      "epoch": 2.3506050876759694,
+      "grad_norm": 0.45078645033514275,
+      "learning_rate": 4.851793003042328e-06,
+      "loss": 0.4801,
+      "step": 5992
+    },
+    {
+      "epoch": 2.3510002469745617,
+      "grad_norm": 0.46083912000399313,
+      "learning_rate": 4.851739856453685e-06,
+      "loss": 0.506,
+      "step": 5993
+    },
+    {
+      "epoch": 2.351395406273154,
+      "grad_norm": 0.456612483620041,
+      "learning_rate": 4.851686700628834e-06,
+      "loss": 0.5031,
+      "step": 5994
+    },
+    {
+      "epoch": 2.351790565571746,
+      "grad_norm": 0.4617876715575406,
+      "learning_rate": 4.851633535567987e-06,
+      "loss": 0.5115,
+      "step": 5995
+    },
+    {
+      "epoch": 2.3521857248703384,
+      "grad_norm": 0.5326660248295265,
+      "learning_rate": 4.851580361271351e-06,
+      "loss": 0.5117,
+      "step": 5996
+    },
+    {
+      "epoch": 2.3525808841689306,
+      "grad_norm": 0.4613788668226646,
+      "learning_rate": 4.851527177739135e-06,
+      "loss": 0.5209,
+      "step": 5997
+    },
+    {
+      "epoch": 2.352976043467523,
+      "grad_norm": 0.44443451469751716,
+      "learning_rate": 4.851473984971549e-06,
+      "loss": 0.4932,
+      "step": 5998
+    },
+    {
+      "epoch": 2.353371202766115,
+      "grad_norm": 0.45353874492848084,
+      "learning_rate": 4.851420782968801e-06,
+      "loss": 0.512,
+      "step": 5999
+    },
+    {
+      "epoch": 2.3537663620647074,
+      "grad_norm": 0.45136448249781186,
+      "learning_rate": 4.8513675717311e-06,
+      "loss": 0.4939,
+      "step": 6000
+    },
+    {
+      "epoch": 2.3541615213632996,
+      "grad_norm": 0.4516803569929796,
+      "learning_rate": 4.851314351258654e-06,
+      "loss": 0.5113,
+      "step": 6001
+    },
+    {
+      "epoch": 2.354556680661892,
+      "grad_norm": 0.4561308566297529,
+      "learning_rate": 4.851261121551674e-06,
+      "loss": 0.4967,
+      "step": 6002
+    },
+    {
+      "epoch": 2.354951839960484,
+      "grad_norm": 0.5260342783696013,
+      "learning_rate": 4.8512078826103675e-06,
+      "loss": 0.4982,
+      "step": 6003
+    },
+    {
+      "epoch": 2.3553469992590763,
+      "grad_norm": 0.47897271506812045,
+      "learning_rate": 4.8511546344349444e-06,
+      "loss": 0.5193,
+      "step": 6004
+    },
+    {
+      "epoch": 2.3557421585576686,
+      "grad_norm": 0.4638833627455354,
+      "learning_rate": 4.851101377025614e-06,
+      "loss": 0.4941,
+      "step": 6005
+    },
+    {
+      "epoch": 2.356137317856261,
+      "grad_norm": 0.4427664156009943,
+      "learning_rate": 4.8510481103825845e-06,
+      "loss": 0.4878,
+      "step": 6006
+    },
+    {
+      "epoch": 2.356532477154853,
+      "grad_norm": 0.44716700666783465,
+      "learning_rate": 4.850994834506065e-06,
+      "loss": 0.4942,
+      "step": 6007
+    },
+    {
+      "epoch": 2.3569276364534453,
+      "grad_norm": 0.45290206397069754,
+      "learning_rate": 4.850941549396267e-06,
+      "loss": 0.504,
+      "step": 6008
+    },
+    {
+      "epoch": 2.3573227957520375,
+      "grad_norm": 0.4624224635836012,
+      "learning_rate": 4.850888255053398e-06,
+      "loss": 0.5128,
+      "step": 6009
+    },
+    {
+      "epoch": 2.3577179550506298,
+      "grad_norm": 0.4685159236259062,
+      "learning_rate": 4.850834951477666e-06,
+      "loss": 0.5056,
+      "step": 6010
+    },
+    {
+      "epoch": 2.358113114349222,
+      "grad_norm": 0.4637190744227209,
+      "learning_rate": 4.850781638669283e-06,
+      "loss": 0.4789,
+      "step": 6011
+    },
+    {
+      "epoch": 2.3585082736478142,
+      "grad_norm": 0.4519731150582127,
+      "learning_rate": 4.850728316628457e-06,
+      "loss": 0.5129,
+      "step": 6012
+    },
+    {
+      "epoch": 2.3589034329464065,
+      "grad_norm": 0.4538513437911314,
+      "learning_rate": 4.8506749853553974e-06,
+      "loss": 0.4994,
+      "step": 6013
+    },
+    {
+      "epoch": 2.3592985922449987,
+      "grad_norm": 0.46979489205654373,
+      "learning_rate": 4.850621644850314e-06,
+      "loss": 0.5151,
+      "step": 6014
+    },
+    {
+      "epoch": 2.359693751543591,
+      "grad_norm": 0.45746167900887663,
+      "learning_rate": 4.850568295113416e-06,
+      "loss": 0.5168,
+      "step": 6015
+    },
+    {
+      "epoch": 2.360088910842183,
+      "grad_norm": 0.4575481082495061,
+      "learning_rate": 4.850514936144913e-06,
+      "loss": 0.5072,
+      "step": 6016
+    },
+    {
+      "epoch": 2.3604840701407754,
+      "grad_norm": 0.45975740355910427,
+      "learning_rate": 4.850461567945015e-06,
+      "loss": 0.5,
+      "step": 6017
+    },
+    {
+      "epoch": 2.3608792294393677,
+      "grad_norm": 0.4748694051073642,
+      "learning_rate": 4.850408190513931e-06,
+      "loss": 0.5029,
+      "step": 6018
+    },
+    {
+      "epoch": 2.36127438873796,
+      "grad_norm": 0.4914980276998342,
+      "learning_rate": 4.850354803851871e-06,
+      "loss": 0.5254,
+      "step": 6019
+    },
+    {
+      "epoch": 2.361669548036552,
+      "grad_norm": 0.4800284534033041,
+      "learning_rate": 4.850301407959045e-06,
+      "loss": 0.5161,
+      "step": 6020
+    },
+    {
+      "epoch": 2.3620647073351444,
+      "grad_norm": 0.4462205733041414,
+      "learning_rate": 4.8502480028356615e-06,
+      "loss": 0.5042,
+      "step": 6021
+    },
+    {
+      "epoch": 2.3624598666337366,
+      "grad_norm": 0.4709204676815143,
+      "learning_rate": 4.850194588481931e-06,
+      "loss": 0.502,
+      "step": 6022
+    },
+    {
+      "epoch": 2.362855025932329,
+      "grad_norm": 0.4568772714716113,
+      "learning_rate": 4.8501411648980635e-06,
+      "loss": 0.4653,
+      "step": 6023
+    },
+    {
+      "epoch": 2.363250185230921,
+      "grad_norm": 0.4547015608724961,
+      "learning_rate": 4.850087732084269e-06,
+      "loss": 0.5,
+      "step": 6024
+    },
+    {
+      "epoch": 2.3636453445295134,
+      "grad_norm": 0.480326169527228,
+      "learning_rate": 4.850034290040756e-06,
+      "loss": 0.5118,
+      "step": 6025
+    },
+    {
+      "epoch": 2.3640405038281056,
+      "grad_norm": 0.4598744970305476,
+      "learning_rate": 4.849980838767736e-06,
+      "loss": 0.498,
+      "step": 6026
+    },
+    {
+      "epoch": 2.364435663126698,
+      "grad_norm": 0.4600510631142906,
+      "learning_rate": 4.849927378265418e-06,
+      "loss": 0.4725,
+      "step": 6027
+    },
+    {
+      "epoch": 2.36483082242529,
+      "grad_norm": 0.4633225031808518,
+      "learning_rate": 4.8498739085340125e-06,
+      "loss": 0.5029,
+      "step": 6028
+    },
+    {
+      "epoch": 2.3652259817238823,
+      "grad_norm": 0.4497673903824722,
+      "learning_rate": 4.849820429573729e-06,
+      "loss": 0.5061,
+      "step": 6029
+    },
+    {
+      "epoch": 2.3656211410224746,
+      "grad_norm": 0.45283626682133,
+      "learning_rate": 4.849766941384777e-06,
+      "loss": 0.5035,
+      "step": 6030
+    },
+    {
+      "epoch": 2.366016300321067,
+      "grad_norm": 0.45318111987548476,
+      "learning_rate": 4.8497134439673685e-06,
+      "loss": 0.5059,
+      "step": 6031
+    },
+    {
+      "epoch": 2.366411459619659,
+      "grad_norm": 0.45785207671037326,
+      "learning_rate": 4.849659937321713e-06,
+      "loss": 0.527,
+      "step": 6032
+    },
+    {
+      "epoch": 2.3668066189182513,
+      "grad_norm": 0.45427024013798384,
+      "learning_rate": 4.849606421448018e-06,
+      "loss": 0.5253,
+      "step": 6033
+    },
+    {
+      "epoch": 2.3672017782168435,
+      "grad_norm": 0.45732742980462926,
+      "learning_rate": 4.849552896346497e-06,
+      "loss": 0.504,
+      "step": 6034
+    },
+    {
+      "epoch": 2.3675969375154358,
+      "grad_norm": 0.5128918032034719,
+      "learning_rate": 4.849499362017359e-06,
+      "loss": 0.5012,
+      "step": 6035
+    },
+    {
+      "epoch": 2.367992096814028,
+      "grad_norm": 0.46442492874120134,
+      "learning_rate": 4.8494458184608135e-06,
+      "loss": 0.4985,
+      "step": 6036
+    },
+    {
+      "epoch": 2.3683872561126202,
+      "grad_norm": 0.44382409744050355,
+      "learning_rate": 4.849392265677072e-06,
+      "loss": 0.4984,
+      "step": 6037
+    },
+    {
+      "epoch": 2.3687824154112125,
+      "grad_norm": 0.4417781251327167,
+      "learning_rate": 4.8493387036663445e-06,
+      "loss": 0.4904,
+      "step": 6038
+    },
+    {
+      "epoch": 2.3691775747098047,
+      "grad_norm": 0.46452188611272743,
+      "learning_rate": 4.84928513242884e-06,
+      "loss": 0.5057,
+      "step": 6039
+    },
+    {
+      "epoch": 2.369572734008397,
+      "grad_norm": 0.4939272964413806,
+      "learning_rate": 4.849231551964771e-06,
+      "loss": 0.5275,
+      "step": 6040
+    },
+    {
+      "epoch": 2.369967893306989,
+      "grad_norm": 0.4656217747759951,
+      "learning_rate": 4.849177962274348e-06,
+      "loss": 0.4959,
+      "step": 6041
+    },
+    {
+      "epoch": 2.3703630526055814,
+      "grad_norm": 0.45451303351162,
+      "learning_rate": 4.8491243633577785e-06,
+      "loss": 0.4827,
+      "step": 6042
+    },
+    {
+      "epoch": 2.3707582119041737,
+      "grad_norm": 0.46682609497280303,
+      "learning_rate": 4.849070755215276e-06,
+      "loss": 0.4905,
+      "step": 6043
+    },
+    {
+      "epoch": 2.371153371202766,
+      "grad_norm": 0.451485975019632,
+      "learning_rate": 4.849017137847049e-06,
+      "loss": 0.5045,
+      "step": 6044
+    },
+    {
+      "epoch": 2.371548530501358,
+      "grad_norm": 0.4631255854420321,
+      "learning_rate": 4.84896351125331e-06,
+      "loss": 0.5326,
+      "step": 6045
+    },
+    {
+      "epoch": 2.3719436897999504,
+      "grad_norm": 0.4571223855093204,
+      "learning_rate": 4.848909875434269e-06,
+      "loss": 0.4987,
+      "step": 6046
+    },
+    {
+      "epoch": 2.3723388490985426,
+      "grad_norm": 0.4680677582589837,
+      "learning_rate": 4.848856230390137e-06,
+      "loss": 0.5158,
+      "step": 6047
+    },
+    {
+      "epoch": 2.372734008397135,
+      "grad_norm": 0.4446257712595702,
+      "learning_rate": 4.848802576121122e-06,
+      "loss": 0.5033,
+      "step": 6048
+    },
+    {
+      "epoch": 2.373129167695727,
+      "grad_norm": 0.4630473866895991,
+      "learning_rate": 4.848748912627438e-06,
+      "loss": 0.5042,
+      "step": 6049
+    },
+    {
+      "epoch": 2.3735243269943194,
+      "grad_norm": 0.44778383563697305,
+      "learning_rate": 4.848695239909295e-06,
+      "loss": 0.4975,
+      "step": 6050
+    },
+    {
+      "epoch": 2.3739194862929116,
+      "grad_norm": 0.4385248274134849,
+      "learning_rate": 4.848641557966902e-06,
+      "loss": 0.4859,
+      "step": 6051
+    },
+    {
+      "epoch": 2.374314645591504,
+      "grad_norm": 0.4527513035838103,
+      "learning_rate": 4.848587866800472e-06,
+      "loss": 0.4834,
+      "step": 6052
+    },
+    {
+      "epoch": 2.374709804890096,
+      "grad_norm": 0.45987272190820844,
+      "learning_rate": 4.8485341664102146e-06,
+      "loss": 0.4934,
+      "step": 6053
+    },
+    {
+      "epoch": 2.3751049641886883,
+      "grad_norm": 0.44214190083964766,
+      "learning_rate": 4.84848045679634e-06,
+      "loss": 0.4863,
+      "step": 6054
+    },
+    {
+      "epoch": 2.3755001234872806,
+      "grad_norm": 0.47023720851459105,
+      "learning_rate": 4.848426737959062e-06,
+      "loss": 0.5147,
+      "step": 6055
+    },
+    {
+      "epoch": 2.375895282785873,
+      "grad_norm": 0.4536759734264281,
+      "learning_rate": 4.848373009898589e-06,
+      "loss": 0.5058,
+      "step": 6056
+    },
+    {
+      "epoch": 2.3762904420844655,
+      "grad_norm": 0.46339461388038883,
+      "learning_rate": 4.848319272615134e-06,
+      "loss": 0.5292,
+      "step": 6057
+    },
+    {
+      "epoch": 2.3766856013830577,
+      "grad_norm": 0.43958334885581213,
+      "learning_rate": 4.848265526108906e-06,
+      "loss": 0.485,
+      "step": 6058
+    },
+    {
+      "epoch": 2.37708076068165,
+      "grad_norm": 0.4442145240300568,
+      "learning_rate": 4.848211770380117e-06,
+      "loss": 0.4961,
+      "step": 6059
+    },
+    {
+      "epoch": 2.377475919980242,
+      "grad_norm": 0.47607365592400747,
+      "learning_rate": 4.848158005428978e-06,
+      "loss": 0.5151,
+      "step": 6060
+    },
+    {
+      "epoch": 2.3778710792788345,
+      "grad_norm": 0.4512904547194257,
+      "learning_rate": 4.8481042312557e-06,
+      "loss": 0.4923,
+      "step": 6061
+    },
+    {
+      "epoch": 2.3782662385774267,
+      "grad_norm": 0.44456571012970686,
+      "learning_rate": 4.8480504478604946e-06,
+      "loss": 0.5055,
+      "step": 6062
+    },
+    {
+      "epoch": 2.378661397876019,
+      "grad_norm": 0.4560502078542866,
+      "learning_rate": 4.847996655243572e-06,
+      "loss": 0.4988,
+      "step": 6063
+    },
+    {
+      "epoch": 2.379056557174611,
+      "grad_norm": 0.45340332096358,
+      "learning_rate": 4.847942853405146e-06,
+      "loss": 0.4981,
+      "step": 6064
+    },
+    {
+      "epoch": 2.3794517164732034,
+      "grad_norm": 0.4581291968585293,
+      "learning_rate": 4.847889042345425e-06,
+      "loss": 0.4771,
+      "step": 6065
+    },
+    {
+      "epoch": 2.3798468757717957,
+      "grad_norm": 0.459995603273202,
+      "learning_rate": 4.8478352220646215e-06,
+      "loss": 0.4929,
+      "step": 6066
+    },
+    {
+      "epoch": 2.380242035070388,
+      "grad_norm": 0.4474153131280138,
+      "learning_rate": 4.847781392562948e-06,
+      "loss": 0.5069,
+      "step": 6067
+    },
+    {
+      "epoch": 2.38063719436898,
+      "grad_norm": 0.4493358041327097,
+      "learning_rate": 4.847727553840615e-06,
+      "loss": 0.4881,
+      "step": 6068
+    },
+    {
+      "epoch": 2.3810323536675724,
+      "grad_norm": 0.45233974550022343,
+      "learning_rate": 4.847673705897832e-06,
+      "loss": 0.5081,
+      "step": 6069
+    },
+    {
+      "epoch": 2.3814275129661646,
+      "grad_norm": 0.5070469825872892,
+      "learning_rate": 4.847619848734814e-06,
+      "loss": 0.5149,
+      "step": 6070
+    },
+    {
+      "epoch": 2.381822672264757,
+      "grad_norm": 0.4565568022870259,
+      "learning_rate": 4.84756598235177e-06,
+      "loss": 0.4859,
+      "step": 6071
+    },
+    {
+      "epoch": 2.382217831563349,
+      "grad_norm": 0.4871291260740165,
+      "learning_rate": 4.847512106748912e-06,
+      "loss": 0.5239,
+      "step": 6072
+    },
+    {
+      "epoch": 2.3826129908619413,
+      "grad_norm": 0.4504152728422771,
+      "learning_rate": 4.847458221926453e-06,
+      "loss": 0.5099,
+      "step": 6073
+    },
+    {
+      "epoch": 2.3830081501605336,
+      "grad_norm": 0.4591280231930706,
+      "learning_rate": 4.847404327884603e-06,
+      "loss": 0.4842,
+      "step": 6074
+    },
+    {
+      "epoch": 2.383403309459126,
+      "grad_norm": 0.45851366936986665,
+      "learning_rate": 4.847350424623574e-06,
+      "loss": 0.5088,
+      "step": 6075
+    },
+    {
+      "epoch": 2.383798468757718,
+      "grad_norm": 0.4634869558586333,
+      "learning_rate": 4.847296512143577e-06,
+      "loss": 0.5131,
+      "step": 6076
+    },
+    {
+      "epoch": 2.3841936280563103,
+      "grad_norm": 0.45304642062746575,
+      "learning_rate": 4.847242590444826e-06,
+      "loss": 0.4977,
+      "step": 6077
+    },
+    {
+      "epoch": 2.3845887873549025,
+      "grad_norm": 0.4561346179180274,
+      "learning_rate": 4.847188659527532e-06,
+      "loss": 0.5047,
+      "step": 6078
+    },
+    {
+      "epoch": 2.384983946653495,
+      "grad_norm": 0.4526743884370499,
+      "learning_rate": 4.847134719391905e-06,
+      "loss": 0.4914,
+      "step": 6079
+    },
+    {
+      "epoch": 2.385379105952087,
+      "grad_norm": 0.46407896971341867,
+      "learning_rate": 4.847080770038158e-06,
+      "loss": 0.5089,
+      "step": 6080
+    },
+    {
+      "epoch": 2.3857742652506793,
+      "grad_norm": 0.45258104010987427,
+      "learning_rate": 4.847026811466504e-06,
+      "loss": 0.5022,
+      "step": 6081
+    },
+    {
+      "epoch": 2.3861694245492715,
+      "grad_norm": 0.461969126357832,
+      "learning_rate": 4.846972843677153e-06,
+      "loss": 0.497,
+      "step": 6082
+    },
+    {
+      "epoch": 2.3865645838478637,
+      "grad_norm": 0.4492623434194907,
+      "learning_rate": 4.846918866670318e-06,
+      "loss": 0.4979,
+      "step": 6083
+    },
+    {
+      "epoch": 2.386959743146456,
+      "grad_norm": 0.4571175682930661,
+      "learning_rate": 4.846864880446211e-06,
+      "loss": 0.5196,
+      "step": 6084
+    },
+    {
+      "epoch": 2.387354902445048,
+      "grad_norm": 0.45913437019988257,
+      "learning_rate": 4.8468108850050436e-06,
+      "loss": 0.5046,
+      "step": 6085
+    },
+    {
+      "epoch": 2.3877500617436405,
+      "grad_norm": 0.45074777593560994,
+      "learning_rate": 4.846756880347029e-06,
+      "loss": 0.4989,
+      "step": 6086
+    },
+    {
+      "epoch": 2.3881452210422327,
+      "grad_norm": 0.44092077546541725,
+      "learning_rate": 4.846702866472377e-06,
+      "loss": 0.4794,
+      "step": 6087
+    },
+    {
+      "epoch": 2.388540380340825,
+      "grad_norm": 0.46155684618109705,
+      "learning_rate": 4.846648843381302e-06,
+      "loss": 0.5089,
+      "step": 6088
+    },
+    {
+      "epoch": 2.388935539639417,
+      "grad_norm": 0.46885706820416534,
+      "learning_rate": 4.846594811074015e-06,
+      "loss": 0.5171,
+      "step": 6089
+    },
+    {
+      "epoch": 2.3893306989380094,
+      "grad_norm": 0.43931037237601184,
+      "learning_rate": 4.846540769550728e-06,
+      "loss": 0.4882,
+      "step": 6090
+    },
+    {
+      "epoch": 2.3897258582366017,
+      "grad_norm": 0.4353274011341564,
+      "learning_rate": 4.8464867188116545e-06,
+      "loss": 0.482,
+      "step": 6091
+    },
+    {
+      "epoch": 2.390121017535194,
+      "grad_norm": 0.47037433987558747,
+      "learning_rate": 4.846432658857006e-06,
+      "loss": 0.4994,
+      "step": 6092
+    },
+    {
+      "epoch": 2.390516176833786,
+      "grad_norm": 0.4593847264977308,
+      "learning_rate": 4.846378589686995e-06,
+      "loss": 0.507,
+      "step": 6093
+    },
+    {
+      "epoch": 2.3909113361323784,
+      "grad_norm": 0.4601985413381888,
+      "learning_rate": 4.846324511301834e-06,
+      "loss": 0.5023,
+      "step": 6094
+    },
+    {
+      "epoch": 2.3913064954309706,
+      "grad_norm": 0.452640709655825,
+      "learning_rate": 4.846270423701734e-06,
+      "loss": 0.4995,
+      "step": 6095
+    },
+    {
+      "epoch": 2.391701654729563,
+      "grad_norm": 0.45638264701889014,
+      "learning_rate": 4.846216326886909e-06,
+      "loss": 0.4959,
+      "step": 6096
+    },
+    {
+      "epoch": 2.392096814028155,
+      "grad_norm": 0.4826711924842887,
+      "learning_rate": 4.846162220857571e-06,
+      "loss": 0.5177,
+      "step": 6097
+    },
+    {
+      "epoch": 2.3924919733267473,
+      "grad_norm": 0.4888215683792986,
+      "learning_rate": 4.846108105613932e-06,
+      "loss": 0.5008,
+      "step": 6098
+    },
+    {
+      "epoch": 2.3928871326253396,
+      "grad_norm": 0.46320608794522933,
+      "learning_rate": 4.8460539811562055e-06,
+      "loss": 0.5089,
+      "step": 6099
+    },
+    {
+      "epoch": 2.393282291923932,
+      "grad_norm": 0.45725756454431987,
+      "learning_rate": 4.845999847484604e-06,
+      "loss": 0.5203,
+      "step": 6100
+    },
+    {
+      "epoch": 2.393677451222524,
+      "grad_norm": 0.44991333274030526,
+      "learning_rate": 4.84594570459934e-06,
+      "loss": 0.4848,
+      "step": 6101
+    },
+    {
+      "epoch": 2.3940726105211163,
+      "grad_norm": 0.4600972177592437,
+      "learning_rate": 4.845891552500625e-06,
+      "loss": 0.5093,
+      "step": 6102
+    },
+    {
+      "epoch": 2.3944677698197085,
+      "grad_norm": 0.5860812755960613,
+      "learning_rate": 4.8458373911886716e-06,
+      "loss": 0.5069,
+      "step": 6103
+    },
+    {
+      "epoch": 2.394862929118301,
+      "grad_norm": 0.4538835124457885,
+      "learning_rate": 4.845783220663694e-06,
+      "loss": 0.4874,
+      "step": 6104
+    },
+    {
+      "epoch": 2.395258088416893,
+      "grad_norm": 0.45218392628786025,
+      "learning_rate": 4.845729040925905e-06,
+      "loss": 0.496,
+      "step": 6105
+    },
+    {
+      "epoch": 2.3956532477154853,
+      "grad_norm": 0.4493156953117064,
+      "learning_rate": 4.845674851975516e-06,
+      "loss": 0.5225,
+      "step": 6106
+    },
+    {
+      "epoch": 2.3960484070140775,
+      "grad_norm": 0.4721248920273722,
+      "learning_rate": 4.845620653812742e-06,
+      "loss": 0.503,
+      "step": 6107
+    },
+    {
+      "epoch": 2.3964435663126697,
+      "grad_norm": 0.4702634244511809,
+      "learning_rate": 4.845566446437793e-06,
+      "loss": 0.509,
+      "step": 6108
+    },
+    {
+      "epoch": 2.396838725611262,
+      "grad_norm": 0.4517371761872178,
+      "learning_rate": 4.845512229850883e-06,
+      "loss": 0.5005,
+      "step": 6109
+    },
+    {
+      "epoch": 2.3972338849098542,
+      "grad_norm": 0.45784931465490497,
+      "learning_rate": 4.845458004052226e-06,
+      "loss": 0.5292,
+      "step": 6110
+    },
+    {
+      "epoch": 2.3976290442084465,
+      "grad_norm": 0.47366039215477546,
+      "learning_rate": 4.845403769042034e-06,
+      "loss": 0.4974,
+      "step": 6111
+    },
+    {
+      "epoch": 2.3980242035070387,
+      "grad_norm": 0.45191947098684676,
+      "learning_rate": 4.8453495248205205e-06,
+      "loss": 0.5141,
+      "step": 6112
+    },
+    {
+      "epoch": 2.398419362805631,
+      "grad_norm": 0.4713607350755517,
+      "learning_rate": 4.845295271387897e-06,
+      "loss": 0.4976,
+      "step": 6113
+    },
+    {
+      "epoch": 2.398814522104223,
+      "grad_norm": 0.4660999045638153,
+      "learning_rate": 4.84524100874438e-06,
+      "loss": 0.5113,
+      "step": 6114
+    },
+    {
+      "epoch": 2.3992096814028154,
+      "grad_norm": 0.47015668850310877,
+      "learning_rate": 4.845186736890179e-06,
+      "loss": 0.4907,
+      "step": 6115
+    },
+    {
+      "epoch": 2.3996048407014077,
+      "grad_norm": 0.46766792350712855,
+      "learning_rate": 4.845132455825508e-06,
+      "loss": 0.513,
+      "step": 6116
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.0836917520596936,
+      "learning_rate": 4.8450781655505815e-06,
+      "loss": 0.4935,
+      "step": 6117
+    },
+    {
+      "epoch": 2.400395159298592,
+      "grad_norm": 0.46151727429931927,
+      "learning_rate": 4.845023866065612e-06,
+      "loss": 0.5011,
+      "step": 6118
+    },
+    {
+      "epoch": 2.4007903185971844,
+      "grad_norm": 0.4605620408152997,
+      "learning_rate": 4.844969557370813e-06,
+      "loss": 0.5045,
+      "step": 6119
+    },
+    {
+      "epoch": 2.4011854778957766,
+      "grad_norm": 0.45858291441150556,
+      "learning_rate": 4.844915239466398e-06,
+      "loss": 0.5035,
+      "step": 6120
+    },
+    {
+      "epoch": 2.401580637194369,
+      "grad_norm": 0.4572405133813424,
+      "learning_rate": 4.844860912352579e-06,
+      "loss": 0.4987,
+      "step": 6121
+    },
+    {
+      "epoch": 2.401975796492961,
+      "grad_norm": 0.4532962343436789,
+      "learning_rate": 4.844806576029571e-06,
+      "loss": 0.5168,
+      "step": 6122
+    },
+    {
+      "epoch": 2.4023709557915534,
+      "grad_norm": 0.44424364135152433,
+      "learning_rate": 4.844752230497586e-06,
+      "loss": 0.4874,
+      "step": 6123
+    },
+    {
+      "epoch": 2.4027661150901456,
+      "grad_norm": 0.45178951408739765,
+      "learning_rate": 4.844697875756837e-06,
+      "loss": 0.5028,
+      "step": 6124
+    },
+    {
+      "epoch": 2.403161274388738,
+      "grad_norm": 0.46131443978892384,
+      "learning_rate": 4.844643511807539e-06,
+      "loss": 0.4997,
+      "step": 6125
+    },
+    {
+      "epoch": 2.40355643368733,
+      "grad_norm": 0.44557753095639435,
+      "learning_rate": 4.844589138649906e-06,
+      "loss": 0.4961,
+      "step": 6126
+    },
+    {
+      "epoch": 2.4039515929859223,
+      "grad_norm": 0.44660632251012705,
+      "learning_rate": 4.84453475628415e-06,
+      "loss": 0.5055,
+      "step": 6127
+    },
+    {
+      "epoch": 2.4043467522845146,
+      "grad_norm": 0.46297338232365337,
+      "learning_rate": 4.844480364710486e-06,
+      "loss": 0.5106,
+      "step": 6128
+    },
+    {
+      "epoch": 2.404741911583107,
+      "grad_norm": 0.4546931439519957,
+      "learning_rate": 4.844425963929126e-06,
+      "loss": 0.507,
+      "step": 6129
+    },
+    {
+      "epoch": 2.405137070881699,
+      "grad_norm": 0.4393631016625778,
+      "learning_rate": 4.844371553940284e-06,
+      "loss": 0.5011,
+      "step": 6130
+    },
+    {
+      "epoch": 2.4055322301802913,
+      "grad_norm": 0.44198752628335675,
+      "learning_rate": 4.844317134744174e-06,
+      "loss": 0.5045,
+      "step": 6131
+    },
+    {
+      "epoch": 2.4059273894788835,
+      "grad_norm": 0.4529021301640311,
+      "learning_rate": 4.844262706341011e-06,
+      "loss": 0.5139,
+      "step": 6132
+    },
+    {
+      "epoch": 2.4063225487774758,
+      "grad_norm": 0.46321469858495373,
+      "learning_rate": 4.844208268731007e-06,
+      "loss": 0.4949,
+      "step": 6133
+    },
+    {
+      "epoch": 2.406717708076068,
+      "grad_norm": 0.5352727595143272,
+      "learning_rate": 4.8441538219143765e-06,
+      "loss": 0.5112,
+      "step": 6134
+    },
+    {
+      "epoch": 2.4071128673746602,
+      "grad_norm": 0.45404512767050736,
+      "learning_rate": 4.844099365891333e-06,
+      "loss": 0.5071,
+      "step": 6135
+    },
+    {
+      "epoch": 2.4075080266732525,
+      "grad_norm": 0.44661826142521477,
+      "learning_rate": 4.844044900662091e-06,
+      "loss": 0.4842,
+      "step": 6136
+    },
+    {
+      "epoch": 2.4079031859718447,
+      "grad_norm": 0.4866476660219383,
+      "learning_rate": 4.843990426226864e-06,
+      "loss": 0.4978,
+      "step": 6137
+    },
+    {
+      "epoch": 2.408298345270437,
+      "grad_norm": 0.48353551024417507,
+      "learning_rate": 4.843935942585865e-06,
+      "loss": 0.5364,
+      "step": 6138
+    },
+    {
+      "epoch": 2.408693504569029,
+      "grad_norm": 0.4710485363082892,
+      "learning_rate": 4.84388144973931e-06,
+      "loss": 0.503,
+      "step": 6139
+    },
+    {
+      "epoch": 2.409088663867622,
+      "grad_norm": 0.4574698914104126,
+      "learning_rate": 4.843826947687412e-06,
+      "loss": 0.5185,
+      "step": 6140
+    },
+    {
+      "epoch": 2.409483823166214,
+      "grad_norm": 0.46016949246716193,
+      "learning_rate": 4.843772436430384e-06,
+      "loss": 0.5148,
+      "step": 6141
+    },
+    {
+      "epoch": 2.4098789824648064,
+      "grad_norm": 0.4482322493892279,
+      "learning_rate": 4.843717915968442e-06,
+      "loss": 0.5046,
+      "step": 6142
+    },
+    {
+      "epoch": 2.4102741417633986,
+      "grad_norm": 0.45116521545047983,
+      "learning_rate": 4.843663386301799e-06,
+      "loss": 0.515,
+      "step": 6143
+    },
+    {
+      "epoch": 2.410669301061991,
+      "grad_norm": 0.4590353660506984,
+      "learning_rate": 4.843608847430669e-06,
+      "loss": 0.5015,
+      "step": 6144
+    },
+    {
+      "epoch": 2.411064460360583,
+      "grad_norm": 0.4509801127921544,
+      "learning_rate": 4.843554299355267e-06,
+      "loss": 0.5036,
+      "step": 6145
+    },
+    {
+      "epoch": 2.4114596196591753,
+      "grad_norm": 0.46483479028215885,
+      "learning_rate": 4.8434997420758065e-06,
+      "loss": 0.5002,
+      "step": 6146
+    },
+    {
+      "epoch": 2.4118547789577676,
+      "grad_norm": 0.4594113983493731,
+      "learning_rate": 4.843445175592502e-06,
+      "loss": 0.5072,
+      "step": 6147
+    },
+    {
+      "epoch": 2.41224993825636,
+      "grad_norm": 0.45010607482721837,
+      "learning_rate": 4.843390599905568e-06,
+      "loss": 0.509,
+      "step": 6148
+    },
+    {
+      "epoch": 2.412645097554952,
+      "grad_norm": 0.47368631625572943,
+      "learning_rate": 4.843336015015218e-06,
+      "loss": 0.5363,
+      "step": 6149
+    },
+    {
+      "epoch": 2.4130402568535443,
+      "grad_norm": 0.4583173953744174,
+      "learning_rate": 4.843281420921668e-06,
+      "loss": 0.4834,
+      "step": 6150
+    },
+    {
+      "epoch": 2.4134354161521365,
+      "grad_norm": 0.46218829266694594,
+      "learning_rate": 4.843226817625132e-06,
+      "loss": 0.5132,
+      "step": 6151
+    },
+    {
+      "epoch": 2.4138305754507288,
+      "grad_norm": 0.46198156224588854,
+      "learning_rate": 4.843172205125824e-06,
+      "loss": 0.5132,
+      "step": 6152
+    },
+    {
+      "epoch": 2.414225734749321,
+      "grad_norm": 0.4703069352755098,
+      "learning_rate": 4.843117583423957e-06,
+      "loss": 0.5222,
+      "step": 6153
+    },
+    {
+      "epoch": 2.4146208940479132,
+      "grad_norm": 0.4614134551221174,
+      "learning_rate": 4.843062952519748e-06,
+      "loss": 0.4887,
+      "step": 6154
+    },
+    {
+      "epoch": 2.4150160533465055,
+      "grad_norm": 0.4697970415651049,
+      "learning_rate": 4.843008312413409e-06,
+      "loss": 0.5093,
+      "step": 6155
+    },
+    {
+      "epoch": 2.4154112126450977,
+      "grad_norm": 0.45775748905364666,
+      "learning_rate": 4.842953663105158e-06,
+      "loss": 0.4919,
+      "step": 6156
+    },
+    {
+      "epoch": 2.41580637194369,
+      "grad_norm": 0.45387013696855416,
+      "learning_rate": 4.8428990045952075e-06,
+      "loss": 0.4933,
+      "step": 6157
+    },
+    {
+      "epoch": 2.416201531242282,
+      "grad_norm": 0.49320450897769486,
+      "learning_rate": 4.842844336883772e-06,
+      "loss": 0.5058,
+      "step": 6158
+    },
+    {
+      "epoch": 2.4165966905408744,
+      "grad_norm": 0.4594972163377635,
+      "learning_rate": 4.842789659971065e-06,
+      "loss": 0.4956,
+      "step": 6159
+    },
+    {
+      "epoch": 2.4169918498394667,
+      "grad_norm": 0.44799652445453403,
+      "learning_rate": 4.842734973857305e-06,
+      "loss": 0.5003,
+      "step": 6160
+    },
+    {
+      "epoch": 2.417387009138059,
+      "grad_norm": 0.4555779832663966,
+      "learning_rate": 4.842680278542704e-06,
+      "loss": 0.4988,
+      "step": 6161
+    },
+    {
+      "epoch": 2.417782168436651,
+      "grad_norm": 0.46273844976885875,
+      "learning_rate": 4.8426255740274776e-06,
+      "loss": 0.4881,
+      "step": 6162
+    },
+    {
+      "epoch": 2.4181773277352434,
+      "grad_norm": 0.44452340365885856,
+      "learning_rate": 4.84257086031184e-06,
+      "loss": 0.5114,
+      "step": 6163
+    },
+    {
+      "epoch": 2.4185724870338356,
+      "grad_norm": 0.45595741947905993,
+      "learning_rate": 4.842516137396007e-06,
+      "loss": 0.4976,
+      "step": 6164
+    },
+    {
+      "epoch": 2.418967646332428,
+      "grad_norm": 0.45112785524674676,
+      "learning_rate": 4.842461405280192e-06,
+      "loss": 0.5021,
+      "step": 6165
+    },
+    {
+      "epoch": 2.41936280563102,
+      "grad_norm": 0.4540156886101404,
+      "learning_rate": 4.842406663964612e-06,
+      "loss": 0.4895,
+      "step": 6166
+    },
+    {
+      "epoch": 2.4197579649296124,
+      "grad_norm": 0.4514068135997412,
+      "learning_rate": 4.842351913449481e-06,
+      "loss": 0.5139,
+      "step": 6167
+    },
+    {
+      "epoch": 2.4201531242282046,
+      "grad_norm": 0.4612849080685796,
+      "learning_rate": 4.842297153735014e-06,
+      "loss": 0.4983,
+      "step": 6168
+    },
+    {
+      "epoch": 2.420548283526797,
+      "grad_norm": 0.47339152666834156,
+      "learning_rate": 4.842242384821426e-06,
+      "loss": 0.4881,
+      "step": 6169
+    },
+    {
+      "epoch": 2.420943442825389,
+      "grad_norm": 0.46244595327643095,
+      "learning_rate": 4.842187606708932e-06,
+      "loss": 0.5034,
+      "step": 6170
+    },
+    {
+      "epoch": 2.4213386021239813,
+      "grad_norm": 0.4386337093959035,
+      "learning_rate": 4.8421328193977475e-06,
+      "loss": 0.4869,
+      "step": 6171
+    },
+    {
+      "epoch": 2.4217337614225736,
+      "grad_norm": 0.4718123255716426,
+      "learning_rate": 4.842078022888088e-06,
+      "loss": 0.4913,
+      "step": 6172
+    },
+    {
+      "epoch": 2.422128920721166,
+      "grad_norm": 0.4612919926951648,
+      "learning_rate": 4.8420232171801675e-06,
+      "loss": 0.5129,
+      "step": 6173
+    },
+    {
+      "epoch": 2.422524080019758,
+      "grad_norm": 0.4486144324992224,
+      "learning_rate": 4.841968402274202e-06,
+      "loss": 0.5147,
+      "step": 6174
+    },
+    {
+      "epoch": 2.4229192393183503,
+      "grad_norm": 0.46891228322282674,
+      "learning_rate": 4.841913578170407e-06,
+      "loss": 0.5172,
+      "step": 6175
+    },
+    {
+      "epoch": 2.4233143986169425,
+      "grad_norm": 0.44889237277545324,
+      "learning_rate": 4.841858744868998e-06,
+      "loss": 0.511,
+      "step": 6176
+    },
+    {
+      "epoch": 2.4237095579155348,
+      "grad_norm": 0.45184987151406364,
+      "learning_rate": 4.841803902370189e-06,
+      "loss": 0.4915,
+      "step": 6177
+    },
+    {
+      "epoch": 2.424104717214127,
+      "grad_norm": 0.45539717660361717,
+      "learning_rate": 4.841749050674196e-06,
+      "loss": 0.508,
+      "step": 6178
+    },
+    {
+      "epoch": 2.4244998765127193,
+      "grad_norm": 0.46183897777708616,
+      "learning_rate": 4.841694189781235e-06,
+      "loss": 0.5147,
+      "step": 6179
+    },
+    {
+      "epoch": 2.4248950358113115,
+      "grad_norm": 0.4596909143183604,
+      "learning_rate": 4.841639319691522e-06,
+      "loss": 0.506,
+      "step": 6180
+    },
+    {
+      "epoch": 2.4252901951099037,
+      "grad_norm": 0.4446113633044188,
+      "learning_rate": 4.841584440405271e-06,
+      "loss": 0.5064,
+      "step": 6181
+    },
+    {
+      "epoch": 2.425685354408496,
+      "grad_norm": 0.45743332618448895,
+      "learning_rate": 4.841529551922699e-06,
+      "loss": 0.504,
+      "step": 6182
+    },
+    {
+      "epoch": 2.426080513707088,
+      "grad_norm": 0.4816180094305801,
+      "learning_rate": 4.84147465424402e-06,
+      "loss": 0.4925,
+      "step": 6183
+    },
+    {
+      "epoch": 2.4264756730056805,
+      "grad_norm": 0.4543170154356122,
+      "learning_rate": 4.84141974736945e-06,
+      "loss": 0.5244,
+      "step": 6184
+    },
+    {
+      "epoch": 2.4268708323042727,
+      "grad_norm": 0.448689306635801,
+      "learning_rate": 4.841364831299206e-06,
+      "loss": 0.511,
+      "step": 6185
+    },
+    {
+      "epoch": 2.427265991602865,
+      "grad_norm": 0.45167259627359596,
+      "learning_rate": 4.8413099060335026e-06,
+      "loss": 0.5084,
+      "step": 6186
+    },
+    {
+      "epoch": 2.427661150901457,
+      "grad_norm": 0.4610385677246653,
+      "learning_rate": 4.841254971572555e-06,
+      "loss": 0.5097,
+      "step": 6187
+    },
+    {
+      "epoch": 2.4280563102000494,
+      "grad_norm": 0.4709029211327468,
+      "learning_rate": 4.84120002791658e-06,
+      "loss": 0.5125,
+      "step": 6188
+    },
+    {
+      "epoch": 2.4284514694986417,
+      "grad_norm": 0.4585385642897026,
+      "learning_rate": 4.841145075065793e-06,
+      "loss": 0.5255,
+      "step": 6189
+    },
+    {
+      "epoch": 2.428846628797234,
+      "grad_norm": 0.45124097243480904,
+      "learning_rate": 4.841090113020409e-06,
+      "loss": 0.4851,
+      "step": 6190
+    },
+    {
+      "epoch": 2.429241788095826,
+      "grad_norm": 0.4589307579275915,
+      "learning_rate": 4.841035141780645e-06,
+      "loss": 0.5237,
+      "step": 6191
+    },
+    {
+      "epoch": 2.4296369473944184,
+      "grad_norm": 0.4590381604206053,
+      "learning_rate": 4.840980161346717e-06,
+      "loss": 0.502,
+      "step": 6192
+    },
+    {
+      "epoch": 2.4300321066930106,
+      "grad_norm": 0.4575028892209658,
+      "learning_rate": 4.84092517171884e-06,
+      "loss": 0.5317,
+      "step": 6193
+    },
+    {
+      "epoch": 2.430427265991603,
+      "grad_norm": 0.4410480910569929,
+      "learning_rate": 4.840870172897231e-06,
+      "loss": 0.5079,
+      "step": 6194
+    },
+    {
+      "epoch": 2.430822425290195,
+      "grad_norm": 0.45485313052571497,
+      "learning_rate": 4.840815164882104e-06,
+      "loss": 0.5069,
+      "step": 6195
+    },
+    {
+      "epoch": 2.4312175845887873,
+      "grad_norm": 0.45726694565940096,
+      "learning_rate": 4.840760147673678e-06,
+      "loss": 0.4904,
+      "step": 6196
+    },
+    {
+      "epoch": 2.4316127438873796,
+      "grad_norm": 0.4591479178991767,
+      "learning_rate": 4.8407051212721664e-06,
+      "loss": 0.5311,
+      "step": 6197
+    },
+    {
+      "epoch": 2.432007903185972,
+      "grad_norm": 0.45199105003630935,
+      "learning_rate": 4.8406500856777875e-06,
+      "loss": 0.5167,
+      "step": 6198
+    },
+    {
+      "epoch": 2.432403062484564,
+      "grad_norm": 0.4629676193358175,
+      "learning_rate": 4.840595040890756e-06,
+      "loss": 0.5048,
+      "step": 6199
+    },
+    {
+      "epoch": 2.4327982217831563,
+      "grad_norm": 0.4476163591715802,
+      "learning_rate": 4.840539986911288e-06,
+      "loss": 0.4954,
+      "step": 6200
+    },
+    {
+      "epoch": 2.4331933810817485,
+      "grad_norm": 0.44798378662618715,
+      "learning_rate": 4.8404849237396005e-06,
+      "loss": 0.4933,
+      "step": 6201
+    },
+    {
+      "epoch": 2.433588540380341,
+      "grad_norm": 0.4571828487256416,
+      "learning_rate": 4.840429851375909e-06,
+      "loss": 0.5018,
+      "step": 6202
+    },
+    {
+      "epoch": 2.433983699678933,
+      "grad_norm": 0.4487082083135661,
+      "learning_rate": 4.840374769820432e-06,
+      "loss": 0.4855,
+      "step": 6203
+    },
+    {
+      "epoch": 2.4343788589775253,
+      "grad_norm": 0.5361808616989676,
+      "learning_rate": 4.840319679073382e-06,
+      "loss": 0.5127,
+      "step": 6204
+    },
+    {
+      "epoch": 2.4347740182761175,
+      "grad_norm": 0.4556095333028292,
+      "learning_rate": 4.840264579134978e-06,
+      "loss": 0.5055,
+      "step": 6205
+    },
+    {
+      "epoch": 2.4351691775747097,
+      "grad_norm": 0.4582035559570844,
+      "learning_rate": 4.840209470005436e-06,
+      "loss": 0.5032,
+      "step": 6206
+    },
+    {
+      "epoch": 2.435564336873302,
+      "grad_norm": 0.4848275765020931,
+      "learning_rate": 4.840154351684973e-06,
+      "loss": 0.515,
+      "step": 6207
+    },
+    {
+      "epoch": 2.4359594961718942,
+      "grad_norm": 0.46189716770494005,
+      "learning_rate": 4.840099224173803e-06,
+      "loss": 0.5158,
+      "step": 6208
+    },
+    {
+      "epoch": 2.4363546554704865,
+      "grad_norm": 0.4586463067687978,
+      "learning_rate": 4.840044087472145e-06,
+      "loss": 0.5169,
+      "step": 6209
+    },
+    {
+      "epoch": 2.4367498147690787,
+      "grad_norm": 0.45983904272371406,
+      "learning_rate": 4.839988941580216e-06,
+      "loss": 0.4927,
+      "step": 6210
+    },
+    {
+      "epoch": 2.437144974067671,
+      "grad_norm": 0.479677905098288,
+      "learning_rate": 4.83993378649823e-06,
+      "loss": 0.5466,
+      "step": 6211
+    },
+    {
+      "epoch": 2.437540133366263,
+      "grad_norm": 0.4605259736107779,
+      "learning_rate": 4.839878622226405e-06,
+      "loss": 0.5026,
+      "step": 6212
+    },
+    {
+      "epoch": 2.4379352926648554,
+      "grad_norm": 0.45622589883484194,
+      "learning_rate": 4.839823448764957e-06,
+      "loss": 0.5026,
+      "step": 6213
+    },
+    {
+      "epoch": 2.4383304519634477,
+      "grad_norm": 0.4780280716829919,
+      "learning_rate": 4.839768266114105e-06,
+      "loss": 0.515,
+      "step": 6214
+    },
+    {
+      "epoch": 2.43872561126204,
+      "grad_norm": 0.46441875870580496,
+      "learning_rate": 4.839713074274064e-06,
+      "loss": 0.508,
+      "step": 6215
+    },
+    {
+      "epoch": 2.439120770560632,
+      "grad_norm": 0.45047431718755854,
+      "learning_rate": 4.83965787324505e-06,
+      "loss": 0.5011,
+      "step": 6216
+    },
+    {
+      "epoch": 2.4395159298592244,
+      "grad_norm": 0.45440270281711137,
+      "learning_rate": 4.83960266302728e-06,
+      "loss": 0.5024,
+      "step": 6217
+    },
+    {
+      "epoch": 2.4399110891578166,
+      "grad_norm": 0.4638601787236139,
+      "learning_rate": 4.839547443620972e-06,
+      "loss": 0.5113,
+      "step": 6218
+    },
+    {
+      "epoch": 2.440306248456409,
+      "grad_norm": 0.4488674824729464,
+      "learning_rate": 4.839492215026342e-06,
+      "loss": 0.4858,
+      "step": 6219
+    },
+    {
+      "epoch": 2.440701407755001,
+      "grad_norm": 0.4536981413225659,
+      "learning_rate": 4.839436977243608e-06,
+      "loss": 0.4959,
+      "step": 6220
+    },
+    {
+      "epoch": 2.4410965670535933,
+      "grad_norm": 0.4614565162626784,
+      "learning_rate": 4.839381730272985e-06,
+      "loss": 0.5012,
+      "step": 6221
+    },
+    {
+      "epoch": 2.4414917263521856,
+      "grad_norm": 0.48492633372884375,
+      "learning_rate": 4.839326474114692e-06,
+      "loss": 0.5146,
+      "step": 6222
+    },
+    {
+      "epoch": 2.441886885650778,
+      "grad_norm": 0.4704218385973438,
+      "learning_rate": 4.839271208768945e-06,
+      "loss": 0.5129,
+      "step": 6223
+    },
+    {
+      "epoch": 2.44228204494937,
+      "grad_norm": 0.45965041302697135,
+      "learning_rate": 4.839215934235961e-06,
+      "loss": 0.5075,
+      "step": 6224
+    },
+    {
+      "epoch": 2.4426772042479623,
+      "grad_norm": 0.4561316353748442,
+      "learning_rate": 4.839160650515957e-06,
+      "loss": 0.4964,
+      "step": 6225
+    },
+    {
+      "epoch": 2.4430723635465545,
+      "grad_norm": 0.4590933327579659,
+      "learning_rate": 4.839105357609151e-06,
+      "loss": 0.5038,
+      "step": 6226
+    },
+    {
+      "epoch": 2.443467522845147,
+      "grad_norm": 0.47776100605193667,
+      "learning_rate": 4.839050055515759e-06,
+      "loss": 0.49,
+      "step": 6227
+    },
+    {
+      "epoch": 2.443862682143739,
+      "grad_norm": 0.43282187857296134,
+      "learning_rate": 4.838994744236e-06,
+      "loss": 0.4863,
+      "step": 6228
+    },
+    {
+      "epoch": 2.4442578414423313,
+      "grad_norm": 0.45445540357185477,
+      "learning_rate": 4.838939423770088e-06,
+      "loss": 0.5021,
+      "step": 6229
+    },
+    {
+      "epoch": 2.4446530007409235,
+      "grad_norm": 0.4562136501460368,
+      "learning_rate": 4.838884094118244e-06,
+      "loss": 0.5307,
+      "step": 6230
+    },
+    {
+      "epoch": 2.4450481600395157,
+      "grad_norm": 0.5480503669106829,
+      "learning_rate": 4.8388287552806825e-06,
+      "loss": 0.5152,
+      "step": 6231
+    },
+    {
+      "epoch": 2.445443319338108,
+      "grad_norm": 0.4651543841687929,
+      "learning_rate": 4.838773407257622e-06,
+      "loss": 0.5007,
+      "step": 6232
+    },
+    {
+      "epoch": 2.4458384786367002,
+      "grad_norm": 0.4685356376510576,
+      "learning_rate": 4.8387180500492795e-06,
+      "loss": 0.5048,
+      "step": 6233
+    },
+    {
+      "epoch": 2.4462336379352925,
+      "grad_norm": 0.4586342865694105,
+      "learning_rate": 4.838662683655872e-06,
+      "loss": 0.5004,
+      "step": 6234
+    },
+    {
+      "epoch": 2.4466287972338847,
+      "grad_norm": 0.44817924080411603,
+      "learning_rate": 4.83860730807762e-06,
+      "loss": 0.5024,
+      "step": 6235
+    },
+    {
+      "epoch": 2.447023956532477,
+      "grad_norm": 0.45665521330466075,
+      "learning_rate": 4.838551923314736e-06,
+      "loss": 0.5106,
+      "step": 6236
+    },
+    {
+      "epoch": 2.447419115831069,
+      "grad_norm": 0.46636023225981704,
+      "learning_rate": 4.838496529367441e-06,
+      "loss": 0.5154,
+      "step": 6237
+    },
+    {
+      "epoch": 2.4478142751296614,
+      "grad_norm": 0.47231613865652516,
+      "learning_rate": 4.8384411262359525e-06,
+      "loss": 0.5131,
+      "step": 6238
+    },
+    {
+      "epoch": 2.4482094344282537,
+      "grad_norm": 0.4613639527463871,
+      "learning_rate": 4.838385713920486e-06,
+      "loss": 0.5084,
+      "step": 6239
+    },
+    {
+      "epoch": 2.448604593726846,
+      "grad_norm": 0.4604238114892905,
+      "learning_rate": 4.838330292421262e-06,
+      "loss": 0.5065,
+      "step": 6240
+    },
+    {
+      "epoch": 2.448999753025438,
+      "grad_norm": 0.4633869738387279,
+      "learning_rate": 4.838274861738494e-06,
+      "loss": 0.5095,
+      "step": 6241
+    },
+    {
+      "epoch": 2.4493949123240304,
+      "grad_norm": 0.47977255402722213,
+      "learning_rate": 4.838219421872405e-06,
+      "loss": 0.5242,
+      "step": 6242
+    },
+    {
+      "epoch": 2.4497900716226226,
+      "grad_norm": 0.4649910129362038,
+      "learning_rate": 4.8381639728232075e-06,
+      "loss": 0.5007,
+      "step": 6243
+    },
+    {
+      "epoch": 2.450185230921215,
+      "grad_norm": 0.44567230911849703,
+      "learning_rate": 4.838108514591124e-06,
+      "loss": 0.4886,
+      "step": 6244
+    },
+    {
+      "epoch": 2.450580390219807,
+      "grad_norm": 0.4802947747613605,
+      "learning_rate": 4.838053047176368e-06,
+      "loss": 0.5182,
+      "step": 6245
+    },
+    {
+      "epoch": 2.4509755495184,
+      "grad_norm": 0.4529418893441322,
+      "learning_rate": 4.83799757057916e-06,
+      "loss": 0.4978,
+      "step": 6246
+    },
+    {
+      "epoch": 2.451370708816992,
+      "grad_norm": 0.4707140210320241,
+      "learning_rate": 4.837942084799717e-06,
+      "loss": 0.5148,
+      "step": 6247
+    },
+    {
+      "epoch": 2.4517658681155843,
+      "grad_norm": 0.4514623045887543,
+      "learning_rate": 4.837886589838259e-06,
+      "loss": 0.5095,
+      "step": 6248
+    },
+    {
+      "epoch": 2.4521610274141765,
+      "grad_norm": 0.4760633546771482,
+      "learning_rate": 4.837831085695e-06,
+      "loss": 0.5098,
+      "step": 6249
+    },
+    {
+      "epoch": 2.4525561867127688,
+      "grad_norm": 0.4602919310882103,
+      "learning_rate": 4.8377755723701614e-06,
+      "loss": 0.5025,
+      "step": 6250
+    },
+    {
+      "epoch": 2.452951346011361,
+      "grad_norm": 0.5205070905111004,
+      "learning_rate": 4.837720049863959e-06,
+      "loss": 0.5558,
+      "step": 6251
+    },
+    {
+      "epoch": 2.4533465053099532,
+      "grad_norm": 0.44480305897623773,
+      "learning_rate": 4.837664518176613e-06,
+      "loss": 0.4903,
+      "step": 6252
+    },
+    {
+      "epoch": 2.4537416646085455,
+      "grad_norm": 0.45504363171585244,
+      "learning_rate": 4.837608977308339e-06,
+      "loss": 0.5183,
+      "step": 6253
+    },
+    {
+      "epoch": 2.4541368239071377,
+      "grad_norm": 0.44828627366707774,
+      "learning_rate": 4.837553427259356e-06,
+      "loss": 0.5075,
+      "step": 6254
+    },
+    {
+      "epoch": 2.45453198320573,
+      "grad_norm": 0.4429586869046653,
+      "learning_rate": 4.837497868029884e-06,
+      "loss": 0.495,
+      "step": 6255
+    },
+    {
+      "epoch": 2.454927142504322,
+      "grad_norm": 0.4580942265610526,
+      "learning_rate": 4.837442299620139e-06,
+      "loss": 0.5026,
+      "step": 6256
+    },
+    {
+      "epoch": 2.4553223018029144,
+      "grad_norm": 0.46738087782305204,
+      "learning_rate": 4.83738672203034e-06,
+      "loss": 0.5316,
+      "step": 6257
+    },
+    {
+      "epoch": 2.4557174611015067,
+      "grad_norm": 0.47534561406612363,
+      "learning_rate": 4.837331135260705e-06,
+      "loss": 0.529,
+      "step": 6258
+    },
+    {
+      "epoch": 2.456112620400099,
+      "grad_norm": 0.45632596134859843,
+      "learning_rate": 4.837275539311454e-06,
+      "loss": 0.5025,
+      "step": 6259
+    },
+    {
+      "epoch": 2.456507779698691,
+      "grad_norm": 0.43916671865176493,
+      "learning_rate": 4.837219934182803e-06,
+      "loss": 0.4894,
+      "step": 6260
+    },
+    {
+      "epoch": 2.4569029389972834,
+      "grad_norm": 0.4685573051031802,
+      "learning_rate": 4.837164319874972e-06,
+      "loss": 0.5134,
+      "step": 6261
+    },
+    {
+      "epoch": 2.4572980982958756,
+      "grad_norm": 0.45202371145633197,
+      "learning_rate": 4.8371086963881774e-06,
+      "loss": 0.4923,
+      "step": 6262
+    },
+    {
+      "epoch": 2.457693257594468,
+      "grad_norm": 0.46730026620993986,
+      "learning_rate": 4.83705306372264e-06,
+      "loss": 0.5016,
+      "step": 6263
+    },
+    {
+      "epoch": 2.45808841689306,
+      "grad_norm": 0.4626170175822174,
+      "learning_rate": 4.836997421878577e-06,
+      "loss": 0.5234,
+      "step": 6264
+    },
+    {
+      "epoch": 2.4584835761916524,
+      "grad_norm": 0.44361760739699646,
+      "learning_rate": 4.836941770856207e-06,
+      "loss": 0.499,
+      "step": 6265
+    },
+    {
+      "epoch": 2.4588787354902446,
+      "grad_norm": 0.45702240750866957,
+      "learning_rate": 4.83688611065575e-06,
+      "loss": 0.5058,
+      "step": 6266
+    },
+    {
+      "epoch": 2.459273894788837,
+      "grad_norm": 0.47299501346012857,
+      "learning_rate": 4.836830441277422e-06,
+      "loss": 0.5066,
+      "step": 6267
+    },
+    {
+      "epoch": 2.459669054087429,
+      "grad_norm": 0.4727679722857182,
+      "learning_rate": 4.836774762721443e-06,
+      "loss": 0.5219,
+      "step": 6268
+    },
+    {
+      "epoch": 2.4600642133860213,
+      "grad_norm": 0.4561162349721409,
+      "learning_rate": 4.836719074988033e-06,
+      "loss": 0.5123,
+      "step": 6269
+    },
+    {
+      "epoch": 2.4604593726846136,
+      "grad_norm": 0.46185475580783847,
+      "learning_rate": 4.836663378077408e-06,
+      "loss": 0.5057,
+      "step": 6270
+    },
+    {
+      "epoch": 2.460854531983206,
+      "grad_norm": 0.4589633018472339,
+      "learning_rate": 4.836607671989789e-06,
+      "loss": 0.491,
+      "step": 6271
+    },
+    {
+      "epoch": 2.461249691281798,
+      "grad_norm": 0.4674016116064247,
+      "learning_rate": 4.836551956725394e-06,
+      "loss": 0.505,
+      "step": 6272
+    },
+    {
+      "epoch": 2.4616448505803903,
+      "grad_norm": 0.44019704083575656,
+      "learning_rate": 4.836496232284441e-06,
+      "loss": 0.4899,
+      "step": 6273
+    },
+    {
+      "epoch": 2.4620400098789825,
+      "grad_norm": 0.4560950335962957,
+      "learning_rate": 4.8364404986671495e-06,
+      "loss": 0.5086,
+      "step": 6274
+    },
+    {
+      "epoch": 2.4624351691775748,
+      "grad_norm": 0.45987663956371605,
+      "learning_rate": 4.83638475587374e-06,
+      "loss": 0.5023,
+      "step": 6275
+    },
+    {
+      "epoch": 2.462830328476167,
+      "grad_norm": 0.45218362910357657,
+      "learning_rate": 4.836329003904429e-06,
+      "loss": 0.5049,
+      "step": 6276
+    },
+    {
+      "epoch": 2.4632254877747592,
+      "grad_norm": 0.46082337545098706,
+      "learning_rate": 4.836273242759436e-06,
+      "loss": 0.4906,
+      "step": 6277
+    },
+    {
+      "epoch": 2.4636206470733515,
+      "grad_norm": 0.4484085796656676,
+      "learning_rate": 4.83621747243898e-06,
+      "loss": 0.5003,
+      "step": 6278
+    },
+    {
+      "epoch": 2.4640158063719437,
+      "grad_norm": 0.4561063177663292,
+      "learning_rate": 4.836161692943282e-06,
+      "loss": 0.4872,
+      "step": 6279
+    },
+    {
+      "epoch": 2.464410965670536,
+      "grad_norm": 0.4635792252049459,
+      "learning_rate": 4.836105904272558e-06,
+      "loss": 0.4891,
+      "step": 6280
+    },
+    {
+      "epoch": 2.464806124969128,
+      "grad_norm": 0.44874718083152354,
+      "learning_rate": 4.836050106427029e-06,
+      "loss": 0.504,
+      "step": 6281
+    },
+    {
+      "epoch": 2.4652012842677204,
+      "grad_norm": 0.4542288616082571,
+      "learning_rate": 4.835994299406914e-06,
+      "loss": 0.5117,
+      "step": 6282
+    },
+    {
+      "epoch": 2.4655964435663127,
+      "grad_norm": 0.46392560474828354,
+      "learning_rate": 4.835938483212431e-06,
+      "loss": 0.5032,
+      "step": 6283
+    },
+    {
+      "epoch": 2.465991602864905,
+      "grad_norm": 0.461032498595409,
+      "learning_rate": 4.835882657843801e-06,
+      "loss": 0.5041,
+      "step": 6284
+    },
+    {
+      "epoch": 2.466386762163497,
+      "grad_norm": 0.43775177363788437,
+      "learning_rate": 4.835826823301242e-06,
+      "loss": 0.5057,
+      "step": 6285
+    },
+    {
+      "epoch": 2.4667819214620894,
+      "grad_norm": 0.45447705767758345,
+      "learning_rate": 4.835770979584974e-06,
+      "loss": 0.5059,
+      "step": 6286
+    },
+    {
+      "epoch": 2.4671770807606817,
+      "grad_norm": 0.4820043582122839,
+      "learning_rate": 4.835715126695216e-06,
+      "loss": 0.5025,
+      "step": 6287
+    },
+    {
+      "epoch": 2.467572240059274,
+      "grad_norm": 0.45364882204768436,
+      "learning_rate": 4.835659264632186e-06,
+      "loss": 0.5121,
+      "step": 6288
+    },
+    {
+      "epoch": 2.467967399357866,
+      "grad_norm": 0.44392647087381,
+      "learning_rate": 4.835603393396106e-06,
+      "loss": 0.4973,
+      "step": 6289
+    },
+    {
+      "epoch": 2.4683625586564584,
+      "grad_norm": 0.4702335158316592,
+      "learning_rate": 4.835547512987194e-06,
+      "loss": 0.5056,
+      "step": 6290
+    },
+    {
+      "epoch": 2.4687577179550506,
+      "grad_norm": 0.46379114054816933,
+      "learning_rate": 4.835491623405669e-06,
+      "loss": 0.5062,
+      "step": 6291
+    },
+    {
+      "epoch": 2.469152877253643,
+      "grad_norm": 0.46254378434485627,
+      "learning_rate": 4.835435724651753e-06,
+      "loss": 0.5034,
+      "step": 6292
+    },
+    {
+      "epoch": 2.469548036552235,
+      "grad_norm": 0.4435217656925511,
+      "learning_rate": 4.8353798167256615e-06,
+      "loss": 0.5178,
+      "step": 6293
+    },
+    {
+      "epoch": 2.4699431958508273,
+      "grad_norm": 0.44400526861468553,
+      "learning_rate": 4.835323899627616e-06,
+      "loss": 0.4828,
+      "step": 6294
+    },
+    {
+      "epoch": 2.4703383551494196,
+      "grad_norm": 0.4661317214500771,
+      "learning_rate": 4.835267973357837e-06,
+      "loss": 0.5096,
+      "step": 6295
+    },
+    {
+      "epoch": 2.470733514448012,
+      "grad_norm": 0.4599470856465283,
+      "learning_rate": 4.835212037916545e-06,
+      "loss": 0.5178,
+      "step": 6296
+    },
+    {
+      "epoch": 2.471128673746604,
+      "grad_norm": 0.44508708208240416,
+      "learning_rate": 4.835156093303956e-06,
+      "loss": 0.5011,
+      "step": 6297
+    },
+    {
+      "epoch": 2.4715238330451963,
+      "grad_norm": 0.45991594887386883,
+      "learning_rate": 4.835100139520292e-06,
+      "loss": 0.5066,
+      "step": 6298
+    },
+    {
+      "epoch": 2.4719189923437885,
+      "grad_norm": 0.4473801495463778,
+      "learning_rate": 4.8350441765657736e-06,
+      "loss": 0.5124,
+      "step": 6299
+    },
+    {
+      "epoch": 2.4723141516423808,
+      "grad_norm": 0.45643508613471995,
+      "learning_rate": 4.834988204440619e-06,
+      "loss": 0.4926,
+      "step": 6300
+    },
+    {
+      "epoch": 2.472709310940973,
+      "grad_norm": 0.4504940699709523,
+      "learning_rate": 4.834932223145049e-06,
+      "loss": 0.5032,
+      "step": 6301
+    },
+    {
+      "epoch": 2.4731044702395653,
+      "grad_norm": 0.4426038077719436,
+      "learning_rate": 4.834876232679283e-06,
+      "loss": 0.5104,
+      "step": 6302
+    },
+    {
+      "epoch": 2.4734996295381575,
+      "grad_norm": 0.46223284198355835,
+      "learning_rate": 4.83482023304354e-06,
+      "loss": 0.5081,
+      "step": 6303
+    },
+    {
+      "epoch": 2.4738947888367497,
+      "grad_norm": 0.4770885110104904,
+      "learning_rate": 4.834764224238042e-06,
+      "loss": 0.5098,
+      "step": 6304
+    },
+    {
+      "epoch": 2.474289948135342,
+      "grad_norm": 0.4732407782266299,
+      "learning_rate": 4.834708206263008e-06,
+      "loss": 0.4808,
+      "step": 6305
+    },
+    {
+      "epoch": 2.474685107433934,
+      "grad_norm": 0.4613411578962985,
+      "learning_rate": 4.834652179118657e-06,
+      "loss": 0.5215,
+      "step": 6306
+    },
+    {
+      "epoch": 2.4750802667325265,
+      "grad_norm": 0.4626350011702521,
+      "learning_rate": 4.83459614280521e-06,
+      "loss": 0.5143,
+      "step": 6307
+    },
+    {
+      "epoch": 2.4754754260311187,
+      "grad_norm": 0.4492080049375109,
+      "learning_rate": 4.834540097322888e-06,
+      "loss": 0.5151,
+      "step": 6308
+    },
+    {
+      "epoch": 2.475870585329711,
+      "grad_norm": 0.46276098431405666,
+      "learning_rate": 4.834484042671909e-06,
+      "loss": 0.5107,
+      "step": 6309
+    },
+    {
+      "epoch": 2.476265744628303,
+      "grad_norm": 0.4600080634693739,
+      "learning_rate": 4.834427978852495e-06,
+      "loss": 0.5183,
+      "step": 6310
+    },
+    {
+      "epoch": 2.4766609039268954,
+      "grad_norm": 0.4560521377408553,
+      "learning_rate": 4.834371905864865e-06,
+      "loss": 0.515,
+      "step": 6311
+    },
+    {
+      "epoch": 2.4770560632254877,
+      "grad_norm": 0.4680722643265072,
+      "learning_rate": 4.83431582370924e-06,
+      "loss": 0.5268,
+      "step": 6312
+    },
+    {
+      "epoch": 2.47745122252408,
+      "grad_norm": 0.45788403308191006,
+      "learning_rate": 4.83425973238584e-06,
+      "loss": 0.5185,
+      "step": 6313
+    },
+    {
+      "epoch": 2.477846381822672,
+      "grad_norm": 0.4489353107357901,
+      "learning_rate": 4.834203631894885e-06,
+      "loss": 0.5245,
+      "step": 6314
+    },
+    {
+      "epoch": 2.4782415411212644,
+      "grad_norm": 0.4769728044328352,
+      "learning_rate": 4.834147522236595e-06,
+      "loss": 0.5034,
+      "step": 6315
+    },
+    {
+      "epoch": 2.4786367004198566,
+      "grad_norm": 0.4531931819815472,
+      "learning_rate": 4.8340914034111916e-06,
+      "loss": 0.4785,
+      "step": 6316
+    },
+    {
+      "epoch": 2.479031859718449,
+      "grad_norm": 0.436765563760745,
+      "learning_rate": 4.834035275418895e-06,
+      "loss": 0.5044,
+      "step": 6317
+    },
+    {
+      "epoch": 2.479427019017041,
+      "grad_norm": 0.45553402918025715,
+      "learning_rate": 4.833979138259923e-06,
+      "loss": 0.5099,
+      "step": 6318
+    },
+    {
+      "epoch": 2.4798221783156333,
+      "grad_norm": 0.45738260128460106,
+      "learning_rate": 4.8339229919345e-06,
+      "loss": 0.5028,
+      "step": 6319
+    },
+    {
+      "epoch": 2.4802173376142256,
+      "grad_norm": 0.45623700990486077,
+      "learning_rate": 4.833866836442844e-06,
+      "loss": 0.5015,
+      "step": 6320
+    },
+    {
+      "epoch": 2.480612496912818,
+      "grad_norm": 0.44512331672110883,
+      "learning_rate": 4.833810671785177e-06,
+      "loss": 0.5099,
+      "step": 6321
+    },
+    {
+      "epoch": 2.48100765621141,
+      "grad_norm": 0.4552888276049071,
+      "learning_rate": 4.833754497961719e-06,
+      "loss": 0.5012,
+      "step": 6322
+    },
+    {
+      "epoch": 2.4814028155100023,
+      "grad_norm": 0.46342152696551503,
+      "learning_rate": 4.83369831497269e-06,
+      "loss": 0.5272,
+      "step": 6323
+    },
+    {
+      "epoch": 2.4817979748085945,
+      "grad_norm": 0.44267854664405953,
+      "learning_rate": 4.833642122818311e-06,
+      "loss": 0.4853,
+      "step": 6324
+    },
+    {
+      "epoch": 2.482193134107187,
+      "grad_norm": 0.43631582113931716,
+      "learning_rate": 4.833585921498802e-06,
+      "loss": 0.4987,
+      "step": 6325
+    },
+    {
+      "epoch": 2.482588293405779,
+      "grad_norm": 0.44520693389382443,
+      "learning_rate": 4.8335297110143854e-06,
+      "loss": 0.5021,
+      "step": 6326
+    },
+    {
+      "epoch": 2.4829834527043713,
+      "grad_norm": 0.4611485355941225,
+      "learning_rate": 4.833473491365281e-06,
+      "loss": 0.5156,
+      "step": 6327
+    },
+    {
+      "epoch": 2.4833786120029635,
+      "grad_norm": 0.45289421087393966,
+      "learning_rate": 4.833417262551711e-06,
+      "loss": 0.5095,
+      "step": 6328
+    },
+    {
+      "epoch": 2.483773771301556,
+      "grad_norm": 0.4566101862712613,
+      "learning_rate": 4.833361024573893e-06,
+      "loss": 0.5237,
+      "step": 6329
+    },
+    {
+      "epoch": 2.4841689306001484,
+      "grad_norm": 0.4564858320090929,
+      "learning_rate": 4.833304777432051e-06,
+      "loss": 0.492,
+      "step": 6330
+    },
+    {
+      "epoch": 2.4845640898987407,
+      "grad_norm": 0.4555944378026437,
+      "learning_rate": 4.8332485211264035e-06,
+      "loss": 0.5221,
+      "step": 6331
+    },
+    {
+      "epoch": 2.484959249197333,
+      "grad_norm": 0.4640570525721282,
+      "learning_rate": 4.833192255657173e-06,
+      "loss": 0.5265,
+      "step": 6332
+    },
+    {
+      "epoch": 2.485354408495925,
+      "grad_norm": 0.4450100165471149,
+      "learning_rate": 4.833135981024581e-06,
+      "loss": 0.4968,
+      "step": 6333
+    },
+    {
+      "epoch": 2.4857495677945174,
+      "grad_norm": 0.44268155087230027,
+      "learning_rate": 4.833079697228847e-06,
+      "loss": 0.4893,
+      "step": 6334
+    },
+    {
+      "epoch": 2.4861447270931096,
+      "grad_norm": 0.446170036400285,
+      "learning_rate": 4.833023404270193e-06,
+      "loss": 0.4945,
+      "step": 6335
+    },
+    {
+      "epoch": 2.486539886391702,
+      "grad_norm": 0.46363170302080026,
+      "learning_rate": 4.8329671021488385e-06,
+      "loss": 0.5106,
+      "step": 6336
+    },
+    {
+      "epoch": 2.486935045690294,
+      "grad_norm": 0.46125075416833533,
+      "learning_rate": 4.832910790865007e-06,
+      "loss": 0.5099,
+      "step": 6337
+    },
+    {
+      "epoch": 2.4873302049888864,
+      "grad_norm": 0.4666268360674439,
+      "learning_rate": 4.832854470418918e-06,
+      "loss": 0.4968,
+      "step": 6338
+    },
+    {
+      "epoch": 2.4877253642874786,
+      "grad_norm": 0.45259995470398834,
+      "learning_rate": 4.8327981408107945e-06,
+      "loss": 0.5231,
+      "step": 6339
+    },
+    {
+      "epoch": 2.488120523586071,
+      "grad_norm": 0.45257495176751394,
+      "learning_rate": 4.832741802040856e-06,
+      "loss": 0.51,
+      "step": 6340
+    },
+    {
+      "epoch": 2.488515682884663,
+      "grad_norm": 0.4702986428108045,
+      "learning_rate": 4.8326854541093235e-06,
+      "loss": 0.511,
+      "step": 6341
+    },
+    {
+      "epoch": 2.4889108421832553,
+      "grad_norm": 0.45287635582005137,
+      "learning_rate": 4.832629097016419e-06,
+      "loss": 0.5194,
+      "step": 6342
+    },
+    {
+      "epoch": 2.4893060014818476,
+      "grad_norm": 0.44097483735488313,
+      "learning_rate": 4.832572730762364e-06,
+      "loss": 0.4886,
+      "step": 6343
+    },
+    {
+      "epoch": 2.48970116078044,
+      "grad_norm": 0.4963761696743815,
+      "learning_rate": 4.83251635534738e-06,
+      "loss": 0.5144,
+      "step": 6344
+    },
+    {
+      "epoch": 2.490096320079032,
+      "grad_norm": 0.4721389393948357,
+      "learning_rate": 4.832459970771688e-06,
+      "loss": 0.5196,
+      "step": 6345
+    },
+    {
+      "epoch": 2.4904914793776243,
+      "grad_norm": 0.4607493050777102,
+      "learning_rate": 4.83240357703551e-06,
+      "loss": 0.5065,
+      "step": 6346
+    },
+    {
+      "epoch": 2.4908866386762165,
+      "grad_norm": 0.4603650627094426,
+      "learning_rate": 4.8323471741390656e-06,
+      "loss": 0.4982,
+      "step": 6347
+    },
+    {
+      "epoch": 2.4912817979748088,
+      "grad_norm": 0.4620527083371952,
+      "learning_rate": 4.832290762082579e-06,
+      "loss": 0.5114,
+      "step": 6348
+    },
+    {
+      "epoch": 2.491676957273401,
+      "grad_norm": 0.46945952654863243,
+      "learning_rate": 4.8322343408662705e-06,
+      "loss": 0.5093,
+      "step": 6349
+    },
+    {
+      "epoch": 2.4920721165719932,
+      "grad_norm": 0.45597389461472576,
+      "learning_rate": 4.8321779104903616e-06,
+      "loss": 0.5103,
+      "step": 6350
+    },
+    {
+      "epoch": 2.4924672758705855,
+      "grad_norm": 0.4517530087892166,
+      "learning_rate": 4.832121470955074e-06,
+      "loss": 0.504,
+      "step": 6351
+    },
+    {
+      "epoch": 2.4928624351691777,
+      "grad_norm": 0.44430291926385046,
+      "learning_rate": 4.832065022260629e-06,
+      "loss": 0.4967,
+      "step": 6352
+    },
+    {
+      "epoch": 2.49325759446777,
+      "grad_norm": 0.46115093494647114,
+      "learning_rate": 4.83200856440725e-06,
+      "loss": 0.5082,
+      "step": 6353
+    },
+    {
+      "epoch": 2.493652753766362,
+      "grad_norm": 0.45214398265733147,
+      "learning_rate": 4.831952097395156e-06,
+      "loss": 0.503,
+      "step": 6354
+    },
+    {
+      "epoch": 2.4940479130649544,
+      "grad_norm": 0.4607263777785255,
+      "learning_rate": 4.831895621224571e-06,
+      "loss": 0.5115,
+      "step": 6355
+    },
+    {
+      "epoch": 2.4944430723635467,
+      "grad_norm": 0.45430159930327796,
+      "learning_rate": 4.8318391358957156e-06,
+      "loss": 0.4955,
+      "step": 6356
+    },
+    {
+      "epoch": 2.494838231662139,
+      "grad_norm": 0.45283670210193705,
+      "learning_rate": 4.831782641408812e-06,
+      "loss": 0.5071,
+      "step": 6357
+    },
+    {
+      "epoch": 2.495233390960731,
+      "grad_norm": 0.4598695429896044,
+      "learning_rate": 4.831726137764082e-06,
+      "loss": 0.518,
+      "step": 6358
+    },
+    {
+      "epoch": 2.4956285502593234,
+      "grad_norm": 0.4399372220031,
+      "learning_rate": 4.831669624961748e-06,
+      "loss": 0.5051,
+      "step": 6359
+    },
+    {
+      "epoch": 2.4960237095579156,
+      "grad_norm": 0.467578857452426,
+      "learning_rate": 4.831613103002032e-06,
+      "loss": 0.4944,
+      "step": 6360
+    },
+    {
+      "epoch": 2.496418868856508,
+      "grad_norm": 0.46610183550354434,
+      "learning_rate": 4.831556571885155e-06,
+      "loss": 0.5046,
+      "step": 6361
+    },
+    {
+      "epoch": 2.4968140281551,
+      "grad_norm": 0.4589963237813121,
+      "learning_rate": 4.831500031611339e-06,
+      "loss": 0.5007,
+      "step": 6362
+    },
+    {
+      "epoch": 2.4972091874536924,
+      "grad_norm": 0.45341734760729985,
+      "learning_rate": 4.831443482180808e-06,
+      "loss": 0.5097,
+      "step": 6363
+    },
+    {
+      "epoch": 2.4976043467522846,
+      "grad_norm": 0.4737045186230201,
+      "learning_rate": 4.831386923593781e-06,
+      "loss": 0.5105,
+      "step": 6364
+    },
+    {
+      "epoch": 2.497999506050877,
+      "grad_norm": 0.4641942011706654,
+      "learning_rate": 4.831330355850484e-06,
+      "loss": 0.5005,
+      "step": 6365
+    },
+    {
+      "epoch": 2.498394665349469,
+      "grad_norm": 0.4856421198337944,
+      "learning_rate": 4.831273778951135e-06,
+      "loss": 0.5245,
+      "step": 6366
+    },
+    {
+      "epoch": 2.4987898246480613,
+      "grad_norm": 0.44305020823239166,
+      "learning_rate": 4.831217192895959e-06,
+      "loss": 0.4882,
+      "step": 6367
+    },
+    {
+      "epoch": 2.4991849839466536,
+      "grad_norm": 0.4582781517686158,
+      "learning_rate": 4.831160597685178e-06,
+      "loss": 0.4903,
+      "step": 6368
+    },
+    {
+      "epoch": 2.499580143245246,
+      "grad_norm": 0.4525149145967303,
+      "learning_rate": 4.8311039933190136e-06,
+      "loss": 0.4834,
+      "step": 6369
+    },
+    {
+      "epoch": 2.499975302543838,
+      "grad_norm": 0.4575043422893562,
+      "learning_rate": 4.831047379797687e-06,
+      "loss": 0.4897,
+      "step": 6370
+    },
+    {
+      "epoch": 2.5003704618424303,
+      "grad_norm": 0.46250674104664735,
+      "learning_rate": 4.830990757121424e-06,
+      "loss": 0.506,
+      "step": 6371
+    },
+    {
+      "epoch": 2.5007656211410225,
+      "grad_norm": 0.4619201958339098,
+      "learning_rate": 4.830934125290443e-06,
+      "loss": 0.4947,
+      "step": 6372
+    },
+    {
+      "epoch": 2.5011607804396148,
+      "grad_norm": 0.4603940568427686,
+      "learning_rate": 4.830877484304969e-06,
+      "loss": 0.5052,
+      "step": 6373
+    },
+    {
+      "epoch": 2.501555939738207,
+      "grad_norm": 0.4671784400513492,
+      "learning_rate": 4.830820834165223e-06,
+      "loss": 0.4942,
+      "step": 6374
+    },
+    {
+      "epoch": 2.5019510990367992,
+      "grad_norm": 0.46267953291708924,
+      "learning_rate": 4.830764174871429e-06,
+      "loss": 0.5049,
+      "step": 6375
+    },
+    {
+      "epoch": 2.5023462583353915,
+      "grad_norm": 0.4670573153772785,
+      "learning_rate": 4.830707506423807e-06,
+      "loss": 0.5071,
+      "step": 6376
+    },
+    {
+      "epoch": 2.5027414176339837,
+      "grad_norm": 0.448679724311928,
+      "learning_rate": 4.830650828822583e-06,
+      "loss": 0.5036,
+      "step": 6377
+    },
+    {
+      "epoch": 2.503136576932576,
+      "grad_norm": 0.4638582201810059,
+      "learning_rate": 4.830594142067977e-06,
+      "loss": 0.4958,
+      "step": 6378
+    },
+    {
+      "epoch": 2.503531736231168,
+      "grad_norm": 0.46498455456100324,
+      "learning_rate": 4.8305374461602115e-06,
+      "loss": 0.4994,
+      "step": 6379
+    },
+    {
+      "epoch": 2.5039268955297604,
+      "grad_norm": 0.47076449596566666,
+      "learning_rate": 4.830480741099511e-06,
+      "loss": 0.5077,
+      "step": 6380
+    },
+    {
+      "epoch": 2.5043220548283527,
+      "grad_norm": 0.4620871607194632,
+      "learning_rate": 4.830424026886098e-06,
+      "loss": 0.5003,
+      "step": 6381
+    },
+    {
+      "epoch": 2.504717214126945,
+      "grad_norm": 0.46348419234538263,
+      "learning_rate": 4.8303673035201935e-06,
+      "loss": 0.5286,
+      "step": 6382
+    },
+    {
+      "epoch": 2.505112373425537,
+      "grad_norm": 0.45149330232898727,
+      "learning_rate": 4.830310571002022e-06,
+      "loss": 0.5099,
+      "step": 6383
+    },
+    {
+      "epoch": 2.5055075327241294,
+      "grad_norm": 0.4375735232605822,
+      "learning_rate": 4.830253829331805e-06,
+      "loss": 0.4983,
+      "step": 6384
+    },
+    {
+      "epoch": 2.5059026920227216,
+      "grad_norm": 0.4469286754515496,
+      "learning_rate": 4.830197078509766e-06,
+      "loss": 0.4992,
+      "step": 6385
+    },
+    {
+      "epoch": 2.506297851321314,
+      "grad_norm": 0.46197178445812176,
+      "learning_rate": 4.830140318536128e-06,
+      "loss": 0.497,
+      "step": 6386
+    },
+    {
+      "epoch": 2.506693010619906,
+      "grad_norm": 0.4619654942592018,
+      "learning_rate": 4.830083549411114e-06,
+      "loss": 0.4983,
+      "step": 6387
+    },
+    {
+      "epoch": 2.5070881699184984,
+      "grad_norm": 0.4522982728309289,
+      "learning_rate": 4.830026771134947e-06,
+      "loss": 0.4994,
+      "step": 6388
+    },
+    {
+      "epoch": 2.5074833292170906,
+      "grad_norm": 0.4628104450392865,
+      "learning_rate": 4.82996998370785e-06,
+      "loss": 0.5056,
+      "step": 6389
+    },
+    {
+      "epoch": 2.507878488515683,
+      "grad_norm": 0.45715011969492436,
+      "learning_rate": 4.829913187130044e-06,
+      "loss": 0.5084,
+      "step": 6390
+    },
+    {
+      "epoch": 2.508273647814275,
+      "grad_norm": 0.4620237998725422,
+      "learning_rate": 4.8298563814017555e-06,
+      "loss": 0.5114,
+      "step": 6391
+    },
+    {
+      "epoch": 2.5086688071128673,
+      "grad_norm": 0.4632610101741824,
+      "learning_rate": 4.829799566523205e-06,
+      "loss": 0.5141,
+      "step": 6392
+    },
+    {
+      "epoch": 2.5090639664114596,
+      "grad_norm": 0.4641510104801741,
+      "learning_rate": 4.829742742494616e-06,
+      "loss": 0.5042,
+      "step": 6393
+    },
+    {
+      "epoch": 2.509459125710052,
+      "grad_norm": 0.4628296152429653,
+      "learning_rate": 4.829685909316214e-06,
+      "loss": 0.4989,
+      "step": 6394
+    },
+    {
+      "epoch": 2.509854285008644,
+      "grad_norm": 0.47161142330182587,
+      "learning_rate": 4.829629066988219e-06,
+      "loss": 0.4889,
+      "step": 6395
+    },
+    {
+      "epoch": 2.5102494443072363,
+      "grad_norm": 0.4703058506566186,
+      "learning_rate": 4.829572215510856e-06,
+      "loss": 0.5248,
+      "step": 6396
+    },
+    {
+      "epoch": 2.5106446036058285,
+      "grad_norm": 0.4487390067504666,
+      "learning_rate": 4.829515354884348e-06,
+      "loss": 0.4989,
+      "step": 6397
+    },
+    {
+      "epoch": 2.5110397629044208,
+      "grad_norm": 0.46578439069427224,
+      "learning_rate": 4.829458485108918e-06,
+      "loss": 0.5157,
+      "step": 6398
+    },
+    {
+      "epoch": 2.511434922203013,
+      "grad_norm": 0.4589142880562439,
+      "learning_rate": 4.8294016061847895e-06,
+      "loss": 0.5113,
+      "step": 6399
+    },
+    {
+      "epoch": 2.5118300815016052,
+      "grad_norm": 0.43436790417000837,
+      "learning_rate": 4.829344718112186e-06,
+      "loss": 0.4928,
+      "step": 6400
+    },
+    {
+      "epoch": 2.5122252408001975,
+      "grad_norm": 0.454548256257017,
+      "learning_rate": 4.829287820891332e-06,
+      "loss": 0.4983,
+      "step": 6401
+    },
+    {
+      "epoch": 2.5126204000987897,
+      "grad_norm": 0.44998917234532554,
+      "learning_rate": 4.829230914522449e-06,
+      "loss": 0.502,
+      "step": 6402
+    },
+    {
+      "epoch": 2.513015559397382,
+      "grad_norm": 0.4419759844210292,
+      "learning_rate": 4.82917399900576e-06,
+      "loss": 0.4961,
+      "step": 6403
+    },
+    {
+      "epoch": 2.513410718695974,
+      "grad_norm": 0.4424432885292469,
+      "learning_rate": 4.829117074341492e-06,
+      "loss": 0.4962,
+      "step": 6404
+    },
+    {
+      "epoch": 2.5138058779945665,
+      "grad_norm": 0.47103994040807695,
+      "learning_rate": 4.829060140529866e-06,
+      "loss": 0.5016,
+      "step": 6405
+    },
+    {
+      "epoch": 2.5142010372931587,
+      "grad_norm": 0.4583560717436245,
+      "learning_rate": 4.829003197571106e-06,
+      "loss": 0.5003,
+      "step": 6406
+    },
+    {
+      "epoch": 2.514596196591751,
+      "grad_norm": 0.4511981257492769,
+      "learning_rate": 4.828946245465435e-06,
+      "loss": 0.5092,
+      "step": 6407
+    },
+    {
+      "epoch": 2.514991355890343,
+      "grad_norm": 0.44378556134382374,
+      "learning_rate": 4.828889284213078e-06,
+      "loss": 0.4868,
+      "step": 6408
+    },
+    {
+      "epoch": 2.5153865151889354,
+      "grad_norm": 0.4638400372635114,
+      "learning_rate": 4.828832313814258e-06,
+      "loss": 0.5157,
+      "step": 6409
+    },
+    {
+      "epoch": 2.5157816744875277,
+      "grad_norm": 0.4555708689326073,
+      "learning_rate": 4.828775334269198e-06,
+      "loss": 0.503,
+      "step": 6410
+    },
+    {
+      "epoch": 2.51617683378612,
+      "grad_norm": 0.4740137430827515,
+      "learning_rate": 4.828718345578124e-06,
+      "loss": 0.5012,
+      "step": 6411
+    },
+    {
+      "epoch": 2.516571993084712,
+      "grad_norm": 0.440175406189311,
+      "learning_rate": 4.828661347741258e-06,
+      "loss": 0.4972,
+      "step": 6412
+    },
+    {
+      "epoch": 2.5169671523833044,
+      "grad_norm": 0.44634195808516225,
+      "learning_rate": 4.828604340758824e-06,
+      "loss": 0.521,
+      "step": 6413
+    },
+    {
+      "epoch": 2.5173623116818966,
+      "grad_norm": 0.44706452475534614,
+      "learning_rate": 4.828547324631045e-06,
+      "loss": 0.5099,
+      "step": 6414
+    },
+    {
+      "epoch": 2.517757470980489,
+      "grad_norm": 0.45566928817560864,
+      "learning_rate": 4.828490299358148e-06,
+      "loss": 0.4969,
+      "step": 6415
+    },
+    {
+      "epoch": 2.518152630279081,
+      "grad_norm": 0.45962733757335644,
+      "learning_rate": 4.828433264940354e-06,
+      "loss": 0.4894,
+      "step": 6416
+    },
+    {
+      "epoch": 2.5185477895776733,
+      "grad_norm": 0.4699440047924458,
+      "learning_rate": 4.828376221377889e-06,
+      "loss": 0.5016,
+      "step": 6417
+    },
+    {
+      "epoch": 2.5189429488762656,
+      "grad_norm": 0.4398573036728772,
+      "learning_rate": 4.828319168670974e-06,
+      "loss": 0.4954,
+      "step": 6418
+    },
+    {
+      "epoch": 2.519338108174858,
+      "grad_norm": 0.4553277215876072,
+      "learning_rate": 4.828262106819837e-06,
+      "loss": 0.5103,
+      "step": 6419
+    },
+    {
+      "epoch": 2.51973326747345,
+      "grad_norm": 0.455563395903747,
+      "learning_rate": 4.8282050358247e-06,
+      "loss": 0.5047,
+      "step": 6420
+    },
+    {
+      "epoch": 2.5201284267720423,
+      "grad_norm": 0.4614817667592641,
+      "learning_rate": 4.828147955685787e-06,
+      "loss": 0.5015,
+      "step": 6421
+    },
+    {
+      "epoch": 2.5205235860706345,
+      "grad_norm": 0.4540776683434743,
+      "learning_rate": 4.8280908664033225e-06,
+      "loss": 0.5004,
+      "step": 6422
+    },
+    {
+      "epoch": 2.5209187453692268,
+      "grad_norm": 0.4427896768855242,
+      "learning_rate": 4.828033767977531e-06,
+      "loss": 0.4898,
+      "step": 6423
+    },
+    {
+      "epoch": 2.521313904667819,
+      "grad_norm": 0.4546512843332004,
+      "learning_rate": 4.8279766604086365e-06,
+      "loss": 0.5219,
+      "step": 6424
+    },
+    {
+      "epoch": 2.5217090639664113,
+      "grad_norm": 0.46014478411090454,
+      "learning_rate": 4.827919543696863e-06,
+      "loss": 0.5003,
+      "step": 6425
+    },
+    {
+      "epoch": 2.5221042232650035,
+      "grad_norm": 0.5757044021010417,
+      "learning_rate": 4.827862417842435e-06,
+      "loss": 0.489,
+      "step": 6426
+    },
+    {
+      "epoch": 2.5224993825635957,
+      "grad_norm": 0.44379865154981046,
+      "learning_rate": 4.827805282845577e-06,
+      "loss": 0.5096,
+      "step": 6427
+    },
+    {
+      "epoch": 2.522894541862188,
+      "grad_norm": 0.4678140092742878,
+      "learning_rate": 4.827748138706514e-06,
+      "loss": 0.5189,
+      "step": 6428
+    },
+    {
+      "epoch": 2.52328970116078,
+      "grad_norm": 0.45998044186365117,
+      "learning_rate": 4.827690985425469e-06,
+      "loss": 0.5091,
+      "step": 6429
+    },
+    {
+      "epoch": 2.5236848604593725,
+      "grad_norm": 0.49126830773446944,
+      "learning_rate": 4.827633823002669e-06,
+      "loss": 0.5154,
+      "step": 6430
+    },
+    {
+      "epoch": 2.5240800197579647,
+      "grad_norm": 0.5033898050900029,
+      "learning_rate": 4.827576651438335e-06,
+      "loss": 0.504,
+      "step": 6431
+    },
+    {
+      "epoch": 2.524475179056557,
+      "grad_norm": 0.4401770069375346,
+      "learning_rate": 4.827519470732693e-06,
+      "loss": 0.5029,
+      "step": 6432
+    },
+    {
+      "epoch": 2.524870338355149,
+      "grad_norm": 0.45085555469804983,
+      "learning_rate": 4.82746228088597e-06,
+      "loss": 0.4777,
+      "step": 6433
+    },
+    {
+      "epoch": 2.5252654976537414,
+      "grad_norm": 0.4711645476173809,
+      "learning_rate": 4.827405081898387e-06,
+      "loss": 0.5107,
+      "step": 6434
+    },
+    {
+      "epoch": 2.5256606569523337,
+      "grad_norm": 0.46331381932207494,
+      "learning_rate": 4.82734787377017e-06,
+      "loss": 0.5287,
+      "step": 6435
+    },
+    {
+      "epoch": 2.526055816250926,
+      "grad_norm": 0.4595073465716432,
+      "learning_rate": 4.827290656501544e-06,
+      "loss": 0.5141,
+      "step": 6436
+    },
+    {
+      "epoch": 2.526450975549518,
+      "grad_norm": 0.45432945562079763,
+      "learning_rate": 4.827233430092733e-06,
+      "loss": 0.4814,
+      "step": 6437
+    },
+    {
+      "epoch": 2.5268461348481104,
+      "grad_norm": 0.4531070026892637,
+      "learning_rate": 4.827176194543963e-06,
+      "loss": 0.5066,
+      "step": 6438
+    },
+    {
+      "epoch": 2.5272412941467026,
+      "grad_norm": 0.5299603756232713,
+      "learning_rate": 4.8271189498554575e-06,
+      "loss": 0.5128,
+      "step": 6439
+    },
+    {
+      "epoch": 2.527636453445295,
+      "grad_norm": 0.444567842697348,
+      "learning_rate": 4.827061696027442e-06,
+      "loss": 0.4827,
+      "step": 6440
+    },
+    {
+      "epoch": 2.528031612743887,
+      "grad_norm": 0.4590491574983818,
+      "learning_rate": 4.827004433060142e-06,
+      "loss": 0.4784,
+      "step": 6441
+    },
+    {
+      "epoch": 2.52842677204248,
+      "grad_norm": 0.46222176709133134,
+      "learning_rate": 4.826947160953781e-06,
+      "loss": 0.5197,
+      "step": 6442
+    },
+    {
+      "epoch": 2.528821931341072,
+      "grad_norm": 0.4553604192939767,
+      "learning_rate": 4.826889879708585e-06,
+      "loss": 0.4899,
+      "step": 6443
+    },
+    {
+      "epoch": 2.5292170906396643,
+      "grad_norm": 0.4485178175280675,
+      "learning_rate": 4.826832589324778e-06,
+      "loss": 0.5062,
+      "step": 6444
+    },
+    {
+      "epoch": 2.5296122499382565,
+      "grad_norm": 0.45899668741167826,
+      "learning_rate": 4.8267752898025855e-06,
+      "loss": 0.5224,
+      "step": 6445
+    },
+    {
+      "epoch": 2.5300074092368487,
+      "grad_norm": 0.45782021791332567,
+      "learning_rate": 4.826717981142233e-06,
+      "loss": 0.5074,
+      "step": 6446
+    },
+    {
+      "epoch": 2.530402568535441,
+      "grad_norm": 0.471834837281007,
+      "learning_rate": 4.8266606633439445e-06,
+      "loss": 0.5152,
+      "step": 6447
+    },
+    {
+      "epoch": 2.5307977278340332,
+      "grad_norm": 0.48656419425692243,
+      "learning_rate": 4.826603336407945e-06,
+      "loss": 0.5209,
+      "step": 6448
+    },
+    {
+      "epoch": 2.5311928871326255,
+      "grad_norm": 0.45199569225901554,
+      "learning_rate": 4.826546000334462e-06,
+      "loss": 0.5035,
+      "step": 6449
+    },
+    {
+      "epoch": 2.5315880464312177,
+      "grad_norm": 0.4505649136262649,
+      "learning_rate": 4.826488655123719e-06,
+      "loss": 0.5034,
+      "step": 6450
+    },
+    {
+      "epoch": 2.53198320572981,
+      "grad_norm": 0.4750069460993798,
+      "learning_rate": 4.826431300775941e-06,
+      "loss": 0.5377,
+      "step": 6451
+    },
+    {
+      "epoch": 2.532378365028402,
+      "grad_norm": 0.464527033598869,
+      "learning_rate": 4.826373937291353e-06,
+      "loss": 0.4931,
+      "step": 6452
+    },
+    {
+      "epoch": 2.5327735243269944,
+      "grad_norm": 0.4530418607627022,
+      "learning_rate": 4.826316564670181e-06,
+      "loss": 0.5072,
+      "step": 6453
+    },
+    {
+      "epoch": 2.5331686836255867,
+      "grad_norm": 0.45364038939024304,
+      "learning_rate": 4.82625918291265e-06,
+      "loss": 0.5041,
+      "step": 6454
+    },
+    {
+      "epoch": 2.533563842924179,
+      "grad_norm": 0.47698190204477,
+      "learning_rate": 4.8262017920189864e-06,
+      "loss": 0.5116,
+      "step": 6455
+    },
+    {
+      "epoch": 2.533959002222771,
+      "grad_norm": 0.4731322955615775,
+      "learning_rate": 4.826144391989414e-06,
+      "loss": 0.5182,
+      "step": 6456
+    },
+    {
+      "epoch": 2.5343541615213634,
+      "grad_norm": 0.46876448274990934,
+      "learning_rate": 4.8260869828241595e-06,
+      "loss": 0.5053,
+      "step": 6457
+    },
+    {
+      "epoch": 2.5347493208199556,
+      "grad_norm": 0.4565534077873607,
+      "learning_rate": 4.826029564523447e-06,
+      "loss": 0.4996,
+      "step": 6458
+    },
+    {
+      "epoch": 2.535144480118548,
+      "grad_norm": 0.47123001937860665,
+      "learning_rate": 4.825972137087504e-06,
+      "loss": 0.5209,
+      "step": 6459
+    },
+    {
+      "epoch": 2.53553963941714,
+      "grad_norm": 0.4620746913890331,
+      "learning_rate": 4.825914700516553e-06,
+      "loss": 0.5306,
+      "step": 6460
+    },
+    {
+      "epoch": 2.5359347987157324,
+      "grad_norm": 0.4431055424606069,
+      "learning_rate": 4.825857254810823e-06,
+      "loss": 0.4777,
+      "step": 6461
+    },
+    {
+      "epoch": 2.5363299580143246,
+      "grad_norm": 0.4670518314206856,
+      "learning_rate": 4.8257997999705365e-06,
+      "loss": 0.4896,
+      "step": 6462
+    },
+    {
+      "epoch": 2.536725117312917,
+      "grad_norm": 0.46517370738592695,
+      "learning_rate": 4.825742335995922e-06,
+      "loss": 0.514,
+      "step": 6463
+    },
+    {
+      "epoch": 2.537120276611509,
+      "grad_norm": 0.46818923159639003,
+      "learning_rate": 4.825684862887204e-06,
+      "loss": 0.5092,
+      "step": 6464
+    },
+    {
+      "epoch": 2.5375154359101013,
+      "grad_norm": 0.47402007401065194,
+      "learning_rate": 4.825627380644607e-06,
+      "loss": 0.4979,
+      "step": 6465
+    },
+    {
+      "epoch": 2.5379105952086936,
+      "grad_norm": 0.45423949379520695,
+      "learning_rate": 4.825569889268359e-06,
+      "loss": 0.4905,
+      "step": 6466
+    },
+    {
+      "epoch": 2.538305754507286,
+      "grad_norm": 0.44881199031238245,
+      "learning_rate": 4.825512388758684e-06,
+      "loss": 0.4949,
+      "step": 6467
+    },
+    {
+      "epoch": 2.538700913805878,
+      "grad_norm": 0.43610597989809763,
+      "learning_rate": 4.825454879115808e-06,
+      "loss": 0.4913,
+      "step": 6468
+    },
+    {
+      "epoch": 2.5390960731044703,
+      "grad_norm": 0.46713538360387696,
+      "learning_rate": 4.8253973603399585e-06,
+      "loss": 0.5091,
+      "step": 6469
+    },
+    {
+      "epoch": 2.5394912324030625,
+      "grad_norm": 0.457557105090376,
+      "learning_rate": 4.825339832431359e-06,
+      "loss": 0.5067,
+      "step": 6470
+    },
+    {
+      "epoch": 2.5398863917016548,
+      "grad_norm": 0.44744590808245827,
+      "learning_rate": 4.8252822953902374e-06,
+      "loss": 0.5041,
+      "step": 6471
+    },
+    {
+      "epoch": 2.540281551000247,
+      "grad_norm": 0.44535213213717323,
+      "learning_rate": 4.825224749216819e-06,
+      "loss": 0.5114,
+      "step": 6472
+    },
+    {
+      "epoch": 2.5406767102988392,
+      "grad_norm": 0.4457813945339316,
+      "learning_rate": 4.825167193911329e-06,
+      "loss": 0.5075,
+      "step": 6473
+    },
+    {
+      "epoch": 2.5410718695974315,
+      "grad_norm": 0.45529691510876213,
+      "learning_rate": 4.825109629473995e-06,
+      "loss": 0.491,
+      "step": 6474
+    },
+    {
+      "epoch": 2.5414670288960237,
+      "grad_norm": 0.4716080108006373,
+      "learning_rate": 4.825052055905043e-06,
+      "loss": 0.5155,
+      "step": 6475
+    },
+    {
+      "epoch": 2.541862188194616,
+      "grad_norm": 0.4566158583555428,
+      "learning_rate": 4.8249944732046975e-06,
+      "loss": 0.522,
+      "step": 6476
+    },
+    {
+      "epoch": 2.542257347493208,
+      "grad_norm": 0.46881103956534304,
+      "learning_rate": 4.8249368813731845e-06,
+      "loss": 0.5274,
+      "step": 6477
+    },
+    {
+      "epoch": 2.5426525067918004,
+      "grad_norm": 0.45250162234599905,
+      "learning_rate": 4.824879280410733e-06,
+      "loss": 0.5005,
+      "step": 6478
+    },
+    {
+      "epoch": 2.5430476660903927,
+      "grad_norm": 0.47056029175688424,
+      "learning_rate": 4.824821670317566e-06,
+      "loss": 0.5215,
+      "step": 6479
+    },
+    {
+      "epoch": 2.543442825388985,
+      "grad_norm": 0.4597434281516728,
+      "learning_rate": 4.824764051093912e-06,
+      "loss": 0.5006,
+      "step": 6480
+    },
+    {
+      "epoch": 2.543837984687577,
+      "grad_norm": 0.45131921755337623,
+      "learning_rate": 4.824706422739996e-06,
+      "loss": 0.5014,
+      "step": 6481
+    },
+    {
+      "epoch": 2.5442331439861694,
+      "grad_norm": 0.4435352090089486,
+      "learning_rate": 4.824648785256045e-06,
+      "loss": 0.4683,
+      "step": 6482
+    },
+    {
+      "epoch": 2.5446283032847616,
+      "grad_norm": 0.4361623565598776,
+      "learning_rate": 4.824591138642285e-06,
+      "loss": 0.4739,
+      "step": 6483
+    },
+    {
+      "epoch": 2.545023462583354,
+      "grad_norm": 0.4528329358816713,
+      "learning_rate": 4.824533482898943e-06,
+      "loss": 0.4936,
+      "step": 6484
+    },
+    {
+      "epoch": 2.545418621881946,
+      "grad_norm": 0.457804851022941,
+      "learning_rate": 4.824475818026244e-06,
+      "loss": 0.5038,
+      "step": 6485
+    },
+    {
+      "epoch": 2.5458137811805384,
+      "grad_norm": 0.47354522749284367,
+      "learning_rate": 4.824418144024416e-06,
+      "loss": 0.5182,
+      "step": 6486
+    },
+    {
+      "epoch": 2.5462089404791306,
+      "grad_norm": 0.4695229394887001,
+      "learning_rate": 4.824360460893686e-06,
+      "loss": 0.4991,
+      "step": 6487
+    },
+    {
+      "epoch": 2.546604099777723,
+      "grad_norm": 0.46327892852606095,
+      "learning_rate": 4.824302768634279e-06,
+      "loss": 0.502,
+      "step": 6488
+    },
+    {
+      "epoch": 2.546999259076315,
+      "grad_norm": 0.45297480248431404,
+      "learning_rate": 4.824245067246422e-06,
+      "loss": 0.5066,
+      "step": 6489
+    },
+    {
+      "epoch": 2.5473944183749073,
+      "grad_norm": 0.46517129424956516,
+      "learning_rate": 4.824187356730341e-06,
+      "loss": 0.4846,
+      "step": 6490
+    },
+    {
+      "epoch": 2.5477895776734996,
+      "grad_norm": 0.4668569089467578,
+      "learning_rate": 4.824129637086264e-06,
+      "loss": 0.5238,
+      "step": 6491
+    },
+    {
+      "epoch": 2.548184736972092,
+      "grad_norm": 0.47720546968786415,
+      "learning_rate": 4.824071908314417e-06,
+      "loss": 0.5293,
+      "step": 6492
+    },
+    {
+      "epoch": 2.548579896270684,
+      "grad_norm": 0.4729585091074834,
+      "learning_rate": 4.824014170415027e-06,
+      "loss": 0.5163,
+      "step": 6493
+    },
+    {
+      "epoch": 2.5489750555692763,
+      "grad_norm": 0.4559767498901322,
+      "learning_rate": 4.8239564233883205e-06,
+      "loss": 0.5028,
+      "step": 6494
+    },
+    {
+      "epoch": 2.5493702148678685,
+      "grad_norm": 0.46193288182539766,
+      "learning_rate": 4.823898667234525e-06,
+      "loss": 0.5036,
+      "step": 6495
+    },
+    {
+      "epoch": 2.5497653741664608,
+      "grad_norm": 0.50276777112018,
+      "learning_rate": 4.823840901953865e-06,
+      "loss": 0.5199,
+      "step": 6496
+    },
+    {
+      "epoch": 2.550160533465053,
+      "grad_norm": 0.44571450965752996,
+      "learning_rate": 4.823783127546571e-06,
+      "loss": 0.4981,
+      "step": 6497
+    },
+    {
+      "epoch": 2.5505556927636452,
+      "grad_norm": 0.4545230887683023,
+      "learning_rate": 4.823725344012866e-06,
+      "loss": 0.5014,
+      "step": 6498
+    },
+    {
+      "epoch": 2.5509508520622375,
+      "grad_norm": 0.45011277157561497,
+      "learning_rate": 4.8236675513529804e-06,
+      "loss": 0.5016,
+      "step": 6499
+    },
+    {
+      "epoch": 2.5513460113608297,
+      "grad_norm": 0.45347326578518127,
+      "learning_rate": 4.823609749567139e-06,
+      "loss": 0.486,
+      "step": 6500
+    },
+    {
+      "epoch": 2.551741170659422,
+      "grad_norm": 0.49671184627419557,
+      "learning_rate": 4.823551938655569e-06,
+      "loss": 0.5229,
+      "step": 6501
+    },
+    {
+      "epoch": 2.552136329958014,
+      "grad_norm": 0.4630157588304572,
+      "learning_rate": 4.823494118618499e-06,
+      "loss": 0.4994,
+      "step": 6502
+    },
+    {
+      "epoch": 2.5525314892566064,
+      "grad_norm": 0.44595435111081594,
+      "learning_rate": 4.8234362894561544e-06,
+      "loss": 0.4913,
+      "step": 6503
+    },
+    {
+      "epoch": 2.5529266485551987,
+      "grad_norm": 0.46456032038638434,
+      "learning_rate": 4.823378451168763e-06,
+      "loss": 0.4989,
+      "step": 6504
+    },
+    {
+      "epoch": 2.553321807853791,
+      "grad_norm": 0.4741181020350032,
+      "learning_rate": 4.8233206037565515e-06,
+      "loss": 0.5169,
+      "step": 6505
+    },
+    {
+      "epoch": 2.553716967152383,
+      "grad_norm": 0.4586086884230933,
+      "learning_rate": 4.823262747219749e-06,
+      "loss": 0.5044,
+      "step": 6506
+    },
+    {
+      "epoch": 2.5541121264509754,
+      "grad_norm": 0.4594391518699296,
+      "learning_rate": 4.823204881558579e-06,
+      "loss": 0.5095,
+      "step": 6507
+    },
+    {
+      "epoch": 2.5545072857495676,
+      "grad_norm": 0.4520677455107984,
+      "learning_rate": 4.8231470067732726e-06,
+      "loss": 0.5007,
+      "step": 6508
+    },
+    {
+      "epoch": 2.55490244504816,
+      "grad_norm": 0.4685641116507666,
+      "learning_rate": 4.823089122864055e-06,
+      "loss": 0.513,
+      "step": 6509
+    },
+    {
+      "epoch": 2.555297604346752,
+      "grad_norm": 0.45202109312605754,
+      "learning_rate": 4.8230312298311535e-06,
+      "loss": 0.5163,
+      "step": 6510
+    },
+    {
+      "epoch": 2.5556927636453444,
+      "grad_norm": 0.4521725537088804,
+      "learning_rate": 4.822973327674796e-06,
+      "loss": 0.4907,
+      "step": 6511
+    },
+    {
+      "epoch": 2.556087922943937,
+      "grad_norm": 0.44282430032215886,
+      "learning_rate": 4.82291541639521e-06,
+      "loss": 0.5044,
+      "step": 6512
+    },
+    {
+      "epoch": 2.5564830822425293,
+      "grad_norm": 0.4656119245187877,
+      "learning_rate": 4.822857495992623e-06,
+      "loss": 0.5138,
+      "step": 6513
+    },
+    {
+      "epoch": 2.5568782415411215,
+      "grad_norm": 0.46796730981011786,
+      "learning_rate": 4.8227995664672625e-06,
+      "loss": 0.5162,
+      "step": 6514
+    },
+    {
+      "epoch": 2.5572734008397138,
+      "grad_norm": 0.4491042247333808,
+      "learning_rate": 4.822741627819355e-06,
+      "loss": 0.5012,
+      "step": 6515
+    },
+    {
+      "epoch": 2.557668560138306,
+      "grad_norm": 0.46834449695954017,
+      "learning_rate": 4.82268368004913e-06,
+      "loss": 0.5263,
+      "step": 6516
+    },
+    {
+      "epoch": 2.5580637194368983,
+      "grad_norm": 0.44103675945686854,
+      "learning_rate": 4.822625723156813e-06,
+      "loss": 0.4862,
+      "step": 6517
+    },
+    {
+      "epoch": 2.5584588787354905,
+      "grad_norm": 0.45017086475460627,
+      "learning_rate": 4.822567757142634e-06,
+      "loss": 0.4989,
+      "step": 6518
+    },
+    {
+      "epoch": 2.5588540380340827,
+      "grad_norm": 0.44261210213602237,
+      "learning_rate": 4.822509782006817e-06,
+      "loss": 0.5209,
+      "step": 6519
+    },
+    {
+      "epoch": 2.559249197332675,
+      "grad_norm": 0.45761518270513285,
+      "learning_rate": 4.822451797749592e-06,
+      "loss": 0.5203,
+      "step": 6520
+    },
+    {
+      "epoch": 2.559644356631267,
+      "grad_norm": 0.46985808636516363,
+      "learning_rate": 4.822393804371188e-06,
+      "loss": 0.5149,
+      "step": 6521
+    },
+    {
+      "epoch": 2.5600395159298595,
+      "grad_norm": 0.4526976621493061,
+      "learning_rate": 4.822335801871832e-06,
+      "loss": 0.5091,
+      "step": 6522
+    },
+    {
+      "epoch": 2.5604346752284517,
+      "grad_norm": 0.46492146610005,
+      "learning_rate": 4.822277790251749e-06,
+      "loss": 0.5276,
+      "step": 6523
+    },
+    {
+      "epoch": 2.560829834527044,
+      "grad_norm": 0.46271285637553683,
+      "learning_rate": 4.82221976951117e-06,
+      "loss": 0.4965,
+      "step": 6524
+    },
+    {
+      "epoch": 2.561224993825636,
+      "grad_norm": 0.4503037126759286,
+      "learning_rate": 4.822161739650322e-06,
+      "loss": 0.516,
+      "step": 6525
+    },
+    {
+      "epoch": 2.5616201531242284,
+      "grad_norm": 0.443470486159481,
+      "learning_rate": 4.822103700669432e-06,
+      "loss": 0.4974,
+      "step": 6526
+    },
+    {
+      "epoch": 2.5620153124228207,
+      "grad_norm": 0.43703146172923873,
+      "learning_rate": 4.82204565256873e-06,
+      "loss": 0.4952,
+      "step": 6527
+    },
+    {
+      "epoch": 2.562410471721413,
+      "grad_norm": 0.45292345235815396,
+      "learning_rate": 4.821987595348442e-06,
+      "loss": 0.5271,
+      "step": 6528
+    },
+    {
+      "epoch": 2.562805631020005,
+      "grad_norm": 0.45759067526669045,
+      "learning_rate": 4.821929529008797e-06,
+      "loss": 0.5063,
+      "step": 6529
+    },
+    {
+      "epoch": 2.5632007903185974,
+      "grad_norm": 0.47064593711048897,
+      "learning_rate": 4.821871453550023e-06,
+      "loss": 0.5057,
+      "step": 6530
+    },
+    {
+      "epoch": 2.5635959496171896,
+      "grad_norm": 0.47813721885154226,
+      "learning_rate": 4.821813368972347e-06,
+      "loss": 0.539,
+      "step": 6531
+    },
+    {
+      "epoch": 2.563991108915782,
+      "grad_norm": 0.4573359139734747,
+      "learning_rate": 4.821755275275998e-06,
+      "loss": 0.5119,
+      "step": 6532
+    },
+    {
+      "epoch": 2.564386268214374,
+      "grad_norm": 0.5039910077391357,
+      "learning_rate": 4.821697172461205e-06,
+      "loss": 0.5096,
+      "step": 6533
+    },
+    {
+      "epoch": 2.5647814275129663,
+      "grad_norm": 0.45480012799103875,
+      "learning_rate": 4.821639060528194e-06,
+      "loss": 0.5073,
+      "step": 6534
+    },
+    {
+      "epoch": 2.5651765868115586,
+      "grad_norm": 0.43914393122344525,
+      "learning_rate": 4.821580939477195e-06,
+      "loss": 0.5173,
+      "step": 6535
+    },
+    {
+      "epoch": 2.565571746110151,
+      "grad_norm": 0.4512267600492565,
+      "learning_rate": 4.821522809308436e-06,
+      "loss": 0.5046,
+      "step": 6536
+    },
+    {
+      "epoch": 2.565966905408743,
+      "grad_norm": 0.45888241462106466,
+      "learning_rate": 4.821464670022146e-06,
+      "loss": 0.5159,
+      "step": 6537
+    },
+    {
+      "epoch": 2.5663620647073353,
+      "grad_norm": 0.4441592838891458,
+      "learning_rate": 4.821406521618551e-06,
+      "loss": 0.5057,
+      "step": 6538
+    },
+    {
+      "epoch": 2.5667572240059275,
+      "grad_norm": 0.45376803364604695,
+      "learning_rate": 4.821348364097882e-06,
+      "loss": 0.4884,
+      "step": 6539
+    },
+    {
+      "epoch": 2.56715238330452,
+      "grad_norm": 0.46633304742246534,
+      "learning_rate": 4.821290197460366e-06,
+      "loss": 0.5307,
+      "step": 6540
+    },
+    {
+      "epoch": 2.567547542603112,
+      "grad_norm": 0.43666480002999797,
+      "learning_rate": 4.821232021706231e-06,
+      "loss": 0.4926,
+      "step": 6541
+    },
+    {
+      "epoch": 2.5679427019017043,
+      "grad_norm": 0.4605237411052962,
+      "learning_rate": 4.8211738368357065e-06,
+      "loss": 0.5049,
+      "step": 6542
+    },
+    {
+      "epoch": 2.5683378612002965,
+      "grad_norm": 0.462050066279481,
+      "learning_rate": 4.821115642849021e-06,
+      "loss": 0.5165,
+      "step": 6543
+    },
+    {
+      "epoch": 2.5687330204988887,
+      "grad_norm": 0.45750052243887834,
+      "learning_rate": 4.821057439746402e-06,
+      "loss": 0.4976,
+      "step": 6544
+    },
+    {
+      "epoch": 2.569128179797481,
+      "grad_norm": 0.44555516105490556,
+      "learning_rate": 4.820999227528079e-06,
+      "loss": 0.5186,
+      "step": 6545
+    },
+    {
+      "epoch": 2.5695233390960732,
+      "grad_norm": 0.46285962401306396,
+      "learning_rate": 4.820941006194281e-06,
+      "loss": 0.4978,
+      "step": 6546
+    },
+    {
+      "epoch": 2.5699184983946655,
+      "grad_norm": 0.4469392519255687,
+      "learning_rate": 4.820882775745236e-06,
+      "loss": 0.4863,
+      "step": 6547
+    },
+    {
+      "epoch": 2.5703136576932577,
+      "grad_norm": 0.4519758312975245,
+      "learning_rate": 4.8208245361811724e-06,
+      "loss": 0.5094,
+      "step": 6548
+    },
+    {
+      "epoch": 2.57070881699185,
+      "grad_norm": 0.46503941371816515,
+      "learning_rate": 4.820766287502319e-06,
+      "loss": 0.5133,
+      "step": 6549
+    },
+    {
+      "epoch": 2.571103976290442,
+      "grad_norm": 0.4600897389443086,
+      "learning_rate": 4.820708029708905e-06,
+      "loss": 0.5137,
+      "step": 6550
+    },
+    {
+      "epoch": 2.5714991355890344,
+      "grad_norm": 0.4588463556332273,
+      "learning_rate": 4.820649762801159e-06,
+      "loss": 0.5092,
+      "step": 6551
+    },
+    {
+      "epoch": 2.5718942948876267,
+      "grad_norm": 0.46703432271255124,
+      "learning_rate": 4.820591486779312e-06,
+      "loss": 0.4949,
+      "step": 6552
+    },
+    {
+      "epoch": 2.572289454186219,
+      "grad_norm": 0.4489108181770717,
+      "learning_rate": 4.820533201643588e-06,
+      "loss": 0.4944,
+      "step": 6553
+    },
+    {
+      "epoch": 2.572684613484811,
+      "grad_norm": 0.4608905598839847,
+      "learning_rate": 4.82047490739422e-06,
+      "loss": 0.5049,
+      "step": 6554
+    },
+    {
+      "epoch": 2.5730797727834034,
+      "grad_norm": 0.46983121539687306,
+      "learning_rate": 4.820416604031435e-06,
+      "loss": 0.4909,
+      "step": 6555
+    },
+    {
+      "epoch": 2.5734749320819956,
+      "grad_norm": 0.4533141543179746,
+      "learning_rate": 4.820358291555462e-06,
+      "loss": 0.525,
+      "step": 6556
+    },
+    {
+      "epoch": 2.573870091380588,
+      "grad_norm": 0.4619074947389451,
+      "learning_rate": 4.820299969966532e-06,
+      "loss": 0.5169,
+      "step": 6557
+    },
+    {
+      "epoch": 2.57426525067918,
+      "grad_norm": 0.44500184074671084,
+      "learning_rate": 4.820241639264872e-06,
+      "loss": 0.5106,
+      "step": 6558
+    },
+    {
+      "epoch": 2.5746604099777723,
+      "grad_norm": 0.45059185521325373,
+      "learning_rate": 4.820183299450713e-06,
+      "loss": 0.5029,
+      "step": 6559
+    },
+    {
+      "epoch": 2.5750555692763646,
+      "grad_norm": 0.45616010644680055,
+      "learning_rate": 4.820124950524282e-06,
+      "loss": 0.5018,
+      "step": 6560
+    },
+    {
+      "epoch": 2.575450728574957,
+      "grad_norm": 0.4572348903384075,
+      "learning_rate": 4.820066592485809e-06,
+      "loss": 0.5139,
+      "step": 6561
+    },
+    {
+      "epoch": 2.575845887873549,
+      "grad_norm": 0.4421219919687656,
+      "learning_rate": 4.8200082253355226e-06,
+      "loss": 0.4971,
+      "step": 6562
+    },
+    {
+      "epoch": 2.5762410471721413,
+      "grad_norm": 0.4666909383851814,
+      "learning_rate": 4.819949849073654e-06,
+      "loss": 0.5117,
+      "step": 6563
+    },
+    {
+      "epoch": 2.5766362064707335,
+      "grad_norm": 0.46477953473232203,
+      "learning_rate": 4.8198914637004305e-06,
+      "loss": 0.4949,
+      "step": 6564
+    },
+    {
+      "epoch": 2.577031365769326,
+      "grad_norm": 0.46799345127111774,
+      "learning_rate": 4.819833069216081e-06,
+      "loss": 0.4962,
+      "step": 6565
+    },
+    {
+      "epoch": 2.577426525067918,
+      "grad_norm": 0.4431351713952274,
+      "learning_rate": 4.819774665620837e-06,
+      "loss": 0.4857,
+      "step": 6566
+    },
+    {
+      "epoch": 2.5778216843665103,
+      "grad_norm": 0.45369599827793194,
+      "learning_rate": 4.819716252914927e-06,
+      "loss": 0.5163,
+      "step": 6567
+    },
+    {
+      "epoch": 2.5782168436651025,
+      "grad_norm": 0.4736993413815267,
+      "learning_rate": 4.81965783109858e-06,
+      "loss": 0.5077,
+      "step": 6568
+    },
+    {
+      "epoch": 2.5786120029636947,
+      "grad_norm": 0.4767138840762411,
+      "learning_rate": 4.819599400172025e-06,
+      "loss": 0.5151,
+      "step": 6569
+    },
+    {
+      "epoch": 2.579007162262287,
+      "grad_norm": 0.46197056590790053,
+      "learning_rate": 4.819540960135493e-06,
+      "loss": 0.494,
+      "step": 6570
+    },
+    {
+      "epoch": 2.5794023215608792,
+      "grad_norm": 0.4719226833563452,
+      "learning_rate": 4.819482510989211e-06,
+      "loss": 0.5235,
+      "step": 6571
+    },
+    {
+      "epoch": 2.5797974808594715,
+      "grad_norm": 0.47446676249677355,
+      "learning_rate": 4.8194240527334115e-06,
+      "loss": 0.5077,
+      "step": 6572
+    },
+    {
+      "epoch": 2.5801926401580637,
+      "grad_norm": 0.4660645331065908,
+      "learning_rate": 4.819365585368322e-06,
+      "loss": 0.5013,
+      "step": 6573
+    },
+    {
+      "epoch": 2.580587799456656,
+      "grad_norm": 0.45366183792375775,
+      "learning_rate": 4.819307108894173e-06,
+      "loss": 0.4923,
+      "step": 6574
+    },
+    {
+      "epoch": 2.580982958755248,
+      "grad_norm": 0.43736598970253193,
+      "learning_rate": 4.819248623311195e-06,
+      "loss": 0.5013,
+      "step": 6575
+    },
+    {
+      "epoch": 2.5813781180538404,
+      "grad_norm": 0.46508435134068205,
+      "learning_rate": 4.819190128619617e-06,
+      "loss": 0.4959,
+      "step": 6576
+    },
+    {
+      "epoch": 2.5817732773524327,
+      "grad_norm": 0.46567765929034605,
+      "learning_rate": 4.819131624819667e-06,
+      "loss": 0.5076,
+      "step": 6577
+    },
+    {
+      "epoch": 2.582168436651025,
+      "grad_norm": 0.4707030207964489,
+      "learning_rate": 4.8190731119115766e-06,
+      "loss": 0.5174,
+      "step": 6578
+    },
+    {
+      "epoch": 2.582563595949617,
+      "grad_norm": 0.46834324977420716,
+      "learning_rate": 4.819014589895575e-06,
+      "loss": 0.5154,
+      "step": 6579
+    },
+    {
+      "epoch": 2.5829587552482094,
+      "grad_norm": 0.4620901719740855,
+      "learning_rate": 4.818956058771893e-06,
+      "loss": 0.4916,
+      "step": 6580
+    },
+    {
+      "epoch": 2.5833539145468016,
+      "grad_norm": 0.4619346062310383,
+      "learning_rate": 4.81889751854076e-06,
+      "loss": 0.5024,
+      "step": 6581
+    },
+    {
+      "epoch": 2.583749073845394,
+      "grad_norm": 0.4704870295417848,
+      "learning_rate": 4.818838969202405e-06,
+      "loss": 0.507,
+      "step": 6582
+    },
+    {
+      "epoch": 2.584144233143986,
+      "grad_norm": 0.48508779424763204,
+      "learning_rate": 4.818780410757059e-06,
+      "loss": 0.512,
+      "step": 6583
+    },
+    {
+      "epoch": 2.5845393924425784,
+      "grad_norm": 0.44361406120966196,
+      "learning_rate": 4.818721843204951e-06,
+      "loss": 0.4971,
+      "step": 6584
+    },
+    {
+      "epoch": 2.5849345517411706,
+      "grad_norm": 0.46045806937302997,
+      "learning_rate": 4.818663266546312e-06,
+      "loss": 0.5146,
+      "step": 6585
+    },
+    {
+      "epoch": 2.585329711039763,
+      "grad_norm": 0.503532295954558,
+      "learning_rate": 4.818604680781372e-06,
+      "loss": 0.5202,
+      "step": 6586
+    },
+    {
+      "epoch": 2.585724870338355,
+      "grad_norm": 0.4809956728838628,
+      "learning_rate": 4.8185460859103596e-06,
+      "loss": 0.5341,
+      "step": 6587
+    },
+    {
+      "epoch": 2.5861200296369473,
+      "grad_norm": 0.4452613791874474,
+      "learning_rate": 4.818487481933507e-06,
+      "loss": 0.4868,
+      "step": 6588
+    },
+    {
+      "epoch": 2.5865151889355396,
+      "grad_norm": 0.4645895638161033,
+      "learning_rate": 4.818428868851042e-06,
+      "loss": 0.5162,
+      "step": 6589
+    },
+    {
+      "epoch": 2.586910348234132,
+      "grad_norm": 0.45690373781668653,
+      "learning_rate": 4.818370246663199e-06,
+      "loss": 0.5186,
+      "step": 6590
+    },
+    {
+      "epoch": 2.587305507532724,
+      "grad_norm": 0.4624829183292078,
+      "learning_rate": 4.818311615370204e-06,
+      "loss": 0.5169,
+      "step": 6591
+    },
+    {
+      "epoch": 2.5877006668313163,
+      "grad_norm": 0.48491493887224846,
+      "learning_rate": 4.818252974972288e-06,
+      "loss": 0.5092,
+      "step": 6592
+    },
+    {
+      "epoch": 2.5880958261299085,
+      "grad_norm": 0.4642333592429095,
+      "learning_rate": 4.818194325469683e-06,
+      "loss": 0.532,
+      "step": 6593
+    },
+    {
+      "epoch": 2.5884909854285008,
+      "grad_norm": 0.48132123097086577,
+      "learning_rate": 4.818135666862618e-06,
+      "loss": 0.5152,
+      "step": 6594
+    },
+    {
+      "epoch": 2.588886144727093,
+      "grad_norm": 0.4736558582523291,
+      "learning_rate": 4.818076999151323e-06,
+      "loss": 0.5074,
+      "step": 6595
+    },
+    {
+      "epoch": 2.5892813040256852,
+      "grad_norm": 0.45982193687887485,
+      "learning_rate": 4.81801832233603e-06,
+      "loss": 0.5136,
+      "step": 6596
+    },
+    {
+      "epoch": 2.5896764633242775,
+      "grad_norm": 0.459117193820494,
+      "learning_rate": 4.817959636416969e-06,
+      "loss": 0.5178,
+      "step": 6597
+    },
+    {
+      "epoch": 2.5900716226228697,
+      "grad_norm": 0.45795808917106084,
+      "learning_rate": 4.817900941394369e-06,
+      "loss": 0.497,
+      "step": 6598
+    },
+    {
+      "epoch": 2.590466781921462,
+      "grad_norm": 0.4961771265409982,
+      "learning_rate": 4.817842237268463e-06,
+      "loss": 0.5025,
+      "step": 6599
+    },
+    {
+      "epoch": 2.590861941220054,
+      "grad_norm": 0.4517244851535084,
+      "learning_rate": 4.817783524039479e-06,
+      "loss": 0.496,
+      "step": 6600
+    },
+    {
+      "epoch": 2.5912571005186464,
+      "grad_norm": 0.4606749996141201,
+      "learning_rate": 4.8177248017076496e-06,
+      "loss": 0.4885,
+      "step": 6601
+    },
+    {
+      "epoch": 2.5916522598172387,
+      "grad_norm": 0.44595584652993003,
+      "learning_rate": 4.817666070273203e-06,
+      "loss": 0.5178,
+      "step": 6602
+    },
+    {
+      "epoch": 2.592047419115831,
+      "grad_norm": 0.4361569136817302,
+      "learning_rate": 4.817607329736373e-06,
+      "loss": 0.4927,
+      "step": 6603
+    },
+    {
+      "epoch": 2.592442578414423,
+      "grad_norm": 0.48331893774269574,
+      "learning_rate": 4.817548580097389e-06,
+      "loss": 0.4918,
+      "step": 6604
+    },
+    {
+      "epoch": 2.5928377377130154,
+      "grad_norm": 0.4480574187648233,
+      "learning_rate": 4.81748982135648e-06,
+      "loss": 0.4898,
+      "step": 6605
+    },
+    {
+      "epoch": 2.5932328970116076,
+      "grad_norm": 0.4574403926622687,
+      "learning_rate": 4.817431053513879e-06,
+      "loss": 0.5153,
+      "step": 6606
+    },
+    {
+      "epoch": 2.5936280563102,
+      "grad_norm": 0.4464682971660518,
+      "learning_rate": 4.8173722765698165e-06,
+      "loss": 0.4879,
+      "step": 6607
+    },
+    {
+      "epoch": 2.594023215608792,
+      "grad_norm": 0.4433326822459801,
+      "learning_rate": 4.817313490524523e-06,
+      "loss": 0.4725,
+      "step": 6608
+    },
+    {
+      "epoch": 2.5944183749073844,
+      "grad_norm": 0.4707423582400784,
+      "learning_rate": 4.817254695378228e-06,
+      "loss": 0.5221,
+      "step": 6609
+    },
+    {
+      "epoch": 2.5948135342059766,
+      "grad_norm": 0.47295746581167725,
+      "learning_rate": 4.8171958911311646e-06,
+      "loss": 0.5108,
+      "step": 6610
+    },
+    {
+      "epoch": 2.595208693504569,
+      "grad_norm": 0.45263375041631376,
+      "learning_rate": 4.817137077783563e-06,
+      "loss": 0.492,
+      "step": 6611
+    },
+    {
+      "epoch": 2.595603852803161,
+      "grad_norm": 0.4552715976251914,
+      "learning_rate": 4.817078255335653e-06,
+      "loss": 0.496,
+      "step": 6612
+    },
+    {
+      "epoch": 2.5959990121017533,
+      "grad_norm": 0.4595060840449498,
+      "learning_rate": 4.817019423787667e-06,
+      "loss": 0.5155,
+      "step": 6613
+    },
+    {
+      "epoch": 2.5963941714003456,
+      "grad_norm": 0.45945144727738035,
+      "learning_rate": 4.8169605831398355e-06,
+      "loss": 0.5292,
+      "step": 6614
+    },
+    {
+      "epoch": 2.596789330698938,
+      "grad_norm": 0.45474069846107534,
+      "learning_rate": 4.81690173339239e-06,
+      "loss": 0.5158,
+      "step": 6615
+    },
+    {
+      "epoch": 2.59718448999753,
+      "grad_norm": 0.44549691265764996,
+      "learning_rate": 4.816842874545562e-06,
+      "loss": 0.4949,
+      "step": 6616
+    },
+    {
+      "epoch": 2.5975796492961223,
+      "grad_norm": 0.4465165484665156,
+      "learning_rate": 4.816784006599582e-06,
+      "loss": 0.5135,
+      "step": 6617
+    },
+    {
+      "epoch": 2.5979748085947145,
+      "grad_norm": 0.46690949256258174,
+      "learning_rate": 4.81672512955468e-06,
+      "loss": 0.4823,
+      "step": 6618
+    },
+    {
+      "epoch": 2.5983699678933068,
+      "grad_norm": 0.44898509566584294,
+      "learning_rate": 4.81666624341109e-06,
+      "loss": 0.5126,
+      "step": 6619
+    },
+    {
+      "epoch": 2.598765127191899,
+      "grad_norm": 0.4486657739928971,
+      "learning_rate": 4.816607348169041e-06,
+      "loss": 0.4901,
+      "step": 6620
+    },
+    {
+      "epoch": 2.5991602864904912,
+      "grad_norm": 0.4527214681354412,
+      "learning_rate": 4.816548443828765e-06,
+      "loss": 0.5123,
+      "step": 6621
+    },
+    {
+      "epoch": 2.5995554457890835,
+      "grad_norm": 0.4529843101656628,
+      "learning_rate": 4.8164895303904935e-06,
+      "loss": 0.4951,
+      "step": 6622
+    },
+    {
+      "epoch": 2.5999506050876757,
+      "grad_norm": 0.4564936706989616,
+      "learning_rate": 4.816430607854458e-06,
+      "loss": 0.5113,
+      "step": 6623
+    },
+    {
+      "epoch": 2.600345764386268,
+      "grad_norm": 0.4940966671369367,
+      "learning_rate": 4.816371676220889e-06,
+      "loss": 0.4969,
+      "step": 6624
+    },
+    {
+      "epoch": 2.60074092368486,
+      "grad_norm": 0.45722310993861126,
+      "learning_rate": 4.81631273549002e-06,
+      "loss": 0.5132,
+      "step": 6625
+    },
+    {
+      "epoch": 2.6011360829834524,
+      "grad_norm": 0.45286575656466427,
+      "learning_rate": 4.816253785662079e-06,
+      "loss": 0.5084,
+      "step": 6626
+    },
+    {
+      "epoch": 2.6015312422820447,
+      "grad_norm": 0.445135014900183,
+      "learning_rate": 4.816194826737302e-06,
+      "loss": 0.5027,
+      "step": 6627
+    },
+    {
+      "epoch": 2.601926401580637,
+      "grad_norm": 0.46012984610648944,
+      "learning_rate": 4.816135858715917e-06,
+      "loss": 0.4891,
+      "step": 6628
+    },
+    {
+      "epoch": 2.602321560879229,
+      "grad_norm": 0.46640326101065027,
+      "learning_rate": 4.816076881598156e-06,
+      "loss": 0.5077,
+      "step": 6629
+    },
+    {
+      "epoch": 2.6027167201778214,
+      "grad_norm": 0.4395627749768447,
+      "learning_rate": 4.816017895384253e-06,
+      "loss": 0.5029,
+      "step": 6630
+    },
+    {
+      "epoch": 2.603111879476414,
+      "grad_norm": 0.4471917295160825,
+      "learning_rate": 4.815958900074437e-06,
+      "loss": 0.4987,
+      "step": 6631
+    },
+    {
+      "epoch": 2.6035070387750063,
+      "grad_norm": 0.4577848836839979,
+      "learning_rate": 4.815899895668941e-06,
+      "loss": 0.5088,
+      "step": 6632
+    },
+    {
+      "epoch": 2.6039021980735986,
+      "grad_norm": 0.44634805154755014,
+      "learning_rate": 4.815840882167997e-06,
+      "loss": 0.5121,
+      "step": 6633
+    },
+    {
+      "epoch": 2.604297357372191,
+      "grad_norm": 0.4506256280957546,
+      "learning_rate": 4.815781859571835e-06,
+      "loss": 0.5081,
+      "step": 6634
+    },
+    {
+      "epoch": 2.604692516670783,
+      "grad_norm": 0.44851796039012143,
+      "learning_rate": 4.815722827880689e-06,
+      "loss": 0.5061,
+      "step": 6635
+    },
+    {
+      "epoch": 2.6050876759693753,
+      "grad_norm": 0.4632284380664699,
+      "learning_rate": 4.81566378709479e-06,
+      "loss": 0.5179,
+      "step": 6636
+    },
+    {
+      "epoch": 2.6054828352679675,
+      "grad_norm": 0.4495346025856113,
+      "learning_rate": 4.8156047372143695e-06,
+      "loss": 0.5185,
+      "step": 6637
+    },
+    {
+      "epoch": 2.6058779945665598,
+      "grad_norm": 0.45146252001014836,
+      "learning_rate": 4.815545678239659e-06,
+      "loss": 0.4782,
+      "step": 6638
+    },
+    {
+      "epoch": 2.606273153865152,
+      "grad_norm": 0.473770212276825,
+      "learning_rate": 4.8154866101708925e-06,
+      "loss": 0.5126,
+      "step": 6639
+    },
+    {
+      "epoch": 2.6066683131637443,
+      "grad_norm": 0.4485774694573647,
+      "learning_rate": 4.8154275330083e-06,
+      "loss": 0.5119,
+      "step": 6640
+    },
+    {
+      "epoch": 2.6070634724623365,
+      "grad_norm": 0.49678656209215,
+      "learning_rate": 4.815368446752114e-06,
+      "loss": 0.5094,
+      "step": 6641
+    },
+    {
+      "epoch": 2.6074586317609287,
+      "grad_norm": 0.470453453281169,
+      "learning_rate": 4.815309351402568e-06,
+      "loss": 0.5265,
+      "step": 6642
+    },
+    {
+      "epoch": 2.607853791059521,
+      "grad_norm": 0.46330047850218925,
+      "learning_rate": 4.815250246959891e-06,
+      "loss": 0.5093,
+      "step": 6643
+    },
+    {
+      "epoch": 2.608248950358113,
+      "grad_norm": 0.4646962109211404,
+      "learning_rate": 4.815191133424318e-06,
+      "loss": 0.5302,
+      "step": 6644
+    },
+    {
+      "epoch": 2.6086441096567055,
+      "grad_norm": 0.45067265265161177,
+      "learning_rate": 4.815132010796079e-06,
+      "loss": 0.5108,
+      "step": 6645
+    },
+    {
+      "epoch": 2.6090392689552977,
+      "grad_norm": 0.4679320453519832,
+      "learning_rate": 4.815072879075409e-06,
+      "loss": 0.5069,
+      "step": 6646
+    },
+    {
+      "epoch": 2.60943442825389,
+      "grad_norm": 0.4633004240529549,
+      "learning_rate": 4.815013738262537e-06,
+      "loss": 0.5084,
+      "step": 6647
+    },
+    {
+      "epoch": 2.609829587552482,
+      "grad_norm": 0.4579293641594263,
+      "learning_rate": 4.8149545883576974e-06,
+      "loss": 0.508,
+      "step": 6648
+    },
+    {
+      "epoch": 2.6102247468510744,
+      "grad_norm": 0.44095006698161243,
+      "learning_rate": 4.814895429361122e-06,
+      "loss": 0.5141,
+      "step": 6649
+    },
+    {
+      "epoch": 2.6106199061496667,
+      "grad_norm": 0.45014307724452574,
+      "learning_rate": 4.814836261273043e-06,
+      "loss": 0.5071,
+      "step": 6650
+    },
+    {
+      "epoch": 2.611015065448259,
+      "grad_norm": 0.5653488468963971,
+      "learning_rate": 4.814777084093692e-06,
+      "loss": 0.4967,
+      "step": 6651
+    },
+    {
+      "epoch": 2.611410224746851,
+      "grad_norm": 0.4715720317621717,
+      "learning_rate": 4.814717897823303e-06,
+      "loss": 0.5069,
+      "step": 6652
+    },
+    {
+      "epoch": 2.6118053840454434,
+      "grad_norm": 0.4575070392319098,
+      "learning_rate": 4.8146587024621075e-06,
+      "loss": 0.5,
+      "step": 6653
+    },
+    {
+      "epoch": 2.6122005433440356,
+      "grad_norm": 0.449501214753915,
+      "learning_rate": 4.814599498010338e-06,
+      "loss": 0.5035,
+      "step": 6654
+    },
+    {
+      "epoch": 2.612595702642628,
+      "grad_norm": 0.4541220531434449,
+      "learning_rate": 4.814540284468227e-06,
+      "loss": 0.4795,
+      "step": 6655
+    },
+    {
+      "epoch": 2.61299086194122,
+      "grad_norm": 0.4527150728896562,
+      "learning_rate": 4.814481061836008e-06,
+      "loss": 0.4958,
+      "step": 6656
+    },
+    {
+      "epoch": 2.6133860212398123,
+      "grad_norm": 0.44537076654968044,
+      "learning_rate": 4.814421830113913e-06,
+      "loss": 0.494,
+      "step": 6657
+    },
+    {
+      "epoch": 2.6137811805384046,
+      "grad_norm": 0.4538622914542414,
+      "learning_rate": 4.814362589302174e-06,
+      "loss": 0.5144,
+      "step": 6658
+    },
+    {
+      "epoch": 2.614176339836997,
+      "grad_norm": 0.4585896393617423,
+      "learning_rate": 4.8143033394010245e-06,
+      "loss": 0.5091,
+      "step": 6659
+    },
+    {
+      "epoch": 2.614571499135589,
+      "grad_norm": 0.47488969909117135,
+      "learning_rate": 4.814244080410695e-06,
+      "loss": 0.5268,
+      "step": 6660
+    },
+    {
+      "epoch": 2.6149666584341813,
+      "grad_norm": 0.44323004486266726,
+      "learning_rate": 4.814184812331422e-06,
+      "loss": 0.4918,
+      "step": 6661
+    },
+    {
+      "epoch": 2.6153618177327735,
+      "grad_norm": 0.4707926165710535,
+      "learning_rate": 4.814125535163435e-06,
+      "loss": 0.5096,
+      "step": 6662
+    },
+    {
+      "epoch": 2.615756977031366,
+      "grad_norm": 0.446777983459746,
+      "learning_rate": 4.814066248906969e-06,
+      "loss": 0.5048,
+      "step": 6663
+    },
+    {
+      "epoch": 2.616152136329958,
+      "grad_norm": 0.4423589321854824,
+      "learning_rate": 4.8140069535622555e-06,
+      "loss": 0.4932,
+      "step": 6664
+    },
+    {
+      "epoch": 2.6165472956285503,
+      "grad_norm": 0.4578414653275977,
+      "learning_rate": 4.813947649129528e-06,
+      "loss": 0.5018,
+      "step": 6665
+    },
+    {
+      "epoch": 2.6169424549271425,
+      "grad_norm": 0.45309280010222674,
+      "learning_rate": 4.8138883356090196e-06,
+      "loss": 0.4921,
+      "step": 6666
+    },
+    {
+      "epoch": 2.6173376142257347,
+      "grad_norm": 0.45045781550198477,
+      "learning_rate": 4.813829013000963e-06,
+      "loss": 0.4947,
+      "step": 6667
+    },
+    {
+      "epoch": 2.617732773524327,
+      "grad_norm": 0.4581432466232565,
+      "learning_rate": 4.81376968130559e-06,
+      "loss": 0.5086,
+      "step": 6668
+    },
+    {
+      "epoch": 2.6181279328229192,
+      "grad_norm": 0.4746838428529649,
+      "learning_rate": 4.813710340523135e-06,
+      "loss": 0.5008,
+      "step": 6669
+    },
+    {
+      "epoch": 2.6185230921215115,
+      "grad_norm": 0.44025887937767083,
+      "learning_rate": 4.813650990653831e-06,
+      "loss": 0.4871,
+      "step": 6670
+    },
+    {
+      "epoch": 2.6189182514201037,
+      "grad_norm": 0.467331378827747,
+      "learning_rate": 4.813591631697912e-06,
+      "loss": 0.5059,
+      "step": 6671
+    },
+    {
+      "epoch": 2.619313410718696,
+      "grad_norm": 0.44926649544211894,
+      "learning_rate": 4.813532263655608e-06,
+      "loss": 0.5107,
+      "step": 6672
+    },
+    {
+      "epoch": 2.619708570017288,
+      "grad_norm": 0.4601513297437074,
+      "learning_rate": 4.813472886527155e-06,
+      "loss": 0.5235,
+      "step": 6673
+    },
+    {
+      "epoch": 2.6201037293158804,
+      "grad_norm": 0.5023320218764915,
+      "learning_rate": 4.813413500312785e-06,
+      "loss": 0.5387,
+      "step": 6674
+    },
+    {
+      "epoch": 2.6204988886144727,
+      "grad_norm": 0.45407121529937644,
+      "learning_rate": 4.813354105012732e-06,
+      "loss": 0.4986,
+      "step": 6675
+    },
+    {
+      "epoch": 2.620894047913065,
+      "grad_norm": 0.4758694797793042,
+      "learning_rate": 4.813294700627229e-06,
+      "loss": 0.5081,
+      "step": 6676
+    },
+    {
+      "epoch": 2.621289207211657,
+      "grad_norm": 0.477442584592846,
+      "learning_rate": 4.8132352871565085e-06,
+      "loss": 0.5098,
+      "step": 6677
+    },
+    {
+      "epoch": 2.6216843665102494,
+      "grad_norm": 0.4635000681832021,
+      "learning_rate": 4.813175864600805e-06,
+      "loss": 0.4921,
+      "step": 6678
+    },
+    {
+      "epoch": 2.6220795258088416,
+      "grad_norm": 0.5493110100084927,
+      "learning_rate": 4.813116432960351e-06,
+      "loss": 0.5086,
+      "step": 6679
+    },
+    {
+      "epoch": 2.622474685107434,
+      "grad_norm": 0.4593734891855389,
+      "learning_rate": 4.813056992235381e-06,
+      "loss": 0.5297,
+      "step": 6680
+    },
+    {
+      "epoch": 2.622869844406026,
+      "grad_norm": 0.4578122007316964,
+      "learning_rate": 4.812997542426126e-06,
+      "loss": 0.4811,
+      "step": 6681
+    },
+    {
+      "epoch": 2.6232650037046183,
+      "grad_norm": 0.46842927501510234,
+      "learning_rate": 4.812938083532822e-06,
+      "loss": 0.5119,
+      "step": 6682
+    },
+    {
+      "epoch": 2.6236601630032106,
+      "grad_norm": 0.46888027270874133,
+      "learning_rate": 4.812878615555702e-06,
+      "loss": 0.4904,
+      "step": 6683
+    },
+    {
+      "epoch": 2.624055322301803,
+      "grad_norm": 0.45800572704355974,
+      "learning_rate": 4.812819138495e-06,
+      "loss": 0.4936,
+      "step": 6684
+    },
+    {
+      "epoch": 2.624450481600395,
+      "grad_norm": 0.4643329241994898,
+      "learning_rate": 4.812759652350947e-06,
+      "loss": 0.5101,
+      "step": 6685
+    },
+    {
+      "epoch": 2.6248456408989873,
+      "grad_norm": 0.47372550653573653,
+      "learning_rate": 4.81270015712378e-06,
+      "loss": 0.5129,
+      "step": 6686
+    },
+    {
+      "epoch": 2.6252408001975795,
+      "grad_norm": 0.4539448907813019,
+      "learning_rate": 4.81264065281373e-06,
+      "loss": 0.5008,
+      "step": 6687
+    },
+    {
+      "epoch": 2.625635959496172,
+      "grad_norm": 0.4678129441000318,
+      "learning_rate": 4.812581139421033e-06,
+      "loss": 0.5341,
+      "step": 6688
+    },
+    {
+      "epoch": 2.626031118794764,
+      "grad_norm": 0.4425885751871649,
+      "learning_rate": 4.812521616945921e-06,
+      "loss": 0.5244,
+      "step": 6689
+    },
+    {
+      "epoch": 2.6264262780933563,
+      "grad_norm": 0.4674942573227224,
+      "learning_rate": 4.8124620853886285e-06,
+      "loss": 0.5266,
+      "step": 6690
+    },
+    {
+      "epoch": 2.6268214373919485,
+      "grad_norm": 0.4509317914278103,
+      "learning_rate": 4.8124025447493885e-06,
+      "loss": 0.5086,
+      "step": 6691
+    },
+    {
+      "epoch": 2.6272165966905408,
+      "grad_norm": 0.465509813792353,
+      "learning_rate": 4.8123429950284365e-06,
+      "loss": 0.522,
+      "step": 6692
+    },
+    {
+      "epoch": 2.627611755989133,
+      "grad_norm": 0.46175300779476025,
+      "learning_rate": 4.812283436226004e-06,
+      "loss": 0.506,
+      "step": 6693
+    },
+    {
+      "epoch": 2.6280069152877252,
+      "grad_norm": 0.455507151711426,
+      "learning_rate": 4.8122238683423276e-06,
+      "loss": 0.5059,
+      "step": 6694
+    },
+    {
+      "epoch": 2.6284020745863175,
+      "grad_norm": 0.45167680557954504,
+      "learning_rate": 4.812164291377639e-06,
+      "loss": 0.4937,
+      "step": 6695
+    },
+    {
+      "epoch": 2.6287972338849097,
+      "grad_norm": 0.462114757199228,
+      "learning_rate": 4.812104705332174e-06,
+      "loss": 0.5205,
+      "step": 6696
+    },
+    {
+      "epoch": 2.629192393183502,
+      "grad_norm": 0.4513431799853371,
+      "learning_rate": 4.812045110206165e-06,
+      "loss": 0.492,
+      "step": 6697
+    },
+    {
+      "epoch": 2.629587552482094,
+      "grad_norm": 0.44284132274625015,
+      "learning_rate": 4.811985505999846e-06,
+      "loss": 0.4751,
+      "step": 6698
+    },
+    {
+      "epoch": 2.6299827117806864,
+      "grad_norm": 0.4750386854078683,
+      "learning_rate": 4.811925892713452e-06,
+      "loss": 0.5302,
+      "step": 6699
+    },
+    {
+      "epoch": 2.6303778710792787,
+      "grad_norm": 0.4651936596747168,
+      "learning_rate": 4.811866270347219e-06,
+      "loss": 0.5187,
+      "step": 6700
+    },
+    {
+      "epoch": 2.630773030377871,
+      "grad_norm": 0.45466631519979056,
+      "learning_rate": 4.811806638901378e-06,
+      "loss": 0.4861,
+      "step": 6701
+    },
+    {
+      "epoch": 2.6311681896764636,
+      "grad_norm": 0.4480365156981654,
+      "learning_rate": 4.8117469983761636e-06,
+      "loss": 0.4928,
+      "step": 6702
+    },
+    {
+      "epoch": 2.631563348975056,
+      "grad_norm": 0.45596295905822454,
+      "learning_rate": 4.811687348771811e-06,
+      "loss": 0.5143,
+      "step": 6703
+    },
+    {
+      "epoch": 2.631958508273648,
+      "grad_norm": 0.4612335510523021,
+      "learning_rate": 4.811627690088555e-06,
+      "loss": 0.5039,
+      "step": 6704
+    },
+    {
+      "epoch": 2.6323536675722403,
+      "grad_norm": 0.4495637737856094,
+      "learning_rate": 4.811568022326628e-06,
+      "loss": 0.5014,
+      "step": 6705
+    },
+    {
+      "epoch": 2.6327488268708326,
+      "grad_norm": 0.4571383147738367,
+      "learning_rate": 4.811508345486267e-06,
+      "loss": 0.4987,
+      "step": 6706
+    },
+    {
+      "epoch": 2.633143986169425,
+      "grad_norm": 0.46620616655944114,
+      "learning_rate": 4.811448659567703e-06,
+      "loss": 0.5157,
+      "step": 6707
+    },
+    {
+      "epoch": 2.633539145468017,
+      "grad_norm": 0.692201266331602,
+      "learning_rate": 4.811388964571173e-06,
+      "loss": 0.502,
+      "step": 6708
+    },
+    {
+      "epoch": 2.6339343047666093,
+      "grad_norm": 0.452425821090267,
+      "learning_rate": 4.811329260496911e-06,
+      "loss": 0.5082,
+      "step": 6709
+    },
+    {
+      "epoch": 2.6343294640652015,
+      "grad_norm": 0.4484831548580827,
+      "learning_rate": 4.811269547345151e-06,
+      "loss": 0.4953,
+      "step": 6710
+    },
+    {
+      "epoch": 2.6347246233637938,
+      "grad_norm": 0.5286805004399712,
+      "learning_rate": 4.8112098251161275e-06,
+      "loss": 0.5059,
+      "step": 6711
+    },
+    {
+      "epoch": 2.635119782662386,
+      "grad_norm": 0.47242345842337413,
+      "learning_rate": 4.811150093810076e-06,
+      "loss": 0.515,
+      "step": 6712
+    },
+    {
+      "epoch": 2.6355149419609782,
+      "grad_norm": 0.45817282138512155,
+      "learning_rate": 4.81109035342723e-06,
+      "loss": 0.4959,
+      "step": 6713
+    },
+    {
+      "epoch": 2.6359101012595705,
+      "grad_norm": 0.47033111569825325,
+      "learning_rate": 4.811030603967824e-06,
+      "loss": 0.5047,
+      "step": 6714
+    },
+    {
+      "epoch": 2.6363052605581627,
+      "grad_norm": 0.4574013018432309,
+      "learning_rate": 4.810970845432094e-06,
+      "loss": 0.5121,
+      "step": 6715
+    },
+    {
+      "epoch": 2.636700419856755,
+      "grad_norm": 0.47733427395045364,
+      "learning_rate": 4.810911077820273e-06,
+      "loss": 0.4993,
+      "step": 6716
+    },
+    {
+      "epoch": 2.637095579155347,
+      "grad_norm": 0.45438462788409517,
+      "learning_rate": 4.8108513011325965e-06,
+      "loss": 0.496,
+      "step": 6717
+    },
+    {
+      "epoch": 2.6374907384539394,
+      "grad_norm": 0.46176365204286124,
+      "learning_rate": 4.8107915153693e-06,
+      "loss": 0.5186,
+      "step": 6718
+    },
+    {
+      "epoch": 2.6378858977525317,
+      "grad_norm": 0.46309358892460495,
+      "learning_rate": 4.810731720530617e-06,
+      "loss": 0.5083,
+      "step": 6719
+    },
+    {
+      "epoch": 2.638281057051124,
+      "grad_norm": 0.4712899021409001,
+      "learning_rate": 4.810671916616783e-06,
+      "loss": 0.5113,
+      "step": 6720
+    },
+    {
+      "epoch": 2.638676216349716,
+      "grad_norm": 0.4621511341873441,
+      "learning_rate": 4.810612103628033e-06,
+      "loss": 0.505,
+      "step": 6721
+    },
+    {
+      "epoch": 2.6390713756483084,
+      "grad_norm": 0.4484479850838346,
+      "learning_rate": 4.810552281564602e-06,
+      "loss": 0.498,
+      "step": 6722
+    },
+    {
+      "epoch": 2.6394665349469006,
+      "grad_norm": 0.4447686728319573,
+      "learning_rate": 4.8104924504267245e-06,
+      "loss": 0.4786,
+      "step": 6723
+    },
+    {
+      "epoch": 2.639861694245493,
+      "grad_norm": 0.4587588719398148,
+      "learning_rate": 4.810432610214636e-06,
+      "loss": 0.5008,
+      "step": 6724
+    },
+    {
+      "epoch": 2.640256853544085,
+      "grad_norm": 0.46171851458075386,
+      "learning_rate": 4.81037276092857e-06,
+      "loss": 0.5104,
+      "step": 6725
+    },
+    {
+      "epoch": 2.6406520128426774,
+      "grad_norm": 0.47362051128187516,
+      "learning_rate": 4.810312902568763e-06,
+      "loss": 0.5179,
+      "step": 6726
+    },
+    {
+      "epoch": 2.6410471721412696,
+      "grad_norm": 0.4707818961914056,
+      "learning_rate": 4.81025303513545e-06,
+      "loss": 0.5152,
+      "step": 6727
+    },
+    {
+      "epoch": 2.641442331439862,
+      "grad_norm": 0.4487464299804814,
+      "learning_rate": 4.810193158628867e-06,
+      "loss": 0.4998,
+      "step": 6728
+    },
+    {
+      "epoch": 2.641837490738454,
+      "grad_norm": 0.4550736327280535,
+      "learning_rate": 4.810133273049247e-06,
+      "loss": 0.5179,
+      "step": 6729
+    },
+    {
+      "epoch": 2.6422326500370463,
+      "grad_norm": 0.47050227006026263,
+      "learning_rate": 4.810073378396827e-06,
+      "loss": 0.5214,
+      "step": 6730
+    },
+    {
+      "epoch": 2.6426278093356386,
+      "grad_norm": 0.4519033384468275,
+      "learning_rate": 4.81001347467184e-06,
+      "loss": 0.4998,
+      "step": 6731
+    },
+    {
+      "epoch": 2.643022968634231,
+      "grad_norm": 0.451105123921464,
+      "learning_rate": 4.809953561874525e-06,
+      "loss": 0.5057,
+      "step": 6732
+    },
+    {
+      "epoch": 2.643418127932823,
+      "grad_norm": 0.4556875869733165,
+      "learning_rate": 4.8098936400051145e-06,
+      "loss": 0.5156,
+      "step": 6733
+    },
+    {
+      "epoch": 2.6438132872314153,
+      "grad_norm": 0.45194215128861687,
+      "learning_rate": 4.809833709063844e-06,
+      "loss": 0.5221,
+      "step": 6734
+    },
+    {
+      "epoch": 2.6442084465300075,
+      "grad_norm": 0.4695069511637479,
+      "learning_rate": 4.809773769050948e-06,
+      "loss": 0.5203,
+      "step": 6735
+    },
+    {
+      "epoch": 2.6446036058285998,
+      "grad_norm": 0.4792987985028994,
+      "learning_rate": 4.809713819966665e-06,
+      "loss": 0.5284,
+      "step": 6736
+    },
+    {
+      "epoch": 2.644998765127192,
+      "grad_norm": 0.45769602212488414,
+      "learning_rate": 4.809653861811228e-06,
+      "loss": 0.4851,
+      "step": 6737
+    },
+    {
+      "epoch": 2.6453939244257842,
+      "grad_norm": 0.4424740957506516,
+      "learning_rate": 4.809593894584873e-06,
+      "loss": 0.4982,
+      "step": 6738
+    },
+    {
+      "epoch": 2.6457890837243765,
+      "grad_norm": 0.4696493235047474,
+      "learning_rate": 4.809533918287836e-06,
+      "loss": 0.4989,
+      "step": 6739
+    },
+    {
+      "epoch": 2.6461842430229687,
+      "grad_norm": 0.46456830773516294,
+      "learning_rate": 4.809473932920352e-06,
+      "loss": 0.4941,
+      "step": 6740
+    },
+    {
+      "epoch": 2.646579402321561,
+      "grad_norm": 0.47014248694056754,
+      "learning_rate": 4.809413938482657e-06,
+      "loss": 0.5156,
+      "step": 6741
+    },
+    {
+      "epoch": 2.646974561620153,
+      "grad_norm": 0.4356406681706022,
+      "learning_rate": 4.809353934974987e-06,
+      "loss": 0.4908,
+      "step": 6742
+    },
+    {
+      "epoch": 2.6473697209187455,
+      "grad_norm": 0.4862338209729092,
+      "learning_rate": 4.809293922397576e-06,
+      "loss": 0.5048,
+      "step": 6743
+    },
+    {
+      "epoch": 2.6477648802173377,
+      "grad_norm": 0.4575448187275317,
+      "learning_rate": 4.80923390075066e-06,
+      "loss": 0.5134,
+      "step": 6744
+    },
+    {
+      "epoch": 2.64816003951593,
+      "grad_norm": 0.4564366850228013,
+      "learning_rate": 4.809173870034477e-06,
+      "loss": 0.5085,
+      "step": 6745
+    },
+    {
+      "epoch": 2.648555198814522,
+      "grad_norm": 0.48219599498932664,
+      "learning_rate": 4.809113830249261e-06,
+      "loss": 0.5221,
+      "step": 6746
+    },
+    {
+      "epoch": 2.6489503581131144,
+      "grad_norm": 0.4453567920915501,
+      "learning_rate": 4.809053781395248e-06,
+      "loss": 0.509,
+      "step": 6747
+    },
+    {
+      "epoch": 2.6493455174117067,
+      "grad_norm": 0.4724611482066963,
+      "learning_rate": 4.8089937234726734e-06,
+      "loss": 0.4851,
+      "step": 6748
+    },
+    {
+      "epoch": 2.649740676710299,
+      "grad_norm": 0.4444318838338744,
+      "learning_rate": 4.808933656481774e-06,
+      "loss": 0.5097,
+      "step": 6749
+    },
+    {
+      "epoch": 2.650135836008891,
+      "grad_norm": 0.46372690064711497,
+      "learning_rate": 4.808873580422785e-06,
+      "loss": 0.5011,
+      "step": 6750
+    },
+    {
+      "epoch": 2.6505309953074834,
+      "grad_norm": 0.45321412299435093,
+      "learning_rate": 4.808813495295942e-06,
+      "loss": 0.4916,
+      "step": 6751
+    },
+    {
+      "epoch": 2.6509261546060756,
+      "grad_norm": 0.47068167177315234,
+      "learning_rate": 4.808753401101483e-06,
+      "loss": 0.5067,
+      "step": 6752
+    },
+    {
+      "epoch": 2.651321313904668,
+      "grad_norm": 0.4635431012232405,
+      "learning_rate": 4.808693297839642e-06,
+      "loss": 0.5182,
+      "step": 6753
+    },
+    {
+      "epoch": 2.65171647320326,
+      "grad_norm": 0.4412239947020369,
+      "learning_rate": 4.8086331855106546e-06,
+      "loss": 0.5035,
+      "step": 6754
+    },
+    {
+      "epoch": 2.6521116325018523,
+      "grad_norm": 0.43995031164220066,
+      "learning_rate": 4.80857306411476e-06,
+      "loss": 0.5028,
+      "step": 6755
+    },
+    {
+      "epoch": 2.6525067918004446,
+      "grad_norm": 0.4591057467664625,
+      "learning_rate": 4.808512933652191e-06,
+      "loss": 0.5049,
+      "step": 6756
+    },
+    {
+      "epoch": 2.652901951099037,
+      "grad_norm": 0.4571034390087959,
+      "learning_rate": 4.808452794123184e-06,
+      "loss": 0.5272,
+      "step": 6757
+    },
+    {
+      "epoch": 2.653297110397629,
+      "grad_norm": 0.4518130469202257,
+      "learning_rate": 4.8083926455279775e-06,
+      "loss": 0.5165,
+      "step": 6758
+    },
+    {
+      "epoch": 2.6536922696962213,
+      "grad_norm": 0.439662265546044,
+      "learning_rate": 4.808332487866806e-06,
+      "loss": 0.4949,
+      "step": 6759
+    },
+    {
+      "epoch": 2.6540874289948135,
+      "grad_norm": 0.46867256715880273,
+      "learning_rate": 4.808272321139907e-06,
+      "loss": 0.5098,
+      "step": 6760
+    },
+    {
+      "epoch": 2.6544825882934058,
+      "grad_norm": 0.4407389861497981,
+      "learning_rate": 4.808212145347515e-06,
+      "loss": 0.4803,
+      "step": 6761
+    },
+    {
+      "epoch": 2.654877747591998,
+      "grad_norm": 0.45779304183120295,
+      "learning_rate": 4.808151960489867e-06,
+      "loss": 0.4899,
+      "step": 6762
+    },
+    {
+      "epoch": 2.6552729068905903,
+      "grad_norm": 0.4565973133474136,
+      "learning_rate": 4.808091766567201e-06,
+      "loss": 0.5085,
+      "step": 6763
+    },
+    {
+      "epoch": 2.6556680661891825,
+      "grad_norm": 0.468953636638347,
+      "learning_rate": 4.8080315635797515e-06,
+      "loss": 0.5232,
+      "step": 6764
+    },
+    {
+      "epoch": 2.6560632254877747,
+      "grad_norm": 0.44984729188558875,
+      "learning_rate": 4.807971351527755e-06,
+      "loss": 0.52,
+      "step": 6765
+    },
+    {
+      "epoch": 2.656458384786367,
+      "grad_norm": 0.46345262945947907,
+      "learning_rate": 4.807911130411449e-06,
+      "loss": 0.4932,
+      "step": 6766
+    },
+    {
+      "epoch": 2.656853544084959,
+      "grad_norm": 0.46323751843274685,
+      "learning_rate": 4.80785090023107e-06,
+      "loss": 0.5168,
+      "step": 6767
+    },
+    {
+      "epoch": 2.6572487033835515,
+      "grad_norm": 0.4626841241252618,
+      "learning_rate": 4.807790660986854e-06,
+      "loss": 0.5224,
+      "step": 6768
+    },
+    {
+      "epoch": 2.6576438626821437,
+      "grad_norm": 0.4413542814112758,
+      "learning_rate": 4.807730412679037e-06,
+      "loss": 0.5105,
+      "step": 6769
+    },
+    {
+      "epoch": 2.658039021980736,
+      "grad_norm": 0.45515913790444384,
+      "learning_rate": 4.807670155307857e-06,
+      "loss": 0.4853,
+      "step": 6770
+    },
+    {
+      "epoch": 2.658434181279328,
+      "grad_norm": 0.45080574051033717,
+      "learning_rate": 4.807609888873548e-06,
+      "loss": 0.5232,
+      "step": 6771
+    },
+    {
+      "epoch": 2.6588293405779204,
+      "grad_norm": 0.45056542118018017,
+      "learning_rate": 4.807549613376351e-06,
+      "loss": 0.503,
+      "step": 6772
+    },
+    {
+      "epoch": 2.6592244998765127,
+      "grad_norm": 0.45084980258696417,
+      "learning_rate": 4.8074893288164995e-06,
+      "loss": 0.5082,
+      "step": 6773
+    },
+    {
+      "epoch": 2.659619659175105,
+      "grad_norm": 0.46000748014035947,
+      "learning_rate": 4.80742903519423e-06,
+      "loss": 0.5126,
+      "step": 6774
+    },
+    {
+      "epoch": 2.660014818473697,
+      "grad_norm": 0.6125057638224274,
+      "learning_rate": 4.807368732509782e-06,
+      "loss": 0.516,
+      "step": 6775
+    },
+    {
+      "epoch": 2.6604099777722894,
+      "grad_norm": 0.4465782067051833,
+      "learning_rate": 4.8073084207633895e-06,
+      "loss": 0.5017,
+      "step": 6776
+    },
+    {
+      "epoch": 2.6608051370708816,
+      "grad_norm": 0.44598562336646147,
+      "learning_rate": 4.807248099955291e-06,
+      "loss": 0.4966,
+      "step": 6777
+    },
+    {
+      "epoch": 2.661200296369474,
+      "grad_norm": 0.4617546372416006,
+      "learning_rate": 4.807187770085724e-06,
+      "loss": 0.5003,
+      "step": 6778
+    },
+    {
+      "epoch": 2.661595455668066,
+      "grad_norm": 0.4676309244036775,
+      "learning_rate": 4.807127431154923e-06,
+      "loss": 0.5039,
+      "step": 6779
+    },
+    {
+      "epoch": 2.6619906149666583,
+      "grad_norm": 0.5277411293225999,
+      "learning_rate": 4.807067083163127e-06,
+      "loss": 0.518,
+      "step": 6780
+    },
+    {
+      "epoch": 2.6623857742652506,
+      "grad_norm": 0.45429895891466116,
+      "learning_rate": 4.8070067261105725e-06,
+      "loss": 0.4982,
+      "step": 6781
+    },
+    {
+      "epoch": 2.662780933563843,
+      "grad_norm": 0.4464728729832292,
+      "learning_rate": 4.806946359997496e-06,
+      "loss": 0.5189,
+      "step": 6782
+    },
+    {
+      "epoch": 2.663176092862435,
+      "grad_norm": 0.4715484427485262,
+      "learning_rate": 4.806885984824136e-06,
+      "loss": 0.5117,
+      "step": 6783
+    },
+    {
+      "epoch": 2.6635712521610273,
+      "grad_norm": 0.45900610189295865,
+      "learning_rate": 4.8068256005907275e-06,
+      "loss": 0.5023,
+      "step": 6784
+    },
+    {
+      "epoch": 2.6639664114596195,
+      "grad_norm": 0.45019303096695007,
+      "learning_rate": 4.80676520729751e-06,
+      "loss": 0.5113,
+      "step": 6785
+    },
+    {
+      "epoch": 2.664361570758212,
+      "grad_norm": 0.46748176211191267,
+      "learning_rate": 4.806704804944719e-06,
+      "loss": 0.5164,
+      "step": 6786
+    },
+    {
+      "epoch": 2.664756730056804,
+      "grad_norm": 0.46855906191928937,
+      "learning_rate": 4.8066443935325926e-06,
+      "loss": 0.5002,
+      "step": 6787
+    },
+    {
+      "epoch": 2.6651518893553963,
+      "grad_norm": 0.452910775505456,
+      "learning_rate": 4.806583973061367e-06,
+      "loss": 0.5156,
+      "step": 6788
+    },
+    {
+      "epoch": 2.6655470486539885,
+      "grad_norm": 0.47440038216416675,
+      "learning_rate": 4.80652354353128e-06,
+      "loss": 0.5303,
+      "step": 6789
+    },
+    {
+      "epoch": 2.6659422079525807,
+      "grad_norm": 0.5924097415433517,
+      "learning_rate": 4.806463104942569e-06,
+      "loss": 0.5167,
+      "step": 6790
+    },
+    {
+      "epoch": 2.666337367251173,
+      "grad_norm": 0.438886847398818,
+      "learning_rate": 4.806402657295472e-06,
+      "loss": 0.5003,
+      "step": 6791
+    },
+    {
+      "epoch": 2.6667325265497652,
+      "grad_norm": 0.46609047736390946,
+      "learning_rate": 4.806342200590227e-06,
+      "loss": 0.5128,
+      "step": 6792
+    },
+    {
+      "epoch": 2.6671276858483575,
+      "grad_norm": 0.4632346721128658,
+      "learning_rate": 4.8062817348270684e-06,
+      "loss": 0.4954,
+      "step": 6793
+    },
+    {
+      "epoch": 2.6675228451469497,
+      "grad_norm": 0.4608854558412661,
+      "learning_rate": 4.806221260006237e-06,
+      "loss": 0.505,
+      "step": 6794
+    },
+    {
+      "epoch": 2.667918004445542,
+      "grad_norm": 0.46630121452068407,
+      "learning_rate": 4.806160776127968e-06,
+      "loss": 0.5108,
+      "step": 6795
+    },
+    {
+      "epoch": 2.668313163744134,
+      "grad_norm": 0.45531938998001714,
+      "learning_rate": 4.806100283192501e-06,
+      "loss": 0.4968,
+      "step": 6796
+    },
+    {
+      "epoch": 2.6687083230427264,
+      "grad_norm": 0.4555861650279177,
+      "learning_rate": 4.806039781200071e-06,
+      "loss": 0.5088,
+      "step": 6797
+    },
+    {
+      "epoch": 2.6691034823413187,
+      "grad_norm": 0.4803868219645436,
+      "learning_rate": 4.805979270150918e-06,
+      "loss": 0.5122,
+      "step": 6798
+    },
+    {
+      "epoch": 2.669498641639911,
+      "grad_norm": 0.47581987750590815,
+      "learning_rate": 4.805918750045278e-06,
+      "loss": 0.4929,
+      "step": 6799
+    },
+    {
+      "epoch": 2.669893800938503,
+      "grad_norm": 0.449650149880212,
+      "learning_rate": 4.80585822088339e-06,
+      "loss": 0.5197,
+      "step": 6800
+    },
+    {
+      "epoch": 2.6702889602370954,
+      "grad_norm": 0.43855063136998546,
+      "learning_rate": 4.8057976826654906e-06,
+      "loss": 0.5101,
+      "step": 6801
+    },
+    {
+      "epoch": 2.6706841195356876,
+      "grad_norm": 0.4720486302585973,
+      "learning_rate": 4.805737135391818e-06,
+      "loss": 0.5262,
+      "step": 6802
+    },
+    {
+      "epoch": 2.67107927883428,
+      "grad_norm": 0.44945472494966127,
+      "learning_rate": 4.80567657906261e-06,
+      "loss": 0.5048,
+      "step": 6803
+    },
+    {
+      "epoch": 2.671474438132872,
+      "grad_norm": 0.4669554725656825,
+      "learning_rate": 4.8056160136781055e-06,
+      "loss": 0.5115,
+      "step": 6804
+    },
+    {
+      "epoch": 2.6718695974314643,
+      "grad_norm": 0.4403150506327582,
+      "learning_rate": 4.805555439238541e-06,
+      "loss": 0.5031,
+      "step": 6805
+    },
+    {
+      "epoch": 2.6722647567300566,
+      "grad_norm": 0.5000895923389528,
+      "learning_rate": 4.805494855744154e-06,
+      "loss": 0.5302,
+      "step": 6806
+    },
+    {
+      "epoch": 2.672659916028649,
+      "grad_norm": 0.47576343934993737,
+      "learning_rate": 4.8054342631951836e-06,
+      "loss": 0.4895,
+      "step": 6807
+    },
+    {
+      "epoch": 2.673055075327241,
+      "grad_norm": 0.477265617906093,
+      "learning_rate": 4.8053736615918675e-06,
+      "loss": 0.5007,
+      "step": 6808
+    },
+    {
+      "epoch": 2.6734502346258333,
+      "grad_norm": 0.4485511317732285,
+      "learning_rate": 4.8053130509344434e-06,
+      "loss": 0.5081,
+      "step": 6809
+    },
+    {
+      "epoch": 2.6738453939244256,
+      "grad_norm": 0.45923352589367195,
+      "learning_rate": 4.8052524312231494e-06,
+      "loss": 0.5056,
+      "step": 6810
+    },
+    {
+      "epoch": 2.674240553223018,
+      "grad_norm": 0.4661199256713386,
+      "learning_rate": 4.8051918024582235e-06,
+      "loss": 0.506,
+      "step": 6811
+    },
+    {
+      "epoch": 2.67463571252161,
+      "grad_norm": 0.4610083722334732,
+      "learning_rate": 4.8051311646399045e-06,
+      "loss": 0.5033,
+      "step": 6812
+    },
+    {
+      "epoch": 2.6750308718202023,
+      "grad_norm": 0.43652772614403434,
+      "learning_rate": 4.80507051776843e-06,
+      "loss": 0.5071,
+      "step": 6813
+    },
+    {
+      "epoch": 2.6754260311187945,
+      "grad_norm": 0.4691550437232749,
+      "learning_rate": 4.805009861844038e-06,
+      "loss": 0.527,
+      "step": 6814
+    },
+    {
+      "epoch": 2.6758211904173868,
+      "grad_norm": 0.4560452636416422,
+      "learning_rate": 4.804949196866967e-06,
+      "loss": 0.4908,
+      "step": 6815
+    },
+    {
+      "epoch": 2.676216349715979,
+      "grad_norm": 0.48133219998650245,
+      "learning_rate": 4.8048885228374556e-06,
+      "loss": 0.5118,
+      "step": 6816
+    },
+    {
+      "epoch": 2.6766115090145712,
+      "grad_norm": 0.452233986235823,
+      "learning_rate": 4.804827839755741e-06,
+      "loss": 0.4915,
+      "step": 6817
+    },
+    {
+      "epoch": 2.6770066683131635,
+      "grad_norm": 0.4631514836434119,
+      "learning_rate": 4.804767147622062e-06,
+      "loss": 0.529,
+      "step": 6818
+    },
+    {
+      "epoch": 2.6774018276117557,
+      "grad_norm": 0.4502944964366507,
+      "learning_rate": 4.804706446436658e-06,
+      "loss": 0.5077,
+      "step": 6819
+    },
+    {
+      "epoch": 2.6777969869103484,
+      "grad_norm": 0.46421201437573734,
+      "learning_rate": 4.8046457361997655e-06,
+      "loss": 0.5185,
+      "step": 6820
+    },
+    {
+      "epoch": 2.6781921462089406,
+      "grad_norm": 0.4613890649783742,
+      "learning_rate": 4.804585016911625e-06,
+      "loss": 0.5138,
+      "step": 6821
+    },
+    {
+      "epoch": 2.678587305507533,
+      "grad_norm": 0.4456774139985832,
+      "learning_rate": 4.8045242885724735e-06,
+      "loss": 0.5217,
+      "step": 6822
+    },
+    {
+      "epoch": 2.678982464806125,
+      "grad_norm": 0.47237905847343165,
+      "learning_rate": 4.80446355118255e-06,
+      "loss": 0.5191,
+      "step": 6823
+    },
+    {
+      "epoch": 2.6793776241047174,
+      "grad_norm": 0.4974035566555643,
+      "learning_rate": 4.804402804742093e-06,
+      "loss": 0.5512,
+      "step": 6824
+    },
+    {
+      "epoch": 2.6797727834033096,
+      "grad_norm": 0.47581595796135917,
+      "learning_rate": 4.804342049251341e-06,
+      "loss": 0.5303,
+      "step": 6825
+    },
+    {
+      "epoch": 2.680167942701902,
+      "grad_norm": 0.47309784832193486,
+      "learning_rate": 4.804281284710534e-06,
+      "loss": 0.5315,
+      "step": 6826
+    },
+    {
+      "epoch": 2.680563102000494,
+      "grad_norm": 0.44291199384154933,
+      "learning_rate": 4.804220511119908e-06,
+      "loss": 0.5095,
+      "step": 6827
+    },
+    {
+      "epoch": 2.6809582612990863,
+      "grad_norm": 0.4418663874638194,
+      "learning_rate": 4.804159728479703e-06,
+      "loss": 0.5,
+      "step": 6828
+    },
+    {
+      "epoch": 2.6813534205976786,
+      "grad_norm": 0.46327736667282093,
+      "learning_rate": 4.804098936790158e-06,
+      "loss": 0.5141,
+      "step": 6829
+    },
+    {
+      "epoch": 2.681748579896271,
+      "grad_norm": 0.4479767367906708,
+      "learning_rate": 4.804038136051512e-06,
+      "loss": 0.5011,
+      "step": 6830
+    },
+    {
+      "epoch": 2.682143739194863,
+      "grad_norm": 0.46247506860437304,
+      "learning_rate": 4.803977326264003e-06,
+      "loss": 0.5114,
+      "step": 6831
+    },
+    {
+      "epoch": 2.6825388984934553,
+      "grad_norm": 0.46141174971832466,
+      "learning_rate": 4.803916507427869e-06,
+      "loss": 0.5141,
+      "step": 6832
+    },
+    {
+      "epoch": 2.6829340577920475,
+      "grad_norm": 0.4873722164912079,
+      "learning_rate": 4.803855679543352e-06,
+      "loss": 0.4998,
+      "step": 6833
+    },
+    {
+      "epoch": 2.6833292170906398,
+      "grad_norm": 0.46403561177788066,
+      "learning_rate": 4.803794842610687e-06,
+      "loss": 0.5164,
+      "step": 6834
+    },
+    {
+      "epoch": 2.683724376389232,
+      "grad_norm": 0.4456637637514536,
+      "learning_rate": 4.803733996630116e-06,
+      "loss": 0.5018,
+      "step": 6835
+    },
+    {
+      "epoch": 2.6841195356878242,
+      "grad_norm": 0.4563297591462206,
+      "learning_rate": 4.803673141601876e-06,
+      "loss": 0.4958,
+      "step": 6836
+    },
+    {
+      "epoch": 2.6845146949864165,
+      "grad_norm": 0.46433291682819583,
+      "learning_rate": 4.803612277526207e-06,
+      "loss": 0.5041,
+      "step": 6837
+    },
+    {
+      "epoch": 2.6849098542850087,
+      "grad_norm": 0.45384038987818476,
+      "learning_rate": 4.803551404403348e-06,
+      "loss": 0.5029,
+      "step": 6838
+    },
+    {
+      "epoch": 2.685305013583601,
+      "grad_norm": 0.4396868856560902,
+      "learning_rate": 4.803490522233538e-06,
+      "loss": 0.4979,
+      "step": 6839
+    },
+    {
+      "epoch": 2.685700172882193,
+      "grad_norm": 0.45017056466155314,
+      "learning_rate": 4.803429631017016e-06,
+      "loss": 0.5048,
+      "step": 6840
+    },
+    {
+      "epoch": 2.6860953321807854,
+      "grad_norm": 0.4574075515113022,
+      "learning_rate": 4.8033687307540214e-06,
+      "loss": 0.5155,
+      "step": 6841
+    },
+    {
+      "epoch": 2.6864904914793777,
+      "grad_norm": 0.4618670159293962,
+      "learning_rate": 4.803307821444793e-06,
+      "loss": 0.516,
+      "step": 6842
+    },
+    {
+      "epoch": 2.68688565077797,
+      "grad_norm": 0.4464411508006451,
+      "learning_rate": 4.803246903089569e-06,
+      "loss": 0.5171,
+      "step": 6843
+    },
+    {
+      "epoch": 2.687280810076562,
+      "grad_norm": 0.45337545621080044,
+      "learning_rate": 4.80318597568859e-06,
+      "loss": 0.5238,
+      "step": 6844
+    },
+    {
+      "epoch": 2.6876759693751544,
+      "grad_norm": 0.4706459371294516,
+      "learning_rate": 4.803125039242096e-06,
+      "loss": 0.5214,
+      "step": 6845
+    },
+    {
+      "epoch": 2.6880711286737466,
+      "grad_norm": 0.444744828944862,
+      "learning_rate": 4.8030640937503245e-06,
+      "loss": 0.5177,
+      "step": 6846
+    },
+    {
+      "epoch": 2.688466287972339,
+      "grad_norm": 0.4650828362886886,
+      "learning_rate": 4.803003139213517e-06,
+      "loss": 0.5075,
+      "step": 6847
+    },
+    {
+      "epoch": 2.688861447270931,
+      "grad_norm": 0.45639383031758396,
+      "learning_rate": 4.802942175631911e-06,
+      "loss": 0.5011,
+      "step": 6848
+    },
+    {
+      "epoch": 2.6892566065695234,
+      "grad_norm": 0.466428988530897,
+      "learning_rate": 4.802881203005746e-06,
+      "loss": 0.5139,
+      "step": 6849
+    },
+    {
+      "epoch": 2.6896517658681156,
+      "grad_norm": 0.4525428401741202,
+      "learning_rate": 4.802820221335263e-06,
+      "loss": 0.4978,
+      "step": 6850
+    },
+    {
+      "epoch": 2.690046925166708,
+      "grad_norm": 0.45716169122612815,
+      "learning_rate": 4.8027592306206995e-06,
+      "loss": 0.501,
+      "step": 6851
+    },
+    {
+      "epoch": 2.6904420844653,
+      "grad_norm": 0.48662071727200495,
+      "learning_rate": 4.802698230862296e-06,
+      "loss": 0.5265,
+      "step": 6852
+    },
+    {
+      "epoch": 2.6908372437638923,
+      "grad_norm": 0.4623332955672627,
+      "learning_rate": 4.802637222060293e-06,
+      "loss": 0.5163,
+      "step": 6853
+    },
+    {
+      "epoch": 2.6912324030624846,
+      "grad_norm": 0.45648458131398784,
+      "learning_rate": 4.802576204214928e-06,
+      "loss": 0.515,
+      "step": 6854
+    },
+    {
+      "epoch": 2.691627562361077,
+      "grad_norm": 0.4551657895457975,
+      "learning_rate": 4.802515177326444e-06,
+      "loss": 0.5282,
+      "step": 6855
+    },
+    {
+      "epoch": 2.692022721659669,
+      "grad_norm": 0.4475910585637045,
+      "learning_rate": 4.802454141395076e-06,
+      "loss": 0.5113,
+      "step": 6856
+    },
+    {
+      "epoch": 2.6924178809582613,
+      "grad_norm": 0.4558226749142037,
+      "learning_rate": 4.802393096421068e-06,
+      "loss": 0.5123,
+      "step": 6857
+    },
+    {
+      "epoch": 2.6928130402568535,
+      "grad_norm": 0.4486376587821075,
+      "learning_rate": 4.802332042404657e-06,
+      "loss": 0.5117,
+      "step": 6858
+    },
+    {
+      "epoch": 2.6932081995554458,
+      "grad_norm": 0.44884818601499643,
+      "learning_rate": 4.8022709793460846e-06,
+      "loss": 0.5084,
+      "step": 6859
+    },
+    {
+      "epoch": 2.693603358854038,
+      "grad_norm": 0.5475296537368842,
+      "learning_rate": 4.8022099072455896e-06,
+      "loss": 0.5086,
+      "step": 6860
+    },
+    {
+      "epoch": 2.6939985181526303,
+      "grad_norm": 0.4747377379257253,
+      "learning_rate": 4.802148826103412e-06,
+      "loss": 0.5109,
+      "step": 6861
+    },
+    {
+      "epoch": 2.6943936774512225,
+      "grad_norm": 0.47187115055855244,
+      "learning_rate": 4.802087735919792e-06,
+      "loss": 0.5107,
+      "step": 6862
+    },
+    {
+      "epoch": 2.6947888367498147,
+      "grad_norm": 0.45581653066216854,
+      "learning_rate": 4.802026636694969e-06,
+      "loss": 0.5092,
+      "step": 6863
+    },
+    {
+      "epoch": 2.695183996048407,
+      "grad_norm": 0.45531582354330513,
+      "learning_rate": 4.8019655284291825e-06,
+      "loss": 0.4925,
+      "step": 6864
+    },
+    {
+      "epoch": 2.695579155346999,
+      "grad_norm": 0.5224115146598348,
+      "learning_rate": 4.801904411122675e-06,
+      "loss": 0.5203,
+      "step": 6865
+    },
+    {
+      "epoch": 2.6959743146455915,
+      "grad_norm": 0.4448869457231406,
+      "learning_rate": 4.8018432847756825e-06,
+      "loss": 0.4948,
+      "step": 6866
+    },
+    {
+      "epoch": 2.6963694739441837,
+      "grad_norm": 0.46002467750035936,
+      "learning_rate": 4.801782149388448e-06,
+      "loss": 0.5056,
+      "step": 6867
+    },
+    {
+      "epoch": 2.696764633242776,
+      "grad_norm": 0.47172532896971114,
+      "learning_rate": 4.801721004961213e-06,
+      "loss": 0.5075,
+      "step": 6868
+    },
+    {
+      "epoch": 2.697159792541368,
+      "grad_norm": 0.45878868214555063,
+      "learning_rate": 4.8016598514942135e-06,
+      "loss": 0.4925,
+      "step": 6869
+    },
+    {
+      "epoch": 2.6975549518399604,
+      "grad_norm": 0.4544323936183125,
+      "learning_rate": 4.801598688987692e-06,
+      "loss": 0.5035,
+      "step": 6870
+    },
+    {
+      "epoch": 2.6979501111385527,
+      "grad_norm": 0.4774600541049669,
+      "learning_rate": 4.801537517441889e-06,
+      "loss": 0.5106,
+      "step": 6871
+    },
+    {
+      "epoch": 2.698345270437145,
+      "grad_norm": 0.4625044065364173,
+      "learning_rate": 4.801476336857043e-06,
+      "loss": 0.4934,
+      "step": 6872
+    },
+    {
+      "epoch": 2.698740429735737,
+      "grad_norm": 0.45276021101004094,
+      "learning_rate": 4.801415147233397e-06,
+      "loss": 0.4948,
+      "step": 6873
+    },
+    {
+      "epoch": 2.6991355890343294,
+      "grad_norm": 0.46016046750830586,
+      "learning_rate": 4.801353948571189e-06,
+      "loss": 0.5105,
+      "step": 6874
+    },
+    {
+      "epoch": 2.6995307483329216,
+      "grad_norm": 0.46937028530160213,
+      "learning_rate": 4.801292740870661e-06,
+      "loss": 0.5325,
+      "step": 6875
+    },
+    {
+      "epoch": 2.699925907631514,
+      "grad_norm": 0.43934115205425334,
+      "learning_rate": 4.801231524132052e-06,
+      "loss": 0.5062,
+      "step": 6876
+    },
+    {
+      "epoch": 2.700321066930106,
+      "grad_norm": 0.4469967985779925,
+      "learning_rate": 4.8011702983556026e-06,
+      "loss": 0.4971,
+      "step": 6877
+    },
+    {
+      "epoch": 2.7007162262286983,
+      "grad_norm": 0.46826293402204716,
+      "learning_rate": 4.801109063541554e-06,
+      "loss": 0.4992,
+      "step": 6878
+    },
+    {
+      "epoch": 2.7011113855272906,
+      "grad_norm": 0.5113971704976565,
+      "learning_rate": 4.801047819690146e-06,
+      "loss": 0.5298,
+      "step": 6879
+    },
+    {
+      "epoch": 2.701506544825883,
+      "grad_norm": 0.4645313397142929,
+      "learning_rate": 4.80098656680162e-06,
+      "loss": 0.5112,
+      "step": 6880
+    },
+    {
+      "epoch": 2.701901704124475,
+      "grad_norm": 0.47144303739327625,
+      "learning_rate": 4.800925304876215e-06,
+      "loss": 0.5002,
+      "step": 6881
+    },
+    {
+      "epoch": 2.7022968634230673,
+      "grad_norm": 0.4780153361617559,
+      "learning_rate": 4.800864033914173e-06,
+      "loss": 0.5317,
+      "step": 6882
+    },
+    {
+      "epoch": 2.7026920227216595,
+      "grad_norm": 0.46311336498971456,
+      "learning_rate": 4.800802753915735e-06,
+      "loss": 0.5098,
+      "step": 6883
+    },
+    {
+      "epoch": 2.7030871820202518,
+      "grad_norm": 0.4704994698152214,
+      "learning_rate": 4.8007414648811405e-06,
+      "loss": 0.4842,
+      "step": 6884
+    },
+    {
+      "epoch": 2.703482341318844,
+      "grad_norm": 0.44615978605663914,
+      "learning_rate": 4.80068016681063e-06,
+      "loss": 0.5057,
+      "step": 6885
+    },
+    {
+      "epoch": 2.7038775006174363,
+      "grad_norm": 0.4582730966673267,
+      "learning_rate": 4.800618859704445e-06,
+      "loss": 0.4998,
+      "step": 6886
+    },
+    {
+      "epoch": 2.7042726599160285,
+      "grad_norm": 0.47695840090832414,
+      "learning_rate": 4.800557543562827e-06,
+      "loss": 0.5027,
+      "step": 6887
+    },
+    {
+      "epoch": 2.7046678192146207,
+      "grad_norm": 0.464279484547299,
+      "learning_rate": 4.800496218386015e-06,
+      "loss": 0.4821,
+      "step": 6888
+    },
+    {
+      "epoch": 2.705062978513213,
+      "grad_norm": 0.4627589954096053,
+      "learning_rate": 4.800434884174251e-06,
+      "loss": 0.5229,
+      "step": 6889
+    },
+    {
+      "epoch": 2.705458137811805,
+      "grad_norm": 0.46481102969321464,
+      "learning_rate": 4.800373540927776e-06,
+      "loss": 0.4879,
+      "step": 6890
+    },
+    {
+      "epoch": 2.705853297110398,
+      "grad_norm": 0.46484595991261723,
+      "learning_rate": 4.800312188646831e-06,
+      "loss": 0.5025,
+      "step": 6891
+    },
+    {
+      "epoch": 2.70624845640899,
+      "grad_norm": 0.4679261982088802,
+      "learning_rate": 4.800250827331656e-06,
+      "loss": 0.5031,
+      "step": 6892
+    },
+    {
+      "epoch": 2.7066436157075824,
+      "grad_norm": 0.45277952364743757,
+      "learning_rate": 4.800189456982492e-06,
+      "loss": 0.5067,
+      "step": 6893
+    },
+    {
+      "epoch": 2.7070387750061746,
+      "grad_norm": 0.4581955934170635,
+      "learning_rate": 4.800128077599581e-06,
+      "loss": 0.498,
+      "step": 6894
+    },
+    {
+      "epoch": 2.707433934304767,
+      "grad_norm": 0.4728192631389551,
+      "learning_rate": 4.800066689183164e-06,
+      "loss": 0.5116,
+      "step": 6895
+    },
+    {
+      "epoch": 2.707829093603359,
+      "grad_norm": 0.4742126911889233,
+      "learning_rate": 4.800005291733482e-06,
+      "loss": 0.5127,
+      "step": 6896
+    },
+    {
+      "epoch": 2.7082242529019513,
+      "grad_norm": 0.4515853200904919,
+      "learning_rate": 4.7999438852507745e-06,
+      "loss": 0.4993,
+      "step": 6897
+    },
+    {
+      "epoch": 2.7086194122005436,
+      "grad_norm": 0.4613616972466868,
+      "learning_rate": 4.799882469735285e-06,
+      "loss": 0.5032,
+      "step": 6898
+    },
+    {
+      "epoch": 2.709014571499136,
+      "grad_norm": 0.45816883607410985,
+      "learning_rate": 4.799821045187254e-06,
+      "loss": 0.5137,
+      "step": 6899
+    },
+    {
+      "epoch": 2.709409730797728,
+      "grad_norm": 0.44940851329013853,
+      "learning_rate": 4.7997596116069215e-06,
+      "loss": 0.5187,
+      "step": 6900
+    },
+    {
+      "epoch": 2.7098048900963203,
+      "grad_norm": 0.4502469570380973,
+      "learning_rate": 4.79969816899453e-06,
+      "loss": 0.5025,
+      "step": 6901
+    },
+    {
+      "epoch": 2.7102000493949125,
+      "grad_norm": 0.4481223650440485,
+      "learning_rate": 4.799636717350321e-06,
+      "loss": 0.4923,
+      "step": 6902
+    },
+    {
+      "epoch": 2.710595208693505,
+      "grad_norm": 0.45792623854887804,
+      "learning_rate": 4.7995752566745345e-06,
+      "loss": 0.5043,
+      "step": 6903
+    },
+    {
+      "epoch": 2.710990367992097,
+      "grad_norm": 0.4601271483289185,
+      "learning_rate": 4.7995137869674135e-06,
+      "loss": 0.509,
+      "step": 6904
+    },
+    {
+      "epoch": 2.7113855272906893,
+      "grad_norm": 0.4476972091826272,
+      "learning_rate": 4.799452308229199e-06,
+      "loss": 0.5014,
+      "step": 6905
+    },
+    {
+      "epoch": 2.7117806865892815,
+      "grad_norm": 0.4600909924968701,
+      "learning_rate": 4.7993908204601315e-06,
+      "loss": 0.5073,
+      "step": 6906
+    },
+    {
+      "epoch": 2.7121758458878737,
+      "grad_norm": 0.45934839488330736,
+      "learning_rate": 4.799329323660453e-06,
+      "loss": 0.5173,
+      "step": 6907
+    },
+    {
+      "epoch": 2.712571005186466,
+      "grad_norm": 0.45337135497372977,
+      "learning_rate": 4.799267817830406e-06,
+      "loss": 0.501,
+      "step": 6908
+    },
+    {
+      "epoch": 2.7129661644850582,
+      "grad_norm": 0.4500761726544255,
+      "learning_rate": 4.7992063029702304e-06,
+      "loss": 0.5018,
+      "step": 6909
+    },
+    {
+      "epoch": 2.7133613237836505,
+      "grad_norm": 0.45950279024379764,
+      "learning_rate": 4.799144779080169e-06,
+      "loss": 0.517,
+      "step": 6910
+    },
+    {
+      "epoch": 2.7137564830822427,
+      "grad_norm": 0.4730730331136291,
+      "learning_rate": 4.799083246160463e-06,
+      "loss": 0.4923,
+      "step": 6911
+    },
+    {
+      "epoch": 2.714151642380835,
+      "grad_norm": 0.43830507511052397,
+      "learning_rate": 4.799021704211354e-06,
+      "loss": 0.4934,
+      "step": 6912
+    },
+    {
+      "epoch": 2.714546801679427,
+      "grad_norm": 0.48563994353988965,
+      "learning_rate": 4.798960153233084e-06,
+      "loss": 0.5061,
+      "step": 6913
+    },
+    {
+      "epoch": 2.7149419609780194,
+      "grad_norm": 0.4792693935069234,
+      "learning_rate": 4.798898593225894e-06,
+      "loss": 0.5025,
+      "step": 6914
+    },
+    {
+      "epoch": 2.7153371202766117,
+      "grad_norm": 0.4535898076683153,
+      "learning_rate": 4.798837024190027e-06,
+      "loss": 0.5132,
+      "step": 6915
+    },
+    {
+      "epoch": 2.715732279575204,
+      "grad_norm": 0.47116866983420425,
+      "learning_rate": 4.798775446125723e-06,
+      "loss": 0.4975,
+      "step": 6916
+    },
+    {
+      "epoch": 2.716127438873796,
+      "grad_norm": 0.4460352697228603,
+      "learning_rate": 4.7987138590332264e-06,
+      "loss": 0.5062,
+      "step": 6917
+    },
+    {
+      "epoch": 2.7165225981723884,
+      "grad_norm": 0.4520031645447902,
+      "learning_rate": 4.798652262912776e-06,
+      "loss": 0.512,
+      "step": 6918
+    },
+    {
+      "epoch": 2.7169177574709806,
+      "grad_norm": 0.44599164428046767,
+      "learning_rate": 4.798590657764617e-06,
+      "loss": 0.4947,
+      "step": 6919
+    },
+    {
+      "epoch": 2.717312916769573,
+      "grad_norm": 0.4462324282376401,
+      "learning_rate": 4.798529043588989e-06,
+      "loss": 0.4997,
+      "step": 6920
+    },
+    {
+      "epoch": 2.717708076068165,
+      "grad_norm": 0.4517482214846663,
+      "learning_rate": 4.798467420386133e-06,
+      "loss": 0.5082,
+      "step": 6921
+    },
+    {
+      "epoch": 2.7181032353667574,
+      "grad_norm": 0.45386053523383296,
+      "learning_rate": 4.798405788156295e-06,
+      "loss": 0.4976,
+      "step": 6922
+    },
+    {
+      "epoch": 2.7184983946653496,
+      "grad_norm": 0.46008987922377587,
+      "learning_rate": 4.7983441468997134e-06,
+      "loss": 0.493,
+      "step": 6923
+    },
+    {
+      "epoch": 2.718893553963942,
+      "grad_norm": 0.4481062465022577,
+      "learning_rate": 4.798282496616633e-06,
+      "loss": 0.5143,
+      "step": 6924
+    },
+    {
+      "epoch": 2.719288713262534,
+      "grad_norm": 0.4505251384368751,
+      "learning_rate": 4.7982208373072936e-06,
+      "loss": 0.5175,
+      "step": 6925
+    },
+    {
+      "epoch": 2.7196838725611263,
+      "grad_norm": 0.47657228107417726,
+      "learning_rate": 4.798159168971938e-06,
+      "loss": 0.5243,
+      "step": 6926
+    },
+    {
+      "epoch": 2.7200790318597186,
+      "grad_norm": 0.46381523563325006,
+      "learning_rate": 4.798097491610809e-06,
+      "loss": 0.4926,
+      "step": 6927
+    },
+    {
+      "epoch": 2.720474191158311,
+      "grad_norm": 0.4925744506297022,
+      "learning_rate": 4.798035805224149e-06,
+      "loss": 0.4848,
+      "step": 6928
+    },
+    {
+      "epoch": 2.720869350456903,
+      "grad_norm": 0.44864029022686136,
+      "learning_rate": 4.797974109812199e-06,
+      "loss": 0.4967,
+      "step": 6929
+    },
+    {
+      "epoch": 2.7212645097554953,
+      "grad_norm": 0.49217622158783636,
+      "learning_rate": 4.797912405375203e-06,
+      "loss": 0.5014,
+      "step": 6930
+    },
+    {
+      "epoch": 2.7216596690540875,
+      "grad_norm": 0.4551306861648084,
+      "learning_rate": 4.797850691913402e-06,
+      "loss": 0.5291,
+      "step": 6931
+    },
+    {
+      "epoch": 2.7220548283526798,
+      "grad_norm": 0.4637807915278924,
+      "learning_rate": 4.797788969427039e-06,
+      "loss": 0.5012,
+      "step": 6932
+    },
+    {
+      "epoch": 2.722449987651272,
+      "grad_norm": 0.4512850558545194,
+      "learning_rate": 4.797727237916355e-06,
+      "loss": 0.5125,
+      "step": 6933
+    },
+    {
+      "epoch": 2.7228451469498642,
+      "grad_norm": 0.4594283870187807,
+      "learning_rate": 4.7976654973815955e-06,
+      "loss": 0.5108,
+      "step": 6934
+    },
+    {
+      "epoch": 2.7232403062484565,
+      "grad_norm": 0.5129146142831824,
+      "learning_rate": 4.797603747823e-06,
+      "loss": 0.5151,
+      "step": 6935
+    },
+    {
+      "epoch": 2.7236354655470487,
+      "grad_norm": 0.4624544070980088,
+      "learning_rate": 4.797541989240812e-06,
+      "loss": 0.5022,
+      "step": 6936
+    },
+    {
+      "epoch": 2.724030624845641,
+      "grad_norm": 0.44163885935392466,
+      "learning_rate": 4.797480221635276e-06,
+      "loss": 0.4944,
+      "step": 6937
+    },
+    {
+      "epoch": 2.724425784144233,
+      "grad_norm": 0.4428403204450448,
+      "learning_rate": 4.7974184450066305e-06,
+      "loss": 0.5108,
+      "step": 6938
+    },
+    {
+      "epoch": 2.7248209434428254,
+      "grad_norm": 0.46568478000032276,
+      "learning_rate": 4.7973566593551216e-06,
+      "loss": 0.4976,
+      "step": 6939
+    },
+    {
+      "epoch": 2.7252161027414177,
+      "grad_norm": 0.45155171800762267,
+      "learning_rate": 4.7972948646809906e-06,
+      "loss": 0.5149,
+      "step": 6940
+    },
+    {
+      "epoch": 2.72561126204001,
+      "grad_norm": 0.46921689781602866,
+      "learning_rate": 4.797233060984481e-06,
+      "loss": 0.5185,
+      "step": 6941
+    },
+    {
+      "epoch": 2.726006421338602,
+      "grad_norm": 0.4584753926896495,
+      "learning_rate": 4.797171248265833e-06,
+      "loss": 0.5043,
+      "step": 6942
+    },
+    {
+      "epoch": 2.7264015806371944,
+      "grad_norm": 0.4457236150051623,
+      "learning_rate": 4.797109426525293e-06,
+      "loss": 0.5003,
+      "step": 6943
+    },
+    {
+      "epoch": 2.7267967399357866,
+      "grad_norm": 0.4577197175651194,
+      "learning_rate": 4.797047595763101e-06,
+      "loss": 0.5268,
+      "step": 6944
+    },
+    {
+      "epoch": 2.727191899234379,
+      "grad_norm": 0.4574229252485062,
+      "learning_rate": 4.796985755979502e-06,
+      "loss": 0.498,
+      "step": 6945
+    },
+    {
+      "epoch": 2.727587058532971,
+      "grad_norm": 0.4596664767781495,
+      "learning_rate": 4.796923907174737e-06,
+      "loss": 0.5038,
+      "step": 6946
+    },
+    {
+      "epoch": 2.7279822178315634,
+      "grad_norm": 0.4458255538392768,
+      "learning_rate": 4.79686204934905e-06,
+      "loss": 0.5137,
+      "step": 6947
+    },
+    {
+      "epoch": 2.7283773771301556,
+      "grad_norm": 0.45280689074026337,
+      "learning_rate": 4.796800182502683e-06,
+      "loss": 0.5208,
+      "step": 6948
+    },
+    {
+      "epoch": 2.728772536428748,
+      "grad_norm": 0.47554043866753837,
+      "learning_rate": 4.7967383066358795e-06,
+      "loss": 0.5044,
+      "step": 6949
+    },
+    {
+      "epoch": 2.72916769572734,
+      "grad_norm": 0.4607340942101068,
+      "learning_rate": 4.796676421748884e-06,
+      "loss": 0.5169,
+      "step": 6950
+    },
+    {
+      "epoch": 2.7295628550259323,
+      "grad_norm": 0.44291773525236144,
+      "learning_rate": 4.796614527841937e-06,
+      "loss": 0.4949,
+      "step": 6951
+    },
+    {
+      "epoch": 2.7299580143245246,
+      "grad_norm": 0.46432616537422905,
+      "learning_rate": 4.796552624915283e-06,
+      "loss": 0.5058,
+      "step": 6952
+    },
+    {
+      "epoch": 2.730353173623117,
+      "grad_norm": 0.4633027740708962,
+      "learning_rate": 4.796490712969165e-06,
+      "loss": 0.5167,
+      "step": 6953
+    },
+    {
+      "epoch": 2.730748332921709,
+      "grad_norm": 0.5118956983442591,
+      "learning_rate": 4.796428792003826e-06,
+      "loss": 0.5229,
+      "step": 6954
+    },
+    {
+      "epoch": 2.7311434922203013,
+      "grad_norm": 0.4543120210417002,
+      "learning_rate": 4.796366862019508e-06,
+      "loss": 0.4928,
+      "step": 6955
+    },
+    {
+      "epoch": 2.7315386515188935,
+      "grad_norm": 0.47789407600911105,
+      "learning_rate": 4.796304923016456e-06,
+      "loss": 0.5188,
+      "step": 6956
+    },
+    {
+      "epoch": 2.7319338108174858,
+      "grad_norm": 0.46242168805945055,
+      "learning_rate": 4.796242974994913e-06,
+      "loss": 0.4952,
+      "step": 6957
+    },
+    {
+      "epoch": 2.732328970116078,
+      "grad_norm": 0.4566066993777903,
+      "learning_rate": 4.796181017955122e-06,
+      "loss": 0.5047,
+      "step": 6958
+    },
+    {
+      "epoch": 2.7327241294146702,
+      "grad_norm": 0.47657476604138915,
+      "learning_rate": 4.796119051897327e-06,
+      "loss": 0.5012,
+      "step": 6959
+    },
+    {
+      "epoch": 2.7331192887132625,
+      "grad_norm": 0.46254465859951066,
+      "learning_rate": 4.79605707682177e-06,
+      "loss": 0.4958,
+      "step": 6960
+    },
+    {
+      "epoch": 2.7335144480118547,
+      "grad_norm": 0.45509248055911644,
+      "learning_rate": 4.795995092728694e-06,
+      "loss": 0.5063,
+      "step": 6961
+    },
+    {
+      "epoch": 2.733909607310447,
+      "grad_norm": 0.4460935553959142,
+      "learning_rate": 4.795933099618344e-06,
+      "loss": 0.5105,
+      "step": 6962
+    },
+    {
+      "epoch": 2.734304766609039,
+      "grad_norm": 0.4477458914794227,
+      "learning_rate": 4.795871097490964e-06,
+      "loss": 0.5147,
+      "step": 6963
+    },
+    {
+      "epoch": 2.7346999259076314,
+      "grad_norm": 0.4423880972213776,
+      "learning_rate": 4.795809086346796e-06,
+      "loss": 0.5089,
+      "step": 6964
+    },
+    {
+      "epoch": 2.7350950852062237,
+      "grad_norm": 0.4391301789018041,
+      "learning_rate": 4.795747066186083e-06,
+      "loss": 0.5023,
+      "step": 6965
+    },
+    {
+      "epoch": 2.735490244504816,
+      "grad_norm": 0.45270875027903384,
+      "learning_rate": 4.79568503700907e-06,
+      "loss": 0.5027,
+      "step": 6966
+    },
+    {
+      "epoch": 2.735885403803408,
+      "grad_norm": 0.4548394809042717,
+      "learning_rate": 4.795622998816001e-06,
+      "loss": 0.4998,
+      "step": 6967
+    },
+    {
+      "epoch": 2.7362805631020004,
+      "grad_norm": 0.4500140263131528,
+      "learning_rate": 4.795560951607118e-06,
+      "loss": 0.4956,
+      "step": 6968
+    },
+    {
+      "epoch": 2.7366757224005926,
+      "grad_norm": 0.45444549472429036,
+      "learning_rate": 4.795498895382667e-06,
+      "loss": 0.4999,
+      "step": 6969
+    },
+    {
+      "epoch": 2.737070881699185,
+      "grad_norm": 0.45943067355398565,
+      "learning_rate": 4.795436830142888e-06,
+      "loss": 0.5079,
+      "step": 6970
+    },
+    {
+      "epoch": 2.737466040997777,
+      "grad_norm": 0.4583842097104527,
+      "learning_rate": 4.795374755888028e-06,
+      "loss": 0.5035,
+      "step": 6971
+    },
+    {
+      "epoch": 2.7378612002963694,
+      "grad_norm": 0.44524024874479756,
+      "learning_rate": 4.7953126726183305e-06,
+      "loss": 0.4885,
+      "step": 6972
+    },
+    {
+      "epoch": 2.7382563595949616,
+      "grad_norm": 0.5259627464540994,
+      "learning_rate": 4.795250580334038e-06,
+      "loss": 0.5078,
+      "step": 6973
+    },
+    {
+      "epoch": 2.738651518893554,
+      "grad_norm": 0.44629367850165574,
+      "learning_rate": 4.795188479035395e-06,
+      "loss": 0.5132,
+      "step": 6974
+    },
+    {
+      "epoch": 2.739046678192146,
+      "grad_norm": 0.4738145104107334,
+      "learning_rate": 4.7951263687226444e-06,
+      "loss": 0.5144,
+      "step": 6975
+    },
+    {
+      "epoch": 2.7394418374907383,
+      "grad_norm": 0.4536390577628937,
+      "learning_rate": 4.795064249396032e-06,
+      "loss": 0.4981,
+      "step": 6976
+    },
+    {
+      "epoch": 2.7398369967893306,
+      "grad_norm": 0.48887491920473725,
+      "learning_rate": 4.795002121055802e-06,
+      "loss": 0.529,
+      "step": 6977
+    },
+    {
+      "epoch": 2.740232156087923,
+      "grad_norm": 0.45703040398721106,
+      "learning_rate": 4.794939983702196e-06,
+      "loss": 0.4995,
+      "step": 6978
+    },
+    {
+      "epoch": 2.740627315386515,
+      "grad_norm": 0.4688231922181613,
+      "learning_rate": 4.7948778373354585e-06,
+      "loss": 0.5007,
+      "step": 6979
+    },
+    {
+      "epoch": 2.7410224746851073,
+      "grad_norm": 0.4466851138075401,
+      "learning_rate": 4.794815681955836e-06,
+      "loss": 0.496,
+      "step": 6980
+    },
+    {
+      "epoch": 2.7414176339836995,
+      "grad_norm": 0.4893848636599617,
+      "learning_rate": 4.79475351756357e-06,
+      "loss": 0.5012,
+      "step": 6981
+    },
+    {
+      "epoch": 2.7418127932822918,
+      "grad_norm": 0.4883005520320106,
+      "learning_rate": 4.794691344158906e-06,
+      "loss": 0.5085,
+      "step": 6982
+    },
+    {
+      "epoch": 2.742207952580884,
+      "grad_norm": 0.45293984493757544,
+      "learning_rate": 4.794629161742088e-06,
+      "loss": 0.5008,
+      "step": 6983
+    },
+    {
+      "epoch": 2.7426031118794763,
+      "grad_norm": 0.5501230497941952,
+      "learning_rate": 4.79456697031336e-06,
+      "loss": 0.5128,
+      "step": 6984
+    },
+    {
+      "epoch": 2.7429982711780685,
+      "grad_norm": 0.4639739628186388,
+      "learning_rate": 4.794504769872966e-06,
+      "loss": 0.5169,
+      "step": 6985
+    },
+    {
+      "epoch": 2.7433934304766607,
+      "grad_norm": 0.4576900039164546,
+      "learning_rate": 4.794442560421151e-06,
+      "loss": 0.4819,
+      "step": 6986
+    },
+    {
+      "epoch": 2.743788589775253,
+      "grad_norm": 0.4954297664525487,
+      "learning_rate": 4.794380341958158e-06,
+      "loss": 0.5166,
+      "step": 6987
+    },
+    {
+      "epoch": 2.744183749073845,
+      "grad_norm": 0.46126145279470626,
+      "learning_rate": 4.794318114484233e-06,
+      "loss": 0.5118,
+      "step": 6988
+    },
+    {
+      "epoch": 2.7445789083724375,
+      "grad_norm": 0.4584209830526114,
+      "learning_rate": 4.79425587799962e-06,
+      "loss": 0.5086,
+      "step": 6989
+    },
+    {
+      "epoch": 2.7449740676710297,
+      "grad_norm": 0.4831191153527967,
+      "learning_rate": 4.794193632504561e-06,
+      "loss": 0.5441,
+      "step": 6990
+    },
+    {
+      "epoch": 2.745369226969622,
+      "grad_norm": 0.4797130021650316,
+      "learning_rate": 4.794131377999305e-06,
+      "loss": 0.5279,
+      "step": 6991
+    },
+    {
+      "epoch": 2.745764386268214,
+      "grad_norm": 0.4699057790970996,
+      "learning_rate": 4.794069114484092e-06,
+      "loss": 0.5147,
+      "step": 6992
+    },
+    {
+      "epoch": 2.7461595455668064,
+      "grad_norm": 0.46046778025908625,
+      "learning_rate": 4.79400684195917e-06,
+      "loss": 0.5086,
+      "step": 6993
+    },
+    {
+      "epoch": 2.7465547048653987,
+      "grad_norm": 0.4383439181821961,
+      "learning_rate": 4.793944560424782e-06,
+      "loss": 0.4921,
+      "step": 6994
+    },
+    {
+      "epoch": 2.746949864163991,
+      "grad_norm": 0.4643419519593291,
+      "learning_rate": 4.7938822698811725e-06,
+      "loss": 0.5024,
+      "step": 6995
+    },
+    {
+      "epoch": 2.747345023462583,
+      "grad_norm": 0.45885771648686097,
+      "learning_rate": 4.793819970328586e-06,
+      "loss": 0.4996,
+      "step": 6996
+    },
+    {
+      "epoch": 2.7477401827611754,
+      "grad_norm": 0.47670938755540426,
+      "learning_rate": 4.793757661767268e-06,
+      "loss": 0.5024,
+      "step": 6997
+    },
+    {
+      "epoch": 2.7481353420597676,
+      "grad_norm": 0.45854522700324046,
+      "learning_rate": 4.7936953441974624e-06,
+      "loss": 0.4988,
+      "step": 6998
+    },
+    {
+      "epoch": 2.74853050135836,
+      "grad_norm": 0.4518882922405723,
+      "learning_rate": 4.793633017619415e-06,
+      "loss": 0.5026,
+      "step": 6999
+    },
+    {
+      "epoch": 2.748925660656952,
+      "grad_norm": 0.4721585044201508,
+      "learning_rate": 4.793570682033368e-06,
+      "loss": 0.5066,
+      "step": 7000
+    },
+    {
+      "epoch": 2.7493208199555443,
+      "grad_norm": 0.45566875735908363,
+      "learning_rate": 4.7935083374395694e-06,
+      "loss": 0.4957,
+      "step": 7001
+    },
+    {
+      "epoch": 2.7497159792541366,
+      "grad_norm": 0.46827293594382546,
+      "learning_rate": 4.793445983838263e-06,
+      "loss": 0.5161,
+      "step": 7002
+    },
+    {
+      "epoch": 2.750111138552729,
+      "grad_norm": 0.45535123480528633,
+      "learning_rate": 4.793383621229694e-06,
+      "loss": 0.4846,
+      "step": 7003
+    },
+    {
+      "epoch": 2.750506297851321,
+      "grad_norm": 0.4640975956499678,
+      "learning_rate": 4.7933212496141055e-06,
+      "loss": 0.5196,
+      "step": 7004
+    },
+    {
+      "epoch": 2.7509014571499133,
+      "grad_norm": 0.45391152016709185,
+      "learning_rate": 4.793258868991743e-06,
+      "loss": 0.5044,
+      "step": 7005
+    },
+    {
+      "epoch": 2.7512966164485055,
+      "grad_norm": 0.456017711396762,
+      "learning_rate": 4.793196479362854e-06,
+      "loss": 0.5073,
+      "step": 7006
+    },
+    {
+      "epoch": 2.751691775747098,
+      "grad_norm": 0.457250650843779,
+      "learning_rate": 4.793134080727682e-06,
+      "loss": 0.5072,
+      "step": 7007
+    },
+    {
+      "epoch": 2.75208693504569,
+      "grad_norm": 0.5241068945571156,
+      "learning_rate": 4.79307167308647e-06,
+      "loss": 0.5143,
+      "step": 7008
+    },
+    {
+      "epoch": 2.7524820943442827,
+      "grad_norm": 0.44978349825521247,
+      "learning_rate": 4.793009256439466e-06,
+      "loss": 0.491,
+      "step": 7009
+    },
+    {
+      "epoch": 2.752877253642875,
+      "grad_norm": 0.445219206652602,
+      "learning_rate": 4.792946830786914e-06,
+      "loss": 0.4946,
+      "step": 7010
+    },
+    {
+      "epoch": 2.753272412941467,
+      "grad_norm": 0.4728591513095926,
+      "learning_rate": 4.792884396129059e-06,
+      "loss": 0.5342,
+      "step": 7011
+    },
+    {
+      "epoch": 2.7536675722400594,
+      "grad_norm": 0.4494210336181794,
+      "learning_rate": 4.792821952466146e-06,
+      "loss": 0.4868,
+      "step": 7012
+    },
+    {
+      "epoch": 2.7540627315386517,
+      "grad_norm": 0.46547211537802097,
+      "learning_rate": 4.7927594997984215e-06,
+      "loss": 0.491,
+      "step": 7013
+    },
+    {
+      "epoch": 2.754457890837244,
+      "grad_norm": 0.47364471713178513,
+      "learning_rate": 4.7926970381261295e-06,
+      "loss": 0.506,
+      "step": 7014
+    },
+    {
+      "epoch": 2.754853050135836,
+      "grad_norm": 0.4508628675635779,
+      "learning_rate": 4.7926345674495155e-06,
+      "loss": 0.5066,
+      "step": 7015
+    },
+    {
+      "epoch": 2.7552482094344284,
+      "grad_norm": 0.47245052795730946,
+      "learning_rate": 4.792572087768825e-06,
+      "loss": 0.509,
+      "step": 7016
+    },
+    {
+      "epoch": 2.7556433687330206,
+      "grad_norm": 0.49318688651877063,
+      "learning_rate": 4.792509599084304e-06,
+      "loss": 0.4953,
+      "step": 7017
+    },
+    {
+      "epoch": 2.756038528031613,
+      "grad_norm": 0.45637468572032197,
+      "learning_rate": 4.792447101396197e-06,
+      "loss": 0.5149,
+      "step": 7018
+    },
+    {
+      "epoch": 2.756433687330205,
+      "grad_norm": 0.4682354739864325,
+      "learning_rate": 4.79238459470475e-06,
+      "loss": 0.5186,
+      "step": 7019
+    },
+    {
+      "epoch": 2.7568288466287973,
+      "grad_norm": 0.4568829125381733,
+      "learning_rate": 4.7923220790102084e-06,
+      "loss": 0.5055,
+      "step": 7020
+    },
+    {
+      "epoch": 2.7572240059273896,
+      "grad_norm": 0.45180194260707746,
+      "learning_rate": 4.792259554312817e-06,
+      "loss": 0.4921,
+      "step": 7021
+    },
+    {
+      "epoch": 2.757619165225982,
+      "grad_norm": 0.4668813422184499,
+      "learning_rate": 4.7921970206128235e-06,
+      "loss": 0.5301,
+      "step": 7022
+    },
+    {
+      "epoch": 2.758014324524574,
+      "grad_norm": 0.4487332150132883,
+      "learning_rate": 4.7921344779104705e-06,
+      "loss": 0.5031,
+      "step": 7023
+    },
+    {
+      "epoch": 2.7584094838231663,
+      "grad_norm": 0.45510934399651354,
+      "learning_rate": 4.7920719262060055e-06,
+      "loss": 0.5386,
+      "step": 7024
+    },
+    {
+      "epoch": 2.7588046431217585,
+      "grad_norm": 0.44723741022980124,
+      "learning_rate": 4.792009365499674e-06,
+      "loss": 0.5155,
+      "step": 7025
+    },
+    {
+      "epoch": 2.759199802420351,
+      "grad_norm": 0.4504980939404593,
+      "learning_rate": 4.791946795791721e-06,
+      "loss": 0.5161,
+      "step": 7026
+    },
+    {
+      "epoch": 2.759594961718943,
+      "grad_norm": 0.44795581514294985,
+      "learning_rate": 4.791884217082394e-06,
+      "loss": 0.5015,
+      "step": 7027
+    },
+    {
+      "epoch": 2.7599901210175353,
+      "grad_norm": 0.45748558259217437,
+      "learning_rate": 4.791821629371936e-06,
+      "loss": 0.5038,
+      "step": 7028
+    },
+    {
+      "epoch": 2.7603852803161275,
+      "grad_norm": 0.47531123398447556,
+      "learning_rate": 4.791759032660596e-06,
+      "loss": 0.4986,
+      "step": 7029
+    },
+    {
+      "epoch": 2.7607804396147198,
+      "grad_norm": 0.46673985312016447,
+      "learning_rate": 4.7916964269486165e-06,
+      "loss": 0.4956,
+      "step": 7030
+    },
+    {
+      "epoch": 2.761175598913312,
+      "grad_norm": 0.4629491572931826,
+      "learning_rate": 4.791633812236245e-06,
+      "loss": 0.4946,
+      "step": 7031
+    },
+    {
+      "epoch": 2.7615707582119042,
+      "grad_norm": 0.4525334948265311,
+      "learning_rate": 4.791571188523729e-06,
+      "loss": 0.5164,
+      "step": 7032
+    },
+    {
+      "epoch": 2.7619659175104965,
+      "grad_norm": 0.45400679108436304,
+      "learning_rate": 4.7915085558113115e-06,
+      "loss": 0.4999,
+      "step": 7033
+    },
+    {
+      "epoch": 2.7623610768090887,
+      "grad_norm": 0.47943008845611956,
+      "learning_rate": 4.791445914099241e-06,
+      "loss": 0.5032,
+      "step": 7034
+    },
+    {
+      "epoch": 2.762756236107681,
+      "grad_norm": 0.4452226257043686,
+      "learning_rate": 4.791383263387761e-06,
+      "loss": 0.501,
+      "step": 7035
+    },
+    {
+      "epoch": 2.763151395406273,
+      "grad_norm": 0.48784364857172546,
+      "learning_rate": 4.7913206036771195e-06,
+      "loss": 0.5172,
+      "step": 7036
+    },
+    {
+      "epoch": 2.7635465547048654,
+      "grad_norm": 0.47792955328117803,
+      "learning_rate": 4.791257934967563e-06,
+      "loss": 0.5341,
+      "step": 7037
+    },
+    {
+      "epoch": 2.7639417140034577,
+      "grad_norm": 0.4432418586956085,
+      "learning_rate": 4.791195257259335e-06,
+      "loss": 0.4891,
+      "step": 7038
+    },
+    {
+      "epoch": 2.76433687330205,
+      "grad_norm": 0.4712175445574134,
+      "learning_rate": 4.791132570552685e-06,
+      "loss": 0.4999,
+      "step": 7039
+    },
+    {
+      "epoch": 2.764732032600642,
+      "grad_norm": 0.4439252343190045,
+      "learning_rate": 4.791069874847857e-06,
+      "loss": 0.5085,
+      "step": 7040
+    },
+    {
+      "epoch": 2.7651271918992344,
+      "grad_norm": 0.5317426037392942,
+      "learning_rate": 4.791007170145097e-06,
+      "loss": 0.5129,
+      "step": 7041
+    },
+    {
+      "epoch": 2.7655223511978266,
+      "grad_norm": 0.4587796397905756,
+      "learning_rate": 4.790944456444653e-06,
+      "loss": 0.5116,
+      "step": 7042
+    },
+    {
+      "epoch": 2.765917510496419,
+      "grad_norm": 0.4797852387845019,
+      "learning_rate": 4.7908817337467695e-06,
+      "loss": 0.5254,
+      "step": 7043
+    },
+    {
+      "epoch": 2.766312669795011,
+      "grad_norm": 0.4558663094889101,
+      "learning_rate": 4.790819002051694e-06,
+      "loss": 0.5199,
+      "step": 7044
+    },
+    {
+      "epoch": 2.7667078290936034,
+      "grad_norm": 0.44832319567068757,
+      "learning_rate": 4.790756261359673e-06,
+      "loss": 0.515,
+      "step": 7045
+    },
+    {
+      "epoch": 2.7671029883921956,
+      "grad_norm": 0.4482969941450535,
+      "learning_rate": 4.7906935116709505e-06,
+      "loss": 0.5046,
+      "step": 7046
+    },
+    {
+      "epoch": 2.767498147690788,
+      "grad_norm": 0.4559545857856152,
+      "learning_rate": 4.790630752985776e-06,
+      "loss": 0.4937,
+      "step": 7047
+    },
+    {
+      "epoch": 2.76789330698938,
+      "grad_norm": 0.44267716011895897,
+      "learning_rate": 4.790567985304396e-06,
+      "loss": 0.508,
+      "step": 7048
+    },
+    {
+      "epoch": 2.7682884662879723,
+      "grad_norm": 0.464044432217731,
+      "learning_rate": 4.790505208627055e-06,
+      "loss": 0.5148,
+      "step": 7049
+    },
+    {
+      "epoch": 2.7686836255865646,
+      "grad_norm": 0.4467116038672829,
+      "learning_rate": 4.790442422954e-06,
+      "loss": 0.498,
+      "step": 7050
+    },
+    {
+      "epoch": 2.769078784885157,
+      "grad_norm": 0.4798202007038068,
+      "learning_rate": 4.790379628285479e-06,
+      "loss": 0.5087,
+      "step": 7051
+    },
+    {
+      "epoch": 2.769473944183749,
+      "grad_norm": 0.44503110439284255,
+      "learning_rate": 4.790316824621736e-06,
+      "loss": 0.5137,
+      "step": 7052
+    },
+    {
+      "epoch": 2.7698691034823413,
+      "grad_norm": 0.46457729311037305,
+      "learning_rate": 4.79025401196302e-06,
+      "loss": 0.5195,
+      "step": 7053
+    },
+    {
+      "epoch": 2.7702642627809335,
+      "grad_norm": 0.44066547002420625,
+      "learning_rate": 4.790191190309578e-06,
+      "loss": 0.4925,
+      "step": 7054
+    },
+    {
+      "epoch": 2.7706594220795258,
+      "grad_norm": 0.45430470319634064,
+      "learning_rate": 4.790128359661654e-06,
+      "loss": 0.495,
+      "step": 7055
+    },
+    {
+      "epoch": 2.771054581378118,
+      "grad_norm": 0.4414603456774517,
+      "learning_rate": 4.790065520019498e-06,
+      "loss": 0.5038,
+      "step": 7056
+    },
+    {
+      "epoch": 2.7714497406767102,
+      "grad_norm": 0.4922018456920078,
+      "learning_rate": 4.790002671383354e-06,
+      "loss": 0.5155,
+      "step": 7057
+    },
+    {
+      "epoch": 2.7718448999753025,
+      "grad_norm": 0.453999025325055,
+      "learning_rate": 4.789939813753471e-06,
+      "loss": 0.5008,
+      "step": 7058
+    },
+    {
+      "epoch": 2.7722400592738947,
+      "grad_norm": 0.4617428583463696,
+      "learning_rate": 4.789876947130095e-06,
+      "loss": 0.4974,
+      "step": 7059
+    },
+    {
+      "epoch": 2.772635218572487,
+      "grad_norm": 0.464515686848104,
+      "learning_rate": 4.789814071513472e-06,
+      "loss": 0.514,
+      "step": 7060
+    },
+    {
+      "epoch": 2.773030377871079,
+      "grad_norm": 0.44904340222359973,
+      "learning_rate": 4.78975118690385e-06,
+      "loss": 0.5063,
+      "step": 7061
+    },
+    {
+      "epoch": 2.7734255371696714,
+      "grad_norm": 0.44818575851748277,
+      "learning_rate": 4.789688293301477e-06,
+      "loss": 0.5114,
+      "step": 7062
+    },
+    {
+      "epoch": 2.7738206964682637,
+      "grad_norm": 0.4394864938402016,
+      "learning_rate": 4.789625390706597e-06,
+      "loss": 0.4861,
+      "step": 7063
+    },
+    {
+      "epoch": 2.774215855766856,
+      "grad_norm": 0.4476957402660163,
+      "learning_rate": 4.789562479119459e-06,
+      "loss": 0.5039,
+      "step": 7064
+    },
+    {
+      "epoch": 2.774611015065448,
+      "grad_norm": 0.44866741905501123,
+      "learning_rate": 4.789499558540311e-06,
+      "loss": 0.4856,
+      "step": 7065
+    },
+    {
+      "epoch": 2.7750061743640404,
+      "grad_norm": 0.4657641738772069,
+      "learning_rate": 4.7894366289693984e-06,
+      "loss": 0.4909,
+      "step": 7066
+    },
+    {
+      "epoch": 2.7754013336626326,
+      "grad_norm": 0.43895805450614483,
+      "learning_rate": 4.789373690406969e-06,
+      "loss": 0.4977,
+      "step": 7067
+    },
+    {
+      "epoch": 2.775796492961225,
+      "grad_norm": 0.45870187996681017,
+      "learning_rate": 4.789310742853269e-06,
+      "loss": 0.4998,
+      "step": 7068
+    },
+    {
+      "epoch": 2.776191652259817,
+      "grad_norm": 0.45232602161964686,
+      "learning_rate": 4.789247786308548e-06,
+      "loss": 0.5082,
+      "step": 7069
+    },
+    {
+      "epoch": 2.7765868115584094,
+      "grad_norm": 0.4533454018434586,
+      "learning_rate": 4.789184820773052e-06,
+      "loss": 0.4736,
+      "step": 7070
+    },
+    {
+      "epoch": 2.7769819708570016,
+      "grad_norm": 0.4473360662431684,
+      "learning_rate": 4.7891218462470264e-06,
+      "loss": 0.4908,
+      "step": 7071
+    },
+    {
+      "epoch": 2.777377130155594,
+      "grad_norm": 0.4682252863668288,
+      "learning_rate": 4.7890588627307214e-06,
+      "loss": 0.5016,
+      "step": 7072
+    },
+    {
+      "epoch": 2.777772289454186,
+      "grad_norm": 0.4457755613086908,
+      "learning_rate": 4.788995870224382e-06,
+      "loss": 0.5183,
+      "step": 7073
+    },
+    {
+      "epoch": 2.7781674487527783,
+      "grad_norm": 0.46757695064672683,
+      "learning_rate": 4.788932868728258e-06,
+      "loss": 0.514,
+      "step": 7074
+    },
+    {
+      "epoch": 2.7785626080513706,
+      "grad_norm": 0.4538274673145632,
+      "learning_rate": 4.788869858242595e-06,
+      "loss": 0.5092,
+      "step": 7075
+    },
+    {
+      "epoch": 2.778957767349963,
+      "grad_norm": 0.4561259698590066,
+      "learning_rate": 4.788806838767642e-06,
+      "loss": 0.5331,
+      "step": 7076
+    },
+    {
+      "epoch": 2.779352926648555,
+      "grad_norm": 0.4589851990064589,
+      "learning_rate": 4.788743810303644e-06,
+      "loss": 0.5087,
+      "step": 7077
+    },
+    {
+      "epoch": 2.7797480859471473,
+      "grad_norm": 0.4587981064535018,
+      "learning_rate": 4.788680772850852e-06,
+      "loss": 0.5064,
+      "step": 7078
+    },
+    {
+      "epoch": 2.7801432452457395,
+      "grad_norm": 0.4495946809709087,
+      "learning_rate": 4.78861772640951e-06,
+      "loss": 0.5094,
+      "step": 7079
+    },
+    {
+      "epoch": 2.780538404544332,
+      "grad_norm": 0.4687277276868725,
+      "learning_rate": 4.788554670979868e-06,
+      "loss": 0.4998,
+      "step": 7080
+    },
+    {
+      "epoch": 2.7809335638429244,
+      "grad_norm": 0.4624450226026877,
+      "learning_rate": 4.7884916065621735e-06,
+      "loss": 0.5357,
+      "step": 7081
+    },
+    {
+      "epoch": 2.7813287231415167,
+      "grad_norm": 0.44737808383227995,
+      "learning_rate": 4.788428533156673e-06,
+      "loss": 0.5085,
+      "step": 7082
+    },
+    {
+      "epoch": 2.781723882440109,
+      "grad_norm": 0.4435365459154493,
+      "learning_rate": 4.788365450763614e-06,
+      "loss": 0.4855,
+      "step": 7083
+    },
+    {
+      "epoch": 2.782119041738701,
+      "grad_norm": 0.47758039605363733,
+      "learning_rate": 4.788302359383247e-06,
+      "loss": 0.5147,
+      "step": 7084
+    },
+    {
+      "epoch": 2.7825142010372934,
+      "grad_norm": 0.47467035291327786,
+      "learning_rate": 4.788239259015817e-06,
+      "loss": 0.5085,
+      "step": 7085
+    },
+    {
+      "epoch": 2.7829093603358857,
+      "grad_norm": 0.44586423467559405,
+      "learning_rate": 4.788176149661572e-06,
+      "loss": 0.4858,
+      "step": 7086
+    },
+    {
+      "epoch": 2.783304519634478,
+      "grad_norm": 0.45852446260824725,
+      "learning_rate": 4.7881130313207615e-06,
+      "loss": 0.4946,
+      "step": 7087
+    },
+    {
+      "epoch": 2.78369967893307,
+      "grad_norm": 0.4608076484794306,
+      "learning_rate": 4.7880499039936315e-06,
+      "loss": 0.5079,
+      "step": 7088
+    },
+    {
+      "epoch": 2.7840948382316624,
+      "grad_norm": 0.4653013987128913,
+      "learning_rate": 4.787986767680431e-06,
+      "loss": 0.5409,
+      "step": 7089
+    },
+    {
+      "epoch": 2.7844899975302546,
+      "grad_norm": 0.4563060161643037,
+      "learning_rate": 4.787923622381409e-06,
+      "loss": 0.4793,
+      "step": 7090
+    },
+    {
+      "epoch": 2.784885156828847,
+      "grad_norm": 0.474596234725646,
+      "learning_rate": 4.787860468096811e-06,
+      "loss": 0.5247,
+      "step": 7091
+    },
+    {
+      "epoch": 2.785280316127439,
+      "grad_norm": 0.4340035243691619,
+      "learning_rate": 4.787797304826887e-06,
+      "loss": 0.5038,
+      "step": 7092
+    },
+    {
+      "epoch": 2.7856754754260313,
+      "grad_norm": 0.45248266773447615,
+      "learning_rate": 4.787734132571884e-06,
+      "loss": 0.5287,
+      "step": 7093
+    },
+    {
+      "epoch": 2.7860706347246236,
+      "grad_norm": 0.45357497159528226,
+      "learning_rate": 4.7876709513320506e-06,
+      "loss": 0.4866,
+      "step": 7094
+    },
+    {
+      "epoch": 2.786465794023216,
+      "grad_norm": 0.471725493815604,
+      "learning_rate": 4.787607761107634e-06,
+      "loss": 0.509,
+      "step": 7095
+    },
+    {
+      "epoch": 2.786860953321808,
+      "grad_norm": 0.4550489319224383,
+      "learning_rate": 4.7875445618988846e-06,
+      "loss": 0.5069,
+      "step": 7096
+    },
+    {
+      "epoch": 2.7872561126204003,
+      "grad_norm": 0.4480112491649345,
+      "learning_rate": 4.787481353706049e-06,
+      "loss": 0.5214,
+      "step": 7097
+    },
+    {
+      "epoch": 2.7876512719189925,
+      "grad_norm": 0.46020685458924593,
+      "learning_rate": 4.787418136529376e-06,
+      "loss": 0.4913,
+      "step": 7098
+    },
+    {
+      "epoch": 2.7880464312175848,
+      "grad_norm": 0.47097392767674306,
+      "learning_rate": 4.787354910369113e-06,
+      "loss": 0.5046,
+      "step": 7099
+    },
+    {
+      "epoch": 2.788441590516177,
+      "grad_norm": 0.6635008391885756,
+      "learning_rate": 4.787291675225508e-06,
+      "loss": 0.5148,
+      "step": 7100
+    },
+    {
+      "epoch": 2.7888367498147693,
+      "grad_norm": 0.4483283359174607,
+      "learning_rate": 4.7872284310988115e-06,
+      "loss": 0.4828,
+      "step": 7101
+    },
+    {
+      "epoch": 2.7892319091133615,
+      "grad_norm": 0.45492414146412075,
+      "learning_rate": 4.78716517798927e-06,
+      "loss": 0.493,
+      "step": 7102
+    },
+    {
+      "epoch": 2.7896270684119537,
+      "grad_norm": 0.46950627047824917,
+      "learning_rate": 4.787101915897133e-06,
+      "loss": 0.5081,
+      "step": 7103
+    },
+    {
+      "epoch": 2.790022227710546,
+      "grad_norm": 0.46505419717794555,
+      "learning_rate": 4.787038644822649e-06,
+      "loss": 0.5116,
+      "step": 7104
+    },
+    {
+      "epoch": 2.790417387009138,
+      "grad_norm": 0.4544759503964953,
+      "learning_rate": 4.786975364766064e-06,
+      "loss": 0.5137,
+      "step": 7105
+    },
+    {
+      "epoch": 2.7908125463077305,
+      "grad_norm": 0.4615992806036096,
+      "learning_rate": 4.786912075727631e-06,
+      "loss": 0.5141,
+      "step": 7106
+    },
+    {
+      "epoch": 2.7912077056063227,
+      "grad_norm": 0.44988730317410236,
+      "learning_rate": 4.786848777707594e-06,
+      "loss": 0.5142,
+      "step": 7107
+    },
+    {
+      "epoch": 2.791602864904915,
+      "grad_norm": 0.46645509824578013,
+      "learning_rate": 4.786785470706204e-06,
+      "loss": 0.5091,
+      "step": 7108
+    },
+    {
+      "epoch": 2.791998024203507,
+      "grad_norm": 0.4714405835689825,
+      "learning_rate": 4.78672215472371e-06,
+      "loss": 0.5221,
+      "step": 7109
+    },
+    {
+      "epoch": 2.7923931835020994,
+      "grad_norm": 0.4614676395031843,
+      "learning_rate": 4.78665882976036e-06,
+      "loss": 0.5036,
+      "step": 7110
+    },
+    {
+      "epoch": 2.7927883428006917,
+      "grad_norm": 0.45921742192416665,
+      "learning_rate": 4.786595495816402e-06,
+      "loss": 0.5073,
+      "step": 7111
+    },
+    {
+      "epoch": 2.793183502099284,
+      "grad_norm": 0.46441158320725995,
+      "learning_rate": 4.786532152892086e-06,
+      "loss": 0.5343,
+      "step": 7112
+    },
+    {
+      "epoch": 2.793578661397876,
+      "grad_norm": 0.4473283825529127,
+      "learning_rate": 4.78646880098766e-06,
+      "loss": 0.4916,
+      "step": 7113
+    },
+    {
+      "epoch": 2.7939738206964684,
+      "grad_norm": 0.4565617472568001,
+      "learning_rate": 4.786405440103372e-06,
+      "loss": 0.4896,
+      "step": 7114
+    },
+    {
+      "epoch": 2.7943689799950606,
+      "grad_norm": 0.4484189292638863,
+      "learning_rate": 4.786342070239473e-06,
+      "loss": 0.4876,
+      "step": 7115
+    },
+    {
+      "epoch": 2.794764139293653,
+      "grad_norm": 0.4561057568670873,
+      "learning_rate": 4.78627869139621e-06,
+      "loss": 0.5226,
+      "step": 7116
+    },
+    {
+      "epoch": 2.795159298592245,
+      "grad_norm": 0.45392026125118257,
+      "learning_rate": 4.786215303573834e-06,
+      "loss": 0.5162,
+      "step": 7117
+    },
+    {
+      "epoch": 2.7955544578908373,
+      "grad_norm": 0.44001406352495115,
+      "learning_rate": 4.7861519067725904e-06,
+      "loss": 0.5084,
+      "step": 7118
+    },
+    {
+      "epoch": 2.7959496171894296,
+      "grad_norm": 0.45328763812652456,
+      "learning_rate": 4.786088500992732e-06,
+      "loss": 0.5035,
+      "step": 7119
+    },
+    {
+      "epoch": 2.796344776488022,
+      "grad_norm": 0.47975356308620704,
+      "learning_rate": 4.786025086234505e-06,
+      "loss": 0.5178,
+      "step": 7120
+    },
+    {
+      "epoch": 2.796739935786614,
+      "grad_norm": 0.4358024295607447,
+      "learning_rate": 4.78596166249816e-06,
+      "loss": 0.508,
+      "step": 7121
+    },
+    {
+      "epoch": 2.7971350950852063,
+      "grad_norm": 0.45560496266499945,
+      "learning_rate": 4.785898229783946e-06,
+      "loss": 0.522,
+      "step": 7122
+    },
+    {
+      "epoch": 2.7975302543837985,
+      "grad_norm": 0.4493585711669279,
+      "learning_rate": 4.785834788092112e-06,
+      "loss": 0.5086,
+      "step": 7123
+    },
+    {
+      "epoch": 2.797925413682391,
+      "grad_norm": 0.4518530962962152,
+      "learning_rate": 4.785771337422906e-06,
+      "loss": 0.5163,
+      "step": 7124
+    },
+    {
+      "epoch": 2.798320572980983,
+      "grad_norm": 0.44724673545538657,
+      "learning_rate": 4.7857078777765796e-06,
+      "loss": 0.5082,
+      "step": 7125
+    },
+    {
+      "epoch": 2.7987157322795753,
+      "grad_norm": 0.46761704301589785,
+      "learning_rate": 4.785644409153379e-06,
+      "loss": 0.514,
+      "step": 7126
+    },
+    {
+      "epoch": 2.7991108915781675,
+      "grad_norm": 0.4533222753692102,
+      "learning_rate": 4.785580931553556e-06,
+      "loss": 0.5179,
+      "step": 7127
+    },
+    {
+      "epoch": 2.7995060508767597,
+      "grad_norm": 0.4559014721619301,
+      "learning_rate": 4.7855174449773595e-06,
+      "loss": 0.508,
+      "step": 7128
+    },
+    {
+      "epoch": 2.799901210175352,
+      "grad_norm": 0.4677071703771446,
+      "learning_rate": 4.785453949425038e-06,
+      "loss": 0.5001,
+      "step": 7129
+    },
+    {
+      "epoch": 2.8002963694739442,
+      "grad_norm": 0.4786331564258494,
+      "learning_rate": 4.785390444896841e-06,
+      "loss": 0.5206,
+      "step": 7130
+    },
+    {
+      "epoch": 2.8006915287725365,
+      "grad_norm": 0.4518046203867029,
+      "learning_rate": 4.7853269313930175e-06,
+      "loss": 0.5311,
+      "step": 7131
+    },
+    {
+      "epoch": 2.8010866880711287,
+      "grad_norm": 0.4493408611983403,
+      "learning_rate": 4.785263408913818e-06,
+      "loss": 0.5078,
+      "step": 7132
+    },
+    {
+      "epoch": 2.801481847369721,
+      "grad_norm": 0.4413293231411184,
+      "learning_rate": 4.7851998774594915e-06,
+      "loss": 0.5015,
+      "step": 7133
+    },
+    {
+      "epoch": 2.801877006668313,
+      "grad_norm": 0.4405329418854599,
+      "learning_rate": 4.7851363370302875e-06,
+      "loss": 0.4971,
+      "step": 7134
+    },
+    {
+      "epoch": 2.8022721659669054,
+      "grad_norm": 0.47861777339015044,
+      "learning_rate": 4.785072787626456e-06,
+      "loss": 0.505,
+      "step": 7135
+    },
+    {
+      "epoch": 2.8026673252654977,
+      "grad_norm": 0.4540541349102917,
+      "learning_rate": 4.785009229248246e-06,
+      "loss": 0.5346,
+      "step": 7136
+    },
+    {
+      "epoch": 2.80306248456409,
+      "grad_norm": 0.4660733299034907,
+      "learning_rate": 4.784945661895907e-06,
+      "loss": 0.5054,
+      "step": 7137
+    },
+    {
+      "epoch": 2.803457643862682,
+      "grad_norm": 0.454650518694851,
+      "learning_rate": 4.784882085569689e-06,
+      "loss": 0.5111,
+      "step": 7138
+    },
+    {
+      "epoch": 2.8038528031612744,
+      "grad_norm": 2.699233093548558,
+      "learning_rate": 4.784818500269842e-06,
+      "loss": 0.512,
+      "step": 7139
+    },
+    {
+      "epoch": 2.8042479624598666,
+      "grad_norm": 0.44347795457544964,
+      "learning_rate": 4.7847549059966144e-06,
+      "loss": 0.509,
+      "step": 7140
+    },
+    {
+      "epoch": 2.804643121758459,
+      "grad_norm": 0.44865460293087767,
+      "learning_rate": 4.784691302750257e-06,
+      "loss": 0.4965,
+      "step": 7141
+    },
+    {
+      "epoch": 2.805038281057051,
+      "grad_norm": 0.44186758621797884,
+      "learning_rate": 4.78462769053102e-06,
+      "loss": 0.4935,
+      "step": 7142
+    },
+    {
+      "epoch": 2.8054334403556433,
+      "grad_norm": 0.4690246679204338,
+      "learning_rate": 4.784564069339154e-06,
+      "loss": 0.4998,
+      "step": 7143
+    },
+    {
+      "epoch": 2.8058285996542356,
+      "grad_norm": 0.4413211990503885,
+      "learning_rate": 4.7845004391749065e-06,
+      "loss": 0.5136,
+      "step": 7144
+    },
+    {
+      "epoch": 2.806223758952828,
+      "grad_norm": 0.4477276180507849,
+      "learning_rate": 4.784436800038528e-06,
+      "loss": 0.4832,
+      "step": 7145
+    },
+    {
+      "epoch": 2.80661891825142,
+      "grad_norm": 0.4624922297012007,
+      "learning_rate": 4.784373151930269e-06,
+      "loss": 0.5086,
+      "step": 7146
+    },
+    {
+      "epoch": 2.8070140775500123,
+      "grad_norm": 0.452716126237242,
+      "learning_rate": 4.78430949485038e-06,
+      "loss": 0.5171,
+      "step": 7147
+    },
+    {
+      "epoch": 2.8074092368486046,
+      "grad_norm": 0.4531177548368098,
+      "learning_rate": 4.78424582879911e-06,
+      "loss": 0.5133,
+      "step": 7148
+    },
+    {
+      "epoch": 2.807804396147197,
+      "grad_norm": 0.45184974430917935,
+      "learning_rate": 4.7841821537767095e-06,
+      "loss": 0.5062,
+      "step": 7149
+    },
+    {
+      "epoch": 2.808199555445789,
+      "grad_norm": 0.45422086326819416,
+      "learning_rate": 4.784118469783429e-06,
+      "loss": 0.5039,
+      "step": 7150
+    },
+    {
+      "epoch": 2.8085947147443813,
+      "grad_norm": 0.4253947576601844,
+      "learning_rate": 4.784054776819517e-06,
+      "loss": 0.4924,
+      "step": 7151
+    },
+    {
+      "epoch": 2.8089898740429735,
+      "grad_norm": 0.4578816649664364,
+      "learning_rate": 4.7839910748852255e-06,
+      "loss": 0.5276,
+      "step": 7152
+    },
+    {
+      "epoch": 2.8093850333415658,
+      "grad_norm": 0.4483518419377659,
+      "learning_rate": 4.7839273639808035e-06,
+      "loss": 0.4853,
+      "step": 7153
+    },
+    {
+      "epoch": 2.809780192640158,
+      "grad_norm": 0.46266492861443637,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.5165,
+      "step": 7154
+    },
+    {
+      "epoch": 2.8101753519387502,
+      "grad_norm": 0.45201465369049515,
+      "learning_rate": 4.783799915262571e-06,
+      "loss": 0.5112,
+      "step": 7155
+    },
+    {
+      "epoch": 2.8105705112373425,
+      "grad_norm": 0.47492599053950296,
+      "learning_rate": 4.783736177449262e-06,
+      "loss": 0.5106,
+      "step": 7156
+    },
+    {
+      "epoch": 2.8109656705359347,
+      "grad_norm": 0.46952297692513184,
+      "learning_rate": 4.783672430666822e-06,
+      "loss": 0.5118,
+      "step": 7157
+    },
+    {
+      "epoch": 2.811360829834527,
+      "grad_norm": 0.45235950979008643,
+      "learning_rate": 4.783608674915505e-06,
+      "loss": 0.5173,
+      "step": 7158
+    },
+    {
+      "epoch": 2.811755989133119,
+      "grad_norm": 0.4900581695225324,
+      "learning_rate": 4.783544910195559e-06,
+      "loss": 0.506,
+      "step": 7159
+    },
+    {
+      "epoch": 2.8121511484317114,
+      "grad_norm": 0.4777901985820845,
+      "learning_rate": 4.783481136507236e-06,
+      "loss": 0.5085,
+      "step": 7160
+    },
+    {
+      "epoch": 2.8125463077303037,
+      "grad_norm": 0.5041776097941606,
+      "learning_rate": 4.783417353850785e-06,
+      "loss": 0.5097,
+      "step": 7161
+    },
+    {
+      "epoch": 2.812941467028896,
+      "grad_norm": 0.45039977578331913,
+      "learning_rate": 4.7833535622264565e-06,
+      "loss": 0.4931,
+      "step": 7162
+    },
+    {
+      "epoch": 2.813336626327488,
+      "grad_norm": 0.46033667197541245,
+      "learning_rate": 4.783289761634502e-06,
+      "loss": 0.5218,
+      "step": 7163
+    },
+    {
+      "epoch": 2.8137317856260804,
+      "grad_norm": 0.5045096295509437,
+      "learning_rate": 4.783225952075173e-06,
+      "loss": 0.506,
+      "step": 7164
+    },
+    {
+      "epoch": 2.8141269449246726,
+      "grad_norm": 0.46808375423952775,
+      "learning_rate": 4.783162133548718e-06,
+      "loss": 0.5239,
+      "step": 7165
+    },
+    {
+      "epoch": 2.814522104223265,
+      "grad_norm": 0.47727715778274427,
+      "learning_rate": 4.783098306055389e-06,
+      "loss": 0.5242,
+      "step": 7166
+    },
+    {
+      "epoch": 2.814917263521857,
+      "grad_norm": 0.49741116135472996,
+      "learning_rate": 4.7830344695954356e-06,
+      "loss": 0.5105,
+      "step": 7167
+    },
+    {
+      "epoch": 2.8153124228204494,
+      "grad_norm": 0.4765653183745411,
+      "learning_rate": 4.78297062416911e-06,
+      "loss": 0.5158,
+      "step": 7168
+    },
+    {
+      "epoch": 2.8157075821190416,
+      "grad_norm": 0.441604632522944,
+      "learning_rate": 4.782906769776661e-06,
+      "loss": 0.4992,
+      "step": 7169
+    },
+    {
+      "epoch": 2.816102741417634,
+      "grad_norm": 0.7388460812324761,
+      "learning_rate": 4.782842906418341e-06,
+      "loss": 0.5187,
+      "step": 7170
+    },
+    {
+      "epoch": 2.816497900716226,
+      "grad_norm": 0.4647214598823728,
+      "learning_rate": 4.7827790340944e-06,
+      "loss": 0.5028,
+      "step": 7171
+    },
+    {
+      "epoch": 2.8168930600148183,
+      "grad_norm": 0.46169878022399585,
+      "learning_rate": 4.7827151528050894e-06,
+      "loss": 0.5102,
+      "step": 7172
+    },
+    {
+      "epoch": 2.8172882193134106,
+      "grad_norm": 0.46029778083095246,
+      "learning_rate": 4.782651262550661e-06,
+      "loss": 0.51,
+      "step": 7173
+    },
+    {
+      "epoch": 2.817683378612003,
+      "grad_norm": 0.45078052262587254,
+      "learning_rate": 4.782587363331363e-06,
+      "loss": 0.5207,
+      "step": 7174
+    },
+    {
+      "epoch": 2.818078537910595,
+      "grad_norm": 0.46397038369117694,
+      "learning_rate": 4.782523455147448e-06,
+      "loss": 0.5054,
+      "step": 7175
+    },
+    {
+      "epoch": 2.8184736972091873,
+      "grad_norm": 0.45332697701497565,
+      "learning_rate": 4.782459537999168e-06,
+      "loss": 0.4939,
+      "step": 7176
+    },
+    {
+      "epoch": 2.8188688565077795,
+      "grad_norm": 0.45625065796075703,
+      "learning_rate": 4.782395611886771e-06,
+      "loss": 0.5069,
+      "step": 7177
+    },
+    {
+      "epoch": 2.8192640158063718,
+      "grad_norm": 0.45551406905446623,
+      "learning_rate": 4.7823316768105115e-06,
+      "loss": 0.5109,
+      "step": 7178
+    },
+    {
+      "epoch": 2.819659175104964,
+      "grad_norm": 0.46505352472091316,
+      "learning_rate": 4.782267732770639e-06,
+      "loss": 0.5063,
+      "step": 7179
+    },
+    {
+      "epoch": 2.8200543344035562,
+      "grad_norm": 0.46864702173796885,
+      "learning_rate": 4.782203779767404e-06,
+      "loss": 0.4998,
+      "step": 7180
+    },
+    {
+      "epoch": 2.8204494937021485,
+      "grad_norm": 0.44137492397186884,
+      "learning_rate": 4.782139817801059e-06,
+      "loss": 0.5154,
+      "step": 7181
+    },
+    {
+      "epoch": 2.8208446530007407,
+      "grad_norm": 0.457276576966862,
+      "learning_rate": 4.782075846871855e-06,
+      "loss": 0.506,
+      "step": 7182
+    },
+    {
+      "epoch": 2.821239812299333,
+      "grad_norm": 0.4617777248102637,
+      "learning_rate": 4.782011866980042e-06,
+      "loss": 0.5193,
+      "step": 7183
+    },
+    {
+      "epoch": 2.821634971597925,
+      "grad_norm": 0.4495925503990227,
+      "learning_rate": 4.781947878125872e-06,
+      "loss": 0.5091,
+      "step": 7184
+    },
+    {
+      "epoch": 2.8220301308965174,
+      "grad_norm": 0.4464844531883246,
+      "learning_rate": 4.781883880309597e-06,
+      "loss": 0.5096,
+      "step": 7185
+    },
+    {
+      "epoch": 2.8224252901951097,
+      "grad_norm": 0.4553422801335472,
+      "learning_rate": 4.781819873531467e-06,
+      "loss": 0.4845,
+      "step": 7186
+    },
+    {
+      "epoch": 2.822820449493702,
+      "grad_norm": 0.508345334941812,
+      "learning_rate": 4.781755857791734e-06,
+      "loss": 0.512,
+      "step": 7187
+    },
+    {
+      "epoch": 2.823215608792294,
+      "grad_norm": 0.46314240936594236,
+      "learning_rate": 4.78169183309065e-06,
+      "loss": 0.5283,
+      "step": 7188
+    },
+    {
+      "epoch": 2.8236107680908864,
+      "grad_norm": 0.45552834604576314,
+      "learning_rate": 4.781627799428466e-06,
+      "loss": 0.4988,
+      "step": 7189
+    },
+    {
+      "epoch": 2.8240059273894786,
+      "grad_norm": 0.43823171263924665,
+      "learning_rate": 4.781563756805434e-06,
+      "loss": 0.4898,
+      "step": 7190
+    },
+    {
+      "epoch": 2.824401086688071,
+      "grad_norm": 0.4425700045729354,
+      "learning_rate": 4.781499705221805e-06,
+      "loss": 0.4902,
+      "step": 7191
+    },
+    {
+      "epoch": 2.824796245986663,
+      "grad_norm": 0.46454792833992753,
+      "learning_rate": 4.7814356446778294e-06,
+      "loss": 0.5339,
+      "step": 7192
+    },
+    {
+      "epoch": 2.8251914052852554,
+      "grad_norm": 0.4766460659681996,
+      "learning_rate": 4.781371575173762e-06,
+      "loss": 0.5107,
+      "step": 7193
+    },
+    {
+      "epoch": 2.8255865645838476,
+      "grad_norm": 0.46813319069331144,
+      "learning_rate": 4.78130749670985e-06,
+      "loss": 0.5201,
+      "step": 7194
+    },
+    {
+      "epoch": 2.82598172388244,
+      "grad_norm": 0.4545253394863707,
+      "learning_rate": 4.781243409286349e-06,
+      "loss": 0.5225,
+      "step": 7195
+    },
+    {
+      "epoch": 2.826376883181032,
+      "grad_norm": 0.439864191746294,
+      "learning_rate": 4.781179312903509e-06,
+      "loss": 0.5041,
+      "step": 7196
+    },
+    {
+      "epoch": 2.8267720424796243,
+      "grad_norm": 0.45086076483594706,
+      "learning_rate": 4.781115207561582e-06,
+      "loss": 0.5045,
+      "step": 7197
+    },
+    {
+      "epoch": 2.827167201778217,
+      "grad_norm": 0.46994834299009564,
+      "learning_rate": 4.781051093260819e-06,
+      "loss": 0.4944,
+      "step": 7198
+    },
+    {
+      "epoch": 2.8275623610768092,
+      "grad_norm": 0.4455482853226493,
+      "learning_rate": 4.7809869700014726e-06,
+      "loss": 0.5059,
+      "step": 7199
+    },
+    {
+      "epoch": 2.8279575203754015,
+      "grad_norm": 0.45802471378032417,
+      "learning_rate": 4.7809228377837934e-06,
+      "loss": 0.5053,
+      "step": 7200
+    },
+    {
+      "epoch": 2.8283526796739937,
+      "grad_norm": 0.4642886398985806,
+      "learning_rate": 4.780858696608036e-06,
+      "loss": 0.5204,
+      "step": 7201
+    },
+    {
+      "epoch": 2.828747838972586,
+      "grad_norm": 0.4489929183982641,
+      "learning_rate": 4.78079454647445e-06,
+      "loss": 0.5073,
+      "step": 7202
+    },
+    {
+      "epoch": 2.829142998271178,
+      "grad_norm": 0.4626978965499321,
+      "learning_rate": 4.7807303873832875e-06,
+      "loss": 0.5068,
+      "step": 7203
+    },
+    {
+      "epoch": 2.8295381575697705,
+      "grad_norm": 0.45282255975863683,
+      "learning_rate": 4.780666219334802e-06,
+      "loss": 0.5095,
+      "step": 7204
+    },
+    {
+      "epoch": 2.8299333168683627,
+      "grad_norm": 0.45313611105031126,
+      "learning_rate": 4.780602042329244e-06,
+      "loss": 0.5198,
+      "step": 7205
+    },
+    {
+      "epoch": 2.830328476166955,
+      "grad_norm": 0.4584691584839706,
+      "learning_rate": 4.7805378563668655e-06,
+      "loss": 0.5147,
+      "step": 7206
+    },
+    {
+      "epoch": 2.830723635465547,
+      "grad_norm": 0.46044793656981825,
+      "learning_rate": 4.780473661447921e-06,
+      "loss": 0.5205,
+      "step": 7207
+    },
+    {
+      "epoch": 2.8311187947641394,
+      "grad_norm": 0.4583164398209612,
+      "learning_rate": 4.7804094575726585e-06,
+      "loss": 0.5222,
+      "step": 7208
+    },
+    {
+      "epoch": 2.8315139540627317,
+      "grad_norm": 0.4498029349438353,
+      "learning_rate": 4.780345244741333e-06,
+      "loss": 0.4927,
+      "step": 7209
+    },
+    {
+      "epoch": 2.831909113361324,
+      "grad_norm": 0.44328976919231206,
+      "learning_rate": 4.780281022954196e-06,
+      "loss": 0.5102,
+      "step": 7210
+    },
+    {
+      "epoch": 2.832304272659916,
+      "grad_norm": 0.44810747427037423,
+      "learning_rate": 4.7802167922115e-06,
+      "loss": 0.5071,
+      "step": 7211
+    },
+    {
+      "epoch": 2.8326994319585084,
+      "grad_norm": 0.44384368117197937,
+      "learning_rate": 4.780152552513499e-06,
+      "loss": 0.4983,
+      "step": 7212
+    },
+    {
+      "epoch": 2.8330945912571006,
+      "grad_norm": 0.44676501579869293,
+      "learning_rate": 4.7800883038604404e-06,
+      "loss": 0.5002,
+      "step": 7213
+    },
+    {
+      "epoch": 2.833489750555693,
+      "grad_norm": 0.44242213343088016,
+      "learning_rate": 4.780024046252581e-06,
+      "loss": 0.5114,
+      "step": 7214
+    },
+    {
+      "epoch": 2.833884909854285,
+      "grad_norm": 0.44041006597665694,
+      "learning_rate": 4.779959779690171e-06,
+      "loss": 0.5059,
+      "step": 7215
+    },
+    {
+      "epoch": 2.8342800691528773,
+      "grad_norm": 0.4440298358961055,
+      "learning_rate": 4.779895504173464e-06,
+      "loss": 0.5156,
+      "step": 7216
+    },
+    {
+      "epoch": 2.8346752284514696,
+      "grad_norm": 0.45776869409256077,
+      "learning_rate": 4.779831219702712e-06,
+      "loss": 0.5051,
+      "step": 7217
+    },
+    {
+      "epoch": 2.835070387750062,
+      "grad_norm": 0.43770623591530705,
+      "learning_rate": 4.7797669262781665e-06,
+      "loss": 0.4988,
+      "step": 7218
+    },
+    {
+      "epoch": 2.835465547048654,
+      "grad_norm": 0.4442467675600192,
+      "learning_rate": 4.779702623900082e-06,
+      "loss": 0.5167,
+      "step": 7219
+    },
+    {
+      "epoch": 2.8358607063472463,
+      "grad_norm": 0.4453421691195409,
+      "learning_rate": 4.779638312568708e-06,
+      "loss": 0.498,
+      "step": 7220
+    },
+    {
+      "epoch": 2.8362558656458385,
+      "grad_norm": 0.43148125749848565,
+      "learning_rate": 4.779573992284301e-06,
+      "loss": 0.4994,
+      "step": 7221
+    },
+    {
+      "epoch": 2.8366510249444308,
+      "grad_norm": 0.4384294284300938,
+      "learning_rate": 4.779509663047111e-06,
+      "loss": 0.4924,
+      "step": 7222
+    },
+    {
+      "epoch": 2.837046184243023,
+      "grad_norm": 0.450005448688682,
+      "learning_rate": 4.779445324857391e-06,
+      "loss": 0.5015,
+      "step": 7223
+    },
+    {
+      "epoch": 2.8374413435416153,
+      "grad_norm": 0.4383831955520918,
+      "learning_rate": 4.779380977715394e-06,
+      "loss": 0.5075,
+      "step": 7224
+    },
+    {
+      "epoch": 2.8378365028402075,
+      "grad_norm": 0.44872433012548524,
+      "learning_rate": 4.7793166216213725e-06,
+      "loss": 0.4907,
+      "step": 7225
+    },
+    {
+      "epoch": 2.8382316621387997,
+      "grad_norm": 0.4486414248911323,
+      "learning_rate": 4.77925225657558e-06,
+      "loss": 0.5103,
+      "step": 7226
+    },
+    {
+      "epoch": 2.838626821437392,
+      "grad_norm": 0.4505470486129958,
+      "learning_rate": 4.7791878825782675e-06,
+      "loss": 0.4949,
+      "step": 7227
+    },
+    {
+      "epoch": 2.839021980735984,
+      "grad_norm": 0.45187161145250815,
+      "learning_rate": 4.77912349962969e-06,
+      "loss": 0.5004,
+      "step": 7228
+    },
+    {
+      "epoch": 2.8394171400345765,
+      "grad_norm": 0.4532129642345992,
+      "learning_rate": 4.779059107730099e-06,
+      "loss": 0.4977,
+      "step": 7229
+    },
+    {
+      "epoch": 2.8398122993331687,
+      "grad_norm": 0.4463390504441417,
+      "learning_rate": 4.7789947068797474e-06,
+      "loss": 0.5049,
+      "step": 7230
+    },
+    {
+      "epoch": 2.840207458631761,
+      "grad_norm": 0.44255699499296886,
+      "learning_rate": 4.7789302970788895e-06,
+      "loss": 0.4965,
+      "step": 7231
+    },
+    {
+      "epoch": 2.840602617930353,
+      "grad_norm": 0.46126441571913374,
+      "learning_rate": 4.7788658783277765e-06,
+      "loss": 0.5196,
+      "step": 7232
+    },
+    {
+      "epoch": 2.8409977772289454,
+      "grad_norm": 0.4545844577709667,
+      "learning_rate": 4.778801450626662e-06,
+      "loss": 0.5099,
+      "step": 7233
+    },
+    {
+      "epoch": 2.8413929365275377,
+      "grad_norm": 0.45959479224584476,
+      "learning_rate": 4.7787370139758e-06,
+      "loss": 0.5083,
+      "step": 7234
+    },
+    {
+      "epoch": 2.84178809582613,
+      "grad_norm": 0.4506555503493617,
+      "learning_rate": 4.7786725683754415e-06,
+      "loss": 0.5267,
+      "step": 7235
+    },
+    {
+      "epoch": 2.842183255124722,
+      "grad_norm": 0.4569111610608343,
+      "learning_rate": 4.7786081138258414e-06,
+      "loss": 0.5184,
+      "step": 7236
+    },
+    {
+      "epoch": 2.8425784144233144,
+      "grad_norm": 0.4436242554559036,
+      "learning_rate": 4.778543650327252e-06,
+      "loss": 0.4825,
+      "step": 7237
+    },
+    {
+      "epoch": 2.8429735737219066,
+      "grad_norm": 0.4452045626348084,
+      "learning_rate": 4.778479177879928e-06,
+      "loss": 0.5106,
+      "step": 7238
+    },
+    {
+      "epoch": 2.843368733020499,
+      "grad_norm": 0.44570436675947306,
+      "learning_rate": 4.77841469648412e-06,
+      "loss": 0.5197,
+      "step": 7239
+    },
+    {
+      "epoch": 2.843763892319091,
+      "grad_norm": 0.44104364743730273,
+      "learning_rate": 4.778350206140083e-06,
+      "loss": 0.4942,
+      "step": 7240
+    },
+    {
+      "epoch": 2.8441590516176833,
+      "grad_norm": 0.470164189035705,
+      "learning_rate": 4.77828570684807e-06,
+      "loss": 0.52,
+      "step": 7241
+    },
+    {
+      "epoch": 2.8445542109162756,
+      "grad_norm": 0.45844457094618113,
+      "learning_rate": 4.778221198608333e-06,
+      "loss": 0.4999,
+      "step": 7242
+    },
+    {
+      "epoch": 2.844949370214868,
+      "grad_norm": 0.45188438120573965,
+      "learning_rate": 4.778156681421129e-06,
+      "loss": 0.5137,
+      "step": 7243
+    },
+    {
+      "epoch": 2.84534452951346,
+      "grad_norm": 0.4865736628712067,
+      "learning_rate": 4.778092155286707e-06,
+      "loss": 0.52,
+      "step": 7244
+    },
+    {
+      "epoch": 2.8457396888120523,
+      "grad_norm": 0.4618516637382695,
+      "learning_rate": 4.778027620205323e-06,
+      "loss": 0.5022,
+      "step": 7245
+    },
+    {
+      "epoch": 2.8461348481106445,
+      "grad_norm": 0.4518757121486141,
+      "learning_rate": 4.77796307617723e-06,
+      "loss": 0.513,
+      "step": 7246
+    },
+    {
+      "epoch": 2.846530007409237,
+      "grad_norm": 0.45286344906811393,
+      "learning_rate": 4.777898523202681e-06,
+      "loss": 0.4984,
+      "step": 7247
+    },
+    {
+      "epoch": 2.846925166707829,
+      "grad_norm": 0.46241029465759526,
+      "learning_rate": 4.777833961281929e-06,
+      "loss": 0.5241,
+      "step": 7248
+    },
+    {
+      "epoch": 2.8473203260064213,
+      "grad_norm": 0.4572057988626441,
+      "learning_rate": 4.7777693904152295e-06,
+      "loss": 0.4991,
+      "step": 7249
+    },
+    {
+      "epoch": 2.8477154853050135,
+      "grad_norm": 0.46688359869533963,
+      "learning_rate": 4.7777048106028345e-06,
+      "loss": 0.5258,
+      "step": 7250
+    },
+    {
+      "epoch": 2.8481106446036057,
+      "grad_norm": 0.4419843786987222,
+      "learning_rate": 4.777640221844998e-06,
+      "loss": 0.5003,
+      "step": 7251
+    },
+    {
+      "epoch": 2.848505803902198,
+      "grad_norm": 0.4509604918984796,
+      "learning_rate": 4.777575624141975e-06,
+      "loss": 0.4943,
+      "step": 7252
+    },
+    {
+      "epoch": 2.8489009632007902,
+      "grad_norm": 0.4544386957164005,
+      "learning_rate": 4.777511017494017e-06,
+      "loss": 0.5191,
+      "step": 7253
+    },
+    {
+      "epoch": 2.8492961224993825,
+      "grad_norm": 0.46217731638209764,
+      "learning_rate": 4.777446401901378e-06,
+      "loss": 0.5026,
+      "step": 7254
+    },
+    {
+      "epoch": 2.8496912817979747,
+      "grad_norm": 0.46198789235520654,
+      "learning_rate": 4.777381777364314e-06,
+      "loss": 0.5046,
+      "step": 7255
+    },
+    {
+      "epoch": 2.850086441096567,
+      "grad_norm": 0.467501208212079,
+      "learning_rate": 4.777317143883076e-06,
+      "loss": 0.5054,
+      "step": 7256
+    },
+    {
+      "epoch": 2.850481600395159,
+      "grad_norm": 0.45572805493163815,
+      "learning_rate": 4.77725250145792e-06,
+      "loss": 0.502,
+      "step": 7257
+    },
+    {
+      "epoch": 2.8508767596937514,
+      "grad_norm": 0.4585262214570415,
+      "learning_rate": 4.777187850089098e-06,
+      "loss": 0.5001,
+      "step": 7258
+    },
+    {
+      "epoch": 2.8512719189923437,
+      "grad_norm": 0.46257434246120355,
+      "learning_rate": 4.777123189776865e-06,
+      "loss": 0.4884,
+      "step": 7259
+    },
+    {
+      "epoch": 2.851667078290936,
+      "grad_norm": 0.44787734910546845,
+      "learning_rate": 4.777058520521476e-06,
+      "loss": 0.4854,
+      "step": 7260
+    },
+    {
+      "epoch": 2.852062237589528,
+      "grad_norm": 0.4679108827035021,
+      "learning_rate": 4.7769938423231825e-06,
+      "loss": 0.4984,
+      "step": 7261
+    },
+    {
+      "epoch": 2.8524573968881204,
+      "grad_norm": 0.5824913189397212,
+      "learning_rate": 4.776929155182241e-06,
+      "loss": 0.5165,
+      "step": 7262
+    },
+    {
+      "epoch": 2.8528525561867126,
+      "grad_norm": 0.47000334347932204,
+      "learning_rate": 4.776864459098904e-06,
+      "loss": 0.5063,
+      "step": 7263
+    },
+    {
+      "epoch": 2.853247715485305,
+      "grad_norm": 0.46271998546318105,
+      "learning_rate": 4.776799754073425e-06,
+      "loss": 0.4995,
+      "step": 7264
+    },
+    {
+      "epoch": 2.853642874783897,
+      "grad_norm": 0.46020642262681605,
+      "learning_rate": 4.776735040106061e-06,
+      "loss": 0.5026,
+      "step": 7265
+    },
+    {
+      "epoch": 2.8540380340824894,
+      "grad_norm": 0.4570714611160204,
+      "learning_rate": 4.776670317197063e-06,
+      "loss": 0.4976,
+      "step": 7266
+    },
+    {
+      "epoch": 2.8544331933810816,
+      "grad_norm": 0.45694522553571065,
+      "learning_rate": 4.776605585346687e-06,
+      "loss": 0.4947,
+      "step": 7267
+    },
+    {
+      "epoch": 2.854828352679674,
+      "grad_norm": 0.45434827116015725,
+      "learning_rate": 4.776540844555186e-06,
+      "loss": 0.5041,
+      "step": 7268
+    },
+    {
+      "epoch": 2.8552235119782665,
+      "grad_norm": 0.4449151544183041,
+      "learning_rate": 4.776476094822815e-06,
+      "loss": 0.4964,
+      "step": 7269
+    },
+    {
+      "epoch": 2.8556186712768588,
+      "grad_norm": 0.455998711224259,
+      "learning_rate": 4.7764113361498284e-06,
+      "loss": 0.5014,
+      "step": 7270
+    },
+    {
+      "epoch": 2.856013830575451,
+      "grad_norm": 0.4336150479227637,
+      "learning_rate": 4.776346568536481e-06,
+      "loss": 0.4975,
+      "step": 7271
+    },
+    {
+      "epoch": 2.8564089898740432,
+      "grad_norm": 0.46393244313991583,
+      "learning_rate": 4.776281791983026e-06,
+      "loss": 0.5004,
+      "step": 7272
+    },
+    {
+      "epoch": 2.8568041491726355,
+      "grad_norm": 0.4738720171848614,
+      "learning_rate": 4.776217006489719e-06,
+      "loss": 0.5105,
+      "step": 7273
+    },
+    {
+      "epoch": 2.8571993084712277,
+      "grad_norm": 0.4533385874249988,
+      "learning_rate": 4.776152212056813e-06,
+      "loss": 0.4789,
+      "step": 7274
+    },
+    {
+      "epoch": 2.85759446776982,
+      "grad_norm": 0.45076697220608347,
+      "learning_rate": 4.7760874086845635e-06,
+      "loss": 0.5222,
+      "step": 7275
+    },
+    {
+      "epoch": 2.857989627068412,
+      "grad_norm": 0.45274267999042384,
+      "learning_rate": 4.7760225963732255e-06,
+      "loss": 0.5075,
+      "step": 7276
+    },
+    {
+      "epoch": 2.8583847863670044,
+      "grad_norm": 0.4543131577956172,
+      "learning_rate": 4.775957775123052e-06,
+      "loss": 0.495,
+      "step": 7277
+    },
+    {
+      "epoch": 2.8587799456655967,
+      "grad_norm": 0.4551171658276694,
+      "learning_rate": 4.775892944934299e-06,
+      "loss": 0.5094,
+      "step": 7278
+    },
+    {
+      "epoch": 2.859175104964189,
+      "grad_norm": 0.45140156710288193,
+      "learning_rate": 4.77582810580722e-06,
+      "loss": 0.5102,
+      "step": 7279
+    },
+    {
+      "epoch": 2.859570264262781,
+      "grad_norm": 0.45003747644498915,
+      "learning_rate": 4.7757632577420696e-06,
+      "loss": 0.4983,
+      "step": 7280
+    },
+    {
+      "epoch": 2.8599654235613734,
+      "grad_norm": 0.45665797218831644,
+      "learning_rate": 4.775698400739104e-06,
+      "loss": 0.5201,
+      "step": 7281
+    },
+    {
+      "epoch": 2.8603605828599656,
+      "grad_norm": 0.4621874107147548,
+      "learning_rate": 4.775633534798576e-06,
+      "loss": 0.5044,
+      "step": 7282
+    },
+    {
+      "epoch": 2.860755742158558,
+      "grad_norm": 0.45736986361350984,
+      "learning_rate": 4.775568659920742e-06,
+      "loss": 0.4892,
+      "step": 7283
+    },
+    {
+      "epoch": 2.86115090145715,
+      "grad_norm": 0.44664628991297056,
+      "learning_rate": 4.775503776105857e-06,
+      "loss": 0.5025,
+      "step": 7284
+    },
+    {
+      "epoch": 2.8615460607557424,
+      "grad_norm": 0.44279821627241905,
+      "learning_rate": 4.775438883354173e-06,
+      "loss": 0.5267,
+      "step": 7285
+    },
+    {
+      "epoch": 2.8619412200543346,
+      "grad_norm": 0.4555964442573869,
+      "learning_rate": 4.775373981665949e-06,
+      "loss": 0.5362,
+      "step": 7286
+    },
+    {
+      "epoch": 2.862336379352927,
+      "grad_norm": 0.4616612464408454,
+      "learning_rate": 4.775309071041435e-06,
+      "loss": 0.5024,
+      "step": 7287
+    },
+    {
+      "epoch": 2.862731538651519,
+      "grad_norm": 0.44173733289424433,
+      "learning_rate": 4.7752441514808905e-06,
+      "loss": 0.4976,
+      "step": 7288
+    },
+    {
+      "epoch": 2.8631266979501113,
+      "grad_norm": 0.44729707048188977,
+      "learning_rate": 4.775179222984568e-06,
+      "loss": 0.5,
+      "step": 7289
+    },
+    {
+      "epoch": 2.8635218572487036,
+      "grad_norm": 0.45155830390167095,
+      "learning_rate": 4.775114285552723e-06,
+      "loss": 0.5023,
+      "step": 7290
+    },
+    {
+      "epoch": 2.863917016547296,
+      "grad_norm": 0.4777436940373346,
+      "learning_rate": 4.7750493391856116e-06,
+      "loss": 0.5002,
+      "step": 7291
+    },
+    {
+      "epoch": 2.864312175845888,
+      "grad_norm": 0.45674967310156706,
+      "learning_rate": 4.7749843838834865e-06,
+      "loss": 0.5122,
+      "step": 7292
+    },
+    {
+      "epoch": 2.8647073351444803,
+      "grad_norm": 0.4399469754211526,
+      "learning_rate": 4.774919419646605e-06,
+      "loss": 0.5023,
+      "step": 7293
+    },
+    {
+      "epoch": 2.8651024944430725,
+      "grad_norm": 0.4405627072122233,
+      "learning_rate": 4.774854446475221e-06,
+      "loss": 0.4848,
+      "step": 7294
+    },
+    {
+      "epoch": 2.8654976537416648,
+      "grad_norm": 0.4654307974697821,
+      "learning_rate": 4.7747894643695904e-06,
+      "loss": 0.5148,
+      "step": 7295
+    },
+    {
+      "epoch": 2.865892813040257,
+      "grad_norm": 0.4502154850100158,
+      "learning_rate": 4.774724473329968e-06,
+      "loss": 0.4916,
+      "step": 7296
+    },
+    {
+      "epoch": 2.8662879723388492,
+      "grad_norm": 0.45464786407421726,
+      "learning_rate": 4.7746594733566085e-06,
+      "loss": 0.5084,
+      "step": 7297
+    },
+    {
+      "epoch": 2.8666831316374415,
+      "grad_norm": 0.46932700358577506,
+      "learning_rate": 4.774594464449769e-06,
+      "loss": 0.521,
+      "step": 7298
+    },
+    {
+      "epoch": 2.8670782909360337,
+      "grad_norm": 0.455144441625113,
+      "learning_rate": 4.774529446609703e-06,
+      "loss": 0.5042,
+      "step": 7299
+    },
+    {
+      "epoch": 2.867473450234626,
+      "grad_norm": 0.456638540958973,
+      "learning_rate": 4.7744644198366665e-06,
+      "loss": 0.5098,
+      "step": 7300
+    },
+    {
+      "epoch": 2.867868609533218,
+      "grad_norm": 0.4731909071584723,
+      "learning_rate": 4.774399384130916e-06,
+      "loss": 0.4896,
+      "step": 7301
+    },
+    {
+      "epoch": 2.8682637688318104,
+      "grad_norm": 0.45219579881485555,
+      "learning_rate": 4.774334339492704e-06,
+      "loss": 0.4995,
+      "step": 7302
+    },
+    {
+      "epoch": 2.8686589281304027,
+      "grad_norm": 0.47541582012310035,
+      "learning_rate": 4.774269285922289e-06,
+      "loss": 0.5225,
+      "step": 7303
+    },
+    {
+      "epoch": 2.869054087428995,
+      "grad_norm": 0.4562665607670454,
+      "learning_rate": 4.774204223419925e-06,
+      "loss": 0.4862,
+      "step": 7304
+    },
+    {
+      "epoch": 2.869449246727587,
+      "grad_norm": 0.47218102712930754,
+      "learning_rate": 4.774139151985867e-06,
+      "loss": 0.5075,
+      "step": 7305
+    },
+    {
+      "epoch": 2.8698444060261794,
+      "grad_norm": 0.4403127626196469,
+      "learning_rate": 4.774074071620372e-06,
+      "loss": 0.5105,
+      "step": 7306
+    },
+    {
+      "epoch": 2.8702395653247716,
+      "grad_norm": 0.45708262550702466,
+      "learning_rate": 4.7740089823236955e-06,
+      "loss": 0.4972,
+      "step": 7307
+    },
+    {
+      "epoch": 2.870634724623364,
+      "grad_norm": 0.46018852248502795,
+      "learning_rate": 4.773943884096091e-06,
+      "loss": 0.4945,
+      "step": 7308
+    },
+    {
+      "epoch": 2.871029883921956,
+      "grad_norm": 0.4601257085296602,
+      "learning_rate": 4.773878776937817e-06,
+      "loss": 0.5051,
+      "step": 7309
+    },
+    {
+      "epoch": 2.8714250432205484,
+      "grad_norm": 0.45920489045937124,
+      "learning_rate": 4.7738136608491284e-06,
+      "loss": 0.505,
+      "step": 7310
+    },
+    {
+      "epoch": 2.8718202025191406,
+      "grad_norm": 0.45371392464155347,
+      "learning_rate": 4.77374853583028e-06,
+      "loss": 0.4837,
+      "step": 7311
+    },
+    {
+      "epoch": 2.872215361817733,
+      "grad_norm": 0.45429850072636996,
+      "learning_rate": 4.773683401881527e-06,
+      "loss": 0.4857,
+      "step": 7312
+    },
+    {
+      "epoch": 2.872610521116325,
+      "grad_norm": 0.4680648093877002,
+      "learning_rate": 4.773618259003127e-06,
+      "loss": 0.5074,
+      "step": 7313
+    },
+    {
+      "epoch": 2.8730056804149173,
+      "grad_norm": 0.461576978903435,
+      "learning_rate": 4.773553107195336e-06,
+      "loss": 0.513,
+      "step": 7314
+    },
+    {
+      "epoch": 2.8734008397135096,
+      "grad_norm": 0.5184320923986061,
+      "learning_rate": 4.773487946458407e-06,
+      "loss": 0.4973,
+      "step": 7315
+    },
+    {
+      "epoch": 2.873795999012102,
+      "grad_norm": 0.43754112690923064,
+      "learning_rate": 4.7734227767926e-06,
+      "loss": 0.4991,
+      "step": 7316
+    },
+    {
+      "epoch": 2.874191158310694,
+      "grad_norm": 0.4487036188195782,
+      "learning_rate": 4.773357598198167e-06,
+      "loss": 0.5125,
+      "step": 7317
+    },
+    {
+      "epoch": 2.8745863176092863,
+      "grad_norm": 0.4574040806296256,
+      "learning_rate": 4.773292410675366e-06,
+      "loss": 0.5261,
+      "step": 7318
+    },
+    {
+      "epoch": 2.8749814769078785,
+      "grad_norm": 0.4490619381032573,
+      "learning_rate": 4.773227214224454e-06,
+      "loss": 0.4986,
+      "step": 7319
+    },
+    {
+      "epoch": 2.8753766362064708,
+      "grad_norm": 0.47037528439860465,
+      "learning_rate": 4.773162008845685e-06,
+      "loss": 0.4982,
+      "step": 7320
+    },
+    {
+      "epoch": 2.875771795505063,
+      "grad_norm": 0.4506229178725294,
+      "learning_rate": 4.773096794539317e-06,
+      "loss": 0.4922,
+      "step": 7321
+    },
+    {
+      "epoch": 2.8761669548036553,
+      "grad_norm": 0.43652933566594937,
+      "learning_rate": 4.773031571305604e-06,
+      "loss": 0.5004,
+      "step": 7322
+    },
+    {
+      "epoch": 2.8765621141022475,
+      "grad_norm": 0.5540614478567485,
+      "learning_rate": 4.7729663391448035e-06,
+      "loss": 0.5186,
+      "step": 7323
+    },
+    {
+      "epoch": 2.8769572734008397,
+      "grad_norm": 0.46194230756962723,
+      "learning_rate": 4.772901098057172e-06,
+      "loss": 0.5071,
+      "step": 7324
+    },
+    {
+      "epoch": 2.877352432699432,
+      "grad_norm": 0.465341882765266,
+      "learning_rate": 4.772835848042965e-06,
+      "loss": 0.5202,
+      "step": 7325
+    },
+    {
+      "epoch": 2.877747591998024,
+      "grad_norm": 0.4778559900906391,
+      "learning_rate": 4.772770589102438e-06,
+      "loss": 0.4969,
+      "step": 7326
+    },
+    {
+      "epoch": 2.8781427512966165,
+      "grad_norm": 0.4405625019192915,
+      "learning_rate": 4.772705321235849e-06,
+      "loss": 0.4917,
+      "step": 7327
+    },
+    {
+      "epoch": 2.8785379105952087,
+      "grad_norm": 0.4506373137457211,
+      "learning_rate": 4.772640044443454e-06,
+      "loss": 0.507,
+      "step": 7328
+    },
+    {
+      "epoch": 2.878933069893801,
+      "grad_norm": 0.46863725137476087,
+      "learning_rate": 4.772574758725507e-06,
+      "loss": 0.4913,
+      "step": 7329
+    },
+    {
+      "epoch": 2.879328229192393,
+      "grad_norm": 0.4987365734752338,
+      "learning_rate": 4.772509464082269e-06,
+      "loss": 0.5125,
+      "step": 7330
+    },
+    {
+      "epoch": 2.8797233884909854,
+      "grad_norm": 0.45031884644226655,
+      "learning_rate": 4.772444160513992e-06,
+      "loss": 0.497,
+      "step": 7331
+    },
+    {
+      "epoch": 2.8801185477895777,
+      "grad_norm": 0.453093686573893,
+      "learning_rate": 4.772378848020935e-06,
+      "loss": 0.5252,
+      "step": 7332
+    },
+    {
+      "epoch": 2.88051370708817,
+      "grad_norm": 0.46311837414420265,
+      "learning_rate": 4.772313526603354e-06,
+      "loss": 0.5245,
+      "step": 7333
+    },
+    {
+      "epoch": 2.880908866386762,
+      "grad_norm": 0.5599787194673955,
+      "learning_rate": 4.772248196261504e-06,
+      "loss": 0.492,
+      "step": 7334
+    },
+    {
+      "epoch": 2.8813040256853544,
+      "grad_norm": 0.47600078027671755,
+      "learning_rate": 4.7721828569956435e-06,
+      "loss": 0.5181,
+      "step": 7335
+    },
+    {
+      "epoch": 2.8816991849839466,
+      "grad_norm": 0.4672363357538037,
+      "learning_rate": 4.772117508806029e-06,
+      "loss": 0.5156,
+      "step": 7336
+    },
+    {
+      "epoch": 2.882094344282539,
+      "grad_norm": 0.44982543608975994,
+      "learning_rate": 4.7720521516929155e-06,
+      "loss": 0.5017,
+      "step": 7337
+    },
+    {
+      "epoch": 2.882489503581131,
+      "grad_norm": 0.47014208289943477,
+      "learning_rate": 4.7719867856565615e-06,
+      "loss": 0.5021,
+      "step": 7338
+    },
+    {
+      "epoch": 2.8828846628797233,
+      "grad_norm": 0.4641826462044799,
+      "learning_rate": 4.771921410697224e-06,
+      "loss": 0.51,
+      "step": 7339
+    },
+    {
+      "epoch": 2.8832798221783156,
+      "grad_norm": 0.4480770317445626,
+      "learning_rate": 4.771856026815157e-06,
+      "loss": 0.4978,
+      "step": 7340
+    },
+    {
+      "epoch": 2.883674981476908,
+      "grad_norm": 0.4764205030319496,
+      "learning_rate": 4.77179063401062e-06,
+      "loss": 0.5137,
+      "step": 7341
+    },
+    {
+      "epoch": 2.8840701407755,
+      "grad_norm": 0.44175592798798174,
+      "learning_rate": 4.771725232283869e-06,
+      "loss": 0.5218,
+      "step": 7342
+    },
+    {
+      "epoch": 2.8844653000740923,
+      "grad_norm": 0.4522738770160621,
+      "learning_rate": 4.771659821635161e-06,
+      "loss": 0.5073,
+      "step": 7343
+    },
+    {
+      "epoch": 2.8848604593726845,
+      "grad_norm": 0.45063177134700066,
+      "learning_rate": 4.771594402064752e-06,
+      "loss": 0.5103,
+      "step": 7344
+    },
+    {
+      "epoch": 2.885255618671277,
+      "grad_norm": 0.4519278901335532,
+      "learning_rate": 4.7715289735729e-06,
+      "loss": 0.508,
+      "step": 7345
+    },
+    {
+      "epoch": 2.885650777969869,
+      "grad_norm": 0.45610027675210985,
+      "learning_rate": 4.771463536159861e-06,
+      "loss": 0.4991,
+      "step": 7346
+    },
+    {
+      "epoch": 2.8860459372684613,
+      "grad_norm": 0.4657657398587538,
+      "learning_rate": 4.771398089825893e-06,
+      "loss": 0.4981,
+      "step": 7347
+    },
+    {
+      "epoch": 2.8864410965670535,
+      "grad_norm": 0.4504836071125689,
+      "learning_rate": 4.771332634571252e-06,
+      "loss": 0.5194,
+      "step": 7348
+    },
+    {
+      "epoch": 2.8868362558656457,
+      "grad_norm": 0.4482132370137858,
+      "learning_rate": 4.771267170396197e-06,
+      "loss": 0.5019,
+      "step": 7349
+    },
+    {
+      "epoch": 2.887231415164238,
+      "grad_norm": 0.44368133807741067,
+      "learning_rate": 4.771201697300982e-06,
+      "loss": 0.5046,
+      "step": 7350
+    },
+    {
+      "epoch": 2.88762657446283,
+      "grad_norm": 0.4506387198321668,
+      "learning_rate": 4.7711362152858665e-06,
+      "loss": 0.5078,
+      "step": 7351
+    },
+    {
+      "epoch": 2.8880217337614225,
+      "grad_norm": 0.4570475595802833,
+      "learning_rate": 4.771070724351108e-06,
+      "loss": 0.5087,
+      "step": 7352
+    },
+    {
+      "epoch": 2.8884168930600147,
+      "grad_norm": 0.45887991966607716,
+      "learning_rate": 4.771005224496962e-06,
+      "loss": 0.5119,
+      "step": 7353
+    },
+    {
+      "epoch": 2.888812052358607,
+      "grad_norm": 0.4556415441058092,
+      "learning_rate": 4.770939715723686e-06,
+      "loss": 0.5164,
+      "step": 7354
+    },
+    {
+      "epoch": 2.889207211657199,
+      "grad_norm": 0.44832761334800575,
+      "learning_rate": 4.7708741980315386e-06,
+      "loss": 0.4931,
+      "step": 7355
+    },
+    {
+      "epoch": 2.8896023709557914,
+      "grad_norm": 0.47312914667107786,
+      "learning_rate": 4.770808671420775e-06,
+      "loss": 0.5121,
+      "step": 7356
+    },
+    {
+      "epoch": 2.8899975302543837,
+      "grad_norm": 0.6480164808981641,
+      "learning_rate": 4.770743135891656e-06,
+      "loss": 0.5064,
+      "step": 7357
+    },
+    {
+      "epoch": 2.890392689552976,
+      "grad_norm": 0.4502180251195585,
+      "learning_rate": 4.770677591444434e-06,
+      "loss": 0.5109,
+      "step": 7358
+    },
+    {
+      "epoch": 2.890787848851568,
+      "grad_norm": 0.44215136184674697,
+      "learning_rate": 4.770612038079372e-06,
+      "loss": 0.5187,
+      "step": 7359
+    },
+    {
+      "epoch": 2.8911830081501604,
+      "grad_norm": 0.46977522141656114,
+      "learning_rate": 4.770546475796724e-06,
+      "loss": 0.5058,
+      "step": 7360
+    },
+    {
+      "epoch": 2.8915781674487526,
+      "grad_norm": 0.44217645544646506,
+      "learning_rate": 4.770480904596747e-06,
+      "loss": 0.4908,
+      "step": 7361
+    },
+    {
+      "epoch": 2.891973326747345,
+      "grad_norm": 0.45598608295802123,
+      "learning_rate": 4.770415324479701e-06,
+      "loss": 0.5157,
+      "step": 7362
+    },
+    {
+      "epoch": 2.892368486045937,
+      "grad_norm": 0.4412753609610493,
+      "learning_rate": 4.770349735445841e-06,
+      "loss": 0.4907,
+      "step": 7363
+    },
+    {
+      "epoch": 2.8927636453445293,
+      "grad_norm": 0.4688090294235565,
+      "learning_rate": 4.770284137495428e-06,
+      "loss": 0.5032,
+      "step": 7364
+    },
+    {
+      "epoch": 2.8931588046431216,
+      "grad_norm": 0.46203038169470984,
+      "learning_rate": 4.770218530628716e-06,
+      "loss": 0.5207,
+      "step": 7365
+    },
+    {
+      "epoch": 2.893553963941714,
+      "grad_norm": 0.45196862440782837,
+      "learning_rate": 4.770152914845964e-06,
+      "loss": 0.498,
+      "step": 7366
+    },
+    {
+      "epoch": 2.893949123240306,
+      "grad_norm": 0.45872640459487846,
+      "learning_rate": 4.77008729014743e-06,
+      "loss": 0.5073,
+      "step": 7367
+    },
+    {
+      "epoch": 2.8943442825388983,
+      "grad_norm": 0.4534048366751859,
+      "learning_rate": 4.770021656533372e-06,
+      "loss": 0.4985,
+      "step": 7368
+    },
+    {
+      "epoch": 2.8947394418374905,
+      "grad_norm": 0.443264773459016,
+      "learning_rate": 4.769956014004047e-06,
+      "loss": 0.5142,
+      "step": 7369
+    },
+    {
+      "epoch": 2.895134601136083,
+      "grad_norm": 0.4379839420016731,
+      "learning_rate": 4.769890362559714e-06,
+      "loss": 0.4825,
+      "step": 7370
+    },
+    {
+      "epoch": 2.895529760434675,
+      "grad_norm": 0.4481743034206296,
+      "learning_rate": 4.769824702200629e-06,
+      "loss": 0.5016,
+      "step": 7371
+    },
+    {
+      "epoch": 2.8959249197332673,
+      "grad_norm": 0.4451918949234217,
+      "learning_rate": 4.769759032927051e-06,
+      "loss": 0.4956,
+      "step": 7372
+    },
+    {
+      "epoch": 2.8963200790318595,
+      "grad_norm": 0.5979190982944923,
+      "learning_rate": 4.7696933547392375e-06,
+      "loss": 0.5125,
+      "step": 7373
+    },
+    {
+      "epoch": 2.8967152383304517,
+      "grad_norm": 0.4563147906713146,
+      "learning_rate": 4.769627667637448e-06,
+      "loss": 0.5096,
+      "step": 7374
+    },
+    {
+      "epoch": 2.897110397629044,
+      "grad_norm": 0.46718974725304174,
+      "learning_rate": 4.7695619716219384e-06,
+      "loss": 0.5175,
+      "step": 7375
+    },
+    {
+      "epoch": 2.8975055569276362,
+      "grad_norm": 0.44977361032355456,
+      "learning_rate": 4.7694962666929674e-06,
+      "loss": 0.4988,
+      "step": 7376
+    },
+    {
+      "epoch": 2.8979007162262285,
+      "grad_norm": 0.4770742328128973,
+      "learning_rate": 4.769430552850793e-06,
+      "loss": 0.5208,
+      "step": 7377
+    },
+    {
+      "epoch": 2.8982958755248207,
+      "grad_norm": 0.46699375367596374,
+      "learning_rate": 4.769364830095674e-06,
+      "loss": 0.5051,
+      "step": 7378
+    },
+    {
+      "epoch": 2.898691034823413,
+      "grad_norm": 0.4793154561864407,
+      "learning_rate": 4.769299098427868e-06,
+      "loss": 0.505,
+      "step": 7379
+    },
+    {
+      "epoch": 2.899086194122005,
+      "grad_norm": 0.45166395843198986,
+      "learning_rate": 4.769233357847633e-06,
+      "loss": 0.5283,
+      "step": 7380
+    },
+    {
+      "epoch": 2.8994813534205974,
+      "grad_norm": 0.47128699616743297,
+      "learning_rate": 4.769167608355227e-06,
+      "loss": 0.516,
+      "step": 7381
+    },
+    {
+      "epoch": 2.8998765127191897,
+      "grad_norm": 0.4582075932359091,
+      "learning_rate": 4.769101849950909e-06,
+      "loss": 0.5087,
+      "step": 7382
+    },
+    {
+      "epoch": 2.900271672017782,
+      "grad_norm": 0.4557041584886058,
+      "learning_rate": 4.7690360826349365e-06,
+      "loss": 0.5215,
+      "step": 7383
+    },
+    {
+      "epoch": 2.900666831316374,
+      "grad_norm": 0.45326360410121896,
+      "learning_rate": 4.768970306407569e-06,
+      "loss": 0.5068,
+      "step": 7384
+    },
+    {
+      "epoch": 2.9010619906149664,
+      "grad_norm": 0.44111797702117767,
+      "learning_rate": 4.7689045212690625e-06,
+      "loss": 0.4926,
+      "step": 7385
+    },
+    {
+      "epoch": 2.9014571499135586,
+      "grad_norm": 0.47593611801487395,
+      "learning_rate": 4.7688387272196775e-06,
+      "loss": 0.5163,
+      "step": 7386
+    },
+    {
+      "epoch": 2.901852309212151,
+      "grad_norm": 0.4473189561304879,
+      "learning_rate": 4.768772924259671e-06,
+      "loss": 0.5233,
+      "step": 7387
+    },
+    {
+      "epoch": 2.9022474685107436,
+      "grad_norm": 0.44995418334579357,
+      "learning_rate": 4.768707112389303e-06,
+      "loss": 0.5283,
+      "step": 7388
+    },
+    {
+      "epoch": 2.902642627809336,
+      "grad_norm": 0.45948641092246273,
+      "learning_rate": 4.768641291608831e-06,
+      "loss": 0.5166,
+      "step": 7389
+    },
+    {
+      "epoch": 2.903037787107928,
+      "grad_norm": 0.4697983204082374,
+      "learning_rate": 4.768575461918513e-06,
+      "loss": 0.5185,
+      "step": 7390
+    },
+    {
+      "epoch": 2.9034329464065203,
+      "grad_norm": 0.4654187783348642,
+      "learning_rate": 4.768509623318609e-06,
+      "loss": 0.5268,
+      "step": 7391
+    },
+    {
+      "epoch": 2.9038281057051125,
+      "grad_norm": 0.4558337696980285,
+      "learning_rate": 4.768443775809376e-06,
+      "loss": 0.5024,
+      "step": 7392
+    },
+    {
+      "epoch": 2.9042232650037048,
+      "grad_norm": 0.4606281586552123,
+      "learning_rate": 4.768377919391074e-06,
+      "loss": 0.514,
+      "step": 7393
+    },
+    {
+      "epoch": 2.904618424302297,
+      "grad_norm": 0.4613832493128262,
+      "learning_rate": 4.768312054063961e-06,
+      "loss": 0.5055,
+      "step": 7394
+    },
+    {
+      "epoch": 2.9050135836008892,
+      "grad_norm": 0.455634197555287,
+      "learning_rate": 4.768246179828295e-06,
+      "loss": 0.5112,
+      "step": 7395
+    },
+    {
+      "epoch": 2.9054087428994815,
+      "grad_norm": 0.45367728348365555,
+      "learning_rate": 4.768180296684335e-06,
+      "loss": 0.5233,
+      "step": 7396
+    },
+    {
+      "epoch": 2.9058039021980737,
+      "grad_norm": 0.4571715325974132,
+      "learning_rate": 4.768114404632341e-06,
+      "loss": 0.5039,
+      "step": 7397
+    },
+    {
+      "epoch": 2.906199061496666,
+      "grad_norm": 0.45459414076829735,
+      "learning_rate": 4.768048503672571e-06,
+      "loss": 0.4977,
+      "step": 7398
+    },
+    {
+      "epoch": 2.906594220795258,
+      "grad_norm": 0.476609792116574,
+      "learning_rate": 4.7679825938052825e-06,
+      "loss": 0.4998,
+      "step": 7399
+    },
+    {
+      "epoch": 2.9069893800938504,
+      "grad_norm": 0.44935073335968,
+      "learning_rate": 4.7679166750307364e-06,
+      "loss": 0.5068,
+      "step": 7400
+    },
+    {
+      "epoch": 2.9073845393924427,
+      "grad_norm": 0.44816683458804896,
+      "learning_rate": 4.767850747349191e-06,
+      "loss": 0.5195,
+      "step": 7401
+    },
+    {
+      "epoch": 2.907779698691035,
+      "grad_norm": 0.514040565789535,
+      "learning_rate": 4.767784810760905e-06,
+      "loss": 0.4983,
+      "step": 7402
+    },
+    {
+      "epoch": 2.908174857989627,
+      "grad_norm": 0.4678938516683875,
+      "learning_rate": 4.767718865266136e-06,
+      "loss": 0.5241,
+      "step": 7403
+    },
+    {
+      "epoch": 2.9085700172882194,
+      "grad_norm": 0.44463113513631825,
+      "learning_rate": 4.767652910865146e-06,
+      "loss": 0.4928,
+      "step": 7404
+    },
+    {
+      "epoch": 2.9089651765868116,
+      "grad_norm": 0.44305031644764015,
+      "learning_rate": 4.767586947558191e-06,
+      "loss": 0.5274,
+      "step": 7405
+    },
+    {
+      "epoch": 2.909360335885404,
+      "grad_norm": 0.4608795710666698,
+      "learning_rate": 4.767520975345533e-06,
+      "loss": 0.496,
+      "step": 7406
+    },
+    {
+      "epoch": 2.909755495183996,
+      "grad_norm": 0.45454535548418995,
+      "learning_rate": 4.767454994227428e-06,
+      "loss": 0.5294,
+      "step": 7407
+    },
+    {
+      "epoch": 2.9101506544825884,
+      "grad_norm": 0.4438403467716369,
+      "learning_rate": 4.767389004204137e-06,
+      "loss": 0.5058,
+      "step": 7408
+    },
+    {
+      "epoch": 2.9105458137811806,
+      "grad_norm": 0.45975886229728474,
+      "learning_rate": 4.76732300527592e-06,
+      "loss": 0.4937,
+      "step": 7409
+    },
+    {
+      "epoch": 2.910940973079773,
+      "grad_norm": 0.4679652680374663,
+      "learning_rate": 4.767256997443034e-06,
+      "loss": 0.5338,
+      "step": 7410
+    },
+    {
+      "epoch": 2.911336132378365,
+      "grad_norm": 0.4443491416640828,
+      "learning_rate": 4.767190980705739e-06,
+      "loss": 0.5009,
+      "step": 7411
+    },
+    {
+      "epoch": 2.9117312916769573,
+      "grad_norm": 0.4466231335445614,
+      "learning_rate": 4.767124955064295e-06,
+      "loss": 0.5008,
+      "step": 7412
+    },
+    {
+      "epoch": 2.9121264509755496,
+      "grad_norm": 0.45416773779205705,
+      "learning_rate": 4.767058920518961e-06,
+      "loss": 0.5097,
+      "step": 7413
+    },
+    {
+      "epoch": 2.912521610274142,
+      "grad_norm": 0.4407003267094362,
+      "learning_rate": 4.766992877069996e-06,
+      "loss": 0.5013,
+      "step": 7414
+    },
+    {
+      "epoch": 2.912916769572734,
+      "grad_norm": 0.4526490214558496,
+      "learning_rate": 4.76692682471766e-06,
+      "loss": 0.5065,
+      "step": 7415
+    },
+    {
+      "epoch": 2.9133119288713263,
+      "grad_norm": 0.4384526778991886,
+      "learning_rate": 4.766860763462211e-06,
+      "loss": 0.5202,
+      "step": 7416
+    },
+    {
+      "epoch": 2.9137070881699185,
+      "grad_norm": 0.4554119386883933,
+      "learning_rate": 4.76679469330391e-06,
+      "loss": 0.4985,
+      "step": 7417
+    },
+    {
+      "epoch": 2.9141022474685108,
+      "grad_norm": 0.4364078258328405,
+      "learning_rate": 4.766728614243016e-06,
+      "loss": 0.4893,
+      "step": 7418
+    },
+    {
+      "epoch": 2.914497406767103,
+      "grad_norm": 0.4588538320640383,
+      "learning_rate": 4.766662526279788e-06,
+      "loss": 0.521,
+      "step": 7419
+    },
+    {
+      "epoch": 2.9148925660656952,
+      "grad_norm": 0.45670544657223855,
+      "learning_rate": 4.766596429414487e-06,
+      "loss": 0.5111,
+      "step": 7420
+    },
+    {
+      "epoch": 2.9152877253642875,
+      "grad_norm": 0.4417258587433614,
+      "learning_rate": 4.76653032364737e-06,
+      "loss": 0.5018,
+      "step": 7421
+    },
+    {
+      "epoch": 2.9156828846628797,
+      "grad_norm": 0.4365589726517331,
+      "learning_rate": 4.7664642089787e-06,
+      "loss": 0.5026,
+      "step": 7422
+    },
+    {
+      "epoch": 2.916078043961472,
+      "grad_norm": 0.4556876909416063,
+      "learning_rate": 4.766398085408734e-06,
+      "loss": 0.5208,
+      "step": 7423
+    },
+    {
+      "epoch": 2.916473203260064,
+      "grad_norm": 0.4496719110515717,
+      "learning_rate": 4.766331952937732e-06,
+      "loss": 0.5158,
+      "step": 7424
+    },
+    {
+      "epoch": 2.9168683625586564,
+      "grad_norm": 0.45674338189085323,
+      "learning_rate": 4.7662658115659546e-06,
+      "loss": 0.5155,
+      "step": 7425
+    },
+    {
+      "epoch": 2.9172635218572487,
+      "grad_norm": 0.4365519880167459,
+      "learning_rate": 4.766199661293662e-06,
+      "loss": 0.5067,
+      "step": 7426
+    },
+    {
+      "epoch": 2.917658681155841,
+      "grad_norm": 0.46044332574529667,
+      "learning_rate": 4.766133502121113e-06,
+      "loss": 0.509,
+      "step": 7427
+    },
+    {
+      "epoch": 2.918053840454433,
+      "grad_norm": 0.4523808596214587,
+      "learning_rate": 4.766067334048567e-06,
+      "loss": 0.5011,
+      "step": 7428
+    },
+    {
+      "epoch": 2.9184489997530254,
+      "grad_norm": 0.4552082949507314,
+      "learning_rate": 4.766001157076284e-06,
+      "loss": 0.515,
+      "step": 7429
+    },
+    {
+      "epoch": 2.9188441590516176,
+      "grad_norm": 0.45129094467367264,
+      "learning_rate": 4.765934971204526e-06,
+      "loss": 0.5026,
+      "step": 7430
+    },
+    {
+      "epoch": 2.91923931835021,
+      "grad_norm": 0.4615128953141365,
+      "learning_rate": 4.765868776433551e-06,
+      "loss": 0.5293,
+      "step": 7431
+    },
+    {
+      "epoch": 2.919634477648802,
+      "grad_norm": 0.44445080798757497,
+      "learning_rate": 4.765802572763619e-06,
+      "loss": 0.5084,
+      "step": 7432
+    },
+    {
+      "epoch": 2.9200296369473944,
+      "grad_norm": 0.5535941391787863,
+      "learning_rate": 4.76573636019499e-06,
+      "loss": 0.5165,
+      "step": 7433
+    },
+    {
+      "epoch": 2.9204247962459866,
+      "grad_norm": 0.4647127083547014,
+      "learning_rate": 4.765670138727925e-06,
+      "loss": 0.5072,
+      "step": 7434
+    },
+    {
+      "epoch": 2.920819955544579,
+      "grad_norm": 0.44196536115349694,
+      "learning_rate": 4.765603908362683e-06,
+      "loss": 0.5036,
+      "step": 7435
+    },
+    {
+      "epoch": 2.921215114843171,
+      "grad_norm": 0.45968440262090315,
+      "learning_rate": 4.765537669099525e-06,
+      "loss": 0.4972,
+      "step": 7436
+    },
+    {
+      "epoch": 2.9216102741417633,
+      "grad_norm": 0.444196485819374,
+      "learning_rate": 4.765471420938711e-06,
+      "loss": 0.5025,
+      "step": 7437
+    },
+    {
+      "epoch": 2.9220054334403556,
+      "grad_norm": 0.4480022299436206,
+      "learning_rate": 4.7654051638805e-06,
+      "loss": 0.4847,
+      "step": 7438
+    },
+    {
+      "epoch": 2.922400592738948,
+      "grad_norm": 0.4567439332123647,
+      "learning_rate": 4.765338897925154e-06,
+      "loss": 0.5128,
+      "step": 7439
+    },
+    {
+      "epoch": 2.92279575203754,
+      "grad_norm": 0.458511120503876,
+      "learning_rate": 4.765272623072932e-06,
+      "loss": 0.5166,
+      "step": 7440
+    },
+    {
+      "epoch": 2.9231909113361323,
+      "grad_norm": 0.44860217503885985,
+      "learning_rate": 4.765206339324095e-06,
+      "loss": 0.5045,
+      "step": 7441
+    },
+    {
+      "epoch": 2.9235860706347245,
+      "grad_norm": 0.4546548390510289,
+      "learning_rate": 4.765140046678903e-06,
+      "loss": 0.4965,
+      "step": 7442
+    },
+    {
+      "epoch": 2.9239812299333168,
+      "grad_norm": 0.4492830754296067,
+      "learning_rate": 4.765073745137616e-06,
+      "loss": 0.5065,
+      "step": 7443
+    },
+    {
+      "epoch": 2.924376389231909,
+      "grad_norm": 0.4605719668288093,
+      "learning_rate": 4.765007434700495e-06,
+      "loss": 0.5078,
+      "step": 7444
+    },
+    {
+      "epoch": 2.9247715485305013,
+      "grad_norm": 0.4640714655281598,
+      "learning_rate": 4.7649411153678e-06,
+      "loss": 0.523,
+      "step": 7445
+    },
+    {
+      "epoch": 2.9251667078290935,
+      "grad_norm": 0.45659739128055815,
+      "learning_rate": 4.764874787139792e-06,
+      "loss": 0.4993,
+      "step": 7446
+    },
+    {
+      "epoch": 2.9255618671276857,
+      "grad_norm": 0.4728553705365856,
+      "learning_rate": 4.764808450016731e-06,
+      "loss": 0.5352,
+      "step": 7447
+    },
+    {
+      "epoch": 2.925957026426278,
+      "grad_norm": 0.43793250070587963,
+      "learning_rate": 4.764742103998877e-06,
+      "loss": 0.4935,
+      "step": 7448
+    },
+    {
+      "epoch": 2.92635218572487,
+      "grad_norm": 0.4423604050440384,
+      "learning_rate": 4.7646757490864926e-06,
+      "loss": 0.5088,
+      "step": 7449
+    },
+    {
+      "epoch": 2.9267473450234625,
+      "grad_norm": 0.4501666815149439,
+      "learning_rate": 4.764609385279836e-06,
+      "loss": 0.5102,
+      "step": 7450
+    },
+    {
+      "epoch": 2.9271425043220547,
+      "grad_norm": 0.4574235898623384,
+      "learning_rate": 4.764543012579169e-06,
+      "loss": 0.5032,
+      "step": 7451
+    },
+    {
+      "epoch": 2.927537663620647,
+      "grad_norm": 0.460475598434439,
+      "learning_rate": 4.764476630984752e-06,
+      "loss": 0.4958,
+      "step": 7452
+    },
+    {
+      "epoch": 2.927932822919239,
+      "grad_norm": 0.43089611921617016,
+      "learning_rate": 4.764410240496846e-06,
+      "loss": 0.4941,
+      "step": 7453
+    },
+    {
+      "epoch": 2.9283279822178314,
+      "grad_norm": 0.45244880455514785,
+      "learning_rate": 4.764343841115712e-06,
+      "loss": 0.5068,
+      "step": 7454
+    },
+    {
+      "epoch": 2.9287231415164237,
+      "grad_norm": 0.4572924724707516,
+      "learning_rate": 4.76427743284161e-06,
+      "loss": 0.509,
+      "step": 7455
+    },
+    {
+      "epoch": 2.929118300815016,
+      "grad_norm": 0.4535852769301378,
+      "learning_rate": 4.764211015674801e-06,
+      "loss": 0.5127,
+      "step": 7456
+    },
+    {
+      "epoch": 2.929513460113608,
+      "grad_norm": 0.5430024075970933,
+      "learning_rate": 4.764144589615547e-06,
+      "loss": 0.5065,
+      "step": 7457
+    },
+    {
+      "epoch": 2.929908619412201,
+      "grad_norm": 0.44556126301612464,
+      "learning_rate": 4.764078154664107e-06,
+      "loss": 0.5309,
+      "step": 7458
+    },
+    {
+      "epoch": 2.930303778710793,
+      "grad_norm": 0.4710840587154262,
+      "learning_rate": 4.764011710820743e-06,
+      "loss": 0.5104,
+      "step": 7459
+    },
+    {
+      "epoch": 2.9306989380093853,
+      "grad_norm": 0.45571321032211237,
+      "learning_rate": 4.763945258085716e-06,
+      "loss": 0.5092,
+      "step": 7460
+    },
+    {
+      "epoch": 2.9310940973079775,
+      "grad_norm": 0.44102051744598414,
+      "learning_rate": 4.763878796459287e-06,
+      "loss": 0.5168,
+      "step": 7461
+    },
+    {
+      "epoch": 2.93148925660657,
+      "grad_norm": 0.4470640541086543,
+      "learning_rate": 4.7638123259417166e-06,
+      "loss": 0.5013,
+      "step": 7462
+    },
+    {
+      "epoch": 2.931884415905162,
+      "grad_norm": 0.43554949182172414,
+      "learning_rate": 4.763745846533265e-06,
+      "loss": 0.4982,
+      "step": 7463
+    },
+    {
+      "epoch": 2.9322795752037543,
+      "grad_norm": 0.4720396479178599,
+      "learning_rate": 4.763679358234196e-06,
+      "loss": 0.5106,
+      "step": 7464
+    },
+    {
+      "epoch": 2.9326747345023465,
+      "grad_norm": 0.4602272736857191,
+      "learning_rate": 4.763612861044768e-06,
+      "loss": 0.5108,
+      "step": 7465
+    },
+    {
+      "epoch": 2.9330698938009387,
+      "grad_norm": 0.4552466120844723,
+      "learning_rate": 4.763546354965244e-06,
+      "loss": 0.5034,
+      "step": 7466
+    },
+    {
+      "epoch": 2.933465053099531,
+      "grad_norm": 0.4575835047170144,
+      "learning_rate": 4.763479839995883e-06,
+      "loss": 0.5096,
+      "step": 7467
+    },
+    {
+      "epoch": 2.9338602123981232,
+      "grad_norm": 0.4678916204593777,
+      "learning_rate": 4.763413316136949e-06,
+      "loss": 0.5274,
+      "step": 7468
+    },
+    {
+      "epoch": 2.9342553716967155,
+      "grad_norm": 0.4452559996447758,
+      "learning_rate": 4.7633467833887015e-06,
+      "loss": 0.4957,
+      "step": 7469
+    },
+    {
+      "epoch": 2.9346505309953077,
+      "grad_norm": 0.44208744364576574,
+      "learning_rate": 4.763280241751402e-06,
+      "loss": 0.495,
+      "step": 7470
+    },
+    {
+      "epoch": 2.9350456902939,
+      "grad_norm": 0.46246549564900696,
+      "learning_rate": 4.763213691225313e-06,
+      "loss": 0.5021,
+      "step": 7471
+    },
+    {
+      "epoch": 2.935440849592492,
+      "grad_norm": 0.44656255715995435,
+      "learning_rate": 4.763147131810693e-06,
+      "loss": 0.5128,
+      "step": 7472
+    },
+    {
+      "epoch": 2.9358360088910844,
+      "grad_norm": 0.4612831877698752,
+      "learning_rate": 4.7630805635078065e-06,
+      "loss": 0.5001,
+      "step": 7473
+    },
+    {
+      "epoch": 2.9362311681896767,
+      "grad_norm": 0.4956157029905282,
+      "learning_rate": 4.763013986316914e-06,
+      "loss": 0.5342,
+      "step": 7474
+    },
+    {
+      "epoch": 2.936626327488269,
+      "grad_norm": 0.4505671941107815,
+      "learning_rate": 4.762947400238276e-06,
+      "loss": 0.4923,
+      "step": 7475
+    },
+    {
+      "epoch": 2.937021486786861,
+      "grad_norm": 0.45390395120855775,
+      "learning_rate": 4.762880805272155e-06,
+      "loss": 0.5247,
+      "step": 7476
+    },
+    {
+      "epoch": 2.9374166460854534,
+      "grad_norm": 0.44602033162244886,
+      "learning_rate": 4.762814201418813e-06,
+      "loss": 0.4961,
+      "step": 7477
+    },
+    {
+      "epoch": 2.9378118053840456,
+      "grad_norm": 0.45061560756472163,
+      "learning_rate": 4.76274758867851e-06,
+      "loss": 0.5152,
+      "step": 7478
+    },
+    {
+      "epoch": 2.938206964682638,
+      "grad_norm": 0.4472180927534863,
+      "learning_rate": 4.762680967051509e-06,
+      "loss": 0.5113,
+      "step": 7479
+    },
+    {
+      "epoch": 2.93860212398123,
+      "grad_norm": 0.4582451501185086,
+      "learning_rate": 4.762614336538071e-06,
+      "loss": 0.5157,
+      "step": 7480
+    },
+    {
+      "epoch": 2.9389972832798223,
+      "grad_norm": 0.4383818777857067,
+      "learning_rate": 4.762547697138458e-06,
+      "loss": 0.4909,
+      "step": 7481
+    },
+    {
+      "epoch": 2.9393924425784146,
+      "grad_norm": 0.46044036455758336,
+      "learning_rate": 4.762481048852931e-06,
+      "loss": 0.5176,
+      "step": 7482
+    },
+    {
+      "epoch": 2.939787601877007,
+      "grad_norm": 0.4559104153811503,
+      "learning_rate": 4.762414391681753e-06,
+      "loss": 0.5117,
+      "step": 7483
+    },
+    {
+      "epoch": 2.940182761175599,
+      "grad_norm": 0.4345018162819212,
+      "learning_rate": 4.762347725625185e-06,
+      "loss": 0.4939,
+      "step": 7484
+    },
+    {
+      "epoch": 2.9405779204741913,
+      "grad_norm": 0.44329288256437194,
+      "learning_rate": 4.7622810506834885e-06,
+      "loss": 0.4959,
+      "step": 7485
+    },
+    {
+      "epoch": 2.9409730797727835,
+      "grad_norm": 0.4719182053165318,
+      "learning_rate": 4.762214366856925e-06,
+      "loss": 0.5199,
+      "step": 7486
+    },
+    {
+      "epoch": 2.941368239071376,
+      "grad_norm": 0.500449067523404,
+      "learning_rate": 4.762147674145759e-06,
+      "loss": 0.5036,
+      "step": 7487
+    },
+    {
+      "epoch": 2.941763398369968,
+      "grad_norm": 0.44559700179627093,
+      "learning_rate": 4.762080972550249e-06,
+      "loss": 0.5127,
+      "step": 7488
+    },
+    {
+      "epoch": 2.9421585576685603,
+      "grad_norm": 0.45054565116407974,
+      "learning_rate": 4.762014262070659e-06,
+      "loss": 0.5329,
+      "step": 7489
+    },
+    {
+      "epoch": 2.9425537169671525,
+      "grad_norm": 0.4530812436143374,
+      "learning_rate": 4.761947542707251e-06,
+      "loss": 0.4968,
+      "step": 7490
+    },
+    {
+      "epoch": 2.9429488762657448,
+      "grad_norm": 0.45123498001248574,
+      "learning_rate": 4.761880814460286e-06,
+      "loss": 0.5331,
+      "step": 7491
+    },
+    {
+      "epoch": 2.943344035564337,
+      "grad_norm": 0.46119671535259266,
+      "learning_rate": 4.761814077330027e-06,
+      "loss": 0.5299,
+      "step": 7492
+    },
+    {
+      "epoch": 2.9437391948629292,
+      "grad_norm": 0.4371101220621986,
+      "learning_rate": 4.7617473313167365e-06,
+      "loss": 0.4896,
+      "step": 7493
+    },
+    {
+      "epoch": 2.9441343541615215,
+      "grad_norm": 0.45650104982890227,
+      "learning_rate": 4.761680576420674e-06,
+      "loss": 0.5266,
+      "step": 7494
+    },
+    {
+      "epoch": 2.9445295134601137,
+      "grad_norm": 0.45022327882961727,
+      "learning_rate": 4.761613812642105e-06,
+      "loss": 0.5307,
+      "step": 7495
+    },
+    {
+      "epoch": 2.944924672758706,
+      "grad_norm": 0.45404058586100055,
+      "learning_rate": 4.76154703998129e-06,
+      "loss": 0.5091,
+      "step": 7496
+    },
+    {
+      "epoch": 2.945319832057298,
+      "grad_norm": 0.45618375047378157,
+      "learning_rate": 4.761480258438491e-06,
+      "loss": 0.5167,
+      "step": 7497
+    },
+    {
+      "epoch": 2.9457149913558904,
+      "grad_norm": 0.44164866519004375,
+      "learning_rate": 4.761413468013972e-06,
+      "loss": 0.4942,
+      "step": 7498
+    },
+    {
+      "epoch": 2.9461101506544827,
+      "grad_norm": 0.43860964167840893,
+      "learning_rate": 4.761346668707993e-06,
+      "loss": 0.5087,
+      "step": 7499
+    },
+    {
+      "epoch": 2.946505309953075,
+      "grad_norm": 0.4453511072236731,
+      "learning_rate": 4.7612798605208175e-06,
+      "loss": 0.5224,
+      "step": 7500
+    },
+    {
+      "epoch": 2.946900469251667,
+      "grad_norm": 0.44709440039596826,
+      "learning_rate": 4.761213043452708e-06,
+      "loss": 0.5303,
+      "step": 7501
+    },
+    {
+      "epoch": 2.9472956285502594,
+      "grad_norm": 0.454881212603844,
+      "learning_rate": 4.761146217503927e-06,
+      "loss": 0.523,
+      "step": 7502
+    },
+    {
+      "epoch": 2.9476907878488516,
+      "grad_norm": 0.43851122242734974,
+      "learning_rate": 4.761079382674737e-06,
+      "loss": 0.5026,
+      "step": 7503
+    },
+    {
+      "epoch": 2.948085947147444,
+      "grad_norm": 0.47892329084918284,
+      "learning_rate": 4.761012538965399e-06,
+      "loss": 0.5178,
+      "step": 7504
+    },
+    {
+      "epoch": 2.948481106446036,
+      "grad_norm": 0.4698801672891979,
+      "learning_rate": 4.760945686376178e-06,
+      "loss": 0.5098,
+      "step": 7505
+    },
+    {
+      "epoch": 2.9488762657446284,
+      "grad_norm": 0.48150219567951025,
+      "learning_rate": 4.760878824907335e-06,
+      "loss": 0.4947,
+      "step": 7506
+    },
+    {
+      "epoch": 2.9492714250432206,
+      "grad_norm": 0.4417389041761658,
+      "learning_rate": 4.7608119545591326e-06,
+      "loss": 0.5044,
+      "step": 7507
+    },
+    {
+      "epoch": 2.949666584341813,
+      "grad_norm": 0.44156128051674065,
+      "learning_rate": 4.760745075331833e-06,
+      "loss": 0.4993,
+      "step": 7508
+    },
+    {
+      "epoch": 2.950061743640405,
+      "grad_norm": 0.43631000583026247,
+      "learning_rate": 4.7606781872257e-06,
+      "loss": 0.4978,
+      "step": 7509
+    },
+    {
+      "epoch": 2.9504569029389973,
+      "grad_norm": 0.4570064542158102,
+      "learning_rate": 4.760611290240996e-06,
+      "loss": 0.5007,
+      "step": 7510
+    },
+    {
+      "epoch": 2.9508520622375896,
+      "grad_norm": 0.45615671034916505,
+      "learning_rate": 4.760544384377984e-06,
+      "loss": 0.4908,
+      "step": 7511
+    },
+    {
+      "epoch": 2.951247221536182,
+      "grad_norm": 0.4343490552276528,
+      "learning_rate": 4.760477469636926e-06,
+      "loss": 0.4896,
+      "step": 7512
+    },
+    {
+      "epoch": 2.951642380834774,
+      "grad_norm": 0.4459408199880296,
+      "learning_rate": 4.760410546018085e-06,
+      "loss": 0.5091,
+      "step": 7513
+    },
+    {
+      "epoch": 2.9520375401333663,
+      "grad_norm": 0.4636503598906805,
+      "learning_rate": 4.760343613521724e-06,
+      "loss": 0.4895,
+      "step": 7514
+    },
+    {
+      "epoch": 2.9524326994319585,
+      "grad_norm": 0.4473832087140672,
+      "learning_rate": 4.7602766721481055e-06,
+      "loss": 0.5059,
+      "step": 7515
+    },
+    {
+      "epoch": 2.9528278587305508,
+      "grad_norm": 0.45903466878021015,
+      "learning_rate": 4.760209721897493e-06,
+      "loss": 0.5259,
+      "step": 7516
+    },
+    {
+      "epoch": 2.953223018029143,
+      "grad_norm": 0.4389672467155343,
+      "learning_rate": 4.76014276277015e-06,
+      "loss": 0.5145,
+      "step": 7517
+    },
+    {
+      "epoch": 2.9536181773277352,
+      "grad_norm": 0.44946325074519894,
+      "learning_rate": 4.760075794766338e-06,
+      "loss": 0.5092,
+      "step": 7518
+    },
+    {
+      "epoch": 2.9540133366263275,
+      "grad_norm": 0.44012358398055595,
+      "learning_rate": 4.76000881788632e-06,
+      "loss": 0.4943,
+      "step": 7519
+    },
+    {
+      "epoch": 2.9544084959249197,
+      "grad_norm": 0.45946199113007924,
+      "learning_rate": 4.75994183213036e-06,
+      "loss": 0.5014,
+      "step": 7520
+    },
+    {
+      "epoch": 2.954803655223512,
+      "grad_norm": 0.45502527214353133,
+      "learning_rate": 4.759874837498721e-06,
+      "loss": 0.4964,
+      "step": 7521
+    },
+    {
+      "epoch": 2.955198814522104,
+      "grad_norm": 0.465012979637925,
+      "learning_rate": 4.759807833991667e-06,
+      "loss": 0.4897,
+      "step": 7522
+    },
+    {
+      "epoch": 2.9555939738206964,
+      "grad_norm": 0.4467396954982866,
+      "learning_rate": 4.759740821609459e-06,
+      "loss": 0.5003,
+      "step": 7523
+    },
+    {
+      "epoch": 2.9559891331192887,
+      "grad_norm": 0.4425638158476036,
+      "learning_rate": 4.759673800352362e-06,
+      "loss": 0.4933,
+      "step": 7524
+    },
+    {
+      "epoch": 2.956384292417881,
+      "grad_norm": 0.4616979621992214,
+      "learning_rate": 4.759606770220638e-06,
+      "loss": 0.5142,
+      "step": 7525
+    },
+    {
+      "epoch": 2.956779451716473,
+      "grad_norm": 0.4722218812872345,
+      "learning_rate": 4.759539731214549e-06,
+      "loss": 0.5087,
+      "step": 7526
+    },
+    {
+      "epoch": 2.9571746110150654,
+      "grad_norm": 0.439973697685221,
+      "learning_rate": 4.759472683334362e-06,
+      "loss": 0.4875,
+      "step": 7527
+    },
+    {
+      "epoch": 2.9575697703136576,
+      "grad_norm": 0.4571911157675649,
+      "learning_rate": 4.759405626580338e-06,
+      "loss": 0.513,
+      "step": 7528
+    },
+    {
+      "epoch": 2.95796492961225,
+      "grad_norm": 0.4657839715911801,
+      "learning_rate": 4.7593385609527406e-06,
+      "loss": 0.5174,
+      "step": 7529
+    },
+    {
+      "epoch": 2.958360088910842,
+      "grad_norm": 0.4530863288631989,
+      "learning_rate": 4.759271486451833e-06,
+      "loss": 0.4956,
+      "step": 7530
+    },
+    {
+      "epoch": 2.9587552482094344,
+      "grad_norm": 0.4722880955783831,
+      "learning_rate": 4.759204403077879e-06,
+      "loss": 0.5187,
+      "step": 7531
+    },
+    {
+      "epoch": 2.9591504075080266,
+      "grad_norm": 0.4567472933113083,
+      "learning_rate": 4.7591373108311425e-06,
+      "loss": 0.5272,
+      "step": 7532
+    },
+    {
+      "epoch": 2.959545566806619,
+      "grad_norm": 0.46118458955762015,
+      "learning_rate": 4.759070209711886e-06,
+      "loss": 0.515,
+      "step": 7533
+    },
+    {
+      "epoch": 2.959940726105211,
+      "grad_norm": 0.4337420420723696,
+      "learning_rate": 4.759003099720373e-06,
+      "loss": 0.4868,
+      "step": 7534
+    },
+    {
+      "epoch": 2.9603358854038033,
+      "grad_norm": 0.4597390934949171,
+      "learning_rate": 4.758935980856868e-06,
+      "loss": 0.524,
+      "step": 7535
+    },
+    {
+      "epoch": 2.9607310447023956,
+      "grad_norm": 0.4520311626789291,
+      "learning_rate": 4.758868853121635e-06,
+      "loss": 0.5164,
+      "step": 7536
+    },
+    {
+      "epoch": 2.961126204000988,
+      "grad_norm": 0.4604721101236667,
+      "learning_rate": 4.758801716514935e-06,
+      "loss": 0.5022,
+      "step": 7537
+    },
+    {
+      "epoch": 2.96152136329958,
+      "grad_norm": 0.4419481364317015,
+      "learning_rate": 4.758734571037035e-06,
+      "loss": 0.4752,
+      "step": 7538
+    },
+    {
+      "epoch": 2.9619165225981723,
+      "grad_norm": 0.4449236247292902,
+      "learning_rate": 4.758667416688197e-06,
+      "loss": 0.5091,
+      "step": 7539
+    },
+    {
+      "epoch": 2.9623116818967645,
+      "grad_norm": 0.4553128748466083,
+      "learning_rate": 4.758600253468684e-06,
+      "loss": 0.5,
+      "step": 7540
+    },
+    {
+      "epoch": 2.9627068411953568,
+      "grad_norm": 0.46612998739450867,
+      "learning_rate": 4.758533081378762e-06,
+      "loss": 0.5018,
+      "step": 7541
+    },
+    {
+      "epoch": 2.963102000493949,
+      "grad_norm": 0.46551416452290106,
+      "learning_rate": 4.7584659004186924e-06,
+      "loss": 0.508,
+      "step": 7542
+    },
+    {
+      "epoch": 2.9634971597925412,
+      "grad_norm": 0.452825185453952,
+      "learning_rate": 4.758398710588741e-06,
+      "loss": 0.4992,
+      "step": 7543
+    },
+    {
+      "epoch": 2.9638923190911335,
+      "grad_norm": 0.4604901561144933,
+      "learning_rate": 4.758331511889171e-06,
+      "loss": 0.5071,
+      "step": 7544
+    },
+    {
+      "epoch": 2.9642874783897257,
+      "grad_norm": 0.4700510709171447,
+      "learning_rate": 4.7582643043202445e-06,
+      "loss": 0.5159,
+      "step": 7545
+    },
+    {
+      "epoch": 2.964682637688318,
+      "grad_norm": 0.4702968346285271,
+      "learning_rate": 4.758197087882228e-06,
+      "loss": 0.5211,
+      "step": 7546
+    },
+    {
+      "epoch": 2.96507779698691,
+      "grad_norm": 0.45518608070060007,
+      "learning_rate": 4.758129862575386e-06,
+      "loss": 0.4996,
+      "step": 7547
+    },
+    {
+      "epoch": 2.9654729562855024,
+      "grad_norm": 0.46985798766213305,
+      "learning_rate": 4.758062628399979e-06,
+      "loss": 0.5043,
+      "step": 7548
+    },
+    {
+      "epoch": 2.9658681155840947,
+      "grad_norm": 0.47941413887771617,
+      "learning_rate": 4.7579953853562744e-06,
+      "loss": 0.5225,
+      "step": 7549
+    },
+    {
+      "epoch": 2.966263274882687,
+      "grad_norm": 0.44370397690809943,
+      "learning_rate": 4.757928133444534e-06,
+      "loss": 0.508,
+      "step": 7550
+    },
+    {
+      "epoch": 2.966658434181279,
+      "grad_norm": 0.5281064344206505,
+      "learning_rate": 4.757860872665024e-06,
+      "loss": 0.5019,
+      "step": 7551
+    },
+    {
+      "epoch": 2.9670535934798714,
+      "grad_norm": 0.46395863123158015,
+      "learning_rate": 4.757793603018007e-06,
+      "loss": 0.5176,
+      "step": 7552
+    },
+    {
+      "epoch": 2.9674487527784636,
+      "grad_norm": 0.4640272614666952,
+      "learning_rate": 4.757726324503749e-06,
+      "loss": 0.5029,
+      "step": 7553
+    },
+    {
+      "epoch": 2.967843912077056,
+      "grad_norm": 0.443178090717011,
+      "learning_rate": 4.757659037122511e-06,
+      "loss": 0.5143,
+      "step": 7554
+    },
+    {
+      "epoch": 2.968239071375648,
+      "grad_norm": 0.4562032262581782,
+      "learning_rate": 4.75759174087456e-06,
+      "loss": 0.4894,
+      "step": 7555
+    },
+    {
+      "epoch": 2.9686342306742404,
+      "grad_norm": 0.4505861613127547,
+      "learning_rate": 4.75752443576016e-06,
+      "loss": 0.511,
+      "step": 7556
+    },
+    {
+      "epoch": 2.9690293899728326,
+      "grad_norm": 0.46040086407622155,
+      "learning_rate": 4.757457121779575e-06,
+      "loss": 0.5134,
+      "step": 7557
+    },
+    {
+      "epoch": 2.969424549271425,
+      "grad_norm": 0.46009994634444334,
+      "learning_rate": 4.757389798933069e-06,
+      "loss": 0.4881,
+      "step": 7558
+    },
+    {
+      "epoch": 2.969819708570017,
+      "grad_norm": 0.48161543752949787,
+      "learning_rate": 4.757322467220906e-06,
+      "loss": 0.4888,
+      "step": 7559
+    },
+    {
+      "epoch": 2.9702148678686093,
+      "grad_norm": 0.4468364867174091,
+      "learning_rate": 4.7572551266433506e-06,
+      "loss": 0.5166,
+      "step": 7560
+    },
+    {
+      "epoch": 2.9706100271672016,
+      "grad_norm": 0.451144658830222,
+      "learning_rate": 4.757187777200669e-06,
+      "loss": 0.497,
+      "step": 7561
+    },
+    {
+      "epoch": 2.971005186465794,
+      "grad_norm": 0.4435308430537957,
+      "learning_rate": 4.757120418893124e-06,
+      "loss": 0.5151,
+      "step": 7562
+    },
+    {
+      "epoch": 2.971400345764386,
+      "grad_norm": 0.4757327743330427,
+      "learning_rate": 4.7570530517209815e-06,
+      "loss": 0.5268,
+      "step": 7563
+    },
+    {
+      "epoch": 2.9717955050629783,
+      "grad_norm": 0.46010095557816083,
+      "learning_rate": 4.756985675684504e-06,
+      "loss": 0.5048,
+      "step": 7564
+    },
+    {
+      "epoch": 2.9721906643615705,
+      "grad_norm": 0.4457407014760488,
+      "learning_rate": 4.756918290783957e-06,
+      "loss": 0.5066,
+      "step": 7565
+    },
+    {
+      "epoch": 2.9725858236601628,
+      "grad_norm": 0.4797320041985171,
+      "learning_rate": 4.756850897019606e-06,
+      "loss": 0.5166,
+      "step": 7566
+    },
+    {
+      "epoch": 2.972980982958755,
+      "grad_norm": 0.44199710687789934,
+      "learning_rate": 4.756783494391716e-06,
+      "loss": 0.487,
+      "step": 7567
+    },
+    {
+      "epoch": 2.9733761422573473,
+      "grad_norm": 0.4426609628978589,
+      "learning_rate": 4.7567160829005496e-06,
+      "loss": 0.5036,
+      "step": 7568
+    },
+    {
+      "epoch": 2.9737713015559395,
+      "grad_norm": 0.4505400917951688,
+      "learning_rate": 4.756648662546373e-06,
+      "loss": 0.5125,
+      "step": 7569
+    },
+    {
+      "epoch": 2.9741664608545317,
+      "grad_norm": 0.4676898947705984,
+      "learning_rate": 4.756581233329451e-06,
+      "loss": 0.529,
+      "step": 7570
+    },
+    {
+      "epoch": 2.974561620153124,
+      "grad_norm": 0.44793604608650994,
+      "learning_rate": 4.756513795250048e-06,
+      "loss": 0.5314,
+      "step": 7571
+    },
+    {
+      "epoch": 2.974956779451716,
+      "grad_norm": 0.4257182384070774,
+      "learning_rate": 4.756446348308429e-06,
+      "loss": 0.4938,
+      "step": 7572
+    },
+    {
+      "epoch": 2.9753519387503085,
+      "grad_norm": 0.4541639462382242,
+      "learning_rate": 4.7563788925048596e-06,
+      "loss": 0.5049,
+      "step": 7573
+    },
+    {
+      "epoch": 2.9757470980489007,
+      "grad_norm": 0.45049135883048075,
+      "learning_rate": 4.7563114278396025e-06,
+      "loss": 0.4962,
+      "step": 7574
+    },
+    {
+      "epoch": 2.976142257347493,
+      "grad_norm": 0.44642306301580476,
+      "learning_rate": 4.756243954312926e-06,
+      "loss": 0.4906,
+      "step": 7575
+    },
+    {
+      "epoch": 2.976537416646085,
+      "grad_norm": 0.4551265744121246,
+      "learning_rate": 4.756176471925092e-06,
+      "loss": 0.5114,
+      "step": 7576
+    },
+    {
+      "epoch": 2.976932575944678,
+      "grad_norm": 0.44159910731274965,
+      "learning_rate": 4.756108980676367e-06,
+      "loss": 0.5182,
+      "step": 7577
+    },
+    {
+      "epoch": 2.97732773524327,
+      "grad_norm": 0.43027421694295137,
+      "learning_rate": 4.756041480567017e-06,
+      "loss": 0.4889,
+      "step": 7578
+    },
+    {
+      "epoch": 2.9777228945418623,
+      "grad_norm": 0.4337873859510924,
+      "learning_rate": 4.755973971597305e-06,
+      "loss": 0.4941,
+      "step": 7579
+    },
+    {
+      "epoch": 2.9781180538404546,
+      "grad_norm": 0.45031945697341647,
+      "learning_rate": 4.7559064537674975e-06,
+      "loss": 0.4998,
+      "step": 7580
+    },
+    {
+      "epoch": 2.978513213139047,
+      "grad_norm": 0.4712226590155607,
+      "learning_rate": 4.755838927077859e-06,
+      "loss": 0.5111,
+      "step": 7581
+    },
+    {
+      "epoch": 2.978908372437639,
+      "grad_norm": 0.4409693860661984,
+      "learning_rate": 4.755771391528655e-06,
+      "loss": 0.5069,
+      "step": 7582
+    },
+    {
+      "epoch": 2.9793035317362313,
+      "grad_norm": 0.4640142389222412,
+      "learning_rate": 4.755703847120152e-06,
+      "loss": 0.5164,
+      "step": 7583
+    },
+    {
+      "epoch": 2.9796986910348235,
+      "grad_norm": 0.45381043122965947,
+      "learning_rate": 4.7556362938526124e-06,
+      "loss": 0.5041,
+      "step": 7584
+    },
+    {
+      "epoch": 2.980093850333416,
+      "grad_norm": 0.4587590971637779,
+      "learning_rate": 4.755568731726304e-06,
+      "loss": 0.5203,
+      "step": 7585
+    },
+    {
+      "epoch": 2.980489009632008,
+      "grad_norm": 0.4644178501879429,
+      "learning_rate": 4.755501160741491e-06,
+      "loss": 0.5328,
+      "step": 7586
+    },
+    {
+      "epoch": 2.9808841689306003,
+      "grad_norm": 0.44776614224075106,
+      "learning_rate": 4.755433580898439e-06,
+      "loss": 0.5036,
+      "step": 7587
+    },
+    {
+      "epoch": 2.9812793282291925,
+      "grad_norm": 0.43844012754058254,
+      "learning_rate": 4.7553659921974134e-06,
+      "loss": 0.4832,
+      "step": 7588
+    },
+    {
+      "epoch": 2.9816744875277847,
+      "grad_norm": 0.45286284274574534,
+      "learning_rate": 4.75529839463868e-06,
+      "loss": 0.5185,
+      "step": 7589
+    },
+    {
+      "epoch": 2.982069646826377,
+      "grad_norm": 0.46049052268894997,
+      "learning_rate": 4.755230788222504e-06,
+      "loss": 0.5378,
+      "step": 7590
+    },
+    {
+      "epoch": 2.9824648061249692,
+      "grad_norm": 0.4485615523268441,
+      "learning_rate": 4.755163172949151e-06,
+      "loss": 0.5009,
+      "step": 7591
+    },
+    {
+      "epoch": 2.9828599654235615,
+      "grad_norm": 0.45034777636007545,
+      "learning_rate": 4.755095548818886e-06,
+      "loss": 0.5236,
+      "step": 7592
+    },
+    {
+      "epoch": 2.9832551247221537,
+      "grad_norm": 0.4531560958419335,
+      "learning_rate": 4.755027915831975e-06,
+      "loss": 0.5101,
+      "step": 7593
+    },
+    {
+      "epoch": 2.983650284020746,
+      "grad_norm": 0.452406575756383,
+      "learning_rate": 4.754960273988684e-06,
+      "loss": 0.51,
+      "step": 7594
+    },
+    {
+      "epoch": 2.984045443319338,
+      "grad_norm": 0.4404650508761369,
+      "learning_rate": 4.754892623289279e-06,
+      "loss": 0.498,
+      "step": 7595
+    },
+    {
+      "epoch": 2.9844406026179304,
+      "grad_norm": 0.45105701449061847,
+      "learning_rate": 4.754824963734024e-06,
+      "loss": 0.5247,
+      "step": 7596
+    },
+    {
+      "epoch": 2.9848357619165227,
+      "grad_norm": 0.4544785032506456,
+      "learning_rate": 4.754757295323186e-06,
+      "loss": 0.5006,
+      "step": 7597
+    },
+    {
+      "epoch": 2.985230921215115,
+      "grad_norm": 0.44648615459844027,
+      "learning_rate": 4.7546896180570305e-06,
+      "loss": 0.5139,
+      "step": 7598
+    },
+    {
+      "epoch": 2.985626080513707,
+      "grad_norm": 0.46035057899107057,
+      "learning_rate": 4.754621931935823e-06,
+      "loss": 0.5129,
+      "step": 7599
+    },
+    {
+      "epoch": 2.9860212398122994,
+      "grad_norm": 0.468475157789722,
+      "learning_rate": 4.75455423695983e-06,
+      "loss": 0.5284,
+      "step": 7600
+    },
+    {
+      "epoch": 2.9864163991108916,
+      "grad_norm": 0.455654479720452,
+      "learning_rate": 4.7544865331293175e-06,
+      "loss": 0.5141,
+      "step": 7601
+    },
+    {
+      "epoch": 2.986811558409484,
+      "grad_norm": 0.43877314753613916,
+      "learning_rate": 4.75441882044455e-06,
+      "loss": 0.5102,
+      "step": 7602
+    },
+    {
+      "epoch": 2.987206717708076,
+      "grad_norm": 0.44066884059445266,
+      "learning_rate": 4.754351098905795e-06,
+      "loss": 0.5096,
+      "step": 7603
+    },
+    {
+      "epoch": 2.9876018770066683,
+      "grad_norm": 0.455561747753586,
+      "learning_rate": 4.754283368513317e-06,
+      "loss": 0.5194,
+      "step": 7604
+    },
+    {
+      "epoch": 2.9879970363052606,
+      "grad_norm": 0.4532632517976643,
+      "learning_rate": 4.754215629267384e-06,
+      "loss": 0.5145,
+      "step": 7605
+    },
+    {
+      "epoch": 2.988392195603853,
+      "grad_norm": 0.4469539479625391,
+      "learning_rate": 4.75414788116826e-06,
+      "loss": 0.5018,
+      "step": 7606
+    },
+    {
+      "epoch": 2.988787354902445,
+      "grad_norm": 0.45541116390015846,
+      "learning_rate": 4.754080124216212e-06,
+      "loss": 0.4859,
+      "step": 7607
+    },
+    {
+      "epoch": 2.9891825142010373,
+      "grad_norm": 0.4496311984557622,
+      "learning_rate": 4.754012358411506e-06,
+      "loss": 0.5018,
+      "step": 7608
+    },
+    {
+      "epoch": 2.9895776734996296,
+      "grad_norm": 0.4530431443004678,
+      "learning_rate": 4.753944583754408e-06,
+      "loss": 0.489,
+      "step": 7609
+    },
+    {
+      "epoch": 2.989972832798222,
+      "grad_norm": 0.4475247994930158,
+      "learning_rate": 4.753876800245186e-06,
+      "loss": 0.4946,
+      "step": 7610
+    },
+    {
+      "epoch": 2.990367992096814,
+      "grad_norm": 0.44677019584511535,
+      "learning_rate": 4.753809007884103e-06,
+      "loss": 0.5155,
+      "step": 7611
+    },
+    {
+      "epoch": 2.9907631513954063,
+      "grad_norm": 0.4491717613748345,
+      "learning_rate": 4.753741206671426e-06,
+      "loss": 0.489,
+      "step": 7612
+    },
+    {
+      "epoch": 2.9911583106939985,
+      "grad_norm": 0.4718935616370409,
+      "learning_rate": 4.753673396607423e-06,
+      "loss": 0.5084,
+      "step": 7613
+    },
+    {
+      "epoch": 2.9915534699925908,
+      "grad_norm": 0.4526066949527833,
+      "learning_rate": 4.7536055776923596e-06,
+      "loss": 0.5151,
+      "step": 7614
+    },
+    {
+      "epoch": 2.991948629291183,
+      "grad_norm": 0.4506375450777332,
+      "learning_rate": 4.753537749926502e-06,
+      "loss": 0.5064,
+      "step": 7615
+    },
+    {
+      "epoch": 2.9923437885897752,
+      "grad_norm": 0.45034114642880246,
+      "learning_rate": 4.753469913310116e-06,
+      "loss": 0.5166,
+      "step": 7616
+    },
+    {
+      "epoch": 2.9927389478883675,
+      "grad_norm": 0.4407471329242633,
+      "learning_rate": 4.753402067843469e-06,
+      "loss": 0.5098,
+      "step": 7617
+    },
+    {
+      "epoch": 2.9931341071869597,
+      "grad_norm": 0.4761783641018543,
+      "learning_rate": 4.753334213526827e-06,
+      "loss": 0.5045,
+      "step": 7618
+    },
+    {
+      "epoch": 2.993529266485552,
+      "grad_norm": 0.4549004819674926,
+      "learning_rate": 4.7532663503604566e-06,
+      "loss": 0.5019,
+      "step": 7619
+    },
+    {
+      "epoch": 2.993924425784144,
+      "grad_norm": 0.43524384467655386,
+      "learning_rate": 4.753198478344624e-06,
+      "loss": 0.4918,
+      "step": 7620
+    },
+    {
+      "epoch": 2.9943195850827364,
+      "grad_norm": 0.45140783826694003,
+      "learning_rate": 4.753130597479596e-06,
+      "loss": 0.5067,
+      "step": 7621
+    },
+    {
+      "epoch": 2.9947147443813287,
+      "grad_norm": 0.4481553706282452,
+      "learning_rate": 4.753062707765639e-06,
+      "loss": 0.5192,
+      "step": 7622
+    },
+    {
+      "epoch": 2.995109903679921,
+      "grad_norm": 0.44777175097676636,
+      "learning_rate": 4.7529948092030204e-06,
+      "loss": 0.5129,
+      "step": 7623
+    },
+    {
+      "epoch": 2.995505062978513,
+      "grad_norm": 0.4479442802590514,
+      "learning_rate": 4.752926901792006e-06,
+      "loss": 0.4877,
+      "step": 7624
+    },
+    {
+      "epoch": 2.9959002222771054,
+      "grad_norm": 0.446793554178131,
+      "learning_rate": 4.752858985532862e-06,
+      "loss": 0.5011,
+      "step": 7625
+    },
+    {
+      "epoch": 2.9962953815756976,
+      "grad_norm": 0.45259107436589663,
+      "learning_rate": 4.7527910604258575e-06,
+      "loss": 0.4992,
+      "step": 7626
+    },
+    {
+      "epoch": 2.99669054087429,
+      "grad_norm": 0.4445697317032285,
+      "learning_rate": 4.752723126471257e-06,
+      "loss": 0.5161,
+      "step": 7627
+    },
+    {
+      "epoch": 2.997085700172882,
+      "grad_norm": 0.46322756286869876,
+      "learning_rate": 4.752655183669327e-06,
+      "loss": 0.5021,
+      "step": 7628
+    },
+    {
+      "epoch": 2.9974808594714744,
+      "grad_norm": 0.4483882532172694,
+      "learning_rate": 4.752587232020337e-06,
+      "loss": 0.4896,
+      "step": 7629
+    },
+    {
+      "epoch": 2.9978760187700666,
+      "grad_norm": 0.4474447067182486,
+      "learning_rate": 4.7525192715245505e-06,
+      "loss": 0.5,
+      "step": 7630
+    },
+    {
+      "epoch": 2.998271178068659,
+      "grad_norm": 0.46092387125951806,
+      "learning_rate": 4.752451302182237e-06,
+      "loss": 0.5081,
+      "step": 7631
+    },
+    {
+      "epoch": 2.998666337367251,
+      "grad_norm": 0.44451538632392057,
+      "learning_rate": 4.752383323993663e-06,
+      "loss": 0.506,
+      "step": 7632
+    },
+    {
+      "epoch": 2.9990614966658433,
+      "grad_norm": 0.44559816658293144,
+      "learning_rate": 4.752315336959094e-06,
+      "loss": 0.4877,
+      "step": 7633
+    },
+    {
+      "epoch": 2.9994566559644356,
+      "grad_norm": 0.4729115645935803,
+      "learning_rate": 4.752247341078798e-06,
+      "loss": 0.5108,
+      "step": 7634
+    },
+    {
+      "epoch": 2.999851815263028,
+      "grad_norm": 0.4634195899356373,
+      "learning_rate": 4.752179336353043e-06,
+      "loss": 0.507,
+      "step": 7635
+    },
+    {
+      "epoch": 3.0002469745616205,
+      "grad_norm": 0.44336385846437987,
+      "learning_rate": 4.752111322782095e-06,
+      "loss": 0.4778,
+      "step": 7636
+    },
+    {
+      "epoch": 3.0006421338602127,
+      "grad_norm": 0.4516001642956355,
+      "learning_rate": 4.752043300366222e-06,
+      "loss": 0.5019,
+      "step": 7637
+    },
+    {
+      "epoch": 3.001037293158805,
+      "grad_norm": 0.45717790797056945,
+      "learning_rate": 4.751975269105689e-06,
+      "loss": 0.5165,
+      "step": 7638
+    },
+    {
+      "epoch": 3.001432452457397,
+      "grad_norm": 0.4584460619343716,
+      "learning_rate": 4.751907229000765e-06,
+      "loss": 0.5139,
+      "step": 7639
+    },
+    {
+      "epoch": 3.0018276117559894,
+      "grad_norm": 0.4505928965434931,
+      "learning_rate": 4.751839180051717e-06,
+      "loss": 0.4879,
+      "step": 7640
+    },
+    {
+      "epoch": 3.0022227710545817,
+      "grad_norm": 0.4550585108130313,
+      "learning_rate": 4.751771122258812e-06,
+      "loss": 0.518,
+      "step": 7641
+    },
+    {
+      "epoch": 3.002617930353174,
+      "grad_norm": 0.4641841812808131,
+      "learning_rate": 4.751703055622317e-06,
+      "loss": 0.5098,
+      "step": 7642
+    },
+    {
+      "epoch": 3.003013089651766,
+      "grad_norm": 0.4619224304919582,
+      "learning_rate": 4.7516349801424995e-06,
+      "loss": 0.492,
+      "step": 7643
+    },
+    {
+      "epoch": 3.0034082489503584,
+      "grad_norm": 0.45126132444360667,
+      "learning_rate": 4.751566895819628e-06,
+      "loss": 0.5081,
+      "step": 7644
+    },
+    {
+      "epoch": 3.0038034082489506,
+      "grad_norm": 0.4397768300217093,
+      "learning_rate": 4.7514988026539686e-06,
+      "loss": 0.4917,
+      "step": 7645
+    },
+    {
+      "epoch": 3.004198567547543,
+      "grad_norm": 0.45619992289347067,
+      "learning_rate": 4.75143070064579e-06,
+      "loss": 0.4798,
+      "step": 7646
+    },
+    {
+      "epoch": 3.004593726846135,
+      "grad_norm": 0.4498728600811053,
+      "learning_rate": 4.751362589795358e-06,
+      "loss": 0.4953,
+      "step": 7647
+    },
+    {
+      "epoch": 3.0049888861447274,
+      "grad_norm": 0.451458140159821,
+      "learning_rate": 4.751294470102941e-06,
+      "loss": 0.4945,
+      "step": 7648
+    },
+    {
+      "epoch": 3.0053840454433196,
+      "grad_norm": 0.44656047500039336,
+      "learning_rate": 4.751226341568806e-06,
+      "loss": 0.4945,
+      "step": 7649
+    },
+    {
+      "epoch": 3.005779204741912,
+      "grad_norm": 0.44674554953771023,
+      "learning_rate": 4.75115820419322e-06,
+      "loss": 0.5066,
+      "step": 7650
+    },
+    {
+      "epoch": 3.006174364040504,
+      "grad_norm": 0.5042336016390171,
+      "learning_rate": 4.751090057976453e-06,
+      "loss": 0.5143,
+      "step": 7651
+    },
+    {
+      "epoch": 3.0065695233390963,
+      "grad_norm": 0.4703862669838775,
+      "learning_rate": 4.751021902918771e-06,
+      "loss": 0.5049,
+      "step": 7652
+    },
+    {
+      "epoch": 3.0069646826376886,
+      "grad_norm": 0.44280468476812773,
+      "learning_rate": 4.750953739020441e-06,
+      "loss": 0.5,
+      "step": 7653
+    },
+    {
+      "epoch": 3.007359841936281,
+      "grad_norm": 0.4423726279571279,
+      "learning_rate": 4.7508855662817325e-06,
+      "loss": 0.5138,
+      "step": 7654
+    },
+    {
+      "epoch": 3.007755001234873,
+      "grad_norm": 0.4668415881963999,
+      "learning_rate": 4.750817384702912e-06,
+      "loss": 0.5075,
+      "step": 7655
+    },
+    {
+      "epoch": 3.0081501605334653,
+      "grad_norm": 0.4395316160725124,
+      "learning_rate": 4.750749194284248e-06,
+      "loss": 0.4892,
+      "step": 7656
+    },
+    {
+      "epoch": 3.0085453198320575,
+      "grad_norm": 0.4523097575996378,
+      "learning_rate": 4.750680995026007e-06,
+      "loss": 0.5155,
+      "step": 7657
+    },
+    {
+      "epoch": 3.0089404791306498,
+      "grad_norm": 0.43939023061801286,
+      "learning_rate": 4.7506127869284585e-06,
+      "loss": 0.4957,
+      "step": 7658
+    },
+    {
+      "epoch": 3.009335638429242,
+      "grad_norm": 0.46277152916188,
+      "learning_rate": 4.7505445699918695e-06,
+      "loss": 0.4936,
+      "step": 7659
+    },
+    {
+      "epoch": 3.0097307977278343,
+      "grad_norm": 0.46615632583881483,
+      "learning_rate": 4.750476344216508e-06,
+      "loss": 0.5201,
+      "step": 7660
+    },
+    {
+      "epoch": 3.0101259570264265,
+      "grad_norm": 0.4914314812774924,
+      "learning_rate": 4.750408109602641e-06,
+      "loss": 0.5254,
+      "step": 7661
+    },
+    {
+      "epoch": 3.0105211163250187,
+      "grad_norm": 0.459305044150223,
+      "learning_rate": 4.7503398661505386e-06,
+      "loss": 0.5076,
+      "step": 7662
+    },
+    {
+      "epoch": 3.010916275623611,
+      "grad_norm": 0.46075158979928715,
+      "learning_rate": 4.750271613860468e-06,
+      "loss": 0.5264,
+      "step": 7663
+    },
+    {
+      "epoch": 3.011311434922203,
+      "grad_norm": 0.450653814827898,
+      "learning_rate": 4.750203352732696e-06,
+      "loss": 0.5151,
+      "step": 7664
+    },
+    {
+      "epoch": 3.0117065942207955,
+      "grad_norm": 0.458283864427654,
+      "learning_rate": 4.750135082767492e-06,
+      "loss": 0.5008,
+      "step": 7665
+    },
+    {
+      "epoch": 3.0121017535193877,
+      "grad_norm": 0.4685611930993023,
+      "learning_rate": 4.750066803965124e-06,
+      "loss": 0.5187,
+      "step": 7666
+    },
+    {
+      "epoch": 3.01249691281798,
+      "grad_norm": 0.47791128196843446,
+      "learning_rate": 4.749998516325859e-06,
+      "loss": 0.5019,
+      "step": 7667
+    },
+    {
+      "epoch": 3.012892072116572,
+      "grad_norm": 0.4872509174602875,
+      "learning_rate": 4.749930219849967e-06,
+      "loss": 0.5231,
+      "step": 7668
+    },
+    {
+      "epoch": 3.0132872314151644,
+      "grad_norm": 0.4684674472270326,
+      "learning_rate": 4.749861914537715e-06,
+      "loss": 0.5076,
+      "step": 7669
+    },
+    {
+      "epoch": 3.0136823907137567,
+      "grad_norm": 0.4421540054512149,
+      "learning_rate": 4.749793600389372e-06,
+      "loss": 0.528,
+      "step": 7670
+    },
+    {
+      "epoch": 3.014077550012349,
+      "grad_norm": 0.46434385398269223,
+      "learning_rate": 4.749725277405205e-06,
+      "loss": 0.5189,
+      "step": 7671
+    },
+    {
+      "epoch": 3.014472709310941,
+      "grad_norm": 0.45238786433034117,
+      "learning_rate": 4.749656945585484e-06,
+      "loss": 0.4985,
+      "step": 7672
+    },
+    {
+      "epoch": 3.0148678686095334,
+      "grad_norm": 0.4740826575920035,
+      "learning_rate": 4.749588604930476e-06,
+      "loss": 0.4961,
+      "step": 7673
+    },
+    {
+      "epoch": 3.0152630279081256,
+      "grad_norm": 0.4356051895472756,
+      "learning_rate": 4.749520255440451e-06,
+      "loss": 0.5006,
+      "step": 7674
+    },
+    {
+      "epoch": 3.015658187206718,
+      "grad_norm": 0.44354071567582887,
+      "learning_rate": 4.749451897115675e-06,
+      "loss": 0.4959,
+      "step": 7675
+    },
+    {
+      "epoch": 3.01605334650531,
+      "grad_norm": 0.4561647234991173,
+      "learning_rate": 4.749383529956419e-06,
+      "loss": 0.4928,
+      "step": 7676
+    },
+    {
+      "epoch": 3.0164485058039023,
+      "grad_norm": 0.4883457852318272,
+      "learning_rate": 4.74931515396295e-06,
+      "loss": 0.5144,
+      "step": 7677
+    },
+    {
+      "epoch": 3.0168436651024946,
+      "grad_norm": 0.44882770949870965,
+      "learning_rate": 4.749246769135537e-06,
+      "loss": 0.4993,
+      "step": 7678
+    },
+    {
+      "epoch": 3.000222277105458,
+      "grad_norm": 0.7010342589000016,
+      "learning_rate": 4.749178375474448e-06,
+      "loss": 0.473,
+      "step": 7679
+    },
+    {
+      "epoch": 3.0006174364040503,
+      "grad_norm": 0.915590098727133,
+      "learning_rate": 4.749109972979953e-06,
+      "loss": 0.45,
+      "step": 7680
+    },
+    {
+      "epoch": 3.0010125957026426,
+      "grad_norm": 0.6897379840405359,
+      "learning_rate": 4.74904156165232e-06,
+      "loss": 0.4431,
+      "step": 7681
+    },
+    {
+      "epoch": 3.001407755001235,
+      "grad_norm": 0.5249346162970593,
+      "learning_rate": 4.748973141491816e-06,
+      "loss": 0.4328,
+      "step": 7682
+    },
+    {
+      "epoch": 3.001802914299827,
+      "grad_norm": 0.9709224152272217,
+      "learning_rate": 4.748904712498712e-06,
+      "loss": 0.4507,
+      "step": 7683
+    },
+    {
+      "epoch": 3.0021980735984193,
+      "grad_norm": 0.9257083172881498,
+      "learning_rate": 4.748836274673275e-06,
+      "loss": 0.4171,
+      "step": 7684
+    },
+    {
+      "epoch": 3.0025932328970115,
+      "grad_norm": 0.6654293170865759,
+      "learning_rate": 4.748767828015777e-06,
+      "loss": 0.4453,
+      "step": 7685
+    },
+    {
+      "epoch": 3.0029883921956038,
+      "grad_norm": 0.5713696408874206,
+      "learning_rate": 4.7486993725264824e-06,
+      "loss": 0.4407,
+      "step": 7686
+    },
+    {
+      "epoch": 3.003383551494196,
+      "grad_norm": 0.6629984546742934,
+      "learning_rate": 4.748630908205663e-06,
+      "loss": 0.4266,
+      "step": 7687
+    },
+    {
+      "epoch": 3.0037787107927882,
+      "grad_norm": 0.7031964288055295,
+      "learning_rate": 4.748562435053587e-06,
+      "loss": 0.4571,
+      "step": 7688
+    },
+    {
+      "epoch": 3.0041738700913805,
+      "grad_norm": 0.8005034174097257,
+      "learning_rate": 4.748493953070522e-06,
+      "loss": 0.441,
+      "step": 7689
+    },
+    {
+      "epoch": 3.0045690293899727,
+      "grad_norm": 0.5656503152080794,
+      "learning_rate": 4.748425462256739e-06,
+      "loss": 0.4482,
+      "step": 7690
+    },
+    {
+      "epoch": 3.004964188688565,
+      "grad_norm": 0.5805979657928287,
+      "learning_rate": 4.748356962612506e-06,
+      "loss": 0.4308,
+      "step": 7691
+    },
+    {
+      "epoch": 3.005359347987157,
+      "grad_norm": 0.6569090685400837,
+      "learning_rate": 4.7482884541380915e-06,
+      "loss": 0.4519,
+      "step": 7692
+    },
+    {
+      "epoch": 3.0057545072857494,
+      "grad_norm": 0.6217734324861036,
+      "learning_rate": 4.748219936833766e-06,
+      "loss": 0.4349,
+      "step": 7693
+    },
+    {
+      "epoch": 3.0061496665843417,
+      "grad_norm": 0.5796522323129869,
+      "learning_rate": 4.748151410699798e-06,
+      "loss": 0.4479,
+      "step": 7694
+    },
+    {
+      "epoch": 3.006544825882934,
+      "grad_norm": 0.5356679945336663,
+      "learning_rate": 4.7480828757364555e-06,
+      "loss": 0.4209,
+      "step": 7695
+    },
+    {
+      "epoch": 3.006939985181526,
+      "grad_norm": 0.5544815975849889,
+      "learning_rate": 4.7480143319440094e-06,
+      "loss": 0.4321,
+      "step": 7696
+    },
+    {
+      "epoch": 3.0073351444801184,
+      "grad_norm": 0.5615701121311352,
+      "learning_rate": 4.747945779322727e-06,
+      "loss": 0.4498,
+      "step": 7697
+    },
+    {
+      "epoch": 3.0077303037787106,
+      "grad_norm": 0.5591511570043203,
+      "learning_rate": 4.747877217872879e-06,
+      "loss": 0.4434,
+      "step": 7698
+    },
+    {
+      "epoch": 3.008125463077303,
+      "grad_norm": 0.5231502732146598,
+      "learning_rate": 4.747808647594735e-06,
+      "loss": 0.4454,
+      "step": 7699
+    },
+    {
+      "epoch": 3.008520622375895,
+      "grad_norm": 0.5749999982521911,
+      "learning_rate": 4.747740068488563e-06,
+      "loss": 0.4335,
+      "step": 7700
+    },
+    {
+      "epoch": 3.0089157816744874,
+      "grad_norm": 0.5594741958286711,
+      "learning_rate": 4.747671480554633e-06,
+      "loss": 0.4479,
+      "step": 7701
+    },
+    {
+      "epoch": 3.0093109409730796,
+      "grad_norm": 0.5229530513831508,
+      "learning_rate": 4.747602883793215e-06,
+      "loss": 0.4206,
+      "step": 7702
+    },
+    {
+      "epoch": 3.009706100271672,
+      "grad_norm": 0.5269098553306883,
+      "learning_rate": 4.747534278204576e-06,
+      "loss": 0.4259,
+      "step": 7703
+    },
+    {
+      "epoch": 3.010101259570264,
+      "grad_norm": 0.5743914914071026,
+      "learning_rate": 4.747465663788989e-06,
+      "loss": 0.4304,
+      "step": 7704
+    },
+    {
+      "epoch": 3.0104964188688563,
+      "grad_norm": 0.5032330836219913,
+      "learning_rate": 4.747397040546721e-06,
+      "loss": 0.4369,
+      "step": 7705
+    },
+    {
+      "epoch": 3.0108915781674486,
+      "grad_norm": 0.5170294237714139,
+      "learning_rate": 4.747328408478042e-06,
+      "loss": 0.442,
+      "step": 7706
+    },
+    {
+      "epoch": 3.011286737466041,
+      "grad_norm": 0.5325099151723756,
+      "learning_rate": 4.747259767583221e-06,
+      "loss": 0.4423,
+      "step": 7707
+    },
+    {
+      "epoch": 3.011681896764633,
+      "grad_norm": 0.5263448352266229,
+      "learning_rate": 4.7471911178625285e-06,
+      "loss": 0.4553,
+      "step": 7708
+    },
+    {
+      "epoch": 3.0120770560632253,
+      "grad_norm": 0.5935074925411176,
+      "learning_rate": 4.747122459316235e-06,
+      "loss": 0.4524,
+      "step": 7709
+    },
+    {
+      "epoch": 3.0124722153618175,
+      "grad_norm": 0.5162191886079706,
+      "learning_rate": 4.747053791944607e-06,
+      "loss": 0.4536,
+      "step": 7710
+    },
+    {
+      "epoch": 3.0128673746604098,
+      "grad_norm": 0.48679105836545394,
+      "learning_rate": 4.746985115747918e-06,
+      "loss": 0.4128,
+      "step": 7711
+    },
+    {
+      "epoch": 3.013262533959002,
+      "grad_norm": 0.5158017050260368,
+      "learning_rate": 4.746916430726435e-06,
+      "loss": 0.4413,
+      "step": 7712
+    },
+    {
+      "epoch": 3.0136576932575943,
+      "grad_norm": 0.5255387088096938,
+      "learning_rate": 4.746847736880429e-06,
+      "loss": 0.4331,
+      "step": 7713
+    },
+    {
+      "epoch": 3.014052852556187,
+      "grad_norm": 0.49833669861127644,
+      "learning_rate": 4.746779034210169e-06,
+      "loss": 0.4546,
+      "step": 7714
+    },
+    {
+      "epoch": 3.014448011854779,
+      "grad_norm": 0.5013002795411056,
+      "learning_rate": 4.746710322715926e-06,
+      "loss": 0.4338,
+      "step": 7715
+    },
+    {
+      "epoch": 3.0148431711533714,
+      "grad_norm": 0.47380997163306915,
+      "learning_rate": 4.746641602397969e-06,
+      "loss": 0.452,
+      "step": 7716
+    },
+    {
+      "epoch": 3.0152383304519637,
+      "grad_norm": 0.48815835539987246,
+      "learning_rate": 4.746572873256568e-06,
+      "loss": 0.4325,
+      "step": 7717
+    },
+    {
+      "epoch": 3.015633489750556,
+      "grad_norm": 0.5234825947101446,
+      "learning_rate": 4.746504135291992e-06,
+      "loss": 0.4627,
+      "step": 7718
+    },
+    {
+      "epoch": 3.016028649049148,
+      "grad_norm": 0.46528150114040756,
+      "learning_rate": 4.746435388504513e-06,
+      "loss": 0.4135,
+      "step": 7719
+    },
+    {
+      "epoch": 3.0164238083477404,
+      "grad_norm": 0.5186401009350103,
+      "learning_rate": 4.746366632894399e-06,
+      "loss": 0.4434,
+      "step": 7720
+    },
+    {
+      "epoch": 3.0168189676463326,
+      "grad_norm": 0.48449546596605336,
+      "learning_rate": 4.746297868461922e-06,
+      "loss": 0.4275,
+      "step": 7721
+    },
+    {
+      "epoch": 3.017214126944925,
+      "grad_norm": 0.4972566211579247,
+      "learning_rate": 4.74622909520735e-06,
+      "loss": 0.4352,
+      "step": 7722
+    },
+    {
+      "epoch": 3.017609286243517,
+      "grad_norm": 0.49817262550108005,
+      "learning_rate": 4.746160313130955e-06,
+      "loss": 0.4427,
+      "step": 7723
+    },
+    {
+      "epoch": 3.0180044455421093,
+      "grad_norm": 0.4807900493796438,
+      "learning_rate": 4.746091522233006e-06,
+      "loss": 0.4458,
+      "step": 7724
+    },
+    {
+      "epoch": 3.0183996048407016,
+      "grad_norm": 0.486178408026208,
+      "learning_rate": 4.746022722513772e-06,
+      "loss": 0.4375,
+      "step": 7725
+    },
+    {
+      "epoch": 3.018794764139294,
+      "grad_norm": 0.47609321433727503,
+      "learning_rate": 4.745953913973526e-06,
+      "loss": 0.439,
+      "step": 7726
+    },
+    {
+      "epoch": 3.019189923437886,
+      "grad_norm": 0.48600695149007456,
+      "learning_rate": 4.745885096612537e-06,
+      "loss": 0.427,
+      "step": 7727
+    },
+    {
+      "epoch": 3.0195850827364783,
+      "grad_norm": 0.509073376294911,
+      "learning_rate": 4.745816270431075e-06,
+      "loss": 0.4388,
+      "step": 7728
+    },
+    {
+      "epoch": 3.0199802420350705,
+      "grad_norm": 0.5057289776617495,
+      "learning_rate": 4.74574743542941e-06,
+      "loss": 0.4388,
+      "step": 7729
+    },
+    {
+      "epoch": 3.020375401333663,
+      "grad_norm": 0.506257402966116,
+      "learning_rate": 4.745678591607813e-06,
+      "loss": 0.4337,
+      "step": 7730
+    },
+    {
+      "epoch": 3.020770560632255,
+      "grad_norm": 0.5051592228444768,
+      "learning_rate": 4.745609738966554e-06,
+      "loss": 0.4439,
+      "step": 7731
+    },
+    {
+      "epoch": 3.0211657199308473,
+      "grad_norm": 0.5014567224089678,
+      "learning_rate": 4.745540877505904e-06,
+      "loss": 0.4341,
+      "step": 7732
+    },
+    {
+      "epoch": 3.0215608792294395,
+      "grad_norm": 0.5016734580118337,
+      "learning_rate": 4.745472007226133e-06,
+      "loss": 0.437,
+      "step": 7733
+    },
+    {
+      "epoch": 3.0219560385280317,
+      "grad_norm": 0.5017266865893028,
+      "learning_rate": 4.7454031281275105e-06,
+      "loss": 0.4297,
+      "step": 7734
+    },
+    {
+      "epoch": 3.022351197826624,
+      "grad_norm": 0.4925195971174838,
+      "learning_rate": 4.745334240210309e-06,
+      "loss": 0.4397,
+      "step": 7735
+    },
+    {
+      "epoch": 3.022746357125216,
+      "grad_norm": 0.47846300333645714,
+      "learning_rate": 4.745265343474797e-06,
+      "loss": 0.4587,
+      "step": 7736
+    },
+    {
+      "epoch": 3.0231415164238085,
+      "grad_norm": 0.4619490312270768,
+      "learning_rate": 4.745196437921247e-06,
+      "loss": 0.4376,
+      "step": 7737
+    },
+    {
+      "epoch": 3.0235366757224007,
+      "grad_norm": 0.4863155493658263,
+      "learning_rate": 4.745127523549928e-06,
+      "loss": 0.486,
+      "step": 7738
+    },
+    {
+      "epoch": 3.023931835020993,
+      "grad_norm": 0.5009948111270373,
+      "learning_rate": 4.7450586003611124e-06,
+      "loss": 0.4433,
+      "step": 7739
+    },
+    {
+      "epoch": 3.024326994319585,
+      "grad_norm": 0.49175225322505095,
+      "learning_rate": 4.744989668355069e-06,
+      "loss": 0.4466,
+      "step": 7740
+    },
+    {
+      "epoch": 3.0247221536181774,
+      "grad_norm": 0.4997137132091148,
+      "learning_rate": 4.744920727532069e-06,
+      "loss": 0.4422,
+      "step": 7741
+    },
+    {
+      "epoch": 3.0251173129167697,
+      "grad_norm": 0.49782214688948523,
+      "learning_rate": 4.744851777892386e-06,
+      "loss": 0.4215,
+      "step": 7742
+    },
+    {
+      "epoch": 3.025512472215362,
+      "grad_norm": 0.4878799150602817,
+      "learning_rate": 4.744782819436287e-06,
+      "loss": 0.4689,
+      "step": 7743
+    },
+    {
+      "epoch": 3.025907631513954,
+      "grad_norm": 0.4801197425799182,
+      "learning_rate": 4.7447138521640435e-06,
+      "loss": 0.4133,
+      "step": 7744
+    },
+    {
+      "epoch": 3.0263027908125464,
+      "grad_norm": 0.5153722734142074,
+      "learning_rate": 4.744644876075926e-06,
+      "loss": 0.4328,
+      "step": 7745
+    },
+    {
+      "epoch": 3.0266979501111386,
+      "grad_norm": 0.48115731489986924,
+      "learning_rate": 4.744575891172209e-06,
+      "loss": 0.4477,
+      "step": 7746
+    },
+    {
+      "epoch": 3.027093109409731,
+      "grad_norm": 0.5030409292416705,
+      "learning_rate": 4.74450689745316e-06,
+      "loss": 0.458,
+      "step": 7747
+    },
+    {
+      "epoch": 3.027488268708323,
+      "grad_norm": 0.47695935750014723,
+      "learning_rate": 4.7444378949190505e-06,
+      "loss": 0.4647,
+      "step": 7748
+    },
+    {
+      "epoch": 3.0278834280069153,
+      "grad_norm": 0.4898314891709876,
+      "learning_rate": 4.744368883570152e-06,
+      "loss": 0.4581,
+      "step": 7749
+    },
+    {
+      "epoch": 3.0282785873055076,
+      "grad_norm": 0.5003360236778067,
+      "learning_rate": 4.7442998634067356e-06,
+      "loss": 0.4249,
+      "step": 7750
+    },
+    {
+      "epoch": 3.0286737466041,
+      "grad_norm": 0.5091015290593945,
+      "learning_rate": 4.744230834429071e-06,
+      "loss": 0.4614,
+      "step": 7751
+    },
+    {
+      "epoch": 3.029068905902692,
+      "grad_norm": 0.49509617467245814,
+      "learning_rate": 4.744161796637432e-06,
+      "loss": 0.4538,
+      "step": 7752
+    },
+    {
+      "epoch": 3.0294640652012843,
+      "grad_norm": 0.49300299029182093,
+      "learning_rate": 4.7440927500320875e-06,
+      "loss": 0.4303,
+      "step": 7753
+    },
+    {
+      "epoch": 3.0298592244998765,
+      "grad_norm": 0.49820644043525975,
+      "learning_rate": 4.74402369461331e-06,
+      "loss": 0.4286,
+      "step": 7754
+    },
+    {
+      "epoch": 3.030254383798469,
+      "grad_norm": 0.5079411624836448,
+      "learning_rate": 4.743954630381369e-06,
+      "loss": 0.4414,
+      "step": 7755
+    },
+    {
+      "epoch": 3.030649543097061,
+      "grad_norm": 0.4842722586826298,
+      "learning_rate": 4.743885557336537e-06,
+      "loss": 0.4296,
+      "step": 7756
+    },
+    {
+      "epoch": 3.0310447023956533,
+      "grad_norm": 0.48422624836220324,
+      "learning_rate": 4.743816475479086e-06,
+      "loss": 0.4456,
+      "step": 7757
+    },
+    {
+      "epoch": 3.0314398616942455,
+      "grad_norm": 0.5143163370660467,
+      "learning_rate": 4.743747384809286e-06,
+      "loss": 0.4369,
+      "step": 7758
+    },
+    {
+      "epoch": 3.0318350209928377,
+      "grad_norm": 0.48054033529354206,
+      "learning_rate": 4.743678285327409e-06,
+      "loss": 0.4414,
+      "step": 7759
+    },
+    {
+      "epoch": 3.03223018029143,
+      "grad_norm": 0.47545118742637443,
+      "learning_rate": 4.743609177033725e-06,
+      "loss": 0.4391,
+      "step": 7760
+    },
+    {
+      "epoch": 3.0326253395900222,
+      "grad_norm": 0.500071350352214,
+      "learning_rate": 4.7435400599285075e-06,
+      "loss": 0.4591,
+      "step": 7761
+    },
+    {
+      "epoch": 3.0330204988886145,
+      "grad_norm": 0.49736275171993133,
+      "learning_rate": 4.743470934012026e-06,
+      "loss": 0.4451,
+      "step": 7762
+    },
+    {
+      "epoch": 3.0334156581872067,
+      "grad_norm": 0.48902933971503654,
+      "learning_rate": 4.7434017992845536e-06,
+      "loss": 0.444,
+      "step": 7763
+    },
+    {
+      "epoch": 3.033810817485799,
+      "grad_norm": 0.5017317318287491,
+      "learning_rate": 4.743332655746362e-06,
+      "loss": 0.4433,
+      "step": 7764
+    },
+    {
+      "epoch": 3.034205976784391,
+      "grad_norm": 0.49852541132644085,
+      "learning_rate": 4.743263503397721e-06,
+      "loss": 0.443,
+      "step": 7765
+    },
+    {
+      "epoch": 3.0346011360829834,
+      "grad_norm": 0.49514495883125453,
+      "learning_rate": 4.743194342238904e-06,
+      "loss": 0.4434,
+      "step": 7766
+    },
+    {
+      "epoch": 3.0349962953815757,
+      "grad_norm": 0.5065607691043884,
+      "learning_rate": 4.743125172270181e-06,
+      "loss": 0.4557,
+      "step": 7767
+    },
+    {
+      "epoch": 3.035391454680168,
+      "grad_norm": 0.500892295601084,
+      "learning_rate": 4.743055993491824e-06,
+      "loss": 0.438,
+      "step": 7768
+    },
+    {
+      "epoch": 3.03578661397876,
+      "grad_norm": 0.4840852701954542,
+      "learning_rate": 4.7429868059041065e-06,
+      "loss": 0.4334,
+      "step": 7769
+    },
+    {
+      "epoch": 3.0361817732773524,
+      "grad_norm": 0.5112220036213672,
+      "learning_rate": 4.742917609507298e-06,
+      "loss": 0.4411,
+      "step": 7770
+    },
+    {
+      "epoch": 3.0365769325759446,
+      "grad_norm": 0.48860572596776,
+      "learning_rate": 4.742848404301671e-06,
+      "loss": 0.4502,
+      "step": 7771
+    },
+    {
+      "epoch": 3.036972091874537,
+      "grad_norm": 0.4819193894547075,
+      "learning_rate": 4.742779190287497e-06,
+      "loss": 0.4429,
+      "step": 7772
+    },
+    {
+      "epoch": 3.037367251173129,
+      "grad_norm": 0.4943547849084416,
+      "learning_rate": 4.742709967465049e-06,
+      "loss": 0.4421,
+      "step": 7773
+    },
+    {
+      "epoch": 3.0377624104717214,
+      "grad_norm": 0.4917250752903298,
+      "learning_rate": 4.742640735834599e-06,
+      "loss": 0.4425,
+      "step": 7774
+    },
+    {
+      "epoch": 3.0381575697703136,
+      "grad_norm": 0.46606336140539156,
+      "learning_rate": 4.742571495396415e-06,
+      "loss": 0.4378,
+      "step": 7775
+    },
+    {
+      "epoch": 3.038552729068906,
+      "grad_norm": 0.47545852725279386,
+      "learning_rate": 4.742502246150775e-06,
+      "loss": 0.4333,
+      "step": 7776
+    },
+    {
+      "epoch": 3.038947888367498,
+      "grad_norm": 0.48667928367880847,
+      "learning_rate": 4.742432988097946e-06,
+      "loss": 0.4436,
+      "step": 7777
+    },
+    {
+      "epoch": 3.0393430476660903,
+      "grad_norm": 0.4984876343406036,
+      "learning_rate": 4.742363721238203e-06,
+      "loss": 0.4413,
+      "step": 7778
+    },
+    {
+      "epoch": 3.0397382069646826,
+      "grad_norm": 0.4810472084691766,
+      "learning_rate": 4.742294445571817e-06,
+      "loss": 0.4297,
+      "step": 7779
+    },
+    {
+      "epoch": 3.040133366263275,
+      "grad_norm": 0.49611193782024826,
+      "learning_rate": 4.742225161099059e-06,
+      "loss": 0.4536,
+      "step": 7780
+    },
+    {
+      "epoch": 3.040528525561867,
+      "grad_norm": 0.5043827439085407,
+      "learning_rate": 4.7421558678202025e-06,
+      "loss": 0.4358,
+      "step": 7781
+    },
+    {
+      "epoch": 3.0409236848604593,
+      "grad_norm": 0.48913922839386753,
+      "learning_rate": 4.7420865657355195e-06,
+      "loss": 0.4348,
+      "step": 7782
+    },
+    {
+      "epoch": 3.0413188441590515,
+      "grad_norm": 0.5045463746963759,
+      "learning_rate": 4.742017254845282e-06,
+      "loss": 0.4381,
+      "step": 7783
+    },
+    {
+      "epoch": 3.0417140034576438,
+      "grad_norm": 0.4955607123203069,
+      "learning_rate": 4.741947935149762e-06,
+      "loss": 0.439,
+      "step": 7784
+    },
+    {
+      "epoch": 3.042109162756236,
+      "grad_norm": 0.49656116436176884,
+      "learning_rate": 4.741878606649232e-06,
+      "loss": 0.4373,
+      "step": 7785
+    },
+    {
+      "epoch": 3.0425043220548282,
+      "grad_norm": 0.48383805086599924,
+      "learning_rate": 4.741809269343964e-06,
+      "loss": 0.4417,
+      "step": 7786
+    },
+    {
+      "epoch": 3.0428994813534205,
+      "grad_norm": 0.48644485198914195,
+      "learning_rate": 4.741739923234231e-06,
+      "loss": 0.4348,
+      "step": 7787
+    },
+    {
+      "epoch": 3.0432946406520127,
+      "grad_norm": 0.5051982041050768,
+      "learning_rate": 4.741670568320304e-06,
+      "loss": 0.4462,
+      "step": 7788
+    },
+    {
+      "epoch": 3.043689799950605,
+      "grad_norm": 0.5048908640256319,
+      "learning_rate": 4.741601204602457e-06,
+      "loss": 0.4518,
+      "step": 7789
+    },
+    {
+      "epoch": 3.044084959249197,
+      "grad_norm": 0.4944058502455239,
+      "learning_rate": 4.741531832080961e-06,
+      "loss": 0.4481,
+      "step": 7790
+    },
+    {
+      "epoch": 3.0444801185477894,
+      "grad_norm": 0.5375482439420921,
+      "learning_rate": 4.741462450756089e-06,
+      "loss": 0.4556,
+      "step": 7791
+    },
+    {
+      "epoch": 3.0448752778463817,
+      "grad_norm": 0.483762343858191,
+      "learning_rate": 4.741393060628115e-06,
+      "loss": 0.4258,
+      "step": 7792
+    },
+    {
+      "epoch": 3.045270437144974,
+      "grad_norm": 0.47995611425569656,
+      "learning_rate": 4.741323661697308e-06,
+      "loss": 0.4364,
+      "step": 7793
+    },
+    {
+      "epoch": 3.045665596443566,
+      "grad_norm": 0.47885796380315865,
+      "learning_rate": 4.741254253963944e-06,
+      "loss": 0.4335,
+      "step": 7794
+    },
+    {
+      "epoch": 3.0460607557421584,
+      "grad_norm": 0.5094045890304211,
+      "learning_rate": 4.741184837428294e-06,
+      "loss": 0.4501,
+      "step": 7795
+    },
+    {
+      "epoch": 3.0464559150407506,
+      "grad_norm": 0.4896089467966352,
+      "learning_rate": 4.741115412090631e-06,
+      "loss": 0.448,
+      "step": 7796
+    },
+    {
+      "epoch": 3.046851074339343,
+      "grad_norm": 0.4870799703211307,
+      "learning_rate": 4.7410459779512276e-06,
+      "loss": 0.4455,
+      "step": 7797
+    },
+    {
+      "epoch": 3.047246233637935,
+      "grad_norm": 0.5102806429622584,
+      "learning_rate": 4.740976535010355e-06,
+      "loss": 0.431,
+      "step": 7798
+    },
+    {
+      "epoch": 3.0476413929365274,
+      "grad_norm": 0.4899387354627377,
+      "learning_rate": 4.740907083268289e-06,
+      "loss": 0.4552,
+      "step": 7799
+    },
+    {
+      "epoch": 3.0480365522351196,
+      "grad_norm": 0.4979089977106902,
+      "learning_rate": 4.740837622725301e-06,
+      "loss": 0.4383,
+      "step": 7800
+    },
+    {
+      "epoch": 3.048431711533712,
+      "grad_norm": 0.49833292526111833,
+      "learning_rate": 4.7407681533816624e-06,
+      "loss": 0.4373,
+      "step": 7801
+    },
+    {
+      "epoch": 3.048826870832304,
+      "grad_norm": 0.49821399658843996,
+      "learning_rate": 4.7406986752376475e-06,
+      "loss": 0.4479,
+      "step": 7802
+    },
+    {
+      "epoch": 3.0492220301308963,
+      "grad_norm": 0.4744560438026494,
+      "learning_rate": 4.740629188293529e-06,
+      "loss": 0.4276,
+      "step": 7803
+    },
+    {
+      "epoch": 3.0496171894294886,
+      "grad_norm": 0.47806331822046966,
+      "learning_rate": 4.740559692549579e-06,
+      "loss": 0.4253,
+      "step": 7804
+    },
+    {
+      "epoch": 3.050012348728081,
+      "grad_norm": 0.48603559179075806,
+      "learning_rate": 4.7404901880060725e-06,
+      "loss": 0.4541,
+      "step": 7805
+    },
+    {
+      "epoch": 3.050407508026673,
+      "grad_norm": 0.4910984242297495,
+      "learning_rate": 4.74042067466328e-06,
+      "loss": 0.4299,
+      "step": 7806
+    },
+    {
+      "epoch": 3.0508026673252653,
+      "grad_norm": 0.4851618865150965,
+      "learning_rate": 4.740351152521475e-06,
+      "loss": 0.4374,
+      "step": 7807
+    },
+    {
+      "epoch": 3.0511978266238575,
+      "grad_norm": 0.4889763206502265,
+      "learning_rate": 4.740281621580932e-06,
+      "loss": 0.436,
+      "step": 7808
+    },
+    {
+      "epoch": 3.0515929859224498,
+      "grad_norm": 0.5012899815645088,
+      "learning_rate": 4.740212081841924e-06,
+      "loss": 0.4457,
+      "step": 7809
+    },
+    {
+      "epoch": 3.0519881452210424,
+      "grad_norm": 0.5082796466392632,
+      "learning_rate": 4.7401425333047215e-06,
+      "loss": 0.4409,
+      "step": 7810
+    },
+    {
+      "epoch": 3.0523833045196347,
+      "grad_norm": 0.4982145445419337,
+      "learning_rate": 4.7400729759696e-06,
+      "loss": 0.4436,
+      "step": 7811
+    },
+    {
+      "epoch": 3.052778463818227,
+      "grad_norm": 0.4956899805603572,
+      "learning_rate": 4.7400034098368325e-06,
+      "loss": 0.4269,
+      "step": 7812
+    },
+    {
+      "epoch": 3.053173623116819,
+      "grad_norm": 0.48480193413501993,
+      "learning_rate": 4.739933834906692e-06,
+      "loss": 0.4479,
+      "step": 7813
+    },
+    {
+      "epoch": 3.0535687824154114,
+      "grad_norm": 0.4806050679449309,
+      "learning_rate": 4.73986425117945e-06,
+      "loss": 0.4367,
+      "step": 7814
+    },
+    {
+      "epoch": 3.0539639417140036,
+      "grad_norm": 0.4855626404004344,
+      "learning_rate": 4.739794658655383e-06,
+      "loss": 0.4482,
+      "step": 7815
+    },
+    {
+      "epoch": 3.054359101012596,
+      "grad_norm": 0.5103646887287417,
+      "learning_rate": 4.739725057334762e-06,
+      "loss": 0.4485,
+      "step": 7816
+    },
+    {
+      "epoch": 3.054754260311188,
+      "grad_norm": 0.48745678622976013,
+      "learning_rate": 4.7396554472178615e-06,
+      "loss": 0.4548,
+      "step": 7817
+    },
+    {
+      "epoch": 3.0551494196097804,
+      "grad_norm": 0.512946784611527,
+      "learning_rate": 4.739585828304953e-06,
+      "loss": 0.4438,
+      "step": 7818
+    },
+    {
+      "epoch": 3.0555445789083726,
+      "grad_norm": 0.49787233339609754,
+      "learning_rate": 4.739516200596313e-06,
+      "loss": 0.4412,
+      "step": 7819
+    },
+    {
+      "epoch": 3.055939738206965,
+      "grad_norm": 0.4923640029971268,
+      "learning_rate": 4.739446564092213e-06,
+      "loss": 0.4266,
+      "step": 7820
+    },
+    {
+      "epoch": 3.056334897505557,
+      "grad_norm": 0.4922770070573321,
+      "learning_rate": 4.739376918792926e-06,
+      "loss": 0.4234,
+      "step": 7821
+    },
+    {
+      "epoch": 3.0567300568041493,
+      "grad_norm": 0.48256515830579066,
+      "learning_rate": 4.7393072646987266e-06,
+      "loss": 0.4283,
+      "step": 7822
+    },
+    {
+      "epoch": 3.0571252161027416,
+      "grad_norm": 0.4871301212135088,
+      "learning_rate": 4.739237601809889e-06,
+      "loss": 0.4347,
+      "step": 7823
+    },
+    {
+      "epoch": 3.057520375401334,
+      "grad_norm": 0.48509606892345153,
+      "learning_rate": 4.739167930126684e-06,
+      "loss": 0.4345,
+      "step": 7824
+    },
+    {
+      "epoch": 3.057915534699926,
+      "grad_norm": 0.6165399120475508,
+      "learning_rate": 4.739098249649388e-06,
+      "loss": 0.4507,
+      "step": 7825
+    },
+    {
+      "epoch": 3.0583106939985183,
+      "grad_norm": 0.4921339621782017,
+      "learning_rate": 4.739028560378274e-06,
+      "loss": 0.4406,
+      "step": 7826
+    },
+    {
+      "epoch": 3.0587058532971105,
+      "grad_norm": 0.5054560093119399,
+      "learning_rate": 4.738958862313615e-06,
+      "loss": 0.4542,
+      "step": 7827
+    },
+    {
+      "epoch": 3.0591010125957028,
+      "grad_norm": 0.4886778814315178,
+      "learning_rate": 4.7388891554556845e-06,
+      "loss": 0.4393,
+      "step": 7828
+    },
+    {
+      "epoch": 3.059496171894295,
+      "grad_norm": 0.4860239144323546,
+      "learning_rate": 4.7388194398047585e-06,
+      "loss": 0.4282,
+      "step": 7829
+    },
+    {
+      "epoch": 3.0598913311928873,
+      "grad_norm": 0.4924325835486165,
+      "learning_rate": 4.738749715361108e-06,
+      "loss": 0.4533,
+      "step": 7830
+    },
+    {
+      "epoch": 3.0602864904914795,
+      "grad_norm": 0.48932648303037984,
+      "learning_rate": 4.738679982125008e-06,
+      "loss": 0.4453,
+      "step": 7831
+    },
+    {
+      "epoch": 3.0606816497900717,
+      "grad_norm": 0.48237789375682943,
+      "learning_rate": 4.738610240096733e-06,
+      "loss": 0.4447,
+      "step": 7832
+    },
+    {
+      "epoch": 3.061076809088664,
+      "grad_norm": 0.506439179071494,
+      "learning_rate": 4.7385404892765565e-06,
+      "loss": 0.4499,
+      "step": 7833
+    },
+    {
+      "epoch": 3.061471968387256,
+      "grad_norm": 0.4977813662126531,
+      "learning_rate": 4.738470729664753e-06,
+      "loss": 0.4369,
+      "step": 7834
+    },
+    {
+      "epoch": 3.0618671276858485,
+      "grad_norm": 0.5122710738027177,
+      "learning_rate": 4.738400961261594e-06,
+      "loss": 0.4395,
+      "step": 7835
+    },
+    {
+      "epoch": 3.0622622869844407,
+      "grad_norm": 0.4846403050899975,
+      "learning_rate": 4.7383311840673565e-06,
+      "loss": 0.4371,
+      "step": 7836
+    },
+    {
+      "epoch": 3.062657446283033,
+      "grad_norm": 0.5032833387552864,
+      "learning_rate": 4.738261398082313e-06,
+      "loss": 0.4389,
+      "step": 7837
+    },
+    {
+      "epoch": 3.063052605581625,
+      "grad_norm": 0.4923216050566678,
+      "learning_rate": 4.738191603306738e-06,
+      "loss": 0.4339,
+      "step": 7838
+    },
+    {
+      "epoch": 3.0634477648802174,
+      "grad_norm": 0.4909668236344491,
+      "learning_rate": 4.738121799740904e-06,
+      "loss": 0.4379,
+      "step": 7839
+    },
+    {
+      "epoch": 3.0638429241788097,
+      "grad_norm": 0.5040118071240738,
+      "learning_rate": 4.738051987385088e-06,
+      "loss": 0.44,
+      "step": 7840
+    },
+    {
+      "epoch": 3.064238083477402,
+      "grad_norm": 0.5125484179414564,
+      "learning_rate": 4.737982166239563e-06,
+      "loss": 0.4341,
+      "step": 7841
+    },
+    {
+      "epoch": 3.064633242775994,
+      "grad_norm": 0.5251112266930515,
+      "learning_rate": 4.737912336304602e-06,
+      "loss": 0.4307,
+      "step": 7842
+    },
+    {
+      "epoch": 3.0650284020745864,
+      "grad_norm": 0.481537716871614,
+      "learning_rate": 4.737842497580482e-06,
+      "loss": 0.4342,
+      "step": 7843
+    },
+    {
+      "epoch": 3.0654235613731786,
+      "grad_norm": 0.48948821736693127,
+      "learning_rate": 4.737772650067474e-06,
+      "loss": 0.439,
+      "step": 7844
+    },
+    {
+      "epoch": 3.065818720671771,
+      "grad_norm": 0.5023038703887029,
+      "learning_rate": 4.737702793765855e-06,
+      "loss": 0.4525,
+      "step": 7845
+    },
+    {
+      "epoch": 3.066213879970363,
+      "grad_norm": 0.494523096453894,
+      "learning_rate": 4.737632928675897e-06,
+      "loss": 0.4439,
+      "step": 7846
+    },
+    {
+      "epoch": 3.0666090392689553,
+      "grad_norm": 0.48065114804411224,
+      "learning_rate": 4.7375630547978764e-06,
+      "loss": 0.4238,
+      "step": 7847
+    },
+    {
+      "epoch": 3.0670041985675476,
+      "grad_norm": 0.49790523236786666,
+      "learning_rate": 4.737493172132067e-06,
+      "loss": 0.4309,
+      "step": 7848
+    },
+    {
+      "epoch": 3.06739935786614,
+      "grad_norm": 0.4989996884805779,
+      "learning_rate": 4.737423280678742e-06,
+      "loss": 0.4512,
+      "step": 7849
+    },
+    {
+      "epoch": 3.067794517164732,
+      "grad_norm": 0.5016208814260822,
+      "learning_rate": 4.737353380438178e-06,
+      "loss": 0.4539,
+      "step": 7850
+    },
+    {
+      "epoch": 3.0681896764633243,
+      "grad_norm": 0.5134079709987914,
+      "learning_rate": 4.737283471410649e-06,
+      "loss": 0.457,
+      "step": 7851
+    },
+    {
+      "epoch": 3.0685848357619165,
+      "grad_norm": 0.5067457933406915,
+      "learning_rate": 4.737213553596428e-06,
+      "loss": 0.4481,
+      "step": 7852
+    },
+    {
+      "epoch": 3.068979995060509,
+      "grad_norm": 0.4776716218044217,
+      "learning_rate": 4.73714362699579e-06,
+      "loss": 0.4383,
+      "step": 7853
+    },
+    {
+      "epoch": 3.069375154359101,
+      "grad_norm": 0.48957989485980835,
+      "learning_rate": 4.737073691609012e-06,
+      "loss": 0.4357,
+      "step": 7854
+    },
+    {
+      "epoch": 3.0697703136576933,
+      "grad_norm": 0.47676778659524477,
+      "learning_rate": 4.737003747436366e-06,
+      "loss": 0.4386,
+      "step": 7855
+    },
+    {
+      "epoch": 3.0701654729562855,
+      "grad_norm": 0.6191504502491905,
+      "learning_rate": 4.736933794478128e-06,
+      "loss": 0.4543,
+      "step": 7856
+    },
+    {
+      "epoch": 3.0705606322548777,
+      "grad_norm": 0.4942925858350992,
+      "learning_rate": 4.736863832734573e-06,
+      "loss": 0.4353,
+      "step": 7857
+    },
+    {
+      "epoch": 3.07095579155347,
+      "grad_norm": 0.5216216205160417,
+      "learning_rate": 4.736793862205974e-06,
+      "loss": 0.4269,
+      "step": 7858
+    },
+    {
+      "epoch": 3.0713509508520622,
+      "grad_norm": 0.4838627707703042,
+      "learning_rate": 4.7367238828926075e-06,
+      "loss": 0.4415,
+      "step": 7859
+    },
+    {
+      "epoch": 3.0717461101506545,
+      "grad_norm": 0.48668009920048644,
+      "learning_rate": 4.736653894794748e-06,
+      "loss": 0.4303,
+      "step": 7860
+    },
+    {
+      "epoch": 3.0721412694492467,
+      "grad_norm": 0.5167979871026906,
+      "learning_rate": 4.7365838979126696e-06,
+      "loss": 0.4665,
+      "step": 7861
+    },
+    {
+      "epoch": 3.072536428747839,
+      "grad_norm": 0.5067480201831851,
+      "learning_rate": 4.736513892246648e-06,
+      "loss": 0.4592,
+      "step": 7862
+    },
+    {
+      "epoch": 3.072931588046431,
+      "grad_norm": 0.503091144854992,
+      "learning_rate": 4.736443877796959e-06,
+      "loss": 0.4509,
+      "step": 7863
+    },
+    {
+      "epoch": 3.0733267473450234,
+      "grad_norm": 0.4966903825101766,
+      "learning_rate": 4.736373854563875e-06,
+      "loss": 0.4393,
+      "step": 7864
+    },
+    {
+      "epoch": 3.0737219066436157,
+      "grad_norm": 0.5053933985734216,
+      "learning_rate": 4.736303822547673e-06,
+      "loss": 0.4505,
+      "step": 7865
+    },
+    {
+      "epoch": 3.074117065942208,
+      "grad_norm": 0.47840381289455225,
+      "learning_rate": 4.736233781748627e-06,
+      "loss": 0.4346,
+      "step": 7866
+    },
+    {
+      "epoch": 3.0745122252408,
+      "grad_norm": 0.5006357088048857,
+      "learning_rate": 4.736163732167014e-06,
+      "loss": 0.4538,
+      "step": 7867
+    },
+    {
+      "epoch": 3.0749073845393924,
+      "grad_norm": 0.47835652192235645,
+      "learning_rate": 4.736093673803106e-06,
+      "loss": 0.4365,
+      "step": 7868
+    },
+    {
+      "epoch": 3.0753025438379846,
+      "grad_norm": 0.5171050130456775,
+      "learning_rate": 4.736023606657181e-06,
+      "loss": 0.4422,
+      "step": 7869
+    },
+    {
+      "epoch": 3.075697703136577,
+      "grad_norm": 0.4937403544438971,
+      "learning_rate": 4.735953530729514e-06,
+      "loss": 0.4316,
+      "step": 7870
+    },
+    {
+      "epoch": 3.076092862435169,
+      "grad_norm": 0.4959244380471904,
+      "learning_rate": 4.735883446020377e-06,
+      "loss": 0.4553,
+      "step": 7871
+    },
+    {
+      "epoch": 3.0764880217337613,
+      "grad_norm": 0.5016759159070576,
+      "learning_rate": 4.7358133525300484e-06,
+      "loss": 0.4445,
+      "step": 7872
+    },
+    {
+      "epoch": 3.0768831810323536,
+      "grad_norm": 0.48705287229754113,
+      "learning_rate": 4.735743250258803e-06,
+      "loss": 0.4469,
+      "step": 7873
+    },
+    {
+      "epoch": 3.077278340330946,
+      "grad_norm": 0.5092392088990833,
+      "learning_rate": 4.735673139206915e-06,
+      "loss": 0.4592,
+      "step": 7874
+    },
+    {
+      "epoch": 3.077673499629538,
+      "grad_norm": 0.47669040144749475,
+      "learning_rate": 4.735603019374661e-06,
+      "loss": 0.4275,
+      "step": 7875
+    },
+    {
+      "epoch": 3.0780686589281303,
+      "grad_norm": 0.4924439478857504,
+      "learning_rate": 4.735532890762316e-06,
+      "loss": 0.4395,
+      "step": 7876
+    },
+    {
+      "epoch": 3.0784638182267225,
+      "grad_norm": 0.5123885580239934,
+      "learning_rate": 4.735462753370156e-06,
+      "loss": 0.4501,
+      "step": 7877
+    },
+    {
+      "epoch": 3.078858977525315,
+      "grad_norm": 0.4823099934221924,
+      "learning_rate": 4.735392607198455e-06,
+      "loss": 0.4275,
+      "step": 7878
+    },
+    {
+      "epoch": 3.079254136823907,
+      "grad_norm": 0.4857701642431153,
+      "learning_rate": 4.735322452247489e-06,
+      "loss": 0.4373,
+      "step": 7879
+    },
+    {
+      "epoch": 3.0796492961224993,
+      "grad_norm": 0.49163839390605096,
+      "learning_rate": 4.7352522885175345e-06,
+      "loss": 0.4455,
+      "step": 7880
+    },
+    {
+      "epoch": 3.0800444554210915,
+      "grad_norm": 0.4884031270257798,
+      "learning_rate": 4.735182116008866e-06,
+      "loss": 0.4582,
+      "step": 7881
+    },
+    {
+      "epoch": 3.0804396147196837,
+      "grad_norm": 0.4899553724792646,
+      "learning_rate": 4.7351119347217585e-06,
+      "loss": 0.4355,
+      "step": 7882
+    },
+    {
+      "epoch": 3.080834774018276,
+      "grad_norm": 0.4949749498592674,
+      "learning_rate": 4.73504174465649e-06,
+      "loss": 0.4525,
+      "step": 7883
+    },
+    {
+      "epoch": 3.0812299333168682,
+      "grad_norm": 0.4988182734709485,
+      "learning_rate": 4.734971545813334e-06,
+      "loss": 0.4498,
+      "step": 7884
+    },
+    {
+      "epoch": 3.0816250926154605,
+      "grad_norm": 0.49792382572578914,
+      "learning_rate": 4.734901338192567e-06,
+      "loss": 0.4562,
+      "step": 7885
+    },
+    {
+      "epoch": 3.0820202519140527,
+      "grad_norm": 0.4915536561332589,
+      "learning_rate": 4.734831121794464e-06,
+      "loss": 0.4463,
+      "step": 7886
+    },
+    {
+      "epoch": 3.082415411212645,
+      "grad_norm": 0.4970561396674575,
+      "learning_rate": 4.734760896619302e-06,
+      "loss": 0.4476,
+      "step": 7887
+    },
+    {
+      "epoch": 3.082810570511237,
+      "grad_norm": 0.4997419639637845,
+      "learning_rate": 4.734690662667356e-06,
+      "loss": 0.4337,
+      "step": 7888
+    },
+    {
+      "epoch": 3.0832057298098294,
+      "grad_norm": 0.48823979322917044,
+      "learning_rate": 4.734620419938902e-06,
+      "loss": 0.4684,
+      "step": 7889
+    },
+    {
+      "epoch": 3.0836008891084217,
+      "grad_norm": 0.5036100358314696,
+      "learning_rate": 4.734550168434216e-06,
+      "loss": 0.4457,
+      "step": 7890
+    },
+    {
+      "epoch": 3.083996048407014,
+      "grad_norm": 0.4941812609213103,
+      "learning_rate": 4.734479908153574e-06,
+      "loss": 0.4448,
+      "step": 7891
+    },
+    {
+      "epoch": 3.084391207705606,
+      "grad_norm": 0.5061906057278033,
+      "learning_rate": 4.734409639097253e-06,
+      "loss": 0.4486,
+      "step": 7892
+    },
+    {
+      "epoch": 3.0847863670041984,
+      "grad_norm": 0.5100975450233716,
+      "learning_rate": 4.734339361265526e-06,
+      "loss": 0.4425,
+      "step": 7893
+    },
+    {
+      "epoch": 3.0851815263027906,
+      "grad_norm": 0.49861233440246766,
+      "learning_rate": 4.7342690746586714e-06,
+      "loss": 0.4356,
+      "step": 7894
+    },
+    {
+      "epoch": 3.085576685601383,
+      "grad_norm": 0.574904471137688,
+      "learning_rate": 4.734198779276964e-06,
+      "loss": 0.4409,
+      "step": 7895
+    },
+    {
+      "epoch": 3.085971844899975,
+      "grad_norm": 0.4873731771876203,
+      "learning_rate": 4.734128475120681e-06,
+      "loss": 0.4405,
+      "step": 7896
+    },
+    {
+      "epoch": 3.0863670041985674,
+      "grad_norm": 0.4951592433754978,
+      "learning_rate": 4.7340581621900985e-06,
+      "loss": 0.4501,
+      "step": 7897
+    },
+    {
+      "epoch": 3.0867621634971596,
+      "grad_norm": 0.4766631688452397,
+      "learning_rate": 4.733987840485491e-06,
+      "loss": 0.4333,
+      "step": 7898
+    },
+    {
+      "epoch": 3.087157322795752,
+      "grad_norm": 0.4923906837500623,
+      "learning_rate": 4.733917510007137e-06,
+      "loss": 0.4383,
+      "step": 7899
+    },
+    {
+      "epoch": 3.087552482094344,
+      "grad_norm": 0.48241702014723586,
+      "learning_rate": 4.73384717075531e-06,
+      "loss": 0.4313,
+      "step": 7900
+    },
+    {
+      "epoch": 3.0879476413929363,
+      "grad_norm": 0.48868586859443763,
+      "learning_rate": 4.733776822730289e-06,
+      "loss": 0.4409,
+      "step": 7901
+    },
+    {
+      "epoch": 3.088342800691529,
+      "grad_norm": 0.5123299558820024,
+      "learning_rate": 4.733706465932349e-06,
+      "loss": 0.4523,
+      "step": 7902
+    },
+    {
+      "epoch": 3.0887379599901212,
+      "grad_norm": 0.4933673322520625,
+      "learning_rate": 4.733636100361766e-06,
+      "loss": 0.4312,
+      "step": 7903
+    },
+    {
+      "epoch": 3.0891331192887135,
+      "grad_norm": 0.5081997308648453,
+      "learning_rate": 4.733565726018817e-06,
+      "loss": 0.4357,
+      "step": 7904
+    },
+    {
+      "epoch": 3.0895282785873057,
+      "grad_norm": 0.4975799368385473,
+      "learning_rate": 4.733495342903778e-06,
+      "loss": 0.4392,
+      "step": 7905
+    },
+    {
+      "epoch": 3.089923437885898,
+      "grad_norm": 0.5025137130985103,
+      "learning_rate": 4.733424951016925e-06,
+      "loss": 0.4734,
+      "step": 7906
+    },
+    {
+      "epoch": 3.09031859718449,
+      "grad_norm": 0.4977488175268691,
+      "learning_rate": 4.733354550358536e-06,
+      "loss": 0.4412,
+      "step": 7907
+    },
+    {
+      "epoch": 3.0907137564830824,
+      "grad_norm": 0.48503718006898866,
+      "learning_rate": 4.733284140928886e-06,
+      "loss": 0.4577,
+      "step": 7908
+    },
+    {
+      "epoch": 3.0911089157816747,
+      "grad_norm": 0.49140511027996614,
+      "learning_rate": 4.733213722728251e-06,
+      "loss": 0.4249,
+      "step": 7909
+    },
+    {
+      "epoch": 3.091504075080267,
+      "grad_norm": 0.4854895181626003,
+      "learning_rate": 4.73314329575691e-06,
+      "loss": 0.4417,
+      "step": 7910
+    },
+    {
+      "epoch": 3.091899234378859,
+      "grad_norm": 0.49472508088997874,
+      "learning_rate": 4.733072860015138e-06,
+      "loss": 0.4447,
+      "step": 7911
+    },
+    {
+      "epoch": 3.0922943936774514,
+      "grad_norm": 0.5209744346067812,
+      "learning_rate": 4.7330024155032115e-06,
+      "loss": 0.4517,
+      "step": 7912
+    },
+    {
+      "epoch": 3.0926895529760436,
+      "grad_norm": 0.5376195020380707,
+      "learning_rate": 4.732931962221407e-06,
+      "loss": 0.4473,
+      "step": 7913
+    },
+    {
+      "epoch": 3.093084712274636,
+      "grad_norm": 0.48849504260879484,
+      "learning_rate": 4.732861500170003e-06,
+      "loss": 0.4362,
+      "step": 7914
+    },
+    {
+      "epoch": 3.093479871573228,
+      "grad_norm": 0.5121173737122551,
+      "learning_rate": 4.732791029349274e-06,
+      "loss": 0.4364,
+      "step": 7915
+    },
+    {
+      "epoch": 3.0938750308718204,
+      "grad_norm": 0.5013991610345373,
+      "learning_rate": 4.7327205497594975e-06,
+      "loss": 0.4413,
+      "step": 7916
+    },
+    {
+      "epoch": 3.0942701901704126,
+      "grad_norm": 0.5132348413742815,
+      "learning_rate": 4.732650061400951e-06,
+      "loss": 0.4609,
+      "step": 7917
+    },
+    {
+      "epoch": 3.094665349469005,
+      "grad_norm": 0.5140445000952766,
+      "learning_rate": 4.73257956427391e-06,
+      "loss": 0.4384,
+      "step": 7918
+    },
+    {
+      "epoch": 3.095060508767597,
+      "grad_norm": 0.4805026819173038,
+      "learning_rate": 4.732509058378653e-06,
+      "loss": 0.4353,
+      "step": 7919
+    },
+    {
+      "epoch": 3.0954556680661893,
+      "grad_norm": 0.4998603496611781,
+      "learning_rate": 4.732438543715456e-06,
+      "loss": 0.4513,
+      "step": 7920
+    },
+    {
+      "epoch": 3.0958508273647816,
+      "grad_norm": 0.5345924785899526,
+      "learning_rate": 4.732368020284596e-06,
+      "loss": 0.4452,
+      "step": 7921
+    },
+    {
+      "epoch": 3.096245986663374,
+      "grad_norm": 0.5074168698408299,
+      "learning_rate": 4.732297488086349e-06,
+      "loss": 0.4556,
+      "step": 7922
+    },
+    {
+      "epoch": 3.096641145961966,
+      "grad_norm": 0.4959304495710035,
+      "learning_rate": 4.732226947120995e-06,
+      "loss": 0.4633,
+      "step": 7923
+    },
+    {
+      "epoch": 3.0970363052605583,
+      "grad_norm": 0.5729612706103528,
+      "learning_rate": 4.732156397388807e-06,
+      "loss": 0.4323,
+      "step": 7924
+    },
+    {
+      "epoch": 3.0974314645591505,
+      "grad_norm": 0.48429647430521955,
+      "learning_rate": 4.732085838890064e-06,
+      "loss": 0.4401,
+      "step": 7925
+    },
+    {
+      "epoch": 3.0978266238577428,
+      "grad_norm": 0.49722917342119355,
+      "learning_rate": 4.732015271625045e-06,
+      "loss": 0.4468,
+      "step": 7926
+    },
+    {
+      "epoch": 3.098221783156335,
+      "grad_norm": 0.5045870249115458,
+      "learning_rate": 4.731944695594024e-06,
+      "loss": 0.4511,
+      "step": 7927
+    },
+    {
+      "epoch": 3.0986169424549272,
+      "grad_norm": 0.5022592905161292,
+      "learning_rate": 4.731874110797281e-06,
+      "loss": 0.4506,
+      "step": 7928
+    },
+    {
+      "epoch": 3.0990121017535195,
+      "grad_norm": 0.5111603683659854,
+      "learning_rate": 4.73180351723509e-06,
+      "loss": 0.4524,
+      "step": 7929
+    },
+    {
+      "epoch": 3.0994072610521117,
+      "grad_norm": 0.5267537367598453,
+      "learning_rate": 4.731732914907731e-06,
+      "loss": 0.4576,
+      "step": 7930
+    },
+    {
+      "epoch": 3.099802420350704,
+      "grad_norm": 0.5723867018529993,
+      "learning_rate": 4.731662303815479e-06,
+      "loss": 0.4464,
+      "step": 7931
+    },
+    {
+      "epoch": 3.100197579649296,
+      "grad_norm": 0.5002231014728444,
+      "learning_rate": 4.7315916839586144e-06,
+      "loss": 0.4518,
+      "step": 7932
+    },
+    {
+      "epoch": 3.1005927389478884,
+      "grad_norm": 0.49456541023282674,
+      "learning_rate": 4.731521055337412e-06,
+      "loss": 0.4395,
+      "step": 7933
+    },
+    {
+      "epoch": 3.1009878982464807,
+      "grad_norm": 0.48028211739122306,
+      "learning_rate": 4.73145041795215e-06,
+      "loss": 0.4491,
+      "step": 7934
+    },
+    {
+      "epoch": 3.101383057545073,
+      "grad_norm": 0.49265412815653176,
+      "learning_rate": 4.731379771803106e-06,
+      "loss": 0.4479,
+      "step": 7935
+    },
+    {
+      "epoch": 3.101778216843665,
+      "grad_norm": 0.488033887759252,
+      "learning_rate": 4.731309116890556e-06,
+      "loss": 0.433,
+      "step": 7936
+    },
+    {
+      "epoch": 3.1021733761422574,
+      "grad_norm": 0.4979653413020749,
+      "learning_rate": 4.731238453214781e-06,
+      "loss": 0.4585,
+      "step": 7937
+    },
+    {
+      "epoch": 3.1025685354408497,
+      "grad_norm": 0.5024424180826177,
+      "learning_rate": 4.731167780776055e-06,
+      "loss": 0.4695,
+      "step": 7938
+    },
+    {
+      "epoch": 3.102963694739442,
+      "grad_norm": 0.49050531548527476,
+      "learning_rate": 4.731097099574656e-06,
+      "loss": 0.4434,
+      "step": 7939
+    },
+    {
+      "epoch": 3.103358854038034,
+      "grad_norm": 0.4993603539723156,
+      "learning_rate": 4.731026409610863e-06,
+      "loss": 0.4562,
+      "step": 7940
+    },
+    {
+      "epoch": 3.1037540133366264,
+      "grad_norm": 0.4852146754915784,
+      "learning_rate": 4.7309557108849535e-06,
+      "loss": 0.4528,
+      "step": 7941
+    },
+    {
+      "epoch": 3.1041491726352186,
+      "grad_norm": 0.47989325470415545,
+      "learning_rate": 4.730885003397204e-06,
+      "loss": 0.4435,
+      "step": 7942
+    },
+    {
+      "epoch": 3.104544331933811,
+      "grad_norm": 0.5012231199309242,
+      "learning_rate": 4.730814287147893e-06,
+      "loss": 0.4349,
+      "step": 7943
+    },
+    {
+      "epoch": 3.104939491232403,
+      "grad_norm": 0.5077310439859108,
+      "learning_rate": 4.730743562137299e-06,
+      "loss": 0.4444,
+      "step": 7944
+    },
+    {
+      "epoch": 3.1053346505309953,
+      "grad_norm": 0.5353226020763748,
+      "learning_rate": 4.7306728283656976e-06,
+      "loss": 0.4208,
+      "step": 7945
+    },
+    {
+      "epoch": 3.1057298098295876,
+      "grad_norm": 0.4789749092687642,
+      "learning_rate": 4.7306020858333685e-06,
+      "loss": 0.4351,
+      "step": 7946
+    },
+    {
+      "epoch": 3.10612496912818,
+      "grad_norm": 0.49123192197023,
+      "learning_rate": 4.730531334540589e-06,
+      "loss": 0.4511,
+      "step": 7947
+    },
+    {
+      "epoch": 3.106520128426772,
+      "grad_norm": 0.503885581742825,
+      "learning_rate": 4.730460574487636e-06,
+      "loss": 0.4593,
+      "step": 7948
+    },
+    {
+      "epoch": 3.1069152877253643,
+      "grad_norm": 0.4910430506949811,
+      "learning_rate": 4.7303898056747895e-06,
+      "loss": 0.4606,
+      "step": 7949
+    },
+    {
+      "epoch": 3.1073104470239565,
+      "grad_norm": 0.4997712623504369,
+      "learning_rate": 4.730319028102326e-06,
+      "loss": 0.4524,
+      "step": 7950
+    },
+    {
+      "epoch": 3.1077056063225488,
+      "grad_norm": 0.4998843657763657,
+      "learning_rate": 4.730248241770523e-06,
+      "loss": 0.4575,
+      "step": 7951
+    },
+    {
+      "epoch": 3.108100765621141,
+      "grad_norm": 0.5030067437860954,
+      "learning_rate": 4.730177446679659e-06,
+      "loss": 0.4403,
+      "step": 7952
+    },
+    {
+      "epoch": 3.1084959249197333,
+      "grad_norm": 0.5010628496154773,
+      "learning_rate": 4.730106642830013e-06,
+      "loss": 0.4491,
+      "step": 7953
+    },
+    {
+      "epoch": 3.1088910842183255,
+      "grad_norm": 0.49750569778277126,
+      "learning_rate": 4.730035830221862e-06,
+      "loss": 0.4488,
+      "step": 7954
+    },
+    {
+      "epoch": 3.1092862435169177,
+      "grad_norm": 0.5133141372969336,
+      "learning_rate": 4.729965008855485e-06,
+      "loss": 0.4502,
+      "step": 7955
+    },
+    {
+      "epoch": 3.10968140281551,
+      "grad_norm": 0.47517558280691385,
+      "learning_rate": 4.729894178731159e-06,
+      "loss": 0.4254,
+      "step": 7956
+    },
+    {
+      "epoch": 3.110076562114102,
+      "grad_norm": 0.5159606553817537,
+      "learning_rate": 4.7298233398491625e-06,
+      "loss": 0.4435,
+      "step": 7957
+    },
+    {
+      "epoch": 3.1104717214126945,
+      "grad_norm": 0.4979703809694994,
+      "learning_rate": 4.729752492209774e-06,
+      "loss": 0.459,
+      "step": 7958
+    },
+    {
+      "epoch": 3.1108668807112867,
+      "grad_norm": 0.49169143247666647,
+      "learning_rate": 4.729681635813272e-06,
+      "loss": 0.454,
+      "step": 7959
+    },
+    {
+      "epoch": 3.111262040009879,
+      "grad_norm": 0.5197648875864174,
+      "learning_rate": 4.729610770659934e-06,
+      "loss": 0.4627,
+      "step": 7960
+    },
+    {
+      "epoch": 3.111657199308471,
+      "grad_norm": 0.5146456770573503,
+      "learning_rate": 4.729539896750039e-06,
+      "loss": 0.4386,
+      "step": 7961
+    },
+    {
+      "epoch": 3.1120523586070634,
+      "grad_norm": 0.5152953570832932,
+      "learning_rate": 4.729469014083865e-06,
+      "loss": 0.4338,
+      "step": 7962
+    },
+    {
+      "epoch": 3.1124475179056557,
+      "grad_norm": 0.5029822022584626,
+      "learning_rate": 4.729398122661692e-06,
+      "loss": 0.4484,
+      "step": 7963
+    },
+    {
+      "epoch": 3.112842677204248,
+      "grad_norm": 0.48762300399924735,
+      "learning_rate": 4.729327222483795e-06,
+      "loss": 0.4572,
+      "step": 7964
+    },
+    {
+      "epoch": 3.11323783650284,
+      "grad_norm": 0.5078757083645274,
+      "learning_rate": 4.7292563135504545e-06,
+      "loss": 0.4435,
+      "step": 7965
+    },
+    {
+      "epoch": 3.1136329958014324,
+      "grad_norm": 0.4895329614863369,
+      "learning_rate": 4.72918539586195e-06,
+      "loss": 0.4424,
+      "step": 7966
+    },
+    {
+      "epoch": 3.1140281551000246,
+      "grad_norm": 0.48685840804033426,
+      "learning_rate": 4.729114469418559e-06,
+      "loss": 0.4529,
+      "step": 7967
+    },
+    {
+      "epoch": 3.114423314398617,
+      "grad_norm": 0.5011426297857062,
+      "learning_rate": 4.729043534220559e-06,
+      "loss": 0.4412,
+      "step": 7968
+    },
+    {
+      "epoch": 3.114818473697209,
+      "grad_norm": 0.5086734828450132,
+      "learning_rate": 4.728972590268229e-06,
+      "loss": 0.4444,
+      "step": 7969
+    },
+    {
+      "epoch": 3.1152136329958013,
+      "grad_norm": 0.4872906939792074,
+      "learning_rate": 4.728901637561849e-06,
+      "loss": 0.4318,
+      "step": 7970
+    },
+    {
+      "epoch": 3.1156087922943936,
+      "grad_norm": 0.4894151559799161,
+      "learning_rate": 4.7288306761016976e-06,
+      "loss": 0.4377,
+      "step": 7971
+    },
+    {
+      "epoch": 3.116003951592986,
+      "grad_norm": 0.5055954518080974,
+      "learning_rate": 4.7287597058880516e-06,
+      "loss": 0.4496,
+      "step": 7972
+    },
+    {
+      "epoch": 3.116399110891578,
+      "grad_norm": 0.5148624398253207,
+      "learning_rate": 4.728688726921191e-06,
+      "loss": 0.4341,
+      "step": 7973
+    },
+    {
+      "epoch": 3.1167942701901703,
+      "grad_norm": 0.5155539975890019,
+      "learning_rate": 4.728617739201396e-06,
+      "loss": 0.4545,
+      "step": 7974
+    },
+    {
+      "epoch": 3.1171894294887625,
+      "grad_norm": 0.4864073054526971,
+      "learning_rate": 4.728546742728941e-06,
+      "loss": 0.4356,
+      "step": 7975
+    },
+    {
+      "epoch": 3.117584588787355,
+      "grad_norm": 0.48348184096928454,
+      "learning_rate": 4.728475737504109e-06,
+      "loss": 0.4346,
+      "step": 7976
+    },
+    {
+      "epoch": 3.117979748085947,
+      "grad_norm": 0.4769632382044736,
+      "learning_rate": 4.728404723527178e-06,
+      "loss": 0.4325,
+      "step": 7977
+    },
+    {
+      "epoch": 3.1183749073845393,
+      "grad_norm": 0.4977904405800571,
+      "learning_rate": 4.728333700798427e-06,
+      "loss": 0.4431,
+      "step": 7978
+    },
+    {
+      "epoch": 3.1187700666831315,
+      "grad_norm": 0.4786988797880768,
+      "learning_rate": 4.728262669318133e-06,
+      "loss": 0.4377,
+      "step": 7979
+    },
+    {
+      "epoch": 3.1191652259817237,
+      "grad_norm": 0.49376051451383535,
+      "learning_rate": 4.728191629086576e-06,
+      "loss": 0.4405,
+      "step": 7980
+    },
+    {
+      "epoch": 3.119560385280316,
+      "grad_norm": 0.49204273770267093,
+      "learning_rate": 4.728120580104036e-06,
+      "loss": 0.4542,
+      "step": 7981
+    },
+    {
+      "epoch": 3.1199555445789082,
+      "grad_norm": 0.47099976061807647,
+      "learning_rate": 4.728049522370791e-06,
+      "loss": 0.4385,
+      "step": 7982
+    },
+    {
+      "epoch": 3.1203507038775005,
+      "grad_norm": 0.5025481626978274,
+      "learning_rate": 4.727978455887121e-06,
+      "loss": 0.462,
+      "step": 7983
+    },
+    {
+      "epoch": 3.1207458631760927,
+      "grad_norm": 0.5017733662110474,
+      "learning_rate": 4.727907380653305e-06,
+      "loss": 0.4428,
+      "step": 7984
+    },
+    {
+      "epoch": 3.121141022474685,
+      "grad_norm": 0.483833290731428,
+      "learning_rate": 4.7278362966696204e-06,
+      "loss": 0.431,
+      "step": 7985
+    },
+    {
+      "epoch": 3.121536181773277,
+      "grad_norm": 0.5083971021184884,
+      "learning_rate": 4.727765203936348e-06,
+      "loss": 0.4337,
+      "step": 7986
+    },
+    {
+      "epoch": 3.1219313410718694,
+      "grad_norm": 0.49782182938515346,
+      "learning_rate": 4.727694102453767e-06,
+      "loss": 0.4353,
+      "step": 7987
+    },
+    {
+      "epoch": 3.1223265003704617,
+      "grad_norm": 0.5092377334599689,
+      "learning_rate": 4.727622992222156e-06,
+      "loss": 0.4453,
+      "step": 7988
+    },
+    {
+      "epoch": 3.122721659669054,
+      "grad_norm": 0.4911234394591529,
+      "learning_rate": 4.7275518732417945e-06,
+      "loss": 0.4348,
+      "step": 7989
+    },
+    {
+      "epoch": 3.123116818967646,
+      "grad_norm": 0.4745444703266985,
+      "learning_rate": 4.727480745512962e-06,
+      "loss": 0.4433,
+      "step": 7990
+    },
+    {
+      "epoch": 3.1235119782662384,
+      "grad_norm": 0.4882371574102391,
+      "learning_rate": 4.727409609035938e-06,
+      "loss": 0.444,
+      "step": 7991
+    },
+    {
+      "epoch": 3.1239071375648306,
+      "grad_norm": 0.49239501433303656,
+      "learning_rate": 4.727338463811002e-06,
+      "loss": 0.4373,
+      "step": 7992
+    },
+    {
+      "epoch": 3.124302296863423,
+      "grad_norm": 0.48170385902308394,
+      "learning_rate": 4.727267309838432e-06,
+      "loss": 0.442,
+      "step": 7993
+    },
+    {
+      "epoch": 3.124697456162015,
+      "grad_norm": 0.49440788201479674,
+      "learning_rate": 4.727196147118509e-06,
+      "loss": 0.4366,
+      "step": 7994
+    },
+    {
+      "epoch": 3.1250926154606073,
+      "grad_norm": 0.48255932560969295,
+      "learning_rate": 4.727124975651512e-06,
+      "loss": 0.4361,
+      "step": 7995
+    },
+    {
+      "epoch": 3.1254877747591996,
+      "grad_norm": 0.49476276790071294,
+      "learning_rate": 4.727053795437721e-06,
+      "loss": 0.4573,
+      "step": 7996
+    },
+    {
+      "epoch": 3.125882934057792,
+      "grad_norm": 0.49896678550201234,
+      "learning_rate": 4.726982606477414e-06,
+      "loss": 0.4553,
+      "step": 7997
+    },
+    {
+      "epoch": 3.126278093356384,
+      "grad_norm": 0.47886145773353866,
+      "learning_rate": 4.7269114087708714e-06,
+      "loss": 0.4255,
+      "step": 7998
+    },
+    {
+      "epoch": 3.1266732526549763,
+      "grad_norm": 0.50038796905005,
+      "learning_rate": 4.7268402023183736e-06,
+      "loss": 0.444,
+      "step": 7999
+    },
+    {
+      "epoch": 3.1270684119535685,
+      "grad_norm": 0.49176360838395583,
+      "learning_rate": 4.7267689871201995e-06,
+      "loss": 0.4473,
+      "step": 8000
+    },
+    {
+      "epoch": 3.1274635712521612,
+      "grad_norm": 0.4995031801626146,
+      "learning_rate": 4.72669776317663e-06,
+      "loss": 0.4583,
+      "step": 8001
+    },
+    {
+      "epoch": 3.1278587305507535,
+      "grad_norm": 0.48056503087598096,
+      "learning_rate": 4.726626530487943e-06,
+      "loss": 0.4377,
+      "step": 8002
+    },
+    {
+      "epoch": 3.1282538898493457,
+      "grad_norm": 0.5014416609905591,
+      "learning_rate": 4.726555289054419e-06,
+      "loss": 0.4422,
+      "step": 8003
+    },
+    {
+      "epoch": 3.128649049147938,
+      "grad_norm": 0.49552073766113186,
+      "learning_rate": 4.726484038876338e-06,
+      "loss": 0.4479,
+      "step": 8004
+    },
+    {
+      "epoch": 3.12904420844653,
+      "grad_norm": 0.5238464019017981,
+      "learning_rate": 4.726412779953979e-06,
+      "loss": 0.4569,
+      "step": 8005
+    },
+    {
+      "epoch": 3.1294393677451224,
+      "grad_norm": 0.5077356869403138,
+      "learning_rate": 4.726341512287623e-06,
+      "loss": 0.4641,
+      "step": 8006
+    },
+    {
+      "epoch": 3.1298345270437147,
+      "grad_norm": 0.5100490606005,
+      "learning_rate": 4.72627023587755e-06,
+      "loss": 0.4478,
+      "step": 8007
+    },
+    {
+      "epoch": 3.130229686342307,
+      "grad_norm": 0.48960448306534565,
+      "learning_rate": 4.726198950724039e-06,
+      "loss": 0.437,
+      "step": 8008
+    },
+    {
+      "epoch": 3.130624845640899,
+      "grad_norm": 0.483619465818084,
+      "learning_rate": 4.726127656827371e-06,
+      "loss": 0.4476,
+      "step": 8009
+    },
+    {
+      "epoch": 3.1310200049394914,
+      "grad_norm": 0.5085324589140995,
+      "learning_rate": 4.726056354187825e-06,
+      "loss": 0.4336,
+      "step": 8010
+    },
+    {
+      "epoch": 3.1314151642380836,
+      "grad_norm": 0.5078606194501115,
+      "learning_rate": 4.725985042805681e-06,
+      "loss": 0.4465,
+      "step": 8011
+    },
+    {
+      "epoch": 3.131810323536676,
+      "grad_norm": 0.49723065197368327,
+      "learning_rate": 4.725913722681219e-06,
+      "loss": 0.4538,
+      "step": 8012
+    },
+    {
+      "epoch": 3.132205482835268,
+      "grad_norm": 0.5007007217832884,
+      "learning_rate": 4.72584239381472e-06,
+      "loss": 0.4738,
+      "step": 8013
+    },
+    {
+      "epoch": 3.1326006421338604,
+      "grad_norm": 0.4898205436347479,
+      "learning_rate": 4.725771056206464e-06,
+      "loss": 0.4344,
+      "step": 8014
+    },
+    {
+      "epoch": 3.1329958014324526,
+      "grad_norm": 0.5166831822498245,
+      "learning_rate": 4.725699709856731e-06,
+      "loss": 0.4412,
+      "step": 8015
+    },
+    {
+      "epoch": 3.133390960731045,
+      "grad_norm": 0.48591617847536656,
+      "learning_rate": 4.7256283547658e-06,
+      "loss": 0.4383,
+      "step": 8016
+    },
+    {
+      "epoch": 3.133786120029637,
+      "grad_norm": 0.49374057323016374,
+      "learning_rate": 4.725556990933953e-06,
+      "loss": 0.4509,
+      "step": 8017
+    },
+    {
+      "epoch": 3.1341812793282293,
+      "grad_norm": 0.48914627951452694,
+      "learning_rate": 4.72548561836147e-06,
+      "loss": 0.432,
+      "step": 8018
+    },
+    {
+      "epoch": 3.1345764386268216,
+      "grad_norm": 0.48980290819549344,
+      "learning_rate": 4.72541423704863e-06,
+      "loss": 0.4362,
+      "step": 8019
+    },
+    {
+      "epoch": 3.134971597925414,
+      "grad_norm": 0.49924757675061804,
+      "learning_rate": 4.7253428469957144e-06,
+      "loss": 0.4405,
+      "step": 8020
+    },
+    {
+      "epoch": 3.135366757224006,
+      "grad_norm": 0.5080074429751118,
+      "learning_rate": 4.725271448203003e-06,
+      "loss": 0.4477,
+      "step": 8021
+    },
+    {
+      "epoch": 3.1357619165225983,
+      "grad_norm": 0.5140176692672513,
+      "learning_rate": 4.7252000406707775e-06,
+      "loss": 0.4527,
+      "step": 8022
+    },
+    {
+      "epoch": 3.1361570758211905,
+      "grad_norm": 0.506032454034494,
+      "learning_rate": 4.725128624399318e-06,
+      "loss": 0.4488,
+      "step": 8023
+    },
+    {
+      "epoch": 3.1365522351197828,
+      "grad_norm": 0.49308027132547305,
+      "learning_rate": 4.725057199388903e-06,
+      "loss": 0.4351,
+      "step": 8024
+    },
+    {
+      "epoch": 3.136947394418375,
+      "grad_norm": 0.5104007123637837,
+      "learning_rate": 4.724985765639815e-06,
+      "loss": 0.4602,
+      "step": 8025
+    },
+    {
+      "epoch": 3.1373425537169672,
+      "grad_norm": 0.48895688295994194,
+      "learning_rate": 4.7249143231523345e-06,
+      "loss": 0.4409,
+      "step": 8026
+    },
+    {
+      "epoch": 3.1377377130155595,
+      "grad_norm": 0.5064246682765903,
+      "learning_rate": 4.724842871926741e-06,
+      "loss": 0.4379,
+      "step": 8027
+    },
+    {
+      "epoch": 3.1381328723141517,
+      "grad_norm": 0.5069265300363764,
+      "learning_rate": 4.724771411963316e-06,
+      "loss": 0.4647,
+      "step": 8028
+    },
+    {
+      "epoch": 3.138528031612744,
+      "grad_norm": 0.5030257723209636,
+      "learning_rate": 4.72469994326234e-06,
+      "loss": 0.4447,
+      "step": 8029
+    },
+    {
+      "epoch": 3.138923190911336,
+      "grad_norm": 0.5051475543956928,
+      "learning_rate": 4.724628465824093e-06,
+      "loss": 0.4457,
+      "step": 8030
+    },
+    {
+      "epoch": 3.1393183502099284,
+      "grad_norm": 0.47809073280195014,
+      "learning_rate": 4.724556979648856e-06,
+      "loss": 0.4483,
+      "step": 8031
+    },
+    {
+      "epoch": 3.1397135095085207,
+      "grad_norm": 0.5238988459465759,
+      "learning_rate": 4.724485484736911e-06,
+      "loss": 0.4503,
+      "step": 8032
+    },
+    {
+      "epoch": 3.140108668807113,
+      "grad_norm": 0.4964702713402975,
+      "learning_rate": 4.724413981088537e-06,
+      "loss": 0.4453,
+      "step": 8033
+    },
+    {
+      "epoch": 3.140503828105705,
+      "grad_norm": 0.49721437049653233,
+      "learning_rate": 4.724342468704016e-06,
+      "loss": 0.4476,
+      "step": 8034
+    },
+    {
+      "epoch": 3.1408989874042974,
+      "grad_norm": 0.47752057571805373,
+      "learning_rate": 4.724270947583628e-06,
+      "loss": 0.4351,
+      "step": 8035
+    },
+    {
+      "epoch": 3.1412941467028896,
+      "grad_norm": 0.4937090612310495,
+      "learning_rate": 4.724199417727654e-06,
+      "loss": 0.4579,
+      "step": 8036
+    },
+    {
+      "epoch": 3.141689306001482,
+      "grad_norm": 0.48947682743996235,
+      "learning_rate": 4.724127879136377e-06,
+      "loss": 0.4379,
+      "step": 8037
+    },
+    {
+      "epoch": 3.142084465300074,
+      "grad_norm": 0.5001718495358742,
+      "learning_rate": 4.7240563318100755e-06,
+      "loss": 0.438,
+      "step": 8038
+    },
+    {
+      "epoch": 3.1424796245986664,
+      "grad_norm": 0.5014707511961164,
+      "learning_rate": 4.723984775749031e-06,
+      "loss": 0.4457,
+      "step": 8039
+    },
+    {
+      "epoch": 3.1428747838972586,
+      "grad_norm": 0.5028261244621974,
+      "learning_rate": 4.7239132109535245e-06,
+      "loss": 0.4604,
+      "step": 8040
+    },
+    {
+      "epoch": 3.143269943195851,
+      "grad_norm": 0.4934107578955984,
+      "learning_rate": 4.723841637423837e-06,
+      "loss": 0.4377,
+      "step": 8041
+    },
+    {
+      "epoch": 3.143665102494443,
+      "grad_norm": 0.48165968429853356,
+      "learning_rate": 4.723770055160251e-06,
+      "loss": 0.4519,
+      "step": 8042
+    },
+    {
+      "epoch": 3.1440602617930353,
+      "grad_norm": 0.49320305911875595,
+      "learning_rate": 4.723698464163046e-06,
+      "loss": 0.4551,
+      "step": 8043
+    },
+    {
+      "epoch": 3.1444554210916276,
+      "grad_norm": 0.5151414275313321,
+      "learning_rate": 4.723626864432504e-06,
+      "loss": 0.4411,
+      "step": 8044
+    },
+    {
+      "epoch": 3.14485058039022,
+      "grad_norm": 0.48841016631789297,
+      "learning_rate": 4.723555255968906e-06,
+      "loss": 0.4375,
+      "step": 8045
+    },
+    {
+      "epoch": 3.145245739688812,
+      "grad_norm": 0.49755737281596757,
+      "learning_rate": 4.723483638772532e-06,
+      "loss": 0.4511,
+      "step": 8046
+    },
+    {
+      "epoch": 3.1456408989874043,
+      "grad_norm": 0.5133698387807278,
+      "learning_rate": 4.723412012843666e-06,
+      "loss": 0.4507,
+      "step": 8047
+    },
+    {
+      "epoch": 3.1460360582859965,
+      "grad_norm": 0.5113782564797575,
+      "learning_rate": 4.723340378182587e-06,
+      "loss": 0.4471,
+      "step": 8048
+    },
+    {
+      "epoch": 3.1464312175845888,
+      "grad_norm": 0.5025143915593622,
+      "learning_rate": 4.7232687347895775e-06,
+      "loss": 0.4506,
+      "step": 8049
+    },
+    {
+      "epoch": 3.146826376883181,
+      "grad_norm": 0.5090332482162311,
+      "learning_rate": 4.7231970826649185e-06,
+      "loss": 0.4601,
+      "step": 8050
+    },
+    {
+      "epoch": 3.1472215361817732,
+      "grad_norm": 0.5129767234895644,
+      "learning_rate": 4.7231254218088906e-06,
+      "loss": 0.4339,
+      "step": 8051
+    },
+    {
+      "epoch": 3.1476166954803655,
+      "grad_norm": 0.4931726199572382,
+      "learning_rate": 4.723053752221777e-06,
+      "loss": 0.4628,
+      "step": 8052
+    },
+    {
+      "epoch": 3.1480118547789577,
+      "grad_norm": 0.4761021361758315,
+      "learning_rate": 4.722982073903857e-06,
+      "loss": 0.4111,
+      "step": 8053
+    },
+    {
+      "epoch": 3.14840701407755,
+      "grad_norm": 0.49121750684116305,
+      "learning_rate": 4.722910386855414e-06,
+      "loss": 0.4332,
+      "step": 8054
+    },
+    {
+      "epoch": 3.148802173376142,
+      "grad_norm": 0.49386756521790587,
+      "learning_rate": 4.722838691076729e-06,
+      "loss": 0.432,
+      "step": 8055
+    },
+    {
+      "epoch": 3.1491973326747345,
+      "grad_norm": 0.4962985239368982,
+      "learning_rate": 4.722766986568083e-06,
+      "loss": 0.4472,
+      "step": 8056
+    },
+    {
+      "epoch": 3.1495924919733267,
+      "grad_norm": 0.49652014240893927,
+      "learning_rate": 4.722695273329758e-06,
+      "loss": 0.4585,
+      "step": 8057
+    },
+    {
+      "epoch": 3.149987651271919,
+      "grad_norm": 0.4984442502510313,
+      "learning_rate": 4.722623551362036e-06,
+      "loss": 0.4285,
+      "step": 8058
+    },
+    {
+      "epoch": 3.150382810570511,
+      "grad_norm": 0.4923794585945551,
+      "learning_rate": 4.7225518206651975e-06,
+      "loss": 0.451,
+      "step": 8059
+    },
+    {
+      "epoch": 3.1507779698691034,
+      "grad_norm": 0.492056259052196,
+      "learning_rate": 4.722480081239527e-06,
+      "loss": 0.4423,
+      "step": 8060
+    },
+    {
+      "epoch": 3.1511731291676957,
+      "grad_norm": 0.4936357538899125,
+      "learning_rate": 4.7224083330853025e-06,
+      "loss": 0.4449,
+      "step": 8061
+    },
+    {
+      "epoch": 3.151568288466288,
+      "grad_norm": 0.4891054280607878,
+      "learning_rate": 4.722336576202808e-06,
+      "loss": 0.4607,
+      "step": 8062
+    },
+    {
+      "epoch": 3.15196344776488,
+      "grad_norm": 0.5006228316341875,
+      "learning_rate": 4.722264810592325e-06,
+      "loss": 0.4578,
+      "step": 8063
+    },
+    {
+      "epoch": 3.1523586070634724,
+      "grad_norm": 0.49583466507150603,
+      "learning_rate": 4.722193036254135e-06,
+      "loss": 0.4505,
+      "step": 8064
+    },
+    {
+      "epoch": 3.1527537663620646,
+      "grad_norm": 0.5029343842600947,
+      "learning_rate": 4.722121253188521e-06,
+      "loss": 0.4449,
+      "step": 8065
+    },
+    {
+      "epoch": 3.153148925660657,
+      "grad_norm": 0.5118215295573837,
+      "learning_rate": 4.722049461395763e-06,
+      "loss": 0.434,
+      "step": 8066
+    },
+    {
+      "epoch": 3.153544084959249,
+      "grad_norm": 0.4861987832676822,
+      "learning_rate": 4.721977660876144e-06,
+      "loss": 0.4308,
+      "step": 8067
+    },
+    {
+      "epoch": 3.1539392442578413,
+      "grad_norm": 0.5021670172513321,
+      "learning_rate": 4.721905851629947e-06,
+      "loss": 0.447,
+      "step": 8068
+    },
+    {
+      "epoch": 3.1543344035564336,
+      "grad_norm": 0.4869993622406486,
+      "learning_rate": 4.721834033657452e-06,
+      "loss": 0.4596,
+      "step": 8069
+    },
+    {
+      "epoch": 3.154729562855026,
+      "grad_norm": 0.515713215408121,
+      "learning_rate": 4.721762206958943e-06,
+      "loss": 0.4579,
+      "step": 8070
+    },
+    {
+      "epoch": 3.155124722153618,
+      "grad_norm": 0.5056235944038723,
+      "learning_rate": 4.7216903715347005e-06,
+      "loss": 0.4692,
+      "step": 8071
+    },
+    {
+      "epoch": 3.1555198814522103,
+      "grad_norm": 0.506279979442091,
+      "learning_rate": 4.721618527385008e-06,
+      "loss": 0.4411,
+      "step": 8072
+    },
+    {
+      "epoch": 3.1559150407508025,
+      "grad_norm": 0.507576731306934,
+      "learning_rate": 4.721546674510146e-06,
+      "loss": 0.4476,
+      "step": 8073
+    },
+    {
+      "epoch": 3.1563102000493948,
+      "grad_norm": 0.4959496994776162,
+      "learning_rate": 4.721474812910398e-06,
+      "loss": 0.4491,
+      "step": 8074
+    },
+    {
+      "epoch": 3.156705359347987,
+      "grad_norm": 0.48307075497150315,
+      "learning_rate": 4.721402942586046e-06,
+      "loss": 0.4518,
+      "step": 8075
+    },
+    {
+      "epoch": 3.1571005186465793,
+      "grad_norm": 0.5009750695914092,
+      "learning_rate": 4.721331063537372e-06,
+      "loss": 0.4293,
+      "step": 8076
+    },
+    {
+      "epoch": 3.1574956779451715,
+      "grad_norm": 0.5279332021896078,
+      "learning_rate": 4.721259175764659e-06,
+      "loss": 0.4506,
+      "step": 8077
+    },
+    {
+      "epoch": 3.1578908372437637,
+      "grad_norm": 0.5166606359510496,
+      "learning_rate": 4.721187279268189e-06,
+      "loss": 0.4785,
+      "step": 8078
+    },
+    {
+      "epoch": 3.158285996542356,
+      "grad_norm": 0.4957762206244525,
+      "learning_rate": 4.721115374048243e-06,
+      "loss": 0.4429,
+      "step": 8079
+    },
+    {
+      "epoch": 3.158681155840948,
+      "grad_norm": 0.5130296220814176,
+      "learning_rate": 4.721043460105106e-06,
+      "loss": 0.4582,
+      "step": 8080
+    },
+    {
+      "epoch": 3.1590763151395405,
+      "grad_norm": 0.5084021122302967,
+      "learning_rate": 4.720971537439058e-06,
+      "loss": 0.4468,
+      "step": 8081
+    },
+    {
+      "epoch": 3.1594714744381327,
+      "grad_norm": 0.5024239703569898,
+      "learning_rate": 4.720899606050382e-06,
+      "loss": 0.4391,
+      "step": 8082
+    },
+    {
+      "epoch": 3.159866633736725,
+      "grad_norm": 0.49038806705570115,
+      "learning_rate": 4.720827665939362e-06,
+      "loss": 0.4273,
+      "step": 8083
+    },
+    {
+      "epoch": 3.160261793035317,
+      "grad_norm": 0.5006231223006984,
+      "learning_rate": 4.720755717106278e-06,
+      "loss": 0.4549,
+      "step": 8084
+    },
+    {
+      "epoch": 3.1606569523339094,
+      "grad_norm": 0.5146917436143081,
+      "learning_rate": 4.7206837595514155e-06,
+      "loss": 0.4446,
+      "step": 8085
+    },
+    {
+      "epoch": 3.1610521116325017,
+      "grad_norm": 0.502995584917966,
+      "learning_rate": 4.720611793275055e-06,
+      "loss": 0.4453,
+      "step": 8086
+    },
+    {
+      "epoch": 3.161447270931094,
+      "grad_norm": 0.49610937331042165,
+      "learning_rate": 4.7205398182774806e-06,
+      "loss": 0.4585,
+      "step": 8087
+    },
+    {
+      "epoch": 3.161842430229686,
+      "grad_norm": 0.4843164971279493,
+      "learning_rate": 4.720467834558973e-06,
+      "loss": 0.4481,
+      "step": 8088
+    },
+    {
+      "epoch": 3.1622375895282784,
+      "grad_norm": 0.5128044968276422,
+      "learning_rate": 4.720395842119817e-06,
+      "loss": 0.4687,
+      "step": 8089
+    },
+    {
+      "epoch": 3.162632748826871,
+      "grad_norm": 0.48798274809495723,
+      "learning_rate": 4.7203238409602936e-06,
+      "loss": 0.4476,
+      "step": 8090
+    },
+    {
+      "epoch": 3.1630279081254633,
+      "grad_norm": 0.4928931238418614,
+      "learning_rate": 4.720251831080687e-06,
+      "loss": 0.4345,
+      "step": 8091
+    },
+    {
+      "epoch": 3.1634230674240555,
+      "grad_norm": 0.5669879530273516,
+      "learning_rate": 4.720179812481279e-06,
+      "loss": 0.4412,
+      "step": 8092
+    },
+    {
+      "epoch": 3.163818226722648,
+      "grad_norm": 0.5067692284751482,
+      "learning_rate": 4.720107785162353e-06,
+      "loss": 0.4427,
+      "step": 8093
+    },
+    {
+      "epoch": 3.16421338602124,
+      "grad_norm": 0.48670841000596843,
+      "learning_rate": 4.7200357491241925e-06,
+      "loss": 0.4391,
+      "step": 8094
+    },
+    {
+      "epoch": 3.1646085453198323,
+      "grad_norm": 0.49696643464950435,
+      "learning_rate": 4.71996370436708e-06,
+      "loss": 0.4545,
+      "step": 8095
+    },
+    {
+      "epoch": 3.1650037046184245,
+      "grad_norm": 0.5074530215661209,
+      "learning_rate": 4.719891650891296e-06,
+      "loss": 0.4583,
+      "step": 8096
+    },
+    {
+      "epoch": 3.1653988639170167,
+      "grad_norm": 0.49354704231252805,
+      "learning_rate": 4.719819588697127e-06,
+      "loss": 0.4461,
+      "step": 8097
+    },
+    {
+      "epoch": 3.165794023215609,
+      "grad_norm": 0.4963673335562395,
+      "learning_rate": 4.719747517784854e-06,
+      "loss": 0.4349,
+      "step": 8098
+    },
+    {
+      "epoch": 3.1661891825142012,
+      "grad_norm": 0.4904154243050707,
+      "learning_rate": 4.719675438154761e-06,
+      "loss": 0.4493,
+      "step": 8099
+    },
+    {
+      "epoch": 3.1665843418127935,
+      "grad_norm": 0.47953064206220075,
+      "learning_rate": 4.719603349807132e-06,
+      "loss": 0.4484,
+      "step": 8100
+    },
+    {
+      "epoch": 3.1669795011113857,
+      "grad_norm": 0.47907940594638293,
+      "learning_rate": 4.719531252742246e-06,
+      "loss": 0.4405,
+      "step": 8101
+    },
+    {
+      "epoch": 3.167374660409978,
+      "grad_norm": 0.4898293576798917,
+      "learning_rate": 4.7194591469603915e-06,
+      "loss": 0.4465,
+      "step": 8102
+    },
+    {
+      "epoch": 3.16776981970857,
+      "grad_norm": 0.4894519791280581,
+      "learning_rate": 4.7193870324618486e-06,
+      "loss": 0.4411,
+      "step": 8103
+    },
+    {
+      "epoch": 3.1681649790071624,
+      "grad_norm": 0.5194593246729222,
+      "learning_rate": 4.7193149092469e-06,
+      "loss": 0.4674,
+      "step": 8104
+    },
+    {
+      "epoch": 3.1685601383057547,
+      "grad_norm": 0.4838702824968695,
+      "learning_rate": 4.719242777315831e-06,
+      "loss": 0.4505,
+      "step": 8105
+    },
+    {
+      "epoch": 3.168955297604347,
+      "grad_norm": 0.506659585187492,
+      "learning_rate": 4.719170636668924e-06,
+      "loss": 0.4573,
+      "step": 8106
+    },
+    {
+      "epoch": 3.169350456902939,
+      "grad_norm": 0.487503557571576,
+      "learning_rate": 4.719098487306463e-06,
+      "loss": 0.4403,
+      "step": 8107
+    },
+    {
+      "epoch": 3.1697456162015314,
+      "grad_norm": 0.4876179319034556,
+      "learning_rate": 4.7190263292287296e-06,
+      "loss": 0.4379,
+      "step": 8108
+    },
+    {
+      "epoch": 3.1701407755001236,
+      "grad_norm": 0.5106552416620477,
+      "learning_rate": 4.718954162436008e-06,
+      "loss": 0.4282,
+      "step": 8109
+    },
+    {
+      "epoch": 3.170535934798716,
+      "grad_norm": 0.48302060495170895,
+      "learning_rate": 4.718881986928583e-06,
+      "loss": 0.4424,
+      "step": 8110
+    },
+    {
+      "epoch": 3.170931094097308,
+      "grad_norm": 0.4958854915432569,
+      "learning_rate": 4.718809802706736e-06,
+      "loss": 0.4507,
+      "step": 8111
+    },
+    {
+      "epoch": 3.1713262533959004,
+      "grad_norm": 0.4932554091257428,
+      "learning_rate": 4.7187376097707515e-06,
+      "loss": 0.4481,
+      "step": 8112
+    },
+    {
+      "epoch": 3.1717214126944926,
+      "grad_norm": 0.48904571684029674,
+      "learning_rate": 4.718665408120913e-06,
+      "loss": 0.4415,
+      "step": 8113
+    },
+    {
+      "epoch": 3.172116571993085,
+      "grad_norm": 0.5168961069991919,
+      "learning_rate": 4.718593197757505e-06,
+      "loss": 0.4599,
+      "step": 8114
+    },
+    {
+      "epoch": 3.172511731291677,
+      "grad_norm": 0.5063108372391634,
+      "learning_rate": 4.718520978680809e-06,
+      "loss": 0.4517,
+      "step": 8115
+    },
+    {
+      "epoch": 3.1729068905902693,
+      "grad_norm": 0.5007293842416772,
+      "learning_rate": 4.71844875089111e-06,
+      "loss": 0.4267,
+      "step": 8116
+    },
+    {
+      "epoch": 3.1733020498888616,
+      "grad_norm": 0.48180519881194206,
+      "learning_rate": 4.718376514388691e-06,
+      "loss": 0.4348,
+      "step": 8117
+    },
+    {
+      "epoch": 3.173697209187454,
+      "grad_norm": 0.5089904727964285,
+      "learning_rate": 4.718304269173837e-06,
+      "loss": 0.4541,
+      "step": 8118
+    },
+    {
+      "epoch": 3.174092368486046,
+      "grad_norm": 0.48877774632457816,
+      "learning_rate": 4.718232015246831e-06,
+      "loss": 0.4376,
+      "step": 8119
+    },
+    {
+      "epoch": 3.1744875277846383,
+      "grad_norm": 0.4891031842232021,
+      "learning_rate": 4.718159752607955e-06,
+      "loss": 0.4461,
+      "step": 8120
+    },
+    {
+      "epoch": 3.1748826870832305,
+      "grad_norm": 0.5040292533478185,
+      "learning_rate": 4.718087481257496e-06,
+      "loss": 0.4475,
+      "step": 8121
+    },
+    {
+      "epoch": 3.1752778463818228,
+      "grad_norm": 0.4981159124676147,
+      "learning_rate": 4.718015201195736e-06,
+      "loss": 0.45,
+      "step": 8122
+    },
+    {
+      "epoch": 3.175673005680415,
+      "grad_norm": 0.489987198323066,
+      "learning_rate": 4.717942912422958e-06,
+      "loss": 0.4469,
+      "step": 8123
+    },
+    {
+      "epoch": 3.1760681649790072,
+      "grad_norm": 0.5132570158383588,
+      "learning_rate": 4.717870614939449e-06,
+      "loss": 0.4671,
+      "step": 8124
+    },
+    {
+      "epoch": 3.1764633242775995,
+      "grad_norm": 0.4838179306406595,
+      "learning_rate": 4.717798308745489e-06,
+      "loss": 0.4483,
+      "step": 8125
+    },
+    {
+      "epoch": 3.1768584835761917,
+      "grad_norm": 0.49436482089246664,
+      "learning_rate": 4.717725993841366e-06,
+      "loss": 0.4505,
+      "step": 8126
+    },
+    {
+      "epoch": 3.177253642874784,
+      "grad_norm": 0.4927294185258886,
+      "learning_rate": 4.71765367022736e-06,
+      "loss": 0.4607,
+      "step": 8127
+    },
+    {
+      "epoch": 3.177648802173376,
+      "grad_norm": 0.5023211446663597,
+      "learning_rate": 4.717581337903759e-06,
+      "loss": 0.4412,
+      "step": 8128
+    },
+    {
+      "epoch": 3.1780439614719684,
+      "grad_norm": 0.4962184544429156,
+      "learning_rate": 4.717508996870843e-06,
+      "loss": 0.4472,
+      "step": 8129
+    },
+    {
+      "epoch": 3.1784391207705607,
+      "grad_norm": 0.5060665727054257,
+      "learning_rate": 4.717436647128899e-06,
+      "loss": 0.4483,
+      "step": 8130
+    },
+    {
+      "epoch": 3.178834280069153,
+      "grad_norm": 0.4856256691010014,
+      "learning_rate": 4.717364288678211e-06,
+      "loss": 0.4531,
+      "step": 8131
+    },
+    {
+      "epoch": 3.179229439367745,
+      "grad_norm": 0.5031699859568227,
+      "learning_rate": 4.717291921519062e-06,
+      "loss": 0.4601,
+      "step": 8132
+    },
+    {
+      "epoch": 3.1796245986663374,
+      "grad_norm": 0.49801437420123507,
+      "learning_rate": 4.717219545651736e-06,
+      "loss": 0.4482,
+      "step": 8133
+    },
+    {
+      "epoch": 3.1800197579649296,
+      "grad_norm": 0.5050121211282538,
+      "learning_rate": 4.71714716107652e-06,
+      "loss": 0.4642,
+      "step": 8134
+    },
+    {
+      "epoch": 3.180414917263522,
+      "grad_norm": 0.5261624095797713,
+      "learning_rate": 4.717074767793695e-06,
+      "loss": 0.4363,
+      "step": 8135
+    },
+    {
+      "epoch": 3.180810076562114,
+      "grad_norm": 0.5003002025261412,
+      "learning_rate": 4.717002365803547e-06,
+      "loss": 0.443,
+      "step": 8136
+    },
+    {
+      "epoch": 3.1812052358607064,
+      "grad_norm": 0.5162727913171745,
+      "learning_rate": 4.716929955106359e-06,
+      "loss": 0.4556,
+      "step": 8137
+    },
+    {
+      "epoch": 3.1816003951592986,
+      "grad_norm": 0.50383251369054,
+      "learning_rate": 4.716857535702417e-06,
+      "loss": 0.4476,
+      "step": 8138
+    },
+    {
+      "epoch": 3.181995554457891,
+      "grad_norm": 0.4989731953152965,
+      "learning_rate": 4.716785107592005e-06,
+      "loss": 0.4505,
+      "step": 8139
+    },
+    {
+      "epoch": 3.182390713756483,
+      "grad_norm": 0.4985898123702961,
+      "learning_rate": 4.716712670775407e-06,
+      "loss": 0.4513,
+      "step": 8140
+    },
+    {
+      "epoch": 3.1827858730550753,
+      "grad_norm": 0.4949371574614923,
+      "learning_rate": 4.7166402252529075e-06,
+      "loss": 0.4486,
+      "step": 8141
+    },
+    {
+      "epoch": 3.1831810323536676,
+      "grad_norm": 0.5050331893640804,
+      "learning_rate": 4.716567771024792e-06,
+      "loss": 0.457,
+      "step": 8142
+    },
+    {
+      "epoch": 3.18357619165226,
+      "grad_norm": 0.4860144273306709,
+      "learning_rate": 4.716495308091343e-06,
+      "loss": 0.4353,
+      "step": 8143
+    },
+    {
+      "epoch": 3.183971350950852,
+      "grad_norm": 0.48399721239140636,
+      "learning_rate": 4.716422836452846e-06,
+      "loss": 0.4373,
+      "step": 8144
+    },
+    {
+      "epoch": 3.1843665102494443,
+      "grad_norm": 0.4806840238583811,
+      "learning_rate": 4.716350356109588e-06,
+      "loss": 0.4395,
+      "step": 8145
+    },
+    {
+      "epoch": 3.1847616695480365,
+      "grad_norm": 0.4905696278198881,
+      "learning_rate": 4.716277867061851e-06,
+      "loss": 0.4521,
+      "step": 8146
+    },
+    {
+      "epoch": 3.1851568288466288,
+      "grad_norm": 0.5052154287634858,
+      "learning_rate": 4.716205369309919e-06,
+      "loss": 0.4654,
+      "step": 8147
+    },
+    {
+      "epoch": 3.185551988145221,
+      "grad_norm": 0.4976670668769747,
+      "learning_rate": 4.716132862854079e-06,
+      "loss": 0.4656,
+      "step": 8148
+    },
+    {
+      "epoch": 3.1859471474438132,
+      "grad_norm": 0.4856606394398254,
+      "learning_rate": 4.716060347694615e-06,
+      "loss": 0.4312,
+      "step": 8149
+    },
+    {
+      "epoch": 3.1863423067424055,
+      "grad_norm": 0.47616575166124564,
+      "learning_rate": 4.7159878238318116e-06,
+      "loss": 0.4321,
+      "step": 8150
+    },
+    {
+      "epoch": 3.1867374660409977,
+      "grad_norm": 0.4993342307580878,
+      "learning_rate": 4.715915291265954e-06,
+      "loss": 0.4534,
+      "step": 8151
+    },
+    {
+      "epoch": 3.18713262533959,
+      "grad_norm": 0.5057600356168487,
+      "learning_rate": 4.715842749997325e-06,
+      "loss": 0.454,
+      "step": 8152
+    },
+    {
+      "epoch": 3.187527784638182,
+      "grad_norm": 0.5076197913058856,
+      "learning_rate": 4.715770200026213e-06,
+      "loss": 0.449,
+      "step": 8153
+    },
+    {
+      "epoch": 3.1879229439367744,
+      "grad_norm": 0.47904704822140237,
+      "learning_rate": 4.7156976413529e-06,
+      "loss": 0.4405,
+      "step": 8154
+    },
+    {
+      "epoch": 3.1883181032353667,
+      "grad_norm": 0.478740457872185,
+      "learning_rate": 4.7156250739776725e-06,
+      "loss": 0.4527,
+      "step": 8155
+    },
+    {
+      "epoch": 3.188713262533959,
+      "grad_norm": 0.4865375368517737,
+      "learning_rate": 4.715552497900815e-06,
+      "loss": 0.4664,
+      "step": 8156
+    },
+    {
+      "epoch": 3.189108421832551,
+      "grad_norm": 0.49057170271705297,
+      "learning_rate": 4.715479913122613e-06,
+      "loss": 0.4498,
+      "step": 8157
+    },
+    {
+      "epoch": 3.1895035811311434,
+      "grad_norm": 0.48932211501442835,
+      "learning_rate": 4.715407319643352e-06,
+      "loss": 0.438,
+      "step": 8158
+    },
+    {
+      "epoch": 3.1898987404297356,
+      "grad_norm": 0.4986907746805224,
+      "learning_rate": 4.715334717463314e-06,
+      "loss": 0.4437,
+      "step": 8159
+    },
+    {
+      "epoch": 3.190293899728328,
+      "grad_norm": 0.5107324129303034,
+      "learning_rate": 4.715262106582788e-06,
+      "loss": 0.4663,
+      "step": 8160
+    },
+    {
+      "epoch": 3.19068905902692,
+      "grad_norm": 0.4816038125829416,
+      "learning_rate": 4.715189487002057e-06,
+      "loss": 0.4459,
+      "step": 8161
+    },
+    {
+      "epoch": 3.1910842183255124,
+      "grad_norm": 0.48589022124625236,
+      "learning_rate": 4.715116858721408e-06,
+      "loss": 0.4596,
+      "step": 8162
+    },
+    {
+      "epoch": 3.1914793776241046,
+      "grad_norm": 0.5059117225587179,
+      "learning_rate": 4.715044221741125e-06,
+      "loss": 0.4424,
+      "step": 8163
+    },
+    {
+      "epoch": 3.191874536922697,
+      "grad_norm": 0.4902259551321021,
+      "learning_rate": 4.714971576061492e-06,
+      "loss": 0.4436,
+      "step": 8164
+    },
+    {
+      "epoch": 3.192269696221289,
+      "grad_norm": 0.49921816165596244,
+      "learning_rate": 4.714898921682797e-06,
+      "loss": 0.4341,
+      "step": 8165
+    },
+    {
+      "epoch": 3.1926648555198813,
+      "grad_norm": 0.4873589691710802,
+      "learning_rate": 4.714826258605323e-06,
+      "loss": 0.447,
+      "step": 8166
+    },
+    {
+      "epoch": 3.1930600148184736,
+      "grad_norm": 0.4947509154503643,
+      "learning_rate": 4.714753586829357e-06,
+      "loss": 0.4531,
+      "step": 8167
+    },
+    {
+      "epoch": 3.193455174117066,
+      "grad_norm": 0.49979557568227245,
+      "learning_rate": 4.714680906355184e-06,
+      "loss": 0.4499,
+      "step": 8168
+    },
+    {
+      "epoch": 3.193850333415658,
+      "grad_norm": 0.47965466768323384,
+      "learning_rate": 4.714608217183088e-06,
+      "loss": 0.4294,
+      "step": 8169
+    },
+    {
+      "epoch": 3.1942454927142503,
+      "grad_norm": 0.5380227850408551,
+      "learning_rate": 4.7145355193133566e-06,
+      "loss": 0.44,
+      "step": 8170
+    },
+    {
+      "epoch": 3.1946406520128425,
+      "grad_norm": 0.4778666409695035,
+      "learning_rate": 4.714462812746275e-06,
+      "loss": 0.4518,
+      "step": 8171
+    },
+    {
+      "epoch": 3.1950358113114348,
+      "grad_norm": 0.498563645161947,
+      "learning_rate": 4.714390097482127e-06,
+      "loss": 0.4512,
+      "step": 8172
+    },
+    {
+      "epoch": 3.195430970610027,
+      "grad_norm": 0.502694986736014,
+      "learning_rate": 4.714317373521199e-06,
+      "loss": 0.4703,
+      "step": 8173
+    },
+    {
+      "epoch": 3.1958261299086193,
+      "grad_norm": 0.4955421090319872,
+      "learning_rate": 4.714244640863778e-06,
+      "loss": 0.4601,
+      "step": 8174
+    },
+    {
+      "epoch": 3.1962212892072115,
+      "grad_norm": 0.47703481341356674,
+      "learning_rate": 4.714171899510148e-06,
+      "loss": 0.4363,
+      "step": 8175
+    },
+    {
+      "epoch": 3.1966164485058037,
+      "grad_norm": 0.488778683807109,
+      "learning_rate": 4.714099149460596e-06,
+      "loss": 0.4309,
+      "step": 8176
+    },
+    {
+      "epoch": 3.197011607804396,
+      "grad_norm": 0.49243205733047013,
+      "learning_rate": 4.714026390715407e-06,
+      "loss": 0.4421,
+      "step": 8177
+    },
+    {
+      "epoch": 3.197406767102988,
+      "grad_norm": 0.47901350686825167,
+      "learning_rate": 4.7139536232748665e-06,
+      "loss": 0.4475,
+      "step": 8178
+    },
+    {
+      "epoch": 3.1978019264015805,
+      "grad_norm": 0.4873600139430021,
+      "learning_rate": 4.71388084713926e-06,
+      "loss": 0.4365,
+      "step": 8179
+    },
+    {
+      "epoch": 3.1981970857001727,
+      "grad_norm": 0.4750978959852484,
+      "learning_rate": 4.713808062308874e-06,
+      "loss": 0.4329,
+      "step": 8180
+    },
+    {
+      "epoch": 3.198592244998765,
+      "grad_norm": 0.49005747817347883,
+      "learning_rate": 4.713735268783995e-06,
+      "loss": 0.44,
+      "step": 8181
+    },
+    {
+      "epoch": 3.198987404297357,
+      "grad_norm": 0.4897517319745767,
+      "learning_rate": 4.713662466564908e-06,
+      "loss": 0.4418,
+      "step": 8182
+    },
+    {
+      "epoch": 3.1993825635959494,
+      "grad_norm": 0.4937199361984085,
+      "learning_rate": 4.713589655651898e-06,
+      "loss": 0.445,
+      "step": 8183
+    },
+    {
+      "epoch": 3.1997777228945417,
+      "grad_norm": 0.5006257765582313,
+      "learning_rate": 4.713516836045253e-06,
+      "loss": 0.4578,
+      "step": 8184
+    },
+    {
+      "epoch": 3.200172882193134,
+      "grad_norm": 0.5102363063233484,
+      "learning_rate": 4.713444007745258e-06,
+      "loss": 0.4649,
+      "step": 8185
+    },
+    {
+      "epoch": 3.200568041491726,
+      "grad_norm": 0.5619926568257341,
+      "learning_rate": 4.713371170752198e-06,
+      "loss": 0.4512,
+      "step": 8186
+    },
+    {
+      "epoch": 3.2009632007903184,
+      "grad_norm": 0.49017178589540195,
+      "learning_rate": 4.713298325066361e-06,
+      "loss": 0.4482,
+      "step": 8187
+    },
+    {
+      "epoch": 3.2013583600889106,
+      "grad_norm": 0.5075841792186128,
+      "learning_rate": 4.713225470688032e-06,
+      "loss": 0.4637,
+      "step": 8188
+    },
+    {
+      "epoch": 3.201753519387503,
+      "grad_norm": 0.4851504244889839,
+      "learning_rate": 4.713152607617497e-06,
+      "loss": 0.4354,
+      "step": 8189
+    },
+    {
+      "epoch": 3.2021486786860955,
+      "grad_norm": 0.49603485658612206,
+      "learning_rate": 4.713079735855043e-06,
+      "loss": 0.4669,
+      "step": 8190
+    },
+    {
+      "epoch": 3.202543837984688,
+      "grad_norm": 0.5131941929666564,
+      "learning_rate": 4.713006855400955e-06,
+      "loss": 0.4363,
+      "step": 8191
+    },
+    {
+      "epoch": 3.20293899728328,
+      "grad_norm": 0.5327331196587508,
+      "learning_rate": 4.712933966255521e-06,
+      "loss": 0.4517,
+      "step": 8192
+    },
+    {
+      "epoch": 3.2033341565818723,
+      "grad_norm": 0.47967217544059954,
+      "learning_rate": 4.7128610684190255e-06,
+      "loss": 0.4363,
+      "step": 8193
+    },
+    {
+      "epoch": 3.2037293158804645,
+      "grad_norm": 0.5032954106990305,
+      "learning_rate": 4.712788161891755e-06,
+      "loss": 0.4555,
+      "step": 8194
+    },
+    {
+      "epoch": 3.2041244751790567,
+      "grad_norm": 0.49443372271224756,
+      "learning_rate": 4.712715246673997e-06,
+      "loss": 0.4432,
+      "step": 8195
+    },
+    {
+      "epoch": 3.204519634477649,
+      "grad_norm": 0.5048761667286308,
+      "learning_rate": 4.712642322766037e-06,
+      "loss": 0.4636,
+      "step": 8196
+    },
+    {
+      "epoch": 3.204914793776241,
+      "grad_norm": 0.5114603132884818,
+      "learning_rate": 4.71256939016816e-06,
+      "loss": 0.4678,
+      "step": 8197
+    },
+    {
+      "epoch": 3.2053099530748335,
+      "grad_norm": 0.4783233894068138,
+      "learning_rate": 4.712496448880656e-06,
+      "loss": 0.4546,
+      "step": 8198
+    },
+    {
+      "epoch": 3.2057051123734257,
+      "grad_norm": 0.489295057789308,
+      "learning_rate": 4.712423498903809e-06,
+      "loss": 0.4391,
+      "step": 8199
+    },
+    {
+      "epoch": 3.206100271672018,
+      "grad_norm": 0.4876753870656324,
+      "learning_rate": 4.712350540237906e-06,
+      "loss": 0.4552,
+      "step": 8200
+    },
+    {
+      "epoch": 3.20649543097061,
+      "grad_norm": 0.4833416370700385,
+      "learning_rate": 4.712277572883233e-06,
+      "loss": 0.4461,
+      "step": 8201
+    },
+    {
+      "epoch": 3.2068905902692024,
+      "grad_norm": 0.4966315262459833,
+      "learning_rate": 4.712204596840077e-06,
+      "loss": 0.464,
+      "step": 8202
+    },
+    {
+      "epoch": 3.2072857495677947,
+      "grad_norm": 0.477804878380041,
+      "learning_rate": 4.712131612108724e-06,
+      "loss": 0.4378,
+      "step": 8203
+    },
+    {
+      "epoch": 3.207680908866387,
+      "grad_norm": 0.4842718555370584,
+      "learning_rate": 4.7120586186894626e-06,
+      "loss": 0.4488,
+      "step": 8204
+    },
+    {
+      "epoch": 3.208076068164979,
+      "grad_norm": 0.48752539718009136,
+      "learning_rate": 4.711985616582578e-06,
+      "loss": 0.4583,
+      "step": 8205
+    },
+    {
+      "epoch": 3.2084712274635714,
+      "grad_norm": 0.493674287690774,
+      "learning_rate": 4.711912605788357e-06,
+      "loss": 0.4461,
+      "step": 8206
+    },
+    {
+      "epoch": 3.2088663867621636,
+      "grad_norm": 0.5940079173022275,
+      "learning_rate": 4.711839586307086e-06,
+      "loss": 0.4796,
+      "step": 8207
+    },
+    {
+      "epoch": 3.209261546060756,
+      "grad_norm": 0.4727016450653654,
+      "learning_rate": 4.711766558139053e-06,
+      "loss": 0.4313,
+      "step": 8208
+    },
+    {
+      "epoch": 3.209656705359348,
+      "grad_norm": 0.4985753825854796,
+      "learning_rate": 4.711693521284543e-06,
+      "loss": 0.4622,
+      "step": 8209
+    },
+    {
+      "epoch": 3.2100518646579403,
+      "grad_norm": 0.5947833225156005,
+      "learning_rate": 4.711620475743845e-06,
+      "loss": 0.4576,
+      "step": 8210
+    },
+    {
+      "epoch": 3.2104470239565326,
+      "grad_norm": 0.4991842972271697,
+      "learning_rate": 4.711547421517244e-06,
+      "loss": 0.4429,
+      "step": 8211
+    },
+    {
+      "epoch": 3.210842183255125,
+      "grad_norm": 0.49646570990011646,
+      "learning_rate": 4.711474358605027e-06,
+      "loss": 0.4589,
+      "step": 8212
+    },
+    {
+      "epoch": 3.211237342553717,
+      "grad_norm": 0.49028663463091515,
+      "learning_rate": 4.711401287007482e-06,
+      "loss": 0.4321,
+      "step": 8213
+    },
+    {
+      "epoch": 3.2116325018523093,
+      "grad_norm": 0.5228254931374301,
+      "learning_rate": 4.711328206724897e-06,
+      "loss": 0.457,
+      "step": 8214
+    },
+    {
+      "epoch": 3.2120276611509015,
+      "grad_norm": 0.48805769076484595,
+      "learning_rate": 4.7112551177575564e-06,
+      "loss": 0.4406,
+      "step": 8215
+    },
+    {
+      "epoch": 3.212422820449494,
+      "grad_norm": 0.48960247238176013,
+      "learning_rate": 4.711182020105748e-06,
+      "loss": 0.4567,
+      "step": 8216
+    },
+    {
+      "epoch": 3.212817979748086,
+      "grad_norm": 0.4988948172821916,
+      "learning_rate": 4.71110891376976e-06,
+      "loss": 0.4662,
+      "step": 8217
+    },
+    {
+      "epoch": 3.2132131390466783,
+      "grad_norm": 0.502843794988854,
+      "learning_rate": 4.711035798749879e-06,
+      "loss": 0.4696,
+      "step": 8218
+    },
+    {
+      "epoch": 3.2136082983452705,
+      "grad_norm": 0.4867119474038113,
+      "learning_rate": 4.710962675046392e-06,
+      "loss": 0.4547,
+      "step": 8219
+    },
+    {
+      "epoch": 3.2140034576438627,
+      "grad_norm": 0.496603147591944,
+      "learning_rate": 4.710889542659586e-06,
+      "loss": 0.4643,
+      "step": 8220
+    },
+    {
+      "epoch": 3.214398616942455,
+      "grad_norm": 0.5108147174349558,
+      "learning_rate": 4.710816401589748e-06,
+      "loss": 0.4659,
+      "step": 8221
+    },
+    {
+      "epoch": 3.2147937762410472,
+      "grad_norm": 0.5331702363488228,
+      "learning_rate": 4.710743251837166e-06,
+      "loss": 0.4392,
+      "step": 8222
+    },
+    {
+      "epoch": 3.2151889355396395,
+      "grad_norm": 0.48329930143617106,
+      "learning_rate": 4.710670093402127e-06,
+      "loss": 0.4466,
+      "step": 8223
+    },
+    {
+      "epoch": 3.2155840948382317,
+      "grad_norm": 0.5172967429519062,
+      "learning_rate": 4.7105969262849185e-06,
+      "loss": 0.4705,
+      "step": 8224
+    },
+    {
+      "epoch": 3.215979254136824,
+      "grad_norm": 0.47907504810620555,
+      "learning_rate": 4.710523750485827e-06,
+      "loss": 0.435,
+      "step": 8225
+    },
+    {
+      "epoch": 3.216374413435416,
+      "grad_norm": 0.4878844771396085,
+      "learning_rate": 4.71045056600514e-06,
+      "loss": 0.4418,
+      "step": 8226
+    },
+    {
+      "epoch": 3.2167695727340084,
+      "grad_norm": 0.4937392208878633,
+      "learning_rate": 4.710377372843147e-06,
+      "loss": 0.4466,
+      "step": 8227
+    },
+    {
+      "epoch": 3.2171647320326007,
+      "grad_norm": 0.5193745711511666,
+      "learning_rate": 4.710304171000133e-06,
+      "loss": 0.445,
+      "step": 8228
+    },
+    {
+      "epoch": 3.217559891331193,
+      "grad_norm": 0.5017673107772458,
+      "learning_rate": 4.7102309604763865e-06,
+      "loss": 0.4535,
+      "step": 8229
+    },
+    {
+      "epoch": 3.217955050629785,
+      "grad_norm": 0.47724701872972525,
+      "learning_rate": 4.710157741272195e-06,
+      "loss": 0.4236,
+      "step": 8230
+    },
+    {
+      "epoch": 3.2183502099283774,
+      "grad_norm": 0.5029142492827028,
+      "learning_rate": 4.710084513387846e-06,
+      "loss": 0.46,
+      "step": 8231
+    },
+    {
+      "epoch": 3.2187453692269696,
+      "grad_norm": 0.4976804393277097,
+      "learning_rate": 4.710011276823627e-06,
+      "loss": 0.4601,
+      "step": 8232
+    },
+    {
+      "epoch": 3.219140528525562,
+      "grad_norm": 0.5261258641015674,
+      "learning_rate": 4.709938031579825e-06,
+      "loss": 0.4587,
+      "step": 8233
+    },
+    {
+      "epoch": 3.219535687824154,
+      "grad_norm": 0.5200984670015669,
+      "learning_rate": 4.70986477765673e-06,
+      "loss": 0.4571,
+      "step": 8234
+    },
+    {
+      "epoch": 3.2199308471227464,
+      "grad_norm": 0.4894476159865528,
+      "learning_rate": 4.709791515054627e-06,
+      "loss": 0.4442,
+      "step": 8235
+    },
+    {
+      "epoch": 3.2203260064213386,
+      "grad_norm": 0.49521421527173004,
+      "learning_rate": 4.709718243773805e-06,
+      "loss": 0.4359,
+      "step": 8236
+    },
+    {
+      "epoch": 3.220721165719931,
+      "grad_norm": 0.4927693522476051,
+      "learning_rate": 4.7096449638145504e-06,
+      "loss": 0.445,
+      "step": 8237
+    },
+    {
+      "epoch": 3.221116325018523,
+      "grad_norm": 0.5172144682625665,
+      "learning_rate": 4.709571675177154e-06,
+      "loss": 0.4584,
+      "step": 8238
+    },
+    {
+      "epoch": 3.2215114843171153,
+      "grad_norm": 0.5154003710753569,
+      "learning_rate": 4.709498377861901e-06,
+      "loss": 0.4433,
+      "step": 8239
+    },
+    {
+      "epoch": 3.2219066436157076,
+      "grad_norm": 0.4855890206340475,
+      "learning_rate": 4.709425071869079e-06,
+      "loss": 0.4501,
+      "step": 8240
+    },
+    {
+      "epoch": 3.2223018029143,
+      "grad_norm": 0.4830024442204952,
+      "learning_rate": 4.709351757198979e-06,
+      "loss": 0.4416,
+      "step": 8241
+    },
+    {
+      "epoch": 3.222696962212892,
+      "grad_norm": 0.5008192768633752,
+      "learning_rate": 4.7092784338518856e-06,
+      "loss": 0.4403,
+      "step": 8242
+    },
+    {
+      "epoch": 3.2230921215114843,
+      "grad_norm": 0.5174554129182379,
+      "learning_rate": 4.7092051018280886e-06,
+      "loss": 0.4436,
+      "step": 8243
+    },
+    {
+      "epoch": 3.2234872808100765,
+      "grad_norm": 0.49631332057852817,
+      "learning_rate": 4.709131761127875e-06,
+      "loss": 0.466,
+      "step": 8244
+    },
+    {
+      "epoch": 3.2238824401086688,
+      "grad_norm": 0.4879530471104053,
+      "learning_rate": 4.709058411751533e-06,
+      "loss": 0.4498,
+      "step": 8245
+    },
+    {
+      "epoch": 3.224277599407261,
+      "grad_norm": 0.5093815291047132,
+      "learning_rate": 4.708985053699351e-06,
+      "loss": 0.4828,
+      "step": 8246
+    },
+    {
+      "epoch": 3.2246727587058532,
+      "grad_norm": 0.48848100859749605,
+      "learning_rate": 4.708911686971618e-06,
+      "loss": 0.4388,
+      "step": 8247
+    },
+    {
+      "epoch": 3.2250679180044455,
+      "grad_norm": 0.49097358270784774,
+      "learning_rate": 4.708838311568621e-06,
+      "loss": 0.4417,
+      "step": 8248
+    },
+    {
+      "epoch": 3.2254630773030377,
+      "grad_norm": 0.5073726289639382,
+      "learning_rate": 4.708764927490648e-06,
+      "loss": 0.4715,
+      "step": 8249
+    },
+    {
+      "epoch": 3.22585823660163,
+      "grad_norm": 0.49923080085782845,
+      "learning_rate": 4.708691534737987e-06,
+      "loss": 0.4672,
+      "step": 8250
+    },
+    {
+      "epoch": 3.226253395900222,
+      "grad_norm": 0.4983917363183614,
+      "learning_rate": 4.7086181333109286e-06,
+      "loss": 0.4526,
+      "step": 8251
+    },
+    {
+      "epoch": 3.2266485551988144,
+      "grad_norm": 0.5111375734969307,
+      "learning_rate": 4.708544723209758e-06,
+      "loss": 0.452,
+      "step": 8252
+    },
+    {
+      "epoch": 3.2270437144974067,
+      "grad_norm": 0.48716751738116665,
+      "learning_rate": 4.708471304434765e-06,
+      "loss": 0.4564,
+      "step": 8253
+    },
+    {
+      "epoch": 3.227438873795999,
+      "grad_norm": 0.5069026621643061,
+      "learning_rate": 4.708397876986238e-06,
+      "loss": 0.4528,
+      "step": 8254
+    },
+    {
+      "epoch": 3.227834033094591,
+      "grad_norm": 0.5152378887836355,
+      "learning_rate": 4.708324440864465e-06,
+      "loss": 0.4355,
+      "step": 8255
+    },
+    {
+      "epoch": 3.2282291923931834,
+      "grad_norm": 0.4962324262994311,
+      "learning_rate": 4.7082509960697345e-06,
+      "loss": 0.442,
+      "step": 8256
+    },
+    {
+      "epoch": 3.2286243516917756,
+      "grad_norm": 0.5028634303397124,
+      "learning_rate": 4.708177542602335e-06,
+      "loss": 0.4553,
+      "step": 8257
+    },
+    {
+      "epoch": 3.229019510990368,
+      "grad_norm": 0.5001078547401838,
+      "learning_rate": 4.708104080462555e-06,
+      "loss": 0.4557,
+      "step": 8258
+    },
+    {
+      "epoch": 3.22941467028896,
+      "grad_norm": 0.4960036879535328,
+      "learning_rate": 4.708030609650683e-06,
+      "loss": 0.448,
+      "step": 8259
+    },
+    {
+      "epoch": 3.2298098295875524,
+      "grad_norm": 0.5050479554627619,
+      "learning_rate": 4.7079571301670076e-06,
+      "loss": 0.4655,
+      "step": 8260
+    },
+    {
+      "epoch": 3.2302049888861446,
+      "grad_norm": 0.5006203296479564,
+      "learning_rate": 4.707883642011818e-06,
+      "loss": 0.4653,
+      "step": 8261
+    },
+    {
+      "epoch": 3.230600148184737,
+      "grad_norm": 0.4999078554437849,
+      "learning_rate": 4.707810145185401e-06,
+      "loss": 0.4574,
+      "step": 8262
+    },
+    {
+      "epoch": 3.230995307483329,
+      "grad_norm": 0.47956978838455655,
+      "learning_rate": 4.707736639688047e-06,
+      "loss": 0.4366,
+      "step": 8263
+    },
+    {
+      "epoch": 3.2313904667819213,
+      "grad_norm": 0.5070009335133143,
+      "learning_rate": 4.707663125520044e-06,
+      "loss": 0.4479,
+      "step": 8264
+    },
+    {
+      "epoch": 3.2317856260805136,
+      "grad_norm": 0.5087445068694287,
+      "learning_rate": 4.707589602681681e-06,
+      "loss": 0.4596,
+      "step": 8265
+    },
+    {
+      "epoch": 3.232180785379106,
+      "grad_norm": 0.5718442258693521,
+      "learning_rate": 4.707516071173246e-06,
+      "loss": 0.4542,
+      "step": 8266
+    },
+    {
+      "epoch": 3.232575944677698,
+      "grad_norm": 0.5011292864269645,
+      "learning_rate": 4.707442530995029e-06,
+      "loss": 0.4525,
+      "step": 8267
+    },
+    {
+      "epoch": 3.2329711039762903,
+      "grad_norm": 0.5026210431492106,
+      "learning_rate": 4.707368982147318e-06,
+      "loss": 0.4431,
+      "step": 8268
+    },
+    {
+      "epoch": 3.2333662632748825,
+      "grad_norm": 0.5127765787850526,
+      "learning_rate": 4.707295424630402e-06,
+      "loss": 0.4665,
+      "step": 8269
+    },
+    {
+      "epoch": 3.2337614225734748,
+      "grad_norm": 0.5086511264451964,
+      "learning_rate": 4.707221858444569e-06,
+      "loss": 0.45,
+      "step": 8270
+    },
+    {
+      "epoch": 3.234156581872067,
+      "grad_norm": 0.5232352547690865,
+      "learning_rate": 4.70714828359011e-06,
+      "loss": 0.4685,
+      "step": 8271
+    },
+    {
+      "epoch": 3.2345517411706592,
+      "grad_norm": 0.49163545831580857,
+      "learning_rate": 4.707074700067312e-06,
+      "loss": 0.4508,
+      "step": 8272
+    },
+    {
+      "epoch": 3.2349469004692515,
+      "grad_norm": 0.5145261335572505,
+      "learning_rate": 4.707001107876466e-06,
+      "loss": 0.4693,
+      "step": 8273
+    },
+    {
+      "epoch": 3.2353420597678437,
+      "grad_norm": 0.5026474655303093,
+      "learning_rate": 4.7069275070178586e-06,
+      "loss": 0.4629,
+      "step": 8274
+    },
+    {
+      "epoch": 3.235737219066436,
+      "grad_norm": 0.486565673024086,
+      "learning_rate": 4.706853897491781e-06,
+      "loss": 0.4526,
+      "step": 8275
+    },
+    {
+      "epoch": 3.236132378365028,
+      "grad_norm": 0.47935306570145486,
+      "learning_rate": 4.70678027929852e-06,
+      "loss": 0.4466,
+      "step": 8276
+    },
+    {
+      "epoch": 3.2365275376636204,
+      "grad_norm": 0.4991778284300973,
+      "learning_rate": 4.706706652438368e-06,
+      "loss": 0.4549,
+      "step": 8277
+    },
+    {
+      "epoch": 3.2369226969622127,
+      "grad_norm": 0.5067651241856679,
+      "learning_rate": 4.706633016911611e-06,
+      "loss": 0.4634,
+      "step": 8278
+    },
+    {
+      "epoch": 3.2373178562608054,
+      "grad_norm": 0.49959265144088416,
+      "learning_rate": 4.70655937271854e-06,
+      "loss": 0.4477,
+      "step": 8279
+    },
+    {
+      "epoch": 3.2377130155593976,
+      "grad_norm": 0.4730123377567548,
+      "learning_rate": 4.706485719859443e-06,
+      "loss": 0.4348,
+      "step": 8280
+    },
+    {
+      "epoch": 3.23810817485799,
+      "grad_norm": 0.5011163271286927,
+      "learning_rate": 4.706412058334611e-06,
+      "loss": 0.469,
+      "step": 8281
+    },
+    {
+      "epoch": 3.238503334156582,
+      "grad_norm": 0.5118563019208421,
+      "learning_rate": 4.706338388144331e-06,
+      "loss": 0.4536,
+      "step": 8282
+    },
+    {
+      "epoch": 3.2388984934551743,
+      "grad_norm": 0.49785560194791934,
+      "learning_rate": 4.706264709288894e-06,
+      "loss": 0.4398,
+      "step": 8283
+    },
+    {
+      "epoch": 3.2392936527537666,
+      "grad_norm": 0.48918931904116053,
+      "learning_rate": 4.7061910217685895e-06,
+      "loss": 0.458,
+      "step": 8284
+    },
+    {
+      "epoch": 3.239688812052359,
+      "grad_norm": 0.4991786615442405,
+      "learning_rate": 4.7061173255837054e-06,
+      "loss": 0.4522,
+      "step": 8285
+    },
+    {
+      "epoch": 3.240083971350951,
+      "grad_norm": 0.48813754093841744,
+      "learning_rate": 4.706043620734533e-06,
+      "loss": 0.454,
+      "step": 8286
+    },
+    {
+      "epoch": 3.2404791306495433,
+      "grad_norm": 0.6162776390081206,
+      "learning_rate": 4.705969907221361e-06,
+      "loss": 0.4252,
+      "step": 8287
+    },
+    {
+      "epoch": 3.2408742899481355,
+      "grad_norm": 0.4916252788914689,
+      "learning_rate": 4.705896185044477e-06,
+      "loss": 0.4541,
+      "step": 8288
+    },
+    {
+      "epoch": 3.2412694492467278,
+      "grad_norm": 0.49006414916965146,
+      "learning_rate": 4.705822454204173e-06,
+      "loss": 0.4498,
+      "step": 8289
+    },
+    {
+      "epoch": 3.24166460854532,
+      "grad_norm": 0.5058325689652357,
+      "learning_rate": 4.705748714700739e-06,
+      "loss": 0.45,
+      "step": 8290
+    },
+    {
+      "epoch": 3.2420597678439123,
+      "grad_norm": 0.4876491731015844,
+      "learning_rate": 4.705674966534462e-06,
+      "loss": 0.4163,
+      "step": 8291
+    },
+    {
+      "epoch": 3.2424549271425045,
+      "grad_norm": 0.5021867760606261,
+      "learning_rate": 4.705601209705635e-06,
+      "loss": 0.441,
+      "step": 8292
+    },
+    {
+      "epoch": 3.2428500864410967,
+      "grad_norm": 0.5038352555082803,
+      "learning_rate": 4.705527444214543e-06,
+      "loss": 0.4469,
+      "step": 8293
+    },
+    {
+      "epoch": 3.243245245739689,
+      "grad_norm": 0.5013117536018146,
+      "learning_rate": 4.705453670061481e-06,
+      "loss": 0.4492,
+      "step": 8294
+    },
+    {
+      "epoch": 3.243640405038281,
+      "grad_norm": 0.47282887919141153,
+      "learning_rate": 4.705379887246735e-06,
+      "loss": 0.4431,
+      "step": 8295
+    },
+    {
+      "epoch": 3.2440355643368735,
+      "grad_norm": 0.49337905075174315,
+      "learning_rate": 4.705306095770596e-06,
+      "loss": 0.4572,
+      "step": 8296
+    },
+    {
+      "epoch": 3.2444307236354657,
+      "grad_norm": 0.49412301657250673,
+      "learning_rate": 4.705232295633355e-06,
+      "loss": 0.4463,
+      "step": 8297
+    },
+    {
+      "epoch": 3.244825882934058,
+      "grad_norm": 0.5031343987855506,
+      "learning_rate": 4.7051584868353e-06,
+      "loss": 0.4502,
+      "step": 8298
+    },
+    {
+      "epoch": 3.24522104223265,
+      "grad_norm": 0.5181152188198329,
+      "learning_rate": 4.705084669376721e-06,
+      "loss": 0.4397,
+      "step": 8299
+    },
+    {
+      "epoch": 3.2456162015312424,
+      "grad_norm": 0.4882928530529042,
+      "learning_rate": 4.7050108432579075e-06,
+      "loss": 0.4515,
+      "step": 8300
+    },
+    {
+      "epoch": 3.2460113608298347,
+      "grad_norm": 0.6015511647794045,
+      "learning_rate": 4.704937008479152e-06,
+      "loss": 0.4449,
+      "step": 8301
+    },
+    {
+      "epoch": 3.246406520128427,
+      "grad_norm": 0.5137381986143884,
+      "learning_rate": 4.704863165040742e-06,
+      "loss": 0.4537,
+      "step": 8302
+    },
+    {
+      "epoch": 3.246801679427019,
+      "grad_norm": 0.5065385080859512,
+      "learning_rate": 4.704789312942969e-06,
+      "loss": 0.4507,
+      "step": 8303
+    },
+    {
+      "epoch": 3.2471968387256114,
+      "grad_norm": 0.5006646215975414,
+      "learning_rate": 4.704715452186122e-06,
+      "loss": 0.4413,
+      "step": 8304
+    },
+    {
+      "epoch": 3.2475919980242036,
+      "grad_norm": 0.4925281165489277,
+      "learning_rate": 4.704641582770492e-06,
+      "loss": 0.4418,
+      "step": 8305
+    },
+    {
+      "epoch": 3.247987157322796,
+      "grad_norm": 0.4935778234372788,
+      "learning_rate": 4.704567704696368e-06,
+      "loss": 0.465,
+      "step": 8306
+    },
+    {
+      "epoch": 3.248382316621388,
+      "grad_norm": 0.5077023224379424,
+      "learning_rate": 4.704493817964041e-06,
+      "loss": 0.4327,
+      "step": 8307
+    },
+    {
+      "epoch": 3.2487774759199803,
+      "grad_norm": 0.5015739358541178,
+      "learning_rate": 4.7044199225738e-06,
+      "loss": 0.4464,
+      "step": 8308
+    },
+    {
+      "epoch": 3.2491726352185726,
+      "grad_norm": 0.5022006581016725,
+      "learning_rate": 4.704346018525937e-06,
+      "loss": 0.4532,
+      "step": 8309
+    },
+    {
+      "epoch": 3.249567794517165,
+      "grad_norm": 0.5067055633348441,
+      "learning_rate": 4.704272105820741e-06,
+      "loss": 0.4597,
+      "step": 8310
+    },
+    {
+      "epoch": 3.249962953815757,
+      "grad_norm": 0.4829044817680482,
+      "learning_rate": 4.704198184458503e-06,
+      "loss": 0.4245,
+      "step": 8311
+    },
+    {
+      "epoch": 3.2503581131143493,
+      "grad_norm": 0.502784983843347,
+      "learning_rate": 4.704124254439512e-06,
+      "loss": 0.4682,
+      "step": 8312
+    },
+    {
+      "epoch": 3.2507532724129415,
+      "grad_norm": 0.5025327631494758,
+      "learning_rate": 4.704050315764062e-06,
+      "loss": 0.4532,
+      "step": 8313
+    },
+    {
+      "epoch": 3.251148431711534,
+      "grad_norm": 0.4827438502841128,
+      "learning_rate": 4.703976368432438e-06,
+      "loss": 0.4582,
+      "step": 8314
+    },
+    {
+      "epoch": 3.251543591010126,
+      "grad_norm": 0.5001402118199502,
+      "learning_rate": 4.703902412444935e-06,
+      "loss": 0.4512,
+      "step": 8315
+    },
+    {
+      "epoch": 3.2519387503087183,
+      "grad_norm": 0.5073201743090425,
+      "learning_rate": 4.7038284478018405e-06,
+      "loss": 0.458,
+      "step": 8316
+    },
+    {
+      "epoch": 3.2523339096073105,
+      "grad_norm": 0.5044450071518284,
+      "learning_rate": 4.703754474503446e-06,
+      "loss": 0.4518,
+      "step": 8317
+    },
+    {
+      "epoch": 3.2527290689059027,
+      "grad_norm": 0.5019360105195412,
+      "learning_rate": 4.703680492550043e-06,
+      "loss": 0.4544,
+      "step": 8318
+    },
+    {
+      "epoch": 3.253124228204495,
+      "grad_norm": 0.49174623590569816,
+      "learning_rate": 4.703606501941921e-06,
+      "loss": 0.4361,
+      "step": 8319
+    },
+    {
+      "epoch": 3.2535193875030872,
+      "grad_norm": 0.5000322070806993,
+      "learning_rate": 4.70353250267937e-06,
+      "loss": 0.4647,
+      "step": 8320
+    },
+    {
+      "epoch": 3.2539145468016795,
+      "grad_norm": 0.4949672092349716,
+      "learning_rate": 4.7034584947626815e-06,
+      "loss": 0.4498,
+      "step": 8321
+    },
+    {
+      "epoch": 3.2543097061002717,
+      "grad_norm": 0.48215269520295684,
+      "learning_rate": 4.703384478192146e-06,
+      "loss": 0.4461,
+      "step": 8322
+    },
+    {
+      "epoch": 3.254704865398864,
+      "grad_norm": 0.4874422999560892,
+      "learning_rate": 4.703310452968055e-06,
+      "loss": 0.4434,
+      "step": 8323
+    },
+    {
+      "epoch": 3.255100024697456,
+      "grad_norm": 0.4842366003915545,
+      "learning_rate": 4.703236419090699e-06,
+      "loss": 0.4338,
+      "step": 8324
+    },
+    {
+      "epoch": 3.2554951839960484,
+      "grad_norm": 0.47835189607454215,
+      "learning_rate": 4.7031623765603665e-06,
+      "loss": 0.4463,
+      "step": 8325
+    },
+    {
+      "epoch": 3.2558903432946407,
+      "grad_norm": 0.48308358621451114,
+      "learning_rate": 4.7030883253773504e-06,
+      "loss": 0.455,
+      "step": 8326
+    },
+    {
+      "epoch": 3.256285502593233,
+      "grad_norm": 0.5094922118559343,
+      "learning_rate": 4.703014265541942e-06,
+      "loss": 0.447,
+      "step": 8327
+    },
+    {
+      "epoch": 3.256680661891825,
+      "grad_norm": 0.49550690331076896,
+      "learning_rate": 4.702940197054431e-06,
+      "loss": 0.4543,
+      "step": 8328
+    },
+    {
+      "epoch": 3.2570758211904174,
+      "grad_norm": 0.49464702815935846,
+      "learning_rate": 4.702866119915108e-06,
+      "loss": 0.443,
+      "step": 8329
+    },
+    {
+      "epoch": 3.2574709804890096,
+      "grad_norm": 0.48344035746473757,
+      "learning_rate": 4.702792034124265e-06,
+      "loss": 0.4422,
+      "step": 8330
+    },
+    {
+      "epoch": 3.257866139787602,
+      "grad_norm": 0.49796852073676307,
+      "learning_rate": 4.702717939682193e-06,
+      "loss": 0.4518,
+      "step": 8331
+    },
+    {
+      "epoch": 3.258261299086194,
+      "grad_norm": 0.49891388022773675,
+      "learning_rate": 4.702643836589182e-06,
+      "loss": 0.4536,
+      "step": 8332
+    },
+    {
+      "epoch": 3.2586564583847863,
+      "grad_norm": 0.5017290426062102,
+      "learning_rate": 4.702569724845523e-06,
+      "loss": 0.4492,
+      "step": 8333
+    },
+    {
+      "epoch": 3.2590516176833786,
+      "grad_norm": 0.5055247142680904,
+      "learning_rate": 4.702495604451508e-06,
+      "loss": 0.4646,
+      "step": 8334
+    },
+    {
+      "epoch": 3.259446776981971,
+      "grad_norm": 0.48822789965751323,
+      "learning_rate": 4.702421475407428e-06,
+      "loss": 0.4509,
+      "step": 8335
+    },
+    {
+      "epoch": 3.259841936280563,
+      "grad_norm": 0.4932198897185812,
+      "learning_rate": 4.7023473377135735e-06,
+      "loss": 0.4464,
+      "step": 8336
+    },
+    {
+      "epoch": 3.2602370955791553,
+      "grad_norm": 0.5137369946580147,
+      "learning_rate": 4.702273191370236e-06,
+      "loss": 0.452,
+      "step": 8337
+    },
+    {
+      "epoch": 3.2606322548777475,
+      "grad_norm": 0.5049765007477618,
+      "learning_rate": 4.702199036377707e-06,
+      "loss": 0.4344,
+      "step": 8338
+    },
+    {
+      "epoch": 3.26102741417634,
+      "grad_norm": 0.48715675795622143,
+      "learning_rate": 4.702124872736277e-06,
+      "loss": 0.4688,
+      "step": 8339
+    },
+    {
+      "epoch": 3.261422573474932,
+      "grad_norm": 0.5058798989467141,
+      "learning_rate": 4.702050700446238e-06,
+      "loss": 0.461,
+      "step": 8340
+    },
+    {
+      "epoch": 3.2618177327735243,
+      "grad_norm": 0.4875539335588554,
+      "learning_rate": 4.701976519507881e-06,
+      "loss": 0.4517,
+      "step": 8341
+    },
+    {
+      "epoch": 3.2622128920721165,
+      "grad_norm": 0.49540858670627236,
+      "learning_rate": 4.7019023299214974e-06,
+      "loss": 0.4436,
+      "step": 8342
+    },
+    {
+      "epoch": 3.2626080513707088,
+      "grad_norm": 0.48452284870049883,
+      "learning_rate": 4.701828131687378e-06,
+      "loss": 0.4563,
+      "step": 8343
+    },
+    {
+      "epoch": 3.263003210669301,
+      "grad_norm": 0.5064967286836504,
+      "learning_rate": 4.701753924805815e-06,
+      "loss": 0.4517,
+      "step": 8344
+    },
+    {
+      "epoch": 3.2633983699678932,
+      "grad_norm": 0.5049107650163969,
+      "learning_rate": 4.7016797092771e-06,
+      "loss": 0.4641,
+      "step": 8345
+    },
+    {
+      "epoch": 3.2637935292664855,
+      "grad_norm": 0.5286126343914365,
+      "learning_rate": 4.701605485101524e-06,
+      "loss": 0.4564,
+      "step": 8346
+    },
+    {
+      "epoch": 3.2641886885650777,
+      "grad_norm": 0.5224526170379453,
+      "learning_rate": 4.701531252279379e-06,
+      "loss": 0.4418,
+      "step": 8347
+    },
+    {
+      "epoch": 3.26458384786367,
+      "grad_norm": 0.5085290561882606,
+      "learning_rate": 4.701457010810955e-06,
+      "loss": 0.4561,
+      "step": 8348
+    },
+    {
+      "epoch": 3.264979007162262,
+      "grad_norm": 0.4860433808053769,
+      "learning_rate": 4.7013827606965446e-06,
+      "loss": 0.4365,
+      "step": 8349
+    },
+    {
+      "epoch": 3.2653741664608544,
+      "grad_norm": 0.5362280369050675,
+      "learning_rate": 4.70130850193644e-06,
+      "loss": 0.453,
+      "step": 8350
+    },
+    {
+      "epoch": 3.2657693257594467,
+      "grad_norm": 0.5026775324033125,
+      "learning_rate": 4.7012342345309316e-06,
+      "loss": 0.4318,
+      "step": 8351
+    },
+    {
+      "epoch": 3.266164485058039,
+      "grad_norm": 0.48284490669371327,
+      "learning_rate": 4.701159958480312e-06,
+      "loss": 0.4375,
+      "step": 8352
+    },
+    {
+      "epoch": 3.266559644356631,
+      "grad_norm": 0.511556437758763,
+      "learning_rate": 4.701085673784874e-06,
+      "loss": 0.4607,
+      "step": 8353
+    },
+    {
+      "epoch": 3.2669548036552234,
+      "grad_norm": 0.4829094276892418,
+      "learning_rate": 4.701011380444907e-06,
+      "loss": 0.4292,
+      "step": 8354
+    },
+    {
+      "epoch": 3.2673499629538156,
+      "grad_norm": 0.51144228894575,
+      "learning_rate": 4.700937078460704e-06,
+      "loss": 0.4603,
+      "step": 8355
+    },
+    {
+      "epoch": 3.267745122252408,
+      "grad_norm": 0.4906533832218431,
+      "learning_rate": 4.700862767832557e-06,
+      "loss": 0.4582,
+      "step": 8356
+    },
+    {
+      "epoch": 3.268140281551,
+      "grad_norm": 0.4930805349262283,
+      "learning_rate": 4.7007884485607565e-06,
+      "loss": 0.458,
+      "step": 8357
+    },
+    {
+      "epoch": 3.2685354408495924,
+      "grad_norm": 0.5009831870584155,
+      "learning_rate": 4.700714120645596e-06,
+      "loss": 0.4687,
+      "step": 8358
+    },
+    {
+      "epoch": 3.2689306001481846,
+      "grad_norm": 0.49092555374069324,
+      "learning_rate": 4.700639784087366e-06,
+      "loss": 0.4359,
+      "step": 8359
+    },
+    {
+      "epoch": 3.269325759446777,
+      "grad_norm": 0.500292655543431,
+      "learning_rate": 4.700565438886361e-06,
+      "loss": 0.4487,
+      "step": 8360
+    },
+    {
+      "epoch": 3.269720918745369,
+      "grad_norm": 0.4855070468612074,
+      "learning_rate": 4.70049108504287e-06,
+      "loss": 0.4411,
+      "step": 8361
+    },
+    {
+      "epoch": 3.2701160780439613,
+      "grad_norm": 0.5023057055550552,
+      "learning_rate": 4.700416722557186e-06,
+      "loss": 0.4455,
+      "step": 8362
+    },
+    {
+      "epoch": 3.2705112373425536,
+      "grad_norm": 0.509131608944125,
+      "learning_rate": 4.700342351429601e-06,
+      "loss": 0.4516,
+      "step": 8363
+    },
+    {
+      "epoch": 3.270906396641146,
+      "grad_norm": 0.49337058153224017,
+      "learning_rate": 4.700267971660408e-06,
+      "loss": 0.4675,
+      "step": 8364
+    },
+    {
+      "epoch": 3.271301555939738,
+      "grad_norm": 0.4860684297948216,
+      "learning_rate": 4.700193583249899e-06,
+      "loss": 0.4467,
+      "step": 8365
+    },
+    {
+      "epoch": 3.2716967152383303,
+      "grad_norm": 0.49228337368850594,
+      "learning_rate": 4.700119186198365e-06,
+      "loss": 0.4579,
+      "step": 8366
+    },
+    {
+      "epoch": 3.2720918745369225,
+      "grad_norm": 0.4845358764762339,
+      "learning_rate": 4.700044780506099e-06,
+      "loss": 0.449,
+      "step": 8367
+    },
+    {
+      "epoch": 3.2724870338355148,
+      "grad_norm": 0.5068451001379258,
+      "learning_rate": 4.699970366173393e-06,
+      "loss": 0.4348,
+      "step": 8368
+    },
+    {
+      "epoch": 3.272882193134107,
+      "grad_norm": 0.4914022566989148,
+      "learning_rate": 4.699895943200539e-06,
+      "loss": 0.4503,
+      "step": 8369
+    },
+    {
+      "epoch": 3.2732773524326992,
+      "grad_norm": 0.4992738042869505,
+      "learning_rate": 4.69982151158783e-06,
+      "loss": 0.4432,
+      "step": 8370
+    },
+    {
+      "epoch": 3.2736725117312915,
+      "grad_norm": 0.4938604279465015,
+      "learning_rate": 4.6997470713355574e-06,
+      "loss": 0.4414,
+      "step": 8371
+    },
+    {
+      "epoch": 3.2740676710298837,
+      "grad_norm": 0.4839800775793246,
+      "learning_rate": 4.699672622444015e-06,
+      "loss": 0.4349,
+      "step": 8372
+    },
+    {
+      "epoch": 3.274462830328476,
+      "grad_norm": 0.5024998565000468,
+      "learning_rate": 4.699598164913493e-06,
+      "loss": 0.4402,
+      "step": 8373
+    },
+    {
+      "epoch": 3.274857989627068,
+      "grad_norm": 0.49637224096799104,
+      "learning_rate": 4.699523698744286e-06,
+      "loss": 0.4483,
+      "step": 8374
+    },
+    {
+      "epoch": 3.2752531489256604,
+      "grad_norm": 0.49260375994924005,
+      "learning_rate": 4.699449223936686e-06,
+      "loss": 0.4474,
+      "step": 8375
+    },
+    {
+      "epoch": 3.2756483082242527,
+      "grad_norm": 0.6533471619862731,
+      "learning_rate": 4.699374740490984e-06,
+      "loss": 0.4655,
+      "step": 8376
+    },
+    {
+      "epoch": 3.276043467522845,
+      "grad_norm": 0.5234139605987889,
+      "learning_rate": 4.699300248407474e-06,
+      "loss": 0.4672,
+      "step": 8377
+    },
+    {
+      "epoch": 3.276438626821437,
+      "grad_norm": 0.4971511724210842,
+      "learning_rate": 4.6992257476864475e-06,
+      "loss": 0.4598,
+      "step": 8378
+    },
+    {
+      "epoch": 3.2768337861200294,
+      "grad_norm": 0.49100800872845307,
+      "learning_rate": 4.699151238328198e-06,
+      "loss": 0.47,
+      "step": 8379
+    },
+    {
+      "epoch": 3.2772289454186216,
+      "grad_norm": 0.4796505235609134,
+      "learning_rate": 4.6990767203330185e-06,
+      "loss": 0.4321,
+      "step": 8380
+    },
+    {
+      "epoch": 3.277624104717214,
+      "grad_norm": 0.5035218618276991,
+      "learning_rate": 4.6990021937012e-06,
+      "loss": 0.4544,
+      "step": 8381
+    },
+    {
+      "epoch": 3.278019264015806,
+      "grad_norm": 0.4912630675638253,
+      "learning_rate": 4.6989276584330365e-06,
+      "loss": 0.4311,
+      "step": 8382
+    },
+    {
+      "epoch": 3.278414423314399,
+      "grad_norm": 0.5050629349004524,
+      "learning_rate": 4.6988531145288204e-06,
+      "loss": 0.4416,
+      "step": 8383
+    },
+    {
+      "epoch": 3.278809582612991,
+      "grad_norm": 0.5106795613363712,
+      "learning_rate": 4.698778561988845e-06,
+      "loss": 0.4728,
+      "step": 8384
+    },
+    {
+      "epoch": 3.2792047419115833,
+      "grad_norm": 0.4961344250576797,
+      "learning_rate": 4.698704000813403e-06,
+      "loss": 0.4555,
+      "step": 8385
+    },
+    {
+      "epoch": 3.2795999012101755,
+      "grad_norm": 0.5105963304245603,
+      "learning_rate": 4.698629431002786e-06,
+      "loss": 0.4758,
+      "step": 8386
+    },
+    {
+      "epoch": 3.2799950605087678,
+      "grad_norm": 0.49603685469527325,
+      "learning_rate": 4.698554852557288e-06,
+      "loss": 0.4394,
+      "step": 8387
+    },
+    {
+      "epoch": 3.28039021980736,
+      "grad_norm": 0.51768754874356,
+      "learning_rate": 4.698480265477201e-06,
+      "loss": 0.4477,
+      "step": 8388
+    },
+    {
+      "epoch": 3.2807853791059522,
+      "grad_norm": 0.4988292744521313,
+      "learning_rate": 4.698405669762819e-06,
+      "loss": 0.4575,
+      "step": 8389
+    },
+    {
+      "epoch": 3.2811805384045445,
+      "grad_norm": 0.4906957305488342,
+      "learning_rate": 4.6983310654144345e-06,
+      "loss": 0.4463,
+      "step": 8390
+    },
+    {
+      "epoch": 3.2815756977031367,
+      "grad_norm": 0.4974751379220848,
+      "learning_rate": 4.6982564524323396e-06,
+      "loss": 0.4454,
+      "step": 8391
+    },
+    {
+      "epoch": 3.281970857001729,
+      "grad_norm": 0.4872843211030378,
+      "learning_rate": 4.698181830816829e-06,
+      "loss": 0.4311,
+      "step": 8392
+    },
+    {
+      "epoch": 3.282366016300321,
+      "grad_norm": 0.49418859390881315,
+      "learning_rate": 4.698107200568195e-06,
+      "loss": 0.4415,
+      "step": 8393
+    },
+    {
+      "epoch": 3.2827611755989135,
+      "grad_norm": 0.5037917521978581,
+      "learning_rate": 4.698032561686731e-06,
+      "loss": 0.4554,
+      "step": 8394
+    },
+    {
+      "epoch": 3.2831563348975057,
+      "grad_norm": 0.643760942178025,
+      "learning_rate": 4.697957914172728e-06,
+      "loss": 0.444,
+      "step": 8395
+    },
+    {
+      "epoch": 3.283551494196098,
+      "grad_norm": 0.49670328228290767,
+      "learning_rate": 4.697883258026483e-06,
+      "loss": 0.4514,
+      "step": 8396
+    },
+    {
+      "epoch": 3.28394665349469,
+      "grad_norm": 0.49682146132488997,
+      "learning_rate": 4.697808593248287e-06,
+      "loss": 0.4398,
+      "step": 8397
+    },
+    {
+      "epoch": 3.2843418127932824,
+      "grad_norm": 0.5166333354777709,
+      "learning_rate": 4.697733919838433e-06,
+      "loss": 0.4618,
+      "step": 8398
+    },
+    {
+      "epoch": 3.2847369720918747,
+      "grad_norm": 0.4806098054963327,
+      "learning_rate": 4.697659237797214e-06,
+      "loss": 0.4501,
+      "step": 8399
+    },
+    {
+      "epoch": 3.285132131390467,
+      "grad_norm": 0.49062496009639534,
+      "learning_rate": 4.697584547124925e-06,
+      "loss": 0.4476,
+      "step": 8400
+    },
+    {
+      "epoch": 3.285527290689059,
+      "grad_norm": 0.5057710961550459,
+      "learning_rate": 4.697509847821858e-06,
+      "loss": 0.4487,
+      "step": 8401
+    },
+    {
+      "epoch": 3.2859224499876514,
+      "grad_norm": 0.5096432840136818,
+      "learning_rate": 4.697435139888307e-06,
+      "loss": 0.4468,
+      "step": 8402
+    },
+    {
+      "epoch": 3.2863176092862436,
+      "grad_norm": 0.5099785625996293,
+      "learning_rate": 4.697360423324564e-06,
+      "loss": 0.4736,
+      "step": 8403
+    },
+    {
+      "epoch": 3.286712768584836,
+      "grad_norm": 0.4840586188305653,
+      "learning_rate": 4.6972856981309245e-06,
+      "loss": 0.4569,
+      "step": 8404
+    },
+    {
+      "epoch": 3.287107927883428,
+      "grad_norm": 0.49475695398984937,
+      "learning_rate": 4.697210964307681e-06,
+      "loss": 0.465,
+      "step": 8405
+    },
+    {
+      "epoch": 3.2875030871820203,
+      "grad_norm": 0.5199796400658511,
+      "learning_rate": 4.697136221855127e-06,
+      "loss": 0.4667,
+      "step": 8406
+    },
+    {
+      "epoch": 3.2878982464806126,
+      "grad_norm": 0.5003237789978858,
+      "learning_rate": 4.697061470773556e-06,
+      "loss": 0.457,
+      "step": 8407
+    },
+    {
+      "epoch": 3.288293405779205,
+      "grad_norm": 0.49819670220659273,
+      "learning_rate": 4.696986711063261e-06,
+      "loss": 0.4494,
+      "step": 8408
+    },
+    {
+      "epoch": 3.288688565077797,
+      "grad_norm": 0.49271546941029926,
+      "learning_rate": 4.696911942724538e-06,
+      "loss": 0.458,
+      "step": 8409
+    },
+    {
+      "epoch": 3.2890837243763893,
+      "grad_norm": 0.5002339935967224,
+      "learning_rate": 4.6968371657576774e-06,
+      "loss": 0.4463,
+      "step": 8410
+    },
+    {
+      "epoch": 3.2894788836749815,
+      "grad_norm": 0.49334667002247623,
+      "learning_rate": 4.696762380162975e-06,
+      "loss": 0.4486,
+      "step": 8411
+    },
+    {
+      "epoch": 3.2898740429735738,
+      "grad_norm": 0.496514797921124,
+      "learning_rate": 4.696687585940724e-06,
+      "loss": 0.4538,
+      "step": 8412
+    },
+    {
+      "epoch": 3.290269202272166,
+      "grad_norm": 0.4964892015506934,
+      "learning_rate": 4.696612783091218e-06,
+      "loss": 0.4445,
+      "step": 8413
+    },
+    {
+      "epoch": 3.2906643615707583,
+      "grad_norm": 0.5015291860594303,
+      "learning_rate": 4.696537971614751e-06,
+      "loss": 0.4537,
+      "step": 8414
+    },
+    {
+      "epoch": 3.2910595208693505,
+      "grad_norm": 0.4896264954525798,
+      "learning_rate": 4.696463151511617e-06,
+      "loss": 0.4516,
+      "step": 8415
+    },
+    {
+      "epoch": 3.2914546801679427,
+      "grad_norm": 0.5160967672675386,
+      "learning_rate": 4.696388322782108e-06,
+      "loss": 0.472,
+      "step": 8416
+    },
+    {
+      "epoch": 3.291849839466535,
+      "grad_norm": 0.5176580403198294,
+      "learning_rate": 4.69631348542652e-06,
+      "loss": 0.4505,
+      "step": 8417
+    },
+    {
+      "epoch": 3.292244998765127,
+      "grad_norm": 0.5070289013247464,
+      "learning_rate": 4.696238639445147e-06,
+      "loss": 0.4678,
+      "step": 8418
+    },
+    {
+      "epoch": 3.2926401580637195,
+      "grad_norm": 0.5280810466683207,
+      "learning_rate": 4.696163784838282e-06,
+      "loss": 0.4643,
+      "step": 8419
+    },
+    {
+      "epoch": 3.2930353173623117,
+      "grad_norm": 0.5182007906795619,
+      "learning_rate": 4.696088921606219e-06,
+      "loss": 0.4367,
+      "step": 8420
+    },
+    {
+      "epoch": 3.293430476660904,
+      "grad_norm": 0.49458421124688884,
+      "learning_rate": 4.696014049749252e-06,
+      "loss": 0.464,
+      "step": 8421
+    },
+    {
+      "epoch": 3.293825635959496,
+      "grad_norm": 0.5055857636265902,
+      "learning_rate": 4.695939169267676e-06,
+      "loss": 0.4492,
+      "step": 8422
+    },
+    {
+      "epoch": 3.2942207952580884,
+      "grad_norm": 0.5004862392256088,
+      "learning_rate": 4.695864280161784e-06,
+      "loss": 0.4587,
+      "step": 8423
+    },
+    {
+      "epoch": 3.2946159545566807,
+      "grad_norm": 0.484703488880412,
+      "learning_rate": 4.695789382431871e-06,
+      "loss": 0.4476,
+      "step": 8424
+    },
+    {
+      "epoch": 3.295011113855273,
+      "grad_norm": 0.4945224351875782,
+      "learning_rate": 4.6957144760782305e-06,
+      "loss": 0.4487,
+      "step": 8425
+    },
+    {
+      "epoch": 3.295406273153865,
+      "grad_norm": 0.47918483340578955,
+      "learning_rate": 4.695639561101156e-06,
+      "loss": 0.4509,
+      "step": 8426
+    },
+    {
+      "epoch": 3.2958014324524574,
+      "grad_norm": 0.523558981494999,
+      "learning_rate": 4.695564637500944e-06,
+      "loss": 0.4584,
+      "step": 8427
+    },
+    {
+      "epoch": 3.2961965917510496,
+      "grad_norm": 0.5135262004505718,
+      "learning_rate": 4.695489705277886e-06,
+      "loss": 0.4713,
+      "step": 8428
+    },
+    {
+      "epoch": 3.296591751049642,
+      "grad_norm": 0.5032512376213754,
+      "learning_rate": 4.695414764432278e-06,
+      "loss": 0.4609,
+      "step": 8429
+    },
+    {
+      "epoch": 3.296986910348234,
+      "grad_norm": 0.515575257988458,
+      "learning_rate": 4.6953398149644135e-06,
+      "loss": 0.4535,
+      "step": 8430
+    },
+    {
+      "epoch": 3.2973820696468263,
+      "grad_norm": 0.4783220407113566,
+      "learning_rate": 4.695264856874589e-06,
+      "loss": 0.4284,
+      "step": 8431
+    },
+    {
+      "epoch": 3.2977772289454186,
+      "grad_norm": 0.4996331261411462,
+      "learning_rate": 4.695189890163095e-06,
+      "loss": 0.4642,
+      "step": 8432
+    },
+    {
+      "epoch": 3.298172388244011,
+      "grad_norm": 0.4859730761571242,
+      "learning_rate": 4.6951149148302285e-06,
+      "loss": 0.4564,
+      "step": 8433
+    },
+    {
+      "epoch": 3.298567547542603,
+      "grad_norm": 0.49508609585829344,
+      "learning_rate": 4.695039930876285e-06,
+      "loss": 0.4478,
+      "step": 8434
+    },
+    {
+      "epoch": 3.2989627068411953,
+      "grad_norm": 0.5073958753301407,
+      "learning_rate": 4.694964938301556e-06,
+      "loss": 0.4498,
+      "step": 8435
+    },
+    {
+      "epoch": 3.2993578661397875,
+      "grad_norm": 0.5042008611601112,
+      "learning_rate": 4.694889937106338e-06,
+      "loss": 0.4743,
+      "step": 8436
+    },
+    {
+      "epoch": 3.29975302543838,
+      "grad_norm": 0.4995949580337922,
+      "learning_rate": 4.694814927290926e-06,
+      "loss": 0.4444,
+      "step": 8437
+    },
+    {
+      "epoch": 3.300148184736972,
+      "grad_norm": 0.5010429776804977,
+      "learning_rate": 4.694739908855613e-06,
+      "loss": 0.4519,
+      "step": 8438
+    },
+    {
+      "epoch": 3.3005433440355643,
+      "grad_norm": 0.49717228341091885,
+      "learning_rate": 4.6946648818006944e-06,
+      "loss": 0.4661,
+      "step": 8439
+    },
+    {
+      "epoch": 3.3009385033341565,
+      "grad_norm": 0.4941824510019,
+      "learning_rate": 4.694589846126465e-06,
+      "loss": 0.4616,
+      "step": 8440
+    },
+    {
+      "epoch": 3.3013336626327487,
+      "grad_norm": 0.4962369271830349,
+      "learning_rate": 4.69451480183322e-06,
+      "loss": 0.4617,
+      "step": 8441
+    },
+    {
+      "epoch": 3.301728821931341,
+      "grad_norm": 0.4973622352887221,
+      "learning_rate": 4.694439748921253e-06,
+      "loss": 0.458,
+      "step": 8442
+    },
+    {
+      "epoch": 3.3021239812299332,
+      "grad_norm": 0.5023833682592578,
+      "learning_rate": 4.694364687390858e-06,
+      "loss": 0.4733,
+      "step": 8443
+    },
+    {
+      "epoch": 3.3025191405285255,
+      "grad_norm": 0.48155367010251726,
+      "learning_rate": 4.694289617242331e-06,
+      "loss": 0.4431,
+      "step": 8444
+    },
+    {
+      "epoch": 3.3029142998271177,
+      "grad_norm": 0.494415452731697,
+      "learning_rate": 4.694214538475969e-06,
+      "loss": 0.4283,
+      "step": 8445
+    },
+    {
+      "epoch": 3.30330945912571,
+      "grad_norm": 0.47365239654389407,
+      "learning_rate": 4.694139451092062e-06,
+      "loss": 0.4505,
+      "step": 8446
+    },
+    {
+      "epoch": 3.303704618424302,
+      "grad_norm": 0.485640275190784,
+      "learning_rate": 4.6940643550909096e-06,
+      "loss": 0.4496,
+      "step": 8447
+    },
+    {
+      "epoch": 3.3040997777228944,
+      "grad_norm": 0.5680281189182768,
+      "learning_rate": 4.693989250472804e-06,
+      "loss": 0.472,
+      "step": 8448
+    },
+    {
+      "epoch": 3.3044949370214867,
+      "grad_norm": 0.5020600876380309,
+      "learning_rate": 4.693914137238042e-06,
+      "loss": 0.4475,
+      "step": 8449
+    },
+    {
+      "epoch": 3.304890096320079,
+      "grad_norm": 0.4919611952382774,
+      "learning_rate": 4.693839015386916e-06,
+      "loss": 0.4579,
+      "step": 8450
+    },
+    {
+      "epoch": 3.305285255618671,
+      "grad_norm": 0.48835887965110875,
+      "learning_rate": 4.6937638849197225e-06,
+      "loss": 0.455,
+      "step": 8451
+    },
+    {
+      "epoch": 3.3056804149172634,
+      "grad_norm": 0.5031462098533465,
+      "learning_rate": 4.6936887458367565e-06,
+      "loss": 0.4576,
+      "step": 8452
+    },
+    {
+      "epoch": 3.3060755742158556,
+      "grad_norm": 0.6061653188739415,
+      "learning_rate": 4.693613598138314e-06,
+      "loss": 0.4332,
+      "step": 8453
+    },
+    {
+      "epoch": 3.306470733514448,
+      "grad_norm": 0.4888084682391102,
+      "learning_rate": 4.693538441824689e-06,
+      "loss": 0.4495,
+      "step": 8454
+    },
+    {
+      "epoch": 3.30686589281304,
+      "grad_norm": 0.4905941903714758,
+      "learning_rate": 4.693463276896177e-06,
+      "loss": 0.4529,
+      "step": 8455
+    },
+    {
+      "epoch": 3.3072610521116323,
+      "grad_norm": 0.49490350717422127,
+      "learning_rate": 4.693388103353073e-06,
+      "loss": 0.4439,
+      "step": 8456
+    },
+    {
+      "epoch": 3.3076562114102246,
+      "grad_norm": 0.5077115025169423,
+      "learning_rate": 4.693312921195673e-06,
+      "loss": 0.4602,
+      "step": 8457
+    },
+    {
+      "epoch": 3.308051370708817,
+      "grad_norm": 0.49228501239192973,
+      "learning_rate": 4.693237730424272e-06,
+      "loss": 0.4479,
+      "step": 8458
+    },
+    {
+      "epoch": 3.308446530007409,
+      "grad_norm": 0.4929223392479759,
+      "learning_rate": 4.693162531039163e-06,
+      "loss": 0.4597,
+      "step": 8459
+    },
+    {
+      "epoch": 3.3088416893060013,
+      "grad_norm": 0.48557109126722053,
+      "learning_rate": 4.693087323040646e-06,
+      "loss": 0.4667,
+      "step": 8460
+    },
+    {
+      "epoch": 3.3092368486045936,
+      "grad_norm": 0.4829968156083153,
+      "learning_rate": 4.693012106429012e-06,
+      "loss": 0.446,
+      "step": 8461
+    },
+    {
+      "epoch": 3.309632007903186,
+      "grad_norm": 0.48460924487669355,
+      "learning_rate": 4.692936881204558e-06,
+      "loss": 0.4668,
+      "step": 8462
+    },
+    {
+      "epoch": 3.310027167201778,
+      "grad_norm": 0.4841591054021867,
+      "learning_rate": 4.69286164736758e-06,
+      "loss": 0.4478,
+      "step": 8463
+    },
+    {
+      "epoch": 3.3104223265003703,
+      "grad_norm": 0.48269716253485223,
+      "learning_rate": 4.692786404918374e-06,
+      "loss": 0.437,
+      "step": 8464
+    },
+    {
+      "epoch": 3.310817485798963,
+      "grad_norm": 0.47804342056190563,
+      "learning_rate": 4.692711153857233e-06,
+      "loss": 0.4483,
+      "step": 8465
+    },
+    {
+      "epoch": 3.311212645097555,
+      "grad_norm": 0.52764555981506,
+      "learning_rate": 4.6926358941844556e-06,
+      "loss": 0.4563,
+      "step": 8466
+    },
+    {
+      "epoch": 3.3116078043961474,
+      "grad_norm": 0.5246940150795155,
+      "learning_rate": 4.692560625900335e-06,
+      "loss": 0.4497,
+      "step": 8467
+    },
+    {
+      "epoch": 3.3120029636947397,
+      "grad_norm": 0.489990665703525,
+      "learning_rate": 4.6924853490051666e-06,
+      "loss": 0.4486,
+      "step": 8468
+    },
+    {
+      "epoch": 3.312398122993332,
+      "grad_norm": 0.4932042814676327,
+      "learning_rate": 4.692410063499249e-06,
+      "loss": 0.4372,
+      "step": 8469
+    },
+    {
+      "epoch": 3.312793282291924,
+      "grad_norm": 0.5058430347314362,
+      "learning_rate": 4.692334769382874e-06,
+      "loss": 0.4471,
+      "step": 8470
+    },
+    {
+      "epoch": 3.3131884415905164,
+      "grad_norm": 0.4913325565934741,
+      "learning_rate": 4.6922594666563405e-06,
+      "loss": 0.4635,
+      "step": 8471
+    },
+    {
+      "epoch": 3.3135836008891086,
+      "grad_norm": 0.4926565499584854,
+      "learning_rate": 4.692184155319943e-06,
+      "loss": 0.4682,
+      "step": 8472
+    },
+    {
+      "epoch": 3.313978760187701,
+      "grad_norm": 0.5075289488806716,
+      "learning_rate": 4.692108835373977e-06,
+      "loss": 0.4619,
+      "step": 8473
+    },
+    {
+      "epoch": 3.314373919486293,
+      "grad_norm": 0.4971824312123781,
+      "learning_rate": 4.692033506818739e-06,
+      "loss": 0.4394,
+      "step": 8474
+    },
+    {
+      "epoch": 3.3147690787848854,
+      "grad_norm": 0.4789459129843913,
+      "learning_rate": 4.691958169654524e-06,
+      "loss": 0.4312,
+      "step": 8475
+    },
+    {
+      "epoch": 3.3151642380834776,
+      "grad_norm": 0.4948838888029013,
+      "learning_rate": 4.691882823881629e-06,
+      "loss": 0.439,
+      "step": 8476
+    },
+    {
+      "epoch": 3.31555939738207,
+      "grad_norm": 0.4995649269101974,
+      "learning_rate": 4.691807469500349e-06,
+      "loss": 0.459,
+      "step": 8477
+    },
+    {
+      "epoch": 3.315954556680662,
+      "grad_norm": 0.48840216224269173,
+      "learning_rate": 4.69173210651098e-06,
+      "loss": 0.4614,
+      "step": 8478
+    },
+    {
+      "epoch": 3.3163497159792543,
+      "grad_norm": 0.49025639250296205,
+      "learning_rate": 4.691656734913818e-06,
+      "loss": 0.4378,
+      "step": 8479
+    },
+    {
+      "epoch": 3.3167448752778466,
+      "grad_norm": 0.4776483977009846,
+      "learning_rate": 4.69158135470916e-06,
+      "loss": 0.429,
+      "step": 8480
+    },
+    {
+      "epoch": 3.317140034576439,
+      "grad_norm": 0.4836933504478722,
+      "learning_rate": 4.6915059658973e-06,
+      "loss": 0.4487,
+      "step": 8481
+    },
+    {
+      "epoch": 3.317535193875031,
+      "grad_norm": 0.47926982677697055,
+      "learning_rate": 4.691430568478536e-06,
+      "loss": 0.4369,
+      "step": 8482
+    },
+    {
+      "epoch": 3.3179303531736233,
+      "grad_norm": 0.4824835398113746,
+      "learning_rate": 4.691355162453164e-06,
+      "loss": 0.4546,
+      "step": 8483
+    },
+    {
+      "epoch": 3.3183255124722155,
+      "grad_norm": 0.4920472483420201,
+      "learning_rate": 4.691279747821479e-06,
+      "loss": 0.443,
+      "step": 8484
+    },
+    {
+      "epoch": 3.3187206717708078,
+      "grad_norm": 0.5110190442880285,
+      "learning_rate": 4.691204324583777e-06,
+      "loss": 0.4742,
+      "step": 8485
+    },
+    {
+      "epoch": 3.3191158310694,
+      "grad_norm": 0.4917474251438137,
+      "learning_rate": 4.691128892740356e-06,
+      "loss": 0.4451,
+      "step": 8486
+    },
+    {
+      "epoch": 3.3195109903679922,
+      "grad_norm": 0.5022323736398868,
+      "learning_rate": 4.691053452291511e-06,
+      "loss": 0.4384,
+      "step": 8487
+    },
+    {
+      "epoch": 3.3199061496665845,
+      "grad_norm": 0.5093861229795713,
+      "learning_rate": 4.690978003237538e-06,
+      "loss": 0.4592,
+      "step": 8488
+    },
+    {
+      "epoch": 3.3203013089651767,
+      "grad_norm": 0.49345040595298023,
+      "learning_rate": 4.690902545578735e-06,
+      "loss": 0.4566,
+      "step": 8489
+    },
+    {
+      "epoch": 3.320696468263769,
+      "grad_norm": 0.4910573607099517,
+      "learning_rate": 4.690827079315397e-06,
+      "loss": 0.4521,
+      "step": 8490
+    },
+    {
+      "epoch": 3.321091627562361,
+      "grad_norm": 0.4847495876246576,
+      "learning_rate": 4.690751604447819e-06,
+      "loss": 0.4428,
+      "step": 8491
+    },
+    {
+      "epoch": 3.3214867868609534,
+      "grad_norm": 0.48758326536965474,
+      "learning_rate": 4.6906761209763e-06,
+      "loss": 0.4521,
+      "step": 8492
+    },
+    {
+      "epoch": 3.3218819461595457,
+      "grad_norm": 0.4916431762270958,
+      "learning_rate": 4.690600628901136e-06,
+      "loss": 0.4488,
+      "step": 8493
+    },
+    {
+      "epoch": 3.322277105458138,
+      "grad_norm": 0.4858608753350482,
+      "learning_rate": 4.690525128222622e-06,
+      "loss": 0.4405,
+      "step": 8494
+    },
+    {
+      "epoch": 3.32267226475673,
+      "grad_norm": 0.4863633181170981,
+      "learning_rate": 4.690449618941056e-06,
+      "loss": 0.4491,
+      "step": 8495
+    },
+    {
+      "epoch": 3.3230674240553224,
+      "grad_norm": 0.49316937629362895,
+      "learning_rate": 4.690374101056734e-06,
+      "loss": 0.4523,
+      "step": 8496
+    },
+    {
+      "epoch": 3.3234625833539146,
+      "grad_norm": 0.5007036354318656,
+      "learning_rate": 4.690298574569952e-06,
+      "loss": 0.4493,
+      "step": 8497
+    },
+    {
+      "epoch": 3.323857742652507,
+      "grad_norm": 0.4945456330336425,
+      "learning_rate": 4.690223039481008e-06,
+      "loss": 0.4582,
+      "step": 8498
+    },
+    {
+      "epoch": 3.324252901951099,
+      "grad_norm": 0.48559472918192703,
+      "learning_rate": 4.690147495790197e-06,
+      "loss": 0.454,
+      "step": 8499
+    },
+    {
+      "epoch": 3.3246480612496914,
+      "grad_norm": 0.5121803446284094,
+      "learning_rate": 4.690071943497818e-06,
+      "loss": 0.4726,
+      "step": 8500
+    },
+    {
+      "epoch": 3.3250432205482836,
+      "grad_norm": 0.4910672144158454,
+      "learning_rate": 4.689996382604165e-06,
+      "loss": 0.4425,
+      "step": 8501
+    },
+    {
+      "epoch": 3.325438379846876,
+      "grad_norm": 0.5094226150602034,
+      "learning_rate": 4.689920813109537e-06,
+      "loss": 0.4625,
+      "step": 8502
+    },
+    {
+      "epoch": 3.325833539145468,
+      "grad_norm": 0.48636282211631227,
+      "learning_rate": 4.6898452350142296e-06,
+      "loss": 0.4548,
+      "step": 8503
+    },
+    {
+      "epoch": 3.3262286984440603,
+      "grad_norm": 0.4995052919549064,
+      "learning_rate": 4.68976964831854e-06,
+      "loss": 0.4523,
+      "step": 8504
+    },
+    {
+      "epoch": 3.3266238577426526,
+      "grad_norm": 0.501508067836005,
+      "learning_rate": 4.6896940530227645e-06,
+      "loss": 0.4596,
+      "step": 8505
+    },
+    {
+      "epoch": 3.327019017041245,
+      "grad_norm": 0.4929069745339944,
+      "learning_rate": 4.6896184491272e-06,
+      "loss": 0.4766,
+      "step": 8506
+    },
+    {
+      "epoch": 3.327414176339837,
+      "grad_norm": 0.48439758395238547,
+      "learning_rate": 4.689542836632144e-06,
+      "loss": 0.4558,
+      "step": 8507
+    },
+    {
+      "epoch": 3.3278093356384293,
+      "grad_norm": 0.5043297803420848,
+      "learning_rate": 4.689467215537894e-06,
+      "loss": 0.4545,
+      "step": 8508
+    },
+    {
+      "epoch": 3.3282044949370215,
+      "grad_norm": 0.4918720068999591,
+      "learning_rate": 4.689391585844745e-06,
+      "loss": 0.4474,
+      "step": 8509
+    },
+    {
+      "epoch": 3.3285996542356138,
+      "grad_norm": 0.49767114366249415,
+      "learning_rate": 4.689315947552997e-06,
+      "loss": 0.44,
+      "step": 8510
+    },
+    {
+      "epoch": 3.328994813534206,
+      "grad_norm": 0.47267654346989246,
+      "learning_rate": 4.689240300662944e-06,
+      "loss": 0.444,
+      "step": 8511
+    },
+    {
+      "epoch": 3.3293899728327983,
+      "grad_norm": 0.502912384208219,
+      "learning_rate": 4.6891646451748855e-06,
+      "loss": 0.4432,
+      "step": 8512
+    },
+    {
+      "epoch": 3.3297851321313905,
+      "grad_norm": 0.49847552956035984,
+      "learning_rate": 4.6890889810891175e-06,
+      "loss": 0.4619,
+      "step": 8513
+    },
+    {
+      "epoch": 3.3301802914299827,
+      "grad_norm": 0.4986927619398265,
+      "learning_rate": 4.689013308405936e-06,
+      "loss": 0.4815,
+      "step": 8514
+    },
+    {
+      "epoch": 3.330575450728575,
+      "grad_norm": 0.481380549992846,
+      "learning_rate": 4.688937627125641e-06,
+      "loss": 0.4617,
+      "step": 8515
+    },
+    {
+      "epoch": 3.330970610027167,
+      "grad_norm": 0.4999972238019487,
+      "learning_rate": 4.6888619372485275e-06,
+      "loss": 0.4521,
+      "step": 8516
+    },
+    {
+      "epoch": 3.3313657693257595,
+      "grad_norm": 0.49828427414950494,
+      "learning_rate": 4.688786238774893e-06,
+      "loss": 0.4558,
+      "step": 8517
+    },
+    {
+      "epoch": 3.3317609286243517,
+      "grad_norm": 0.49747513646012703,
+      "learning_rate": 4.688710531705036e-06,
+      "loss": 0.4397,
+      "step": 8518
+    },
+    {
+      "epoch": 3.332156087922944,
+      "grad_norm": 0.4845769352628467,
+      "learning_rate": 4.688634816039253e-06,
+      "loss": 0.4477,
+      "step": 8519
+    },
+    {
+      "epoch": 3.332551247221536,
+      "grad_norm": 0.4955335961994695,
+      "learning_rate": 4.688559091777841e-06,
+      "loss": 0.4429,
+      "step": 8520
+    },
+    {
+      "epoch": 3.3329464065201284,
+      "grad_norm": 0.5002101662039317,
+      "learning_rate": 4.6884833589210984e-06,
+      "loss": 0.4557,
+      "step": 8521
+    },
+    {
+      "epoch": 3.3333415658187207,
+      "grad_norm": 0.5012403566636711,
+      "learning_rate": 4.688407617469321e-06,
+      "loss": 0.4539,
+      "step": 8522
+    },
+    {
+      "epoch": 3.333736725117313,
+      "grad_norm": 0.4846544545345259,
+      "learning_rate": 4.688331867422809e-06,
+      "loss": 0.4452,
+      "step": 8523
+    },
+    {
+      "epoch": 3.334131884415905,
+      "grad_norm": 0.4966381233591069,
+      "learning_rate": 4.688256108781858e-06,
+      "loss": 0.4487,
+      "step": 8524
+    },
+    {
+      "epoch": 3.3345270437144974,
+      "grad_norm": 0.5829217411242708,
+      "learning_rate": 4.688180341546765e-06,
+      "loss": 0.4482,
+      "step": 8525
+    },
+    {
+      "epoch": 3.3349222030130896,
+      "grad_norm": 0.511353842249708,
+      "learning_rate": 4.688104565717828e-06,
+      "loss": 0.4382,
+      "step": 8526
+    },
+    {
+      "epoch": 3.335317362311682,
+      "grad_norm": 0.5083341149647729,
+      "learning_rate": 4.688028781295346e-06,
+      "loss": 0.4576,
+      "step": 8527
+    },
+    {
+      "epoch": 3.335712521610274,
+      "grad_norm": 0.537344546705706,
+      "learning_rate": 4.687952988279615e-06,
+      "loss": 0.4469,
+      "step": 8528
+    },
+    {
+      "epoch": 3.3361076809088663,
+      "grad_norm": 0.5068930887011414,
+      "learning_rate": 4.687877186670934e-06,
+      "loss": 0.458,
+      "step": 8529
+    },
+    {
+      "epoch": 3.3365028402074586,
+      "grad_norm": 0.5136759214514139,
+      "learning_rate": 4.687801376469599e-06,
+      "loss": 0.4687,
+      "step": 8530
+    },
+    {
+      "epoch": 3.336897999506051,
+      "grad_norm": 0.48837862440654095,
+      "learning_rate": 4.687725557675909e-06,
+      "loss": 0.4405,
+      "step": 8531
+    },
+    {
+      "epoch": 3.337293158804643,
+      "grad_norm": 0.48565670852398274,
+      "learning_rate": 4.687649730290162e-06,
+      "loss": 0.4434,
+      "step": 8532
+    },
+    {
+      "epoch": 3.3376883181032353,
+      "grad_norm": 0.5031092299127129,
+      "learning_rate": 4.6875738943126544e-06,
+      "loss": 0.4517,
+      "step": 8533
+    },
+    {
+      "epoch": 3.3380834774018275,
+      "grad_norm": 0.5215776103880958,
+      "learning_rate": 4.687498049743685e-06,
+      "loss": 0.4554,
+      "step": 8534
+    },
+    {
+      "epoch": 3.3384786367004198,
+      "grad_norm": 0.5102826341112808,
+      "learning_rate": 4.687422196583552e-06,
+      "loss": 0.4708,
+      "step": 8535
+    },
+    {
+      "epoch": 3.338873795999012,
+      "grad_norm": 0.4978853219738474,
+      "learning_rate": 4.6873463348325535e-06,
+      "loss": 0.4542,
+      "step": 8536
+    },
+    {
+      "epoch": 3.3392689552976043,
+      "grad_norm": 0.5025990366613098,
+      "learning_rate": 4.687270464490986e-06,
+      "loss": 0.4417,
+      "step": 8537
+    },
+    {
+      "epoch": 3.3396641145961965,
+      "grad_norm": 0.49855784405646736,
+      "learning_rate": 4.687194585559148e-06,
+      "loss": 0.4601,
+      "step": 8538
+    },
+    {
+      "epoch": 3.3400592738947887,
+      "grad_norm": 0.4932797881481111,
+      "learning_rate": 4.687118698037338e-06,
+      "loss": 0.4471,
+      "step": 8539
+    },
+    {
+      "epoch": 3.340454433193381,
+      "grad_norm": 0.4854332267118237,
+      "learning_rate": 4.687042801925854e-06,
+      "loss": 0.4452,
+      "step": 8540
+    },
+    {
+      "epoch": 3.340849592491973,
+      "grad_norm": 0.5037097432876378,
+      "learning_rate": 4.686966897224995e-06,
+      "loss": 0.4677,
+      "step": 8541
+    },
+    {
+      "epoch": 3.3412447517905655,
+      "grad_norm": 0.4856578335596684,
+      "learning_rate": 4.686890983935057e-06,
+      "loss": 0.4449,
+      "step": 8542
+    },
+    {
+      "epoch": 3.3416399110891577,
+      "grad_norm": 0.507337201279644,
+      "learning_rate": 4.686815062056338e-06,
+      "loss": 0.4636,
+      "step": 8543
+    },
+    {
+      "epoch": 3.34203507038775,
+      "grad_norm": 0.5033310923165123,
+      "learning_rate": 4.686739131589139e-06,
+      "loss": 0.4504,
+      "step": 8544
+    },
+    {
+      "epoch": 3.342430229686342,
+      "grad_norm": 0.5337152714631989,
+      "learning_rate": 4.686663192533756e-06,
+      "loss": 0.4927,
+      "step": 8545
+    },
+    {
+      "epoch": 3.3428253889849344,
+      "grad_norm": 0.47904040874221143,
+      "learning_rate": 4.686587244890488e-06,
+      "loss": 0.4503,
+      "step": 8546
+    },
+    {
+      "epoch": 3.3432205482835267,
+      "grad_norm": 0.49230347712013833,
+      "learning_rate": 4.686511288659633e-06,
+      "loss": 0.4568,
+      "step": 8547
+    },
+    {
+      "epoch": 3.343615707582119,
+      "grad_norm": 0.5166799488026463,
+      "learning_rate": 4.686435323841489e-06,
+      "loss": 0.4562,
+      "step": 8548
+    },
+    {
+      "epoch": 3.344010866880711,
+      "grad_norm": 0.4809789988181581,
+      "learning_rate": 4.686359350436355e-06,
+      "loss": 0.4509,
+      "step": 8549
+    },
+    {
+      "epoch": 3.3444060261793034,
+      "grad_norm": 0.4984285848482146,
+      "learning_rate": 4.686283368444529e-06,
+      "loss": 0.4519,
+      "step": 8550
+    },
+    {
+      "epoch": 3.3448011854778956,
+      "grad_norm": 0.4914707828734001,
+      "learning_rate": 4.68620737786631e-06,
+      "loss": 0.4478,
+      "step": 8551
+    },
+    {
+      "epoch": 3.345196344776488,
+      "grad_norm": 0.5038394475829153,
+      "learning_rate": 4.6861313787019955e-06,
+      "loss": 0.4666,
+      "step": 8552
+    },
+    {
+      "epoch": 3.34559150407508,
+      "grad_norm": 0.48139717719915776,
+      "learning_rate": 4.686055370951884e-06,
+      "loss": 0.4516,
+      "step": 8553
+    },
+    {
+      "epoch": 3.3459866633736723,
+      "grad_norm": 0.48724474236998466,
+      "learning_rate": 4.685979354616275e-06,
+      "loss": 0.4405,
+      "step": 8554
+    },
+    {
+      "epoch": 3.3463818226722646,
+      "grad_norm": 0.4840451154072024,
+      "learning_rate": 4.685903329695467e-06,
+      "loss": 0.4385,
+      "step": 8555
+    },
+    {
+      "epoch": 3.346776981970857,
+      "grad_norm": 0.4881270901311099,
+      "learning_rate": 4.685827296189757e-06,
+      "loss": 0.4506,
+      "step": 8556
+    },
+    {
+      "epoch": 3.347172141269449,
+      "grad_norm": 0.4989552343260758,
+      "learning_rate": 4.6857512540994456e-06,
+      "loss": 0.4698,
+      "step": 8557
+    },
+    {
+      "epoch": 3.3475673005680413,
+      "grad_norm": 0.4808629699114711,
+      "learning_rate": 4.68567520342483e-06,
+      "loss": 0.4413,
+      "step": 8558
+    },
+    {
+      "epoch": 3.3479624598666335,
+      "grad_norm": 0.5190496818983991,
+      "learning_rate": 4.68559914416621e-06,
+      "loss": 0.4611,
+      "step": 8559
+    },
+    {
+      "epoch": 3.348357619165226,
+      "grad_norm": 0.5037425270647897,
+      "learning_rate": 4.685523076323882e-06,
+      "loss": 0.4509,
+      "step": 8560
+    },
+    {
+      "epoch": 3.348752778463818,
+      "grad_norm": 0.510585938637351,
+      "learning_rate": 4.685446999898148e-06,
+      "loss": 0.4438,
+      "step": 8561
+    },
+    {
+      "epoch": 3.3491479377624103,
+      "grad_norm": 0.4974626944458801,
+      "learning_rate": 4.685370914889305e-06,
+      "loss": 0.4606,
+      "step": 8562
+    },
+    {
+      "epoch": 3.3495430970610025,
+      "grad_norm": 0.7314950180759313,
+      "learning_rate": 4.685294821297652e-06,
+      "loss": 0.4673,
+      "step": 8563
+    },
+    {
+      "epoch": 3.3499382563595947,
+      "grad_norm": 0.49161103652361915,
+      "learning_rate": 4.685218719123488e-06,
+      "loss": 0.4424,
+      "step": 8564
+    },
+    {
+      "epoch": 3.350333415658187,
+      "grad_norm": 0.5024992958323747,
+      "learning_rate": 4.685142608367112e-06,
+      "loss": 0.4678,
+      "step": 8565
+    },
+    {
+      "epoch": 3.3507285749567792,
+      "grad_norm": 0.4938192969589111,
+      "learning_rate": 4.685066489028823e-06,
+      "loss": 0.4492,
+      "step": 8566
+    },
+    {
+      "epoch": 3.3511237342553715,
+      "grad_norm": 0.4803879916882299,
+      "learning_rate": 4.684990361108919e-06,
+      "loss": 0.4561,
+      "step": 8567
+    },
+    {
+      "epoch": 3.3515188935539637,
+      "grad_norm": 0.4949401035088218,
+      "learning_rate": 4.6849142246077e-06,
+      "loss": 0.4442,
+      "step": 8568
+    },
+    {
+      "epoch": 3.351914052852556,
+      "grad_norm": 0.5976486404806604,
+      "learning_rate": 4.684838079525465e-06,
+      "loss": 0.4277,
+      "step": 8569
+    },
+    {
+      "epoch": 3.352309212151148,
+      "grad_norm": 0.49601388755837394,
+      "learning_rate": 4.684761925862512e-06,
+      "loss": 0.4531,
+      "step": 8570
+    },
+    {
+      "epoch": 3.3527043714497404,
+      "grad_norm": 0.5152413906273675,
+      "learning_rate": 4.684685763619141e-06,
+      "loss": 0.4492,
+      "step": 8571
+    },
+    {
+      "epoch": 3.353099530748333,
+      "grad_norm": 0.49341639261673015,
+      "learning_rate": 4.684609592795651e-06,
+      "loss": 0.4489,
+      "step": 8572
+    },
+    {
+      "epoch": 3.3534946900469254,
+      "grad_norm": 0.48592662223802807,
+      "learning_rate": 4.684533413392341e-06,
+      "loss": 0.455,
+      "step": 8573
+    },
+    {
+      "epoch": 3.3538898493455176,
+      "grad_norm": 0.49332844351682664,
+      "learning_rate": 4.684457225409511e-06,
+      "loss": 0.4657,
+      "step": 8574
+    },
+    {
+      "epoch": 3.35428500864411,
+      "grad_norm": 0.5037336410016675,
+      "learning_rate": 4.684381028847459e-06,
+      "loss": 0.4659,
+      "step": 8575
+    },
+    {
+      "epoch": 3.354680167942702,
+      "grad_norm": 0.506139866036012,
+      "learning_rate": 4.684304823706484e-06,
+      "loss": 0.4483,
+      "step": 8576
+    },
+    {
+      "epoch": 3.3550753272412943,
+      "grad_norm": 0.4942492969355824,
+      "learning_rate": 4.6842286099868864e-06,
+      "loss": 0.4454,
+      "step": 8577
+    },
+    {
+      "epoch": 3.3554704865398866,
+      "grad_norm": 0.49335672939762426,
+      "learning_rate": 4.684152387688966e-06,
+      "loss": 0.4433,
+      "step": 8578
+    },
+    {
+      "epoch": 3.355865645838479,
+      "grad_norm": 0.4969284399321435,
+      "learning_rate": 4.6840761568130204e-06,
+      "loss": 0.4584,
+      "step": 8579
+    },
+    {
+      "epoch": 3.356260805137071,
+      "grad_norm": 0.4873906665965171,
+      "learning_rate": 4.68399991735935e-06,
+      "loss": 0.4597,
+      "step": 8580
+    },
+    {
+      "epoch": 3.3566559644356633,
+      "grad_norm": 0.4886120652074471,
+      "learning_rate": 4.6839236693282544e-06,
+      "loss": 0.4442,
+      "step": 8581
+    },
+    {
+      "epoch": 3.3570511237342555,
+      "grad_norm": 0.4899352867723243,
+      "learning_rate": 4.683847412720033e-06,
+      "loss": 0.4521,
+      "step": 8582
+    },
+    {
+      "epoch": 3.3574462830328478,
+      "grad_norm": 0.4933462350667174,
+      "learning_rate": 4.683771147534985e-06,
+      "loss": 0.4438,
+      "step": 8583
+    },
+    {
+      "epoch": 3.35784144233144,
+      "grad_norm": 0.4898866455801216,
+      "learning_rate": 4.68369487377341e-06,
+      "loss": 0.4564,
+      "step": 8584
+    },
+    {
+      "epoch": 3.3582366016300322,
+      "grad_norm": 0.5332125037232016,
+      "learning_rate": 4.683618591435607e-06,
+      "loss": 0.4515,
+      "step": 8585
+    },
+    {
+      "epoch": 3.3586317609286245,
+      "grad_norm": 0.5359679858136057,
+      "learning_rate": 4.683542300521876e-06,
+      "loss": 0.4486,
+      "step": 8586
+    },
+    {
+      "epoch": 3.3590269202272167,
+      "grad_norm": 0.4767570634929505,
+      "learning_rate": 4.6834660010325175e-06,
+      "loss": 0.4387,
+      "step": 8587
+    },
+    {
+      "epoch": 3.359422079525809,
+      "grad_norm": 0.5056533475750209,
+      "learning_rate": 4.68338969296783e-06,
+      "loss": 0.4585,
+      "step": 8588
+    },
+    {
+      "epoch": 3.359817238824401,
+      "grad_norm": 0.48897355907247364,
+      "learning_rate": 4.683313376328113e-06,
+      "loss": 0.4395,
+      "step": 8589
+    },
+    {
+      "epoch": 3.3602123981229934,
+      "grad_norm": 0.5203028354845993,
+      "learning_rate": 4.683237051113669e-06,
+      "loss": 0.4614,
+      "step": 8590
+    },
+    {
+      "epoch": 3.3606075574215857,
+      "grad_norm": 0.4925183013708426,
+      "learning_rate": 4.683160717324794e-06,
+      "loss": 0.4436,
+      "step": 8591
+    },
+    {
+      "epoch": 3.361002716720178,
+      "grad_norm": 0.4909350641925199,
+      "learning_rate": 4.683084374961789e-06,
+      "loss": 0.4459,
+      "step": 8592
+    },
+    {
+      "epoch": 3.36139787601877,
+      "grad_norm": 0.4990261941077317,
+      "learning_rate": 4.683008024024955e-06,
+      "loss": 0.4662,
+      "step": 8593
+    },
+    {
+      "epoch": 3.3617930353173624,
+      "grad_norm": 0.5042484386336988,
+      "learning_rate": 4.6829316645145905e-06,
+      "loss": 0.4454,
+      "step": 8594
+    },
+    {
+      "epoch": 3.3621881946159546,
+      "grad_norm": 0.5110820201232344,
+      "learning_rate": 4.682855296430997e-06,
+      "loss": 0.461,
+      "step": 8595
+    },
+    {
+      "epoch": 3.362583353914547,
+      "grad_norm": 0.5119373520344455,
+      "learning_rate": 4.6827789197744725e-06,
+      "loss": 0.4351,
+      "step": 8596
+    },
+    {
+      "epoch": 3.362978513213139,
+      "grad_norm": 0.4864074295817185,
+      "learning_rate": 4.682702534545318e-06,
+      "loss": 0.4648,
+      "step": 8597
+    },
+    {
+      "epoch": 3.3633736725117314,
+      "grad_norm": 0.5170136763217118,
+      "learning_rate": 4.682626140743833e-06,
+      "loss": 0.4485,
+      "step": 8598
+    },
+    {
+      "epoch": 3.3637688318103236,
+      "grad_norm": 0.5124926082111486,
+      "learning_rate": 4.682549738370319e-06,
+      "loss": 0.4701,
+      "step": 8599
+    },
+    {
+      "epoch": 3.364163991108916,
+      "grad_norm": 0.4920001975170289,
+      "learning_rate": 4.682473327425074e-06,
+      "loss": 0.4648,
+      "step": 8600
+    },
+    {
+      "epoch": 3.364559150407508,
+      "grad_norm": 0.4937153316476773,
+      "learning_rate": 4.682396907908399e-06,
+      "loss": 0.442,
+      "step": 8601
+    },
+    {
+      "epoch": 3.3649543097061003,
+      "grad_norm": 0.48028050782341225,
+      "learning_rate": 4.682320479820595e-06,
+      "loss": 0.4459,
+      "step": 8602
+    },
+    {
+      "epoch": 3.3653494690046926,
+      "grad_norm": 0.5216497378987135,
+      "learning_rate": 4.68224404316196e-06,
+      "loss": 0.4581,
+      "step": 8603
+    },
+    {
+      "epoch": 3.365744628303285,
+      "grad_norm": 0.5014704905655514,
+      "learning_rate": 4.682167597932797e-06,
+      "loss": 0.4563,
+      "step": 8604
+    },
+    {
+      "epoch": 3.366139787601877,
+      "grad_norm": 0.478995486285583,
+      "learning_rate": 4.682091144133404e-06,
+      "loss": 0.4382,
+      "step": 8605
+    },
+    {
+      "epoch": 3.3665349469004693,
+      "grad_norm": 0.5182832142253619,
+      "learning_rate": 4.682014681764082e-06,
+      "loss": 0.459,
+      "step": 8606
+    },
+    {
+      "epoch": 3.3669301061990615,
+      "grad_norm": 0.4961453547341973,
+      "learning_rate": 4.6819382108251316e-06,
+      "loss": 0.4501,
+      "step": 8607
+    },
+    {
+      "epoch": 3.3673252654976538,
+      "grad_norm": 0.5111669516032571,
+      "learning_rate": 4.681861731316852e-06,
+      "loss": 0.4616,
+      "step": 8608
+    },
+    {
+      "epoch": 3.367720424796246,
+      "grad_norm": 0.49667736409059116,
+      "learning_rate": 4.681785243239546e-06,
+      "loss": 0.4476,
+      "step": 8609
+    },
+    {
+      "epoch": 3.3681155840948382,
+      "grad_norm": 0.509991731084816,
+      "learning_rate": 4.681708746593511e-06,
+      "loss": 0.4551,
+      "step": 8610
+    },
+    {
+      "epoch": 3.3685107433934305,
+      "grad_norm": 0.4861532008939771,
+      "learning_rate": 4.6816322413790495e-06,
+      "loss": 0.4551,
+      "step": 8611
+    },
+    {
+      "epoch": 3.3689059026920227,
+      "grad_norm": 0.47927743937701206,
+      "learning_rate": 4.6815557275964605e-06,
+      "loss": 0.4407,
+      "step": 8612
+    },
+    {
+      "epoch": 3.369301061990615,
+      "grad_norm": 0.48329305038335735,
+      "learning_rate": 4.681479205246047e-06,
+      "loss": 0.4317,
+      "step": 8613
+    },
+    {
+      "epoch": 3.369696221289207,
+      "grad_norm": 0.5557400243821855,
+      "learning_rate": 4.681402674328106e-06,
+      "loss": 0.4298,
+      "step": 8614
+    },
+    {
+      "epoch": 3.3700913805877994,
+      "grad_norm": 0.516015662888361,
+      "learning_rate": 4.681326134842941e-06,
+      "loss": 0.4399,
+      "step": 8615
+    },
+    {
+      "epoch": 3.3704865398863917,
+      "grad_norm": 0.51146416218406,
+      "learning_rate": 4.681249586790851e-06,
+      "loss": 0.4549,
+      "step": 8616
+    },
+    {
+      "epoch": 3.370881699184984,
+      "grad_norm": 0.5148904737293851,
+      "learning_rate": 4.681173030172138e-06,
+      "loss": 0.4503,
+      "step": 8617
+    },
+    {
+      "epoch": 3.371276858483576,
+      "grad_norm": 0.47920716436293814,
+      "learning_rate": 4.681096464987101e-06,
+      "loss": 0.4567,
+      "step": 8618
+    },
+    {
+      "epoch": 3.3716720177821684,
+      "grad_norm": 0.48667328112391944,
+      "learning_rate": 4.681019891236042e-06,
+      "loss": 0.4422,
+      "step": 8619
+    },
+    {
+      "epoch": 3.3720671770807606,
+      "grad_norm": 0.48135580887321766,
+      "learning_rate": 4.680943308919261e-06,
+      "loss": 0.4675,
+      "step": 8620
+    },
+    {
+      "epoch": 3.372462336379353,
+      "grad_norm": 0.5032541309680445,
+      "learning_rate": 4.680866718037058e-06,
+      "loss": 0.4648,
+      "step": 8621
+    },
+    {
+      "epoch": 3.372857495677945,
+      "grad_norm": 0.4811650608559039,
+      "learning_rate": 4.680790118589737e-06,
+      "loss": 0.4447,
+      "step": 8622
+    },
+    {
+      "epoch": 3.3732526549765374,
+      "grad_norm": 0.5466985608523096,
+      "learning_rate": 4.680713510577594e-06,
+      "loss": 0.457,
+      "step": 8623
+    },
+    {
+      "epoch": 3.3736478142751296,
+      "grad_norm": 0.5592565607971153,
+      "learning_rate": 4.680636894000935e-06,
+      "loss": 0.4449,
+      "step": 8624
+    },
+    {
+      "epoch": 3.374042973573722,
+      "grad_norm": 0.504991945313021,
+      "learning_rate": 4.680560268860057e-06,
+      "loss": 0.4542,
+      "step": 8625
+    },
+    {
+      "epoch": 3.374438132872314,
+      "grad_norm": 0.48890150098264806,
+      "learning_rate": 4.680483635155263e-06,
+      "loss": 0.4502,
+      "step": 8626
+    },
+    {
+      "epoch": 3.3748332921709063,
+      "grad_norm": 0.4943593439450679,
+      "learning_rate": 4.680406992886854e-06,
+      "loss": 0.4404,
+      "step": 8627
+    },
+    {
+      "epoch": 3.3752284514694986,
+      "grad_norm": 0.5332119219072962,
+      "learning_rate": 4.680330342055129e-06,
+      "loss": 0.4288,
+      "step": 8628
+    },
+    {
+      "epoch": 3.375623610768091,
+      "grad_norm": 0.4841185179406501,
+      "learning_rate": 4.68025368266039e-06,
+      "loss": 0.4506,
+      "step": 8629
+    },
+    {
+      "epoch": 3.376018770066683,
+      "grad_norm": 0.523621408569861,
+      "learning_rate": 4.68017701470294e-06,
+      "loss": 0.4586,
+      "step": 8630
+    },
+    {
+      "epoch": 3.3764139293652753,
+      "grad_norm": 0.46451406844891363,
+      "learning_rate": 4.680100338183078e-06,
+      "loss": 0.4332,
+      "step": 8631
+    },
+    {
+      "epoch": 3.3768090886638675,
+      "grad_norm": 0.4905181028363597,
+      "learning_rate": 4.6800236531011055e-06,
+      "loss": 0.4411,
+      "step": 8632
+    },
+    {
+      "epoch": 3.3772042479624598,
+      "grad_norm": 0.502652462267723,
+      "learning_rate": 4.679946959457325e-06,
+      "loss": 0.4693,
+      "step": 8633
+    },
+    {
+      "epoch": 3.377599407261052,
+      "grad_norm": 0.4911574856877145,
+      "learning_rate": 4.6798702572520345e-06,
+      "loss": 0.44,
+      "step": 8634
+    },
+    {
+      "epoch": 3.3779945665596443,
+      "grad_norm": 0.5052945482135779,
+      "learning_rate": 4.6797935464855385e-06,
+      "loss": 0.4649,
+      "step": 8635
+    },
+    {
+      "epoch": 3.3783897258582365,
+      "grad_norm": 0.48569037287035166,
+      "learning_rate": 4.679716827158137e-06,
+      "loss": 0.4587,
+      "step": 8636
+    },
+    {
+      "epoch": 3.3787848851568287,
+      "grad_norm": 0.5041362195084776,
+      "learning_rate": 4.6796400992701314e-06,
+      "loss": 0.4767,
+      "step": 8637
+    },
+    {
+      "epoch": 3.379180044455421,
+      "grad_norm": 0.4813960654671574,
+      "learning_rate": 4.679563362821823e-06,
+      "loss": 0.4568,
+      "step": 8638
+    },
+    {
+      "epoch": 3.379575203754013,
+      "grad_norm": 0.5088131032990694,
+      "learning_rate": 4.679486617813513e-06,
+      "loss": 0.4435,
+      "step": 8639
+    },
+    {
+      "epoch": 3.3799703630526055,
+      "grad_norm": 0.4861775857012652,
+      "learning_rate": 4.679409864245503e-06,
+      "loss": 0.4632,
+      "step": 8640
+    },
+    {
+      "epoch": 3.3803655223511977,
+      "grad_norm": 0.490586606299979,
+      "learning_rate": 4.679333102118095e-06,
+      "loss": 0.4594,
+      "step": 8641
+    },
+    {
+      "epoch": 3.38076068164979,
+      "grad_norm": 0.4932899496063389,
+      "learning_rate": 4.67925633143159e-06,
+      "loss": 0.4725,
+      "step": 8642
+    },
+    {
+      "epoch": 3.381155840948382,
+      "grad_norm": 0.5069803709886499,
+      "learning_rate": 4.679179552186289e-06,
+      "loss": 0.4608,
+      "step": 8643
+    },
+    {
+      "epoch": 3.3815510002469744,
+      "grad_norm": 0.4788990174095867,
+      "learning_rate": 4.679102764382494e-06,
+      "loss": 0.4577,
+      "step": 8644
+    },
+    {
+      "epoch": 3.3819461595455667,
+      "grad_norm": 0.49268285139092627,
+      "learning_rate": 4.6790259680205064e-06,
+      "loss": 0.463,
+      "step": 8645
+    },
+    {
+      "epoch": 3.382341318844159,
+      "grad_norm": 0.48644575261707157,
+      "learning_rate": 4.678949163100629e-06,
+      "loss": 0.4608,
+      "step": 8646
+    },
+    {
+      "epoch": 3.382736478142751,
+      "grad_norm": 0.4884652879626933,
+      "learning_rate": 4.678872349623161e-06,
+      "loss": 0.4498,
+      "step": 8647
+    },
+    {
+      "epoch": 3.3831316374413434,
+      "grad_norm": 0.4981392757751588,
+      "learning_rate": 4.678795527588407e-06,
+      "loss": 0.4554,
+      "step": 8648
+    },
+    {
+      "epoch": 3.3835267967399356,
+      "grad_norm": 0.4825160629921625,
+      "learning_rate": 4.678718696996666e-06,
+      "loss": 0.447,
+      "step": 8649
+    },
+    {
+      "epoch": 3.383921956038528,
+      "grad_norm": 0.496410984109344,
+      "learning_rate": 4.678641857848241e-06,
+      "loss": 0.4894,
+      "step": 8650
+    },
+    {
+      "epoch": 3.38431711533712,
+      "grad_norm": 0.48757622681326085,
+      "learning_rate": 4.6785650101434335e-06,
+      "loss": 0.4523,
+      "step": 8651
+    },
+    {
+      "epoch": 3.3847122746357123,
+      "grad_norm": 0.49958339915552713,
+      "learning_rate": 4.678488153882546e-06,
+      "loss": 0.4837,
+      "step": 8652
+    },
+    {
+      "epoch": 3.3851074339343046,
+      "grad_norm": 0.48280083745483665,
+      "learning_rate": 4.67841128906588e-06,
+      "loss": 0.4733,
+      "step": 8653
+    },
+    {
+      "epoch": 3.385502593232897,
+      "grad_norm": 0.4827487271721312,
+      "learning_rate": 4.6783344156937375e-06,
+      "loss": 0.4551,
+      "step": 8654
+    },
+    {
+      "epoch": 3.3858977525314895,
+      "grad_norm": 0.5135331354845594,
+      "learning_rate": 4.67825753376642e-06,
+      "loss": 0.4505,
+      "step": 8655
+    },
+    {
+      "epoch": 3.3862929118300817,
+      "grad_norm": 0.4950284195537903,
+      "learning_rate": 4.678180643284229e-06,
+      "loss": 0.4456,
+      "step": 8656
+    },
+    {
+      "epoch": 3.386688071128674,
+      "grad_norm": 0.49134952539015186,
+      "learning_rate": 4.678103744247468e-06,
+      "loss": 0.4536,
+      "step": 8657
+    },
+    {
+      "epoch": 3.3870832304272662,
+      "grad_norm": 0.5177519290341787,
+      "learning_rate": 4.678026836656437e-06,
+      "loss": 0.442,
+      "step": 8658
+    },
+    {
+      "epoch": 3.3874783897258585,
+      "grad_norm": 0.5228271177583046,
+      "learning_rate": 4.6779499205114406e-06,
+      "loss": 0.4477,
+      "step": 8659
+    },
+    {
+      "epoch": 3.3878735490244507,
+      "grad_norm": 0.48410785050252503,
+      "learning_rate": 4.677872995812778e-06,
+      "loss": 0.4512,
+      "step": 8660
+    },
+    {
+      "epoch": 3.388268708323043,
+      "grad_norm": 0.4710781351956366,
+      "learning_rate": 4.677796062560753e-06,
+      "loss": 0.4506,
+      "step": 8661
+    },
+    {
+      "epoch": 3.388663867621635,
+      "grad_norm": 0.48111306966615963,
+      "learning_rate": 4.677719120755669e-06,
+      "loss": 0.4598,
+      "step": 8662
+    },
+    {
+      "epoch": 3.3890590269202274,
+      "grad_norm": 0.4720947283760529,
+      "learning_rate": 4.6776421703978245e-06,
+      "loss": 0.4468,
+      "step": 8663
+    },
+    {
+      "epoch": 3.3894541862188197,
+      "grad_norm": 0.49393506201536114,
+      "learning_rate": 4.677565211487526e-06,
+      "loss": 0.4379,
+      "step": 8664
+    },
+    {
+      "epoch": 3.389849345517412,
+      "grad_norm": 0.48345686255173453,
+      "learning_rate": 4.677488244025072e-06,
+      "loss": 0.458,
+      "step": 8665
+    },
+    {
+      "epoch": 3.390244504816004,
+      "grad_norm": 0.490727505465697,
+      "learning_rate": 4.677411268010768e-06,
+      "loss": 0.4772,
+      "step": 8666
+    },
+    {
+      "epoch": 3.3906396641145964,
+      "grad_norm": 0.5034772743200914,
+      "learning_rate": 4.677334283444913e-06,
+      "loss": 0.4652,
+      "step": 8667
+    },
+    {
+      "epoch": 3.3910348234131886,
+      "grad_norm": 0.49435834675571205,
+      "learning_rate": 4.677257290327812e-06,
+      "loss": 0.4475,
+      "step": 8668
+    },
+    {
+      "epoch": 3.391429982711781,
+      "grad_norm": 0.5301178874307636,
+      "learning_rate": 4.677180288659766e-06,
+      "loss": 0.4683,
+      "step": 8669
+    },
+    {
+      "epoch": 3.391825142010373,
+      "grad_norm": 0.5010243323234773,
+      "learning_rate": 4.677103278441079e-06,
+      "loss": 0.4642,
+      "step": 8670
+    },
+    {
+      "epoch": 3.3922203013089653,
+      "grad_norm": 0.4885105136781966,
+      "learning_rate": 4.67702625967205e-06,
+      "loss": 0.4411,
+      "step": 8671
+    },
+    {
+      "epoch": 3.3926154606075576,
+      "grad_norm": 0.5011611192314379,
+      "learning_rate": 4.6769492323529865e-06,
+      "loss": 0.4517,
+      "step": 8672
+    },
+    {
+      "epoch": 3.39301061990615,
+      "grad_norm": 0.4939700998530884,
+      "learning_rate": 4.6768721964841865e-06,
+      "loss": 0.4498,
+      "step": 8673
+    },
+    {
+      "epoch": 3.393405779204742,
+      "grad_norm": 0.4954624935938974,
+      "learning_rate": 4.676795152065955e-06,
+      "loss": 0.4407,
+      "step": 8674
+    },
+    {
+      "epoch": 3.3938009385033343,
+      "grad_norm": 0.5119224790406571,
+      "learning_rate": 4.676718099098594e-06,
+      "loss": 0.4485,
+      "step": 8675
+    },
+    {
+      "epoch": 3.3941960978019265,
+      "grad_norm": 0.49705423434443835,
+      "learning_rate": 4.676641037582407e-06,
+      "loss": 0.4382,
+      "step": 8676
+    },
+    {
+      "epoch": 3.394591257100519,
+      "grad_norm": 0.4752730371851217,
+      "learning_rate": 4.676563967517694e-06,
+      "loss": 0.4412,
+      "step": 8677
+    },
+    {
+      "epoch": 3.394986416399111,
+      "grad_norm": 0.49012430064946205,
+      "learning_rate": 4.67648688890476e-06,
+      "loss": 0.4519,
+      "step": 8678
+    },
+    {
+      "epoch": 3.3953815756977033,
+      "grad_norm": 0.5050443098707427,
+      "learning_rate": 4.676409801743908e-06,
+      "loss": 0.4619,
+      "step": 8679
+    },
+    {
+      "epoch": 3.3957767349962955,
+      "grad_norm": 0.5049807653407213,
+      "learning_rate": 4.676332706035439e-06,
+      "loss": 0.4584,
+      "step": 8680
+    },
+    {
+      "epoch": 3.3961718942948877,
+      "grad_norm": 0.4838150340724149,
+      "learning_rate": 4.676255601779656e-06,
+      "loss": 0.4453,
+      "step": 8681
+    },
+    {
+      "epoch": 3.39656705359348,
+      "grad_norm": 0.4860542133541303,
+      "learning_rate": 4.676178488976864e-06,
+      "loss": 0.4621,
+      "step": 8682
+    },
+    {
+      "epoch": 3.3969622128920722,
+      "grad_norm": 0.4982361920417968,
+      "learning_rate": 4.676101367627364e-06,
+      "loss": 0.4705,
+      "step": 8683
+    },
+    {
+      "epoch": 3.3973573721906645,
+      "grad_norm": 0.4923037972317949,
+      "learning_rate": 4.676024237731459e-06,
+      "loss": 0.4545,
+      "step": 8684
+    },
+    {
+      "epoch": 3.3977525314892567,
+      "grad_norm": 0.5050487317688797,
+      "learning_rate": 4.675947099289452e-06,
+      "loss": 0.4528,
+      "step": 8685
+    },
+    {
+      "epoch": 3.398147690787849,
+      "grad_norm": 0.506471973724114,
+      "learning_rate": 4.675869952301646e-06,
+      "loss": 0.4753,
+      "step": 8686
+    },
+    {
+      "epoch": 3.398542850086441,
+      "grad_norm": 0.4709433859416677,
+      "learning_rate": 4.675792796768344e-06,
+      "loss": 0.445,
+      "step": 8687
+    },
+    {
+      "epoch": 3.3989380093850334,
+      "grad_norm": 0.49103621457028523,
+      "learning_rate": 4.67571563268985e-06,
+      "loss": 0.4651,
+      "step": 8688
+    },
+    {
+      "epoch": 3.3993331686836257,
+      "grad_norm": 0.490208526893036,
+      "learning_rate": 4.675638460066465e-06,
+      "loss": 0.4492,
+      "step": 8689
+    },
+    {
+      "epoch": 3.399728327982218,
+      "grad_norm": 0.48711510354854665,
+      "learning_rate": 4.675561278898494e-06,
+      "loss": 0.4417,
+      "step": 8690
+    },
+    {
+      "epoch": 3.40012348728081,
+      "grad_norm": 0.48830265701534586,
+      "learning_rate": 4.675484089186239e-06,
+      "loss": 0.449,
+      "step": 8691
+    },
+    {
+      "epoch": 3.4005186465794024,
+      "grad_norm": 0.4779197088460178,
+      "learning_rate": 4.6754068909300044e-06,
+      "loss": 0.4609,
+      "step": 8692
+    },
+    {
+      "epoch": 3.4009138058779946,
+      "grad_norm": 0.5000217624862712,
+      "learning_rate": 4.675329684130091e-06,
+      "loss": 0.4776,
+      "step": 8693
+    },
+    {
+      "epoch": 3.401308965176587,
+      "grad_norm": 0.515046165717711,
+      "learning_rate": 4.675252468786805e-06,
+      "loss": 0.4721,
+      "step": 8694
+    },
+    {
+      "epoch": 3.401704124475179,
+      "grad_norm": 0.4887957168364599,
+      "learning_rate": 4.675175244900447e-06,
+      "loss": 0.4457,
+      "step": 8695
+    },
+    {
+      "epoch": 3.4020992837737714,
+      "grad_norm": 0.48483411104988006,
+      "learning_rate": 4.675098012471322e-06,
+      "loss": 0.459,
+      "step": 8696
+    },
+    {
+      "epoch": 3.4024944430723636,
+      "grad_norm": 0.49370445185103784,
+      "learning_rate": 4.675020771499733e-06,
+      "loss": 0.4517,
+      "step": 8697
+    },
+    {
+      "epoch": 3.402889602370956,
+      "grad_norm": 0.5232425366645843,
+      "learning_rate": 4.6749435219859825e-06,
+      "loss": 0.4522,
+      "step": 8698
+    },
+    {
+      "epoch": 3.403284761669548,
+      "grad_norm": 0.47812677963839323,
+      "learning_rate": 4.674866263930375e-06,
+      "loss": 0.4373,
+      "step": 8699
+    },
+    {
+      "epoch": 3.4036799209681403,
+      "grad_norm": 0.50114037166455,
+      "learning_rate": 4.674788997333214e-06,
+      "loss": 0.4761,
+      "step": 8700
+    },
+    {
+      "epoch": 3.4040750802667326,
+      "grad_norm": 0.4990732251105035,
+      "learning_rate": 4.674711722194802e-06,
+      "loss": 0.4305,
+      "step": 8701
+    },
+    {
+      "epoch": 3.404470239565325,
+      "grad_norm": 0.4919053285526375,
+      "learning_rate": 4.674634438515443e-06,
+      "loss": 0.4596,
+      "step": 8702
+    },
+    {
+      "epoch": 3.404865398863917,
+      "grad_norm": 0.500445117775505,
+      "learning_rate": 4.6745571462954395e-06,
+      "loss": 0.4375,
+      "step": 8703
+    },
+    {
+      "epoch": 3.4052605581625093,
+      "grad_norm": 0.48686796875270305,
+      "learning_rate": 4.674479845535097e-06,
+      "loss": 0.4523,
+      "step": 8704
+    },
+    {
+      "epoch": 3.4056557174611015,
+      "grad_norm": 0.5030201027639913,
+      "learning_rate": 4.6744025362347175e-06,
+      "loss": 0.4715,
+      "step": 8705
+    },
+    {
+      "epoch": 3.4060508767596938,
+      "grad_norm": 0.4858539869831347,
+      "learning_rate": 4.674325218394606e-06,
+      "loss": 0.4542,
+      "step": 8706
+    },
+    {
+      "epoch": 3.406446036058286,
+      "grad_norm": 0.48744591028154344,
+      "learning_rate": 4.674247892015065e-06,
+      "loss": 0.4632,
+      "step": 8707
+    },
+    {
+      "epoch": 3.4068411953568782,
+      "grad_norm": 0.5305591582260777,
+      "learning_rate": 4.674170557096398e-06,
+      "loss": 0.4471,
+      "step": 8708
+    },
+    {
+      "epoch": 3.4072363546554705,
+      "grad_norm": 0.4896593869094789,
+      "learning_rate": 4.674093213638909e-06,
+      "loss": 0.4546,
+      "step": 8709
+    },
+    {
+      "epoch": 3.4076315139540627,
+      "grad_norm": 0.5193419476995951,
+      "learning_rate": 4.674015861642903e-06,
+      "loss": 0.4572,
+      "step": 8710
+    },
+    {
+      "epoch": 3.408026673252655,
+      "grad_norm": 0.5029397565244852,
+      "learning_rate": 4.673938501108684e-06,
+      "loss": 0.4718,
+      "step": 8711
+    },
+    {
+      "epoch": 3.408421832551247,
+      "grad_norm": 0.4966168365599225,
+      "learning_rate": 4.673861132036552e-06,
+      "loss": 0.4393,
+      "step": 8712
+    },
+    {
+      "epoch": 3.4088169918498394,
+      "grad_norm": 0.48279757458338685,
+      "learning_rate": 4.673783754426815e-06,
+      "loss": 0.4432,
+      "step": 8713
+    },
+    {
+      "epoch": 3.4092121511484317,
+      "grad_norm": 0.47848792729340384,
+      "learning_rate": 4.673706368279775e-06,
+      "loss": 0.4654,
+      "step": 8714
+    },
+    {
+      "epoch": 3.409607310447024,
+      "grad_norm": 0.48502060720503026,
+      "learning_rate": 4.6736289735957365e-06,
+      "loss": 0.4351,
+      "step": 8715
+    },
+    {
+      "epoch": 3.410002469745616,
+      "grad_norm": 0.517902551374955,
+      "learning_rate": 4.673551570375003e-06,
+      "loss": 0.4608,
+      "step": 8716
+    },
+    {
+      "epoch": 3.4103976290442084,
+      "grad_norm": 0.4959576763379469,
+      "learning_rate": 4.673474158617879e-06,
+      "loss": 0.4575,
+      "step": 8717
+    },
+    {
+      "epoch": 3.4107927883428006,
+      "grad_norm": 0.487828895869595,
+      "learning_rate": 4.673396738324668e-06,
+      "loss": 0.4381,
+      "step": 8718
+    },
+    {
+      "epoch": 3.411187947641393,
+      "grad_norm": 0.4921131542323153,
+      "learning_rate": 4.6733193094956755e-06,
+      "loss": 0.4618,
+      "step": 8719
+    },
+    {
+      "epoch": 3.411583106939985,
+      "grad_norm": 0.489989732392488,
+      "learning_rate": 4.6732418721312036e-06,
+      "loss": 0.4712,
+      "step": 8720
+    },
+    {
+      "epoch": 3.4119782662385774,
+      "grad_norm": 0.48021030964681694,
+      "learning_rate": 4.6731644262315575e-06,
+      "loss": 0.4398,
+      "step": 8721
+    },
+    {
+      "epoch": 3.4123734255371696,
+      "grad_norm": 0.5196748841661029,
+      "learning_rate": 4.673086971797041e-06,
+      "loss": 0.4569,
+      "step": 8722
+    },
+    {
+      "epoch": 3.412768584835762,
+      "grad_norm": 0.6391029638584853,
+      "learning_rate": 4.673009508827958e-06,
+      "loss": 0.4584,
+      "step": 8723
+    },
+    {
+      "epoch": 3.413163744134354,
+      "grad_norm": 0.49223316026383734,
+      "learning_rate": 4.672932037324614e-06,
+      "loss": 0.4523,
+      "step": 8724
+    },
+    {
+      "epoch": 3.4135589034329463,
+      "grad_norm": 0.4865710928676197,
+      "learning_rate": 4.672854557287312e-06,
+      "loss": 0.4411,
+      "step": 8725
+    },
+    {
+      "epoch": 3.4139540627315386,
+      "grad_norm": 0.5065983068219876,
+      "learning_rate": 4.672777068716357e-06,
+      "loss": 0.4726,
+      "step": 8726
+    },
+    {
+      "epoch": 3.414349222030131,
+      "grad_norm": 0.49208422147691183,
+      "learning_rate": 4.672699571612053e-06,
+      "loss": 0.4772,
+      "step": 8727
+    },
+    {
+      "epoch": 3.414744381328723,
+      "grad_norm": 0.4883146901724943,
+      "learning_rate": 4.6726220659747035e-06,
+      "loss": 0.4645,
+      "step": 8728
+    },
+    {
+      "epoch": 3.4151395406273153,
+      "grad_norm": 0.4873411772334921,
+      "learning_rate": 4.672544551804615e-06,
+      "loss": 0.453,
+      "step": 8729
+    },
+    {
+      "epoch": 3.4155346999259075,
+      "grad_norm": 0.4812888051291896,
+      "learning_rate": 4.67246702910209e-06,
+      "loss": 0.4401,
+      "step": 8730
+    },
+    {
+      "epoch": 3.4159298592244998,
+      "grad_norm": 0.48596379658909483,
+      "learning_rate": 4.6723894978674344e-06,
+      "loss": 0.4595,
+      "step": 8731
+    },
+    {
+      "epoch": 3.416325018523092,
+      "grad_norm": 0.48806164094782617,
+      "learning_rate": 4.672311958100952e-06,
+      "loss": 0.4559,
+      "step": 8732
+    },
+    {
+      "epoch": 3.4167201778216842,
+      "grad_norm": 0.5278943342209049,
+      "learning_rate": 4.672234409802946e-06,
+      "loss": 0.4598,
+      "step": 8733
+    },
+    {
+      "epoch": 3.4171153371202765,
+      "grad_norm": 0.48766632378299773,
+      "learning_rate": 4.672156852973724e-06,
+      "loss": 0.447,
+      "step": 8734
+    },
+    {
+      "epoch": 3.4175104964188687,
+      "grad_norm": 0.4998558934415389,
+      "learning_rate": 4.672079287613588e-06,
+      "loss": 0.466,
+      "step": 8735
+    },
+    {
+      "epoch": 3.417905655717461,
+      "grad_norm": 0.5161927705054742,
+      "learning_rate": 4.672001713722844e-06,
+      "loss": 0.4512,
+      "step": 8736
+    },
+    {
+      "epoch": 3.418300815016053,
+      "grad_norm": 0.4993835199850496,
+      "learning_rate": 4.671924131301795e-06,
+      "loss": 0.4667,
+      "step": 8737
+    },
+    {
+      "epoch": 3.4186959743146454,
+      "grad_norm": 0.5111345200683314,
+      "learning_rate": 4.671846540350748e-06,
+      "loss": 0.4669,
+      "step": 8738
+    },
+    {
+      "epoch": 3.4190911336132377,
+      "grad_norm": 0.4968151170656032,
+      "learning_rate": 4.6717689408700065e-06,
+      "loss": 0.4415,
+      "step": 8739
+    },
+    {
+      "epoch": 3.41948629291183,
+      "grad_norm": 0.48927357809806327,
+      "learning_rate": 4.671691332859875e-06,
+      "loss": 0.4454,
+      "step": 8740
+    },
+    {
+      "epoch": 3.419881452210422,
+      "grad_norm": 0.4986943442895406,
+      "learning_rate": 4.671613716320658e-06,
+      "loss": 0.4578,
+      "step": 8741
+    },
+    {
+      "epoch": 3.4202766115090144,
+      "grad_norm": 0.507640735251368,
+      "learning_rate": 4.671536091252662e-06,
+      "loss": 0.4708,
+      "step": 8742
+    },
+    {
+      "epoch": 3.4206717708076066,
+      "grad_norm": 0.4864374605398443,
+      "learning_rate": 4.671458457656191e-06,
+      "loss": 0.4459,
+      "step": 8743
+    },
+    {
+      "epoch": 3.421066930106199,
+      "grad_norm": 0.48587924372525426,
+      "learning_rate": 4.671380815531549e-06,
+      "loss": 0.4556,
+      "step": 8744
+    },
+    {
+      "epoch": 3.421462089404791,
+      "grad_norm": 0.49368596535285864,
+      "learning_rate": 4.671303164879043e-06,
+      "loss": 0.4518,
+      "step": 8745
+    },
+    {
+      "epoch": 3.4218572487033834,
+      "grad_norm": 0.49585373553596823,
+      "learning_rate": 4.671225505698975e-06,
+      "loss": 0.4557,
+      "step": 8746
+    },
+    {
+      "epoch": 3.4222524080019756,
+      "grad_norm": 0.48924156726844276,
+      "learning_rate": 4.671147837991653e-06,
+      "loss": 0.4598,
+      "step": 8747
+    },
+    {
+      "epoch": 3.422647567300568,
+      "grad_norm": 0.5066355289955341,
+      "learning_rate": 4.67107016175738e-06,
+      "loss": 0.4619,
+      "step": 8748
+    },
+    {
+      "epoch": 3.42304272659916,
+      "grad_norm": 0.4998574490378027,
+      "learning_rate": 4.670992476996462e-06,
+      "loss": 0.4628,
+      "step": 8749
+    },
+    {
+      "epoch": 3.4234378858977523,
+      "grad_norm": 0.4926776349012955,
+      "learning_rate": 4.670914783709203e-06,
+      "loss": 0.4533,
+      "step": 8750
+    },
+    {
+      "epoch": 3.4238330451963446,
+      "grad_norm": 0.48942538427023524,
+      "learning_rate": 4.6708370818959105e-06,
+      "loss": 0.4497,
+      "step": 8751
+    },
+    {
+      "epoch": 3.424228204494937,
+      "grad_norm": 0.48790555324765683,
+      "learning_rate": 4.670759371556886e-06,
+      "loss": 0.4497,
+      "step": 8752
+    },
+    {
+      "epoch": 3.424623363793529,
+      "grad_norm": 0.5083233533489905,
+      "learning_rate": 4.670681652692439e-06,
+      "loss": 0.4488,
+      "step": 8753
+    },
+    {
+      "epoch": 3.4250185230921213,
+      "grad_norm": 0.5216951615222701,
+      "learning_rate": 4.670603925302871e-06,
+      "loss": 0.4747,
+      "step": 8754
+    },
+    {
+      "epoch": 3.4254136823907135,
+      "grad_norm": 0.49070908525677087,
+      "learning_rate": 4.6705261893884904e-06,
+      "loss": 0.4622,
+      "step": 8755
+    },
+    {
+      "epoch": 3.4258088416893058,
+      "grad_norm": 0.4914126193159798,
+      "learning_rate": 4.6704484449496e-06,
+      "loss": 0.4628,
+      "step": 8756
+    },
+    {
+      "epoch": 3.426204000987898,
+      "grad_norm": 0.49544497930427484,
+      "learning_rate": 4.670370691986507e-06,
+      "loss": 0.4443,
+      "step": 8757
+    },
+    {
+      "epoch": 3.4265991602864903,
+      "grad_norm": 0.509745151929745,
+      "learning_rate": 4.670292930499515e-06,
+      "loss": 0.4639,
+      "step": 8758
+    },
+    {
+      "epoch": 3.4269943195850825,
+      "grad_norm": 0.4992599417209327,
+      "learning_rate": 4.67021516048893e-06,
+      "loss": 0.4588,
+      "step": 8759
+    },
+    {
+      "epoch": 3.4273894788836747,
+      "grad_norm": 0.48349844003574116,
+      "learning_rate": 4.670137381955058e-06,
+      "loss": 0.4485,
+      "step": 8760
+    },
+    {
+      "epoch": 3.4277846381822674,
+      "grad_norm": 0.492075989647956,
+      "learning_rate": 4.670059594898204e-06,
+      "loss": 0.4503,
+      "step": 8761
+    },
+    {
+      "epoch": 3.4281797974808597,
+      "grad_norm": 0.48934117903372404,
+      "learning_rate": 4.669981799318673e-06,
+      "loss": 0.4554,
+      "step": 8762
+    },
+    {
+      "epoch": 3.428574956779452,
+      "grad_norm": 0.4878416543021339,
+      "learning_rate": 4.669903995216772e-06,
+      "loss": 0.4563,
+      "step": 8763
+    },
+    {
+      "epoch": 3.428970116078044,
+      "grad_norm": 0.4967785153968765,
+      "learning_rate": 4.669826182592806e-06,
+      "loss": 0.4613,
+      "step": 8764
+    },
+    {
+      "epoch": 3.4293652753766364,
+      "grad_norm": 0.4892337802144703,
+      "learning_rate": 4.669748361447081e-06,
+      "loss": 0.4604,
+      "step": 8765
+    },
+    {
+      "epoch": 3.4297604346752286,
+      "grad_norm": 0.5013718521706811,
+      "learning_rate": 4.6696705317799e-06,
+      "loss": 0.4558,
+      "step": 8766
+    },
+    {
+      "epoch": 3.430155593973821,
+      "grad_norm": 0.5039804223697372,
+      "learning_rate": 4.669592693591571e-06,
+      "loss": 0.4654,
+      "step": 8767
+    },
+    {
+      "epoch": 3.430550753272413,
+      "grad_norm": 0.5028272637505543,
+      "learning_rate": 4.669514846882401e-06,
+      "loss": 0.4487,
+      "step": 8768
+    },
+    {
+      "epoch": 3.4309459125710053,
+      "grad_norm": 0.47464547478018637,
+      "learning_rate": 4.669436991652693e-06,
+      "loss": 0.4574,
+      "step": 8769
+    },
+    {
+      "epoch": 3.4313410718695976,
+      "grad_norm": 0.4998838948907234,
+      "learning_rate": 4.6693591279027535e-06,
+      "loss": 0.461,
+      "step": 8770
+    },
+    {
+      "epoch": 3.43173623116819,
+      "grad_norm": 0.4955573106154294,
+      "learning_rate": 4.669281255632889e-06,
+      "loss": 0.4541,
+      "step": 8771
+    },
+    {
+      "epoch": 3.432131390466782,
+      "grad_norm": 0.5161307923252046,
+      "learning_rate": 4.669203374843405e-06,
+      "loss": 0.4616,
+      "step": 8772
+    },
+    {
+      "epoch": 3.4325265497653743,
+      "grad_norm": 0.4873616543396202,
+      "learning_rate": 4.669125485534608e-06,
+      "loss": 0.4531,
+      "step": 8773
+    },
+    {
+      "epoch": 3.4329217090639665,
+      "grad_norm": 0.4914250914065892,
+      "learning_rate": 4.669047587706803e-06,
+      "loss": 0.4371,
+      "step": 8774
+    },
+    {
+      "epoch": 3.433316868362559,
+      "grad_norm": 0.49013758880173275,
+      "learning_rate": 4.668969681360295e-06,
+      "loss": 0.455,
+      "step": 8775
+    },
+    {
+      "epoch": 3.433712027661151,
+      "grad_norm": 0.4935136679299746,
+      "learning_rate": 4.668891766495393e-06,
+      "loss": 0.4581,
+      "step": 8776
+    },
+    {
+      "epoch": 3.4341071869597433,
+      "grad_norm": 0.5068462598993336,
+      "learning_rate": 4.6688138431124e-06,
+      "loss": 0.4592,
+      "step": 8777
+    },
+    {
+      "epoch": 3.4345023462583355,
+      "grad_norm": 0.5140101571522524,
+      "learning_rate": 4.668735911211623e-06,
+      "loss": 0.44,
+      "step": 8778
+    },
+    {
+      "epoch": 3.4348975055569277,
+      "grad_norm": 0.48453766176716734,
+      "learning_rate": 4.668657970793369e-06,
+      "loss": 0.4403,
+      "step": 8779
+    },
+    {
+      "epoch": 3.43529266485552,
+      "grad_norm": 0.5039231232251424,
+      "learning_rate": 4.668580021857943e-06,
+      "loss": 0.4624,
+      "step": 8780
+    },
+    {
+      "epoch": 3.4356878241541122,
+      "grad_norm": 0.5007826918099934,
+      "learning_rate": 4.668502064405651e-06,
+      "loss": 0.4575,
+      "step": 8781
+    },
+    {
+      "epoch": 3.4360829834527045,
+      "grad_norm": 0.4847249820915745,
+      "learning_rate": 4.6684240984368005e-06,
+      "loss": 0.4504,
+      "step": 8782
+    },
+    {
+      "epoch": 3.4364781427512967,
+      "grad_norm": 0.5305724135544032,
+      "learning_rate": 4.668346123951696e-06,
+      "loss": 0.4301,
+      "step": 8783
+    },
+    {
+      "epoch": 3.436873302049889,
+      "grad_norm": 0.4873670692094404,
+      "learning_rate": 4.668268140950646e-06,
+      "loss": 0.4645,
+      "step": 8784
+    },
+    {
+      "epoch": 3.437268461348481,
+      "grad_norm": 0.4961011367709567,
+      "learning_rate": 4.668190149433955e-06,
+      "loss": 0.4574,
+      "step": 8785
+    },
+    {
+      "epoch": 3.4376636206470734,
+      "grad_norm": 0.519896601726298,
+      "learning_rate": 4.668112149401928e-06,
+      "loss": 0.4673,
+      "step": 8786
+    },
+    {
+      "epoch": 3.4380587799456657,
+      "grad_norm": 0.5092646275519088,
+      "learning_rate": 4.6680341408548746e-06,
+      "loss": 0.4497,
+      "step": 8787
+    },
+    {
+      "epoch": 3.438453939244258,
+      "grad_norm": 0.5057955855549762,
+      "learning_rate": 4.667956123793099e-06,
+      "loss": 0.4783,
+      "step": 8788
+    },
+    {
+      "epoch": 3.43884909854285,
+      "grad_norm": 0.4985817701072453,
+      "learning_rate": 4.667878098216909e-06,
+      "loss": 0.4517,
+      "step": 8789
+    },
+    {
+      "epoch": 3.4392442578414424,
+      "grad_norm": 0.5134738759251121,
+      "learning_rate": 4.66780006412661e-06,
+      "loss": 0.4479,
+      "step": 8790
+    },
+    {
+      "epoch": 3.4396394171400346,
+      "grad_norm": 0.5255474891390775,
+      "learning_rate": 4.667722021522508e-06,
+      "loss": 0.4562,
+      "step": 8791
+    },
+    {
+      "epoch": 3.440034576438627,
+      "grad_norm": 0.49695920779912806,
+      "learning_rate": 4.667643970404911e-06,
+      "loss": 0.4648,
+      "step": 8792
+    },
+    {
+      "epoch": 3.440429735737219,
+      "grad_norm": 0.49907335777305206,
+      "learning_rate": 4.667565910774124e-06,
+      "loss": 0.4706,
+      "step": 8793
+    },
+    {
+      "epoch": 3.4408248950358113,
+      "grad_norm": 0.4884559789494324,
+      "learning_rate": 4.667487842630455e-06,
+      "loss": 0.4514,
+      "step": 8794
+    },
+    {
+      "epoch": 3.4412200543344036,
+      "grad_norm": 0.5000777793061358,
+      "learning_rate": 4.667409765974209e-06,
+      "loss": 0.4539,
+      "step": 8795
+    },
+    {
+      "epoch": 3.441615213632996,
+      "grad_norm": 0.5069843087753504,
+      "learning_rate": 4.667331680805694e-06,
+      "loss": 0.4659,
+      "step": 8796
+    },
+    {
+      "epoch": 3.442010372931588,
+      "grad_norm": 0.5197003132246248,
+      "learning_rate": 4.6672535871252165e-06,
+      "loss": 0.4652,
+      "step": 8797
+    },
+    {
+      "epoch": 3.4424055322301803,
+      "grad_norm": 0.4871319544582793,
+      "learning_rate": 4.6671754849330834e-06,
+      "loss": 0.4448,
+      "step": 8798
+    },
+    {
+      "epoch": 3.4428006915287725,
+      "grad_norm": 0.5028109247529198,
+      "learning_rate": 4.6670973742296e-06,
+      "loss": 0.4618,
+      "step": 8799
+    },
+    {
+      "epoch": 3.443195850827365,
+      "grad_norm": 0.49091531311391173,
+      "learning_rate": 4.667019255015075e-06,
+      "loss": 0.4409,
+      "step": 8800
+    },
+    {
+      "epoch": 3.443591010125957,
+      "grad_norm": 0.5051259627632506,
+      "learning_rate": 4.666941127289813e-06,
+      "loss": 0.4567,
+      "step": 8801
+    },
+    {
+      "epoch": 3.4439861694245493,
+      "grad_norm": 0.49791345108367346,
+      "learning_rate": 4.666862991054123e-06,
+      "loss": 0.4553,
+      "step": 8802
+    },
+    {
+      "epoch": 3.4443813287231415,
+      "grad_norm": 0.4943684112194807,
+      "learning_rate": 4.666784846308311e-06,
+      "loss": 0.4501,
+      "step": 8803
+    },
+    {
+      "epoch": 3.4447764880217338,
+      "grad_norm": 0.598269397017305,
+      "learning_rate": 4.666706693052684e-06,
+      "loss": 0.4722,
+      "step": 8804
+    },
+    {
+      "epoch": 3.445171647320326,
+      "grad_norm": 0.5033122711335563,
+      "learning_rate": 4.666628531287548e-06,
+      "loss": 0.4559,
+      "step": 8805
+    },
+    {
+      "epoch": 3.4455668066189182,
+      "grad_norm": 0.47928044079152565,
+      "learning_rate": 4.666550361013211e-06,
+      "loss": 0.4556,
+      "step": 8806
+    },
+    {
+      "epoch": 3.4459619659175105,
+      "grad_norm": 0.49275067830564856,
+      "learning_rate": 4.66647218222998e-06,
+      "loss": 0.4271,
+      "step": 8807
+    },
+    {
+      "epoch": 3.4463571252161027,
+      "grad_norm": 0.49320681766088637,
+      "learning_rate": 4.666393994938162e-06,
+      "loss": 0.4565,
+      "step": 8808
+    },
+    {
+      "epoch": 3.446752284514695,
+      "grad_norm": 0.48982527817865373,
+      "learning_rate": 4.666315799138064e-06,
+      "loss": 0.459,
+      "step": 8809
+    },
+    {
+      "epoch": 3.447147443813287,
+      "grad_norm": 0.48836849709394153,
+      "learning_rate": 4.6662375948299924e-06,
+      "loss": 0.4543,
+      "step": 8810
+    },
+    {
+      "epoch": 3.4475426031118794,
+      "grad_norm": 0.49635265949726676,
+      "learning_rate": 4.666159382014255e-06,
+      "loss": 0.4618,
+      "step": 8811
+    },
+    {
+      "epoch": 3.4479377624104717,
+      "grad_norm": 0.4953454582630569,
+      "learning_rate": 4.66608116069116e-06,
+      "loss": 0.4572,
+      "step": 8812
+    },
+    {
+      "epoch": 3.448332921709064,
+      "grad_norm": 0.5006722224939579,
+      "learning_rate": 4.6660029308610125e-06,
+      "loss": 0.4508,
+      "step": 8813
+    },
+    {
+      "epoch": 3.448728081007656,
+      "grad_norm": 0.5043235685637225,
+      "learning_rate": 4.665924692524121e-06,
+      "loss": 0.462,
+      "step": 8814
+    },
+    {
+      "epoch": 3.4491232403062484,
+      "grad_norm": 0.5053267442560526,
+      "learning_rate": 4.665846445680792e-06,
+      "loss": 0.4496,
+      "step": 8815
+    },
+    {
+      "epoch": 3.4495183996048406,
+      "grad_norm": 0.4917987457397994,
+      "learning_rate": 4.665768190331334e-06,
+      "loss": 0.4662,
+      "step": 8816
+    },
+    {
+      "epoch": 3.449913558903433,
+      "grad_norm": 0.48570702363194657,
+      "learning_rate": 4.665689926476054e-06,
+      "loss": 0.4525,
+      "step": 8817
+    },
+    {
+      "epoch": 3.450308718202025,
+      "grad_norm": 0.49691878545305157,
+      "learning_rate": 4.665611654115258e-06,
+      "loss": 0.4573,
+      "step": 8818
+    },
+    {
+      "epoch": 3.4507038775006174,
+      "grad_norm": 0.4895661716841793,
+      "learning_rate": 4.665533373249255e-06,
+      "loss": 0.4411,
+      "step": 8819
+    },
+    {
+      "epoch": 3.4510990367992096,
+      "grad_norm": 0.49186441261086594,
+      "learning_rate": 4.665455083878352e-06,
+      "loss": 0.4692,
+      "step": 8820
+    },
+    {
+      "epoch": 3.451494196097802,
+      "grad_norm": 0.5099516410323954,
+      "learning_rate": 4.665376786002856e-06,
+      "loss": 0.4656,
+      "step": 8821
+    },
+    {
+      "epoch": 3.451889355396394,
+      "grad_norm": 0.5169744344300137,
+      "learning_rate": 4.665298479623075e-06,
+      "loss": 0.4556,
+      "step": 8822
+    },
+    {
+      "epoch": 3.4522845146949863,
+      "grad_norm": 0.5171843190465332,
+      "learning_rate": 4.665220164739316e-06,
+      "loss": 0.4769,
+      "step": 8823
+    },
+    {
+      "epoch": 3.4526796739935786,
+      "grad_norm": 0.4974279869698287,
+      "learning_rate": 4.665141841351888e-06,
+      "loss": 0.439,
+      "step": 8824
+    },
+    {
+      "epoch": 3.453074833292171,
+      "grad_norm": 0.50434560452558,
+      "learning_rate": 4.665063509461098e-06,
+      "loss": 0.4594,
+      "step": 8825
+    },
+    {
+      "epoch": 3.453469992590763,
+      "grad_norm": 0.4955408379140396,
+      "learning_rate": 4.664985169067251e-06,
+      "loss": 0.4758,
+      "step": 8826
+    },
+    {
+      "epoch": 3.4538651518893553,
+      "grad_norm": 0.48902071003257797,
+      "learning_rate": 4.664906820170658e-06,
+      "loss": 0.4516,
+      "step": 8827
+    },
+    {
+      "epoch": 3.4542603111879475,
+      "grad_norm": 0.4937855562200361,
+      "learning_rate": 4.664828462771627e-06,
+      "loss": 0.4646,
+      "step": 8828
+    },
+    {
+      "epoch": 3.4546554704865398,
+      "grad_norm": 0.5082998875173607,
+      "learning_rate": 4.664750096870463e-06,
+      "loss": 0.4603,
+      "step": 8829
+    },
+    {
+      "epoch": 3.455050629785132,
+      "grad_norm": 0.47237718060013906,
+      "learning_rate": 4.664671722467475e-06,
+      "loss": 0.4576,
+      "step": 8830
+    },
+    {
+      "epoch": 3.4554457890837242,
+      "grad_norm": 0.5038153640837024,
+      "learning_rate": 4.664593339562971e-06,
+      "loss": 0.4622,
+      "step": 8831
+    },
+    {
+      "epoch": 3.4558409483823165,
+      "grad_norm": 0.49668508531522104,
+      "learning_rate": 4.66451494815726e-06,
+      "loss": 0.4477,
+      "step": 8832
+    },
+    {
+      "epoch": 3.4562361076809087,
+      "grad_norm": 0.5133159759954414,
+      "learning_rate": 4.664436548250646e-06,
+      "loss": 0.4596,
+      "step": 8833
+    },
+    {
+      "epoch": 3.456631266979501,
+      "grad_norm": 0.49905932346964094,
+      "learning_rate": 4.664358139843442e-06,
+      "loss": 0.4759,
+      "step": 8834
+    },
+    {
+      "epoch": 3.457026426278093,
+      "grad_norm": 0.4849036414485639,
+      "learning_rate": 4.664279722935953e-06,
+      "loss": 0.4609,
+      "step": 8835
+    },
+    {
+      "epoch": 3.4574215855766854,
+      "grad_norm": 0.5058385856975989,
+      "learning_rate": 4.664201297528487e-06,
+      "loss": 0.4706,
+      "step": 8836
+    },
+    {
+      "epoch": 3.4578167448752777,
+      "grad_norm": 0.4967750894306117,
+      "learning_rate": 4.6641228636213524e-06,
+      "loss": 0.4501,
+      "step": 8837
+    },
+    {
+      "epoch": 3.45821190417387,
+      "grad_norm": 0.5136748603140383,
+      "learning_rate": 4.664044421214857e-06,
+      "loss": 0.4634,
+      "step": 8838
+    },
+    {
+      "epoch": 3.458607063472462,
+      "grad_norm": 0.4914768586557601,
+      "learning_rate": 4.663965970309311e-06,
+      "loss": 0.4609,
+      "step": 8839
+    },
+    {
+      "epoch": 3.4590022227710544,
+      "grad_norm": 0.5032307100674864,
+      "learning_rate": 4.6638875109050184e-06,
+      "loss": 0.454,
+      "step": 8840
+    },
+    {
+      "epoch": 3.4593973820696466,
+      "grad_norm": 0.5373774375004284,
+      "learning_rate": 4.663809043002291e-06,
+      "loss": 0.4558,
+      "step": 8841
+    },
+    {
+      "epoch": 3.459792541368239,
+      "grad_norm": 0.5050016062307993,
+      "learning_rate": 4.663730566601436e-06,
+      "loss": 0.4626,
+      "step": 8842
+    },
+    {
+      "epoch": 3.460187700666831,
+      "grad_norm": 0.4891364435706425,
+      "learning_rate": 4.6636520817027596e-06,
+      "loss": 0.4506,
+      "step": 8843
+    },
+    {
+      "epoch": 3.460582859965424,
+      "grad_norm": 0.48998437910287085,
+      "learning_rate": 4.6635735883065734e-06,
+      "loss": 0.4529,
+      "step": 8844
+    },
+    {
+      "epoch": 3.460978019264016,
+      "grad_norm": 0.5234452315475704,
+      "learning_rate": 4.663495086413183e-06,
+      "loss": 0.4699,
+      "step": 8845
+    },
+    {
+      "epoch": 3.4613731785626083,
+      "grad_norm": 0.49371396515639815,
+      "learning_rate": 4.663416576022898e-06,
+      "loss": 0.4706,
+      "step": 8846
+    },
+    {
+      "epoch": 3.4617683378612005,
+      "grad_norm": 0.49709209591769404,
+      "learning_rate": 4.663338057136025e-06,
+      "loss": 0.476,
+      "step": 8847
+    },
+    {
+      "epoch": 3.4621634971597928,
+      "grad_norm": 0.49578446171291213,
+      "learning_rate": 4.663259529752876e-06,
+      "loss": 0.4444,
+      "step": 8848
+    },
+    {
+      "epoch": 3.462558656458385,
+      "grad_norm": 0.49126172062385615,
+      "learning_rate": 4.663180993873756e-06,
+      "loss": 0.4489,
+      "step": 8849
+    },
+    {
+      "epoch": 3.4629538157569772,
+      "grad_norm": 0.49180903216506305,
+      "learning_rate": 4.663102449498974e-06,
+      "loss": 0.4546,
+      "step": 8850
+    },
+    {
+      "epoch": 3.4633489750555695,
+      "grad_norm": 0.5005694477594985,
+      "learning_rate": 4.66302389662884e-06,
+      "loss": 0.4512,
+      "step": 8851
+    },
+    {
+      "epoch": 3.4637441343541617,
+      "grad_norm": 0.49774390912434563,
+      "learning_rate": 4.6629453352636615e-06,
+      "loss": 0.4651,
+      "step": 8852
+    },
+    {
+      "epoch": 3.464139293652754,
+      "grad_norm": 0.5149141552536891,
+      "learning_rate": 4.662866765403747e-06,
+      "loss": 0.4473,
+      "step": 8853
+    },
+    {
+      "epoch": 3.464534452951346,
+      "grad_norm": 0.5072099362598289,
+      "learning_rate": 4.662788187049405e-06,
+      "loss": 0.4772,
+      "step": 8854
+    },
+    {
+      "epoch": 3.4649296122499385,
+      "grad_norm": 0.49560828556589115,
+      "learning_rate": 4.6627096002009445e-06,
+      "loss": 0.4537,
+      "step": 8855
+    },
+    {
+      "epoch": 3.4653247715485307,
+      "grad_norm": 0.4920647341464689,
+      "learning_rate": 4.662631004858674e-06,
+      "loss": 0.4422,
+      "step": 8856
+    },
+    {
+      "epoch": 3.465719930847123,
+      "grad_norm": 0.4992878751198371,
+      "learning_rate": 4.6625524010229026e-06,
+      "loss": 0.4608,
+      "step": 8857
+    },
+    {
+      "epoch": 3.466115090145715,
+      "grad_norm": 0.49054926255487874,
+      "learning_rate": 4.662473788693938e-06,
+      "loss": 0.4539,
+      "step": 8858
+    },
+    {
+      "epoch": 3.4665102494443074,
+      "grad_norm": 0.511990187730294,
+      "learning_rate": 4.6623951678720894e-06,
+      "loss": 0.4829,
+      "step": 8859
+    },
+    {
+      "epoch": 3.4669054087428997,
+      "grad_norm": 0.5042603888878492,
+      "learning_rate": 4.662316538557666e-06,
+      "loss": 0.4613,
+      "step": 8860
+    },
+    {
+      "epoch": 3.467300568041492,
+      "grad_norm": 0.4969317296679051,
+      "learning_rate": 4.662237900750976e-06,
+      "loss": 0.445,
+      "step": 8861
+    },
+    {
+      "epoch": 3.467695727340084,
+      "grad_norm": 0.503339914538068,
+      "learning_rate": 4.662159254452329e-06,
+      "loss": 0.4607,
+      "step": 8862
+    },
+    {
+      "epoch": 3.4680908866386764,
+      "grad_norm": 0.47526527789470935,
+      "learning_rate": 4.662080599662032e-06,
+      "loss": 0.4332,
+      "step": 8863
+    },
+    {
+      "epoch": 3.4684860459372686,
+      "grad_norm": 0.48910025218721065,
+      "learning_rate": 4.662001936380397e-06,
+      "loss": 0.4532,
+      "step": 8864
+    },
+    {
+      "epoch": 3.468881205235861,
+      "grad_norm": 0.5023800565124126,
+      "learning_rate": 4.661923264607731e-06,
+      "loss": 0.4635,
+      "step": 8865
+    },
+    {
+      "epoch": 3.469276364534453,
+      "grad_norm": 0.49650951915124897,
+      "learning_rate": 4.661844584344342e-06,
+      "loss": 0.4726,
+      "step": 8866
+    },
+    {
+      "epoch": 3.4696715238330453,
+      "grad_norm": 0.4968947894789248,
+      "learning_rate": 4.661765895590541e-06,
+      "loss": 0.4736,
+      "step": 8867
+    },
+    {
+      "epoch": 3.4700666831316376,
+      "grad_norm": 0.5009727916810776,
+      "learning_rate": 4.661687198346636e-06,
+      "loss": 0.459,
+      "step": 8868
+    },
+    {
+      "epoch": 3.47046184243023,
+      "grad_norm": 0.48976101872113026,
+      "learning_rate": 4.661608492612937e-06,
+      "loss": 0.4467,
+      "step": 8869
+    },
+    {
+      "epoch": 3.470857001728822,
+      "grad_norm": 0.5016589813587168,
+      "learning_rate": 4.661529778389752e-06,
+      "loss": 0.4542,
+      "step": 8870
+    },
+    {
+      "epoch": 3.4712521610274143,
+      "grad_norm": 0.4861886122067568,
+      "learning_rate": 4.66145105567739e-06,
+      "loss": 0.461,
+      "step": 8871
+    },
+    {
+      "epoch": 3.4716473203260065,
+      "grad_norm": 0.641821102486397,
+      "learning_rate": 4.661372324476161e-06,
+      "loss": 0.4541,
+      "step": 8872
+    },
+    {
+      "epoch": 3.4720424796245988,
+      "grad_norm": 0.4940555527439461,
+      "learning_rate": 4.6612935847863746e-06,
+      "loss": 0.4645,
+      "step": 8873
+    },
+    {
+      "epoch": 3.472437638923191,
+      "grad_norm": 0.5017974488912654,
+      "learning_rate": 4.661214836608339e-06,
+      "loss": 0.4536,
+      "step": 8874
+    },
+    {
+      "epoch": 3.4728327982217833,
+      "grad_norm": 0.4969241194719381,
+      "learning_rate": 4.661136079942364e-06,
+      "loss": 0.4661,
+      "step": 8875
+    },
+    {
+      "epoch": 3.4732279575203755,
+      "grad_norm": 0.497649831354656,
+      "learning_rate": 4.661057314788758e-06,
+      "loss": 0.4413,
+      "step": 8876
+    },
+    {
+      "epoch": 3.4736231168189677,
+      "grad_norm": 0.485793011113323,
+      "learning_rate": 4.6609785411478326e-06,
+      "loss": 0.4653,
+      "step": 8877
+    },
+    {
+      "epoch": 3.47401827611756,
+      "grad_norm": 0.49424185636318685,
+      "learning_rate": 4.6608997590198945e-06,
+      "loss": 0.4475,
+      "step": 8878
+    },
+    {
+      "epoch": 3.474413435416152,
+      "grad_norm": 0.48959550866721235,
+      "learning_rate": 4.6608209684052555e-06,
+      "loss": 0.455,
+      "step": 8879
+    },
+    {
+      "epoch": 3.4748085947147445,
+      "grad_norm": 0.4878411516995577,
+      "learning_rate": 4.660742169304223e-06,
+      "loss": 0.4612,
+      "step": 8880
+    },
+    {
+      "epoch": 3.4752037540133367,
+      "grad_norm": 0.5094157531998392,
+      "learning_rate": 4.660663361717107e-06,
+      "loss": 0.4718,
+      "step": 8881
+    },
+    {
+      "epoch": 3.475598913311929,
+      "grad_norm": 0.4989284958044305,
+      "learning_rate": 4.660584545644218e-06,
+      "loss": 0.4602,
+      "step": 8882
+    },
+    {
+      "epoch": 3.475994072610521,
+      "grad_norm": 0.48462148426022633,
+      "learning_rate": 4.660505721085865e-06,
+      "loss": 0.4506,
+      "step": 8883
+    },
+    {
+      "epoch": 3.4763892319091134,
+      "grad_norm": 0.4881796143310651,
+      "learning_rate": 4.660426888042356e-06,
+      "loss": 0.4601,
+      "step": 8884
+    },
+    {
+      "epoch": 3.4767843912077057,
+      "grad_norm": 0.49485495431393517,
+      "learning_rate": 4.660348046514004e-06,
+      "loss": 0.4462,
+      "step": 8885
+    },
+    {
+      "epoch": 3.477179550506298,
+      "grad_norm": 0.48394677880720927,
+      "learning_rate": 4.660269196501116e-06,
+      "loss": 0.4463,
+      "step": 8886
+    },
+    {
+      "epoch": 3.47757470980489,
+      "grad_norm": 0.4855215947244085,
+      "learning_rate": 4.660190338004002e-06,
+      "loss": 0.4465,
+      "step": 8887
+    },
+    {
+      "epoch": 3.4779698691034824,
+      "grad_norm": 0.5086993405331509,
+      "learning_rate": 4.660111471022973e-06,
+      "loss": 0.456,
+      "step": 8888
+    },
+    {
+      "epoch": 3.4783650284020746,
+      "grad_norm": 0.49724371282452545,
+      "learning_rate": 4.660032595558337e-06,
+      "loss": 0.4611,
+      "step": 8889
+    },
+    {
+      "epoch": 3.478760187700667,
+      "grad_norm": 0.5029381576180928,
+      "learning_rate": 4.659953711610405e-06,
+      "loss": 0.4807,
+      "step": 8890
+    },
+    {
+      "epoch": 3.479155346999259,
+      "grad_norm": 0.48536678216800494,
+      "learning_rate": 4.659874819179486e-06,
+      "loss": 0.4673,
+      "step": 8891
+    },
+    {
+      "epoch": 3.4795505062978513,
+      "grad_norm": 0.48336927893573745,
+      "learning_rate": 4.65979591826589e-06,
+      "loss": 0.4539,
+      "step": 8892
+    },
+    {
+      "epoch": 3.4799456655964436,
+      "grad_norm": 0.4937274142048356,
+      "learning_rate": 4.659717008869928e-06,
+      "loss": 0.4763,
+      "step": 8893
+    },
+    {
+      "epoch": 3.480340824895036,
+      "grad_norm": 0.4928360042158491,
+      "learning_rate": 4.659638090991909e-06,
+      "loss": 0.4613,
+      "step": 8894
+    },
+    {
+      "epoch": 3.480735984193628,
+      "grad_norm": 0.530269437282756,
+      "learning_rate": 4.659559164632142e-06,
+      "loss": 0.4487,
+      "step": 8895
+    },
+    {
+      "epoch": 3.4811311434922203,
+      "grad_norm": 0.4910271446685397,
+      "learning_rate": 4.659480229790938e-06,
+      "loss": 0.4719,
+      "step": 8896
+    },
+    {
+      "epoch": 3.4815263027908125,
+      "grad_norm": 0.4777584233118303,
+      "learning_rate": 4.659401286468609e-06,
+      "loss": 0.4439,
+      "step": 8897
+    },
+    {
+      "epoch": 3.481921462089405,
+      "grad_norm": 0.48331532023965773,
+      "learning_rate": 4.659322334665461e-06,
+      "loss": 0.454,
+      "step": 8898
+    },
+    {
+      "epoch": 3.482316621387997,
+      "grad_norm": 0.49680547638456035,
+      "learning_rate": 4.659243374381806e-06,
+      "loss": 0.4589,
+      "step": 8899
+    },
+    {
+      "epoch": 3.4827117806865893,
+      "grad_norm": 0.4759886694048339,
+      "learning_rate": 4.659164405617955e-06,
+      "loss": 0.453,
+      "step": 8900
+    },
+    {
+      "epoch": 3.4831069399851815,
+      "grad_norm": 0.48347058134167675,
+      "learning_rate": 4.6590854283742175e-06,
+      "loss": 0.4649,
+      "step": 8901
+    },
+    {
+      "epoch": 3.4835020992837737,
+      "grad_norm": 0.5021839134977939,
+      "learning_rate": 4.659006442650903e-06,
+      "loss": 0.4534,
+      "step": 8902
+    },
+    {
+      "epoch": 3.483897258582366,
+      "grad_norm": 0.48132025196008,
+      "learning_rate": 4.658927448448323e-06,
+      "loss": 0.4478,
+      "step": 8903
+    },
+    {
+      "epoch": 3.4842924178809582,
+      "grad_norm": 0.49135070095173283,
+      "learning_rate": 4.658848445766786e-06,
+      "loss": 0.4506,
+      "step": 8904
+    },
+    {
+      "epoch": 3.4846875771795505,
+      "grad_norm": 0.48125094395751417,
+      "learning_rate": 4.6587694346066035e-06,
+      "loss": 0.4377,
+      "step": 8905
+    },
+    {
+      "epoch": 3.4850827364781427,
+      "grad_norm": 0.49750742233825657,
+      "learning_rate": 4.658690414968086e-06,
+      "loss": 0.4655,
+      "step": 8906
+    },
+    {
+      "epoch": 3.485477895776735,
+      "grad_norm": 0.48458376051136504,
+      "learning_rate": 4.658611386851543e-06,
+      "loss": 0.4559,
+      "step": 8907
+    },
+    {
+      "epoch": 3.485873055075327,
+      "grad_norm": 0.5115369273188421,
+      "learning_rate": 4.658532350257285e-06,
+      "loss": 0.4611,
+      "step": 8908
+    },
+    {
+      "epoch": 3.4862682143739194,
+      "grad_norm": 0.48861679265318614,
+      "learning_rate": 4.658453305185623e-06,
+      "loss": 0.4476,
+      "step": 8909
+    },
+    {
+      "epoch": 3.4866633736725117,
+      "grad_norm": 0.4877426200944746,
+      "learning_rate": 4.658374251636867e-06,
+      "loss": 0.4544,
+      "step": 8910
+    },
+    {
+      "epoch": 3.487058532971104,
+      "grad_norm": 0.4828165510082438,
+      "learning_rate": 4.658295189611327e-06,
+      "loss": 0.4535,
+      "step": 8911
+    },
+    {
+      "epoch": 3.487453692269696,
+      "grad_norm": 0.6245926944397081,
+      "learning_rate": 4.658216119109314e-06,
+      "loss": 0.4151,
+      "step": 8912
+    },
+    {
+      "epoch": 3.4878488515682884,
+      "grad_norm": 0.4874104433869976,
+      "learning_rate": 4.65813704013114e-06,
+      "loss": 0.4456,
+      "step": 8913
+    },
+    {
+      "epoch": 3.4882440108668806,
+      "grad_norm": 0.5145352743350671,
+      "learning_rate": 4.658057952677113e-06,
+      "loss": 0.4768,
+      "step": 8914
+    },
+    {
+      "epoch": 3.488639170165473,
+      "grad_norm": 0.49016021595770465,
+      "learning_rate": 4.657978856747546e-06,
+      "loss": 0.4567,
+      "step": 8915
+    },
+    {
+      "epoch": 3.489034329464065,
+      "grad_norm": 0.48831218665648374,
+      "learning_rate": 4.6578997523427475e-06,
+      "loss": 0.4356,
+      "step": 8916
+    },
+    {
+      "epoch": 3.4894294887626573,
+      "grad_norm": 0.5010113609737595,
+      "learning_rate": 4.657820639463029e-06,
+      "loss": 0.4799,
+      "step": 8917
+    },
+    {
+      "epoch": 3.4898246480612496,
+      "grad_norm": 0.4907345989010224,
+      "learning_rate": 4.657741518108702e-06,
+      "loss": 0.4628,
+      "step": 8918
+    },
+    {
+      "epoch": 3.490219807359842,
+      "grad_norm": 0.49499089035592075,
+      "learning_rate": 4.657662388280076e-06,
+      "loss": 0.4615,
+      "step": 8919
+    },
+    {
+      "epoch": 3.490614966658434,
+      "grad_norm": 0.5066619543747422,
+      "learning_rate": 4.657583249977463e-06,
+      "loss": 0.4592,
+      "step": 8920
+    },
+    {
+      "epoch": 3.4910101259570263,
+      "grad_norm": 0.4958302058204061,
+      "learning_rate": 4.657504103201173e-06,
+      "loss": 0.4485,
+      "step": 8921
+    },
+    {
+      "epoch": 3.4914052852556186,
+      "grad_norm": 0.4944424623211386,
+      "learning_rate": 4.657424947951517e-06,
+      "loss": 0.4522,
+      "step": 8922
+    },
+    {
+      "epoch": 3.491800444554211,
+      "grad_norm": 0.6267993198451537,
+      "learning_rate": 4.657345784228805e-06,
+      "loss": 0.462,
+      "step": 8923
+    },
+    {
+      "epoch": 3.492195603852803,
+      "grad_norm": 0.548874751841137,
+      "learning_rate": 4.657266612033348e-06,
+      "loss": 0.4649,
+      "step": 8924
+    },
+    {
+      "epoch": 3.4925907631513953,
+      "grad_norm": 0.5118844974346741,
+      "learning_rate": 4.65718743136546e-06,
+      "loss": 0.4482,
+      "step": 8925
+    },
+    {
+      "epoch": 3.4929859224499875,
+      "grad_norm": 0.5033343474006436,
+      "learning_rate": 4.657108242225449e-06,
+      "loss": 0.4731,
+      "step": 8926
+    },
+    {
+      "epoch": 3.4933810817485798,
+      "grad_norm": 0.4861803347304094,
+      "learning_rate": 4.657029044613627e-06,
+      "loss": 0.4503,
+      "step": 8927
+    },
+    {
+      "epoch": 3.493776241047172,
+      "grad_norm": 0.4791250625600111,
+      "learning_rate": 4.656949838530304e-06,
+      "loss": 0.4428,
+      "step": 8928
+    },
+    {
+      "epoch": 3.4941714003457642,
+      "grad_norm": 0.4863617818870344,
+      "learning_rate": 4.656870623975791e-06,
+      "loss": 0.4659,
+      "step": 8929
+    },
+    {
+      "epoch": 3.4945665596443565,
+      "grad_norm": 0.5127343716951415,
+      "learning_rate": 4.656791400950401e-06,
+      "loss": 0.461,
+      "step": 8930
+    },
+    {
+      "epoch": 3.4949617189429487,
+      "grad_norm": 0.49777034576549223,
+      "learning_rate": 4.656712169454444e-06,
+      "loss": 0.4585,
+      "step": 8931
+    },
+    {
+      "epoch": 3.495356878241541,
+      "grad_norm": 0.48877150097142436,
+      "learning_rate": 4.656632929488231e-06,
+      "loss": 0.4547,
+      "step": 8932
+    },
+    {
+      "epoch": 3.495752037540133,
+      "grad_norm": 0.4927474974697246,
+      "learning_rate": 4.656553681052074e-06,
+      "loss": 0.4581,
+      "step": 8933
+    },
+    {
+      "epoch": 3.4961471968387254,
+      "grad_norm": 0.4849494940227447,
+      "learning_rate": 4.656474424146283e-06,
+      "loss": 0.4459,
+      "step": 8934
+    },
+    {
+      "epoch": 3.4965423561373177,
+      "grad_norm": 0.48094079718336596,
+      "learning_rate": 4.6563951587711706e-06,
+      "loss": 0.4617,
+      "step": 8935
+    },
+    {
+      "epoch": 3.49693751543591,
+      "grad_norm": 0.5234897546830422,
+      "learning_rate": 4.6563158849270465e-06,
+      "loss": 0.4462,
+      "step": 8936
+    },
+    {
+      "epoch": 3.497332674734502,
+      "grad_norm": 0.49375252338582026,
+      "learning_rate": 4.656236602614223e-06,
+      "loss": 0.4497,
+      "step": 8937
+    },
+    {
+      "epoch": 3.4977278340330944,
+      "grad_norm": 0.5267848201795143,
+      "learning_rate": 4.656157311833013e-06,
+      "loss": 0.442,
+      "step": 8938
+    },
+    {
+      "epoch": 3.4981229933316866,
+      "grad_norm": 0.5060587932376777,
+      "learning_rate": 4.656078012583725e-06,
+      "loss": 0.4711,
+      "step": 8939
+    },
+    {
+      "epoch": 3.498518152630279,
+      "grad_norm": 0.5025260956706785,
+      "learning_rate": 4.655998704866672e-06,
+      "loss": 0.4468,
+      "step": 8940
+    },
+    {
+      "epoch": 3.498913311928871,
+      "grad_norm": 0.687091654136015,
+      "learning_rate": 4.655919388682166e-06,
+      "loss": 0.4844,
+      "step": 8941
+    },
+    {
+      "epoch": 3.4993084712274634,
+      "grad_norm": 0.48026680469874616,
+      "learning_rate": 4.655840064030517e-06,
+      "loss": 0.4412,
+      "step": 8942
+    },
+    {
+      "epoch": 3.4997036305260556,
+      "grad_norm": 0.4942720809704896,
+      "learning_rate": 4.655760730912038e-06,
+      "loss": 0.4645,
+      "step": 8943
+    },
+    {
+      "epoch": 3.500098789824648,
+      "grad_norm": 0.509126848920385,
+      "learning_rate": 4.655681389327039e-06,
+      "loss": 0.4497,
+      "step": 8944
+    },
+    {
+      "epoch": 3.50049394912324,
+      "grad_norm": 0.7110801116137855,
+      "learning_rate": 4.655602039275833e-06,
+      "loss": 0.4598,
+      "step": 8945
+    },
+    {
+      "epoch": 3.5008891084218323,
+      "grad_norm": 0.5019582976854259,
+      "learning_rate": 4.6555226807587304e-06,
+      "loss": 0.4657,
+      "step": 8946
+    },
+    {
+      "epoch": 3.5012842677204246,
+      "grad_norm": 0.47619258987485347,
+      "learning_rate": 4.655443313776045e-06,
+      "loss": 0.4483,
+      "step": 8947
+    },
+    {
+      "epoch": 3.501679427019017,
+      "grad_norm": 0.5029273839023065,
+      "learning_rate": 4.655363938328086e-06,
+      "loss": 0.4669,
+      "step": 8948
+    },
+    {
+      "epoch": 3.502074586317609,
+      "grad_norm": 0.5028619187317396,
+      "learning_rate": 4.655284554415167e-06,
+      "loss": 0.4582,
+      "step": 8949
+    },
+    {
+      "epoch": 3.5024697456162013,
+      "grad_norm": 0.49264262307149015,
+      "learning_rate": 4.655205162037598e-06,
+      "loss": 0.4606,
+      "step": 8950
+    },
+    {
+      "epoch": 3.5028649049147935,
+      "grad_norm": 0.4708730042896239,
+      "learning_rate": 4.655125761195694e-06,
+      "loss": 0.4526,
+      "step": 8951
+    },
+    {
+      "epoch": 3.5032600642133858,
+      "grad_norm": 0.4861263707563721,
+      "learning_rate": 4.655046351889763e-06,
+      "loss": 0.4414,
+      "step": 8952
+    },
+    {
+      "epoch": 3.503655223511978,
+      "grad_norm": 0.4893316048388238,
+      "learning_rate": 4.654966934120118e-06,
+      "loss": 0.453,
+      "step": 8953
+    },
+    {
+      "epoch": 3.5040503828105702,
+      "grad_norm": 0.5139962221803164,
+      "learning_rate": 4.654887507887073e-06,
+      "loss": 0.465,
+      "step": 8954
+    },
+    {
+      "epoch": 3.5044455421091625,
+      "grad_norm": 0.47724658926285185,
+      "learning_rate": 4.654808073190937e-06,
+      "loss": 0.4418,
+      "step": 8955
+    },
+    {
+      "epoch": 3.504840701407755,
+      "grad_norm": 0.48987603430328075,
+      "learning_rate": 4.654728630032024e-06,
+      "loss": 0.4561,
+      "step": 8956
+    },
+    {
+      "epoch": 3.5052358607063474,
+      "grad_norm": 0.49005477886288673,
+      "learning_rate": 4.654649178410645e-06,
+      "loss": 0.4554,
+      "step": 8957
+    },
+    {
+      "epoch": 3.5056310200049396,
+      "grad_norm": 0.49167524318701145,
+      "learning_rate": 4.654569718327113e-06,
+      "loss": 0.4432,
+      "step": 8958
+    },
+    {
+      "epoch": 3.506026179303532,
+      "grad_norm": 0.48832196263631317,
+      "learning_rate": 4.654490249781739e-06,
+      "loss": 0.4567,
+      "step": 8959
+    },
+    {
+      "epoch": 3.506421338602124,
+      "grad_norm": 0.5195513981086755,
+      "learning_rate": 4.654410772774836e-06,
+      "loss": 0.4682,
+      "step": 8960
+    },
+    {
+      "epoch": 3.5068164979007164,
+      "grad_norm": 0.4883940235257672,
+      "learning_rate": 4.654331287306715e-06,
+      "loss": 0.4648,
+      "step": 8961
+    },
+    {
+      "epoch": 3.5072116571993086,
+      "grad_norm": 0.4939222640936546,
+      "learning_rate": 4.65425179337769e-06,
+      "loss": 0.4555,
+      "step": 8962
+    },
+    {
+      "epoch": 3.507606816497901,
+      "grad_norm": 0.507042086996444,
+      "learning_rate": 4.654172290988071e-06,
+      "loss": 0.4688,
+      "step": 8963
+    },
+    {
+      "epoch": 3.508001975796493,
+      "grad_norm": 0.4851747453407202,
+      "learning_rate": 4.654092780138172e-06,
+      "loss": 0.4667,
+      "step": 8964
+    },
+    {
+      "epoch": 3.5083971350950853,
+      "grad_norm": 0.49132032301674466,
+      "learning_rate": 4.654013260828304e-06,
+      "loss": 0.4558,
+      "step": 8965
+    },
+    {
+      "epoch": 3.5087922943936776,
+      "grad_norm": 0.5020222923780524,
+      "learning_rate": 4.653933733058781e-06,
+      "loss": 0.4471,
+      "step": 8966
+    },
+    {
+      "epoch": 3.50918745369227,
+      "grad_norm": 0.4894974516850361,
+      "learning_rate": 4.653854196829913e-06,
+      "loss": 0.442,
+      "step": 8967
+    },
+    {
+      "epoch": 3.509582612990862,
+      "grad_norm": 0.4868608182493534,
+      "learning_rate": 4.653774652142014e-06,
+      "loss": 0.4481,
+      "step": 8968
+    },
+    {
+      "epoch": 3.5099777722894543,
+      "grad_norm": 0.7314629885353082,
+      "learning_rate": 4.6536950989953965e-06,
+      "loss": 0.4573,
+      "step": 8969
+    },
+    {
+      "epoch": 3.5103729315880465,
+      "grad_norm": 0.48728983834155626,
+      "learning_rate": 4.653615537390371e-06,
+      "loss": 0.4548,
+      "step": 8970
+    },
+    {
+      "epoch": 3.5107680908866388,
+      "grad_norm": 0.49949151444479584,
+      "learning_rate": 4.653535967327253e-06,
+      "loss": 0.463,
+      "step": 8971
+    },
+    {
+      "epoch": 3.511163250185231,
+      "grad_norm": 0.4859042126598325,
+      "learning_rate": 4.6534563888063535e-06,
+      "loss": 0.4636,
+      "step": 8972
+    },
+    {
+      "epoch": 3.5115584094838233,
+      "grad_norm": 0.49909226950739194,
+      "learning_rate": 4.653376801827983e-06,
+      "loss": 0.4647,
+      "step": 8973
+    },
+    {
+      "epoch": 3.5119535687824155,
+      "grad_norm": 0.4841418346190016,
+      "learning_rate": 4.653297206392458e-06,
+      "loss": 0.4487,
+      "step": 8974
+    },
+    {
+      "epoch": 3.5123487280810077,
+      "grad_norm": 0.5017750095021966,
+      "learning_rate": 4.653217602500089e-06,
+      "loss": 0.4772,
+      "step": 8975
+    },
+    {
+      "epoch": 3.5127438873796,
+      "grad_norm": 0.5054355471499746,
+      "learning_rate": 4.653137990151188e-06,
+      "loss": 0.442,
+      "step": 8976
+    },
+    {
+      "epoch": 3.513139046678192,
+      "grad_norm": 0.5124016068780876,
+      "learning_rate": 4.6530583693460685e-06,
+      "loss": 0.4693,
+      "step": 8977
+    },
+    {
+      "epoch": 3.5135342059767845,
+      "grad_norm": 0.4884638972810405,
+      "learning_rate": 4.6529787400850435e-06,
+      "loss": 0.4549,
+      "step": 8978
+    },
+    {
+      "epoch": 3.5139293652753767,
+      "grad_norm": 0.484980903508112,
+      "learning_rate": 4.652899102368425e-06,
+      "loss": 0.4508,
+      "step": 8979
+    },
+    {
+      "epoch": 3.514324524573969,
+      "grad_norm": 0.4750814543537716,
+      "learning_rate": 4.652819456196527e-06,
+      "loss": 0.4455,
+      "step": 8980
+    },
+    {
+      "epoch": 3.514719683872561,
+      "grad_norm": 0.4787417445872702,
+      "learning_rate": 4.65273980156966e-06,
+      "loss": 0.4497,
+      "step": 8981
+    },
+    {
+      "epoch": 3.5151148431711534,
+      "grad_norm": 0.49080099109165837,
+      "learning_rate": 4.6526601384881396e-06,
+      "loss": 0.4487,
+      "step": 8982
+    },
+    {
+      "epoch": 3.5155100024697457,
+      "grad_norm": 0.47651127208635596,
+      "learning_rate": 4.652580466952277e-06,
+      "loss": 0.4488,
+      "step": 8983
+    },
+    {
+      "epoch": 3.515905161768338,
+      "grad_norm": 0.47966488268224355,
+      "learning_rate": 4.6525007869623845e-06,
+      "loss": 0.4513,
+      "step": 8984
+    },
+    {
+      "epoch": 3.51630032106693,
+      "grad_norm": 0.4878014502335229,
+      "learning_rate": 4.652421098518777e-06,
+      "loss": 0.4429,
+      "step": 8985
+    },
+    {
+      "epoch": 3.5166954803655224,
+      "grad_norm": 0.48298319426845326,
+      "learning_rate": 4.652341401621766e-06,
+      "loss": 0.448,
+      "step": 8986
+    },
+    {
+      "epoch": 3.5170906396641146,
+      "grad_norm": 0.47671515243568685,
+      "learning_rate": 4.652261696271665e-06,
+      "loss": 0.4482,
+      "step": 8987
+    },
+    {
+      "epoch": 3.517485798962707,
+      "grad_norm": 0.5012073706299464,
+      "learning_rate": 4.652181982468787e-06,
+      "loss": 0.4517,
+      "step": 8988
+    },
+    {
+      "epoch": 3.517880958261299,
+      "grad_norm": 0.48952067375088115,
+      "learning_rate": 4.652102260213445e-06,
+      "loss": 0.4697,
+      "step": 8989
+    },
+    {
+      "epoch": 3.5182761175598913,
+      "grad_norm": 0.5014312771832417,
+      "learning_rate": 4.652022529505953e-06,
+      "loss": 0.4617,
+      "step": 8990
+    },
+    {
+      "epoch": 3.5186712768584836,
+      "grad_norm": 0.48760811032686774,
+      "learning_rate": 4.651942790346622e-06,
+      "loss": 0.4775,
+      "step": 8991
+    },
+    {
+      "epoch": 3.519066436157076,
+      "grad_norm": 0.4804722298357196,
+      "learning_rate": 4.6518630427357666e-06,
+      "loss": 0.4565,
+      "step": 8992
+    },
+    {
+      "epoch": 3.519461595455668,
+      "grad_norm": 0.49171418019719676,
+      "learning_rate": 4.6517832866737e-06,
+      "loss": 0.4605,
+      "step": 8993
+    },
+    {
+      "epoch": 3.5198567547542603,
+      "grad_norm": 0.49918333071216625,
+      "learning_rate": 4.651703522160736e-06,
+      "loss": 0.463,
+      "step": 8994
+    },
+    {
+      "epoch": 3.5202519140528525,
+      "grad_norm": 0.49170780602129877,
+      "learning_rate": 4.651623749197187e-06,
+      "loss": 0.4683,
+      "step": 8995
+    },
+    {
+      "epoch": 3.520647073351445,
+      "grad_norm": 0.5033774543953301,
+      "learning_rate": 4.651543967783366e-06,
+      "loss": 0.4509,
+      "step": 8996
+    },
+    {
+      "epoch": 3.521042232650037,
+      "grad_norm": 0.6184118730606496,
+      "learning_rate": 4.651464177919587e-06,
+      "loss": 0.4398,
+      "step": 8997
+    },
+    {
+      "epoch": 3.5214373919486293,
+      "grad_norm": 0.5583292211707592,
+      "learning_rate": 4.6513843796061624e-06,
+      "loss": 0.441,
+      "step": 8998
+    },
+    {
+      "epoch": 3.5218325512472215,
+      "grad_norm": 0.5111210467079835,
+      "learning_rate": 4.651304572843407e-06,
+      "loss": 0.4463,
+      "step": 8999
+    },
+    {
+      "epoch": 3.5222277105458137,
+      "grad_norm": 0.4925902941362768,
+      "learning_rate": 4.651224757631634e-06,
+      "loss": 0.4565,
+      "step": 9000
+    },
+    {
+      "epoch": 3.522622869844406,
+      "grad_norm": 0.5053536793843054,
+      "learning_rate": 4.651144933971156e-06,
+      "loss": 0.454,
+      "step": 9001
+    },
+    {
+      "epoch": 3.523018029142998,
+      "grad_norm": 0.4939136118476549,
+      "learning_rate": 4.651065101862287e-06,
+      "loss": 0.4509,
+      "step": 9002
+    },
+    {
+      "epoch": 3.5234131884415905,
+      "grad_norm": 0.5038896891132291,
+      "learning_rate": 4.65098526130534e-06,
+      "loss": 0.438,
+      "step": 9003
+    },
+    {
+      "epoch": 3.5238083477401827,
+      "grad_norm": 0.5009202891855576,
+      "learning_rate": 4.65090541230063e-06,
+      "loss": 0.4411,
+      "step": 9004
+    },
+    {
+      "epoch": 3.524203507038775,
+      "grad_norm": 0.498130501386656,
+      "learning_rate": 4.650825554848468e-06,
+      "loss": 0.4614,
+      "step": 9005
+    },
+    {
+      "epoch": 3.524598666337367,
+      "grad_norm": 0.5107785107225102,
+      "learning_rate": 4.650745688949171e-06,
+      "loss": 0.4641,
+      "step": 9006
+    },
+    {
+      "epoch": 3.5249938256359594,
+      "grad_norm": 0.4973684335934365,
+      "learning_rate": 4.65066581460305e-06,
+      "loss": 0.4511,
+      "step": 9007
+    },
+    {
+      "epoch": 3.5253889849345517,
+      "grad_norm": 0.5120248873540496,
+      "learning_rate": 4.650585931810421e-06,
+      "loss": 0.4767,
+      "step": 9008
+    },
+    {
+      "epoch": 3.525784144233144,
+      "grad_norm": 0.5032821292378913,
+      "learning_rate": 4.6505060405715944e-06,
+      "loss": 0.4502,
+      "step": 9009
+    },
+    {
+      "epoch": 3.526179303531736,
+      "grad_norm": 0.4956958147919807,
+      "learning_rate": 4.6504261408868875e-06,
+      "loss": 0.4561,
+      "step": 9010
+    },
+    {
+      "epoch": 3.5265744628303284,
+      "grad_norm": 0.4957187392523741,
+      "learning_rate": 4.6503462327566115e-06,
+      "loss": 0.4681,
+      "step": 9011
+    },
+    {
+      "epoch": 3.5269696221289206,
+      "grad_norm": 0.4952159771925259,
+      "learning_rate": 4.650266316181082e-06,
+      "loss": 0.4461,
+      "step": 9012
+    },
+    {
+      "epoch": 3.527364781427513,
+      "grad_norm": 0.5672869678695379,
+      "learning_rate": 4.650186391160611e-06,
+      "loss": 0.4461,
+      "step": 9013
+    },
+    {
+      "epoch": 3.527759940726105,
+      "grad_norm": 0.5080704469053954,
+      "learning_rate": 4.650106457695515e-06,
+      "loss": 0.4708,
+      "step": 9014
+    },
+    {
+      "epoch": 3.5281551000246973,
+      "grad_norm": 0.4936767560240317,
+      "learning_rate": 4.650026515786106e-06,
+      "loss": 0.4585,
+      "step": 9015
+    },
+    {
+      "epoch": 3.5285502593232896,
+      "grad_norm": 0.5008042637689482,
+      "learning_rate": 4.649946565432698e-06,
+      "loss": 0.4692,
+      "step": 9016
+    },
+    {
+      "epoch": 3.528945418621882,
+      "grad_norm": 0.4962751637351177,
+      "learning_rate": 4.649866606635605e-06,
+      "loss": 0.4354,
+      "step": 9017
+    },
+    {
+      "epoch": 3.529340577920474,
+      "grad_norm": 0.5008934519010351,
+      "learning_rate": 4.649786639395142e-06,
+      "loss": 0.4699,
+      "step": 9018
+    },
+    {
+      "epoch": 3.5297357372190663,
+      "grad_norm": 0.48596615093437223,
+      "learning_rate": 4.649706663711623e-06,
+      "loss": 0.461,
+      "step": 9019
+    },
+    {
+      "epoch": 3.5301308965176585,
+      "grad_norm": 0.49089635795809994,
+      "learning_rate": 4.649626679585361e-06,
+      "loss": 0.4614,
+      "step": 9020
+    },
+    {
+      "epoch": 3.530526055816251,
+      "grad_norm": 0.4946185325035558,
+      "learning_rate": 4.64954668701667e-06,
+      "loss": 0.4487,
+      "step": 9021
+    },
+    {
+      "epoch": 3.530921215114843,
+      "grad_norm": 0.5126278761898772,
+      "learning_rate": 4.6494666860058655e-06,
+      "loss": 0.4568,
+      "step": 9022
+    },
+    {
+      "epoch": 3.5313163744134353,
+      "grad_norm": 0.5285391391199131,
+      "learning_rate": 4.649386676553262e-06,
+      "loss": 0.4557,
+      "step": 9023
+    },
+    {
+      "epoch": 3.5317115337120275,
+      "grad_norm": 0.5038626087100095,
+      "learning_rate": 4.649306658659172e-06,
+      "loss": 0.4707,
+      "step": 9024
+    },
+    {
+      "epoch": 3.5321066930106197,
+      "grad_norm": 0.4859360226761564,
+      "learning_rate": 4.649226632323911e-06,
+      "loss": 0.4837,
+      "step": 9025
+    },
+    {
+      "epoch": 3.532501852309212,
+      "grad_norm": 0.4966668712438498,
+      "learning_rate": 4.649146597547792e-06,
+      "loss": 0.4648,
+      "step": 9026
+    },
+    {
+      "epoch": 3.5328970116078047,
+      "grad_norm": 0.48166044464094737,
+      "learning_rate": 4.649066554331131e-06,
+      "loss": 0.4455,
+      "step": 9027
+    },
+    {
+      "epoch": 3.533292170906397,
+      "grad_norm": 0.49847140402641615,
+      "learning_rate": 4.648986502674241e-06,
+      "loss": 0.4745,
+      "step": 9028
+    },
+    {
+      "epoch": 3.533687330204989,
+      "grad_norm": 0.48375147299765764,
+      "learning_rate": 4.6489064425774375e-06,
+      "loss": 0.448,
+      "step": 9029
+    },
+    {
+      "epoch": 3.5340824895035814,
+      "grad_norm": 0.501937133054762,
+      "learning_rate": 4.648826374041034e-06,
+      "loss": 0.4857,
+      "step": 9030
+    },
+    {
+      "epoch": 3.5344776488021736,
+      "grad_norm": 0.5027543628434751,
+      "learning_rate": 4.648746297065345e-06,
+      "loss": 0.4561,
+      "step": 9031
+    },
+    {
+      "epoch": 3.534872808100766,
+      "grad_norm": 0.48644400230014095,
+      "learning_rate": 4.648666211650686e-06,
+      "loss": 0.4511,
+      "step": 9032
+    },
+    {
+      "epoch": 3.535267967399358,
+      "grad_norm": 0.492175793746889,
+      "learning_rate": 4.648586117797371e-06,
+      "loss": 0.4451,
+      "step": 9033
+    },
+    {
+      "epoch": 3.5356631266979504,
+      "grad_norm": 0.4854671322850227,
+      "learning_rate": 4.648506015505714e-06,
+      "loss": 0.4627,
+      "step": 9034
+    },
+    {
+      "epoch": 3.5360582859965426,
+      "grad_norm": 0.4972183962198117,
+      "learning_rate": 4.64842590477603e-06,
+      "loss": 0.4699,
+      "step": 9035
+    },
+    {
+      "epoch": 3.536453445295135,
+      "grad_norm": 0.5200452269905167,
+      "learning_rate": 4.648345785608633e-06,
+      "loss": 0.4788,
+      "step": 9036
+    },
+    {
+      "epoch": 3.536848604593727,
+      "grad_norm": 0.530977112773409,
+      "learning_rate": 4.64826565800384e-06,
+      "loss": 0.4598,
+      "step": 9037
+    },
+    {
+      "epoch": 3.5372437638923193,
+      "grad_norm": 0.48919921086325696,
+      "learning_rate": 4.648185521961963e-06,
+      "loss": 0.4605,
+      "step": 9038
+    },
+    {
+      "epoch": 3.5376389231909116,
+      "grad_norm": 0.48321014692774794,
+      "learning_rate": 4.648105377483318e-06,
+      "loss": 0.4562,
+      "step": 9039
+    },
+    {
+      "epoch": 3.538034082489504,
+      "grad_norm": 0.5096165386730921,
+      "learning_rate": 4.648025224568219e-06,
+      "loss": 0.4709,
+      "step": 9040
+    },
+    {
+      "epoch": 3.538429241788096,
+      "grad_norm": 0.4992741223806897,
+      "learning_rate": 4.647945063216981e-06,
+      "loss": 0.4531,
+      "step": 9041
+    },
+    {
+      "epoch": 3.5388244010866883,
+      "grad_norm": 0.5075195911086815,
+      "learning_rate": 4.64786489342992e-06,
+      "loss": 0.4482,
+      "step": 9042
+    },
+    {
+      "epoch": 3.5392195603852805,
+      "grad_norm": 0.4817099394481762,
+      "learning_rate": 4.64778471520735e-06,
+      "loss": 0.4566,
+      "step": 9043
+    },
+    {
+      "epoch": 3.5396147196838728,
+      "grad_norm": 0.4936601585082596,
+      "learning_rate": 4.6477045285495845e-06,
+      "loss": 0.4618,
+      "step": 9044
+    },
+    {
+      "epoch": 3.540009878982465,
+      "grad_norm": 0.4892077110552015,
+      "learning_rate": 4.647624333456941e-06,
+      "loss": 0.4588,
+      "step": 9045
+    },
+    {
+      "epoch": 3.5404050382810572,
+      "grad_norm": 0.4969970473771617,
+      "learning_rate": 4.647544129929733e-06,
+      "loss": 0.4653,
+      "step": 9046
+    },
+    {
+      "epoch": 3.5408001975796495,
+      "grad_norm": 0.4819114190057846,
+      "learning_rate": 4.647463917968275e-06,
+      "loss": 0.4526,
+      "step": 9047
+    },
+    {
+      "epoch": 3.5411953568782417,
+      "grad_norm": 0.48764776197722826,
+      "learning_rate": 4.647383697572883e-06,
+      "loss": 0.4338,
+      "step": 9048
+    },
+    {
+      "epoch": 3.541590516176834,
+      "grad_norm": 0.4818100110717255,
+      "learning_rate": 4.647303468743873e-06,
+      "loss": 0.4755,
+      "step": 9049
+    },
+    {
+      "epoch": 3.541985675475426,
+      "grad_norm": 0.5005030935665212,
+      "learning_rate": 4.647223231481557e-06,
+      "loss": 0.4614,
+      "step": 9050
+    },
+    {
+      "epoch": 3.5423808347740184,
+      "grad_norm": 0.5150064517997461,
+      "learning_rate": 4.647142985786254e-06,
+      "loss": 0.4648,
+      "step": 9051
+    },
+    {
+      "epoch": 3.5427759940726107,
+      "grad_norm": 0.48527068890730546,
+      "learning_rate": 4.647062731658276e-06,
+      "loss": 0.4558,
+      "step": 9052
+    },
+    {
+      "epoch": 3.543171153371203,
+      "grad_norm": 0.5093185573266966,
+      "learning_rate": 4.6469824690979394e-06,
+      "loss": 0.4848,
+      "step": 9053
+    },
+    {
+      "epoch": 3.543566312669795,
+      "grad_norm": 0.5371264924132514,
+      "learning_rate": 4.64690219810556e-06,
+      "loss": 0.4721,
+      "step": 9054
+    },
+    {
+      "epoch": 3.5439614719683874,
+      "grad_norm": 0.49196835315592014,
+      "learning_rate": 4.646821918681451e-06,
+      "loss": 0.4584,
+      "step": 9055
+    },
+    {
+      "epoch": 3.5443566312669796,
+      "grad_norm": 0.5095176634298171,
+      "learning_rate": 4.64674163082593e-06,
+      "loss": 0.454,
+      "step": 9056
+    },
+    {
+      "epoch": 3.544751790565572,
+      "grad_norm": 0.48759629545526767,
+      "learning_rate": 4.646661334539312e-06,
+      "loss": 0.4679,
+      "step": 9057
+    },
+    {
+      "epoch": 3.545146949864164,
+      "grad_norm": 0.4870613246032581,
+      "learning_rate": 4.646581029821912e-06,
+      "loss": 0.4512,
+      "step": 9058
+    },
+    {
+      "epoch": 3.5455421091627564,
+      "grad_norm": 0.4823673270563382,
+      "learning_rate": 4.646500716674044e-06,
+      "loss": 0.4649,
+      "step": 9059
+    },
+    {
+      "epoch": 3.5459372684613486,
+      "grad_norm": 0.4977917296918362,
+      "learning_rate": 4.646420395096025e-06,
+      "loss": 0.4615,
+      "step": 9060
+    },
+    {
+      "epoch": 3.546332427759941,
+      "grad_norm": 0.4981541055070267,
+      "learning_rate": 4.646340065088169e-06,
+      "loss": 0.4505,
+      "step": 9061
+    },
+    {
+      "epoch": 3.546727587058533,
+      "grad_norm": 0.4768432018824217,
+      "learning_rate": 4.646259726650795e-06,
+      "loss": 0.4401,
+      "step": 9062
+    },
+    {
+      "epoch": 3.5471227463571253,
+      "grad_norm": 0.49995907837696146,
+      "learning_rate": 4.646179379784215e-06,
+      "loss": 0.4566,
+      "step": 9063
+    },
+    {
+      "epoch": 3.5475179056557176,
+      "grad_norm": 0.529211500161168,
+      "learning_rate": 4.646099024488745e-06,
+      "loss": 0.4578,
+      "step": 9064
+    },
+    {
+      "epoch": 3.54791306495431,
+      "grad_norm": 0.49321795179014727,
+      "learning_rate": 4.646018660764702e-06,
+      "loss": 0.4441,
+      "step": 9065
+    },
+    {
+      "epoch": 3.548308224252902,
+      "grad_norm": 0.49786103565113565,
+      "learning_rate": 4.6459382886123996e-06,
+      "loss": 0.4731,
+      "step": 9066
+    },
+    {
+      "epoch": 3.5487033835514943,
+      "grad_norm": 0.48219336482722647,
+      "learning_rate": 4.645857908032156e-06,
+      "loss": 0.4381,
+      "step": 9067
+    },
+    {
+      "epoch": 3.5490985428500865,
+      "grad_norm": 0.4923319509953273,
+      "learning_rate": 4.645777519024285e-06,
+      "loss": 0.4582,
+      "step": 9068
+    },
+    {
+      "epoch": 3.5494937021486788,
+      "grad_norm": 0.4865553175937021,
+      "learning_rate": 4.645697121589103e-06,
+      "loss": 0.4423,
+      "step": 9069
+    },
+    {
+      "epoch": 3.549888861447271,
+      "grad_norm": 0.48095897113056707,
+      "learning_rate": 4.645616715726926e-06,
+      "loss": 0.448,
+      "step": 9070
+    },
+    {
+      "epoch": 3.5502840207458632,
+      "grad_norm": 0.5001311843314722,
+      "learning_rate": 4.645536301438069e-06,
+      "loss": 0.4466,
+      "step": 9071
+    },
+    {
+      "epoch": 3.5506791800444555,
+      "grad_norm": 0.477997428256851,
+      "learning_rate": 4.645455878722848e-06,
+      "loss": 0.4532,
+      "step": 9072
+    },
+    {
+      "epoch": 3.5510743393430477,
+      "grad_norm": 0.49193722985280725,
+      "learning_rate": 4.64537544758158e-06,
+      "loss": 0.4599,
+      "step": 9073
+    },
+    {
+      "epoch": 3.55146949864164,
+      "grad_norm": 0.4920447859862548,
+      "learning_rate": 4.645295008014579e-06,
+      "loss": 0.4532,
+      "step": 9074
+    },
+    {
+      "epoch": 3.551864657940232,
+      "grad_norm": 0.49558747516046264,
+      "learning_rate": 4.645214560022162e-06,
+      "loss": 0.4736,
+      "step": 9075
+    },
+    {
+      "epoch": 3.5522598172388244,
+      "grad_norm": 0.4762337243641712,
+      "learning_rate": 4.6451341036046455e-06,
+      "loss": 0.4517,
+      "step": 9076
+    },
+    {
+      "epoch": 3.5526549765374167,
+      "grad_norm": 0.5013245757181765,
+      "learning_rate": 4.6450536387623444e-06,
+      "loss": 0.4648,
+      "step": 9077
+    },
+    {
+      "epoch": 3.553050135836009,
+      "grad_norm": 0.5403068478658516,
+      "learning_rate": 4.644973165495576e-06,
+      "loss": 0.4449,
+      "step": 9078
+    },
+    {
+      "epoch": 3.553445295134601,
+      "grad_norm": 0.4931560024394474,
+      "learning_rate": 4.644892683804653e-06,
+      "loss": 0.4646,
+      "step": 9079
+    },
+    {
+      "epoch": 3.5538404544331934,
+      "grad_norm": 0.5078764038379829,
+      "learning_rate": 4.644812193689897e-06,
+      "loss": 0.4648,
+      "step": 9080
+    },
+    {
+      "epoch": 3.5542356137317856,
+      "grad_norm": 0.506957616825252,
+      "learning_rate": 4.64473169515162e-06,
+      "loss": 0.4636,
+      "step": 9081
+    },
+    {
+      "epoch": 3.554630773030378,
+      "grad_norm": 0.507432762723005,
+      "learning_rate": 4.644651188190139e-06,
+      "loss": 0.4732,
+      "step": 9082
+    },
+    {
+      "epoch": 3.55502593232897,
+      "grad_norm": 0.48687923691371193,
+      "learning_rate": 4.6445706728057705e-06,
+      "loss": 0.4512,
+      "step": 9083
+    },
+    {
+      "epoch": 3.5554210916275624,
+      "grad_norm": 0.6304721111076129,
+      "learning_rate": 4.64449014899883e-06,
+      "loss": 0.4679,
+      "step": 9084
+    },
+    {
+      "epoch": 3.5558162509261546,
+      "grad_norm": 0.4888618812137909,
+      "learning_rate": 4.644409616769635e-06,
+      "loss": 0.458,
+      "step": 9085
+    },
+    {
+      "epoch": 3.556211410224747,
+      "grad_norm": 0.4859315158657899,
+      "learning_rate": 4.644329076118502e-06,
+      "loss": 0.4544,
+      "step": 9086
+    },
+    {
+      "epoch": 3.556606569523339,
+      "grad_norm": 0.4927974306644837,
+      "learning_rate": 4.6442485270457464e-06,
+      "loss": 0.4546,
+      "step": 9087
+    },
+    {
+      "epoch": 3.5570017288219313,
+      "grad_norm": 0.46978089261360473,
+      "learning_rate": 4.644167969551683e-06,
+      "loss": 0.4435,
+      "step": 9088
+    },
+    {
+      "epoch": 3.5573968881205236,
+      "grad_norm": 0.5002112740039599,
+      "learning_rate": 4.644087403636631e-06,
+      "loss": 0.4575,
+      "step": 9089
+    },
+    {
+      "epoch": 3.557792047419116,
+      "grad_norm": 0.4816381481041411,
+      "learning_rate": 4.644006829300906e-06,
+      "loss": 0.4441,
+      "step": 9090
+    },
+    {
+      "epoch": 3.558187206717708,
+      "grad_norm": 0.5106665981705985,
+      "learning_rate": 4.643926246544823e-06,
+      "loss": 0.4611,
+      "step": 9091
+    },
+    {
+      "epoch": 3.5585823660163003,
+      "grad_norm": 0.49510570450790387,
+      "learning_rate": 4.6438456553687e-06,
+      "loss": 0.4687,
+      "step": 9092
+    },
+    {
+      "epoch": 3.5589775253148925,
+      "grad_norm": 0.507508378297751,
+      "learning_rate": 4.6437650557728535e-06,
+      "loss": 0.4652,
+      "step": 9093
+    },
+    {
+      "epoch": 3.5593726846134848,
+      "grad_norm": 0.49897649507220837,
+      "learning_rate": 4.643684447757599e-06,
+      "loss": 0.45,
+      "step": 9094
+    },
+    {
+      "epoch": 3.559767843912077,
+      "grad_norm": 0.5042685832105795,
+      "learning_rate": 4.643603831323255e-06,
+      "loss": 0.447,
+      "step": 9095
+    },
+    {
+      "epoch": 3.5601630032106693,
+      "grad_norm": 0.5209604652999842,
+      "learning_rate": 4.643523206470135e-06,
+      "loss": 0.4588,
+      "step": 9096
+    },
+    {
+      "epoch": 3.5605581625092615,
+      "grad_norm": 0.49949487365809175,
+      "learning_rate": 4.6434425731985585e-06,
+      "loss": 0.4741,
+      "step": 9097
+    },
+    {
+      "epoch": 3.5609533218078537,
+      "grad_norm": 0.5157946856320857,
+      "learning_rate": 4.643361931508841e-06,
+      "loss": 0.4709,
+      "step": 9098
+    },
+    {
+      "epoch": 3.561348481106446,
+      "grad_norm": 0.5076379876922849,
+      "learning_rate": 4.6432812814013e-06,
+      "loss": 0.4518,
+      "step": 9099
+    },
+    {
+      "epoch": 3.561743640405038,
+      "grad_norm": 0.48357121942351644,
+      "learning_rate": 4.64320062287625e-06,
+      "loss": 0.4689,
+      "step": 9100
+    },
+    {
+      "epoch": 3.5621387997036305,
+      "grad_norm": 0.4851016675612043,
+      "learning_rate": 4.64311995593401e-06,
+      "loss": 0.4623,
+      "step": 9101
+    },
+    {
+      "epoch": 3.5625339590022227,
+      "grad_norm": 0.5032142946703332,
+      "learning_rate": 4.643039280574897e-06,
+      "loss": 0.4569,
+      "step": 9102
+    },
+    {
+      "epoch": 3.562929118300815,
+      "grad_norm": 0.4981056358231815,
+      "learning_rate": 4.642958596799226e-06,
+      "loss": 0.4652,
+      "step": 9103
+    },
+    {
+      "epoch": 3.563324277599407,
+      "grad_norm": 0.5045031493173247,
+      "learning_rate": 4.642877904607316e-06,
+      "loss": 0.4757,
+      "step": 9104
+    },
+    {
+      "epoch": 3.5637194368979994,
+      "grad_norm": 0.49929741491236107,
+      "learning_rate": 4.642797203999482e-06,
+      "loss": 0.4553,
+      "step": 9105
+    },
+    {
+      "epoch": 3.5641145961965917,
+      "grad_norm": 0.48493063212795195,
+      "learning_rate": 4.6427164949760415e-06,
+      "loss": 0.4519,
+      "step": 9106
+    },
+    {
+      "epoch": 3.564509755495184,
+      "grad_norm": 0.4825224454556553,
+      "learning_rate": 4.642635777537312e-06,
+      "loss": 0.4657,
+      "step": 9107
+    },
+    {
+      "epoch": 3.564904914793776,
+      "grad_norm": 0.4923598417193523,
+      "learning_rate": 4.6425550516836106e-06,
+      "loss": 0.4535,
+      "step": 9108
+    },
+    {
+      "epoch": 3.5653000740923684,
+      "grad_norm": 0.5002848790092057,
+      "learning_rate": 4.6424743174152544e-06,
+      "loss": 0.4617,
+      "step": 9109
+    },
+    {
+      "epoch": 3.5656952333909606,
+      "grad_norm": 0.5246927195871764,
+      "learning_rate": 4.642393574732559e-06,
+      "loss": 0.4581,
+      "step": 9110
+    },
+    {
+      "epoch": 3.566090392689553,
+      "grad_norm": 0.5054517893159672,
+      "learning_rate": 4.642312823635843e-06,
+      "loss": 0.4571,
+      "step": 9111
+    },
+    {
+      "epoch": 3.566485551988145,
+      "grad_norm": 0.483406457593407,
+      "learning_rate": 4.642232064125424e-06,
+      "loss": 0.4415,
+      "step": 9112
+    },
+    {
+      "epoch": 3.5668807112867373,
+      "grad_norm": 0.49885049133184317,
+      "learning_rate": 4.642151296201617e-06,
+      "loss": 0.45,
+      "step": 9113
+    },
+    {
+      "epoch": 3.5672758705853296,
+      "grad_norm": 0.6059844092001336,
+      "learning_rate": 4.6420705198647405e-06,
+      "loss": 0.4584,
+      "step": 9114
+    },
+    {
+      "epoch": 3.567671029883922,
+      "grad_norm": 0.49181406902845737,
+      "learning_rate": 4.641989735115112e-06,
+      "loss": 0.4593,
+      "step": 9115
+    },
+    {
+      "epoch": 3.568066189182514,
+      "grad_norm": 0.4942661823361688,
+      "learning_rate": 4.64190894195305e-06,
+      "loss": 0.4683,
+      "step": 9116
+    },
+    {
+      "epoch": 3.5684613484811063,
+      "grad_norm": 0.4954476984028424,
+      "learning_rate": 4.641828140378868e-06,
+      "loss": 0.4657,
+      "step": 9117
+    },
+    {
+      "epoch": 3.5688565077796985,
+      "grad_norm": 0.4937799659377194,
+      "learning_rate": 4.641747330392886e-06,
+      "loss": 0.4503,
+      "step": 9118
+    },
+    {
+      "epoch": 3.569251667078291,
+      "grad_norm": 0.4902163898739017,
+      "learning_rate": 4.641666511995422e-06,
+      "loss": 0.463,
+      "step": 9119
+    },
+    {
+      "epoch": 3.569646826376883,
+      "grad_norm": 0.5043714006810839,
+      "learning_rate": 4.641585685186792e-06,
+      "loss": 0.4624,
+      "step": 9120
+    },
+    {
+      "epoch": 3.5700419856754753,
+      "grad_norm": 0.5004840014877152,
+      "learning_rate": 4.641504849967315e-06,
+      "loss": 0.4879,
+      "step": 9121
+    },
+    {
+      "epoch": 3.5704371449740675,
+      "grad_norm": 0.5077783702630878,
+      "learning_rate": 4.6414240063373065e-06,
+      "loss": 0.4456,
+      "step": 9122
+    },
+    {
+      "epoch": 3.5708323042726597,
+      "grad_norm": 0.4824382259117086,
+      "learning_rate": 4.6413431542970845e-06,
+      "loss": 0.4638,
+      "step": 9123
+    },
+    {
+      "epoch": 3.571227463571252,
+      "grad_norm": 0.4871993531897932,
+      "learning_rate": 4.641262293846966e-06,
+      "loss": 0.4599,
+      "step": 9124
+    },
+    {
+      "epoch": 3.571622622869844,
+      "grad_norm": 0.49457463356784,
+      "learning_rate": 4.641181424987271e-06,
+      "loss": 0.4611,
+      "step": 9125
+    },
+    {
+      "epoch": 3.5720177821684365,
+      "grad_norm": 0.48893854230152123,
+      "learning_rate": 4.641100547718314e-06,
+      "loss": 0.4459,
+      "step": 9126
+    },
+    {
+      "epoch": 3.5724129414670287,
+      "grad_norm": 0.4814786016167628,
+      "learning_rate": 4.641019662040417e-06,
+      "loss": 0.4518,
+      "step": 9127
+    },
+    {
+      "epoch": 3.572808100765621,
+      "grad_norm": 0.4990303217153392,
+      "learning_rate": 4.6409387679538925e-06,
+      "loss": 0.4668,
+      "step": 9128
+    },
+    {
+      "epoch": 3.573203260064213,
+      "grad_norm": 0.5001213765361451,
+      "learning_rate": 4.640857865459061e-06,
+      "loss": 0.4632,
+      "step": 9129
+    },
+    {
+      "epoch": 3.5735984193628054,
+      "grad_norm": 0.47937361186876326,
+      "learning_rate": 4.6407769545562395e-06,
+      "loss": 0.4403,
+      "step": 9130
+    },
+    {
+      "epoch": 3.5739935786613977,
+      "grad_norm": 0.4886845046562659,
+      "learning_rate": 4.6406960352457476e-06,
+      "loss": 0.4671,
+      "step": 9131
+    },
+    {
+      "epoch": 3.57438873795999,
+      "grad_norm": 0.6230155087433286,
+      "learning_rate": 4.6406151075279e-06,
+      "loss": 0.4663,
+      "step": 9132
+    },
+    {
+      "epoch": 3.574783897258582,
+      "grad_norm": 0.49211316752499784,
+      "learning_rate": 4.640534171403017e-06,
+      "loss": 0.4662,
+      "step": 9133
+    },
+    {
+      "epoch": 3.5751790565571744,
+      "grad_norm": 0.47931129273412515,
+      "learning_rate": 4.640453226871415e-06,
+      "loss": 0.462,
+      "step": 9134
+    },
+    {
+      "epoch": 3.5755742158557666,
+      "grad_norm": 0.4801125074561214,
+      "learning_rate": 4.640372273933412e-06,
+      "loss": 0.4514,
+      "step": 9135
+    },
+    {
+      "epoch": 3.575969375154359,
+      "grad_norm": 0.48855690347192954,
+      "learning_rate": 4.6402913125893275e-06,
+      "loss": 0.4655,
+      "step": 9136
+    },
+    {
+      "epoch": 3.576364534452951,
+      "grad_norm": 0.4902447553553269,
+      "learning_rate": 4.640210342839479e-06,
+      "loss": 0.4594,
+      "step": 9137
+    },
+    {
+      "epoch": 3.5767596937515433,
+      "grad_norm": 0.48368454552996915,
+      "learning_rate": 4.640129364684182e-06,
+      "loss": 0.4566,
+      "step": 9138
+    },
+    {
+      "epoch": 3.5771548530501356,
+      "grad_norm": 0.514151993001096,
+      "learning_rate": 4.640048378123757e-06,
+      "loss": 0.4547,
+      "step": 9139
+    },
+    {
+      "epoch": 3.577550012348728,
+      "grad_norm": 0.4776789489243305,
+      "learning_rate": 4.639967383158523e-06,
+      "loss": 0.4598,
+      "step": 9140
+    },
+    {
+      "epoch": 3.57794517164732,
+      "grad_norm": 0.48525569302537913,
+      "learning_rate": 4.639886379788794e-06,
+      "loss": 0.4418,
+      "step": 9141
+    },
+    {
+      "epoch": 3.5783403309459123,
+      "grad_norm": 0.478721665258136,
+      "learning_rate": 4.6398053680148926e-06,
+      "loss": 0.4651,
+      "step": 9142
+    },
+    {
+      "epoch": 3.5787354902445045,
+      "grad_norm": 0.47444493198435933,
+      "learning_rate": 4.639724347837135e-06,
+      "loss": 0.4424,
+      "step": 9143
+    },
+    {
+      "epoch": 3.579130649543097,
+      "grad_norm": 0.5070133096361706,
+      "learning_rate": 4.639643319255838e-06,
+      "loss": 0.466,
+      "step": 9144
+    },
+    {
+      "epoch": 3.5795258088416895,
+      "grad_norm": 0.4866231994143781,
+      "learning_rate": 4.639562282271323e-06,
+      "loss": 0.4792,
+      "step": 9145
+    },
+    {
+      "epoch": 3.5799209681402817,
+      "grad_norm": 0.513869457185633,
+      "learning_rate": 4.6394812368839055e-06,
+      "loss": 0.4557,
+      "step": 9146
+    },
+    {
+      "epoch": 3.580316127438874,
+      "grad_norm": 0.48227115983762886,
+      "learning_rate": 4.639400183093905e-06,
+      "loss": 0.4516,
+      "step": 9147
+    },
+    {
+      "epoch": 3.580711286737466,
+      "grad_norm": 0.5869304828960821,
+      "learning_rate": 4.63931912090164e-06,
+      "loss": 0.4585,
+      "step": 9148
+    },
+    {
+      "epoch": 3.5811064460360584,
+      "grad_norm": 0.4882734722747367,
+      "learning_rate": 4.639238050307428e-06,
+      "loss": 0.4565,
+      "step": 9149
+    },
+    {
+      "epoch": 3.5815016053346507,
+      "grad_norm": 0.5044115242028842,
+      "learning_rate": 4.639156971311589e-06,
+      "loss": 0.4693,
+      "step": 9150
+    },
+    {
+      "epoch": 3.581896764633243,
+      "grad_norm": 0.5019659383311873,
+      "learning_rate": 4.63907588391444e-06,
+      "loss": 0.4629,
+      "step": 9151
+    },
+    {
+      "epoch": 3.582291923931835,
+      "grad_norm": 0.5141519049977628,
+      "learning_rate": 4.638994788116299e-06,
+      "loss": 0.4677,
+      "step": 9152
+    },
+    {
+      "epoch": 3.5826870832304274,
+      "grad_norm": 0.4785979998489708,
+      "learning_rate": 4.638913683917486e-06,
+      "loss": 0.4595,
+      "step": 9153
+    },
+    {
+      "epoch": 3.5830822425290196,
+      "grad_norm": 0.4922496588594111,
+      "learning_rate": 4.638832571318319e-06,
+      "loss": 0.4422,
+      "step": 9154
+    },
+    {
+      "epoch": 3.583477401827612,
+      "grad_norm": 0.49580896733041097,
+      "learning_rate": 4.6387514503191165e-06,
+      "loss": 0.448,
+      "step": 9155
+    },
+    {
+      "epoch": 3.583872561126204,
+      "grad_norm": 0.48090076896396333,
+      "learning_rate": 4.638670320920196e-06,
+      "loss": 0.4618,
+      "step": 9156
+    },
+    {
+      "epoch": 3.5842677204247964,
+      "grad_norm": 0.4790421513834484,
+      "learning_rate": 4.638589183121879e-06,
+      "loss": 0.467,
+      "step": 9157
+    },
+    {
+      "epoch": 3.5846628797233886,
+      "grad_norm": 0.4878225988687124,
+      "learning_rate": 4.63850803692448e-06,
+      "loss": 0.4849,
+      "step": 9158
+    },
+    {
+      "epoch": 3.585058039021981,
+      "grad_norm": 0.4870640632594355,
+      "learning_rate": 4.638426882328322e-06,
+      "loss": 0.4645,
+      "step": 9159
+    },
+    {
+      "epoch": 3.585453198320573,
+      "grad_norm": 0.5050638549536798,
+      "learning_rate": 4.638345719333721e-06,
+      "loss": 0.4572,
+      "step": 9160
+    },
+    {
+      "epoch": 3.5858483576191653,
+      "grad_norm": 0.5069748386116536,
+      "learning_rate": 4.638264547940996e-06,
+      "loss": 0.4553,
+      "step": 9161
+    },
+    {
+      "epoch": 3.5862435169177576,
+      "grad_norm": 0.5578612016671592,
+      "learning_rate": 4.6381833681504675e-06,
+      "loss": 0.4625,
+      "step": 9162
+    },
+    {
+      "epoch": 3.58663867621635,
+      "grad_norm": 0.4656285059162261,
+      "learning_rate": 4.638102179962452e-06,
+      "loss": 0.4616,
+      "step": 9163
+    },
+    {
+      "epoch": 3.587033835514942,
+      "grad_norm": 0.4795082800303555,
+      "learning_rate": 4.63802098337727e-06,
+      "loss": 0.4658,
+      "step": 9164
+    },
+    {
+      "epoch": 3.5874289948135343,
+      "grad_norm": 0.4989309789649853,
+      "learning_rate": 4.637939778395239e-06,
+      "loss": 0.4533,
+      "step": 9165
+    },
+    {
+      "epoch": 3.5878241541121265,
+      "grad_norm": 0.4790162592288028,
+      "learning_rate": 4.637858565016679e-06,
+      "loss": 0.4608,
+      "step": 9166
+    },
+    {
+      "epoch": 3.5882193134107188,
+      "grad_norm": 0.4881752317773561,
+      "learning_rate": 4.6377773432419105e-06,
+      "loss": 0.4633,
+      "step": 9167
+    },
+    {
+      "epoch": 3.588614472709311,
+      "grad_norm": 0.4892512885912421,
+      "learning_rate": 4.637696113071249e-06,
+      "loss": 0.4697,
+      "step": 9168
+    },
+    {
+      "epoch": 3.5890096320079032,
+      "grad_norm": 0.4723594520462734,
+      "learning_rate": 4.637614874505016e-06,
+      "loss": 0.4436,
+      "step": 9169
+    },
+    {
+      "epoch": 3.5894047913064955,
+      "grad_norm": 0.4893238305591922,
+      "learning_rate": 4.637533627543529e-06,
+      "loss": 0.4557,
+      "step": 9170
+    },
+    {
+      "epoch": 3.5897999506050877,
+      "grad_norm": 0.48322974662494034,
+      "learning_rate": 4.637452372187109e-06,
+      "loss": 0.4547,
+      "step": 9171
+    },
+    {
+      "epoch": 3.59019510990368,
+      "grad_norm": 0.5042287973047576,
+      "learning_rate": 4.6373711084360725e-06,
+      "loss": 0.4595,
+      "step": 9172
+    },
+    {
+      "epoch": 3.590590269202272,
+      "grad_norm": 0.48463044178358355,
+      "learning_rate": 4.637289836290741e-06,
+      "loss": 0.4553,
+      "step": 9173
+    },
+    {
+      "epoch": 3.5909854285008644,
+      "grad_norm": 0.48284154312198374,
+      "learning_rate": 4.6372085557514335e-06,
+      "loss": 0.4398,
+      "step": 9174
+    },
+    {
+      "epoch": 3.5913805877994567,
+      "grad_norm": 0.48052961472036454,
+      "learning_rate": 4.637127266818467e-06,
+      "loss": 0.4667,
+      "step": 9175
+    },
+    {
+      "epoch": 3.591775747098049,
+      "grad_norm": 0.5272462072802832,
+      "learning_rate": 4.637045969492164e-06,
+      "loss": 0.4617,
+      "step": 9176
+    },
+    {
+      "epoch": 3.592170906396641,
+      "grad_norm": 0.47896935915209116,
+      "learning_rate": 4.636964663772841e-06,
+      "loss": 0.4543,
+      "step": 9177
+    },
+    {
+      "epoch": 3.5925660656952334,
+      "grad_norm": 0.5590814683075651,
+      "learning_rate": 4.636883349660819e-06,
+      "loss": 0.4606,
+      "step": 9178
+    },
+    {
+      "epoch": 3.5929612249938256,
+      "grad_norm": 0.5042016745245929,
+      "learning_rate": 4.6368020271564166e-06,
+      "loss": 0.4789,
+      "step": 9179
+    },
+    {
+      "epoch": 3.593356384292418,
+      "grad_norm": 0.47545093889683404,
+      "learning_rate": 4.636720696259954e-06,
+      "loss": 0.4471,
+      "step": 9180
+    },
+    {
+      "epoch": 3.59375154359101,
+      "grad_norm": 0.4886174481015884,
+      "learning_rate": 4.636639356971749e-06,
+      "loss": 0.4501,
+      "step": 9181
+    },
+    {
+      "epoch": 3.5941467028896024,
+      "grad_norm": 0.4909615597451317,
+      "learning_rate": 4.6365580092921224e-06,
+      "loss": 0.4536,
+      "step": 9182
+    },
+    {
+      "epoch": 3.5945418621881946,
+      "grad_norm": 0.5011647702759715,
+      "learning_rate": 4.6364766532213936e-06,
+      "loss": 0.4594,
+      "step": 9183
+    },
+    {
+      "epoch": 3.594937021486787,
+      "grad_norm": 0.4891039483973173,
+      "learning_rate": 4.636395288759881e-06,
+      "loss": 0.4428,
+      "step": 9184
+    },
+    {
+      "epoch": 3.595332180785379,
+      "grad_norm": 0.49238304772818725,
+      "learning_rate": 4.6363139159079056e-06,
+      "loss": 0.4598,
+      "step": 9185
+    },
+    {
+      "epoch": 3.5957273400839713,
+      "grad_norm": 0.499060773425193,
+      "learning_rate": 4.636232534665787e-06,
+      "loss": 0.4816,
+      "step": 9186
+    },
+    {
+      "epoch": 3.5961224993825636,
+      "grad_norm": 0.4876846908708361,
+      "learning_rate": 4.636151145033844e-06,
+      "loss": 0.4695,
+      "step": 9187
+    },
+    {
+      "epoch": 3.596517658681156,
+      "grad_norm": 0.5010421129141813,
+      "learning_rate": 4.636069747012395e-06,
+      "loss": 0.4674,
+      "step": 9188
+    },
+    {
+      "epoch": 3.596912817979748,
+      "grad_norm": 0.4742596611960192,
+      "learning_rate": 4.6359883406017625e-06,
+      "loss": 0.4597,
+      "step": 9189
+    },
+    {
+      "epoch": 3.5973079772783403,
+      "grad_norm": 0.5045304178288424,
+      "learning_rate": 4.635906925802264e-06,
+      "loss": 0.4566,
+      "step": 9190
+    },
+    {
+      "epoch": 3.5977031365769325,
+      "grad_norm": 0.49285646311806497,
+      "learning_rate": 4.635825502614221e-06,
+      "loss": 0.4506,
+      "step": 9191
+    },
+    {
+      "epoch": 3.5980982958755248,
+      "grad_norm": 1.3189611198454247,
+      "learning_rate": 4.635744071037952e-06,
+      "loss": 0.4826,
+      "step": 9192
+    },
+    {
+      "epoch": 3.598493455174117,
+      "grad_norm": 0.4978191275680937,
+      "learning_rate": 4.6356626310737774e-06,
+      "loss": 0.4644,
+      "step": 9193
+    },
+    {
+      "epoch": 3.5988886144727092,
+      "grad_norm": 0.48939510446702533,
+      "learning_rate": 4.635581182722017e-06,
+      "loss": 0.4695,
+      "step": 9194
+    },
+    {
+      "epoch": 3.5992837737713015,
+      "grad_norm": 0.5054530113910645,
+      "learning_rate": 4.635499725982989e-06,
+      "loss": 0.4633,
+      "step": 9195
+    },
+    {
+      "epoch": 3.5996789330698937,
+      "grad_norm": 0.48017657962523336,
+      "learning_rate": 4.6354182608570155e-06,
+      "loss": 0.4469,
+      "step": 9196
+    },
+    {
+      "epoch": 3.600074092368486,
+      "grad_norm": 0.5007634832293895,
+      "learning_rate": 4.635336787344416e-06,
+      "loss": 0.4594,
+      "step": 9197
+    },
+    {
+      "epoch": 3.600469251667078,
+      "grad_norm": 0.5032668243356092,
+      "learning_rate": 4.635255305445511e-06,
+      "loss": 0.4653,
+      "step": 9198
+    },
+    {
+      "epoch": 3.6008644109656704,
+      "grad_norm": 0.49875765737609773,
+      "learning_rate": 4.635173815160619e-06,
+      "loss": 0.4605,
+      "step": 9199
+    },
+    {
+      "epoch": 3.6012595702642627,
+      "grad_norm": 0.5062930640900483,
+      "learning_rate": 4.635092316490061e-06,
+      "loss": 0.4514,
+      "step": 9200
+    },
+    {
+      "epoch": 3.601654729562855,
+      "grad_norm": 0.4974728918315835,
+      "learning_rate": 4.635010809434157e-06,
+      "loss": 0.4514,
+      "step": 9201
+    },
+    {
+      "epoch": 3.602049888861447,
+      "grad_norm": 0.49648655888984,
+      "learning_rate": 4.634929293993226e-06,
+      "loss": 0.4744,
+      "step": 9202
+    },
+    {
+      "epoch": 3.6024450481600394,
+      "grad_norm": 0.4824566489860067,
+      "learning_rate": 4.634847770167591e-06,
+      "loss": 0.4635,
+      "step": 9203
+    },
+    {
+      "epoch": 3.6028402074586316,
+      "grad_norm": 0.4979705347753646,
+      "learning_rate": 4.6347662379575685e-06,
+      "loss": 0.451,
+      "step": 9204
+    },
+    {
+      "epoch": 3.603235366757224,
+      "grad_norm": 0.5062537574749927,
+      "learning_rate": 4.634684697363482e-06,
+      "loss": 0.4744,
+      "step": 9205
+    },
+    {
+      "epoch": 3.603630526055816,
+      "grad_norm": 0.5738607455547821,
+      "learning_rate": 4.634603148385649e-06,
+      "loss": 0.4592,
+      "step": 9206
+    },
+    {
+      "epoch": 3.6040256853544084,
+      "grad_norm": 1.189807804448986,
+      "learning_rate": 4.6345215910243915e-06,
+      "loss": 0.4479,
+      "step": 9207
+    },
+    {
+      "epoch": 3.6044208446530006,
+      "grad_norm": 0.4879581966562651,
+      "learning_rate": 4.634440025280029e-06,
+      "loss": 0.4646,
+      "step": 9208
+    },
+    {
+      "epoch": 3.604816003951593,
+      "grad_norm": 0.5497629591149095,
+      "learning_rate": 4.634358451152883e-06,
+      "loss": 0.4546,
+      "step": 9209
+    },
+    {
+      "epoch": 3.605211163250185,
+      "grad_norm": 0.5194558268040678,
+      "learning_rate": 4.634276868643273e-06,
+      "loss": 0.4611,
+      "step": 9210
+    },
+    {
+      "epoch": 3.6056063225487773,
+      "grad_norm": 0.4845956782876475,
+      "learning_rate": 4.634195277751518e-06,
+      "loss": 0.4445,
+      "step": 9211
+    },
+    {
+      "epoch": 3.6060014818473696,
+      "grad_norm": 0.4940635660411414,
+      "learning_rate": 4.634113678477942e-06,
+      "loss": 0.4524,
+      "step": 9212
+    },
+    {
+      "epoch": 3.606396641145962,
+      "grad_norm": 0.5070909576956948,
+      "learning_rate": 4.634032070822862e-06,
+      "loss": 0.4833,
+      "step": 9213
+    },
+    {
+      "epoch": 3.606791800444554,
+      "grad_norm": 0.490670333205641,
+      "learning_rate": 4.633950454786601e-06,
+      "loss": 0.4624,
+      "step": 9214
+    },
+    {
+      "epoch": 3.6071869597431463,
+      "grad_norm": 0.4835269632726879,
+      "learning_rate": 4.633868830369477e-06,
+      "loss": 0.4555,
+      "step": 9215
+    },
+    {
+      "epoch": 3.607582119041739,
+      "grad_norm": 0.4992739285137002,
+      "learning_rate": 4.633787197571813e-06,
+      "loss": 0.4769,
+      "step": 9216
+    },
+    {
+      "epoch": 3.607977278340331,
+      "grad_norm": 0.5402074831328878,
+      "learning_rate": 4.633705556393928e-06,
+      "loss": 0.4751,
+      "step": 9217
+    },
+    {
+      "epoch": 3.6083724376389235,
+      "grad_norm": 0.48961767171363246,
+      "learning_rate": 4.633623906836144e-06,
+      "loss": 0.4565,
+      "step": 9218
+    },
+    {
+      "epoch": 3.6087675969375157,
+      "grad_norm": 0.4838103684730782,
+      "learning_rate": 4.63354224889878e-06,
+      "loss": 0.4506,
+      "step": 9219
+    },
+    {
+      "epoch": 3.609162756236108,
+      "grad_norm": 0.4925277287138113,
+      "learning_rate": 4.633460582582157e-06,
+      "loss": 0.4488,
+      "step": 9220
+    },
+    {
+      "epoch": 3.6095579155347,
+      "grad_norm": 0.4825571381803209,
+      "learning_rate": 4.633378907886597e-06,
+      "loss": 0.4607,
+      "step": 9221
+    },
+    {
+      "epoch": 3.6099530748332924,
+      "grad_norm": 0.48936877599995243,
+      "learning_rate": 4.633297224812422e-06,
+      "loss": 0.4578,
+      "step": 9222
+    },
+    {
+      "epoch": 3.6103482341318847,
+      "grad_norm": 0.4905493219692705,
+      "learning_rate": 4.633215533359949e-06,
+      "loss": 0.4848,
+      "step": 9223
+    },
+    {
+      "epoch": 3.610743393430477,
+      "grad_norm": 0.5041306176026158,
+      "learning_rate": 4.633133833529501e-06,
+      "loss": 0.4625,
+      "step": 9224
+    },
+    {
+      "epoch": 3.611138552729069,
+      "grad_norm": 0.47807489803501046,
+      "learning_rate": 4.633052125321399e-06,
+      "loss": 0.4579,
+      "step": 9225
+    },
+    {
+      "epoch": 3.6115337120276614,
+      "grad_norm": 0.4905504604932367,
+      "learning_rate": 4.632970408735963e-06,
+      "loss": 0.4677,
+      "step": 9226
+    },
+    {
+      "epoch": 3.6119288713262536,
+      "grad_norm": 0.48751702506148886,
+      "learning_rate": 4.632888683773515e-06,
+      "loss": 0.4787,
+      "step": 9227
+    },
+    {
+      "epoch": 3.612324030624846,
+      "grad_norm": 0.497021089178736,
+      "learning_rate": 4.6328069504343745e-06,
+      "loss": 0.4421,
+      "step": 9228
+    },
+    {
+      "epoch": 3.612719189923438,
+      "grad_norm": 0.5243484285892979,
+      "learning_rate": 4.632725208718864e-06,
+      "loss": 0.451,
+      "step": 9229
+    },
+    {
+      "epoch": 3.6131143492220303,
+      "grad_norm": 0.49379034366259994,
+      "learning_rate": 4.6326434586273035e-06,
+      "loss": 0.4446,
+      "step": 9230
+    },
+    {
+      "epoch": 3.6135095085206226,
+      "grad_norm": 0.5142924559989173,
+      "learning_rate": 4.632561700160015e-06,
+      "loss": 0.4488,
+      "step": 9231
+    },
+    {
+      "epoch": 3.613904667819215,
+      "grad_norm": 0.504763239789171,
+      "learning_rate": 4.632479933317319e-06,
+      "loss": 0.4603,
+      "step": 9232
+    },
+    {
+      "epoch": 3.614299827117807,
+      "grad_norm": 0.503230561587256,
+      "learning_rate": 4.632398158099537e-06,
+      "loss": 0.4544,
+      "step": 9233
+    },
+    {
+      "epoch": 3.6146949864163993,
+      "grad_norm": 0.490744168584867,
+      "learning_rate": 4.63231637450699e-06,
+      "loss": 0.4409,
+      "step": 9234
+    },
+    {
+      "epoch": 3.6150901457149915,
+      "grad_norm": 0.5095413791531018,
+      "learning_rate": 4.6322345825399985e-06,
+      "loss": 0.4655,
+      "step": 9235
+    },
+    {
+      "epoch": 3.615485305013584,
+      "grad_norm": 0.4937333674079909,
+      "learning_rate": 4.6321527821988845e-06,
+      "loss": 0.4571,
+      "step": 9236
+    },
+    {
+      "epoch": 3.615880464312176,
+      "grad_norm": 0.4898084499608632,
+      "learning_rate": 4.632070973483969e-06,
+      "loss": 0.4839,
+      "step": 9237
+    },
+    {
+      "epoch": 3.6162756236107683,
+      "grad_norm": 0.5007373601317516,
+      "learning_rate": 4.631989156395574e-06,
+      "loss": 0.4541,
+      "step": 9238
+    },
+    {
+      "epoch": 3.6166707829093605,
+      "grad_norm": 0.48723321374477363,
+      "learning_rate": 4.631907330934019e-06,
+      "loss": 0.48,
+      "step": 9239
+    },
+    {
+      "epoch": 3.6170659422079527,
+      "grad_norm": 0.48448261567750023,
+      "learning_rate": 4.631825497099627e-06,
+      "loss": 0.4543,
+      "step": 9240
+    },
+    {
+      "epoch": 3.617461101506545,
+      "grad_norm": 0.5316625962607656,
+      "learning_rate": 4.63174365489272e-06,
+      "loss": 0.4764,
+      "step": 9241
+    },
+    {
+      "epoch": 3.6178562608051372,
+      "grad_norm": 0.49953626646231936,
+      "learning_rate": 4.6316618043136165e-06,
+      "loss": 0.4518,
+      "step": 9242
+    },
+    {
+      "epoch": 3.6182514201037295,
+      "grad_norm": 0.5003965600017454,
+      "learning_rate": 4.631579945362641e-06,
+      "loss": 0.4579,
+      "step": 9243
+    },
+    {
+      "epoch": 3.6186465794023217,
+      "grad_norm": 0.48092823514138977,
+      "learning_rate": 4.631498078040114e-06,
+      "loss": 0.448,
+      "step": 9244
+    },
+    {
+      "epoch": 3.619041738700914,
+      "grad_norm": 0.49710416001614555,
+      "learning_rate": 4.631416202346357e-06,
+      "loss": 0.4595,
+      "step": 9245
+    },
+    {
+      "epoch": 3.619436897999506,
+      "grad_norm": 0.49446651343934783,
+      "learning_rate": 4.631334318281691e-06,
+      "loss": 0.4602,
+      "step": 9246
+    },
+    {
+      "epoch": 3.6198320572980984,
+      "grad_norm": 0.5009277044883268,
+      "learning_rate": 4.631252425846439e-06,
+      "loss": 0.4408,
+      "step": 9247
+    },
+    {
+      "epoch": 3.6202272165966907,
+      "grad_norm": 0.4948355571473325,
+      "learning_rate": 4.63117052504092e-06,
+      "loss": 0.4506,
+      "step": 9248
+    },
+    {
+      "epoch": 3.620622375895283,
+      "grad_norm": 0.48896770764585235,
+      "learning_rate": 4.631088615865458e-06,
+      "loss": 0.455,
+      "step": 9249
+    },
+    {
+      "epoch": 3.621017535193875,
+      "grad_norm": 0.48712330415560645,
+      "learning_rate": 4.631006698320374e-06,
+      "loss": 0.4579,
+      "step": 9250
+    },
+    {
+      "epoch": 3.6214126944924674,
+      "grad_norm": 0.5013178593300339,
+      "learning_rate": 4.630924772405989e-06,
+      "loss": 0.4721,
+      "step": 9251
+    },
+    {
+      "epoch": 3.6218078537910596,
+      "grad_norm": 0.691080349270438,
+      "learning_rate": 4.630842838122627e-06,
+      "loss": 0.4728,
+      "step": 9252
+    },
+    {
+      "epoch": 3.622203013089652,
+      "grad_norm": 0.4974090024974102,
+      "learning_rate": 4.630760895470607e-06,
+      "loss": 0.4682,
+      "step": 9253
+    },
+    {
+      "epoch": 3.622598172388244,
+      "grad_norm": 0.4738706863887212,
+      "learning_rate": 4.630678944450253e-06,
+      "loss": 0.43,
+      "step": 9254
+    },
+    {
+      "epoch": 3.6229933316868363,
+      "grad_norm": 0.4996067080433458,
+      "learning_rate": 4.630596985061886e-06,
+      "loss": 0.4625,
+      "step": 9255
+    },
+    {
+      "epoch": 3.6233884909854286,
+      "grad_norm": 0.49310669745464825,
+      "learning_rate": 4.630515017305827e-06,
+      "loss": 0.4723,
+      "step": 9256
+    },
+    {
+      "epoch": 3.623783650284021,
+      "grad_norm": 0.5093600935457048,
+      "learning_rate": 4.630433041182398e-06,
+      "loss": 0.4711,
+      "step": 9257
+    },
+    {
+      "epoch": 3.624178809582613,
+      "grad_norm": 0.48115236735295236,
+      "learning_rate": 4.630351056691923e-06,
+      "loss": 0.4537,
+      "step": 9258
+    },
+    {
+      "epoch": 3.6245739688812053,
+      "grad_norm": 0.5011382943630356,
+      "learning_rate": 4.630269063834723e-06,
+      "loss": 0.458,
+      "step": 9259
+    },
+    {
+      "epoch": 3.6249691281797976,
+      "grad_norm": 0.5079718914818618,
+      "learning_rate": 4.630187062611119e-06,
+      "loss": 0.4779,
+      "step": 9260
+    },
+    {
+      "epoch": 3.62536428747839,
+      "grad_norm": 0.4980405547820634,
+      "learning_rate": 4.630105053021433e-06,
+      "loss": 0.4665,
+      "step": 9261
+    },
+    {
+      "epoch": 3.625759446776982,
+      "grad_norm": 0.4893026369877525,
+      "learning_rate": 4.6300230350659885e-06,
+      "loss": 0.4648,
+      "step": 9262
+    },
+    {
+      "epoch": 3.6261546060755743,
+      "grad_norm": 0.5555460867365863,
+      "learning_rate": 4.629941008745108e-06,
+      "loss": 0.4647,
+      "step": 9263
+    },
+    {
+      "epoch": 3.6265497653741665,
+      "grad_norm": 0.5759949696380431,
+      "learning_rate": 4.629858974059111e-06,
+      "loss": 0.4673,
+      "step": 9264
+    },
+    {
+      "epoch": 3.6269449246727588,
+      "grad_norm": 0.48052405561183786,
+      "learning_rate": 4.629776931008322e-06,
+      "loss": 0.458,
+      "step": 9265
+    },
+    {
+      "epoch": 3.627340083971351,
+      "grad_norm": 0.5010331429705553,
+      "learning_rate": 4.629694879593062e-06,
+      "loss": 0.4754,
+      "step": 9266
+    },
+    {
+      "epoch": 3.6277352432699432,
+      "grad_norm": 0.5007942666179961,
+      "learning_rate": 4.6296128198136545e-06,
+      "loss": 0.4696,
+      "step": 9267
+    },
+    {
+      "epoch": 3.6281304025685355,
+      "grad_norm": 0.5016428896270185,
+      "learning_rate": 4.62953075167042e-06,
+      "loss": 0.4672,
+      "step": 9268
+    },
+    {
+      "epoch": 3.6285255618671277,
+      "grad_norm": 0.49272882028841225,
+      "learning_rate": 4.629448675163682e-06,
+      "loss": 0.4639,
+      "step": 9269
+    },
+    {
+      "epoch": 3.62892072116572,
+      "grad_norm": 0.48986056236330106,
+      "learning_rate": 4.629366590293763e-06,
+      "loss": 0.4695,
+      "step": 9270
+    },
+    {
+      "epoch": 3.629315880464312,
+      "grad_norm": 0.4962133184327803,
+      "learning_rate": 4.629284497060985e-06,
+      "loss": 0.4502,
+      "step": 9271
+    },
+    {
+      "epoch": 3.6297110397629044,
+      "grad_norm": 0.4898941204572704,
+      "learning_rate": 4.629202395465672e-06,
+      "loss": 0.4609,
+      "step": 9272
+    },
+    {
+      "epoch": 3.6301061990614967,
+      "grad_norm": 0.49506032850463966,
+      "learning_rate": 4.629120285508143e-06,
+      "loss": 0.4482,
+      "step": 9273
+    },
+    {
+      "epoch": 3.630501358360089,
+      "grad_norm": 0.5171758595413785,
+      "learning_rate": 4.629038167188723e-06,
+      "loss": 0.4501,
+      "step": 9274
+    },
+    {
+      "epoch": 3.630896517658681,
+      "grad_norm": 0.4967738395377754,
+      "learning_rate": 4.628956040507734e-06,
+      "loss": 0.4649,
+      "step": 9275
+    },
+    {
+      "epoch": 3.6312916769572734,
+      "grad_norm": 0.5063747871234539,
+      "learning_rate": 4.628873905465498e-06,
+      "loss": 0.4629,
+      "step": 9276
+    },
+    {
+      "epoch": 3.6316868362558656,
+      "grad_norm": 0.47510127574337424,
+      "learning_rate": 4.628791762062338e-06,
+      "loss": 0.4449,
+      "step": 9277
+    },
+    {
+      "epoch": 3.632081995554458,
+      "grad_norm": 0.49070050113003616,
+      "learning_rate": 4.628709610298578e-06,
+      "loss": 0.4599,
+      "step": 9278
+    },
+    {
+      "epoch": 3.63247715485305,
+      "grad_norm": 0.513370408704164,
+      "learning_rate": 4.628627450174537e-06,
+      "loss": 0.463,
+      "step": 9279
+    },
+    {
+      "epoch": 3.6328723141516424,
+      "grad_norm": 0.5091150770823012,
+      "learning_rate": 4.628545281690541e-06,
+      "loss": 0.4746,
+      "step": 9280
+    },
+    {
+      "epoch": 3.6332674734502346,
+      "grad_norm": 0.5151701240747677,
+      "learning_rate": 4.628463104846912e-06,
+      "loss": 0.4705,
+      "step": 9281
+    },
+    {
+      "epoch": 3.633662632748827,
+      "grad_norm": 0.49056073566435143,
+      "learning_rate": 4.628380919643972e-06,
+      "loss": 0.4638,
+      "step": 9282
+    },
+    {
+      "epoch": 3.634057792047419,
+      "grad_norm": 0.5144421824248759,
+      "learning_rate": 4.6282987260820445e-06,
+      "loss": 0.4778,
+      "step": 9283
+    },
+    {
+      "epoch": 3.6344529513460113,
+      "grad_norm": 0.49744475854641185,
+      "learning_rate": 4.6282165241614515e-06,
+      "loss": 0.458,
+      "step": 9284
+    },
+    {
+      "epoch": 3.6348481106446036,
+      "grad_norm": 0.5035190140052788,
+      "learning_rate": 4.628134313882518e-06,
+      "loss": 0.4693,
+      "step": 9285
+    },
+    {
+      "epoch": 3.635243269943196,
+      "grad_norm": 0.5615476058939156,
+      "learning_rate": 4.6280520952455635e-06,
+      "loss": 0.4932,
+      "step": 9286
+    },
+    {
+      "epoch": 3.635638429241788,
+      "grad_norm": 0.49189317167932184,
+      "learning_rate": 4.627969868250912e-06,
+      "loss": 0.4635,
+      "step": 9287
+    },
+    {
+      "epoch": 3.6360335885403803,
+      "grad_norm": 0.4912590147231549,
+      "learning_rate": 4.6278876328988885e-06,
+      "loss": 0.4639,
+      "step": 9288
+    },
+    {
+      "epoch": 3.6364287478389725,
+      "grad_norm": 2.521359709853842,
+      "learning_rate": 4.627805389189814e-06,
+      "loss": 0.4611,
+      "step": 9289
+    },
+    {
+      "epoch": 3.6368239071375648,
+      "grad_norm": 0.486064042323978,
+      "learning_rate": 4.627723137124012e-06,
+      "loss": 0.4412,
+      "step": 9290
+    },
+    {
+      "epoch": 3.637219066436157,
+      "grad_norm": 0.5048179415790461,
+      "learning_rate": 4.627640876701806e-06,
+      "loss": 0.4571,
+      "step": 9291
+    },
+    {
+      "epoch": 3.6376142257347492,
+      "grad_norm": 0.4959460801045448,
+      "learning_rate": 4.627558607923517e-06,
+      "loss": 0.4677,
+      "step": 9292
+    },
+    {
+      "epoch": 3.6380093850333415,
+      "grad_norm": 0.48066916313414726,
+      "learning_rate": 4.627476330789471e-06,
+      "loss": 0.4447,
+      "step": 9293
+    },
+    {
+      "epoch": 3.6384045443319337,
+      "grad_norm": 0.4973239424582944,
+      "learning_rate": 4.62739404529999e-06,
+      "loss": 0.4672,
+      "step": 9294
+    },
+    {
+      "epoch": 3.638799703630526,
+      "grad_norm": 0.48199497014952,
+      "learning_rate": 4.627311751455397e-06,
+      "loss": 0.4517,
+      "step": 9295
+    },
+    {
+      "epoch": 3.639194862929118,
+      "grad_norm": 0.49279718443129383,
+      "learning_rate": 4.627229449256014e-06,
+      "loss": 0.4693,
+      "step": 9296
+    },
+    {
+      "epoch": 3.6395900222277104,
+      "grad_norm": 0.5010681919023393,
+      "learning_rate": 4.627147138702166e-06,
+      "loss": 0.4743,
+      "step": 9297
+    },
+    {
+      "epoch": 3.6399851815263027,
+      "grad_norm": 0.47695376136618906,
+      "learning_rate": 4.627064819794177e-06,
+      "loss": 0.454,
+      "step": 9298
+    },
+    {
+      "epoch": 3.640380340824895,
+      "grad_norm": 0.4888943085386212,
+      "learning_rate": 4.626982492532368e-06,
+      "loss": 0.4456,
+      "step": 9299
+    },
+    {
+      "epoch": 3.640775500123487,
+      "grad_norm": 0.4771083873842363,
+      "learning_rate": 4.626900156917064e-06,
+      "loss": 0.468,
+      "step": 9300
+    },
+    {
+      "epoch": 3.6411706594220794,
+      "grad_norm": 0.48136951242486825,
+      "learning_rate": 4.626817812948586e-06,
+      "loss": 0.4516,
+      "step": 9301
+    },
+    {
+      "epoch": 3.6415658187206716,
+      "grad_norm": 0.4979788453537473,
+      "learning_rate": 4.6267354606272605e-06,
+      "loss": 0.4723,
+      "step": 9302
+    },
+    {
+      "epoch": 3.641960978019264,
+      "grad_norm": 0.5102309614105011,
+      "learning_rate": 4.62665309995341e-06,
+      "loss": 0.4598,
+      "step": 9303
+    },
+    {
+      "epoch": 3.642356137317856,
+      "grad_norm": 0.49649828248075617,
+      "learning_rate": 4.6265707309273565e-06,
+      "loss": 0.4686,
+      "step": 9304
+    },
+    {
+      "epoch": 3.6427512966164484,
+      "grad_norm": 0.48125108712989917,
+      "learning_rate": 4.626488353549425e-06,
+      "loss": 0.4482,
+      "step": 9305
+    },
+    {
+      "epoch": 3.6431464559150406,
+      "grad_norm": 0.5210918594356808,
+      "learning_rate": 4.626405967819938e-06,
+      "loss": 0.459,
+      "step": 9306
+    },
+    {
+      "epoch": 3.643541615213633,
+      "grad_norm": 0.47329837989409934,
+      "learning_rate": 4.626323573739219e-06,
+      "loss": 0.4465,
+      "step": 9307
+    },
+    {
+      "epoch": 3.643936774512225,
+      "grad_norm": 0.5042256332277069,
+      "learning_rate": 4.626241171307593e-06,
+      "loss": 0.4573,
+      "step": 9308
+    },
+    {
+      "epoch": 3.6443319338108173,
+      "grad_norm": 1.0400744964458153,
+      "learning_rate": 4.626158760525383e-06,
+      "loss": 0.4835,
+      "step": 9309
+    },
+    {
+      "epoch": 3.6447270931094096,
+      "grad_norm": 0.5001511613882816,
+      "learning_rate": 4.6260763413929124e-06,
+      "loss": 0.4533,
+      "step": 9310
+    },
+    {
+      "epoch": 3.645122252408002,
+      "grad_norm": 0.4940591531345029,
+      "learning_rate": 4.625993913910505e-06,
+      "loss": 0.4646,
+      "step": 9311
+    },
+    {
+      "epoch": 3.645517411706594,
+      "grad_norm": 0.4942019788158831,
+      "learning_rate": 4.625911478078484e-06,
+      "loss": 0.4551,
+      "step": 9312
+    },
+    {
+      "epoch": 3.6459125710051863,
+      "grad_norm": 0.4884265738310717,
+      "learning_rate": 4.6258290338971735e-06,
+      "loss": 0.4517,
+      "step": 9313
+    },
+    {
+      "epoch": 3.6463077303037785,
+      "grad_norm": 0.4908742478353482,
+      "learning_rate": 4.625746581366898e-06,
+      "loss": 0.4711,
+      "step": 9314
+    },
+    {
+      "epoch": 3.6467028896023708,
+      "grad_norm": 0.5066441924253818,
+      "learning_rate": 4.625664120487981e-06,
+      "loss": 0.4672,
+      "step": 9315
+    },
+    {
+      "epoch": 3.647098048900963,
+      "grad_norm": 0.4975632434645502,
+      "learning_rate": 4.625581651260745e-06,
+      "loss": 0.4463,
+      "step": 9316
+    },
+    {
+      "epoch": 3.6474932081995552,
+      "grad_norm": 0.5054480921919917,
+      "learning_rate": 4.625499173685516e-06,
+      "loss": 0.4676,
+      "step": 9317
+    },
+    {
+      "epoch": 3.6478883674981475,
+      "grad_norm": 0.4881775251448275,
+      "learning_rate": 4.6254166877626175e-06,
+      "loss": 0.4441,
+      "step": 9318
+    },
+    {
+      "epoch": 3.6482835267967397,
+      "grad_norm": 0.48222427669341467,
+      "learning_rate": 4.625334193492371e-06,
+      "loss": 0.4665,
+      "step": 9319
+    },
+    {
+      "epoch": 3.648678686095332,
+      "grad_norm": 0.4936417971367663,
+      "learning_rate": 4.625251690875104e-06,
+      "loss": 0.4888,
+      "step": 9320
+    },
+    {
+      "epoch": 3.649073845393924,
+      "grad_norm": 0.49953652158686007,
+      "learning_rate": 4.6251691799111376e-06,
+      "loss": 0.4675,
+      "step": 9321
+    },
+    {
+      "epoch": 3.6494690046925164,
+      "grad_norm": 0.49771002562217137,
+      "learning_rate": 4.625086660600798e-06,
+      "loss": 0.468,
+      "step": 9322
+    },
+    {
+      "epoch": 3.6498641639911087,
+      "grad_norm": 0.49436016460202736,
+      "learning_rate": 4.625004132944409e-06,
+      "loss": 0.4492,
+      "step": 9323
+    },
+    {
+      "epoch": 3.650259323289701,
+      "grad_norm": 0.46676714681432757,
+      "learning_rate": 4.624921596942292e-06,
+      "loss": 0.4241,
+      "step": 9324
+    },
+    {
+      "epoch": 3.650654482588293,
+      "grad_norm": 0.49467615491708383,
+      "learning_rate": 4.6248390525947755e-06,
+      "loss": 0.4691,
+      "step": 9325
+    },
+    {
+      "epoch": 3.6510496418868854,
+      "grad_norm": 0.4938231345250084,
+      "learning_rate": 4.624756499902181e-06,
+      "loss": 0.46,
+      "step": 9326
+    },
+    {
+      "epoch": 3.6514448011854777,
+      "grad_norm": 0.4838739325764053,
+      "learning_rate": 4.624673938864832e-06,
+      "loss": 0.4513,
+      "step": 9327
+    },
+    {
+      "epoch": 3.65183996048407,
+      "grad_norm": 0.49112971366065933,
+      "learning_rate": 4.6245913694830545e-06,
+      "loss": 0.4524,
+      "step": 9328
+    },
+    {
+      "epoch": 3.652235119782662,
+      "grad_norm": 0.5011715788855052,
+      "learning_rate": 4.624508791757173e-06,
+      "loss": 0.4856,
+      "step": 9329
+    },
+    {
+      "epoch": 3.6526302790812544,
+      "grad_norm": 0.4993605096743152,
+      "learning_rate": 4.62442620568751e-06,
+      "loss": 0.471,
+      "step": 9330
+    },
+    {
+      "epoch": 3.6530254383798466,
+      "grad_norm": 0.48611034727334546,
+      "learning_rate": 4.624343611274391e-06,
+      "loss": 0.4552,
+      "step": 9331
+    },
+    {
+      "epoch": 3.653420597678439,
+      "grad_norm": 0.5253566397582367,
+      "learning_rate": 4.624261008518141e-06,
+      "loss": 0.4566,
+      "step": 9332
+    },
+    {
+      "epoch": 3.653815756977031,
+      "grad_norm": 0.5175702806785849,
+      "learning_rate": 4.624178397419083e-06,
+      "loss": 0.4533,
+      "step": 9333
+    },
+    {
+      "epoch": 3.6542109162756238,
+      "grad_norm": 0.4994237852186553,
+      "learning_rate": 4.624095777977543e-06,
+      "loss": 0.447,
+      "step": 9334
+    },
+    {
+      "epoch": 3.654606075574216,
+      "grad_norm": 0.4940996164170102,
+      "learning_rate": 4.624013150193844e-06,
+      "loss": 0.4544,
+      "step": 9335
+    },
+    {
+      "epoch": 3.6550012348728083,
+      "grad_norm": 0.4931404991433969,
+      "learning_rate": 4.623930514068311e-06,
+      "loss": 0.4364,
+      "step": 9336
+    },
+    {
+      "epoch": 3.6553963941714005,
+      "grad_norm": 0.5060095886918009,
+      "learning_rate": 4.623847869601269e-06,
+      "loss": 0.4622,
+      "step": 9337
+    },
+    {
+      "epoch": 3.6557915534699927,
+      "grad_norm": 0.4864732607657538,
+      "learning_rate": 4.623765216793042e-06,
+      "loss": 0.4508,
+      "step": 9338
+    },
+    {
+      "epoch": 3.656186712768585,
+      "grad_norm": 0.515417673885724,
+      "learning_rate": 4.623682555643955e-06,
+      "loss": 0.4746,
+      "step": 9339
+    },
+    {
+      "epoch": 3.656581872067177,
+      "grad_norm": 0.5093440038900928,
+      "learning_rate": 4.623599886154333e-06,
+      "loss": 0.4586,
+      "step": 9340
+    },
+    {
+      "epoch": 3.6569770313657695,
+      "grad_norm": 0.4884238293387281,
+      "learning_rate": 4.623517208324499e-06,
+      "loss": 0.4611,
+      "step": 9341
+    },
+    {
+      "epoch": 3.6573721906643617,
+      "grad_norm": 0.4949252091120524,
+      "learning_rate": 4.623434522154779e-06,
+      "loss": 0.453,
+      "step": 9342
+    },
+    {
+      "epoch": 3.657767349962954,
+      "grad_norm": 0.4996973803416822,
+      "learning_rate": 4.623351827645498e-06,
+      "loss": 0.4615,
+      "step": 9343
+    },
+    {
+      "epoch": 3.658162509261546,
+      "grad_norm": 0.504144710893261,
+      "learning_rate": 4.623269124796981e-06,
+      "loss": 0.4673,
+      "step": 9344
+    },
+    {
+      "epoch": 3.6585576685601384,
+      "grad_norm": 0.4969157578415948,
+      "learning_rate": 4.623186413609552e-06,
+      "loss": 0.4487,
+      "step": 9345
+    },
+    {
+      "epoch": 3.6589528278587307,
+      "grad_norm": 0.4932886314806929,
+      "learning_rate": 4.623103694083535e-06,
+      "loss": 0.456,
+      "step": 9346
+    },
+    {
+      "epoch": 3.659347987157323,
+      "grad_norm": 0.5148738766834107,
+      "learning_rate": 4.623020966219257e-06,
+      "loss": 0.4471,
+      "step": 9347
+    },
+    {
+      "epoch": 3.659743146455915,
+      "grad_norm": 0.5830202209682454,
+      "learning_rate": 4.622938230017041e-06,
+      "loss": 0.4643,
+      "step": 9348
+    },
+    {
+      "epoch": 3.6601383057545074,
+      "grad_norm": 0.48992602819472203,
+      "learning_rate": 4.622855485477214e-06,
+      "loss": 0.4474,
+      "step": 9349
+    },
+    {
+      "epoch": 3.6605334650530996,
+      "grad_norm": 0.48284568742478173,
+      "learning_rate": 4.622772732600098e-06,
+      "loss": 0.4623,
+      "step": 9350
+    },
+    {
+      "epoch": 3.660928624351692,
+      "grad_norm": 0.48801605211010135,
+      "learning_rate": 4.622689971386021e-06,
+      "loss": 0.4426,
+      "step": 9351
+    },
+    {
+      "epoch": 3.661323783650284,
+      "grad_norm": 0.49203131886308554,
+      "learning_rate": 4.6226072018353055e-06,
+      "loss": 0.4389,
+      "step": 9352
+    },
+    {
+      "epoch": 3.6617189429488763,
+      "grad_norm": 0.48655905238073444,
+      "learning_rate": 4.622524423948279e-06,
+      "loss": 0.4602,
+      "step": 9353
+    },
+    {
+      "epoch": 3.6621141022474686,
+      "grad_norm": 0.4934300579478431,
+      "learning_rate": 4.6224416377252645e-06,
+      "loss": 0.4439,
+      "step": 9354
+    },
+    {
+      "epoch": 3.662509261546061,
+      "grad_norm": 0.4858789552341792,
+      "learning_rate": 4.622358843166589e-06,
+      "loss": 0.4555,
+      "step": 9355
+    },
+    {
+      "epoch": 3.662904420844653,
+      "grad_norm": 0.4899213923525825,
+      "learning_rate": 4.622276040272576e-06,
+      "loss": 0.4529,
+      "step": 9356
+    },
+    {
+      "epoch": 3.6632995801432453,
+      "grad_norm": 0.494097810082973,
+      "learning_rate": 4.622193229043552e-06,
+      "loss": 0.4511,
+      "step": 9357
+    },
+    {
+      "epoch": 3.6636947394418375,
+      "grad_norm": 0.4844491051320266,
+      "learning_rate": 4.622110409479842e-06,
+      "loss": 0.4587,
+      "step": 9358
+    },
+    {
+      "epoch": 3.66408989874043,
+      "grad_norm": 0.4914228114049162,
+      "learning_rate": 4.622027581581771e-06,
+      "loss": 0.4487,
+      "step": 9359
+    },
+    {
+      "epoch": 3.664485058039022,
+      "grad_norm": 0.49201477673440297,
+      "learning_rate": 4.6219447453496626e-06,
+      "loss": 0.4654,
+      "step": 9360
+    },
+    {
+      "epoch": 3.6648802173376143,
+      "grad_norm": 0.5030420734743306,
+      "learning_rate": 4.621861900783845e-06,
+      "loss": 0.4556,
+      "step": 9361
+    },
+    {
+      "epoch": 3.6652753766362065,
+      "grad_norm": 0.49226581282451,
+      "learning_rate": 4.621779047884642e-06,
+      "loss": 0.4662,
+      "step": 9362
+    },
+    {
+      "epoch": 3.6656705359347987,
+      "grad_norm": 0.485604702744728,
+      "learning_rate": 4.621696186652379e-06,
+      "loss": 0.4651,
+      "step": 9363
+    },
+    {
+      "epoch": 3.666065695233391,
+      "grad_norm": 0.4947225613132439,
+      "learning_rate": 4.621613317087382e-06,
+      "loss": 0.45,
+      "step": 9364
+    },
+    {
+      "epoch": 3.6664608545319832,
+      "grad_norm": 0.49964170799077484,
+      "learning_rate": 4.6215304391899765e-06,
+      "loss": 0.4368,
+      "step": 9365
+    },
+    {
+      "epoch": 3.6668560138305755,
+      "grad_norm": 0.4640684797492653,
+      "learning_rate": 4.621447552960488e-06,
+      "loss": 0.4426,
+      "step": 9366
+    },
+    {
+      "epoch": 3.6672511731291677,
+      "grad_norm": 0.493726848514641,
+      "learning_rate": 4.621364658399241e-06,
+      "loss": 0.4724,
+      "step": 9367
+    },
+    {
+      "epoch": 3.66764633242776,
+      "grad_norm": 0.48809324691574657,
+      "learning_rate": 4.621281755506562e-06,
+      "loss": 0.4362,
+      "step": 9368
+    },
+    {
+      "epoch": 3.668041491726352,
+      "grad_norm": 0.4939624266923956,
+      "learning_rate": 4.621198844282777e-06,
+      "loss": 0.4726,
+      "step": 9369
+    },
+    {
+      "epoch": 3.6684366510249444,
+      "grad_norm": 0.49641926622094995,
+      "learning_rate": 4.62111592472821e-06,
+      "loss": 0.4717,
+      "step": 9370
+    },
+    {
+      "epoch": 3.6688318103235367,
+      "grad_norm": 0.5050735506985684,
+      "learning_rate": 4.6210329968431876e-06,
+      "loss": 0.4674,
+      "step": 9371
+    },
+    {
+      "epoch": 3.669226969622129,
+      "grad_norm": 0.4909447418334573,
+      "learning_rate": 4.620950060628037e-06,
+      "loss": 0.451,
+      "step": 9372
+    },
+    {
+      "epoch": 3.669622128920721,
+      "grad_norm": 0.48904389952363275,
+      "learning_rate": 4.620867116083081e-06,
+      "loss": 0.4497,
+      "step": 9373
+    },
+    {
+      "epoch": 3.6700172882193134,
+      "grad_norm": 0.5054031870535084,
+      "learning_rate": 4.620784163208647e-06,
+      "loss": 0.4945,
+      "step": 9374
+    },
+    {
+      "epoch": 3.6704124475179056,
+      "grad_norm": 0.5433428790690614,
+      "learning_rate": 4.6207012020050614e-06,
+      "loss": 0.4533,
+      "step": 9375
+    },
+    {
+      "epoch": 3.670807606816498,
+      "grad_norm": 0.5002618872303594,
+      "learning_rate": 4.620618232472649e-06,
+      "loss": 0.4489,
+      "step": 9376
+    },
+    {
+      "epoch": 3.67120276611509,
+      "grad_norm": 0.48690513054549206,
+      "learning_rate": 4.620535254611735e-06,
+      "loss": 0.4678,
+      "step": 9377
+    },
+    {
+      "epoch": 3.6715979254136824,
+      "grad_norm": 0.48402593380651104,
+      "learning_rate": 4.6204522684226475e-06,
+      "loss": 0.4506,
+      "step": 9378
+    },
+    {
+      "epoch": 3.6719930847122746,
+      "grad_norm": 0.5103203925733566,
+      "learning_rate": 4.620369273905711e-06,
+      "loss": 0.4528,
+      "step": 9379
+    },
+    {
+      "epoch": 3.672388244010867,
+      "grad_norm": 0.48564066203388767,
+      "learning_rate": 4.620286271061251e-06,
+      "loss": 0.4622,
+      "step": 9380
+    },
+    {
+      "epoch": 3.672783403309459,
+      "grad_norm": 0.47915158304914873,
+      "learning_rate": 4.620203259889593e-06,
+      "loss": 0.4528,
+      "step": 9381
+    },
+    {
+      "epoch": 3.6731785626080513,
+      "grad_norm": 0.4852717018202506,
+      "learning_rate": 4.620120240391065e-06,
+      "loss": 0.4716,
+      "step": 9382
+    },
+    {
+      "epoch": 3.6735737219066436,
+      "grad_norm": 0.496307917359212,
+      "learning_rate": 4.620037212565992e-06,
+      "loss": 0.4636,
+      "step": 9383
+    },
+    {
+      "epoch": 3.673968881205236,
+      "grad_norm": 0.49948922206920776,
+      "learning_rate": 4.6199541764147e-06,
+      "loss": 0.45,
+      "step": 9384
+    },
+    {
+      "epoch": 3.674364040503828,
+      "grad_norm": 0.4906585809016417,
+      "learning_rate": 4.619871131937516e-06,
+      "loss": 0.4709,
+      "step": 9385
+    },
+    {
+      "epoch": 3.6747591998024203,
+      "grad_norm": 0.49580349029874954,
+      "learning_rate": 4.619788079134766e-06,
+      "loss": 0.4575,
+      "step": 9386
+    },
+    {
+      "epoch": 3.6751543591010125,
+      "grad_norm": 0.4877988171055439,
+      "learning_rate": 4.619705018006775e-06,
+      "loss": 0.4606,
+      "step": 9387
+    },
+    {
+      "epoch": 3.6755495183996048,
+      "grad_norm": 0.47635905394834055,
+      "learning_rate": 4.619621948553869e-06,
+      "loss": 0.4598,
+      "step": 9388
+    },
+    {
+      "epoch": 3.675944677698197,
+      "grad_norm": 0.4848704937642032,
+      "learning_rate": 4.619538870776375e-06,
+      "loss": 0.4531,
+      "step": 9389
+    },
+    {
+      "epoch": 3.6763398369967892,
+      "grad_norm": 0.5132723431379773,
+      "learning_rate": 4.61945578467462e-06,
+      "loss": 0.4813,
+      "step": 9390
+    },
+    {
+      "epoch": 3.6767349962953815,
+      "grad_norm": 0.49446717879197793,
+      "learning_rate": 4.61937269024893e-06,
+      "loss": 0.4732,
+      "step": 9391
+    },
+    {
+      "epoch": 3.6771301555939737,
+      "grad_norm": 0.48783487200107034,
+      "learning_rate": 4.619289587499631e-06,
+      "loss": 0.4642,
+      "step": 9392
+    },
+    {
+      "epoch": 3.677525314892566,
+      "grad_norm": 0.5095115851096903,
+      "learning_rate": 4.619206476427049e-06,
+      "loss": 0.456,
+      "step": 9393
+    },
+    {
+      "epoch": 3.677920474191158,
+      "grad_norm": 0.4947934593811385,
+      "learning_rate": 4.619123357031511e-06,
+      "loss": 0.4721,
+      "step": 9394
+    },
+    {
+      "epoch": 3.6783156334897504,
+      "grad_norm": 0.5075806473304344,
+      "learning_rate": 4.619040229313343e-06,
+      "loss": 0.4786,
+      "step": 9395
+    },
+    {
+      "epoch": 3.6787107927883427,
+      "grad_norm": 0.5801842763429373,
+      "learning_rate": 4.618957093272872e-06,
+      "loss": 0.4517,
+      "step": 9396
+    },
+    {
+      "epoch": 3.679105952086935,
+      "grad_norm": 0.5005276061566315,
+      "learning_rate": 4.618873948910425e-06,
+      "loss": 0.454,
+      "step": 9397
+    },
+    {
+      "epoch": 3.679501111385527,
+      "grad_norm": 0.5031083933218373,
+      "learning_rate": 4.618790796226327e-06,
+      "loss": 0.4526,
+      "step": 9398
+    },
+    {
+      "epoch": 3.6798962706841194,
+      "grad_norm": 0.4919801264259388,
+      "learning_rate": 4.618707635220905e-06,
+      "loss": 0.4524,
+      "step": 9399
+    },
+    {
+      "epoch": 3.6802914299827116,
+      "grad_norm": 0.4978593733121662,
+      "learning_rate": 4.6186244658944865e-06,
+      "loss": 0.4633,
+      "step": 9400
+    },
+    {
+      "epoch": 3.680686589281304,
+      "grad_norm": 0.4998560933361007,
+      "learning_rate": 4.618541288247397e-06,
+      "loss": 0.4571,
+      "step": 9401
+    },
+    {
+      "epoch": 3.681081748579896,
+      "grad_norm": 0.4916373727411663,
+      "learning_rate": 4.618458102279964e-06,
+      "loss": 0.4585,
+      "step": 9402
+    },
+    {
+      "epoch": 3.6814769078784884,
+      "grad_norm": 0.4888930129846756,
+      "learning_rate": 4.6183749079925145e-06,
+      "loss": 0.4678,
+      "step": 9403
+    },
+    {
+      "epoch": 3.6818720671770806,
+      "grad_norm": 0.4956175383165202,
+      "learning_rate": 4.618291705385374e-06,
+      "loss": 0.4691,
+      "step": 9404
+    },
+    {
+      "epoch": 3.6822672264756733,
+      "grad_norm": 0.5119254964059565,
+      "learning_rate": 4.61820849445887e-06,
+      "loss": 0.4503,
+      "step": 9405
+    },
+    {
+      "epoch": 3.6826623857742655,
+      "grad_norm": 0.49166332352889675,
+      "learning_rate": 4.61812527521333e-06,
+      "loss": 0.4644,
+      "step": 9406
+    },
+    {
+      "epoch": 3.6830575450728578,
+      "grad_norm": 0.4813292349115858,
+      "learning_rate": 4.61804204764908e-06,
+      "loss": 0.448,
+      "step": 9407
+    },
+    {
+      "epoch": 3.68345270437145,
+      "grad_norm": 0.49544114801840133,
+      "learning_rate": 4.6179588117664465e-06,
+      "loss": 0.4531,
+      "step": 9408
+    },
+    {
+      "epoch": 3.6838478636700422,
+      "grad_norm": 0.5062682853242432,
+      "learning_rate": 4.6178755675657565e-06,
+      "loss": 0.4658,
+      "step": 9409
+    },
+    {
+      "epoch": 3.6842430229686345,
+      "grad_norm": 0.4885565367565571,
+      "learning_rate": 4.617792315047338e-06,
+      "loss": 0.4576,
+      "step": 9410
+    },
+    {
+      "epoch": 3.6846381822672267,
+      "grad_norm": 0.4927003220978129,
+      "learning_rate": 4.6177090542115176e-06,
+      "loss": 0.4545,
+      "step": 9411
+    },
+    {
+      "epoch": 3.685033341565819,
+      "grad_norm": 0.5342756099499717,
+      "learning_rate": 4.617625785058622e-06,
+      "loss": 0.4444,
+      "step": 9412
+    },
+    {
+      "epoch": 3.685428500864411,
+      "grad_norm": 0.5191362354914825,
+      "learning_rate": 4.617542507588977e-06,
+      "loss": 0.4632,
+      "step": 9413
+    },
+    {
+      "epoch": 3.6858236601630034,
+      "grad_norm": 0.5005797890017317,
+      "learning_rate": 4.6174592218029115e-06,
+      "loss": 0.4565,
+      "step": 9414
+    },
+    {
+      "epoch": 3.6862188194615957,
+      "grad_norm": 0.5115586690359692,
+      "learning_rate": 4.617375927700752e-06,
+      "loss": 0.4716,
+      "step": 9415
+    },
+    {
+      "epoch": 3.686613978760188,
+      "grad_norm": 0.5168058976627257,
+      "learning_rate": 4.617292625282826e-06,
+      "loss": 0.4794,
+      "step": 9416
+    },
+    {
+      "epoch": 3.68700913805878,
+      "grad_norm": 0.5038401934419339,
+      "learning_rate": 4.617209314549459e-06,
+      "loss": 0.4795,
+      "step": 9417
+    },
+    {
+      "epoch": 3.6874042973573724,
+      "grad_norm": 0.49671493204697803,
+      "learning_rate": 4.617125995500981e-06,
+      "loss": 0.4742,
+      "step": 9418
+    },
+    {
+      "epoch": 3.6877994566559646,
+      "grad_norm": 0.5245802827596923,
+      "learning_rate": 4.617042668137717e-06,
+      "loss": 0.4655,
+      "step": 9419
+    },
+    {
+      "epoch": 3.688194615954557,
+      "grad_norm": 0.49027669847543004,
+      "learning_rate": 4.616959332459995e-06,
+      "loss": 0.4685,
+      "step": 9420
+    },
+    {
+      "epoch": 3.688589775253149,
+      "grad_norm": 0.5318803917463871,
+      "learning_rate": 4.616875988468142e-06,
+      "loss": 0.4566,
+      "step": 9421
+    },
+    {
+      "epoch": 3.6889849345517414,
+      "grad_norm": 0.47512179395486337,
+      "learning_rate": 4.616792636162486e-06,
+      "loss": 0.4508,
+      "step": 9422
+    },
+    {
+      "epoch": 3.6893800938503336,
+      "grad_norm": 0.4995389465446858,
+      "learning_rate": 4.616709275543353e-06,
+      "loss": 0.4573,
+      "step": 9423
+    },
+    {
+      "epoch": 3.689775253148926,
+      "grad_norm": 0.4952456922866283,
+      "learning_rate": 4.616625906611072e-06,
+      "loss": 0.4449,
+      "step": 9424
+    },
+    {
+      "epoch": 3.690170412447518,
+      "grad_norm": 0.4681009974875698,
+      "learning_rate": 4.61654252936597e-06,
+      "loss": 0.4443,
+      "step": 9425
+    },
+    {
+      "epoch": 3.6905655717461103,
+      "grad_norm": 0.49698293961782763,
+      "learning_rate": 4.616459143808374e-06,
+      "loss": 0.4504,
+      "step": 9426
+    },
+    {
+      "epoch": 3.6909607310447026,
+      "grad_norm": 0.4799614987205371,
+      "learning_rate": 4.616375749938612e-06,
+      "loss": 0.4827,
+      "step": 9427
+    },
+    {
+      "epoch": 3.691355890343295,
+      "grad_norm": 0.4935543789709199,
+      "learning_rate": 4.61629234775701e-06,
+      "loss": 0.454,
+      "step": 9428
+    },
+    {
+      "epoch": 3.691751049641887,
+      "grad_norm": 0.48312786289813203,
+      "learning_rate": 4.616208937263897e-06,
+      "loss": 0.4618,
+      "step": 9429
+    },
+    {
+      "epoch": 3.6921462089404793,
+      "grad_norm": 0.48779193972998197,
+      "learning_rate": 4.616125518459601e-06,
+      "loss": 0.4664,
+      "step": 9430
+    },
+    {
+      "epoch": 3.6925413682390715,
+      "grad_norm": 0.5102169263992223,
+      "learning_rate": 4.616042091344449e-06,
+      "loss": 0.4528,
+      "step": 9431
+    },
+    {
+      "epoch": 3.6929365275376638,
+      "grad_norm": 0.5028415523931336,
+      "learning_rate": 4.615958655918768e-06,
+      "loss": 0.4761,
+      "step": 9432
+    },
+    {
+      "epoch": 3.693331686836256,
+      "grad_norm": 0.49065489093573394,
+      "learning_rate": 4.615875212182887e-06,
+      "loss": 0.4392,
+      "step": 9433
+    },
+    {
+      "epoch": 3.6937268461348483,
+      "grad_norm": 0.49732637769698324,
+      "learning_rate": 4.615791760137133e-06,
+      "loss": 0.4633,
+      "step": 9434
+    },
+    {
+      "epoch": 3.6941220054334405,
+      "grad_norm": 0.49238892823032243,
+      "learning_rate": 4.615708299781833e-06,
+      "loss": 0.4622,
+      "step": 9435
+    },
+    {
+      "epoch": 3.6945171647320327,
+      "grad_norm": 0.49343740327676344,
+      "learning_rate": 4.615624831117316e-06,
+      "loss": 0.4568,
+      "step": 9436
+    },
+    {
+      "epoch": 3.694912324030625,
+      "grad_norm": 0.4861385224097656,
+      "learning_rate": 4.615541354143908e-06,
+      "loss": 0.4503,
+      "step": 9437
+    },
+    {
+      "epoch": 3.695307483329217,
+      "grad_norm": 0.497943873992621,
+      "learning_rate": 4.61545786886194e-06,
+      "loss": 0.4495,
+      "step": 9438
+    },
+    {
+      "epoch": 3.6957026426278095,
+      "grad_norm": 0.48616932102253096,
+      "learning_rate": 4.615374375271738e-06,
+      "loss": 0.476,
+      "step": 9439
+    },
+    {
+      "epoch": 3.6960978019264017,
+      "grad_norm": 0.49036991811170894,
+      "learning_rate": 4.615290873373629e-06,
+      "loss": 0.4574,
+      "step": 9440
+    },
+    {
+      "epoch": 3.696492961224994,
+      "grad_norm": 0.4874642622968561,
+      "learning_rate": 4.615207363167943e-06,
+      "loss": 0.4534,
+      "step": 9441
+    },
+    {
+      "epoch": 3.696888120523586,
+      "grad_norm": 0.4874523989618741,
+      "learning_rate": 4.615123844655006e-06,
+      "loss": 0.4669,
+      "step": 9442
+    },
+    {
+      "epoch": 3.6972832798221784,
+      "grad_norm": 0.4893447874095786,
+      "learning_rate": 4.615040317835147e-06,
+      "loss": 0.4441,
+      "step": 9443
+    },
+    {
+      "epoch": 3.6976784391207707,
+      "grad_norm": 0.5035498076080915,
+      "learning_rate": 4.614956782708694e-06,
+      "loss": 0.4679,
+      "step": 9444
+    },
+    {
+      "epoch": 3.698073598419363,
+      "grad_norm": 0.4973304541825886,
+      "learning_rate": 4.614873239275976e-06,
+      "loss": 0.4513,
+      "step": 9445
+    },
+    {
+      "epoch": 3.698468757717955,
+      "grad_norm": 0.5653732922507139,
+      "learning_rate": 4.6147896875373185e-06,
+      "loss": 0.4583,
+      "step": 9446
+    },
+    {
+      "epoch": 3.6988639170165474,
+      "grad_norm": 0.47730565163897437,
+      "learning_rate": 4.614706127493052e-06,
+      "loss": 0.4706,
+      "step": 9447
+    },
+    {
+      "epoch": 3.6992590763151396,
+      "grad_norm": 0.4900913046606925,
+      "learning_rate": 4.614622559143504e-06,
+      "loss": 0.4713,
+      "step": 9448
+    },
+    {
+      "epoch": 3.699654235613732,
+      "grad_norm": 0.4807624939670291,
+      "learning_rate": 4.614538982489003e-06,
+      "loss": 0.462,
+      "step": 9449
+    },
+    {
+      "epoch": 3.700049394912324,
+      "grad_norm": 0.5255103013851058,
+      "learning_rate": 4.614455397529876e-06,
+      "loss": 0.4596,
+      "step": 9450
+    },
+    {
+      "epoch": 3.7004445542109163,
+      "grad_norm": 0.49339176253468786,
+      "learning_rate": 4.614371804266453e-06,
+      "loss": 0.4446,
+      "step": 9451
+    },
+    {
+      "epoch": 3.7008397135095086,
+      "grad_norm": 0.479984900218154,
+      "learning_rate": 4.614288202699061e-06,
+      "loss": 0.4539,
+      "step": 9452
+    },
+    {
+      "epoch": 3.701234872808101,
+      "grad_norm": 0.4906333419555256,
+      "learning_rate": 4.6142045928280284e-06,
+      "loss": 0.4602,
+      "step": 9453
+    },
+    {
+      "epoch": 3.701630032106693,
+      "grad_norm": 0.47786003426235846,
+      "learning_rate": 4.6141209746536855e-06,
+      "loss": 0.451,
+      "step": 9454
+    },
+    {
+      "epoch": 3.7020251914052853,
+      "grad_norm": 0.505494984357728,
+      "learning_rate": 4.614037348176358e-06,
+      "loss": 0.4777,
+      "step": 9455
+    },
+    {
+      "epoch": 3.7024203507038775,
+      "grad_norm": 0.5053597345252311,
+      "learning_rate": 4.613953713396376e-06,
+      "loss": 0.4582,
+      "step": 9456
+    },
+    {
+      "epoch": 3.70281551000247,
+      "grad_norm": 0.5125562819742167,
+      "learning_rate": 4.613870070314067e-06,
+      "loss": 0.4626,
+      "step": 9457
+    },
+    {
+      "epoch": 3.703210669301062,
+      "grad_norm": 0.5017156061662159,
+      "learning_rate": 4.6137864189297595e-06,
+      "loss": 0.451,
+      "step": 9458
+    },
+    {
+      "epoch": 3.7036058285996543,
+      "grad_norm": 0.47667609480628925,
+      "learning_rate": 4.613702759243784e-06,
+      "loss": 0.4517,
+      "step": 9459
+    },
+    {
+      "epoch": 3.7040009878982465,
+      "grad_norm": 0.4825374845237388,
+      "learning_rate": 4.613619091256466e-06,
+      "loss": 0.4474,
+      "step": 9460
+    },
+    {
+      "epoch": 3.7043961471968387,
+      "grad_norm": 0.47835338941796834,
+      "learning_rate": 4.6135354149681365e-06,
+      "loss": 0.4469,
+      "step": 9461
+    },
+    {
+      "epoch": 3.704791306495431,
+      "grad_norm": 0.9294106269951347,
+      "learning_rate": 4.6134517303791235e-06,
+      "loss": 0.4607,
+      "step": 9462
+    },
+    {
+      "epoch": 3.705186465794023,
+      "grad_norm": 0.4930233385165018,
+      "learning_rate": 4.613368037489756e-06,
+      "loss": 0.4506,
+      "step": 9463
+    },
+    {
+      "epoch": 3.7055816250926155,
+      "grad_norm": 0.49884851479906,
+      "learning_rate": 4.613284336300361e-06,
+      "loss": 0.4536,
+      "step": 9464
+    },
+    {
+      "epoch": 3.7059767843912077,
+      "grad_norm": 0.47765620112543045,
+      "learning_rate": 4.613200626811268e-06,
+      "loss": 0.4554,
+      "step": 9465
+    },
+    {
+      "epoch": 3.7063719436898,
+      "grad_norm": 0.49830740900398934,
+      "learning_rate": 4.613116909022807e-06,
+      "loss": 0.4485,
+      "step": 9466
+    },
+    {
+      "epoch": 3.706767102988392,
+      "grad_norm": 0.5144490254440631,
+      "learning_rate": 4.613033182935306e-06,
+      "loss": 0.4664,
+      "step": 9467
+    },
+    {
+      "epoch": 3.7071622622869844,
+      "grad_norm": 0.5067754911121317,
+      "learning_rate": 4.6129494485490935e-06,
+      "loss": 0.4505,
+      "step": 9468
+    },
+    {
+      "epoch": 3.7075574215855767,
+      "grad_norm": 0.49318464717590493,
+      "learning_rate": 4.612865705864499e-06,
+      "loss": 0.4718,
+      "step": 9469
+    },
+    {
+      "epoch": 3.707952580884169,
+      "grad_norm": 0.5308810953951024,
+      "learning_rate": 4.612781954881851e-06,
+      "loss": 0.4685,
+      "step": 9470
+    },
+    {
+      "epoch": 3.708347740182761,
+      "grad_norm": 0.47912651709226944,
+      "learning_rate": 4.6126981956014775e-06,
+      "loss": 0.4418,
+      "step": 9471
+    },
+    {
+      "epoch": 3.7087428994813534,
+      "grad_norm": 0.48759890811022727,
+      "learning_rate": 4.612614428023709e-06,
+      "loss": 0.451,
+      "step": 9472
+    },
+    {
+      "epoch": 3.7091380587799456,
+      "grad_norm": 0.5022546921872667,
+      "learning_rate": 4.612530652148875e-06,
+      "loss": 0.4441,
+      "step": 9473
+    },
+    {
+      "epoch": 3.709533218078538,
+      "grad_norm": 0.48883903463677697,
+      "learning_rate": 4.6124468679773015e-06,
+      "loss": 0.44,
+      "step": 9474
+    },
+    {
+      "epoch": 3.70992837737713,
+      "grad_norm": 0.4925590816151284,
+      "learning_rate": 4.61236307550932e-06,
+      "loss": 0.4625,
+      "step": 9475
+    },
+    {
+      "epoch": 3.7103235366757223,
+      "grad_norm": 0.5102359588145222,
+      "learning_rate": 4.612279274745259e-06,
+      "loss": 0.4532,
+      "step": 9476
+    },
+    {
+      "epoch": 3.7107186959743146,
+      "grad_norm": 0.5016564703231445,
+      "learning_rate": 4.612195465685448e-06,
+      "loss": 0.4611,
+      "step": 9477
+    },
+    {
+      "epoch": 3.711113855272907,
+      "grad_norm": 0.4985293696763766,
+      "learning_rate": 4.612111648330216e-06,
+      "loss": 0.4617,
+      "step": 9478
+    },
+    {
+      "epoch": 3.711509014571499,
+      "grad_norm": 0.49461850359770454,
+      "learning_rate": 4.612027822679892e-06,
+      "loss": 0.4549,
+      "step": 9479
+    },
+    {
+      "epoch": 3.7119041738700913,
+      "grad_norm": 0.4879647278277727,
+      "learning_rate": 4.611943988734805e-06,
+      "loss": 0.4686,
+      "step": 9480
+    },
+    {
+      "epoch": 3.7122993331686835,
+      "grad_norm": 0.502142235265202,
+      "learning_rate": 4.611860146495284e-06,
+      "loss": 0.468,
+      "step": 9481
+    },
+    {
+      "epoch": 3.712694492467276,
+      "grad_norm": 0.5078297162937819,
+      "learning_rate": 4.611776295961659e-06,
+      "loss": 0.4844,
+      "step": 9482
+    },
+    {
+      "epoch": 3.713089651765868,
+      "grad_norm": 0.6651109373307685,
+      "learning_rate": 4.611692437134259e-06,
+      "loss": 0.4576,
+      "step": 9483
+    },
+    {
+      "epoch": 3.7134848110644603,
+      "grad_norm": 0.48181360567693204,
+      "learning_rate": 4.611608570013414e-06,
+      "loss": 0.4633,
+      "step": 9484
+    },
+    {
+      "epoch": 3.7138799703630525,
+      "grad_norm": 0.48089336396003285,
+      "learning_rate": 4.611524694599452e-06,
+      "loss": 0.4548,
+      "step": 9485
+    },
+    {
+      "epoch": 3.7142751296616447,
+      "grad_norm": 0.5565739216162738,
+      "learning_rate": 4.611440810892703e-06,
+      "loss": 0.4592,
+      "step": 9486
+    },
+    {
+      "epoch": 3.714670288960237,
+      "grad_norm": 0.5089387807662913,
+      "learning_rate": 4.611356918893497e-06,
+      "loss": 0.4673,
+      "step": 9487
+    },
+    {
+      "epoch": 3.7150654482588292,
+      "grad_norm": 0.4849507219283539,
+      "learning_rate": 4.611273018602164e-06,
+      "loss": 0.4579,
+      "step": 9488
+    },
+    {
+      "epoch": 3.7154606075574215,
+      "grad_norm": 0.4957394966619531,
+      "learning_rate": 4.611189110019032e-06,
+      "loss": 0.4722,
+      "step": 9489
+    },
+    {
+      "epoch": 3.7158557668560137,
+      "grad_norm": 0.4907758996231714,
+      "learning_rate": 4.6111051931444304e-06,
+      "loss": 0.4597,
+      "step": 9490
+    },
+    {
+      "epoch": 3.716250926154606,
+      "grad_norm": 0.5101532752645882,
+      "learning_rate": 4.61102126797869e-06,
+      "loss": 0.4828,
+      "step": 9491
+    },
+    {
+      "epoch": 3.716646085453198,
+      "grad_norm": 0.5008474308167519,
+      "learning_rate": 4.610937334522141e-06,
+      "loss": 0.4722,
+      "step": 9492
+    },
+    {
+      "epoch": 3.7170412447517904,
+      "grad_norm": 0.49062418438118194,
+      "learning_rate": 4.610853392775111e-06,
+      "loss": 0.4542,
+      "step": 9493
+    },
+    {
+      "epoch": 3.7174364040503827,
+      "grad_norm": 0.5014221426388613,
+      "learning_rate": 4.61076944273793e-06,
+      "loss": 0.4843,
+      "step": 9494
+    },
+    {
+      "epoch": 3.717831563348975,
+      "grad_norm": 0.4915499915118497,
+      "learning_rate": 4.61068548441093e-06,
+      "loss": 0.4587,
+      "step": 9495
+    },
+    {
+      "epoch": 3.718226722647567,
+      "grad_norm": 0.5132337480297305,
+      "learning_rate": 4.610601517794437e-06,
+      "loss": 0.4743,
+      "step": 9496
+    },
+    {
+      "epoch": 3.7186218819461594,
+      "grad_norm": 0.5016517224129324,
+      "learning_rate": 4.610517542888785e-06,
+      "loss": 0.462,
+      "step": 9497
+    },
+    {
+      "epoch": 3.7190170412447516,
+      "grad_norm": 0.5060950647207751,
+      "learning_rate": 4.6104335596943004e-06,
+      "loss": 0.4641,
+      "step": 9498
+    },
+    {
+      "epoch": 3.719412200543344,
+      "grad_norm": 0.5062740065120886,
+      "learning_rate": 4.610349568211314e-06,
+      "loss": 0.4592,
+      "step": 9499
+    },
+    {
+      "epoch": 3.719807359841936,
+      "grad_norm": 0.4930087960545402,
+      "learning_rate": 4.6102655684401575e-06,
+      "loss": 0.47,
+      "step": 9500
+    },
+    {
+      "epoch": 3.7202025191405284,
+      "grad_norm": 0.4898865945969133,
+      "learning_rate": 4.6101815603811576e-06,
+      "loss": 0.4437,
+      "step": 9501
+    },
+    {
+      "epoch": 3.7205976784391206,
+      "grad_norm": 0.4795656726959586,
+      "learning_rate": 4.610097544034647e-06,
+      "loss": 0.4453,
+      "step": 9502
+    },
+    {
+      "epoch": 3.720992837737713,
+      "grad_norm": 0.49426700494295145,
+      "learning_rate": 4.610013519400954e-06,
+      "loss": 0.4864,
+      "step": 9503
+    },
+    {
+      "epoch": 3.721387997036305,
+      "grad_norm": 0.4887121669736692,
+      "learning_rate": 4.609929486480409e-06,
+      "loss": 0.4573,
+      "step": 9504
+    },
+    {
+      "epoch": 3.7217831563348973,
+      "grad_norm": 0.4905352215891702,
+      "learning_rate": 4.609845445273343e-06,
+      "loss": 0.4582,
+      "step": 9505
+    },
+    {
+      "epoch": 3.7221783156334896,
+      "grad_norm": 0.5105211554553892,
+      "learning_rate": 4.6097613957800845e-06,
+      "loss": 0.4693,
+      "step": 9506
+    },
+    {
+      "epoch": 3.722573474932082,
+      "grad_norm": 0.4987264748579873,
+      "learning_rate": 4.6096773380009655e-06,
+      "loss": 0.4679,
+      "step": 9507
+    },
+    {
+      "epoch": 3.722968634230674,
+      "grad_norm": 0.49566139627836764,
+      "learning_rate": 4.609593271936313e-06,
+      "loss": 0.4483,
+      "step": 9508
+    },
+    {
+      "epoch": 3.7233637935292663,
+      "grad_norm": 0.49187996336706025,
+      "learning_rate": 4.609509197586461e-06,
+      "loss": 0.467,
+      "step": 9509
+    },
+    {
+      "epoch": 3.7237589528278585,
+      "grad_norm": 0.4820091496779555,
+      "learning_rate": 4.609425114951737e-06,
+      "loss": 0.4536,
+      "step": 9510
+    },
+    {
+      "epoch": 3.7241541121264508,
+      "grad_norm": 0.5021340173430089,
+      "learning_rate": 4.609341024032472e-06,
+      "loss": 0.4824,
+      "step": 9511
+    },
+    {
+      "epoch": 3.724549271425043,
+      "grad_norm": 0.49803869540293166,
+      "learning_rate": 4.609256924828997e-06,
+      "loss": 0.4547,
+      "step": 9512
+    },
+    {
+      "epoch": 3.7249444307236352,
+      "grad_norm": 0.5128440482147195,
+      "learning_rate": 4.60917281734164e-06,
+      "loss": 0.4608,
+      "step": 9513
+    },
+    {
+      "epoch": 3.7253395900222275,
+      "grad_norm": 0.4913277569926873,
+      "learning_rate": 4.609088701570735e-06,
+      "loss": 0.4537,
+      "step": 9514
+    },
+    {
+      "epoch": 3.7257347493208197,
+      "grad_norm": 0.495974924098836,
+      "learning_rate": 4.609004577516609e-06,
+      "loss": 0.4734,
+      "step": 9515
+    },
+    {
+      "epoch": 3.726129908619412,
+      "grad_norm": 0.49140903339070413,
+      "learning_rate": 4.608920445179594e-06,
+      "loss": 0.4442,
+      "step": 9516
+    },
+    {
+      "epoch": 3.726525067918004,
+      "grad_norm": 0.5217558182771603,
+      "learning_rate": 4.60883630456002e-06,
+      "loss": 0.4896,
+      "step": 9517
+    },
+    {
+      "epoch": 3.7269202272165964,
+      "grad_norm": 0.49384622356957825,
+      "learning_rate": 4.608752155658218e-06,
+      "loss": 0.4412,
+      "step": 9518
+    },
+    {
+      "epoch": 3.7273153865151887,
+      "grad_norm": 0.4857726615455569,
+      "learning_rate": 4.608667998474518e-06,
+      "loss": 0.4497,
+      "step": 9519
+    },
+    {
+      "epoch": 3.727710545813781,
+      "grad_norm": 0.4963906295933952,
+      "learning_rate": 4.60858383300925e-06,
+      "loss": 0.4613,
+      "step": 9520
+    },
+    {
+      "epoch": 3.728105705112373,
+      "grad_norm": 0.4838697905450302,
+      "learning_rate": 4.608499659262745e-06,
+      "loss": 0.4584,
+      "step": 9521
+    },
+    {
+      "epoch": 3.7285008644109654,
+      "grad_norm": 0.48800127051436243,
+      "learning_rate": 4.608415477235334e-06,
+      "loss": 0.4589,
+      "step": 9522
+    },
+    {
+      "epoch": 3.7288960237095576,
+      "grad_norm": 0.48246388572291404,
+      "learning_rate": 4.6083312869273475e-06,
+      "loss": 0.451,
+      "step": 9523
+    },
+    {
+      "epoch": 3.7292911830081503,
+      "grad_norm": 0.4820982897305016,
+      "learning_rate": 4.608247088339116e-06,
+      "loss": 0.4481,
+      "step": 9524
+    },
+    {
+      "epoch": 3.7296863423067426,
+      "grad_norm": 0.4782123289577348,
+      "learning_rate": 4.60816288147097e-06,
+      "loss": 0.4613,
+      "step": 9525
+    },
+    {
+      "epoch": 3.730081501605335,
+      "grad_norm": 0.4893516657625509,
+      "learning_rate": 4.60807866632324e-06,
+      "loss": 0.4555,
+      "step": 9526
+    },
+    {
+      "epoch": 3.730476660903927,
+      "grad_norm": 0.489071023151812,
+      "learning_rate": 4.607994442896257e-06,
+      "loss": 0.4623,
+      "step": 9527
+    },
+    {
+      "epoch": 3.7308718202025193,
+      "grad_norm": 0.49687366002746247,
+      "learning_rate": 4.607910211190353e-06,
+      "loss": 0.4398,
+      "step": 9528
+    },
+    {
+      "epoch": 3.7312669795011115,
+      "grad_norm": 0.47881395393697335,
+      "learning_rate": 4.607825971205857e-06,
+      "loss": 0.4469,
+      "step": 9529
+    },
+    {
+      "epoch": 3.7316621387997038,
+      "grad_norm": 0.5130814385073522,
+      "learning_rate": 4.6077417229430995e-06,
+      "loss": 0.4628,
+      "step": 9530
+    },
+    {
+      "epoch": 3.732057298098296,
+      "grad_norm": 0.48593565257627325,
+      "learning_rate": 4.607657466402414e-06,
+      "loss": 0.4533,
+      "step": 9531
+    },
+    {
+      "epoch": 3.7324524573968882,
+      "grad_norm": 0.4875631077933674,
+      "learning_rate": 4.607573201584129e-06,
+      "loss": 0.4579,
+      "step": 9532
+    },
+    {
+      "epoch": 3.7328476166954805,
+      "grad_norm": 0.4957260704045972,
+      "learning_rate": 4.607488928488576e-06,
+      "loss": 0.4563,
+      "step": 9533
+    },
+    {
+      "epoch": 3.7332427759940727,
+      "grad_norm": 0.4860278696039461,
+      "learning_rate": 4.607404647116087e-06,
+      "loss": 0.4664,
+      "step": 9534
+    },
+    {
+      "epoch": 3.733637935292665,
+      "grad_norm": 0.4845068061589555,
+      "learning_rate": 4.607320357466992e-06,
+      "loss": 0.4614,
+      "step": 9535
+    },
+    {
+      "epoch": 3.734033094591257,
+      "grad_norm": 0.4786694984990013,
+      "learning_rate": 4.607236059541622e-06,
+      "loss": 0.4601,
+      "step": 9536
+    },
+    {
+      "epoch": 3.7344282538898494,
+      "grad_norm": 0.5088132150023401,
+      "learning_rate": 4.6071517533403085e-06,
+      "loss": 0.4538,
+      "step": 9537
+    },
+    {
+      "epoch": 3.7348234131884417,
+      "grad_norm": 0.4908292637789205,
+      "learning_rate": 4.6070674388633825e-06,
+      "loss": 0.4686,
+      "step": 9538
+    },
+    {
+      "epoch": 3.735218572487034,
+      "grad_norm": 0.5142726426805816,
+      "learning_rate": 4.606983116111175e-06,
+      "loss": 0.4719,
+      "step": 9539
+    },
+    {
+      "epoch": 3.735613731785626,
+      "grad_norm": 0.49016313325540406,
+      "learning_rate": 4.606898785084017e-06,
+      "loss": 0.4535,
+      "step": 9540
+    },
+    {
+      "epoch": 3.7360088910842184,
+      "grad_norm": 0.49285245557129015,
+      "learning_rate": 4.60681444578224e-06,
+      "loss": 0.4645,
+      "step": 9541
+    },
+    {
+      "epoch": 3.7364040503828106,
+      "grad_norm": 0.4839808550700243,
+      "learning_rate": 4.6067300982061754e-06,
+      "loss": 0.4491,
+      "step": 9542
+    },
+    {
+      "epoch": 3.736799209681403,
+      "grad_norm": 0.484570336078592,
+      "learning_rate": 4.606645742356155e-06,
+      "loss": 0.4594,
+      "step": 9543
+    },
+    {
+      "epoch": 3.737194368979995,
+      "grad_norm": 0.48518657604346466,
+      "learning_rate": 4.606561378232508e-06,
+      "loss": 0.4701,
+      "step": 9544
+    },
+    {
+      "epoch": 3.7375895282785874,
+      "grad_norm": 0.48628553075992975,
+      "learning_rate": 4.606477005835568e-06,
+      "loss": 0.4533,
+      "step": 9545
+    },
+    {
+      "epoch": 3.7379846875771796,
+      "grad_norm": 0.5123338585282978,
+      "learning_rate": 4.6063926251656656e-06,
+      "loss": 0.5095,
+      "step": 9546
+    },
+    {
+      "epoch": 3.738379846875772,
+      "grad_norm": 0.5102529842049373,
+      "learning_rate": 4.6063082362231306e-06,
+      "loss": 0.4866,
+      "step": 9547
+    },
+    {
+      "epoch": 3.738775006174364,
+      "grad_norm": 0.4795300343781412,
+      "learning_rate": 4.606223839008297e-06,
+      "loss": 0.4565,
+      "step": 9548
+    },
+    {
+      "epoch": 3.7391701654729563,
+      "grad_norm": 0.500458305229408,
+      "learning_rate": 4.6061394335214945e-06,
+      "loss": 0.4546,
+      "step": 9549
+    },
+    {
+      "epoch": 3.7395653247715486,
+      "grad_norm": 0.49447185079146605,
+      "learning_rate": 4.606055019763056e-06,
+      "loss": 0.4711,
+      "step": 9550
+    },
+    {
+      "epoch": 3.739960484070141,
+      "grad_norm": 0.4837507662358641,
+      "learning_rate": 4.6059705977333116e-06,
+      "loss": 0.4492,
+      "step": 9551
+    },
+    {
+      "epoch": 3.740355643368733,
+      "grad_norm": 0.6482460092048694,
+      "learning_rate": 4.605886167432595e-06,
+      "loss": 0.4591,
+      "step": 9552
+    },
+    {
+      "epoch": 3.7407508026673253,
+      "grad_norm": 0.48576022911474664,
+      "learning_rate": 4.605801728861235e-06,
+      "loss": 0.4561,
+      "step": 9553
+    },
+    {
+      "epoch": 3.7411459619659175,
+      "grad_norm": 0.4901974973756972,
+      "learning_rate": 4.6057172820195635e-06,
+      "loss": 0.4713,
+      "step": 9554
+    },
+    {
+      "epoch": 3.7415411212645098,
+      "grad_norm": 0.49639360986444836,
+      "learning_rate": 4.605632826907915e-06,
+      "loss": 0.4759,
+      "step": 9555
+    },
+    {
+      "epoch": 3.741936280563102,
+      "grad_norm": 0.4931648478189971,
+      "learning_rate": 4.605548363526619e-06,
+      "loss": 0.4587,
+      "step": 9556
+    },
+    {
+      "epoch": 3.7423314398616943,
+      "grad_norm": 0.4688955095749135,
+      "learning_rate": 4.605463891876006e-06,
+      "loss": 0.4648,
+      "step": 9557
+    },
+    {
+      "epoch": 3.7427265991602865,
+      "grad_norm": 0.49745821587087424,
+      "learning_rate": 4.605379411956411e-06,
+      "loss": 0.4603,
+      "step": 9558
+    },
+    {
+      "epoch": 3.7431217584588787,
+      "grad_norm": 0.46605724180238206,
+      "learning_rate": 4.605294923768164e-06,
+      "loss": 0.4536,
+      "step": 9559
+    },
+    {
+      "epoch": 3.743516917757471,
+      "grad_norm": 0.48666332590044686,
+      "learning_rate": 4.605210427311596e-06,
+      "loss": 0.4794,
+      "step": 9560
+    },
+    {
+      "epoch": 3.743912077056063,
+      "grad_norm": 0.49885552329714344,
+      "learning_rate": 4.605125922587041e-06,
+      "loss": 0.451,
+      "step": 9561
+    },
+    {
+      "epoch": 3.7443072363546555,
+      "grad_norm": 0.5000142082417693,
+      "learning_rate": 4.6050414095948294e-06,
+      "loss": 0.4529,
+      "step": 9562
+    },
+    {
+      "epoch": 3.7447023956532477,
+      "grad_norm": 0.4862982293683471,
+      "learning_rate": 4.604956888335292e-06,
+      "loss": 0.4614,
+      "step": 9563
+    },
+    {
+      "epoch": 3.74509755495184,
+      "grad_norm": 0.48972690615641834,
+      "learning_rate": 4.604872358808764e-06,
+      "loss": 0.4696,
+      "step": 9564
+    },
+    {
+      "epoch": 3.745492714250432,
+      "grad_norm": 0.48344315601917986,
+      "learning_rate": 4.604787821015575e-06,
+      "loss": 0.4618,
+      "step": 9565
+    },
+    {
+      "epoch": 3.7458878735490244,
+      "grad_norm": 0.501362692935816,
+      "learning_rate": 4.604703274956057e-06,
+      "loss": 0.4465,
+      "step": 9566
+    },
+    {
+      "epoch": 3.7462830328476167,
+      "grad_norm": 0.4869221765616658,
+      "learning_rate": 4.604618720630542e-06,
+      "loss": 0.4519,
+      "step": 9567
+    },
+    {
+      "epoch": 3.746678192146209,
+      "grad_norm": 0.49044552780260375,
+      "learning_rate": 4.604534158039364e-06,
+      "loss": 0.4575,
+      "step": 9568
+    },
+    {
+      "epoch": 3.747073351444801,
+      "grad_norm": 0.5337351622614847,
+      "learning_rate": 4.604449587182854e-06,
+      "loss": 0.4757,
+      "step": 9569
+    },
+    {
+      "epoch": 3.7474685107433934,
+      "grad_norm": 0.5142889870437418,
+      "learning_rate": 4.6043650080613434e-06,
+      "loss": 0.4368,
+      "step": 9570
+    },
+    {
+      "epoch": 3.7478636700419856,
+      "grad_norm": 0.5002069423060246,
+      "learning_rate": 4.604280420675165e-06,
+      "loss": 0.4662,
+      "step": 9571
+    },
+    {
+      "epoch": 3.748258829340578,
+      "grad_norm": 0.5012425468189804,
+      "learning_rate": 4.60419582502465e-06,
+      "loss": 0.4689,
+      "step": 9572
+    },
+    {
+      "epoch": 3.74865398863917,
+      "grad_norm": 0.48744682141535584,
+      "learning_rate": 4.6041112211101325e-06,
+      "loss": 0.4783,
+      "step": 9573
+    },
+    {
+      "epoch": 3.7490491479377623,
+      "grad_norm": 0.49788598446759974,
+      "learning_rate": 4.604026608931943e-06,
+      "loss": 0.4668,
+      "step": 9574
+    },
+    {
+      "epoch": 3.7494443072363546,
+      "grad_norm": 0.4964539038735511,
+      "learning_rate": 4.6039419884904155e-06,
+      "loss": 0.4768,
+      "step": 9575
+    },
+    {
+      "epoch": 3.749839466534947,
+      "grad_norm": 0.5034149134145032,
+      "learning_rate": 4.603857359785881e-06,
+      "loss": 0.4654,
+      "step": 9576
+    },
+    {
+      "epoch": 3.750234625833539,
+      "grad_norm": 0.5102791225039909,
+      "learning_rate": 4.6037727228186715e-06,
+      "loss": 0.4767,
+      "step": 9577
+    },
+    {
+      "epoch": 3.7506297851321313,
+      "grad_norm": 0.49641266816152013,
+      "learning_rate": 4.603688077589121e-06,
+      "loss": 0.4692,
+      "step": 9578
+    },
+    {
+      "epoch": 3.7510249444307235,
+      "grad_norm": 0.5170425495731209,
+      "learning_rate": 4.603603424097561e-06,
+      "loss": 0.4437,
+      "step": 9579
+    },
+    {
+      "epoch": 3.751420103729316,
+      "grad_norm": 0.4903857984331723,
+      "learning_rate": 4.603518762344325e-06,
+      "loss": 0.4665,
+      "step": 9580
+    },
+    {
+      "epoch": 3.751815263027908,
+      "grad_norm": 0.49410476736023734,
+      "learning_rate": 4.603434092329743e-06,
+      "loss": 0.4648,
+      "step": 9581
+    },
+    {
+      "epoch": 3.7522104223265003,
+      "grad_norm": 0.47620798317913865,
+      "learning_rate": 4.603349414054151e-06,
+      "loss": 0.4558,
+      "step": 9582
+    },
+    {
+      "epoch": 3.7526055816250925,
+      "grad_norm": 0.4909803148027834,
+      "learning_rate": 4.603264727517879e-06,
+      "loss": 0.4728,
+      "step": 9583
+    },
+    {
+      "epoch": 3.7530007409236847,
+      "grad_norm": 0.4841546764667915,
+      "learning_rate": 4.60318003272126e-06,
+      "loss": 0.4384,
+      "step": 9584
+    },
+    {
+      "epoch": 3.753395900222277,
+      "grad_norm": 0.49138074599478704,
+      "learning_rate": 4.603095329664627e-06,
+      "loss": 0.4509,
+      "step": 9585
+    },
+    {
+      "epoch": 3.7537910595208692,
+      "grad_norm": 0.4916572867617533,
+      "learning_rate": 4.6030106183483135e-06,
+      "loss": 0.4493,
+      "step": 9586
+    },
+    {
+      "epoch": 3.7541862188194615,
+      "grad_norm": 0.4867907743227322,
+      "learning_rate": 4.60292589877265e-06,
+      "loss": 0.4561,
+      "step": 9587
+    },
+    {
+      "epoch": 3.7545813781180537,
+      "grad_norm": 0.5044721101451728,
+      "learning_rate": 4.6028411709379715e-06,
+      "loss": 0.4525,
+      "step": 9588
+    },
+    {
+      "epoch": 3.754976537416646,
+      "grad_norm": 0.47514228642887263,
+      "learning_rate": 4.602756434844609e-06,
+      "loss": 0.4616,
+      "step": 9589
+    },
+    {
+      "epoch": 3.755371696715238,
+      "grad_norm": 0.776026961750764,
+      "learning_rate": 4.6026716904928965e-06,
+      "loss": 0.4883,
+      "step": 9590
+    },
+    {
+      "epoch": 3.7557668560138304,
+      "grad_norm": 0.4937700305585744,
+      "learning_rate": 4.602586937883167e-06,
+      "loss": 0.4647,
+      "step": 9591
+    },
+    {
+      "epoch": 3.7561620153124227,
+      "grad_norm": 0.47957905951413793,
+      "learning_rate": 4.6025021770157516e-06,
+      "loss": 0.4446,
+      "step": 9592
+    },
+    {
+      "epoch": 3.756557174611015,
+      "grad_norm": 0.5017133949581569,
+      "learning_rate": 4.602417407890984e-06,
+      "loss": 0.4884,
+      "step": 9593
+    },
+    {
+      "epoch": 3.756952333909607,
+      "grad_norm": 0.4870056949571294,
+      "learning_rate": 4.602332630509199e-06,
+      "loss": 0.4602,
+      "step": 9594
+    },
+    {
+      "epoch": 3.7573474932082,
+      "grad_norm": 0.5202279557981537,
+      "learning_rate": 4.602247844870728e-06,
+      "loss": 0.4776,
+      "step": 9595
+    },
+    {
+      "epoch": 3.757742652506792,
+      "grad_norm": 0.4973526737448953,
+      "learning_rate": 4.602163050975903e-06,
+      "loss": 0.4423,
+      "step": 9596
+    },
+    {
+      "epoch": 3.7581378118053843,
+      "grad_norm": 0.4983329516809982,
+      "learning_rate": 4.602078248825058e-06,
+      "loss": 0.4657,
+      "step": 9597
+    },
+    {
+      "epoch": 3.7585329711039765,
+      "grad_norm": 0.49221414055030543,
+      "learning_rate": 4.601993438418527e-06,
+      "loss": 0.4544,
+      "step": 9598
+    },
+    {
+      "epoch": 3.758928130402569,
+      "grad_norm": 0.4865035442026118,
+      "learning_rate": 4.601908619756642e-06,
+      "loss": 0.4511,
+      "step": 9599
+    },
+    {
+      "epoch": 3.759323289701161,
+      "grad_norm": 0.48356478090463534,
+      "learning_rate": 4.601823792839735e-06,
+      "loss": 0.4738,
+      "step": 9600
+    },
+    {
+      "epoch": 3.7597184489997533,
+      "grad_norm": 0.49314242592096336,
+      "learning_rate": 4.601738957668142e-06,
+      "loss": 0.463,
+      "step": 9601
+    },
+    {
+      "epoch": 3.7601136082983455,
+      "grad_norm": 0.5078494446402765,
+      "learning_rate": 4.601654114242194e-06,
+      "loss": 0.4618,
+      "step": 9602
+    },
+    {
+      "epoch": 3.7605087675969378,
+      "grad_norm": 0.5028843214675345,
+      "learning_rate": 4.6015692625622255e-06,
+      "loss": 0.4608,
+      "step": 9603
+    },
+    {
+      "epoch": 3.76090392689553,
+      "grad_norm": 0.5008323117874007,
+      "learning_rate": 4.601484402628569e-06,
+      "loss": 0.4549,
+      "step": 9604
+    },
+    {
+      "epoch": 3.7612990861941222,
+      "grad_norm": 0.4850454741183229,
+      "learning_rate": 4.6013995344415565e-06,
+      "loss": 0.4627,
+      "step": 9605
+    },
+    {
+      "epoch": 3.7616942454927145,
+      "grad_norm": 0.48808279245655617,
+      "learning_rate": 4.601314658001524e-06,
+      "loss": 0.4535,
+      "step": 9606
+    },
+    {
+      "epoch": 3.7620894047913067,
+      "grad_norm": 0.5066989763656677,
+      "learning_rate": 4.601229773308802e-06,
+      "loss": 0.4751,
+      "step": 9607
+    },
+    {
+      "epoch": 3.762484564089899,
+      "grad_norm": 0.4805137326607607,
+      "learning_rate": 4.601144880363726e-06,
+      "loss": 0.4638,
+      "step": 9608
+    },
+    {
+      "epoch": 3.762879723388491,
+      "grad_norm": 0.4922622961214075,
+      "learning_rate": 4.60105997916663e-06,
+      "loss": 0.4686,
+      "step": 9609
+    },
+    {
+      "epoch": 3.7632748826870834,
+      "grad_norm": 0.4810705349543995,
+      "learning_rate": 4.600975069717846e-06,
+      "loss": 0.4461,
+      "step": 9610
+    },
+    {
+      "epoch": 3.7636700419856757,
+      "grad_norm": 0.4773354945166677,
+      "learning_rate": 4.6008901520177065e-06,
+      "loss": 0.4516,
+      "step": 9611
+    },
+    {
+      "epoch": 3.764065201284268,
+      "grad_norm": 0.49206349143578293,
+      "learning_rate": 4.600805226066547e-06,
+      "loss": 0.4485,
+      "step": 9612
+    },
+    {
+      "epoch": 3.76446036058286,
+      "grad_norm": 0.4814701688954984,
+      "learning_rate": 4.6007202918647e-06,
+      "loss": 0.4489,
+      "step": 9613
+    },
+    {
+      "epoch": 3.7648555198814524,
+      "grad_norm": 0.5022048385321567,
+      "learning_rate": 4.600635349412499e-06,
+      "loss": 0.4721,
+      "step": 9614
+    },
+    {
+      "epoch": 3.7652506791800446,
+      "grad_norm": 0.48243788870808746,
+      "learning_rate": 4.600550398710278e-06,
+      "loss": 0.4464,
+      "step": 9615
+    },
+    {
+      "epoch": 3.765645838478637,
+      "grad_norm": 0.48757340703408836,
+      "learning_rate": 4.600465439758371e-06,
+      "loss": 0.4703,
+      "step": 9616
+    },
+    {
+      "epoch": 3.766040997777229,
+      "grad_norm": 0.48716159886857296,
+      "learning_rate": 4.600380472557112e-06,
+      "loss": 0.4519,
+      "step": 9617
+    },
+    {
+      "epoch": 3.7664361570758214,
+      "grad_norm": 0.5069818045941086,
+      "learning_rate": 4.600295497106833e-06,
+      "loss": 0.4943,
+      "step": 9618
+    },
+    {
+      "epoch": 3.7668313163744136,
+      "grad_norm": 0.5022666320922664,
+      "learning_rate": 4.6002105134078675e-06,
+      "loss": 0.4673,
+      "step": 9619
+    },
+    {
+      "epoch": 3.767226475673006,
+      "grad_norm": 0.4807723287562549,
+      "learning_rate": 4.600125521460552e-06,
+      "loss": 0.4646,
+      "step": 9620
+    },
+    {
+      "epoch": 3.767621634971598,
+      "grad_norm": 0.4912477141689686,
+      "learning_rate": 4.600040521265219e-06,
+      "loss": 0.4641,
+      "step": 9621
+    },
+    {
+      "epoch": 3.7680167942701903,
+      "grad_norm": 0.49929865985825106,
+      "learning_rate": 4.599955512822201e-06,
+      "loss": 0.4823,
+      "step": 9622
+    },
+    {
+      "epoch": 3.7684119535687826,
+      "grad_norm": 0.49230115169608873,
+      "learning_rate": 4.599870496131833e-06,
+      "loss": 0.4593,
+      "step": 9623
+    },
+    {
+      "epoch": 3.768807112867375,
+      "grad_norm": 0.48105300330907896,
+      "learning_rate": 4.59978547119445e-06,
+      "loss": 0.444,
+      "step": 9624
+    },
+    {
+      "epoch": 3.769202272165967,
+      "grad_norm": 0.4818330655568901,
+      "learning_rate": 4.599700438010382e-06,
+      "loss": 0.4473,
+      "step": 9625
+    },
+    {
+      "epoch": 3.7695974314645593,
+      "grad_norm": 0.4963785455532329,
+      "learning_rate": 4.599615396579968e-06,
+      "loss": 0.4342,
+      "step": 9626
+    },
+    {
+      "epoch": 3.7699925907631515,
+      "grad_norm": 0.4830399833200486,
+      "learning_rate": 4.5995303469035406e-06,
+      "loss": 0.44,
+      "step": 9627
+    },
+    {
+      "epoch": 3.7703877500617438,
+      "grad_norm": 0.47990178930501065,
+      "learning_rate": 4.59944528898143e-06,
+      "loss": 0.4504,
+      "step": 9628
+    },
+    {
+      "epoch": 3.770782909360336,
+      "grad_norm": 0.4786217792852282,
+      "learning_rate": 4.599360222813975e-06,
+      "loss": 0.4661,
+      "step": 9629
+    },
+    {
+      "epoch": 3.7711780686589282,
+      "grad_norm": 0.4805124331616531,
+      "learning_rate": 4.599275148401507e-06,
+      "loss": 0.4583,
+      "step": 9630
+    },
+    {
+      "epoch": 3.7715732279575205,
+      "grad_norm": 0.4879125432986269,
+      "learning_rate": 4.599190065744362e-06,
+      "loss": 0.4528,
+      "step": 9631
+    },
+    {
+      "epoch": 3.7719683872561127,
+      "grad_norm": 0.48314579798707913,
+      "learning_rate": 4.5991049748428725e-06,
+      "loss": 0.4537,
+      "step": 9632
+    },
+    {
+      "epoch": 3.772363546554705,
+      "grad_norm": 0.4903701935863649,
+      "learning_rate": 4.599019875697374e-06,
+      "loss": 0.4512,
+      "step": 9633
+    },
+    {
+      "epoch": 3.772758705853297,
+      "grad_norm": 0.5023398611383797,
+      "learning_rate": 4.598934768308199e-06,
+      "loss": 0.4546,
+      "step": 9634
+    },
+    {
+      "epoch": 3.7731538651518894,
+      "grad_norm": 0.6910082967623805,
+      "learning_rate": 4.598849652675683e-06,
+      "loss": 0.454,
+      "step": 9635
+    },
+    {
+      "epoch": 3.7735490244504817,
+      "grad_norm": 0.48924919480622175,
+      "learning_rate": 4.598764528800161e-06,
+      "loss": 0.4573,
+      "step": 9636
+    },
+    {
+      "epoch": 3.773944183749074,
+      "grad_norm": 0.5032570900259695,
+      "learning_rate": 4.598679396681964e-06,
+      "loss": 0.4735,
+      "step": 9637
+    },
+    {
+      "epoch": 3.774339343047666,
+      "grad_norm": 0.4884216635553054,
+      "learning_rate": 4.598594256321431e-06,
+      "loss": 0.4593,
+      "step": 9638
+    },
+    {
+      "epoch": 3.7747345023462584,
+      "grad_norm": 0.4839339436338744,
+      "learning_rate": 4.598509107718894e-06,
+      "loss": 0.4734,
+      "step": 9639
+    },
+    {
+      "epoch": 3.7751296616448506,
+      "grad_norm": 0.48156978064749584,
+      "learning_rate": 4.598423950874687e-06,
+      "loss": 0.4514,
+      "step": 9640
+    },
+    {
+      "epoch": 3.775524820943443,
+      "grad_norm": 0.4866270246823189,
+      "learning_rate": 4.598338785789144e-06,
+      "loss": 0.4615,
+      "step": 9641
+    },
+    {
+      "epoch": 3.775919980242035,
+      "grad_norm": 0.47743928665454505,
+      "learning_rate": 4.598253612462601e-06,
+      "loss": 0.4578,
+      "step": 9642
+    },
+    {
+      "epoch": 3.7763151395406274,
+      "grad_norm": 0.47599434611222147,
+      "learning_rate": 4.598168430895392e-06,
+      "loss": 0.4495,
+      "step": 9643
+    },
+    {
+      "epoch": 3.7767102988392196,
+      "grad_norm": 0.49548131853505706,
+      "learning_rate": 4.598083241087852e-06,
+      "loss": 0.4481,
+      "step": 9644
+    },
+    {
+      "epoch": 3.777105458137812,
+      "grad_norm": 0.536021204426832,
+      "learning_rate": 4.597998043040315e-06,
+      "loss": 0.4548,
+      "step": 9645
+    },
+    {
+      "epoch": 3.777500617436404,
+      "grad_norm": 0.49859452013864736,
+      "learning_rate": 4.597912836753116e-06,
+      "loss": 0.4472,
+      "step": 9646
+    },
+    {
+      "epoch": 3.7778957767349963,
+      "grad_norm": 0.4904930266960114,
+      "learning_rate": 4.597827622226588e-06,
+      "loss": 0.4407,
+      "step": 9647
+    },
+    {
+      "epoch": 3.7782909360335886,
+      "grad_norm": 0.5632944952584257,
+      "learning_rate": 4.597742399461067e-06,
+      "loss": 0.4677,
+      "step": 9648
+    },
+    {
+      "epoch": 3.778686095332181,
+      "grad_norm": 0.4825786026169142,
+      "learning_rate": 4.597657168456889e-06,
+      "loss": 0.4711,
+      "step": 9649
+    },
+    {
+      "epoch": 3.779081254630773,
+      "grad_norm": 0.48663207299892375,
+      "learning_rate": 4.5975719292143865e-06,
+      "loss": 0.4562,
+      "step": 9650
+    },
+    {
+      "epoch": 3.7794764139293653,
+      "grad_norm": 0.48151353926840385,
+      "learning_rate": 4.5974866817338955e-06,
+      "loss": 0.4491,
+      "step": 9651
+    },
+    {
+      "epoch": 3.7798715732279575,
+      "grad_norm": 0.490395399162099,
+      "learning_rate": 4.597401426015751e-06,
+      "loss": 0.4708,
+      "step": 9652
+    },
+    {
+      "epoch": 3.7802667325265498,
+      "grad_norm": 0.48180099915518,
+      "learning_rate": 4.597316162060287e-06,
+      "loss": 0.4571,
+      "step": 9653
+    },
+    {
+      "epoch": 3.780661891825142,
+      "grad_norm": 0.5093098515857098,
+      "learning_rate": 4.597230889867837e-06,
+      "loss": 0.4601,
+      "step": 9654
+    },
+    {
+      "epoch": 3.7810570511237342,
+      "grad_norm": 0.5002845474985712,
+      "learning_rate": 4.5971456094387395e-06,
+      "loss": 0.4705,
+      "step": 9655
+    },
+    {
+      "epoch": 3.7814522104223265,
+      "grad_norm": 0.49887974718155487,
+      "learning_rate": 4.597060320773327e-06,
+      "loss": 0.4552,
+      "step": 9656
+    },
+    {
+      "epoch": 3.7818473697209187,
+      "grad_norm": 0.4844913686111668,
+      "learning_rate": 4.596975023871935e-06,
+      "loss": 0.4494,
+      "step": 9657
+    },
+    {
+      "epoch": 3.782242529019511,
+      "grad_norm": 0.49008324798281255,
+      "learning_rate": 4.596889718734898e-06,
+      "loss": 0.4534,
+      "step": 9658
+    },
+    {
+      "epoch": 3.782637688318103,
+      "grad_norm": 0.48905798875896944,
+      "learning_rate": 4.5968044053625525e-06,
+      "loss": 0.4652,
+      "step": 9659
+    },
+    {
+      "epoch": 3.7830328476166954,
+      "grad_norm": 0.48024108850578784,
+      "learning_rate": 4.596719083755231e-06,
+      "loss": 0.4592,
+      "step": 9660
+    },
+    {
+      "epoch": 3.7834280069152877,
+      "grad_norm": 0.4900301758693202,
+      "learning_rate": 4.596633753913272e-06,
+      "loss": 0.4565,
+      "step": 9661
+    },
+    {
+      "epoch": 3.78382316621388,
+      "grad_norm": 0.4802953667319636,
+      "learning_rate": 4.596548415837007e-06,
+      "loss": 0.4543,
+      "step": 9662
+    },
+    {
+      "epoch": 3.784218325512472,
+      "grad_norm": 0.4837608289095752,
+      "learning_rate": 4.596463069526775e-06,
+      "loss": 0.4546,
+      "step": 9663
+    },
+    {
+      "epoch": 3.7846134848110644,
+      "grad_norm": 0.5182415469829378,
+      "learning_rate": 4.596377714982907e-06,
+      "loss": 0.4844,
+      "step": 9664
+    },
+    {
+      "epoch": 3.7850086441096567,
+      "grad_norm": 0.4966888453091403,
+      "learning_rate": 4.5962923522057415e-06,
+      "loss": 0.4576,
+      "step": 9665
+    },
+    {
+      "epoch": 3.785403803408249,
+      "grad_norm": 0.49753630366467044,
+      "learning_rate": 4.596206981195611e-06,
+      "loss": 0.468,
+      "step": 9666
+    },
+    {
+      "epoch": 3.785798962706841,
+      "grad_norm": 0.48874836080231554,
+      "learning_rate": 4.596121601952854e-06,
+      "loss": 0.4405,
+      "step": 9667
+    },
+    {
+      "epoch": 3.7861941220054334,
+      "grad_norm": 0.5076579580062619,
+      "learning_rate": 4.596036214477804e-06,
+      "loss": 0.467,
+      "step": 9668
+    },
+    {
+      "epoch": 3.7865892813040256,
+      "grad_norm": 0.49135910265658483,
+      "learning_rate": 4.595950818770796e-06,
+      "loss": 0.4639,
+      "step": 9669
+    },
+    {
+      "epoch": 3.786984440602618,
+      "grad_norm": 0.4840783957984401,
+      "learning_rate": 4.595865414832166e-06,
+      "loss": 0.4864,
+      "step": 9670
+    },
+    {
+      "epoch": 3.78737959990121,
+      "grad_norm": 0.5059648003209108,
+      "learning_rate": 4.59578000266225e-06,
+      "loss": 0.4623,
+      "step": 9671
+    },
+    {
+      "epoch": 3.7877747591998023,
+      "grad_norm": 0.5155562419723234,
+      "learning_rate": 4.595694582261382e-06,
+      "loss": 0.4552,
+      "step": 9672
+    },
+    {
+      "epoch": 3.7881699184983946,
+      "grad_norm": 0.49286246868120626,
+      "learning_rate": 4.595609153629899e-06,
+      "loss": 0.4531,
+      "step": 9673
+    },
+    {
+      "epoch": 3.788565077796987,
+      "grad_norm": 0.5030907783859865,
+      "learning_rate": 4.5955237167681356e-06,
+      "loss": 0.4809,
+      "step": 9674
+    },
+    {
+      "epoch": 3.788960237095579,
+      "grad_norm": 0.48208063419912994,
+      "learning_rate": 4.595438271676427e-06,
+      "loss": 0.4621,
+      "step": 9675
+    },
+    {
+      "epoch": 3.7893553963941713,
+      "grad_norm": 0.48623500793987223,
+      "learning_rate": 4.595352818355109e-06,
+      "loss": 0.4628,
+      "step": 9676
+    },
+    {
+      "epoch": 3.7897505556927635,
+      "grad_norm": 0.5114092445947878,
+      "learning_rate": 4.595267356804518e-06,
+      "loss": 0.4678,
+      "step": 9677
+    },
+    {
+      "epoch": 3.7901457149913558,
+      "grad_norm": 0.4776370080310583,
+      "learning_rate": 4.595181887024989e-06,
+      "loss": 0.4466,
+      "step": 9678
+    },
+    {
+      "epoch": 3.790540874289948,
+      "grad_norm": 0.5818700802644523,
+      "learning_rate": 4.595096409016858e-06,
+      "loss": 0.4536,
+      "step": 9679
+    },
+    {
+      "epoch": 3.7909360335885403,
+      "grad_norm": 0.4740846191844771,
+      "learning_rate": 4.595010922780461e-06,
+      "loss": 0.4419,
+      "step": 9680
+    },
+    {
+      "epoch": 3.7913311928871325,
+      "grad_norm": 0.49737461925124105,
+      "learning_rate": 4.594925428316132e-06,
+      "loss": 0.458,
+      "step": 9681
+    },
+    {
+      "epoch": 3.7917263521857247,
+      "grad_norm": 0.5051660541685692,
+      "learning_rate": 4.594839925624209e-06,
+      "loss": 0.4409,
+      "step": 9682
+    },
+    {
+      "epoch": 3.792121511484317,
+      "grad_norm": 0.4984405654390054,
+      "learning_rate": 4.594754414705027e-06,
+      "loss": 0.4665,
+      "step": 9683
+    },
+    {
+      "epoch": 3.792516670782909,
+      "grad_norm": 0.48241664062480294,
+      "learning_rate": 4.594668895558921e-06,
+      "loss": 0.4474,
+      "step": 9684
+    },
+    {
+      "epoch": 3.7929118300815015,
+      "grad_norm": 0.5191185844568438,
+      "learning_rate": 4.594583368186228e-06,
+      "loss": 0.4513,
+      "step": 9685
+    },
+    {
+      "epoch": 3.7933069893800937,
+      "grad_norm": 0.48750409539612494,
+      "learning_rate": 4.594497832587283e-06,
+      "loss": 0.4593,
+      "step": 9686
+    },
+    {
+      "epoch": 3.793702148678686,
+      "grad_norm": 0.5133154479299648,
+      "learning_rate": 4.594412288762423e-06,
+      "loss": 0.4735,
+      "step": 9687
+    },
+    {
+      "epoch": 3.794097307977278,
+      "grad_norm": 0.47361385821885527,
+      "learning_rate": 4.594326736711983e-06,
+      "loss": 0.4412,
+      "step": 9688
+    },
+    {
+      "epoch": 3.7944924672758704,
+      "grad_norm": 0.48210172755908104,
+      "learning_rate": 4.5942411764362985e-06,
+      "loss": 0.4547,
+      "step": 9689
+    },
+    {
+      "epoch": 3.7948876265744627,
+      "grad_norm": 0.4960098779386097,
+      "learning_rate": 4.5941556079357076e-06,
+      "loss": 0.4625,
+      "step": 9690
+    },
+    {
+      "epoch": 3.795282785873055,
+      "grad_norm": 0.4903029773758132,
+      "learning_rate": 4.5940700312105444e-06,
+      "loss": 0.4712,
+      "step": 9691
+    },
+    {
+      "epoch": 3.795677945171647,
+      "grad_norm": 0.5136839009508477,
+      "learning_rate": 4.593984446261146e-06,
+      "loss": 0.473,
+      "step": 9692
+    },
+    {
+      "epoch": 3.7960731044702394,
+      "grad_norm": 0.4747156730400875,
+      "learning_rate": 4.5938988530878485e-06,
+      "loss": 0.4611,
+      "step": 9693
+    },
+    {
+      "epoch": 3.7964682637688316,
+      "grad_norm": 0.4870902649544296,
+      "learning_rate": 4.5938132516909865e-06,
+      "loss": 0.4524,
+      "step": 9694
+    },
+    {
+      "epoch": 3.796863423067424,
+      "grad_norm": 0.5010946794796335,
+      "learning_rate": 4.593727642070899e-06,
+      "loss": 0.4849,
+      "step": 9695
+    },
+    {
+      "epoch": 3.797258582366016,
+      "grad_norm": 0.49536194899113356,
+      "learning_rate": 4.593642024227919e-06,
+      "loss": 0.4797,
+      "step": 9696
+    },
+    {
+      "epoch": 3.7976537416646083,
+      "grad_norm": 0.4930240021909693,
+      "learning_rate": 4.593556398162386e-06,
+      "loss": 0.4596,
+      "step": 9697
+    },
+    {
+      "epoch": 3.7980489009632006,
+      "grad_norm": 0.4921928001724572,
+      "learning_rate": 4.5934707638746344e-06,
+      "loss": 0.4532,
+      "step": 9698
+    },
+    {
+      "epoch": 3.798444060261793,
+      "grad_norm": 0.4973760960380266,
+      "learning_rate": 4.5933851213650005e-06,
+      "loss": 0.462,
+      "step": 9699
+    },
+    {
+      "epoch": 3.798839219560385,
+      "grad_norm": 0.47361402891352083,
+      "learning_rate": 4.593299470633821e-06,
+      "loss": 0.4658,
+      "step": 9700
+    },
+    {
+      "epoch": 3.7992343788589773,
+      "grad_norm": 0.48887488140182345,
+      "learning_rate": 4.593213811681433e-06,
+      "loss": 0.455,
+      "step": 9701
+    },
+    {
+      "epoch": 3.7996295381575695,
+      "grad_norm": 0.4867609427093806,
+      "learning_rate": 4.593128144508171e-06,
+      "loss": 0.4464,
+      "step": 9702
+    },
+    {
+      "epoch": 3.800024697456162,
+      "grad_norm": 0.5026991707874338,
+      "learning_rate": 4.593042469114374e-06,
+      "loss": 0.4899,
+      "step": 9703
+    },
+    {
+      "epoch": 3.800419856754754,
+      "grad_norm": 0.47389953125387,
+      "learning_rate": 4.592956785500376e-06,
+      "loss": 0.4522,
+      "step": 9704
+    },
+    {
+      "epoch": 3.8008150160533463,
+      "grad_norm": 0.49268631953734526,
+      "learning_rate": 4.5928710936665156e-06,
+      "loss": 0.4552,
+      "step": 9705
+    },
+    {
+      "epoch": 3.8012101753519385,
+      "grad_norm": 0.4743740110089471,
+      "learning_rate": 4.592785393613128e-06,
+      "loss": 0.4472,
+      "step": 9706
+    },
+    {
+      "epoch": 3.8016053346505307,
+      "grad_norm": 0.4888493962366472,
+      "learning_rate": 4.59269968534055e-06,
+      "loss": 0.4569,
+      "step": 9707
+    },
+    {
+      "epoch": 3.802000493949123,
+      "grad_norm": 0.49012796067105,
+      "learning_rate": 4.592613968849119e-06,
+      "loss": 0.4644,
+      "step": 9708
+    },
+    {
+      "epoch": 3.8023956532477152,
+      "grad_norm": 0.48032600347511467,
+      "learning_rate": 4.592528244139171e-06,
+      "loss": 0.4587,
+      "step": 9709
+    },
+    {
+      "epoch": 3.8027908125463075,
+      "grad_norm": 0.49803327331966196,
+      "learning_rate": 4.5924425112110425e-06,
+      "loss": 0.4694,
+      "step": 9710
+    },
+    {
+      "epoch": 3.8031859718448997,
+      "grad_norm": 0.4876381618700325,
+      "learning_rate": 4.59235677006507e-06,
+      "loss": 0.4651,
+      "step": 9711
+    },
+    {
+      "epoch": 3.803581131143492,
+      "grad_norm": 0.4865814270109722,
+      "learning_rate": 4.592271020701591e-06,
+      "loss": 0.4516,
+      "step": 9712
+    },
+    {
+      "epoch": 3.8039762904420846,
+      "grad_norm": 0.47309099910390273,
+      "learning_rate": 4.592185263120942e-06,
+      "loss": 0.4569,
+      "step": 9713
+    },
+    {
+      "epoch": 3.804371449740677,
+      "grad_norm": 0.4828676962421864,
+      "learning_rate": 4.592099497323459e-06,
+      "loss": 0.4503,
+      "step": 9714
+    },
+    {
+      "epoch": 3.804766609039269,
+      "grad_norm": 0.5044657636177922,
+      "learning_rate": 4.592013723309481e-06,
+      "loss": 0.4534,
+      "step": 9715
+    },
+    {
+      "epoch": 3.8051617683378613,
+      "grad_norm": 0.4741720246957415,
+      "learning_rate": 4.591927941079341e-06,
+      "loss": 0.4707,
+      "step": 9716
+    },
+    {
+      "epoch": 3.8055569276364536,
+      "grad_norm": 0.49614095314653145,
+      "learning_rate": 4.591842150633381e-06,
+      "loss": 0.4493,
+      "step": 9717
+    },
+    {
+      "epoch": 3.805952086935046,
+      "grad_norm": 0.4882532495153286,
+      "learning_rate": 4.5917563519719334e-06,
+      "loss": 0.4653,
+      "step": 9718
+    },
+    {
+      "epoch": 3.806347246233638,
+      "grad_norm": 0.4949527043726383,
+      "learning_rate": 4.5916705450953384e-06,
+      "loss": 0.4679,
+      "step": 9719
+    },
+    {
+      "epoch": 3.8067424055322303,
+      "grad_norm": 0.5602238728690052,
+      "learning_rate": 4.591584730003931e-06,
+      "loss": 0.4557,
+      "step": 9720
+    },
+    {
+      "epoch": 3.8071375648308226,
+      "grad_norm": 0.4912568517475018,
+      "learning_rate": 4.591498906698048e-06,
+      "loss": 0.4709,
+      "step": 9721
+    },
+    {
+      "epoch": 3.807532724129415,
+      "grad_norm": 0.4813539895906584,
+      "learning_rate": 4.591413075178029e-06,
+      "loss": 0.4531,
+      "step": 9722
+    },
+    {
+      "epoch": 3.807927883428007,
+      "grad_norm": 0.49909964708037335,
+      "learning_rate": 4.591327235444209e-06,
+      "loss": 0.4624,
+      "step": 9723
+    },
+    {
+      "epoch": 3.8083230427265993,
+      "grad_norm": 0.4814723511010151,
+      "learning_rate": 4.591241387496925e-06,
+      "loss": 0.4705,
+      "step": 9724
+    },
+    {
+      "epoch": 3.8087182020251915,
+      "grad_norm": 0.4827612230715431,
+      "learning_rate": 4.591155531336514e-06,
+      "loss": 0.4532,
+      "step": 9725
+    },
+    {
+      "epoch": 3.8091133613237838,
+      "grad_norm": 0.4938055389724764,
+      "learning_rate": 4.591069666963315e-06,
+      "loss": 0.4554,
+      "step": 9726
+    },
+    {
+      "epoch": 3.809508520622376,
+      "grad_norm": 0.4852708740829828,
+      "learning_rate": 4.590983794377664e-06,
+      "loss": 0.4645,
+      "step": 9727
+    },
+    {
+      "epoch": 3.8099036799209682,
+      "grad_norm": 0.48556679669930314,
+      "learning_rate": 4.590897913579898e-06,
+      "loss": 0.4488,
+      "step": 9728
+    },
+    {
+      "epoch": 3.8102988392195605,
+      "grad_norm": 0.48745710424116256,
+      "learning_rate": 4.590812024570355e-06,
+      "loss": 0.4391,
+      "step": 9729
+    },
+    {
+      "epoch": 3.8106939985181527,
+      "grad_norm": 0.4941001569984601,
+      "learning_rate": 4.590726127349372e-06,
+      "loss": 0.4688,
+      "step": 9730
+    },
+    {
+      "epoch": 3.811089157816745,
+      "grad_norm": 0.5066586936353458,
+      "learning_rate": 4.5906402219172865e-06,
+      "loss": 0.4672,
+      "step": 9731
+    },
+    {
+      "epoch": 3.811484317115337,
+      "grad_norm": 0.49499770510893454,
+      "learning_rate": 4.590554308274435e-06,
+      "loss": 0.4665,
+      "step": 9732
+    },
+    {
+      "epoch": 3.8118794764139294,
+      "grad_norm": 0.48385700939951193,
+      "learning_rate": 4.5904683864211564e-06,
+      "loss": 0.4287,
+      "step": 9733
+    },
+    {
+      "epoch": 3.8122746357125217,
+      "grad_norm": 0.5006235625670952,
+      "learning_rate": 4.590382456357787e-06,
+      "loss": 0.4763,
+      "step": 9734
+    },
+    {
+      "epoch": 3.812669795011114,
+      "grad_norm": 0.5063019446978001,
+      "learning_rate": 4.5902965180846645e-06,
+      "loss": 0.4849,
+      "step": 9735
+    },
+    {
+      "epoch": 3.813064954309706,
+      "grad_norm": 0.47648185972596263,
+      "learning_rate": 4.590210571602127e-06,
+      "loss": 0.4518,
+      "step": 9736
+    },
+    {
+      "epoch": 3.8134601136082984,
+      "grad_norm": 0.48236587628734506,
+      "learning_rate": 4.590124616910511e-06,
+      "loss": 0.4613,
+      "step": 9737
+    },
+    {
+      "epoch": 3.8138552729068906,
+      "grad_norm": 0.4854824711766063,
+      "learning_rate": 4.590038654010155e-06,
+      "loss": 0.4611,
+      "step": 9738
+    },
+    {
+      "epoch": 3.814250432205483,
+      "grad_norm": 0.47570054733499323,
+      "learning_rate": 4.5899526829013966e-06,
+      "loss": 0.4569,
+      "step": 9739
+    },
+    {
+      "epoch": 3.814645591504075,
+      "grad_norm": 0.47994844957076466,
+      "learning_rate": 4.589866703584573e-06,
+      "loss": 0.465,
+      "step": 9740
+    },
+    {
+      "epoch": 3.8150407508026674,
+      "grad_norm": 0.49181023857768696,
+      "learning_rate": 4.589780716060022e-06,
+      "loss": 0.4519,
+      "step": 9741
+    },
+    {
+      "epoch": 3.8154359101012596,
+      "grad_norm": 0.4784710573418661,
+      "learning_rate": 4.589694720328081e-06,
+      "loss": 0.4663,
+      "step": 9742
+    },
+    {
+      "epoch": 3.815831069399852,
+      "grad_norm": 0.4852518817476975,
+      "learning_rate": 4.589608716389088e-06,
+      "loss": 0.4718,
+      "step": 9743
+    },
+    {
+      "epoch": 3.816226228698444,
+      "grad_norm": 0.4869284641618619,
+      "learning_rate": 4.589522704243381e-06,
+      "loss": 0.4675,
+      "step": 9744
+    },
+    {
+      "epoch": 3.8166213879970363,
+      "grad_norm": 0.4895242207948993,
+      "learning_rate": 4.589436683891299e-06,
+      "loss": 0.4726,
+      "step": 9745
+    },
+    {
+      "epoch": 3.8170165472956286,
+      "grad_norm": 0.49471582563789906,
+      "learning_rate": 4.589350655333177e-06,
+      "loss": 0.4548,
+      "step": 9746
+    },
+    {
+      "epoch": 3.817411706594221,
+      "grad_norm": 0.48302731444823493,
+      "learning_rate": 4.589264618569355e-06,
+      "loss": 0.453,
+      "step": 9747
+    },
+    {
+      "epoch": 3.817806865892813,
+      "grad_norm": 0.4991821360059649,
+      "learning_rate": 4.5891785736001696e-06,
+      "loss": 0.4618,
+      "step": 9748
+    },
+    {
+      "epoch": 3.8182020251914053,
+      "grad_norm": 0.5035434112433709,
+      "learning_rate": 4.58909252042596e-06,
+      "loss": 0.4794,
+      "step": 9749
+    },
+    {
+      "epoch": 3.8185971844899975,
+      "grad_norm": 0.5381210527580598,
+      "learning_rate": 4.589006459047063e-06,
+      "loss": 0.457,
+      "step": 9750
+    },
+    {
+      "epoch": 3.8189923437885898,
+      "grad_norm": 0.5042220819531829,
+      "learning_rate": 4.588920389463817e-06,
+      "loss": 0.4713,
+      "step": 9751
+    },
+    {
+      "epoch": 3.819387503087182,
+      "grad_norm": 0.49824542001689043,
+      "learning_rate": 4.5888343116765604e-06,
+      "loss": 0.459,
+      "step": 9752
+    },
+    {
+      "epoch": 3.8197826623857742,
+      "grad_norm": 0.4812925277928597,
+      "learning_rate": 4.5887482256856305e-06,
+      "loss": 0.4468,
+      "step": 9753
+    },
+    {
+      "epoch": 3.8201778216843665,
+      "grad_norm": 0.4867523713344396,
+      "learning_rate": 4.588662131491367e-06,
+      "loss": 0.4725,
+      "step": 9754
+    },
+    {
+      "epoch": 3.8205729809829587,
+      "grad_norm": 0.5045119465234934,
+      "learning_rate": 4.588576029094107e-06,
+      "loss": 0.4651,
+      "step": 9755
+    },
+    {
+      "epoch": 3.820968140281551,
+      "grad_norm": 0.49383831121424904,
+      "learning_rate": 4.588489918494188e-06,
+      "loss": 0.4575,
+      "step": 9756
+    },
+    {
+      "epoch": 3.821363299580143,
+      "grad_norm": 0.49047441119572965,
+      "learning_rate": 4.588403799691949e-06,
+      "loss": 0.4587,
+      "step": 9757
+    },
+    {
+      "epoch": 3.8217584588787354,
+      "grad_norm": 0.513029731134795,
+      "learning_rate": 4.5883176726877276e-06,
+      "loss": 0.4473,
+      "step": 9758
+    },
+    {
+      "epoch": 3.8221536181773277,
+      "grad_norm": 0.5013246021960757,
+      "learning_rate": 4.588231537481863e-06,
+      "loss": 0.4535,
+      "step": 9759
+    },
+    {
+      "epoch": 3.82254877747592,
+      "grad_norm": 0.4916166616553396,
+      "learning_rate": 4.5881453940746925e-06,
+      "loss": 0.4626,
+      "step": 9760
+    },
+    {
+      "epoch": 3.822943936774512,
+      "grad_norm": 0.49314443616382275,
+      "learning_rate": 4.588059242466555e-06,
+      "loss": 0.4711,
+      "step": 9761
+    },
+    {
+      "epoch": 3.8233390960731044,
+      "grad_norm": 0.5008098059952933,
+      "learning_rate": 4.587973082657789e-06,
+      "loss": 0.445,
+      "step": 9762
+    },
+    {
+      "epoch": 3.8237342553716966,
+      "grad_norm": 0.4838807035601777,
+      "learning_rate": 4.587886914648733e-06,
+      "loss": 0.456,
+      "step": 9763
+    },
+    {
+      "epoch": 3.824129414670289,
+      "grad_norm": 0.48847805670933026,
+      "learning_rate": 4.587800738439725e-06,
+      "loss": 0.461,
+      "step": 9764
+    },
+    {
+      "epoch": 3.824524573968881,
+      "grad_norm": 0.4836684897311038,
+      "learning_rate": 4.587714554031103e-06,
+      "loss": 0.4676,
+      "step": 9765
+    },
+    {
+      "epoch": 3.8249197332674734,
+      "grad_norm": 0.4744461410812829,
+      "learning_rate": 4.587628361423205e-06,
+      "loss": 0.4467,
+      "step": 9766
+    },
+    {
+      "epoch": 3.8253148925660656,
+      "grad_norm": 0.5046875959722604,
+      "learning_rate": 4.587542160616372e-06,
+      "loss": 0.4594,
+      "step": 9767
+    },
+    {
+      "epoch": 3.825710051864658,
+      "grad_norm": 0.4894718925080993,
+      "learning_rate": 4.587455951610941e-06,
+      "loss": 0.4596,
+      "step": 9768
+    },
+    {
+      "epoch": 3.82610521116325,
+      "grad_norm": 0.6451510629529777,
+      "learning_rate": 4.58736973440725e-06,
+      "loss": 0.456,
+      "step": 9769
+    },
+    {
+      "epoch": 3.8265003704618423,
+      "grad_norm": 0.49698088184421574,
+      "learning_rate": 4.587283509005638e-06,
+      "loss": 0.4729,
+      "step": 9770
+    },
+    {
+      "epoch": 3.8268955297604346,
+      "grad_norm": 0.4744119187512547,
+      "learning_rate": 4.587197275406444e-06,
+      "loss": 0.4611,
+      "step": 9771
+    },
+    {
+      "epoch": 3.827290689059027,
+      "grad_norm": 0.48047611658306816,
+      "learning_rate": 4.587111033610007e-06,
+      "loss": 0.4404,
+      "step": 9772
+    },
+    {
+      "epoch": 3.827685848357619,
+      "grad_norm": 0.5074066632267403,
+      "learning_rate": 4.587024783616665e-06,
+      "loss": 0.4595,
+      "step": 9773
+    },
+    {
+      "epoch": 3.8280810076562113,
+      "grad_norm": 0.4886772671378936,
+      "learning_rate": 4.586938525426756e-06,
+      "loss": 0.4584,
+      "step": 9774
+    },
+    {
+      "epoch": 3.8284761669548035,
+      "grad_norm": 0.4855093381504905,
+      "learning_rate": 4.586852259040621e-06,
+      "loss": 0.4727,
+      "step": 9775
+    },
+    {
+      "epoch": 3.8288713262533958,
+      "grad_norm": 0.47748604681856477,
+      "learning_rate": 4.586765984458597e-06,
+      "loss": 0.441,
+      "step": 9776
+    },
+    {
+      "epoch": 3.829266485551988,
+      "grad_norm": 0.48420777791114294,
+      "learning_rate": 4.586679701681024e-06,
+      "loss": 0.4587,
+      "step": 9777
+    },
+    {
+      "epoch": 3.8296616448505802,
+      "grad_norm": 0.4833361501646878,
+      "learning_rate": 4.586593410708239e-06,
+      "loss": 0.4474,
+      "step": 9778
+    },
+    {
+      "epoch": 3.8300568041491725,
+      "grad_norm": 0.49869671048495123,
+      "learning_rate": 4.586507111540583e-06,
+      "loss": 0.4707,
+      "step": 9779
+    },
+    {
+      "epoch": 3.8304519634477647,
+      "grad_norm": 0.4879810325262599,
+      "learning_rate": 4.586420804178394e-06,
+      "loss": 0.4575,
+      "step": 9780
+    },
+    {
+      "epoch": 3.830847122746357,
+      "grad_norm": 0.48913117133273337,
+      "learning_rate": 4.58633448862201e-06,
+      "loss": 0.4654,
+      "step": 9781
+    },
+    {
+      "epoch": 3.831242282044949,
+      "grad_norm": 0.5998061332057096,
+      "learning_rate": 4.586248164871773e-06,
+      "loss": 0.4651,
+      "step": 9782
+    },
+    {
+      "epoch": 3.8316374413435415,
+      "grad_norm": 0.5325641228028988,
+      "learning_rate": 4.5861618329280185e-06,
+      "loss": 0.4674,
+      "step": 9783
+    },
+    {
+      "epoch": 3.832032600642134,
+      "grad_norm": 0.48101331017897647,
+      "learning_rate": 4.586075492791088e-06,
+      "loss": 0.4709,
+      "step": 9784
+    },
+    {
+      "epoch": 3.8324277599407264,
+      "grad_norm": 0.48810663701291834,
+      "learning_rate": 4.585989144461319e-06,
+      "loss": 0.4682,
+      "step": 9785
+    },
+    {
+      "epoch": 3.8328229192393186,
+      "grad_norm": 0.49088685212482186,
+      "learning_rate": 4.585902787939052e-06,
+      "loss": 0.4667,
+      "step": 9786
+    },
+    {
+      "epoch": 3.833218078537911,
+      "grad_norm": 0.4816964576299058,
+      "learning_rate": 4.585816423224625e-06,
+      "loss": 0.4619,
+      "step": 9787
+    },
+    {
+      "epoch": 3.833613237836503,
+      "grad_norm": 0.49202140202010347,
+      "learning_rate": 4.585730050318378e-06,
+      "loss": 0.4692,
+      "step": 9788
+    },
+    {
+      "epoch": 3.8340083971350953,
+      "grad_norm": 0.49988725949905755,
+      "learning_rate": 4.585643669220648e-06,
+      "loss": 0.4513,
+      "step": 9789
+    },
+    {
+      "epoch": 3.8344035564336876,
+      "grad_norm": 0.4826706645801356,
+      "learning_rate": 4.585557279931779e-06,
+      "loss": 0.4474,
+      "step": 9790
+    },
+    {
+      "epoch": 3.83479871573228,
+      "grad_norm": 0.49913183679571577,
+      "learning_rate": 4.585470882452106e-06,
+      "loss": 0.4592,
+      "step": 9791
+    },
+    {
+      "epoch": 3.835193875030872,
+      "grad_norm": 0.5102617938895149,
+      "learning_rate": 4.58538447678197e-06,
+      "loss": 0.4622,
+      "step": 9792
+    },
+    {
+      "epoch": 3.8355890343294643,
+      "grad_norm": 0.5093288577737429,
+      "learning_rate": 4.58529806292171e-06,
+      "loss": 0.4876,
+      "step": 9793
+    },
+    {
+      "epoch": 3.8359841936280565,
+      "grad_norm": 0.4847749359173575,
+      "learning_rate": 4.585211640871665e-06,
+      "loss": 0.4527,
+      "step": 9794
+    },
+    {
+      "epoch": 3.836379352926649,
+      "grad_norm": 0.4967128561121068,
+      "learning_rate": 4.5851252106321755e-06,
+      "loss": 0.4512,
+      "step": 9795
+    },
+    {
+      "epoch": 3.836774512225241,
+      "grad_norm": 0.5122771054364694,
+      "learning_rate": 4.585038772203581e-06,
+      "loss": 0.4581,
+      "step": 9796
+    },
+    {
+      "epoch": 3.8371696715238333,
+      "grad_norm": 0.4778663343535554,
+      "learning_rate": 4.584952325586219e-06,
+      "loss": 0.455,
+      "step": 9797
+    },
+    {
+      "epoch": 3.8375648308224255,
+      "grad_norm": 0.4730286915929293,
+      "learning_rate": 4.584865870780431e-06,
+      "loss": 0.4438,
+      "step": 9798
+    },
+    {
+      "epoch": 3.8379599901210177,
+      "grad_norm": 0.49474641210541287,
+      "learning_rate": 4.584779407786556e-06,
+      "loss": 0.4647,
+      "step": 9799
+    },
+    {
+      "epoch": 3.83835514941961,
+      "grad_norm": 0.4979187357021784,
+      "learning_rate": 4.5846929366049316e-06,
+      "loss": 0.469,
+      "step": 9800
+    },
+    {
+      "epoch": 3.838750308718202,
+      "grad_norm": 0.4969648880958739,
+      "learning_rate": 4.584606457235901e-06,
+      "loss": 0.472,
+      "step": 9801
+    },
+    {
+      "epoch": 3.8391454680167945,
+      "grad_norm": 0.48419610177762074,
+      "learning_rate": 4.584519969679803e-06,
+      "loss": 0.4609,
+      "step": 9802
+    },
+    {
+      "epoch": 3.8395406273153867,
+      "grad_norm": 0.47519923964504873,
+      "learning_rate": 4.584433473936975e-06,
+      "loss": 0.4407,
+      "step": 9803
+    },
+    {
+      "epoch": 3.839935786613979,
+      "grad_norm": 0.5036626686511386,
+      "learning_rate": 4.584346970007758e-06,
+      "loss": 0.4666,
+      "step": 9804
+    },
+    {
+      "epoch": 3.840330945912571,
+      "grad_norm": 0.49885414130181815,
+      "learning_rate": 4.584260457892492e-06,
+      "loss": 0.4486,
+      "step": 9805
+    },
+    {
+      "epoch": 3.8407261052111634,
+      "grad_norm": 0.4973984116213614,
+      "learning_rate": 4.584173937591516e-06,
+      "loss": 0.4716,
+      "step": 9806
+    },
+    {
+      "epoch": 3.8411212645097557,
+      "grad_norm": 0.47438026051623494,
+      "learning_rate": 4.584087409105171e-06,
+      "loss": 0.4508,
+      "step": 9807
+    },
+    {
+      "epoch": 3.841516423808348,
+      "grad_norm": 0.48007747354455976,
+      "learning_rate": 4.5840008724337955e-06,
+      "loss": 0.4631,
+      "step": 9808
+    },
+    {
+      "epoch": 3.84191158310694,
+      "grad_norm": 0.587833445348625,
+      "learning_rate": 4.583914327577731e-06,
+      "loss": 0.4741,
+      "step": 9809
+    },
+    {
+      "epoch": 3.8423067424055324,
+      "grad_norm": 0.499951286009737,
+      "learning_rate": 4.583827774537316e-06,
+      "loss": 0.456,
+      "step": 9810
+    },
+    {
+      "epoch": 3.8427019017041246,
+      "grad_norm": 0.4848679924648173,
+      "learning_rate": 4.583741213312891e-06,
+      "loss": 0.4729,
+      "step": 9811
+    },
+    {
+      "epoch": 3.843097061002717,
+      "grad_norm": 0.48840107615604933,
+      "learning_rate": 4.583654643904796e-06,
+      "loss": 0.4642,
+      "step": 9812
+    },
+    {
+      "epoch": 3.843492220301309,
+      "grad_norm": 0.4968970099252351,
+      "learning_rate": 4.5835680663133705e-06,
+      "loss": 0.4621,
+      "step": 9813
+    },
+    {
+      "epoch": 3.8438873795999013,
+      "grad_norm": 0.47240176053138544,
+      "learning_rate": 4.583481480538955e-06,
+      "loss": 0.458,
+      "step": 9814
+    },
+    {
+      "epoch": 3.8442825388984936,
+      "grad_norm": 0.4873267364456109,
+      "learning_rate": 4.583394886581889e-06,
+      "loss": 0.4614,
+      "step": 9815
+    },
+    {
+      "epoch": 3.844677698197086,
+      "grad_norm": 0.4813398493399383,
+      "learning_rate": 4.5833082844425135e-06,
+      "loss": 0.4543,
+      "step": 9816
+    },
+    {
+      "epoch": 3.845072857495678,
+      "grad_norm": 0.49742091745038736,
+      "learning_rate": 4.583221674121167e-06,
+      "loss": 0.4437,
+      "step": 9817
+    },
+    {
+      "epoch": 3.8454680167942703,
+      "grad_norm": 0.4937773614181849,
+      "learning_rate": 4.5831350556181934e-06,
+      "loss": 0.4668,
+      "step": 9818
+    },
+    {
+      "epoch": 3.8458631760928625,
+      "grad_norm": 0.4807185417064237,
+      "learning_rate": 4.583048428933928e-06,
+      "loss": 0.4303,
+      "step": 9819
+    },
+    {
+      "epoch": 3.846258335391455,
+      "grad_norm": 0.49996366474147874,
+      "learning_rate": 4.582961794068714e-06,
+      "loss": 0.463,
+      "step": 9820
+    },
+    {
+      "epoch": 3.846653494690047,
+      "grad_norm": 0.4767782183375261,
+      "learning_rate": 4.582875151022891e-06,
+      "loss": 0.4725,
+      "step": 9821
+    },
+    {
+      "epoch": 3.8470486539886393,
+      "grad_norm": 0.48139899842359823,
+      "learning_rate": 4.582788499796798e-06,
+      "loss": 0.4533,
+      "step": 9822
+    },
+    {
+      "epoch": 3.8474438132872315,
+      "grad_norm": 0.4963731934676299,
+      "learning_rate": 4.582701840390778e-06,
+      "loss": 0.4667,
+      "step": 9823
+    },
+    {
+      "epoch": 3.8478389725858237,
+      "grad_norm": 0.48437203938631246,
+      "learning_rate": 4.5826151728051696e-06,
+      "loss": 0.4576,
+      "step": 9824
+    },
+    {
+      "epoch": 3.848234131884416,
+      "grad_norm": 0.4788898724749113,
+      "learning_rate": 4.582528497040313e-06,
+      "loss": 0.4588,
+      "step": 9825
+    },
+    {
+      "epoch": 3.8486292911830082,
+      "grad_norm": 0.4864655138587333,
+      "learning_rate": 4.5824418130965485e-06,
+      "loss": 0.4664,
+      "step": 9826
+    },
+    {
+      "epoch": 3.8490244504816005,
+      "grad_norm": 0.491011327584925,
+      "learning_rate": 4.582355120974218e-06,
+      "loss": 0.443,
+      "step": 9827
+    },
+    {
+      "epoch": 3.8494196097801927,
+      "grad_norm": 0.5037804637537733,
+      "learning_rate": 4.582268420673661e-06,
+      "loss": 0.496,
+      "step": 9828
+    },
+    {
+      "epoch": 3.849814769078785,
+      "grad_norm": 0.4911686450431375,
+      "learning_rate": 4.582181712195218e-06,
+      "loss": 0.4648,
+      "step": 9829
+    },
+    {
+      "epoch": 3.850209928377377,
+      "grad_norm": 0.4885405950646796,
+      "learning_rate": 4.582094995539229e-06,
+      "loss": 0.4663,
+      "step": 9830
+    },
+    {
+      "epoch": 3.8506050876759694,
+      "grad_norm": 0.46999551043751275,
+      "learning_rate": 4.582008270706035e-06,
+      "loss": 0.4411,
+      "step": 9831
+    },
+    {
+      "epoch": 3.8510002469745617,
+      "grad_norm": 0.4882884684859506,
+      "learning_rate": 4.581921537695978e-06,
+      "loss": 0.4626,
+      "step": 9832
+    },
+    {
+      "epoch": 3.851395406273154,
+      "grad_norm": 0.4913585267933321,
+      "learning_rate": 4.581834796509397e-06,
+      "loss": 0.457,
+      "step": 9833
+    },
+    {
+      "epoch": 3.851790565571746,
+      "grad_norm": 0.4659923322842953,
+      "learning_rate": 4.581748047146633e-06,
+      "loss": 0.4434,
+      "step": 9834
+    },
+    {
+      "epoch": 3.8521857248703384,
+      "grad_norm": 0.49559627762410036,
+      "learning_rate": 4.581661289608027e-06,
+      "loss": 0.4724,
+      "step": 9835
+    },
+    {
+      "epoch": 3.8525808841689306,
+      "grad_norm": 0.4881260286548596,
+      "learning_rate": 4.581574523893919e-06,
+      "loss": 0.4519,
+      "step": 9836
+    },
+    {
+      "epoch": 3.852976043467523,
+      "grad_norm": 0.488542461742271,
+      "learning_rate": 4.581487750004651e-06,
+      "loss": 0.4687,
+      "step": 9837
+    },
+    {
+      "epoch": 3.853371202766115,
+      "grad_norm": 0.47946619013094255,
+      "learning_rate": 4.581400967940562e-06,
+      "loss": 0.4507,
+      "step": 9838
+    },
+    {
+      "epoch": 3.8537663620647074,
+      "grad_norm": 0.49113873280049447,
+      "learning_rate": 4.581314177701994e-06,
+      "loss": 0.4652,
+      "step": 9839
+    },
+    {
+      "epoch": 3.8541615213632996,
+      "grad_norm": 0.506053760875361,
+      "learning_rate": 4.581227379289288e-06,
+      "loss": 0.4565,
+      "step": 9840
+    },
+    {
+      "epoch": 3.854556680661892,
+      "grad_norm": 0.498570278337444,
+      "learning_rate": 4.581140572702785e-06,
+      "loss": 0.4572,
+      "step": 9841
+    },
+    {
+      "epoch": 3.854951839960484,
+      "grad_norm": 0.4750541834581741,
+      "learning_rate": 4.5810537579428255e-06,
+      "loss": 0.4602,
+      "step": 9842
+    },
+    {
+      "epoch": 3.8553469992590763,
+      "grad_norm": 0.48656050386521776,
+      "learning_rate": 4.580966935009751e-06,
+      "loss": 0.4811,
+      "step": 9843
+    },
+    {
+      "epoch": 3.8557421585576686,
+      "grad_norm": 0.48249156384469394,
+      "learning_rate": 4.580880103903901e-06,
+      "loss": 0.4764,
+      "step": 9844
+    },
+    {
+      "epoch": 3.856137317856261,
+      "grad_norm": 0.4880019438321168,
+      "learning_rate": 4.580793264625618e-06,
+      "loss": 0.4573,
+      "step": 9845
+    },
+    {
+      "epoch": 3.856532477154853,
+      "grad_norm": 0.5055590149018785,
+      "learning_rate": 4.5807064171752426e-06,
+      "loss": 0.4484,
+      "step": 9846
+    },
+    {
+      "epoch": 3.8569276364534453,
+      "grad_norm": 0.48601910111621366,
+      "learning_rate": 4.580619561553116e-06,
+      "loss": 0.4391,
+      "step": 9847
+    },
+    {
+      "epoch": 3.8573227957520375,
+      "grad_norm": 0.5571216265744174,
+      "learning_rate": 4.5805326977595784e-06,
+      "loss": 0.4402,
+      "step": 9848
+    },
+    {
+      "epoch": 3.8577179550506298,
+      "grad_norm": 0.49789507388969795,
+      "learning_rate": 4.5804458257949725e-06,
+      "loss": 0.4753,
+      "step": 9849
+    },
+    {
+      "epoch": 3.858113114349222,
+      "grad_norm": 0.493498154898902,
+      "learning_rate": 4.580358945659639e-06,
+      "loss": 0.4788,
+      "step": 9850
+    },
+    {
+      "epoch": 3.8585082736478142,
+      "grad_norm": 0.48765613657828505,
+      "learning_rate": 4.580272057353918e-06,
+      "loss": 0.4624,
+      "step": 9851
+    },
+    {
+      "epoch": 3.8589034329464065,
+      "grad_norm": 0.4844714607389518,
+      "learning_rate": 4.580185160878151e-06,
+      "loss": 0.459,
+      "step": 9852
+    },
+    {
+      "epoch": 3.8592985922449987,
+      "grad_norm": 0.497325111713686,
+      "learning_rate": 4.580098256232681e-06,
+      "loss": 0.4842,
+      "step": 9853
+    },
+    {
+      "epoch": 3.859693751543591,
+      "grad_norm": 0.48442884762655214,
+      "learning_rate": 4.5800113434178485e-06,
+      "loss": 0.4475,
+      "step": 9854
+    },
+    {
+      "epoch": 3.860088910842183,
+      "grad_norm": 0.4942069483791756,
+      "learning_rate": 4.579924422433993e-06,
+      "loss": 0.458,
+      "step": 9855
+    },
+    {
+      "epoch": 3.8604840701407754,
+      "grad_norm": 0.48273095849038933,
+      "learning_rate": 4.579837493281459e-06,
+      "loss": 0.459,
+      "step": 9856
+    },
+    {
+      "epoch": 3.8608792294393677,
+      "grad_norm": 0.48744132992622485,
+      "learning_rate": 4.579750555960585e-06,
+      "loss": 0.466,
+      "step": 9857
+    },
+    {
+      "epoch": 3.86127438873796,
+      "grad_norm": 0.4947598696132769,
+      "learning_rate": 4.5796636104717155e-06,
+      "loss": 0.4701,
+      "step": 9858
+    },
+    {
+      "epoch": 3.861669548036552,
+      "grad_norm": 0.4856546204635866,
+      "learning_rate": 4.579576656815188e-06,
+      "loss": 0.4595,
+      "step": 9859
+    },
+    {
+      "epoch": 3.8620647073351444,
+      "grad_norm": 0.484668155652294,
+      "learning_rate": 4.579489694991347e-06,
+      "loss": 0.452,
+      "step": 9860
+    },
+    {
+      "epoch": 3.8624598666337366,
+      "grad_norm": 0.49823648475531956,
+      "learning_rate": 4.579402725000534e-06,
+      "loss": 0.4525,
+      "step": 9861
+    },
+    {
+      "epoch": 3.862855025932329,
+      "grad_norm": 0.4879242829207528,
+      "learning_rate": 4.579315746843088e-06,
+      "loss": 0.4682,
+      "step": 9862
+    },
+    {
+      "epoch": 3.863250185230921,
+      "grad_norm": 0.5043668954023681,
+      "learning_rate": 4.579228760519354e-06,
+      "loss": 0.4696,
+      "step": 9863
+    },
+    {
+      "epoch": 3.8636453445295134,
+      "grad_norm": 0.47800525043818226,
+      "learning_rate": 4.579141766029672e-06,
+      "loss": 0.4392,
+      "step": 9864
+    },
+    {
+      "epoch": 3.8640405038281056,
+      "grad_norm": 0.4946490914928116,
+      "learning_rate": 4.579054763374383e-06,
+      "loss": 0.4515,
+      "step": 9865
+    },
+    {
+      "epoch": 3.864435663126698,
+      "grad_norm": 0.48212116695919593,
+      "learning_rate": 4.578967752553829e-06,
+      "loss": 0.4771,
+      "step": 9866
+    },
+    {
+      "epoch": 3.86483082242529,
+      "grad_norm": 0.49582742338022634,
+      "learning_rate": 4.578880733568353e-06,
+      "loss": 0.4477,
+      "step": 9867
+    },
+    {
+      "epoch": 3.8652259817238823,
+      "grad_norm": 0.49339493973167836,
+      "learning_rate": 4.578793706418295e-06,
+      "loss": 0.4487,
+      "step": 9868
+    },
+    {
+      "epoch": 3.8656211410224746,
+      "grad_norm": 0.4929894857650072,
+      "learning_rate": 4.578706671103998e-06,
+      "loss": 0.4811,
+      "step": 9869
+    },
+    {
+      "epoch": 3.866016300321067,
+      "grad_norm": 0.5010690182198145,
+      "learning_rate": 4.578619627625803e-06,
+      "loss": 0.4644,
+      "step": 9870
+    },
+    {
+      "epoch": 3.866411459619659,
+      "grad_norm": 0.48769438400413784,
+      "learning_rate": 4.578532575984053e-06,
+      "loss": 0.4517,
+      "step": 9871
+    },
+    {
+      "epoch": 3.8668066189182513,
+      "grad_norm": 0.49383384006167214,
+      "learning_rate": 4.5784455161790895e-06,
+      "loss": 0.4656,
+      "step": 9872
+    },
+    {
+      "epoch": 3.8672017782168435,
+      "grad_norm": 0.4810665135549552,
+      "learning_rate": 4.578358448211253e-06,
+      "loss": 0.4454,
+      "step": 9873
+    },
+    {
+      "epoch": 3.8675969375154358,
+      "grad_norm": 0.48861310498403004,
+      "learning_rate": 4.578271372080888e-06,
+      "loss": 0.4605,
+      "step": 9874
+    },
+    {
+      "epoch": 3.867992096814028,
+      "grad_norm": 0.49304147135775594,
+      "learning_rate": 4.5781842877883335e-06,
+      "loss": 0.4559,
+      "step": 9875
+    },
+    {
+      "epoch": 3.8683872561126202,
+      "grad_norm": 0.4901516373069663,
+      "learning_rate": 4.578097195333935e-06,
+      "loss": 0.4488,
+      "step": 9876
+    },
+    {
+      "epoch": 3.8687824154112125,
+      "grad_norm": 0.47112670377763216,
+      "learning_rate": 4.57801009471803e-06,
+      "loss": 0.4447,
+      "step": 9877
+    },
+    {
+      "epoch": 3.8691775747098047,
+      "grad_norm": 0.48837740071566554,
+      "learning_rate": 4.577922985940965e-06,
+      "loss": 0.457,
+      "step": 9878
+    },
+    {
+      "epoch": 3.869572734008397,
+      "grad_norm": 0.4965735724057039,
+      "learning_rate": 4.57783586900308e-06,
+      "loss": 0.4508,
+      "step": 9879
+    },
+    {
+      "epoch": 3.869967893306989,
+      "grad_norm": 0.4896825295302401,
+      "learning_rate": 4.577748743904717e-06,
+      "loss": 0.4706,
+      "step": 9880
+    },
+    {
+      "epoch": 3.8703630526055814,
+      "grad_norm": 0.48477203855856776,
+      "learning_rate": 4.577661610646219e-06,
+      "loss": 0.4486,
+      "step": 9881
+    },
+    {
+      "epoch": 3.8707582119041737,
+      "grad_norm": 0.4772573464731468,
+      "learning_rate": 4.577574469227928e-06,
+      "loss": 0.4639,
+      "step": 9882
+    },
+    {
+      "epoch": 3.871153371202766,
+      "grad_norm": 0.47562723439852267,
+      "learning_rate": 4.577487319650186e-06,
+      "loss": 0.4738,
+      "step": 9883
+    },
+    {
+      "epoch": 3.871548530501358,
+      "grad_norm": 0.48073672972235093,
+      "learning_rate": 4.577400161913335e-06,
+      "loss": 0.4711,
+      "step": 9884
+    },
+    {
+      "epoch": 3.8719436897999504,
+      "grad_norm": 0.4937434966089234,
+      "learning_rate": 4.577312996017718e-06,
+      "loss": 0.4678,
+      "step": 9885
+    },
+    {
+      "epoch": 3.8723388490985426,
+      "grad_norm": 0.5076968652598326,
+      "learning_rate": 4.5772258219636765e-06,
+      "loss": 0.4633,
+      "step": 9886
+    },
+    {
+      "epoch": 3.872734008397135,
+      "grad_norm": 0.4945776677023478,
+      "learning_rate": 4.5771386397515535e-06,
+      "loss": 0.4664,
+      "step": 9887
+    },
+    {
+      "epoch": 3.873129167695727,
+      "grad_norm": 0.49792744846750037,
+      "learning_rate": 4.577051449381691e-06,
+      "loss": 0.4664,
+      "step": 9888
+    },
+    {
+      "epoch": 3.8735243269943194,
+      "grad_norm": 0.48812305721614563,
+      "learning_rate": 4.576964250854432e-06,
+      "loss": 0.4683,
+      "step": 9889
+    },
+    {
+      "epoch": 3.8739194862929116,
+      "grad_norm": 0.48682855568601446,
+      "learning_rate": 4.5768770441701184e-06,
+      "loss": 0.4594,
+      "step": 9890
+    },
+    {
+      "epoch": 3.874314645591504,
+      "grad_norm": 0.48935463641111665,
+      "learning_rate": 4.576789829329093e-06,
+      "loss": 0.4752,
+      "step": 9891
+    },
+    {
+      "epoch": 3.874709804890096,
+      "grad_norm": 0.5341391098277775,
+      "learning_rate": 4.576702606331698e-06,
+      "loss": 0.4527,
+      "step": 9892
+    },
+    {
+      "epoch": 3.8751049641886883,
+      "grad_norm": 0.49067023477198496,
+      "learning_rate": 4.5766153751782775e-06,
+      "loss": 0.4546,
+      "step": 9893
+    },
+    {
+      "epoch": 3.8755001234872806,
+      "grad_norm": 0.4953575355699284,
+      "learning_rate": 4.576528135869171e-06,
+      "loss": 0.4609,
+      "step": 9894
+    },
+    {
+      "epoch": 3.875895282785873,
+      "grad_norm": 0.5064570635778317,
+      "learning_rate": 4.576440888404724e-06,
+      "loss": 0.4615,
+      "step": 9895
+    },
+    {
+      "epoch": 3.876290442084465,
+      "grad_norm": 0.4852237088041872,
+      "learning_rate": 4.576353632785278e-06,
+      "loss": 0.4504,
+      "step": 9896
+    },
+    {
+      "epoch": 3.8766856013830573,
+      "grad_norm": 0.4896826011480754,
+      "learning_rate": 4.576266369011175e-06,
+      "loss": 0.46,
+      "step": 9897
+    },
+    {
+      "epoch": 3.8770807606816495,
+      "grad_norm": 0.48478132345598957,
+      "learning_rate": 4.576179097082759e-06,
+      "loss": 0.4634,
+      "step": 9898
+    },
+    {
+      "epoch": 3.8774759199802418,
+      "grad_norm": 0.49074759563963166,
+      "learning_rate": 4.576091817000372e-06,
+      "loss": 0.4631,
+      "step": 9899
+    },
+    {
+      "epoch": 3.877871079278834,
+      "grad_norm": 0.5444614925031207,
+      "learning_rate": 4.576004528764358e-06,
+      "loss": 0.4646,
+      "step": 9900
+    },
+    {
+      "epoch": 3.8782662385774263,
+      "grad_norm": 0.4886325205119538,
+      "learning_rate": 4.575917232375058e-06,
+      "loss": 0.4634,
+      "step": 9901
+    },
+    {
+      "epoch": 3.878661397876019,
+      "grad_norm": 0.49095261197771417,
+      "learning_rate": 4.575829927832816e-06,
+      "loss": 0.465,
+      "step": 9902
+    },
+    {
+      "epoch": 3.879056557174611,
+      "grad_norm": 0.49252455851339666,
+      "learning_rate": 4.575742615137973e-06,
+      "loss": 0.4713,
+      "step": 9903
+    },
+    {
+      "epoch": 3.8794517164732034,
+      "grad_norm": 0.5467856159576922,
+      "learning_rate": 4.575655294290875e-06,
+      "loss": 0.479,
+      "step": 9904
+    },
+    {
+      "epoch": 3.8798468757717957,
+      "grad_norm": 0.49019837168996383,
+      "learning_rate": 4.575567965291864e-06,
+      "loss": 0.4722,
+      "step": 9905
+    },
+    {
+      "epoch": 3.880242035070388,
+      "grad_norm": 0.5341818933608717,
+      "learning_rate": 4.575480628141281e-06,
+      "loss": 0.4527,
+      "step": 9906
+    },
+    {
+      "epoch": 3.88063719436898,
+      "grad_norm": 0.4913343796849772,
+      "learning_rate": 4.575393282839471e-06,
+      "loss": 0.473,
+      "step": 9907
+    },
+    {
+      "epoch": 3.8810323536675724,
+      "grad_norm": 0.48134695756624346,
+      "learning_rate": 4.575305929386776e-06,
+      "loss": 0.4438,
+      "step": 9908
+    },
+    {
+      "epoch": 3.8814275129661646,
+      "grad_norm": 0.5029217370787735,
+      "learning_rate": 4.57521856778354e-06,
+      "loss": 0.4536,
+      "step": 9909
+    },
+    {
+      "epoch": 3.881822672264757,
+      "grad_norm": 0.4807948637994951,
+      "learning_rate": 4.5751311980301064e-06,
+      "loss": 0.4405,
+      "step": 9910
+    },
+    {
+      "epoch": 3.882217831563349,
+      "grad_norm": 0.47962412886103184,
+      "learning_rate": 4.5750438201268165e-06,
+      "loss": 0.4619,
+      "step": 9911
+    },
+    {
+      "epoch": 3.8826129908619413,
+      "grad_norm": 0.49393796080145874,
+      "learning_rate": 4.574956434074014e-06,
+      "loss": 0.4589,
+      "step": 9912
+    },
+    {
+      "epoch": 3.8830081501605336,
+      "grad_norm": 0.4946297200387813,
+      "learning_rate": 4.574869039872044e-06,
+      "loss": 0.4811,
+      "step": 9913
+    },
+    {
+      "epoch": 3.883403309459126,
+      "grad_norm": 0.5028163739738554,
+      "learning_rate": 4.574781637521247e-06,
+      "loss": 0.4621,
+      "step": 9914
+    },
+    {
+      "epoch": 3.883798468757718,
+      "grad_norm": 0.4846491793605942,
+      "learning_rate": 4.5746942270219686e-06,
+      "loss": 0.461,
+      "step": 9915
+    },
+    {
+      "epoch": 3.8841936280563103,
+      "grad_norm": 0.5158987625564292,
+      "learning_rate": 4.574606808374551e-06,
+      "loss": 0.4566,
+      "step": 9916
+    },
+    {
+      "epoch": 3.8845887873549025,
+      "grad_norm": 0.4833525715221507,
+      "learning_rate": 4.574519381579337e-06,
+      "loss": 0.4601,
+      "step": 9917
+    },
+    {
+      "epoch": 3.884983946653495,
+      "grad_norm": 0.5076912429801129,
+      "learning_rate": 4.574431946636671e-06,
+      "loss": 0.4689,
+      "step": 9918
+    },
+    {
+      "epoch": 3.885379105952087,
+      "grad_norm": 0.4863106371729655,
+      "learning_rate": 4.574344503546896e-06,
+      "loss": 0.4532,
+      "step": 9919
+    },
+    {
+      "epoch": 3.8857742652506793,
+      "grad_norm": 0.49043260914664416,
+      "learning_rate": 4.5742570523103555e-06,
+      "loss": 0.4664,
+      "step": 9920
+    },
+    {
+      "epoch": 3.8861694245492715,
+      "grad_norm": 0.5123972555223707,
+      "learning_rate": 4.574169592927392e-06,
+      "loss": 0.4707,
+      "step": 9921
+    },
+    {
+      "epoch": 3.8865645838478637,
+      "grad_norm": 0.47548790562191307,
+      "learning_rate": 4.5740821253983505e-06,
+      "loss": 0.4617,
+      "step": 9922
+    },
+    {
+      "epoch": 3.886959743146456,
+      "grad_norm": 0.489095797302651,
+      "learning_rate": 4.573994649723575e-06,
+      "loss": 0.4711,
+      "step": 9923
+    },
+    {
+      "epoch": 3.887354902445048,
+      "grad_norm": 0.48921262328932935,
+      "learning_rate": 4.573907165903406e-06,
+      "loss": 0.4555,
+      "step": 9924
+    },
+    {
+      "epoch": 3.8877500617436405,
+      "grad_norm": 0.49537271238528213,
+      "learning_rate": 4.57381967393819e-06,
+      "loss": 0.4711,
+      "step": 9925
+    },
+    {
+      "epoch": 3.8881452210422327,
+      "grad_norm": 0.5259888620610288,
+      "learning_rate": 4.573732173828269e-06,
+      "loss": 0.4623,
+      "step": 9926
+    },
+    {
+      "epoch": 3.888540380340825,
+      "grad_norm": 0.48638729424662486,
+      "learning_rate": 4.573644665573987e-06,
+      "loss": 0.4726,
+      "step": 9927
+    },
+    {
+      "epoch": 3.888935539639417,
+      "grad_norm": 0.503911496590745,
+      "learning_rate": 4.573557149175689e-06,
+      "loss": 0.4711,
+      "step": 9928
+    },
+    {
+      "epoch": 3.8893306989380094,
+      "grad_norm": 0.47957742258960145,
+      "learning_rate": 4.573469624633717e-06,
+      "loss": 0.4574,
+      "step": 9929
+    },
+    {
+      "epoch": 3.8897258582366017,
+      "grad_norm": 0.49719338473032687,
+      "learning_rate": 4.5733820919484165e-06,
+      "loss": 0.4637,
+      "step": 9930
+    },
+    {
+      "epoch": 3.890121017535194,
+      "grad_norm": 0.49630282899421835,
+      "learning_rate": 4.573294551120129e-06,
+      "loss": 0.4621,
+      "step": 9931
+    },
+    {
+      "epoch": 3.890516176833786,
+      "grad_norm": 0.48727643540534094,
+      "learning_rate": 4.573207002149199e-06,
+      "loss": 0.4585,
+      "step": 9932
+    },
+    {
+      "epoch": 3.8909113361323784,
+      "grad_norm": 0.4809569000720417,
+      "learning_rate": 4.573119445035972e-06,
+      "loss": 0.4586,
+      "step": 9933
+    },
+    {
+      "epoch": 3.8913064954309706,
+      "grad_norm": 0.47571185854215003,
+      "learning_rate": 4.57303187978079e-06,
+      "loss": 0.4683,
+      "step": 9934
+    },
+    {
+      "epoch": 3.891701654729563,
+      "grad_norm": 0.48390173237622336,
+      "learning_rate": 4.5729443063839984e-06,
+      "loss": 0.4748,
+      "step": 9935
+    },
+    {
+      "epoch": 3.892096814028155,
+      "grad_norm": 0.49110153521074534,
+      "learning_rate": 4.57285672484594e-06,
+      "loss": 0.4597,
+      "step": 9936
+    },
+    {
+      "epoch": 3.8924919733267473,
+      "grad_norm": 0.4952646122846938,
+      "learning_rate": 4.572769135166959e-06,
+      "loss": 0.4827,
+      "step": 9937
+    },
+    {
+      "epoch": 3.8928871326253396,
+      "grad_norm": 0.4823294860273062,
+      "learning_rate": 4.572681537347398e-06,
+      "loss": 0.47,
+      "step": 9938
+    },
+    {
+      "epoch": 3.893282291923932,
+      "grad_norm": 0.48698397822277945,
+      "learning_rate": 4.572593931387604e-06,
+      "loss": 0.4602,
+      "step": 9939
+    },
+    {
+      "epoch": 3.893677451222524,
+      "grad_norm": 0.48400546395610927,
+      "learning_rate": 4.572506317287921e-06,
+      "loss": 0.4481,
+      "step": 9940
+    },
+    {
+      "epoch": 3.8940726105211163,
+      "grad_norm": 0.48847518203566603,
+      "learning_rate": 4.572418695048689e-06,
+      "loss": 0.4573,
+      "step": 9941
+    },
+    {
+      "epoch": 3.8944677698197085,
+      "grad_norm": 0.4854571876960505,
+      "learning_rate": 4.572331064670257e-06,
+      "loss": 0.4626,
+      "step": 9942
+    },
+    {
+      "epoch": 3.894862929118301,
+      "grad_norm": 0.47898682929482006,
+      "learning_rate": 4.572243426152965e-06,
+      "loss": 0.44,
+      "step": 9943
+    },
+    {
+      "epoch": 3.895258088416893,
+      "grad_norm": 0.491532931524844,
+      "learning_rate": 4.57215577949716e-06,
+      "loss": 0.4807,
+      "step": 9944
+    },
+    {
+      "epoch": 3.8956532477154853,
+      "grad_norm": 0.47872724576808984,
+      "learning_rate": 4.572068124703185e-06,
+      "loss": 0.4603,
+      "step": 9945
+    },
+    {
+      "epoch": 3.8960484070140775,
+      "grad_norm": 0.45451419510929936,
+      "learning_rate": 4.571980461771386e-06,
+      "loss": 0.458,
+      "step": 9946
+    },
+    {
+      "epoch": 3.8964435663126697,
+      "grad_norm": 0.4762087573565171,
+      "learning_rate": 4.571892790702105e-06,
+      "loss": 0.4539,
+      "step": 9947
+    },
+    {
+      "epoch": 3.896838725611262,
+      "grad_norm": 0.501694628462472,
+      "learning_rate": 4.571805111495687e-06,
+      "loss": 0.4593,
+      "step": 9948
+    },
+    {
+      "epoch": 3.8972338849098542,
+      "grad_norm": 0.49270030012761595,
+      "learning_rate": 4.571717424152476e-06,
+      "loss": 0.4679,
+      "step": 9949
+    },
+    {
+      "epoch": 3.8976290442084465,
+      "grad_norm": 0.4925249641495412,
+      "learning_rate": 4.5716297286728184e-06,
+      "loss": 0.4769,
+      "step": 9950
+    },
+    {
+      "epoch": 3.8980242035070387,
+      "grad_norm": 0.4858992891832617,
+      "learning_rate": 4.571542025057057e-06,
+      "loss": 0.4795,
+      "step": 9951
+    },
+    {
+      "epoch": 3.898419362805631,
+      "grad_norm": 0.4786925181433357,
+      "learning_rate": 4.571454313305536e-06,
+      "loss": 0.4415,
+      "step": 9952
+    },
+    {
+      "epoch": 3.898814522104223,
+      "grad_norm": 0.48023530140553916,
+      "learning_rate": 4.5713665934186005e-06,
+      "loss": 0.4498,
+      "step": 9953
+    },
+    {
+      "epoch": 3.8992096814028154,
+      "grad_norm": 0.48453549831266507,
+      "learning_rate": 4.571278865396594e-06,
+      "loss": 0.4644,
+      "step": 9954
+    },
+    {
+      "epoch": 3.8996048407014077,
+      "grad_norm": 0.4900914360178932,
+      "learning_rate": 4.571191129239863e-06,
+      "loss": 0.474,
+      "step": 9955
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.5041682526200453,
+      "learning_rate": 4.57110338494875e-06,
+      "loss": 0.4554,
+      "step": 9956
+    },
+    {
+      "epoch": 3.900395159298592,
+      "grad_norm": 0.5410219421410024,
+      "learning_rate": 4.571015632523601e-06,
+      "loss": 0.4793,
+      "step": 9957
+    },
+    {
+      "epoch": 3.9007903185971844,
+      "grad_norm": 0.4825869143751156,
+      "learning_rate": 4.57092787196476e-06,
+      "loss": 0.482,
+      "step": 9958
+    },
+    {
+      "epoch": 3.9011854778957766,
+      "grad_norm": 0.48251364045941925,
+      "learning_rate": 4.5708401032725725e-06,
+      "loss": 0.4482,
+      "step": 9959
+    },
+    {
+      "epoch": 3.901580637194369,
+      "grad_norm": 0.47369158864484645,
+      "learning_rate": 4.570752326447382e-06,
+      "loss": 0.4565,
+      "step": 9960
+    },
+    {
+      "epoch": 3.901975796492961,
+      "grad_norm": 0.4777236188587447,
+      "learning_rate": 4.5706645414895335e-06,
+      "loss": 0.4592,
+      "step": 9961
+    },
+    {
+      "epoch": 3.9023709557915534,
+      "grad_norm": 0.4818832878308696,
+      "learning_rate": 4.5705767483993725e-06,
+      "loss": 0.4561,
+      "step": 9962
+    },
+    {
+      "epoch": 3.9027661150901456,
+      "grad_norm": 0.4728202329441403,
+      "learning_rate": 4.570488947177243e-06,
+      "loss": 0.4611,
+      "step": 9963
+    },
+    {
+      "epoch": 3.903161274388738,
+      "grad_norm": 0.4774824700271554,
+      "learning_rate": 4.570401137823491e-06,
+      "loss": 0.4667,
+      "step": 9964
+    },
+    {
+      "epoch": 3.90355643368733,
+      "grad_norm": 0.5201325910426959,
+      "learning_rate": 4.57031332033846e-06,
+      "loss": 0.4556,
+      "step": 9965
+    },
+    {
+      "epoch": 3.9039515929859223,
+      "grad_norm": 0.49100204946745024,
+      "learning_rate": 4.570225494722495e-06,
+      "loss": 0.4605,
+      "step": 9966
+    },
+    {
+      "epoch": 3.9043467522845146,
+      "grad_norm": 0.4933350216968649,
+      "learning_rate": 4.5701376609759415e-06,
+      "loss": 0.465,
+      "step": 9967
+    },
+    {
+      "epoch": 3.904741911583107,
+      "grad_norm": 0.49608467793393873,
+      "learning_rate": 4.570049819099145e-06,
+      "loss": 0.4673,
+      "step": 9968
+    },
+    {
+      "epoch": 3.905137070881699,
+      "grad_norm": 0.49885302185060626,
+      "learning_rate": 4.569961969092449e-06,
+      "loss": 0.4731,
+      "step": 9969
+    },
+    {
+      "epoch": 3.9055322301802913,
+      "grad_norm": 0.4765724417728314,
+      "learning_rate": 4.569874110956201e-06,
+      "loss": 0.4664,
+      "step": 9970
+    },
+    {
+      "epoch": 3.9059273894788835,
+      "grad_norm": 0.48345495768456714,
+      "learning_rate": 4.569786244690743e-06,
+      "loss": 0.4747,
+      "step": 9971
+    },
+    {
+      "epoch": 3.9063225487774758,
+      "grad_norm": 0.48003753403167,
+      "learning_rate": 4.569698370296421e-06,
+      "loss": 0.445,
+      "step": 9972
+    },
+    {
+      "epoch": 3.9067177080760684,
+      "grad_norm": 0.4923494525260739,
+      "learning_rate": 4.5696104877735815e-06,
+      "loss": 0.4575,
+      "step": 9973
+    },
+    {
+      "epoch": 3.9071128673746607,
+      "grad_norm": 0.5211429128554534,
+      "learning_rate": 4.569522597122569e-06,
+      "loss": 0.4637,
+      "step": 9974
+    },
+    {
+      "epoch": 3.907508026673253,
+      "grad_norm": 0.4818937227907757,
+      "learning_rate": 4.569434698343727e-06,
+      "loss": 0.4429,
+      "step": 9975
+    },
+    {
+      "epoch": 3.907903185971845,
+      "grad_norm": 0.5046848124987253,
+      "learning_rate": 4.569346791437403e-06,
+      "loss": 0.4938,
+      "step": 9976
+    },
+    {
+      "epoch": 3.9082983452704374,
+      "grad_norm": 0.5027445720769897,
+      "learning_rate": 4.5692588764039415e-06,
+      "loss": 0.4727,
+      "step": 9977
+    },
+    {
+      "epoch": 3.9086935045690296,
+      "grad_norm": 0.5039278354823773,
+      "learning_rate": 4.569170953243688e-06,
+      "loss": 0.4559,
+      "step": 9978
+    },
+    {
+      "epoch": 3.909088663867622,
+      "grad_norm": 0.4843847065341945,
+      "learning_rate": 4.569083021956987e-06,
+      "loss": 0.4564,
+      "step": 9979
+    },
+    {
+      "epoch": 3.909483823166214,
+      "grad_norm": 0.49170737209975196,
+      "learning_rate": 4.568995082544184e-06,
+      "loss": 0.4752,
+      "step": 9980
+    },
+    {
+      "epoch": 3.9098789824648064,
+      "grad_norm": 0.4784679535834229,
+      "learning_rate": 4.568907135005625e-06,
+      "loss": 0.4719,
+      "step": 9981
+    },
+    {
+      "epoch": 3.9102741417633986,
+      "grad_norm": 0.48953143430090623,
+      "learning_rate": 4.5688191793416545e-06,
+      "loss": 0.4604,
+      "step": 9982
+    },
+    {
+      "epoch": 3.910669301061991,
+      "grad_norm": 0.5026359012231226,
+      "learning_rate": 4.56873121555262e-06,
+      "loss": 0.4494,
+      "step": 9983
+    },
+    {
+      "epoch": 3.911064460360583,
+      "grad_norm": 0.48889537939386374,
+      "learning_rate": 4.568643243638864e-06,
+      "loss": 0.4452,
+      "step": 9984
+    },
+    {
+      "epoch": 3.9114596196591753,
+      "grad_norm": 0.4968415694067955,
+      "learning_rate": 4.568555263600735e-06,
+      "loss": 0.4645,
+      "step": 9985
+    },
+    {
+      "epoch": 3.9118547789577676,
+      "grad_norm": 0.4877243861936003,
+      "learning_rate": 4.568467275438575e-06,
+      "loss": 0.4587,
+      "step": 9986
+    },
+    {
+      "epoch": 3.91224993825636,
+      "grad_norm": 0.49223731516425107,
+      "learning_rate": 4.568379279152733e-06,
+      "loss": 0.459,
+      "step": 9987
+    },
+    {
+      "epoch": 3.912645097554952,
+      "grad_norm": 0.49250738243308856,
+      "learning_rate": 4.568291274743553e-06,
+      "loss": 0.4778,
+      "step": 9988
+    },
+    {
+      "epoch": 3.9130402568535443,
+      "grad_norm": 0.5009850640273527,
+      "learning_rate": 4.56820326221138e-06,
+      "loss": 0.4748,
+      "step": 9989
+    },
+    {
+      "epoch": 3.9134354161521365,
+      "grad_norm": 0.4961840767254751,
+      "learning_rate": 4.568115241556562e-06,
+      "loss": 0.4531,
+      "step": 9990
+    },
+    {
+      "epoch": 3.9138305754507288,
+      "grad_norm": 0.48894424493050087,
+      "learning_rate": 4.5680272127794424e-06,
+      "loss": 0.4451,
+      "step": 9991
+    },
+    {
+      "epoch": 3.914225734749321,
+      "grad_norm": 0.4811481982683269,
+      "learning_rate": 4.567939175880367e-06,
+      "loss": 0.4579,
+      "step": 9992
+    },
+    {
+      "epoch": 3.9146208940479132,
+      "grad_norm": 0.4924442883893573,
+      "learning_rate": 4.567851130859683e-06,
+      "loss": 0.4714,
+      "step": 9993
+    },
+    {
+      "epoch": 3.9150160533465055,
+      "grad_norm": 0.5035466446279564,
+      "learning_rate": 4.567763077717735e-06,
+      "loss": 0.4947,
+      "step": 9994
+    },
+    {
+      "epoch": 3.9154112126450977,
+      "grad_norm": 0.4995580198600491,
+      "learning_rate": 4.56767501645487e-06,
+      "loss": 0.4711,
+      "step": 9995
+    },
+    {
+      "epoch": 3.91580637194369,
+      "grad_norm": 0.495893119902799,
+      "learning_rate": 4.5675869470714314e-06,
+      "loss": 0.4749,
+      "step": 9996
+    },
+    {
+      "epoch": 3.916201531242282,
+      "grad_norm": 0.4723772655931205,
+      "learning_rate": 4.567498869567769e-06,
+      "loss": 0.4362,
+      "step": 9997
+    },
+    {
+      "epoch": 3.9165966905408744,
+      "grad_norm": 0.4718587728404546,
+      "learning_rate": 4.567410783944225e-06,
+      "loss": 0.4638,
+      "step": 9998
+    },
+    {
+      "epoch": 3.9169918498394667,
+      "grad_norm": 0.4928852611306575,
+      "learning_rate": 4.567322690201147e-06,
+      "loss": 0.4487,
+      "step": 9999
+    },
+    {
+      "epoch": 3.917387009138059,
+      "grad_norm": 0.47713196411455694,
+      "learning_rate": 4.5672345883388816e-06,
+      "loss": 0.455,
+      "step": 10000
+    },
+    {
+      "epoch": 3.917782168436651,
+      "grad_norm": 0.49819690808918476,
+      "learning_rate": 4.567146478357773e-06,
+      "loss": 0.4426,
+      "step": 10001
+    },
+    {
+      "epoch": 3.9181773277352434,
+      "grad_norm": 0.5062977312521194,
+      "learning_rate": 4.5670583602581685e-06,
+      "loss": 0.4653,
+      "step": 10002
+    },
+    {
+      "epoch": 3.9185724870338356,
+      "grad_norm": 0.5003782785606375,
+      "learning_rate": 4.566970234040415e-06,
+      "loss": 0.4615,
+      "step": 10003
+    },
+    {
+      "epoch": 3.918967646332428,
+      "grad_norm": 0.5189545106568986,
+      "learning_rate": 4.566882099704857e-06,
+      "loss": 0.4731,
+      "step": 10004
+    },
+    {
+      "epoch": 3.91936280563102,
+      "grad_norm": 0.5035785229251792,
+      "learning_rate": 4.566793957251841e-06,
+      "loss": 0.4779,
+      "step": 10005
+    },
+    {
+      "epoch": 3.9197579649296124,
+      "grad_norm": 0.4800415368289926,
+      "learning_rate": 4.566705806681712e-06,
+      "loss": 0.4539,
+      "step": 10006
+    },
+    {
+      "epoch": 3.9201531242282046,
+      "grad_norm": 0.4882424905193737,
+      "learning_rate": 4.56661764799482e-06,
+      "loss": 0.4584,
+      "step": 10007
+    },
+    {
+      "epoch": 3.920548283526797,
+      "grad_norm": 0.47308810535562973,
+      "learning_rate": 4.566529481191507e-06,
+      "loss": 0.4741,
+      "step": 10008
+    },
+    {
+      "epoch": 3.920943442825389,
+      "grad_norm": 0.4867276304401251,
+      "learning_rate": 4.566441306272123e-06,
+      "loss": 0.4541,
+      "step": 10009
+    },
+    {
+      "epoch": 3.9213386021239813,
+      "grad_norm": 0.6343188880548906,
+      "learning_rate": 4.56635312323701e-06,
+      "loss": 0.4842,
+      "step": 10010
+    },
+    {
+      "epoch": 3.9217337614225736,
+      "grad_norm": 0.5071802879941949,
+      "learning_rate": 4.5662649320865186e-06,
+      "loss": 0.489,
+      "step": 10011
+    },
+    {
+      "epoch": 3.922128920721166,
+      "grad_norm": 0.5255039030846776,
+      "learning_rate": 4.566176732820991e-06,
+      "loss": 0.4627,
+      "step": 10012
+    },
+    {
+      "epoch": 3.922524080019758,
+      "grad_norm": 0.4971011698886485,
+      "learning_rate": 4.566088525440778e-06,
+      "loss": 0.4599,
+      "step": 10013
+    },
+    {
+      "epoch": 3.9229192393183503,
+      "grad_norm": 0.4852264543790252,
+      "learning_rate": 4.566000309946223e-06,
+      "loss": 0.4602,
+      "step": 10014
+    },
+    {
+      "epoch": 3.9233143986169425,
+      "grad_norm": 0.49830525957051336,
+      "learning_rate": 4.565912086337674e-06,
+      "loss": 0.4736,
+      "step": 10015
+    },
+    {
+      "epoch": 3.9237095579155348,
+      "grad_norm": 0.4773373018311242,
+      "learning_rate": 4.565823854615477e-06,
+      "loss": 0.466,
+      "step": 10016
+    },
+    {
+      "epoch": 3.924104717214127,
+      "grad_norm": 0.4831436103078164,
+      "learning_rate": 4.565735614779977e-06,
+      "loss": 0.4643,
+      "step": 10017
+    },
+    {
+      "epoch": 3.9244998765127193,
+      "grad_norm": 0.4802473860556205,
+      "learning_rate": 4.565647366831522e-06,
+      "loss": 0.4721,
+      "step": 10018
+    },
+    {
+      "epoch": 3.9248950358113115,
+      "grad_norm": 0.49874045500383063,
+      "learning_rate": 4.5655591107704595e-06,
+      "loss": 0.4702,
+      "step": 10019
+    },
+    {
+      "epoch": 3.9252901951099037,
+      "grad_norm": 0.47831567102450434,
+      "learning_rate": 4.565470846597135e-06,
+      "loss": 0.474,
+      "step": 10020
+    },
+    {
+      "epoch": 3.925685354408496,
+      "grad_norm": 0.49992203811316205,
+      "learning_rate": 4.565382574311894e-06,
+      "loss": 0.4655,
+      "step": 10021
+    },
+    {
+      "epoch": 3.926080513707088,
+      "grad_norm": 0.4926517107007104,
+      "learning_rate": 4.565294293915086e-06,
+      "loss": 0.4675,
+      "step": 10022
+    },
+    {
+      "epoch": 3.9264756730056805,
+      "grad_norm": 0.5084730405938702,
+      "learning_rate": 4.565206005407055e-06,
+      "loss": 0.4822,
+      "step": 10023
+    },
+    {
+      "epoch": 3.9268708323042727,
+      "grad_norm": 0.4931236099060291,
+      "learning_rate": 4.565117708788149e-06,
+      "loss": 0.4437,
+      "step": 10024
+    },
+    {
+      "epoch": 3.927265991602865,
+      "grad_norm": 0.483301233803827,
+      "learning_rate": 4.565029404058715e-06,
+      "loss": 0.4709,
+      "step": 10025
+    },
+    {
+      "epoch": 3.927661150901457,
+      "grad_norm": 0.47819265448253306,
+      "learning_rate": 4.5649410912191e-06,
+      "loss": 0.4658,
+      "step": 10026
+    },
+    {
+      "epoch": 3.9280563102000494,
+      "grad_norm": 0.4822872193597864,
+      "learning_rate": 4.564852770269648e-06,
+      "loss": 0.4385,
+      "step": 10027
+    },
+    {
+      "epoch": 3.9284514694986417,
+      "grad_norm": 0.48138130491052755,
+      "learning_rate": 4.5647644412107104e-06,
+      "loss": 0.4603,
+      "step": 10028
+    },
+    {
+      "epoch": 3.928846628797234,
+      "grad_norm": 0.4945851911204728,
+      "learning_rate": 4.564676104042631e-06,
+      "loss": 0.4505,
+      "step": 10029
+    },
+    {
+      "epoch": 3.929241788095826,
+      "grad_norm": 0.48293991390518787,
+      "learning_rate": 4.564587758765759e-06,
+      "loss": 0.45,
+      "step": 10030
+    },
+    {
+      "epoch": 3.9296369473944184,
+      "grad_norm": 0.5024502851465005,
+      "learning_rate": 4.5644994053804384e-06,
+      "loss": 0.4603,
+      "step": 10031
+    },
+    {
+      "epoch": 3.9300321066930106,
+      "grad_norm": 0.47694264905270695,
+      "learning_rate": 4.564411043887018e-06,
+      "loss": 0.4547,
+      "step": 10032
+    },
+    {
+      "epoch": 3.930427265991603,
+      "grad_norm": 0.5039734950596827,
+      "learning_rate": 4.564322674285845e-06,
+      "loss": 0.4551,
+      "step": 10033
+    },
+    {
+      "epoch": 3.930822425290195,
+      "grad_norm": 0.48332955949025447,
+      "learning_rate": 4.564234296577266e-06,
+      "loss": 0.4636,
+      "step": 10034
+    },
+    {
+      "epoch": 3.9312175845887873,
+      "grad_norm": 0.487981969360117,
+      "learning_rate": 4.564145910761627e-06,
+      "loss": 0.4659,
+      "step": 10035
+    },
+    {
+      "epoch": 3.9316127438873796,
+      "grad_norm": 0.4884183172019331,
+      "learning_rate": 4.564057516839277e-06,
+      "loss": 0.4776,
+      "step": 10036
+    },
+    {
+      "epoch": 3.932007903185972,
+      "grad_norm": 0.484407990579762,
+      "learning_rate": 4.563969114810563e-06,
+      "loss": 0.4549,
+      "step": 10037
+    },
+    {
+      "epoch": 3.932403062484564,
+      "grad_norm": 0.49302390973656574,
+      "learning_rate": 4.563880704675831e-06,
+      "loss": 0.461,
+      "step": 10038
+    },
+    {
+      "epoch": 3.9327982217831563,
+      "grad_norm": 0.4824245546460462,
+      "learning_rate": 4.56379228643543e-06,
+      "loss": 0.4456,
+      "step": 10039
+    },
+    {
+      "epoch": 3.9331933810817485,
+      "grad_norm": 0.524281421582032,
+      "learning_rate": 4.563703860089705e-06,
+      "loss": 0.4631,
+      "step": 10040
+    },
+    {
+      "epoch": 3.933588540380341,
+      "grad_norm": 0.48699863426336437,
+      "learning_rate": 4.563615425639005e-06,
+      "loss": 0.4554,
+      "step": 10041
+    },
+    {
+      "epoch": 3.933983699678933,
+      "grad_norm": 0.4956289418570272,
+      "learning_rate": 4.5635269830836764e-06,
+      "loss": 0.4691,
+      "step": 10042
+    },
+    {
+      "epoch": 3.9343788589775253,
+      "grad_norm": 0.48487716834135475,
+      "learning_rate": 4.563438532424067e-06,
+      "loss": 0.4785,
+      "step": 10043
+    },
+    {
+      "epoch": 3.9347740182761175,
+      "grad_norm": 0.47820983244486903,
+      "learning_rate": 4.563350073660524e-06,
+      "loss": 0.4445,
+      "step": 10044
+    },
+    {
+      "epoch": 3.9351691775747097,
+      "grad_norm": 0.5073931859582581,
+      "learning_rate": 4.5632616067933944e-06,
+      "loss": 0.4777,
+      "step": 10045
+    },
+    {
+      "epoch": 3.935564336873302,
+      "grad_norm": 0.473757508440239,
+      "learning_rate": 4.563173131823026e-06,
+      "loss": 0.4494,
+      "step": 10046
+    },
+    {
+      "epoch": 3.9359594961718942,
+      "grad_norm": 0.4807056489036854,
+      "learning_rate": 4.563084648749767e-06,
+      "loss": 0.4452,
+      "step": 10047
+    },
+    {
+      "epoch": 3.9363546554704865,
+      "grad_norm": 0.4878554376029113,
+      "learning_rate": 4.562996157573964e-06,
+      "loss": 0.4536,
+      "step": 10048
+    },
+    {
+      "epoch": 3.9367498147690787,
+      "grad_norm": 0.5087371100450915,
+      "learning_rate": 4.562907658295966e-06,
+      "loss": 0.488,
+      "step": 10049
+    },
+    {
+      "epoch": 3.937144974067671,
+      "grad_norm": 0.48438840021575363,
+      "learning_rate": 4.562819150916118e-06,
+      "loss": 0.4857,
+      "step": 10050
+    },
+    {
+      "epoch": 3.937540133366263,
+      "grad_norm": 0.49057136917157657,
+      "learning_rate": 4.562730635434768e-06,
+      "loss": 0.4643,
+      "step": 10051
+    },
+    {
+      "epoch": 3.9379352926648554,
+      "grad_norm": 0.47526482048284263,
+      "learning_rate": 4.562642111852266e-06,
+      "loss": 0.4482,
+      "step": 10052
+    },
+    {
+      "epoch": 3.9383304519634477,
+      "grad_norm": 0.5014131338073492,
+      "learning_rate": 4.562553580168958e-06,
+      "loss": 0.4565,
+      "step": 10053
+    },
+    {
+      "epoch": 3.93872561126204,
+      "grad_norm": 0.5031712717781439,
+      "learning_rate": 4.562465040385193e-06,
+      "loss": 0.4495,
+      "step": 10054
+    },
+    {
+      "epoch": 3.939120770560632,
+      "grad_norm": 1.0950220813960667,
+      "learning_rate": 4.562376492501316e-06,
+      "loss": 0.4635,
+      "step": 10055
+    },
+    {
+      "epoch": 3.9395159298592244,
+      "grad_norm": 0.4839972284867313,
+      "learning_rate": 4.5622879365176775e-06,
+      "loss": 0.4449,
+      "step": 10056
+    },
+    {
+      "epoch": 3.9399110891578166,
+      "grad_norm": 0.4914974711515712,
+      "learning_rate": 4.562199372434624e-06,
+      "loss": 0.4587,
+      "step": 10057
+    },
+    {
+      "epoch": 3.940306248456409,
+      "grad_norm": 0.4876711211009558,
+      "learning_rate": 4.562110800252504e-06,
+      "loss": 0.4591,
+      "step": 10058
+    },
+    {
+      "epoch": 3.940701407755001,
+      "grad_norm": 0.47546627363978317,
+      "learning_rate": 4.5620222199716645e-06,
+      "loss": 0.4416,
+      "step": 10059
+    },
+    {
+      "epoch": 3.9410965670535933,
+      "grad_norm": 0.48586322704048324,
+      "learning_rate": 4.561933631592453e-06,
+      "loss": 0.4549,
+      "step": 10060
+    },
+    {
+      "epoch": 3.9414917263521856,
+      "grad_norm": 0.4926171275617641,
+      "learning_rate": 4.56184503511522e-06,
+      "loss": 0.4598,
+      "step": 10061
+    },
+    {
+      "epoch": 3.941886885650778,
+      "grad_norm": 0.5123664239699524,
+      "learning_rate": 4.561756430540311e-06,
+      "loss": 0.4851,
+      "step": 10062
+    },
+    {
+      "epoch": 3.94228204494937,
+      "grad_norm": 0.5199067760226548,
+      "learning_rate": 4.5616678178680744e-06,
+      "loss": 0.4807,
+      "step": 10063
+    },
+    {
+      "epoch": 3.9426772042479623,
+      "grad_norm": 0.49024172756251366,
+      "learning_rate": 4.561579197098858e-06,
+      "loss": 0.4632,
+      "step": 10064
+    },
+    {
+      "epoch": 3.9430723635465545,
+      "grad_norm": 0.4814153973868991,
+      "learning_rate": 4.561490568233013e-06,
+      "loss": 0.4495,
+      "step": 10065
+    },
+    {
+      "epoch": 3.943467522845147,
+      "grad_norm": 0.47441139134775145,
+      "learning_rate": 4.561401931270882e-06,
+      "loss": 0.4604,
+      "step": 10066
+    },
+    {
+      "epoch": 3.943862682143739,
+      "grad_norm": 0.4924678869233399,
+      "learning_rate": 4.561313286212817e-06,
+      "loss": 0.4699,
+      "step": 10067
+    },
+    {
+      "epoch": 3.9442578414423313,
+      "grad_norm": 0.48654953928643213,
+      "learning_rate": 4.561224633059166e-06,
+      "loss": 0.4489,
+      "step": 10068
+    },
+    {
+      "epoch": 3.9446530007409235,
+      "grad_norm": 0.4963720302187141,
+      "learning_rate": 4.561135971810275e-06,
+      "loss": 0.4721,
+      "step": 10069
+    },
+    {
+      "epoch": 3.9450481600395157,
+      "grad_norm": 0.5091750814580799,
+      "learning_rate": 4.561047302466494e-06,
+      "loss": 0.4736,
+      "step": 10070
+    },
+    {
+      "epoch": 3.945443319338108,
+      "grad_norm": 0.49308737751715986,
+      "learning_rate": 4.56095862502817e-06,
+      "loss": 0.4517,
+      "step": 10071
+    },
+    {
+      "epoch": 3.9458384786367002,
+      "grad_norm": 0.4895075330718553,
+      "learning_rate": 4.5608699394956525e-06,
+      "loss": 0.4472,
+      "step": 10072
+    },
+    {
+      "epoch": 3.9462336379352925,
+      "grad_norm": 0.5031481365671037,
+      "learning_rate": 4.56078124586929e-06,
+      "loss": 0.4716,
+      "step": 10073
+    },
+    {
+      "epoch": 3.9466287972338847,
+      "grad_norm": 0.4990715567557606,
+      "learning_rate": 4.560692544149429e-06,
+      "loss": 0.4452,
+      "step": 10074
+    },
+    {
+      "epoch": 3.947023956532477,
+      "grad_norm": 0.4795562000811165,
+      "learning_rate": 4.56060383433642e-06,
+      "loss": 0.4639,
+      "step": 10075
+    },
+    {
+      "epoch": 3.947419115831069,
+      "grad_norm": 0.4809738041038454,
+      "learning_rate": 4.5605151164306095e-06,
+      "loss": 0.4589,
+      "step": 10076
+    },
+    {
+      "epoch": 3.9478142751296614,
+      "grad_norm": 0.5003235218268587,
+      "learning_rate": 4.5604263904323474e-06,
+      "loss": 0.4782,
+      "step": 10077
+    },
+    {
+      "epoch": 3.9482094344282537,
+      "grad_norm": 0.49029440459937,
+      "learning_rate": 4.560337656341981e-06,
+      "loss": 0.45,
+      "step": 10078
+    },
+    {
+      "epoch": 3.948604593726846,
+      "grad_norm": 0.48335540139031397,
+      "learning_rate": 4.560248914159861e-06,
+      "loss": 0.4601,
+      "step": 10079
+    },
+    {
+      "epoch": 3.948999753025438,
+      "grad_norm": 0.4893171254147366,
+      "learning_rate": 4.560160163886332e-06,
+      "loss": 0.4486,
+      "step": 10080
+    },
+    {
+      "epoch": 3.9493949123240304,
+      "grad_norm": 0.48236278214675543,
+      "learning_rate": 4.560071405521746e-06,
+      "loss": 0.4755,
+      "step": 10081
+    },
+    {
+      "epoch": 3.9497900716226226,
+      "grad_norm": 0.49023533737652236,
+      "learning_rate": 4.559982639066451e-06,
+      "loss": 0.457,
+      "step": 10082
+    },
+    {
+      "epoch": 3.950185230921215,
+      "grad_norm": 0.4812818608327351,
+      "learning_rate": 4.559893864520795e-06,
+      "loss": 0.4616,
+      "step": 10083
+    },
+    {
+      "epoch": 3.950580390219807,
+      "grad_norm": 0.4809646923115125,
+      "learning_rate": 4.559805081885126e-06,
+      "loss": 0.4535,
+      "step": 10084
+    },
+    {
+      "epoch": 3.9509755495183994,
+      "grad_norm": 0.5242179398310719,
+      "learning_rate": 4.559716291159793e-06,
+      "loss": 0.4589,
+      "step": 10085
+    },
+    {
+      "epoch": 3.9513707088169916,
+      "grad_norm": 0.48289980979582037,
+      "learning_rate": 4.559627492345147e-06,
+      "loss": 0.4572,
+      "step": 10086
+    },
+    {
+      "epoch": 3.951765868115584,
+      "grad_norm": 0.4805787354967715,
+      "learning_rate": 4.5595386854415335e-06,
+      "loss": 0.4507,
+      "step": 10087
+    },
+    {
+      "epoch": 3.952161027414176,
+      "grad_norm": 0.4853162285383262,
+      "learning_rate": 4.5594498704493025e-06,
+      "loss": 0.473,
+      "step": 10088
+    },
+    {
+      "epoch": 3.9525561867127683,
+      "grad_norm": 0.5148358522338808,
+      "learning_rate": 4.559361047368803e-06,
+      "loss": 0.4558,
+      "step": 10089
+    },
+    {
+      "epoch": 3.9529513460113606,
+      "grad_norm": 0.49389057275002457,
+      "learning_rate": 4.559272216200385e-06,
+      "loss": 0.4529,
+      "step": 10090
+    },
+    {
+      "epoch": 3.9533465053099532,
+      "grad_norm": 0.4800782868223698,
+      "learning_rate": 4.559183376944395e-06,
+      "loss": 0.4654,
+      "step": 10091
+    },
+    {
+      "epoch": 3.9537416646085455,
+      "grad_norm": 0.49900896659440425,
+      "learning_rate": 4.559094529601183e-06,
+      "loss": 0.4511,
+      "step": 10092
+    },
+    {
+      "epoch": 3.9541368239071377,
+      "grad_norm": 0.5014788964906836,
+      "learning_rate": 4.5590056741711e-06,
+      "loss": 0.4565,
+      "step": 10093
+    },
+    {
+      "epoch": 3.95453198320573,
+      "grad_norm": 0.49370738821746696,
+      "learning_rate": 4.558916810654491e-06,
+      "loss": 0.442,
+      "step": 10094
+    },
+    {
+      "epoch": 3.954927142504322,
+      "grad_norm": 0.49314453829677907,
+      "learning_rate": 4.558827939051707e-06,
+      "loss": 0.4664,
+      "step": 10095
+    },
+    {
+      "epoch": 3.9553223018029144,
+      "grad_norm": 0.4850261607507114,
+      "learning_rate": 4.558739059363098e-06,
+      "loss": 0.4599,
+      "step": 10096
+    },
+    {
+      "epoch": 3.9557174611015067,
+      "grad_norm": 0.5150383652631196,
+      "learning_rate": 4.558650171589012e-06,
+      "loss": 0.479,
+      "step": 10097
+    },
+    {
+      "epoch": 3.956112620400099,
+      "grad_norm": 0.49222732950123294,
+      "learning_rate": 4.558561275729798e-06,
+      "loss": 0.46,
+      "step": 10098
+    },
+    {
+      "epoch": 3.956507779698691,
+      "grad_norm": 0.5036146912323076,
+      "learning_rate": 4.558472371785805e-06,
+      "loss": 0.476,
+      "step": 10099
+    },
+    {
+      "epoch": 3.9569029389972834,
+      "grad_norm": 0.49324709847980996,
+      "learning_rate": 4.558383459757383e-06,
+      "loss": 0.4721,
+      "step": 10100
+    },
+    {
+      "epoch": 3.9572980982958756,
+      "grad_norm": 0.4688106876677952,
+      "learning_rate": 4.55829453964488e-06,
+      "loss": 0.4553,
+      "step": 10101
+    },
+    {
+      "epoch": 3.957693257594468,
+      "grad_norm": 0.4864105983815226,
+      "learning_rate": 4.558205611448646e-06,
+      "loss": 0.4507,
+      "step": 10102
+    },
+    {
+      "epoch": 3.95808841689306,
+      "grad_norm": 0.5127931976516101,
+      "learning_rate": 4.5581166751690306e-06,
+      "loss": 0.4646,
+      "step": 10103
+    },
+    {
+      "epoch": 3.9584835761916524,
+      "grad_norm": 0.493192790264334,
+      "learning_rate": 4.558027730806383e-06,
+      "loss": 0.4634,
+      "step": 10104
+    },
+    {
+      "epoch": 3.9588787354902446,
+      "grad_norm": 0.4995732048308892,
+      "learning_rate": 4.557938778361052e-06,
+      "loss": 0.4685,
+      "step": 10105
+    },
+    {
+      "epoch": 3.959273894788837,
+      "grad_norm": 0.4833683966329439,
+      "learning_rate": 4.557849817833386e-06,
+      "loss": 0.4548,
+      "step": 10106
+    },
+    {
+      "epoch": 3.959669054087429,
+      "grad_norm": 0.5172522955731031,
+      "learning_rate": 4.5577608492237365e-06,
+      "loss": 0.4625,
+      "step": 10107
+    },
+    {
+      "epoch": 3.9600642133860213,
+      "grad_norm": 0.48501688273496174,
+      "learning_rate": 4.557671872532452e-06,
+      "loss": 0.4537,
+      "step": 10108
+    },
+    {
+      "epoch": 3.9604593726846136,
+      "grad_norm": 0.5012142376692591,
+      "learning_rate": 4.557582887759881e-06,
+      "loss": 0.464,
+      "step": 10109
+    },
+    {
+      "epoch": 3.960854531983206,
+      "grad_norm": 0.49905859888997944,
+      "learning_rate": 4.557493894906375e-06,
+      "loss": 0.4688,
+      "step": 10110
+    },
+    {
+      "epoch": 3.961249691281798,
+      "grad_norm": 0.48977789559433366,
+      "learning_rate": 4.5574048939722825e-06,
+      "loss": 0.4545,
+      "step": 10111
+    },
+    {
+      "epoch": 3.9616448505803903,
+      "grad_norm": 0.47820284946504593,
+      "learning_rate": 4.557315884957952e-06,
+      "loss": 0.4509,
+      "step": 10112
+    },
+    {
+      "epoch": 3.9620400098789825,
+      "grad_norm": 0.5033813687269575,
+      "learning_rate": 4.557226867863734e-06,
+      "loss": 0.4541,
+      "step": 10113
+    },
+    {
+      "epoch": 3.9624351691775748,
+      "grad_norm": 0.48559157024559013,
+      "learning_rate": 4.5571378426899784e-06,
+      "loss": 0.4678,
+      "step": 10114
+    },
+    {
+      "epoch": 3.962830328476167,
+      "grad_norm": 0.48951951068737737,
+      "learning_rate": 4.557048809437034e-06,
+      "loss": 0.4826,
+      "step": 10115
+    },
+    {
+      "epoch": 3.9632254877747592,
+      "grad_norm": 0.5107355330569413,
+      "learning_rate": 4.556959768105253e-06,
+      "loss": 0.4643,
+      "step": 10116
+    },
+    {
+      "epoch": 3.9636206470733515,
+      "grad_norm": 0.4827625154057386,
+      "learning_rate": 4.556870718694981e-06,
+      "loss": 0.4592,
+      "step": 10117
+    },
+    {
+      "epoch": 3.9640158063719437,
+      "grad_norm": 0.5016649652795582,
+      "learning_rate": 4.55678166120657e-06,
+      "loss": 0.4668,
+      "step": 10118
+    },
+    {
+      "epoch": 3.964410965670536,
+      "grad_norm": 0.4945290431200768,
+      "learning_rate": 4.55669259564037e-06,
+      "loss": 0.4726,
+      "step": 10119
+    },
+    {
+      "epoch": 3.964806124969128,
+      "grad_norm": 0.48945605735322806,
+      "learning_rate": 4.55660352199673e-06,
+      "loss": 0.4581,
+      "step": 10120
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 50600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 2530,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.540274358583296e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}